From meh9 at cornell.edu  Tue Apr 11 15:25:50 2000
From: meh9 at cornell.edu (Matthew Hirsch)
Date: Tue, 11 Apr 2000 15:25:50 -0400
Subject: Functions and objects
Message-ID: <meh9-C7461E.15255011042000@news.cit.cornell.edu>

Hi All,

I'm confused.  Take a look at the following:

>>> def f(x):
...     return [x]
... 
>>> a=f(5)
>>> a
[5]
>>> for x in range(5):     # Case A
...     temp=a
...     print a
...     temp.append(1)
... 
[5]
[5, 1]
[5, 1, 1]
[5, 1, 1, 1]
[5, 1, 1, 1, 1]

>>> a=f(5)
>>> for x in range(5):     # Case B
...     temp=a[:]
...     print a
...     temp.append(1)
... 
[5]
[5]
[5]
[5]
[5]
>>> 

In case A, why doesn't temp reset itself to the value of a, [5], that 
was predetermined before it entered the loop?  Why do you need to copy 
the list to get the behavior I'm looking for in Case B?  Doesn't a 
always equal [5]?

Thanks for your help,
Matt


From rcunning at acm.org  Mon Apr 10 08:16:08 2000
From: rcunning at acm.org (Robert W. Cunningham)
Date: Mon, 10 Apr 2000 12:16:08 GMT
Subject: Python paradigms
References: <Pine.LNX.4.21.0004100757580.19783-100000@fep132.fep.ru> <8cs70f$ojf$1@pegasus.csx.cam.ac.uk>
Message-ID: <38F1C600.3FD9878B@acm.org>

Nick Maclaren wrote:

> In article <Pine.LNX.4.21.0004100757580.19783-100000 at fep132.fep.ru>, Oleg Broytmann <phd at phd.russ.ru> writes:
> |> On Mon, 10 Apr 2000, Robert W. Cunningham wrote:
> |>
> |> > Oleg Broytmann wrote:
> |> > > > python's "and" and "or" operators doesn't evaluate the
> |> > > > second argument unless they really have to...
> |> > >
> |> > >    Yes, I missed that, sorry.
> |> >
> |> > So, it works just fine, right?
> |>
> |>    No. Code is intended not only to beeing run on processor, but also (and
> |> often more important) to beeing read by human. Code is speech, it is the
> |> way to interchange ideas.
> |>    In that sense ugly code is not working even if you add three books of
> |> comments. Code should be readable. Period.
>
> Yes, precisely.  In the real world, programs often have long lifetimes,
> are worked on by many programmers (in parallel and sequentially), are
> sometimes transliterated between languages, and it is common to need
> to modify code which was written 20 years ago by some long-retired
> programmer.
>
> At best, confusing code wastes effort.  At worst, it causes mistakes
> to be introduced.  I have seen the latter effect many times.
>
> That is why I was talking about the benefits of those paradigms for
> code clarification, and neither efficiency nor compactness.  Python
> has some very useful facilities, but is a bit tedious for some kinds
> of scripting.

Again, comments in the code can cover all these bases.  A good "paradigm" is a good thing to have, but it is no
Holy Grail.  It is but one step in the creation of a good program.  If a sound and reliable implementation is
possible, then the paradigm is probably a good fit.  Paradigms that encourage "slow" or "fat" code merely to
improve the self-expressiveness of the code is a dubious goal in practical terms, but certainly has value in
abstract terms (i.e., to extract reusable patterns - patterns are *very* expressive).

Have any of you ever had to port a product written in a "dead" computer language?  I have, more than once.  The
comments were the ONLY thing to make the task possible, since even texts adequately documenting the old
languages (especially the libraries) were extremely difficult to find.  Many such languages seemed to have come
into existence with the PDP-11 and the Intel 8080, and never migrated far from those platforms.  At times, I was
reduced to staring at assembler dumps to figure out what a line of supposedly "high-level code" was actually
trying to do.  And the code, I was told, had an excellent reputation for being "well written".  It had merely
gotten old, and the original authors had literally died.

Human language, written well, is far more expressive than computer languages.  That's why computer hardware and
software is documented with books, and not in their "own" languages.  Remember, code maintainers are also
"users" of a product, and deserve to have the support of equal documentation.  If the code could do it alone,
I'd say go for it.  But it cannot, and never will.  To believe otherwise is sheer foolishness that displays a
blatant ignorance of how the "real world" works.

The greatest programmers of our age manage to include the equivalent of a "Rosetta Stone" in their work.  It
could be the Weave and Tangle of Knuth's "Literate Programming", or merely inline comments and a good set of
design documents.  But never bare code, all alone.  Never.  A single representation, no matter how elegant,
cannot be sufficient.

Unless, of course, you write code that will never need maintenance: never have a bug, never need a new feature,
never have any other developer than yourself, and your code is used by nobody.  In that case, sure, leave out
the comments!

Computer programming is far more than a form of communication between a programmer and a computer.  It
implicitly involves a virtual army of other people, users and maintainers chief among them.  Remember, the goal
of good communication is NOT "making it possible for another party to understand you".  No, good communication
is "making it DIFFICULT for another party to MISUNDERSTAND you".  Which means, to be understood, you usually
need to say a given thing more than once, in more than one way.  Code is one way.  Comments is another.  They
both (and more) are needed if the source code is to communicate effectively throughout time.

At best, sound paradigms can simplify code and reduce the level of commenting needed, but can in no way
eliminate the need for thorough commenting and other documentation.

And the occasional efficient and effective (but(t) ugly) one-liner that saves significant processor time or lots
of code deserves to be commented to death.

And if you can't write good comments, how can you POSSIBLY believe you can write good code?


-BobC




From gmcm at hypernet.com  Fri Apr 21 17:05:38 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Fri, 21 Apr 2000 17:05:38 -0400
Subject: The Simple Economics of Open Source
In-Reply-To: <f4JdfuAkLKA5EwcQ@jessikat.demon.co.uk>
Message-ID: <1255771356-36910442@hypernet.com>

Robin Becker wrote:

> In article <1255786824-35980039 at hypernet.com>, Gordon McMillan
> <gmcm at hypernet.com> writes
> >Why does everyone miss the main point?
> >
> >I (and you) are literally 100s of times more productive because 
> >of open source. Imagine having to either invent it all yourself, 
> >or duct tape together expensive half-assed closed "solutions". 
> >Shudder.
> 
> Developers who share code may actually do so for other than friendly
> reasons. If M$ released win2000 source into the world it might actually
> slow down Linux development.

Doubt it. It would probably yield an emulation layer (c runtime 
lib) that worked.
 
> I prefer not to worry about the exact electronics/chemistry behind my
> computer. So in that sense my 'open sourceness' has limits. I guess this
> argument has to do with the relative value 'developers' put on the
> tools. On win32 VC++ is valuable; while gcc/egcs is really useful on
> unices.

What problems would you be solving without Python (TCL ...)? 
Remember the late 80s / early 90s when big companies were 
pushing 7 figure middleware that is now obsolete (thanks to 
internet technologies)? Remember when IBM truly believed the 
world would communicate through SNA? Remember when 
damn near any real app cost at least $10K?
  
> >
> >Developers who share code are *enormously* more productive 
> >than those who don't. So the range of problems that can be 
> >attacked increases geometrically, so computers penetrate 
> >new areas, so geeks stand a better chance of making a living.
> 
> This is a competitive position to take. Geeks presumably compete with
> non-geeks for resources.

So? The article was about why developers like open source. 
Those wierd people with real lives didn't enter into it.
 
> In addition the inter geek co-operation may actually be a form of
> competition for other things like 'mindset' 'noosphere' etc. Painters
> exchange ideas, concepts and criticism freely, but compete for
> customers.

Sure. And my competitive position is greatly enhanced vs 
those developers who rely on commercial products. I never 
used the word "altruism". I was just pointing out that the 
article and the initial responses there to completely missed 
what I see as the main benefit. In most cases, cooperation 
remains a good thing, whatever the motivation.
 
> >
> >Win-win situations are rarely profitably analyzed by 
> >competitive models.

> Wrong! Non-zero sum games are widely studied in many areas of
> mathematics, economics and control theory. 

Did you see any in that article?

> Just because we can both win
> doesn't mean we're not in competition. 

And just because we're in competition doesn't mean we 
should shoot each other <wink>.

> Win-Win games tend not to be so
> stable as minimal-loss minimal-loss (Nash) games and for that reason are
> of less interest. It requires a constant flow of information (co-
> operation) to achieve best-best so that is rarely achieved or sustained.
> 
> Win-Win for the group of geeks is a sub-game in the wider community. As
> a group the geeks strive to turn their knowledge into something useful
> to get income. It may be that inter geek co-operation aids the group as
> a whole; certainly that cannot be simply asserted as there exists a
> large number of geeks who work only for $/? and who co-operate with the
> prince of darkness in Seattle. I don't believe any comparative study of
> commercial vs opensourcerer geeks has yet been carried out to see which
> group does better.

It doesn't take a comparative study for me to know that, 
despite having an NT box and MSVC etc., I do 90% of what I 
do with open source stuff.

Whether mankind is inherently selfish and evil is another 
question.

i-think-they're-just-sleeping-ly y'rs

- Gordon



From artswan at ktc.com  Wed Apr  5 12:33:03 2000
From: artswan at ktc.com (Art Swan)
Date: Wed, 5 Apr 2000 11:33:03 -0500
Subject: Clear screen?
Message-ID: <38eb6a36@207.71.36.3>

In Perl I can clear the screen using system("cls"). What is the equivalent
in
Python?





From aahz at netcom.com  Sun Apr  2 12:53:19 2000
From: aahz at netcom.com (Aahz Maruch)
Date: 2 Apr 2000 16:53:19 GMT
Subject: Paid to work in Python? (was: Re: Choice Of Language)
References: <Uqvw4.2261$yL2.28868@typhoon.southeast.rr.com> <8ajmib$elo$1@nnrp1.deja.com> <8c3ats$b5f$1@nntp.Stanford.EDU> <slrn8eact2.sq7.nobody@bambi.visi.com>
Message-ID: <8c7ttv$ge7$1@slb2.atl.mindspring.net>

In article <slrn8eact2.sq7.nobody at bambi.visi.com>,
Grant Edwards <nobody at nowhere.nohow> wrote:
>
>Pointless tanget: Is it just me, or is Google really the last usable search
>engine?

I still use Yahoo! as a backup for Google, but yeah.  I never did think
AltaVista was usable, nor NorthernLight, Lycos, or Hotbot.  Infoseek was
pretty good for a long time, but it's gone downhill and I'm not sure
why; I suspect it simply isn't scaling.
--
                      --- Aahz (Copyright 2000 by aahz at netcom.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

"It's 106 miles to Chicago.  We have a full tank of gas, a half-pack of
cigarettes, it's dark, and we're wearing sunglasses."  "Hit it."


From milesthompson at sprint.ca  Thu Apr 20 16:51:46 2000
From: milesthompson at sprint.ca (Miles Thompson)
Date: Thu, 20 Apr 2000 17:51:46 -0300
Subject: How2 use curses for data input
Message-ID: <38FF6DE2.5751423A@sprint.ca>

I 've read the curses part of the manual, and the HOWTO, but I'm not
much wiser. I need to capture some user input (name of a new directory
for archiving some web files), and I understand how to define a window,
etc. What I don't know to do is how to edit the lines, and loop from
last to first.

In FoxPro/dBase one could issue
    @ 5,10 say "Source file:"
    @ 5, 23 get csource
    @ 6,10 say "Destination:"
    @ 6,23 get cDestination
    READ CYCLE
and the user could loop around the fields until a (not displayed here)
OK button was clicked on.

Can the same thing be done with curses and Python, or do all the
movement keys, etc. have to be trapped and the correct movement of the
cursor applied?

I know this seems a bit antedeluvian in an age of GUI's, but the box is
a web server, I don't want to run X, and it's a very simple task. I
would like the user to be able to look at what's been typed in before
commiting to doing it.

Regards - Miles Thompson
milesthompson at sprint.ca



From gmcm at hypernet.com  Fri Apr  7 14:05:44 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Fri, 7 Apr 2000 14:05:44 -0400
Subject: Compile Gadfly?
In-Reply-To: <E%kH4.4341$HG1.126509@nnrp1.uunet.ca>
Message-ID: <1256991746-21080406@hypernet.com>

Warren Postma wrote:

> Has anyone tried compiling Gadfly and 'embedding' it into an executable, for
> easier distribution? Does it work!?

Yup. The only trick is putting the marshalled grammar file 
where the app will find it.

- Gordon



From aa8vb at yahoo.com  Wed Apr 12 11:29:35 2000
From: aa8vb at yahoo.com (Randall Hopper)
Date: Wed, 12 Apr 2000 11:29:35 -0400
Subject: comparing strings and ints
In-Reply-To: <jr%I4.643$rc9.190628352@newsb.telia.net>
References: <2YmI4.366$rc9.188014080@newsb.telia.net>
 <20000411134358.A2352304@vislab.epa.gov> <1256641282-42160637@hypernet.com>
 <20000412075812.A2449004@vislab.epa.gov>
 <jr%I4.643$rc9.190628352@newsb.telia.net>
Message-ID: <20000412112935.A2446939@vislab.epa.gov>

Fredrik Lundh:
 |Randall Hopper:
 |>Gordon McMillan:
 |> |Randall Hopper wrote:
 |> |> Fredrik Lundh:
 |> |>  |
 |> |>  |    Objects of different types always compare unequal, and are
 |> |>  |    ordered consistently but arbitrarily.
 |> |> 
 |> |> Is there a case where this could be useful (with its undefined
 |> |> behavior)?  I believe that is the root of the question.
 |> |
 |> |I'd happily bet your wife and firstborn child that there's code that
 |> |relies on it <wink>, if only because after sorting a list, all the ints
 |> |will end up in a clump, and all the strings in another, etc. 
 |>
 |>Oh, I also have __no doubt__ that someone is using this unofficial,
 |>undocumented behavior.  But they don't have a leg to stand on if the
 |>ordering breaks in a future Python version ;-)
 |
 |interestingly enough, you included my documentation quote in
 |a post claiming that this is an undocumented feature...

I was of course referring to the behavior Gordon McMillan mentioned in the
message I replied to.  Namely, that sorting a list of objects of
hetrogenous primitive types "groups" the instances of a particular type
together (ints together, strings together, etc.).

I'm taking Gordon's word that that's what it does.  The documentation you
cited simply said "arbitrary" ordering.

Unless it's documented as official behavior, I would think that expecting
it is probably not a safe bet.

-- 
Randall Hopper
aa8vb at yahoo.com



From P.J.W.S.VRIJLANDT at INT.azg.nl  Wed Apr 19 04:39:35 2000
From: P.J.W.S.VRIJLANDT at INT.azg.nl (P.J.W.S. Vrijlandt)
Date: Wed, 19 Apr 2000 09:39:35 +0100
Subject: Telnet client using curses
In-Reply-To: <8dj02n$61k$1@kryogenix.yi.org>
Message-ID: <1E0DE27726A@nds-3.azg.nl>

On 19 Apr 00 at 1:53, Aquarius wrote:

> Having used Python for a few brief scripts (the sort of thing I used
> to use Perl for :), I thought I'd try something a bit more
> ambitious, to teach myself the language. What I decided on, because
> I need one, is a telnet client. The key thing is that I'd like to
> use curses to give me one input line into which text it typed, while
> data returned from the connection goes into the window above.
> Something like TinyFugue.
> 
> Looking at telnetlib, it uses select to poll stdin. However, I'll be
> getting user input using curses input functions, not input() or
> raw_input(), so the user input doesn't go into stdin. Is the best
> way to handle this: 1. get user input with the curses function and
> then somehow slip this input into stdin, so it looks like it came
> from raw_input() or whatever? Obviously, raw_input() gets its data
> into stdin, but it's a built-in function, so I don't know how it
> does it :-) 2. write a file object which I feed data *into* from the
> curses input function, and which then feeds that data back out when
> requested to do so, and poll *that* instead of stdin? 3. A better
> solution?
> 
> If your answer is "3", some kind of suggestion as to the Right Way
> to do it would be appreciated :-)
> 
> Aquarius
> 
> -- 
> Well *done*, Aquarius -- a phrase we love typing because it makes us
> feel like the boss-character from a 1980s Glen A Larson action
> series...
>   NTK,
>   http://www.ntk.net/index.cgi?back=archive99/now0618.txt&line=292#l
> 
> -- 
> http://www.python.org/mailman/listinfo/python-list

I would opt for "3". The Right Thing (tm) requires 
rewriting/modifying of telnetlib in order to seperate the terminal 
functions from the protocol handling, like it was intended in de rfc.

I have done such a thing and I will mail it to you privately when I'm 
home  (it's not ready for publication). 

You then get a TTY object with read and write methods and you can use 
it as you like. On windows, you can't use select with stdin and you 
have to use seperate threads for input and output to get an elegant 
solution.

The elegance of this solution is that you don't have to modify 
telnetlib again if you want to write logfiles or implement a terminal 
that uses another gui instead of curses.

Patrick Vrijlandt





From rhicks at nospam.rma.edu  Sat Apr  1 14:28:16 2000
From: rhicks at nospam.rma.edu (Robert Hicks)
Date: Sat, 01 Apr 2000 19:28:16 GMT
Subject: Python 1.6 alpha 1 released
References: <200003312130.QAA04361@eric.cnri.reston.va.us> <38E5205F.DE811F61@roguewave.com> <Wf9F4.4423$74.70231@newsc.telia.net> <38E58D2D.384F1546@Lugoj.Com> <S9kF4.4472$74.71414@newsc.telia.net> <38E646B6.12011A4E@Lugoj.Com>
Message-ID: <k5sF4.2$O9.2705@news>

> James Logajan <JamesL at Lugoj.Com> wrote:
> The benefit/pain ratio for by this change is close to zero.

I am sure that this will be thorougly thought out before the official
release
of v1.6.

> Well then I hope it gets changed due to those polite requests. Hopefully
> that old extra code can be patched back in.

same answer as previous

 > now how come I find your "guido the god" argument to be
> directly offensive?
>
> Because it was meant to be. Glad to see it worked! The motivation for it
was
> because even though Bjorn directed his post to Guido, you and one or two
> others responded. Almost like high priests guarding the temple. Any reason
> Guido doesn't post to this newsgroup like he used to?

Why would you possibly want to offend someone, whoever they are, with a
blatantly
attacking statement?

Why does it matter WHO gives the answers as long as the answer is well
thought out and
correct? I for one am hoping that Guido is working hard on the 1.6 release.
Mr. Lundh has
always actively participated in this newsgroup and I appreciate his time for
doing so.

I wonder if someone in the perl group ever said "Larry Wall the god" to make
anyone upset? If so, then Guido stands in good company.

Bob




From hamish_lawson at yahoo.co.uk  Mon Apr 10 17:08:14 2000
From: hamish_lawson at yahoo.co.uk (Hamish Lawson)
Date: Mon, 10 Apr 2000 14:08:14 -0700
Subject: indexed dictionaries?
References: <Pine.BSF.4.10.10004092225040.1361-100000@chi.pair.com> <B516CDD9.17E6%dgoodger@bigfoot.com> <8cta0u$2bg$2@eeyore.INS.CWRU.Edu>
Message-ID: <0062c1d4.5891a1da@usw-ex0104-033.remarq.com>

Opening up the Vaults of Parnassus we find the seqdict package:

http://www.vex.net/parnassus/apyllo.py/514463245.769244789.833538
28

Hamish Lawson


* Sent from RemarQ http://www.remarq.com The Internet's Discussion Network *
The fastest and easiest way to search and participate in Usenet - Free!




From mjhand at concentric.net  Fri Apr 14 16:55:30 2000
From: mjhand at concentric.net (Manus Hand)
Date: Fri, 14 Apr 2000 14:55:30 -0600
Subject: Questions on 1.6a2's string methods
References: <38F6C904.76B538E5@concentric.net> <38F7598E.97A07412@aw.sgi.com>
Message-ID: <38F785C2.7087957C@concentric.net>

Thanks, Gary!

"".title() was exactly what I was hoping for!

Manus




From johngrayson at home.com  Sun Apr 23 16:07:18 2000
From: johngrayson at home.com (johngrayson at home.com)
Date: Sun, 23 Apr 2000 20:07:18 GMT
Subject: Radio-button in Tkinter
References: <39028096.24B972B4@swt.edu>
Message-ID: <8dvl5e$ff7$1@nnrp1.deja.com>

In article <39028096.24B972B4 at swt.edu>,
  ns56645 <ns56645 at swt.edu> wrote:
>
> --------------834CABF3AAE289AC3BC56CC4
> Content-Type: text/plain; charset=us-ascii
> Content-Transfer-Encoding: 7bit
>
> In the following code the radiobuttons that have been set are not seen
> as default selected in the output.
> For e.g         v1.set(3) # initialize should set the radiobutton with
> text 'BOTH' as default.
>
> What could be the wrong with the code?
> Please help?
>
> Shah
>

Nothing is wrong. I tried your code in a little harness and it
worked correctly. 'BOTH' was selected for each radiobutton group.
Changing the variable to 1 or 2 for v1 and 11 or 12 for v2 selected
the corresponding button.

What is it that you're expecting the widgets to do?

    John


Sent via Deja.com http://www.deja.com/
Before you buy.


From effbot at telia.com  Sat Apr 15 07:27:39 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Sat, 15 Apr 2000 11:27:39 GMT
Subject: General question of "SystemError:NULL result..."
References: <8d9645$fsa$1@nnrp1.deja.com>
Message-ID: <LmYJ4.1127$rc9.190032896@newsb.telia.net>

gmol at my-deja.com wrote:
> Hi all, disclaimer, I am new to python so I may be doing something
> severly stupid.

unless you've written the extension module, it's not
really your fault...  read on.

> I have installed VTK and python on my sys,and some of the python
> examples work, and some don't...the ones that don't say always say
> something like:
>
>   File "KSpline.py", line 106, in ?
>     profileTubes.SetInput(profileData)
> SystemError: NULL result without error in call_object
>
> Now, no one has been able to figure this out on the VTK list...but I do
> know that people can run the same examples on the same python under the
> same platform (linux) and the same Mesa...

technically, this means that an extension library returns NULL
to Python, without setting the exception status.  it's a bug in
the extension library.

(it's probably caused by a configuration problem of some kind,
but the library should tell you so, rather than returning bogus
values to the python interpreter...)

</F>




From dln at jump.net  Wed Apr 19 13:39:46 2000
From: dln at jump.net (Dick Norton)
Date: Wed, 19 Apr 2000 12:39:46 -0500
Subject: need non-blocking keyboard read on dos - like kbhit?
Message-ID: <38FDEF62.92171124@jump.net>

Hello,

I've gotten pydos.zip from 

http://www.python.org/ftp/python/wpy/dos.html

and it runs but I need either threads (these seem absent)
or a non-blocking
keyboard char in routine to return an otherwise unsed 
thinkpad back to some level of dignity as a shop timer.

Is there a routine like kbhit without msvcrt (dos, not windows)?

Thanks,
Dick Norton


From bozon at natur.cuni.cz  Tue Apr 11 11:40:59 2000
From: bozon at natur.cuni.cz (Michal Bozon)
Date: Tue, 11 Apr 2000 17:40:59 +0200
Subject: starship.python.net
In-Reply-To: <Pine.OSF.4.10.10004111721150.9584-100000@prfdec.natur.cuni.cz>
References: <Pine.OSF.4.10.10004111721150.9584-100000@prfdec.natur.cuni.cz>
Message-ID: <Pine.OSF.4.10.10004111740110.11294-100000@prfdec.natur.cuni.cz>

> What the hell is with the starship.python.net server? Today and yesterday
> I cannont connect (from Czech Republic).
> Someone knows?
> 
> Michal Bozon
> 

Just after I posted this message I tried connection again, and it is OK.

MB



From somaking at yoyo.cc.monash.edu.au  Sun Apr 23 23:28:45 2000
From: somaking at yoyo.cc.monash.edu.au (Linus Chang)
Date: Mon, 24 Apr 2000 13:28:45 +1000 (EST)
Subject: dbm string lengths
Message-ID: <Pine.OSF.4.21.0004241254240.410935-100000@yoyo.cc.monash.edu.au>

Hi all,

I'm trying to use anydbm to store potentially long strings, and am running
into problems which seem to be caused by the length of the strings. (The
Python Ref Manual sec 3.11 on Shelve mentions that the length of pickled
objects should be small - how small?)

eg. the following code fragment:

import anydbm
import cPickle

db = anydbm.open("testingdb", "c")

print "Keys in db are: ",
for k in db.keys():
    print k,
list = []

for x in range(10000):
    list.append("Item number " + `x`)

db["item"] = cPickle.dumps(list)
db.sync()
db["item"] = "Small string"
db.sync()

-- Run the first time (no testingdb file exists), output as expected:
D:\Program Files\Python>python dbtest.py
Keys in db are:
-- Run a second time, the name of the key gets confused:
D:\Program Files\Python>python dbtest.py
Keys in db are:  itemm

(Changing the list to say a length of 10 yields results as expected.)

Alternatively, removing the line db["item"] = "Small string" causes the
following output on the second run:

D:\Program Files\Python>python dbtest.py
Keys in db are:
Traceback (innermost last):
  File "dbtest.py", line 8, in ?
    for k in db.keys():
bsddb.error: (22, 'Invalid argument')

-- Can anyone please help? I'm trying to store potentially long strings in
a dictionary, and need to store them to disk. I expect that entries will
be updated at up to 20 per second, so it's not really feasible to pickle
the whole dictionary to disk every time an individual entry is updated. In
addition, saving to disk asynchronously (eg. every X seconds) really isn't
my preference...

Thanks in advance,

Linus Chang




From exarkun at flashmail.com  Thu Apr 27 12:41:37 2000
From: exarkun at flashmail.com (Jp Calderone)
Date: Thu, 27 Apr 2000 16:41:37 +0000
Subject: Converting = to ==
References: <n01O4.12725$HG1.340768@nnrp1.uunet.ca>
Message-ID: <39086DC1.E2699CDE@flashmail.com>

Warren Postma wrote:
> 
> I have an "end user query tool" that uses Python expressions, but I want
> comparisons to be able to be written either as A=B or A==B then I want to
> compare them using Python's Eval function.
> 
> Problem is if you search and replace all = with == then users have to use
> the single-equals convention, and some of them will be python programmers,
> and will want to use the "non idiot mode" of the query tool.
> 
> So i did this, but is there a better way?
> 
>     >>> string.replace( string.replace("a=b a==b", "=", "=="), "====","==")
>     a==b a==b
> 
> Warren

 You can use regular expressions.  I'm not going to try to give you the
exact one off hand, because I'll get it wrong, but basically replace
"[anything-but-equals]=[anything-but-equals]" with "[first part]==[second
part]"
-- 
-----BEGIN GEEK CODE BLOCK-----
Version: 3.12
GCS dpu>-- s+: a--- C++++ UL++++ P+ L+++ E---- W++ N++ o+ K- w--- 
O-- M- V- PS+++ PE- Y+ PGP t+ 5+ X++ R+ tv+ b++ DI++ D--- 
G e* h! !r y-- 
------END GEEK CODE BLOCK------
--
 4:40pm up 29 days, 21:47, 0 users, load average: 0.11, 0.07, 0.01


From bawolk at pacbell.net  Thu Apr  6 01:43:02 2000
From: bawolk at pacbell.net (Bruce Wolk)
Date: Wed, 05 Apr 2000 22:43:02 -0700
Subject: Widget for plotting graphs?
Message-ID: <38EC23E6.9975FD7D@pacbell.net>

I am exploring Tkinter and Pmw and wondering whether anyone has created
a widget for graphing points and functions.  Also, is there a way to
place one pixel (i.e., a point) on a Canvas widget?  Thanks.

Bruce


From rgb122 at my-deja.com  Thu Apr  6 15:07:14 2000
From: rgb122 at my-deja.com (rgb122 at my-deja.com)
Date: Thu, 06 Apr 2000 19:07:14 GMT
Subject: Problem calling COM Object written in Python
Message-ID: <8cin8n$dhk$1@nnrp1.deja.com>

I have a COM object I wrote in Python that needs to be called from an
ASP script (VBScript). I can create the object just fine, but then when
I try to call a method on it:

Python COM Server Internal Error error '80004005'

Unexpected Python Error: exceptions.TypeError: Objects for SAFEARRAYS
must be sequences (of sequences), or a buffer object.

The COM object is a pretty simple wrapper for a Python class. It's got
one factory method that takes a string as a parameter

import Publisher

class PublisherCOM:

	_public_methods_ = ['loadSite']
	_reg_progid_ = "ECS.Publisher"
	_reg_clsid_ = "{5E633A5D-D0BA-40A4-B30D-54F7B29CCD23}"

	def loadSite (self, filename):

		print "I'm here\n"
		print str(filename)

		site = Publisher.loadSite(str(filename))
		return site

The PythonWin Trace Collector gives this friendly output:

Object with win32trace dispatcher created (object=None)
in _GetDispID_ with LoadSite 8
in _InvokeEx_ with 1000 1033L 3
(L'D:\\rbender\\Desktop\\pyTemplate\\brands.xml',) None
<PyIServiceProvider at 0x210ebfc with obj at 0x202f130>

Is there some sort of problem passing the string as a parameter? I'm
running this under Win2000, so perhaps there's a lovely new "feature"
that's cropped up.

Any and all help is much appreciated. Thanks!

-Rob Bender


Sent via Deja.com http://www.deja.com/
Before you buy.


From hamish_lawson at yahoo.co.uk  Sat Apr  1 16:41:49 2000
From: hamish_lawson at yahoo.co.uk (Hamish Lawson)
Date: Sat, 01 Apr 2000 13:41:49 -0800
Subject: Why should I switch to Python?
References: <seas5ia89177@corp.supernews.com><00033120055300.03687@quadra.teleo.net> <secfidq8921@corp.supernews.com>
Message-ID: <170a3a86.966ca715@usw-ex0102-014.remarq.com>

Hi Aaron

Here are the reasons I switched to Python.

I had used Perl for several years and I think I had gotten to
quite an advanced stage with it. However what convinced me to
switch to Python was the recurring difficulties I saw my less-
experienced colleagues having with things like references and
other dark corners of Perl. Sure, it would have been much harder
to use C or C++ for the kinds of things we were using Perl for,
but I felt that the continual taking of references and
dereferencing was still too much like C's pointer hassle. In
Python, since all variables hold references, references can be
transparently handled.

Compare this in Perl:

    @people = ('Joe', 'Pat', 'Sue');
    @stuff = (5, \@people, 7);
    dosomething(\@stuff);
    print $people[2]

with this in Python (square brackets are used to make lists):

    people = ['Joe', 'Pat', 'Sue']
    stuff = [5, people, 7]
    dosomething(stuff)
    print people[2]

The first line of the Perl version causes @people to contain a
list, but in the Python version it causes the people variable to
hold a *reference* to the list. A great deal of simplification
follows on from that.

Or take another example, adapted from 'Programming Perl' (I know
my use of printf here isn't entirely idiomatic in Perl, but I'm
trying to make that aspect look as much like Python in order to
highlight the actual comparison I'm making):

    foreach $family (keys %TV) {
        printf "Family: %s", $family;
        foreach $who (@{$TV{$family}{'members'}}) {
            printf "%s (%s)\n", $who->{'name'}, $who->{'role'};
        }
    }

I found it much easier to explain to my less-experienced
colleagues the Python version below:

    for family in TV.keys():
        print "Family: %s" % family
        for who in TV[family]['members']:
            print "%s (%s)" % (who['name'], who['role'])

Neither "who" or "TV[family]['members']" needed to be explicitly
dereferenced.

I also found Python's handling of OO more elegant and explicable:

    class MyClass(AnotherClass):

        def __init__(self, name):
            self.name = name

        def greet(self):
            print "Hi, I'm", self.name

    m = MyClass('Bob')
    m.greet()

compared to this:

    package MyClass;
    @ISA = ('AnotherClass');

    sub new {
       ($class, $name) = @_;
       $self = {};
       bless $self, $class;
       $self->{'name'} = $name;
       return $self;
    }

    sub greet {
        ($self) = @_;
        print "Hi, I'm ", $self->{'name'}, "\n";
    }

    $a = MyClass->new('Bob');
    $a->greet();

Once I switched to Python I found other gems in the everything-
is-an-object philosophy: even modules (packages in Perl) and
classes are objects, so you can do things like this:

    try:
        import optional_fast_native_stuffmodule
        stuffmodule = optional_fast_native_stuffmodule
    except ImportError:
        import standard_slow_portable_stuffmodule
        stuffmodule = standard_slow_portable_stuffmodule

    stuffmodule.dostuff()

So while there may not be many things that you can do in Python
can't in Perl (but treating classes and modules as objects and
reflection would be among them), I found that the real advantage
of Python is that most things will be simpler and more
understandable. This makes it so much easier when working with
others or when revisiting my code months later.

Hamish Lawson


* Sent from RemarQ http://www.remarq.com The Internet's Discussion Network *
The fastest and easiest way to search and participate in Usenet - Free!



From effbot at telia.com  Wed Apr 26 16:42:11 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Wed, 26 Apr 2000 20:42:11 GMT
Subject: newbie regexp question
References: <39075648.CA622A7D@imeme.net>
Message-ID: <DwIN4.4305$Za1.64176@newsc.telia.net>

mindlace <mindlace at imeme.net> wrote:
> I've got a re.compile('\W') that I can use to search for non-alphanumeric
> characters in a variable.  However, if the variable has ( in it, for
example,
> the search breaks.

breaks?

> is there a quoted_var = string.quote(somevar) or similar? Do I want to
> use string.maketrans(somevar, quoted_var) ?

have no idea what you're talking about here, but maybe
re.escape does what you wants?

>>> import re
>>> print re.escape.__doc__
Escape all non-alphanumeric characters in pattern.
>>> print re.escape("(")
\(

</F>




From m.faassen at vet.uu.nl  Fri Apr 21 15:15:59 2000
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 21 Apr 2000 19:15:59 GMT
Subject: Using python on the web
References: <MPG.1365adcfa8470eef9896aa@news> <8di2bk$l1s$1@newshost.accu.uu.nl> <MPG.136631472f892a119896ac@news> <_W2L4.986$GK5.20898@news2-win.server.ntlworld.com> <MPG.13665ef93eedbda29896ad@news> <9N4L4.499$wYl.191816704@newsb.telia.net> <8do2te$kpr$1@nnrp1.deja.com>
Message-ID: <8dq9df$pjg$1@newshost.accu.uu.nl>

spenrose at my-deja.com wrote:
> We build medium-busy, CGI-intensive sites in Python. We were briefly
> quite excited about Zope, but once we realized it locked us into Yet
> Another Programming Language, we lost interest. DHTML is particularly
> problematic, AFAI can tell, because it doesn't appear to allow the
> segregation of HTML development from coding the way Python string
> substitution does. We hire very skilled HTML people to do HTML and tell
> them to leave our sustitution tags alone, rather than taking the time
> and money to create novice DHTML people to do the same work.

If I understand you right, you seem to be incorrect about Zope. I've worked
together with a HTML designer and he had no trouble with DTML tags.
You can also tell the designer to leave the DTML tags alone, though
usually you just factor out pieces of HTML into fragments, so your
designer isn't even going to see much of DTML beyond the occasional
<dtml-var> tag.

> If we were
> going to learn another language for production purposes, Perl, Java, and
> C++ all have vastly larger infrastructures (code bases, labor pools,
> documentation, etc.) than DHTML ever will.

Note: DTML is named confusingly, and has nothing to do with DHTML. And
I'd just like to see a comparable infrastructure to Zope in Perl or C++.
*Maybe* Java.

DTML should be used as a glue language. Real code should be moved to
Python products. Currently in Zope this is more of a hassle than it needs
to be, but there are already PythonMethods which you can edit through the
web (which you can install seperately). Use minor amounts of DTML only!
(or do away with DTML completely and just use ZPublisher to publish
Python objects)

> There are enough hassles in
> using Python, which I love, because of its relatively small installed
> base. No way am I going to something which is an order of magnitude less
> supported.

That's your choice, of course. I don't think Zope is badly supported,
myself.

Regards,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From anders.eriksson at morateknikutveckling.se  Mon Apr  3 04:23:46 2000
From: anders.eriksson at morateknikutveckling.se (Anders M Eriksson)
Date: Mon, 03 Apr 2000 10:23:46 +0200
Subject: examples for HTMLgen
References: <14564.28003.931035.44380@gargle.gargle.HOWL>
Message-ID: <60kgesct0e4arv70qgbch9telqpc3tu3qc@4ax.com>

On Fri, 31 Mar 2000 18:18:27 +0900 (JST), Christian Stolte
<stolte at fuchinobe.skk.slb.com> wrote:

>
>I'd like to web-present a file structure (=pseudo database) in different sort
>orders with info and pics, nice and webby.
>
>a. project view - split frame  left:  all projs to select from
>                               right: detailed info with short descr,
>                                      pics etc
>b. data view    - split frame  left:  grouped data to select from
>                               right: detailed info with project information
>
>I know there is Zope, but that's too much. So what about picking the files
>into a directory like list, then sort that list in different ways and
>display it with HTMLGen???
>

I would use os.listdir() to create a list of files and then sort it
the way I want it. If you have several 'levels' of subdirectories you
want to look up os.path.walk()

Using the Table class in HTMLgen you just use the list of files and
create a table which you then can insert into your HTML page using
TemplateDocument

The only example that I have found (.although I haven't look very
hard.) is the HTMLtest.py. Here you will find alot of things!

// Anders



From tim_one at email.msn.com  Sat Apr  1 00:29:02 2000
From: tim_one at email.msn.com (Tim Peters)
Date: Sat, 1 Apr 2000 00:29:02 -0500
Subject: Function override?
In-Reply-To: <38E4D7A4.4C83CFF7@math.okstate.edu>
Message-ID: <000101bf9b9b$312f8fe0$2f2d153f@tim>

[David C. Ullrich]
> ...
> This is certainly one of the things that _I_ find keen,
> amyway - I hope the people who are talking about adding
> stricter typing to Python make it optional somehow.

Yup, not to worry.  All proposals to date make static typing declarations
optional, and by default not needed.  It may be that the choice will apply
to entire modules, though (that is, that if you want to use static
declarations in one part of a module, you'll need to use it everywhere in
that module; other modules make their own choices).

> There are obvious advantages to stricter typing but you lose
> something as well. (For example, I get a big kick out of
> the fact that exactly the  same euclidean-algorithm code
> works to find the GCD of two integers or the GCD of
> two polynomials. Etc.)

This isn't a problem with static typing per se, provided the system is rich
enough to let you define a suitable EuclideanDomain type.  The system is not
rich enough today, but people throughout this industry hold a charming faith
that tomorrow's solutions won't suffer today's problems <wink>.

just-a-fan-of-the-illusion-of-progress-ly y'rs  - tim





From lewst at yahoo.com  Wed Apr  5 00:58:55 2000
From: lewst at yahoo.com (lewst)
Date: Tue, 4 Apr 2000 21:58:55 -0700 (PDT)
Subject: Starving for an Advanced Python Book
Message-ID: <20000405045855.26682.qmail@web2101.mail.yahoo.com>

Jeff Bauer <jbauer at rubic.com> wrote:

> I would second Bjorn's recommendation.  It's a useful
> book for anyone who writes Python code and wants quick 
> access to library examples.

Thanks Jeff and Bjorn, now I want the book even more.
 
> That said, the electronic distribution mechanism
> will try your patience.  When you get it to work, 
> the first thing you'll want to do is follow Bjorn's
> advice and print it.

Yes, even though I won't get to install it on the 
workstation of my choice (Grr..), I think I'll try 
to find some way to get the book printed.

> purchasing-the-book-is-like-taking-a-test-but-it's-
>   all-downhill-from-there-ly y'rs,

This is off-topic, but what is with this funky way
of signing articles?  I've now seen many people do 
this on the Python list but am not quite getting it.


__________________________________________________
Do You Yahoo!?
Talk to your friends online with Yahoo! Messenger.
http://im.yahoo.com



From jcmllc at earthlink.net  Sat Apr 15 12:35:41 2000
From: jcmllc at earthlink.net (Pepper-Emporium)
Date: Sat, 15 Apr 2000 09:35:41 -0700
Subject: (no subject)
Message-ID: <001f01bfa6f8$b05dc9e0$69ee85ce@default>

PLEASE HELP:

I'm getting irate e-mails from people on my list indicating that when ever they send a message is bounces back to them instead of me. My list is called "hotlist" and I have about 27,000 members.

I want to REMOVE ALL the members and start with a clean slate. 

Whenever I try to login to MAILMAN I get the following error message:
Bug in Mailman version 2.0beta2

We're sorry, we hit a bug!
If you would like to help us identify the problem, please email a copy of this page to the webmaster for this site with a description of what happened. Thanks! 

Traceback:

Traceback (innermost last):
  File "/usr/local/mailman/scripts/driver", line 89, in run_main
    main()
  File "/usr/local/mailman/Mailman/Cgi/admin.py", line 68, in main
    mlist = MailList.MailList(listname)
  File "/usr/local/mailman/Mailman/MailList.py", line 69, in __init__
    self.Load()
  File "/usr/local/mailman/Mailman/MailList.py", line 858, in Load
    self.Lock()
  File "/usr/local/mailman/Mailman/MailList.py", line 1300, in Lock
    self.__lock.lock()
  File "/usr/local/mailman/Mailman/LockFile.py", line 209, in lock
    os.link(self.__lockfile, self.__tmpfname)
OSError: [Errno 28] No space left on device



--------------------------------------------------------------------------------

Python information:
Variable Value 
            sys.version 1.5.2 (#1, Jan 20 2000, 22:40:08) [GCC 2.8.1]  
            sys.executable /usr/local/bin/python  
            sys.prefix /usr/local  
            sys.exec_prefix /usr/local  
            sys.path /usr/local  
            sys.platform sunos5  
    



--------------------------------------------------------------------------------

Environment variables:
Variable Value 
            DOCUMENT_ROOT  /home3/awdesigns/public_html  
            SERVER_ADDR  209.40.121.211  
            HTTP_ACCEPT_ENCODING  gzip, deflate  
            SERVER_PORT  80  
            PATH_TRANSLATED  /home3/awdesigns/public_html/hotlist  
            REMOTE_ADDR  206.133.238.105  
            HTTP_ACCEPT_LANGUAGE  en-us  
            GATEWAY_INTERFACE  CGI/1.1  
            SERVER_NAME  awdesigns.com  
            TZ  US/Eastern  
            HTTP_USER_AGENT  Mozilla/4.0 (compatible; MSIE 4.01; Windows 95)  
            HTTP_ACCEPT  image/gif, image/x-xbitmap, image/jpeg, image/pjpeg, application/msword, */*  
            REQUEST_URI  /mailman/admin/hotlist  
            PATH  /usr/sbin:/usr/bin  
            QUERY_STRING   
            SCRIPT_FILENAME  /usr/local/mailman/cgi-bin/admin  
            PATH_INFO  /hotlist  
            HTTP_HOST  www.a-webdesigns.com  
            REQUEST_METHOD  GET  
            SERVER_SIGNATURE  Apache/1.3.12 Server at awdesigns.com Port 80 
            SCRIPT_NAME  /mailman/admin  
            SERVER_ADMIN  webmaster at awdesigns.com  
            SERVER_SOFTWARE  Apache/1.3.12 (Unix) ApacheJServ/1.1.1b1 PHP/3.0.14 FrontPage/4.0.4.3 mod_ssl/2.6.2 OpenSSL/0.9.5  
            PYTHONPATH  /usr/local/mailman  
            REMOTE_PORT  1390  
            SERVER_PROTOCOL  HTTP/1.1  
            HTTP_CONNECTION  Keep-Alive  


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20000415/1dcab26f/attachment.html>

From fig at oreilly.com  Wed Apr 19 11:15:00 2000
From: fig at oreilly.com (Stephen R. Figgins)
Date: Wed, 19 Apr 2000 15:15:00 GMT
Subject: Writing for the python devcenter
References: <Pine.BSF.4.10.10004181241280.15729-100000@chi.pair.com> <38FCF25E.4D945D81@oreilly.com> <Gx7L4.176$Pa.3485@news>
Message-ID: <38FDCD2B.3CCB4A84@oreilly.com>

Robert Hicks wrote:
> 
> Do I write an article and then just email it to you?
> 
> Bob

It would be best if you checked in with me first.  I would hate to have
you spend all that effort for something I might reject.  Send me an
outline of what you want to do.  If I like it, we will talk about when
you can deliver it and I will get you contract. 

I haven't got author's guidelines for the network up yet (for some
reason it keeps slipping down my to do list) but you should take a look
at the author's guidelines at WebReview:

  http://www.webreview.com/pub/universal/authorguide.html

The audience isn't exactly the same but the information on writing for
the web is dead on. 

Stephen Figgins
fig at oreilly.com


From dgreen at kryten.kitenet.net  Wed Apr  5 00:07:13 2000
From: dgreen at kryten.kitenet.net (Dan Green)
Date: Wed, 05 Apr 2000 04:07:13 GMT
Subject: ASP web-type stuff
Message-ID: <RZyG4.3061$9o.5812531@news.magma.ca>

Ahoy, folks.
	I was looking over things like Embperl and Apache::ASP and liked the
idea. I would like the idea more if I could do with python. So I had a
couple of questions:

Is there such a thing?
Is there interest in such a thing?
Does anbody want to hear about my 75 lines of code hack that basically does
the trick, but needs a lot of work?

If nobody cares at all, I'll just shut up and go home. :)

For those of you who have no idea what I'm talking about, it's basically an
alternative to CGI that allows you to embed the code right into the HTML.

something like this:

<HTML>
	<BODY>
<%  for i in range(5): %>
	Loop item <%= i %><BR>
	</BODY>
</HTML>

would look like this when called from a browser:

<HTML>
	<BODY>
	Loop item 1<BR>
	Loop item 2<BR>
	Loop item 3<BR>
	Loop item 4<BR>
	Loop item 5<BR>
	</BODY>
</HTML>

Let me know. If you're interested, please e-mail me.

--
Dan Green           | "So What? It's your problem to learn to live with,
Developer           |  destroy us, or make us saints."
Rewind Technologies |  -- Ministry, "So What?"


From mal at lemburg.com  Wed Apr 26 14:01:48 2000
From: mal at lemburg.com (M.-A. Lemburg)
Date: Wed, 26 Apr 2000 20:01:48 +0200
Subject: [Python-Dev] Re: Python 1.6a2 Unicode bug (was Re: comparing strings 
 and ints)
References: <l03102805b52ca7830b18@[193.78.237.154]>
Message-ID: <39072F0C.5214E339@lemburg.com>

Just van Rossum wrote:
> 
> I wrote:
> >A utf-8-encoded 8-bit string in Python is *not* a string, but a "ByteArray".
> 
> Another way of putting this is:
> - utf-8 in an 8-bit string is to a unicode string what a pickle is to an
> object.
> - defaulting to utf-8 upon coercing is like implicitly trying to unpickle
> an 8-bit string when comparing it to an instance. Bad idea.
> 
> Defaulting to Latin-1 is the only logical choice, no matter how
> western-culture-centric this may seem.

Please note that the support for mixing strings and Unicode
objects is really only there to aid porting applications
to Unicode.

New code should use Unicode directly and apply all needed
conversions explicitly using one of the many ways to
encode or decode Unicode data. The auto-conversions are
only there to help out and provide some convenience.

-- 
Marc-Andre Lemburg
______________________________________________________________________
Business:                                      http://www.lemburg.com/
Python Pages:                           http://www.lemburg.com/python/




From nascheme at enme.ucalgary.ca  Mon Apr  3 21:38:41 2000
From: nascheme at enme.ucalgary.ca (Neil Schemenauer)
Date: 4 Apr 2000 01:38:41 -0000
Subject: [Python-Dev] New Features in Python 1.6
In-Reply-To: <20000403165621.A9955@cnri.reston.va.us>
References: <200004011740.MAA04675@eric.cnri.reston.va.us> <200004031604.MAA05283@eric.cnri.reston.va.us> <20000403165621.A9955@cnri.reston.va.us>
Message-ID: <20000404013841.15629.qmail@cranky.arctrix.com>

In comp.lang.python, you wrote:
>You people are *so* gullible!  ;-)

Well done.  You had me going for a while.  You had just enough
truth in there.  Guido releasing the alpha at that time helped
your cause as well.


    Neil

-- 
Tact is the ability to tell a man he has an open mind when
he has a hole in his head.



From alwagner at tcac.net  Thu Apr 13 13:03:48 2000
From: alwagner at tcac.net (Albert Wagner)
Date: Thu, 13 Apr 2000 12:03:48 -0500
Subject: Doesn't anybody write to paper anymore?
Message-ID: <38F5FDF3.EE2FBA01@tcac.net>

I've searched the docs and my "Learning Python" book and I can't find
anything on printing to a real printer.  Through trial and error I
finally made the following work:

printer = open('/dev/lp0', 'w')
printer.write("page 1\n\r\f")
printer.write("page 2\n\r\f")
printer.close

But it writes directly to the printer port.  I would like to pass it
into the queue.  IO redirection in the shell is ugly and awkward.  Also,
assigning "printer" to "os.stdout" does not work.  Someone please help a
newbie.
-- 
Small is Beautiful


From warlock at eskimo.com  Tue Apr  4 01:14:07 2000
From: warlock at eskimo.com (Jim Richardson)
Date: Tue, 04 Apr 2000 05:14:07 GMT
Subject: Advice requested: GUI project beginning 
References: <m12avmV-000wcEC@swing.co.at>
Message-ID: <slrn8eiuu1.b3t.warlock@raven.myth>

On Fri, 31 Mar 2000 09:23:47 +0200, 
 Christian Tanzer, in the persona of <tanzer at swing.co.at>,
 brought forth the following words...:

>
>"Russell E. Owen" <owen at astroNOJNK.washington.edu.invalid> wrote:
>
>> I am starting work on a telescope remote control user interface. It's a 
>> big program with lots of windows and controls and display graphics 
>> supporting the telescope and numerous instruments.
>
>Nice project.
>
>> Python is a leading contender because I'd like a system that offers:
>> - rapid development/change/test
>> - robust
>> - safe (automatic garbage collection and bounds checking)
>> - runs under both unix and MacOS, if possible (windows wouldn't hurt, 
>>   either)
>> - will probably be around in 10 years
>
>IMHO, Python is a very good bet on all counts.
>
>> Any comments? If I do switch, is Python development or support notably 
>> superior under unix or windows -- aside from OS preferences? (Please no 
>> OS flame wars.)
>
>I'm developing on Linux while most of the production code runs on
>Win32.
>
>If you have free choice, Linux would be the better platform for
>controlling a telescope than either Mac or Win32. The observers
>wouldn't like to have to reboot in the middle of the night on a
>regular basis.

Not discourage you from writing the app yourself, but there is a 
linux Telescope control app at http://206.169.225.102/eots.html
certainly worth a look at least. I know nothing of the app itself, don't
know if it would fit the bill for you. 

-- 
Jim Richardson
	Anarchist, pagan and proud of it
WWW.eskimo.com/~warlock
	Linux, because life's too short for a buggy OS.



From tismer at tismer.com  Sun Apr 16 12:53:47 2000
From: tismer at tismer.com (Christian Tismer)
Date: Sun, 16 Apr 2000 18:53:47 +0200
Subject: [ANNOUNCE] Garbage collection for Python
References: <20000407221750.A29279@acs.ucalgary.ca>
	 <m3r9cagx3c.fsf@atrus.jesus.cam.ac.uk> <l03102809b51d41895a80@[193.78.237.135]>
Message-ID: <38F9F01B.57273052@tismer.com>


Just van Rossum wrote:
> 
> Michael Hudson <mwh21 at cam.ac.uk> wrote:
> > To someone who might know: what hope is there for this to go
> > into 1.6?
> [ ... ]
> > It would be, in my opinion, be a really really really good thing; for
> > one thing it would help in language wars...
> 
> Fredrik Lundh wrote:
> > sorry, but stackless python is the only really really really good
> > thing that hasn't yet been added to the core.  this is just really
> > really good.

Hey! If Stackless goes with GC, can we get 5 times "really"? :-))

> Oh how I agree... (Now lets hope the new Python GC isn't
> incompatible with Stackless... Project anyone? ;-)

I'm currently looking into that. Seems to be *very* possible
and *very* worthy to be done. Many cyclic issues with
continuations are much easier and probably more efficient
to solve with *the* GC at hand, similar to Mikecode Hackson's
observation.

> While were at it, here's a little Stackless news:
> - Christian is redesigning the continuation module, it already has
>   improved quite a bit over the last two weeks.

It will become simpler, simpler and simpler.

> - Even so, Chris and I agree that continuations should eventually
>   be completely hidden from the user, since it's so incredibly
>   obscure, hardly anyone will understand it. Not good for CP4E ;-)

Today I'm saying so. But I think there will come a time
where people are told "look, a continuation is as simple as 1+1",
and they'll say "sure? I thought a co is simpler."

...
> - Dunno about coroutines: it appears they are much harder to get
>   right, and the people on the python-coro at egroups.com list only
>   seem to make slow progress. (But then again, I've only been
>   there for a week or so, so who am I to say that...)

That's not a continuation matter. The problem is to get
a really really really (really? :) good coroutine design
together that fits everybody's expectations. Expressing
them with continuation is just a homework. Doing the
right design is very hard.

<snip>

until-Tim-Peters-tells-us-the-way-hopefully-<0.90000000000000002 wink>

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaunstr. 26                  :    *Starship* http://starship.python.net
14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     where do you want to jump today?   http://www.stackless.com



From glyph at twistedmatrix.com  Fri Apr 21 15:06:28 2000
From: glyph at twistedmatrix.com (Glyph Lefkowitz)
Date: 21 Apr 2000 14:06:28 -0500
Subject: Pickling Robustly
In-Reply-To: m.faassen@vet.uu.nl's message of "21 Apr 2000 13:11:14 GMT"
References: <jak8huf1of.fsf@helix.twistedmatrix.com> <8dpk1i$7de$1@newshost.accu.uu.nl>
Message-ID: <jazoqn2s0b.fsf@helix.twistedmatrix.com>

m.faassen at vet.uu.nl (Martijn Faassen) writes:

> Glyph Lefkowitz <glyph at twistedmatrix.com> wrote:
> 
> The Digicool people must've run into similar problems with their
> Zope Object Database, which is based on Pickle. Perhaps looking at
> the ZODB wouldn't be a bad idea, if you haven't already, therefore:
> 
> http://www.python.org/workshops/2000-01/proceedings/papers/fulton/zodb3.html
> http://www.zope.org/Members/michel/HowTos/ZODB-How-To
> http://www.zope.org/Documentation/Developer/Models/ZODB
> 
> Though I'm not sure if they indeed have solved your problem, as far
> as I understand it.

Thanks for the link!  This is some great reading, and gives me a
number of things to think about.  However, one of the first things I
noticed was:

      The ZODB stores objects in Python pickle format [van
      Rossum99]. All of the rules for pickling objects apply. See the
      documentation for the Python pickle module for more details.

... so I don't think it *quite* solves my problems :)

(It also imposes some other inconvenient restrictions, such as the necessity of subclassing from Persistent and allowing it to override __setattr__ and __getattr__, which I need to have overridden for various other reasons).

Perhaps I should ask this question in a different way.  How does
Pickle tell if an object is pickleable or not?  Is there some
pickle.ispicklable method I could refer to?

-- 
                  __________________________________________
                 |    ______      __   __  _____  _     _   |
                 |   |  ____ |      \_/   |_____] |_____|   |
                 |   |_____| |_____  |    |       |     |   |
                 |   @ t w i s t e d m a t r i x  . c o m   |
                 |   http://www.twistedmatrix.com/~glyph/   |
                 `__________________________________________'



From pinard at iro.umontreal.ca  Wed Apr 26 11:17:40 2000
From: pinard at iro.umontreal.ca (=?ISO-8859-1?Q?Fran=E7ois_Pinard?=)
Date: 26 Apr 2000 11:17:40 -0400
Subject: The Simple Economics of Open Source
In-Reply-To: Albert Wagner's message of "Fri, 21 Apr 2000 11:56:03 -0500"
References: <14591.15630.192472.440375@goon.cnri.reston.va.us> <39008823.A3B67E8@tcac.net>
Message-ID: <oqitx4yjqj.fsf@titan.progiciels-bpi.ca>

Albert Wagner <alwagner at tcac.net> writes:

> An elegant algorithm can be as beautiful as a painting or a song.
> The truely creative desperately need to share their creations.

This is a bit like I feel :-).  Just introspecting myself a little, I would
say that my real initial motivation for sharing software was my mere way
to thank the community who gave so much good software to me.  It spared
me a lot of time, and I felt I should return part of this spared time to
the community.  It then became sort of an habit in the long run.

About why I feel like thanking people?  It's just that I've been educated
that way.  It's builtin, kind of hardware.  A factory defect! :-)

-- 
Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard





From gregm at iname.com  Thu Apr 27 08:38:47 2000
From: gregm at iname.com (Greg McFarlane)
Date: Thu, 27 Apr 2000 22:38:47 +1000
Subject: Scrollbar needed to get attached to Canvas
In-Reply-To: <3903BA47.A7CABC48@swt.edu>; from Shah, Navneet on 23 Apr 2000 at 10:06:47PM
References: <3903BA47.A7CABC48@swt.edu>
Message-ID: <20000427223847.59255@nms.otc.telstra.com.au>

You should not pack widgets into a canvas, you should use
create_window, like this:
    self.sc.create_window(200, 200, anchor='nw', window = button)

But maybe you should be using a Pmw.ScrolledFrame instead of a
Pmw.ScrolledCanvas if all you want to do scroll around a bunch of
widgets.


On 23 Apr, Shah, Navneet wrote:
> I am using Pmw Mega widget. I  already  have canvas with buttons in it.
> How can I attached scrollbar to the canvas. Any help will be
> appreciated.
> 
> Shah
> 
> My code goes as follows:
> 
> class Demo:
>     def __int__(self,parent)
>         #Create the canvas
>         self.sc=Pmw.ScrolledCanvas(parent,
>                      borderframe=1,
>                      labelpos ='n',
>                      usehullsize = 1,
>                      hull_width = 400,
>                      hull_height = 300,
>            )
> 
>           self.sc.pack()
>           button=Tkinter.Button(self.sc.interior(), text='Hello')
>           button.pack(side='left')
> 
> 
> 
> 
> -- 
> http://www.python.org/mailman/listinfo/python-list
> 

-- 
Greg McFarlane     INMS Telstra Australia     gregm at iname.com



From jae at ilk.de  Fri Apr 21 23:43:13 2000
From: jae at ilk.de (Juergen A. Erhard)
Date: Sat, 22 Apr 2000 05:43:13 +0200
Subject: The Simple Economics of Open Source
In-Reply-To: <slrn8g26tu.p7o.neelk@brick.cswv.com>
References: <14591.15630.192472.440375@goon.cnri.reston.va.us> <slrn8fv2rd.1kt.neelk@brick.cswv.com> <3900611F.3B525BDA@cfdrc.com>  <FtDG4z.774@world.std.com> <slrn8g26tu.p7o.neelk@brick.cswv.com>
Message-ID: <22042000.2@sanctum.jae.ddns.org>

----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

>>>>> "Neel" == Neel Krishnaswami <neelk at brick.cswv.com> writes:

    Neel> Will Ware <wware at world.std.com> wrote:
    >> Ed (elb at cfdrc.com) wrote:
    >> > ... I didn't think much of the article...
    >> > ...they gave the impression that programmers behave
    >> > somewhat like monkeys, motivated primarily by concerns
    >> > about group status and dominance...
    >> 
    >> They also mentioned, but immediately belittled, the possibility
    >> that altruism might be a real motivation. They may have been
    >> projecting their own thinking upon programmers in doing so. 

    Neel> One of the interesting things about this thread is that it has given
    Neel> me a bit of insight into what economics looks like to people who
    Neel> haven't internalized its assumptions. I'm afraid that I'm tainted
    Neel> enough that it no longer looks like the dismal science to me. :)

"Internalized its assumptions".  Well put... now hold that thought...

    Neel> Seriously, the rejection of altruism as a possible motive strikes me
    Neel> as a very reasonable decision. Altruism is generally not sufficient
    Neel> motive in other aspects of life -- [...]

Same goes for egoism... see below.

    Neel> Second, economists assume that everyone is basically self interested,
    Neel> including themselves -- it's a standard analytic assumption, not
    Neel> snobbery.

But it is not just a `basic analytic assumption' in most cases... most
people in the field, as you yourself have stated above, have
`internalized its assumptions'.  Now, you could argue that you didn't
mean it like *that*... but my impression of people in that field of
science[1] is that the *did* indeed internalize this... which means
that *everything*, ever single point of data that goes into an
analysis, into a study, is colored and filtered by this worldview.

Now, I think the fundamental flaw here is that people only see this in
black and white: it's either altruistic or egoistic.  "Okay, so it
might be some shade in between."  Which still is basically black and
white... just a mix of the two.

No, I think (and this thought just occurred to me, so it's pretty
fresh) that there's a third thing... it's the answer to the old
question asked of a mountaineers: "Why did you climb that mountain?".
"Well... because it was there."  Or, in other words, "because I had
to".[4]

Egoism is often equated with the profit motive these days (little
wonder in our money-based society...[2])

Now this is a nice explanation for many things... but not for all.
I'm certain that the greatest thinkers, the greatest musicians,
artists... the greatest innovators were *not* motivated by that...

For example, when Einstein wrote his famouse relativity
theories... did he do it for the money?  Well, certainly not (he was a
teacher at the time, I think, and did it in his spare time).

Did he do for fame?  Hmmm... I don't do something for fame that comes
years later... if I want fame, I want it *now*!

So... why did he do it?

Well, I think because he *had* *to*.

Some writer once said that there are books that one writes... and then
there are books that *demand* to be written.  It's something you
because you want to... but because you have to.

I think this can be compared to egoism and altruism like this:

  Egoism is "I get something".

  Altruism is "`They' get something".

  Xyzzy (for lack of a name[3]) is "I get *rid* of something".  Or "I
  get something of my chest".

Ranting-is-my-middle-name-(Chorus:-"Obviously!")-ly y'rs, J

[1] If it can be called a science... in the end, you have to deal with
people (because even big corporations are evantually directed by
people), and then you get into the uncertainty areas of psychology and
sociology (which are sciences, IMHO... they're just not as hard as
things like physics or chemistry (though some areas of physics are
even softer than psychology ;-)

[2] I can rant at length about that, too... ;-)

[3] Actually, I guess the old greeks probably had a word for *that*,
too...  I'm just not that well-read in (greek) philosophy... I just
dabble in it (because I must ;-)

[4] I'm sure someone has found some weird explanation that makes this
something egoistical...

- -- 
J?rgen A. Erhard      eMail: jae at ilk.de      phone: (GERMANY) 0721 27326
     MARS: http://members.tripod.com/Juergen_Erhard/mars_index.html
                    "Ever wonder why the SAME PEOPLE
      make up ALL the conspiracy theories?" -- Michael K. Johnson
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.1 (GNU/Linux)
Comment: Use Mailcrypt and GnuPG <http://www.gnupg.org/>

iEYEARECAAYFAjkBH84ACgkQN0B+CS56qs256ACeKvBlNlglVqK2jsWsMCjUV/IH
Hm4An2VP08D5RFccgGmGw4FohYwvPHHg
=SV1u
-----END PGP SIGNATURE-----



From ajung at sz-sb.de  Tue Apr  4 16:58:10 2000
From: ajung at sz-sb.de (Andreas Jung)
Date: Tue, 4 Apr 2000 22:58:10 +0200
Subject: Getting the argument names passed to a function ?
Message-ID: <20000404225810.A12156@sz-sb.de>

def myfunc(arg1,arg2):
	...


a=1;b=2
myfunc(a,b)

Is it possible to get the name of variables that  were
used as arguments to call myfunc(). myfunc() should be
able to determine the names 'a' and 'b'. Don't ask
me why I need this - it's for a worse Python hack :-)

Cheers,
Andreas




From greg at cosc.canterbury.ac.nz  Wed Apr 12 22:56:48 2000
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Thu, 13 Apr 2000 14:56:48 +1200
Subject: Musing - Algebraic types? (was Tuples...)
Message-ID: <38F53770.230E6DA@cosc.canterbury.ac.nz>

The recent discussion about lists vs. tuples got me
thinking about algebraic types. It was mentioned that
an alternative to tuples was to define a class with
an appropriate __init__ method. This is true, but it
can get tedious writing __init__ methods that do
nothing but stuff their arguments into instance
variable.

So I was thinking about a shorthand syntax for
defining such classes, based on the way that algebraic
types are defined in Haskell-like languages. Maybe
something like

  class Statement(ParseNode) = (
    IfStatement(expr, then_stmts, else_stmts),
    WhileStatement(expr, body),
    ForStatement(var, expr, body)
  )

which would be equivalent to writing

  class Statement(ParseNode):
    pass

  class IfStatement(Statement):
    def __init__(self, expr, then_stmts, else_stmts):
      self.expr = expr
      self.then_stmts = then_stmts
      self.else_stmts = else_stmts

etc. etc.

Then, of course, we're going to want to be able to
unpack these:

  IfStatement(b, t, e) = my_if_statement

would be equivalent to

  b = my_if_statement.expr
  t = my_if_statement.then_stmts
  e = my_if_statement.else_stmts

and to be really useful we want some sort of case-statement
with unpacking:

  switch my_statement:
    case IfStatement(b, t, e):
      ...
    case WhileStatement(b, s):
      ...
    case ForStatement(v, e, s):
      ...

(expansion left as an exercise for the reader - this post
is long enough already!)

What think ye all?

-- 
Greg Ewing, Computer Science Dept,
+--------------------------------------+
University of Canterbury,	   | A citizen of NewZealandCorp, a	  |
Christchurch, New Zealand	   | wholly-owned subsidiary of USA Inc.  |
greg at cosc.canterbury.ac.nz	   +--------------------------------------+


From gklix at hassler.de  Thu Apr 13 05:39:43 2000
From: gklix at hassler.de (Gerald Klix)
Date: Thu, 13 Apr 2000 11:39:43 +0200
Subject: No method overloadin I suppose ?
References: <38F4938E.8F245B68@altern.org>
Message-ID: <38F595DE.B413D191@hassler.de>

Thomas SMETS wrote:

> ## Beginning of Python program ...
> class Lister:
>   def __init__(self):
>     print 'In constructor ...'
>     __init__(1)
>
>   def __init__(self, lValue):
>     print 'In constructor with parameter'
>     self.l = lValue
>
>   def __repr__(self):
>     print '\n******\n\tValue is : ', l, '\n********'
>
> if __name__ == "__main__":
>   x = Lister()
>   y = Lister (2)
> ## End of python program
>
> Trace of execution is :
> Traceback (innermost last):
>   File "Minheritence.py", line 14, in ?
>   x = Lister()
> TypeError: not enough arguments; expected 2, got 1
> End of trace ...
>
> I guess I can't do methods overloadin', as shown... WHY is that ? I'm
> wrong ?
>
> Thomas,
>
> --

Yes Thomas, you can't do overloading. The second __init__ method
overides the first one.
But you can do:
def __init__( self, lValue = 1 ):
 self.l = lValue

Gerald



From dworkin at ccs.neu.edu  Thu Apr  6 12:47:23 2000
From: dworkin at ccs.neu.edu (Justin Sheehy)
Date: 06 Apr 2000 12:47:23 -0400
Subject: More about variables
In-Reply-To: Michael Scharf's message of "Thu, 06 Apr 2000 18:22:38 +0200"
References: <meh9-5C17A0.10445406042000@news.cit.cornell.edu> <38ECB9CE.44A73C13@gmx.de>
Message-ID: <vndk8ib9nxg.fsf@camelot-new.ccs.neu.edu>

Michael Scharf <Michael.Scharf at gmx.de> writes:

> Just do:
> 
> for i in range(number_of_variables):
>     vars()["variable%s" %(i+1)]=[]

Don't do that unless you don't mind the strong possibility that a
future version of Python will cause your code to break.  The above
relies on an implementation detail that is subject to change.

>From the standard reference:

vars ([object]) 
       Without arguments, return a dictionary corresponding to the
       current local symbol table. With a module, class or class
       instance object as argument (or anything else that has a
       __dict__ attribute), returns a dictionary corresponding to the
       object's symbol table. The returned dictionary should not be
       modified: the effects on the corresponding symbol table are
       undefined.

-Justin

 



From rcunning at acm.org  Thu Apr 13 06:17:47 2000
From: rcunning at acm.org (Robert W. Cunningham)
Date: Thu, 13 Apr 2000 10:17:47 GMT
Subject: 1st non-trivial program - gentle criticism requested
References: <1e90r16.1sd2ts41e7ur6kN@gershwin.bio.ic.ac.uk>
Message-ID: <38F59E2E.308BD8C4@acm.org>

I couldn't get it to run under Linux until I trivially initialized the
following variables:

gTarget = ""
gNumGenerations = gMutationRate = gStepSize = None


-BobC




From effbot at telia.com  Thu Apr 13 16:30:51 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Thu, 13 Apr 2000 20:30:51 GMT
Subject: Simple formatting string question
References: <meh9-637743.14563813042000@news.cit.cornell.edu> <8d57gp$1om@gap.cco.caltech.edu>
Message-ID: <%7qJ4.929$rc9.193537024@newsb.telia.net>

Robert Kern <kern at caltech.edu> wrote:
> GuruQuestion:
> Should implicit string concatenation have a different precedence than
> explicit string concatenation?  Perhaps only a mention in
> http://www.python.org/doc/current/ref/string-catenation.html ?

what do you mean?

that page describes "string literal concatenation", and clearly
says that multiple adjacent string literals are concatenated at
*compile* time.

operator precedence has nothing to do with it -- it's not an
operator, and it's not documented as being one.

</F>




From tismer at tismer.com  Wed Apr 19 06:51:45 2000
From: tismer at tismer.com (Christian Tismer)
Date: Wed, 19 Apr 2000 12:51:45 +0200
Subject: BAD starship.python.net news!
References: <38FC852A.9B11F458@concentric.net> <38FCBD77.A5925D65@darwin.in-berlin.de>
Message-ID: <38FD8FC1.EC5DDCD7@tismer.com>


"Dinu C. Gherman" wrote:
> 
> Manus Hand wrote:
> >
> > This from Mike McLay.  Everyone, let's pony up what it takes to recover
> > our starship!
> >
> > Manus (who will backup more and more often than he has until now)
> 
> Not a bad idea! Sigh...
> 
> I think Christian Tismer used to make regular copies of
> large parts of the Starship, at least in former times...
> 
> Thinking of the not so long ago shut-down of Starship Py-
> thon I (starship.skyport.net)... couldn't that be a source
> to recover substantial parts of the previous data??

Ironically, yes it could have been used. My last copy wasn't older
than a month. But I forgot about that when I gave up
starship.skyport, and I asked my ISP about it:
They wiped the machine last Friday.

cheerio, Starship "Titanic" Python - chris

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaunstr. 26                  :    *Starship* http://starship.python.net
14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     where do you want to jump today?   http://www.stackless.com



From ljz at asfast.com  Sat Apr  1 23:10:47 2000
From: ljz at asfast.com (Lloyd Zusman)
Date: 01 Apr 2000 23:10:47 -0500
Subject: New Features in Python 1.6
References: <200004011740.MAA04675@eric.cnri.reston.va.us> <_pyF4.25562$E85.640966@news1.rdc1.md.home.com> <slrn8edfmk.185.wtanksle@dolphin.openprojects.net>
Message-ID: <ltvh21ywfs.fsf@asfast.com>

wtanksle at dolphin.openprojects.net (William Tanksley) writes:

> On Sun, 02 Apr 2000 02:39:54 GMT, Eric Hagemann wrote:
> >What are the options for users wanting 1.6 that may not have access to
> >modules.python.org
> >on a regular basis or at all.  I'd hate to be disconnected from the net and
> >find I need that
> >'one' function that used to be in the standard library ......... e.g. can I
> >down load the
> >entire library and forgo the net check ?  What happens if there is no net --
> >will the import function
> >get unhappy ?
> 
> It's actually not a problem -- all modern OSes support TCP/IP even when
> you're disconnected from the net, and one particular address, 127.0.0.1,
> is always live.  This happens to be the site from which Python loads its
> modules (we have an agreement with the owner of that site, which also
> hosts many other products of the same nature, including some marketing
> apps).

Hey!  Wait a minute ... that's *my* site!  I never agreed to this!

Oh ... now I get it ... my cat went and signed another contract behind
my back.  You guys outbid me on catnip, didn't you?!

I've been having enough bandwidth problems lately ... and now thousands
of module downloads every day ... <sigh>

OK ... this is the last straw ... no more Mr. Nice Guy!  Here kitty
.. here kitty kitty kitty ... it's neutering time!


Well, I guess this isn't all that bad, because I can cut down on the
download bandwidth by compressing all those Python modules.  I'll use
the new compression algorithm I recently invented.  It's amazing!
.. it compresses *anything* down to one bit.

I'm still working on the de-compression algorithm, though ...

-- 
 Lloyd Zusman
 ljz at asfast.com


From m.faassen at vet.uu.nl  Thu Apr 27 14:24:12 2000
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 27 Apr 2000 18:24:12 GMT
Subject: Newbie Question: Is there a Python equiv to Perl's "do" statement?
References: <8e9rna$ajh$1@nnrp1.deja.com> <390881D9.5661AD51@tismer.com>
Message-ID: <8ea0kc$kik$2@newshost.accu.uu.nl>

Christian Tismer <tismer at tismer.com> wrote:


> mdefreitas at sikorsky.com wrote:
>> 
>> I'm trying to call a python script from another python script (given
>> the filename).

> Yes, you can use the execfile function to do that.

> execfile("otherscript.py")

And please do consider breaking up your system into modules which
get imported, instead of using execfile(). Generally a much cleaner
way to organize your code. And more powerful, to boot.

Regards,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From jmassung at magpiesystems.com  Wed Apr 26 16:41:26 2000
From: jmassung at magpiesystems.com (Jeff Massung)
Date: Wed, 26 Apr 2000 14:41:26 -0600
Subject: global variables or inheritence
References: <28p8zy0irgc.fsf@lada.kom.auc.dk> <eRHN4.1113$3K4.19980@news.uswest.net> <28p4s8oipzh.fsf@lada.kom.auc.dk> <j6IN4.1168$3K4.20204@news.uswest.net> <28pzoqghaoh.fsf@lada.kom.auc.dk>
Message-ID: <yqIN4.116$qj1.6200@news.uswest.net>

Thomas Rasmussen wrote in message <28pzoqghaoh.fsf at lada.kom.auc.dk>...
>  Jeff> NOTE: myvar is local to main() (not to the file)
>Well I still want to be able to do this....

Well, you can't :), so you're going to have to find another way to do what
you want.

>Well i'm programming a small admin tool, which must use the users
>$EDITOR for editing files. This works fine if the EDITOR sysenv is
>set, if not, then i want to ask the user which editor he want to use,
>and then pass this variable to the module that handles
>filemanipulation. This modules is also in another file if it
>matters... I must add that this is my first real program made in
>python...

Definitely no need for circular code there, just rethink it through some
more - try a simple program of what you want with only 1 file for now.

As for circular code, think of it like this:

def CodeA():
    def CodeB():
        CodeA()
    CodeB()

What happens? Ouch ;) Stay away... very bad...

Jeff




From JohnH at PHM.GOV.AU  Sun Apr  2 21:38:07 2000
From: JohnH at PHM.GOV.AU (Hirsch, John)
Date: Mon, 3 Apr 2000 11:38:07 +1000 
Subject: controlling Excel2000 with COM
Message-ID: <A90DD8D7B82FD311B4D00008C7DBD6EAB5EB5C@phm-exch.phm.gov.au>

Christer Fernstrom wrote:

	I tried the most obvious thing with COM with Excel 20000 in Windows
2000 (p.
	67 in "Programming on 32")

	>> import win32com.client
	>> xl = win32com.client.Dispatch("Excel.Application")
	  -- This launches Excel (as verified with the windows task manager)
	>> xl.visible =1 (I also tried to capitalise Visible)
	Result:
	Traceback (innermost last):
	  File "<pyshell#2>", line 1, in ?
	    xl.visible =1
	  File "D:\tools\Python\win32com\client\dynamic.py", line 453, in
	__setattr__
	    raise AttributeError, "Property '%s.%s' can not be set." %
	(self._username_, attr)
	AttributeError: Property 'Excel.Application.visible' can not be set.

	Anybody got a clue? Is this a problem related to Office2000?

	    christer
____________________________________________________________________________
______________

This works for me.

>>> import win32com.client
>>> xl = win32com.client.Dispatch("Excel.Application")
>>> xl.Visible = 1

Must be capital 'V'.

john





From smigowsky at dspace.de  Mon Apr  3 05:53:11 2000
From: smigowsky at dspace.de (Stefan Migowsky)
Date: Mon, 3 Apr 2000 10:53:11 +0100 
Subject: Q: List and Dicts as default args
Message-ID: <326536345498D311B3BC00105A39802A074654@newsintern.dspace.de>

Hi,

I was just wondering how to handle lists and dictionaries in 
a simple way as default arguments to functions. Since following
strange behaviour occured :

>>> def f(Index,List = [], Dict = {}):
...     List.append(1)
...     List.append(2)
...     print List
...     Dict[Index] = Index
...     print Dict
>>> f(1)
[1, 2]
{1: 1}
>>> f(2)
[1, 2, 1, 2]
{2: 2, 1: 1}
>>> f(3)
[1, 2, 1, 2, 1, 2]
{3: 3, 2: 2, 1: 1}

This behaviour only occurs with dictionaries and list. All other
types are "well" behaved. I could try the following but that 
doesn't look so nice: 

def f(Index,List = None, Dict = None):
    if not List: List = []
    List.append(1)
    List.append(2)
    print List
    if not Dict: Dict = {}
    Dict[Index] = Index
    print Dict


Thanks for any advice.

	Stefan 



From mhammond at skippinet.com.au  Fri Apr 28 19:26:12 2000
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Fri, 28 Apr 2000 23:26:12 GMT
Subject: win32all-131 at python.org
References: <1UDL4.539$v85.2477@news-server.bigpond.net.au> <8ecf70$766$1@nnrp1.deja.com>
Message-ID: <o6pO4.6819$v85.40065@news-server.bigpond.net.au>

"Matthias" <matthias_w at my-deja.com> wrote in message
news:8ecf70$766$1 at nnrp1.deja.com...
> Hy Mark :-)
> Are the sources anywhere ??
> (You told me to include the new startup-code in my project)

Only from CVS - available via the web at
http://www.pythonpros.com/cgi-bin/cvsweb.cgi/PyWin32

David Ascher creates a nightly tarball:
http://starship.python.net/crew/da/pywin32dists/


> BTW
> I have the following problems:
> (1) I want to use python-win for the interface in my programm, but it
> is most important to hide the interactive windows for most of the users
> (they should not know that they are powered by python), and

Ultimately you will end up with your own "intpyapp.py" (hopefully named
something else!)  This gives you full control over the app - eg, do you
want an editor, interactive, debugger, etc.

> (2) I have a conflict with my main windows. In the InitInstance() from
> my application the pyglue-code creates a new window and now visualc
> does not know which of them is the main window. (e.g. I open
> splittst.py in my window, have to choose debug (go is disabled ??) from
> the python-window and the splitter - test window is a mdi-child in my
> own main window again (little confuseing, isn't it ;-) - the menus are
> quite the same, but in my window all python-related menus are diasbled)

You _definately_ need a new intpyapp.py - either drop the creation of the
main frame from your C++ code, or from the custom app.py

Here is some code taken from a real project that does the exact same
thing - the MFC app creates the frame, and Pythonwin attaches to it!

Mark.

# initapp.py
#
"The main startup code for my custom Pythonwin App."
#
import win32ui
try:
 debugging = win32ui.GetProfileVal("Whatever", "Debugging", 0)
except:
 debugging=1
# Keep this as short as possible, cos error output is only redirected to
# the interactive window if this runs OK.
# Keep as much logic in imported modules, as errors in these are much
better -
# the messages go somewhere!

# Redirect to the debugging device!
if debugging:
 import win32traceutil
 import sys
 print "Starting up:", sys.path

import sys
import win32con

import pywin
from pywin.framework import intpyapp, app
from pywin.mfc import afxres

class MyApp(intpyapp.InteractivePythonApp):
 def HookCommands(self):
  f = win32ui.GetMainFrame()
  # Probably dont want most of these if you are not exposing the
  # interactive Window, or any Python IDE features.
  f.HookCommand(self.OnViewBrowse,win32ui.ID_VIEW_BROWSE)
  f.HookCommand(self.OnFileImport,win32ui.ID_FILE_IMPORT)
  f.HookCommand(self.OnFileCheck,win32ui.ID_FILE_CHECK)
  f.HookCommandUpdate(self.OnUpdateFileCheck, win32ui.ID_FILE_CHECK)
  f.HookCommand(self.OnFileRun,win32ui.ID_FILE_RUN)
  f.HookCommand(self.OnFileLocate,win32ui.ID_FILE_LOCATE)
  f.HookCommand(self.OnInteractiveWindow, win32ui.ID_VIEW_INTERACTIVE)
  f.HookCommandUpdate(self.OnUpdateInteractiveWindow,
win32ui.ID_VIEW_INTERACTIVE)
  f.HookCommand(self.OnViewOptions, win32ui.ID_VIEW_OPTIONS)
  f.HookCommand(self.OnHelpIndex, afxres.ID_HELP_INDEX)
  f.HookCommand(self.OnFileSaveAll, win32ui.ID_FILE_SAVE_ALL)
  app.CApp.HookCommands(self)

 def InitInstance(self):
  self.ddeServer = None
  # Create the taskbar icon if necessary.
  if debugging:
   win32ui.CreateDebuggerThread()

  # Attach to (NOT CREATE) the main frame.
  # This code assumes the MFC app has ALREADY created the main frame
window,
  # so win32ui.GetMainFrame() will return this window.  By passing this
window to the
  # app.MainFrame() constructor, we attach to the existing window.
  self.frame = app.MainFrame(win32ui.GetMainFrame())
  # However, in this example, our main app does not create the status bar!
  self.frame._CreateStatusBar()

  self.HookCommands()

  from editor import editor
  # Allow HTML files to be edited by Pythonwin!
  # ('something' is a module that exposes an MFC resource ID - there
  # should be a string, icon, menu etc entries in the resources for this
ID)
  self.AddDocTemplate( editor.EditorTemplate(something.IDR_HTML_FILE) )

  # Create the interactive window if we want.
  from pywin.framework import interact
  interact.CreateInteractiveWindow()

  # Load internal modules we need (if you want :-)
  self.LoadSystemModules()

  # Minimise the interactive window.
  win32ui.PumpWaitingMessages()
  interactiveFrame = interact.edit.currentView.GetParentFrame()
  if interactiveFrame != win32ui.GetMainFrame():
   interactiveFrame.ShowWindow(win32con.SW_MINIMIZE)

appobj = MyApp()





From PClaerhout at CREO.BE  Wed Apr 26 10:11:20 2000
From: PClaerhout at CREO.BE (Pieter Claerhout)
Date: Wed, 26 Apr 2000 16:11:20 +0200
Subject: Saving pieces of the registry
Message-ID: <2B1262E83448D211AE4B00A0C9D61B03011152B9@MSGEURO1>

Hi,

does anyone know is there is a command in Python to save a part from the registry
to a .reg file, like you can do with regedit.exe? I would like for example like to save
all keys under HKEY_LOCAL_MACHINE\Software to a file. Anyone know if this is
possible or not?

Kind regards,


Pieter




From c.evans at clear.net.nz  Thu Apr 20 03:57:15 2000
From: c.evans at clear.net.nz (Carey Evans)
Date: Thu, 20 Apr 2000 07:57:15 GMT
Subject: Pythonware Website???? Gone on vacation with Starship? ;-))
References: <20000419154202.B480@owl.rhein-zeitung.de>
	<JCjL4.540$wYl.187815424@newsb.telia.net>
	<3d4s8yotiy.fsf@amarok.cnri.reston.va.us>
Message-ID: <87zoqp8ask.fsf@psyche.evansnet>

akuchlin at mems-exchange.org (Andrew M. Kuchling) writes:

> First Zope.org (down because an Ethernet cable got pulled out), then
> Starship (disk crash), now pythonware.com.  Has someone released evil
> whitespace-eating nanobots that also attack computer hardware?

>From my perspective, the problem's obvious.

Debian is going to release version 2.2 sometime soon.  We've had most
of the usual symptoms (new version of XFree86 released, new version of
Apache released, even a new version of Perl); but we're still waiting
for the requisite disk crash on master.debian.org.

Obviously when Guido announced that 1.6's release schedule was being
changed, he jinxed all the Python machines with the same problem
Debian has.

-- 
	 Carey Evans  http://home.clear.net.nz/pages/c.evans/

		 "I can't believe it's not Cthulhu!"


From aa8vb at yahoo.com  Wed Apr 26 06:39:27 2000
From: aa8vb at yahoo.com (Randall Hopper)
Date: Wed, 26 Apr 2000 06:39:27 -0400
Subject: what does XDR stand for? and what is it?
In-Reply-To: <002101bfaeb5$b57f5dc0$2801000a@iel.ie>
References: <002101bfaeb5$b57f5dc0$2801000a@iel.ie>
Message-ID: <20000426063927.A1240847@vislab.epa.gov>

Shaun:
 |im here looking and trying to figure out the xdrlib module, 
...
 |my question is what is XDR? what does it stand for?

Fredrik and Richard already gave you the what's-it-stand-for and Richard
mentioned RPC (a Sun protocol using XDR).  Here's a bit more detail on the
what's-it-for.  XDR is for transparently exchanging data between systems
which may have different representations for basic data types.

"Data types" being things like ints, floats, double-precision floats,
characters, etc.  "Representations" meaning things like:  word sizes, byte
orderings, floating-point formats, etc.

You use XDR to "flatten your data structures" into stream of bytes so you
can send it over the net or write it to a file (similar in concept to
Python's marshal and pickle modules).  On the other end, you "build them
back up" from the data stream to construct the heirarchy of data structures
originally sent.  These two processes are called serialization and
deserialization, respectively.

As Richard mentioned, the main protocol Sun developed using XDR for a data
representation was RPC (remote procedure call), which is used by (among
other things) NFS -- Network File System.  

Most UNICes nowadays (and maybe MSWin too) provide XDR C APIs with the OS,
since they have to have them for NFS anyways.

-- 
Randall Hopper
aa8vb at yahoo.com



From jwbnews at scandaroon.com  Wed Apr 19 14:45:06 2000
From: jwbnews at scandaroon.com (John W. Baxter)
Date: Wed, 19 Apr 2000 11:45:06 -0700
Subject: Python in the future
References: <38FDB8A7.E701F81F@sec.noaa.gov> <m3itxei05c.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <jwbnews-4F034E.11450619042000@news.olympus.net>

In article <m3itxei05c.fsf at atrus.jesus.cam.ac.uk>, Michael Hudson 
<mwh21 at cam.ac.uk> wrote:

> The point being that Python hasn't changed fast, and probably won't
> change that fast in the next five years either.

There's something of a bump in the road ahead..."Python3000" will break 
some existing scripts to change some things.  Aside from that bump, I'd 
expect the same sort of slow, careful, don't break the old stuff 
progression which Guido has wisely given us.

   --John

-- 
John W. Baxter   Port Ludlow, WA USA  jwbnews at scandaroon.com


From ullrich at math.okstate.edu  Tue Apr  4 14:21:19 2000
From: ullrich at math.okstate.edu (David C. Ullrich)
Date: Tue, 04 Apr 2000 13:21:19 -0500
Subject: Learning in Stereo:  Math + Python
References: <ua1fes81hmb12dh1klbv0p7d0373qhleps@4ax.com> <ur3fessapua3bi1k4e5g89i3ndbuejhabo@4ax.com>
Message-ID: <38EA329F.4D407309@math.okstate.edu>

Kirby Urner wrote:

> [...]
> And here's another fast converger [to pi] using every other
> Fibonacci number:
>
> def pifib(n):
>    # return approximation for pi using
>    # arctan(1) = 4 * SIGMA [ arctan(1.0/Fib(2i+1))]
>    sum = 0
>    for i in range(1,n+1):
>       sum = sum + math.atan(1.0/fibo(2*i+1))
>    return 4*sum

    I didn't know that arctan(1) was equal to that sum, fascinating.
But I don't quite see the point to using that as a way to "calculate"
arctan(1) - seems like the call to math.atan is cheating, if we're
gonna do that why not just call math.atan(1)?

    Now if we were calculating the arctangents by hand that might
be different - there's an easy and fairly efficient way to find
arctan(x) for |x| < 1 that doesn't work for |x| > 1 (and just
barely works for |x| = 1), ie the power series; so there's
a possible point to finding arctan(1) using arctan(1/fib) in that
case.
    You might note that pi/6 = arctan(1/sqrt(3)). The power series
for arctan(1/sqrt(3)) converges pretty fast, and you can make
the calculations go pretty quick as well, if you note that
(sqrt(3))**(2*k+1) = sqrt(3)*(3**k); you factor out one
sqrt(3) and in the inner loop you have almost nothing but
integer computations. (I got 7,000 digits of pi this morning
in about two minutes this way. Using ordinary floats it's

def floatpi(n):
  """Gets pi from the power series for arctan(1/sqrt(3)) -
factors out some stuff and combined pairs of terms... """
  eps = 10.0**(-n)
  pow = 1L
  term = 1.0 #just to make the first test succeed
  p=0.0
  j=0
  while abs(term) > eps:
    term=float(j+1)/((4*j+1)*(4*j+3)*pow)
    p=p+term
    pow=9*pow
    j=j+1
  return (16/math.sqrt(3))*p

The error's just a little larger than the first omitted term,
so floatpi(n) should give pi to n places. And it only takes
about n terms to get there. Probably worrying about the
"optimizations" doesn't make much sense here - makes
more sense when dealing with high-precision (slow) reals.)

    Or you could use someone's formula

pi = 16*atan(1/5) - 4*atan(1/239)

to get faster convergence (although explaining why the
formula is valid might not be quite as straightforward as
pi = 6*atan(1/sqrt(3)) - dunno if we mean to include
explanations of that sort of thing or not).

DU




From gmuller at worldonline.nl  Wed Apr  5 14:59:29 2000
From: gmuller at worldonline.nl (gerrit)
Date: Wed, 05 Apr 2000 20:59:29 +0200
Subject: A Mountain of Perl Books + Python Advocacy
References: <20000405053920.20579.qmail@web2102.mail.yahoo.com> <8cfe42$o02$1@nnrp1.deja.com> <3d3dp01t13.fsf@amarok.cnri.reston.va.us>
Message-ID: <38EB8D10.2BA39F8D@worldonline.nl>

"Andrew M. Kuchling" wrote:

> rzantow at my-deja.com writes:
> > the whole language, and buying more books. I'm really surprised there
> > are as many as 13 Python books on the market. I'd have thought there
> > was a need for maybe three, along with about two pamphlets and a poster.
>
> I'm not surprised, and think there's room for still more.  Hopefully
> the next wave of books will aim at specific topics: Python and
> {Tkinter,KDE,GTk+,wxWindows}, Python for numeric work, for Web stuff,
> for bioinformatics, for teaching programming, inside Zope, JPython for
> Java programmers who need to embed a scripting language, etc.

<...snip...>
I would like to see a book "Python for children". My children (9 and 11 years
old) are now learning Superlogo, with a dutch book "Programmeren voor Kinderen
deel 2 met Superlogo" (Programming for children part 2 with Superlogo). It is
quite detailed step by detailed step, too much detail for my taste.

Python plus tkinter could be as simple and (more) educational than Superlogo.
Of course the right level of interfaces should be supplied to make it succesful
(turtles, which is available, but also images sound and so on)

Regards Gerrit




From jason-usenet at mastaler.com  Mon Apr  3 03:36:12 2000
From: jason-usenet at mastaler.com (Jason R Mastaler)
Date: Mon, 3 Apr 2000 01:36:12 -0600 (MDT)
Subject: Integrating Pmw into Python dist
Message-ID: <20000403073612.5943.qmail@ashanti.in.mastaler.com>

Have the Python developers considered integrating Greg McFarlane's
wonderful Python megawidgets work into the core Python distribution?
I feel this would be an excellent addition.

Pmw offers a nice set of extra widgets and an easy way to add
functionality to Tkinter applications.  The only reservation I have
with using Pmw, is that it adds yet another package to the list of
extras that users must download and install before they can use my
software.  If Pmw was integrated, obviously applications which rely on
it would suddenly become much more portable.

What do others think of this suggestion?

---Jason




From hathawa2 at marshall.edu  Tue Apr 11 21:17:10 2000
From: hathawa2 at marshall.edu (Mark Hathaway)
Date: Tue, 11 Apr 2000 21:17:10 -0400
Subject: Huh? func_defaults, default values in function calls
References: <8cv5qj$35p$1@zinc.intern.netconnect.no>
 <8cv8ra$hio$1@pegasus.csx.cam.ac.uk> <nMGI4.42394$U4.182659@news1.rdc1.az.home.
Message-ID: <38F3EBBF.7B23@marshall.edu>

>>> I don't get it. Why isn't the dictionary reset to {'x':'a'} when i call
>>> the function the last time?
>>>
>>> >>> a.func_defaults
>>> ('', '', {'x': 'a', 'y': 'c'})
>>>
>>> Why is not {} the default for the z parameter?

>> According to my understanding of the Python docs, defaults are only
>> evaluated at function definition time, not at function invokation time.

Definately goofy!

>> As a result, if you set a mutable object (list, dictionary, etc) as a
>> default, it'll be set when the function is defined and then if the
>> function changes it, it'll stay changed for subsequent invokations.
>> The function effectively has a "side-effect", of modifying it's own
>> default!

This way you'll never really know what the default is going to be.
Ha ha ha ha. that's pathetic.

> That's exactly right. The idiomatic Python way of doing what Roger wants
> here is:
>
> def a(x='',y='',z=None):
>    if z is None:
>       z = {}
>    if x:
>       z['x'] = x
>    if y:
>       z['y'] = y
>    print str(z)
>
> I think this is the first time I've seen show up with dictionaries, usually
> is shows up with lists.

At least this idiom gets the job done, but it sincerely defeats the
idea of having default values to have to say "z=None".


Mark S. Hathaway
email: hathawa2 at marshall.edu


From hamish_lawson at yahoo.co.uk  Wed Apr  5 08:24:23 2000
From: hamish_lawson at yahoo.co.uk (Hamish Lawson)
Date: Wed, 05 Apr 2000 05:24:23 -0700
Subject: UserDict's has_key() and get() should use __getitem__()
References: <0535db98.ec878f67@usw-ex0101-005.remarq.com> <m3u2hgn5jg.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <29328840.06b5b87b@usw-ex0101-005.remarq.com>

Michael Hudson wrote:

>Hmm. These are likely to be rather less efficient than the
>current implementation, so you want to penalise those who use
>UserDict for the sake of those who don't subclass it
>thoroughly - do you really want to do this?

I grant that my proposal is a bit less efficient, but I see that
as part of the cost for the proper abstraction. Given that
UserDict exists primarily to be subclassed, and usually for the
purpose of providing new definitions of setting and getting
items, I do think that UserDict should be implemented in such a
way as not to require those methods which are conceptually
dependent to be redefined in the derived class. If they are
conceptually dependent they should be made so in the code.

>Let's face it, you're unlikely to make this mistake again.

Except that I think subclassers should be able to make
this 'mistake'.

Hamish



* Sent from RemarQ http://www.remarq.com The Internet's Discussion Network *
The fastest and easiest way to search and participate in Usenet - Free!



From mwh21 at cam.ac.uk  Wed Apr 12 20:32:24 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 13 Apr 2000 01:32:24 +0100
Subject: comparing strings and ints
References: <2YmI4.366$rc9.188014080@newsb.telia.net> <20000411134358.A2352304@vislab.epa.gov> <1256641282-42160637@hypernet.com> <20000412075812.A2449004@vislab.epa.gov> <jr%I4.643$rc9.190628352@newsb.telia.net> <20000412112935.A2446939@vislab.epa.gov> <slrn8f9r9j.in3.grant@grante.comtrol.com>
Message-ID: <m3u2h6hmcn.fsf@atrus.jesus.cam.ac.uk>

grant at nowhere. (Grant Edwards) writes:

> No, it said "_consistently_ but arbitrarily" [emph added].
> That means that you don't which result you will get when a
> string and an int are compared, but you know you will always
> get the _same_ result.

That's one reading; however I'd have thought a more likely one would
be to specify that if x>y is true once for some x and y of differing
types, then x>y is true each subsequent time it is evaluated.

can-I-have-a-language-lawyer-brownie-point?-ly y'rs
M.

-- 
  please realize that the Common  Lisp community is more than 40 
  years old.  collectively, the community has already been where 
  every clueless newbie  will be going for the next three years.  
  so relax, please.                 -- Erik Naggum, comp.lang.lisp


From ddt92 at home.com  Tue Apr 11 18:55:35 2000
From: ddt92 at home.com (HC2k)
Date: Tue, 11 Apr 2000 22:55:35 GMT
Subject: I need Help
Message-ID: <38F3AD34.E6C88698@home.com>

Hello I am new to programming and someone said start with Python it's a
great learning language and can be very useful what excactly does python
program? And how long did it take anyone else to learn thanks !


From neilh at hare.net.au  Sun Apr  2 21:25:14 2000
From: neilh at hare.net.au (Neil Hodgson)
Date: Mon, 3 Apr 2000 11:25:14 +1000
Subject: Unicode program representation
Message-ID: <015c01bf9d0b$76f9b2a0$59dd3fcb@neil>

  I've been playing around with the Unicode support in Python 1.6.

  IDLE uses a Unicode enabled text widget so a script like the following:
print "hi ?"
   works well. There is a good chance this didn't travel over mail well, so
the string is 'h', 'i', ' '(space), '', (U+0429 cyrillic capital letter
shcha). The last letter looks like an angular 'W' with a tail.

   This saves as UTF-8, so it looks like IDLE's text widget stores data in
UTF-8 form. In my opinion, this is very much the right thing to do and I am
thinking of adding Unicode support to Scintilla so Pythonwin will also be
able to work like this.

   This means that if you want to use non-roman characters in string
constants then you should use 'ASCII' strings rather than the new u" form.
With:

i = 0
s = "hi ?"
print s
for x in s:
    print i, x
    i = i + 1
i = 0
su = u"hi ?"
print su
for x in su:
    print i, x
    i = i + 1

   Printing the 'ASCII' string did what I wanted but the Unicode form did
not.

   This leads to the question of what the use of the u" form is. The current
answer is that u" takes the ASCII string and makes a Unicode string object
by extending each byte with another zero byte. Because its a Unicode string
object it behaves appropriately with Unicode aware functions.

   I think this should be changed to interpreting the literal as a UTF-8
literal. The advantage here is that non-roman string literals become a
natural part of the language.

   The example scripts do not work well from the command line. Thought
should be given to whether on NT, the console should be reopened in wide
character mode so that Unicode I/O will work well.

   Saving the scripts as UCS-2 shows that the interpreter is unable to deal
with UCS-2 scripts, which is what I expected. I think very few people will
be creating script files in UCS-2, instead preferring to keep source code in
UTF-8 files.

   Neil





From culliton at clark.net  Sun Apr  9 12:55:34 2000
From: culliton at clark.net (Tom Culliton)
Date: Sun, 09 Apr 2000 16:55:34 GMT
Subject: awk-like usage of Python
References: <8cq5os$8b9$1@nnrp1.deja.com>
Message-ID: <aC2I4.16577$0o4.110574@iad-read.news.verio.net>

In article <8cq5os$8b9$1 at nnrp1.deja.com>,  <ejbaker_seekonk at my-deja.com> wrote:
>I'm a relative newbie to Python. But, I have extensive experience with
>awk (it's an old stand-by that I rely on over and over).
>
>Are there any library modules that make python behave more like awk?
>Lacking that, what's the best way to achieve line by line parsing
>and pattern matching with Python?

Here's an awk like base class written when I first started using
Python.  It still sees heavy use today as the base class for many
filters which summarize data from log files.  I've seen other more
elaborate ones but this has continued to be useful for me and others.

--------------------------------filter.py--------------------------------
import sys
import os

def perror(*elist) :
	for e in elist :
		sys.stderr.write(str(e) + " ")
	sys.stderr.write("\n")

class file_filter :

	def __init__(self, filenames) :
		self.filenames = filenames
		self.filename = "stdin"
		self.file = sys.stdin
		self.line = ""
		self.nr = 0
		self.fnr = 0

	def run(self) :
		self.nr = 0
		self.fnr = 0
		self.begin()
		if not self.filenames :
			self.filename = "stdin"
			self.file = sys.stdin
			self.do_file()
		else :
		    for self.filename in self.filenames :
			self.fnr = 0
			try :
				ext = os.path.splitext(self.filename)[1]
				if ext == ".Z":
					self.file = os.popen("zcat " + self.filename, "rb")
				elif ext == ".gz":
					self.file = os.popen("gzip -dc " + self.filename, "rb")
				elif ext == ".bz2":
					self.file = os.popen("bzip2 -dc " + self.filename, "rb")
				else:
					self.file = open(self.filename, "rb")
			except IOError, msg :
				perror("Cannot open:", self.filename, msg)
				continue
			self.do_file()
			self.file.close()
		self.end()

	def do_file(self) :
		while 1 :
			self.line = self.file.readline()
			if self.line == "" : break
			self.nr = self.nr + 1
			self.fnr = self.fnr + 1
			self.do_line()

	def do_line(self) :
		print self.line,

	def begin(self) :
		pass

	def end(self) :
		pass


From soundhack at hotmail.com  Thu Apr 20 11:23:52 2000
From: soundhack at hotmail.com (Robert)
Date: Thu, 20 Apr 2000 11:23:52 -0400
Subject: Why MFC instead of Tkinter -- was Re: Pythonwin?
References: <eMz2Nmiq$GA.73@net025s> <4QtL4.24$v85.41@news-server.bigpond.net.au> <chuL4.8483$O7.17315@typhoon.southeast.rr.com> <7LDL4.534$v85.2563@news-server.bigpond.net.au> <sfu6d657l6e184@corp.supernews.com>
Message-ID: <38ff21b8$0$10972@senator-bedfellow.mit.edu>

To each his own I guess. I work mostly on Windows platforms, and for me
MFC is a lot better than Tk for a couple of reasons.
1) I can port MFC/C++ programs a lot easier
2) I may be mistaken, but MFC based guis are faster than Tk
3) I really do not like using layout managers, and using Visual C++'s dialog
editor
is the best thing since sliced bread IMHO.

One suggestion for learning MFC fresh: don't read books that basically teach
you how to use Visual C++'s AppWizard. AppWizard is really nice and easy to
use but it basically automates so much of creating MFC programs that it was
hard (at least for me) to make the transition to a non-wizard MFC
environment in python.

Robert

Michael A. Crawford wrote in message ...
>I'm curious.  Why did you not steer him in the Tkinter direction for his
GUI
>needs?  Is MFC better?  I learned (and forgot MFC) some time ago, years in
>fact.  Same with X-Windows programming.  Now I'm working my way throught
>"Python and Tkinter Programming" by Grayson.  I thought Tkinter was the
>preferred method for Pythong GUI programming.  Am I mistaken?
>
>-Michael





From tjg at avalongroup.net  Tue Apr 25 00:20:27 2000
From: tjg at avalongroup.net (Timothy Grant)
Date: Mon, 24 Apr 2000 21:20:27 -0700
Subject: Anyone generating SWF content with Python?
References: <FVNM4.664$nb6.400276@news.pacbell.net>
Message-ID: <39051D0B.6FC168F2@exceptionalminds.com>

Grant Munsey wrote:

And here I am all excited that Python could possibly create something
having to do with Single White Females...

Imagine my disappointment!

-- 
Stand Fast,
    tjg.

Timothy Grant                         tjg at exceptionalminds.com
Chief Technology Officer              www.exceptionalminds.com
Red Hat Certified Engineer                      (503) 246-3630
Avalon Technology Group, Inc.              fax  (503) 246-3124
>>>>>>>>>>>>Linux...Because rebooting isn't normal<<<<<<<<<<<<



From btribble at blueshiftinc.com  Wed Apr  5 18:19:59 2000
From: btribble at blueshiftinc.com (Brett Tribble)
Date: Wed, 5 Apr 2000 15:19:59 -0700
Subject: os.utime limitation on Windows?
References: <senek66krfg118@corp.supernews.com>
Message-ID: <seneufksrfg51@corp.supernews.com>

Oh, I should mention that no error is reported by the second os.error
catch...

Brett

"Brett Tribble" <btribble at blueshiftinc.com> wrote in message
news:senek66krfg118 at corp.supernews.com...
> I have the following routine:
>
> def synctime(self, source, target):
>     try:
>         syncstat = os.stat(source)
>     except os.error:
>         self.statuswindow.insert('end', source + ': cannot get stats\n')
>     try:
>         os.utime(target, (syncstat[ST_ATIME], syncstat[ST_MTIME]))
>     except os.error:
>         self.statuswindow.insert('end', target + ': cannot change time\n')
>     return
>
> This works fine under NT on an NTFS drive, but doesn't seem to do anything
> on a 95 box on a FAT16 partition. Is this a limitation with 95, FAT16,
> Python, or a combination therein?
>
>
> Thanks in advance,
>
> Brett
>
>
>
>




From hoehn at cs.utk.edu  Wed Apr 12 14:01:53 2000
From: hoehn at cs.utk.edu (Jacob A Hoehn)
Date: Wed, 12 Apr 2000 14:01:53 -0400 (EDT)
Subject: a question about flushing input buffer
Message-ID: <Pine.GSO.4.10.10004121340380.4990-100000@castor4.cs.utk.edu>

I am expanding on the rolodex example given in the learning python book,
as a project for me to better learn python. I have run into a problem. I
have a class named Rolodex that uses the cmd module. One of the methods
for the class is to add a user to the rolodex. In the book example it just
has you enter the person's name and phone number. I want to alter this
so that you can also enter in additional information. So, in the add
method I have the following statement:

                print "Enter additional information (to stop, hit ^D):"
                try:
                        self.info[name] = sys.stdin.readlines()
                        sys.stdin.flush()
                except EOFError: return

The problem is, when I run the program I get the following:

ROLODEX> add foo
Enter phone number for foo: 911
Enter additional information (to stop, hit ^D):
hello
world
ROLODEX> list
=========================================
                 foo : 911                 
=========================================
ROLODEX> find foo
=========================================
The number for foo is 911.

Additional information:

hello
world
=========================================
ROLODEX> add guido
Enter phone number for guido: 111-1111
Enter additional information (to stop, hit ^D):
ROLODEX> list
=========================================
                 foo : 911                 
               guido : 111-1111            
=========================================
ROLODEX> find guido
=========================================
The number for guido is 111-1111.

Additional information:

=========================================
ROLODEX> 



It is not letting me enter in the additional information. It skips it as
if I had typed cntrl D. So, is there a buffer that I need to be flushing?
I thought sys.stdin.flush() would do that. Below is all of the code.
Thanks for any help you can give me and I appologize for the length of
this email.

Jacob Hoehn



#!/usr/local/bin/python

# rolodex, a simpler version of this exists in the oreilly
# learning python book... i added some stuff

import string, sys, pickle, cmd

class Rolodex(cmd.Cmd):

        def __init__(self):
                cmd.Cmd.__init__(self)
                self.prompt = "ROLODEX> "
                self.people = {}
                self.info = {}

        def help_add(self):
                print "Adds an entry (specifiy a name)"
        def do_add(self, name):
                if name == "": name = raw_input("Enter name: ")
                name = string.lower(name)
                if self.people.has_key(name):
                        option = ""
                        while 1:
                                if option == 'y': break
                                if option == 'n': return
                                option = raw_input("Name exists! " \
                                         "Overwrite? Enter [y/n]  ")
                phone = raw_input("Enter phone number for "+ name+": ")
                self.people[name] = phone
                print "Enter additional information (to stop, hit ^D):"
                try: 
                        self.info[name] = sys.stdin.readlines()
                        sys.stdin.flush()
                except EOFError: return

        def help_delete(self):
                print "Delete an entry (specify a name)"
        def do_delete(self, name):
                if name == "": name = raw_input("Enter name: ")
                name = string.lower(name)
                try:
                        del self.people[name]
                        del self.info[name]
                except KeyError:
                        print "Name not found"

        def help_find(self):
                print "Find an entry (specifiy a name)"
        def do_find(self, name):
                if name == "": name = raw_input("Enter name: ")
                name = string.lower(name)
                if self.people.has_key(name):
                        print '='*41
                        print "The number for %s is %s." % \
                        (name, self.people[name])
                        print "\nAdditional information:\n"
                        for line in self.info[name]: print line,
                        print '='*41
                else:
                        print "We have no record for %s." % (name,)

        def help_list(self):
                print "Prints the contents of the directory"
        def do_list(self, line):
                names = self.people.keys()
                if names == []: return
                names.sort()
                print '='*41
                for name in names:
                        print string.rjust(name, 20), ":", \
                        string.ljust(self.people[name], 20) 
                print '='*41

        def help_EOF(self):
                print "Quits the program"
        def do_EOF(self, line):
                sys.exit()

        def help_save(self):
                print "save the current state of affairs"
        def do_save(self, filename):
                if filename == "": 
                        filename = raw_input("Enter filename: ")
                saveFile = open(filename, 'w')
                pickle.dump(self.people, saveFile)
                pickle.dump(self.info, saveFile)

        def help_load(self):
                print "load a directory"
        def do_load(self, filename):
                if filename == "":
                        filename = raw_input("Enter filename: ")
                saveFile = open(filename, 'r')
                self.people = pickle.load(saveFile)
                self.info = pickle.load(saveFile)

if __name__ == '__main__':
        rolo = Rolodex()
        rolo.cmdloop()





From paul.magwene at yale.edu  Sun Apr 23 16:19:04 2000
From: paul.magwene at yale.edu (Paul Magwene)
Date: Sun, 23 Apr 2000 16:19:04 -0400
Subject: setting up Tkinter
References: <KlHM4.8299$4w1.119156@typhoon.nyroc.rr.com>
Message-ID: <39035AB8.5845D1EA@yale.edu>

Andy Mullen wrote:
> 
> As i understand from the website (www.python.com), this program doesn't run
> on windows 98. Since i was already thinking of putting linux on my machine
> under a partition of the HD, i was wondering if someone could point me in
> the right direction as to where to get linux online, how to go about
> install, and how to run the 2 separate OS's on one machine.  any response
> would be appreciated


If I read your email correctly, you're confused. Tk (and hence Tkinter)
DOES run on Windows (95/98,NT,2000, etc),various Unix and Unix-like
systems, and the Mac as well (though recent posts seem to indicate some
minor bugs in the current installer).  You don't need to install Linux
to work with Tkinter. 

--Paul


From tuttledon at hotmail.com  Thu Apr 20 08:55:15 2000
From: tuttledon at hotmail.com (Don Tuttle)
Date: Thu, 20 Apr 2000 12:55:15 GMT
Subject: [ANNOUNCE] Webware for Python 0.1
References: <5650A1190E4FD111BC7E0000F8034D26A0F40D@huina.oceanic.com> <38FEF62B.2A4402CD@mindspring.com>
Message-ID: <T6DL4.25197$cZ.34941@typhoon.southeast.rr.com>

WinZip is having problems with your tar.  It's giving me the error "Error
reading header after processing 0 times."

Don




From robin at jessikat.demon.co.uk  Sat Apr 22 14:45:30 2000
From: robin at jessikat.demon.co.uk (Robin Becker)
Date: Sat, 22 Apr 2000 19:45:30 +0100
Subject: The Simple Economics of Open Source
References: <14591.15630.192472.440375@goon.cnri.reston.va.us>
 <3900611F.3B525BDA@cfdrc.com> <FtDG4z.774@world.std.com>
 <slrn8g26tu.p7o.neelk@brick.cswv.com>
 <14837CB4E98A72FE.8FD547BA6E928594.5A103CFE68327EB7@lp.airnews.net>
Message-ID: <VLvUljAKNfA5EwPM@jessikat.demon.co.uk>

In article <14837CB4E98A72FE.8FD547BA6E928594.5A103CFE68327EB7 at lp.airnew
s.net>, Tres Seaver <tseaver at starbase.neosoft.com> writes
...
>"That dog won't hunt!"  Your claim that "basically no one makes the
>effort" is easily refuted:  consider the League of Women Voters,
>and the guides they put out before each election -- that effort is
>well-organized, and involves coordinated, non-partisan effort on the
>scale of tens of thousands of volunteers (much like a large Open Source
>project).  The League's explicit motive is to educate themselves and
>the population at large on the issues and candidates;  that at least
>some of the effort here is altruistic is obvious to anyone who hasn't
>"internalized [economics'] assumptions."
>
>Altruism-is-a-shibboleth'ly,
>
>Tres.
I don't think the pessimists amongst whom I count myself reject altruism
totally. A purely altruistic action has benefit for others and none for
the actor. Very few of the altruistic examples in the thread come close.

A lot of claims are made, but few are verifiable. I think the
internalized utility assumptions are a bit spurious as there is evidence
that many of the assumptions that economists make (about human
behaviour) are not verifiable and in some cases have been shown to be
true; a classic example is the assumption of risk averse behaviour in
investment decisions. I feel that the assumption that economic agents
(people) always behave as though they have a well defined utility is
false. I think that what economists are really saying is that they have
a model where the 'average' agent has such a utility and that the model
gives good results for certain problems. In this sense the average
volunteer feels they behave in a non-partisan altruistic way, but the
ensemble actually behaves selfishly although this is not selfishness in
the way we normally think of it. 
-- 
Robin Becker


From fwang2 at yahoo.com  Tue Apr  4 16:06:06 2000
From: fwang2 at yahoo.com (oliver)
Date: Tue, 4 Apr 2000 16:06:06 -0400
Subject: ldap support?
Message-ID: <Pine.LNX.4.10.10004041604050.10324-100000@worf.mcnc.org>

hi,

I wonder is there any standard LDAP support module in python? I looked up
SIGs, doesn't show anything there. I appreciate your help.

oliver



From sabren at manifestation.com  Thu Apr 20 03:29:59 2000
From: sabren at manifestation.com (Michal Wallace (sabren))
Date: Thu, 20 Apr 2000 03:29:59 -0400 (EDT)
Subject: globals: a dirty little secret?
Message-ID: <Pine.BSF.4.10.10004200321050.29174-100000@chi.pair.com>


Globals have come up a couple times recently.. If someone posted the
following approach, I missed it...


### step 1: globals.py #######################

import __builtin__

def setGlobal(key, value):
   __builtin__.__dict__[key] = value

__builtin__.setGlobal = setGlobal # makes the above function global



## step 2: your program ######################

import globals
setGlobal("x", 5)

print x

##############################################


Whatcha think?


Cheers,

- Michal
-------------------------------------------------------------------------
http://www.manifestation.com/         http://www.linkwatcher.com/metalog/
-------------------------------------------------------------------------




From hamish_lawson at yahoo.co.uk  Wed Apr  5 06:44:01 2000
From: hamish_lawson at yahoo.co.uk (Hamish Lawson)
Date: Wed, 05 Apr 2000 03:44:01 -0700
Subject: UserDict's has_key() and get() should use __getitem__()
Message-ID: <0535db98.ec878f67@usw-ex0101-005.remarq.com>

[This is a modified version of an earlier posting for which I
didn't get any responses. I'm trying again with what I'm hoping
might be a clearer exposition.]

The UserDict class in the standard library provides a framework
for defining your own dictionary-like classes. However I think
their may be an anomaly in how it is currently implemented (up
to 1.5.2). I think that most people would expect that if the
following statement succeeded:

    print d['Joe']

then the following line would evaluate true:

    print d.has_key('Joe')

However, under the current implementation of UserDict, if you
redefine __getitem__() in your own UserDict-derived class, you
will probably also have to redefine has_key() and get() (and
maybe others) in order to have them behave as expected. This
wholescale redefining is caused by the fact that UserDict's
has_key() and get() methods don't make use of the class's own
__getitem__() method in determining whether a key exists, as
shown in the extract below:

    def __getitem__(self, key):
        return self.data[key]
    def has_key(self, key):
        return self.data.has_key(key)
    def get(self, key, failobj=None):
        return self.data.get(key, failobj)

My contention is that the behaviour we expect from get() and
has_key() is intimately bound up with how __getitem__() behaves,
and this should be reflected in get() and has_key() making use
of __getitem__():

    def has_key(self, key):
        try:
            self[key]
            return 1
        except KeyError:
            return 0
    def get(self, key, failobj=None):
        try:
            return self[key]
        except KeyError:
            return failobj

It is now possible to redefine __getitem__() in a derived class
and have has_key() and get() behave as expected, without having
to redefine them also.

(It may be that other methods in UserDict should also be
reimplemented along these lines - I've not looked yet.)

Is there some reason why this isn't the right thing to do?
Assuming it is, I'd like to get this change into Python 1.6.

Hamish Lawson

-----------------------------------------------------------

PS: Below is an example demonstrating how the semantics of
has_key() and get() get broken under the current version of
UserDict when you redefine only __getitem__() in a derived class.


    import UserDict, string

    class CaselessDict(UserDict):
        "A dictionary class with case-insensitive keys"

        def __setitem__(self, key, value):
            self.data[string.lower(key)] = value

        def __getitem__(self, key):
            return self.data[string.lower(key)]


    d = CaselessDict()
    d['Joe'] = "Sue"
    print d['jOE']             # prints "Sue"
    print d.has_key('jOE')     # prints 0 but expected 1
    print d.get('jOE', None)   # prints None but expected "Sue"


* Sent from RemarQ http://www.remarq.com The Internet's Discussion Network *
The fastest and easiest way to search and participate in Usenet - Free!



From joachim at medien.tecmath.de  Tue Apr  4 05:28:37 2000
From: joachim at medien.tecmath.de (Joachim Koenig-Baltes)
Date: Tue,  4 Apr 2000 11:28:37 +0200 (MEST)
Subject: New Features in Python 1.6
In-Reply-To: <20000403131720.A10313@sz-sb.de>
References: <200004011740.MAA04675@eric.cnri.reston.va.us> <20000403131720.A10313@sz-sb.de>
Message-ID: <20000404092837.944E889@tmpc200.medien.tecmath.de>

In comp.lang.python, you wrote:
>On Sat, Apr 01, 2000 at 12:00:00PM -0500, Guido van Rossum wrote:
>> 
>> Python strings can now be stored as Unicode strings.  To make it easier
>> to type Unicode strings, the single-quote character defaults to creating
>> a Unicode string, while the double-quote character defaults to ASCII
>> strings.  If you need to create a Unicode string with double quotes,
>> just preface it with the letter "u"; likewise, an ASCII string can be
>> created by prefacing single quotes with the letter "a".  For example:
>> 
>>    foo = 'hello'                # Unicode
>>    foo = "hello"                # ASCII
>
>Is single-quoting for creating unicode clever ? I think there might be a problem
>with old code when the operations on unicode strings are not 100% compatible to
>the standard string operations. I don't know if this is a real problem - it's
>just a point for discussion.
>
>Cheers,
>Andreas 
>
Hallo Andreas,

hast Du mal auf das Datum des Beitrages von Guido geschaut? Echt guter April-
Scherz, da er die Scherze sehr gut mit der Realit?t mischt.

Liebe Gr??e, auch an die anderen,

Joachim




From robin at alldunn.com  Tue Apr 18 16:42:35 2000
From: robin at alldunn.com (Robin Dunn)
Date: Tue, 18 Apr 2000 13:42:35 -0700
Subject: Starship update: April 18
References: <3du2gzmiki.fsf@amarok.cnri.reston.va.us>
Message-ID: <AO3L4.63$Lwf1.5898538@news.randori.com>

>    * If you had downloadable files on Starship (such as software)
>      which you no longer have copies of, you might want to post to
>      comp.lang.python or the appropriate mailing list and ask if
>      anyone grabbed a copy recently.
>

The only thing of value that I have (potentially) lost is the archives of
the wxPython-users mailman list.  Did anybody happen to have downloaded
them?


--
Robin Dunn
Software Craftsman
robin at AllDunn.com
http://AllDunn.com/robin/
http://AllDunn.com/wxPython/  Check it out!





From dan at cgsoftware.com  Wed Apr 12 09:16:00 2000
From: dan at cgsoftware.com (Daniel Berlin+list.python)
Date: 12 Apr 2000 09:16:00 -0400
Subject: Python/Perl Popularity (Re: A Mountain of Perl...)
In-Reply-To: tom__98@my-deja.com's message of "Wed, 12 Apr 2000 06:15:34 GMT"
References: <200004102111.HAA03380@envy.fulcrum.com.au> <8d0pcq$7ti$1@nnrp1.deja.com> <g0ssm03o.fsf@dan.resnet.rochester.edu> <8d149t$jo6$1@nnrp1.deja.com>
Message-ID: <vh1nlasv.fsf@dan.resnet.rochester.edu>

>>>>> "tom" == tom 98 <tom__98 at my-deja.com> writes:

    tom> -- How can I browse the Python documentation without firing
    tom> up a web browser?
    >>  Convert it to text format. It's latex, you can make it into
    >> anything you want, and do anything you want with it.

    tom> Well, what I'm suggesting is that it might be a good idea to
    tom> make a command-line accessible version of the documentation,
    tom> together with a command line tool, available as part of the
    tom> standard distribution.
Sure, i'm all for it, and i'm sure it's being worked on.


    tom> -- Is there a Python equivalent of CPAN and the Perl CPAN
    tom> module?  The closest I know of is Parnassus.  But Parnassus
    tom> is merely a collection of links, not an archive, and it
    tom> doesn't have any facilities (AFAIK) for automatic
    tom> installation.
    >>  So, who gives a crap?  I try not to be that lazy.

    tom> So, you are saying that you basically agree: compiling and
    tom> installing modules is more work on Python, but that really
    tom> shouldn't matter because expecting to install modules quickly
    tom> is just laziness.  Well, I admit it, I'm lazy.
You'll have to excuse me, i have a problem with the CPAN module.
Namely, i have yet to install a single PERL module, and it claims
there are 203 installed modules with no parseable version number.
So there are 203 installed modules, which i didn't install, that it
can't tell me anything about.
That's not very helpful.

Whenever i try to upgrade any of my installed modules, they install
fine, yet the CPAN module never notices. I have to manually go and do
some cleanup before it notices the new versions.
It's more of an annoyance to me than anything else.


    tom> perldoc and the Perl CPAN module are really nice and well
    tom> done,
perldoc yes.
    tom> and as a user, I miss that functionality when I use
    tom> Python.
    tom>  Why not learn from Perl where they did something
    tom> good?
Because i don't feel they did it all that well.
    tom> This antagonism between the two camps, expressed in
    tom> your and other posts, just seems really silly to me as a user
    tom> of both languages.

You'll have to pardon me. I have PERL angst from trying to write an
HMM based Part of Speech tagger in PERL.


    tom> Tom.

    tom> Sent via Deja.com http://www.deja.com/ Before you buy.  --
    tom> http://www.python.org/mailman/listinfo/python-list



From laird at tsunami.ecn.purdue.edu  Sat Apr 22 09:04:47 2000
From: laird at tsunami.ecn.purdue.edu (Kyler Laird)
Date: 22 Apr 2000 13:04:47 GMT
Subject: Using python on the web
References: <MPG.1365adcfa8470eef9896aa@news> <MPG.136631472f892a119896ac@news> <_W2L4.986$GK5.20898@news2-win.server.ntlworld.com> <MPG.13665ef93eedbda29896ad@news> <BC4B61746187C623.3FA0E7D58DF4EA7E.ECC1D97F577DEDE5@lp.airnews.net> <slrn8g28ah.cc.amphgobb@eccie.blackflag.net>
Message-ID: <8ds81f$qne$1@mozo.cc.purdue.edu>

amphgobb at bjord.org (amphgobb) writes:

>I am not aware of many sysadmins who allow universal CGI access without
>mod_perl, and a cgi-wrapper. 

Hmmm...cgi-wrapper doesn't work with mod_perl (or
mod_php, or ...), right?

(This is why I built my own system to do the job.)

--kyler


From akira.kiyomiya at autodesk.com  Wed Apr 26 14:44:45 2000
From: akira.kiyomiya at autodesk.com (Akira Kiyomiya)
Date: Wed, 26 Apr 2000 11:44:45 -0700
Subject: Simple Function Question
Message-ID: <8e7dh3$1gc@autodesk.autodesk.com>

Hi,  Could you explain this in dummy's language?

This function prints out '4' and I have a feeling that x=1 and n=3, so
add1(3) prints '4'.

However, I have a difficlut time trying to see how these two arguments
assigned to both x (especially x) and n.

Thanks

Akira


# Return a function that returns its argument incremented by 'n'
def make_incrementer(n):
    def increment(x, n=n):
        return x+n
    return increment

add1 = make_incrementer(1)
print add1(3)  # This prints '4'







From claird at starbase.neosoft.com  Fri Apr 21 09:48:03 2000
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 21 Apr 2000 08:48:03 -0500
Subject: Good style for multi-valued returns
References: <889A6EB6AE399393.619B0539BFCAAE0F.5C6D110C3DA908F7@lp.airnews.net> <m3hfcwh6rd.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <43E1059BD6C613C1.19161A8DC1C439F1.3E710DAA95C4F8D8@lp.airnews.net>

In article <m3hfcwh6rd.fsf at atrus.jesus.cam.ac.uk>,
Michael Hudson  <mwh21 at cam.ac.uk> wrote:
>claird at starbase.neosoft.com (Cameron Laird) writes:
>
>> Let me guess:  folks generally pass tuples for small and
>> determinate-in-number multi-values, and dictionaries other-
>> wise, 'cept some do it all with tuples.  Accurate?
>
>I've always used tuples; I wouldn't have thought of something that
>returns a dictionary as return multiple values.
>
>Any reason for asking?
			.
			.
			.
"Reason"?  You want a *reason*?  Well, sure, there
are plenty of those:  it's time for me to be more
of a producer than consumer (of methods); I've been
thinking about decomposition; I didn't notice any-
thing in Guido's style guide; I'm eager to learn
more; ...

Particular thanks to the three people who've already
gently written to remind me that what I really want
in the complex-and-big case is to return an *object*
(but why do you all call it a class?  To my ear,
it truly sounds like a confusion to say, "class",
rather than, "instance of a class").  I think I still
have special cases where I want to move dictionaries
around, but I'm learning that their proper domain is
much smaller than I used to believe.
-- 

Cameron Laird <claird at NeoSoft.com>
Business:  http://www.Phaseit.net
Personal:  http://starbase.neosoft.com/~claird/home.html


From chapman at bioreason.com  Mon Apr 17 15:21:40 2000
From: chapman at bioreason.com (Mitch Chapman)
Date: Mon, 17 Apr 2000 13:21:40 -0600
Subject: Scripting and Gnome and KDE
References: <Pine.GSO.4.10.10004150741520.11261-100000@sundial> <8d95m0$9j6$1@news1.xs4all.nl> <8damat$4ja$1@slb2.atl.mindspring.net> <8dbsgc$35o$1@news1.xs4all.nl> <38FB4336.2BC6C235@bioreason.com> <87bt38h8wq.fsf@parus.parus.no>
Message-ID: <38FB6444.939CE6D6@bioreason.com>

Jon K Hellan wrote:
> You should definitely have mentioned libglade, which in most cases
> makes it unnecessary to code the Gtk GUI at all. You still have to do
> the callbacks, of course.

Yep, for static UIs (ones in which user interface elements
are not added dynamically at runtime), libglade saves all kinds of
effort.  We had not used it at the time the article was written
(September 1999), which is one reason why it was not mentioned.

-- 
Mitch Chapman
chapman at bioreason.com

Return-Path: <phord at mailgizmo.com>
Delivered-To: python-list at python.org
Received: from mailgizmo.com (105.164.200.216.fastpoint.net [216.200.164.105])
	by dinsdale.python.org (Postfix) with ESMTP id 995E51CE50
	for <python-list at python.org>; Mon, 17 Apr 2000 15:03:55 -0400 (EDT)
Received: (from phord at localhost)
	by mailgizmo.com (8.9.3/8.9.3) id PAA26768;
	Mon, 17 Apr 2000 15:00:59 -0400
From: nathan_christensen at peoplesoft.com
X-Envelope-To: <funny at phord.com>
X-Server-Uuid: b85f21a3-cfd1-11d3-8401-00104bf46ab7
Cc: AMitchell at origin.ea.com, bryanandlacey at earthlink.net,
	carma at carmapro.com, cujorover at earthlink.net, Disalphafu at aol.com,
	eovelman at origin.ea.com, falconson at nettally.com,
	fischer at tsgc.utexas.edu, funny at phord.com, goldie.mcelroy at attws.com,
	joe.maloukis at mail.va.gov, jtaylor at gfbank.com, jttaylor at flash.net,
	karyn at nightfire.com, kevin at aestechnology.com,
	Kizzy_Williams at yahoo.com, kmagill at prysm.net, knaack at gwatt.com,
	lachtaylor at yahoo.com, lbyrne at creativelabs.com, lindt at mail.utexas.edu,
	lmwwelsh at yahoo.com, lrosens at home.com, mar4 at texas.net,
	matsaleh at texas.net, matw at player1.com, mbakonyi at pencom.com,
	mbutler at esginc.com, mcginnis at texas.net, melanie at vsearch.com,
	mhornbuc at etec.com, mikeb at pencom.com, mike.cato at amd.com,
	mjerman at itrsjobs.com, mjgreathouse at juno.com, mkasper at mshow.com,
	monty_hornbuckle at etec.com, mrcambron at earthlink.net, MThor at aol.com,
	pbuchorn at gator1.brazosport.cc.tx.us, postmaster at novacell.com,
	python-list at python.org, ra6200 at email.sps.mot.com, rayjunem at juno.com,
	Ray_Grosvenor at Dell.com, rbarton at bbs.wiztower.com, rds at attymail.com,
	rgrosvenor at austin.rr.com, rong at cmsopen.com, rstidolph at austin.rr.com,
	san2edb at ups.com, scotts at illuminactive.com, seans at cmsopen.com,
	sitnalta at cableone.net, skip at Aureal.com, Srhughes30 at aol.com,
	stidolph at origin.ea.com, stone at golden-gate.org, tandmel at juno.com,
	Tanea_Greenwood at amat.com, tech_support at Quantum3D.com,
	terri.calloway at stanford.edu, tgif at iconn.net, Tiggerree at aol.com,
	tjhuff at msn.com, tlujan1 at hotmail.com, Tonya.Ochiltree at ci.austin.tx.us,
	tracie at austin.rr.com, tsmith at cs.utexas.edu, tt at peakware.com,
	Veronicahu at aol.com, vocooper at juno.com, WanChef at aol.com,
	wayne.stidolph at lutris.com, y2krdi at aol.com
X-Mailer: Lotus Notes Release 5.0.1 July 16, 1999
Message-ID: <OF25503401.CB160CEC-ON882568C4.00680F11 at peoplesoft.com>
Date: Mon, 17 Apr 2000 11:56:55 -0700
X-MIMETrack: Serialize by Router on SMTPOUT/Server/PSExternal(Release
 5.0.2b |December 16, 1999) at 04/17/2000 12:05:20 PM
MIME-Version: 1.0
X-WSS-ID: 14E580871000386-01-01
Content-Type: text/plain; 
 charset=iso-8859-1
Content-Transfer-Encoding: 8bit
X-MIME-Autoconverted: from quoted-printable to 8bit by mailgizmo.com id OAA26751
X-Loop: funny at phord.com
To: funny at phord.com
Subject: funny:  Re: May 27th PARTY!
Sender: python-list-admin at python.org
Errors-To: python-list-admin at python.org
X-BeenThere: python-list at python.org
X-Mailman-Version: 2.0beta3
Precedence: bulk
List-Id: General discussion list for the Python programming language <python-list.python.org>


Sounds like a blast...What should I bring?

Nate


                                                                                                                     
                    "Rebecca                                                                                         
                    Stidolph"              To:     "Wayne Stidolph" <wayne.stidolph at lutris.com>, <WanChef at aol.com>,  
                    <rstidolph at aust        "Viele Humor" <funny at phord.com>, "Veronica Hughes" <Veronicahu at aol.com>,  
                    in.rr.com>             "Val Cooper" <vocooper at juno.com>, "Trevor Lind" <lindt at mail.utexas.edu>,  
                                           "Tracie Thor" <tracie at austin.rr.com>, <Tonya.Ochiltree at ci.austin.tx.us>,  
                    04/15/00 08:47         "Tom Taylor" <jttaylor at flash.net>, "Todd Hughes" <Srhughes30 at aol.com>,    
                    PM                     "Tim Huffman" <tjhuff at msn.com>, "Tim and Melody Martin"                   
                                           <tandmel at juno.com>, "Tim and Cindy Martin" <mar4 at texas.net>, "Tim"        
                                           <tgif at iconn.net>, <Tiggerree at aol.com>, "Thad Smith"                       
                                           <tsmith at cs.utexas.edu>, "Terry Thompson" <tt at peakware.com>, "Terri        
                                           Calloway" <terri.calloway at stanford.edu>, "tech_support"                   
                                           <tech_support at Quantum3D.com>, "Taylor, Jacki" <jtaylor at gfbank.com>,       
                                           "Tanya Lujan" <tlujan1 at hotmail.com>, <Tanea_Greenwood at amat.com>, "Stone   
                                           Eleazer" <stone at golden-gate.org>, "Stidolph, David"                       
                                           <stidolph at origin.ea.com>, "Steven Jumper" <sitnalta at cableone.net>, "Stan  
                                           Edwards" <cujorover at earthlink.net>, "Skip McIlvaine" <skip at Aureal.com>,   
                                           "Sheridan Layman" <falconson at nettally.com>, "Sean Stanley"                
                                           <seans at cmsopen.com>, "Scott Schimpf" <scotts at illuminactive.com>,          
                                           <san2edb at ups.com>, "Ron Gilbert" <rong at cmsopen.com>, "Roja, Elena"        
                                           <ra6200 at email.sps.mot.com>, "Robert Knaack" <knaack at gwatt.com>, "Richard  
                                           Barton" <rbarton at bbs.wiztower.com>, <rgrosvenor at austin.rr.com>, "Rebecca  
                                           Stidolph" <rstidolph at austin.rr.com>, <rds at attymail.com>, "Ray O.          
                                           Grosvenor" <Ray_Grosvenor at Dell.com>, "Ray and June Minyard"               
                                           <rayjunem at juno.com>, "Python-List @ python. org"                          
                                           <python-list at python.org>, "Python Mailing List" <python-list at cwi.nl>,     
                                           "postmaster" <postmaster at novacell.com>, "'Peggy Buchorn'"                 
                                           <pbuchorn at gator1.brazosport.cc.tx.us>, "Ovelman, Elizabeth"               
                                           <eovelman at origin.ea.com>, "Nathan Christensen"                            
                                           <Nathan_Christensen at peoplesoft.com>, "Monty Hornbuckle"                   
                                           <monty_hornbuckle at etec.com>, "Monty Hornbuckle" <mhornbuc at etec.com>,      
                                           "Mitchell, Andy" <AMitchell at origin.ea.com>, "Mike Jerman"                 
                                           <mjerman at itrsjobs.com>, "Mike Greathouse" <mjgreathouse at juno.com>, "Mike  
                                           Cato" <mike.cato at amd.com>, "Michael Bakonyi" <mikeb at pencom.com>, "Michael 
                                           & Treasa McGinnis" <mcginnis at texas.net>, "Melinda Kasper"                 
                                           <mkasper at mshow.com>, "Melanie R. Cambron" <melanie at vsearch.com>, "Melanie 
                                           Cambron" <mrcambron at earthlink.net>, "Mcelroy, Goldie"                     
                                           <goldie.mcelroy at attws.com>, <mbutler at esginc.com>, <mbakonyi at pencom.com>,  
                                           "Matt Weathers" <matw at player1.com>, "Matt Walker" <matsaleh at texas.net>,   
                                           "Mark Thor" <MThor at aol.com>, "Mark Fischer" <fischer at tsgc.utexas.edu>,    
                                           "Mark Fischer" <fischer at csr.utexas.edu>, "Margaret Warren"                
                                           <carma at carmapro.com>, "Maloukis, Joe" <joe.maloukis at mail.va.gov>, "Leon   
                                           Rosenhein" <lrosens at home.com>, "Lbyrne" <lbyrne at creativelabs.com>,        
                                           "Laurie Welsh" <lmwwelsh at yahoo.com>, "Lach Taylor"                        
                                           <lachtaylor at yahoo.com>, "Kurt Schmidt" <y2krdi at aol.com>,                  
                                           "Kizzy_Williams at amat.com" <Kizzy_Williams at yahoo.com>, "Kevin Cooper"      
                                           <kevin at aestechnology.com>, "Kelsi" <Disalphafu at aol.com>, "Kelley Magill"  
                                           <kmagill at prysm.net>, "Karyn Christensen" <karyn at nightfire.com>, "A1C      
                                           Bryan A. and Mrs. Lacey D. Buchorn" <bryanandlacey at earthlink.net>         
                                           cc:                                                                       
                                           Subject:     May 27th PARTY!                                              
                                                                                                                     






ATTENTION FRIENDS AND FAMILY AROUND THE  NATION!!

We (David and Rebecca Stidolph) are celebrating our  17th wedding
anniversary with a big BBQ bash on Saturday, May 27th!? We  want everyone
to come, so we're giving plenty of notice.?  Please set aside the late
afternoon and evening to have lots of fun!

Please reply if you would like to come, so we can  include you in future
e-mails for more information.

Love,
David and Rebecca, Kelley and Kenneth

PS Sorry for any duplicate e-mails.? Let us  know if you got more than two.
Thanks!






Return-Path: <rstidolph at austin.rr.com>
Delivered-To: python-list at python.org
Received: from mail.austin.rr.com (sm1.texas.rr.com [24.93.35.54])
	by dinsdale.python.org (Postfix) with ESMTP id 3E0881CDB8
	for <python-list at python.org>; Sat, 15 Apr 2000 23:51:23 -0400 (EDT)
Received: from play1 ([24.160.3.64]) by mail.austin.rr.com  with Microsoft SMTPSVC(5.5.1877.357.35);
	 Sat, 15 Apr 2000 22:48:00 -0500
Message-ID: <006201bfa756$896e5a80$da00a8c0 at austin.rr.com>
From: "Rebecca Stidolph" <rstidolph at austin.rr.com>
To: "Wayne Stidolph" <wayne.stidolph at lutris.com>, <WanChef at aol.com>,
	"Viele Humor" <funny at phord.com>,
	"Veronica Hughes" <Veronicahu at aol.com>,
	"Val Cooper" <vocooper at juno.com>,
	"Trevor Lind" <lindt at mail.utexas.edu>,
	"Tracie Thor" <tracie at austin.rr.com>,
	<Tonya.Ochiltree at ci.austin.tx.us>, "Tom Taylor" <jttaylor at flash.net>,
	"Todd Hughes" <Srhughes30 at aol.com>, "Tim Huffman" <tjhuff at msn.com>,
	"Tim and Melody Martin" <tandmel at juno.com>,
	"Tim and Cindy Martin" <mar4 at texas.net>, "Tim" <tgif at iconn.net>,
	<Tiggerree at aol.com>, "Thad Smith" <tsmith at cs.utexas.edu>,
	"Terry Thompson" <tt at peakware.com>,
	"Terri Calloway" <terri.calloway at stanford.edu>,
	"tech_support" <tech_support at Quantum3D.com>,
	"Taylor, Jacki" <jtaylor at gfbank.com>,
	"Tanya Lujan" <tlujan1 at hotmail.com>, <Tanea_Greenwood at amat.com>,
	"Stone Eleazer" <stone at golden-gate.org>,
	"Stidolph, David" <stidolph at origin.ea.com>,
	"Steven Jumper" <sitnalta at cableone.net>,
	"Stan Edwards" <cujorover at earthlink.net>,
	"Skip McIlvaine" <skip at Aureal.com>,
	"Sheridan Layman" <falconson at nettally.com>,
	"Sean Stanley" <seans at cmsopen.com>,
	"Scott Schimpf" <scotts at illuminactive.com>, <san2edb at ups.com>,
	"Ron Gilbert" <rong at cmsopen.com>,
	"Roja, Elena" <ra6200 at email.sps.mot.com>,
	"Robert Knaack" <knaack at gwatt.com>,
	"Richard Barton" <rbarton at bbs.wiztower.com>,
	<rgrosvenor at austin.rr.com>,
	"Rebecca Stidolph" <rstidolph at austin.rr.com>, <rds at attymail.com>,
	"Ray O. Grosvenor" <Ray_Grosvenor at Dell.com>,
	"Ray and June Minyard" <rayjunem at juno.com>,
	"Python-List @ python. org" <python-list at python.org>,
	"Python Mailing List" <python-list at cwi.nl>,
	"postmaster" <postmaster at novacell.com>,
	"'Peggy Buchorn'" <pbuchorn at gator1.brazosport.cc.tx.us>,
	"Ovelman, Elizabeth" <eovelman at origin.ea.com>,
	"Nathan Christensen" <Nathan_Christensen at peoplesoft.com>,
	"Monty Hornbuckle" <monty_hornbuckle at etec.com>,
	"Monty Hornbuckle" <mhornbuc at etec.com>,
	"Mitchell, Andy" <AMitchell at origin.ea.com>,
	"Mike Jerman" <mjerman at itrsjobs.com>,
	"Mike Greathouse" <mjgreathouse at juno.com>,
	"Mike Cato" <mike.cato at amd.com>,
	"Michael Bakonyi" <mikeb at pencom.com>,
	"Michael & Treasa McGinnis" <mcginnis at texas.net>,
	"Melinda Kasper" <mkasper at mshow.com>,
	"Melanie R. Cambron" <melanie at vsearch.com>,
	"Melanie Cambron" <mrcambron at earthlink.net>,
	"Mcelroy, Goldie" <goldie.mcelroy at attws.com>, <mbutler at esginc.com>,
	<mbakonyi at pencom.com>, "Matt Weathers" <matw at player1.com>,
	"Matt Walker" <matsaleh at texas.net>, "Mark Thor" <MThor at aol.com>,
	"Mark Fischer" <fischer at tsgc.utexas.edu>,
	"Mark Fischer" <fischer at csr.utexas.edu>,
	"Margaret Warren" <carma at carmapro.com>,
	"Maloukis, Joe" <joe.maloukis at mail.va.gov>,
	"Leon Rosenhein" <lrosens at home.com>,
	"Lbyrne" <lbyrne at creativelabs.com>,
	"Laurie Welsh" <lmwwelsh at yahoo.com>,
	"Lach Taylor" <lachtaylor at yahoo.com>,
	"Kurt Schmidt" <y2krdi at aol.com>,
	"Kizzy_Williams at amat.com" <Kizzy_Williams at yahoo.com>,
	"Kevin Cooper" <kevin at aestechnology.com>,
	"Kelsi" <Disalphafu at aol.com>, "Kelley Magill" <kmagill at prysm.net>,
	"Karyn Christensen" <karyn at nightfire.com>,
	"A1C Bryan A. and Mrs. Lacey D. Buchorn" <bryanandlacey at earthlink.net>
Subject: May 27th PARTY!
Date: Sat, 15 Apr 2000 22:47:16 -0500
MIME-Version: 1.0
Content-Type: multipart/alternative;
	boundary="----=_NextPart_000_005C_01BFA72C.8C651E60"
X-Priority: 3
X-MSMail-Priority: Normal
X-Mailer: Microsoft Outlook Express 5.00.2919.6600
X-MimeOLE: Produced By Microsoft MimeOLE V5.00.2919.6600
Sender: python-list-admin at python.org
Errors-To: python-list-admin at python.org
X-BeenThere: python-list at python.org
X-Mailman-Version: 2.0beta3
Precedence: bulk
List-Id: General discussion list for the Python programming language <python-list.python.org>

This is a multi-part message in MIME format.

------=_NextPart_000_005C_01BFA72C.8C651E60
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable

ATTENTION FRIENDS AND FAMILY AROUND THE NATION!!

We (David and Rebecca Stidolph) are celebrating our 17th wedding =
anniversary with a big BBQ bash on Saturday, May 27th!  We want everyone =
to come, so we're giving plenty of notice.  Please set aside the late =
afternoon and evening to have lots of fun! =20

Please reply if you would like to come, so we can include you in future =
e-mails for more information.

Love,=20
David and Rebecca, Kelley and Kenneth

PS Sorry for any duplicate e-mails.  Let us know if you got more than =
two.  Thanks!

------=_NextPart_000_005C_01BFA72C.8C651E60
Content-Type: text/html;
	charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML><HEAD>
<META content=3D"text/html; charset=3Diso-8859-1" =
http-equiv=3DContent-Type>
<META content=3D"MSHTML 5.00.3013.2600" name=3DGENERATOR>
<STYLE></STYLE>
</HEAD>
<BODY bgColor=3D#ffffff>
<DIV><FONT face=3DArial size=3D2>ATTENTION FRIENDS AND FAMILY AROUND THE =

NATION!!</FONT></DIV>
<DIV>&nbsp;</DIV>
<DIV><FONT face=3DArial size=3D2>We (David and Rebecca Stidolph) are =
celebrating our=20
17th wedding anniversary with a big BBQ bash on Saturday, May =
27th!&nbsp; We=20
want everyone to come, so we're giving <STRONG>plenty </STRONG>of =
notice.&nbsp;=20
Please set aside the late afternoon and evening to have lots of =
fun!&nbsp;=20
</FONT></DIV>
<DIV>&nbsp;</DIV>
<DIV><FONT face=3DArial size=3D2>Please reply if you would like to come, =
so we can=20
include you in future e-mails for more information.</FONT></DIV>
<DIV>&nbsp;</DIV>
<DIV><FONT face=3DArial size=3D2>Love, </FONT></DIV>
<DIV><FONT face=3DArial size=3D2>David and Rebecca, Kelley and =
Kenneth</FONT></DIV>
<DIV>&nbsp;</DIV>
<DIV><FONT face=3DArial size=3D2>PS Sorry for any duplicate =
e-mails.&nbsp; Let us=20
know if you got more than two.&nbsp; Thanks!</FONT></DIV></BODY></HTML>

------=_NextPart_000_005C_01BFA72C.8C651E60--



Return-Path: <dixon at channel21.com>
Delivered-To: python-list at python.org
Received: from mail.channel21.com (unknown [209.139.165.10])
	by dinsdale.python.org (Postfix) with SMTP
	id 1FFFB1CD0E; Fri, 14 Apr 2000 17:45:58 -0400 (EDT)
Received: by XENON with Internet Mail Service (5.5.2650.21)
	id <2PBTHPD9>; Fri, 14 Apr 2000 18:08:00 -0400
Message-ID: <81DD38F43381D111BA3E00A076A0998A45A00A at XENON>
From: Channel21 Python Team <python at channel21.com>
Reply-To: Channel21 Python Team <python at channel21.com>
To: "'python-list at python.org'" <python-list at python.org>,
	"'python-sig at python.org'" <python-sig at python.org>,
	"'thread-sig at python.org'" <thread-sig at python.org>
Subject: FW: problems w/ IIS - Microsoft says it's python...
Date: Fri, 14 Apr 2000 18:07:54 -0400
Importance: high
X-Priority: 1
MIME-Version: 1.0
X-Mailer: Internet Mail Service (5.5.2650.21)
Content-Type: multipart/mixed;
	boundary="----_=_NextPart_000_01BFA65D.E540DE90"
Sender: python-list-admin at python.org
Errors-To: python-list-admin at python.org
X-BeenThere: python-list at python.org
X-Mailman-Version: 2.0beta3
Precedence: bulk
List-Id: General discussion list for the Python programming language <python-list.python.org>

This message is in MIME format. Since your mail reader does not understand
this format, some or all of this message may not be legible.

------_=_NextPart_000_01BFA65D.E540DE90
Content-Type: text/plain

something I sent to mark, just FYI to you guys (maybe you have some
suggestions)
> ----------
> From: 	Jason Nadler
> Sent: 	Friday, April 14, 2000 6:02 PM
> To: 	'mhammond at skippinet.com.au'
> Subject: 	problems w/ python and IIS - Microsoft says it's python...
> Importance: 	High
> 
My IIS has been crashing more frequently lately w/ more traffic to the
machine,  so I opened up a ticket w/ the IIS guys at Microsoft and they gave
me some software (IIS Exception Monitor) which trapped the stuff loaded im
memory and it points to the python interpreter. They say the problem is
definitely w/ python and they can't (won't) fix it, or even address it, any
further. The IIS services are now crashing about once an hour... I attached
2 of the log files here and an excerpt.
>  <<IIS_173-153131-413-2000.dbl>>  <<IIS_332-224129-413-2000.dbl>> 
> You can open the attached files w/ notepad... (look for instance 39 in the
> 173 file, and instance 22 in the 332 file - they are actually the same...)
> 
> here is what instance 39 from the 173 file looks like:
> 
>  39  id: ad.5d   Suspend: 0 Teb 7ff93000 Unfrozen
> Stack frame count of 0x300 is too high.  Changing to 500.
> ChildEBP RetAddr  Args to Child
> 01bafd68 1e142143 0ae13b80 00000000 0aee1f60
> python15!PyThreadState_Clear+0x6d
> 01bafd78 1e14267f 0aee1f60 00000000 1e2a2401
> python15!PyInterpreterState_Clear+0x13
> 01bafd84 1e2a2401 01bafda4 1e2c4fd8 00000000 python15!Py_Finalize+0x3f
> 01bafdec 680c3b32 01bafe6c 00139248 77f050f0
> pythoncom15!PyCom_DLLReleaseRef+0x71
> 01bafe34 680c2f38 000dfa58 00000000 01bafec0 asp!CAppConfig__SetValue+0xc8
> 01bafe7c 680c78c0 000dfa58 000c4698 01bafec0 asp!CPerfMainBlock__Init+0x47
> 01bafecc 680e25c6 00000000 697e1898 00313ea0
> asp!CConnectionPoint__~CConnectionPoint+0x6
> 01bafee4 697e67cc 00000000 697e1898 0aceefc0
> asp!_IASPObjectContextCustomStubVtbl+0x95
> 01baff24 697e57ce 00148598 697f21de 77f0178b
> MTxEx!CActivity__CallWithNewContext+0xac
> 77f67610 4affc033 89257508 fff00c42 037d044a
> MTxEx!CQueuedRequest__~CQueuedRequest+0x5e
> 0424548b 000003e6 0006ef1c 0006ee60 00000008 0x4affc033
> 
> 
> 
> PLEASE HELP!!! I am having the worst day...
> 

------_=_NextPart_000_01BFA65D.E540DE90
Content-Type: application/octet-stream;
	name="IIS_173-153131-413-2000.dbl"
Content-Transfer-Encoding: quoted-printable
Content-Disposition: attachment;
	filename="IIS_173-153131-413-2000.dbl"

<EMSessionType>Auto</EMSessionType><EMStartTime>4/13/00 3:31:31 =
PM</EMStartTime><EMEndTime>4/13/00 10:39:41 =
PM</EMEndTime><EMException>Manual</EMException>

log file opened
0:044> !emdbg.tag IIS Exception Monitor 6.1.0.0

Loaded emdbg extension DLL
IIS Exception Monitor 6.1.0.0
0:044> !emdbg.tag =
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D

=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D
0:044> !emdbg.tag =3D=3D Your Web Server should be running now, DON'T =
press Ctl-C Now.

=3D=3D Your Web Server should be running now, DON'T press Ctl-C Now.
0:044> !emdbg.tag =
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D

=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D
0:044> !emdbg.tag =3D=3D Wait until the application stops serving pages =
and then press Ctl-C in this window

=3D=3D Wait until the application stops serving pages and then press =
Ctl-C in this window
0:044> !emdbg.tag =3D=3D If you press Ctl-C, you can use the log to =
look for 100% CPU utilization

=3D=3D If you press Ctl-C, you can use the log to look for 100% CPU =
utilization
0:044> !emdbg.tag =3D=3D however it may not be able to tell you which =
module is causing your instability

=3D=3D however it may not be able to tell you which module is causing =
your instability
0:044> !emdbg.tag =
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D

=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D
0:044> sxd ch

0:044> g

CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 0e030000 0e041000   C:\Perl\bin\perlis.dll
LDR: Automatic DLL Relocation in inetinfo.exe
LDR: Dll PerlCRT.dll base 78000000 relocated due to collision with =
C:\WINNT\system32\MSVCRT.dll
CDB ModLoad: 02140000 02188000   C:\WINNT\system32\PerlCRT.dll
LDR: Automatic DLL Relocation in inetinfo.exe
LDR: Dll PerlCore.dll base 10000000 relocated due to collision with =
C:\Program Files\WEBTREND\Plugins\IIS4\RevDNS.dll
CDB ModLoad: 025a0000 02627000   C:\Perl\bin\PerlCore.dll
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
LDR: Automatic DLL Relocation in inetinfo.exe
LDR: Dll Socket.dll base 10000000 relocated due to collision with =
C:\Program Files\WEBTREND\Plugins\IIS4\RevDNS.dll
CDB ModLoad: 02a70000 02a77000   C:\Perl\lib\auto\Socket\Socket.dll
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB: exception number 8000d001
CDB: exception number 8000d001
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB: exception number 8000d001
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 68e50000 68e56000   C:\WINNT\system32\WAMREGPS.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB: exception number 8000d001
CDB: exception number 8000d001
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB: exception number 8000d001
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB: exception number 8000d001
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB: exception number 8000d001
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
LDR: Automatic DLL Relocation in inetinfo.exe
LDR: Dll IO.dll base 10000000 relocated due to collision with =
C:\Program Files\WEBTREND\Plugins\IIS4\RevDNS.dll
CDB ModLoad: 07d10000 07d16000   C:\Perl\lib\auto\IO\IO.dll
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB: exception number 8000d001
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB: exception number 8000d001
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB: exception number 8000d001
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB: exception number 8000d001
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB: exception number 8000d001
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB: exception number 8000d001
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB: exception number 8000d001
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB: exception number 8000d001
CDB: exception number 800706be
CDB: exception number 800706ba
CDB: exception number 800706ba
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB: exception number 800706ba
CDB: exception number 800706ba
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB: exception number 8000d001
CDB: exception number 8000d001
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB: access violation
eax=3D00000002 ebx=3D680ec470 ecx=3Dffffffff edx=3D7803c120 =
esi=3D0ae13b80 edi=3D00000000
eip=3D1e1422ed esp=3D01bafd64 ebp=3D01bafdec iopl=3D0         nv up ei =
pl nz na pe nc
cs=3D001b  ss=3D0023  ds=3D0023  es=3D0023  fs=3D003b  gs=3D0000        =
     efl=3D00010202
*** WARNING: Using export symbols for python15.dll
python15!PyThreadState_Clear+0x6d:
1e1422ed 8b08             mov     ecx,[eax]         =
ds:0023:00000002=3D????????
0:039> kv

*** WARNING: Using export symbols for pythoncom15.dll
*** WARNING: symbols checksum and timestamp is wrong 0x0005847a =
0x000512f3 for asp.dll
*** WARNING: symbols checksum and timestamp is wrong 0x000784a0 =
0x00078fb6 for mtxex.dll
ChildEBP RetAddr  Args to Child
01bafd68 1e142143 0ae13b80 00000000 0aee1f60 =
python15!PyThreadState_Clear+0x6d (FPO: [1,0,2])
01bafd78 1e14267f 0aee1f60 00000000 1e2a2401 =
python15!PyInterpreterState_Clear+0x13 (FPO: [1,0,2])
01bafd84 1e2a2401 01bafda4 1e2c4fd8 00000000 python15!Py_Finalize+0x3f =
(FPO: [0,0,1])
01bafdec 680c3b32 01bafe6c 00139248 77f050f0 =
pythoncom15!PyCom_DLLReleaseRef+0x71(FPO: [Non-Fpo]
01bafe34 680c2f38 000dfa58 00000000 01bafec0 =
asp!CAppConfig__SetValue+0xc8(FPO: [Non-Fpo]
01bafe7c 680c78c0 000dfa58 000c4698 01bafec0 =
asp!CPerfMainBlock__Init+0x47(FPO: [Non-Fpo]
01bafecc 680e25c6 00000000 697e1898 00313ea0 =
asp!CConnectionPoint__~CConnectionPoint+0x6 (FPO: [0,0,2])
01bafee4 697e67cc 00000000 697e1898 0aceefc0 =
asp!_IASPObjectContextCustomStubVtbl+0x95(FPO: [Non-Fpo]
01baff24 697e57ce 00148598 697f21de 77f0178b =
MTxEx!CActivity__CallWithNewContext+0xac(FPO: [Non-Fpo]
77f67610 4affc033 89257508 fff00c42 037d044a =
MTxEx!CQueuedRequest__~CQueuedRequest+0x5e(FPO: [Non-Fpo]
0424548b 000003e6 0006ef54 0006ee98 00000008 0x4affc033 [Stdcall: 20]
0:039> !emdbg.sysinfo

<EMXML><SYSINFO>
	<ENVIRONMENT>
		<COMPUTER-NAME>OBERON</COMPUTER-NAME>
		<SYS-DIR>C:\WINNT\System32</SYS-DIR>
	</ENVIRONMENT>
	<OS>
		<OS-VERSION>4.0 (Build 1381) Service Pack 6</OS-VERSION>
	</OS>
	<PROCESSOR>
		<NUM-PROC>2</NUM-PROC>
		<PROC-ARCH>x86</PROC-ARCH>
		<PROC-LEVEL>Pentium Pro or PII</PROC-LEVEL>
	</PROCESSOR>
	<MEMORY>
		<PHYSICAL-MEMORY>1073135616 bytes</PHYSICAL-MEMORY>
		<PHYSICAL-MEMORY-FREE>920686592 bytes</PHYSICAL-MEMORY-FREE>
		<PAGE-FILE>1304231936 bytes</PAGE-FILE>
		<PAGE-FILE-FREE>1181814784 bytes</PAGE-FILE-FREE>
	</MEMORY>
	<DISK>
		<LOGICAL-DRIVE>
			<DRIVE-LETTER>A:\</DRIVE-LETTER>
			<DRIVE-TYPE>Removable</DRIVE-TYPE>
		</LOGICAL-DRIVE>
		<LOGICAL-DRIVE>
			<DRIVE-LETTER>C:\</DRIVE-LETTER>
			<DRIVE-TYPE>Fixed Drive</DRIVE-TYPE>
			<DRIVE-SIZE>4194860032 bytes</DRIVE-SIZE>
			<BYTES-USED>1352962560 bytes</BYTES-USED>
			<BYTES-FREE>2841897472 bytes</BYTES-FREE>
			<VOLUME-NAME></VOLUME-NAME>
			<SERIAL-NUMBER>4c8e8b33</SERIAL-NUMBER>
			<FILE-SYSTEM>NTFS</FILE-SYSTEM>
			<FILE-SYSTEM-FLAGS>
			FS_CASE_IS_PRESERVED
			FS_CASE_SENSITIVE
			FS_UNICODE_STORED_ON_DISK
			FS_PERSISTENT_ACLS
			FS_FILE_COMPRESSION
			</FILE-SYSTEM-FLAGS>
		</LOGICAL-DRIVE>
		<LOGICAL-DRIVE>
			<DRIVE-LETTER>D:\</DRIVE-LETTER>
			<DRIVE-TYPE>Fixed Drive</DRIVE-TYPE>
			<DRIVE-SIZE>36413280256 bytes</DRIVE-SIZE>
			<BYTES-USED>702312448 bytes</BYTES-USED>
			<BYTES-FREE>35710967808 bytes</BYTES-FREE>
			<VOLUME-NAME></VOLUME-NAME>
			<SERIAL-NUMBER>342d9cd8</SERIAL-NUMBER>
			<FILE-SYSTEM>NTFS</FILE-SYSTEM>
			<FILE-SYSTEM-FLAGS>
			FS_CASE_IS_PRESERVED
			FS_CASE_SENSITIVE
			FS_UNICODE_STORED_ON_DISK
			FS_PERSISTENT_ACLS
			FS_FILE_COMPRESSION
			</FILE-SYSTEM-FLAGS>
		</LOGICAL-DRIVE>
		<LOGICAL-DRIVE>
			<DRIVE-LETTER>E:\</DRIVE-LETTER>
			<DRIVE-TYPE>Fixed Drive</DRIVE-TYPE>
			<DRIVE-SIZE>36413280256 bytes</DRIVE-SIZE>
			<BYTES-USED>9957376 bytes</BYTES-USED>
			<BYTES-FREE>36403322880 bytes</BYTES-FREE>
			<VOLUME-NAME></VOLUME-NAME>
			<SERIAL-NUMBER>50397a26</SERIAL-NUMBER>
			<FILE-SYSTEM>NTFS</FILE-SYSTEM>
			<FILE-SYSTEM-FLAGS>
			FS_CASE_IS_PRESERVED
			FS_CASE_SENSITIVE
			FS_UNICODE_STORED_ON_DISK
			FS_PERSISTENT_ACLS
			FS_FILE_COMPRESSION
			</FILE-SYSTEM-FLAGS>
		</LOGICAL-DRIVE>
		<LOGICAL-DRIVE>
			<DRIVE-LETTER>F:\</DRIVE-LETTER>
			<DRIVE-TYPE>Removable</DRIVE-TYPE>
		</LOGICAL-DRIVE>
		<LOGICAL-DRIVE>
			<DRIVE-LETTER>G:\</DRIVE-LETTER>
			<DRIVE-TYPE>CDROM</DRIVE-TYPE>
		</LOGICAL-DRIVE>
		<LOGICAL-DRIVE>
			<DRIVE-LETTER>L:\</DRIVE-LETTER>
			<DRIVE-TYPE>Fixed Drive</DRIVE-TYPE>
			<DRIVE-SIZE>4902232064 bytes</DRIVE-SIZE>
			<BYTES-USED>1175777280 bytes</BYTES-USED>
			<BYTES-FREE>3726454784 bytes</BYTES-FREE>
			<VOLUME-NAME></VOLUME-NAME>
			<SERIAL-NUMBER>70239c30</SERIAL-NUMBER>
			<FILE-SYSTEM>NTFS</FILE-SYSTEM>
			<FILE-SYSTEM-FLAGS>
			FS_CASE_IS_PRESERVED
			FS_CASE_SENSITIVE
			FS_UNICODE_STORED_ON_DISK
			FS_PERSISTENT_ACLS
			FS_FILE_COMPRESSION
			</FILE-SYSTEM-FLAGS>
		</LOGICAL-DRIVE>
	</DISK>
</SYSINFO></EMXML>

0:039> !inetdbg.mod

Loaded inetdbg extension DLL
Start    End      Entry    Path
01000000 01005000 01001d80 C:\WINNT\System32\inetsrv\inetinfo.exe
77f60000 77fbe000 00000000 C:\WINNT\System32\ntdll.dll
78000000 78045000 7800256b C:\WINNT\system32\MSVCRT.dll
77f00000 77f5e000 77f01000 C:\WINNT\system32\KERNEL32.dll
77dc0000 77dff000 77dc1000 C:\WINNT\system32\ADVAPI32.dll
77e70000 77ec5000 77e781d7 C:\WINNT\system32\USER32.dll
77ed0000 77efc000 00000000 C:\WINNT\system32\GDI32.dll
77e10000 77e67000 77e186bf C:\WINNT\system32\RPCRT4.dll
77b20000 77bd7000 77b24d6e C:\WINNT\system32\ole32.dll
68e80000 68e86000 68e810bc C:\WINNT\System32\inetsrv\rpcref.dll
77bf0000 77bf7000 00000000 C:\WINNT\system32\rpcltc1.dll
683a0000 683a7000 683a15a8 C:\WINNT\System32\inetsrv\iisadmin.dll
68220000 6822c000 682269e0 C:\WINNT\System32\inetsrv\COADMIN.dll
70980000 70b28000 709868fb C:\WINNT\system32\SHELL32.dll
710d0000 710f3000 710d14ce C:\WINNT\system32\SHLWAPI.dll
77aa0000 77b15000 77aa1b4d C:\WINNT\system32\COMCTL32.dll
68020000 6802a000 68022a91 C:\WINNT\system32\ADMWPROX.dll
74880000 74894000 74881000 C:\WINNT\system32\userenv.dll
77800000 7783a000 77801000 C:\WINNT\system32\NETAPI32.dll
77840000 77849000 00000000 C:\WINNT\system32\NETRAP.dll
777e0000 777ed000 00000000 C:\WINNT\system32\SAMLIB.dll
68900000 6890d000 68906180 C:\WINNT\System32\inetsrv\nsepm.dll
68440000 68450000 68448800 C:\WINNT\system32\IISMAP.dll
77400000 77429000 7740382b C:\WINNT\system32\schannel.dll
5e380000 5e3a5000 5e3a0920 C:\WINNT\system32\MSOSS.dll
776d0000 776d8000 00000000 C:\WINNT\system32\WSOCK32.dll
776b0000 776c4000 776b36b5 C:\WINNT\system32\WS2_32.dll
776a0000 776a7000 776a110e C:\WINNT\system32\WS2HELP.dll
5cf00000 5cf5e000 5cf06b44 C:\WINNT\system32\CRYPT32.dll
68df0000 68e03000 68df2b1f C:\WINNT\System32\inetsrv\metadata.dll
68c90000 68c9b000 68c91f80 C:\WINNT\System32\inetsrv\wamreg.dll
5f400000 5f4f2000 5f4065ff C:\WINNT\system32\MFC42.DLL
65340000 653d2000 6534ee21 C:\WINNT\system32\OLEAUT32.dll
7ca00000 7ca1b000 7ca0b2cc C:\WINNT\system32\rsabase.dll
5a800000 5a80f000 5a802004 C:\WINNT\system32\pstorec.dll
68db0000 68db7000 68db2b10 C:\WINNT\System32\inetsrv\admexs.dll
68ea0000 68ea7000 68ea2350 C:\WINNT\System32\inetsrv\svcext.dll
68330000 68345000 68336930 C:\WINNT\System32\inetsrv\ftpsvc2.dll
68590000 685bf000 68594a93 C:\WINNT\System32\inetsrv\INFOCOMM.dll
68620000 6862e000 68625f71 C:\WINNT\System32\inetsrv\ISATQ.dll
682b0000 682b6000 682b17b4 C:\WINNT\System32\inetsrv\IISFECNV.dll
678a0000 67923000 678ef3b0 C:\WINNT\System32\inetsrv\SMTPSVC.dll
77a90000 77a9b000 77a92fd0 C:\WINNT\system32\VERSION.dll
779c0000 779c8000 779c1881 C:\WINNT\system32\LZ32.dll
679a0000 679aa000 679a1230 C:\WINNT\system32\exstrace.dll
67780000 67786000 677811c0 C:\WINNT\system32\RWNH.dll
68c00000 68c39000 68c1c344 C:\WINNT\System32\inetsrv\w3svc.dll
68560000 68568000 68562f47 C:\WINNT\system32\inetsloc.dll
77670000 77685000 77672599 C:\WINNT\system32\MSWSOCK.dll
76e70000 76e82000 76e71000 C:\WINNT\system32\security.dll
68820000 68827000 6882132a C:\WINNT\System32\inetsrv\lonsint.dll
47600000 4760e000 47608a10 C:\WINNT\system32\wintrust.dll
74fa0000 74fab000 00000000 C:\WINNT\system32\RpcLtScm.Dll
77660000 7766f000 77663bc8 C:\WINNT\system32\msafd.dll
77690000 77699000 77691000 C:\WINNT\System32\wshtcpip.dll
74ff0000 74ffe000 74ff7b68 C:\WINNT\System32\rnr20.dll
60000000 60011000 60006440 C:\WINNT\system32\msapsspc.dll
779d0000 779e5000 779d1000 C:\WINNT\system32\MSVCRT40.dll
780a0000 780b2000 780a1000 C:\WINNT\system32\MSVCIRT.dll
011a0000 011bf000 011a52c0 C:\WINNT\system32\msnsspc.dll
77e00000 77e06000 00000000 C:\WINNT\system32\rpclts1.dll
74fc0000 74fcf000 00000000 C:\WINNT\system32\RpcLtCcm.Dll
68650000 68659000 686519e8 C:\WINNT\System32\inetsrv\iscomlog.dll
677b0000 67802000 677b6b30 C:\WINNT\System32\inetsrv\seo.dll
68a10000 68a19000 68a126d6 C:\WINNT\System32\inetsrv\sspifilt.dll
74480000 74488000 00000000 C:\WINNT\System32\wshisn.dll
67860000 67868000 678613e0 c:\program files\microsoft =
frontpage\bin\fpexedll.dll
75830000 7585a000 7583636b C:\WINNT\System32\nwprovau.dll
77720000 77731000 777278db C:\WINNT\system32\MPR.dll
67ec0000 67ec5000 67ec1460 C:\Program Files\Microsoft =
FrontPage\version3.0\bin\fpexedll.dll
74470000 74477000 74471000 C:\WINNT\system32\wshnetbs.dll
68410000 68421000 684160c8 C:\WINNT\System32\inetsrv\iislog.dll
10000000 10008000 100025f0 C:\Program =
Files\WEBTREND\Plugins\IIS4\RevDNS.dll
017d0000 017d8000 017d27a0 C:\Program =
Files\WEBTREND\Plugins\IIS4\wtCookie.dll
68890000 688a6000 6889244e C:\WINNT\System32\inetsrv\wam.dll
697e0000 6984c000 6981c900 C:\WINNT\System32\mtxex.dll
69d00000 69d1a000 69d10490 C:\WINNT\System32\MTSEvents.DLL
69a90000 69a9a000 69a94690 C:\WINNT\System32\MTxInfr1.dll
69ab0000 69abd000 69ab5cf0 C:\WINNT\System32\MTxInfr2.dll
69620000 69630000 696289a0 C:\WINNT\System32\MTxRn.DLL
6b550000 6b556000 6b551480 C:\WINNT\System32\MTxPerf.dll
69a70000 69a7a000 69a74220 C:\WINNT\System32\MfcSubs.dll
690b0000 690ce000 690c5a00 C:\WINNT\System32\DTCCM.dll
69000000 6900a000 69004180 C:\WINNT\System32\DTCUtil.dll
68ff0000 68ff7000 68ff1420 C:\WINNT\System32\DTCTRACE.dll
69790000 6979d000 69798460 C:\WINNT\System32\MTXCLU.DLL
699c0000 699cc000 699c4f00 C:\WINNT\System32\MTXTRK.DLL
69360000 69369000 69361d40 C:\WINNT\System32\XOLEHLP.dll
69120000 69133000 6912d300 C:\WINNT\System32\ADME.dll
69a20000 69a2b000 69a24a30 C:\WINNT\System32\MtxDm.dll
69640000 69671000 69656720 C:\WINNT\System32\MtxCat.dll
1f7d0000 1f804000 1f7d12d4 C:\WINNT\System32\ODBC32.dll
77d80000 77db2000 77d81000 C:\WINNT\system32\comdlg32.dll
7f230000 7f23d000 00000000 C:\WINNT\system32\CLUSAPI.DLL
7f250000 7f25a000 7f2516f8 C:\WINNT\system32\RESUTILS.DLL
1f8c0000 1f8d6000 00000000 C:\WINNT\System32\odbcint.dll
697c0000 697ca000 697c4d10 C:\WINNT\System32\Mts\mtxact.dll
69a40000 69a5b000 69a4f730 C:\WINNT\system32\mtxoci.dll
680b0000 680ff000 680e2c80 C:\WINNT\System32\inetsrv\asp.dll
6b600000 6b66b000 6b605359 C:\WINNT\System32\vbscript.dll
1e2a0000 1e2e2000 1e2c4d6d C:\WINNT\system32\pythoncom15.dll
1e600000 1e611000 1e606f5a C:\WINNT\system32\PyWinTypes15.dll
1e100000 1e186000 1e1583c0 C:\WINNT\system32\python15.dll
77fd0000 77ffa000 77fd5640 C:\WINNT\system32\WINMM.dll
1e200000 1e20e000 1e206fa3 C:\Program Files\Python\win32\win32api.pyd
01d30000 01d42000 01d37a18 C:\Program =
Files\Python\win32comext\axscript\axscript.pyd
1e3b0000 1e3e8000 1e3d1b6f C:\Program =
Files\Python\win32comext\axdebug\axdebug.pyd
1e360000 1e366000 1e361855 C:\Program Files\Python\win32\dbi.dll
1e350000 1e358000 1e3534c5 C:\Program Files\Python\win32\odbc.pyd
41230000 412ae000 41231000 C:\WINNT\System32\sqlsrv32.dll
41100000 4110c000 411061e0 C:\WINNT\System32\SQLWOA.dll
1f820000 1f83a000 1f82c63c C:\WINNT\system32\odbccp32.dll
68c80000 68c86000 68c81510 C:\WINNT\System32\inetsrv\wamps.dll
68690000 68696000 68691970 C:\WINNT\System32\inetsrv\iwrps.dll
0e030000 0e041000 0e039d30 C:\Perl\bin\perlis.dll
02140000 02188000 02149010 C:\WINNT\system32\PerlCRT.dll
025a0000 02627000 0260c040 C:\Perl\bin\PerlCore.dll
02a70000 02a77000 02a73c80 C:\Perl\lib\auto\Socket\Socket.dll
68e50000 68e56000 68e514e0 C:\WINNT\system32\WAMREGPS.DLL
07d10000 07d16000 07d12610 C:\Perl\lib\auto\IO\IO.dll
0:039> ~*kb 300


  0  id: ad.ac   Suspend: 0 Teb 7ffde000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
*** WARNING: Using export symbols for advapi32.dll
*** WARNING: Using export symbols for kernel32.dll
ChildEBP RetAddr  Args to Child
0006fdb0 77dd8a35 00000050 00073880 0000021a ntdll!ZwReadFile+0xb
0006fde4 77dd8481 00000050 00073880 0000021a =
ADVAPI32!RegisterServiceCtrlHandlerA+0x352
0006fe48 77dd829e 00000050 00073880 0000021a =
ADVAPI32!StartServiceCtrlDispatcherW+0x184
0006fe6c 0100182c 00073278 68e80000 77f13fb3 =
ADVAPI32!StartServiceCtrlDispatcherA+0x82
0006fea4 010015d5 00000000 00000000 7ffdf000 =
inetinfo!StartDispatchTable+0x61
0006ff70 01001e7f 00000001 00240550 002401a0 inetinfo!main+0x2a5
0006ffc0 77f1b9ea 00000000 00000000 7ffdf000 =
inetinfo!mainCRTStartup+0xff
0006fff0 00000000 01001d80 00000000 000000b0 =
KERNEL32!GetProcessPriorityBoost+0x117

  1  id: ad.ae   Suspend: 0 Teb 7ffdd000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
0087fe54 77f04f37 000000b4 ffffffff 00000000 =
ntdll!NtWaitForSingleObject+0xb
00000001 000003e6 00000088 77f04f37 0000000f =
KERNEL32!WaitForSingleObject+0xf

  2  id: ad.34   Suspend: 0 Teb 7ffdc000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
*** WARNING: Using export symbols for rpcrt4.dll
ChildEBP RetAddr  Args to Child
0d25ff90 77e15a1d 77e160f7 0007f878 0d25ffec =
ntdll!ZwReplyWaitReceivePort+0xb
00003a98 000003e6 00000088 77e15a1d 00000024 RPCRT4!NdrOleAllocate+0x24

  3  id: ad.6a   Suspend: 0 Teb 7ffdb000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
*** WARNING: Using export symbols for user32.dll
*** WARNING: symbols checksum and timestamp is wrong 0x0003b1c7 =
0x0003704a for infocomm.dll
ChildEBP RetAddr  Args to Child
00e4fd80 77e79de5 00000002 00e4fda8 00000000 =
ntdll!NtWaitForMultipleObjects+0xb
00e4fddc 77e79e31 00000001 0028030c ffffffff =
USER32!MsgWaitForMultipleObjectsEx+0xa8
00e4fdf8 68599007 00000001 0028030c 00000000 =
USER32!MsgWaitForMultipleObjects+0x1d
0028030c 00000000 00000000 00000000 68333581 =
INFOCOMM!INetpCreateSecurityDescriptor+0x1cf

  4  id: ad.af   Suspend: 0 Teb 7ffda000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
00e8fd8c 77e79de5 00000002 00e8fdb4 00000000 =
ntdll!NtWaitForMultipleObjects+0xb
00e8fde8 77e79e31 00000001 0023ed44 ffffffff =
USER32!MsgWaitForMultipleObjectsEx+0xa8
00e8fe04 68599007 00000001 0023ed44 00000000 =
USER32!MsgWaitForMultipleObjects+0x1d
0023ed44 00000000 00000000 00000000 678ae3e0 =
INFOCOMM!INetpCreateSecurityDescriptor+0x1cf

  5  id: ad.b0   Suspend: 0 Teb 7ffd9000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
00ecfd80 77e79de5 00000002 00ecfda8 00000000 =
ntdll!NtWaitForMultipleObjects+0xb
00ecfddc 77e79e31 00000001 0023e864 ffffffff =
USER32!MsgWaitForMultipleObjectsEx+0xa8
00ecfdf8 68599007 00000001 0023e864 00000000 =
USER32!MsgWaitForMultipleObjects+0x1d
0023e864 00000000 00000000 00000800 68c09c95 =
INFOCOMM!INetpCreateSecurityDescriptor+0x1cf

  6  id: ad.b1   Suspend: 0 Teb 7ffd8000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
*** WARNING: symbols checksum and timestamp is wrong 0x00013a0e =
0x00011cc1 for isatq.dll
ChildEBP RetAddr  Args to Child
00f0fef4 77e79de5 00000002 00f0ff1c 00000000 =
ntdll!NtWaitForMultipleObjects+0xb
00f0ff50 77e79e31 00000001 68629668 000010e9 =
USER32!MsgWaitForMultipleObjectsEx+0xa8
00f0ff6c 68625cce 00000001 68629668 00000000 =
USER32!MsgWaitForMultipleObjects+0x1d
00000001 000003e6 00000088 68622913 00000000 =
ISATQ!W95CPORT__PrepareDescriptorArrays+0x58

  7  id: ad.b2   Suspend: 0 Teb 7ffd7000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
00f4fef4 77e79de5 00000002 00f4ff1c 00000000 =
ntdll!NtWaitForMultipleObjects+0xb
00f4ff50 77e79e31 00000001 68629668 000010e9 =
USER32!MsgWaitForMultipleObjectsEx+0xa8
00f4ff6c 68625cce 00000001 68629668 00000000 =
USER32!MsgWaitForMultipleObjects+0x1d
00000001 000003e6 00000088 68622913 00000000 =
ISATQ!W95CPORT__PrepareDescriptorArrays+0x58

  8  id: ad.b3   Suspend: 0 Teb 7ffd6000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
00fcff84 686248e5 0000016c 00fcffac 00fcffb4 =
ntdll!ZwRemoveIoCompletion+0xb
00fcffb8 77f04ede abcdef01 00e4fc8c 00e4fcf0 ISATQ!AtqSetInfo+0xbb
00fcffec 00000000 686248a8 abcdef01 00000000 KERNEL32!lstrcmpiW+0xbe

  9  id: ad.b4   Suspend: 0 Teb 7ffd5000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
0104ff84 686248e5 0000016c 0104ffac 0104ffb4 =
ntdll!ZwRemoveIoCompletion+0xb
0104ffb8 77f04ede abcdef01 00e4fc8c 00e4fcf0 ISATQ!AtqSetInfo+0xbb
0104ffec 00000000 686248a8 abcdef01 00000000 KERNEL32!lstrcmpiW+0xbe

 10  id: ad.b5   Suspend: 0 Teb 7ffd4000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
0108ff88 685a75d5 00000002 0108ffb4 00000000 =
ntdll!NtWaitForMultipleObjects+0xb

 11  id: ad.b6   Suspend: 0 Teb 7ffaf000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
010cff90 77e15a1d 77e160f7 00084d18 010cffec =
ntdll!ZwReplyWaitReceivePort+0xb
00003a98 000003e6 00000088 77e15a1d 00000024 RPCRT4!NdrOleAllocate+0x24

 12  id: ad.b7   Suspend: 0 Teb 7ffae000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
*** WARNING: Using export symbols for msafd.dll
*** WARNING: Using export symbols for ws2_32.dll
*** WARNING: Using export symbols for rpcltscm.dll
ChildEBP RetAddr  Args to Child
0118fd78 77664a12 000001e0 000001e8 00000001 =
ntdll!NtWaitForSingleObject+0xb
0118fea0 776b9f5f 00000000 000854f0 00000000 MSAFD+0x4a12
0118fef0 74fa21e2 00000000 000854f0 00000000 WS2_32!select+0xfc
0118ff2c 77e1b0dc 00086c00 0118ff58 0118ff88 RPCLTSCM+0x21e2
0118ff60 77e1cb96 0118ff8c 0118ff88 0118ff84 =
RPCRT4!I_RpcTransServerUnprotectThread+0x48f
0118ff90 77e1aa95 77e160f7 00086b20 0118ffec =
RPCRT4!I_RpcTransServerProtectThread+0xcc
00003a98 000003e6 00000088 77e1aa95 000000c8 =
RPCRT4!I_RpcTransServerReallocBuffer+0xc8

 13  id: ad.b8   Suspend: 0 Teb 7ffad000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
011fff2c 77e1b0dc 00089280 011fff58 011fff88 =
ntdll!NtWaitForMultipleObjects+0xb
011fff60 77e1cb96 011fff8c 011fff88 011fff84 =
RPCRT4!I_RpcTransServerUnprotectThread+0x48f
011fff90 77e1aa95 77e160f7 000891a0 011fffec =
RPCRT4!I_RpcTransServerProtectThread+0xcc
00003a98 000003e6 00000088 77e1aa95 000000c8 =
RPCRT4!I_RpcTransServerReallocBuffer+0xc8

 14  id: ad.ba   Suspend: 0 Teb 7ffac000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
0123ff64 77f1cebe 0000ea60 00000000 7583a68a ntdll!NtDelayExecution+0xb
0123ffb8 77f04ede 00000000 00000001 00000000 KERNEL32!Sleep+0xb
0123ffec 00000000 7583a65e 00000000 00000000 KERNEL32!lstrcmpiW+0xbe

 15  id: ad.bb   Suspend: 0 Teb 7ffab000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
*** WARNING: symbols checksum and timestamp is wrong 0x0003e35a =
0x00039d63 for w3svc.dll
ChildEBP RetAddr  Args to Child
0127fec8 77e79de5 00000002 0127fef0 00000000 =
ntdll!NtWaitForMultipleObjects+0xb
0127ff24 77e79e31 00000001 0028186c ffffffff =
USER32!MsgWaitForMultipleObjectsEx+0xa8
0127ff40 68c243f1 00000001 0028186c 00000000 =
USER32!MsgWaitForMultipleObjects+0x1d

 16  id: ad.bc   Suspend: 0 Teb 7ffaa000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
012bff00 77e79de5 00000002 012bff28 00000000 =
ntdll!NtWaitForMultipleObjects+0xb
012bff5c 77e79e31 00000001 68c2c25c ffffffff =
USER32!MsgWaitForMultipleObjectsEx+0xa8
012bff78 68c1c55e 00000001 68c2c25c 00000000 =
USER32!MsgWaitForMultipleObjects+0x1d
012bffb8 77f04ede 00000000 00000003 00230000 =
W3SVC!HTTP_FILTER__ParseSendHeaders+0x46
012bffec 00000000 68c1c52a 00000000 00000000 KERNEL32!lstrcmpiW+0xbe

 17  id: ad.e1   Suspend: 0 Teb 7ffa9000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
*** WARNING: symbols checksum and timestamp is wrong 0x000077d7 =
0x0000908e for inetsloc.dll
ChildEBP RetAddr  Args to Child
012ffcf8 77664a12 00000340 00000240 00000001 =
ntdll!NtWaitForSingleObject+0xb
012ffe20 776b9f5f 00000000 012ffe98 00000000 MSAFD+0x4a12
012ffe70 68563c07 00000000 012ffe98 00000000 WS2_32!select+0xfc
012fffb8 77f04ede 00000000 00000000 00000000 =
INETSLOC!SocketListenThread+0x51
012fffec 00000000 68563bb6 00000000 00000000 KERNEL32!lstrcmpiW+0xbe

 18  id: ad.c7   Suspend: 0 Teb 7ffa8000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
0133ff90 77e15a1d 77e160f7 00094390 0133ffec =
ntdll!ZwReplyWaitReceivePort+0xb
00003a98 000003e6 00000088 77e15a1d 00000024 RPCRT4!NdrOleAllocate+0x24

 19  id: ad.cb   Suspend: 0 Teb 7ffa7000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
0147ff2c 77e1b0dc 00097d58 0147ff58 0147ff88 =
ntdll!NtWaitForMultipleObjects+0xb
0147ff60 77e1cb96 0147ff8c 0147ff88 0147ff84 =
RPCRT4!I_RpcTransServerUnprotectThread+0x48f
0147ff90 77e1aa95 77e160f7 00097c78 0147ffec =
RPCRT4!I_RpcTransServerProtectThread+0xcc
00003a98 000003e6 00000088 77e1aa95 000000c8 =
RPCRT4!I_RpcTransServerReallocBuffer+0xc8

 20  id: ad.c2   Suspend: 0 Teb 7ffa6000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
*** WARNING: symbols checksum and timestamp is wrong 0x00087d67 =
0x0007a2f5 for smtpsvc.dll
ChildEBP RetAddr  Args to Child
014cff68 77f1cd76 00000002 014cffac 00000000 =
ntdll!NtWaitForMultipleObjects+0xb
014cff84 678b9333 00000002 014cffac 00000000 =
KERNEL32!WaitForMultipleObjects+0x16

 21  id: ad.de   Suspend: 0 Teb 7ffa5000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
0150ff68 77f1cd76 00000002 0150ffac 00000000 =
ntdll!NtWaitForMultipleObjects+0xb
0150ff84 678b9333 00000002 0150ffac 00000000 =
KERNEL32!WaitForMultipleObjects+0x16

 22  id: ad.dc   Suspend: 0 Teb 7ffa4000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
0154ff68 77f1cd76 00000002 0154ffac 00000000 =
ntdll!NtWaitForMultipleObjects+0xb
0154ff84 678b9333 00000002 0154ffac 00000000 =
KERNEL32!WaitForMultipleObjects+0x16

 23  id: ad.e2   Suspend: 0 Teb 7ffa3000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
0158ff68 77f1cd76 00000002 0158ffac 00000000 =
ntdll!NtWaitForMultipleObjects+0xb
0158ff84 678b9333 00000002 0158ffac 00000000 =
KERNEL32!WaitForMultipleObjects+0x16

 24  id: ad.ef   Suspend: 0 Teb 7ffa2000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
015cff68 77f1cd76 00000002 015cffac 00000000 =
ntdll!NtWaitForMultipleObjects+0xb
015cff84 678b9333 00000002 015cffac 00000000 =
KERNEL32!WaitForMultipleObjects+0x16

 25  id: ad.f0   Suspend: 0 Teb 7ffa1000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
0160ff8c 77f04f37 00000484 ffffffff 00000000 =
ntdll!NtWaitForSingleObject+0xb
77f67610 4affc033 89257508 fff00c42 037d044a =
KERNEL32!WaitForSingleObject+0xf
0424548b 000003e6 0006ef1c 0006ee60 00000008 0x4affc033

 26  id: ad.f1   Suspend: 0 Teb 7ffa0000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
0164ff8c 77f04f37 00000490 ffffffff 00000000 =
ntdll!NtWaitForSingleObject+0xb
77f67610 4affc033 89257508 fff00c42 037d044a =
KERNEL32!WaitForSingleObject+0xf
0424548b 000003e6 0006ef1c 0006ee60 00000008 0x4affc033

 27  id: ad.f2   Suspend: 0 Teb 7ff9f000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
0168ff8c 77f04f37 0000049c ffffffff 00000000 =
ntdll!NtWaitForSingleObject+0xb
77f67610 4affc033 89257508 fff00c42 037d044a =
KERNEL32!WaitForSingleObject+0xf
0424548b 000003e6 0006ef1c 0006ee60 00000008 0x4affc033

 28  id: ad.f3   Suspend: 0 Teb 7ff9e000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
016cff94 77f04f37 000004a8 ffffffff 00000000 =
ntdll!NtWaitForSingleObject+0xb
016cffec 00000000 678d5c30 002834e0 00000000 =
KERNEL32!WaitForSingleObject+0xf

 29  id: ad.f5   Suspend: 0 Teb 7ff9c000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
0174ff8c 77f04f37 000004c8 ffffffff 00000000 =
ntdll!NtWaitForSingleObject+0xb
77f67610 4affc033 89257508 fff00c42 037d044a =
KERNEL32!WaitForSingleObject+0xf
0424548b 000003e6 0006ef1c 0006ee60 00000008 0x4affc033

 30  id: ad.f7   Suspend: 0 Teb 7ff9d000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
*** WARNING: Using export symbols for revdns.dll
ChildEBP RetAddr  Args to Child
017cfba4 10001a2b 017cfd70 00000000 00000000 USER32!InvalidateRect+0x19
017cffb8 77f04ede 00000000 00070860 00070000 =
REVDNS!CreateConfigTab+0x84e
017cffec 00000000 10001755 00000000 00000000 KERNEL32!lstrcmpiW+0xbe

 31  id: ad.f8   Suspend: 0 Teb 7ff9a000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
0181ff84 686248e5 0000016c 0181ffac 0181ffb4 =
ntdll!ZwRemoveIoCompletion+0xb
0181ffb8 77f04ede 00000000 001f0003 00000000 ISATQ!AtqSetInfo+0xbb
0181ffec 00000000 686248a8 00000000 00000000 KERNEL32!lstrcmpiW+0xbe

 32  id: ad.f9   Suspend: 0 Teb 7ff99000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
0185ff84 686248e5 0000016c 0185ffac 0185ffb4 =
ntdll!ZwRemoveIoCompletion+0xb
0185ffb8 77f04ede 00000000 001f0003 00000000 ISATQ!AtqSetInfo+0xbb
0185ffec 00000000 686248a8 00000000 00000000 KERNEL32!lstrcmpiW+0xbe

 33  id: ad.124   Suspend: 0 Teb 7ff9b000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
08c2feb4 77e79de5 00000003 08c2fedc 00000000 =
ntdll!NtWaitForMultipleObjects+0xb
08c2ff10 77e79e31 00000002 08c2ff58 000003e8 =
USER32!MsgWaitForMultipleObjectsEx+0xa8
08c2ff2c 697f1dbe 00000002 08c2ff58 00000000 =
USER32!MsgWaitForMultipleObjects+0x1d
69a92ea0 750169a9 24448b22 244c8b14 24548b10 =
MTxEx!ContextWrapper__Recycle+0xe
60b03d83 000003e6 0006ef1c 0006ee60 00000008 0x750169a9

 34  id: ad.fd   Suspend: 0 Teb 7ff98000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
01a1ff80 77f04f37 00000b20 00002710 00000000 =
ntdll!NtWaitForSingleObject+0xb
00000000 000003e6 00000088 77f04f37 0000000f =
KERNEL32!WaitForSingleObject+0xf

 35  id: ad.fe   Suspend: 0 Teb 7ff97000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
*** WARNING: Using export symbols for msvcrt.dll
ChildEBP RetAddr  Args to Child
01a8fec8 77e79de5 00000002 01a8fef0 00000000 =
ntdll!NtWaitForMultipleObjects+0xb
01a8ff24 77e79e31 00000001 680ebddc ffffffff =
USER32!MsgWaitForMultipleObjectsEx+0xa8
01a8ff40 680ca30c 00000001 680ebddc 00000000 =
USER32!MsgWaitForMultipleObjects+0x1d
01a8ff84 7800be50 680ebdc0 00000008 0000021c =
asp!CCookieSupportErr__`vftable'+0xb
01a8ffb8 77f04ede 00293f30 00000008 0000021c MSVCRT!endthread+0x96
01a8ffec 00000000 7800bdf9 00293f30 00000000 KERNEL32!lstrcmpiW+0xbe

 36  id: ad.31   Suspend: 0 Teb 7ff96000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
01aefeb4 77e79de5 00000003 01aefedc 00000000 =
ntdll!NtWaitForMultipleObjects+0xb
01aeff10 77e79e31 00000002 01aeff58 000003e8 =
USER32!MsgWaitForMultipleObjectsEx+0xa8
01aeff2c 697f1dbe 00000002 01aeff58 00000000 =
USER32!MsgWaitForMultipleObjects+0x1d
69a92ea0 750169a9 24448b22 244c8b14 24548b10 =
MTxEx!ContextWrapper__Recycle+0xe
60b03d83 000003e6 0006ef1c 0006ee60 00000008 0x750169a9

 37  id: ad.c6   Suspend: 0 Teb 7ff95000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
01b2feb4 77e79de5 00000003 01b2fedc 00000000 =
ntdll!NtWaitForMultipleObjects+0xb
01b2ff10 77e79e31 00000002 01b2ff58 000003e8 =
USER32!MsgWaitForMultipleObjectsEx+0xa8
01b2ff2c 697f1dbe 00000002 01b2ff58 00000000 =
USER32!MsgWaitForMultipleObjects+0x1d
69a92ea0 750169a9 24448b22 244c8b14 24548b10 =
MTxEx!ContextWrapper__Recycle+0xe
60b03d83 000003e6 0006ef1c 0006ee60 00000008 0x750169a9

 38  id: ad.bf   Suspend: 0 Teb 7ff94000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
01b6ff88 77f04f37 00000bc4 ffffffff 00000000 =
ntdll!NtWaitForSingleObject+0xb
01b6ffb8 77f04ede 001522a8 ffffffff 00fcf1fc =
KERNEL32!WaitForSingleObject+0xf
01b6ffec 00000000 776b1bdd 001522a8 00000000 KERNEL32!lstrcmpiW+0xbe

 39  id: ad.5d   Suspend: 0 Teb 7ff93000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
01bafd68 1e142143 0ae13b80 00000000 0aee1f60 =
python15!PyThreadState_Clear+0x6d
01bafd78 1e14267f 0aee1f60 00000000 1e2a2401 =
python15!PyInterpreterState_Clear+0x13
01bafd84 1e2a2401 01bafda4 1e2c4fd8 00000000 python15!Py_Finalize+0x3f
01bafdec 680c3b32 01bafe6c 00139248 77f050f0 =
pythoncom15!PyCom_DLLReleaseRef+0x71
01bafe34 680c2f38 000dfa58 00000000 01bafec0 =
asp!CAppConfig__SetValue+0xc8
01bafe7c 680c78c0 000dfa58 000c4698 01bafec0 =
asp!CPerfMainBlock__Init+0x47
01bafecc 680e25c6 00000000 697e1898 00313ea0 =
asp!CConnectionPoint__~CConnectionPoint+0x6
01bafee4 697e67cc 00000000 697e1898 0aceefc0 =
asp!_IASPObjectContextCustomStubVtbl+0x95
01baff24 697e57ce 00148598 697f21de 77f0178b =
MTxEx!CActivity__CallWithNewContext+0xac
77f67610 4affc033 89257508 fff00c42 037d044a =
MTxEx!CQueuedRequest__~CQueuedRequest+0x5e
0424548b 000003e6 0006ef1c 0006ee60 00000008 0x4affc033

 40  id: ad.73   Suspend: 0 Teb 7ff92000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
01dcff84 686248e5 0000016c 01dcffac 01dcffb4 =
ntdll!ZwRemoveIoCompletion+0xb
01dcffb8 77f04ede 00000000 00000133 00000000 ISATQ!AtqSetInfo+0xbb
01dcffec 00000000 686248a8 00000000 00000000 KERNEL32!lstrcmpiW+0xbe

 41  id: ad.4b   Suspend: 0 Teb 7ff6f000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
01e3ff84 686248e5 0000016c 01e3ffac 01e3ffb4 =
ntdll!ZwRemoveIoCompletion+0xb
01e3ffb8 77f04ede 00000000 ffffffff 0104fe68 ISATQ!AtqSetInfo+0xbb
01e3ffec 00000000 686248a8 00000000 00000000 KERNEL32!lstrcmpiW+0xbe

 42  id: ad.8a   Suspend: 0 Teb 7ff6e000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
01e7ff84 686248e5 0000016c 01e7ffac 01e7ffb4 =
ntdll!ZwRemoveIoCompletion+0xb
01e7ffb8 77f04ede 00000000 00000187 00001000 ISATQ!AtqSetInfo+0xbb
01e7ffec 00000000 686248a8 00000000 00000000 KERNEL32!lstrcmpiW+0xbe

 43  id: ad.144   Suspend: 0 Teb 7ff6d000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
034dff84 686248e5 0000016c 034dffac 034dffb4 =
ntdll!ZwRemoveIoCompletion+0xb
034dffb8 77f04ede 00000000 ffffffff 0104fe68 ISATQ!AtqSetInfo+0xbb
034dffec 00000000 686248a8 00000000 00000000 KERNEL32!lstrcmpiW+0xbe

 44  id: ad.147   Suspend: 0 Teb 7ff6c000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
0351ff84 686248e5 0000016c 0351ffac 0351ffb4 =
ntdll!ZwRemoveIoCompletion+0xb
0351ffb8 77f04ede 00000000 001f0003 00000000 ISATQ!AtqSetInfo+0xbb
0351ffec 00000000 686248a8 00000000 00000000 KERNEL32!lstrcmpiW+0xbe

 45  id: ad.149   Suspend: 0 Teb 7ff6b000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
0355ff84 686248e5 0000016c 0355ffac 0355ffb4 =
ntdll!ZwRemoveIoCompletion+0xb
0355ffb8 77f04ede 00000000 001f0003 00000000 ISATQ!AtqSetInfo+0xbb
0355ffec 00000000 686248a8 00000000 00000000 KERNEL32!lstrcmpiW+0xbe

 46  id: ad.101   Suspend: 0 Teb 7ff6a000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
0359ff84 686248e5 0000016c 0359ffac 0359ffb4 =
ntdll!ZwRemoveIoCompletion+0xb
0359ffb8 77f04ede 00000000 0000003d 00001000 ISATQ!AtqSetInfo+0xbb
0359ffec 00000000 686248a8 00000000 00000000 KERNEL32!lstrcmpiW+0xbe

 47  id: ad.136   Suspend: 0 Teb 7fdaf000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
0361ff84 686248e5 0000016c 0361ffac 0361ffb4 =
ntdll!ZwRemoveIoCompletion+0xb
0361ffb8 77f04ede 00000000 001f0003 00000000 ISATQ!AtqSetInfo+0xbb
0361ffec 00000000 686248a8 00000000 00000000 KERNEL32!lstrcmpiW+0xbe

 48  id: ad.13e   Suspend: 0 Teb 7ff69000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
035dff84 686248e5 0000016c 035dffac 035dffb4 =
ntdll!ZwRemoveIoCompletion+0xb
035dffb8 77f04ede 00000000 001f0003 00000000 ISATQ!AtqSetInfo+0xbb
035dffec 00000000 686248a8 00000000 00000000 KERNEL32!lstrcmpiW+0xbe

 49  id: ad.14e   Suspend: 0 Teb 7fdae000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
0365ff84 686248e5 0000016c 0365ffac 0365ffb4 =
ntdll!ZwRemoveIoCompletion+0xb
0365ffb8 77f04ede 00000000 001f0003 00000000 ISATQ!AtqSetInfo+0xbb
0365ffec 00000000 686248a8 00000000 00000000 KERNEL32!lstrcmpiW+0xbe

 50  id: ad.16f   Suspend: 0 Teb 7f81f000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
0584ff84 686248e5 0000016c 0584ffac 0584ffb4 =
ntdll!ZwRemoveIoCompletion+0xb
0584ffb8 77f04ede 00000000 00000000 00000000 ISATQ!AtqSetInfo+0xbb
0584ffec 00000000 686248a8 00000000 00000000 KERNEL32!lstrcmpiW+0xbe

 51  id: ad.135   Suspend: 0 Teb 7f81e000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
048cff84 686248e5 0000016c 048cffac 048cffb4 =
ntdll!ZwRemoveIoCompletion+0xb
048cffb8 77f04ede 00000000 00908e50 000003e5 ISATQ!AtqSetInfo+0xbb
048cffec 00000000 686248a8 00000000 00000000 KERNEL32!lstrcmpiW+0xbe

 52  id: ad.e0   Suspend: 0 Teb 7f81d000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
0588ff84 686248e5 0000016c 0588ffac 0588ffb4 =
ntdll!ZwRemoveIoCompletion+0xb
0588ffb8 77f04ede 00000000 ffffffff 0104fe68 ISATQ!AtqSetInfo+0xbb
0588ffec 00000000 686248a8 00000000 00000000 KERNEL32!lstrcmpiW+0xbe

 53  id: ad.ec   Suspend: 0 Teb 7f81c000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
058cff84 686248e5 0000016c 058cffac 058cffb4 =
ntdll!ZwRemoveIoCompletion+0xb
058cffb8 77f04ede 00000000 ffffffff 0104fe68 ISATQ!AtqSetInfo+0xbb
058cffec 00000000 686248a8 00000000 00000000 KERNEL32!lstrcmpiW+0xbe

 54  id: ad.169   Suspend: 0 Teb 7f81b000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
0590ff84 686248e5 0000016c 0590ffac 0590ffb4 =
ntdll!ZwRemoveIoCompletion+0xb
0590ffb8 77f04ede 00000000 0090af28 000003e5 ISATQ!AtqSetInfo+0xbb
0590ffec 00000000 686248a8 00000000 00000000 KERNEL32!lstrcmpiW+0xbe

 55  id: ad.158   Suspend: 0 Teb 7f81a000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
0594ff84 686248e5 0000016c 0594ffac 0594ffb4 =
ntdll!ZwRemoveIoCompletion+0xb
0594ffb8 77f04ede 00000000 001f0003 00000000 ISATQ!AtqSetInfo+0xbb
0594ffec 00000000 686248a8 00000000 00000000 KERNEL32!lstrcmpiW+0xbe

 56  id: ad.167   Suspend: 0 Teb 7f819000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
0598ff84 686248e5 0000016c 0598ffac 0598ffb4 =
ntdll!ZwRemoveIoCompletion+0xb
0598ffb8 77f04ede 00000000 ffffffff 0590fe68 ISATQ!AtqSetInfo+0xbb
0598ffec 00000000 686248a8 00000000 00000000 KERNEL32!lstrcmpiW+0xbe

 57  id: ad.162   Suspend: 0 Teb 7f818000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
059cff84 686248e5 0000016c 059cffac 059cffb4 =
ntdll!ZwRemoveIoCompletion+0xb
059cffb8 77f04ede 00000000 ffffffff 0590fe68 ISATQ!AtqSetInfo+0xbb
059cffec 00000000 686248a8 00000000 00000000 KERNEL32!lstrcmpiW+0xbe

 58  id: ad.6d   Suspend: 0 Teb 7f817000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
05a0ff84 686248e5 0000016c 05a0ffac 05a0ffb4 =
ntdll!ZwRemoveIoCompletion+0xb
05a0ffb8 77f04ede 00000000 00000000 00000000 ISATQ!AtqSetInfo+0xbb
05a0ffec 00000000 686248a8 00000000 00000000 KERNEL32!lstrcmpiW+0xbe

 59  id: ad.44   Suspend: 0 Teb 7f816000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
05a4ff84 686248e5 0000016c 05a4ffac 05a4ffb4 =
ntdll!ZwRemoveIoCompletion+0xb
05a4ffb8 77f04ede 00000000 0000000f 00001000 ISATQ!AtqSetInfo+0xbb
05a4ffec 00000000 686248a8 00000000 00000000 KERNEL32!lstrcmpiW+0xbe

 60  id: ad.178   Suspend: 0 Teb 7f815000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
05a8ff84 686248e5 0000016c 05a8ffac 05a8ffb4 =
ntdll!ZwRemoveIoCompletion+0xb
05a8ffb8 77f04ede 00000000 0000020c 00001000 ISATQ!AtqSetInfo+0xbb
05a8ffec 00000000 686248a8 00000000 00000000 KERNEL32!lstrcmpiW+0xbe

 61  id: ad.176   Suspend: 0 Teb 7f814000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
05acff84 686248e5 0000016c 05acffac 05acffb4 =
ntdll!ZwRemoveIoCompletion+0xb
05acffb8 77f04ede 00000000 00000009 00001000 ISATQ!AtqSetInfo+0xbb
05acffec 00000000 686248a8 00000000 00000000 KERNEL32!lstrcmpiW+0xbe

 62  id: ad.d4   Suspend: 0 Teb 7f813000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
05b0ff84 686248e5 0000016c 05b0ffac 05b0ffb4 =
ntdll!ZwRemoveIoCompletion+0xb
05b0ffb8 77f04ede 00000000 001f0003 00000000 ISATQ!AtqSetInfo+0xbb
05b0ffec 00000000 686248a8 00000000 00000000 KERNEL32!lstrcmpiW+0xbe

 63  id: ad.156   Suspend: 0 Teb 7f812000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
05b4ff84 686248e5 0000016c 05b4ffac 05b4ffb4 =
ntdll!ZwRemoveIoCompletion+0xb
05b4ffb8 77f04ede 00000000 00000003 00001000 ISATQ!AtqSetInfo+0xbb
05b4ffec 00000000 686248a8 00000000 00000000 KERNEL32!lstrcmpiW+0xbe

 64  id: ad.166   Suspend: 0 Teb 7f811000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
05b8ff84 686248e5 0000016c 05b8ffac 05b8ffb4 =
ntdll!ZwRemoveIoCompletion+0xb
05b8ffb8 77f04ede 00000000 001f0003 00000000 ISATQ!AtqSetInfo+0xbb
05b8ffec 00000000 686248a8 00000000 00000000 KERNEL32!lstrcmpiW+0xbe

 65  id: ad.68   Suspend: 0 Teb 7f810000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
05bcff84 686248e5 0000016c 05bcffac 05bcffb4 =
ntdll!ZwRemoveIoCompletion+0xb
05bcffb8 77f04ede 00000000 00000000 00000000 ISATQ!AtqSetInfo+0xbb
05bcffec 00000000 686248a8 00000000 00000000 KERNEL32!lstrcmpiW+0xbe

 66  id: ad.13d   Suspend: 0 Teb 7f80f000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
0d2cff90 77e15a1d 77e160f7 0007f878 0d2cffec =
ntdll!ZwReplyWaitReceivePort+0xb
00003a98 000003e6 00000088 77e15a1d 00000024 RPCRT4!NdrOleAllocate+0x24
0:039> !locks

0:039> !inetdbg.atq -g

Atq Globals:
	isatq!g_cConcurrency          =3D        1 (0x       1)
	isatq!g_cThreads              =3D        2 (0x       2)
	isatq!g_cAvailableThreads     =3D        0 (0x       0)
	isatq!g_cMaxThreads           =3D        0 (0x       0)

	isatq!g_fUseAcceptEx          =3D        0 (0x       0)
	isatq!g_cbXmitBufferSize      =3D        0 (0x       0)
	isatq!g_cbMinKbSec            =3D        0 (0x       0)
	isatq!g_cCPU                  =3D        0 (0x       0)
	isatq!g_fShutdown             =3D        2 (0x       2)

	isatq!g_msThreadTimeout       =3D        0 (0x       0)
	isatq!g_dwTimeoutCookie       =3D        0 (0x       0)
	isatq!g_cListenBacklog        =3D        0 (0x       0)
	isatq!AtqGlobalContextCount   =3D        0 (0x       0)
	sizeof(ATQ_CONTEXT) =3D 212
0:039> !inetdbg.asp -g

Asp Globals:
	asp!g_nSessions                 =3D        0 (0x       0)
	asp!g_nApplications             =3D        0 (0x       0)
	asp!g_nApplicationsRestarting   =3D        0 (0x       0)
	asp!g_nBrowserRequests          =3D        0 (0x       0)
	asp!g_nSessionCleanupRequests   =3D        0 (0x       0)
	asp!g_nApplnCleanupRequests     =3D        0 (0x       0)
	asp!g_fShutDownInProgress       =3D 1868967740 (0x6f662f3c)
0:039> !inetdbg.ds -v esp

01bafd6c : 1e142143 : python15!PyInterpreterState_Clear+0x13
01bafd7c : 1e14267f : python15!Py_Finalize+0x3f
01bafd88 : 1e2a2401 : pythoncom15!PyCom_DLLReleaseRef+0x71
01bafd90 : 1e2c4fd8 : pythoncom15!PyITypeComp__Bind+0x1c28
01bafd98 : 1e2a6795 : pythoncom15!PyGatewayBase__~PyGatewayBase+0xa5
01bafda8 : 1e2c5108 : pythoncom15!PyITypeComp__Bind+0x1d58
01bafdb0 : 1e2a6a41 : pythoncom15!PyGatewayBase__Release+0x21
*** WARNING: Using export symbols for axscript.pyd
01bafdbc : 01d3245b : axscript!PyGActiveScript__Release+0xb
01bafdc4 : 680d3924 : asp!STR__Copy+0x17
01bafdcc : 680ec330 : asp!g_szErrTemplate+0x17f
01bafdd4 : 680d50cd : asp!CCertRequest__ParseRDNS+0x43
01bafdf0 : 680c3b32 : asp!CAppConfig__SetValue+0xc8
01bafdfc : 77f050f0 : KERNEL32!InterlockedIncrement
01bafe04 : 680ebe20 : asp!g_PerfMain+0x7
01bafe08 : 680c31e0 : asp!CPerfMainBlock__AddProcess+0x49
01bafe24 : 680c3b58 : asp!CAppConfig__SetValue+0x11a
01bafe2c : 680ec6d0 : asp!g_szErrTemplate+0x51f
01bafe38 : 680c2f38 : asp!CPerfMainBlock__Init+0x47
01bafe4c : 77f050f0 : KERNEL32!InterlockedIncrement
01bafe54 : 680ebe20 : asp!g_PerfMain+0x7
01bafe80 : 680c78c0 : asp!CConnectionPoint__~CConnectionPoint+0x6
01bafed0 : 680e25c6 : asp!_IASPObjectContextCustomStubVtbl+0x95
01bafed8 : 697e1898 : MTxEx!IID_IRegNode+0x8
01bafee8 : 697e67cc : MTxEx!CActivity__CallWithNewContext+0xac
01bafef0 : 697e1898 : MTxEx!IID_IRegNode+0x8
01baff18 : 6981ca16 : MTxEx!CAspSecurity__Release+0x1d6
01baff1c : 697e18b0 : MTxEx!IID_IRegNodeDispenser
01baff24 : 77f67610 : ntdll!RtlLeaveCriticalSection
01baff28 : 697e57ce : MTxEx!CQueuedRequest__~CQueuedRequest+0x5e
01baff30 : 697f21de : MTxEx!ContextWrapper__QueryInterface+0x352
01baff34 : 77f0178b : KERNEL32!GetLastError
01baff3c : 69a92ea0 : MTxInfr1!VipThunkPeekMessageW
01baff44 : 697f1ed1 : MTxEx!ContextWrapper__QueryInterface+0x45
01baff50 : 697e21f8 : MTxEx!IID_IJavaObjectPrivate
01baff7c : 697f22d2 : MTxEx!ContextWrapper__Release+0x6e
01baff88 : 7800be50 : MSVCRT!endthread+0x96
*** WARNING: Using export symbols for ole32.dll
01baff94 : 77b2aa4d : ole32!CreateBindCtx+0x2d10
01baffac : 7800ef57 : MSVCRT!except_handler3
01baffb0 : 78033200 : MSVCRT!__non_rtti_object__`vftable'+0x33c
01baffbc : 77f04ede : KERNEL32!lstrcmpiW+0xbe
01baffc8 : 77b2aa4d : ole32!CreateBindCtx+0x2d10
01baffd0 : 77b2aa4d : ole32!CreateBindCtx+0x2d10
01baffd8 : 77bc7d60 : ole32+0xa7d60
01baffe0 : 77f3b944 : KERNEL32!GetDateFormatW+0x411f
01baffe4 : 77f3d238 : KERNEL32!GetDateFormatW+0x5a13
01bafff4 : 7800bdf9 : MSVCRT!endthread+0x3f
inetdbg!ds: cannot read memory @ 1bb0000
0:039> !inetdbg.ds esp

01bafd64 : 0aee1f60
01bafd68 : 0ae13b80
01bafd6c : 1e142143 : python15!PyInterpreterState_Clear+0x13
01bafd70 : 0ae13b80
01bafd74 : 00000000
01bafd78 : 0aee1f60
01bafd7c : 1e14267f : python15!Py_Finalize+0x3f
01bafd80 : 0aee1f60
01bafd84 : 00000000
01bafd88 : 1e2a2401 : pythoncom15!PyCom_DLLReleaseRef+0x71
01bafd8c : 01bafda4
01bafd90 : 1e2c4fd8 : pythoncom15!PyITypeComp__Bind+0x1c28
01bafd94 : 00000000
01bafd98 : 1e2a6795 : pythoncom15!PyGatewayBase__~PyGatewayBase+0xa5
01bafd9c : 0ae631b0
01bafda0 : 00000001
01bafda4 : 01baff14
01bafda8 : 1e2c5108 : pythoncom15!PyITypeComp__Bind+0x1d58
01bafdac : ffffffff
01bafdb0 : 1e2a6a41 : pythoncom15!PyGatewayBase__Release+0x21
01bafdb4 : 00084fd0
01bafdb8 : 00084f50
01bafdbc : 01d3245b : axscript!PyGActiveScript__Release+0xb
01bafdc0 : 0ae631b0
01bafdc4 : 680d3924 : asp!STR__Copy+0x17
01bafdc8 : 0ae631b0
01bafdcc : 680ec330 : asp!g_szErrTemplate+0x17f
01bafdd0 : 00084f50
01bafdd4 : 680d50cd : asp!CCertRequest__ParseRDNS+0x43
01bafdd8 : 00000000
01bafddc : 01bafe58
01bafde0 : 00000000
01bafde4 : 00000000
01bafde8 : 01c3ace0
01bafdec : 01bafe34
01bafdf0 : 680c3b32 : asp!CAppConfig__SetValue+0xc8
01bafdf4 : 01bafe6c
01bafdf8 : 00139248
01bafdfc : 77f050f0 : KERNEL32!InterlockedIncrement
01bafe00 : 000c4698
01bafe04 : 680ebe20 : asp!g_PerfMain+0x7
01bafe08 : 680c31e0 : asp!CPerfMainBlock__AddProcess+0x49
01bafe0c : 01bafe58
01bafe10 : 00139248
01bafe14 : 01bafec0
01bafe18 : 000c4698
01bafe1c : 00000000
01bafe20 : 0013f741
!inetdbg.ds 1bafe24 to dump next block
0:039> u eip-50 eip+10

python15!PyThreadState_Clear+0x1d:
1e14229d 2d171e83c0       sub     eax,0xc0831e17
1e1422a2 40               inc     eax
1e1422a3 50               push    eax
1e1422a4 ff15fce4171e     call    dword ptr [python15+0x7e4fc =
(1e17e4fc)]
1e1422aa 83c408           add     esp,0x8
1e1422ad 8b4608           mov     eax,[esi+0x8]
1e1422b0 897e08           mov     [esi+0x8],edi
1e1422b3 3bc7             cmp     eax,edi
1e1422b5 7411             jz     python15!PyThreadState_Clear+0x48 =
(1e1422c8)
1e1422b7 8b08             mov     ecx,[eax]
1e1422b9 49               dec     ecx
1e1422ba 8908             mov     [eax],ecx
1e1422bc 750a             jnz    python15!PyThreadState_Clear+0x48 =
(1e1422c8)
1e1422be 8b4804           mov     ecx,[eax+0x4]
1e1422c1 50               push    eax
1e1422c2 ff5118           call    dword ptr [ecx+0x18]
1e1422c5 83c404           add     esp,0x4
1e1422c8 8b4638           mov     eax,[esi+0x38]
1e1422cb 897e38           mov     [esi+0x38],edi
1e1422ce 3bc7             cmp     eax,edi
1e1422d0 7411             jz     python15!PyThreadState_Clear+0x63 =
(1e1422e3)
1e1422d2 8b08             mov     ecx,[eax]
1e1422d4 49               dec     ecx
1e1422d5 8908             mov     [eax],ecx
1e1422d7 750a             jnz    python15!PyThreadState_Clear+0x63 =
(1e1422e3)
1e1422d9 8b5004           mov     edx,[eax+0x4]
1e1422dc 50               push    eax
1e1422dd ff5218           call    dword ptr [edx+0x18]
1e1422e0 83c404           add     esp,0x4
1e1422e3 8b4620           mov     eax,[esi+0x20]
1e1422e6 897e20           mov     [esi+0x20],edi
1e1422e9 3bc7             cmp     eax,edi
1e1422eb 7411             jz     python15!PyThreadState_Clear+0x7e =
(1e1422fe)
1e1422ed 8b08             mov     ecx,[eax]
1e1422ef 49               dec     ecx
1e1422f0 8908             mov     [eax],ecx
1e1422f2 750a             jnz    python15!PyThreadState_Clear+0x7e =
(1e1422fe)
1e1422f4 8b4804           mov     ecx,[eax+0x4]
1e1422f7 50               push    eax
1e1422f8 ff5118           call    dword ptr [ecx+0x18]
1e1422fb 83c404           add     esp,0x4
0:039> dc eax-100 eax+100

                        ^ Range error in 'dc eax-100 eax+100'
0:039> dc ebx-100 ebx+100

680ec370  00000000 00000000 00000000 00000000  ................
680ec380  00000000 00000000 00000000 00000000  ................
680ec390  00000000 00000000 00000000 00000000  ................
680ec3a0  00000000 00000000 00000000 00000000  ................
680ec3b0  00000000 00000000 00000000 680b1b10  ...............h
680ec3c0  00159014 00159014 00140478 ffffffff  ........x.......
680ec3d0  00000000 00000000 00000000 00000000  ................
680ec3e0  680b53d0 00000001 00000000 00000000  .S.h............
680ec3f0  680ec400 680c57c0 00000003 00000000  ...h.W.h........
680ec400  00000000 00000000 00000000 00000000  ................
680ec410  00000000 00000000 00000000 00000000  ................
680ec420  00000000 00000000 00000000 00000000  ................
680ec430  00000000 00000000 00000000 00000000  ................
680ec440  00000000 00000000 00000000 00000000  ................
680ec450  00000000 00000000 00000000 00000000  ................
680ec460  00000000 680b1b10 680ec464 680ec464  .......hd..hd..h
680ec470  00140498 ffffffff 00000000 00000000  ................
680ec480  00000000 00000000 680b3340 00000001  ........ at 3.h....
680ec490  0014a2a8 00146d28 680ec4a8 680c57c0  ....(m.....h.W.h
680ec4a0  0000000b 00000002 00000000 00146d28  ............(m..
680ec4b0  00000000 0014a2a8 00000000 00000000  ................
680ec4c0  00000000 00000000 00000000 00000000  ................
680ec4d0  00000000 00000000 00000000 00000000  ................
680ec4e0  00000000 00000000 00000000 00000000  ................
680ec4f0  00000000 00000000 00000000 00000000  ................
680ec500  00000000 00000000 00000000 00140458  ............X...
680ec510  ffffffff 00000000 00000000 00000000  ................
680ec520  00000000 000000a2 00000002 00000008  ................
680ec530  00000007 00293db8 00000002 00000008  .....=3D).........
680ec540  00000007 00000000 00000002 00000000  ................
680ec550  00000001 00140378 ffffffff 00000000  ....x...........
680ec560  00000000 00000000 00000000 3101f691  ...............1
680ec570  3101fa55                             U..1
0:039> dc ecx-100 ecx+100

                        ^ Range error in 'dc ecx-100 ecx+100'
0:039> dc edx-100 edx+100

7803c020  00000000 00000000 00000000 00000000  ................
7803c030  00000000 7803c0f0 00299890 00000000  .......x..).....
7803c040  00000000 00000000 00000000 00000000  ................
7803c050  00000000 7803c120 00000000 0029acd0  .... ..x......).
7803c060  00296a30 7803c138 00000000 00000000  0j).8..x........
7803c070  00000000 7803c108 00299900 00240600  .......x..)...$.
7803c080  00000000 00000000 00000000 00000000  ................
7803c090  00000000 00240180 00000000 00000000  ......$.........
7803c0a0  00000000 003cbfc0 002414b0 002998e0  ......<...$...).
7803c0b0  0029acf0 002a0b70 002b0910 002b5ad0  ..).p.*...+..Z+.
7803c0c0  002bbd30 00354970 00000000 00000000  0.+.pI5.........
7803c0d0  00000000 00000000 00000000 00000000  ................
7803c0e0  00000000 00000000 00000000 00000000  ................
7803c0f0  77fa7790 ffffffff 00000000 00000000  .w.w............
7803c100  00000000 00000000 77fa7730 ffffffff  ........0w.w....
7803c110  00000000 00000000 00000000 00000000  ................
7803c120  77fa7770 ffffffff 00000000 00000000  pw.w............
7803c130  000003e0 00000000 77fa7750 ffffffff  ........Pw.w....
7803c140  00000000 00000000 00000000 00000000  ................
7803c150  000001e0 04e91ff0 00000000 0000001e  ................
7803c160  00200000 00200020 00200020 00200020  .. . . . . . . .
7803c170  00200020 00280028 00280028 00200028   . .(.(.(.(.(. .
7803c180  00200020 00200020 00200020 00200020   . . . . . . . .
7803c190  00200020 00200020 00200020 00200020   . . . . . . . .
7803c1a0  00480020 00100010 00100010 00100010   .H.............
7803c1b0  00100010 00100010 00100010 00100010  ................
7803c1c0  00840010 00840084 00840084 00840084  ................
7803c1d0  00840084 00100084 00100010 00100010  ................
7803c1e0  00100010 00810081 00810081 00810081  ................
7803c1f0  00010001 00010001 00010001 00010001  ................
7803c200  00010001 00010001 00010001 00010001  ................
7803c210  00010001 00010001 00100010 00100010  ................
7803c220  00100010                             ....
0:039> dc esi-100 esi+100

0ae13a80  00000002 1e176bd8 0000001a e91943f6  .....k.......C..
0ae13a90  0ae13a80 5f4e4150 49454858 5f544847  .:..PAN_XHEIGHT_
0ae13aa0  534e4f43 544e4154 52414c5f 44004547  CONSTANT_LARGE.D
0ae13ab0  5954535f 4f00454c 00000057 2920bfe4  _STYLE.OW..... )
0ae13ac0  00000002 1e176bd8 0000000c fb80250b  .....k.......%..
0ae13ad0  0ae13ac0 454c4150 49455454 5845444e  .:..PALETTEINDEX
0ae13ae0  004e5200 72610000 00007367 0ae14460  .RN...args..`D..
0ae13af0  00000001 1e176bd8 00000005 ffffffff  .....k..........
0ae13b00  00000000 6f727265 695f0072 00000074  ....error._it...
0ae13b10  00000002 1e176bd8 00000019 9e835d28  .....k......(]..
0ae13b20  0ae13b10 5f4e4150 49454858 5f544847  .;..PAN_XHEIGHT_
0ae13b30  4b435544 5f474e49 4c414d53 5902004c  DUCKING_SMALL..Y
0ae13b40  00000000 1e176bd8 00000011 174c09dc  .....k........L.
0ae13b50  00000002 1e176bd8 00000017 7ba26528  .....k......(e.{
0ae13b60  0ae13b50 5f4e4150 49454858 5f544847  P;..PAN_XHEIGHT_
0ae13b70  4b435544 5f474e49 00445453 a91a4b8b  DUCKING_STD..K..
0ae13b80  00000000 1e2d9bf8 00000000 00000000  ......-.........
0ae13b90  ffffffff ffffffff 00000004 00000003  ................
0ae13ba0  00000000 00000000 00000001 00000024  ............$...
0ae13bb0  00000004 00001100 00000000 00000000  ................
0ae13bc0  1e170238 0ada1390 00004445 0ae13e00  8.......ED...>..
0ae13bd0  00000001 1e176bd8 00000020 ffffffff  .....k.. .......
0ae13be0  00000000 62654449 6f446775 656d7563  ....IDebugDocume
0ae13bf0  6554746e 78457478 6e726574 75416c61  ntTextExternalAu
0ae13c00  726f6874 1e176b00 00000001 3f9b43ec  thor.k.......C.?
0ae13c10  00000002 1e176bd8 00000019 52e821e0  .....k.......!.R
0ae13c20  0ae13c10 5f4e4150 49454858 5f544847  .<..PAN_XHEIGHT_
0ae13c30  4b435544 5f474e49 4752414c 00540045  DUCKING_LARGE.T.
0ae13c40  0ae13ad0 7261776b 0a007367 0ad267f0  .:..kwargs...g..
0ae13c50  00000001 1e176bd8 00000015 ffffffff  .....k..........
0ae13c60  00000000 62654449 6f446775 656d7563  ....IDebugDocume
0ae13c70  6f43746e 7865746e 726f0074 0ad1d200  ntContext.or....
0ae13c80  00000002                             ....
0:039> dc edi-100 edi+100

                        ^ Range error in 'dc edi-100 edi+100'
0:039> dc esp-100 esp+100

01bafc64  00230000 00001000 05054000 00230548  ..#...... at ..H.#.
01bafc74  01d702f0 77f6447e 04d30000 01d702f0  ....~D.w........
01bafc84  77f6447e 04d30000 01bafcd8 05055000  ~D.w.........P..
01bafc94  05053aa0 00230000 77f650a8 04d30000  .:....#..P.w....
01bafca4  01bafcd4 00000000 77f8f2ad 00230000  ...........w..#.
01bafcb4  04ef5c38 00000000 01bafcb0 00000000  8\..............
01bafcc4  01bafd34 77f92cd4 77f9d870 ffffffff  4....,.wp..w....
01bafcd4  01bafcfc 77f64ddc 00231a90 04ef5c40  .....M.w..#.@\..
01bafce4  0000000c 04ef5c40 06062c50 0504ab20  ....@\..P,.. ...
01bafcf4  780012a6 780012a6 7803c120 780012a6  ...x...x ..x...x
01bafd04  7803c120 01bafd4c 78001434 00000009   ..xL...4..x....
01bafd14  780012a6 7803c120 01bafd60 78001434  ...x ..x`...4..x
01bafd24  00000009 78001405 00000020 78001385  .......x ......x
01bafd34  0aef4580 0aef4050 04e91ff0 0aef4000  .E..P at .......@..
01bafd44  78032250 ffffffff 01bafdec 01bafd8c  P".x............
01bafd54  7800ef57 78032250 ffffffff 01bafdec  W..xP".x........
01bafd64  0aee1f60 0ae13b80 1e142143 0ae13b80  `....;..C!...;..
01bafd74  00000000 0aee1f60 1e14267f 0aee1f60  ....`....&..`...
01bafd84  00000000 1e2a2401 01bafda4 1e2c4fd8  .....$*......O,.
01bafd94  00000000 1e2a6795 0ae631b0 00000001  .....g*..1......
01bafda4  01baff14 1e2c5108 ffffffff 1e2a6a41  .....Q,.....Aj*.
01bafdb4  00084fd0 00084f50 01d3245b 0ae631b0  .O..PO..[$...1..
01bafdc4  680d3924 0ae631b0 680ec330 00084f50  $9.h.1..0..hPO..
01bafdd4  680d50cd 00000000 01bafe58 00000000  .P.h....X.......
01bafde4  00000000 01c3ace0 01bafe34 680c3b32  ........4...2;.h
01bafdf4  01bafe6c 00139248 77f050f0 000c4698  l...H....P.w.F..
01bafe04  680ebe20 680c31e0 01bafe58 00139248   ..h.1.hX...H...
01bafe14  01bafec0 000c4698 00000000 0013f741  .....F......A...
01bafe24  680c3b58 00000000 680ec6d0 000000c0  X;.h.......h....
01bafe34  01bafe7c 680c2f38 000dfa58 00000000  |...8/.hX.......
01bafe44  01bafec0 01bafe58 77f050f0 000c4698  ....X....P.w.F..
01bafe54  680ebe20 00000001 00000001 01bafe64   ..h........d...
01bafe64  04198054                             T...
0:039> dc ebp-100 ebp+100

01bafcec  06062c50 0504ab20 780012a6 780012a6  P,.. ......x...x
01bafcfc  7803c120 780012a6 7803c120 01bafd4c   ..x...x ..xL...
01bafd0c  78001434 00000009 780012a6 7803c120  4..x.......x ..x
01bafd1c  01bafd60 78001434 00000009 78001405  `...4..x.......x
01bafd2c  00000020 78001385 0aef4580 0aef4050   ......x.E..P at ..
01bafd3c  04e91ff0 0aef4000 78032250 ffffffff  ..... at ..P".x....
01bafd4c  01bafdec 01bafd8c 7800ef57 78032250  ........W..xP".x
01bafd5c  ffffffff 01bafdec 0aee1f60 0ae13b80  ........`....;..
01bafd6c  1e142143 0ae13b80 00000000 0aee1f60  C!...;......`...
01bafd7c  1e14267f 0aee1f60 00000000 1e2a2401  .&..`........$*.
01bafd8c  01bafda4 1e2c4fd8 00000000 1e2a6795  .....O,......g*.
01bafd9c  0ae631b0 00000001 01baff14 1e2c5108  .1...........Q,.
01bafdac  ffffffff 1e2a6a41 00084fd0 00084f50  ....Aj*..O..PO..
01bafdbc  01d3245b 0ae631b0 680d3924 0ae631b0  [$...1..$9.h.1..
01bafdcc  680ec330 00084f50 680d50cd 00000000  0..hPO...P.h....
01bafddc  01bafe58 00000000 00000000 01c3ace0  X...............
01bafdec  01bafe34 680c3b32 01bafe6c 00139248  4...2;.hl...H...
01bafdfc  77f050f0 000c4698 680ebe20 680c31e0  .P.w.F.. ..h.1.h
01bafe0c  01bafe58 00139248 01bafec0 000c4698  X...H........F..
01bafe1c  00000000 0013f741 680c3b58 00000000  ....A...X;.h....
01bafe2c  680ec6d0 000000c0 01bafe7c 680c2f38  ...h....|...8/.h
01bafe3c  000dfa58 00000000 01bafec0 01bafe58  X...........X...
01bafe4c  77f050f0 000c4698 680ebe20 00000001  .P.w.F.. ..h....
01bafe5c  00000001 01bafe64 04198054 0419805c  ....d...T...\...
01bafe6c  00084f50 00000000 000dfa58 00000000  PO......X.......
01bafe7c  00000000 680c78c0 000dfa58 000c4698  .....x.hX....F..
01bafe8c  01bafec0 00000000 00148598 01bafee4  ................
01bafe9c  0013f6a8 00000000 00000000 00152678  ............x&..
01bafeac  000dfa58 00000000 00000000 fffffffe  X...............
01bafebc  00000001 00157258 00157280 001572a8  ....Xr...r...r..
01bafecc  001572d0 680e25c6 00000000 697e1898  .r...%.h......~i
01bafedc  00313ea0 0031caec 01baff24 697e67cc  .>1...1.$....g~i
01bafeec  00000000                             ....
0:039> dc esp+0x04

01bafd68  0ae13b80 1e142143 0ae13b80 00000000  .;..C!...;......
01bafd78  0aee1f60 1e14267f 0aee1f60 00000000  `....&..`.......
01bafd88  1e2a2401 01bafda4 1e2c4fd8 00000000  .$*......O,.....
01bafd98  1e2a6795 0ae631b0 00000001 01baff14  .g*..1..........
01bafda8  1e2c5108 ffffffff 1e2a6a41 00084fd0  .Q,.....Aj*..O..
01bafdb8  00084f50 01d3245b 0ae631b0 680d3924  PO..[$...1..$9.h
01bafdc8  0ae631b0 680ec330 00084f50 680d50cd  .1..0..hPO...P.h
01bafdd8  00000000 01bafe58 00000000 00000000  ....X...........
0:039> dc esp+0x08

01bafd6c  1e142143 0ae13b80 00000000 0aee1f60  C!...;......`...
01bafd7c  1e14267f 0aee1f60 00000000 1e2a2401  .&..`........$*.
01bafd8c  01bafda4 1e2c4fd8 00000000 1e2a6795  .....O,......g*.
01bafd9c  0ae631b0 00000001 01baff14 1e2c5108  .1...........Q,.
01bafdac  ffffffff 1e2a6a41 00084fd0 00084f50  ....Aj*..O..PO..
01bafdbc  01d3245b 0ae631b0 680d3924 0ae631b0  [$...1..$9.h.1..
01bafdcc  680ec330 00084f50 680d50cd 00000000  0..hPO...P.h....
01bafddc  01bafe58 00000000 00000000 01c3ace0  X...............
0:039> dc esp+0x0c

01bafd70  0ae13b80 00000000 0aee1f60 1e14267f  .;......`....&..
01bafd80  0aee1f60 00000000 1e2a2401 01bafda4  `........$*.....
01bafd90  1e2c4fd8 00000000 1e2a6795 0ae631b0  .O,......g*..1..
01bafda0  00000001 01baff14 1e2c5108 ffffffff  .........Q,.....
01bafdb0  1e2a6a41 00084fd0 00084f50 01d3245b  Aj*..O..PO..[$..
01bafdc0  0ae631b0 680d3924 0ae631b0 680ec330  .1..$9.h.1..0..h
01bafdd0  00084f50 680d50cd 00000000 01bafe58  PO...P.h....X...
01bafde0  00000000 00000000 01c3ace0 01bafe34  ............4...
0:039> dc esp+0x10

01bafd74  00000000 0aee1f60 1e14267f 0aee1f60  ....`....&..`...
01bafd84  00000000 1e2a2401 01bafda4 1e2c4fd8  .....$*......O,.
01bafd94  00000000 1e2a6795 0ae631b0 00000001  .....g*..1......
01bafda4  01baff14 1e2c5108 ffffffff 1e2a6a41  .....Q,.....Aj*.
01bafdb4  00084fd0 00084f50 01d3245b 0ae631b0  .O..PO..[$...1..
01bafdc4  680d3924 0ae631b0 680ec330 00084f50  $9.h.1..0..hPO..
01bafdd4  680d50cd 00000000 01bafe58 00000000  .P.h....X.......
01bafde4  00000000 01c3ace0 01bafe34 680c3b32  ........4...2;.h
0:039> dc ebp+0x04

01bafdf0  680c3b32 01bafe6c 00139248 77f050f0  2;.hl...H....P.w
01bafe00  000c4698 680ebe20 680c31e0 01bafe58  .F.. ..h.1.hX...
01bafe10  00139248 01bafec0 000c4698 00000000  H........F......
01bafe20  0013f741 680c3b58 00000000 680ec6d0  A...X;.h.......h
01bafe30  000000c0 01bafe7c 680c2f38 000dfa58  ....|...8/.hX...
01bafe40  00000000 01bafec0 01bafe58 77f050f0  ........X....P.w
01bafe50  000c4698 680ebe20 00000001 00000001  .F.. ..h........
01bafe60  01bafe64 04198054 0419805c 00084f50  d...T...\...PO..
0:039> dc ebp+0x08

01bafdf4  01bafe6c 00139248 77f050f0 000c4698  l...H....P.w.F..
01bafe04  680ebe20 680c31e0 01bafe58 00139248   ..h.1.hX...H...
01bafe14  01bafec0 000c4698 00000000 0013f741  .....F......A...
01bafe24  680c3b58 00000000 680ec6d0 000000c0  X;.h.......h....
01bafe34  01bafe7c 680c2f38 000dfa58 00000000  |...8/.hX.......
01bafe44  01bafec0 01bafe58 77f050f0 000c4698  ....X....P.w.F..
01bafe54  680ebe20 00000001 00000001 01bafe64   ..h........d...
01bafe64  04198054 0419805c 00084f50 00000000  T...\...PO......
0:039> dc ebp+0x0c

01bafdf8  00139248 77f050f0 000c4698 680ebe20  H....P.w.F.. ..h
01bafe08  680c31e0 01bafe58 00139248 01bafec0  .1.hX...H.......
01bafe18  000c4698 00000000 0013f741 680c3b58  .F......A...X;.h
01bafe28  00000000 680ec6d0 000000c0 01bafe7c  .......h....|...
01bafe38  680c2f38 000dfa58 00000000 01bafec0  8/.hX...........
01bafe48  01bafe58 77f050f0 000c4698 680ebe20  X....P.w.F.. ..h
01bafe58  00000001 00000001 01bafe64 04198054  ........d...T...
01bafe68  0419805c 00084f50 00000000 000dfa58  \...PO......X...
0:039> dc ebp+0x10

01bafdfc  77f050f0 000c4698 680ebe20 680c31e0  .P.w.F.. ..h.1.h
01bafe0c  01bafe58 00139248 01bafec0 000c4698  X...H........F..
01bafe1c  00000000 0013f741 680c3b58 00000000  ....A...X;.h....
01bafe2c  680ec6d0 000000c0 01bafe7c 680c2f38  ...h....|...8/.h
01bafe3c  000dfa58 00000000 01bafec0 01bafe58  X...........X...
01bafe4c  77f050f0 000c4698 680ebe20 00000001  .P.w.F.. ..h....
01bafe5c  00000001 01bafe64 04198054 0419805c  ....d...T...\...
01bafe6c  00084f50 00000000 000dfa58 00000000  PO......X.......
0:039> dc poi(esp+0x04)

0ae13b80  00000000 1e2d9bf8 00000000 00000000  ......-.........
0ae13b90  ffffffff ffffffff 00000004 00000003  ................
0ae13ba0  00000000 00000000 00000001 00000024  ............$...
0ae13bb0  00000004 00001100 00000000 00000000  ................
0ae13bc0  1e170238 0ada1390 00004445 0ae13e00  8.......ED...>..
0ae13bd0  00000001 1e176bd8 00000020 ffffffff  .....k.. .......
0ae13be0  00000000 62654449 6f446775 656d7563  ....IDebugDocume
0ae13bf0  6554746e 78457478 6e726574 75416c61  ntTextExternalAu
0:039> dc poi(esp+0x08)

1e142143  c483368b 75f68504 08478bf1 000847c7  .6.....u..G..G..
1e142153  85000000 8b1174c0 08894908 488b0a75  .....t...I..u..H
1e142163  51ff5004 04c48318 c70c478b 00000c47  .P.Q.....G..G...
1e142173  c0850000 088b1174 75088949 04508b0a  ....t...I..u..P.
1e142183  1852ff50 8b04c483 47c71047 00000010  P.R.....G..G....
1e142193  74c08500 49088b11 0a750889 5004488b  ...t...I..u..H.P
1e1421a3  831851ff 5e5f04c4 909090c3 8b575690  .Q...._^.....VW.
1e1421b3  570c247c 000054e8 04c48300 17cd34be  |$.W.T.......4..
0:039> dc poi(esp+0x0c)

0ae13b80  00000000 1e2d9bf8 00000000 00000000  ......-.........
0ae13b90  ffffffff ffffffff 00000004 00000003  ................
0ae13ba0  00000000 00000000 00000001 00000024  ............$...
0ae13bb0  00000004 00001100 00000000 00000000  ................
0ae13bc0  1e170238 0ada1390 00004445 0ae13e00  8.......ED...>..
0ae13bd0  00000001 1e176bd8 00000020 ffffffff  .....k.. .......
0ae13be0  00000000 62654449 6f446775 656d7563  ....IDebugDocume
0ae13bf0  6554746e 78457478 6e726574 75416c61  ntTextExternalAu
0:039> dc poi(esp+0x10)

00000000  ???????? ???????? ???????? ????????  ????????????????
00000010  ???????? ???????? ???????? ????????  ????????????????
00000020  ???????? ???????? ???????? ????????  ????????????????
00000030  ???????? ???????? ???????? ????????  ????????????????
00000040  ???????? ???????? ???????? ????????  ????????????????
00000050  ???????? ???????? ???????? ????????  ????????????????
00000060  ???????? ???????? ???????? ????????  ????????????????
00000070  ???????? ???????? ???????? ????????  ????????????????
0:039> dc poi(ebp+0x04)

680c3b32  c7b70f47 7c04463b 045e89de 7e013e83  G...;F.|..^..>.~
680c3b42  ff53530a 15e80876 89000050 085e891e  .SS.v...P.....^.
680c3b52  c25b5e5f 4c8b0008 448b0824 413b0424  _^[....L$..D$.;A
680c3b62  33047204 8b0aebc0 048d0849 81448b80  .r.3....I.....D.
680c3b72  0008c208 8bec8b55 4d8b1845 00208014  ....U...E..M.. .
680c3b82  1f74c985 08758b56 0f118b50 ff0c45bf  ..t.V.u.P....E..
680c3b92  768b1075 80048d08 088674ff 1452ff51  u..v.....t..Q.R.
680c3ba2  0f19eb5e 8b0c45bf 75ff084d 80048d10  ^....E..M..u....
0:039> dc poi(ebp+0x08)

01bafe6c  00084f50 00000000 000dfa58 00000000  PO......X.......
01bafe7c  00000000 680c78c0 000dfa58 000c4698  .....x.hX....F..
01bafe8c  01bafec0 00000000 00148598 01bafee4  ................
01bafe9c  0013f6a8 00000000 00000000 00152678  ............x&..
01bafeac  000dfa58 00000000 00000000 fffffffe  X...............
01bafebc  00000001 00157258 00157280 001572a8  ....Xr...r...r..
01bafecc  001572d0 680e25c6 00000000 697e1898  .r...%.h......~i
01bafedc  00313ea0 0031caec 01baff24 697e67cc  .>1...1.$....g~i
0:039> dc poi(ebp+0x0c)

00139248  680b1aa8 680b3bc0 00000000 0014411c  ...h.;.h.....A..
00139258  680b1ae4 00146930 00000011 00156340  ...h0i...... at c..
00139268  00000000 00000001 00000006 00000001  ................
00139278  00000001 00146930 001441f8 00146ca8  ....0i...A...l..
00139288  00000000 00000000 00146df0 000a2200  .........m..."..
00139298  0007a800 001434c0 00146da0 00000000  .....4...m......
001392a8  00140578 ffffffff 00000000 00000000  x...............
001392b8  00000000 00000000 001405b8 ffffffff  ................
0:039> dc poi(ebp+0x10)

77f050f0  04244c8b 000001b8 c10ff000 04c24001  .L$.......... at ..
77f05100  c08b2e00 04244c8b ffffffb8 c10ff0ff  .....L$.........
77f05110  04c24801 c08b2e00 04244c8b 0824548b  .H.......L$..T$.
77f05120  0ff0018b fa7511b1 900008c2 04244c8b  ......u......L$.
77f05130  0824548b 0c24448b 11b10ff0 90000cc2  .T$..D$.........
77f05140  04244c8b 0824448b 01c10ff0 900008c2  .L$..D$.........
77f05150  0424448b 3878c00b 0824548b 880fd20b  .D$...x8.T$.....
77f05160  000000b2 4c8be2f7 c90b0c24 00e6880f  .......L$.......
0:039> !inetdbg.ver

Module @ 0x01000000 =3D inetinfo.exe
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D Internet Information Services
FileVersion         =3D 4.02.0622
InternalName        =3D INETINFO.EXE
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1997
OriginalFilename    =3D INETINFO.EXE
ProductName         =3D Internet Information Server
ProductVersion      =3D 4.02.0622

Module @ 0x77f60000 =3D ntdll.dll
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D NT Layer DLL
FileVersion         =3D 4.00
InternalName        =3D ntdll.dll
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1996
OriginalFilename    =3D ntdll.dll
ProductName         =3D Microsoft(R) Windows NT(TM) Operating System
ProductVersion      =3D 4.00

Module @ 0x78000000 =3D MSVCRT.dll
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D Microsoft (R) C Runtime Library
FileVersion         =3D 6.10.8528.0
InternalName        =3D MSVCRT.DLL
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1998
OriginalFilename    =3D MSVCRT.DLL
ProductName         =3D Microsoft (R) Visual C++
ProductVersion      =3D 6.10.8528.0

Module @ 0x77f00000 =3D KERNEL32.dll
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D Windows NT BASE API Client DLL
FileVersion         =3D 4.00
InternalName        =3D kernel32
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1996
OriginalFilename    =3D kernel32
ProductName         =3D Microsoft(R) Windows NT(TM) Operating System
ProductVersion      =3D 4.00

Module @ 0x77dc0000 =3D ADVAPI32.dll
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D Advanced Windows 32 Base API
FileVersion         =3D 4.00
InternalName        =3D advapi32.dll
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1996
OriginalFilename    =3D advapi32.dll
ProductName         =3D Microsoft(R) Windows NT(TM) Operating System
ProductVersion      =3D 4.00

Module @ 0x77e70000 =3D USER32.dll
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D Windows NT USER API Client DLL
FileVersion         =3D 4.00
InternalName        =3D user32
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1996
OriginalFilename    =3D user32
ProductName         =3D Microsoft(R) Windows NT(TM) Operating System
ProductVersion      =3D 4.00

Module @ 0x77ed0000 =3D GDI32.dll
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D GDI Client DLL
FileVersion         =3D 4.00
InternalName        =3D gdi32
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1996
OriginalFilename    =3D gdi32
ProductName         =3D Microsoft(R) Windows NT(TM) Operating System
ProductVersion      =3D 4.00

Module @ 0x77e10000 =3D RPCRT4.dll
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D Remote Procedure Call Runtime
FileVersion         =3D 4.00
InternalName        =3D rpcrt4.dll
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1996
OriginalFilename    =3D rpcrt4.dll
ProductName         =3D Microsoft(R) Windows NT(TM) Operating System
ProductVersion      =3D 4.00

Module @ 0x77b20000 =3D ole32.dll
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D Microsoft OLE for Windows and Windows NT
FileVersion         =3D 4.00
InternalName        =3D OLE32.DLL
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1996
OriginalFilename    =3D OLE32.DLL
ProductName         =3D Microsoft(R) Windows NT(TM) Operating System
ProductVersion      =3D 4.00

Module @ 0x68e80000 =3D rpcref.dll
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D Microsoft Internet Information Services RPC =
helper library
FileVersion         =3D 4.02.0622
InternalName        =3D RPCREF.DLL
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1997
OriginalFilename    =3D RPCREF.DLL
ProductName         =3D Internet Information Server
ProductVersion      =3D 4.02.0622

Module @ 0x77bf0000 =3D rpcltc1.dll
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D Remote Process Control LTC1 DLL
FileVersion         =3D 4.00
InternalName        =3D rpcltc1.dll
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1996
OriginalFilename    =3D rpcltc1.dll
ProductName         =3D Microsoft(R) Windows NT(TM) Operating System
ProductVersion      =3D 4.00

Module @ 0x683a0000 =3D iisadmin.dll
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D Metadata and Admin Service
FileVersion         =3D 4.02.0720
InternalName        =3D mdadmin.dll
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1997
OriginalFilename    =3D mdadmin.dll
ProductName         =3D Internet Information Server
CDB: c0000005 Exception in inetdbg.ver debugger extension.
      PC: 77f66ebf  VA: 00078000  R/W: 1  Parameter: 1001f
0:039> !inetdbg.ver

Module @ 0x01000000 =3D inetinfo.exe
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D Internet Information Services
FileVersion         =3D 4.02.0622
InternalName        =3D INETINFO.EXE
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1997
OriginalFilename    =3D INETINFO.EXE
ProductName         =3D Internet Information Server
ProductVersion      =3D 4.02.0622

Module @ 0x77f60000 =3D ntdll.dll
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D NT Layer DLL
FileVersion         =3D 4.00
InternalName        =3D ntdll.dll
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1996
OriginalFilename    =3D ntdll.dll
ProductName         =3D Microsoft(R) Windows NT(TM) Operating System
ProductVersion      =3D 4.00

Module @ 0x78000000 =3D MSVCRT.dll
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D Microsoft (R) C Runtime Library
FileVersion         =3D 6.10.8528.0
InternalName        =3D MSVCRT.DLL
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1998
OriginalFilename    =3D MSVCRT.DLL
ProductName         =3D Microsoft (R) Visual C++
ProductVersion      =3D 6.10.8528.0

Module @ 0x77f00000 =3D KERNEL32.dll
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D Windows NT BASE API Client DLL
FileVersion         =3D 4.00
InternalName        =3D kernel32
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1996
OriginalFilename    =3D kernel32
ProductName         =3D Microsoft(R) Windows NT(TM) Operating System
ProductVersion      =3D 4.00

Module @ 0x77dc0000 =3D ADVAPI32.dll
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D Advanced Windows 32 Base API
FileVersion         =3D 4.00
InternalName        =3D advapi32.dll
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1996
OriginalFilename    =3D advapi32.dll
ProductName         =3D Microsoft(R) Windows NT(TM) Operating System
ProductVersion      =3D 4.00

Module @ 0x77e70000 =3D USER32.dll
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D Windows NT USER API Client DLL
FileVersion         =3D 4.00
InternalName        =3D user32
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1996
OriginalFilename    =3D user32
ProductName         =3D Microsoft(R) Windows NT(TM) Operating System
ProductVersion      =3D 4.00

Module @ 0x77ed0000 =3D GDI32.dll
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D GDI Client DLL
FileVersion         =3D 4.00
InternalName        =3D gdi32
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1996
OriginalFilename    =3D gdi32
ProductName         =3D Microsoft(R) Windows NT(TM) Operating System
ProductVersion      =3D 4.00

Module @ 0x77e10000 =3D RPCRT4.dll
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D Remote Procedure Call Runtime
FileVersion         =3D 4.00
InternalName        =3D rpcrt4.dll
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1996
OriginalFilename    =3D rpcrt4.dll
ProductName         =3D Microsoft(R) Windows NT(TM) Operating System
ProductVersion      =3D 4.00

Module @ 0x77b20000 =3D ole32.dll
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D Microsoft OLE for Windows and Windows NT
FileVersion         =3D 4.00
InternalName        =3D OLE32.DLL
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1996
OriginalFilename    =3D OLE32.DLL
ProductName         =3D Microsoft(R) Windows NT(TM) Operating System
ProductVersion      =3D 4.00

Module @ 0x68e80000 =3D rpcref.dll
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D Microsoft Internet Information Services RPC =
helper library
FileVersion         =3D 4.02.0622
InternalName        =3D RPCREF.DLL
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1997
OriginalFilename    =3D RPCREF.DLL
ProductName         =3D Internet Information Server
ProductVersion      =3D 4.02.0622

Module @ 0x77bf0000 =3D rpcltc1.dll
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D Remote Process Control LTC1 DLL
FileVersion         =3D 4.00
InternalName        =3D rpcltc1.dll
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1996
OriginalFilename    =3D rpcltc1.dll
ProductName         =3D Microsoft(R) Windows NT(TM) Operating System
ProductVersion      =3D 4.00

Module @ 0x683a0000 =3D iisadmin.dll
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D Metadata and Admin Service
FileVersion         =3D 4.02.0720
InternalName        =3D mdadmin.dll
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1997
OriginalFilename    =3D mdadmin.dll
ProductName         =3D Internet Information Server
ProductVersion      =3D 4.02.0720

Module @ 0x68220000 =3D COADMIN.dll
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D IIS CoAdmin DLL
FileVersion         =3D 4.02.0720
InternalName        =3D coadmin.dll
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1997
OriginalFilename    =3D coadmin.dll
ProductName         =3D Internet Information Server
ProductVersion      =3D 4.02.0720

Module @ 0x70980000 =3D SHELL32.dll
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D Windows Shell Common Dll
FileVersion         =3D 4.72.3612.1700
InternalName        =3D SHELL32
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1997
OriginalFilename    =3D SHELL32.DLL
ProductName         =3D Microsoft(R) Windows NT(R) Operating System
ProductVersion      =3D 4.72.3612.1700

Module @ 0x710d0000 =3D SHLWAPI.dll
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D Shell Light-weight Utility Library
FileVersion         =3D 4.72.3110.0
InternalName        =3D SHLWAPI
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1997
OriginalFilename    =3D SHLWAPI.DLL
ProductName         =3D Microsoft(R) Windows NT(R) Operating System
ProductVersion      =3D 4.72.3110.0

Module @ 0x77aa0000 =3D COMCTL32.dll
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D Custom Controls Library
FileVersion         =3D 4.72.3612.1702
InternalName        =3D COMMCTRL
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1999
OriginalFilename    =3D COMMCTRL.DLL
ProductName         =3D Microsoft(R) Windows NT(R) Operating System
ProductVersion      =3D 4.72.3612.1702

Module @ 0x68020000 =3D ADMWPROX.dll
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D IIS Admin Com API Proxy dll
FileVersion         =3D 4.02.0720
InternalName        =3D admwprox.dll
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1997
OriginalFilename    =3D admwprox.dll
ProductName         =3D Internet Information Server
ProductVersion      =3D 4.02.0720

Module @ 0x74880000 =3D userenv.dll
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D Userenv
FileVersion         =3D 4.00
InternalName        =3D userenv
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1996
OriginalFilename    =3D userenv.dll
ProductName         =3D Microsoft(R) Windows NT(TM) Operating System
ProductVersion      =3D 4.00

Module @ 0x77800000 =3D NETAPI32.dll
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D Net Win32 API DLL
FileVersion         =3D 4.00
InternalName        =3D NetApi32.DLL
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1996
OriginalFilename    =3D NetApi32.DLL
ProductName         =3D Microsoft(R) Windows NT(TM) Operating System
ProductVersion      =3D 4.00

Module @ 0x77840000 =3D NETRAP.dll
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D Net Remote Admin Protocol DLL
FileVersion         =3D 4.00
InternalName        =3D NetRap.DLL
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1996
OriginalFilename    =3D NetRap.DLL
ProductName         =3D Microsoft(R) Windows NT(TM) Operating System
ProductVersion      =3D 4.00

Module @ 0x777e0000 =3D SAMLIB.dll
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D SAM Library DLL
FileVersion         =3D 4.00
InternalName        =3D SAMLib.DLL
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1996
OriginalFilename    =3D SAMLib.DLL
ProductName         =3D Microsoft(R) Windows NT(TM) Operating System
ProductVersion      =3D 4.00

Module @ 0x68900000 =3D nsepm.dll
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D IIS NSEP mapping DLL
FileVersion         =3D 4.02.0720
InternalName        =3D nsepm.dll
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1997
OriginalFilename    =3D nsepm.dll
ProductName         =3D Internet Information Server
ProductVersion      =3D 4.02.0720

Module @ 0x68440000 =3D IISMAP.dll
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D Microsoft IIS mapper
FileVersion         =3D 4.02.0720
InternalName        =3D iismap.dll
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1997
OriginalFilename    =3D iismap.dll
ProductName         =3D Internet Information Server
ProductVersion      =3D 4.02.0720

Module @ 0x77400000 =3D schannel.dll
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
CDB: c0000005 Exception in inetdbg.ver debugger extension.
      PC: 77f66edd  VA: 00079000  R/W: 1  Parameter: 1001f
0:039> q

quit:

------_=_NextPart_000_01BFA65D.E540DE90
Content-Type: application/octet-stream;
	name="IIS_332-224129-413-2000.dbl"
Content-Transfer-Encoding: quoted-printable
Content-Disposition: attachment;
	filename="IIS_332-224129-413-2000.dbl"

<EMSessionType>Auto</EMSessionType><EMStartTime>4/13/00 10:41:29 =
PM</EMStartTime><EMEndTime>4/14/00 11:10:49 =
AM</EMEndTime><EMException>Manual</EMException>

log file opened
0:017> !emdbg.tag IIS Exception Monitor 6.1.0.0

Loaded emdbg extension DLL
IIS Exception Monitor 6.1.0.0
0:017> !emdbg.tag =
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D

=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D
0:017> !emdbg.tag =3D=3D Your Web Server should be running now, DON'T =
press Ctl-C Now.

=3D=3D Your Web Server should be running now, DON'T press Ctl-C Now.
0:017> !emdbg.tag =
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D

=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D
0:017> !emdbg.tag =3D=3D Wait until the application stops serving pages =
and then press Ctl-C in this window

=3D=3D Wait until the application stops serving pages and then press =
Ctl-C in this window
0:017> !emdbg.tag =3D=3D If you press Ctl-C, you can use the log to =
look for 100% CPU utilization

=3D=3D If you press Ctl-C, you can use the log to look for 100% CPU =
utilization
0:017> !emdbg.tag =3D=3D however it may not be able to tell you which =
module is causing your instability

=3D=3D however it may not be able to tell you which module is causing =
your instability
0:017> !emdbg.tag =
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D

=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D
0:017> sxd ch

0:017> g

CDB ModLoad: 68890000 688a6000   C:\WINNT\System32\inetsrv\wam.dll
CDB ModLoad: 697e0000 6984c000   C:\WINNT\System32\mtxex.dll
CDB ModLoad: 69d00000 69d1a000   MTSEvents.DLL
CDB ModLoad: 69a90000 69a9a000   C:\WINNT\System32\MTxInfr1.dll
CDB ModLoad: 69ab0000 69abd000   C:\WINNT\System32\MTxInfr2.dll
CDB ModLoad: 69620000 69630000   C:\WINNT\System32\MTxRn.DLL
CDB ModLoad: 6b550000 6b556000   C:\WINNT\System32\MTxPerf.dll
CDB ModLoad: 69a70000 69a7a000   C:\WINNT\System32\MfcSubs.dll
CDB ModLoad: 690b0000 690ce000   C:\WINNT\System32\DTCCM.dll
CDB ModLoad: 69000000 6900a000   C:\WINNT\System32\DTCUtil.dll
CDB ModLoad: 68ff0000 68ff7000   C:\WINNT\System32\DTCTRACE.dll
CDB ModLoad: 69790000 6979d000   C:\WINNT\System32\MTXCLU.DLL
CDB ModLoad: 77a90000 77a9b000   C:\WINNT\system32\VERSION.dll
CDB ModLoad: 779c0000 779c8000   C:\WINNT\system32\LZ32.dll
CDB ModLoad: 699c0000 699cc000   C:\WINNT\System32\MTXTRK.DLL
CDB ModLoad: 69360000 69369000   C:\WINNT\System32\XOLEHLP.dll
CDB ModLoad: 69120000 69133000   C:\WINNT\System32\ADME.dll
CDB ModLoad: 69a20000 69a2b000   C:\WINNT\System32\MtxDm.dll
CDB ModLoad: 69640000 69671000   C:\WINNT\System32\MtxCat.dll
CDB ModLoad: 1f7d0000 1f804000   C:\WINNT\System32\ODBC32.dll
CDB ModLoad: 77d80000 77db2000   C:\WINNT\system32\comdlg32.dll
CDB ModLoad: 7f230000 7f23d000   C:\WINNT\system32\CLUSAPI.DLL
CDB: exception number 000006ba
CDB ModLoad: 7f250000 7f25a000   C:\WINNT\system32\RESUTILS.DLL
CDB ModLoad: 1f8c0000 1f8d6000   C:\WINNT\System32\odbcint.dll
CDB ModLoad: 697c0000 697ca000   C:\WINNT\System32\Mts\mtxact.dll
CDB ModLoad: 69a40000 69a5b000   C:\WINNT\system32\mtxoci.dll
CDB ModLoad: 0e030000 0e041000   C:\Perl\bin\perlis.dll
LDR: Automatic DLL Relocation in inetinfo.exe
LDR: Dll PerlCRT.dll base 78000000 relocated due to collision with =
C:\WINNT\system32\MSVCRT.dll
CDB ModLoad: 015e0000 01628000   C:\WINNT\system32\PerlCRT.dll
LDR: Automatic DLL Relocation in inetinfo.exe
LDR: Dll PerlCore.dll base 10000000 relocated due to collision with =
C:\Program Files\WEBTREND\Plugins\IIS4\RevDNS.dll
CDB ModLoad: 01a40000 01ac7000   C:\Perl\bin\PerlCore.dll
CDB ModLoad: 680b0000 680ff000   C:\WINNT\System32\inetsrv\asp.dll
CDB ModLoad: 6b600000 6b66b000   C:\WINNT\System32\vbscript.dll
LDR: Automatic DLL Relocation in inetinfo.exe
LDR: Dll Socket.dll base 10000000 relocated due to collision with =
C:\Program Files\WEBTREND\Plugins\IIS4\RevDNS.dll
CDB ModLoad: 01c60000 01c67000   C:\Perl\lib\auto\Socket\Socket.dll
CDB ModLoad: 1e2a0000 1e2e2000   C:\WINNT\system32\pythoncom15.dll
CDB ModLoad: 1e600000 1e611000   C:\WINNT\system32\PyWinTypes15.dll
CDB ModLoad: 1e100000 1e186000   python15.dll
CDB ModLoad: 77fd0000 77ffa000   C:\WINNT\system32\WINMM.dll
CDB ModLoad: 77fc0000 77fc8000   C:\WINNT\system32\mmdrv.dll
CDB ModLoad: 1e200000 1e20e000   C:\Program =
Files\Python\win32\win32api.pyd
LDR: Automatic DLL Relocation in inetinfo.exe
LDR: Dll axscript.pyd base 1e2b0000 relocated due to collision with =
C:\WINNT\system32\pythoncom15.dll
CDB ModLoad: 02080000 02092000   C:\Program =
Files\Python\win32comext\axscript\axscript.pyd
CDB ModLoad: 1e3b0000 1e3e8000   C:\Program =
Files\Python\win32comext\axdebug\axdebug.pyd
CDB ModLoad: 68c80000 68c86000   C:\WINNT\System32\inetsrv\wamps.dll
CDB ModLoad: 68690000 68696000   C:\WINNT\System32\inetsrv\iwrps.dll
CDB ModLoad: 1e360000 1e366000   C:\Program Files\Python\win32\dbi.dll
CDB ModLoad: 1e350000 1e358000   C:\Program Files\Python\win32\odbc.pyd
CDB ModLoad: 41230000 412ae000   C:\WINNT\System32\sqlsrv32.dll
CDB ModLoad: 41100000 4110c000   C:\WINNT\System32\SQLWOA.dll
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 1f820000 1f83a000   C:\WINNT\system32\odbccp32.dll
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
LDR: Automatic DLL Relocation in inetinfo.exe
LDR: Dll IO.dll base 10000000 relocated due to collision with =
C:\Program Files\WEBTREND\Plugins\IIS4\RevDNS.dll
CDB ModLoad: 02ec0000 02ec6000   C:\Perl\lib\auto\IO\IO.dll
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB: exception number 8000d001
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB: exception number 8000d001
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB: exception number 8000d001
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 689f0000 689fb000   C:\WINNT\System32\inetsrv\ssinc.dll
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB ModLoad: 75a80000 75a87000   C:\WINNT\system32\NDDEAPI.DLL
CDB ModLoad: 77c00000 77c18000   C:\WINNT\system32\WINSPOOL.DRV
CDB ModLoad: 73310000 73318000   C:\WINNT\system32\DBNMPNTW.DLL
CDB: access violation
eax=3D5c6e6f68 ebx=3D680ec470 ecx=3D020ffd8c edx=3D7803c120 =
esi=3D0a5734f0 edi=3D00000000
eip=3D1e1422b7 esp=3D020ffd64 ebp=3D020ffdec iopl=3D0         nv up ei =
pl nz na pe nc
cs=3D001b  ss=3D0023  ds=3D0023  es=3D0023  fs=3D003b  gs=3D0000        =
     efl=3D00010202
*** WARNING: Using export symbols for python15.dll
python15!PyThreadState_Clear+0x37:
1e1422b7 8b08             mov     ecx,[eax]         =
ds:0023:5c6e6f68=3D????????
0:022> kv

*** WARNING: Using export symbols for c:\winnt\system32\pythoncom15.dll
*** WARNING: symbols checksum and timestamp is wrong 0x0005847a =
0x000512f3 for c:\winnt\system32\inetsrv\asp.dll
*** WARNING: symbols checksum and timestamp is wrong 0x000784a0 =
0x00078fb6 for c:\winnt\system32\mtxex.dll
ChildEBP RetAddr  Args to Child
020ffd68 1e142143 0a5734f0 00000000 0a6b31f0 =
python15!PyThreadState_Clear+0x37 (FPO: [1,0,2])
020ffd78 1e14267f 0a6b31f0 00000000 1e2a2401 =
python15!PyInterpreterState_Clear+0x13 (FPO: [1,0,2])
020ffd84 1e2a2401 020ffda4 1e2c4fd8 00000000 python15!Py_Finalize+0x3f =
(FPO: [0,0,1])
020ffdec 680c3b32 020ffe6c 00134748 77f050f0 =
pythoncom15!PyCom_DLLReleaseRef+0x71(FPO: [Non-Fpo]
020ffe34 680c2f38 0009f0f0 00000000 020ffec0 =
asp!CAppConfig__SetValue+0xc8(FPO: [Non-Fpo]
020ffe7c 680c78c0 0009f0f0 02216400 020ffec0 =
asp!CPerfMainBlock__Init+0x47(FPO: [Non-Fpo]
020ffecc 680e25c6 00000000 697e1898 003ce8e0 =
asp!CConnectionPoint__~CConnectionPoint+0x6 (FPO: [0,0,2])
020ffee4 697e67cc 00000000 697e1898 0a6a5c90 =
asp!_IASPObjectContextCustomStubVtbl+0x95(FPO: [Non-Fpo]
020fff24 697e57ce 02228310 697f21de 77f0178b =
mtxex!CActivity__CallWithNewContext+0xac(FPO: [Non-Fpo]
77f67610 4affc033 89257508 fff00c42 037d044a =
mtxex!CQueuedRequest__~CQueuedRequest+0x5e(FPO: [Non-Fpo]
0424548b 000003e6 0006ef54 0006ee98 00000008 0x4affc033 [Stdcall: 20]
0:022> !emdbg.sysinfo

<EMXML><SYSINFO>
	<ENVIRONMENT>
		<COMPUTER-NAME>OBERON</COMPUTER-NAME>
		<SYS-DIR>C:\WINNT\System32</SYS-DIR>
	</ENVIRONMENT>
	<OS>
		<OS-VERSION>4.0 (Build 1381) Service Pack 6</OS-VERSION>
	</OS>
	<PROCESSOR>
		<NUM-PROC>2</NUM-PROC>
		<PROC-ARCH>x86</PROC-ARCH>
		<PROC-LEVEL>Pentium Pro or PII</PROC-LEVEL>
	</PROCESSOR>
	<MEMORY>
		<PHYSICAL-MEMORY>1073135616 bytes</PHYSICAL-MEMORY>
		<PHYSICAL-MEMORY-FREE>898994176 bytes</PHYSICAL-MEMORY-FREE>
		<PAGE-FILE>1304231936 bytes</PAGE-FILE>
		<PAGE-FILE-FREE>1161953280 bytes</PAGE-FILE-FREE>
	</MEMORY>
	<DISK>
		<LOGICAL-DRIVE>
			<DRIVE-LETTER>A:\</DRIVE-LETTER>
			<DRIVE-TYPE>Removable</DRIVE-TYPE>
		</LOGICAL-DRIVE>
		<LOGICAL-DRIVE>
			<DRIVE-LETTER>C:\</DRIVE-LETTER>
			<DRIVE-TYPE>Fixed Drive</DRIVE-TYPE>
			<DRIVE-SIZE>4194860032 bytes</DRIVE-SIZE>
			<BYTES-USED>1353121792 bytes</BYTES-USED>
			<BYTES-FREE>2841738240 bytes</BYTES-FREE>
			<VOLUME-NAME></VOLUME-NAME>
			<SERIAL-NUMBER>4c8e8b33</SERIAL-NUMBER>
			<FILE-SYSTEM>NTFS</FILE-SYSTEM>
			<FILE-SYSTEM-FLAGS>
			FS_CASE_IS_PRESERVED
			FS_CASE_SENSITIVE
			FS_UNICODE_STORED_ON_DISK
			FS_PERSISTENT_ACLS
			FS_FILE_COMPRESSION
			</FILE-SYSTEM-FLAGS>
		</LOGICAL-DRIVE>
		<LOGICAL-DRIVE>
			<DRIVE-LETTER>D:\</DRIVE-LETTER>
			<DRIVE-TYPE>Fixed Drive</DRIVE-TYPE>
			<DRIVE-SIZE>36413280256 bytes</DRIVE-SIZE>
			<BYTES-USED>702562304 bytes</BYTES-USED>
			<BYTES-FREE>35710717952 bytes</BYTES-FREE>
			<VOLUME-NAME></VOLUME-NAME>
			<SERIAL-NUMBER>342d9cd8</SERIAL-NUMBER>
			<FILE-SYSTEM>NTFS</FILE-SYSTEM>
			<FILE-SYSTEM-FLAGS>
			FS_CASE_IS_PRESERVED
			FS_CASE_SENSITIVE
			FS_UNICODE_STORED_ON_DISK
			FS_PERSISTENT_ACLS
			FS_FILE_COMPRESSION
			</FILE-SYSTEM-FLAGS>
		</LOGICAL-DRIVE>
		<LOGICAL-DRIVE>
			<DRIVE-LETTER>E:\</DRIVE-LETTER>
			<DRIVE-TYPE>Fixed Drive</DRIVE-TYPE>
			<DRIVE-SIZE>36413280256 bytes</DRIVE-SIZE>
			<BYTES-USED>9957376 bytes</BYTES-USED>
			<BYTES-FREE>36403322880 bytes</BYTES-FREE>
			<VOLUME-NAME></VOLUME-NAME>
			<SERIAL-NUMBER>50397a26</SERIAL-NUMBER>
			<FILE-SYSTEM>NTFS</FILE-SYSTEM>
			<FILE-SYSTEM-FLAGS>
			FS_CASE_IS_PRESERVED
			FS_CASE_SENSITIVE
			FS_UNICODE_STORED_ON_DISK
			FS_PERSISTENT_ACLS
			FS_FILE_COMPRESSION
			</FILE-SYSTEM-FLAGS>
		</LOGICAL-DRIVE>
		<LOGICAL-DRIVE>
			<DRIVE-LETTER>F:\</DRIVE-LETTER>
			<DRIVE-TYPE>Removable</DRIVE-TYPE>
		</LOGICAL-DRIVE>
		<LOGICAL-DRIVE>
			<DRIVE-LETTER>G:\</DRIVE-LETTER>
			<DRIVE-TYPE>CDROM</DRIVE-TYPE>
		</LOGICAL-DRIVE>
		<LOGICAL-DRIVE>
			<DRIVE-LETTER>L:\</DRIVE-LETTER>
			<DRIVE-TYPE>Fixed Drive</DRIVE-TYPE>
			<DRIVE-SIZE>4902232064 bytes</DRIVE-SIZE>
			<BYTES-USED>1180934144 bytes</BYTES-USED>
			<BYTES-FREE>3721297920 bytes</BYTES-FREE>
			<VOLUME-NAME></VOLUME-NAME>
			<SERIAL-NUMBER>70239c30</SERIAL-NUMBER>
			<FILE-SYSTEM>NTFS</FILE-SYSTEM>
			<FILE-SYSTEM-FLAGS>
			FS_CASE_IS_PRESERVED
			FS_CASE_SENSITIVE
			FS_UNICODE_STORED_ON_DISK
			FS_PERSISTENT_ACLS
			FS_FILE_COMPRESSION
			</FILE-SYSTEM-FLAGS>
		</LOGICAL-DRIVE>
	</DISK>
</SYSINFO></EMXML>

0:022> !inetdbg.mod

Loaded inetdbg extension DLL
Start    End      Entry    Path
01000000 01005000 01001d80 C:\WINNT\System32\inetsrv\inetinfo.exe
77f60000 77fbe000 00000000 C:\WINNT\System32\ntdll.dll
78000000 78045000 7800256b C:\WINNT\system32\MSVCRT.dll
77f00000 77f5e000 77f01000 C:\WINNT\system32\KERNEL32.dll
77dc0000 77dff000 77dc1000 C:\WINNT\system32\ADVAPI32.dll
77e70000 77ec5000 77e781d7 C:\WINNT\system32\USER32.dll
77ed0000 77efc000 00000000 C:\WINNT\system32\GDI32.dll
77e10000 77e67000 77e186bf C:\WINNT\system32\RPCRT4.dll
77b20000 77bd7000 77b24d6e C:\WINNT\system32\ole32.dll
68e80000 68e86000 68e810bc C:\WINNT\System32\inetsrv\rpcref.dll
77bf0000 77bf7000 00000000 C:\WINNT\system32\rpcltc1.dll
683a0000 683a7000 683a15a8 C:\WINNT\System32\inetsrv\iisadmin.dll
68220000 6822c000 682269e0 C:\WINNT\System32\inetsrv\COADMIN.dll
70980000 70b28000 709868fb C:\WINNT\system32\SHELL32.dll
710d0000 710f3000 710d14ce C:\WINNT\system32\SHLWAPI.dll
77aa0000 77b15000 77aa1b4d C:\WINNT\system32\COMCTL32.dll
68020000 6802a000 68022a91 C:\WINNT\system32\ADMWPROX.dll
74880000 74894000 74881000 C:\WINNT\system32\userenv.dll
77800000 7783a000 77801000 C:\WINNT\system32\NETAPI32.dll
77840000 77849000 00000000 C:\WINNT\system32\NETRAP.dll
777e0000 777ed000 00000000 C:\WINNT\system32\SAMLIB.dll
68900000 6890d000 68906180 C:\WINNT\System32\inetsrv\nsepm.dll
68440000 68450000 68448800 C:\WINNT\system32\IISMAP.dll
77400000 77429000 7740382b C:\WINNT\system32\schannel.dll
5e380000 5e3a5000 5e3a0920 C:\WINNT\system32\MSOSS.dll
776d0000 776d8000 00000000 C:\WINNT\system32\WSOCK32.dll
776b0000 776c4000 776b36b5 C:\WINNT\system32\WS2_32.dll
776a0000 776a7000 776a110e C:\WINNT\system32\WS2HELP.dll
5cf00000 5cf5e000 5cf06b44 C:\WINNT\system32\CRYPT32.dll
68df0000 68e03000 68df2b1f C:\WINNT\System32\inetsrv\metadata.dll
68c90000 68c9b000 68c91f80 C:\WINNT\System32\inetsrv\wamreg.dll
5f400000 5f4f2000 5f4065ff C:\WINNT\system32\MFC42.DLL
65340000 653d2000 6534ee21 C:\WINNT\system32\OLEAUT32.dll
7ca00000 7ca1b000 7ca0b2cc C:\WINNT\system32\rsabase.dll
5a800000 5a80f000 5a802004 C:\WINNT\system32\pstorec.dll
68db0000 68db7000 68db2b10 C:\WINNT\System32\inetsrv\admexs.dll
68ea0000 68ea7000 68ea2350 C:\WINNT\System32\inetsrv\svcext.dll
68c00000 68c39000 68c1c344 C:\WINNT\System32\inetsrv\w3svc.dll
68590000 685bf000 68594a93 C:\WINNT\System32\inetsrv\INFOCOMM.dll
68620000 6862e000 68625f71 C:\WINNT\System32\inetsrv\ISATQ.dll
682b0000 682b6000 682b17b4 C:\WINNT\System32\inetsrv\IISFECNV.dll
68560000 68568000 68562f47 C:\WINNT\system32\inetsloc.dll
77670000 77685000 77672599 C:\WINNT\system32\MSWSOCK.dll
76e70000 76e82000 76e71000 C:\WINNT\system32\security.dll
68820000 68827000 6882132a C:\WINNT\System32\inetsrv\lonsint.dll
47600000 4760e000 47608a10 C:\WINNT\system32\wintrust.dll
74fa0000 74fab000 00000000 C:\WINNT\system32\RpcLtScm.Dll
77660000 7766f000 77663bc8 C:\WINNT\system32\msafd.dll
77690000 77699000 77691000 C:\WINNT\System32\wshtcpip.dll
74ff0000 74ffe000 74ff7b68 C:\WINNT\System32\rnr20.dll
60000000 60011000 60006440 C:\WINNT\system32\msapsspc.dll
779d0000 779e5000 779d1000 C:\WINNT\system32\MSVCRT40.dll
780a0000 780b2000 780a1000 C:\WINNT\system32\MSVCIRT.dll
01120000 0113f000 011252c0 C:\WINNT\system32\msnsspc.dll
77e00000 77e06000 00000000 C:\WINNT\system32\rpclts1.dll
74fc0000 74fcf000 00000000 C:\WINNT\system32\RpcLtCcm.Dll
68650000 68659000 686519e8 C:\WINNT\System32\inetsrv\iscomlog.dll
68a10000 68a19000 68a126d6 C:\WINNT\System32\inetsrv\sspifilt.dll
67860000 67868000 678613e0 c:\program files\microsoft =
frontpage\bin\fpexedll.dll
67ec0000 67ec5000 67ec1460 C:\Program Files\Microsoft =
FrontPage\version3.0\bin\fpexedll.dll
74480000 74488000 00000000 C:\WINNT\System32\wshisn.dll
75830000 7585a000 7583636b C:\WINNT\System32\nwprovau.dll
77720000 77731000 777278db C:\WINNT\system32\MPR.dll
74470000 74477000 74471000 C:\WINNT\system32\wshnetbs.dll
68410000 68421000 684160c8 C:\WINNT\System32\inetsrv\iislog.dll
10000000 10008000 100025f0 C:\Program =
Files\WEBTREND\Plugins\IIS4\RevDNS.dll
013c0000 013c8000 013c27a0 C:\Program =
Files\WEBTREND\Plugins\IIS4\wtCookie.dll
68890000 688a6000 6889244e C:\WINNT\System32\inetsrv\wam.dll
697e0000 6984c000 6981c900 C:\WINNT\System32\mtxex.dll
69d00000 69d1a000 69d10490 C:\WINNT\System32\MTSEvents.DLL
69a90000 69a9a000 69a94690 C:\WINNT\System32\MTxInfr1.dll
69ab0000 69abd000 69ab5cf0 C:\WINNT\System32\MTxInfr2.dll
69620000 69630000 696289a0 C:\WINNT\System32\MTxRn.DLL
6b550000 6b556000 6b551480 C:\WINNT\System32\MTxPerf.dll
69a70000 69a7a000 69a74220 C:\WINNT\System32\MfcSubs.dll
690b0000 690ce000 690c5a00 C:\WINNT\System32\DTCCM.dll
69000000 6900a000 69004180 C:\WINNT\System32\DTCUtil.dll
68ff0000 68ff7000 68ff1420 C:\WINNT\System32\DTCTRACE.dll
69790000 6979d000 69798460 C:\WINNT\System32\MTXCLU.DLL
77a90000 77a9b000 77a92fd0 C:\WINNT\system32\VERSION.dll
779c0000 779c8000 779c1881 C:\WINNT\system32\LZ32.dll
699c0000 699cc000 699c4f00 C:\WINNT\System32\MTXTRK.DLL
69360000 69369000 69361d40 C:\WINNT\System32\XOLEHLP.dll
69120000 69133000 6912d300 C:\WINNT\System32\ADME.dll
69a20000 69a2b000 69a24a30 C:\WINNT\System32\MtxDm.dll
69640000 69671000 69656720 C:\WINNT\System32\MtxCat.dll
1f7d0000 1f804000 1f7d12d4 C:\WINNT\System32\ODBC32.dll
77d80000 77db2000 77d81000 C:\WINNT\system32\comdlg32.dll
7f230000 7f23d000 00000000 C:\WINNT\system32\CLUSAPI.DLL
7f250000 7f25a000 7f2516f8 C:\WINNT\system32\RESUTILS.DLL
1f8c0000 1f8d6000 00000000 C:\WINNT\System32\odbcint.dll
697c0000 697ca000 697c4d10 C:\WINNT\System32\Mts\mtxact.dll
69a40000 69a5b000 69a4f730 C:\WINNT\system32\mtxoci.dll
0e030000 0e041000 0e039d30 C:\Perl\bin\perlis.dll
015e0000 01628000 015e9010 C:\WINNT\system32\PerlCRT.dll
01a40000 01ac7000 01aac040 C:\Perl\bin\PerlCore.dll
680b0000 680ff000 680e2c80 C:\WINNT\System32\inetsrv\asp.dll
6b600000 6b66b000 6b605359 C:\WINNT\System32\vbscript.dll
01c60000 01c67000 01c63c80 C:\Perl\lib\auto\Socket\Socket.dll
1e2a0000 1e2e2000 1e2c4d6d C:\WINNT\system32\pythoncom15.dll
1e600000 1e611000 1e606f5a C:\WINNT\system32\PyWinTypes15.dll
1e100000 1e186000 1e1583c0 C:\WINNT\system32\python15.dll
77fd0000 77ffa000 77fd5640 C:\WINNT\system32\WINMM.dll
1e200000 1e20e000 1e206fa3 C:\Program Files\Python\win32\win32api.pyd
02080000 02092000 02087a18 C:\Program =
Files\Python\win32comext\axscript\axscript.pyd
1e3b0000 1e3e8000 1e3d1b6f C:\Program =
Files\Python\win32comext\axdebug\axdebug.pyd
68c80000 68c86000 68c81510 C:\WINNT\System32\inetsrv\wamps.dll
68690000 68696000 68691970 C:\WINNT\System32\inetsrv\iwrps.dll
1e360000 1e366000 1e361855 C:\Program Files\Python\win32\dbi.dll
1e350000 1e358000 1e3534c5 C:\Program Files\Python\win32\odbc.pyd
41230000 412ae000 41231000 C:\WINNT\System32\sqlsrv32.dll
41100000 4110c000 411061e0 C:\WINNT\System32\SQLWOA.dll
1f820000 1f83a000 1f82c63c C:\WINNT\system32\odbccp32.dll
02ec0000 02ec6000 02ec2610 C:\Perl\lib\auto\IO\IO.dll
689f0000 689fb000 689f3b28 C:\WINNT\System32\inetsrv\ssinc.dll
0:022> ~*kb 300


  0  id: 14c.128   Suspend: 0 Teb 7ffde000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
*** WARNING: Using export symbols for advapi32.dll
*** WARNING: Using export symbols for kernel32.dll
ChildEBP RetAddr  Args to Child
0006fdb0 77dd8a35 00000050 00073880 0000021a ntdll!ZwReadFile+0xb
0006fde4 77dd8481 00000050 00073880 0000021a =
ADVAPI32!RegisterServiceCtrlHandlerA+0x352
0006fe48 77dd829e 00000050 00073880 0000021a =
ADVAPI32!StartServiceCtrlDispatcherW+0x184
0006fe6c 0100182c 00073278 68e80000 77f13fb3 =
ADVAPI32!StartServiceCtrlDispatcherA+0x82
0006fea4 010015d5 00000000 00000000 7ffdf000 =
inetinfo!StartDispatchTable+0x61
0006ff70 01001e7f 00000001 00240550 002401a0 inetinfo!main+0x2a5
0006ffc0 77f1b9ea 00000000 00000000 7ffdf000 =
inetinfo!mainCRTStartup+0xff
0006fff0 00000000 01001d80 00000000 000000b0 =
KERNEL32!GetProcessPriorityBoost+0x117

  1  id: 14c.ac   Suspend: 0 Teb 7ffdd000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
0087fe54 77f04f37 000000b4 ffffffff 00000000 =
ntdll!NtWaitForSingleObject+0xb
00000001 000003e6 00000088 77f04f37 0000000f =
KERNEL32!WaitForSingleObject+0xf

  2  id: 14c.d4   Suspend: 0 Teb 7ffdc000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
*** WARNING: Using export symbols for user32.dll
ChildEBP RetAddr  Args to Child
0252feb4 77e79de5 00000003 0252fedc 00000000 =
ntdll!NtWaitForMultipleObjects+0xb
0252ff10 77e79e31 00000002 0252ff58 000003e8 =
USER32!MsgWaitForMultipleObjectsEx+0xa8
0252ff2c 697f1dbe 00000002 0252ff58 00000000 =
USER32!MsgWaitForMultipleObjects+0x1d
69a92ea0 750169a9 24448b22 244c8b14 24548b10 =
mtxex!ContextWrapper__Recycle+0xe
60b03d83 000003e6 0006ef1c 0006ee60 00000008 0x750169a9

  3  id: 14c.102   Suspend: 0 Teb 7ffdb000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
*** WARNING: symbols checksum and timestamp is wrong 0x0003b1c7 =
0x0003704a for infocomm.dll
ChildEBP RetAddr  Args to Child
00e4fd80 77e79de5 00000002 00e4fda8 00000000 =
ntdll!NtWaitForMultipleObjects+0xb
00e4fddc 77e79e31 00000001 0023e864 ffffffff =
USER32!MsgWaitForMultipleObjectsEx+0xa8
00e4fdf8 68599007 00000001 0023e864 00000000 =
USER32!MsgWaitForMultipleObjects+0x1d
0023e864 00000000 00000000 00000800 68c09c95 =
INFOCOMM!INetpCreateSecurityDescriptor+0x1cf

  4  id: 14c.eb   Suspend: 0 Teb 7ffda000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
*** WARNING: symbols checksum and timestamp is wrong 0x00013a0e =
0x00011cc1 for isatq.dll
ChildEBP RetAddr  Args to Child
00e8fef4 77e79de5 00000002 00e8ff1c 00000000 =
ntdll!NtWaitForMultipleObjects+0xb
00e8ff50 77e79e31 00000001 68629668 00003ea0 =
USER32!MsgWaitForMultipleObjectsEx+0xa8
00e8ff6c 68625cce 00000001 68629668 00000000 =
USER32!MsgWaitForMultipleObjects+0x1d
00000001 000003e6 00000088 68622913 00000000 =
ISATQ!W95CPORT__PrepareDescriptorArrays+0x58

  5  id: 14c.df   Suspend: 0 Teb 7ffd9000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
00ecfef4 77e79de5 00000002 00ecff1c 00000000 =
ntdll!NtWaitForMultipleObjects+0xb
00ecff50 77e79e31 00000001 68629668 00003ea0 =
USER32!MsgWaitForMultipleObjectsEx+0xa8
00ecff6c 68625cce 00000001 68629668 00000000 =
USER32!MsgWaitForMultipleObjects+0x1d
00000001 000003e6 00000088 68622913 00000000 =
ISATQ!W95CPORT__PrepareDescriptorArrays+0x58

  6  id: 14c.165   Suspend: 0 Teb 7ffd8000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
00f4ff84 686248e5 000000d8 00f4ffac 00f4ffb4 =
ntdll!ZwRemoveIoCompletion+0xb
00f4ffb8 77f04ede abcdef01 00e4fc8c 00e4fcf0 ISATQ!AtqSetInfo+0xbb
00f4ffec 00000000 686248a8 abcdef01 00000000 KERNEL32!lstrcmpiW+0xbe

  7  id: 14c.108   Suspend: 0 Teb 7ffd7000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
00f8ff84 686248e5 000000d8 00f8ffac 00f8ffb4 =
ntdll!ZwRemoveIoCompletion+0xb
00f8ffb8 77f04ede abcdef01 00e4fc8c 00e4fcf0 ISATQ!AtqSetInfo+0xbb
00f8ffec 00000000 686248a8 abcdef01 00000000 KERNEL32!lstrcmpiW+0xbe

  8  id: 14c.157   Suspend: 0 Teb 7ffd6000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
00fcff88 685a75d5 00000002 00fcffb4 00000000 =
ntdll!NtWaitForMultipleObjects+0xb

  9  id: 14c.177   Suspend: 0 Teb 7ffd5000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
*** WARNING: Using export symbols for rpcrt4.dll
ChildEBP RetAddr  Args to Child
0104ff90 77e15a1d 77e160f7 000afab0 0104ffec =
ntdll!ZwReplyWaitReceivePort+0xb
00003a98 000003e6 00000088 77e15a1d 00000024 RPCRT4!NdrOleAllocate+0x24

 10  id: 14c.155   Suspend: 0 Teb 7ffd4000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
*** WARNING: Using export symbols for msafd.dll
*** WARNING: Using export symbols for ws2_32.dll
*** WARNING: Using export symbols for rpcltscm.dll
ChildEBP RetAddr  Args to Child
0110fd78 77664a12 00000158 00000140 00000001 =
ntdll!NtWaitForSingleObject+0xb
0110fea0 776b9f5f 00000000 00084050 00000000 MSAFD+0x4a12
0110fef0 74fa21e2 00000000 00084050 00000000 WS2_32!select+0xfc
0110ff2c 77e1b0dc 000b01c0 0110ff58 0110ff88 RPCLTSCM+0x21e2
0110ff60 77e1cb96 0110ff8c 0110ff88 0110ff84 =
RPCRT4!I_RpcTransServerUnprotectThread+0x48f
0110ff90 77e1aa95 77e160f7 000b00e0 0110ffec =
RPCRT4!I_RpcTransServerProtectThread+0xcc
00003a98 000003e6 00000088 77e1aa95 000000c8 =
RPCRT4!I_RpcTransServerReallocBuffer+0xc8

 11  id: 14c.172   Suspend: 0 Teb 7ffaf000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
0117ff2c 77e1b0dc 00087c78 0117ff58 0117ff88 =
ntdll!NtWaitForMultipleObjects+0xb
0117ff60 77e1cb96 0117ff8c 0117ff88 0117ff84 =
RPCRT4!I_RpcTransServerUnprotectThread+0x48f
0117ff90 77e1aa95 77e160f7 00087b98 0117ffec =
RPCRT4!I_RpcTransServerProtectThread+0xcc
00003a98 000003e6 00000088 77e1aa95 000000c8 =
RPCRT4!I_RpcTransServerReallocBuffer+0xc8

 12  id: 14c.111   Suspend: 0 Teb 7ffae000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
*** WARNING: symbols checksum and timestamp is wrong 0x0003e35a =
0x00039d63 for w3svc.dll
ChildEBP RetAddr  Args to Child
011bfec8 77e79de5 00000002 011bfef0 00000000 =
ntdll!NtWaitForMultipleObjects+0xb
011bff24 77e79e31 00000001 002812ac ffffffff =
USER32!MsgWaitForMultipleObjectsEx+0xa8
011bff40 68c243f1 00000001 002812ac 00000000 =
USER32!MsgWaitForMultipleObjects+0x1d

 13  id: 14c.116   Suspend: 0 Teb 7ffad000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
011fff00 77e79de5 00000002 011fff28 00000000 =
ntdll!NtWaitForMultipleObjects+0xb
011fff5c 77e79e31 00000001 68c2c25c ffffffff =
USER32!MsgWaitForMultipleObjectsEx+0xa8
011fff78 68c1c55e 00000001 68c2c25c 00000000 =
USER32!MsgWaitForMultipleObjects+0x1d
011fffb8 77f04ede 00000000 10231850 00e4fd68 =
W3SVC!HTTP_FILTER__ParseSendHeaders+0x46
011fffec 00000000 68c1c52a 00000000 00000000 KERNEL32!lstrcmpiW+0xbe

 14  id: 14c.118   Suspend: 0 Teb 7ffac000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
0123ff64 77f1cebe 0000ea60 00000000 7583a68a ntdll!NtDelayExecution+0xb
0123ffb8 77f04ede 00000000 00000001 00000000 KERNEL32!Sleep+0xb
0123ffec 00000000 7583a65e 00000000 00000000 KERNEL32!lstrcmpiW+0xbe

 15  id: 14c.112   Suspend: 0 Teb 7ffab000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
*** WARNING: symbols checksum and timestamp is wrong 0x000077d7 =
0x0000908e for inetsloc.dll
ChildEBP RetAddr  Args to Child
0127fcf8 77664a12 00000204 000001b0 00000001 =
ntdll!NtWaitForSingleObject+0xb
0127fe20 776b9f5f 00000000 0127fe98 00000000 MSAFD+0x4a12
0127fe70 68563c07 00000000 0127fe98 00000000 WS2_32!select+0xfc
0127ffb8 77f04ede 00000000 00000000 00000000 =
INETSLOC!SocketListenThread+0x51
0127ffec 00000000 68563bb6 00000000 00000000 KERNEL32!lstrcmpiW+0xbe

 16  id: 14c.11e   Suspend: 0 Teb 7ffaa000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
*** WARNING: Using export symbols for revdns.dll
ChildEBP RetAddr  Args to Child
013bfba4 10001a2b 013bfd70 00000000 00000000 USER32!InvalidateRect+0x19
013bffb8 77f04ede 00000000 00070860 00070000 =
REVDNS!CreateConfigTab+0x84e
013bffec 00000000 10001755 00000000 00000000 KERNEL32!lstrcmpiW+0xbe

 17  id: 14c.10b   Suspend: 0 Teb 7ffa9000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
0275ff84 686248e5 000000d8 0275ffac 0275ffb4 =
ntdll!ZwRemoveIoCompletion+0xb
0275ffb8 77f04ede 00000000 00000000 00000000 ISATQ!AtqSetInfo+0xbb
0275ffec 00000000 686248a8 00000000 00000000 KERNEL32!lstrcmpiW+0xbe

 18  id: 14c.e2   Suspend: 0 Teb 7ffa8000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
0159ff84 686248e5 000000d8 0159ffac 0159ffb4 =
ntdll!ZwRemoveIoCompletion+0xb
0159ffb8 77f04ede 00000000 00000000 00000000 ISATQ!AtqSetInfo+0xbb
0159ffec 00000000 686248a8 00000000 00000000 KERNEL32!lstrcmpiW+0xbe

 19  id: 14c.c2   Suspend: 0 Teb 7ffa7000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
015dff80 77f04f37 00000874 00002710 00000000 =
ntdll!NtWaitForSingleObject+0xb
00000000 000003e6 00000088 77f04f37 0000000f =
KERNEL32!WaitForSingleObject+0xf

 20  id: 14c.de   Suspend: 0 Teb 7ffa6000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
*** WARNING: Using export symbols for msvcrt.dll
ChildEBP RetAddr  Args to Child
01c5fec8 77e79de5 00000002 01c5fef0 00000000 =
ntdll!NtWaitForMultipleObjects+0xb
01c5ff24 77e79e31 00000001 680ebddc ffffffff =
USER32!MsgWaitForMultipleObjectsEx+0xa8
01c5ff40 680ca30c 00000001 680ebddc 00000000 =
USER32!MsgWaitForMultipleObjects+0x1d
01c5ff84 7800be50 680ebdc0 00000008 0000021c =
asp!CCookieSupportErr__`vftable'+0xb
01c5ffb8 77f04ede 00292950 00000008 0000021c MSVCRT!endthread+0x96
01c5ffec 00000000 7800bdf9 00292950 00000000 KERNEL32!lstrcmpiW+0xbe

 21  id: 14c.f0   Suspend: 0 Teb 7ffa5000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
01e6feb4 77e79de5 00000003 01e6fedc 00000000 =
ntdll!NtWaitForMultipleObjects+0xb
01e6ff10 77e79e31 00000002 01e6ff58 000003e8 =
USER32!MsgWaitForMultipleObjectsEx+0xa8
01e6ff2c 697f1dbe 00000002 01e6ff58 00000000 =
USER32!MsgWaitForMultipleObjects+0x1d
69a92ea0 750169a9 24448b22 244c8b14 24548b10 =
mtxex!ContextWrapper__Recycle+0xe
60b03d83 000003e6 0006ef1c 0006ee60 00000008 0x750169a9

 22  id: 14c.f2   Suspend: 0 Teb 7ffa4000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
020ffd68 1e142143 0a5734f0 00000000 0a6b31f0 =
python15!PyThreadState_Clear+0x37
020ffd78 1e14267f 0a6b31f0 00000000 1e2a2401 =
python15!PyInterpreterState_Clear+0x13
020ffd84 1e2a2401 020ffda4 1e2c4fd8 00000000 python15!Py_Finalize+0x3f
020ffdec 680c3b32 020ffe6c 00134748 77f050f0 =
pythoncom15!PyCom_DLLReleaseRef+0x71
020ffe34 680c2f38 0009f0f0 00000000 020ffec0 =
asp!CAppConfig__SetValue+0xc8
020ffe7c 680c78c0 0009f0f0 02216400 020ffec0 =
asp!CPerfMainBlock__Init+0x47
020ffecc 680e25c6 00000000 697e1898 003ce8e0 =
asp!CConnectionPoint__~CConnectionPoint+0x6
020ffee4 697e67cc 00000000 697e1898 0a6a5c90 =
asp!_IASPObjectContextCustomStubVtbl+0x95
020fff24 697e57ce 02228310 697f21de 77f0178b =
mtxex!CActivity__CallWithNewContext+0xac
77f67610 4affc033 89257508 fff00c42 037d044a =
mtxex!CQueuedRequest__~CQueuedRequest+0x5e
0424548b 000003e6 0006ef1c 0006ee60 00000008 0x4affc033

 23  id: 14c.f4   Suspend: 0 Teb 7ffa3000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
0214ff88 77f04f37 00000994 ffffffff 00000000 =
ntdll!NtWaitForSingleObject+0xb
0214ffb8 77f04ede 0015d260 ffffffff 00f4f234 =
KERNEL32!WaitForSingleObject+0xf
0214ffec 00000000 776b1bdd 0015d260 00000000 KERNEL32!lstrcmpiW+0xbe

 25  id: 14c.dc   Suspend: 0 Teb 7ffa1000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
0314ff84 686248e5 000000d8 0314ffac 0314ffb4 =
ntdll!ZwRemoveIoCompletion+0xb
0314ffb8 77f04ede 00000000 00000000 00000000 ISATQ!AtqSetInfo+0xbb
0314ffec 00000000 686248a8 00000000 00000000 KERNEL32!lstrcmpiW+0xbe

 26  id: 14c.115   Suspend: 0 Teb 7ffa0000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
025aff84 686248e5 000000d8 025affac 025affb4 =
ntdll!ZwRemoveIoCompletion+0xb
025affb8 77f04ede 00000000 ffffffff 00f4ff28 ISATQ!AtqSetInfo+0xbb
025affec 00000000 686248a8 00000000 00000000 KERNEL32!lstrcmpiW+0xbe

 27  id: 14c.113   Suspend: 0 Teb 7ff9f000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
025fff84 686248e5 000000d8 025fffac 025fffb4 =
ntdll!ZwRemoveIoCompletion+0xb
025fffb8 77f04ede 00000000 00000000 00000000 ISATQ!AtqSetInfo+0xbb
025fffec 00000000 686248a8 00000000 00000000 KERNEL32!lstrcmpiW+0xbe

 28  id: 14c.ff   Suspend: 0 Teb 7ff9e000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
0263ff84 686248e5 000000d8 0263ffac 0263ffb4 =
ntdll!ZwRemoveIoCompletion+0xb
0263ffb8 77f04ede 00000000 00000000 00000000 ISATQ!AtqSetInfo+0xbb
0263ffec 00000000 686248a8 00000000 00000000 KERNEL32!lstrcmpiW+0xbe

 29  id: 14c.132   Suspend: 0 Teb 7ff9d000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
0279ff84 686248e5 000000d8 0279ffac 0279ffb4 =
ntdll!ZwRemoveIoCompletion+0xb
0279ffb8 77f04ede 00000000 00000000 00000000 ISATQ!AtqSetInfo+0xbb
0279ffec 00000000 686248a8 00000000 00000000 KERNEL32!lstrcmpiW+0xbe

 30  id: 14c.127   Suspend: 0 Teb 7ff9c000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
0b89ff90 77e15a1d 77e160f7 0007f878 0b89ffec =
ntdll!ZwReplyWaitReceivePort+0xb
00003a98 000003e6 00000088 77e15a1d 00000024 RPCRT4!NdrOleAllocate+0x24

 31  id: 14c.dd   Suspend: 0 Teb 7ff9b000 Unfrozen
Stack frame count of 0x300 is too high.  Changing to 500.
ChildEBP RetAddr  Args to Child
0b92ff90 77e15a1d 77e160f7 0007f878 0b92ffec =
ntdll!ZwReplyWaitReceivePort+0xb
00003a98 000003e6 00000088 77e15a1d 00000024 RPCRT4!NdrOleAllocate+0x24
0:022> !locks

0:022> !inetdbg.atq -g

Atq Globals:
	isatq!g_cConcurrency          =3D        1 (0x       1)
	isatq!g_cThreads              =3D        2 (0x       2)
	isatq!g_cAvailableThreads     =3D        0 (0x       0)
	isatq!g_cMaxThreads           =3D        0 (0x       0)

	isatq!g_fUseAcceptEx          =3D        0 (0x       0)
	isatq!g_cbXmitBufferSize      =3D        0 (0x       0)
	isatq!g_cbMinKbSec            =3D        0 (0x       0)
	isatq!g_cCPU                  =3D        0 (0x       0)
	isatq!g_fShutdown             =3D        2 (0x       2)

	isatq!g_msThreadTimeout       =3D        0 (0x       0)
	isatq!g_dwTimeoutCookie       =3D        0 (0x       0)
	isatq!g_cListenBacklog        =3D        0 (0x       0)
	isatq!AtqGlobalContextCount   =3D        0 (0x       0)
	sizeof(ATQ_CONTEXT) =3D 212
0:022> !inetdbg.asp -g

Asp Globals:
	asp!g_nSessions                 =3D        0 (0x       0)
	asp!g_nApplications             =3D        0 (0x       0)
	asp!g_nApplicationsRestarting   =3D        0 (0x       0)
	asp!g_nBrowserRequests          =3D        0 (0x       0)
	asp!g_nSessionCleanupRequests   =3D        0 (0x       0)
	asp!g_nApplnCleanupRequests     =3D        0 (0x       0)
	asp!g_fShutDownInProgress       =3D 1868967740 (0x6f662f3c)
0:022> !inetdbg.ds -v esp

020ffd6c : 1e142143 : python15!PyInterpreterState_Clear+0x13
020ffd7c : 1e14267f : python15!Py_Finalize+0x3f
020ffd88 : 1e2a2401 : pythoncom15!PyCom_DLLReleaseRef+0x71
020ffd90 : 1e2c4fd8 : pythoncom15!PyITypeComp__Bind+0x1c28
020ffd98 : 1e2a6795 : pythoncom15!PyGatewayBase__~PyGatewayBase+0xa5
020ffda8 : 1e2c5108 : pythoncom15!PyITypeComp__Bind+0x1d58
020ffdb0 : 1e2a6a41 : pythoncom15!PyGatewayBase__Release+0x21
*** WARNING: Using export symbols for c:\program =
files\python\win32comext\axscript\axscript.pyd
020ffdbc : 0208245b : axscript!PyGActiveScript__Release+0xb
020ffdc4 : 680d3924 : asp!STR__Copy+0x17
020ffdcc : 680ec330 : asp!g_szErrTemplate+0x17f
020ffdd4 : 680d50cd : asp!CCertRequest__ParseRDNS+0x43
020ffdf0 : 680c3b32 : asp!CAppConfig__SetValue+0xc8
020ffdfc : 77f050f0 : KERNEL32!InterlockedIncrement
020ffe04 : 680ebe20 : asp!g_PerfMain+0x7
020ffe08 : 680c31e0 : asp!CPerfMainBlock__AddProcess+0x49
020ffe24 : 680c3b58 : asp!CAppConfig__SetValue+0x11a
020ffe2c : 680ec6d0 : asp!g_szErrTemplate+0x51f
020ffe38 : 680c2f38 : asp!CPerfMainBlock__Init+0x47
020ffe4c : 77f050f0 : KERNEL32!InterlockedIncrement
020ffe54 : 680ebe20 : asp!g_PerfMain+0x7
020ffe80 : 680c78c0 : asp!CConnectionPoint__~CConnectionPoint+0x6
020ffed0 : 680e25c6 : asp!_IASPObjectContextCustomStubVtbl+0x95
020ffed8 : 697e1898 : mtxex!IID_IRegNode+0x8
020ffee8 : 697e67cc : mtxex!CActivity__CallWithNewContext+0xac
020ffef0 : 697e1898 : mtxex!IID_IRegNode+0x8
020fff18 : 6981ca16 : mtxex!CAspSecurity__Release+0x1d6
020fff1c : 697e18b0 : mtxex!IID_IRegNodeDispenser
020fff24 : 77f67610 : ntdll!RtlLeaveCriticalSection
020fff28 : 697e57ce : mtxex!CQueuedRequest__~CQueuedRequest+0x5e
020fff30 : 697f21de : mtxex!ContextWrapper__QueryInterface+0x352
020fff34 : 77f0178b : KERNEL32!GetLastError
020fff3c : 69a92ea0 : MTxInfr1!VipThunkPeekMessageW
020fff44 : 697f1ed1 : mtxex!ContextWrapper__QueryInterface+0x45
020fff50 : 697e21f8 : mtxex!IID_IJavaObjectPrivate
020fff7c : 697f22d2 : mtxex!ContextWrapper__Release+0x6e
020fff88 : 7800be50 : MSVCRT!endthread+0x96
*** WARNING: Using export symbols for ole32.dll
020fff94 : 77b2aa4d : ole32!CreateBindCtx+0x2d10
020fffac : 7800ef57 : MSVCRT!except_handler3
020fffb0 : 78033200 : MSVCRT!__non_rtti_object__`vftable'+0x33c
020fffbc : 77f04ede : KERNEL32!lstrcmpiW+0xbe
020fffc8 : 77b2aa4d : ole32!CreateBindCtx+0x2d10
020fffd0 : 77b2aa4d : ole32!CreateBindCtx+0x2d10
020fffd8 : 77bc7d60 : ole32+0xa7d60
020fffe0 : 77f3b944 : KERNEL32!GetDateFormatW+0x411f
020fffe4 : 77f3d238 : KERNEL32!GetDateFormatW+0x5a13
020ffff4 : 7800bdf9 : MSVCRT!endthread+0x3f
inetdbg!ds: cannot read memory @ 2101000
0:022> !inetdbg.ds esp

020ffd64 : 0a6b31f0
020ffd68 : 0a5734f0
020ffd6c : 1e142143 : python15!PyInterpreterState_Clear+0x13
020ffd70 : 0a5734f0
020ffd74 : 00000000
020ffd78 : 0a6b31f0
020ffd7c : 1e14267f : python15!Py_Finalize+0x3f
020ffd80 : 0a6b31f0
020ffd84 : 00000000
020ffd88 : 1e2a2401 : pythoncom15!PyCom_DLLReleaseRef+0x71
020ffd8c : 020ffda4
020ffd90 : 1e2c4fd8 : pythoncom15!PyITypeComp__Bind+0x1c28
020ffd94 : 00000000
020ffd98 : 1e2a6795 : pythoncom15!PyGatewayBase__~PyGatewayBase+0xa5
020ffd9c : 0a661f70
020ffda0 : 00000001
020ffda4 : 020fff14
020ffda8 : 1e2c5108 : pythoncom15!PyITypeComp__Bind+0x1d58
020ffdac : ffffffff
020ffdb0 : 1e2a6a41 : pythoncom15!PyGatewayBase__Release+0x21
020ffdb4 : 0008faf8
020ffdb8 : 0008fa78
020ffdbc : 0208245b : axscript!PyGActiveScript__Release+0xb
020ffdc0 : 0a661f70
020ffdc4 : 680d3924 : asp!STR__Copy+0x17
020ffdc8 : 0a661f70
020ffdcc : 680ec330 : asp!g_szErrTemplate+0x17f
020ffdd0 : 0008fa78
020ffdd4 : 680d50cd : asp!CCertRequest__ParseRDNS+0x43
020ffdd8 : 00000000
020ffddc : 020ffe58
020ffde0 : 00000000
020ffde4 : 00000000
020ffde8 : 0014ce18
020ffdec : 020ffe34
020ffdf0 : 680c3b32 : asp!CAppConfig__SetValue+0xc8
020ffdf4 : 020ffe6c
020ffdf8 : 00134748
020ffdfc : 77f050f0 : KERNEL32!InterlockedIncrement
020ffe00 : 02216400
020ffe04 : 680ebe20 : asp!g_PerfMain+0x7
020ffe08 : 680c31e0 : asp!CPerfMainBlock__AddProcess+0x49
020ffe0c : 020ffe58
020ffe10 : 00134748
020ffe14 : 020ffec0
020ffe18 : 02216400
020ffe1c : 00000000
020ffe20 : 0012c241
!inetdbg.ds 20ffe24 to dump next block
0:022> u eip-50 eip+10

python15!PyThreadState_New+0x37:
1e142267 48               dec     eax
1e142268 3489             xor     al,0x89
1e14226a 48               dec     eax
1e14226b 1889481c8b4a     sbb     [ecx+0x4a8b1c48],cl
1e142271 0489             add     al,0x89
1e142273 08894204c390     or      [ecx+0x90c30442],cl
1e142279 90               nop
1e14227a 90               nop
1e14227b 90               nop
1e14227c 90               nop
1e14227d 90               nop
1e14227e 90               nop
1e14227f 90               nop
python15!PyThreadState_Clear:
1e142280 a180d6171e       mov     eax,[python15!Py_VerboseFlag =
(1e17d680)]
1e142285 56               push    esi
1e142286 8b742408         mov     esi,[esp+0x8]
1e14228a 57               push    edi
1e14228b 33ff             xor     edi,edi
1e14228d 3bc7             cmp     eax,edi
1e14228f 741c             jz     python15!PyThreadState_Clear+0x2d =
(1e1422ad)
1e142291 397e08           cmp     [esi+0x8],edi
1e142294 7417             jz     python15!PyThreadState_Clear+0x2d =
(1e1422ad)
1e142296 a100e5171e       mov     eax,[python15+0x7e500 (1e17e500)]
1e14229b 68ec2d171e       push    0x1e172dec
1e1422a0 83c040           add     eax,0x40
1e1422a3 50               push    eax
1e1422a4 ff15fce4171e     call    dword ptr [python15+0x7e4fc =
(1e17e4fc)]
1e1422aa 83c408           add     esp,0x8
1e1422ad 8b4608           mov     eax,[esi+0x8]
1e1422b0 897e08           mov     [esi+0x8],edi
1e1422b3 3bc7             cmp     eax,edi
1e1422b5 7411             jz     python15!PyThreadState_Clear+0x48 =
(1e1422c8)
1e1422b7 8b08             mov     ecx,[eax]
1e1422b9 49               dec     ecx
1e1422ba 8908             mov     [eax],ecx
1e1422bc 750a             jnz    python15!PyThreadState_Clear+0x48 =
(1e1422c8)
1e1422be 8b4804           mov     ecx,[eax+0x4]
1e1422c1 50               push    eax
1e1422c2 ff5118           call    dword ptr [ecx+0x18]
1e1422c5 83c404           add     esp,0x4
0:022> dc eax-100 eax+100

5c6e6e68  ???????? ???????? ???????? ????????  ????????????????
5c6e6e78  ???????? ???????? ???????? ????????  ????????????????
5c6e6e88  ???????? ???????? ???????? ????????  ????????????????
5c6e6e98  ???????? ???????? ???????? ????????  ????????????????
5c6e6ea8  ???????? ???????? ???????? ????????  ????????????????
5c6e6eb8  ???????? ???????? ???????? ????????  ????????????????
5c6e6ec8  ???????? ???????? ???????? ????????  ????????????????
5c6e6ed8  ???????? ???????? ???????? ????????  ????????????????
5c6e6ee8  ???????? ???????? ???????? ????????  ????????????????
5c6e6ef8  ???????? ???????? ???????? ????????  ????????????????
5c6e6f08  ???????? ???????? ???????? ????????  ????????????????
5c6e6f18  ???????? ???????? ???????? ????????  ????????????????
5c6e6f28  ???????? ???????? ???????? ????????  ????????????????
5c6e6f38  ???????? ???????? ???????? ????????  ????????????????
5c6e6f48  ???????? ???????? ???????? ????????  ????????????????
5c6e6f58  ???????? ???????? ???????? ????????  ????????????????
5c6e6f68  ???????? ???????? ???????? ????????  ????????????????
5c6e6f78  ???????? ???????? ???????? ????????  ????????????????
5c6e6f88  ???????? ???????? ???????? ????????  ????????????????
5c6e6f98  ???????? ???????? ???????? ????????  ????????????????
5c6e6fa8  ???????? ???????? ???????? ????????  ????????????????
5c6e6fb8  ???????? ???????? ???????? ????????  ????????????????
5c6e6fc8  ???????? ???????? ???????? ????????  ????????????????
5c6e6fd8  ???????? ???????? ???????? ????????  ????????????????
5c6e6fe8  ???????? ???????? ???????? ????????  ????????????????
5c6e6ff8  ???????? ???????? ???????? ????????  ????????????????
5c6e7008  ???????? ???????? ???????? ????????  ????????????????
5c6e7018  ???????? ???????? ???????? ????????  ????????????????
5c6e7028  ???????? ???????? ???????? ????????  ????????????????
5c6e7038  ???????? ???????? ???????? ????????  ????????????????
5c6e7048  ???????? ???????? ???????? ????????  ????????????????
5c6e7058  ???????? ???????? ???????? ????????  ????????????????
5c6e7068  ????????                             ????
0:022> dc ebx-100 ebx+100

680ec370  00000000 00000000 00000000 00000000  ................
680ec380  00000000 00000000 00000000 00000000  ................
680ec390  00000000 00000000 00000000 00000000  ................
680ec3a0  00000000 00000000 00000000 00000000  ................
680ec3b0  00000000 00000000 00000000 680b1b10  ...............h
680ec3c0  680ec3bc 680ec3bc 0012d4a0 ffffffff  ...h...h........
680ec3d0  00000000 00000000 00000000 00000000  ................
680ec3e0  680b53d0 00000001 00000000 00000000  .S.h............
680ec3f0  680ec400 680c57c0 00000003 00000000  ...h.W.h........
680ec400  00000000 00000000 00000000 00000000  ................
680ec410  00000000 00000000 00000000 00000000  ................
680ec420  00000000 00000000 00000000 00000000  ................
680ec430  00000000 00000000 00000000 00000000  ................
680ec440  00000000 00000000 00000000 00000000  ................
680ec450  00000000 00000000 00000000 00000000  ................
680ec460  00000000 680b1b10 680ec464 680ec464  .......hd..hd..h
680ec470  0012d4c0 ffffffff 00000000 00000000  ................
680ec480  00000000 00000000 680b3340 00000001  ........ at 3.h....
680ec490  001377c0 001348e8 680ec4a8 680c57c0  .w...H.....h.W.h
680ec4a0  0000000b 00000002 00000000 001348e8  .............H..
680ec4b0  00000000 001377c0 00000000 00000000  .....w..........
680ec4c0  00000000 00000000 00000000 00000000  ................
680ec4d0  00000000 00000000 00000000 00000000  ................
680ec4e0  00000000 00000000 00000000 00000000  ................
680ec4f0  00000000 00000000 00000000 00000000  ................
680ec500  00000000 00000000 00000000 0012d480  ................
680ec510  ffffffff 00000000 00000000 00000000  ................
680ec520  00000000 0000008a 00000002 00000008  ................
680ec530  00000007 002927d8 00000002 00000008  .....').........
680ec540  00000007 00000000 00000002 00000000  ................
680ec550  00000001 0012d3a0 ffffffff 00000000  ................
680ec560  00000000 00000000 00000000 342ad0bc  ..............*4
680ec570  342ad6c0                             ..*4
0:022> dc ecx-100 ecx+100

020ffc8c  020ffcd8 07879000 07877e00 00230000  .........~....#.
020ffc9c  77f650a8 072f0000 020ffcd4 00000000  .P.w../.........
020ffcac  77f8f2ad 00230000 078528a8 00000000  ...w..#..(......
020ffcbc  020ffcb0 00000000 020ffd34 77f92cd4  ........4....,.w
020ffccc  77f9d870 ffffffff 020ffcfc 77f64ddc  p..w.........M.w
020ffcdc  00231a90 078528b0 0000000c 078528b0  ..#..(.......(..
020ffcec  08983200 078726d8 780012a6 780012a6  .2...&.....x...x
020ffcfc  7803c120 780012a6 7803c120 020ffd4c   ..x...x ..xL...
020ffd0c  78001434 00000009 780012a6 7803c120  4..x.......x ..x
020ffd1c  020ffd60 78001434 00000009 78001405  `...4..x.......x
020ffd2c  00000020 78001385 0a6de250 0a6de01d   ......xP.m...m.
020ffd3c  0767df18 0a6de000 78032250 ffffffff  ..g...m.P".x....
020ffd4c  020ffdec 020ffd8c 7800ef57 78032250  ........W..xP".x
020ffd5c  ffffffff 020ffdec 0a6b31f0 0a5734f0  .........1k..4W.
020ffd6c  1e142143 0a5734f0 00000000 0a6b31f0  C!...4W......1k.
020ffd7c  1e14267f 0a6b31f0 00000000 1e2a2401  .&...1k......$*.
020ffd8c  020ffda4 1e2c4fd8 00000000 1e2a6795  .....O,......g*.
020ffd9c  0a661f70 00000001 020fff14 1e2c5108  p.f..........Q,.
020ffdac  ffffffff 1e2a6a41 0008faf8 0008fa78  ....Aj*.....x...
020ffdbc  0208245b 0a661f70 680d3924 0a661f70  [$..p.f.$9.hp.f.
020ffdcc  680ec330 0008fa78 680d50cd 00000000  0..hx....P.h....
020ffddc  020ffe58 00000000 00000000 0014ce18  X...............
020ffdec  020ffe34 680c3b32 020ffe6c 00134748  4...2;.hl...HG..
020ffdfc  77f050f0 02216400 680ebe20 680c31e0  .P.w.d!. ..h.1.h
020ffe0c  020ffe58 00134748 020ffec0 02216400  X...HG.......d!.
020ffe1c  00000000 0012c241 680c3b58 00000000  ....A...X;.h....
020ffe2c  680ec6d0 000000c0 020ffe7c 680c2f38  ...h....|...8/.h
020ffe3c  0009f0f0 00000000 020ffec0 020ffe58  ............X...
020ffe4c  77f050f0 02216400 680ebe20 00000001  .P.w.d!. ..h....
020ffe5c  00000001 020ffe64 045dc31c 045dc324  ....d.....].$.].
020ffe6c  0008fa78 00000000 0009f0f0 00000000  x...............
020ffe7c  00000000 680c78c0 0009f0f0 02216400  .....x.h.....d!.
020ffe8c  020ffec0                             ....
0:022> dc edx-100 edx+100

7803c020  00000000 00000000 00000000 00000000  ................
7803c030  00000000 7803c0f0 00298880 00000000  .......x..).....
7803c040  00000000 00000000 00000000 00000000  ................
7803c050  00000000 7803c120 00000000 002988e0  .... ..x......).
7803c060  00294aa0 7803c138 00000000 00000000  .J).8..x........
7803c070  00000000 7803c108 002988c0 00240600  .......x..)...$.
7803c080  00000000 00000000 00000000 00000000  ................
7803c090  00000000 00240180 00000000 00000000  ......$.........
7803c0a0  00000000 003c2ba0 0029de50 002988a0  .....+<.P.)...).
7803c0b0  00298900 0029eb70 002ae910 002b3ad0  ..).p.)...*..:+.
7803c0c0  00348f80 00350970 063274c0 00000000  ..4.p.5..t2.....
7803c0d0  00000000 00000000 00000000 00000000  ................
7803c0e0  00000000 00000000 00000000 00000000  ................
7803c0f0  77fa7790 ffffffff 00000000 00000000  .w.w............
7803c100  00000000 00000000 77fa7730 ffffffff  ........0w.w....
7803c110  00000000 00000000 00000000 00000000  ................
7803c120  77fa7770 ffffffff 00000000 00000000  pw.w............
7803c130  00000a70 00000000 77fa7750 ffffffff  p.......Pw.w....
7803c140  00000000 00000000 00000000 00000000  ................
7803c150  000001e0 0767df18 00000000 00000018  ......g.........
7803c160  00200000 00200020 00200020 00200020  .. . . . . . . .
7803c170  00200020 00280028 00280028 00200028   . .(.(.(.(.(. .
7803c180  00200020 00200020 00200020 00200020   . . . . . . . .
7803c190  00200020 00200020 00200020 00200020   . . . . . . . .
7803c1a0  00480020 00100010 00100010 00100010   .H.............
7803c1b0  00100010 00100010 00100010 00100010  ................
7803c1c0  00840010 00840084 00840084 00840084  ................
7803c1d0  00840084 00100084 00100010 00100010  ................
7803c1e0  00100010 00810081 00810081 00810081  ................
7803c1f0  00010001 00010001 00010001 00010001  ................
7803c200  00010001 00010001 00010001 00010001  ................
7803c210  00010001 00010001 00100010 00100010  ................
7803c220  00100010                             ....
0:022> dc esi-100 esi+100

0a5733f0  00000000 1e168fd0 0a54f7a0 0a5717f0  ..........T...W.
0a573400  00000000 1e170238 0a54f760 414b4341  ....8...`.T.ACKA
0a573410  00000000 1e168fd0 0a5502f0 0a5717f0  ..........U...W.
0a573420  00000000 1e170238 0a54f8d0 69725065  ....8.....T.ePri
0a573430  697e3768 697e3740 697e3728 697e3708  h7~i at 7~i(7~i.7~i
0a573440  697e36f8 00000003 00293230 00000000  .6~i....02).....
0a573450  00000002 1e176bd8 00000008 e1d9e008  .....k..........
0a573460  0a573450 65707954 65707954 45544f00  P4W.TypeType.OTE
0a573470  53490138 00000000 00010003 0a5735b0  8.IS.........5W.
0a573480  00000001 1e176bd8 00000014 ffffffff  .....k..........
0a573490  00000000 63614d49 656e6968 75626544  ....IMachineDebu
0a5734a0  6e614d67 72656761 51973c00 11d0cb0c  gManager.<.Q....
0a5734b0  00000002 1e176bd8 00000007 3c0dad84  .....k.........<
0a5734c0  0a5734b0 54746e49 00657079 69766900  .4W.IntType..ivi
0a5734d0  00000000 1e176bd8 00000037 ffffffff  .....k..7.......
0a5734e0  00000000 505c3a43 72676f72 46206d61  ....C:\Program F
0a5734f0  73656c69 7479505c 00000000 336e6977  iles\Pyt....win3
0a573500  6d6f6332 5c747865 65647861 5c677562  2comext\axdebug\
0a573510  65746167 73796177 0079702e 00000000  gateways.py.....
0a573520  00000000 00000000 00000000 3674e0d9  ..............t6
0a573530  00000002 1e176bd8 00000018 65c3ff27  .....k......'..e
0a573540  0a573530 336e6977 6d6f6332 7265732e  05W.win32com.ser
0a573550  2e726576 656e6977 726f7272 65534200  ver.winerror.BSe
0a573560  00000002 1e176bd8 00000008 44a2c362  .....k......b..D
0a573570  0a573560 676e6f4c 65707954 0a573600  `5W.LongType.6W.
0a573580  00000001 1e176bd8 00000011 ffffffff  .....k..........
0a573590  00000000 62654449 70416775 63696c70  ....IDebugApplic
0a5735a0  6f697461 6e69006e 78655f74 68740063  ation.int_exc.th
0a5735b0  00000000 1e176bd8 00000014 ffffffff  .....k..........
0a5735c0  00000000 43746547 65746e6f 664f7478  ....GetContextOf
0a5735d0  69736f50 6e6f6974 00010000 0a5735e0  Position.....5W.
0a5735e0  00000123 00000000 00010003 0a573610  #............6W.
0a5735f0  00000000                             ....
0:022> dc edi-100 edi+100

                        ^ Range error in 'dc edi-100 edi+100'
0:022> dc esp-100 esp+100

020ffc64  00230000 00001000 0786b000 00230548  ..#.........H.#.
020ffc74  072f0030 77f6447e 072f0000 072f0030  0./.~D.w../.0./.
020ffc84  77f6447e 072f0000 020ffcd8 07879000  ~D.w../.........
020ffc94  07877e00 00230000 77f650a8 072f0000  .~....#..P.w../.
020ffca4  020ffcd4 00000000 77f8f2ad 00230000  ...........w..#.
020ffcb4  078528a8 00000000 020ffcb0 00000000  .(..............
020ffcc4  020ffd34 77f92cd4 77f9d870 ffffffff  4....,.wp..w....
020ffcd4  020ffcfc 77f64ddc 00231a90 078528b0  .....M.w..#..(..
020ffce4  0000000c 078528b0 08983200 078726d8  .....(...2...&..
020ffcf4  780012a6 780012a6 7803c120 780012a6  ...x...x ..x...x
020ffd04  7803c120 020ffd4c 78001434 00000009   ..xL...4..x....
020ffd14  780012a6 7803c120 020ffd60 78001434  ...x ..x`...4..x
020ffd24  00000009 78001405 00000020 78001385  .......x ......x
020ffd34  0a6de250 0a6de01d 0767df18 0a6de000  P.m...m...g...m.
020ffd44  78032250 ffffffff 020ffdec 020ffd8c  P".x............
020ffd54  7800ef57 78032250 ffffffff 020ffdec  W..xP".x........
020ffd64  0a6b31f0 0a5734f0 1e142143 0a5734f0  .1k..4W.C!...4W.
020ffd74  00000000 0a6b31f0 1e14267f 0a6b31f0  .....1k..&...1k.
020ffd84  00000000 1e2a2401 020ffda4 1e2c4fd8  .....$*......O,.
020ffd94  00000000 1e2a6795 0a661f70 00000001  .....g*.p.f.....
020ffda4  020fff14 1e2c5108 ffffffff 1e2a6a41  .....Q,.....Aj*.
020ffdb4  0008faf8 0008fa78 0208245b 0a661f70  ....x...[$..p.f.
020ffdc4  680d3924 0a661f70 680ec330 0008fa78  $9.hp.f.0..hx...
020ffdd4  680d50cd 00000000 020ffe58 00000000  .P.h....X.......
020ffde4  00000000 0014ce18 020ffe34 680c3b32  ........4...2;.h
020ffdf4  020ffe6c 00134748 77f050f0 02216400  l...HG...P.w.d!.
020ffe04  680ebe20 680c31e0 020ffe58 00134748   ..h.1.hX...HG..
020ffe14  020ffec0 02216400 00000000 0012c241  .....d!.....A...
020ffe24  680c3b58 00000000 680ec6d0 000000c0  X;.h.......h....
020ffe34  020ffe7c 680c2f38 0009f0f0 00000000  |...8/.h........
020ffe44  020ffec0 020ffe58 77f050f0 02216400  ....X....P.w.d!.
020ffe54  680ebe20 00000001 00000001 020ffe64   ..h........d...
020ffe64  045dc31c                             ..].
0:022> dc ebp-100 ebp+100

020ffcec  08983200 078726d8 780012a6 780012a6  .2...&.....x...x
020ffcfc  7803c120 780012a6 7803c120 020ffd4c   ..x...x ..xL...
020ffd0c  78001434 00000009 780012a6 7803c120  4..x.......x ..x
020ffd1c  020ffd60 78001434 00000009 78001405  `...4..x.......x
020ffd2c  00000020 78001385 0a6de250 0a6de01d   ......xP.m...m.
020ffd3c  0767df18 0a6de000 78032250 ffffffff  ..g...m.P".x....
020ffd4c  020ffdec 020ffd8c 7800ef57 78032250  ........W..xP".x
020ffd5c  ffffffff 020ffdec 0a6b31f0 0a5734f0  .........1k..4W.
020ffd6c  1e142143 0a5734f0 00000000 0a6b31f0  C!...4W......1k.
020ffd7c  1e14267f 0a6b31f0 00000000 1e2a2401  .&...1k......$*.
020ffd8c  020ffda4 1e2c4fd8 00000000 1e2a6795  .....O,......g*.
020ffd9c  0a661f70 00000001 020fff14 1e2c5108  p.f..........Q,.
020ffdac  ffffffff 1e2a6a41 0008faf8 0008fa78  ....Aj*.....x...
020ffdbc  0208245b 0a661f70 680d3924 0a661f70  [$..p.f.$9.hp.f.
020ffdcc  680ec330 0008fa78 680d50cd 00000000  0..hx....P.h....
020ffddc  020ffe58 00000000 00000000 0014ce18  X...............
020ffdec  020ffe34 680c3b32 020ffe6c 00134748  4...2;.hl...HG..
020ffdfc  77f050f0 02216400 680ebe20 680c31e0  .P.w.d!. ..h.1.h
020ffe0c  020ffe58 00134748 020ffec0 02216400  X...HG.......d!.
020ffe1c  00000000 0012c241 680c3b58 00000000  ....A...X;.h....
020ffe2c  680ec6d0 000000c0 020ffe7c 680c2f38  ...h....|...8/.h
020ffe3c  0009f0f0 00000000 020ffec0 020ffe58  ............X...
020ffe4c  77f050f0 02216400 680ebe20 00000001  .P.w.d!. ..h....
020ffe5c  00000001 020ffe64 045dc31c 045dc324  ....d.....].$.].
020ffe6c  0008fa78 00000000 0009f0f0 00000000  x...............
020ffe7c  00000000 680c78c0 0009f0f0 02216400  .....x.h.....d!.
020ffe8c  020ffec0 00000000 02228310 020ffee4  ..........".....
020ffe9c  0012c110 00000000 00000000 0013cba0  ................
020ffeac  0009f0f0 00000000 00000000 fffffffe  ................
020ffebc  00000001 0014ba80 0014baa8 0014bad0  ................
020ffecc  0014baf8 680e25c6 00000000 697e1898  .....%.h......~i
020ffedc  003ce8e0 003ce3ec 020fff24 697e67cc  ..<...<.$....g~i
020ffeec  00000000                             ....
0:022> dc esp+0x04

020ffd68  0a5734f0 1e142143 0a5734f0 00000000  .4W.C!...4W.....
020ffd78  0a6b31f0 1e14267f 0a6b31f0 00000000  .1k..&...1k.....
020ffd88  1e2a2401 020ffda4 1e2c4fd8 00000000  .$*......O,.....
020ffd98  1e2a6795 0a661f70 00000001 020fff14  .g*.p.f.........
020ffda8  1e2c5108 ffffffff 1e2a6a41 0008faf8  .Q,.....Aj*.....
020ffdb8  0008fa78 0208245b 0a661f70 680d3924  x...[$..p.f.$9.h
020ffdc8  0a661f70 680ec330 0008fa78 680d50cd  p.f.0..hx....P.h
020ffdd8  00000000 020ffe58 00000000 00000000  ....X...........
0:022> dc esp+0x08

020ffd6c  1e142143 0a5734f0 00000000 0a6b31f0  C!...4W......1k.
020ffd7c  1e14267f 0a6b31f0 00000000 1e2a2401  .&...1k......$*.
020ffd8c  020ffda4 1e2c4fd8 00000000 1e2a6795  .....O,......g*.
020ffd9c  0a661f70 00000001 020fff14 1e2c5108  p.f..........Q,.
020ffdac  ffffffff 1e2a6a41 0008faf8 0008fa78  ....Aj*.....x...
020ffdbc  0208245b 0a661f70 680d3924 0a661f70  [$..p.f.$9.hp.f.
020ffdcc  680ec330 0008fa78 680d50cd 00000000  0..hx....P.h....
020ffddc  020ffe58 00000000 00000000 0014ce18  X...............
0:022> dc esp+0x0c

020ffd70  0a5734f0 00000000 0a6b31f0 1e14267f  .4W......1k..&..
020ffd80  0a6b31f0 00000000 1e2a2401 020ffda4  .1k......$*.....
020ffd90  1e2c4fd8 00000000 1e2a6795 0a661f70  .O,......g*.p.f.
020ffda0  00000001 020fff14 1e2c5108 ffffffff  .........Q,.....
020ffdb0  1e2a6a41 0008faf8 0008fa78 0208245b  Aj*.....x...[$..
020ffdc0  0a661f70 680d3924 0a661f70 680ec330  p.f.$9.hp.f.0..h
020ffdd0  0008fa78 680d50cd 00000000 020ffe58  x....P.h....X...
020ffde0  00000000 00000000 0014ce18 020ffe34  ............4...
0:022> dc esp+0x10

020ffd74  00000000 0a6b31f0 1e14267f 0a6b31f0  .....1k..&...1k.
020ffd84  00000000 1e2a2401 020ffda4 1e2c4fd8  .....$*......O,.
020ffd94  00000000 1e2a6795 0a661f70 00000001  .....g*.p.f.....
020ffda4  020fff14 1e2c5108 ffffffff 1e2a6a41  .....Q,.....Aj*.
020ffdb4  0008faf8 0008fa78 0208245b 0a661f70  ....x...[$..p.f.
020ffdc4  680d3924 0a661f70 680ec330 0008fa78  $9.hp.f.0..hx...
020ffdd4  680d50cd 00000000 020ffe58 00000000  .P.h....X.......
020ffde4  00000000 0014ce18 020ffe34 680c3b32  ........4...2;.h
0:022> dc ebp+0x04

020ffdf0  680c3b32 020ffe6c 00134748 77f050f0  2;.hl...HG...P.w
020ffe00  02216400 680ebe20 680c31e0 020ffe58  .d!. ..h.1.hX...
020ffe10  00134748 020ffec0 02216400 00000000  HG.......d!.....
020ffe20  0012c241 680c3b58 00000000 680ec6d0  A...X;.h.......h
020ffe30  000000c0 020ffe7c 680c2f38 0009f0f0  ....|...8/.h....
020ffe40  00000000 020ffec0 020ffe58 77f050f0  ........X....P.w
020ffe50  02216400 680ebe20 00000001 00000001  .d!. ..h........
020ffe60  020ffe64 045dc31c 045dc324 0008fa78  d.....].$.].x...
0:022> dc ebp+0x08

020ffdf4  020ffe6c 00134748 77f050f0 02216400  l...HG...P.w.d!.
020ffe04  680ebe20 680c31e0 020ffe58 00134748   ..h.1.hX...HG..
020ffe14  020ffec0 02216400 00000000 0012c241  .....d!.....A...
020ffe24  680c3b58 00000000 680ec6d0 000000c0  X;.h.......h....
020ffe34  020ffe7c 680c2f38 0009f0f0 00000000  |...8/.h........
020ffe44  020ffec0 020ffe58 77f050f0 02216400  ....X....P.w.d!.
020ffe54  680ebe20 00000001 00000001 020ffe64   ..h........d...
020ffe64  045dc31c 045dc324 0008fa78 00000000  ..].$.].x.......
0:022> dc ebp+0x0c

020ffdf8  00134748 77f050f0 02216400 680ebe20  HG...P.w.d!. ..h
020ffe08  680c31e0 020ffe58 00134748 020ffec0  .1.hX...HG......
020ffe18  02216400 00000000 0012c241 680c3b58  .d!.....A...X;.h
020ffe28  00000000 680ec6d0 000000c0 020ffe7c  .......h....|...
020ffe38  680c2f38 0009f0f0 00000000 020ffec0  8/.h............
020ffe48  020ffe58 77f050f0 02216400 680ebe20  X....P.w.d!. ..h
020ffe58  00000001 00000001 020ffe64 045dc31c  ........d.....].
020ffe68  045dc324 0008fa78 00000000 0009f0f0  $.].x...........
0:022> dc ebp+0x10

020ffdfc  77f050f0 02216400 680ebe20 680c31e0  .P.w.d!. ..h.1.h
020ffe0c  020ffe58 00134748 020ffec0 02216400  X...HG.......d!.
020ffe1c  00000000 0012c241 680c3b58 00000000  ....A...X;.h....
020ffe2c  680ec6d0 000000c0 020ffe7c 680c2f38  ...h....|...8/.h
020ffe3c  0009f0f0 00000000 020ffec0 020ffe58  ............X...
020ffe4c  77f050f0 02216400 680ebe20 00000001  .P.w.d!. ..h....
020ffe5c  00000001 020ffe64 045dc31c 045dc324  ....d.....].$.].
020ffe6c  0008fa78 00000000 0009f0f0 00000000  x...............
0:022> dc poi(esp+0x04)

0a5734f0  73656c69 7479505c 00000000 336e6977  iles\Pyt....win3
0a573500  6d6f6332 5c747865 65647861 5c677562  2comext\axdebug\
0a573510  65746167 73796177 0079702e 00000000  gateways.py.....
0a573520  00000000 00000000 00000000 3674e0d9  ..............t6
0a573530  00000002 1e176bd8 00000018 65c3ff27  .....k......'..e
0a573540  0a573530 336e6977 6d6f6332 7265732e  05W.win32com.ser
0a573550  2e726576 656e6977 726f7272 65534200  ver.winerror.BSe
0a573560  00000002 1e176bd8 00000008 44a2c362  .....k......b..D
0:022> dc poi(esp+0x08)

1e142143  c483368b 75f68504 08478bf1 000847c7  .6.....u..G..G..
1e142153  85000000 8b1174c0 08894908 488b0a75  .....t...I..u..H
1e142163  51ff5004 04c48318 c70c478b 00000c47  .P.Q.....G..G...
1e142173  c0850000 088b1174 75088949 04508b0a  ....t...I..u..P.
1e142183  1852ff50 8b04c483 47c71047 00000010  P.R.....G..G....
1e142193  74c08500 49088b11 0a750889 5004488b  ...t...I..u..H.P
1e1421a3  831851ff 5e5f04c4 909090c3 8b575690  .Q...._^.....VW.
1e1421b3  570c247c 000054e8 04c48300 17cd34be  |$.W.T.......4..
0:022> dc poi(esp+0x0c)

0a5734f0  73656c69 7479505c 00000000 336e6977  iles\Pyt....win3
0a573500  6d6f6332 5c747865 65647861 5c677562  2comext\axdebug\
0a573510  65746167 73796177 0079702e 00000000  gateways.py.....
0a573520  00000000 00000000 00000000 3674e0d9  ..............t6
0a573530  00000002 1e176bd8 00000018 65c3ff27  .....k......'..e
0a573540  0a573530 336e6977 6d6f6332 7265732e  05W.win32com.ser
0a573550  2e726576 656e6977 726f7272 65534200  ver.winerror.BSe
0a573560  00000002 1e176bd8 00000008 44a2c362  .....k......b..D
0:022> dc poi(esp+0x10)

00000000  ???????? ???????? ???????? ????????  ????????????????
00000010  ???????? ???????? ???????? ????????  ????????????????
00000020  ???????? ???????? ???????? ????????  ????????????????
00000030  ???????? ???????? ???????? ????????  ????????????????
00000040  ???????? ???????? ???????? ????????  ????????????????
00000050  ???????? ???????? ???????? ????????  ????????????????
00000060  ???????? ???????? ???????? ????????  ????????????????
00000070  ???????? ???????? ???????? ????????  ????????????????
0:022> dc poi(ebp+0x04)

680c3b32  c7b70f47 7c04463b 045e89de 7e013e83  G...;F.|..^..>.~
680c3b42  ff53530a 15e80876 89000050 085e891e  .SS.v...P.....^.
680c3b52  c25b5e5f 4c8b0008 448b0824 413b0424  _^[....L$..D$.;A
680c3b62  33047204 8b0aebc0 048d0849 81448b80  .r.3....I.....D.
680c3b72  0008c208 8bec8b55 4d8b1845 00208014  ....U...E..M.. .
680c3b82  1f74c985 08758b56 0f118b50 ff0c45bf  ..t.V.u.P....E..
680c3b92  768b1075 80048d08 088674ff 1452ff51  u..v.....t..Q.R.
680c3ba2  0f19eb5e 8b0c45bf 75ff084d 80048d10  ^....E..M..u....
0:022> dc poi(ebp+0x08)

020ffe6c  0008fa78 00000000 0009f0f0 00000000  x...............
020ffe7c  00000000 680c78c0 0009f0f0 02216400  .....x.h.....d!.
020ffe8c  020ffec0 00000000 02228310 020ffee4  ..........".....
020ffe9c  0012c110 00000000 00000000 0013cba0  ................
020ffeac  0009f0f0 00000000 00000000 fffffffe  ................
020ffebc  00000001 0014ba80 0014baa8 0014bad0  ................
020ffecc  0014baf8 680e25c6 00000000 697e1898  .....%.h......~i
020ffedc  003ce8e0 003ce3ec 020fff24 697e67cc  ..<...<.$....g~i
0:022> dc poi(ebp+0x0c)

00134748  680b1aa8 680b3bc0 00000000 0013143c  ...h.;.h....<...
00134758  680b1ae4 00134398 00000011 00134160  ...h.C......`A..
00134768  00000000 00000001 00000006 00000001  ................
00134778  00000001 00134398 00134840 00134358  .....C.. at H..XC..
00134788  00000000 00000000 001349b0 000826b8  .........I...&..
00134798  000bb0c0 00134088 00134960 00000000  ..... at ..`I......
001347a8  0012d660 ffffffff 00000000 00000000  `...............
001347b8  00000000 00000000 0012d680 ffffffff  ................
0:022> dc poi(ebp+0x10)

77f050f0  04244c8b 000001b8 c10ff000 04c24001  .L$.......... at ..
77f05100  c08b2e00 04244c8b ffffffb8 c10ff0ff  .....L$.........
77f05110  04c24801 c08b2e00 04244c8b 0824548b  .H.......L$..T$.
77f05120  0ff0018b fa7511b1 900008c2 04244c8b  ......u......L$.
77f05130  0824548b 0c24448b 11b10ff0 90000cc2  .T$..D$.........
77f05140  04244c8b 0824448b 01c10ff0 900008c2  .L$..D$.........
77f05150  0424448b 3878c00b 0824548b 880fd20b  .D$...x8.T$.....
77f05160  000000b2 4c8be2f7 c90b0c24 00e6880f  .......L$.......
0:022> !inetdbg.ver

Module @ 0x01000000 =3D inetinfo.exe
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D Internet Information Services
FileVersion         =3D 4.02.0622
InternalName        =3D INETINFO.EXE
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1997
OriginalFilename    =3D INETINFO.EXE
ProductName         =3D Internet Information Server
ProductVersion      =3D 4.02.0622

Module @ 0x77f60000 =3D ntdll.dll
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D NT Layer DLL
FileVersion         =3D 4.00
InternalName        =3D ntdll.dll
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1996
OriginalFilename    =3D ntdll.dll
ProductName         =3D Microsoft(R) Windows NT(TM) Operating System
ProductVersion      =3D 4.00

Module @ 0x78000000 =3D MSVCRT.dll
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D Microsoft (R) C Runtime Library
FileVersion         =3D 6.10.8528.0
InternalName        =3D MSVCRT.DLL
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1998
OriginalFilename    =3D MSVCRT.DLL
ProductName         =3D Microsoft (R) Visual C++
ProductVersion      =3D 6.10.8528.0

Module @ 0x77f00000 =3D KERNEL32.dll
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D Windows NT BASE API Client DLL
FileVersion         =3D 4.00
InternalName        =3D kernel32
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1996
OriginalFilename    =3D kernel32
ProductName         =3D Microsoft(R) Windows NT(TM) Operating System
ProductVersion      =3D 4.00

Module @ 0x77dc0000 =3D ADVAPI32.dll
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D Advanced Windows 32 Base API
FileVersion         =3D 4.00
InternalName        =3D advapi32.dll
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1996
OriginalFilename    =3D advapi32.dll
ProductName         =3D Microsoft(R) Windows NT(TM) Operating System
ProductVersion      =3D 4.00

Module @ 0x77e70000 =3D USER32.dll
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D Windows NT USER API Client DLL
FileVersion         =3D 4.00
InternalName        =3D user32
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1996
OriginalFilename    =3D user32
ProductName         =3D Microsoft(R) Windows NT(TM) Operating System
ProductVersion      =3D 4.00

Module @ 0x77ed0000 =3D GDI32.dll
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D GDI Client DLL
FileVersion         =3D 4.00
InternalName        =3D gdi32
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1996
OriginalFilename    =3D gdi32
ProductName         =3D Microsoft(R) Windows NT(TM) Operating System
ProductVersion      =3D 4.00

Module @ 0x77e10000 =3D RPCRT4.dll
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D Remote Procedure Call Runtime
FileVersion         =3D 4.00
InternalName        =3D rpcrt4.dll
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1996
OriginalFilename    =3D rpcrt4.dll
ProductName         =3D Microsoft(R) Windows NT(TM) Operating System
ProductVersion      =3D 4.00

Module @ 0x77b20000 =3D ole32.dll
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D Microsoft OLE for Windows and Windows NT
FileVersion         =3D 4.00
InternalName        =3D OLE32.DLL
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1996
OriginalFilename    =3D OLE32.DLL
ProductName         =3D Microsoft(R) Windows NT(TM) Operating System
ProductVersion      =3D 4.00

Module @ 0x68e80000 =3D rpcref.dll
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D Microsoft Internet Information Services RPC =
helper library
FileVersion         =3D 4.02.0622
InternalName        =3D RPCREF.DLL
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1997
OriginalFilename    =3D RPCREF.DLL
ProductName         =3D Internet Information Server
ProductVersion      =3D 4.02.0622

Module @ 0x77bf0000 =3D rpcltc1.dll
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D Remote Process Control LTC1 DLL
FileVersion         =3D 4.00
InternalName        =3D rpcltc1.dll
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1996
OriginalFilename    =3D rpcltc1.dll
ProductName         =3D Microsoft(R) Windows NT(TM) Operating System
ProductVersion      =3D 4.00

Module @ 0x683a0000 =3D iisadmin.dll
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D Metadata and Admin Service
FileVersion         =3D 4.02.0720
InternalName        =3D mdadmin.dll
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1997
OriginalFilename    =3D mdadmin.dll
ProductName         =3D Internet Information Server
CDB: c0000005 Exception in inetdbg.ver debugger extension.
      PC: 77f66ebf  VA: 00078000  R/W: 1  Parameter: 1001f
0:022> !inetdbg.ver

Module @ 0x01000000 =3D inetinfo.exe
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D Internet Information Services
FileVersion         =3D 4.02.0622
InternalName        =3D INETINFO.EXE
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1997
OriginalFilename    =3D INETINFO.EXE
ProductName         =3D Internet Information Server
ProductVersion      =3D 4.02.0622

Module @ 0x77f60000 =3D ntdll.dll
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D NT Layer DLL
FileVersion         =3D 4.00
InternalName        =3D ntdll.dll
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1996
OriginalFilename    =3D ntdll.dll
ProductName         =3D Microsoft(R) Windows NT(TM) Operating System
ProductVersion      =3D 4.00

Module @ 0x78000000 =3D MSVCRT.dll
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D Microsoft (R) C Runtime Library
FileVersion         =3D 6.10.8528.0
InternalName        =3D MSVCRT.DLL
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1998
OriginalFilename    =3D MSVCRT.DLL
ProductName         =3D Microsoft (R) Visual C++
ProductVersion      =3D 6.10.8528.0

Module @ 0x77f00000 =3D KERNEL32.dll
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D Windows NT BASE API Client DLL
FileVersion         =3D 4.00
InternalName        =3D kernel32
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1996
OriginalFilename    =3D kernel32
ProductName         =3D Microsoft(R) Windows NT(TM) Operating System
ProductVersion      =3D 4.00

Module @ 0x77dc0000 =3D ADVAPI32.dll
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D Advanced Windows 32 Base API
FileVersion         =3D 4.00
InternalName        =3D advapi32.dll
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1996
OriginalFilename    =3D advapi32.dll
ProductName         =3D Microsoft(R) Windows NT(TM) Operating System
ProductVersion      =3D 4.00

Module @ 0x77e70000 =3D USER32.dll
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D Windows NT USER API Client DLL
FileVersion         =3D 4.00
InternalName        =3D user32
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1996
OriginalFilename    =3D user32
ProductName         =3D Microsoft(R) Windows NT(TM) Operating System
ProductVersion      =3D 4.00

Module @ 0x77ed0000 =3D GDI32.dll
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D GDI Client DLL
FileVersion         =3D 4.00
InternalName        =3D gdi32
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1996
OriginalFilename    =3D gdi32
ProductName         =3D Microsoft(R) Windows NT(TM) Operating System
ProductVersion      =3D 4.00

Module @ 0x77e10000 =3D RPCRT4.dll
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D Remote Procedure Call Runtime
FileVersion         =3D 4.00
InternalName        =3D rpcrt4.dll
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1996
OriginalFilename    =3D rpcrt4.dll
ProductName         =3D Microsoft(R) Windows NT(TM) Operating System
ProductVersion      =3D 4.00

Module @ 0x77b20000 =3D ole32.dll
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D Microsoft OLE for Windows and Windows NT
FileVersion         =3D 4.00
InternalName        =3D OLE32.DLL
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1996
OriginalFilename    =3D OLE32.DLL
ProductName         =3D Microsoft(R) Windows NT(TM) Operating System
ProductVersion      =3D 4.00

Module @ 0x68e80000 =3D rpcref.dll
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D Microsoft Internet Information Services RPC =
helper library
FileVersion         =3D 4.02.0622
InternalName        =3D RPCREF.DLL
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1997
OriginalFilename    =3D RPCREF.DLL
ProductName         =3D Internet Information Server
ProductVersion      =3D 4.02.0622

Module @ 0x77bf0000 =3D rpcltc1.dll
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D Remote Process Control LTC1 DLL
FileVersion         =3D 4.00
InternalName        =3D rpcltc1.dll
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1996
OriginalFilename    =3D rpcltc1.dll
ProductName         =3D Microsoft(R) Windows NT(TM) Operating System
ProductVersion      =3D 4.00

Module @ 0x683a0000 =3D iisadmin.dll
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D Metadata and Admin Service
FileVersion         =3D 4.02.0720
InternalName        =3D mdadmin.dll
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1997
OriginalFilename    =3D mdadmin.dll
ProductName         =3D Internet Information Server
ProductVersion      =3D 4.02.0720

Module @ 0x68220000 =3D COADMIN.dll
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D IIS CoAdmin DLL
FileVersion         =3D 4.02.0720
InternalName        =3D coadmin.dll
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1997
OriginalFilename    =3D coadmin.dll
ProductName         =3D Internet Information Server
ProductVersion      =3D 4.02.0720

Module @ 0x70980000 =3D SHELL32.dll
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D Windows Shell Common Dll
FileVersion         =3D 4.72.3612.1700
InternalName        =3D SHELL32
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1997
OriginalFilename    =3D SHELL32.DLL
ProductName         =3D Microsoft(R) Windows NT(R) Operating System
ProductVersion      =3D 4.72.3612.1700

Module @ 0x710d0000 =3D SHLWAPI.dll
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D Shell Light-weight Utility Library
FileVersion         =3D 4.72.3110.0
InternalName        =3D SHLWAPI
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1997
OriginalFilename    =3D SHLWAPI.DLL
ProductName         =3D Microsoft(R) Windows NT(R) Operating System
ProductVersion      =3D 4.72.3110.0

Module @ 0x77aa0000 =3D COMCTL32.dll
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D Custom Controls Library
FileVersion         =3D 4.72.3612.1702
InternalName        =3D COMMCTRL
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1999
OriginalFilename    =3D COMMCTRL.DLL
ProductName         =3D Microsoft(R) Windows NT(R) Operating System
ProductVersion      =3D 4.72.3612.1702

Module @ 0x68020000 =3D ADMWPROX.dll
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D IIS Admin Com API Proxy dll
FileVersion         =3D 4.02.0720
InternalName        =3D admwprox.dll
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1997
OriginalFilename    =3D admwprox.dll
ProductName         =3D Internet Information Server
ProductVersion      =3D 4.02.0720

Module @ 0x74880000 =3D userenv.dll
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D Userenv
FileVersion         =3D 4.00
InternalName        =3D userenv
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1996
OriginalFilename    =3D userenv.dll
ProductName         =3D Microsoft(R) Windows NT(TM) Operating System
ProductVersion      =3D 4.00

Module @ 0x77800000 =3D NETAPI32.dll
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D Net Win32 API DLL
FileVersion         =3D 4.00
InternalName        =3D NetApi32.DLL
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1996
OriginalFilename    =3D NetApi32.DLL
ProductName         =3D Microsoft(R) Windows NT(TM) Operating System
ProductVersion      =3D 4.00

Module @ 0x77840000 =3D NETRAP.dll
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D Net Remote Admin Protocol DLL
FileVersion         =3D 4.00
InternalName        =3D NetRap.DLL
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1996
OriginalFilename    =3D NetRap.DLL
ProductName         =3D Microsoft(R) Windows NT(TM) Operating System
ProductVersion      =3D 4.00

Module @ 0x777e0000 =3D SAMLIB.dll
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D SAM Library DLL
FileVersion         =3D 4.00
InternalName        =3D SAMLib.DLL
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1996
OriginalFilename    =3D SAMLib.DLL
ProductName         =3D Microsoft(R) Windows NT(TM) Operating System
ProductVersion      =3D 4.00

Module @ 0x68900000 =3D nsepm.dll
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D IIS NSEP mapping DLL
FileVersion         =3D 4.02.0720
InternalName        =3D nsepm.dll
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1997
OriginalFilename    =3D nsepm.dll
ProductName         =3D Internet Information Server
ProductVersion      =3D 4.02.0720

Module @ 0x68440000 =3D IISMAP.dll
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
FileDescription     =3D Microsoft IIS mapper
FileVersion         =3D 4.02.0720
InternalName        =3D iismap.dll
LegalCopyright      =3D Copyright (C) Microsoft Corp. 1981-1997
OriginalFilename    =3D iismap.dll
ProductName         =3D Internet Information Server
ProductVersion      =3D 4.02.0720

Module @ 0x77400000 =3D schannel.dll
dwFileFlags         =3D 0x00000000 (FREE)
CompanyName         =3D Microsoft Corporation
CDB: c0000005 Exception in inetdbg.ver debugger extension.
      PC: 77f66edd  VA: 00079000  R/W: 1  Parameter: 1001f
0:022> q

quit:

------_=_NextPart_000_01BFA65D.E540DE90--


Return-Path: <nathan_christensen at peoplesoft.com>
Delivered-To: python-list at python.org
Received: from mail03-oak.pilot.net (mail-oak-3.pilot.net [198.232.147.18])
	by dinsdale.python.org (Postfix) with ESMTP id 84E6F1CE4E
	for <python-list at python.org>; Mon, 17 Apr 2000 14:56:58 -0400 (EDT)
Received: from psh-wss-01.peoplesoft.com (unk-131-192-163.peoplesoft.com [216.131.192.163] (may be forged)) by mail03-oak.pilot.net with SMTP id LAA11818 for <python-list at python.org>; Mon, 17 Apr 2000 11:59:35 -0700 (PDT)
From: nathan_christensen at peoplesoft.com
Received: from 216.131.192.133 by psh-wss-01.peoplesoft.com with ESMTP (
 WorldSecure Server SMTP Relay(WSS) v4.3); Mon, 17 Apr 00 11:59:25 -0700
X-Server-Uuid: b85f21a3-cfd1-11d3-8401-00104bf46ab7
Subject: Re: May 27th PARTY!
To: rstidolph at austin.rr.com
Cc: AMitchell at origin.ea.com, bryanandlacey at earthlink.net,
	carma at carmapro.com, cujorover at earthlink.net, Disalphafu at aol.com,
	eovelman at origin.ea.com, falconson at nettally.com,
	fischer at tsgc.utexas.edu, funny at phord.com, goldie.mcelroy at attws.com,
	joe.maloukis at mail.va.gov, jtaylor at gfbank.com, jttaylor at flash.net,
	karyn at nightfire.com, kevin at aestechnology.com,
	Kizzy_Williams at yahoo.com, kmagill at prysm.net, knaack at gwatt.com,
	lachtaylor at yahoo.com, lbyrne at creativelabs.com, lindt at mail.utexas.edu,
	lmwwelsh at yahoo.com, lrosens at home.com, mar4 at texas.net,
	matsaleh at texas.net, matw at player1.com, mbakonyi at pencom.com,
	mbutler at esginc.com, mcginnis at texas.net, melanie at vsearch.com,
	mhornbuc at etec.com, mikeb at pencom.com, mike.cato at amd.com,
	mjerman at itrsjobs.com, mjgreathouse at juno.com, mkasper at mshow.com,
	monty_hornbuckle at etec.com, mrcambron at earthlink.net, MThor at aol.com,
	pbuchorn at gator1.brazosport.cc.tx.us, postmaster at novacell.com,
	python-list at python.org, ra6200 at email.sps.mot.com, rayjunem at juno.com,
	Ray_Grosvenor at Dell.com, rbarton at bbs.wiztower.com, rds at attymail.com,
	rgrosvenor at austin.rr.com, rong at cmsopen.com, rstidolph at austin.rr.com,
	san2edb at ups.com, scotts at illuminactive.com, seans at cmsopen.com,
	sitnalta at cableone.net, skip at Aureal.com, Srhughes30 at aol.com,
	stidolph at origin.ea.com, stone at golden-gate.org, tandmel at juno.com,
	Tanea_Greenwood at amat.com, tech_support at Quantum3D.com,
	terri.calloway at stanford.edu, tgif at iconn.net, Tiggerree at aol.com,
	tjhuff at msn.com, tlujan1 at hotmail.com, Tonya.Ochiltree at ci.austin.tx.us,
	tracie at austin.rr.com, tsmith at cs.utexas.edu, tt at peakware.com,
	Veronicahu at aol.com, vocooper at juno.com, WanChef at aol.com,
	wayne.stidolph at lutris.com, y2krdi at aol.com
X-Mailer: Lotus Notes Release 5.0.1 July 16, 1999
Message-ID: <OF25503401.CB160CEC-ON882568C4.00680F11 at peoplesoft.com>
Date: Mon, 17 Apr 2000 11:56:55 -0700
X-MIMETrack: Serialize by Router on SMTPOUT/Server/PSExternal(Release
 5.0.2b |December 16, 1999) at 04/17/2000 12:05:20 PM
MIME-Version: 1.0
X-WSS-ID: 14E580871000386-01-01
Content-Type: text/plain; 
 charset=iso-8859-1
Content-Transfer-Encoding: quoted-printable
Sender: python-list-admin at python.org
Errors-To: python-list-admin at python.org
X-BeenThere: python-list at python.org
X-Mailman-Version: 2.0beta3
Precedence: bulk
List-Id: General discussion list for the Python programming language <python-list.python.org>

Sounds like a blast...What should I bring?

Nate


                                                                       =
                                             =20
                    "Rebecca                                           =
                                             =20
                    Stidolph"              To:     "Wayne Stidolph" <wa=
yne.stidolph at lutris.com>, <WanChef at aol.com>, =20
                    <rstidolph at aust        "Viele Humor" <funny at phord.c=
om>, "Veronica Hughes" <Veronicahu at aol.com>, =20
                    in.rr.com>             "Val Cooper" <vocooper at juno.=
com>, "Trevor Lind" <lindt at mail.utexas.edu>, =20
                                           "Tracie Thor" <tracie at austin=
.rr.com>, <Tonya.Ochiltree at ci.austin.tx.us>, =20
                    04/15/00 08:47         "Tom Taylor" <jttaylor at flash=
.net>, "Todd Hughes" <Srhughes30 at aol.com>,   =20
                    PM                     "Tim Huffman" <tjhuff at msn.co=
m>, "Tim and Melody Martin"                  =20
                                           <tandmel at juno.com>, "Tim and=
 Cindy Martin" <mar4 at texas.net>, "Tim"       =20
                                           <tgif at iconn.net>, <Tiggerree=
@aol.com>, "Thad Smith"                      =20
                                           <tsmith at cs.utexas.edu>, "Ter=
ry Thompson" <tt at peakware.com>, "Terri       =20
                                           Calloway" <terri.calloway at st=
anford.edu>, "tech_support"                  =20
                                           <tech_support at Quantum3D.com>=
, "Taylor, Jacki" <jtaylor at gfbank.com>,      =20
                                           "Tanya Lujan" <tlujan1 at hotma=
il.com>, <Tanea_Greenwood at amat.com>, "Stone  =20
                                           Eleazer" <stone at golden-gate.=
org>, "Stidolph, David"                      =20
                                           <stidolph at origin.ea.com>, "S=
teven Jumper" <sitnalta at cableone.net>, "Stan =20
                                           Edwards" <cujorover at earthlin=
k.net>, "Skip McIlvaine" <skip at Aureal.com>,  =20
                                           "Sheridan Layman" <falconson=
@nettally.com>, "Sean Stanley"               =20
                                           <seans at cmsopen.com>, "Scott =
Schimpf" <scotts at illuminactive.com>,         =20
                                           <san2edb at ups.com>, "Ron Gilb=
ert" <rong at cmsopen.com>, "Roja, Elena"       =20
                                           <ra6200 at email.sps.mot.com>, =
"Robert Knaack" <knaack at gwatt.com>, "Richard =20
                                           Barton" <rbarton at bbs.wiztowe=
r.com>, <rgrosvenor at austin.rr.com>, "Rebecca =20
                                           Stidolph" <rstidolph at austin.=
rr.com>, <rds at attymail.com>, "Ray O.         =20
                                           Grosvenor" <Ray_Grosvenor at De=
ll.com>, "Ray and June Minyard"              =20
                                           <rayjunem at juno.com>, "Python=
-List @ python. org"                         =20
                                           <python-list at python.org>, "P=
ython Mailing List" <python-list at cwi.nl>,    =20
                                           "postmaster" <postmaster at nov=
acell.com>, "'Peggy Buchorn'"                =20
                                           <pbuchorn at gator1.brazosport.=
cc.tx.us>, "Ovelman, Elizabeth"              =20
                                           <eovelman at origin.ea.com>, "N=
athan Christensen"                           =20
                                           <Nathan_Christensen at peopleso=
ft.com>, "Monty Hornbuckle"                  =20
                                           <monty_hornbuckle at etec.com>,=
 "Monty Hornbuckle" <mhornbuc at etec.com>,     =20
                                           "Mitchell, Andy" <AMitchell@=
origin.ea.com>, "Mike Jerman"                =20
                                           <mjerman at itrsjobs.com>, "Mik=
e Greathouse" <mjgreathouse at juno.com>, "Mike =20
                                           Cato" <mike.cato at amd.com>, "=
Michael Bakonyi" <mikeb at pencom.com>, "Michael=20
                                           & Treasa McGinnis" <mcginnis=
@texas.net>, "Melinda Kasper"                =20
                                           <mkasper at mshow.com>, "Melani=
e R. Cambron" <melanie at vsearch.com>, "Melanie=20
                                           Cambron" <mrcambron at earthlin=
k.net>, "Mcelroy, Goldie"                    =20
                                           <goldie.mcelroy at attws.com>, =
<mbutler at esginc.com>, <mbakonyi at pencom.com>, =20
                                           "Matt Weathers" <matw at player=
1.com>, "Matt Walker" <matsaleh at texas.net>,  =20
                                           "Mark Thor" <MThor at aol.com>,=
 "Mark Fischer" <fischer at tsgc.utexas.edu>,   =20
                                           "Mark Fischer" <fischer at csr.=
utexas.edu>, "Margaret Warren"               =20
                                           <carma at carmapro.com>, "Malou=
kis, Joe" <joe.maloukis at mail.va.gov>, "Leon  =20
                                           Rosenhein" <lrosens at home.com=
>, "Lbyrne" <lbyrne at creativelabs.com>,       =20
                                           "Laurie Welsh" <lmwwelsh at yah=
oo.com>, "Lach Taylor"                       =20
                                           <lachtaylor at yahoo.com>, "Kur=
t Schmidt" <y2krdi at aol.com>,                 =20
                                           "Kizzy_Williams at amat.com" <K=
izzy_Williams at yahoo.com>, "Kevin Cooper"     =20
                                           <kevin at aestechnology.com>, "=
Kelsi" <Disalphafu at aol.com>, "Kelley Magill" =20
                                           <kmagill at prysm.net>, "Karyn =
Christensen" <karyn at nightfire.com>, "A1C     =20
                                           Bryan A. and Mrs. Lacey D. B=
uchorn" <bryanandlacey at earthlink.net>        =20
                                           cc:                         =
                                             =20
                                           Subject:     May 27th PARTY!=
                                             =20
                                                                       =
                                             =20






ATTENTION FRIENDS AND FAMILY AROUND THE  NATION!!

We (David and Rebecca Stidolph) are celebrating our  17th wedding
anniversary with a big BBQ bash on Saturday, May 27th!=A0 We  want ever=
yone
to come, so we're giving plenty of notice.=A0  Please set aside the lat=
e
afternoon and evening to have lots of fun!

Please reply if you would like to come, so we can  include you in futur=
e
e-mails for more information.

Love,
David and Rebecca, Kelley and Kenneth

PS Sorry for any duplicate e-mails.=A0 Let us  know if you got more tha=
n two.
Thanks!

=





Return-Path: <rstidolph at austin.rr.com>
Delivered-To: python-list at python.org
Received: from hera.cwi.nl (hera.cwi.nl [192.16.191.1])
	by dinsdale.python.org (Postfix) with ESMTP id 6BEBB1CDBA
	for <python-list at python.org>; Sat, 15 Apr 2000 23:45:05 -0400 (EDT)
Received: from mail.austin.rr.com (sm1.texas.rr.com [24.93.35.54]) by hera.cwi.nl with ESMTP
	id FAA22703 for <python-list at cwi.nl>; Sun, 16 Apr 2000 05:47:35 +0200 (MET DST)
Received: from play1 ([24.160.3.64]) by mail.austin.rr.com  with Microsoft SMTPSVC(5.5.1877.357.35);
	 Sat, 15 Apr 2000 22:48:00 -0500
Message-ID: <006201bfa756$896e5a80$da00a8c0 at austin.rr.com>
From: "Rebecca Stidolph" <rstidolph at austin.rr.com>
To: "Wayne Stidolph" <wayne.stidolph at lutris.com>, <WanChef at aol.com>,
	"Viele Humor" <funny at phord.com>,
	"Veronica Hughes" <Veronicahu at aol.com>,
	"Val Cooper" <vocooper at juno.com>,
	"Trevor Lind" <lindt at mail.utexas.edu>,
	"Tracie Thor" <tracie at austin.rr.com>,
	<Tonya.Ochiltree at ci.austin.tx.us>, "Tom Taylor" <jttaylor at flash.net>,
	"Todd Hughes" <Srhughes30 at aol.com>, "Tim Huffman" <tjhuff at msn.com>,
	"Tim and Melody Martin" <tandmel at juno.com>,
	"Tim and Cindy Martin" <mar4 at texas.net>, "Tim" <tgif at iconn.net>,
	<Tiggerree at aol.com>, "Thad Smith" <tsmith at cs.utexas.edu>,
	"Terry Thompson" <tt at peakware.com>,
	"Terri Calloway" <terri.calloway at stanford.edu>,
	"tech_support" <tech_support at Quantum3D.com>,
	"Taylor, Jacki" <jtaylor at gfbank.com>,
	"Tanya Lujan" <tlujan1 at hotmail.com>, <Tanea_Greenwood at amat.com>,
	"Stone Eleazer" <stone at golden-gate.org>,
	"Stidolph, David" <stidolph at origin.ea.com>,
	"Steven Jumper" <sitnalta at cableone.net>,
	"Stan Edwards" <cujorover at earthlink.net>,
	"Skip McIlvaine" <skip at Aureal.com>,
	"Sheridan Layman" <falconson at nettally.com>,
	"Sean Stanley" <seans at cmsopen.com>,
	"Scott Schimpf" <scotts at illuminactive.com>, <san2edb at ups.com>,
	"Ron Gilbert" <rong at cmsopen.com>,
	"Roja, Elena" <ra6200 at email.sps.mot.com>,
	"Robert Knaack" <knaack at gwatt.com>,
	"Richard Barton" <rbarton at bbs.wiztower.com>,
	<rgrosvenor at austin.rr.com>,
	"Rebecca Stidolph" <rstidolph at austin.rr.com>, <rds at attymail.com>,
	"Ray O. Grosvenor" <Ray_Grosvenor at Dell.com>,
	"Ray and June Minyard" <rayjunem at juno.com>,
	"Python-List @ python. org" <python-list at python.org>,
	"Python Mailing List" <python-list at cwi.nl>,
	"postmaster" <postmaster at novacell.com>,
	"'Peggy Buchorn'" <pbuchorn at gator1.brazosport.cc.tx.us>,
	"Ovelman, Elizabeth" <eovelman at origin.ea.com>,
	"Nathan Christensen" <Nathan_Christensen at peoplesoft.com>,
	"Monty Hornbuckle" <monty_hornbuckle at etec.com>,
	"Monty Hornbuckle" <mhornbuc at etec.com>,
	"Mitchell, Andy" <AMitchell at origin.ea.com>,
	"Mike Jerman" <mjerman at itrsjobs.com>,
	"Mike Greathouse" <mjgreathouse at juno.com>,
	"Mike Cato" <mike.cato at amd.com>,
	"Michael Bakonyi" <mikeb at pencom.com>,
	"Michael & Treasa McGinnis" <mcginnis at texas.net>,
	"Melinda Kasper" <mkasper at mshow.com>,
	"Melanie R. Cambron" <melanie at vsearch.com>,
	"Melanie Cambron" <mrcambron at earthlink.net>,
	"Mcelroy, Goldie" <goldie.mcelroy at attws.com>, <mbutler at esginc.com>,
	<mbakonyi at pencom.com>, "Matt Weathers" <matw at player1.com>,
	"Matt Walker" <matsaleh at texas.net>, "Mark Thor" <MThor at aol.com>,
	"Mark Fischer" <fischer at tsgc.utexas.edu>,
	"Mark Fischer" <fischer at csr.utexas.edu>,
	"Margaret Warren" <carma at carmapro.com>,
	"Maloukis, Joe" <joe.maloukis at mail.va.gov>,
	"Leon Rosenhein" <lrosens at home.com>,
	"Lbyrne" <lbyrne at creativelabs.com>,
	"Laurie Welsh" <lmwwelsh at yahoo.com>,
	"Lach Taylor" <lachtaylor at yahoo.com>,
	"Kurt Schmidt" <y2krdi at aol.com>,
	"Kizzy_Williams at amat.com" <Kizzy_Williams at yahoo.com>,
	"Kevin Cooper" <kevin at aestechnology.com>,
	"Kelsi" <Disalphafu at aol.com>, "Kelley Magill" <kmagill at prysm.net>,
	"Karyn Christensen" <karyn at nightfire.com>,
	"A1C Bryan A. and Mrs. Lacey D. Buchorn" <bryanandlacey at earthlink.net>
Subject: May 27th PARTY!
Date: Sat, 15 Apr 2000 22:47:16 -0500
MIME-Version: 1.0
Content-Type: multipart/alternative;
	boundary="----=_NextPart_000_005C_01BFA72C.8C651E60"
X-Priority: 3
X-MSMail-Priority: Normal
X-Mailer: Microsoft Outlook Express 5.00.2919.6600
X-MimeOLE: Produced By Microsoft MimeOLE V5.00.2919.6600
Sender: python-list-admin at python.org
Errors-To: python-list-admin at python.org
X-BeenThere: python-list at python.org
X-Mailman-Version: 2.0beta3
Precedence: bulk
List-Id: General discussion list for the Python programming language <python-list.python.org>

This is a multi-part message in MIME format.

------=_NextPart_000_005C_01BFA72C.8C651E60
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable

ATTENTION FRIENDS AND FAMILY AROUND THE NATION!!

We (David and Rebecca Stidolph) are celebrating our 17th wedding =
anniversary with a big BBQ bash on Saturday, May 27th!  We want everyone =
to come, so we're giving plenty of notice.  Please set aside the late =
afternoon and evening to have lots of fun! =20

Please reply if you would like to come, so we can include you in future =
e-mails for more information.

Love,=20
David and Rebecca, Kelley and Kenneth

PS Sorry for any duplicate e-mails.  Let us know if you got more than =
two.  Thanks!

------=_NextPart_000_005C_01BFA72C.8C651E60
Content-Type: text/html;
	charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML><HEAD>
<META content=3D"text/html; charset=3Diso-8859-1" =
http-equiv=3DContent-Type>
<META content=3D"MSHTML 5.00.3013.2600" name=3DGENERATOR>
<STYLE></STYLE>
</HEAD>
<BODY bgColor=3D#ffffff>
<DIV><FONT face=3DArial size=3D2>ATTENTION FRIENDS AND FAMILY AROUND THE =

NATION!!</FONT></DIV>
<DIV>&nbsp;</DIV>
<DIV><FONT face=3DArial size=3D2>We (David and Rebecca Stidolph) are =
celebrating our=20
17th wedding anniversary with a big BBQ bash on Saturday, May =
27th!&nbsp; We=20
want everyone to come, so we're giving <STRONG>plenty </STRONG>of =
notice.&nbsp;=20
Please set aside the late afternoon and evening to have lots of =
fun!&nbsp;=20
</FONT></DIV>
<DIV>&nbsp;</DIV>
<DIV><FONT face=3DArial size=3D2>Please reply if you would like to come, =
so we can=20
include you in future e-mails for more information.</FONT></DIV>
<DIV>&nbsp;</DIV>
<DIV><FONT face=3DArial size=3D2>Love, </FONT></DIV>
<DIV><FONT face=3DArial size=3D2>David and Rebecca, Kelley and =
Kenneth</FONT></DIV>
<DIV>&nbsp;</DIV>
<DIV><FONT face=3DArial size=3D2>PS Sorry for any duplicate =
e-mails.&nbsp; Let us=20
know if you got more than two.&nbsp; Thanks!</FONT></DIV></BODY></HTML>

------=_NextPart_000_005C_01BFA72C.8C651E60--




From loredo at spacenet.tn.cornell.edu  Thu Apr 13 16:22:47 2000
From: loredo at spacenet.tn.cornell.edu (Tom Loredo)
Date: Thu, 13 Apr 2000 16:22:47 -0400
Subject: Need full example of C extension on Macintosh (CW)
References: <120420000714412887%pecora@anvil.nrl.navy.mil>
Message-ID: <38F62C97.FD5073B@spacenet.tn.cornell.edu>

Lewis-

I'm putting together some project stationery for CodeWarrior 5 for
simple numerical extensions (i.e., just returning the value of
a function; or using NumPy arrays).  Are you using CW5?  If so,
I can send you something.

To get this to work, I had to "borrow" some stuff from the source
distribution; some files were missing from the binary distribution,
even though it is advertised as suitable for "plugin" development.
But I think what was missing only affects NumPy extensions.

-Tom Loredo


From alwagner at tcac.net  Thu Apr 20 21:03:35 2000
From: alwagner at tcac.net (Albert Wagner)
Date: Thu, 20 Apr 2000 20:03:35 -0500
Subject: os.listdir & os.stat fail on windows zipdrive
References: <38FF9F53.A4698EE6@tcac.net>
Message-ID: <38FFA8E7.8D926F32@tcac.net>

Crap, I forgot the code:

    def descendTree(self, dir):
        '''
        recursively descend the directory tree rooted at dir,
        save each directory and its files as an entry in a Dictionary.
        Each directory pathname is a key with a List for a value.
        Each filename is placed in the list for its directory.
        '''
        self.tree[dir] = []
        for aName in os.listdir(dir):
            dirName = '%s/%s' % (dir, aName)
            try: 
                mode = os.stat(dirName)[ST_MODE]
            except:
                self.tree[dir].append("(%s)" % aName)
                continue           
            if S_ISDIR(mode):            
                if not self.tree.has_key(dirName):
                    self.tree[dirName] = []
                # recurse into directory
                self.descendTree(dirName)           
            elif S_ISREG(mode):
                self.tree[dir].append(aName)
            else:
                self.tree[dir].append("(%s)" % aName)
   

-- 
Small is Beautiful


From smalleys at gte.net  Sat Apr 15 14:20:19 2000
From: smalleys at gte.net (sue)
Date: Sat, 15 Apr 2000 18:20:19 GMT
Subject: Building Core Python
References: <213CDEBFDDA7D3118081009027E40C56071139@SHEF-MS1>
Message-ID: <38F8A4E4.582F5451@gte.net>

Have you checked the "small python" from
http://www.lyra.org/greg/small/

Sue.


From tseaver at starbase.neosoft.com  Fri Apr 21 20:21:52 2000
From: tseaver at starbase.neosoft.com (Tres Seaver)
Date: 21 Apr 2000 19:21:52 -0500
Subject: Using python on the web
References: <EDFD2A95EE7DD31187350090279C6767754BA8@THRESHER> <8dq9lj$pjg$2@newshost.accu.uu.nl>
Message-ID: <30317B4110C2CF2B.69B375EBD430B2E8.1EDF515EC1E9B219@lp.airnews.net>

In article <8dq9lj$pjg$2 at newshost.accu.uu.nl>,
Martijn Faassen <m.faassen at vet.uu.nl> wrote:
>Brent Fulgham <brent.fulgham at xpsystems.com> wrote:
>> A couple of questions:
>
>>> We build medium-busy, CGI-intensive sites in Python. We were briefly
>>> quite excited about Zope, but once we realized it locked us into Yet
>>> Another Programming Language, we lost interest. DHTML is particularly
>>> problematic,
>> 1.  I'm not familiar with Zope.  Is DHTML specific to Zope?  It seems
>> to be proudly advertised on a lot of "webware" brochures I see these
>> days.
>
>Acronym mixup. DHTML (Dynamic HTML) and DTML (Document Template Markup Language)
>have little in common besides that they both are used with HTML. DHTML is
>client side (in the browser), DTML server side (in Zope). DHTML is to do
>fancy graphics effects, DTML is a reporting language that talks to Zope
>(read Python) objects and usually generates HTML.

The closest cognate I know of to DTML in the non-Zope world is "server-side
includes," and their siblings.  It is not a coincidence that Zope supports
three distinct DTML syntaxes:

  * One based on Python "%"-substitution (the oldest, I think; almost nobody
	uses this now);

  * One using the same '<!--#...-->' style of syntax as SSI: this is
	comfortable for developers used to SSI, but doesn't play nice with
	design tools;

  * The now-dominant "pseudo-XML" style ('<dtml-var...>'), specifically
	designed to be teachable to DreamWeaver, et al.

DTML is a "document template" language, logically equivalent to building
a document via Python string substituion, but with conditional and iterative
processing constructs added in.  It is *not* adequate for doing "real"
programming, although many early Zope apps used it so, because it was
the only way to do logic "through-the-web."

What's-one-more-letter-between-friends'ly,

Tres.
-- 
---------------------------------------------------------------
Tres Seaver        tseaver at digicool.com     http://www.zope.org
Digital Creations


From Alessandro.Bottoni at think3.com  Mon Apr 17 11:48:47 2000
From: Alessandro.Bottoni at think3.com (Alessandro Bottoni)
Date: Mon, 17 Apr 2000 17:48:47 +0200
Subject: Any Web-Mail Gateway (written in Python)?
Message-ID: <6D8A17398E28D3119F860090274DD7DB4986A3@pces.cadlab.it>

I'm looking for a web-mail gateway (written in Python, if possible) for
making accessible the Postfix mail boxes of my users through my PyApache web
server (both running under Linux, of course).
Any suggestion?
TIA
----------------------------------------------------------------
Alessandro Bottoni (Alessandro.Bottoni at Think3.com)
Web Programmer @ Think3 inc. (www.think3.com)
I do not speak for think3 and they return the favour




From andrew at andrewcooke.free-online.co.uk  Sat Apr 22 04:33:42 2000
From: andrew at andrewcooke.free-online.co.uk (Andrew Cooke)
Date: Sat, 22 Apr 2000 08:33:42 GMT
Subject: LAmbda calculus evaluator (interactive)
Message-ID: <8dro55$eq6$1@nnrp1.deja.com>


Hi,

Does anyone, by any chance, know of an interactive lambda fuinction
evaluator/editor?  Preferably in Python as that's a language I
understand that's common to both platforms I'd like to run this on.

I've found source to a simple evaluator on this ng, but am looking for
something that lets you single step, do abstractions as well as
reductions, step different evaluation strategies (lazy or by value etc).

This is just for untyped calculus (at the moment - I see that in a few
chapters more I might need something similar with types...).  I know
where I can get various (Oca)ML programs, but I'm not sure they're
interactive and for various reasons I don't want to install that
language at the moment.

Thanks,
Andrew

http://www.andrewcooke.free-online.co.uk/index.html


Sent via Deja.com http://www.deja.com/
Before you buy.


From kilian.lorenz at dkfz-heidelberg.de  Sun Apr  9 17:01:47 2000
From: kilian.lorenz at dkfz-heidelberg.de (Kilian Lorenz)
Date: Sun, 09 Apr 2000 23:01:47 +0200
Subject: Tkinter:  centering windows on screen
References: <ak1I4.121418$Dv1.1306619@news1.rdc1.bc.home.com> <38F0C6A7.10878609@dkfz-heidelberg.de> <38F0C95D.1F7E853@dkfz-heidelberg.de> <xR4I4.274$rc9.188784640@newsb.telia.net>
Message-ID: <38F0EFBB.F39F4A6F@dkfz-heidelberg.de>

Fredrik Lundh wrote:
> 
> if you browsed the source, how come you didn't find the
> corresponding Tkinter methods?
> 

Ooooops, because I'm kind of an idiot ...
(and as a paltry excuse I started using Tkinter just yesterday)

windows geometry:
   I didn't realize what "geometry = wm_geometry" means
file dialogues:
   I only searched in Tkinter.py itself

... sorry ... you may torture me with soft cussions ...


From aa8vb at yahoo.com  Wed Apr  5 12:42:58 2000
From: aa8vb at yahoo.com (Randall Hopper)
Date: Wed, 05 Apr 2000 12:42:58 -0400
Subject: Tkinter:  "Get root for a widget" is private?
In-Reply-To: <wk8zysd6zc.fsf@SNIPE.maya.com>
References: <20000331112048.A1257432@vislab.epa.gov>
 <QdEF4.892$n68.190327808@newsb.telia.net> <m3snx3n1x0.fsf@devel.no.rwd.dinet>
 <20000331112048.A1257432@vislab.epa.gov>
 <QdEF4.892$n68.190327808@newsb.telia.net>
 <20000403064335.A1560771@vislab.epa.gov>
 <2Q_F4.1092$n68.187097600@newsb.telia.net>
 <20000405073526.A1760238@vislab.epa.gov> <wk8zysd6zc.fsf@SNIPE.maya.com>
Message-ID: <20000405124258.A1747369@vislab.epa.gov>

Jeff Senn:
 |Randall Hopper <aa8vb at yahoo.com> writes:
 |> I have a timer class built around Tk timers.  I realize I could just pass
 |> any_old_widget into this routine, and it would store it off into state data
 |> and use it periodically to access the 'after' and 'after_cancel' methods.
 |> 
 |> However, this timer has nothing to do with this widget in particular, so
 |> why should it have a reference to it (versus the button next to it, or the
 |> list widget below it)?  It makes more sense for the timer class to accept
 |> and bookmark the root object.  This root is associated with the global Tk
 |> state, and it makes sense for a timer to bookmark this to use global Tk
 |> state commands.
 |
 |Unless you're doing something wacky (creating multiple roots??), you
 |could just do what Tkinter itself (generally) does and use
 |Tkinter._default_root to get the instance of the root object....

Right.  I'm using _root() now which works.  My original point wasn't that I
couldn't get root from a widget, but that to do it I violate the declared
public/private/protected policy for Tkinter.  Since Python has no enforced
access control, I can poke at object internals anytime, but I really want
to avoid this.

My understanding is that .<symbol> is public, ._<symbol> is protected, and
.__<symbol> is private.

If _root() was ripped out in the next version of Python, I wouldn't have a
leg to stand on.  And by all rights it's protected so removing it should
not affect me (or any other external code), unless I (or it) was deriving
off this class.

As Fredrik mentioned I could just snitch global Tk commands off
any_old_widget, but an event loop timer object that has nothing to do with
"any" widget shouldn't be keeping a handle to any_old_widget.  That's not
clean code.

An alternative is just to pass root around everywhere, but that seems silly
if every widget already knows its root.

So my request was for _root() to be declared public (e.g. as .root())

-- 
Randall Hopper
aa8vb at yahoo.com



From mwh21 at cam.ac.uk  Sat Apr  1 09:33:27 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 01 Apr 2000 15:33:27 +0100
Subject: bytecodehacks-April2000
Message-ID: <m3wvmhopqg.fsf@atrus.jesus.cam.ac.uk>

Well, it's April the first, so what better time than this to announce
the release of a new version of bytecodehacks?

Not all that much has changed in the code; a bug fix here and there.

* The mechanism for calculating the co_stacklevel attribute of code
  objects actually gets it right (at least in cases I have tested).

* xapply.xapply has gotten *much* cleverer, and much hairier.  It now
  supports such delights as keyword arguments, trimming the default
  argument list correctly, assignments to arguments and handling
  **-style arguments (but not *-style, yet).  It's approaching the
  point of being actually useful (scary!).

* there's the beginnings of a test suite.

Administrivia-wise, there's much more shaking up.

* bytecodehacks now live on sourceforge; start here to find the
  tarballs and updated docs.

    http://bytecodehacks.sourceforge.net

  There's a mailing list, which I don't envision being very high
  volume:

    http://lists.sourceforge.net/mailman/listinfo/bytecodehacks-discuss

  but I invite anyone who's interested to join for banter about the
  internal structures of Python and how to (ab)use them.

* bytecodehacks now has a license, based on the deliberately liberal
  Python license; please read the file LICENSE for details.

* Documentation is now being distributed separately; a tarball of the
  HTML can be found on the project homepage.

As ever, all comments are welcome!

Cheers,
Michael

PS: I haven't sent this to python-announce, as that list appears to be
entirely dead.  News to the contrary will be gratefully accepted.

-- 
well, take it from an old hand: the only reason it would be easier
to program in C is that you can't easily express complex  problems
in C, so you don't.                 -- Erik Naggum, comp.lang.lisp


From ulf.engstrom at b2b-link.com  Wed Apr 12 07:26:12 2000
From: ulf.engstrom at b2b-link.com (=?iso-8859-1?Q?Ulf_Engstr=F8m?=)
Date: Wed, 12 Apr 2000 13:26:12 +0200
Subject: Getting an overview if all available driveletters on NT
References: <2B1262E83448D211AE4B00A0C9D61B03BA7294@MSGEURO1> <_FGI4.1194$Za1.20497@newsc.telia.net> <00ed01bfa442$53c66150$858647c1@DEVELOPMENT> <8d1hak$25p$1@nnrp1.deja.com>
Message-ID: <007501bfa471$e89bfd60$858647c1@DEVELOPMENT>

Well, Yes, the original question was that, but there's no such thing as a
GetDriveType in my win32api at least. As I said, that's all I find in the
win32api that has to do with the drives, but someone else might now better
:)
Some digging around and I found the right module, the GetDriveType is in the
win32file package :) And there we go :)
Regards
Ulf

----- Original Message -----
From: <fredrik at pythonware.com>
Newsgroups: comp.lang.python
To: <python-list at python.org>
Sent: Wednesday, April 12, 2000 11:57 AM
Subject: Re: Getting an overview if all available driveletters on NT


> Ulf Engstrom wrote:
> > As the effbot told you the win32api is probably the one you wanna
> use, but
> > the function is GetLogicalDriveStrings() This won't gives you all the
> the
> > driveletters, but not their labels, there might be something like that
> > there, just never have had to use it. Someone else might now that
> part?
> > >>> win32api.GetLogicalDriveStrings()
> > 'A:\\\000C:\\\000D:\\\000E:\\\000'
>
> umm.  didn't the original poster ask for a way to
> tell ordinary drives from network drives?  that's
> exactly what GetDriveType does...
>
> </F>
>
>
> Sent via Deja.com http://www.deja.com/
> Before you buy.
> --
> http://www.python.org/mailman/listinfo/python-list




From darrell at dorb.com  Fri Apr 28 20:22:43 2000
From: darrell at dorb.com (Darrell)
Date: Sat, 29 Apr 2000 00:22:43 GMT
Subject: command pattern/framework in python?
References: <8ebpl4$eq9$1@nnrp1.deja.com>
Message-ID: <nXpO4.17469$GP3.124141@typhoon.nyroc.rr.com>

<tiddlerdeja at my-deja.com> wrote in message
> Anyone got an example of a Command pattern/framework that they have
> used in python?
>

This pattern wraps an action or command in an instance with an execute
method.
The instance can be passed around like a function pointer.

def demo(a,b,c):
    print 'a:',a
    print 'b:',b
    print 'c:',c

class Command:
    def __init__(self, cmd, *args):
        self._cmd=cmd
        self._args=args

    def __call__(self, *args):
       return apply(self._cmd, self._args+args)


cmd=Command(dir,__builtins__)
print cmd()

cmd=Command(demo,1,2)
cmd(3)

--Darrell Gallion




From order at ownbox.com  Mon Apr 10 02:19:38 2000
From: order at ownbox.com (order at ownbox.com)
Date: Mon, 10 Apr 2000 06:19:38 GMT
Subject: Free Money Making Reports
Message-ID: <_neI4.24192$d5.66171@typhoon1.san.rr.com>

Check it out for yourself http://www.geocities.com/freemoneymakingsystem/index.html


From garry at sage.att.com  Fri Apr 28 09:44:41 2000
From: garry at sage.att.com (Garry Hodgson)
Date: Fri, 28 Apr 2000 13:44:41 GMT
Subject: Running shell programs from Python
References: <38FC64DA.7D297563@ffi.no> <8di47s$mkg$1@nntp6.u.washington.edu> <390485A8.B5ADAAE9@sage.att.com> <slrn8g92q5.5d6.grant@grante.comtrol.com>
Message-ID: <390995C9.F49AF78A@sage.att.com>

Grant Edwards wrote:
> 
> In article <390485A8.B5ADAAE9 at sage.att.com>, Garry Hodgson wrote:
> >Donn Cave wrote:
> >
> >> The references are mainly UNIX man pages pipe(2), fork(2), fcntl(2)
> >> (look for dup2), exec(2) and wait(2).
> >>
> >> Basically, the sequence is repeat ( open pipes, fork, close pipes )
> >> After a fork, the parent and child both have both ends of each pipe
> >> open, and it's important to close the unused "write" ends so the
> >> pipe will generate an EOF when its real writer exits.
> >
> >i've always been curious why this wasn't encapsulated as a single
> >command in the os module.  i use it infrequently enough that i always
> >need to go look the idiom, or skip it for simpler things.  seems like
> >a single command that returned a process id, and a pair of file
> >descriptors would have lots of uses.
> 
> You mean like popen2?

well...yeah, kinda like that, except...
uhhhh...one that i'd know about.

thanks for the clue.  i needed one.

-- 
Garry Hodgson                   Every night 
garry at sage.att.com                a child is born
Software Innovation Services        is a Holy Night.
AT&T Labs                         - Sophia Lyon Fahs


From tjg at avalongroup.net  Tue Apr  4 17:52:34 2000
From: tjg at avalongroup.net (Timothy Grant)
Date: Tue, 04 Apr 2000 14:52:34 -0700
Subject: Two questions
References: <38EA5FE8.853E9B75@callware.com> <38EA6508.90BF630C@roguewave.com> <38EA6E87.40EDB542@callware.com>
Message-ID: <38EA6422.61054B11@exceptionalminds.com>

Ivan Van Laningham wrote:
> Well, that's two recommendations.  And judging from previous Marc-Andre
> packages I've used, it'll be a snap.
> 
> Steve Holden, in another post, said that Tkinter should work fine along
> with mxODBC.  That's a relief, but does anyone have direct personal
> experience with doing so?

Ivan,

mxODBC works beautifully, but I've never used it to access and access
database, only SQL database. Perchance do you know how Windows
advertises an Access database to an ODBC driver?

It's amazing how often I come here with a question only to find someone
else asking the same question I came with!

-- 
Stand Fast,
    tjg.

Timothy Grant                         tjg at exceptionalminds.com
Chief Technology Officer              www.exceptionalminds.com
Red Hat Certified Engineer                      (503) 246-3630
Avalon Technology Group, Inc.              fax  (503) 246-3124
>>>>>>>>>>>>Linux...Because rebooting isn't normal<<<<<<<<<<<<



From jon at dgs.monash.edu.au  Wed Apr 12 02:50:52 2000
From: jon at dgs.monash.edu.au (Jonathan Giddy)
Date: Wed, 12 Apr 2000 16:50:52 +1000 (EST)
Subject: Calling os.waitpid() From Secondary Thread Under Linux 2.2.12-20?
In-Reply-To: <38f363e5.2931715373@news> from "Randy Wiser" at Apr 11, 2000 05:48:11 PM
Message-ID: <200004120650.QAA22353@nexus.csse.monash.edu.au>

Randy Wiser declared:
>
>I've written a small python program that spawns one child process from
>its main thread, and then creates a second thread.  The sole purpose
>of the second thread (at this point) is to call os.waitpid(processID,
>0) to wait for the child that the main thread created.  This works
>fine under Solaris when I call os.waitpid() from the second thread. 
>
>But, under Linux 2.2.12-20 I get the following OSError exception:
>
>[Errno 10] No child processes

I recently had this reported as a problem in code I developed on a Sun,
which was then run on Linux.  I believe it is due to Linux threads being
separate processes, hence the second thread is not the spawned process's
parent and cannot wait for it.  (Presumably, if pthread_join() works, then
waitpid() should be able to work too, since they must be the same function 
on Linux).

Anyway, my (non-)solution was to redesign my code to wait from the same 
thread as the fork().  If you can't move the waitpid() into the main thread,
maybe you can move the fork() into the second thread.

Jon.



From c.hintze at gmx.net  Sat Apr 22 10:36:55 2000
From: c.hintze at gmx.net (Clemens Hintze)
Date: Sat, 22 Apr 2000 14:36:55 GMT
Subject: Distribution
References: <39013f48@flexnet239.uunt.net>
Message-ID: <m3u2gurym0.fsf@qiao.localnet>

>>>>> "Hdman" == Highdesertman  <HighdesertmanREMOVE at yahoo.com> writes:

...

    Hdman> Now to my question:

    Hdman> If I use python to develop an application for the windows
    Hdman> 9x environment, I have already downloaded the appropriate
    Hdman> libraries and version) what form does the finished program
    Hdman> take? How does it get distributed? I remember in the early
    Hdman> days of BASIC programming, (before compiled basic) your
    Hdman> program was just raw code that could be (often
    Hdman> accidentally) altered or otherwise corrupted. In order for
    Hdman> your basic program to be used, the user had to have a basic
    Hdman> compiler on board and they loaded your raw source from a
    Hdman> text file and executed it. Is this how Python is used?

    I feel that this is similar to how Python is used. The Python
interpreter accepts to sources of input (actually three, to speak the
truth), 1st is plain Python source, 2nd is a bytecode file created by
the interpreter. In detail:

    If you execute a Python source with the interpreter, it will be
compiled to bytecode (machine code of a virtual CPU); this machine
code will be executed by the interpreter simulating that virtual CPU
here. You could also store that machine code into another file and
hand this to the interpreter. The compilation process is not necessary 
then. Heared about Java? Python is similar here. Only compiler and
virtual machine (you may call it 'interpreter') are not separated.

    But that also mean, that your client must have Python installed
normally. But IIRC somebody (Chris Tismer?) has shown a way how to
compile the whole program into an standalone executable with linked-in 
interpreter. Here you would not have to pass the whole Python together 
with your proggy. Perhaps others could clarify this?

...

    Hdman> Cheers,

    Hdman> Mathew

HTH,
Clemens.


From jae at ilk.de  Sat Apr 15 11:51:44 2000
From: jae at ilk.de (Juergen A. Erhard)
Date: Sat, 15 Apr 2000 17:51:44 +0200
Subject: A Mountain of Perl Books + Python Advocacy
In-Reply-To: <8cr4an$j40$2@newshost.accu.uu.nl> (m.faassen@vet.uu.nl)
References: <000f01bf9eca$2eec8c60$3e2d153f@tim> <38ECF4E6.B85C62D5@sightreader.com> <8cr4an$j40$2@newshost.accu.uu.nl>
Message-ID: <15042000.2@sanctum.jae.ddns.org>

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

>>>>> "Martijn" == Martijn Faassen <m.faassen at vet.uu.nl> writes:

    Martijn> [...] Well, but Perl isn't based on a wild, revolutionary
    Martijn> idea either. [...]

Well, maybe not revolutionary, but certainly wild: "Everybody should
program as they speak".  If that ain't wild (and the results even more
so), I don't know what is.

";-) " * 1000.

Bye, J

PS: ";-) " x 999 so the perlers out there get it too ;-).

- -- 
J?rgen A. Erhard      eMail: jae at ilk.de      phone: (GERMANY) 0721 27326
          My WebHome: http://members.tripod.com/Juergen_Erhard
             Internet Movie Database (http://www.imdb.com)
  "Windows NT" is an acronym for "Windows? No thanks." -- Russ McManus
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.1 (GNU/Linux)
Comment: Use Mailcrypt and GnuPG <http://www.gnupg.org/>

iEYEARECAAYFAjj4kA0ACgkQN0B+CS56qs3YhgCcDsNMqbKFIBZQXtPzoZobe/SJ
WMsAnR7z6o72b0sOO1hIlc6245Dpzv7n
=0bQH
-----END PGP SIGNATURE-----



From gmcm at hypernet.com  Thu Apr  6 09:32:10 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Thu, 6 Apr 2000 09:32:10 -0400
Subject: Installer / mxDateTime problem
In-Reply-To: <8chm9o$qbv$1@gxsn.com>
Message-ID: <1257094563-14896354@hypernet.com>

Richard Chamberlain wrote:

> I've got a simple script that uses mxDateTime that I want to run as an exe.
> I run it through simple.py and I get the following error:
> 
> NameError: Case mismatch for module name feasts
> (filename C:\Program Files\Python\DateTime\Feasts.py)

If I attempt to import DateTime.feasts interactively, I get the 
same error. Do you have PYTHONCASEOK set? The exe isn't 
going to do that for you.

- Gordon



From ngps at madcap.dyndns.org  Tue Apr 18 11:57:40 2000
From: ngps at madcap.dyndns.org (Ng Pheng Siong)
Date: 18 Apr 2000 15:57:40 GMT
Subject: Pliant?
Message-ID: <8di0lk$4h0$1@coco.singnet.com.sg>

Hello,

Saw a reference to the language Pliant on Freshmeat today.
Looks pretty interesting. What are your views on Pliant,
as a Python programmer?

Pliant is here: http://pliant.cams.ehess.fr

TIA. Cheers.

-- 
Ng Pheng Siong <ngps at post1.com> * http://www.post1.com/home/ngps



From emile at fenx.com  Wed Apr  5 12:03:03 2000
From: emile at fenx.com (Emile van Sebille)
Date: Wed, 5 Apr 2000 09:03:03 -0700
Subject: reload imported module
Message-ID: <05fd01bf9f18$6e5a33e0$01ffffc0@worldnet.att.net>

I'm testing a new class in its own module from within
idle.  Is there a way to reload y after having done:

    from x import y

or is it easiest to just import x and reload it?

Emile van Sebille
emile at fenx.com
-------------------







From m.faassen at vet.uu.nl  Tue Apr 25 08:35:09 2000
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 25 Apr 2000 12:35:09 GMT
Subject: plugins for browsers?
References: <3904ccb0.32134281@news.demon.co.uk>
Message-ID: <8e43dt$ods$1@newshost.accu.uu.nl>

greg at quokka.demon.co.uk wrote:
> 	Are there any plugins or other means of running Python in Web
> browsers? I seem to recall that there were a while ago ...

JPython:

http://www.jpython.org

Regards,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From ulf.engstrom at b2b-link.com  Wed Apr 12 01:45:36 2000
From: ulf.engstrom at b2b-link.com (=?iso-8859-1?Q?Ulf_Engstr=F8m?=)
Date: Wed, 12 Apr 2000 07:45:36 +0200
Subject: Getting an overview if all available driveletters on NT
References: <2B1262E83448D211AE4B00A0C9D61B03BA7294@MSGEURO1> <_FGI4.1194$Za1.20497@newsc.telia.net>
Message-ID: <00ed01bfa442$53c66150$858647c1@DEVELOPMENT>

As the effbot told you the win32api is probably the one you wanna use, but
the function is GetLogicalDriveStrings() This won't gives you all the the
driveletters, but not their labels, there might be something like that
there, just never have had to use it. Someone else might now that part?
>>> win32api.GetLogicalDriveStrings()
'A:\\\000C:\\\000D:\\\000E:\\\000'
Regards
Ulf

> Pieter Claerhout <PClaerhout at CREO.BE> wrote:
> > is there a module in Python, which can list all available driveletters,
> together with
> > their label. I'm using Windows NT? I already thought of iterating over
all
> possible
> > driveletters, and see if the path exists or not, but I also want to
> separate the
> > network volumes from the other volumes. Any ideas?
>
> just guessing here (don't have win32api on this box),
> but this might do the trick:
>
>     type = win32api.GetDriveType("%s:\\" % drive)
>
> where drive is 'A' through 'Z'.
>
> </F>
>
>
> --
> http://www.python.org/mailman/listinfo/python-list




From cer at tutopia.com  Thu Apr 13 11:55:15 2000
From: cer at tutopia.com (cer at tutopia.com)
Date: Thu, 13 Apr 2000 15:55:15 GMT
Subject: ANN: Tkinter Look
Message-ID: <8d4qkq$l7s$1@nnrp1.deja.com>

I want to share this work related with a Look&Feel package
for Tkinter.

For info and download see :
http://www.geocities.com/cristian_echeverria/Chevegui.html

This has been very usefull to my and maybe somebody can do
a better job with it.

Bye.......Cristian Echeverria


Sent via Deja.com http://www.deja.com/
Before you buy.


From PClaerhout at CREO.BE  Mon Apr 17 22:08:18 2000
From: PClaerhout at CREO.BE (Pieter Claerhout)
Date: Tue, 18 Apr 2000 04:08:18 +0200
Subject: Upgrading IIS to 1.6 problem....
Message-ID: <2B1262E83448D211AE4B00A0C9D61B03BA72C9@MSGEURO1>

If you go to the Internet Services Manager, and you select a web
application, you can edit the application options (in IIS 4 or higher,
a webapp is a folder with the blue icon). If you go to the home directory
tab, you can click on the configure button to change the settings for your
application. If you go there to the App Mapping tabs, to what points the
extension .cgi? My guess it that it's pointing to the python 1.5 
executable and not to the 1.6 executable.

Kind regards,


Pieter

-----Original Message-----
From: Manus Hand
To: python-list at python.org
Sent: 4/17/00 7:58 PM
Subject: Upgrading IIS to 1.6 problem....

I run CGI scripts on a real machine, but I also run some on IIS on an NT
machine.

To configure IIS to know that .cgi files should run Python, I (long ago)
followed Aaron Watters' instructions, and added a "ScriptMap" parameter
in the Windows registry:
    HKEY_LOCAL_MACHINE
        System
            CurrentControlSet
                W3SVC
                    Parameters
                        ScriptMap
                            .cgi -> C:\Program Files\Python\Python.exe
%s %s

Worked just fine for a good long time.

The other day, I downloaded 1.6a2 and installed it on the PC.  BUT...not
at the
same location -- I installed it in the directory D:\Python16

I figured that changing the above registry entry to
"D:\Python16\\python.exe %s %s"
would be all I needed to do and my Python .cgi code would run using 1.6.

Long story short(er), it didn't work.  My .cgi code still runs using the
1.5.2 version
that is in C:\Program Files\Python.  Honest.  I made the registry change
(I promise) and even rebooted.  Twice to be sure.

Anyone know why my registry change hasn't affected IIS yet?

(Oh, I also -- though this shouldn't matter -- made sure to update my
%PATH%
in my autoexec.bat -- now when I type Python at an MS-DOS prompt, I get
1.6a2.)

Let me know.  Thanks,
Manus

-- 
http://www.python.org/mailman/listinfo/python-list



From bill at libc.org  Tue Apr 25 01:18:01 2000
From: bill at libc.org (Bill Anderson)
Date: Mon, 24 Apr 2000 23:18:01 -0600
Subject: Using python on the web
References: <EDFD2A95EE7DD31187350090279C6767754BA8@THRESHER> <8dq6k3$qmb$1@nnrp1.deja.com>
Message-ID: <39052A89.19DC9417@libc.org>

spenrose at my-deja.com wrote:
> 
> In article <EDFD2A95EE7DD31187350090279C6767754BA8 at THRESHER>,
>   Brent Fulgham <brent.fulgham at xpsystems.com> wrote:
> > A couple of questions:
> >
> > > We build medium-busy, CGI-intensive sites in Python. We were briefly
> > > quite excited about Zope, but once we realized it locked us into Yet
> > > Another Programming Language, we lost interest. DHTML is
> > > particularly
> > > problematic,
> > 1.  I'm not familiar with Zope.  Is DHTML specific to Zope?  It seems
> > to be proudly advertised on a lot of "webware" brochures I see these
> > days.
> 
> My limited understanding of Zope is that to run Zope is to use DHTML and
> vice-versa. I welcome corrections on this point.

Nope.

DHTML: Dynamic HTML
Translation: Usually Javascript and/or javascript with css

You may be thinking of DTML: Document Template Markup language


-- 
In flying I have learned that carelessness and overconfidence are 
usually far more dangerous than deliberately accepted risks. 
          -- Wilbur Wright in a letter to his father, September 1900


From effbot at telia.com  Mon Apr 24 03:51:38 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Mon, 24 Apr 2000 07:51:38 GMT
Subject: Tkinter.Button error
References: <3903E8C1.39B3E3D7@swt.edu>
Message-ID: <e2TM4.3781$Za1.55405@newsc.telia.net>

ABC <ns566645 at swt.edu> wrote:
> button=Tkinter.Button(group1.interior(), text='8.00')
> button.pack()
>
> The above program gives an error in TKinter in the second last line.
> I have no clue of what the error is?

how about posting the error message?

</F>




From embed at geocities.com  Wed Apr 26 09:05:32 2000
From: embed at geocities.com (Warren Postma)
Date: Wed, 26 Apr 2000 09:05:32 -0400
Subject: Why doesn't O'Reilly use parrots on its python books?
References: <etdvh186ztw.fsf@w20-575-115.mit.edu> <B5293E3C.255C%dgoodger@bigfoot.com> <FEAC06345F08C2BE.74A6EF6C28FDB863.AED7C7EC5CB151E7@lp.airnews.net>
Message-ID: <NTBN4.12149$HG1.327272@nnrp1.uunet.ca>

> >If they ever run out of fauna and need to go flora, the only choice would
> >be: the larch.
> >
>
> Pruned into shrubbery.

Let us not forget the Lofty Scotts Pine.


Warren




From m.faassen at vet.uu.nl  Thu Apr  6 06:24:38 2000
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 6 Apr 2000 10:24:38 GMT
Subject: Tuples -- who needs 'em
References: <AF223B12F1FDD2119253006097C4FCEC0AD357@cats.pdri.xerox.com>
Message-ID: <8chol6$6em$1@newshost.accu.uu.nl>

Alexander, Bob <BobAlex at uppercase.xerox.com> wrote:

> m.faassen at vet.uu.nl [mailto:m.faassen at vet.uu.nl] wrote:

>> But tuples are nice in particular for their *syntax*. With tuples, you
>> can do:
>> 
>> def foo():
>>     return 1, 2, 3, 4
>> 
>> first, second, third, fourth = foo()
>> 
>> I think this is nice syntax. Without tuples it'd not be as 
>> nice. It's not so much about efficiency. This pattern occurs all
>> over Python.

> I like that syntax too. If tuples and lists were unified, that bracket-free
> syntax could also be used for the list/tuple datatype, and nice syntax you
> mention would not change.

Except that you can suddenly do this:

results = foo()
...
results.append(5)
...
first, second, third, fourth = results # error

That's could be a bad thing, though of course Python isn't typesafe
in any case. The usage of tuples is however different, supported by 
syntactic difference. A tuple tends to contain a few heterogenous things,
while a list tends to contain a larger number of homogenous things.

[snip]
>> But this is less symmetrical, and since lists are mutable, 
>> this may go wrong far more easily.

> I am wondering where the serious concern about mutable lists comes from, as
> it is a pattern in the responses I've gotten on this topic. For me, the
> immutable list is a relatively new concept, since I can't think of many
> other languages that have it.

Functional programming languages have the concept. At least the
*spirit* that you shouldn't be changing a list. This is different from Python,
as we do change (homogenous) lists. The pattern in functional programming,
which *does* occur often in Python code as well, is some operation over
all elements of a list creating a new list (map(), though this is often
expressed in Python as a 'for' loop, partially due to the nastiness of
'lambda'). The old list isn't changed.

> I'm a programming-language-design kind of
> person, and I never really saw mutable lists as a problem that needed
> solving. I've used many Python-type languages that have mutable lists but
> not immutable ones, and if I don't want the elements to change they don't
> change -- I simply avoid the x[i] = y construct :-) And I still haven't
> heard from anyone about why there isn't similar concern about immutable
> dictionaries.

People tend to use classes for this purpose. You can give a class a public
interface (enforced or not) with only get_() methods, for instance. That
gets you just about the same thing as an immutable dictionary would give
you. Again, a dictionary tends to contain a larger number of homogenous
objects, while an object tends to contain a smaller number of heterogenous
objects.

> Anyway, it seems to me that most of the tuple advantages we enjoy could and
> should still exist with a merged tuple-list type, and the only possible
> drawback is the loss of an immutable sequence type. The gain, though, is one
> less decision that we must frequently make (between tuple and list) while
> programming.

Do you really feel that pain, then? Why don't you always use lists in your
Python code, if you want to avoid the decision? That'd be effectively the
same as what you're proposing? If you don't do this, is this because you're
afraid other programmers will complain about your code, or because you do
believe there's something to decide about?

If-everything-were-blocks-of-bytes-we-would-never-need-to-decide-ly yours,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From rkhale at earthlink.net  Sat Apr 29 11:04:35 2000
From: rkhale at earthlink.net (Richard Hale)
Date: Sat, 29 Apr 2000 15:04:35 GMT
Subject: The Simple Economics of Open Source
References: <1255228265-1994958@hypernet.com> <1255108931-9171946@hypernet.com>
Message-ID: <MPG.1374c41fa6c09b7198970f@news.earthlink.net>

In article <1255108931-9171946 at hypernet.com>, gmcm at hypernet.com says...
> If that were the model, we'd all be using Apples. MS got where 
> they are by outmanuvering IBM and DR. If Gary Kildall had 
> been willing to play hardball with a bunch of IBM suits, the 
> world would be quite different.
> 
> 
> 
You must mean "play ball".

Kildall "played hardball" much to his (past) regret.  He missed the 
appointment with IBM suits and left his wife to greet them.  He wanted 
lawyers and contracts and a too favorable deal.

Gates on the other hand took risks and "played ball" on IBM's terms.  No 
lawyers, no complicated contracts or deals.  He offered a system he 
didn't have (bought it afterward for $50,000 from ? Mountain View ? 
software).  Then, he out-maneuvered the IBM suits and ended up with all 
the marbles.

CP/M was the OS "monopoly" of its day.  A perfect illustration of how 
success breeds failure and how every trend has within it the seeds of its 
own destruction.

It must be Microsoft's turn now!?

Richard


From m.faassen at vet.uu.nl  Sun Apr  2 14:49:13 2000
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 2 Apr 2000 18:49:13 GMT
Subject: New Features in Python 1.6
References: <200004011740.MAA04675@eric.cnri.reston.va.us>
Message-ID: <8c84n9$pf0$3@newshost.accu.uu.nl>

Guido van Rossum <guido at python.org> wrote:
> New Features in Python 1.6
> ==========================

> With the recent release of Python 1.6 alpha 1, a lot of people have
> been wondering what's new.  This short note aims to explain the major
> changes in Python 1.6.

You almost got me! I didn't read the rest of the thread, and I was already
thinking about why all these would be bad ideas, and then I realized it
was posted on April 1. Good one!

Think-up-the-bad-stuff-so-as-to-avoid-it-ly yours,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From mhammond at skippinet.com.au  Mon Apr 17 20:20:19 2000
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Tue, 18 Apr 2000 00:20:19 GMT
Subject: Microsoft ActiveX Scripting
References: <8dg0rg$afj$1@cubacola.tninet.se>
Message-ID: <7TNK4.3520$1D.25703@news-server.bigpond.net.au>

"Erik Englund" <erikenglund at hotmail.com> wrote in message
news:8dg0rg$afj$1 at cubacola.tninet.se...
> "You create script components using any scripting language that supports
the
> Microsoft? ActiveX? Scripting interfaces. Script languages that support
> these interfaces include JScript, Microsoft? Visual Basic? Scripting
Edition
> (VBScript), PERLScript, PScript, and Python."
>
> If I want to use vbscript with the activeX control I just set the
property
> 'Languange' to 'vbscript', and it works. Does anyone know what I need to
do
> wo use it with Python? I have downloaded and installed python for
windows.

All you need to do then is say 'LANGUAGE="Python"' :-)

Check out the win32com reademe from the Python for Windows extensions.
This has a link to the Active Scripting pages - some basic demos and info
about using Active Scripting...

[Actually, Ive never personally used the scripting control, but I believe
it works!]

Mark.




From cut_me_out at hotmail.com  Thu Apr 13 10:30:10 2000
From: cut_me_out at hotmail.com (Alex)
Date: 13 Apr 2000 10:30:10 -0400
Subject: 1st non-trivial program - gentle criticism requested
References: <1e90r16.1sd2ts41e7ur6kN@gershwin.bio.ic.ac.uk>
Message-ID: <etdk8i2t6od.fsf@w20-575-109.mit.edu>

> With some trepidation, I post it here in the hopes that experts will
> be able to coach me in "The Python Way".

I'm by no means an expert.  Python is the first language I've used
extensively since I was a kid.  But I've been using it for close to a
year, and have acquired a couple of prejudices.

If I was going to write a program like this, practically everything
would have been inside a class.  This would have avoided the need for
global variables (looking at this script, I realized that I have never,
ever used a global variable in a python script.  I've never written a
really complex application, though.)  

Also, for extended calculations like yours, having everything in a class
makes it possible to stop the calculation, tinker with the innards of
the class, and restart it.  E.g.

class Weasle:
    '''etc. etc.'''
    pass

if ('weasle' not in dir ()):
    weasle = Weasle ()
else:
    weasle.__class__ = Weasle

weasle.run ()

...then you pump this into an interactive interpreter (I use emacs
python-mode for this), stop it whenever it seems to be going awry, edit
the script, and pump it back into the interpreter.  The weasle instance
gets all the new methods from your editing but keeps its old data, so
you don't have to start it up over and over and wait till it reaches the
point you want to study.  

If you want to speed up MutateIndividual, you might try something like
this untested code:

import random

def MutateIndividual (theTargetIndiv):
   """Iterates over symbols of individual changing their 'genes'"""
   # Must be a faster way of doing this
   theNewIndiv = []
   for theSymbol in theTargetIndiv:
      if (gRng.random() < gMutationRate):
         theNewIndiv.append (random.choice (gAlphabet))
      else:
         theNewIndiv.append (theSymbol)
   return string.join (theNewIndiv)

also, your findMax method seems redundant.  I think you could do
something like:

best_score = max (theFitness)
# Index of first element == best_score:
best_index = theFitness.index (best_score) 
theWinner = gPopulation[best_index]

Alex.


From effbot at telia.com  Tue Apr 25 03:56:01 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Tue, 25 Apr 2000 07:56:01 GMT
Subject: Simple list.append() question
References: <Pine.GSO.4.10.10004250813240.20737-100000@sundial> <0CF260C495FED111A6610000F866308D0D893596@mail3.ntu.edu.sg>
Message-ID: <lccN4.1530$wYl.189935616@newsb.telia.net>

#VEDALA NARASIMHA KUMAR# <PA0215481 at ntu.edu.sg> wrote:
> l=[]
> i=0
> while 1:
>     if i <3:
>         l.append([])
>     else:
>        break
>     i=i+1
>
>  crude..but works

tip: using for-in and range is faster, easier
to read, and definitely more "pythonish":

    l = []
    for i in range(3):
        l.append([])

</F>




From noreply at antispam.com  Mon Apr  3 02:16:21 2000
From: noreply at antispam.com (Brian Schmidt)
Date: Mon, 03 Apr 2000 06:16:21 GMT
Subject: embeddable Python web server
Message-ID: <VGWF4.1395$wo4.236156@paloalto-snr1.gtei.net>

Does anyone know of a webserver written in Python that is embeddable into a
Python application, something similar to tclhttpd available for Tcl?

Thanks!





From pinard at iro.umontreal.ca  Sun Apr  2 08:15:08 2000
From: pinard at iro.umontreal.ca (=?ISO-8859-1?Q?Fran=E7ois_Pinard?=)
Date: 02 Apr 2000 08:15:08 -0400
Subject: Why should I switch to Python?
In-Reply-To: "Tim Peters"'s message of "Sat, 1 Apr 2000 14:54:07 -0500"
References: <000201bf9c14$0a79d9e0$182d153f@tim>
Message-ID: <oqn1nc66nn.fsf@titan.progiciels-bpi.ca>

"Tim Peters" <tim_one at email.msn.com> writes:

> Python's [...] philosophy of raising exceptions is "gripe whenever there's
> a reasonable doubt" rather than "assume the programmer meant whatever
> they wrote".

That one is particularly valuable to me.  When I write C, it is very tedious
having to check all return codes and take a diagnostic action for most cases.
Of course, I could do like many and just check here and there, but then,
error handling has no spinal chord.  This aspect of C programming has been
bothering me for the last twelve years :-).  In Python, I only have to
handle errors for which I want something else than a diagnostic action,
which is much less burden while programming.

-- 
Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard





From tismer at tismer.com  Thu Apr 20 11:34:06 2000
From: tismer at tismer.com (Christian Tismer)
Date: Thu, 20 Apr 2000 17:34:06 +0200
Subject: __del__ problem - would adopting Garbage Collection fix this?
References: <000801bfaa9c$c9957aa0$d62d153f@tim> <vdFL4.9899$HG1.265876@nnrp1.uunet.ca>
Message-ID: <38FF236E.E34E64AB@tismer.com>


Warren Postma wrote:
> 
> > BTW, this was all discussed at extreme length on Python-Dev -- check it
> out!
> > I doubt anyone wants to repeat all of that whlie struggling to pretend
> it's
> > new & interesting <wink>.
> 
> Is there an archive on the web somewhere of Python-Dev discussions?! Sorry
> if that's a FAQ.

Not a FAQ, andthe archives are not advertized, but they're public and here:
http://www.python.org/pipermail/python-dev/

ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaunstr. 26                  :    *Starship* http://starship.python.net
14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     where do you want to jump today?   http://www.stackless.com



From tim_one at email.msn.com  Thu Apr  6 22:18:49 2000
From: tim_one at email.msn.com (Tim Peters)
Date: Thu, 6 Apr 2000 22:18:49 -0400
Subject: Some suggestion for a GUI kit evolved on Perl
In-Reply-To: <8cijoc$gfs$1@newsg3.svr.pol.co.uk>
Message-ID: <000101bfa037$9c1c2980$6c2d153f@tim>

[Makhno]
>> label = OrcLabel
>> NewOrcWindow(
>>     title = "Test"
>> ).Add(
>>     OrcVGroup(
>>         OrcButton,
>>         label
>>     )
>> ).Show()

> Yes, this appears to be the best way. I thought that Python
> statements could only be continued onto a new line if a '\' was
> supplied, but perhaps I was wrong.

An unclosed bracketing structure (left paren, square or curly bracket)
suffices.

> I don't like the way the Label has to be pre-made (not created in the
> constructor) in order to keep a reference.

It's thoroughly idiomatic Python, though, so even if you could dream up a
way around that it wouldn't be accepted -- Python programmers aren't
accustomed to staring at the insides of expressions for side-effects.  This
is considered to be a feature <wink>.

clash-of-philosophies-ly y'rs  - tim





From dbeagle at my-deja.com  Tue Apr 18 09:27:19 2000
From: dbeagle at my-deja.com (dbeagle at my-deja.com)
Date: Tue, 18 Apr 2000 13:27:19 GMT
Subject: telnetlib and NT vs. AIX
Message-ID: <8dhnr5$glm$1@nnrp1.deja.com>

I am trying to figure out what the difference is
between telnet to Windows NT and AIX.  As you can
see below the calls made to each telnet server
are the same but for some reason (unknown to me)
after I enter username in the NT telnet session
it never returns asking for the password.
Testing on the AIX system worked as expected. The
printout from both sessions is shown below.

Thank you for your help
db

**************************************************
************
This is the attempt to log into NT.
**************************************************
************

C:\_work\logparse\telnet>c:\_langs\python\python
Python 1.5.2 (#0, Apr 13 1999, 10:51:12) [MSC 32
bit (Intel)] on win32
Copyright 1991-1995 Stichting Mathematisch
Centrum, Amsterdam
>>> import telnetlib
>>> tn = telnetlib.Telnet("WinNT4SVR_1")
>>> tn.set_debuglevel(10)
>>> tn.interact()
Telnet(WinNT4SVR_1,23): recv '\377\375\030\377
\375 \377\375#\377\375$'
Telnet(WinNT4SVR_1,23): IAC DO 253
Telnet(WinNT4SVR_1,23): IAC DO 253
Telnet(WinNT4SVR_1,23): IAC DO 253
Telnet(WinNT4SVR_1,23): IAC DO 253
Telnet(WinNT4SVR_1,23): recv '\377\373\003\377\375
\001\377\375\037\377\373\005\377\375!'
Telnet(WinNT4SVR_1,23): IAC WILL 251
Telnet(WinNT4SVR_1,23): IAC DO 253
Telnet(WinNT4SVR_1,23): IAC DO 253
Telnet(WinNT4SVR_1,23): IAC WILL 251
Telnet(WinNT4SVR_1,23): IAC DO 253
Telnet(WinNT4SVR_1,23): recv '\377\373\001'
Telnet(WinNT4SVR_1,23): IAC WILL 251
Telnet(WinNT4SVR_1,23): recv '\033[?7h\015
\012Welcome to WinNT4SVR_1\015\012(C) Copyright
1994-1'
?[?7h
Welcome to WinNT4SVR_1
(C) Copyright 1994-1Telnet(WinNT4SVR_1,23): recv
'998 Pragma Systems, Inc.\015\012\015\012login
name:
 '
998 Pragma Systems, Inc.

login name: username
Telnet(WinNT4SVR_1,23): send 'username\012'
Telnet(WinNT4SVR_1,23): recv 'u'
cTelnet(WinNT4SVR_1,23): recv 'sername'
ondtt
Telnet(WinNT4SVR_1,23): send '\012'

**************************************************
************
This is a far as I get on NT.  I can never get to
the password.
Below is the attempt to log into an AIX Sever.
**************************************************
************

>>> tn = telnetlib.Telnet("AIXSvr_2")
>>> tn.set_debuglevel(10)
>>> tn.interact()
Telnet(AIXSvr_2,23): recv '\377\376%\377\375\030'
Telnet(AIXSvr_2,23): IAC DONT 254
Telnet(AIXSvr_2,23): IAC DO 253
Telnet(AIXSvr_2,23): recv '\377\376\030\377\375
\030\377\373\001\377\373\003\377\375\037\377\374
\310\37
7\375\001\015\012\015\012telnet (AIXSvr_2)\015\012
\015\000\015\012\015\000'
Telnet(AIXSvr_2,23): IAC DONT 254
Telnet(AIXSvr_2,23): IAC DO 253
Telnet(AIXSvr_2,23): IAC WILL 251
Telnet(AIXSvr_2,23): IAC WILL 251
Telnet(AIXSvr_2,23): IAC DO 253
Telnet(AIXSvr_2,23): IAC WONT 252
Telnet(AIXSvr_2,23): IAC DO 253


telnet (AIXSvr_2)

Telnet(AIXSvr_2,23): recv '\015\012\015\012\015
\012\015\012\015\012\015\012\015\012\015\012\015
\012\01
5\012\015\012\015\012\015\012\015\012\015\012\015
\012\015\012\015\012\015\012\015\012\015\012\015
\01
2\015\012\015\012\015\012'

Telnet(AIXSvr_2,23): recv '\015\000AIX Version 4
\015\012\015\000(C) Copyrights by IBM and by ot'
AIX Version 4
(C) Copyrights by IBM and by otTelnet
(AIXSvr_2,23): recv 'hers 1982, 1996.\015\012\015
\000login: \377\
376\037\377\376\001'
Telnet(AIXSvr_2,23): IAC DONT 254
Telnet(AIXSvr_2,23): IAC DONT 254
hers 1982, 1996.
login: username
Telnet(AIXSvr_2,23): send 'username\012'
Telnet(AIXSvr_2,23): recv 'username\015\012'
username
Telnet(AIXSvr_2,23): recv "username's Password: "
username's Password: myPassword
Telnet(AIXSvr_2,23): send 'myPassword\012'
Telnet(AIXSvr_2,23): recv '\015\012'

Telnet(AIXSvr_2,23): recv
'*************************************************
*'
**************************************************
Telnet(AIXSvr_2,23): recv
'*************************
****\015\012*                  '
*****************************
*
Telnet(AIXSvr_2,23): recv
'
 '
Telnet(AIXSvr_2,23): recv '         *\015\012*

// Lots of garbage but the connection is made!!

Telnet(AIXSvr_2,23): recv
'AIXSvr_2:/u001/username>'
AIXSvr_2:/u001/username>exit




Sent via Deja.com http://www.deja.com/
Before you buy.


From c.evans at clear.net.nz  Thu Apr 13 06:13:23 2000
From: c.evans at clear.net.nz (Carey Evans)
Date: Thu, 13 Apr 2000 10:13:23 GMT
Subject: Why should I switch to Python?
References: <seas5ia89177@corp.supernews.com>
	<slrn8f4d0j.1t0.bwinton@tor.dhs.org>
	<slrn8f4tfr.6hh.quinn@mono.ugcs.caltech.edu>
	<IvFI4.5759$HG1.168717@nnrp1.uunet.ca>
	<etdu2h8btwj.fsf@w20-575-109.mit.edu>
Message-ID: <87u2h6nwak.fsf@psyche.evansnet>

Alex <alex at somewhere.round.here> writes:

> Catchy title.

Good sigfodder too, if you don't mind.

> Mail-copies-to: [nobody|poster]

Uh, I think that's supposed to be _one_ of:

Mail-copies-to: nobody
Mail-copies-to: poster

-- 
	 Carey Evans  http://home.clear.net.nz/pages/c.evans/

"Perl Programmers are from Mars, Python Programmers are from Yorkshire?"
Catchy title.                                              - Alex in c.l.py


From gresham at mediavisual.com  Sat Apr  8 08:36:37 2000
From: gresham at mediavisual.com (Paul Gresham)
Date: Sat, 8 Apr 2000 20:36:37 +0800
Subject: Socket Programming Question
References: <aZdH4.95109$AT6.106753@dfw-read.news.verio.net>
Message-ID: <8cn94p$je01@imsp212.netvigator.com>

Have a look at www.freshmeat.net search for ACHAT. Someone has written it
already ... I believe this is using sockets.


"Ray Van Dolson" <rayvd at nospam.firetail.org> wrote in message
news:aZdH4.95109$AT6.106753 at dfw-read.news.verio.net...
> I'm trying to write a simple chat server w/ Python sockets that supports
> multiple connections and am having a little trouble... here's my server
code so
> far:
>
> #!/usr/bin/python
>
> from socket import *
> from threading import *
>
> serversocket = socket(AF_INET,SOCK_STREAM)
> serversocket.setsockopt(SOL_SOCKET,SO_REUSEADDR,1)
> serversocket.bind((gethostname(), 9001))
> serversocket.listen(5)
>
> def client_thread(clientsocket):
>         clientsocket.send("Welcome to the chat server...\r\n")
>         while 1:
>                 data = clientsocket.recv(1024)
>                 if data:
>                         clientsocket.send(data)
>                 else:
>                         pass
>
> def main_server_thread():
>         while 1:
>                 (clientsocket, address) = serversocket.accept()
>                 ct = Thread(target=client_thread,args=((clientsocket),))
>                 ct.run()
>
> mainThread = Thread(target=main_server_thread,args=())
> mainThread.start()
> while 1:
>         pass
>
> However, this code only allows one connection at a time... the other can
connect
> to the socket but is simply queued.  I would think the
main_server_thread()
> therad would create a new thread for each connection?
>
> What do I need to do to support multiple, simultaneous connections with
this
> code?
>
> Thanks,
> Ray Van Dolson




From garry at sage.att.com  Fri Apr 14 12:11:13 2000
From: garry at sage.att.com (Garry Hodgson)
Date: Fri, 14 Apr 2000 16:11:13 GMT
Subject: Java
References: <8d56qk$qdn1@imsp212.netvigator.com>
Message-ID: <38F74321.465CCA09@sage.att.com>

Gresh wrote:

>     but there are few experiences in life;
>     where you realize what you are missing;
>     i.e {
>         programming in python, then {
>             returning to a language with curly braces && semicolons;
>     }

last year i wrote a testing framework for a client-server TTS
engine in python.  i later decided to rewrite it in c++ for some
stupid reason or another.  drove me nuts.  i gave it a day or
so, then gave up.  i've done a *lot* of c++ in the past, but
python really spoiled me, and i couldn't (or wouldn't) deal
with all of the low level minutia.

-- 
Garry Hodgson                   Every night 
garry at sage.att.com                a child is born
Software Innovation Services        is a Holy Night.
AT&T Labs                         - Sophia Lyon Fahs


From nmm1 at cus.cam.ac.uk  Sun Apr  9 15:04:30 2000
From: nmm1 at cus.cam.ac.uk (Nick Maclaren)
Date: 9 Apr 2000 19:04:30 GMT
Subject: Python paradigms
References: <8cmr70$b0e$1@pegasus.csx.cam.ac.uk> <8cntqp$upi$1@slb0.atl.mindspring.net> <8covpn$4jq$1@slb6.atl.mindspring.net>
Message-ID: <8cqk7u$dmc$1@pegasus.csx.cam.ac.uk>

In article <8covpn$4jq$1 at slb6.atl.mindspring.net>,
Andrew Dalke <dalke at acm.org> wrote:
>
>I wrote:
>>Nick Maclaren wrote
>>>    x = (a != NULL ? a[i]->weeble : 0) + (b != NULL ? b[i]->wombat : 0)
>
>>How about
>>  x = getattr(a, "weeble", 0) + getattr(b, "wombat", 0)
>
>My code, BTW, is wrong.  I didn't see the "[i]" part of the original
>expression.  To my view of things, this means the expression is
>too complex.

Yes.  That is the point.  The pity is that the expression that I am
trying to evaluate is EXACTLY what the logic of the program requires,
but it cannot be written in Python (or Fortran) without artificial
variables.

>You asked about other idioms.  You could have a class wrapper
>for accessing elements, like:

Again, that works only in a few cases.  When the code is such that
there is not a common pattern to the test condition and evaluated
expression, you have to create a class for every instance.  That
doesn't exactly clarify your code ....


To the several people who suggested:

    (a() and [b()] or [c()])[0]

Thank you for the suggestion.  I have occasionally used perversities
like that (such as '*(x ? &a : &b) = y;' in C to mimic the Algol 68
facility), but I don't regard them as clarifying the code, either!


Regards,
Nick Maclaren,
University of Cambridge Computing Service,
New Museums Site, Pembroke Street, Cambridge CB2 3QG, England.
Email:  nmm1 at cam.ac.uk
Tel.:  +44 1223 334761    Fax:  +44 1223 334679


From jeremy at cnri.reston.va.us  Sun Apr  2 17:32:10 2000
From: jeremy at cnri.reston.va.us (jeremy at cnri.reston.va.us)
Date: Sun, 02 Apr 2000 21:32:10 GMT
Subject: Python 1.6 alpha 1 released
References: <200003312130.QAA04361@eric.cnri.reston.va.us> <38E5205F.DE811F61@roguewave.com>
Message-ID: <8c8e8q$27v$1@nnrp1.deja.com>

> Just read the announcement page, and found that socket.connect() no
> longer takes two arguments as was previously documented. If this
> change is staying I'm assuming the examples in the manual that uses
> a two argument socket.connect() will be changed?
Yes. The source and documentation releases are decoupled. The doc
release will take a bit longer, because Fred just finished a big
revision of the 1.5.2 docs.
> A quick look shows that this breaks all the network scripts I have
> installed (at least the ones that I found, undoubtedly there are many
> more). Because of this I will put any upgrade plans on hold.
I'm surprised at all the vituperative responses you got on this one.
I like the change, but agree that in some environments it makes a lot
of sense to delay upgrades. If you're project is mission critical
and its not broken, why upgrade? In the long run, you'll probably
benefit from all the little changes in 1.6 even if Unicode isn't
important, but do it on your own schedule.
Jeremy


Sent via Deja.com http://www.deja.com/
Before you buy.


From mwh21 at cam.ac.uk  Fri Apr 28 19:17:53 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 29 Apr 2000 00:17:53 +0100
Subject: Linux Processes from Python HOW-TO?
References: <8ecpch$j73$1@nnrp1.deja.com> <20000428234608.L11481@xs4all.nl>
Message-ID: <m38zxxolwe.fsf@atrus.jesus.cam.ac.uk>

Thomas Wouters <thomas at xs4all.net> writes:

> > I'm really hoping someopne has written this up!
> 
> Well, I just did, didn't I ? ;) 

While I think this was a great write-up, I'll just add that Python has
fairly recently gained, in os.py, a bunch of functions that act a bit
like a cross between the os.exec* functions and os.system; the spawn*
functions.  They take a file and a list of arguments (like the exec*
functions) but the subprocess des not replace the calling process.  It
can either be run as a subprocess (a bit like popen, but you do not
control it's stdin or stdout) or you can ask the spawn* function to
wait for the process to return (a bit like system, but there is no
chance of the shell doing strange things to the arguments).  On
unices, these functions will most likely be implemented using
fork&execv, but they save some tedious boilerplate.

Cheers,
M.

-- 
6. Symmetry is a complexity-reducing concept (co-routines include
   subroutines); seek it everywhere.
      -- Alan Perlis, http://www.cs.yale.edu/~perlis-alan/quotes.html


From frohne at gci.net  Tue Apr 18 17:40:05 2000
From: frohne at gci.net (Ivan Frohne)
Date: Tue, 18 Apr 2000 13:40:05 -0800
Subject: o'reilly python devcenter
References: <Pine.BSF.4.10.10004181241280.15729-100000@chi.pair.com> <38FC984D.59407BCB@callware.com>
Message-ID: <sfpliip4db865@corp.supernews.com>

"Ivan Van Laningham" <ivanlan at callware.com> wrote in message news:38FC984D.59407BCB at callware.com...

> >      http://www.oreillynet.com/python/

> Hmmm.  I looked it over, admittedly not in great depth, but:
> 
> 1)  There seems to be no mention of www.python.org
> 2)  There seems to be no mention of the python mailing lists
> 3)  The only books mentioned are O'Reilly books
> 4)  There seem to be no links to other websites

http://www.oreillynet.com/python/ is an outstanding web site.  
The design is professional, the home page loads quickly, and the 
material is well-organized. Almost every click either either leads to 
http://www.Python.org or displays the "Python-powered" ikon,
which also clicks to Mecca. 

Gee, I wonder if it's even legal for O'Reilly to flog other publishers' 
books?  And even if it is, ...?

So cut O'Reilly some slack.  Python can use the web presence.

--Ivan Frohne





From fw at deneb.cygnus.argh.org  Mon Apr 10 17:10:59 2000
From: fw at deneb.cygnus.argh.org (Florian Weimer)
Date: 10 Apr 2000 23:10:59 +0200
Subject: [ANNOUNCE] Garbage collection for Python
References: <20000407221750.A29279@acs.ucalgary.ca> <38F13FB9.D1D16AF@mindspring.com> <slrn8f2if1.dop.nascheme@cranky.arctrix.com>
Message-ID: <87r9cdhdb0.fsf@deneb.cygnus.argh.org>

nascheme at enme.ucalgary.ca (Neil Schemenauer) writes:

> I would rather get the bugs stomped out of my patch and get it
> included in 1.6 or 1.7.  We can then remove the FAQ entry.  :)

Is it possible to extend your GC in a way which would permit the
execution of several bytecode interpreters in multiple threads?


From Wolfgang.Strobl at gmd.de  Wed Apr  5 06:45:32 2000
From: Wolfgang.Strobl at gmd.de (Wolfgang Strobl)
Date: Wed, 05 Apr 2000 12:45:32 +0200
Subject: win32all-130 for Python 1.6 available
References: <_ziF4.1488$1b4.6926@news-server.bigpond.net.au> <d5rjeskov6c6drsok4n7vvcvvi74155lna@4ax.com> <fqvG4.5062$1b4.24253@news-server.bigpond.net.au>
Message-ID: <ph5mes8vkdpgk05lgneus29fi3apbvtofp@4ax.com>

"Mark Hammond" <mhammond at skippinet.com.au> wrote:

>Delete _all_ "Toolbar*" subkeys - probably 20 or so in total!

Worked, Thanks.

The editors new "folding" feature is great, btw.  Im currently using it
for browsing the Zope sources, and especially like the "Toggle
top-level" key. Its a great tool for viewing a programs overall
structure. 

May I take the opportunity to suggest another feature? The Zope sources,
like many others packages, have each and every file starting with a
lengthy, boring copyright message, license, or something similar, which
doesn't aid in understanding a programs structure in any way.  Wouldn't
it be nice if "Toggle top-level" and "Collapse all" would collapse
comment blocks, too?

-- 
      o      (    Wolfgang.Strobl at gmd.de          (+49 2241) 14-2394
     /\        *  GMD Forschungszentrum Informationstechnik GmbH
   _`\ `_<===     Schloss Birlinghoven,           | #include
__(_)/_(_)___.-._ D-53754 Sankt Augustin, Germany | <std.disclaimer>


From matheusrodrigues at hotmail.com  Wed Apr 19 16:36:24 2000
From: matheusrodrigues at hotmail.com (Matheus Rodrigues)
Date: Wed, 19 Apr 2000 20:36:24 GMT
Subject: ADO Question
Message-ID: <20000419203624.53393.qmail@hotmail.com>

Hi all,
I have a piece of code that runs fine in my machine (windows 98).
I am trying to create a package to install this application in other 
machines just by unziping a directory structure. But I am getting the 
following traceback message. I know it has something to do with the ADO 
configuration. Btw, the other machine has the same DNS entry that mine has. 
Any suggestions ??

Cheers,
Matheus


Traceback (innermost last):
  File "c:\python\win32com\client\dynamic.py", line 242, in _make_method_
    codeObject = compile(methodCode, "<COMObject %s>" % 
self._username_,"exec")
  File "<string>", line 1
     def Open(self, ConnectionString=L'', UserID=L'', Password=L'', 
Options=-1):

                                       ^
SyntaxError: invalid syntax
Traceback (innermost last):
  File "App/launch.py", line 18, in ?
    main.main()
  File "c:\python\App\main.py", line 33, in main
    adoConn.Open("data source=dns_sql_server;")
TypeError: call of non-function (type None)


______________________________________________________
Get Your Private, Free Email at http://www.hotmail.com




From bingdo19 at earthlink.net  Thu Apr 20 11:52:20 2000
From: bingdo19 at earthlink.net (Bob Langdon)
Date: Thu, 20 Apr 2000 15:52:20 GMT
Subject: Test_time Fails? / Installing on Virtual DOmain
References: <bingdo19-5535DF.10511520042000@news-server.twcny.rr.com> <bingdo19-8B25AD.11352820042000@news-server.twcny.rr.com>
Message-ID: <bingdo19-313AB5.12013820042000@news-server.twcny.rr.com>

In article <bingdo19-8B25AD.11352820042000 at news-server.twcny.rr.com>, 
Bob Langdon <bingdo19 at earthlink.net> wrote:

Anyone know why "altzone" won't work on Linux2?


> In article <bingdo19-5535DF.10511520042000 at news-server.twcny.rr.com>, 
> Bob Langdon <bingdo19 at earthlink.net> wrote:
> 
> > Hi,
> > 
> > Upon installing Python in the home directory of a virtual domain (after 
> > making the necessary configure file changes; i.e.: replacing all 
> > instances of "/usr/" with: "/home/username/python/usr/")
> > 
> > "make test" results in 1 error:
> > 
> > >>1 test failed: test_time
> > 19 tests skipped: test_al test_audioop test_bsddb test_cd test_cl 
> > test_crypt test_dbm test_dl test_gdbm test_gl test_gzip test_imageop 
> > test_imgfile test_nis test_rgbimg test_sunaudiodev test_thread 
> > test_timing test_zlib<<
> > 
> > Running Lib/test/test_time.py results in:
> > >>Lib/test/test_time.py: import: command not found
> > Lib/test/test_time.py: time.altzone: command not found
> > Lib/test/test_time.py: line 5: syntax error near unexpected token `t'
> > Lib/test/test_time.py: line 5: `t = time.time()'<<
> > 
> > Is this something I've failed to install, or something missing from the 
> > system?
> 
> 
> 
> Of course if I add the correct path test_time.pl
> 
> then I get the following error:
> 
> 
> >>Traceback (innermost last):
>   File "Lib/test/test_time.py", line 5, in ?
>     time.altzone
> AttributeError: altzone<<


From cdh at mail.ala.net  Wed Apr  5 22:53:03 2000
From: cdh at mail.ala.net (cdh)
Date: Wed, 5 Apr 2000 21:53:03 -0500
Subject: special attributes (and greetings from a new list user)  (repost, hopefully in plain text)
Message-ID: <002301bf9f73$3cff8f60$4f186bce@ala.net>

-- sorry for the repost, but outlook rendered my last one nearly unreadable:(

Hi,
My name is Chris Hickman and though I've been using python for a while now,
I'm new to the list. I've been a member on another prog-lang list before and
found it very fun and interesting.

Below is a module I'd like to submit for comments. It make's attribute handling
a bit easier as it takes care of __getattr__ and __setattr__ for you. The document
string should explain it fairly well.

Happy python programming,
Christoper D. Hickman

###############################################################################

# Author: Christopher D. Hickman
# Date: 4/3/2000

"""
  Module: spec_attr.py
    Simplifies adding special get and set handlers for attributes.

  Exports:
    class spec_attr:
      Defines the methods __getattr__ and __setattr__ implement magic handlers for special attributes.
      The special attribute must be defined with an _ (underscore) prepended.
      The methods are defined as getattr<special attr> and setattr<special attr>.
      The special attributes will then automatically call their handlers whenever accessed without
      the leading underscore.
      You may also implement "virtual attributes" which are handlers only with no special attribute
      supporting them.
      See example below for usage.

    class spec_attr_declarative:
      Same as the spec_attr class except it disallows implicit attribute creation.

    class SpecialAttributeError(AttributeError):
      Error class raised by spec_attr_declarative.__setattr__() when attempting to implicitly
      create a new attribute by assigning a value to a non-existant attribute.

  Example:
    >>> from spec_attr import spec_attr
    >>> class foo(spec_attr):
    ...  _x = 0 # special attribute x
    ...  _y = 0
    ...  _z = 0
    ...  count = 5
    ...  def getattr_x(self): # special get method for x
    ...    print "x attribute of %s was read" % repr(self) # verbalize all accesses to x
    ...    return self._x # caveat: you must use the underscore prepended form for all "gets"
    ...                   #         of a special attribute when defining it's get method.
    ...                   #         otherwise, infinite recursion will ensue.
    ...  def setattr_x(self, value):
    ...    print "x attribute of %s was changed from %s to %s" % (self, self._x, value)
    ...    self._x = value # same caveat applies for sets in set methods as for gets in get methods.
    ...  def setattr_y(self, value):
    ...    if value % 2:
    ...      raise "y must be even!"
    ...    self._y = value
    ...  def getattr_xyz(self): # a virtual get method (there's no xyz attribute)
    ...    return self.x, self.y, self.z
    ...  def setattr_xyz(self, value): # a virtual set method
    ...    self.x, self.y, self.z = value
    ...  def getattr_decr_count(self):
    ...    val = self.count
    ...    self.count = self.count - 1
    ...    return val
    ...
    >>> a = foo()
    >>> a.x
    x attribute of <__main__.foo instance at 10e2bf0> was read
    0
    >>> a.x = 12
    x attribute of <__main__.foo instance at 10e2bf0> was changed from 0 to 12
    >>> a.xyz = (1,2,3)
    x attribute of <__main__.foo instance at 10e2bf0> was changed from 12 to 1
    >>> a.xyz
    x attribute of <__main__.foo instance at 10e2bf0> was read
    (1, 2, 3)
    >>> while a.decr_count:
    ...  print a.count
    ...
    4
    3
    2
    1
    0
    >>> a.xyz = (6, 7, 8)
    x attribute of <__main__.foo instance at 10e2bf0> was changed from 1 to 6
    Traceback (innermost last):
      File "<interactive input>", line 1, in ?
      .
      .
      .
      File "<interactive input>", line 16, in setattr_y
    y must be even!
    >>>

"""

class SpecialAttributeError(AttributeError):
  pass

class spec_attr_declarative:

  # important things to remember about __getattr__:
  #   1. __getattr__ is *only* called when a "real" attr is not present
  #   2. hasattr() can cause recursion to __getattr__,  and traps any exceptions generated
  def __getattr__(self, attr):
    # see if there is a method defined to handle gets for attr
    if attr[:8] == 'getattr_':
      # we've either recursed because getattr_<attr> doesn't exist, (in which case the exception
      # generated here is trapped by the hasattr() call from the caller), or we tried to call
      # the non-existant getattr_<attr> directly.
      raise AttributeError, attr
    if hasattr(self, 'getattr_' + attr):
      return eval ('self.getattr_' + attr + '()')
    # if no method is present, try to return the value of the special attribute behind the one
    # we're trying to access
    if attr[:2] == '__':
      # we're probably checking for ('_' + _<attr>) due to recursive calls, a private attribute,
      # or an undefined instance of one of python's "magic" attributes (like __repr__),
      # so give up.
      raise AttributeError, attr
    if hasattr(self, '_' + attr):
      return eval ('self._' + attr)
    # otherwise, it simply doesn't exist
    raise AttributeError, attr

  # unlike __getattr__, __setattr__ is always called
  def __setattr__(self, attr, value):
    # if it's in the __dict__ it's not special, so process it quickly
    if self.__dict__.has_key(attr):
      self.__dict__[attr] = value
      return
    # see if there is a method defined to handle sets for attr
    if hasattr(self, 'setattr_' + attr):
      exec ('self.setattr_' + attr + '(value)')
      return
    # if no method is present, try to set the value of the special attribute behind the one
    # we're trying to set
    if hasattr(self, '_' + attr):
      exec ('self._' + attr + ' = value')
      return
    # otherwise attr could be present in the <instance>.__class__.__dict__ (or somewhere), in which
    # case hasattr() will return true even though the above tests failed. Create a real attribute
    # in the instance's __dict__. (this mirrors the default behavior in this case when no __setattr__
    # method is defined)
    if hasattr(self, attr):
      self.__dict__[attr] = value
      return
    # for the declarative version, don't allow de-facto creation of a new attribute. (explicitly use
    # __dict__ to circumvent this)
    raise SpecialAttributeError, "Attempted to create a new attribute %s for the declarative class instance %s" % (attr, repr(self))

class spec_attr(spec_attr_declarative):
  def __setattr__(self, attr, value):
    try:
      spec_attr_declarative.__setattr__(self, attr, value)
    except SpecialAttributeError:
      # allow creation of new normal attributes (without explicitly using __dict__) in the
      # non-declaritive version.
      self.__dict__[attr] = value






From dalke at acm.org  Sat Apr  1 01:48:15 2000
From: dalke at acm.org (Andrew Dalke)
Date: Fri, 31 Mar 2000 23:48:15 -0700
Subject: "tostring" as a string method?
Message-ID: <8c45ue$b7u$1@slb1.atl.mindspring.net>

Hello,

  Strings in Python 1.6 will have methods.  I would like
them to also have a "tostring()" method, which returns
itself.

  I've been working on classes for storing and analyzing
protein and dna sequences.  There are four ways I can
think of to store the data: a string, array.array,
Numeric.array, and a [] of single character strings.
Because of space requirements, [] is out, and Numeric.array
isn't available on all machines (and its length can't be
changed in-place).

  Strings are more natural, because of the large number
of string operations, including regular expressions.  An
array.array of characters is nice since it's mutable.
I would like to write functions which work on both strings
and array.arrays.

  Here's an example, which is used to translate DNA to
protein:

def translate(dna, table = Translation.standard_table):
  s = dna    # or "dna.tostring()" for array.array
  x = []
  for i in range(0, len(s), 3):
    x.append(table[s[i:i+3]])
  return string.join(x, "")

This would work with both strings and arrays, if strings
supported a "tostring()" method.  (I could give an example
with needing strings for regular expressions, but it would
be more complicated.)

Would there be any reason not to have tostring() as a method?
If not, I can supply a patch off the CVS version for it.

                    Andrew
                    dalke at acm.org





From tim.hochberg at ieee.org  Thu Apr  6 11:08:10 2000
From: tim.hochberg at ieee.org (Tim Hochberg)
Date: Thu, 06 Apr 2000 15:08:10 GMT
Subject: More about variables
References: <meh9-5C17A0.10445406042000@news.cit.cornell.edu>
Message-ID: <uL1H4.30328$U4.130371@news1.rdc1.az.home.com>


> Hi All,
>
> Let's say I had something like:
>
> number_of_variables=int(raw_input('Enter number of variables: '))
>
> if number_of_variables==1:
>    variable1=[]
> elif number_of_variables==2:
>    variable1=[]
>    variable2=[]
[snip....]


you could do:

for i in range(number_of_variables):
    exec "variable%s = []" % i

Or, if you wanted to about exec, and you didn't mind your variables being
stored in a dictionary or class, you could do:

varDict = {}
for i in range(number_of_variables):
    varDict["variable%s" % i] = []
varDict["variable0"].append(something)
#...

or,

class VariableHolder:
    pass
varClass = VariableHolder()
for i in range(number_of_variables):
    setattr(varClass, "variable%s" % i, [])
varClass.variable0.append(something)
#...


-tim




From c.hintze at gmx.net  Thu Apr 27 01:30:15 2000
From: c.hintze at gmx.net (Clemens Hintze)
Date: Thu, 27 Apr 2000 05:30:15 GMT
Subject: defining __getitem__
References: <WiMN4.1$7C.101@news.hananet.net>
Message-ID: <m3r9bsnmag.fsf@qiao.localnet>

"Gang" Gang Seong has written:
   Gang> I'd like to have a method __getitem__ in a class which acts
   Gang> like other sequence type.  For example,
   Gang> ------------------------
   Gang> class data:
   Gang> 	  def __init__(self, n):
   Gang> 	      self.maxsize = n
   Gang> 
   Gang> 	  def __getitem__(self, k):
   Gang> 	      if k >= self.maxsize:
                            raise IndexError  # !!!

The 'for' loop works that way! It calls __getitem__ until it catched
an IndexError, then it stops.

   Gang> 		    # I want to stop here
   Gang> 		    # ????
   Gang> 	      return get_data_from_some_where_else(k)
   Gang> 
   Gang> d = data(10)
   Gang> for el in d:
   Gang> 	  print el
   Gang> ------------------------
   Gang> I want to make it stop at some point automatically, like
   Gang> range() does. I don't want to use try: exception clause for
   Gang> this. Is there any way for this?

No try ... except clause necessary, but a 'raise' is! ;-)

   Gang> 
   Gang> Thanks
   
   Gang> Gang Seong

You're welcome,
Clemens.



From jwbnews at scandaroon.com  Sat Apr 22 16:38:35 2000
From: jwbnews at scandaroon.com (John W. Baxter)
Date: Sat, 22 Apr 2000 13:38:35 -0700
Subject: code not true?
References: <Pine.GSO.4.10.10004221525010.8110-100000@sundial>
Message-ID: <jwbnews-6EFA9D.13383522042000@news.olympus.net>

In article <Pine.GSO.4.10.10004221525010.8110-100000 at sundial>, Moshe 
Zadka <mzadka at geocities.com> wrote:

> On Sat, 22 Apr 2000 jeff_islay at my-deja.com wrote:
> 
> > Why does the second example return false?
> > 
> > >>> print 7.0 == 7
> > 1
> > >>> print (.07 * 100) == 7
> > 0
> 
> Because computers represent floats with too little bits.
> Never ever compare two floats for equality!
> 
Not quite.  No number of bits is sufficient to represent .07 exactly in 
the typical (and IEEE standard) representation.  

Various alternate representations are possible which are exact (such as 
BCD floating point of sufficient precision, rational numbers with 
(potentially) Python longs as numerator and/or denominator, some 
interesting recent work I haven't followed, etc etc etc).

  --John

-- 
John W. Baxter   Port Ludlow, WA USA  jwbnews at scandaroon.com


From scherbi at bam.com  Mon Apr 24 13:17:19 2000
From: scherbi at bam.com (Bill Scherer)
Date: Mon, 24 Apr 2000 13:17:19 -0400
Subject: ftplib sendcmd
References: <1255528387-51524611@hypernet.com>
Message-ID: <3904819F.E7FEEFE0@bam.com>

Thanks Gordon! This is the answer I needed.  It all works great now.

Thanks again,

Bill

Gordon McMillan wrote:

> Bill Scherer wrote
>
> > I'm having trouble with ftplib's sendcmd method (Python 1.5.2 on
> > RH linux 6.1):
> >
> > >>> ftp = ftplib.FTP(host, user, password)
> > >>> ftp.sendcmd('umask 002')
> > Traceback (innermost last):
> >   File "<stdin>", line 1, in ?
> >   File "/usr/local/lib/python1.5/ftplib.py", line 228, in sendcmd
> >     return self.getresp()
> >   File "/usr/local/lib/python1.5/ftplib.py", line 201, in getresp
> >     raise error_perm, resp
> > ftplib.error_perm: 500 'UMASK 002': command not understood.
> > >>>
> >
> > The umask command does, of course, work using a standard ftp
> > client to the same host.
> >
> > Other commands, chmod, etc, suffer equal fate.
> >
> > Is this not what sendcmd is for? What am I doing wrong?  Thanks!
>
> These are not rcf 959 commands. GNU ftpd will support them
> as SITE UMASK x and SITE CHMOD mod file.
>
> - Gordon




From chris_barker at my-deja.com  Fri Apr 28 15:39:00 2000
From: chris_barker at my-deja.com (chris_barker at my-deja.com)
Date: Fri, 28 Apr 2000 19:39:00 GMT
Subject: Linux Processes from Python HOW-TO?
Message-ID: <8ecpch$j73$1@nnrp1.deja.com>

Hi all,

Has anyone written a HOW-TO or tutorial that explains how to start and
manipulate other processes with Python. I imagine it's trivial if you
are familiar with the Linux/Unix process model, but I get very confused
about when to use:

system
popen and it's variants
exec and it's variants
fork
etc.

I'm really hoping someopne has written this up!

If you just have a subset of the info, send it to me, and maybe I'll
write it up!

-Chris Barker
cbarker at jps.net





Sent via Deja.com http://www.deja.com/
Before you buy.


From wfouche45 at hotmail.com  Thu Apr 20 13:49:56 2000
From: wfouche45 at hotmail.com (Werner Fouche)
Date: Thu, 20 Apr 2000 13:49:56 -0400
Subject: XML 1.0 DTD parser for SPARK.
Message-ID: <8e0c1u$hqc$1@slb1.atl.mindspring.net>

If anybody has a XML 1.0 DTD parser for SPARK
(http://csr.uvic.ca/~aycock/python) please drop me an e-mail.

Thanks,

Werner






From embed at geocities.com  Fri Apr  7 09:02:21 2000
From: embed at geocities.com (Warren Postma)
Date: Fri, 7 Apr 2000 09:02:21 -0400
Subject: Compile Gadfly?
Message-ID: <E%kH4.4341$HG1.126509@nnrp1.uunet.ca>

Has anyone tried compiling Gadfly and 'embedding' it into an executable, for
easier distribution? Does it work!?

Warren




From effbot at telia.com  Sun Apr  2 15:19:32 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Sun, 02 Apr 2000 19:19:32 GMT
Subject: Python 1.6 alpha 1 released
References: <Pine.LNX.4.10.10003311651590.22919-100000@starship.python.net> <00033115131700.03666@quadra.teleo.net> <38E7922D.52166F1D@bellatlantic.net>
Message-ID: <83NF4.1011$n68.190636032@newsb.telia.net>

Steve Holden wrote:
> Maybe someone could borrw the time machine and visit the future to let us
> know how this one is going to come out? :-)

guido didn't mention this in yesterday's 1.6 final release note,
so it's probably gone.

note, though, that socket.connect doesn't support unicode in
1.6 final, so you may be SOL anyway.

</F>




From DOUGS at oceanic.com  Thu Apr 20 00:17:24 2000
From: DOUGS at oceanic.com (Doug Stanfield)
Date: Wed, 19 Apr 2000 18:17:24 -1000
Subject: [ANNOUNCE] Webware for Python 0.1
Message-ID: <5650A1190E4FD111BC7E0000F8034D26A0F40D@huina.oceanic.com>

The following intrigues me but its not probably the best endorsement.  When
I hit your published link I get:

ERROR 


Traceback (innermost last):
  File "MainPage.py", line 55, in main
    _main()
  File "MainPage.py", line 49, in _main
    inc_counter()
  File "MainPage.py", line 23, in inc_counter
    f.write(repr(counter+1))
NameError: counter

But I guess you're at V0.1 so proving the error reporting works is a good
thing. :-)

This was from a redirect to http://writewithme.com/Webware/ by the way, so
maybe therin lies the problem.  I am still interested in looking at the
package.

-Doug-

> -----Original Message-----
> From: Chuck Esterbrook [mailto:echuck at mindspring.com]
> Sent: Wednesday, April 19, 2000 5:52 PM
> To: python-list at python.org
> Subject: [ANNOUNCE] Webware for Python 0.1
> 
> 
> 
> Webware for Python is a definitive suite of Python software 
> components for use in web development. Each component has a 
> focused purpose and can often be used on it's own. At the 
> same time, the components are designed to work together where 
> appropriate. As a user of Webware, you can bite off as much 
> or as little as you want to chew.
> 
> You can find more and download at:
> 
>       http://www.writers-webshop.com/Webware
> 
> 
> -Chuck
> -- 
> http://www.python.org/mailman/listinfo/python-list
> 



From Jonathan_Epstein at nih.gov  Mon Apr 10 10:25:17 2000
From: Jonathan_Epstein at nih.gov (Jonathan Epstein)
Date: Mon, 10 Apr 2000 10:25:17 -0400
Subject: Visual Studio add-ins in Python
Message-ID: <38F1E44D.13BE21D0@nih.gov>

Some colleagues and I are interested in writing some fairly complex
Visual Studio add-ins in Python.  Before digging into writing a complex
add-in, we will probably try a simple one, analogous to what appears at:
 
http://msdn.microsoft.com/library/devprods/vs6/visualc/vcsample/vcsmpdeveloperstudioaddinsamples.htm

Has somebody already written an add-in using Python?  I have the
_Programming Python_ book, but gather that it would be worthwhile to
also purchase _Python Programming on Win32_ for this project.  Correct?
  http://www.amazon.com/exec/obidos/ASIN/1565926218/104-2140334-1904426

Thanks for any advice,

- Jonathan


From genius at idirect.com  Fri Apr 28 13:38:58 2000
From: genius at idirect.com (Snoopy :-)))
Date: Fri, 28 Apr 2000 17:38:58 GMT
Subject: Help Needed-IDLE (Linux)Upgrading
Message-ID: <3909CC26.D684AD92@idirect.com>

No Flames Please. I am a NEWBIE :-))

Using Linux (RH-6.1) I have downloaded the Idle-0.5 Upgrade "tgz".
package.But I don't know where and how to install the package.

After unzipping the package I've read the README
file which doesn't seem togive clear instructions regarding installation
of the upgrade. On my system Idle is in /usr/bin 
but this is only the executable file without subdirectories where I
would be able to put the untared and uzippedfiles.

I would appreciate a helping hand
Best regards
Charles Takacs


From aahz at netcom.com  Thu Apr 20 16:25:42 2000
From: aahz at netcom.com (Aahz Maruch)
Date: 20 Apr 2000 20:25:42 GMT
Subject: Good style for multi-valued returns
References: <889A6EB6AE399393.619B0539BFCAAE0F.5C6D110C3DA908F7@lp.airnews.net>
Message-ID: <8dnp46$23t$1@slb0.atl.mindspring.net>

In article <889A6EB6AE399393.619B0539BFCAAE0F.5C6D110C3DA908F7 at lp.airnews.net>,
Cameron Laird <claird at starbase.neosoft.com> wrote:
>
>Examples of multi-valued returns: (head, tail); (quotient, remainder);
>(left, right); (prologue, body, epilogue); (data, code); (CAR, CDR);
>(count, size); (handle, status); (decomposition; remainder); ...
>
>Let me guess: folks generally pass tuples for small and
>determinate-in-number multi-values, and dictionaries otherwise, 'cept
>some do it all with tuples.  Accurate?

Somewhat.  There's also the option of returning a class (though that
usually ends up being a class instantiation followed by a method call).
--
                      --- Aahz (Copyright 2000 by aahz at netcom.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

Why is this newsgroup different from all other newsgroups?  --Aahz


From mdefreitas at sikorsky.com  Sun Apr 30 19:15:31 2000
From: mdefreitas at sikorsky.com (mdefreitas at sikorsky.com)
Date: Sun, 30 Apr 2000 23:15:31 GMT
Subject: MORE INFO
References: <8ea0ao$fro$1@nnrp1.deja.com> <8ea9uu$rh0$1@nnrp1.deja.com> <8ecad6$bun$1@ites.inria.fr> <8ecref$lck$1@nnrp1.deja.com> <8ehnvs$m0h$1@ites.inria.fr>
Message-ID: <8eieq8$ees$1@nnrp1.deja.com>

In article <8ehnvs$m0h$1 at ites.inria.fr>,
  Cedric Adjih <adjih at crepuscule.com> wrote:
> mdefreitas at sikorsky.com wrote:
>
> >> Do you want to run several python scripts (and several ui.commands
> >> C code) concurently (how would you synchronize?) ?
>
> > I do not want to run concurrently, I just want to run in one thread
of
> > execution, but recursively.
>
> Ok.
>
> >> Note that if all you want is to have C code that calls Python
> >> code that calls C code (through a module) that calls Python code
> >> without threads, PyRun_XXX functions are re-entrant.
>
> > Thanks! That's what I needed to know. I now modified (and
simplified)
> > my code as follows:
>
> > void interp(char *script) {
> >    static int nest_level = 0;
> >    FILE *fd = fopen(script, "r");
> >    if (nest_level == 0) {
> >       Py_Initialize();
> >       PyRun_SimpleString("import ui\n"); // import my ui
> >    }
> >    nest_level++;
> >    PyRun_SimpleFile(fd, script);
> >    nest_level--;
> >    if (nest_level == 0) Py_Finalize();
> > }
>
> > The only problem is that the nested script that is run "remembers"
the
> > variable settings from it's caller. Is there a way to give each
nested
> > python script a fresh, clean environment? The nested script should
not
> > effect the environment of it's caller either. Is that what the
global
> > and local dictionaries for the API function PyRun_File are for? If
so,
> > is there any examples on their usage. The C/API doc isn't all that
> > explicit.
>
>   Indeed.
>   PyRun_File is actually implementing most of the "execfile(...)"
> code (see http://www.python.org/doc/current/lib/built-in-funcs.html).
> Python source code for execfile is also an example of use for
> PyRun_File.
>   Passing empty dictionaries would be a way to have a "clean"
> environment except if a script imports a Python module and fiddles
> with its internals. Because modules will be shared among scripts.
> But fiddling with internals is very bad style.
>
> -- Cedric

Hmmm... I looked at the source code for execfile and it looks like it
either passes PyRun_File the dictionaries that were passed into
execfile itself, or if none was passed in, it passed PyRun_File the
current dictionaries as default. I believe that what I need to do is
pass in EMPTY dictionaries for the globals and locals. I tried the
following:

PyRun_File(fd, script, Py_file_input, PyDict_New(), PyDict_New());

But that didn't seem to work. The top level script seemed OK, but the
nested one didn't seem to execute. Is there something else that needs
to be done? Is it bad to pass in an empty global dictionary? Also, I
noticed that there was a PyDict_New function, but not a PyDict_Delete
function... I assume memory cleanup is automatic?

As always, thanks for the help...


Sent via Deja.com http://www.deja.com/
Before you buy.


From gmcm at hypernet.com  Tue Apr  4 11:53:50 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Tue, 4 Apr 2000 11:53:50 -0400
Subject: Python and .HTA files in IE5
In-Reply-To: <38E9EBA2.CF62323A@bellatlantic.net>
Message-ID: <1257258851-5014402@hypernet.com>

Richard M. Smith wrote:
[snip]
> I am having trouble executing the following simple Python script
> to read the contents of AUTOEXEC.BAT and output it on
> the .HTA page:
> 
> <script language=python>
> myfile = open("c:\\autoexec.bat")
> document.write(str(myfile.readlines()))
> myfile.close()
> </script>

readlines() returns a list of lines, which str() is not going to be 
happy about. Try using read().

- Gordon



From nascheme at enme.ucalgary.ca  Thu Apr 20 00:50:09 2000
From: nascheme at enme.ucalgary.ca (Neil Schemenauer)
Date: Thu, 20 Apr 2000 04:50:09 GMT
Subject: __del__ problem - would adopting Garbage Collection fix this?
References: <aw2L4.8907$HG1.245743@nnrp1.uunet.ca> <7ojL4.9424$HG1.254543@nnrp1.uunet.ca> <Z%oL4.9643$HG1.257533@nnrp1.uunet.ca> <m31z41iyhg.fsf@atrus.jesus.cam.ac.uk> <38FE7F1E.168E914A@cosc.canterbury.ac.nz>
Message-ID: <slrn8ft35a.uqs.nascheme@cranky.arctrix.com>

Greg Ewing <greg at cosc.canterbury.ac.nz> wrote:
>By the way, does anyone know what the recently
>announced GC patch does about __del__ methods?

Well, I guess I should. :)

When garbage is found by the collector (ie. reference cycles not
reachable from Python), all objects reachable from __del__
methods are moved into a separate set.  This is what I call
uncollectable garbage.  Uncollectable instances are currently
added to a global list of uncollectable objects reachable from
within Python.  If the programmer creates such garbage it is up
to them to deal with it.

The bottom line is don't add __del__ methods to objects that form
cycles.  It is, however, okay to reference objects with __del__
methods from objects involved in reference cycles.  If that's not
clear perhaps an example will help:

    >>> import gc
    >>> class A:
    ...     pass
    ... 
    >>> class B:
    ...     def __del__(self):
    ...         pass
    ... 
    >>> a = A()
    >>> a.b = B() # referenced from, not involved in cycle
    >>> c = A()
    >>> a.c = c
    >>> c.a = a
    >>> del a, c # collectable garbage created
    >>> gc.collect()
    gc: collectable <Object 0x81c6a24>
    gc: collectable <A instance at 81c6a04>
    gc: collectable <Object 0x81c699c>
    gc: collectable <A instance at 81c697c>
    4
    >>> a = A()
    >>> b = B()
    >>> a.b = b
    >>> b.a = a
    >>> del a, b # uncollectable garbage created
    >>> gc.collect()
    gc: uncollectable <Object 0x81c351c>
    gc: uncollectable <A instance at 81abbc4>
    gc: uncollectable <Object 0x81abbe4>
    gc: uncollectable <B instance at 81c50a4>
    4
    >>> gc.garbage
    [<__main__.A instance at 81abbc4>, <__main__.B instance at
    81c50a4>]


There was quite a bit of discussion about this issue on the
python-dev list.

Getting back on topic, GC does not help with this problem.  In
fact it is hard to make __del__ methods behave sanely with GC.
Nice __del__ semantics are one of the benefits of reference
counting.  __del__ methods are strong medicine, use them
sparingly.


    Neil

-- 
"A language that doesn't affect the way you think about programming, is
not worth knowing" -- Alan Perlis


From ivanlan at callware.com  Fri Apr  7 08:55:40 2000
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Fri, 07 Apr 2000 06:55:40 -0600
Subject: Teaching the art of programming, in python
References: <20000405053920.20579.qmail@web2102.mail.yahoo.com>    <8cfe42$o02$1@nnrp1.deja.com> <dpJG4.3744$HG1.106895@nnrp1.uunet.ca>    <38EB68A3.B1B5809F@callware.com>    <8cfqa7$4ml$3@newsserver.rrzn.uni-hannover.de>    <38EB73B6.6C0913A7@callware.com> <8chsij$k5r$1@newsserver.rrzn.uni-hannover.de> <slrn8er3fd.fcd.r2d2@shaka.acc.umu.se>
Message-ID: <38EDDACC.8641FF8C@callware.com>

Hi All

Niklas Frykholm wrote:
> 
[snip]
> 
> As an interesting side note, a publisher here in Sweden has
> recently started a 10-minute series, with titles such as
> "Teach your self Internet in 10 minutes". Seems there is
> inflation (or perhaps deflation) in these things.
> 

Probably SAMS.  They have a 10-minutes series.  The books are comprised
of *lots* of 10-minute lessons; there's no expectation that you would
learn the whole area of study in ten minutes;-)

> And it _is_ a silly practice. But then again, corporate
> silliness doesn't surprise me that much anymore.
> 

<g>

<ministry-of-silly-walks>-ly y'rs,
Ivan;-)
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
http://www.pauahtun.org and
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
Author:  Teach Yourself Python in 24 Hours



From nascheme at enme.ucalgary.ca  Sat Apr  8 00:17:50 2000
From: nascheme at enme.ucalgary.ca (Neil Schemenauer)
Date: Fri, 7 Apr 2000 22:17:50 -0600
Subject: [ANNOUNCE] Garbage collection for Python
Message-ID: <20000407221750.A29279@acs.ucalgary.ca>

Garbage collection for Python:

    http://www.enme.ucalgary.ca/~nascheme/python/gc.html

I think I have all of the wrinkles ironed out of this patch.
Reference cycles involving lists, tuples, instances, classes,
dictionaries, and functions are found.  Instances with __del__
methods are handled in a sane way.  It is easy to add GC support
to new types.  GC enabled Python is binary compatible with
regular Python.  

Generational collection works (currently three generations).  The
overhead measured by pybench is less than 1 percent and is not
measurable with pystone (on my machine).  Virually all extension
modules should work unchanged (I had to modify new and cPickle in
the standard distribution).  A new module called gc is available
for tuning the collector and setting debugging options.

The collector should be portable across all platforms but I have
only tested on my Linux machine.  The patched version of Python
passes all regression tests and runs Grail, Idle and Sketch
without problems.

The patch is against the current CVS version of Python.  If
enough people ask for a patch against the stable version of
Python I will make one.  Please send comments and bug reports.


    Neil

-- 
"Simplicity does not precede complexity, but follows it." -- Alan Perlis



From ppi at searchy.net  Fri Apr  7 10:34:21 2000
From: ppi at searchy.net (Penpal International)
Date: Fri, 07 Apr 2000 16:34:21 +0200
Subject: PPython
References: <38ede548.353776233@news.oh.verio.com>
Message-ID: <38EDF1ED.D102B7C9@searchy.net>

I bet also your web hoster has support for python. It comes along with
almost every ditribution of Linux. (95 % of all webservers are working
with linux).
Why you don't try it out. Make a little script, put it in the cgi-bin
directory (Don't forget to chmod it 0755). you also can login with
telnet and type: python  . If you get a sort of shell you know for sure
you can execute python scripts.


Kevin Dahlhausen wrote:
> 
> So many web hosting sites only allow Perl scripting, I got to thinking.  That's
> often a dangerous thing.   Would it be possible to implement the Python virtual
> machine in Perl?   OK, anything is possible,  but I guess the real question is
> does anyone have a feel for the scope of such a project?  Say using a native
> interpreter to pre-compile the Python code to pyc files and the perl cgi vm to
> run them?      Feasible, or just a silly idea?

-- 
Penpal International
http://ppi.searchy.net/
ppi at searchy.net


From neilh at hare.net.au  Sat Apr  1 09:07:38 2000
From: neilh at hare.net.au (Neil Hodgson)
Date: Sun, 2 Apr 2000 00:07:38 +1000
Subject: Why should I switch to Python?
References: <seas5ia89177@corp.supernews.com>
Message-ID: <011d01bf9be3$a3dca0d0$4ddd3fcb@neil>

> What I'd like to hear from people is what technical
> resons I should switch to Python.  Are there any
> features that Python has (perhaps in it's OO
> offering) that Perl lacks?  From what I can tell
> people tend to prefer Python over Perl mostly
> for its syntax, which I find very likable.

   If you are happy with Perl's syntax and semantics, I think you should
stay with it. There are cases where particular libraries are available in
one or the other language which may make it more usable on a particular
project but overall the languages cover similar ground. If you are curious,
it may be worthwhile trying a small task with Python. It shouldn't take very
long to learn enough Python to have a go.

   I moved from Perl to Python about 2 years ago because of the then better
Windows COM support and stayed because it suited me better. To explain why,
I'd have to repeat some of the common subjective arguments that don't seem
to satisfy Perlers.

   Neil





From chris.lada at westgroup.com  Tue Apr 11 11:01:57 2000
From: chris.lada at westgroup.com (Chris Lada)
Date: Tue, 11 Apr 2000 11:01:57 -0400
Subject: Processing BCD and Signed Integer Fields with Python
References: <38f334ea@wwwproxy3.westgroup.com> <8cvdel$lb9$1@pegasus.csx.cam.ac.uk>
Message-ID: <38f33dca@wwwproxy3.westgroup.com>

Nick,

 Being a newbie in python, I have to ask these questions: what functions can
I use to process these fields ?  What can I use to separate the number from
the sign ?   Is there a conversion for this type of data to strings ? What
data types would they be ?

Thanks,
Chris

Nick Maclaren wrote in message <8cvdel$lb9$1 at pegasus.csx.cam.ac.uk>...
>
>In article <38f334ea at wwwproxy3.westgroup.com>, "Chris Lada"
<chris.lada at westgroup.com> writes:
>|> We have been receiving record/field-based information from a data
provider.
>|> Some of the fields are BCD (Binary Coded Decimal), and some are signed
>|> integer. The data originates from a mainframe/Cobol(?) system. The
record
>|> layouts have definitions like "PIC S99999" (signed integer), and "PIC
S99999
>|> Comp-3" (BCD). Has anyone had to process this sort of data with Python ?
>|> Can Python handle this sort of data type ?  Anyone have suggestions on
how
>|> to approach this problem ?
>
>It's pretty easy handling anything like that, PROVIDED that you have
>the data intact.  Let us assume that is IBM 360/370/390 MVS or CMS,
>then it originated in a record-oriented system, where ALL characters
>are valid in records.  99% of people manage to mangle the data when
>transferring it to Unix or Unix-like systems (including Microsoft
>systems), and lose information.
>
>If you can avoid that, it is just like handling any field-oriented
>data, except in binary rather than human-readable text.  This remark
>applies to any sufficiently powerful language - C, Perl, Python and
>so on.
>
>
>Regards,
>Nick Maclaren,
>University of Cambridge Computing Service,
>New Museums Site, Pembroke Street, Cambridge CB2 3QG, England.
>Email:  nmm1 at cam.ac.uk
>Tel.:  +44 1223 334761    Fax:  +44 1223 334679




From dalke at acm.org  Mon Apr 24 04:33:38 2000
From: dalke at acm.org (Andrew Dalke)
Date: Mon, 24 Apr 2000 02:33:38 -0600
Subject: partial parsing?
References: <8dvilp$88s$1@slb6.atl.mindspring.net>
Message-ID: <8e10n7$rkl$1@slb3.atl.mindspring.net>

It looks like I wasn't clear in my post, so let me start over.
This time I'll explain the problem domain first, and end up with
the question part.

I need to parse data files of two broad clases, data files and
output files.  The first contains many sets of data records, which
were designed to be parsed by hand-coded parsers written by non-CS
people (no CF grammers here).  These files range from 10K up to
several hundred MB.  The second class, output from executables,
weren't really designed to be machine parsable, but since they are
machine generated, they can be parsed.

In both cases, the formats contain more information than is usually
needed.  I often just want the record identifier and one or two other
fields, and not the author's name, the publication data, the date
of submission, the ... I think you get it.  And in some cases, I just
want the start/end position of the record to make an index.

Additionally, all of the formats are line oriented.  Nearly all are
machine generated, so there's a good likelyhood that the data is in
the right format.  The lexical elements are quite stateful; the word
'sequence' may mean three or more things in the same record, depending
on the line and position.

On the simplification side, I believe I can get away with parsing
regular grammers only.  Some of the formats, about 25%, are non-regular.
Indded, expect three lines in a row to have the same length, which is
a context sensitive grammer, but the assertion that they are in the
right format means a regular grammer will work.  (Eg, a[n]b[n]c[n]
can be parsed with a*b*c* if it is guaranteed that the input grammer
is correct.)

Examples of the database files I need to parse are:
 220MB at ftp://ftp.expasy.ch/databases/swiss-prot/release/sprot38.dat
 21MB (compressed) at ftp://ncbi.nlm.nih.gov/genbank/gbmam.seq.Z
 55K at http://www.expasy.ch/cgi-bin/get-pdb-entry.pl?9pti
HTML marked-up records from the first two formats are at:
 http://www.expasy.ch/cgi-bin/get-sprot-entry?P80222

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=Nucleotide&lis
t_uids=3822546&dopt=GenBank
Links from the "DR" lines of the SwissProt record (the first one) gives
pointers to another dozen database formats.

Examples of executable output I need to parse are at:
 490K at ftp://bio.perl.org/pub/katel/Blast/Nucleic/wisteria.txt
 3K (a small one!) at ftp://bio.perl.org/pub/katel/Blast/Nucleic/lupine.txt

I usually write event driven parsers, in the style of SAX for XML
parsing.  This works well if I want to parse everything, but there's
a lot of overhead if I only need the couple of fields - I end up
ignoring most of them.  Indeed, most people usually write specialized
parsers to extract just the needed data.  Since the formats are line
oriented, that means something like:

  find the line starting with 'ID'; read the first word,
  find the lines starting with 'SQ'; get rest of the data in the line,
  record ends at the line equal to '//'.

In perl, this is ridiculously easy;
  $id = ''; $sq = '';
  while (<>) {
    if (/^ID (\w+)/) { $id = $1; next; }
    if (/^SQ (.*)/) { $sq .= $1; next; }
    if ($_ eq '//\n') {
       #&do_something($id, $sq)
       $id = ''; $sq = '';
    }
  }
That's one reason why perl is popular in this field.

So I want some way to start with the full grammer for the format,
then generate a parser which sends events for only those terms
I'm interested in, and allows optimizations which assume the format
is correct, so it can do thing like skip some bytes, or read a few
lines without checking what's on them.

Say, perhaps, like:

  import sys, ParserGen
  from Formats import GivenFormat
  parser = ParserGen.LenientParser(GivenFormat,
                           tags = ("identifier", "sq_data", "end") )
  class SQHandler:
    def __init__(self, func):
      self.reset()
      self.func = func
    def reset(self):
      self.id = self.sq = ''
    def parse(self, tag, text):
      if tag == "identifier":
        self.id = text
      elif tag == "sq_data":
        self.sq = self.sq + text
      elif tag == "end":
        self.func(self.id, self.sq)
        self.reset()
  def do_something(id, seq):
    pass
  parser.set_handler(SQHandler(do_something))
  parser.parse(sys.stdin)

I know, it looks more complicated than the perl version, but it
allows reads from any file-like object, and once created, the
SQHandler object can be reused, so the amortized cost is quite
small.  Plus, the same handler can work with related formats, perhaps
with the right adapter.

I've been looking at existing parser generation systems, such as
lex/yacc, SPARK and Plex.  They all seem to be designed to verify
every character in the file, so I lose some of the performance
given the guarantee of having a valid format.

They also prefer having a good context free grammer, as compared
to the stateful one I have.  While they do allow the lexer to have
different states, the specification for doing so isn't very natural
(IMHO).  Or to put it another way, the author of the lex/yacc
O'Rielly book says he did a FORTRAN parser purely in lex because
yacc didn't work very well.  I think that's relevant, since the
formats I deal with have the same sort of feel as I expect a
FORTRAN parser to work.

Indeed, what I'm envisioning is more like a super-scanner than a
parser, since the formats I need this system for are almost all
regular in nature, rather than context free.


So!  Since what I think I'm looking for doesn't seems to exist in
the toolsets I'm used to using, I'm going to have to write something
(most likely writing to mxTextTools or Plex for the back end).
Since I doubt I'm working on a unique situation, I figure there
must be existing code that does this which I haven't heard about,
or at least papers describing it.  I don't know exactly what I'm
looking for, so any pointer would be nice.  BTW, the implementation
for what I'm thinking about feels pretty straight-forward, but I
think it's best to compare what others have done.

"It", to reiterate, is a generator for event-based parsers which
are optimized to create events only for selected subsets of a given
(regular?) grammer and which can make the assumption that the input
file is in the correct format.

                    Andrew Dalke
                    dalke at acm.org





From robin at jessikat.demon.co.uk  Fri Apr 21 05:55:26 2000
From: robin at jessikat.demon.co.uk (Robin Becker)
Date: Fri, 21 Apr 2000 10:55:26 +0100
Subject: finding real $0
Message-ID: <OWeCyjAOWCA5Ew86@jessikat.demon.co.uk>

Is there any pythonic way to locate the original argument zero of the
process? I mean the location of python itself not the script. This would
allow me to launch subscripts without having to rely on the path being
set up correctly.
-- 
Robin Becker


From scarblac-spamtrap at pino.selwerd.nl  Sun Apr 30 18:39:20 2000
From: scarblac-spamtrap at pino.selwerd.nl (Remco Gerlich)
Date: 30 Apr 2000 22:39:20 GMT
Subject: Do I always have to write "self." ?
References: <270420001706386648%pecora@anvil.nrl.navy.mil> <slrn8ghc6t.f01.scarblac-spamtrap@flits104-37.flits.rug.nl> <8ecsgj$6lp$1@slb0.atl.mindspring.net> <m3bt2tonyg.fsf@atrus.jesus.cam.ac.uk> <300420001008012011%pecora@anvil.nrl.navy.mil> <slrn8gp6c4.ilf.kc5tja@garnet.armored.net> <bz1P4.75789$uF1.160076@afrodite.telenet-ops.be> <eqbpgsobm30cb50u4ctfclobr8qu6816j1@4ax.com>
Message-ID: <slrn8gpder.kp0.scarblac-spamtrap@flits104-37.flits.rug.nl>

Ben Wolfson wrote in comp.lang.python:
> On Sun, 30 Apr 2000 21:27:35 GMT, ben at co.and.co wrote:
> 
> >Samuel A. Falvo II <kc5tja at garnet.armored.net> wrote:
> >> I would write something like this:
> >> 
> >> 	uint32 WriteData( void *buffer, char *pString, uint32 cbString )
> >> 
> >> This tells me that WriteData() takes as parameters a buffer, a pointer to a
> >> string, and the number of bytes in the said string.  I fail to see how this
> >> is somehow hard to read.  :)
> >
> >OK, let's bite :-)
> >
> >    int WriteData(void *buffer, char *str, int len)
> >
> >I fail to see how this is somehow hard to read :-)
> 
> For me, actually, it's easier to read, since I don't have to wonder
> why the int parameter is "cbString".  That "cb" means "count of bytes"
> is not intuitively obvious.  "len" or "length" is much easier.

For me too. "cbString" sounds like some sort of string. Why is it pString,
but not pBuffer, btw?


-- 
Remco Gerlich,  scarblac at pino.selwerd.nl
 12:38am  up 55 days, 12:54,  7 users,  load average: 0.08, 0.02, 0.01


From python at rose164.wuh.wustl.edu  Sat Apr  1 13:40:16 2000
From: python at rose164.wuh.wustl.edu (David Fisher)
Date: Sat, 1 Apr 2000 12:40:16 -0600
Subject: Python Threading and sys.exit() codes
References: <0532446FBA0ED31193BD00A0C92C53AB0134460A@billtcat.serviceware.com>
Message-ID: <004701bf9c0c$f84b7aa0$0201a8c0@spkydomain>

----- Original Message -----
From: "Bob Dickenson" <bdickenson at serviceware.com>
To: "'David Fisher'" <python at rose164.wuh.wustl.edu>; "Bob Dickenson"
<bdickenson at serviceware.com>
Cc: <python-list at python.org>
Sent: Saturday, April 01, 2000 9:56 AM
Subject: RE: Python Threading and sys.exit() codes


> These are good suggestions.  A couple of questions/comments.

You got lucky, normally my suggestions are pretty lame. :)

>
> I thought about using the Queue model, but I was experimenting with
MxStack
> at the time.  It seemed to work fine, so...
>
> I think I need the global lock to get proper thread sync because I have to
> pop values off two stacks for each "job" - the file to be ftp'd and it's
> time stamp on the source machine (for resetting it locally after the
> ftp.retrbinary so that the next time the job runs it does valid date
> compares.)  If I'm missing something with this logic, I'd appreciate a
> clarification.  (I suppose I could compose a tuple and push/pop that, but
> otherwise....)
>

Yes, that was exactly what I was thinking of, pushing a tuple or list.  Hey,
if you've already done the work and it's doing the job, don't mess with it.
I just think Queues are cool.  I'll attach a program that was inspired by
your first post.

> I have read the thread doc and know that the sys.exit() in the FTPthread
> class def is bogus on my part--part of the "prettifying" the code needs
(mea
> culpa).
>
> It's the sys.exit() (explicit or implied) from the __main__ thread after
the
> iterative thread.join of the ftp worker threads  finishes that seems to be
> causing me problems. The process isn't "finishing" from the perspective of
> the job scheduler.  The single-threaded version surfaces both an implied
or
> explicit exit code to the OS, the multi-threaded version does not surface
> anything.  The SMTP message gets sent (last item before the sys.exit() in
> __main__) AND the command shell window closes, but something seems to be
at
> loose ends on the process termination cleanup side.

Well one thing you might try it to call setDaemon(1) for the worker threads
before start().  In threading.py sys.exitfunc() is overriden for the main
thread.  Sys.exitfunc() is called by the interpreter for cleanup after a
sys.exit().  It sounds like your program is hanging inside there.  Setting
all the threads to daemonic should bypass the logic in the new
sys.exitfunc().  I'm still a little confused though.  The only thing
sys.exitfunc() seems to do it join() to all the non-daemon threads, and
you've already done that, so it shouldn't hang.  But it's worth a try.  If
it were me, I would hack threading.py to not override sys.exitfunc() to see
if that fixed the problem.

>
> (BTW - the single threaded version of this program had about 60 files per
> minute throughput on the ftp; with 12 worker threads I'm getting between
> 800-1000 files per minute.  The flux rate on the target site is several
> thousand files per day, so it's really worth it to use the multi-threaded
> version)

Yeah, threads kick butt.
After your first post, I got to thinkin' about Queue and threads and wrote
this program to scan the internet randomly for NNTP servers that allow
public access.  So thanks, I got some inspiration from you.

Good luck,
David

from nntplib import NNTP
from Queue import Queue
import thread
import whrandom
##import daemonize

def checkserv(servq,repq):
    while 1:
        serv = servq.get()
        try:
            nn = NNTP(serv)
        except:
            repq.put((serv,''))
        else:
            repq.put((serv,'connect'))

def randurl(servq):
    rip = str(whrandom.randint(0,255))
    for i in range(3):
        rip = rip + '.' + str(whrandom.randint(0,255))
    try:
        rurl = socket.gethostbyaddr(rip)[0]
    except:
        servq.put(rip)
    else:
        servq.put(rurl)

def main():
    N_THREADS = 100
##    daemonize.become_daemon()
    f = open('openserv.txt','w')
    servq = Queue(0)
    repq = Queue(0)
    for i in range(N_THREADS * 2):
        thread.start_new_thread(randurl,(servq,))
    for i in range(N_THREADS):
        thread.start_new_thread(checkserv,(servq,repq))
    while 1:
        resp = repq.get()
        if resp[1]:
            f.write(`resp[0]`)
            f.flush()
        thread.start_new_thread(randurl,(servq,))





From karl at ulbrich.org  Wed Apr 26 13:57:24 2000
From: karl at ulbrich.org (Karl G. Ulbrich)
Date: Wed, 26 Apr 2000 13:57:24 -0400
Subject: PIL compile/make bombs out at step #7 -- help?
Message-ID: <20000426135724.A26663@exeter.exeter.org>

I'm stuck trying to compile PIL.  My environment is 
a standard Slackware 7.0 Linux instalation (Python
1.5.2, but the distribution version, not custom).  

The install bombs out in step #7 of the readme,
which is mentioned as a problem spot... however,
I modifed Setup.in to point to /usr/include,
and checked to see that yes, there's a python1.5
directory in there and yes, it has lots of .h files.

make -f Makefile.pre.in boot blows up as shown below
(sorry for the lengthy garbage, but I'm -very- new 
th this whole "compile" world and figured the extra
debug info couldn't hurt much).

Thanks for any suggestions,

Karl Ulbrich
karl at ulbrich.org
......................................
rm -f *.o *~
rm -f *.a tags TAGS config.c Makefile.pre python sedscript
rm -f *.so *.sl so_locations
VERSION=`python -c "import sys; print sys.version[:3]"`; \
installdir=`python -c "import sys; print sys.prefix"`; \
exec_installdir=`python -c "import sys; print sys.exec_prefix"`; \
make -f ./Makefile.pre.in VPATH=. srcdir=. \
	VERSION=$VERSION \
	installdir=$installdir \
	exec_installdir=$exec_installdir \
	Makefile
make[1]: Entering directory `/home/caribou/proj/pil/Imaging-1.0'
sed -n \
 -e '1s/.*/1i\\/p' \
 -e '2s%.*%# Generated automatically from Makefile.pre.in by sedscript.%p' \
 -e '/^VERSION=/s/^VERSION=[ 	]*\(.*\)/s%@VERSION[@]%\1%/p' \
 -e '/^CC=/s/^CC=[ 	]*\(.*\)/s%@CC[@]%\1%/p' \
 -e '/^CCC=/s/^CCC=[ 	]*\(.*\)/s%#@SET_CCC[@]%CCC=\1%/p' \
 -e '/^LINKCC=/s/^LINKCC=[ 	]*\(.*\)/s%@LINKCC[@]%\1%/p' \
 -e '/^OPT=/s/^OPT=[ 	]*\(.*\)/s%@OPT[@]%\1%/p' \
 -e '/^LDFLAGS=/s/^LDFLAGS=[ 	]*\(.*\)/s%@LDFLAGS[@]%\1%/p' \
 -e '/^DEFS=/s/^DEFS=[ 	]*\(.*\)/s%@DEFS[@]%\1%/p' \
 -e '/^LIBS=/s/^LIBS=[ 	]*\(.*\)/s%@LIBS[@]%\1%/p' \
 -e '/^LIBM=/s/^LIBM=[ 	]*\(.*\)/s%@LIBM[@]%\1%/p' \
 -e '/^LIBC=/s/^LIBC=[ 	]*\(.*\)/s%@LIBC[@]%\1%/p' \
 -e '/^RANLIB=/s/^RANLIB=[ 	]*\(.*\)/s%@RANLIB[@]%\1%/p' \
 -e '/^MACHDEP=/s/^MACHDEP=[ 	]*\(.*\)/s%@MACHDEP[@]%\1%/p' \
 -e '/^SO=/s/^SO=[ 	]*\(.*\)/s%@SO[@]%\1%/p' \
 -e '/^LDSHARED=/s/^LDSHARED=[ 	]*\(.*\)/s%@LDSHARED[@]%\1%/p' \
 -e '/^CCSHARED=/s/^CCSHARED=[ 	]*\(.*\)/s%@CCSHARED[@]%\1%/p' \
 -e '/^LINKFORSHARED=/s/^LINKFORSHARED=[ 	]*\(.*\)/s%@LINKFORSHARED[@]%\1%/p' \
 -e '/^prefix=/s/^prefix=\(.*\)/s%^prefix=.*%prefix=\1%/p' \
 -e '/^exec_prefix=/s/^exec_prefix=\(.*\)/s%^exec_prefix=.*%exec_prefix=\1%/p' \
 /usr/lib/python1.5/config/Makefile >sedscript
echo "/^#@SET_CCC@/d" >>sedscript
echo "/^installdir=/s%=.*%=	/usr%" >>sedscript
echo "/^exec_installdir=/s%=.*%=/usr%" >>sedscript
echo "/^srcdir=/s%=.*%=		.%" >>sedscript
echo "/^VPATH=/s%=.*%=		.%" >>sedscript
echo "/^LINKPATH=/s%=.*%=	%" >>sedscript
echo "/^BASELIB=/s%=.*%=	%" >>sedscript
echo "/^BASESETUP=/s%=.*%=	%" >>sedscript
sed -f sedscript ./Makefile.pre.in >Makefile.pre
/usr/lib/python1.5/config/makesetup \
	 -m Makefile.pre -c /usr/lib/python1.5/config/config.c.in Setup -n  /usr/lib/python1.5/config/Setup
bad word *noobjects* in _imaging  _imaging.c decode.c encode.c map.c display.c outline.c path.c 	-IlibImaging libImaging/libImaging.a 	-I/usr/include -L/usr/lib -ljpeg 	-I/usr/include -L/usr/lib -lz *noobjects*
make[1]: *** [Makefile] Error 1
make[1]: Leaving directory `/home/caribou/proj/pil/Imaging-1.0'
make: *** [boot] Error 2

----- End forwarded message -----



From effbot at telia.com  Tue Apr 11 12:23:38 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Tue, 11 Apr 2000 16:23:38 GMT
Subject: issubclass funny business
References: <m2snwsd5ah.fsf@netmeg.net>
Message-ID: <ekII4.1212$Za1.21055@newsc.telia.net>

Les Schaffer <godzilla at netmeg.net> wrote:
> but when i need to test whther an object is an instance of a subclass
> of StandardArrayFileScanner:
>
>         # make sure std is of the correct type
>         if not issubclass(std.__class__, StandardArrayFileScanner):

how about:

        if not isinstance(std, StandardArrayFileScanner):

>             import cie
>             print std.__class__, cie.CIE.__bases__
>     raise NotStandardArrayError
>
>
> but it doesnt work, even though std is of type cie.CIE :
>
> (gustav)~/Engineering/dspring/stoplite/matlab/Jue-Data/: python
FileScanner.py
>
> cie.CIE (<class cie.CIE_Standards at 813b320>,)  <=== result of print stmt
>
> huh??????

you have two different StandardArrayFileScanner
classes.

when you run FileScanner as a script, it's evaluated
in the __main__ namespace.  when you import it, it's
evaluated again, in the FileScanner namespace.

solution: never import the script you use as the main
program.

also see the last section on this page:

    http://www.pythonware.com/people/fredrik/fyi/fyi06.htm

</F>




From Henning.Schroeder at Informatik.Uni-Oldenburg.DE  Thu Apr  6 15:53:41 2000
From: Henning.Schroeder at Informatik.Uni-Oldenburg.DE (Henning Schroeder)
Date: Thu, 06 Apr 2000 21:53:41 +0200
Subject: alphanumerics in mails
In-Reply-To: <029501bf9eda$0dd8e2c0$858647c1@DEVELOPMENT>
References: <029501bf9eda$0dd8e2c0$858647c1@DEVELOPMENT>
Message-ID: <E12dILV-0003a7-00@riga.Informatik.Uni-Oldenburg.DE>

In comp.lang.python, you wrote:
>I've made a small mail-client which is working very good, except for when I
>get mails with alphanumerics (which happens quite often since I'm in
>scandinavia;).
>If I send a mail with subject:
>'Ett ?'
>The subject line in my mailclient will say:
>'=?iso-8859-1?Q?Ett_=E5?='
>Is there some kind of parser for this, has someone made anything like this
>or do I have to make one?
Use mime_decode_header from the mimify module

Greets from Germany
  Henning



From moshez at math.huji.ac.il  Tue Apr 25 01:13:43 2000
From: moshez at math.huji.ac.il (Moshe Zadka)
Date: Tue, 25 Apr 2000 08:13:43 +0300 (IDT)
Subject: Simple list.append() question
In-Reply-To: <IL3N4.4283$L3.47222@typhoon.midsouth.rr.com>
Message-ID: <Pine.GSO.4.10.10004250813240.20737-100000@sundial>

On Mon, 24 Apr 2000, Bill Wilkinson wrote:

> l = [[][:]] * 3

No, this won't work.




From tismer at tismer.com  Mon Apr  3 05:43:39 2000
From: tismer at tismer.com (Christian Tismer)
Date: Mon, 03 Apr 2000 11:43:39 +0200
Subject: Q: List and Dicts as default args
References: <326536345498D311B3BC00105A39802A074654@newsintern.dspace.de>
Message-ID: <38E867CB.D47BD1D2@tismer.com>


Stefan Migowsky wrote:
> 
> Hi,
> 
> I was just wondering how to handle lists and dictionaries in
> a simple way as default arguments to functions. Since following
> strange behaviour occured :

[Stefan, exploring the side effects of mutable default parameters]

> This behaviour only occurs with dictionaries and list. All other
> types are "well" behaved. I could try the following but that
> doesn't look so nice:
> 
> def f(Index,List = None, Dict = None):
>     if not List: List = []
>     List.append(1)
>     List.append(2)
>     print List
>     if not Dict: Dict = {}
>     Dict[Index] = Index
>     print Dict

The above is correct. When supplying default args, you need
to take care of possible side effects. Lists and dicts are
mutable objects, and the default assignment is done at
compile time.
You can use "real" values as defaults if:
- they are immutable, like ints, strings, tuples
- they are only used for lookup (not changed)
- you *want* to change state permanently

In all other cases, use something like you did above.

ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaunstr. 26                  :    *Starship* http://starship.python.net
14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     we're tired of banana software - shipped green, ripens at home



From Peter.Desaulniers at pahv.xerox.com  Mon Apr 10 14:30:23 2000
From: Peter.Desaulniers at pahv.xerox.com (Desaulniers, Peter)
Date: Mon, 10 Apr 2000 11:30:23 -0700
Subject: Actually Running a function from a DLL
Message-ID: <51B8ABCE456FD111899900805F6FD6EE071E9108@mercury.ADOC.xerox.com>

Writing C extensions for the first time is a daunting task.

The best source of information that I am aware of are:

1. The API is described, minimally, on the Python web site

2. Some of the python source header files have lengthy comments that explain
some important concepts.  

3. Internet Programming with Python by Watters, van Rossum and Ahlstrom has
a decent chapter on C extensions.  But you will still need to read the
headers indicated in #2.

Getting the call mechanism will be fairly straightforward with the above
information.  The challenge will be understanding object management.  The
proper use of INCREF and DECREF is a difficult concept to get right.
Improper INCREF may leak (maybe you don't care) and improper DECREF may
crash you due to a prematurely cleaned up reference (you will care about
that one).

In terms of how rapidly you can learn to do C extensions (that run correctly
on all systems... beware, want may work on Windows NT, no longer works on
Win2000 in some complex models), you may need to save your appetite unless
you have some really savvy developers. 

Good luck.

> -----Original Message-----
> From: Peter Olsen [mailto:pcolsen at draper.com]
> Sent: Friday, April 07, 2000 6:13 PM
> To: python-list at cwi.nl
> Subject: Actually Running a function from a DLL
> 
> 
> I'm writing ask for a quick explanation of how to load and 
> use dll files 
> from Python running under MS Windows.
> 
> I have installed PythonWin and win32all, but I seem to be completely 
> unable to run a function from a dll.  I'm loading the dll with 
> win32api.LoadLibrary, but I can't figure out how to access the 
> functions.  I've made some statements about how fast we can 
> learn Python 
> and the start development, but if I don't find an answer 
> quickly,  I may 
> have to eat my words.  If the best answer is "RTFM," please give me a 
> hint as to where in TFM I should R.
> 
> I would greatly appreciate receiving answers by email because 
> my company 
> makes it hard to read news.
> 
> Thanks!
> 
> Peter
> 
> 
> -- 
> http://www.python.org/mailman/listinfo/python-list
> 



From jnw22 at cam.ac.uk  Tue Apr 11 09:21:33 2000
From: jnw22 at cam.ac.uk (James "Wez" Weatherall)
Date: Tue, 11 Apr 2000 14:21:33 +0100
Subject: Why should I switch to Python?
References: <seas5ia89177@corp.supernews.com> <slrn8f4d0j.1t0.bwinton@tor.dhs.org> <slrn8f4tfr.6hh.quinn@mono.ugcs.caltech.edu> <IvFI4.5759$HG1.168717@nnrp1.uunet.ca>
Message-ID: <8cv8st$hjc$1@pegasus.csx.cam.ac.uk>

> Yorkshire.

Which is a great place to be from.  In particular, Knaresborough. :)




From arta at x-stream.nl  Thu Apr 13 08:41:27 2000
From: arta at x-stream.nl (A[r]TA)
Date: Thu, 13 Apr 2000 14:41:27 +0200
Subject: windows registry
Message-ID: <era0iGYp$GA.273@net003s>

Heyz..

How do I edit the windows registry with Python?

Thnx, A[r]TA


--
We shall not cease from exploration, and the end of all our exploring will
be to arrive where we started and know the place for the first time.
- T. S. Eliot





From Andreas.Cardeneo at mach.uni-karlsruhe.de  Fri Apr  7 08:54:19 2000
From: Andreas.Cardeneo at mach.uni-karlsruhe.de (Andreas Cardeneo)
Date: Fri, 07 Apr 2000 14:54:19 +0200
Subject: CGI Help Needed
References: <8ck9uo$ga08@imsp212.netvigator.com>
Message-ID: <38EDDA7B.47FC3FDE@mach.uni-karlsruhe.de>

> 
> What is the normal and respected method of User authentication on a web
> site, how should I present a login box when the user is not authenticated
> already, and how do I maintain a session with an authenticated user, without
> using cookies, allowing them to continue surfing around the web site ?

Hi,

well. I think you're right that this is not exactly Python specific, but
you might
use Python in a CGI-script to look up session-keys. If you have a look
at PHP, then
you will see that there is a function to generate a session key (some
large number
that can not be guessed easily and has some checksum in it) that you
pass to your pages
as a URL paramter in the form "?session=4242424242424242&nextpar=....".
You then evaluate
the key in your script looking it up in a database or file and checking
if it is still
valid.
I would certainly not use JavaScipt to authenticate the user because the
implementations
have some security problems. If I remember correctly Microsoft had
problems on their
Hotmail-system with Javascript. One possibility would be to use the
.htaccess file of
a directory and let the browser/server check authorization. But that's
probably a bad idea
for larger sites. You could better use a form within a frame on your
page so the URL is
not being displayed in the browser location line (but in the HTML code
it is).

Hope this helps,

Andreas


From snoopyy at ms11.hinet.net  Tue Apr 25 11:10:25 2000
From: snoopyy at ms11.hinet.net (¤ý©y»Ê)
Date: Tue, 25 Apr 2000 23:10:25 +0800
Subject: ·Q·Q ·í¦ÑÁó ¥s§A±N ¸ê®Æ®wªº¸ê®Æ ¿é¤J¨ì ¬Y­Óºô­¶¥h ¡A¤Ñ¨º ?? À°À°§Ú§a!!
Message-ID: <8e4cc2$d5t@netnews.hinet.net>

?Q?Q ?????? ?s?A?N ?????w?????? ???J?? ?Y???????h ?A???? ?? ???????a!!


--
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

    ??       ?] ?? ?^ ?]???^          ???^     ??    ?]??    ??????  ?]??
????
???U??    ???U??    ?]?_?_   ?]?U     ?]?U?^    ?U?^    ?U      ?]??
 ??
  ???@       ??  ??      ?]             ?? ?^    ?] ?^   ?] ?@   ?? ??
???`?`  ????
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
?{?????n?g???h..???????n?????[..?_???????a!!                        ?i?????j
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~





From phd at phd.russ.ru  Thu Apr 27 06:03:54 2000
From: phd at phd.russ.ru (Oleg Broytmann)
Date: Thu, 27 Apr 2000 10:03:54 +0000 (GMT)
Subject: How can I get the name of an object???
In-Reply-To: <8e9090$c4d$1@nnrp1.deja.com>
Message-ID: <Pine.LNX.4.21.0004271002510.9760-100000@fep132.fep.ru>

On Thu, 27 Apr 2000 spex66 at my-deja.com wrote:
> > An object could be referred to by any number of names:
> >
> > dd = zz ()
> > ee = dd
> >
> > Which name for the object do you wish to know?  ee or dd?
> dd or in other words, the name of the FIRST made instance... I tried to

dd = zz ()
ee = dd
del dd

   What is the name now? (ee still holds the reference to the object)

Oleg.            (All opinions are mine and not of my employer)
---- 
    Oleg Broytmann      Foundation for Effective Policies      phd at phd.russ.ru
           Programmers don't die, they just GOSUB without RETURN.




From m.faassen at vet.uu.nl  Wed Apr 26 10:16:09 2000
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 26 Apr 2000 14:16:09 GMT
Subject: how would i make a HTML type UI?
References: <000f01bfaf6b$0c9c6200$2801000a@iel.ie>
Message-ID: <8e6tn9$gvq$1@newshost.accu.uu.nl>

Shaun <shogan at iel.ie> wrote:

> how would i create a HTML type user interface using python which would take
> in say, -name-address-date of birth- and record it with the intention
> of struct.pack-ing the values entered,

I'm not quite sure what you mean by struct.pack-ing the values entered, but
for information on web programming with Python look here:

http://www.python.org/topics/web/

For a very powerful system based on Python for building web applications,
look at Zope:

http://www.zope.org

Good luck!

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From embed at geocities.com  Fri Apr  7 09:07:16 2000
From: embed at geocities.com (Warren Postma)
Date: Fri, 7 Apr 2000 09:07:16 -0400
Subject: "tostring" as a string method?
References: <8c45ue$b7u$1@slb1.atl.mindspring.net> <Lp9H4.3526$y4.52453@newsread1.prod.itd.earthlink.net> <8ck7kq$1cu$1@slb3.atl.mindspring.net>
Message-ID: <f4lH4.4346$HG1.126863@nnrp1.uunet.ca>

>
> Meanwhile, on the biopthon.org list, Jeff Chang pointed out that
> __str__ is in general an ill-defined quantity.  The usual
> interpretation is that __str__ returns a human-representation while
> __repr__ returns a computer representation.  So using __str__
> to return the character stream is not the right interface.  We're
> now using "tostring()", which is more obvious.
>

Hmm. Have you considered marshalling everything instead?

Warren




From kragen at dnaco.net  Sat Apr 22 19:49:03 2000
From: kragen at dnaco.net (Kragen Sitaker)
Date: Sat, 22 Apr 2000 23:49:03 GMT
Subject: code not true?
References: <Pine.GSO.4.10.10004221525010.8110-100000@sundial> <jwbnews-6EFA9D.13383522042000@news.olympus.net>
Message-ID: <PTqM4.490$RM6.95483@news-east.usenetserver.com>

In article <jwbnews-6EFA9D.13383522042000 at news.olympus.net>,
John W. Baxter <jwbnews at scandaroon.com> wrote:
>Various alternate representations are possible which are exact (such as 
>BCD floating point of sufficient precision, rational numbers with 
>(potentially) Python longs as numerator and/or denominator, some 
>interesting recent work I haven't followed, etc etc etc).

Well, back in the 1970s, Gosper proposed to represent all computable
numbers exactly as continued fractions --- see HAKMEM for the details.
I haven't implemented it, but it looks like it should only be one or
two orders of magnitude slower than using standard integers or
floating-point (for normal uses, of course --- printing out pi to the
thousandth place will be slower in the continued-fraction form than in
the floating-point form, but that's because the continued-fraction
output is correct.)

There has been other, more recent work I haven't followed either.
-- 
<kragen at pobox.com>       Kragen Sitaker     <http://www.pobox.com/~kragen/>
The Internet stock bubble didn't burst on 1999-11-08.  Hurrah!
<URL:http://www.pobox.com/~kragen/bubble.html>
The power didn't go out on 2000-01-01 either.  :)


From kajiyama at grad.sccs.chukyo-u.ac.jp  Thu Apr  6 10:44:22 2000
From: kajiyama at grad.sccs.chukyo-u.ac.jp (Tamito Kajiyama)
Date: 06 Apr 2000 14:44:22 GMT
Subject: Codecs for Japanese character encodings
Message-ID: <KAJIYAMA.00Apr6234422@pseudo.grad.sccs.chukyo-u.ac.jp>

Based on the Python Unicode support proposal, I wrote codecs for
two Japanese character encodings EUC-JP and Shift_JIS.  The codecs
are available at the following location:

http://pseudo.grad.sccs.chukyo-u.ac.jp/~kajiyama/tmp/japanese-codecs.tar.gz

Comments and suggestions are very appreciated.

Enjoy,

--
KAJIYAMA, Tamito <kajiyama at grad.sccs.chukyo-u.ac.jp>
--
KAJIYAMA, Tamito <kajiyama at grad.sccs.chukyo-u.ac.jp>


From nmm1 at cus.cam.ac.uk  Sat Apr 15 05:19:02 2000
From: nmm1 at cus.cam.ac.uk (Nick Maclaren)
Date: 15 Apr 2000 09:19:02 GMT
Subject: else clauses in while and for loops
References: <Zv_C4.6415$4N.169440@news.tpnet.pl> <v0RD4.13$n68.186989568@newsb.telia.net> <8c7b29$37v$1@inputplus.demon.co.uk> <38F79B93.B03F0A6B@mindspring.com>
Message-ID: <8d9c66$el7$1@pegasus.csx.cam.ac.uk>

In article <38F79B93.B03F0A6B at mindspring.com>,
Chuck Esterbrook  <echuck at mindspring.com> wrote:
>So did anyone present, in this thread, an example that shows a real need?

Whether or not they have one, there is a fairly common case:

    Perform an action XXX provided that no member of sequence YYY
fulfils a particular criterion.

Yes, you can also use exceptions for this, but the 'else' construct
is clearer and probably more efficient.

Actually, I would prefer a different way of doing this, but that is
another matter.  Python is Python and not some other language.


Regards,
Nick Maclaren,
University of Cambridge Computing Service,
New Museums Site, Pembroke Street, Cambridge CB2 3QG, England.
Email:  nmm1 at cam.ac.uk
Tel.:  +44 1223 334761    Fax:  +44 1223 334679


From Michael.Scharf at gmx.de  Wed Apr 19 16:40:45 2000
From: Michael.Scharf at gmx.de (Michael Scharf)
Date: Wed, 19 Apr 2000 22:40:45 +0200
Subject: SNiFF+ for Linux
References: <UGnL4.1$kM.566@news>
Message-ID: <38FE19CD.385CCD64@gmx.de>

Robert Hicks wrote:
> 
> There was a "how-to" to integrate Python into SNiFF+. I have 2 questions
> regarding this:

> 1. Does anyone have this info?

The how-to is at 
  http://starship.python.net/crew/scharf/sn4py

Unfortunately, starship is down at the moment. 
I did the integration for an older version of SNiFF and
Python. I don't think that the debugger integration runs
out of the box.

> 2. Does it work with the free Linux version that only supports C/C++?

I simply don't know what is in the free linux version. But I think
it is the full distribution with a restricted license. Fortunately,
the pythonparser has no license (since it is not officially supported).

So, add the Python file-type to new projects and it should work fine.
I think the restriction is a few hundred files. That will hit you, if
you want to browse the entire python-lib.

But yes, it works fine for browsing and editing. Cross-ref doesn't work,
but that is hard (probably impossible) for a dynamically typed language.

If you start sniff as
  % sniff --GUI.Tools.ED.SmartComments 

even "comment region" will work for python. (It's enough to
call sniff once with this option, open the preferences and
hit OK: sniff will then remember the option in the user
preferences.)

Michael
-- 
     ''''\     Michael Scharf
    ` c-@@     TakeFive Software GmbH, a Wind River Company
    `    >     http://www.WindRiver.com
     \_ V      mailto:Michael.Scharf at gmx.de


From mwh21 at cam.ac.uk  Sat Apr  1 07:43:42 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 01 Apr 2000 13:43:42 +0100
Subject: Paid to work in Python? (was: Re: Choice Of Language)
References: <Uqvw4.2261$yL2.28868@typhoon.southeast.rr.com> <8ajmib$elo$1@nnrp1.deja.com> <8c3ats$b5f$1@nntp.Stanford.EDU> <slrn8eact2.sq7.nobody@bambi.visi.com>
Message-ID: <m37leioutd.fsf@atrus.jesus.cam.ac.uk>

nobody at nowhere.nohow (Grant Edwards) writes:

> In article <8c3ats$b5f$1 at nntp.Stanford.EDU>, Amit Patel wrote:
> 
> >I do 100% of my professional development in Python, too.  Google's
> >using a lot of Python and C++, and fortunately I get to do my work in
> >Python.  :-)
> >
> >I should post Google's Python-related job openings to
> >          http://www.python.org/Jobs.html
> 
> Pointless tanget: Is it just me, or is Google really the last usable search
> engine?
> 

I'd have said that it was the first, last and only really usable
search engine.

I used to be amazed at its accuracy, but now I just get narked by the
inaccuracy of every other engine I use (not very many...).

Not affiliated with google, just a very satisfied user (though if they
*want* to give me a job...).

Cheers,
M.

-- 
well, take it from an old hand: the only reason it would be easier
to program in C is that you can't easily express complex  problems
in C, so you don't.                 -- Erik Naggum, comp.lang.lisp


From aa8vb at yahoo.com  Tue Apr 18 08:25:04 2000
From: aa8vb at yahoo.com (Randall Hopper)
Date: Tue, 18 Apr 2000 08:25:04 -0400
Subject: IDLE 0.5:  Odd NumTut import results
Message-ID: <20000418082504.A417909@vislab.epa.gov>

>From the command-line:

   > python
   Python 1.6a2 (#2, Apr 17 2000, 12:17:27) [C] on irix646-n32
   Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
   >>> from Numeric import *
   >>> from NumTut import *
   >>>

No problem.

>From within IDLE (started from the same shell):

   Python 1.6a2 (#2, Apr 17 2000, 12:17:27) [C] on irix646-n32
   Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
   IDLE 0.5 -- press F1 for help
   >>> from Numeric import *
   >>> from NumTut import *
   Traceback (innermost last):
     File "<pyshell#1>", line 1, in ?
       from NumTut import *
     File "NumTut/__init__.py", line 11, in ?
       from view import *
     File "/home/rhh/software/pythonutil/NumTut/view.py", line 146, in ?
       class ThreadedTk(Thread):
   NameError: Thread
   >>> 


-- 
Randall Hopper
aa8vb at yahoo.com



From smalleys at gte.net  Thu Apr 20 00:35:29 2000
From: smalleys at gte.net (sue)
Date: Thu, 20 Apr 2000 04:35:29 GMT
Subject: Very Horrible Question About Goto's
References: <6325E90C8B6ED21189080001FA7E2ADA0D4051@IBA1><2hkL4.8514$5D.18609@ozemail.com.au> <sfruuci2eoc31@corp.supernews.com>
Message-ID: <38FE8924.F2F9FFF0@gte.net>

Simple blocks can be translated, which is probably the
infamous 80% of your code.  The remaining 10-20% probably
isn't worth the effort; rewrite it.  We did have a bit
more luck when we added "hints", based on data, context,
and so on, but 5-10% needed a redesign.  Good luck.

Sue.


From lexberezhny at email.msn.com  Sat Apr  1 16:53:56 2000
From: lexberezhny at email.msn.com (lexberezhny)
Date: Sat, 1 Apr 2000 16:53:56 -0500
Subject: Python/GTK on Windows?
References: <#XSVXB5m$GA.240@cpmsnbbsa03> <8c5m42$1dv@gap.cco.caltech.edu>
Message-ID: <ewzoHQCn$GA.361@cpmsnbbsa04>

Thanks you very much.

 - Lex

Robert Kern <kern at caltech.edu> wrote in message
news:8c5m42$1dv at gap.cco.caltech.edu...
> In article <#XSVXB5m$GA.240 at cpmsnbbsa03>,
> "lexberezhny" <lexberezhny at email.msn.com> writes:
> > Hi,
> >   I was wondering if anyone was able to get PyGTK to work on windows,
and if
> > its possible at all. I am trying to port a Linux app to windows that
uses
> > PyGTK. If you know anything about this or can point me in the right
> > direction, I would really appreciate it. Thanks.
>
> http://theopenlab.uml.edu/pygtkwin/
>
> >  - Lex Berezhny
>
> --
> Robert Kern
> kern at caltech.edu
>
> "In the fields of hell where the grass grows high
>  Are the graves of dreams allowed to die."
>   -- Richard Harter




From Lisowski.Tomasz at sssa.nospam.pl  Fri Apr  7 00:23:52 2000
From: Lisowski.Tomasz at sssa.nospam.pl (Tomek Lisowski)
Date: Fri, 07 Apr 2000 04:23:52 GMT
Subject: Odp: Creating variables on the fly...
References: <meh9-25D0C5.15160303042000@news.cit.cornell.edu> <38E8EFF0.D71D0415@gmd.de>
Message-ID: <spdH4.43688$a01.958441@news.tpnet.pl>

U?ytkownik Joachim Kaeber <kaeber at gmd.de> w wiadomooci do grup dyskusyjnych
napisa?:38E8EFF0.D71D0415 at gmd.de...
> Hi,
>
> what about this:
>
> def add(*args):
>    sum=0
>    for arg in args:
>      sum=sum+arg
>    return sum
>
> add(1,2,3,4)
> => 10
>
> add (1,2,3,4,5,6,7)
> => 28

Why not:

def add(*args):
  return reduce(lambda x, y: x+y, args)

I am not completely sure, but it may be faster, than a Python for loop

Tomasz Lisowski

>
>
> Matthew Hirsch wrote:
> >
> > Hi All,
> >
> > Let's say I have four variables:
> >
> > a=1
> > b=2
> > c=3
> > d=4
> >
> > And I have a function that adds these variables together:
> >
> > def add(a,b,c,d):
> >    return a+b+c+d
> >
> > But now let's say I have twenty variables that I want to add together.
> > This function will no longer work.  I would have to rewrite it as
> > a+b+c+d+e+f+g+...+(20th letter).  Is there a way to dynamically create a
> > variable name?  So that my add function can automatically determine how
> > many variable names to create and then add the values together.  In
> > other words, I'd ideally like something like:
> >
> > def add(number_of_variables):
> >    return a+b+c+...+(letter corresponding to number_of_variables)
> >
> > Thanks for your help,
> >
> > Matt
>
> --
> Joachim Kaeber (kaeber at gmd.de) (http://imk.gmd.de/~kaeber)
> GMD - Forschungszentrum Informationstechnik GmbH
>     - German National Research Center for Information Technology
> Schloss Birlinghoven                    email: kaeber at gmd.de
> D-53757 Sankt Augustin 1, Germany       phone: (+49 2241) 14-2546



From bwilk_97 at yahoo.com  Sat Apr  8 18:11:58 2000
From: bwilk_97 at yahoo.com (Bill Wilkinson)
Date: Sat, 08 Apr 2000 22:11:58 GMT
Subject: tabular data
References: <8co4l2$5r2$1@nnrp1.deja.com>
Message-ID: <O8OH4.46089$Uz3.487916@typhoon.midsouth.rr.com>

<<<For example, say that I have the following tabular data: [(2,1,"REC02"),
<<<(1,3,"REC03"),(3,1,"REC01")], and I want to obtain: [(1,3,"REC03"),
<<<(2,1,"REC02"),(3,1,"REC01")], that is, the tabular data should be
<<<sorted on the first and then the second column.

You can get lists.py from:
http://starship.python.net/crew/lemburg/

It has this routine, which is attributed to Guido and others.

def sortedby(list,*indices):
   nlist = map( lambda x,indices=indices:
   map( lambda i,x=x: x[i], indices) + [x],
   list)
    nlist.sort()
    return map( lambda l: l[-1], nlist)

The list in not sorted in place though.

<<<By the way, should I represent rows by tuples or by embedded lists?

They need to be lists if you want to edit them... In my opinion, a list of
dictionaries is the best way to go.....

<<<Another problem: to visually represent/edit tabular data, I need an
<<<editable grid widget (with possibly comboboxes as cell editors).
<<<TkInter does not seem to have an appropriate grid widget.

wxPython has a grid that is ok... There is a new grid that will be available
soon that will be a bit better.




From jwbnews at scandaroon.com  Sun Apr  2 18:07:26 2000
From: jwbnews at scandaroon.com (John W. Baxter)
Date: Sun, 02 Apr 2000 15:07:26 -0700
Subject: Emacs Python mode
References: <yayk8iii0ex.fsf@ganglot.ifi.uio.no>                                        <8c794n$mjq$1@zingo.tninet.se>
Message-ID: <jwbnews-C0A1B0.15072602042000@news.olympus.net>

In article <8c794n$mjq$1 at zingo.tninet.se>, Andr? Dahlqvist 
<andre at beta.telenordia.se> wrote:

> > all in all, *emacs* has lots of shortcuts for handy things. I wish I
> > knew more about it.. :)
> 
> While we're on the subject of Emacs Python mode, why is it that I can't 
> indent a line under a line which has been commanted out? Emacs 
> refuses to let me tab in there, but if I remove the comment it will 
> happily 
> let me do it.

Experiment with whether or not there is a space following the #

  --John (who didn't fully understand that part of the fine manual and 
hasn't experimented much)

-- 
John W. Baxter   Port Ludlow, WA USA  jwbnews at scandaroon.com


From donn at u.washington.edu  Mon Apr 17 17:53:34 2000
From: donn at u.washington.edu (Donn Cave)
Date: 17 Apr 2000 21:53:34 GMT
Subject: else clauses in while and for loops
References: <2JKK4.325$vr2.19414@news.uswest.net> <Zv_C4.6415$4N.169440@news.tpnet.pl> <38DD2EE3.C7F3CB54@webamused.com> <8boe2k$gmh$1@inputplus.demon.co.uk> <v0RD4.13$n68.186989568@newsb.telia.net> <8c7b29$37v$1@inputplus.demon.co.uk> <38F79B93.B03F0A6B@mindspring.com>
Message-ID: <8dg14u$f08$1@nntp6.u.washington.edu>

Quoth "Jeff Massung" <jmassung at magpiesystems.com>:
| I'm new to Python, so perhaps I'm miss reading something in these posts, but
| it appears as though the "else" statement (especially in while loops -
| unless infinite) always get executed regardless.
|
| i = 0
| while i<=5:
|     i = i + 1
|     print i
| else:
|     print 10
|
| should spit out "1 2 3 4 5 10" right?

Yes, but if you look at the examples, they also use "break".  After a
loop, "else" basically means "if no iteration of the loop met the
conditions to break out."  It's a common usage, where in C we'd compare
the final loop index with the maximum (or whatever) loop index to see
if the loop ran all the way through.

| Now a use that I could definitely see for the "else" would be if the loop
| were never entered at all! Example:
|
| i = 6
| while i<=5:
|     i = i+1
|     print i
| else:
|     print 10
|
| would print "10", and if 'i' started at 0, would print "1 2 3 4 5".
|
| That could definitely be useful, because it would eliminate an "if"
| statement right before the "while" statement. Extremely useful for debugging
| purposes.

I think I would go along with that if you took out "Extremely" - and then
your paragraph would justify better on the right margin, too!  Seriously,
"else" does invariably execute if the loop was never run once, but as
mentioned above it also executes in other circumstances.

	Donn Cave, University Computing Services, University of Washington
	donn at u.washington.edu


From gmcm at hypernet.com  Tue Apr 11 10:29:11 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Tue, 11 Apr 2000 10:29:11 -0400
Subject: Thread-safe atomic operation
In-Reply-To: <8cuf0o$p9a$1@news1.kornet.net>
Message-ID: <1256659126-41087547@hypernet.com>

Dong-gweon Oh writes:

> Are these statements thread-safe?
> 1. import xxx
> 2. xxx = __import__('xxx')
> 3. reload(xxx)
> 4. dict['xxx'] = sys.modules['xxx']
> 5. dict.has_key('xxx')

Numbers 2 and 4 could be interrupted between evaluation of 
the RHS and assigment to LHS. For #2 that would be of 
concern only if xxx was a global.



- Gordon



From hdiwan at rpi.edu  Wed Apr  5 10:31:38 2000
From: hdiwan at rpi.edu (Hasan Diwan)
Date: 5 Apr 2000 14:31:38 GMT
Subject: Why should I switch to Python? - Infinity of Primes
References: <NDBBKEGCNONMNKGDINPFGEJDDFAA.infonuovo@email.com> <8ccoug$f4o$1@pegasus.csx.cam.ac.uk> <7F51E902E09F3F60.8F612607812ED16B.82CB8B41435E14FD@lp.airnews.net> <38EA369C.26A1C25@math.okstate.edu> <38EAC8C4.ADC25640@cosc.canterbury.ac.nz>
Message-ID: <8cfioa$qvc$1@newsfeeds.rpi.edu>

Proof of infinity of primes by contradiction:=
	Suppose there is a finite number of primes p, where p is the set
	of all primes up to p[k]. Now multiply all the elements of p
	together. You now have a number that is a product of primes, add 1
	to this number, and you have a new prime, so my assumption is
	bogus and there are an infinite number of primes.
--
Hasan Diwan [hdiwan at pobox.com] :)
Rensselaer Polytechnic Institute 
Computer Science Department
http://forsythe.dhs.org/~hdiwan 


From mhammond at skippinet.com.au  Thu Apr 13 23:09:49 2000
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Fri, 14 Apr 2000 03:09:49 GMT
Subject: windows registry
References: <era0iGYp$GA.273@net003s> <8d5h78$fpq$1@nnrp1.deja.com>
Message-ID: <1_vJ4.809$1D.5094@news-server.bigpond.net.au>

Note that in Python 1.6, there is a "winreg" standard module - ie, no need
for the win32 extensions.

This is taken from the win32api sources.  The relevant differences are:

* The module name has obviously changed!  The winreg module will raise
winreg.error (which happens to be thenew standard WindowsError exception
object, derived from OSError, derived from EnvironmentError, derived from
StandardError - so any of these error objects will catch the exceptions!)

* The function names have changed (eg, win32api.RegCreateKey ->
winreg.CreateKey)

* Security information is not supported - none must be passed to all
winreg functions that accept security information.

* winreg has docstrings, win32api has help files :-)

Apart from that, all args and usage are identical to the win32api
functions.  Thus, in the general case, code can be changed to use winreg
from win32api with a couple of search-and-replaces!

Mark.




From cut_me_out at hotmail.com  Wed Apr 26 09:52:43 2000
From: cut_me_out at hotmail.com (Alex)
Date: 26 Apr 2000 09:52:43 -0400
Subject: Metaclasses?
References: <3905F4A0.2B39C705@darmstadt.gmd.de> <slrn8gc6qj.p7e.quinn@hedono.ugcs.caltech.edu> <200004261201.IAA05678@zippy.init.com> <m3k8hlou1f.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <etdaeih2clw.fsf@w20-575-108.mit.edu>

Such a fine comrade you are Michael!  Raising (meta)class consciousness
and speeding the arrival of universal selflessness in one fell swoop!

Alex.


From bawolk at pacbell.net  Thu Apr 20 00:19:39 2000
From: bawolk at pacbell.net (Bruce Wolk)
Date: Wed, 19 Apr 2000 21:19:39 -0700
Subject: Tkinter aspect question
Message-ID: <38FE855B.79F7881A@pacbell.net>

I am trying to create a square window that will remain square when its
size is adjusted.  I found the aspect method, which looked promising,
and tried the following:

from Tkinter import *
master=Tk()
win = Toplevel(master)
win.geometry('400x400+100+100')
win.aspect(1.0,1.0,1.0,1.0)

This does not work, i.e., the window does not maintain its aspect.  Can
anyone help me out?  Thanks.

Bruce


From jmassung at magpiesystems.com  Wed Apr 19 11:12:55 2000
From: jmassung at magpiesystems.com (Jeff Massung)
Date: Wed, 19 Apr 2000 09:12:55 -0600
Subject: Python in the future
References: <38FDB8A7.E701F81F@sec.noaa.gov>
Message-ID: <8ZjL4.273$CP3.3311@news.uswest.net>

IMHO, C++ will DEFINITELY be here 10 years from now, as will assembler -
they are just inherent in computers - the core languages (used to make
Python, Java, Basic, etc). But I personally think that as computers grow in
speed exponentially, and the internet becomes much faster (10 years, maybe
everyone will have T1 access?) interpreted languages will be increasingly
more important and used widespread - not for sale (because the code is
needed), but rather for distribution within a business, or with other
programs, the web, etc.

Of course I'm refering to interpreted languages as a whole. Now to just
Python. I am new to Python, but being an accomplished C/C++ programmer, I
have already begun seeing the power and use of Python. And like Java (and I
believe unlike Perl) has a great following and good programmers commited to
its success (Perl I feel is more of a "hack"ing language that people that
don't code in other languages use).

So to conclude - I feel Python and Java will have a strong presence in the
near and far futures. Of course they will be different from how they are
today - and possibly in how they are used, but they will be there.

Jeff

j vickroy wrote in message <38FDB8A7.E701F81F at sec.noaa.gov>...
>Hello all,
>
>I'm relatively new to Python and am trying to gain support for it at my
>work place.  To that end, I'm giving a brief (overview) presentation.
>One of the items I have been asked to consider is where will Python be
>5-10 years from now.  I certainly don't know where C++ or Java will be
>in that time frame!  Any ideas on where to research the future plans for
>Python would be appreciated.
>
>Thanks for your time,
>
>- jv
>




From bjorn at roguewave.com  Tue Apr  4 14:27:19 2000
From: bjorn at roguewave.com (Bjorn Pettersen)
Date: Tue, 04 Apr 2000 12:27:19 -0600
Subject: Tuples -- who needs 'em
References: <00e101bf9d80$a8384d20$74eb0b18@stcla1.sfba.home.com> <va3G4.1110$n68.190454272@newsb.telia.net> <037f01bf9e55$6d328760$74eb0b18@stcla1.sfba.home.com>
Message-ID: <38EA3407.534827DF@roguewave.com>

Bob Alexander wrote:
[...]
> However, too bad about that nasty syntax irregularity that using parens for
> tuples presents  (1,)  (Couldn't resist  ;-)
[...]

Contrary to popular belief, it's actually the comma that creates the
tuple, not the parens...  As an example, I recently copied code from a
dictionary to an assignment a little to hasty:

   foo = {
      'a' : 1,
      'b' : 2,
   }

to:

   a = 1,
   b = 2,

and it took me way to long to figure out where the eventual "unpack
tuple of wrong size" error was coming from... *sigh*   Serves me righ, I
suppose <wink>

--bjorn



From spex66 at my-deja.com  Thu Apr 27 05:25:42 2000
From: spex66 at my-deja.com (spex66 at my-deja.com)
Date: Thu, 27 Apr 2000 09:25:42 GMT
Subject: Namespace Problem with global declaration in module
References: <8e6vv9$4n8$1@nnrp1.deja.com>
Message-ID: <8e912d$cpi$1@nnrp1.deja.com>

hmm, deja is not so fast as my newsreader :), so here is a reply of
Gordons message

>spex66 at my-deja.com wrote:
>
>> Hi,
>> it seemed very simple but don't worked like expected:
>>
>> #FILE: test.py
>> dd = None #INIT a shadow instance for automated access
>> class zz:
>>     #ONLY one class will be instantiated
>>     def __init__(self):
>>         self.name = 'hallo'
>>         global dd
>>         dd = self
>> #END OF FILE
>>
>> >>> from test import *
>
>There's your problem. "from test import *" copies names from
>test to your namespace. Now dd is another name for None.
>yy() affects the dd in test's namespace, not yours.
>
>- Gordon

I see the point, but I like the from statement :)
Is it a bug? It's not the way I expected this code to work... and I
cannot find a hook in the documentation that explicit explained this
misbehaviour.

Any idea for a workaround (I need the from-statement for further code
readability)?

thanks
Peter
(=PA=)

>
>> >>> yy = zz()
>> >>> dd.name
>>     ...
>>     AttributeError
>>     ...
>



Sent via Deja.com http://www.deja.com/
Before you buy.


From pinard at iro.umontreal.ca  Tue Apr 25 19:46:59 2000
From: pinard at iro.umontreal.ca (=?ISO-8859-1?Q?Fran=E7ois_Pinard?=)
Date: 25 Apr 2000 19:46:59 -0400
Subject: Sourceforge and Python
Message-ID: <oq8zy1zqto.fsf@titan.progiciels-bpi.ca>

Hi, people.

Bruno Haible is convincing me to put `recode' in an open source environment,
like `sourceforge', say.  I have been reluctant to CVS so far (for various
reasons), and if I was doing such a move, I would like to bring `recode'
a bit closer to Python, in one way or another, given this is easily doable.
Just to make things clear, `recode' is C code currently, this is not Python.

So, merely beginning to explore the possibility, I'm asking for advice
on this list, as to if there is environments offering a set of services
similar to `sourceforge', but that would be especially Python-using,
or Python-aware.  Of course, and also given some recent saga reported on
this list, stability and availability are parts of the requirements :-).

It would be tempted to think such a site might be a better investment of
energies in the long run, presumably.  But I may also be wrong, and it might
not be an important criterion, after all.  Tell me if you have an opinion.

-- 
Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard





From jbauer at rubic.com  Wed Apr 26 23:24:17 2000
From: jbauer at rubic.com (Jeff Bauer)
Date: Wed, 26 Apr 2000 22:24:17 -0500
Subject: Python compiler ? (was Re: Python for drivers?)
References: <3900EDBE.282BC85@dsuper.net> <8dr2pn$ec8$1@newshost.accu.uu.nl> <3906f804@nntp.server.uni-frankfurt.de> <015b01bfaf8b$9cd03930$0500a8c0@secret.pythonware.com> <39076316@nntp.server.uni-frankfurt.de> <8e7sop$5io$1@nnrp1.deja.com> <FtnGu1.Dso@world.std.com> <slrn8gf9ah.lia.bgdarnel@Sull-04-227.rh.ncsu.edu>
Message-ID: <3907B2E1.DF6B19B4@rubic.com>

[ Discussion about sending somebody $60K to build 
  a better Python mousetrap. ]


I haven't kept up with this thread, but from what 
I've skimmed, most of the posts appear to assume 
we'd have a useful technology if this project was 
financed.  I don't think this is conclusive, however.  
I can think of a few ways $60K could be more 
effectively spent on improving Python.  

after-budgeting-for-Marketing-we'd-*still*-have-almost-
  $10,000-to-get-some-work-done-ly y'rs,

Jeff Bauer
Rubicon Research



From mwh21 at cam.ac.uk  Tue Apr 11 20:48:19 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 12 Apr 2000 01:48:19 +0100
Subject: issubclass funny business
References: <m2snwsd5ah.fsf@netmeg.net> <ekII4.1212$Za1.21055@newsc.telia.net> <m2purwcfjf.fsf@netmeg.net>
Message-ID: <m3itxojga4.fsf@atrus.jesus.cam.ac.uk>

Les Schaffer <godzilla at netmeg.net> writes:

> </F> said:
> 
> > how about:
> >        if not isinstance(std, StandardArrayFileScanner):
> 
> i tried that first, but same problem.
> 
> > when you run FileScanner as a script, it's evaluated in the __main__
> > namespace.  when you import it, it's evaluated again, in the
> > FileScanner namespace.
> 
> groan.... its not intuitive (to me). because i can make CIE_Standards
> class inherit from FileScanner classes in the udder module no problem
> at runtime, so its a surprise to see it won't trace the tree properly
> at runtime for isinstance issubclass.
> 
> > solution: never import the script you use as the main program.
> 
> i am only at the testing stage. neither of these modules will be main
> at "usage" time.

This is definitely a pain.  It'd be nice if when you `reload'ed
modules that contain classes you could "smash" extant instances of
those classes so that they were instances of the new class, but I
can't for the life of me see how you might do this.

wishful-ly y'rs
Michael

-- 
  On the other hand,  the following areas are subject to boycott
  in reaction to the rampant impurity of design or execution, as
  determined after a period of study, in no particular order: 
    ...                            http://www.naggum.no/profile.html


From anthony at dstc.edu.au  Wed Apr 26 05:37:52 2000
From: anthony at dstc.edu.au (Anthony J Wilkinson)
Date: Wed, 26 Apr 2000 19:37:52 +1000 (EST)
Subject: huh? Python says that AttributeError: 'string' object has no
 attribute 'split'
In-Reply-To: <8e62qm$511$1@nnrp1.deja.com>
Message-ID: <Pine.OSF.4.20.0004261936280.21454-100000@azure.dstc.edu.au>

On Wed, 26 Apr 2000 schmitthead at my-deja.com wrote:

[...]
> Traceback (innermost last):
[...]  
>   File "C:\bin\mac2dos.py", line 9, in mac2dosfile
>     lines = myline.split( myline, chr(10) )
> AttributeError: 'string' object has no attribute 'split'

[...]

>  lines = myline.split( myline, chr(10) )

The above should be:

   lines = string.split(myline, chr(10))

Regards,
Anthony





From smalleys at gte.net  Thu Apr 20 00:43:13 2000
From: smalleys at gte.net (sue)
Date: Thu, 20 Apr 2000 04:43:13 GMT
Subject: A tkMessageBox question!
References: <38FDC335.B9FDCFE4@icestormfx.com>
Message-ID: <38FE8AF6.2A4CDBDA@gte.net>

I don't remember the exact syntax, but try:

msg = TkMessageBox("text")
msg.configure( font = "something_decent" ).

You might have to check that the font exists.

Sue.


From phd at phd.russ.ru  Thu Apr 27 10:46:46 2000
From: phd at phd.russ.ru (Oleg Broytmann)
Date: Thu, 27 Apr 2000 14:46:46 +0000 (GMT)
Subject: DCOracle
In-Reply-To: <NEBBIEPHGLOHPEDKJLGJAEBICAAA.youngd@digiship.com>
Message-ID: <Pine.LNX.4.21.0004271445360.11789-100000@fep132.fep.ru>

On Thu, 27 Apr 2000, Darren Young wrote:
> Any pointers on examples of working code out there?

from DCOracle import Connect
jwt = Connect("%s/%s@%s" % (login, password, SID))

cursor = jwt.cursor()                                                           
format = "%s | "*13 + "%s\n"                                                    
                                                                                
cursor.execute("SELECT * FROM sides_view")                                      
while 1:                                                                        
   olist = cursor.fetchmany(10)                                                 
   if not olist: break                                                          
   for row in olist:                                                            
      sys.stdout.write(format % row)                                            
                                                                                
cursor.close()                                                                  
jwt.close()

Oleg.            (All opinions are mine and not of my employer)
---- 
    Oleg Broytmann      Foundation for Effective Policies      phd at phd.russ.ru
           Programmers don't die, they just GOSUB without RETURN.




From echuck at mindspring.com  Tue Apr 25 10:13:34 2000
From: echuck at mindspring.com (Chuck Esterbrook)
Date: Tue, 25 Apr 2000 10:13:34 -0400
Subject: sys.exit
References: <3904C35C.B16BDE75@mindspring.com> <8e2hss$1250$1@nntp6.u.washington.edu>
Message-ID: <3905A80E.7A8323F9@mindspring.com>

Donn Cave wrote:
> 
> Quoth Chuck Esterbrook <echuck at mindspring.com>:
> | The short question:
> |
> | If I catch a SystemExit exception (because of an invocation of
> | sys.exit(code)), how do I examine the argument to sys.exit()?
> 
> ...
> | Upon printing the "excInfo" tuple I get:
> |
> |     (<class exceptions.SystemExit at 80afc70>, <exceptions.SystemExit instance at 8110f58>, <traceback object at 811f540>)
> |
> | And the attributes of the SystemExit instance are:
> |
> |     ['__doc__', '__init__', '__module__']
> |
> | I don't know how to get my hands on the exit code.
> 
> That's not the instance you're looking at there, but the class, so
> that's kind of a red herring.  The instance has a 'code' attribute
> for you.
> 
> So you need something like
> 
>         eclass, einstance, etrace = sys.exc_info()
>         if eclass == SystemExit:
>                 if einstance.code == 0:
>                         self._errorOccurred = 0

Silly me: I was printing [0] of the tuple rather than [1] by mistake. Don & Tim are both right. The instance has attributes "args" and "code". For my purpose, I prefer using "code".

Thanks!

-Chuck


From danzap_python at yahoo.com.br  Thu Apr 13 12:38:28 2000
From: danzap_python at yahoo.com.br (=?iso-8859-1?q?Daniel=20Zapata?=)
Date: Thu, 13 Apr 2000 13:38:28 -0300 (ART)
Subject: How do I execute a program ?
Message-ID: <20000413163828.28296.qmail@web4303.mail.yahoo.com>

How do I execute a program or a batch file (under
Windows) using python and get the return information
???


Thanks,

Daniel

____________________________________________________________________________
YAHOO! BATE-PAPO
Conhe?a gente nova, fa?a amigos e divirta-se! - http://www.yahoo.com.br



From shae at webwitches.com  Tue Apr 18 18:13:03 2000
From: shae at webwitches.com (Shae Erisson)
Date: Tue, 18 Apr 2000 22:13:03 GMT
Subject: gaming and python
References: <000c01bfa91c$f421d6a0$0f8d0b3f@oemcomputer> <8dhifa$aom$1@nnrp1.deja.com>
Message-ID: <38FCDF99.352876CE@webwitches.com>

sp00fD wrote:
> 
> Well, I'd have a hard time believing that you could program something
> like quake, but you could probably do a decent job with tetris or the
<snippage>

Maybe you could write quake in python....
Crystalspace is an open source 3D engine that claims to be scriptable
with Python.

this ripped without permission from http://crystal.linuxgames.com :

Crystal Space is a free (LGPL) and portable 3D engine written in C++. It
supports: true six degree's of freedom, colored lighting, mipmapping,
portals, mirrors, alpha transparency, reflective surfaces, 3D sprites,
scripting (using Python or other languages), 8-bit, 16-bit, and 32-bit
display support, Direct3D under Windows, Glide and OpenGL on Windows,
Linux, OS/2, Macintosh, BeOS, ... See the extensive list of features for
more details.

Crystal Space currently runs on Linux, general Unix, Windows, Windows
NT, OS/2, BeOS, NextStep, OpenStep, MacOS/X Server, DOS, and Macintosh.
It can optionally use OpenGL (Windows, Linux, Mac, OS/2, BeOS), Direct3D
(Windows), GGI (Linux), Allegro (DOS), X11 (Unix/Linux) and SVGALIB
(Linux). It can also optionally use assembler routines (GCC/DJGPP,
Visual C++ and Intel only currently) and MMX if the processor supports
it.
-- 
sHae mAtijs eRisson (sHae at wEbwitchEs.coM) gEnius fOr hIre
   bRing mE fOur sQuirrels aNd nO oNe wIll gEt hUrt
13:00pm up 1 season, 1 squirrel, load average: 1+1j 0.5+2j 0.2+0.5j


From gmunsey at adobe.com  Mon Apr 17 00:44:02 2000
From: gmunsey at adobe.com (Grant Munsey)
Date: Sun, 16 Apr 2000 21:44:02 -0700
Subject: How to fire off a web browser process from Python in Win?
Message-ID: <aDwK4.1834$lM4.436461@news.pacbell.net>

Is there a call somewhere in the std library for staring up the default
( the one that the user has set for opening .htm and .html files)
web browser on Win 95/8/NT/2K?




From un4e at rzstud1.rz.uni-karlsruhe.de  Thu Apr 27 09:30:53 2000
From: un4e at rzstud1.rz.uni-karlsruhe.de (Bjoern Giesler)
Date: 27 Apr 2000 13:30:53 GMT
Subject: PyOpenGL
Message-ID: <8e9fed$lm4$1@news.rz.uni-karlsruhe.de>

Hi,

is there an archive that carries a newer version of PyOpenGL? The newest I
have is from the contrib-09-Dec-99 directory on ftp.python.org, and that
seems to require tcl7.6, which I don't want to use anymore.

TIA,
		--Bj?rn


From dwhite2 at seas.upenn.edu  Tue Apr  4 21:20:36 2000
From: dwhite2 at seas.upenn.edu (dave white)
Date: Tue, 04 Apr 2000 21:20:36 -0400
Subject: reading data from a file
Message-ID: <38EA94E3.7B249C3@seas.upenn.edu>

Hello, can anyone please help with these basic questions:

1)  I'm using file.readlines() to read in some data.  Is there any quick
and efficient way to automatically get rid of the endofline '\012'
character that gets appended to the strings at the end of every line?

2)  Is there a way for python to read a line from a file in such a way
that it knows its a collection data type and not just a string? i.e.
read something like [[1,2][3,4][5,6]] and know its a list of lists?

thanks...
david


From calishar at *NOSPAM*home.com  Tue Apr 18 23:26:26 2000
From: calishar at *NOSPAM*home.com (Calishar)
Date: Wed, 19 Apr 2000 03:26:26 GMT
Subject: win32all-128.exe availability?
References: <38fd01b2$0$10982@senator-bedfellow.mit.edu>
Message-ID: <3k9qfs0ropdsv515f9he7lkpdrf1qfputv@4ax.com>

On Tue, 18 Apr 2000 20:49:08 -0400, "Robert" <soundhack at hotmail.com>
wrote:

>Now that starship has crashed, I was wondering if it was possible to
>get the 128 build of Pythonwin extensions off the net? By an unfortunate
>coincidence, my own hard drive chose to crash around the time starship
>did.

Don't know if any sites have a mirror of win32all-128, but I have it
on my HD (love being a packrat, still have -127 too). If you want I
can try e-mailing it to you. I say try because the file is 3.2M and
@home gives me a 5M message size limit.

Let me know if you want it via mail.

Calishar

calishar at home.com


From sabren at manifestation.com  Thu Apr 20 13:18:13 2000
From: sabren at manifestation.com (Michal Wallace (sabren))
Date: Thu, 20 Apr 2000 13:18:13 -0400 (EDT)
Subject: python templating system
In-Reply-To: <CDA59A302BEB26AC.DE30D79F63ED12C2.5C8017889EA62DAB@lp.airnews.net>
Message-ID: <Pine.BSF.4.10.10004201316530.13031-100000@chi.pair.com>

On 20 Apr 2000, Cameron Laird wrote:

> In article <Pine.BSF.4.10.10004191555250.14584-100000 at chi.pair.com>,
> Michal Wallace (sabren) <sabren at manifestation.com> wrote:
> >On Wed, 19 Apr 2000 pozman at my-deja.com wrote:
> >
> >> Does anyone know of a templating system (other than zope) that is
> >> implemented with python [ comparisons would be embperl with perl]
> > 
> >http://zebra.sourceforge.net/
> 			.
> 			.
> 			.
> According to the text currently at that page, Zebra does *not* now
> interpret Python.  However, I collect systems which do at <URL:http://
> starbase.neosoft.com/~claird/comp.lang.python/web_python.html>.

Uh. Oops. Must be that the idiot who wrote it (me) didn't update the
documentation yet. Zebra exports Python well and PHP poorly.

Cheers,

- Michal
-------------------------------------------------------------------------
http://www.manifestation.com/         http://www.linkwatcher.com/metalog/
-------------------------------------------------------------------------




From bjorn at roguewave.com  Wed Apr  5 13:19:12 2000
From: bjorn at roguewave.com (Bjorn Pettersen)
Date: Wed, 05 Apr 2000 11:19:12 -0600
Subject: Teaching the art of programming, in python
References: <20000405053920.20579.qmail@web2102.mail.yahoo.com>
	    <8cfe42$o02$1@nnrp1.deja.com> <dpJG4.3744$HG1.106895@nnrp1.uunet.ca> <8cfqf7$4ml$4@newsserver.rrzn.uni-hannover.de>
Message-ID: <38EB7590.1827F01C@roguewave.com>

Bernhard Reiter wrote:
> 
> In article <dpJG4.3744$HG1.106895 at nnrp1.uunet.ca>,
>         "Warren Postma" <embed at geocities.com> writes:
> 
> > The Python texts I've read are good References to Getting Stuff Done, but
> > they Don't Teach the Art. Speaking of Dummies, was there/is there a LISP for
> > Dummies, or Smalltalk for Dummies?  Can Python for Dummies be far off?
> 
> Yes, because people who do Python are no dummies.
> At least I hope so.

Reminds me of something Stroustrup wrote:

  "...  In a small dedicated community, life is relatively easy. people
do their homework,
people have access to reasonable sources of information, gross errors
and misconceptions are corrected before they can cause significant harm,
compilers and
teaching materials are up-to-date, etc. 

This is not and cannot be the case in a multi-hundred-thousand member
community: Some will be taught out of outdated or unsuitable books, some
will use
antiquated compilers and tools, some will be taught by charlatans, some
will be remote from current and reliable news-sources, some will have
unsuitable rules
and regulations imposed on their work, etc. Also, in a rapidly growing
community, most users will be novices. "

For the full article, check out his website -- it's in the section
titled "C++ programmers are idiots" :-)
(http://www.research.att.com/~bs/blast.html).

--bjorn



From guido at python.org  Tue Apr 11 16:29:16 2000
From: guido at python.org (Guido van Rossum)
Date: Tue, 11 Apr 2000 16:29:16 -0400
Subject: ANNOUNCE: Python 1.6 alpha 2
Message-ID: <200004112029.QAA09762@eric.cnri.reston.va.us>

I've just released a source tarball and a Windows installer for Python
1.6 alpha 2 to the Python website:

  http://www.python.org/1.6/

If you missed the announcement for 1.6a1, probably the biggest news is
Unicode support.  More news is on the above webpage; Unicode is being
discussed in the i18n-sig.

Most changes since 1.6a1 affect either details of the Unicode support,
or details of what the Windows installer installs where.


Note: this is an alpha release.  Some of the code is very rough!
Please give it a try with your favorite Python application, but don't
trust it for production use yet.  I plan to release several more alpha
and beta releases over the next two months, culminating in an 1.6
final release before June first.

We need your help to make the final 1.6 release as robust as possible
-- please test this alpha release!!!

--Guido van Rossum (home page: http://www.python.org/~guido/)



From rjroy at takingcontrol.com  Wed Apr 12 22:45:58 2000
From: rjroy at takingcontrol.com (Robert Roy)
Date: Thu, 13 Apr 2000 02:45:58 GMT
Subject: Using xmllib
References: <2B1262E83448D211AE4B00A0C9D61B03BA72A1@MSGEURO1>
Message-ID: <38f533f1.513177687@news1.on.sympatico.ca>

On Wed, 12 Apr 2000 13:38:41 +0200, Pieter Claerhout
<PClaerhout at CREO.BE> wrote:

>Hello,
>
>does anyone has an example on how to use the xmllib to parse an xmlfile?
>I have an xml-file, which looks a bit like this:
>
><!-- a general comment -->
><COLLECTION VERSION="1">
>    <GENERAL PATH="S:\" />
>    <ITEM NAME="Item1">
>        <DESCRIPTION>Item 1 description.</DESCRIPTION>
>        <DATE>Thu Dec 16 16:35:05 1999</DATE>
>    </ITEM>
>    <ITEM NAME="Item2">
>        <DESCRIPTION>Item 2 description.</DESCRIPTION>
>        <DATE>Wed Mar 15 11:54:27 2000</DATE>
>    </ITEM>
></COLLECTION>
>
>I want to convert it to something like this:
>
>Collection from path s:\
>
>- Item 1 (Thu Dec 16 16:35:05 1999)
>  Item 1 description.
>- Item 2 (Wed Mar 15 11:54:27 2000)
>  Item 2 description.
>
>Anyone an idea on how to do that with XMLlib?
>
>Kind regards,
>
>
>Pieter
>Pieter_Claerhout at creoscitex.com
>

This is a fairly trivial example based on what you posted.

Bob
++++++++++

import xmllib, string

class myparser(xmllib.XMLParser):
    def __init__(self):
        xmllib.XMLParser.__init__(self)
        self.currentdescrip = ''
        self.currentdate = ''
        self.currentitem = ''
        self.currentdata = []
        
    def handle_data(self, data):
        self.currentdata.append(data)

    def start_COLLECTION(self, attrs):
        print "Collection from path",
                
    def end_COLLECTION(self):
        pass

    def start_GENERAL(self, attrs):
        print attrs['PATH']
                
    def end_GENERAL(self):
        pass

    def start_ITEM(self, attrs):
        self.currentitem = attrs['NAME']
                
    def end_ITEM(self):
        print '-%(currentitem)s (%(currentdate)s)\n
%(currentdescrip)s' % self.__dict__

    def start_DESCRIPTION(self, attrs):
        self.currentdata = []
                
    def end_DESCRIPTION(self):
        self.currentdescrip = string.join(self.currentdata,'')
        
    def start_DATE(self, attrs):
        self.currentdata = []
                
    def end_DATE(self):
        self.currentdate = string.join(self.currentdata,'')


### some test code
data= """\
<COLLECTION VERSION="1">
    <GENERAL PATH="S:\\" />
    <ITEM NAME="Item1">
        <DESCRIPTION>Item 1 description.</DESCRIPTION>
        <DATE>Thu Dec 16 16:35:05 1999</DATE>
    </ITEM>
    <ITEM NAME="Item2">
        <DESCRIPTION>Item 2 description.</DESCRIPTION>
        <DATE>Wed Mar 15 11:54:27 2000</DATE>
    </ITEM>
</COLLECTION>

"""

if __name__ == "__main__":
    p=myparser()
    p.feed(data)
    p.close()





From tismer at tismer.com  Sun Apr 16 14:26:17 2000
From: tismer at tismer.com (Christian Tismer)
Date: Sun, 16 Apr 2000 20:26:17 +0200
Subject: [ANNOUNCE] Garbage collection for Python
References: <20000407221750.A29279@acs.ucalgary.ca>  <m3r9cagx3c.fsf@atrus.jesus.cam.ac.uk> <l03102809b51d41895a80@[193.78.237.135]> <38F9F01B.57273052@tismer.com> <slrn8fjtao.2d4.quinn@drachma.ugcs.caltech.edu>
Message-ID: <38FA05C9.442868E4@tismer.com>


Quinn Dunkan wrote:
> 
> >> - Even so, Chris and I agree that continuations should eventually
> >>   be completely hidden from the user, since it's so incredibly
> >>   obscure, hardly anyone will understand it. Not good for CP4E ;-)
> >
> >Today I'm saying so. But I think there will come a time
> >where people are told "look, a continuation is as simple as 1+1",
> >and they'll say "sure? I thought a co is simpler."
> 
> I'm not sure I understand correctly, but does "completely hidden from the
> user" mean "only visible (in the "raw" form) from C"?  In that case, I think
> it should still be visible from python.

Right. One should perhaps not advertize continuations in the first
place, until people know them from school.
That's why we try to build really useful stuff from them, familiar
and understandable for many people, like threads.
This also means, after some stabilization phase and iterations,
most of these applications on top of co's will get optimized
back into C, dealing with the co stuff from there.
This doesn't mean to remove the co's Python interface,
but to improve and simplify it further.

> I don't see how it's different from metaclasses or the "new" module:  just
> put big red stickers on it.  It'd be nice to have around for people who want
> to experiment, write obfuscated code, or maybe even write a useful
> continuation-using extension.  I don't think there's much trouble with people
> abusing the Beaudry hook to explode anyone's head other than their own.

Thank you for reminding me that I have to write a real C interface
now.

ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaunstr. 26                  :    *Starship* http://starship.python.net
14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     where do you want to jump today?   http://www.stackless.com



From parker at gol.com  Fri Apr 21 21:10:01 2000
From: parker at gol.com (Ian Parker)
Date: Sat, 22 Apr 2000 01:10:01 GMT
Subject: editor for win32
References: <NpHL4.779$fK2.113609@news.flash.net>
 <QqYL4.1341$fK2.161623@news.flash.net>
Message-ID: <1vhEiDAdtPA5EwJ3@gol.com>

In article <QqYL4.1341$fK2.161623 at news.flash.net>, Scott Hathaway
<slhath at flash.net> writes
>Thanks for all the good input!
>
>I am new to python and excited about giving it a whirl.
>
>Scott
>
>"Scott Hathaway" <slhath at flash.net> wrote in message
>news:NpHL4.779$fK2.113609 at news.flash.net...
>> Can anyone recommend a python editor for Windows?
>>
>> If there is an IDE for windows (other than IDLE), even better.
>>
>> Thanks,
>> Scott
>>
>>
>
>

I use PFE - Programmer's File Editor.   It's simple, programmable and
with optional line numbers.

Regards

-- 
Ian Parker


From zorro at zipzap.ch  Sat Apr 15 04:24:50 2000
From: zorro at zipzap.ch (Boris Borcic)
Date: Sat, 15 Apr 2000 10:24:50 +0200
Subject: starship.python.net
References: <Pine.OSF.4.10.10004111721150.9584-100000@prfdec.natur.cuni.cz> <Pine.OSF.4.10.10004111740110.11294-100000@prfdec.natur.cuni.cz> <m3itxkg41s.fsf@atrus.jesus.cam.ac.uk> <38F763CA.43F8EA85@eq.uc.pt>
Message-ID: <38F82752.E7FBE4E6@zipzap.ch>

Pedro Vale Lima wrote:
> 
> I still can't connect.

Same here.

BB


From meh9 at cornell.edu  Thu Apr  6 14:24:00 2000
From: meh9 at cornell.edu (Matthew Hirsch)
Date: Thu, 06 Apr 2000 14:24:00 -0400
Subject: More about variables
References: <meh9-5C17A0.10445406042000@news.cit.cornell.edu> <3X3H4.16435$0o4.108310@iad-read.news.verio.net>
Message-ID: <meh9-158D1A.14240006042000@news.cit.cornell.edu>

Why create named variables?

I always thought that having 20 lists saved more space than having one 
list of 20 lists.  Maybe I'm wrong.

Thanks for your help,
Matt






In article <3X3H4.16435$0o4.108310 at iad-read.news.verio.net>, 
culliton at clark.net (Tom Culliton) wrote:

> What (or maybe what do you think) you are trying to do?  Most likely
> you would be better off with just a list or tuple if you're dealing
> with and unknown number of data items.  Not everything needs a name,
> and there isn't much difference between:
> 
> v0, v1, v2, v3, v4, v5
> 
> and
> 
> v[0], v[1], v[2], v[3], v[4], v[5]
> 
> Why do you think this variable length collection of data needs
> individual names?
> 
> Now, if after consideration there is still some compelling reason to
> generate names, you can use something like this with a dictionary
> (possibly the one returned by globals()):
> 
> v = {}
> for i in range(number_of_variables):
>         name = "variable%d" % i
>         v[name] = []
> 
> But again, Why?
> 
> In article <meh9-5C17A0.10445406042000 at news.cit.cornell.edu>,
> Matthew Hirsch  <meh9 at cornell.edu> wrote:
> >Hi All,
> >
> >Let's say I had something like:
> >
> >number_of_variables=int(raw_input('Enter number of variables: '))
> >
> >if number_of_variables==1:
> >   variable1=[]
> >elif number_of_variables==2:
> >   variable1=[]
> >   variable2=[]
> >elif number_of_variables==3:
> >   variable1=[]
> >   variable2=[]
> >   variable3=[]
> >elif number_of_variables==4:
> >   variable1=[]
> >   variable2=[]
> >   variable3=[]
> >   variable4=[]
> >elif number_of_variables==5:
> >   variable1=[]
> >   variable2=[]
> >   variable3=[]
> >   variable4=[]
> >   variable5=[]
> >else:
> >   print 'else'
> >
> >Is there a more efficient way of doing this? What if I wanted to create 
> >100 variables (lists).  It doesn't make sense to write an if statement 
> >for that many conditions.
> >
> >Thanks for your help,
> >Matt


From bingdo19 at earthlink.net  Sat Apr 29 11:25:03 2000
From: bingdo19 at earthlink.net (Bob Langdon)
Date: Sat, 29 Apr 2000 15:25:03 GMT
Subject: Need to Hire Freelance Python Expert
Message-ID: <bingdo19-D1D62A.11344229042000@news-server.twcny.rr.com>

Hi,

I've recently installed Python 1.5.2 in the home directopry on a virtual 
domain name. All seems to work well, but having trouble with a series of 
scripts moved from another server; e.g.: name errors and key errors.

Need to hire someone to trouble shoot the setup.

Call Bob Langdon at 800-982-0548 as soon as possible.

Will be in Sat & Sunday 10-4 eastern time.

Thanks.


From phd at phd.russ.ru  Thu Apr 27 08:56:04 2000
From: phd at phd.russ.ru (Oleg Broytmann)
Date: Thu, 27 Apr 2000 12:56:04 +0000 (GMT)
Subject: PyApache memory leak?
In-Reply-To: <Pine.LNX.4.21.0004270834490.27709-100000@saturn.ispol.com>
Message-ID: <Pine.LNX.4.21.0004271254330.11789-100000@fep132.fep.ru>

Hello!

On Thu, 27 Apr 2000, Gregory Trubetskoy wrote:
> On Fri, 21 Apr 2000, Oleg Broytmann wrote:
> 
> >    This don't help saving memory. Other way round - PyApache destroys
> > subinterpreter after processing a request, thus saving memory; but httpdapy
> > keeps the interpreter with all (unneccessary and superfluos) imported
> > modules, thus leaking memory.
> 
> I wouldn't call this "leaking". "Leaking" means forgetting to free memory
> where you should be freeing it (like overlooking a Py_DECREF), loosing a
> pointer to memory, etc. To the best of my knowledge, httpdapy 1.7b does
> not leak memory. I've done a lot of work to make sure of this.
> 
> Another note to consider - I am not 100% sure that Py_EndInterpreter()
> actually frees all memory. In fact I'm almost sure it doesn't, and you
> still need to watch your reference counts if you create object via the
> C-API. If that is the case, then PyApache has a higher likelyhood of
> actually leaking memory.

   May be...

> BTW, if I remember it correctly, the decision by Lele (the author of
> PyApache) to recreate the interpreter was based not on memory usage, but
> because the CGI wouldn't work as expected otherwise and for security
> reasons.

   I am pretty satisfied with PyApache. (Note: I didn't tried httpdapy yet,
so I cannot say bad words about it; I belive it's good product anyway).

Oleg.            (All opinions are mine and not of my employer)
---- 
    Oleg Broytmann      Foundation for Effective Policies      phd at phd.russ.ru
           Programmers don't die, they just GOSUB without RETURN.




From tim_one at email.msn.com  Tue Apr 11 00:09:26 2000
From: tim_one at email.msn.com (Tim Peters)
Date: Tue, 11 Apr 2000 00:09:26 -0400
Subject: Python paradigms
In-Reply-To: <slrn8f3paf.opc.grant@grante.comtrol.com>
Message-ID: <001b01bfa36b$bafc7a80$27a2143f@tim>

[Grant Edwards]
> ...
> In assembly comments are probably required, but in a language
> like Python, if code needs comments other than the occasionaly
> "doc" string, then it's written badly.

Depends on how hard the algorithm is.  I have, on occasion, and even in
Python, written two pages of comments to explain one line of code; also one
line of comment to explain two pages.  Not all algorithms are trivial!
Although more *should* be in most code I see <0.5 wink>.

The first compiler I worked on was written entirely in Cray assembly
language, for a RISC vector machine warped toward supernaturally fast 64-bit
floating-point arithmetic in all possible ways.  It was a nightmare in most
respects for what a compiler needs to do.  The original author gave all
functions 4-letter names, and arranged them in strict alphabetical order.
Each function used its own calling and return conventions, undocumented
except by reverse-engineering the code.  Over time you just *learned*, e.g.,
that function CPMX didn't touch register S5 and that callers relied on that,
while function CX00 expected its input in S5 and put its output in S3 (names
& register numbers disguised to protect the guilty).  Etc.  In many respects
it was a magnificent house of soggy cards balanced on grains of sand <wink>.

One early day I was tracking down a particularly nasty bug, scratching my
head in the emulator, stepping thru page after page of this uncommented
code.  Then, like a long-sought lighthouse in an oceanic storm, a single
comment appeared on a line of code in the middle of the page:

        S1   -3     ; because 15 didn't work

Oh -- of course <wink>.

bad-comments-are-worse-than-none-ly y'rs  - tim





From chambert at borel2.institut.math.jussieu.fr  Sat Apr 22 03:57:22 2000
From: chambert at borel2.institut.math.jussieu.fr (Antoine Chambert-Loir)
Date: 22 Apr 2000 09:57:22 +0200
Subject: Mac Python Questions
References: <IRjL4.271$CP3.3183@news.uswest.net> <956247497.216189@fezzik.endicor.com> <Pine.SO4.4.02.10004210015280.25675-100000@ub.d.umn.edu> <956347062.17103@fezzik.endicor.com>
Message-ID: <8drm12$v1v$1@borel2.institut.math.jussieu.fr>

In article <956347062.17103 at fezzik.endicor.com>,
Kent Polk <kent at goatnospamhill.org> wrote:

>I have no idea where things are actually *supposed* to be, but the
>installer placed _tkinter.CFM68k.slb in :Mac:PlugIns: and did not
>install the _tkinter.ppc.slb there also. From what I saw, you can
>either copy _tkinter.ppc.slb into :Mac:PlugIns: or add its location
>to the Path in order to be able to import _tkinter. We are copying
>to :Mac:PlugIns: as it appears that was the installer's intent.


In my distribution, I got 2 copies of _tkinter.ppc.slb
and 2 (or maybe even 3) of _tkinter.CFM68k.slb !
They look different. Maybe one of them is bugged ?

It would be great to have a full listing of the PYTHON directory
on a mac where it works well. (Plus some stats as, e.g. the
memory available, the advised memory, etc.)

Regards,



-- 
Antoine Chambert-Loir  o-o  Institut de Math?matiques de Jussieu (Chevaleret)
Tel: 01 44 27 75 20     |   Universit? Pierre et Marie Curie
Fax: 01 44 27 48 44     -   Boite 247, 4 place Jussieu, F-75252 Paris Cedex 05
http://www.math.jussieu.fr/~chambert


From emile at fenx.com  Thu Apr 13 13:14:25 2000
From: emile at fenx.com (Emile van Sebille)
Date: Thu, 13 Apr 2000 10:14:25 -0700
Subject: List mapping?
References: <38f5d68e@news.xtml.co.uk> <newscache$u6nysf$1l1$1@jenpc07.jennic.co.uk> <38f5e714@news.xtml.co.uk>
Message-ID: <010a01bfa56b$b93364a0$01ffffc0@worldnet.att.net>

While this does work, it is slower and harder
to see what's going on.

Emile van Sebille
emile at fenx.com
-------------------


----- Original Message -----
From: Nick Trout <nick at spam_me_notvideosystem.co.uk>
Newsgroups: comp.lang.python
To: <python-list at python.org>
Sent: Thursday, April 13, 2000 8:20 AM
Subject: Re: List mapping?


> > Try this:
> > mylist = ['/a','1','/b','2','/c','3']
> > map(lambda x: reduce(lambda a, b: (a,b), mylist[x:x+2]), range(0,
> > len(mylist), 2))
>
> Very good! I wasnt familiar with the reduce function. I would
certainly not have
> come up with something of this complexity.
>
> This begs the question: Is there an archive of code snippets for
Python? These
> could form an excellent tutorial. Quite a few messages arrive on this
newsgroup
> asking for good starting points for Python (and thats just the people
who come
> here and ask).
>
> N
>
>
> --
> http://www.python.org/mailman/listinfo/python-list





From effbot at telia.com  Thu Apr 13 15:04:18 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Thu, 13 Apr 2000 19:04:18 GMT
Subject: Simple formatting string question
References: <meh9-637743.14563813042000@news.cit.cornell.edu>
Message-ID: <SSoJ4.908$rc9.190465024@newsb.telia.net>

Matthew Hirsch <meh9 at cornell.edu> wrote:
> I'm trying to print something like this:
>
> >>> print 6*'%2i' '%1i' % tuple([5,5,5,5,5,5,1])
>
> but I'm getting an error.

how about:

>>> print (6*'%2i' + '%1i') % tuple([5,5,5,5,5,5,1])

</F>




From dieter at handshake.de  Sat Apr  1 03:22:34 2000
From: dieter at handshake.de (Dieter Maurer)
Date: 01 Apr 2000 10:22:34 +0200
Subject: pdb will not continue
References: <Pine.GSO.4.10.10003301115570.22353-100000@xmission.xmission.com>
Message-ID: <x7n1netelx.fsf@lindm.dm>

Tim Egbert <tegbert at xmission.xmission.com> writes on Thu, 30 Mar 2000 11:30:17 -0700:
> I'm new to the pbd module, but think I've got most of it working. The
> problem I'm having, however, is that even though I can set break points, I
> can't get pdb to 'continue' ('cont' or 'c') to the break point.  All it
> will do is advance one line.  What am I doing wrong?
This is a bug in PDB:
  Pdb.runcall cannot continue in the top level function
  (it can inside a called function).

You can use "Pdb.run" instead of "Pdb.runcall" (it takes a string
with the function call rather than the function and the arguments).

Dieter.


From chapman at bioreason.com  Wed Apr  5 11:10:52 2000
From: chapman at bioreason.com (Mitch Chapman)
Date: Wed, 05 Apr 2000 09:10:52 -0600
Subject: opening a serial port at a given baudrate. How?
References: <200004050319.NAA08789@piglet.dstc.edu.au>
Message-ID: <38EB577C.E1EC5704@bioreason.com>


David Arnold wrote:
> 
> -->"Jim" == Jim Richardson <warlock at eskimo.com> writes:
> 
>   Jim> I have been using open('/dev/ttyS0','r+') but need to control
>   Jim> baud rate. fcntl.ioctl() will do this?
> 
> as someone else pointed out (sorry, lost the mail), the termios module
> is the go for setting baud rates, parity, etc.

Here's an old answer, courtesy  of www.deja.com:
http://www.deja.com/[ST_rn=ps]/getdoc.xp?AN=309748544&fmt=text

Search for "# Set the baud rate."

--
Mitch Chapman
chapman at bioreason.com


From dalke at acm.org  Fri Apr  7 09:27:26 2000
From: dalke at acm.org (Andrew Dalke)
Date: Fri, 7 Apr 2000 07:27:26 -0600
Subject: "tostring" as a string method?
References: <8c45ue$b7u$1@slb1.atl.mindspring.net> <Lp9H4.3526$y4.52453@newsread1.prod.itd.earthlink.net> <8ck7kq$1cu$1@slb3.atl.mindspring.net> <f4lH4.4346$HG1.126863@nnrp1.uunet.ca>
Message-ID: <8cknio$1v8$1@slb1.atl.mindspring.net>

Warren Postma asked:
>Hmm. Have you considered marshalling everything instead?


I think I didn't explain the context well enough.  The result
of this string (either from __str__, tostring() or whichever)
will be sent to one of the string.* or re.* functions.  It
really needs the string representation of the underlying
sequence (assuming it can be encoded into a single-letter
alphabet).

Eg, if
  seq = Seq("GGATACTG", Alphabet.dna)
then
  string.count(seq.tostring(), "G")
works.  As would
  seq = MutableSeq(array.array("c", "GGATACTG"))

(or as /F pointed out, so does string.count(buffer(seq.data), "G"))

Marshall would be used to get a string encoded representation
of the object which isn't (easily) usable by, say, a regular
expression.


                    Andrew
                    dalke at acm.org





From lewst at yahoo.com  Wed Apr  5 00:52:27 2000
From: lewst at yahoo.com (lewst)
Date: Tue, 4 Apr 2000 21:52:27 -0700 (PDT)
Subject: Starving for an Advanced Python Book
Message-ID: <20000405045227.29911.qmail@web2105.mail.yahoo.com>

Lee Joramo <ljoramo.list at nickads.com> wrote:

> 'Python Essential Reference'; David Beazley; New Riders 
> An excellent reworking of the online documentation. This is a
> reference book with out tutorial, but I have learned much from
> it. HIGHLY RECOMMENDED

I also decided to buy this book tonight after having a chance to look
at it.  This will be much more convenient than looking up things in
the online library reference manual which I have been doing thus far.


__________________________________________________
Do You Yahoo!?
Talk to your friends online with Yahoo! Messenger.
http://im.yahoo.com



From guido at python.org  Sat Apr  1 12:00:00 2000
From: guido at python.org (Guido van Rossum)
Date: Sat, 1 Apr 2000 12:00:00 -0500 EST
Subject: New Features in Python 1.6
Message-ID: <200004011740.MAA04675@eric.cnri.reston.va.us>

New Features in Python 1.6
==========================

With the recent release of Python 1.6 alpha 1, a lot of people have
been wondering what's new.  This short note aims to explain the major
changes in Python 1.6.


Core Language
-------------

1. Unicode strings

Python strings can now be stored as Unicode strings.  To make it easier
to type Unicode strings, the single-quote character defaults to creating
a Unicode string, while the double-quote character defaults to ASCII
strings.  If you need to create a Unicode string with double quotes,
just preface it with the letter "u"; likewise, an ASCII string can be
created by prefacing single quotes with the letter "a".  For example:

   foo = 'hello'                # Unicode
   foo = "hello"                # ASCII
   foo = a'hello'               # ASCII
   foo = u"hello"               # Unicode

You can still use the "r" character to quote strings in a manner
convenient for the regular expression engine, but with subtle changes in
semantics: see "New regular expression engine" below for more
information.

Also, for compatibility with most editors and operating systems, 
Python source code is still 7-bit ASCII.  Thus, for portability it's
best to write Unicode strings using one of two new escapes: \u and \N.
\u lets you specify a Unicode character as a 16-bit hexadecimal number,
and \N lets you specify it by name:

    message = 'Bienvenue \N{LATIN SMALL LETTER A WITH GRAVE} ' \ +
              'Python fran\N{LATIN SMALL LETTER C WITH CEDILLA}ais!'
    message = 'Bienvenue \u00E0 Python fran\u00E7ais!'


2. string methods

Python strings have grown methods, just like lists and dictionaries.
For instance, to split a string on spaces, you can now say this:

    tokens = "foo bar baz".split(" ")

Or, equivalently, this:

    tokens = " ".split("foo bar baz")

(Python figures out which string is the delimiter and which is the
string to split by examining both strings to see which one occurs more
frequently inside the other.)

Be careful not to mix Unicode and ASCII strings when doing this, though.

Other examples:

    foo = "The quick red fox jumped over the lazy brown dog."
    foo.find("dog")
    foo.strip()    
    foo.lower()

Note that use of any string method on a particular string renders it
mutable.  This is for consistency with lists, which are mutable and have
methods like 'append()' and 'sort()' that modify the list.  Thus,
"foo.strip()" modifies the string 'foo' in-place.  "strip(foo)" retains
its old behavior of returning a modified copy of 'foo'.


3. extended call syntax

The variable argument list and keyword argument syntax introduced in
Python 1.3 has been extended.  Previously, it only worked in
function/method signatures; calling other functions with the same
arguments required the use of 'apply()'

    def spam(arg1,arg2,*more_args,**keyword_args):
        # ...
        apply(foo,(arg1,arg2) + more_args,keyword_args)

Now it works for calling functions too.  For consistency with C and C++,
asterisks in the function signature become ampersands in the function
body:

        foo(arg1,arg2,&more_args,&&keyword_args)


4. assignment to None now works

In previous version of Python, values assigned to None were lost.  For
example, this code:

    (username,None,None,None,realname,homedir,None) = getpwuid(uid)

would only preserve the user name, real name, and home directory fields
from a password file entry -- everything else of interest was lost.

In Python 1.6, you can meaningfully assign to None.  In the above
example, None would be replaced by a tuple containing the four values of
interest.

You can also use the variable argument list syntax here, for example:

    (username,password,uid,uid,*None) = getpwuid(uid)

would set None to a tuple containing the last three elements of the
tuple returned by getpwuid.


Library
-------

1. Distutils

In the past, lots of people have complained about the lack of a standard
mechanism for distributing and installing Python modules.  This has been
fixed by the Distutils, or Distribution Utilities.  We took the approach
of leveraging past efforts in this area rather than reinventing a number
of perfectly good wheels.

Thus, the Distutils take advantage of a number of "best-of-breed" tools
for distributing, configuring, building, and installing software.  The
core of the system is a set of m4 macros that augment the standard
macros supplied by GNU Autoconf.  Where the Autoconf macros generate
shell code that becomes a configure script, the Distutils macros
generate Python code that creates a Makefile.  (This is a similar idea
to Perl's MakeMaker system, but of course this Makefile builds Python
modules and extensions!)

Using the Distutils is easy: you write a script called "setup.in" which
contains both Autoconf and Distutils m4 macros; the Autoconf macros are
used to create a "configure" script which examines the target system to
find out how to build your extensions there, and the Distutils macros
create a "setup.py" script, which generates a Makefile that knows how to
build your particular collection of modules.  You process "setup.in"
before distributing your modules, and bundle the resulting "configure"
and "setup.py" with your modules.  Then, the user just has to run
"configure", "setup.py", and "make" to build everything.

For example, here's a small, simple "setup.in" for a hypothetical module
distribution that uses Autoconf to check for a C library "frob" and
builds a Python extension called "_frob" and a pure Python module
"frob":

    AC_INIT(frobmodule.c)
    AC_CHECK_HEADER(frob.h)
    AC_HAVE_LIBRARY(frob)
    AC_OUTPUT()

    DU_INIT(Frob,1.0)
    DU_EXTENSION(_frob,frobmodule.c,-lfrob)
    DU_MODULE(frob,frob.py)
    DU_OUTPUT(setup.py)

First, you run this setup.in using the "prepare_dist" script; this
creates "configure" and "setup.py":

    % prepare_dist

Next, you configure the package and create a makefile:

    % ./configure
    % ./setup.py

Finally, to create a source distribution, use the "sdist" target of the
generated Makefile:

    % make sdist

This creates Frob-1.0.tar.gz, which you can then share with the world.
A user who wishes to install your extension would download
Frob-1.0.tar.gz and create local, custom versions of the "configure" and
"setup.py" scripts:

    % gunzip -c Frob-1.0.tar.gz | tar xf -
    % cd Frob-1.0
    % ./configure
    % ./setup.py

Then, she can build and install your modules:

    % make
    % make install

Hopefully this will foster even more code sharing in the Python
community, and prevent unneeded duplication of effort by module
developers.

Note that the Python installer for Windows now installs GNU m4, the bash
shell, and Autoconf, so that Windows users will be able to use the
Distutils just like on Unix.


2. Imputils

Complementary to the Distutils are the Imputils, or Import Utilities.
Python's import mechanism has been reworked to make it easy for Python
programmers to put "hooks" into the code that finds and loads modules.
The default import mechanism now includes hooks, written in Python, to
load modules via HTTP from a known URL.

This has allowed us to drop most of the standard library from the
distribution.  Now, for example, when you import a less-commonly-needed
module from the standard library, Python fetches the code for you.  For
example, if you say

    import tokenize

then Python -- via the Imputils -- will fetch
http://modules.python.org/lib/tokenize.py for you and install it on your
system for future use.  (This is why the Python interpreter is now
installed as a setuid binary under Unix -- if you turn off this bit, you 
will be unable to load modules from the standard library!)

If you try to import a module that's not part of the standard library,
then the Imputils will find out -- again from modules.python.org --
where it can find this module.  It then downloads the entire relevant
module distribution, and uses the Distutils to build and install it on
your system.  It then loads the module you requested.  Simplicity
itself!


3. New regular expression engine

Python 1.6 includes a new regular expression engine, accessed through
the "sre" module, to support Unicode strings.  Be sure to use the *old*
engine for ASCII strings, though:

    import re, sre
    # ...
    re.match(r"(\d+)", "The number is 42.")   # ASCII
    sre.match(r'(\d+)', 'The number is \N{SUPERSCRIPT TWO}')  # Unicode

If you're not sure whether a string is ASCII or Unicode, you can always
determine this at runtime:

    from types import *
    # ...
    if type(s) is StringType:
        m = re.match(r"...", s)
    elif type(s) is UnicodeType:
        m = sre.match(r'...', s)



From jamarijr at hotmail.com  Mon Apr 17 12:22:05 2000
From: jamarijr at hotmail.com (Arinté)
Date: Mon, 17 Apr 2000 16:22:05 GMT
Subject: Python flak???
Message-ID: <8dfdmq$v1o$1@nnrp1.deja.com>

What are the reprecussions <sp?> of doing a PyErr_Clear() when there is
no error?


Sent via Deja.com http://www.deja.com/
Before you buy.


From pony at nmt.edu  Mon Apr 10 14:24:23 2000
From: pony at nmt.edu (Enrique Herrera)
Date: Mon, 10 Apr 2000 12:24:23 -0600
Subject: Help on Big Python Program
References: <38E8DB50.A79CB88@nmt.edu>
Message-ID: <38F21C57.F6DD2CFA@nmt.edu>

I am not going to relse this software to the public. Is for my
department. I just need to find a way to combind all my modeles and the
main program into one bin exec. Any Ideas?


From darrell at dorb.com  Sun Apr  2 15:48:59 2000
From: darrell at dorb.com (Darrell)
Date: Sun, 2 Apr 2000 12:48:59 -0700
Subject: Debugging confusion -- too many stacks!
References: <000a01bf9c6b$3f80c2a0$752d153f@tim> <38E739A1.45F4B199@tismer.com> <030301bf9cd9$eb6786d0$6401a8c0@dorb> <38E77676.4599978B@tismer.com>
Message-ID: <035a01bf9cdc$7d433d40$6401a8c0@dorb>

"Christian Tismer"
> Oops, sorry! I confused PDB with GDB.
> Probably since the issue was about stacks...
>
Oh...
Still when it comes to walking the stack and inspecting things PDB is hard
to beat. The trick being how to best use it with the likes of GDB.

--Darrell





From JamesL at Lugoj.Com  Sat Apr  1 13:57:58 2000
From: JamesL at Lugoj.Com (James Logajan)
Date: Sat, 01 Apr 2000 10:57:58 -0800
Subject: Python 1.6 alpha 1 released
References: <200003312130.QAA04361@eric.cnri.reston.va.us> <38E5205F.DE811F61@roguewave.com> <Wf9F4.4423$74.70231@newsc.telia.net> <38E58D2D.384F1546@Lugoj.Com> <S9kF4.4472$74.71414@newsc.telia.net>
Message-ID: <38E646B6.12011A4E@Lugoj.Com>

Fredrik Lundh wrote:
> 
> James Logajan <JamesL at Lugoj.Com> wrote:
> > Also, how good are you at math? You seem to feel that 5 minutes of "Guido
> > the god's" time is of much more value than that of 1000 knuckle-dragging
> > Neanderthal programmers like myself spending a mere 10 minutes each to
> make
> > sure their code doesn't break. Of course, if we happen to take more than
> 10
> > minutes its our own damn fault I'm sure you'll agree.
> 
> the strange thing here is that there are changes listed on
> that page that are far more difficult to spot and fix, but for
> some reason, neither you or bjorn seem to care.

To be honest, I have so little time these days that I only spot-check this
newsgroup and have not even checked into what is happening with Python's
latest version. The only reason I even read Bjorn's post is because I saw
Guido's announcement and read it and started reading some of the followups.
While Bjorn came across as a bit irritating I had to reluctantly agree that
he had a valid point: one should not change API signatures unless there is
something concretely wrong with them. The benefit/pain ratio for by this
change is close to zero.

> others have politely asked for backwards compatibility on
> this one (which would mean *adding* extra code to handle
> multiple calling syntaxes).  someone will probably come up
> with a patch well before the next alpha, but probably not
> because you guys are flaming away on this newsgroup.

Well then I hope it gets changed due to those polite requests. Hopefully
that old extra code can be patched back in.

> 
> besides, the notion that 10 minutes spent on fixing this will
> outweigh all others costs involved in upgrading to a new
> release of the core interpreter strikes me as rather odd...
> 
> (heck, both of you have probably spent far more than ten
> minutes each writing your posts...)

I'm not sure if you realize that the reference to ten minutes was an
absurdism. If 1.6 includes a number of function signature changes or
semantic changes to existing functionality and not just bug fixes and benign
additions then I know of at least two firms that will be unwilling to use
1.6 or later versions. In fact I may be in the unfortunate position of
getting calls on it even though I am no longer consulting to them. I had
told them that as far as I knew code that worked for version 1.N should in
*general* work for version 1.(N+1) but that all bets were off for N.X to
(N+1).X. It was how I managed to sell Python into them. Ouch; guess I shoud
have read the fine-print.

> 
> > Now how did I get so pissed over this exchange of posts?
> 
> now how come I find your "guido the god" argument to be
> directly offensive?

Because it was meant to be. Glad to see it worked! The motivation for it was
because even though Bjorn directed his post to Guido, you and one or two
others responded. Almost like high priests guarding the temple. Any reason
Guido doesn't post to this newsgroup like he used to?

P.S. I'm not pissed anymore about it. Since it is clear I can't influence
things I have calmed down and will move on to things that I can influence
(to stay the same) or that no one can influence. The pragmatist in me says I
should look for a language that abides by a standard that can not be easily
changed.


From ullrich at math.okstate.edu  Sun Apr 23 14:14:13 2000
From: ullrich at math.okstate.edu (David C. Ullrich)
Date: Sun, 23 Apr 2000 18:14:13 GMT
Subject: LAmbda calculus evaluator (interactive)
References: <000301bfacdf$1f5c7260$212d153f@tim>
Message-ID: <39033c95.437294824@news>

On Sun, 23 Apr 2000 00:48:08 -0400, "Tim Peters"
<tim_one at email.msn.com> wrote:

>[Kragen Sitaker]
>> ...
>> Many years ago, Colby Kraybill suggested to me that it would be nice
>> to have something similar for mathematical expressions --- something
>> that would do nothing on its own, but would do the grunt-work involved
>> in an action like "take the reciprocal of both sides" or "divide both
>> the top and bottom by (x-1)".
>
>Hyland Software's "Merlyn" program does that; see
>
>    http://www.hylandsoftware.com/
>
>It's based on a general term-rewriting system, and is programmable by the
>end-user (indeed, almost all the builtin rules are exposed in files you can
>edit too).  However, it's proprietary, costs money, source for the engine
>itself is not available, and it works only on Windows.  I often use it
>instead of, e.g., Macsyma, because the high-powered systems drive me nuts
>with the pervasive & often unhelpful transformations they make in an
>eyeblink; Merlyn never does anything unless it's told to.
>
>Similar but weaker symbolic facilities can be found, e.g., in Dave
>Gillespie's Emacs calc, and high-end HP calculators.
>
>If the world were sane, they would have written all these things in Python
><wink>.

	So if they'd written them in Python what would the source
code look like? 

>the-temptation-to-use-the-time-machine-is-often-overwhelming-ly y'rs  - tim

	So you have a time machine - so "too busy" is no excuse...

DU


From ljoramo.list at nickads.com  Mon Apr  3 18:19:59 2000
From: ljoramo.list at nickads.com (Lee Joramo)
Date: Mon, 3 Apr 2000 16:19:59 -0600
Subject: PyGreSQL errors
Message-ID: <v04210108b50ec6c8e9bf@[192.168.1.2]>

I am trying to get pygresql to work. I have a Cobalt RaQ3 server which runs
a slightly modified version of RedHat. The system came with PostgreSQL
pre-installed. When I import the pygresql's basics.py tutorial at pythons
command line, I follow the instructions to create a connection and get an
error:


>>> cnx = basics.DB('mytest')
Traceback (innermost last):
  File "<stdin>", line 1, in ?
  File "/usr/lib/python1.5/pg.py", line 34, in __init__
    self.db = apply(connect, args, kw)
pg.error: FATAL 1:  Database mytest does not exist in pg_database



What has gone wrong?

--
Lee A. Joramo                      ljoramo at nickads.com
The Nickel Want Ads                www.nickads.com
Internet Manager                   970-242-5555

--

Onward and Upward!

Lee A. Joramo }}-----> Cowboy on the Electronic Frontier

lee at joramo.com

Alamar Ranch  -  Palisade  -  Colorado

ThreeRivers|Eagle|UCD|D&D|TNC|DX|Mac|221B|Cannondale|Deutsch




From pete at bluesq2.com  Thu Apr  6 06:40:00 2000
From: pete at bluesq2.com (David Grove)
Date: Thu, 06 Apr 2000 10:40:00 GMT
Subject: ANNOUNCE: CodeMagicCD Version 1.0
Message-ID: <38ec6964.14922499@news.rmi.net>

ANNOUNCE: CodeMagicCD Version 1.0

New York based software developer, The Blue Square Group,
announces the release of CodeMagicCD Ver 1.0., a three-level
toolkit for professional programmers. This software is now
available at www.CodeMagicCD.com.

CodeMagicCD, The Complete Programmer's Toolkit For Win32,
CDRom Version 1.0 includes:


***CODEMAGIC UNIVERSAL IDE***
This high-quality, multi-language, professional programmer's
editor proved itself by being used by more than 12,000 users
before coming out of its beta stage. It has a thriving
community of users, its own mailing lists, and the best
programming user interface of any editor currently on the
market.

Now in version 1.0, this editor supports over 12 languages,
and more are being added regularly.

Among the features of this editor, you will find:

* User configurable syntax highlighting for multiple languages
* Extremely configurable and flexible editing environment
* A pure build of core Perl 5.005_03, with more than a hundred
  of the most popular perl modules built in (yes, including Tk,
  GD, DBD/DBI, Win32::ODBC, and DB_File). Although CodeMagic can
  be used with other versions of Win32-based perl, this standard
  build is used because it's embedding and extension mechanisms
  are not corrupted or C++ based, and can therefore be used as
  a macro language and API extension language for the editor
  itself, or from any other language that supports DLL linkage.
* Code templates for lightning fast coding
* Per-user configuration, for multi-user machines
* The ability to work with any text file, and these languages:
  Perl, Python, C, C++, Java, JavaScript, Html, Html with
  embedded JavaScript/VBScript/PerlScript/PML, ASP, VBScript,
  BASIC, Visual Basic, Pascal (Delphi), REBOL, Tcl and TclTk,
  SQL, Lisp, PML (PerlMagic/Lightning or the "Perl Markup
  Language"), and DOS/4DOS/NT Batch Files
* Revision control
* FTP communication, excellent for telecommuters, web program-
  mers, and remote projects
* Project management
* Online customer support

(Guido, et. al.) NOTE: Python syntax highlighting is now fully
supported in this version, as promised. As of the first service
pack, Python will be embedded into CodeMagic as its primary API
and macro facility using PyDelphi. Also please note that the
CodeMagicCD Environment is _NOT_ a "professional" version of the
CodeMagic editor, which remains free and fully functional. The
CodeMagicCD Environment is a full working environment that
contains the CodeMagic editor.


***PACKAGED SOFTWARE***
Included on this CD is the software that makes Microsoft run
and hide. Free tools, freely distributed, hand picked from
around the world. These tools allow for a 100% complete
programming environment that VisualStudio could never beat.

Any kind of programming currently in existence is included,
from basic learning to web programming to advanced COM and
device driver development. (Microsoft doesn't want you to
know that these tools are available. They have nothing that
can compete.) These tools include:

* Perl 5.005_03, PerlMagic port. (Perl 5.6 is not stable at
  this time.) This port of perl contains only the true Win32
  port of Perl. This means that embedding and extending are
  not corrupted, as they are in other Win32 ports of Perl.
  As the hands down industry standard language for web and
  administration programming, and as the most powerful script
  programming and text programming language in existence
  today... not to mention the fastest, Perl is a "must have"
  for everyone, even if it's to use it as a super-advanced
  command line tool.
* TclTk, the original "popular" scripting language, and the
  first scripting language (that we know of) to put GUI
  programming in reach of entry-level programmers.
* Python, the up-and-coming scripting language preferred
  by its several-hundred-thousand-member community.
* A large set of GNU programming tools, such as grep, diff,
  patch, awk, sed, merge, and many others required for full-
  featured programming.
* The REBOL language, an extremely high-level language that
  allows you to send an email in one line of code. Of course,
  this is just an example. If it can be done on the internet,
  REBOL can do it in the blink of an eye. No complicated
  sockets or protocol programming to get in the way.
* The wxWindows GUI toolkit for C/C++ brings simple, object-
  oriented Windows programs within easy reach of the average
  Joe or Josephine. No longer do poor-quality COM and ActiveX
  components rule the Windows GUI world.
* The industry standard GNU C++ complier for Win32 (MinGWin),
  with its accompanying Assembler, Fortran77 compiler, and C
  and Objective C Compliers and tools. (Who said programming
  in C took a $1000 set of do-little wizards? Oh, that's right,
  Microsoft does!)
* The DJGPP DOC C compiler used for many years by C programmers
  around the world.
* Cint, C Interpreter powerful enough to run itself running a
  program written in C. (Yes, you read that right, it is powerful
  enough to run its own source code, and have that running code
  run another program.) There are times when the code-compile-
  link-run steps need simplification. Just write it and run
  it with this tool.
* YaBasic, proof that BASIC is not a Microsoft commodity.
* Other tools and languages, more of which will be added until
  no language is unavailable to our users, and people begin to
  realize that they really do have better choices than the
  supposed de-facto standards dictated by Redmond.


***THE PACKAGE***
All of this comes in a tidy package to tie everything together.
New programmers can use just one language, and not have to worry
about any of the others (but can graduate to them when the
need arises). Hackers (the guys who wear the propeller hats)
can enjoy learning and using everything in the most complete
programming package available anywhere for the Win32 platform.

One of the most pervasive problems with getting these superior
tools into the public workplace is the lack of wizards, gadgets,
buttons, gizmos, and little spinning icons. In other words,
management has always liked these languages, but couldn't find
themselves accepting them, because of the lack of any congruity
between them, or of any interface for them. This package
provides that interface once and for all.


***SUPPORT***
CodeMagicCD includes free installation support, per-incident
phone support, free email support, and mailing list support.

The languages contained in the package collectively have over
100 methods of free and commercial support. Nothing at Microsoft
can top that.




From tom-main at REMOVEME.home.com  Mon Apr 24 19:57:15 2000
From: tom-main at REMOVEME.home.com (Tom)
Date: Mon, 24 Apr 2000 23:57:15 GMT
Subject: OOP books that compare - what about Python?
Message-ID: <vb5N4.247204$Hq3.6239116@news2.rdc1.on.home.com>

I've been browsing two books that compare and/or teach OO languages: An
Intro to OOP by Timothy Budd, and Objects Unencapsulated by Ian Joyner.
Both seem good, both cover C++ and Java, but neither consider Python.

Any recommandations?

Thanks,
Tom




From glhansen at steel.ucs.indiana.edu  Wed Apr  5 13:39:08 2000
From: glhansen at steel.ucs.indiana.edu (Gregory L. Hansen)
Date: 5 Apr 2000 17:39:08 GMT
Subject: MacPython and Tkinter
Message-ID: <8cftns$udt$1@flotsam.uits.indiana.edu>

I've just installed Python 1.5.2 for Macintosh, and it seems to work well
enough except that none of the Tkinter demos work.  When I try to run a
demo script like Guido's "brownian.py", I get "ImportError: No module
named Tkinter".  The ConfigurePython app throws up a dialog "Locating
plugin modules..." that disappears so fast I hadto run it many times to
figure out what it said.

What could be the problem?  Do I need to download the Tkinter module
seperately from somewhere?
-- 
"Ballpark nachos -- you don't buy them, you just rent them." -- Hank Hill



From debl.nospamm at world.std.com  Sun Apr 30 21:31:18 2000
From: debl.nospamm at world.std.com (David Lees)
Date: Mon, 01 May 2000 01:31:18 GMT
Subject: COM question
Message-ID: <390CDC8B.779DB0C3@world.std.com>

I am working examples in Hammond book on Python.  How do I get a list of
names of COM objects?  I am on page 67 and and to know how to figure out
what is the name for a Microsoft Word object.  In the example it is
referred to as:
"Word.Application", but this gives a name error.  I am doing the client
call correctly, because when I substitute "Excel.Application" or
"Access.Application"  I have not problem opening and closing Excel or
Access.  Apparently Word must have a different name for COM stuff on my
system, so I thought that if someone can tell me how to look at the list
of COM objects the name will be obvious.  I have Norton Utility under
Win95 OSR2, so I can easily look at the Registry (though as a Win95
newbie, much of it is not intelligible)

Thanks in advance.

David Lees


From tim_one at email.msn.com  Tue Apr 18 02:43:14 2000
From: tim_one at email.msn.com (Tim Peters)
Date: Tue, 18 Apr 2000 02:43:14 -0400
Subject: Another question
In-Reply-To: <m3ln2cckg6.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <000201bfa901$5f32f620$642d153f@tim>

[Michael Hudson]
> ...
> For that last dreg of performance you might try:
>
> import random
> def do_it(randint=random.randint,xrange=xrange):
>     r = [0,] * 100000
>     for i in xrange(100000):
>        r[i] = randint(1,10)
> do_it()

Or even make the guts:

    r = range(100000)
    for i in r:
        r[i] = randint(1, 10)

> But really, this is getting silly; if performance is this much of an
> obsession you aren't going to have that much fun with Python.

To the contrary, speeding a language that fights you each step of the way is
rather akin to seduction -- what could be more fun than that <wink>?

python-is-the-language-of-lovers-ly y'rs  - tim





From mwh21 at cam.ac.uk  Fri Apr 28 09:05:57 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 28 Apr 2000 14:05:57 +0100
Subject: Do I always have to write "self." ?
References: <270420001706386648%pecora@anvil.nrl.navy.mil> <8eacdf$nnm$1@newshost.accu.uu.nl> <280420000751353947%pecora@anvil.nrl.navy.mil>
Message-ID: <m3wvlinznu.fsf@atrus.jesus.cam.ac.uk>

"Louis M. Pecora" <pecora at anvil.nrl.navy.mil> writes:

> In the above I would describe things as follows (am I right or wrong?):
> 
> (1) data and self.data point to the same object, a list, so when a
> method is called it is the method that works on that instance of a list
> and either data or self.data will append 'foo' to the same list object.
> 
> (2) when you do more=more+1 an new object (the more+1 value) is created
> and more then points to it and no longer to the old value of self.more. 
> The latter still exists and points to the original 10 object.
> 
> (3) Similar story here for interesting where a new object
> interesting[10:] is created and interesting is "switched" to point to
> it.  self.interesting still points tohe "foo" object.
> 
> I hope I got that right, because they have tripped me up a few times.

Yes!  That's right.  Simple isn't it (once you understand, anyway)?

Cheers,
M.

-- 
  If i don't understand lisp,  it would be wise to not bray about
  how lisp is stupid or otherwise criticize, because my stupidity 
  would be archived and open for all in the know to see.
                                               -- Xah, comp.lang.lisp


From a.d.stribblehill at dur.ac.uk  Fri Apr  7 07:10:07 2000
From: a.d.stribblehill at dur.ac.uk (Andrew Stribblehill)
Date: 07 Apr 2000 12:10:07 +0100
Subject: CGI extension problem
References: <UX3UiBAPET54Ewi5@jessikat.demon.co.uk>
Message-ID: <87ya6q5fqo.fsf@womble.dur.ac.uk>

Robin Becker <robin at jessikat.demon.co.uk> writes:

> I don't have root privilege, my python cgi script needs access to an extension which I have
> built; the extension in turn needs a shared lib which I have built. I install in a directory say
> $HOME/lib. When I run the cgi script I need to set the LD_LIBRARY_PATH to at least include
> $HOME/lib. I have use the following idiom
> 
> import sys, os
> HOME_LIB='/home/me/lib'
> if sys.platform != 'win32' and not os.environ.has_key('LD_LIBRARY_PATH'):
>         os.environ['LD_LIBRARY_PATH'] = HOME_LIB
>         os.execve(sys.argv[0],sys.argv,os.environ)
> 
> # rest of cgi script
> .....
> 
> I can think of a few ways in which this might break, but is there anything obviously easier less
> breakable?

Wouldn't it be better to use LD_RUN_PATH when /compiling/ the
extension? Then it won't need a LD_LIBRARY_PATH variable at
runtime.

Obviously you then have to make sure that this location stays
the same else you'll need to resort to the LD_LIBRARY_PATH hack (or
recompile) again!

Cheerio,

Andrew Stribblehill
Systems Programmer, IT Service, University of Durham, England


From effbot at telia.com  Wed Apr 12 10:39:33 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Wed, 12 Apr 2000 14:39:33 GMT
Subject: Compiling Python Imaging under LinuxPPC
References: <20000412234630.21095@nms.otc.telstra.com.au> <20000412102120.18661.00002626@ng-cg1.aol.com>
Message-ID: <FU%I4.646$rc9.190630912@newsb.telia.net>

William F. Adams wrote:
> Greg McFarlane was kind enough to clue me in to BLT being:
> >A graphics extension to Tk:
> >    http://www.tcltk.com/blt/
>
> I've got that, but still having trouble with python imaging, now because
> imaging.h isn't appropriately installed---I found a message on the Sketch
list
> which had a patch, but that won't work on my LinuxPPC install, reports it
as
> being garbage....
>
> Can I just d/l Imaging.h manually from somewhere or copy it in to
> /usr/include/python1.5 by hand?

eh?  Imaging.h is part of the PIL distribution (it's in the
libImaging directory) and the standard Setup file already
points to it.

as long as you follow the instructions in the README, you
shouldn't really have to move it around...

</F>




From pecora at anvil.nrl.navy.mil  Fri Apr 14 07:17:19 2000
From: pecora at anvil.nrl.navy.mil (Louis M. Pecora)
Date: Fri, 14 Apr 2000 11:17:19 GMT
Subject: Need full example of C extension on Macintosh (CW)
References: <120420000714412887%pecora@anvil.nrl.navy.mil> <38F62C97.FD5073B@spacenet.tn.cornell.edu>
Message-ID: <140420000717194110%pecora@anvil.nrl.navy.mil>

In article <38F62C97.FD5073B at spacenet.tn.cornell.edu>, Tom Loredo
<loredo at spacenet.tn.cornell.edu> wrote:


Tom,

I got a small extension to work (! succeess).  I have written it up and
will pass it by the Mac-Python crowd to see if there are any mistakes
(I am still unsure of some rules about naming things), but I think I
have the general scheme down.  I am using CW3 (plan to upgrade in a few
months).  I will send you the text file of my write up separately. 
It's not ready for prime time.  Let me know what you think and if there
are differences for CW5.


From javanet at dynacap.com  Wed Apr 12 17:34:58 2000
From: javanet at dynacap.com (Arnaldo Riquelme)
Date: Wed, 12 Apr 2000 13:34:58 -0800
Subject: Python 4 DOS
References: <sf7hsua14nk34@corp.supernews.com> <38F48C6D.FD671082@interet.com> <sf99mtkv4nk80@corp.supernews.com> <Pine.SO4.4.02.10004121254400.29583-100000@ub.d.umn.edu>
Message-ID: <sf9qqtt0aln46@corp.supernews.com>

> Did you try running it *with* the DOS extender?  I have no experience
> using Python on DOS with the extender, but I've used other prorgams which
> required it (DooM, DJGPP [I think]), and they won't work without it.
> GO32.exe, or something like that?
>
>
> Aaron
> ----

Thanks Aaron, I looked into DJGPP and it all makes sense now.

Arnaldo

Python rules!!




From effbot at telia.com  Thu Apr 13 15:17:45 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Thu, 13 Apr 2000 19:17:45 GMT
Subject: 1st non-trivial program - gentle criticism requested
References: <1e90r16.1sd2ts41e7ur6kN@gershwin.bio.ic.ac.uk> <8d4tgu$11qa$1@nntp6.u.washington.edu> <38F6193A.66284778@gte.com>
Message-ID: <t3pJ4.913$rc9.192648192@newsb.telia.net>

Gary D. Duzan <gduzan at gte.com> wrote:
> This potential stumbling block makes me wonder if allowing the
> singleton case was really wise.

stumbling block?

what kind of programmer changes his code without thinking,
reading what he just typed, or running the resulting program?

if you want to optimize python for that kind of programmer,
you have to change a whole lot more than the string formatting
operator...

</F>




From johnm at magnet.com  Thu Apr 20 13:12:45 2000
From: johnm at magnet.com (John Mitchell)
Date: Thu, 20 Apr 2000 13:12:45 -0400
Subject: Creating Classes Dynamically?
In-Reply-To: <eviltofu-783F16.12334920042000@news.earthlink.net>
Message-ID: <Pine.SGI.4.05.10004201301200.6784-100000@lemur.magnet.com>

On Thu, 20 Apr 2000, Jerome Chan wrote:

> Is there a way to create an object of a class from a string?

You can save and restore objects using the 'pickle' module:

import pickle

class Struct:
    pass

# create new object 'j'
#
j = Struct()
j.name = 'john'
j.title = 'stud'

# package it up
#
x = pickle.dumps(j)
print x

# "x" is a semi-printable representation of object "j":
# => looks something like "(i__main__  Struct  p0  
# (dp1  S'title'  p2  S'stud'  p3  sS'name'  p4  S'john'  p5  sb."
#
z = pickle.loads(x)

# "z" is a new object, loaded from "x", which is a
# copy of our original object.  To test it, print out its name.
#
print z.name

# => "john"


Also check out the shelve and binascii modules.


- j






From dan at cgsoftware.com  Wed Apr 12 09:11:52 2000
From: dan at cgsoftware.com (Daniel Berlin+list.python)
Date: 12 Apr 2000 09:11:52 -0400
Subject: Python/Perl Popularity (Re: A Mountain of Perl...)
In-Reply-To: m.faassen@vet.uu.nl's message of "12 Apr 2000 11:09:18 GMT"
References: <200004102111.HAA03380@envy.fulcrum.com.au> <8d0pcq$7ti$1@nnrp1.deja.com> <g0ssm03o.fsf@dan.resnet.rochester.edu> <8d1lgu$ga0$1@newshost.accu.uu.nl>
Message-ID: <zoqzlazr.fsf@dan.resnet.rochester.edu>

>>>>> "Martijn" == Martijn Faassen <m.faassen at vet.uu.nl> writes:


    Martijn> Daniel Berlin+list.python <dan at cgsoftware.com> wrote:
    tom> -- Is there a Python equivalent of CPAN and the Perl CPAN
    tom> module?  The closest I know of is Parnassus.  But Parnassus
    tom> is merely a collection of links, not an archive, and it
    tom> doesn't have any facilities (AFAIK) for automatic
    tom> installation.

    >> So, who gives a crap?  I try not to be that lazy.

    Martijn> I care, for instance. The distutils people care. Please
    Martijn> don't assume nobody cares when there's a whole SIG about
    Martijn> this issue.
I didn't say nobody cares, i was basically saying i don't see all that many complaints
about this on the python lists.
Maybe i'm just missing them.
I realize their is a whole SIG about the issue, but it's not a problem
for me.


    Martijn> It's not only about laziness (though that's a good trait
    Martijn> in a programmer :), it's about accessibility. Python
    Martijn> modules should be very easy to get and install. 
No doubt this is true.
    Martijn> This
    Martijn> makes Python more useable by people who don't know
    Martijn> Python, such as sysadmins, and users. It also is good for
    Martijn> Computer Programming for Everybody.

Sure, i'm all for it too, I just don't see why it should be all that
high of a priority.
I also don't see why Tom praises the CPAN module so much.
My main problem with it is
"
203 installed modules have no parseable version number     
"
Yet i haven't installed any modules for PERL on my own.

    Martijn> [snip]
    >> Why do i get the feeling you'll never be satisfied?

    Martijn> Why attacking Tom so much? 
I've not attacked him at all.
I simply asked a question.
    Martijn> What has Tom done to you in
    Martijn> the past? Why does he get the bad reception he does?

Because he comes in here with both barrels firing.
What reception did he expect?
    Martijn>  He
    Martijn> isn't spreading misinformation, is he? He has reasonable
    Martijn> requests, in my opinion.


    Martijn> Quick-someone-start-a-math-discussion!-ly yours,




From rhicks at nospam.rma.edu  Fri Apr 14 19:41:21 2000
From: rhicks at nospam.rma.edu (Robert Hicks)
Date: Fri, 14 Apr 2000 23:41:21 GMT
Subject: TCP/IP protocols
References: <ZSMJ4.1099$Y14.27933@news.uswest.net>
Message-ID: <B0OJ4.4$Pa.1446@news>

Did you look at the LIB documentation? Python has a plethora of
documentation (in PDF) on the website http://www.python.org/doc/

Have fun!

Bob

"Jeff Massung" <jmassung at magpiesystems.com> wrote in message
news:ZSMJ4.1099$Y14.27933 at news.uswest.net...
> I'm new to Python programming, but am very experienced in C/C++. Could
> someone point me to a good reference for TCP/IP programming with Python
> (pdf, book, anything :]) thanks.
>
> Jeff
>
>




From ullrich at math.okstate.edu  Sun Apr  2 15:43:47 2000
From: ullrich at math.okstate.edu (David C. Ullrich)
Date: Sun, 02 Apr 2000 14:43:47 -0500
Subject: Does Python have a long floating data type please?
References: <000601bf9c18$4d147e00$182d153f@tim>
Message-ID: <38E7A2F3.566CC0EE@math.okstate.edu>

Tim Peters wrote:

> [David C. Ullrich]
> > ...
> >     Exactly what convinced me that a real should be a pair of
> > longRationals (giving a lower bound and an upper bound for the
> > "real" value). At least the basic arithmetic operations seem
> > simple enough that a person could get them right, and they'd be
> > obviously right.
> > ...
> [Stuff that Tim claimed he'd find exciting snipped]
>
> Yes, this is what I'd do too -- you'll get something you can trust much
> faster this way.  Jurjen's real.py can be viewed as space optimizations of
> this approach, first representing the interval by a midpoint and a delta,
> then fixing the delta at 1 by introducing a power-of-2 scale factor instead.
> The details are tricky, though!
>
> Something all "interval" approaches have problems with is systematic
> pessimism.  The classic example is
>
>     y = x - x
>
> y is in fact exactly 0 no matter how fuzzy x is, but straightforward
> interval subtraction yields a result centered at 0 and twice as fuzzy as x.
>
> In your original post you had a milder example, x*x.  This has got to be >=
> 0 in actuality, but if x is e.g. [-2, 2] then straightforward interval
> multiplication yields [-4, 4] instead of [0, 4].
>
> Because of this pessimism, some simple series approaches to building
> functions (like exp and sin etc) out of the arithmetic primitives can fail
> to converge; this is where proofs start to get hard despite (or perhaps
> because of <wink>) the simplicity of the primitives.

    I guess I shouldn't say I don't see the problem until I've done it.
Exp is increasing, so it seems easy: given that a <= x <= b what
follows about exp(x) is exactly exp(a) <= exp(x) <= exp(b); you
calculate exp(a) and exp(b) as precisely as needed, insert the
appropriate fudge factor and there you are. (If you're talking about
summing the power series with an interval-valued argument I
see why that would be tricky. But I don't see why you'd do that.)

    Sin is not quite so simple, since it's not increasing. But given
a <= x <= b what follows about sin(x) is A <= sin(x) <= B,
where each of A and B is either sin(a), sin(b), 1, or -1, depending
on where the multiples of Pi lie wrt the interval [a,b]. Seems
like a few cases to consider (and we have to start by getting
(a divmod pi) to the appropriate precision) but I don't see
any big problems.

    Easy to say, I guess.

>
> doable-in-the-end-but-it-doesn't-stay-trivial-ly y'rs  - tim



From A.Weis at t-online.de  Fri Apr  7 19:14:09 2000
From: A.Weis at t-online.de (Arnold Weis)
Date: Fri, 07 Apr 2000 23:14:09 GMT
Subject: Paid to work in Python? (was: Re: Choice Of Language)
References: <Uqvw4.2261$yL2.28868@typhoon.southeast.rr.com> <8c3ats$b5f$1@nntp.Stanford.EDU> <slrn8eact2.sq7.nobody@bambi.visi.com> <38e5f8ed.29659873@news.nikoma.de> <8cl2mq$971$1@mawar.singnet.com.sg>
Message-ID: <38ee674b.1229749@news.main-echo.net>

ngps at madcap.dyndns.org (Ng Pheng Siong) wrote:

>I once saw a German term that was translated to "improve-worse-ment."
>
>Can some one provide the German word, please?

Verschlimmbesserung?

(Trying to make something better, but making it worse instead)

Regards

	Arnold


From exarkun at flashmail.com  Sun Apr 16 17:44:48 2000
From: exarkun at flashmail.com (Jp Calderone)
Date: Sun, 16 Apr 2000 21:44:48 +0000
Subject: Python w/ SSL
Message-ID: <38FA3450.1DD845B2@flashmail.com>

 Can anyone point me to an *up to date* Python/SSL interface?  I've
found about 4 w/ search engines, but the most recent is from 1996 and
has a link to Python 1.3 in case you are without an interpreter ;)
(It also doesn't help that starship is still down - again :( )

 Thanks in advance
 Jp


-- 
If the automobile had followed the same development as the computer, a 
Rolls-Royce would today cost $100, get a million miles per per gallon,
and explode once a year killing everyone inside.
                -- Robert Cringely, InfoWorld
--
 9:43pm up 19 days, 2:50, 0 users, load average: 0.02, 0.07, 0.09


From twcook at iswt.com  Mon Apr  3 20:12:00 2000
From: twcook at iswt.com (Tim Cook)
Date: Tue, 04 Apr 2000 00:12:00 +0000
Subject: PyGreSQL errors
References: <v04210108b50ec6c8e9bf@[192.168.1.2]>
Message-ID: <38E93350.4BFAF363@iswt.com>

Lee Joramo wrote:
> 
> I am trying to get pygresql to work. I have a Cobalt RaQ3 server which runs
> a slightly modified version of RedHat. The system came with PostgreSQL
> pre-installed. When I import the pygresql's basics.py tutorial at pythons

  I have no experience with PygreSQL BUT! I suggest that you
download and install PostgreSQL according to their documentation.
I tried to go searching around for where the RH RPM put things
and it was very different from the PostgreSQL docs. I just found
it easier to do a full build from source.

Hope this helps,
-- Tim --
Open Source Health Care Alliance
Join us to help build solutions!
http://www.openhealth.com/en/healthcare.html



From effbot at telia.com  Thu Apr 13 06:33:56 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Thu, 13 Apr 2000 10:33:56 GMT
Subject: Why is if a in dict.keys() so slow.....
References: <38F5A075.CDFA32D9@bedford.waii.com>
Message-ID: <oohJ4.1587$Za1.26310@newsc.telia.net>

Martin Franklin <martin.franklin at waii.com> wrote:
> Just thought i'd throw this out there....
>
> This is very slow:-
>
> if split[0] in map_dict.keys():
>     print 'found it!', map_dict[split[0]]
> else:
>     print 'not found'
>
> Compared to:-
>
> try:
>     print 'found it!',map_dict[split[0]]
>
> except:
>     print 'Not found'
>
> The dictionary had about 40000 key:value pairs and I needed to
> look through it 250000 times.....
>
> Question is why?

any special reason you couldn't figure this out from
the documentation? ;-)

if you read up on dictionaries and the "in" operator,
you'll find that:

    keys() creates a list of 40,000 items, and "in"
    does a linear search in that list...

    ...while [key] looks the key up in a data structure
    which is carefully designed for fast lookups.

(btw, the try/except version is faster only if you have
more hits than misses.  otherwise, you may get better
performance by using "if dict.has_key(key)".)

</F>




From sholden at bellatlantic.net  Tue Apr  4 17:55:03 2000
From: sholden at bellatlantic.net (Steve Holden)
Date: Tue, 04 Apr 2000 21:55:03 GMT
Subject: Two questions
References: <38EA5FE8.853E9B75@callware.com>
Message-ID: <38EA640F.FBC878CB@bellatlantic.net>

Ivan Van Laningham wrote:
> 
> Hi All--
> I'm finally getting a chance to use Python at work.  I have two
> questions:
> 
> 1)  Is there a way to run silent (i.e., we determine where things go and
> the customer never sees any wizards) installs of Python, Tcl/Tk and
> Win32all?
> 
I'll have to leave this to the cognoscenti: I just ran the standard install.

> 2)  What stuff do I need to use msaccess with Python?  Do I need to use
> Win32all, or can I use Tkinter for my (minimal) GUI, with msaccess?
> 
Win32all might be more appropriate if you want to drive Access as a COM
object, but if you just need access to it as a data source, ODCB is going
to give you most of what you want.

I would recommend, from personal experience, the mxODBC module, written
by Marc-Andre Lemburg.  There's no reason why you shouldn't mix this
with Tkinter GUI code.  It works fine alongside wxPython window interfaces.

See http://starship.python.net/crew/lemburg/mxExtensions.html as your
starting point.  I also like mxDateTime, which is required anyway by
mxODBC.

regards
 Steve

> <there-that-didn't-hurt-did-it>-ly y'rs,
> Ivan
> ----------------------------------------------
> Ivan Van Laningham
> Callware Technologies, Inc.
> http://www.pauahtun.org
> http://www.foretec.com/python/workshops/1998-11/proceedings.html
> Army Signal Corps:  Cu Chi, Class of '70
> Author:  Teach Yourself Python in 24 Hours

--
"Bulding information systems just because it's fun."
Steve Holden  sholden at bellatlantic.net  703 716 7275


From sholden at bellatlantic.net  Sun Apr  2 14:41:29 2000
From: sholden at bellatlantic.net (Steve Holden)
Date: Sun, 02 Apr 2000 18:41:29 GMT
Subject: was Re: Why should I switch to Python?
References: <seas5ia89177@corp.supernews.com> <1EeF4.43$CA.2422@news> <oqzord6h5i.fsf@titan.progiciels-bpi.ca> <38E618D7.A15F9D9B@webwitches.com> <05AB7F747D3FC9FA.FF1C6A9D99274EAE.D8581C4B2696A7E4@lp.airnews.net>
Message-ID: <38E79424.F10BA58@bellatlantic.net>

Sounds like the time machine could come in handy here too :-)

Cameron Laird wrote:
> 
> In article <38E618D7.A15F9D9B at webwitches.com>,
> Shae Erisson  <shae at webwitches.com> wrote:
>                         .
>                         .
>                         .
> >maybe a newsgroup bot that finds common text patterns and posts FAQtual
> >replies?
>                         .
>                         .
>                         .
> Comp.lang.tcl does this, by the way; a bot emits
> <URL:http://www.westend.com/~kupries/c.l.t.welcome.html>
> to first-time posters.  There is discussion now about
> refinements involving multi-posting, posts that mention
> "newbie", automated answers to common questions, ...
> We're thinking as seriously as possible about how to
> send it out *before* they first ask for it ...
> --
> 
> Cameron Laird <claird at NeoSoft.com>
> Business:  http://www.Phaseit.net
> Personal:  http://starbase.neosoft.com/~claird/home.html

--
"Bulding information systems just because it's fun."
Steve Holden  sholden at bellatlantic.net  703 716 7275


From rhicks at nospam.rma.edu  Wed Apr 19 15:21:24 2000
From: rhicks at nospam.rma.edu (Robert Hicks)
Date: Wed, 19 Apr 2000 19:21:24 GMT
Subject: SNiFF+ for Linux
Message-ID: <UGnL4.1$kM.566@news>

There was a "how-to" to integrate Python into SNiFF+. I have 2 questions
regarding this:

1. Does anyone have this info?
2. Does it work with the free Linux version that only supports C/C++?

Thanks,

Bob




From hniksic at iskon.hr  Tue Apr  4 09:04:19 2000
From: hniksic at iskon.hr (Hrvoje Niksic)
Date: 04 Apr 2000 15:04:19 +0200
Subject: Tuples -- who needs 'em
References: <00e101bf9d80$a8384d20$74eb0b18@stcla1.sfba.home.com> <va3G4.1110$n68.190454272@newsb.telia.net>
Message-ID: <dpn1nauiek.fsf@mraz.iskon.hr>

"Fredrik Lundh" <effbot at telia.com> writes:

> -- tuples can be used as dictionary keys.  mutable collections can
>    not.

This is wrong.

> -- good designers know that hypergeneralization tends to be a bad
>    idea.

...and this is arm-waiving.

Sorry, couldn't resist.  :-)


From me at nospam.edu  Wed Apr 26 16:32:43 2000
From: me at nospam.edu (Me)
Date: Wed, 26 Apr 2000 20:32:43 GMT
Subject: <nifty hieroglyphics here>
References: <8e4cc2$d5t@netnews.hinet.net> <BWEN4.12235$HG1.329193@nnrp1.uunet.ca> <Pine.WNT.4.21.0004261229180.246-100000@nt_alester> <slrn8gei3a.2um.kc5tja@garnet.armored.net>
Message-ID: <LnIN4.14$kM.3741@news>

right...but since this newsgroup doesn't have the special designation...it
is in English.

Samuel A. Falvo II <kc5tja at garnet.armored.net> wrote in message
news:slrn8gei3a.2um.kc5tja at garnet.armored.net...
> In article <Pine.WNT.4.21.0004261229180.246-100000 at nt_alester>, Andy
Lester
> wrote:
>
> >> The official language of usenet newsgroups is English,
> >
> >Oh?  And where's THAT defined?
>
> If memory serves me correctly, I believe that it is a *recommendation*
(not
> necessarily a strict standard) that English be the `official' language of
> Usenet newsgroups, in the RFCs.  I recall seeing a statement to the effect
> of (paraphrased), "Newsgroups which are not in English should append their
> two-character ISO country identifier to the newsgroup name." Thus, the
> French-speaking comp.lang.python would appear as comp.lang.python.fr;
> likewise, an Italian alt.rec.radio.amateur would be formed as
> alt.rec.radio.amateur.it.
>
> --
> KC5TJA/6, DM13, QRP-L #1447
> Samuel A. Falvo II
> Oceanside, CA




From nascheme at enme.ucalgary.ca  Thu Apr 20 02:37:49 2000
From: nascheme at enme.ucalgary.ca (Neil Schemenauer)
Date: Thu, 20 Apr 2000 00:37:49 -0600
Subject: __del__ problem - would adopting Garbage Collection fix this?
In-Reply-To: <l03102805b52461abc14d@[193.78.237.146]>; from just@letterror.com on Thu, Apr 20, 2000 at 08:35:13AM +0100
References: <aw2L4.8907$HG1.245743@nnrp1.uunet.ca> <7ojL4.9424$HG1.254543@nnrp1.uunet.ca> <Z%oL4.9643$HG1.257533@nnrp1.uunet.ca> <m31z41iyhg.fsf@atrus.jesus.cam.ac.uk> <38FE7F1E.168E914A@cosc.canterbury.ac.nz><slrn8ft35a.uqs.nascheme@cranky.a <B5240D22.4AFD%abrahams@mediaone.net> <slrn8ft7qc.u9n.nascheme@cranky.arctrix.com> <l03102805b52461abc14d@[193.78.237.146]>
Message-ID: <20000420003749.A31345@acs.ucalgary.ca>

On Thu, Apr 20, 2000 at 08:35:13AM +0100, Just van Rossum wrote:
> Maybe you're right, but to me it feels this is in the exact
> same category as a __del__ exception, so ignoring & collecting
> seems more logical to me as a default.

It shouldn't be hard to change.  I will wait until I get some
more feedback though.

    Neil

-- 
"More computing sins are committed in the name of efficiency (without
necessarily achieving it) than for any other single reason, including
blind stupidity." -- W.A. Wulf



From donn at u.washington.edu  Mon Apr 24 18:29:48 2000
From: donn at u.washington.edu (Donn Cave)
Date: 24 Apr 2000 22:29:48 GMT
Subject: sys.exit
References: <3904C35C.B16BDE75@mindspring.com>
Message-ID: <8e2hss$1250$1@nntp6.u.washington.edu>

Quoth Chuck Esterbrook <echuck at mindspring.com>:
| The short question:
|
| If I catch a SystemExit exception (because of an invocation of
| sys.exit(code)), how do I examine the argument to sys.exit()?

...
| Upon printing the "excInfo" tuple I get:
|
|     (<class exceptions.SystemExit at 80afc70>, <exceptions.SystemExit instance at 8110f58>, <traceback object at 811f540>) 
|
| And the attributes of the SystemExit instance are:
|
|     ['__doc__', '__init__', '__module__'] 
|
| I don't know how to get my hands on the exit code.

That's not the instance you're looking at there, but the class, so
that's kind of a red herring.  The instance has a 'code' attribute
for you.

So you need something like

	eclass, einstance, etrace = sys.exc_info()
	if eclass == SystemExit:
		if einstance.code == 0:
			self._errorOccurred = 0

	Donn Cave, University Computing Services, University of Washington
	donn at u.washington.edu


From thomas at cintra.no  Thu Apr 20 16:02:29 2000
From: thomas at cintra.no (thomas at cintra.no)
Date: Thu, 20 Apr 2000 20:02:29 GMT
Subject: Sax installed, but not working
Message-ID: <38ff5f04.1047500@news.c2i.net>

Hi,

I`m using Python 1.5.2 on a Mandrake 7.0 Linux box. Installed
python-xml-0_5_1-2_i386.rpm and xml.sax exists in the site-packages
folder. But when I try 

from xml.sax import *

it fails and says there`s no module called sax. All the demo-scripts
and tutorials I`ve tried fails too.

What`s wrong? I`ve got the saxlib.zip and saxdrv.zip too, but they
seem to be installed correctly. 

All I actually need is to traverse a tree generated from a
xml-document and stuff each data-element into a database, based on
tag-location and sibling-relations.

Hints???


Thomas


From gaiusm at my-deja.com  Thu Apr 20 14:14:03 2000
From: gaiusm at my-deja.com (gaiusm at my-deja.com)
Date: Thu, 20 Apr 2000 18:14:03 GMT
Subject: pythonware site is down??
Message-ID: <8dnhcr$104$1@nnrp1.deja.com>

I was trying to download the Python Imaging Library (PIL).
But the pythonware site seems t be down.
I'm seeing a directory tree and not much in it when I point my browser
at www.pythonware.com.
Puzzled,
Gaius


Sent via Deja.com http://www.deja.com/
Before you buy.


From emile at fenx.com  Thu Apr 20 22:55:37 2000
From: emile at fenx.com (Emile van Sebille)
Date: Thu, 20 Apr 2000 19:55:37 -0700
Subject: editor for win32
References: <NpHL4.779$fK2.113609@news.flash.net> <PhOL4.9343$wM3.107684@typhoon.nyroc.rr.com>
Message-ID: <01b301bfab3d$161fc9c0$01ffffc0@worldnet.att.net>

And it supports line numbers (easier debugging from tracebacks)
visible whitespace (no tab vs space issues)
workspaces (multiple associated source files)
find on disk with jump to's (great for finding examples of usage or
source)
Hex display mode (analyzing bianry content)
save in dos/mac/unix format (edit other sources)

and-reg-ex-search-and-replace-too ly y'rs,

Emile van Sebille
emile at fenx.com
-------------------


----- Original Message -----
From: news-server <KingArthur at roundtable.dyndns.org>
Newsgroups: comp.lang.python
To: <python-list at python.org>
Sent: Thursday, April 20, 2000 6:37 PM
Subject: Re: editor for win32


> Textpad is a great programming text-editor. It has Syntax highlighting
for
> all the popular languages and various other languages. Even has syntax
> highlighting for SQL and IDL.
>
> You can get it at www.textpad.com (you can also add your python
interpreter
> to the tools and run your programs from within textpad).
>
>
> Scott Hathaway wrote in message ...
> >Can anyone recommend a python editor for Windows?
> >
> >If there is an IDE for windows (other than IDLE), even better.
> >
> >Thanks,
> >Scott
> >
> >
>
>
> --
> http://www.python.org/mailman/listinfo/python-list
>
>





From just at letterror.com  Thu Apr 20 13:16:38 2000
From: just at letterror.com (Just van Rossum)
Date: Thu, 20 Apr 2000 18:16:38 +0100
Subject: __del__ problem - would adopting Garbage Collection fix this?
In-Reply-To: <000801bfaa9c$c9957aa0$d62d153f@tim>
References: <l03102806b5246c57437f@[193.78.237.146]>
Message-ID: <l03102802b524eaecfc41@[193.78.237.179]>

At 3:48 AM -0400 20-04-2000, Tim Peters wrote:
>BTW, this was all discussed at extreme length on Python-Dev -- check it out!
>I doubt anyone wants to repeat all of that whlie struggling to pretend it's
>new & interesting <wink>.

Last time I looked, I couldn't find a *single* link on python.org that
points to python-dev subscription info. Even a search for "python-dev"
reveals nothing...

but-you've-convinced-me-anyway-ly yrs  - just





From jschmitt at vmlabs.com  Mon Apr 24 14:28:23 2000
From: jschmitt at vmlabs.com (jschmitt at vmlabs.com)
Date: Mon, 24 Apr 2000 18:28:23 GMT
Subject: regulare expression searches and newlines
References: <8e0oco$k75$1@nnrp1.deja.com> <f2TM4.3782$Za1.55405@newsc.telia.net>
Message-ID: <8e23nr$21d$1@nnrp1.deja.com>

I guess I forgot to mention that I got it to work by not looking at the
beginning of the line.  It turns out that for my data set, not looking
for targets only at the beginning of a line was ok.

I guess if I'm running Python in Windows and I'm searching inside (I'm
assuming) a Linux or a Mac text file, Python thinks the file is just
one long line.

For the general case, the answer is to make the text files native to
your platform.

John

In article <f2TM4.3782$Za1.55405 at newsc.telia.net>,
  "Fredrik Lundh" <effbot at telia.com> wrote:
> schmitthead at my-deja.com wrote:
> > I was doing a search like this:
> >
> > searchobj = re.compile( r'^foobarbaz' )
> >
> > where I was looking for something at the beginning of a line.  This
> > works well for native text files, but not when I was searching
through
> > Linux files running on NT or NT files on Linux or some other
variation
> > because of the newline difference, I guess.  Is there any way to
handle
> > the general case?
>
> there's nothing wrong with your regular expression.
>
> I suggest printing the strings you're trying to match against,
> to see if you can figure out what's wrong.
>
> if that doesn't help, post the code.
>
> </F>
>
>


Sent via Deja.com http://www.deja.com/
Before you buy.


From moshez at math.huji.ac.il  Mon Apr 10 01:57:14 2000
From: moshez at math.huji.ac.il (Moshe Zadka)
Date: Mon, 10 Apr 2000 07:57:14 +0200 (IST)
Subject: A Mountain of Perl Books + Python Advocacy
In-Reply-To: <8cr4an$j40$2@newshost.accu.uu.nl>
Message-ID: <Pine.GSO.4.10.10004100754430.2656-100000@sundial>

On 9 Apr 2000, Martijn Faassen wrote:

> Anyway, yeah, that's what I noticed in this newsgroup. Whenever someone asks
> if Python is a good language Tim Peters tells them to check out Haskell, or
> someone mentions Smalltalk or Scheme or Lisp or whatnot. It's rare that
> people are being sent to C or C++ (unless it's for Python extensions) or
> Perl though (but it does always get the 'good for fast text processing'
> mention). It's nice to see so many comparisons between languages here.

Ah, Martijn, this is all part of the standard Timbot trick to prove Python
rules -- in his on words, about how to provide an objective comparisons
between languages, "just say 5 good points about every language, and
conclude 'it obvious that overall, Python wins'".
(He also gives a few nice examples)

world-domination-through-brain-washing-ly y'rs, Z.

--
Moshe Zadka <mzadka at geocities.com>. 
http://www.oreilly.com/news/prescod_0300.html
http://www.linux.org.il -- we put the penguin in .com




From johann at physics.berkeley.edu  Wed Apr  5 13:08:42 2000
From: johann at physics.berkeley.edu (Johann Hibschman)
Date: 05 Apr 2000 10:08:42 -0700
Subject: Why should I switch to Python?
References: <000701bf9c5a$b806b420$752d153f@tim> <oqity066km.fsf@titan.progiciels-bpi.ca> <slrn8ej2gc.b3t.warlock@raven.myth> <m3itxyz744.fsf@atrus.jesus.cam.ac.uk> <050420000914595917%pecora@anvil.nrl.navy.mil>
Message-ID: <mt1z4k8oh1.fsf@astron.berkeley.edu>

Louis M Pecora writes:

> In article <m3itxyz744.fsf at atrus.jesus.cam.ac.uk>, Michael Hudson
> <mwh21 at cam.ac.uk> wrote:

>> don't know it; I think it involves slinging Riemann zeta functions).
>                                      ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

> There's a good phrase.

Holy slinging Riemann zeta functions, Batman!

Quick, Robin, fetch the Bat-Conformal-Mapper.

-- 
Johann Hibschman                           johann at physics.berkeley.edu


From exarkun at flashmail.com  Sat Apr 15 07:30:33 2000
From: exarkun at flashmail.com (Jp Calderone)
Date: Sat, 15 Apr 2000 11:30:33 +0000
Subject: Python CGI
Message-ID: <38F852D9.E2144F17@flashmail.com>

 I recently converted one of my perl CGI's to Python, and came across an
bizarre and extremely undesirable side affect - Whenever a form that uses
the Python script as its ACTION, the python scripts seems to get executed
about 8 times.  At first I thought there was some logic error in my program
making it octuple all the output, but I can't find anything that might do
this.  Has anyone seen this before, or know how to fix it?

 Thanks in advance,
  Jp

-- 
Where a calculator on the ENIAC is equipped with 18,000 vacuum tubes and
weighs 30 tons, computers in the future may have only 1,000 vacuum tubes and
weigh only 1.5 tons.    -- Popular Mechanics, March 1949
--
 11:27am up 17 days, 16:34, 1 user, load average: 0.23, 0.11, 0.04


From shecter at darmstadt.gmd.de  Tue Apr 25 22:27:57 2000
From: shecter at darmstadt.gmd.de (Robb Shecter)
Date: Wed, 26 Apr 2000 04:27:57 +0200
Subject: Metaclasses?
References: <3905F4A0.2B39C705@darmstadt.gmd.de> <m3em7tg8ei.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <3906542D.62B94F17@darmstadt.gmd.de>

Michael Hudson wrote:
> 
> b) How, when you've "fixed" Python do you propose, in method bodies,
>    to distinguish between local and instance variables?  I might drop
>    my longstanding habit of complaining about people complaining about
>    the necessity of "self" if you can answer this question...
> 

Ah hah.  I see the problem.  The language would need to be extended in
one of three ways, right?  Either:

Choice A:  Have a smalltalk-like instance variable declaration.
Choice B:  Have a smalltalk-like local declaration in each method.

...and then, depending on which way the language changed, this would
determine the lookup rules for the variable.  (ie. with choice A, it'd
be: "if it's not explicitly declared, then it's a local variable.")

Choice C:  Have both A & B (like smalltalk does).

This last one is probably not necessary, and not very much like a
scripting language anymore. Now that I've written this all out, I have
to say that I like "Choice A".  I come from Java, and it did bug me a
bit that there was no one easy place to see what instance variables a
class has.  It made understanding the XML parser source code harder, for
example.

PS: I didn't mean to imply that Python is "broken".  I guess that the
repeated "self" just rubs me the wrong way.

- Robb


From culliton at clark.net  Sat Apr  1 15:28:01 2000
From: culliton at clark.net (Tom Culliton)
Date: Sat, 01 Apr 2000 20:28:01 GMT
Subject: Why should I switch to Python?
References: <seas5ia89177@corp.supernews.com> <00033120055300.03687@quadra.teleo.net> <secfidq8921@corp.supernews.com>
Message-ID: <lZsF4.16089$0o4.101396@iad-read.news.verio.net>

In article <secfidq8921 at corp.supernews.com>,
Aaron Turner  <aturner_ at _pobox.com> wrote:
>In article <00033120055300.03687 at quadra.teleo.net>, Patrick Phalen <python-list at teleo.net> wrote:
>Yeah, I found those before I posted (as well as reading the archives of
>this group on deja news).   Honestly I found them for the
>most part very lacking in technical information.  Mostly things like:
>
>- Perl code is ugly, Python code is easy to read
>- Python was OO from the start, Perl's OO was a bolt on
>
>But things like this aren't particularly useful in trying to determine
>which language gets the job done better.  Code readibilty has 
>historically been a human problem which is solved with coding
>standards set for the programmers IMHO.   So I'm more interested in 
>things like:

Calling readability a "human problem" begs the issue, some languages
encourage readble code and discourage unreadble code while others make
it almost painful to write stuff that makes sense a few days later.

This issue is especially critical given the statistical trusim that
half of the people developing code have skills which are below
average.  I'll go further and say that about 70-80% just want to get
through their day without putting out anymore effort than absolutely
necessary.  A language that still makes it possible for this segment
of the community to produce maintainable code is an enormous win.

In my experience Python is such a language, and Perl most definitely
is not.

>- What is the impact of Python's OO being there from the start?
>- Does it perform better?
>- Is it significantly easier to use?  
>- More powerful/support features that aren't in Perl yet?
>
>Honestly the more I reasearch, the more it seems that it's purely 
>a matter of taste.  If you like Perl, there's nothing inherient about
>Python that makes it better.   Sorry if this just sounds like a troll, but
>I'm really trying to get a feel about something I know nothing about 
>from personal experiance.

You're dismissing subjective arguments ("easy to read") in one breath
and then asking subjective questions ("significantly easier to use")
in the next.  Do you honestly want to be convinced or are you just
looking for rationalizations not to switch?

>Basically if you know both languages, and were planning a mid-sized 
>project, why would you choose Python?  It's hard to take my friend
>seriously in his evaluation when he says he'd rather do this project
>in C++  than in Perl (especially when he doesn't know Perl).

We choose Python after doing a survey of "scripting"/VHLL languages,
with Perl, TCL, Scheme, and Python being the main contenders.  We also
looked at (and continued to look at) a variety of less well know
options.  Going into the evaluation we had strong backgrounds in C,
C++, shell, awk, ... and substantial experience with all of the better
known languages but Python.  Python still won going away.

Perl was slightly faster but was just too cryptic and made it too hard
to write maintainable code.

TCL was smaller, a bit easier to extend/embed, and had tools like TK
and Expect, but was just too slow, only really had two types (strings
and lists od strings) and didn't catch typos that the Python compiler
yelled about right away.

Scheme was... well... scheme.  Lisp fans to the contrary it just
wasn't for us.  It didn't have the libraries of modules we needed, we
got lost in the parenthesis even with the editors help, ...  Scheme is
a beautiful elegant powerful expressive language, but it just isn't
most peoples cup of tea.

Python was a delight.  At first we all pulled faces at the significant
whitespace, but it grew on us since it enforced uniform layout.  The
libraries and builtin types were a treasure chest, the syntax clean
and elegant, the support for OOP amazingly straight forward.

The documentation was generally excellent, written with a clarity and
lucidity which would put most native speakers to shame. ("This guy is
Dutch?  He writes like Brian Kernighan." (We'd all been struggling
with Bjarne's 1st edition a bit too long.))

We were intrigued.  We grew enthusiastic.  We became rabid fans.

There were things that we still wished for, more speed (but it was
fast enough), the ability to do more static type checking (none of the
langauages on our short list really provided this though), the ability
to do more stringent encapsulation, ...  But in practice we seldom
found these to be real problems.

You can teach the basics of Python to someone in an afternoon.  I've
done it many times.  After getting past the inital resistance to
learning something new, people with programming experience seem to
"get it" amazingly fast.  There are remarkably few puzzled or pained
looks, and many smiles of delight.  The enthusiasm grows quickly,
especially when you show them things like a full blown multi-user chat
server in a page of easily understood code, or a utility of roughly
the same size and clarity to do DB to DB transfers without fighting
with dump formats and the like, or when they whip together a clean
solution to some otherwise difficult problem on their own in a couple
hours.

THIS is why the buzz about Python is on the verge of becoming
deafening.

Try it for a week or two.  Do some small projects or utilities with
it.  Read a couple of the tutorials and the examples in the Lib, Demo,
and Tools directory.  You'll probably still find your self doing quick
hacks in perl, but you'll probably also find yourself seeing your
friends point.


From smalleys at gte.net  Wed Apr 12 23:01:15 2000
From: smalleys at gte.net (sue)
Date: Thu, 13 Apr 2000 03:01:15 GMT
Subject: Need an enum type extension
Message-ID: <38F52A7A.D8841724@gte.net>

Has anyone done an enum implementation?  I'd prefer one like
Ada (with attributes), but I can settle for one that works
more like C.  The main thing an enum_literal can't do is be
equal to an integer.  This rules out several of the simple
implementations I found on DejaNews.  Performance is not an
issue at this time.

Sue.


From effbot at telia.com  Tue Apr 18 18:13:39 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Tue, 18 Apr 2000 22:13:39 GMT
Subject: Split
References: <38FCC20D.624AB2F6@searchy.net>
Message-ID: <n65L4.501$wYl.187474432@newsb.telia.net>

Penpal International <ppi at searchy.net> wrote:
> I've already managed to split a string with this:
>
> split = regsub.split(line,'\s+')

footnote:

    split = string.split(line)

is faster (in 1.6, line.split() is even better)

> I should get something like this as result:
>
> ['This', 'is', 'a', 'test', 'short', 'and', 'sweet', 'of', 'split', '']
>
> But how can I print only the 5th for example?

print split[4]

for more on lists, read section 3.1.3 of the tutorial
(again?):

    http://www.python.org/doc/current/tut/node5.html

</F>




From jstok at bluedog.apana.org.au  Sun Apr  2 11:36:48 2000
From: jstok at bluedog.apana.org.au (Jason Stokes)
Date: Mon, 3 Apr 2000 01:36:48 +1000
Subject: New Features in Python 1.6
References: <200004011740.MAA04675@eric.cnri.reston.va.us>
Message-ID: <fPJF4.61343$3b6.238942@ozemail.com.au>

Guido van Rossum wrote in message
<200004011740.MAA04675 at eric.cnri.reston.va.us>...

>Complementary to the Distutils are the Imputils, or Import Utilities.
>Python's import mechanism has been reworked to make it easy for Python
>programmers to put "hooks" into the code that finds and loads modules.
>The default import mechanism now includes hooks, written in Python, to
>load modules via HTTP from a known URL.
>
>This has allowed us to drop most of the standard library from the
>distribution.  Now, for example, when you import a less-commonly-needed
>module from the standard library, Python fetches the code for you.  For
>example, if you say
>
>    import tokenize
>
>then Python -- via the Imputils -- will fetch
>http://modules.python.org/lib/tokenize.py for you and install it on your
>system for future use.  (This is why the Python interpreter is now
>installed as a setuid binary under Unix -- if you turn off this bit, you
>will be unable to load modules from the standard library!)


This is the only item silly enough to prove conclusively that we have an
April Fool's joke!  The problem with all these items is that they are all
just too plausible.  Double quote for unicode, single quote for ASCII -- I
can just imagine someone coming up with that!  Likewise the bizarre string
mutability semantics, and the self correcting "split" method that decides
which string is the split method by looking at which one occurs most
frequently inside the other!




From tjreedy at udel.edu  Sat Apr 29 00:05:34 2000
From: tjreedy at udel.edu (Terry Reedy)
Date: Sat, 29 Apr 2000 00:05:34 -0400
Subject: The Simple Economics of Open Source
References: <1255227918-2015166@hypernet.com>
Message-ID: <8edmnr$od7$1@news.udel.edu>

"Gordon McMillan" <gmcm at hypernet.com> wrote in message
news:1255227918-2015166 at hypernet.com...
> Terry Reedy wrote:

> > "Gordon McMillan" <gmcm at hypernet.com> wrote in message news:1255272357-
> > > One of those open source delusions is that being open source
> > > somehow makes a difference to the end consumer. Bunk.

> > Whether most consumers yet know it or not, there is all the
> > difference in the world.

> Of course there is; all your points are correct. But no end user
> will read the source and discover the bug / trojan horse /
> whatever. It's another developer that will.

Which is what I thought I said, and certainly meant to say.

> My point was that to the end user, those files are a waste of  disk
space.

When binaries are available, agreed (and getting the source should then be
optional).
 Like most Windows/Pythons users, I do not currently have the source files.

> They get value from the fact that it's open sourced, not from the source.

And my point was that they do get such value with respect to aspect of
software
which has been much in the news lately.  In other words, I was trying to
present
an open-source selling point that I have not seen much discussed or
exploited yet.

I can imagine, sometime in the future, the existence of an independent,
third-party, open-source
examination and certification service.

> - Gordon

Terry





From ppetru at coltronix.com  Wed Apr 12 11:28:44 2000
From: ppetru at coltronix.com (Petru Paler)
Date: Wed, 12 Apr 2000 18:28:44 +0300
Subject: No method overloadin I suppose ?
In-Reply-To: <38F4938E.8F245B68@altern.org>; from tsmets@altern.org on Wed, Apr 12, 2000 at 03:15:42PM +0000
References: <38F4938E.8F245B68@altern.org>
Message-ID: <20000412182844.G7538@coltronix.com>

On Wed, Apr 12, 2000 at 03:15:42PM +0000, Thomas SMETS wrote:
> ## Beginning of Python program ...
> class Lister:
>   def __init__(self):
>     print 'In constructor ...'
>     __init__(1)
> 
>   def __init__(self, lValue):
>     print 'In constructor with parameter'
>     self.l = lValue
> 
>   def __repr__(self):
>     print '\n******\n\tValue is : ', l, '\n********'
> 
> if __name__ == "__main__":
>   x = Lister()
>   y = Lister (2)
> ## End of python program
> 
> Trace of execution is :
> Traceback (innermost last):
>   File "Minheritence.py", line 14, in ?
>   x = Lister()
> TypeError: not enough arguments; expected 2, got 1
> End of trace ...
> 
> I guess I can't do methods overloadin', as shown... WHY is that ? I'm
> wrong ?

    Yes, you are wrong. Try:

class Lister:
	def __init__(self, lValue=1)
		self.l = lValue

-Petru



From raffael at mediaone.net  Sat Apr 22 23:35:02 2000
From: raffael at mediaone.net (Raffael Cavallaro)
Date: Sat, 22 Apr 2000 23:35:02 -0400
Subject: The Simple Economics of Open Source
In-Reply-To: <3901247A.E87B5D0A@rubic.com>
Message-ID: <B527E7A5.51AA%raffael@mediaone.net>

> And the success of that approach has made painting
> a profitable and lucrative career choice, in contrast
> to the thousands of struggling software engineers who
> can't even qualify for an NEA grant. <wink>
> 
> If I didn't have my part-time performance art
> income to help pay the bills, I could never
> afford to support my programming lifestyle.
> 
> -Jeff

Sarcasm aside, I think you miss the point. Those artists least willing to
share, most jealous of their "secrets" have always been those who were most
successful. Unsucessful young artists are, by and large, quite willing to
share ideas, working techniques, etc. It's actually the more established,
financially succesful ones, who are not.

I see a parallael to the software industry here as well. Those with a
lucrative cash cow (e.g., MS) are reluctant to share code, but those who
just make a living programming are happy to share source. It would seem that
sharing is inversely proportional to the value of that which is shared.

Ralph




From dalke at acm.org  Thu Apr 20 23:14:34 2000
From: dalke at acm.org (Andrew Dalke)
Date: Thu, 20 Apr 2000 21:14:34 -0600
Subject: __del__ problem - would adopting Garbage Collection fix this?
References: <aw2L4.8907$HG1.245743@nnrp1.uunet.ca> <7ojL4.9424$HG1.254543@nnrp1.uunet.ca> <Z%oL4.9643$HG1.257533@nnrp1.uunet.ca>  <l03102802b5245023a247@[193.78.237.147]> <slrn8fv4eo.1kt.neelk@brick.cswv.com>
Message-ID: <8doh1p$3sk$1@slb2.atl.mindspring.net>

Neel Krishnaswami wrote:
> An even more sensible idea would be to drop __del__ methods
> altogether, since IMO you deserve to lose if you try to use
> the same deallocator for arbitrary resources that you use for
> memory, but that idea is probably not going to fly. ;)


Whenever this topic comes up, I bring out my case.  I've got
a Python wrapped interface to a C library.  The library also
works with FORTRAN, so stores a table mapping externally
exposed integer handles to internally used pointers.  In other
words, every object created is accessible, so even a C based
GC cannot delete all unused objects.  (At least to my rather
limited understanding of how automatic garbage collectors work
with non-cooperative C libraries.)

My wrapper code, however, does know when it's appropriate
to delete/free objects, and I've hooked that in using __del__.

This, BTW, is why my code doesn't work with JPython using a
JNI version of the C library instead of a SWIG one.

                    Andrew
                    dalke at acm.org





From R.Brodie at rl.ac.uk  Mon Apr 10 06:44:26 2000
From: R.Brodie at rl.ac.uk (Richard Brodie)
Date: Mon, 10 Apr 2000 11:44:26 +0100
Subject: Python paradigms
References: <Pine.LNX.4.21.0004100757580.19783-100000@fep132.fep.ru>
Message-ID: <8csbab$1422@newton.cc.rl.ac.uk>

"Oleg Broytmann" <phd at phd.russ.ru> wrote in message
news:Pine.LNX.4.21.0004100757580.19783-100000 at fep132.fep.ru...

>    No. Code is intended not only to beeing run on processor, but also (and
> often more important) to beeing read by human. Code is speech, it is the
> way to interchange ideas.
>    In that sense ugly code is not working even if you add three books of
> comments. Code should be readable. Period.

If this were always the case, why does Python have an interactive mode?
Python is a useful language to try out a few rough drafts in too. And I
disagree that elegance and transparency are the same thing. The Fast
Fourier Transform is one of the classic algorithmic optimisations but
produces obscure code. Likewise most of the standard fast sorting
algorithms.




From michael at camelot-it.com  Tue Apr  4 10:10:08 2000
From: michael at camelot-it.com (Michael Tiomkin)
Date: Tue, 04 Apr 2000 16:10:08 +0200
Subject: Determining Classname of an object
References: <8ccbk4$b6p$1@nnrp1.deja.com>
Message-ID: <38E9F7C0.54899FE4@camelot-it.com>

a.__class__.__name__

mpetroni at my-deja.com wrote:
> 
> Is there a way to find out from which class an object has been created?


From pinard at iro.umontreal.ca  Wed Apr 26 09:07:53 2000
From: pinard at iro.umontreal.ca (=?ISO-8859-1?Q?Fran=E7ois_Pinard?=)
Date: 26 Apr 2000 09:07:53 -0400
Subject: Need advice performing a search
In-Reply-To: scarblac-spamtrap@pino.selwerd.nl's message of "26 Apr 2000 08:48:44 GMT"
References: <3905F795.A536CFA7@stsci.edu> <3906944F.F718FE91@teleatlas.com> <slrn8gdbak.bp5.scarblac-spamtrap@flits104-37.flits.rug.nl>
Message-ID: <oq8zy1ypqu.fsf@titan.progiciels-bpi.ca>

scarblac-spamtrap at pino.selwerd.nl (Remco Gerlich) writes:

> from string import find
> def testline(line):
>   return find(line, "first_string") or find(line, "second_string")

Shouldn't we have:

   return find(line, "first_string") >= 0 or find(line, "second_string") >= 0

-- 
Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard





From tismer at tismer.com  Thu Apr 27 14:07:21 2000
From: tismer at tismer.com (Christian Tismer)
Date: Thu, 27 Apr 2000 20:07:21 +0200
Subject: Newbie Question: Is there a Python equiv to Perl's "do" statement?
References: <8e9rna$ajh$1@nnrp1.deja.com>
Message-ID: <390881D9.5661AD51@tismer.com>


mdefreitas at sikorsky.com wrote:
> 
> I'm trying to call a python script from another python script (given
> the filename).

Yes, you can use the execfile function to do that.

execfile("otherscript.py")

ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaunstr. 26                  :    *Starship* http://starship.python.net
14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     where do you want to jump today?   http://www.stackless.com



From phd at phd.russ.ru  Mon Apr 10 04:00:03 2000
From: phd at phd.russ.ru (Oleg Broytmann)
Date: Mon, 10 Apr 2000 08:00:03 +0000 (GMT)
Subject: Python paradigms
In-Reply-To: <38F138CD.38983629@acm.org>
Message-ID: <Pine.LNX.4.21.0004100757580.19783-100000@fep132.fep.ru>

On Mon, 10 Apr 2000, Robert W. Cunningham wrote:

> Oleg Broytmann wrote:
> > > python's "and" and "or" operators doesn't evaluate the
> > > second argument unless they really have to...
> >
> >    Yes, I missed that, sorry.
> 
> So, it works just fine, right?

   No. Code is intended not only to beeing run on processor, but also (and
often more important) to beeing read by human. Code is speech, it is the
way to interchange ideas.
   In that sense ugly code is not working even if you add three books of
comments. Code should be readable. Period.

Oleg.
---- 
    Oleg Broytmann    http://members.xoom.com/phd2.1/    phd2 at earthling.net
           Programmers don't die, they just GOSUB without RETURN.




From akuchlin at mems-exchange.org  Wed Apr  5 11:12:08 2000
From: akuchlin at mems-exchange.org (Andrew M. Kuchling)
Date: 05 Apr 2000 11:12:08 -0400
Subject: A Mountain of Perl Books + Python Advocacy
References: <20000405053920.20579.qmail@web2102.mail.yahoo.com> <8cfe42$o02$1@nnrp1.deja.com>
Message-ID: <3d3dp01t13.fsf@amarok.cnri.reston.va.us>

rzantow at my-deja.com writes:
> the whole language, and buying more books. I'm really surprised there
> are as many as 13 Python books on the market. I'd have thought there
> was a need for maybe three, along with about two pamphlets and a poster.

I'm not surprised, and think there's room for still more.  Hopefully
the next wave of books will aim at specific topics: Python and
{Tkinter,KDE,GTk+,wxWindows}, Python for numeric work, for Web stuff,
for bioinformatics, for teaching programming, inside Zope, JPython for
Java programmers who need to embed a scripting language, etc.

Introductory books are a filled niche, IMHO, and they all sort of look
the same, with the same topics and organization: here are the basic
types, here are lists and dictionaries, here are modules, here are
classes, etc... Probably this all stems from the Tutorial's ordering
of topics.  Doubtless there's some iconoclast who'd start with classes
and work downward (or a Knuth who'd start with bytecode and then work
up).  Those variations would be interesting to see, but ringing more
tiny changes on the same introductory structure seems like a poor way
to spend time.

-- 
A.M. Kuchling			http://starship.python.net/crew/amk/
Nobody can find fault with legitimate ambition, but when the wealth of the
spiritual and intellectual life is reduced to a formula for overcoming sales
resistance, we protest.
  -- Robertson Davies, _A Voice from the Attic_


From sabren at manifestation.com  Thu Apr 13 21:38:44 2000
From: sabren at manifestation.com (Michal Wallace (sabren))
Date: Thu, 13 Apr 2000 21:38:44 -0400 (EDT)
Subject: how do I know to what module an object belongs?
Message-ID: <Pine.BSF.4.10.10004132122420.4951-100000@chi.pair.com>

Hey All,

   I'm not having much luck with the Python docs tonight. :)

   I'm building a wrapper around the Python DB-API. Each module that
implements the API is supposed to define a bunch of objects like
STRING, DATE, BINARY, etc, so you can tell what kind of field you're
dealing with. But those values change from module to module.. Eg, the
XYZ database module, might use strings, whereas the ABC module might
use numeric codes. So basically, I need to test for something like:

   if fieldtype == somemodule.STRING:
       pass

   Now, because this is an abstraction layer, I don't know which module
I'm dealing with. The user creates a connection object from whatever
module they want, and then passes just that connection object to my
class.

   I know I can just as easily have them pass a reference to the
module along with it, but I'd be happier if I could just look at the
connection object and know which module it came from. Can I do this?

Cheers,

- Michal
-------------------------------------------------------------------------
http://www.manifestation.com/         http://www.linkwatcher.com/metalog/
-------------------------------------------------------------------------




From dave at dogwood.com  Fri Apr 28 09:43:33 2000
From: dave at dogwood.com (Dave Cornejo)
Date: Fri, 28 Apr 2000 13:43:33 GMT
Subject: RESULT: comp.lang.ruby passes 200:16
References: <954456262.23656@isc.org> <955393816.29413@isc.org>
Message-ID: <956929413.9394@isc.org>

                                RESULT
            unmoderated group comp.lang.ruby passes 200:16

There were 200 YES votes and 16 NO votes, for a total of 216 valid
votes.  There were 5 abstentions and 2 invalid ballots.

For a group to pass, YES votes must be at least 2/3 of all valid
(YES and NO) votes.  There must also be at least 100 more YES votes
than NO votes.

A five day discussion period follows this announcement.  If no
serious allegations of voting irregularities are raised, the
moderator of news.announce.newgroups will create the group shortly
thereafter.

Newsgroups line:
comp.lang.ruby		The Ruby dynamic OO programming language.

Voting closed at 23:59:59 UTC, 20 Apr 2000.

This vote was conducted by a neutral third party.  Questions
about the proposed group should be directed to the proponent.

Proponent: Conrad Schneiker <schneiker at jump.net>
Votetaker: Dave Cornejo <dave at dogwood.com>

RATIONALE: comp.lang.ruby

Ruby is a very-high-level, dynamic object oriented programming language.
Ruby provides many of the best-liked power and convenience features of Perl
and Python on one hand, with an elegantly simple and powerful syntax that
was partly inspired by Eiffel and Ada on the other hand.  Ruby is a more
thoroughly objected oriented language than Python, and some Ruby features
were inspired by Smalltalk. Ruby is an open source language.

There are over 1000 members of 4 Japanese Ruby mail lists (general,
developers, math, extension writers), with a peak volume topping 100
messages per day. At present, there is one English Ruby mail list with
about 175 members, with a recent volume often exceeding 10 messages per
day and growing.  Many users of the Japanese Ruby mail lists have some
understanding of English and are also prospective comp.lang.ruby users
as well.

Much of 1999's newsgroup discussion of Ruby took place on comp.lang.python,
which is not optimal, since comp.lang.python is dedicated to the Python
language. For obvious reasons, Ruby users do not post there frequently. In
2000, most Ruby discussion has occurred in comp.lang.misc. This again is not
ideal since other comp.lang.misc regulars are typically more interested in
issues that transcend Ruby, and have said as much.

In Japan, Ruby has overtaken Python in terms of popular usage. During 1999
and into early 2000, Ruby has been featured in several articles in the
English language software trade press, indicating it is at the takeoff
threshold that Perl and Python attained years ago.  (Please see
http://www-4.ibm.com/software/developer/library/ruby.html for the latest
example.) The English language Ruby home page is
http://www.ruby-lang.org/en/, which provides links to documentation and
download pages for Ruby.

CHARTER: comp.lang.ruby

The comp.lang.ruby newsgroup is devoted to discussions of the Ruby
programming language and related issues.

Examples of relevant postings include, but are not be limited to,
the following subjects:

- Bug reports
- Announcements of software written with Ruby
- Examples of Ruby code
- Suggestions for Ruby developers
- Requests for help from new Ruby programmers

The newsgroup is not moderated.  Binaries are prohibited (except the
small PGP type). Advertising is prohibited (except for announcements
of new Ruby-related products).

END CHARTER.


comp.lang.ruby Final Voter list

Voted YES
-------------------------------------------------------------------------------
andy#research.canon.com.au                                          Andy Newman
joandvan#ihug.com.au                                                   joandvan
alxx#tig.com.au                                                              al
jim_g#tig.com.au                                                            Jim
p.fuller#student.murdoch.edu.au                                     Paul Fuller
geldridg#progsoc.uts.edu.au                                      Geoff Eldridge
toivo#ucs.uwa.edu.au                                              Toivo Pedaste
gcross#netspace.net.au                                             Graeme Cross
doctor#doctor.nl2k.ab.ca                                          Dave Yadallee
trevort#vianet.on.ca                                             Trevor Tymchuk
MSteed#altiris.com                                                   Mike Steed
ml#au.com                                                       Mike Linksvayer
kimdv#best.com                                                     Kim DeVaughn
dgoodger#bigfoot.com                                              David Goodger
jcm#bigskytel.com                                                  David Porter
Eric.Peterson2#PSS.Boeing.com                                     Eric Peterson
Makoto.Nakagawa#jp.compaq.com                                   Nakagawa Makoto
fenghou#erols.com                                         Jack Hou</FONT></DIV>
julesd#erols.com                                                   Jules Dubois
smlucas#famvid.com                                                 Steven Lucas
emile#fenx.com                                                Emile van Sebille
mzadka#geocities.com                                                Moshe Zadka
Henry#hdsSystems.com                                            Henry Schneiker
SchaeferFFM#hotmail.com                                         Oliver Schaefer
thucdat#hotmail.com                                                  Dat Nguyen
tetsu#jpn.hp.com                                               Tetsuya WATANABE
gnhurst#hurstlinks.com                                             Guy N. Hurst
sullivan#Mathcom.COM                                             Steve Sullivan
petelyons#mindspring.com                                             Pete Lyons
Taka.Adachi#msdw.com                                            Takanori Adachi
tim_one#email.msn.com                                                Tim Peters
jolomo#netcom.com                                                   Joey Morris
tgm#netcom.com                                             Thomas G. McWilliams
Ralph.Amissah#operamail.com                                       Ralph Amissah
tseaver#palladion.com                                               Tres Seaver
apter#panix.com                                                    stevan apter
Tbone#pimpdaddy.com                                               Tushar Samant
jimrtex#pipeline.com                                                  Jim Riley
wknaka#pobox.com                                                   Wes Nakamura
wang#rjka.com                                                       Huayin Wang
jbauer#rubic.com                                                     Jeff Bauer
chris.browne#sabre.com                                       Christopher Browne
takaaki.higuchi#Japan.Sun.COM                                   Takaaki Higuchi
peter#taronga.com                                                Peter da Silva
Dave#Thomases.com                                                   Dave Thomas
juliet#thomases.com                                               Juliet Thomas
andy#Toolshed.Com                                                   Andrew Hunt
ellie#Toolshed.Com                                               Ellie Callahan
joe31416#voicenet.com                                                 Joe Myers
Abhay.Natu#wcom.com                                                  Abhay Natu
xanthian#well.com                                               Kent Paul Dolan
rascproponents#yahoo.com                                       Brian P. Sweeney
wpostma#ztr.com                                 <SPAN class=485464314-31032000>
mrilu#ale.cx                                                              mrilu
mertz#gnosis.cx                                                     David Mertz
Juergen.Cohrs#alcatel.de                                          Juergen Cohrs
Ekkehard.Uthke#gmx.de                                            Ekkehard Uthke
peter.sommerfeld#gmx.de                                        Peter Sommerfeld
burow#ifh.de                                      Burkhard D. Steinmacher-Burow
sn#neopoly.de                                                      Sven Neuhaus
vote-de#kholdan.snafu.de                                            Tobias Erle
Walter.Zettel#t-online.de                                         Walter Zettel
bouvin#daimi.au.dk                                            Niels Olof Bouvin
jpsecher#diku.dk                                              Jens Peter Secher
quinn#groat.ugcs.caltech.edu                                       Quinn Dunkan
cjc26#cornell.edu                                                Cliff Crawford
adam#mlug.missouri.edu                                             Adam Procter
neelk#alum.mit.edu                                            Neel Krishnaswami
behrends#cse.msu.edu                                            Reimer Behrends
shippy#cs.nmt.edu                                                  Jeff Shipman
rcl211#is9.nyu.edu                                             Robert C Linwood
graham#ccs.ucsb.edu                                               Graham Hughes
reic0024#d.umn.edu                                              Aaron J Reichow
uscfadsl#ds.cesga.es                                             David S de Lis
hcjake#cs.hut.fi                                                Jaakko Vayrynen
decoux#moulon.inra.fr                                                Guy Decoux
dcalvelo#pharion.univ-lille2.fr                                  Daniel Calvelo
drew.mcdowell#msfc.nasa.gov                                  Andrew D. McDowell
legradi#alpha1.obuda.kando.hu                                     Gabor Legradi
ariels#compugen.co.il                                           Ariel Scolnicov
masato#csys.ce.hiroshima-cu.ac.jp                                 Masato KIYAMA
gotoken#math.sci.hokudai.ac.jp                                     GOTO Kentaro
masa#koala.astro.isas.ac.jp                                     Masahiro Tanaka
thitoshi#ne.scphys.kyoto-u.ac.jp                              TAKAHASHI Hitoshi
nagai#ai.kyutech.ac.jp                                          Hidetoshi NAGAI
sinara#blade.nagaokaut.ac.jp                                   Shin-ichiro HARA
ogino#coop.nagoya-u.ac.jp                                         Mitsuru Ogino
naosone#naruto-u.ac.jp                                               Naoto SONE
ohdachi#nifs.ac.jp                                              Satoshi Ohdachi
washizu#kkc1.c.u-tokyo.ac.jp                                    Hitoshi Washizu
fukusima#goto.info.waseda.ac.jp                                Masaki Fukushima
igarashi#ueda.info.waseda.ac.jp                                Hiroshi Igarashi
os#iij.ad.jp                                                      OHARA Shigeki
ksasaki#kddnet.ad.jp                                                 Kan Sasaki
Taiji.Can#atesoft.advantest.co.jp                                     Taiji Kan
suenaga#birds.co.jp                                            SUENAGA Tomohiro
p95#brain-tokyo.co.jp                                             Taku Nakajima
kimura#sqa.canon.co.jp                                            KIMURA Koichi
fit0298#fitec.co.jp                                              ARIMA Yasuhiro
inoue#ftv.co.jp                                                    Hirosi Inoue
ogi#open.nm.fujitsu.co.jp                                      Takayuki Ogihara
tachino#open.nm.fujitsu.co.jp                                  Tachino Nobuhiro
uno#open.nm.fujitsu.co.jp                                            Shunji Uno
ito#htk.hitachi-cable.co.jp                                       Ito Kazumitsu
kyamada#hitachi-densa.co.jp                                        Yamada Kenji
agawa#tir.hitachi-sk.co.jp                                          Norio Agawa
mt#izu.co.jp                                                   Masato Toyoshima
taca#kyoto.jepro.co.jp                                           Takahiro Kambe
Shinichi_Hiramoto#justsystem.co.jp                           HIRAMOTO Shin-ichi
kitamura#jzf.co.jp                                              Mikiya Kitamura
sa-sakamoto#kddwinstar.co.jp                                  Sadayuki Sakamoto
hide#lisa.co.jp                                                  Hideyuki Ezaki
inouesin#pic.melco.co.jp                                         Shinichi Inoue
kima#np.bs1.fc.nec.co.jp                                            KIMATA Isao
h_inoue#netlab.co.jp                                              Hiroshi Inoue
matz#netlab.co.jp                                            Yukihiro Matsumoto
shugo#netlab.co.jp                                                  Shugo Maeda
kondo#nik-prt.co.jp                                             Mitsuhiro Kondo
hmorita#ohmsha.co.jp                                             Hisashi Morita
adachi#tpegasus.lab.okidata.co.jp                                    Jun Adachi
miyagawa#sanritz.co.jp                                         Seiichi Miyagawa
komatsu#sarion.co.jp                                          Katsuyuki Komatsu
thata#net.sony.co.jp                                         Takashi Hatakeyama
tamori#sm.sony.co.jp                                            Hirofumi Tamori
kanematu#sra.co.jp                                              Daiji KANEMATSU
aoki#technoa.co.jp                                                Hirotaka AOKI
kawata#titan.co.jp                                              Masahiro Kawata
anakata#vector.co.jp                                               Ataru NAKATA
hengsu#winningrun.co.jp                                              HengSu Kim
takagi#etl.go.jp                                               Hiromitsu Takagi
irie#hri.pref.hokkaido.jp                                                  Yuji
simtak#mua.biglobe.ne.jp                                        Takashi Shimizu
junji999#mud.biglobe.ne.jp                                      Junzi TAKAHASHI
kyano#mva.biglobe.ne.jp                                              YANO Kyoji
ikarashi#d3.dion.ne.jp                                           IKARASHI Akira
marutomo#jade.dti.ne.jp                                          INOUE Tomoichi
masao-k#a-net.email.ne.jp                                        $B6b8w2mIW(B
kai#star.email.ne.jp                                             Kazuhiko IZAWA
ymsd#ns.m-net.ne.jp                                               Yuichi Masuda
nnakamur#mxq.mesh.ne.jp                                      Noritsugu Nakamura
nobu.nakada#nifty.ne.jp                                        Nobuyoshi Nakada
felio#peanet.ne.jp                                           YAMASHIRO Shunsuke
greentea#fa2.so-net.ne.jp                                     Tomoyuki Kosimizu
tomokazu#pc4.so-net.ne.jp                                   MATSUSHIMA Tomokazu
toshirok#yb3.so-net.ne.jp                                      Toshiro Kuwabara
ozawa#mba.sphere.ne.jp                                             Makoto Ozawa
wakou#fsinet.or.jp                                                 Wakou Aoyama
earth#hokuto7.or.jp                                        Hisashi Wakai(earth)
shoch#jsdi.or.jp                                                  Shoichi OZAWA
ogochan#linux.or.jp
platypus#a1.mbn.or.jp                                          Yoshinori Tahara
kaneoka#grn.mmtr.or.jp                                           KANEOKA Shinji
noroi#mob.or.jp                                                     IKEDA Kenji
aygoofy#kt.rim.or.jp                                           Akihiro Yamauchi
hisashim#kt.rim.or.jp                                         hisashi y. morita
tadf#kt.rim.or.jp                                              Tadayoshi Funaba
inaba#st.rim.or.jp                                                 Inaba Hiroto
hattori#ppp.upload.or.jp                                        Masashi Hattori
Suketa<CQN02273#nifty.ne.jp>                                      Masaki Suketa
h.fulton#att.net                                                     Hal Fulton
jaredr#bellsouth.net                                           jared richardson
n8gray#earthlink.net                                             Nathaniel Gray
rzimmerle#earthlink.net                                         Robert Zimmerle
kikutani#galaxy.net                                            Kikutani, Makoto
c.hintze#gmx.net                                                 Clemens Hintze
d.hintze#gmx.net                                                    Dirk Hintze
jmh#gmx.net                                                            Jan Haar
shrogers#ionet.net                                             Steven H. Rogers
schneiker#jump.net                                             Conrad Schneiker
MAP2303#mapletown.net                                             Daisuke Kanda
franke#meso.net                                                   Stefan Franke
cartisan#pacwestracing.net                                          Neal Cappel
edwards#rahul.net                                               Gregory Edwards
evert#royal.net                                                Evert Meulie esq
cind0revs#toadmail.toad.net                                     Arthur J Mongan
scarblac-rt#pino.selwerd.nl                                       Remco Gerlich
boud#rempt.xs4all.nl                                            Boudewijn Rempt
kai.andresen#chello.no                                             Kai Andresen
jaj39#student.canterbury.ac.nz                                   J Alan Jackson
John_David_Galt#acm.org                                         John David Galt
perin#acm.org                                                       Lewis Perin
sferris#acm.org                                                 Scott M. Ferris
darkskyz#cyberspace.org                                           Tomer Brisker
crouton#duelists.org                                               OZAWA Sakuro
quinet#gamers.org                                                Raphael Quinet
imain#gtk.org                                                          Ian Main
h_inoue#hi-fun.org                                                   Hiro Inoue
koyama#hoge.org                                                  KOYAMA Tetsuji
jc#joerch.org                                                    Jvrg Czeranski
t#kondara.org                                                      Toru Hoshina
mentifex#scn.org                                               Arthur T. Murray
larry#smith-house.org                                               Larry Smith
dave#technopagan.org                                             David E. Smith
ysjj#unixuser.org                                               Junji Yamashita
wormboyslim#wiggly.org                                             Nigel Rantor
mak#mikroplan.com.pl                                        Grzegorz Makarewicz
qrczak#knm.org.pl                                     Marcin 'Qrczak' Kowalczyk
hgs#dmu.ac.uk                                                        Hugh Sasse
james#obeah.demon.co.uk                                             James Coupe
bill#xencat.co.uk                                              William Boughton

Voted NO
-------------------------------------------------------------------------------
yan#storm.ca                                                  Yves Bellefeuille
ru#cls.usask.ca                                                     Ru Igarashi
stainles#bga.com                                                   Dwight Brown
mcclenon#erols.com                                                 Bob McClenon
gprrspw#mindspring.com                                                G.P. Ryan
rootbeer#redcat.com                                                 Tom Phoenix
pimentel#ultranet.com                                                J Pimentel
Thilo.Ernst#gmd.de                                                  Thilo Ernst
naddy#mips.rhein-neckar.de                                 Christian Weisgerber
rick#bcm.tmc.edu                                              Richard H. Miller
4ws#gibsoncounty.net                                                 Kurt Weber
mmontcha#OregonVOS.net                                       Matthew Montchalin
denebeim#deepthot.org                                              Jay Denebeim
sthoenna#efn.org                                        Yitzchak Scott-Thoennes
phd#sun.med.ru                                                   Oleg Broytmann
chriseb#ukshells.co.uk                                           Chris Ebenezer

Abstained
-------------------------------------------------------------------------------
chris#kzim.com                                      Christopher Robin Zimmerman
sleepy#os.com                                                 Alexander Pennace
r.vanette#pink.nl                                           Robert-Jan van Ette
james#sfgiants22.freeserve.co.uk                                   James Farrar
merlin#netlink.co.uk                                            Darren Wyn Rees

Invalid ballots
-------------------------------------------------------------------------------
mntsui#netvigator.com                                             tsui man ngar
    ! No vote statement in message
mjd#plover.com
    ! No vote statement in message


To restore the email addresses above, pipe the ack list through the
following command:

    sed -e 's/#/@/g'

-- 
Voting question & problems: Dave Cornejo <dave at dogwood.com>
Voting address: vote at dogwood.com


From abouf066 at aix2.uottawa.ca  Sat Apr 29 09:29:34 2000
From: abouf066 at aix2.uottawa.ca (abouf066 at aix2.uottawa.ca)
Date: Sat, 29 Apr 2000 09:29:34 -0400
Subject: GUI developping with Python (Any tutorial )
Message-ID: <Pine.A41.3.96.1000429092757.42286A-100000@aix2.uottawa.ca>

Greetings, 

	I wonder if anybody can point me to a site where I could find the
starting steps to develop a GUI using python. 

	Thnaks for your time and help. 



_____________________________________________________________
							
					LaBoufarikoise




From cut_me_out at hotmail.com  Tue Apr 25 13:28:13 2000
From: cut_me_out at hotmail.com (Alex)
Date: 25 Apr 2000 13:28:13 -0400
Subject: Don't understand error message
References: <lukN4.1574$wYl.191480320@newsb.telia.net>
Message-ID: <etd1z3uf5ua.fsf@w20-575-109.mit.edu>

> It goes like this:
> 
> import sys, glob, operator
> print sys.argv[1:]
> sys.argv = reduce(operator.add, map(glob.glob, sys.argv))
> print sys.argv[1:]
> 
> When called with "python glob.py *.py" on Win98 I'm supposed to get a list
> of all .py files in the current directory but instead I get:
> 
> Traceback (innermost last):
>   File "glob.py", line 1, in ?
>     import sys, glob, operator
>   File "glob.py", line 3, in ?
>     list = reduce(operator.add, map(glob.glob, '*.py'))
> TypeError: call of non-function (type module)
> 
> Anyone know why?

Yeah, the problem is, your file is called glob.py, and that is loaded
when you import glob.  Instead, you should be loading the glob from the
standard python library.  To get around this, one thing you could do is
rename your glob.py file to my_glob.py, and remove your old glob.py and
glob.pyc.  Then 'python my_glob.py *.py' should work fine.  At least,
that solved the problem for me.

Alex.


From jhefferon at my-deja.com  Mon Apr 10 17:03:40 2000
From: jhefferon at my-deja.com (jhefferon at my-deja.com)
Date: Mon, 10 Apr 2000 21:03:40 GMT
Subject: How to compile Doc from tex format to pdf in win32
References: <8crdm2$2bff$1@news.cz.js.cn>
Message-ID: <8ctfj0$qeo$1@nnrp1.deja.com>

  "fmr" <mrfx at 263.net> wrote:
> I want to compile Python Doc from Tex format to
> Pdf or HTML format, Who can help me?
What are your resources?

If you have a standard TeX distribution (teTeX on a UNIX, MikTeX on a
PC, for instance, then you should be able to just say:
  > pdflatex <fn>
instead of `latex <fn>' and it will probably work.  (It might not
work if, for instance, the documents use certain kinds of PostScript
code, etc., but without looking at the documents (lately), I'd
guess there is a 99.9% chance that what I said above works.)

As for HTML, that's also possible.  But I very much like the pdf
myself.

If you do not have one of these distributions, you'll have to be more
specific about what you do have.

Jim Hefferon tex at joshua.smcvt.edu


Sent via Deja.com http://www.deja.com/
Before you buy.


From skip at mojam.com  Sat Apr  1 15:46:34 2000
From: skip at mojam.com (Skip Montanaro)
Date: Sat, 01 Apr 2000 20:46:34 GMT
Subject: Confusion about re lookahead assertions
Message-ID: <8c5n76$9le$1@nnrp1.deja.com>

I'm trying to use re's lookahead assertion constructs for the first time
and am more than a bit confused. If I understand them correctly, the
regular expression r'(?![.])([0-9]+)\s+' should only match digits if they
are not preceeded by a dot, yet the example below clearly contradicts
that.
>>> pat = re.compile(r'(?![.])([0-9]+)\s+')
>>> print pat.search("$5.00 mem")
<re.MatchObject instance at 1ff99a0>
>>> print pat.search("$5.00 mem").group(1)
00
>>> mat = pat.search("$5.00 mem")
>>> print mat
<re.MatchObject instance at 1ff93e0>
>>> print mat.group(1)
00
>>> pat = re.compile(r'(?![-A-Za-z0-9:._])([0-9]+)\s+')
>>> mat = pat.search("$5.00 mem")
>>> print mat
None
I think mat should be None or mat.group(1) should be '0' in both cases.
What am I missing?
--
Skip Montanaro (skip at mojam.com)
Mojam: http://www.mojam.com/
Musi-Cal: http://www.musi-cal.com/


Sent via Deja.com http://www.deja.com/
Before you buy.


From tseaver at starbase.neosoft.com  Sat Apr 22 23:42:11 2000
From: tseaver at starbase.neosoft.com (Tres Seaver)
Date: 22 Apr 2000 22:42:11 -0500
Subject: The Simple Economics of Open Source
References: <14591.15630.192472.440375@goon.cnri.reston.va.us> <slrn8g26tu.p7o.neelk@brick.cswv.com> <14837CB4E98A72FE.8FD547BA6E928594.5A103CFE68327EB7@lp.airnews.net> <VLvUljAKNfA5EwPM@jessikat.demon.co.uk>
Message-ID: <7B7C00D8D0547D72.DD1773F4502EF503.0B129F97531509B3@lp.airnews.net>

In article <VLvUljAKNfA5EwPM at jessikat.demon.co.uk>,
Robin Becker  <robin at jessikat.demon.co.uk> wrote:
>In article <14837CB4E98A72FE.8FD547BA6E928594.5A103CFE68327EB7 at lp.airnew
>s.net>, Tres Seaver <tseaver at starbase.neosoft.com> writes
>...
>>"That dog won't hunt!"  Your claim that "basically no one makes the
>>effort" is easily refuted:  consider the League of Women Voters,
>>and the guides they put out before each election -- that effort is
>>well-organized, and involves coordinated, non-partisan effort on the
>>scale of tens of thousands of volunteers (much like a large Open Source
>>project).  The League's explicit motive is to educate themselves and
>>the population at large on the issues and candidates;  that at least
>>some of the effort here is altruistic is obvious to anyone who hasn't
>>"internalized [economics'] assumptions."
>>
>>Altruism-is-a-shibboleth'ly,
>>
>>Tres.
>I don't think the pessimists amongst whom I count myself reject altruism
>totally. A purely altruistic action has benefit for others and none for
>the actor. Very few of the altruistic examples in the thread come close.
>
>A lot of claims are made, but few are verifiable. I think the
>internalized utility assumptions are a bit spurious as there is evidence
>that many of the assumptions that economists make (about human
>behaviour) are not verifiable and in some cases have been shown to be
>true; a classic example is the assumption of risk averse behaviour in
>investment decisions. I feel that the assumption that economic agents
>(people) always behave as though they have a well defined utility is
>false. I think that what economists are really saying is that they have
>a model where the 'average' agent has such a utility and that the model
>gives good results for certain problems. In this sense the average
>volunteer feels they behave in a non-partisan altruistic way, but the
>ensemble actually behaves selfishly although this is not selfishness in
>the way we normally think of it. 

OK, I'll by the "significant empirical predictive power" bit;  I was
actually objecting to the specific example (voter apathy), and then
vented my distaste for the prevailing mechanistic assumptions.

Utility-is-for-ants,-give-me-virtue-anytime'ly :)

Tres.
-- 
---------------------------------------------------------------
Tres Seaver        tseaver at digicool.com     http://www.zope.org
Digital Creations


From tim_one at email.msn.com  Thu Apr 20 03:48:16 2000
From: tim_one at email.msn.com (Tim Peters)
Date: Thu, 20 Apr 2000 03:48:16 -0400
Subject: __del__ problem - would adopting Garbage Collection fix this?
In-Reply-To: <l03102806b5246c57437f@[193.78.237.146]>
Message-ID: <000801bfaa9c$c9957aa0$d62d153f@tim>

[Just van Rossum, unaccountably arguing for ignoring __del__ methods]
> ...
> There's an additional argument, though: symmetry with JPython.

That JPython is currenty diseased is no excuse either.  Java certainly
doesn't ignore finalization methods *no matter what*; JPython ignores
__del__ mostly because Java's implementation of (non-default) finalizers was
painfully slow.

BTW, this was all discussed at extreme length on Python-Dev -- check it out!
I doubt anyone wants to repeat all of that whlie struggling to pretend it's
new & interesting <wink>.

don't-do-the-crime-if-you-can't-pay-the-time-ly y'rs  - baretta





From thomas at xs4all.net  Fri Apr 28 17:46:08 2000
From: thomas at xs4all.net (Thomas Wouters)
Date: Fri, 28 Apr 2000 23:46:08 +0200
Subject: Linux Processes from Python HOW-TO?
In-Reply-To: <8ecpch$j73$1@nnrp1.deja.com>; from chris_barker@my-deja.com on Fri, Apr 28, 2000 at 07:39:00PM +0000
References: <8ecpch$j73$1@nnrp1.deja.com>
Message-ID: <20000428234608.L11481@xs4all.nl>

On Fri, Apr 28, 2000 at 07:39:00PM +0000, chris_barker at my-deja.com wrote:

> Has anyone written a HOW-TO or tutorial that explains how to start and
> manipulate other processes with Python. I imagine it's trivial if you
> are familiar with the Linux/Unix process model, but I get very confused
> about when to use:

> system
> popen and it's variants
> exec and it's variants
> fork
> etc.

Well, it's detailed in most books about UNIX, and in most books about
operating systems in general. (At least on the programming level.) I dont
know any HOWTO-level explanation of it, but I'm happy to supply you with
some short descriptions. I'll do it in the reverse order of what you listed,
though, because it's easier that way ;)

First, the basics: under UNIX, and most current operating systems, a running
program consists of one or more processes. A single program, or a single
task, is usually a single process, but it's possible for several processes
to communicate with eachother, to share workload. Each process has its own
memoryspace, invisible to other processes, and it can only access its own
memoryspace.

To create a new process, you use the system call fork(). It duplicates the
current process, copies all of its memory, instruction pointers, function
stack, everything. The only difference between the two processes is the
process-id (pid), and the return code of fork(): In the new process, the
'child', fork returns 0. In the original process, the 'parent', fork()
returns the pid of the newly created child process. (If fork() fails, it
wont create a child process, of course, and the return code will signify an
error.)

Because fork() creates an exact duplicate of the old process, you can't use
it to start a new program. You use one of the 'exec*()' calls for that. This
system call *replaces* the calling process by a new process, running a new
program as indicated by the first argument to the exec() call. If exec()
returns anything, it's always an error, because if it succeeds, the process
that called it will be gone. The new process isn't really a new process,
though, because it will have the pid of the old one. It'll just be running a
different program ;)

The difference between the various exec*() calls is fairly subtle, and a bit
strange from a Python point of view: they are mostly about how to pass the
arguments: the execl*() calls take a variable number of arguments,
whereas the execv*() calls take a pointer to a list of arguments. The
exec*p() calls search the environment variable PATH for the name of the
executable, whereas the other exec*() functions expect the name of the
executable to be a full path to the program. And lastly, the exec*e()
calls also take an argument that list the environment variables that
should be passed to the new program, whereas the other calls pass the same
environment as the current process has.

These two system calls are all that is necessary to create as many and as
varying processes as you wish. None the less, the Standard C Library
provides a few funky and helpful functions that save you a lot of time in
the common cases:

system() is a library call that takes a single string describing the program it
should start, and the arguments it should get. It forks() and exec()s for
you, and then waits for the child to exit (using one of the wait() calls)
and returns the exit code. You dont have particularly much control over how
the string is executed, other than via the standard 'shell' commands: system
always starts /bin/sh with your string as argument.

popen() and it's friends are similar functions, but instead of running the
string you pass and returning the exitcode, it starts the new process with
its standard input or standard output (or both, in the case of popen2)
connected to the calling process (using a UNIX 'pipe', hence the first p in
popen()). This way, the calling process can pass data to the program just as
a user would type it on a computer, or read data that the new program writes
to the terminal.

So, basically, if you simply want to run a command and dont care much about
runtime influence, you usually use system(). If you need to pass it some
data for it to work on, for example 'gzip' or some other program that works
on the input stream, you use popen(). And if you want more control over
things, or if you simply want your python process to do two tasks at once,
or you wish to start a lot of programs in a specific order, or other
complicated things, you use fork() and exec() manually.

> I'm really hoping someopne has written this up!

Well, I just did, didn't I ? ;) There are a lot more details I didn't
mention, by the way: threads (processes that share their memory space);
redirecting stdin, stdout and stderr somewhere specific, or to another
process; creating pipes and sockets yourself; how to avoid deadlocks in
popen2; shared memory; controling ttys, how to see the difference between a
simple pipe and a real user, and how to fool programs that try to see that
difference; UNIX session ids; setuid() and other ways to change the current
process's permissions; And probably a lot more ;P Nevertheless, I hope this
helps a bit :-)

If you need more info than this, either read a UNIX manualpage for one of
those functions, or if you prefer readable material, buy a good book on UNIX
programming. They usually start with a good explanation of processes, the
memory layout, pipes, sockets and networking. Some even cover shared memory.

Drained-ly y'rs,
-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From parkw at freewwweb.com  Wed Apr 26 23:09:22 2000
From: parkw at freewwweb.com (William Park)
Date: Wed, 26 Apr 2000 23:09:22 -0400
Subject: regsub backreference question
In-Reply-To: <8e89g7$dfg$0@216.39.162.232>; from dante@oz.net on Thu, Apr 27, 2000 at 02:43:19AM +0000
References: <8e89g7$dfg$0@216.39.162.232>
Message-ID: <20000426230922.A1536@freewwweb.com>

On Thu, Apr 27, 2000 at 02:43:19AM +0000, Michael Esveldt wrote:
> I'm trying to figure out how back referencing works in python's regsub 
> module. I thought the following code would simply double the string 
> "foo" but it's just returning "foo".
> 
> regsub.sub(r"(.*)", r"\1\1", "foo")
> 
> What am I doing wrong here?

Try 
    regsub.sub(r"\(.*\)", r"\1\1", "foo")
    re.sub("(.*)", r"\1\1", "foo")

--William



From sbarron at ansic.net  Mon Apr  3 10:49:55 2000
From: sbarron at ansic.net (Scott)
Date: Mon, 3 Apr 2000 14:49:55 GMT
Subject: Platf. indep. console lib.
References: <8c03u1$igk$1@news.hit.fi>
Message-ID: <38E8AF93.70F4CCF7@ansic.net>

Hi,

There is an slang module which wraps the slang library.  It's available
for Unix and Win32 though I haven't had the time to actually use it
yet.  Here is an ftp site for it: 
ftp://quasar.ipa.nw.ru/pub/avv/Python/pyslang .  If you give it a try,
let me know how it turns out!

Scott

"Matti Johannes K?rki" wrote:
> 
> Hi!
> 
> It seems that there isn't any platform independent console (text graphics)
> library available in core Python distribution (or am I wrong?). Is there
> some specific reason for that? As far as I can tell, it should be quite
> easy to implement one with NCurses (*nix) and Conio (Windows: Watcom and
> Borland directly, Visual C with Win32API?), other platforms, I don't know.
> 
> I know that there is some kind of NCurses module for Python but I would
> like to see something simple (such as Conio) which also would be supported
> under Linux and Win32 in core Python package. It would be nice to know
> that when I install the Python package to the computer, I'll be able to
> write and run Python programs with text graphics user interfaces. I know
> that Tkinter rules, but there has been some situations when I've been
> missing console library with move/gotoxy, addch/putch and getch (and with
> certainty that my programs will work both under Win32 and Linux terminal).
> 
> What do you think?
> 
> Sincerely,
> 
>     Matti


From richard_chamberlain at ntlworld.com  Thu Apr 27 05:45:22 2000
From: richard_chamberlain at ntlworld.com (Richard Chamberlain)
Date: Thu, 27 Apr 2000 10:45:22 +0100
Subject: IP Address
Message-ID: <WZTN4.135$E44.2126@news6-win.server.ntlworld.com>

Is there a way of returning the IP address of the machine that a script is
running on?

Thanks

Richard




From quinn at riyal.ugcs.caltech.edu  Tue Apr  4 20:21:45 2000
From: quinn at riyal.ugcs.caltech.edu (Quinn Dunkan)
Date: 5 Apr 2000 00:21:45 GMT
Subject: blessed be Guido, for "5 <= X <= 10" does what it should...
References: <3TpG4.3349$HG1.98231@nnrp1.uunet.ca>
Message-ID: <slrn8el1su.3h1.quinn@riyal.ugcs.caltech.edu>

On Tue, 4 Apr 2000 13:42:41 -0500, Warren Postma <embed at geocities.com> wrote:
><X-DISCLAIMER: I'm in a weird mood.>
>
>I was just playing around today,and it occurred to me to test the
>'transitive' properties of comparison operators. Ie, writing on a
>chalkboard, you might say X is in the range 5..9 this way:
>
>        5 <= X <= 9

Yes, this has always struck me as one of the weirder corners of python.
There's a bit of magic for comparison operators:

>From the lang ref 5.9:

Formally, if a, b, c, ..., y, z are expressions and opa, opb, ..., opy are
comparison operators, then a opa b opb c ...y opy z is equivalent to a opa b
and b opb c and ... y opy z, except that each expression is evaluated at most
once. 


It's kinda cool, I guess.  Maybe ABC did it first?


From willadams at aol.com  Wed Apr 12 10:21:20 2000
From: willadams at aol.com (William F. Adams)
Date: 12 Apr 2000 14:21:20 GMT
Subject: Compiling Python Imaging under LinuxPPC
References: <20000412234630.21095@nms.otc.telstra.com.au>
Message-ID: <20000412102120.18661.00002626@ng-cg1.aol.com>

Greg McFarlane was kind enough to clue me in to BLT being:
>A graphics extension to Tk:
>    http://www.tcltk.com/blt/

I've got that, but still having trouble with python imaging, now because
imaging.h isn't appropriately installed---I found a message on the Sketch list
which had a patch, but that won't work on my LinuxPPC install, reports it as
being garbage....

Can I just d/l Imaging.h manually from somewhere or copy it in to
/usr/include/python1.5 by hand?

William


-- 
William Adams
http://members.aol.com/willadams
Sphinx of black quartz, judge my vow.



From emile at fenx.com  Fri Apr  7 10:35:20 2000
From: emile at fenx.com (Emile van Sebille)
Date: Fri, 7 Apr 2000 07:35:20 -0700
Subject: Creating variables on the fly...
References: <meh9-25D0C5.15160303042000@news.cit.cornell.edu> <38E8EFF0.D71D0415@gmd.de> <spdH4.43688$a01.958441@news.tpnet.pl>
Message-ID: <085f01bfa09e$83285a40$01ffffc0@worldnet.att.net>

Lately, I've become skeptical of faster claims
when they point away from the obvious.  I know
that there are exceptions, but this is not one.

For a small number of arguments, the difference
is negligible but measurable at a high number of
iterations.  The gap widens with more arguments.

Emile van Sebille
emile at fenx.com
--------------------
from time import time

def add1(*args):
   sum=0
   for arg in args:
     sum=sum+arg
   return sum

def add2(*args):
   return reduce(lambda x, y: x+y, args)

items = (107,  132,  147,   159,   183,
         194,  204,  285,   294,  1107,
         2132, 3147, 4159,  5183,  6194,
         7204, 8285, 9294, 10483, 14572)

for count in (3, 8, 14, 20):
   print '\n---%s---' % count
   for iterations in (5000, 15000, 50000):
      print "    ", iterations,":\n",
      for func in (add1, add2):
         print '  ',func.__name__,

         start = time()
         for i in range(iterations):
            apply(func, items[:count])
         print '%6.2f' % (time()-start)


-------------------


----- Original Message -----
From: Tomek Lisowski <Lisowski.Tomasz at sssa.nospam.pl>
Newsgroups: comp.lang.python
To: <python-list at python.org>
Sent: Thursday, April 06, 2000 9:23 PM
Subject: Odp: Creating variables on the fly...


> U?ytkownik Joachim Kaeber <kaeber at gmd.de> w wiadomooci do grup
dyskusyjnych
> napisa?:38E8EFF0.D71D0415 at gmd.de...
> > Hi,
> >
> > what about this:
> >
> > def add(*args):
> >    sum=0
> >    for arg in args:
> >      sum=sum+arg
> >    return sum
> >
> > add(1,2,3,4)
> > => 10
> >
> > add (1,2,3,4,5,6,7)
> > => 28
>
> Why not:
>
> def add(*args):
>   return reduce(lambda x, y: x+y, args)
>
> I am not completely sure, but it may be faster, than a Python for loop
>
> Tomasz Lisowski
>






From cpu at bigfoot.com  Sat Apr 29 00:49:53 2000
From: cpu at bigfoot.com (S. Hoon Yoon)
Date: Sat, 29 Apr 2000 04:49:53 GMT
Subject: ODBC / DBI help (via Win32 Extensions)
References: <rNHvjEdhm5Pp-pn2-RytgtQHbLg9d@d185d18f2.rochester.rr.com>
Message-ID: <RRtO4.26899$R4.157537@news1.rdc1.nj.home.com>

I think that's because you should use format #date# on update. Unless, I did
not understand your problems.
That's really more of SQL and access convention rather than Python.
Pls look at the docs for Access and SQL.

UPDATE [BT Dates] SET [BT Dates].[date] = #12/12/2000#;

This is what I get if I use Query wizard and you should follow it.
Hope it helps,


"Benjamin Schollnick" <junkster at nospam.rochester.rr.com> wrote in message
news:rNHvjEdhm5Pp-pn2-RytgtQHbLg9d at d185d18f2.rochester.rr.com...
> ====================================================
>           (Remove "NoSpam" to Email me)
> ====================================================
> Hi All,
>
> I'm doing some MS Access/SQL work via the ODBC & DBI
> modules... and it's working great...
>
> I've got the read, and most of the "write" functions working
> fine, except when it comes to utilizing the date values...
> I can read and convert the date information into a python readable
> value.... But I can't seem to do an UPDATE or INSERT INTO command
> with a date value inside it.
>
> Can someone possibly post a example, or a fragment of a
> working  UPDATE / INSERT INTO command?  Or point me to some
> python resources that might assist?
>
> I've checked the ODBC how-to's on the python web site, and
> they don't even touch on the subject of WRITING a date... The examples
> on reading were very informative, but obviously I'm doing something
> wrong.
>
> - Benjamin
>
> ====================================================
>           (Remove "NoSpam" to Email me)
> ====================================================
> Please feel free to copy any and or all of this sig.
> A little something for spam bots:
>
> root at localhost postmaster at localhost admin at localhost
> abuse at localhost postmaster at 127.0.0.1
>
> Chairman William Kennard: bkennard at fcc.gov
> Commissioner Susan Ness: sness at fcc.gov
> Commissioner Harold Furchtgott-Roth: hfurchtg at fcc.gov
> Commissioner Michael Powell: mpowell at fcc.gov
> Commissioner Gloria Tristani: gtristan at fcc.gov
> consumerline at ftc.gov
> fccinfo at fcc.gov
> ssegal at fcc.gov
>




From rhundt at cup.hp.com  Wed Apr  5 22:26:54 2000
From: rhundt at cup.hp.com (Robert Hundt)
Date: Wed, 05 Apr 2000 19:26:54 -0700
Subject: Error in "pow" on Itanium
Message-ID: <38EBF5EE.98B905A@cup.hp.com>

Hi @ll,

Running Python 1.5.2 as a 64-bit process on Itanium delivers an error in
test_pow. The test delivers different results whenever run! Sometimes it
even succeeds!

I looked into the sources of do_pow etc. but was not able to ficure out,
what exactly is happening there. It seems not to be safe for every
64-bit architecture. Is that right? Any pointers are highly appreciated.

Thanks in anticipation
--
Robert Hundt  HP  408-447-0546  rhundt at cup.hp.com




From icarr at bullfrog-tech.com  Mon Apr 17 18:31:16 2000
From: icarr at bullfrog-tech.com (icarr at bullfrog-tech.com)
Date: Mon, 17 Apr 2000 16:31:16 -0600
Subject: Here is some starship.python.net news
References: <38FA9A30.33DD1994@concentric.net>
Message-ID: <38FB90B4.35E9CA1B@bullfrog-tech.com>

I heard once more from the hosts of starship.python.net

Apparently, the "responsible (for fixing it) person" is our very own
Mike McLay, who is theoretically on-site with the machine today
(Monday) and I got a "hopefully, it will be back up today" from the
mscp guys.

I say theoretically and hopefully because according to the news on
my pager, today is the worst day in the history of Washington, D.C.
traffic.  So who knows if Mike McLay even made it to the machine.

If he did or once he does, let us hope that he can put it back into
working order.

Manus Hand
posting from God-knows-whose address, so please don't reply



From matthias_w at my-deja.com  Wed Apr 12 07:58:33 2000
From: matthias_w at my-deja.com (Matthias)
Date: Wed, 12 Apr 2000 11:58:33 GMT
Subject: Embedding Python in Visual C++ 6.0
References: <8d1he4$26d$1@nnrp1.deja.com> <N2YI4.2030$UB.13624@news-server.bigpond.net.au>
Message-ID: <8d1od2$9fl$1@nnrp1.deja.com>

Sorry - but I realy tried. I have the same probleme with your
pythonwin.cpp sourcecode when I try to debug it. In
<b>glue.DynamicApplicationInit(startup)</b>everything looks nice, but
if I step into <b>glue.InitInstance()</b> I get the unhandled exception
(see below). The Address where the debugger "land's" seems right
(1E401F69) it should be the address of <b>BOOL Win32uiInitInstance()
</b>, but there is no code :.(  and if I try to place a break-point in
this function, it is deactivated from VisualC (white dot with red frame)
I really do not have any idea what I am doing wrong - If a start
pythonwin_d.exe from the explorer there is no problem!

Thank you for your help
Matthias :-)

In article <N2YI4.2030$UB.13624 at news-server.bigpond.net.au>,
  "Mark Hammond" <mhammond at skippinet.com.au> wrote:
> You also need to override InitApplication().
>
> I should update the docs, but until then, simply look at
Pythonwin.cpp and
> see what it does - whatever it does, it obviously works :-)
>
> Mark.
>
> <matthias_w at my-deja.com> wrote in message
> news:8d1he4$26d$1 at nnrp1.deja.com...
> > Hy! :-)
> >
> > Maybe someone (Mark Hammond??) can help me. I
> > tried to embedd the win32 extentions from Mark in
> > a visual c++ project, but everytime the
> > <b>Win32uiInitInstance()</b> is called (via
> > <b>glue.InitInstance()</b>) I get an unhandled
> > exception at 0xC0000005 Access Violation and the
> > debugger jumps to 1E401F69 - the address of the
> > Win32uiInitInstance() function. When i try create
> > a break-point in this function this break-point
> > is disabled %-( by the compiler. The same is true
> > if i try to debug the original pythonwin project.
> >
> > Help please!
> >
> > Matthias
> >
> >
> > Sent via Deja.com http://www.deja.com/
> > Before you buy.
>
>


Sent via Deja.com http://www.deja.com/
Before you buy.


From jon_mclin at attglobal.net  Tue Apr 25 10:15:06 2000
From: jon_mclin at attglobal.net (Jon McLin)
Date: Tue, 25 Apr 2000 14:15:06 +0000
Subject: Python C++ Bindings
Message-ID: <3905A86A.39C372FB@attglobal.net>

Several links, from c.l.p

From niels at endea.demon.nl  Fri Apr 14 10:56:15 2000
From: niels at endea.demon.nl (Niels Diepeveen)
Date: Fri, 14 Apr 2000 16:56:15 +0200
Subject: an intern-like memory saver
References: <8d5ing$heb$1@nnrp1.deja.com>
Message-ID: <38F7318F.AE4F0EE5@endea.demon.nl>


sjmachin at lexicon.net schreef:
> 
> Problem:
> I have an application that works with words and can have millions of them
> in memory at one time. Apart from the main data structure, a dictionary is
> used to maintain frequencies. As the words are loaded from files, multiple
> instances of the same word don't share memory. The memory savings
> could be huge --- see how frequent "the" is in English text, or "Smith" in
> an Anglo telephone directory.
> 
> Trial solution:
> (1) Clone dictobject.c. Make it into an extension module for a type called
> "mydict". Add a method called "key_ref".with one argument:
> adict.key_ref(obj). If adict.has_key(obj) is true, this returns a reference to
> the key value inside the dictionary; else it returns "obj".
> (2) Make simple changes to the application:
> (a) Change
>    freq_dict = {}
> to
>    freq_dict = mydict.mydict()
> (b) assuming for purposes of exposition that words are stored simply in a
> list, after
>    freq_dict[w] = freq_dict.get(w, 0) + 1
> change
>    word_list.append(w)
> to
>    word_list.append(freq_dict.key_ref(w))
> 
> Results:
> Gratifying. An exercise that was running out of real memory (384 MB) and
> taking a day now takes an hour or so.
> 
> Questions:
> (1) Would this be sufficiently generally useful to make it a method of the
> standard dictionary object in Python?

It seems to make sense. I don't know if it would be widely used though.
I can think of a number of ways to do very much the same thing in
Python. I don't know whether you've tried, but I think simply using
    w = intern(w)
might be even faster than what you describe.

> (2) As methods seem to be found by sequential search, wouldn't it be a
> good idea to move "get" a bit higher up the method_def list in
> dictobject.c? At the end, after "update" and "copy", doesn't seem like a
> good idea.
> (3) Has anyone had any success in compiling Python on WinNT 4.0 with
> gcc 2.95.2? It's just fine for making extension modules; I haven't tried
> compiling the whole Python yet.

I'm working on 1.6a2 now (for Win95, should work for NT too). I've got
the
core to compile and run, but it takes some patches. I haven't got
Tkinter working yet (problems linking the Tcl/Tk libaries, any
suggestions other than building those from source welcome).
If you like, I can send you what I have so far.

> (4) Has anyone any better ideas for gauging Python memory usage than
> sitting watching the graphical display in WinNT's Task Manager? Might an
> instrumented or instrumentable malloc/free package (like Doug Lea's) that
> permitted implementation of a Python builtin memused() be the way to go,
> or is there a policy of using the standard malloc from the C library on each
> platform?

As far as I can tell, you can use any malloc package you like; there are
no assumptions other than the C standard.

-- 
Niels Diepeveen
Endea automatisering




From USENET at questionexchange.com  Fri Apr 14 09:42:28 2000
From: USENET at questionexchange.com (QuestionExchange)
Date: 14 Apr 2000 13:42:28 GMT
Subject: Python equivalent of lynx -dump?
References: <20000327200829.29200.qmail@web2101.mail.yahoo.com>
Message-ID: <3310qx@questionexchange.com>

The server only sends the raw HTML. If you want it formatted,
you need to format it yourself --
sort of. To retrieve the data from the server, you can use
urlopen from urllib. You could
alternatively use httplib, but that's generally only necessary
if you're doing something really
fancy and HTTP specific.
Once you've got the HTML, you can use htmllib to do the
parsing. It needs a "formatter",
which in turn needs a "writer" (see the fomatter module at
http://www.python.org/doc/current/lib/module-formatter.html).
The
formatter module has an AbstractFormatter and a DumbWriter,
which are both pretty basic,
but reasonably close to what "lynx -dump" does. If you want
better formatting, you can
write your own formatter and/or writer.
Here's some sample code that does basically what you want. Not
that I use a StringIO,
since DumbWriter wants to write to a file, but you want the
value in a string:
    from urllib import urlopen
    import htmllib
    import formatter
    # first, retrieve the HTML...
    html = urlopen(url).read()
    # create a "string file"...
    outfile = StringIO()
    # create a writer and formatter...
    myWriter = formatter.DumbWriter(outfile)
    myFormatter = formatter.AbstractFormatter(myWriter)
    # now parse and format the HTML...
    parser = htmllib.HTMLParser(myFormatter)
    parser.feed(html)
    parser.close()
    # get the formatted output
    data = outfile.getValue()

-- 
  This answer is courtesy of QuestionExchange.com
  http://www.questionexchange.com/servlet3/qx.usenetGuest.showUsenetGuest?ans_id=13247&cus_id=USENET


From jmassung at magpiesystems.com  Wed Apr 26 16:20:28 2000
From: jmassung at magpiesystems.com (Jeff Massung)
Date: Wed, 26 Apr 2000 14:20:28 -0600
Subject: global variables or inheritence
References: <28p8zy0irgc.fsf@lada.kom.auc.dk> <eRHN4.1113$3K4.19980@news.uswest.net> <28p4s8oipzh.fsf@lada.kom.auc.dk>
Message-ID: <j6IN4.1168$3K4.20204@news.uswest.net>

>testprog1.py:
>import os, sys, string, testprog2
>def main():
>    print 'hello'
>    myvar = 2
>    testprog2.second()
>if __name__ == '__main__':
>    main()

NOTE: myvar is local to main() (not to the file)

>
>testprog2.py:
>import os, sys, string, testprog1
>def second():
>    print testprog1.myvar
>if __name__ == '__main__':
>    print 'nope'
>

Also, just as coding practice - you're cirular use of the files is VERY
bad... it can cause many problems (and is illegal in many languages - I
can't believe Python would allow it). What is your program that you need to
do this with? Just curious.

Jeff




From bozon at natur.cuni.cz  Tue Apr 11 10:42:57 2000
From: bozon at natur.cuni.cz (Michal Bozon)
Date: Tue, 11 Apr 2000 16:42:57 +0200
Subject: IDLE lability
Message-ID: <Pine.OSF.4.10.10004111635070.6088-100000@prfdec.natur.cuni.cz>

I use IDLE as an Python IDE on my computer (which is pow(d)ered by Win98).
I really like it (IDLE, not Windows), but it is little bit unstable 
(Windows much more). I worked in 0.4 and 0.5 versions. Sometimes after
opening new window (or closing) it yielded an error. But it is caused by
tk80.dll module.
Now, after Python 1.6 release I tried New IDLE 0.6 and it seems to be even
more unstable (it falls always after closing a window). Again the error 
was in tk module, I thing tk83.dll or so.


Michal Bozon
- - - - -
bozon at natur.cuni.cz
http://www.natur.cuni.cz/~bozon



From grant at nowhere.  Fri Apr 14 15:38:19 2000
From: grant at nowhere. (Grant Edwards)
Date: Fri, 14 Apr 2000 19:38:19 GMT
Subject: How do I undo an import??
References: <7QrJ4.185421$Hq3.4353581@news2.rdc1.on.home.com> <m3d7nthbbj.fsf@atrus.jesus.cam.ac.uk> <bkKJ4.190780$Hq3.4442202@news2.rdc1.on.home.com>
Message-ID: <slrn8festb.8ld.grant@grante.comtrol.com>

In article <bkKJ4.190780$Hq3.4442202 at news2.rdc1.on.home.com>, Tom wrote:

>After importing NetCfg I type:
>
>>>> del sys.modules['NetCfg']
>
>Python responds as follows:
>
>Traceback (innermost last):
>  File "<stdin>", line 1, in ?
>NameError: sys
>
>Any ideas??

Did you do an "import sys"?

-- 
Grant Edwards                   grante             Yow!  You mean you don't
                                  at               want to watch WRESTLING
                               visi.com            from ATLANTA?


From PClaerhout at CREO.BE  Thu Apr 13 04:22:29 2000
From: PClaerhout at CREO.BE (Pieter Claerhout)
Date: Thu, 13 Apr 2000 10:22:29 +0200
Subject: Problems with PythonWin
Message-ID: <2B1262E83448D211AE4B00A0C9D61B03BA72AF@MSGEURO1>

Hello,

I tried to reinstall PythonWin build 128 today, but I got errors
when he was trying to register the Python AXScript engine.
Any idea what could be wrong? Any help would be appreciated.

Kind regards,


Pieter
Pieter_Claerhout at creoscitex.com



From mfletch at tpresence.com  Tue Apr  4 09:13:15 2000
From: mfletch at tpresence.com (Mike Fletcher)
Date: Tue, 4 Apr 2000 09:13:15 -0400 
Subject: Tuples -- who needs 'em
Message-ID: <CC59868A0A76D311B1F50090277B321C1F324E@VRTMAIL>

I think the point that is being made is this:
	The _value_ of a mutable object cannot (easily) be used as a
dictionary key...
	i.e. in this sequence, what is the value of the final line...

s = [ 1,2,3 ]
p = { s: 4, [2,3,4]: 5 }
s[:] = [2,3,4]
p[ [2,3,4] ]

You can get around that by saying (effectively):
	When a mutable object's value is used as a dictionary key, a copy of
the object is used, rather than the object itself (that is, further changes
to the object do not alter the key mapping.

Which is done, today, as:
	p = { tuple(s):4, tuple([2,3,4]):5 }

The problem is, you still don't have anything like what you really want
(that is, store the objects by the keys and somehow do magic behind the
scenes to decide which instance is going to be "the one" when keys are
equal.  Rigorously define that in a simple-to-understand and teach way, with
plausible implementation details, and you might get the powers to agree, no?

Of course, you could always use id( listvalue) as the key, but that doesn't
have the value, only the identity of the object, so two equal lists won't
give the same key.

Enjoy all,
Mike

-----Original Message-----
From: Hrvoje Niksic [mailto:hniksic at iskon.hr]
Sent: Tuesday, April 04, 2000 9:04 AM
To: python-list at python.org
Subject: Re: Tuples -- who needs 'em


"Fredrik Lundh" <effbot at telia.com> writes:

> -- tuples can be used as dictionary keys.  mutable collections can
>    not.

This is wrong.
...



From urner at alumni.princeton.edu  Sun Apr  2 14:33:37 2000
From: urner at alumni.princeton.edu (Kirby Urner)
Date: Sun, 02 Apr 2000 11:33:37 -0700
Subject: Learning in Stereo:  Math + Python
References: <ua1fes81hmb12dh1klbv0p7d0373qhleps@4ax.com>
Message-ID: <ur3fessapua3bi1k4e5g89i3ndbuejhabo@4ax.com>

>  >>> pi(100)
>  3.04936163598
>
>  >>> pi(10000)
>  3.14149716395
>

Of course you can push this slowly converging series even
further by rewriting f(n) to multiply with long integers:

 >>> def f(n):
       return 1.0/long(x)*x

 >>> pi(100000)
 3.14158310433
 >>> pi(500000)
 3.14159074373 

Never use this as an efficient way to get PI!!

Here's a continued fraction method:

def pi(depth):
   # use continued fraction of user-supplied depth to refine pi
   n = 1.0
   return 4.0 * 1.0/(1 + 1.0/pifract(n,depth))

def pifract(n,depth):
   if n<depth:
      return 2 + ((2*n+1)**2)/pifract(n+1,depth)
   else: return 2.0

And here's another fast converger using every other 
Fibonacci number:

def pifib(n):
   # return approximation for pi using
   # arctan(1) = 4 * SIGMA [ arctan(1.0/Fib(2i+1))]
   sum = 0
   for i in range(1,n+1):
      sum = sum + math.atan(1.0/fibo(2*i+1))
   return 4*sum

meaning you need and algorithm for the Fibonacci's, like:

fibcache = {}          # thanks to Tim Peters
def fibo(n):
    # return nth fibonacci number (recursive, cached)   
    if n == 0: return 0L
    elif n == 1: return 1L
    elif fibcache.has_key(n): return fibcache[n]        
    else: 
      result = fibo(n-1) + fibo(n-2)
      fibcache[n] = result
      return result

or 
      
root5   = 5.0**0.5
gold    = (1+root5)/2.0 # same as phi, named to not conflict with phi()
tau     = 1/gold

def fibo(n): # thanks to David Zachmann
    # return nth fibonacci number as a floating point
    return (gold**n - (-tau)**n)/root5
    

Then of course there's always:

  >>> import math
  >>> math.pi
  3.14159265359

  :-D

Kirby



From tim_one at email.msn.com  Sun Apr 23 00:48:08 2000
From: tim_one at email.msn.com (Tim Peters)
Date: Sun, 23 Apr 2000 00:48:08 -0400
Subject: LAmbda calculus evaluator (interactive)
In-Reply-To: <uQqM4.461$RM6.88244@news-east.usenetserver.com>
Message-ID: <000301bfacdf$1f5c7260$212d153f@tim>

[Kragen Sitaker]
> ...
> Many years ago, Colby Kraybill suggested to me that it would be nice
> to have something similar for mathematical expressions --- something
> that would do nothing on its own, but would do the grunt-work involved
> in an action like "take the reciprocal of both sides" or "divide both
> the top and bottom by (x-1)".

Hyland Software's "Merlyn" program does that; see

    http://www.hylandsoftware.com/

It's based on a general term-rewriting system, and is programmable by the
end-user (indeed, almost all the builtin rules are exposed in files you can
edit too).  However, it's proprietary, costs money, source for the engine
itself is not available, and it works only on Windows.  I often use it
instead of, e.g., Macsyma, because the high-powered systems drive me nuts
with the pervasive & often unhelpful transformations they make in an
eyeblink; Merlyn never does anything unless it's told to.

Similar but weaker symbolic facilities can be found, e.g., in Dave
Gillespie's Emacs calc, and high-end HP calculators.

If the world were sane, they would have written all these things in Python
<wink>.

the-temptation-to-use-the-time-machine-is-often-overwhelming-ly y'rs  - tim





From mhammond at skippinet.com.au  Fri Apr 21 20:55:32 2000
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Sat, 22 Apr 2000 00:55:32 GMT
Subject: win32all build 125 debug builds?
References: <8dpsik$fsu$1@nnrp1.deja.com>
Message-ID: <8M6M4.1334$v85.7864@news-server.bigpond.net.au>

<wilkes at memphisonline.com> wrote in message
news:8dpsik$fsu$1 at nnrp1.deja.com...
> Does anyone have the debug builds for win32all build 125, or the source
> so that I may build them.

The source can always be extracted from CVS - use the "b125" tag.

Mark.




From pinard at iro.umontreal.ca  Wed Apr 26 09:26:44 2000
From: pinard at iro.umontreal.ca (=?ISO-8859-1?Q?Fran=E7ois_Pinard?=)
Date: 26 Apr 2000 09:26:44 -0400
Subject: Sourceforge and Python
In-Reply-To: yawyi@gmx.de's message of "Wed, 26 Apr 2000 12:42:13 +0200"
References: <oq8zy1zqto.fsf@titan.progiciels-bpi.ca> <3906c7e8$0$29200@personalnews.de.uu.net>
Message-ID: <oq4s8pyovf.fsf@titan.progiciels-bpi.ca>

yawyi at gmx.de (Johannes Stezenbach) ?crit:

> Fran?ois Pinard <pinard at iro.umontreal.ca> wrote:
> > I would like to bring `recode' a bit closer to Python

> There already exists a GNU recode wrapper module for python:

> http://www.vex.net/parnassus/apyllo.py?find=recode
> http://www.suxers.de/python/pyrecode.htm

Thanks for the references.  File `contrib/README' for the next `recode'
distribution already has some of them, and I completed it from your post.

By "bringing closer to Python", I was more thinking more in an administrative
than a technical way.

Technically, `recode' already uses Python to generate its C or Texinfo
tables.  Given some free time, I'll rewrite and maintain a parallel version
`recode' in Python, just to make prototyping easier.  Python 1.6 already has
a lot of bundled facilities so simple recodings are very easy to do, to the
point that in many cases, a `recode' interface might be seen as overkill.
If I re-implement a parallel version of `recode' in Python, I would prefer
to use bundled facilities, merely checking that they are congruent with
`recode' knowledge, and try to get differences resolved if not.

Administratively, I was more thinking of a site offering Mailman, and mostly
based of Python scripting, or Python synchronisation methods if any.  A way
to me, and even for my users, to get closer to Python, even if indirectly.
A kind of fuzzy move in a probably good direction :-).

-- 
Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard





From kc5tja at garnet.armored.net  Wed Apr 26 12:16:13 2000
From: kc5tja at garnet.armored.net (Samuel A. Falvo II)
Date: 26 Apr 2000 16:16:13 GMT
Subject: plugins for browsers?
References: <3904ccb0.32134281@news.demon.co.uk> <etd7ldk325j.fsf@w20-575-97.mit.edu>
Message-ID: <slrn8gf18d.3m1.kc5tja@garnet.armored.net>

In article <etd7ldk325j.fsf at w20-575-97.mit.edu>, Alex wrote:
>I guess if you're using Netscape, you can specify the command yourself.
>Go to Edit -> Preferences, then in the dialogue that brings up, go to
>Navigator -> Applications.  Then go to New, and in the dialogue that
>brings up, enter .py as the suffix, and python %s as the command.

That's not a plug-in though -- that's just a helper program, and will launch
Python applications outside of the browser window.

-- 
KC5TJA/6, DM13, QRP-L #1447
Samuel A. Falvo II
Oceanside, CA


From kcazabon at home.com  Sun Apr  9 19:38:38 2000
From: kcazabon at home.com (Kevin Cazabon)
Date: Sun, 09 Apr 2000 23:38:38 GMT
Subject: More Tkinter... tkFileDialog to select a directory instead of a file?
Message-ID: <2w8I4.122246$Dv1.1320515@news1.rdc1.bc.home.com>

Is there any easy way to get tkFileDialog.selectopenfile() to select a
directory instead of a file?

I need an easy way to select a directory, even if it contains no files...

Any suggestions are appreciated, as always!

Kevin.




From greg at cosc.canterbury.ac.nz  Wed Apr  5 00:42:44 2000
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Wed, 05 Apr 2000 16:42:44 +1200
Subject: Tuples -- who needs 'em
References: <CC59868A0A76D311B1F50090277B321C1F324F@VRTMAIL>
Message-ID: <38EAC444.9994D2E0@cosc.canterbury.ac.nz>

Mike Fletcher wrote:
> 
> The question of what a "meaningful" hash function would be for a mutable
> sequence is the crux of the issue

The important thing is whether the object *behaves* immutably
as far as comparison is concerned. If two objects a and b of a type T
have the property that a == b always returns the same result regardless
of what is done to a or b, then it makes sense to use objects of type T
as dictionary keys. But there has to be *something* about them that can 
be regarded as immutable. So, in that broader sense, the statement that 
"only immutable objects can be used as dictionary keys" remains correct.

-- 
Greg Ewing, Computer Science Dept,
+--------------------------------------+
University of Canterbury,	   | A citizen of NewZealandCorp, a	  |
Christchurch, New Zealand	   | wholly-owned subsidiary of USA Inc.  |
greg at cosc.canterbury.ac.nz	   +--------------------------------------+


From sblakey at freei.com  Thu Apr 13 15:06:37 2000
From: sblakey at freei.com (Sean Blakey)
Date: Thu, 13 Apr 2000 12:06:37 -0700 (PDT)
Subject: Simple formatting string question
In-Reply-To: <meh9-637743.14563813042000@news.cit.cornell.edu>
Message-ID: <20000413190220.C11CB1CDBA@dinsdale.python.org>

On 13 Apr, Matthew Hirsch wrote:
> Hi All,
> 
> I'm trying to print something like this:
> 
>>>> print 6*'%2i' '%1i' % tuple([5,5,5,5,5,5,1])
> 
> but I'm getting an error.
> 
> Do you know what the correct syntax should be?
> 
> Thanks,
> Matt
Two issues:
First, you have to concatenate the format stings with plus.
Second, formating binds more tightly than string formatting, so you
will need parentheses around "6x'%2i' + '%1i'".

What I got to work was:
>>> print (6*'%2i' + '%1i') % tuple([5,5,5,5,5,5,1])
 5 5 5 5 5 51

-- 

Sean Blakey
FreeInternet.com
sblakey at freei.com
(253)796-6500x1025




From thamelry at vub.ac.be  Tue Apr 25 06:34:40 2000
From: thamelry at vub.ac.be (Thomas Hamelryck)
Date: 25 Apr 2000 10:34:40 GMT
Subject: how to wrap c++ software to python?
References: <8e3ggq$2r1d$1@news.cz.js.cn>
Message-ID: <8e3sc0$jss$1@mach.vub.ac.be>

fmr <mrfx at 263.net> wrote:
: I will call c++ routine in python,how to wrap
: c++ routine for python.

: Is there any utils to work except SWIG.

Try:

http://www.acl.lanl.gov/siloon/

There's also something called CXX that 
might be of interest to you.

Cheers,

---
Thomas Hamelryck
Institute of Molecular and Structural Biology
Aarhus University Gustav Wieds Vej 10C
DK-8000 Aarhus C Denmark          






From mrfx at 263.net  Sun Apr  9 22:27:26 2000
From: mrfx at 263.net (fmr)
Date: Mon, 10 Apr 2000 10:27:26 +0800
Subject: How to compile Doc from tex format to pdf in win32
Message-ID: <8crdm2$2bff$1@news.cz.js.cn>

I want to compile Python Doc from Tex format to
Pdf or HTML format, Who can help me?




From cut_me_out at hotmail.com  Thu Apr 13 09:35:08 2000
From: cut_me_out at hotmail.com (Alex)
Date: 13 Apr 2000 09:35:08 -0400
Subject: Why should I switch to Python?
References: <seas5ia89177@corp.supernews.com> 	<slrn8f4d0j.1t0.bwinton@tor.dhs.org> 	<slrn8f4tfr.6hh.quinn@mono.ugcs.caltech.edu> 	<IvFI4.5759$HG1.168717@nnrp1.uunet.ca> 	<etdu2h8btwj.fsf@w20-575-109.mit.edu> <87u2h6nwak.fsf@psyche.evansnet>
Message-ID: <etdn1myt983.fsf@w20-575-109.mit.edu>

> > Mail-copies-to: [nobody|poster]
> 
> Uh, I think that's supposed to be _one_ of:

Oh, yeah, thanks.  Forgot to change my .gnus.el file on this account.

Alex.


From bwinton at tor.dhs.org  Mon Apr 10 15:35:47 2000
From: bwinton at tor.dhs.org (Blake Winton)
Date: Mon, 10 Apr 2000 19:35:47 GMT
Subject: Python paradigms
References: <8cmr70$b0e$1@pegasus.csx.cam.ac.uk> <8covpn$4jq$1@slb6.atl.mindspring.net> <8cqk7u$dmc$1@pegasus.csx.cam.ac.uk> <3H4I4.709$4w1.17557@typhoon.nyroc.rr.com> <8cs3ho$l06$1@pegasus.csx.cam.ac.uk>
Message-ID: <slrn8f4be7.1t0.bwinton@tor.dhs.org>

On 10 Apr 2000 08:31:52 GMT, Nick Maclaren wrote:
>In article <3H4I4.709$4w1.17557 at typhoon.nyroc.rr.com>,
>Darrell <darrell at dorb.com> wrote:
>>> x = (a != NULL ? a[i]->weeble : 0) + (b != NULL ? b[i]->wombat : 0)
>>
>>x=0
>>if a != NULL:
>>    x=a[i].weeble
>>
>>if b != NULL:
>>    x=x+b[i].wombat
>>
>>Isn't this more clear?  In fact isn't not much if any longer than the
>>"C" version when you count the parens used.
>
>No.  Firstly, it relies on an artifact of my example (i.e. that it is
>in an assignment, rather than a test condition or function call) and,
>secondly, the more lines that code takes the harder it is to see the
>function in its entirety.

Now perhaps it's just because I've been forced to switch brace styles at
work, but this comment kind of rubs me the wrong way.  Even on the linux
console (possibly the least configurable text mode I know), you can get
80 lines on the screen at a time.  It's my (not so) humble opinion that
if your functions are 80 lines long, then you're probably doing
something wrong.  Even if you have to break your single line up into
four or five, surely it's not causing your functions to run off the end
of the screen, or if it is, perhaps it's time to break your functions up
a little.

And in my case, I far prefer the second version.  One concept per line
seems like a decent metric to go by, and "x = (a?b:c) + (d?e:f)" seems
like a lot more than one concept to me...

Finally, think of all the whitespace you've saved by not having to use
"}"s all over the place.  ;)

Later,
Blake.
-- 
2:54pm up 12 days, 15:28, 1 user, load average: 1.08, 1.02, 1.01


From effbot at telia.com  Mon Apr 24 12:35:04 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Mon, 24 Apr 2000 16:35:04 GMT
Subject: Simple list.append() question
References: <meh9-5ED8CD.11192424042000@news.cit.cornell.edu>
Message-ID: <YI_M4.1471$wYl.192602112@newsb.telia.net>

Matthew Hirsch <meh9 at cornell.edu> wrote:
> Hi All,
>
> I have this:
> >>> entire_list=[[]]*3
> >>> entire_list
> [[], [], []]
> >>> entire_list[0].append(1.23)
> >>> entire_list
> [[1.23], [1.23], [1.23]]
>
> I only want to append 1.23 to entire_list[0].  Why did it also append it
> to entire_list[1] and entire_list[2]?

entire_list contains three references to the
same list.

for more info, see:
http://www.python.org/doc/FAQ.html#4.50

</F>




From just at letterror.com  Thu Apr 20 04:12:58 2000
From: just at letterror.com (Just van Rossum)
Date: Thu, 20 Apr 2000 09:12:58 +0100
Subject: __del__ problem - would adopting Garbage Collection fix this?
In-Reply-To: <20000420003749.A31345@acs.ucalgary.ca>
References: <l03102805b52461abc14d@[193.78.237.146]>; from
 just@letterror.com on Thu, Apr 20, 2000 at 08:35:13AM +0100
 <aw2L4.8907$HG1.245743@nnrp1.uunet.ca>
 <7ojL4.9424$HG1.254543@nnrp1.uunet.ca>
 <Z%oL4.9643$HG1.257533@nnrp1.uunet.ca>
 <m31z41iyhg.fsf@atrus.jesus.cam.ac.uk>
 <38FE7F1E.168E914A@cosc.canterbury.ac.nz><slrn8ft35a.uqs.nascheme@cranky.a
 <B5240D22.4AFD%abrahams@mediaone.net>
 <slrn8ft7qc.u9n.nascheme@cranky.arctrix.com>
 <l03102805b52461abc14d@[193.78.237.146]>
Message-ID: <l03102806b5246c57437f@[193.78.237.146]>

At 12:37 AM -0600 20-04-2000, Neil Schemenauer wrote:
>On Thu, Apr 20, 2000 at 08:35:13AM +0100, Just van Rossum wrote:
>> Maybe you're right, but to me it feels this is in the exact
>> same category as a __del__ exception, so ignoring & collecting
>> seems more logical to me as a default.
>
>It shouldn't be hard to change.  I will wait until I get some
>more feedback though.

Sure, I don't expect you to take my word for it ;-)

There's an additional argument, though: symmetry with JPython.

Just





From hamish_lawson at yahoo.co.uk  Fri Apr  7 10:02:04 2000
From: hamish_lawson at yahoo.co.uk (Hamish Lawson)
Date: Fri, 07 Apr 2000 07:02:04 -0700
Subject: PPython
References: <38ede548.353776233@news.oh.verio.com>
Message-ID: <02d80e56.59818fe0@usw-ex0108-063.remarq.com>

You may instead be interested in the mxCGIPython tool, which
helps you to install Python on your ISP when your ISP
won't/can't.

http://starship.skyport.net/~lemburg/mxCGIPython.html


* Sent from RemarQ http://www.remarq.com The Internet's Discussion Network *
The fastest and easiest way to search and participate in Usenet - Free!



From mikael at isy.liu.se  Fri Apr 14 15:41:48 2000
From: mikael at isy.liu.se (Mikael Olofsson)
Date: Fri, 14 Apr 2000 21:41:48 +0200 (MET DST)
Subject: How do I undo an import??
In-Reply-To: <bkKJ4.190780$Hq3.4442202@news2.rdc1.on.home.com>
Message-ID: <XFMail.000414214148.mikael@isy.liu.se>

On 14-Apr-00 Tom wrote:
 >  After importing NetCfg I type:
 >  
 > >>> del sys.modules['NetCfg']
 >  
 >  Python responds as follows:
 >  
 >  Traceback (innermost last):
 >    File "<stdin>", line 1, in ?
 >  NameError: sys
 >  
 >  Any ideas??

Have you imported sys? I don't think so.

/Mikael

-----------------------------------------------------------------------
E-Mail:  Mikael Olofsson <mikael at isy.liu.se>
WWW:     http://www.dtr.isy.liu.se/dtr/staff/mikael
Phone:   +46 - (0)13 - 28 1343
Telefax: +46 - (0)13 - 28 1339
Date:    14-Apr-00
Time:    21:41:15

This message was sent by XF-Mail.
-----------------------------------------------------------------------


From effbot at telia.com  Mon Apr  3 06:59:10 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Mon, 03 Apr 2000 10:59:10 GMT
Subject: Tkinter:  "Get root for a widget" is private?
References: <20000331112048.A1257432@vislab.epa.gov><QdEF4.892$n68.190327808@newsb.telia.net> <m3snx3n1x0.fsf@devel.no.rwd.dinet><20000331112048.A1257432@vislab.epa.gov><QdEF4.892$n68.190327808@newsb.telia.net> <20000403064335.A1560771@vislab.epa.gov>
Message-ID: <2Q_F4.1092$n68.187097600@newsb.telia.net>

Randall Hopper <aa8vb at yahoo.com> wrote:
> Given this it seems we should have a public root() method.  Could we
> add a:
>
>     root = _root
>
> definition to Tkinter's Misc class, to make this a public feature?

the question is why?

if you're only out to get an interpreter handle (e.g. to
instantiate a variable or image object), any widget will
do.

if you wish to find the original application window, I'd say
it's much cleaner if you keep track of that by yourself

(IMO and IME, a good GUI design has one application level
object, which should *not* be a subclass of Tk).

so why, exactly, do you need the root?

</F>




From che at debian.org  Wed Apr 19 17:18:55 2000
From: che at debian.org (Ben Gertzfield)
Date: 19 Apr 2000 14:18:55 -0700
Subject: [Announce] CDDB 1.0 Python module released!
Message-ID: <ytt8zy923i8.fsf@gilgamesh.cse.ucsc.edu>

Hi everyone! I've received several patches for the CDDB module and
lots of good feedback since the original 0.5 release, and so now it's
time for a new version, 1.0!

You can fetch it from the usual place:

http://csl.cse.ucsc.edu/~ben/python/

CDDB.py provides an easy way for Python programs to fetch track and
disc information on audio CDs. This information is acquired from CDDB,
a very large online database of track listings and other information
on audio CDs. Included is a C extension module to enable Python to
read track listings from audio CDs under Linux, FreeBSD, Solaris, and
Win32. The interface to this extension module is portable and is
intended to be ported to other OSes easily.

Brief summary of changes in the 1.0 release:

This release includes a cdrom module that builds under Linux, Solaris,
and FreeBSD, along with a Win32 cdrom module!

There were several bugs in the old code related to multiple match
returns from CDDB

Please let me know if you find any bugs! :)

Thanks!

-- 
Brought to you by the letters S and O and the number 17.
"A kinkle is a clever method."
Debian GNU/Linux maintainer of Gimp and GTK+ -- http://www.debian.org/


From alander at nortelnetworks.com  Thu Apr 13 10:18:33 2000
From: alander at nortelnetworks.com (Anthony Lander)
Date: 13 Apr 2000 14:18:33 GMT
Subject: Thanks
Message-ID: <8F15632D3anthonylander@47.80.11.122>

Thanks guys. I think I've got it all sorted out now.

   peace,

   -anth


From claird at starbase.neosoft.com  Mon Apr 10 13:41:00 2000
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 10 Apr 2000 12:41:00 -0500
Subject: Migrating Tcl system to Python
References: <Pine.GSO.4.10.10004100709320.26767-100000@sundial>
Message-ID: <5C5D50759ACE4423.57E1CE451FB57C9E.2C33FE35E7FD9271@lp.airnews.net>

In article <Pine.GSO.4.10.10004100709320.26767-100000 at sundial>,
Moshe Zadka  <mzadka at geocities.com> wrote:
			.
			.
			.
>Tcl has exactly two types: strings, and arrays mapping strings to
>strings. The problem is that those are often used to create
>data-structures in strange and unpredictable ways. Hence, no good
			.
			.
			.
1.  There are different ways of counting.  It
    was long traditional to say that Tcl had
    only strings.
2.  That "strange and unpredictable" stuff is
    what Tclers think of as programming.

Mr. Zadka's right:  the prospects for general autotranslation
are bleak (you don't want to know about [trace] or [uplevel]),
but simple marshalling schemes generally work quite well.
-- 

Cameron Laird <claird at NeoSoft.com>
Business:  http://www.Phaseit.net
Personal:  http://starbase.neosoft.com/~claird/home.html


From gmcm at hypernet.com  Thu Apr 20 08:14:18 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Thu, 20 Apr 2000 08:14:18 -0400
Subject: Confused by List behavior
In-Reply-To: <8dmqsd$6m0$1@nnrp1.deja.com>
Message-ID: <1255889636-29796319@hypernet.com>

ejbaker_seekonk at my-deja.com wrote:

> As a python newbie, I entered this script to explore list processing
> behaviours.
> 
> yy = dir()
> for element in yy:
>   print element
> 
> for element in yy:
>   print dir(element)
> 
> The first 'for' loop works as expected. It returns,
> ['__builtins__', '__doc__', '__name__'].
> 
> The second loop, however, returns three empty lists. For the life of
> me, I can't understand that. Becauese, if I type 'dir(__builtins__)',
> in a script, I get all of the elements of __builtins__.

You're confused by dictionaries, not lists. dir() returns the 
names of the objects (that is, a list of strings), not the objects 
themselves. 

One way:
  for element in dir():
    print dir(eval(element))

More sensibly:
  for key, val in vars().items():
    print key, val, dir(val)

- Gordon



From guido at python.org  Thu Apr  6 17:30:21 2000
From: guido at python.org (Guido van Rossum)
Date: Thu, 06 Apr 2000 17:30:21 -0400
Subject: [Python-Dev] Round Bug in Python 1.6?
In-Reply-To: Your message of "Thu, 06 Apr 2000 23:22:30 +0200."
             <38ED0016.E1C4A26C@tismer.com> 
References: <38ED0016.E1C4A26C@tismer.com> 
Message-ID: <200004062130.RAA26273@eric.cnri.reston.va.us>

> asa side effect, I happened to observe the following rounding bug.
> It happens in Stackless Python, which is built against the
> pre-unicode CVS branch.
> 
> Is this changed for 1.6, or might it be my bug?
> 
> D:\python\spc>python
> Python 1.5.42+ (#0, Mar 29 2000, 20:23:26) [MSC 32 bit (Intel)] on win32
> Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
> >>> round(3.1415926585, 4)
> 3.1415999999999999
> >>> ^Z
> 
> D:\python>python
> Python 1.5.2 (#0, Apr 13 1999, 10:51:12) [MSC 32 bit (Intel)] on win32
> Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
> >>> round(3.1415926585, 4)
> 3.1416
> >>> ^Z

This is because repr() now uses full precision for floating point
numbers.  round() does what it can, but 3.1416 just can't be
represented exactly, and "%.17g" gives 3.1415999999999999.

This is definitely the right thing to do for repr() -- ask Tim.

However, it may be time to switch so that "immediate expression"
values are printed as str() instead of as repr()...

--Guido van Rossum (home page: http://www.python.org/~guido/)



From gresham at mediavisual.com  Wed Apr 12 06:14:40 2000
From: gresham at mediavisual.com (Paul Gresham)
Date: Wed, 12 Apr 2000 18:14:40 +0800
Subject: Classes, Inheritance - Stupid lazy question
References: <8d1cfh$hme7@imsp212.netvigator.com> <874s9766r6.fsf@ifm.uni-kiel.de> <m3itxnis32.fsf@atrus.jesus.cam.ac.uk> <87zoqz4p9n.fsf@ifm.uni-kiel.de>
Message-ID: <8d1iak$r1u2@imsp212.netvigator.com>

"Janko Hauser" <jhauser at ifm.uni-kiel.de> wrote in message
news:87zoqz4p9n.fsf at ifm.uni-kiel.de...
> Michael Hudson <mwh21 at cam.ac.uk> writes:
>
> > Nononono; think about what happens if you do this:
> >
> > class bob(paul):
> >    pass
> >
> > bob("adam").fired()
> >
> > What *I* think you want is:
> >
> > class paul(lazy):
> >     def __init__(self, name):
> >         lazy.__init__(self,name)
> >         self.april_pay = 0
> >
> >     def fired(self):
> >         print "Here's a box, collect the things from your desk"
> >         lazy.fired(self)
> >
> > After all, if you're going to inherit from a class, you must have
> > access to it, right?
> >
>
> I see, this is clearly better. But I do not understand, where the
> class/instance?? of lazy.__init__() is stored.

In this way the lazy class does not have an instance, but we can still call
its member functions, as if they were a normal function. By passing them our
instance of self, then our self will also be initialised. In the above
example we can do this without risk, as we are inheriting from lazy, so we
have all attributes.

>
> Some more slaps on the head, please :-)
>
> __Janko
>
> --
>   Institut fuer Meereskunde             phone: 49-431-597 3989
>   Dept. Theoretical Oceanography        fax  : 49-431-565876
>   Duesternbrooker Weg 20                email: jhauser at ifm.uni-kiel.de
>   24105 Kiel, Germany




From hamish_lawson at yahoo.co.uk  Sun Apr 23 12:50:27 2000
From: hamish_lawson at yahoo.co.uk (Hamish Lawson)
Date: Sun, 23 Apr 2000 09:50:27 -0700
Subject: Overloading the = operator?
References: Jerome Chan's message of "Sun, 23 Apr 2000 00:14:02 GMT" <1255667027-43185679@hypernet.com>
Message-ID: <089fe23e.38bfd6ac@usw-ex0103-019.remarq.com>

Jerome

There seem to be two issues here: copying the contents of one
object to another and adding two objects together. I'll use a
cut-down version of your example.

    class Base:
        def __init__(self, q):
            self.list = q
        def __add__(self,other):
            return self.list + other.list

    x = Base([1,2])
    y = Base([2,4,3])
    z = Base([])
    z = x + y

I'm guessing that your coming at this from a C++ viewpoint, so
I'll try to relate the discussion to C++ concepts.

Let's deal with copying objects first. Whereas in C++ assignment
is a copying operation, you have learnt by now that in Python it
is a name binding operation. Thus the statement z = x doesn't
copy the contents of x over to z, but instead makes the name z
be another reference to the object that x refers to. If we want
z to have the same value as x but be an independent object, then
we first need to make a new object that is a clone of x and then
assign z to this new object. We can clone objects using the copy
module; we'll also define a __copy__() method for the Base class:

    class Base:
        def __init__(self, q):
            self.list = q
        def __add__(self,other):
            return self.list + other.list
        def __copy__(self):
            copiedlist = self.list[:]   # make a copy of the list
            return Base(copiedlist)

    import copy
    x = Base([1,2])
    z = copy.copy(x)   # calls x.__copy__()

In Python as in C++ the question of shallow and deep copying
crops up. Our __copy__() method has been defined so that the
cloned object gets its own copy of the list attribute (see the
copy module for a discussion of shallow and deep copying).

However you may not need to bother with copying if your real
concern is with adding objects together, which we'll now
consider. The __add__() method as currently defined returns a
list, but it should really return a new Base object (this is in
fact what an overloaded operator+ would be expected to do in C++
too).

    class Base:
        def __init__(self, q):
            self.list = q
        def __add__(self, other):
            return Base(self.list + other.list)   # new Base obj
        def __copy__(self):
            copiedlist = self.list[:]
            return Base(copiedlist)

    x = Base([1,2])
    y = Base([2,4,3])
    z = x + y

Since x + y produces a *new* Base object, this can simply be
assigned to z without the need to make a copy.

Hamish Lawson

* Sent from RemarQ http://www.remarq.com The Internet's Discussion Network *
The fastest and easiest way to search and participate in Usenet - Free!



From nascheme at enme.ucalgary.ca  Wed Apr 19 11:44:50 2000
From: nascheme at enme.ucalgary.ca (Neil Schemenauer)
Date: Wed, 19 Apr 2000 15:44:50 GMT
Subject: Python 1.6a2 crashes in test_fork1 on my Mandrake Linux 7 machine
References: <20000413010344.A13931@owl.rhein-zeitung.de> <38F658C6.66AD9CA5@clark.net> <20000414154718.A15664@xs4all.nl> <20000414190308.A30502@owl.rhein-zeitung.de> <38F91D1B.3D663B0C@stanford.edu> <38FD72B1.DBB2F731@stanford.edu>
Message-ID: <slrn8frl4s.fte.nascheme@cranky.arctrix.com>

Chad Netzer <cnetzer at stanford.edu> wrote:
>Neil Schemenauer sent me the following code, which demonstrates
>a problem quite clearly. It locks up on me in dual-CPU mode, but
>works fine in single-CPU mode.

Interesting.  My machine is a single CPU and still locks up with
that code.

    Neil

-- 
"A designer knows he has achieved perfection not when there is nothing
left to add, but when there is nothing left to take away."
        -- Antoine de Saint-Exuper


From bgdarnel at my-deja.com  Wed Apr 26 22:26:24 2000
From: bgdarnel at my-deja.com (Ben Darnell)
Date: 27 Apr 2000 02:26:24 GMT
Subject: Python compiler ? (was Re: Python for drivers?)
References: <3900EDBE.282BC85@dsuper.net> <8dr2pn$ec8$1@newshost.accu.uu.nl> <3906f804@nntp.server.uni-frankfurt.de> <015b01bfaf8b$9cd03930$0500a8c0@secret.pythonware.com> <39076316@nntp.server.uni-frankfurt.de> <8e7sop$5io$1@nnrp1.deja.com> <FtnGu1.Dso@world.std.com>
Message-ID: <slrn8gf9ah.lia.bgdarnel@Sull-04-227.rh.ncsu.edu>

In article <FtnGu1.Dso at world.std.com>, Will Ware wrote:
>gchiaramonte at my-deja.com wrote:
>> How about $100 each? Then we only need 600 users. If his compiler
>> really works, count me in.
>
>That's not a bad price per-person, but what are the chances that he'd
>be willing to open-source it? He seems to regard this as a revenue
>generator, so he'd likely consider the source code (or at least a
>large fraction of it) a trade secret. I'd also be willing to kick in
>$100 for an efficient Python compiler, but only if it were open source.

You might want to post a request on cosource.com, a service which is
designed to do just what you're talking about here - collect aggregate 
payments from the community for open source projects.  You first post the
request and get indications of interest from sponsors then see if any
developer will submit a proposal.

-Ben

-- 
Ben Darnell              bgdarnel at unity.nc su.edu
http://thoughtstream.org



From effbot at pythonware.com  Fri Apr  7 13:59:10 2000
From: effbot at pythonware.com (effbot at pythonware.com)
Date: Fri, 07 Apr 2000 17:59:10 GMT
Subject: Python on Google?
References: <8F0ED22EBjoshtinamecom@209.208.0.7> <8cjfk8$qtlo$1@fido.engr.sgi.com>
Message-ID: <8cl7l3$4uk$1@nnrp1.deja.com>

Paul Jackson wrote:
>
> |> to what extent Python is used on www.google.com?
>
> My vague impression is that quite a bit of the
> original "interesting" code was done in Python, but
> that most of the serious stuff now is being done
> in C or C-like C++, due to the serious performance
> challenges handling their current large and growing
> load.
>
> Perhaps some of the non-critical path stuff is still
> in Python.

iirc, their crawlers have been python-powered from the
very beginning.

if you skip over to their jobs page:

    http://www.google.com/jobs.html

you'll find that they're looking for programmers with
various levels of Python experience.

</F>


Sent via Deja.com http://www.deja.com/
Before you buy.


From dfan at harmonixmusic.com  Fri Apr 28 09:22:08 2000
From: dfan at harmonixmusic.com (Dan Schmidt)
Date: 28 Apr 2000 09:22:08 -0400
Subject: Metaclasses?
References: <Pine.GSO.4.10.10004280810510.8383-100000@sundial>
Message-ID: <wk66t2gy2n.fsf@turangalila.harmonixmusic.com>

Moshe Zadka <moshez at math.huji.ac.il> writes:

| On Thu, 27 Apr 2000, GTE wrote:
| 
| > On the other hand I kinda like Visual Basic's 'with' construct. Consider:
| > def method(self,arg1,arg2):
| >     with self:
| >         .spam = "spam"
| >         .eggs = arg1+arg2
| 
| Here's a way to get almost that effect:
| 
| def method(self, arg1, arg2):
| 	_ = self
| 	_.spam = "spam"
| 	_.egss = arg1 + arg2

I just do this:

def method(_, arg1, arg2):
	_.spam = "spam"
	_.eggs = arg1 + arg2

Will this practice get me kicked out of any Python Users Groups?

-- 
Dan Schmidt | http://www.dfan.org


From Jhlee at Brooks.com  Tue Apr  4 17:25:26 2000
From: Jhlee at Brooks.com (Lee, Jaeho)
Date: Tue, 4 Apr 2000 17:25:26 -0400 
Subject: complete novice question
Message-ID: <857F15D7E3D8D3118D290008C7CF058627071E@mail-naeast1.brooks.com>

> I've run 'configure' and it seems to have worked. Now, according to the
> Python package's README, I should be able to type 'make' for the rest of
> the install. If I don't have a 'make' file, did the configure portion
> not work?

You don't need 'make' file to make. Don't worry. Just type 'make'. It needs
Makefile or makefile, not make. 



From sctheagn at kraken.itc.gu.edu.au  Thu Apr 20 01:30:11 2000
From: sctheagn at kraken.itc.gu.edu.au (Joal Heagney)
Date: Thu, 20 Apr 2000 15:30:11 +1000
Subject: Testing a module's return
Message-ID: <38FE95E3.1A10D5B0@kraken.itc.gu.edu.au>

I've found that I have to define several of my modules such that they
return a string if they fail, rather than 0. (Let's face it, there are a
lot more things that can go wrong than right). Now I originally wanted
to have these modules return the values I was after. eg.

def function_name(arg1, arg2, arg3, .... argN):
    do something with the args, return value

But as a beginner, *grins* I've found returning error strings a little
more important

def function_name(arg1, arg2, arg3, .... argN):
    check for error condition 1:
        return 'Whoops'
    check for error condition 2:
        return 'What are you doing, Dave?'
    check for error condition 3:
        return 'That ain't right, or humanly decent either'
    do something with the args, return something.

I can't sort out, should I return 0 for a success, and give the answer
as a set attribute (Eg. a class variable)? Or is there a much neater way
of doing things? There'd have to be.

I'd appreciate if answers were posted by email as well as newsgroup. It
could take me a while to get back to them. Thank you,

Joal Heagney/AncientHart



From gmcm at hypernet.com  Mon Apr 24 12:35:07 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Mon, 24 Apr 2000 12:35:07 -0400
Subject: ftplib sendcmd
In-Reply-To: <3904722D.12EE9927@bam.com>
Message-ID: <1255528387-51524611@hypernet.com>

Bill Scherer wrote


> I'm having trouble with ftplib's sendcmd method (Python 1.5.2 on
> RH linux 6.1):
> 
> >>> ftp = ftplib.FTP(host, user, password)
> >>> ftp.sendcmd('umask 002')
> Traceback (innermost last):
>   File "<stdin>", line 1, in ?
>   File "/usr/local/lib/python1.5/ftplib.py", line 228, in sendcmd
>     return self.getresp()
>   File "/usr/local/lib/python1.5/ftplib.py", line 201, in getresp
>     raise error_perm, resp
> ftplib.error_perm: 500 'UMASK 002': command not understood.
> >>>
> 
> The umask command does, of course, work using a standard ftp
> client to the same host.
> 
> Other commands, chmod, etc, suffer equal fate.
> 
> Is this not what sendcmd is for? What am I doing wrong?  Thanks!

These are not rcf 959 commands. GNU ftpd will support them 
as SITE UMASK x and SITE CHMOD mod file.


- Gordon



From jon at dgs.monash.edu.au  Fri Apr  7 01:35:40 2000
From: jon at dgs.monash.edu.au (Jonathan Giddy)
Date: Fri, 7 Apr 2000 15:35:40 +1000 (EST)
Subject: Socket Programming Question
In-Reply-To: <aZdH4.95109$AT6.106753@dfw-read.news.verio.net> from "Ray Van Dolson" at Apr 07, 2000 05:01:58 AM
Message-ID: <200004070535.PAA18944@nexus.csse.monash.edu.au>

>def main_server_thread():
>        while 1:
>                (clientsocket, address) = serversocket.accept()
>                ct = Thread(target=client_thread,args=((clientsocket),))
>                ct.run()

You need to call ct.start(), not ct.run().  The start method creates the
new thread to execute the run method..

Jon.



From scarblac-spamtrap at pino.selwerd.nl  Mon Apr 17 12:37:20 2000
From: scarblac-spamtrap at pino.selwerd.nl (Remco Gerlich)
Date: 17 Apr 2000 16:37:20 GMT
Subject: How do I make a constant global variable in Python Please
References: <vCj7OA2LfMfb=znR7yTHOvmiZAkd@4ax.com> <rZGK4.4613$Uq2.3181@news.uswest.net>
Message-ID: <slrn8fmimh.bbo.scarblac-spamtrap@flits104-37.flits.rug.nl>

Jeff Massung wrote in comp.lang.python:
> Along the same lines... is there a way to say a variable must be declared
> before use? I know Perl has something like this. I just want to avoid
> something like this following:

No, simply because there is no way to declare a variable.

> ate = 1
> while ate:
>     # do something here
>     ...
>     are = 1
> 
> This will loop infinitely just because of a typo (that if the above were
> added I would get an error saying that "are" wasn't declared).

It's happened a few times to me, but you notice that sort of error almost
immediately, in my experience.

Note that it would loop too if it said "ate = 1" as you seemed to intend -
you need to test to find errors anyway...

Also, this problem only occurs with assignment. If you try to use a new
variable in an expression, an exception will be raised, because the variable
can't be found.

ate = 1
x = are

gives an exception.


Python has no declarations, and nothing is a 'constant' as in other languages.

-- 
Remco Gerlich,  scarblac at pino.selwerd.nl

   This is no way to be
     Man ought to be free      -- Ted Bundy
       That man should be me


From quinn at zloty.ugcs.caltech.edu  Wed Apr 12 16:22:44 2000
From: quinn at zloty.ugcs.caltech.edu (Quinn Dunkan)
Date: 12 Apr 2000 20:22:44 GMT
Subject: modifying os.environ
Message-ID: <slrn8f9mrj.ven.quinn@zloty.ugcs.caltech.edu>

Is there any way to delete things from the environment in python?  A look at
os.py reveals that _Environ only overrides __setitem__ to putenv.  So how do
you *delete* environment values?  perl does it, and a look at the source
seems to indicate that it doodles with the C environ directly.  So perhaps
python needs a delenv _Environ can use or something, or implement os.environ
in C?

Or perhaps there's a better way to do the (common) thing I'm trying to do:
clean the environment to insure some minimum sane values.  I can't make a copy
of the environ and then pass to execve() since that doesn't help much for
popen and system.  I can set values to '', but that doesn't seem as safe as
just removing them entirely (IFS, LD_LIBRARY_PATH, etc.?)


From akuchlin at mems-exchange.org  Mon Apr 24 11:27:50 2000
From: akuchlin at mems-exchange.org (Andrew M. Kuchling)
Date: 24 Apr 2000 11:27:50 -0400
Subject: The Simple Economics of Open Source
References: <B527E7A5.51AA%raffael@mediaone.net> <3903316C.5D6F1611@tcac.net> <LHEoAJAXWFB5EwF6@jessikat.demon.co.uk> <FtIznM.EH1@world.std.com>
Message-ID: <3dya63y0ww.fsf@amarok.cnri.reston.va.us>

wware at world.std.com (Will Ware) writes:
> very small. Are there other talented artists with distinctive styles
> for whom it wouldn't be very unreasonable for there to be undiscovered
> works appearing?

Probably not.  The Dutch artist Vermeer is sort of the opposite case;
the number of actual Vermeers is slowly being whittled away by
scholarship as paintings are attributed to his students.  I think the
number of original Vermeer paintings is now put at around 30.  A
forger named Han van Meegeren painted several forged Vermeers, some of
which fooled people for many years.  (http://www.mystudios.com/gallery/han/)

This connects to two obsessions of mine.  Robertson Davies's excellent
novel _What's Bred in The Bone_ has a subplot based on van Meegeren.
And Peter Greenaway's film _A Zed and Two Noughts_ features a
character named van Meegeren who's a doctor obsessed with Vermeer, and
lighting design and photography influenced by Vermeer (always lit from
the left, for example.  In the screenplay PG gives a far lower number
of actual Vermeers than any other source: 23.

-- 
A.M. Kuchling			http://starship.python.net/crew/amk/
I sit here for hours. It's like sitting amongst lighthouses, each lighthouse
giving you a bearing on lost spaces of time...
  -- Peter Greenaway, _A Zed and Two Noughts_ (1986)


From m.faassen at vet.uu.nl  Tue Apr 11 09:24:12 2000
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 11 Apr 2000 13:24:12 GMT
Subject: Python paradigms
References: <8cmr70$b0e$1@pegasus.csx.cam.ac.uk> <slrn8f4be7.1t0.bwinton@tor.dhs.org> <8ctc9m$sp2$1@pegasus.csx.cam.ac.uk> <8ctkm5$bko$2@newshost.accu.uu.nl> <8cur12$5pf$1@pegasus.csx.cam.ac.uk>
Message-ID: <8cv91s$q2i$1@newshost.accu.uu.nl>

Nick Maclaren <nmm1 at cus.cam.ac.uk> wrote:
> In article <8ctkm5$bko$2 at newshost.accu.uu.nl>,
> Martijn Faassen <m.faassen at vet.uu.nl> wrote:
>>Nick Maclaren <nmm1 at cus.cam.ac.uk> wrote:
>>[if your function gets longer than a screenfull, it's time to start
>>thinking about splitting it up]
>>
>>> It really is quite amusing to see these same old myths coming back
>>> every few decades :-)
>>
>>You think splitting a function up is a myth? I find it helps 
>>me and my code. It becomes more readable and writeable. I'm genuinely
>>surprised by your assertion this is an 'old myth'.

> The myth is that splitting functions up necessarily improves clarity.
> Sometimes it does; sometimes it doesn't.

It depends on how you do the splitting, I'd still say. I think the
cases are rare that one long complicated function is more readable
than two or three or more *well-chosen* shorter ones.

>>> That particular dogma was very popular back in the late 1960s and
>>> early 1970s, when practical programmers pointed out that it was a
>>> virtually impossibility for a normal person to keep track of 50
>>> trivial functions in their head, but it was quite easy to look
>>> at 60 lines of sequential code.  That didn't get listened to
>>> then, either!
>>
>>Of course if you go about randomly chopping up your function into
>>multiple ones, that's not going to improve matters. The idea is
>>to conceptually split your function into multiple smaller ones
>>that make more sense, are more easy to adapt, are coherent by 
>>themselves, and make the original big function more readable, as
>>by writing all the smaller functions you've hopefully also given
>>them names. It's called refactoring, of course. :)

> The dogma is that there is a maximum size of function that should
> ever be written,

Putting a strict limit on the number of lines is of course silly.

> and that complex functions necessarily split up
> cleanly into smaller sections.

I think they do; perhaps not multiple functions per se, but multiple
classes, definitely. Particularly in a language like Python!

> Sometimes they do; sometimes they
> don't.  I have seen codes where the median size of functions was
> below 5 lines, and you had to keep the specification of over 50 in
> your head just to read the simplest function!

Didn't the functions have readable names then? Of course it's silly
to take all this to ridiculous extremes, but if the naming was
done well you have a lot of contextual information when you're reading
a function which you wouldn't have otherwise.

>>My guidelines are:
>>
>>If your function is long and complicated, too many concepts
>>and abstractions are mingling in the same place. Do a better job
>>of abstracting stuff, and you'll end up with shorter, clearer
>>functions.

> Sometimes.  I have very often done that and backed off, because the
> complex function was clearer.  Another good rule is, if auxiliary
> functions need to access more arguments and artificial global
> variables than they have lines of code, consider whether they would
> be better written inline.

If you run into that, use records or classes. OO is pretty good at 
this; you can still bundle a set of methods into an object. The
object can do complicated things, but its parts are simple.
I don't recall cases where this was impossible, but then again my experience
may be too limited.

>>If your function is long and repetitive, you're not being a lazy
>>enough programmer. Write some smaller functions that generate the
>>stuff, read it from a file or database, whatever.

> If the repetitiveness is not quite regular enough to make that easy,
> you are making unnecessary work for yourself.  Sometimes that is the
> right solution; sometimes it isn't.

The repetitiveness is always regular enough to abstract it into
a bunch of classes, make a dictionaries of lists or whatnot. If there
is no regularity at all repetitiveness isn't the problem.

>>Are you saying these guidelines are based on a myth?

> No - the myth is that they invariably improve things.  Life is not
> that simple.

No, they don't improve anything if you don't think well before you
refactor. But I do believe it's always possible to split up longer
complicated functions into smaller ones (or objects), given enough
thought, and that this makes the code more clear.

>>> My current infliction is a accounting file format, and the the
>>> function to check the input (you DO check your input, don't you?)
>>> needs to perform 100 tests in series.  Yes, I am inventing
>>> trivial functions, but it would be clearer to put more of the
>>> tests inline.
>>
>>Why not split this up into a bunch of validation objects or 
>>something, and then check the input that way? And aren't at least
>>some of these tests similar enough to abstract into a function
>>(such as tests for maximum various lengths of strings). Of course
>>the accounting file format may be so baroque your function just
>>has to be baroque, but that sort of thing *ought* to be rare. :)

> That is, of course, what I am doing.  But designing in enough
> commonality takes a lot longer (and is more complex) than simply
> repeating the very simple tests.  After all, it gains you nothing
> by calling an auxiliary that then has a massive switch statement
> with special code for each of its calls - and some people do write
> code like that!

> In this particular case, there are 50-60 entries, which need
> 40-50 separate test conditions.  By thinking, I can common up
> quite a lot of that code, but there is some which is not amenable
> to that.

Yes, but even that code could probably be split out into some
way so that it's clear that something special, uncommon is happening, 
right? And that would likely be clearer than it just being thrown
in a big function, I think? More work, yes, perhaps. At least more
up-front work.

> If you remember, I asked whether there were any equivalent to a
> couple of very common paradigms, that have been used to clarify
> code for over 30 years.

Obfuscate code at least as often as clarify.

> The answer appears to be "no", so I have
> to use somewhat less satisfactory ones.  This is not a major
> problem - been there, done that, in dozens of languages.

In some cases, perhaps you'll have to use less satisfactory ones. I think
in most cases the alternatives are *more* satisfactory, however.

> What isn't acceptable, however, is putting up with people saying
> that it is heretical to admit that such requirements exist!  Fads
> are one thing, but dogma is harmful.

Fads, dogma, but what about wisdom, though? Python's idea is that ?: and
assignment-in-expression both tend to obfuscate the code. It'd be nice
to have something replacing the assignment-in-expression in the example
in your first post, but I disagree that your ?: construction makes the
code more easy to read than than various alternatives. I can grok C's
(?:), but its main advantage is that it makes code more writeable, not
more readable, in my opinion. Likewise any replacement to enable the
niceness assignment-in-expression allows (in 'if' and 'while' loops) will
*not* be assignment-in-expression but something more readable and 
less general.
 
In general, the pattern in Python is that it avoids several handy but
easy to convolute constructs; for instance we have no C style 
for(;;) loop either, while it's clearly the general case. Instead in
Python you tend to work with more specific idioms like its own 'for' loop, 
'range', and lists. This appears to make the code more readable.

People have been showing you some examples of special case idioms that
in part take over the functionality of (?:). I think the idea is that
such special case idioms help in making the code more readable, and catch
90% or more of the cases.

No 'heretical'. Just trade-offs. Python's idioms for writing Python. The
strategy does have a pay-off: Python code tends to be fairly readable.

Regards,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From thomas at xs4all.net  Tue Apr 11 07:51:03 2000
From: thomas at xs4all.net (Thomas Wouters)
Date: Tue, 11 Apr 2000 13:51:03 +0200
Subject: Thread-safe atomic operation
In-Reply-To: <8cuf0o$p9a$1@news1.kornet.net>; from sadwolf@chollian.net on Tue, Apr 11, 2000 at 03:02:28PM +0900
References: <8cuf0o$p9a$1@news1.kornet.net>
Message-ID: <20000411135103.B28687@xs4all.nl>

On Tue, Apr 11, 2000 at 03:02:28PM +0900, Dong-gweon Oh wrote:

> Is there any documents which describe Python's atomic operation in
> multithreaded environment. I guess Python virtual instruction is atomic.
> But, without definitive documentation in high level Python statements, I
> tend to over-lock my programs.

> Are these statements thread-safe?
> 1. import xxx
> 2. xxx = __import__('xxx')
> 3. reload(xxx)
> 4. dict['xxx'] = sys.modules['xxx']
> 5. dict.has_key('xxx')

Well, Python uses a global interpreter lock. This means that no matter howmany
threads are instantiated, only one of them will be executing Python
code. Inside python code, and in C modules and extensions, however, the C
code itself takes care to allow/disallow multiple threads to run a block of
code.

So, the answer to your question is: Yes, all of them are thread-safe, but
none of them will benifit from threading. Some of them may block and allow
threads, though, so you might get *some* aditional performance if you are
doing something else in another thread, at the same time. You can safely
assume all your python code is always thread-safe, and if you do not bother
with the 'Py_BEGIN_ALLOW_THREADS'/'Py_END_ALLOW_THREADS' C macros, so will
all your C extension code (as long as you dont do anything funky with
cloning threads yourself.)

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From greg at quokka.demon.co.uk  Mon Apr 24 18:40:07 2000
From: greg at quokka.demon.co.uk (greg at quokka.demon.co.uk)
Date: Mon, 24 Apr 2000 22:40:07 GMT
Subject: plugins for browsers?
Message-ID: <3904ccb0.32134281@news.demon.co.uk>

	Are there any plugins or other means of running Python in Web
browsers? I seem to recall that there were a while ago ...

Greg Lewin


From pecora at anvil.nrl.navy.mil  Sun Apr 30 09:55:47 2000
From: pecora at anvil.nrl.navy.mil (Louis M. Pecora)
Date: Sun, 30 Apr 2000 13:55:47 GMT
Subject: Do I always have to write "self." ?
References: <270420001706386648%pecora@anvil.nrl.navy.mil> <slrn8ghc6t.f01.scarblac-spamtrap@flits104-37.flits.rug.nl> <280420000746597329%pecora@anvil.nrl.navy.mil> <Ftq7Fo.z7@world.std.com>
Message-ID: <300420000955487866%pecora@anvil.nrl.navy.mil>

[[ This message was both posted and mailed: see
   the "To," "Cc," and "Newsgroups" headers for details. ]]

In article <Ftq7Fo.z7 at world.std.com>, Will Ware <wware at world.std.com>
wrote:

> Louis M. Pecora (pecora at anvil.nrl.navy.mil) wrote:
> > I'm sure some way of declaring variables 'local' could be done in the
> > class defintion, but I doubt Guido would do that now.
> 
> It wouldn't be too hard to write a preprocessor that would do this
> for you. You might put tags in your code such as
> #Locals: foo bar xyzzy
> and
> #Endlocals

[cut]

> The choice to use the word 'self' is not imposed by Python, it's just
> a convention for readability. You can use 's' instead if you wish.

[cut]

Thanks for both ideas and pointers.  Writing a preprocessor is out of
the question for me.  I doubt I'm that good or even have the time.  I
like the second suggestion.  Thanks, again.


From bboard at parc.xerox.com  Tue Apr 18 19:43:08 2000
From: bboard at parc.xerox.com (Mail Archiving Daemon)
Date: Tue, 18 Apr 2000 16:43:08 PDT
Subject: does Python actually handle SIGTERM and call sys.exitfunc?
Message-ID: <00Apr18.164300pdt."3438"@watson.parc.xerox.com>

I'm looking at the documentation for the signal module in the 1.5.2
Library document, which says:

     Python installs a small number of signal handlers by default:
     `SIGPIPE' is ignored (so write errors on pipes and sockets can be
     reported as ordinary Python exceptions), `SIGINT' is translated
     into a `KeyboardInterrupt' exception, and `SIGTERM' is caught so
     that necessary cleanup (especially `sys.exitfunc') can be
     performed before actually terminating.  All of these can be
     overridden.

I can find no evidence that SIGTERM is ever caught anywhere in the
interpreter.  Can someone point me to the place where this is done?

Also, when I try the following small program:


   import sys, time, signal

   def writeit (msg):
      sys.stdout.write("exiting (" + msg + ")...\n")
      sys.stdout.flush()

   sys.exitfunc = lambda x="foo": writeit(x)
   time.sleep(60)


running it thusly:

   /tmp 359 % python test.py &
   [2] 11067
   /tmp 360 % kill 11067
   /tmp 361 % 
   [2]    Terminated           python test.py
   /tmp 361 % 

I see no output saying "exiting (foo)...".

This might be a documentation bug.

Bill




From dgreen at kryten.kitenet.net  Wed Apr  5 13:51:04 2000
From: dgreen at kryten.kitenet.net (Dan Green)
Date: Wed, 05 Apr 2000 17:51:04 GMT
Subject: ASP web-type stuff
References: <RZyG4.3061$9o.5812531@news.magma.ca> <8cfea6$71$1@gxsn.com>
Message-ID: <c2LG4.3228$9o.6039280@news.magma.ca>

Ah. I suppose I should have made myself more clear. What I meant was that
I was interested in writing a CGI/FastCGI type script so that the same
code could be used under Apache/Netscape Enterprise/insert your favorite
unix web server here. I already have a prototype working under Apache with
a CGI redirect. I'd like to write interfaces to httpdapy and PyApache as
well.

In article <8cfea6$71$1 at gxsn.com>, Richard Chamberlain wrote:
>Hi Dan,
>
>ASP is not language dependent, most of the stuff you'll see is written in
>VBScript or JScript but you can actually configure it to use other
>languages.
>
>See http://starship.python.net/crew/pirx/asp/py_asp.html to show you how to
>use Python with ASP.
>
>Richard
>
>"Dan Green" <dgreen at kryten.kitenet.net> wrote in message
>news:RZyG4.3061$9o.5812531 at news.magma.ca...
>> Ahoy, folks.
>> I was looking over things like Embperl and Apache::ASP and liked the
>> idea. I would like the idea more if I could do with python. So I had a
>> couple of questions:
>>
>> Is there such a thing?
>> Is there interest in such a thing?
>> Does anbody want to hear about my 75 lines of code hack that basically
>does
>> the trick, but needs a lot of work?
>>
>> If nobody cares at all, I'll just shut up and go home. :)
>>
>> For those of you who have no idea what I'm talking about, it's basically
>an
>> alternative to CGI that allows you to embed the code right into the HTML.
>>
>> something like this:
>>
>> <HTML>
>> <BODY>
>> <%  for i in range(5): %>
>> Loop item <%= i %><BR>
>> </BODY>
>> </HTML>
>>
>> would look like this when called from a browser:
>>
>> <HTML>
>> <BODY>
>> Loop item 1<BR>
>> Loop item 2<BR>
>> Loop item 3<BR>
>> Loop item 4<BR>
>> Loop item 5<BR>
>> </BODY>
>> </HTML>
>>
>> Let me know. If you're interested, please e-mail me.
>>
>> --
>> Dan Green           | "So What? It's your problem to learn to live with,
>> Developer           |  destroy us, or make us saints."
>> Rewind Technologies |  -- Ministry, "So What?"
>
>


-- 
Dan Green           | "So What? It's your problem to learn to live with,
Developer           |  destroy us, or make us saints."
Rewind Technologies |  -- Ministry, "So What?"


From python-list at teleo.net  Mon Apr 10 14:59:56 2000
From: python-list at teleo.net (Patrick Phalen)
Date: Mon, 10 Apr 2000 11:59:56 -0700
Subject: Problem running Python script as cron job
In-Reply-To: <wrd-1E3688.12433910042000@news.awenet.com>
References: <wrd-0AEDDA.12282507042000@news.awenet.com> <0004071555090M.01234@quadra.teleo.net> <wrd-1E3688.12433910042000@news.awenet.com>
Message-ID: <0004101202210J.00831@quadra.teleo.net>

[William R. Dickson, on Mon, 10 Apr 2000]
:: In article <0004071555090M.01234 at quadra.teleo.net>, Patrick Phalen 
:: <python-list at teleo.net> wrote:
:: 
:: > [William R. Dickson, on Fri, 07 Apr 2000]
:: > :: Anybody know of any issues with running python scripts from Cron?
:: > 
:: > No. I've had multiple Python scripts running every one, two, five and
:: > ten minutes around the clock for years on multiple servers without
:: > problems.
:: > 
:: > I'd suspect you might have a bug lurking.
:: 
:: I think so too, but I can't find it, because the script runs flawlessly 
:: every time I run it by hand, and the buffer never flushes the output to 
:: the log when it's run by Cron.

Are you "running by hand" as the same user as your script? Maybe you
have a permissions problem, assuming the script needs to write to a
file or whatever.



From fredrik at pythonware.com  Thu Apr 27 05:44:16 2000
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 27 Apr 2000 11:44:16 +0200
Subject: xmlrpclib problem
References: <8e8591$2ro$0@216.39.162.232>
Message-ID: <00c801bfb02e$091d69d0$0500a8c0@secret.pythonware.com>

Michael Esveldt <dante at oz.net> wrote:
> I'm working with xmlrpclib to talk to Userland's Manila application and 
> I'm having a problems sending html via xmlrpc. Here's the code I'm 
> working with
> 
> mysite = xmlrpclib.Server("http://mysite.editthispage.com") 
> mysite.manila.message.set(username, password, sitename, msgNum, "my 
> subject", "body containing html", "text/x-outline-tabbed", {}, 
> {"name":"pikeRenderer", "flRenderOnEntry":"true"})
> 
> I've tried changing "text/x-outline-tabbed" to "text/plain" and 
> "text/html", none of these have any effect. What's happening is that my 
> "body containing html" is coming out on the other side with &lt; and 
> &gt; instead of actual tags. Obviously this results in displaying html 
> instead of rendering it.
> 
> Is this a problem with xmlrpclib, how I'm using it, or with the server 
> decoding my xmlrpc messages?

it's a bug in Manila:
http://discuss.userland.com/msgReader$16672

here's a workaround for xmlrpclib.py:
http://zopefish.weblogs.com/discuss/msgReader$19

</F>




From mhammond at skippinet.com.au  Sat Apr 22 10:02:21 2000
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Sat, 22 Apr 2000 14:02:21 GMT
Subject: code not true?
References: <8ds3ku$q0l$1@nnrp1.deja.com> <3901aa0f.0@news.proweb.co.uk> <M3iM4.518$wJ1.17829@ptah.visi.com>
Message-ID: <NhiM4.1704$v85.10013@news-server.bigpond.net.au>

"Grant Edwards" <nobody at nowhere.nohow> wrote in message
news:M3iM4.518$wJ1.17829 at ptah.visi.com...

> This seems to be something that we were all told as freshmen
> but stumble over later anyway.  One wonders if a safe language
> intended for use by beginners should even have an equality
> operator for floating point objects.

Of all the solutions to this problem, this does strike me as one of the
best - simply remove equality testing al together for floating point:

>>> 1.0==1.0
Traceback (whatever it says today)
File <etc etc> in ?
Floating points numbers cant be compared.
In fact, life would be better if you could pretend they didnt even exist.
Dont even ask.  You were warned.


Mark.




From effbot at telia.com  Mon Apr 24 03:51:39 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Mon, 24 Apr 2000 07:51:39 GMT
Subject: regulare expression searches and newlines
References: <8e0oco$k75$1@nnrp1.deja.com>
Message-ID: <f2TM4.3782$Za1.55405@newsc.telia.net>

schmitthead at my-deja.com wrote:
> I was doing a search like this:
>
> searchobj = re.compile( r'^foobarbaz' )
>
> where I was looking for something at the beginning of a line.  This
> works well for native text files, but not when I was searching through
> Linux files running on NT or NT files on Linux or some other variation
> because of the newline difference, I guess.  Is there any way to handle
> the general case?

there's nothing wrong with your regular expression.

I suggest printing the strings you're trying to match against,
to see if you can figure out what's wrong.

if that doesn't help, post the code.

</F>




From claird at starbase.neosoft.com  Tue Apr  4 15:37:39 2000
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 4 Apr 2000 14:37:39 -0500
Subject: What is python's language level?
References: <38EA1B8F.A797A130@lmco.com> <seqG4.3372$HG1.98334@nnrp1.uunet.ca>
Message-ID: <D14AF0462357D8F5.ECF3E63D386E365A.638CE468B4CA9C9D@lp.airnews.net>

In article <seqG4.3372$HG1.98334 at nnrp1.uunet.ca>,
Warren Postma <embed at geocities.com> wrote:
			.
			.
			.
>2. What's a function point?  Who decides what a "function point" is, and
>says that it applies across all the various kinds of people who write
>software as part of their professional duties?  How many lines of COBOL does
			.
			.
			.
The International Function Point Users Group
<URL:http://www.ifpug.org/> is made up of the
kinds of people who decide what a function
point is.  They take it quite seriously.
-- 

Cameron Laird <claird at NeoSoft.com>
Business:  http://www.Phaseit.net
Personal:  http://starbase.neosoft.com/~claird/home.html


From m69ph83q at students.su.se  Sun Apr  2 09:48:32 2000
From: m69ph83q at students.su.se (Per Hogberg)
Date: 2 Apr 2000 13:48:32 GMT
Subject: Some bugs in Pythonwin b.128
References: <Pine.OSF.4.10.10003281852310.6437-100000@prfdec.natur.cuni.cz> <_NaE4.6119$V56.24967@news-server.bigpond.net.au>
Message-ID: <8F0AAC6BEm69ph83qstudentssuse@127.0.0.1>

mhammond at skippinet.com.au (Mark Hammond) wrote in 
<_NaE4.6119$V56.24967 at news-server.bigpond.net.au>:

>"Michal Bozon" <bozon at natur.cuni.cz> wrote in message
>news:Pine.OSF.4.10.10003281852310.6437-100000 at prfdec.natur.cuni.c
>z...
>> known from IDLE 0.5, I mean when I write a function name and
>> the beginning bracket, it appears a tip "window" with list of
>>arguments
>> of function and the docstring.

>
>This works for every one else :-)


Actually, if you're one of those cursed few with a non-US/Brit./Australian 
keyboard you may have to tinker a bit. The events for parenthesis in the 
config file are wired to the Shift-9 & 0 keys. My keyboard uses Shift-8 & 
9. 

Try changing these settings in your config file:
    Shift-9           = <<paren-open>>
    Shift-0           = <<paren-close>>


BTW, I never managed to put my changes in a separate file and use fallback 
to the default file. Not even idle.cfg in the distribution worked this way. 
I ended up copying the whole default.cfg file. I didn't try hard so this 
just might work for everyone else :-)

  Per.


From m.faassen at vet.uu.nl  Fri Apr 14 15:27:33 2000
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 14 Apr 2000 19:27:33 GMT
Subject: Complaint Dept: My shoes!
References: <Pine.GSO.4.10.10004141925330.7167-100000@sundial>
Message-ID: <8d7rf5$odi$1@newshost.accu.uu.nl>

Moshe Zadka <moshez at math.huji.ac.il> wrote:
> On Thu, 13 Apr 2000 musingattheruins at my-deja.com wrote:

>> If shoe is None then len(shoe) raises an exception.  Would be nice if
>> len(None) was 0 and NOT raise an exeption.

> Huh? Why not have len(None) 4? Or -1?
> If you want an empty string, use an empty string. The good thing
> about None is that it supports so few operations, functions not expecting
> it would break right away, instead of having a hidden bugs.

I agree completely. You DON'T want len(None) to not raise an exception.
It's good that it does. I agree with all posters. Don't do it. Don't.

...

And now I'm going to tell you how to make len(None) return 0. This
is an implementation detail. It's a *bug* that ought to be fixed.

Put this in the module where you want len(None) to be 0:

None = ""

Now it works as you want. But you *don't* want to want this. If you go
down on this path, the Python Secret Underground will come and destroy you.
Perhaps retroactively wipe you out. The only good thing that may come
out of this is Guido going back in his time machine to fix this. Just don't
do it.

This-is-not-what-G*d*-indented!-ly yours,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From effbot at telia.com  Tue Apr 18 12:45:20 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Tue, 18 Apr 2000 16:45:20 GMT
Subject: Lists
References: <A63AFB20111ED311AC5500A0C96B7BF5015F0F2C@orsmsx36.jf.intel.com>
Message-ID: <Ai0L4.467$wYl.186956288@newsb.telia.net>

Daley, MarkX <markx.daley at intel.com> wrote:
> This feels like a real newbie question, but here is the code that is
causing
> my question:
>
> # Learning exceptions
>
> def test():
>    a = [-5, -4, -3, -2, -1, 0, 1, 2, 3, 4, 5]
>    print a
>    b = 5
>    for item in a:
>        try:
>            print b / a[item]
>        except ZeroDivisionError:
>            pass
>
> Here is the output of this code under IDLE:
>
> [-5, -4, -3, -2, -1, 0, 1, 2, 3, 4, 5]
> 5
> 2
> 1
> 1
> 1
> -1
> -2
> -2
> -3
> -5
>
> Why is the list being processed in reverse?  The exception works fine, but
> the math is backwards.  I don't recall this happening before.

note that you divide by a[item], not by item.

try printing item and a[item] in the loop and see if you
can figure out what's going on...

(where did you find this script, btw?  "100 obfuscated
python scripts for beginners?" ;-)

</F>




From donb at init.com  Tue Apr 18 09:45:51 2000
From: donb at init.com (Donald Beaudry)
Date: Tue, 18 Apr 2000 09:45:51 -0400
Subject: Another question 
References: <dVFK4.41$vr2.5817@news.uswest.net> <38FB2EEB.A32AF88C@bam.com> <38FB339F.604364AD@bam.com> <200004171605.MAA04929@zippy.init.com> <38FB7610.80EAF965@ichips.intel.com> <200004172120.RAA11102@zippy.init.com> <jWNK4.432$vr2.32822@news.uswest.net>
Message-ID: <200004181345.JAA24254@zippy.init.com>

"Jeff Massung" <jmassung at magpiesystems.com> wrote,
> 
> Donald Beaudry wrote in message <200004172120.RAA11102 at zippy.init.com>...
> >and this one is better still (at least on my machine):
> >
> >  def do_it():
> >      from random import randint
> >      r = [0,] * 100000
> >      for i in xrange(100000):
> >         r[i] = randint(1,10)
> >  do_it()
> >
> 
> With the "from/import" being local, does the namespace go away after the
> function is done?

No, only the local binding to it goes away.

> How fast is importing (say I called do_it() 10000 times - then how
> efficient is your function?)

A from/import, or just an import for that matter, can be slow the
first time but very fast every time thereafter.  Modules (under normal
circumstances) only get loaded once and that's what takes the time.
Regardless, if I knew I was going to be calling do_it() many times in
a performance critical application I would not put the import inside
the function.  In that case, I would write:

  import random

  def do_it(randint = random.randint):
      r = [0,] * 100000
      for i in xrange(100000):
         r[i] = randint(1,10)

In my original example I put the from/import inside the function
because I felt that it was a less obscure way to get randint into the
function's local name space (and I knew it would only get called
once).  Getting randint into the local name space is a big win for two
reasons.  First, the dictionary look-up in the random name space is
eliminated.  Second, variable references inside functions are
optimised to array references thus they are not dictionary look-ups as
they are when outside a function.

--
Donald Beaudry                                     Ab Initio Software Corp.
                                                   201 Spring Street
donb at init.com                                      Lexington, MA 02421
                  ...So much code, so little time...





From python-list at teleo.net  Mon Apr  3 02:30:12 2000
From: python-list at teleo.net (Patrick Phalen)
Date: Sun, 2 Apr 2000 23:30:12 -0700
Subject: embeddable Python web server
In-Reply-To: <VGWF4.1395$wo4.236156@paloalto-snr1.gtei.net>
References: <VGWF4.1395$wo4.236156@paloalto-snr1.gtei.net>
Message-ID: <00040223335404.04535@quadra.teleo.net>

[Brian Schmidt, on Sun, 02 Apr 2000]
:: Does anyone know of a webserver written in Python that is embeddable into a
:: Python application, something similar to tclhttpd available for Tcl?

There's SimpleHTTPServer.py (in the distribution) and also (free for
non-commercial use) Medusa, available from http://www.nightmare.com.



From bingdo19 at earthlink.net  Tue Apr 25 12:06:54 2000
From: bingdo19 at earthlink.net (Bob Langdon)
Date: Tue, 25 Apr 2000 16:06:54 GMT
Subject: €€Newbie Needs Help -- Name and Key errors
Message-ID: <bingdo19-4B8F57.12162225042000@news-server.twcny.rr.com>

Hi,

New to Python, and moving a site for a client. The backend sas developed 
by another developer in Python.  I've i9nstalled Python1.5.2 in my home 
direcpory and tested it. The interpreter seems to functioning properly.

However, some paths seem to be defined incorrectly.

When running a script "getresume.py" for example, 

>>try:
   import sys
   sys.stderr = sys.stdout
   sys.path.insert(0,'.')
   sys.path.append('/usr/local/lib/python1.5')
   sys.path.append('/usr/local/lib/python1.5/site-packages')
   import os
   TEMPLATES=os.environ['DOCUMENT_ROOT']
   from calendar import month_name
   from types import *
   from formhandler import FormHandler
   from string import strip, split, join, atol, atof, upper, lower, 
capitalize
   from refer import getReferer, addGetResume
   from time import asctime, localtime, time, strftime
   from urllib import quote_plus, unquote_plus
   from fcsetup import LOCALHOSTNAME
   from fcsetup import *
   import Mail
except:
   print '<pre>\n'
   import traceback
   traceback.print_exc()
   print sys.path

resumeDIR = DEPOT

      
class GetResumeHandler(FormHandler):
yada yafda yada...<<


I get the following errors:


Traceback (innermost last):
  File "./getresume.py", line 10, in ?
    TEMPLATES=os.environ['DOCUMENT_ROOT']
  File "/home/fcinc/python/Python-1.5.2/lib/python1.5/UserDict.py", line 
14, in __getitem__
    def __getitem__(self, key): return self.data[key]
KeyError: DOCUMENT_ROOT
['.', '.', '/home/fcinc/python/Python-1.5.2/lib/python1.5/', 
'/home/fcinc/python/Python-1.5.2/lib/python1.5/plat-linux2', 
'/home/fcinc/python/Python-1.5.2/lib/python1.5/lib-tk', 
'/home/fcinc/python/Python-1.5.2/lib/python1.5/lib-dynload', 
'/home/fcinc/python/Python-1.5.2/lib/python1.5/site-packages', 
'/home/fcinc/python/Python-1.5.2/lib/python1.5/site-packages/Gadfly', 
'/home/fcinc/python/Python-1.5.2/lib/python1.5/site-packages/HTMLgen', 
'/home/fcinc/python/Python-1.5.2/lib/python1.5/site-packages/MySQL', 
'/home/fcinc/python/Python-1.5.2/lib/python1.5/site-packages/kwParsing', 
'/home/fcinc/python/Python-1.5.2', 
'/home/fcinc/python/Python-1.5.2/lib/Python1.5/site-packages']
Traceback (innermost last):
  File "./getresume.py", line 28, in ?
    resumeDIR = DEPOT
NameError: DEPOT

I think DEPOT might be a reference to the MySQL database, using the 
Berkley dbmodule (which I can't find). I certainly can't find a 
directory named DEPOT anywhere on the old host's server.

Neither can I fathom where "DOCUMENT_ROOT" is defined, or find a file 
called "os.environ."

Please help!

TIA,


From chafik at nevrax.com  Thu Apr  6 04:00:14 2000
From: chafik at nevrax.com (Sameh chafik pro)
Date: Thu, 6 Apr 2000 10:00:14 +0200
Subject: looking for LL(1), LR(1), or LALR(1) syntaxic analysis in python.
Message-ID: <00bc01bf9f9e$240c53a0$1001a8c0@nevrax.net>

Hello evry body,
At present i develop an open source in agent IA, i write for you because i'm looking for any LL(1), LR(1), or LALR(1) syntaxic analysis in python implementation.
thanks for your answer.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20000406/dfc1377b/attachment.html>

From kern at caltech.edu  Sun Apr 23 05:57:10 2000
From: kern at caltech.edu (Robert Kern)
Date: 23 Apr 2000 09:57:10 GMT
Subject: [Ann] PyFT 0.65 -- Python Bindings for FreeType
Message-ID: <8duhdm$ipr@gap.cco.caltech.edu>

I have released a new version of the Python wrappers of FreeType, the
TrueType rendering engine.

New in this version (besides bugfixes) is PyFT.py, a high(er)-level
interface for rendering text.

While I wait for my Starship account to be restored, the files are located
on my own computer.  Beware, though: I try to keep my computer up 24-7, but
sometimes it does come down.  It is not a dedicated server, either.

http://taliesen.caltech.edu:9673
http://taliesen.caltech.edu:9673/PyFT-0.65.tar.gz

-- 
Robert Kern
kern at caltech.edu

"In the fields of hell where the grass grows high
 Are the graves of dreams allowed to die."
  -- Richard Harter


From remove.me.wwicker at spectratechnologies.com  Wed Apr 26 14:34:15 2000
From: remove.me.wwicker at spectratechnologies.com (William Wicker)
Date: Wed, 26 Apr 2000 18:34:15 GMT
Subject: Python COM troubles
Message-ID: <390a2f04.149999918@news.onlineisp.com>

I'm trying to use Python with SolidWorks, a really cool CAD package.
SolidWorks has a COM interface which plays reasonably well with
Python.

Unfortunately, the interface does not seem to provide run-time type
info. This appears to confuse the gen_py interface, which means that
all of my dispatch objects are dynamic. (bummer)
In turn, this looks like it means that I cannot use
DispatchWithEvents, as it seems to rely on the gen_py library being
recognizable.

Is there a way around this appearance of typelessness? There is a type
library, and I've run make-py successfully. I just can't see an easy
way into it.

It should by now be evident that I know little about Python, and less
about COM. Enlightenment welcome.

And now for something completely different:
COM dispatch methods that are documented to return SAFEARRAYS of
dispatch pointers fail with pythonwin v125 and 128 (the only ones I've
felt safe enough to try). The traceback is something like this:
>>> import win32com.client
>>> swx = win32com.client.Dispatch("SldWorks.Application")
>>> adoc = swx.ActiveDoc	#get the active model
>>> sel = adoc.SelectionManager #get the current list of selected objects
>>> sel.GetSelectedObjectCount #how many are there?
1
>>> face = sel.GetSelectedObject3(1) #I already know that I selected a face, so lets get it.
>>> face
<COMObject <unknown>>
>>> edges = face.GetEdges() #should return a tuple full of dispatch pointers to edge objects...
Traceback (innermost last):
  File "<interactive input>", line 1, in ?
  File "C:\tools\Python\win32com\client\dynamic.py", line 393, in
__getattr__
    ret =
self._oleobj_.Invoke(retEntry.dispid,0,pythoncom.DISPATCH_PROPERTYGET,1)
SystemError: NULL result without error in call_object
>>> 

Suggestions?

	William.


From wrd at awenet.com  Mon Apr 10 16:24:12 2000
From: wrd at awenet.com (William R. Dickson)
Date: Mon, 10 Apr 2000 13:24:12 -0700
Subject: Problem running Python script as cron job
References: <wrd-0AEDDA.12282507042000@news.awenet.com> <38EE41F4.8050708@gmx.net> <wrd-70F5CD.12423810042000@news.awenet.com> <m3k8i5los8.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <wrd-22C351.13241210042000@news.awenet.com>

In article <m3k8i5los8.fsf at atrus.jesus.cam.ac.uk>, Michael Hudson 
<mwh21 at cam.ac.uk> wrote:

> > > Try stderr?
> > > /path/prog > /path/log 2>&1
> > > 
> > > Maybe it spits some errors?
> > 
> > Tried that, but I still wind up with an empty log.  Anyway to force it 
> > to write immediately rather than buffering?
> 
> [mwh21 at atrus mwh21]$ python --bleargh
> Unknown option: --
> usage: python [option] ... [-c cmd | file | -] [arg] ...
> Options and arguments (and corresponding environment variables):
> -d     : debug output from parser (also PYTHONDEBUG=x)
> -i     : inspect interactively after running script, (also 
> PYTHONINSPECT=x)
>          and force prompts, even if stdin does not appear to be a 
>          terminal
> -O     : optimize generated bytecode (a tad; also PYTHONOPTIMIZE=x)
> -OO    : remove doc-strings in addition to the -O optimizations
> -S     : don't imply 'import site' on initialization
> -t     : issue warnings about inconsistent tab usage (-tt: issue errors)
> -u     : unbuffered binary stdout and stderr (also PYTHONUNBUFFERED=x)
> -v     : verbose (trace import statements) (also PYTHONVERBOSE=x)
> -x     : skip first line of source, allowing use of non-Unix forms of 
> #!cmd
> -X     : disable class based built-in exceptions
> -c cmd : program passed in as string (terminates option list)
> file   : program read from script file
> -      : program read from stdin (default; interactive mode if a tty)
> arg ...: arguments passed to program in sys.argv[1:]
> Other environment variables:
> PYTHONSTARTUP: file executed on interactive startup (no default)
> PYTHONPATH   : ':'-separated list of directories prefixed to the
>                default module search path.  The result is sys.path.
> PYTHONHOME   : alternate <prefix> directory (or <prefix>:<exec_prefix>).
>                The default module search path uses <prefix>/python1.5.
> 
> so that'll be `python -u' then ...
> 
> Cheers,
> M.

Heh...dowh.  If you're ever in Seattle, I owe you a beer :).

-Bill


From aahz at netcom.com  Tue Apr  4 16:11:54 2000
From: aahz at netcom.com (Aahz Maruch)
Date: 4 Apr 2000 20:11:54 GMT
Subject: Debugging Python 1.3 -> 1.5.2 (was Re: Python 1.6 alpha 1 released)
References: <Pine.GSO.4.10.10004040035420.21084-100000@sundial> <8cca93$9oh$1@nnrp1.deja.com> <B964E942A897FF5F.529887C74CF2FC7E.977238F9475EAC23@lp.airnews.net> <FdoG4.5023$74.81126@newsc.telia.net>
Message-ID: <8cdiaa$44i$1@slb7.atl.mindspring.net>

In article <FdoG4.5023$74.81126 at newsc.telia.net>,
Fredrik Lundh <effbot at telia.com> wrote:
>
>   [....]
>I suggest you dig up my list of "change cost" criterias, and apply
>them to all non-compatible changes (possibly starting with 1.5.2's
>class object change, which broke *lots* of code in really naughty
>ways [1]).  I've done that, and I still think that "connect" is the
>*least* critical of all code-breaking changes in the python core,
>since 1.3 or so.
>
>   [....]
>1) stumbled upon one today: it turned out that by changing from 1.3
>to 1.5.2, large data transfers from a high-performance server to clients
>on an 10 Mbit Ethernet were unexpectedly terminated.  the same trans-
>fers worked just fine with high-performance workstations, on an FDDI
>network.  ten guru points to anyone who can explain why.

Not without a bit more info about the data transfer mechanism and the
error reported by client & server.
--
                      --- Aahz (Copyright 2000 by aahz at netcom.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

"It's 106 miles to Chicago.  We have a full tank of gas, a half-pack of
cigarettes, it's dark, and we're wearing sunglasses."  "Hit it."


From mwh21 at cam.ac.uk  Sun Apr 16 15:59:23 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 16 Apr 2000 20:59:23 +0100
Subject: [ANNOUNCE] Garbage collection for Python
References: <20000407221750.A29279@acs.ucalgary.ca> <m3r9cagx3c.fsf@atrus.jesus.cam.ac.uk> <l03102809b51d41895a80@[193.78.237.135]> <38F9F01B.57273052@tismer.com> <m33domez7n.fsf@atrus.jesus.cam.ac.uk> <38FA05FA.B0526CAB@tismer.com>
Message-ID: <m3hfd1es10.fsf@atrus.jesus.cam.ac.uk>

Christian Tismer <tismer at tismer.com> writes:

> > I cam across a great quote the other day; I think it was attributed to
> > Alan Perlis, and it was:
> > 
> >     Simplicity follows complexity, not the other way round.
> 
> Thank you a lot, it was exactly the quote I was trying
> to look up, but I didn't find it again. It is very true.

Actually from

    http://www.cs.yale.edu/~perlis-alan/quotes.html

we find:

31. Simplicity does not precede complexity, but follows it. 

so I got the wording slightly wrong.  The intent's clear though.

now-wait-while-I-add-most-of-those-to-my-sig-file-ly y'rs
Michael

-- 
  On the other hand,  the following areas are subject to boycott
  in reaction to the rampant impurity of design or execution, as
  determined after a period of study, in no particular order: 
    ...                            http://www.naggum.no/profile.html


From bparsia at email.unc.edu  Wed Apr  5 20:16:02 2000
From: bparsia at email.unc.edu (Bijan Parsia)
Date: Wed, 5 Apr 2000 20:16:02 -0400
Subject: Teaching the art of programming, in python
References: <20000405053920.20579.qmail@web2102.mail.yahoo.com> <8cfe42$o02$1@nnrp1.deja.com> <dpJG4.3744$HG1.106895@nnrp1.uunet.ca>
Message-ID: <1e8mov8.1yh8ryv1kp6z0lN%bparsia@email.unc.edu>

Warren Postma <embed at geocities.com> wrote:

[snip]
> You see, I have this idea of 'elegance' in my head which Python satisfies,
> but I haven't found a book that really captures that. I have read a few (4),
> and none really do justice to the Art side.
> 
> Remember the famous 'Smalltalk-80 - The Language and Implementation'
> purple-book?
[snip]

Pedantic correction: The Language and Implementation version is the
*blue* book. The Blue Book is no longer in print, although the *purple*
book: Smalltalk-80 The Language is still in print (and you can get the
implementation chapters on the web) and it still a wonderful, wonderful,
wonderful read. It is beautiful to look at, a pleasure to read, and
deeply instructive.

-- 
Bijan Parsia
http://monkeyfist.com/
...among many things.


From dennis_marti at yahoo.com  Sat Apr  1 01:06:02 2000
From: dennis_marti at yahoo.com (Dennis Marti)
Date: Sat, 01 Apr 2000 01:06:02 -0500
Subject: Qutting the interpreter
References: <38E42BC6.4591E2E7@simplyweb.net>
Message-ID: <dennis_marti-C29703.01060201042000@news.starpower.net>

In article <38E42BC6.4591E2E7 at simplyweb.net>, Michael Ackerman 
<ack at simplyweb.net> wrote:

>    I am trying to find a way of quitting the interpreter. On p.3 of the
> tutorial it says this can be done by 'import sys; sys.exit()', but this
> produces an error message when I try it. I used to be able to quit by
> 'raise SystemExit', but that no longer works either. So how is this done?

File -> Quit works, as does command-q.

control-d, "raise SystemExit", and "import sys; ..." work when the 
interpreter is running by itself, but not with the IDE.

>    I am using version 1.5.2c.1 of Python on a Mac.

me too. an iBook, in case Matheus is reading.

Dennis


From python at rose164.wuh.wustl.edu  Sat Apr  8 06:33:26 2000
From: python at rose164.wuh.wustl.edu (David Fisher)
Date: Sat, 8 Apr 2000 05:33:26 -0500
Subject: Refeing to substring in list of strings
References: <MPG.1359171bf0a65199989683@nntp.server.uni-frankfurt.de>
Message-ID: <00c301bfa145$dff11f80$0201a8c0@spkydomain>

Referencing a substring works fine for me:

>>> ll = ['1234','asdf','qwer']
>>> for i in range(len(ll)):
    for j in range(len(ll[i]) - 1):
        print ll[i][j:j+2]


12
23
34
as
sd
df
qw
we
er
>>>

If you're trying to assign to the substring, like s[1:2] = 'e',  you can't,
strings are immutable.
dnf

----- Original Message -----
From: "Peter Kovalsky" <julschae at usa.net>
Newsgroups: comp.lang.python
To: <python-list at python.org>
Sent: Saturday, April 08, 2000 4:15 AM
Subject: Refeing to substring in list of strings


> Hi,
>
> I am new to Python and experienced a little problem with referencing
> lately. I have a list of strings that I have to modify. Thereofre I chose
> a for-construct (for index in range(len(mylist)):). I found certain
> substrings that i want to extract - but don' t know how to adress these
> substrings. I tried a
> mylist[index][lborder:rborder]
> which won' r work for some reason - and I don' t know why. Can anybody
> help???





From tiddlerdeja at my-deja.com  Tue Apr 11 12:59:57 2000
From: tiddlerdeja at my-deja.com (tiddlerdeja at my-deja.com)
Date: Tue, 11 Apr 2000 16:59:57 GMT
Subject: Free Python magazine
References: <8bauug$9m7$0@216.39.159.17>
Message-ID: <8cvlm7$sa$1@nnrp1.deja.com>

I can't see this on the site. Could you give a bit more information?

Thanks.


In article <8bauug$9m7$0 at 216.39.159.17>,
  fyl at ssc.com wrote:
> With the May issue of Linux Journal we are including a 28-page
supplement on
> Python.  As a special deal for Python fans, we are making this
issue/supplement
> available for free.  The info is on our main web page at www.ssc.com.
>
> The only catch is that we ask a few questions as we want to see what
> your interests may be.  No other strings.
>
> If you are into Monty Python I think you will really enjoy the
supplement.
> The parody starts with the cover and continues throughout the
supplement.
> Besides the humor (which is in the form of graphics), you will see
> the following articles:
>   * Computer Porgramming for Everybody by Guido van Rossum
>   * Why Python? by Eric Raymond
>   * JPython by David Ascher
>   * Mailman by Barry A. Warsaw
>   * Book reviews on three new Python books
> Plus you get a whole (180 page) issue of Linux Journal.
>
> --
> Phil Hughes, SSC, Inc. P.O. Box 55549, Seattle, WA 98155  (206)PUBS-
REF
> >>> Publishers of pocket references for UNIX, C, ..., Linux Journal
<<<
>     E-mail: fyl at ssc.com  Phone: (206)782-7733 x104  Fax: (206)782-7191
> SSC/Linux Journal web site: http://www.ssc.com/
>


Sent via Deja.com http://www.deja.com/
Before you buy.


From samschul at pacbell.net  Sat Apr  1 12:41:38 2000
From: samschul at pacbell.net (Sam Schulenburg)
Date: Sat, 01 Apr 2000 17:41:38 GMT
Subject: Python1.6 alpha bugs
Message-ID: <8c5ccg$u96$1@nnrp1.deja.com>

I am posting this in the news group as I can not access python.org.

The following problems were seen while testing Python1.6a:

1) I had to recompile the source to python1.6a under windows to get the
python16.lib file to link with my extended c routines. The default lib
file with the distrubution would not link.

2) IDLE and python crashes when trying to close an edit window that has
been opened in IDLE.

3) The help menu in IDLE, F1 key shows the default stderr color to be
dark green instead of red.

Sam Schulenburg


Sent via Deja.com http://www.deja.com/
Before you buy.


From gmcm at hypernet.com  Thu Apr 27 09:23:52 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Thu, 27 Apr 2000 09:23:52 -0400
Subject: Namespace Problem with global declaration in module
In-Reply-To: <8e912d$cpi$1@nnrp1.deja.com>
Message-ID: <1255280660-66425116@hypernet.com>

spex66 at my-deja.com wrote:

> hmm, deja is not so fast as my newsreader :), so here is a reply
> of Gordons message
> 
> >spex66 at my-deja.com wrote:
> >
> >> Hi,
> >> it seemed very simple but don't worked like expected:
> >>
> >> #FILE: test.py
> >> dd = None #INIT a shadow instance for automated access
[snip]
> >>
> >> >>> from test import *
> >
> >There's your problem. "from test import *" copies names from
> >test to your namespace. Now dd is another name for None. yy()
> >affects the dd in test's namespace, not yours.
> >
> >- Gordon
> 
> I see the point, but I like the from statement :)

Tough :-). import * is mostly a shortcut for interactive work. Look at the 
top of Tkinter.py to see what a module needs to do to make import * 
safe.

> Is it a bug? 

No.

> It's not the way I expected this code to work... and
> I cannot find a hook in the documentation that explicit explained
> this misbehaviour.

It's there, in 6.11 of the Language Ref:

The from form does not bind the module name: it goes through the 
list of identifiers, looks each one of them up in the module found in 
step (1), and binds the name in the local namespace to the object 
thus found. If a name is not found, ImportError is raised. If the list 
of identifiers is replaced by a star ("*"), all names defined in the 
module are bound, except those beginning with an underscore ("_"). 

No, the implications of that statement are not immediately obvious.

> Any idea for a workaround (I need the from-statement for further
> code readability)?

No, you want it for type-ability. Come back to it in a few months and 
you'll wish you'd spelled out where all those names came from.

- Gordon



From eviltofu at rocketmail.com  Sun Apr  2 11:23:42 2000
From: eviltofu at rocketmail.com (Jerome Chan)
Date: Sun, 02 Apr 2000 15:23:42 GMT
Subject: Is www.python.org down?
Message-ID: <eviltofu-C63EA3.11234102042000@news.earthlink.net>

Have not been able to connect to the web server although the machine 
responds to pings. Is it just me?


From daniels at mindspring.com  Thu Apr 20 23:05:52 2000
From: daniels at mindspring.com (Alan Daniels)
Date: 21 Apr 2000 03:05:52 GMT
Subject: editor for win32
References: <NpHL4.779$fK2.113609@news.flash.net>
Message-ID: <slrn8fv3a3.kk.daniels@localhost.localdomain>

On Thu, 20 Apr 2000 17:48:29 GMT, Scott Hathaway <slhath at flash.net>
wrote:

>Can anyone recommend a python editor for Windows?

VIM, available at http://www.vim.org.

That is, if you enjoy the VI way of doing things. :-) Even if you
haven't tried it before, I highly recommend it: Once I learned it, I
realized I'm able to blaze through text now because I'm never reaching
for the mouse (or ctrl-key combinations <0.5 wink>). I made the switch
because VIM is available on every machine I've ever come in contact
with (Windows, Linux, Solaris, yada yada).

Haven't-looked-back-since-ly your's, Alan.

-- 
============================================================
Alan Daniels
daniels at mindspring.com


From mikael at isy.liu.se  Tue Apr  4 11:57:48 2000
From: mikael at isy.liu.se (Mikael Olofsson)
Date: Tue, 04 Apr 2000 17:57:48 +0200 (MET DST)
Subject: New Features in Python 1.6
In-Reply-To: <ltvh21ywfs.fsf@asfast.com>
Message-ID: <XFMail.000404175748.mikael@isy.liu.se>

On 02-Apr-00 Lloyd Zusman wrote:
 >  Well, I guess this isn't all that bad, because I can cut down on the
 >  download bandwidth by compressing all those Python modules.  I'll use
 >  the new compression algorithm I recently invented.  It's amazing!
 >  .. it compresses *anything* down to one bit.
 >  
 >  I'm still working on the de-compression algorithm, though ...

Ahh, the famous BlackHole alpha release...

/Mikael

-----------------------------------------------------------------------
E-Mail:  Mikael Olofsson <mikael at isy.liu.se>
WWW:     http://www.dtr.isy.liu.se/dtr/staff/mikael
Phone:   +46 - (0)13 - 28 1343
Telefax: +46 - (0)13 - 28 1339
Date:    04-Apr-00
Time:    17:57:18

This message was sent by XF-Mail.
-----------------------------------------------------------------------


From fredrik at pythonware.com  Wed Apr 26 10:27:53 2000
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 26 Apr 2000 16:27:53 +0200
Subject: Python compiler ? (was Re: Python for drivers?)
References: <3900EDBE.282BC85@dsuper.net> <8dr2pn$ec8$1@newshost.accu.uu.nl> <3906f804@nntp.server.uni-frankfurt.de>
Message-ID: <015b01bfaf8b$9cd03930$0500a8c0@secret.pythonware.com>

mlauer at trollinger-fe.rz.uni-frankfurt.de wrote:
> Speaking if this... there was a guy in this newsgroup claiming
> that he could write a Python compiler if there would be enough
> interest. What has become of this (guy | project) ?

if you have $60,000 to spare:
http://home.t-online.de/home/aotto/compiler.html

</F>




From bjorn at roguewave.com  Tue Apr  4 16:32:31 2000
From: bjorn at roguewave.com (Bjorn Pettersen)
Date: Tue, 04 Apr 2000 14:32:31 -0600
Subject: Python 1.6 alpha 1 released
References: <Pine.GSO.4.10.10004040035420.21084-100000@sundial> <5E877FD7637AB46E.DABBED9CF3C1D64B.CDDDC05AE7374A78@lp.airnews.net> <8cca93$9oh$1@nnrp1.deja.com> <B964E942A897FF5F.529887C74CF2FC7E.977238F9475EAC23@lp.airnews.net> <FdoG4.5023$74.81126@news
Message-ID: <38EA515F.CC257802@roguewave.com>

Fredrik Lundh wrote:
> 
> Bjorn Pettersen <bjorn at roguewave.com> wrote:
> > Your definition of broken seems very peculiar.  Correct me if I'm wrong,
> > but it seems to go something like:
> 
> I've explained all this to you already.  lousy newsfeed? :-)

you keep repeating yourself, but you're certainly not explaining... not
having a good stance to argue from?

> >   - you don't know if examples are binding documentation

I take it this is still an unknown?  It's a pretty easy question.  When
example code in the documentation works as advertised, should people be
able to rely on it?  You can repeat that examples have been broken in
the past as much as you want, that's not what I'm asking about.

> >   - pair in the description of connect _must_ mean tuple
> >     and _not_ a pair of arguments because ???
> 
> I've explained that in another post -- the library reference use
> a special syntax for optional arguments.  the description of the
> connect method doesn't use that syntax.  other method
> descriptions do (this includes descriptions on the same page).
> 
> besides, "pair" is used quite extensively in the library reference,
> usually with exactly the same meaning as here: a 2-tuple.  see,
> for example, the description of "socket.error", "accept", and
> "recvfrom".

Apparently it is only used extensively enough that a large segment of
the Python community still thought the example was perfectly valid.

> >     (I would have thought it would have said tuple if it meant
> >     tuple -- tuples being Python concept as opposed to pairs)
> 
> I've explained that in another post -- it tells you that the
> address format is (host, port) -- which is perfectly valid tuple
> syntax.

...in context it can also be viewed as a perfectly valid parenthentical
note telling the reader what each of the two arguments to connect are --
especially since the "host, port" part is in italics and the parens
aren't.

> but alright, let's change "pair" to "2-tuple" everywhere.  will
> fixing this make you accept the change?

That would be clear, but it wouldn't change the fact that this change
involves changing a documented API. (And please argue why the example is
not documentation instead of repeating "the example is wrong, it
conflicts with the documentation", a perfectly reasonable reading of the
documentation, both description and example, gives you a two-argument
connect.

> >   - the example in question is broken because, even though
> >     it works, and seems to follow the description from the
> >     previous page, ...
> 
> I've explained that in another post -- it worked because the
> connect implementation didn't handle its argument properly,
> not because it was designed that way.  it stopped working
> when Guido fixed the argument handling bug.

that is purely an implementation detail. Any reasonable reader would
assume it worked because the implementation followed the documented
interface.

> and yes, this is exactly the same reason why append, insert,
> etc. did work in 1.5.2, and why they don't work in 1.6.
> yet, you think there is a difference.

That's also an implementation detail, but it's not documented. _That_ is
what the difference is.

> >   - ... other examples have been broken in the past...
> 
> I've explained that in another post -- broken examples are
> obviously not formally binding.  You replied that broken
> examples should be fixed.  Of course it should.  So should
> all broken demo code and the broken examples, just like
> all code that uses "append" etc must be fixed.

I haven't heard anyone argue that non-working examples should be
binding, but you seem to argue that working examples shouldn't be
binding either.  It's a perfectly valid opinion, and it is a stance that
other official programming language documentation has taken, I just
don't think it would be a good idea for Python -- and besides you, I
haven't heard anyone else argue that it would be.

[...]
> > The good news is that Guido realized how much code this would break, and
> > will fix the code.
> 
> yeah, and he'll probably break it again in 1.7.

I'm sorry you hold Guido in such low regard.  Waffling back and forth is
not a Guido characteristic as far as I can tell.

> just like he'll break append and a few other things in 1.6,
> and just like he broke stuff in 1.5.  and in 1.4.  etc.
> 
> you can keep on arguing until the end of c.l.py, but there's
> nothing special about "connect".  except, maybe, that you
> *know* that you're affected by this change, and you think
> you're not affected by any of the others.
> 
> if so, that's not only a very egocentric view, it's also rather
> naive.

So, I have to read your posts but not vice versa?  (I guess I'm in good
company with my egocentric views...)

> (no, don't reply.  just read my other posts in this thread. or
> if you think I don't have a clue, don't.  I don't really care)

You're sure arguing hard for a person who doesn't care. (and regressing
to schoolyard argumenting tactics certainly indicates that this thread
has lost any usefulness it may have had... not that I'm not expecting
you to come back with a 'last word' though <wink>)

> </F>

-b



From pinard at iro.umontreal.ca  Thu Apr 20 14:07:39 2000
From: pinard at iro.umontreal.ca (=?ISO-8859-1?Q?Fran=E7ois_Pinard?=)
Date: 20 Apr 2000 14:07:39 -0400
Subject: docstrings
In-Reply-To: Ben Wolfson's message of "Thu, 20 Apr 2000 16:05:38 GMT"
References: <00bufsonasp7bgkf86lfb6sv5np5v8mqlj@4ax.com>
Message-ID: <oqya684pec.fsf@titan.progiciels-bpi.ca>

Ben Wolfson <rumjuggler at home.com> writes:

> is there a way to use %c and %s within a docstring, or to add strings
> together in a docstring, so that if some strings are used in all
> docstrings, but might be changed, I can simply change one or two
> variables?

You could merely write docstrings as formats, using %(VARIABLE)s notation.
Then, whenever you want to process it, just use __doc__ % locals(), or
more generally __doc__ % DICTIONARY, with a suitable DICTIONARY.

-- 
Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard





From jcm at bigskytel.com  Tue Apr 25 00:35:42 2000
From: jcm at bigskytel.com (David Porter)
Date: Mon, 24 Apr 2000 22:35:42 -0600
Subject: The Simple Economics of Open Source
In-Reply-To: <24042000.3@sanctum.jae.ddns.org>; from jae@ilk.de on Mon, Apr 24, 2000 at 11:57:47PM +0200
References: <B527E7A5.51AA%raffael@mediaone.net> <3903316C.5D6F1611@tcac.net> <raffael-34658F.23400223042000@news.ne.mediaone.net> <390451C0.4B5D3196@tcac.net> <LHEoAJAXWFB5EwF6@jessikat.demon.co.uk> <24042000.3@sanctum.jae.ddns.org>
Message-ID: <20000424223542.B4648@bigskytel.com>

* Juergen A. Erhard <jae at ilk.de>:

> Fundamental question is: Are we just a complex arrangement of
> bio-chemical (and thus ultimately physical) processes... or is there
> more to being `human beings'?

Both. There is no "just" about complex arrangements of bio-chemical
processes, and who cares whether we are just physical... Of course there is
much more to being human beings, unless you are the greediest of
reductionists. 

> If we are just a collection of bio-chemical processes, it's simple to
> `simulate' or replicate us in machine form[1].

I disagree. Especially with the simple part. Ask someone in AI or cognitive
science whether they feel it is simple. We had better understand how those
processes work before we call them simple!

> If we are more... than it will not be possible.

But that's asking for magic and, that has to be explained too.

> - -- 

Is this type of discussion regular fare for c.l.py? 

  David

 



From gregm at iname.com  Thu Apr 20 10:47:52 2000
From: gregm at iname.com (Greg McFarlane)
Date: Fri, 21 Apr 2000 00:47:52 +1000
Subject: Lazy imports (was Re: module naming)
In-Reply-To: <m12hWfB-000CnvC@artcom0.artcom-gmbh.de>; from Peter Funk on 18 Apr 2000 at 01:59:29PM
References: <Pine.BSF.4.10.10004162056040.25069-100000@chi.pair.com> <m12hWfB-000CnvC@artcom0.artcom-gmbh.de>
Message-ID: <20000421004752.05406@nms.otc.telstra.com.au>

On 18 Apr, Peter Funk wrote:
> You could try to adopt the lazy loader concept implemented in 
> the Pmw package.  ...


I think there is a bug in your code.  Several lines have the wrong
indentation.  The following:
>         for name in dict.keys():
>             self.__dict__[name] = {}
>             d = {}
>             execfile(os.path.join(self._path, _EXP_DEF), d)
>             for k,v in d.items():
>                 if dict.has_key(k):
>                     if type(v) == types.TupleType:
>                         for item in v:
>                             ## modpath = self._package + item
>                             modpath = item
>                             dict[k][item] = modpath
>                     elif type(v) == types.DictionaryType:
>                         for k1, v1 in v.items():
>                             ## modpath = '_'+self._package +'.'+ v1
>                             modpath = v1
>                             dict[k][k1] = modpath

should be:
>         for name in dict.keys():
>             self.__dict__[name] = {}
>         d = {}
>         execfile(os.path.join(self._path, _EXP_DEF), d)
>         for k,v in d.items():
>             if dict.has_key(k):
>                 if type(v) == types.TupleType:
>                     for item in v:
>                         ## modpath = self._package + item
>                         modpath = item
>                         dict[k][item] = modpath
>                 elif type(v) == types.DictionaryType:
>                     for k1, v1 in v.items():
>                         ## modpath = '_'+self._package +'.'+ v1
>                         modpath = v1
>                         dict[k][k1] = modpath

-- 
Greg McFarlane     INMS Telstra Australia     gregm at iname.com



From eviltofu at rocketmail.com  Fri Apr 21 17:27:41 2000
From: eviltofu at rocketmail.com (Jerome Chan)
Date: Fri, 21 Apr 2000 21:27:41 GMT
Subject: Overloading methods?
Message-ID: <eviltofu-3DE33B.17272521042000@news.earthlink.net>

I have the following:

class mc:
    def __init__(self,a):
        self.a = a
    def dosomething(self,a):
        self.a = self.a + a

class yc(mc):
    def __init__(self,a,b):
        mc.__init__(self,a)
        self.b = b
    def dosomething(self,a,b):
        mc.dosomething(self,a)
        self.b = self.b + b

x = mc(1)
y = yc(1,2)

x.dosomething(1)
y.dosomething(1,2)

y = yc(1) <--- Got expected 3 but got 2 error
y.dosomething(1) <--- Got exptected 3 but got 2 error

Does this mean that method overloading is not allowed?

I tried the following workaround, by dispatching the call myself:

class mc:
    def __init__(self,a1):
        self.a = a1
    def dosomething(self,a1):
        self.a = self.a + a1

class yc(mc):
    def __init__(self,a1=None,b1=None):
        if a1:
            mc.__init__(self,a1)
        if b1:
            self.b = b1
    def dosomething(self,a1=None,b1=None):
        if a1:
            mc.dosomething(self,a1)
        if b1:
            self.b = self.b + b1

x = mc(1)
y = yc(1)

x.dosomething(1)
y.dosomething(b=1,a=2) <-- got attribute error: b

So it seems that if if you can't use b until it has been instantiated 
previously.

By the way, being able to call methods with keyword arguments in any 
order is one of the neatess things I like about python.


From ullrich at math.okstate.edu  Sun Apr  2 15:29:45 2000
From: ullrich at math.okstate.edu (David C. Ullrich)
Date: Sun, 02 Apr 2000 14:29:45 -0500
Subject: Function override?
References: <000101bf9b9b$312f8fe0$2f2d153f@tim> <38E64275.CF8F97DC@math.okstate.edu> <m37leh4fx4.fsf@greebo.nodomain.de>
Message-ID: <38E79FA8.4FB26C1@math.okstate.edu>

Bernhard Herzog wrote:

> "David C. Ullrich" <ullrich at math.okstate.edu> writes:
>
> >     Or I could if I hadn't just read the bit about how in 1.6 strings
> > have a split method, and if s1 and s2 are strings then
> >
> > s1.split(s2)
> >
> > either splits s1 using s2 as the delimiter or the other way around,
> > depending on which occurs more frequently in the other. I mean
> > surely it will happen that we want to split a bunch of strings
> > using 'delimiter' as a delimiter, and then it just happens that
> > the routine that does that gets passed 'lim' as an argument and
> > does the opposite of what we want?
>
> Oh, that's easy to work around. I'm sure we'll all get used to the
> following new idiom:
>
> (text + delim).split(delim)[:-1]

    What puzzles me about _this_ is that it was posted on
April 2. (So I'm slow - they don't call me "DU" for nothing...)

DU




From remove.me.wwicker at spectratechnologies.com  Thu Apr 27 12:34:34 2000
From: remove.me.wwicker at spectratechnologies.com (William Wicker)
Date: Thu, 27 Apr 2000 16:34:34 GMT
Subject: Python COM troubles
References: <390a2f04.149999918@news.onlineisp.com> <L5MN4.5088$v85.30985@news-server.bigpond.net.au>
Message-ID: <390864e0.229263112@news.onlineisp.com>

On Thu, 27 Apr 2000 00:46:35 GMT, "Mark Hammond"
<mhammond at skippinet.com.au> wrote:

>"William Wicker" <remove.me.wwicker at spectratechnologies.com> wrote in
>message news:390a2f04.149999918 at news.onlineisp.com...
>
>> Is there a way around this appearance of typelessness? There is a type
>> library, and I've run make-py successfully. I just can't see an easy
>> way into it.
>
>You can try running "makepy.py -i", and select your type library.  This
>will spit out some code, something like:
>
>gencache.EnsureModule(...)
>
>change this to:
>
>mod = gencache.EnsureModule(...)
>
>And you should be able to create any of the class objects in the makepy
>module - eg:
>
>ob = mod.TheClassName()
>

This works great!

There is one hitch (maybe). At some point, as I dive deeper into the
object structure (SolidWorks -> Document -> Selection ->
SelectedObject), I get a handle to an object which exists only to
super-class others. Can I pull some trick to force an unknown COM
object to use a specific class interface?

>If you search deja news, there should be lots of info about this - but
>generally relating to "Lotus Notes", which also has the same problem.
>
>[BTW - drop them a line, and tell them they should fix their objects :-]
>
>> COM dispatch methods that are documented to return SAFEARRAYS of
>> dispatch pointers fail with pythonwin v125 and 128 (the only ones I've
>> felt safe enough to try). The traceback is something like this:
>
>> SystemError: NULL result without error in call_object
>> >>>
>
>From a CVS checkin message on 11th of Feb:
>Can you believe that we have never been able to recieve arrays of
>IDispatch or IUnknown objects?!?!?
>
>If you tried, you would have got "SystemError: NULL result without error
>in call_object"
>
>:-)  This was fixed in 129.  But 129 hasnt made it public yet - I will do
>this as soon as I can log back in to starship!
>
>Mark.

Wonderful! I'm looking forward to it.


From amitp at Xenon.Stanford.EDU  Thu Apr 13 21:59:35 2000
From: amitp at Xenon.Stanford.EDU (Amit Patel)
Date: 14 Apr 2000 01:59:35 GMT
Subject: readline skips on line
References: <8culp8$23a$1@porthos.nl.uu.net>
Message-ID: <8d5u27$fsu$1@nntp.Stanford.EDU>

 Soundwave <maybe at lucky.com> wrote:
| I got a part of a programm looking like this
| 
| f=open('test.txt', 'r')
| while f.readline()!='':
|     print f.readline()
| print 'ready'
| 

You're calling readline() twice.  The first line is checked to see if
it is != '', and the second line is printed.  The third line is
checked to see if it is != '', and the fourth line is printed.  And so
on.  Try this:

f = open('test.txt', 'r')
while 1:
    line = f.readline()
    if line == '': break
    print line
print 'ready'


      - Amit

--
Amit J Patel, Computer Science Department, Stanford University
http://www-cs-students.stanford.edu/~amitp/




From tim_one at email.msn.com  Wed Apr 19 00:56:04 2000
From: tim_one at email.msn.com (Tim Peters)
Date: Wed, 19 Apr 2000 00:56:04 -0400
Subject: IDLE lability
In-Reply-To: <Pine.OSF.4.10.10004111635070.6088-100000@prfdec.natur.cuni.cz>
Message-ID: <000401bfa9bb$90ed1be0$d42d153f@tim>

[Michal Bozon]
> I use IDLE as an Python IDE ...
> Win98 ...
> ...
> Now, after Python 1.6 release I tried New IDLE 0.6 and it seems to be even
> more unstable (it falls always after closing a window). Again the error
> was in tk module, I thing tk83.dll or so.

You didn't say which version of 1.6.  If 1.6a1, many people had this problem
on Windows flavors.  1.6a2 backed off to an earlier release of Tcl/Tk, and
that fixed the "death on window closing" problem for many.

more-info-for-more-info-ly y'rs  - tim





From thomas at xs4all.net  Wed Apr 19 15:15:40 2000
From: thomas at xs4all.net (Thomas Wouters)
Date: Wed, 19 Apr 2000 21:15:40 +0200
Subject: Python in the future
In-Reply-To: <38FDB8A7.E701F81F@sec.noaa.gov>; from jvickroy@sec.noaa.gov on Wed, Apr 19, 2000 at 07:46:15AM -0600
References: <38FDB8A7.E701F81F@sec.noaa.gov>
Message-ID: <20000419211540.P15664@xs4all.nl>

On Wed, Apr 19, 2000 at 07:46:15AM -0600, j vickroy wrote:

> I'm relatively new to Python and am trying to gain support for it at my
> work place.  To that end, I'm giving a brief (overview) presentation.
> One of the items I have been asked to consider is where will Python be
> 5-10 years from now.  I certainly don't know where C++ or Java will be
> in that time frame!  Any ideas on where to research the future plans for
> Python would be appreciated.

Well, if you want to see glimpes of Pythons' future, what better way than
Guido's very own timemachine ? Guido is being the true visionary, saying
more about the *direction* in which Python is going (or he wishes it to go,
which ammounts much to the same thing, currently) than the lowlevel,
technical, unimportant bits. And Guido is working on CP4E, Computer
Programming For Everyone -- making programming easy, for everyone, even
those with no affinity for computers what so ever. Python is a good step in
that direction, and hopefully other tools will follow. Other languages,
because diversity is good, but also good developement tools, interfaces, new
toys to play with, new areas for Python to concur.

Ten years, even five years, is a long way for computers. (For humans too,
seeing that five years ago i'd dropped out of school and was hauling pallets
of boxes at the local hospital, and today I could get any job I would wish
for.) We can say very little about how technology will look in ten years.
It'll still be recognizable, but the cutting edge of the techworld will
definately be new and exciting. And Python will definately be there, even if
only propelled by momentum. I, for one, would probably still port Python to
every OS I can get my hands on. Maybe I can even be the first for once, then
;)

For more information about CP4E, follow the 'essays' link on Guido's
homepage (http://www.python.org/~guido) or go directly to
http://www.python.org/doc/essays/. You can probably find some more
information in the edu-sig, and perhaps even some in other places by using a
normal search engine.

Gee-i-hope-i-understood-that-cp4e-thing-correct-ly y'rs

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From hudson_melville at my-deja.com  Mon Apr 17 20:35:39 2000
From: hudson_melville at my-deja.com (hudson_melville at my-deja.com)
Date: Tue, 18 Apr 2000 00:35:39 GMT
Subject: Beginner: string to long
Message-ID: <8dgakr$2i$1@nnrp1.deja.com>

I'm a very beginner, and I have a program with
nested loops that put numeric operators (+,-) in
a string of numbers ('3564752617') The outputs
are similar to '3564-752+617' How do I convert it
so it can evaluate into a long integer? Is there
a built-in that I don't know about that would do
this?


Sent via Deja.com http://www.deja.com/
Before you buy.


From fredrik at pythonware.com  Tue Apr  4 04:57:45 2000
From: fredrik at pythonware.com (fredrik at pythonware.com)
Date: Tue, 04 Apr 2000 08:57:45 GMT
Subject: Images (.bmp,.jpg,gif) in Tkinter windows (Tk newbie)
References: <38df6cfe.2185312@news.integrityonline.com> <8bo8gc$gic$1@nnrp1.deja.com> <38e8dd4f.15889708@news.integrityonline.com>
Message-ID: <8ccaq5$ab1$1@nnrp1.deja.com>

Tim Ottinger wrote:
> >   image = PhotoImage(file="myfile.gif")
> >   widget.config(image=image)
> >   widget.image = image # make sure to hang on to the image
>
> Well, that would be okay, except that it doesn't actually display the
> image on the screen (for me). Neither in '98 (boo) or 'NT (boo).

this definitely works, given a label or button widget.

maybe you should post your code, so we can figure out
what the problem really is.

</F>


Sent via Deja.com http://www.deja.com/
Before you buy.


From aa8vb at yahoo.com  Fri Apr  7 08:44:10 2000
From: aa8vb at yahoo.com (Randall Hopper)
Date: Fri, 07 Apr 2000 08:44:10 -0400
Subject: Beginning to be disapointed
In-Reply-To: <Aw4H4.8078$pR.68745@carnaval.risq.qc.ca>
References: <Aw4H4.8078$pR.68745@carnaval.risq.qc.ca>
Message-ID: <20000407084410.A1757097@vislab.epa.gov>

Hugues Demers:
 |Can somebody help me on this? I'm using python to read big arrays (100x2000)
...
 |The cpu percentage used by python goes up and there is nothing else to be
 |done except to kill the interpreter. I have no idea what's going on. Do I
 |need to free memory in some way ? Am I doing something wrong ? 

This may not be it, but if you compiled Python with threading support,
try compiling it out.  Seems a big buggy.

On both FreeBSD and IRIX, when I hit Ctrl-C in the Python interpreter, I
get a hung Python that maxes out the CPU.

-- 
Randall Hopper
aa8vb at yahoo.com



From ivanlan at callware.com  Tue Apr  4 23:35:24 2000
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Tue, 04 Apr 2000 21:35:24 -0600
Subject: What is python's language level?
References: <38EA1B8F.A797A130@lmco.com> <38E9DDD3.3313FF6@bioreason.com> <slrn8ekddq.gm3.zawrotny@gecko.sb.fsu.edu> <roy-D857DA.21212304042000@netnews.nyu.edu> <38EA9CD4.3473B910@callware.com> <roy-CF57ED.23102504042000@netnews.nyu.edu>
Message-ID: <38EAB47C.F792326C@callware.com>

Hola, Listeros--

Roy Smith wrote:
> 
> Ivan Van Laningham <ivanlan at callware.com> wrote:
> > Yes indeed.  Any n/troff source file could qualify.  Macro files, by
> > definition, transcend obfuscated.  I wonder what's next up that ladder?
> 
> Have you ever looked at the output of a bib | tbl | eqn pipeline?
> 

Yes!  I've *read* it.  Even worse, I've understood it.

<the-new-mystery-wrapped-in-the-Enigma-machine-is-a-masterful-robbery-that-foiled-alarm-systems-and-guards-at-the-Bletchley-museum-dedicated-to-WWII's-gay-codebreaker>-ly
y'rs,
Ivan
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
http://www.pauahtun.org and
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
Author:  Teach Yourself Python in 24 Hours



From hniksic at iskon.hr  Thu Apr 13 03:59:31 2000
From: hniksic at iskon.hr (Hrvoje Niksic)
Date: 13 Apr 2000 09:59:31 +0200
Subject: Python/Perl Popularity (Re: A Mountain of Perl...)
References: <200004102111.HAA03380@envy.fulcrum.com.au> <38F4996F.E3BD2626@acm.org> <8d2p3q$2vj9o$1@fido.engr.sgi.com> <snwrf1ls.fsf@dan.resnet.rochester.edu> <8d2uji$301cu$1@fido.engr.sgi.com>
Message-ID: <dp1z4av3bw.fsf@mraz.iskon.hr>

pj at sgi.com (Paul Jackson) writes:

> |> Nope, both lynx, and links, exist.
> 
> oh - could you offer a link to "links" -- it's a hard name to search
> for -- too many false hits.

Not so on freshmeat -- comes up first.


From dave at humancode.com  Mon Apr  3 19:05:11 2000
From: dave at humancode.com (Dave Barrett)
Date: Mon, 3 Apr 2000 18:05:11 -0500
Subject: embedding python question
Message-ID: <38e922ef$0$27307@wodc7nh7.news.uu.net>

Hola,

    Here is what I would like to do...

    Let's say I have N python scripts i would like to run.  I have a set
amount of time I would like them to run (eg.  500 milliisecs), splitting the
time between them. In other words, I would like each script to run for
approx N/time millisecs, then suspend itslef return to the C++ main
procedure to call the next one in order.  During the next timeslot I would
like to resume running the scripts where they stopped.  The timing does not
have to be exact, just close.  How would I go about doing this with the
embedded api ?  Is it possible ?  If not, are they any similiar things I can
try ?


Thanks,

Dave





From smalleys at gte.net  Wed Apr 12 22:53:57 2000
From: smalleys at gte.net (sue)
Date: Thu, 13 Apr 2000 02:53:57 GMT
Subject: Python and VTK and TK
References: <38F51AB1.29FB4C1D@earthlink.net>
Message-ID: <38F528C8.2C18364F@gte.net>

Scott -

Unfortunately, it depends on your definition of moderate.
Can you elaborate a bit?

You might want to start a dirty prototype in TK, to check
your overall design, and go to VTK/FLTK when time (speed or
due date :) becomes an issue.

You might want to check pyFLTK as well.  I've compiled all
three without any major problems, but if I remember
correctly, VTK ends up wanting several other small libs.

I think someone was working on a "generic GUI" file that
could map to any one of the three, trading off recoding for
a bit of a speed hit.

Sue.


From kdart at techie.com  Mon Apr  3 20:48:21 2000
From: kdart at techie.com (Keith Dart)
Date: Tue, 04 Apr 2000 00:48:21 +0000 (GMT)
Subject: info to spice up web pages with python
In-Reply-To: <008e01bf9b79$3b2f7fe0$0101a8c0@highwayi.com>
References: <008e01bf9b79$3b2f7fe0$0101a8c0@highwayi.com>
Message-ID: <20000404.482100@leviathan.kdart.com>


>>>>>>>>>>>>>>>>>> Original Message <<<<<<<<<<<<<<<<<<

On 3/31/00, 5:25:56 PM, "Geoff Nordli" <geoff at gnaa.net> wrote regarding 
info to spice up web pages with python:


> Hello All.

> I normally just use Python to script things, and help me with
> admin tasks.  I know very little about web page development.

> I have a young nephew.  He is really interested in web page
> development.  He wanted to learn JavaScript, but I convinced
> him to give Python a go.

Well, there is actually one good reason to use Javascript: it is run 
by the client (browser). This
allows small scripts to be embedded in web pages. This can be done 
because all modern browsers 
have embedded Javascript (also called Ecmascript) interpreters. To do 
the same thing with Python
would require a Python browser plug-in, and I'm not aware of any that 
exist. There is a Tcl plug-in,
however. 

On the other hand, if you want to do server-side scripting (e.g. CGI), 
then Python works just fine with
your Apache server. In fact, there is a mod_python (from PyApache 
package) available for Apache. This
makes server/CGI scripts run faster, but won't really be necessary for 
CGI experimenting.

HTH,
Keith Dart






From ralf.claus at teleatlas.com  Mon Apr 17 05:18:02 2000
From: ralf.claus at teleatlas.com (Ralf Claus)
Date: Mon, 17 Apr 2000 11:18:02 +0200
Subject: a progress bar in tkinter ?
Message-ID: <38FAD6CA.6198CF2A@teleatlas.com>

Hello,
i search about a "widget", that works like a ProgressBar in VB.
Is there a possibility given by Tkinter?

thanks

Ralf


From markus.oberhumer at jk.uni-linz.ac.at  Wed Apr 12 16:29:12 2000
From: markus.oberhumer at jk.uni-linz.ac.at (Markus F.X.J. Oberhumer)
Date: Wed, 12 Apr 2000 22:29:12 +0200
Subject: [ANNOUNCE] PySol 4.00 - a solitaire game collection
Message-ID: <20000412222912.A10580@laetitia.oberhumer.com>

-----BEGIN PGP SIGNED MESSAGE-----


                    PySol - a Solitaire Game Collection
                              Version 4.00

                          http://pysol.tsx.org

           Copyright (C) 1998, 1999, 2000 Markus F.X.J. Oberhumer
                   <markus.oberhumer at jk.uni-linz.ac.at>


What is PySol ?
===============

PySol is an exciting collection of 173 solitaire card games.

Among the supported games are classics like Aces Up, Baker's Game,
Canfield, FreeCell, Forty Thieves, Golf, Klondike, Monte Carlo,
Osmosis, Pyramid, Scorpion, Spider, Yukon, and many more...

PySol may be freely distributed under the terms of the GNU GPL.

PySol aims to be Commercial Quality Freeware.


Why yet another solitaire game ?
================================

Here are some highlights of PySol:

  - freely available
  - distributed under the GNU GPL with full source code
  - currently supports 173 (!) distinct solitaire variants
  - based upon an extensible solitaire engine
  - very nice look and feel including multiple cardsets
    and background table tiles
  - sound samples and background music
  - unlimited undo & redo
  - load & save games
  - player statistics and log files
  - hint system and demo games
  - support for user written plug-ins - add your own solitaire variants
  - integrated HTML help browser
  - lots of documentation
  - fully portable across Unix/X11, Windows 95/98/2000/NT and MacOS
  - written in 100% pure Python
  - just run it - no need to compile anything


Nice. And what's new ?
======================

  * PySol now supports Mahjongg games

  * PySol now plays Tarot type games

  * Added a floating "Demo" logo

  * Fixed a number of minor problems


Cool. Where can I get it ?
==========================

Point your browser to http://pysol.tsx.org
The PySol Gallery is awaiting your visit as well.


What do I need to start playing ?
=================================

PySol requires Python 1.5.2 and Tcl/Tk 8.0.5. Both packages are
freely available for Unix, Windows 95/98/NT and Macintosh.

BTW, there is no need to compile anything since the whole program is just
a Python script. Just run it, and that's all.

The Windows version ships as a completely self-contained setup file.


License terms
=============

PySol is Copyright (C) 1998, 1999, 2000 Markus Franz Xaver Johannes Oberhumer

PySol is distributed under the terms of the GNU General Public License (GPL).
See the file COPYING.


Have fun,
Markus

http://pysol.tsx.org



-----BEGIN PGP SIGNATURE-----
Version: 2.6.3ia
Charset: noconv

iQCVAwUBOPTcWG10fyLu8beJAQH6wQP/SR6LJX2TqGaaRZ5kdiGvtxzX0HroPmke
kZRq3sb2NxZmwfhqLH/Bz4W0aP6pAhrA7VyRiM8kyLUSR8Mlu0CSD60bNzUxKAur
BxhCgw7iNcHkTBR3oQL1squBzqolQNDS+fuAmPNIszQY8u7KQ66/jOsXhzXjIfBT
dtXVi1jNCqE=
=Hi3/
-----END PGP SIGNATURE-----



From M_Pope at shef.rage.co.uk  Thu Apr 20 04:54:03 2000
From: M_Pope at shef.rage.co.uk (Mark Pope)
Date: Thu, 20 Apr 2000 09:54:03 +0100
Subject: dir(..)   vs.   PyList_GetItem
Message-ID: <213CDEBFDDA7D3118081009027E40C56071163@SHEF-MS1>

Rollocks.

Yes, I understand now. Thanks very much, you've saved me from yet more cloud
gazing...

mark.

> -----Original Message-----
> From: Greg Ewing [mailto:greg at cosc.canterbury.ac.nz]
> Sent: 20 April 2000 3:22 AM
> To: python-list at python.org
> Subject: Re: dir(..) vs. PyList_GetItem
> 
> 
> Mark Pope wrote:
> > 
> > The code I posted was just a snippet.
> > I have already successfully imported the module. I now need 
> to get a list of
> > the functions it exposes.
> 
> Are you *sure* you have imported the module properly?
> Because the result you posted looked an awful lot like
> you were looking at an empty module. Note that even if
> you imported the module previously, PyModule_New will
> *still* create a new module object. 
> 
> Can you post the code you actually ran?
> 
> If you've got the right module object, I can't imagine
> why looking at the keys of the module dict wouldn't
> work -- particularly since it works in Python, too:
> 
> >>> import string
> >>> dir(string)
> ['__builtins__', '__doc__', '__file__', '__name__', '_idmap', 
> '_idmapL',
> '_lower', '_re', '_safe_env', '_swapcase', '_upper', 'atof',
> 'atof_error', 'atoi', 'atoi_error', 'atol', 'atol_error', 
> 'capitalize',
> 'capwords', 'center', 'count', 'digits', 'expandtabs', 'find',
> 'hexdigits', 'index', 'index_error', 'join', 'joinfields', 'letters',
> 'ljust', 'lower', 'lowercase', 'lstrip', 'maketrans', 'octdigits',
> 'replace', 'rfind', 'rindex', 'rjust', 'rstrip', 'split', 
> 'splitfields',
> 'strip', 'swapcase', 'translate', 'upper', 'uppercase', 'whitespace',
> 'zfill']
> >>> string.__dict__.keys()
> ['_safe_env', '_idmap', '_swapcase', 'capitalize', 'atoi_error',
> 'octdigits', 'rfind', 'index', 'expandtabs', 'center', 
> 'join', 'lower',
> 'digits', 'atof', '__name__', 'hexdigits', 'atol', 'zfill', 'count',
> 'atoi', 'atof_error', '_re', 'swapcase', '__file__', '__builtins__',
> 'joinfields', 'lowercase', '__doc__', 'translate', 'rstrip', 'split',
> 'splitfields', '_lower', 'atol_error', 'rindex', 'whitespace',
> '_idmapL', '_upper', 'rjust', 'find', 'ljust', 'strip', 'index_error',
> 'capwords', 'replace', 'maketrans', 'uppercase', 'lstrip', 'letters',
> 'upper']
> >>>
> 
> The only difference seems to be that dir() sorts the
> list.
> 
> -- 
> Greg Ewing, Computer Science Dept,
> +--------------------------------------+
> University of Canterbury,	   | A citizen of 
> NewZealandCorp, a	  |
> Christchurch, New Zealand	   | wholly-owned subsidiary of 
> USA Inc.  |
> greg at cosc.canterbury.ac.nz	   
+--------------------------------------+
> -- 
> http://www.python.org/mailman/listinfo/python-list
> 



From bingdo19 at earthlink.net  Thu Apr 20 11:26:10 2000
From: bingdo19 at earthlink.net (Bob Langdon)
Date: Thu, 20 Apr 2000 15:26:10 GMT
Subject: Test_time Fails? / Installing on Virtual DOmain
References: <bingdo19-5535DF.10511520042000@news-server.twcny.rr.com>
Message-ID: <bingdo19-8B25AD.11352820042000@news-server.twcny.rr.com>

In article <bingdo19-5535DF.10511520042000 at news-server.twcny.rr.com>, 
Bob Langdon <bingdo19 at earthlink.net> wrote:

> Hi,
> 
> Upon installing Python in the home directory of a virtual domain (after 
> making the necessary configure file changes; i.e.: replacing all 
> instances of "/usr/" with: "/home/username/python/usr/")
> 
> "make test" results in 1 error:
> 
> >>1 test failed: test_time
> 19 tests skipped: test_al test_audioop test_bsddb test_cd test_cl 
> test_crypt test_dbm test_dl test_gdbm test_gl test_gzip test_imageop 
> test_imgfile test_nis test_rgbimg test_sunaudiodev test_thread 
> test_timing test_zlib<<
> 
> Running Lib/test/test_time.py results in:
> >>Lib/test/test_time.py: import: command not found
> Lib/test/test_time.py: time.altzone: command not found
> Lib/test/test_time.py: line 5: syntax error near unexpected token `t'
> Lib/test/test_time.py: line 5: `t = time.time()'<<
> 
> Is this something I've failed to install, or something missing from the 
> system?



Of course if I add the correct path test_time.pl

then I get the following error:


>>Traceback (innermost last):
  File "Lib/test/test_time.py", line 5, in ?
    time.altzone
AttributeError: altzone<<


From aahz at netcom.com  Tue Apr 11 16:22:54 2000
From: aahz at netcom.com (Aahz Maruch)
Date: 11 Apr 2000 20:22:54 GMT
Subject: [ANNOUNCE] Garbage collection for Python
References: <20000407221750.A29279@acs.ucalgary.ca> <Fsuuw7.Inu@world.std.com> <dpk8i4646r.fsf@mraz.iskon.hr> <87em8ccywq.fsf@deneb.cygnus.argh.org>
Message-ID: <8d01iu$ca0$1@slb2.atl.mindspring.net>

In article <87em8ccywq.fsf at deneb.cygnus.argh.org>,
Florian Weimer  <fw-usenet at deneb.cygnus.argh.org> wrote:
>
>Yes.  AFAIK, Python doesn't perform well in a SMP environment, because
>the interpreter only runs on one CPU at a given time, even if you
>create multiple Python threads.

That is true if you stick to pure Python code and extensions that don't
release the global interpreter lock.  For example, if you're using HTTP,
the socket library releases the global lock and threads are extremely
effective.
--
                      --- Aahz (Copyright 2000 by aahz at netcom.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

Why doesn't "Just Say NO" include caffeine, nicotine, alcohol, Prozac,
and Ritalin?  --Aahz


From mlauer at trollinger-fe.rz.uni-frankfurt.de  Wed Apr 26 17:43:50 2000
From: mlauer at trollinger-fe.rz.uni-frankfurt.de (mlauer at trollinger-fe.rz.uni-frankfurt.de)
Date: 26 Apr 2000 23:43:50 +0200
Subject: Python compiler ? (was Re: Python for drivers?)
References: <3900EDBE.282BC85@dsuper.net> <8dr2pn$ec8$1@newshost.accu.uu.nl> <3906f804@nntp.server.uni-frankfurt.de> <015b01bfaf8b$9cd03930$0500a8c0@secret.pythonware.com>
Message-ID: <39076316@nntp.server.uni-frankfurt.de>

Fredrik Lundh (fredrik at pythonware.com) wrote:
> mlauer at trollinger-fe.rz.uni-frankfurt.de wrote:
> > Speaking if this... there was a guy in this newsgroup claiming
> > that he could write a Python compiler if there would be enough
> > interest. What has become of this (guy | project) ?

> if you have $60,000 to spare:
> http://home.t-online.de/home/aotto/compiler.html

Hmm... $60.000 - this is a funny guy... however,
do we have 60.000 python users ? If everyone was
to donate 1$... :))

--
--
    Regards & Gruesse from Mickey @ http://www.Vanille.de
  ---------------------------------------------------------
  How could anyone know me - when I don't even know myself ?




From a.eyre at optichrome.com  Tue Apr  4 07:31:07 2000
From: a.eyre at optichrome.com (Adrian Eyre)
Date: Tue, 4 Apr 2000 12:31:07 +0100
Subject: builtding library .pyd file format
In-Reply-To: <010f01bf9e0d$78af96c0$1001a8c0@nevrax.net>
Message-ID: <000b01bf9e29$4490a0c0$3acbd9c2@optichrome.com>

> Hello evry body,

Can you not post HTML in future, please?

> I want to know how to do to built a .pyd library file format form
> the windows .dll file format

Just rename the .dll to .pyd

No... really. It's *that* simple.




From donb at init.com  Wed Apr 26 08:01:01 2000
From: donb at init.com (Donald Beaudry)
Date: Wed, 26 Apr 2000 08:01:01 -0400
Subject: Metaclasses? 
References: <3905F4A0.2B39C705@darmstadt.gmd.de> <slrn8gc6qj.p7e.quinn@hedono.ugcs.caltech.edu>
Message-ID: <200004261201.IAA05678@zippy.init.com>

quinn at hedono.ugcs.caltech.edu (Quinn Dunkan) wrote,
> Anyway, even if you could figure out the ambiguity problem, I don't
> think metaclasses would help you, bytecodehacks might.

*If* bytecodehacks could help, metaclasses could be used to make
calling them neat and clean.  The idea is that during class
instantiation the metaclass would, using bytecodehacks, rewrite all of
the class's methods.  This might confuse the poor guy who comes along
and tries to derive a new class from one of these 'self less'.

--
Donald Beaudry                                     Ab Initio Software Corp.
                                                   201 Spring Street
donb at init.com                                      Lexington, MA 02421
                  ...So much code, so little time...




From peter at whetstonelogic.com  Mon Apr 17 19:06:25 2000
From: peter at whetstonelogic.com (Peter Halliday)
Date: Mon, 17 Apr 2000 19:06:25 -0400
Subject: Pointers
References: <38D00DCE.9740380F@be-research.ucsd.edu> <8bfsqo$kje$1@tribune.oar.net> <dp4s9wl9df.fsf@mraz.iskon.hr> <uPYC4.55946$3b6.224410@ozemail.com.au> <4k5D4.7566$0o4.54442@iad-read.news.verio.net>
Message-ID: <38FB98F1.239EC0FB@whetstonelogic.com>

I just started learning python.  I just want to make sure that I understand the
object interaction you were discussion here.  To use your statement.  So x + 1
is an object, x is an object, and y is an object.  Y and x happen to both be
pointing to x + 1.  Is this right?  If you then did something to y would you be
doing something to x + 1?  I would say, no because it is a seperate object
right!  You can't have a dangling pointer like in C++ right?  Because of the
full functional objects.  If this is true, and I understand this right, then
Python rocks.


>
>         x = x + 1
>         y = x
>
>

Peter Halliday



From sabren at manifestation.com  Sun Apr 23 02:54:27 2000
From: sabren at manifestation.com (Michal Wallace (sabren))
Date: Sun, 23 Apr 2000 02:54:27 -0400 (EDT)
Subject: code not true?
In-Reply-To: <Z%vM4.1491$RM6.3518433@news-east.usenetserver.com>
Message-ID: <Pine.BSF.4.10.10004230244460.5642-100000@chi.pair.com>

On Sun, 23 Apr 2000, Kragen Sitaker wrote:

> In base 3, one third is 0.1.  Two thirds is 0.2.

hey, I was right! :)

[snipped analogy of why 0.1 in base 3 is hard to write
in decimal]
 
> That said, it *is* just a speed issue --- there are representations
> other than floating point, like the ordered-pair rational-number
> representation, that give the right answer but are slower to work
> with.

But... Python is a scripting language.. It's okay to be slow.  From
what I'm hearing, many people on this list shy away from floats simply
because of the innacuracy.. Why don't we use the way that gets the
right answer?

Isn't unicode much slower to work with than regular strings? But
the benefit outweighs the cost.

Other than speed, is there a pro side to floats? Would switching
to altnerate representations break people's code?

I'm just wondering... :)

Cheers,

- Michal
-------------------------------------------------------------------------
http://www.manifestation.com/         http://www.linkwatcher.com/metalog/
-------------------------------------------------------------------------




From dalke at acm.org  Mon Apr 24 05:04:35 2000
From: dalke at acm.org (Andrew Dalke)
Date: Mon, 24 Apr 2000 03:04:35 -0600
Subject: Sorting a list of classes
References: <OmnqwHar$GA.233@cpmsnbbsa03>
Message-ID: <8e12gp$qpa$1@slb6.atl.mindspring.net>

lexberezhny wrote:
>If i have a python list of classes, can i do something like
classList.sort()
>and how can i control this behavior within the classes. ie
>i have a address book, and a have a class called 'Person', and it has an
>attribute Person.Name, so i have a list of 'Person's how can i sort the
>list, and maybe sort it by name? do i have to write my own function or can
i
>use list.sort() some how?


I wrote a mini-HOWTO on this subject.  It's at
http://python.org/doc/howto/sorting/sorting.html

If you will only ever sort on name, you can implement the __cmp__
method in your Person

class Person:
  ...
  def __cmp__(self, other):
    return cmp(self.Name, other.Name)

Most likely you will want to pass the comparison function to sort(),
as in

def byName(x, y):
   return cmp(x.Name, y.Name)

nameList.sort(byName)

There are more variations in that HOWTO.

                    Andrew
                    dalke at acm.org





From hamish_lawson at yahoo.co.uk  Fri Apr  7 20:04:46 2000
From: hamish_lawson at yahoo.co.uk (Hamish Lawson)
Date: Fri, 07 Apr 2000 17:04:46 -0700
Subject: Meaning of __rlshift__ and __rrshift__ ?
References: <20000407192251.A1974@yetix.sz-sb.de>
Message-ID: <07b20beb.2fe8f249@usw-ex0103-019.remarq.com>

__rrshift__ is the reversed-operand version of __rshift__. There
are reversed-operand versions of the other binary operators.
They are invoked for the object on the right-hand side of a
binary operator when the left-hand-side object doesn't support
the operator in question.

For example, consider the expression

    x / y

If x has a __div__() method defined then x.__div__(y) will be
invoked. But if x doesn't have __div__() defined then an attempt
will be made to invoke y.__rdiv__(x) instead.

Hamish Lawson


* Sent from RemarQ http://www.remarq.com The Internet's Discussion Network *
The fastest and easiest way to search and participate in Usenet - Free!



From sabren at manifestation.com  Tue Apr 18 12:42:50 2000
From: sabren at manifestation.com (Michal Wallace (sabren))
Date: Tue, 18 Apr 2000 12:42:50 -0400 (EDT)
Subject: o'rielly python devcenter
Message-ID: <Pine.BSF.4.10.10004181241280.15729-100000@chi.pair.com>

I didn't see this posted already, but sorry if its old news.
O'Reilly started a new python site:

     http://www.oreillynet.com/python/


Cheers,

- Michal
-------------------------------------------------------------------------
http://www.manifestation.com/         http://www.linkwatcher.com/metalog/
-------------------------------------------------------------------------




From moshez at math.huji.ac.il  Thu Apr 20 16:19:12 2000
From: moshez at math.huji.ac.il (Moshe Zadka)
Date: Thu, 20 Apr 2000 23:19:12 +0300 (IDT)
Subject: Good style for multi-valued returns
In-Reply-To: <889A6EB6AE399393.619B0539BFCAAE0F.5C6D110C3DA908F7@lp.airnews.net>
Message-ID: <Pine.GSO.4.10.10004202317390.21090-100000@sundial>

On 20 Apr 2000, Cameron Laird wrote:

> Examples of multi-valued returns:  
...
> (CAR, CDR)
<snipped>

Ummm....wouldn't that be a bit twisted? Just return the pair!

sorry-couldn't-resist-ly y'rs, Z.
--
Moshe Zadka <mzadka at geocities.com>. 
http://www.oreilly.com/news/prescod_0300.html
http://www.linux.org.il -- we put the penguin in .com




From someone at somewhere.com  Tue Apr 18 15:58:59 2000
From: someone at somewhere.com (Charley Horse)
Date: Tue, 18 Apr 2000 19:58:59 GMT
Subject: Using python on the web
References: <MPG.1365adcfa8470eef9896aa@news> <8di2bk$l1s$1@newshost.accu.uu.nl> <MPG.136631472f892a119896ac@news> <_W2L4.986$GK5.20898@news2-win.server.ntlworld.com>
Message-ID: <MPG.13665ef93eedbda29896ad@news>

It was just an unexpected surprise. I'm fairly new to all of this and 
always heard of CGI in the context of "you don't have to user CGI 
anymore because of the fancy new blah blah technology we offer!" I'll 
try it and I'm sure it'll be fine. If all of you clean-code loving 
python users can live with it then I'm sure I can too. <g>

In article <_W2L4.986$GK5.20898 at news2-win.server.ntlworld.com>, 
richard_chamberlain at ntlworld.com says...
> I'm curious why you would be put off using cgi?
> 
> having used python and cgi, plus java servlets/jsp and asp at various
> different times I cannot say I see your point.
> 
> In many ways my preferred method would be python because I can get something
> working in such a short space of time, and there's no significant speed
> difference between, for instance, java and python.
> 
> cgi has to be about the most common way of construction on the fly web
> pages, and I don't think it's going away just yet.
> 
> Richard
> 
> Charley Horse <someone at somewhere.com> wrote in message
> news:MPG.136631472f892a119896ac at news...
> > Thanks to all of those pointers. Finding out that CGI plays an important
> > role in the python world was a shock to me as I only started working
> > with web stuff in the last couple of years. I've just never touched a
> > CGI script, shockingly I suppose <g>. ASP/CF and PHP are what I've
> > worked with and CGI was just the "old world" technology for
> > interactivity that I read about but never had to work with.
> >
> > So coming at this as a newbie, and for years hearing about python's
> > elegance and oo-ness, it just seemed a strange pairing, python and CGI.
> > Puzzling in fact! I see why it makes sense, but it's just not what I
> > expected.
> >
> > I'll fool with CGi and perhaps it won't be enough of a turn off to
> > affect my interest.
> >
> > In article <8di2bk$l1s$1 at newshost.accu.uu.nl>, m.faassen at vet.uu.nl
> > says...
> > > Charley Horse <someone at somewhere.com> wrote:
> > > > I've just started fooling with python. I've looked around the python
> > > > site and a number of the other sites referenced there and am a bit
> > > > puzzled. Umm... how does one go about using python for web
> applications,
> > > > especially database driven apps? Almost everything deals with python
> > > > just as a language, as C++ would be.
> > >
> > > That's because Python is a language. Python isn't a web-specific
> language,
> > > after all.
> > >
> > > > Not much about the web specifically.
> > > > I was a bit shocked to find the only direct refs on about
> > > > this at python.org are entwined with CGI (unless I missed it).
> > >
> > > Why is CGI so shocking? It's used quite a bit, after all? You can write


From effbot at telia.com  Sun Apr 23 05:50:33 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Sun, 23 Apr 2000 09:50:33 GMT
Subject: code not true?
References: <Pine.BSF.4.10.10004230244460.5642-100000@chi.pair.com>
Message-ID: <JHzM4.1116$wYl.189856768@newsb.telia.net>

Michal Wallace (sabren) <sabren at manifestation.com> wrote:
> Isn't unicode much slower to work with than regular strings?

nope.

</F>




From dworkin at ccs.neu.edu  Wed Apr  5 13:38:08 2000
From: dworkin at ccs.neu.edu (Justin Sheehy)
Date: 05 Apr 2000 13:38:08 -0400
Subject: A Mountain of Perl Books + Python Advocacy
In-Reply-To: "Jason Stokes"'s message of "Wed, 5 Apr 2000 16:00:20 +1000"
References: <20000405053920.20579.qmail@web2102.mail.yahoo.com> <3FAG4.62951$3b6.243372@ozemail.com.au>
Message-ID: <vnditxwfny7.fsf@camelot-new.ccs.neu.edu>

"Jason Stokes" <jstok at bluedog.apana.org.au> writes:

> The market for Perl is 10-100 times bigger than Python.

The "market", as defined how?

-Justin

 



From moshez at math.huji.ac.il  Fri Apr 28 01:06:58 2000
From: moshez at math.huji.ac.il (Moshe Zadka)
Date: Fri, 28 Apr 2000 08:06:58 +0300 (IDT)
Subject: Converting = to ==
In-Reply-To: <n01O4.12725$HG1.340768@nnrp1.uunet.ca>
Message-ID: <Pine.GSO.4.10.10004280805520.8383-100000@sundial>

On Thu, 27 Apr 2000, Warren Postma wrote:

> I have an "end user query tool" that uses Python expressions, but I want
> comparisons to be able to be written either as A=B or A==B then I want to
> compare them using Python's Eval function.

Try using the tokenize model to tokenize the string, and replacing "="
with "==" at the token level. This will also gurrantee you won't be fooled
by strings containing "=" or other evils.

--
Moshe Zadka <mzadka at geocities.com>. 
http://www.oreilly.com/news/prescod_0300.html
http://www.linux.org.il -- we put the penguin in .com




From g_will at cyberus.ca  Wed Apr  5 00:17:31 2000
From: g_will at cyberus.ca (Gordon Williams)
Date: Wed, 05 Apr 2000 04:17:31 GMT
Subject: Installer Problem Finding Modules
Message-ID: <01bf9eb5$7ee884e0$1e41c3d1@Cyberus>

Hi,

I am trying to run Gordon McMillan's Installer (3f on W95) and I am having
a problem with it finding the modules that I am importing into the main
program.  The first part of my program goes like

"The main module for running the bitscope program"
import traceback
import GlobalVar
from Menu import myMenu
from MyCanvas import plotCanvas
from OperationMode import opModePanel
from TriggerPanel import triggerPanel
from SampleRatePanel import sampleRatePanel
from AnalogSourcePanel import analogSourcePanel
from SampleButtonPanel import sampleButtonPanel
import Registers
from wxPython.wx import *

reload(GlobalVar)
g= GlobalVar

Registers.openPort()		#sets up the port for IO


... and then goes on to define classes etc.

The installer ignores the all the imports except for the wxPython stuff
which it appears to get all of.  Any ideas where I may be going wrong.  I
hope that I dont have to list all these modules as dependancies as some of
these modules also call other modules.

Maybe this is a related problem, but I dont think so:
All the modules along with the main module are in a sub directory of the
installer program.  When I run python ..\simple.py bs_main.py from a dos
box
I was geting an error "can't open 'E:Program" when it it was trying to
start build.py.  I ended up changing simple.py line 74 to
cmd = '"%s" %s %s %s' % (sys.executable,
                           '../Builder.py',                      <<< change
is here
                           cfgfile, tkcfg)

to get it to work.

Thanks

Gordon Williams


From landauer at apple.com  Fri Apr  7 15:10:03 2000
From: landauer at apple.com (Doug Landauer)
Date: Fri, 07 Apr 2000 12:10:03 -0700
Subject: Teaching the art of programming, in python
References: <20000405053920.20579.qmail@web2102.mail.yahoo.com> <8cfe42$o02$1@nnrp1.deja.com> <dpJG4.3744$HG1.106895@nnrp1.uunet.ca> <38EB68A3.B1B5809F@callware.com> <8cfqa7$4ml$3@newsserver.rrzn.uni-hannover.de> <38EB73B6.6C0913A7@callware.com> <8chsij$k5r$1@newsserver.rrzn.uni-hannover.de> <slrn8er3fd.fcd.r2d2@shaka.acc.umu.se> <38EDDACC.8641FF8C@callware.com>
Message-ID: <landauer-D0C8CF.12100307042000@news.apple.com>

> "Teach your self Internet in 10 minutes". Seems there is
> inflation (or perhaps deflation) in these things.

I think I'll wait for the Cliff's Notes to come out ... that
might get it down to 2 minutes.


From robin at jessikat.demon.co.uk  Tue Apr 25 03:02:29 2000
From: robin at jessikat.demon.co.uk (Robin Becker)
Date: Tue, 25 Apr 2000 08:02:29 +0100
Subject: new linereading standard?
References: <3904D21B.333DD930@visionart.com> <390510C9.A744D22E@yahoo.com>
Message-ID: <R2g+MIAFMUB5EwUl@jessikat.demon.co.uk>

In article <390510C9.A744D22E at yahoo.com>, Al-Amerrho H. Amerin
<al_amerin at yahoo.com> writes
>Pete Shinners wrote:
>
>>
.....
>_
>>    pete at visionart.com
>
>A beginner myself, I would write it like this:
>
>inputlines = open('myfile.txt').readlines()
>for line in inputlines:
>    print line
>
>Instead of the 'while 1: ....break' thingy.
>
>Al-
>

am I wrong for writing

for line in open('myfile.txt').readlines():
    print line

-- 
Robin Becker


From gherman at darwin.in-berlin.de  Tue Apr 18 15:54:31 2000
From: gherman at darwin.in-berlin.de (Dinu C. Gherman)
Date: Tue, 18 Apr 2000 21:54:31 +0200
Subject: BAD starship.python.net news!
References: <38FC852A.9B11F458@concentric.net>
Message-ID: <38FCBD77.A5925D65@darwin.in-berlin.de>

Manus Hand wrote:
> 
> This from Mike McLay.  Everyone, let's pony up what it takes to recover
> our starship!
> 
> Manus (who will backup more and more often than he has until now)

Not a bad idea! Sigh...

I think Christian Tismer used to make regular copies of
large parts of the Starship, at least in former times...

Thinking of the not so long ago shut-down of Starship Py-
thon I (starship.skyport.net)... couldn't that be a source 
to recover substantial parts of the previous data??

Sigh!

Dinu

-- 
Dinu C. Gherman
................................................................
"The thing about Linux or open software in general is that 
it actually tries to move software from being witchcraft to 
being a science," [...] "A lot of the programs you see today 
are actually put together by shamans, and you just take it and 
if the computer crashes you walk around it three times... and 
maybe it's OK." (Linus Thorvalds, LinuxWorld 2000, NYC)


From klaus-gerd.meyer at gmx.de  Fri Apr 21 18:36:35 2000
From: klaus-gerd.meyer at gmx.de (Klaus-Gerd Meyer)
Date: Sat, 22 Apr 2000 00:36:35 +0200
Subject: Tkinter crash with Cleansweep
Message-ID: <8dql2j$1k52$1@msheas02.msh.de>

I'm new to Python. I installed Python 1.52 and TCL/TK on a Win98 PC.

On exit from IDLE and other Tkinter programs the programs often hangs and i
have to reboot.

To keep my PC clean, i have installed Norton CleanSweep 4.5.
I found, that CleanSweep runs always a program CSINJECT.EXE at startup, even
if i have disabled all the Cleansweep "Monitoring" options.
CSINJECT loads QDCSPI.DLL and that seems the problem(?).

If i remove CSINJECT.EXE from startup, all the Tkinter programs run fine!

Maybe this helps other users...

(Excuse my bad english)
--
MfG   K.- G. Meyer




From fig at oreilly.com  Fri Apr 14 13:45:32 2000
From: fig at oreilly.com (Stephen R. Figgins)
Date: Fri, 14 Apr 2000 17:45:32 GMT
Subject: Python tools for PDE and Linear algebra
References: <ya0vh1kzsaf.fsf@grindaflet.ifi.uio.no>
Message-ID: <38F75987.26AD3FEA@oreilly.com>

> Does anyone know
> some PDE tools or are working on such projects? Support for different
> numerical methods as Finite Difference and Finite Element Methods
> would be nice.

I don't know of any in Python, Roger, but there are some free libraries
available for C++.  You can find them listed at: 

  http://oonumerics.org/oon/

You might be able to adapt one of these, rewrite it in python or maybe
just call it from python. 

I expect these would be rather computationally expensive, and so you
might be better off with having them written in C or C++.  

Stephen Figgins


From jpb at mycroft.communiweb.net  Wed Apr 26 23:22:01 2000
From: jpb at mycroft.communiweb.net (jpb)
Date: Thu, 27 Apr 2000 03:22:01 GMT
Subject: python-only crypt()?
References: <Pine.BSF.4.10.10004230627580.5642-100000@chi.pair.com> <39065A08.F776CBE1@home.com> <87d7nd444b.fsf@psyche.evansnet>
Message-ID: <jpb-F69F87.23220226042000@news-server.cfl.rr.com>

In article <87d7nd444b.fsf at psyche.evansnet>, Carey Evans 
<c.evans at clear.net.nz> wrote:

>bo Vandenberg <famvandenberg at home.com> writes:
>
>> "a pure Python implementation of the Unix crypt(3) algorithm"
>> 
>> http://home.clear.net.nz/pages/c.evans/sw/
>
>Wow.  I finished writing that and uploaded it less than six hours
>before your post.  Would you mind letting me know how you found it so
>quickly?

Guido has been lending out the time machine again, I guess.

jpb


From clarence at netlojix.com  Sat Apr 15 16:59:20 2000
From: clarence at netlojix.com (Clarence Gardner)
Date: Sat, 15 Apr 2000 13:59:20 -0700
Subject: else clauses in while and for loops
References: <Zv_C4.6415$4N.169440@news.tpnet.pl> <v0RD4.13$n68.186989568@newsb.telia.net> <8c7b29$37v$1@inputplus.demon.co.uk> <38F79B93.B03F0A6B@mindspring.com> <8d9c66$el7$1@pegasus.csx.cam.ac.uk>
Message-ID: <955832508.2143160134@news.silcom.com>

On Sat, 15 Apr 2000, Nick Maclaren wrote:
>In article <38F79B93.B03F0A6B at mindspring.com>,
>Chuck Esterbrook  <echuck at mindspring.com> wrote:
>>So did anyone present, in this thread, an example that shows a real need?
>
>Whether or not they have one, there is a fairly common case:
>
>    Perform an action XXX provided that no member of sequence YYY
>fulfils a particular criterion.
>
I never thought of that general wording, but it certainly fits.
I did this just yesterday:

for fname in ('keyword','terms','title','journal','author','pubdate','thd','entrynumber'):
    if Form.has_key(fname):
        break
else:
    Errs.append('No search field supplied')


From paul.magwene at yale.edu  Tue Apr 11 10:38:06 2000
From: paul.magwene at yale.edu (Paul Magwene)
Date: Tue, 11 Apr 2000 10:38:06 -0400
Subject: Tcl vs Tkinter
References: <38f31c65$0$27527@wodc7nh0.news.uu.net>
Message-ID: <38F338CE.1507B802@yale.edu>

T Zimmerman wrote:
> 
>     I again am looking for help on the use of 'alt-*' to access the menus in
> any application written with Tkinter. I have tried the tcl examples that
> were included with the download of Python and they work as expected. So, I
> am confused as to why it does not work for Tkinter widgets.  The following
> script should work and should open the 'File' menu when 'alt-f' is pressed.
> On NT this does nothing.
> 
> ==================================================================
> import Tkinter
> 
> root = Tkinter.Tk()
> 
> button = Tkinter.Menubutton(text = 'File', underline = 0)
> button.pack()
> menu = Tkinter.Menu(button)
> button.configure(menu = menu)
> menu.add_command(label = 'Hello World', underline = 0)
> 
> root.mainloop()
> ==================================================================

Odd, your code works as expected using Python 1.5.2 on FreeBSD (3.x
stable), but doesn't work right on NT.

However, the following, which uses only Menu objects and cascades, DOES
work on both FreeBSD and NT:

#==================================================================
import Tkinter

root = Tkinter.Tk()

MnMenu = Tkinter.Menu()

menu = Tkinter.Menu(MnMenu)
menu.add_command(label = 'Hello World', underline = 0)

MnMenu.add_cascade(label='File',underline=0,menu=menu)

root.config(menu=MnMenu)

root.mainloop()
#==================================================================


Go figure.  Maybe the effbot has some insight?

--Paul


From javanet at dynacap.com  Tue Apr 11 20:50:01 2000
From: javanet at dynacap.com (Arnaldo Riquelme)
Date: Tue, 11 Apr 2000 16:50:01 -0800
Subject: Python 4 DOS
Message-ID: <sf7hsua14nk34@corp.supernews.com>

Can anyone tell me if I can get python binaries for DOS (16-bit).
I heard of  Hans Nowak 's  Python-DX but the links to his website is broken?
Where are you Hans?

Thanks



ajr


have fun





From exarkun at flashmail.com  Sun Apr  2 08:55:19 2000
From: exarkun at flashmail.com (Jp Calderone)
Date: Sun, 02 Apr 2000 12:55:19 +0000
Subject: Hmm, little Tk question.
References: <20000402165146.20535.cpmta@c000.sfo.cp.net>
Message-ID: <38E74337.35A912A8@flashmail.com>

Falknor wrote:
> 
> Allo peoples :)
> 
>     I was wonderin, is it possible to initiate a Tk fully functional Tk window across a socket connection?  Ie they are connected to me through a generic telnet app and have the Tk libs installed and operational.  Mudding is one of my hobbies and I had an idea last night while talkin to a player, got me thinkin bout this.  I believe I've heard someone talk about it on one of the mud mailing lists a good while ago but I can't seem to recall.  Ah well, any input is appreciated. :)
> 

export DISPLAY="my.ip.address:0.0"
./spiffyTkMudClient.py &

This assumes, of course, that they are running an X-server (Generally not
a safe assumption for the average MUD user)

> Falknor of the WoT MUD project,
> http://www.tarmongaidon.org/wotmud
> ______________________________________________________________
> Get free Internet service and email at http://www.worldspy.com

-- 
It is practically impossible to teach good programming style to
students that have had prior exposure to BASIC: as potential
programmers they are mentally mutilated beyond hope of
regeneration.        -- Dijkstra
--
 12:53pm up 4 days, 18:00, 0 users, load average: 0.21, 0.07, 0.02


From lutz at rmi.net  Tue Apr 25 12:18:54 2000
From: lutz at rmi.net (lutz at rmi.net)
Date: Tue Apr 25 12:18:54 2000
Subject: Nudge, nudge (LJ May issue)
Message-ID: <200004251821.MAA00625@chevalier.rmi.net>

I can't believe nobody has posted this yet, but 
just in case...

   http://www2.linuxjournal.com/lj-issues/issue73/lstoc.html

--Mark Lutz  (http://rmi.net/~lutz)  [PyMail 1.0]




From akuchlin at mems-exchange.org  Sat Apr 22 10:57:33 2000
From: akuchlin at mems-exchange.org (Andrew M. Kuchling)
Date: 22 Apr 2000 10:57:33 -0400
Subject: Starship status, April 22
Message-ID: <3dln26ji8y.fsf@amarok.cnri.reston.va.us>

More progress: 

* User accounts have been recreated. 

* Mail service has restored to most users.  Exceptions are:

No address found: tdfunk, stevel, sdm7g, moshez, mclay, bwilk
Forwarded back to starship (no non-starship address): bhoel,
manus, marduk, scharf, sk, tbryan, tommy, xoltar

* The contents of the FTP server have been restored, as have the Web
  pages.

* If you set up an SSH key for your account, you should be able to log
  in again.  IMPORTANT: We are still waiting to hear from the data
  recovery firm about the contents of the old disk.  Don't embark on
  any massive updating of your Web pages, in case we have to restore
  the more recent data.

  Just update any old software releases that will bury you in support
  e-mail, fix any really critical problems 

Still to do: Mailman, Zope, Web site config tweaks, passwords for
people who didn't use SSH.

-- 
A.M. Kuchling			http://starship.python.net/crew/amk/
If you can, and if you are a playgoer and a filmgoer, you should be able to
find voices for all the characters in the books you read.
  -- Robertson Davies, _A Voice from the Attic_



From mhammond at skippinet.com.au  Tue Apr 11 20:21:56 2000
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Wed, 12 Apr 2000 00:21:56 GMT
Subject: ANNOUNCE: Python 1.6 alpha 2
References: <200004112029.QAA09762@eric.cnri.reston.va.us>
Message-ID: <EkPI4.1681$UB.11570@news-server.bigpond.net.au>

Note that win32all-131 for this version of Python is available via
http://starship.python.net/crew/mhammond/win32

Works like a charm :-)

Mark.

"Guido van Rossum" <guido at python.org> wrote in message
news:200004112029.QAA09762 at eric.cnri.reston.va.us...
> I've just released a source tarball and a Windows installer for Python
> 1.6 alpha 2 to the Python website:
>
>   http://www.python.org/1.6/
>
...





From darrell at funkomite.com  Thu Apr 13 00:20:52 2000
From: darrell at funkomite.com (darrell)
Date: Wed, 12 Apr 2000 21:20:52 -0700
Subject: simple httplib and urllib timeout question
References: <Pine.LNX.4.21.0004120814500.6832-100000@fep132.fep.ru>
Message-ID: <1WbJ4.86$9B.7202@news.bc.tac.net>

Yeah I would love to:
a) check out your Internet Robot
b) help out with any code

Ive already made a database module using mysqldb for inserting urls
into a table and updating the db relationships etc, for classifying the
urls.

I played a bit today with Michal's botlib , it seems to work really well ,
and increases my bots speed alot (being that it doesn't hang on any urls) so
far. Ive left it running for the evening , and see how it goes.

Anyway , I would love to help out, if I can. My goal is to build a fast(as
possible) robot in python that will be ultra flexible and hopefully
intelligent as well that can add & classify urls.

So count me in if you want some help.

Darrell

"Oleg Broytmann" <phd at phd.russ.ru> wrote in message
news:Pine.LNX.4.21.0004120814500.6832-100000 at fep132.fep.ru...
> Hi, there!
>
> On Tue, 11 Apr 2000, Michal Wallace (sabren) wrote:
> > meanwhile... HEY OLEG! -> have you done anything w/ this?
>
>    I am about to release the third version of my "Bookmarks database and
> Internet robot" project. It is 98% finished, I just need a bit more
> testing. If you are interested, I can publish it in current beta state.
>
>    The third version is a complete rewrite. All things had been revamped,
> and rewriteen in OO manner, and made more modular and pluggable. I don't
> value OO by itself, but this time the OO approach helped me to make the
> code simpler, smaller, more readable and reusable.
>
>    After releasing it, I'll continue to work on new modules and plugins.
> There are thre areas for plugins - database managers, so called writers
> (data dumpers that could dump data to different text formats) and robots.
>    Currently I (re)implemented 2 robots - simple and forking, Forking
robot
> is fine, but slow - it checks one URL at a time.
>
>    In the field of your interest - yes, I want to write more robots, at
> least two - I want a threading robot, but much more than that I want an
> asyncore-based robot. I haven't done anything yet; it seems I need to
> rewrite the entire urllib to use asyncore, and it is hard work. Anyway I
> need the work to be done, so if no one would help me - I will do it
myself.
> At least will TRY to do it :)
>
> Oleg.
> ----
>     Oleg Broytmann    http://members.xoom.com/phd2.1/
phd2 at earthling.net
>            Programmers don't die, they just GOSUB without RETURN.
>
>




From joacim at home.se  Thu Apr 27 10:09:30 2000
From: joacim at home.se (JJ)
Date: Thu, 27 Apr 2000 16:09:30 +0200
Subject: Concatening string and integer
References: <8e9gco$jlt$1@vg170.it.volvo.se> <slrn8gghgk.el7.scarblac-spamtrap@flits104-37.flits.rug.nl>
Message-ID: <8e9hnv$ru9$1@vg170.it.volvo.se>

Thanks for all your fast answers

BR
Joacim

"Remco Gerlich" <scarblac-spamtrap at pino.selwerd.nl> wrote in message
news:slrn8gghgk.el7.scarblac-spamtrap at flits104-37.flits.rug.nl...
> JJ wrote in comp.lang.python:
> > Hi!
> >
> > I want to concatenate a string containing both strings and integers
like:
> >
> > strTemp = 'I make test number: ' + iMyNumber + ' once again'
> >
> > It doesn't work, how should I do this?
>
> 1) You make a string of the integer:
>
> strTemp = 'I make test number: ' + str(iMyNumber) + ' once again'
>
> 2) You use string formatting:
>
> strTemp = 'I make test number: %d once again' % iMyNumber
>
> --
> Remco Gerlich,  scarblac at pino.selwerd.nl
> Hi! I'm a .sig virus! Join the fun and copy me into yours!




From mwh21 at cam.ac.uk  Tue Apr  4 08:39:50 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 04 Apr 2000 13:39:50 +0100
Subject: Bugs in Python 1.6 alpha 1
References: <Pine.LNX.4.10.10003311651590.22919-100000@starship.python.net> <8c8vfj$5n6$1@news.udel.edu> <8ccmah$chj$1@pegasus.csx.cam.ac.uk>
Message-ID: <m33dp211m1.fsf@atrus.jesus.cam.ac.uk>

nmm1 at cus.cam.ac.uk (Nick Maclaren) writes:

>     6) test_unicode fails on Hitachi systems -- Writing: '*', expected:
> 'T'.  But only if -v is NOT specified!  Again, I failed to tie it down,
> and it looks like a test harness problem.

That's the way regrtest works; if you specify `-v' all output goes to
stdout, if you specify `-g' it goes into the output directory and if
you specify neither it goes into some internal buffer to be compared
with what's in the output directory.

A bit odd, yes.  I don't know if it's a deliberate behaviour (doubt
it).

Cheers,
Michael

-- 
well, take it from an old hand: the only reason it would be easier
to program in C is that you can't easily express complex  problems
in C, so you don't.                 -- Erik Naggum, comp.lang.lisp


From grisha at ispol.com  Thu Apr 27 08:25:06 2000
From: grisha at ispol.com (Gregory Trubetskoy)
Date: Thu, 27 Apr 2000 08:25:06 -0400
Subject: when to use intern()
Message-ID: <Pine.LNX.4.21.0004270820120.27672-100000@saturn.ispol.com>

It says in the manual that some things in Python get interned
automatically. I'm not sure where intern() is necessary - 

if I have a statement like:

if self.config.has_key("PythonOption"):
   ...

that gets executed a lot - should I consider interning "PythonOption" or
is it done automatically?

what about

PYTHON_OPTION = "PythonOption"
if self.config.has_key(PYTHON_OPTION):
   ...

is this interned?


Thanks!


--
  Gregory (Grisha) Trubetskoy
       grisha at ispol.com



From rob at hooft.net  Tue Apr 25 06:23:53 2000
From: rob at hooft.net (Rob W. W. Hooft)
Date: 25 Apr 2000 12:23:53 +0200
Subject: Good style for multi-valued returns
References: <889A6EB6AE399393.619B0539BFCAAE0F.5C6D110C3DA908F7@lp.airnews.net>
Message-ID: <m2wvlmcwcm.fsf@hooft.net>

>>>>> "CL" == Cameron Laird <claird at starbase.neosoft.com> writes:

 CL> Let me guess: folks generally pass tuples for small and
 CL> determinate-in-number multi-values, and dictionaries other- wise,
 CL> 'cept some do it all with tuples.  Accurate?

I sometimes use:

class Record:
    """Empty class for misc use"""
    _format={}
    
    def __init__(self,**kw):
        for key,value in kw.items():
            setattr(self,key,value)

    def __repr__(self):
        t=[]
        for i in dir(self):
            if i[0]!="_":
                if self._format.has_key(i):
                    if self._format[i]:
                        t.append("%s=%s"%(i,self._format[i]%getattr(self,i)))
                else:
                    t.append("%s=%s"%(i,repr(getattr(self,i))))
        import string
        return "Record: "+string.join(t)

    def copy(self):
        import copy
        return copy.copy(self)
    

-- 
=====   rob at hooft.net          http://www.xs4all.nl/~hooft/rob/  =====
=====   R&D, Nonius BV, Delft  http://www.nonius.nl/             =====
===== PGPid 0xFA19277D ========================== Use Linux! =========


From mark at chem.uwa.edu.au  Thu Apr 27 19:52:13 2000
From: mark at chem.uwa.edu.au (Mark C Favas)
Date: 27 Apr 00 23:52:13 GMT
Subject: Can there be scrollbar attached to group widget
References: <3903C15D.E33126@swt.edu>
Message-ID: <mark.956879533@declan>

ABC <ns566645 at swt.edu> writes:

>I have a notebook and the group is embedded in the notebook. can the
>scrollbar can be attached to the group.

You could try putting the group inside a PmwScrolledFrame, which is itself
embedded in the notebook...

Mark
--
Email  - mark at chem.uwa.edu.au      ,-_|\                           Mark C Favas
Phone  - +61 9 380 3482           /     \               Department of Chemistry
Fax    - +61 9 380 1005      ---> *_,-._/   The University of Western Australia
                                       v                               Nedlands
Loc    - 31.97 S, 115.81 E                               Western Australia 6009


From dgoodger at bigfoot.com  Sat Apr  8 23:34:58 2000
From: dgoodger at bigfoot.com (David Goodger)
Date: Sat, 08 Apr 2000 23:34:58 -0400
Subject: Python paradigms
References: <8cmr70$b0e$1@pegasus.csx.cam.ac.uk>
Message-ID: <B515729E.1745%dgoodger@bigfoot.com>

on 2000-04-08 04:38, Nick Maclaren (nmm1 at cus.cam.ac.uk) wrote:
> x = (a != NULL ? a[i]->weeble : 0) + (b != NULL ? b[i]->wombat : 0)

C's ?: operator, ie:

    test ? true : false

can be done in Python, with:

    ((test and [true]) or [false])[0]

Note that I said *can* be done, not *should* be done! It is very ugly and
counter-intuitive and naughty. But sometimes necessary. Use sparingly,
especially if your code will ever be read (even by you!).

I got this expression from Mark Lutz' book "Programming Python", end of
Chapter 5, page 135 (under "The Naughty Bits", "if as an expression").

-- 
David Goodger    dgoodger at bigfoot.com    Open-source projects:
 - The Go Tools Project: http://gotools.sourceforge.net
 (more to come!)



From rurban at x-ray.at  Sun Apr  9 11:16:13 2000
From: rurban at x-ray.at (Reini Urban)
Date: Sun, 09 Apr 2000 15:16:13 GMT
Subject: Python Language Support for ArcView GIS
References: <q0MH4.2398$Bq2.38636@newscontent-01.sprint.ca> <38EFAC08.2A10932D@verecomm.com>
Message-ID: <38f09fee.21313387@judy>

Peter Kropf wrote:
>Just an FYI: You might want to consider moving AVPython to
>sourceforge.net. The entire environment is designed for 
>various levels of open source development... 

When AVPython will be open source...


>Bruce Dodson wrote:
>> Python is free and open source.  I hope AVPython will eventually be the
>> same, but we haven't yet settled on an acceptable open source license, so
>> for now the best I can do is provide a binary that is freely
>> redistributable.

--
Reini Urban
http://xarch.tu-graz.ac.at/autocad/news/faq/autolisp.html


From pf at artcom-gmbh.de  Thu Apr 20 12:53:26 2000
From: pf at artcom-gmbh.de (Peter Funk)
Date: Thu, 20 Apr 2000 18:53:26 +0200 (MEST)
Subject: Lazy imports (was Re: module naming) (corrected version)
Message-ID: <m12iKFm-000CnvC@artcom0.artcom-gmbh.de>

Hi!

Two days ago I posted a followup to a question by Michal Wallace, 
which included some Python code, which was BUGGY!  But see below.

> Michal Wallace (sabren) asks:
> >   That's certainly a lot cleaner! But is there a way to do that and
> > not have it load each one of those files every time? There's actually
> > more than 3, and even though they're all fairly small, I'd rather not
> > have the overhead of loading them until they're needed.
> 
> You could try to adopt the lazy loader concept implemented in 
> the Pmw package.  The idea is to use a small tool, that builds a 
> dictionary of exported features (classes and more) by investigating
> all module files in the package and places this in a file called
> for example 'exports.def'.  Than you can use the following
> __init__.py file to load features (classes, constants, functions,
> or modules, what you want) on demand.  The idea is "stolen" from
> Greg McFarlanes wonderful Pmw package:

As Greg McFarlane pointed out to me, a block of my program got the
wrong indentation, so it will not work as I intended it to.  Here is
an updated (and hopefully more correct version).  Many thanks to
Greg McFarlane for his original work and the time he spent to review
my code.  This time I also included a little helper script used to
automatically generate the 'exports.def' file used by the lazy importer.

Regards, Peter
---- 8< ---- 8< ---- cut here ---- 8< ---- schnipp ---- 8< ---- schnapp ----
#!/usr/bin/env python
## vim:ts=4:et:nowrap
"""__init__ : This file is executed when the package is imported.  It creates
a lazy importer/dynamic loader for the package and replaces the package 
module with it.  This is a very simplified version of the loader supplied 
with Pmw.  All the package version management has been stripped off."""

import sys, os, string, types

_EXP_DEF = 'exports.def'       # export definition file
_BASEMODULE = 'base'           # Name of Base module for the package

class _Loader:
    """An instance of this class will replace the module in sys.modules"""

    def __init__(self, path, package):
        self._path, self._package = path, package
        self._initialised = 0
        
    def _getmodule(self, modpath):
        __import__(modpath)
        mod = sys.modules[modpath]
        return mod

    def _initialise(self):
        # Create attributes for the Base classes and functions.
        basemodule = self._getmodule('_'+self._package+'.'+_BASEMODULE)
        for k,v in basemodule.__dict__.items():
            if k[0] is not '_' and type(v) != types.ModuleType:
                self.__dict__[k] = v
        # Set the package definitions from the exports.def file.
        dict = {
            '_features'     : {},
            '_modules'      : {},
        }
        for name in dict.keys():
            self.__dict__[name] = {}
            d = {}
            execfile(os.path.join(self._path, _EXP_DEF), d)
            for k,v in d.items():
                if dict.has_key(k):
                    if type(v) == types.TupleType:
                        for item in v:
                            ## modpath = self._package + item
                            modpath = item
                            dict[k][item] = modpath
                    elif type(v) == types.DictionaryType:
                        for k1, v1 in v.items():
                            ## modpath = '_'+self._package +'.'+ v1
                            modpath = v1
                            dict[k][k1] = modpath
        self.__dict__.update(dict)
        self._initialised = 1

    def __getattr__(self, name):
        """This will solve references to not yet used features"""
        if not self._initialised:
            self._initialise()
            # Beware: _initialise may have defined 'name'
            if self.__dict__.has_key(name):
                return self.__dict__[name]
        # The requested feature is not yet set. Look it up in the
        # tables set by exports.def, import the appropriate module and
        # set the attribute so that it will be found next time.
        if self._features.has_key(name):
            # The attribute is a feature from one of the modules.
            modname = self._features[name]
            mod  = self._getmodule('_'+self._package+'.'+modname)
            feature = getattr(mod, name)
            self.__dict__[name] = feature
            return feature
        elif self._modules.has_key(name):
            # The attribute is a module
            mod = self._getmodule('_'+self._package+'.'+name)
            self.__dict__[name] = mod
            return mod
        else:
            # The attribute is not known by the package, report an error.
            raise AttributeError, name

# Retrieve the name of the package:
_package = os.path.split(__path__[0])[1]
# Rename (hide) the original package for later perusual:
sys.modules['_'+_package] = sys.modules[_package]
# Create the dynamic loader and install it into sys.modules:
sys.modules[_package] = _Loader(__path__[0], _package)
---- 8< ---- 8< ---- cut here ---- 8< ---- schnipp ---- 8< ---- schnapp ----
#!/usr/bin/env python
## vim:ts=4:et:nowrap                    
"""build_exports.py -- create 'exports.def' helper file for lazy importer

This is a hack to aid changing the substructure of a ggui
package quickly and painless.
"""
import sys, types, pprint

modules = []
features = {}

multiple_defined = {}

template = '''## vim:ts=4:et:nowrap                     
# [Emacs: -*- python -*-]
"""export.def --- This is an exports definition file ---

this was automatically created by %(prog)s

It is invoked by a dynamic import loader in __init__.

    features      : dictionary from feature names to modules names.
    modules       : tuple of module names 
"""
#

_features = %(features)s

_modules = %(modules)s

'''

def spewout(stream=sys.stdout, modules=(), features={}):
    pp = pprint.PrettyPrinter(indent=4)
    d = { 'prog': sys.argv[0], 
          'modules': pp.pformat(tuple(modules)), 
          'features': pp.pformat(features),
    }
    stream.write(template % d)

def inspect(modulename, modules, features, multiple_defined):
    if modulename[-3:] == ".py": modulename = modulename[:-3]
    __import__(modulename)
    mod = sys.modules[modulename]
    for symbol in dir(mod):
        if symbol[:1] != '_' or (symbol == '_' and modulename == 'base'):
            obj = mod.__dict__[symbol]
            if type(obj) == types.ModuleType or symbol == "Pmw":
                if not symbol in modules:
                    modules.append(symbol)
            else:
                if features.has_key(symbol):
                    if multiple_defined.has_key(symbol):
                        multiple_defined[symbol] = multiple_defined[symbol] + \
                            " " + features[symbol]
                    else:
                        multiple_defined[symbol] = features[symbol]
                features[symbol] = modulename

if __name__ == "__main__":
    sys.path.insert(0, '.')
    if len(sys.argv) > 1:
        for arg in sys.argv[1:]:
            inspect(arg, modules, features, multiple_defined)
        outfile = sys.stdout
    else:
        import glob
        l = glob.glob("[a-z]*.py")
        print l
        for module in l:
            inspect(module, modules, features, multiple_defined)
        if multiple_defined == {}:
            outfile = open("exports.def", "w")
    if multiple_defined == {}:
        spewout(outfile, modules,  features)
    else:
        for k, v in multiple_defined.items():
            print k, "has multiple definitions in:", v, features[k]




From sabren at manifestation.com  Wed Apr 26 06:50:04 2000
From: sabren at manifestation.com (Michal Wallace (sabren))
Date: Wed, 26 Apr 2000 06:50:04 -0400 (EDT)
Subject: python-only crypt()?
In-Reply-To: <87d7nd444b.fsf@psyche.evansnet>
Message-ID: <Pine.BSF.4.10.10004260634470.23888-100000@chi.pair.com>

On Wed, 26 Apr 2000, Carey Evans wrote:

> > http://home.clear.net.nz/pages/c.evans/sw/
> 
> Wow.  I finished writing that and uploaded it less than six hours
> before your post.  Would you mind letting me know how you found it so
> quickly?
> 
> > Post back if this is helpful
> 
> Given that I ported it because I couldn't get Michal's port of
> UnixCrypt going, it probably is.

:) Yup! Thanks again, Carey!

Just to clarify for everyone, there are at least two styles of crypt()
routines floating around out there:

http://home.clear.net.nz/pages/c.evans/sw/ is the original, DES-style crypt().
http://www.sabren.com/code/python/crypt/ is the other, MD5-style crypt().

Cheers,

- Michal
-------------------------------------------------------------------------
http://www.manifestation.com/         http://www.linkwatcher.com/metalog/
-------------------------------------------------------------------------





From mhammond at skippinet.com.au  Fri Apr 28 19:35:05 2000
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Fri, 28 Apr 2000 23:35:05 GMT
Subject: Newbie - COM parameters problem
References: <3909F954.9B51E117@cti.ecp.fr>
Message-ID: <JepO4.6825$v85.40009@news-server.bigpond.net.au>

"Martin Rehak" <rehakm1 at cti.ecp.fr> wrote in message
news:3909F954.9B51E117 at cti.ecp.fr...

> In one method , we have to pass the reference to the object as an [in]
> parameter.

Have you tried it?  You should just be able to pass the object!

> And my question is, how can we actually do it, when the only VARIANT
> "sub-type" fit for this is an IDispatch * and in the whole Python
> documentation, that I found on the net,  is no mention of some pointer
> type or its substitution ?

Python fully supports the VARIANT, including all the funky pointers it
contains.  It is not exposed to the Python programmer - it "just works"
(or should :-).  Passing a COM object will result in the IDispatch * of
the VARIANT getting correctly set - and if the param was [in,out], would
result in the IDispatch ** getting correctly set!

If you have problems, post the complete traceback, but passing other COM
objects works just fine...

Mark.





From pennybridge_pioneer at my-deja.com  Thu Apr 13 23:32:05 2000
From: pennybridge_pioneer at my-deja.com (pennybridge_pioneer at my-deja.com)
Date: Fri, 14 Apr 2000 03:32:05 GMT
Subject: intelligent reply please
Message-ID: <8d63fl$3ji$1@nnrp1.deja.com>

I am looking for someone knowledgable to help me decompile command
interpeter, ie command.com.  I wish to translate it from assembly to
machine language to readable ascii to be able to read the information
contained within.  If you think this can be done and you are someone
who can help me invent a decompiler for this purpose or if you have any
questions mail me at <mailto: fenlxxtx at aol.com>


Sent via Deja.com http://www.deja.com/
Before you buy.


From thanos at 0x01.com  Mon Apr  3 10:32:40 2000
From: thanos at 0x01.com (thanos)
Date: Mon, 03 Apr 2000 10:32:40 -0400
Subject: CGI - Hidden Fields (NS4.7/IE5)
References: <38D625B2.5E0E88FD@student.tuwien.ac.at>
Message-ID: <38E8AB88.5E5E60E5@0x01.com>

Make shure you are using POST and not GET. EI handle form method GET
badly!!!!

Tamer Fahmy wrote:

> Hi!
>
> I have got a strange problem with my CGI-script. If I call my
> page from Netscape through Apache I get all hidden Fields and
> values. But if I instead call it from Internet Explorer 5 the
> hidden fields don't show up in my FieldStorage.
>
> I noticed in the apache Log Files that IE5 claims itself to be
> HTTP/1.1 comformant and Netscape says HTTP/1.0. Could this be
> the reason or am I missing something?
>
> A hidden line in my CGI-script looks like this:
> <input type="hidden" name="EXPIRED">
> and I use method="post" in my form tag.
>
> Any hints?
>
> Thanks,
>   Tamer.
>
> --
> ------------------------------------------------------------------------
>        -////-         | Tamer FAHMY, computer science student at the
>       --o o --        |              university of technology in Vienna
>      --- <  ---       |       aka: Tut-Ench-Tamer       irc: tuhtah
>     ---  v   ---      | eMail:    e9526976 at student.tuwien.ac.at
>      ---    ---       | home page: http://stud4.tuwien.ac.at/~e9526976
>  ^^^  --    --  ^^^   |        Support bacteria -- it's the only culture
> :) Tut-Ench-Tamer (:  |          some people have !    -- Linux fortunes
> ------------------------------------------------------------------------



From miles at caddr.com  Sat Apr 15 11:47:29 2000
From: miles at caddr.com (miles)
Date: Sat, 15 Apr 2000 15:47:29 GMT
Subject: Nasty Zope & Python Imaging Library name collision
References: <slrn8fg5pq.td6.miles@caddr.com> <KmYJ4.1126$rc9.190032896@newsb.telia.net>
Message-ID: <slrn8fh3p3.tsr.miles@caddr.com>

In article <KmYJ4.1126$rc9.190032896 at newsb.telia.net>, Fredrik Lundh wrote:
>miles <miles at caddr.com> wrote:
>> I've finally traced a mysterious failure in some Zope external methods
>> I've been working on to a module name collision between Zope.  In a
>> nutshell, they both contain an 'ImageFile' module and the wrong one
>> gets loaded.
>>
>> This is essentially what my external method does:
>>
>> import Image
>> i = Image.open(filename.jpg)
>> i = i.resize((x,y))
>> i.save(newfilename.jpg)
>>
>> This eventually results in the construction of a Jpeg image object which
>> is supposed to derive from ImageFile.ImageFile from the Imaging library.
>> Sadly, the __init__ method from Zope's ImageFile module is called instead,
>> which promptly goes berserk.
>>
>> Is there anything I can do to fix this?  How do people do deal with these
>> kinds of problems in general?  It seems to be very easy to inadvertently
>> collide with another module.  Perl avoids this by allowing arbitrarily
>nested
>> modules.  For example, in this case I'd have Zope::ImageFile and
>> Imaging::ImageFile and everything would be roses.
>
>assuming you have PIL 1.0, just make sure you've installed
>all the PIL files in a directory named PIL, and then do:
>
>    from PIL import Image
>
>instead of "import Image".
>
>if you have an older version, you may have to add an
>__init__.py file to the PIL directory, and explicitly import
>format handlers for the file formats you want to use (e.g.
>import PIL.JpegImagePlugin, PIL.GifImagePlugin)
>
></F>

That did the trick!  Thanks for your help & thanks for PIL!

-- 
miles at caddr dot com


From gmcm at hypernet.com  Mon Apr 24 10:02:49 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Mon, 24 Apr 2000 10:02:49 -0400
Subject: Tkinter+win32 Installer question
In-Reply-To: <8e1hg2$dq2$1@nnrp1.deja.com>
Message-ID: <1255537525-50974959@hypernet.com>

gregholmes at my-deja.com wrote:

> I am developing multiple programs using Tkinter and win32 extensions.
> 
> Is there a reliable, brute force way to use the Installer to make a
> distribution that will work? I realize I am not posting specifics here,
> but that is kind of the point; is there a general solution? The default
> behavior seems to exclude necessary things.
> 
> I don't mind reinventing the wheel if needed ;) but has this wheel
> already been invented?

You sure are short on specifics: I guess you're talking about 
my Installer. I also guess you're talking about the default 
exclude of win32api and PyWinTypes15.dll. Simply remove 
the offending excludes line from the config and run Builder on 
it, (or remove it from the templates in Simple, Standalone and 
or Freeze, whichever you're using).

They are excluded because most people don't specifically use 
them, but the framework uses the os module, which will 
include these. But os uses them only for the abspath function, 
and has a reasonable fallback if they're not installed. So for 
most users they would be excess baggage that they didn't 
ask for.


- Gordon



From robin at jessikat.demon.co.uk  Fri Apr 21 07:15:27 2000
From: robin at jessikat.demon.co.uk (Robin Becker)
Date: Fri, 21 Apr 2000 12:15:27 +0100
Subject: finding real $0
References: <OWeCyjAOWCA5Ew86@jessikat.demon.co.uk>
 <qlWL4.1038$v85.6158@news-server.bigpond.net.au>
Message-ID: <0H+c2XAPhDA5EwLw@jessikat.demon.co.uk>

In article <qlWL4.1038$v85.6158 at news-server.bigpond.net.au>, Mark
Hammond <mhammond at skippinet.com.au> writes
>Isnt sys.executable what you are after?
>
>Mark.
>
>"Robin Becker" <robin at jessikat.demon.co.uk> wrote in message
>news:OWeCyjAOWCA5Ew86 at jessikat.demon.co.uk...
>> Is there any pythonic way to locate the original argument zero of the
>> process? I mean the location of python itself not the script. This would
>> allow me to launch subscripts without having to rely on the path being
>> set up correctly.
>> --
>> Robin Becker
>
>
yes I'm just stupid.
-- 
Robin Becker


From darrell at dorb.com  Tue Apr 11 19:51:59 2000
From: darrell at dorb.com (Darrell)
Date: Tue, 11 Apr 2000 23:51:59 GMT
Subject: Filter
References: <955427129.619305@neptune.uniserve.ca>
Message-ID: <zUOI4.3561$wM3.31782@typhoon.nyroc.rr.com>

"Matthew" <shayman at uniserve.com> wrote in message
news:955427129.619305 at neptune.uniserve.ca...
> I need to write a filter that inserts tabs between Log values .  So that
> lines 1 to 3 come out looking something like lines 4 to 6.

import re, string

def tabIt(buf):
        # Markup that single letter
    buf=re.sub("(L11.*)([A-Za-z]) ","\g<1>\t\g<2>",buf)

    def twos(mo):
        """
        Add tabs to groups of two then remove extra tabs
        """
        g=mo.groups()
        g1=re.sub("(\w\w)","\t\g<1>\t",g[1])
        g1=re.sub("\t\t+","\t",g1)
        return g[0]+g1

    buf=re.sub("(L11)(.*)",twos,buf)
    return buf


print tabIt(open(fname).read())


--Darrell





From hniksic at iskon.hr  Tue Apr 11 11:35:40 2000
From: hniksic at iskon.hr (Hrvoje Niksic)
Date: 11 Apr 2000 17:35:40 +0200
Subject: [ANNOUNCE] Garbage collection for Python
References: <20000407221750.A29279@acs.ucalgary.ca> <38F13FB9.D1D16AF@mindspring.com> <slrn8f2if1.dop.nascheme@cranky.arctrix.com> <87r9cdhdb0.fsf@deneb.cygnus.argh.org> <FsunGI.BtB@world.std.com> <dphfd87s1h.fsf@mraz.iskon.hr> <Fsuuw7.Inu@world.std.com>
Message-ID: <dpk8i4646r.fsf@mraz.iskon.hr>

wware at world.std.com (Will Ware) writes:

> Hrvoje Niksic (hniksic at iskon.hr) wrote:
> > This is still different from being able to embed multiple Python
> > instances in threads.
> 
> Are we talking about multiple Python instances running in a single
> thread?

No, we're talking (I think) about multiple Python instances running in
a single process, each in its own thread.  To quote Florian:

    > Is it possible to extend your GC in a way which would permit the
    > execution of several bytecode interpreters in multiple threads?

That may be hard, but it's still easier than what you're talking about
-- you needn't handle context switches, etc.


From ajung at sz-sb.de  Fri Apr 28 08:41:19 2000
From: ajung at sz-sb.de (Andreas Jung)
Date: Fri, 28 Apr 2000 14:41:19 +0200
Subject: sleep function?
In-Reply-To: <8ebvud$l6h$1@nnrp1.deja.com>; from tiddlerdeja@my-deja.com on Fri, Apr 28, 2000 at 12:24:57PM +0000
References: <8ebvud$l6h$1@nnrp1.deja.com>
Message-ID: <20000428144118.B20818@sz-sb.de>

On Fri, Apr 28, 2000 at 12:24:57PM +0000, tiddlerdeja at my-deja.com wrote:
> Does python have a sleep function? e.g. sleep(60) to wait 60 seconds?
> 
> I realise this is a lame question, but I have no python documentation
> around just now.

Yeah - it?s quite lame :)

import time
time.sleep(60)

Andreas 



From mwh21 at cam.ac.uk  Mon Apr 17 20:38:17 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 18 Apr 2000 01:38:17 +0100
Subject: Another question
References: <dVFK4.41$vr2.5817@news.uswest.net> <38FB2EEB.A32AF88C@bam.com> <38FB339F.604364AD@bam.com> <200004171605.MAA04929@zippy.init.com> <38FB7610.80EAF965@ichips.intel.com> <200004172120.RAA11102@zippy.init.com> <jWNK4.432$vr2.32822@news.uswest.net>
Message-ID: <m3ln2cckg6.fsf@atrus.jesus.cam.ac.uk>

"Jeff Massung" <jmassung at magpiesystems.com> writes:

> Donald Beaudry wrote in message <200004172120.RAA11102 at zippy.init.com>...
> >and this one is better still (at least on my machine):
> >
> >  def do_it():
> >      from random import randint
> >      r = [0,] * 100000
> >      for i in xrange(100000):
> >         r[i] = randint(1,10)
> >  do_it()
> >
> 
> With the "from/import" being local, does the namespace go away after the
> function is done? How fast is importing (say I called do_it() 10000 times -
> then how efficient is your function?)

Pretty fast; two dictionary lookups (once in sys.modules, another in
random.__dict__), for strings that will have been interned.

For that last dreg of performance you might try:

import random
def do_it(randint=random.randint,xrange=xrange):
    r = [0,] * 100000
    for i in xrange(100000):
       r[i] = randint(1,10)
do_it()

But really, this is getting silly; if performance is this much of an
obsession you aren't going to have that much fun with Python.

Cheers,
M.

-- 
59. In English every word can be verbed. Would that it were so in 
    our programming languages.
     -- Alan Perlis, http://www.cs.yale.edu/~perlis-alan/quotes.html


From tom-main at REMOVEME.home.com  Thu Apr 13 22:17:38 2000
From: tom-main at REMOVEME.home.com (Tom)
Date: Fri, 14 Apr 2000 02:17:38 GMT
Subject: How do I undo an import??
References: <7QrJ4.185421$Hq3.4353581@news2.rdc1.on.home.com> <m3d7nthbbj.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <6dvJ4.186934$Hq3.4382652@news2.rdc1.on.home.com>

I tried the environment variable as you suggested, pointing to a file
containing the following:

print "about to ... from NetCfg import *"
from NetCfg import *

This partially works.  Now, when I run the python shell I see the printed
text twice, but the functions in NetCfg aren't added to the namespace.

Tom.

"Michael Hudson" <mwh21 at cam.ac.uk> wrote in message
news:m3d7nthbbj.fsf at atrus.jesus.cam.ac.uk...
> "Tom" <tom-main at REMOVEME.home.com> writes:
>
> > I'm a C++ programmer, new to Python, working with v1.5.2 on Windows.
> >
> > I run the command-line interpreter, then I type:
> >
> > >>> from NetCfg import *
> >
> > to import my extension.  Now I want to unload my NetCfg DLL without
exiting
> > the interpreter.  How do I do this?
>
> You don't, in general.
>
> del sys.modules['NetCfg']
>
> goes some of the way, but it's unlikely it goes far enough to let the
> dll be unloaded from memory.
>
> > Also, I type the above import command every time I start the
interpreter.
> > Is there some way to get have this command executed automatically?
>
> Set the environment variable "PYTHONSTARTUP" to point to a file
> containing commands you want executed.
>
> At least, that's what I do on Linux.
>
> Cheers,
> M.
>
> --
>   ... but I guess there are some things that are so gross you just have
>   to forget,  or it'll destroy something within you.  perl is the first
>   such thing I have known.                 -- Erik Naggum, comp.lang.lisp




From willadams at aol.com  Tue Apr 11 17:39:04 2000
From: willadams at aol.com (William F. Adams)
Date: 11 Apr 2000 21:39:04 GMT
Subject: Compiling Python Imaging under LinuxPPC
Message-ID: <20000411173904.18637.00002261@ng-cg1.aol.com>

I'm trying to do this, and not getting far.

Here's what I get:
.0 -L/usr/X11R6/lib -lX11 -lXext -lXmu -ltiff -ljpeg -lz -lXpm -0 _tkinter.so
/usr/bin/ld: cannot find -lBLT
collect2: ld returned 1 exit status
make: *** [_tkinter.so] Error 1
Bad exit status from /var/tmp/rpm-tmp.39884 (%build)

What's BLT?

Thanks!

William

-- 
William Adams
http://members.aol.com/willadams
Sphinx of black quartz, judge my vow.



From moshez at math.huji.ac.il  Mon Apr  3 18:37:35 2000
From: moshez at math.huji.ac.il (Moshe Zadka)
Date: Tue, 4 Apr 2000 00:37:35 +0200 (IST)
Subject: Python 1.6 alpha 1 released
In-Reply-To: <8catgc$5i84$1@uvaix7e1.comp.UVic.CA>
Message-ID: <Pine.GSO.4.10.10004040035420.21084-100000@sundial>

On 3 Apr 2000, John Aycock wrote:

> I'm not going to take sides here, but I coincidentally just finished
> reading something highly relevant:
<snipped>

I beg to differ: it's not highly relevant. What Guido did was correct
a *bug* in the implementation. That people depended on it is sad, and
just goes to show it should be corrected ASAP before more people depend
on it.

--
Moshe Zadka <mzadka at geocities.com>. 
http://www.oreilly.com/news/prescod_0300.html
http://www.linux.org.il -- we put the penguin in .com




From paro at ieee.org  Mon Apr 17 11:59:29 2000
From: paro at ieee.org (Parag Shah)
Date: Mon, 17 Apr 2000 15:59:29 GMT
Subject: Project idea - python
Message-ID: <BxGK4.811$9B.103178@news.bc.tac.net>

hi,
   i am planning to do a project for my bachelors where i wish to write a translator for a language (undecided at 
the moment ) which generates java bytecode. how do i go about doing it???
thanx in advance
cheers
parag
paro at ieee.org


==================================
Posted via http://nodevice.com
Linux Programmer's Site


From scarblac-spamtrap at pino.selwerd.nl  Wed Apr 26 04:48:44 2000
From: scarblac-spamtrap at pino.selwerd.nl (Remco Gerlich)
Date: 26 Apr 2000 08:48:44 GMT
Subject: Need advice performing a search
References: <3905F795.A536CFA7@stsci.edu> <3906944F.F718FE91@teleatlas.com>
Message-ID: <slrn8gdbak.bp5.scarblac-spamtrap@flits104-37.flits.rug.nl>

Klaus Bruns wrote in comp.lang.python:
> mybe try this:import re
> 
> file = open("file","r")
> 
> alllines = file.readlines()
> pattern1 = re.compile("first_string") ## first_string as regualr
> expression
> pattern2 = re.compile("secound_string") ## secound_string as regular
> expression
> 
> result = []
> 
> for line in alllines:
>     if pattern1.search(line) or pattern2.search(line):
>         result.append(line)
> 
> print result ## result is maybe ['line1','line2','line3']

I think using regular expressions is overkill for simply searching a
substring. Use string.find(string, substring) instead?

file = open("file","r")
alllines = file.readlines()
file.close()

from string import find
def testline(line):
  return find(line, "first_string") or find(line, "second_string")

print filter(testline, alllines)

Is another way to do it.
-- 
Remco Gerlich,  scarblac at pino.selwerd.nl
Hi! I'm a .sig virus! Join the fun and copy me into yours!


From tismer at tismer.com  Tue Apr 25 20:09:19 2000
From: tismer at tismer.com (Christian Tismer)
Date: Wed, 26 Apr 2000 02:09:19 +0200
Subject: Metaclasses?
References: <3905F4A0.2B39C705@darmstadt.gmd.de>
Message-ID: <390633AE.FB7FEA83@tismer.com>


Robb Shecter wrote:
> 
> Hi,
> 
> I read an old post about metaclasses (included on the cd-rom with the
> book "Mit Python programmieren"), and got very interested, because I
> imagine that by making a new metaclass, I could maybe fix the one thing
> about Python that annoys me: the obligatory use of "self" in method
> declarations and variable access.  (Python's syntax is so clean and neat
> in all other areas, I feel like it's a shame that there's excess
> verbosity in this one place.)

No-no. No? No!
There are reasons why people might want to use metaclasses.
But not just to get rid of that verbose "self" argument.

I'd suggest to stay with both the shame and classes as they
are now. This is the cheapest possible way here: Change your
mind about self --
-- it is not *the* problem. it is *really* no problem.

The problem is very much different and harder.
Ask me via private mails if you don't get it.

ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaunstr. 26                  :    *Starship* http://starship.python.net
14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     where do you want to jump today?   http://www.stackless.com



From richard at cogsci.ed.ac.uk  Wed Apr  5 12:40:15 2000
From: richard at cogsci.ed.ac.uk (Richard Tobin)
Date: 5 Apr 2000 16:40:15 GMT
Subject: Why should I switch to Python? - Infinity of Primes
References: <NDBBKEGCNONMNKGDINPFGEJDDFAA.infonuovo@email.com> <7F51E902E09F3F60.8F612607812ED16B.82CB8B41435E14FD@lp.airnews.net> <38EA369C.26A1C25@math.okstate.edu> <38EAC8C4.ADC25640@cosc.canterbury.ac.nz>
Message-ID: <8cfq9f$2q5e$1@pc-news.cogsci.ed.ac.uk>

In article <38EAC8C4.ADC25640 at cosc.canterbury.ac.nz>,
Greg Ewing  <greg at cosc.canterbury.ac.nz> wrote:

>Um, no it doesn't - it constructs a number which is
>*either* prime *or* divisible by some other prime bigger
>than the one you started with.

It trivially gives you a procedure for finding a larger prime: you
just find the factors of the constructed number (eg by trying all
possible divisors), and if there are any then one will be a new prime,
otherwise the constructed number is a new prime.

-- Richard
-- 
Spam filter: to mail me from a .com/.net site, put my surname in the headers.

"The Internet is really just a series of bottlenecks joined by high
speed networks." - Sam Wilson


From dgoodger at bigfoot.com  Wed Apr 12 23:21:54 2000
From: dgoodger at bigfoot.com (David Goodger)
Date: Wed, 12 Apr 2000 23:21:54 -0400
Subject: Python idiom: Multiple search-and-replace
References: <20000412100816.A2431031@vislab.epa.gov><5C%I4.645$rc9.190288896@newsb.telia.net> <20000412120012.D2446939@vislab.epa.gov><0s1J4.655$rc9.190631424@newsb.telia.net>
Message-ID: <B51AB591.1BAC%dgoodger@bigfoot.com>

on 2000-04-12 12:25, Fredrik Lundh (effbot at telia.com) wrote:
> while you're at it, try replacing the original readline loop with:
> 
> while 1:
> lines = fp.readlines(BUFFERSIZE)
> if not lines:
> break
> lines = string.join(lines, "")
> lines = re.sub(...)
> out_fp.write(lines)
> 
> where BUFFERSIZE is 1000000 or so...

why not just,

    while 1:
        lines = fp.read(BUFFERSIZE)
        if not lines:
            break
        lines = re.sub(...)
        out_fp.write(lines)

? Saves the string.join() step. Or am I missing something? (I await
enlightenment...)

BTW, excellent idiom (especially the first one, represented by '...' above).
I can put it to immediate use. You've just sold another copy of your book
... as soon as it's Mac-friendly. And can I copy & paste from it? If not,
maybe I'll wait for the tree-killing edition.

-- 
David Goodger    dgoodger at bigfoot.com    Open-source projects:
 - The Go Tools Project: http://gotools.sourceforge.net
 (more to come!)



From meh9 at cornell.edu  Thu Apr  6 10:44:54 2000
From: meh9 at cornell.edu (Matthew Hirsch)
Date: Thu, 06 Apr 2000 10:44:54 -0400
Subject: More about variables
Message-ID: <meh9-5C17A0.10445406042000@news.cit.cornell.edu>

Hi All,

Let's say I had something like:

number_of_variables=int(raw_input('Enter number of variables: '))

if number_of_variables==1:
   variable1=[]
elif number_of_variables==2:
   variable1=[]
   variable2=[]
elif number_of_variables==3:
   variable1=[]
   variable2=[]
   variable3=[]
elif number_of_variables==4:
   variable1=[]
   variable2=[]
   variable3=[]
   variable4=[]
elif number_of_variables==5:
   variable1=[]
   variable2=[]
   variable3=[]
   variable4=[]
   variable5=[]
else:
   print 'else'

Is there a more efficient way of doing this? What if I wanted to create 
100 variables (lists).  It doesn't make sense to write an if statement 
for that many conditions.

Thanks for your help,
Matt


From rcunning at acm.org  Tue Apr  4 01:31:14 2000
From: rcunning at acm.org (Robert W. Cunningham)
Date: Tue, 04 Apr 2000 05:31:14 GMT
Subject: Advice requested: GUI project beginning
References: <m12avmV-000wcEC@swing.co.at> <slrn8eiuu1.b3t.warlock@raven.myth>
Message-ID: <38E97E29.2687864D@acm.org>

Jim Richardson wrote:

> On Fri, 31 Mar 2000 09:23:47 +0200,
>  Christian Tanzer, in the persona of <tanzer at swing.co.at>,
>  brought forth the following words...:
>
> >
> >"Russell E. Owen" <owen at astroNOJNK.washington.edu.invalid> wrote:
> >
> >> I am starting work on a telescope remote control user interface. It's a
> >> big program with lots of windows and controls and display graphics
> >> supporting the telescope and numerous instruments.
>
> Not discourage you from writing the app yourself, but there is a
> linux Telescope control app at http://206.169.225.102/eots.html
> certainly worth a look at least. I know nothing of the app itself, don't
> know if it would fit the bill for you.

Programming?  Telescope control?  Doesn't this thread belong in a Forth NG?

I mean, wasn't Forth CREATED to control a telescope?

-BobC




From tseaver at starbase.neosoft.com  Sat Apr 15 14:24:46 2000
From: tseaver at starbase.neosoft.com (Tres Seaver)
Date: 15 Apr 2000 13:24:46 -0500
Subject: TCP/IP protocols
References: <ZSMJ4.1099$Y14.27933@news.uswest.net>
Message-ID: <5911665EC85DE246.5C7B608EA19D735B.E388A3CBF3EE1E96@lp.airnews.net>

In article <ZSMJ4.1099$Y14.27933 at news.uswest.net>,
Jeff Massung <jmassung at magpiesystems.com> wrote:
>I'm new to Python programming, but am very experienced in C/C++. Could
>someone point me to a good reference for TCP/IP programming with Python
>(pdf, book, anything :]) thanks.
>
>Jeff
>
>

Aaron Watters, Guido van Rossum, James Ahlstrom. _Internet Programming
      with Pyton_. M&T Books, 1996. ISBN 1-55851-484-8.

Now out-of-print (I think) -- you may be able to find it through the net.
-- 
---------------------------------------------------------------
Tres Seaver        tseaver at digicool.com     http://www.zope.org
Digital Creations


From chapman at bioreason.com  Thu Apr 20 16:30:17 2000
From: chapman at bioreason.com (Mitch Chapman)
Date: Thu, 20 Apr 2000 14:30:17 -0600
Subject: Good style for multi-valued returns
References: <889A6EB6AE399393.619B0539BFCAAE0F.5C6D110C3DA908F7@lp.airnews.net>
Message-ID: <38FF68D9.6D23FB38@bioreason.com>


Cameron Laird wrote:
> 
> Examples of multi-valued returns:  (head, tail); (quotient,
> remainder); (left, right); (prologue, body, epilogue);
> (data, code); (CAR, CDR); (count, size); (handle, status);
> (decomposition; remainder); ...
> 
> Let me guess:  folks generally pass tuples for small and
> determinate-in-number multi-values, and dictionaries other-
> wise, 'cept some do it all with tuples.  Accurate?

If there are many returned values, I tend to define a new class 
for the return type rather than returning a dictionary.  This
usually pays off, because the consumers of those return values 
often want to perform the same sorts of operations on the values.  
If I start off representing the return value as a dictionary,
clients duplicate the logic for those operations.  If I start off
with a class for the new return type, those operations get
expressed once, as methods of the new class.

--
Mitch Chapman
chapman at bioreason.com


From smalleys at gte.net  Sun Apr 30 17:18:20 2000
From: smalleys at gte.net (sue)
Date: Sun, 30 Apr 2000 21:18:20 GMT
Subject: Reading in strings -> numbers ??
References: <300420001033289484%pecora@anvil.nrl.navy.mil>
Message-ID: <390CA332.E8EA45B6@gte.net>

Check the HOWTO and topic guides for databases.  I remember
references to table.py, csv.py and a bunch of others.

Sue.


From hamish_lawson at yahoo.co.uk  Fri Apr  7 12:24:17 2000
From: hamish_lawson at yahoo.co.uk (Hamish Lawson)
Date: Fri, 07 Apr 2000 09:24:17 -0700
Subject: Migrating Tcl system to Python
Message-ID: <095cc48a.7e9b36dd@usw-ex0108-063.remarq.com>

I'd be interested in anyone's experience of migrating a (several-
thousand line) system written in Tcl to Python. And since we'll
have to deploy new functionality while we're in the process of
migrating, I'd particularly like to hear of experiences with
having Tcl code using Python code (and vice versa). I'm
considering XML-RPC/SOAP as a possible way to do that, but what
other approaches might there be?

Hamish Lawson


* Sent from RemarQ http://www.remarq.com The Internet's Discussion Network *
The fastest and easiest way to search and participate in Usenet - Free!



From cbbrowne at news.hex.net  Thu Apr 27 21:04:51 2000
From: cbbrowne at news.hex.net (Christopher Browne)
Date: Fri, 28 Apr 2000 01:04:51 GMT
Subject: The Simple Economics of Open Source
References: <B529350E.51B9%raffael@mediaone.net> <390534BB.3D6CEBDF@libc.org> <raffael-01CD89.20575826042000@news.ne.mediaone.net> <3907AD52.FC0630FD@tcac.net>
Message-ID: <Ts5O4.1292$F.37918@news4.giganews.com>

Centuries ago, Nostradamus foresaw a time when Albert Wagner would say:
>Raffael Cavallaro wrote:
>> 
><snip>
>> I think that many open source advocates have
>> failed to apply simple logic to the economics of open source, believing
>> that somehow, the internet, or software, is immune from simple laws of
>> supply and demand. They are not.
>> 
>I am going to have to go back and reread all of your previous posts. 
>You seem to be rebutting a point that was never made.  Of course, Open
>Source software is not immune to the laws of supply and demand.  Who
>said it was?  It is in response to those laws that it is free.  But free
>only in the world of economics where dollar worth is all that matters.
>It is this very world, where the only motivation taken into account is
>money profit, that Open Source has escaped.  And the escape has been
>dazzling.  

_Economics_ is not the study of money.

_Finance theory_ is the study of money.

Economics, in contrast, has to do with the study of the allocation of
resources.

The theory generally regards the allocation of _scarce_ resources.

It is entirely common for the basic "denomination" to be treated as
money, but the analytical tools economics provides are equally
applicable when the "coin of the realm" is the _time developers spend_
as it is when the "coin" represents dollars and cents.
-- 
Rule of Scarcity
"If it's not allocated by a market, then it's more expensive than
money."
-- Mark Miller
cbbrowne at ntlug.org- <http://www.ntlug.org/~cbbrowne/lsf.html>


From nmm1 at cus.cam.ac.uk  Mon Apr 10 04:36:22 2000
From: nmm1 at cus.cam.ac.uk (Nick Maclaren)
Date: 10 Apr 2000 08:36:22 GMT
Subject: Python paradigms
References: <8cmr70$b0e$1@pegasus.csx.cam.ac.uk> <8covpn$4jq$1@slb6.atl.mindspring.net> <8cqk7u$dmc$1@pegasus.csx.cam.ac.uk> <38F13CA1.1A91D85@mindspring.com>
Message-ID: <8cs3q6$l4e$1@pegasus.csx.cam.ac.uk>

In article <38F13CA1.1A91D85 at mindspring.com>,
Chuck Esterbrook  <echuck at mindspring.com> wrote:
>Nick Maclaren wrote:
>
>> To the several people who suggested:
>>
>>     (a() and [b()] or [c()])[0]
>>
>> Thank you for the suggestion.  I have occasionally used perversities
>> like that (such as '*(x ? &a : &b) = y;' in C to mimic the Algol 68
>> facility), but I don't regard them as clarifying the code, either!
>
>OK, so what does the Algol 68 code look like for that C code and any other interesting cases?

If I remember correctly (and I may well be misremembering), an Algol 68
unit could return a name and not just an expression - i.e.:

    (a | b | c) := y;

or:

    if a then b else c fi := y;

However, I may be getting confused with which language it was.  As I
said, it is not a construction that I regard as particularly clear,
and therefore almost never used it.  I use the expression form very
heavily, in any language that has it, because so many operations are
precisely that construction.


Regards,
Nick Maclaren,
University of Cambridge Computing Service,
New Museums Site, Pembroke Street, Cambridge CB2 3QG, England.
Email:  nmm1 at cam.ac.uk
Tel.:  +44 1223 334761    Fax:  +44 1223 334679


From karl-sve at dsv.su.se  Mon Apr  3 12:33:22 2000
From: karl-sve at dsv.su.se (Kalle Svensson)
Date: Mon, 3 Apr 2000 18:33:22 +0200
Subject: Platf. indep. console lib.
In-Reply-To: <38E8AF93.70F4CCF7@ansic.net>
References: <8c03u1$igk$1@news.hit.fi> <38E8AF93.70F4CCF7@ansic.net>
Message-ID: <Pine.LNX.4.10.10004031747160.18441-100000@l166.dsv.su.se>

On Mon, 3 Apr 2000, Scott wrote:

> There is an slang module which wraps the slang library.  It's available
> for Unix and Win32 though I haven't had the time to actually use it
> yet.  Here is an ftp site for it: 
> ftp://quasar.ipa.nw.ru/pub/avv/Python/pyslang .  If you give it a try,
> let me know how it turns out!

It's been a while since I used it, but at the time it wasn't perfectly
cross-platform, (for example, I couldn't get the arrow keys to work in
Win98 with a Swedish keyboard).
Anyway, the program I wrote can be found on 
http://www.dsv.su.se/~karl-sve/tdsweeper.shtml but beware, it's ugly...

HTH,
  Kalle

--

"When faced with a problem, some people think.
Now they have three problems."  -- The timbot

/------------------------------------------------------\
| karl-sve at dsv.su.se   http://www.dsv.su.se/~karl-sve/ |
| Proud member of Technological Dimensions:            |
|                 http://hem.passagen.se/tech/?noframe |
\------------------------------------------------------/



From clee at thalamus.wustl.edu  Fri Apr 28 11:12:42 2000
From: clee at thalamus.wustl.edu (clee at thalamus.wustl.edu)
Date: Fri, 28 Apr 2000 15:12:42 GMT
Subject: memory leak debugging in C modules
Message-ID: <8ec9oq$k7$1@nnrp1.deja.com>

I'm looking for advice on fixing memory leaks in a python C module.
How do the C modules gurus out there do it?

Thanks in advance for any help,

-chris lee

P.S.  What I've done so far:

- working on a linux machine primarily
- using gcc with -g flag
- debugging with gdb python <testscript.py>
- I've done some preliminary experiments with linking my module with a
utility called electric fence, and I've tried running my module under
the gnome memory checker, memprof.  Both of these options show some
promise but clearly give problematic results.  (I won't go into all the
problems---I suspect the cause of many of them may be that my python
interpreter is compiled with the --with-threads options.  I'm going to
recompile python next.)




Sent via Deja.com http://www.deja.com/
Before you buy.


From tjg at avalongroup.net  Tue Apr  4 18:35:22 2000
From: tjg at avalongroup.net (Timothy Grant)
Date: Tue, 04 Apr 2000 15:35:22 -0700
Subject: Two questions
References: <38EA5FE8.853E9B75@callware.com> <38EA6508.90BF630C@roguewave.com> <38EA6E87.40EDB542@callware.com> <38EA6422.61054B11@exceptionalminds.com> <38EA7C18.8F3150D1@roguewave.com>
Message-ID: <38EA6E2A.B7ACEF91@exceptionalminds.com>

Bjorn Pettersen wrote:

> > mxODBC works beautifully, but I've never used it to access and access
> > database, only SQL database. Perchance do you know how Windows
> > advertises an Access database to an ODBC driver?
> [snip]
 
> In the control panel there should be a Odbc 32 control (should be
> installed with ms access).  It has a simple point-and-click interface
> where you have to choose the ms access driver, give the filename for the
> database, and choose a DSN (Data Source Name?).  The DSN is what you use
> to connect with.

I just fired up my trusty VMWare and figured this out. I wasn't thinking
about the driver interface being setup differently for different types
of drivers, but sure enough it was. Every-so-often this Windows stuff
actually surprises me in a positive sense.

-- 
Stand Fast,
    tjg.

Timothy Grant                         tjg at exceptionalminds.com
Chief Technology Officer              www.exceptionalminds.com
Red Hat Certified Engineer                      (503) 246-3630
Avalon Technology Group, Inc.              fax  (503) 246-3124
>>>>>>>>>>>>Linux...Because rebooting isn't normal<<<<<<<<<<<<



From glyph at twistedmatrix.com  Sat Apr 22 17:24:39 2000
From: glyph at twistedmatrix.com (Glyph Lefkowitz)
Date: 22 Apr 2000 16:24:39 -0500
Subject: The Simple Economics of Open Source
In-Reply-To: m.faassen@vet.uu.nl's message of "22 Apr 2000 14:18:41 GMT"
References: <14591.15630.192472.440375@goon.cnri.reston.va.us> <slrn8fv2rd.1kt.neelk@brick.cswv.com> <3900611F.3B525BDA@cfdrc.com> <FtDG4z.774@world.std.com> <slrn8g26tu.p7o.neelk@brick.cswv.com> <22042000.2@sanctum.jae.ddns.org> <ja1z3yog61.fsf@helix.twistedmatrix.com> <8dscc1$n1o$2@newshost.accu.uu.nl>
Message-ID: <jar9bxygko.fsf@helix.twistedmatrix.com>

m.faassen at vet.uu.nl (Martijn Faassen) writes:

> Glyph Lefkowitz <glyph at twistedmatrix.com> wrote:
> 
> [snip]
> > For example, if you run across some dusty two-year-old python code
> > that (f'rinstance) facilitates time-travel, and you hack on it for a
> > while, bring it up to date, but get bored or disinterested because of
> > more pressing concerns, like the upcoming alien invasion, you can put
> > your code out there and somebody else with a like mind might come
> > along and pick up where you left off.
> 
> Heh. *Everybody* reads those off-topic threads, don't they? :)

Off-topic?  I thought they were the point of the newsgroup!! They're
better than most net comics... is someone going to be publishing a
"guido mythos" compilation anytime soon? :)

-- 
                  __________________________________________
                 |    ______      __   __  _____  _     _   |
                 |   |  ____ |      \_/   |_____] |_____|   |
                 |   |_____| |_____  |    |       |     |   |
                 |   @ t w i s t e d m a t r i x  . c o m   |
                 |   http://www.twistedmatrix.com/~glyph/   |
                 `__________________________________________'



From ullrich at math.okstate.edu  Sun Apr 23 14:02:33 2000
From: ullrich at math.okstate.edu (David C. Ullrich)
Date: Sun, 23 Apr 2000 18:02:33 GMT
Subject: code not true?
References: <Pine.GSO.4.10.10004230946161.17585-100000@sundial>
Message-ID: <39033a68.436738407@news>

On Sun, 23 Apr 2000 09:47:00 +0300 (IDT), Moshe Zadka
<moshez at math.huji.ac.il> wrote:

>On Sun, 23 Apr 2000, Kragen Sitaker wrote:
>
>> How many decimal digits are needed to make sure that 3 * (1/3) is
>> indistinguishable from 1?
>
>moshe-bombei:/home/moshe>python
>Python 1.5.2 (#1, Feb 21 2000, 14:52:33)  [GCC 2.95.2 19991024 (release)]
>on sunos5
>Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>>> (1.0/3)*3 == 1
>1
>
>
>Less then 17

	Aha. And now how many digits do we need for
x=1.0/3 to ensure that all calculations we do in the
future with x will come out right?

DU



From tim_one at email.msn.com  Wed Apr  5 02:07:38 2000
From: tim_one at email.msn.com (Tim Peters)
Date: Wed, 5 Apr 2000 02:07:38 -0400
Subject: Why should I switch to Python? - Infinity of Primes
In-Reply-To: <38EAC8C4.ADC25640@cosc.canterbury.ac.nz>
Message-ID: <000d01bf9ec5$3fb90820$3e2d153f@tim>

[Greg Ewing]
> ...
> If someone actually came up with a formula for constructing
> primes, it would be rather large news -- isn't that one of
> the Big Unsolved Problems?

Several such formulas were developed in the 70's, although they're strained.
Paulo Ribenboim's accessible "The Little Book of Big Primes"
(Springer-Verlag, 1991) gives a bunch of 'em in chapter 3, along with 6
other proofs of the infinitude of the primes.

not-that-archimedes-plutonium-will-be-fooled-ly y'rs  - tim





From thomas at xs4all.net  Mon Apr 17 03:47:38 2000
From: thomas at xs4all.net (Thomas Wouters)
Date: Mon, 17 Apr 2000 09:47:38 +0200
Subject: building on linux w/o threads?
In-Reply-To: <8ddokt$65t$1@nnrp1.deja.com>; from firefly64@my-deja.com on Mon, Apr 17, 2000 at 01:16:24AM +0000
References: <8ddokt$65t$1@nnrp1.deja.com>
Message-ID: <20000417094738.E15664@xs4all.nl>

On Mon, Apr 17, 2000 at 01:16:24AM +0000, firefly64 at my-deja.com wrote:

> I've done everything I can think of, including
> ./configure --without-thread and editing the
> Makefiles to include -DWITH_THREAD=0.

Dont do that :) Just do a 'make distclean' (possibly backupping your
Modules/Setup file first, if you dont want to edit it again) and do
./configure without *any* thread related options (you can use --prefix= and
such, of course.)

> The build process still always either produces a python linked to
> libpthreads or dies complaining about not being able to find various
> thread functions. (Yes, I've done make clean).

The latter, complaints about the thread functions, is probably because you
changed the WITH_THREAD option in the Makefile without editing the
Modules/Setup.thread file. If you disable threading by hand, be sure to
comment the line about the thread module, in that file. (It's the only
uncommented line there ;)

Make distclean && ./configure should do the trick, though.

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From effbot at telia.com  Thu Apr 20 04:33:07 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Thu, 20 Apr 2000 08:33:07 GMT
Subject: help learning "re"
References: <38FE6E5A.E7673A83@visionart.com>
Message-ID: <7hzL4.3189$Za1.47032@newsc.telia.net>

Pete J Shinners <pete at visionart.com> wrote:
> i'm trying to get the hang of these regular expression
>
> i'm trying to write a function that takes two filenames
> and copies one to the other. while it is copying i want
> it to scan for any environment variable names and substitute
> the expanded value.

os.path.expandvars(name)

> sounded like a fabulous time to learn the regexp.
> sadly it doesn't work. the code runs, but none of the
> variables are found or substituted
>
> here's what i expected to work
> ----------------------------------------------
> import os, re
>
>
> def subvar(regfind):
>     "callback for regex substitution"
>     var = regfind.group(0)[1:]
>     print 'REGEX found var:', var #temp debugging. never gets here
>     return os.environ.get(var, '$'+var)
>
>
> def smartcopy(source, dest):
>     "copies files. substites env.variables with values"
>     srcfile = open(source, 'r')
>     dstfile = open(dest, 'w')
>     regexp = re.compile(r"\$[0-9A-Z_]+")
>     while 1:
>         line = srcfile.readline()
>         if not line: break
>         regexp.sub(subvar, line)

regexp.sub returns the new value, it doesn't modify
line in place.

>         dstfile.write(line)
>
> -----------------------------------------------
> surely this can be done correctly, but how?

the only bug I can spot is the regexp.sub return value.  the
expression and the substitution function works just fine for
me...

(try adding "print regexp.sub(subvar, '$HOME')" just after
the re.compile to see what I mean)

</F>




From aa8vb at yahoo.com  Wed Apr 12 12:00:12 2000
From: aa8vb at yahoo.com (Randall Hopper)
Date: Wed, 12 Apr 2000 12:00:12 -0400
Subject: Python idiom: Multiple search-and-replace
In-Reply-To: <5C%I4.645$rc9.190288896@newsb.telia.net>
References: <20000412100816.A2431031@vislab.epa.gov>
 <5C%I4.645$rc9.190288896@newsb.telia.net>
Message-ID: <20000412120012.D2446939@vislab.epa.gov>

Fredrik Lundh:
 |Randall Hopper <aa8vb at yahoo.com> wrote:
 |> Is there a Python feature or standard library API that will get me less
 |> Python code spinning inside this loop?   re.multisub or equivalent? :-)
 |
 |haven't benchmarked it, but I suspect that this approach
 |is more efficient:
 |
 |<re.sub algorithm with replacement function>

Thanks!  It's much more efficient.  The 140 seconds original running time
was reduced to 11.6 seconds.  I can certainly live with that.

-- 
Randall Hopper
aa8vb at yahoo.com



From lvw at lfpt.rwth-aachen.de  Fri Apr 28 06:25:15 2000
From: lvw at lfpt.rwth-aachen.de (Lars von Wedel)
Date: Fri, 28 Apr 2000 12:25:15 +0200 (MET DST)
Subject: Using BaseHTTPServer
In-Reply-To: <Pine.GSO.4.10.10004281007030.10249-100000@sundial>
References: <200004280649.IAA00894@kazul.rwth-aachen.de>
	<Pine.GSO.4.10.10004281007030.10249-100000@sundial>
Message-ID: <200004281025.MAA02435@kazul.rwth-aachen.de>

Hello,

 > > Within any of the do_GET methods, how do I access data that has been
 > > sent by the client (...&data1=val1?data2=val2?...)?
 >
 > Use the CGI module <wink>. Just copy the code from CGIHTTPServer to create
 > the dictionary, (not as os.environ), and use the cgi module to parse
 > that dictionary.

I have been looking into CGIHTTPServer.py a little. I think, it should
be possible to modify run_cgi() (in a subclass) in the following way:

 - Decode self.cgi_info as shown additionally use cgi.parse_qs
   to write results (directory with CGI arguments) into
   e.g. self.cgi_args

 - fork the process (same as original run_cgi())

 - Associate stdin with self.rfile and stdout with self.wfile

 - Instead of running a script with name 'scriptname'
   call a method on 'self' called 'scriptname'.

 - The method called can access self.cgi_args to access
   the data passed and it can simply write to stdout
   using print.


Lars





From lewst at yahoo.com  Wed Apr  5 01:07:09 2000
From: lewst at yahoo.com (lewst)
Date: Tue, 4 Apr 2000 22:07:09 -0700 (PDT)
Subject: Starving for an Advanced Python Book
Message-ID: <20000405050709.25794.qmail@web2106.mail.yahoo.com>

Willie Peloquin <wpeloqui at ismi.net> wrote:

> Try "Python and Tkinter Programming" by John E. Grayson. It is an
> excellent book, full of useful information on using Python in the
> real world.

I already own this one, and I agree it's a really comprehensive and
nicely organized book.  Just make sure you are fairly-to-very
comfortable with Python before taking it on.  It jumps right into some
(IMO) complicated example code and doesn't let up.  I figure this text
will become more useful to me as I become more comfortable and
experienced with Python itself.


__________________________________________________
Do You Yahoo!?
Talk to your friends online with Yahoo! Messenger.
http://im.yahoo.com



From andymac at bullseye.apana.org.au  Thu Apr  6 06:44:41 2000
From: andymac at bullseye.apana.org.au (Andrew MacIntyre)
Date: Thu, 6 Apr 2000 20:44:41 +1000 (EST)
Subject: Tkinter and C extensions
In-Reply-To: <m3pus4w4cz.fsf@greebo.nodomain.de>
Message-ID: <Pine.OS2.3.95.1000406201722.1213B-100000@CENTRAL>

On 6 Apr 2000, Bernhard Herzog wrote:

> Andrew MacIntyre <andymac at bullseye.apana.org.au> writes:
> 
> [...]
> > My research has turned up ways of determining these via Tk's API
> > (Tk_Display(), Tk_Window() in tk.h) provided I can extract the tkwin
> > handle from Tkinter's canvas widget - something I've yet to figure out how
> > to do from within a C extension.
> > 
> > I'm wondering whether its possible to get at this information via Tkinter,
> > which would simplify the interactions between the various components.
> 
> Tkinter widgets have a tk attribute which is basically a tcl
> interpreter. It's interpaddr() method returns the TclInterp* pointer as
> a python int object. Just pass that and the name of the widget to a
> C-extension function and the rest follows from that...

Thanks for this thought - it fills in another gap.

After more research and another read of Fredrik's Tkinter docs, I've
reached the conclusion that its only the X Display pointer thats a problem
- the Window handle is available from the widget's winfo_id() method.  Oh,
and a Frame would be better than a Canvas for the intended purpose too.

AFAICT the only way to convert a widget's name to a pointer to its
Tk_Window structure, so that the Display pointer can be retrieved, 
requires a pointer to _another_ Tk_Window structure in the same widget
tree :-( Somewhat of a recursive problem when just one such pointer would
probably give me the info I require....

This is probably one of the very few situations where Tkinter's "arms
length" usage of Tk is an impediment.

I've still a couple of ideas to follow up though....

> > Any hints on additional sources of enlightment, such as other modules
> > which have similarities, would be appreciated.
> 
> Sketch's paxmodule does something similar. Sketch uses it to implement
> widgets in Python, which allow you to handle expose events yourself and
> use Xlib functions to draw the widgets.

Thanks for this suggestion as well.

BTW, it was SVGAmodule rather than GDmodule whose link from the Vaults of
Parnassus led to a dead end.

--
Andrew I MacIntyre                     "These thoughts are mine alone..."
E-mail: andrew.macintyre at aba.gov.au    (work) | Snail: PO Box 370
        andymac at bullseye.apana.org.au  (play) |        Belconnen  ACT  2616
        andymac at pcug.org.au           (play2) |        Australia




From gmcm at hypernet.com  Wed Apr 12 14:29:16 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Wed, 12 Apr 2000 14:29:16 -0400
Subject: static attributes & __getattr__ override
In-Reply-To: <38f4b2e0@news.raccoon.com>
Message-ID: <1256558337-47149947@hypernet.com>

Vadim Suvorov wrote:

> I am trying to redefine class attributes access, to include static
> attributes definitions from class (shared by all class instances). It is ok
> for instance to shade class attribute. So I wrote
> 
> class MyClass:
>     var1 = 1
>     var2 = 2
>     def __init__(self, x3):
>         self.var3 = x3
> 
>     def __getattr__(self, name):
>         return self.__class__.__dict__.get(name, None)
> 
> MyObj = MyClass(3)
> print MyObj.var1, MyObj.var2, MyObj.var3, MyObj.var4
> 
> (expected result) >>> 1 2 3 None
> 
> It works fine for attributes. 

Are you sure? Did you realize that only MyObj.var4 invokes 
__getattr__?

> However, I can not access standard methods of
> MyClass. I resorted to define __str__ and __repr__ (to be able to 'print
> MyObj'), and even then I could not use MyObj <> None which really hurts me.
> 
> Can you advice me on best way to proceed, and explain briefly what is going
> on? I do not think it is a bug, I probably missing something in the way
> attributes/methods work.

"obj.var" is the same as "getattr(obj, 'var')". Where obj is of 
InstanceType, that involves looking in the instance __dict__, 
the obj.__class__.__dict__, and then a search of the 
__dict__s of obj.__class__.__bases__. Only if all of that fails 
is the __getattr__ method tried.

Methods normally live in obj.__class__.__dict__, or in the 
__dict__ of one of the classes in __bases__. The only 
difference between a variable and a method is that the getattr 
code, upon finding it, will magically wrap it up into a bound 
method (squirrelling away the "self" pointer where the code 
object can make use of it).

- Gordon



From effbot at telia.com  Tue Apr  4 11:53:09 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Tue, 04 Apr 2000 15:53:09 GMT
Subject: Python 1.6 alpha 1 released
References: <Pine.GSO.4.10.10004040035420.21084-100000@sundial> <5E877FD7637AB46E.DABBED9CF3C1D64B.CDDDC05AE7374A78@lp.airnews.net> <8cca93$9oh$1@nnrp1.deja.com> <B964E942A897FF5F.529887C74CF2FC7E.977238F9475EAC23@lp.airnews.net>
Message-ID: <FdoG4.5023$74.81126@newsc.telia.net>

Tres Seaver wrote:
> >not true.  the documented interface explicitly says that
> >there is one single argument.  it does not mention optional
> >arguments.
>
> Your insistence that the "documented interface" consists only of the
> description in the module reference, and not of the example on the next
> page or of the way the API is actually used in the standard libraries
> baffles me.

I thought I explained that in the next sentence:

> >(the example was broken, but so were other examples in earlier
> >releases of the documentation.  see other messages in this
> >thread for details)

> >> No Python module in the standard library uses the single-argument
> >> connect, period.
> >
> >*all* python modules in the 1.6 standard library use
> >single-argument socket connect.
>
> Invalid argument -- nobody depends on the 1.6 library yet.

umm.  if you plan to use the 1.5.2 standard library with 1.6,
you're in for some *serious* surprises...

> Here is a check from my (unfiddled) Python distro -- its the same on
Windows,
> too:

[only of few of the examples are socket.connect calls...]

> So don't tell me that the 1.5.2 library used one-argument connect().

didn't think I did.  I thought you told me that *all* connect
calls in the standard library used two arguments, but maybe
you didn't.  in either case, as you just noticed yourself, it
wasn't true.

> The append() change is hard to find, but fixes something that shouldn't
work.

why shouldn't that work?  after all, lots of sample code use that
form.  library code use that form (popular libraries like PIL, Medusa,
Zope, just to name a few).  books use that form.

and as I've said a million times, code using "append" is much harder
to fix, since "append" is 1) far more likely to be used in contexts
where a TypeError is expected, but for other reasons, and 2) often
used as a bound method in performance critical code.  the former
makes it hard to find this by running the code, the second makes
it hard to find it by grepping through the sources.

I suggest you dig up my list of "change cost" criterias, and apply
them to all non-compatible changes (possibly starting with 1.5.2's
class object change, which broke *lots* of code in really naughty
ways [1]).  I've done that, and I still think that "connect" is the
*least* critical of all code-breaking changes in the python core,
since 1.3 or so.

I still don't understand why others seem to think that this is the
most critical of them all, and nobody seem to be capable of ex-
plaining that to me.

(I guess some people will change their minds when other changes
in 1.6 causes their code to explode.  are you sure your code doesn't
use multi-argument appends, or expect str(1L) to include the L, or
expect certain operations to raise exceptions if given longs instead
of integers, etc.  and if you think you are, are you sure all non-std
libraries you depend on doesn't do that?  thought so ;-)

> In fact, I think one could make a reasonable case that the real bug here
> is in the API doc:  single-argument connect() is unintuitive for AF_INET
> sockets, and exists primarily to shoehorn orthagonality with AF_UNIX ones
> (where a default port argument would do just as well.)

if you look at the socket API a bit more carefully, you'll notice that
there are a number of functions in there which can *only* deal with
(host, port) tuples.  so in other words, it's internal API consistency
we're talking about here:  if someone returns an address, it's reason-
able to be able to pass it on to anyone expecting an address, no
matter what format the address structure happens to have.  heck,
the C socket API handles that, so why shouldn't Python do that too?

...

but who cares?  as just announced, 1.6 will support connect(host, port)
as a special case -- and it won't surprise me if it's gone again in 1.7.

until then, over and out.

</F>

1) stumbled upon one today: it turned out that by changing from 1.3
to 1.5.2, large data transfers from a high-performance server to clients
on an 10 Mbit Ethernet were unexpectedly terminated.  the same trans-
fers worked just fine with high-performance workstations, on an FDDI
network.  ten guru points to anyone who can explain why.




From aturner_ at _pobox.com  Sat Apr  1 13:21:13 2000
From: aturner_ at _pobox.com (Aaron Turner)
Date: Sat, 01 Apr 2000 10:21:13 -0800
Subject: Why should I switch to Python?
References: <seas5ia89177@corp.supernews.com><00033120055300.03687@quadra.teleo.net>
Message-ID: <secfidq8921@corp.supernews.com>

In article <00033120055300.03687 at quadra.teleo.net>, Patrick Phalen <python-list at teleo.net> wrote:
> [Aaron Turner, on Fri, 31 Mar 2000]
> :: A friend of mine is trying to convince me we should 
> :: do a project using ORBit in Python.  I've been 
> :: coding Perl for over 3 years and really like it.  
> :: (Especially it's syntax and ability to grok my 
> :: programming style.)  Now, I understand the Perl
> :: vs. Python war and everyone has their preferred
> :: taste in the matter.  
> :: 
> :: What I'd like to hear from people is what technical 
> :: resons I should switch to Python.  Are there any 
> :: features that Python has (perhaps in it's OO 
> :: offering) that Perl lacks?  From what I can tell
> :: people tend to prefer Python over Perl mostly
> :: for its syntax, which I find very likable.
> 
> 
> Aaron,
> 
> Please see:
> 
> http://www.python.org/doc/Comparisons.html
> 
> http://www.perl.com/pub/language/versus/python.html
> 
> Plus a Google search on "perl vs python" should give you plenty more to
> chew on without reviving that tiresome thread here.
> 

Yeah, I found those before I posted (as well as reading the archives of
this group on deja news).   Honestly I found them for the
most part very lacking in technical information.  Mostly things like:

- Perl code is ugly, Python code is easy to read
- Python was OO from the start, Perl's OO was a bolt on

But things like this aren't particularly useful in trying to determine
which language gets the job done better.  Code readibilty has 
historically been a human problem which is solved with coding
standards set for the programmers IMHO.   So I'm more interested in 
things like:

- What is the impact of Python's OO being there from the start?
- Does it perform better?
- Is it significantly easier to use?  
- More powerful/support features that aren't in Perl yet?

Honestly the more I reasearch, the more it seems that it's purely 
a matter of taste.  If you like Perl, there's nothing inherient about
Python that makes it better.   Sorry if this just sounds like a troll, but
I'm really trying to get a feel about something I know nothing about 
from personal experiance.

Basically if you know both languages, and were planning a mid-sized 
project, why would you choose Python?  It's hard to take my friend
seriously in his evaluation when he says he'd rather do this project
in C++  than in Perl (especially when he doesn't know Perl).

Thank you for your patience.

-Aaron


From moshez at math.huji.ac.il  Fri Apr 28 01:12:09 2000
From: moshez at math.huji.ac.il (Moshe Zadka)
Date: Fri, 28 Apr 2000 08:12:09 +0300 (IDT)
Subject: Metaclasses?
In-Reply-To: <Ok2O4.726$4A6.145248@dfiatx1-snr1.gtei.net>
Message-ID: <Pine.GSO.4.10.10004280810510.8383-100000@sundial>

On Thu, 27 Apr 2000, GTE wrote:

> On the other hand I kinda like Visual Basic's 'with' construct. Consider:
> def method(self,arg1,arg2):
>     with self:
>         .spam = "spam"
>         .eggs = arg1+arg2

Here's a way to get almost that effect:

def method(self, arg1, arg2):
	_ = self
	_.spam = "spam"
	_.egss = arg1 + arg2

--
Moshe Zadka <mzadka at geocities.com>. 
http://www.oreilly.com/news/prescod_0300.html
http://www.linux.org.il -- we put the penguin in .com




From gmol at my-deja.com  Sat Apr 15 03:35:33 2000
From: gmol at my-deja.com (gmol at my-deja.com)
Date: Sat, 15 Apr 2000 07:35:33 GMT
Subject: General question of "SystemError:NULL result..."
Message-ID: <8d9645$fsa$1@nnrp1.deja.com>

Hi all, disclaimer, I am new to python so I may be doing something
severly stupid.

I have a problem, while it is with VTK, I was wondering if someone could
tell me if it is in fact a general one.

I have installed VTK and python on my sys,and some of the python
examples work, and some don't...the ones that don't say always say
something like:

  File "KSpline.py", line 106, in ?
    profileTubes.SetInput(profileData)
SystemError: NULL result without error in call_object

Now, no one has been able to figure this out on the VTK list...but I do
know that people can run the same examples on the same python under the
same platform (linux) and the same Mesa...

Has anyone ever had problems like this in another package; where some
examples do work (showing that is it not a problem about compilation
with shared suppourt or anything)? Could someone let me in on what I
could do to find out exactly what call is being made in the compiled lib
that is returning the NULL result?

Sigh, I realize this is a shot in the dark,but I am desparate and
someone knows whats going on :)


Sent via Deja.com http://www.deja.com/
Before you buy.


From JamesL at Lugoj.Com  Sat Apr  1 00:46:21 2000
From: JamesL at Lugoj.Com (James Logajan)
Date: Fri, 31 Mar 2000 21:46:21 -0800
Subject: Python 1.6 alpha 1 released
References: <200003312130.QAA04361@eric.cnri.reston.va.us> <38E5205F.DE811F61@roguewave.com> <Wf9F4.4423$74.70231@newsc.telia.net>
Message-ID: <38E58D2D.384F1546@Lugoj.Com>

Fredrik Lundh wrote:
> 
> Bjorn Pettersen <bjorn at roguewave.com> wrote:
> > Just read the announcement page, and found that socket.connect() no
> > longer takes two arguments as was previously documented.
> 
> note that the actual reference page says that connect takes
> just one argument, and it has done so since at least 1.4.
> 
> > A quick look shows that this breaks all the network scripts I have
> > installed (at least the ones that I found, undoubtedly there are many
> > more).  Because of this I will put any upgrade plans on hold.
> 
> were you really planning to *upgrade* to a rough alpha?
> 
> </F>

Sorry, but Bjorn Pettersen has a perfectly valid point. The copy of the
reference page that I have on hand at the moment (1.5b1) uses the word
"pair", not "tuple" and the example clearly shows a two argument call, not a
one argument call using a tuple. In the face of conflicting information in
the Python library reference manual (which has been significantly less than
perfect), I have tended to use examples as a guide. I should also note that
a quick check of some old code of mine indicates use of the two argument
form for connect. So you have at least two idiots who read the library
reference and still managed to do the wrong thing (and made it work; I
wonder how...). Worse, the idiot implementor of the connect interface
managed to support the same thing! I wonder who the third idiot is? I've
noticed the world is full of idiots like me and I bet there is a lot more
code out there that will break due to this.

Also, how good are you at math? You seem to feel that 5 minutes of "Guido
the god's" time is of much more value than that of 1000 knuckle-dragging
Neanderthal programmers like myself spending a mere 10 minutes each to make
sure their code doesn't break. Of course, if we happen to take more than 10
minutes its our own damn fault I'm sure you'll agree.

Now how did I get so pissed over this exchange of posts?


From olivier at phgroup.com  Tue Apr 18 03:41:25 2000
From: olivier at phgroup.com (Olivier Migeon)
Date: Tue, 18 Apr 2000 09:41:25 +0200
Subject: readline skips on line
References: <8culp8$23a$1@porthos.nl.uu.net>
Message-ID: <38FC11A5.627E1EA9@phgroup.com>

readline() actually reads a line and put the cursor at the beginning of the
next one each time you invoke it. And you invoke it in the loop.
I suggest:

f=open('test.txt','r')
while 1:
    mline=f.readline()
    if not mline : break
    print mline

> ----------------------------------Oo
>
> I got a part of a programm looking like this
>
> f=open('test.txt', 'r')
> while f.readline()!='':
>     print f.readline()
> print 'ready'
>
> but when it is runned, it wil print out the lines in the text file, but it
> skips all the the odd line, it only displays line 2,4,6,8,10 and further
>
> How can I solve this problem ?
>



 Sent via Deja.com http://www.deja.com/
 Before you buy.


From akuchlin at mems-exchange.org  Sat Apr 15 12:34:47 2000
From: akuchlin at mems-exchange.org (Andrew Kuchling)
Date: 15 Apr 2000 12:34:47 -0400
Subject: TCP/IP protocols
References: <ZSMJ4.1099$Y14.27933@news.uswest.net>
Message-ID: <3dn1mvb9w8.fsf@newcnri.cnri.reston.va.us>

"Jeff Massung" <jmassung at magpiesystems.com> writes:
> I'm new to Python programming, but am very experienced in C/C++. Could
> someone point me to a good reference for TCP/IP programming with Python
> (pdf, book, anything :]) thanks.

Try Gordon McMillan's HOWTO on socket programming:
http://www.python.org/doc/howto/sockets/

--amk


From lenny at squiggie.com  Mon Apr 10 13:35:42 2000
From: lenny at squiggie.com (Lenny Self)
Date: Mon, 10 Apr 2000 10:35:42 -0700
Subject: using httpdlib with a proxy server
Message-ID: <ChoI4.844$g54.17363@news.uswest.net>

I am attempting to write a small script that sends a list of URLs though a
proxy server and then records a bit of information on the respose.  I though
I could use the httplib module to help with this, however, I don't see where
it can send a request though a proxy server. Does anyone know of a module
that will allow me to do this?


    -- Lenny Self
        lenny at squiggie.com




From sabren at manifestation.com  Thu Apr 13 21:01:22 2000
From: sabren at manifestation.com (Michal Wallace (sabren))
Date: Thu, 13 Apr 2000 21:01:22 -0400 (EDT)
Subject: IndexError and for x in list
Message-ID: <Pine.BSF.4.10.10004132041530.4951-100000@chi.pair.com>

Hey all,

Thanks everyone who helped with the indexed dictionary thing..

What I wanted to be able to do was simulate Evilsoft's "Collection"
concept from Visual Basic. Anyway, I have it working (and will release
the code soon as part of a larger project)... But..

I wanted to be able to do this:

idic = IdxDict()

for item in idic:
    print item


which should print the value, whereas:

for key in idic.keys():
    print key

would print the key.

The seqdict class on Parnassus couldn't do this, but the standard
UserList can... I thought this magical behavior might from
__getslice__ or even __len__ or count()... But it doesn't..

for x in list:
   print x


does something like this:


@@@ i = 0
@@@ while true:
@@@     try:
@@@         x = list[i]
@@@         print x
@@@     except IndexError:
@@@         break
@@@     i = i+1


So in order for my object to allow "for x in idic" I needed to
have __getitem__ raise an IndexError, like so:


    def _toStringKey(self, key):
        """Convert numeric keys into string keys. Leaves string keys as is"""
        # handle numeric keys:
        if _isNum(key):
            if not (0 <= key < len(self.idx)):
                ## oddly enough, it is this IndexError here
                ## that allows you to do "for x in myIdxDict:"
                raise IndexError, `key` + " is out of bounds."
            # convert it to a string key
            key = self.idx[key]

        return key


    def __getitem__(self, key):
        key = self._toStringKey(key)
        return self.data[key]



Anyway, aside from informing you all of some Python Trivia :), my
purpose in writing all this is to ask: where is this documented?
The docs for "for" say:

> When the items are exhausted (which is immediately when the sequence
> is empty), the suite in the else clause, if present, is executed, and
> the loop terminates

But it doesn't mention HOW it knows when the list is exhausted.

>From the library reference:

> IndexError
>
> Raised when a sequence subscript is out of range. (Slice indices are
> silently truncated to fall in the allowed range; if an index is not a
> plain integer, TypeError is raised.)

Does "for x in y" have a hidden slice in there?

Basically, I don't want to rely on an undocumented feature that could
go away sometime. Is IndexError the right way to simulate a list?


Cheers,

- Michal
-------------------------------------------------------------------------
http://www.manifestation.com/         http://www.linkwatcher.com/metalog/
-------------------------------------------------------------------------




From lewst at yahoo.com  Wed Apr  5 00:41:54 2000
From: lewst at yahoo.com (lewst)
Date: Tue, 4 Apr 2000 21:41:54 -0700 (PDT)
Subject: Starving for an Advanced Python Book
Message-ID: <20000405044154.10369.qmail@web2103.mail.yahoo.com>

Don Tuttle <tuttledon at hotmail.com> wrote:

> I went ahead and bought it because Oreilly's cheif editor says, "We're
> timing the new edition of Programming Python to follow shortly after the
> release of Python 2.0."(aka Python 3K).  If that turns out to be the case,
> it won't published until sometime in 2001.  See his complete (and somewhat
> dated) remarks at  http://www.oreilly.com/frank/python.html

Those remarks are misleading and should probably be removed.  I won't
disclose my source, but I'll you bet dollars to donuts that the 2nd
edition of PP will be out this summer -- much closer to the release of
Python 1.6 than P3K.


__________________________________________________
Do You Yahoo!?
Talk to your friends online with Yahoo! Messenger.
http://im.yahoo.com



From Lisowski.Tomasz at sssa.nospam.pl  Fri Apr  7 02:45:29 2000
From: Lisowski.Tomasz at sssa.nospam.pl (Tomek Lisowski)
Date: Fri, 07 Apr 2000 06:45:29 GMT
Subject: Odp: Odp: Python on Alpha OpenVMS
References: <z4_C4.6214$4N.168038@news.tpnet.pl> <hDlBpKAS7T34Ew1e@gol.com><_KCD4.2797$a01.54385@news.tpnet.pl> <ZRhvtDAzBR74EwzH@gol.com>
Message-ID: <dufH4.43997$a01.966703@news.tpnet.pl>

U?ytkownik Ian Parker <parker at gol.com> w wiadomooci do grup dyskusyjnych
napisa?:ZRhvtDAzBR74EwzH at gol.com...
> In article <_KCD4.2797$a01.54385 at news.tpnet.pl>, Tomek Lisowski
> <Lisowski.Tomasz at sssa.nospam.pl> writes
> >U?ytkownik Ian Parker <parker at gol.com> w wiadomo?ci do grup dyskusyjnych
> >napisa?:hDlBpKAS7T34Ew1e at gol.com...
> >> In article <z4_C4.6214$4N.168038 at news.tpnet.pl>, Tomek Lisowski
> >> <Lisowski.Tomasz at sssa.nospam.pl> writes
> >> >Hi!
> >> >
> >> >I have tried to compile Python sources on my VAX OpenVMS system, but
soon
> >I
> >> >realized, that instead of my trusty DEC C++ compiler I need a native C
> >> >compiler. Unfortunately my C++ compiler cannot be reverted to a native
C
> >> >language mode. Has anyone succeeded to create VAX OpenVMS Python
> >binaries?
> >> >
> >> >Note: Apart from VAX OpenVMS I have also an Alpha OpenVMS server, but
> >here I
> >> >don't have any C/C++ compiler available. If anyone has created Alpha
> >OpenVMS
> >> >Python binaries, please contact me too.
> >> >
> >> >Thanks in advance
> >> >
> >> >Tomasz Lisowski
> >> >
> >>
> >> Tomasz
> >>
> >> If you aren't planning changes to the source, why not just download and
> >> link the object modules from:
> >> http://decus.decus.de/~zessin/python/obj.html
> >>
> >> I'm so inspired by this solution, I'm at last downloading them myself.
> >
> >I am just not sure, for which platform were these modules compiled, and
for
> >which version of OpenVMS system. I suppose, there could be some
differences
> >between the OpenVMS versions. At least you can see them in the source
code,
> >where there are some #ifdef and #if directives concerning the OpenVMS
> >version string.
> >
> >Please report, if you succeeded in making your Python running properly,
> >including the platform and OS info.
> >
> >Regards
> >
> >Tomasz Lisowski
> >
>
> Tomasz
>
> I'm glad to say that the VMS objects worked perfectly on both Alpha and
> VAX VMS systems.  They're both V7.x - I didn't bother to check specific
> version numbers.  I'm planning to deploy to about 20 VMS Alpha and VAX
> boxes.  I even built a backup saveset of the resultant directory tree
> and a have a DCL script that does the VAX or Alpha assembly and linking
> on the target box - this keeps it very simple.   I may even go all the
> way and make a VMSINSTALL/PCSI kit.
>
> I'm really impressed with the OpenVMS port (Uwe, please keep up the good
> work!).  Only one qualification, at the moment I think the objects don't
> support threading, but haven't really investigated yet.

Good to hear that!

I will try too those object files, as I am using the same Open VMS version.
Thanks for the information





From aa8vb at yahoo.com  Mon Apr 17 08:57:33 2000
From: aa8vb at yahoo.com (Randall Hopper)
Date: Mon, 17 Apr 2000 08:57:33 -0400
Subject: Imaging Library & Numerical Python
In-Reply-To: <38F883B6.A8ADF538@engineering.uiowa.edu>
References: <38F7F865.9C86399C@engineering.uiowa.edu>
 <PjTJ4.137181$Dv1.1488334@news1.rdc1.bc.home.com>
 <38F883B6.A8ADF538@engineering.uiowa.edu>
Message-ID: <20000417085733.A301135@vislab.epa.gov>

Juerg Tschirren:
>
> Does anybody have experience with getting the Imaging Library and
> Numerical Python working together? What I would like to do is importing
> an imaging with the Imaging Library and then using functions from both
> -- the Imaging Library and Numerical Python -- on this image.

I just recently did some list searches on this.  Past posts indicate that
if you're careful about formats, you can use the tostring() and
fromstring() methods of each library to shuffle data back and forth between
the two extensions.

You can also push the data across one pixel at a time using
getpixel()/putpixel(), but it's very slow.

-- 
Randall Hopper
aa8vb at yahoo.com



From craniac at datapro.co.za  Sun Apr 30 11:09:52 2000
From: craniac at datapro.co.za (Steve Crane)
Date: Sun, 30 Apr 2000 17:09:52 +0200
Subject: Do I always have to write "self." ?
In-Reply-To: <300420001008012011%pecora@anvil.nrl.navy.mil>; from pecora@anvil.nrl.navy.mil on Sun, Apr 30, 2000 at 02:08:00PM +0000
References: <270420001706386648%pecora@anvil.nrl.navy.mil> <slrn8ghc6t.f01.scarblac-spamtrap@flits104-37.flits.rug.nl> <8ecsgj$6lp$1@slb0.atl.mindspring.net> <m3bt2tonyg.fsf@atrus.jesus.cam.ac.uk> <300420001008012011%pecora@anvil.nrl.navy.mil>
Message-ID: <20000430170952.A502@datapro.co.za>

On Sun, Apr 30, 2000 at 02:08:00PM +0000, Louis M. Pecora wrote:

> Isn't this called something like "Hungarian" variable naming?

Hungarian Notation was introduced by Charles Simonyi (who happened to
work for Microsoft).  In it's initial form it used a prefix to indicate
the data type of the variable but not the scope.  Later people started
to add a scope prefix but I don't think that Microsoft uses this as a
standard.

While it may seem to make code harder to read it is in fact very useful
when working on large projects involving may programmers.  However if
you're writing a program yourself and don't think that anyone else will
ever have to read the code you can probably steer clear of Hungarian
Notation in either form.

For an explanation of the notation by Charles Simonyi himself, see
http://msdn.microsoft.com/library/techart/hunganotat.htm.
-- 
Steve Crane
http://craniac.daveworld.org



From m.faassen at vet.uu.nl  Wed Apr 26 06:18:54 2000
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 26 Apr 2000 10:18:54 GMT
Subject: Simple list.append() question
References: <meh9-5ED8CD.11192424042000@news.cit.cornell.edu> <ZHZM4.360$qg7.2949@news.uswest.net> <39046FA2.603EAFE5@roguewave.com> <PsaN4.174166$Dv1.2059732@news1.rdc1.bc.home.com> <8e4akc$p7o$1@newshost.accu.uu.nl> <dcuN4.177124$Dv1.2114909@news1.rdc1.bc.home.com>
Message-ID: <8e6fqe$f0q$1@newshost.accu.uu.nl>

[I spoil Kevin's fun]
Kevin Cazabon <kcazabon at home.com> wrote:
> Aw, comon, I was having fun... q:]

Oops, sorry to spoil your fun. :)

Regards,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From g_will at cyberus.ca  Mon Apr 10 17:24:56 2000
From: g_will at cyberus.ca (Gordon Williams)
Date: Mon, 10 Apr 2000 21:24:56 GMT
Subject: Installer Producing Page Fault
References: <1256735085-36518465@hypernet.com>
Message-ID: <01bfa332$df04f6c0$5240c3d1@Cyberus>


Gordon McMillan <gmcm at hypernet.com> wrote in article
<1256735085-36518465 at hypernet.com>...
> Gordon Williams wrote:
> 
> > I am using Gordon's installer program to make a standalone program that
> > self installs.  After running the install_BS_main.exe created,
everything
> > is put into the proper directory (I think).  Then running BS_main.exe I
> > immediately get a crash with the following:
> 
> Never tried it myself, but I got this note once from Gary Herron:
> 
> > ** Numeric Python.  The 'import Numeric' statment imports _numpy.pyd,
> > but this is not caught by builder.py.  Add
> >   misc= C:\Program Files\Python\Numeric\_numpy.pyd
> > in the appropriate target of the cfg file.
> 

Yes that did the trick and it is now working.  The wxPython stuff was all
found (a bit to my supprise) but the serial IO module had a problem and
required a few extra lines in the .cfg file.

Thanks

Gordon


From gmcm at hypernet.com  Sat Apr 22 09:10:07 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Sat, 22 Apr 2000 09:10:07 -0400
Subject: The Simple Economics of Open Source
In-Reply-To: <22042000.3@sanctum.jae.ddns.org>
References: <39011EEF.4BAEC3C9@tcac.net> (message from Albert Wagner on Fri,	21 Apr 2000 22:39:27 -0500)
Message-ID: <1255713487-40391203@hypernet.com>

Juergen A. Erhard wrote:

>     Albert> Those with cooperative strategies have discovered that
>     Albert> there are an infinite number of pies.
> 
> Wrong.[1] They simply discovered (or, rather, understood) that there
> are enough pies so that *everyone* can be more than satiated ("Jeder
> kann sich bis zum Erbrechen den Bauch vollschlagen" I'd say in
> german).

> [1] The human-accessible part of this solar system is finite, right?
> So there simple cannot be an infinite number of pies, even if you
> could bake a pie out of one atom... ;-)

But you can grow more wheat and grow more apples and raise 
more cows and make more pies tomorrow.

"Infinite", in an economic context, simply means "more than 
what Bill Gates has" <wink>.

- Gordon



From python-list at teleo.net  Thu Apr 13 23:39:28 2000
From: python-list at teleo.net (Patrick Phalen)
Date: Thu, 13 Apr 2000 20:39:28 -0700
Subject: *total* newbie help
In-Reply-To: <sfd5gp9ealn159@corp.supernews.com>
References: <sfd5gp9ealn159@corp.supernews.com>
Message-ID: <00041321053602.02223@quadra.teleo.net>

[Collin Greene, on Thu, 13 Apr 2000]
:: I am new to to programming but it would seem to em that you could white some
:: code and then run it in a different action, does python ingrate these into
:: one action? I am just a little confused

Collin,

I'm a little confused too. When you say "write some code and then run it
in a different action," I think of several possible meanings. Maybe
you're talking about creating a function and then invoking it. Have you
read in the tutorial about functions?

How about giving us a more specific example of an action you have in
mind?

Or, since you're new to programming, I bet you'll find some useful
information here:
http://members.xoom.com/alan_gauld/tutor/tutindex.htm



From smalleys at gte.net  Sat Apr 15 13:48:21 2000
From: smalleys at gte.net (sue)
Date: Sat, 15 Apr 2000 17:48:21 GMT
Subject: Python and VTK and TK
References: <38F51AB1.29FB4C1D@earthlink.net> <38F528C8.2C18364F@gte.net> <38F7815C.57EC8A83@boi.hp.com>
Message-ID: <38F89D69.1CC83E26@gte.net>

Yes, I ended up reading header files for some things.  But
the speed and size tradeoff is very nice.

Sue.


From timo at iol32.com  Mon Apr  3 18:55:31 2000
From: timo at iol32.com (Tim Ottinger)
Date: Mon, 03 Apr 2000 22:55:31 GMT
Subject: Images (.bmp,.jpg,gif) in Tkinter windows (Tk newbie)
References: <38df6cfe.2185312@news.integrityonline.com> <8bo8gc$gic$1@nnrp1.deja.com>
Message-ID: <38e8dd4f.15889708@news.integrityonline.com>

On Mon, 27 Mar 2000 18:16:17 GMT, effbot at pythonware.com wrote:

>Tim Ottinger wrote:
>> I'm in Win32. I need an example (that works) of loading a graphical
>> image using Tkinter and python. Looking at the image and bitmap
>> options hasn't helped me, because I'm reading win32 .bmp files, not
>> xbm files. I'd like to read and display gifs, jpgs, and the like as
>> part of my Tkinter dialogs.
>
>native Tkinter supports GIF and PPM files (use the PhotoImage
>class to load them).
>
>   image = PhotoImage(file="myfile.gif")
>   widget.config(image=image)
>   widget.image = image # make sure to hang on to the image

Well, that would be okay, except that it doesn't actually display the
image on the screen (for me). Neither in '98 (boo) or 'NT (boo).

Tim


From mlh at idi.ntnu.no  Thu Apr  6 12:39:33 2000
From: mlh at idi.ntnu.no (Magnus Lie Hetland)
Date: Thu, 6 Apr 2000 18:39:33 +0200
Subject: NumPy modelling of Belief Tables?
Message-ID: <8ciem1$scb$1@kopp.stud.ntnu.no>

Hi!

I'm dabbling in Bayesian Networks at the moment, and have considered
implementing a system/framework in Python, using NymPy for the
Belief Table calculations... The calculations are quite simple, but I'm not
sure how I should do them all in Numeric Python...

In short, a Belief Table represents a discrete joint probability function
over
a set of variables, where each dimension of the table represents a variable,
and the indices of the dimension represent different values it may take.

When manipulating information in Bayesian Networks, you need to do some
operations on these tables. For instance, you may want to multiply them.
If the tables have the same dimensions (variables) then the standard NumPy
multiplication (i.e. *not* matrix multiplication) applies. Now that's
fine...
But if they *don't* have the same dimensions, then this won't work
directly...

If, for instance, the table t1 has the dimensions X and Y, while t2 has the
dimensions X and Z, then for each x, y, z in X, Y, X,

        (t1 * t2)[x,y,z] = t1[x,y] * t2[x,z].

This may be simple to do in NumPy (it is certainly easy to do
algorithmically) - I just can't see how...

Division is about the same thing, except that 0/0 = 0... For tables with the
same dimensions, I guess one could use something like

        t1 / where(equal(t2,0),1,t2)

where 1 is just an arbitrary value. (I'm not sure how this solution
generalizes when the dimensions differ...)

The other operation needed is marginalization of the tables. When a table
is marginalized over some of its dimensions, the elements having equal
coordinates in the remaining dimensions (i.e. the ones not being
marginalized
over) are added. For instance, if you have a two-dimensional table, t, over
the variables X and Y, and you want to marginalize out Y, then for each
x in X,

        m[x] = sum with x in Y { t[x,y] },

where m is the marginalized table.

If anyone finds these problems interesting, I would appreciate any hints
or solutions.

--

  Magnus
  Lie
  Hetland                            magnus @ hetland . org





From josht at iname.com  Fri Apr  7 21:03:40 2000
From: josht at iname.com (Josh Tompkins)
Date: Sat, 08 Apr 2000 01:03:40 GMT
Subject: Searching for Python Material...
References: <8F0ED18C1joshtinamecom@209.208.0.7> <oFkH4.203$0p1.45386@typhoon.austin.rr.com>
Message-ID: <8F0FD068Fjoshtinamecom@209.208.0.7>

Thanks for the reply...

I've got Learning Python on order at Barnes & Noble, and I'm finding some 
of the stuff at Parnassus informative.

Thanks again.

Josue
 
________________________________________________________________

"Destined For Great Things -- but pacing myself."
- From a t-shirt.

E-Mail:  josht at iname.com 
ICQ:  21219667
AIM:  JosueTheGreat
Web:  http://www.crosswinds.net/~josht
_________________________________________________________________


From mhammond at skippinet.com.au  Wed Apr 12 06:15:36 2000
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Wed, 12 Apr 2000 10:15:36 GMT
Subject: Seg fault in Python/C API: PyString_FromString(NULL)
References: <38F422C5293.A67BYEMARTIN@smtp.garage.co.jp>
Message-ID: <c1YI4.2029$UB.13736@news-server.bigpond.net.au>

"Yves-Eric Martin" <yemartin at garage.co.jp> wrote in message
news:38F422C5293.A67BYEMARTIN at smtp.garage.co.jp...

>     First, I am new to the Python/C API, so correct me if I am wrong:
> shouldn't PyString_FromString(NULL) return Py_None? Or am I out to
> lunch?

Nope - you must check for NULL before making the call.

Mark.





From noreply at antispam.com  Sat Apr  1 17:06:15 2000
From: noreply at antispam.com (Brian Schmidt)
Date: Sat, 01 Apr 2000 22:06:15 GMT
Subject: Why should I switch to Python?
References: <seas5ia89177@corp.supernews.com>
Message-ID: <rpuF4.1284$HB2.127422@paloalto-snr1.gtei.net>

Hi Aaron,

I don't know if anyone has pointed you to this yet, but here's a very nice
paper entitled "An empirical comparison of C, C++, Java, Perl, Python, Rexx
and Tcl ..." that I found to be very informative:
http://wwwipd.ira.uka.de/~prechelt/Biblio/jccpprtTR.pdf    (PDF format)

HTH,
Brian


"Aaron Turner" <aturner_ at _pobox.com> wrote in message
news:seas5ia89177 at corp.supernews.com...
>
> A friend of mine is trying to convince me we should
> do a project using ORBit in Python.  I've been
> coding Perl for over 3 years and really like it.
> (Especially it's syntax and ability to grok my
> programming style.)  Now, I understand the Perl
> vs. Python war and everyone has their preferred
> taste in the matter.
>
> What I'd like to hear from people is what technical
> resons I should switch to Python.  Are there any
> features that Python has (perhaps in it's OO
> offering) that Perl lacks?  From what I can tell
> people tend to prefer Python over Perl mostly
> for its syntax, which I find very likable.
>
> Thanks.




From tom-main at REMOVEME.home.com  Thu Apr 13 18:26:11 2000
From: tom-main at REMOVEME.home.com (Tom)
Date: Thu, 13 Apr 2000 22:26:11 GMT
Subject: How do I undo an import??
Message-ID: <7QrJ4.185421$Hq3.4353581@news2.rdc1.on.home.com>

I'm a C++ programmer, new to Python, working with v1.5.2 on Windows.

I run the command-line interpreter, then I type:

>>> from NetCfg import *

to import my extension.  Now I want to unload my NetCfg DLL without exiting
the interpreter.  How do I do this?

Also, I type the above import command every time I start the interpreter.
Is there some way to get have this command executed automatically?

Thanks,

Tom.





From aahz at netcom.com  Sun Apr  2 13:18:24 2000
From: aahz at netcom.com (Aahz Maruch)
Date: 2 Apr 2000 17:18:24 GMT
Subject: Why should I switch to Python?
References: <seas5ia89177@corp.supernews.com><00033120055300.03687@quadra.teleo.net> <secfidq8921@corp.supernews.com>
Message-ID: <8c7vd0$67o$1@slb0.atl.mindspring.net>

In article <secfidq8921 at corp.supernews.com>,
Aaron Turner  <aturner_ at _pobox.com> wrote:
>
>Honestly the more I reasearch, the more it seems that it's purely 
>a matter of taste.  If you like Perl, there's nothing inherient about
>Python that makes it better.   Sorry if this just sounds like a troll, but
>I'm really trying to get a feel about something I know nothing about 
>from personal experiance.
>
>Basically if you know both languages, and were planning a mid-sized 
>project, why would you choose Python?  It's hard to take my friend
>seriously in his evaluation when he says he'd rather do this project
>in C++  than in Perl (especially when he doesn't know Perl).

As a person who has now used Perl and Python for roughly equal amounts
of time (more than a year), I think the most important distinction is
the way that Perl's TMTOWTDI destroys multi-person projects.  Not just
at the pretty-printing level, but all the way up to module layout and
general code design.

Perl is the result of attempting to make code "readable" by combining
the syntax and semantics of popular Unix utilities.  Python is the
result of attempting to make code readable by designing a language based
on real programming usability studies.  If you're already a Unix shell
hacker, Perl is likely to make you more immediately productive, but it
simply doesn't scale.  In all other cases, I think Python wins.
--
                      --- Aahz (Copyright 2000 by aahz at netcom.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

"It's 106 miles to Chicago.  We have a full tank of gas, a half-pack of
cigarettes, it's dark, and we're wearing sunglasses."  "Hit it."


From dwhite2 at seas.upenn.edu  Fri Apr 14 19:26:35 2000
From: dwhite2 at seas.upenn.edu (dave white)
Date: Fri, 14 Apr 2000 19:26:35 -0400
Subject: Debugging with IDLE
Message-ID: <38F7A90A.5019937D@seas.upenn.edu>

Hello,

I'd like to debug a python script using IDLE.  The way my script is
written, it requires command like arguments when run -- e.g.  I run my
script on the command line by typing :

>python script.py input.txt

and 'input.txt' goes into sys.argv[1].  

Is there a way to provide a command line value to IDLE so that when I
run script.py, it will be evaluated with 'input.txt' in sys.argv[1]?

thanks,
david


From vetler at ifi.uio.no  Mon Apr  3 04:02:31 2000
From: vetler at ifi.uio.no (Vetle Roeim)
Date: 03 Apr 2000 10:02:31 +0200
Subject: Emacs Python mode
References: <yayk8iii0ex.fsf@ganglot.ifi.uio.no> <m3zorendyk.fsf@atrus.jesus.cam.ac.uk> <yayzorcj1ox.fsf@ganglot.ifi.uio.no> <8c794n$mjq$1@zingo.tninet.se>
Message-ID: <yayu2hja9yg.fsf@ganglot.ifi.uio.no>

* Andr? Dahlqvist
> > all in all, *emacs* has lots of shortcuts for handy things. I wish I
> > knew more about it.. :)
> 
> While we're on the subject of Emacs Python mode, why is it that I can't 
> indent a line under a line which has been commanted out? Emacs 
> refuses to let me tab in there, but if I remove the comment it will happily 
> let me do it.

try writing:

def foobar(zot):
    print "and now for something completely different"
## this is a comment
    print "ni!"



vr


From hinsen at cnrs-orleans.fr  Wed Apr 19 05:40:50 2000
From: hinsen at cnrs-orleans.fr (Konrad Hinsen)
Date: 19 Apr 2000 11:40:50 +0200
Subject: Standard Makefile for embedding Python?
Message-ID: <m3itxea0nx.fsf@chinon.cnrs-orleans.fr>

Is there anything similar to the standard Makefile for modules
(Misc/Makefile.pre.in) for making executables that embed Python?
Ideally with the option of adding built-in modules...

Konrad.
-- 
-------------------------------------------------------------------------------
Konrad Hinsen                            | E-Mail: hinsen at cnrs-orleans.fr
Centre de Biophysique Moleculaire (CNRS) | Tel.: +33-2.38.25.55.69
Rue Charles Sadron                       | Fax:  +33-2.38.63.15.17
45071 Orleans Cedex 2                    | Deutsch/Esperanto/English/
France                                   | Nederlands/Francais
-------------------------------------------------------------------------------


From tim.hochberg at ieee.org  Tue Apr 25 10:01:59 2000
From: tim.hochberg at ieee.org (Tim Hochberg)
Date: Tue, 25 Apr 2000 14:01:59 GMT
Subject: Simple list.append() question
References: <Pine.GSO.4.10.10004250813240.20737-100000@sundial> <0CF260C495FED111A6610000F866308D0D893596@mail3.ntu.edu.sg>
Message-ID: <rzhN4.75410$U4.570041@news1.rdc1.az.home.com>

For those who _need_ a one-liner, the following works:

map(list, [()]*3)

as does

map(list, [[]]*3)

and

map(copy.copy, [[]]*3)


-tim





From yawyi at gmx.de  Tue Apr  4 19:45:38 2000
From: yawyi at gmx.de (Johannes Stezenbach)
Date: Wed, 5 Apr 2000 01:45:38 +0200
Subject: ldap support?
References: <Pine.LNX.4.10.10004041604050.10324-100000@worf.mcnc.org>
Message-ID: <38ea8c40$0$28633@personalnews.de.uu.net>

oliver <fwang2 at yahoo.com> wrote:
>I wonder is there any standard LDAP support module in python? I looked up
>SIGs, doesn't show anything there. I appreciate your help.

SuSE Linux (6.3) comes with a python ldap module written by
David Leonard and downloadable from
http://www.csee.uq.edu.au/~leonard/dc-prj/ldapmodule/
It works with OpenLDAP (www.openldap.org).
See http://www.web2ldap.de/ for example usage.



From mhammond at skippinet.com.au  Sat Apr  1 06:49:47 2000
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Sat, 01 Apr 2000 11:49:47 GMT
Subject: arraymodule questions
References: <8c4iqv$2na$1@nntp9.atl.mindspring.net>
Message-ID: <vnlF4.1712$1b4.7736@news-server.bigpond.net.au>

> So my questions are about changes for Python 1.6:
>   1) would it be okay to make the arrayobject C API available?
>   2) could I add the .append optimizations mentioned above?
>   3) should I also add the code to preallocated space?

How about if we better expose a "buffer" object.  There is a
Python C API for a new read-write buffer object, but it is not
exposed to Python code.

It is basically a fixed array("c"), and supports all slicing and
assignment operations.  The difference from an array object is
that its size is fixed, but that can sometimes be a bonus.

I wouldnt mind seeing the builtin buffer() function check for an
integer argument, and have it return a read-write buffer object
of that size.

eg:
buffer(string_object) now returns a read-only buffer into the
string object.  As buffer(10) doesnt make sense in that context,
buffer(10) could return a read-write buffer of size 10.

>From you description, this would be an alternative solution?

Mark.




From arthur.haas at westgeo.com  Thu Apr 27 18:35:46 2000
From: arthur.haas at westgeo.com (Art Haas)
Date: 27 Apr 2000 17:35:46 -0500
Subject: Newbie Question
References: <sghdv57tqtj79@corp.supernews.com>
Message-ID: <lr1z3rtbnh.fsf@yoda.wg.waii.com>

"Arnaldo Riquelme" <javanet at dynacap.com> writes:

> Suppose I have a file name tick.tkr that looks like this
> 
> SUNW.
> MFST.
> FOO.
> SPAM.
> MOO.
> End_Of_File
> 
> I want to read the file into a list
> 
> myfile = open("tick.tkr", "r")
> mylist = myfile.readlines()
> 
> Now I have mylist with the contents of tick.tkr. "mylist" looks like this
> ['SUNW.\012', 'MSFT.\012', ..........'End_of_File/012']
> 
> How do I read the file into a list without getting the newline character, so
> I can accomplish this
> 


> for tic in mylist:
>     #create file tic.prn
> 
> I know the other P language has something like 'chop' or something like
> that, but I'm not going there.

In the other `P' language you'd probably use chomp() ...

my $file = 'tick.tkr';
open(TICK,"<$file") || die "Can't open '$file'! $!\n"
chomp(my @lines = <TICK>);
close(TICK);

In python ...

>>> import string
>>> myfile = open("tick.tkr", "r")
>>> mylist = map(string.strip, myfiles.readlines())
>>> myfile.close()
>>> mylist
['SUNW.', 'MFST.', 'FOO.', 'SPAM.', 'MOO.']
>>>

-- 
###############################
# Art Haas
# (713) 689-2417
###############################


From mpetroni at my-deja.com  Fri Apr  7 04:52:07 2000
From: mpetroni at my-deja.com (mpetroni at my-deja.com)
Date: Fri, 07 Apr 2000 08:52:07 GMT
Subject: Determining Classname of an object
References: <8ccbk4$b6p$1@nnrp1.deja.com> <38E9F7C0.54899FE4@camelot-it.com>
Message-ID: <8ck7ji$20s$1@nnrp1.deja.com>

In article <38E9F7C0.54899FE4 at camelot-it.com>,
  Michael Tiomkin <michael at camelot-it.com> wrote:
> a.__class__.__name__
>
> mpetroni at my-deja.com wrote:
> >
> > Is there a way to find out from which class an object has been
created?
>

yope, thanx, got it!


Sent via Deja.com http://www.deja.com/
Before you buy.


From quinn at drachma.ugcs.caltech.edu  Sun Apr 16 13:13:47 2000
From: quinn at drachma.ugcs.caltech.edu (Quinn Dunkan)
Date: 16 Apr 2000 17:13:47 GMT
Subject: [ANNOUNCE] Garbage collection for Python
References: <20000407221750.A29279@acs.ucalgary.ca>  <m3r9cagx3c.fsf@atrus.jesus.cam.ac.uk> <l03102809b51d41895a80@[193.78.237.135]> <38F9F01B.57273052@tismer.com>
Message-ID: <slrn8fjtao.2d4.quinn@drachma.ugcs.caltech.edu>

>> - Even so, Chris and I agree that continuations should eventually
>>   be completely hidden from the user, since it's so incredibly
>>   obscure, hardly anyone will understand it. Not good for CP4E ;-)
>
>Today I'm saying so. But I think there will come a time
>where people are told "look, a continuation is as simple as 1+1",
>and they'll say "sure? I thought a co is simpler."

I'm not sure I understand correctly, but does "completely hidden from the
user" mean "only visible (in the "raw" form) from C"?  In that case, I think
it should still be visible from python.

I don't see how it's different from metaclasses or the "new" module:  just
put big red stickers on it.  It'd be nice to have around for people who want
to experiment, write obfuscated code, or maybe even write a useful
continuation-using extension.  I don't think there's much trouble with people
abusing the Beaudry hook to explode anyone's head other than their own.


From sabren at manifestation.com  Thu Apr 27 17:49:21 2000
From: sabren at manifestation.com (Michal Wallace (sabren))
Date: Thu, 27 Apr 2000 17:49:21 -0400 (EDT)
Subject: web site test framework? in python?
In-Reply-To: <8e9uor$eaa$1@nnrp1.deja.com>
Message-ID: <Pine.BSF.4.10.10004271745230.28211-100000@chi.pair.com>

On Thu, 27 Apr 2000 tiddlerdeja at my-deja.com wrote:

> I'd also like to code the testing framework in Python.
> 
> Again, has anyone any experience in this testing websites arena?

Well, there's http://pyunit.sourceforge.net/ .. I't s a pretty good
test suite manager.


> I was thinking of having JSP or ASP pages check for a testing variable
> (registry or environment). If pages were called in test mode they would
> return XML to be parsed which would can then be used as input data to
> call the page again (this way you could store the test data with the
> page you are testing). The XML would also contain the desired correct
> data to check against to ensure the page is working correctly.

Perhaps you should stick all this code in an object so you don't have
to complicate your tests by communicating via http. You can set it
up so you just pass objects around..

Cheers,

- Michal
-------------------------------------------------------------------------
http://www.manifestation.com/         http://www.linkwatcher.com/metalog/
-------------------------------------------------------------------------




From matsaleh at my-deja.com  Sat Apr 22 16:52:55 2000
From: matsaleh at my-deja.com (matsaleh at my-deja.com)
Date: Sat, 22 Apr 2000 20:52:55 GMT
Subject: pickle.py and cPickle.c - persistent_id() is always called - why?
Message-ID: <8dt3f2$r0v$1@nnrp1.deja.com>

Hello fellow .py types...

I am doing a bit of work using pickle/cpickle and am trying to
optimize. I found that my user-defined persistent_id() function is
being called for (just about) every attribute (name and value) in my
object that is being pickled.

My specific (somewhat ad-hoc) test case pickles ~100 objects in a
containment hierarchy, using persistent_id to break the containment
relationship and replace the references with a proprietary object id.
Although I have only ~100 object references to resolve, persistent_id()
is called ~6400 times. In tracing the code, it appears to be called for
every attribute and value in my objects.

This appears to be caused by the fact that pickle.py:save() is called
with the default pers_save flag of 0 in all cases except for when it is
called by save_pers(). This appears to indicate that all types: tuples,
dicts, sequences, longs, strings, etc, cause my persistent_id() to be
called, when all I want is for it to be called for object instances.

I modified pickle.py to change the default of the pers_save flag to 1
in the save() method, and then call it with a 0 only from within the
save_inst() method. This amounts to invoking my persistent_id()
function only when a reference to an object instance is encountered as
my objects are being pickled.

My pickled objects do not appear to be adversely  affected by this
change, and the number of calls to persistent_id() was reduced from
~6400 to ~500, reducing the time spent in this method by an order of
magnitude (0.37 sec to 0.03 sec).

I have not yet tested this change in cPickle.c, but in my original
tests, persistent_id() was a much more significant factor in my runs
using cPickle, because all the pickling code is in C and is relatively
much faster than my persistent_id() method, which is in Python. I
expect the performance boost my making the change in cPickle to be even
greater, relatively speaking.

My question is, why is persistent_id() being invoked so often? I do not
see the reason for calling it for basic python types and structures
such as tuples, dicts, and the like. Is this a reasonable change to
make to the python source, or is it not a safe change for the general
pickling cases that the pickle/cPickle modules have to handle? I'm sure
this code has been scruitinized by many folks much more experienced
than I - I would be grateful for your insights.

Regards.



Sent via Deja.com http://www.deja.com/
Before you buy.


From mjhand at concentric.net  Tue Apr 18 10:42:02 2000
From: mjhand at concentric.net (Manus Hand)
Date: Tue, 18 Apr 2000 08:42:02 -0600
Subject: Upgrading IIS to 1.6 problem....
References: <2B1262E83448D211AE4B00A0C9D61B03BA72C9@MSGEURO1>
Message-ID: <38FC743A.86008959@concentric.net>

Thanks, Pieter!

That was indeed the problem.  My IIS Mgr settings still had the 1.5.2
installation listed.  (Makes me wonder how important the registry
setting is in NT.)

Manus




From root at sirius.ftsm.ukm.my  Mon Apr 10 05:41:03 2000
From: root at sirius.ftsm.ukm.my (root)
Date: Mon, 10 Apr 2000 17:41:03 +0800
Subject: statistic on python (Rpy)
In-Reply-To: <20000410084948.7D7D01CDE1@dinsdale.python.org>; from python-list-request@python.org on Mon, Apr 10, 2000 at 04:49:48AM -0400
References: <20000410084948.7D7D01CDE1@dinsdale.python.org>
Message-ID: <20000410174103.A3558@sirius.ftsm.ukm.my>

hai,

/* I am not an expert on C or python, but know enough to make my way around C
 * programs. This is very preliminary, and is primary to generate interest.
 */
 
currently there is not much support for doing statistical computing on
python (distributions, random numbers ). as an experiment, I took R src
code, (http://www.r-project.org),(R is  a computer language for doing
statistic , based heavily on S and Scheme), hack a bit and use SWIG to
implement the basic statistical distributions as python modules.

after a few days, I finally able to build a python module consist of basic
distributions. (the distributions computing is wrote in C)

my opinion is that, I would nice to port R statistical engine into python
module, and the user could use python as the front-end for statistical
programming. anybody willing to try this? this would greatly complement
Numpy, and help establish python as a computing platform.

R also have graphical interface.

Some interesting notes;

1. the basic distributions computation is in C, this can be imported into
      Python as module. however, the C src is a bit difficult to follow. I
      have to resort to try-and-error to finally get the module done.

2. the graphic is nice, not sure how port this to python.

3. some function is based on lapack. use in linear aljebra.

4. there quite a lot of R specific code that you have to consider in making
     the module. this is hard an time consuming.

5. most statistical function is writing in S, it is probably possible to do
     R2Py. Some of them are straight forward, and can be reimplemented using
     Numpy.

zamri



From thomas at xs4all.net  Sun Apr 30 06:15:44 2000
From: thomas at xs4all.net (Thomas Wouters)
Date: Sun, 30 Apr 2000 12:15:44 +0200
Subject: How do I make list global please?
In-Reply-To: <t=0LOZtIsnos1t9tIo=000hy2MF1@4ax.com>; from cmfinlay@SPAMmagnet.com.au on Sun, Apr 30, 2000 at 07:42:59PM +1000
References: <t=0LOZtIsnos1t9tIo=000hy2MF1@4ax.com>
Message-ID: <20000430121544.M11481@xs4all.nl>

On Sun, Apr 30, 2000 at 07:42:59PM +1000, cmfinlay at SPAMmagnet.com.au wrote:
> How do I make list global please?
> 
> I have a list which comprise of mags[0] and mags[1]
> I want to pass and return it from a subroutine.
> my code is (in two text files)
> 
> 
> from inittest import InitTest
> mags = [333]
> mags.remove(333)
> mags.append(0) # (Big) array of magnets
> mags.append(1) # (Big) array of magnets
> print mags[0]
> print mags[1]
> InitTest()
> print mags[0]
> print mags[1]
> 
> # file :inittest.py
> def InitTest():
>    global mags
>    print "I am in the Sub"
>    Print mags[0]
>    Print mags[1]
>    print "Leaving Sub"
>    return
> 
> I get the error message
> QUOTE
> Traceback (innermost last):
>   File "inittest1.py", line 1, in ?
>     from inittest import InitTest
>   File "inittest.py", line 5
>     Print mags[0]
> SyntaxError: invalid syntax

You actually have 2 problems: One is typing 'Print' instead of 'print';
Python is case-sensitive, which causes the SyntaxError message you see. The
other problem is your real problem, you dont define 'mags' anywhere in your
inittest.py file. The 'global' statement doesn't do what you expect it to do
(it makes the variables global even if you assign to them. Normally
variables you assign to are always local, whereas variables you only read
from are read globally only if they dont exist locally.)

Your problem is not that 'mags' is global, not local, but that 'mags' is
undefined completely. The 'inittest.py' file doesn't know what it should be,
where it should come from, etc. You can do two things: in inittest.py,
import the first file (which creates a circular dependancy, never a good
thing) or (much better) pass InitTest the list to work on, like so:

from inittest import InitTest
mags = [333]
mags.remove(333)
mags.append(0) # (Big) array of magnets
mags.append(1) # (Big) array of magnets
print mags[0]
print mags[1]
InitTest(mags)
print mags[0]
print mags[1]

# file :inittest.py
def InitTest(mags):
    print "I am in the Sub"
    Print mags[0]
    Print mags[1]
    print "Leaving Sub"
    return

You can change the name of 'mags' in the 2nd file, of course.

> my code works when it is all in the main routine.

Yes, it would (barring the Print/print thing) -- but you wouldn't need the
'global' statement, as you dont assign to mags, you only modify mags.

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From tom-main at REMOVEME.home.com  Thu Apr 13 22:23:19 2000
From: tom-main at REMOVEME.home.com (Tom)
Date: Fri, 14 Apr 2000 02:23:19 GMT
Subject: How do I undo an import??
References: <7QrJ4.185421$Hq3.4353581@news2.rdc1.on.home.com> <m3d7nthbbj.fsf@atrus.jesus.cam.ac.uk> <6dvJ4.186934$Hq3.4382652@news2.rdc1.on.home.com>
Message-ID: <rivJ4.186967$Hq3.4379319@news2.rdc1.on.home.com>

I had given my startup file the same name as my extension ... so it was
finding itself.

Tom.

"Tom" <tom-main at REMOVEME.home.com> wrote in message
news:6dvJ4.186934$Hq3.4382652 at news2.rdc1.on.home.com...
> I tried the environment variable as you suggested, pointing to a file
> containing the following:
>
> print "about to ... from NetCfg import *"
> from NetCfg import *
>
> This partially works.  Now, when I run the python shell I see the printed
> text twice, but the functions in NetCfg aren't added to the namespace.
>
> Tom.
>
> "Michael Hudson" <mwh21 at cam.ac.uk> wrote in message
> news:m3d7nthbbj.fsf at atrus.jesus.cam.ac.uk...
> > "Tom" <tom-main at REMOVEME.home.com> writes:
> >
> > > I'm a C++ programmer, new to Python, working with v1.5.2 on Windows.
> > >
> > > I run the command-line interpreter, then I type:
> > >
> > > >>> from NetCfg import *
> > >
> > > to import my extension.  Now I want to unload my NetCfg DLL without
> exiting
> > > the interpreter.  How do I do this?
> >
> > You don't, in general.
> >
> > del sys.modules['NetCfg']
> >
> > goes some of the way, but it's unlikely it goes far enough to let the
> > dll be unloaded from memory.
> >
> > > Also, I type the above import command every time I start the
> interpreter.
> > > Is there some way to get have this command executed automatically?
> >
> > Set the environment variable "PYTHONSTARTUP" to point to a file
> > containing commands you want executed.
> >
> > At least, that's what I do on Linux.
> >
> > Cheers,
> > M.
> >
> > --
> >   ... but I guess there are some things that are so gross you just have
> >   to forget,  or it'll destroy something within you.  perl is the first
> >   such thing I have known.                 -- Erik Naggum,
comp.lang.lisp
>
>




From m.faassen at vet.uu.nl  Mon Apr 10 18:30:29 2000
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 10 Apr 2000 22:30:29 GMT
Subject: Python paradigms
References: <8cmr70$b0e$1@pegasus.csx.cam.ac.uk> <3H4I4.709$4w1.17557@typhoon.nyroc.rr.com> <8cs3ho$l06$1@pegasus.csx.cam.ac.uk> <slrn8f4be7.1t0.bwinton@tor.dhs.org> <8ctc9m$sp2$1@pegasus.csx.cam.ac.uk>
Message-ID: <8ctkm5$bko$2@newshost.accu.uu.nl>

Nick Maclaren <nmm1 at cus.cam.ac.uk> wrote:
[if your function gets longer than a screenfull, it's time to start
thinking about splitting it up]

> It really is quite amusing to see these same old myths coming back
> every few decades :-)

You think splitting a function up is a myth? I find it helps 
me and my code. It becomes more readable and writeable. I'm genuinely
surprised by your assertion this is an 'old myth'.

> That particular dogma was very popular back in the late 1960s and
> early 1970s, when practical programmers pointed out that it was a
> virtually impossibility for a normal person to keep track of 50
> trivial functions in their head, but it was quite easy to look
> at 60 lines of sequential code.  That didn't get listened to
> then, either!

Of course if you go about randomly chopping up your function into
multiple ones, that's not going to improve matters. The idea is
to conceptually split your function into multiple smaller ones
that make more sense, are more easy to adapt, are coherent by 
themselves, and make the original big function more readable, as
by writing all the smaller functions you've hopefully also given
them names. It's called refactoring, of course. :)

My guidelines are:

If your function is long and complicated, too many concepts
and abstractions are mingling in the same place. Do a better job
of abstracting stuff, and you'll end up with shorter, clearer
functions.

If your function is long and repetitive, you're not being a lazy
enough programmer. Write some smaller functions that generate the
stuff, read it from a file or database, whatever.

If your function is long and simple..do these happen? They don't seem
to happen in my code. Then again, my "this is complicated" reflex is
triggered pretty quickly; I'm just not too smart. :)

Are you saying these guidelines are based on a myth?

[snip]
> My current infliction is a accounting file format, and the the
> function to check the input (you DO check your input, don't you?)
> needs to perform 100 tests in series.  Yes, I am inventing
> trivial functions, but it would be clearer to put more of the
> tests inline.

Why not split this up into a bunch of validation objects or 
something, and then check the input that way? And aren't at least
some of these tests similar enough to abstract into a function
(such as tests for maximum various lengths of strings). Of course
the accounting file format may be so baroque your function just
has to be baroque, but that sort of thing *ought* to be rare. :)

Regards,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From Gareth.McCaughan at pobox.com  Wed Apr  5 14:59:47 2000
From: Gareth.McCaughan at pobox.com (Gareth McCaughan)
Date: 05 Apr 2000 19:59:47 +0100
Subject: Why should I switch to Python?
References: <000701bf9c5a$b806b420$752d153f@tim>
	<oqity066km.fsf@titan.progiciels-bpi.ca>
	<slrn8ej2gc.b3t.warlock@raven.myth>
	<m3itxyz744.fsf@atrus.jesus.cam.ac.uk>
	<050420000914595917%pecora@anvil.nrl.navy.mil>
	<mt1z4k8oh1.fsf@astron.berkeley.edu>
	<m3pus4lbej.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <867lec74rg.fsf@g.local>

Michael Hudson wrote:

> <grin> - mind you "Riemann zeta function" is a pretty cool name for
> any function.  I also like "Weierstrass p-function", but mainly
> because the p is usually drawn in an entertainly deranged way.

Which, incidentally, Weierstrass himself invented.

-- 
Gareth McCaughan  Gareth.McCaughan at pobox.com
sig under construction


From robin at jessikat.demon.co.uk  Fri Apr 21 09:43:20 2000
From: robin at jessikat.demon.co.uk (Robin Becker)
Date: Fri, 21 Apr 2000 14:43:20 +0100
Subject: IDLE patch to use py2pdf
Message-ID: <zxD9rgA4rFA5EwpH@jessikat.demon.co.uk>

A minimal patch to allow the latest CVS idle to use Dinu Gherman's
py2pdf for printing is available at ftp://ftp.reportlab.com/idle.pat;
this patch hasn't been thoroughly tested yet, but I would like some
input on all the problems. 

After applying the patch using

cd idle
patch -p1 <idle.pat

you will need to extract ftp://ftp.reportlab.com/py2pdf.tgz (or .zip)
in the tools directory. This will create py2pdf in tools; edit the file
py2pdf/idle_print.py to reflect your local system.

The patch adds a file.print menu item which just copies the window to a
temporary file and then uses idle_print.py to convert to pdf and
executes acrobat reader on the result. It works quite well for me on my
win32 and suse systems, but will no doubt need some hacking to get the
print styling and or other things to suit.
-- 
Robin Becker


From brent.fulgham at xpsystems.com  Wed Apr  5 12:46:30 2000
From: brent.fulgham at xpsystems.com (Brent Fulgham)
Date: Wed, 5 Apr 2000 09:46:30 -0700 
Subject: Extending Python with C-Libs
Message-ID: <EDFD2A95EE7DD31187350090279C67676B49AB@THRESHER>

> I have a Library written in C with its Headerfiles but without the
> Sourcecode.
> Does anybody know an easy way to extend Python with this Library?
> 

Sure -- it's no different than adding hooks for the Win32 stuff or
whatnot.  Just create a new Python C extension module with an API
corresponding to the interesting pieces of the header file you have.

Then your C-extension is just a jump-off point where Python objects
come in, you convert them to whatever the C-library needs, and you
provide a valid Python-style return type.

Review the examples in the source code for details, or see Mark Lutz's
"Programming Python" book.

Regards,

-Brent



From gmcm at hypernet.com  Thu Apr 27 23:57:08 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Thu, 27 Apr 2000 23:57:08 -0400
Subject: The Simple Economics of Open Source
In-Reply-To: <B52E6E0E.521E%raffael@mediaone.net>
References: <1255272357-66924550@hypernet.com>
Message-ID: <1255228265-1994958@hypernet.com>

Raffael Cavallaro wrote:

[I wrote]
> > The closed source world is based on a diseased model
> > (pretending software is a product).
> 
> This "diseased model" has made billions for a number of firms,
> so, from an economic standpoint, your argument rings rather
> hollow, especially when compared to numerous open source ventures
> which have yet to post a cent of profit.

So has bashing people over the head and taking all their 
money. And many proprietary ventures have yet to make a 
cent of profit (like all those dot coms with ridiculous stock 
prices).

> 
> > Scarcity of an intellectual secret may make a software
> > package containing that secret more valuable, but scarcity of a
> > software package is likely to make it less valuable (since
> > support will be hard to find, and upgrades unlikely). OTOH it's
> > scarcity of a piece of artwork that makes it valuable.
> 
> You're conflating separate markets. Unique artworks (paintings,
> for example) have no real equivalent in the software world,
> because software sales are based on essentially unlimited
> identical copies.

'Twas your analogy.
 
> My example referred to IP in the art world - secret working
> methods that allowed artists to produce works their
> contemporaries could not - methods of bronze casting, for
> example, or formulas for oil painting media. These IPs are like
> closed source in that, without them, competitors are shut out of
> a lucrative market.

There's very little of that in code. The question "how do they 
do that?" is rarely unanswered for longer than a few months.
 
> My point was, and is, that it makes no economic sense to open
> source an IP that could be used to create a monopoly. MS
> understands this, and works ceaslessly to create such monopolies.
> Note that this has *nothing whatsoever* to do with what is best
> for consumers. 

You really think MS's "monopoly" is based on IP?

> I think part of the open source community's problem is this naive
> belief that just because a certain model (open source) leads to
> products that benefit consumers more, markets will allow
> consumers to choose this production model. This belief is naive
> because it fails to consider that the producers' interests are in
> no way congruent with the consumer's.

An indicator that the model is diseased.

> The producer has every
> reason to go closed source and attempt to secure a quick
> monopoly. We live in this world now, and there's no reason to
> believe that it's going to go away any time soon. I think open
> source advocates grossly underestimate the power of monopolies,
> and the unwillingness of consumers to break ranks even though it
> *may* provide long term benefits, because breaking ranks entails
> large short term costs, especially in network utility - i.e., the
> loss of interoperability with the rest of the monopolist's
> customers.

They are not "monopolies". They are large companies with a 
lot of money, salesman, distribution channels, lawyers and 
programmers. The competition has quite a few programmer, 
and just started to get a little money.

If instead of analyzing profits you looked at market share, in 
the area of web servers at least, you would have to draw the 
opposite conclusion.
 
> BTW, all software is *not* support. This is another open source
> community bias which comes from living in the server room. Lots
> of software is *client* software, which requires little or no
> support from the vendor. 

And client software is often free. Where software has client 
and server, the money's all made from the server. If the 
vendor's smart, they'll open up the client side so they won't 
have to waste time developing it.

> Instead, these support services are
> provided by full time IT people, who may be certified by the
> vendor, but who are paid by the enterprise. In other words, MS
> makes their money on enterprise licences for Office, not for
> troubleshooting the enterprise's problems with Office.

Bzzt. A 10,000 seat license complains that X is broken, X will 
get fixed. (And when enterprise learns that Y requires only 
50% the support costs, they will switch.)

>  For the
> most lucrative software businesses, software *is* a product, not
> a service.

No. The only thing they share is that they both require 
investment, as does almost anything else.

Congress does not understand software. Lawyers and judges 
do not. Most businesses do not. Most people do not. The 
world has never before encountered anything with a 
vanishingly small marginal cost of production, but something 
that has so much "value".

Software companies adopted the "product" model. The fact 
that some of them make money does not make the model 
sane. It just means software can be lucrative.

In fact, let's compare Apple and MS. Apple could be where 
MS is now, but Steve Jobs repeatedly insisted on shutting out 
the competition, keeping developers ignorant and making 
them pay. MS, believe it or not, gives away a great deal. 
Anybody can browse all the MSDN docs. If you buy MSVC, 
you have the source to the c runtime library. MS publishes 
large amounts of specs. Yes, they usually withhold something 
- about a six month to one year lead on the market. IP in this 
market has a very short shelf life.

They're commanding position has nothing to do with IP. Most 
of the maybe-almost-original things they've done (ODBC and 
COM) are pretty thoroughly documented. Their success is due 
to business practices that are common in, say, the food 
industry or the mega-store segment, but are (rightfully) 
considered evil in the software market.

The open source model(s) will almost certainly require 
adjustment - they are mostly a reflection of the fact that the 
current model is warped. But the big boys are dinosaurs. They 
will change their models, or die (Bill won't make it, but neither 
will Scott or Larry).

But before things become sane, people will have to figure out 
a "sale" is; and we're nowhere close to that :-(.

- Gordon



From richard_chamberlain at ntlworld.com  Wed Apr 19 06:51:20 2000
From: richard_chamberlain at ntlworld.com (Richard Chamberlain)
Date: Wed, 19 Apr 2000 11:51:20 +0100
Subject: testing true condition
References: <Pine.BSF.4.10.10004191138390.292-100000@hercule.perceval.be>
Message-ID: <8dk337$349$1@gxsn.com>

You could use something like this:

file=open("test.txt","r")
for line in file.readlines():
    print line

Richard


"Gregoire Welraeds" <helpdesk at perceval.be> wrote in message
news:Pine.BSF.4.10.10004191138390.292-100000 at hercule.perceval.be...
> Hi,
>
> I know that I can't do something like the following:
>
> >>> while (line= f.readline()):
> ...
>
> as i could in C.  Is there any replacement tricks ?
> I know that I could use something with f.read() but i don't want to have
> the file completly loaded in memory
>
> Best regards,
>
> --
> Life is not fair
> But the root password helps
> --
>
> Gregoire Welraeds
> greg at perceval.be
> Perceval Development team
> --------------------------------------------------------------------------
-----
> Perceval Technologies sa/nv Tel: +32-2-6409194
> Rue Tenbosch, 9 Fax: +32-2-6403154
> B-1000 Brussels general information:   info at perceval.net
> BELGIUM technical information: helpdesk at perceval.net
> URL: http://www.perceval.be/
> --------------------------------------------------------------------------
-----
>
>




From tjg at avalongroup.net  Mon Apr  3 16:53:17 2000
From: tjg at avalongroup.net (Timothy Grant)
Date: Mon, 03 Apr 2000 13:53:17 -0700
Subject: calendar.py
References: <38E901BB.A41CEB25@rtpro.net>
Message-ID: <38E904BD.BD0274F9@exceptionalminds.com>

Tesla Coil wrote:
> 
> Who maintains calendar.py?  I'd like to propose inclusion
> of functions to facilitate such calculations as "what will the
> date be in 45 days," or "print dates for twelve 28 day cycles
> beginning from 3/1/2000."   Excuse the undercooked code,
> but something like the following:

Not directly on point, but you might like to look at M.-A. Lemburg's
mxDateTime module which--I believe--does exactly what you are looking
for.

-- 
Stand Fast,
    tjg.

Timothy Grant                         tjg at exceptionalminds.com
Chief Technology Officer              www.exceptionalminds.com
Red Hat Certified Engineer                      (503) 246-3630
Avalon Technology Group, Inc.              fax  (503) 246-3124
>>>>>>>>>>>>Linux...Because rebooting isn't normal<<<<<<<<<<<<



From hschroeder at gmx.net  Sun Apr 30 12:05:45 2000
From: hschroeder at gmx.net (Henning Schroeder)
Date: Sun, 30 Apr 2000 18:05:45 +0200
Subject: [Announce] BoboMail 0.3 - a webmailer
Message-ID: <390C59D9.797EFAEC@gmx.net>

Hi,
I'm pleased to announce:

BoboMail Version 0.3
====================

BoboMail is a so-called webmail application using Zope's ZPublisher and
ZTemplates. It provides mail access through a web interface and is 
distributed under the GNU GPL.

Current features:
 * read, write, reply, forward and delete messages
 * direct support for the following MIME types:
   message/rfc822, message/delivery-status, multipart/alternative,
   multipart/mixed, multipart/report, multipart/related,
   text/plain, text/html, text/x-vcard, image/jpeg, 
   image/png, image/x-xpixmap, image/gif
 * works with POP3-servers
 * easy to modify and to customize through HTML-templates
 * modular design provides easy interface for other authentication
   methods and mail protocols
 * provides optionally its own fast multithreading webserver
(BoboMailHTTPD)
 * multi-lingual support (i18n) - currently English and German

For more information go to http://bobomail.sourceforge.net

Call for help: Can anyone help to improve the speed by making a Python-C
module
for a RFC822-/MIME-parser with Swig (e.g. of the c-client library) ?

  Henning





From emile at fenx.com  Wed Apr 12 06:02:13 2000
From: emile at fenx.com (Emile van Sebille)
Date: Wed, 12 Apr 2000 03:02:13 -0700
Subject: Python/Perl Popularity (Re: A Mountain of Perl...)
References: <200004102111.HAA03380@envy.fulcrum.com.au> <8d0pcq$7ti$1@nnrp1.deja.com> <87og7fopvs.fsf@ifm.uni-kiel.de>
Message-ID: <003401bfa466$2f5e6c80$01ffffc0@worldnet.att.net>

Nice utility Janko.  Thank-you.

For anyone interested, here are the diffs I threw at this
for use on my Win95 system:

Compare: (<)D:\Program Files\Python\Lib\Fenx\Lib\docindex.py (12132l
bytes)
   with: (>)D:\Program Files\Python\temp.py (12063l bytes)

51c51
< pager = None # 'less'
---
> pager = 'less'
318,321c318
<    try:
<     self.docdb = os.path.join(os.environ['HOME'],'.ipphelp')
<    except:
<     self.docdb = os.path.join(os.environ['TMP'],'.ipphelp')
---
>    self.docdb = os.path.join(os.environ['HOME'],'.ipphelp')
421c418,422
<  help = docreader((r'D:\Program Files\Python\Doc_1.5.2\lib',))
---
>  help = docreader(('/home/user/Doc/html/NumPy/',
>        '/home/user/PYHTML/lib/'))
>
>  help('round')
>
423,425d424
<  help('round')
<
<


Emile van Sebille
emile at fenx.com
-------------------


----- Original Message -----
From: Janko Hauser <jhauser at ifm.uni-kiel.de>
Newsgroups: comp.lang.python
To: <python-list at python.org>
Sent: Tuesday, April 11, 2000 10:21 PM
Subject: Re: Python/Perl Popularity (Re: A Mountain of Perl...)


> Use docindex.py!
>
> This module is definetly no substitute for perldoc but a start that
> answers your question. It gets the relevant information from the html
> tree which needs to be installed on your system.
>
> A benefit over docstrings is, that one does not need to import a
> module. This is a prototype, but shows that it is possible. Not that
> this will change your mind about Python :-)
>
> HTH,
> __Janko
>






From z3penguin at penguinpowered.com  Sun Apr 23 11:19:58 2000
From: z3penguin at penguinpowered.com (Z 3 Penguin)
Date: Sun, 23 Apr 2000 11:19:58 -0400
Subject: Posting CGI
Message-ID: <8dv4j8$2bs$1@bob.news.rcn.net>

How would I go about submiting a CGI post form in python?  (This sounds like
it should be simple, but I can't figure it out :-) Thanks.

-Z




From gherron at aw.sgi.com  Fri Apr 14 13:46:54 2000
From: gherron at aw.sgi.com (Gary Herron)
Date: Fri, 14 Apr 2000 10:46:54 -0700
Subject: Questions on 1.6a2's string methods
References: <38F6C904.76B538E5@concentric.net>
Message-ID: <38F7598E.97A07412@aw.sgi.com>

Manus Hand wrote:
> 
> 1.  I see that string objects now support (as methods) most of the
>     functions from the string module.  Among these are:
>         upper, lower, split, strip, find
>     and surely others.  My question is, why is capwords (which seems
>     to be in the same genre as upper and lower) not a method?
> 
> 2.  Along these same lines, since split() became a method of the
>     string type, wouldn't it make sense to make join() a method of
>     the list type?
> 
> 3.  Are the classlike standard types (list, dictionary, and now
>     string) equipped with a __dict__ attribute?  I can see the names
>     of all functions supported by a user-defined class by saying
>     className.__dict__.keys(), but I cannot see the list of methods
>     for the string type (at least not in the same way).  Thus my
>     need to ask silly questions like #1 above (maybe capwords is
>     there by some other name??)
> 
 
It looks like `capwords' has changed name to `title' with a slight
change in functionality.

>>> '  This    is   a test'.title()
'  This    Is   A Test'

This capitalizes words, but does not change white space as `capwords'
did.

Take a look at the other functions:

>>> ''.__methods__
['capitalize', 'center', 'count', 'endswith', 'expandtabs', 'find',
'index', 'isdigit', 'islower', 'isspace', 'istitle', 'isupper', 'join',
'ljust', 'lower', 'lstrip', 'replace', 'rfind', 'rindex', 'rjust',
'rstrip', 'split', 'splitlines', 'startswith', 'strip', 'swapcase',
'title', 'translate', 'upper']

Most are familiar.   I'm happy to see 'startswith' and 'endswith'.

-- 
Dr. Gary Herron <gherron at aw.sgi.com>
206-287-5616
Alias | Wavefront
1218 3rd Ave, Suite 800, Seattle WA 98101



From jwbnews at scandaroon.com  Sun Apr 23 01:41:49 2000
From: jwbnews at scandaroon.com (John W. Baxter)
Date: Sat, 22 Apr 2000 22:41:49 -0700
Subject: code not true?
References: <Pine.GSO.4.10.10004230758200.15546-100000@sundial>
Message-ID: <jwbnews-291BC8.22414922042000@news.olympus.net>

In article <Pine.GSO.4.10.10004230758200.15546-100000 at sundial>, Moshe 
Zadka <mzadka at geocities.com> wrote:

> On Sat, 22 Apr 2000, John W. Baxter wrote:
> 
> > > > >>> print 7.0 == 7
> > > > 1
> > > > >>> print (.07 * 100) == 7
> > > > 0
> > > 
> > > Because computers represent floats with too little bits.
> > > Never ever compare two floats for equality!
> > > 
> > Not quite.  No number of bits is sufficient to represent .07 exactly in 
> > the typical (and IEEE standard) representation.  
> 
> Yes, but enough bits could make sure that 0.07*100 is indistinguishable
> from 7.0

We have that many bits now...we just have to define "indistinguishable" 
properly.  For example

Python 1.5.2 (#2, Dec 23 1999, 13:16:10)  [GCC egcs-2.91.66 19990314 
(egcs-1.1.2  on bsdos4
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> a = .07 * 100
>>> b = 7
>>> epsilon = 1e-15
>>> abs(a-b) < epsilon
1
>>> a == b
0
>>> 

The above is not how == is defined.  It is one way (with a suitable 
epsilon...this one may not be suitable for one's particular problem) to 
test two numbers for "close enough to be indistinguishable, so we'll 
call them equal".  (And don't forget the abs()...forgetting that can 
ruin an evening.)

   --John (who also fought with plus zero and minus zero, both in 
integers and floating point)

-- 
John W. Baxter   Port Ludlow, WA USA  jwbnews at scandaroon.com


From exarkun at flashmail.com  Fri Apr 21 11:03:14 2000
From: exarkun at flashmail.com (Jp Calderone)
Date: Fri, 21 Apr 2000 15:03:14 +0000
Subject: The Simple Economics of Open Source
References: <14591.15630.192472.440375@goon.cnri.reston.va.us>
	 <slrn8fv2rd.1kt.neelk@brick.cswv.com> <3900611F.3B525BDA@cfdrc.com>
	 <FtDG4z.774@world.std.com> <uFaBgiAr5HA5EwJt@jessikat.demon.co.uk>
Message-ID: <39006DB2.FD215253@flashmail.com>

Robin Becker wrote:
> 
> In article <FtDG4z.774 at world.std.com>, Will Ware <wware at world.std.com>
> writes
> >Ed (elb at cfdrc.com) wrote:
> >> ... I didn't think much of the article...
> >> ...they gave the impression that programmers behave
> >> somewhat like monkeys, motivated primarily by concerns
> >> about group status and dominance...
> >
> >They also mentioned, but immediately belittled, the possibility
> >that altruism might be a real motivation. They may have been
> >projecting their own thinking upon programmers in doing so. I'm
> >not an open source purist, but I can easily see that a future world
> >with a lot of free software will be a lot more pleasant for me to
> >live in than one from which free software is absent. Benefits to
> >the rest of society and to myself are not in conflict here. Maybe
> >that's not really altruism, maybe it's just long-term thinking, a
> >motivation they didn't mention at all.
> 
> I'm afraid your desire for a better world reduces your altruism
> quotient.
> Doing it because it helps others is altruism, doing it to help your self
> and or relatives isn't really. The indian otters which help injured
> friends were thought to be altruists until it was shown that by doing so
> they reduced their own risk of predation (the slow one attracted the
> predators most). Real altruism is rare.
> 

 Of course; for the reason you allude to, it isn't evolutionarily
advantagious to be altruistic.  

The phrase "enlightened self-interest" comes to mind :)

 Jp

> [snip]

> --
> Robin Becker

-- 
 3:01pm up 23 days, 20:08, 2 users, load average: 1.29, 0.90, 0.47


From meh9 at cornell.edu  Fri Apr  7 11:08:33 2000
From: meh9 at cornell.edu (Matthew Hirsch)
Date: Fri, 07 Apr 2000 11:08:33 -0400
Subject: Algorithm: combinations of (k) taken from (n) values
References: <meh9-ACCD3B.16041406042000@news.cit.cornell.edu> <38ED5285.719F26D7@udel.edu>
Message-ID: <meh9-00C86E.11083307042000@news.cit.cornell.edu>

Thanks for your help.  This is exactly how I pictured it, starting with 
lists of numbers up to n, and then adding on to them from there.

Matt




In article <38ED5285.719F26D7 at udel.edu>, Charles Boncelet 
<boncelet at udel.edu> wrote:

> Matthew Hirsch wrote:
> 
> > Hi All,
> >
> > Onto the next question...
> >
> > Can anyone think of an algorithm to store as lists all possible
> > combinations of k numbers taken from n possible numbers.  For example,
> > given 5 values, I want to choose 2.  The number of possible combinations
> > is given by 5!/(3!2!)=10.  They are:
> >
> 
> Funny you should ask, since I just wrote one two days ago:
> 
> def combs(n,k=None):
>     """returns sorted list of k items taken from n."""
>     if k==None:
>         k=n
>     l = []
>     for c in range(n):
>         l.append([c])
>     for i in range(k-1):
>         li = []
>         for c in l:
>             for j in range(c[-1]+1,n):
>                 li.append(c+[j])
>         l = li
>     return l
> 
>     Charlie Boncelet
> ------
> Charles Boncelet, University of Delaware,
> On sabbatical at ADFA, Canberra Australia,
> Home Page: http://www.ece.udel.edu/~boncelet/
> 
>


From roy at popmail.med.nyu.edu  Tue Apr  4 21:21:23 2000
From: roy at popmail.med.nyu.edu (Roy Smith)
Date: Tue, 04 Apr 2000 21:21:23 -0400
Subject: What is python's language level?
References: <38EA1B8F.A797A130@lmco.com> <38E9DDD3.3313FF6@bioreason.com> <slrn8ekddq.gm3.zawrotny@gecko.sb.fsu.edu>
Message-ID: <roy-D857DA.21212304042000@netnews.nyu.edu>

zawrotny at gecko.sb.fsu.edu (Michael Zawrotny) wrote:
> Lets start with the fact that nroff and html are markup
> languages not programming languages.

HTML certainly is markup, but I don't know if I'd call nroff markup.  
It's much more like a programing language.

Nroff has arithmetic expressions and operators, conditional execution, 
several different forms of data structures (number registers, text 
diversions), user-definable functions with argument passing, traps, etc.  
I don't remember if it supports recursion or not, but I think so.

I will admit, the syntax leaves something to be desired (an obfuscated 
nroff contest would be pointless because it'd be too hard to pick one 
winner).  I'll leave it to the theory wonks to figure out if it's Turing 
complete or what not, but it's very much more than just a markup 
language.


From ajung at sz-sb.de  Fri Apr  7 13:22:51 2000
From: ajung at sz-sb.de (Andreas Jung)
Date: Fri, 7 Apr 2000 19:22:51 +0200
Subject: Meaning of __rlshift__ and __rrshift__ ?
Message-ID: <20000407192251.A1974@yetix.sz-sb.de>

The documentation say nothing concerning the meaning
of the functions above ?

Any comments ?

Cheers,
Andreas



From rogerha at ifi.uio.no  Fri Apr 14 10:11:20 2000
From: rogerha at ifi.uio.no (Roger Hansen)
Date: 14 Apr 2000 16:11:20 +0200
Subject: Python tools for PDE and Linear algebra
Message-ID: <ya0vh1kzsaf.fsf@grindaflet.ifi.uio.no>

Hi everyone.

I'm looking for some tools for solving Partial Differential Equations,
and I want to program in Python of course. I've searched the net and
found some Linear Algebra packages but not PDE tools. Does anyone know
some PDE tools or are working on such projects? Support for different
numerical methods as Finite Difference and Finite Element Methods
would be nice.


TTFN,

Roger


From aahz at netcom.com  Sun Apr 16 00:04:04 2000
From: aahz at netcom.com (Aahz Maruch)
Date: 16 Apr 2000 04:04:04 GMT
Subject: Python idiom: Multiple search-and-replace
References: <20000412100816.A2431031@vislab.epa.gov> <20000412120012.D2446939@vislab.epa.gov> <0s1J4.655$rc9.190631424@newsb.telia.net> <20000414084023.A2012@vislab.epa.gov>
Message-ID: <8dbe3k$tkt$1@nntp9.atl.mindspring.net>

In article <20000414084023.A2012 at vislab.epa.gov>,
Randall Hopper  <aa8vb at yahoo.com> wrote:
>
>It's hard to believe we're burning 8.9 sec out of 11.6 seconds (77%) just
>in the simple loop overhead of iterating over lines.  Where's that JIT
>Python interpreter when you need it ;-)  

It's not the loop overhead per se so much as the readlines() overhead.
readline() is relatively inefficient.
--
                      --- Aahz (Copyright 2000 by aahz at netcom.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

"First one back from the bathroom gets to be the bottom."


From Vladimir.Marangozov at inrialpes.fr  Sat Apr 15 11:49:58 2000
From: Vladimir.Marangozov at inrialpes.fr (Vladimir Marangozov)
Date: Sat, 15 Apr 2000 17:49:58 +0200
Subject: an intern-like memory saver
References: <8d5ing$heb$1@nnrp1.deja.com> <38F7318F.AE4F0EE5@endea.demon.nl> <8d8fe0$nr5$1@nnrp1.deja.com>
Message-ID: <38F88FA6.35ED9CB9@inrialpes.fr>

sjmachin at lexicon.net wrote:
> 
> I was happy enough with the idea of having an instrumented malloc in my
> own Python (when I get around to compiling it with gcc), but was
> wondering about the possibility of having an instrumented malloc as
> part of the *standard* Python distribution.

Heads up! It's on the way -- Python 1.6 is expected to be user-malloc
friendly, so you'll be able to link your executable with the malloc of
your choice. Actually, this is painful and you have few chances to do
it right if you're not a Python guru.

No default replacement of libc malloc is envisioned for the distribution,
though. At least not yet.

encouraging'ly y'rs
-- 
       Vladimir MARANGOZOV          | Vladimir.Marangozov at inrialpes.fr
http://sirac.inrialpes.fr/~marangoz | tel:(+33-4)76615277 fax:76615252


From sabren at manifestation.com  Sun Apr  9 22:30:24 2000
From: sabren at manifestation.com (Michal Wallace (sabren))
Date: Sun, 9 Apr 2000 22:30:24 -0400 (EDT)
Subject: indexed dictionaries?
Message-ID: <Pine.BSF.4.10.10004092225040.1361-100000@chi.pair.com>

Hey all,

  I just KNOW this has been done before, but I can't seem to find an
example... I'm looking for a cross between a dictionary and a list,
such that keys() always returns keys in the order in which they were
added, and that instance[0] refers to the first key...

eg:

>>> import IdxDict
>>> x = IdxDict()
>>> x["a"] = "first one"
>>> x["b"] = "second one"
>>> x[0]
'first one'
>>> x["a"]
'first one'
>>> x.keys()
('a', 'b')

Anyone got something like this?  

Thanks,

- Michal
-------------------------------------------------------------------------
http://www.manifestation.com/         http://www.linkwatcher.com/metalog/
-------------------------------------------------------------------------




From zeitlin at seth.lpthe.jussieu.fr  Wed Apr 19 11:55:18 2000
From: zeitlin at seth.lpthe.jussieu.fr (Vadim Zeitlin)
Date: 19 Apr 2000 15:55:18 GMT
Subject: How to compile wxpython in win98 using vc6.0 for wxWindows2.1.15!
References: <8djn4c$22ch$1@news.cz.js.cn>
Message-ID: <slrn8frm0r.d5m.zeitlin@seth.lpthe.jussieu.fr>

On Wed, 19 Apr 2000 15:37:12 +0800, fmr <mrfx at 263.net> wrote:
>wx.cpp:
>wchar.h(700) : error C2733: second C linkage of overloaded function
>'wmemchr' not allowed
>wchar.h(699) : see declaration of 'wmemchr'

 This is really weird because:

1. my wx/wxchar.h doesn't have 700 lines (it's about 650)
2. there is no mention of wmemchr in it at all
3. are you trying to compile it in Unicode mode (won't work under Win98
   anyhow then)? Otherwise, wxxx() functions shouldn't be referenced
   at all.

 Check your installation of wxWin, in particular check that you didn't
mix the new headers with some left overs of a previous version.

 Regards,
VZ


From pj at sgi.com  Wed Apr 12 18:50:26 2000
From: pj at sgi.com (Paul Jackson)
Date: 12 Apr 2000 22:50:26 GMT
Subject: Python/Perl Popularity (Re: A Mountain of Perl...)
References: <200004102111.HAA03380@envy.fulcrum.com.au> <38F4996F.E3BD2626@acm.org> <8d2p3q$2vj9o$1@fido.engr.sgi.com> <snwrf1ls.fsf@dan.resnet.rochester.edu>
Message-ID: <8d2uji$301cu$1@fido.engr.sgi.com>

|> Nope, both lynx, and links, exist.

oh - could you offer a link to "links" -- it's
a hard name to search for -- too many false hits.

-- 

=======================================================================
I won't rest till it's the best ...	   Software Production Engineer
Paul Jackson (pj at sgi.com; pj at usa.net) 3x1373 http://sam.engr.sgi.com/pj


From kc5tja at garnet.armored.net  Sun Apr 30 20:19:53 2000
From: kc5tja at garnet.armored.net (Samuel A. Falvo II)
Date: 1 May 2000 00:19:53 GMT
Subject: Do I always have to write "self." ?
References: <270420001706386648%pecora@anvil.nrl.navy.mil> <slrn8ghc6t.f01.scarblac-spamtrap@flits104-37.flits.rug.nl> <8ecsgj$6lp$1@slb0.atl.mindspring.net> <m3bt2tonyg.fsf@atrus.jesus.cam.ac.uk> <300420001008012011%pecora@anvil.nrl.navy.mil><20000430170952.A502@datapro.co.za> <B531EC9B.57C1%abrahams@mediaone.net><slrn8gp7ei.ilf.kc5tja@garnet.armored.net> <B53241E5.580F%abrahams@mediaone.net>
Message-ID: <slrn8gqf3a.k65.kc5tja@garnet.armored.net>

In article <B53241E5.580F%abrahams at mediaone.net>, Dave Abrahams wrote:
>As Simonyi, the author of the Hungarian notation concept writes, "the basic
>idea is to name all quantities by their types". If you are naming quantities
>by their interface, you are taking a step away from Hungarian. Do as I do,
>naming quantities by their *role*, and you have taken a long leap away from
>Hungarian.

If an object exposes a particular interface, then it is guaranteed to be of
a particular type.  Defining parameters in terms of their roles is
analogous.  Again, I just don't see the distinction.

>I can see how it might help to name a variable by its interface, especially
>in Python if you don't write comments. But: do you write listDogs, tupleDogs
>or sequenceDogs when you're being passed a sequence of dogs? Don't the

Yes.  In particular, listDogs, tupDogs, or seqDogs.  Depending on certain
issues, I may well go so far as to write listOfDogs, tupleOfDogs, or even
sequenceOfDogs (<- especially when dealing with CORBA).

>Suppose you're writing UserDict. What do you name your function arguments
>which represent keys and values? Not too many requirements on the value
>type, are there? Is Hungarian applied only selectively?

The names would be key and value, accordingly.  Variables which hold keys
would be called keyX, keyUser, etc.  Likewise with values.

Finally, yes, Hungarian is best used only selectively.  I strongly dislike
redundant information source code.  For example:

	def PrintName( name ):
	   print "The name is %s" % name

	.
	.
	.

	for listOfNames in listOfListOfNames:
	   for name in listOfNames:
	      PrintName( name )

Note that I do NOT use Hungarian inside of PrintName(), because it would be
completely redundant.  Within the context of the function, the type of a
name would obviously have to be a string.

But outside the function, where the context could be *anything*, I use
Hungarian notation to make it patently obvious the types of data I'm dealing
with.

Some might say, "listOfListOfNames" isn't Hungarian notation.  Isn't it?
Isn't Hungarian notation just a consistent system by which you tag variable
names with the intended datatypes of the objects they hold/refer to?  If we
chose to use an abbreviation, we'd get llNames, which is decidedly closer to
what Microsoft uses in their software.

This technique is not new, nor was it "invented" at Microsoft either.
Mathematicians have often used the "Factor Label" method for solving
equations for eons.  There's a whole algebra of working with units.  For
example, if you're going X miles/hour, and you've been travelling for Y
hours, then the distance you've travelled is Z=(X*Y) miles.  Or:

	distanceZ = speedX * timeY

Why can't this be applied to computer programming too?  Given a pointer to a
pointer to an object,

	pObject = *ppObject;	/* The * operator cancels out a p */

or:

	pArgument = argv[2];
	
Oh, here's a beautiful example of 'reverse' Hungarian notation, which
someone else on this newsgroup recommended/inquired about:

	void main( int argc, char *argv[] );
			  ^	      ^
-- 
KC5TJA/6, DM13, QRP-L #1447
Samuel A. Falvo II
Oceanside, CA


From philh at vision25.demon.co.uk  Sun Apr 30 16:11:15 2000
From: philh at vision25.demon.co.uk (phil hunt)
Date: Sun, 30 Apr 2000 21:11:15 +0100
Subject: The Simple Economics of Open Source
References: <B527E7A5.51AA%raffael@mediaone.net> <MPG.136cc01af4ab703b989709@news.earthlink.net>  <raffael-4899DC.23562823042000@news.ne.mediaone.net>
Message-ID: <slrn8gp4r2.eb1.philh@vision25.demon.co.uk>

On Mon, 24 Apr 2000 03:57:00 GMT, 
Raffael Cavallaro <raffael at mediaone.net> wrote:
>In article <MPG.136cc01af4ab703b989709 at news.earthlink.net>, Richard 
>Hale <rkhale at earthlink.net> wrote:
>
>>The Open Source software that I have used has been very valuable to me.  
>>Did those who produced it not see as much value in it?
>
>Well, they probably knew they couldn't sell it as a closed source 
>product and make much money from it.
>
>Fact is, only commodity items, that are known to many programmers, are 
>open sourced. 

Eh?

Consider Parrot and leafwa, two open source packages I am writing.
When I put them up on my website, they were known to exactly *one* 
programmer.

>If there were sufficient demand to support a closed source 
>business, then it would have happened that way.

Demonstrably wrong.

At the time Linus wrote his kernel, there were several closed source
Unices for x86.

>The closed source biggies are not open sourcing the crown jewels. Why? 
>Cause there's plenty of cash still left to milk from those cows. Notice 
>that Apple has open sourced only those parts of MacOS X that *aren't* 
>lucrative consumer software. Where's the Darwin QuickTime (not the 
>streaming server, but the client?) Where's Darwin Quartz, Darwin Aqua, 
>Darwin Cocoa? Don't hold your breath for these, because these parts of 
>the OS are what allows Apple to turn a profit. If they became open 
>source commodities, Apple would become just another clone maker.

This is a reasonably comment.

If Apple open sourced the Mac user interface, they'd be worse off.

What they could do however, is disclose the source (so you can look
at it, but not do anything with it) and and put it on a time-delayed open
source licence (so any release will become open source after X years).

At the same time, they could release a version of MacOS for Wintel
boxes. They could base it on the Linux kernel; this would mean they 
wouldn't have to go to their own separate effort to write device drivers
etc. (They could of course do this without open sourcing the Mac
user interface code).


-- 
***** Phil Hunt ***** send email to phil at comuno.com *****
Moore's Law: hardware speed doubles every 18 months
Gates' Law: software speed halves every 18 months 


From fredrik at pythonware.com  Tue Apr  4 05:32:38 2000
From: fredrik at pythonware.com (fredrik at pythonware.com)
Date: Tue, 04 Apr 2000 09:32:38 GMT
Subject: Unicode program representation
References: <015c01bf9d0b$76f9b2a0$59dd3fcb@neil> <3xXF4.1060$n68.186533888@newsb.telia.net> <001201bf9d62$3e1ac1c0$36dd3fcb@neil> <%o3G4.1111$n68.191610880@newsb.telia.net> <oqn1nb11l3.fsf@titan.progiciels-bpi.ca>
Message-ID: <8cccrm$cc3$1@nnrp1.deja.com>

Francois Pinard wrote:
> > hmm.  wouldn't that mean that we end up using different encodings in
> > different parts of the script?  feels a little scary, to say the
least...
>
> This is very likely to happen, but this should not be seen as scary.
>
> Modules handling linguistic or cultural matters might come from
various
> sources, which sources do not necessarily use the same encodings, but
yet
> be used simultaneously in a program building on them all.

sure, I'm all for letting different *scripts* (or different
modules in a larger program) use different encodings; say
utf-8 in some files, iso latin 2 in another, chinese in a
third one...

what scares me is if we end up with multiple encodings in the
same source file.

(and no, it's not because we're developing a python
development environment -- supporting this in our python
editor would be fairly trivial ;-)

</F>


Sent via Deja.com http://www.deja.com/
Before you buy.


From m.faassen at vet.uu.nl  Mon Apr  3 12:42:22 2000
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 3 Apr 2000 16:42:22 GMT
Subject: Who is this years winner?
References: <000101bf9d2d$b724e1c0$e02d153f@tim>
Message-ID: <8cahle$356$1@newshost.accu.uu.nl>

Tim Peters <tim_one at email.msn.com> wrote:
> Since then, potential winners other than ME have,
> umm, gracefully withdrawn.  I'm sure it's just coincidence that the PSA
> Steering Committee has never been heard from again.

That's susp[PSU/UsenetInterrupt$_]
Mr. Peters, Nr. Timbot,

The PSU notifies you for the first and last time have not paid our fees for:

  * Use of Guido's time machine to wipe out the existence of Peter Tims,
    2nd winner of the Pythonic Award.

  * Liquidation and replacement with bot of F. Lundth, who threatened to
    win the Pythonic Award.

Please transfer the money to our Swiss bank account immediately, or you will
be discontinued.
[PSU/UsenetInterruptEnd$_]ious, Tim! Something to do with the PSU? :)

There-is-no-such-thing-as-the-Python-Secret-Underground-ly yours,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From R.Brodie at rl.ac.uk  Tue Apr 25 09:25:19 2000
From: R.Brodie at rl.ac.uk (Richard Brodie)
Date: Tue, 25 Apr 2000 14:25:19 +0100
Subject: what does XDR stand for? and what is it?
References: <002101bfaeb5$b57f5dc0$2801000a@iel.ie>
Message-ID: <8e46c6$1aoa@newton.cc.rl.ac.uk>

"Shaun" <shogan at iel.ie> wrote in message news:002101bfaeb5$b57f5dc0$2801000a at iel.ie...
>
> im here looking and trying to figure out the xdrlib module,
> it says here - http://www.python.org/doc/current/lib/module-xdrlib.html
> "The xdrlib module defines two classes,
> one for packing variables into XDR representation,
> and another for unpacking from XDR representation.
> There are also two exception classes."
> my question is what is XDR? what does it stand for?

XDR stands for 'external data representation'. It has a few simple
encoding rules: use IEEE floating point, send integers big-endian etc.
It is used primarily in Sun style remote procedure calls - those
associated with the portmapper (e.g. NFS), though it sees some use
in portable file formats also.

Sun produced an informational RFC about it, which should be on your
nearest repository.





From dennis_marti at yahoo.com  Mon Apr 24 10:53:14 2000
From: dennis_marti at yahoo.com (dennis_marti at yahoo.com)
Date: Mon, 24 Apr 2000 14:53:14 GMT
Subject: crypt() results
References: <Pine.BSF.4.10.10004232258410.24749-100000@chi.pair.com>
Message-ID: <8e1n4k$jt7$1@nnrp1.deja.com>

In article <Pine.BSF.4.10.10004232258410.24749-100000 at chi.pair.com>,
  "Michal Wallace (sabren)" <sabren at manifestation.com> wrote:
>
>
> Well. I tried porting both Crypt::UnixCrypt and Crypt::PasswdMD5 to
> python.  And neither one of the python versions give the same results
> as the perl versions. They come out in the same format, but with
> different values.
>
> I've pounded my head, pulled out my hair, and finally given up. I
> simply don't know enough about the algorithms or perl's bit twiddling
> to understand what these modules are actually doing.

It wasn't in the bit twiddling. You reversed the conditionals in the
first for loop in unix_md5_crypt().

/home/dennis $ diff md5crypt.py md5crypt.py.orig
73,74d72
<             ctx = ctx + final[:16]
<         else:
75a74,75
>         else:
>             ctx = ctx + final[:16]

Hope you didn't lose too much hair.

Dennis Marti


Sent via Deja.com http://www.deja.com/
Before you buy.


From mskott at image.dk  Mon Apr 24 08:05:12 2000
From: mskott at image.dk (Martin Skøtt)
Date: Mon, 24 Apr 2000 12:05:12 GMT
Subject: Trouble compiling PyGreSQL 2.4 on Linux
References: <Pine.LNX.4.21.0004241126280.6019-100000@fep132.fep.ru>
Message-ID: <k8hnhfh0.fsf@jumbo.skott.dk>

Oleg Broytmann <phd at phd.russ.ru> writes:

> On Sun, 23 Apr 2000, Martin Sk?tt wrote:
>    You forget to mention path to postgres libs (-L /usr/local/pgsql/lib).

I knew it was something like that, thanks :-)
> > I can't use the make command mentioned in the readme.linux file
> > because the simply isn't any make file in the archive. Can you help me
> > getting it compiled?
> 
>    There is no Makefile 'cause the module is intended to use standard
> Python Makefile.pre.in/Setup.in scheme.

That is onlu if you chose the standard Python module installation. Red
Hat has made something special where modules are stored in a directory
called lib-dyn/ some where. To me it seems like the modules then are
loaded automaticly.
As I mention in my first posting the readme.linux file describes how
to do an red hat install (make redhat) but it doesn't seem to work. An
address is also mentioned for an RPM but the file doesn't exist on the server.

-- 
Martin Sk?tt
mskott at image.dk

'In a certain sense, all decent programming languages are equally powerfull'
Dr. Alan Turing


From amused at webamused.com  Sun Apr 30 14:43:40 2000
From: amused at webamused.com (Joshua Macy)
Date: Sun, 30 Apr 2000 18:43:40 GMT
Subject: Comments/Help on Index-module
References: <390c564b.102144@news.c2i.net>
Message-ID: <390C721D.24380F19@webamused.com>

 Without spending much time looking at your script (to say the least), I
suspect it would be easier and more robust to use pickle to store and
retrieve the data, instead of trying to do it yourself through struct
packing, shelving and the like.

 Joshua


From effbot at telia.com  Tue Apr 18 10:39:14 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Tue, 18 Apr 2000 14:39:14 GMT
Subject: TKinter-destroying a window
References: <8dhk0j$cj5$1@nnrp1.deja.com>
Message-ID: <ms_K4.456$wYl.186843648@newsb.telia.net>

gregholmes at my-deja.com wrote:
> Sorry, I'm reading TFM but haven't figured this
> out yet. If I am done with a secondary (or
> tertiary, or whatever) window (in TKinter), can I
> just destroy the window or do I need to get rid
> of any class instantiation that is
> attached/associated to it? i.e.
>
> ....
> theWindow=Toplevel()
> solveThis=SolveThis(theWindow)
> ....
> theWindow.destroy()
>
> Will the class instantiation and its variables
> and everything still be laying around?

not everything -- destroy releases all the Tk specific
stuff, but the proxy object hangs around until you've
removed all references to it.

(since Python objects don't know who's referring to
them, they cannot destroy all references by them-
selves).

in your case, assuming that "theWindow" is a global
variable, you can use:

    del theWindow

or

    theWindow = None

to get rid of it (assuming SolveThis doesn't create
a circual reference, that is).

if the code is local scope (inside a function or a
method), the variable will be deleted by Python
when you leave that scope.

</F>




From nospam at nowhere.net  Fri Apr 21 00:05:19 2000
From: nospam at nowhere.net (Ray&Maria)
Date: Thu, 20 Apr 2000 22:05:19 -0600
Subject: Tkinter Canvas -- help?
References: <8ddtdm$b2s$1@nnrp1.deja.com> <dRyK4.2542$Za1.37532@newsc.telia.net>
Message-ID: <XrQL4.6240$lw2.184908@news.corecomm.net>

Whatever happen to your excellant (and only) Tkinter manual site Fredrik?
The pages are gone for the last 2 days, and python.org does not respond to
my e-mail except the thank you robot.  It appears that pythonware no longer
holds any content for Tkinter.  Does anyone know if I can download a copy
from a mirror site?

Jueiming Chen
NOSPAMjchen at NOSPAMquark.com


"Fredrik Lundh" <effbot at telia.com> wrote in message
news:dRyK4.2542$Za1.37532 at newsc.telia.net...
> <lbear11 at my-deja.com> wrote:
> > I'm just starting to learn Python/tkinter, and I have found that there
> > is virtually _no_ documentation for the tkinter canvas. Could anybody
> > help? I'm specifically looking for stuff like creating an image by a
> > grid in canvas on "mouseDown", and how to read the items in the canvas,
> > and write them to a text file, in their proper positions. I would
> > _really_ appreciate it if anybody could help me with this.
>
> how to capture mouse down:
>
http://www.pythonware.com/library/tkinter/introduction/events-and-bindings.h
> tm
>
> how to create image items:
> http://www.pythonware.com/library/tkinter/introduction/canvas-image.htm
>
> how to read items in the canvas:
> http://www.pythonware.com/library/tkinter/introduction/canvas.htm
> (look under methods, for find_all, coords, etc)
>
> > P.S. The only tkinter canvas documentation that I have is the
> > "Introduction to Tkinter" from pythonware.com
>
> </F>
>
>




From dgallion1 at yahoo.com  Wed Apr 12 10:46:09 2000
From: dgallion1 at yahoo.com (Darrell Gallion)
Date: Wed, 12 Apr 2000 07:46:09 -0700 (PDT)
Subject: Python idiom: Multiple search-and-replace
Message-ID: <20000412144609.7026.qmail@web2006.mail.yahoo.com>

Been there done that.

My solution is more work than Fredrik's, unless you
plan on making a lot of changes to the buffer.
Like many hundreds of patterns on a large buffer.

Keep a list of changes to be performed.
(start, seqNumber, text, end)

seqNumber stabilizes the sort of this list.

If the start==end then delete a section.

Sort the list and perform all operations at once.
This is like a transaction, which allows for some
other cool features. Such as detecting conflicting or
overlapping changes. 

I have code if you want it.










=====
Darrell Gallion




__________________________________________________
Do You Yahoo!?
Send online invitations with Yahoo! Invites.
http://invites.yahoo.com



From kno at jtan.com  Wed Apr 12 20:02:32 2000
From: kno at jtan.com (Matt Dunford)
Date: Thu, 13 Apr 2000 00:02:32 GMT
Subject: global variables - how??
References: <8d2bne$qn7$1@news101.telia.com>
Message-ID: <FsxHG8.o8q.0.sinope@news.jtan.com>

I usually set a module aside for global variables:

import glob_vars
glob_vars.count = 6

def printme( num ):
  print num

printme( glob_vars.count )

you get the idea...

"Anders Eggers-Krag" <aek at mail1.stofanet.dk> writes:

>how do I define a propper global variable that works in all functions in all
>modules?

>I have tried simply
>defining one outside of any function in the main module,
>I have tried using the global keyword under the same cicumstances,
>but they are not even avialible inside of functions within the main
>module...


>--
>Anders Eggers - Krag


--


From michael.stroeder at inka.de  Thu Apr 27 18:01:47 2000
From: michael.stroeder at inka.de (Michael =?iso-8859-1?Q?Str=F6der?=)
Date: Fri, 28 Apr 2000 00:01:47 +0200
Subject: run detached
Message-ID: <3908B8CB.97A0755D@inka.de>

HI!

How can I automatically detach from console after starting up with
SocketServer? (E.g. web server with access and error logging to
files.)

Ciao, Michael.


From ejbaker_seekonk at my-deja.com  Sun Apr  9 10:57:37 2000
From: ejbaker_seekonk at my-deja.com (ejbaker_seekonk at my-deja.com)
Date: Sun, 09 Apr 2000 14:57:37 GMT
Subject: awk-like usage of Python
Message-ID: <8cq5os$8b9$1@nnrp1.deja.com>

I'm a relative newbie to Python. But, I have extensive experience with
awk (it's an old stand-by that I rely on over and over).

Are there any library modules that make python behave more like awk?
Lacking that, what's the best way to achieve line by line parsing
and pattern matching with Python?

Thanks


Sent via Deja.com http://www.deja.com/
Before you buy.


From pointal at lure.u-psud.fr  Thu Apr 27 11:04:37 2000
From: pointal at lure.u-psud.fr (Laurent POINTAL)
Date: Thu, 27 Apr 2000 15:04:37 GMT
Subject: global variables or inheritence
References: <28p8zy0irgc.fsf@lada.kom.auc.dk>
Message-ID: <39085541.2079513161@news.u-psud.fr>

On Wed, 26 Apr 2000 19:37:40 GMT, Thomas Rasmussen
<simpsons at kom.auc.dk> wrote:

>: Hi
>: 
>: I have quite a big problem. I have two different files, where I in one
>: file want to declare a variable and then in the other file be able to
>: access and use this variable (no need to change it), but I just can't
>: seem to get it to work... Is there any nice way of doing this easily?
>: I have tried the global var but this doesn't seem to work.
>: 
>: Anyone got a good idea?

Be careful that when you write "from module import *", you get
duplication of the imported module public names into your namespace,
with their current referenced objects.

If one module change(*) the object which is referenced in its
namespace, the other module dont see the modification.

Ex.
    #mod1
    var = 12

    #mod2
    from mod1 import *
    # Here you get var with value 12.
    var = 18
    # Here you have var with value 18 and mod1 has var with value 12.


To avoid this problem, you must access globals by their modules names.

Ex.
    #mod2
    import mod1
    mod1.var = 18
    # And its ok.

Here, to force the use of module prefix, we have adopted the rule to
all times use a _ before globals names, so that they MUST be accessed
by their modules names.

A+

Laurent.

(*) Change the object instance!!! If you have a 'shared' list between
modules, you can do operations on it and all modules ill see them...
except the bad module which create a new list instance and assign it
to its namespace'name.

---
Laurent POINTAL - CNRS/LURE - Service Informatique Experiences
Tel/fax: 01 64 46 82 80 / 01 64 46 41 48
email  : pointal at lure.u-psud.fr  ou  lpointal at planete.net 


From python at rose164.wuh.wustl.edu  Tue Apr 25 07:34:27 2000
From: python at rose164.wuh.wustl.edu (David Fisher)
Date: Tue, 25 Apr 2000 06:34:27 -0500
Subject: urllib (sort of reload) question
References: <m3purjdj2b.fsf@localhost.localdomain>
Message-ID: <001801bfaeaa$3727ece0$0201a8c0@spkydomain>

Hi,
I've never actually done this so caveat emptor.  A glance at rfc2616 tells
me you're going to need to add a "Pragma: no-cache" to the outgoing header.
Add the line:

h.putheader("Pramga","no-cache")

somewhere before h.endheaders() in the method open_http() of class
URLopener.  Right next to the GET would probably be good.
Good luck,
David


----- Original Message -----
From: "John Littler" <linuxmusic at crosswinds.net>
Newsgroups: comp.lang.python
To: <python-list at python.org>
Sent: Friday, April 21, 2000 8:24 PM
Subject: urllib (sort of reload) question


>
> Hi,
> I use urllib to get a number of URLs and parse the
> results. One pesky URL doesn't expire it's content
> correctly and as an interim measure I need to reload
> the page in my browser before getting the data with
> python. I'm going through squid so I guess that's where
> the old stuff is being picked up.
> Does anyone know how to emulate browser "reload" in
> python? Looking at the urllib code didn't provide me
> with any clues.
> TIA
> John
>
>
> --
> pgp public key @ http://www.crosswinds.net/~linuxmusic/pubring.html
> --
> http://www.python.org/mailman/listinfo/python-list




From nospam.newton at gmx.li  Sat Apr  1 13:57:47 2000
From: nospam.newton at gmx.li (Philip 'Yes, that's my address' Newton)
Date: Sat, 01 Apr 2000 18:57:47 GMT
Subject: Paid to work in Python? (was: Re: Choice Of Language)
References: <Uqvw4.2261$yL2.28868@typhoon.southeast.rr.com> <8ajmib$elo$1@nnrp1.deja.com> <8c3ats$b5f$1@nntp.Stanford.EDU> <slrn8eact2.sq7.nobody@bambi.visi.com>
Message-ID: <38e5f8ed.29659873@news.nikoma.de>

On Fri, 31 Mar 2000 23:24:18 GMT, nobody at nowhere.nohow (Grant Edwards)
wrote:

>Pointless tanget: Is it just me, or is Google really the last usable search
>engine?

Or possibly the last search engine -- all other previous search engines
seeming to have turned into portals or other advertisement-laden,
all-singing-all-dancing sites. (The German term "eierlegende
Wollmilchsau"[1] comes to mind.)

Cheers,
Philip

[1] literally, "egg-laying wool-milk-sow [female pig]", i. e., something
that tries to do too much at once.
-- 
Philip Newton <nospam.newton at gmx.li>


From mindlace at imeme.net  Wed Apr 26 16:49:12 2000
From: mindlace at imeme.net (mindlace)
Date: Wed, 26 Apr 2000 14:49:12 -0600
Subject: newbie regexp question
Message-ID: <39075648.CA622A7D@imeme.net>

Flying Circus,

I've got a re.compile('\W') that I can use to search for non-alphanumeric characters in a
variable.  However, if the variable has ( in it, for example, the search breaks. is there a
quoted_var = string.quote(somevar) or similar? Do I want to use string.maketrans(somevar,
quoted_var) ?
 
Thanks,
-- 
ethan mindlace fremen        mindlace at imeme.net
zope    -&-     imap email   -&-   mailing list
weave your web with the web at http://imeme.net



From dalke at acm.org  Wed Apr  5 18:11:45 2000
From: dalke at acm.org (Andrew Dalke)
Date: Wed, 5 Apr 2000 16:11:45 -0600
Subject: UserDict's has_key() and get() should use __getitem__()
References: <0535db98.ec878f67@usw-ex0101-005.remarq.com> <m3u2hgn5jg.fsf@atrus.jesus.cam.ac.uk> <29328840.06b5b87b@usw-ex0101-005.remarq.com>
Message-ID: <8cgdp3$nm5$1@slb6.atl.mindspring.net>

Hamish Lawson wrote:
>I grant that my proposal is a bit less efficient, but I see that
>as part of the cost for the proper abstraction. Given that
>UserDict exists primarily to be subclassed, and usually for the
>purpose of providing new definitions of setting and getting
>items, I do think that UserDict should be implemented in such a
>way as not to require those methods which are conceptually
>dependent to be redefined in the derived class.

I don't believe your estimate, that UserDicts are usually derived
"for the purpose of providing new definitions of setting and
getting items", is correct.

Most of the times I've used UserDict/UserList were to add extra
functionality.  For example, a UserDict which preserves the
input order (modifies __setitem__ and keys; __delitem__ and
items/values should be added, but just needed a quick hack).

Or the Dict is supposed to emulate a dictionary-like format,
so I've had __str__ return the properly formatted string.
(Eg, consider a dictionary for HTML-like tags.)

Only rarely have it used it to do normalization - I try to
push that into the I/O routines.

In addition, in a good OO design, you should be able to use
an instance of a derived class anywhere you use an instance
of a base class.  Your example (a caseless Dict), changes
how/when lookup failures occur.  Thus, a "CaselessDict" really
isn't a type of UserDict, you're just using it for some
implementation reuse.

Instead, you want a minimal interface, which puts the least
amount of assumptions on the class as possible.  This corresponds
to something like:

class AbstractUserDict:
  ...
  def __getitem__(self, key):
    raise NotImplementedError
  def __setitem__(self, key, val):
    raise NotImplementedError
  def keys(self):
    raise NotImplemenetedError
  def has_key(self, key):
    try:
      self[key]
      return 1
    except KeyError:
      return 0
  def items(self):
    items = []
    for key in self.keys():
        items.append(self[key])
    return tuple(items)
  ...

Then Python's UserDict is just an implementation of the
AbstractUserDict base class.

This is the sort of design you need in a stricter compile-time
language like Java or C++.  Because this is Python, you don't
need to tell the complier you're implementing an interface -
you just implement it.  And because this is Python, it's easy to
implement.

                    Andrew
                    dalke at acm.org





From USENET at questionexchange.com  Tue Apr  4 18:54:53 2000
From: USENET at questionexchange.com (QuestionExchange)
Date: 4 Apr 2000 22:54:53 GMT
Subject: Does Python have a long floating data type please?
References: <Mh3eOLPZHxSW8Cz6lxvoeAN6vlAH@4ax.com>
Message-ID: <2728qx@questionexchange.com>

http://www.python.org/doc/current/lib/typesnumeric.html
This link descirbes the python data types. It looks like
a float in python will be the same size as a C double. Keep
in mind this is machine dependant.
However, third party modules exist.
http://www.vex.net/parnassus/apyllo.py/684222876.3267957

-- 
  This answer is courtesy of QuestionExchange.com
  http://www.questionexchange.com/servlet3/qx.usenetGuest.showUsenetGuest?ans_id=12552&cus_id=USENET


From echuck at mindspring.com  Sun Apr  9 22:43:05 2000
From: echuck at mindspring.com (Chuck Esterbrook)
Date: Sun, 09 Apr 2000 22:43:05 -0400
Subject: [ANNOUNCE] Garbage collection for Python
References: <20000407221750.A29279@acs.ucalgary.ca>
Message-ID: <38F13FB9.D1D16AF@mindspring.com>

Neil Schemenauer wrote:

> Garbage collection for Python:
>
>     http://www.enme.ucalgary.ca/~nascheme/python/gc.html
>
> I think I have all of the wrinkles ironed out of this patch.
> Reference cycles involving lists, tuples, instances, classes,
> dictionaries, and functions are found.  Instances with __del__
> methods are handled in a sane way.  It is easy to add GC support
> to new types.  GC enabled Python is binary compatible with
> regular Python.
>
> Generational collection works (currently three generations).  The
> overhead measured by pybench is less than 1 percent and is not
> measurable with pystone (on my machine).  Virually all extension
> modules should work unchanged (I had to modify new and cPickle in
> the standard distribution).  A new module called gc is available
> for tuning the collector and setting debugging options.
>
> The collector should be portable across all platforms but I have
> only tested on my Linux machine.  The patched version of Python
> passes all regression tests and runs Grail, Idle and Sketch
> without problems.
>
> The patch is against the current CVS version of Python.  If
> enough people ask for a patch against the stable version of
> Python I will make one.  Please send comments and bug reports.
>
>     Neil
>
> --
> "Simplicity does not precede complexity, but follows it." -- Alan Perlis

Looking at Python FAQ 6.14 on GC, I see another one as well:

http://starship.python.net/crew/gandalf/gc-ss.html


Can you compare/contrast the two?  Also, you might want to get yours mentioned in the FAQ.


-Chuck



From scarblac-spamtrap at pino.selwerd.nl  Fri Apr 21 14:06:22 2000
From: scarblac-spamtrap at pino.selwerd.nl (Remco Gerlich)
Date: 21 Apr 2000 18:06:22 GMT
Subject: List comprehensions?
Message-ID: <slrn8g19cn.mpr.scarblac-spamtrap@flits104-37.flits.rug.nl>

I thought list comprehensions had been in CVS for quite a while, so I
expected them to be in 1.6. But I can't find anything about them, and the
syntax as I find it on Deja doesn't work.

What's the deal? Features that are in CVS but don't make it into the next
release?

-- 
Remco Gerlich,  scarblac at pino.selwerd.nl
-rwxr-xr-x  1 root  5.89824e37 Oct 22  1990 /usr/bin/emacs [STR]


From ngps at madcap.dyndns.org  Mon Apr 17 10:55:03 2000
From: ngps at madcap.dyndns.org (Ng Pheng Siong)
Date: 17 Apr 2000 14:55:03 GMT
Subject: Filtering web proxy
References: <Pine.LNX.4.21.0004171406250.30999-100000@fep132.fep.ru>
Message-ID: <8df8k7$rqp$1@coco.singnet.com.sg>

According to Oleg Broytmann  <phd2 at mail.com>:
> Once I tried Junkbuster, but found it inadequate. My need is simple -
> just a short list of "white" sites. Never tried adzapper...

I use Alfajor and am happy with it. I only need cookie filtering,
which Alfajor does very well.

Cheers.

-- 
Ng Pheng Siong <ngps at post1.com> * http://www.post1.com/home/ngps



From phd at phd.russ.ru  Mon Apr 24 07:28:21 2000
From: phd at phd.russ.ru (Oleg Broytmann)
Date: Mon, 24 Apr 2000 11:28:21 +0000 (GMT)
Subject: Trouble compiling PyGreSQL 2.4 on Linux
In-Reply-To: <k8hoeidh.fsf@jumbo.skott.dk>
Message-ID: <Pine.LNX.4.21.0004241126280.6019-100000@fep132.fep.ru>

On Sun, 23 Apr 2000, Martin Sk?tt wrote:
> I'm trying to compile the PyGreSQL 2.4 module on my Red Hat Linux 6.1
> box. I use the following command:
> 
> gcc -fpic -shared -o _pg.so -I/usr/include/python1.5
> -I/usr/local/pgsql/include pgmodule.c -lpq
> 
> With the following result:
> pgmodule.c: In function `pgconnect':
> pgmodule.c:694: warning: passing arg 4 of `PyArg_ParseTupleAndKeywords' from incompatible pointer type
> /usr/bin/ld: cannot open -lpq: No such file or directory

   You forget to mention path to postgres libs (-L /usr/local/pgsql/lib).

> I can't use the make command mentioned in the readme.linux file
> because the simply isn't any make file in the archive. Can you help me
> getting it compiled?

   There is no Makefile 'cause the module is intended to use standard
Python Makefile.pre.in/Setup.in scheme.

> And a little addon queston:
> How stable is the DB-API version of the PostgreSQL module? Can it
> be used for production work (or semi serious)?

   It was relased few days ago, so it just hadn't been tested...

Oleg.            (All opinions are mine and not of my employer)
---- 
    Oleg Broytmann      Foundation for Effective Policies      phd at phd.russ.ru
           Programmers don't die, they just GOSUB without RETURN.




From echuck at mindspring.com  Fri Apr 14 20:09:52 2000
From: echuck at mindspring.com (Chuck Esterbrook)
Date: Fri, 14 Apr 2000 20:09:52 -0400
Subject: else clauses in while and for loops
References: <200004142351.JAA02824@nexus.csse.monash.edu.au>
Message-ID: <38F7B350.5D6548E1@mindspring.com>

Jonathan Giddy wrote:

> Chuck Esterbrook declared:
> >
> >So did anyone present, in this thread, an example that shows a real need?
>
> The difficulty is what will you accept as a "real need"?  It sounds like you
> want an example that cannot be done any other way.  Even those who like else
> clauses in for and while will admit they're not essential.

I don't know why you think it 'sounds like' that. Most of the posts didn't have any examples and the one that did wasn't all that cogent. I'm just looking for something that is [a] an example and [b] shows some obvious advantage.


> Just like there's no "real need" for tuples (use lists), for loops (use
> while), or elif (use else: if).
>
> All of these concepts are redundant in the sense that they can be mimicked
> using an alternative sequence of code.  But they're nice because they
> express higher-level concepts about the algorithm.

I'm aware of all of this. In fact, I think we could all get our work done with ADD, NEG, JMP and JMPZ.   :-)

But no thank you.



> OTOH, a quick look at my code shows that I regularly use tuples, for
> loops, and elif, but I rarely use else with anything but if and
> try..except clauses.
>
> The only example I found creates a list of usable temporary directories on
> Unix, checking the environment first.  Certainly no real need here.
>
> _tmpdirs = ['/tmp', '/var/tmp']
>
> for envname in ['TMPDIR', 'TMP', 'TEMP']:
>     try:
>         paths = [os.environ[envname]]
>     except KeyError:
>         pass
>     else:
>         basepaths = validate_paths(paths)
>         if basepaths:
>             break
> else:
>     basepaths = validate_paths(_tmpdirs)

This example looks good to me. It eliminates a redundant "if" after the "for" loop to see if "basepaths" still needs to be set.

Thanks.


On another topic: With regards to the "try...except" technique above (which I've seen quite a bit in Python) does the following code do the same thing and if so, which do you find more readable?

_tmpdirs = ['/tmp', '/var/tmp']

for envname in ['TMPDIR', 'TMP', 'TEMP']:
    if os.environ.has_key(envname):
        basepaths = validate_paths(path)
        if basepaths:
            break
else:
    basepaths = validate_paths(_tmpdirs)


-Chuck




From tim_one at email.msn.com  Wed Apr 26 23:12:14 2000
From: tim_one at email.msn.com (Tim Peters)
Date: Wed, 26 Apr 2000 23:12:14 -0400
Subject: arithmetic in longobject.c
In-Reply-To: <39071752.51DB78B9@stud.ntnu.no>
Message-ID: <000401bfaff6$632a5ec0$272d153f@tim>

[posted & mailed]

[Peter Schneider-Kamp]
> I got a few questions about longobject.c which a quick
> glance at the source could not answer.
>
> What algorithms does longobject.c implement for
> multiplication and division of long integers?

Knuth's.

> What representation is used for long integers?

See the comment block near the top of Include/longintrepr.h (signed
magnitude, base 2**15, using only 15 of each 16 bits -- or however long a C
short happens to be on your platform).

> I'd like to add a rns representation option in python
> for speedups in multiplication.  Internal knowledge
> of the long integer arithmetic would be helpful
> to make the right design choices.

Unless you're just having fun, read about the mpz module in the Library
Reference Manual (which allows use of GMP from within Python).  "Optimized"
bigint arithmetic is unlikely to get into the Python core.

too-little-demand-for-so-much-hairy-code-ly y'rs  - tim





From hgoldste at mpcs.com  Sun Apr  9 08:53:27 2000
From: hgoldste at mpcs.com (Howard Goldstein)
Date: Sun, 09 Apr 2000 12:53:27 GMT
Subject: Python Language Support for ArcView GIS
References: <q0MH4.2398$Bq2.38636@newscontent-01.sprint.ca>
Message-ID: <8914.955284807.daemon@warez.bofh.sebastian.fl.us>

On Sat, 8 Apr 2000 16:43:23 -0300, Bruce Dodson <bruce_dodson.nospam at bigfoot.com> wrote:
 : Python Language Support for ArcView GIS (aka AVPython) is now available
 : online as a free download.  I promised this several months ago but never
 : managed to find a home for it (too busy) so finally I decided to just slap
 : together a page for it on GeoCities:
 : www.geocities.com/brucedodson.rm/avpython.htm

Sweet!  Thank you for making this available to the GIS and python
community


From shogan at iel.ie  Thu Apr 20 09:20:56 2000
From: shogan at iel.ie (Shaun)
Date: Thu, 20 Apr 2000 14:20:56 +0100
Subject: how do i pack and unpack messages in python?
Message-ID: <008301bfaacb$42fe2c60$2801000a@iel.ie>

im still new to python and i have this problem to solve and i have not a
clue where to start can anyone please put me on the right track...

say i have a form, with the values:

Name: Shaun Hogan
Address: Anytown, Anywhere
Credit Card Number: 1234567890

so i have 3 values that i want to pack into a message so as the info can be
sent (but dont worry about th3 sending bit for the mo)

i have a message that looks like this when broken down:
(makes more sense if you read from the bottom up!)

{
Message:
    (type)int(value)id            // the message has its id
    (type)int(value)subid       //and subid
     NVsequence                 //and list of NV pairs
NVsequence:
     NV pair []                   //NVsequence consists of an array/list of
NV
                                         // pairs
NV pair:
     int name                       //we assign id names to the values
     value v                          //eg. [14:"shaun"], [20:"Any..."],
[34:1234...]
                                         //creating [name:value] pairs
Value:
     char type                    //here we determine and state the type of
the
     void *pVal                 //values, "shaun"=string,
"Anyt...,Anyw.."=list,
                                       //1234567890=int, and we establish a
*pointer
                                       //to the values.

Value List:                       //here we have a list of values
     Value[]
                //"shaun","Anytown,Anywhere",1234567890]       }

so i now have a message that looks like:

id                       : 10
subid                  :13
NV sequence :   {14, "shaun"-string
                         {20, ["Anytown, Anywhere"]-list
                         {34, 1234567890-int


i want to pack the message to look like this:

10 fs 13 fs 14 fs s shaun fs 20 fs l s anytown fs s anywhere fs i 1234567890
gs gs

(NV=Name Value Pair, fs=field seperator, s=string, i=int, l=list, group
seperator)

can anyone please tell me how id go about manipulating the strings so as i
can join them using field seperators, any ideas anyone, or even tell me
where i might find out how to do this problem.

Thanks a Million
Shaun

=====================
Shaun Hogan
Interactive Enterprise Ltd.
alt. E-mail : shaun_hogan at yahoo.com
+353 86 8342529





From tseaver at starbase.neosoft.com  Sat Apr 22 12:50:56 2000
From: tseaver at starbase.neosoft.com (Tres Seaver)
Date: 22 Apr 2000 11:50:56 -0500
Subject: The Simple Economics of Open Source
References: <14591.15630.192472.440375@goon.cnri.reston.va.us> <3900611F.3B525BDA@cfdrc.com> <FtDG4z.774@world.std.com> <slrn8g26tu.p7o.neelk@brick.cswv.com>
Message-ID: <14837CB4E98A72FE.8FD547BA6E928594.5A103CFE68327EB7@lp.airnews.net>

In article <slrn8g26tu.p7o.neelk at brick.cswv.com>,
Neel Krishnaswami <neelk at alum.mit.edu> wrote:
>
>One of the interesting things about this thread is that it has given
>me a bit of insight into what economics looks like to people who
>haven't internalized its assumptions. I'm afraid that I'm tainted
>enough that it no longer looks like the dismal science to me. :)
>
>Seriously, the rejection of altruism as a possible motive strikes me
>as a very reasonable decision. Altruism is generally not sufficient
>motive in other aspects of life -- and since there's no particular
>reason that programmers in particular are more likely to be altruistic
>than the rest of the population, positing that hackers have greater
>virtue is at best a dubious proposition. Consider, for example, how
>much better government would be if all the voters studied the issues
>and took care to be informed voters, and that basically no one makes
>the effort. Most people don't even bother digging up evidence to
>support their prejudices, let alone try to come to a careful decision.
>And searching the Census and BLS websites is *easier* than hacking on
>the Python core.

"That dog won't hunt!"  Your claim that "basically no one makes the
effort" is easily refuted:  consider the League of Women Voters,
and the guides they put out before each election -- that effort is
well-organized, and involves coordinated, non-partisan effort on the
scale of tens of thousands of volunteers (much like a large Open Source
project).  The League's explicit motive is to educate themselves and
the population at large on the issues and candidates;  that at least
some of the effort here is altruistic is obvious to anyone who hasn't
"internalized [economics'] assumptions."

Altruism-is-a-shibboleth'ly,

Tres.
-- 
---------------------------------------------------------------
Tres Seaver        tseaver at digicool.com     http://www.zope.org
Digital Creations


From breiter at usf.Uni-Osnabrueck.DE  Mon Apr  3 15:43:32 2000
From: breiter at usf.Uni-Osnabrueck.DE (Bernhard Reiter)
Date: 3 Apr 2000 19:43:32 GMT
Subject: Python 1.6 alpha 1 released
References: <200004010159.LAA00403@piglet.dstc.edu.au>
    <8c9nq3$ghk$4@newsserver.rrzn.uni-hannover.de>
    <m3bt3rcweu.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <8cas94$po5$1@newsserver.rrzn.uni-hannover.de>

In article <m3bt3rcweu.fsf at atrus.jesus.cam.ac.uk>,
	Michael Hudson <mwh21 at cam.ac.uk> writes:
> breiter at usf.Uni-Osnabrueck.DE (Bernhard Reiter) writes:
> 
>> In article <200004010159.LAA00403 at piglet.dstc.edu.au>,
>> 	David Arnold <arnold at dstc.edu.au> writes:
>> > -->"Bjorn" == Bjorn Pettersen <bjorn at roguewave.com> writes:
>> > 
>> >   Bjorn> The more time I spend on fixing
>> >   Bjorn> compatibility issues in Python scripts, the less time I spend
>> >   Bjorn> on "real" work.
>> > 
>> > you have been using incorrect arguments.  the fact that they worked
>> > was unfortunate because it leads to circumstances like this.  but the
>> > time you have spent on these emails would have been enough to run a
>> > simple script over your code to fix the problem ... ;-)
>> 
>> Hmm how simple would a script be to get all occurences of:
>> 
>> 	socket.connect(what, \ 
>> 			else \
>> 			)
>> 
>> 	sc=socket.connect
>> 
>> 	sc(what,else)
>> 	sc(what, \
>> 		else)
>> 
> 
> Yes, but `grep -w connect' and the minimal application of a human
> eyeball is surely enough to get 90%+ of the cases sorted out.  I can't
> imagine sock.connect is called from all that many places in one's code
> (but I might be wrong).  And, to answer you're question, about as
> complicated as:
> 
> http://x42.deja.com/[ST_rn=ps]/getdoc.xp?AN=591243549
> 
> after `sed -e s/append/connect/', which doesn't get the bound method
> case, but I am having trouble imagining circumstances in which you'd
> reasonably do that.

True the whole discussion is kind of pointless.
Nontheless it is not an easy taks to just rewrite your code with
a simple regular expression replacement. ;-) (And that was my point.)

	Bernhard

-- 
Professional Service around Free Software                (intevation.net)  
The FreeGIS Project				            (freegis.org)
Association for a Free Informational Infrastructure            (ffii.org)


From nick at spam_me_notvideosystem.co.uk  Thu Apr 13 11:20:24 2000
From: nick at spam_me_notvideosystem.co.uk (Nick Trout)
Date: Thu, 13 Apr 2000 16:20:24 +0100
Subject: List mapping?
References: <38f5d68e@news.xtml.co.uk> <newscache$u6nysf$1l1$1@jenpc07.jennic.co.uk>
Message-ID: <38f5e714@news.xtml.co.uk>

> Try this:
> mylist = ['/a','1','/b','2','/c','3']
> map(lambda x: reduce(lambda a, b: (a,b), mylist[x:x+2]), range(0,
> len(mylist), 2))

Very good! I wasnt familiar with the reduce function. I would certainly not have
come up with something of this complexity.

This begs the question: Is there an archive of code snippets for Python? These
could form an excellent tutorial. Quite a few messages arrive on this newsgroup
asking for good starting points for Python (and thats just the people who come
here and ask).

N




From mdefreitas at sikorsky.com  Thu Apr 27 17:03:37 2000
From: mdefreitas at sikorsky.com (mdefreitas at sikorsky.com)
Date: Thu, 27 Apr 2000 21:03:37 GMT
Subject: MORE INFO
References: <8ea0ao$fro$1@nnrp1.deja.com>
Message-ID: <8ea9uu$rh0$1@nnrp1.deja.com>

By the way... I finally tested the code I wrote and it evidently is
incorrect. It works fine on the first nest level, but my application
aborts on the second nest level.

The error message is:
Fatal Python Error: PyThreadState_Get: no current thread

OK... I'm lost. Any ideas? Anybody? Buelher?


Sent via Deja.com http://www.deja.com/
Before you buy.


From mjhand at concentric.net  Tue Apr 18 10:42:49 2000
From: mjhand at concentric.net (Manus Hand)
Date: Tue, 18 Apr 2000 08:42:49 -0600
Subject: Upgrading IIS to 1.6 problem....
References: <2B1262E83448D211AE4B00A0C9D61B03BA72C9@MSGEURO1>
Message-ID: <38FC7469.79082530@concentric.net>

Thanks, Pieter!

That was indeed the problem.  My IIS Mgr settings still had the 1.5.2
installation listed.  (Makes me wonder how important the registry
setting is in NT.)

Manus




From donn at u.washington.edu  Thu Apr 27 12:05:46 2000
From: donn at u.washington.edu (Donn Cave)
Date: 27 Apr 2000 16:05:46 GMT
Subject: IP Address
References: <014801bfb030$141df690$0500a8c0@secret.pythonware.com> <WZTN4.135$E44.2126@news6-win.server.ntlworld.com>
Message-ID: <8e9ogq$hqi$1@nntp6.u.washington.edu>

Quoth "Fredrik Lundh" <fredrik at pythonware.com>:
| Richard Chamberlain wrote:
| > Is there a way of returning the IP address of the machine that a =
| script is
| > running on?
|
| here's one way to do it:
| ip = socket.gethostbyname(socket.gethostname())

Took the words out of my mouth.  But I would add, while this is
the right thing for general use, there are situations where it's
not reliable, where it might return an IP address that is in some
sense valid but not useful.  Consider this as an alternative for
those situations:

s = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
s.connect(('123.345.567.789', 53))
ip, localport = s.getsockname()
s.close()

This establishes an IP connection to a remote host.  It happens to
be to the DNS service port -- thinking this remote host would be the
primary DNS for your host, chosen because it's one of those network
identities without which little else has meaning.  But any host and
service will do, the point is that you know it will be ``out there''
and accepting connections.  The getsockname() on this connection
tells you what your IP address is in the context of ``out there'',
as opposed to one of perhaps several IP addresses that are or were
valid in some back door network shown in /etc/hosts.  It also will
get your network up first, if you're in an ``on demand'' situation
where the address is dynamically assigned.

	Donn Cave, University Computing Services, University of Washington
	donn at u.washington.edu


From herzog at online.de  Wed Apr  5 18:49:16 2000
From: herzog at online.de (Bernhard Herzog)
Date: 06 Apr 2000 00:49:16 +0200
Subject: Tkinter and C extensions
References: <Pine.OS2.3.95.1000405222214.1184B-100000@CENTRAL>
Message-ID: <m3pus4w4cz.fsf@greebo.nodomain.de>

Andrew MacIntyre <andymac at bullseye.apana.org.au> writes:

[...]
> My research has turned up ways of determining these via Tk's API
> (Tk_Display(), Tk_Window() in tk.h) provided I can extract the tkwin
> handle from Tkinter's canvas widget - something I've yet to figure out how
> to do from within a C extension.
> 
> I'm wondering whether its possible to get at this information via Tkinter,
> which would simplify the interactions between the various components.

Tkinter widgets have a tk attribute which is basically a tcl
interpreter. It's interpaddr() method returns the TclInterp* pointer as
a python int object. Just pass that and the name of the widget to a
C-extension function and the rest follows from that...

> Any hints on additional sources of enlightment, such as other modules
> which have similarities, would be appreciated.

Sketch's paxmodule does something similar. Sketch uses it to implement
widgets in Python, which allow you to handle expose events yourself and
use Xlib functions to draw the widgets.

-- 
Bernhard Herzog   | Sketch, a drawing program for Unix
herzog at online.de  | http://sketch.sourceforge.net/


From just at letterror.com  Tue Apr  4 19:23:01 2000
From: just at letterror.com (Just van Rossum)
Date: Wed, 5 Apr 2000 00:23:01 +0100
Subject: ANN: Stackless Python for MacOS
Message-ID: <l03102815b5101e19e027@[193.78.237.148]>

I Proudly Present a Not-For-Everybody(tm) Release:

   S T A C K L E S S   P Y T H O N   F O R   M A C O S

I've uploaded a "drop-in" replacement for the PythonCore 1.5.2 shared
library here:
  <http://starship.python.net/~just/PythonCoreStackless.sit.hqx>

It should be fully compatible with the latest MacPython release.


What it is?

Well, it's a modified version of Python which avoids the C stack, written
by Christian Tismer.


But, what *is* it?

Erm, for one, see <http://www.stackless.com/> and then, for it's killer app
-- Micro Threads -- see <http://world.std.com/~wware/uthread.html>.
Stackless Python adds cool, but mind boggling features to Python:
continuations & co-routines.

I'm playing with MicroThread additions to the MacPython IDE, using separate
threads for scripts. This looks *very* promising. If I manage to make
something stable-ish, I'll post IDE patches to the MacPython sig.

General Stackless talk should probably take place on comp.lang.python, Mac
specific issues (as IDE support) should be discussed on the MacPython sig.

Even though I still hardly understand the implications of it all yet,
Stackless is IHMO the most exciting new Python development at the time,
together with Unicode support in 1.6. Stackless will most likely not be
incorporated into Python 1.6, but *maybe* if we make enough noise we can
convince Guido it should go into 1.7... More stackless killer apps please!


Have fun,

Just





From shogan at iel.ie  Fri Apr 28 12:10:27 2000
From: shogan at iel.ie (Shaun)
Date: Fri, 28 Apr 2000 17:10:27 +0100
Subject: struct, xdrlib and array examples?
Message-ID: <00ca01bfb12c$448e0a20$2801000a@iel.ie>

where would i find examples of the struct, xdrlib and array modules in
action?

=====================
Shaun Hogan
Interactive Enterprise Ltd.
alt. E-mail : shaun_hogan at yahoo.com
+353 86 8342529






From hamish_lawson at yahoo.co.uk  Fri Apr  7 05:13:20 2000
From: hamish_lawson at yahoo.co.uk (Hamish Lawson)
Date: Fri, 07 Apr 2000 02:13:20 -0700
Subject: Cacheing sys.path rather than building it each time?
Message-ID: <2e070214.0e300ac8@usw-ex0108-063.remarq.com>

Some overhead in Python's startup time is caused by building
sys.path. Of course you could use the -S option, but then you
don't get a proper sys.path! What about cacheing the constructed
path and a command-line option to specify use of that rather
than building it each time? Presumably you'd also need a tool to
rebuild the path when necessary - perhaps it could be invoked
automatically by the distribution tools whenever a new module
gets installed.

Hamish Lawson


* Sent from RemarQ http://www.remarq.com The Internet's Discussion Network *
The fastest and easiest way to search and participate in Usenet - Free!



From thomas at xs4all.net  Fri Apr 21 06:48:17 2000
From: thomas at xs4all.net (Thomas Wouters)
Date: Fri, 21 Apr 2000 12:48:17 +0200
Subject: finding real $0
In-Reply-To: <OWeCyjAOWCA5Ew86@jessikat.demon.co.uk>; from robin@jessikat.demon.co.uk on Fri, Apr 21, 2000 at 10:55:26AM +0100
References: <OWeCyjAOWCA5Ew86@jessikat.demon.co.uk>
Message-ID: <20000421124817.C11481@xs4all.nl>

On Fri, Apr 21, 2000 at 10:55:26AM +0100, Robin Becker wrote:

> Is there any pythonic way to locate the original argument zero of the
> process? I mean the location of python itself not the script. This would
> allow me to launch subscripts without having to rely on the path being
> set up correctly.

Use 'sys.executable':

centurion:~ > python
>>> import sys 
>>> sys.executable
'/usr/bin/python'

centurion:~ > ln -s /usr/bin/python penguin-on-the-telly
centurion:~ > ./penguin-on-the-telly 
>>> import sys
>>> sys.executable
'./penguin-on-the-telly'

It behaves the same way in a script (it returns the python executable
started, not the script.) Dont forget about relative paths though, if you
are going to do stuff based on sys.executable. You probably want to get the
cwd at the earliest possible moment, and translate sys.executable into an
absolute path.

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From tjreedy at udel.edu  Thu Apr 27 23:25:57 2000
From: tjreedy at udel.edu (Terry Reedy)
Date: Thu, 27 Apr 2000 23:25:57 -0400
Subject: How can I get the name of an object???
References: <8e712k$65c$1@nnrp1.deja.com> <etd8zy03m44.fsf@w20-575-108.mit.edu> <8e9090$c4d$1@nnrp1.deja.com>
Message-ID: <8eb01p$kmk$1@news.udel.edu>

As noted already, PyObjects do not have *a* name.
 They do have *an* id (a unique integer),
which is simply retrieved with the id() function.




From glen at electricorb.com  Thu Apr 20 08:03:28 2000
From: glen at electricorb.com (Glen Starchman)
Date: Thu, 20 Apr 2000 12:03:28 GMT
Subject: Java editor for Python
References: <8atgsd$ava$1@vg170.it.volvo.se> <pXrA4.24285$W5.355862@typhoon.nyroc.rr.com>
Message-ID: <38FF1CC5.4A0F7DAF@electricorb.com>

> Be there is a colonizing editor that comes with Python for Python.
> Checkout Idle in the tools directory or Pythonwin.
> 
OUCH! A colonizing editor sounds very painful. Could one consider vi a
colonizing editor?

;-)


> --Darrell


From gberger at channel1.com  Mon Apr 10 21:26:19 2000
From: gberger at channel1.com (Gabriel Berger)
Date: Mon, 10 Apr 2000 21:26:19 -0400
Subject: downloading jpeg using python--extra bytes?
Message-ID: <sf5001f0cqv98@corp.supernews.com>

Hello,

I wrote a python script to download jpegs from web sites.
(I realize that it's possible to do this using my browser, but
as an exercise, I'd like to be able to do it directly.)

The script works, with one exception: the jpeg image has
extra bytes in it that corrupt the file.  I'm not sure exactly
how the bytes are dispersed, but they're not all at the
beginning or at the end.

I would imagine that I need to modify my http request, but
I don't know how to do so.  Any help would be appreciated.

Here is the script:

import httplib
h = httplib.HTTP('www.something.com')
h.putrequest('GET','/directory/picture.jpg')
h.putheader('Accept','image/jpeg')
h.endheaders()
errcode,errmsg,headers = h.getreply()
f=h.getfile()
data=f.read()
f.close()
g = open("myjpeg.jpg","w")
g.write(data)
g.close()

Thanks in advance,

Gabriel Berger




From tiskander at hologic.com  Mon Apr 24 11:28:29 2000
From: tiskander at hologic.com (Tim Iskander x7483)
Date: Mon, 24 Apr 2000 11:28:29 -0400
Subject: questions about imports and classes
Message-ID: <3904681D.FE39B41A@hologic.com>

Hi,
as a relatively new python user I have a couple of questions...

1 - is it possisble to get the instance name of an object (similar to
the class.__name__
        attribute)

2 - has there been any/much discussion on extending the import command
to allow
        renaming the module at import time,
        ie: import myLongWindedModuleName as myShortName

        using the (new) keyword "as" to rename the module. We are
dealing with a multi
       developer, multitask application using lots of python modules and
we would like to
        give the files non-ambiguous names without having unruly module
names in our code.


Thanks
/Tim

p.s. I am not very vi - literate and cannot get it to use spaces instead
of tabs for autoindent... I will go
and read the FAQ etc... but if its a simple answer that someone can send
me I'd be really happy :-)
-------------- next part --------------
A non-text attachment was scrubbed...
Name: tiskander.vcf
Type: text/x-vcard
Size: 335 bytes
Desc: Card for Tim Iskander x7483
URL: <http://mail.python.org/pipermail/python-list/attachments/20000424/de188073/attachment.vcf>

From brent.fulgham at xpsystems.com  Wed Apr 19 18:35:54 2000
From: brent.fulgham at xpsystems.com (Brent Fulgham)
Date: Wed, 19 Apr 2000 15:35:54 -0700
Subject: Dynamic web pages! not Dynamic HTML 
Message-ID: <EDFD2A95EE7DD31187350090279C67677548C5@THRESHER>

> [Sunil Hadap]
> > I am really dumb in databases and perl. I am good at python and it's
> > my scripting language for whatever I do in graphics. I want to start
> > doing my homepage which is inspired from
> > 
> > http://www.photo.net/photo/
> > 
> [snip] 
> > 
> > I want to do this in Python. Is zope useful for this, I 
> > don't want to learn a big system as I only want few specific 
> > functions only. Which free database?
> 

Sunil,

The photo.net site is based on a product called the ACS.  It's based on
the open-source AOLserver running in conjunction with an Oracle RDBMS.
It runs on "big iron" and is in nearly all respects beyond the resources
of the average joe for a "live" site because Oracle licenses are so
costly.  The ACS software, however, is GPL'd and can be downloaded from
http://arsdigita.com.

For us mortals, a port of the ACS to run on the PostgreSQL database has 
been accomplished.  See http://acspg.benadida.com.  I have some preliminary
debian packages to set up a skeleton system, which I can make available
if anyone cares.

The only problem with the ACS-pg is that you still need to do your
scripting in Tcl.

So, Titus Brown and I have successfully embedded Python in the AOLserver.
See our project at http://pywx.sourceforge.net.  It is now actually
semi-stable for simple CGI-style operations, and provides a 4-5x speedup
over the straight CGI.

Titus has gotten most of the database-related Tcl functions from the
AOLserver
Tcl API ported to Python.  I'm working on a few others.  Eventually we'll
get them all, but for now it's still a bit limited.

Our near-term goal is to provide access to the ACS and AOLserver through
Python so extensions can be made to a "skeleton" ACS without having to
expend a lot of effort learning Tcl.

Thanks,

-Brent



From mjhand at concentric.net  Fri Apr 21 11:54:12 2000
From: mjhand at concentric.net (Manus Hand)
Date: Fri, 21 Apr 2000 09:54:12 -0600
Subject: Starship update: April 21
References: <3daeirht58.fsf@amarok.cnri.reston.va.us>
Message-ID: <390079A4.73E66400@concentric.net>

Given the update you just put out (that there is only a 25% chance
of recovery!  Aauugghh!!) do you think a proper course of
action is to restore from the 1 Feb backup to the new disk, and
then later -- if the 25% chance comes through -- offer the shipmates
a chance to do a massive diff or something on their files compared
to those that are recovered?

That way at least the ship would start moving again.

Just an idea, and please do not hesitate to call on me for funds
if and when they are needed.

Manus (virtually in tears now, but 1 Feb is a lot better than nothing)




From huy at nsw.bigpond.net.au  Tue Apr 25 08:13:22 2000
From: huy at nsw.bigpond.net.au (Bigpond)
Date: Tue, 25 Apr 2000 12:13:22 GMT
Subject: downloading jpeg using python--extra bytes?
References: <sf5001f0cqv98@corp.supernews.com>
Message-ID: <CZfN4.3758$v85.24236@news-server.bigpond.net.au>

Try using the urllib module,

it handles all this low level stuff for you.

one line can do all that you have specified below, and it works...


Gabriel Berger <gberger at channel1.com> wrote in message
news:sf5001f0cqv98 at corp.supernews.com...
> Hello,
>
> I wrote a python script to download jpegs from web sites.
> (I realize that it's possible to do this using my browser, but
> as an exercise, I'd like to be able to do it directly.)
>
> The script works, with one exception: the jpeg image has
> extra bytes in it that corrupt the file.  I'm not sure exactly
> how the bytes are dispersed, but they're not all at the
> beginning or at the end.
>
> I would imagine that I need to modify my http request, but
> I don't know how to do so.  Any help would be appreciated.
>
> Here is the script:
>
> import httplib
> h = httplib.HTTP('www.something.com')
> h.putrequest('GET','/directory/picture.jpg')
> h.putheader('Accept','image/jpeg')
> h.endheaders()
> errcode,errmsg,headers = h.getreply()
> f=h.getfile()
> data=f.read()
> f.close()
> g = open("myjpeg.jpg","w")
> g.write(data)
> g.close()
>
> Thanks in advance,
>
> Gabriel Berger
>
>




From mwh21 at cam.ac.uk  Thu Apr 13 05:37:59 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 13 Apr 2000 10:37:59 +0100
Subject: [ANNOUNCE] Garbage collection for Python
References: <20000407221750.A29279@acs.ucalgary.ca>
Message-ID: <m3r9cagx3c.fsf@atrus.jesus.cam.ac.uk>

"Neil Schemenauer" <nascheme at enme.ucalgary.ca> writes:

> Garbage collection for Python:
> 
>     http://www.enme.ucalgary.ca/~nascheme/python/gc.html
> 
> I think I have all of the wrinkles ironed out of this patch.
> Reference cycles involving lists, tuples, instances, classes,
> dictionaries, and functions are found.  Instances with __del__
> methods are handled in a sane way.  It is easy to add GC support
> to new types.  GC enabled Python is binary compatible with
> regular Python.  

Woohoo!

> Generational collection works (currently three generations).  The
> overhead measured by pybench is less than 1 percent and is not
> measurable with pystone (on my machine).  Virually all extension
> modules should work unchanged (I had to modify new and cPickle in
> the standard distribution).  A new module called gc is available
> for tuning the collector and setting debugging options.

And in some cases it can be faster; eg. bytecodehacks does a lot of
scary hackery to avoid producing cycles, and if I take the hackery out
and run it under a patched python, it goes quicker (test case in 7.2s
vs 12.5s, so not to be sniffed at).

> The patch is against the current CVS version of Python.  If
> enough people ask for a patch against the stable version of
> Python I will make one.  Please send comments and bug reports.

Haven't found a problem yet.


To someone who might know: what hope is there for this to go into 1.6?


It would be, in my opinion, be a really really really good thing; for
one thing it would help in language wars...

hopeful-ly y'rs
Michael

-- 
  it's not that perl programmers are idiots, it's that the language
  rewards idiotic behavior in a  way that no other language or tool 
  has ever done                        -- Erik Naggum, comp.lang.lisp


From ivanlan at callware.com  Wed Apr 12 15:40:13 2000
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Wed, 12 Apr 2000 13:40:13 -0600
Subject: Off-topic
References: <38F4BE12.A0C73D1D@callware.com> <m3zoqzgois.fsf@atrus.jesus.cam.ac.uk> <38F4C52C.450F611F@callware.com> <38F4C8D6.1AF36727@rubic.com>
Message-ID: <38F4D11D.569AC449@callware.com>

Hi Jeff, All--

Jeff Bauer wrote:
> 
> Ivan Van Laningham wrote:
> > Sh scripts do work, as long as I remember to do ./script ;-).  I'm
> > running as root (but the same thing happens when I run as myself, too).
> 
> Okay, my last post to you went out before I saw the above
> on c.l.py.
> 
> Is there an issue with your PATH environ?
> 
> $ hello.py
> bash: hello.py: command not found
> 
> $ echo $PATH
> /usr/local/bin:/bin:/usr/bin:/usr/X11R6/bin:/home/jbauer/bin:\
> /usr/local/mysql/bin:/usr/local/jdk/bin:/var/share/s0/bin
> 
> $ export PATH=.:$PATH
> $ hello.py
> Hello, nerd.
> 
> Note that it's probably dangerous to use '.' in your root path.
> 

Yup, of course.

The problem turned out to be two things:  the "h2n" command had not
after all been changed to unix line endings, which made it not work
unless the interpreter was explicitly invoked on it; and forgetting to
/test.py instead of test.py.  Even though I had *just* ./test.sh 5
seconds before, I forgot on test.py.  Bleaugh.

Sorry to waste bandwidth.

<craft>ly y'rs,
Ivan;-(
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
http://www.pauahtun.org 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
Author:  Teach Yourself Python in 24 Hours



From jmg at ecs.soton.ac.uk  Wed Apr  5 04:08:38 2000
From: jmg at ecs.soton.ac.uk (Jacek Generowicz)
Date: Wed, 05 Apr 2000 09:08:38 +0100
Subject: Why should I switch to Python? - Infinity of Primes
References: <1257244378-5884705@hypernet.com>
Message-ID: <38EAF486.D5CA8A75@ecs.soton.ac.uk>

Gordon McMillan wrote:

> David C. Ullrich wrote:
>
> > ... Could be that there's a proof of the existence
> > of infinitely many primes using FTA that has some property
> > not shared by Euclid's proof, but "constructive" isn't it.
>
> Oh, come on. There are at least 3 well known constructive
> proofs that all odd numbers are prime:
>
> The mathematician's:
>  3 is prime,
>  5 is prime,
>  7 is prime
>  as so on, by induction.
>
> The physicist's:
>  3 is prime,
>  5 is prime,
>  7 is prime,
>  9 is prime <experimental error>
>  11 is prime...
>
> The computer scientist's:
>  3 is prime,
>  3 is prime,
>  3 is prime ....
>
> - Gordon

The engineer's:
3 is prime,
5 is prime,
7 is prime,
9 is prime,
11 is prime,
13 is prime,
15 is prime,
.
.
.





From brokeninside at my-deja.com  Mon Apr 24 16:36:50 2000
From: brokeninside at my-deja.com (Lee Malatesta)
Date: Mon, 24 Apr 2000 20:36:50 GMT
Subject: Installing Python on NT
Message-ID: <8e2b92$9n6$1@nnrp1.deja.com>

I'm interested in installing Python on a properly battened down NT
workstation for which I do not have administrative priveledges.  This is
my workstation at work.  Mostly I just want to play around with
SGMLtools which is largely written in Python.  The installer told me
that I have to install logged in as administrator.  This is the first
program I've installed where this was necessary.  Unfortuntely for me to
install this as admin, I'd have to put in a request to PC support, they
would have to review to see if there is a need for me to have Python
installed and then send one of the PC support personelle out to install
for me.

Anyone know of anyway I can circumvent the need to be root to install?

BTW, for the most part I use the Workstation as an Xterminal to do work
on our Solaris boxes, but I'm not even going to start on the
administrative nightmare that would be required to get SGMLtools
installed on our Solaris machines which already have python installed.

Thanks much,

-lee

--
====================================================
Lee Malatesta
Unix Programmer by Day, Wild Eyed Mystic by Night


Sent via Deja.com http://www.deja.com/
Before you buy.


From ljz at asfast.com  Sat Apr  1 10:39:21 2000
From: ljz at asfast.com (Lloyd Zusman)
Date: 01 Apr 2000 10:39:21 -0500
Subject: Why should I switch to Python?
References: <seas5ia89177@corp.supernews.com> <1EeF4.43$CA.2422@news> <oqzord6h5i.fsf@titan.progiciels-bpi.ca>
Message-ID: <ltzordzv86.fsf@asfast.com>

Fran?ois Pinard <pinard at iro.umontreal.ca> writes:

> "Chris Ryland" <cpr at emsoftware.com> writes:
> 
> > Oh, no, not another Perl vs. Python war! ;-)
> 
> That remembers me of a little story that made me smile, years ago.
> This was the comment of a discouraged nurse at a recruiting centre for
> the army, at the entrance of the medical examination room.  "I told
> them for _years_ to remove their shirt, and they still do not know
> they should."
> 
> There will be newcomers to Python all the time, [ ... ]
> [ ... ]
> You people received me with an exquisite
> politeness and patience, and my liking of Python deepened instantly as
> it got kind of fuzzily intermixed, in my fragile mind, with the
> feeling of good human values, as well.

I share your appreciation for the civility and humaneness here in
c.l.python, and I would also add that the lack of this is one of the
main reasons I stopped contributing to a certain other newsgroup that
used to be one of my main haunts for years.  There are literally
dozens of times when I saw newcomers to that newsgroup being treated
with a mind-set that matched that of the nurse in your little story
(and often much worse).

Your story is a gem and I wish I had known it back then, when I had
been trying many other ways to get through to the regulars in that
infamous other newsgroup.

> We should not see the potential for a war every time Perl is
> discussed, and be patient instead.  It looks all pretty bearable so
> far.  And if it was becoming too heavy, we could prepare pre-canned
> replies for the most questions, and even suggest a pointer to this FAQ
> right into the message footer generated for this list.  Welcoming
> people in person is better!

I agree.  The problem will be that once Python reaches a certain
critical mass of popularity, a steady stream of interested newcomers
will start showing up with these FAQ's.  It appears that Python is
nearing that critical mass, and I think that the ensuing influx of
newcomers might very well tax the patience of many of us.  This could
cause many of us to find ourselves temporarily suffering from the
dreaded malady called "ANFM" ("Anti-Newbie Flame Mode").  I would like
to nominate your wonderful little story about the nurse in the
recruiting station as the official Pythonic remedy and vaccine for
ANFM.

> Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard

-- 
 Lloyd Zusman
 ljz at asfast.com


From donn at oz.net  Sun Apr 23 00:04:55 2000
From: donn at oz.net (Donn Cave)
Date: 23 Apr 2000 04:04:55 GMT
Subject: executable "wrapper" gui -- advice sought
References: <xhnM4.1153$PP5.17158@news.swbell.net>
Message-ID: <8dtsp7$e7h$0@216.39.151.169>

Quoth "Grant Goodyear" <g2boojum at hotmail.com>:
| I'm in the process of writing an exceedingly simple gui that will take an
| already existing binary executable and input text file (I don't want to
| break the
| working system!) and allow the user to modify the input arguments,
| if desired, and then start the executable.  That part is straightforward.
| What I'm not sure how to do is to capture the executable's std out as
| it's generated.  My goal is to echo that output to a seperate window, and
| also to use it to generate a periodically updated html "status report" so
| that the program's progress can be monitored remotely.  Working
| through the gui aspects isn't all that difficult, but I have _no_ idea how
| to grab the executable's std out _while it is running_ and then have the
| python gui do something (anything!) with that data.  Any advice?

To start with, have you decided what operating system you're
going to use for this project?  That will matter a lot!  It
sort of sounds like we're on some type of UNIX, but it may
be important which one.

Can you modify the ``existing binary executable'', if it turns
out to have some minor problems with this arrangement?

Does it have a generally teletype-like interface, where its
output appears at the bottom and everything eventually will
scroll off the top, or does is it a ``full screen'' system?

If you'd rather just start bashing around on something, you
might look at os.popen().  Or depending on your UNIX flavor,
the "pty" module might carry the day.

	Donn Cave, donn at oz.net


From robin at alldunn.com  Thu Apr 27 11:07:33 2000
From: robin at alldunn.com (Robin Dunn)
Date: Thu, 27 Apr 2000 08:07:33 -0700
Subject: [ANNOUNCE] New wxPython
Message-ID: <00be01bfb05a$5178ce00$3225d2d1@ARES>

wxPython 2.1.15 is now available for download from
http://alldunn.com/wxPython/  This release includes new features such as a
new and improved wxGrid, ActiveX wrappers and a wxStyledTextCtrl class.

--
Robin Dunn
Software Craftsman
robin at AllDunn.com
http://AllDunn.com/robin/
http://AllDunn.com/wxPython/  Check it out!





From charlie at intelligenesis.net  Fri Apr 14 12:10:16 2000
From: charlie at intelligenesis.net (Charlie Derr)
Date: Fri, 14 Apr 2000 12:10:16 -0400
Subject: Comparing PythonWin and IDLE?
In-Reply-To: <38F66A4E.92E3BDCA@gmx.de>
Message-ID: <NDBBLJJLLFOJMLCMJPLOGEEADJAA.charlie@intelligenesis.net>

 	instead of:

import myModule

	try this:

reload(myModule)

	and you should be able to get your recently modified changes to take effect
without restarting python

	~c


Michael Scharf wrote:
|
|David Smith wrote:
|>
|> Warren Postma wrote:
|>
|> >  I find once a module has
|> > been imported once, it often will not import again, even if I
|have changed
|> > the source (.py) file, the .pyc file does not appear to be
|regenerated, and
|> > any errors in the .py file can only be fixed by editing and
|saving the .py
|> > file, quitting the IDE and restarting. This appears to happen
|for both IDLE
|> > and Python. Anybody else seen this?
|>
|> When you edit a file under Idle, Save it and then hit Control-F5.  Then
|> you can re-import the file.
|
|But a re-import should also change existing instances of
|functions, classes and methods defined in the module. OK,
|there are some cases, where a new version is incompatible
|with the old data members, but sometimes it would be
|helpful....
|
|
|Michael
|--
|     ''''\     Michael Scharf
|    ` c-@@     TakeFive Software GmbH, a Wind River Company
|    `    >     http://www.WindRiver.com
|     \_ V      mailto:Michael.Scharf at gmx.de
|--
|http://www.python.org/mailman/listinfo/python-list
|




From mwh21 at cam.ac.uk  Sun Apr  2 16:16:09 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 02 Apr 2000 21:16:09 +0100
Subject: Debugging confusion -- too many stacks!
References: <000a01bf9c6b$3f80c2a0$752d153f@tim> <38E739A1.45F4B199@tismer.com> <m37legk3kx.fsf@atrus.jesus.cam.ac.uk> <38E75E9C.7501756C@tismer.com>
Message-ID: <m34s9kjm2e.fsf@atrus.jesus.cam.ac.uk>

Christian Tismer <tismer at tismer.com> writes:

> Hey, we forgot about a stack :-))
> 
> Michael Hudson wrote:
> > 
> > Christian Tismer <tismer at tismer.com> writes:
> > 
> > > This VM stack could also be replaced by a set of registers, since
> > > the maximum stack size is always known at compile time.
> > 
> > My gut tells me that making this work with exceptions would be a dog.
> 
> I don't think so. Exception handlers have their own little stack
> in each frame. It is currently fixed to 20 items, wasting quite
> a lot of space most of the time (and this will vanish in
> Stackless "Heresy" 1.2), but it won't cause problems with a
> register machine.

Hmm.  You still need a evalutaion stack though, don't you?  I mean,
implementing function call without would one mean a pretty radical
change to the interpreter.

> > It's the sort of thing that might be fun to play with if the compiler
> > and eval loop were written in Python, but if I have to hack C to play
> > with these things, it just becomes tedious.  Still, maybe if I have a
> > month or two with nothing else to do...
> 
> A while ago, I considered this, too. Meanwhile I don't, since
> it doesn't bear a very reasonable speed gain. The win would be
> to have less of the tiny operations occouring all the time,
> and to have less of stack pointer operations. But when looking
> closer to it, you will find that the most slowdown lies in
> the outmost interpreter loop, which checks for periodic things
> for every opcode, regardless wether it is very fast or a slow
> function call.
> In Stackless Python, I have stolen these cycles already. The eval
> loop has some internal shortcuts to avoid this many checks. The fast
> opcodes are all running as fast as possible, avoiding these
> checks. One could still save some jumps and some assignments,
> but this is not enough to justify the lots of changes that a
> register implementation would impose to other modules.

Oh right.  I haven't looked at stackless python's ceval.c for a while.
A register machine might be a bit more elegant, but not much, I
suspect.

> It can still make sense to provide a hand-full of combined
> opcodes which have high frequency. jump_if_true combined
> with pop_top for instance, or some load/store combinations.
> If you have a proposal, and if you will support them through
> the bytecodehacks, I'd be happy to add them to Stackless Heresy.

Another thing that would make sense is typed bytecodes and typed
registers or a typed stack.  So if you can prove a and b are integers,
say, then you can stick a,b in registers 1,2 and compile

 a = a + b

into 

 IBINARY_ADD 1 2 1

Another observation: all the bytecodes currently understood by the
Python VM have length 1 or 3.  There's no hard limit though;
IBINARY_ADD above would have to be at least four bytes long.  Further,
there's no reason you can't embed chunks of another bytecode in the
midsts of the one the Python VM understands, such as one that operates
on typed registers.  Then you use bytecodehacks to pick functions
apart, do some cfa and dfa to find which variables are floats or ints
and generate typed bytecode for operations on them.

I did some of the above once, then it became apparent that the flow
analysis was far too hard for me.  I might give it another whack
sometime...

Cheers,
M.

-- 
well, take it from an old hand: the only reason it would be easier
to program in C is that you can't easily express complex  problems
in C, so you don't.                 -- Erik Naggum, comp.lang.lisp


From linuxmusic at crosswinds.net  Fri Apr 21 21:24:12 2000
From: linuxmusic at crosswinds.net (John Littler)
Date: 22 Apr 2000 11:24:12 +1000
Subject: urllib (sort of reload) question
Message-ID: <m3purjdj2b.fsf@localhost.localdomain>

Hi,
I use urllib to get a number of URLs and parse the
results. One pesky URL doesn't expire it's content
correctly and as an interim measure I need to reload
the page in my browser before getting the data with
python. I'm going through squid so I guess that's where
the old stuff is being picked up.
Does anyone know how to emulate browser "reload" in
python? Looking at the urllib code didn't provide me
with any clues.
TIA
John


-- 
pgp public key @ http://www.crosswinds.net/~linuxmusic/pubring.html


From c.evans at clear.net.nz  Fri Apr 28 05:41:51 2000
From: c.evans at clear.net.nz (Carey Evans)
Date: Fri, 28 Apr 2000 09:41:51 GMT
Subject: Python COM troubles
References: <390a2f04.149999918@news.onlineisp.com>
	<L5MN4.5088$v85.30985@news-server.bigpond.net.au>
Message-ID: <877ldi36lc.fsf@psyche.evansnet>

"Mark Hammond" <mhammond at skippinet.com.au> writes:

> If you search deja news, there should be lots of info about this - but
> generally relating to "Lotus Notes", which also has the same problem.
> 
> [BTW - drop them a line, and tell them they should fix their objects :-]

Good news!  From 5.0.2b onwards the Notes client comes with a new COM
interface, in addition to the old `OLE Automation' that doesn't work
very well.  It's only an interface to the backend objects, i.e. things
like NotesSession and NotesDatabase, not the objects with "UI" in
their names, but it's better than LotusScript.

-- 
	 Carey Evans  http://home.clear.net.nz/pages/c.evans/

		 "I can't believe it's not Cthulhu!"


From gmcm at hypernet.com  Wed Apr  5 10:41:20 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Wed, 5 Apr 2000 10:41:20 -0400
Subject: Installer Problem Finding Modules
In-Reply-To: <01bf9eb5$7ee884e0$1e41c3d1@Cyberus>
Message-ID: <1257176798-9949680@hypernet.com>

Gordon Williams wrote:

> I am trying to run Gordon McMillan's Installer (3f on W95) and I am having
> a problem with it finding the modules that I am importing into the main
> program. 
[snip]
> The installer ignores the all the imports except for the wxPython stuff
> which it appears to get all of.  Any ideas where I may be going wrong.  I
> hope that I dont have to list all these modules as dependancies as some of
> these modules also call other modules.

These modules are in the same directory as your main script? 
For some reason Win95 fails here, though NT succeeds. The 
workaround seems to be to set 
PYTHONPATH=<absolute path to your directory>
in the DOS box before running the Installer.
 
> Maybe this is a related problem, but I dont think so:
> All the modules along with the main module are in a sub directory of the
> installer program.  When I run python ..\simple.py bs_main.py from a dos
> box
> I was geting an error "can't open 'E:Program" when it it was trying to
> start build.py.  I ended up changing simple.py line 74 to
> cmd = '"%s" %s %s %s' % (sys.executable,
>                            '../Builder.py',                      <<< change
> is here
>                            cfgfile, tkcfg)
> 
> to get it to work.

Grrr. The fact that Windows allows spaces in directory names 
doesn't mean the "feature" should be used.

A better fix, I think, is to quote all the %s's.

- Gordon



From effbot at telia.com  Tue Apr 25 10:06:21 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Tue, 25 Apr 2000 14:06:21 GMT
Subject: Question about exception-handling mechanism
References: <8e46mt$4nh$1@nnrp1.deja.com>
Message-ID: <xDhN4.4004$Za1.59528@newsc.telia.net>

willfg at my-deja.com wrote:
>    I'm new to Python, but find its exception handling mechanism much
> more powerful than comparable languages; but a question was put to me
> that being new I can't come up with a very articulate answer. A
> colleague asked why in an exception handling mechanism you'd want the
> ELSE block to be executed if you don't throw an exception as opposed to
> a FINALLY block. Anyone used this feature in practice? Thanks in
> advance for your input.

the standard pattern for try/except/else is:

    try:
        do something
    except SomeError:
        deal with error
    else:
        deal with success
    proceed

this could also be written:

    try:
        do something
        deal with success
    except SomeError:
        deal with error
    proceed

but this only works if the "deal with success"
code cannot raise a SomeError exception...

...

try-finally is different; the code in the finally
clause will be executed in either case, but if
there was an exception, it's reraised after the
finally clause has finished:

    create a mess
    try:
        raise SomeError
    finally:
        clean up
    this line will never be reached

hope this helps!

</F>

<!-- (the eff-bot guide to) the standard python library:
http://www.pythonware.com/people/fredrik/librarybook.htm
-->






From dworkin at ccs.neu.edu  Tue Apr 25 19:36:51 2000
From: dworkin at ccs.neu.edu (Justin Sheehy)
Date: 25 Apr 2000 19:36:51 -0400
Subject: os.path.walk() question.
In-Reply-To: "Andrew McDowell"'s message of "25 Apr 2000 22:19:44 GMT"
References: <8e55m0$pvl$1@hammer.msfc.nasa.gov>
Message-ID: <vndpurdycq4.fsf@camelot-new.ccs.neu.edu>

"Andrew McDowell" <drew.mcdowell at msfc.nasa.gov> writes:

> I need to find out if a file exists under a directory hierarchy.
> The hierarchy could be quite large so I don't want to continue recursing
> through the hierarchy if the file is found one or two directories down.   Is
> there a way that you can tell os.path.walk() to stop searching?

You could put the os.path.walk() call inside a try/except, and raise
an appropriate exception when the desired item is found.

-Justin

 



From tony at lsl.co.uk  Tue Apr 11 10:41:11 2000
From: tony at lsl.co.uk (Tony J Ibbs (Tibs))
Date: Tue, 11 Apr 2000 15:41:11 +0100
Subject: Regular Expression Help for Python Newbie.
In-Reply-To: <N5AI4.1125$Za1.20080@newsc.telia.net>
Message-ID: <000901bfa3c3$fabd3f80$f0c809c0@lslp7o.lsl.co.uk>

I was being a pedant about HTML allowing <TR>'s to be absent or not...

> Fredrik Lundh replied:
> you're right -- but only in theory ;-)

Well, yes, I knew that!

> so when the browser stumbles upon a <TABLE> followed by
> a <TD>, it can easily figure the rest out itself, based on the
> DTD.

in that case, and in the example given, because it's not in a record 'cos
it's just ended one, yes, it's obviously easy.

Then, on trying to "guess" what to do next:
> > And obviously that's what some browsers do. This way also lies
> > madness.
>
> some?  I suspect 99.99% of browsers used for serious surfing
> do things like this -- and *much* more...  (if you want to see
> some real madness, go browse the original navigator sources ;-)

as I said, madness did lie that way! (you perhaps haven't spent as much time
as I have trying to get browsers to behave correctly when they have
*correct* HTML, or striving to find out what is wrong when they guess wildly
incorrectly because of a stupid typo - "where did the rest of my page go?".
I seem to be fated that way sometimes).

I don't mind the browser attempting to continue, what I mind is wild
unpredictability...

> (the real design mistake is of course to do the fixes in the
> browser, instead of doing them in the authoring system or
> the server, but that's another story).

on which we agree entirely. Unfortunately, given all the people who don't
know what they're doing when they write authoring (well, I hesitate to say
"systems") I guess we're stuck. But it's still a pity, and I can't help
grumbling.

And (as you gave an example) obscure messages from validators don't help at
all.

Tibs

--
Tony J Ibbs (Tibs)      http://www.tibsnjoan.demon.co.uk/
Feet first with 5 wheels...
My views! Mine! Mine! (Unless Laser-Scan ask nicely to borrow them.)




From dan at cgsoftware.com  Sun Apr  9 12:22:17 2000
From: dan at cgsoftware.com (Daniel Berlin)
Date: Sun, 9 Apr 2000 09:22:17 -0700 (PDT)
Subject: tabular data
In-Reply-To: <8cq9io$ca0$1@nnrp1.deja.com>
Message-ID: <Pine.LNX.4.10.10004090919320.30545-100000@propylaea.anduin.com>


On Sun, 9 Apr 2000 poupaerte at my-deja.com wrote:

> Hi
> 
> > def sortedby(list,*indices):
> 
> I didn't know Python worked with pointers, or else what does *indices
> mean? The fine manuals do not explain the *var construct ...

It's not a pointer, it's a tuple of the rest of the arguments passed in.
def test(a,*b):
	print "a",a,"b",b

>>> test(5,6,7,8,9,10,11)
a 5 b (6, 7, 8, 9, 10, 11)


> 
> >    nlist = map( lambda x,indices=indices:
> 
> The expression "indices=indices" seems absolutely strange to me. I
> don't have a clue of what it could be doing, except for being true at
> all times.
> 
i missed the rest of the message, so i can't tell without the rest of the
map command.

> Is there some kind of lengthier tutorial on the lambda construct; it
> seems quite an advanced construct not easily to be found elsewhere in
> other programming languages, a bit like regexes ...
I want to say lambda originated in LISP, but i could be wrng.
It's used extensively in lisp and scheme.

> 
> Greetings
> Erik
> 




From mdefreitas at sikorsky.com  Thu Apr 27 14:19:22 2000
From: mdefreitas at sikorsky.com (mdefreitas at sikorsky.com)
Date: Thu, 27 Apr 2000 18:19:22 GMT
Subject: Q: Recursive embedded Python interpreters?
Message-ID: <8ea0ao$fro$1@nnrp1.deja.com>

I am working on a helicopter simulation user-interface that takes
commands such as:

> start_sim
> freeze_sim
> altitude = 100.0
> end_sim

I would like to add scripting capabilities to this user-interface, so I
am attempting to extend and embed Python. I am adding a new command
(python) to my user interface that will run a python script with
optional arguments:

> python  script.py  arg1  arg2

My first step was to extend python with a module ?ui? that python
scripts can import to execute various user interface commands. This was
fairly easy. I am now trying to embed the python interpreter into my
user interface. I anticipate a possible problem when a python scripts
executes the ?ui? command to run another python script. I anticipate re-
entry issues. I assume I have to create a new interpreter for each
nested call to the python interpreter. I really don?t understand the
mechanics of how this works, but from comp.lang.python archives, and
from the Demos/pysvr example that came bundled with the python1.5
distribution this is what I think I have to do:

static int nest_level = 0;
static PyThreadState *gtstate = 0;

void interp(char *script) {
   if (nest_level == 0) {
      nest_level++;  // init topmost interp
      Py_Initialize();
      PyEval_InitThreads();
      gtstate = PyEval_SaveThread();
   } else {
      PyEval_ReleaseLock();  // nested interp needs to relinquish lock?
   }

   PyEval_AcquireLock();
   PyThreadState *tstate = Py_NewInterpreter();
   FILE *fd = fopen(script, ?r?);
   PyRun_SimpleFile(fd, script);
   Py_EndInterpreter(tstate);
   PyEval_ReleaseLock();
   nest_level--;
   if (nest_level == 0) {
      PyEval_AcquireThread(gtstate); // top-most interp... clean up
      gtstate = 0;
      Py_Finalize();
   } else {
      PyEval_AcquireLock();  // need to reacquire lock if nested?
   }
}

Is this right? Did I miss anything? Also? how do I pass arguments to
the (possibly nested) scripts.

Thanks in advance for any help.





Sent via Deja.com http://www.deja.com/
Before you buy.


From donn at u.washington.edu  Wed Apr 26 18:03:52 2000
From: donn at u.washington.edu (Donn Cave)
Date: 26 Apr 2000 22:03:52 GMT
Subject: global variables or inheritence
References: <28pzoqghaoh.fsf@lada.kom.auc.dk> <28p8zy0irgc.fsf@lada.kom.auc.dk> <eRHN4.1113$3K4.19980@news.uswest.net> <28p4s8oipzh.fsf@lada.kom.auc.dk> <j6IN4.1168$3K4.20204@news.uswest.net>
Message-ID: <8e7p48$ljg$1@nntp6.u.washington.edu>

Quoth Thomas Rasmussen <simpsons at kom.auc.dk>:
...
| Well i'm programming a small admin tool, which must use the users
| $EDITOR for editing files. This works fine if the EDITOR sysenv is
| set, if not, then i want to ask the user which editor he want to use,
| and then pass this variable to the module that handles
| filemanipulation. This modules is also in another file if it
| matters... I must add that this is my first real program made in
| python...

When things get difficult in Python, my reflex is to write a
big hairy class.  Maybe something like this would work, or at
least give you an idea:

  import os
  import sys

  class UserEditor:
          def __init__(self, default_editor = None):
                  self.program = default_editor
          def get(self):
                  if not self.program:
                           try:
                                    self.program = os.environ['EDITOR']
                           except KeyError:
                                    sys.stdout.write('Editor? ')
                                    ... etc.
                  return self.program
          def set(self, editor):
                  self.program = editor
          def invoke(self, file):
                  program = self.get()
                  return os.system('%s %s' % (program, file)) == 0

  editor = UserEditor()


Let's say that's at module scope in the "interfaces" module.  In other
modules you may say

  filemanipulation.xyzfile(interfaces.editor.get(), ...)

or maybe this class will turn out to be a convenient place to do some
stuff that is now cluttering up filemanipulation, so you should say

  filemanipulation.xyzfile(interfaces.editor, ...)

	Donn Cave, University Computing Services, University of Washington
	donn at u.washington.edu


From aahz at netcom.com  Sun Apr  9 01:45:09 2000
From: aahz at netcom.com (Aahz Maruch)
Date: 9 Apr 2000 05:45:09 GMT
Subject: Windows quotes (was Re: Installer Problem Finding Modules)
References: <1257176798-9949680@hypernet.com>
Message-ID: <8cp5d5$qcu$1@slb6.atl.mindspring.net>

In article <1257176798-9949680 at hypernet.com>,
Gordon McMillan <gmcm at hypernet.com> wrote:
>
>Grrr. The fact that Windows allows spaces in directory names doesn't
>mean the "feature" should be used.
>
>A better fix, I think, is to quote all the %s's.

I re-discovered recently that you need to use "-style quotes on Windows;
'-style quotes don't work.
--
                      --- Aahz (Copyright 2000 by aahz at netcom.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

Why doesn't "Just Say NO" include caffeine, nicotine, alcohol, Prozac,
and Ritalin?  --Aahz


From conversy at lri.fr  Thu Apr 20 13:10:17 2000
From: conversy at lri.fr (Stephane Conversy)
Date: Thu, 20 Apr 2000 19:10:17 +0200
Subject: problem implementing a threaded module
References: <38FF2069.1F25@marshall.edu>
Message-ID: <38FF39F9.6334571F@lri.fr>

Hi. I'm trying to implement a C module that launches a pthread
and calla a python call-back.

I think it's not a big deal for those who know how to do it, but
my implementation does not work (seg fault).

I didn't really understand what to do exaclty:
should I create an interpreter state for each thread, using
PyInterpreterState_New ?
should I create a thread state for each thread with  PyThreadState_New and
swap
it with the current thread state ?

I can't figure out what to do. Is there any example that just creates a
pthread
and calls a python func ?

Thank you.

--
St?phane Conversy
http://www-ihm.lri.fr/~conversy/
mailto:conversy at lri.fr





From warlock at eskimo.com  Sun Apr  9 16:08:11 2000
From: warlock at eskimo.com (Jim Richardson)
Date: Sun, 09 Apr 2000 20:08:11 GMT
Subject: Why should I switch to Python? - Infinity of Primes
References: <Pine.GSO.4.10.10004060914420.3428-100000@sundial>  <8coh4i$hnu$1@newshost.accu.uu.nl>
Message-ID: <slrn8f1p6d.rdg.warlock@raven.myth>

On 8 Apr 2000 23:59:14 GMT, 
 Martijn Faassen, in the persona of <m.faassen at vet.uu.nl>,
 brought forth the following words...:

>Summary of the thread:
>
>Why should I switch to Python from Perl?
>[usenet lore knows that huge flamewar may be about to explode]
>Some discussion on Python versus Perl.
>Under what circumstances Python could be better.
>Several people say: stick to Perl if you're already comfortable with it.
>[usenet lore scratches its head]
>Tim Peters joins in.
>thread segues into math.
>[usenet lore looks flabbergasted] 
>long discussion on math, formal proof of the infinity of primes ensues
>Constructivist math, Brouwer, half of Pythondom joins into the discussion
>[usenet lore panics!]
>Discussion mentiones Haskell and Scheme, as they always do.
>[usenet lore goes mad and concludes original poster should switch to
>Haskell, *now*]
>
>I just think this thread is very funny. It should count as a reason to
>use Python. Although the place has become a bit hectic at times in this
>newsgroups, we seem to have various antibodies against flamewars here. :)
>

It's Python-jitsu
 Don't fight the flamers, use their energy for constuctive discussions :)

-- 
Jim Richardson
	Anarchist, pagan and proud of it
WWW.eskimo.com/~warlock
	Linux, because life's too short for a buggy OS.



From effbot at telia.com  Mon Apr  3 17:45:41 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Mon, 03 Apr 2000 21:45:41 GMT
Subject: Python 1.6 alpha 1 released
References: <200003312130.QAA04361@eric.cnri.reston.va.us> <38E52716.A1C63B5D@roguewave.com> <rJ9F4.549$n68.187227648@newsb.telia.net> <38E52D76.1820FC32@roguewave.com> <8c7tfd$k80$1@nntp9.atl.mindspring.net> <8catgc$5i84$1@uvaix7e1.comp.UVic.CA>
Message-ID: <9i8G4.1196$n68.192123904@newsb.telia.net>

John Aycock wrote:
> Although I don't have time to volunteer, it wouldn't be too difficult
> to write a program which would look through existing Python code and
> warn of potential uses of these changed constructs.

after reading your post, tim peters used barry's new "post
via the time machine" facility:

http://www.deja.com/=dnc/getdoc.xp?AN=591243549

changing it to look for "connect" should be trivial (he's probably
done that by now).

fwiw, while it may miss many real appends (see comments in
script), I'm pretty the number of missed connect calls are very
close to zero.

</F>




From neelk at brick.cswv.com  Fri Apr 28 19:44:20 2000
From: neelk at brick.cswv.com (Neel Krishnaswami)
Date: 28 Apr 2000 23:44:20 GMT
Subject: What does "first class" mean?
References: <3909F2F5.A6F91256@spacenet.tn.cornell.edu>
Message-ID: <slrn8gkalp.3bk.neelk@brick.cswv.com>

Tom Loredo <loredo at spacenet.tn.cornell.edu> wrote:
> 
> I've often seen reference here (and on the Ruby page) to Python's
> treatment of functions and methods as "first class objects."  What
> does this mean? 

It means that functions and methods are values just like like lists
and numbers. Basically, you can write functions that accept functions
as arguments, return them as values, create new ones and bind them to
variables just like any other object. 

In Pascal or C, for example, you can't write a procedure that creates
a function and returns it to its callers, whereas in Scheme or Python
this is trivial.


Neel


From duncan at rcp.co.uk  Wed Apr 19 07:11:52 2000
From: duncan at rcp.co.uk (Duncan Booth)
Date: 19 Apr 2000 11:11:52 GMT
Subject: WBMP support for PIL
Message-ID: <8F1B74290duncanrcpcouk@194.238.50.13>

I just wrote a WBMP module for PIL. If anyone wants it they can download it 
from http://www.rcp.co.uk/distributed/Downloads (the file is 
wbmpconvsrc.zip). The download includes a script for converting between 
WBMP and any other PIL supported bitmap format.

For those that dont know: WBMP is wireless bitmap format, the graphic 
format used by WAP mobile phones. PIL is Python Imaging Library, but you 
all knew that anyway.


From mhammond at skippinet.com.au  Wed Apr 12 06:17:17 2000
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Wed, 12 Apr 2000 10:17:17 GMT
Subject: Embedding Python in Visual C++ 6.0
References: <8d1he4$26d$1@nnrp1.deja.com>
Message-ID: <N2YI4.2030$UB.13624@news-server.bigpond.net.au>

You also need to override InitApplication().

I should update the docs, but until then, simply look at Pythonwin.cpp and
see what it does - whatever it does, it obviously works :-)

Mark.

<matthias_w at my-deja.com> wrote in message
news:8d1he4$26d$1 at nnrp1.deja.com...
> Hy! :-)
>
> Maybe someone (Mark Hammond??) can help me. I
> tried to embedd the win32 extentions from Mark in
> a visual c++ project, but everytime the
> <b>Win32uiInitInstance()</b> is called (via
> <b>glue.InitInstance()</b>) I get an unhandled
> exception at 0xC0000005 Access Violation and the
> debugger jumps to 1E401F69 - the address of the
> Win32uiInitInstance() function. When i try create
> a break-point in this function this break-point
> is disabled %-( by the compiler. The same is true
> if i try to debug the original pythonwin project.
>
> Help please!
>
> Matthias
>
>
> Sent via Deja.com http://www.deja.com/
> Before you buy.




From none at none  Wed Apr 19 15:43:20 2000
From: none at none (Eric Jacobs)
Date: Wed, 19 Apr 2000 14:43:20 -0500
Subject: Very Horrible Question About Goto's
References: <6325E90C8B6ED21189080001FA7E2ADA0D4051@IBA1><2hkL4.8514$5D.18609@ozemail.com.au>
Message-ID: <sfruuci2eoc31@corp.supernews.com>

In article <2hkL4.8514$5D.18609 at ozemail.com.au>,
"Jason Stokes" <jstok at bluedog.apana.org.au> wrote:
> 
>     >Richard Jones wrote in message
>><6325E90C8B6ED21189080001FA7E2ADA0D4051 at IBA1>...
>     >I'm trying to create a converter from our own internal language
> (horrid) >into python.  My only problem so far is that the original language
> uses >*LOTS* of goto's.  Is there any way that this can be replicated within
>>python ?
> 
> 
>     Yes and no.  It should be possible to translate arbitary gotos into
> Python constructs, but the results won't be readable by a human programmer.
> I'm not quite sure how right now, but I could think about it.  Because the
> Goto is such an unstructured construct, it's hard to translate it into
> structured terms without bizarre circumlocutions.

In fact, it may not even be directly possible in a general case. Here's
a loop with two entry points:

     if (something) goto entry2
  entry1:
     ... (1)
  entry2:
     ... (2)
     goto entry1

The idea here is called reducibility; it means the edges in the flow graph
which create cycles must point to blocks which have already been encountered
on the path to that edge ("back edges"). Here, the second goto violates that
requirement. This means that it can't be reduced to a hierarchy of blocks
(whether Pythonic blocks or otherwise.)

It IS possible to translate it by duplicating code:

      if (something) {
         ... (2)
      }
   entry1:
      ... (1)
      ... (2)
      goto entry1

and that's how most C compilers would probably handle it anyway (for
optimization purposes). Formally, it's not the same flow graph, but it
does the same things.


From phd at phd.russ.ru  Sun Apr  9 12:37:48 2000
From: phd at phd.russ.ru (Oleg Broytmann)
Date: Sun, 9 Apr 2000 16:37:48 +0000 (GMT)
Subject: Python paradigms
In-Reply-To: <1j2I4.252$rc9.186596864@newsb.telia.net>
Message-ID: <Pine.LNX.4.21.0004091636070.13883-100000@fep132.fep.ru>

On Sun, 9 Apr 2000, Fredrik Lundh wrote:
> Oleg Broytmann wrote:
> >    blah = ((foo() and [poo()]) or [boo()])[0]
> >
> >    In C, "foo() ? poo() : boo()" will not call boo() if foo() returns
> true;
> > but in Python, you'll have a side-effect of calling boo() even if foo()
> > returns true - what is wrong, usually.
> 
> did you try running the code?
> 
> python's "and" and "or" operators doesn't evaluate the
> second argument unless they really have to...

   Yes, I missed that, sorry.
   Anyway, it is strange (not to say "ugly") construct; I prefer not to see
it in a code...

Oleg.            (All opinions are mine and not of my employer)
---- 
    Oleg Broytmann      Foundation for Effective Policies      phd at phd.russ.ru
           Programmers don't die, they just GOSUB without RETURN.




From schmitthead at my-deja.com  Mon Apr 24 02:08:29 2000
From: schmitthead at my-deja.com (schmitthead at my-deja.com)
Date: Mon, 24 Apr 2000 06:08:29 GMT
Subject: regulare expression searches and newlines
Message-ID: <8e0oco$k75$1@nnrp1.deja.com>

I was doing a search like this:

searchobj = re.compile( r'^foobarbaz' )

where I was looking for something at the beginning of a line.  This
works well for native text files, but not when I was searching through
Linux files running on NT or NT files on Linux or some other variation
because of the newline difference, I guess.  Is there any way to handle
the general case?


Sent via Deja.com http://www.deja.com/
Before you buy.


From gregm at iname.com  Thu Apr 27 08:08:21 2000
From: gregm at iname.com (Greg McFarlane)
Date: Thu, 27 Apr 2000 22:08:21 +1000
Subject: accessing BLT vectors from C
In-Reply-To: <8dq957$tc8$1@nnrp1.deja.com>; from genehilton@my-deja.com on 21 Apr 2000 at 07:11:47PM
References: <8dq957$tc8$1@nnrp1.deja.com>
Message-ID: <20000427220821.63091@nms.otc.telstra.com.au>

Gene,

What is the value of bltname?  The names that Pmw give to Blt vectors
are PY_VEC0, PY_VEC1, PY_VEC2, etc.  You can see what Tcl calls are
being made by calling Pmw.tracetk().  For example, this Pmw:
    v = Pmw.Blt.Vector(1)
prints:
    CALL  TK> 1:  ('::blt::vector', 'create', 'PY_VEC0(1)') -> '::PY_VEC0'
Which means, in tcl:
    ::blt::vector create PY_VEC0(1)
	--> ::PY_VEC0

So Blt_VectorExists(interp,"PY_VEC0") should return true.  If not, I'm
not sure what the problem is.

Greg

On 21 Apr, genehilton at my-deja.com wrote:
> Has anyone out there written anything that will enable me to access
> the BLT C api.  I am using BLT as a part of a data acquisition
> system.  It works great, but filling the vectors is really slow in
> python.  I wrote a python module in C to use the Blt c api, but it can't
> find the vectors I created with Pmw.Blt.Vector().  I can access (in the
> c module) vectors that I create in the c-module.  This must be some sort
> of TCL thing I don't understand.  Here is a snippet from the C module
> 
> 
>         if (!PyArg_ParseTuple(args, "OO",&master,&bltobj)) return NULL;
> 
> /* get the name of the vector - PMW stores it as the repr. */
>         bltobjname=PyObject_Str(bltobj);
>         if (!bltobjname) return NULL;
>         bltname=PyString_AsString(bltobjname);
>         if (!bltname) return NULL;
>         printf("%s\n",bltname);
> /* get the interpreter from an argument (for now) */
>          interp=((TkappObject *)master)->interp;
> 
> /* this doesn't work (returns a 0) */
> 	printf("%d\n",Blt_VectorExists(interp,bltname));
>         printf("%d\n",Blt_CreateVector(interp,"myvec",1,&vec));
> /* this works (returns a 1) */
>         printf("%d\n",Blt_VectorExists(interp,"myvec"));
> 
> I call it with the following python code
> 
> import Tkinter
> import Pmw
> import mymod
> v=Pmw.Blt.Vector(1)
> mymod.myroutine(master.tk,v)
> v=Pmw.Blt.Vector(1)
> 
> I get an output like
> 0     (couldn't find it - v created in python)
> 0     (created ok in c)
> 1     (found the one created in c)
> 
> Any pointers would be real helpful
> 
> Gene Hilton
> hilton at boulder.nist.gov
> 
> 
> Sent via Deja.com http://www.deja.com/
> Before you buy.
> -- 
> http://www.python.org/mailman/listinfo/python-list
> 

-- 
Greg McFarlane     INMS Telstra Australia     gregm at iname.com



From sabren at manifestation.com  Sun Apr 23 02:42:43 2000
From: sabren at manifestation.com (Michal Wallace (sabren))
Date: Sun, 23 Apr 2000 02:42:43 -0400 (EDT)
Subject: code not true?
In-Reply-To: <B527F43C.24AD%dgoodger@bigfoot.com>
Message-ID: <Pine.BSF.4.10.10004230234050.5642-100000@chi.pair.com>

On Sun, 23 Apr 2000, David Goodger wrote:

> > Why is this the case? I mean.. I understand why decimals are hard to
> > show in binary... But why can't we use variably long strings, or
> > something? Is it just a speed issue?
> 
> Try writing 1/3 in decimal. How many digits do you need to represent it
> *exactly*? Same issue for 0.07 in binary floating point.

Hmm... I see your point.. I'm going to go out on a limb here, though..
wouldn't "0.1" be a simple string representation of that number in
base 3? We could make it: "3:0.1" or something to show which base its
in..

If python is going to have radix methods for numbers, then it would seem
to me that you could store floats in whichever system made it easiest to
express them.

I guess there's still numbers like pi out there... But even so, if
you're only talking about adding and subtracting and multiplying
floating points..  Hmm... Well, okay... I give up. :)

Cheers,

- Michal
-------------------------------------------------------------------------
http://www.manifestation.com/         http://www.linkwatcher.com/metalog/
-------------------------------------------------------------------------




From paulb at infercor.no  Tue Apr  4 04:37:45 2000
From: paulb at infercor.no (Paul Boddie)
Date: Tue, 04 Apr 2000 10:37:45 +0200
Subject: Python 1.6 alpha 1 released
References: <Pine.GSO.4.10.10004040035420.21084-100000@sundial> <5E877FD7637AB46E.DABBED9CF3C1D64B.CDDDC05AE7374A78@lp.airnews.net>
Message-ID: <38E9A9D9.20556D1C@infercor.no>

Tres Seaver wrote:
> 

[Single argument 'connect' and the standard libraries]

> This is *not* the same case as the list.append() one, where the
> published interface was itself consistent -- I was quite surprised to
> learn just this spring that multi-argument append was even possible.

Same here.

> But the socket module is one of the first I learned in Python, and I
> have consistenly used the two-argument connect.  Constructing a spurious
> "address" object (which has no behavior, and exists only to be torn
> apart inside the implementation) seems a foolish consistency, beyond
> doubt.

Yes, without defining the type of the argument there and then, the reader is
going to start looking around for some kind of hint, and what do they find? And
I can hardly see any positive aesthetic in the use of a tuple as an argument
when "normal" arguments would be more "Pythonic" - they would at least allow you
to write this:

  s.connect(address="123.45.678.9", port=1234)

Although I hear that some code uses a tuple for 'address'... ;-)

> Rough-consensus-and-running-code-wins-every-time'ly

Indeed.

Paul


From thomas at xs4all.net  Mon Apr 17 16:12:13 2000
From: thomas at xs4all.net (Thomas Wouters)
Date: Mon, 17 Apr 2000 22:12:13 +0200
Subject: Another question
In-Reply-To: <dVFK4.41$vr2.5817@news.uswest.net>; from jmassung@magpiesystems.com on Mon, Apr 17, 2000 at 09:22:05AM -0600
References: <dVFK4.41$vr2.5817@news.uswest.net>
Message-ID: <20000417221213.I15664@xs4all.nl>

On Mon, Apr 17, 2000 at 09:22:05AM -0600, Jeff Massung wrote:
> Okay, another quick question. If it is quicker to do:

> import random
> from random import randint
> for i in range(1,1000000):
>     print randint(1,10)

> than this:

> import random
> for i in range(1,1000000):
>     print random.randint(1,10)

> Why would I not do "from <module> import *" with every module I load?

Everyone else already gave you good answers, but since you're primarily
concerned about minor speedups: Using 'from module import *' is a bad idea
if you want speed; it turns off certain optmizations ! Exactly the oposite
of what you intended :) See a posting by Tim Peters a couple of months back,
on optimizations.

The gist of the message was something like this: In absense of unpredicably
namespace modifying code ('from module import *' and 'exec something' in the
local namespace, if I recall correctly) all local variable lookups are
translated into a simple vector into the local namespace. So, 'dont do
that'. If you *insist* on that kind of speedup, import all items by name.

(But dont forget: if your module namespace gets too large, you'll end up
paying a price in performance ;)

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From jon at dgs.monash.edu.au  Tue Apr 11 22:40:23 2000
From: jon at dgs.monash.edu.au (Jonathan Giddy)
Date: Wed, 12 Apr 2000 12:40:23 +1000 (EST)
Subject: ANNOUNCE: Python 1.6 alpha 2
In-Reply-To: <"00Apr11.191729pdt.3438"@watson.parc.xerox.com> from "Bill Janssen" at Apr 11, 2000 07:17:38 PM
Message-ID: <200004120240.MAA11342@nexus.csse.monash.edu.au>

Bill Janssen declared:
>
>ILU seems to work fine with it.
>
>Bill

Without wishing to jinx this good news, isn't the release of 1.6 the
appropriate time to remove the redundant thread.h file?

Jon.



From bobalex at home.com  Mon Apr  3 11:24:08 2000
From: bobalex at home.com (Bob Alexander)
Date: Mon, 3 Apr 2000 08:24:08 -0700
Subject: Tuples -- who needs 'em
Message-ID: <00e101bf9d80$a8384d20$74eb0b18@stcla1.sfba.home.com>

This is not an [intentional] flame bait, nor an April Fool's message, but
think about it: Python has two "sequence" data types, lists and tuples,
which from the Python programmer's perspective are only slightly different.
How does having both types significantly improve life for Python
programmers?

I can think of a few ways in which it makes life worse. Although I doubt
that having both types prevents anyone from mastering the language, it does
add some cognitive load in choosing whether to use a list or a tuple when a
sequence data type is needed. There are times when I started with the
"tuple" choice (since it's more "minimal" than list one should choose it
when one doesn't need the additional list features, right?) but later
changed it to list because I needed a list-type feature. Usually that would
be the need for mutability, but sometimes it's unrelated to mutability, like
wanting a method to search for a specific element (list.index()). (Can
someone tell me why index() and count() are not available for tuples?)

Even small reductions in cognitive load can yield significant benefits in
ease of use of a language.

Suppose Python had only one sequence data type: list, and that we could use
it in all places where tuples are needed now. I would never have to think
about whether to use tuple or list. I would never have to convert from tuple
to list or vice versa. Just one more thing I don't have to bother with.

I suspect there will be several answers offered as to why we need both
types, but I also suspect most of the benefits are for the Python
implementor(s), not programmers. Yes, there are probably a few times where
the ability to create an immutable sequence is useful, but is that enough to
warrant the additional complexity of a whole additional datatype? If
immutable sequences are so useful, then why not immutable mapping types,
etc. etc.?

Note that this message is coming from a Python fan, not a critic. This is
just an issue that has piqued my curiosity. Python is for the most part a
language where design choices have been in favor of ease of programming,
with few concessions to making life easier for language implementors. Having
both lists and tuples seems to be an exception.

-- Bob





From gregm at iname.com  Thu Apr 20 11:25:47 2000
From: gregm at iname.com (Greg McFarlane)
Date: Fri, 21 Apr 2000 01:25:47 +1000
Subject: A tkMessageBox question!
In-Reply-To: <38FDC335.B9FDCFE4@icestormfx.com>; from Sylvie Bruneau on 19 Apr 2000 at 10:31:18AM
References: <38FDC335.B9FDCFE4@icestormfx.com>
Message-ID: <20000421012547.10596@nms.otc.telstra.com.au>

Instead of using the built-in Tk dialogs you may like to try the Pmw
dialoges.  All options of all component widgets can be fully
configured.  For example:
    dialog = Pmw.MessageDialog(
	title = 'Simple message dialog',
	message_font = 'Times 12',
	defaultbutton = 0,
	message_text = 'A simple message dialog\nwith no callback.'
    )

The 'message_font' component option accesses the 'font' option of the
'message' component (which is a Tkinter.Label widget), so you can
reconfigure to your heart's content.

Pmw home page: http://www.dscpl.com.au/pmw/
    

On 19 Apr, Sylvie Bruneau wrote:
> Hi Everybody.
> 
>     I am using Python for an inhouse tool that runs on irix and WinNT.  I needed to
> popup some message dialogs so I used the dialogs provided by Tkinter, ie, the
> tkMessageBox.  On irix, the appearance of the dialogs are really ugly since the font
> is 'Times 18'.  I want to change the font of the dialogs but I can't find how I can
> do it !
> 
>     I look at the available options for the tkMessageBox at
> http://w1.132.telia.com/~u13208596/tkintrobook/standard-dialogs.htm.  None of them
> let me change the font.   Then I looked in the code of tkMessageBox.py to find that
> the dialogs are completely done in Tcl/Tk and the 'Times 18' is hardcoded in a file
> called msgbox.tcl.
> 
>     How can I change this font ?  Thank you for you help !
> 
> Sly
> 
> --
> /=-/=-/=-/=-/=-/=-/=-/=-/=-/=-/=-/=-/=-/=-/=-/=-/=-/=-/=-/=-/
> Sylvie Bruneau ( sylvieb at icestormfx.com ) Tel: (514) 527-3963
> Software Developer, IceStorm Digital.     Fax: (514) 527-5165
> 2595 Place Chasse, Montreal, Quebec, H1Y 2C3.
> 
> La vie est un cadeau parfois difficile a deballer. (P. Brassard)
> 
> 

-- 
Greg McFarlane     INMS Telstra Australia     gregm at iname.com



From warlock at eskimo.com  Sat Apr  8 00:50:36 2000
From: warlock at eskimo.com (Jim Richardson)
Date: Sat, 08 Apr 2000 04:50:36 GMT
Subject: opening a serial port at a given baudrate. How?
References: <200004050319.NAA08789@piglet.dstc.edu.au>  <38EB577C.E1EC5704@bioreason.com>
Message-ID: <slrn8etf2b.137.warlock@raven.myth>

On Wed, 05 Apr 2000 09:10:52 -0600, 
 Mitch Chapman, in the persona of <chapman at bioreason.com>,
 brought forth the following words...:

>
>
>David Arnold wrote:
>> 
>> -->"Jim" == Jim Richardson <warlock at eskimo.com> writes:
>> 
>>   Jim> I have been using open('/dev/ttyS0','r+') but need to control
>>   Jim> baud rate. fcntl.ioctl() will do this?
>> 
>> as someone else pointed out (sorry, lost the mail), the termios module
>> is the go for setting baud rates, parity, etc.
>
>Here's an old answer, courtesy  of www.deja.com:
>http://www.deja.com/[ST_rn=ps]/getdoc.xp?AN=309748544&fmt=text
>
>Search for "# Set the baud rate."
>
>--

Thanks, this is helpful :)


-- 
Jim Richardson
	Anarchist, pagan and proud of it
WWW.eskimo.com/~warlock
	Linux, because life's too short for a buggy OS.



From gresham at mediavisual.com  Fri Apr 14 14:30:32 2000
From: gresham at mediavisual.com (Gresh)
Date: Sat, 15 Apr 2000 02:30:32 +0800
Subject: problem with os.system calls under win32
References: <1256392132-57146878@hypernet.com>
Message-ID: <8d7o4d$c8h1@imsp212.netvigator.com>

Actually I was trying a similar thing, and decided to be smart and use
os.path.join
but this is what happened:

>>> import os
>>> os.system(os.path.join("D:","My Documents","test.bat"))
'D:\My' is not recognized as an internal or external command,
operable program or batch file.
1

However:
>>> os.path.exists(os.path.join("D:","My Documents","test.bat"))
1

That's a bit confusing !

Of course from Gordon's example:
>>> os.system('"%s"' % (os.path.join("D:","My Documents","test.bat")))
0

works

Regards,
Paul


"Gordon McMillan" <gmcm at hypernet.com> wrote in message
news:1256392132-57146878 at hypernet.com...
> william_93402 at my-deja.com wrote:
>
> > I am attempting to do a system call in a directory with spaces in the
> > name. For example:
> >
> > >>>import os
> > >>>os.system('c:\\f o o\\bar.exe')
> > 1
> >
> > when I place the executable in a dir such as c:\\foo I can successfully
> > execute the system call. Is there something special I have to do when
> > directory names contain spaces? Just for notes I am running by example
> > in build 125 of pythonwin.
>
> Follow MS's rules:
> os.system('"c:\\f o o\\bar.exe"')
>
> (must be double-quotes).
>
> - Gordon
>




From ullrich at math.okstate.edu  Sat Apr  1 14:07:35 2000
From: ullrich at math.okstate.edu (David C. Ullrich)
Date: Sat, 01 Apr 2000 13:07:35 -0600
Subject: Does Python have a long floating data type please?
References: <000201bf9b9b$32560e80$2f2d153f@tim>
Message-ID: <38E648F7.D4AB7F7F@math.okstate.edu>

Tim Peters wrote:

> [David C. Ullrich, notes some examples of dubious intervals computed
>  by Jurjen N.E. Bos's real.py module]

    One was dubious, the other was wrong. (Never mind...)

> Yes, and I've posted others in the past (including cases that aren't near
> 0).

    Musta been before my time.

> Jurjen used a very clever storage-efficient representation, but didn't
> record correspondingly clever correctness arguments <wink>.  Sometimes it's
> easy to see a bug, sometimes I have no idea what he was thinking.
>
> This isn't unusual for these kinds of pkgs, though!  Correctness proofs can
> be quite involved.  Val?rie M?nissier-Morain has submitted a 59-page article
> giving correctness proofs for her version of this quest, available somewhere
> more-or-less obvious under
>
>     http://pauillac.inria.fr/~menissier/
>
> > ... - at this point I decided to give up on real.py and make my own.
>
> Which is why there are a dozen packages like this out in the world each with
> its own set of unique subtle bugs <wink>.  Since this stuff runs too slow to
> use for truly large problems anyway, I never understood why pkgs try to
> conserve storage at the cost of vastly subtler reasoning.

    Exactly what convinced me that a real should be a pair of longRationals
(giving a lower bound and an upper bound for the "real" value). At least
the basic arithmetic operations seem simple enough that a person could
get them right, and they'd be obviously right.

> Guess it's seen
> as a challenge.
>
> personally-more-excited-by-things-that-are-obviously-right-than-by-
>     things-that-aren't-obviously-wrong-ly y'rs  - tim

class real:
  def __init__(self, l, u):
    self.l = l
    self.u = u
  def __add__(self, other):
    return real(self.l + other.l, self.u + other.u)

Pretty exciting stuff. (Well, the correctness of __mul__
might not be obvious at one glance like __add__ above,
because of what happens if l < 0 and u > 0, but it should
nonetheless be extremely trivial to verify.)

DU



From effbot at telia.com  Mon Apr 17 12:45:26 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Mon, 17 Apr 2000 16:45:26 GMT
Subject: Python flak???
References: <8dfdmq$v1o$1@nnrp1.deja.com>
Message-ID: <GcHK4.2615$Za1.39072@newsc.telia.net>

Arint? <jamarijr at hotmail.com> wrote:
> What are the reprecussions <sp?> of doing a PyErr_Clear() when there is
> no error?

also see:
http://www.python.org/doc/current/api/exceptionHandling.html

    void PyErr_Clear ()

    Clear the error indicator. If the error indicator is
    not set, there is no effect.

</F>




From MSteed at altiris.com  Tue Apr 25 13:24:57 2000
From: MSteed at altiris.com (Mike Steed)
Date: Tue, 25 Apr 2000 11:24:57 -0600
Subject: Don't understand error message
Message-ID: <65118AEEFF5AD3118E8300508B1248774CB2B4@ALTNET>

> From: Michael Hudson [mailto:mwh21 at cam.ac.uk]
> Sent: Tuesday, April 25, 2000 11:22 AM
> To: python-list at python.org
> Subject: Re: Don't understand error message
> 
> 
> "Lars Lundstedt" <lars.lundstedt at telia.com> writes:
> 
> > import sys, glob, operator
> > print sys.argv[1:]
> > sys.argv = reduce(operator.add, map(glob.glob, sys.argv))
> > print sys.argv[1:]
> > 
> > When called with "python glob.py *.py" on Win98 I'm 
> supposed to get a list
> > of all .py files in the current directory but instead I get:
> > 
> > Traceback (innermost last):
> >   File "glob.py", line 1, in ?
> >     import sys, glob, operator
> >   File "glob.py", line 3, in ?
> >     list = reduce(operator.add, map(glob.glob, '*.py'))
> > TypeError: call of non-function (type module)
> > 
> > Anyone know why?
> 
> "Works here".  Is that *really* your *entire* script?  You haven't
> snuck a module called "reduce" in anywhere?

It works on WinNT 4 too.  I see that the line that Python is complaining
about does not appear anywhere in the script in your message (close, but not
exactly).  Something is missing.
 
> Mind you, I'm on Linux - but that's shouldn't really affect this.
> 
> Cheers,
> M.
  ^^
(I thought that was my .sig)
M.



From M_Pope at shef.rage.co.uk  Mon Apr 10 08:41:10 2000
From: M_Pope at shef.rage.co.uk (Mark Pope)
Date: Mon, 10 Apr 2000 13:41:10 +0100
Subject: Extending built-in language (Win 95)
Message-ID: <213CDEBFDDA7D3118081009027E40C56071113@SHEF-MS1>

Hi,

I would like to extend the objects available in python, however It is not
appropriate to build normal extension dll's.

The file:
\python-1.5.3\objects\xxobject.c

Appears to be a template for defining new built in objects which will be
compiled into the Python static lib. However there must be other steps to
making the new object reachable from python script.


There are plenty of examples on how to make the normal DLL extensions, but
this doesn't help.
Can somebody point me at a tutorial for extending the built-in
functionality.

thanks,

mark.



From bobalex at home.com  Thu Apr  6 11:14:30 2000
From: bobalex at home.com (Bob Alexander)
Date: Thu, 6 Apr 2000 08:14:30 -0700
Subject: Tuples -- who needs 'em
References: <200004060137.LAA27607@nexus.csse.monash.edu.au>
Message-ID: <009a01bf9fda$ce66e540$74eb0b18@stcla1.sfba.home.com>

"Jonathan Giddy" <jon at dgs.monash.edu.au> wrote:

> I nearly fell out of my tree when Alexander, Bob declared:
>
> >I am wondering where the serious concern about mutable lists comes from,
as
> >it is a pattern in the responses I've gotten on this topic. For me, the
> >immutable list is a relatively new concept, since I can't think of many
> >other languages that have it. I'm a programming-language-design kind of
> >person, and I never really saw mutable lists as a problem that needed
> >solving.
>
> How about C?  Surely 'const' is an attempt to mark arrays as immutable.

Okay, that's true. Clearly my statement is exaggerated. Still, I can't
recall many instances where having only mutable sequences has caused me
trouble -- and this is over 30+ years of programming. And sometimes I pine
for the days before "const" was introduced  :-)

> Avoiding the x[i] = y construct is much harder in non-trivial programs,
> where assumptions about mutability need to be specified as part of the
call
> interface.  Hence the proliferation of const in the C string library.

Agreed -- any function that modifies a structure should spell that out
loudly in its interface description.

Programming languages make tradeoffs between easy coding vs. early error
detection. For example, the greater annotation required in C programs (e.g.
type declarations) makes coding harder, but allows detection of many errors
at compile time that would wait until run time in Python. In this case, the
distinction between tuples and lists is an "annotation" that allows us to
detect improper modification of a sequence earlier -- at the point of
attempted modification instead of later when it causes some other failure.
Most of Python opts for less annotation for easier coding vs. early error
detection -- having both mutable and immutable sequences is a move in the
other direction.

By the way, I personally prefer the low annotation level of Python-type
languages over C-type languages. It seems to me that programming is more
pleasant and overall the job gets done faster.

-- Bob





From M_Pope at shef.rage.co.uk  Tue Apr 18 11:29:54 2000
From: M_Pope at shef.rage.co.uk (Mark Pope)
Date: Tue, 18 Apr 2000 16:29:54 +0100
Subject: dir(..)   vs.   PyList_GetItem
Message-ID: <213CDEBFDDA7D3118081009027E40C56071155@SHEF-MS1>

Hi,

Can someone explain the difference between calling 'dir(moduleName)' from
python, and 'PyModule_GetDict( pModule )' from the C API.

I'd presumed they'd be the same thing, but appear to return different lists.


Details:

In Python, I enter the following:

	import myScript
	dir( myScript )

and receive the output:

	['myFunc1', 'myFunc2', '__builtins__', '__doc__', '__file__',
'__name__']


>From the C API, I call:

	PyObject *pModule, *pDictionary, *pKeys;
	pModule = PyModule_New( "myScript" );
	if( pModule )
	{
		pDictionary = PyModule_GetDict( pModule ) ;

		pKeys = PyDict_Keys( pDictionary );
	}

and the 'keys' list just holds:

	__doc__     and     __name__


I would like to have a list of the script functions available in C.


thanks,
mark.





From wayne.izatt at myself.com  Sat Apr  8 02:01:55 2000
From: wayne.izatt at myself.com (Wayne Izatt)
Date: Sat, 08 Apr 2000 06:01:55 GMT
Subject: MySQL-module
References: <55jH4.145$rD3.2312@news1.online.no>
Message-ID: <38EECB5D.328B141E@myself.com>

MySQLdb is the way to go. MySQLmodule is no longer well supported, and MySQLdb
adheres to the Python Database API 2. I've used it extensively, and you can do
anything you need to do, including creating you tables etc...


Thomas Weholt wrote:

> Hi,
>
> I`m going to try MySQL and want to access stuff using Python. At Python.org
> I found to links to two different modules, one written in python, the other
> C. Is there a special reason for choosing one over the other?
>
> I`d like to use all python, not compile anything cuz I`m scared python won`t
> work anymore. Had this problem earlier with re-compilation of python. zlib
> etc. didn`t work anymore.
>
> Any tips? I can create the actual databases/tables manually, but need to
> acces/insert/manipulate the data from a python-script.
>
> Thomas



From shapr at uab.edu  Sat Apr 29 06:24:44 2000
From: shapr at uab.edu (Shae Erisson)
Date: Sat, 29 Apr 2000 10:24:44 GMT
Subject: other patterns in python?
References: <8ebpl4$eq9$1@nnrp1.deja.com> <nXpO4.17469$GP3.124141@typhoon.nyroc.rr.com>
Message-ID: <390ABB26.3831DDD1@uab.edu>

Darrell wrote:
> 
> <tiddlerdeja at my-deja.com> wrote in message
> > Anyone got an example of a Command pattern/framework that they have
> > used in python?
> >
> 
> This pattern wraps an action or command in an instance with an execute
> method.
> The instance can be passed around like a function pointer.

<snip>

Has anyone else written pattern demos in python?
I'd like to see more if they're out there.
-- 
sHae mAtijs eRisson (sHae at wEbwitchEs.coM) gEnius fOr hIre
 		blah. mornings suck.


From djc at itga.com.au  Wed Apr 12 03:07:11 2000
From: djc at itga.com.au (Dave Cole)
Date: 12 Apr 2000 17:07:11 +1000
Subject: Sybase module - release 0.5
Message-ID: <873dor3ihs.fsf@heresy.itga.com.au>

Sybase is a module which provides a Python interface to the Sybase
relational database system (surprise). The Sybase package supports
almost all of the Python Database API, version 2.0 with extensions.

The Sybase package is designed to work with Python versions 1.5.2 and
later and Sybase versions 11.0.3 and later. It is based on the Sybase
Client Library (ct_* API), and the Bulk-Library Client (blk_* API)
interfaces.

The module contains the following items: 

      Module interface 
      Connection objects 
      Cursor objects 
      Bulkcopy objects 

The homepage http://www.itga.com.au/~djc/sybase.html contains a
description of the deviations from the Database API specification.

The module can also be downloaded from
        ftp://ftp.itga.com.au/pub/python/Sybase-0.5.tar.gz

Changes in this release:

        - DST bug fix.  datetime fields were always converted to
          daylight saving time for some bizarre reason ('cos I am an
          idiot).
        
        - Numeric columns with scale == 0 are now retrieved as Python
          longs.
        
        - You can bulkcopy Python long values into numeric columns
          when the numeric has scale == 0.

- Dave


From schmitthead at my-deja.com  Wed Apr 26 02:37:11 2000
From: schmitthead at my-deja.com (schmitthead at my-deja.com)
Date: Wed, 26 Apr 2000 06:37:11 GMT
Subject: huh? Python says that AttributeError: 'string' object has no attribute 'split'
Message-ID: <8e62qm$511$1@nnrp1.deja.com>

I received a whole folder full of files with MacOS type text files, with
different newline characters.  I thought, a perfect job for python.
Below is my attempt, but it generates the following error message:

>mac2dos tt_prvt.h
Traceback (innermost last):
  File "C:\bin\mac2dos.py", line 33, in ?
    mac2dos( directory )
  File "C:\bin\mac2dos.py", line 24, in mac2dos
    os.path.walk( os.getcwd(), callbackfunc, '' )
  File "d:\python\Lib\ntpath.py", line 265, in walk
    func(arg, top, names)
  File "C:\bin\mac2dos.py", line 18, in callbackfunc
    mac2dosfile( file )
  File "C:\bin\mac2dos.py", line 9, in mac2dosfile
    lines = myline.split( myline, chr(10) )
AttributeError: 'string' object has no attribute 'split'


#mac2dos.py-------------------------------------------------------------
import os
import sys
import string


#-----------------------------------------------------------------------
def mac2dosfile( file ):
 myline = file.read()
 lines = myline.split( myline, chr(10) )
 for l in lines:
  l = l.strip( l )
 file.writelines( lines )

#-----------------------------------------------------------------------
def callbackfunc( junk, dir, names ):
 for name in names:
  file = open( name, 'r+' )
  mac2dosfile( file )
  close( file )


#-----------------------------------------------------------------------
def mac2dos( dir ):
 os.path.walk( os.getcwd(), callbackfunc, '' )


#-----------------------------------------------------------------------
if __name__ == '__main__':
 if ( len(sys.argv) < 2 ):
  directory = os.getcwd()
 else:
  directory = sys.argv[1]
 mac2dos( directory )
#-end-of-mac2dos.py-----------------------------------------------------


Sent via Deja.com http://www.deja.com/
Before you buy.


From rumjuggler at home.com  Fri Apr 21 14:14:36 2000
From: rumjuggler at home.com (Ben Wolfson)
Date: Fri, 21 Apr 2000 18:14:36 GMT
Subject: docstrings
References: <00bufsonasp7bgkf86lfb6sv5np5v8mqlj@4ax.com> <8dnasp$iep$1@news.wrc.xerox.com>
Message-ID: <6v61gs8b1op8esg3o7hf0hii97jlvma01r@4ax.com>

On 20 Apr 2000 16:22:49 GMT, mjackson at wc.eso.mc.xerox.com (Mark
Jackson) wrote:

>Ben Wolfson <rumjuggler at home.com> writes:
>> is there a way to use %c and %s within a docstring, or to add strings
>> together in a docstring, so that if some strings are used in all
>> docstrings, but might be changed, I can simply change one or two
>> variables?
>
>This might do what you want, although it's ugly (in the sense that it
>serves the code-documentation - as opposed to user-documentation -
>function less well):
>
>    cat > spam.py
>    """I am a docstring"""
>    
>    sharedoc = """ and I am a widely-reused docstring clause"""
>    
>    __doc__ = __doc__ + sharedoc
>    
>    del sharedoc
>    ^D
>    yngvi> python
>    Python 1.5.2 (#7, May  6 1999, 14:39:45)  [GCC 2.8.1] on sunos5
>    Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>    >>> import spam
>    >>> print spam.__doc__
>    I am a docstring and I am a widely-reused docstring clause
>    >>> 

This approach doesn't work for function docstrings, unless I tried it
wrong.

-- 
Barnabas T. Rumjuggler

He still loves life,
But O O O O how he wishes
The good Lord would take him.
 -- W. H. Auden


From cgw at fnal.gov  Fri Apr 14 16:42:19 2000
From: cgw at fnal.gov (Charles G Waldman)
Date: Fri, 14 Apr 2000 15:42:19 -0500 (CDT)
Subject: Memory leaks in Python 1.6
Message-ID: <14583.33451.120821.695270@buffalo.fnal.gov>

I've been running a small script I wrote called "leaktest" (attached
below) which runs tests from the the Python/Lib/test directory in a
loop, while watching the memory use of the process.  I've found quite
a few memory leaks this way, and am working on fixing them.  I will be
submitting a bunch of patches to the patches list over the next
several days.




Script: leaktest.py

#!/usr/bin/env python

"""
Runs tests repeatedly, looking for memory leaks
If you have gnuplot installed you can also get nice plots of memory
usage vs. time

Usage:  leaktest [-max max] [-min min] [-check check] [testname] [testname]...

runs one or more tests repeatedly, looking for memory leaks
The tests can be specified as module name (test_gzip) or file name
(/path/to/test_gzip.py)

memory usage is checked every "check" iterations of the test.

at least "min" readings are taken, and at most "max" readings.  If no
memory leak is detected after "min", the test stops.

If "DO_LOG" is set to 1, memory usage will be logged to a file.
If "DO_PLOT" is set to 1, pretty pictures will be drawn.
If "VERBOSE" is set to 1, memory usage will be printed during the test
"""


import sys
import os
import time
import string

DO_LOG=1
DO_PLOT=1
VERBOSE=0

def open_file(name):
    if os.path.exists(name):
        os.system("cp -b %s %s.bak"%(name,name))
    f=open(name,'w')
    return f

def mem_usage(pid):
    p=os.popen('ps uwp %s'%pid)
    lines=p.readlines()
    status=p.close()
    if status or len(lines)!=2:
        return None
    return int(string.split(lines[1])[4])

def no_increase(l):
    for i in xrange(1,len(l)):
        if l[i]>l[i-1]:
            return 0
    return 1
        
def run_test(testname, min_readings=10, max_readings=50, check_interval=10):
    print "RUN %s min %d max %d check %d"%(testname,min_readings,
                                           max_readings, check_interval)
    if DO_LOG:
        log = open_file('%s.log'%testname)
    logdata=[]
    ret=1
    pid=os.getpid()

    stdout, stderr = sys.stdout, sys.stderr
    sys.stdout, sys.stderr = open_file('%s.out'%testname), open_file('%s.err'%testname)
    
    for loop_readings in xrange(max_readings):
        mem = mem_usage(pid)
        if not mem:
            break
        if VERBOSE:
            stdout.write('%s\n'%mem)
        logdata.append(mem)
        if DO_LOG:
            log.write('%s\n'%mem)
            log.flush()

        for n in xrange(check_interval):
            try:
                if loop_readings==0:
                    exec "import test.%s"%testname
                else:
                    exec('reload(test.%s)'%testname)
            except:
                ret=-1
                break

        if ret<0:
            break

        if len(logdata)>=min_readings and no_increase(logdata[-min_readings:]):
            ret=0
            break

    sys.stdout.close()
    sys.stderr.close()
    sys.stdout, sys.stderr = stdout, stderr
    
    if DO_LOG:
        log.close()

    return ret, logdata
            
def do_plot(testname):
    cmdfile=open('%s.gnuplot' % testname, 'w')
    template='set nokey\nset xlabel "Iterations"\nset ylabel "Memory Use (KB)"\n\
                              set title "%s\nplot "%s" with lines\npause 60\n'
    cmdfile.write(template % (testname, testname+'.log'))
    cmdfile.close()
    os.system("gnuplot %s.gnuplot &"%testname)
    
def main(args):
    check_interval=10
    min_readings=10
    max_readings=50
    while args:
        arg = args.pop(0)
        if arg=='-min':
            min_readings = int(args.pop(0))
            continue
        if arg=='-max':
            max_readings = int(args.pop(0))
            continue
        if arg=='-check':
            check_interval = int(args.pop(0))
            continue
        ## allow the test module names to be specified as full pathnames
        ## for ease in calling this from shellscripts
        arg = os.path.split(arg)[1]
        arg = os.path.splitext(arg)[0]
        status, data=run_test(arg, min_readings, max_readings, check_interval)
        if status<0:
            print "test", arg, "did not run"
        elif status==0:
            print arg, "no memory leak detected", data
        else:
            print arg, "leaks memory", data
            if DO_PLOT:
                do_plot(arg)
    
if __name__ == '__main__':
    main(sys.argv[1:])
    
       



From ivanlan at callware.com  Tue Apr  4 18:36:55 2000
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Tue, 04 Apr 2000 16:36:55 -0600
Subject: Two questions
References: <38EA5FE8.853E9B75@callware.com> <38EA6508.90BF630C@roguewave.com>
Message-ID: <38EA6E87.40EDB542@callware.com>

Hi All--

Bjorn Pettersen wrote:
> 
[bobbit]
> 
> If you want to package up your program into an exe and execute it,
> Installer from Gordon's starship pages can do this
> (http://starship.python.net/crew/gmcm/distribute.html).  It's actually
> _much_ easier than it looks ;-)
> 

<g>  This sounds like the right course to me.  I'd like to force
customers to install Python on their machines, but my boss doesn't think
that's all that great an idea (dunno why).  My only concern is
size--just how big are the executables produced by Gordon's
flummery^H^H^H^H^H^H^H^Hmagic?

> 
> You'll probably want to check out one of the odbc packages.  (I haven't
> used any of them myself, but mxODBC seems to be highly recomended...)
> 

Well, that's two recommendations.  And judging from previous Marc-Andre
packages I've used, it'll be a snap.

Steve Holden, in another post, said that Tkinter should work fine along
with mxODBC.  That's a relief, but does anyone have direct personal
experience with doing so?

<floccinausinihilipilification>-ly y'rs,
Ivan

(I think I've spelled that right--I don't have the OED online.)
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
http://www.pauahtun.org 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
Author:  Teach Yourself Python in 24 Hours



From ivanlan at callware.com  Tue Apr 18 17:44:22 2000
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Tue, 18 Apr 2000 15:44:22 -0600
Subject: [Fwd: o'reilly python devcenter]
Message-ID: <38FCD736.13282680@callware.com>

Hi All--
Dennis Hamilton, in this forwarded letter, has kindly corrected my
misapprehensions regarding O'Reilly's Python Devcenter.  However, he
replied only to me, so I attach his response for your collective
delight.

But, regarding point 3 (books), Dennis says:

> 3) Why would they do that?  You can do that and perhaps be seen to have
> objectivity in providing it.  I would not expect O'Reilly to promote the
> books of others.  Would you?  (Expect it or use your own site to do it?)
> There is already python.org and starship, yes?

Actually, I _would_ expect O'Reilly to well, not exactly promote, but
certainly _mention_ worthwhile books from other publishers.  They can
certainly afford it, and it would do nothing except bolster their
already exemplary reputation.


Dennis also asks,

> Regarding hard sell, what were you looking at?

Well, since I was so wrong about the other points, this really doesn't
hold water anymore.  My opinion that it was "hard sell" was based on the
(incorrect) presumption that they were trying to control everything
Python.  That made me cranky.  Since they're not, I withdraw all
complaints but number three (which is a vestige of its former self).

<now-where'd-i-put-that-recipe-for-veggie-crow-pie?>-ly y'rs,
Ivan
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
http://www.pauahtun.org 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
Author:  Teach Yourself Python in 24 Hours
-------------- next part --------------
An embedded message was scrubbed...
From: "Dennis E. Hamilton" <infonuovo at email.com>
Subject: RE: o'reilly python devcenter
Date: Tue, 18 Apr 2000 12:04:50 -0700
Size: 4255
URL: <http://mail.python.org/pipermail/python-list/attachments/20000418/0dbb2e75/attachment.mht>

From bragib at my-deja.com  Mon Apr  3 13:58:50 2000
From: bragib at my-deja.com (bragib at my-deja.com)
Date: Mon, 03 Apr 2000 17:58:50 GMT
Subject: Text editing/replacing
Message-ID: <8cam4g$gdj$1@nnrp1.deja.com>

Are there any good algorithms available for text editing/replacing
of words in a huge text file.

If anyone knows of any sample codes, books, algorithms etc.
I would appreciate it.

Thanks, Bragi


Sent via Deja.com http://www.deja.com/
Before you buy.


From mjhand at concentric.net  Fri Apr 14 11:44:26 2000
From: mjhand at concentric.net (Manus Hand)
Date: Fri, 14 Apr 2000 09:44:26 -0600
Subject: starship.python.net
References: <Pine.OSF.4.10.10004111721150.9584-100000@prfdec.natur.cuni.cz> <Pine.OSF.4.10.10004111740110.11294-100000@prfdec.natur.cuni.cz>
Message-ID: <38F73CDA.40958FEE@concentric.net>

python.net is once again down -- this time for over a day.  Anyone have any
prospects for us?

Manus




From dgoodger at bigfoot.com  Sat Apr  1 10:54:50 2000
From: dgoodger at bigfoot.com (David Goodger)
Date: Sat, 01 Apr 2000 10:54:50 -0500
Subject: Why should I switch to Python?
References: <seas5ia89177@corp.supernews.com>
Message-ID: <B50B85F9.10E9%dgoodger@bigfoot.com>

on 2000-03-31 22:43, Aaron Turner (aturner_ at _pobox.com) wrote:
> What I'd like to hear from people is what technical
> resons I should switch to Python.

For me, readability is the ultimate factor. I've written thousands of lines
of Perl code, and I still had trouble understanding others' code, and even
my own! (Even though I tried to write readable stuff, revisiting code from
weeks or months earlier was an exercise in frustration.)

Python works the way I do, and that's why I like it. It forces readability
through its syntax rules (block start/end via indentation, etc.), which I
followed anyhow; now I don't have to worry about missing semicolons.

I love Perl for what it is, but now that I've used Python (about 1.5 years
now), I can't imagine going back. Functionally, they both do the job. But
Python's "interface" suits me better.

Isn't it great that there are so many tools out there? Everyone can find one
that suits their style.

-- 
David Goodger    dgoodger at bigfoot.com    Open-source projects:
 - The Go Tools Project: http://gotools.sourceforge.net
 (more to come)



From tismer at tismer.com  Fri Apr 28 10:19:46 2000
From: tismer at tismer.com (Christian Tismer)
Date: Fri, 28 Apr 2000 16:19:46 +0200
Subject: Simple list.append() question
References: <meh9-5ED8CD.11192424042000@news.cit.cornell.edu> <ZHZM4.360$qg7.2949@news.uswest.net> <39046FA2.603EAFE5@roguewave.com> <PsaN4.174166$Dv1.2059732@news1.rdc1.bc.home.com> <8e4akc$p7o$1@newshost.accu.uu.nl> <m366t6ozqx.fsf@greebo.nodomain.de> <3908E6EC.B9E6B2E8@cosc.canterbury.ac.nz> <m3zoqenzt6.fsf@atrus.jesus.cam.ac.uk> <39099069.7BFD8387@tismer.com> <m3puranyyp.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <39099E02.E68483E7@tismer.com>


Michael Hudson wrote:
> 
> Christian Tismer <tismer at tismer.com> writes:
> > > Greg Ewing <greg at cosc.canterbury.ac.nz> writes:
> > > > Maybe the LC syntax should include something like
> > > >
> > > >    [[] times 10]
> > >
> > > But wouldn't that make "times" a keyword?  Not sure I'd want that...
> >
> > Currently, yes, but it is absolutely not necessary to be so.
> > It requires to change the grammar to treat operators
> > differently.
> 
> But then what happens to things like
> 
>   times = 10
>   return [[] times times]

No problem for me. I wouldn't write this, for clarity.
No reason to forbid it, anyway.

> ?  I'm not sure context senstive keywords are a good idea (in any
> language, ever).  Currently a naked identifier in Python is either a
> variable reference or a keyword, and you can tell which just by
> looking at *it*, not at its surroundings - and I think I like that.
> 
> Though it would be occasionally useful to be able to call a variable
> `class'.

Well, we would get into trouble anyway, if keywords appear
in other contexts but as attributes. self.class would be ok,
but class as a parameter, and you can't access it.

Algol68's-bold-stropping-had-its-benefits - ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaunstr. 26                  :    *Starship* http://starship.python.net
14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     where do you want to jump today?   http://www.stackless.com



From thomas at cintra.no  Sat Apr 22 13:36:20 2000
From: thomas at cintra.no (thomas at cintra.no)
Date: Sat, 22 Apr 2000 17:36:20 GMT
Subject: Sax installed, but not working
References: <38ff5f04.1047500@news.c2i.net> <3dya68o3yb.fsf@amarok.cnri.reston.va.us>
Message-ID: <3901df21.7429914@news.c2i.net>

On 20 Apr 2000 17:30:36 -0400, akuchlin at mems-exchange.org (Andrew M.
Kuchling) wrote:

>thomas at cintra.no writes:
>> from xml.sax import *
>> 
>> it fails and says there`s no module called sax. All the demo-scripts
>> and tutorials I`ve tried fails too.
>
>Check that xml/__init__.py and xml/sax/__init__.py exist.  (They might
>be of length 0, which is OK.)  Does 'import xml' work?
>
>-- 
>A.M. Kuchling			http://starship.python.net/crew/amk/
>I wrote my name, but I can find it no longer; my ashes blow like dust around
>the invisible labyrinth.
>  -- Zatara, in BOOKS OF MAGIC #1
>
>

Hi,

__ini__.py in both folders. 

other stuff seem to work, like xml.dom etc. but not xml.sax. Hm ...
can`t see anything wrong.

Ex. 'from xml.sax import saxlib' will fail.

Anything else I can check?

Thomas




From scott at csihq.com  Mon Apr 24 11:35:38 2000
From: scott at csihq.com (Scott Bortman)
Date: Mon, 24 Apr 2000 11:35:38 -0400
Subject: Need Help w/JPython deployment
Message-ID: <sg8pqclrcdo26@corp.supernews.com>

Hello,

I am having a bit of difficulty using jpythonc.

I have a script I want to deploy and I want to
have everything it needs in a single jar file.  This
is because JPython is NOT on the machine I want
to run the script (but Java is).

Here's the options I've tried using:

--deep
--all
--compiler <path to my javac>
--jar <target jar filename>

And then the name of my script:

<scriptname>.py

But, I get the following error:

processing npp_lab_status_mai

Required packages:
  java.lang
  java.text*
  java.sql*
  java.util*

Creating adapters:

Creating .java files:
  npp_lab_status_mai module

Compiling .java to .class...
Compiling with args: ['h:\\jdk1.3\\bin\\javac', '-classpath',
'H:\\JPython-1.1\\jpython.jar;I:\\JClass36J\\lib\\jcchart361J.jar;I:\\JClass
36J;;h:\\jbuilder3\\beans\\classes12.zip',
'.\\jpywork\\npp_lab_status_mai.java']
0
Building archive: npp_lab_status_mai.jar
Tracking java dependencies:
Traceback (innermost last):
  File "H:\JPython-1.1\Tools\jpythonc2\jpythonc.py", line 5, in ?
  File "H:\JPython-1.1\Tools\jpythonc2\main.py", line 291, in main
  File "H:\JPython-1.1\Tools\jpythonc2\main.py", line 284, in writeResults
  File "H:\JPython-1.1\Tools\jpythonc2\jar.py", line 86, in dump
  File "H:\JPython-1.1\Tools\jpythonc2\jar.py", line 80, in dumpFiles
  File "H:\JPython-1.1\Tools\jpythonc2\jar.py", line 94, in addOneClass
AttributeError: java package 'org.python.modules' has no attribute
'__path__'


Is there something I missed?

Thanks in advance,
scottie





From gdeeringNONEOFTHAT at acslink.SPAMTHANKS.aone.net.au  Wed Apr  5 05:00:36 2000
From: gdeeringNONEOFTHAT at acslink.SPAMTHANKS.aone.net.au (Geoff Deering)
Date: Wed, 05 Apr 2000 09:00:36 GMT
Subject: map.yahoo.com uses Python?
References: <Pine.SO4.4.02.10002132333440.18438-100000@ub.d.umn.edu>
Message-ID: <38eb00af.8905516@news.mel.aone.net.au>

I've heard all of Yahoo is Python based.




From JoeSmith at bogusaddress.com  Sat Apr 15 16:18:46 2000
From: JoeSmith at bogusaddress.com (Joe Smith)
Date: Sat, 15 Apr 2000 20:18:46 GMT
Subject: TCP/IP protocols
References: <Pine.SGI.4.05.10004141855400.4276-100000@lemur.magnet.com>
Message-ID: <38F8CF63.40771176@bogusaddress.com>

I get a name error for p.  What type is p supposed to be?

John Mitchell wrote:

> On Sat, 15 Apr 2000, Thomas Wouters wrote:
>
> > On Fri, Apr 14, 2000 at 04:28:03PM -0600, Jeff Massung wrote:
> >
> > > I'm new to Python programming, but am very experienced in C/C++. Could
> > > someone point me to a good reference for TCP/IP programming with Python
> > > (pdf, book, anything :]) thanks.
>
> > If you're looking for some higher-level programming, see libraries like
> > httplib, ftplib, telnetlib, urllib, SocketServer, *HTTPServer, etc.
>
> These are actually just two libraries: client-side (urllib handles FTP,
> HTTP, local files), plus server-side (Socket/HTTPServer).
>
> For example, to fetch a web page is this:
>
> import urllib
> data = urllib.urlopen('http://www.magnet.com/')
>
> To get some information out, it's much more complicated: 3 lines:
>
> import htmllib, formatter
> p.feed(data)
> p.title
> # =>  'Magnet Interactive Communications'
>
> Show this to your local Java weenie.  He'll whine like a stuck pig!
>
> > Documentation for all these modules comes with the python source (in the
> > Doc/lib/ directory) in TeX format, with most binary distributions in some
> > form or another, or can be downloaded or browsed online at
> > http://www.python.org/doc/lib/ (There's an entire chapter, 11, dedicated to
> > 'Internet Protocols and Support'.)
>
> oh yeah.  But the above is your 60-second introduction.
>
> - j



From dalke at acm.org  Sun Apr 23 01:29:22 2000
From: dalke at acm.org (Andrew Dalke)
Date: Sat, 22 Apr 2000 23:29:22 -0600
Subject: The Simple Economics of Open Source
References: <14591.15630.192472.440375@goon.cnri.reston.va.us><3900611F.3B525BDA@cfdrc.com> <FtDG4z.774@world.std.com><slrn8g26tu.p7o.neelk@brick.cswv.com><14837CB4E98A72FE.8FD547BA6E928594.5A103CFE68327EB7@lp.airnews.net> <VLvUljAKNfA5EwPM@jessikat.demon.co.uk>
Message-ID: <8du1in$hth$1@slb7.atl.mindspring.net>

Robin Becker wrote:
>I don't think the pessimists amongst whom I count myself reject altruism
>totally. A purely altruistic action has benefit for others and none for
>the actor. Very few of the altruistic examples in the thread come close.


The one I came up with, the first time I heard this argument (freshman
year of college, I think), was the following:

  I'm walking down the street and there is a hole in my front
pocket.  I keep money in that pocket.  I have a hand in that pocket
because it's cold, and I'm jostling the money.  As I pass a beggar,
some money falls out through the hole.  It is altruistic or an
accident?
  The only way to find out is to ask me, and I can lie, so
there's no observational way to tell the difference.  Thus, I
considered the definition of "altruistic", when used in this
extreme case, to be useless.

                    Andrew
                    dalke at acm.org





From aek at mail1.stofanet.dk  Wed Apr 12 15:47:19 2000
From: aek at mail1.stofanet.dk (Anders Eggers-Krag)
Date: Wed, 12 Apr 2000 21:47:19 +0200
Subject: global variables - how??
References: <20000412175218.784841CE49@dinsdale.python.org>
Message-ID: <8d2jsh$3a9$1@news101.telia.com>

it is obvious not to use global variables except when the use of them is
obvious,
in my case I could otherwise send them along with every function, creating a
huge
overhead passing a million pointers around, using exactly the same amount of
memory,
or in fact more, than I would using globals.

I did figure out how to do it myself after a lot of trying, but as you said
it is ugly,
but this really anoys me, and I *was* starting to like python...


--
Anders Eggers - Krag


"Sean Blakey" <sblakey at freei.com> wrote in message
news:20000412175218.784841CE49 at dinsdale.python.org...
> On 12 Apr, William Park wrote:
> > On Wed, Apr 12, 2000 at 07:28:37PM +0200, Anders Eggers-Krag wrote:
> >> how do I define a propper global variable that works in all functions
in all
> >> modules?
> >
> > As far as I know, you can't.  You have to use 'global ...' inside all
> > your functions.
> >
> > --William
> >
> >>
> >> I have tried simply
> >> defining one outside of any function in the main module,
> >> I have tried using the global keyword under the same cicumstances,
> >> but they are not even avialible inside of functions within the main
> >> module...
> >>
> >>
> >> --
> >> Anders Eggers - Krag
> >>
> >>
> >> --
> >> http://www.python.org/mailman/listinfo/python-list
> >
>
> Short answer: Python discourages globals.  This is probably intentional.
>
> Long answer:  You can do some work with globals, but it is painful and
> ugly.  You also need to keep track of namespaces (globals from your
> main module will be in the __main__ namespace).
>
> For example:
> -----file one.py-----
> TEST_VALUE = 'spam'
>
> def printTest():
>     print TEST_VALUE
>
> def changeTest():
>     global TEST_VALUE # Without this, the next line
>     # will simply assign 'spamspam'
>     # to the local variable
>     # TEST_VALUE
>     TEST_VALUE = TEST_VALUE + TEST_VALUE
>
> if __name__ == '__main__':
>     import two
>     printTest()
>     two.printTest()
>     changeTest()
>     printTest()
>     two.printTest()
>
> -----file two.py-----
> import __main__
>
> def printTest():
>     print __main__.TEST_VALUE
>
>
> Now, from the shell:
> seanb at seanb:~$ python one.py
> spam
> spam
> spamspam
> spamspam
>
>
> --
>
> Sean Blakey
> FreeInternet.com
> sblakey at freei.com
> (253)796-6500x1025
>
>




From m.faassen at vet.uu.nl  Wed Apr 12 13:04:08 2000
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 12 Apr 2000 17:04:08 GMT
Subject: Why should I switch to Python?
References: <seas5ia89177@corp.supernews.com> <slrn8f4d0j.1t0.bwinton@tor.dhs.org>
Message-ID: <8d2aa8$n58$1@newshost.accu.uu.nl>

Blake Winton <bwinton at tor.dhs.org> wrote:
> On Fri, 31 Mar 2000 19:43:58 -0800, Aaron Turner wrote:
>>I've been coding Perl for over 3 years and really like it.  (Especially
>>it's syntax and ability to grok my programming style.)

> Now, just as a side note, why is it that the sets of The Perl Hackers I
> Know, and The People I Know Who Use The Word "Grok" are the same set?

It's in the hacker's dictionary:

http://www.tuxedo.org/~esr/jargon/html/entry/grok.html

And common in hacker slang (from the science fiction world).

Actually, Python 3000 will contain the 'grok' statement (I posted
about this long ago). It works like this:

grok:
   <arbitrary indented text>

So, you can do things like this:

grok:
    Hello, please give me a web page out of the
    financial reports of the third quarter 1999. It should be fast!

And your Python program will grok it and do it.

In fact, I lied: the indentation rules in the 'grok' statement can be broken
and Python 3k will grok that too:

if a == 1:
    grok:
 Hello this isn't really indented right
  but Python can figure it out
anyway...give me a 3d pacman game if I press the 'pacman' button
    a = a + 1 # continue suite with normal indentation
    
But the recommendation is not to do this as it will result in unreadable
code. Wrong indentation may not work in future versions of Python 3k 
either, as it's an implementation detail -- if you use it anyway expect
the effbot to snip at you. Then again, the timbot will release 
groknanny to fix this, so you are probably safe. It'll look like this:

(grokindent.py)
grok:
    Please fix the indentation of grok code in all
    files which have wrong indentation of grok code.
    now-this-is-easy-ly y'rs  - tim

As you see, it has all the usual timbot elegance and charm. :)

Yes-I've-been-using-the-time-machine-ly yours,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From effbot at telia.com  Sun Apr 23 12:31:37 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Sun, 23 Apr 2000 16:31:37 GMT
Subject: Posting CGI
References: <8dv4j8$2bs$1@bob.news.rcn.net>
Message-ID: <JzFM4.3672$Za1.53909@newsc.telia.net>

Z 3 Penguin <z3penguin at penguinpowered.com> wrote:
> How would I go about submiting a CGI post form in python?  (This sounds
like
> it should be simple, but I can't figure it out :-) Thanks.

http://www.python.org/doc/current/lib/module-urllib.html

    import urllib

    request = urllib.urlencode({
        "name": "effbot",
        "email": "effbot at telia.com"
    })

    fp = urllib.urlopen("http://www.spam.egg", request)

    response = fp.read()

</F>

<!-- (the eff-bot guide to) the standard python library:
http://www.pythonware.com/people/fredrik/librarybook.htm
-->




From effbot at telia.com  Tue Apr 11 11:39:52 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Tue, 11 Apr 2000 15:39:52 GMT
Subject: AF_INET error
References: <8cvero$g8k$1@porthos.nl.uu.net>
Message-ID: <cHHI4.1202$Za1.20911@newsc.telia.net>

Soundwave <maybe at lucky.com> wrote:
> I want to learne some things about socket programming, so I started the
> socet programming tutorial on www.python.org
>
> but when I copy te code from some sample programm's and run it, it doesn't
> work, it don't recognize the AF_INET thing, is this because I am using
> Windows ??
>
> can somebody please explain this to me ?
>
> Or what am I doeing wrong ?

well, since you forgot to tell us what example you tried to
run, how you typed it in, and what error you got, I can only
guess.

-- make sure to import the socket module

-- make sure to use the "socket." prefix, so Python
   knows where to look for the AF_INET "thing" (it's
   a variable).

</F>




From matt at mondoinfo.com  Mon Apr 24 20:25:54 2000
From: matt at mondoinfo.com (Matthew Dixon Cowles)
Date: Mon, 24 Apr 2000 19:25:54 -0500
Subject: sys.exit
References: <3904C35C.B16BDE75@mindspring.com>
Message-ID: <240420001925542853%matt@mondoinfo.com>

In article <3904C35C.B16BDE75 at mindspring.com>, Chuck Esterbrook
<echuck at mindspring.com> wrote:

> The short question:
> 
> If I catch a SystemExit exception (because of an invocation of
> sys.exit(code)), how do I examine the argument to sys.exit()?

[snip]

try:
  sys.exit(42)
except SystemExit,detail:
  print detail

Getting access to many exceptions' arguments works similarly.

Regards,
Matt


From spex66 at my-deja.com  Fri Apr 28 12:22:36 2000
From: spex66 at my-deja.com (spex66 at my-deja.com)
Date: Fri, 28 Apr 2000 16:22:36 GMT
Subject: Namespace Problem with global declaration in module
References: <1255280660-66425116@hypernet.com>
Message-ID: <8ecdro$5gf$1@nnrp1.deja.com>

In article <1255280660-66425116 at hypernet.com>,
  gmcm at hypernet.com wrote:
[snip]
> > I see the point, but I like the from statement :)
>
> Tough :-). import * is mostly a shortcut for interactive work. Look
at the
> top of Tkinter.py to see what a module needs to do to make import *
> safe.
>
> > Is it a bug?
>
> No.
>
> > It's not the way I expected this code to work... and
> > I cannot find a hook in the documentation that explicit explained
> > this misbehaviour.
>
> It's there, in 6.11 of the Language Ref:
>
> The from form does not bind the module name: it goes through the
> list of identifiers, looks each one of them up in the module found in
> step (1), and binds the name in the local namespace to the object
> thus found. If a name is not found, ImportError is raised. If the
list
> of identifiers is replaced by a star ("*"), all names defined in the
> module are bound, except those beginning with an underscore ("_").
>
> No, the implications of that statement are not immediately obvious.
>
> > Any idea for a workaround (I need the from-statement for further
> > code readability)?
>
> No, you want it for type-ability. Come back to it in a few months and
> you'll wish you'd spelled out where all those names came from.
Not only, it is for automated code-generation *and* simple interactive
work :-)

I've found a solution working for me, I posted it in my other actual
thread "How can I get the name of an object???" 28/4/00 17:20

probably it's getting a little bit clearer why I'm riding that horse :)

thanks for your input
Peter
(=PA=)


Sent via Deja.com http://www.deja.com/
Before you buy.


From chega_ at yahoo.com  Fri Apr 21 03:29:02 2000
From: chega_ at yahoo.com (Vadim Chugunov)
Date: Fri, 21 Apr 2000 07:29:02 GMT
Subject: __del__ problem - would adopting Garbage Collection fix this?
References: <aw2L4.8907$HG1.245743@nnrp1.uunet.ca><7ojL4.9424$HG1.254543@nnrp1.uunet.ca><Z%oL4.9643$HG1.257533@nnrp1.uunet.ca> <l03102802b5245023a247@[193.78.237.147]>
Message-ID: <2rTL4.887$a6.294989@paloalto-snr1.gtei.net>

"Just van Rossum" <just at letterror.com> wrote in message
news:l03102802b5245023a247@[193.78.237.147]...
> >"Warren Postma" <embed at geocities.com> writes:
> >I don't think so.  You still have to destroy things; true gc might
> >help I suppose, but I'm fairly sure it doesn't solve the problem
> >completely.
>
> I'm not sure if I understood it correctly, but I think the newly proposed
> garbage collector does *not* call __del__ if the object in question is part
> of a cycle. It seems indeed impossible to do so correctly. I think the site
> explained quite well why this is, but I can't seem to repeat it off the top
> of my head.

If you are not going to settle for anything less than a perfect solution, than
this is indeed impossible.
On the other hand, I do remember Python prophets saying that "practicality beats
purity" :-)).

Ask yourself: "What kind of stuff do I typically do in __del__ methods?".
Close files, close sockets, write "Good-bye" to stdout... SWIG generated code
would call a destructor of the wrapped object.
And, I think, that's about it.

Note, that __del__ does not usually involve access to sibling objects
participating in the cycle.  (assumption)
IMHO, this means that we can:
1. pick any PyInstanceObject that comes along and call it's __del__ method.
2. change object's __class__ to, say, DeletedObject (which is defined as   class
DeletedObject: pass)
3. clean out instance __dict__, which will hopefully break the cycle.
4. repeat as needed.
If anyone *does* break the above assumption, well, life is tough...  They will
get an AttributeError exception.
Actually, step 2 above is optional.  It is there just to make possible an
explicit check for deleted objects.

Vadim



From gmcm at hypernet.com  Tue Apr 18 08:47:55 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Tue, 18 Apr 2000 08:47:55 -0400
Subject: Using python on the web
In-Reply-To: <MPG.1365adcfa8470eef9896aa@news>
Message-ID: <1256060415-19523925@hypernet.com>

Charley Horse  wrote:

> I've just started fooling with python. I've looked around the python 
> site and a number of the other sites referenced there and am a bit 
> puzzled. Umm... how does one go about using python for web applications, 
> especially database driven apps? Almost everything deals with python 
> just as a language, as C++ would be. Not much about the web 
> specifically. I was a bit shocked to find the only direct refs on about 
> this at python.org are entwined with CGI (unless I missed it). I have no 
> exposure to CGI, but rather to php, CF, and ASP, so python has gone from 
> looking very attractive to looking... well, like a puzzle. I know about 
> Zope, but am interested in how one uses python on the web just by 
> itself.

ASP, php and CF are all descended from CGI. By creating a 
CGI framework and / or putting hooks into the web server 
(there are a number of pure Python ones to play with) you can 
create environments that look like any one of these. Almost 
any web server will run CGI (with sometimes a bit of tweaking) 
so the docs focus on that. At the other end of the scale, Zope 
does everything those guys do, and usually better. IIS is 
widespread enough that there's a significant Python 
subculture doing ASP in Python. But the grandaddy of them 
all (and still the only portable one) is CGI.

- Gordon



From brent.fulgham at xpsystems.com  Thu Apr 13 17:40:36 2000
From: brent.fulgham at xpsystems.com (Brent Fulgham)
Date: Thu, 13 Apr 2000 14:40:36 -0700
Subject: Module Loading Issues
Message-ID: <EDFD2A95EE7DD31187350090279C67677079B4@THRESHER>

Is there any disadvantage to having many modules loaded
at the same time?  I.e., is it possible for badly-written
modules to write entries into the "__main__" namespace
that might clash with other entries?

Or is the only concern memory use?

Thanks,

-Brent



From spex66 at my-deja.com  Wed Apr 26 11:13:41 2000
From: spex66 at my-deja.com (spex66 at my-deja.com)
Date: Wed, 26 Apr 2000 15:13:41 GMT
Subject: How can I get the name of an object???
Message-ID: <8e712k$65c$1@nnrp1.deja.com>

Hi again,

other trivial (looking) question:

>>> class zz: pass
>>> dd = zz()

I cannot find the hook to get back the name of 'dd' as a string.
Why? Cause I need the names for code generation :-)

like estimated
>>> dd.__name__
>>> 'dd' #DON'T WORK

any hacks available?

Peter
(=PA=)


Sent via Deja.com http://www.deja.com/
Before you buy.


From smalleys at gte.net  Wed Apr 12 23:09:19 2000
From: smalleys at gte.net (sue)
Date: Thu, 13 Apr 2000 03:09:19 GMT
Subject: pattern-based string expansion
Message-ID: <38F52C61.A5D72F89@gte.net>

I need to generate a (largish) set of strings from a
pattern, sort of like glob with csh-style expansion.

Given "foo{,bar,baz/[aced]{,x,z}}", generate:

foo
foobar
foobaz/a
foobaz/c
foobaz/e
foobaz/d
foobaz/ax
foobaz/az

etc.

The manual doesn't seem to have this glob(3) option.

Sue.


From wayne.izatt at myself.com  Wed Apr  5 04:21:54 2000
From: wayne.izatt at myself.com (Wayne Izatt)
Date: Wed, 05 Apr 2000 08:21:54 GMT
Subject: A Mountain of Perl Books + Python Advocacy
References: <20000405053920.20579.qmail@web2102.mail.yahoo.com>
Message-ID: <38EAF7A0.E372EB2C@myself.com>

Books are one thing, but a regular printed journal would be especially
useful. A journal is visible, creates a strong sense of community, and is
an easy way to combine advanced and general Python topics in one tight
package. I wonder what the minimum number of Pythoners it would take to
keep something like that afloat?

jm2cw

lewst wrote:

> Before I launch into another question and gripe, I'd like to thank
> everyone who offered comments and suggestions on the advanced Python
> books that are out there.
>
> Now something I can't quite figure out is: why are there so many more
> books on Perl out there than on Python?
>
> Searching through Fatbrain.COM (which is where I order my books from),
> I found 68 books, 4 training manuals, and 2 eMatter documents on Perl.
> Compare this to Python's 13 books and 2 eMatter documents.
>
> What is it about Perl that makes it so much more popular and have such
> a huge grassroots swell?  I personally find Perl an abomination and
> Python a breath of fresh air.  Perl has that first mover advantage I
> suppose, but should that really make such a hugh difference?
>
> I'll admit that Perl is what led me to Python in the first place.
> After hearing about how great Perl was several years ago from the
> local sysadmin, I learned it and started using it for my scripting
> needs.  If there is one thing that sums up my Perl experience, it is
> that it always took me too long write the programs that I needed to
> write.  The syntax was always no unnatural for me that I could never
> get my head out of the reference manual, and errors were always so
> tough to track down.  In the end was frustration albeit a working
> result.  This frustration led me to look into Python and I'll never
> touch Perl again.  Python was so natural for me I often found myself
> correctly "guessing" at the syntax as I learned it.  I rewrote all my
> Perl scripts in one weekend and most of them worked on the first try.
>
> At this point I wondered if my brain was just different than all those
> Perl junkies out there.  But now I really don't think so; I think it's
> a question of awareness.  Perl is very publicized and well-known while
> the better language is sitting here a dark corner unnoticed.  Sure
> there will always be some religious fanatics that won't even give
> Python a try, but I think Python's popularity could be vastly improved
> with some serious advocacy work.  CNRI and/or PSA should seriously
> look into funding a Python "marketing" campaign of sorts.  I think the
> result would be allot of converts and more understanding and respect
> for Python.
>
> With this in mind, let me include one of my favorite pro-Python quotes
> of all time.  This is from a message to the fetchmail-announce mailing
> list by Eric S. Raymond <http://tuxedo.org/~esr>, the famous
> open-source advocate and author of many popular software programs.  He
> is discussing "fetchmailconf" which is the Tkinter GUI for his
> fetchmail POP/IMAP mail client.  I think it demonstrates the point I
> make above perfectly.
>
>   "A note about fetchmailconf. It took me approximately six days to
>   write this elaborate multi-paneled GUI -- that's counting the four
>   days it took me to learn the implementation language in the process.
>   This could easily have been a two-month project in C (with six weeks
>   of that spent debugging and bugs still left).  Or a week-long
>   project in Perl, with working but ugly and unmaintainable results."
>
>   "The verdict: Python is *waaaay* cool!  I'm sold.  It's clean, it's
>   elegant, it's easy, and it's astonishingly powerful.  I'm not going
>   to program anything longer than one screenful of script in Perl
>   anymore.  I love Larry Wall dearly, but Guido van Rossum is the
>   better designer -- I haven't had this much fun with a language since
>   the glory days of LISP.  Eric sez check it out."
>
> Regards.
>
> __________________________________________________
> Do You Yahoo!?
> Talk to your friends online with Yahoo! Messenger.
> http://im.yahoo.com



From kc5tja at garnet.armored.net  Sun Apr 30 09:00:22 2000
From: kc5tja at garnet.armored.net (Samuel A. Falvo II)
Date: 30 Apr 2000 13:00:22 GMT
Subject: Do I always have to write "self." ?
References: <270420001706386648%pecora@anvil.nrl.navy.mil> <slrn8ghc6t.f01.scarblac-spamtrap@flits104-37.flits.rug.nl> <8ecsgj$6lp$1@slb0.atl.mindspring.net> <m3bt2tonyg.fsf@atrus.jesus.cam.ac.uk> <300420001008012011%pecora@anvil.nrl.navy.mil> <20000430170952.A502@datapro.co.za> <8ehrb9$mm5$1@ites.inria.fr>
Message-ID: <slrn8gp797.ilf.kc5tja@garnet.armored.net>

In article <8ehrb9$mm5$1 at ites.inria.fr>, Cedric Adjih wrote:
>  If you can read, you should be able to read the definition of 
>any variable, and there are serious drawbacks. A flame of this
>notation is for instance in "Developing Windows NT Device Drivers":
>"The so-called `Hungarian Notation' (developped by a Hungarian
>programmer at Microsoft) is one of the worst ideas to have hit
>software development in many years"

"All good ideas are judged from the enemies they make." -- I forgot who.

>  The main problem, is that when you change the type of a structure
>you _cannot_ change the name (because other people are using it),
>so the type specification becomes plain wrong. The 'wParam'
>is reported to be such an example (now unsigned 32 bits), there
>are other examples in Microsoft API (should be worse with Win64).

Hungarian notation is fine for the definition of an API, as long as its used
to describe abstract concepts, like arrays, pointers to things, a count of
whatever, etc.  Hardcoding the bit-widths of types is obviously going to get
you into trouble in the future.  Just because Microsoft fscked up the
execution doesn't mean the idea is somehow "bad."

Like I said earlier, I use it only when it makes *sense* to.  Like anything
else you have in your toolshed, Hungarian notation is a tool.  Those who are
skilled in its use will not have the problems you cite above.  :)  Those who
aren't only use it because it's a fad.

>  Also each time you change the type of one variable/parameter, 
>you have change all the variable names of all the code under
>your control.

In my work with Dolphin, I've found, on several occasions, where I needed to
change symbol names pervasively throughout the source.  I have found #MORE
BUGS# just waiting to crop up and bite me in the butt this way than through
any other technique.  The very thing you cite as a "problem" is the very
thing that will help reduce the number of bugs in Dolphin upon its final
release.  Is it painful?  You bet.  Just ask Billy -- the translation from
old symbol names to new names broke the compile for weeks.  But once it was
all over, I had fixed no less than 14 bugs that would *never* become
apparent until after I'd released the code to the public.  Most of the bugs
were bound to be of the "intermittent" type too, so I was particularly
pleased at the end result.

I view it as sheer laziness on the programmers part.  You get what you pay
for.

>  And finally, for multi-platform code, it makes less sense:
>what happens if you don't have 2-bytes integers and you
>have code with wXXXX variables (unsigned int 16 bits) ?

On PowerPCs, a word is 32-bits.  On Intels, 16-bits.  So what's the lesson
to be learned here?  Don't hardcode widths in the meanings of your symbols.

Hungarian notation as used by Microsoft is by far NOT the only way to use
it.  For example, I *never* identify word-widths in my Hungarian notated
variable names.  Why?  Because of the very problem you cite above.  However,
I do indicate higher-level meanings with prefixes: p for pointer, c for a
count (e.g., cb for a count of bytes, cuint32 for a count of 32-bit unsigned
integers, etc), a for array, etc.

Again, it takes skill to use a tool well.

-- 
KC5TJA/6, DM13, QRP-L #1447
Samuel A. Falvo II
Oceanside, CA


From jimc at regnoc.com  Sun Apr  2 19:23:59 2000
From: jimc at regnoc.com (Jim Conger)
Date: Sun, 02 Apr 2000 23:23:59 GMT
Subject: Just doing local file testing
References: <38E796BC.23191B2@regnoc.com> <00040214141502.04535@quadra.teleo.net>
Message-ID: <38E7D6DE.EA09C891@regnoc.com>

An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20000402/87a1e7ef/attachment.html>

From mwh21 at cam.ac.uk  Mon Apr  3 06:26:49 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 03 Apr 2000 11:26:49 +0100
Subject: Python 1.6 alpha 1 released
References: <200004010159.LAA00403@piglet.dstc.edu.au> <8c9nq3$ghk$4@newsserver.rrzn.uni-hannover.de>
Message-ID: <m3bt3rcweu.fsf@atrus.jesus.cam.ac.uk>

breiter at usf.Uni-Osnabrueck.DE (Bernhard Reiter) writes:

> In article <200004010159.LAA00403 at piglet.dstc.edu.au>,
> 	David Arnold <arnold at dstc.edu.au> writes:
> > -->"Bjorn" == Bjorn Pettersen <bjorn at roguewave.com> writes:
> > 
> >   Bjorn> The more time I spend on fixing
> >   Bjorn> compatibility issues in Python scripts, the less time I spend
> >   Bjorn> on "real" work.
> > 
> > you have been using incorrect arguments.  the fact that they worked
> > was unfortunate because it leads to circumstances like this.  but the
> > time you have spent on these emails would have been enough to run a
> > simple script over your code to fix the problem ... ;-)
> 
> Hmm how simple would a script be to get all occurences of:
> 
> 	socket.connect(what, \ 
> 			else \
> 			)
> 
> 	sc=socket.connect
> 
> 	sc(what,else)
> 	sc(what, \
> 		else)
> 

Yes, but `grep -w connect' and the minimal application of a human
eyeball is surely enough to get 90%+ of the cases sorted out.  I can't
imagine sock.connect is called from all that many places in one's code
(but I might be wrong).  And, to answer you're question, about as
complicated as:

http://x42.deja.com/[ST_rn=ps]/getdoc.xp?AN=591243549

after `sed -e s/append/connect/', which doesn't get the bound method
case, but I am having trouble imagining circumstances in which you'd
reasonably do that.

Cheers,
M.

-- 
well, take it from an old hand: the only reason it would be easier
to program in C is that you can't easily express complex  problems
in C, so you don't.                 -- Erik Naggum, comp.lang.lisp


From ng at hardlight.couk.com  Mon Apr 10 22:40:38 2000
From: ng at hardlight.couk.com (.:|:.)
Date: Tue, 11 Apr 2000 03:40:38 +0100
Subject: downloading jpeg using python--extra bytes?
References: <sf5001f0cqv98@corp.supernews.com>
Message-ID: <38f2919b.0@news.proweb.co.uk>

Hi Gabriel,

> The script works, with one exception: the jpeg image has
> extra bytes in it that corrupt the file.  I'm not sure exactly
> how the bytes are dispersed, but they're not all at the
> beginning or at the end.

> g = open("myjpeg.jpg","w")


I can only presume you are in an MS-DOS environment where you need to use

g = open("myjpeg.jpg","wb")

instead of

g = open("myjpeg.jpg","w")

otherwise linefeeds are converted into carriage return / linefeed pairs

(it does tell you this in the manual page for open :-)

Matt






From matheusrodrigues at hotmail.com  Sat Apr  8 12:28:15 2000
From: matheusrodrigues at hotmail.com (Matheus Rodrigues)
Date: Sat, 08 Apr 2000 16:28:15 GMT
Subject: I need MacPython screenshots!
Message-ID: <20000408162815.50658.qmail@hotmail.com>

Hello everybody

I need some screenshots of a Macintosh running
MacPython. If possible I would like to receive one
screenshot of the MacPython Installation too. The
faster you can provide some material to me the more I
thank you.

Please send the files (zipped) to my email as we dont
want to create huge traffic in the list. It's not
necessary to crop the images.

btw, where is the Macintosh discussion list ?

Thank you
Matheus
______________________________________________________
Get Your Private, Free Email at http://www.hotmail.com




From maybe at lucky.com  Sat Apr  8 11:51:11 2000
From: maybe at lucky.com (Soundwave)
Date: Sat, 8 Apr 2000 17:51:11 +0200
Subject: starting python
Message-ID: <8cnkfi$at7$1@porthos.nl.uu.net>

Hi, I want to learn a bit python, I know a bit C and PASCAL, but I am
intrested in other languages.

where can I start ????

Do you naad a compiler or something for python, and where can I find one ?

also, if anyone got a good tutor, please, give me the URL

thanx

Niels




From helge.hess at mdlink.de  Sat Apr  8 09:48:44 2000
From: helge.hess at mdlink.de (Helge Hess)
Date: Sat, 08 Apr 2000 15:48:44 +0200
Subject: calendar/scheduling app
References: <38EE332E.2954EA49@fizzylab.com>
Message-ID: <38EF38BC.EE32E983@mdlink.de>

Hi,

you can try out SKYRIX, which is a set of groupware applications with a
wholly HTML based GUI. You can find a demo version at

  http://www.skyrix.com/de/products/trial_download.html

or on every SuSE Linux CD starting with version 6.2.

(parts of it are even implemented in Python ;-)

best regards
  Helge

paul beard wrote:
> 
> I am interesting in having a meeting scheduler/calendar app that does
> what something like Exchange purports to do without requiring me to use
> a client that doesn't exist. In other words, it should be
> platform-agnostic and use webforms/cookies to handle transactions and
> display options.
> 
> Possible in Zope? How hard? How long to do, given a baseline knowledge
> store of zero?
> 
> --
> Paul Beard / 206 728 4282 x138
> Fizzylab / 2025 First Ave, Suite 1150
> Seattle WA 98121


From jbauer at rubic.com  Wed Apr 19 11:07:28 2000
From: jbauer at rubic.com (Jeff Bauer)
Date: Wed, 19 Apr 2000 10:07:28 -0500
Subject: Playing wave files on Windows
References: <38FB6ED7.4C72773E@callware.com> <38FB7624.943A8C35@rubic.com> <38FB7862.7B15633@callware.com> <38FCE40A.1D882566@quasar.ipa.nw.ru>
Message-ID: <38FDCBB0.46FF3DAF@rubic.com>

[ Me & Ivan ]
>>> Then I can't interest you in my Python interface to
>>> the Diamond Rio MP3 player?  <0.5 wink>
>> 
>> Not unless everyone that I'll be distributing my application 
>> to will have one plugged in and ready to go;-)

[ Alexander Voinov ]
> But indeed, is it inevitable to have a specific module for 
> one specific MP3 player? 

It's inevitable if it's the only MP3 I own. <0.5 wink>

Seriously, I'm likely to add support for additional
players, assuming the Snowblind Alliance remains
intact and adds driver support for additional platforms.

> I have no ideas about MP3, but it's strange that 
> there are no abstraction layers.

I think we may be talking about two different 
issues. The manipulation and conversion of MP3 
files (e.g. .WAV -> MP3 conversion) is unrelated 
to copying MP3's to MP3 players via the parallel 
or USB port.  Though I would agree it would be 
nice to have a single abstraction layer for the 
download interface.  OTOH, if we don't add support 
for other players, we may never need it. ;-)

Jeff Bauer
Rubicon Research



From gmcm at hypernet.com  Wed Apr 12 12:30:23 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Wed, 12 Apr 2000 12:30:23 -0400
Subject: comparing strings and ints
In-Reply-To: <20000412112935.A2446939@vislab.epa.gov>
References: <jr%I4.643$rc9.190628352@newsb.telia.net>
Message-ID: <1256565470-46720619@hypernet.com>

Randall Hopper writes:
[vast snippery]
> I was of course referring to the behavior Gordon McMillan mentioned in the
> message I replied to.  Namely, that sorting a list of objects of
> hetrogenous primitive types "groups" the instances of a particular type
> together (ints together, strings together, etc.).

Back to the original subject: should comparing unlike types 
throw an exception.

The argument for seems to be that comparing unlike types is 
*probably* a programmer mistake. Certainly sometimes it is a 
programmer mistake (else this thread would not exist). Other 
times it is not. (I know I have made use of this behavior, 
though I think only in ad-hoc interpreter usage.)

I find myself aware that all Python objects (implicit in Python, 
but explicit in implementation) have a common root (in C, 
PyObject). In that sense, it's perfectly natural to be able to 
compare arbitrary objects safely.

In fact, current behavior already relies on "types" (eg, 
"Number") that are not explicit:
>>> .9 < 1 < 1.1
1
>>>

Certainly nothing in Doc/lib/built-in-funcs.html mentions any 
danger in comparing arbitrary objects, nor the discussion in 
2.1.5.2 ( Lib | Built-in Types | Mutable Sequence Types | sort).

Personally, I read Doc/ref/comparisons.html as warning that 
the ordering of types may change, not that comparing different 
types might become illegal.

So, I'd think it is safe to rely on the types being ordered; even 
if it is not safe to rely on the ordering.

- Gordon



From gregm at iname.com  Thu Apr 20 10:58:45 2000
From: gregm at iname.com (Greg McFarlane)
Date: Fri, 21 Apr 2000 00:58:45 +1000
Subject: Combobox in Tkinter
In-Reply-To: <8db0b1$b3o$1@nnrp1.deja.com>; from johngrayson@home.com on 16 Apr 2000 at 12:09:13AM
References: <38F8F133.54FCC64B@swt.edu> <8db0b1$b3o$1@nnrp1.deja.com>
Message-ID: <20000421005845.36765@nms.otc.telstra.com.au>

On 16 Apr, johngrayson at home.com wrote:
> In article <38F8F133.54FCC64B at swt.edu>,
>   "Shah, Navneet" <ns566645 at swt.edu> wrote:
> > How to decrease the width of the combobox in Tkinter? Do we have any
> > option for that
> 
> set the width of the entryfield component:
> 
> assuming the combobox id is 'xxx':
> 
>    xxx.component('entryfield').configure(width=4)

This is not quite correct.  The Pmw.ComboBox has an Pmw.EntryField
component, named 'entryfield', which has a Tkinter.Entry component,
named 'entry'. So, more correctly:
   xxx.component('entryfield').component('entry').configure(width = 4)

Since this is long winded, you can use Pmw's component options:
   xxx.configure(entryfield_entry_width = 4)

Or even shorter, since the Tkinter.Entry sub-sub-component of the
Pmw.ComboBox has an alias named 'entry':
   xxx.configure(entry_width = 4)

Component options (and aliases) can also be used during construction:
   xxx = Pmw.ComboBox(entry_width = 4)

For more detail, see the Pmw documentation.

-- 
Greg McFarlane     INMS Telstra Australia     gregm at iname.com



From rcunning at acm.org  Fri Apr  7 03:34:43 2000
From: rcunning at acm.org (Robert W. Cunningham)
Date: Fri, 07 Apr 2000 07:34:43 GMT
Subject: Tuples -- who needs 'em
References: <00e101bf9d80$a8384d20$74eb0b18@stcla1.sfba.home.com> <va3G4.1110$n68.190454272@newsb.telia.net> <037f01bf9e55$6d328760$74eb0b18@stcla1.sfba.home.com> <8cfa75$fho$1@newshost.accu.uu.nl> <38ED4292.2EE8D3CA@cosc.canterbury.ac.nz>
Message-ID: <38ED8EF2.B0835BE1@acm.org>

Greg Ewing wrote:

> Unfortunately, Python *does* need 1-tuples. If
> only ASCII had more kinds of brackets, Guido might
> have chosen differently. Maybe if Python 3000 uses
> Unicode for source files...

Well, if they can be syntactically disambiguated from Python relational
comparison operators, "<" and ">" can be used as "brokets".

Then again, we can always resort to something like trigraphs... <gag>  Whoops,
Python has already started down that road:  """

-BobC




From greg at cosc.canterbury.ac.nz  Mon Apr 17 21:41:44 2000
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Tue, 18 Apr 2000 13:41:44 +1200
Subject: module naming?
References: <Pine.BSF.4.10.10004162018190.25069-100000@chi.pair.com>
Message-ID: <38FBBD58.D1AF7A77@cosc.canterbury.ac.nz>

"Michal Wallace (sabren)" wrote:
> 
> def Record(dbc, table):
>    import zdc.record, zdc
>    zdc.Record = zdc.record.Record
>    return zdc.record.Record(dbc, table)
> 
> Is there a better way to do this?

Yes. In your __init__.py, simply say

   from Record import Record

That will put the Record class in the zdc package
namespace. 

The only drawback is that all the submodules will
be loaded as soon as zdc is imported, even if they're
not all used.

-- 
Greg Ewing, Computer Science Dept,
+--------------------------------------+
University of Canterbury,	   | A citizen of NewZealandCorp, a	  |
Christchurch, New Zealand	   | wholly-owned subsidiary of USA Inc.  |
greg at cosc.canterbury.ac.nz	   +--------------------------------------+


From bingdo19 at earthlink.net  Tue Apr 25 15:01:02 2000
From: bingdo19 at earthlink.net (Bob Langdon)
Date: Tue, 25 Apr 2000 19:01:02 GMT
Subject: Newbie Needs Help -- Name and Key errors
References: <bingdo19-4B8F57.12162225042000@news-server.twcny.rr.com> <SIjN4.1570$wYl.193603072@newsb.telia.net>
Message-ID: <bingdo19-4A0DE0.15103025042000@news-server.twcny.rr.com>

Thanks Fredrik!

Dumb ?, but how do I set the DOCUMENT_ROOT environment variable?
>From the fcsetup module?


> just guessing here, but it looks like
> 
> 1) you need to the DOCUMENT_ROOT environment variable
> before running the script
> 
> 2) the fcsetup.py module should define a variable called DEPOT,
> but doesn't.
> 
> > I think DEPOT might be a reference to the MySQL database, using the
> > Berkley dbmodule (which I can't find). I certainly can't find a
> > directory named DEPOT anywhere on the old host's server.
> 
> DEPT is a variable name, not a directory name.  the "from
> fcsetup import *" line imports all variables from the fcsetup
> module.
> 
> > Neither can I fathom where "DOCUMENT_ROOT" is defined, or find a file
> > called "os.environ."
> 
> "os.environ" refers to a variable in the standard "os" module,
> not a file.  see the library reference for more info.
> 
> </F>
> 
>


From lewst at yahoo.com  Sun Apr  2 16:33:38 2000
From: lewst at yahoo.com (lewst)
Date: Sun, 2 Apr 2000 13:33:38 -0700 (PDT)
Subject: NameError when assigning dictionary values
Message-ID: <20000402203338.21701.qmail@web2105.mail.yahoo.com>

This simple problem is driving my crazy.  I can't figure out why I get
a NameError with this program.

  # add each key-value pair to the dictionary and
  # then for each value, create an empty list.
  mydict = {}                         
  for key,value in [ (".", "one"), ("+", "two") ]:
      mydict[key] = value
      value = []
  print mydict
  print one,two

Whe I run this I get "NameError: one".  Why aren't my 2 empty lists
named "one" and "two" being created?


__________________________________________________
Do You Yahoo!?
Talk to your friends online with Yahoo! Messenger.
http://im.yahoo.com



From pete at visionart.com  Thu Apr 27 13:49:13 2000
From: pete at visionart.com (Pete Shinners)
Date: Thu, 27 Apr 2000 10:49:13 -0700
Subject: Concatening string and integer
References: <8e9gco$jlt$1@vg170.it.volvo.se> <390846EF.79A67E02@yale.edu>
Message-ID: <39087D99.3CAEF5AA@visionart.com>

Paul Magwene wrote:
> 
> JJ wrote:
> >
> > I want to concatenate a string containing both strings and integers like:
> >
> > strTemp = 'I make test number: ' + iMyNumber + ' once again'
> >
> > It doesn't work, how should I do this?
> 
> Use the str() function (one of the builtins), like so:
> 
> strTemp = "I make a test number: " + str(iMyNumber) + " once again"


you can also use the back ticks as a shortcut for the str()
builtin. you can do it like this;

strTemp = "I make a test number: " + `iMyNumber` + " once again"


From rcunning at acm.org  Tue Apr  4 05:23:27 2000
From: rcunning at acm.org (Robert W. Cunningham)
Date: Tue, 04 Apr 2000 09:23:27 GMT
Subject: From Python Newbie to Python Guru on the Fast Track:  HOW?
Message-ID: <38E9B48F.F1827543@acm.org>

Well, I just bit off more than I can chew.  While I know barely enough
Python to patch the occasional errant script, I've never written a thing
of my own.

Now I need to get up to speed FAST to design and develop a
cross-platform UI interoperability environment for Windows-to-Linux that
will allow much tighter integration of Windows apps running on a Linux
desktop, no matter if the Windows app is running locally under Win4Lin
or VMware, or is running remotely on a machine on a LAN or WAN or even
the Internet (via a remote-control app such as VNC).

I've selected Python simply due to the massive support it enjoys on both
platforms (Windows Python is nothing less than amazing), the simplicity
of the language itself, and it's ability to be used for fairly large
distributed programs (via DOPY or PYRO or Fnorb).  However, when the
Windows app is running locally, all unnecessary communication layers
should be eliminated, since there is only one CPU available to run
Linux, the Windows app, and this interface (making for one very busy
CPU).

The goal will be to provide only the connective interface needed to
bridge "the User Interface Gap" in the simplest way possible, even if
that way is not optimal at the start.  For example, all shared
filesystems will be either vfat (if local) and/or NFS and Samba (until
Windows gets native ext2 support ;^)).  This will allow the
implementation of services such as Drag&Drop to avoid having to
transport the dragged item itself, if it is a file, and will allow files
to be used as an intermediary mechanism (including on RAM disks, to
simplify or eliminate any shared memory issues).

Ultimately, I hope to expose Python Linux interfaces to many Windows
services, ones that would permit dynamic application interaction (such
as via ActiveX or OLE/COM), and maybe even to give Linux access to the
Windows Media Player.  The primary focus will be two-fold:  1)  Make it
easier for Windows apps to be sued from the Linux desktop.  2)  Give
Linux full access to Windows applications that have no true Linux
equivalent (such as MathCad, LabView, TurboTax, Quicken, FaxTalk
[answering machine + speakerphone + fax + BBS + ...], etc.).  One thing
I like about this entire approach is that it makes Windows seem more
like an optional Linux add-on than a competitor!

I'm an embedded real-time systems wonk, with lots of hardware how-to (I
hate device drivers with a passion - I've written far too many),
fantastic debugging skills, a decade of design and programming, a thin
layer of OOP, and very little Python experience.  I already program in
at least 10 languages, so picking up a new one won't be hard.  But
learning to use it well can be a bit harder, not to mention all the
extension packages.

As for the time angle:  I'm self-employed, and presently between
clients.  I have at least a few weeks to devote 100% to this effort,
maybe more.  At *least* a hundred hours or so, just to get started.

Then there's the money angle:  I don't mind spending $100 on this, but
I'd like to keep it minimal (I also need to get a new system this month
- this one is on its last legs).

Given where I am, and where I hope to go, what is the fastest way to
fill in the gaps in my knowledge?

I'm looking for the best books and tutorials that will do the job, with
my only access to other Python expertise being this newsgroup (or any
other similar forum that may be recommended).  I will be completely
unable to evaluate these resources until AFTER I've put substantial time
into the effort, so I'd really like to avoid the lemon I'd likely wind
up with if I shop randomly on FatBrain.  And I need to proceed to solid
discussions of (and practice with) the various Python packages fairly
quickly.  Please also include any on-line classes that are known to be
fast-paced and effective.

Right now, I'm reading everything on Python.org, since I'll need to be
familiar with that documentation no matter what else I use.

Suggestions?


TIA,

-BobC




From wrd at awenet.com  Mon Apr 10 16:23:25 2000
From: wrd at awenet.com (William R. Dickson)
Date: Mon, 10 Apr 2000 13:23:25 -0700
Subject: Problem running Python script as cron job
References: <wrd-0AEDDA.12282507042000@news.awenet.com> <0004071555090M.01234@quadra.teleo.net> <wrd-1E3688.12433910042000@news.awenet.com> <0004101202210J.00831@quadra.teleo.net>
Message-ID: <wrd-4AA63F.13232510042000@news.awenet.com>

In article <0004101202210J.00831 at quadra.teleo.net>, Patrick Phalen 
<python-list at teleo.net> wrote:

> :: > :: Anybody know of any issues with running python scripts from Cron?
> :: > 
> :: > No. I've had multiple Python scripts running every one, two, five 
> :: > and
> :: > ten minutes around the clock for years on multiple servers without
> :: > problems.
> :: > 
> :: > I'd suspect you might have a bug lurking.
> :: 
> :: I think so too, but I can't find it, because the script runs 
> :: flawlessly 
> :: every time I run it by hand, and the buffer never flushes the output 
> :: to 
> :: the log when it's run by Cron.
> 
> Are you "running by hand" as the same user as your script? Maybe you
> have a permissions problem, assuming the script needs to write to a
> file or whatever.

In both cases, it's running as root.

Thanks,

-Bill


From ljmacken at charter.net  Tue Apr 25 06:31:59 2000
From: ljmacken at charter.net (Luke macken)
Date: Tue, 25 Apr 2000 06:31:59 -0400
Subject: Newbie Question
Message-ID: <sgat0detcdo94@corp.supernews.com>

Hello, i just started teaching myself Python the other day, and I have been
getting very into it. I have both the Linux version of it and the Windows
version. I just want to know how i make executable files with it.
Thanks




From Michael.Scharf at gmx.de  Thu Apr  6 12:22:38 2000
From: Michael.Scharf at gmx.de (Michael Scharf)
Date: Thu, 06 Apr 2000 18:22:38 +0200
Subject: More about variables
References: <meh9-5C17A0.10445406042000@news.cit.cornell.edu>
Message-ID: <38ECB9CE.44A73C13@gmx.de>

Matthew Hirsch wrote:
> 
> Hi All,
> 
> Let's say I had something like:
> 
> number_of_variables=int(raw_input('Enter number of variables: '))
> 
> if number_of_variables==1:
>    variable1=[]
> elif number_of_variables==2:
>    variable1=[]
>    variable2=[]
> elif number_of_variables==3:
>    variable1=[]
>    variable2=[]
>    variable3=[]
> elif number_of_variables==4:
>    variable1=[]
>    variable2=[]
>    variable3=[]
>    variable4=[]
> elif number_of_variables==5:
>    variable1=[]
>    variable2=[]
>    variable3=[]
>    variable4=[]
>    variable5=[]
> else:
>    print 'else'
> 
> Is there a more efficient way of doing this? What if I wanted to create
> 100 variables (lists).  It doesn't make sense to write an if statement
> for that many conditions.

Just do:

for i in range(number_of_variables):
    vars()["variable%s" %(i+1)]=[]

Michael
-- 
     ''''\     Michael Scharf
    ` c-@@     TakeFive Software GmbH, a Wind River Company
    `    >     http://www.WindRiver.com
     \_ V      mailto:Michael.Scharf at gmx.de


From alberanid at bigfoot.togli.questo.com  Sun Apr 30 07:54:34 2000
From: alberanid at bigfoot.togli.questo.com (Davide Alberani)
Date: Sun, 30 Apr 2000 11:54:34 GMT
Subject: Problem running NNTP example from docs.
References: <87hfckk6fv.fsf@tea.thpoon.com>
Message-ID: <0nvge8.7ua.ln@snoopy.mio>

Arcady Genkin <a.genkin at utoronto.ca> wrote:
>
> Could somebody point out what I am missing?

Try:
  s = nntp.NNTP('news')

Or using this 'import' statement:
  from nntplib import NNTP

and than:
  s = NNTP('news')

-- 
Cinereview: Stigmate: "tecnicamente" non male, un po' meno   _/Davide\_
gli altri aspetti troppo superficiali.  Voto: 5.5            \Alberani/


From dworkin at ccs.neu.edu  Thu Apr 27 12:31:33 2000
From: dworkin at ccs.neu.edu (Justin Sheehy)
Date: 27 Apr 2000 12:31:33 -0400
Subject: Python 2.0
In-Reply-To: m.faassen@vet.uu.nl's message of "27 Apr 2000 12:36:45 GMT"
References: <3907B52D.B01D5748@myself.com> <8e9c8t$hi6$1@newshost.accu.uu.nl>
Message-ID: <vndln1zv72y.fsf@camelot-new.ccs.neu.edu>

m.faassen at vet.uu.nl (Martijn Faassen) writes:

>       * case insensitivity. foo = Foo() would break
>  
>     Opinions on them vary; they may reek a bit like Do-What-I-Mean,
>     but nobody thinks either change would be a disaster.

Yow.  I would personally consider that one a disaster, but my vote
doesn't count for much.  I very much like the idiom of an instance
being a lowercased version of its class name, in some cases.

I guess it depends on one's definition of "distaster"...

There will be plenty of time to argue over this stuff before it
becomes really relevant, in any case.

-Justin

 



From jnw22 at cam.ac.uk  Tue Apr 11 08:31:51 2000
From: jnw22 at cam.ac.uk (James "Wez" Weatherall)
Date: Tue, 11 Apr 2000 13:31:51 +0100
Subject: WinNT and TCL/TK
References: <eviltofu-9BCDF3.08165411042000@news.earthlink.net>
Message-ID: <8cv5vo$f0v$1@pegasus.csx.cam.ac.uk>

"Jerome Chan" <eviltofu at rocketmail.com> wrote in message
news:eviltofu-9BCDF3.08165411042000 at news.earthlink.net...
> Have installed Python 1.5.2 on a WinNT Server but it says it can't find
> a TCL lib when I start it up. Do I have to add something?

This is a path-setting bug, as far as I can tell.  Under the My Computer
Properties tab in Windows, change the system's PATH to include the directory
containing the Tcl library (usually c:\program files\tcl\bin).

The other option is to copy the tcl80.dll (?) file from that directory into
the system32 directory.  That way you needn't alter the system PATH but you
will have a rogue file in your system directory!

Hope that helps,

--
James "Wez" Weatherall





From PClaerhout at CREO.BE  Wed Apr 12 06:30:01 2000
From: PClaerhout at CREO.BE (Pieter Claerhout)
Date: Wed, 12 Apr 2000 12:30:01 +0200
Subject: Saving a copy of a remote NT event log
Message-ID: <2B1262E83448D211AE4B00A0C9D61B03BA72A0@MSGEURO1>

It does indeed, and it works like this:

    import win32evtlog

    evtLogSys = win32evtlog.OpenEventLog('pclaerhout', 'system')
    win32evtlog.BackupEventLog(evtLogSys, 'c:\\system.evt')
    win32evtlog.CloseEventLog(evtLogSys)

Kind regards,


Pieter
Pieter_Claerhout at creoscitex.com


-----Original Message-----
From: Mark Hammond [mailto:mhammond at skippinet.com.au]
Sent: Wednesday, April 12, 2000 2:16 AM
To: python-list at python.org
Subject: Re: Saving a copy of a remote NT event log


Im fairly sure the win32evtlog module will do this for you.  Check the
docs.

Mark.

"Pieter Claerhout" <PClaerhout at CREO.BE> wrote in message
news:2B1262E83448D211AE4B00A0C9D61B03BA7295 at MSGEURO1...
> Hello,
>
> does anyone know if it's possible to save a copy of the NT-event log
> from a remote machine using Python? Normally, you have to do a
> save as from that machine, and then copy over the logfile, but I want
> to automate this using Python. Any ideas?
>
> Kind regards,
>
>
> Pieter
>
> [ Pieter Claerhout :: CreoScitex Response Centre :: Applications
support ]
> [ Pieter_Claerhout at creoscitex.com    ::    http://www.creoscitex.com/ ]
> [ CreoScitex Europe  ::   Excelsiorlaan 21   ::  1930 Zaventem  ::
Belgium ]
>
>


-- 
http://www.python.org/mailman/listinfo/python-list



From wrd at awenet.com  Fri Apr  7 15:28:25 2000
From: wrd at awenet.com (William R. Dickson)
Date: Fri, 07 Apr 2000 12:28:25 -0700
Subject: Problem running Python script as cron job
Message-ID: <wrd-0AEDDA.12282507042000@news.awenet.com>

Hi all,

I've got a Python script that runs very nicely from the command line, 
but seems to hang up and loaf around forever if I run it as a cron job.  
>From the command line, I do:

     /path/to/statsthing -r

That works well.  The script prints many status messages to STDOUT as it 
runs.  I want to preserve these messages when running the script 
automatically, so I do this in the crontab:

     /path/to/statsthing -r > /usr/local/temp/.log

This is a script I rewrote from an old Perl version as a learning 
exercise to teach myself Python.  The Perl version runs from Cron very 
happily with that command.  The Python version runs to a point, then 
simply stops, and I need to kill the processes.  The log contains some 
data, but I'm pretty sure some of it is never being flushed to disk, so 
I'm not sure exactly where it's halting.

Anybody know of any issues with running python scripts from Cron?

Thanks,

-Bill


From kent at tiamat.goathill.org  Thu Apr 20 12:44:24 2000
From: kent at tiamat.goathill.org (Kent Polk)
Date: 20 Apr 2000 16:44:24 GMT
Subject: Mac Python Questions
References: <IRjL4.271$CP3.3183@news.uswest.net> <38FE8032.DE233C29@cosc.canterbury.ac.nz>
Message-ID: <956247497.216189@fezzik.endicor.com>

On Thu, 20 Apr 2000 15:57:38 +1200, Greg Ewing wrote:
>Jeff Massung wrote:
>> 
>> b) No matter what I do, I can't get Tkinter to work :(.
>
>Neither can I, and from what I've seen here, neither
>can anyone else. Tkinter just seems to be broken on
>the Mac at present.

I have the distributed Python 1.5.2 with tkinter (precompiled)
working on some older PPC macs and even some G3's (or whatever they
are).  I know nothing about Macs, but I finally got around to
writing a tkinter frontend for a PCR data postprocessor that had
to run on the wetlab Macs that run the ABI 'Genotyper' software.
I developed on unix and just had to adapt some printer support code
that I found in the included mac-specific stuff for it to run on
the Macs.

The standard Python installation program installed the wrong version
of the _tkinter shared library and(/or) didn't set the path to the
correct (ppc) version, so Python couldn't load _tkinter. If that
is the problem, there are several ways to solve it.  If that isn't
the problem, I'm not sure what to tell you as it's working on I
think 3 or 4 different mac platforms here.



From tim_one at email.msn.com  Sat Apr  1 14:54:07 2000
From: tim_one at email.msn.com (Tim Peters)
Date: Sat, 1 Apr 2000 14:54:07 -0500
Subject: Why should I switch to Python?
In-Reply-To: <secfidq8921@corp.supernews.com>
Message-ID: <000201bf9c14$0a79d9e0$182d153f@tim>

[Aaron Turner]
> ...
> Basically if you know both languages [Perl & Python], and were planning
> a mid-sized project, why would you choose Python?

When I was very young, I wondered why my dad's electric razor didn't rip his
face to shreds.  As a budding empirical scientist, I tested my theories by
trying it first on my cheek, and then on my tongue.  This was very
educational -- although it's still painful to talk about <wink>.

In other words, "voice of experience".  There aren't a lot of bilingual P
programmers out there, but I expect you'll find the majority of them hanging
out on c.l.py rather than c.l.p.m.  That should tell you the conclusion
you're likely to reach if you decide to become one of them too.  Nobody is
going to argue you into it, though, and Python is easy enough to learn that
there's no significant barrier against just giving it a try.

>From what I've seen-- which matches my own experience --people generally
switch over gradually.  The prime motivation seems to be that, time after
time, when tracking down a bug in a Perl program, you increasingly realize
"hmm -- if I had written this in Python, this wouldn't have happened".
Whether it's the 200th time you've been shredded by an array-vs-scalar
context surprise, or the 12th time an implicit coercion guessed the wrong
way, they're traps you *can't* fall into in Python.

Python's syntax is simpler, its semantics more regular, and its philosophy
of raising exceptions is "gripe whenever there's a reasonable doubt" rather
than "assume the programmer meant whatever they wrote".  These conspire to
make larger Python systems less work over the long haul, and especially in a
team project (TMTOWTDI loses its cowboy charm when you have to decipher 6
other peoples' code every day too).

> It's hard to take my friend seriously in his evaluation when he says
> he'd rather do this project in C++ than in Perl (especially when he
> doesn't know Perl).

Depends on the project.  I make my living these days on a project that's
about a half million lines of C++ and assembler -- Perl and Python are both
too fat and slow to have much use in it apart from automating testing.  Perl
gets used a lot for the latter, but other groups where I work, with large
projects that can afford the fat of either P language, all use Python now,
or Java + JPython.  There's no company policy on this, it's a group-by-group
decision.  And despite that we now have many more lines of Python than Perl,
I still spend most of my "help others debug their scripts" time staring at
Perl.

a-day-of-experience-is-worth-a-decade-of-usenet<wink>-ly y'rs  - tim





From mjhand at concentric.net  Sun Apr 16 03:22:26 2000
From: mjhand at concentric.net (Manus Hand)
Date: 16 Apr 2000 03:22:26 EDT
Subject: Python Starship Grounded??
References: <38F6B2DF.940E7C23@concentric.net> <OqQJ4.8866$fV.715336@bgtnsc05-news.ops.worldnet.att.net> <8db5sb$gsc$1@nnrp1.deja.com>
Message-ID: <38F969B5.D2B201E@concentric.net>

I too live my whole life on the ship.  And many other people do too,
through the use of my Websites.  I certainly hope someone is reading
these cries for assistance and doing what can be done about putting
the starship back in working order.

I have twice mailed the mcsp folks who host the machine, but no answer
as of yet.

Manus


From rumjuggler at home.com  Sun Apr 16 13:35:46 2000
From: rumjuggler at home.com (Ben Wolfson)
Date: Sun, 16 Apr 2000 17:35:46 GMT
Subject: Python paradigms
References: <8cmr70$b0e$1@pegasus.csx.cam.ac.uk> <8dbdj3$dgt$1@slb2.atl.mindspring.net> <8dcpom$fbc$1@slb3.atl.mindspring.net>
Message-ID: <9nujfs0p7vn2idr0b7buls12fvdq19llhh@4ax.com>

On Sun, 16 Apr 2000 04:17:22 -0600, "Andrew Dalke" <dalke at acm.org>
wrote:

>You can imagine:
>
>def expression_if(test, true_string, false_string):
>  # some nasty code to get the locals of the calling function
>  if test:
>    return eval(true_string with the given locals)
>  else:
>    return eval(false_string with the given locals
>
>x = expression_if(a is not None, "a[i].weeble", "0") + ...

I did this.  Instead of the nasty code to get the locals of the
calling function, you can change the parameter list to

def expression_if(test, true_string, false_string, gbals, lcals):

and pass in globals and locals when you call it.

-- 
Barnabas T. Rumjuggler

Half our days we pass in the shadow of the earth; and the brother of
death exacteth a third part of our lives.
 -- Thomas Browne.


From paul.robinson at quantisci.co.uk  Thu Apr 13 09:18:48 2000
From: paul.robinson at quantisci.co.uk (Paul Robinson)
Date: Thu, 13 Apr 2000 14:18:48 +0100
Subject: Python CGI running commands under NT
References: <2B1262E83448D211AE4B00A0C9D61B03BA72B1@MSGEURO1>
Message-ID: <38F5C938.8278EBFA@quantisci.co.uk>


Pieter Claerhout wrote:
> 
> Are there any other ways to accomplish this?

There is always:

import os, tempfile
t = tempfile.mktemp()
command = 'ipconfig > ' + t

os.system(command)

p = open(t, 'r')

But I'm sure someone can think of a better way of doing this.

This question has definately been asked before (possibly by me!) so see
if you can find a fuller answer at dejanews...

	Paul.

-----------------------------------
Business Collaborator Team
Enviros Software Solutions
http://www.businesscollaborator.com
-----------------------------------



From tismer at tismer.com  Mon Apr 17 13:38:57 2000
From: tismer at tismer.com (Christian Tismer)
Date: Mon, 17 Apr 2000 19:38:57 +0200
Subject: How do I make a constant global variable in Python Please
References: <vCj7OA2LfMfb=znR7yTHOvmiZAkd@4ax.com>
Message-ID: <38FB4C31.734DFBB3@tismer.com>

Hi,

cmfinlay at magnet.com.au wrote:
> 
> How do I make a constant global variable in Python Please.
> or just a constant.
> or just a global.
> 
> I bought a Python Book. however I can't find this info in it.
> 
> Please e-mail me at cmfinlay at magnet.com.au
> --
> http://www.python.org/mailman/listinfo/python-list

There is no such concept as you required, but it can be modelled.
If you can live with the rule that you need to use one global
thunk that holds your real vars/consts, then the following will
work for you.

The advantage of this approach:
After the debugging phase, you can replace the "declared" class
with an empty class, and things will work just faster.
You can also use it to monitor usage of certain variables
if you like.

ciao - chris

file declared.py
--------------------------------------------------------------------
# declared global variables
# done with a class.
#
# Christian Tismer, Y2k0417

class declared:
    """A class that mimicks declaration of variables before use.
    Usage example:
        g = declared("hi there")
    gives you write access to the attributes "hi" and "there".
    Note that they cannot be read before they have been set.
    """
    def __init__(self, namestring=""):
        self.__dict__["__names__"] = {}
        self.declare(namestring)
    def __setattr__(self, key, value):
        dic = self.__dict__
        nam = dic["__names__"]
        dic[nam[key]] = value

    def __getattr__(self, key):
        dic = self.__dict__
        nam = dic["__names__"]
        return dic[nam[key]]

    def vars(self):
        return self.__names__.keys()

    def declare(self, namestring):
        import string
        for name in string.split(namestring):
            self.__names__[name] = name

sample_session = """
>>> g=declared("hi there")
>>> g.vars()
['hi', 'there']
>>> g.hi
Traceback (innermost last):
  File "<interactive input>", line 1, in ?
  File "D:\python\spc\declared.py", line 24, in __getattr__
    return dic[nam[key]]
KeyError: hi
>>> g.hi=4
>>> g.hi
4
>>> del g.hi
>>> g.hi
Traceback (innermost last):
  File "<interactive input>", line 1, in ?
  File "D:\python\spc\declared.py", line 24, in __getattr__
    return dic[nam[key]]
KeyError: hi
>>>
"""
------------------------------------------------------------------

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaunstr. 26                  :    *Starship* http://starship.python.net
14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     where do you want to jump today?   http://www.stackless.com



From hakan at netg.se  Thu Apr 13 11:50:19 2000
From: hakan at netg.se (Hakan Nilsson)
Date: Thu, 13 Apr 2000 17:50:19 +0200 (CEST)
Subject: List mapping?
In-Reply-To: <newscache$wroysf$2s2$1@jenpc07.jennic.co.uk>
Message-ID: <Pine.LNX.3.96.1000413174501.3264B-100000@valdez.netg.se>

On Thu, 13 Apr 2000, Robert Cragie wrote:

> > Try this:
> > mylist = ['/a','1','/b','2','/c','3']
> > map(lambda x: reduce(lambda a, b: (a,b), mylist[x:x+2]), range(0,
> > len(mylist), 2))
> 
> Sorry, forgot to assign...
> 
> mylist = ['/a','1','/b','2','/c','3']
> tuplist = map(lambda x: reduce(lambda a, b: (a,b), mylist[x:x+2]), range(0,
> len(mylist), 2))
> 
> Robert Cragie
> 
> 

Howcome you use reduce at all?

Works fine with just doing 

tuplist = map(lambda x: (mylist[x], mylist[x+1]), range(0, len(mylist),
2))

Reduce has many uses, but I can't see why this should be one of them ;o)

(as mentioned, getopts would be even better though;)

/H

H?kan Nilsson                   NetGuide Scandinavia
hakan at netg.se                   031-50 79 40




From jay_love at mindspring.com  Mon Apr  3 19:27:57 2000
From: jay_love at mindspring.com (Jay Love)
Date: Mon, 03 Apr 2000 18:27:57 -0500
Subject: PyGreSQL errors
References: <v04210108b50ec6c8e9bf@[192.168.1.2]>
Message-ID: <8cb9db$p0k$1@slb3.atl.mindspring.net>

The first thing that comes to mind is, have you created the test database?

Jay

In article <v04210108b50ec6c8e9bf@[192.168.1.2]>, Lee Joramo <ljoramo.list at nickads.com> wrote:
> I am trying to get pygresql to work. I have a Cobalt RaQ3 server which runs
> a slightly modified version of RedHat. The system came with PostgreSQL
> pre-installed. When I import the pygresql's basics.py tutorial at pythons
> command line, I follow the instructions to create a connection and get an
> error:
> 
> 
>>>> cnx = basics.DB('mytest')
> Traceback (innermost last):
>   File "<stdin>", line 1, in ?
>   File "/usr/lib/python1.5/pg.py", line 34, in __init__
>     self.db = apply(connect, args, kw)
> pg.error: FATAL 1:  Database mytest does not exist in pg_database
> 
> 
> 
> What has gone wrong?
> 
> --
> Lee A. Joramo                      ljoramo at nickads.com
> The Nickel Want Ads                www.nickads.com
> Internet Manager                   970-242-5555
> 
> --
> 
> Onward and Upward!
> 
> Lee A. Joramo }}-----> Cowboy on the Electronic Frontier
> 
> lee at joramo.com
> 
> Alamar Ranch  -  Palisade  -  Colorado
> 
> ThreeRivers|Eagle|UCD|D&D|TNC|DX|Mac|221B|Cannondale|Deutsch
> 
> 




From nick at spam_me_notvideosystem.co.uk  Thu Apr 13 11:01:38 2000
From: nick at spam_me_notvideosystem.co.uk (Nick Trout)
Date: Thu, 13 Apr 2000 16:01:38 +0100
Subject: List mapping?
References: <CC59868A0A76D311B1F50090277B321C1F326A@VRTMAIL>
Message-ID: <38f5e2b0@news.xtml.co.uk>

I came up with something similar:

    processfn = { 'del':deletefiles }
    comm = None
    flist = []
    fileopts = sys.argv[2:]
    while fileopts:
        if fileopts[0]=='/c':
            comm = fileopts[1]
            fileopts = fileopts[2:]
        elif fileopts[0]=='/d':
            fdir = fileopts[1]
            ext = fileopts[2]
            fileopts = fileopts[3:]
            flist.extend( readdir(fdir,ext) )
        else:
            raise "Unknown flag:",fileopts[0]
    processfn[comm](flist)

There are obviously a lot of ways to do this in Python, the most elegant way
doesnt always seem obvious at first! Just wondered how others would approach
this problem.

Cheers,
Nick.


"Mike Fletcher" <mfletch at tpresence.com> wrote in message
news:CC59868A0A76D311B1F50090277B321C1F326A at VRTMAIL...
> not elegant, but here's another way...
>
> original = string.split( x, ' ')
> result = []
> while original:
> result.append( tuple(original[:2]) )
> del original[:2]
>
> and another...
>
> original = string.split( x, ' ')
> result = []
> for x in range(0, len(original), 2 ):
> result.append( tuple( original[x:x+2]))
>
> HTH,
> Mike
>
> -----Original Message-----
> From: Nick Trout [mailto:nick at spam_me_notvideosystem.co.uk]
> Sent: Thursday, April 13, 2000 10:10 AM
> To: python-list at python.org
> Subject: List mapping?
>
>
> Is there a nice way of pairing up the members to form tuple pairs in another
> list? ie. like you might to processing input args.
>
> ie. changing ['/a','1','/b','2','/c','3'] into [ ('/a','1'), ('/b','2'),
> ('/c','3') ]
>
> eg:
>
> args = '/a 1 /b 2 /c 3'
> listargs= string.split(args)
> tuplist = []
> i=0
> while i < len(listargs):
>     tuplist.append( (listargs[i],listargs[i+1]) )
>     i = i+2
> for i in tuplist:
>     processargs(i)
>
> or:
>
> args = '/a 1 /b 2 /c 3'
> for i in string.split(args):
>     i = argflag
>     # i want next arg now as a parameter!!!
>
> Can I (you!) get rid of the long winded while loop cleverly?!!  :-)
>
> Thanks in advance,
> Nick.
>
>
>
>
>
> --
> http://www.python.org/mailman/listinfo/python-list
>




From neelk at brick.cswv.com  Thu Apr 13 19:12:18 2000
From: neelk at brick.cswv.com (Neel Krishnaswami)
Date: 13 Apr 2000 23:12:18 GMT
Subject: Why is if a in dict.keys() so slow.....
References: <38F5A075.CDFA32D9@bedford.waii.com>
Message-ID: <slrn8fcn3s.2vl.neelk@brick.cswv.com>

Martin Franklin <martin.franklin at waii.com> wrote:
> Hi,
> 
> Just thought i'd throw this out there....
> 
> This is very slow:-
> 
> if split[0] in map_dict.keys():
>     print 'found it!', map_dict[split[0]]
> else:
>     print 'not found'
> 
> Question is why?


dict.keys() generates a list of the keys each time it is called, so
you are generating a list of 40K elements 250,000 times. Then, to do
the "split[0] in map_dict.keys()" test, you are searching through that
list one element at a time. So on average you'll search 40K/2 = 20K
elements before finding the key. Lookup for dictionaries (ie, writing
dict[foo]) is very fast -- the time to do a lookup doesn't go up, no
matter how many items you put in the dictionary. So the try/except
statement should run around 20K times faster.

If you just want to check whether an element is in a dictionary, there
is a method called dict.has_key(foo); it will return true if foo is a
key in the dictionary and false if not. This is also a very fast
lookup.

So you could also write:

  if map_dict.has_key(split[0])
      print "Found it!"
  else:
      print "Not found!"

The choice of try/except versus dict.has_key() is mostly one of taste;
the speed differences between the two is quite small, so you should
pick whichever one makes your intent clearer. (Incidentally, this will
also usually be the faster choice.)


Neel


From sabren at manifestation.com  Sun Apr 23 23:02:50 2000
From: sabren at manifestation.com (Michal Wallace (sabren))
Date: Sun, 23 Apr 2000 23:02:50 -0400 (EDT)
Subject: crypt() results
Message-ID: <Pine.BSF.4.10.10004232258410.24749-100000@chi.pair.com>


Well. I tried porting both Crypt::UnixCrypt and Crypt::PasswdMD5 to
python.  And neither one of the python versions give the same results
as the perl versions. They come out in the same format, but with
different values.

I've pounded my head, pulled out my hair, and finally given up. I
simply don't know enough about the algorithms or perl's bit twiddling
to understand what these modules are actually doing.

If anyone else wants to take a stab at it, my code's at:

http://www.sabren.com/code/python/crypt/

(That page also has links to the perl versions on CPAN)

Cheers,

- Michal
-------------------------------------------------------------------------
http://www.manifestation.com/         http://www.linkwatcher.com/metalog/
-------------------------------------------------------------------------




From kragen at dnaco.net  Sun Apr 23 01:33:47 2000
From: kragen at dnaco.net (Kragen Sitaker)
Date: Sun, 23 Apr 2000 05:33:47 GMT
Subject: The Simple Economics of Open Source
References: <B527E7A5.51AA%raffael@mediaone.net>
Message-ID: <%WvM4.1478$RM6.3452288@news-east.usenetserver.com>

In article <B527E7A5.51AA%raffael at mediaone.net>,
Raffael Cavallaro  <raffael at mediaone.net> wrote:
>Sarcasm aside, I think you miss the point. Those artists least willing to
>share, most jealous of their "secrets" have always been those who were most
>successful. Unsucessful young artists are, by and large, quite willing to
>share ideas, working techniques, etc. It's actually the more established,
>financially succesful ones, who are not.
>
>I see a parallael to the software industry here as well. Those with a
>lucrative cash cow (e.g., MS) are reluctant to share code, but those who
>just make a living programming are happy to share source. It would seem that
>sharing is inversely proportional to the value of that which is shared.

<spoof>
Those spenders most willing to spend profligately have always been
those who had the most money.  Unsuccessful young spenders are, by and
large, unwilling to spend money on luxuries like expensive cars, large
houses, etc.  It's actually the more established, financially
successful ones who do.

Those with a lucrative trust fund (e.g. Bill Gates) are happy to spend
lots of money on nonessential things, but those who just make a living
flipping burgers at McDonald's are reluctant to do so.  It would seem
that spending is directly proportional to the value of that which is to
be spent.
</spoof>

Correlation doesn't imply causality, and when it does, it's easy to get
causality backwards if you're not careful.  You could get misled into
spending lots of money to get rich if you just look at the
correlations.
-- 
<kragen at pobox.com>       Kragen Sitaker     <http://www.pobox.com/~kragen/>
The Internet stock bubble didn't burst on 1999-11-08.  Hurrah!
<URL:http://www.pobox.com/~kragen/bubble.html>
The power didn't go out on 2000-01-01 either.  :)


From tseaver at starbase.neosoft.com  Tue Apr  4 00:58:49 2000
From: tseaver at starbase.neosoft.com (Tres Seaver)
Date: 3 Apr 2000 23:58:49 -0500
Subject: Python 1.6 alpha 1 released
References: <Pine.GSO.4.10.10004040035420.21084-100000@sundial>
Message-ID: <5E877FD7637AB46E.DABBED9CF3C1D64B.CDDDC05AE7374A78@lp.airnews.net>

In article <Pine.GSO.4.10.10004040035420.21084-100000 at sundial>,
Moshe Zadka  <mzadka at geocities.com> wrote:
>On 3 Apr 2000, John Aycock wrote:
>
>> I'm not going to take sides here, but I coincidentally just finished
>> reading something highly relevant:
><snipped>
>
>I beg to differ: it's not highly relevant. What Guido did was correct
>a *bug* in the implementation. That people depended on it is sad, and
>just goes to show it should be corrected ASAP before more people depend
>on it.

The point of this debate is that classifying it as a *bug* is, well,
debateable.  Years of socket programming experience makes:

  s.connect( host, port )

seem like a perfectly reasonable, sane interface;  disallowing it
may preserve "orthagonality," or some other academic purity, but
it *breaks existing applications coded to the published interface*,
with no appreciable real-world benefit (yes, I realize that the published
interface is inconsistent;  I do not accept the deprecation of the
examples as a reasonable argument.)  Consider that *every one* of the
protocol modules which offers a connect() method uses a two-argument
signature for it:  ftplib, gopherlib, httplib, imaplib, nntplib, poplib,
smtplib, telnetlib, urllib.  And they all *use* the two-argument connect
when invoking their underlying socket!  No Python module in the standard
library uses the single-argument connect, period.

This is *not* the same case as the list.append() one, where the
published interface was itself consistent -- I was quite surprised to
learn just this spring that multi-argument append was even possible.
But the socket module is one of the first I learned in Python, and I
have consistenly used the two-argument connect.  Constructing a spurious
"address" object (which has no behavior, and exists only to be torn
apart inside the implementation) seems a foolish consistency, beyond
doubt.

Rough-consensus-and-running-code-wins-every-time'ly

Tres.
-- 
---------------------------------------------------------------
Tres Seaver        tseaver at digicool.com     http://www.zope.org
Digital Creations


From jon at dgs.monash.edu.au  Mon Apr 10 21:13:20 2000
From: jon at dgs.monash.edu.au (Jonathan Giddy)
Date: Tue, 11 Apr 2000 11:13:20 +1000 (EST)
Subject: Frames in TK
In-Reply-To: <38F266FB.2363E980@swt.edu> from "Shah, Navneet" at Apr 10, 2000 07:42:51 PM
Message-ID: <200004110113.LAA17487@nexus.csse.monash.edu.au>

Shah, Navneet declared:
>
>Can anyone let me know the code for making frames in TK. I need to have
>two frames in the window with the top frame should cover 65% of the
>window while the remaining 35% should be covered by the second frame.

For that level of control, use the Tk place command.  Alternatively, the Pmw
PanedWidget gives the user more control.

from Tkinter import *

def create_frames(master):
    # relief and borderwidth arguments make frames visible
    frame1 = Frame(master, relief='sunken', borderwidth=2)
    frame2 = Frame(master, relief='sunken', borderwidth=2)

    frame1.place(relx=0, rely=0, relwidth=1, relheight=0.65)
    frame2.place(relx=0, rely=0.65, relwidth=1, relheight=0.35)

    # put something in the frames
    label1 = Label(frame1, text="Hello")
    label1.grid(row=0, column=0)
    label2 = Label(frame2, text="World!")
    label2.grid(row=0, column=0)

root = Tk()
create_frames(root)
root.mainloop()



From gmcm at hypernet.com  Thu Apr 20 11:12:45 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Thu, 20 Apr 2000 11:12:45 -0400
Subject: Using popen in windows
In-Reply-To: <newscache$y7jbtf$od$1@jenpc07.jennic.co.uk>
Message-ID: <1255878929-30440209@hypernet.com>

Robert Cragie

> I tried this and it still hangs; cvs.exe is definitely not a built in:
> 
> from win32pipe import popen
> p = popen('c:\\cygwin\\contrib\\bin\\cvs -v','r')
> str = p.read()
> print str
> 
> Again, the equivalent works fine on Linux. 

Works fine on NT, too. Bill Tutt put out something that works 
around the Win9x bug. Searching dejanews on popen and 
Win95 should find it.

> So how do I run a command in
> Windows and capture the output? The following works, but is clumsy, and you
> get the annoying DOS box while it runs - is there a better way?:
> 
> import os
> os.system('dir >temp')
> f = open('temp', 'r')
> f.read()

Yeah, os.listdir('.') <wink>.

To get around the singing, dancing consoles, you need to use  
os.spawnv, which is another whole can of worms.

Python 1.5.42+ (#0, Mar 28 2000, 20:39:47) [MSC 32 bit 
(Intel)] on win32
Copyright 1991-1995 Stichting Mathematisch Centrum, 
Amsterdam
>>> import os
>>> os.spawnv(os.P_WAIT, 'd:\\util\\bin\\cvs.exe', 
('d:\\util\\bin\\cvs.exe', '-v'))

Concurrent Versions System (CVS) 1.9 (client)

Copyright (c) 1993-1994 Brian Berliner
Copyright (c) 1993-1994 david d `zoo' zuhn
Copyright (c) 1992, Brian Berliner and Jeff Polk
Copyright (c) 1989-1992, Brian Berliner

CVS may be copied only under the terms of the GNU General 
Public License,
a copy of which can be found with the CVS distribution kit.
0
>>>

If you really want to understand this crap, look at 
win32process.CreateProcess(), because every method of 
method of starting a new process on Windows ends up 
making this call; and the cruntime lib (*nix emulation) calls are 
all hopelessly braindead in some way or other.

- Gordon



From aa8vb at yahoo.com  Sat Apr  1 08:25:13 2000
From: aa8vb at yahoo.com (Randall Hopper)
Date: Sat, 01 Apr 2000 08:25:13 -0500
Subject: Emacs Python mode
In-Reply-To: <yayk8iii0ex.fsf@ganglot.ifi.uio.no>
References: <yayk8iii0ex.fsf@ganglot.ifi.uio.no>
Message-ID: <20000401082513.B1353928@vislab.epa.gov>

Vetle Roeim:
 |I just had an idea. Using the Emacs python-mode, would it be possible
 |to somehow hide the function-bodies?
 |
 |Sometimes it would be useful to get an overview of the functions
 |provided by a class.

Your prayers have been answered.  See Ken Manheimer's outdent.el for Emacs
(http://www.python.org/emacs/).  Collapsable Python blocks, block skips,
etc.

...that is, once they get http://www.python.org back up:

   TCP connection to 'www.python.org' failed: Broken pipe. 

-- 
Randall Hopper
aa8vb at yahoo.com



From gary_thomson at yahoo.com  Sun Apr 30 07:06:23 2000
From: gary_thomson at yahoo.com (gary thomson)
Date: Sun, 30 Apr 2000 11:06:23 GMT
Subject: IDLE & Win 95
Message-ID: <390c1265.4105818@news.force9.net>

I've installed Python 1.5.2 and Tcl/Tk on my machine. 

I wanted to use IDLE. Downloaded it from the net and overwrote the
IDLE folder in the Tools directory. When I go to IDLE (Python GUI) in
the Start menu, nothing happens. Similarly double clicking idle.pyw. 

What makes this more mysterious is that I managed to do this at work
(under Windows 98) without any problems. 

IDLE is cool! Hence I'd like to be able to work with it at home. can
anyone see what I need to do to get it working?

TIA
Gary Thomson

garyt at garyt.force9.co.uk


Try MindProbe. Let your PC read your mind

Visit http://www.garyt.f9.co.uk


From python at rose164.wuh.wustl.edu  Tue Apr 25 06:32:09 2000
From: python at rose164.wuh.wustl.edu (David Fisher)
Date: Tue, 25 Apr 2000 05:32:09 -0500
Subject: serial port and event loop question
References: <8doe40$h8$1@nnrp1.deja.com>
Message-ID: <003101bfaea1$82a3be00$0201a8c0@spkydomain>

----- Original Message -----
From: <bgue at my-deja.com>
Newsgroups: comp.lang.python
To: <python-list at python.org>
Sent: Thursday, April 20, 2000 9:24 PM
Subject: serial port and event loop question


> Hi everyone. This is mostly a general programming question, but since
> I'm using Python for this project, I thought I'd ask here... I wrote
> database written with Python and Gadfly, and I've got a nice little
> Tkinter GUI for it. I need to make it monitor the serial port for
> incoming data (from a barcode scanner). I'm using the Serial module,
> but I don't know how to have my program watch for new data
> constantly...any help?
>


Hi,
If you're talking about the win32 serial io module, you don't need to do
anything special for it to watch for new data.  It buffers the incoming data
until you read() or readTerminated() to get the data.
Good luck,
David




From chafik at nevrax.com  Mon Apr 10 04:11:25 2000
From: chafik at nevrax.com (Sameh chafik pro)
Date: Mon, 10 Apr 2000 10:11:25 +0200
Subject: (no subject)
Message-ID: <017d01bfa2c4$5d4d4ce0$1001a8c0@nevrax.net>

Hello i want to erase my existing subscription from the the mail list
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20000410/cdb5624e/attachment.html>

From jschmittNOjsSPAM at vmlabs.com.invalid  Fri Apr  7 14:31:29 2000
From: jschmittNOjsSPAM at vmlabs.com.invalid (John Schmitt)
Date: Fri, 07 Apr 2000 11:31:29 -0700
Subject: Tkinter install?
Message-ID: <081c4ffb.329993cd@usw-ex0102-014.remarq.com>

What environment variables are needed to use Tk/Tkinter?  Are
you supposed to be able to edit the environment variables from
within PythonWin's Tools/EditPytonPath menu?  I can't seem to
change anything from there.  I really want to point the relevant
variables to the right path, but don't know how to do that with
PythonWin.  Any advice?

What brought this question on is this:  I did a clean install of
py152 (NT4, if it matters) and did the install of tcl/tk.  I
have Grayson's Tkinter book, and I'm trying to run one of the
examples.  I get this message in Tkinter.py:886:

-------------------
TclError: Can't find a usable init.tcl in the following
directories:
    {} ./lib/tcl8.0 D:/pyprojects/grayson/tcl8.0/library
D:/pyprojects/grayson/Examples/library

This probably means that Tcl wasn't installed properly.
-------------------

Thanks.

John


* Sent from RemarQ http://www.remarq.com The Internet's Discussion Network *
The fastest and easiest way to search and participate in Usenet - Free!



From tjg at avalongroup.net  Tue Apr  4 17:49:32 2000
From: tjg at avalongroup.net (Timothy Grant)
Date: Tue, 04 Apr 2000 14:49:32 -0700
Subject: Tkinter, <select> style menu
References: <8cdqom$aqp$1@news101.telia.com>
Message-ID: <38EA636C.EB62A9BF@exceptionalminds.com>

Anders Eggers-Krag wrote:
> 
> Is there really no reasoanble way to make a box similar to that of the HTML
> <SELECT>
> in tkinter?
> 
> the nearest I have been able to find is from Pmw, and it doesn't really look
> like windows...

Hmm, I thought that the Pmw.ComboBox() when used in drop down mode
looked almost just like the Windows Drop Down box.

What is your problem with it?

-- 
Stand Fast,
    tjg.

Timothy Grant                         tjg at exceptionalminds.com
Chief Technology Officer              www.exceptionalminds.com
Red Hat Certified Engineer                      (503) 246-3630
Avalon Technology Group, Inc.              fax  (503) 246-3124
>>>>>>>>>>>>Linux...Because rebooting isn't normal<<<<<<<<<<<<



From m.faassen at vet.uu.nl  Wed Apr  5 14:00:45 2000
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 5 Apr 2000 18:00:45 GMT
Subject: Two questions
References: <38EA5FE8.853E9B75@callware.com> <01f801bf9e80$5a522a80$29dc9f26@unitedelectric> <38EA6517.978953E4@callware.com>
Message-ID: <8cfv0d$l0e$1@newshost.accu.uu.nl>

Ivan Van Laningham <ivanlan at callware.com> wrote:
> Query for information, using a not-very-complicated schema.  Read a file
> and, after massaging the data, stuff the information into the db.

> As reading the file progresses and the information gets transferred to
> the db, update a progress indicator (that's the extent of the GUI;-)

I have been using Win32all's ODBC support successfully for a long time
now. Just install Win32all, create a system DSN for your access database
with the Windows odbc manager, and do:

import odbc, dbi

..after that you can use the Python DB API (version 1) to do SQL queries.

Regards,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From claird at starbase.neosoft.com  Mon Apr 24 18:46:11 2000
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 24 Apr 2000 17:46:11 -0500
Subject: Why doesn't O'Reilly use parrots on its python books?
References: <etdvh186ztw.fsf@w20-575-115.mit.edu> <B5293E3C.255C%dgoodger@bigfoot.com>
Message-ID: <FEAC06345F08C2BE.74A6EF6C28FDB863.AED7C7EC5CB151E7@lp.airnews.net>

In article <B5293E3C.255C%dgoodger at bigfoot.com>,
David Goodger  <dgoodger at bigfoot.com> wrote:
>How about half a bee?
>
>If they ever run out of fauna and need to go flora, the only choice would
>be: the larch.
>

Pruned into shrubbery.
-- 

Cameron Laird <claird at NeoSoft.com>
Business:  http://www.Phaseit.net
Personal:  http://starbase.neosoft.com/~claird/home.html


From thomas at xs4all.net  Wed Apr 12 01:58:15 2000
From: thomas at xs4all.net (Thomas Wouters)
Date: Wed, 12 Apr 2000 07:58:15 +0200
Subject: Huh? func_defaults, default values in function calls
In-Reply-To: <38F3EBBF.7B23@marshall.edu>; from hathawa2@marshall.edu on Tue, Apr 11, 2000 at 09:17:10PM -0400
References: <8cv5qj$35p$1@zinc.intern.netconnect.no> <8cv8ra$hio$1@pegasus.csx.cam.ac.uk> <nMGI4.42394$U4.182659@news1.rdc1.az.home. <38F3EBBF.7B23@marshall.edu>
Message-ID: <20000412075815.A14644@xs4all.nl>

On Tue, Apr 11, 2000 at 09:17:10PM -0400, Mark Hathaway wrote:

> > I think this is the first time I've seen show up with dictionaries, usually
> > is shows up with lists.

> At least this idiom gets the job done, but it sincerely defeats the
> idea of having default values to have to say "z=None".

Not really. I agree that it is slightly counter-intuitive, which is very
suprising in Python, but it adds a pretty valuable construct -- that of a
'static' variable in C. I dont think there is another way to do that, except
global variables or using an explicit namespace. As for the use of default
values, the use stays in the case of immutable values, and also for mutable
values which you do not mutilate.

mutable-as-in-mutiny?-ly y'rs,
-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From soundhack at hotmail.com  Thu Apr 20 07:14:31 2000
From: soundhack at hotmail.com (Robert)
Date: Thu, 20 Apr 2000 07:14:31 -0400
Subject: win32all-128.exe availability?
References: <38fd01b2$0$10982@senator-bedfellow.mit.edu> <3k9qfs0ropdsv515f9he7lkpdrf1qfputv@4ax.com>
Message-ID: <38fee74c$0$10971@senator-bedfellow.mit.edu>

Well  thanks to calishar's patience, I finally have it. It's available to
everyone
on www.idrive.com/soundhack  Thanks again to calishar!

Robert
Calishar wrote in message <3k9qfs0ropdsv515f9he7lkpdrf1qfputv at 4ax.com>...
>On Tue, 18 Apr 2000 20:49:08 -0400, "Robert" <soundhack at hotmail.com>
>wrote:
>
>>Now that starship has crashed, I was wondering if it was possible to
>>get the 128 build of Pythonwin extensions off the net? By an unfortunate
>>coincidence, my own hard drive chose to crash around the time starship
>>did.
>
>Don't know if any sites have a mirror of win32all-128, but I have it
>on my HD (love being a packrat, still have -127 too). If you want I
>can try e-mailing it to you. I say try because the file is 3.2M and
>@home gives me a 5M message size limit.
>
>Let me know if you want it via mail.
>
>Calishar
>
>calishar at home.com




From dalcolmo at mediaone.net  Sat Apr 22 13:43:21 2000
From: dalcolmo at mediaone.net (Josef Dalcolmo)
Date: Sat, 22 Apr 2000 17:43:21 GMT
Subject: editor for win32
References: <NpHL4.779$fK2.113609@news.flash.net>
Message-ID: <ZwlM4.11399$QX1.482571@typhoon.ne.mediaone.net>

Scott Hathaway <slhath at flash.net> wrote in message
news:NpHL4.779$fK2.113609 at news.flash.net...
> Can anyone recommend a python editor for Windows?
>
> If there is an IDE for windows (other than IDLE), even better.

Yes, there is Pythonwin (download and install(run) Win32all.exe). Thats a
full IDE

If you just want an editor any of the ones others already recommended are
good, but beware that Emacs is huge (20+ MB), and not trivial to learn,
although powerful.

Vim is good too, and some versions can use Python as an extension language,
but again, unless you come from the unix world it may appear rather strange
to you and your fingers.

PFE is a bit less powerful than the two above, but is a real Windows
application, easy to learn and has some quite neat features, I use it a lot.

If you like Emacs, but don't want the 20MB monster and don't need Lisp, you
can use Jed, which comes with a good Python mode too.

I use all of the above at times. PFE is the only one without a specific
Python mode, but I still use it the most, since I find it is the best
integrated with Windows (I use it as a Notepad replacement). For Python
programming I use mostly Win32all from the python site.

- Josef




From akuchlin at mems-exchange.org  Wed Apr 19 14:05:57 2000
From: akuchlin at mems-exchange.org (Andrew M. Kuchling)
Date: 19 Apr 2000 14:05:57 -0400
Subject: Pythonware Website???? Gone on vacation with Starship? ;-))
References: <20000419154202.B480@owl.rhein-zeitung.de> <JCjL4.540$wYl.187815424@newsb.telia.net>
Message-ID: <3d4s8yotiy.fsf@amarok.cnri.reston.va.us>

"Fredrik Lundh" <effbot at telia.com> writes:
>     www.pythonware.com is down
>     the reason appears to be hardware problems on
>     the ISP site. ...

First Zope.org (down because an Ethernet cable got pulled out), then
Starship (disk crash), now pythonware.com.  Has someone released evil
whitespace-eating nanobots that also attack computer hardware?

--amk


From mfletch at tpresence.com  Wed Apr 26 17:09:45 2000
From: mfletch at tpresence.com (Mike Fletcher)
Date: Wed, 26 Apr 2000 17:09:45 -0400
Subject: [Win32] Retrieve icon for file type?
Message-ID: <CC59868A0A76D311B1F50090277B321C1F32B6@VRTMAIL>

Well, I think I'm almost there, but I can't figure out how to create a
colour device context, and drawing a colour icon into the black-and-white
context just gives me a failed message.  Anyone know any tricks.

Incidentally, the IShellFolder/IShellIcon stuff looks like the "right" way
to do this, I just don't see any way to touch it from Python (apparently you
need the functions "SHGetDesktopFolder" etceteras in shell32.dll to get
anywhere with it.

Here's what I've got so far (based on Thomas' work).  Suggestions
appreciated...
Mike

def getIcon(srcpath):
	import os.path, string
	index = None
	try:
		srcpath, index = map (string.strip, string.split (srcpath,
','))
		index = int (index)
	except:
		pass
	print "PATH, INDEX", srcpath, index
	srcext = os.path.splitext (srcpath)[1]
	if string.lower (srcext) == '.ico':
		return None # CopyIcons_FromIco (dstpath, srcpath)
	import win32api, win32ui #, win32con
##    hdst = win32api.BeginUpdateResource (dstpath, 0)
	hsrc = win32api.LoadLibraryEx (srcpath, 0, LOAD_LIBRARY_AS_DATAFILE)
	if index is None:
		grpname = win32api.EnumResourceNames (hsrc,
RT_GROUP_ICON)[0]
	elif index >= 0:
		grpname = win32api.EnumResourceNames (hsrc,
RT_GROUP_ICON)[index]
	else:
		grpname = -index
	data = win32api.LoadResource (hsrc, RT_GROUP_ICON, grpname)
	for iconname in win32api.EnumResourceNames (hsrc, RT_ICON):
		icon = win32ui.GetApp().LoadIcon( iconname )
		bitmap = getBitmapFromIcon( icon )
	win32api.FreeLibrary (hsrc)
	return data

def getBitmapFromIcon( iconresource ):
	import win32ui
	try:
		dc = win32ui.GetActiveWindow().GetDC()
	except win32ui.error:
		dc = win32ui.CreateWnd().GetDC()
##	dc = win32ui.CreateDC()
	# create in-memory dc with display params...
	mdc = dc.CreateCompatibleDC( None )
	bitmap = win32ui.CreateBitmap()	bitmap.CreateCompatibleBitmap( mdc,
32, 32 )
	mdc.SelectObject( bitmap )
##	mdc.DrawIcon( (0,32), iconresource ) # fails :(
	# just for testing...
	bitmap.SaveBitmapFile( mdc, "test4.bmp" )
	return bitmap

if __name__ == "__main__":
	import sys
	getIcon( 'd:\\bin\\lang\\python\\python.exe' )

-----Original Message-----
From: Thomas Heller [mailto:thomas.heller at ion-tof.com]
Sent: Wednesday, April 26, 2000 12:22 PM
To: Mike Fletcher; Python Listserv (E-mail)
Subject: Re: [Win32] Retrieve icon for file type?


> I'm attempting to use the Windows registry to retrieve (and convert to a
> PNG/JPEG) the Windows icons for given files (so that a Windows-based
server
> can display file-type-specific icons).  I do not seem to be able to find
the
> function shellapi.ExtractIcon (
> http://msdn.microsoft.com/library/psdk/winui/icons_7h2m.htm ) in the Win32
> extensions.

You might want to look at
MEInc/Dist/icon.py in Gordon's win32 installer package.
This file uses LoadResource to extract icons from executable files.
(Note however that you get the raw data, and not an HICON).

Thomas


-- 
http://www.python.org/mailman/listinfo/python-list



From shapr at uab.edu  Fri Apr 28 04:53:51 2000
From: shapr at uab.edu (Shae Erisson)
Date: Fri, 28 Apr 2000 08:53:51 GMT
Subject: Another question
References: <dVFK4.41$vr2.5817@news.uswest.net> <38FB2EEB.A32AF88C@bam.com> <38FB339F.604364AD@bam.com> <200004171605.MAA04929@zippy.init.com> <38FB7610.80EAF965@ichips.intel.com> <200004172120.RAA11102@zippy.init.com> <jWNK4.432$vr2.32822@news.uswest.net> <m3ln2cckg6.fsf@atrus.jesus.cam.ac.uk> <p53L4.1217$zk4.22638@news.uswest.net> <20000418234932.L15664@xs4all.nl>
Message-ID: <3909543E.C775CAD6@uab.edu>

Thomas Wouters wrote:

> Of course not all Python code can be boiled down to these two things (they
> aren't really 'rules', by the way, merely descriptions of the Python view of
> the world) but understanding these subtle differences can really make you
> grok Python ;) Tim Peters once wrote down the Python Philosophy (or zen):

Can you tell me where you found the python philosophy quote? I'd like to
read more along these lines.
thanks,
-- 
sHae mAtijs eRisson (sHae at wEbwitchEs.coM) gEnius fOr hIre
   bRing mE fIve sQuirrels aNd nO oNe wIll gEt hUrt
13:00pm up 1 season, 1 squirrel, load average: 1+1j 0.5+2j 0.2+0.5j


From richard_chamberlain at ntlworld.com  Thu Apr  6 05:45:08 2000
From: richard_chamberlain at ntlworld.com (Richard Chamberlain)
Date: Thu, 6 Apr 2000 10:45:08 +0100
Subject: Installer / mxDateTime problem
Message-ID: <8chm9o$qbv$1@gxsn.com>

Hi!

I've got a simple script that uses mxDateTime that I want to run as an exe.
I run it through simple.py and I get the following error:

NameError: Case mismatch for module name feasts
(filename C:\Program Files\Python\DateTime\Feasts.py)

Sorry if this has been covered before, I'm a little slow to catch on!

Thanks,

Richard




From mhammond at skippinet.com.au  Wed Apr 12 10:00:18 2000
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Wed, 12 Apr 2000 14:00:18 GMT
Subject: Embedding Python in Visual C++ 6.0
References: <8d1he4$26d$1@nnrp1.deja.com> <N2YI4.2030$UB.13624@news-server.bigpond.net.au> <8d1od2$9fl$1@nnrp1.deja.com>
Message-ID: <Sj%I4.2225$UB.14895@news-server.bigpond.net.au>

"Matthias" <matthias_w at my-deja.com> wrote in message
news:8d1od2$9fl$1 at nnrp1.deja.com...

> Sorry - but I realy tried. I have the same probleme with your
> pythonwin.cpp sourcecode when I try to debug it. In
> <b>glue.DynamicApplicationInit(startup)</b>everything looks nice, but
> if I step into <b>glue.InitInstance()</b> I get the unhandled exception
> (see below). The Address where the debugger "land's" seems right
> (1E401F69) it should be the address of <b>BOOL Win32uiInitInstance()

You are probably loading win32ui.pyd instead of win32ui_d.pyd - this would
explain the crash and the lack of source code.

You may have win32ui.lib on the command line for your builds - if so, you
should remove it and it will magically work correctly.

Mark.





From tim.hochberg at ieee.org  Mon Apr 24 12:02:38 2000
From: tim.hochberg at ieee.org (Tim Hochberg)
Date: Mon, 24 Apr 2000 16:02:38 GMT
Subject: Need Help w/JPython deployment
References: <sg8pqclrcdo26@corp.supernews.com>
Message-ID: <ye_M4.72468$U4.433321@news1.rdc1.az.home.com>

"Scott Bortman" <scott at csihq.com> wrote in message
news:sg8pqclrcdo26 at corp.supernews.com...


> I am having a bit of difficulty using jpythonc.

[SNIP]

> But, I get the following error:

[SNIP]

> AttributeError: java package 'org.python.modules' has no attribute
> '__path__'
>
>
> Is there something I missed?

This is a jpythonc bug. You should download Finn Bock's errata which
contains a set of patches to fix various problems in jpythonc that have
surfaced since the last release. It can be found at
http://sourceforge.net/project/?group_id=1842. Once you have this it should
just be a matter of figuring out the correct set of options and it sounds
like you were most of the way there....

-tim






From oli at rz-online.net  Wed Apr 12 19:03:44 2000
From: oli at rz-online.net (Oliver Andrich)
Date: Thu, 13 Apr 2000 01:03:44 +0200
Subject: Python 1.6a2 crashes in test_fork1 on my Mandrake Linux 7 machine
Message-ID: <20000413010344.A13931@owl.rhein-zeitung.de>

Hi,

I just tried to build a first alpha release of the Linux distribution of
Python, and the testall.py run crashes on the test_fork1 script. Sometimes it
just hangs, as soon as I switch on full optimizations during compilation I get
a nice seg fault. The freezes occurs when I compile python without any but the
standard optimizations on my machine. I am running glibc 2.1.2 and gcc 2.95.2.
With 1.5.2 this works really fine.

I will try to investigate this issue. But for this I have to check the CVS
logs of the relevant files. Anybody with a useful hint should feel free to
mail me. ;-))

Best regards,

    Oliver



From moeller at networksplus.net  Mon Apr 10 17:42:29 2000
From: moeller at networksplus.net (Derek Moeller)
Date: Mon, 10 Apr 2000 16:42:29 -0500
Subject: Teaching the art of programming, in python
In-Reply-To: <landauer-D0C8CF.12100307042000@news.apple.com>; from landauer@apple.com on Fri, Apr 07, 2000 at 12:10:03PM -0700
References: <20000405053920.20579.qmail@web2102.mail.yahoo.com> <8cfe42$o02$1@nnrp1.deja.com> <dpJG4.3744$HG1.106895@nnrp1.uunet.ca> <38EB68A3.B1B5809F@callware.com> <8cfqa7$4ml$3@newsserver.rrzn.uni-hannover.de> <38EB73B6.6C0913A7@callware.com> <8chsij$k5r$1@newsserver.rrzn.uni-hannover.de> <slrn8er3fd.fcd.r2d2@shaka.acc.umu.se> <38EDDACC.8641FF8C@callware.com> <landauer-D0C8CF.12100307042000@news.apple.com>
Message-ID: <20000410163700.B20635@networksplus.net>

On Fri, Apr 07, 2000 at 12:10:03PM -0700, Doug Landauer wrote:
> > "Teach your self Internet in 10 minutes". Seems there is
> > inflation (or perhaps deflation) in these things.
> 
> I think I'll wait for the Cliff's Notes to come out ... that
> might get it down to 2 minutes.

Remember Number 7's reading performance with dictionaries in "Shortcircuit"?
That's my excuse for the title of my next vaporbook.

-- 
Derek Moeller
Learn Python 20 times faster
than those silly SAMS books
       promise!



From greg at cosc.canterbury.ac.nz  Wed Apr  5 00:28:50 2000
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Wed, 05 Apr 2000 16:28:50 +1200
Subject: Tuples -- who needs 'em
References: <00e101bf9d80$a8384d20$74eb0b18@stcla1.sfba.home.com> <va3G4.1110$n68.190454272@newsb.telia.net> <037f01bf9e55$6d328760$74eb0b18@stcla1.sfba.home.com>
Message-ID: <38EAC102.7D5DF504@cosc.canterbury.ac.nz>

Bob Alexander wrote:
> 
> BTW, both tuples and lists are ordered collections of non-homogeneous values
> (i.e. the elements can be if different types).

In Haskell, a statically-typed language which also has
both tuples and lists, the two types play distinctly
different roles, and both are needed. A tuple is a 
fixed-size structure whose elements may be of different 
types, whereas a list is a variable-size structure whose 
elements must be of the *same* type.

This is the way I tend to use them in Python as well --
lists for homogeneous collections, tuples for "records".

If static typing is introduced into Python, a more rigid
distinction may have to be made, in which case we may be
glad that we already have the two kinds of structure.

-- 
Greg Ewing, Computer Science Dept,
+--------------------------------------+
University of Canterbury,	   | A citizen of NewZealandCorp, a	  |
Christchurch, New Zealand	   | wholly-owned subsidiary of USA Inc.  |
greg at cosc.canterbury.ac.nz	   +--------------------------------------+


From aa8vb at yahoo.com  Mon Apr 17 12:13:50 2000
From: aa8vb at yahoo.com (Randall Hopper)
Date: Mon, 17 Apr 2000 12:13:50 -0400
Subject: Numeric-13 build failure on Python 1.6a2
Message-ID: <20000417121350.A316132@vislab.epa.gov>

     I'm getting a build failure with Numeric-13 with Python 1.6a2.  There
seem to be two problems at work here:

   ---------------------------------------------------------------------------
1) The first problem is the linker tries to link for the wrong architecture.

   When makesetup writes $(LDSHARED) command-lines, it doesn't include 
   $(LDFLAGS) on that command-line.  In Python's config/Makefile,
   LDFLAGS=-n32 here on IRIX.

   It's this line in makesetup that writes these link lines:

                rule="$rule; \$(LDSHARED) $objs $libs -o $file"

   If I add \$(LDFLAGS) to it, we're past that hurdle.

   ---------------------------------------------------------------------------
2) The second problem is that in Python's config/Makefile, it has:

      LDSHARED=       ld

   But it should be:

      LDSHARED=       ld -shared

   As a result, when Numeric goes to link a shared object, it inadvertantly
   tries to link it as an executable, which of course fails.
   Changing this declaration as described above fixes problem #2.

   ---------------------------------------------------------------------------

Before I add this to the bug database, could someone confirm that these are
the right fixes for this problem?

Thanks,

Randall



==============================================================================

ERROR #1:
...
        cc -n32   -I./Include -O -KPIC -OPT:Olimit=0 -I/home/rhh/bin/../software/python-1.6/include/python1.6 -I/home/rhh/bin/../software/python-1.6/PLATFORM/plat-irix646-n32/include/python1.6 -DHAVE_CONFIG_H -c ./Src/ufuncobject.c
        ld  _numpymodule.o arrayobject.o ufuncobject.o  -o _numpymodule.so
ld: FATAL   12 : Expecting o32 objects: _numpymodule.o is n32.
*** Error code 4 (bu21)

==============================================================================
ERROR #2:
...
        ld -n32   _numpymodule.o arrayobject.o ufuncobject.o  -o _numpymodule.so
ld32: ERROR   33 : Unresolved text symbol "Py_InitModule4" -- 1st referenced by _numpymodule.o.
ld32: ERROR   33 : Unresolved text symbol "PyModule_GetDict" -- 1st referenced by _numpymodule.o.
ld32: ERROR   33 : Unresolved text symbol "PyCObject_FromVoidPtr" -- 1st referenced by _numpymodule.o.
...

==============================================================================



From phd at phd.russ.ru  Mon Apr 24 12:15:19 2000
From: phd at phd.russ.ru (Oleg Broytmann)
Date: Mon, 24 Apr 2000 16:15:19 +0000 (GMT)
Subject: Trouble compiling PyGreSQL 2.4 on Linux
In-Reply-To: <g0sbh8jd.fsf@jumbo.skott.dk>
Message-ID: <Pine.LNX.4.21.0004241614020.7343-100000@fep132.fep.ru>

On Mon, 24 Apr 2000, Martin Sk?tt wrote:
> Seems like a great time for me to learn how to write a make file. I
> will post it to him one of the following days.

   You couldn't write a plain makefile - there are many different systems,
and installations, etc.
   Better, learn to write Distutil, and write one for PyGreSQL - that's a
Real Challenge!

Oleg.            (All opinions are mine and not of my employer)
---- 
    Oleg Broytmann      Foundation for Effective Policies      phd at phd.russ.ru
           Programmers don't die, they just GOSUB without RETURN.




From hanche at math.ntnu.no  Thu Apr  6 01:19:19 2000
From: hanche at math.ntnu.no (Harald Hanche-Olsen)
Date: 06 Apr 2000 01:19:19 -0400
Subject: Why should I switch to Python? - Infinity of Primes
References: <NDBBKEGCNONMNKGDINPFGEJDDFAA.infonuovo@email.com> <7F51E902E09F3F60.8F612607812ED16B.82CB8B41435E14FD@lp.airnews.net> <38EA369C.26A1C25@math.okstate.edu> <38EAC8C4.ADC25640@cosc.canterbury.ac.nz> <8cfq9f$2q5e$1@pc-news.cogsci.ed.ac.uk>
Message-ID: <pcoog7najs8.fsf@math.ntnu.no>

+ richard at cogsci.ed.ac.uk (Richard Tobin):

| In article <38EAC8C4.ADC25640 at cosc.canterbury.ac.nz>,
| Greg Ewing  <greg at cosc.canterbury.ac.nz> wrote:
| 
| >Um, no it doesn't - it constructs a number which is
| >*either* prime *or* divisible by some other prime bigger
| >than the one you started with.
| 
| It trivially gives you a procedure for finding a larger prime: you
| just find the factors of the constructed number (eg by trying all
| possible divisors), and if there are any then one will be a new prime,
| otherwise the constructed number is a new prime.

This thread is already much too long, but I can't resist pointing out
that the good old Erathostenes' sieve together with the ideas
introduced earlier in the thread provides a splendid constructive
proof:  Essentially, if the sieve has produced primes 2, 3, ..., p
then let q be one plus the product of 2, 3, ..., p.  Then the sieve
will produce a new prime no later than q.

Since this is a posted to a computer language group, after all, I
include my own simple implementation of the Sieve in Haskell:

primes :: [Integer]
primes = 2 : ([3..] `except` composites)

composites :: [Integer]
composites  = flatten (map (\ p -> [p*p, p*(p+1)..]) primes)

union :: [Integer] -> [Integer] -> [Integer]
xl@(x:xs) `union` yl@(y:ys) | x<y       = x : xs `union` yl
                            | x>y       = y : xl `union` ys
                            | otherwise = x : xs `union` ys

except :: [Integer] -> [Integer] -> [Integer]
xl@(x:xs) `except` yl@(y:ys) | x<y       = x : xs `except` yl
                             | x==y      = xs `except` ys
                             | otherwise = xl `except` ys

flatten :: [[Integer]] -> [Integer]
flatten ((x:xs):zz) = x : xs `union` (flatten zz)

I wonder if someone can come up with an equally immediate solution
(i.e., a straight translation from the mathematical formulation) in
python?

Indeed, I would be tempted to say that Haskell is better suited for
learning (some kinds of) mathematics than python.

but-I-wouldn't-dare-say-so-in-comp.lang.python-ly y'rs,
-- 
* Harald Hanche-Olsen     <URL:http://www.math.ntnu.no/~hanche/>
- "There arises from a bad and unapt formation of words
   a wonderful obstruction to the mind."  - Francis Bacon



From bjorn at roguewave.com  Tue Apr  4 17:56:24 2000
From: bjorn at roguewave.com (Bjorn Pettersen)
Date: Tue, 04 Apr 2000 15:56:24 -0600
Subject: Two questions
References: <38EA5FE8.853E9B75@callware.com>
Message-ID: <38EA6508.90BF630C@roguewave.com>

Ivan Van Laningham wrote:
> 
> Hi All--
> I'm finally getting a chance to use Python at work.  I have two
> questions:

Congratulations :-)

> 1)  Is there a way to run silent (i.e., we determine where things go and
> the customer never sees any wizards) installs of Python, Tcl/Tk and
> Win32all?

If you want to package up your program into an exe and execute it,
Installer from Gordon's starship pages can do this
(http://starship.python.net/crew/gmcm/distribute.html).  It's actually
_much_ easier than it looks ;-)

> 2)  What stuff do I need to use msaccess with Python?  Do I need to use
> Win32all, or can I use Tkinter for my (minimal) GUI, with msaccess?

You'll probably want to check out one of the odbc packages.  (I haven't
used any of them myself, but mxODBC seems to be highly recomended...)

-- bjorn



From m.faassen at vet.uu.nl  Fri Apr 21 09:11:14 2000
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 21 Apr 2000 13:11:14 GMT
Subject: Pickling Robustly
References: <jak8huf1of.fsf@helix.twistedmatrix.com>
Message-ID: <8dpk1i$7de$1@newshost.accu.uu.nl>

Glyph Lefkowitz <glyph at twistedmatrix.com> wrote:

[snip pickle is cool]
> Except...

> It appears that Pickle is not terribly robusta

[snip]
> Due to the extremely self-referential nature of the data structures
> I'm using, it is much easier to keep everything in one pickle (and
> that effectively happens anyway, even if I don't try to, thanks to the
> dependency resolution that pickle does). The problem is that pickle
> (as far as I can tell, anyway) does not allow one to register error
> handlers to keep track of when a problem occurrs.  If an unpicklable
> datatype is attempted to pickle, EVERYTHING previously pickled will
> suddenly vanish and no indication will be given about where the
> offending data was.

> If you've got a large data structure like I
> anticipate, this means writing and loading code on-the-fly to attempt
> a traversal of that structure to determine where the unpicklable type
> is.

The Digicool people must've run into similar problems with their Zope Object
Database, which is based on Pickle. Perhaps looking at the ZODB wouldn't
be a bad idea, if you haven't already, therefore:

http://www.python.org/workshops/2000-01/proceedings/papers/fulton/zodb3.html
http://www.zope.org/Members/michel/HowTos/ZODB-How-To
http://www.zope.org/Documentation/Developer/Models/ZODB

Though I'm not sure if they indeed have solved your problem, as far
as I understand it.

Regards,

Martijn



From lvw at lfpt.rwth-aachen.de  Wed Apr 19 08:01:01 2000
From: lvw at lfpt.rwth-aachen.de (Lars von Wedel)
Date: Wed, 19 Apr 2000 14:01:01 +0200 (MET DST)
Subject: Using CGI module with 'canned queries'
In-Reply-To: <38FD8C36.4242425F@infercor.no>
References: <200004190828.KAA03403@gromit.rwth-aachen.de>
	<38FD8C36.4242425F@infercor.no>
Message-ID: <200004191201.OAA04426@gromit.rwth-aachen.de>

Paul,

The %'s were obviously wrong, but they got mixed up when I
transferred the string from '%s:%s...' % (a,b,c) into the e-mail
where I wrote some bogus stuff.

Indeed I overlookes the .value attribute of the keys.

Thanks,
Lars


Paul Boddie writes:
 > Lars von Wedel wrote:
 > > 
 > > Hello,
 > > 
 > > I want to use Python to implement some CGI scripts. I want these
 > > scripts to be callable from any link (not only via a form),
 > > e.g. the user shall be able to enter something like
 > > 
 > > http://.../test.py?name=Bill
 > > 
 > > into Netscape directly.
 > 
 > The parameter(s) should still be received correctly in the CGI script.
 > 
 > > I expected that such a URL could be processed using
 > > 
 > >     form = cgi.FieldStorage()    print '<UL>'
 > >     for k in form.keys:
 > >         print '<LI>' % k % ':' % form[k] % '</LI>'
 > 
 > Surely you mean...
 > 
 >           print '<LI>' + k + ':' + form[k].value + '</LI>'
 > 
 > > (after printing an html header of course).
 > > 
 > > However, the dictionary seems to be empty when the http daemon
 > > starts my script.
 > 
 > Are you sure that the problem doesn't lie in the fact that form[k] is not likely
 > to be a string, but a FieldStorage object instead? Thus, you must refer to the
 > value attribute of form[k].
 > 
 > > How do I access parameters within the script if not started via a form?
 > 
 > You should be able to do just that with the 'cgi' module.
 > 
 > Paul
 > -- 
 > http://www.python.org/mailman/listinfo/python-list
 > 



From bwinton at tor.dhs.org  Tue Apr 18 12:53:11 2000
From: bwinton at tor.dhs.org (Blake Winton)
Date: Tue, 18 Apr 2000 16:53:11 GMT
Subject: Lists
References: <A63AFB20111ED311AC5500A0C96B7BF5015F0F2C@orsmsx36.jf.intel.com>
Message-ID: <slrn8fp4tv.i3d.bwinton@tor.dhs.org>

On Tue, 18 Apr 2000 09:26:34 -0700, Daley, MarkX wrote:
>	a = [-5, -4, -3, -2, -1, 0, 1, 2, 3, 4, 5]
>	for item in a:
>		try:
>			print b / a[item]
>		except ZeroDivisionError:
>			pass
>
>Why is the list being processed in reverse?

Hmmm... good question.
I don't think it is.  I think you're falling into a classic C trap.

Let's look at what the value of item and a[item] are in the loop.

               item     a[item]
Iteration 1     -5        1
Iteration 2     -4        2
Iteration 3     -3        3
...

The line you really wanted to write wasn't
  print b / a[item]
but instead was
  print b / item

Later,
Blake.
-- 
12:46pm up 20 days, 13:21, 1 user, load average: 1.00, 1.00, 1.00


From johngrayson at home.com  Sat Apr 29 09:18:11 2000
From: johngrayson at home.com (johngrayson at home.com)
Date: Sat, 29 Apr 2000 13:18:11 GMT
Subject: widget resizing
References: <8edrn7$nvh$1@nnrp1.deja.com>
Message-ID: <8eened$jns$1@nnrp1.deja.com>

In article <8edrn7$nvh$1 at nnrp1.deja.com>,
  ekw1 at my-deja.com wrote:
> can anyone point me to an example of how to have widgets resize to
> expand/contract when the window is resized?  Or tell me what are the
> relevant sections are in the python documentation?  are events and
> bindings the sections i should study?  i'm not familiar with the terms
> involved in doing this.  i am using python1.5.2 on win98se.  thanks
for
> any help.
>
> ekw
>
> Sent via Deja.com http://www.deja.com/
> Before you buy.
>

Look at the geometry managers, pack and grid.

Or chapter 5 of python and Tkinter Programming!

   jeg


Sent via Deja.com http://www.deja.com/
Before you buy.


From effbot at telia.com  Sun Apr  9 15:28:29 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Sun, 09 Apr 2000 19:28:29 GMT
Subject: tabular data
References: <8co4l2$5r2$1@nnrp1.deja.com> <O8OH4.46089$Uz3.487916@typhoon.midsouth.rr.com> <8cq9io$ca0$1@nnrp1.deja.com>
Message-ID: <xR4I4.275$rc9.188784640@newsb.telia.net>

poupaerte at my-deja.com wrote:
> > def sortedby(list,*indices):
>
> I didn't know Python worked with pointers, or else what does *indices
> mean? The fine manuals do not explain the *var construct ...

of course they do:

    http://www.python.org/doc/current/ref/function.html

    If the form "*identifier'' is present, it is initialized to a tuple
    receiving any excess positional parameters, defaulting to
    the empty tuple. If the form "**identifier'' is present, it is
    initialized to a new dictionary receiving any excess key-
    word arguments, defaulting to a new empty dictionary."

</F>




From dreid at jetnet.co.uk  Sat Apr 15 12:22:48 2000
From: dreid at jetnet.co.uk (David Reid)
Date: Sat, 15 Apr 2000 16:22:48 GMT
Subject: BSDi and PIL
Message-ID: <sH0K4.13$Ny3.1761@newreader.ukcore.bt.net>

Has anyone had any joy in getting Python 1.5.2 and PIL 1.0 to build together
on BSDi 4.0.1?

If so, how?

david




From scarblac-spamtrap at pino.selwerd.nl  Sat Apr 22 09:50:43 2000
From: scarblac-spamtrap at pino.selwerd.nl (Remco Gerlich)
Date: 22 Apr 2000 13:50:43 GMT
Subject: code not true?
References: <8ds3ku$q0l$1@nnrp1.deja.com> <3901aa0f.0@news.proweb.co.uk>
Message-ID: <slrn8g3bhf.1pb.scarblac-spamtrap@flits104-37.flits.rug.nl>

ng at hardlight.couk.com wrote in comp.lang.python:
> interestingly enough on both Win32 and Linux
> 
> >>> a = 0.07 * 100
> 
> >>> a
> 7.0
> 
> >>> a == 7.0
> 0

This is fixed in 1.6. It will show the "full" value of a, i.e.
>>> a = 0.07 * 100
>>> a
7.0000000000000009

Older versions tried to simplify but that just makes it harder to
understand...

-- 
Remco Gerlich,  scarblac at pino.selwerd.nl
Hi! I'm a .sig virus! Join the fun and copy me into yours!


From hniksic at iskon.hr  Tue Apr 11 08:15:06 2000
From: hniksic at iskon.hr (Hrvoje Niksic)
Date: 11 Apr 2000 14:15:06 +0200
Subject: [ANNOUNCE] Garbage collection for Python
References: <20000407221750.A29279@acs.ucalgary.ca> <38F13FB9.D1D16AF@mindspring.com> <slrn8f2if1.dop.nascheme@cranky.arctrix.com> <87r9cdhdb0.fsf@deneb.cygnus.argh.org> <FsunGI.BtB@world.std.com>
Message-ID: <dphfd87s1h.fsf@mraz.iskon.hr>

wware at world.std.com (Will Ware) writes:

> Florian Weimer (fw at deneb.cygnus.argh.org) wrote:
> > Is it possible to extend your GC in a way which would permit the
> > execution of several bytecode interpreters in multiple threads?
> 
> It sounds like you might want to simply run several copies of the
> Python interpreter in separate processes, which isn't a GC issue.

Maybe he might want that, but that's not what he said.  Separate
processes will not provide the sharing semantics needed for a
multi-threaded program to work.

C programs can at least fake that by using shared memory or System V
message IPC, but that won't work in Python.

> If memory isn't really that big a deal, and you don't need thousands
> of threads running simultaneously, you can use the thread/threading
> modules, which use whatever threading is provided by your OS.

This is still different from being able to embed multiple Python
instances in threads.


From breiter at usf.Uni-Osnabrueck.DE  Thu Apr  6 07:31:31 2000
From: breiter at usf.Uni-Osnabrueck.DE (Bernhard Reiter)
Date: 6 Apr 2000 11:31:31 GMT
Subject: Teaching the art of programming, in python
References: <20000405053920.20579.qmail@web2102.mail.yahoo.com>
    <8cfe42$o02$1@nnrp1.deja.com> <dpJG4.3744$HG1.106895@nnrp1.uunet.ca>
    <38EB68A3.B1B5809F@callware.com>
    <8cfqa7$4ml$3@newsserver.rrzn.uni-hannover.de>
    <38EB73B6.6C0913A7@callware.com>
Message-ID: <8chsij$k5r$1@newsserver.rrzn.uni-hannover.de>

In article <38EB73B6.6C0913A7 at callware.com>,
	Ivan Van Laningham <ivanlan at callware.com> writes:
> Bernhard Reiter wrote:

>> And all in 24 hours?
>> This is as misleading as the dummy attribute.

> I don't think so.  You can go at your own speed.  The book is broken up
> into 24 chapters, each taking more or less than an hour to read.  If you
> pay attention and apply yourself, you could, I think, arrive at some
> degree of understanding in the overall 24-hour time frame--give or take
> a few hours.

Well I guess that a good speed would be 14 days, then. :)
With two hour lections plus time to exercise.

> I picked up the basics of Python in perhaps four hours.  That's a
> tribute to the language, not me.  Given an interested, attentive newbie,
> I think it's entirely possible to teach programming and expose the
> newbie to some of the finer things in life in 24 hours or so--with
> Python.
> 
>> Ranting-about-dummy-booktitles-ly-yours,
Well I am just ranting about book title giving wrong impressions.
24 times a one hours session can be quite a bit to teach people the
basics, I agree with that. But for all the philosophy I guess that
you need to spread the lectures about a larger time period. :)

> <ranting-is-misleading>-ly y'rs,
Splitting-hair-for-less-marketing-hype-ly y'rs,
	Bernhard
-- 
Professional Service around Free Software                (intevation.net)  
The FreeGIS Project				            (freegis.org)
Association for a Free Informational Infrastructure            (ffii.org)


From jorn at spekter.no  Fri Apr 14 04:59:48 2000
From: jorn at spekter.no (=?iso-8859-1?Q?J=F8rn?= Eilertsen)
Date: Fri, 14 Apr 2000 10:59:48 +0200
Subject: Cron problems with pipes
Message-ID: <38F6DE03.6EB8B08B@spekter.no>

Hi,

I have this little program which includes the following line og code:
 mounts = commands.getoutput('df | grep eggum | cut -d 1 -f 1 | cut -d /
-f 4')

this runs fine from the commandline but when run as a cronjob I get the
following error:

File "/usr/lib/python1.5/commands.py", line 42, in getoutput
    return getstatusoutput(cmd)[1]
  File "/usr/lib/python1.5/commands.py", line 53, in getstatusoutput
    sts = pipe.close()
IOError: (10, 'No child processes')

Any help on how the get around this would be greatly appreciated.

Thanks,
Jorn



From alwagner at tcac.net  Fri Apr 21 23:39:27 2000
From: alwagner at tcac.net (Albert Wagner)
Date: Fri, 21 Apr 2000 22:39:27 -0500
Subject: The Simple Economics of Open Source
References: <14591.15630.192472.440375@goon.cnri.reston.va.us> <slrn8fv2rd.1kt.neelk@brick.cswv.com> <3900611F.3B525BDA@cfdrc.com>  <FtDG4z.774@world.std.com> <slrn8g26tu.p7o.neelk@brick.cswv.com>
Message-ID: <39011EEF.4BAEC3C9@tcac.net>

You are absolutely right about altruism.  Fortunately, altruism isn't
required for cooperative strategies to evolve in a society.  All that is
required is that a "critical mass" of participants realize that the
shared rewards of long term cooperation exceed the rewards of short term
reward at the expense of others.  Those with uncooperative strategies
view the pie as finite and feel that they must optimize their slice by
diminishing the share of others.  Those with cooperative strategies have
discovered that there are an infinite number of pies. 
-- 
Small is Beautiful


From ngps at madcap.dyndns.org  Fri Apr 21 05:52:22 2000
From: ngps at madcap.dyndns.org (Ng Pheng Siong)
Date: 21 Apr 2000 09:52:22 GMT
Subject: socket.SO_REUSEADDR
References: <38FEEA44.6C05E2B0@inka.de> <8dn9o5$jsp$1@mawar.singnet.com.sg> <38FF58F5.E1464322@inka.de>
Message-ID: <8dp8cm$1cm$1@violet.singnet.com.sg>

According to Michael Str?der  <michael.stroeder at inka.de>:
> Ng Pheng Siong wrote:
> >     self.socket.setsockopt(socket.SOL_SOCKET, socket.SO_REUSEADDR, 1)
> 
> And with M2Crypto? (AttributeError: setsockopt)

Hmm, gotta add the method, then.

But M2Crypto already does it for you:

class Connection:
    def __init__(self, ctx, sock=None):
        self.ctx=ctx
        self.ssl=m2.ssl_new(self.ctx.ctx)
        if sock is not None:
            self.socket=sock
        else:
            self.socket=socket.socket(socket.AF_INET, socket.SOCK_STREAM)
            # XXX debugging
            self.socket.setsockopt(socket.SOL_SOCKET, socket.SO_REUSEADDR, 1)

Cheers.

-- 
Ng Pheng Siong <ngps at post1.com> * http://www.post1.com/home/ngps



From mwh21 at cam.ac.uk  Sat Apr  1 08:18:48 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 01 Apr 2000 14:18:48 +0100
Subject: Emacs Python mode
References: <yayk8iii0ex.fsf@ganglot.ifi.uio.no>
Message-ID: <m34s9mot6v.fsf@atrus.jesus.cam.ac.uk>

Vetle Roeim <vetler at ifi.uio.no> writes:

> I just had an idea. Using the Emacs python-mode, would it be possible
> to somehow hide the function-bodies?

Try "M-8 C-x $", which isn't the easiest thing to type, but it works
reasonably well.

It's worth readong through the extended help on python-mode (type "C-h
C-m"), which I did yesterday and realised that there were lots of
shortcuts for handy things I never knew existed.  Of course I now
can't remember what the key combinations for any of these cool things
were.

Cheers,
M.

-- 
well, take it from an old hand: the only reason it would be easier
to program in C is that you can't easily express complex  problems
in C, so you don't.                 -- Erik Naggum, comp.lang.lisp


From bawolk at pacbell.net  Tue Apr 18 12:26:55 2000
From: bawolk at pacbell.net (Bruce Wolk)
Date: Tue, 18 Apr 2000 09:26:55 -0700
Subject: Integer solutions to linear equation?
References: <slrn8fovo2.nq.grant@grante.comtrol.com>
Message-ID: <38FC8CCF.B8B786CB@pacbell.net>

ax + by = c, where a, b, and c are integers, and the solution (x,y) must
be an integer, is one of the simplest Diophantine equations.  There is
an algorithm, but at the moment I can't remember it.  I do recall that
if the greatest common divisor of a and b divides c, then there are
infinitely many solutions, otherwise there are none.  Thus, if c = 1,
there are always infintely many solutions if a and b are relatively
prime, as in your example.

You probably can find the algorithm in any basic book on number theory.

Bruce



Grant Edwards wrote:
> 
> This isn't really a Python question, but my example is in
> Python, and there seem to be plenty of people who know a fair
> bit of math here....
> 
> A friend of mine ran across a brain-teaser involving a bunch of
> flowers, some magic bowls and some other camoflage text.  What
> you end up with is having to solve the equation
> 
>      64x = 41y + 1
> 
> Where x and y are both integers.  After scratching our heads
> for a while, we used the brute force approach:
> 
> for x in range(1,100):
>     y = ((64*x)-1)/41
>     if 64*x == 41*y+1:
>         print (x,y)
> 
> The results:
> 
>   (25, 39)
>   (66, 103)
> 
> 25 was the expected solution, so we got both the equation and
> the Python snippet correct.  Is there a non-iterative way to
> solve a linear equation when the results are contrained to be
> integers?  I don't remember anything from any of my math
> classes that seems relevent, but I didn't take any anything
> beyond what is required for all undergrad engineers.
> 
> --
> Grant Edwards                   grante             Yow!  I didn't order
>                                   at               any WOO-WOO... Maybe a
>                                visi.com            YUBBA... But no WOO-WOO!


From jon at dgs.monash.edu.au  Sun Apr 16 20:28:46 2000
From: jon at dgs.monash.edu.au (Jonathan Giddy)
Date: Mon, 17 Apr 2000 10:28:46 +1000 (EST)
Subject: else clauses in while and for loops
In-Reply-To: <38F7B350.5D6548E1@mindspring.com> from "Chuck Esterbrook" at Apr 14, 2000 08:09:52 PM
Message-ID: <200004170028.KAA07736@nexus.csse.monash.edu.au>

Chuck Esterbrook declared:
>
>On another topic: With regards to the "try...except" technique above 
>(which I've seen quite a bit in Python) does the following code do the 
>same thing and if so, which do you find more readable?

It does do the same thing with the extra line included below.  Now that
I've looked at this code in more detail than usual, I would probably choose
your form over my original, mainly on the grounds that it's quite likely
for the test to fail, so it's not really "exceptional" behaviour.

However, for code that is executed once over a limited number of
iterations, I'm not losing sleep either way.
>
>_tmpdirs = ['/tmp', '/var/tmp']
>
>for envname in ['TMPDIR', 'TMP', 'TEMP']:
>    if os.environ.has_key(envname):
	 path = [os.environ[envname]]
>        basepaths = validate_paths(path)
>        if basepaths:
>            break
>else:
>    basepaths = validate_paths(_tmpdirs)

Jon.



From mra at pobox.com  Sun Apr 23 23:20:54 2000
From: mra at pobox.com (Mark Atwood)
Date: 23 Apr 2000 20:20:54 -0700
Subject: Python - Next Release Questions
References: <NDBBKEGCNONMNKGDINPFEEJCDEAA.infonuovo@email.com>
Message-ID: <m3ya64nq09.fsf@flash.localdomain>

"Dennis E. Hamilton" <infonuovo at email.com> writes:
> issues and locale issues to deal with.  There are also language communities
> that have difficulty with Unicode for cultural as well as application
> reasons.

I.E., the various cultural groups that dislike the Unicode Han
unification because they wish to continue their racial illusion that
C, J, & V characters are completely different and seperate sets, and
that it's somehow corrupting to stir them together.

-- 
Mark Atwood   | It is the hardest thing for intellectuals to understand, that
mra at pobox.com | just because they haven't thought of something, somebody else
              | might. <http://www.friesian.com/rifkin.htm>
http://www.pobox.com/~mra


From portaldata at my-deja.com  Tue Apr 25 12:12:37 2000
From: portaldata at my-deja.com (portaldata at my-deja.com)
Date: Tue, 25 Apr 2000 16:12:37 GMT
Subject: how to instance shared objects ?
Message-ID: <8e4g5b$efs$1@nnrp1.deja.com>

I am trying to have python call a tibco api, either dll or shared
object.  How do I instance these shared objects (load them) ?

This is a C API, so I guess python can just call these api's once
loaded, right ?

thanks!


Sent via Deja.com http://www.deja.com/
Before you buy.


From gmcm at hypernet.com  Sat Apr 29 09:06:04 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Sat, 29 Apr 2000 09:06:04 -0400
Subject: The Simple Economics of Open Source
In-Reply-To: <B52E8CEC.5222%raffael@mediaone.net>
References: <1255228265-1994958@hypernet.com>
Message-ID: <1255108931-9171946@hypernet.com>

Raffael Cavallaro wrote:


> Oh really? Then why has no competitor yet figured out a
> *reliable* way to read and edit MS Word format documents? Just
> because the IP involved is intentional obfuscation, doesn't mean
> that "there's very little of that in code."

http://www.sped.ukans.edu/~justin/word/wword8.html
which is taken from (online) msdn docs.


 
> .... From an economic
> standpoint, the closed source model is a big winner. It is the
> anemic open source model, which must rely on service revenues for
> any profit that looks sick when it comes to profitability.

Consumer behavior (particularly what they're willing to pay for 
a product that which will no longer be supported) indicates 
that the value is based on future support and upgrades. 

Consumers are used to paying X up front, then something like 
50% of X in few years for the upgrade. Corps are used to 
paying X up front and 15% of X annually and then renogotiate 
at upgrade time. The open source model is strange to both of 
them.

Eventually, buyers and sellers will evolve to a common 
understanding of their common business. I don't know that the 
result will be the current open source model; but this model is 
a large advance in that understanding.

That doesn't mean there won't be rip-offs and gouging. In 
estabished markets, that's normally done by controlling the 
distribution chain. Software presents a unique challenge in 
this regard, but people are creative.

> If they produce software that:
> 
> 1. cannot be easily reverse engineered
> 
> 2. achievs sufficient market penetration that interoperability
> with others becomes a significant feature
> 
> then yes, they are effective monopolies. This is *precisely* how
> MS got where they are.

If that were the model, we'd all be using Apples. MS got where 
they are by outmanuvering IBM and DR. If Gary Kildall had 
been willing to play hardball with a bunch of IBM suits, the 
world would be quite different.

Corp's standardize on office suites for support, licensing and 
interoperability. Some standardize on Lotus SmartSuite. Face 
it, MS Office is significantly better, and is more open to 
working with third party software.
 
> >> BTW, all software is *not* support. This is another open
> >> source community bias which comes from living in the server
> >> room. Lots of software is *client* software, which requires
> >> little or no support from the vendor.
> > 
> > And client software is often free.
> 
> Only as a loss leader, or a monopolist's attempt to undercut
> potential competition (Netscape for the former, MSIE for the
> latter)

So can I conclude that even according to the "monopolists", 
client software is better considered a service? That's a start...


- Gordon



From eq3pvl at eq.uc.pt  Fri Apr 14 14:30:34 2000
From: eq3pvl at eq.uc.pt (Pedro Vale Lima)
Date: Fri, 14 Apr 2000 18:30:34 +0000
Subject: starship.python.net
References: <Pine.OSF.4.10.10004111721150.9584-100000@prfdec.natur.cuni.cz> <Pine.OSF.4.10.10004111740110.11294-100000@prfdec.natur.cuni.cz> <m3itxkg41s.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <38F763CA.43F8EA85@eq.uc.pt>

Michael Hudson wrote:

> Michal Bozon <bozon at natur.cuni.cz> writes:
>
> > > What the hell is with the starship.python.net server? Today and yesterday
> > > I cannont connect (from Czech Republic).
> > > Someone knows?
> > >
> > > Michal Bozon
> > >
> >
> > Just after I posted this message I tried connection again, and it is OK.
>
> Really?  Still seems to be wedged from here.  I can ping it, but can't
> get it to play on any other port.

I still can't connect. I need to get a module from there... I had it in my
computer, then
I deleted 'cause I could always go back to starship when I needed it.

Python needs something like CPAN, with lot's of mirrors. I know it's comming...
but
faster please ...

pedro lima



From effbot at telia.com  Tue Apr 25 09:25:38 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Tue, 25 Apr 2000 13:25:38 GMT
Subject: Efficient coding advice?
References: <8e2sjn$te4l$1@swen.emba.uvm.edu>
Message-ID: <m1hN4.3997$Za1.59327@newsc.telia.net>

Kelvin Chu <kelvin.chu at nospam.uvm.edu> wrote:
> 2.  Is there an easy way to turn an entire array of characters into
>     their ord value?

http://www.python.org/doc/essays/list2str.html
(see the epilogue)

</F>




From hathawa2 at marshall.edu  Tue Apr 11 21:10:52 2000
From: hathawa2 at marshall.edu (Mark Hathaway)
Date: Tue, 11 Apr 2000 21:10:52 -0400
Subject: Functions and objects
References: <meh9-C7461E.15255011042000@news.cit.cornell.edu>
 <38F385D1.9ED3FACC@yale.edu>
Message-ID: <38F3EA43.2C6E@marshall.edu>

>> Matthew wrote:
>>
>> I'm confused.  Take a look at the following:

a = [5]

>> for x in range(5):     # Case A
>>     temp=a
>>     print a
>>     temp.append(1)
       print temp        # to see what's happening to it

>> [5]
>> [5, 1]
>> [5, 1, 1]
>> [5, 1, 1, 1]
>> [5, 1, 1, 1, 1]

a = [5]

>> for x in range(5):     # Case B
>>     temp=a[:]
>>     print a
>>     temp.append(1)
       print temp        # to see what's happening to it

>> [5]
>> [5]
>> [5]
>> [5]
>> [5]


>> In case A, why doesn't temp reset itself to the value of a, [5], that
>> was predetermined before it entered the loop?  Why do you need to copy
>> the list to get the behavior I'm looking for in Case B?  Doesn't a
>> always equal [5]?

> Assignment always stores a reference to an object, it DOES NOT make a
> copy of the object.

It doesn't default to making a copy. you have to specify if you want a
copy!

So,

    temp = a

isn't the same as

    temp = a[:]

The first copies the address, stored as a simple variable in a, which
points to the object [5]

The second doesn't copy the address, it creates a new object with a copy
of what a pointed to.

This is one of the oddities (IMO) of Python.

It's about like saying temp = a[:] is temp = newOjbect([5])
wheras                 temp = a    is temp = addressOf([5])

I know it was all simple when you could write (in most languages)

    a = 5
    temp = a

and temp is a new separate different variable with a value of 5.
That's not quite what's happening in Python. Study the difference
carefully.


Mark S. Hathaway
email: hathawa2 at marshall.edu


From moshez at math.huji.ac.il  Sat Apr 15 01:36:59 2000
From: moshez at math.huji.ac.il (Moshe Zadka)
Date: Sat, 15 Apr 2000 07:36:59 +0200 (IST)
Subject: for item in some_class_object...
In-Reply-To: <4j6ffs0b6ins12qq559q07u28pc6o0mc9l@4ax.com>
Message-ID: <Pine.GSO.4.10.10004150736390.11261-100000@sundial>

On Fri, 14 Apr 2000, Ben Wolfson wrote:

> What special methods should I define for a class if I want to support
> a for item in ... syntax for it?

__getitem__
--
Moshe Zadka <mzadka at geocities.com>. 
http://www.oreilly.com/news/prescod_0300.html
http://www.linux.org.il -- we put the penguin in .com




From helpdesk at perceval.be  Wed Apr 19 05:58:09 2000
From: helpdesk at perceval.be (Gregoire Welraeds)
Date: Wed, 19 Apr 2000 11:58:09 +0200 (CEST)
Subject: testing true condition
Message-ID: <Pine.BSF.4.10.10004191138390.292-100000@hercule.perceval.be>

Hi,

I know that I can't do something like the following:

>>> while (line= f.readline()):
	...

as i could in C.  Is there any replacement tricks ?
I know that I could use something with f.read() but i don't want to have
the file completly loaded in memory

Best regards,

--
Life is not fair
But the root password helps
--

Gregoire Welraeds
greg at perceval.be
Perceval Development team
-------------------------------------------------------------------------------
Perceval Technologies sa/nv	Tel: +32-2-6409194		
Rue Tenbosch, 9			Fax: +32-2-6403154		
B-1000 Brussels			general information:   info at perceval.net
BELGIUM				technical information: helpdesk at perceval.net
URL: http://www.perceval.be/
-------------------------------------------------------------------------------




From mwh21 at cam.ac.uk  Wed Apr 19 11:24:47 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 19 Apr 2000 16:24:47 +0100
Subject: Python in the future
References: <38FDB8A7.E701F81F@sec.noaa.gov>
Message-ID: <m3itxei05c.fsf@atrus.jesus.cam.ac.uk>

j vickroy <jvickroy at sec.noaa.gov> writes:

> I'm relatively new to Python and am trying to gain support for it at
> my work place.  To that end, I'm giving a brief (overview)
> presentation.  One of the items I have been asked to consider is
> where will Python be 5-10 years from now.  I certainly don't know
> where C++ or Java will be in that time frame!  Any ideas on where to
> research the future plans for Python would be appreciated.

Well, one of the ways I've been avoiding work in the last few weeks
hass been to read some ancient archives of this list - and one thing
that is obvious is that Python hasn't changed very much in the last
five years.  There's a quality quote from Tim Peters in the Python
quotations file:

   In many ways, it's a dull language, borrowing solid old concepts
   from many other languages & styles: boring syntax, unsurprising
   semantics, few automatic coercions, etc etc.

The point being that Python hasn't changed fast, and probably won't
change that fast in the next five years either.

What I think will/would like to happen in the next couple of years is
that the implementations of Python will get more solid and faster;
things like stackless and Neil Schemenauer's (who's name I can now
almost spell without looking it up...) gc patch may move into the
core.

I'd guess that some kind of optional static typing will be present,
which should enable Python to get significantly faster - and as that
happens, I'd expect more of Python to be written in Python (this trend
is already beginning - see the import- and compiler- sigs).

Indentation based structure is here to stay, I'd think!

But 5-10 years is an almost absurdly long time in computing.  I hope
Python is still here then! (C/C++ can go take a leap, though).

-- 
  well, take it from an old hand: the only reason it would be easier
  to program in C is that you can't easily express complex  problems
  in C, so you don't.                   -- Erik Naggum, comp.lang.lisp


From owen at astroNOJNK.washington.edu.invalid  Tue Apr  4 15:37:07 2000
From: owen at astroNOJNK.washington.edu.invalid (Russell E. Owen)
Date: Tue, 04 Apr 2000 12:37:07 -0700
Subject: Python 1.6 alpha 1 released
References: <Pine.GSO.4.10.10004040035420.21084-100000@sundial> <5E877FD7637AB46E.DABBED9CF3C1D64B.CDDDC05AE7374A78@lp.airnews.net> <5l7leern5v.fsf@eric.cnri.reston.va.us>
Message-ID: <8cdg92$j0m$1@nntp6.u.washington.edu>

>The argument that the address object is spurious doesn't hold water.
>There are many places where it is convenient to hold an address in a
>single variable, and a tuple is the obvious way to represent this.
>(The alternative, a "host:port" string, seems too Perlish to me.)
>Note also that recvfrom() returns the address as a tuple (if it is an
>IP address) and sendto() requires the address to be passed as a tuple
>-- the canonical form of an address is clearly a tuple.

I was following along happily until now, but just got a bit lost.

I completely agree with you that:
- Holding an address in a single variable is a Good Thing
- It pays to support protocols other than IP
- A "host:port:" string would be dreadful

But tuples seem a less than wonderful choice to me. Protocol flexibility 
presumably requires each protocol to have its own tuple with its own 
numer of fields in its own order. That's a lot to remember and try not 
to get mixed up.

The lightest weight alternative I can think of is a dictionary. Each 
protocol would require its own dictionary, but at least they are fairly 
self-documenting once they exist. (And presumably one could get some 
class or other to create a suitable dictionary for a particular 
protocol).

Another alternative is a hierarchy of classes: a network address base 
class with a subclass for each protocol. This could easily get 
heavyweight and clumsy, but I'm not convinced it needs to do so. Imagine 
connection calls that accept appropriate arguments (like file open 
calls) with named arguments (or dictionaries or tuples) and return an 
instance of the network class to be passed around for later calls.

Regards,

-- Russell


From mfletch at tpresence.com  Thu Apr 27 01:31:26 2000
From: mfletch at tpresence.com (Mike Fletcher)
Date: Thu, 27 Apr 2000 01:31:26 -0400
Subject: [PythonWin] VSS fix/workaround for database in non-standard locat
	ion
Message-ID: <CC59868A0A76D311B1F50090277B321C1F32BA@VRTMAIL>

Upgraded to build 128 today, realized that my SourceSafe integration was no
longer functional.  System kept complaining that the database did not exist.
On reading source of vss.py, discovered that it allows for a database =
"inifilename" key in the scc file, but that it ignores the value. I altered
vss.py to use the value, added the database key to the INI file, and was
able to continue working.

Background info: I do not use the default database location for visual
SourceSafe, which has in the past messed up vss.py (I don't actually recall
the fix  in that case (one of the reasons I'm posting)).

In case someone else needs it, find attached.
Enjoy all,
Mike

-------------- next part --------------
A non-text attachment was scrubbed...
Name: vss.py
Type: application/octet-stream
Size: 3092 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20000427/773930ce/attachment.obj>

From dirck at pacbell.net  Wed Apr 19 11:39:59 2000
From: dirck at pacbell.net (Dirck Blaskey)
Date: Wed, 19 Apr 2000 08:39:59 -0700
Subject: module __call__
References: <LX2L4.1275$B43.102174@news.pacbell.net> <7N4L4.497$wYl.191816704@newsb.telia.net> <286L4.1328$B43.118787@news.pacbell.net> <20000419102700.M15664@xs4all.nl>
Message-ID: <jrkL4.684$vE6.202902@news.pacbell.net>

Thomas Wouters <thomas at xs4all.net> wrote in message
news:20000419102700.M15664 at xs4all.nl...
> On Tue, Apr 18, 2000 at 04:23:43PM -0700, Dirck Blaskey wrote:
> > Fredrik Lundh <effbot at telia.com> wrote in message
> > news:7N4L4.497$wYl.191816704 at newsb.telia.net...
> > > Dirck Blaskey <dirck at pacbell.net> wrote:
> > > > ...Gee-Wouldn't-It-Be-Nice-If...
> > > >
> > > > modules could def __call__():
> > > >
> > > > so you could call a module, i.e.
> > > >
> > > >     import StringIO
> > > >     f = StringIO()
> > >
> > > and
> > >
> > > class myStringIO(StringIO):
> > >     ...
> > >
> > > would explode in your face.  lousy idea.
> >
> > ...
> > you have to
> >
> >     class myStringIO(StringIO.StringIO):
> >
> > as currently implemented.
>
> Yes, but 'f = StringIO()' doesn't work either, in that case. Fredrik's
> point, though a bit blunt, is that it would be confusing and thus unwise
to
> functionally separate
>
> f = StringIO()
>
> and
>
> class myStringIO(StringIO)
>
> ...

Thank you for the relevant explanation.
Actually, I was initially suprised to discover that
this feature wasn't already in the language.
If there was a 'module' class it would already work, but
I can't see any other benefit to implementing a 'module' class.

Personally, I don't think it would be confusing, because
I instantiate classes a lot more frequently than I subclass.
However, I can relate to the motivation to keep
magic tricks to a minimum.

(Restated:)
#=======
import mt

def x(s): print s

mt.__call__ = x
mt.__call__("ok")
mt("not ok")
#=======

still-adverb-less

d





From pecora at anvil.nrl.navy.mil  Sun Apr 30 10:08:00 2000
From: pecora at anvil.nrl.navy.mil (Louis M. Pecora)
Date: Sun, 30 Apr 2000 14:08:00 GMT
Subject: Do I always have to write "self." ?
References: <270420001706386648%pecora@anvil.nrl.navy.mil> <slrn8ghc6t.f01.scarblac-spamtrap@flits104-37.flits.rug.nl> <8ecsgj$6lp$1@slb0.atl.mindspring.net> <m3bt2tonyg.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <300420001008012011%pecora@anvil.nrl.navy.mil>

In article <m3bt2tonyg.fsf at atrus.jesus.cam.ac.uk>, Michael Hudson
<mwh21 at cam.ac.uk> wrote:


> > Thinking about readability, I wrote the C++ coding standard here
> > where I work.  When we programmers reviewed it, a co-worker
> > insisted that part of our standard be to prefix the names of all
> > member variables in any class with "m_" so that we could
> > distinguish them from other variables.  I believe that "m_" is a
> > Microsoft-ism and in my humble opinion is just another way to spell
> > "self.".  I think this shows that even C++ programmers are capable
> > of seeing the wisdom of the Pythonic way.
> 
> I have unpleasant memories of trying to understand C++ code that named
> a *local* variable m_something.  Bleargh.  Unambiguous is good.

Isn't this called something like "Hungarian" variable naming?  It has
somethig to do with the Hungarian language IIRC.  There was a whole
movement (maybe it took refuge at Microsoft) which prepended all
variable names with a letter that referred to it's "namespace."  E.g.
use g for global variables:  gChannel, gSocket1, etc.  I thought it was
an aberration. Ugly!  A suffix would be much more readable:  Channelg,
Socket1g etc.  But even then...


From culliton at clark.net  Thu Apr  6 13:37:03 2000
From: culliton at clark.net (Tom Culliton)
Date: Thu, 06 Apr 2000 17:37:03 GMT
Subject: More about variables
References: <meh9-5C17A0.10445406042000@news.cit.cornell.edu>
Message-ID: <3X3H4.16435$0o4.108310@iad-read.news.verio.net>

What (or maybe what do you think) you are trying to do?  Most likely
you would be better off with just a list or tuple if you're dealing
with and unknown number of data items.  Not everything needs a name,
and there isn't much difference between:

v0, v1, v2, v3, v4, v5

and

v[0], v[1], v[2], v[3], v[4], v[5]

Why do you think this variable length collection of data needs
individual names?

Now, if after consideration there is still some compelling reason to
generate names, you can use something like this with a dictionary
(possibly the one returned by globals()):

v = {}
for i in range(number_of_variables):
	name = "variable%d" % i
	v[name] = []

But again, Why?

In article <meh9-5C17A0.10445406042000 at news.cit.cornell.edu>,
Matthew Hirsch  <meh9 at cornell.edu> wrote:
>Hi All,
>
>Let's say I had something like:
>
>number_of_variables=int(raw_input('Enter number of variables: '))
>
>if number_of_variables==1:
>   variable1=[]
>elif number_of_variables==2:
>   variable1=[]
>   variable2=[]
>elif number_of_variables==3:
>   variable1=[]
>   variable2=[]
>   variable3=[]
>elif number_of_variables==4:
>   variable1=[]
>   variable2=[]
>   variable3=[]
>   variable4=[]
>elif number_of_variables==5:
>   variable1=[]
>   variable2=[]
>   variable3=[]
>   variable4=[]
>   variable5=[]
>else:
>   print 'else'
>
>Is there a more efficient way of doing this? What if I wanted to create 
>100 variables (lists).  It doesn't make sense to write an if statement 
>for that many conditions.
>
>Thanks for your help,
>Matt




From raffael at mediaone.net  Thu Apr 27 10:48:29 2000
From: raffael at mediaone.net (Raffael Cavallaro)
Date: Thu, 27 Apr 2000 14:48:29 GMT
Subject: The Simple Economics of Open Source
References: <B527E7A5.51AA%raffael@mediaone.net> <MPG.136cc01af4ab703b989709@news.earthlink.net> <raffael-4899DC.23562823042000@news.ne.mediaone.net> <vndya62ykbi.fsf@camelot-new.ccs.neu.edu>
Message-ID: <raffael-21A8E7.10481427042000@news.ne.mediaone.net>

In article <vndya62ykbi.fsf at camelot-new.ccs.neu.edu>, Justin Sheehy 
<dworkin at ccs.neu.edu> wrote:

>You mean like BSD/OS, from BSDI?  
>
>It happened 9 years ago, and they've been selling it ever since.

Yeah, and BSDI is just tearing up the software world. BSDI sells 
*support* not software, because the software they sell is essentially 
interchangable with the free BSD variants. BSDI have *themselves* 
recognized this fact by merging with Walnut Creek. In future, they'll be 
merging the code of FreeBSD and BSD/OS.

In other words, the history of BSDI is a perfect example why you can't 
succeed at a closed source business when your product is a software 
commodity. BSDI tried, and have, essentially, failed. In a market where 
other companies have billions in revenues for *software alone* BSDI is a 
minor player dependent on a support business.


Ralph

-- 

Raffael Cavallaro, Ph.D.
raffael at mediaone.net


From andy at ictibroker.com  Sat Apr 29 11:46:01 2000
From: andy at ictibroker.com (Andy Smith)
Date: Sat, 29 Apr 2000 11:46:01 -0400
Subject: ODBC / DBI help (via Win32 Extensions)
References: <rNHvjEdhm5Pp-pn2-RytgtQHbLg9d@d185d18f2.rochester.rr.com>
Message-ID: <8ef01r$3c0$1@bob.news.rcn.net>

Hi,

Calling the str method on a dbiDate will return a human-readable date.
Also, calling int(myDbiDate) will return a 'seconds since the epoch'
integer, which you can then easily work with using the functions in the
'time' module. I confess there may be methods of a dbiDate that can achieve
the same, but I'm happy enough to just convert them to integers ;)

Here's an operation on dbiDates that should help you get an access-friendly
format - (but remember to wrap in #'s cos that's crucial too!)

Regards,

Andy.

def dbToReuters(x):
    a = time.localtime(int(x))
    return time.strftime("%m/%d/%Y", time.localtime(int(x)))


Benjamin Schollnick <junkster at nospam.rochester.rr.com> wrote in message
news:rNHvjEdhm5Pp-pn2-RytgtQHbLg9d at d185d18f2.rochester.rr.com...
> ====================================================
>           (Remove "NoSpam" to Email me)
> ====================================================
> Hi All,
>
> I'm doing some MS Access/SQL work via the ODBC & DBI
> modules... and it's working great...
>
> I've got the read, and most of the "write" functions working
> fine, except when it comes to utilizing the date values...
> I can read and convert the date information into a python readable
> value.... But I can't seem to do an UPDATE or INSERT INTO command
> with a date value inside it.
>
> Can someone possibly post a example, or a fragment of a
> working  UPDATE / INSERT INTO command?  Or point me to some
> python resources that might assist?
>
> I've checked the ODBC how-to's on the python web site, and
> they don't even touch on the subject of WRITING a date... The examples
> on reading were very informative, but obviously I'm doing something
> wrong.
>
> - Benjamin
>
> ====================================================
>           (Remove "NoSpam" to Email me)
> ====================================================
> Please feel free to copy any and or all of this sig.
> A little something for spam bots:
>
> root at localhost postmaster at localhost admin at localhost
> abuse at localhost postmaster at 127.0.0.1
>
> Chairman William Kennard: bkennard at fcc.gov
> Commissioner Susan Ness: sness at fcc.gov
> Commissioner Harold Furchtgott-Roth: hfurchtg at fcc.gov
> Commissioner Michael Powell: mpowell at fcc.gov
> Commissioner Gloria Tristani: gtristan at fcc.gov
> consumerline at ftc.gov
> fccinfo at fcc.gov
> ssegal at fcc.gov
>




From mark at chem.uwa.edu.au  Mon Apr 17 16:58:35 2000
From: mark at chem.uwa.edu.au (Mark C Favas)
Date: 17 Apr 00 20:58:35 GMT
Subject: How to fire off a web browser process from Python in Win?
References: <aDwK4.1834$lM4.436461@news.pacbell.net>
Message-ID: <mark.956005115@declan>

"Grant Munsey" <gmunsey at adobe.com> writes:

>Is there a call somewhere in the std library for staring up the default
>( the one that the user has set for opening .htm and .html files)
>web browser on Win 95/8/NT/2K?

No, but I have some code that does this, using Mark Hammond's win32 extensions.
Drop me some email if you want it.

Mark

--
Email  - mark at chem.uwa.edu.au      ,-_|\                           Mark C Favas
Phone  - +61 9 380 3482           /     \               Department of Chemistry
Fax    - +61 9 380 1005      ---> *_,-._/   The University of Western Australia
                                       v                               Nedlands
Loc    - 31.97 S, 115.81 E                               Western Australia 6009


From dalke at acm.org  Fri Apr 14 22:38:36 2000
From: dalke at acm.org (Andrew Dalke)
Date: Fri, 14 Apr 2000 20:38:36 -0600
Subject: FAQ questions
Message-ID: <8d8khb$7e2$1@slb6.atl.mindspring.net>

Just decided to reread the FAQ.  Came across a few things:

section 4.26:
> For more information, read the source of the standard (but
> undocumented) module newdir.

I've never heard if it and it's not present in my install of
PythonWin 1.5.2.  Ahh, it's in Lib/lib-old/newdir.py .

Searching for 'newdir',
http://www.python.org/workshops/1995-05/sftmgmtsess.html
it looks like newdir is now the standard dir?

4.77:
> DocumentTemplate is part of the Bobo objects publishing system
> (http:/www.digicool.com/releases) but can be used independantly
> of course!

I assume that should read Zope now?

6.4:
>  "Why don't strings have methods like index() or sort(), like lists?"

I would like to add:
] String methods will be added in Python 1.6.  Besides the various
] requests, they were also added to support both 8-bit and
] unicode strings.

but I've lost Guido's PSA email about the FAQ password.  :(

6.8:
> Can't you emulate threads in the interpreter instead of relying
> on an OS-specific thread implementation?

Do the microthreads people want to add something to this?

                    Andrew
                    dalke at acm.org






From scarblac-spamtrap at pino.selwerd.nl  Fri Apr 28 07:43:28 2000
From: scarblac-spamtrap at pino.selwerd.nl (Remco Gerlich)
Date: 28 Apr 2000 11:43:28 GMT
Subject: Reading user input
References: <39096d88.3112113532@news.ramk.fi>
Message-ID: <slrn8giu9f.fl9.scarblac-spamtrap@flits104-37.flits.rug.nl>

Teemu Keskinarkaus wrote in comp.lang.python:
> I'm doing some studing about python and MySQL. I've done little test
> program where I use database from python. Now I need get user input to
> do some queries(and other stuff) to database. I've read Python manual
> but haven't found simple solution to do that. I can read something
> using curses but I don't like it. Is there other(and simpler) way to
> do that??

Uhm, something like this is in the Tutorial:

x = raw_input("What is your name? ")
print "Hello, %s!" % x

That's as simple as it gets :)

> Of course I could do html-page and do my test program as CGI and use
> it that way but it's too much trouble when I'm only trying to learn
> databases and how to use them from python.

For things that are more complicated, you could make a little GUI with
Tkinter or so.

And, of course, you could run your tests from the interpreter, and just fill
in the arguments...

-- 
Remco Gerlich,  scarblac at pino.selwerd.nl
Hi! I'm a .sig virus! Join the fun and copy me into yours!


From hanche at math.ntnu.no  Thu Apr  6 22:14:18 2000
From: hanche at math.ntnu.no (Harald Hanche-Olsen)
Date: 06 Apr 2000 22:14:18 -0400
Subject: Why should I switch to Python? - Infinity of Primes
References: <000601bf9f8e$bd31e780$ea2d153f@tim>
Message-ID: <pcoln2q3bet.fsf@math.ntnu.no>

+ "Tim Peters" <tim_one at email.msn.com>:

| Hmm.  Try out this one-liner Haskell defn, which is so pretty it *should*
| bring tears of joy to everyone's eyes <blink>:

Beautiful <sniff>, and almost certainly even easier to prove correct
than mine.  But it is also inefficient compared to my more clunky
implementation (statistics from hugs):

Primes> take 100 primes
[2,3,5,7,...,523,541]
(27698 reductions, 42791 cells)

which, by the way, is brought down by a factor of ten to
(2720 reductions, 4094 cells)
by employing a few more fairly obvious optimizations.

Primes> take 100 (sieve [2..]) where sieve (p:rest) = p : sieve [n | n <- rest, mod n p /= 0]
[2,3,5,7,...,523,541]
(394280 reductions, 573890 cells, 2 garbage collections)

beauty-hath-its-price-but-it's-worth-it-ly y'rs,
-- 
* Harald Hanche-Olsen     <URL:http://www.math.ntnu.no/~hanche/>
- "There arises from a bad and unapt formation of words
   a wonderful obstruction to the mind."  - Francis Bacon


From ngps at madcap.dyndns.org  Mon Apr 17 13:25:35 2000
From: ngps at madcap.dyndns.org (Ng Pheng Siong)
Date: 17 Apr 2000 17:25:35 GMT
Subject: [ANNOUNCE] M2Crypto 0.04
Message-ID: <8dfhef$99b$1@mawar.singnet.com.sg>

Hello,

I am pleased to announce the release of M2Crypto 0.04. This release
offers the following:

-   Based on OpenSSL 0.9.5. Tested on FreeBSD, Linux and Win NT.

-   Much improved SSL:

        -   SSL certificate-based client authentication with 
        callbacks in Python.

        -   An interface to create SSL sessions. An SSL session 
        caches security parameters negotiated during SSL/TLS handshaking. 
        Many SSL connections can be instantiated over an SSL session; 
        this reduces the frequency of expensive public-key cryptographic 
        operations and thereby improves performance.

        -   More robust SSL socket implementation - raises Python 
        exceptions, not dumps core. (Mostly. ;-)

-   S/MIME: signs, encrypts, decrypts, verifies. 

-   More X.509 certificate manipulation, in support of PyCA.

-   Various bug fixes and minor enhancements.


Thanks to Ray Soursa <res at loudcloud.com>, Johannes Stezenbach 
<yawyi at gmx.net>, Michael Str?der <michael.stroeder at inka.de> and 
others for bug reports and suggestions.


The release is here: 

    http://www.post1.com/home/ngps/m2/m2crypto-0.04.zip
    http://www.post1.com/home/ngps/m2/m2crypto-0.04.zip.asc


As usual, feedback is very much appreciated. Cheers.
-- 
Ng Pheng Siong <ngps at post1.com> * http://www.post1.com/home/ngps
-- 
Ng Pheng Siong <ngps at post1.com> * http://www.post1.com/home/ngps



From dworkin at ccs.neu.edu  Tue Apr 18 21:15:05 2000
From: dworkin at ccs.neu.edu (Justin Sheehy)
Date: 18 Apr 2000 21:15:05 -0400
Subject: Python paradigms
In-Reply-To: "Robert W. Cunningham"'s message of "Sun, 09 Apr 2000 04:38:22 GMT"
References: <8cmr70$b0e$1@pegasus.csx.cam.ac.uk> <B515729E.1745%dgoodger@bigfoot.com> <38F00890.3C98EBDB@acm.org>
Message-ID: <vndya6ac2na.fsf@camelot-new.ccs.neu.edu>

"Robert W. Cunningham" <rcunning at acm.org> writes:

> # blah = (test ? true : false)

> Are there ANY Python programmers who don't also know at least this much C?

Yes, many.  And the number of them keeps growing...

-Justin

 



From flu at iki.fi  Mon Apr 24 16:14:54 2000
From: flu at iki.fi (Ulf Betlehem)
Date: 24 Apr 2000 23:14:54 +0300
Subject: invisible cursor in ncurses5
Message-ID: <87wvlnjlxd.fsf@lemon.tky.hut.fi>

        Using curses modules included with most python distributions,
        there seems to be no way to turn off the cursor if the module
        is linked against ncurses5. (earlier versions are ok)

        Even though the included manpage for leaveok() states that "If
        possible, the cursor is made invisible when this option is
        enabled.", ncurses5 makes no such effort. Instead, the user is
        expected to call curs_set() explicitly, which is NOT supported
        by the python curses module.

        I have no idea why this behaviour was changed. The NEWS file
        for ncurses5 only says: "remove spurious curs_set() call from
        leaveok() (J.T.Conklin)."

        Is there a good solution to this problem?

-- 
 flu


From darcy at vex.net  Thu Apr 27 06:08:37 2000
From: darcy at vex.net (D'Arcy J.M. Cain)
Date: 27 Apr 2000 10:08:37 GMT
Subject: Best database to use with Python
References: <3906ea27.698463777@news.online.no> <39070DC7.37F8089D@NOSPAM.linuxsolutions.demon.co.uk>
Message-ID: <8e93j5$qba$2@news.tht.net>

Kevin McDermott <kevinmcd at nospam.linuxsolutions.demon.co.uk> wrote:
> Depends on what you want to use it for...if it's to connect CGI programs
> to, then I'd recommend MySQL.

Please don't take this as argumentative but I am curious why you think
that being used with CGI programs per se makes one database better
than the other.  It seems to me that there are a lot more issues to
consider such as what kind of data, what kind of data relationships,
how the data needs to be accessed, etc.  I use PostgreSQL for all my
database work, CGI accessed or not, except for where the data sets
are simple enough to use shelve.

-- 
D'Arcy J.M. Cain <darcy at vex.net>           |  Democracy is three wolves
http://www.vex.net/                        |  and a sheep voting on         
+1 416 425 1212     (DoD#0082)    (eNTP)   |  what's for dinner.


From akuchlin at mems-exchange.org  Fri Apr 21 10:13:17 2000
From: akuchlin at mems-exchange.org (Andrew M. Kuchling)
Date: 21 Apr 2000 10:13:17 -0400
Subject: Starship update, April 21
Message-ID: <3d1z3zzgn6.fsf@amarok.cnri.reston.va.us>

The data recovery firm has received the disk, and after a preliminary
examination, things aren't looking very promising.  It was explained
to me that the actuator coil expanded and began to rub up against the
magnet, which in turn made the disk heads rub against the platter.
The fix is to buy a new actuator coil, replace the heads (which is the
most delicate and costly operation they do) and try reading data off
the disk.  They estimate there is only a 25% chance of recovering any
data at all.

Further news will be posted as events warrant.

-- 
A.M. Kuchling			http://starship.python.net/crew/amk/
Yes: the code in ntpath.split is too clever to have any hope of working
correctly <wink>.
  -- Tim Peters, 19 Mar 99



From gregholmes at my-deja.com  Mon Apr 24 09:17:00 2000
From: gregholmes at my-deja.com (gregholmes at my-deja.com)
Date: Mon, 24 Apr 2000 13:17:00 GMT
Subject: Tkinter+win32 Installer question
Message-ID: <8e1hg2$dq2$1@nnrp1.deja.com>

I am developing multiple programs using Tkinter and win32 extensions.

Is there a reliable, brute force way to use the Installer to make a
distribution that will work? I realize I am not posting specifics here,
but that is kind of the point; is there a general solution? The default
behavior seems to exclude necessary things.

I don't mind reinventing the wheel if needed ;) but has this wheel
already been invented?


Sent via Deja.com http://www.deja.com/
Before you buy.


From nobody at nowhere.nohow  Sun Apr 23 22:57:24 2000
From: nobody at nowhere.nohow (Grant Edwards)
Date: Mon, 24 Apr 2000 02:57:24 GMT
Subject: setting up Tkinter
References: <KlHM4.8299$4w1.119156@typhoon.nyroc.rr.com>
Message-ID: <oKOM4.1157$wJ1.23728@ptah.visi.com>

On Sun, 23 Apr 2000 18:33:14 GMT, Andy Mullen <pickaxe22 at stny.rr.com> wrote:

>As i understand from the website (www.python.com), this program
>doesn't run on windows 98. Since i was already thinking of
>putting linux on my machine under a partition of the HD, i was
>wondering if someone could point me in the right direction as
>to where to get linux online, how to go about install, and how
>to run the 2 separate OS's on one machine.  any response would
>be appreciated

By far the easiest way to install Linux is from a CD.  You can
get them for a couple dollars from places like
www.cheapbytes.com.  

Or you can download and burn one yourself. Check out
www.linux.org, www.linux.com, www.linuxdoc.org

-- 
Grant Edwards                   grante             Yow!  .. If I cover this
                                  at               entire WALL with MAZOLA,
                               visi.com            wdo I have to give my AGENT
                                                   ten per cent??


From dirck at pacbell.net  Tue Apr 18 19:23:43 2000
From: dirck at pacbell.net (Dirck Blaskey)
Date: Tue, 18 Apr 2000 16:23:43 -0700
Subject: module __call__
References: <LX2L4.1275$B43.102174@news.pacbell.net> <7N4L4.497$wYl.191816704@newsb.telia.net>
Message-ID: <286L4.1328$B43.118787@news.pacbell.net>

Fredrik Lundh <effbot at telia.com> wrote in message
news:7N4L4.497$wYl.191816704 at newsb.telia.net...
> Dirck Blaskey <dirck at pacbell.net> wrote:
> > ...Gee-Wouldn't-It-Be-Nice-If...
> >
> > modules could def __call__():
> >
> > so you could call a module, i.e.
> >
> >     import StringIO
> >     f = StringIO()
>
> and
>
> class myStringIO(StringIO):
>     ...
>
> would explode in your face.  lousy idea.

Hmm...

    class myStringIO(StringIO):

already doesn't work;
you have to

    class myStringIO(StringIO.StringIO):

as currently implemented.

d





From olsen at umbc.edu  Tue Apr 11 20:02:41 2000
From: olsen at umbc.edu (Peter Olsen)
Date: 11 Apr 2000 20:02:41 -0400
Subject: Actually Running a function from a DLL
References: <38EE87A3.4080802@draper.com>
Message-ID: <99ir9cc9of2.fsf@sunserver1.cs.umbc.edu>

I'm writing to ask a question about calling a dll file with the
windll/calldll package.

First, thanks to Mark Hammond, Neil Hodgson, and Peter Desaulniers for
their responses and help with my previous question.

I have a DLL file, ETIEncr.dll, with documentation purporting to show
it contains a function GetVersion() (of no arguments) which returns
the version number as a string, e.g. "1.1". I want to call this
function and several others from the dll using windll/calldll.

After putting windll.py and calldll.pyd in the directory
..\Python\Lib\Plat-Win (a directory which is already in sys.path), I
have tried to execute GetVersion() in the PythonWin environment
following the commented directions in windll.py.

This is what happens

>>> import windll
>>> from windll import *
>>> etie = module("ETIEncr")
>>> etie
<win32 module 'ETIEncr' (0 functions)>
>>> etie.GetVersion()
Traceback (innermost last)
  File "<interactive input>", line 0 in ?
  File "dynwin\windll.py", line 91 in __getattr__
    raise AttributeError 'GetProcAddress failed for %s' % name
AttributeError: GetProcAddress failed for GetVersion
>>>

I think I've matched the calls exactly with those in the kernel32.Beep
example in windll.py.  There is one possible difference between that
example and my call here.  The PythonWin COMbrowser shows an
"intermediate layer" between the ETIEncr library and the GetVersion()
function.  The hierarchy is

ETIEncy
 |
 +--Type Library
       |
       +--IEtiEncrypt - Dispatch
             |
             +--GetVersion - Function

The COMBrowser shows these details for GetVersion:

Dispatch ID = 3
Return Type = 'String'
Function Kind = 'Dispatch'
Invoke Kind = 'Function'
Number Optional Parameters = 0


If possible, I would appreciate it if you would send any responses to
pcolsen at draper.com.  Company policy makes it hard to read news.

Thanks!

Peter
-- 
   Peter Olsen, PE,            n2ell          engineer at draper.com
     P.O. Box 410, Simpsonville, MD 21150-0410 USA, 410-997-8584
   "Engineering is the art of applying a professional knowledge of
mathematics and the physical sciences to improve the quality of life."


From cestluk at polyu.edu.hk  Sat Apr 29 06:12:35 2000
From: cestluk at polyu.edu.hk (cestluk)
Date: Sat, 29 Apr 2000 18:12:35 +0800
Subject: Environmental variable $PYTHONPATH
Message-ID: <77363B9E386.AAA2589@smtp1.polyu.edu.hk>

Please excuse me if this rather dumb question has been answered before. It 
does not seem to be in the FAQ, though.

It seems that each time you add a new library/module to your current python 
installation, you have to make it known to python manually, for example, by 
editing $PYTHONPATH or including it in the search path by invoking sys.path(). 
I am using the W32 version of python 1.6 and the sys.path output looks 
something like

'', 'C\\Python16\\win32', 'C\\Python16\\win32\\lib', 'C\\Python16', 
'C\\Python16\\Pythonwin', 'C\\PYTHON16\\DLLs', 'C\\PYTHON16\\lib', 
'C\\PYTHON16\\lib\\plat-win', 'C\\PYTHON16\\lib\\lib-tk', 
'C\\PYTHON16\\PYTHONWIN', 'C\\PYTHON16']

It seems you have to specify the name of _every_ directory and subdirectory. 
Doing this manually would be quite tedious if you have a lot of modules to 
install. Is there a simpler/less manual way to do it?

I am thinking of the case of TeX (in tetex) where the system installation is 
under a certain texmf tree and the user's own packages are under another and 
running kpathsea will set up the search path for the packages.

Regards,
ST
--




From julschae at usa.net  Sat Apr  8 05:15:32 2000
From: julschae at usa.net (Peter Kovalsky)
Date: Sat, 8 Apr 2000 11:15:32 +0200
Subject: Refeing to substring in list of strings
Message-ID: <MPG.1359171bf0a65199989683@nntp.server.uni-frankfurt.de>

Hi,

I am new to Python and experienced a little problem with referencing 
lately. I have a list of strings that I have to modify. Thereofre I chose 
a for-construct (for index in range(len(mylist)):). I found certain 
substrings that i want to extract - but don' t know how to adress these 
substrings. I tried a
	mylist[index][lborder:rborder]
which won' r work for some reason - and I don' t know why. Can anybody 
help???

Thanx in advance,

Peter Kovalsky


From embed at geocities.com  Fri Apr  7 09:18:49 2000
From: embed at geocities.com (Warren Postma)
Date: Fri, 7 Apr 2000 09:18:49 -0400
Subject: Embedding Python: How To Hardwire Module Search Path
References: <38EDDB40.FC715534@gmx.net>
Message-ID: <4flH4.4358$HG1.126878@nnrp1.uunet.ca>

> Does anybody know what to do prior to Py_Initialize(). Or am I
> completely wrong?
> PySys_SetPath() crashes when called before Py_Initialize().

First just set the environment variable PYTHONPATH, but if you want to have
it ignore the registry on Windows, I think you may have to resort to
modifying source code.

I did this. Look for registry code in import_nt.c. Email me if you want to
see what I did, but since they are trivial you can probably do it yourself.

My modified version ignores the registry, uses the PYTHONPATH, as well as
certain 'hardcoded' sensibiel defaults such as looking in the Lib
subdirectory from wherever your program is running from for the 'site.py'
and 'exceptions.py' modules. From there, set up your additional module
search paths in the PYTHONPATH environment variable.

Warren




From aek at mail1.stofanet.dk  Sun Apr 30 09:53:43 2000
From: aek at mail1.stofanet.dk (Anders Eggers-Krag)
Date: Sun, 30 Apr 2000 15:53:43 +0200
Subject: ASP - Collection value assignment
Message-ID: <8ehds8$2ms$1@news101.telia.com>

I have the following code

def Application_OnStart():
 Application("THcolor") = "#DDDDDD"    <---
 Application("color1") = "infobackground"
 Application("color2") = "#DDDDDD"

I get an unknown error, in the line pointet to with the error

(
actually with this code I get a syntax error for some wierd reason,
but that can be solved placing the first line on the same line as the
function
as in
def Application_OnStart(): Application("THcolor") = "#DDDDDD"
 Application("color1") = "infobackground"
 Application("color2") = "#DDDDDD"
)

how do I assign values to items in collections?


--
Anders Eggers - Krag




From rupole at compaq.net  Fri Apr 14 23:20:11 2000
From: rupole at compaq.net (Roger Upole)
Date: Sat, 15 Apr 2000 03:20:11 GMT
Subject: Getting file version information
References: <20000413163629.28006.qmail@web4303.mail.yahoo.com>
Message-ID: <LdRJ4.120194$ra.2817147@news.easynews.com>

I don't know if this COM object is available on your system, but on win98 I
can:
   1) run makepy on "Microsoft Scripting Runtime"
   2) x=win32com.client.Dispatch('Scripting.FileSystemObject')
   3) x.GetFileVersion('somefilename')

                 Roger Upole



"Daniel Zapata" <danzap_python at yahoo.com.br> wrote in message
news:20000413163629.28006.qmail at web4303.mail.yahoo.com...
> I?d like to know how to get the file version
> information in Windows. I?m using pythonwin and
> couldn?t find the GetFileVersionInfo API on the
> win32api module.
>
> Thanks
>
> Daniel Zapata
>
>
____________________________________________________________________________
> YAHOO! BATE-PAPO
> Conhe?a gente nova, fa?a amigos e divirta-se! - http://www.yahoo.com.br
>




From fredrik at pythonware.com  Thu Apr 27 06:05:04 2000
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 27 Apr 2000 12:05:04 +0200
Subject: IP Address
References: <WZTN4.135$E44.2126@news6-win.server.ntlworld.com>
Message-ID: <014801bfb030$141df690$0500a8c0@secret.pythonware.com>

Richard Chamberlain wrote:
> Is there a way of returning the IP address of the machine that a script is
> running on?

here's one way to do it:
ip = socket.gethostbyname(socket.gethostname())

</F>




From darrell at dorb.com  Sat Apr 29 11:01:25 2000
From: darrell at dorb.com (Darrell)
Date: Sat, 29 Apr 2000 15:01:25 GMT
Subject: other patterns in python?
References: <8ebpl4$eq9$1@nnrp1.deja.com> <nXpO4.17469$GP3.124141@typhoon.nyroc.rr.com> <390ABB26.3831DDD1@uab.edu>
Message-ID: <9PCO4.18553$GP3.127275@typhoon.nyroc.rr.com>

"Shae Erisson" <shapr at uab.edu> wrote:
>
> Has anyone else written pattern demos in python?
> I'd like to see more if they're out there.
>
There was a site that talked about the Gang of four patterns in Python.
Although I can't find it now.
Try this search +"design pattern" on python.org and deja.com
Also check this out. http://www.python.org/doc/essays/

--Darrell Gallion







From bjorn at roguewave.com  Tue Apr  4 19:30:36 2000
From: bjorn at roguewave.com (Bjorn Pettersen)
Date: Tue, 04 Apr 2000 17:30:36 -0600
Subject: Two questions
References: <38EA5FE8.853E9B75@callware.com> <38EA6508.90BF630C@roguewave.com> <38EA6E87.40EDB542@callware.com>
Message-ID: <38EA7B1C.95420D07@roguewave.com>

Ivan Van Laningham wrote:
[snip]
> <g>  This sounds like the right course to me.  I'd like to force
> customers to install Python on their machines, but my boss doesn't think
> that's all that great an idea (dunno why).  My only concern is
> size--just how big are the executables produced by Gordon's
> flummery^H^H^H^H^H^H^H^Hmagic?

A simple "hello world" type app is about 285K (not too unreasonable for
a windows app ;-)

There are various options that will install various subsets of Python. 
Half an hour of playing around should give you a good indication of
which route you want to take...

my-boss-won't-let-me-install-Python-on-everyone's-desktop-either'ly y'rs
--bjorn



From btang at pacific.jpl.nasa.gov  Tue Apr 18 11:33:59 2000
From: btang at pacific.jpl.nasa.gov (Benyang Tang)
Date: Tue, 18 Apr 2000 08:33:59 -0700
Subject: Running shell programs from Python
References: <38FC64DA.7D297563@ffi.no>
Message-ID: <38FC8067.59000C47@pacific.jpl.nasa.gov>

seh at ffi.no wrote:
> 
> I have just started with Python and have some questions.. I have some
> compiled c programs that I normally run from the shell. I can use | to
> pipe the output from one as input to the next. I would like to call the
> programs from Python. Could someone please indicate how this is done?
> Any references?
> 
> Regards
> Svein-Erik Hamran

Use os.popen() or popen2.popen2(). 

Reference:
http://www.python.org/doc/current/lib/os-newstreams.html
http://www.python.org/doc/current/lib/module-popen2.html

Here is an example (for Unix):

import os,popen2
str1 = os.popen('ls','r').read()
print str1
out1,in1 = popen2.popen2('cat')
in1.write(str1)
in1.close()
str2 = out1.read()
out1.close()
print str2

===============
Benyang Tang


From jhauser at ifm.uni-kiel.de  Wed Apr 12 01:21:59 2000
From: jhauser at ifm.uni-kiel.de (Janko Hauser)
Date: 12 Apr 2000 07:21:59 +0200
Subject: Python/Perl Popularity (Re: A Mountain of Perl...)
References: <200004102111.HAA03380@envy.fulcrum.com.au> <8d0pcq$7ti$1@nnrp1.deja.com>
Message-ID: <87og7fopvs.fsf@ifm.uni-kiel.de>

Use docindex.py!

This module is definetly no substitute for perldoc but a start that
answers your question. It gets the relevant information from the html
tree which needs to be installed on your system.

A benefit over docstrings is, that one does not need to import a
module. This is a prototype, but shows that it is possible. Not that
this will change your mind about Python :-)

HTH,
__Janko

#!/usr/bin/env python
#
# docindex -- get HTML documentation for an index of keywords.
#

__version__ = '0.1.0'
__author__ = 'Janko Hauser <jhauser at starship.python.net>'

import re, htmllib, formatter
import os, string, cPickle
import glob

# A simple help system which builds on top of docreader.py from
# Jim Eggleston  <jimegg at gol.com>. These classes are included here to
# have only one file. The interface is changed a little bit and the
# indices are generated once and can then be stored in a pickle.
#
#
## START docreader.py -- read Python HTML docs from the command line.

## docreader builds a dictionary of terms from the index of the Python Library
## HTML documentation. The user is prompted to enter a term. If there is only
## one entry for the term, the section of the relevant document node is displayed.
## If there is more than one entry, a numbered list of index entries for that term
## is displayed. Individual entries can be read by entering the list item number.

## docreader will exit if a null string is entered as a search term.

## Setup: 
## The global variables docdir, contents_doc and pager must be set according to
## your particular configuration.

## Changes:
## 14-Mar-1998 Created
## 15-Mar-1998 Added MyWriter class taken from Tony Ibbs 'html2text.py'

## Jim Eggleston  <jimegg at gol.com>
#
# Change these according to your configuration
# (This is now done in the shell class 
#

# directory where the 'lib' HTML doc files are kept
# docdir = '/home/hauser/public_html/doc/python/lib'

# the index document
# index_doc = 'genindex.html' # 'node268.html'

# An external pager program (e.g. 'more', 'less') which accepts stdin
# (Set to None if no pager available)
pager = 'less'

#
# Sub-class HTMLParser to add minimal support for tables, and also for
# (future) support of links.
#
class MyHTMLParser(htmllib.HTMLParser):

    def __init__(self, formatter, verbose=0):
        htmllib.HTMLParser.__init__(self, formatter, verbose)

    def anchor_bgn(self, href, name, type):
        self.anchor = href
        self.anchorname = name
        if self.anchor:
            self.anchorlist.append(href, name)

    def anchor_end(self):
        if self.anchor:
            self.handle_data("[%d]" % len(self.anchorlist))
        
        # (Look at handling anchor names later)
        # elif self.anchorname:
        #     self.handle_data("[#%s]" % self.anchorname)
    
    # Send a line break before each table row       
    def do_tr(self, attr):
        self.formatter.add_line_break()


#
# Based on text2html.py, by Tony Ibbs (Tibbs).
#
DEBUG = 0

class MyWriter(formatter.NullWriter):

    def __init__(self, file=None, maxcol=72):
        self.file = file or sys.stdout
        self.maxcol = maxcol
        self.margin = 0         # margin
        self.extra_margin = 0   # extra margin in a list element
        self.indent = 3         # increment for margin
        self.extra_indent = 2   # indent for extra margin
        self.level = {}         # dictionary of levels
        formatter.NullWriter.__init__(self)
        self.reset()

    def reset(self):
        self.col = 0            # current column
        self.atbreak = 0        # is this a good place to linebreak?

    def new_alignment(self, align):
        if DEBUG: print "new_alignment(%s)" % `align`

    def new_font(self, font):
        if DEBUG: print "new_font(%s)" % `font`

    def new_margin(self, margin, level):
        if DEBUG: print "new_margin(%s, %d)" % (`margin`, level)

        if margin == None:
            self.margin = 0
        elif self.level.has_key(margin):
            if level == 1:
                self.level[margin] = level
                self.margin = self.margin + self.indent
            elif level == 2:
                self.level[margin] = level
                self.margin = self.margin - self.indent
                if self.margin < 0: self.margin = 0
        else:
            self.level[margin] = level
            self.margin = self.margin + self.indent
        if DEBUG: print "   margin =",self.margin
        self.extra_margin = 0
        return

        if margin == None:
            self.margin = 0
        elif level == 1:
            self.margin = self.margin + self.indent
        elif level == 2:
            self.margin = self.margin - self.indent
            if self.margin < 0: self.margin = 0
        else:
            print "***Margin %s level %d (not 1 or 2)"%(margin,level)
            self.margin = 0
        if DEBUG: print "   margin =",self.margin
        self.extra_margin = 0
        #self.reset()  #???
        #self.start_margin()

    def new_spacing(self, spacing):
        if DEBUG: print "new_spacing(%s)" % `spacing`

    def new_styles(self, styles):
        if DEBUG: print "new_styles(%s)" % `styles`

    def send_paragraph(self, blankline):
        if DEBUG: print "send_paragraph(%s)" % `blankline`
        if blankline > 1:
            self.file.write('\n'*(blankline-1))
        self.reset()
        self.file.write("\n")
        ##self.start_line()

    def send_line_break(self):
        if DEBUG: print "send_line_break()"
        #self.file.write('\n')
        self.reset()
        self.file.write("\n")
        ##self.start_line()

    def send_hor_rule(self, *args, **kw):
        if DEBUG: print "send_hor_rule(%s,%s)"%(args,kw)
        self.file.write('\n')
        self.file.write('-'*self.maxcol)
        self.file.write('\n')
        self.reset()

    def start_margin(self):
        if DEBUG: print "   start_margin()"
        self.file.write(" "*self.margin + " "*self.extra_margin)
        self.col = self.col + self.margin

    def start_line(self):
        if DEBUG: print "   start_line()"
        self.file.write("\n")
        self.start_margin()
        #self.file.write("\n" + " "*self.margin + " "*self.extra_margin)
        #self.col = self.col + self.margin

    def send_label_data(self, data):
        if DEBUG: print "send_label_data(%s)" % `data`
        self.extra_margin = 0
        self.start_line()
        self.file.write(data+" ")
        self.extra_margin = self.extra_indent

    def send_flowing_data(self, data):
        if DEBUG: print "send_flowing_data(%s)" % `data`
        if not data: return
        atbreak = self.atbreak or data[0] in string.whitespace
        col = self.col

        if col == 0:
            self.start_margin()

        maxcol = self.maxcol
        write = self.file.write
        #self.start_line()
        for word in string.split(data):
            if atbreak:
                if col + len(word) >= maxcol:
                    #write('\n')
                    col = 0
                    self.start_line()
                else:
                    write(' ')
                    col = col + 1
            write(word)
            if DEBUG: print "   ",word
            col = col + len(word)
            atbreak = 1
        self.col = col
        self.atbreak = data[-1] in string.whitespace

    def send_literal_data(self, data):
        if DEBUG: print "send_literal_data(%s)" % `data`
        self.file.write(data)
        i = string.rfind(data, '\n')
        if i >= 0:
            self.col = 0
            data = data[i+1:]
        data = string.expandtabs(data)
        self.col = self.col + len(data)
        self.atbreak = 0


######
# Helper functions for building the index datastructure
######
# Scan through all nodes and build a dictionary mapping node file names
# to document titles.
#
def build_std_nodes(docdir):
    lower = string.lower
    node_path = os.path.join(docdir, "*.html")
    title_re = r'(<TITLE>|<title>)([0-9\.]*) (.*)((--)|(</TITLE>|</title>))'  
    node_titles = {}
    for path in glob.glob(node_path):
        node = os.path.split(path)[1]
        f = open(path)
        while 1:
            line = f.readline()
            if not line: break
            m = re.match(title_re, line)
            if m:
                #print m.groups()
                node_titles[node] = string.strip(m.group(2))
                break
    return node_titles

#
# Create a dictionary from the index document that maps index terms
# to node documents and anchors.
#
def build_std_index(docdir, docindex):
    index_path = os.path.join(docdir, docindex)
    #index_re = r'<dt>.*HREF="([^#]*)#(\d*)"><tt>(.*)</tt>'
    index_re = r'<dt><a href=["|\']([^#]*)#(l2h-(\d*))["|\']>(.*)</a>'
    extra_re = r'<A NAME=.*HREF="([^#]*)#(\d*)">'  
    index = {}
    nterm = ''
    lines = open(index_path).readlines()
    for line in lines:
        m = re.match(index_re, line)
        if m:
            term = m.group(4)
            nterm = re.match(r'([^ ()]*)', term).group(1)
            if index.has_key(nterm):
                index[nterm].append((m.group(1),m.group(2)))
            else:
                index[nterm] = [(m.group(1),m.group(2))]
        m = re.match(extra_re, line)
        if nterm and m and index.has_key(nterm):
            index[nterm].append((m.group(1),m.group(2)))
    return index

#
# Extract a section of a node referenced by an anchor.
#
class docreader:
    """Get documentation from an index of keywords"""
    def __init__(self,stdroot, docdb=''):
        self.std_docdir = stdroot
        self.std_index_file = 'genindex.html'
        self.pager = 'less'
        self.use_pager = 1
        self.page_size = 1000
        self.docdb = docdb
        self.ask_fullview = 1
        self.show_url = 1
        self.load_db()
        
    def show(self,data):
        """ Format HTML and write it to the screen.
        A pager is use if documentation text is to long"""
        if self.pager and self.use_pager:
            out = os.popen(pager, 'w')
        else:
            import sys
            out = sys.stdout   
        fmtr = formatter.AbstractFormatter(MyWriter(out))
        p = MyHTMLParser(fmtr)
        try:
            p.feed(data)
            p.close()
        except IOError:
            pass

    def load_db(self):
        """Get the indices"""
        # Look first if there is already a pickle of the indices, currently no
        # global place but in the home of the user. Needs to be fixed!
        if not self.docdb:
            self.docdb = os.path.join(os.environ['HOME'],'.ipphelp')
        has_docdb = os.path.isfile(self.docdb)
        if has_docdb:
            self.index_cont = cPickle.load(open(self.docdb,'r'))[0]
            self.hasdb = 1
        elif self.std_docdir:
            # Build an index for every directory
            self.index_cont={}
            for help_dir in self.std_docdir:
                cont={}

                nodes = build_std_nodes( help_dir)
                index = build_std_index( help_dir, self.std_index_file)

                cont['directory'] = help_dir
                cont['nodes']=nodes
                cont['indices']=index
                self.index_cont[help_dir]=cont
            self.help_hasdb = 1

    def save(self):
        """Saves the indices for later use"""
        cPickle.dump((self.index_cont,), open(self.docdb, 'w'))
        return

    def __call__(self, name):
        self.get(name)
        
    def get(self, name):
        """Get info for the given name"""
        entries = []
        for dsource in self.index_cont.values():
            if dsource['indices'].has_key(name):
                for entry in dsource['indices'][name]:
                    found = [dsource['directory'], entry[0], entry[1]]
                    entries.append(found)
        if len(entries) > 1:
            sdir, node, anchor = self.display_node_select(name, entries)
        elif entries:
            sdir, node, anchor = entries[0]
        else:
            print 'No help available'
            
        if anchor:
            self.display_node_section(sdir, node, anchor)
            if self.ask_fullview:
                resp = raw_input('View entire node [yN]? ')
                if resp and resp[0] in 'yY':
                    self.display_node(sdir, node)
        else:
            self.display_node(sdir, node)

    def display_node_section(self, sdir, node, anchor):
        node_path = os.path.join(sdir, node)
        anchor_re = re.compile(r'<a name=["|\']%s["|\']>(.*?)(?:<a name|<HR>|</BODY>)' % \
                               anchor, re.DOTALL)
        data = open(node_path).read()
        m = anchor_re.search(data)
        if m:
            if self.show_url:
                data = 'file:%s%s#%s<P>%s' % (sdir, node, anchor, m.group(1))
            else:
                data = 'Lookup in %s<P>%s' % (node, m.group(1))
        else:
            data = 'No help available'
        if len(data) > self.page_size: # page_size in character
            self.use_pager=1
        else:
            self.use_pager=0

        self.show(data)

    def display_node(self, sdir, node):
        """Display an entire node"""
        self.use_pager = 1 # always use the pager for a full node
        node_path = os.path.join(sdir, node)
        data = open(node_path).read()
        self.show(data)

    def display_node_select(self, name, entries):
        print "\nEntries for '%s':" % name
        i = 0
        for entry in entries:
            sdir, node, anchor=entry
            i = i + 1
            fnode = node
            if fnode[:7] == 'module-':
                fnode = fnode[7:-5]
            print "    %s: %s" % (i, fnode)
        print
        item = raw_input('Lookup in module? ')
        if item:
            i = string.atoi(item) - 1
            sdir, node, anchor = entries[i]
        return sdir, node, anchor


if __name__ == '__main__':

    # Put the path to the HTML library documentation on your system here.
    help = docreader(('/home/user/Doc/html/NumPy/',
                   '/home/user/PYHTML/lib/'))

    help('round')


-- 
  Institut fuer Meereskunde             phone: 49-431-597 3989
  Dept. Theoretical Oceanography        fax  : 49-431-565876
  Duesternbrooker Weg 20                email: jhauser at ifm.uni-kiel.de
  24105 Kiel, Germany

-- 
  Institut fuer Meereskunde             phone: 49-431-597 3989
  Dept. Theoretical Oceanography        fax  : 49-431-565876
  Duesternbrooker Weg 20                email: jhauser at ifm.uni-kiel.de
  24105 Kiel, Germany


From poebblo at tjohoo.se  Wed Apr 26 10:00:03 2000
From: poebblo at tjohoo.se (Per Wigren)
Date: Wed, 26 Apr 2000 15:00:03 +0100
Subject: Sv: huh? Python says that AttributeError: 'string' object has no attribute 'split'
Message-ID: <20000426125950.2A6181CD60@dinsdale.python.org>

>  for l in lines:
>   l = l.strip( l )

l = string.strip(l)

____________________________________
Gratis epost med TJOHOOMAIL
http://www.tjohoo.se





From exarkun at flashmail.com  Sun Apr 23 13:04:38 2000
From: exarkun at flashmail.com (Jp Calderone)
Date: Sun, 23 Apr 2000 17:04:38 +0000
Subject: SSL w/ urllib
References: <3901D800.DB50D968@flashmail.com> <3ditx9v851.fsf@newcnri.cnri.reston.va.us> <390249B5.F5D545F7@flashmail.com> <3dhfcsnbz5.fsf@newcnri.cnri.reston.va.us>
Message-ID: <39032D26.1F2FAA5C@flashmail.com>

Andrew Kuchling wrote:
> 
> Oops!  I wrote the patch too quickly and forgot one more piece; it
> never sends the body of the POST.  <bullwinkle>This time for
> sure!</bullwinkle>
> 
> --amk
> 
> [snip patch]

 Excellent, it works now :)  Thanks again

 Jp

-- 
-----BEGIN GEEK CODE BLOCK-----
Version: 3.12
GCS dpu>-- s+: a--- C++++ UL++++ P+ L+++ E---- W++ N++ o+ K- w--- 
O-- M- V- PS+++ PE- Y+ PGP t+ 5+ X++ R+ tv+ b++ DI++ D--- 
G e* h! !r y-- 
------END GEEK CODE BLOCK------
--
 5:03pm up 25 days, 22:10, 1 user, load average: 0.00, 0.02, 0.00


From aa8vb at yahoo.com  Wed Apr 12 10:08:16 2000
From: aa8vb at yahoo.com (Randall Hopper)
Date: Wed, 12 Apr 2000 10:08:16 -0400
Subject: Python idiom: Multiple search-and-replace
Message-ID: <20000412100816.A2431031@vislab.epa.gov>

There's got to be a better way.  Is there a Python idiom I'm missing?

I want to do search-and-replace of multiple symbols on each line of a file.
But the simple-minded code below takes a while.  It simply uses
string.replace N times per line (N is 240 in this case).  There are 33994
lines (1.1Meg).

Total time: 140.7 seconds.

I stopped to investigate.  What was slowing it up so much?

- Comment out the inner loop, and Python completes in 0.9 sec.  No prob
  there. It can read and write-the data very quickly.
- It's not dictionary lookups.  Converted to a tuple-list and took 3 sec
  longer.  
- I tried a few other things but it only made it take longer than
  140 sec.

Is there a Python feature or standard library API that will get me less
Python code spinning inside this loop?   re.multisub or equivalent? :-)

Thanks,

Randall


------------------------------------------------------------------------------

  symbol_map = { 'oldsym1' : 'newsym1', oldsym2' : 'newsym2', ... }
  fp = open( net_path, "r" )

  while 1:
    line = fp.readline()
    if not line: break

    for old_sym in symbol_map.keys():
      line = string.replace( line, old_sym, symbol_map[ old_sym ] )

    out_fp.write( line )

------------------------------------------------------------------------------

-- 
Randall Hopper
aa8vb at yahoo.com



From alan.gauld at gssec.bt.co.uk  Tue Apr  4 10:34:32 2000
From: alan.gauld at gssec.bt.co.uk (Alan Gauld)
Date: Tue, 04 Apr 2000 15:34:32 +0100
Subject: From Python Newbie to Python Guru on the Fast Track:  HOW?
References: <38E9B48F.F1827543@acm.org>
Message-ID: <38E9FD78.C740CF4E@gssec.bt.co.uk>

"Robert W. Cunningham" wrote:
> Given where I am, and where I hope to go, what is the fastest way to
> fill in the gaps in my knowledge?

Read the official tutorial for Python and Frederik's Tkinter 
tutor (for cross platform portability Tkinters probably the 
easiest way to go for now, GTk and Glade notwithstanding)

There are several good Python books for existing programmers
'Programming Python'(now a wee bit old), 
Essential Python(?), 
Quick Python and 
Learning Python all get good marks.
'Internet Programming with Python' may be relevant since 
it includes stuff on using sockets etc but its hard to find.

There is only 1 that I know of for Tkinter - Manning Press, 
written by John Grayson. Its probably worth buying.
Frederick is apparently publishing his tutor but I don't 
know when....

The O'Reilly 'Python for win32' may be a must have given 
your target environment.

> my only access to other Python expertise being this newsgroup 

Check out the mailing lists too - the tutor list would be 
a good starter but you will outgrow it pretty fast I suspect....

Also try some of the SIGs, they might be relevant.

Finally consider a web based approach to what you want, that 
might be simpler albeit more retricted in the long term...
It all looks mightily ambitious to me.

Alan G.
-- 
=================================================
This post represents the views of the author 
and does not necessarily accurately represent 
the views of BT.


From S.I.Reynolds at cs.bham.ac.uk  Wed Apr  5 09:01:28 2000
From: S.I.Reynolds at cs.bham.ac.uk (Stuart Reynolds)
Date: Wed, 05 Apr 2000 14:01:28 +0100
Subject: reading data from a file
References: <38EA94E3.7B249C3@seas.upenn.edu>
Message-ID: <38EB3928.267E@cs.bham.ac.uk>

Hi Dave,

dave white wrote:
> 
> Hello, can anyone please help with these basic questions:
> 
> 1)  I'm using file.readlines() to read in some data.  Is there any quick
> and efficient way to automatically get rid of the endofline '\012'
> character that gets appended to the strings at the end of every line?
 
  text = text[:-1]

Will chop off the last character of any sequence object.


> 2)  Is there a way for python to read a line from a file in such a way
> that it knows its a collection data type and not just a string? i.e.
> read something like [[1,2][3,4][5,6]] and know its a list of lists?
> 


--- pyfile.txt ----
"Nobody expects the Spanish Inquisition"
["Cat","Dog", 42]
99
99.0
----- read it in

for line in  open("pyfile.txt").readlines():
    obj = eval(line)
    print type(obj), obj
----

This prints out,

 <type 'string'> Nobody expects the Spanish Inquisition
 <type 'list'> ['Cat', 'Dog', 42]
 <type 'int'> 99
 <type 'float'> 99.0




Stuart


From wware at world.std.com  Thu Apr 20 16:03:24 2000
From: wware at world.std.com (Will Ware)
Date: Thu, 20 Apr 2000 20:03:24 GMT
Subject: Good style for multi-valued returns
References: <889A6EB6AE399393.619B0539BFCAAE0F.5C6D110C3DA908F7@lp.airnews.net>
Message-ID: <FtBzpo.92y@world.std.com>

Cameron Laird (claird at starbase.neosoft.com) wrote:
> Let me guess:  folks generally pass tuples for small and
> determinate-in-number multi-values

Yup, this is what I've always seen others do, and done myself.
This makes it easy to take advantage of tuple-wise assignments
from return values, e.g.:

	x, y, z = multiValuedFunction(arg1, arg2)

The dictionaries-for-larger-cases idea sounds good, but it almost
sounds like you're solving a different problem. My inexperienced
guess is that tuples would offer better performance than dictionaries,
but a 62-element tuple wouldn't be particularly self-documenting.
-- 
 - - - - - - - - - - - - - - - - - - - - - - - -
Resistance is futile. Capacitance is efficacious.
Will Ware	email:    wware @ world.std.com


From jvickroy at sec.noaa.gov  Fri Apr 21 15:53:20 2000
From: jvickroy at sec.noaa.gov (j vickroy)
Date: Fri, 21 Apr 2000 13:53:20 -0600
Subject: using PyArg_ParseTuple with a list
Message-ID: <3900B1B0.1CFD748A@sec.noaa.gov>

Hello all,

I wish to pass a Python list [] to a C extension module, but I am unsure
of the
proper format to specify in PyArg_ParseTuple (PyObject *args, char
*format, ...).

Should it be:

        (1) PyArg_ParseTuple (PyObject *args, "s#", ...) in which case
the list would be converted to a string with a length attribute
        (2) PyArg_ParseTuple (PyObject *args, "S", ...)
        (3) PyArg_ParseTuple (PyObject *args, "O&", ...) in which case I
would supply a converter function for the list

or something else?

Thanks in advance for your response.

- jv



From mjhand at concentric.net  Tue Apr 18 14:53:47 2000
From: mjhand at concentric.net (Manus Hand)
Date: Tue, 18 Apr 2000 12:53:47 -0600
Subject: BAD starship.python.net news!
References: <38FC852A.9B11F458@concentric.net> <sfp64err7l573@corp.supernews.com>
Message-ID: <38FCAF3B.497E1B72@concentric.net>

Thanks, Arnaldo!

To contribute funds, get in touch with Michael McLay (mclay at nist.gov).

With the support of the Python community, we will recover that disk!
(The power of positive thinking.)

Manus




From petersc at stud.ntnu.no  Wed Apr 26 12:20:34 2000
From: petersc at stud.ntnu.no (Peter Schneider-Kamp)
Date: Wed, 26 Apr 2000 18:20:34 +0200
Subject: arithmetic in longobject.c
Message-ID: <39071752.51DB78B9@stud.ntnu.no>

I got a few questions about longobject.c which a quick
glance at the source could not answer.

What algorithms does longobject.c implement for
multiplication and division of long integers?

What representation is used for long integers?

I'd like to add a rns representation option in python
for speedups in multiplication. Internal knowledge
of the long integer arithmetic would be helpful
to make the right design choices.

thnx for a quick reply,
Peter 
--
Peter Schneider-Kamp          ++47-7388-7331
Herman Krags veg 51-11        mailto:peter at schneider-kamp.de
N-7050 Trondheim              http://schneider-kamp.de


From gmcm at hypernet.com  Tue Apr 11 15:26:52 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Tue, 11 Apr 2000 15:26:52 -0400
Subject: AF_INET error
In-Reply-To: <8cvnve$l40$1@porthos.nl.uu.net>
Message-ID: <1256641280-42160733@hypernet.com>

Soundwave wrote:

> On www.python.org is a HOWTO guide for creating sockets, i'll give an
> example from what is said in the guide
> 
> This is the first example
> 
> import socket
> #create an INET, STREAMing socket
> s = socket.socket(
>     socket.AF_INET, socket.SOCK_STREAM)
> #now connect to the web server on port 80
> # - the normal http port
> s.connect("www.mcmillan-inc.com", 80)
> 
> when I run this code, I get the following output
> 
> Traceback (innermost last):
>   File "C:\PROGRA~1\PYTHON\TOOLS\IDLE\ScriptBinding.py", line 131, in
> run_module_event
>     execfile(filename, mod.__dict__)
>   File "C:/Program Files/Python/socket.py", line 3, in ?
>     s = socket.socket(
> AttributeError: AF_INET

It seems you have a module named socket.py which is being 
imported instead of the standard one (which would be 
Lib/socket.py).

- Gordon



From someone at somewhere.com  Tue Apr 18 12:44:54 2000
From: someone at somewhere.com (Charley Horse)
Date: Tue, 18 Apr 2000 16:44:54 GMT
Subject: Using python on the web
References: <MPG.1365adcfa8470eef9896aa@news> <8di2bk$l1s$1@newshost.accu.uu.nl>
Message-ID: <MPG.136631472f892a119896ac@news>

Thanks to all of those pointers. Finding out that CGI plays an important 
role in the python world was a shock to me as I only started working 
with web stuff in the last couple of years. I've just never touched a 
CGI script, shockingly I suppose <g>. ASP/CF and PHP are what I've 
worked with and CGI was just the "old world" technology for 
interactivity that I read about but never had to work with.

So coming at this as a newbie, and for years hearing about python's 
elegance and oo-ness, it just seemed a strange pairing, python and CGI. 
Puzzling in fact! I see why it makes sense, but it's just not what I 
expected.

I'll fool with CGi and perhaps it won't be enough of a turn off to 
affect my interest.

In article <8di2bk$l1s$1 at newshost.accu.uu.nl>, m.faassen at vet.uu.nl 
says...
> Charley Horse <someone at somewhere.com> wrote:
> > I've just started fooling with python. I've looked around the python 
> > site and a number of the other sites referenced there and am a bit 
> > puzzled. Umm... how does one go about using python for web applications, 
> > especially database driven apps? Almost everything deals with python 
> > just as a language, as C++ would be.
> 
> That's because Python is a language. Python isn't a web-specific language,
> after all. 
> 
> > Not much about the web specifically.
> > I was a bit shocked to find the only direct refs on about 
> > this at python.org are entwined with CGI (unless I missed it).
> 
> Why is CGI so shocking? It's used quite a bit, after all? You can write
> a database driven web application using Python with CGI. CGI from Python
> isn't hard either.
> 
> > I have no exposure to CGI, but rather to php, CF, and ASP, so python has
> > gone from looking very attractive to looking... well, like a puzzle.
> 
> I'm not sure why your exposure to PHP, ColdFusion or ASP instead of CGI
> would turn Python into a puzzle.
> 
> There _is_ ASP support for Python, enabling you to use Python as a
> ASP language. Others will have to explain to you how and where to look
> for information, as I've never used it.
> 
> > I know about Zope, but am interested in how one uses python on the web just
> > by itself.
> 
> Zope would be the equivalent of ColdFusion for Python, as it's a web
> application server.
> 
> Besides using Python with CGI or ASP, you can use Zope's ZPublisher
> (formerly Bobo) independently to publish arbitrary Python objects to the
> web. I've never used this, and the documentation may be scarce, but it
> looks quite neat. You may want to look that up.
> 
> Database integration in Python isn't bad either, look at the database
> topic guide:
> 
> http://www.python.org/topics/database/
> 
> And look at the web topic guide if you haven't seen it:
> 
> http://www.python.org/topics/web/
> 
> The vaults of Parnassus are also a good resource, here is for
> instance a link to its web category:


From jraven at psu.edu  Thu Apr  6 19:59:29 2000
From: jraven at psu.edu (Jeff Raven)
Date: 6 Apr 2000 23:59:29 GMT
Subject: Algorithm: combinations of (k) taken from (n) values
References: <meh9-ACCD3B.16041406042000@news.cit.cornell.edu>
Message-ID: <slrn8eq90h.le.jraven@corvus.local>

On Thu, 06 Apr 2000 16:04:14 -0400, Matthew Hirsch <meh9 at cornell.edu> wrote:
>Hi All,
>
>Onto the next question...
>
>Can anyone think of an algorithm to store as lists all possible 
>combinations of k numbers taken from n possible numbers.  For example, 
>given 5 values, I want to choose 2.  The number of possible combinations 
>is given by 5!/(3!2!)=10.  They are:
>
>[0,1], [1,2], [2,3], [3,4]
>[0,2], [1,3], [2,4]
>[0,3], [1,4]
>[0,4]
>
>I have something like:
>
>n=5
>
>for first_choice in range(n):
>   for second_choice in range(first_choice+1,n):
>      for third_choice in range(second_choice+1,n):
>         print first_choice, second_choice, third_choice
>
>
>The problem is that I'll be dealing with large combinations of numbers 
>and it doesn't make sense to continue writing nested for loops.  I'm 
>trying to figure out something that generates all combinations, just 
>based on values of k and n.  Any ideas?
>
>Thanks for your help,
>Matt

Well, the following seems to work. Just call choices(5,2), etc.

The main idea is just to define a function which will let you
iterate through all the combinations (that's succ()). Once you've
got that, it's easy.

def succ(choice, n, k) :
    # Returns the 'next' increasing k-element subset of
    # {0 ... n-1} after choice. It does so by working from
    # the end of the list, trying to increment elements.
    # Once it finds one it can increment, it then works
    # back towards the end filling in the remaining
    # values.
    index = k - 1
    while index >= 0 :
	choice[index] = choice[index] + 1
	if choice[index] <= n - k + index :
	    break
	index = index - 1
    else:
	return 0
    while index < k - 1 :
	choice[index + 1] = choice[index] + 1
	index = index + 1
    return 1

def choices(n, k) :
    # Returns a list consisting of all k-element (ordered)
    # subsets of {0 ... n-1}.
    result = []
    if k > n :
	return result

    current = range(k)
    while 1 :
	result.append(current[:])
	if not succ(current, n, k) :
	    break
    return result

Hope that helps,
Jeff Raven


From ns566645 at swt.edu  Sat Apr 15 18:46:11 2000
From: ns566645 at swt.edu (Shah, Navneet)
Date: Sat, 15 Apr 2000 17:46:11 -0500
Subject: Combobox in Tkinter
Message-ID: <38F8F133.54FCC64B@swt.edu>

How to decrease the width of the combobox in Tkinter? Do we have any
option for that

Shah



From frankr at first-mind.de  Wed Apr  5 12:29:42 2000
From: frankr at first-mind.de (Frank Rotermund)
Date: Wed, 5 Apr 2000 18:29:42 +0200
Subject: Extending Python with C-Libs
Message-ID: <8cfpl8$9b2$1@newnews.ilk.de>

Hello

I have a Library written in C with its Headerfiles but without the
Sourcecode.
Does anybody know an easy way to extend Python with this Library?

Thanks
Frank Rotermund







From news at heitlager.com  Tue Apr 11 05:17:10 2000
From: news at heitlager.com (Ilja Heitlager)
Date: Tue, 11 Apr 2000 11:17:10 +0200
Subject: using httpdlib with a proxy server
References: <ChoI4.844$g54.17363@news.uswest.net>
Message-ID: <8cuqhm$3f$1@weber.a2000.nl>

Lenny Self wrote in message ...
>I am attempting to write a small script that sends a list of URLs though a
>proxy server and then records a bit of information on the respose.  I
though
>I could use the httplib module to help with this, however, I don't see
where
>it can send a request though a proxy server. Does anyone know of a module
>that will allow me to do this?

You can use the httplib. You need to open the connection to the proxy server
and insert
a Host header variable with the requested site as value.

HTH,
Ilja Heitlager




From smalleys at gte.net  Mon Apr 24 17:54:15 2000
From: smalleys at gte.net (sue)
Date: Mon, 24 Apr 2000 21:54:15 GMT
Subject: A tkMessageBox question!
References: <38FDC335.B9FDCFE4@icestormfx.com> <20000421012547.10596@nms.otc.telstra.com.au>
Message-ID: <3904C29E.62231CC4@gte.net>

I know I've done this, so it probably was PMW.
Sue.


From greg at perceval.be  Thu Apr 27 04:07:06 2000
From: greg at perceval.be (Gregoire Welraeds)
Date: Thu, 27 Apr 2000 10:07:06 +0200 (CEST)
Subject: Simple del[] list question
In-Reply-To: <8e7b9g$r2s$1@newshost.accu.uu.nl>
Message-ID: <Pine.BSF.4.10.10004271004170.27597-100000@hercule.perceval.be>

In reply to the message of Martijn Faassen sent on Apr 26 (see below) :

> Note that another strategy which tends to avoid these troubles is to
> construct a new list if you modify one. That is, usually if you're going

Isn't it time and memory consuming when working on big list (eg with 1000
or above elements) ?


--
Life is not fair
But the root password helps
--

Gregoire Welraeds
greg at perceval.be
Perceval Development team
-------------------------------------------------------------------------------
Perceval Technologies sa/nv	Tel: +32-2-6409194		
Rue Tenbosch, 9			Fax: +32-2-6403154		
B-1000 Brussels			general information:   info at perceval.net
BELGIUM				technical information: helpdesk at perceval.net
URL: http://www.perceval.be/
-------------------------------------------------------------------------------

On 26 Apr 2000, Martijn Faassen wrote:

> Date: 26 Apr 2000 18:07:44 GMT
> From: Martijn Faassen <m.faassen at vet.uu.nl>
> To: python-list at python.org
> Newsgroups: comp.lang.python
> Subject: Re: Simple del[] list question
> 
> Matthew Hirsch <meh9 at cornell.edu> wrote:
> [snip problem description]
> > Why did a change?  Why suddenly when you go to two dimensions does the 
> > behavior change?
> 
> It's because [:] only does a shallow copy; that is, it copies the list,
> the individual elements are *not* copied; only new references are made.
> So when you delete an element of a sublist, both lists will now have 
> a reference to the same sublist which now misses an element.
> 
> Python is really very consistent about this, but tricky sometimes, indeed.
> 
> In order to make a deep copy of the list, use the deepcopy() function from
> the standard copy module.
> 
> Note that another strategy which tends to avoid these troubles is to
> construct a new list if you modify one. That is, usually if you're going
> to alter some list, you're going to alter multiple elements, or perhaps
> not include multiple elements, and not just a single one. In that case it's
> easier and possibly even faster to do something like this:
> 
> mylist = [...] 
> newlist = []
> for element in mylist:
>     if some_criterion(element):
>         newlist.append(element)
> mylist = newlist
> 
> By constructing a new list, you avoid the potential problems of some other
> variable still pointing to your modified list.
> 
> Regards,
> 
> Martijn
> -- 
> History of the 20th Century: WW1, WW2, WW3?
> No, WWW -- Could we be going in the right direction?
> -- 
> http://www.python.org/mailman/listinfo/python-list
> 
> 




From ivanlan at callware.com  Mon Apr 17 16:06:47 2000
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Mon, 17 Apr 2000 14:06:47 -0600
Subject: Playing wave files on Windows
Message-ID: <38FB6ED7.4C72773E@callware.com>

Hi All--
This is a FAQ.  I know it's a FAQ.  But I can't find the answer now.

How can I play a sound file on Windows.  I don't want to do anything
else, just play a wave file. ...

I will summarize the answers and compose a FAQ entry for addition to the
FAQ.

<eh-eh?-what'd-you-say-sonny?>-ly y'rs,
Ivan
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
http://www.pauahtun.org 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
Author:  Teach Yourself Python in 24 Hours



From jnw22 at cam.ac.uk  Tue Apr 11 06:22:47 2000
From: jnw22 at cam.ac.uk (James "Wez" Weatherall)
Date: Tue, 11 Apr 2000 11:22:47 +0100
Subject: Deadlocking IDLE under Win32
Message-ID: <8cuudo$8pj$1@pegasus.csx.cam.ac.uk>

I've noticed a bug with which a simple multithreaded test program such as:

import threading

class MyThread(threading.Thread):
    def run(self):
        while self.a < 1000:
            print self.a,
            self.a=self.a+2

thread1=MyThread()
thread2=MyThread()
thread1.a, thread2.a = 0, 1
thread1.start()
thread2.start()

This program will print out about twenty numbers and then hang the whole
interpreter if you run it from IDLE but it runs fine from the command-line
version, suggesting some locking problem in the Tk-based print command.

Does anyone have any ideas about this?

Cheers,

--
James "Wez" Weatherall





From paul.robinson at quantisci.co.uk  Thu Apr 13 04:53:57 2000
From: paul.robinson at quantisci.co.uk (Paul Robinson)
Date: Thu, 13 Apr 2000 09:53:57 +0100
Subject: Python CGI running commands under NT
References: <2B1262E83448D211AE4B00A0C9D61B03BA72AC@MSGEURO1>
Message-ID: <38F58B25.12E5FF7D@quantisci.co.uk>

Pieter Claerhout wrote:
> 
> Hello,
> 
> I made a Python CGI-script that executes a DOS-command and
> should return the result in a HTML page. The script looks
> as follows:
> 
>   import os
> 
>   pipe = os.popen("ipconfig", 'r')
>   text = pipe.read()

Pieter,
By default IIS does not create a console associated with a CGI process.
This means that the pipe has nowhere to read from!

See: http://support.microsoft.com/support/kb/articles/Q194/9/48.ASP

For some info on how you can change this behaviour. I've tried this in
the past with limited success (and I also found it hard to turn it off
again!).

	Paul.


-----------------------------------
Business Collaborator Team
Enviros Software Solutions
http://www.businesscollaborator.com
-----------------------------------



From cut_me_out at hotmail.com  Fri Apr 28 08:44:32 2000
From: cut_me_out at hotmail.com (Alex)
Date: 28 Apr 2000 08:44:32 -0400
Subject: sleep function?
References: <8ebvud$l6h$1@nnrp1.deja.com>
Message-ID: <etdu2gm5r9r.fsf@w20-575-109.mit.edu>

> Does python have a sleep function? e.g. sleep(60) to wait 60 seconds?

Yeah, in the time module.

import time
time.sleep (10)

> I have no python documentation around just now.

Do you know about the online documentation at python.org/doc?  
http://python.org/doc/current/lib/genindex.html would have been a good
place to start, in this case

Alex.


From moshez at math.huji.ac.il  Sat Apr 22 00:53:53 2000
From: moshez at math.huji.ac.il (Moshe Zadka)
Date: Sat, 22 Apr 2000 07:53:53 +0300 (IDT)
Subject: Python for drivers?
In-Reply-To: <3900EDBE.282BC85@dsuper.net>
Message-ID: <Pine.GSO.4.10.10004220752330.5248-100000@sundial>

On Sat, 22 Apr 2000, Dave wrote:

> Can Python be used for writing device drivers? I just got BeOS 5 and it
> doesn't support some video cards. I'd like to write one for the Cirrus
> Logic card and I want a language that's clear and concise- Python. I've
> never written device drivers. Any suggestions?

Yes. Python is a wonderful language, but don't write device drivers with
it. Use a language that allows you to get at the raw metal -- C (with some
assembly mixed in for good measure)

the-one-thing-Python-isn't-good-at-ly y'rs, Z.
--
Moshe Zadka <mzadka at geocities.com>. 
http://www.oreilly.com/news/prescod_0300.html
http://www.linux.org.il -- we put the penguin in .com




From dirck at pacbell.net  Fri Apr  7 22:53:28 2000
From: dirck at pacbell.net (Dirck Blaskey)
Date: Fri, 7 Apr 2000 19:53:28 -0700
Subject: Tkinter install?
References: <081c4ffb.329993cd@usw-ex0102-014.remarq.com>
Message-ID: <G8xH4.627$Qy3.199594@news.pacbell.net>

John Schmitt <jschmittNOjsSPAM at vmlabs.com.invalid> wrote in message
news:081c4ffb.329993cd at usw-ex0102-014.remarq.com...
> ...  I get this message in Tkinter.py:886:
>
> -------------------
> TclError: Can't find a usable init.tcl in the following
> directories:
>     {} ./lib/tcl8.0 D:/pyprojects/grayson/tcl8.0/library
> D:/pyprojects/grayson/Examples/library
>
> This probably means that Tcl wasn't installed properly.
> -------------------

This usually means you have extra copies of the tcl/tk dll's floating
around your system.  They should only be present in the tcl/tk install,
get rid of the others.  Here's a fix for the PATH issue, as well:
http://www.danbala.com/python/fixtk/

This has popped up several times recently -
if you need more details, search in
http://www.deja.com/group/comp.lang.python

good luck

d




From ullrich at math.okstate.edu  Fri Apr  7 12:16:41 2000
From: ullrich at math.okstate.edu (David C. Ullrich)
Date: Fri, 07 Apr 2000 11:16:41 -0500
Subject: Long Multiplication is not commutative.
References: <38ED0291.65CB7B5F@tismer.com>
Message-ID: <38EE09E9.40EBDAF7@math.okstate.edu>


Christian Tismer wrote:

> Yikes!
>
> No, it is computatively commutative, just not in terms
> of computation time. :-))
> [...]
> Swapping the multiplicands would be a 5 line patch.
> Should I submit it?

    My vote would be yes - long multiplication is slow enough
that it's hard to see why anyone would object to the tiny slowdown
in the case where the swap is not needed. (Hard for _me_ to see,
anyway, which proves very little...)

    (I've noticed the same phenomenon in my silly little "numeric"
types - always assumed it was fixed long ago in grownup types.)

>
> ciao - chris
>
> --
> Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
> Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
> Kaunstr. 26                  :    *Starship* http://starship.python.net
> 14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
> PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
>      where do you want to jump today?   http://www.stackless.com



From mhammond at skippinet.com.au  Fri Apr 28 19:09:14 2000
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Fri, 28 Apr 2000 23:09:14 GMT
Subject: Python COM troubles
References: <390a2f04.149999918@news.onlineisp.com> <L5MN4.5088$v85.30985@news-server.bigpond.net.au> <390acf5e.62946872@news.onlineisp.com>
Message-ID: <uSoO4.6813$v85.39864@news-server.bigpond.net.au>

"William Wicker" <remove.me.wwicker at spectratechnologies.com> wrote in
message news:390acf5e.62946872 at news.onlineisp.com...
> On Thu, 27 Apr 2000 00:46:35 GMT, "Mark Hammond"
> <mhammond at skippinet.com.au> wrote:
>
> >"William Wicker" <remove.me.wwicker at spectratechnologies.com> wrote in
> >message news:390a2f04.149999918 at news.onlineisp.com...

> >[BTW - drop them a line, and tell them they should fix their objects
:-]
> >
>
> Can you fill me in with what in particular I should complain about?
> Comments that RTTI was lacking brought the response that
> QueryInterface was supported.

Their IDispatch implementation should implement GetTypeInfo() - it should
be quite trivial to do, as they already have the type info available, it
just isnt connected to the objects at runtime.

> Also, (did I already ask this?) I'm guessing that using the
> EnsureModule approach will make DispatchWithEvents unavailable.
> Reading dynamic.py suggests that there is an alternate (more complex?)
> way.

It _should_ work - but instead of passing a CLSID, pass the object
itself - eg, instead of passing:

ob = DispatchWithEvents("Foo.Bar", your_class)

use:
mod = gencache.EnsureModule(...)
ob = mod.FooBar() # To work around the lack of type info.
ob = DispatchWithEvents(ob, your_class)

Mark.





From cgw at alum.mit.edu  Thu Apr  6 16:57:03 2000
From: cgw at alum.mit.edu (Charles G Waldman)
Date: Thu, 06 Apr 2000 14:57:03 -0600
Subject: Beginning to be disapointed
References: <Aw4H4.8078$pR.68745@carnaval.risq.qc.ca>
Message-ID: <8ciq6f$q5a$1@info3.fnal.gov>

How about posting some code?  Try to whittle down your 
program to the smallest fragment which can reproduce
the offending behavior, then post that code to the 
newsgroup.



From tismer at tismer.com  Sun Apr  2 12:33:58 2000
From: tismer at tismer.com (Christian Tismer)
Date: Sun, 02 Apr 2000 18:33:58 +0200
Subject: Debugging confusion -- too many stacks!
References: <000a01bf9c6b$3f80c2a0$752d153f@tim> <38E739A1.45F4B199@tismer.com> <030301bf9cd9$eb6786d0$6401a8c0@dorb>
Message-ID: <38E77676.4599978B@tismer.com>


Darrell wrote:
> 
> From: "Christian Tismer"
> > The following watches are always active when I'm debuging:
> > (are there similar ways to do this with PDB?)
> >
> > (char*)(f->f_code->co_name)+20       # current function name
> > (char*)(f->f_code->co_filename)+20   # current file name
> > f->f_lineno                          # line number
> >
> > In order to see which frame is calling you, do the same, but
> > replace "f" by "f->f_back".
> 
> pdb works pretty good.
> I use it all the time. You can get the call stack or what ever you want to
> look at.

Oops, sorry! I confused PDB with GDB.
Probably since the issue was about stacks...

ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaunstr. 26                  :    *Starship* http://starship.python.net
14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     we're tired of banana software - shipped green, ripens at home



From rgb122 at my-deja.com  Mon Apr  3 16:16:45 2000
From: rgb122 at my-deja.com (rgb122 at my-deja.com)
Date: Mon, 03 Apr 2000 20:16:45 GMT
Subject: Loading image from URL with PIL
Message-ID: <8cau6t$pph$1@nnrp1.deja.com>

Is it possible to load an image (GIF or JPEG)
straight from a URL without saving it to a file
first. I tried this:

u = urllib.urlopen
("http://localhost/cache/test.jpg")
data = u.read()
s = array.array("b", data).tostring()

But then how can I get the dimensions of the
image to pass to the fromstring function?

i = Image.fromstring("RGB", (???, ???), s)

Also, how can I convert a GIF to a JPEG using PIL?

Thanks,
Rob


Sent via Deja.com http://www.deja.com/
Before you buy.


From thomas at xs4all.net  Wed Apr 19 14:45:33 2000
From: thomas at xs4all.net (Thomas Wouters)
Date: Wed, 19 Apr 2000 20:45:33 +0200
Subject: python-list/mailman (Was: module __call__)
In-Reply-To: <jrkL4.684$vE6.202902@news.pacbell.net>; from dirck@pacbell.net on Wed, Apr 19, 2000 at 08:39:59AM -0700
References: <LX2L4.1275$B43.102174@news.pacbell.net> <7N4L4.497$wYl.191816704@newsb.telia.net> <286L4.1328$B43.118787@news.pacbell.net> <20000419102700.M15664@xs4all.nl> <jrkL4.684$vE6.202902@news.pacbell.net>
Message-ID: <20000419204532.O15664@xs4all.nl>

On Wed, Apr 19, 2000 at 08:39:59AM -0700, Dirck Blaskey wrote:

> Thank you for the relevant explanation.

No problem, though I should point out to the list that this message probably
never made it here: it was bounced by postfix because the mailman list
delivery command was 'running too long'. This happened before, either
yesterday morning or monday morning, I forget.

It might be a good idea for whoever operates postfix on the relevant
machine(s) to check wether it was a deadlock in mailman, too high a load on
the machine or just a too impatient postfix ? Barry, would that be you ? (I
can find you exact times for the error messages, or mail you the bounces, if
you wish, by the way.)

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From quinn at steal.ugcs.caltech.edu  Fri Apr 21 20:49:54 2000
From: quinn at steal.ugcs.caltech.edu (Quinn Dunkan)
Date: 22 Apr 2000 00:49:54 GMT
Subject: Overloading methods?
References: <eviltofu-3DE33B.17272521042000@news.earthlink.net>
Message-ID: <slrn8g1tu1.4uj.quinn@steal.ugcs.caltech.edu>

On Fri, 21 Apr 2000 21:27:41 GMT, Jerome Chan <eviltofu at rocketmail.com> wrote:
>Does this mean that method overloading is not allowed?

Yes.  python doesn't do method overloading (it does overload operators with
some magic method names, though).  Keyword args are quite nice, though.  It's
also possible to use the general purpose **kwargs and then have each method
pick out the args it's interested in.  It's probably safer and nicer to just
have the same signature for all methods of the same name.


From guido at python.org  Mon Apr  3 12:04:02 2000
From: guido at python.org (Guido van Rossum)
Date: Mon, 03 Apr 2000 12:04:02 -0400
Subject: [Python-Dev] New Features in Python 1.6
In-Reply-To: Your message of "Sat, 01 Apr 2000 12:00:00 EST."
             <200004011740.MAA04675@eric.cnri.reston.va.us> 
References: <200004011740.MAA04675@eric.cnri.reston.va.us> 
Message-ID: <200004031604.MAA05283@eric.cnri.reston.va.us>

Not only was it an April fool's joke, but it wasn't mine!  It was
forged by an insider.  I know by who, but won't tell, because it was
so good.  It shows that I can trust to delegate way more to the Python
community than I think I can! :-)

BTW, the biggest give-away that it wasn't mine was the absence of my
standard sign-off line:

--Guido van Rossum (home page: http://www.python.org/~guido/)



From tismer at tismer.com  Thu Apr 20 13:03:12 2000
From: tismer at tismer.com (Christian Tismer)
Date: Thu, 20 Apr 2000 19:03:12 +0200
Subject: Creating Classes Dynamically?
References: <eviltofu-783F16.12334920042000@news.earthlink.net>
Message-ID: <38FF3850.E1053517@tismer.com>


Jerome Chan wrote:
> 
> Is there a way to create an object of a class from a string?

Yes, you can use exec on the string if it evaluates as a call
to the class.
If you don't want to use exec, you also have the option to use the
new module. Import it and read the doc strings.

ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaunstr. 26                  :    *Starship* http://starship.python.net
14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     where do you want to jump today?   http://www.stackless.com



From tom at parlant.com  Tue Apr  4 16:22:33 2000
From: tom at parlant.com (Thomas Lane)
Date: Tue, 04 Apr 2000 14:22:33 -0600
Subject: tkinter - changing the mouse pointer?
References: <38E91666.D3B87B4C@parlant.com> <%IoG4.5028$74.81199@newsc.telia.net>
Message-ID: <38EA4F09.2D2A24A3@parlant.com>

root.config(cursor="wait")
root.config(cursor="")

So simple. Thank you Fredrik.

-Tom

Fredrik Lundh wrote:
> 
> Thomas Lane <tom at parlant.com> wrote:
> > Does anyone out there know how to change the mouse pointer in tkinter? I'd
> like
> > to change the mouse pointer to an hourglass (or the default 'wait' icon
> for the
> > os) and back to normal after a short operation. How do I go about doing
> this?
> > I've hunted through the Grayson book without success.
> 
> this might help:
> http://www.deja.com/getdoc.xp?AN=392027946
> 
> </F>


From tom.kralidis at ccrs.nrcanDOTgc.ca  Fri Apr 21 08:45:47 2000
From: tom.kralidis at ccrs.nrcanDOTgc.ca (Tom Kralidis)
Date: Fri, 21 Apr 2000 08:45:47 -0400
Subject: Python Language Support for ArcView GIS
References: <q0MH4.2398$Bq2.38636@newscontent-01.sprint.ca>
Message-ID: <39004D7B.A1FFA037@ccrs.nrcanDOTgc.ca>

Bruce Dodson wrote:
> 
> Python Language Support for ArcView GIS (aka AVPython) is now available
> online as a free download.  I promised this several months ago but never
> managed to find a home for it (too busy) so finally I decided to just slap
> together a page for it on GeoCities:
> www.geocities.com/brucedodson.rm/avpython.htm
> 
> If you're an ArcView developer and want to be able to control COM-enabled
> programs like Excel or ArcInfo, or if you want to be able to interoperate
> with distributed systems through standards like XML and HTTP, you might want
> to give it a try.  The same goes for Python programmers who happen to be
> working with ArcView GIS (hey, there must be a few!), and want to be able to
> apply their Python skills (and use a better language) when developing
> ArcView applications.
> 
> Python is free and open source.  I hope AVPython will eventually be the
> same, but we haven't yet settled on an acceptable open source license, so
> for now the best I can do is provide a binary that is freely
> redistributable.
> 
> If you use AVPython on any interesting projects, or if you have any positive
> feedback or things that you'd like to change, feed my hubris: tell me about
> it!  Just take nospam out of my address, or use the one listed in the
> documentation.
> 
> Sincerely,
> Bruce Dodson

How about Perl (or C) language support?
-- 
=================================
Tom Kralidis
Geo-Spatial Developer
Canada Centre for Remote Sensing
Tel: (613) 947-1828
Fax: (613) 947-1408
http://www.ccrs.nrcan.gc.ca
=================================


From ke at gnu.franken.de  Fri Apr  7 13:29:40 2000
From: ke at gnu.franken.de (Karl EICHWALDER)
Date: 07 Apr 2000 19:29:40 +0200
Subject: Paid to work in Python? (was: Re: Choice Of Language)
References: <Uqvw4.2261$yL2.28868@typhoon.southeast.rr.com> <8c3ats$b5f$1@nntp.Stanford.EDU> <slrn8eact2.sq7.nobody@bambi.visi.com> <38e5f8ed.29659873@news.nikoma.de> <8cl2mq$971$1@mawar.singnet.com.sg>
Message-ID: <shsnwxn7jv.fsf@tux.gnu.franken.de>

ngps at madcap.dyndns.org (Ng Pheng Siong) writes:

|   I once saw a German term that was translated to "improve-worse-ment."
|   
|   Can some one provide the German word, please?

Most probably, you're talking about "Verschlimmbesserung" (literally: ~=
"worse-improve-ment").

-- 
work : ke at suse.de                          |
     : http://www.suse.de/~ke/             |          ------    ,__o
home : ke at gnu.franken.de                   |         ------   _-\_<,
     : http://www.franken.de/users/gnu/ke/ |        ------   (*)/'(*)


From janeb at stud.ntnu.no  Sat Apr 29 13:05:59 2000
From: janeb at stud.ntnu.no (Jan Erik Breimo)
Date: 29 Apr 2000 17:05:59 GMT
Subject: Where can I find NumPy
Message-ID: <8ef4pn$jge$1@kopp.stud.ntnu.no>

The official site (ftp-icf.llnl.gov) hasn't been responding for more than
a week, so are there any other sites where I can find the source
files for NumPy?

--
Jan Erik Breimo
email: janeb at stud.ntnu.no


From gansevle at cs.utwente.nl  Thu Apr 13 07:28:26 2000
From: gansevle at cs.utwente.nl (Fred Gansevles)
Date: Thu, 13 Apr 2000 13:28:26 +0200
Subject: nisplus -- help needed
Message-ID: <38F5AF5A.168E7C71@cs.utwente.nl>

A few weeks ago, Guido asked me to write a NIS+ module for python.
I wrote the original NIS module (that's why he asked me) and I know how
NIS works.
The past few weeks I've been reading documentation about NIS+ and I
*think* I know
how to implement a nisplus extention module, but before I start to
implement this
module, I want to 'feel' a NIS+ environment.

My question to the python community is:

    can anyone help me to get *access* to a NIS+ domain?

Adding me and my system to such a domain woul help me a lot.
Having an account on a computer in a NIS+ domain would also be very
usefull.

So, are there any NIS+ administrators out there who can help me  (and
the python
community as a whole)?

Thanks in advance.

P.S.: Please mail your replies, since I don't have the time to regulary
read the c.l.p. newsgroup.
____________________________________________________________________________

 Fred Gansevles <mailto:Fred.Gansevles at cs.utwente.nl> Phone: +31 53 489
4613
        >>> Your one-stop-shop for Linux/WinNT/NetWare <<<
 Org.: Twente University, Fac. of CS, Box 217, 7500 AE Enschede,
Netherlands
            "Bill needs more time to learn Linux" - Steve B.



From rhicks at nospam.rma.edu  Mon Apr 17 11:08:46 2000
From: rhicks at nospam.rma.edu (Robert Hicks)
Date: Mon, 17 Apr 2000 15:08:46 GMT
Subject: a progress bar in tkinter ?
References: <38FAD6CA.6198CF2A@teleatlas.com>
Message-ID: <2OFK4.9$Pa.1970@news>

A Tkinter progress bar...and I did not create this nor have I used it.  I am
only sharing it. It comes from John Graysons book "Python and Tkinter
programming".

"""
A  basic widget for showing the progress
being made in a task.

"""

from Tkinter import *

class ProgressBar:
    def __init__(self, master=None, orientation="horizontal",
                 min=0, max=100, width=100, height=18,
                 doLabel=1, appearance="sunken",
   fillColor="blue", background="gray",
                 labelColor="yellow", labelFont="Verdana",
                 labelText="", labelFormat="%d%%",
                 value=50, bd=2):
 # preserve various values
 self.master=master
 self.orientation=orientation
 self.min=min
 self.max=max
 self.width=width
 self.height=height
 self.doLabel=doLabel
 self.fillColor=fillColor
        self.labelFont= labelFont
 self.labelColor=labelColor
 self.background=background
 self.labelText=labelText
 self.labelFormat=labelFormat
 self.value=value
 self.frame=Frame(master, relief=appearance, bd=bd)
 self.canvas=Canvas(self.frame, height=height, width=width, bd=0,
                           highlightthickness=0, background=background)
 self.scale=self.canvas.create_rectangle(0, 0, width, height,
      fill=fillColor)
 self.label=self.canvas.create_text(self.canvas.winfo_reqwidth() / 2,
        height / 2, text=labelText,
        anchor="c", fill=labelColor,
        font=self.labelFont)
 self.update()
 self.canvas.pack(side='top', fill='x', expand='no')

    def updateProgress(self, newValue, newMax=None):
        if newMax:
            self.max = newMax
        self.value = newValue
        self.update()

    def update(self):
 # Trim the values to be between min and max
 value=self.value
 if value > self.max:
     value = self.max
 if value < self.min:
     value = self.min
 # Adjust the rectangle
 if self.orientation == "horizontal":
     self.canvas.coords(self.scale, 0, 0,
       float(value) / self.max * self.width, self.height)
 else:
     self.canvas.coords(self.scale, 0,
                     self.height - (float(value) / self.max*self.height),
       self.width, self.height)
 # Now update the colors
 self.canvas.itemconfig(self.scale, fill=self.fillColor)
 self.canvas.itemconfig(self.label, fill=self.labelColor)
 # And update the label
 if self.doLabel:
            if value:
                if value >= 0:
                    pvalue = int((float(value) / float(self.max)) * 100.0)
                else:
                    pvalue = 0
                self.canvas.itemconfig(self.label, text=self.labelFormat %
pvalue)
            else:
                self.canvas.itemconfig(self.label, text='')
 else:
     self.canvas.itemconfig(self.label, text=self.labelFormat %
self.labelText)
 self.canvas.update_idletasks()


Hope this helps!

Bob

"Ralf Claus" <ralf.claus at teleatlas.com> wrote in message
news:38FAD6CA.6198CF2A at teleatlas.com...
> Hello,
> i search about a "widget", that works like a ProgressBar in VB.
> Is there a possibility given by Tkinter?
>
> thanks
>
> Ralf




From jrl at clark.net  Thu Apr 13 19:31:18 2000
From: jrl at clark.net (John Lloyd)
Date: Thu, 13 Apr 2000 18:31:18 -0500
Subject: Python 1.6a2 crashes in test_fork1 on my Mandrake Linux 7 machine
References: <20000413010344.A13931@owl.rhein-zeitung.de>
Message-ID: <38F658C6.66AD9CA5@clark.net>

Oliver Andrich wrote:

> Hi,
>
> I just tried to build a first alpha release of the Linux distribution of
> Python, and the testall.py run crashes on the test_fork1 script. Sometimes it
> just hangs, as soon as I switch on full optimizations during compilation I get
> a nice seg fault. The freezes occurs when I compile python without any but the
> standard optimizations on my machine. I am running glibc 2.1.2 and gcc 2.95.2.
> With 1.5.2 this works really fine.
>
> I will try to investigate this issue. But for this I have to check the CVS
> logs of the relevant files. Anybody with a useful hint should feel free to
> mail me. ;-))
>
> Best regards,
>
>     Oliver

I had the same fault but when I compiled without thread support it works OK.
I have not any fork opperations however.

HTH,
John




From nobody at nowhere.nohow  Tue Apr 25 00:22:17 2000
From: nobody at nowhere.nohow (Grant Edwards)
Date: Tue, 25 Apr 2000 04:22:17 GMT
Subject: Tkinter
References: <A63AFB20111ED311AC5500A0C96B7BF5015F0F44@orsmsx36.jf.intel.com> <vD5N4.1521$wYl.189935104@newsb.telia.net> <39050A2A.D71E9EFD@callware.com>
Message-ID: <slrn8ga7bp.87r.nobody@isis.visi.com>

In article <39050A2A.D71E9EFD at callware.com>, Ivan Van Laningham wrote:

>2) If you're at a more advanced level, where the /F documentation doesn't
>cover enough, or not in enough depth, take a look at John Grayson's _Python
>and Tkinter Programming_.  John's writing style is appealing, he covers the
>topics in admirable depth, and he gives clever and entertaining examples.

I just bought a copy on Saturday, and have been skipping around reading a
section here and there over the past couple days.  I'm very impressed so
far.  I like the writing style, and the examples are well thought out and
detailed enough to show real-world usage.

-- 
Grant Edwards                   grante             Yow!  Mary Tyler Moore's
                                  at               SEVENTH HUSBAND is wearing
                               visi.com            my DACRON TANK TOP in a
                                                   cheap hotel in HONOLULU!


From oli at rz-online.net  Wed Apr 19 09:42:03 2000
From: oli at rz-online.net (Oliver Andrich)
Date: Wed, 19 Apr 2000 15:42:03 +0200
Subject: Pythonware Website???? Gone on vacation with Starship? ;-))
Message-ID: <20000419154202.B480@owl.rhein-zeitung.de>

Hi,

I can't access the www.pythonware.com webpages since yesterday evening. Does
someelse also have the problems or is it just me?

Bye, Oliver




From kjl at inforonics.com  Thu Apr 20 16:51:43 2000
From: kjl at inforonics.com (Kurt J. Lanza)
Date: Thu, 20 Apr 2000 16:51:43 -0400
Subject: Using python on the web
References: <MPG.1365adcfa8470eef9896aa@news> <MPG.136631472f892a119896ac@news> <_W2L4.986$GK5.20898@news2-win.server.ntlworld.com> <MPG.13665ef93eedbda29896ad@news> <BC4B61746187C623.3FA0E7D58DF4EA7E.ECC1D97F577DEDE5@lp.airnews.net>
Message-ID: <38FF6DDF.1284E054@inforonics.com>

Cameron Laird wrote:
[ snip ]
> 
> I recognize no one here is disputing the proposition.  It's
> just struck me recently how pervasive the propaganda on this
> score is.  CGI has an *abyssmal* popular reputation.

I kinda like it...


From claird at starbase.neosoft.com  Thu Apr 20 16:10:17 2000
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 20 Apr 2000 15:10:17 -0500
Subject: Using python on the web
References: <MPG.1365adcfa8470eef9896aa@news> <MPG.136631472f892a119896ac@news> <_W2L4.986$GK5.20898@news2-win.server.ntlworld.com> <MPG.13665ef93eedbda29896ad@news>
Message-ID: <BC4B61746187C623.3FA0E7D58DF4EA7E.ECC1D97F577DEDE5@lp.airnews.net>

In article <MPG.13665ef93eedbda29896ad at news>,
Charley Horse  <someone at somewhere.com> wrote:
>It was just an unexpected surprise. I'm fairly new to all of this and 
>always heard of CGI in the context of "you don't have to user CGI 
>anymore because of the fancy new blah blah technology we offer!" I'll 
>try it and I'm sure it'll be fine. If all of you clean-code loving 
>python users can live with it then I'm sure I can too. <g>
			.
			.
			.
CGI has been dead for at least five years.

That is, in 1995 already, knowledgeable people were saying
that there was no good reason to start a new project with
CGI.

I was briefly one of them.

We were wrong.

There certainly are fancy new technologies.  I use several
of them.  In many cases, though, they don't even do as ad-
vertised (that is, perform better, code with more ease, ...
than CGI-based alternatives), let alone compete with CGI's
portability, ...

I recognize no one here is disputing the proposition.  It's
just struck me recently how pervasive the propaganda on this
score is.  CGI has an *abyssmal* popular reputation.
-- 

Cameron Laird <claird at NeoSoft.com>
Business:  http://www.Phaseit.net
Personal:  http://starbase.neosoft.com/~claird/home.html


From ng at hardlight.couk.com  Fri Apr 14 05:50:58 2000
From: ng at hardlight.couk.com (.:|:.)
Date: Fri, 14 Apr 2000 10:50:58 +0100
Subject: Complaint Dept: My shoes!
References: <8d4s4d$mvr$1@nnrp1.deja.com>
Message-ID: <38f6eb03.0@news.proweb.co.uk>

<musingattheruins at my-deja.com> wrote in message
news:8d4s4d$mvr$1 at nnrp1.deja.com...
> If shoe is None then len(shoe) raises an exception.  Would be nice if
> len(None) was 0 and NOT raise an exeption.
>
> Example: (Python 1.5.2)
>
> def show_shoe_wear_to_complaint_dept(name):
>    if name == "Idle":
>       return None
>    else:
>       return "pick up hammer..."
>
> def hit_on_head(): return "POW!!!"
>
> shoe = ""
> shoe = show_shoe_wear_to_complaint_dept("Idle")
> if len(shoe)>0: print shoe, hit_on_head()
>
> If len(shoe) returned 0 then the example would work without error, the
> example above raises an exception whenever Idle shows his shoe to the
> complaint department.


try

if shoe:
    print shoe, hit_on_head()

an exception is appropriate

Matt




From jbauer at rubic.com  Thu Apr 20 14:22:14 2000
From: jbauer at rubic.com (Jeff Bauer)
Date: Thu, 20 Apr 2000 13:22:14 -0500
Subject: editor for win32
References: <NpHL4.779$fK2.113609@news.flash.net>
Message-ID: <38FF4AD6.78C72B34@rubic.com>

Scott Hathaway wrote:
> Can anyone recommend a python editor for Windows?

I'm quite fond of using XEmacs on Windows:

  http://www.xemacs.org

though-fonder-of-it-running-on-Solaris-ly y'rs,

Jeff Bauer
Rubicon Research



From mhammond at skippinet.com.au  Wed Apr 26 20:49:55 2000
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Thu, 27 Apr 2000 00:49:55 GMT
Subject: Need help with COMObject for Zope
References: <8e7jeo$s8c$1@nnrp1.deja.com>
Message-ID: <T8MN4.5091$v85.31034@news-server.bigpond.net.au>

<deepak_marda at my-deja.com> wrote in message
news:8e7jeo$s8c$1 at nnrp1.deja.com...

> I created a COMObject with id = word and Dispatch
> = "Word.Application". WHen I try to connect this
> COM object, it raises the following exception
> Error Type: com_error
> Error Value: (-2147221008, 'CoInitialize has not
> been called.', None, None)

Add the 2 lines:
import pythoncom
pythoncom.CoInitialize()

before your code, and it will almost certainly work.

The Zope guys should document this fairly well.  Note that Zope is almost
certainly initializing COM on a different thread than your code is being
called on (hence you see this message) - but there are some serious
implications with COM and threading.  I dont believe you will hit them
here, but someone will sometime :-)

Mark.





From cjc26 at nospam.cornell.edu  Sat Apr  8 16:15:43 2000
From: cjc26 at nospam.cornell.edu (Cliff Crawford)
Date: Sat, 08 Apr 2000 20:15:43 GMT
Subject: Searching for Python Material...
References: <8F0ED18C1joshtinamecom@209.208.0.7>
Message-ID: <slrn8ev51o.2u85.cjc26@ith1-491.twcny.rr.com>

Try the Topic Guides and HOWTOs:
<http://www.python.org/topics/>
<http://www.python.org/doc/howto/>

* Josh Tompkins <josht at iname.com> menulis:
| [..]
| So, anyway, I was wondering if anyone could point me to some websites that 
| cover material that is more advanced than the stuff in most of the 
| tutorials out there.  I'm actually looking for something along the list of:
| 
| to do (something cool in python) do this...
| 
| where something cool in python is a variable with values such as "Accessing 
| webpages and/or e-mail", "Connecting to a database", etc, while still 
| keeping the deep magic stuff to a relative minimum.


-- 
cliff crawford    -><-    http://www.people.cornell.edu/pages/cjc26/
                                                       icq 68165166
"IS I yes wardrobe yield [the] evaluation."     member of A.H.M.A.D.


From glhansen at steel.ucs.indiana.edu  Fri Apr  7 14:02:25 2000
From: glhansen at steel.ucs.indiana.edu (Gregory L. Hansen)
Date: 7 Apr 2000 18:02:25 GMT
Subject: No threads in MacPython 1.5.2?
Message-ID: <8cl7rh$kvm$3@flotsam.uits.indiana.edu>

My Python can find Tkinter now, thanks for the help.  But now it can't
find the module named "thread".  And yet there it is in the Lib folder,
thread.py and threading.py, and the search path $(PYTHON):Lib is set for
it.

Now what's wrong?

-- 
"Ballpark nachos -- you don't buy them, you just rent them." -- Hank Hill



From junkster at nospam.rochester.rr.com  Sat Apr 29 03:38:56 2000
From: junkster at nospam.rochester.rr.com (Benjamin Schollnick)
Date: Sat, 29 Apr 2000 07:38:56 GMT
Subject: ODBC / DBI help (via Win32 Extensions)
References: <rNHvjEdhm5Pp-pn2-RytgtQHbLg9d@d185d18f2.rochester.rr.com> <RRtO4.26899$R4.157537@news1.rdc1.nj.home.com>
Message-ID: <rNHvjEdhm5Pp-pn2-dzBINcOCBoWi@d185d18f2.rochester.rr.com>

On Sun, 29 Apr 3900 04:49:53, "S. Hoon Yoon" <cpu at bigfoot.com> wrote:

> I think that's because you should use format #date# on update. Unless, I did
> not understand your problems.
> That's really more of SQL and access convention rather than Python.
> Pls look at the docs for Access and SQL.
> 
> UPDATE [BT Dates] SET [BT Dates].[date] = #12/12/2000#;
> 
> This is what I get if I use Query wizard and you should follow it.
> Hope it helps,

The real problem is that I can't find any documentation on the DBI 
module...

The pound convention I found out about...
Unfornately, I don't have my code with me... But...(From memory)

sql_cmd = """
		INSERT INTO logfile (datetime, loglevel,, logdata)
		VALUES ('#12/15/2000#', 1, 'test')
	"""
	# I've tried 12/15/00, 09/09/99, 09/09/1999, and a few other
	# date variants....

	# And as far as I can tell the ODBC execute command
	# appends a semicolon?  Or at least everything works
	# without having me add one...

If I attempt to execute that command, I get a SQL error from the ODBC
driver.

If I remove the datetime field and just submit using the loglevel, 
logdata
it works fine... (But that sort of defeats the purpose of a log 
file....)

The datetime field is setup in Access as a standard "Date/Time"
field, so I figure I might need to convert the datetime field into a 
dbiDate
record, but as I mentioned I can't find any instructions on the fields
&
methods for the dbiDate field(s).... (I checked the Win 32 
documentation
without any luck).

		- Benjamin	


From ejbaker_seekonk at my-deja.com  Thu Apr 20 07:49:38 2000
From: ejbaker_seekonk at my-deja.com (ejbaker_seekonk at my-deja.com)
Date: Thu, 20 Apr 2000 11:49:38 GMT
Subject: Confused by List behavior
Message-ID: <8dmqsd$6m0$1@nnrp1.deja.com>

As a python newbie, I entered this script to explore list processing
behaviours.

yy = dir()
for element in yy:
  print element

for element in yy:
  print dir(element)

The first 'for' loop works as expected. It returns,
['__builtins__', '__doc__', '__name__'].

The second loop, however, returns three empty lists. For the life of
me, I can't understand that. Becauese, if I type 'dir(__builtins__)',
in a script, I get all of the elements of __builtins__.

Can anyone explain this?


Sent via Deja.com http://www.deja.com/
Before you buy.


From vanandel at atd.ucar.edu  Mon Apr  3 17:38:27 2000
From: vanandel at atd.ucar.edu (Joe Van Andel)
Date: Mon, 03 Apr 2000 15:38:27 -0600
Subject: How to build shared Python binaries under Solaris 7 (sparc) ?
References: <200003291605.SAA04572@mail.inhh.de>
Message-ID: <38E90F53.CF9571F1@atd.ucar.edu>

I've discovered a problem with the configure script for Python 1.5.2
under Solaris 7 with gcc 2.95 .  configure should set the make variable
LINKFORSHARED=-Xlinker --export-dynamic

Until I made this change, I could not load any dynamicly linked library
into Python.  imports would fail - I would get errors like:

ImportError: ld.so.1: python: fatal: relocation error: file
/usr/local/lib/python1.5/site-packages/Perp/util/du_SweepUtilc.so:
symbol PyExc_RuntimeError: referenced symbol not found


Once I manually set this variable in
Modules/Makefile, I could build a Python executable that could load
shared library extensions (xxx.so) .


-- 
Joe VanAndel  	          
National Center for Atmospheric Research
http://www.atd.ucar.edu/~vanandel/
Internet: vanandel at ucar.edu


From wware at world.std.com  Wed Apr 12 15:28:35 2000
From: wware at world.std.com (Will Ware)
Date: Wed, 12 Apr 2000 19:28:35 GMT
Subject: Off-topic
References: <38F4BE12.A0C73D1D@callware.com>
Message-ID: <Fsx4rn.MpD@world.std.com>

Ivan Van Laningham (ivanlan at callware.com) wrote:
> Why don't #! scripts seem to work anymore?  #!/usr/bin/python and
> #!/usr/bin/perl scripts don't work.  Error is

> bash:  ./h2n:  No such file or directory

This usually means that bash can't find the executable mentioned
in the first line of your script. On RH6.1 you should find python
in /usr/bin, but if you're seeing this message maybe something
happened to it. Just for yucks, type "which python", then it will
tell you where it expects to find python.

Since I sometimes run scripts on different machines, I've found
that using this first line works better:

#!/usr/bin/env python

This of course assumes that "env" is in /usr/bin, but that seems
to be a more consistent rule than finding python in /usr/bin.
-- 
 - - - - - - - - - - - - - - - - - - - - - - - -
Resistance is futile. Capacitance is efficacious.
Will Ware	email:    wware @ world.std.com


From grant at nowhere.  Mon Apr 10 10:29:35 2000
From: grant at nowhere. (Grant Edwards)
Date: Mon, 10 Apr 2000 14:29:35 GMT
Subject: Python paradigms
References: <Pine.LNX.4.21.0004100757580.19783-100000@fep132.fep.ru> <8cs70f$ojf$1@pegasus.csx.cam.ac.uk> <38F1C600.3FD9878B@acm.org>
Message-ID: <slrn8f3paf.opc.grant@grante.comtrol.com>

In article <38F1C600.3FD9878B at acm.org>, Robert W. Cunningham wrote:
>
>> |> > So, it works just fine, right?
>> |>
>> |>    No. Code is intended not only to beeing run on processor, but also (and
>> |> often more important) to beeing read by human. Code is speech, it is the
>> |> way to interchange ideas.
>> |>    In that sense ugly code is not working even if you add three books of
>> |> comments. Code should be readable. Period.
>>
>> Yes, precisely.  In the real world, programs often have long
>> lifetimes, are worked on by many programmers (in parallel and
>> sequentially), are sometimes transliterated between languages,
>> and it is common to need to modify code which was written 20
>> years ago by some long-retired programmer.
>>
>> At best, confusing code wastes effort.  At worst, it causes
>> mistakes to be introduced.  I have seen the latter effect many
>> times.

>Again, comments in the code can cover all these bases.  

Not in my experience.  There's one problem with comments:
They're wrong.  Either a) they never matched the code, or b)
somebody changed the code and forgot to fix the comments.

On the legacy stuff I've worked on what the comments say rarely
matches what the code does.  Perhaps you're experience with the
veracity of comments is better than mine, but after 20 years of
working on other people's (and my) old code, I've learned one
thing the hard way:

**Ignore the comments.  They're misleading more often than not.**

Sometimes I fix the comments, usually I just delete them.  A
sentance or two to describe what a function is to be used for
is about the only useful comment I have run into (the "doc"
strings for example).  Comments on individual variables and
lines of code are usually worse than useless.

In assembly comments are probably required, but in a language
like Python, if code needs comments other than the occasionaly
"doc" string, then it's written badly.

-- 
Grant Edwards                   grante             Yow!  Did YOU find a
                                  at               DIGITAL WATCH in YOUR box
                               visi.com            of VELVEETA?


From effbot at telia.com  Sat Apr  1 14:43:59 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Sat, 01 Apr 2000 19:43:59 GMT
Subject: Python 1.6 alpha 1 released
References: <200003312130.QAA04361@eric.cnri.reston.va.us> <38E5205F.DE811F61@roguewave.com> <Wf9F4.4423$74.70231@newsc.telia.net> <38E58D2D.384F1546@Lugoj.Com> <S9kF4.4472$74.71414@newsc.telia.net> <38E646B6.12011A4E@Lugoj.Com>
Message-ID: <3ksF4.4558$74.72866@newsc.telia.net>

James Logajan wrote:
> Any reason Guido doesn't post to this newsgroup like he used to?

well, what do *you* think is the reason?

> The pragmatist in me says I should look for a language that abides
> by a standard that can not be easily changed.

ahem.  GvR changed the CPython implementation to adhere
to the standard.  *you* want him to change the standard.

...

but sure, if you know of a language which is strictly specified,
never changes, and for which all implementations are truly strict
(that is, they shouldn't just accept all valid constructs, but also
flag all possible invalid constructs -- and be completely bug free,
of course), let us know.

</F>

<!-- Ythonpay 1.6 Eleaseray Eduleschay:
http://slashdot.org/articles/00/03/31/1734205.shtml
-->





From phd at phd.russ.ru  Thu Apr 13 05:07:16 2000
From: phd at phd.russ.ru (Oleg Broytmann)
Date: Thu, 13 Apr 2000 09:07:16 +0000 (GMT)
Subject: simple httplib and urllib timeout question
In-Reply-To: <8d404q$oee$1@nnrp1.deja.com>
Message-ID: <Pine.LNX.4.21.0004130904470.16470-100000@fep132.fep.ru>

On Thu, 13 Apr 2000, Georg Mischler wrote:
> What exactly is your problem with signal.alarm() ?

   Python delivers alarm only when in Python VM. If the script is blocked
in C (even worse - in system call), the signal will NOT be delivered.
   Of course, URL checker, while running through a big list of URLs,
certainly spend most of the time in connect() and recv() system calls. This
makes signals unusable.

Oleg.            (All opinions are mine and not of my employer)
---- 
    Oleg Broytmann      Foundation for Effective Policies      phd at phd.russ.ru
           Programmers don't die, they just GOSUB without RETURN.




From sara12 at hotmail.com  Mon Apr 17 14:52:47 2000
From: sara12 at hotmail.com (sara12 at hotmail.com)
Date: Mon, 17 Apr 2000 18:52:47 GMT
Subject: !!!Bluemountain -PRIVACY ALERT!!  6862
Message-ID: <34JK4.14054$K72.4855@typhoon2.gnilink.net>

Bluemountain says they protect your privacy, but they don't.  They use cookies to track everything you are doing on their site.  How do you think they remember your e-mail address when you come back to their site?  I found this great greeting card site that doesn't use cookies www.cardgirl.com
 Cardgirl and a bunch of her friends put together a really cool site.   It's a lot of fun and 100% free!  And they really protect our privacy.  And they have really good cards.  If you find any other good card sites that don't use cookies, please let me know
nxleoftyytzpgqlgqtrgbvsmkslldrfgwj



From ajung at sz-sb.de  Fri Apr 28 06:48:08 2000
From: ajung at sz-sb.de (Andreas Jung)
Date: Fri, 28 Apr 2000 12:48:08 +0200
Subject: How to use DCOracle and LOBs in Oracle 8i ????
Message-ID: <20000428124808.A21389@sz-sb.de>

After reading old postings concerning DCOracle and LOBs I
can not solve the following problem:

I have  a table under Oracle 8i with a NCLOB field. The base
character set for the complete database it UTF8. My first
problem with DCOracle was an ORA-12704 character set mismatch error when
I tried to insert an instance of DCOracle.dbi.dbiRaw() in to the database.


My current code looks like this (the table contains the NCLOB field 'content'):


lob = 'very long text .....'
raw = DCOracle.dbi.dbiRaw(lob)

com = 'insert into tab (docnum,content) values(:x,translate(:y using nchar_cs))'
curs.execute(com,x='1234',y=raw)


This works for LOBs with less than 4000 bytes. For longer LOBs I get the following error
message: ORA-01461 can bind a LONG value only for insert into a LONG column.

Now...how can I store larger LOBs with more than 4000 bytes ? I have read something
about a dbms_lob package but I have really no idea how to use it. 

Has anyone a working solution or a code example ?!


Thanks a lot for any hint,
Andreas 





From rcc at nospamthanks_jennic.com  Wed Apr 12 06:23:25 2000
From: rcc at nospamthanks_jennic.com (Robert Cragie)
Date: Wed, 12 Apr 2000 10:23:25 GMT
Subject: Printing a long
Message-ID: <newscache$tmfwsf$bz8$1@jenpc07.jennic.co.uk>

I can't seem to print a long. The following shows the inconsistencies:

print '%08x' % 0xffffffff
ffffffff
>>> print '%08x' % 0xffffffffL
Traceback (innermost last):
  File "<pyshell#1>", line 1, in ?
    print '%08x' % 0xffffffffL
OverflowError: long int too long to convert
>>> import string
>>> print '%08x' % string.atoi(hex(0xffffffffL)[:-1], 0)
ffffffff
>>> print '%08x' % string.atoi(hex(0xffffffffL)[2:-1], 16)
Traceback (innermost last):
  File "<pyshell#4>", line 1, in ?
    print '%08x' % string.atoi(hex(0xffffffffL)[2:-1], 16)
ValueError: atoi() literal too large: FFFFFFFF
>>>

I realise 0xffffffff is a (signed) int (i.e. -1), whereas 0xffffffffL is a
(signed) long int (i.e. +4294967295)

The problem originally arose from using struct.pack() and struct.unpack()
with 'unsigned' parameters - these are represented by Python longs, but C is
only really concerned about the number of bits, so I have used the signed
representation (which then uses Python ints); this solves my original
problem.

However, I'd be interested to see comments about these apparent
inconsistencies.

Robert Cragie




From quinn at auto.ugcs.caltech.edu  Wed Apr  5 15:08:41 2000
From: quinn at auto.ugcs.caltech.edu (Quinn Dunkan)
Date: 5 Apr 2000 19:08:41 GMT
Subject: Getting the argument names passed to a function ?
References: <slrn8el5t7.40k.quinn@riyal.ugcs.caltech.edu> <8ce6dm$4vp$1@news.wrc.xerox.com> <slrn8elpu4.kjk.quinn@zloty.ugcs.caltech.edu> <m3r9ckn58n.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <slrn8en3tu.l27.quinn@auto.ugcs.caltech.edu>

On 05 Apr 2000 12:42:48 +0100, Michael Hudson <mwh21 at cam.ac.uk> wrote:
>Have you downloaded bytecodehacks yet?

Yeah, I've looked at it.  Nice work.  But I fear if I actually used it I might
be tempted to do this sort of thing more often :)


From thehaas at binary.net  Mon Apr 24 10:01:02 2000
From: thehaas at binary.net (Mike Hostetler)
Date: Mon, 24 Apr 2000 09:01:02 -0500
Subject: compress_web.py
References: <Pine.LNX.4.21.0004241105310.6019-100000@fep132.fep.ru>
Message-ID: <3904539E.A5EA518F@binary.net>

Oleg Broytmann wrote:
> 
> On Fri, 21 Apr 2000, Mike Hostetler wrote:
> 
> >               for i in newfiles:
> >                       tar = tar+ " " + i
> 
>    This is just
> 
> tar = string.join(newfiles, " ")
> 

Well, I agree that it is simpiler and and faster, the above doesn't
quite work - you lose the present value of tar in the process.

But this works instead:

    tar = tar+ " "+ string.join(newfiles," ")

Thanks for the help.

- mikeh
-- 
Mike Hostetler          
thehaas at binary.net 
http://www.binary.net/thehaas 
GnuPG key: http://www.binary.net/thehaas/mikeh.gpg


From lewst at yahoo.com  Sun Apr  2 18:03:51 2000
From: lewst at yahoo.com (lewst)
Date: Sun, 2 Apr 2000 15:03:51 -0700 (PDT)
Subject: NameError when assigning dictionary values
Message-ID: <20000402220351.9903.qmail@web2103.mail.yahoo.com>

Gordon McMillan <gmcm at hypernet.com> wrote:

> >   # add each key-value pair to the dictionary and
> >   # then for each value, create an empty list.
> >   mydict = {}                         
> >   for key,value in [ (".", "one"), ("+", "two") ]:
> >       mydict[key] = value
> >       value = []
> >   print mydict
> >   print one,two
> > 
> > Whe I run this I get "NameError: one".  Why aren't my 2 empty lists
> > named "one" and "two" being created?
> 
> You're creating two empty lists, both named "value".The first 
> is thown away by creating the second. Nowhere are "one" and 
> "two" variable names, hence the NameError. 

Ok.  So now how do I get the program to do what I want?


__________________________________________________
Do You Yahoo!?
Talk to your friends online with Yahoo! Messenger.
http://im.yahoo.com



From tim_one at email.msn.com  Sun Apr  9 00:39:21 2000
From: tim_one at email.msn.com (Tim Peters)
Date: Sun, 9 Apr 2000 00:39:21 -0400
Subject: Python paradigms
In-Reply-To: <8cmr70$b0e$1@pegasus.csx.cam.ac.uk>
Message-ID: <000401bfa1dd$930fe500$172d153f@tim>

[Nick Maclaren]
> Well, starting to use Python, there are two Algol 68 constructions
> that I sorely miss.  Here is an example of them in C:
>
>     if ((x = non_trivial_expression) != NULL) {
>         do_something_using_x;
>     } elif ((x = another_expression) != NULL) {
>         do_something_else_using_x;
>     } ...
>
> and:
>
>     x = (a != NULL ? a[i]->weeble : 0) + (b != NULL ? b[i]->wombat : 0)
>
> It is pretty obvious why Python doesn't have the former, but I am
> a bit puzzled as to the reason for the lack of the latter.  Or
> maybe I just haven't found it yet :-)

It's just a FIR (Frequently Ignored Request <wink>).  3-argument getattr and
dict.get() and builtin max() and min() cover many common cases, and nobody
has screamed loud enough to overcome Guido's reluctance to add more syntax
for others, which *can* be done easily with obvious if/else sequences.  Note
that the C

    a() ? b() : c()

is equivalent (including evaluating only one of b and c) to the Python
atrocity

   (a() and [b()] or [c()])[0]

But you'll be shot if you ever use the latter, so forget I said that.

no-relief-in-sight-ly y'rs  - tim






From bdickenson at serviceware.com  Sat Apr  1 10:56:50 2000
From: bdickenson at serviceware.com (Bob Dickenson)
Date: Sat, 1 Apr 2000 10:56:50 -0500 
Subject: Python Threading and sys.exit() codes
Message-ID: <0532446FBA0ED31193BD00A0C92C53AB0134460A@billtcat.serviceware.com>

These are good suggestions.  A couple of questions/comments.

I thought about using the Queue model, but I was experimenting with MxStack
at the time.  It seemed to work fine, so...

I think I need the global lock to get proper thread sync because I have to
pop values off two stacks for each "job" - the file to be ftp'd and it's
time stamp on the source machine (for resetting it locally after the
ftp.retrbinary so that the next time the job runs it does valid date
compares.)  If I'm missing something with this logic, I'd appreciate a
clarification.  (I suppose I could compose a tuple and push/pop that, but
otherwise....)

I have read the thread doc and know that the sys.exit() in the FTPthread
class def is bogus on my part--part of the "prettifying" the code needs (mea
culpa).  

It's the sys.exit() (explicit or implied) from the __main__ thread after the
iterative thread.join of the ftp worker threads  finishes that seems to be
causing me problems. The process isn't "finishing" from the perspective of
the job scheduler.  The single-threaded version surfaces both an implied or
explicit exit code to the OS, the multi-threaded version does not surface
anything.  The SMTP message gets sent (last item before the sys.exit() in
__main__) AND the command shell window closes, but something seems to be at
loose ends on the process termination cleanup side.

(BTW - the single threaded version of this program had about 60 files per
minute throughput on the ftp; with 12 worker threads I'm getting between
800-1000 files per minute.  The flux rate on the target site is several
thousand files per day, so it's really worth it to use the multi-threaded
version)

Bob
"For a list of the ways which technology has failed to improve our
quality of life, press 3"



-----Original Message-----
From: David Fisher [mailto:python at rose164.wuh.wustl.edu]
Sent: Friday, March 31, 2000 11:52 PM
To: dickenson at serviceware.com
Cc: python-list at python.org
Subject: Re: Python Threading and sys.exit() codes


----- Original Message -----
From: <dickenson at serviceware.com>
Newsgroups: comp.lang.python
To: <python-list at python.org>
Sent: Friday, March 31, 2000 1:26 PM
Subject: Python Threading and sys.exit() codes

> My problem is this: I want to run this program as a scheduled task
> under a job scheduler (currently testing cronDsys from #ifdef). The
> single-threaded version of the program posts an appropriate code to the
> environment via sys.exit(). The job scheduler traps this and takes
> conditional action. So far, so good.
> In the multi-threaded version of the program, the exit codes are not
> getting posted and the main process is not terminating cleanly, at
> least from the perspective of the job scheduler -- it appears to run
> until manually terminated from the scheduler control console although
> the NT command window disappears from the target desktop as expected.

Well for one thing, sys.exit() doesn't do what you think it does.
Sys.exit() raises an exception SystemExit which if not caught will cause the
thread to exit silently.  Read the docs for thread which threading is built
on.  I don't know of a way to catch an exception that another thread raises.
Doesn't mean there isn't one, just that I don't know one.

Another thing you look at is the Queue module.  That's what got my attention
because you are setting a global lock and poping stuff off of a fifo stack,
and Queue does all that for you. Here's some code:

import Queue
import thread    #yes, I know I'm supposed to use threading

def worker(jobqueue):
    while 1:
        job = jobqueue.get()    # this blocks until a job is available
        print job    # or whatever

def main(joblist):
    jq = Queue.Queue(0)
    for i in range(10):
        thread.start_new_thread(worker,(jq,))
    for job in joblist:
        jq.put(job)

There are lots of things that could be prettified in this code, but I hope
it gives you some ideas.  You might try a second Queue to gather return
information.

Good luck,
David



From John.Grayson at GD-CS.COM  Wed Apr 12 10:13:40 2000
From: John.Grayson at GD-CS.COM (Grayson, John)
Date: Wed, 12 Apr 2000 10:13:40 -0400
Subject: Widget for plotting graphs?
Message-ID: <3774EF539472D211B98F0008C7F468A1033B6812@TNTNEX01.tntn.gtegsc.com>

>> Bruce Wolk wrote:
>>
>> I am exploring Tkinter and Pmw and wondering whether anyone has created
>> a widget for graphing points and functions. Also, is there a way to
>> place one pixel (i.e., a point) on a Canvas widget? Thanks.
>>

Take a look at http://www.manning.com/Grayson/Chapters.html

In the sample chapter 11, you'll find a graph widget derived from
one written by Konrad Hinsen. I've used this for quite a number
of applications... It does line plots, bargraphs and pie charts.

    John Grayson



From jsolbrig at my-deja.com  Fri Apr 28 20:58:36 2000
From: jsolbrig at my-deja.com (jsolbrig at my-deja.com)
Date: Sat, 29 Apr 2000 00:58:36 GMT
Subject: Do I always have to write "self." ?
References: <270420001706386648%pecora@anvil.nrl.navy.mil>
Message-ID: <8edc3l$7sj$1@nnrp1.deja.com>

Actually,
What I'd like to see would for "self" to an official keyword rather than
simply a convention. The more things that are "conventional" but not
required, the more potential there is for occasional bugs.
And occasional bugs are the WORST.

Of course, a real programmer would never make a mistake like that -
that's why all real programmers are still using assembler. But for the
rest of us small-brained folks, more discipline is good.

jsolbrig at webcom.com


In article <270420001706386648%pecora at anvil.nrl.navy.mil>,
  "Louis M. Pecora" <pecora at anvil.nrl.navy.mil> wrote:
> I am two weeks plus into learning Python and have started on a first
> real project.  Amazing language.  BUT, now that I've gotten going with
> classes, I'm finding I hate seeing all those  "self."  prefixes on the
> class variables.  Python is generally very clean, but the classes I
> write use a lot of "per-instance" variables, hence seem to require
> self.  As a result the expressions can become cluttered and harder to
> read than C++ in this particular case -- imagine that!  Is there any
> relief here?  Any (simple) way around writing that self. every time?
>
> Thanks.  Love Python so far otherwise.
>


Sent via Deja.com http://www.deja.com/
Before you buy.


From walter at data.franken.de  Fri Apr 28 03:56:27 2000
From: walter at data.franken.de (Walter Doerwald)
Date: Fri, 28 Apr 2000 09:56:27 +0200
Subject: [Python-Dev] Re: Python 1.6a2 Unicode bug (was Re: comparing strings and ints)
References: <l03102805b52ca7830b18@[193.78.237.154]> <39072F0C.5214E339@lemburg.com> <BlHN4.1719$wYl.195075584@newsb.telia.net> <oqvh14fikb.fsf@titan.progiciels-bpi.ca> <45919986@data.franken.de> <048501bfb0b4$c793d760$4add3fcb@neil>
Message-ID: <45919991@data.franken.de>

On Fri, 28 Apr 2000 11:55:05 +1000 Neil Hodgson wrote:

> > Would it be possible to write the source code in UTF-16 (which is much
> > easier to to under WinNT with NotePad, I know no text editor under that
> > writes UTF-8)
>
>    The W2K version of Notepad writes UTF-8 as well as UCS-2. I don't know
> of
> any editor that understands UTF-16.

Of course I meant UCS-2, but I'm confused now. Isn't UCS-2 and UTF-16 more
or less the same?

> Idle writes UTF-8. Beta versions of
> Scintilla/SciTE and Pythonwin 131 can write UTF-8.

But I find UCS-2 much cleaner, because it's easily distinguished from
ASCII/UTF-8/ISO8859-1/..., distinguishing between UTF-8 and ISO8859-1 has
to be done by "external" methods. How does Python do it?

Bye...
	Walter

--
Walter D?rwald ? walter at data.franken.de ? Kommunikationnetz Franken e.V.



From mjhand at concentric.net  Fri Apr 14 11:53:04 2000
From: mjhand at concentric.net (Manus Hand)
Date: Fri, 14 Apr 2000 09:53:04 -0600
Subject: Questions on 1.6a2's string methods
References: <38F6C904.76B538E5@concentric.net> <m33dopgkgv.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <38F73EE0.35046EC8@concentric.net>

> I think weve had all these discussions recently.
>
Thanks.  Sorry to have wasted bandwidth.

As to "".capwords(),  I hope someone gets around to it.

And as to [].join(), it sure makes sense to me, but if it's been decided otherwise, I'll pull up the thread and see
why.

Manus (who obviously hasn't read the ng for a while, and who's so used to typing __dict__.keys() that he totally
spaced on dir())




From mhammond at skippinet.com.au  Thu Apr 20 09:38:11 2000
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Thu, 20 Apr 2000 13:38:11 GMT
Subject: Pythonwin?
References: <eMz2Nmiq$GA.73@net025s> <4QtL4.24$v85.41@news-server.bigpond.net.au> <chuL4.8483$O7.17315@typhoon.southeast.rr.com>
Message-ID: <7LDL4.534$v85.2563@news-server.bigpond.net.au>

"Don Tuttle" <tuttledon at hotmail.com> wrote in message
news:chuL4.8483$O7.17315 at typhoon.southeast.rr.com...
> [Mark]
> > You are really going to need to go through an MFC tutorial or book
before
> > you can make heads or tails of Pythonwin from that level...
>
> Any recommendations?

Afraid not.  I already new MFC when I started Pythonwin.  I learnt it from
the MFC "scribble" tutorial.  I find I tend to learn best with a tutorial
and then playing - that is what I did with Python (back then there was no
other alternative, but even now that is my recommended strategy for
competant existing programmers learning Python...)

So assuming some C++ knowledge, my suggestion for MFC would still be
scribble tutorial, then invent an app to write in MFC C++, then port that
app to Python.  By then you will be a wiz :-)

Mark.





From greg at cosc.canterbury.ac.nz  Wed Apr  5 01:01:56 2000
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Wed, 05 Apr 2000 17:01:56 +1200
Subject: Why should I switch to Python? - Infinity of Primes
References: <NDBBKEGCNONMNKGDINPFGEJDDFAA.infonuovo@email.com> <8ccoug$f4o$1@pegasus.csx.cam.ac.uk> <7F51E902E09F3F60.8F612607812ED16B.82CB8B41435E14FD@lp.airnews.net> <38EA369C.26A1C25@math.okstate.edu>
Message-ID: <38EAC8C4.ADC25640@cosc.canterbury.ac.nz>

"David C. Ullrich" wrote:
> 
>     There's nothing non-constructive about the traditional
> proof of the infinitude of the sequence of primes - given
> a sequence of primes it _constructs_ a prime not on the
> list.

Um, no it doesn't - it constructs a number which is
*either* prime *or* divisible by some other prime bigger
than the one you started with.

If someone actually came up with a formula for constructing
primes, it would be rather large news -- isn't that one of
the Big Unsolved Problems?

-- 
Greg Ewing, Computer Science Dept,
+--------------------------------------+
University of Canterbury,	   | A citizen of NewZealandCorp, a	  |
Christchurch, New Zealand	   | wholly-owned subsidiary of USA Inc.  |
greg at cosc.canterbury.ac.nz	   +--------------------------------------+


From stevee at redac.co.uk  Mon Apr 10 09:13:14 2000
From: stevee at redac.co.uk (stevee at redac.co.uk)
Date: Mon, 10 Apr 2000 13:13:14 GMT
Subject: Cascading menus
References: <8cko46$j2i$1@nnrp1.deja.com> <38F0DF82.EEAB208D@dkfz-heidelberg.de>
Message-ID: <8csk0v$r44$1@nnrp1.deja.com>

In article <38F0DF82.EEAB208D at dkfz-heidelberg.de>,
  Kilian Lorenz <kilian.lorenz at dkfz-heidelberg.de> wrote:
>
>
> Perhaps you can solve your problem by using the postcommand option.
>
> When the menu is posted this function/method is called and
> probably you can change your sub menu there. I don't expect that
anyone
> estimates code to be run anytime a menu item is selected a very clever
> idea since with an indecisive user this could happen quite frequently
> and I can't think of a good reason why there should be done something
> else in the second call than in the first call (except perhaps making
a
> statistic about menu usage).
>
> But I don't speak for the whole Python/Tkinter/Tk-community and
> who knows ... nobody expects the Spanish Inquisition either ...
>
> Regards, Kilian
>
Using the postcommand hook is another option I am investigating.
However it is looking like a lot of work to get that to function
correctly. For example I might have existing code called by the parent
menu that looks like this:

    function foo
       set some variables
       check a feature is  authorized
       display sub menu
       do some more stuff

To automatically translate this to Python I need to extract the code
before the sub menu display into another function that can be called
by the postcommand hook. Unfortunately there are a lot of cases that
are more complex than this, with several levels of functions before
the sub menu display:

    function foo
       set some variables
       call function bar
       do some more stuff

    function bar
       check a feature is  authorized
       display sub menu

And so on.

In addition in some cases the code actually decides between several
different sub menus at run time:

    function foo
       set some variables
       check a feature is  authorized
       if feature 1 then
           display sub menu for feature 1
       else
           display sub menu for feature 2
       do some more stuff

This would require me to completely redefine the sub menu on the fly!

However you are fundamentally right this is not a clever idea. Thus
the aim is that as we redesign the interface the auto generated code
will be dropped to we end up doing things correctly. Unfortunately the
effort estimate to do the complete UI redesign in one go is so high
that it will not happen, which is why we are trying to change the
current interface to use Python without radically changing the
interface itself. This then allows to gradually move the UI to where
we want it.

But thanks for your reply.

Steve


Sent via Deja.com http://www.deja.com/
Before you buy.


From rurban at x-ray.at  Sun Apr 16 08:37:37 2000
From: rurban at x-ray.at (Reini Urban)
Date: Sun, 16 Apr 2000 12:37:37 GMT
Subject: Future compatibility of the Python C API
References: <lwEJ4.7980$Ib7.101092@typhoon2.kc.rr.com>
Message-ID: <38f9b48f.8998058@judy>

Keith Davidson wrote:
>1) Has there been any discussion about the upwards compatibility of the C
>API in future versions of Python?  Historically our libraries have had a
>7-10+ year lifespan where we provide 95+% backwards compatibility (i.e. we
>don't want our C API users to have to make more than a day or so worth of
>changes between releases).  My big concern is Python 3000- what are the odds
>that its C API will be backwards compatibility with Python 1.x?

Isn't it a waste to even think of a API in C for the next 1000 years?
I would rather be concerned about changing OS and hardware issues. The
API is dependent on that. 
Tried to build something for 64bit? Win64, IA64? C is horror on most
complicated issues.
--
Reini Urban
http://xarch.tu-graz.ac.at/autocad/news/faq/autolisp.html


From rmilner at point-one.net  Thu Apr 20 12:10:43 2000
From: rmilner at point-one.net (Roman Milner)
Date: 20 Apr 2000 11:10:43 -0500
Subject: does any have the python/gtk tutorial? can't reach http://laguna.fmedic.unam.mx/~daniel/pygtutorial/
Message-ID: <A35FD62F631DA692.FC074B8626E0F614.856E75DC531CEC66@lp.airnews.net>

I haven't been able to get through to the site with the python gtk
tutorial for the last couple of days.  Does anyone have a copy they
can share?

Thanks,
^Roman


From gduzan at gte.com  Fri Apr 14 09:58:51 2000
From: gduzan at gte.com (Gary D. Duzan)
Date: Fri, 14 Apr 2000 09:58:51 -0400
Subject: 1st non-trivial program - gentle criticism requested
References: <1e90r16.1sd2ts41e7ur6kN@gershwin.bio.ic.ac.uk> <8d4tgu$11qa$1@nntp6.u.washington.edu> <38F6193A.66284778@gte.com> <t3pJ4.913$rc9.192648192@newsb.telia.net>
Message-ID: <38F7241B.9672E7C2@gte.com>

Fredrik Lundh wrote:
> 
> Gary D. Duzan <gduzan at gte.com> wrote:
> > This potential stumbling block makes me wonder if allowing the
> > singleton case was really wise.
> 
> stumbling block?
> 
> what kind of programmer changes his code without thinking,
> reading what he just typed, or running the resulting program?
> 
> if you want to optimize python for that kind of programmer,
> you have to change a whole lot more than the string formatting
> operator...
> 
> </F>

   My only point would be that if the formatting operator only took
tuples (oh, and dictionaries, of course), everyone would already be
in the habit of using a "(x,)" notation, so such changes wouldn't be
a problem. It is always nice to reduce the number of code points
impacted by a given change. I think it is arguably an easier rule
for newbies, as well, who may be thinking that the formatting string
is matching the variable, not its expansion. After all, "'%s' % x"
works fine when "x" is a string, number, list, dictionary, class,
instance, module, etc., but not if it is a tuple. Encouraging a
style that only mostly works isn't really a good thing and could be
considered a violation of the Principle of Least Astonishment.
   Not that I'm proposing a change, mind you; it is a bit late for
that. I'm just noting it as an issue, and suggesting that it may be
worth considering to discourage the use of the singleton version in
the future. (Of course, I'm being a hypocrite here since I use the
singleton version all the time, but I like to put my idealist/language
purist hat on every once in a while.)

					Gary Duzan
					GTE Laboratories


From python-list at teleo.net  Thu Apr 20 19:24:28 2000
From: python-list at teleo.net (Patrick Phalen)
Date: Thu, 20 Apr 2000 16:24:28 -0700
Subject: Using python on the web
In-Reply-To: <8do2te$kpr$1@nnrp1.deja.com>
References: <MPG.1365adcfa8470eef9896aa@news> <9N4L4.499$wYl.191816704@newsb.telia.net> <8do2te$kpr$1@nnrp1.deja.com>
Message-ID: <00042017440707.02559@quadra.teleo.net>

[spenrose at my-deja.com, on Thu, 20 Apr 2000]
:: We build medium-busy, CGI-intensive sites in Python. We were briefly
:: quite excited about Zope, but once we realized it locked us into Yet
:: Another Programming Language, we lost interest.

The only programming language Zope "locks you into" is Python.

:: DHTML is particularly problematic,
:: AFAI can tell, because it doesn't appear to allow the
:: segregation of HTML development from coding the way Python string
:: substitution does.

How did DHTML enter into this discussion?

Oh -- perhaps you mean DTML? If so, Zope's DTML is best viewed as a
reporting language, certainly not a programming language, although it
has support for variable substitution and conditionals, etc. And it's
aim IS precisely to separate HTML from coding. In fact it's
considerably more powerful in that role than string substitution. Sounds
like you didn't look at it very closely.

For heavy lifting in Zope, or to extend it, you use Python. Having the
combination of DTML and Python is particularly flexible and powerful.

:: We hire very skilled HTML people to do HTML and tell
:: them to leave our sustitution tags alone, rather than taking the time
:: and money to create novice DHTML people to do the same work. If we were
:: going to learn another language for production purposes, Perl, Java, and
:: C++ all have vastly larger infrastructures (code bases, labor pools,
:: documentation, etc.) than DHTML ever will. There are enough hassles in
:: using Python, which I love, because of its relatively small installed
:: base. No way am I going to something which is an order of magnitude less
:: supported.

Zope is an application server and content management framework. On
the one hand, as an application server framework, it seeks to provide a
superior alternative to CGI. It is object oriented from the ground up.
People deploy it because it provides technology to publish Python
objects over the Web, a powerful concept.

On the other hand, as a content management solution, Zope shines in
situations where you need to grant control to web producers in a
fine-grained, hierarchical, and secure way. These producers can use
HTML. They can also, if you like, learn a smattering of DTML, but that
isn't necessary. You can set things up so that they don't even need to
know how to *spell* DTML and they can be very productive -- in HTML.
You, as a developer, can continue to be productive -- in Python.




From hellan at acm.org  Mon Apr 17 14:36:21 2000
From: hellan at acm.org (Jon K Hellan)
Date: 17 Apr 2000 20:36:21 +0200
Subject: Scripting and Gnome and KDE
References: <Pine.GSO.4.10.10004150741520.11261-100000@sundial> <8d95m0$9j6$1@news1.xs4all.nl> <8damat$4ja$1@slb2.atl.mindspring.net> <8dbsgc$35o$1@news1.xs4all.nl> <38FB4336.2BC6C235@bioreason.com>
Message-ID: <87bt38h8wq.fsf@parus.parus.no>

Mitch Chapman <chapman at bioreason.com> writes:

> That's exactly right.  It's fairly easy to get started with pygtk,
> but some things about the toolkit are difficult to discover.  We wanted
> to show how to do the difficult (i.e. obscure) things, in order to 
> make them easy.  The article introduction should have made that more
> clear.

You should definitely have mentioned libglade, which in most cases
makes it unnecessary to code the Gtk GUI at all. You still have to do
the callbacks, of course.

Regards

Jon K?re Hellan


From effbot at telia.com  Mon Apr 24 20:27:07 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Tue, 25 Apr 2000 00:27:07 GMT
Subject: Tkinter
References: <A63AFB20111ED311AC5500A0C96B7BF5015F0F44@orsmsx36.jf.intel.com>
Message-ID: <vD5N4.1521$wYl.189935104@newsb.telia.net>

Daley, MarkX <markx.daley at intel.com> wrote:
> I am trying to use Tkinter for a GUI interface, but I am having a hard
time
> finding manuals for it.  Is there documentation comparable to Python's?

http://www.python.org/topics/tkinter/doc.html

</F>




From jimserac at mindspring.com  Sun Apr  2 11:42:19 2000
From: jimserac at mindspring.com (James Pannozzi)
Date: Sun, 2 Apr 2000 11:42:19 -0400
Subject: New Features in Python 1.6
References: <200004011740.MAA04675@eric.cnri.reston.va.us> <8c6mkg$8i$1@slb2.atl.mindspring.net> <0a509c7e.1fbf25d6@usw-ex0101-006.remarq.com>
Message-ID: <8c7poj$pv2$1@slb0.atl.mindspring.net>

Nah, the only jokes I've noticed are Javascript and VBScript!

Thanks
Jim

"Hamish Lawson" <hamish_lawson at yahoo.co.uk> wrote in message
news:0a509c7e.1fbf25d6 at usw-ex0101-006.remarq.com...
> As I read Guido's announcement an increasing sense of unease
> came over me. Then I saw the posting date. Ah, it must be an
> April Fool joke. No, really ... It MUST be an April Fool Joke!
>
> However (assuming it was a joke) some people seem to have taken
> the bait and are quite concerned. Might it be a good idea now to
> admit that it was a joke before more Fear, Uncertainty and Doubt
> is spread?
>
> Hamish Lawson
>
>
> * Sent from RemarQ http://www.remarq.com The Internet's Discussion Network
*
> The fastest and easiest way to search and participate in Usenet - Free!
>




From breiter at usf.Uni-Osnabrueck.DE  Thu Apr 13 06:20:54 2000
From: breiter at usf.Uni-Osnabrueck.DE (Bernhard Reiter)
Date: 13 Apr 2000 10:20:54 GMT
Subject: Compiling Python Imaging under LinuxPPC
References: <FU%I4.646$rc9.190630912@newsb.telia.net>
    <20000412144842.06030.00002524@ng-ch1.aol.com>
Message-ID: <8d4726$a7n$1@newsserver.rrzn.uni-hannover.de>

[Posted and mailed]

William,
sorry for answering so late. But I have build PIL on LinuxPPC
(LinuxPPC2000 on a Powerbook Lombarde) without major problems.

I downloaded Imaging-1.0 as tarball and just followed the instructions.
The only thing I had to do was to change the Tcl/Tk support and
library location line in "Setup" to include the position of the X11
libraries. ( I just added "-L/usr/X11R6/lib -lX11" )

The rest was just out of the book. :)=
I would suggest that you try building it from the Imaging-1.0.tar.gz.

BTW: I got sketch 0.6.5 running fine here.
	Bernhard

In article <20000412144842.06030.00002524 at ng-ch1.aol.com>,
	willadams at aol.com (William F. Adams) writes:
> When I attempt to compile the Sketch .src.rpm, the compilation fails because
> Imaging.h is not where it's supposed to be.
> 
> I ddin't see much in the way of documentation with the .src.rpm for PIL, sorry.

-- 
Professional Service around Free Software                (intevation.net)  
The FreeGIS Project				            (freegis.org)
Association for a Free Informational Infrastructure            (ffii.org)


From tjreedy at udel.edu  Tue Apr  4 14:51:09 2000
From: tjreedy at udel.edu (Terry Reedy)
Date: Tue, 4 Apr 2000 14:51:09 -0400
Subject: Python 1.6 alpha 1 released
References: <200003312130.QAA04361@eric.cnri.reston.va.us> <38E5205F.DE811F61@roguewave.com> <Wf9F4.4423$74.70231@newsc.telia.net> <38E52716.A1C63B5D@roguewave.com> <38E88E5F.324CFB99@infercor.no>
Message-ID: <8cdd6k$4el$1@news.udel.edu>

"Paul Boddie" <paulb at infercor.no> wrote in message
news:38E88E5F.324CFB99 at infercor.no...

> Socket programming, especially for anyone who tried it in C, can be
daunting to
> those new to the subject and it is only natural that people will make the
most
> out of the examples provided in order to get started. Given that the
example in
> the library reference (even as recently as the one provided for 1.5.2)

and, as I pointed out elsewhere, all the examples I could find in the two
'classic' books
by four of our gurus, including Guido himself,

> contradicts the supposedly correct way of calling 'connect', we can
seemingly
> infer that either the intention was always to support this usage of the
method
> in question, or that the documentation is not to be relied upon






From michael.stroeder at inka.de  Thu Apr 20 15:20:42 2000
From: michael.stroeder at inka.de (Michael =?iso-8859-1?Q?Str=F6der?=)
Date: Thu, 20 Apr 2000 21:20:42 +0200
Subject: socket.SO_REUSEADDR
References: <38FEEA44.6C05E2B0@inka.de> <8dmsrd$8ta$1@nnrp1.deja.com>
Message-ID: <38FF588A.4760CA34@inka.de>

fredrik at pythonware.com wrote:
> 
> Michael Str?der wrote:
> > For avoiding problems with errors like
> >
> >   socket.error: (98, 'Address already in use.')
> 
> that getproto stuff looks strange.  try this instead:
> 
>    self.socket.setsockopt(socket.SOL_SOCKET, socket.SO_REUSEADDR, 1)

Three times the same answer. Must be right. ;-)

Thanks to all.

Ciao, Michael.


From aahz at netcom.com  Sun Apr  2 12:45:33 2000
From: aahz at netcom.com (Aahz Maruch)
Date: 2 Apr 2000 16:45:33 GMT
Subject: Python 1.6 alpha 1 released
References: <200003312130.QAA04361@eric.cnri.reston.va.us> <38E52716.A1C63B5D@roguewave.com> <rJ9F4.549$n68.187227648@newsb.telia.net> <38E52D76.1820FC32@roguewave.com>
Message-ID: <8c7tfd$k80$1@nntp9.atl.mindspring.net>

In article <38E52D76.1820FC32 at roguewave.com>,
Bjorn Pettersen  <bjorn at roguewave.com> wrote:
>
>I'm glad your schedules provide enough time for this, mine doesn't.  In
>addition, Python is starting to gain acceptance here, and having
>"mission critical" parts of the development environment break would be a
>"bad thing" (especially if I happened to be the one who implemented
>them...)

I'm sympathetic.  We only upgraded our production system from 1.5.1 to
1.5.2 last week, because we'd made a number of 1.5.1-specific hacks and
had to back them out correctly for 1.5.2.  Now we'll have to do the same
for 1.6, but as with the 1.5.2 upgrade, we don't have to follow Guido's
timeline.
--
                      --- Aahz (Copyright 2000 by aahz at netcom.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

"It's 106 miles to Chicago.  We have a full tank of gas, a half-pack of
cigarettes, it's dark, and we're wearing sunglasses."  "Hit it."


From lexberezhny at email.msn.com  Mon Apr 24 00:08:49 2000
From: lexberezhny at email.msn.com (lexberezhny)
Date: Mon, 24 Apr 2000 00:08:49 -0400
Subject: Sorting a list of classes
Message-ID: <OmnqwHar$GA.233@cpmsnbbsa03>

If i have a python list of classes, can i do something like classList.sort()
and how can i control this behavior within the classes. ie
i have a address book, and a have a class called 'Person', and it has an
attribute Person.Name, so i have a list of 'Person's how can i sort the
list, and maybe sort it by name? do i have to write my own function or can i
use list.sort() some how?

 thanks,
  lex




From hgg9140 at george.org  Sat Apr  1 18:20:41 2000
From: hgg9140 at george.org (Harry George)
Date: 01 Apr 2000 15:20:41 -0800
Subject: Ann: cgipm, pyperl, mkpythonproj, pdx
Message-ID: <m3aejd9zna.fsf@wilma.george.org>

Finally got the pieces in place (I think -- neither mirror nor
sitecopy seems to work completely on my ISP):

  http://www.seanet.com/~hgg9140/

Also converted all the gif stuff to png.  So pdx's texmath requires
pstogif and then convert from gif to png.


-- 
Harry George
hgg9140 at seanet.com


From jeremy at cnri.reston.va.us  Tue Apr  4 17:55:47 2000
From: jeremy at cnri.reston.va.us (jeremy at cnri.reston.va.us)
Date: Tue, 04 Apr 2000 21:55:47 GMT
Subject: bytecodehacks-April2000
References: <m3wvmhopqg.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <8cdocs$tnd$1@nnrp1.deja.com>

In article <m3wvmhopqg.fsf at atrus.jesus.cam.ac.uk>,
  Michael Hudson <mwh21 at cam.ac.uk> wrote:

> PS: I haven't sent this to python-announce, as that list appears to be
> entirely dead.  News to the contrary will be gratefully accepted.

I have also had problems over the last few months getting anything
through the python-announce moderators in a timely fashion.  It sounds
like it is time to start agitating for something to be done.

Jeremy

PS Sorry I don't have anything to say about your very cool hacks.


Sent via Deja.com http://www.deja.com/
Before you buy.


From tismer at tismer.com  Sun Apr  2 10:58:18 2000
From: tismer at tismer.com (Christian Tismer)
Date: Sun, 02 Apr 2000 16:58:18 +0200
Subject: New Features in Python 1.6
References: <200004011740.MAA04675@eric.cnri.reston.va.us> <8c6mkg$8i$1@slb2.atl.mindspring.net> <0a509c7e.1fbf25d6@usw-ex0101-006.remarq.com>
Message-ID: <38E7600A.F3067AF8@tismer.com>


Hamish Lawson wrote:
> 
> As I read Guido's announcement an increasing sense of unease
> came over me. Then I saw the posting date. Ah, it must be an
> April Fool joke. No, really ... It MUST be an April Fool Joke!
> 
> However (assuming it was a joke) some people seem to have taken
> the bait and are quite concerned. Might it be a good idea now to
> admit that it was a joke before more Fear, Uncertainty and Doubt
> is spread?

People always get what they deserve :-)

PythonWin 1.5.42+ (#0, Mar 29 2000, 20:23:26) [MSC 32 bit (Intel)] on win32
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
Portions Copyright 1994-2000 Mark Hammond (MHammond at skippinet.com.au)
>>> " ".split("ha ha")
[' ']
>>> "ha ha".split(" ")
['ha', 'ha']
>>> 

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaunstr. 26                  :    *Starship* http://starship.python.net
14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     we're tired of banana software - shipped green, ripens at home



From spex66 at my-deja.com  Thu Apr 27 05:12:07 2000
From: spex66 at my-deja.com (spex66 at my-deja.com)
Date: Thu, 27 Apr 2000 09:12:07 GMT
Subject: How can I get the name of an object???
References: <8e712k$65c$1@nnrp1.deja.com> <etd8zy03m44.fsf@w20-575-108.mit.edu>
Message-ID: <8e9090$c4d$1@nnrp1.deja.com>

In article <etd8zy03m44.fsf at w20-575-108.mit.edu>,
  Alex <cut_me_out at hotmail.com> wrote:
>
> An object could be referred to by any number of names:
>
> dd = zz ()
> ee = dd
>
> Which name for the object do you wish to know?  ee or dd?
dd or in other words, the name of the FIRST made instance... I tried to
expand the zz.__init__(self) to save the name of the generated
instance, but I think that the name must given as an extra parameter to
__init__(self, instanceName)

I hoped that there is a way to get it from the context :-|

thanks for all the other answers in the thread!

Peter
(=PA=)

> Alex.
>


Sent via Deja.com http://www.deja.com/
Before you buy.


From quinn at auto.ugcs.caltech.edu  Wed Apr  5 16:00:33 2000
From: quinn at auto.ugcs.caltech.edu (Quinn Dunkan)
Date: 5 Apr 2000 20:00:33 GMT
Subject: Pass by reference ?
References: <38E864C9.E1EE94F7@ecs.soton.ac.uk> <38EAF8D8.4B7C43B4@ecs.soton.ac.uk> <38EB56AE.F938CE06@acm.org> <m33dp0muk9.fsf@atrus.jesus.cam.ac.uk> <38EB63BC.FE03BA31@ecs.soton.ac.uk> <m3ya6sle0j.fsf@atrus.jesus.cam.ac.uk> <38EB6EB1.9759397C@ecs.soton.ac.uk> <38EB838A.53122AF6@acm.org>
Message-ID: <slrn8en6v5.l27.quinn@auto.ugcs.caltech.edu>

On Wed, 05 Apr 2000 18:21:36 GMT, Robert W. Cunningham <rcunning at acm.org> wrote:
>Jacek Generowicz wrote:
>
>> Michael Hudson wrote:
>>
>> > The code you posted only confuses if you have a flawed understanding of
>> > assignment in Python
>>
>> That is exactly the point. My original confusion arose precisely because I

>The CS models of "Pass By Reference" and "Pass By Value" are fairly well
>understood, but they do not seem to map simply and directly to Python.
>Rather than say "Give up on theory, just learn Python", I'd rather EXTEND (or
>clarify) the theory to account for Python!  Then, I'm certain, I will learn
>Python better and faster.

Hmm, that's funny, because under my understanding of "pass by reference",
python fits it exactly.  And all languages I use except C acts like python
(but C does too if you pass by pointer).  I was initially a bit confused
because most languages I use don't have mutation, but it cleared up pretty
fast.

python:
def pass_by_ref(lst):
    lst[0] = 'baz'
    return lst

def pass_by_ref2(a):
    a = a + 1
    return a

x = []
print pass_by_ref(x)
x = 5
print pass_by_ref2(x)
print x

pseudo-C:
char **pass_by_ref(char **lst)
{
    lst[0] = "baz";
    return lst;
}
char *x[1];
printf("%s\n", *pass_by_ref(x)[0]);

int *pass_by_ref2(int *a)
{
    int a[1];
    *a = *a + 5;
    return a;
}
int x[1] = { 5 };
printf("%d\n", *pass_by_ref2(x));
printf("%d\n", *x);

perl:
sub pass_by_ref {
    my $lst = shift;
    $$lst[0] = "baz";
    return $lst;
}
$x = [];
print @{pass_by_ref($x)}, "\n";

sub pass_by_ref2 {
    my $a = shift;
    $a = \($$a + 1);
    return $a;
}
$x = \5;
print ${pass_by_ref2($x)}, "\n";
print $$x, "\n";  # gah, it's worse than C, how does anyone put up with perl?

As you noticed, the important thing is to realize how python's assignment
differs from C's, and that's simply because python doesn't have declarations,
so it assumes when you say 'foo = bar' you are binding to a *local* name,
unless you explicitly say 'global foo' (python also doesn't have C's confusing
magic pointer decay, but that's different).  How I see it, it has *nothing* to
do with references and everything to do with declarations vs. no declarations.
I think any explanation of the concept to confused C people should focus on
that.  And if you really understand it, you'll know why

a = 5
def kaboom():
    a = a + 1

throws a NameError, which is such a common confusion they changed it to
UnboundLocalError or something like that in 1.6.  It's the same reason:

int a[1] = { 5 };
int *kaboom(void)
{
    int a[1];
    *a = *a + 1;
    return a
}

won't give you what you want (at least I assume it won't, I'm too lazy to
see).

>There is not the "World Of All Other Computing Languages" and "Python" off by
>itself.

Exactly.  Python's behaviour is very standard.  The norm, rather than the
rule, IMHO.


From m.faassen at vet.uu.nl  Mon Apr 10 18:05:16 2000
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 10 Apr 2000 22:05:16 GMT
Subject: Tuples -- who needs 'em
References: <AF223B12F1FDD2119253006097C4FCEC0AD357@cats.pdri.xerox.com> <8chol6$6em$1@newshost.accu.uu.nl> <026e01bf9fe7$9554fd20$74eb0b18@stcla1.sfba.home.com> <8cofcf$hns$1@newshost.accu.uu.nl> <007801bfa30e$ffe79560$74eb0b18@stcla1.sfba.home.com>
Message-ID: <8ctj6s$bko$1@newshost.accu.uu.nl>

Bob Alexander <bobalex at home.com> wrote:
> "Martijn Faassen" <m.faassen at vet.uu.nl> wrote:

> Thanks for your well-considered comments, but we still have our differences
> :-)

I didn't expect otherwise. :)

>> the idea of immutable
>> objects is that assigning them is exactly the same behavior as copying
>> them. That's something an easy thing to think about; for instance with
>> strings and numbers. The idea here is that tuples allow you to think that
>> way about lists.

> Well, immutable objects render the copy vs. reference semantics moot, since
> you can't change them anyway. With true copy semantics, the receiver of the
> copy can change the object without affecting the originator's object.

That's exactly what I meant. Since the difference is moot when you have
immutability, you lose all the subtleties involved with reference
semantics versus value semantics if you use immutable objects. Which
can make things easier to think about.

> But I
> agree that you can safely pass around references to immutable objects
> without fear of their being modified -- I guess that's the point of
> immutable.

Right.

>> Okay, but it's generally practiced in idiom. Language constructions can
>> exist that partially support idiom, but not enforce it, right? Here the
>> idea is to bind various immutable things together to something that's
>> still immutable. The immutability of non-tuples isn't there to
>> catch 'mutation' errors (as that there is no way to actually mutate them
>> at all). It's there for the nice semantics. Tuples supply the syntax and
>> semantics to do this with multiple immutable things. Your problem arises
>> because there's something very similar to a tuple, a, list, which does
>> support mutation syntax. Would you have the same trouble with immutable
>> strings if there were mutable strings in the language? And why not?

> You might ask, "why doesn't Python have both mutable strings and
> immutable?".

In fact, you can easily construct mutable strings from for instance
Arrays of characters in Python, but mutable strings are not clearly
present, that's true.

> Some languages (like Java) do. One answer: having both flavors
> doesn't buy us much, even though there might be an occasional use for a
> mutable string. I'm glad we don't have both flavors of string. Similarly, I
> don't think we need both mutable and immutable sequences. But in the case of
> sequences, it's the immutable flavor that doesn't buy us much.

> By the way, for those who contend that tuples are a good way to pass around
> groups of heterogeneous elements, I contend that they have some drawbacks.
> Take the example of "point" that was previously used in this thread. Create
> a point p = (11, 22). How do I refer to the elements? As p[0] and y[1]. Hmm,
> which is x and which is y?

No, you unpack it:

x, y = p

and then you refer to the elements.

> Wouldn't it be better if I could reference the
> elements by name, x.x and x.y? Some languages have record types where you
> can declare "record Point(x, y)", create an object as Point(11, 22), and
> refer to the elements as p.x and p.y.

In Python this is called 'class':

class Point:
    pass

mypoint = Point()
mypoint.x = 1
mypoint.y = 2

If you want convenience, write an __init__ function for Point. :)

> A language supporting this is Icon,
> and records are roughly as efficient as Python's tuples even though the
> elements have names.

I've seen the idea of named tuple addressing thrown around before on
the list; I think Paul Prescod came up with it a while back. It's
an interesting idea, might have some merits, but probably has hassles
too, and it's not certain it's worth those. Look up the thread in deja news
if you want a more detailed discussion.

> Even in JavaScript you can construct something like a
> tuple (but not immutable) with named elements using the same syntax as
> Python dictionary "literals" {x: 11, y: 22}.

What's wrong with using classes? 

> Although I clearly prefer
> Python as a language over JavaScript, that is a nice feature. Of course, you
> can effect the same behavior in Python, but not so conveniently.

Classes may be a trifle more verbose, but I don't really think my example
above was *that* verbose; even if you use an '__init__' method. And if 
they're too heavy weight, tuple unpacking is your friend. :)

> I think
> that tuple-type objects with named elements are superior to numbered
> elements for groups of heterogeneous elements, and that immutability is not
> often necessary and even sometimes an inconvenience.

It should rarely be necessary that you need to access tuple elements by
numbers; it's supposed to be a collection of heterogeneous objects (in
my understanding of idiom), and accessing by number usually doesn't make
any sense. Just unpack them.

[snip]
>> I do think the Python language supports the homogeniety idea. Just because
>> Python supports heterogeneous lists doesn't mean you should be using them
>> all over the place.

> But, why not? How is it damaging?

It's damaging if you go and access those heterogeneous objects by numerical
indexes. It isn't very readable. Why not use a tuple in such a
case and unpack it? Or if you have multiple similar heterogeneous lists,
use a list of tuples. Or of course you can always switch to classes.

> When I first discovered a language that
> did not require homogeniety (is that a word?) in lists, I though of it as an
> advantage, not a drawback. (That was Snobol4, a *long* time ago!) I still
> feel that way.

It's very powerful in OO as you can exploit polymorphism. The way the
.sort() method of lists uses the __cmp__() method is a good example.
I'm not sure what use it is to have a list of say, an integer, 3 strings,
an integer, and a floating point number, etc, etc, unless you're going
to treat this list somehow as a list of structs of objects. Which means
you should've been using tuples or classes in the first place, right?
Could you give me an example of where this is handy?

Of *course* the dynamic typing is an advantage, but still in by far the
most lists you're treating all items as *similar* objects. You assume
they all support the same interface; perhaps you do some checks like
hasattr() or type comparisons, but if that gets out of hand it usually
means your code isn't designed the right way, right?

> Some of my following responses reflect the fact that I haven't bought into
> the if-a-sequence-is-a-list-it-must-be-homogeneous concept.

[snip]

>> > There have been cases where I used tuples, to find out
>> > later that I wanted mutability, so went back and changed.
>>
>> Hm, I rarely if ever seem to run into this. If I need mutability of
>> a set of heterogenous objects I tend to use a class. There may be a
>> switch from tuples to classes, but not often from tuples to lists, I
>> think.

> Well, since I don't have an aversion to heterogeneous lists, it's kind of
> the same thing  :-) Yes, the proper choice is class, but a small list (or
> tuple) is much more convenient (for the coder, BTW, not the reader :-). And,
> after all, Python is a "rapid development" language where convenience is
> highly valued. Just don't forget which element number is which! Or if you
> change the order later, don't forget to change all the references.

I don't really understand how this is more convenient. I need to juggle
element numbers in my head, remember offsets, difficult. Unless you mean
you unpack the tuples. For small enough tuples you can usually remember
what's in them, but I do always need to look up the time tuple when
I use the time module. That might've been nicer as a class. :)

>> Perhaps you shouldn't think 'tuples or lists', but 'tuples or classes',
>> then?

> If tuple is a super-convenient-quickie substitution for class, then, to me,
> list is just a mutable version of the same concept. (And neither of those is
> as expressive as using a class.)

Okay. I can see that usage, but I don't see how much convenience you buy
here; it's not hard to write:

class Foo:
   pass

foo_object = Foo()

instead of:

foo_list = []

Of course the latter is shorter, but you pay heavily when you do 'attribute
access':

foo_object.x, foo_object.y = 10, 11 
foo_object.y = 22
foo_object.z = 14
print foo_object.x, foo_object.y

foo_list = [10, 11]
foo_list[1] = 22
foo_list.append(14)
print foo_list[0], foo_list[1]

The latter is a lot less readable. And less writable, as soon as you start
changing individual elements of the lists.

> You're suggestion is change tuple-to-class,
> in my example I changed tuple-to-list. Either way is a different means to
> the same end (yours is more correct, but more laborious). One might start
> out using tuple not realizing that he/she would later want mutability for
> that particular use. Then he/she has to change existing code. In a language
> focused on rapid development, that is not good. I guess the moral is: if
> you're not 100% sure, use list :-)

My moral would be: if you run into the limitations of tuples, switch to
classes. :) I think that starts saving you hassle almost immediately.
Python is great because classes are *not* laborious!

> And if there were only one tuple/list datatype, we'd never have to decide
> about it during programming, or go back and change it when we don't
> anticipate correctly, or explain the difference to Python newbies, or have
> FAQ items about it, or have discussions such as this :-)

Instead we'd have people straining lists where they should've used 
classes, more weird reference semantics bugs (I passed [x, y] to this
function and then changed it and now nothing works, whaa), and Tim
Peters switching to Haskell because it has tuples. :)

Regards,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From cut_me_out at hotmail.com  Thu Apr 13 13:25:40 2000
From: cut_me_out at hotmail.com (Alex)
Date: 13 Apr 2000 13:25:40 -0400
Subject: 1st non-trivial program - gentle criticism requested
References: <1e90r16.1sd2ts41e7ur6kN@gershwin.bio.ic.ac.uk> <etdk8i2t6od.fsf@w20-575-109.mit.edu>
Message-ID: <etdhfd5ud4b.fsf@w20-575-109.mit.edu>

> I realized that I have never, ever used a global variable in a python
> script.

Oops, I meant I've never used the keyword 'global' 

Not using global variables at all would be a bit of a bore. :)

Alex.


From aaron at reportlab.com  Mon Apr 24 09:38:32 2000
From: aaron at reportlab.com (Aaron Watters)
Date: Mon, 24 Apr 2000 06:38:32 -0700
Subject: ANN: NJ Piggies welcome Andy Robinson
Message-ID: <39044E57.58467ED@reportlab.com>

The NJPiggies Welcome Andy Robinson!

Date: Monday 1 May 2000
Time: 7:00pm to whenever
Place: the Hyatt, New Brunswick, New Jersey
  (intersection of routes 27 and 18)
  meet near the bar in the Atrium.

Fun fun fun!

Please join us as the New Jersey Python interest group (NJ Piggies)
welcome Andy Robinson, Chief Architect and CEO of ReportLab,
visiting from London.

Also attending will be Jim Ahlstrom of Interet, the staff of Quantiva
and Aaron Watters and Et Perold of ReportLab -- in addition to anyone
else who cares to attend.

PLEASE POST!

PDF announcement:
   http://www.chordate.com/map.pdf
program that generated the pdf announcement:
   http://www.chordate.com/map.py

ps: Our lawyers asked us to emphasize
  that Andy Robinson IS NOT Sharon Stone.  Sorry
  for the confusion!





From neilh at hare.net.au  Thu Apr 27 21:55:05 2000
From: neilh at hare.net.au (Neil Hodgson)
Date: Fri, 28 Apr 2000 11:55:05 +1000
Subject: [Python-Dev] Re: Python 1.6a2 Unicode bug (was Re: comparing strings and ints)
References: <l03102805b52ca7830b18@[193.78.237.154]> <39072F0C.5214E339@lemburg.com> <BlHN4.1719$wYl.195075584@newsb.telia.net> <oqvh14fikb.fsf@titan.progiciels-bpi.ca> <45919986@data.franken.de>
Message-ID: <048501bfb0b4$c793d760$4add3fcb@neil>

> Would it be possible to write the source code in UTF-16 (which is much
> easier to to under WinNT with NotePad, I know no text editor under that
> writes UTF-8)

   The W2K version of Notepad writes UTF-8 as well as UCS-2. I don't know of
any editor that understands UTF-16. Idle writes UTF-8. Beta versions of
Scintilla/SciTE and Pythonwin 131 can write UTF-8.

   Neil





From mjhand at concentric.net  Tue Apr 18 01:51:56 2000
From: mjhand at concentric.net (Manus Hand)
Date: Mon, 17 Apr 2000 23:51:56 -0600
Subject: Upgrading IIS to 1.6 problem....
References: <2B1262E83448D211AE4B00A0C9D61B03BA72C9@MSGEURO1>
Message-ID: <38FBF7FC.509A1D37@concentric.net>


Pieter Claerhout wrote:
> 
> If you go to the Internet Services Manager, and you select a web
> application, you can edit the application options (in IIS 4 or higher,
> a webapp is a folder with the blue icon). If you go to the home directory
> tab, you can click on the configure button to change the settings for your
> application. If you go there to the App Mapping tabs, to what points the
> extension .cgi? My guess it that it's pointing to the python 1.5
> executable and not to the 1.6 executable.
> 
I'll check that when I get to work in the morning.  Thanks for the
pointer.
I failed to consider that the IIS Mgr apps might be overriding the
registry setting.  I'll let you know if this turns the trick.

Manus



From kistler at gmx.net  Fri Apr  7 16:15:48 2000
From: kistler at gmx.net (Per Kistler)
Date: Fri, 07 Apr 2000 22:15:48 +0200
Subject: Problem running Python script as cron job
References: <wrd-0AEDDA.12282507042000@news.awenet.com>
Message-ID: <38EE41F4.8050708@gmx.net>

Hi Bill

Try stderr?
/path/prog > /path/log 2>&1

Maybe it spits some errors?

Per.



From wrogers at ektron.com  Sun Apr 16 20:27:02 2000
From: wrogers at ektron.com (William Rogers)
Date: Mon, 17 Apr 2000 00:27:02 GMT
Subject: eWebEditPro  WYSIWYG Text Area Field
Message-ID: <qTsK4.10498$WF.592718@bgtnsc04-news.ops.worldnet.att.net>

Ektron, has been requested to generate sample code for Python.   We are
looking for a python developer to develop sample code like our PHP3,  ASP or
ColdFusion code.  We will trade some ewebeditPro editors for the sample
code.

eWebEditPro is a browser-based, easy-to-use, WYSIWYG Text Area Field.  A
content authoring tool for business users with no knowledge of HTML.  The
WYSIWYG (What You See Is What You Get) Web editor allows content
contributors to easily create and edit content in a word processor-like
environment.  eWebEditPro allows the site manager or Webmaster to set
parameters for look and feel to maintain Web site consistency.  It is the
only browser-based Editor that is XHTML compliant.  XHTML is HTML that
conforms to XML. For more information regarding XHTML, see the official W3C
web page on XHTML at http://www.w3.org/TR/xhtml1/.
Note Netscape support will soon be released.

http://www.ektron.com/products/ewebeditpro.cfm
  a.. WELL-FORMED HTML is produced, no overlapping tags.
  b.. Office 2000 filters.
  c.. Extra font tags are removed.
  d.. INTERNATIONAL LANGUAGES for the menus and dialogs.
Thanks
William Rogers
CEO  Ektron Inc.
http://www.ektron.com
wrogers at ektron.com
Ektron office;  (603)-594-0249 ext 106





From ed at sis.yorku.ca  Tue Apr 18 08:50:39 2000
From: ed at sis.yorku.ca (Ed F. McCurdy)
Date: Tue, 18 Apr 2000 08:50:39 -0400
Subject: Any Web-Mail Gateway (written in Python)?
In-Reply-To: Alessandro Bottoni's message of "17 Apr 2000 11:48:47 -0400"
             <6D8A17398E28D3119F860090274DD7DB4986A3@pces.cadlab.it>
References: <6D8A17398E28D3119F860090274DD7DB4986A3@pces.cadlab.it>
Message-ID: <200004181250.IAA11781@sprawl.sis.yorku.ca>

You should look at mailman; http://www.list.org/

   What is Mailman? 
   Mailman is software to help manage email discussion lists, much like
   Majordomo and Smartmail. Unlike most similar products, Mailman gives
   each mailing list a web page, and allows users to subscribe,
   unsubscribe, etc. over the web. Even the list manager can administer
   his or her list entirely from the web. Mailman also integrates most
   things people want to do with mailing lists, including archiving,
   mail-to-news gateways, and so on. See the features page for a brief
   overview of features.

Hope this helps,
Ed
-- 

Edward F. McCurdy,               | ed at sis.yorku.ca
York University, Toronto, Canada | (416) 736-2100x22685




From chafik at nevrax.com  Tue Apr 11 10:39:34 2000
From: chafik at nevrax.com (Sameh chafik pro)
Date: Tue, 11 Apr 2000 16:39:34 +0200
Subject: Are there any one, now how can i do to run a node table ?
Message-ID: <8cv9iq$2eh$1@gw.localdomain>

Hello evrybody,
I find in the Python C API the folwing function:
struct _node* PyParser_SimpleParseString (char *str, int start);
struct _node* PyParser_SimpleParseFile (FILE *fp, char *filename, int
start);
but i do'nt fin a function like
PyObject* PyRun_String (struct _node *str, int start);
Are there any one, now how can i do to run a node table ?




From dalcolmo at mediaone.net  Wed Apr 19 23:39:08 2000
From: dalcolmo at mediaone.net (Josef Dalcolmo)
Date: Thu, 20 Apr 2000 03:39:08 GMT
Subject: win32all-128.exe availability?
References: <38fd01b2$0$10982@senator-bedfellow.mit.edu>
Message-ID: <wZuL4.8229$QX1.310320@typhoon.ne.mediaone.net>

Robert <soundhack at hotmail.com> wrote in message
news:38fd01b2$0$10982 at senator-bedfellow.mit.edu...
> Now that starship has crashed, I was wondering if it was possible to
> get the 128 build of Pythonwin extensions off the net? By an unfortunate
> coincidence, my own hard drive chose to crash around the time starship
> did.
>

I got the same question. I have build 125, but suspect that the WinComm
package needs something newer.

- Josef




From m.faassen at vet.uu.nl  Tue Apr 18 15:45:02 2000
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 18 Apr 2000 19:45:02 GMT
Subject: gaming and python
References: <000c01bfa91c$f421d6a0$0f8d0b3f@oemcomputer> <8dhifa$aom$1@nnrp1.deja.com>
Message-ID: <8didvv$p4m$1@newshost.accu.uu.nl>

sp00fD <sp00fD at yahoo.com> wrote:
> Well, I'd have a hard time believing that you could program something
> like quake, but you could probably do a decent job with tetris or the
> like.  Check out PySDL http://pysdl.sourceforge.net/, I haven't used
> it, but it's probably exactly what you need, SDL is apparently a cross
> platform library for game programming.

It's way cool! I've been playing it from Python and both SDL and the
wrapper are excellent, and both are improving. It's amazingly easy to do
graphics from Python -- easy to build up your abstractions and so on. And
still pretty fast so far. Though I'm sure I'll run into the limitations
eventually and have to start moving some parts into C.

SDL is quite low-level, PySDL too but a bit less so. You will probably
need to know a little about bitmapped graphics concepts to be able to
use it effectively -- coordinates, pixels, RGB colors, and blitting. It's
not hard to play with it all from Python, though.

Anyway, PySDL is definitely recommended. I heard a Windows port of it
still needs to be tested, perhaps someone could take a look at it? I've
been using it in Linux without problems.

Regards,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From lamzak8642 at my-deja.com  Sat Apr  8 11:37:34 2000
From: lamzak8642 at my-deja.com (lamzak8642 at my-deja.com)
Date: Sat, 08 Apr 2000 15:37:34 GMT
Subject: read _and_ write from/to commandline program
Message-ID: <8cnjnt$k7o$1@nnrp1.deja.com>

Hi,

Is it possible to read from _and_ write to a commandline program (Linux
+ Python)? Apparantly I can't do this with os.popen() because it is
unidirectional.

I'm looking for something like this pseudo code:

	proces = magic.function('myprogram', "wr")
	proces.write('do this')
	result = proces.read

Any options?

LZ


Sent via Deja.com http://www.deja.com/
Before you buy.


From mak at imakhno.freeserve.co.uk  Wed Apr  5 16:08:23 2000
From: mak at imakhno.freeserve.co.uk (Makhno)
Date: Wed, 5 Apr 2000 21:08:23 +0100
Subject: Some suggestion for a GUI kit evolved on Perl
Message-ID: <8cg7oo$m10$1@newsg4.svr.pol.co.uk>

I have written a GUI kit called Orchard. It is essentially a code library
with c-like functions which when called correctly will create windows,
buttons and other controls. I have had great success writing an interface
for this with Perl, and now I'm about to attempt a Python module.
One of the main points of Orchard is to keep the GUI-creation code as quick
and as neat as possible. Because Perl ignores whitespace, I have designed
the module in such a way so that the user can take advantage of this and use
whitespace indicate GUI structure. For example, the following code creates a
window with a button and a label in it:

Orchard::NewOrcWindow
(
    title=>"Test"
)->Add
(
    OrcVGroup
    (
        OrcButton,
        $label=OrcLabel
    )
)->Show;

As you can see, the window consists of a button and a label arranged
horizontally. A ref is kept to the label so it can be changed in later code,
and the window's Add() method returns a ref to the window so that Show() can
be called on it.

The problem I'm having is that Python has reserved whitespace (to indicate
program scope). Having looked at the Tk modules, I see it doing the
equivalent of:

window=Orchard.NewOrcWindow(title="Test")
button=OrcButton
label=Label
vgroup=OrcVGroup
window.add(vgroup)
vgroup.add(button)
vgroup.add(label)
window.Show

The problem is that this code stores references to everything (There's often
no need to keep references to buttons, it's looked after internally), it's
not as easy to read, and is easily complicated by GUI additions.

An alternative would be to use the style I've used with Perl; creating GUI
elements in the contructors of other GUI elements.

Orchard.NewOrcWindow(title=Text).Add(OrcVGroup(...

But without whitespace to help separate the elements I end up with a long
line that looks even worse than before. Separating the long line to several
smaller lines doesn't help readablity either (it just ends up all wrapped
round rather than structured). And the line has to be split up when I want
to keep a reference to the element anyway.

This complication is precisely the sort of thing I wished to avoid when
creating Orchard!
Does anybody have any advice as to how I could 'structure' Python code to
make neat, extendable GUI code? Essentially it comes down to what I can use
to indicate GUI structure. I can't use whitespace, so what can I use?

Thankyou.

(The Library and Perl interface are located here:
http://www.bebits.com/app/829 )











From hamish_lawson at yahoo.co.uk  Fri Apr  7 05:38:03 2000
From: hamish_lawson at yahoo.co.uk (Hamish Lawson)
Date: Fri, 07 Apr 2000 02:38:03 -0700
Subject: Any news of ActiveState's Python developments?
Message-ID: <13caa058.14a1c69c@usw-ex0108-063.remarq.com>

Is there any news of Python developments at ActiveState since
they adopted the language? The only thing I've seen is their
announcement of plans to port Python (and Perl) to the Intel
Itanium.

Hamish Lawson


* Sent from RemarQ http://www.remarq.com The Internet's Discussion Network *
The fastest and easiest way to search and participate in Usenet - Free!



From phd at phd.russ.ru  Tue Apr 11 05:54:12 2000
From: phd at phd.russ.ru (Oleg Broytmann)
Date: Tue, 11 Apr 2000 09:54:12 +0000 (GMT)
Subject: Redirecting the output from sys.stdout
In-Reply-To: <2B1262E83448D211AE4B00A0C9D61B03BA7296@MSGEURO1>
Message-ID: <Pine.LNX.4.21.0004110951300.29384-100000@fep132.fep.ru>

On Tue, 11 Apr 2000, Pieter Claerhout wrote:
> I want to run a DOS command from Python, and I want to redirect the output of
> that command to a variable. Anyone a suggestion on how to accomplish this? The
> output of the command should not be shown on the screen.

   Long time ago I wrote a version of popen for DOS. If popen is not
available in your version of module os - download it here:
   http://sun.med.ru/~phd/Software/Python/
Look for dospopen. These patches are now obsolete, I don't know how well
it'd fitted you.

Faster mirrors: http://members.xoom.com/phd2.1/Software/Python/
http://www.fortunecity.com/skyscraper/unix/797/Software/Python/

Oleg.
---- 
    Oleg Broytmann    http://members.xoom.com/phd2.1/    phd2 at earthling.net
           Programmers don't die, they just GOSUB without RETURN.




From runyaga at thisbox.com  Wed Apr 26 00:52:47 2000
From: runyaga at thisbox.com (alan runyan)
Date: Wed, 26 Apr 2000 04:52:47 GMT
Subject: Python/Zope or better PHP4
References: <8F2172931newsbachich77@192.168.1.1>
Message-ID: <zCuN4.57$JH.4398@typhoon.austin.rr.com>

> However, we're looking for an integrated, object-oriented development
frame
> with some security concepts, so Zope came to our view.

sounds like ZOPE wins within that criteria.  PHP is neither object oriented
nor has a concept of security - at least to the degree of ZOPE.

> My question now, does anyone here have experience with both PHP and Zope?

I've done minor PHP scripting ~ 2 years ago and then did some ASP/COM work
for the past year.  and I feel that ASP/COM is far superior, strictly in
that you can further remove the application (COM objects) logic from the
presentation layer (ASP scripts) - you can do this also w/ ZOPE using
external methods, python methods (application) and DTML
(presentation/reporting).

> Would it be worth take the effort to change to Zope? What about
> advantages/disadvantages?

you would probably be very surprised how fast you can migrate your code from
PHP to ZOPE, especially if you are doing lots of database calls.  but.. this
is comp.lang.python not comp.lang.php, which is a reason to beware of my
bias :) php is a language that can only be exploited in a web environment,
Python is __more__.

advantages:
* ZOPE is more structured than PHP - reusability
* usually easier to 'deploy' products/third party add-ins
* more maintainable
* has concept of 'three tier' - look into ZEO, should be arriving for public
consuption in the next week or two
* community is *really* cool and helpful

disadvantages:
* ZOPE has a steeper learning curve (object oriented, large framework,
acquistion, ZOPE Zen, Python)
* not as fast as PHP, but ZOPE's cacheing mechanism is great! - go read the
zope.org frontpage!
* documentation isnt as plentiful as PHP
* if you dont know python at first -- its kinda weird figuring out DTML vs
Python and you will be reading lots of legible source code (python) if you
are doing *really* interesting stuff.
* doesnt have _ALL_ the example code that PHP has to offer, yet.


> How well does Python work as a web programming
> language compared to PHP (which was designed right from the start for this
> purpose) ?

Python was designed for maintainability.  ZOPE was designed from the
beginning with the web in mind.  ZOPE is written in Python with minor speed
critical pieces in c.  Python can be used alone in a cgi environment as well
w/o ZOPE.  Pieces of ZOPE can be pulled out of the framework and used by
their lonesome

experience: I took two people at work who have never touched python or ZOPE
and within 2 weeks have created a content manager that includes a simple
workflow and publishes ZOPE objects in XML on a staging server.  <- its been
wildly successful for .py beginners to learn and implement *real* projects
quickly w/ Python and ZOPE!

~alan





From mcalla at home.com  Sat Apr 22 21:36:55 2000
From: mcalla at home.com (Mike Callahan)
Date: Sun, 23 Apr 2000 01:36:55 GMT
Subject: Tkinter's OptionMenu
References: <ZiYL4.110745$8S2.1383832@news1.rdc1.tn.home.com> <8dship$83j$1@nnrp1.deja.com>
Message-ID: <XssM4.112726$8S2.1407777@news1.rdc1.tn.home.com>

Thanks, John. I see how to do it now.

<johngrayson at home.com> wrote in message news:8dship$83j$1 at nnrp1.deja.com...
> In article <ZiYL4.110745$8S2.1383832 at news1.rdc1.tn.home.com>,
>   "Mike Callahan" <mcalla at home.com> wrote:
> > How does one change an OptionMenu widget's list? I know that you can
> do this
> > easily using the PMW widget, but I would like to stay with basic
> Tkinter
> > widgets to keep the code as small as possible. I could not find any
> methods
> > in the code and I could not find any examples in either the online
> > documentation or the new book on Tkinter.
> >
> >
>
> I didn't expand on Tkinter's OptionMenu class too much in _Python and
> Tkinter Programming_, since using Pmw's version is usually more
> convenient. The problem is that you really have to 'roll your own'
> to do much with the vanilla widget.
>
> However, if you insist, here is some sample code to manipulate
> Tkinter's OptionMenu:
>
> from Tkinter import *
> root = Tk()
> root.title('OptionMenu')
>
> class setit:
> def __init__(self, var, value):
> self.__value = value
> self.__var = var
> def __call__(self, *args):
> self.__var.set(self.__value)
>
> class OptionM:
>     def __init__(self,parent):
>
>         self.var = StringVar()
>         self.var.set('BRM')
>
>         self.opt_menu = OptionMenu(parent, self.var, 'Stockbroker',
>                                    'Quantity Surveyor', 'Church Warden',
>                                    'BRM' )
>         self.opt_menu.pack(anchor=W, padx=20, pady=30)
>
>         Button(parent, text='Add item', command=self.addItem).pack()
>
>         Button(parent, text='Change items',
>                command=self.changeItems).pack()
>
>     def addItem(self):
>         self.opt_menu["menu"].add_command(label='Dinsdale',
>                                   command=setit(self.var, 'Dinsdale'))
>
>     def changeItems(self):
>         self.opt_menu["menu"].delete(0, END)
>         for item in ['Passion fruit','Loganberries','Mangoes in syrup',
>                       'Oranges','Apples','Grapefruit']:
>             self.opt_menu["menu"].add_command(label=item,
>                                       command=setit(self.var, item))
>         self.var.set('Loganberries')
>
> om = OptionM(root)
> root.mainloop()
>
> A bit crude, but it's a start. Use Pmw if you can!
>
>     John
>
>
> Sent via Deja.com http://www.deja.com/
> Before you buy.




From conversy at lri.fr  Thu Apr 13 07:45:14 2000
From: conversy at lri.fr (Stephane Conversy)
Date: Thu, 13 Apr 2000 13:45:14 +0200
Subject: Python 1.6 -- Any hope for new threading core?
References: <EDFD2A95EE7DD31187350090279C6767706F14@THRESHER>
Message-ID: <38F5B34A.854588C1@lri.fr>

Brent Fulgham wrote:
it's still necessary to grab the global
> interpreter lock, etc., when writing code for a
> multi-threaded C extension.


did you write such an extension ?
if so, I would be glad if you send me what the
necessary steps are, or if you point me to some
docs on how to do that.

-- 
St?phane Conversy
http://www-ihm.lri.fr/~conversy/
mailto:conversy at lri.fr


From samschul at pacbell.net  Sun Apr 30 23:53:25 2000
From: samschul at pacbell.net (Sam Schulenburg)
Date: Sun, 30 Apr 2000 20:53:25 -0700
Subject: COM question
References: <390CDC8B.779DB0C3@world.std.com>
Message-ID: <fc7P4.1061$Q47.842764@news.pacbell.net>

I assume that you are using Pythonwin. If this is the case you can browse
the registery for your COM objects. The particular example that you give
"Word.Application" should work. Verify that when you browse the registery
with the COM Browser provided in the tools menu, that you see Word.
Sam Schulenburg

David Lees <debl.nospamm at world.std.com> wrote in message
news:390CDC8B.779DB0C3 at world.std.com...
> I am working examples in Hammond book on Python.  How do I get a list of
> names of COM objects?  I am on page 67 and and to know how to figure out
> what is the name for a Microsoft Word object.  In the example it is
> referred to as:
> "Word.Application", but this gives a name error.  I am doing the client
> call correctly, because when I substitute "Excel.Application" or
> "Access.Application"  I have not problem opening and closing Excel or
> Access.  Apparently Word must have a different name for COM stuff on my
> system, so I thought that if someone can tell me how to look at the list
> of COM objects the name will be obvious.  I have Norton Utility under
> Win95 OSR2, so I can easily look at the Registry (though as a Win95
> newbie, much of it is not intelligible)
>
> Thanks in advance.
>
> David Lees
>




From just at letterror.com  Thu Apr  6 04:44:19 2000
From: just at letterror.com (Just van Rossum)
Date: Thu, 6 Apr 2000 09:44:19 +0100
Subject: Imaginative solutions to connect. Was: Python 1.6 alpha 1
 released
In-Reply-To: <ya6r3loa.fsf@dan.resnet.rochester.edu>
References: James Logajan's message of "Wed, 05 Apr 2000 21:03:04 -0700"
 <Pine.GSO.4.10.10004040035420.21084-100000@sundial>
 <5E877FD7637AB46E.DABBED9CF3C1D64B.CDDDC05AE7374A78@lp.airnews.net>
 <5l7leern5v.fsf@eric.cnri.reston.va.us> <38EC0C78.C6170660@Lugoj.Com>
Message-ID: <l03102800b511fe32305b@[193.78.237.124]>

At 12:20 AM -0400 06-04-2000, Daniel Berlin+list.python wrote:
>Why don't we introduce an official deprecation mechanism so that
>python will say "Warning, connect(arg1,arg2) is deprecated, use
>connect(arg1) instead" the first time it sees it or something.

Warnings are good. But: where do they go? Quick proposal (before breakfast,
so I can't guarantee it makes sense): a new variable in sys: sys.stdwarn.
Needless to say, it's a file-like object. Ah well, maybe warning can just
go to sys.stderr after all. Coffee!

Just





From jstok at bluedog.apana.org.au  Wed Apr  5 02:00:20 2000
From: jstok at bluedog.apana.org.au (Jason Stokes)
Date: Wed, 5 Apr 2000 16:00:20 +1000
Subject: A Mountain of Perl Books + Python Advocacy
References: <20000405053920.20579.qmail@web2102.mail.yahoo.com>
Message-ID: <3FAG4.62951$3b6.243372@ozemail.com.au>

lewst wrote in message
<20000405053920.20579.qmail at web2102.mail.yahoo.com>...
>Before I launch into another question and gripe, I'd like to thank
>everyone who offered comments and suggestions on the advanced Python
>books that are out there.
>
>Now something I can't quite figure out is: why are there so many more
>books on Perl out there than on Python?


The market for Perl is 10-100 times bigger than Python.




From quinn at photo.ugcs.caltech.edu  Mon Apr 24 13:48:10 2000
From: quinn at photo.ugcs.caltech.edu (Quinn Dunkan)
Date: 24 Apr 2000 17:48:10 GMT
Subject: Simple list.append() question
References: <meh9-5ED8CD.11192424042000@news.cit.cornell.edu>
Message-ID: <slrn8g92bb.h8a.quinn@photo.ugcs.caltech.edu>

On Mon, 24 Apr 2000 11:19:24 -0400, Matthew Hirsch <meh9 at cornell.edu> wrote:
>Hi All,
>
>I have this:
>>>> entire_list=[[]]*3
>>>> entire_list
>[[], [], []]
>>>> entire_list[0].append(1.23)
>>>> entire_list
>[[1.23], [1.23], [1.23]]
>
>I only want to append 1.23 to entire_list[0].  Why did it also append it 
>to entire_list[1] and entire_list[2]?

When you multiply a list, you get multiple references to the *same* object.
See FAQ 4.50



From nmm1 at cus.cam.ac.uk  Wed Apr  5 12:02:40 2000
From: nmm1 at cus.cam.ac.uk (Nick Maclaren)
Date: 5 Apr 2000 16:02:40 GMT
Subject: Pass by reference ?
References: <38E864C9.E1EE94F7@ecs.soton.ac.uk> <38E8D449.FBC1E2AD@roguewave.com> <8caohp$j4v$1@pegasus.csx.cam.ac.uk> <38EAB103.E76E74C2@cosc.canterbury.ac.nz>
Message-ID: <8cfo30$4i9$1@pegasus.csx.cam.ac.uk>

In article <38EAB103.E76E74C2 at cosc.canterbury.ac.nz>, Greg Ewing <greg at cosc.canterbury.ac.nz> writes:
|> Nick Maclaren wrote:
|> > 
|> > Unless I am thoroughly confused, Python uses bog-standard call by
|> > reference, with the wrinkle that assignment is a rebinding rather
|> > than a copying operation.  AM I confused about that?
|> 
|> The trouble with describing Python parameter passing as
|> "call by reference" is that that term implies something
|> like a Pascal VAR parameter, which Python most certainly
|> does *not* have. That's why Mr. Generowicz got confused -
|> he thought it meant that a reference to the *variable*
|> x was being passed.

Maybe, though Pascal VAR parameters are usually regarded as an
aberrant form of call by reference.

|> If I had to pick one of the standard terms, I would describe 
|> Python parameter passing as call-by-value, with the wrinkle 
|> that the value being passed is always a reference to an object.

Well, you could put it that way :-)  That particular viewpoint
is very similar to that of the the base Algol 68 language (i.e.
WITHOUT the syntactic sugar of variable declarations), which had
similar properties except that identifiers (i.e. names) were not
rebindable.

It comes down to whether you think that the argument is the
object, or the reference to the object.


Regards,
Nick Maclaren,
University of Cambridge Computing Service,
New Museums Site, Pembroke Street, Cambridge CB2 3QG, England.
Email:  nmm1 at cam.ac.uk
Tel.:  +44 1223 334761    Fax:  +44 1223 334679


From garry at sage.att.com  Mon Apr 24 13:34:32 2000
From: garry at sage.att.com (Garry Hodgson)
Date: Mon, 24 Apr 2000 17:34:32 GMT
Subject: Running shell programs from Python
References: <38FC64DA.7D297563@ffi.no> <8di47s$mkg$1@nntp6.u.washington.edu>
Message-ID: <390485A8.B5ADAAE9@sage.att.com>

Donn Cave wrote:

> The references are mainly UNIX man pages pipe(2), fork(2), fcntl(2)
> (look for dup2), exec(2) and wait(2).
> 
> Basically, the sequence is repeat ( open pipes, fork, close pipes )
> After a fork, the parent and child both have both ends of each pipe
> open, and it's important to close the unused "write" ends so the
> pipe will generate an EOF when its real writer exits.

i've always been curious why this wasn't encapsulated as a single
command in the os module.  i use it infrequently enough that i always
need to go look the idiom, or skip it for simpler things.  seems like
a single command that returned a process id, and a pair of file
descriptors would have lots of uses.

-- 
Garry Hodgson                   Every night 
garry at sage.att.com                a child is born
Software Innovation Services        is a Holy Night.
AT&T Labs                         - Sophia Lyon Fahs


From effbot at telia.com  Tue Apr 25 09:19:21 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Tue, 25 Apr 2000 13:19:21 GMT
Subject: what does XDR stand for? and what is it?
References: <002101bfaeb5$b57f5dc0$2801000a@iel.ie>
Message-ID: <tXgN4.3995$Za1.59408@newsc.telia.net>

Shaun <shogan at iel.ie> wrote:
> im here looking and trying to figure out the xdrlib module,
> it says here - http://www.python.org/doc/current/lib/module-xdrlib.html
> "The xdrlib module defines two classes,
> one for packing variables into XDR representation,
> and another for unpacking from XDR representation.
> There are also two exception classes."
>
> my question is what is XDR? what does it stand for?

look at that page again:

    "The xdrlib module supports the External Data Representation
    Standard as described in RFC 1014, written by Sun Microsystems,
    Inc. June 987. It supports most of the data types described in
    the RFC."

the xdrlib page also includes a pointer to the RFC 1014 specification.

</F>




From digitome at iol.ie  Fri Apr 28 15:35:56 2000
From: digitome at iol.ie (Sean Mc Grath)
Date: Fri, 28 Apr 2000 19:35:56 GMT
Subject: [Announce]:RAX - a database style API for XML - Python Implementation
Message-ID: <3909e259.55223356@news.iol.ie>

RAX is a record-oriented API for
XML processing suitable to working with the
"flat" XML typically generated from relational
databases.

Article explaining RAX with a RAX implementation
in Python is at http://www.xml.com/pub/2000/04/26/rax/index.html

regards,
Sean McGrath



From dalke at acm.org  Sat Apr  1 06:28:32 2000
From: dalke at acm.org (Andrew Dalke)
Date: Sat, 1 Apr 2000 04:28:32 -0700
Subject: why did string.py change in 1.6?
Message-ID: <8c4mdr$5p7$1@slb6.atl.mindspring.net>

In the CVS version (1.6a1) I see that string.py has been
changed a bit.  For example, string.join used to be implemented
by strop.join, which is a C function.  Now it is something
like:

def join(list, sep = " "):
  return sep.join(list)

The function strop.join still exists (so there is duplicated
code between that and the string module).  Given the extra
method call overhead, can anyone tell me why this change was
made?

                    Andrew
                    dalke at acm.org





From pozman at my-deja.com  Wed Apr 19 15:17:49 2000
From: pozman at my-deja.com (pozman at my-deja.com)
Date: Wed, 19 Apr 2000 19:17:49 GMT
Subject: python templating system
Message-ID: <8dl0oc$7ct$1@nnrp1.deja.com>

Does anyone know of a templating system (other than zope) that is
implemented with python [ comparisons would be embperl with perl]

Thanks!


Sent via Deja.com http://www.deja.com/
Before you buy.


From embed at geocities.com  Thu Apr 27 15:43:35 2000
From: embed at geocities.com (Warren Postma)
Date: Thu, 27 Apr 2000 15:43:35 -0400
Subject: Python 2.0
References: <3907B52D.B01D5748@myself.com>
Message-ID: <XO0O4.12715$HG1.340574@nnrp1.uunet.ca>

> I'm evaluating the viability of developing a number of largish
> applications over the next few months. However, I'm concerned about them
> not being conformant with Python 2.0. I know that the interpreter could
> simply be left at the 1.5.2 or 1.6 level, but this will probably not
> satisfy my customer. I could go with that other scripting language, but
> would rather not.

I am quite happy with 1.5.2, and I am lucky because my customers don't know
anything about Python 2.0, and I haven't mentioned it. I'm not sure even
Guido knows [yet] everything that will be different in 2.0, since they're
really still in the 1.6 alpha stages of development right now.

I suspect that if Python 1.5.2 does everything you need, you should stick to
solving the problems you have right now, and not worry too much about trying
to prevent all future problems before they happen. be careful you don't get
so Future Minded in your approach that you never get anything done. And you
certainly shouldn't choose languages based on whether or not they're going
to change, because either they will change, or they have stopped evolving,
and are basically dead.   Take your pick, but both of those problems are
there, and they won't be easily cheated.

Warren




From aa175 at torfree.net  Fri Apr 21 23:09:10 2000
From: aa175 at torfree.net (Ken McCracken)
Date: Sat, 22 Apr 2000 03:09:10 GMT
Subject: Using python on the web
References: <EDFD2A95EE7DD31187350090279C6767754BA8@THRESHER>
Message-ID: <FtEE3B.Ar2.0.bloor@torfree.net>

Brent Fulgham (brent.fulgham at xpsystems.com) wrote:
: A couple of questions:

:  
: > We build medium-busy, CGI-intensive sites in Python. We were briefly
: > quite excited about Zope, but once we realized it locked us into Yet
: > Another Programming Language, we lost interest. DHTML is particularly

I'm a bit of an armchair programmer, but it would seem that JSP nad
Jpython and a Db could be put together to end up with a zopelike result,
without language lockin. No?

Ken McCracken
Toronto


From johann at physics.berkeley.edu  Wed Apr 12 16:23:11 2000
From: johann at physics.berkeley.edu (Johann Hibschman)
Date: 12 Apr 2000 13:23:11 -0700
Subject: global variables - how??
References: <20000412175218.784841CE49@dinsdale.python.org> <8d2jsh$3a9$1@news101.telia.com>
Message-ID: <mtpurvvzkg.fsf@astron.berkeley.edu>

Anders Eggers-Krag writes:

> it is obvious not to use global variables except when the use of
> them is obvious, in my case I could otherwise send them along with
> every function, creating a huge overhead passing a million pointers
> around, using exactly the same amount of memory, or in fact more,
> than I would using globals.

> I did figure out how to do it myself after a lot of trying, but as
> you said it is ugly, but this really anoys me, and I *was* starting
> to like python...

Weird.  Honestly, in years of writing python code, I've never had
a problem with globals, and I'm more than a little curious as to
what you're doing.

If the globals are only needed inside one module, you can simply refer
to them by adding 'global foo' at the beginning of the module, and you
only have to do that if you want to re-bind the variable.  (i.e. not
even if you're modifying a mutable object itself.)

If you need globals across multiple modules, just define a module to
put them in.  i.e. make an empty file, 'gl.py', possibly just
containing a comment:

gl.py:
--------------------------------------------------
"""
Dummy module to contain global variables.
"""
pass
--------------------------------------------------

Then you can just import gl and modify its contents to you heart's
content.

import gl
gl.x = 10
print gl.x
etc.

This is a bit hacky, but it works fine.

If you want to change variables around in the "if
__name__=='__main__'" block, you're best off simply re-importing the
module, then changing things through that.

ie.
t1.py:
------------------------------
import t2
x = 10
if __name__=="__main__":
  import t1     # pull itself into scope, evading the __main__ hackery
  t1.x = 100    # modify the vars properly
  t2.do_stuff() # go from there

Does this help?

--Johann

-- 
Johann Hibschman                           johann at physics.berkeley.edu


From aahz at netcom.com  Mon Apr 17 10:53:35 2000
From: aahz at netcom.com (Aahz Maruch)
Date: 17 Apr 2000 14:53:35 GMT
Subject: Was Re: [ANNOUNCE] Garbage collection for Python, now Stackless comments
References: <20000407221750.A29279@acs.ucalgary.ca> <slrn8fjtao.2d4.quinn@drachma.ugcs.caltech.edu> <38FA05C9.442868E4@tismer.com> <38FAEEC9.B5CED499@webwitches.com>
Message-ID: <8df8hf$6su$1@nntp9.atl.mindspring.net>

In article <38FAEEC9.B5CED499 at webwitches.com>,
Shae Erisson  <shae at webwitches.com> wrote:
>
>PS. Can someone point me to example code for building smalltalk style
>closures easily with continuations?

I'm guessing that you probably want coroutines?
--
                      --- Aahz (Copyright 2000 by aahz at netcom.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

"First one back from the bathroom gets to be the bottom."


From alwagner at tcac.net  Mon Apr 24 09:53:04 2000
From: alwagner at tcac.net (Albert Wagner)
Date: Mon, 24 Apr 2000 08:53:04 -0500
Subject: The Simple Economics of Open Source
References: <B527E7A5.51AA%raffael@mediaone.net> <3903316C.5D6F1611@tcac.net> <raffael-34658F.23400223042000@news.ne.mediaone.net>
Message-ID: <390451C0.4B5D3196@tcac.net>

You STILL don't understand.  Artifice can only copy.  It cannot invent.
An end product can be copied, even a style can be copied, but not the
creative process itself.

Raffael Cavallaro wrote:
> 
> >Talent is not a "secret" that can
> >be hoarded or shared.  It is visible in the end product, yet no amount
> >of sharing how it was done will allow the process to be duplicated.
> 
> Oh, so that's why museums have been repeatedly fooled to the tune of
> millions of dollars over the years by forgeries of the works of great
> artists.
> 
> Fact is, knowing how great art is made allows reasonably skilled
> individuals to make effectively indistinguishable copies, or fake
> originals. This simple fact was well known to artists in the past, and
> it's why they guarded their secrets well.
> 
> It is only in an era when many artists have little or no technical
> talent that such an obviously false statement as Mr. Wagner's could even
> be thought to be true.
> 
> Ralph
> 
> --
> 
> Raffael Cavallaro, Ph.D.
> raffael at mediaone.net

-- 
Small is Beautiful


From rumjuggler at home.com  Sun Apr 30 18:04:07 2000
From: rumjuggler at home.com (Ben Wolfson)
Date: Sun, 30 Apr 2000 22:04:07 GMT
Subject: Do I always have to write "self." ?
References: <270420001706386648%pecora@anvil.nrl.navy.mil> <slrn8ghc6t.f01.scarblac-spamtrap@flits104-37.flits.rug.nl> <8ecsgj$6lp$1@slb0.atl.mindspring.net> <m3bt2tonyg.fsf@atrus.jesus.cam.ac.uk> <300420001008012011%pecora@anvil.nrl.navy.mil> <slrn8gp6c4.ilf.kc5tja@garnet.armored.net> <bz1P4.75789$uF1.160076@afrodite.telenet-ops.be>
Message-ID: <eqbpgsobm30cb50u4ctfclobr8qu6816j1@4ax.com>

On Sun, 30 Apr 2000 21:27:35 GMT, ben at co.and.co wrote:

>Samuel A. Falvo II <kc5tja at garnet.armored.net> wrote:
> 
>> I always use Hungarian notation when it makes sense to.  For example, if I
>> have a function in C that accepts a string and an explicit length argument,
>> I would write something like this:
>> 
>> 	uint32 WriteData( void *buffer, char *pString, uint32 cbString )
>> 
>> This tells me that WriteData() takes as parameters a buffer, a pointer to a
>> string, and the number of bytes in the said string.  I fail to see how this
>> is somehow hard to read.  :)
>
>OK, let's bite :-)
>
>    int WriteData(void *buffer, char *str, int len)
>
>I fail to see how this is somehow hard to read :-)

For me, actually, it's easier to read, since I don't have to wonder
why the int parameter is "cbString".  That "cb" means "count of bytes"
is not intuitively obvious.  "len" or "length" is much easier.

-- 
Barnabas T. Rumjuggler

I'm sure there isn't a single person alive who hasn't mistaken an
antimetabole for a chiasmus before.
 -- John Flynn


From nascheme at enme.ucalgary.ca  Thu Apr 20 02:09:36 2000
From: nascheme at enme.ucalgary.ca (Neil Schemenauer)
Date: Thu, 20 Apr 2000 06:09:36 GMT
Subject: __del__ problem - would adopting Garbage Collection fix this?
References: <aw2L4.8907$HG1.245743@nnrp1.uunet.ca> <7ojL4.9424$HG1.254543@nnrp1.uunet.ca> <Z%oL4.9643$HG1.257533@nnrp1.uunet.ca> <m31z41iyhg.fsf@atrus.jesus.cam.ac.uk> <38FE7F1E.168E914A@cosc.canterbury.ac.nz><slrn8ft35a.uqs.nascheme@cranky.arctrix.com> <B5240D22.4AFD%abrahams@mediaone.net>
Message-ID: <slrn8ft7qc.u9n.nascheme@cranky.arctrix.com>

Dave Abrahams <abrahams at mediaone.net> wrote:
>Hmm. If the objects are really garbage (i.e. unreferenceable by
>the programmer), then there really is nothing she can do to deal
>with it.

Uncollectable garbage is placed in a list reachable from Python
(ie. gc.garbage).

>course you can't call the __del__ method, but it seems you
>should be able to reclaim the memory occupied by the objects
>without losing anything.

Maybe, but what are the __del__ methods for if they are not to
free resources?  This situation really should not occur.  With
the default debugging options the collector prints a message to
sys.stderr when it finds uncollectable objects.  Redesigning the
class structure should fix the problem.


    Neil

-- 
HTML needs a rant tag. --Alan Cox


From exarkun at flashmail.com  Sat Apr 22 12:49:04 2000
From: exarkun at flashmail.com (Jp Calderone)
Date: Sat, 22 Apr 2000 16:49:04 +0000
Subject: SSL w/ urllib
Message-ID: <3901D800.DB50D968@flashmail.com>

 I'm looking at these two lines of codes, trying to figure out how to
get the second one to work:

 f = urllib.urlopen('http://www.mudprovider.com/cgi-bin/printvars.cgi',
params)
 f = urllib.urlopen('https://www.mudprovider.com/cgi-bin/printvars.cgi',
params)

with params defined as:
 params = urllib.urlencode({'name': 'bob', 'value': 5})

The first works fine, and f.read() gives the expected output.  
The second gives me this, though:

Traceback (most recent call last):
  File "<stdin>", line 1, in ?
  File "/usr/local/lib/python1.6/urllib.py", line 62, in urlopen
    return _urlopener.open(url, data)
  File "/usr/local/lib/python1.6/urllib.py", line 163, in open
    return getattr(self, name)(url, data)
TypeError: too many arguments; expected 2, got 3


 As far as I can tell, this is because of the 'https' instead of 'http'
My python interpreter is this:

  Python 1.6a2 (#7, Apr 17 2000, 11:50:39)  [GCC 2.95.2 19991024 (release)] on
linux2

 There are references to HTTPS in urllib, but nothing I can't tell what, if
anything, I should be doing differently to get https to work.  Can anyone
help me out?

  Thanks in advance, Jp

-- 
No, `Eureka' is Greek for `This bath is too hot.'
                -- Dr. Who
--
 4:44pm up 24 days, 21:51, 0 users, load average: 0.00, 0.02, 0.01


From sblakey at freei.com  Wed Apr 12 20:21:47 2000
From: sblakey at freei.com (Sean Blakey)
Date: Wed, 12 Apr 2000 17:21:47 -0700 (PDT)
Subject: How to escape escape?
In-Reply-To: <8d32kl$sjg$1@nnrp02.primenet.com>
Message-ID: <20000413001731.369831CEB2@dinsdale.python.org>

On 12 Apr, Larry Gantz wrote:
> I'm new to the snake.
> I think I would like to be able to turn off and on the escape character
> processing so the snake would perform escape processing when I want it.
> 
> It seems that Python always has escape processing.
> 
> Thanks for any help.
> 
> 
I am not sure what you want, but you have two options in this wonderful
language named after insane comedians (cover art on documentation
notwithstanding):
You can use repr() or `` to get the representation of an escape
character.
>>> repr('spam\n')
"'spam\\012'"
Or you can use raw string (precede the quotation mark with the letter
r).
>> r'spam\n'
'spam\\n'
-- 

Sean Blakey
FreeInternet.com
sblakey at freei.com
(253)796-6500x1025




From sylvieb at icestormfx.com  Wed Apr 19 10:31:18 2000
From: sylvieb at icestormfx.com (Sylvie Bruneau)
Date: Wed, 19 Apr 2000 10:31:18 -0400
Subject: A tkMessageBox question!
Message-ID: <38FDC335.B9FDCFE4@icestormfx.com>

Hi Everybody.

    I am using Python for an inhouse tool that runs on irix and WinNT.  I needed to
popup some message dialogs so I used the dialogs provided by Tkinter, ie, the
tkMessageBox.  On irix, the appearance of the dialogs are really ugly since the font
is 'Times 18'.  I want to change the font of the dialogs but I can't find how I can
do it !

    I look at the available options for the tkMessageBox at
http://w1.132.telia.com/~u13208596/tkintrobook/standard-dialogs.htm.  None of them
let me change the font.   Then I looked in the code of tkMessageBox.py to find that
the dialogs are completely done in Tcl/Tk and the 'Times 18' is hardcoded in a file
called msgbox.tcl.

    How can I change this font ?  Thank you for you help !

Sly

--
/=-/=-/=-/=-/=-/=-/=-/=-/=-/=-/=-/=-/=-/=-/=-/=-/=-/=-/=-/=-/
Sylvie Bruneau ( sylvieb at icestormfx.com ) Tel: (514) 527-3963
Software Developer, IceStorm Digital.     Fax: (514) 527-5165
2595 Place Chasse, Montreal, Quebec, H1Y 2C3.

La vie est un cadeau parfois difficile a deballer. (P. Brassard)


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20000419/dd64b0a1/attachment.html>

From i_luv_fudge at I_HATE_SPAM.hotmail.com  Mon Apr 10 21:07:21 2000
From: i_luv_fudge at I_HATE_SPAM.hotmail.com (.,.)
Date: Tue, 11 Apr 2000 01:07:21 GMT
Subject: A Mountain of Perl Books + Python Advocacy
References: <20000405053920.20579.qmail@web2102.mail.yahoo.com> <38ED40B0.A48EC4CC@udel.edu>
Message-ID: <tNvI4.18$_N4.1751@vic.nntp.telstra.net>


>> Now something I can't quite figure out is: why are there so many more
>> books on Perl out there than on Python?


>6 months ago, I discovered Python and it immediately clicked.
>Learned most of Python before buying my first Python book.


I don't know about any of the rest of you guys,  but I figure this is where I 
should speak up.  I have just discovered Python in the last week.  I was 
trying to manipulate about 400 smallish text files with identical datafields 
and differing values into a format that I could display on the web.  After 
spending about 80 hours cutting, pasting deleting and hten realizing that it 
would be good to start again,  so I could include that peice of info afterall, 
 I decided there had to be a better way.

 I have done one semester in C++ about 2 years back, and nothing before or 
since.  I have a very great friend who is helping me out with my script and 
what I want to do, but this is because they have also only just discovered 
Python and have a direct need to use it for a project over the next week.  
This is very lucky for me, cause I am so close to forgetting absolutely 
everything I learnt in that semester, except the basic tenets of form and 
structure.

Trying to get information out of the python website is virtually impossible 
for the layman trying to learn Python.   I am finding it very very hard to 
expand on the stuff my mate has shown me already,  and I need to expand so 
much!  I am looking for the Learning Python book now,  but I doubt I will find 
it on any shelves in my town, as I live in a backwater country downunder.

======

Whuddaya know?   I just rang TechBooks and they have 4 in stock  :-)

I am off to buy one. 

It still doesn't change the fact that this is a VERY hard language to learn as 
a layman without a guide or some sort of coding TLA translator.  When and if I 
get proficient I might write a page for people with that in mind.

:-)

I have no doubt you will be hearing from me again as I post newbie 
questions....


*sigh*

Nemir


From effbot at telia.com  Mon Apr 17 12:35:21 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Mon, 17 Apr 2000 16:35:21 GMT
Subject: Python 1.6a2 Unicode bug (was Re: comparing strings and ints)
References: <jr%I4.643$rc9.190628352@newsb.telia.net> <1256565470-46720619@hypernet.com> <6M1J4.662$rc9.209708544@newsb.telia.net> <8daop0$8fk$1@slb6.atl.mindspring.net>
Message-ID: <d3HK4.2612$Za1.38971@newsc.telia.net>

Aahz Maruch <aahz at netcom.com> wrote:
> >Python 1.6a2 (#0, Apr  5 2000, 23:56:55)
> >Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
> >>>> u"abc" < "???"
> >Traceback (most recent call last):
> >  File "<stdin>", line 1, in ?
> >ValueError: UTF-8 decoding error: unexpected code byte
>
> That looks like a bug to me.

as far as I can tell, it's supposed to be a feature.

if you mix 8-bit strings with unicode strings, python 1.6a2
attempts to interpret the 8-bit string as an utf-8 encoded
unicode string.

but yes, I also think it's a bug.  but this far, my attempts
to get someone else to fix it has failed.  might have to do
it myself... ;-)

</F>




From m.faassen at vet.uu.nl  Sat Apr  8 19:29:19 2000
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 8 Apr 2000 23:29:19 GMT
Subject: Tuples -- who needs 'em
References: <AF223B12F1FDD2119253006097C4FCEC0AD357@cats.pdri.xerox.com> <8chol6$6em$1@newshost.accu.uu.nl> <026e01bf9fe7$9554fd20$74eb0b18@stcla1.sfba.home.com>
Message-ID: <8cofcf$hns$1@newshost.accu.uu.nl>

Bob Alexander <bobalex at home.com> wrote:
> "Martijn Faassen" <m.faassen at vet.uu.nl> wrote:

>> Except that you can suddenly do this:
>>
>> results = foo()
>> ...
>> results.append(5)
>> ...
>> first, second, third, fourth = results # error

> True -- but there is a simple solution: omit the second statement  :-)

That's of course true; I already mentioned that the 'protection' argument
isn't that strong, as Python is dynamically typed anyway.

> There are things you can do to sabotage things even with tuples; for example
> replace "results.append(5)" with "results = (1,2)". Is mysterious changing
> of lists a frequently-occurring problem?

Okay, I'll extend my problem description:

results = foo()
bar = results
bar.append(5)

That can't happen with tuples, can happen with lists. Of course I haven't
made any research on whether it does happen, but the idea of immutable
objects is that assigning them is exactly the same behavior as copying 
them. That's something an easy thing to think about; for instance with
strings and numbers. The idea here is that tuples allow you to think that
way about lists.

>> A tuple tends to contain a few heterogenous things,
>> while a list tends to contain a larger number of homogenous things.

> This distinction seems to be in the minds of several in this thread, but is
> not mentioned in any Python doc I've seen, is not enforced by the language,
> and, IMO, is unnecessary.

Okay, but it's generally practiced in idiom. Language constructions can
exist that partially support idiom, but not enforce it, right? Here the
idea is to bind various immutable things together to something that's
still immutable. The immutability of non-tuples isn't there to
catch 'mutation' errors (as that there is no way to actually mutate them
at all). It's there for the nice semantics. Tuples supply the syntax and
semantics to do this with multiple immutable things. Your problem arises
because there's something very similar to a tuple, a, list, which does
support mutation syntax. Would you have the same trouble with immutable
strings if there were mutable strings in the language? And why not?

> In languages with mutable lists but no tuples,
> lists are often used for spiritually-unmutable sequences of a few
> hetrogenious things. I've never though of the syntactic distinction as
> necessary or particularly helpful.

Okay. 

> I think the "homogeniety" idea is a carryover from the many languages that
> do not permit heterogeneous sequences. Homogeneous sequences should be used
> where appropriate, but there are uses for heterogeneous sequences too.
> Flexibility is good.

But the idea here is that heterogeneous sequences of a small number of
immutable things (which behave with a semantics indistinguishable of
pass-by-value) should also be such a thing. Python passes everything
by reference. But for some things, such as numbers and strings, 
pass-by-value is actually more interesting. So there we get immutability
to support that.

I do think the Python language supports the homogeniety idea. Just because
Python supports heterogeneous lists doesn't mean you should be using them
all over the place..

>> Functional programming languages have the concept. At least the
>> *spirit* that you shouldn't be changing a list. This is different from
> Python,
>> as we do change (homogenous) lists. The pattern in functional programming,
>> which *does* occur often in Python code as well, is some operation over
>> all elements of a list creating a new list (map(), though this is often
>> expressed in Python as a 'for' loop, partially due to the nastiness of
>> 'lambda'). The old list isn't changed.

> As you say, the functional programming paradigm can be nicely expressed in
> Python, and is not dependent on tuples.

Some of it can be nicely expressed, some can't be, and as you say, it's 
not dependent on tuples. But you mentioned you knew no languages with
the concept of immutable lists, and I mentioned functional programming
languages. I did already say Python doesn't work this way, so you don't
have to tell me what I just told you. :)

>> > And I still haven't
>> > heard from anyone about why there isn't similar concern about immutable
>> > dictionaries.
>>
>> People tend to use classes for this purpose. You can give a class a public
>> interface (enforced or not) with only get_() methods, for instance. That
>> gets you just about the same thing as an immutable dictionary would give
>> you. Again, a dictionary tends to contain a larger number of homogenous
>> objects, while an object tends to contain a smaller number of heterogenous
>> objects.

> That same technique could be used to argue against the need for tuples. Of
> course, that's not often desirable in casual programs because it's somewhat
> painful, and not all that necessary.

Right, since it's painful to do it for things like x,y coordinates, 
there's this nice thing, called the tuple. :) 

>> Do you really feel that pain, then? Why don't you always use lists in your
>> Python code, if you want to avoid the decision? That'd be effectively the
>> same as what you're proposing

> Well, it hasn't achieved the threshold of "pain", but I do seem to be using
> lists in preference to tuples more often than I used to. And I still
> speculate that life would be a bit better if I never had to think about
> tuples vs. lists.

It's an interesting speculation. I'm learning a lot here by trying to
come up with the various arguments there are for tuples. I think it's more
than just the dictionary key argument.

> There have been cases where I used tuples, to find out
> later that I wanted mutability, so went back and changed.

Hm, I rarely if ever seem to run into this. If I need mutability of
a set of heterogenous objects I tend to use a class. There may be a 
switch from tuples to classes, but not often from tuples to lists, I 
think.

> This sort of thing
> happens in exploratory programming, so when in that mode one might want to
> use lists unless one is really sure in advance that tuples are the right
> thing.

Perhaps you shouldn't think 'tuples or lists', but 'tuples or classes',
then?

> By the way, when I get a run time error about trying to modify a tuple, it
> never seems to be because I'm changing something I shouldn't. It's always
> because it really should of been a list in the first place.

Why a list? Is it a collection of homogeneous objects, then?

>> If you don't do this, is this because you're
>> afraid other programmers will complain about your code, or because you do
>> believe there's something to decide about?

> There are other possible motivations besides fear of criticism, such as the
> desire to observe the spirit of the language.

Okay, right, I should've said that, sorry for expressing myself poorly. 

> In doing that, it has occurred
> to me that tuples might not provide enough benefit to warrant the additional
> cognitive load.

Perhaps you are trying to switch between the wrong things? Could you
supply us with some examples?

Regards,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From phd at phd.russ.ru  Mon Apr 17 09:57:39 2000
From: phd at phd.russ.ru (Oleg Broytmann)
Date: Mon, 17 Apr 2000 13:57:39 +0000 (GMT)
Subject: read/write file problem
In-Reply-To: <38FB1375.2FF65B96@sec.noaa.gov>
Message-ID: <Pine.LNX.4.21.0004171356490.30648-100000@fep132.fep.ru>

On Mon, 17 Apr 2000, j vickroy wrote:
> >>> file = open (file_name, 'r')
> >>> file = open (file_name, 'w')

   Open JPEG (binary) files in BINARY mode!!!

file = open (file_name, 'rb')
file = open (file_name, 'wb')

Oleg.            (All opinions are mine and not of my employer)
---- 
    Oleg Broytmann      Foundation for Effective Policies      phd at phd.russ.ru
           Programmers don't die, they just GOSUB without RETURN.




From anthony at dstc.edu.au  Wed Apr 12 05:47:45 2000
From: anthony at dstc.edu.au (Anthony J Wilkinson)
Date: Wed, 12 Apr 2000 19:47:45 +1000 (EST)
Subject: Announcing the release of Fnorb 1.1
Message-ID: <Pine.OSF.4.20.0004121946290.30873-100000@azure.dstc.edu.au>

Greetings one and all,

Announcing the release of version 1.1 of 'Fnorb' the Python CORBA ORB.

The major change with this version is that there is now type checking
employed in the marshalling process. This has two major ramifications:
1) fnorb will no longer dump core when given incorrect types - an exception
   will be thrown instead.
2) some code that works under 1.01 due to non-standard language mappings
   will encounter problems.

In tightening up the type restrictions some otherwise working code will fail.
This decision has not been made lightly and, after much discussion and debate,
we have opted for a scheme that is a little more tolerant than the language
mapping specification. In this way we hope to achieve a balance between
Python's flexibility and CORBA's type safety.

Along with this there are numerous patches and bug fixes which are detailed
in the CHANGES document.

Things on the To Do RSN list:-

1) Allow fnidl to re-open modules
2) Implement the POA
3) Incrementally implement the rest of the CORBA datatypes

Get more details and download Fnorb from:-

	http://www.fnorb.com/

The Fnorb Development Team

c/o

Anthony
_____________________________________________________________________
Anthony J Wilkinson                                  anthony at dstc.com
Software Engineer                                     http://dstc.com
DSTC Pty Ltd                                     Ph:  +61 7 3365 4310

---- From the Fnorb README ----

What is Fnorb?

Fnorb is a CORBA 2.0 ORB written in Python (with just the eensiest-teensiest
bit of C code for marshalling and parsing ;^). Python is a mature,
interpreted, object-oriented programming language with powerful high-level
datatypes that make it ideally suited as a scripting language for CORBA.

Fnorb gives the Python programmer access to the wonderful world of CORBA,
like ILU from Xerox PARC. It supports all CORBA 2.0 datatypes (including
Any's) and provides a full implementation of IIOP.  Unlike ILU, Fnorb is
Python and CORBA/IDL specific which makes it simple, light-weight, and easy
to install and use.

Using Fnorb, you no longer have to use <insert your least favourite language
here - who said C++ ;^)> to write CORBA clients and servers - you can now
use Python.  This makes Fnorb ideal for prototyping complex CORBA architectures,
for use as a scripting tool, and for building test harnesses for all your CORBA
development projects.

The Python language mapping used by Fnorb is based on the document currently
undergoing the final stages of the OMG standardisation process.

Fnorb is being developed by the Distributed Systems Technology Centre Pty Ltd.
http://www.dstc.edu.au/






From raffael at mediaone.net  Sun Apr 23 23:35:21 2000
From: raffael at mediaone.net (Raffael Cavallaro)
Date: Mon, 24 Apr 2000 03:35:21 GMT
Subject: The Simple Economics of Open Source
References: <B527E7A5.51AA%raffael@mediaone.net> <%WvM4.1478$RM6.3452288@news-east.usenetserver.com>
Message-ID: <raffael-E67BBE.23345523042000@news.ne.mediaone.net>

> Correlation doesn't imply causality, and when it does, it's easy to get
> causality backwards if you're not careful.  You could get misled into
> spending lots of money to get rich if you just look at the
> correlations.

In the case of cash cows, we don't need a causal argument at all. It's
simply enough to note that people don't share them. Whether they became 
cash cows because they weren't shared (possible, though not likely), or 
they are not now shared because they have become cash cows doesn't 
really matter. The lesson for open source is the same - expect only 
commodity items, known to many, to be open sourced. Expect lucrative 
cash cows to remain closed.

Ralph

-- 

Raffael Cavallaro, Ph.D.
raffael at mediaone.net

-- 

Raffael Cavallaro, Ph.D.
raffael at mediaone.net


From mikael at isy.liu.se  Fri Apr 14 03:09:51 2000
From: mikael at isy.liu.se (Mikael Olofsson)
Date: Fri, 14 Apr 2000 09:09:51 +0200 (MET DST)
Subject: need help just begining
In-Reply-To: <11822ca0.dd34a400@usw-ex0109-068.remarq.com>
Message-ID: <XFMail.000414090951.mikael@isy.liu.se>

On 10-Apr-00 SheaMon wrote:
 >  Hi everyone,
 >  My name is Shea and I am just starting out on programming.
 >  I know not one thing about it, and a person told me to try
 >  Python programming. But I have not idea what to do to start
 >  making a program.
 >  
 >  I main question is this. If i got a copy of this tutor can
 >  I just leave the numbers and stuff like that and just edit
 >  the rest? I get so lost in the number part of the
 >  programming. If anyone here has some time can you please
 >  help me and email me at my address here.

Shea,

I usually recommend two introductions written by Magnus Lie Hetland. 

First, there is "Instant Hacking - Learn how to program with Python",
which is an introduction to programming (using Python) for the complete 
novice. You find it at

  http://www.idi.ntnu.no/~mlh/python/programming.html

There are some numerics there, but not that much. Then, there is 
"Instant Python", which is an introduction to Python for the newbie
who has at least som programming experience. You can find that one at

  http://www.idi.ntnu.no/~mlh/python/instant.html

Good luck!

/Mikael

-----------------------------------------------------------------------
E-Mail:  Mikael Olofsson <mikael at isy.liu.se>
WWW:     http://www.dtr.isy.liu.se/dtr/staff/mikael
Phone:   +46 - (0)13 - 28 1343
Telefax: +46 - (0)13 - 28 1339
Date:    14-Apr-00
Time:    08:57:21

This message was sent by XF-Mail.
-----------------------------------------------------------------------


From nmm1 at cus.cam.ac.uk  Tue Apr  4 08:58:56 2000
From: nmm1 at cus.cam.ac.uk (Nick Maclaren)
Date: 4 Apr 2000 12:58:56 GMT
Subject: Why should I switch to Python? - Infinity of Primes
References: <NDBBKEGCNONMNKGDINPFGEJDDFAA.infonuovo@email.com>
Message-ID: <8ccoug$f4o$1@pegasus.csx.cam.ac.uk>

In article <NDBBKEGCNONMNKGDINPFGEJDDFAA.infonuovo at email.com>, "Dennis E. Hamilton" <infonuovo at email.com> writes:
|> The standard approach is a proof by contradiction starting from the
|> assumption that there is a largest prime.

There is also a constructive proof based on the Fundamental
Theorem of Arithmetic, that is little more complex.

But I am at a complete loss to know why you need to be able
to prove that there are an infinity of primes in order to
switch to Python :-)


Regards,
Nick Maclaren,
University of Cambridge Computing Service,
New Museums Site, Pembroke Street, Cambridge CB2 3QG, England.
Email:  nmm1 at cam.ac.uk
Tel.:  +44 1223 334761    Fax:  +44 1223 334679


From jmassung at magpiesystems.com  Tue Apr 18 16:12:13 2000
From: jmassung at magpiesystems.com (Jeff Massung)
Date: Tue, 18 Apr 2000 14:12:13 -0600
Subject: Forward References?
References: <eviltofu-4C5688.21593817042000@news.earthlink.net>
Message-ID: <oh3L4.1227$zk4.23905@news.uswest.net>

Jerome Chan wrote in message ...
>Just a silly question but how do I do forward references?
>
>
>class a:
>   def x(self):
>      y(self)

Should this be a.y(self)?
Also, someone mentioned above having: self.y() (or self.y(self)), my
question is what is the difference between a.y and self.y?

Jeff





From just at letterror.com  Wed Apr 26 08:04:08 2000
From: just at letterror.com (Just van Rossum)
Date: Wed, 26 Apr 2000 13:04:08 +0100
Subject: Python 1.6a2 Unicode bug (was Re: comparing strings and ints)
In-Reply-To: <8e6bsl$f1a$1@nnrp1.deja.com>
References: <jr%I4.643$rc9.190628352@newsb.telia.net>
 <1256565470-46720619@hypernet.com>
 <6M1J4.662$rc9.209708544@newsb.telia.net>
 <8daop0$8fk$1@slb6.atl.mindspring.net>
 <d3HK4.2612$Za1.38971@newsc.telia.net>
Message-ID: <l03102802b52c7df3422a@[193.78.237.154]>

Fredrik Lundh replied to himself in c.l.py:
>> as far as I can tell, it's supposed to be a feature.
>>
>> if you mix 8-bit strings with unicode strings, python 1.6a2
>> attempts to interpret the 8-bit string as an utf-8 encoded
>> unicode string.
>>
>> but yes, I also think it's a bug.  but this far, my attempts
>> to get someone else to fix it has failed.  might have to do
>> it myself... ;-)
>
>postscript: the powers-that-be has decided that this is not
>a bug.  if you thought that strings were just sequences of
>characters, just as in Perl and Tcl, you're in for one big
>surprise in Python 1.6...

I just read the last few posts of the powers-that-be-list on this subject
(Thanks to Christian for pointing out the archives in c.l.py ;-), and I
must say I completely agree with Fredrik. The current situation sucks. A
string should always be a sequence of characters. A utf-8-encoded 8-bit
string in Python is *not* a string, but a "ByteArray". An 8-bit string
should never be assumed to be utf-8 because of that distinction. (The
default encoding for the builtin unicode() function may be another story.)

Just





From akuchlin at mems-exchange.org  Thu Apr 27 10:27:25 2000
From: akuchlin at mems-exchange.org (Andrew M. Kuchling)
Date: 27 Apr 2000 10:27:25 -0400
Subject: Non SSH accounts on starship?
References: <lbMN4.5092$v85.31026@news-server.bigpond.net.au>
Message-ID: <3d4s8n39gy.fsf@amarok.cnri.reston.va.us>

"Mark Hammond" <mhammond at skippinet.com.au> writes:
> But I can't log on.  How can I determine what my account has been reset
> to?  Im getting _lots_ of enquiries about my starship page, so Im keen to
> get it back online ASAP...

We haven't yet set up new passwords for people, though those who use
SSH can already log in.  I think we've figured out *how* they'll be
assigned; now it's just a matter of implementing the required script.

--amk


From abrahams at mediaone.net  Thu Apr 20 01:27:11 2000
From: abrahams at mediaone.net (Dave Abrahams)
Date: Thu, 20 Apr 2000 05:27:11 GMT
Subject: __del__ problem - would adopting Garbage Collection fix this?
References: <aw2L4.8907$HG1.245743@nnrp1.uunet.ca> <7ojL4.9424$HG1.254543@nnrp1.uunet.ca> <Z%oL4.9643$HG1.257533@nnrp1.uunet.ca> <m31z41iyhg.fsf@atrus.jesus.cam.ac.uk> <38FE7F1E.168E914A@cosc.canterbury.ac.nz><slrn8ft35a.uqs.nascheme@cranky.arctrix.com>
Message-ID: <B5240D22.4AFD%abrahams@mediaone.net>

in article slrn8ft35a.uqs.nascheme at cranky.arctrix.com, Neil Schemenauer at
nascheme at enme.ucalgary.ca wrote on 4/19/00 11:50 PM:

> Well, I guess I should. :)
> 
> When garbage is found by the collector (ie. reference cycles not
> reachable from Python), all objects reachable from __del__
> methods are moved into a separate set.  This is what I call
> uncollectable garbage.  Uncollectable instances are currently
> added to a global list of uncollectable objects reachable from
> within Python.  If the programmer creates such garbage it is up
> to them to deal with it.

Hmm. If the objects are really garbage (i.e. unreferenceable by the
programmer), then there really is nothing she can do to deal with it. Of
course you can't call the __del__ method, but it seems you should be able to
reclaim the memory occupied by the objects without losing anything.

-Dave



From kcazabon at home.com  Tue Apr 18 00:06:13 2000
From: kcazabon at home.com (Kevin Cazabon)
Date: Tue, 18 Apr 2000 04:06:13 GMT
Subject: read/write file problem
References: <38FB1375.2FF65B96@sec.noaa.gov>
Message-ID: <VaRK4.147211$Dv1.1584521@news1.rdc1.bc.home.com>

On a simpler note:  the Python Imaging Library has full JPEG support.  I'd
suggest one of two things:

-use PIL instead of trying to do it yourself,
OR
-see the PIL implementation for JPEG for helping hints.

Kevin.


"j vickroy" <jvickroy at sec.noaa.gov> wrote in message
news:38FB1375.2FF65B96 at sec.noaa.gov...
> Hello all,
>
> I am attempting to read and, subsequently, write a jpeg file.
> The following interactive session illustrates the problem I am having
> (i.e., if the source file is read byte-by-byte into memory and
> subsequently
> rewritten using the file.write() method, only the initial portion of the
>
> file appears to be written).
> How should I be doing this in Python 1.5.2.
>
> Thanks for your time.
>
>
> >>> import os.path
> >>> file_name = 'd:/p1010001.jpg'
> >>> source_byte_cnt = os.path.getsize (file_name)
> >>> print source_byte_cnt
> 381965
> >>> buf = []
> >>> file = open (file_name, 'r')
> >>> for i in xrange (0,source_byte_cnt):
> ...  buf .append (file .read(1))
> ...
> >>> file .close()
> >>> print len (buf) == source_byte_cnt
> 1
> >>> file_name = 'd:/test.jpg'
> >>> file = open (file_name, 'w')
> >>> for i in xrange (0,source_byte_cnt):
> ...  file .write (buf[i])
> ...
> >>> file .close()
> >>> target_byte_cnt = os.path.getsize (file_name)
> >>> print target_byte_cnt
> 70
> >>> file = open (file_name, 'w')
> >>> for i in xrange (0,source_byte_cnt):
> ...  file .write (str(buf[i]))
> ...
> >>> file .close()
> >>> target_byte_cnt = os.path.getsize (file_name)
> >>> print target_byte_cnt
> 70
> >>>
>
>




From m.faassen at vet.uu.nl  Wed Apr 26 14:16:25 2000
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 26 Apr 2000 18:16:25 GMT
Subject: Python Time classes
References: <slrn8ge5b0.1gu.philh@vision25.demon.co.uk>
Message-ID: <8e7bpp$r2s$2@newshost.accu.uu.nl>

phil hunt <philh at vision25.demon.co.uk> wrote:
> Is there a high-level Time class for Python?

[snip]
> Does this exist, or shall I write one?

There's something like that in Zope. I don't know if fits your
requirements completely, but you may want to take a look at it.

<zope_directory>/lib/python/DateTime

Regards,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From aa8vb at yahoo.com  Mon Apr  3 06:43:35 2000
From: aa8vb at yahoo.com (Randall Hopper)
Date: Mon, 03 Apr 2000 06:43:35 -0400
Subject: Tkinter:  "Get root for a widget" is private?
In-Reply-To: <QdEF4.892$n68.190327808@newsb.telia.net>
References: <20000331112048.A1257432@vislab.epa.gov>
 <QdEF4.892$n68.190327808@newsb.telia.net> <m3snx3n1x0.fsf@devel.no.rwd.dinet>
 <20000331112048.A1257432@vislab.epa.gov>
 <QdEF4.892$n68.190327808@newsb.telia.net>
Message-ID: <20000403064335.A1560771@vislab.epa.gov>

Fredrik Lundh:
 |Randall Hopper wrote:
 |> I have a widget and want to retrieve the root.  I see the _root() method,
 |> but by Python convention, isn't this method private?
 |>
 |> Should users of Tkinter be calling this?
 |
 |as long as nobody calls the Tkinter police, you're safe.

Well as Python doesn't have enforced data hiding, I try to honor the
conventional access rules.  _<symbol> is private, an implementation detail,
and thus less likely to be present in future versions than <symbol>.

 |    root = widget.winfo_toplevel()
 |    if root.master:
 |        root = root.master # real root

Rob W. W. Hooft:
 |Hm. Except that my Toplevel windows are often nested deeply, and
 |certainly not restricted to children of the root! I think that your
 |approach is distinctly different from _root() except if you severely
 |restrict your toplevel window creation.

Given this it seems we should have a public root() method.  Could we
add a:

    root = _root

definition to Tkinter's Misc class, to make this a public feature?

Thanks,

Randall

-- 
Randall Hopper
aa8vb at yahoo.com



From jimsp at ichips.intel.com  Tue Apr 18 14:10:29 2000
From: jimsp at ichips.intel.com (jimsp)
Date: Tue, 18 Apr 2000 11:10:29 -0700
Subject: How do I catch a Ctrl-Break in an InteractiveInterpreter?
Message-ID: <38FCA515.93EA319A@ichips.intel.com>

I am trying to build a replacement for a test framework that provides
an interactive command shell, providing variables, expressions, control
flow
statements as well as test specific commands. I would like to build a
new
system based on python and the main command language ( that way I don't
have
to maintain the general language core ) combined with C extensions ( I
Love
Swig!) to implement the test specific commands.

I have run into a snag, using the InteractiveInterpreter class I can
catch the KeyboardInterrupt exception and restart the shell just fine,
*But* a
Ctrl-Break still causes the python interpreter to exit. Does anyone have
any
suggestions?

--james




From tescoil at rtpro.net  Mon Apr  3 16:40:27 2000
From: tescoil at rtpro.net (Tesla Coil)
Date: Mon, 03 Apr 2000 15:40:27 -0500
Subject: calendar.py
Message-ID: <38E901BB.A41CEB25@rtpro.net>

Looking at calendar.py and wondering if there's any reason
mdays, day_name, day_abbr, month_name, and month_abbr
wouldn't be better off tuples than lists.  I don't see that they
are anywhere in the module the objects of operations specific
to lists, and there doesn't seem to be any advantage having
them mutable sequences.

Who maintains calendar.py?  I'd like to propose inclusion
of functions to facilitate such calculations as "what will the
date be in 45 days," or "print dates for twelve 28 day cycles
beginning from 3/1/2000."   Excuse the undercooked code,
but something like the following:

mdays = (0, 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31)

def isleap(year):
    return year % 4 == 0 and (year % 100 <> 0 or year % 400 == 0)

# lacks handling targets in excess of 365 (or 366) days
# eventually return, not print, results.
def day2date(target, year):
    for month in range(1, 13, 1):
        if isleap(year) and month==2 and target<=29:
            print month, target
            break
        elif target <= mdays[month]:
            print month, target
            break
        elif isleap(year) and month==2:
            target=target-29
        else:
            target=target-mdays[month]

# lacks handling of nonexistent dates, e.g., 2,29,2001
# eventually return, not print, results.
def date2day(month, day, year):
    def add(x,y): return x+y
    total=reduce(add, mdays[:month])
    total=total+day
    if isleap(year) and month > 2:
        total=total+1
    print total



From scherbi at bam.com  Mon Apr 17 11:34:03 2000
From: scherbi at bam.com (Bill Scherer)
Date: Mon, 17 Apr 2000 11:34:03 -0400
Subject: Another question
References: <dVFK4.41$vr2.5817@news.uswest.net>
Message-ID: <38FB2EEB.A32AF88C@bam.com>

Jeff -

The main reason (AFAIK) you don't want to do 'from <mod> import *' is to
avoid namespace clashing.
A better way is this:

#untested
import random
randint = random.randint

for i in xrange(1,1000000):
    print randint(1,10)

#

This method allows you to control exactly which names you import into
your local namespace form the module.  The above should also be a little
faster due to the use of xrange.  Se the docs for why.

-Bill

Jeff Massung wrote:

> Okay, another quick question. If it is quicker to do:
>
> import random
> from random import randint
> for i in range(1,1000000):
>     print randint(1,10)
>
> than this:
>
> import random
> for i in range(1,1000000):
>     print random.randint(1,10)
>
> Why would I not do "from <module> import *" with every module I load?
>
> Jeff
>
> --
> http://www.python.org/mailman/listinfo/python-list




From hernanf at my-deja.com  Sun Apr 16 14:24:38 2000
From: hernanf at my-deja.com (Hernan M. Foffani)
Date: Sun, 16 Apr 2000 18:24:38 GMT
Subject: Doesn't anybody write to paper anymore?
References: <38F5FDF3.EE2FBA01@tcac.net> <38F94701.CA52FDBB@tcac.net>
Message-ID: <8dd0gq$crn$1@nnrp1.deja.com>

In article <38F94701.CA52FDBB at tcac.net>,
  Albert Wagner <alwagner at tcac.net> wrote:
>......Just one follow-up:  The
> docs say that os.popen() is "unreliable on Windows."  Is this still
true
> in 1.5?  And if so, is there an alternate implementation for
windows?  I
> run on Redhat6.1 Linux at home but need to port to windows98 for work.

Well, there is a reliable "os.popen()" for windows around there. But if
you want to print in windows from python, that will be useless. Unless
you've got a "lpr" for windows (?)
The quick way: output to a text file, then run "notepad /p filename".
The advanced way: use the win32 module from pyhton.

-H.

--
Hern?n Mart?nez Foffani
hernanf at my-deja.com


Sent via Deja.com http://www.deja.com/
Before you buy.


From wware at world.std.com  Thu Apr 13 00:52:49 2000
From: wware at world.std.com (Will Ware)
Date: Thu, 13 Apr 2000 04:52:49 GMT
Subject: Need an enum type extension
References: <38F52A7A.D8841724@gte.net>
Message-ID: <Fsxuw1.EFn@world.std.com>

sue (smalleys at gte.net) wrote:
> Has anyone done an enum implementation?

One C-like idiom I've seen looks like this:

( ALPHA, BRAVO, CHARLIE, DELTA ) = range(4)

-- 
 - - - - - - - - - - - - - - - - - - - - - - - -
Resistance is futile. Capacitance is efficacious.
Will Ware	email:    wware @ world.std.com


From someone at somewhere.com  Tue Apr 18 11:44:03 2000
From: someone at somewhere.com (Charley Horse)
Date: Tue, 18 Apr 2000 15:44:03 GMT
Subject: ASP web-type stuff
References: <RZyG4.3061$9o.5812531@news.magma.ca> <8cfea6$71$1@gxsn.com> <c2LG4.3228$9o.6039280@news.magma.ca>
Message-ID: <MPG.136623215f8755e19896ab@news>

[This followup was posted to comp.lang.python and a copy was sent to the 
cited author.]

Did you make any progress with this project? I'm a total python newbie, 
I started a thread called "Using python on the web" yesterday after 
realizing that learning python itself isn't enough - that CGI or 
something is required in order to use python in a web app. I am more 
familiar with the CF/ASP/PHP way of doing things, and know nothing at 
all of CGI. I still have no idea how I'd acutally use python to power an 
interactive website, any more than I would if I was using BASIC for a 
language. So your idea sounds like something I can relate to, and 
perhaps other neophytes could too. I wonder what would be lost, if 
anything, with this approach. I'm not sure enough of python to even 
begin to evaluate this.

And if you have any links up your sleeve that talk about using python to 
run interactive site on the web, I'd love to see them. Seems like a very 
strange blank spot on the map.

In article <c2LG4.3228$9o.6039280 at news.magma.ca>, 
dgreen at kryten.kitenet.net says...
> Ah. I suppose I should have made myself more clear. What I meant was that
> I was interested in writing a CGI/FastCGI type script so that the same
> code could be used under Apache/Netscape Enterprise/insert your favorite
> unix web server here. I already have a prototype working under Apache with
> a CGI redirect. I'd like to write interfaces to httpdapy and PyApache as
> well.
> 
> In article <8cfea6$71$1 at gxsn.com>, Richard Chamberlain wrote:
> >Hi Dan,
> >
> >ASP is not language dependent, most of the stuff you'll see is written in
> >VBScript or JScript but you can actually configure it to use other
> >languages.
> >
> >See http://starship.python.net/crew/pirx/asp/py_asp.html to show you how to
> >use Python with ASP.
> >
> >Richard
> >
> >"Dan Green" <dgreen at kryten.kitenet.net> wrote in message
> >news:RZyG4.3061$9o.5812531 at news.magma.ca...
> >> Ahoy, folks.
> >> I was looking over things like Embperl and Apache::ASP and liked the
> >> idea. I would like the idea more if I could do with python. So I had a
> >> couple of questions:
> >>
> >> Is there such a thing?
> >> Is there interest in such a thing?
> >> Does anbody want to hear about my 75 lines of code hack that basically
> >does
> >> the trick, but needs a lot of work?
> >>
> >> If nobody cares at all, I'll just shut up and go home. :)
> >>
> >> For those of you who have no idea what I'm talking about, it's basically
> >an
> >> alternative to CGI that allows you to embed the code right into the HTML.
> >>
> >> something like this:
> >>
> >> <HTML>
> >> <BODY>
> >> <%  for i in range(5): %>
> >> Loop item <%= i %><BR>
> >> </BODY>
> >> </HTML>
> >>
> >> would look like this when called from a browser:
> >>
> >> <HTML>


From sholden at bellatlantic.net  Sun Apr  2 14:33:06 2000
From: sholden at bellatlantic.net (Steve Holden)
Date: Sun, 02 Apr 2000 18:33:06 GMT
Subject: Python 1.6 alpha 1 released
References: <Pine.LNX.4.10.10003311651590.22919-100000@starship.python.net> <00033115131700.03666@quadra.teleo.net>
Message-ID: <38E7922D.52166F1D@bellatlantic.net>


Patrick Phalen wrote:
> 
> [Vladimir Ulogov and Bjorn Pettersen's posts quoted]
> 
> Well, these two posts have scratching my head. What sort of
> environment are you two (Vladimir & Bjorn) working in that makes it
> seem so onerous to fix your sock.connect() syntax (considering that
> you literally have months to prepare for it)?
> 
Probably one that likes to move forwards rather than hang around
fixing regression-testing errors induced by language changes :-)

> After all, this isn't like the multi-argument append issue, which defies
> even Tim Peters' estimable skills with regular expressions to arrive at
> a surefire "nanny" treatment. We're just talking about a simple string
> replacement here!
> 
Sadly all the cases Tim quoted which made multi-argument append hard
to detect (especially function assignment) apply here too.  While many
cases will be detectable by string scanning, it's not possible to detect
all possible cases easily.

> In my own case, I administer around 12 servers, geographically
> dispersed around the U.S, plus the workstations in my office. All have
> various Python scripts running on them. Long ago, I hacked a simple
> string replacement script in Python, which takes a directory and two
> strings as arguments and which walks that directory tree replacing one
> string with the other. I have this script on all 12 servers, since I use
> it all the time.
> 
I am sure this is useful.  But even when you've fixed all the cases you
find, surely you have to at least apply a set of standard regression tests
to ensure compatibility?

> Now, I expect it will take me all of 15-20 minutes, total, to
> find and "fix" any such problems on these 12 servers. I know I'll need
> to also fix some other situations where long int "L" is manually
> truncated. Big deal. Since I have 2-3 months warning on this, I feel
> fairly confident I am up to this challenge. In fact, I expect it
> will take me less time to fix sock.connect() than to reply to your
> posts. ;-)
> 
Very gung-ho!

> I must be missing something. Is your situation somehow much more
> complex than that? If so, I'd be curious to know how and why.
> 
I can't comment on others' situations.  Fortunately I'm new enough to
Python that I don't have major systems depending on the language ... yet.

> And as for asking Guido to change his mind ... heh, that was a little
> joke, right?

Seems to me that a legitimate concern has been raised.  There have been
responses which complained about the *tone* of the original posts, but
the bottom line is that this change will require fixes to installed code.
Alpha is surely a better time to discuss this than when the final release
comes along?

Maybe someone could borrw the time machine and visit the future to let us
know how this one is going to come out? :-)

regards
 Steve
--
"Bulding information systems just because it's fun."
Steve Holden  sholden at bellatlantic.net  703 716 7275


From effbot at telia.com  Sat Apr  1 17:00:59 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Sat, 01 Apr 2000 22:00:59 GMT
Subject: New Features in Python 1.6
References: <000301bf9c15$23a18f20$182d153f@tim>
Message-ID: <vkuF4.4584$74.73299@newsc.telia.net>

now posting from Belgium, Tim Peters wrote:
> What we've seen here is simply Guido *finally* embracing a
> policy of giving people exactly what they ask for.

Tim,

I think things are worse than this -- someone just sent me
something that looks like a draft press release, announcing
that O'Reilly had taken over Python development, with Jon
Orwant as CTO, and Larry Wall as the Python 1.6 release
manager.

(it wasn't clear if it was intended for public consumption
at this time, so please don't spread the word around just
yet.  if I get permission, I'll post more info to the list).

given this, Guido's post makes a lot more sense, in a slightly
scary way -- the press release also talks about fusing Python
3000 and Perl's "Topaz" project, rewriting the whole thing in
Scheme.

note that this seems to have been in the work for quite a
while; things like Perl companies hiring python talent, Tom
Christiansen's Python courses [1], Guido schmoozing with
Randal Schwartz on the SD2000 conference [2], the Casbah
folks working on a unified Python/Perl engine, etc.

still think it's kinda weird, though...

</F>

1) http://training.perl.com/open-courses.html
2) http://www.ddj.com/articles/2000/0050/0050toc.htm




From lvw at lfpt.rwth-aachen.de  Fri Apr 28 02:49:22 2000
From: lvw at lfpt.rwth-aachen.de (Lars von Wedel)
Date: Fri, 28 Apr 2000 08:49:22 +0200 (MET DST)
Subject: Using BaseHTTPServer
Message-ID: <200004280649.IAA00894@kazul.rwth-aachen.de>



Hello,

my CGI project suffers from a lengthy delay for starting up
python and loading all required modules. I'm thinking about
setting up a small HTTP server in python itself so that the
interpreter and all the required modules have to be loaded only
once.

>From the module documentation it is not clear to me, what I actually
have to do to catch the GET calls arriving at the server.

I assume that I can use BaseHTTPServer as it is. I just need
to implement a working request handler, right?

Now comes the more fuzzy part:

Since I do not want to execute external files but internal methods,
I think I could modify CGIHTTPServer by subclassing it and overriding
its do_GET() method to decode the request directly? Or would I rather
use SimpleHTTPServer, here?

Within any of the do_GET methods, how do I access data that has been
sent by the client (...&data1=val1?data2=val2?...)?

Thanks,
Lars






From ullrich at math.okstate.edu  Wed Apr  5 11:59:07 2000
From: ullrich at math.okstate.edu (David C. Ullrich)
Date: Wed, 05 Apr 2000 10:59:07 -0500
Subject: Why should I switch to Python? - Infinity of Primes
References: <000d01bf9ec5$3fb90820$3e2d153f@tim>
Message-ID: <38EB62CB.A36DD25F@math.okstate.edu>


Tim Peters wrote:

> [Greg Ewing]
> > ...
> > If someone actually came up with a formula for constructing
> > primes, it would be rather large news -- isn't that one of
> > the Big Unsolved Problems?
>
> Several such formulas were developed in the 70's, although they're strained.

    "Strained" indeed. I've only seen one - something about the integer
part of some power of a certain real number x. The strained part
being that you don't know what x actually is, you need to know
infinitely many primes to "construct" x in the first place.

> Paulo Ribenboim's accessible "The Little Book of Big Primes"
> (Springer-Verlag, 1991) gives a bunch of 'em in chapter 3, along with 6
> other proofs of the infinitude of the primes.
>
> not-that-[AP]-will-be-fooled-ly y'rs  - tim

    You want to be careful mentioning that name. Seriously - you
think having me around is a pain, it could be much worse...

DU



From gmcm at hypernet.com  Wed Apr  5 10:41:20 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Wed, 5 Apr 2000 10:41:20 -0400
Subject: Two questions
In-Reply-To: <38EAA002.D82AFAB8@callware.com>
Message-ID: <1257176812-9948710@hypernet.com>

Ivan Van Laningham wrote:
> Gordon McMillan wrote:
[snip]
> > At the moment, you'll have a bit of trouble with my stuff and
> > the fact that Marc-Andre puts his .pyds inside the packages.
> > There are workarounds, and I have a patch from Thomas Heller
> > which I haven't finished reconciling to some of my other
> > changes.
> > 
> 
> How nasty are the workarounds?  Do they make life annoying for the guy
> who writes the installs?  (Acceptable--possibly even desirable.) 
> Annoying for me?  (Irritating but acceptable.)  Or infuriating for our
> customers?  (Unacceptable.)

A bit of pain for you, but not nearly enough to make me 
happy. The problem is that there's a mxDateTime directory 
and a mxDateTime.pyd.

Here's how Calishar solved it:

> The way I ended up solving it was renaming the mxDateTime directory to
> something else, and copying the mxDateTime.pyd file to the dlls
> directory under (pythonroot).

[Ivan] 
> Of course, if there is a way to make life unbearable for the salesmen
> (and they are all men), that would be ideal.

You were made for each other, I'm sure.

> > > <floccinausinihilipilification>-ly y'rs,
> > > Ivan
> > 
> > Are you describing your hairline, your nervous twitch, or the
> > effect of the latter on your former?
> > 
> 
> C'mon, Gordon.  Get out your OED!  I _know_ you've got one.

Sorry, I don't. I find that "flocci" could be related to the term for 
the hairy stuff growing on vegetable matter, or from a synonym 
for carphology. Either way, I suspect your hopeless quest for 
verisimilitude has you attempting to emulate male pattern 
baldness...

- Gordon



From timehorse at unforgettable.com  Mon Apr 24 21:28:42 2000
From: timehorse at unforgettable.com (timehorse at unforgettable.com)
Date: Tue, 25 Apr 2000 01:28:42 GMT
Subject: Starship News Brief
Message-ID: <8e2sc2$qto$1@nnrp1.deja.com>

To help Andrew out since he is busy at work on the stuff for which he
gets paid, I thought I'd just send off a couple lines to update the
folks on the Starship.python.net status, the long and harrowing saga of
which the recent chapter of the April [Y2K] disk crash will soon be
coming to a close, for better or worse.

Now, for this I have two items, one good an one bad, and what do you
guys want first?

Okay, since no-one spoke up, I'll tell the good first, as it's good for
a little softening up before the bad.  :)

So the good news is, it looks like the 'Mysterious Entity' that may
have stepped in to pay for the old disk recovery will have much less to
pay towards that end...

The reason for this is, THE DISK IS DEAD, LONG LIVE THE DISK!  Yep, the
recovery place could not restore the ond disk, so we'll have to stick
with that old 1 February 2000 backup and hope that folks didn't do too
much work since then.

Of course, this means it is SAFE TO UPDATE now, since there will be no
further updates to the Starship apart from getting the rest of the non-
mission critical systems working such as Telnet and Zope.  Stay tuned
for more information as it appears, and Andrew will no doubt return to
give an update on the final missing services later this week.

Be Seeing You,

Jeffrey.

P.S. Yes, they ARE considering adding MAPS or ORBS to the Starship!
Woo hoo!


Sent via Deja.com http://www.deja.com/
Before you buy.


From tismer at tismer.com  Fri Apr  7 15:52:37 2000
From: tismer at tismer.com (Christian Tismer)
Date: Fri, 07 Apr 2000 21:52:37 +0200
Subject: No threads in MacPython 1.5.2?
References: <8cl7rh$kvm$3@flotsam.uits.indiana.edu>
Message-ID: <38EE3C85.713C1D85@tismer.com>


"Gregory L. Hansen" wrote:
> 
> My Python can find Tkinter now, thanks for the help.  But now it can't
> find the module named "thread".  And yet there it is in the Lib folder,
> thread.py and threading.py, and the search path $(PYTHON):Lib is set for
> it.
> 
> Now what's wrong?

The operating system is wrong. It has no threads.
But watch out: (and see a recent post)

Just "Stackless" v. Rossum is right now building threads
with Stackless Python for the Mac.

and-they-work-with-win2000-as-well -ly y'rs - chris

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaunstr. 26                  :    *Starship* http://starship.python.net
14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     where do you want to jump today?   http://www.stackless.com



From dworkin at ccs.neu.edu  Sat Apr 22 18:27:07 2000
From: dworkin at ccs.neu.edu (Justin Sheehy)
Date: 22 Apr 2000 18:27:07 -0400
Subject: LAmbda calculus evaluator (interactive)
References: <8dro55$eq6$1@nnrp1.deja.com> <vndk8hq3t1z.fsf@camelot-new.ccs.neu.edu> <8dt0cq$nte$1@nnrp1.deja.com>
Message-ID: <vnd8zy54vr8.fsf@camelot-new.ccs.neu.edu>

Andrew Cooke <andrew at andrewcooke.free-online.co.uk> writes:

> ok, is this one of those pedantic geek jokes?

No, it was intended to be helpful, in a rather obvious way.  

It is clear that you wanted something more, but I was answering what
looked like a stand-alone question about the interactiveness of ocaml
in the midst of your post.

> ...but if not, i thought the (unquoted) context made clear that i was
> talking about programs that manipulated expressions in lambda calculus
> (so a "not interactive" program simply takes an expression and follows
> some evaluation strategy as far as possible; an "interactive"
> program allows me to step reduction by reduction, do abstractions, etc).
> in other words, manipulate the appropriate ast.

Ah.  Well, that's not at all what most people that I know mean by the
distinction between "interactive" and "non-interactive" when referring to
a language implementation.

If I refer to an interpreter (or compiler, whatever) as "interactive",
I mean simply that it has an read-eval-print loop that one can
interact with.

The standard Python implementation is like this, as is ocaml (as shown
in my previous response to you) and several other languages.  Many
other languages (C++, Perl, many others) typically do not have an
interactive interpreter.  One can refer to their implementations as
non-interactive.

With regard to the lambda calculus debugger you seem to be looking
for, I'm not sure exactly what you'd want it to look like.  Even if I
did, I don't know if I'd be able to direct you to one.

However, to re-answer the sub-question that I thought I answered
before: ocaml is (or at least can be) interactive.

<wink>

-Justin

 



From alander at nortelnetworks.com  Tue Apr 11 14:51:17 2000
From: alander at nortelnetworks.com (Anthony Lander)
Date: 11 Apr 2000 18:51:17 GMT
Subject: Trouble with Tkinter under IDLE
Message-ID: <8F139460Canthonylander@47.80.11.122>

Hi -

I'm having some trouble with Tkinter windows running under Idle, and I 
wonder if any of you have encountered this:

When I open a window from the Idle shell, it won't return from the mainloop 
when the window closes. Further, the shell won't respond because it 
believes that there's a thread running (and it can't kill it). 

All of this works just fine from the command line, though, so I'm pretty 
convinced it's a problem related to running Idle. 

I'm running Python 1.5.2, idle 0.6 and tcl/tk 8.0 all under NT.

Anyone have any guesses?

Thanks much,

Anth


From ckotso at cs.ntua.gr  Wed Apr 26 07:39:51 2000
From: ckotso at cs.ntua.gr (Constantinos A. Kotsokalis)
Date: 26 Apr 2000 11:39:51 GMT
Subject: ungreedy regular expressions
References: <slrn8gdlaj.k0s.ckotso@theseas.softlab.ece.ntua.gr>
Message-ID: <slrn8gdlc7.lt2.ckotso@theseas.softlab.ece.ntua.gr>

On 26 Apr 2000 11:39:00 GMT, Constantinos A. Kotsokalis wrote:
>[...]

Forgot to mention this, Python version is 1.5.2.

-- 
fsck your brains


From gresham at mediavisual.com  Fri Apr  7 05:32:05 2000
From: gresham at mediavisual.com (Paul Gresham)
Date: Fri, 7 Apr 2000 17:32:05 +0800
Subject: CGI Help Needed
Message-ID: <8ck9uo$ga08@imsp212.netvigator.com>

Hi All,
I have plenty of experience of programming, networks etc, but when it comes
to Brower/Web server protocols I'm pretty lost. I have to set up a
subscriber database, and wish to do so using Python. I can do everything I
need to, I even have Python working on my server as a CGI, it was a breeze
to get working. I can capture input to a form, parse and store it, I can
modify the data via the browser, etc, etc. My question is probably not
Python specific, but as we are by our Pythonness, enlightened programmers, I
thought I'd ask here first.

What is the normal and respected method of User authentication on a web
site, how should I present a login box when the user is not authenticated
already, and how do I maintain a session with an authenticated user, without
using cookies, allowing them to continue surfing around the web site ?

Answers or pointers to relevant information will be most appreciated

Paul Gresham
** Please do not Spam, I have no egg and chips **




From chafik at nevrax.com  Tue Apr  4 03:45:34 2000
From: chafik at nevrax.com (Sameh chafik pro)
Date: Tue, 4 Apr 2000 09:45:34 +0200
Subject: builtding library .pyd file format
Message-ID: <009101bf9e09$c2d20020$1001a8c0@nevrax.net>

Hello evry body,
I want to know how to do to built a .pyd library file format form the windows .dll file format
thanks

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20000404/5fdc4f85/attachment.html>

From willadams at aol.com  Wed Apr 12 14:48:42 2000
From: willadams at aol.com (William F. Adams)
Date: 12 Apr 2000 18:48:42 GMT
Subject: Compiling Python Imaging under LinuxPPC
References: <FU%I4.646$rc9.190630912@newsb.telia.net>
Message-ID: <20000412144842.06030.00002524@ng-ch1.aol.com>

When I attempt to compile the Sketch .src.rpm, the compilation fails because
Imaging.h is not where it's supposed to be.

I ddin't see much in the way of documentation with the .src.rpm for PIL, sorry.

William


-- 
William Adams
http://members.aol.com/willadams
Sphinx of black quartz, judge my vow.



From gmcm at hypernet.com  Sun Apr 23 16:48:40 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Sun, 23 Apr 2000 16:48:40 -0400
Subject: setting up Tkinter
In-Reply-To: <KlHM4.8299$4w1.119156@typhoon.nyroc.rr.com>
Message-ID: <1255599575-47242831@hypernet.com>

Andy Mullen wrote:

> As i understand from the website (www.python.com), this program doesn't run
> on windows 98. Since i was already thinking of putting linux on my machine
> under a partition of the HD, i was wondering if someone could point me in
> the right direction as to where to get linux online, how to go about
> install, and how to run the 2 separate OS's on one machine.  any response
> would be appreciated

You probably picked up a warning about a problem with the 
1.6 alpha. Use the release version (1.5.2) and look at the FAQ 
if you have problems.

- Gordon



From spenrose at my-deja.com  Fri Apr 21 14:28:34 2000
From: spenrose at my-deja.com (spenrose at my-deja.com)
Date: Fri, 21 Apr 2000 18:28:34 GMT
Subject: Using python on the web
References: <EDFD2A95EE7DD31187350090279C6767754BA8@THRESHER>
Message-ID: <8dq6k3$qmb$1@nnrp1.deja.com>

In article <EDFD2A95EE7DD31187350090279C6767754BA8 at THRESHER>,
  Brent Fulgham <brent.fulgham at xpsystems.com> wrote:
> A couple of questions:
>
> > We build medium-busy, CGI-intensive sites in Python. We were briefly
> > quite excited about Zope, but once we realized it locked us into Yet
> > Another Programming Language, we lost interest. DHTML is
> > particularly
> > problematic,
> 1.  I'm not familiar with Zope.  Is DHTML specific to Zope?  It seems
> to be proudly advertised on a lot of "webware" brochures I see these
> days.


My limited understanding of Zope is that to run Zope is to use DHTML and
vice-versa. I welcome corrections on this point.

> > The biggest hassle for us has been the weak MySQL support in
> > Python. The spec keeps changing and the interface is raw.
>
> 2.  Have you tried using PostgreSQL?  MySQL may be fast, but it's
> really little more than a filesystem represented as a database.
> It has no transaction support whatsoever which tends to lead to
> problems when you have a community-style site that has multiple
> parties trying to add messages, etc.  Have you had much success
> using it in a production environment?  I realize Slashdot and I
> think Sourceforge do, but they have their share of troubles.
>
> Of course, the Postgres Python bindings may be just as bad from your
> perspective.  :-)
>
> Thanks,
>
> -Brent
>
If we were going to look at a transaction-enabled system it would
probably be Oracle, since as you note there are hassles with everything.
We might as well get a marketing buzzword for our troubles. But then, I
don't really know....


Sent via Deja.com http://www.deja.com/
Before you buy.


From effbot at telia.com  Wed Apr 26 16:37:26 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Wed, 26 Apr 2000 20:37:26 GMT
Subject: PIL compile/make bombs out at step #7 -- help?
References: <20000426135724.A26663@exeter.exeter.org>
Message-ID: <asIN4.4302$Za1.64185@newsc.telia.net>

Karl G. Ulbrich <karl at ulbrich.org> wrote:
> The install bombs out in step #7 of the readme,
> which is mentioned as a problem spot... however,
> I modifed Setup.in to point to /usr/include,
> and checked to see that yes, there's a python1.5
> directory in there and yes, it has lots of .h files.

from the look of it, you either have a broken /bin/sh,
or you messed something up when editing the Setup
file...

can you perhaps post the Setup file ?

</F>




From poupaerte at my-deja.com  Sat Apr  8 16:26:20 2000
From: poupaerte at my-deja.com (poupaerte at my-deja.com)
Date: Sat, 08 Apr 2000 20:26:20 GMT
Subject: tabular data
Message-ID: <8co4l2$5r2$1@nnrp1.deja.com>

Hi

I would like to start effectively using Python, instead of just looking
at it once in a while. From what I have read in the manuals and so, I
still can't make up how to handle tabular data properly with Python.
Tabular data and their permutations are the mainstay of the
applications I'm usually involved in.

For example, say that I have the following tabular data: [(2,1,"REC02"),
(1,3,"REC03"),(3,1,"REC01")], and I want to obtain: [(1,3,"REC03"),
(2,1,"REC02"),(3,1,"REC01")], that is, the tabular data should be
sorted on the first and then the second column.

How can I have Python perform an in-place sort on a combination of
several columns?
By the way, should I represent rows by tuples or by embedded lists?

Another problem: to visually represent/edit tabular data, I need an
editable grid widget (with possibly comboboxes as cell editors).
TkInter does not seem to have an appropriate grid widget.


Sent via Deja.com http://www.deja.com/
Before you buy.


From kragen at dnaco.net  Sat Apr 22 23:07:53 2000
From: kragen at dnaco.net (Kragen Sitaker)
Date: Sun, 23 Apr 2000 03:07:53 GMT
Subject: code not true?
References: <8ds3ku$q0l$1@nnrp1.deja.com> <M3iM4.518$wJ1.17829@ptah.visi.com> <NhiM4.1704$v85.10013@news-server.bigpond.net.au> <slrn8g45lj.oa.daniels@localhost.localdomain>
Message-ID: <dOtM4.1076$RM6.1566545@news-east.usenetserver.com>

In article <slrn8g45lj.oa.daniels at localhost.localdomain>,
Alan Daniels <daniels at mindspring.com> wrote:
>On Sat, 22 Apr 2000 14:02:21 GMT, Mark Hammond
><mhammond at skippinet.com.au> wrote:
>
>[snip...]
>>Of all the solutions to this problem, this does strike me as one of
>>the best - simply remove equality testing al together for floating
>>point.
>
>Are you sure of this? It would be awful confusing for a programming
>newbie to do:
>
>x = 5.0
>y = 5.0 # Later on down in the code.
>x == y  # And even further...
>
>And get an exception thrown. All the "why doesn't this work?" messages
>we've seen about floating point have to do with doing manipulation of
>the numbers and *then* having them not work.
>
>For 0.07 * 100 to not be the same as 7.0 is understandable, because
>that's just how floating point works on CPUs today: Not all numbers
>are representable in IEEE format. But to outlaw asking if 7.0 is equal
>to 7.0 definitely goes against the principle of least suprise. And
>learning that floating-point math sometimes leads to slight numerical
>errors is an important lesson that newbie programmers shouldn't be
>shielded from.

Almost the only time I care whether two floating-point numbers are
exactly equal is when one of them is exactly zero.  But if I'm throwing
around floor() and ceil(), I might well want to know whether one of
them is exactly an integer.

The Perl way to handle this would be to have an enableable flag that
prints a warning when two numbers that are almost, but not quite equal,
are compared for equality.
-- 
<kragen at pobox.com>       Kragen Sitaker     <http://www.pobox.com/~kragen/>
The Internet stock bubble didn't burst on 1999-11-08.  Hurrah!
<URL:http://www.pobox.com/~kragen/bubble.html>
The power didn't go out on 2000-01-01 either.  :)


From effbot at telia.com  Tue Apr 18 07:55:23 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Tue, 18 Apr 2000 11:55:23 GMT
Subject: Forward References?
References: <eviltofu-4C5688.21593817042000@news.earthlink.net> <m3itxgcggc.fsf@atrus.jesus.cam.ac.uk> <eviltofu-4B17FA.07361018042000@news.earthlink.net>
Message-ID: <L2YK4.445$wYl.183079424@newsb.telia.net>

Jerome Chan <eviltofu at rocketmail.com> wrote:
> > > class a:
> > >    def x(self):
> > >       y(self)
> > >    def y(self):
> > >       print "Hello"
> > >
> > > c = a()
> > > c.x()
> > > c.y()
> > >
> > > gives a NameError: y
> >
> > Your mistake isn;t what you think it is, try this instead:
> >
> > class a:
> >    def x(self):
> >       self.y()
> >    def y(self):
> >       print "Hello"
> >
> > HTH, HAND,
> > M.
>
> Thanks! I also found out that
>
> a.y() works!

really?

Traceback (most recent call last):
  File "broken.py", line 8, in ?
  File "broken.py", line 3, in x
TypeError: unbound method must be called with class instance 1st argument

but maybe you meant "a.y(self)"?

it's not the same thing, and breaks down the day you
start to use inheritance.  as usual, python works best
if you use it to write python code...

</F>




From mwh21 at cam.ac.uk  Mon Apr 17 22:04:35 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 18 Apr 2000 03:04:35 +0100
Subject: Forward References?
References: <eviltofu-4C5688.21593817042000@news.earthlink.net>
Message-ID: <m3itxgcggc.fsf@atrus.jesus.cam.ac.uk>

Jerome Chan <eviltofu at rocketmail.com> writes:

> Just a silly question but how do I do forward references?
> 
> 
> class a:
>    def x(self):
>       y(self)
>    def y(self):
>       print "Hello"
> 
> c = a()
> c.x()
> c.y()
> 
> gives a NameError: y

Your mistake isn;t what you think it is, try this instead:

class a:
   def x(self):
      self.y()
   def y(self):
      print "Hello"

HTH, HAND,
M.

-- 
112. Computer Science is embarrassed by the computer.
     -- Alan Perlis, http://www.cs.yale.edu/~perlis-alan/quotes.html


From effbot at telia.com  Fri Apr 14 16:05:03 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Fri, 14 Apr 2000 20:05:03 GMT
Subject: [ANNOUNCE] Garbage collection for Python
References: <20000407221750.A29279@acs.ucalgary.ca> <m3r9cagx3c.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <PRKJ4.1078$rc9.191689728@newsb.telia.net>

Michael Hudson <mwh21 at cam.ac.uk> wrote:
> To someone who might know: what hope is there for this to go
> into 1.6?

not speaking for Guido here, but I suspect that the accelerated
1.6 schedule means that it's too late.

but I'm pretty confident that this (or something very similar) will
go into 1.7.

> It would be, in my opinion, be a really really really good thing; for
> one thing it would help in language wars...

sorry, but stackless python is the only really really really good
thing that hasn't yet been added to the core.  this is just really
really good.

</F>




From pete at visionart.com  Wed Apr 19 20:11:54 2000
From: pete at visionart.com (Pete J Shinners)
Date: Wed, 19 Apr 2000 17:11:54 -0700
Subject: creating simple threaded socket server
Message-ID: <38FE4B4A.5987EA0A@visionart.com>

still new to python, and this is my first " > hello,world"
project. eventually the project will be a simplish chat-type
server with some extra enhancements. i know, it's been done
to death, but my learning python is the main goal here, not
breaking any new ground.

i'm starting with a simple server that listens on a socket
and spawns threads to handle incoming connections. it is
working, but i'd strongly like to get some peer review at
this point.

am i headed down the road to destruction? does this look
like a positive start? i am also wondering if there is a
better way to 'unblock' my calls like socket.accepts without
silly(??) workarounds like i have with my unplug() method

anyways, like i said, the server does work, you can telnet
into it and it simple echoes the strings it gets as a list.
sending "Quit" will disconnect your thread, sending "Exit"
will shutdown the server (as best i know how)

i've looked all around for some good examples to follow
but could find none (nothing on vaults or starship) so
i had to do a lot of trailblazing for myself. i also started
with 'SocketServer', but was having difficulty, and decided
to "simplify" things be going lower level? perhaps now that
i have a grasp on things i should move back up to SocketServer?

here's the code, all and any insight is welcome!!
i've been running in interactive mode...
import myserver
myserver.test()

--------------------------------myserver.py---
# simple(??) threaded server in python

from socket import *
import string
import threading
import time

networkactive = 1;




class NetworkListener(threading.Thread):
    def __init__(self, socket, address, manager):
	threading.Thread.__init__(self, None)
	self.socket = socket
	self.address = address
	self.manager = manager
	self.start()
    
    def __del__(self):
	self.manager = None
    
    def run(self):
	global networkactive
	print 'Connected from ', self.address
	self.socket.send('Welcome\n')
	while 1:
	    data = string.strip(self.socket.recv(1024))
	    if not data or data == 'Quit': break
	    if data == 'Exit': networkactive = 0; break
	    self.socket.send(`list(data)`+'\n')
	print 'Disconnecting ', self.address
	self.socket.shutdown(2)
	self.socket.close()





class PortManager(threading.Thread):
    def __init__(self, port):
	threading.Thread.__init__(self, None)
	self.port = port
	self.start()
	
    def run(self):
	global networkactive
	s = socket(AF_INET, SOCK_STREAM)
	s.bind('', self.port)
	s.listen(2)
	while networkactive:
	    sock, addr = s.accept()
	    if networkactive:
		NetworkListener(sock, addr, self)
	s.close()
	print 'SOCKETS DONE'

    def unplug(self):
	"""this will unstick the thread from accept"""
	s = socket(AF_INET, SOCK_STREAM)
	s.connect('', self.port)
	s.close()
	print 'Unsticking ACCEPT call'


 
def test():
    server = PortManager(6016)
    while networkactive:
	time.sleep(2)
	print 'Threads: ', threading.enumerate()
    server.unplug()

----------------------------------------------------


From gmcm at hypernet.com  Tue Apr 25 10:06:05 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Tue, 25 Apr 2000 10:06:05 -0400
Subject: Question about exception-handling mechanism
In-Reply-To: <8e46mt$4nh$1@nnrp1.deja.com>
Message-ID: <1255450930-56183831@hypernet.com>

willfg at my-deja.com wrote


> ...A colleague asked why in an exception handling mechanism
> you'd want the ELSE block to be executed if you don't throw an
> exception as opposed to a FINALLY block. Anyone used this feature
> in practice? Thanks in advance for your input, -- Will

A finally clause is *always* executed, so it's not really in 
competition with a try / else. If everything goes fine, these are 
equivalent:

try:
  do_something()
  do_morestuff()
except NoBananasException:
  yesweaint()

try:
  do_something()
except NoBananasException:
  yesweaint()
else:
  do_morestuff()

Now we expect do_something() to sometimes throw a specific 
exception, and we're prepared to deal with it. But we can't deal 
with the same exception coming from do_morestuff(). So the 
first snippet may cause a problem by "recovering" from the 
wrong error. That's the rationale for try / else, (essentially, we 
don't have to use a flag variable).

- Gordon



From hathawa2 at marshall.edu  Thu Apr 20 09:16:31 2000
From: hathawa2 at marshall.edu (Mark Hathaway)
Date: Thu, 20 Apr 2000 09:16:31 -0400
Subject: closures
Message-ID: <38FF2069.1F25@marshall.edu>

Hi all,

I've been discussing OO with Smalltalkers on comp.object. The
issue/topic
of blocks (as used in Smalltalk) comes up frequently. I am wondering how
Python handles 'if's and 'for's which is where Smalltalk commonly uses
it's blocks. How does Python do it?

if <condition>:
    <statements>

for each in list:
    <statements>



Mark S. Hathaway
email: hathawa2 at marshall.edu


From PClaerhout at CREO.BE  Thu Apr 13 05:26:15 2000
From: PClaerhout at CREO.BE (Pieter Claerhout)
Date: Thu, 13 Apr 2000 11:26:15 +0200
Subject: Python CGI running commands under NT
Message-ID: <2B1262E83448D211AE4B00A0C9D61B03BA72B1@MSGEURO1>

Are there any other ways to accomplish this?

Kind regards,


Pieter
Pieter_Claerhout at creoscitex.com


-----Original Message-----
From: Paul Robinson [mailto:paul.robinson at quantisci.co.uk]
Sent: Thursday, April 13, 2000 10:54 AM
To: python-list at python.org
Subject: Re: Python CGI running commands under NT



Pieter Claerhout wrote:
> 
> Hello,
> 
> I made a Python CGI-script that executes a DOS-command and
> should return the result in a HTML page. The script looks
> as follows:
> 
>   import os
> 
>   pipe = os.popen("ipconfig", 'r')
>   text = pipe.read()

Pieter,
By default IIS does not create a console associated with a CGI process.
This means that the pipe has nowhere to read from!

See: http://support.microsoft.com/support/kb/articles/Q194/9/48.ASP

For some info on how you can change this behaviour. I've tried this in
the past with limited success (and I also found it hard to turn it off
again!).

	Paul.


-----------------------------------
Business Collaborator Team
Enviros Software Solutions
http://www.businesscollaborator.com
-----------------------------------

-- 
http://www.python.org/mailman/listinfo/python-list



From gansevle at cs.utwente.nl  Wed Apr 19 03:19:39 2000
From: gansevle at cs.utwente.nl (Fred Gansevles)
Date: Wed, 19 Apr 2000 07:19:39 GMT
Subject: nisplus -- help needed
References: <38F5AF5A.168E7C71@cs.utwente.nl>
Message-ID: <8djmm1$ndv$1@nnrp1.deja.com>

In article <38F5AF5A.168E7C71 at cs.utwente.nl>,
  I <gansevle at cs.utwente.nl> wrote:
> A few weeks ago, Guido asked me to write a NIS+ module for python.
> I wrote the original NIS module (that's why he asked me) and I
> know how
> NIS works.
> The past few weeks I've been reading documentation about NIS+ and I
> *think* I know
> how to implement a nisplus extention module, but before I start to
> implement this
> module, I want to 'feel' a NIS+ environment.
>
> My question to the python community is:
>
>     can anyone help me to get *access* to a NIS+ domain?
>
> Adding me and my system to such a domain woul help me a lot.
> Having an account on a computer in a NIS+ domain would also be very
> usefull.
>
> So, are there any NIS+ administrators out there who can help me  (and
> the python community as a whole)?

Any system administrator with systems in a NIS+ domain, willing
to give me access to such a system (preferable Linux, Solaris is fine
too) helps me a lot.
With an 'ordinary' account I can start on the read-only-access
implementation of nisplus. If a NIS+ administrator grants me more rights
I can also implemment the read-write-access implementation of nisplus.


> Thanks in advance.
>
> P.S.: Please mail your replies, since I don't have the time to
> regulary read the c.l.p. newsgroup.

Posting is o.k. too since "Deja Tracker" sends me an e-mail if someone
responds.

-----------------------------------------------------------------------
----          Linux/Windwos-NT/IntraNetware Administrator          ----
-- Whenever it sounds simple, I've probably missed something! ... Me --
-----------------------------------------------------------------------


Sent via Deja.com http://www.deja.com/
Before you buy.


From mitch.harris at usa.net  Thu Apr  6 17:44:33 2000
From: mitch.harris at usa.net (mitch.harris at usa.net)
Date: Thu, 06 Apr 2000 21:44:33 GMT
Subject: help: build fails on sun in make
Message-ID: <8cj0ft$okg$1@nnrp1.deja.com>



Full_Name: mitch harris
Version: 1.6
OS: sun 5.6
Submission from: nat135022.ericy.com (208.237.135.22)


This is the first time I tried to install Python.
*************************************

/usr2/t_mitchh/python/Python-1.6a1>version
Machine hardware:   sun4u
OS version:         5.6
Processor type:     sparc
Hardware:           SUNW,Ultra-1

The following components are installed on your system:


Sun Visual WorkShop C++ 3.0
        Sun WorkShop Compiler C 4.2
        Sun WorkShop Compiler C++ 4.2
        Sun WorkShop Tools.h++ 7.0
        Sun WorkShop Tools.h++ 6.0.4
        Sun WorkShop Visual 2.0
        Sun WorkShop IPE 4.0
        Sun WorkShop CodeManager 2.0
        Sun WorkShop Distributed Make 2.0
        Sun WorkShop FileMerge 3.0
        Sun WorkShop FreezePoint 2.0
        Sun WorkShop Maketool 2.0
        Sun WorkShop VersionTool 2.0
        Sun WorkShop Dbx 4.0
        Sun WorkShop Performance Analyzer 4.0
        Sun WorkShop LoopTool 2.1
        Sun WorkShop LockLint 2.1
        Sun WorkShop Thread Analyzer 1.2
        Sun WorkShop XEmacs 20.00

/usr2/t_mitchh/python/Python-1.6a1>./configure
creating cache ./config.cache
checking MACHDEP... sunos5
checking for --without-gcc... no
checking for --with-cxx=<compiler>... no
checking for gcc... gcc
checking whether the C compiler (gcc  ) works... yes
checking whether the C compiler (gcc  ) is a cross-compiler... no
checking whether we are using GNU C... yes
checking whether gcc accepts -g... yes
checking LINKCC... $(PURIFY) $(CC)
checking LDLIBRARY...
checking for ranlib... ranlib
checking for ar... ar
checking how to run the C preprocessor... gcc -E
checking for AIX... no
checking for minix/config.h... no
checking whether gcc accepts -OPT:Olimit=0... no
checking whether gcc accepts -Olimit 1500... no
checking for C preprocessor type... ansi
checking for ANSI C header files... yes
checking for dlfcn.h... yes
checking for fcntl.h... yes
checking for limits.h... yes
checking for locale.h... yes
checking for ncurses.h... no
checking for pthread.h... yes
checking for signal.h... yes
checking for stdarg.h... yes
checking for stddef.h... yes
checking for stdlib.h... yes
checking for thread.h... yes
checking for unistd.h... yes
checking for utime.h... yes
checking for sys/audioio.h... yes
checking for sys/file.h... no
checking for sys/lock.h... yes
checking for sys/param.h... no
checking for sys/select.h... yes
checking for sys/time.h... yes
checking for sys/times.h... yes
checking for sys/un.h... yes
checking for sys/utsname.h... yes
checking for sys/wait.h... yes
checking for dirent.h that defines DIR... yes
checking for opendir in -ldir... no
checking for clock_t in time.h... yes
checking for mode_t... yes
checking for off_t... yes
checking for pid_t... yes
checking return type of signal handlers... void
checking for size_t... yes
checking for uid_t in sys/types.h... yes
checking size of int... 4
checking size of long... 4
checking size of void *... 4
checking size of char... 1
checking size of short... 2
checking size of float... 4
checking size of double... 8
checking for long long support... yes
checking size of long long... 8
checking size of off_t... 4
checking whether to enable large file support... no
checking for --with-next-framework... no
checking for --with-dyld... no
checking SO... .so
checking LDSHARED... ld -G
checking CCSHARED...
checking LINKFORSHARED...
checking for dlopen in -ldl... yes
checking for shl_load in -ldld... no
checking for t_open in -lnsl... yes
checking for socket in -lsocket... yes
checking for socket in -lnet... no
checking for --with-libs... no
checking for --with(out)-readline... not specified.
checking for --with-dec-threads... no
checking for --with-threads... no
checking for --with-thread... no
checking for --with-sgi-dl... no
checking for --with-dl-dld... no
checking for dlopen... yes
checking DYNLOADFILE... dynload_shlib.o
checking for alarm... yes
checking for chown... yes
checking for clock... yes
checking for confstr... yes
checking for ctermid... yes
checking for ctermid_r... yes
checking for dlopen... (cached) yes
checking for execv... yes
checking for flock... no
checking for fork... yes
checking for fsync... yes
checking for fdatasync... no
checking for fpathconf... yes
checking for ftime... yes
checking for ftruncate... yes
checking for getgroups... yes
checking for getlogin... yes
checking for getpeername... yes
checking for getpgrp... yes
checking for getpid... yes
checking for getpwent... yes
checking for gettimeofday... yes
checking for getwd... yes
checking for kill... yes
checking for link... yes
checking for lstat... yes
checking for mkfifo... yes
checking for mktime... yes
checking for nice... yes
checking for pathconf... yes
checking for pause... yes
checking for plock... yes
checking for pthread_init... no
checking for putenv... yes
checking for readlink... yes
checking for select... yes
checking for setgid... yes
checking for setlocale... yes
checking for setuid... yes
checking for setsid... yes
checking for setpgid... yes
checking for setpgrp... yes
checking for setvbuf... yes
checking for sigaction... yes
checking for siginterrupt... yes
checking for sigrelse... yes
checking for strftime... yes
checking for strptime... yes
checking for symlink... yes
checking for sysconf... yes
checking for tcgetpgrp... yes
checking for tcsetpgrp... yes
checking for tempnam... yes
checking for timegm... no
checking for times... yes
checking for tmpfile... yes
checking for tmpnam... yes
checking for tmpnam_r... yes
checking for truncate... yes
checking for uname... yes
checking for waitpid... yes
checking for fseek64... no
checking for fseeko... yes
checking for fstatvfs... yes
checking for ftell64... no
checking for ftello... yes
checking for statvfs... yes
checking for dup2... yes
checking for getcwd... yes
checking for strdup... yes
checking for strerror... yes
checking for memmove... yes
checking for getpgrp... (cached) yes
checking for setpgrp... (cached) yes
checking for gettimeofday... (cached) yes
checking whether time.h and sys/time.h may both be included... yes
checking whether struct tm is in sys/time.h or time.h... time.h
checking for tm_zone in struct tm... no
checking for tzname... yes
checking for time.h that defines altzone... yes
checking whether sys/select.h and sys/time.h may both be included... yes
checking whether char is unsigned... no
checking for working const... yes
checking for inline... inline
checking for working volatile... yes
checking for working signed char... yes
checking for prototypes... yes
checking for variable length prototypes and stdarg.h... yes
checking for bad exec* prototypes... no
checking for bad static forward... no
checking whether va_list is an array... no
checking for gethostbyname_r... yes
checking gethostbyname_r with 6 args... no
checking gethostbyname_r with 5 args... no
checking gethostbyname_r with 3 args... no
checking for __fpu_control in -lieee... no
checking for --with-fpectl... no
checking for --with-libm=STRING... default LIBM="-lm"
checking for --with-libc=STRING... default LIBC=""
checking for hypot... yes
checking for hypot... (cached) yes
checking for genuine getopt... yes
checking what malloc(0) returns... nonnull
checking for wchar.h... yes
checking for usable wchar_t... no
checking whether byte ordering is bigendian... yes
checking for --with-wctype-functions... no
updating cache ./config.cache
creating ./config.status
creating Makefile
creating Objects/Makefile
creating Parser/Makefile
creating Python/Makefile
creating Modules/Makefile.pre
creating Modules/Setup.thread
creating config.h


/usr2/t_mitchh/python/Python-1.6a1/Modules>make
gcc  -g -O2 -I./../Include -I.. -DHAVE_CONFIG_H -c ./posixmodule.c
In file included from ./posixmodule.c:3194:
/usr/include/sys/statvfs.h:38: parse error before `fsblkcnt_t'
/usr/include/sys/statvfs.h:38: warning: no semicolon at end of struct or
union
/usr/include/sys/statvfs.h:39: warning: data definition has no type or
storage
class
/usr/include/sys/statvfs.h:40: parse error before `f_bavail'
/usr/include/sys/statvfs.h:40: warning: data definition has no type or
storage
class
/usr/include/sys/statvfs.h:41: parse error before `f_files'
/usr/include/sys/statvfs.h:41: warning: data definition has no type or
storage
class
/usr/include/sys/statvfs.h:42: parse error before `f_ffree'
/usr/include/sys/statvfs.h:42: warning: data definition has no type or
storage
class
/usr/include/sys/statvfs.h:43: parse error before `f_favail'
/usr/include/sys/statvfs.h:43: warning: data definition has no type or
storage
class
/usr/include/sys/statvfs.h:51: parse error before `}'
/usr/include/sys/statvfs.h:51: warning: data definition has no type or
storage
class
./posixmodule.c: In function `posix_fstatvfs':
./posixmodule.c:3207: storage size of `st' isn't known
./posixmodule.c: In function `posix_statvfs':
./posixmodule.c:3259: storage size of `st' isn't known
make: *** [posixmodule.o] Error 1

I tried this to fix it by undefining the definitions
*********************************************************************

#if
defined(HAVE_FSTATVFSxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
x)
#if
defined(HAVE_STATVFSxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
xx)
#ifdef
HAVE_FSTATVFSxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
#ifdef
HAVE_STATVFSxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

and got this
**********************************************************************

/prj/qed2/local//lib/gcc-lib/sparc-sun-solaris2.5/2.7.2.1/include/sys/pa
ram.h:187:
warning: `NBBY' r
/usr/include/sys/select.h:45: warning: this is the location of the
previous
definition
In file included from /usr/include/sys/stream.h:26,
                 from /usr/include/netinet/in.h:38,
                 from /usr/include/netdb.h:96,
                 from ./socketmodule.c:163:
/usr/include/sys/model.h:32: #error "No DATAMODEL_NATIVE specified"
make[1]: *** [socketmodule.o] Error 1
make[1]: Leaving directory `/usr2/t_mitchh/python/Python-1.6a1/Modules'
make: *** [Modules] Error 2

Does anyone know what I am doing wrong and how I should build python?
*********************************************************************

Mitch

mitch.harris at usa.net



Sent via Deja.com http://www.deja.com/
Before you buy.


From gmcm at hypernet.com  Sat Apr 22 09:10:08 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Sat, 22 Apr 2000 09:10:08 -0400
Subject: Pythonware Website???? Gone on vacation with Starship? ;-))
In-Reply-To: <8dqp6u$plm$1@newshost.accu.uu.nl>
Message-ID: <1255713486-40391297@hypernet.com>

Martijn Faassen wrote:

> Gordon McMillan <gmcm at hypernet.com> wrote:
> > Martijn Faassen wrote:
> 
> >> .... Also, since the PSU has access to

[snip]

Why hello Martijn! It's amazing that you have the exact same 
name as another c.l.py poster - one we all remember fondly.


- Gordon



From bjorn at roguewave.com  Thu Apr 27 15:10:21 2000
From: bjorn at roguewave.com (Bjorn Pettersen)
Date: Thu, 27 Apr 2000 13:10:21 -0600
Subject: HTML parser example, anybody?
References: <slrn8gh2a9.oj.haering@sunhalle8.informatik.tu-muenchen.de>
Message-ID: <3908909D.4022A846@roguewave.com>

Gerhard Haering wrote:
> 
> Hi!
> 
> I want to transform my own HTML templates (with a few template tags,
> special comments, and so on) into final HTML for the browser. I would like
> to use Pyhton's htmllib, but from the docs I can get no clue on how to get
> it done. I seem to have to use a writer and a formatter class ...
> 
> I would greatly appreciate an example of HTML -> HTML using
> htmllib/sgmllib. Can anybody provide pointers/code snippets?

Why don't you just use Python's string syntax, e.g. if you have the
following in a template file called foo.templ:

   <html><head><title>%(title)s</title></head>
   <body><h1>%(heading)s</h1></body></html>

you can do something like:

   htmlfile = open('foo.html', 'w')
   replacements = {
     'title':   'my title',
     'heading': 'A Heading',
   }
   htmlfile.write( open(foo.templ).read() % replacements )
   htmlfile.close()

-- bjorn



From jimc at regnoc.com  Sun Apr  2 15:54:18 2000
From: jimc at regnoc.com (Jim Conger)
Date: Sun, 02 Apr 2000 19:54:18 GMT
Subject: HTML Form to Python Script Problem
Message-ID: <38E796BC.23191B2@regnoc.com>

I'm having trouble getting a simple form to launch a python script on my
RH6.1 system.  Python runs fine alone, but I don't get any action when I
launch a script.  Here is the HTML form:

<form method=post action="test1.py">
<input type=text name=field1 size=50 value="Edit me">
<br>
<input type=submit>
</form>

The python script is in the same directory as the HTML file and looks
like:

#!/usr/bin/python
import cgi, sys
sys.stederr = sys.stdout
print "Content-type: text/html"
print
print "<H1>Heading</H1>"
pinrt "text line here."

The script is set to be executable by all and runs fine alone.  However,
nothing happens when I click the submit button on the HTML form.  Any
clues?

--
Jim Conger
Regnoc Software Inc.
jimc at regnoc.com, www.regnoc.com


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20000402/c06b99f8/attachment.html>

From mfletch at tpresence.com  Tue Apr 18 09:40:41 2000
From: mfletch at tpresence.com (Mike Fletcher)
Date: Tue, 18 Apr 2000 09:40:41 -0400
Subject: RE question
Message-ID: <CC59868A0A76D311B1F50090277B321C1F327B@VRTMAIL>

Your regex matches any letter of [abc] at least once, stores that as the
group result, then goes back and tries again.  The last result the group
gets is c...

[abc] -> a
() -> store a
+ -> repeat group
[abc] -> b
() -> store b
+ -> repeat group
[abc] -> c
() -> store c
+ -> repeat group
[abc] -> stop
() -> stop
return, with c as result for group 1

Re's don't allow for storing multiple group match results for the same
group.  For that, get yourself a parsing system :) .

try:
	([abc]+) -> abc
except MikeyMissingPoint:
	shake head in amazement at him

HTH,
Mike


-----Original Message-----
From: Benyang Tang [mailto:btang at pacific.jpl.nasa.gov]
Sent: Monday, April 17, 2000 7:24 PM
To: python-list at python.org
Subject: RE question




>>> m = re.match("([abc])+", "abc"); m.group(1)
'c'

Why is it 'c', instead of 'a'?
-- 
http://www.python.org/mailman/listinfo/python-list



From raffael at mediaone.net  Sun Apr 23 23:57:00 2000
From: raffael at mediaone.net (Raffael Cavallaro)
Date: Mon, 24 Apr 2000 03:57:00 GMT
Subject: The Simple Economics of Open Source
References: <B527E7A5.51AA%raffael@mediaone.net> <MPG.136cc01af4ab703b989709@news.earthlink.net>
Message-ID: <raffael-4899DC.23562823042000@news.ne.mediaone.net>

In article <MPG.136cc01af4ab703b989709 at news.earthlink.net>, Richard 
Hale <rkhale at earthlink.net> wrote:

>The Open Source software that I have used has been very valuable to me.  
>Did those who produced it not see as much value in it?

Well, they probably knew they couldn't sell it as a closed source 
product and make much money from it.

Fact is, only commodity items, that are known to many programmers, are 
open sourced. If there were sufficient demand to support a closed source 
business, then it would have happened that way. The proof of this is the 
BSDs, which would allow a closed source, proprietary fork at any time. 
Hasn't happened 'cause there's no demand for such a product.

The closed source biggies are not open sourcing the crown jewels. Why? 
Cause there's plenty of cash still left to milk from those cows. Notice 
that Apple has open sourced only those parts of MacOS X that *aren't* 
lucrative consumer software. Where's the Darwin QuickTime (not the 
streaming server, but the client?) Where's Darwin Quartz, Darwin Aqua, 
Darwin Cocoa? Don't hold your breath for these, because these parts of 
the OS are what allows Apple to turn a profit. If they became open 
source commodities, Apple would become just another clone maker.

Of course this is even more true for MS. If they open sourced their 
stuff, the only thing they'd have to sell is support. Ever looked at the 
margins MS has on their software? Ever seen the margins that support 
businesses have? Guess which one MS would rather have.

Open source will only ever have commodity items for the simple reason 
that anything truly unique can command a much higher profit margin than 
support and service, which is what you really sell if you're in the open 
source *business* (as opposed to hobby).

It's just simple supply and demand. Unique items have (by definition) 
only one supplier, so prices can ge high - supply is fixed. Commodity 
items have (by definition) many suppliers, so if one supplier raises his 
prices, customers just move to another equivalent vendor - supply is 
abundant.

-- 

Raffael Cavallaro, Ph.D.
raffael at mediaone.net


From cdaniel at Level3.net  Thu Apr  6 15:50:44 2000
From: cdaniel at Level3.net (Cliff Daniel)
Date: Thu, 6 Apr 2000 15:50:44 -0400
Subject: CLI over Sockets?
In-Reply-To: <20000406100529.A13830@xs4all.nl>
References: <20000406002732.D1027@shell2.xcom.net> <20000406100529.A13830@xs4all.nl>
Message-ID: <20000406155044.E1027@shell2.xcom.net>

> It's dead easy to write such a tool though: look at telnetlib for some good
> examples. I ended up overriding the interact() method to add my logging, but
> you still get all the protocol stuff for free. (telnet is more than just a
> TCP socket, you see.) I'd guess that it isn't too much trouble to add
> readline and your own completer function (if necessary, that last bit) but
> i'm not sure if readline works properly with the select() method interact
> uses... you might want to use mt_interact or something similar instead.

   telnetlib provides no helpful information in implementing a daemon
base CLI in which someone uses 'telnet' to connect.  What I'm looking
to implement is just like Cisco's IOS CLI.  The socket code is trivial
but but the vt100 emulation, to me anyway, is non-trivial.  Getting the
client into character mode is a cinch, but implementing the command
line editting features (left arrow, insert, tab, ?, etc...) seems to be
a little more involved.  Basic interaction within the sockets is just
a function of recv()/send() but with that you lack any useful cli
functions.

Regards,
Cliff



From mwh21 at cam.ac.uk  Wed Apr 19 17:15:23 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 19 Apr 2000 22:15:23 +0100
Subject: __del__ problem - would adopting Garbage Collection fix this?
References: <aw2L4.8907$HG1.245743@nnrp1.uunet.ca> <7ojL4.9424$HG1.254543@nnrp1.uunet.ca> <Z%oL4.9643$HG1.257533@nnrp1.uunet.ca>
Message-ID: <m31z41iyhg.fsf@atrus.jesus.cam.ac.uk>

"Warren Postma" <embed at geocities.com> writes:

> My questions:
> 
> 1. Shouldn't freeing modules in Py_Finalize be in the reverse order
> of the way in which the modules were created?  Could a
> module-destruction order be explicitly maintained by the system?

It's not clear that that would help; I'd bet I could come up with a
counter example if I was feeling twisted enough.

> 2. Is this convincing proof, or not, that Garbage Collection would
> be a Good Thing.

I don't think so.  You still have to destroy things; true gc might
help I suppose, but I'm fairly sure it doesn't solve the problem
completely.

> Is JPython immune to these bizarre behaviours?

Well, I don;t think JPython implements __del__ methods, so yes, but
not in a helpful way...

> 3. What about adding a __finalize__ method to objects.  Python could
> then finalize all objects first, and then delete them, and the rule
> is you ONLY EVER USE __del__ when you Absolutely Have to, and with
> some known issues (such as this). In other words, all objects are
> finalized in Py_Finalize, then destroyed. When an object is deleted
> with del(obj), it is finalized first, then freed.

But this would help the problem; if a file object has a finalize
method, and some other object wants to write to one of these file
objects in *its* finalize method... what you need is some way of
spelling "finalize me before him", and that's what we have already
(it's just a bizarre way of spelling it).

> 4. Until such point as there is a __finalize__, I am actually going
> to explicitly call a finalization method, then delete objects. Side
> effects on Destructors can arguably be a Bad Thing.

Surely a side-effect-free destructor is one of the most pointless
concepts possible?

I-think-the-reason-I-don't-know-a-neat-solution-is-because
-there-isn't-one-ly y'rs
M.

-- 
39. Re graphics:  A picture is worth 10K  words - but only those 
    to describe the picture. Hardly any sets of 10K words can be 
    adequately described with pictures.
     -- Alan Perlis, http://www.cs.yale.edu/~perlis-alan/quotes.html


From jepler.lnk at lnk.ispi.net  Mon Apr 24 22:54:21 2000
From: jepler.lnk at lnk.ispi.net (jepler epler)
Date: Tue, 25 Apr 2000 02:54:21 GMT
Subject: Pickling Robustly
References: <jak8huf1of.fsf@helix.twistedmatrix.com> <8dpk1i$7de$1@newshost.accu.uu.nl> <jazoqn2s0b.fsf@helix.twistedmatrix.com> <slrn8g4fs9.33l.jepler.lnk@potty.housenet> <ja3dobqnvm.fsf@helix.twistedmatrix.com>
Message-ID: <slrn8ga28f.1ol.jepler.lnk@potty.housenet>

On 24 Apr 2000 14:50:53 -0500, Glyph Lefkowitz
 <glyph at twistedmatrix.com> wrote:
>It would still be nice if there were an option to register an error
>handler to just ignore attributes which couldn't be pickled, rather
>than failing.  At this point though, I think I'd need to rewrite the
>module to get that.

This tickled another memory of mine, back when I was fiddling with
"POO", a Python-based multi-user game.  I discovered that anybody with
programming ability could add a non-picklable type to a POO object and
prevent the database from being dumped (i.e., saved) ... At the time, I
modified the pickle.py module to store a None when a non-picklable
object was encountered.  I seem to recall that Guido came up with a
superior solution, but unfortunately I can't seem to find it on deja.com
(I'm pretty sure the discussion was on the newsgroup)

Ah, here it is.  If you are using the pickle.py module (as opposed to
the cpickler), you can subclass pickle.Pickler and override the 'save'
method.  In poo, the following is done:

class PooPickler(pickle.Pickler):

        """CLASS PooPickler:
        This class extends the normal Pickler object such that datatypes
        which can't by pickled are set to None, rather than crashing
        the server.
        """

        # RADICALLY CHANGED on 4/20/98 to work with Python 1.5,
        # as per recommendation by our fearless leader (Guido)
        def save(self, object):
                try:
                        pickle.Pickler.save(self, object)
                except pickle.PicklingError:
                        pickle.Pickler.save(self, None)

This doesn't seem to be documented in the manpage for the pickle
module, but the documentation makes it clear that the cPickle.Pickler is
not a class and cannot have its save method overridden in the same way.

Rather than checking for a PicklingError, you probably want to use
type(object) is BlahType  or  isinstance(object, klass) .. but you can
decide that for yourself.

Good luck! I hope my answer is closer to what you're looking for this
time around.

Jeff


From nmm1 at cus.cam.ac.uk  Mon Apr  3 10:09:24 2000
From: nmm1 at cus.cam.ac.uk (Nick Maclaren)
Date: 3 Apr 2000 14:09:24 GMT
Subject: Why should I switch to Python?
References: <000201bf9c14$0a79d9e0$182d153f@tim> <38E7967E.DE18EFA5@bellatlantic.net>
Message-ID: <8ca8mk$4kf$1@pegasus.csx.cam.ac.uk>

In article <38E7967E.DE18EFA5 at bellatlantic.net>, Steve Holden <sholden at bellatlantic.net> writes:
|> 
|> Perl's avowed intention that everything which *might* mean something in
|> the language *should* mean something makes the odds much higher that an
|> incorrect program won't be caught until execution.  This means that for
|> a learner much more time has to be spent testing code to find out just
|> how Perl interprets what *has* to be regarded as a pretty grotty syntax.

What makes you think that it would be caught then?

The first Perl program I wrote was 20-30 lines long and, as is my
wont, I tested every single branch and boundary value.  Everything
finally worked.  I then tried it on real data, and it gave wrong
answers, quietly.

I had misunderstood the syntax description and written incorrect
syntax that just happened to behave in the same way as the correct
syntax on all my tests.  You may be able to guess why I don't
trust Perl for anything critical.

Oh, and by the way, I was the guy who got Perl 4 going under MVS,
and have ported Perl 5 to Unix a couple of times.  I may make rude
remarks about the quality of the C and its design in Python, but
it is wonderful compared to that in Perl.  It is a miracle Perl
works at all.

I am learning Python because I need a scripting language that is
much more flexible and reliable and less feature-ridden than Bourne
shell - I could stick with C, of course, but it is very tedious
even for an old Assembler hacker like me.


Regards,
Nick Maclaren,
University of Cambridge Computing Service,
New Museums Site, Pembroke Street, Cambridge CB2 3QG, England.
Email:  nmm1 at cam.ac.uk
Tel.:  +44 1223 334761    Fax:  +44 1223 334679


From smalleys at gte.net  Sat Apr 15 12:51:29 2000
From: smalleys at gte.net (sue)
Date: Sat, 15 Apr 2000 16:51:29 GMT
Subject: Python paradigms
References: <8cmr70$b0e$1@pegasus.csx.cam.ac.uk> <slrn8f4be7.1t0.bwinton@tor.dhs.org> <8ctc9m$sp2$1@pegasus.csx.cam.ac.uk> <8ctkm5$bko$2@newshost.accu.uu.nl> <8cur12$5pf$1@pegasus.csx.cam.ac.uk> <8cv91s$q2i$1@newshost.accu.uu.nl> <slrn8fdfiu.v9q.spin@Dizzy.vals>
Message-ID: <38F89015.4A359A8F@gte.net>

I think Python is insanely readable.  Yesterday, I had to
revive a package of tools I wrote in 1995 (Python 1.3+) and
hadn't touched since 1998.  It took two hours.  A coworker
is started working on his old Perl-based part of the system
Monday, and isn't done yet.

Sue.


From dalke at acm.org  Mon Apr  3 01:38:35 2000
From: dalke at acm.org (Andrew Dalke)
Date: Sun, 2 Apr 2000 23:38:35 -0600
Subject: new method for string objects
References: <8c44e8$5hj$1@slb6.atl.mindspring.net> <xDnF4.4507$74.72019@newsc.telia.net>
Message-ID: <8c9aj6$u49$1@slb6.atl.mindspring.net>

Fredrik Lundh wrote:
>assuming that "string value" means the "raw contents of the
>sequence object", you should use a buffer objects instead.
>you can use the 'buffer' builtin to turn most sequence objects
>into buffers:


I had never heard of 'buffer' before.  I looked into it a bit.
It seems like it only works with C based structures as there's
no way to get the buffer of an object.

I have a "Seq" class which looks like:

class Seq:
  def __init__(self, letters, alphabet = Alphabet.generic_alphabet):
    assert type(letters) == type("")  # input must be a string
    self.letters = letters
    self.alphabet = alphabet
  def tostring(self):
    return self.letters

and I have a translation function which looks like:

def translate(seq, table = TranslationTable.standard_table):
  s = seq  # or "s = seq.tostring()"
  letters = []
  for i in range(0, len(s), 3):
    letters.append(table[s[i:i+3]])
  return Seq(string.join(letters, table.forward_alphabet))

If the input is a string or an array.array, then I can
replace the "s = seq" with your buffer suggestion.

  s = buffer(seq)

However, if the sequence is a Seq object, then buffer won't
work.  (I added a __getattr__ to the Seq object to see if
the buffer() gets an attribute from an instance; it doesn't.)

On the other hand, if strings have a "tostring" method, then I
can always use:

  s = seq.tostring()

                    Andrew Dalke
                    dalke at acm.org





From robin at jessikat.demon.co.uk  Mon Apr 24 10:09:27 2000
From: robin at jessikat.demon.co.uk (Robin Becker)
Date: Mon, 24 Apr 2000 15:09:27 +0100
Subject: The Simple Economics of Open Source
References: <B527E7A5.51AA%raffael@mediaone.net> <3903316C.5D6F1611@tcac.net>
 <raffael-34658F.23400223042000@news.ne.mediaone.net>
 <390451C0.4B5D3196@tcac.net>
Message-ID: <LHEoAJAXWFB5EwF6@jessikat.demon.co.uk>

In article <390451C0.4B5D3196 at tcac.net>, Albert Wagner
<alwagner at tcac.net> writes
>You STILL don't understand.  Artifice can only copy.  It cannot invent.
>An end product can be copied, even a style can be copied, but not the
>creative process itself.
>
>Raffael Cavallaro wrote:
...
This position is like the weak AI position; you assert there's something
magical about creation. I think A Turing was right; any human behaviour
can be simulated and the best simulators (people) are around already.
-- 
Robin Becker


From tsmets at altern.org  Wed Apr 12 11:15:42 2000
From: tsmets at altern.org (Thomas SMETS)
Date: Wed, 12 Apr 2000 15:15:42 GMT
Subject: No method overloadin I suppose ?
Message-ID: <38F4938E.8F245B68@altern.org>


## Beginning of Python program ...
class Lister:
  def __init__(self):
    print 'In constructor ...'
    __init__(1)

  def __init__(self, lValue):
    print 'In constructor with parameter'
    self.l = lValue

  def __repr__(self):
    print '\n******\n\tValue is : ', l, '\n********'

if __name__ == "__main__":
  x = Lister()
  y = Lister (2)
## End of python program

Trace of execution is :
Traceback (innermost last):
  File "Minheritence.py", line 14, in ?
  x = Lister()
TypeError: not enough arguments; expected 2, got 1
End of trace ...

I guess I can't do methods overloadin', as shown... WHY is that ? I'm
wrong ?

Thomas,


-- 

Thomas SMETS                           Phone : +32 (0)2 742. 05. 94.
Av. de la Braban?onne 133 / 3          e-mail : tsmets @altern.org
1030 Bruxelles

==== Quote of the day =========
1 1 was a race-horse, 2 2 was 1 2. When 1 1 1 1 race, 2 2 1 1 2.
 ======= End of Quote =========
-------------- next part --------------
A non-text attachment was scrubbed...
Name: tsmets.vcf
Type: text/x-vcard
Size: 182 bytes
Desc: Card for Thomas SMETS
URL: <http://mail.python.org/pipermail/python-list/attachments/20000412/1c966570/attachment.vcf>

From reic0024 at ub.d.umn.edu  Wed Apr 12 13:56:24 2000
From: reic0024 at ub.d.umn.edu (Aaron J Reichow)
Date: Wed, 12 Apr 2000 12:56:24 -0500
Subject: Python 4 DOS
In-Reply-To: <sf99mtkv4nk80@corp.supernews.com>
References: <sf7hsua14nk34@corp.supernews.com> <38F48C6D.FD671082@interet.com> <sf99mtkv4nk80@corp.supernews.com>
Message-ID: <Pine.SO4.4.02.10004121254400.29583-100000@ub.d.umn.edu>

On Wed, 12 Apr 2000, Arnaldo Riquelme wrote:

> > www.python.org has a DOS Python running with a DOS extender.
> >
> > Jim
> 
> Thanks Jim
> 
> I want to run Python in just plain dos (6.x), I don't think the version in
> http://www.python.org/ftp/python/wpy/dos.html, can run
> in just DOS. I tried  and it doesn't work.

Did you try running it *with* the DOS extender?  I have no experience
using Python on DOS with the extender, but I've used other prorgams which
required it (DooM, DJGPP [I think]), and they won't work without it.
GO32.exe, or something like that?


Aaron
----
"The best way to predict the future is to invent it." -- Alan Kay





From nmm1 at cus.cam.ac.uk  Fri Apr  7 10:25:25 2000
From: nmm1 at cus.cam.ac.uk (Nick Maclaren)
Date: 7 Apr 2000 14:25:25 GMT
Subject: PPython
References: <38ede548.353776233@news.oh.verio.com> <8ckptt$j8s$1@pegasus.csx.cam.ac.uk> <38EDED43.78A97864@olf.com>
Message-ID: <8ckr4l$kdh$1@pegasus.csx.cam.ac.uk>

In article <38EDED43.78A97864 at olf.com>, Jason Benderly <jbenderl at olf.com> writes:
|> Nick Maclaren wrote:
|> 
|> > If you want a really silly idea, it probably wouldn't be hard to
|> > convert Python's C source to Perl, because the former looks pretty
|> > clean and therefore could be translated semi-automatically without
|> > needing a full-blown C to Perl compiler.  That's the easy bit.
|> >
|> > When you put 130,000 lines of translated source into Perl, you
|> > would discover things about Perl that you didn't want to know and
|> > even its author had not suspected :-)
|> 
|> Could yo be more specific about those things that you wouldn't want to know.
|> I want to know...

No :-)

It is a long time since I looked at the innards of Perl, so I can't
predict any specific source of chaos - but I assuredly can predict
that there would be some!

I suspect that the same would be true of Python, but haven't looked
at it in enough detail to be quite so certain.  But my statement
applies to virtually ANY large program that is pushed so far beyond
its normal limits, so I wasn't exactly sticking my neck out.


Regards,
Nick Maclaren,
University of Cambridge Computing Service,
New Museums Site, Pembroke Street, Cambridge CB2 3QG, England.
Email:  nmm1 at cam.ac.uk
Tel.:  +44 1223 334761    Fax:  +44 1223 334679


From jwmurphy1 at att.com  Thu Apr  6 09:47:24 2000
From: jwmurphy1 at att.com (James W. Murphy)
Date: Thu, 06 Apr 2000 09:47:24 -0400
Subject: MacPython and Tkinter
References: <8cftns$udt$1@flotsam.uits.indiana.edu>
Message-ID: <38EC956C.1318F2F9@att.com>

See http://www.python.org/pipermail/pythonmac-sig/1999-October/002887.html

It's ususally worthwhile to check the Pythonmac SIG list for 
Python/Mac problems.
-- 
jim murphy


From simpsons at kom.auc.dk  Wed Apr 26 16:25:18 2000
From: simpsons at kom.auc.dk (Thomas Rasmussen)
Date: Wed, 26 Apr 2000 20:25:18 GMT
Subject: global variables or inheritence
References: <28p8zy0irgc.fsf@lada.kom.auc.dk> <eRHN4.1113$3K4.19980@news.uswest.net> <28p4s8oipzh.fsf@lada.kom.auc.dk> <j6IN4.1168$3K4.20204@news.uswest.net>
Message-ID: <28pzoqghaoh.fsf@lada.kom.auc.dk>

>>>>> "Jeff" == Jeff Massung <jmassung at magpiesystems.com> writes:

  >> testprog1.py: import os, sys, string, testprog2 def main(): print
  >> 'hello' myvar = 2 testprog2.second() if __name__ == '__main__':
  >> main()

  Jeff> NOTE: myvar is local to main() (not to the file)

Well I still want to be able to do this....

  >>  testprog2.py: import os, sys, string, testprog1 def second():
  >> print testprog1.myvar if __name__ == '__main__': print 'nope'
  >> 

  Jeff> Also, just as coding practice - you're cirular use of the
  Jeff> files is VERY bad... it can cause many problems (and is
  Jeff> illegal in many languages - I can't believe Python would allow
  Jeff> it). What is your program that you need to do this with? Just
  Jeff> curious.

Well i'm programming a small admin tool, which must use the users
$EDITOR for editing files. This works fine if the EDITOR sysenv is
set, if not, then i want to ask the user which editor he want to use,
and then pass this variable to the module that handles
filemanipulation. This modules is also in another file if it
matters... I must add that this is my first real program made in
python...

/Thomas

-- 
KOM Network student helper

"To alcohol! The cause of - and solution to - all of life's problems!"
-- Homer Simpson


From gresham at mediavisual.com  Thu Apr 13 15:22:55 2000
From: gresham at mediavisual.com (Gresh)
Date: Fri, 14 Apr 2000 03:22:55 +0800
Subject: Java
Message-ID: <8d56qk$qdn1@imsp212.netvigator.com>

while (true) {
    using Python considerable these weeks;
    I've just gone and played with Java to build a little applet;
    Now I'm not saying anything bad about Java;
    don't get me wrong;
    but there are few experiences in life;
    where you realize what you are missing;
    i.e {
        wife leaving you;
        being very rich and then suddenly poor;
        programming in python, then {
            returning to a language with curly braces && semicolons;
    }
    if (lucky()){
        I have an IDE;
        that autoformats the code when I ask it too;
    } else {
        I'd be in tears
    }
}

% Message Java/Java [19:1] ';' expected.
    }
    ^

Dammit




From phd at phd.russ.ru  Mon Apr 17 11:48:32 2000
From: phd at phd.russ.ru (Oleg Broytmann)
Date: Mon, 17 Apr 2000 15:48:32 +0000 (GMT)
Subject: Filtering web proxy
In-Reply-To: <00041707431801.00880@quadra.teleo.net>
Message-ID: <Pine.LNX.4.21.0004171547470.31160-100000@fep132.fep.ru>

On Mon, 17 Apr 2000, Patrick Phalen wrote:
> [Oleg Broytmann, on Mon, 17 Apr 2000]
> ::    I want a filtering web proxy. I can write one myself, but if there is a
> :: thing already... well, I don't want to reinvent the wheel. If there is such
> :: thing (free and opensourse, 'course), I'll extend it for my needs.
> 
> Try CTC (Cut the Crap). It's written in Python and is GNU'd.
> 
> http://softlab.ntua.gr/~ckotso/CTC/

   Looks promising, thanks. They also recommends Alfajor:
http://www.andrewcooke.free-online.co.uk/jara/alfajor/. Python, GPL'd...

Oleg.            (All opinions are mine and not of my employer)
---- 
    Oleg Broytmann      Foundation for Effective Policies      phd at phd.russ.ru
           Programmers don't die, they just GOSUB without RETURN.




From matthias_w at my-deja.com  Wed Apr 12 08:36:52 2000
From: matthias_w at my-deja.com (Matthias)
Date: Wed, 12 Apr 2000 12:36:52 GMT
Subject: Embedding Python in Visual C++ 6.0 (Part II)
References: <8d1he4$26d$1@nnrp1.deja.com> <N2YI4.2030$UB.13624@news-server.bigpond.net.au>
Message-ID: <8d1ql3$c14$1@nnrp1.deja.com>

I played a littlet bit in the environment (especially Edit-
>Breakpoints) and if I try to activate the disabled breakpoints VisualC
tells me: "No code exists at the line number for breakpoint at
'{,win32uimodule.cpp,}@2128" - if I say EditCode the compiler jumps to
the correct addressed - I think I'm doing something wrong - but I have
no idea

Thank's for your help
Matthias :-)

In article <N2YI4.2030$UB.13624 at news-server.bigpond.net.au>,
  "Mark Hammond" <mhammond at skippinet.com.au> wrote:
> You also need to override InitApplication().
>
> I should update the docs, but until then, simply look at
Pythonwin.cpp and
> see what it does - whatever it does, it obviously works :-)
>
> Mark.
>
> <matthias_w at my-deja.com> wrote in message
> news:8d1he4$26d$1 at nnrp1.deja.com...
> > Hy! :-)
> >
> > Maybe someone (Mark Hammond??) can help me. I
> > tried to embedd the win32 extentions from Mark in
> > a visual c++ project, but everytime the
> > <b>Win32uiInitInstance()</b> is called (via
> > <b>glue.InitInstance()</b>) I get an unhandled
> > exception at 0xC0000005 Access Violation and the
> > debugger jumps to 1E401F69 - the address of the
> > Win32uiInitInstance() function. When i try create
> > a break-point in this function this break-point
> > is disabled %-( by the compiler. The same is true
> > if i try to debug the original pythonwin project.
> >
> > Help please!
> >
> > Matthias
> >
> >
> > Sent via Deja.com http://www.deja.com/
> > Before you buy.
>
>


Sent via Deja.com http://www.deja.com/
Before you buy.


From m.faassen at vet.uu.nl  Sat Apr 22 10:18:41 2000
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 22 Apr 2000 14:18:41 GMT
Subject: The Simple Economics of Open Source
References: <14591.15630.192472.440375@goon.cnri.reston.va.us> <slrn8fv2rd.1kt.neelk@brick.cswv.com> <3900611F.3B525BDA@cfdrc.com> <FtDG4z.774@world.std.com> <slrn8g26tu.p7o.neelk@brick.cswv.com> <22042000.2@sanctum.jae.ddns.org> <ja1z3yog61.fsf@helix.twistedmatrix.com>
Message-ID: <8dscc1$n1o$2@newshost.accu.uu.nl>

Glyph Lefkowitz <glyph at twistedmatrix.com> wrote:

> I was too tempted by the philosophical sounding thread.  Forgive me
> for jumping in =)

> "Juergen A. Erhard" <jae at ilk.de> writes:

>> Egoism is often equated with the profit motive these days (little
>> wonder in our money-based society...[2])

> Yes, and THIS is where the mistake is from ... there are a LARGE
> number of reasons to do something for personal gain that decidedly do
> NOT involve gaining money.

But I think that the paper did mention this. They sounded like 
economics who did count this non-money gain. Perhaps our trouble is that
they *counted* it, but they're economists after all. Good economists
know that in the end, the true value is not money -- it's happiness,
which includes things like having enough food, other people around, and so
on. They do have to make some assumptions about what makes people happy,
though; if you go into that you're moving into psychology.

[snip]
> For example, if you run across some dusty two-year-old python code
> that (f'rinstance) facilitates time-travel, and you hack on it for a
> while, bring it up to date, but get bored or disinterested because of
> more pressing concerns, like the upcoming alien invasion, you can put
> your code out there and somebody else with a like mind might come
> along and pick up where you left off.

Heh. *Everybody* reads those off-topic threads, don't they? :)

Regards,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From aek at mail1.stofanet.dk  Tue Apr  4 18:31:17 2000
From: aek at mail1.stofanet.dk (Anders Eggers-Krag)
Date: Wed, 5 Apr 2000 00:31:17 +0200
Subject: Python and .HTA files in IE5
References: <38E9EBA2.CF62323A@bellatlantic.net>
Message-ID: <8cdqf1$ano$1@news101.telia.com>

The first problem here is the fact tha the window object overrides the
python objects,
causing open() to be the window.open() method rather than ___builtin___.open

Anders Eggers - Krag



"Richard M. Smith" <rms2000 at bellatlantic.net> wrote in message
news:38E9EBA2.CF62323A at bellatlantic.net...
> Hello,
>
> I am attempting to use Python as a scripting language in a .HTA
> (HTML Application) file.  .HTA files are executed by IE5 just
> like .HTM files except that there are no secrutiy restrictions on
> what script code can do.  In a .HTA files, scripts can run programs,
> read and write files, access the Windows registry, etc.  They are usefull
> for developing Windows applications that use IE5 for their user
> interface.
>
> I am having trouble executing the following simple Python script
> to read the contents of AUTOEXEC.BAT and output it on
> the .HTA page:
>
> <script language=python>
> myfile = open("c:\\autoexec.bat")
> document.write(str(myfile.readlines()))
> myfile.close()
> </script>
>
> I am getting an AttributeError in line 400 of
> the file dynamic.py.  It looks like the __getattr__
> method is failing when the readlines method is
> called by my script code.
>
> Anyone know what is going on here and how to
> fix the proble?.
>
> A couple more data points here.  If I rename the .HTA
> file to have the .HTM extension, I get the identical error
> message.  This leads me to believe that the Python runtime
> library isn't distinguishing between .HTA and .HTM files.
>
> OTOH the equivalent .ASP file works fine:
>
> <script language=python runat=server>
> myfile = open("c:\\autoexec.bat")
> Response.write(str(myfile.readlines()))
> myfile.close()
> </script>
>
> Richard
>
>




From boud at rempt.xs4all.nl  Sun Apr 16 04:09:48 2000
From: boud at rempt.xs4all.nl (Boudewijn Rempt)
Date: 16 Apr 2000 08:09:48 GMT
Subject: Scripting and Gnome and KDE
References: <Pine.GSO.4.10.10004150741520.11261-100000@sundial> <8d95m0$9j6$1@news1.xs4all.nl> <8damat$4ja$1@slb2.atl.mindspring.net>
Message-ID: <8dbsgc$35o$1@news1.xs4all.nl>

Andrew Dalke <dalke at acm.org> wrote:

> Even though it says "The result is a great environment for rapid
> development of GUI-based applications." ?

Yes - the article didn't make clear what was so great, but it did point
out a lot of problems (lack of documentation, inconsistent API's, bugs).

> I used to work at Bioreason, though I wasn't involved in the GUI
> development.  They are pretty happy about PyGTK.  The idea of the
> DDJ article was to point out how to take advantage of the toolkit,
> based on finding out what doesn't work.

Very honest, but it doesn't make me all that eager ;-).

> You could write a similarly bleak picture of any toolkit.

Absolutely, but I don't know that an essay on the problems is exactly
the right way of introducing the toolkit. 

-- 

Boudewijn Rempt  | http://www.valdyas.org


From akuchlin at mems-exchange.org  Fri Apr 14 13:43:29 2000
From: akuchlin at mems-exchange.org (Andrew M. Kuchling)
Date: 14 Apr 2000 13:43:29 -0400
Subject: Python 1.6a2 crashes in test_fork1 on my Mandrake Linux 7 machine
References: <20000413010344.A13931@owl.rhein-zeitung.de> <38F658C6.66AD9CA5@clark.net> <20000414154718.A15664@xs4all.nl> <20000414190308.A30502@owl.rhein-zeitung.de>
Message-ID: <3d4s94h932.fsf@amarok.cnri.reston.va.us>

Oliver Andrich <oli at rz-online.net> writes:
> this is the interesting thing. The system is the same where I run python 1.5.2
> and 1.6a2 on. Threading code seems to be the same in the source distribution.
> As far as I checked it. It is getting stranger and stranger. 

Careful... Fred Drake checked in a change to test_fork1 not too long
ago; perhaps that's causing the problem.  It might also be related to
the problem described in http://x32.deja.com/getdoc.xp?AN=602366237

-- 
A.M. Kuchling			http://starship.python.net/crew/amk/
Well, you are the best I have, Cluracan, and to say that is to say little
indeed.
  -- Queen Titania, in SANDMAN #52: "Cluracan's Tale"




From scarblac-spamtrap at pino.selwerd.nl  Fri Apr 28 12:15:48 2000
From: scarblac-spamtrap at pino.selwerd.nl (Remco Gerlich)
Date: 28 Apr 2000 16:15:48 GMT
Subject: ?sys.argv values more literally?
References: <3909B1AF.32FC8BED@email.sps.mot.com>
Message-ID: <slrn8gje83.fqh.scarblac-spamtrap@flits104-37.flits.rug.nl>

Norman Shelley wrote in comp.lang.python:
> I need   argv to look like this:
> ["frontend", "-ver", "1.0", "word1", "word2",  '"word three"',
> '"word(three)"']
> where words that were quoted on the command line are strings with
> beginning and ending quotes in the argv.
> 
> Any way to do this or get around this problem?

Can't you just quote every argument?

-- 
Remco Gerlich,  scarblac at pino.selwerd.nl
  6:14pm  up 53 days,  6:31,  8 users,  load average: 0.18, 0.24, 0.13


From dworkin at ccs.neu.edu  Thu Apr  6 11:46:26 2000
From: dworkin at ccs.neu.edu (Justin Sheehy)
Date: 06 Apr 2000 11:46:26 -0400
Subject: More about variables
In-Reply-To: Matthew Hirsch's message of "Thu, 06 Apr 2000 10:44:54 -0400"
References: <meh9-5C17A0.10445406042000@news.cit.cornell.edu>
Message-ID: <vndwvmbb5bh.fsf@camelot-new.ccs.neu.edu>

Matthew Hirsch <meh9 at cornell.edu> writes:

> Is there a more efficient way of doing this? What if I wanted to create 
> 100 variables (lists).  

Yes.  Put them in a dictionary.

-Justin

 



From mikael at isy.liu.se  Thu Apr 27 09:54:32 2000
From: mikael at isy.liu.se (Mikael Olofsson)
Date: Thu, 27 Apr 2000 15:54:32 +0200 (MET DST)
Subject: Concatening string and integer
In-Reply-To: <8e9gco$jlt$1@vg170.it.volvo.se>
Message-ID: <XFMail.000427155432.mikael@isy.liu.se>

On 27-Apr-00 JJ wrote:
 >  I want to concatenate a string containing both strings and integers like:
 >  
 >  strTemp = 'I make test number: ' + iMyNumber + ' once again'
 >  
 >  It doesn't work, how should I do this?

You need to convert the integer to a string, for instance as

  strTemp = 'I make test number: ' + str(iMyNumber) + ' once again'

or as

  strTemp = 'I make test number: %s once again'%iMyNumber

HTH

/Mikael

-----------------------------------------------------------------------
E-Mail:  Mikael Olofsson <mikael at isy.liu.se>
WWW:     http://www.dtr.isy.liu.se/dtr/staff/mikael
Phone:   +46 - (0)13 - 28 1343
Telefax: +46 - (0)13 - 28 1339
Date:    27-Apr-00
Time:    15:51:27

This message was sent by XF-Mail.
-----------------------------------------------------------------------


From nmm1 at cus.cam.ac.uk  Fri Apr  7 08:20:45 2000
From: nmm1 at cus.cam.ac.uk (Nick Maclaren)
Date: 7 Apr 2000 12:20:45 GMT
Subject: Newbie: How to do an exit ?
References: <38EDC1DA.62CBFCF3@sxb.bsf.alcatel.fr>
Message-ID: <8ckjqt$ds9$1@pegasus.csx.cam.ac.uk>

In article <38EDC1DA.62CBFCF3 at sxb.bsf.alcatel.fr>, Pascal Fremaux <pf at sxb.bsf.alcatel.fr> writes:
|> How to exit from a program ?
|> I don't find a command such as exit(0) in C

Try one of:

    raise SystemExit(0)

    import sys
    sys.exit(0)

    from sys import exit
    exit(0)


Regards,
Nick Maclaren,
University of Cambridge Computing Service,
New Museums Site, Pembroke Street, Cambridge CB2 3QG, England.
Email:  nmm1 at cam.ac.uk
Tel.:  +44 1223 334761    Fax:  +44 1223 334679


From firefly64 at my-deja.com  Sun Apr 16 21:13:05 2000
From: firefly64 at my-deja.com (firefly64 at my-deja.com)
Date: Mon, 17 Apr 2000 01:13:05 GMT
Subject: threads on linux
Message-ID: <8ddoem$5sl$1@nnrp1.deja.com>

Hi,
I'm trying to build python 1.5.2 on a linux box
but without thread support. I'm using gcc 2.9.5,
on Mandrak Linux 6.1 kernel version 2.2.14

I've done everything I can think of, including
./configure --without-thread and editing the
Makefiles to include -DWITH_THREAD=0. The build
process still always either produces a python with
linked to libpthreads or dies complaining about
not being able to find various thread functions.
(Yes, I've done make clean).

The reason I'm doing this is because I'm trying to
get httpdapy to work. The author suggested that I
try building a python without threads.


Thanks a lot everyone,
mike salib


Sent via Deja.com http://www.deja.com/
Before you buy.


From william_93402 at my-deja.com  Fri Apr 14 11:04:00 2000
From: william_93402 at my-deja.com (william_93402 at my-deja.com)
Date: Fri, 14 Apr 2000 15:04:00 GMT
Subject: problem with os.system calls under win32
Message-ID: <8d7c0k$fjr$1@nnrp1.deja.com>

I am attempting to do a system call in a directory with spaces in the
name. For example:

>>>import os
>>>os.system('c:\\f o o\\bar.exe')
1

when I place the executable in a dir such as c:\\foo I can successfully
execute the system call. Is there something special I have to do when
directory names contain spaces? Just for notes I am running by example
in build 125 of pythonwin.

Thanks,

Bill Browning


Sent via Deja.com http://www.deja.com/
Before you buy.


From nmm1 at cus.cam.ac.uk  Sat Apr  8 18:40:11 2000
From: nmm1 at cus.cam.ac.uk (Nick Maclaren)
Date: 8 Apr 2000 22:40:11 GMT
Subject: Python paradigms
References: <8cmr70$b0e$1@pegasus.csx.cam.ac.uk> <8cntqp$upi$1@slb0.atl.mindspring.net>
Message-ID: <8cocgb$iuk$1@pegasus.csx.cam.ac.uk>

In article <8cntqp$upi$1 at slb0.atl.mindspring.net>,
Andrew Dalke <dalke at acm.org> wrote:
>
>Nick Maclaren wrote
>>    x = (a != NULL ? a[i]->weeble : 0) + (b != NULL ? b[i]->wombat : 0)
>> [...]
>>What paradigms are there for replacing such constructions by?
>
>
>How about
>
>  x = getattr(a, "weeble", 0) + getattr(b, "wombat", 0)

Thank you, yes, that does deal with that specific example.  But
it does not extend to the general case.


Regards,
Nick Maclaren,
University of Cambridge Computing Service,
New Museums Site, Pembroke Street, Cambridge CB2 3QG, England.
Email:  nmm1 at cam.ac.uk
Tel.:  +44 1223 334761    Fax:  +44 1223 334679


From eviltofu at rocketmail.com  Tue Apr 18 07:36:11 2000
From: eviltofu at rocketmail.com (Jerome Chan)
Date: Tue, 18 Apr 2000 11:36:11 GMT
Subject: Forward References?
References: <eviltofu-4C5688.21593817042000@news.earthlink.net> <m3itxgcggc.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <eviltofu-4B17FA.07361018042000@news.earthlink.net>

In article <m3itxgcggc.fsf at atrus.jesus.cam.ac.uk>, Michael Hudson 
<mwh21 at cam.ac.uk> wrote:

> Jerome Chan <eviltofu at rocketmail.com> writes:
> 
> > Just a silly question but how do I do forward references?
> > 
> > 
> > class a:
> >    def x(self):
> >       y(self)
> >    def y(self):
> >       print "Hello"
> > 
> > c = a()
> > c.x()
> > c.y()
> > 
> > gives a NameError: y
> 
> Your mistake isn;t what you think it is, try this instead:
> 
> class a:
>    def x(self):
>       self.y()
>    def y(self):
>       print "Hello"
> 
> HTH, HAND,
> M.

Thanks! I also found out that 


a.y() works!


From moshez at math.huji.ac.il  Thu Apr  6 03:22:38 2000
From: moshez at math.huji.ac.il (Moshe Zadka)
Date: Thu, 6 Apr 2000 09:22:38 +0200 (IST)
Subject: Why should I switch to Python? - Infinity of Primes
In-Reply-To: <000601bf9f8e$bd31e780$ea2d153f@tim>
Message-ID: <Pine.GSO.4.10.10004060914420.3428-100000@sundial>

On Thu, 6 Apr 2000, Tim Peters wrote:

> That's approximately as irritating as it is in Scheme, were laziness also
> needs to be faked.

Tim, you said that just to annoy me, didn't you?
Scheme is much less painful: it has all kinds of macros and special forms,
like "force" and "promise" to make it as irritatingless as just writing
the infrastructure.

If Python has macros and lexical scoping, it would get much simpler.

can't-wait-for-sicp-to-get-here-ly y'rs, Z.
--
Moshe Zadka <mzadka at geocities.com>. 
http://www.oreilly.com/news/prescod_0300.html
http://www.linux.org.il -- we put the penguin in .com




From meh9 at cornell.edu  Wed Apr 26 13:25:16 2000
From: meh9 at cornell.edu (Matthew Hirsch)
Date: Wed, 26 Apr 2000 13:25:16 -0400
Subject: Simple del[] list question
Message-ID: <meh9-9D8292.13251626042000@news.cit.cornell.edu>

Hi All,

What is going on here?

This leaves a intact.

>>> a=[[1,2,3],[4,5,6]]
>>> b=a[:]
>>> del b[0]
>>> b
[[4, 5, 6]]
>>> a
[[1, 2, 3], [4, 5, 6]]


This also changes a.

>>> a=[[1,2,3],[4,5,6]]
>>> b=a[:]
>>> del b[0][0]
>>> b
[[2, 3], [4, 5, 6]]
>>> a
[[2, 3], [4, 5, 6]]

Why did a change?  Why suddenly when you go to two dimensions does the 
behavior change?

Thanks,
Matt


From effbot at telia.com  Tue Apr 18 03:11:58 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Tue, 18 Apr 2000 07:11:58 GMT
Subject: else clauses in while and for loops
References: <2JKK4.325$vr2.19414@news.uswest.net> <Zv_C4.6415$4N.169440@news.tpnet.pl> <38DD2EE3.C7F3CB54@webamused.com> <8boe2k$gmh$1@inputplus.demon.co.uk> <v0RD4.13$n68.186989568@newsb.telia.net> <8c7b29$37v$1@inputplus.demon.co.uk> <38F79B93.B03F0A6B@mindspring.com> <8dg14u$f08$1@nntp6.u.washington.edu> <%TNK4.430$vr2.32615@news.uswest.net>
Message-ID: <2VTK4.421$wYl.187705344@newsb.telia.net>

Jeff Massung <jmassung at magpiesystems.com> wrote:
> >Yes, but if you look at the examples, they also use "break".  After a
> >loop, "else" basically means "if no iteration of the loop met the
> >conditions to break out."  It's a common usage, where in C we'd compare
> >the final loop index with the maximum (or whatever) loop index to see
> >if the loop ran all the way through.
>
> I don't think that I'm following you, give me an example where the loop
> ends, but the else wouldn't be executed and I think I'll better
understand.

how about reading the fine manual?

    http://www.python.org/doc/current/ref/while.html

    "A break statement executed in the first suite terminates
    the loop without executing the else clause's suite."

examples have been given earlier in this thread, but here's
another one:

        s = list_of_strings

        for i in range(len(s)):
            if s[i] == something:
                break
            else:
                print "didn't find anything"
                raise ValueError

        print "index is", i

> >your paragraph would justify better on the right margin, too!  Seriously,
> >"else" does invariably execute if the loop was never run once, but as
> >mentioned above it also executes in other circumstances.
>
> I think it would be better suited if it didn't execute under "other
> circumstances". It would be nice to know _exactly_ when it would execute,
> and only under 1 or 2 conditions (maybe that is the way it is now, I just
> don't see it).

"if", "for", "while", and "try" all execute their "else" suite if
and only if the controlling condition is tested and found to
be false.

for "if" and "while", the controlling condition is the given
expression.  if evaluated, and found to be false, the "else"
suite is executed.

for a "for" statement, the controlling condition is "is the
sequence exhausted?".

for a "try/except" statement, it is "did an exception occur
in the try suite?".

statements like "break", "return", and "raise" can be used to
terminate loops without testing the condition.

</F>

<!-- (the eff-bot guide to) the standard python library:
http://www.pythonware.com/people/fredrik/librarybook.htm
-->




From yemartin at garage.co.jp  Wed Apr 12 03:16:21 2000
From: yemartin at garage.co.jp (Yves-Eric Martin)
Date: Wed, 12 Apr 2000 16:16:21 +0900
Subject: Seg fault in Python/C API: PyString_FromString(NULL)
Message-ID: <38F422C5293.A67BYEMARTIN@smtp.garage.co.jp>

    Hi there,

    While writing a little test unit for a Python gettext
implementation, I ran into a seg fault that I traced to
PyString_FromString being called with a NULL argument.

    First, I am new to the Python/C API, so correct me if I am wrong:
shouldn't PyString_FromString(NULL) return Py_None? Or am I out to
lunch?

    Any information on this segfault is welcome. Here is how I simply
reproduce it: (This was run on Debian GNU/Linux 2.2.14, with gcc
2.95.2 and Python 1.5.2).

---------------- test.c ----------------
#include <stdio.h>
#include "Python.h"

static PyObject*
PyTest_testnull(PyObject* self,PyObject *args)
{
  return PyString_FromString(NULL);
}

static struct PyMethodDef PyTest_Methods[] = {
  {"testnull",(PyCFunction)PyTest_testnull,METH_VARARGS},
  {NULL, NULL}
};

void inittest()
{
  Py_InitModule("test",PyTest_Methods);
}
--------------------------------


---------------- Makefile ----------------
test.so: test.o test.c
        gcc -shared -lc  test.o  -o test.so

test.o: test.c
        gcc -fPIC  -g -O2 -I/usr/include/python1.5 -I/usr/include/python1.5 -DHAVE_CONFIG_H -c ./test.c
--------------------------------


---------------- runtest.py ----------------
import test

print "About to run test..."
test.testnull()
print "If you are reading this, it's working!"
--------------------------------

The result of running "python runtest.py" is:

About to run test...
Segmentation fault



    Thanks.

-- 
Yves-Eric Martin
Digital Garage Inc.
yemartin at garage.co.jp



From rcunning at acm.org  Sun Apr  9 00:47:40 2000
From: rcunning at acm.org (Robert W. Cunningham)
Date: Sun, 09 Apr 2000 04:47:40 GMT
Subject: Cacheing sys.path rather than building it each time?
References: <1256908084-26112509@hypernet.com>
Message-ID: <38F00ABF.3E8A28E2@acm.org>

Gordon McMillan wrote:

> Hamish Lawson wrote:
>
> > Steve Holden wrote:
> >
> > > Probably not an awfully good idea, given that "installation" of
> > > a new module often consists of just dropping a ".py" file into
> > > a library directory.  Without an common "install" process the
> > > work of checking whether the cahced path is up to date would
> > > be almost as much as that of building the path from scratch
> > > as is now done.
> >
> > Had you thought I was proposing that such a check be done
> > whenever Python was started? I'd agree that the work of checking
> > whether the cached path is up to date would be almost as much
> > work as building the path. However that wasn't part of my
> > proposal.
> >
> > Instead when a module is installed, the pathbuilder tool would be
> > run either by the module's install program (perhaps via the
> > distutils) or manually by the user. (You could even run it at
> > scheduled intervals if you thought that necessary to catch
> > omissions.) My goal is to avoid having the path built every time
> > Python starts.
> >
> > A command-line option for the Python interpreter could be used to
> > determine whether it uses the cached path or builds the path from
> > scratch; the decision as to which should be the default I'll
> > leave to others.
>
> There are 3 pieces to the puzzle.
>
> 1) How do you find those things that Python requires
> (bootstrap).
> 2) How do you build the list of places to look for Python
> modules.
> 3) What is the complete universe of importable Python
> modules.

This  discussion seems to have swerved away from the concept of the
feature, and toward the automation of the use of that feature.

For my $0.02, build in the feature, and let it be Caveat Emptor for all
users until a suitable automation and management system is imposed upon
it.  For a fairly static system, it will bring significant speed
benefit, and if things start acting strange, it will soon become
automatic to first invalidate the cache before doing any deeper
debugging.

Those who don't understand the feature, or don't trust it, or are just
afraid to use it, will not be forced to use it.

-BobC




From gregor at hoffleit.de  Tue Apr 25 02:19:32 2000
From: gregor at hoffleit.de (Gregor Hoffleit)
Date: Tue, 25 Apr 2000 08:19:32 +0200
Subject: MetaKit 2.01: problem with g++ 2.95.2
Message-ID: <20000425081932.B29635@53b.hoffleit.de>

When I try to compile the Python extensions for MetaKit 2.01 on Debian
GNU/Linux (using GNU C++ 2.95.2), it fails (in python/PyView.cpp) with what
looks to me like an C++ error (see below for the transcript).

Furthermore, for some reason I'm unable to compile the kit with tcl8.0-dev
(see below for the error log in compiling tcl/mk4tcl.cpp), though, IIRC
tcl8.2-dev worked fine.


The README implies that the build would work with gcc 2.95.2. Has anybody
reports on success or problems with a similar configuration ? My C++
knowledge is so sparse that I don't see where the problem could be.


    Gregor 


PS: I already reported this to the MetaKit Bug Tracking system as #104541
(http://sourceforge.net/bugs/?func=detailbug&bug_id=104541&group_id=736),
but got no response yet.
    


clapton:3> make Mk4py.so
...
/bin/sh ./libtool --mode=compile c++ -c -O3 -fomit-frame-pointer -DNDEBUG \
   -I../include -I../python/scxx -I/usr/include/python1.5 ../python/PyView.cpp
rm -f .libs/PyView.lo
c++ -c -O3 -fomit-frame-pointer -DNDEBUG -I../include -I../python/scxx \
-I/usr/include/python1.5 ../python/PyView.cpp  -fPIC -DPIC -o .libs/PyView.lo
../python/PyView.cpp: In method void PyView::map(const PWOCallable &)':
../python/PyView.cpp:962: 
     initialization of non-const reference type class PWOBase &'
../python/PyView.cpp:962: from rvalue of type PWOBase'
../python/scxx/PWOSequence.h:121: 
     in passing argument 2 of PWOTuple::setItem(int, PWOBase &)'
../python/PyView.cpp: 
     In method void PyView::map(const PWOCallable &, const PyView &)':
../python/PyView.cpp:972: 
     initialization of non-const reference type class PWOBase &'
../python/PyView.cpp:972: from rvalue of type PWOBase'
../python/scxx/PWOSequence.h:121: 
     in passing argument 2 of PWOTuple::setItem(int, PWOBase &)'
../python/PyView.cpp: 
     In method class PyView * PyView::filter(const PWOCallable &)':
../python/PyView.cpp:1000: 
     initialization of non-const reference type class PWOBase &'
../python/PyView.cpp:1000: from rvalue of type PWOBase'
../python/scxx/PWOSequence.h:121: 
     in passing argument 2 of PWOTuple::setItem(int, PWOBase &)'
../python/PyView.cpp: 
 In method struct PyObject * PyView::reduce(const PWOCallable &, PWONumber &)':
../python/PyView.cpp:1015: 
     initialization of non-const reference type class PWOBase &'
../python/PyView.cpp:1015: from rvalue of type PWOBase'
../python/scxx/PWOSequence.h:121: 
     in passing argument 2 of PWOTuple::setItem(int, PWOBase &)'
make: *** [PyView.o] Error 1





freefly;40> make
/bin/sh ./libtool --mode=compile c++ -c -O3 -fomit-frame-pointer -DNDEBUG \
     -I../include -I../../include -I/usr/lib/../generic -DUSE_TCL_STUBS \
     ../tcl/mk4tcl.cpp
rm -f .libs/mk4tcl.lo
c++ -c -O3 -fomit-frame-pointer -DNDEBUG -I../include -I../../include \
     -I/usr/lib/../generic -DUSE_TCL_STUBS ../tcl/mk4tcl.cpp  -fPIC -DPIC -o \
     .libs/mk4tcl.lo
../tcl/mk4tcl.cpp:54: mkMutex' was not declared in this scope
../tcl/mk4tcl.cpp:66: ANSI C++ forbids declaration TCL_DECLARE_MUTEX' with
     no type
../tcl/mk4tcl.cpp:66: syntax error before {'
../tcl/mk4tcl.cpp:69: destructors must be member functions
../tcl/mk4tcl.cpp:71: non-member function GetLength()' cannot have const'
     method qualifier
../tcl/mk4tcl.cpp:74: non-member function GetData(int)' cannot have const'
     method qualifier
../tcl/mk4tcl.cpp:75: new declaration void * GetData(int)'
../tcl/mk4tcl.cpp:74: ambiguates old declaration const void * GetData(int)'
../tcl/mk4tcl.cpp:82: parse error before private'
../tcl/mk4tcl.cpp:85: parse error before }'
../tcl/mk4tcl.cpp:108: syntax error before ;'
../tcl/mk4tcl.cpp: In method int c4_PtrArray::GetSize() const':
../tcl/mk4tcl.cpp:127: _vector' undeclared (first use this function)
../tcl/mk4tcl.cpp:127: (Each undeclared identifier is reported only once
../tcl/mk4tcl.cpp:127: for each function it appears in.)
../tcl/mk4tcl.cpp: In function struct Tcl_Obj * GetAsObj(const c4_RowRef &,
     const c4_Property &, Tcl_Obj * = 0)':
../tcl/mk4tcl.cpp:459: implicit declaration of function int
     Tcl_SetByteArrayObj(...)'
../tcl/mk4tcl.cpp: In function int SetAsObj(Tcl_Interp *, const c4_RowRef &,
     const c4_Property &, Tcl_Obj *)':
../tcl/mk4tcl.cpp:514: implicit declaration of function int
     Tcl_GetByteArrayFromObj(...)'
../tcl/mk4tcl.cpp:514: initialization to const t4_byte *' from int' lacks a
     cast
../tcl/mk4tcl.cpp: In method int TclSelector::MatchOneString(int, const char
     *, const c4_String &)':
../tcl/mk4tcl.cpp:1668: implicit declaration of function int
     Tcl_StringCaseMatch(...)'
../tcl/mk4tcl.cpp:1668: warning: cannot pass objects of type const
     c4_String' through ...'
../tcl/mk4tcl.cpp: In method int MkTcl::Execute(int, Tcl_Obj *const *)':
../tcl/mk4tcl.cpp:2980: mkMutex' undeclared (first use this function)
../tcl/mk4tcl.cpp:2980: implicit declaration of function int
     Tcl_MutexLock(...)'
../tcl/mk4tcl.cpp:2982: implicit declaration of function int
     Tcl_MutexUnlock(...)'
../tcl/mk4tcl.cpp: In function int Mktcl_Cmds(Tcl_Interp *, int)':
../tcl/mk4tcl.cpp:3012: implicit declaration of function int
     Tcl_InitStubs(...)'
../tcl/mk4tcl.cpp:3012: warning: NULL used in arithmetic
make: *** [mk4tcl.o] Error 1
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 240 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20000425/452a7b8b/attachment.sig>

From jcm at bigskytel.com  Mon Apr 17 20:10:37 2000
From: jcm at bigskytel.com (David Porter)
Date: Mon, 17 Apr 2000 18:10:37 -0600
Subject: Filtering web proxy
In-Reply-To: <Pine.LNX.4.21.0004171406250.30999-100000@fep132.fep.ru>; from phd@phd.russ.ru on Mon, Apr 17, 2000 at 02:20:02PM +0000
References: <Pine.LNX.4.21.0004171406250.30999-100000@fep132.fep.ru>
Message-ID: <20000417181037.B8191@bigskytel.com>

* Oleg Broytmann <phd at phd.russ.ru>:
> Hello!
> 
>    I want a filtering web proxy. I can write one myself, but if there is a
> thing already... well, I don't want to reinvent the wheel. If there is such
> thing (free and opensourse, 'course), I'll extend it for my needs.
> 
>    Once I tried Junkbuster, but found it inadequate. My need is simple -
> just a short list of "white" sites. Never tried adzapper...

You tried the *wrong* junkbuster. There is an alternate version at
http://www.waldherr.org/junkbuster/ which is the best ad filter that I have
ever used (and still use). It comes with a very complete ad blocking list
and it turns blocked images into transparent 1x1 gifs, both unlike the
official version. 

I have tried adzapper and CTC (Cut The Crap), but both of them were
unreliable. (I should mail the creators about this RSN.) They would for no
apparent reason, just stop working. That is, I would select a URL and
nothing would happen... I found myself constantly bypassing the proxy
because of this. I also dislike the concept of creating a new file everytime
that I want to block a site, one central config suits me better.

If the junkbuster version I mentioned is not satisfactory to you I would be
highly surprised, give it a go!


    __David__



From gmcm at hypernet.com  Tue Apr  4 11:53:51 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Tue, 4 Apr 2000 11:53:51 -0400
Subject: lightweight embedded database table object
In-Reply-To: <5a6G4.2924$HG1.87456@nnrp1.uunet.ca>
Message-ID: <1257258860-5013801@hypernet.com>

Warren Postma wrote:
> 
> > Hi Warren,
> > Have you looked at MetaKit? http://www.equi4.com/metakit/python.html
> 
> It didn't look 'ready for prime time' -- too many disclaimers of the kind I
> usually put on something when I expect it's still quite buggy or unscalable.
> I suppose I should give it more of a try though, eh!? Is anyone using this
> in a 'production environment'?

Yes. Metakit, extensively. Probably less so for Mk4py.

The problems you'll find aren't stability problems. Some of the 
advanced stuff (the relational algebra) can be tricky to get 
working - it's not documented well, and you may find, for 
example, that some operation returns a read-only view where 
updates silently fail. It should all show up during development, 
though.

- Gordon



From m.faassen at vet.uu.nl  Sun Apr  2 14:16:14 2000
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 2 Apr 2000 18:16:14 GMT
Subject: was Re: Why should I switch to Python?
References: <seas5ia89177@corp.supernews.com> <1EeF4.43$CA.2422@news> <oqzord6h5i.fsf@titan.progiciels-bpi.ca> <38E618D7.A15F9D9B@webwitches.com>
Message-ID: <8c82pe$pf0$2@newshost.accu.uu.nl>

Shae Erisson <shae at webwitches.com> wrote:
[snip]
> I wonder if the next step to CP4E would be online 'classes?' Emacs
> allows for multiple windows on multiple displays. A friend and I have
> used that to try out multi-user editing of a single document across
> 800km, and it works.

How do you do this? I wanna know! :) I'm teaching Python to a friend
across the Atlantic, and this would be a great thing to be able to
use.

Regards,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From echuck at mindspring.com  Fri Apr 14 18:28:35 2000
From: echuck at mindspring.com (Chuck Esterbrook)
Date: Fri, 14 Apr 2000 18:28:35 -0400
Subject: else clauses in while and for loops
References: <Zv_C4.6415$4N.169440@news.tpnet.pl> <38DD2EE3.C7F3CB54@webamused.com> <8boe2k$gmh$1@inputplus.demon.co.uk> <v0RD4.13$n68.186989568@newsb.telia.net> <8c7b29$37v$1@inputplus.demon.co.uk>
Message-ID: <38F79B93.B03F0A6B@mindspring.com>

So did anyone present, in this thread, an example that shows a real need?

That seemed to have gotten lost in the poor example and the strange tangent about removing features from Python...

-Chuck


Ralph Corderoy wrote:

> Hi,
>
> > > I don't think that is a real need.
> >
> > so?  do you suggest that we should remove it, and force people to
> > rewrite their perfectly working pro- grams, just to make you happy?
> > or maybe we could use all that energy to work on something more
> > important?
>
> I suggest you go back and read my post properly.  In particular, the
> original poster's request.
>
> > > Please give some examples showing the real need for else clauses in
> > > this context.
>
> Someone replied giving a pointless use of else.  I said out that this
> was a pointless use of else and therefore didn't match the original
> poster's request of an example `showing the real need'.
>
> I didn't say there wasn't a real need for else, I said `I don't think
> that is a real need' immediately after quoting the pointless example.
>
> Ralph.
>
> -------- Original post follows
>
> > > Please give some examples showing the real need for else clauses in
> > > this context.
> >
> > def lookup(stringList, word):
> >     """Finds the first instance of word in stringList. Returns -1 if not
> > found"""
> >     i = 0
> >     while i < len(stringList):
> >         if stringList[i] == word:
> >             break
> >         i = i + 1
> >     else:
> >       # is only executed if we never hit that break
> >         return -1
> >     return i
>
> I don't think that is a real need.  Wouldn't it normally be written to
> return ASAP rather than jump through hoops.
>
>     def lookup(stringList, word):
>         i = 0
>         while i < len(stringList):
>             if stringList[i] == word:
>                 return i
>             i = i + 1
>         return -1
>
> Ralph.



From mfletch at tpresence.com  Thu Apr 13 10:58:45 2000
From: mfletch at tpresence.com (Mike Fletcher)
Date: Thu, 13 Apr 2000 10:58:45 -0400
Subject: List mapping?
Message-ID: <CC59868A0A76D311B1F50090277B321C1F326B@VRTMAIL>

Oh, how would I actually do the work of parsing command lines?  Probably use
the getopt module :) 

Enjoy,
Mike

-----Original Message-----
From: Nick Trout [mailto:nick at spam_me_notvideosystem.co.uk]
Sent: Thursday, April 13, 2000 11:02 AM
To: python-list at python.org
Subject: Re: List mapping?


I came up with something similar:

    processfn = { 'del':deletefiles }
    comm = None
    flist = []
    fileopts = sys.argv[2:]
    while fileopts:
        if fileopts[0]=='/c':
            comm = fileopts[1]
            fileopts = fileopts[2:]
        elif fileopts[0]=='/d':
            fdir = fileopts[1]
            ext = fileopts[2]
            fileopts = fileopts[3:]
            flist.extend( readdir(fdir,ext) )
        else:
            raise "Unknown flag:",fileopts[0]
    processfn[comm](flist)

There are obviously a lot of ways to do this in Python, the most elegant way
doesnt always seem obvious at first! Just wondered how others would approach
this problem.

Cheers,
Nick.


"Mike Fletcher" <mfletch at tpresence.com> wrote in message
news:CC59868A0A76D311B1F50090277B321C1F326A at VRTMAIL...
> not elegant, but here's another way...
>
> original = string.split( x, ' ')
> result = []
> while original:
> result.append( tuple(original[:2]) )
> del original[:2]
>
> and another...
>
> original = string.split( x, ' ')
> result = []
> for x in range(0, len(original), 2 ):
> result.append( tuple( original[x:x+2]))
>
> HTH,
> Mike
>
> -----Original Message-----
> From: Nick Trout [mailto:nick at spam_me_notvideosystem.co.uk]
> Sent: Thursday, April 13, 2000 10:10 AM
> To: python-list at python.org
> Subject: List mapping?
>
>
> Is there a nice way of pairing up the members to form tuple pairs in
another
> list? ie. like you might to processing input args.
>
> ie. changing ['/a','1','/b','2','/c','3'] into [ ('/a','1'), ('/b','2'),
> ('/c','3') ]
>
> eg:
>
> args = '/a 1 /b 2 /c 3'
> listargs= string.split(args)
> tuplist = []
> i=0
> while i < len(listargs):
>     tuplist.append( (listargs[i],listargs[i+1]) )
>     i = i+2
> for i in tuplist:
>     processargs(i)
>
> or:
>
> args = '/a 1 /b 2 /c 3'
> for i in string.split(args):
>     i = argflag
>     # i want next arg now as a parameter!!!
>
> Can I (you!) get rid of the long winded while loop cleverly?!!  :-)
>
> Thanks in advance,
> Nick.
>
>
>
>
>
> --
> http://www.python.org/mailman/listinfo/python-list
>


-- 
http://www.python.org/mailman/listinfo/python-list



From mhammond at skippinet.com.au  Thu Apr  6 19:45:01 2000
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Thu, 06 Apr 2000 23:45:01 GMT
Subject: Formatting Excel cells with PythonCOM
References: <38ECAF9D.ECA74709@ansic.net>
Message-ID: <1k9H4.28$4v.151@news-server.bigpond.net.au>

This code is from the win32com test suite (in testMSOffice.py)

 # test dates out with Excel
 xl.Cells(5,1).Value = "Excel time"
 xl.Cells(5,2).Formula = "=Now()"

 import time
 xl.Cells(6,1).Value = "Python time"
 xl.Cells(6,2).Value = pythoncom.MakeTime(time.time())
 xl.Cells(6,2).NumberFormat = "d/mm/yy h:mm"
 xl.Columns("A:B").EntireColumn.AutoFit()


"Scott" <sbarron at ansic.net> wrote in message
news:38ECAF9D.ECA74709 at ansic.net...
> Hi,
>
> I am looking for information regarding how to set the Date/Time
format
> for Excel cells.  When I enter my dates/times I use the
DateTime module
> and the COMDate member of my objects.  When I open Excel it
displays
> them as ###### until I manually set the date/time format.  How
can I do
> this from within my Python program?
>
> Thanks,
> Scott




From smalleys at gte.net  Wed Apr 26 23:53:29 2000
From: smalleys at gte.net (sue)
Date: Thu, 27 Apr 2000 03:53:29 GMT
Subject: new linereading standard?
References: <3904D21B.333DD930@visionart.com> <390510C9.A744D22E@yahoo.com> <R2g+MIAFMUB5EwUl@jessikat.demon.co.uk> <mt1z3uyr8k.fsf@astron.berkeley.edu>
Message-ID: <3907B9D0.F5E59295@gte.net>

Maps and lambdas on the command line let you backtrack
without regenerating everything from scratch.  I've thought
about writing a small "map fn" library that I could import
just for CLI.

Sue.


From greg at cosc.canterbury.ac.nz  Wed Apr 19 00:04:29 2000
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Wed, 19 Apr 2000 16:04:29 +1200
Subject: dir(..)   vs.   PyList_GetItem
References: <213CDEBFDDA7D3118081009027E40C56071155@SHEF-MS1>
Message-ID: <38FD304C.8E0990E1@cosc.canterbury.ac.nz>

Mark Pope wrote:
> 
>         pModule = PyModule_New( "myScript" );

That doesn't import the module, it just creates an 
empty module. You want

   pModule = PyImport_ImportModule("myScript");

-- 
Greg Ewing, Computer Science Dept,
+--------------------------------------+
University of Canterbury,	   | A citizen of NewZealandCorp, a	  |
Christchurch, New Zealand	   | wholly-owned subsidiary of USA Inc.  |
greg at cosc.canterbury.ac.nz	   +--------------------------------------+


From beigebox at usa.net  Wed Apr 12 18:28:19 2000
From: beigebox at usa.net (deviant)
Date: 12 Apr 2000 22:28:19 GMT
Subject: Word up
Message-ID: <8d2ta3$qil$6303@testinfo.cs.uoguelph.ca>

Check out my music at http://nav.to/deviant3/

eveything under the sun.. mp3s and impulse tracker files are available... 




From embed at geocities.com  Tue Apr 18 15:13:08 2000
From: embed at geocities.com (Warren Postma)
Date: Tue, 18 Apr 2000 15:13:08 -0400
Subject: __del__ problem (destructor called After members of a class are deleted?)
Message-ID: <aw2L4.8907$HG1.245743@nnrp1.uunet.ca>

What's going on here?

Exception except.AttributeError: "'None' object has no attribute 'time'" in
<met
hod lwdb_base.__del__ of lwdb_hash instance at 5de410> ignored
Exception except.AttributeError: "'None' object has no attribute
'delete_DB'" in
 <method DBPtr.__del__ of DB instance at 5dfd00> ignored
Exception except.AttributeError: "'None' object has no attribute 'time'" in
<met
hod lwdb_base.__del__ of lwdb_base instance at 5de6d0> ignored


Why is __del__ being called with a self = None ?



Oversimplified Code Snippet:

class foo(bar):
    def close(self):
        # pretend I do stuff here

    def __del__(self):
        #print "lwdbtable.__del__"
        self.close()


Warren




From spex66 at my-deja.com  Fri Apr 28 11:20:25 2000
From: spex66 at my-deja.com (spex66 at my-deja.com)
Date: Fri, 28 Apr 2000 15:20:25 GMT
Subject: How can I get the name of an object???
References: <8e712k$65c$1@nnrp1.deja.com> <etd8zy03m44.fsf@w20-575-108.mit.edu> <8e9090$c4d$1@nnrp1.deja.com> <slrn8gglsv.ep7.scarblac-spamtrap@flits104-37.flits.rug.nl>
Message-ID: <8eca76$19c$1@nnrp1.deja.com>

corrected code version if my cancel haven't worked:

In article <slrn8gglsv.ep7.scarblac-spamtrap at flits104-37.flits.rug.nl>,
  scarblac-rt at pino.selwerd.nl wrote:
> spex66 at my-deja.com wrote in comp.lang.python:
> > In article <etd8zy03m44.fsf at w20-575-108.mit.edu>,
> >   Alex <cut_me_out at hotmail.com> wrote:
> > >
> > > An object could be referred to by any number of names:
> > >
> > > dd = zz ()
> > > ee = dd
> > >
> > > Which name for the object do you wish to know?  ee or dd?
> >
> > dd or in other words, the name of the FIRST made instance...
>
> What would happen in the following cases:
> 1)
>   dd = zz()
>   ee = dd
>   del dd
>   print ee.__name__
> 2)
>   dd = [xx(), yy(), zz()]
>   print dd[2].__name__
> 3)
>   print zz().__name__
>
> Classes have a name (the one used in their class: statement).
> Instances have no name, period.
OK, I understand this point very well :-) there are enough good examples

> But why would you need it? In most cases where people try to use
something
I need it to build navigation-code on the fly for "exec" command. Build
on a knitted database API. I have an instance of the navigation-class,
but for the "exec"code I need the name of the *instance*.

Good for me: there is only one instance necessary, but I don't wanted
to hardcode the name, cause an interactive use i sto restricted.

So I have actually a new approach: a "shadow" instance, of the first
navigation instance.

#FILE: shadow.py
"""arghh, last 45min I implemented the __setattr__, I don't needed in
my example, but I think it's nice to play with
"""
class shadow:
    #BUILD for internal access on the given instance with a reliable
name
    #ONLY practible if there is ONE instance of a class
    #Needed for internal code-building purposes
    def hide(self, realinstance):
        self.internal = realinstance

    def __getattr__(self, attrName):
        # Get any other interesting attributes from the base class.
        if (attrName <> 'internal'):
            return self.internal.__dict__[attrName]
        else:
            if self.__dict__.has_key('internal'):
                return self.__dict__['internal']

    def __setattr__(self, attrName, value):
        if (attrName <> 'internal'):
            setattr(self.internal, attrName, value)
        else:
            self.__dict__['internal'] = value

sw = shadow()
shadowName = 'sw'

class navigation:
    def __init__(self):
        self.name = 'hallo'
        sw.hide(self)
#FILE:END

so I can write at the prompt:
>>> from shadow import * #IMPORTANT for me
>>> yy = navigation()
>>> yy.name
>>>  'hallo'
>>> sw.name
>>>  'hallo'
>>> sw.test = 7
>>> yy.test
>>> 5

and I have my fixed access for the navigation instance over shadowName
>>> exec "%s.name = '%s'" % (shadowName, 'hi')
>>> yy.name
>>> 'hi'

hmm, and it work for all classes with one instance (I hope :))
and a nice excercise for __getattr__ and __setattr__

greetings
Peter
(=PA=)

> like this, dictionaries are a solution.
>
> --
> Remco Gerlich


Sent via Deja.com http://www.deja.com/
Before you buy.


From nmm1 at cus.cam.ac.uk  Tue Apr 11 10:39:17 2000
From: nmm1 at cus.cam.ac.uk (Nick Maclaren)
Date: 11 Apr 2000 14:39:17 GMT
Subject: Processing BCD and Signed Integer Fields with Python
References: <38f334ea@wwwproxy3.westgroup.com>
Message-ID: <8cvdel$lb9$1@pegasus.csx.cam.ac.uk>

In article <38f334ea at wwwproxy3.westgroup.com>, "Chris Lada" <chris.lada at westgroup.com> writes:
|> We have been receiving record/field-based information from a data provider.
|> Some of the fields are BCD (Binary Coded Decimal), and some are signed
|> integer. The data originates from a mainframe/Cobol(?) system. The record
|> layouts have definitions like "PIC S99999" (signed integer), and "PIC S99999
|> Comp-3" (BCD). Has anyone had to process this sort of data with Python ?
|> Can Python handle this sort of data type ?  Anyone have suggestions on how
|> to approach this problem ?

It's pretty easy handling anything like that, PROVIDED that you have
the data intact.  Let us assume that is IBM 360/370/390 MVS or CMS,
then it originated in a record-oriented system, where ALL characters
are valid in records.  99% of people manage to mangle the data when
transferring it to Unix or Unix-like systems (including Microsoft
systems), and lose information.

If you can avoid that, it is just like handling any field-oriented
data, except in binary rather than human-readable text.  This remark
applies to any sufficiently powerful language - C, Perl, Python and
so on.


Regards,
Nick Maclaren,
University of Cambridge Computing Service,
New Museums Site, Pembroke Street, Cambridge CB2 3QG, England.
Email:  nmm1 at cam.ac.uk
Tel.:  +44 1223 334761    Fax:  +44 1223 334679


From tim.hochberg at ieee.org  Tue Apr 11 10:37:07 2000
From: tim.hochberg at ieee.org (Tim Hochberg)
Date: Tue, 11 Apr 2000 14:37:07 GMT
Subject: Huh? func_defaults, default values in function calls
References: <8cv5qj$35p$1@zinc.intern.netconnect.no> <8cv8ra$hio$1@pegasus.csx.cam.ac.uk>
Message-ID: <nMGI4.42394$U4.182659@news1.rdc1.az.home.com>

"James "Wez" Weatherall" <jnw22 at cam.ac.uk> wrote in message
news:8cv8ra$hio$1 at pegasus.csx.cam.ac.uk...
> "Roger Baklund" <roger at netconnect.no> wrote in message
> news:8cv5qj$35p$1 at zinc.intern.netconnect.no...
> [snip of some code]
> >
> > I don't get it. Why isn't the dictionary reset to {'x':'a'} when i call
> > the function the last time?
> >
> > >>> a.func_defaults
> > ('', '', {'x': 'a', 'y': 'c'})
> >
> > Why is not {} the default for the z parameter?
>
> According to my understanding of the Python docs, defaults are only
> evaluated at function definition time, not at function invokation time.
>
> As a result, if you set a mutable object (list, dictionary, etc) as a
> default, it'll be set when the function is defined and then if the
function
> changes it, it'll stay changed for subsequent invokations.  The function
> effectively has a "side-effect", of modifying it's own default!


That's exactly right. The idiomatic Python way of doing what Roger wants
here is:

def a(x='',y='',z=None):
   if z is None:
      z = {}
   if x:
      z['x'] = x
   if y:
      z['y'] = y
   print str(z)

I think this is the first time I've seen show up with dictionaries, usually
is shows up with lists.

-tim




From famvandenberg at home.com  Tue Apr 25 22:50:51 2000
From: famvandenberg at home.com (bo Vandenberg)
Date: Wed, 26 Apr 2000 02:50:51 GMT
Subject: python-only crypt()?
References: <Pine.BSF.4.10.10004230627580.5642-100000@chi.pair.com>
Message-ID: <39065A08.F776CBE1@home.com>

I can't tell if this is the sort of thing you want but try:


"a pure Python implementation of the Unix crypt(3) algorithm"

http://home.clear.net.nz/pages/c.evans/sw/



Post back if this is helpful



From smalleys at gte.net  Mon Apr 24 19:21:35 2000
From: smalleys at gte.net (sue)
Date: Mon, 24 Apr 2000 23:21:35 GMT
Subject: SWIG
References: <slrn8g495g.svu.stevey@polgar.cnd.hp.com>
Message-ID: <3904D716.67C19ECD@gte.net>

It's on SourceForge now.


From scarblac-spamtrap at pino.selwerd.nl  Thu Apr 27 10:28:28 2000
From: scarblac-spamtrap at pino.selwerd.nl (Remco Gerlich)
Date: 27 Apr 2000 14:28:28 GMT
Subject: Simple Function Question
References: <8e7ddt$18n@autodesk.autodesk.com>
Message-ID: <slrn8ggjiv.el7.scarblac-spamtrap@flits104-37.flits.rug.nl>

Akira Kiyomiya wrote in comp.lang.python:
> Hi,  Could you explain this in dummy's language?
> 
> 
> # Return a function that returns its argument incremented by 'n'
> def make_incrementer(n):
>     def increment(x, n=n):
>         return x+n
>     return increment
> 
> add1 = make_incrementer(1)
> print add1(3)  # This prints '4'

Let's try simpler versions. Do you understand the following?

# Return a function that returns its argument incremented by 3
def make_incrementer():
   def increment(x):
      return x+3
   return increment
   
Now this one:

# Return a function that adds its two arguments together, or by default adds 3
def make_incrementer():
   def increment(x, y=3):
      return x+y
   return increment
   
And finally, the version where you can give another default value:

# Return a function that adds n to its argument:
def make_incrementer(n):
   def increment(x, y=n):
      return x+y
   return increment
   
The last one is equivalent to your version, except here one variable is
called y instead of n. Maybe it was confusing that the previous one used 'n'
for two different variables.

-- 
Remco Gerlich,  scarblac at pino.selwerd.nl
'Oook?'


From mpetroni at my-deja.com  Tue Apr  4 05:11:38 2000
From: mpetroni at my-deja.com (mpetroni at my-deja.com)
Date: Tue, 04 Apr 2000 09:11:38 GMT
Subject: Determining Classname of an object
Message-ID: <8ccbk4$b6p$1@nnrp1.deja.com>

Is there a way to find out from which class an object has been created?
Tanx!


Sent via Deja.com http://www.deja.com/
Before you buy.


From jcm at bigskytel.com  Wed Apr  5 23:56:04 2000
From: jcm at bigskytel.com (David Porter)
Date: Wed, 5 Apr 2000 21:56:04 -0600
Subject: Python for QNX?
In-Reply-To: <38EBAB2B.BDA6D20F@boeing.com>; from tom.mcdavid@boeing.com on Wed, Apr 05, 2000 at 09:07:55PM +0000
References: <38EBAB2B.BDA6D20F@boeing.com>
Message-ID: <20000405215604.C6713@bigskytel.com>

* Tom McDavid <tom.mcdavid at boeing.com>:
> Has anyone ported python to qnx4?

ftp://ftp.qnx.com/usr/free/qnx4/os/language/python-1.5.tgz

For future reference, go to http://www.python.org/download/ (which is where
I looked).

    /david



From psl at mitre.org  Thu Apr 27 13:06:45 2000
From: psl at mitre.org (Pete)
Date: Thu, 27 Apr 2000 13:06:45 -0400
Subject: Py_BuildValue
Message-ID: <390873A5.E94E5376@mitre.org>

Does anyone know where Py_BuildValue lives,  or what arguments it takes?


From python-list at teleo.net  Fri Apr  7 10:35:22 2000
From: python-list at teleo.net (Patrick Phalen)
Date: Fri, 7 Apr 2000 07:35:22 -0700
Subject: CGI Help Needed
In-Reply-To: <8ck9uo$ga08@imsp212.netvigator.com>
References: <8ck9uo$ga08@imsp212.netvigator.com>
Message-ID: <00040707452909.01234@quadra.teleo.net>

[Paul Gresham, on Fri, 07 Apr 2000]
:: What is the normal and respected method of User authentication on a web
:: site, how should I present a login box when the user is not authenticated
:: already, and how do I maintain a session with an authenticated user, without
:: using cookies, allowing them to continue surfing around the web site ?

Have you looked at Zope (http://www.zope.org)? Zope is essentially a
very complete framework for building web applications, written in
Python.

Zope is a long running process, has a sophisticated
authentication/authorization model, and uses acquisition technology to
manage the sort of thing you're looking for. It also has a SQL Session
product to maintain user state across sessions in a SQL database.



From mwh21 at cam.ac.uk  Thu Apr 20 08:04:05 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 20 Apr 2000 13:04:05 +0100
Subject: Confused by List behavior
References: <8dmqsd$6m0$1@nnrp1.deja.com>
Message-ID: <m3og75geru.fsf@atrus.jesus.cam.ac.uk>

ejbaker_seekonk at my-deja.com writes:

> As a python newbie, I entered this script to explore list processing
> behaviours.
> 
> yy = dir()
> for element in yy:
>   print element
> 
> for element in yy:
>   print dir(element)
> 
> The first 'for' loop works as expected. It returns,
> ['__builtins__', '__doc__', '__name__'].
> 
> The second loop, however, returns three empty lists. For the life of
> me, I can't understand that. Becauese, if I type 'dir(__builtins__)',
> in a script, I get all of the elements of __builtins__.

dir("__builtins__") != dir(__builtins__).

Maybe you want "getattr"?  Eg.

for element in dir(ob):
    print dir(getattr(ob,element))

Cheers,
M.

-- 
  On the other hand,  the following areas are subject to boycott
  in reaction to the rampant impurity of design or execution, as
  determined after a period of study, in no particular order: 
    ...                            http://www.naggum.no/profile.html


From darrell at rubina.funkomite.com  Tue Apr 11 13:28:02 2000
From: darrell at rubina.funkomite.com (darrell at rubina.funkomite.com)
Date: Tue, 11 Apr 2000 17:28:02 GMT
Subject: simple httplib and urllib timeout question
Message-ID: <CgJI4.14892$eh.1040983@news.bc.tac.net>

Just a quick question, im writing a small webbot for a python learning experience and it comes across some sites that will make it hang for a while, if the site is down or extremely slow etc,  is there a way to set a faster timeout for httplib . Or is there a better way to handle it ?


Thanks in advance.

Darrell


From brent.fulgham at xpsystems.com  Tue Apr 11 20:23:58 2000
From: brent.fulgham at xpsystems.com (Brent Fulgham)
Date: Tue, 11 Apr 2000 17:23:58 -0700
Subject: [Request for Information] Python Thread Core
Message-ID: <EDFD2A95EE7DD31187350090279C676770744B@THRESHER>

I heard a rumor that some supercomputing types were contemplating
working on the Python core to enable true multi-threading.  If this
is true, could one of you contact me?  I'd like to discuss some
threading issues/possibly help out.

Thanks,

-Brent



From eviltofu at rocketmail.com  Wed Apr 26 14:59:06 2000
From: eviltofu at rocketmail.com (Jerome Chan)
Date: Wed, 26 Apr 2000 18:59:06 GMT
Subject: How can I get the name of an object???
References: <8e712k$65c$1@nnrp1.deja.com> <390731A2.12788170@freei.com>
Message-ID: <eviltofu-66EBF2.14590526042000@news.earthlink.net>

object.__name__ returns a tuple of all vars referencings the object?


From shae at webwitches.com  Tue Apr 18 03:20:55 2000
From: shae at webwitches.com (Shae Erisson)
Date: Tue, 18 Apr 2000 07:20:55 GMT
Subject: Project idea - python
References: <BxGK4.811$9B.103178@news.bc.tac.net>
Message-ID: <38FC0E72.9AB025DB@webwitches.com>

Parag Shah wrote:
> 
> hi,
>    i am planning to do a project for my bachelors where i wish to write a translator for a language (undecided at
> the moment ) which generates java bytecode. how do i go about doing it???
> thanx in advance
> cheers

read up on jpython, it uses JavaCC.

http://www.jpython.org/docs/compile.html
-- 
sHae mAtijs eRisson (sHae at wEbwitchEs.coM) gEnius fOr hIre
   bRing mE fIve sQuirrels aNd nO oNe wIll gEt hUrt
13:00pm up 1 season, 1 squirrel, load average: 1+1j 0.5+2j 0.2+0.5j


From mwh21 at cam.ac.uk  Wed Apr 12 14:22:52 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 12 Apr 2000 19:22:52 +0100
Subject: confused about path
References: <38F49297.274ABBAE@tcac.net>
Message-ID: <m366tni3gj.fsf@atrus.jesus.cam.ac.uk>

Albert Wagner <alwagner at tcac.net> writes:

> Assuming I am set up correctly, shouldn't I be able to invoke idle like
> this:
> 
> $ python idle.py
> 
> from any directory?

No.

You might want to try "python -c 'import idle'", but that's not quite
the same.  Or put a symlink to /usr/local/whatever/Tools/idle/idle.py
on your path.

Cheers,
M.

-- 
  incidentally, asking why things are "left out of the language" is
  a good sign that the asker is fairly clueless.
                                      -- Erik Naggum, comp.lang.lisp


From mwh21 at cam.ac.uk  Tue Apr  4 06:24:11 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 04 Apr 2000 11:24:11 +0100
Subject: From Python Newbie to Python Guru on the Fast Track:  HOW?
References: <38E9B48F.F1827543@acm.org>
Message-ID: <m3aejayxis.fsf@atrus.jesus.cam.ac.uk>

"Robert W. Cunningham" <rcunning at acm.org> writes:

> Well, I just bit off more than I can chew.  While I know barely enough
> Python to patch the occasional errant script, I've never written a thing
> of my own.
> 
> Now I need to get up to speed FAST to design and develop a
> cross-platform UI interoperability environment for Windows-to-Linux that
> will allow much tighter integration of Windows apps running on a Linux
> desktop, no matter if the Windows app is running locally under Win4Lin
> or VMware, or is running remotely on a machine on a LAN or WAN or even
> the Internet (via a remote-control app such as VNC).
> 
> I've selected Python simply due to the massive support it enjoys on both
> platforms (Windows Python is nothing less than amazing), the simplicity
> of the language itself, and it's ability to be used for fairly large
> distributed programs (via DOPY or PYRO or Fnorb).  However, when the
> Windows app is running locally, all unnecessary communication layers
> should be eliminated, since there is only one CPU available to run
> Linux, the Windows app, and this interface (making for one very busy
> CPU).

However, the Windows world still has to talk to the Linux world;
you're going to need a Python process in each and some way of them
talking to each other - like a socket?  I'd have thought you'd want to
leverage someone else's work in this area though - have you thought
about xmlrpc?

> The goal will be to provide only the connective interface needed to
> bridge "the User Interface Gap" in the simplest way possible, even if
> that way is not optimal at the start.  For example, all shared
> filesystems will be either vfat (if local) and/or NFS and Samba (until
> Windows gets native ext2 support ;^)).  This will allow the
> implementation of services such as Drag&Drop to avoid having to
> transport the dragged item itself, if it is a file, and will allow files
> to be used as an intermediary mechanism (including on RAM disks, to
> simplify or eliminate any shared memory issues).

Wow, you're certainly aiming high.

> Ultimately, I hope to expose Python Linux interfaces to many Windows
> services, ones that would permit dynamic application interaction (such
> as via ActiveX or OLE/COM), and maybe even to give Linux access to the
> Windows Media Player.  The primary focus will be two-fold:  1)  Make it
> easier for Windows apps to be sued from the Linux desktop.  2)  Give
                                ^^^^
Quality typo :-).

> Linux full access to Windows applications that have no true Linux
> equivalent (such as MathCad, LabView, TurboTax, Quicken, FaxTalk
> [answering machine + speakerphone + fax + BBS + ...], etc.).  One thing
> I like about this entire approach is that it makes Windows seem more
> like an optional Linux add-on than a competitor!

Hrm; I can't really see how you'd do the more visual things here; and
they're the things you'd want Windows for the most.

> I'm an embedded real-time systems wonk, with lots of hardware how-to (I
> hate device drivers with a passion - I've written far too many),
> fantastic debugging skills, a decade of design and programming, a thin
> layer of OOP, and very little Python experience.  I already program in
> at least 10 languages, so picking up a new one won't be hard.  But
> learning to use it well can be a bit harder, not to mention all the
> extension packages.

Generally speaking, one of the reasons Python is great is that doesn't
get in the way of what you're trying to do, ie. the difficulty of
programming something in Python is often closer to how hard the
problem actually is than with other languages.  But I think you've
chosen a pretty hard problem...

> As for the time angle:  I'm self-employed, and presently between
> clients.  I have at least a few weeks to devote 100% to this effort,
> maybe more.  At *least* a hundred hours or so, just to get started.
> 
> Then there's the money angle:  I don't mind spending $100 on this, but
> I'd like to keep it minimal (I also need to get a new system this month
> - this one is on its last legs).
> 
> Given where I am, and where I hope to go, what is the fastest way to
> fill in the gaps in my knowledge?

I don't think you could go far wrong by buying the "Python Essential
Reference" by Beazley.

> I'm looking for the best books and tutorials that will do the job, with
> my only access to other Python expertise being this newsgroup (or any
> other similar forum that may be recommended).  I will be completely
> unable to evaluate these resources until AFTER I've put substantial time
> into the effort, so I'd really like to avoid the lemon I'd likely wind
> up with if I shop randomly on FatBrain.  And I need to proceed to solid
> discussions of (and practice with) the various Python packages fairly
> quickly.  Please also include any on-line classes that are known to be
> fast-paced and effective.
> 
> Right now, I'm reading everything on Python.org, since I'll need to be
> familiar with that documentation no matter what else I use.
> 
> Suggestions?

Look for four-leafed clovers?  I wish you luck, but I think you may
have chosen a task for yourself that is pretty close to immpossible.

enjoy-yourself-anyway-ly y'rs
M.

-- 
well, take it from an old hand: the only reason it would be easier
to program in C is that you can't easily express complex  problems
in C, so you don't.                 -- Erik Naggum, comp.lang.lisp


From embed at geocities.com  Mon Apr 17 11:57:35 2000
From: embed at geocities.com (Warren Postma)
Date: Mon, 17 Apr 2000 11:57:35 -0400
Subject: ODMG 3.0 and Python
Message-ID: <QyGK4.8376$HG1.232986@nnrp1.uunet.ca>

I was just reading at the ODMG's home page (www.odmg.org) that they have
released ODMG 3.0, and it represents a not-so-subtle-shift in what they
consider the ODMG standard to involve. I quote from the disclaimer added
above the existing online documentation which covers only the "2.0"
standard.

"Note: This text is for ODMG 2.0.  The new release, ODMG 3.0, was published
in January
2000.   The following text does not reflect the ODMG's charter change.  The
ODMG is now
an Object Storage API standard that can work with any DBMS or tool.  We are
in the process of changing all references to Object Databases (ODBMS) to
Object Data. "

Isn't it neat when a standards body actually ADMITs it is time to
refactor/rename their
standard?

Anyways... Is anybody interested in an ODMG-3.0 compliant database for
Python?
That would be a LARGE undertaking, but perhaps we could Borgle
(borrow+burgle) bits of gadfly to help out, such as the excellent kjbuckets,
parser, and other similar bits. As for the underlying data store, persisting
Python objects is already trivial, so adding a variety of persistence
techniques (b+tree of marshalled or pickled objects, etc) would not exactly
be Difficult.  A certain amount of C++ and Smalltalk Bias would have to be
overcome, but that's not insurmountable.

Warren




From jstok at bluedog.apana.org.au  Wed Apr 19 11:25:08 2000
From: jstok at bluedog.apana.org.au (Jason Stokes)
Date: Thu, 20 Apr 2000 01:25:08 +1000
Subject: Very Horrible Question About Goto's
References: <6325E90C8B6ED21189080001FA7E2ADA0D4051@IBA1>
Message-ID: <2hkL4.8514$5D.18609@ozemail.com.au>

    >Richard Jones wrote in message
><6325E90C8B6ED21189080001FA7E2ADA0D4051 at IBA1>...
    >I'm trying to create a converter from our own internal language
(horrid) >into python.  My only problem so far is that the original language
uses >*LOTS* of goto's.  Is there any way that this can be replicated within
>python ?


    Yes and no.  It should be possible to translate arbitary gotos into
Python constructs, but the results won't be readable by a human programmer.
I'm not quite sure how right now, but I could think about it.  Because the
Goto is such an unstructured construct, it's hard to translate it into
structured terms without bizarre circumlocutions.

    If your desire is simply to use Python as a backend, a translator will
work.  If your desire is to produce something that can be maintained by a
human, I'm afraid you're probably out of luck.  Automatic translation
between computer languages is rarely successful; your best strategy is to
refactor, by

    1) Rewriting the code; or
    2) Encapsulating the legacy code by creating an interface to it that can
be accessed by future Python programs (eg through CORBA or COM, or less
heavyweight-l-y, C bindings).  The facade pattern, basically.





From kuncej at mail.conservation.state.mo.us  Fri Apr 14 12:59:43 2000
From: kuncej at mail.conservation.state.mo.us (Jeffrey Kunce)
Date: Fri, 14 Apr 2000 11:59:43 -0500
Subject: file time function symmetry
Message-ID: <s8f7084d.031@mail.conservation.state.mo.us>

To obtain the time when a file was messed with:
    lastAccessTime = os.path.getatime(path)
    lastModifiedTime = os.path.getmtime(path)

To change the time when a file was messed with:
     os.utime(path,(lastAccessTime,lastModifiedTime))

Is there any reason why these functions
are not more symmetrical?

  --Jeff





From kern at caltech.edu  Thu Apr 13 16:57:33 2000
From: kern at caltech.edu (Robert Kern)
Date: 13 Apr 2000 20:57:33 GMT
Subject: Simple formatting string question
References: <meh9-637743.14563813042000@news.cit.cornell.edu>
    <8d57gp$1om@gap.cco.caltech.edu>
    <%7qJ4.929$rc9.193537024@newsb.telia.net>
Message-ID: <8d5cbt$9vt@gap.cco.caltech.edu>

In article <%7qJ4.929$rc9.193537024 at newsb.telia.net>,
	"Fredrik Lundh" <effbot at telia.com> writes:
> Robert Kern <kern at caltech.edu> wrote:
>> GuruQuestion:
>> Should implicit string concatenation have a different precedence than
>> explicit string concatenation?  Perhaps only a mention in
>> http://www.python.org/doc/current/ref/string-catenation.html ?
> 
> what do you mean?
> 
> that page describes "string literal concatenation", and clearly
> says that multiple adjacent string literals are concatenated at
> *compile* time.
> 
> operator precedence has nothing to do with it -- it's not an
> operator, and it's not documented as being one.

True, and I understand this subtlety.  The difference, however, may be lost
to newbies.  A clear description of the difference in the Reference and 
Tutorial might be benificial.

> </F>

-- 
Robert Kern
kern at caltech.edu

"In the fields of hell where the grass grows high
 Are the graves of dreams allowed to die."
  -- Richard Harter


From shogan at iel.ie  Tue Apr 25 10:30:12 2000
From: shogan at iel.ie (Shaun)
Date: Tue, 25 Apr 2000 15:30:12 +0100
Subject: HTML type form?
Message-ID: <003301bfaec2$c4529c60$2801000a@iel.ie>

how would i create a HTML type user interface using python which would take
in say, -name-address-credit card number- and record it with the intention
of struct.pack-ing the values entered,
thanks
Shaun
=====================
Shaun Hogan
Interactive Enterprise Ltd.
alt. E-mail : shaun_hogan at yahoo.com
+353 86 8342529






From tismer at tismer.com  Thu Apr  6 17:33:05 2000
From: tismer at tismer.com (Christian Tismer)
Date: Thu, 06 Apr 2000 23:33:05 +0200
Subject: Long Multiplication is not commutative.
Message-ID: <38ED0291.65CB7B5F@tismer.com>

Yikes!

No, it is computatively commutative, just not in terms
of computation time. :-))

The following factorial loops differ by a remarkable
factor of 1.8, and we can gain this speed by changing
long_mult to always put the lower multiplicand into the left.

This was reported to me by Lenny Kneler, who thought he had
found a Stackless bug, but he was actually testing long math. :-)

This buddy...

>>> def ifact3(n) :
...  p = 1L
...  for i in range(1,n+1) :
...    p = i*p
...  return p

performs better by a factor of 1.8 than this one:

>>> def ifact1(n) :
...  p = 1L
...  for i in range(1,n+1) :
...    p = p*i
...  return p

The analysis of this behavior is quite simple if you look at the
implementation of long_mult. If the left operand is big and the
right is small, there are much more carry operations performed,
together with more loop overhead.
Swapping the multiplicands would be a 5 line patch.
Should I submit it?

ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaunstr. 26                  :    *Starship* http://starship.python.net
14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     where do you want to jump today?   http://www.stackless.com



From btang at pacific.jpl.nasa.gov  Tue Apr 18 11:13:55 2000
From: btang at pacific.jpl.nasa.gov (Benyang Tang)
Date: Tue, 18 Apr 2000 08:13:55 -0700
Subject: control applications in Windows 95
Message-ID: <38FC7BB3.10BCEB80@pacific.jpl.nasa.gov>

Is there any way for python to log usage of certain applications (say MS Word) in Windows 95?


From embed at geocities.com  Wed Apr 26 12:33:18 2000
From: embed at geocities.com (Warren Postma)
Date: Wed, 26 Apr 2000 12:33:18 -0400
Subject: <nifty hieroglyphics here>
References: <8e4cc2$d5t@netnews.hinet.net>
Message-ID: <BWEN4.12235$HG1.329193@nnrp1.uunet.ca>

The official language of usenet newsgroups is English, and most most
newsreader software can't handle non-european fonts.  So, we see a stream of
meaningles symbols, so we're unable to help you out.




From m.faassen at vet.uu.nl  Fri Apr 21 19:45:34 2000
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 21 Apr 2000 23:45:34 GMT
Subject: Pythonware Website???? Gone on vacation with Starship? ;-))
References: <1255771354-36910521@hypernet.com>
Message-ID: <8dqp6u$plm$1@newshost.accu.uu.nl>

Gordon McMillan <gmcm at hypernet.com> wrote:
> Martijn Faassen wrote:

>> .... Also, since the PSU has access to
>> Guido's time machine and didn't prevent this, the PSU must be in on this.
>> Perhaps they're testing those samples of whitespace-eating nanoviruses that
>> they got from that UFO crash near Roswell back in the 50s.

(oops, it was 1947, not the 50s)

> Oh you silly boy. They're not alien. They simply used the time 
> machine to bring back samples from the future.

> if-i-told-you-how-i-knew-i'd-have-to-kill-you-ly y'rs

[Roswell, New Mexico, July 1947]

Gordon McMillan stood in the desert, looking sadly at the remains of Guido's
time machine.

"Guido won't like this," a slightly metallic voice said behind him.

"Um... no," Gordon said. "I guess he won't."

"Perhaps it wasn't a good idea after all to go back to the time of the
Roswell crash," the voice persisted.

Gordon turned to the bot. "But I wanted to *know*. It might have helped us
during the invasion in 2010. I wanted to know once and for all what had
happened here."

"And now you do! Hehehehe-ly." The bot pronounced its chuckling.

Behind them they heard the sound of a car. They both stood to watch the car
approach from a cloud of dust. "Tim!" Gordon said. "Go invisible, quickly!"

Mumbling to itself, the bot faded into thin air.
 
The car stopped, and a man came out. He stood staring at the wrecked time
machine for a moment, then looked at Gordon, open-mouthed.

"W-what is that?" the man asked. "One of them f-flying saucers?"

"Um," Gordon said, and gathered his mind. He flashed the PSU badge he
was always carrying. "Weather balloon. I'm with the weather bureau."

The man looked dubious, staring at the wreckage of the time machine.

The timbot became visible again after the man had left. "I don't think
you convinced him," it said.

"We'd better go in hiding for a while," Gordon said. 

"For some decades, at least. It will be a long time before technology 
develops far enough for us to build a new time machine. The blueprints
I have call for a Perl interpreter, for instance. We need spacetime
convolution regexes."

"Maybe we can help things move forward a bit, then."

"And we'd better hide the nanovirus samples as well. They're too dangerous,"
the timbot said.

"At least we'll have some time to analyze them now. I still don't think
they're alien."

"I think they are," the timbot said.

Arguing, they walked off into the sunset.

Decades passed while Gordon and the timbot did their work. The timbot sent
off some surreptitious messages to various people. John McCarthy got a
few hints on recursion. Bill Gates was funded by an anonymous investor. 
(so, they needed the money! what?!) Larry Wall received a pamphlet
on postmodern linguistic analysis. And one christmas, Guido van Rossum
received a thick envelope containing the complete plans for a working time
machine.
 
Do-you-have-to-kill-me-now?-ly yours,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From rkiendl at gmx.net  Fri Apr  7 08:57:36 2000
From: rkiendl at gmx.net (Robert Kiendl)
Date: Fri, 07 Apr 2000 14:57:36 +0200
Subject: Embedding Python: How To Hardwire Module Search Path
Message-ID: <38EDDB40.FC715534@gmx.net>

I've embedded python into a C++ - windows app by a
PyImport_AppendInittab() +  Py_Initialize() method.

Py_Initialize() calculates a module search path from the
windows-registry settings.

I want to prevent Python from doing that and want to hardwire the search
path to a app. relative path ("<app>/pythonmod") which points to a
minimal python bunch to avoid the need for installing python.

Does anybody know what to do prior to Py_Initialize(). Or am I
completely wrong?
PySys_SetPath() crashes when called before Py_Initialize(). 

Robert


From johann at physics.berkeley.edu  Wed Apr 26 22:16:55 2000
From: johann at physics.berkeley.edu (Johann Hibschman)
Date: 26 Apr 2000 19:16:55 -0700
Subject: new linereading standard?
References: <3904D21B.333DD930@visionart.com> <390510C9.A744D22E@yahoo.com> <R2g+MIAFMUB5EwUl@jessikat.demon.co.uk> <mt1z3uyr8k.fsf@astron.berkeley.edu> <26042000.2@sanctum.jae.ddns.org>
Message-ID: <mthfcob84o.fsf@astron.berkeley.edu>

Juergen A Erhard writes:

>>>>>> "Johann" == Johann Hibschman <johann at physics.berkeley.edu> writes:
> [J W B]


Johann> So, basically, I find the functional stuff much more
Johann> useful from the command line than I do from actual
Johann> programs.  I think of the right transform, then I apply
Johann> it.  Interesting.

> Actually, I find the third example more readable than the other two...

>    map(string.strip, lines)

> is, IMHO, more readable than a for loop doing the same.

I agree.  For a single operation, that is.  However, I find a for loop
expressing successive transformations of a single data item to be
easier, once things get more complicated than a single map.  Just
taste, though.

> Though it also depends on what you're used to... (maps etc are *more*
> readable in Lisp... than in Python)

But even in Lisp (by which I assume you mean Common Lisp), I find that
dolist and the loop macro are often more readable than using mapcar
and friends.  In Scheme, you don't have dolist and loop, and I find
readability suffers.  (I'm not so certain about loop; it's a monster.
But I like dolist.)

Thinking vaguely that if Common Lisp were more common, python wouldn't
be nearly as clear a win as it is...

--Johann

-- 
Johann Hibschman                           johann at physics.berkeley.edu


From tismer at tismer.com  Tue Apr 25 16:00:11 2000
From: tismer at tismer.com (Christian Tismer)
Date: Tue, 25 Apr 2000 22:00:11 +0200
Subject: XML Schemas and Python?
References: <3905F588.124AF24E@darmstadt.gmd.de>
Message-ID: <3905F94B.9AD9CDB3@tismer.com>


Robb Shecter wrote:
> 
> Hi,
> 
> Has anyone done anything with schemas?  I have a research project
> originally written in Java, but the more I work on it, the more it
> mutates into a Python project... :)
> 
> Anyhow, if there's any sample code or articles out there, I'd like to
> hear about it.

Please refer to the XML sig, where this is just under
discussion.

ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaunstr. 26                  :    *Starship* http://starship.python.net
14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     where do you want to jump today?   http://www.stackless.com



From akuchlin at mems-exchange.org  Thu Apr  6 12:55:29 2000
From: akuchlin at mems-exchange.org (Andrew M. Kuchling)
Date: 06 Apr 2000 12:55:29 -0400
Subject: Error in "pow" on Itanium
References: <38EBF5EE.98B905A@cup.hp.com>
Message-ID: <3dsnwzuq2m.fsf@amarok.cnri.reston.va.us>

Robert Hundt <rhundt at cup.hp.com> writes:
> Running Python 1.5.2 as a 64-bit process on Itanium delivers an error in
> test_pow. The test delivers different results whenever run! Sometimes it
> even succeeds!

pow() does different things for integers, floats, and long integers;
which is the version that fails?  The float version in
Objects/floatobject.c seems to just check for special cases and then
calls pow() from the C library, so that's probably OK.  Either the
integer or long integer versions might well run into trouble on a new
64-bit platform, though.

-- 
A.M. Kuchling			http://starship.python.net/crew/amk/
And I suppose if everybody cut off their heads and stuffed asafoetida into
their mouths and buried their hearts at crossroads, then you'd do it too?
  -- The grandfather in SANDMAN #38: "The Hunt"



From greg at cosc.canterbury.ac.nz  Wed Apr 12 23:26:18 2000
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Thu, 13 Apr 2000 15:26:18 +1200
Subject: need help just begining
References: <11822ca0.dd34a400@usw-ex0109-068.remarq.com> <38F41190.C0BB507B@sympatico.ca>
Message-ID: <38F53E5A.977913FE@cosc.canterbury.ac.nz>

R E Heath wrote:

> SheaMon wrote:
> >
> > can
> > I just leave the numbers and stuff like that and just edit
> > the rest? I get so lost in the number part of the
> > programming.

> Python is an exceedingly good programming language but not really the
> thing for a complete newcomer to start with. I'd suggest you start with
> one of the older versions of BASIC such as MBASIC, BASICA, or QBASIC

I don't think that's necessarily good advice. If
numbers turn you off, traditional BASIC is going to
turn you off even more. You can do quite a lot of
useful things in Python without ever going near a
number -- something that can't be said about 
traditional BASIC!

It's not that Python isn't suitable for newcomers
to programming -- I and many others regard it as even
more suitable than BASIC -- it's that the tutorial
which comes with Python isn't designed for such
people. It's also unfortunate that it introduces
so many key concepts using heavily number-oriented
examples, when other subject matter could just as
easily have been chosen.

There are introductory books to Python around. I
haven't had personal experience of any of them, but
many people seem to think that "Learning Python"
is reasonably good. Whether it's any less biased
towards number-intensive examples I don't know.
I'm sure others will offer advice on the matter.

Anyway, to answer your question, Shea, you can try
skimming over any examples that have too many
numbers in them for your taste, but you may miss
out on some important concepts if you skip too
may examples.

-- 
Greg Ewing, Computer Science Dept,
+--------------------------------------+
University of Canterbury,	   | A citizen of NewZealandCorp, a	  |
Christchurch, New Zealand	   | wholly-owned subsidiary of USA Inc.  |
greg at cosc.canterbury.ac.nz	   +--------------------------------------+


From neelk at brick.cswv.com  Fri Apr 21 22:51:48 2000
From: neelk at brick.cswv.com (Neel Krishnaswami)
Date: 22 Apr 2000 02:51:48 GMT
Subject: The Simple Economics of Open Source
References: <14591.15630.192472.440375@goon.cnri.reston.va.us> <slrn8fv2rd.1kt.neelk@brick.cswv.com> <3900611F.3B525BDA@cfdrc.com>  <FtDG4z.774@world.std.com>
Message-ID: <slrn8g26tu.p7o.neelk@brick.cswv.com>

Will Ware <wware at world.std.com> wrote:
> Ed (elb at cfdrc.com) wrote:
> > ... I didn't think much of the article...
> > ...they gave the impression that programmers behave
> > somewhat like monkeys, motivated primarily by concerns
> > about group status and dominance...
> 
> They also mentioned, but immediately belittled, the possibility
> that altruism might be a real motivation. They may have been
> projecting their own thinking upon programmers in doing so. 

One of the interesting things about this thread is that it has given
me a bit of insight into what economics looks like to people who
haven't internalized its assumptions. I'm afraid that I'm tainted
enough that it no longer looks like the dismal science to me. :)

Seriously, the rejection of altruism as a possible motive strikes me
as a very reasonable decision. Altruism is generally not sufficient
motive in other aspects of life -- and since there's no particular
reason that programmers in particular are more likely to be altruistic
than the rest of the population, positing that hackers have greater
virtue is at best a dubious proposition. Consider, for example, how
much better government would be if all the voters studied the issues
and took care to be informed voters, and that basically no one makes
the effort. Most people don't even bother digging up evidence to
support their prejudices, let alone try to come to a careful decision.
And searching the Census and BLS websites is *easier* than hacking on
the Python core.

Second, economists assume that everyone is basically self interested,
including themselves -- it's a standard analytic assumption, not
snobbery. I read an article last year where two economists developed a
model predicting that a disproportionate fraction of groundbreaking
academic papers should be wrong. Basically, they assumed that since
funding stems from status, and status stems from writing ground
breaking papers, there's an incentive to draw flashy, counterintuitive
conclusions rather than confirming prior work. (They did observe that
their work implied their paper was probably wrong, since it was
advancing a flashy, counterintuitive proposition. :)


Neel


From mwh21 at cam.ac.uk  Wed Apr  5 11:33:26 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 05 Apr 2000 16:33:26 +0100
Subject: Pass by reference ?
References: <38E864C9.E1EE94F7@ecs.soton.ac.uk> <38EAF8D8.4B7C43B4@ecs.soton.ac.uk> <38EB56AE.F938CE06@acm.org>
Message-ID: <m33dp0muk9.fsf@atrus.jesus.cam.ac.uk>

"Robert W. Cunningham" <rcunning at acm.org> writes:

> Python can either 1) Pass a reference to the original data, or 2) Pass a
> reference to a copy of the original data.  At no time does Python "pass
> the data" directly (as in the "C" sense of, say, a register variable),
> since everything is treated as though it were (in some sense) an object
> (even if not "really" an object of the First Water).
> 
> Which, when mapped to conventional CS terminology (as extended by Moi),
> becomes "Pass By Reference" and "Pass By Clone".
> 
> Clearly, "Pass By Clone" makes the original data item completely
> inaccessible within the called function (ignoring bytehacks for now),
> and thus immutable.  For me, this helps to clarify things, since Python
> literally passes only references.
> 
> But, of course, that depends:  Is this a clarification, an obfuscation,
> or Just Plain Wrong?

I'm not sure I'm understanding you, but I'd say "Just Plain Wrong".
Python never ever copies anything when you pass it as an argument to a
function - it's just that sometimes the things you pass are immutable
- which is when it looks like call by value - and sometimes they're
mutable - when it looks a bit like call by reference.

Try to cast Python semantics in terms of what other langauges do is
confusing - I've just successfully confused myself trying (as you
could see if you could read my kill-ring) - and is hence to be
considered bogus, because Python's semantics are actually extremely
straightforward.

HTH,
M.

-- 
  well, take it from an old hand: the only reason it would be easier
  to program in C is that you can't easily express complex  problems
  in C, so you don't.                   -- Erik Naggum, comp.lang.lisp


From cut_me_out at hotmail.com  Thu Apr 13 15:02:49 2000
From: cut_me_out at hotmail.com (Alex)
Date: 13 Apr 2000 15:02:49 -0400
Subject: Simple formatting string question
References: <meh9-637743.14563813042000@news.cit.cornell.edu>
Message-ID: <etdem89u8me.fsf@w20-575-109.mit.edu>

> I'm trying to print something like this:
> 
> >>> print 6*'%2i' '%1i' % tuple([5,5,5,5,5,5,1])
> 
> but I'm getting an error.

This works:

print (6 * ('%2i') + '%1i') % tuple([5,5,5,5,5,5,1])

It's all a matter of operator precedence. :)

Alex.


From wware at world.std.com  Tue Apr 11 07:19:30 2000
From: wware at world.std.com (Will Ware)
Date: Tue, 11 Apr 2000 11:19:30 GMT
Subject: [ANNOUNCE] Garbage collection for Python
References: <20000407221750.A29279@acs.ucalgary.ca> <38F13FB9.D1D16AF@mindspring.com> <slrn8f2if1.dop.nascheme@cranky.arctrix.com> <87r9cdhdb0.fsf@deneb.cygnus.argh.org>
Message-ID: <FsunGI.BtB@world.std.com>

Florian Weimer (fw at deneb.cygnus.argh.org) wrote:
> Is it possible to extend your GC in a way which would permit the
> execution of several bytecode interpreters in multiple threads?

It sounds like you might want to simply run several copies of the
Python interpreter in separate processes, which isn't a GC issue.
Unix will be perfectly happy to do that, and WinNT should be OK.
Win95 and Win98 don't use "real" multitasking but it should be
possible to run at least a few simultaneous instances.

But the fact that you regard this as a GC issue might mean that you
are actually looking for very lightweight threads. Christian Tismer
and I have developed something that might be what you need, called
microthreads, which you can read about at:
http://world.std.com/~wware/uthread.html

If memory isn't really that big a deal, and you don't need thousands
of threads running simultaneously, you can use the thread/threading
modules, which use whatever threading is provided by your OS. These
modules have been around a long time and are much more thoroughly
debugged than microthreads.
-- 
 - - - - - - - - - - - - - - - - - - - - - - - -
Resistance is futile. Capacitance is efficacious.
Will Ware	email:    wware @ world.std.com


From smalleys at gte.net  Mon Apr 24 19:55:45 2000
From: smalleys at gte.net (sue)
Date: Mon, 24 Apr 2000 23:55:45 GMT
Subject: Installing Python on NT
References: <8e2b92$9n6$1@nnrp1.deja.com>
Message-ID: <3904DF17.DD3C79EB@gte.net>

Install on Win9x, rebundle and zip everything?
Sue.


From boncelet at udel.edu  Thu Apr  6 21:41:54 2000
From: boncelet at udel.edu (Charles Boncelet)
Date: Fri, 07 Apr 2000 11:41:54 +1000
Subject: Why should I switch to Python? - Infinity of Primes
References: <NDBBKEGCNONMNKGDINPFGEJDDFAA.infonuovo@email.com> <7F51E902E09F3F60.8F612607812ED16B.82CB8B41435E14FD@lp.airnews.net> <38EA369C.26A1C25@math.okstate.edu> <38EAC8C4.ADC25640@cosc.canterbury.ac.nz> <8cfq9f$2q5e$1@pc-news.cogsci.ed.ac.uk> <pcoog7najs8.fsf@math.ntnu.no>
Message-ID: <38ED3CE2.5503489@udel.edu>

Harald Hanche-Olsen wrote:
> Since this is a posted to a computer language group, after all, I

> include my own simple implementation of the Sieve in Haskell:
>

(Haskell code deleted)

>
> I wonder if someone can come up with an equally immediate solution
> (i.e., a straight translation from the mathematical formulation) in
> python?
>
> Indeed, I would be tempted to say that Haskell is better suited for
> learning (some kinds of) mathematics than python.

I don't know any Haskell, but a quick Python sieve is easy:

def sieve(n):
    """find primes up to n by sieve of Erasthones"""
    primes = [2]
    marked = [0]*n
    for p in primes:
        i = p
        while(i < n):
            marked[i] = 1
            i = i + p
        for q in range(p,n):
            if marked[q] == 0:
                primes.append(q)
                break
    return primes


>
> ------
> Charles Boncelet, University of Delaware,
> On sabbatical at ADFA, Canberra Australia,
> Home Page: http://www.ece.udel.edu/~boncelet/



From tim_one at email.msn.com  Sun Apr 23 23:23:33 2000
From: tim_one at email.msn.com (Tim Peters)
Date: Sun, 23 Apr 2000 23:23:33 -0400
Subject: 'database' search
In-Reply-To: <39036340.BE82BEAD@roguewave.com>
Message-ID: <000101bfad9c$79af53a0$692d153f@tim>

[Bjorn Pettersen, wants to do word completion]
> ...
> The smallest datastructure I've found so far, is a 'prefix dictionary',
> so if e.g. "aardwark" was in the database, the 'prefix dictionary' would
> contain:
>
>  pfdict = {
>   'a':        ['aardwark', ...],
>   'aa':       ['aardwark', ...],
>   'aar':      ['aardwark'],
>   'aard':     ['aardwark'],
>   'aardw':    ['aardwark'],
>   'aardwa':   ['aardwark'],
>   'aardwar':  ['aardwark'],
>   'aardwark': ['aardwark']
>  }
>
> The maximum size for this would be n * avg len of item, with the real
> size being much(?) smaller with sufficient prefix sharing...

Before going nuts with variations on prefix trees, try the attached.  This
uses a sorted list of all words, and does a binary search to find the ones
starting with a given prefix -- dirt simple.  The advantage to that is that
it works <wink>.

BTW, the amount of compression you can get from a tree (or trie) approach
depends on the language and the set of words, but when coding in Python
you're almost certainly going to consume more storage with all the
supporting hair (ancillary dicts and lists) than you'll save by sharing a
few initial bytes in the words.  Trees are used for speed reasons instead,
in intense searching applications (which yours is not).

> I have a relatively stable database of between 1500 and 5000 items

So even if they average 10 letters/item, this won't consume more than 50Kb.

> new insertions are probably going to be less than 100 during the
> lifetime of the database)

In 1.5.2 and 1.6, appending a new word to a sorted list and then sorting
again takes time linear in the # of words -- it's recognized as a special
case (of another special case <wink>), and doesn't do a full sort.  So the
method below is more efficient than you might think.

when-your-database-grows-100x-larger-ask-again-ly y'rs  - tim


from bisect import bisect

class Completer:
    def __init__(self, wordlist):
        """wordlist -> a Completer for the list of words."""
        wordlist.sort()
        self.words = wordlist

    def complete(self, prefix):
        """prefix -> alphabetical list of all words starting w/ prefix."""
        words = self.words
        # Find i, j s.t. all in words[i:j] start w/ prefix.
        i = bisect(words, prefix)
        # Subtle:  if prefix is in the list, i points *after* it.
        # Fiddle i to point before it.  The bisect module should
        # really have more functions to control what happens here.
        while i > 0 and words[i-1] == prefix:
            i = i-1
        j = i
        n = len(words)
        # note that .startswith() is a 1.6 feature
        while j < n and words[j].startswith(prefix):
            j = j+1
        return words[i:j]

    def addword(self, word):
        """Add a word to the list of words."""
        self.words.append(word)
        self.words.sort()





From herzog at online.de  Sun Apr 16 14:33:49 2000
From: herzog at online.de (Bernhard Herzog)
Date: 16 Apr 2000 20:33:49 +0200
Subject: Event triggering and weak references
References: <8dbif3$hm0$1@metro.ucc.usyd.edu.au>
Message-ID: <m3em85aoaa.fsf@greebo.nodomain.de>

"Ricardo Nogueira" <rnog2438 at mail.usyd.edu.au> writes:

[...]
> This is what I would like to do:
> Imagine I could have two or more Widgets open displaying a car speed value,
> say a numerical and a graphic display, if this car speed changes (ex. some
> one pushes the break), I would like to have any open widget on this value
> updated (on opening each window would register interest with this car
> speed).
> 
> Smalltalk implements this with:
> 
> The widget would register interest in one aspect of the car  (on opening
> over aCar):
>   aCar when:#speedChanged send: #updateSpeedDisplay to: self
> 
> And the car would just notify who is interested, if any (when its speed
> changes):
>   self trigger: #speedChanged

I've implemented something like that in Sketch. In that scheme, classes
that trigger events are usually derived from a Publisher class which
implementes two methods Subscribe and Unsubscribe to connect to events
and events are identified by strings and may have additional parameters.
You example would look like this (only the relevant methods):

class Car(Publisher):

	def set_speed(self, speed):
		if self.speed != speed:
			self.speed = speed
			# now issue a speed message. This will call all 
			# callbacks subscribed to "speedChanged"
			self.issue("speedChanged")

class Speedometer:

	def __init__(self, car):
		self.car = car
		# subscribe the update method to the "speedChanged" messages
		self.car.Subscribe("speedChanged", self.update)

	def update(self):
		speed = self.car.Speed()
		# update the display...


> But if any of this windows is not referenced anymore or crashes I would like
> the garbage collector clear it (the car would not be notified and his
> pointer to the window shouldn't keep the garbage collector form freeing the
> window memory).
> 
> Again, Smalltalk implements this with weak references, that doesn't count
> for the garbage collector.

In this particular case the displays are probably connected to some
specific widget, so its probably enough to have a callback for the
widget's "destroy" or "delete" event or whatever it's called in your
widget set and disconnect the event handlers there.
 
> Is there support for this in Python?

Not directly.

> Any suggestions?

The implementation of this in Sketch is pure Python and it has already
been used outside of Sketch in pybliographer with practically no changes
(the only changes were to get rid of some Sketch specific error
reporting).

The current version of this connector module is in GNOME CVS:
http://cvs.gnome.org/lxr/source/sketch/Sketch/Base/connector.py

It's only slightly different from the one in Sketch 0.6.5 or 0.7.6. (for
a URL see my .sig)

-- 
Bernhard Herzog   | Sketch, a drawing program for Unix
herzog at online.de  | http://sketch.sourceforge.net/


From rjroy at takingcontrol.com  Tue Apr 18 15:52:18 2000
From: rjroy at takingcontrol.com (Robert Roy)
Date: Tue, 18 Apr 2000 19:52:18 GMT
Subject: __del__ problem (destructor called After members of a class are deleted?)
References: <aw2L4.8907$HG1.245743@nnrp1.uunet.ca>
Message-ID: <38fcbad8.1006465296@news1.on.sympatico.ca>

Sometimes Oversimplified is wayyyy too oversimplified. 
What is bar? At least post a fragment that demonstrates the exception.
What you have posted is meaningless in context with your question. In
fact with the addition of a trivial bar class (class bar: pass) and a
pass statement in the close method, the code you have posted is
correct.

Throw us a bone to chew on :-)

Bob

On Tue, 18 Apr 2000 15:13:08 -0400, "Warren Postma"
<embed at geocities.com> wrote:

>What's going on here?
>
>Exception except.AttributeError: "'None' object has no attribute 'time'" in
><met
>hod lwdb_base.__del__ of lwdb_hash instance at 5de410> ignored
>Exception except.AttributeError: "'None' object has no attribute
>'delete_DB'" in
> <method DBPtr.__del__ of DB instance at 5dfd00> ignored
>Exception except.AttributeError: "'None' object has no attribute 'time'" in
><met
>hod lwdb_base.__del__ of lwdb_base instance at 5de6d0> ignored
>
>
>Why is __del__ being called with a self = None ?
>
>
>
>Oversimplified Code Snippet:
>
>class foo(bar):
>    def close(self):
>        # pretend I do stuff here
>
>    def __del__(self):
>        #print "lwdbtable.__del__"
>        self.close()
>
>
>Warren
>
>



From djc at itga.com.au  Thu Apr 27 20:37:27 2000
From: djc at itga.com.au (Dave Cole)
Date: 28 Apr 2000 10:37:27 +1000
Subject: Best database to use with Python
References: <1255280638-66426487@hypernet.com> <slrn8ggos5.4un.tim.lavoie@goibniu.beyondtv.net>
Message-ID: <878zxzkqm0.fsf@heresy.itga.com.au>

>>>>> "Tim" == Tim Lavoie <tim.lavoie at mts.net> writes:

>> to lack of transaction support), but the most startlingly fast one
>> was bulk load, which does you no good in a CGI situation. Again,
>> from memory, MySQL was fast at very simple queries, but had a very
>> limited SQL dialect and fell apart with fancier queries.

Tim> I did a work project involving a search engine of sorts, with a
Tim> crawler creeping through the data at night, and a CGI front-end
Tim> to look for one or more words. The first iteration used Postgres,
Tim> but was slow.... The crawler took forever on inserts, and since
Tim> the data changed often, there were a lot of them. Queries for
Tim> single words were acceptable, but multi-word 'OR' searches were
Tim> painfully slow. In this case, MySQL fit the bill very well; data
Tim> was added very quickly, and queries were extremely fast. Also, we
Tim> didn't need transactions or anything of the sort, since at the
Tim> very worst we'd crawl the whole works again.

I have written a mostly DBI 2.0 compliant interface to Sybase.  The
freely available 11.0.3 version of Sybase from linux.sybase.com was
used for development.

One of the things that we needed here was fast inserts of large
amounts of data, so I extended the interface to support the bulkcopy
import interface.  This allows you to do things like:

    db = Sybase.connect(server, user, password, bulkcopy = 1)
    bcp = db.bulkcopy('cli_order')
    file = open(orders_file)
    while 1:
        line = string.rstrip(file.readline())
        if not line:
            break
        bcp.rowxfer(line)

The speed of inserts is blindingly fast.  I just wrote a little test
program to show the speed.

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
import time
import Sybase

start = time.time()
db = Sybase.connect('SYBASE', 'djc', 'password', bulkcopy = 1)
db.execute('create table #test (col1 varchar(80), col2 int, col3 char(20))')
bcp = db.bulkcopy('#test')
for i in xrange(100000):
    bcp.rowxfer(('a string', 20, 'another string'))
bcp.done()
end = time.time()
c = db.cursor()
c.execute('select count(*) from #test')
num, = c.fetchone()
print '%d rows in %.2f seconds, or %.0f rows / second' \
      % (num, end - start, num / (end - start))
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

The program output:

100000 rows in 8.70 seconds, or 11500 rows / second

I am running this on a PC with a P-III/450, 128M RAM, IDE disk and the
Linux 2.2 kernel.

The nice thing about the rowxfer method is that you can pass in a CSV
string which is internally split into column values.  You can also
pass a list or tuple of column values.

For the interested, the module can be found at:

        http://www.itga.com.au/~djc/sybase.html

- Dave


From phd at phd.russ.ru  Fri Apr 21 06:25:11 2000
From: phd at phd.russ.ru (Oleg Broytmann)
Date: Fri, 21 Apr 2000 10:25:11 +0000 (GMT)
Subject: Using python on the web
In-Reply-To: <sssnd8.b91.ln@sleipnir.webthing.com>
Message-ID: <Pine.LNX.4.21.0004211024430.2712-100000@fep132.fep.ru>

On Thu, 20 Apr 2000, Nick Kew wrote:
> I have to confess to some measure of guilt.  I started work on an 
> Apache module to implement non-forking CGI, eliminating the so-called
> CGI overhead.

   New FastCGI/PersistentCGI/SpeedyCGI?

Oleg.            (All opinions are mine and not of my employer)
---- 
    Oleg Broytmann      Foundation for Effective Policies      phd at phd.russ.ru
           Programmers don't die, they just GOSUB without RETURN.




From erikenglund at hotmail.com  Mon Apr 17 17:45:37 2000
From: erikenglund at hotmail.com (Erik Englund)
Date: Mon, 17 Apr 2000 23:45:37 +0200
Subject: Microsoft ActiveX Scripting
Message-ID: <8dg0rg$afj$1@cubacola.tninet.se>

"You create script components using any scripting language that supports the
Microsoft? ActiveX? Scripting interfaces. Script languages that support
these interfaces include JScript, Microsoft? Visual Basic? Scripting Edition
(VBScript), PERLScript, PScript, and Python."

If I want to use vbscript with the activeX control I just set the property
'Languange' to 'vbscript', and it works. Does anyone know what I need to do
wo use it with Python? I have downloaded and installed python for windows.

/ Erik Englund




From python at rose164.wuh.wustl.edu  Thu Apr  6 15:10:04 2000
From: python at rose164.wuh.wustl.edu (David Fisher)
Date: Thu, 6 Apr 2000 14:10:04 -0500
Subject: Newbie: urllib.urlretrieve()
References: <221ee89a.3bf4ca4b@usw-ex0102-016.remarq.com>
Message-ID: <000b01bf9ffb$b73e8a00$0201a8c0@spkydomain>

Check out the websucker module in the tools/webchecker directory of the
source.  It mirrors a remote url locally.  If it doesn't do what you want,
it should at least point you in the right direction.

Good luck,
David

----- Original Message -----
From: "ez4gl" <ez4glNOezSPAM at hotmail.com.invalid>
Newsgroups: comp.lang.python
To: <python-list at python.org>
Sent: Thursday, April 06, 2000 1:54 PM
Subject: Newbie: urllib.urlretrieve()


> I'm trying to clone some Intranet pages for a contract
> assignment and thought Python would be a good tool because of
> it's urllib module.  However, when trying to use
> urllib.urlretrieve(), I have found that if you specify the
> second argument, the file for writing to has to already exist.
> What I want to do instead is create the name of the file on the
> fly, based on the name of the URL I'm retrieving.  I guess what
> I need to do is to create an empty file ahead of time?!  But how
> would I go about doing this under Windows? (Unix I could handle)
>
> Thanks in advance
>
> George
>
>
>
> * Sent from RemarQ http://www.remarq.com The Internet's Discussion Network
*
> The fastest and easiest way to search and participate in Usenet - Free!
>
> --
> http://www.python.org/mailman/listinfo/python-list




From thomas at xs4all.net  Sat Apr 15 05:01:24 2000
From: thomas at xs4all.net (Thomas Wouters)
Date: Sat, 15 Apr 2000 11:01:24 +0200
Subject: The Opposite of list(a,b,c)?
In-Reply-To: <Pine.LNX.4.10.10004141656190.3969-100000@propylaea.anduin.com>; from dan@cgsoftware.com on Fri, Apr 14, 2000 at 05:04:47PM -0700
References: <38F7A07F.9B0F309C@bigfoot.com> <Pine.LNX.4.10.10004141656190.3969-100000@propylaea.anduin.com>
Message-ID: <20000415110123.C15664@xs4all.nl>

On Fri, Apr 14, 2000 at 05:04:47PM -0700, Daniel Berlin wrote:

> <snip>
> > def func(SearchTerm):
> >   <...>
> >     retval = []
> >     for keyval in (dtnA.keys()):
> >         mtch = re.search(patSrchTrm, keyval)
> >         if mtch:
> >             retval.append(UNLIST(dtnA[keyval]))

> Change this to:

> if mtch:
> 	for value in dtnA[keyval]:
> 		retval.append(value)

Dont forget 'extend':

>>> spam = ['spam', 'spam', 'spam', 'spam and eggs']
>>> toast = ['spam and toast', 'toast, eggs and spam', 'eggs, spam and bacon']

>>> spam.extend(toast)
>>> spam
['spam', 'spam', 'spam', 'spam and eggs', 'spam and toast', 'toast, eggs and
spam', 'eggs, spam and bacon']

'spam + toast' works the same, but creates a new list. extend() works as the
for loop above, appending each element of the 2nd list to the first,
in-place.

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From re.heath at sympatico.ca  Wed Apr 12 02:06:16 2000
From: re.heath at sympatico.ca (R E Heath)
Date: Wed, 12 Apr 2000 06:06:16 GMT
Subject: need help just begining
References: <11822ca0.dd34a400@usw-ex0109-068.remarq.com>
Message-ID: <38F41190.C0BB507B@sympatico.ca>

SheaMon wrote:
> 
> Hi everyone,
> My name is Shea and I am just starting out on programming.
> I know not one thing about it, and a person told me to try
> Python programming. But I have not idea what to do to start
> making a program.
> 
> I main question is this. If i got a copy of this tutor can
> I just leave the numbers and stuff like that and just edit
> the rest? I get so lost in the number part of the
> programming. If anyone here has some time can you please
> help me and email me at my address here.
> 
> thanks.
> Shea
> SheaCouture at hotmail.com
> 
> * Sent from AltaVista http://www.altavista.com Where you can also find related Web Pages, Images, Audios, Videos, News, and Shopping.  Smart is Beautiful

Python is an exceedingly good programming language but not really the
thing for a complete newcomer to start with. I'd suggest you start with
one of the older versions of BASIC such as MBASIC, BASICA, or QBASIC (or
others of similar type, but not VisualBasic) which run under DOS. When
you get the hang of that with a few non-trivial programs, then come back
and have a look at Python.

REH


From zawrotny at gecko.sb.fsu.edu  Tue Apr  4 15:32:57 2000
From: zawrotny at gecko.sb.fsu.edu (Michael Zawrotny)
Date: 4 Apr 2000 19:32:57 GMT
Subject: What is python's language level?
References: <38EA1B8F.A797A130@lmco.com> <38E9DDD3.3313FF6@bioreason.com>
Message-ID: <slrn8ekddq.gm3.zawrotny@gecko.sb.fsu.edu>

On Tue, 04 Apr 2000 13:19:31 +0100, Brian Kelley <kelley at bioreason.com> wrote:
> Chuck Meyers wrote:
> 
> > Does anyone know python's language level? I did not find it in the list
> > at:
> >
> > http://www.spr.com/library/0langtbl.htm
> >
> > I would guess that it would be somewhere between 15 and 30. This is an
> > important point, even if language levels are imprecise they can make it
> > easier to convince people to look at or try a language like python.
> >
> > -Thanx
> >
> 
> Okay, I'll bite.  This seems kind of arbitrary to me.  Let me put it this
> way, looking at the lists of "language levels", special purpose languages
> rank highly:  Mathcad, excel, lotus, etc.

Very arbitrary.  I liked the fact that in their table:

common lisp = 5.00
C++ = 6.00
nroff = 6.00
scheme = 6.00
HTML = 20.00 - 22.00

How many things are wrong with this ranking?  It's hard to count all of
the problems.  Lets start with the fact that nroff and html are markup
languages not programming languages.  Then move on to scheme higher than
common lisp.  Likewise C++.  I can't see anyone who has ever used the
three of them ranking them in that order.

Just my $0.02.  Everybody is entitled to their own, probably different,
opinion.


Mike

-- 
Michael Zawrotny
411 Molecular Biophysics Building
Florida State University		| email:  zawrotny at sb.fsu.edu
Tallahassee, FL 32306-4380		| phone:  (850) 644-0069


From abrahams at mediaone.net  Sun Apr 23 17:37:57 2000
From: abrahams at mediaone.net (Dave Abrahams)
Date: Sun, 23 Apr 2000 21:37:57 GMT
Subject: partial parsing?
References: <8dvilp$88s$1@slb6.atl.mindspring.net>
Message-ID: <B528E531.4EF9%abrahams@mediaone.net>

in article 8dvilp$88s$1 at slb6.atl.mindspring.net, Andrew Dalke at
dalke at acm.org wrote on 4/23/00 2:27 PM:

> Can anyone here provide pointers to relevant programs or papers?

You might think about looking at chart parsing technology, sometimes used in
natural language processing.

You can use a chart parser to extract all parseable sub-regions from a
sequence of tokens. Just turn off the grammar rules you're not interested in
and you should be all set (provided the grammar rules you care about don't
generate false positives due to the lack of contextual information that
would be neccessary for a complete parse). Depending on how self-similar
your grammar is, this technique may well end up doing more work than you'd
like. Whether it will work for you sort of depends on your goals and the
problem domain.

Just a caveat: in typical program parsers much more time is spent in lexical
analysis than in actual parsing. This tends to indicate that using YACC is
probably about as well as you'll do for speed and trying to optimize away
parsing time is misguided.

Good luck,
Dave



From effbot at telia.com  Sun Apr  2 13:16:49 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Sun, 02 Apr 2000 17:16:49 GMT
Subject: Python 1.6 alpha 1 released
References: <200003312130.QAA04361@eric.cnri.reston.va.us> <38E52716.A1C63B5D@roguewave.com> <rJ9F4.549$n68.187227648@newsb.telia.net> <38E52D76.1820FC32@roguewave.com> <8c7tfd$k80$1@nntp9.atl.mindspring.net>
Message-ID: <5gLF4.970$n68.190385152@newsb.telia.net>

Aahz Maruch wrote:
> I'm sympathetic.  We only upgraded our production system from 1.5.1 to
> 1.5.2 last week, because we'd made a number of 1.5.1-specific hacks and
> had to back them out correctly for 1.5.2.

we upgraded a customer site from 1.3 to 1.5.2 in november
(distributed, mission critical, many gigabytes per day, custom
interpreters, lots of C and C++ extensions, etc).

catching up with changes in Python and its standard libraries
was a very small part of the total effort.  getting it all to work
on a new version of the operating system was much harder...

</F>




From just at letterror.com  Thu Apr 20 02:16:45 2000
From: just at letterror.com (Just van Rossum)
Date: Thu, 20 Apr 2000 07:16:45 +0100
Subject: __del__ problem - would adopting Garbage Collection fix this?
In-Reply-To: <m31z41iyhg.fsf@atrus.jesus.cam.ac.uk>
References: <aw2L4.8907$HG1.245743@nnrp1.uunet.ca>
 <7ojL4.9424$HG1.254543@nnrp1.uunet.ca>
 <Z%oL4.9643$HG1.257533@nnrp1.uunet.ca>
Message-ID: <l03102802b5245023a247@[193.78.237.147]>

>"Warren Postma" <embed at geocities.com> writes:
>> My questions:
>>
>> 1. Shouldn't freeing modules in Py_Finalize be in the reverse order
>> of the way in which the modules were created?  Could a
>> module-destruction order be explicitly maintained by the system?
>
At 10:15 PM +0100 19-04-2000, Michael Hudson wrote:
>It's not clear that that would help; I'd bet I could come up with a
>counter example if I was feeling twisted enough.
>
>> 2. Is this convincing proof, or not, that Garbage Collection would
>> be a Good Thing.
>
>I don't think so.  You still have to destroy things; true gc might
>help I suppose, but I'm fairly sure it doesn't solve the problem
>completely.

I'm not sure if I understood it correctly, but I think the newly proposed
garbage collector does *not* call __del__ if the object in question is part
of a cycle. It seems indeed impossible to do so correctly. I think the site
explained quite well why this is, but I can't seem to repeat it off the top
of my head.

Just





From moshez at math.huji.ac.il  Wed Apr  5 01:55:24 2000
From: moshez at math.huji.ac.il (Moshe Zadka)
Date: Wed, 5 Apr 2000 07:55:24 +0200 (IST)
Subject: Why should I switch to Python? - Infinity of Primes
In-Reply-To: <7F51E902E09F3F60.8F612607812ED16B.82CB8B41435E14FD@lp.airnews.net>
Message-ID: <Pine.GSO.4.10.10004050749030.24181-100000@sundial>

On 4 Apr 2000, Cameron Laird wrote:

> To what extent are constructive proofs supplanting *... ad
> absurdum* as standards?

I don't know, but the given proof is easily transformed into a
constructive proof:

Here's a way to find one more prime:

Take any n-primes, take the product, add 1, via brute force find a prime
factor. A constructive way to find in a bounded time (though long!) an
infinite sequence of primes,

Note that a similar transformation shows that the diagonal argument is
usually just as constructive.

There are some reductio ad absurdum which cannot be trnsformed thus: this
is usually because of a use of the axiom of choice. So, in a very real
sense, this al boils down to: do you believe the axiom of choice (or some
weaker variant, like the ultra-filter axiom, which is almost always
enough, or the countable axiom of choice, which is enough in many cases)?

Thus, the question was taken out of mathematics and into philosophy, and
hence cannot be answered, so the only answer I can give is my own: I
refuse to use the axiom of choice, but I do use the ultra-filter axiom.
This relates in a deep way to how I view the future of mathematics.

was-that-offtopic-enough-ly y'rs, Z.

--
Moshe Zadka <mzadka at geocities.com>. 
http://www.oreilly.com/news/prescod_0300.html
http://www.linux.org.il -- we put the penguin in .com




From just at letterror.com  Thu Apr 20 03:35:13 2000
From: just at letterror.com (Just van Rossum)
Date: Thu, 20 Apr 2000 08:35:13 +0100
Subject: __del__ problem - would adopting Garbage Collection fix this?
In-Reply-To: <slrn8ft7qc.u9n.nascheme@cranky.arctrix.com>
References: <aw2L4.8907$HG1.245743@nnrp1.uunet.ca>
 <7ojL4.9424$HG1.254543@nnrp1.uunet.ca>
 <Z%oL4.9643$HG1.257533@nnrp1.uunet.ca>
 <m31z41iyhg.fsf@atrus.jesus.cam.ac.uk>
 <38FE7F1E.168E914A@cosc.canterbury.ac.nz><slrn8ft35a.uqs.nascheme@cranky.a
 rctrix.com> <B5240D22.4AFD%abrahams@mediaone.net>
Message-ID: <l03102805b52461abc14d@[193.78.237.146]>

>Dave Abrahams <abrahams at mediaone.net> wrote:
>>Hmm. If the objects are really garbage (i.e. unreferenceable by
>>the programmer), then there really is nothing she can do to deal
>>with it.
>
At 6:09 AM +0000 20-04-2000, Neil Schemenauer wrote:
>Uncollectable garbage is placed in a list reachable from Python
>(ie. gc.garbage).
>
>>course you can't call the __del__ method, but it seems you
>>should be able to reclaim the memory occupied by the objects
>>without losing anything.
>
>Maybe, but what are the __del__ methods for if they are not to
>free resources?  This situation really should not occur.

Sure, but nor should an exception ever occur in a __del__ method.

>With
>the default debugging options the collector prints a message to
>sys.stderr when it finds uncollectable objects.  Redesigning the
>class structure should fix the problem.

Maybe you're right, but to me it feels this is in the exact same category
as a __del__ exception, so ignoring & collecting seems more logical to me
as a default.

Just





From rjroy at takingcontrol.com  Tue Apr  4 16:02:02 2000
From: rjroy at takingcontrol.com (Robert Roy)
Date: Tue, 04 Apr 2000 20:02:02 GMT
Subject: Python 1.6 alpha 1 released
References: <Pine.GSO.4.10.10004040035420.21084-100000@sundial> <5E877FD7637AB46E.DABBED9CF3C1D64B.CDDDC05AE7374A78@lp.airnews.net> <5l7leern5v.fsf@eric.cnri.reston.va.us>
Message-ID: <38ea47ab.498518500@news1.on.sympatico.ca>

On 04 Apr 2000 09:49:48 -0400, Guido van Rossum
<guido at cnri.reston.va.us> wrote:

>Before more people mail this to me, let me apologize.  I didn't
>realize that there were so many examples of the 2-argument form.  I
>don't want to break that much code, so I'll revert the decision in the
>next alpha release.
>
>However, I'll do so for purely backwards compatible reasons, and I
>will continue to press for the tuple address as the correct form.
>This has nothing to do with academic purity.  It is because sockets
>are, by their very nature, address family neutral, and not all address
>families use a host and a port to identify a socket.  For example,
>Unix sockets use just a filename.
>
>The argument that gopherlib c.s. define a 2-arg connect() doesn't mean
>that socket should, to: those are all Internet specific protocols,
>where host+port is the norm; the lower-level socket API must support
>other address families.
>
>The argument that the address object is spurious doesn't hold water.
>There are many places where it is convenient to hold an address in a
>single variable, and a tuple is the obvious way to represent this.
>(The alternative, a "host:port" string, seems too Perlish to me.)
>Note also that recvfrom() returns the address as a tuple (if it is an
>IP address) and sendto() requires the address to be passed as a tuple
>-- the canonical form of an address is clearly a tuple.
>
>Repeat after me: SOCKETS ARE NOT JUST FOR IP CONNECTIONS.
>
>But I will fix the code -- it's the only practical thing to do.
>
>--Guido van Rossum (home page: http://www.python.org/~guido/)

Does this mean that the two argument connect will be officially
deprecated in 1.6,  (and be documented as such) and removed in a
future release say 1.7?

This would serve as fair warning to and give people a release cycle to
clean up their code.

Bob




From mhammond at skippinet.com.au  Tue Apr  4 19:59:50 2000
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Tue, 04 Apr 2000 23:59:50 GMT
Subject: builtding library .pyd file format
References: <000b01bf9e29$4490a0c0$3acbd9c2@optichrome.com>
Message-ID: <WlvG4.5059$1b4.24311@news-server.bigpond.net.au>

"Adrian Eyre" <a.eyre at optichrome.com> wrote in message
news:000b01bf9e29$4490a0c0$3acbd9c2 at optichrome.com...
> > Hello evry body,
>
> Can you not post HTML in future, please?
>
> > I want to know how to do to built a .pyd library file format
form
> > the windows .dll file format
>
> Just rename the .dll to .pyd

Or dont :-)  Python should find it named .dll just fine...

Mark.




From gmcm at hypernet.com  Thu Apr 20 08:05:39 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Thu, 20 Apr 2000 08:05:39 -0400
Subject: Using popen in windows
In-Reply-To: <newscache$vzabtf$ia6$1@jenpc07.jennic.co.uk>
Message-ID: <1255890154-29764960@hypernet.com>

Robert Cragie wrote:

> I tried the following in Windows 98 as an experiment:
> 
> from win32pipe import popen
> p = popen('dir','r')
> str = p.read()
> print str
> 
> It just hangs when trying to do the read, and I have to kill the shell off.
> I have tried IDLE 0.5 and PythonWin and the 'DOS' command line, both do the
> same. I am using Python 1.5.2, and the win32pipe.pyd DLL is dated 3rd May
> 1999.
> 
> I have done the equivalent on Linux and it works fine:
> 
> from os import popen
> p = popen('ls','r')
> str = p.read()
> print str
> 
> Anyone got any ideas?

Ah, but it's not equivalent :-). 'dir' is a cmd.exe builtin, not an 
"external program". You can't (directly) popen any of the 
cmd.exe builtins (and the list changes with each OS release).

on-Windows-you-gotta-think-diphpherent-ly y'rs

- Gordon



From jhauser at ifm.uni-kiel.de  Wed Apr 12 05:55:32 2000
From: jhauser at ifm.uni-kiel.de (Janko Hauser)
Date: 12 Apr 2000 11:55:32 +0200
Subject: Classes, Inheritance - Stupid lazy question
References: <8d1cfh$hme7@imsp212.netvigator.com> <874s9766r6.fsf@ifm.uni-kiel.de> <m3itxnis32.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <87zoqz4p9n.fsf@ifm.uni-kiel.de>

Michael Hudson <mwh21 at cam.ac.uk> writes:

> Nononono; think about what happens if you do this:
> 
> class bob(paul):
>    pass
> 
> bob("adam").fired()
> 
> What *I* think you want is:
> 
> class paul(lazy):
>     def __init__(self, name):
>         lazy.__init__(self,name)
>         self.april_pay = 0
> 
>     def fired(self):
>         print "Here's a box, collect the things from your desk"
>         lazy.fired(self)
> 
> After all, if you're going to inherit from a class, you must have
> access to it, right?
> 

I see, this is clearly better. But I do not understand, where the
class/instance?? of lazy.__init__() is stored. 

Some more slaps on the head, please :-)

__Janko

-- 
  Institut fuer Meereskunde             phone: 49-431-597 3989
  Dept. Theoretical Oceanography        fax  : 49-431-565876
  Duesternbrooker Weg 20                email: jhauser at ifm.uni-kiel.de
  24105 Kiel, Germany


From alwagner at tcac.net  Sun Apr 16 00:52:17 2000
From: alwagner at tcac.net (Albert Wagner)
Date: Sat, 15 Apr 2000 23:52:17 -0500
Subject: Doesn't anybody write to paper anymore?
References: <38F5FDF3.EE2FBA01@tcac.net>
Message-ID: <38F94701.CA52FDBB@tcac.net>

Thanks to everyone who replied.  With your help, I got it working. 
Even, the facetious replies were appreciated. Just one follow-up:  The
docs say that os.popen() is "unreliable on Windows."  Is this still true
in 1.5?  And if so, is there an alternate implementation for windows?  I
run on Redhat6.1 Linux at home but need to port to windows98 for work.

-- 
Small is Beautiful


From tismer at tismer.com  Thu Apr 27 12:18:30 2000
From: tismer at tismer.com (Christian Tismer)
Date: Thu, 27 Apr 2000 18:18:30 +0200
Subject: Schwartzian Transform in Python?
References: <lr3do7tuin.fsf@yoda.wg.waii.com>
Message-ID: <39086856.525D1999@tismer.com>


Art Haas wrote:
...
> The python documentation doesn't make
> clear to me how to do things like "sort a list of lists by the second
> field in each sublist", which is what the Schwartzian transform is
> based on.

I think to remember an FAQ entry.
Here an (older) version of mine.
Not that today I would code that without
"transpose"
(left as an exercise to the reader)

You would use the below by supplying a metric like

sort(list, lambda x:x[1])

--------------------------

# this is taken from the FAQ and modified slightly:
"""
4.51. I want to do a complicated sort: can you do a Schwartzian
     Transform in Python?

     Yes, and in Python you only have to write it once: 

      def st(List, Metric):
          def pairing(element, M = Metric):
                return (M(element), element)
          paired = map(pairing, List)
          paired.sort()
          return map(stripit, paired)

      def stripit(pair):
          return pair[1]
"""

# now, this is my version:

def sort(List, Metric=None, stable=1) :
	"""sort a list by a given metric function. The sort defaults
	to be stable, keeping the order of elements which cannot be
	distinguished by the metric"""

	# transpose has problems with small lists, but in that
	# case we don't need it:
	if len(List) < 2: return List[:]
	
	shaker = [List]
	where = 0
	if Metric :
		# sorter is a list of function results
		shaker.insert(0, map(Metric, List))
		where = -1 # we are at the end
	if stable :
		# always stabilize behind the key
		shaker.insert(1, range(len(List)))
		# and we are still either at front or end
	shaker = transpose(shaker)
	shaker.sort()
	return list(transpose(shaker)[where])

def transpose(x) :
	"""transpose a sequence of lists into a list of tuples. This
	is very fast, but doesn't work for len(x)==1"""
	return apply(map, (None,)+tuple(x))

# Special problem with fast transpose:
# Due to map, transpose of a list with a single element
# returns a list, shape is lost.

# Solution:

def transpose(x) :
	if len(x) == 1 : 
		return map(lambda y : (y,), x[0])
	return apply(map, (None,)+tuple(x))

# General problems with transpose (not solvable here):
# A list of empty tuples returns an empty list (shape lost)

--------------------------

ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaunstr. 26                  :    *Starship* http://starship.python.net
14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     where do you want to jump today?   http://www.stackless.com



From vetler at ifi.uio.no  Sun Apr  2 05:23:58 2000
From: vetler at ifi.uio.no (Vetle Roeim)
Date: 02 Apr 2000 11:23:58 +0200
Subject: Emacs Python mode
References: <yayk8iii0ex.fsf@ganglot.ifi.uio.no> <m3zorendyk.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <yayzorcj1ox.fsf@ganglot.ifi.uio.no>

* Michael Hudson
> Vetle Roeim <vetler at ifi.uio.no> writes:
> 
> > I just had an idea. Using the Emacs python-mode, would it be possible
> > to somehow hide the function-bodies?
> 
> Try "M-8 C-x $", which isn't the easiest thing to type, but it works
> reasonably well.

hey, I didn't know that was possible. thanks! :)

but it looks like it's only possible hide all functions at once, so I
think I'll test the latest emacs-mode as Randal Hopper suggested.

..once python.org gets back up, of course.
 
> It's worth readong through the extended help on python-mode (type "C-c
> C-h m" or "C-c ?"), which I did yesterday and realised that there were
> lots of shortcuts for handy things I never knew existed.  Of course I
> now can't remember what the key combinations for any of these cool
> things were.

all in all, *emacs* has lots of shortcuts for handy things. I wish I
knew more about it.. :)


vr


From sabren at manifestation.com  Sun Apr 23 00:14:43 2000
From: sabren at manifestation.com (Michal Wallace (sabren))
Date: Sun, 23 Apr 2000 00:14:43 -0400 (EDT)
Subject: code not true?
In-Reply-To: <slrn8g35ot.1pb.scarblac-spamtrap@flits104-37.flits.rug.nl>
Message-ID: <Pine.BSF.4.10.10004230013310.29461-100000@chi.pair.com>

On 22 Apr 2000, Remco Gerlich wrote:

> jeff_islay at my-deja.com wrote in comp.lang.python:
> > Why does the second example return false?
> > 
> > >>> print 7.0 == 7
> > 1
> > >>> print (.07 * 100) == 7
> > 0
> 
> Floating point number on a binary computer are strange. They're not that
> accurate. 0.07 * 100 != 7.0.

Why is this the case? I mean.. I understand why decimals are hard to
show in binary... But why can't we use variably long strings, or
something? Is it just a speed issue?

Cheers,

- Michal
-------------------------------------------------------------------------
http://www.manifestation.com/         http://www.linkwatcher.com/metalog/
-------------------------------------------------------------------------




From jmg at ecs.soton.ac.uk  Wed Apr  5 12:03:08 2000
From: jmg at ecs.soton.ac.uk (Jacek Generowicz)
Date: Wed, 05 Apr 2000 17:03:08 +0100
Subject: Pass by reference ?
References: <38E864C9.E1EE94F7@ecs.soton.ac.uk> <38EAF8D8.4B7C43B4@ecs.soton.ac.uk> <38EB56AE.F938CE06@acm.org> <m33dp0muk9.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <38EB63BC.FE03BA31@ecs.soton.ac.uk>

Michael Hudson wrote:

> Python never ever copies anything when you pass it as an argument to a
> function - it's just that sometimes the things you pass are immutable
> - which is when it looks like call by value - and sometimes they're
> mutable - when it looks a bit like call by reference.

Not necessarily; consider the following (and please excuse my Python-naive
coding style):

def looks_like_by_value( a ):
    a = a + 1
    print a

def looks_like_by_reference( a ):
    a.append('x')
    print a

def looks_like_by_value_again( a ):
    a = a + ['x']
    print a

a = 3
print a
looks_like_by_value( a )
print a; print

a = [ 1,2,3 ]
print a
looks_like_by_reference( a )
print a; print

a = [ 1,2,3 ]
print a
looks_like_by_value_again( a )
print a; print
============
output:

3
4
3

[1, 2, 3]
[1, 2, 3, 'x']
[1, 2, 3, 'x']

[1, 2, 3]
[1, 2, 3, 'x']
[1, 2, 3]


So whether the call resembles by-reference or by-value depends on what you do
with the mutable object.

Jacek




From ns56645 at swt.edu  Sun Apr 23 00:56:47 2000
From: ns56645 at swt.edu (ns56645)
Date: Sat, 22 Apr 2000 23:56:47 -0500
Subject: Questions regarding my project
Message-ID: <3902828F.9308BD91@swt.edu>

I am currently doing a project in Tkinter and Python.

The prototype of the project is placed in HTML format at
http://www.swt.edu/~ns56645/gui.

I have almost finished with the bottom part but the top frame seems to
be difficult to me. I am running short of ideas of how to implement the
top frame. I am thinking of using different canvas for each day(i.e.
Mon, Tue etc.)  to show classes held during days at a particular time.
Ofcourse the data willbe taken from the database and will be dynamic. So
I will also need the scrollbar for each day if the classes held at
monday at 8.00 is > than 1.

Some suggestions for implememting the upper frame are welcome.
Also, let me know if it is possible with the canvas and buttons and
writing text on the buttons to represent the classes at  a particular
time.

Shah

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20000422/5693ac5c/attachment.html>

From aek at mail1.stofanet.dk  Tue Apr  4 18:36:33 2000
From: aek at mail1.stofanet.dk (Anders Eggers-Krag)
Date: Wed, 5 Apr 2000 00:36:33 +0200
Subject: Tkinter, <select> style menu
Message-ID: <8cdqom$aqp$1@news101.telia.com>

Is there really no reasoanble way to make a box similar to that of the HTML
<SELECT>
in tkinter?

the nearest I have been able to find is from Pmw, and it doesn't really look
like windows...


Anders Eggers - Krag




From aa8vb at yahoo.com  Wed Apr  5 07:35:26 2000
From: aa8vb at yahoo.com (Randall Hopper)
Date: Wed, 05 Apr 2000 07:35:26 -0400
Subject: Tkinter:  "Get root for a widget" is private?
In-Reply-To: <2Q_F4.1092$n68.187097600@newsb.telia.net>
References: <20000331112048.A1257432@vislab.epa.gov>
 <QdEF4.892$n68.190327808@newsb.telia.net> <m3snx3n1x0.fsf@devel.no.rwd.dinet>
 <20000331112048.A1257432@vislab.epa.gov>
 <QdEF4.892$n68.190327808@newsb.telia.net>
 <20000403064335.A1560771@vislab.epa.gov>
 <2Q_F4.1092$n68.187097600@newsb.telia.net>
Message-ID: <20000405073526.A1760238@vislab.epa.gov>

Fredrik Lundh:
 |Randall Hopper <aa8vb at yahoo.com> wrote:
 |> Given this it seems we should have a public root() method.  Could we
 |> add a:
 |>
 |>     root = _root
 |>
 |> definition to Tkinter's Misc class, to make this a public feature?
 |
 |the question is why?
 |
 |if you're only out to get an interpreter handle (e.g. to
 |instantiate a variable or image object), any widget will
 |do.

I have a timer class built around Tk timers.  I realize I could just pass
any_old_widget into this routine, and it would store it off into state data
and use it periodically to access the 'after' and 'after_cancel' methods.

However, this timer has nothing to do with this widget in particular, so
why should it have a reference to it (versus the button next to it, or the
list widget below it)?  It makes more sense for the timer class to accept
and bookmark the root object.  This root is associated with the global Tk
state, and it makes sense for a timer to bookmark this to use global Tk
state commands.

-- 
Randall Hopper
aa8vb at yahoo.com



From cjc26 at nospam.cornell.edu  Mon Apr  3 20:14:19 2000
From: cjc26 at nospam.cornell.edu (Cliff Crawford)
Date: Tue, 04 Apr 2000 00:14:19 GMT
Subject: Creating variables on the fly...
References: <857F15D7E3D8D3118D290008C7CF058627070F@mail-naeast1.brooks.com>
Message-ID: <slrn8eid4n.i8e.cjc26@ith1-491.twcny.rr.com>

This is the simplest solution, but also the most adaptable.  Because
someday, you're going to want to get the list of numbers from a file, or
from a web page, or some other source external to your program; and if
you use *args, then you're going to have to do apply(add, numlist)
instead of just add(numlist).

hi-Matt-I-go-to-Cornell-too!-<g>-ly yrs, Cliff

* Lee, Jaeho <Jhlee at Brooks.com> menulis:
| I think that this is the place you should use list.
| 
| What about this.
| 
| def add(listVar)
|    r = 0
|    for something in listVar
|       r = r + something
|    return r
| 
| I'm still learning python. So please forgive any syntax or naming convetion
| problem.
| 
| Jaeho


-- 
cliff crawford    -><-    http://www.people.cornell.edu/pages/cjc26/
                                                       icq 68165166
"IS I yes wardrobe yield [the] evaluation."     member of A.H.M.A.D.


From jflores at codeit.com  Fri Apr 14 12:46:36 2000
From: jflores at codeit.com (Grakka)
Date: Fri, 14 Apr 2000 09:46:36 -0700
Subject: Crosswords for Linux Backtracking recursive crossword generator
References: <8d59hk$6to$1@nnrp1.deja.com> <m3g0sphfay.fsf@atrus.jesus.cam.ac.uk> <8d7gnd$l4a$1@nnrp1.deja.com>
Message-ID: <38F74B6C.966EB04F@codeit.com>

What does "hlio" means in spanish also?

Manuel Gutierrez Algaba wrote:

> In article <m3g0sphfay.fsf at atrus.jesus.cam.ac.uk>,
>   Michael Hudson <mwh21 at cam.ac.uk> wrote:
> > Manuel Gutierrez Algaba <algaba at my-deja.com> writes:
> >
> > > Comments/improvements are welcome!
> >
> > I'm just looking at it, but in playing around I got this output:
> >
> > - - - - - - - - - - - - - - -
> > - - - - - - - - - - c - - c -
> > - - - - - - - - - s o l f a -
> > - - - - - - - - - - n - - r -
> > - - - - - - - - - - f - - a -
> > - - - - - - - - - - u - - n -
> > - - - - - - - - - - s - - t -
> > - - - z a l a m e r i a - o -
> > - - - - - - - - - - o - m ? -
> > - - - - c o h e r e n c i a -
> > - - - - a - l - - - - - m - -
> > - - - - p - i n s u l t o - -
> > - - - m a n o - - - - - - - -
> > - - - - - - - - - - - - - - -
> > - - - - - - - - - - - - - - -
> >
> > That's not right is it?  The `m' and `?' being next to each other.
>
> It's not a "bug". It's that way, of course, the algorithm may be
> modified for avoiding such things, but then it's much slower.
>
> --
> -------------
> Manolo
>
> Sent via Deja.com http://www.deja.com/
> Before you buy.
> --
> http://www.python.org/mailman/listinfo/python-list




From echuck at mindspring.com  Sat Apr 15 19:28:23 2000
From: echuck at mindspring.com (Chuck Esterbrook)
Date: Sat, 15 Apr 2000 19:28:23 -0400
Subject: Web hosting python CGI and or Zope
References: <8c8hie$5lo$1@nnrp1.deja.com>
Message-ID: <38F8FB17.7FE5B8D4@mindspring.com>

zsoltman at my-deja.com wrote:

> Question:  Does anyone know of Web Hosting companies that support
> python or Zope?

Check out "Python-friendly ISPs":

http://www.corrt.com/info/pyisp-list.html


-Chuck

From samschul at pacbell.net  Fri Apr 21 21:32:11 2000
From: samschul at pacbell.net (Sam Schulenburg)
Date: Sat, 22 Apr 2000 01:32:11 GMT
Subject: Pythonwin com question
Message-ID: <8dqver$l4v$1@nnrp1.deja.com>

I have a com object that I can excercise with Visual Basic as such:

Dim WApp as New Winx32Lib.Winx32App    ' Get an instance of Winx32App
WApp.ShowDemoBox("This is a demo box") ' Call the client ShowDemoBox()

>From Python I try the following which starts the client application and
then timesout with the following error msg:

>>> import win32com.client
>>> MyCLSID = '{1A762221-D8BA-11CF-AFC2-508201C10000}'
>>> WApp = win32com.client.Dispatch(MyCLSID)
Traceback (innermost last):
  File "<interactive input>", line 1, in ?
  File "D:\Python\win32com\client\__init__.py", line 94, in Dispatch
    dispatch, userName = dynamic._GetGoodDispatchAndUserName
(dispatch,userName,clsctx)
  File "D:\Python\win32com\client\dynamic.py", line 66, in
_GetGoodDispatchAndUserName
    return (_GetGoodDispatch(IDispatch, clsctx), userName)
  File "D:\Python\win32com\client\dynamic.py", line 57, in
_GetGoodDispatch
    IDispatch = pythoncom.CoCreateInstance(IDispatch, None, clsctx,
pythoncom.IID_IDispatch)
com_error: (-2146959355, 'Server execution failed', None, None)
>>>

Any help would be appreciated
Samuel W. Schulenburg


Sent via Deja.com http://www.deja.com/
Before you buy.


From pony at nmt.edu  Mon Apr  3 13:56:32 2000
From: pony at nmt.edu (Enrique Herrera)
Date: Mon, 03 Apr 2000 11:56:32 -0600
Subject: Help on Big Python Program
Message-ID: <38E8DB50.A79CB88@nmt.edu>

	I have been working on a python program that is getting big. When its
all finished it will have several modules that have been written for it. 

	I was wandering if there is a way to "compile" the python source into
one executable program?

THanks.

Enrique


From claird at starbase.neosoft.com  Sun Apr  2 18:56:44 2000
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 2 Apr 2000 17:56:44 -0500
Subject: Smalltalk on the small end (was: Advice requested: GUI project beginning)
References: <m12b4OW-000wcEC@swing.co.at>
Message-ID: <8D79BC9E4960ED70.7B6609342D6EB723.7625773D14AD9140@lp.airnews.net>

In article <m12b4OW-000wcEC at swing.co.at>,
Christian Tanzer <tanzer at swing.co.at> wrote:
			.
			.
			.
>Recently, I find myself writing Python scripts of 5 to 20 lines for
>which I normally would have used the unix shell. I wouldn't use
>Smalltalk for this purpose.
But others do (particularly with GNU Smalltalk),
so this isn't conclusive.

Actually, there's another thread here asking to
be launched.  Python (and Tcl, too, in a somewhat
different fashion is (are) really quite extraor-
dinary for shell-level scripting and operation.
I find myself wondering more and more how people
get along without such (*particularly* on the
conventional desktops--Win* and MacOS).  I do
think Python is clearly superior to Smalltalk
for those one-page scripts that most people do
with /bin/sh; still, I'm willing to consider that
Smalltalk is more capable in this regard than I
yet realize, and, in any case, I can well imagine
Smalltalk proponents favoring it over /bin/sh.
			.
			.
			.
>What is Squeak?
"Squeak is an open, highly-portable Smalltalk-80
implementation whose virtual machine is written
entirely in Smalltalk, making it easy to debug,
analyze, and change."  <URL:http://www.squeak.org/>
It's lightweight in licensing and execution; you
can get Squeaking without a lot of the setup costs
traditionally associated with Smalltalk.
			.
			.
			.
-- 

Cameron Laird <claird at NeoSoft.com>
Business:  http://www.Phaseit.net
Personal:  http://starbase.neosoft.com/~claird/home.html


From effbot at telia.com  Tue Apr 25 17:51:53 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Tue, 25 Apr 2000 21:51:53 GMT
Subject: Strange translation in GET
References: <XFMail.000425112129.mikael@isy.liu.se>
Message-ID: <ZroN4.4172$Za1.60449@newsc.telia.net>

Mikael Olofsson <mikael at isy.liu.se> wrote:
> According to my logs, this link was recently clicked by someone in Japan,
> and my script failed because the argument had somehow been translated to
>
>  a+AD0-A+ACY-b+AD0-B
>
> That is, each '=' was translated to '+AD0-', and each '&' was translated
> to '+ACY-'. This probably has to do with the character set used by the
> caller.

it looks like they're using UTF-7 encoding.  ignore them.

</F>




From froydnj at rose-hulman.edu  Wed Apr 26 23:02:15 2000
From: froydnj at rose-hulman.edu (Nathan Froyd)
Date: 27 Apr 2000 03:02:15 GMT
Subject: defining __getitem__
References: <WiMN4.1$7C.101@news.hananet.net>
Message-ID: <slrn8gfb1c.a3q.froydnj@froydnj.laptop.rose-hulman.edu>

In article <WiMN4.1$7C.101 at news.hananet.net>, Gang Seong Lee wrote:
>I'd like to have a method __getitem__ in a class which acts like other
>sequence type.
>For example,
>------------------------
>class data:
>    def __init__(self, n):
>        self.maxsize = n
>
>    def __getitem__(self, k):
>        if k >= self.maxsize:
>              # I want to stop here
>              # ????
               raise IndexError

I believe...
-- 
</nathan>  froydnj at rose-hulman.edu  |  http://www.rose-hulman.edu/~froydnj/


From nmm1 at cus.cam.ac.uk  Tue Apr  4 08:56:47 2000
From: nmm1 at cus.cam.ac.uk (Nick Maclaren)
Date: 4 Apr 2000 12:56:47 GMT
Subject: Bugs in Python 1.6 alpha 1
References: <Pine.LNX.4.10.10003311651590.22919-100000@starship.python.net> <8c8vfj$5n6$1@news.udel.edu> <8ccmah$chj$1@pegasus.csx.cam.ac.uk> <m33dp211m1.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <8ccoqf$esj$1@pegasus.csx.cam.ac.uk>

In article <m33dp211m1.fsf at atrus.jesus.cam.ac.uk>, Michael Hudson <mwh21 at cam.ac.uk> writes:
|> nmm1 at cus.cam.ac.uk (Nick Maclaren) writes:
|> 
|> >     6) test_unicode fails on Hitachi systems -- Writing: '*', expected:
|> > 'T'.  But only if -v is NOT specified!  Again, I failed to tie it down,
|> > and it looks like a test harness problem.
|> 
|> That's the way regrtest works; if you specify `-v' all output goes to
|> stdout, if you specify `-g' it goes into the output directory and if
|> you specify neither it goes into some internal buffer to be compared
|> with what's in the output directory.
|> 
|> A bit odd, yes.  I don't know if it's a deliberate behaviour (doubt
|> it).

Oh, I see.  I misread the comments at the beginning.  It doesn't
exactly make it easy to tie down problems in the test harness ....
Never mind - it isn't an important problem.


Regards,
Nick Maclaren,
University of Cambridge Computing Service,
New Museums Site, Pembroke Street, Cambridge CB2 3QG, England.
Email:  nmm1 at cam.ac.uk
Tel.:  +44 1223 334761    Fax:  +44 1223 334679


From effbot at telia.com  Tue Apr 25 12:31:17 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Tue, 25 Apr 2000 16:31:17 GMT
Subject: Newbie Needs Help -- Name and Key errors
References: <bingdo19-4B8F57.12162225042000@news-server.twcny.rr.com> <SIjN4.1570$wYl.193603072@newsb.telia.net>
Message-ID: <pLjN4.1571$wYl.193775616@newsb.telia.net>

I wrote:
> 1) you need to the DOCUMENT_ROOT environment variable
> before running the script

guess I gotta fix the "set" key on my keyboard.

</F>




From mdkersey at hal-pc.org  Mon Apr  3 14:57:30 2000
From: mdkersey at hal-pc.org (Michael D. Kersey)
Date: Mon, 03 Apr 2000 13:57:30 -0500
Subject: CGI - Hidden Fields (NS4.7/IE5)
References: <38D625B2.5E0E88FD@student.tuwien.ac.at> <38E8AB88.5E5E60E5@0x01.com>
Message-ID: <38E8E99A.4F457351@hal-pc.org>

thanos wrote:
> Make shure you are using POST and not GET. EI handle form method GET
> badly!!!!
Perhaps you could clarify two questions:
a) what is "EI"?
b) what *exactly*, do you mean by "EI handle form method GET badly!!!!"?
Please provide references to authoritative sources.


From moshez at math.huji.ac.il  Sat Apr 29 07:32:03 2000
From: moshez at math.huji.ac.il (Moshe Zadka)
Date: Sat, 29 Apr 2000 14:32:03 +0300 (IDT)
Subject: Checking Comments
Message-ID: <Pine.GSO.4.10.10004291429230.25462-100000@sundial>

Here's a small script I use to check the comments in a file. It is a nice
tools for Py-teachers, to check students comment their code enough.

Here it is, in all it's glory:

================ cut here =========================
# comments.py
# Released to the Public Domain by Moshe Zadka 
# on 29.4.2000
import sys, tokenize

def print_if_comment(type, token, *args):
        if type==tokenize.COMMENT:
                sys.stdout.write(token)

tokenize.tokenize(sys.stdin.readline, print_if_comment)
====================== cut here ==================

use-and-enjoy-ly y'rs, Z.
--
Moshe Zadka <mzadka at geocities.com>. 
http://www.oreilly.com/news/prescod_0300.html
http://www.linux.org.il -- we put the penguin in .com




From pf at artcom-gmbh.de  Wed Apr 12 09:39:05 2000
From: pf at artcom-gmbh.de (Peter Funk)
Date: Wed, 12 Apr 2000 15:39:05 +0200 (MEST)
Subject: XML-RPC and None objects
Message-ID: <m12fNMH-000CnCC@artcom0.artcom-gmbh.de>

Hi!

XML-RPC <http://www.xmlrpc.com> is a very nice vehicle to build
distributed server/client systems.  Fredrik Lundhs xmlrpc package
available at <ttp://www.pythonware.com/products/xmlrpc/> provides
everything you need to build clients and servers in Python.

But the current XML-RPC specification has a serious problem, which
was pointed out over a year ago by Skip Montanaro in the XML-RPC
discussion forum:

> Some time ago the subject of supporting discriminated null data in
> XML-RPC was brought up and apparently dropped. I just checked the
> spec and didn't see anything new. I scanned back through a couple
> pages of topics but didn't see anything there either.

> I'm busy using XML-RPC to communicate between front-end Python
> and Perl stuff (CGI scripts, modules published through
> ZopeHTTPServer or Mason components) and back-end Python stuff
> (databases). The biggest problem I've encountered when converting
> existing code to use XML-RPC is identifying and converting cases
> where I use Python's None object. I have to figure out how it's used in
> the caller and callee, then figure out a reasonable replacement. The end
> result is usually that my code isn't as clear as it used to be and that I
> wind up adding the occasional extra bug to my code (as if I don't have
> enough bugs to squash already).

> <whine> I would *really* like to see a <nil></nil> or <nil/> data type
> added to XML-RPC. </whine>

I stumbled into the same problem and since I currently use XML-RPC only 
between my own server and clients I decided to ignore the official 
specification and added a 'None'-handler to my version of Fredrik Lundhs 
library.  The needed patch is indeed very very simple:  Add the following 
lines to the class 'Marshaller' in 'xmlrpclib.py' you are done:

    def dump_None(self, none):
        self.write("<value><null></null></value>\n")
    dispatch[NoneType] = dump_None

Since the class 'Unmarshaller' already returns 'None' for all unknown values
a modification there is not required.  This works fine for me.  YMMV.

Regards, Peter
-- 
Peter Funk, Oldenburger Str.86, D-27777 Ganderkesee, Germany, Fax:+49 4222950260
office: +49 421 20419-0 (ArtCom GmbH, Grazer Str.8, D-28359 Bremen)



From bjorn at roguewave.com  Tue Apr  4 19:46:41 2000
From: bjorn at roguewave.com (Bjorn Pettersen)
Date: Tue, 04 Apr 2000 17:46:41 -0600
Subject: Starving for an Advanced Python Book
References: <20000404231358.17447.qmail@web2105.mail.yahoo.com>
Message-ID: <38EA7EE1.457A5533@roguewave.com>

lewst wrote:
> 
> "M.-A. Lemburg" <mal at lemburg.com> wrote:
> 
> > Fredrik Lundh has compiled such a cookbook. I don't have the URL
> > handy, but I'm sure you'll find it on their www.pythonware.com
> > site.
> 
> Yes, this is Fredrik Lundh's eff-bot guide to The Standard Python
> Library which is an on-line book in e-Matter format.  This looks like
> it would be a great book for what I need, but there are a few
> problems.  First, eMatter is not currently available for Unix/Linux.
> While I do have access to Microsoft Windows, I don't own such a
> machine.  Because the eMatter license is tied to the particular
> computer you install the book on, this pretty much rules it out for me
> as a possibility which is unfortunate.  Second, at $12.00, the book
> seems a bit expensive for what it is.  I can understand paying that
> much for an eMatter version of a published book which has gone through
> a rigorous and professional editing process, but not otherwise.
> Nonetheless, I'd still buy it if it supported Linux.

At $12.00 I would say the book is a steal and contains a wealth of
information no serious Python programmer should be without :-)  And if
you're like me, you'll print it out, bind it at Kinko's, and place it
half way between your Linux and win2k boxes <wink>.

-- bjorn



From m.faassen at vet.uu.nl  Fri Apr 28 14:36:04 2000
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 28 Apr 2000 18:36:04 GMT
Subject: Do I always have to write "self." ?
References: <270420001706386648%pecora@anvil.nrl.navy.mil> <8eacdf$nnm$1@newshost.accu.uu.nl> <280420000751353947%pecora@anvil.nrl.navy.mil> <3909817b.2156432776@news.u-psud.fr>
Message-ID: <8eclmk$euk$2@newshost.accu.uu.nl>

Laurent POINTAL <pointal at lure.u-psud.fr> wrote:

> *** Note: taking references copies of the self members is the same
> problem as importing * from a module: its only reference copying... 

It's not a problem, Python *always* copies references unless you explicitly
tell it to make a real value copy. I've found it actually quite rare that I
needed a copy by value in my code. This is also why C and C++ code uses
pointers all over the place. C was designed originally with the
idea that copy by value is the most important thing. It's not (though more 
important in C than in Python, as integers and such aren't objects). Now
they're stuck with a zillion pointers to structures and objects.

[snip hack]
> But this is still only a reference copying... if you wants to change
> of referenced object in a member, you must use self.xxxxxx. 

If the referenced object is mutable, you only need self.xxxx once, 
in the beginning of the method, as I demonstrated. You can continue on
with the local variable from then on. Except when you're programming an
object that needs to be persistent in the Zope Object Database; you have
to treat all objects as immutable to make the persistence machinery pick
up your changes and store them. That is, instead of:

# Python
def mymethod(self, foo):
    self.mylist.append(foo)

or

def mymethod(self, foo):
    mylist = self.mylist
    mylist.append(foo)

You have to do something like:

# Python code for ZODB
def mymethod(self, foo):
    mylist = self.mylist
    mylist.append(foo)
    self.mylist = mylist

or
  
def mymethod(self, foo):
    self.mylist.append(foo)
    self.mylist = self.mylist

Ugly, but it works.

Regards,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From neilh at hare.net.au  Fri Apr  7 23:02:12 2000
From: neilh at hare.net.au (Neil Hodgson)
Date: Sat, 8 Apr 2000 13:02:12 +1000
Subject: Actually Running a function from a DLL
References: <38EE87A3.4080802@draper.com>
Message-ID: <025a01bfa106$e3511200$35dd3fcb@neil>

> I have installed PythonWin and win32all, but I seem to be completely 
> unable to run a function from a dll. I'm loading the dll with 
> win32api.LoadLibrary, but I can't figure out how to access the 
> functions. 

   You need Sam Rushing's calldll module:
http://www.nightmare.com/squirl/python-ext/calldll.zip

   You can often find useful Python code in the Vaults of Parnassus:
http://www.vex.net/parnassus/

   Neil





From rich at anna.UUCP  Mon Apr  3 01:20:30 2000
From: rich at anna.UUCP (Rich James)
Date: Mon, 03 Apr 2000 05:20:30 GMT
Subject: Just doing local file testing
References: <38E796BC.23191B2@regnoc.com> <00040214141502.04535@quadra.teleo.net> <38E7D6DE.EA09C891@regnoc.com> <00040216504403.04535@quadra.teleo.net>
Message-ID: <ySVF4.16805$9m6.677485@newsread1.prod.itd.earthlink.net>

It's well enough that you can open an HTML file with your browser and view it locally, but you cannot launch CGI scripts
this way.  They need to be launched by a web server -- so you will need to get into Apache.  Most likely, on your RH6.1
system, Apache is already set up and running.  Look for httpd.conf in your filesystem and review it.  Look for a line
with ScriptAlias in it and that will tell you where you are configured to run cgi scripts from.  Put your script in that
directory.

According to Patrick Phalen  <python-list at teleo.net>:
> Yeeks. You forgot to turn off HTML formatting for your message.
> 
> :: I'm not even to the point of using Apache yet - I'm just to launch the
> :: script locally from within the browser by having the form element point
> :: to a script on disk.&nbsp; In other words, this is just a development environment
> :: using only local files.
> :: <p>Any other clues?
> 
> My original advice still holds, but now I'll add that you probably need
> to do some additional study on client/server computing.
> 
> IOW, your browser is the client. Where's the server?
-- 
War is the health of the State.  
-- Randolph Bourne (1917)
---
Rich James


From dan at cgsoftware.com  Fri Apr 14 20:04:47 2000
From: dan at cgsoftware.com (Daniel Berlin)
Date: Fri, 14 Apr 2000 17:04:47 -0700 (PDT)
Subject: The Opposite of list(a,b,c)?
In-Reply-To: <38F7A07F.9B0F309C@bigfoot.com>
Message-ID: <Pine.LNX.4.10.10004141656190.3969-100000@propylaea.anduin.com>

On Fri, 14 Apr 2000, Andrew P. Jewell wrote:

<snip>
> def func(SearchTerm):
>   <...>
>     retval = []
>     for keyval in (dtnA.keys()):
>         mtch = re.search(patSrchTrm, keyval)
>         if mtch:
>             retval.append(UNLIST(dtnA[keyval]))

Change this to:

if mtch:
	for value in dtnA[keyval]:
		retval.append(value)



> 
> I want to end up with just a list of matching values from the dictionary,
> ala:
>     ['val1, val2', 'val3, val4']
> 
> But I can only get a list of lists: [[val1,val2],[val3,val4]] - which
> makes perfect sense based on how I'm doing it - but I don't WANT a list
> of lists!  Is there anyway to do this?  I don't really want to tuple-ize
> my dictionary since it's attached to its list values in other ways.  It
> seems ugly to have to iterate my dictionary value list to UN-list them
> manually - and the repr deal is REALLY ugly. What am I missing??  Thanks
> for any help!

Okay, you don't want to iterate?
Let something else iterate for you:
>>> def delist(x, retval):
...     import types
...     for a in x:
...             if type(a)==types.ListType:
...                     delist(a,retval)
...             else:
...                     retval.append(a)
...     return retval
...
>>> delist([5,6],[])
[5, 6]
>>> delist([5,6,[5,6]],[])
[5, 6, 5, 6]
>>> delist([5,6,[5,6],[5,[5,5]]],[])
[5, 6, 5, 6, 5, 5, 5]
>>>
etc
--Dan
> 
> Andy
>  




From gmcm at hypernet.com  Wed Apr 19 10:55:05 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Wed, 19 Apr 2000 10:55:05 -0400
Subject: __del__ problem - more details provided. still stumped though.
In-Reply-To: <7ojL4.9424$HG1.254543@nnrp1.uunet.ca>
Message-ID: <1255966388-25179470@hypernet.com>

Warren Postma wrote:
[vast snippery]
> class DBPtr :
>     def __init__(self,this):
>         self.this = this
>         self.thisown = 0
>     def __del__(self):
>         if self.thisown == 1 :
>             dbc.delete_DB(self.this)  # appears some exceptions being raised
> here.
>     # rest of class DBPtr snipped

Patch that to 
  def __del__(self, dbc=dbc):
     ...etc.

The problem is that the Py_Finalize logic has a real hard time 
knowing in what order to zap things. By the time this __del__ 
runs, the dbc module has already been zapped. Squirrelling 
away a reference in the method definition will ensure it's 
existence.  In general, a __del__ method (or anything called 
from a __del__ method) cannot rely on any references outside 
the object unless it uses this trick.

> That still leaves the two messages regarding unknown attribute named "time".
> the only reference to time in my modules is calls to time.time() in a
> function that could be called in the destructor of the table objects, while
> updating the table headers one last time before closing the file. 

Same thing.

- Gordon



From BgPorter at NOacmSPAM.org  Thu Apr 20 11:16:36 2000
From: BgPorter at NOacmSPAM.org (Brett g Porter)
Date: Thu, 20 Apr 2000 15:16:36 GMT
Subject: Pythonwin?
References: <eMz2Nmiq$GA.73@net025s> <4QtL4.24$v85.41@news-server.bigpond.net.au> <chuL4.8483$O7.17315@typhoon.southeast.rr.com> <7LDL4.534$v85.2563@news-server.bigpond.net.au>
Message-ID: <obFL4.9429$R4.40588@news1.rdc1.nj.home.com>

"Mark Hammond" <mhammond at skippinet.com.au> wrote in message
news:7LDL4.534$v85.2563 at news-server.bigpond.net.au...
> "Don Tuttle" <tuttledon at hotmail.com> wrote in message
> news:chuL4.8483$O7.17315 at typhoon.southeast.rr.com...
> > [Mark]
> > > You are really going to need to go through an MFC tutorial or book
> before
> > > you can make heads or tails of Pythonwin from that level...
> >
> > Any recommendations?
>
> Afraid not.  I already new MFC when I started Pythonwin.  I learnt it from
> the MFC "scribble" tutorial.  I find I tend to learn best with a tutorial
> and then playing - that is what I did with Python (back then there was no
> other alternative, but even now that is my recommended strategy for
> competant existing programmers learning Python...)
>
If you're not of this bent, I recommend these books for learning MFC (a
little light reading):

Programming Windows with MFC, Second Edition
By Prosise, Jeff
1337 Pages
Published by Microsoft Press
Date Published: 05/1999
ISBN: 1572316950

Professional MFC with Visual C++ 6
By Blaszczak, Mike
1200 Pages
Published by Wrox Press Inc.
Date Published: 08/1999
ISBN: 1861000154






From jmassung at magpiesystems.com  Tue Apr 18 16:01:32 2000
From: jmassung at magpiesystems.com (Jeff Massung)
Date: Tue, 18 Apr 2000 14:01:32 -0600
Subject: Another question
References: <dVFK4.41$vr2.5817@news.uswest.net> <38FB2EEB.A32AF88C@bam.com> <38FB339F.604364AD@bam.com> <200004171605.MAA04929@zippy.init.com> <38FB7610.80EAF965@ichips.intel.com> <200004172120.RAA11102@zippy.init.com> <jWNK4.432$vr2.32822@news.uswest.net> <m3ln2cckg6.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <p53L4.1217$zk4.22638@news.uswest.net>

Michael Hudson wrote in message ...
>But really, this is getting silly; if performance is this much of an
>obsession you aren't going to have that much fun with Python.

No, performance isn't a BIG issue for my program - and so far I love Python.
I'm just learning it and trying to stay away from "performance pitfalls".
It's not that I'm trying to squeeze every bit of speed out of Python, just
trying to get the best performance out of my program.

Two things I've learned from this thread:

1. Differences in code don't make THAT much difference in performance
(except in a few of the cases mentioned).

2. There are many different ways to do something (ala Perl) - which has
drawbacks and advantages.

Thanks, everyone!
Jeff




From tismer at tismer.com  Thu Apr 27 15:46:48 2000
From: tismer at tismer.com (Christian Tismer)
Date: Thu, 27 Apr 2000 21:46:48 +0200
Subject: Newbie Question
References: <sgh4qdmbqtj130@corp.supernews.com>
Message-ID: <39089928.47B55E1B@tismer.com>


Arnaldo Riquelme wrote:
> 
> Suppose I have a file name tick.tkr that looks like this
> 
> SUNW.
> MFST.
> FOO.
> SPAM.
> MOO.
> End_Of_File
> 
> I want to read the file into a list
> 
> myfile = open("tick.tkr", "r")
> mylist = myfile.readlines()
> 
> Now I have mylist with the contents of tick.tkr. "mylist" looks like this
> ['SUNW.\012', 'MSFT.\012', ..........'End_of_File/012']

Sure? :-)

> How do I read the file into a list without getting the newline character, so
> I can accomplish this
> 
> for tic in mylist:
>     #create file tic.prn
> 
> I know the other P language has something like 'chop' or something like
> that, but I'm not going there.

If you don't care about leading and trailing whitespicy,
you would be lazy and just do this:

import string
mylist = open("tick.tkr").readlines()
mylist = map(string.split, mylist)

If you just want to loose trailing crap, use string.rstrip .

But if you want to be sure only to loose \n, and also might
want to solve the problem that the list line might have no \n,
you are better off with a loop, like so: (after reading)

for i in range(len(mylist)):
    if mylist[i][-1:] == "\n":
        mylist = mylist[:-1]

Finally, here is what I use when I want to have all the lines,
ignoring line ends, but dropping empty lines:

import string
txt = open("tick.tkr").read()  # slurp it all
mylist = string.split(txt, "\n")
# this splits at line ends and removes them,
# but you get probably an empty entry at the end.
mylist = filter(None, mylist) # remove all empty lines.

ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaunstr. 26                  :    *Starship* http://starship.python.net
14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     where do you want to jump today?   http://www.stackless.com



From effbot at telia.com  Sun Apr  2 18:53:52 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Sun, 02 Apr 2000 22:53:52 GMT
Subject: Some bugs in Pythonwin b.128
References: <Pine.OSF.4.10.10003281852310.6437-100000@prfdec.natur.cuni.cz> <_NaE4.6119$V56.24967@news-server.bigpond.net.au> <8F0AAC6BEm69ph83qstudentssuse@127.0.0.1> <EUPF4.2880$1b4.14419@news-server.bigpond.net.au>
Message-ID: <4cQF4.4791$74.76270@newsc.telia.net>

Mark Hammond wrote:
> > Try changing these settings in your config file:
> >     Shift-9           = <<paren-open>>
> >     Shift-0           = <<paren-close>>
>
> That is excellent to know.  Does anyone know how I can determine
> the keyboard layout?  ie, how can I programatically determine
> where the paren keys are?

umm.  why should you?  shouldn't you either use TranslateMessage
and look for "(" or ")" WM_CHAR events, or use ToAscii/ToUnicode to
do the translation "inline"?

(in Tkinter, all you need to do is to bind to <Shift-(> and <Shift-)>,
but I supports you're not using that one here :-)

</F>




From embed at geocities.com  Mon Apr  3 14:48:22 2000
From: embed at geocities.com (Warren Postma)
Date: Mon, 3 Apr 2000 13:48:22 -0500
Subject: lightweight embedded database table object
Message-ID: <sS4G4.2894$HG1.86777@nnrp1.uunet.ca>

I am trying to find or build a small python database for a deeply embedded
Python application, that runs on an embedded piece of hardware we're
developing. The hardware is
basically PC compatible, and runs an embedded realtime operating system.
Python runs great on it, but I have to have ALL C modules statically linked
in.

I have attached some sample code to show the sorts of things I'm trying to
do.  The performance of the pure-python isn't great, I'd rather use a
compiled C python module
that did similar things. Has anyone written one.

I've looked at Gadfly, and since Gadfly can consume loads of RAM, it's not
exactly ideal for me. I don't really need an SQL grammar, but if it weren't
for the RAM issue, I'd consider it.

The lwdb module I've attached uses BSDDB v2 as the disk-based
store-and-retrieve method.

It adds ascii-delimited (CSV) import and export features (not well tested
yet), and a trivial assortment of types (say, float, string and integer for
now, but I'll be adding others like datetime fields, etc).

I'd like an ultra-compact binary storage method and maybe even a very-fast
compression method, and I'm thinking of adding support for the "ROTOR"
builtin Python module, so that the database store can be somewhat "opaque".
I will be storing user names and passwords in there, so I'll probably want
to add some encryption.

Perhaps all those things I can do with Python plus a little marshalling or
pickling are worth the overhead of a BSDDB file, which seems to use about
200K of disk space for every 100K of data in the variable-length-record
field.

I've considered things like dBase files, but they are such a crusty old
standard, really.  If I had to go that way, I would buy Codebase from
Sequiter Software, and whip up a C to Python interface to them. Anyone else
done any Python+Codebase work?

I don't really need wicked-fast lookup on disk. I'm wondering if I should
just invent my own C module to do this job.  Then again, I could be very
foolish, as any "database" code you write ends up always being more
complicated than you originally thought.

Warren Postma

----

attached sample code:
     lwdb.py - lightweight database table object (stores in a bsd database)
     lwdbrec.py - lightweight database record object (primitive
typed-database-'tuple')

[I have no idea of attachments bounce from the mailing list or newsgroup, or
if they considered In Bad Taste, but as they are small, I'm going ahead
anyways....]





begin 666 lwdbrec.py
M(B(B#0H@;'=D8G)E8RYP>2 M(&1E9FEN97, at 82!L:6=H='=E:6=H="!D871A
M8F%S92!O8FIE8W0 at +2!V97(@,"XP,2 M(&UA<F-H(#(P,# -"B @(" @*$,I
M(#(P,# @5V%R<F5N(%!O<W1M82P at 6E12($-O;G1R;VP at 4WES=&5M<PT*#0H@
M,2X at 8V]N=&%I;G, at 86X@871T<FEB=71E(&YA;65D(')O=RP@=VAI8V@@:7,@
M86X at 87)R87D@;V8@=F%L=65S+"!O;F4 at 9F]R#0H@(" @96%C:"!R;W<N#0H-
M"B R+B!!(&-L87-S+6%T=')I8G5T92!C86QL960@)V%T=')I8G5T97,G(&1E
M9FEN97,@=&AE(&%T=')I8G5T97,@=&AA="!E86-H#0H@(" @<F]W('1Y<&4@
M;75S="!H;VQD+B @268 at 82!R;W<@=F%L=64@:7,@;W!T:6]N86P@=&AE;B!R
M;W=;3ET at 8V%N(&)E($YO;F4-"B @("!W:&EC:"!I<R!E<75I=F%L96YT('1O
M(&$@3E5,3"!D871A8F%S92!F:65L9"X-"@T*(#,N($5A8V@@8VQA<W,@;75S
M="!I;7!L96UE;G0 at 82!K97DH*2!M971H;V0@=&AA="!R971U<FYS(&$@:V5Y
M('9A;'5E+"!W:&EC: T*(" @(&UU<W0 at 96ET:&5R(&)E(&$@<W1R:6YG+"!O
M<B!C;VYV97)T<R!T;R!A('5N:7%U92!K97D@<W1R:6YG('5S:6YG('1H90T*
M(" @('-T86YD87)D('!Y=&AO;B!S='(H*2!F=6YC=&EO;BX at 6VYO=&4Z(&UI
M9VAT(&1E8VED92!L871E<B!T;R!S=&%N9&%R9&EZ90T*(" @(&]N(')E<'(H
M*2!F;W(@:V5Y+7-T<FEN9R!G96YE<F%T:6]N+B!S=&%Y('1U;F5D+ET-"@T*
M(#0N(%-I;F-E('=E(&%R92!0:6-K;&EN9R!T:&5S92!R;W=S+"!W92!D;VXG
M="!W86YT('1O('-T;W)E('1H92!C;VQU;6X@;F%M97,-"B @("!E86-H('1I
M;64@=V4@<W1O<F4 at 82!R;W<L('-O('=E(&]N;'D@<W1O<F4@=&AE(&%R<F%Y
M(&]F(')O=R!V86QU97,L(&%N9 T*(" @('=E)VQL('-T;W)E(&$@<W!E8VEA
M;"!R96-O<F0@<V]M97=H97)E('=I=&@@86QL('1H92!O=&AE<B!I;F9O<FUA
M=&EO;BP-"B @("!I;F-L=61I;F<@=&AE(&1A=&%B87-E+7-C:&5M82!A;F0@
M<V\@;VXN#0H-"B U+B!2;W=S(&%R92!E87-I;'D at 8V]N=F5R=&5D(&9R;VTO
M=&\@87-C:6D at 9&5L:6UI=&5D("AC<W8I(&9O<FUA=',N(" @(" @#0H@#0H@
M5T%23DE.1SH at 2$E'2$Q9(%5.4D5,24%"3$4 at 15A015))345.5$%,($-/1$4A
M(%5312!!5"!/5TX at 4DE32R$-"B -"B(B(@T*#0IF<F]M('1Y<&5S(&EM<&]R
M=" J( T*:6UP;W)T(&5X8V5P=&EO;G,L(&]S+"!S>7,L('1I;64L('-T<FEN
M9RP@<'!R:6YT#0H-"@T*(R!R97%U:7)E9"!F:65L9"!F;&%G<S\-"G)E<75I
M<F5D(#T@,0T*;W!T:6]N86P@/2 P#0H-"F-S=E]S97 @/2 B+"(-"@T*#0HC
M($9I96QD(&1E9FEN:71I;VX at 871T<FEB=71E<RP at 8GD@=&AE:7(@<&]S:71I
M;VXZ#0H-"E]P871T97)N7V9O<F1E<B ](# @(R!W:&%T(&]R9&5R(&1O97,@
M:70 at 9V5T('!R:6YT960@=&\@;VX@=&AE('-C<F5E;B H0U-6(&9O<FUA="D-
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M(&ES('1H92!I;FET:6%L('9A;'5E('=H96X at 82!N97<@;VYE(&ES(&-R96%T
M960-"@T*(R!.86UE<W!A8V4 at 9F]R(%-A9F4 at 179A;'5A=&EO;G,L('5S960@
M:6X@<V5T7V-S=E]R;W<Z#0H-"G-A9F5?979A;" ]('L@(E]?8G5I;'1I;G-?
M7R(@.B![(").54Q,(B Z($YO;F4@?2!]#0H-"F-L87-S($Q71&)296-%<G)O
M<BA!<FET:&UE=&EC17)R;W(I. at T*(" @("(B(EI44B!$871A8F%S92!296-O
M<F0 at 3W!E<F%T:6]N($9A:6QE9"(B(@T*(" @('!A<W,-"@T*9&5F(&-S=E]V
M86PH>"DZ#0H@(" @(F-O;G9E<G0 at 82!V87)I86)L92!T;R!A('-T<F%I9VAT
M($%30TE)(&9O<FUA="P at 8F%C:W-L87-H+7%U;W1E<R!A;&P@<W1R:6YG(&1A
M=&$B#0H@(" @:68@*'@@/3T at 3F]N92DZ#0H@(" @(" @(')E='5R;B B3E5,
M3"(-"B @("!I9B H='EP92AX*2 ]/2!3=')I;F=4>7!E*3H-"B @(" @(" @
M<F5T=7)N("<B)RMS=')I;F<N<F5P;&%C92AX+" G(B<L('(G7"(G*2LG(B<@
M(R!H86YD;&5S(&5M8F5D9&5D('%U;W1E<PT*(" @(&5L<V4Z#0H@(" @(" @
M(')E='5R;B!S='(H>"D@(R!T:&4@<W1R:6YG('1H870@=VEL;"!R97!R;V1U
M8V4 at 6"!I9B!E=F%L=6%T960-"@T*9&5F(&UA:V5?8W-V7VAE861E<BAF:65L
M9',I. at T*(" @(")R971U<FX@<F]W(&%S(&$@8V]M;6$@<V5P87)A=&5D('9A
M<FEA8FQE("AA<V-I:2D@<W1R:6YG(@T*(" @(&X@/2!F:65L9'-;,%T-"B @
M("!F;W(@:2!I;B!F:65L9'-;,3I=. at T*(" @(" @(" @;B ](&X@*R!C<W9?
M<V5P("L@:2 C(&-O;6UA('-E<&%R871E9 T*(" @(')E='5R;B!N#0H-"B @
M(" -"B,@;'=D8E]R96,Z($-L87-S('=I=&@@<W1R:6-T('1Y<&4 at 8VAE8VMI
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M;&0 at 8F4@8V%L;&5D(&)Y(&$@<F]W)W, at 7U]I;FET7U\I(@T*(" @(" @("!I
M9B H;&5N*'-E;&8N871T<FEB=71E<RD]/3 I. at T*(" @(" @(" @(" @<F%I
M<V4 at 3%=$8E)E8T5R<F]R+"!@<V5L9BY?7V-L87-S7U]@*R(N871T<FEB=71E
M<R!N;W0@<V5T(2(-"@T*(" @(" @("!I9B H;&5N*'-E;&8N;W)D97(I/3TP
M*3H-"B @(" @(" @(" @(')A:7-E($Q71&)296-%<G)O<BP at 8'-E;&8N7U]C
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M(&]R('-T<FEN9RP at 9V]T("(K<W1R*'1Y<&4H<F]W*2D-"@T*#0H@(" @(" @
M('-E;&8N7U]D:6-T7U];(E]D:7)T>2)=(#T@, T*#0H@(" @9&5F(')E<V5T
M7W1O7V1E9F%U;'1S*'-E;&8I. at T*(" @(" @(" B9FEL;"!I;B!D969A=6QT
M('9A;'5E<R!F;W(@96%C:"!A='1R:6)U=&4B#0H@(" @(" @(&X@/2!L96XH
M<V5L9BYO<F1E<BD-"B @(" @(" @<V5L9BY?7V1I8W1?7ULB<F]W(ET@/2!;
M3F]N95TJ;B C(&-R96%T92!A;B!E;7!T>2!A<G)A>2QL96YG=&@@;BX@(" @
M( T*(" @(" @("!F;W(@>"!I;B!R86YG92 at P+&XI. at T*(" @(" @(" @(" @
M9FEE;&0@/2!S96QF+F]R9&5R6WA=( T*(" @(" @(" @(" @82 ]('-E;&8N
M871T<FEB=71E<UMF:65L9%T@(R!G970 at 86X@871T<FEB=71E(&)Y(&9I96QD
M(&YA;64-"B @(" @(" @(" @('-E;&8N7U]D:6-T7U];(G)O=R)=6WA=(#T@
M85M?<&%T=&5R;E]F:6YI='9=#0H@(" @(" @(" @(" C<')I;G0 at 9FEE;&0L
M(B ]("(L8&%;7W!A='1E<FY?9FEN:71V76 -"B @(" @(" @<V5L9BY?7V1I
M8W1?7ULB7V1I<G1Y(ET@/2 P#0H-"B @("!D968 at 9V5T:V5Y;F%M92AS96QF
M*3H-"B @(" @(" @(G)E='5R;B!T:&4@<')I;6%R>2!K97D@;F%M92(-"B @
M(" @(" @<F5T=7)N('-E;&8N;W)D97);,%T-"B @(" -"B @("!D968@<V5T
M7V9R;VU?9&EC="AS96QF+&1I8W0I. at T*(" @(" @(" B87-S:6=N('9A;'5E
M<R!T;R!F:65L9',@=VET:"!F:65L9&YA;65S(&%N9"!V86QU97,@:6X@=&AE
M(&1I8W1I;VYA<GD@)V1I8W0G(@T*(" @(" @("!F;W(@82!I;B!D:6-T+FME
M>7,H*3H@(R!G970 at 82!K97D-"B @(" @(" @(" @('-E;&8N7U]S971A='1R
M7U\H82QD:6-T6V%=*0T*( T*(" @(&1E9B!K97DH<V5L9BDZ#0H@(" @(" @
M(")R971U<FX@=&AE(&ME>2!U<V5D('=H96X@<W1O<FEN9R!O<B!L;V]K:6YG
M('5P(&$@<F]W(@T*(" @(" @("!T<GDZ#0H@(" @(" @(" @("!R971U<FX@
M<V5L9BY?7V1I8W1?7ULB<F]W(EU;,%T-"B @(" @(" @97AC97!T($ME>45R
M<F]R. at T*(" @(" @(" @(" @<F%I<V4 at 3%=$8E)E8T5R<F]R+" B36ES<VEN
M9R]I;G9A;&ED(&ME>3H@)R(K<V5L9BYK97EN86UE*R(G(@T*#0H@(" @9&5F
M(&ME>71Y<&4H<V5L9BDZ#0H@(" @(" @(')E='5R;B!S96QF+F%T=')I8G5T
M97-;,%U;7W!A='1E<FY?9G1Y<&5=.PT*(" @(" @(" -"B @("!D968 at 7U]R
M97!R7U\H<V5L9BDZ#0H@(" @(" @(")D969A=6QT(')E<')E<V5N=&%T:6]N
M(&]F(&1A=&%B87-E(')E8V]R9"!I<R!A<R!A(&-S=B!R;W<B#0H@(" @(" @
M(')E='5R;B!S96QF+F-S=E]R;W<H*0T*(" -"B @("!D968 at 7U]S971A='1R
M7U\H<V5L9BQN86UE+'9A;'5E*3H-"B @(" @(" @(F-U<W1O;2!?7W-E=&%T
M=')?('=I=&@@='EP97-A9F5T>2!C:&5C:R(-"B @(" @(" @82 ]('-E;&8N
M871T<FEB=71E<UMN86UE70T*(" @(" @("!I9B H=F%L=64@/3T at 3F]N92DZ
M#0H@(" @(" @(" @("!I9B H85M?<&%T=&5R;E]F<F5Q9%TI. at T*(" @(" @
M(" @(" @(" @(')A:7-E($Q71&)296-%<G)O<BP@(D-A;B=T('-E="!F:65L
M9"!T;R!.54Q,("AI92!.;VYE*2(-"B @(" @(" @(" @('-E;&8N<F]W6V%;
M,%U=(#T at 3F]N90T*(" @(" @("!E;'-E. at T*(" @(" @(" @(" @=')Y.B @
M(" @(" @(" @( T*(" @(" @(" @(" @(" @(&EF("AT>7!E*'9A;'5E*2 ]
M/2!A6U]P871T97)N7V9T>7!E72DZ#0H@(" @(" @(" @(" @(" @(" @('-E
M;&8N<F]W6V%;,%U=(#T@=F%L=64-"B @(" @(" @(" @(" @(" @(" @<V5L
M9BY?7V1I8W1?7ULB7V1I<G1Y(ET@/2 Q#0H@(" @(" @(" @(" @(" @96QS
M93H-"B @(" @(" @(" @(" @(" @(" @<F%I<V4 at 3%=$8E)E8T5R<F]R+" B
M26YV86QI9"!T>7!E(&EN(&%S<VEG;FUE;G0B*V!T>7!E*'9A;'5E*6 -"B @
M(" @(" @(" @(&5X8V5P="!+97E%<G)O<CH-"B @(" @(" @(" @(" @(" C
M<')I;G0@(F%T=')I8G5T97,Z("(L<V5L9BYA='1R:6)U=&5S#0H@(" @(" @
M(" @(" @(" @<F%I<V4 at 2V5Y17)R;W(L("));G9A;&ED(&%T=')I8G5T92 B
M*V!N86UE8 T*(" @(" @(" @(" @#0H@(" @9&5F(%]?9V5T871T<E]?*'-E
M;&8L;F%M92DZ#0H@(" @(" @(")G970 at 871T<FEB=71E(&9R;VT at 8W-V+7)O
M=R!A='1R:6)U=&5S+"!O<B!I;G-T86YC92P@;W(@8VQA<W,B#0H@(" @(" @
M(", at 9FER<W0@<V5A<F-H(')O=R!A='1R:6)U=&5S. at T*(" @(" @("!T<GDZ
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M70T*(" @(" @("!S='(Q(#T at 8W-V7W9A;"@@<F]W6S!=("D@(R!C;VYV97)T
M('1O('-T<FEN9W,-"B @(" @(" @9F]R(&D@:6X@<F]W6S$Z73H-"B @(" @
M(" @(" @('-T<C$@/2!S='(Q("L at 8W-V7W-E<" K(&-S=E]V86PH:2D@(W-E
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M97 ]8W-V7W-E<"DZ#0H@(" @(" @(")C;VYV97)T(&$@<F]W(&9R;VT at 8W-V
M(&9O<FUA="X@:71E;7, at 35535"!B92!S97!A<F%T960 at 8GD@8V]M;6%S("$B
M#0H@(" @(" @('1R>3H-"B @(" @(" @(" @('-E;&8N7U]D:6-T7U];(G)O
M=R)=(#T at 979A;"@@(ELB*W)O=RLB72(L('-A9F5?979A;" @*0T*(" @(" @
M(" @(" @;B ](&QE;BAS96QF+F]R9&5R*2UL96XH<V5L9BYR;W<I#0H@(" @
M(" @(" @("!F;W(@:2!I;B!R86YG92 at P+&XI. at T*(" @(" @(" @(" @(" @
M('-E;&8N7U]D:6-T7U];(G)O=R)=+F%P<&5N9"A.;VYE*0T*(" @(" @("!E
M>&-E<'0 at 3F%M945R<F]R. at T*(" @(" @(" @(" @<F%I<V4 at 3%=$8E)E8T5R
M<F]R+" G<V5T7V-S=E]R;W<@+2!I;G9A;&ED(')O=R M(&-H96-K(&EF('-T
M<FEN9W, at 87)E(&%L;"!Q=6]T960Z("<K<F]W#0H@(" @(" @(" @(" -"B @
M(" @(" @#0H@(" @(" @( T*(" @(" @(" -"@T*(" @(&1E9B!V86QI9&%T
M92AS96QF*3H-"B @(" @(" @(G)A:7-E<R!A;B!E>&-E<'1I;VX@:68@=F%L
M:61A=&EO;B!F86EL<RP@:64@<F5Q=6ER960 at 9FEE;&0@;F]T('!R;W9I9&5D
M(@T*(" @(" @("!F;W(@>"!I;B!S96QF+F%T=')I8G5T97,N:V5Y<R at I.@T*
M(" @(" @(" @(" @:68@<V5L9BYA='1R:6)U=&5S6WA=6U]P871T97)N7V9R
M97%D73H-"B @(" @(" @(" @(" @("!I9B!S96QF+E]?9&EC=%]?6WA=(#T]
M($YO;F4Z#0H@(" @(" @(" @(" @(" @(" @(')A:7-E($Q71&)296-%<G)O
M<BP@(G9A;&ED871E*"D at 9F%I;&5D("T@<F5Q=6ER960 at 9FEE;&0@;F]T('-E
M="(-"@T*#0H-"B,@<')O8FQE;3H at 66]U(&-A;B!H879E(&9I96QD(&YA;64L
M(&]R(&9I96QD(&EN9&5X(&%S('1H92!P<FEM87)Y#0HC(&QO;VMU<"P at 8G5T
M(&YO="!B;W1H+B!3;R!W92!C<F5A=&4 at 8F]T:"!A(&YA;64@;&]O:W5P(&1I
M8W1I;VYA<GD-"B, at 86YD('1H96X@<V]R="!T:&%T(&)Y(&9I96QD(&]R9&5R
M:6YG('1O(&-R96%T92!A(&9I96QD(&]R9&5R:6YG('1H870-"B,@=V]R:W,-
M"F1E9B!L=V1B7V]R9&5R*&%T=')I8G,I. at T*(" @(&ME>7,@/2!A='1R:6)S
M+FME>7,H*3L-"B @("!N(#T@;&5N*&ME>7,I#0H@(" @;W)D97(@/2!;3F]N
M95TJ;B C(&-R96%T92!A<G)A>2!L96YG=&@@3 at T*(" @(&9O<B!I(&EN(&ME
M>7,Z#0H@(" @(" @(&]R9&5R6R!A='1R:6)S6VE=6S!=(%T@/2!I#0H@(" @
M<F5T=7)N(&]R9&5R#0H-"B @(" -"@T*>" ]($YO;F4@(R!M86ME('@@<W1I
M8VL at 87)O=6YD(&%F=&5R('1E<W0H*0T*#0ID968@=&5S="@I. at T*(" @('!R
M:6YT("),:6=H='=E:6=H="!$871A8F%S92!296-O<F0 at 0VQA<W, at 5&5S=',B
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M=')I8G5T92 P('=I;&P at 8F4@=7-E9"$-"B @(" @(" @(" @(&1E9B!?7VEN
M:71?7RAS96QF+&1I8W0]3F]N92DZ#0H@(" @(" @(" @(" @(" @<V5L9BYS
M971U<"AD:6-T*0T*(" @('!R:6YT(")R96-O<F0 at 8W)E871I;VX@=&5S=',N
M+BXB#0H@(" @<')I;G0@(F-S=E]V86P@=&5S=',Z(@T*(" @('!R:6YT(")C
M<W9?=F%L*#4I/2 B+&-S=E]V86PH-2D-"B @("!P<FEN=" B8W-V7W9A;"A<
M(F)O8BP at 7%Q<(G1H92!S;&]B7%Q<(BD](BP at 8W-V7W9A;"@B8F]B+"!<(G1H
M92!S;&]B7"(B*0T*#0H@(" @<')I;G0@(F-R96%T92!P97)S;VX at 8VQA<W,N
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M(F1U;7 @<&5R<V]N(&ME>2!V86QU92XN+B(-"B @("!P<FEN="!X+FME>2 at I
M#0H@(" @<')I;G0@>"YK97EN86UE#0H@(" @<')I;G0@(B B#0H@(" @<')I
M;G0@(F-R96%T92!A;F]T:&5R('!E<G-O;B(-"B @("!Z(#T@;'=D8E]P97)S
M;VXH*0T*(" @('HN<V5T7V-S=E]R;W<H("<@(D)O8B(L(" B,2!3;W5T:&=A
M=&4B+" T,R G("D-"B @("!Z<W1R/2!Z+F-S=E]R;W<H*0T*(" @('!R:6YT
M('IS='(-"B @("!D96P@>@T*(" @('H@/2!L=V1B7W!E<G-O;B at I#0H@(" @
M<')I;G0@(G)E8V]V97(@<F]W(&9R;VT at 8W-V(&]U='!U=#HB#0H@(" @>BYS
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27U\G. at T*(" @('1E<W0H*0T*
`
end

begin 666 lwdb.py
M(B(B#0H@;'=D8BYP>2 M($QI9VAT=V5I9VAT($1A=&%B87-E(%1A8FQE($]B
M:F5C=', at +2!697)S:6]N(# N,#$@+2!-87)C:" R,# P#0H@#0H@(" @(" @
M(" H0RD@,C P,"!787)R96X at 4&]S=&UA+"!:5%(@0V]N=')O;"!3>7-T96US
M#0H-"B!);7!L96UE;G0 at 82!L:6=H='=E:6=H="!R;W<M8F%S960 at 9&%T86)A
M<V4L('5S:6YG(%!I8VML960 at 07)R87ES(&]F(%!Y=&AO;@T*(%-T<FEN9W,L
M($EN=&5G97)S+"!E=&,L(&%S(&$@9F%S="!E9F9I8VEE;G0 at 9&%T82US=&]R
M86=E(&%N9"!L;V]K=7 -"B!T96-H;FEQ=64N#0H-"B!)9B!Y;W4@;F5E9"!T
M;R!P97)S:7-T(&%R8FET<F%R>2!0>71H;VX@;V)J96-T<R!I;B!A(&9I;&4@
M;W(@9&%T86)A<V4L#0H@>6]U(&UI9VAT('=A;G0 at 4VAE;'9E(&]R(&1B4VAE
M;'9E("AW:&EC:"!A;'-O('5S97,@=&AE(&)S9"!D871A8F%S92DN#0H-"B!)
M9B!O;B!T:&4@;W1H97(@:&%N9"!Y;W4@:&%V92!A('-E="!O9B!R;W=S+"!A
M;&P@;V8@=VAI8V@@87)E(&]F(&5X86-T;'D-"B!T:&4@<V%M92!D871A('1Y
M<&4L('=I=&@@=&AE('-A;64 at 871T<FEB=71E<R H8V%N;VYI8V%L(&5X86UP
M;&4@:7, at 82!L:7-T#0H@;V8@<&5O<&QE)W,@;F%M97,L(&%D9')E<W-E<RP@
M86YD('1E;&5L<&AO;F4@;G5M8F5R<RD@=&AE;B!T:&ES(&QI='1L90T*(&UO
M9'5L92!M87D at 8F4@;V8@=7-E('1O('EO=2X-"@T*(%=!4DY)3D<Z($A)1TA,
M62!53E)%3$E!0DQ%($584$5224U%3E1!3"!#3T1%(2!54T4 at 050@3U=.(%))
M4TLA#0H-"B(B(@T*#0II;7!O<G0 at 97AC97!T:6]N<PT*:6UP;W)T('-Y<PT*
M:6UP;W)T('-T<FEN9PT*:6UP;W)T(&]S#0II;7!O<G0@=&EM90T*9G)O;2 @
M('1Y<&5S(&EM<&]R=" J#0IT<GDZ#0H@(" @9G)O;2 @(&)S9&1B,B!I;7!O
M<G0 at 9&(@#0IE>&-E<'0 at 26UP;W)T17)R;W(Z#0H@(" @<F%I<V4 at 26UP;W)T
M17)R;W(L(")L=V1B+G!Y.B!9;W4@<F5A;&QY('-H;W5L9"!I;G-T86QL($)3
M1"!$871A8F%S92!697)S:6]N(#(N(@T*#0IF<F]M(" @8U!I8VML92!I;7!O
M<G0 at 4&EC:VQE<BP at 56YP:6-K;&5R( T*9G)O;2 @(&-3=')I;F=)3R!I;7!O
M<G0 at 4W1R:6YG24\-"F9R;VT@("!L=V1B<F5C("!I;7!O<G0@*@T*#0HC(&1E
M9F%U;'0@;&EG:'1W96EG:'0 at 9&%T86)A<V4 at 97AT96YS:6]N. at T*>G1R9&)?
M97AT(#T@(BYL9&(B#0H-"B, at 9&5F875L="!C;VYD:71I;VYS#0ID969?<W1A
M=&4@/2 P#0ID969?<V5V97)I='D@/2 S#0H-"@T*#0IC;&%S<R!,5T1B17)R
M;W(H07)I=&AM971I8T5R<F]R*3H-"B @(" B(B),:6=H='=E:6=H="!$871A
M8F%S92!/<&5R871I;VX at 1F%I;&5D(B(B#0H@(" @<&%S<PT*#0H-"B,@;7D@
M=F5R>2!O=VX@;&ET=&QE(&)I;F%R>2!S96%R8V at L(')E9'5N9&%N="P@<VEN
M8V4@=&AE<F4G<R!O;F4@:6X-"B, at 8FES96-T+G!Y+ at T*9&5F(%]B:6Y?<V5A
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M:7)S= T*(" @(&EF("AX/# I. at T*(" @(" @("!R86ES92!3=&%N9&%R9$5R
M<F]R+" B7V)I;E]S96%R8V at Z(&YO="!F;W5N9"(-"B @("!I9B H>#T],"DZ
M#0H@(" @(" @(&X@/2!F:7)S= T*(" @(&5L<V4Z#0H@(" @(" @(&X@/2!F
M:7)S=" K("AX*S$I+S(-"B @("!I9B!L:7-T6VY=(#T](&ET96TZ#0H@(" @
M(" @(')E='5R;B!N#0H@(" @96QI9B!I=&5M(#P@;&ES=%MN73H-"B @(" @
M(" @<F5T=7)N(%]B:6Y?<V5A<F-H*&QI<W0L:71E;2QF:7)S="QN+3$L9&5P
M=&@K,2D-"B @("!E;'-E. at T*(" @(" @("!R971U<FX at 7V)I;E]S96%R8V at H
M;&ES="QI=&5M+&XK,2QL87-T+&1E<'1H*S$I#0H@(" @(" @( T*(" @( T*
M(" @( T*#0ID968 at 7V-O;%]M87 H9G)O;5]O<F1E<BQT;U]O<F1E<BDZ( T*
M(" @("(B(@T*(" @(%]C;VQ?;6%P*&9R;VU?;W)D97(L=&]?;W)D97(I("T^
M(&UA<%]A<G)A>2P at 93$L(&4R#0H@(" @#0IF:6YD<R!C;W)R97-P;VYD:6YG
M(&5L96UE;G0@:6X@=&]?;W)D97(@9F]R(&5A8V@@96QE;65N="!I;@T*9G)O
M;5]O<F1E<BP at 86YD(')E='5R;G,@=&AE(&EN9&5X97,@:6X at 86X@87)R87D@
M=&\@<VAO=R!T:&4-"F-O<G)E<W!O;F1A;F-E+B!!;'-O(')E='5R;G,@='=O
M(&5R<F]R(&-O=6YT<RX at 5&AE(&9I<G-T(&ES#0IT:&4@;G5M8F5R(&]F(&ET
M96US(&EN(&9R;VU?;W)D97(@=VET:"!N;R!C;W)R97-P;VYD:6YG(&UA=&-H
M#0II;B!T;U]O<F1E<BP at 86YD('1H92!S96-O;F0@:7,@=&AE(&YU;6)E<B!O
M9B!I=&5M<R!I;B!T;U]O<F1E<@T*=&AA="!A<F5N)W0@:6X at 9G)O;5]O<F1E
M<BX-"@T*4W!E8VEA;"!#87-E.B!I9B!F<F]M7V]R9&5R(&%N9"!T;U]O<F1E
M<B!A<F4@=&AE('-A;64@;&5N9W1H+"!A;F0-"G1H92!I=&5M<R!A<F4@:6X@
M=&AE('-A;64@;W)D97(L('1H96X at 93$L(&%N9"!E,B!A<F4@<F5T=7)N960@
M87,-"BTQ('=H:6-H(&UE86YS(")N;R!C;VYV97)S:6]N(&YE8V5S<V%R>2(N
M#0H-"F5X86UP;&4Z#0H-"B @("!L:7-T+&4Q+&4R/5]C<W9?;6%P*"!;(")A
M(BPB8B(L(F,B+" B9"(@72P at 6R B8R(L(")A(BP@(F(B(%T@*0T*(" @('!R
M:6YT(&QI<W0@#0H@(" @6S$L(#(L(# L("TQ72 C(&UA=&-H97, at 82T^82P@
M8BT^8BP at 8RT^8RP at 8G5T(")D(B!H87,@;F\@;6%T8V@@#0H@(" @<')I;G0@
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M;W)D97(L(&QO;VME9"!U<"!B>2!N86UE. at T*(" @(&9O<B!T;U]I=&5M(&EN
M('1O7V]R9&5R. at T*(" @(" @("!M87 Q6W1O7VET96U=(#T@;@T*(" @(" @
M(" C('1R86-K('-T<F%I9VAT(&UA<'!I;F=S. at T*(" @(" @("!I9B H;&5N
M*&9R;VU?;W)D97(I/3UL96XH=&]?;W)D97(I*3H-"B @(" @(" @(" @(&EF
M("AF<F]M7V]R9&5R6VY=(#T]('1O7VET96TI. at T*(" @(" @(" @(" @(" @
M('-T<F%I9VAT(#T@<W1R86EG:'0@*R Q#0H@(" @(" @(&X];BLQ#0H@(" @
M(" @( T*(" @(",@;6%P,B!C;VYT86EN<R!T:&4@=')A;G-L871I;VX at 9G)O
M;2!S=&%R=&EN9R!O<F1E<B!T;R!F:6YA;"!O<F1E<CH@(" @(" @(" @(" -
M"B @("!M87 R(#T at 6UT-"B @("!F;W(@9G)O;5]I=&5M(&EN(&9R;VU?;W)D
M97(Z#0H@(" @(" @(&EF(&UA<#$N:&%S7VME>2AF<F]M7VET96TI. at T*(" @
M(" @(" @(" @;6%P,BYA<'!E;F0H(&UA<#%;9G)O;5]I=&5M72 I#0H@(" @
M(" @(" @("!V/78K,0T*(" @(" @("!E;'-E. at T*(" @(" @(" @(" @;6%P
M,BYA<'!E;F0H("TQ("D@(R!S:VEP(&ET96TA#0H@(" @:68@*'8]/3 I. at T*
M(" @(" @("!R971U<FX at 6UTL;&5N*&9R;VU?;W)D97(I+&QE;BAT;U]O<F1E
M<BD@(R!N;R!M871C:&5S+"!A;&P at 97)R;W)S(0T*(" @(&5L<V4Z#0H@(" @
M(" @(&EF("AS=')A:6=H=" ]/2!L96XH9G)O;5]O<F1E<BDI. at T*(" @(" @
M(" @(" @93$L93(]+3$L+3$-"B @(" @(" @96QS93H-"B @(" @(" @(" @
M(&4Q(#T@;&5N*&9R;VU?;W)D97(I+78-"B @(" @(" @(" @(&4R(#T at +3$J
M("@@*&QE;BAF<F]M7V]R9&5R*2UL96XH=&]?;W)D97(I*2UE,2 I#0H@(" @
M(" @(')E='5R;B!M87 R+&4Q+&4R#0H@(" -"@T*#0H-"@T*#0H-"@T*#0H@
M(" @#0H@#0H@(" @(" @(" @(" @#0HC(&=E;F5R:6,@=&%B;&4@:&%N9&QI
M;F<@8VQA<W,Z( T*8VQA<W,@;'=D8E]T86)L93H-"B @("!D968 at 7U]I;FET
M7U\H<V5L9BQT86)L96YA;64L9&5S8W)I<'1I;VXL(')O=U]C;&%S<U]O8FHL
M(&9I;&5N86UE+"!U<V5?=6YI<75E8VAE8VL],2QU<V5?875T;VEN8STP+'5S
M95]I;G1K97ES/3 I. at T*(" @(" @(" @<V5L9BYT86)L96YA;64@/2!T86)L
M96YA;64-"B @(" @(" @('-E;&8N9&5S8W)I<'1I;VX@/2!D97-C<FEP=&EO
M;@T*(" @(" @(" @<V5L9BYR;W=?8VQA<W-?;V)J(#T@<F]W7V-L87-S7V]B
M:@T*(" @(" @(" @<V5L9BYA9&1E9" ]($YO;F4@#0H@(" @(" @("!S96QF
M+FUO9&4@/2 P(",@," ](&)R;W=S:6YG+" Q(#T at 861D960@;F5W(')O=RP@
M;F]T('!O<W1E9"!Y970-"B @(" @(" @('-E;&8N8VQO<V5D/3$-"B @(" @
M(" @(&EF("AU<V5?=6YI<75E8VAE8VLI. at T*(" @(" @(" @(" @<V5L9BYU
M;FEQ=65C:&5C:R ](#$-"B @(" @(" @(&5L<V4Z#0H@(" @(" @(" @("!S
M96QF+G5N:7%U96-H96-K(#T@, T*(" @(" @(" @(" @#0H@(" @(" @("!I
M9B H=7-E7V%U=&]I;F,I. at T*(" @(" @(" @(" @('-E;&8N875T;VEN8R ]
M(# @(R!F96%T=7)E(&5N86)L960-"B @(" @(" @(&5L<V4Z#0H@(" @(" @
M(" @(" @<V5L9BYA=71O:6YC(#T at +3$@(R!F96%T=7)E(&1I<V%B;&5D#0H@
M(" @(" @("!I9B H=7-E7VEN=&ME>7,I. at T*(" @(" @(" @(" @('-E;&8N
M:6YT7VME>7,@/2 Q#0H@(" @(" @("!E;'-E. at T*(" @(" @(" @(" @('-E
M;&8N:6YT7VME>7,@/2 P#0H@(" @(" @("!S96QF+G-F;&%G(#T@," C('-T
M;W)E<R!F;&%G<R!V86QU92!F<F]M(&QA<W0 at 8V%L;"!T;R!O<&5N*"D-"B @
M(" @(" @(&EF(&9I;&5N86UE(#T]($YO;F4Z#0H@(" @(" @(" @(" @(" @
M(&9I;&5N86UE(#T@=&%B;&5N86UE*WIT<F1B7V5X= T*(" @(" @(" @<V5L
M9BYO<&5N*&9I;&5N86UE*0T*#0H@(" @9&5F(&-L;W-E*'-E;&8I. at T*(" @
M(" @(" B8VQO<V4 at 9FEL92!A;F0 at 8V]M;6ET(&%L;"!C:&%N9V5S(2(-"B @
M(" @(" @:68@*'-E;&8N;6]D92 ]/2 Q*3H@(R!P;W-T(&)E9F]R92!C;&]S
M:6YG(0T*(" @(" @(" @(" @<V5L9BYP;W-T*"D-"B @(" @(" @<V5L9BYU
M<&1A=&5?:&5A9&5R*"D-"B @(" @(" @<V5L9BYD8BYC;&]S92 at I#0H@(" @
M(" @('-E;&8N9&ER='D@/2 P#0H@(" @(" @('-E;&8N8VQO<V5D/3$-"@T*
M(" @(" @(" -"B @(" @(" @#0H@(" @9&5F(&]P96XH<V5L9BP at 9FEL96YA
M;64L(&9L86=S/61B+D1"7T-214%412P@;6]D93TP-C8V+"!C86-H93TP+"!F
M:6QE='EP93UD8BY$0E](05-(*3H-"B @(" @(" @<V5L9BYF:6QE;F%M92 ]
M(&9I;&5N86UE#0H@(" @(" @('-E;&8N8VQO<V5D/3 -"B @(" @(" @:68@
M='EP92AF;&%G<RD@/3T@='EP92 at G)RDZ#0H@(" @(" @(" @(" @(" @<V5L
M9BYS9FQA9R ](&9L86=S#0H@(" @(" @(" @(" @(" @:68@<V9L86<@/3T@
M)W(G. at T*(" @(" @(" @(" @(" @(" @("!F;&%G<R ](&1B+D1"7U)%041/
M3DQ9#0H@(" @(" @(" @(" @(" @96QI9B!S9FQA9R ]/2 G<G<G. at T*(" @
M(" @(" @(" @(" @(" @("!F;&%G<R ](# -"B @(" @(" @(" @(" @("!E
M;&EF('-F;&%G(#T]("=W)SH-"B @(" @(" @(" @(" @(" @(" @9FQA9W,@
M/2 @9&(N1$)?0U)%051%#0H@(" @(" @(" @(" @(" @96QI9B!S9FQA9R ]
M/2 G8R<Z#0H@(" @(" @(" @(" @(" @(" @(&9L86=S(#T@(&1B+D1"7T-2
M14%410T*(" @(" @(" @(" @(" @(&5L:68@<V9L86<@/3T@)VXG. at T*(" @
M(" @(" @(" @(" @(" @("!F;&%G<R ](&1B+D1"7U1254Y#051%#0H@(" @
M(" @(" @(" @(" @96QS93H-"B @(" @(" @(" @(" @(" @(" @<F%I<V4@
M3%=$8D5R<F]R+" B9FQA9W,@<VAO=6QD(&)E(&]N92!O9B G<B<L("=W)RP@
M)V,G(&]R("=N)R!O<B!U<V4@=&AE(&1B+D1"7RH at 9FQA9W,B#0H@(" @(" @
M('-E;&8N9&(@/2!D8BY$0B at I#0H@(" @(" @('-E;&8N9&(N:6YF;RYD8E]C
M86-H92 ](&-A8VAE#0H@(" @(" @('-E;&8N9&(N;W!E;BAF:6QE;F%M92P@
M9FEL971Y<&4L(&9L86=S+"!M;V1E*0T*(" @(" @("!T<GDZ#0H@(" @(" @
M(" @(" C(&-R96%T92!A('-T<FEN9R!T:&%T(&QO;VMS(&QI:V4 at 82!F:6QE
M#0H@(" @(" @(" @("!B=68]4W1R:6YG24\H<V5L9BYD8ELB(2)=*2 C('-P
M96-I86P@<F]W(&YA;65D("(A(B!I<R!T:&4 at 9&%T86)A<V4@:&5A9&5R+"!E
M=&,N#0H@(" @(" @(" @(" C(&-R96%T92!A(')O=R!C;&%S<R!O8FIE8W0@
M9G)O;2!A('-T86YD87)D(&1I8W1I;VYA<GDZ#0H@(" @(" @(" @("!S96QF
M+F1B:6YF;SU5;G!I8VML97(H8G5F*2YL;V%D*"D-"B @(" @(" @(" @(&EF
M("AS96QF+F%U=&]I;F,^/3 I. at T*(" @(" @(" @(" @(" @('-E;&8N875T
M;VEN8R ]('-E;&8N9&)I;F9O6R)A=71O:6YC(ET-"B @(" @(" @(" @('!R
M:6YT(")L;V%D960@<')E=FEO=7,L('-E;&8N875T;VEN8STB+'-E;&8N875T
M;VEN8PT*(" @(" @("!E>&-E<'0Z#0H@(" @(" @(" @("!P<FEN=" B;F5W
M('1A8FQE(@T*(" @(" @(" @(" @<V5L9BYD8FEN9F\@/2![?0T*(" @(" @
M(" @(" @<V5L9BYD8FEN9F];(F]R9&5R(ET@/2!S96QF+G)O=U]C;&%S<U]O
M8FHN;W)D97(@(R!F:65L9"!N86UE<R!A;F0@=&AE:7(@;W)D97)S(&EN('1H
M92!R;W<@87)R87D-"B @(" @(" @(" @('-E;&8N=7!D871E7VAE861E<B at I
M#0H@(" @(" @(" @("!R971U<FX@(R!D;VXG="!N965D('1O(&-H96-K(&-O
M;'5M;B!M87!P:6YG+"!A<R!W92!J=7-T(&-R96%T960@:70A#0H@(" @(" @
M(", at 8VAE8VL@8V]L=6UN(&UA<'!I;F<Z#0H@(" @(" @(&UA<"QE,2QE,B ]
M(%]C;VQ?;6%P*'-E;&8N9&)I;F9O6R)O<F1E<B)=+"!S96QF+G)O=U]C;&%S
M<U]O8FHN;W)D97(I#0H@(" @(" @(&EF("AE,3XP*6]R*&4R/C I.B C(&5R
M<F]R<R$_#0H@(" @(" @(" @("!R86ES92!,5T1B17)R;W(L(")M=7-T(')E
M<&%I<B at I(&1A=&%B87-E(&)E9F]R92!O<&5N:6YG(2!C;VQU;6X@;F%M97,@
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M:6UE+G1I;64H*0T*(" @(" @(" -"B @(" @(" @8G5F(#T at 4W1R:6YG24\H
M*2 -"B @(" @(" @<&L@/2!0:6-K;&5R*&)U9BPQ*0T*(" @(" @("!P:RYD
M=6UP*'-E;&8N9&)I;F9O*0T*(" @(" @("!S96QF+F1B6R(A(ET@/2!B=68N
M9V5T=F%L=64H*0T*#0H@(" @(" @( T*(R @("!D968 at 7U]D96Q?7RAS96QF
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M<B at Q,"D@*0T*(" @(" @("!K97ES(#T@<V5L9BYK97ES*"D@(R!N;W1E.B!S
M96QF+FME>7, at 86QR96%D>2!R96UO=F5S('1H92!S<&5C:6%L("(A(B!K97DN
M#0H@(" @(" @(&ME>7,N<V]R="@I#0H@(" @(" @("-P<FEN=" B:V5Y<ULP
M73TB+&ME>7-;,%T-"B @(" @(" @9F]R(&D@:6X@:V5Y<SH-"B @(" @(" @
M(" @(&8N=W)I=&4H('-E;&8N9V5T<F]W*&DI+F-S=E]R;W<H*2 K(&-H<B at Q
M,"DI#0H@(" @(" @(" @(" @(" -"B @(" @(" @9BYC;&]S92 at I#0H@(" @
M(" @(')E='5R;B!L96XH:V5Y<RD-"@T*(" @(&1E9B!C<W9?:6UP;W)T*'-E
M;&8L9FEL96YA;64L*3H-"B @(" @(" @(FEM<&]R="!A(&-S=B!F:6QE+"!C
M86X@:6UP;W)T('!A<G1I86P at 9FEL92!I;F9O<FUA=&EO;BP at 86YD(&-A;B!I
M9VYO<F4 at 97AT<F$@8V]L=6UN<R(-"B @(" @(" @9CUO<&5N*&9I;&5N86UE
M+")R(BD[#0H@(" @(" @(&5M<'1Y<F]W/5M.;VYE72IL96XH<V5L9BYR;W=?
M8VQA<W-?;V)J+F%T=')I8G5T97,I(", at 871T<FEB=71E(&-O=6YT(0T*(" @
M(" @("!L:6YE<R ](&8N<F5A9&QI;F5S*"D-"B @(" @(" @8V]L<R ](&5V
M86PH(ELB*VQI;F5S6S!=*R)=(BQS869E7V5V86PI#0H@(" @(" @(&QA<W0@
M/2!L96XH8V]L<RDM,0T*(" @(" @("!N965D;6%P(#T@,0T*(" @(" @("!C
M;VQ?;6%P+&4Q+&4R(#T at 7V-O;%]M87 H8V]L<RD-"B @(" @(" @:68@*&4Q
M/# I86YD*&4R/# I. at T*(" @(" @(" @(" @;F5E9&UA<" ](# @(R!D:7)E
M8W0 at 8V]P>2P@;F\@8V]N=F5R<VEO;@T*(" @(" @("!E;&EF(&4R(#X](&QE
M;BAC;VQS*3H-"B @(" @(" @(" @(')A:7-E($Q71&)%<G)O<BP@(DYO(&-O
M;'5M;B!N86UE<R!M871C:"X at 0V%N)W0@:6UP;W)T(&9I;&4A(@T*(" @(" @
M("!F;W(@;"!I;B!L:6YE<ULQ.ETZ#0H@(" @(" @(" @("!I9B!N965D;6%P
M. at T*(" @(" @(" @(" @(" @('(@/2!E;7!T>7)O=ULZ70T*(" @(" @(" @
M(" @(" @(&-O;',@/2!E=F%L*");(BML:6YE<ULP72LB72(L<V%F95]E=F%L
M*0T*(" @(" @(" @(" @(" @(&9O<B!N(&EN(')A;F=E*# L;&%S="DZ(",@
M;6%P(&EN=&\@<@T*(" @(" @(" @(" @(" @(" @("!M(#T at 8V]L7VUA<%MN
M70T*(" @(" @(" @(" @(" @(" @("!I9B!M(#X@,#H-"B @(" @(" @(" @
M(" @(" @(" @(" @(');;5T@/2!C;VQS6VY=#0H@(" @(" @(" @("!E;'-E
M. at T*(" @(" @(" @(" @(" @('(@/2!C;VQS6SI=(", at 8V]P>2!C;VQS(2!A
M=F]I9"!P<F]B;&5M<R$-"B @(" @(" @(" @(&YE=W)O=R ]('-E;&8N<F]W
M7V-L87-S7V]B:B at I(", at 8W)E871E(&9R;VT at 9&5F875L=',-"B @(" @(" @
M(" @(&YE=W)O=RY?7V1I8W1?7ULB<F5C(ET@/2!R#0H@(" @(" @(" @("!S
M96QF+G!U=')O=RAN97=R;W<I#0H@(" @(" @(" @(" -"B @(" @(" @#0H@
M(" @9&5F(&ME>7,H<V5L9BDZ( T*(" @(" @(" B<F5T=7)N(&ME>7,N(')A
M:7-E<R!E>&-E<'1I;VX@:68 at 96UP='D@;W(@;F]T(&]P96XB#0H@(" @(" @
M(&ME>7,@/2!S96QF+F1B+FME>7,H*0T*(" @(" @("!T<GDZ#0H@(" @(" @
M(" @("!K97ES+G)E;6]V92 at B(2(I(",@<F5M;W9E('-P96-I86P@=F%L=64A
M#0H@(" @(" @(&5X8V5P="!686QU945R<F]R. at T*(" @(" @(" @(" @<&%S
M<PT*(" @(" @(" C(&-O;G9E<G0@:V5Y<R!T;R!I;G0Z#0H@(" @(" @(&EF
M('-E;&8N:6YT7VME>7,Z#0H@(" @(" @(" @("!R971U<FX@;6%P*&EN="QK
M97ES*0T*(" @(" @("!E;'-E. at T*(" @(" @(" @(" @(W!R:6YT(")S='(@
M:V5Y<R(-"B @(" @(" @(" @(')E='5R;B!K97ES#0H@(" @#0H@(" @9&5F
M(&ME>6-O=6YT*'-E;&8I. at T*(" @(" @(" B<F5T=7)N(&ME>2!C;W5N="X@
M<F5T=7)N<R P(&EF(&5M<'1Y(@T*(" @(" @("!I9B!S96QF+F1B(#T]($YO
M;F4Z#0H@(" @(" @(" @("!R971U<FX@, T*(" @(" @("!E;'-E. at T*(" @
M(" @(" @(" @<F5T=7)N(&QE;BAS96QF+F1B+FME>7,H*2DM,0T*#0H-"B @
M("!D968 at 861D<F]W*'-E;&8L*BIO<'1I;VYA;%]A<F=S*3H-"B @(" @(" @
M(F%D9')O=R!C<F5A=&5S(&$@=V]R:VEN9R!R;W<N(&%D9')O=RAF,3U<(F%<
M(BQF,CTU*2!S971S(&9I96QD<R!F,2!A;F0 at 9C(B#0H@(" @(" @("-P<FEN
M=" B861D<F]W+3XB+&]P=&EO;F%L7V%R9W,-"B @(" @(" @:68@*'-E;&8N
M;6]D92 ]/2 Q*3H@(R!I9B!E9&ET:6YG+"!P;W-T('!R979I;W5S(')O=PT*
M(" @(" @(" @(" @(" @('-E;&8N<&]S="@I(" @(" @(" -"B @(" @(" @
M;F5W<F]W(#T@<V5L9BYR;W=?8VQA<W-?;V)J*"D@(R!C<F5A=&4 at 9G)O;2!D
M969A=6QT<PT*(" @(" @("!I9B H<V5L9BYA=71O:6YC(#X](# I. at T*(" @
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M('!O<W0@;W(@8V%N8V5L+B!I9B!Y;W4 at 8V%L;"!C86YC96PL('1H92!P;W-T
M('=I;&P@;F]T(&AA<'!E;B(-"B @(" @(" @:68@*'-E;&8N<&5N9&EN9R!I
M<R!.;VYE*3H-"B @(" @(" @(" @(')E='5R;B!.;VYE#0H@(" @(" @(&EF
M("AS96QF+FUO9&4@/3T@,"DZ#0H@(" @(" @(" @("!R971U<FX at 3F]N90T*
M(" @(" @("!I9B H<V5L9BYU;FEQ=65C:&5C:RDZ#0H@(" @(" @(" @("!I
M9B H<V5L9BYH87-?:V5Y*'-E;&8N<&5N9&EN9RYK97DH*2DI. at T*(" @(" @
M(" @(" @(" @(')A:7-E($Q71&)%<G)O<BPB<&]S="!F86EL960Z(&ME>2!N
M;W0@=6YI<75E.B B*W-T<BAS96QF+G!E;F1I;F<N:V5Y*"DI#0H@(" @(" @
M(" @(" -"B @(" @(" @<V5L9BYP=71R;W<H('-E;&8N<&5N9&EN9R I#0H@
M(" @(" @('-E;&8N<&5N9&EN9R ]($YO;F4-"B @(" @(" @<V5L9BYM;V1E
M(#T@, T*(" @(" @(" -"B @("!D968 at 8V%N8V5L*'-E;&8I. at T*(" @(" @
M("!S96QF+G!E;F1I;F<@/2!.;VYE#0H@(" @(" @('-E;&8N;6]D92 ](# -
M"@T*(" @(&1E9B!D96Q?86QL<F]W<RAS96QF*3H-"B @(" @(" @(F1E;&5T
M97, at 86QL(&1A=&$@<F]W<R!I;B!A(&9I;&4L(&EN8VQU9&EN9R!T:&4 at 8V]L
M=6UN(&YA;65S(@T*(" @(" @("!S96QF+G!E;F1I;F<@/2!.;VYE#0H@(" @
M(" @('-E;&8N;6]D92 ](# -"B @(" @(" @9F]R(&D@:6X@<V5L9BYD8BYK
M97ES*"DZ#0H@(" @(" @(" @("!D96P@<V5L9BYD8EL@:2!=#0H@(" @(" @
M( T*(" @(" @(" -"B @("!D968 at 9&5L<F]W*'-E;&8L9&5L<F]W*3H-"B @
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M(" @(")L;V]K=7 @<F]W(&)Y(&ME>2!V86QU92 H87, at 82!S=')I;F<I+B!R
M971U<FYS(')O=R!O8FIE8W0B#0H@(" @(" @(&EF("AS96QF+FUO9&4@/3T@
M,2DZ(",@=V]R:VEN9R!R;W<_#0H@(" @(" @(" @("!I9B H<V5L9BYP96YD
M:6YG+FME>2 at I(#T](&ME>2DZ#0H@(" @(" @(" @(" @(" @<F5T=7)N('-E
M;&8N<&5N9&EN9R @(R!R971U<FX@=V]R:VEN9R!R;W<-"B @(" @(" @(R!G
M970@=&AE(&-S=B!S=')I;F<@9G)O;2!T:&4 at 9FEL93H-"B @(" @(" @<F]W
M/7-E;&8N9&);<W1R*&ME>2E=#0H@(" @(" @(')E='5R;B!S96QF+G)O=U]C
M;&%S<U]O8FHH<F]W*0T*(" @(" @(" @(" @#0H@(" @9&5F('!U=')O=RAS
M96QF+"!R;W<I. at T*(" @(" @(" B<'5T(&$@<F]W(&EN=&\@=&AE(&1A=&%B
M87-E+B!R;W<@;75S="!H879E(&ME>2 at I(@T*(" @(" @("!R;W=?8W-V/2!R
M;W<N8W-V7W)O=R at I#0H@(" @(" @(&ME>2 ]('-T<BAR;W<N:V5Y*"DI#0H@
M(" @(" @(&EF(&QE;BAK97DI/3TP. at T*(" @(" @(" @(" @<F%I<V4 at 3%=$
M8D5R<F]R+" B4F]W(&AA<R!N;R!K97D@=F%L=64A(@T*(" @(" @("!S96QF
M+F1B6VME>5T@/2!R;W=?8W-V#0H@(" @#0H-"FQW9&)?=&%B;&5S(#T@>WT-
M"@T*(R!E=F5R>2!A9&0@=&\@=&AE('1A8FQE(&%D9', at 82!M971H;V0-"F1E
M9B!L=V1B7W1A8FQE7V%D9"AT86)L96YA;64L9&5S8W)I<'1I;VXL<F]W7V-L
M87-S7V]B:BQF:6QE;F%M93U.;VYE+'5S95]U;FEQ=65C:&5C:STQ+'5S95]A
M=71O:6YC/3 L=7-E7VEN=&ME>7,],"DZ#0H@(" @(F%D9',@:71E;2!T;R!T
M86)L97, at 9&EC=&EO;F%R>2(-"B @("!O8FH@/2!L=V1B7W1A8FQE*'1A8FQE
M;F%M92P-"B @(" @(" @(" @(" @(" @(" @9&5S8W)I<'1I;VXL#0H@(" @
M(" @(" @(" @(" @(" @(')O=U]C;&%S<U]O8FHL#0H@(" @(" @(" @(" @
M(" @(" @(&9I;&5N86UE+ T*(" @(" @(" @(" @(" @(" @("!U<V5?=6YI
M<75E8VAE8VLL#0H@(" @(" @(" @(" @(" @(" @('5S95]A=71O:6YC+ T*
M(" @(" @(" @(" @(" @(" @("!U<V5?:6YT:V5Y<PT*(" @(" @(" @(" @
M(" @(" @("D-"B @("!L=V1B7W1A8FQE<UMT86)L96YA;65=(#T@(&]B:@T*
M(" @(')E='5R;B!O8FH-"@T*#0ID968@=&5S=%]A8W1I=F5?861D<F]W<R at I
M. at T*(" @(")C86QL960 at 8GD@=&5S="@I('1O(&%D9"!S;VUE('-A;7!L92!R
M;W=S('1O(&$@9&%T86)A<V4B#0H@(" @8V]U;G0@/2 Q,# -"B @("!G<F]U
M<',@/2!;(")',2(L(")',B(L(")',R(L(")31R(@70T*(" @('1A9T$@/2!;
M(D-/3TPB+")"4D5!2T52(BPB0558(BPB4D]/32(L(D5154E0(BPB34E30R(L
M(D%,5"(L(DY%6%0B70T*(" @('1A9T(@/2!;(E1%35 B+")44DE0(BPB4U1!
M5$4B+" B5$535"(L(")&055,5"(L(")35TE40T at B(%T-"B @("!C;V1E(#T@
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M90T*(" @('0@/2!;70T*(" @('0N87!P96YD*'1I;64N=&EM92 at I*0T*(" @
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M:&5A9&5R#0H@(" @(" @(" @("!D968 at 7U]I;FET7U\H("!S96QF+" @<F]W
M/4YO;F4@*3H-"B @(" @(" @(" @(" @("!S96QF+G-E='5P*')O=RD@(R!I
M;FET:6%L:7IE('1O(&1E9F%U;'0@=F%L=65S#0H@(" @<')I;G0@(DQI9VAT
M=V5I9VAT($1A=&%B87-E("T at 0F%S:6, at 5&5S=',B#0H@(" @(W!R:6YT(")M
M87 @=&5S=" Q.B(L7V-O;%]M87 H(%L@(F$B+")B(BPB8R(L(")D(B!=+"!;
M(")C(BP@(F$B+" B8B(@72 I#0H@(" @(W!R:6YT(")M87 @=&5S=" R.B(L
M7V-O;%]M87 H(%L@(F$B+")X(BPB>2(L(")Z(B!=+"!;(")P(BP@(F$B("!=
M("D-"B @(" C<')I;G0@(FUA<"!T97-T(#,Z(BQ?8V]L7VUA<"@@6R B02(L
M(E at B+" B4B(L(")1(B!=+"!;(")!(BPB6"(L(")2(BP@(E$B(%T@*0T*(" @
M('!R:6YT(")T86)L92!A8W1I=F4Z(@T*(" @(&%C=&EV92 @(#T@;'=D8E]T
M86)L95]A9&0H(")A8W1I=F4B+" @(")A8W1I=F4 at 86QA<FUS(BP at 86QA<FU?
M979E;G1?<F5C+ T*(" @(" @(" @(" @(" @(" @(" @(" @(" @(" @("!U
M<V5?=6YI<75E8VAE8VL],2QU<V5?875T;VEN8STQ+'5S95]I;G1K97ES/3$I
M#0H-"B @("!T+F%P<&5N9"AT:6UE+G1I;64H*2D-"B @("!P<FEN=" B=&%B
M;&4@:&ES=&]R>3HB#0H@(" @(VAI<W1O<GD@(#T@;'=D8E]T86)L95]A9&0H
M(")H:7-T;W)Y(BP@(")A;&%R;2]E=F5N="!H:7-T;W)Y(BP at 86QA<FU?979E
M;G1?<F5C("D-"B @("!T97-T7V%C=&EV95]A9&1R;W=S*"D[#0H-"B @("!T
M+F%P<&5N9"AT:6UE+G1I;64H*2D-"@T*(" @(&L@/2!A8W1I=F4N:V5Y<R at I
M#0H@(" @:RYS;W)T*"D-"B @("!N/6QE;BAK*0T*(" @('!R:6YT(")A8W1I
M=F4N:V5Y<STB+'-T<BAK6S Z,3!=*5LZ+3%=+"(L("XN+BX at +" B+'-T<BAK
M6VXM,3 Z;ETI6S$Z70T*(" @('!R:6YT(")A8W1I=F4N:V5Y8V]U;G0](BQA
M8W1I=F4N:V5Y8V]U;G0H*0T*(" @(&1I<B at I#0H@(" @<')I;G0@(F%C=&EV
M92!E>'!O<G0Z(@T*(" @(&%C=&EV92YC<W9?97AP;W)T*")A8W1I=F4N8W-V
M(BD-"B @("!T+F%P<&5N9"AT:6UE+G1I;64H*2D-"@T*(" @('!R:6YT(")A
M8W1I=F4 at 8VQO<V4Z(@T*(" @(&%C=&EV92YC;&]S92 at I#0H@(" @<')I;G0@
M(F1O;F4N(@T*(" @('0N87!P96YD*'1I;64N=&EM92 at I*0T*(" @('!R:6YT
M(")T:6UE<SHB#0H@(" @<R ]('1;,%T-"B @("!F;W(@<B!I;B!T6S$Z73H-
M"B @(" @(" @<')I;G0@<BUS#0H@(" @(" @(',]<@T*#0H-"@T*#0II9B!?
J7VYA;65?7R ]/2 G7U]M86EN7U\G. at T*(" @('1E<W0H*0T*(" @( T*
`
end



From ckotso at cs.ntua.gr  Wed Apr 26 07:39:00 2000
From: ckotso at cs.ntua.gr (Constantinos A. Kotsokalis)
Date: 26 Apr 2000 11:39:00 GMT
Subject: ungreedy regular expressions
Message-ID: <slrn8gdlaj.k0s.ckotso@theseas.softlab.ece.ntua.gr>

Disclaimer: I am not sure whether this has been brought up before,
I am not reading the newsgroup very often (I could not find any reference
though, after a subject search).

I recently added two lines of code to the python distribution, in order
to support ungreedy regular expressions (which are supported by pcre
but I could find no way to use them in python code). If there is some
way already that I am not aware of, please e-mail me :-) If you are
interested in having it available, edit Modules/pcremodule.c and add
the line:

insint(d, "UNGREEDY", PCRE_UNGREEDY);

after line 681. Then compile and install python, and right after
edit <python-prefix>/lib/python1.5/re.py, and add the line:

U = UNGREEDY

after line 18.
That's all, you can now use re.UNGREEDY (or re.U) just like re.MULTILINE
etc.

  Regards,
	Costas

-- 
fsck your brains


From mal at lemburg.com  Mon Apr  3 07:48:05 2000
From: mal at lemburg.com (M.-A. Lemburg)
Date: Mon, 03 Apr 2000 13:48:05 +0200
Subject: [Python-Dev] Re: New Features in Python 1.6
References: <200004011740.MAA04675@eric.cnri.reston.va.us> <20000403131720.A10313@sz-sb.de>
Message-ID: <38E884F5.8F2FB271@lemburg.com>

Andreas Jung wrote:
> 
> On Sat, Apr 01, 2000 at 12:00:00PM -0500, Guido van Rossum wrote:

The above line has all the answers ;-) ...

> > Python strings can now be stored as Unicode strings.  To make it easier
> > to type Unicode strings, the single-quote character defaults to creating
> > a Unicode string, while the double-quote character defaults to ASCII
> > strings.  If you need to create a Unicode string with double quotes,
> > just preface it with the letter "u"; likewise, an ASCII string can be
> > created by prefacing single quotes with the letter "a".  For example:
> >
> >    foo = 'hello'                # Unicode
> >    foo = "hello"                # ASCII
> 
> Is single-quoting for creating unicode clever ? I think there might be a problem
> with old code when the operations on unicode strings are not 100% compatible to
> the standard string operations. I don't know if this is a real problem - it's
> just a point for discussion.

-- 
Marc-Andre Lemburg
______________________________________________________________________
Business:                                      http://www.lemburg.com/
Python Pages:                           http://www.lemburg.com/python/



From gmcm at hypernet.com  Wed Apr 26 17:17:57 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Wed, 26 Apr 2000 17:17:57 -0400
Subject: global variables or inheritence
In-Reply-To: <28p4s8oipzh.fsf@lada.kom.auc.dk>
Message-ID: <1255338618-62939035@hypernet.com>

Thomas Rasmussen  wrote:
> >>>>> "Jeff" == Jeff Massung <jmassung at magpiesystems.com> writes:
> 
>   Jeff> Sorry if I get this wrong... but I thought you'd just
>   import Jeff> your other file.
> 
>   Jeff> Example:
> 
>   Jeff> # File1.py myVar = 2
> 
> 
>   Jeff> # File2.py import File1 print File1.myVar #should print 2
> 
> Well that was also what I thought, but the following program wont
> execute: testprog1.py: import os, sys, string, testprog2 def
> main():
>     print 'hello'
>     myvar = 2
>     testprog2.second()
> if __name__ == '__main__':
>     main()
> 
> testprog2.py:
> import os, sys, string, testprog1
> def second():
>     print testprog1.myvar
> if __name__ == '__main__':
>     print 'nope'

Circular imports are fine in Python (well, there are some 
gotchas). Your problem here is that "import testprog1" is 
creating a new instance of testprog1, which is not the same 
one you're executing (which is known as "__main__").

Add the line "global myvar" inside testprog1.main. Now 
execute "import testprog1" from the interactive prompt (or 
another script) and testprog2 will work as expected.

If you really want to get to variables in the executing script, 
you need to "import __main__" in the submodule, and 
schedule yourself for a severe spanking.

> and then I execute python testprog1.py and get the following
> error:
> 
> 22:03 simpsons at lada% python testprog1.py                         
>      ~/python hejsa Traceback (innermost last):
>   File "testprog1.py", line 7, in ?
>     main()
>   File "testprog1.py", line 5, in main
>     testprog2.second()
>   File "testprog2.py", line 3, in second
>     print testprog1.myvar
> AttributeError: myvar
> 
> :-(
> 
> Thomas
> 
> -- 
> KOM Network student helper
> 
> "To alcohol! The cause of - and solution to - all of life's
> problems!" -- Homer Simpson --
> http://www.python.org/mailman/listinfo/python-list



- Gordon



From neelk at brick.cswv.com  Tue Apr 25 18:14:34 2000
From: neelk at brick.cswv.com (Neel Krishnaswami)
Date: 25 Apr 2000 22:14:34 GMT
Subject: new linereading standard?
References: <3904D21B.333DD930@visionart.com>
Message-ID: <slrn8gc8a2.6g.neelk@brick.cswv.com>

Pete Shinners <pete at visionart.com> wrote:
> 
> i'm sure someone has posted a class like the before, but i'm hoping it
> can one day make it in as a python standard someday. (6.0, there's still
> time!)
> 
> class filelines():
>     def __init__(self, file):
>         self.file = file
>     def __getitem__(self, index):
>         line = self.file.readline()
>         if not line: raise IndexError
>         return line
> 
> anyways, i now use this code ALWAYS. unfortunately, i haven't found a
> good way to make this part of my standard environment while working on
> many different machines.

I use something like this whenever the file is too big for
file.readlines(), too.

A couple of notes:

o If you need this to be part of your standard environment, you should
  (as others have mentioned) look at the 'fileinput' standard module.

o If the wrapper approach is too slow, then you might want to look at 
  the QIO module at

  http://members.xoom.com/meowing/python/

  This is -blazingly- fast; it lets you step through files about 2-3
  times faster than using standard file objects, even without wrappers,
  and lets you use friendly syntax. This is especially helpful if you 
  have great big files (that are too big for file.readlines()). 


Neel


From neilh at scintilla.org  Sun Apr 30 21:45:57 2000
From: neilh at scintilla.org (Neil Hodgson)
Date: Mon, 01 May 2000 01:45:57 GMT
Subject: [Python-Dev] Re: Python 1.6a2 Unicode bug (was Re: comparing strings and ints)
References: <l03102805b52ca7830b18@[193.78.237.154]> <39072F0C.5214E339@lemburg.com> <BlHN4.1719$wYl.195075584@newsb.telia.net> <oqvh14fikb.fsf@titan.progiciels-bpi.ca> <45919986@data.franken.de> <048501bfb0b4$c793d760$4add3fcb@neil> <45919991@data.franken.de>
Message-ID: <pl5P4.8409$v85.54219@news-server.bigpond.net.au>

> Of course I meant UCS-2, but I'm confused now. Isn't UCS-2 and UTF-16 more
> or less the same?

   Well, depends on how far you stretch 'more or less'. UTF-16 has room to
encode about 900,000 characters by using two 16 bit elements. UCS-2
currently has room for about 57,000 characters with 39,000 currently
assigned. UTF-16 is mostly of concern for academic use with extinct
languages although it could also be used for Han deunification.

   For an editor to support UTF-16, I would expect it to treat 32 bit UTF-16
characters as indivisible, at least not putting the caret between the two
halves even if it did not have a sensible glyph to display.

   Neil




From gregor at hoffleit.de  Fri Apr 21 17:18:56 2000
From: gregor at hoffleit.de (Gregor Hoffleit)
Date: Fri, 21 Apr 2000 23:18:56 +0200
Subject: How2 use curses for data input
In-Reply-To: <38FF8F9C.9D533E74@sprint.ca>; from milesthompson@sprint.ca on Thu, Apr 20, 2000 at 08:15:40PM -0300
References: <38FF6DE2.5751423A@sprint.ca> <slrn8fusjd.99a.grant@grante.comtrol.com> <38FF8F9C.9D533E74@sprint.ca>
Message-ID: <20000421231856.A22756@53b.hoffleit.de>

On Thu, Apr 20, 2000 at 08:15:40PM -0300, Miles Thompson wrote:
> I located it on the RH system here in the office at home. The target machine
> is Debian, and I think I can install RPM on it so that if need be, I can get
> the RPM. Debian doesn't seem to have a snack package.

Hmm, I never heard about snack, but Debian potato has a python-newt package,
and this package includes the snack module:

  freefly;23> dlocate -S snack
  python-newt: /usr/lib/python1.5/site-packages/_snackmodule.so
  python-newt: /usr/lib/python1.5/site-packages/snack.py

python-newt was not included in Debian 2.1, though, but, if you're lucky,
you may be able to install the python-newt package from potato (probably
together with the libnewt0 from potato) on a Debian 2.1 system, too.

    Gregor
    
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 240 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20000421/a54c8541/attachment.sig>

From rburnham at cri-inc.com  Tue Apr 18 18:26:16 2000
From: rburnham at cri-inc.com (Roger Burnham)
Date: Tue, 18 Apr 2000 22:26:16 GMT
Subject: PythonWin and Dialogs in DLLs
References: <38FCD4FB.6E07A632@callware.com>
Message-ID: <38fcde9b.12797461@news-server.socal.rr.com>

On Tue, 18 Apr 2000 15:34:51 -0600, Ivan Van Laningham
<ivanlan at callware.com> wrote:

>Hi All--
>OK, I followed the instructions in previous messages to this list to
>build a resource-only DLL from which I wish to import custom dialogs. 
>The build works, but using "import migrate1", where the dll is named
>"migrate1.dll" gives me the error message
>
>	dynamic module does not define init function (initmigrate1)
>
>Digging around for that error message seems to indicate that it's
>thinking I want an extension module with code in it, when all I want to
>do is just have *resources only*.
>
>What am I missing?  What's staring me in the face if only I would open
>my eyes?
>

Ivan,

In your pywin.mfc.dialog.Dialog derived class, call the parent init
with the dll:

           Dialog.__init__(self, dllid='pathToYourDll')
Roger Burnham
rburnham at cri-inc.com


From bparsia at email.unc.edu  Fri Apr  7 10:09:10 2000
From: bparsia at email.unc.edu (Bijan Parsia)
Date: Fri, 7 Apr 2000 10:09:10 -0400
Subject: Teaching the art of programming, in python
References: <20000405053920.20579.qmail@web2102.mail.yahoo.com> <8cfe42$o02$1@nnrp1.deja.com> <dpJG4.3744$HG1.106895@nnrp1.uunet.ca> <1e8mov8.1yh8ryv1kp6z0lN%bparsia@email.unc.edu> <06042000.4@sanctum.jae.ddns.org>
Message-ID: <1e8pq1t.3vp46r1cpxokhN%bparsia@email.unc.edu>

Juergen A. Erhard <jae at ilk.de> wrote:

> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
> 
> >>>>> "Bijan" == Bijan Parsia <bparsia at email.unc.edu> writes:
> 
>     Bijan> Warren Postma <embed at geocities.com> wrote:
> 
>     Bijan> [snip]
> [...]
>     >> Remember the famous 'Smalltalk-80 - The Language and Implementation'
>     >> purple-book?
>     Bijan> [snip]
> 
>     Bijan> Pedantic correction: The Language and Implementation version is the
>     Bijan> *blue* book. [...]
> 
> Pedantry time: I remember it as
> 
>   Smalltalk-80: The Language and it's Implementation
> 
> (though I haven't read it... yet)
> 

Pedantry time: I didn't actually refer to the title, though I didn't
correct Warren either :)

It's *still* a great book.

Ah, and I have the reference to the on line Implementation chapters, if
folks want to whet their appetite:
        http://users.ipa.net/~dwighth/

A couple of other cool classic papers are there to, including a
description of Smalltalk-76 which will soon have an image running on a
Squeak VM near you!

-- 
Bijan Parsia
http://monkeyfist.com/
...among many things.


From jlj at cfdrc.com  Mon Apr 24 11:28:40 2000
From: jlj at cfdrc.com (Lyle Johnson)
Date: Mon, 24 Apr 2000 10:28:40 -0500
Subject: can't find SWIG for Python
References: <38FF5A56.F76794D8@sec.noaa.gov>
Message-ID: <39046828.DC71820E@cfdrc.com>

http://swig.sourceforge.net

j vickroy wrote:
> 
> Hello all,
> 
> I have been unable to locate SWIG for Python.  My references list
> www.swig.org but the server can not be found.
> 
> Could someone point me to the site or provide some other Win 32 options
> for easing the process of extending Python to use existing C libraries.
> 
> Thanks for your time.


From johnson2 at hiwaay.net  Sun Apr  2 12:37:49 2000
From: johnson2 at hiwaay.net (Lyle Johnson)
Date: Sun, 2 Apr 2000 11:37:49 -0500
Subject: New Features in Python 1.6
References: <200004011740.MAA04675@eric.cnri.reston.va.us>
Message-ID: <seetkrhr9119@corp.supernews.com>

OK, I'm laughing and breathing a sigh of relief now. But I must admit I fell
for it too.




From cut_me_out at hotmail.com  Thu Apr 20 12:40:06 2000
From: cut_me_out at hotmail.com (Alex)
Date: 20 Apr 2000 12:40:06 -0400
Subject: Creating Classes Dynamically?
References: <eviltofu-783F16.12334920042000@news.earthlink.net>
Message-ID: <etdu2gw915l.fsf@m56-129-11.mit.edu>

> Is there a way to create an object of a class from a string?

Can you give an example of the sort of thing you want to do?

Alex.


From sjoerd at oratrix.nl  Mon Apr 17 06:54:45 2000
From: sjoerd at oratrix.nl (Sjoerd Mullender)
Date: Mon, 17 Apr 2000 12:54:45 +0200
Subject: Looking for IRIX Python programmers for help
In-Reply-To: Your message of Fri, 14 Apr 2000 14:24:32 -0400.
             <38F76260.16603101@heathers.stdio.com> 
References: <38F76260.16603101@heathers.stdio.com> 
Message-ID: <20000417105446.0473E301CF9@bireme.oratrix.nl>

The al (Audio Library) module uses the -laudio library which is in the
dmedia_eoe package.  The cd module uses -lcdaudio which is in the
dmedia_dev package.  The cl (Compression Library) module uses the -lcl
and -lawareaudio libraries which are also in dmedia_eoe.  In order to
use these modules you need to have the abovbe mentioned packages
installed.  You can see whether you have them using the versions
command.

I don't know how relevant the sv (Starter Video) module is.  If I
remeber right, it was for video on an SGI Indigo.

On Fri, Apr 14 2000 "Eric H." wrote:

> Hi,
> 
> While building Python 1.5.2 for IRIX 6.5, I was editing Modules/Setup
> and discovered there is a section that mentions that you need a
> "multimedia development option(I think)" to build the "al" module and a
> couple others. Can someone clarify this? Is this really needed and if
> so, where do I get it? Thanx for your wonderful response and L8a
> 
> 
> P.S. A response via email would be much welcome since I don't get to
> check newsgroups every day. 
> 
> 
> -- 
> Parts of this message, including the email address, ? 2000-Eric Haddix. 
> Use implies agreement to the terms and conditions made available on my 
> homepage. Keep the Internet free, refuse to support the providers that
> copyright YOUR email and messages. All rights reserved. U.C.C. 1-207
> 
> visit my homepage @ http://www.bounce.to/EyesCremeProductions
> 
> 

-- Sjoerd Mullender <sjoerd.mullender at oratrix.com>



From effbot at telia.com  Sun Apr  9 12:35:09 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Sun, 09 Apr 2000 16:35:09 GMT
Subject: Python paradigms
References: <Pine.LNX.4.21.0004091506110.13883-100000@fep132.fep.ru>
Message-ID: <1j2I4.252$rc9.186596864@newsb.telia.net>

Oleg Broytmann wrote:
> On Sun, 9 Apr 2000, Robert W. Cunningham wrote:
> >  blah = (((test and [true]) or [false])[0])             # blah = (test ?
> > true : false)
> >
> > Works for me!  Only if I have a good reason to suspect a problem in this
> > particular line of code will I have to look past the comment.
>
>    There is no problem in this particular line, but you'll have plenty of
> them if you'd try to use this paradigm:
>
>    blah = ((foo() and [poo()]) or [boo()])[0]
>
>    In C, "foo() ? poo() : boo()" will not call boo() if foo() returns
true;
> but in Python, you'll have a side-effect of calling boo() even if foo()
> returns true - what is wrong, usually.

did you try running the code?

python's "and" and "or" operators doesn't evaluate the
second argument unless they really have to...

</F>




From kc5tja at garnet.armored.net  Tue Apr 25 13:26:18 2000
From: kc5tja at garnet.armored.net (Samuel A. Falvo II)
Date: 25 Apr 2000 17:26:18 GMT
Subject: Sourceforge and Python
References: <oq8zy1zqto.fsf@titan.progiciels-bpi.ca> <007b01bfaf14$f33114b0$3cdd3fcb@neil>
Message-ID: <slrn8gcgvp.u19.kc5tja@garnet.armored.net>

In article <007b01bfaf14$f33114b0$3cdd3fcb at neil>, Neil Hodgson wrote:
>   Aspects which are good at SourceForge include having MailMan to set up
>mailing lists with archives and no advertising. Your site may be found by
>people searching or browsing SourceForge. CVS access is normally good
>although there are problems on Windows with SourceForge's choice of SSH
>version _1_ for write access.

How the *hell* did you get your CVS repo to work on SourceForge?  :)

Whenever I try to upload to the repo I've set up for my own project, it
tells me I don't have write permissions after processing about four or five
files (I forget the exact error phrase, but that's basically what it
reports).  Now, I've used CVS before -- I have a repo set up on my box at
home.  What am I missing with respect to CVS with SourceForge?

-- 
KC5TJA/6, DM13, QRP-L #1447
Samuel A. Falvo II
Oceanside, CA


From mhammond at skippinet.com.au  Tue Apr 11 20:16:14 2000
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Wed, 12 Apr 2000 00:16:14 GMT
Subject: Saving a copy of a remote NT event log
References: <2B1262E83448D211AE4B00A0C9D61B03BA7295@MSGEURO1>
Message-ID: <ifPI4.1678$UB.11920@news-server.bigpond.net.au>

Im fairly sure the win32evtlog module will do this for you.  Check the
docs.

Mark.

"Pieter Claerhout" <PClaerhout at CREO.BE> wrote in message
news:2B1262E83448D211AE4B00A0C9D61B03BA7295 at MSGEURO1...
> Hello,
>
> does anyone know if it's possible to save a copy of the NT-event log
> from a remote machine using Python? Normally, you have to do a
> save as from that machine, and then copy over the logfile, but I want
> to automate this using Python. Any ideas?
>
> Kind regards,
>
>
> Pieter
>
> [ Pieter Claerhout :: CreoScitex Response Centre :: Applications
support ]
> [ Pieter_Claerhout at creoscitex.com    ::    http://www.creoscitex.com/ ]
> [ CreoScitex Europe  ::   Excelsiorlaan 21   ::  1930 Zaventem  ::
Belgium ]
>
>




From abrahams at mediaone.net  Sun Apr 30 20:02:44 2000
From: abrahams at mediaone.net (Dave Abrahams)
Date: Mon, 01 May 2000 00:02:44 GMT
Subject: Do I always have to write "self." ?
References: <270420001706386648%pecora@anvil.nrl.navy.mil> <slrn8ghc6t.f01.scarblac-spamtrap@flits104-37.flits.rug.nl> <8ecsgj$6lp$1@slb0.atl.mindspring.net> <m3bt2tonyg.fsf@atrus.jesus.cam.ac.uk> <300420001008012011%pecora@anvil.nrl.navy.mil><20000430170952.A502@datapro.co.za> <B531EC9B.57C1%abrahams@mediaone.net><slrn8gp7ei.ilf.kc5tja@garnet.armored.net>
Message-ID: <B53241E5.580F%abrahams@mediaone.net>

in article slrn8gp7ei.ilf.kc5tja at garnet.armored.net, Samuel A. Falvo II at
kc5tja at garnet.armored.net wrote on 4/30/00 9:03 AM:

> If it's a dict, or if it's a UserDict, or if it's anything else that exports
> the dict interface, then it's perfectly valid to name a parameter
> dictWhatever.  I fail to see what's wrong with this concept.

As Simonyi, the author of the Hungarian notation concept writes, "the basic
idea is to name all quantities by their types". If you are naming quantities
by their interface, you are taking a step away from Hungarian. Do as I do,
naming quantities by their *role*, and you have taken a long leap away from
Hungarian.

I can see how it might help to name a variable by its interface, especially
in Python if you don't write comments. But: do you write listDogs, tupleDogs
or sequenceDogs when you're being passed a sequence of dogs? Don't the
prefix nouns start to sound a bit like adjectives? Or do you come up with
some tongue-twisting acronym for each of these?

Suppose you're writing UserDict. What do you name your function arguments
which represent keys and values? Not too many requirements on the value
type, are there? Is Hungarian applied only selectively?

and-is-my-name-person-dave-or-programmer-dave-ly y'rs,
-pDave



From mhammond at skippinet.com.au  Wed Apr 19 22:28:33 2000
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Thu, 20 Apr 2000 02:28:33 GMT
Subject: Browse dialog in PythonWin
References: <38FE32AB.520C1736@callware.com> <38fe3297.36998571@news-server.socal.rr.com> <38FE3998.1ED7FB24@callware.com>
Message-ID: <lXtL4.32$v85.230@news-server.bigpond.net.au>

"Ivan Van Laningham" <ivanlan at callware.com> wrote in message
news:38FE3998.1ED7FB24 at callware.com...
> Hi All, Roger--
>
> Roger Burnham wrote:
> >
> [snippery]
> >
> > Here's a dialog I adapted from something I found in the PythonWin
> > distribution.
> >
>
> [more ruthless snippery]
>
> That will do *very* nicely!  Thank you!

An alternative to get the real Windows browser dialog:

>>> import win32com.shell.shell
>>> win32com.shell.shell.SHBrowseForFolder(0, None, "Select it", 0, None,
0)
('\024\000\037P\340O\320 \352:i\020\242\330\010\000+00\235\024\000.w
\354!\352:i\020\242\335\010\000+00\235\000\000', 'Control Panel', 21)
>>>

I selected "Control Panel" - You will need to read the docs to see how to
use that funky string (the PIDL)...

No news on starship yet - if its not available for me to re-upload in 24
hours I will do it somewhere else...

Mark.





From andy at reportlab.com  Fri Apr  7 10:22:10 2000
From: andy at reportlab.com (Andy Robinson)
Date: Fri, 07 Apr 2000 14:22:10 GMT
Subject: A Mountain of Perl Books + Python Advocacy
References: <000f01bf9eca$2eec8c60$3e2d153f@tim> <38ECF4E6.B85C62D5@sightreader.com>
Message-ID: <38eeeee0.15607732@news.demon.co.uk>

Ken Seehof <kens at sightreader.com> wrote:

>Indeed.  I've noticed that while just about everyone who I've known who
>has tried python ends up liking it alot, there aren't many fanatics.  I define
>a fanatic as someone who is absolutely certain that their language is by far
>the best language for every possible application, and that other languages
>don't even deserve to exist.  Python programmers tend to acknowledge
>the positive qualities of other languages.
>
which is a nice little argument for Python having the smartest
community of programmers.  Perhaps that's as important as the
language.

Andy Robinson


From moshez at math.huji.ac.il  Fri Apr 14 13:27:29 2000
From: moshez at math.huji.ac.il (Moshe Zadka)
Date: Fri, 14 Apr 2000 19:27:29 +0200 (IST)
Subject: Complaint Dept: My shoes!
In-Reply-To: <8d4s4d$mvr$1@nnrp1.deja.com>
Message-ID: <Pine.GSO.4.10.10004141925330.7167-100000@sundial>

On Thu, 13 Apr 2000 musingattheruins at my-deja.com wrote:

> If shoe is None then len(shoe) raises an exception.  Would be nice if
> len(None) was 0 and NOT raise an exeption.

Huh? Why not have len(None) 4? Or -1?
If you want an empty string, use an empty string. The good thing
about None is that it supports so few operations, functions not expecting
it would break right away, instead of having a hidden bugs.

in-the-face-of-ambiguity-refuse-to-guess-ly y'rs, Z.




From maybe at lucky.com  Tue Apr 11 13:39:45 2000
From: maybe at lucky.com (Soundwave)
Date: Tue, 11 Apr 2000 19:39:45 +0200
Subject: AF_INET error
References: <8cvero$g8k$1@porthos.nl.uu.net> <cHHI4.1202$Za1.20911@newsc.telia.net>
Message-ID: <8cvnve$l40$1@porthos.nl.uu.net>

OK, sorry

here we go

On www.python.org is a HOWTO guide for creating sockets, i'll give an
example from what is said in the guide

This is the first example

import socket
#create an INET, STREAMing socket
s = socket.socket(
    socket.AF_INET, socket.SOCK_STREAM)
#now connect to the web server on port 80
# - the normal http port
s.connect("www.mcmillan-inc.com", 80)

when I run this code, I get the following output

Traceback (innermost last):
  File "C:\PROGRA~1\PYTHON\TOOLS\IDLE\ScriptBinding.py", line 131, in
run_module_event
    execfile(filename, mod.__dict__)
  File "C:/Program Files/Python/socket.py", line 3, in ?
    s = socket.socket(
AttributeError: AF_INET


maybe I look (maybe i even are ?) a bit stupid, but I realy don't know what
it means, or what I am doeing wrong.

Thanx by the way

Fredrik Lundh <effbot at telia.com> schreef in berichtnieuws
cHHI4.1202$Za1.20911 at newsc.telia.net...
> Soundwave <maybe at lucky.com> wrote:
> > I want to learne some things about socket programming, so I started the
> > socet programming tutorial on www.python.org
> >
> > but when I copy te code from some sample programm's and run it, it
doesn't
> > work, it don't recognize the AF_INET thing, is this because I am using
> > Windows ??
> >
> > can somebody please explain this to me ?
> >
> > Or what am I doeing wrong ?
>
> well, since you forgot to tell us what example you tried to
> run, how you typed it in, and what error you got, I can only
> guess.
>
> -- make sure to import the socket module
>
> -- make sure to use the "socket." prefix, so Python
>    knows where to look for the AF_INET "thing" (it's
>    a variable).
>
> </F>
>
>




From effbot at telia.com  Tue Apr 18 12:06:39 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Tue, 18 Apr 2000 16:06:39 GMT
Subject: Project idea - python
References: <BxGK4.811$9B.103178@news.bc.tac.net> <38FC0E72.9AB025DB@webwitches.com> <Pine.GSO.4.21.0004181113110.22094-100000@sun21epsl.wam.umd.edu>
Message-ID: <jK%K4.465$wYl.187703296@newsb.telia.net>

Bob Schmertz <schmertz at wam.umd.edu> wrote:
> SPARK - sounds like a Python implementation of Yacc/Bison, though from
> what I can tell it doesn't handle complex grammars such as LR
> (http://www.csr.UVic.CA/~aycock/python/)

umm.  the doc describes the parser algorithm used in SPARK as "one
of the most general algorithms known; it can parse all context-free
grammars whereas the more popular LL and LR techniques cannot."

maybe some parser guru can sort this out?

</F>




From yawyi at gmx.de  Wed Apr 26 06:42:13 2000
From: yawyi at gmx.de (Johannes Stezenbach)
Date: Wed, 26 Apr 2000 12:42:13 +0200
Subject: Sourceforge and Python
References: <oq8zy1zqto.fsf@titan.progiciels-bpi.ca>
Message-ID: <3906c7e8$0$29200@personalnews.de.uu.net>

Fran?ois Pinard <pinard at iro.umontreal.ca> wrote:
> I would like to bring `recode'
>a bit closer to Python

There already exists a GNU recode wrapper module for python:

http://www.vex.net/parnassus/apyllo.py?find=recode
http://www.suxers.de/python/pyrecode.htm

Johannes



From karl at ulbrich.org  Thu Apr 27 12:20:38 2000
From: karl at ulbrich.org (Karl G. Ulbrich)
Date: Thu, 27 Apr 2000 12:20:38 -0400
Subject: PIL compile/make bombs out at step #7 -- help?
In-Reply-To: <20000426182200.B27107@exeter.exeter.org>; from karl@ulbrich.org on Wed, Apr 26, 2000 at 06:22:00PM -0400
References: <20000426135724.A26663@exeter.exeter.org> <asIN4.4302$Za1.64185@newsc.telia.net> <20000426182200.B27107@exeter.exeter.org>
Message-ID: <20000427122038.B434@exeter.exeter.org>

Found a solution; don't know if it's the right one:

> Here's my setup file -- I disabled the TK part as I neglected
> to install that package with Linux, but I have the python .h
> files, jpeg and libz stuff.
> 
> # Setup file for the Imaging module
> # =================================
[...]
> # *** Tcl/Tk support and library location
> #        Tk/tkImaging.c -DWITH_TKINTER -I/usr/local/include -L/usr/local/lib -ltcl8.0 -ltk8.0
>

Instead of commenting OUT that last line to disable
Tcl/TK support, replace it with a line that tells 
make nothing.  This seems to get around the *noobjects* error.
I used this, and it worked:

        -I/ 

No idea why; a friends' suggestion.

Karl 



From aycock at csc.uvic.ca  Fri Apr 14 02:22:12 2000
From: aycock at csc.uvic.ca (John Aycock)
Date: 14 Apr 2000 06:22:12 GMT
Subject: [ANNOUNCE] SPARK 0.6
Message-ID: <8d6dek$4lv2$1@uvaix7e1.comp.UVic.CA>

This is the latest release of SPARK: the Scanning, Parsing, and Rewriting Kit.  
(It was formerly known as "the little language framework.")                     
                                                                                
SPARK features a collection of classes which can be used to implement           
little languages.  Mixing and matching components is possible; you don't        
have to use the entire collection of classes if you don't want to.              
                                                                                
New in release 0.6 (see CHANGES file for a complete list):

	-  Bug fixes.
	-  Parser speed improvements.  Depends on the grammar and input
		file, but I've seen between 8%-35% improvements on some
		of my standard tests.  (A minor code change is required;
		see the release notes.)

Old in release 0.6:

	-  Generic scanner class.                                               
        -  Generic parser class, which handles any context-free grammar.        
        -  Class to automatically construct parse trees or abstract             
                syntax trees.                                                   
        -  Class for traversing trees, useful for semantic checking and         
                code generation.                                                
        -  Class for tree pattern matching, good for code generation or         
                interpreter implementation.                                     
        -  100% pure Python.                                                    
                                                                                
Available at http://csr.uvic.ca/~aycock/python.                                 
                                                                                
John

From: "T Zimmerman" <tzman at bhmi.com>
Newsgroups: comp.lang.python
Subject: Help
Date: Fri, 7 Apr 2000 08:01:43 -0500
X-Priority: 3
X-MSMail-Priority: Normal
X-Newsreader: Microsoft Outlook Express 5.00.2314.1300
X-MimeOLE: Produced By Microsoft MimeOLE V5.00.2314.1300
Lines: 28
Message-ID: <38eddc38$0$17341 at wodc7nh6.news.uu.net>
NNTP-Posting-Host: 208.248.174.66
X-Trace: reader2.news.uu.net 955112504 17341 208.248.174.66
Path: news!uunet!ffx.uu.net!spool0.news.uu.net!reader2.news.uu.net!not-for-mail
Xref: news comp.lang.python:90831
To: python-list at python.org
Sender: python-list-admin at python.org
Errors-To: python-list-admin at python.org
X-BeenThere: python-list at python.org
X-Mailman-Version: 2.0beta3
Precedence: bulk
List-Id: General discussion list for the Python programming language <python-list.python.org>

I have left several posts concerning Tkinter Meta keys, and yet to get a
response from anyone. If this doesnt work for anyone else than it is a bug.
Otherwise there is something wrong on my end. The following script should
allow the use of 'alt-f' to access the 'File' menu. This does not work on
any of the NT boxes in my office. If this does or does not work for someone
please let me know.

Thanks,
T Zimmerman
tzman at bhmi.com


==================================================================
import Tkinter

root = Tkinter.Tk()

button = Tkinter.Menubutton(text = 'File', underline = 0)
button.pack()
menu = Tkinter.Menu(button)
button.configure(menu = menu)
menu.add_command(label = 'Hello World', underline = 0)

root.mainloop()
==================================================================




Return-Path: <cgw at buffalo.fnal.gov>
Delivered-To: python-list at python.org
Received: from buffalo.fnal.gov (buffalo.fnal.gov [131.225.84.156])
	by dinsdale.python.org (Postfix) with ESMTP id DE44E1CDCF
	for <python-list at python.org>; Tue,  4 Apr 2000 20:11:29 -0400 (EDT)
Received: (from cgw at localhost)
	by buffalo.fnal.gov (8.8.7/8.8.7) id TAA00970
	for python-list at python.org; Tue, 4 Apr 2000 19:13:54 -0500
Date: Tue, 4 Apr 2000 19:13:54 -0500
Resent-From: Charles G Waldman <cgw at fnal.gov>
Message-Id: <200004050013.TAA00970 at buffalo.fnal.gov>
Resent-Message-ID: <14570.34114.546452.113478 at buffalo.fnal.gov>
Resent-Date: Tue, 4 Apr 2000 19:13:54 -0500 (CDT)
Resent-To: python-list at python.org
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
From: Charles G Waldman <cgw at fnal.gov>
To: python-users at python.org
Cc: Dennis E Hamilton <infonuovo at email.com>
Subject: Re: Why should I swich to Python? - Infinity of Primes
X-Mailer: VM 6.72 under 21.1 (patch 9) "Canyonlands" XEmacs Lucid
X-Face: %OO~XPb`a}(s2it:MIMa&Ig&fbz)+h$L,2js]u<DS^QQ0UGR'_Wps8,#>XlS*7R<nj`T(nz
 @}>#!#e{6W^.z~0blXY]guz at qdC;-s>BG`iu,HOP"j\nV_W)'})|,9C>&St4H"\l$&:V;8)"gsPRlH
 S6]sBPDb:f<,&ReiQS59nI;6P{w1kPMSR|`8L6EaC?SBb|ujr$V^C8A+G3Z#'>U.>
Resent-Sender: cgw at buffalo.fnal.gov
Sender: python-list-admin at python.org
Errors-To: python-list-admin at python.org
X-BeenThere: python-list at python.org
X-Mailman-Version: 2.0beta3
Precedence: bulk
List-Id: General discussion list for the Python programming language <python-list.python.org>


In article <NDBBKEGCNONMNKGDINPFGEJDDFAA.infonuovo at email.com>, "Dennis E.
Hamilton" <infonuovo at email.com> wrote:

> 
> It goes something like this (matching the one that was already posted
> for you):
> 
> 1.	Assume P is the largest prime.
> 
> 2.	Let Q be the product of all of the primes up to P: 2*3*5* ... *P
> 
> 3.	The number Q+1 is not divisible by any of the primes that divide Q
> (they
> each leave a remainder of 1).  Therefore Q+1 is a prime or there is some
> other prime, larger than P (i.e., not among 2, 3, 5, ..., P), that
> divides Q+1.  So P is not the largest prime.  QED.
> 
> This all stands on the fundamental theorem of arithmetic: that every
> integer number is a product of primes in exactly one way.  
> That is Exercise 1.2.4-21 in "The Art of Computer Programming, v.1

I'm pretty sure that this last statement is not correct.  You don't
need unique factorization here.  Euclid's proof does not require a
hypothesis of unique factorization into primes, just a well-ordering
of primes, and existence of division-with-remainder.

You appeal to the fact that when you divide A into (A*B)+1, you get a
remainder of 1.  That's enough to show that none of the primes in the
list 2,...,P divide Q=1+(2*3*5*7..*P).  

Since none of these primes divide Q, either Q is prime, or some other
prime not in the list 2,...,P divides Q, either of which options
contradicts the hypothesis that Q is the largest prime.  Where do you
need to make use of the fact that prime factorizations are unique?
Euclid's proof is self-contained and referencing Knuth for a proof of
unique factorization just muddies the waters.

Interestingly, there are some quadratic number fields (discrete
subsets of the complex numbers) which don't have unique factorization,
but still have infinitely many prime elements.



From kelvin.chu at nospam.uvm.edu  Mon Apr 24 21:32:39 2000
From: kelvin.chu at nospam.uvm.edu (Kelvin Chu)
Date: 25 Apr 2000 01:32:39 GMT
Subject: Efficient coding advice?
Message-ID: <8e2sjn$te4l$1@swen.emba.uvm.edu>

Dear Python gurus;

I'm writing a bit of code to read the data from AIF files.  Data are
stored as large strings of characters and alternate channels.  In
addition, I want to translate these strings to integers.  So my code
looks like:

data = fp.readframes(nframes)
left = Numeric.zeros(nframes)
right = Numeric.zeros(nframes)
for i in range(0,nframes):
    left[i] = ord(data[i*2])
    right[i] = ord(data[i*2+1])

This seems to take a while, especially since nframes ~ 10^6.  So I
guess my questions are:

1.  Is there an easy way to split a large string so that every-other
    element gets assigned to an array?

2.  Is there an easy way to turn an entire array of characters into
    their ord value?

Thanks very much, I'd appreciate insight you have into this problem.
 

-- 

  kelvin.chu at nospam.uvm.edu			     (802) 656-0064



From ekw1 at my-deja.com  Sat Apr 29 01:25:07 2000
From: ekw1 at my-deja.com (ekw1 at my-deja.com)
Date: Sat, 29 Apr 2000 05:25:07 GMT
Subject: widget resizing
Message-ID: <8edrn7$nvh$1@nnrp1.deja.com>

can anyone point me to an example of how to have widgets resize to
expand/contract when the window is resized?  Or tell me what are the
relevant sections are in the python documentation?  are events and
bindings the sections i should study?  i'm not familiar with the terms
involved in doing this.  i am using python1.5.2 on win98se.  thanks for
any help.

ekw


Sent via Deja.com http://www.deja.com/
Before you buy.


From adrian at simspon.dyndns.org  Sat Apr  8 12:26:26 2000
From: adrian at simspon.dyndns.org (Adrian Eyre)
Date: Sat, 8 Apr 2000 17:26:26 +0100
Subject: win32all-130 for Python 1.6 available
References: <_ziF4.1488$1b4.6926@news-server.bigpond.net.au>
Message-ID: <Q4JH4.605$hh6.22805@news2-win.server.ntlworld.com>

> To help encourage testing of Python 1.6, I have made a
> version of win32all available.  This should also be considered
> alpha!

Looks good.

> Please let me know of any problems...

1. The following does not fold correctly:

def spam():
    return """spam
is
good"""

2. Similarly, the following should perhaps not fold:

# A comment ending with a colon:
# more comment here

3. I don't suppose it would be possible to make debugged scripts
run in a separate thread, to stop it blocking the program's event
loop?





From thomas at xs4all.net  Thu Apr 13 12:48:54 2000
From: thomas at xs4all.net (Thomas Wouters)
Date: Thu, 13 Apr 2000 18:48:54 +0200
Subject: Complaint Dept: My shoes!
In-Reply-To: <8d4s4d$mvr$1@nnrp1.deja.com>; from musingattheruins@my-deja.com on Thu, Apr 13, 2000 at 04:20:47PM +0000
References: <8d4s4d$mvr$1@nnrp1.deja.com>
Message-ID: <20000413184853.D14644@xs4all.nl>

On Thu, Apr 13, 2000 at 04:20:47PM +0000, musingattheruins at my-deja.com wrote:
> If shoe is None then len(shoe) raises an exception.  Would be nice if
> len(None) was 0 and NOT raise an exeption.
> 
> Example: (Python 1.5.2)
> 
> def show_shoe_wear_to_complaint_dept(name):
>    if name == "Idle":
>       return None
>    else:
>       return "pick up hammer..."
> 
> def hit_on_head(): return "POW!!!"
> 
> shoe = ""
> shoe = show_shoe_wear_to_complaint_dept("Idle")
> if len(shoe)>0: print shoe, hit_on_head()
> 
> If len(shoe) returned 0 then the example would work without error, the
> example above raises an exception whenever Idle shows his shoe to the
> complaint department.
> 
> The work around for the above is....
> shoe = show_shoe_wear_to_complaint_dept("Idle") or ""
> which gets rid of the problem but feels weird.

Or, of course,

    if name == "Idle":
	return ""

You apparently depend on show_shoe_wear_to_complaint_dept returning a string
(or at least something that len() works on... len() only works on sequences
(as it should))

Or perhaps the better solution would be to change:

if len(shoe)>0:

to 

if shoe:

as both None and the empty string are false. It entirely depends on what
else you wish to do with your shoe. (Dont get me started on how Shoe should
actually be a class with it's 'put_on' and 'throw' methods... ;)

Making errors silently go away is not quite the pythonic way. Nothing stops
you, of course, from using:

oldlen = len
def len(object):
    if not object:
        return 0
    return oldlen(object)

or something like

    try:
        return len(object)
    except TypeError:
        return 0

if you insist :P

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From ngps at madcap.dyndns.org  Thu Apr 20 12:03:17 2000
From: ngps at madcap.dyndns.org (Ng Pheng Siong)
Date: 20 Apr 2000 16:03:17 GMT
Subject: socket.SO_REUSEADDR
References: <38FEEA44.6C05E2B0@inka.de>
Message-ID: <8dn9o5$jsp$1@mawar.singnet.com.sg>

According to Michael Str?der  <michael.stroeder at inka.de>:
>     self.socket.setsockopt(
>       socket.getprotobyname('tcp'),
>       socket.SO_REUSEADDR,1
>     )

Michael,

    self.socket.setsockopt(socket.SOL_SOCKET, socket.SO_REUSEADDR, 1)

Cheers.

-- 
Ng Pheng Siong <ngps at post1.com> * http://www.post1.com/home/ngps



From jmassung at magpiesystems.com  Mon Apr 17 11:22:05 2000
From: jmassung at magpiesystems.com (Jeff Massung)
Date: Mon, 17 Apr 2000 09:22:05 -0600
Subject: Another question
Message-ID: <dVFK4.41$vr2.5817@news.uswest.net>

Okay, another quick question. If it is quicker to do:


import random
from random import randint
for i in range(1,1000000):
    print randint(1,10)


than this:


import random
for i in range(1,1000000):
    print random.randint(1,10)


Why would I not do "from <module> import *" with every module I load?

Jeff




From paulb at infercor.no  Fri Apr 28 06:40:37 2000
From: paulb at infercor.no (Paul Boddie)
Date: Fri, 28 Apr 2000 12:40:37 +0200
Subject: ODBC error trap resolved!  Now...
References: <A63AFB20111ED311AC5500A0C96B7BF5015F0F50@orsmsx36.jf.intel.com>
Message-ID: <39096AA5.836C1F36@infercor.no>

"Daley, MarkX" wrote:
> 
> OK, so the error trapping is good, but now I run into an append problem.  I
> have a list defined in the beginning of the function, like so:
> 
> servername = ['ServerA', 'ServerB', 'ServerC']
> for index in servername:
>         try:
>                 #Big routine to collect and write data to a file.
>         except dbi.opError:

This should really be 'ODBC.something.OperationalError' or
'module.OperationalError'.

>                 servername.append(index)
> 
> So, if one of the servers doesn't make the connection, it gets added to the
> end of the list until it does connect.  This worked great in my test
> scripts, but when put to practice, I get this:
> 
> Traceback (innermost last):
>   File "<pyshell#1>", line 1, in ?
>     collect.collect()
>   File "C:\PROGRA~1\Python\collect.py", line 51, in collect
>     servername.append(index)
> AttributeError: 'string' object has no attribute 'append'

It looks like 'servername' has been made to refer to a string value. Since
strings can be treated like other sequence objects, you may be getting single
character strings assigned to 'index' in the loop as well.

> I didn't see this when I tested my append function.  What do I not
> understand about lists in this context?

It looks like an incorrect assignment has happened somewhere, but it is hard to
tell from what you have written. If the above suggestion doesn't help, perhaps
you could provide a bit more detail.

Paul


From nmm1 at cus.cam.ac.uk  Wed Apr  5 13:07:30 2000
From: nmm1 at cus.cam.ac.uk (Nick Maclaren)
Date: 5 Apr 2000 17:07:30 GMT
Subject: Pass by reference ?
References: <38E864C9.E1EE94F7@ecs.soton.ac.uk> <38E8D449.FBC1E2AD@roguewave.com> <8caohp$j4v$1@pegasus.csx.cam.ac.uk> <38EB6797.F03EEFB3@roguewave.com>
Message-ID: <8cfrsi$83a$1@pegasus.csx.cam.ac.uk>

In article <38EB6797.F03EEFB3 at roguewave.com>, Bjorn Pettersen <bjorn at roguewave.com> writes:
|> 
|> Call by reference is (at least now) generally used to refer to Pascal
|> VAR arguments or C++ & arguments, both of which has similar semantics
|> which allow rebinding in the caller's scope.  This is in contrast to
|> many OO languages where arguments can be mutated but not rebound...  I'm
|> not sure if call-by-reference has changed meaning since the 60s, but
|> pass/call-by-object-reference is the term consistently used when I was
|> doing my type theory work in 85-87...  

Pascal VAR arguments don't involve actual rebinding - the method
that I was thinking of would (in Python terms) permit the called
function to change the caller's identifier from pointing to a tuple
to pointing to a dictionary.  Upon return, the caller would see the
new object.

|> Just because something is a bad idea doesn't mean you can't get a
|> research grant for it <wink>

True.  I knew some that were for developing a program to solve
the Halting Problem ....


Regards,
Nick Maclaren,
University of Cambridge Computing Service,
New Museums Site, Pembroke Street, Cambridge CB2 3QG, England.
Email:  nmm1 at cam.ac.uk
Tel.:  +44 1223 334761    Fax:  +44 1223 334679


From raffael at mediaone.net  Fri Apr 21 23:35:34 2000
From: raffael at mediaone.net (Raffael Cavallaro)
Date: Sat, 22 Apr 2000 03:35:34 GMT
Subject: The Simple Economics of Open Source
References: <3900611F.3B525BDA@cfdrc.com> <1255786824-35980039@hypernet.com> <f4JdfuAkLKA5EwcQ@jessikat.demon.co.uk>
Message-ID: <raffael-63EB69.23345321042000@news.ne.mediaone.net>

In article <f4JdfuAkLKA5EwcQ at jessikat.demon.co.uk>, Robin Becker 
<robin at jessikat.demon.co.uk> wrote:

>Painters
>exchange ideas, concepts and criticism freely, but compete for
>customers.

Speaking as a painter, I can tell you that many painters do not share 
ideas and concepts, but are quite secretive about them. In fact, there 
is a long history of artists keeping their working methods secret from 
each other, for precisely the same reason that any other set of business 
competitors do.

Ralph

-- 

Raffael Cavallaro, Ph.D.
raffael at mediaone.net


From jlehmann at nsw.bigpond.net.au  Fri Apr 21 03:37:17 2000
From: jlehmann at nsw.bigpond.net.au (John Lehmann)
Date: Fri, 21 Apr 2000 07:37:17 GMT
Subject: suggestion on an operator
References: <Pine.LNX.4.21.0004202210480.9838-100000@saturn.ispol.com>
Message-ID: <3900043E.78539877@nsw.bigpond.net.au>

Gregory Trubetskoy wrote:
> 
> Hello there -
> 
> I'm looking for suggestions for the best operator to use in deliniating
> modules from what's inside of them in a configuration file for the next
> version of mod_python (a.k.a httpdapy).
> 
> Basically, I'd like to give the user an option to specify a handler for a
> particular Apache event in Apache configuration. A handler will be a
> Python function (perhaps a class, but that's not relevant) in a Python
> module.
> 
> 

Hi,

What is wrong with using the period again?

eg.

PythonAuthenHandle mypackage.mymodule.authenticate

--
john j. lehmann


From ppi at searchy.net  Tue Apr 18 16:14:05 2000
From: ppi at searchy.net (Penpal International)
Date: Tue, 18 Apr 2000 22:14:05 +0200
Subject: Split
Message-ID: <38FCC20D.624AB2F6@searchy.net>

I'm fairly new to the language python. Mostly because it's object
oriented while I'm used to perl-like scripts.



I've already managed to split a string with this:

split = regsub.split(line,'\s+')

(I think...)

I should get something like this as result:

['This', 'is', 'a', 'test', 'short', 'and', 'sweet', 'of', 'split', '']

But how can I print only the 5th for example?


-- 
Penpal International
http://ppi.searchy.net/
ppi at searchy.net


From mbel44 at dial.pipex.net  Tue Apr  4 02:18:58 2000
From: mbel44 at dial.pipex.net (Toby Dickenson)
Date: Tue, 04 Apr 2000 07:18:58 +0100
Subject: Python and COM without Automation (Dispatch)
References: <vphd7o7b1vc.fsf@hiirihaukka.cs.tut.fi>
Message-ID: <382jesk68s64s0njruime7qjrb4usnq06i@4ax.com>

Vespe Savikko <vespe at cs.tut.fi> wrote:
>
>PythonCOM works very well with COM objects that implement IDispatch
>interface. However, I have a number of COM objects that do not
>implement IDispatch and I would like to use them from Python as well.

Unless those are dual interfaces.... pythoncom can happily use the
IDispatch part of a dual interface. (appologies if that's obvious)


Toby Dickenson
tdickenson at geminidataloggers.com


From ralph at inputplus.demon.co.uk  Sun Apr  2 07:24:37 2000
From: ralph at inputplus.demon.co.uk (Ralph Corderoy)
Date: 2 Apr 2000 12:24:37 +0100
Subject: else clauses in while and for loops
References: <Zv_C4.6415$4N.169440@news.tpnet.pl> <38DD2EE3.C7F3CB54@webamused.com> <8boe2k$gmh$1@inputplus.demon.co.uk> <20000327165149.A1525@better.net>
Message-ID: <8c7all$300$1@inputplus.demon.co.uk>

> > > > Please give some examples showing the real need for else
> > > > clauses in this context.
> >
> > I don't think that is a real need.  Wouldn't it normally be written
> > to return ASAP rather than jump through hoops.
>
> True.  But, using "else" avoids having to deal with a local variable
> to find out how it left the loop.

I realise that, but I didn't use a local variable either;  one wasn't
needed because it was the example I was replying to was poor.  The
original poster asked for `some examples showing the real need for else
clauses in this context'.  I was trying to point out that the example
given didn't show a real need.


Ralph.



From grant at nowhere.  Wed Apr 12 17:40:04 2000
From: grant at nowhere. (Grant Edwards)
Date: Wed, 12 Apr 2000 21:40:04 GMT
Subject: comparing strings and ints
References: <2YmI4.366$rc9.188014080@newsb.telia.net> <20000411134358.A2352304@vislab.epa.gov> <1256641282-42160637@hypernet.com> <20000412075812.A2449004@vislab.epa.gov> <jr%I4.643$rc9.190628352@newsb.telia.net> <20000412112935.A2446939@vislab.epa.gov>
Message-ID: <slrn8f9r9j.in3.grant@grante.comtrol.com>

Randall Hopper wrote:
>Fredrik Lundh:
> |Randall Hopper:
> |>Gordon McMillan:
> |> |Randall Hopper wrote:
> |> |> Fredrik Lundh:
> |> |>  |
> |> |>  |    Objects of different types always compare unequal, and are
> |> |>  |    ordered consistently but arbitrarily.
> |> |> 
> |> |> Is there a case where this could be useful (with its undefined
> |> |> behavior)?  I believe that is the root of the question.
> |> |
> |> |I'd happily bet your wife and firstborn child that there's code that
> |> |relies on it <wink>, if only because after sorting a list, all the ints
> |> |will end up in a clump, and all the strings in another, etc. 
> |>
> |>Oh, I also have __no doubt__ that someone is using this unofficial,
> |>undocumented behavior.  But they don't have a leg to stand on if the
> |>ordering breaks in a future Python version ;-)
> |
> |interestingly enough, you included my documentation quote in
> |a post claiming that this is an undocumented feature...
>
>I was of course referring to the behavior Gordon McMillan
>mentioned in the message I replied to.  Namely, that sorting a
>list of objects of hetrogenous primitive types "groups" the
>instances of a particular type together (ints together, strings
>together, etc.).
>
>I'm taking Gordon's word that that's what it does.  The
>documentation you cited simply said "arbitrary" ordering.

No, it said "_consistently_ but arbitrarily" [emph added].
That means that you don't which result you will get when a
string and an int are compared, but you know you will always
get the _same_ result.  That is sufficient to gaurantee that
ints will clump together and strings will clump together in a
sorted list.  You just don't know in what order the clumps are
going be ahead of time.

>Unless it's documented as official behavior, I would think that
>expecting it is probably not a safe bet.

Assuming the quote is correct, it looks like documented
behavior to me.

-- 
Grant Edwards                   grante             Yow!  Dehydrated EGGS are
                                  at               STREWN across ROULETTE
                               visi.com            TABLES...


From python at rose164.wuh.wustl.edu  Wed Apr 26 07:46:17 2000
From: python at rose164.wuh.wustl.edu (David Fisher)
Date: Wed, 26 Apr 2000 06:46:17 -0500
Subject: huh? Python says that AttributeError: 'string' object has no attribute 'split'
References: <8e62qm$511$1@nnrp1.deja.com>
Message-ID: <003101bfaf75$083fda20$0201a8c0@spkydomain>

----- Original Message -----
From: <schmitthead at my-deja.com>
Newsgroups: comp.lang.python
To: <python-list at python.org>
Sent: Wednesday, April 26, 2000 1:37 AM
Subject: huh? Python says that AttributeError: 'string' object has no
attribute 'split'

>
> #mac2dos.py-------------------------------------------------------------
> import os
> import sys
> import string
>
>
> #-----------------------------------------------------------------------
> def mac2dosfile( file ):
>  myline = file.read()
>  lines = myline.split( myline, chr(10) )

change to:

lines = string.split( myline, chr(10) )
           ^^^^^

>  for l in lines:
>   l = l.strip( l )
>  file.writelines( lines )
>
> #-----------------------------------------------------------------------
> def callbackfunc( junk, dir, names ):
>  for name in names:
>   file = open( name, 'r+' )
>   mac2dosfile( file )
>   close( file )
>
>
> #-----------------------------------------------------------------------
> def mac2dos( dir ):
>  os.path.walk( os.getcwd(), callbackfunc, '' )
>
>
> #-----------------------------------------------------------------------
> if __name__ == '__main__':
>  if ( len(sys.argv) < 2 ):
>   directory = os.getcwd()
>  else:
>   directory = sys.argv[1]
>  mac2dos( directory )
> #-end-of-mac2dos.py-----------------------------------------------------
>
>
> Sent via Deja.com http://www.deja.com/
> Before you buy.
> --
> http://www.python.org/mailman/listinfo/python-list




From c.evans at clear.net.nz  Wed Apr 26 07:08:01 2000
From: c.evans at clear.net.nz (Carey Evans)
Date: Wed, 26 Apr 2000 11:08:01 GMT
Subject: Python 1.6a2 Unicode bug (was Re: comparing strings and ints)
References: <jr%I4.643$rc9.190628352@newsb.telia.net>
	<1256565470-46720619@hypernet.com>
	<6M1J4.662$rc9.209708544@newsb.telia.net>
	<8daop0$8fk$1@slb6.atl.mindspring.net>
	<d3HK4.2612$Za1.38971@newsc.telia.net> <8e6bsl$f1a$1@nnrp1.deja.com>
Message-ID: <87ln212k8e.fsf@psyche.evansnet>

fredrik at pythonware.com writes:

> postscript: the powers-that-be has decided that this is not
> a bug.  if you thought that strings were just sequences of
> characters, just as in Perl and Tcl, you're in for one big
> surprise in Python 1.6...

In the latest versions of Perl and Tcl, aren't all strings Unicode
strings anyway?  It seems to me you're still in for _less_ of a
surprise in Python 1.6.

-- 
	 Carey Evans  http://home.clear.net.nz/pages/c.evans/

		 "I can't believe it's not Cthulhu!"


From moshez at math.huji.ac.il  Tue Apr 18 03:08:26 2000
From: moshez at math.huji.ac.il (Moshe Zadka)
Date: Tue, 18 Apr 2000 09:08:26 +0200 (IST)
Subject: Forward References?
In-Reply-To: <eviltofu-4C5688.21593817042000@news.earthlink.net>
Message-ID: <Pine.GSO.4.10.10004180905550.3780-100000@sundial>

On Tue, 18 Apr 2000, Jerome Chan wrote:

> Just a silly question but how do I do forward references?

This has nothing to with your problem.

> class a:
>    def x(self):
>       y(self)
        ^^^
         Change this to self.y()
      
>    def y(self):
>       print "Hello"

Attribute reference is always done explicitly
--
Moshe Zadka <mzadka at geocities.com>. 
http://www.oreilly.com/news/prescod_0300.html
http://www.linux.org.il -- we put the penguin in .com




From johann at physics.berkeley.edu  Thu Apr  6 18:20:35 2000
From: johann at physics.berkeley.edu (Johann Hibschman)
Date: 06 Apr 2000 15:20:35 -0700
Subject: Beginning to be disapointed
References: <Aw4H4.8078$pR.68745@carnaval.risq.qc.ca> <mt3dozt6sh.fsf@astron.berkeley.edu> <mtu2hfrnl7.fsf@astron.berkeley.edu> <87snwydgzh.fsf@lumen.med.iupui.edu>
Message-ID: <mthfdeswgc.fsf@astron.berkeley.edu>

Michael A Miller writes:

> The innards of _tableio doesn't use NumPy at all.  All it does is
> tokenize a file, run atof on it and return a list.  The list is
> stuffed into a dictionary and there certainly could be some
> weirdness in my reference counting there, but I've never
> experienced the sort of problem that Demers reports.  NumPy is
> used in a python wrapper to convert the list to an array.

Ah, that's probably not it, then.  Sorry to malign your module, but my
first reaction in this sort of thing is always to look for bugs in
extensions.  I had some bad C++ code in a module of my own once do
something like this.  It didn't consume memory and thrash the disk,
but it operated for a while, then seg-faulted mysteriously.

Bring on the code samples!  ;-)

-- 
Johann Hibschman                           johann at physics.berkeley.edu


From jmg at ecs.soton.ac.uk  Mon Apr  3 05:30:49 2000
From: jmg at ecs.soton.ac.uk (Jacek Generowicz)
Date: Mon, 03 Apr 2000 10:30:49 +0100
Subject: Pass by reference ?
Message-ID: <38E864C9.E1EE94F7@ecs.soton.ac.uk>

Hi,

I'm just trying to familiarize myself with Python.
In Magnus Lie Hatland's Instant Python, he
mentions that `all parameters in Python are passed
by reference'.

So I tried the following code:

#!/usr/bin/python

def increment(x):
    x = x + 1
    return x

a = 0
print a
print increment(a)
print a
====================

output:

0
1
0

Given my understanding of passing by reference,
the last number should be 1.

What am I missing ?

Thanks,

Jacek




From bsb at winnegan.de  Wed Apr 12 10:37:07 2000
From: bsb at winnegan.de (Siggy Brentrup)
Date: 12 Apr 2000 16:37:07 +0200
Subject: Python idiom: Multiple search-and-replace
In-Reply-To: Randall Hopper's message of "Wed, 12 Apr 2000 10:08:16 -0400"
References: <20000412100816.A2431031@vislab.epa.gov>
Message-ID: <871z4bqtbg.fsf@orge.winnegan.de>

Randall Hopper <aa8vb at yahoo.com> writes:

> There's got to be a better way.  Is there a Python idiom I'm missing?
> 
> I want to do search-and-replace of multiple symbols on each line of a file.
> But the simple-minded code below takes a while.  It simply uses
> string.replace N times per line (N is 240 in this case).  There are 33994
> lines (1.1Meg).
> 
> Total time: 140.7 seconds.
> 
> I stopped to investigate.  What was slowing it up so much?
> 
> - Comment out the inner loop, and Python completes in 0.9 sec.  No prob
>   there. It can read and write-the data very quickly.
> - It's not dictionary lookups.  Converted to a tuple-list and took 3 sec
>   longer.  
> - I tried a few other things but it only made it take longer than
>   140 sec.
> 
> Is there a Python feature or standard library API that will get me less
> Python code spinning inside this loop?   re.multisub or equivalent? :-)
> 
> Thanks,
> 
> Randall
> 
> 
> ------------------------------------------------------------------------------
> 
>   symbol_map = { 'oldsym1' : 'newsym1', oldsym2' : 'newsym2', ... }
>   fp = open( net_path, "r" )
> 
>   while 1:
>     line = fp.readline()
>     if not line: break
> 
>     for old_sym in symbol_map.keys():
>       line = string.replace( line, old_sym, symbol_map[ old_sym ] )
> 
>     out_fp.write( line )
> 
> ------------------------------------------------------------------------------

symbol_items = symbol_map.items()

while 1:
    line = fp.readline()
    if not line: break

    for old, new in symbol_items:
        line = string.replace(line, old, new)

    out_fp.write(line)

springs into mind (untested)

Regards
  Siggy

-- 
Siggy Brentrup - bsb at winnegan.de - http://www.winnegan.de/
****** ceterum censeo javascriptum esse restrictam *******



From sblakey at freei.com  Wed Apr 26 15:10:12 2000
From: sblakey at freei.com (Sean Blakey)
Date: Wed, 26 Apr 2000 19:10:12 +0000
Subject: Simple Function Question
References: <8e7ddt$18n@autodesk.autodesk.com>
Message-ID: <39073F14.2B7F3A3F@freei.com>

def make_incrementer(n):
     def increment(x, n=n):
         return x+n
     return increment

When you call make_incrementer with an argument of 1, the first thing
make_incrementer does is defina a new function, increment.  

Increment takes two arguments, x and n.  n defaults to the same value n
has in make_incrementer (1).  Since n has a default value, increment can
be called as a one-argument function that simply returns whatever it is
called with(x) + 1.

make_incrementer then returns this function.  This is the tricky part:
the return value of the make_incrementer function is itself a function.
When this function is called with one argument, that function becomes x
in increment().  If that function is called with two arguments, the two
arguments will become x and n.

Some people find the OO version of this easier to understand.  Perhaps
this will help:
>>>class Incrementer:
>>>	def __init__(self, increment_amount):
>>>		self.increment_amount = increment_amount
>>>	def increment(self, value):
>>>		return value + self.increment_amount
...
>>>i = Incrementor(1)
>>>val = i.increment(3)
>>>print val
4

Once you understand this, the use of the __call__ method becomes
apparent:
>>>class Incrementer:
>>>	def __init__(self, increment_amount):
>>>		self.increment_amount = increment_amount
>>>	def __call__(self, value):
>>>		return value + self.increment_amount
...
>>>i = Incrementor(1)
>>>val = i(3)
>>>print val
4


Akira Kiyomiya wrote:
> 
> Hi,  Could you explain this in dummy's language?
> 
> This function prints out '4' and I have a feeling that x=1 and n=3, so
> add1(3) prints '4'.
> 
> However, I have a difficlut time trying to see how these two arguments
> assigned to both x (especially x) and n.
> 
> Thanks
> 
> Akira
> 
> # Return a function that returns its argument incremented by 'n'
> def make_incrementer(n):
>     def increment(x, n=n):
>         return x+n
>     return increment
> 
> add1 = make_incrementer(1)
> print add1(3)  # This prints '4'
> 
> --
> http://www.python.org/mailman/listinfo/python-list

-- 
Sean Blakey
FreeInternet.com
sblakey at freei.com
(253)796-6500x1025



From robin at jessikat.demon.co.uk  Thu Apr 27 11:16:55 2000
From: robin at jessikat.demon.co.uk (Robin Becker)
Date: Thu, 27 Apr 2000 16:16:55 +0100
Subject: module case bug in 1.6a2 Win9x
Message-ID: <UGQuwCAnnFC5Ew+e@jessikat.demon.co.uk>

I believe there's a module case bug with long named modules under win98
at least. I believe the bug relates to the length of the filename
somehow.

This is with win98 version 2 + 1.6a2

C:\Python16\Tools\test>copy PWD.py PrintWindowDialog.py
        1 file(s) copied

C:\Python16\Tools\test>dir

 Volume in drive C has no label
 Volume Serial Number is 291A-00FF
 Directory of C:\Python16\Tools\test

.              <DIR>        27/04/00  16:02 .
..             <DIR>        27/04/00  16:02 ..
PWD      PY             71  27/04/00  16:03 PWD.py
PRINTW~1 PY             71  27/04/00  16:03 PrintWindowDialog.py
         2 file(s)            142 bytes
         2 dir(s)        3,314.16 MB free

C:\Python16\Tools\test>..\..\python
Python 1.6a2 (#0, Apr  6 2000, 11:45:12) [MSC 32 bit (Intel)] on win32
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> from pWD import PWD
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
NameError: Case mismatch for module name pWD
(filename PWD.py)
>>> from printWindowDialog import PWD
>>> PWD()
PWD initialized printWindowDialog.py
<printWindowDialog.PWD instance at 7903f0>
>>>


the two files PWD.py and PrintWindowDialog are identical and look like

class PWD:
        def __init__(self):
                print 'PWD initialized', __file__
-- 
Robin Becker


From meh9 at cornell.edu  Mon Apr  3 16:17:17 2000
From: meh9 at cornell.edu (Matthew Hirsch)
Date: Mon, 03 Apr 2000 16:17:17 -0400
Subject: Creating variables on the fly...
References: <meh9-25D0C5.15160303042000@news.cit.cornell.edu> <38E8EFF0.D71D0415@gmd.de>
Message-ID: <meh9-32FDF8.16171703042000@news.cit.cornell.edu>

I like this solution.  How do I use the apply function in the body to 
call add()?  (I won't know ahead of time how many arguments are going in 
the function).

Thanks,
Matt




In article <38E8EFF0.D71D0415 at gmd.de>, Joachim Kaeber <kaeber at gmd.de> 
wrote:

> Hi,
> 
> what about this:
> 
> def add(*args):
>    sum=0
>    for arg in args:
>      sum=sum+arg
>    return sum
> 
> add(1,2,3,4)
> => 10
> 
> add (1,2,3,4,5,6,7)
> => 28
> 
> 
> Matthew Hirsch wrote:
> > 
> > Hi All,
> > 
> > Let's say I have four variables:
> > 
> > a=1
> > b=2
> > c=3
> > d=4
> > 
> > And I have a function that adds these variables together:
> > 
> > def add(a,b,c,d):
> >    return a+b+c+d
> > 
> > But now let's say I have twenty variables that I want to add together.
> > This function will no longer work.  I would have to rewrite it as
> > a+b+c+d+e+f+g+...+(20th letter).  Is there a way to dynamically create a
> > variable name?  So that my add function can automatically determine how
> > many variable names to create and then add the values together.  In
> > other words, I'd ideally like something like:
> > 
> > def add(number_of_variables):
> >    return a+b+c+...+(letter corresponding to number_of_variables)
> > 
> > Thanks for your help,
> > 
> > Matt


From mak at imakhno.freeserve.co.uk  Sun Apr  2 05:04:20 2000
From: mak at imakhno.freeserve.co.uk (Makhno)
Date: Sun, 2 Apr 2000 10:04:20 +0100
Subject: Linking Python modules to C++ modules...a problem
Message-ID: <8c72kc$tmq$1@news5.svr.pol.co.uk>

I'm trying to write an interface to a library written in C++ from Python, on
the BeOS.
The module is called OrchardExt, and I want to link it to a library called
libOrchard.so

So I create an appropriate OrchardExt.c and a OrchardExtwrapper.c, the
wrapper contains the functions (extern'd "C") that simply 'pass on' the
function call to the libOrchard.so library, whereas the OrchardExt.c file is
empty (except for an #include that include the declarations for the
library).

I run
Make -f Makefile.pre.in boot

I edit the Makefile, making it link to the appropriate libraries and
ensuring that it will compile in C++.
then:
make
make install

Unfortuanately, this doesn't work. The output is shown below.

Do I need to recompile the Python binary for the BeOS with a C++ compiler,
or is such module compiling screwed on the BeOS for Python?
I have already written the Perl interface to this library, and it worked a
lot more beautifully than this.

~~~~~~~~~~


Traceback (innermost last):
  File "/boot/home/Py_script", line 2, in ?
    import OrchardExt
ImportError: Unknown error loading
/boot/home/config/lib/python1.5/site-packages/OrchardExt.so







From meh9 at cornell.edu  Fri Apr 14 15:15:58 2000
From: meh9 at cornell.edu (Matthew Hirsch)
Date: Fri, 14 Apr 2000 15:15:58 -0400
Subject: Attributes (+Numerical Python)
Message-ID: <meh9-986AAB.15155814042000@news.cit.cornell.edu>

Hi All,

In Numerical Python, you can say something like vector.imag and it will 
return arrays of the imaginary parts of the elements in the array, 
vector.  But this attribute only exists when the array is complex.  I'd 
like to do something like the following, but I have no idea what the 
syntax should be:

if the .imag attribute exists:
   do something
else
   do something else

How should this be written?

Thanks for your help,
Matt


From musingattheruins at my-deja.com  Thu Apr 13 12:31:49 2000
From: musingattheruins at my-deja.com (musingattheruins at my-deja.com)
Date: Thu, 13 Apr 2000 16:31:49 GMT
Subject: simple httplib and urllib timeout question
References: <Pine.BSF.4.10.10004111337520.29635-100000@chi.pair.com>
Message-ID: <8d4spl$nth$1@nnrp1.deja.com>

> > Just a quick question, im writing a small webbot for a python
> > learning experience and it comes across some sites that will make it
> > hang for a while, if the site is down or extremely slow etc, is
> > there a way to set a faster timeout for httplib . Or is there a
> > better way to handle it ?

I had the same problem and would get back exceptions that said a
timeout had occurred.  I used a retry as a work-around....

class Page:
	"Page: class to get a page from the web."
	def get(self, host, page):
		"""
			p = Page()
			text = p.get(host, page)
		"""
		for i in range(3):
			try:
				import httplib
				http = httplib.HTTP(host)
				http.putrequest("GET", page)
				http.putheader("Accept", "text/html")
				http.putheader("Accept", "text/plain")
				http.endheaders()
				errcode,errmsg,headers = http.getreply()
				fp = http.getfile()
				sData = fp.read() or ""
				fp.close()
				del http
				if len(sData)>0:
					return sData
			except:
				pass

which will retry up to three times if the returned data is empty or if
an exception is raised.

L8R :-)


Sent via Deja.com http://www.deja.com/
Before you buy.


From pinard at iro.umontreal.ca  Mon Apr  3 09:15:19 2000
From: pinard at iro.umontreal.ca (=?ISO-8859-1?Q?Fran=E7ois_Pinard?=)
Date: 03 Apr 2000 09:15:19 -0400
Subject: Unicode program representation
In-Reply-To: "Neil Hodgson"'s message of "Mon, 3 Apr 2000 22:12:13 +1000"
References: <015c01bf9d0b$76f9b2a0$59dd3fcb@neil> <oqem8n4zlx.fsf@titan.progiciels-bpi.ca> <00ac01bf9d65$d988c460$36dd3fcb@neil>
Message-ID: <oqwvmf1g2g.fsf@titan.progiciels-bpi.ca>

"Neil Hodgson" <neilh at hare.net.au> ?crit:

>    But then, what role do you see the u" form having?

First, beware these things are new to me as they are new to you.  People who
went to the miseries and length of getting these ideas adopted, against
surely many competing directions and trends (I know how heavy these games
could get sometimes), are surely in a better position than me to reply.

Of course, u'' allows for \uHHHH sequences to represent Unicode characters
outside Latin-1.  But if we put this aside, I would guess that u'' is a
bit, for strings, what the `L' suffix is for integers: that is, it forces
a wider internal notation, which then propagate through coercion while
using various string operators.

One could do, in Python, without usual integers (only using long ones),
but usual integers have a shorter in-memory representation, and are much
faster to operate upon.  One could do, in Python, without narrow strings
(please do not call them ASCII strings, as ASCII is strictly 7-bits)
(only using wide strings, calling them Unicode strings is OK :-), but
narrow strings have a shorter in-memory representation, and are slightly
faster to operate upon.

I much like the way Python is offering Unicode.  Just write as you always
did.  At various points you really need Unicode strings in your code, just
use them: everything is going to be fairly automatic, for conversions and
all such things (unless you do not like UTF-8, in which case some doing is
needed to state your preference).  Until `sre' definitely replaces `re',
you'll have to specify which one you want, forcing you as a programmer to a
better tracking of narrow vs wide string types: but this is only temporary,
the goal seems to be that the programmer should not worry about these things.

-- 
Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard





From mhammond at skippinet.com.au  Wed Apr 12 06:10:48 2000
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Wed, 12 Apr 2000 10:10:48 GMT
Subject: Visual Studio add-ins in Python
References: <38F1E44D.13BE21D0@nih.gov>
Message-ID: <IYXI4.2025$UB.13727@news-server.bigpond.net.au>

"Jonathan Epstein" <Jonathan_Epstein at nih.gov> wrote in message
news:38F1E44D.13BE21D0 at nih.gov...

> Has somebody already written an add-in using Python?  I have the
> _Programming Python_ book, but gather that it would be worthwhile to
> also purchase _Python Programming on Win32_ for this project.  Correct?

I hope so :-)  You wont find anything specific to your problem, but will
get some good Win32 specific information.

You should also check out Ken Seehof's work at
http://starship.python.net/crew/seehof/ - some time ago he did lots of
work with MSVC.

Mark.




From doesntexist at noemailthanks.net  Wed Apr  5 06:51:47 2000
From: doesntexist at noemailthanks.net (Samu Mielonen)
Date: Wed, 05 Apr 2000 13:51:47 +0300
Subject: Python & Zope professionals in Finland?
Message-ID: <bg6messrsdc2e2228s3i9qileueev0skgc@4ax.com>

I'm looking for Python (and preferably Zope) knowledgeable programmers
and software designers for an open source project from and within Finland.

I know this is a long shot (and yes, I've already posted this in Finnish
groups to no avail ), but I thought this group might also be visited by
Finnish Python developers who don't ready local newgroups.

So, in case you are well versed in Python and would like to work for
an open source project in the public sector, please contact me at
samu at uta.fi (please parse the e-mail address manually, I get
enough spam as it is). Zope skills are a plus, but not required. You'll
have to learn them along the way though :)

I'm preferably looking for a person who could live and work in Helsinki,
Finland, speaks Finnish fluently and is ready to work in a small team
to produce an open source web application. A lot of work has been
done already, but now there's a need to port the application on top
of Zope.

Please let me know if you are interested or know of somebody who might
be. We may not pay the best wages, but you can't beat our athmosphere :)

Regards,
   Samu Mielonen
   University of Art & Design, Helsinki, Finland
   samu at uta.fi (please reconstruct the e-mail address from here instead 
   of replying to this message)

PS Yes, I've also tried the Python Jobs announcement board. No dice.



From kramsay at aol.commangled  Sat Apr  8 16:59:54 2000
From: kramsay at aol.commangled (Keith Ramsay)
Date: 08 Apr 2000 20:59:54 GMT
Subject: Why should I switch to Python? - Infinity of Primes
References: <467ED74C3A44BDFC.B66525D52208472C.27972C76E272D139@lp.airnews.net>
Message-ID: <20000408165954.16420.00000265@ng-cv1.aol.com>

>The constructivist sect most familiar to me says,
...

Not exactly; I have a discussion of the 
matter I'm posting to sci.math, but
probably it is too far afield for
comp.lang.python.

Keith Ramsay



From l_user at yahoo.com  Thu Apr 27 17:31:26 2000
From: l_user at yahoo.com (GTE)
Date: Thu, 27 Apr 2000 21:31:26 GMT
Subject: Metaclasses?
References: <3905F4A0.2B39C705@darmstadt.gmd.de> <m3em7tg8ei.fsf@atrus.jesus.cam.ac.uk> <3906542D.62B94F17@darmstadt.gmd.de> <8e937n$f4p$1@nnrp1.deja.com>
Message-ID: <Ok2O4.726$4A6.145248@dfiatx1-snr1.gtei.net>

Andrew Cooke <andrew at andrewcooke.free-online.co.uk> wrote in message
news:8e937n$f4p$1 at nnrp1.deja.com...
> In article <3906542D.62B94F17 at darmstadt.gmd.de>,
>   Robb Shecter <shecter at darmstadt.gmd.de> wrote:
> > Michael Hudson wrote:
> [...]
> > Choice A:  Have a smalltalk-like instance variable declaration.
> [...]
> > to say that I like "Choice A".  I come from Java, and it did bug me a
> > bit that there was no one easy place to see what instance variables a
> > class has.  It made understanding the XML parser source code harder,
> for
> > example.
>
> this doesn't really save typing - instead of typing self you have to
> type the local declaration.  worse, this introduces extra syntax (after
> all, you still need ".").
>
> but more importantly, having "self" is good for the soul.  in most other
> languages objects function in some "magical" manner.  having self makes
> the mechanism clear - which is helpful both to newbies understanding OOP
> for the first time, and to fools like me that get easily confused when
> things happen behind their back...
>
> andrew
>
> ps it doesn't have to be "self", afaik - you could use just "s" to save
> key presses.... :-)

Not to mention that most C++ coding standards require usage of 'm_' as a member
variable prefix, so if you shorten 'self.' to 's.' (or, even, 'm.'), you'd have
to do exactly the same amount of typing.

On the other hand I kinda like Visual Basic's 'with' construct. Consider:
def method(self,arg1,arg2):
    with self:
        .spam = "spam"
        .eggs = arg1+arg2

Actually, for class methods that 'with ...:' could be <gasp> implicit.

Vadim





From sabren at manifestation.com  Mon Apr 17 08:54:21 2000
From: sabren at manifestation.com (Michal Wallace (sabren))
Date: Mon, 17 Apr 2000 08:54:21 -0400 (EDT)
Subject: How to fire off a web browser process from Python in Win?
In-Reply-To: <aDwK4.1834$lM4.436461@news.pacbell.net>
Message-ID: <Pine.BSF.4.10.10004170853360.25069-100000@chi.pair.com>

On Sun, 16 Apr 2000, Grant Munsey wrote:

> Is there a call somewhere in the std library for staring up the default
> ( the one that the user has set for opening .htm and .html files)
> web browser on Win 95/8/NT/2K?

import os
os.popen("start http://www.python.org/")



.. seems to work for me, anyway..

Cheers,

- Michal
-------------------------------------------------------------------------
http://www.manifestation.com/         http://www.linkwatcher.com/metalog/
-------------------------------------------------------------------------




From boncelet at udel.edu  Thu Apr  6 21:58:09 2000
From: boncelet at udel.edu (Charles Boncelet)
Date: Fri, 07 Apr 2000 11:58:09 +1000
Subject: A Mountain of Perl Books + Python Advocacy
References: <20000405053920.20579.qmail@web2102.mail.yahoo.com>
Message-ID: <38ED40B0.A48EC4CC@udel.edu>

lewst wrote:

>
> Now something I can't quite figure out is: why are there so many more
> books on Perl out there than on Python?

Quick personal note: a few years ago, I realized that scripting languages
were the future and, AFAIK, Perl was it, so I bought
a Perl book.  After reading it, I didn't really "get it", so I
bought another, and another.  I also picked up every issue of
the Perl Journal for awhile.  Wrote a few Perl scripts, but
never really "got it".

6 months ago, I discovered Python and it immediately clicked.
Learned most of Python before buying my first Python book.

Now I use Python for everything (well, ocassionally C or
Matlab).

But I own 3 or 4 Perl books and only 1 Python book.

    Charlie Boncelet
------
Charles Boncelet, University of Delaware,
On sabbatical at ADFA, Canberra Australia,
Home Page: http://www.ece.udel.edu/~boncelet/





From pinard at iro.umontreal.ca  Wed Apr 26 21:17:56 2000
From: pinard at iro.umontreal.ca (=?ISO-8859-1?Q?Fran=E7ois_Pinard?=)
Date: 26 Apr 2000 21:17:56 -0400
Subject: [Python-Dev] Re: Python 1.6a2 Unicode bug (was Re: comparing strings and ints)
In-Reply-To: "Fredrik Lundh"'s message of "Wed, 26 Apr 2000 19:22:09 GMT"
References: <l03102805b52ca7830b18@[193.78.237.154]> <39072F0C.5214E339@lemburg.com> <BlHN4.1719$wYl.195075584@newsb.telia.net>
Message-ID: <oqvh14fikb.fsf@titan.progiciels-bpi.ca>

"Fredrik Lundh" <effbot at telia.com> writes:

> M.-A. Lemburg <mal at lemburg.com> wrote:

> does this mean that the 8-bit string type is deprecated ???

I did not follow the implementation discussions, but from the outside, it
looks like 8-bit strings are kept and will be speedy.  People should either:

1) use them for ASCII without bothering or thinking much,

2) write Python source files in UTF-8, relying on the default encoding,

3) explicitly "cast" the encoding while converting strings to Unicode.

My guess is that 2) is not such good style, even if we know it will likely
be used in programs.  This might be convenient for speed prototyping, or
interactive use, say.  Python already has these two level of languages,
comparable to spoken English versus written English, the former being
more relaxed, and the latter more formal.  There are many things I let me
write in interactive Python that I conscientiously avoid in saved scripts.
That should probably be the case of 2).

Better would be to always use 3) for non-ASCII strings which are not
directly Unicode.  The notion of good style would almost turn invalid UTF-8
in 8-bit strings into an academical problem, and one more reason to avoid 2).

-- 
Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard




From rhicks at nospam.rma.edu  Tue Apr  4 10:16:15 2000
From: rhicks at nospam.rma.edu (Robert Hicks)
Date: Tue, 04 Apr 2000 14:16:15 GMT
Subject: Starving for an Advanced Python Book
References: <20000404040941.7766.qmail@web2106.mail.yahoo.com> <urlG4.2557$d13.25939@typhoon.southeast.rr.com>
Message-ID: <POmG4.6$O9.4610@news>

There are also many titles to look at on the Python bookstore web page.

I would highly recommend to anyone "Python and Tkinter Programming" by John
Grayson. It documents Tkinter and Pmw very well. Pmw should IMHO be
incorporated into the standard Python distribution.

Bob


"Don Tuttle" <tuttledon at hotmail.com> wrote in message
news:urlG4.2557$d13.25939 at typhoon.southeast.rr.com...
> "lewst"
> >ORA's "Programming Python" is a natural choice, but I
> > can't bear buying this outdated book when the 2nd edition is due out
> > in a few months (I'm dying here Lutz, hurry it up already!).
>
> I went ahead and bought it because Oreilly's cheif editor says, "We're
> timing the new edition of Programming Python to follow shortly after the
> release of Python 2.0."(aka Python 3K).  If that turns out to be the case,
> it won't published until sometime in 2001.  See his complete (and somewhat
> dated) remarks at  http://www.oreilly.com/frank/python.html
>
> > Does anyone have any suggestions?  I suppose I'm looking for a text
> > that has examples of how to solve various kinds of programming
> > problems in Python and also examples of code showing how the library
> > and various modules actually work.
>
> Fredrick Lundh's "(the Eff-bot Guide To) The Standard Python Library"
>  http://www1.mightywords.com/asp/bookinfo/bookinfo.asp?theisbn=EB00002582
> has many short examples.
>
> Also "The Python Annotated Achives" is pretty much just examples and
> comments.
>
http://shop.barnesandnoble.com/booksearch/isbnInquiry.asp?userid=589TPW6Z1E&
> mscssid=7T95UX3412S92K8S0017QRP4NV5DCFDD&srefer=&isbn=0072121041
>
> >How is the "Python Essential
> > Reference" on examples?  Is it just a rebinding of the python.org
> > library reference
>
> Excelent reference, but no examples.
>
> Don
>
>
>




From musingattheruins at my-deja.com  Wed Apr 19 17:45:33 2000
From: musingattheruins at my-deja.com (musingattheruins at my-deja.com)
Date: Wed, 19 Apr 2000 21:45:33 GMT
Subject: How to write to printer in Windows?
References: <38FC43DF.78A38D5D@tcac.net> <20000419064729.A517358@vislab.epa.gov>
Message-ID: <8dl9do$he5$1@nnrp1.deja.com>

> From my DOS days, I seem to recall that 'lpt1:' was the correct name
> for the printer device.  lpt1 may be trying to write to a file in a
> directory for which you don't have permission.

You can save the data that you want to print to a text file (say
text.txt) and then call...

   print text.txt

You can do this with a system call...
   import os
   os.system("print text.txt")

or you can go the long way around and try this module....
import printer
p = printer.Printer()
p.open()
p.write("this is some test\nand so is this")
p.close()
del p

You can find it at...
http://musingattheruins.homepage.com/printer.html

L8R,
Bill :-)


Sent via Deja.com http://www.deja.com/
Before you buy.


From aahz at netcom.com  Sun Apr 30 23:11:16 2000
From: aahz at netcom.com (Aahz Maruch)
Date: 1 May 2000 03:11:16 GMT
Subject: Circular Dependancies
References: <8F26D44DDjoshtinamecom@209.208.0.7>
Message-ID: <8eiskk$429$1@slb6.atl.mindspring.net>

In article <8F26D44DDjoshtinamecom at 209.208.0.7>,
Josh Tompkins <josht at iname.com> wrote:
>
>I've read in a couple of places that circular dependancies are not a good 
>thing to have in a program.  I'm actually kind of curious:  why is this?  
>Is it a style thing, or something that can actually confuse the 
>interpreter/compiler?

Circular references cause memory leaks.  That's it.  Now, if you want to
know why memory leaks are bad, that's another question...
--
                      --- Aahz (Copyright 2000 by aahz at netcom.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

"Laws that restrict speech are an abomination for a simple reason.
They're like a poison gas; they seem to be a good weapon to vanquish
your enemy, but the wind has a way of shifting."  -- Ira Glasser


From mwh21 at cam.ac.uk  Tue Apr 18 09:52:27 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 18 Apr 2000 14:52:27 +0100
Subject: RE question
References: <38FB9CF5.444457CF@pacific.jpl.nasa.gov>
Message-ID: <m366tfcy90.fsf@atrus.jesus.cam.ac.uk>

Benyang Tang <btang at pacific.jpl.nasa.gov> writes:

> >>> m = re.match("([abc])+", "abc"); m.group(1)
> 'c'
> 
> Why is it 'c', instead of 'a'?

On

    http://www.python.org/doc/current/lib/match-objects.html

we find:

   If a group is contained in a part of the pattern that matched
   multiple times, the last match is returned.

So that's why.

Cheers,
M.

-- 
  nonono,  while we're making wild  conjectures about the behavior 
  of completely  irrelevant tasks,  we must not  also make serious 
  mistakes, or the data might suddenly become statistically valid.
                                      -- Erik Naggum, comp.lang.lisp


From mwh21 at cam.ac.uk  Tue Apr  4 06:02:54 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 04 Apr 2000 11:02:54 +0100
Subject: Determining Classname of an object
References: <8ccbk4$b6p$1@nnrp1.deja.com>
Message-ID: <m3d7o6yyi9.fsf@atrus.jesus.cam.ac.uk>

mpetroni at my-deja.com writes:

> Is there a way to find out from which class an object has been created?
> Tanx!

instance.__class__?

HTH,
M.

-- 
well, take it from an old hand: the only reason it would be easier
to program in C is that you can't easily express complex  problems
in C, so you don't.                 -- Erik Naggum, comp.lang.lisp


From darcy at vex.net  Thu Apr 27 22:08:41 2000
From: darcy at vex.net (D'Arcy J.M. Cain)
Date: 28 Apr 2000 02:08:41 GMT
Subject: Best database to use with Python
References: <3906ea27.698463777@news.online.no> <8e9334$qba$1@news.tht.net> <39084a70$0$22983@personalnews.de.uu.net> <8ea025$s2c$1@news1.xs4all.nl>
Message-ID: <8earr8$277t$1@news.tht.net>

Boudewijn Rempt <boud at rempt.xs4all.nl> wrote:
> Last time I looked (five minutes ago, that is), there still wasn't
> a DB-API II compliant Python interface to PostgreSQL, which means

Any day now.  It's available as beta.

-- 
D'Arcy J.M. Cain <darcy at vex.net>           |  Democracy is three wolves
http://www.vex.net/                        |  and a sheep voting on         
+1 416 425 1212     (DoD#0082)    (eNTP)   |  what's for dinner.


From aycock at csc.uvic.ca  Thu Apr 27 01:49:52 2000
From: aycock at csc.uvic.ca (John Aycock)
Date: 27 Apr 2000 05:49:52 GMT
Subject: SPARK / Earley Parser question
References: <3907774d.32199072@news.btx.dtag.de>
Message-ID: <8e8ke0$4qb8$1@uvaix7e1.comp.UVic.CA>

Stefan Franke <spamfranke at bigfoot.de> wrote:
> I'm designing a grammar with John Aycock's SPARK toolkit. 

An excellent choice ;-) ;-)

> Are there some rules of thumbs which kind of rules should be avoided
> to retain linear complexity? 

The exact answer is in Earley's 1970 CACM paper.  As a rough guide, however,
an Earley parser has a fair bit in common with an LR parser, so LR grammars
with little in the way of lookahead required (i.e. things you might feed
to yacc or bison) will likely result in the best performance.  Lots of
ambiguity won't help matters either.  If blazing speed is a big issue
for your project, however, you probably don't want to be using SPARK; that's
not what it's designed for.

> In particular, I wonder if using empty right-hand sides for rules has
> any disadvantages.

No more so than for an LR parser.

I've also heard from people who've made a parser class in SPARK, then
subclassed it, rewriting selected rules in a less maintainable but
faster-to-parse fashion.  I don't have any personal experience doing
that, though.

John


From effbot at telia.com  Sun Apr  2 18:39:27 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Sun, 02 Apr 2000 22:39:27 GMT
Subject: NameError when assigning dictionary values
References: <20000402220351.9903.qmail@web2103.mail.yahoo.com>
Message-ID: <z_PF4.4786$74.76672@newsc.telia.net>

lewst <lewst at yahoo.com> wrote:
> > You're creating two empty lists, both named "value".The first
> > is thown away by creating the second. Nowhere are "one" and
> > "two" variable names, hence the NameError.
>
> Ok.  So now how do I get the program to do what I want?

dunno.  what *do* you want?

if you want to access the dictionary members you just
created, use dictionary syntax:

    print mydict["one"], mydict["two"]

if you want to create variables in the local namespace
on the fly, you're obviously trying to emulate some other
programming language (Tcl?) in Python.  that's usually a
rather bad idea.

(there are ways to create variables on the fly, but they're
all rather obscure and not very efficient -- so you don't really
want that.  but sure, if you can give me a very good reason...)

</F>




From robin at jessikat.demon.co.uk  Fri Apr 21 12:14:35 2000
From: robin at jessikat.demon.co.uk (Robin Becker)
Date: Fri, 21 Apr 2000 17:14:35 +0100
Subject: The Simple Economics of Open Source
References: <14591.15630.192472.440375@goon.cnri.reston.va.us>
 <slrn8fv2rd.1kt.neelk@brick.cswv.com> <3900611F.3B525BDA@cfdrc.com>
 <FtDG4z.774@world.std.com>
Message-ID: <uFaBgiAr5HA5EwJt@jessikat.demon.co.uk>

In article <FtDG4z.774 at world.std.com>, Will Ware <wware at world.std.com>
writes
>Ed (elb at cfdrc.com) wrote:
>> ... I didn't think much of the article...
>> ...they gave the impression that programmers behave
>> somewhat like monkeys, motivated primarily by concerns
>> about group status and dominance...
>
>They also mentioned, but immediately belittled, the possibility
>that altruism might be a real motivation. They may have been
>projecting their own thinking upon programmers in doing so. I'm
>not an open source purist, but I can easily see that a future world
>with a lot of free software will be a lot more pleasant for me to
>live in than one from which free software is absent. Benefits to
>the rest of society and to myself are not in conflict here. Maybe
>that's not really altruism, maybe it's just long-term thinking, a
>motivation they didn't mention at all.

I'm afraid your desire for a better world reduces your altruism
quotient.
Doing it because it helps others is altruism, doing it to help your self
and or relatives isn't really. The indian otters which help injured
friends were thought to be altruists until it was shown that by doing so
they reduced their own risk of predation (the slow one attracted the
predators most). Real altruism is rare.

I think the main self benefit that open sourcerers get is simple
enjoyment; rather like doing a cross-word with others. There may be side
benefits like getting kudos or a few million dollars as well. I would
hate to be Linus Torvalds at this point in time. I wonder how many
emails he gets each day; no wonder he now gets snappy with co-
developers.

I think arguing that we get better software is a bit spurious since it's
very difficult to compare things like operating systems or office
packages. It may well be true that we get more secure software as more
eyes get to see inside, but I don't think it's a guarantee that no
backdoors etc will be left in code. Few people other than applied
mathematicians will look at linear programming algorithms, many of the
mathematicians will 'turn off' when they come to deep magic (say related
to memory allocation or shared memory files or threading). In real
science, a rule of thumb is that 3 people read each paper (including
author and reviewers); perhaps one understands it.
-- 
Robin Becker


From michael.stroeder at inka.de  Sun Apr  9 10:58:47 2000
From: michael.stroeder at inka.de (Michael =?iso-8859-1?Q?Str=F6der?=)
Date: Sun, 09 Apr 2000 16:58:47 +0200
Subject: Receiving comp.lang.python.announce
Message-ID: <38F09AA7.B6ED18E1@inka.de>

HI!

like many other users I'm still receiving no articles in the group
news:comp.lang.python.announce since 1st of March, 2000.

Can some people who are receiving articles of 
news:comp.lang.python.announce recently post their Path: and
Message-Id: header of the articles to track the problems with the
news group distribution.

Thanks.

Ciao, Michael.


From bobalex at home.com  Mon Apr 10 13:05:40 2000
From: bobalex at home.com (Bob Alexander)
Date: Mon, 10 Apr 2000 10:05:40 -0700
Subject: Tuples -- who needs 'em
References: <AF223B12F1FDD2119253006097C4FCEC0AD357@cats.pdri.xerox.com> <8chol6$6em$1@newshost.accu.uu.nl> <026e01bf9fe7$9554fd20$74eb0b18@stcla1.sfba.home.com> <8cofcf$hns$1@newshost.accu.uu.nl>
Message-ID: <007801bfa30e$ffe79560$74eb0b18@stcla1.sfba.home.com>

"Martijn Faassen" <m.faassen at vet.uu.nl> wrote:

Thanks for your well-considered comments, but we still have our differences
:-)

> the idea of immutable
> objects is that assigning them is exactly the same behavior as copying
> them. That's something an easy thing to think about; for instance with
> strings and numbers. The idea here is that tuples allow you to think that
> way about lists.

Well, immutable objects render the copy vs. reference semantics moot, since
you can't change them anyway. With true copy semantics, the receiver of the
copy can change the object without affecting the originator's object. But I
agree that you can safely pass around references to immutable objects
without fear of their being modified -- I guess that's the point of
immutable.

> Okay, but it's generally practiced in idiom. Language constructions can
> exist that partially support idiom, but not enforce it, right? Here the
> idea is to bind various immutable things together to something that's
> still immutable. The immutability of non-tuples isn't there to
> catch 'mutation' errors (as that there is no way to actually mutate them
> at all). It's there for the nice semantics. Tuples supply the syntax and
> semantics to do this with multiple immutable things. Your problem arises
> because there's something very similar to a tuple, a, list, which does
> support mutation syntax. Would you have the same trouble with immutable
> strings if there were mutable strings in the language? And why not?

You might ask, "why doesn't Python have both mutable strings and
immutable?". Some languages (like Java) do. One answer: having both flavors
doesn't buy us much, even though there might be an occasional use for a
mutable string. I'm glad we don't have both flavors of string. Similarly, I
don't think we need both mutable and immutable sequences. But in the case of
sequences, it's the immutable flavor that doesn't buy us much.

By the way, for those who contend that tuples are a good way to pass around
groups of heterogeneous elements, I contend that they have some drawbacks.
Take the example of "point" that was previously used in this thread. Create
a point p = (11, 22). How do I refer to the elements? As p[0] and y[1]. Hmm,
which is x and which is y? Wouldn't it be better if I could reference the
elements by name, x.x and x.y? Some languages have record types where you
can declare "record Point(x, y)", create an object as Point(11, 22), and
refer to the elements as p.x and p.y. A language supporting this is Icon,
and records are roughly as efficient as Python's tuples even though the
elements have names. Even in JavaScript you can construct something like a
tuple (but not immutable) with named elements using the same syntax as
Python dictionary "literals" {x: 11, y: 22}. Although I clearly prefer
Python as a language over JavaScript, that is a nice feature. Of course, you
can effect the same behavior in Python, but not so conveniently. I think
that tuple-type objects with named elements are superior to numbered
elements for groups of heterogeneous elements, and that immutability is not
often necessary and even sometimes an inconvenience.

> But the idea here is that heterogeneous sequences of a small number of
> immutable things (which behave with a semantics indistinguishable of
> pass-by-value) should also be such a thing. Python passes everything
> by reference. But for some things, such as numbers and strings,
> pass-by-value is actually more interesting. So there we get immutability
> to support that.

Yes, it's true that if you pass a collection to a function and don't change
it, it's much like pass-by-value.

> I do think the Python language supports the homogeniety idea. Just because
> Python supports heterogeneous lists doesn't mean you should be using them
> all over the place.

But, why not? How is it damaging? When I first discovered a language that
did not require homogeniety (is that a word?) in lists, I though of it as an
advantage, not a drawback. (That was Snobol4, a *long* time ago!) I still
feel that way.

Some of my following responses reflect the fact that I haven't bought into
the if-a-sequence-is-a-list-it-must-be-homogeneous concept.

> > As you say, the functional programming paradigm can be nicely expressed
in
> > Python, and is not dependent on tuples.
>
> Some of it can be nicely expressed, some can't be, and as you say, it's
> not dependent on tuples. But you mentioned you knew no languages with
> the concept of immutable lists, and I mentioned functional programming
> languages. I did already say Python doesn't work this way, so you don't
> have to tell me what I just told you. :)

Oops, sorry. I do that sometimes :-)

> > There have been cases where I used tuples, to find out
> > later that I wanted mutability, so went back and changed.
>
> Hm, I rarely if ever seem to run into this. If I need mutability of
> a set of heterogenous objects I tend to use a class. There may be a
> switch from tuples to classes, but not often from tuples to lists, I
> think.

Well, since I don't have an aversion to heterogeneous lists, it's kind of
the same thing  :-) Yes, the proper choice is class, but a small list (or
tuple) is much more convenient (for the coder, BTW, not the reader :-). And,
after all, Python is a "rapid development" language where convenience is
highly valued. Just don't forget which element number is which! Or if you
change the order later, don't forget to change all the references.

> Perhaps you shouldn't think 'tuples or lists', but 'tuples or classes',
> then?

If tuple is a super-convenient-quickie substitution for class, then, to me,
list is just a mutable version of the same concept. (And neither of those is
as expressive as using a class.) You're suggestion is change tuple-to-class,
in my example I changed tuple-to-list. Either way is a different means to
the same end (yours is more correct, but more laborious). One might start
out using tuple not realizing that he/she would later want mutability for
that particular use. Then he/she has to change existing code. In a language
focused on rapid development, that is not good. I guess the moral is: if
you're not 100% sure, use list :-)

And if there were only one tuple/list datatype, we'd never have to decide
about it during programming, or go back and change it when we don't
anticipate correctly, or explain the difference to Python newbies, or have
FAQ items about it, or have discussions such as this :-)

-- Bob





From yawyi at gmx.de  Tue Apr  4 20:43:19 2000
From: yawyi at gmx.de (Johannes Stezenbach)
Date: Wed, 5 Apr 2000 02:43:19 +0200
Subject: Warning switches (was: Python 1.6 alpha 1 released)
References: <Pine.GSO.4.10.10004040035420.21084-100000@sundial> <5E877FD7637AB46E.DABBED9CF3C1D64B.CDDDC05AE7374A78@lp.airnews.net> <5l7leern5v.fsf@eric.cnri.reston.va.us> <38ea1de0$0$17688@personalnews.de.uu.net> <nctG4.5133$74.83090@newsc.telia.net>
Message-ID: <38ea8c42$0$28633@personalnews.de.uu.net>

Fredrik Lundh <effbot at telia.com> wrote:
>Johannes Stezenbach <yawyi at gmx.de> wrote:
>> Will python then support a -w switch to support people in writing
>> correct programs (i.e. compatible with likely furure language/library
>> changes), just like our-favourite-four-letter-P-language?
>
>how about -v<version> ?
>
>there has been talks about compatibility "modes" (probably an option,
>but probably also a way to set options programmatically), as a way to
>help people prepare for "Python 3000".  I think we might expect some
>of it in 1.7, and more of it in 1.8 (if that ever materializes).
>
>but hopefully, Python 3000 will be completely free from modes.

The next word I had written should have alarmed you that I was joking:

>> Seriously:

I really don't want -w. -t is already enough for me.


>> Just postponing the change is evil since it will only hurt more people
>> in the future, as the number of python users is growing. (CP4E?)
>
>yeah, but at the moment, guido and the rest of the 1.6 team have
>more important things to do than to deal with the spanish inquisition.
>so I'm afraid it'll have to wait until that -v1.6 switch is in place.

Well, this thread certanly got a bit lenghty but I wouldn't call
it "the spanish inquisition". And even if Guido's mailbox overflowed
from complaints, what's there to gain in postponing the connect()
change? He will just get all the complaints again in 1.7, and in
numbers which increase with the number of python users.
And I don't think that claiming "connect(host, post) is deprecated
and won't work in 1.7" in the documentation is a Good Thing,
because there is a lot of code "out there" (e.g. python.org,
parnassus, the starship, ...) which people will download and
use as examples or building blocks for their own stuff -- and
it's all broken and will stop working in 1.7!
My point is simply: Do it now or never [1]. The longer you wait the
worse it gets.

And just to waste some more bandwith let me repeat (yawn) what has
been said before by others in this group:
a) There are much more changes in python 1.6 which will break user code.
b) The socket.connect() stuff does most likely affect only a few
   lines for most python users, and these are most likely easy to
   find and fix.
c) It's still months until 1.6final will be out.
d) No one is forced to upgrade immediately.

>> If it's not broken then don't fix it.
>
>    "I try to avoid breaking user programs, but if something
>    needs to be done it really needs to be done."
>    -- linus torvalds

Again I feel you misunderstood me, maybe because I snipped too much
of Guidos posting. I object to Guido "fixing" the code so that
the broken 2-arg connect() syntax becomes valid again. I'm more
with Linus' quote.

-eot


1) The "never" option would be OK for me. I'm not a language lawyer.



From bjorn at roguewave.com  Tue Apr 11 12:24:16 2000
From: bjorn at roguewave.com (Bjorn Pettersen)
Date: Tue, 11 Apr 2000 10:24:16 -0600
Subject: [ANNOUNCE] Garbage collection for Python
References: <20000407221750.A29279@acs.ucalgary.ca> <38F13FB9.D1D16AF@mindspring.com> <slrn8f2if1.dop.nascheme@cranky.arctrix.com> <87r9cdhdb0.fsf@deneb.cygnus.argh.org> <FsunGI.BtB@world.std.com> <38F319A4.89621192@mdlink.de>
Message-ID: <38F351B0.45E137AA@roguewave.com>

Helge Hess wrote:
> 
> Will Ware wrote:
> > Florian Weimer (fw at deneb.cygnus.argh.org) wrote:
> > > Is it possible to extend your GC in a way which would permit the
> > > execution of several bytecode interpreters in multiple threads?
> >
> > It sounds like you might want to simply run several copies of the
> > Python interpreter in separate processes, which isn't a GC issue.
> 
> I think what he wants to do (me too ;-) is to run several copies of the
> interpreter in *one* process, possibly in multiple threads.
> But I'm not sure whether this is a GC issue though (RC isn't thread-safe
> anyway, right ?).

Well, it actually can be -- at least on Windows :-)

Windows have system calls for atomic increment/decrement
(InterlockedIncrement/InterlockedExchangeAdd), something that is sorely
lacking from Unix (*sigh*)

-- bjorn



From gresham at mediavisual.com  Fri Apr  7 12:37:05 2000
From: gresham at mediavisual.com (Paul Gresham)
Date: Sat, 8 Apr 2000 00:37:05 +0800
Subject: CGI extension problem
References: <UX3UiBAPET54Ewi5@jessikat.demon.co.uk> <87ya6q5fqo.fsf@womble.dur.ac.uk>
Message-ID: <8cl2rk$kfn4@imsp212.netvigator.com>

How about this, taken directly from my own python CGI scripts, I have all my
own python modules in a sub-directory called python, under my cgi-bin:

#!/usr/bin/python

#System imports
import cgi
import sys
import traceback

sys.stderr=sys.stdout
sys.path.insert(0, "./python")

Thats real simple, so maybe I've missed your point !

Regards
Paul

"Andrew Stribblehill" <a.d.stribblehill at dur.ac.uk> wrote in message
news:87ya6q5fqo.fsf at womble.dur.ac.uk...
> Robin Becker <robin at jessikat.demon.co.uk> writes:
>
> > I don't have root privilege, my python cgi script needs access to an
extension which I have
> > built; the extension in turn needs a shared lib which I have built. I
install in a directory say
> > $HOME/lib. When I run the cgi script I need to set the LD_LIBRARY_PATH
to at least include
> > $HOME/lib. I have use the following idiom
> >
> > import sys, os
> > HOME_LIB='/home/me/lib'
> > if sys.platform != 'win32' and not
os.environ.has_key('LD_LIBRARY_PATH'):
> >         os.environ['LD_LIBRARY_PATH'] = HOME_LIB
> >         os.execve(sys.argv[0],sys.argv,os.environ)
> >
> > # rest of cgi script
> > .....
> >
> > I can think of a few ways in which this might break, but is there
anything obviously easier less
> > breakable?
>
> Wouldn't it be better to use LD_RUN_PATH when /compiling/ the
> extension? Then it won't need a LD_LIBRARY_PATH variable at
> runtime.
>
> Obviously you then have to make sure that this location stays
> the same else you'll need to resort to the LD_LIBRARY_PATH hack (or
> recompile) again!
>
> Cheerio,
>
> Andrew Stribblehill
> Systems Programmer, IT Service, University of Durham, England




From m.faassen at vet.uu.nl  Thu Apr 27 08:36:45 2000
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 27 Apr 2000 12:36:45 GMT
Subject: Python 2.0
References: <3907B52D.B01D5748@myself.com>
Message-ID: <8e9c8t$hi6$1@newshost.accu.uu.nl>

Wayne Izatt <wayne.izatt at myself.com> wrote:
[will Python 2.0 cause a maintenance problem]
> So, can anyone give me any advice? Is there any cause for alarm?

First of all, the 1.x line of development is definitely good for a few
years yet at least. I doubt Python 1.6 will be superceded by Python 1.7
within a year, and Python 1.6 isn't even out yet. Python 1.7, which is
being planned, is surely going to last at least a year more. That brings us
to the summer of 2002. That's centuries in computing-history time. 2 years
ago I don't even think C++ was standardized yet, for instance. Or just take
Java back in '98. That should put things in perspective; Python's development
tends to be cautious, the current code base is good for years yet.

Now on to Python 2.0. It's been said before on this newsgroup that while
Python 2.0 is expected to break the C API for Python, it's not thought it'll
break a lot of code written in Python. Perhaps they know more than I do, but
I myself am a bit more dubious about this for two reasons:

  * it might be *good* for the language to ignore backward compatibility for
    a change. Things that are broken now can be fixed for good. Not that
    a lot of things are broken, but there are some candidates for improvement.
    lambdas for instance. We *want* to break existing code. :)

  * we only know a few things that are thought about for Python 2.0, and if
    these are implemented we'd get large backwards incompatibility problems:

      * integer division giving float results. 1/2 might suddenly give us 0.5

      * case insensitivity. foo = Foo() would break
 
    Both of these are catering to beginners. Opinions on them vary; 
    they may reek a bit like Do-What-I-Mean, but nobody thinks either change
    would be a disaster. But both proposed changes would definitely
    break tons of existing Python code.

That said, there will probably be at least some code migration tools. They can
probably not be perfectly automated in all cases; humans will be necessary.

Concluding my analysis:

  * current Python codebase is good for almost the indefinite future. A mature
    Python 2.0 replacing Python 1.x is years away.

  * Python 2.0 is bound to break quite a bit of existing code, when it 
    happens.

  * No cause for alarm yet.

Someone should put this question into a FAQ or something -- it's being asked
quite frequently these days.

Regards,

Martijn 
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From bobalex at home.com  Thu Apr  6 12:45:57 2000
From: bobalex at home.com (Bob Alexander)
Date: Thu, 6 Apr 2000 09:45:57 -0700
Subject: Tuples -- who needs 'em
References: <AF223B12F1FDD2119253006097C4FCEC0AD357@cats.pdri.xerox.com> <8chol6$6em$1@newshost.accu.uu.nl>
Message-ID: <026e01bf9fe7$9554fd20$74eb0b18@stcla1.sfba.home.com>

"Martijn Faassen" <m.faassen at vet.uu.nl> wrote:

> Except that you can suddenly do this:
>
> results = foo()
> ...
> results.append(5)
> ...
> first, second, third, fourth = results # error

True -- but there is a simple solution: omit the second statement  :-)
There are things you can do to sabotage things even with tuples; for example
replace "results.append(5)" with "results = (1,2)". Is mysterious changing
of lists a frequently-occurring problem?

> A tuple tends to contain a few heterogenous things,
> while a list tends to contain a larger number of homogenous things.

This distinction seems to be in the minds of several in this thread, but is
not mentioned in any Python doc I've seen, is not enforced by the language,
and, IMO, is unnecessary. In languages with mutable lists but no tuples,
lists are often used for spiritually-unmutable sequences of a few
hetrogenious things. I've never though of the syntactic distinction as
necessary or particularly helpful.

I think the "homogeniety" idea is a carryover from the many languages that
do not permit heterogeneous sequences. Homogeneous sequences should be used
where appropriate, but there are uses for heterogeneous sequences too.
Flexibility is good.

> Functional programming languages have the concept. At least the
> *spirit* that you shouldn't be changing a list. This is different from
Python,
> as we do change (homogenous) lists. The pattern in functional programming,
> which *does* occur often in Python code as well, is some operation over
> all elements of a list creating a new list (map(), though this is often
> expressed in Python as a 'for' loop, partially due to the nastiness of
> 'lambda'). The old list isn't changed.

As you say, the functional programming paradigm can be nicely expressed in
Python, and is not dependent on tuples.

> > And I still haven't
> > heard from anyone about why there isn't similar concern about immutable
> > dictionaries.
>
> People tend to use classes for this purpose. You can give a class a public
> interface (enforced or not) with only get_() methods, for instance. That
> gets you just about the same thing as an immutable dictionary would give
> you. Again, a dictionary tends to contain a larger number of homogenous
> objects, while an object tends to contain a smaller number of heterogenous
> objects.

That same technique could be used to argue against the need for tuples. Of
course, that's not often desirable in casual programs because it's somewhat
painful, and not all that necessary.

> Do you really feel that pain, then? Why don't you always use lists in your
> Python code, if you want to avoid the decision? That'd be effectively the
> same as what you're proposing

Well, it hasn't achieved the threshold of "pain", but I do seem to be using
lists in preference to tuples more often than I used to. And I still
speculate that life would be a bit better if I never had to think about
tuples vs. lists. There have been cases where I used tuples, to find out
later that I wanted mutability, so went back and changed. This sort of thing
happens in exploratory programming, so when in that mode one might want to
use lists unless one is really sure in advance that tuples are the right
thing.

By the way, when I get a run time error about trying to modify a tuple, it
never seems to be because I'm changing something I shouldn't. It's always
because it really should of been a list in the first place.

> If you don't do this, is this because you're
> afraid other programmers will complain about your code, or because you do
> believe there's something to decide about?

There are other possible motivations besides fear of criticism, such as the
desire to observe the spirit of the language. In doing that, it has occurred
to me that tuples might not provide enough benefit to warrant the additional
cognitive load. So far there have been zero responses in favor of my point
of view, so I'm in a very tiny minority  :-)

> If-everything-were-blocks-of-bytes-we-would-never-need-to-decide-ly yours,

Then we would have tcl :-)

Bob





From nmm1 at cus.cam.ac.uk  Sat Apr  8 04:38:56 2000
From: nmm1 at cus.cam.ac.uk (Nick Maclaren)
Date: 8 Apr 2000 08:38:56 GMT
Subject: Python paradigms
Message-ID: <8cmr70$b0e$1@pegasus.csx.cam.ac.uk>

Well, starting to use Python, there are two Algol 68 constructions
that I sorely miss.  Here is an example of them in C:

    if ((x = non_trivial_expression) != NULL) {
        do_something_using_x;
    } elif ((x = another_expression) != NULL) {
        do_something_else_using_x;
    } ...

and:

    x = (a != NULL ? a[i]->weeble : 0) + (b != NULL ? b[i]->wombat : 0)

It is pretty obvious why Python doesn't have the former, but I am
a bit puzzled as to the reason for the lack of the latter.  Or
maybe I just haven't found it yet :-)

However, my main point is that the above constructions clarify some
types of code a great deal.  Creating ad-hoc functions is all very
well, but too many of them (especially trivial ones, used once)
obscure the program considerably.  And please note that the above
constructions are NOT an artifact of my coding style - in what I
am doing, they arise naturally out of the problem.

What paradigms are there for replacing such constructions by?


Regards,
Nick Maclaren,
University of Cambridge Computing Service,
New Museums Site, Pembroke Street, Cambridge CB2 3QG, England.
Email:  nmm1 at cam.ac.uk
Tel.:  +44 1223 334761    Fax:  +44 1223 334679


From loewis at informatik.hu-berlin.de  Tue Apr 18 13:08:31 2000
From: loewis at informatik.hu-berlin.de (Martin von Loewis)
Date: 18 Apr 2000 19:08:31 +0200
Subject: How to Copy a Dictionary from a C Extension?
References: <JklJ4.6736$HG1.192758@nnrp1.uunet.ca>
Message-ID: <p6qem83nxps.fsf@informatik.hu-berlin.de>

"Warren Postma" <embed at geocities.com> writes:

> Problem: If you want to copy the existing Globals used by the interpreter,
> what is the way to do this?  I found pydict_copy in dictobject.c, but it's
> static, so I can't invoke it directly.

I guess you can use PyObject_CallMethod, and give "copy" as the method
name. This is a couple of indirections more than needed, but should
work right away.

Regards,
Martin



From maybe at lucky.com  Tue Apr 11 17:05:11 2000
From: maybe at lucky.com (Soundwave)
Date: Tue, 11 Apr 2000 23:05:11 +0200
Subject: AF_INET error
References: <1256641280-42160733@hypernet.com>
Message-ID: <8d040e$17s$1@porthos.nl.uu.net>

Thanx, it was a file called socket.py, witch I had made myself, stupid of
me, I could have thought of it myself.

sorry for bothering you guys

thanx
Gordon McMillan <gmcm at hypernet.com> schreef in berichtnieuws
1256641280-42160733 at hypernet.com...
> Soundwave wrote:
>
> > On www.python.org is a HOWTO guide for creating sockets, i'll give an
> > example from what is said in the guide
> >
> > This is the first example
> >
> > import socket
> > #create an INET, STREAMing socket
> > s = socket.socket(
> >     socket.AF_INET, socket.SOCK_STREAM)
> > #now connect to the web server on port 80
> > # - the normal http port
> > s.connect("www.mcmillan-inc.com", 80)
> >
> > when I run this code, I get the following output
> >
> > Traceback (innermost last):
> >   File "C:\PROGRA~1\PYTHON\TOOLS\IDLE\ScriptBinding.py", line 131, in
> > run_module_event
> >     execfile(filename, mod.__dict__)
> >   File "C:/Program Files/Python/socket.py", line 3, in ?
> >     s = socket.socket(
> > AttributeError: AF_INET
>
> It seems you have a module named socket.py which is being
> imported instead of the standard one (which would be
> Lib/socket.py).
>
> - Gordon
>




From kern at caltech.edu  Tue Apr 18 18:14:38 2000
From: kern at caltech.edu (Robert Kern)
Date: 18 Apr 2000 22:14:38 GMT
Subject: Date class?
References: <8dikk2$p4s@flatland.dimensional.com>
Message-ID: <8dimoe$t0o@gap.cco.caltech.edu>

In article <8dikk2$p4s at flatland.dimensional.com>,
	jtc at dimensional.com (Jim Cochrane) writes:

[snip]

> Concidentally, I just read here that the starship server just died; and
> that appears to be the only place where mxDateTime is available.  Does
> anyone know of another site that has the mxDateTime package?

The Debian GNU/Linux distribution has it.  It's the package
"python-mxdatetime".

RPMs can be found from one of Oliver Andrich's mirrors, namely:
ftp://ftp.rhein-zeitung.de/comp/Python/

There are source RPMs there, and there are tools to unpack RPMs on Windows
systems.  If people need Windows binaries, e-mail me, and I'll work on it.

> Thanks!

-- 
Robert Kern
kern at caltech.edu

"In the fields of hell where the grass grows high
 Are the graves of dreams allowed to die."
  -- Richard Harter


From j.spies at hccnet.nl  Wed Apr 26 17:44:44 2000
From: j.spies at hccnet.nl (Jaap Spies)
Date: Wed, 26 Apr 2000 23:44:44 +0200
Subject: translation of troff text to latex
Message-ID: <3907634C.5B166424@hccnet.nl>

Hi,

Is there anybody out there, who has written some Python code
to translate good old troff (mm) to latex?????

Jaap Spies

Keep Peace in Mind


From amitp at Xenon.Stanford.EDU  Tue Apr  4 15:13:55 2000
From: amitp at Xenon.Stanford.EDU (Amit Patel)
Date: 4 Apr 2000 19:13:55 GMT
Subject: Paid to work in Python? (was: Re: Choice Of Language)
References: <NDBBJPNCJLKKIOBLDOMJMEMECDAA.DavidA@ActiveState.com> <8c7tt5$7dl$1@slb3.atl.mindspring.net>
Message-ID: <8cdetj$310$1@nntp.Stanford.EDU>

 Aahz Maruch <aahz at netcom.com> wrote:
| In article <NDBBJPNCJLKKIOBLDOMJMEMECDAA.DavidA at ActiveState.com>,
| David Ascher <DavidA at ActiveState.com> wrote:
| >Grant Edwards
| >>
| >> Pointless tanget: Is it just me, or is Google really the last usable
| >> search engine?
| >
| >And it's _all_ due to their use of Python!  Cool, eh? =)
| 
| That doesn't explain why Infoseek is no longer particularly usable.

Didn't they switch to C / Java?

  http://www.skirsch.com/javase/Javainfo.htm

  http://www.wired.com/news/news/technology/story/15989.html

(I'm not sure.)

- Amit


From tismer at tismer.com  Fri Apr 28 16:32:23 2000
From: tismer at tismer.com (Christian Tismer)
Date: Fri, 28 Apr 2000 22:32:23 +0200
Subject: What does "first class" mean?
References: <3909F2F5.A6F91256@spacenet.tn.cornell.edu>
Message-ID: <3909F557.71C12A49@tismer.com>


Tom Loredo wrote:
> 
> Hi folks-
> 
> I've often seen reference here (and on the Ruby page) to Python's
> treatment of functions and methods as "first class objects."

My understanding:
It means that they are first class citizens, there is no
difference to any other object type. They are allowed in
the same context as other objects of other types.

Example:   you can write

  print f

and Python will try to print this object, regardless
whether it is an int, a string, a function, a traceback, a class,...

Notabene the entry "a function":
f is not called, since it is a first class object.
In order to call f, you need the call syntax, and Python will
try to call the object, if there is a way to do so.

VB for example would always try to call f, with or without
parentheses.

ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaunstr. 26                  :    *Starship* http://starship.python.net
14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     where do you want to jump today?   http://www.stackless.com



From jae at ilk.de  Thu Apr  6 15:51:53 2000
From: jae at ilk.de (Juergen A. Erhard)
Date: Thu, 06 Apr 2000 21:51:53 +0200
Subject: Teaching the art of programming, in python
In-Reply-To: <1e8mov8.1yh8ryv1kp6z0lN%bparsia@email.unc.edu>
References: <20000405053920.20579.qmail@web2102.mail.yahoo.com> <8cfe42$o02$1@nnrp1.deja.com> <dpJG4.3744$HG1.106895@nnrp1.uunet.ca> <1e8mov8.1yh8ryv1kp6z0lN%bparsia@email.unc.edu>
Message-ID: <06042000.4@sanctum.jae.ddns.org>

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

>>>>> "Bijan" == Bijan Parsia <bparsia at email.unc.edu> writes:

    Bijan> Warren Postma <embed at geocities.com> wrote:

    Bijan> [snip]
[...]
    >> Remember the famous 'Smalltalk-80 - The Language and Implementation'
    >> purple-book?
    Bijan> [snip]

    Bijan> Pedantic correction: The Language and Implementation version is the
    Bijan> *blue* book. [...]

Pedantry time: I remember it as

  Smalltalk-80: The Language and it's Implementation

(though I haven't read it... yet)

Bye, J

- -- 
J?rgen A. Erhard      eMail: jae at ilk.de      phone: (GERMANY) 0721 27326
          My WebHome: http://members.tripod.com/Juergen_Erhard
           "Outside of a dog, a man's best friend is a book;
           inside of a dog, it's very dark." --  Groucho Marx
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.1 (GNU/Linux)
Comment: Use Mailcrypt and GnuPG <http://www.gnupg.org/>

iEYEARECAAYFAjjs6tgACgkQN0B+CS56qs2fJwCgn2vKztFbPdPhBI72KXzmuR5s
BZsAn2FVCEGYJfdgmKBImnWkWXSOOkPt
=vOMV
-----END PGP SIGNATURE-----



From scarblac-spamtrap at pino.selwerd.nl  Thu Apr 13 10:34:01 2000
From: scarblac-spamtrap at pino.selwerd.nl (Remco Gerlich)
Date: 13 Apr 2000 14:34:01 GMT
Subject: List mapping?
References: <38f5d68e@news.xtml.co.uk>
Message-ID: <slrn8fbpvv.1d0.scarblac-spamtrap@flits104-37.flits.rug.nl>

Nick Trout wrote in comp.lang.python:
> while i < len(listargs):
>     tuplist.append( (listargs[i],listargs[i+1]) )
>     i = i+2

> Can I (you!) get rid of the long winded while loop cleverly?!!  :-)

That isn't long winded, in my opinion. And not hard to read.

But

for i in range(0, len(listargs), 2):
   ...
   
Is one line shorter. I don't think there's a builtin function for making a
list into tuples like that.

-- 
Remco Gerlich,  scarblac at pino.selwerd.nl
  5:30pm  up 38 days,  4:50,  8 users,  load average: 1.27, 1.42, 1.48


From mhammond at skippinet.com.au  Thu Apr 13 23:13:08 2000
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Fri, 14 Apr 2000 03:13:08 GMT
Subject: Getting file version information
References: <20000413163629.28006.qmail@web4303.mail.yahoo.com>
Message-ID: <81wJ4.811$1D.5094@news-server.bigpond.net.au>

It aint there.  One was promised to me once, but never arrived.  Ive never
needed it personally :-)

You may be forced to using calldll (www.nightmare.com) to call the API
functions directly.

Mark.

"Daniel Zapata" <danzap_python at yahoo.com.br> wrote in message
news:20000413163629.28006.qmail at web4303.mail.yahoo.com...
> I?d like to know how to get the file version
> information in Windows. I?m using pythonwin and
> couldn?t find the GetFileVersionInfo API on the
> win32api module.
>
> Thanks
>
> Daniel Zapata
>
>
__________________________________________________________________________
__
> YAHOO! BATE-PAPO
> Conhe?a gente nova, fa?a amigos e divirta-se! - http://www.yahoo.com.br
>




From ns566645 at swt.edu  Mon Apr 10 19:42:51 2000
From: ns566645 at swt.edu (Shah, Navneet)
Date: Mon, 10 Apr 2000 19:42:51 -0400
Subject: Frames in TK
Message-ID: <38F266FB.2363E980@swt.edu>

Can anyone let me know the code for making frames in TK. I need to have
two frames in the window with the top frame should cover 65% of the
window while the remaining 35% should be covered by the second frame.





From kilian.lorenz at dkfz-heidelberg.de  Sun Apr  9 17:19:56 2000
From: kilian.lorenz at dkfz-heidelberg.de (Kilian Lorenz)
Date: Sun, 09 Apr 2000 23:19:56 +0200
Subject: Tkinter:  centering windows on screen
References: <ak1I4.121418$Dv1.1306619@news1.rdc1.bc.home.com> <38F0C6A7.10878609@dkfz-heidelberg.de> <38F0C95D.1F7E853@dkfz-heidelberg.de> <xR4I4.274$rc9.188784640@newsb.telia.net> <8U5I4.121484$Dv1.1314709@news1.rdc1.bc.home.com>
Message-ID: <38F0F3FC.5ADB3DA4@dkfz-heidelberg.de>

Kevin Cazabon wrote:
> 
> Is there a Python or Tkinter call to return the screen resolution?  I know I
> can probably get it through Win32, but I'd rather make it cross-platform...
> I'll do some digging.
> 

try these:

winfo_screenwidth, winfo_screenheight

http://w1.132.telia.com/~u13208596/tkintrobook/
x9548-window-related-information.htm

:)


From M_Pope at shef.rage.co.uk  Wed Apr 19 04:15:39 2000
From: M_Pope at shef.rage.co.uk (Mark Pope)
Date: Wed, 19 Apr 2000 09:15:39 +0100
Subject: dir(..)   vs.   PyList_GetItem
Message-ID: <213CDEBFDDA7D3118081009027E40C56071159@SHEF-MS1>

The code I posted was just a snippet.
I have already successfully imported the module. I now need to get a list of
the functions it exposes.
I want the equivalent of calling 'dir(myScript)' for the C API

> -----Original Message-----
> From: Greg Ewing [mailto:greg at cosc.canterbury.ac.nz]
> Sent: 19 April 2000 5:04 AM
> To: python-list at python.org
> Subject: Re: dir(..) vs. PyList_GetItem
> 
> 
> Mark Pope wrote:
> > 
> >         pModule = PyModule_New( "myScript" );
> 
> That doesn't import the module, it just creates an 
> empty module. You want
> 
>    pModule = PyImport_ImportModule("myScript");
> 
> -- 
> Greg Ewing, Computer Science Dept,
> +--------------------------------------+
> University of Canterbury,	   | A citizen of 
> NewZealandCorp, a	  |
> Christchurch, New Zealand	   | wholly-owned subsidiary of 
> USA Inc.  |
> greg at cosc.canterbury.ac.nz	   
+--------------------------------------+
> -- 
> http://www.python.org/mailman/listinfo/python-list
> 



From noreply at nospam.org  Thu Apr 20 12:03:46 2000
From: noreply at nospam.org (Brian)
Date: Thu, 20 Apr 2000 16:03:46 GMT
Subject: Why MFC instead of Tkinter -- was Re: Pythonwin?
References: <eMz2Nmiq$GA.73@net025s> <4QtL4.24$v85.41@news-server.bigpond.net.au> <chuL4.8483$O7.17315@typhoon.southeast.rr.com> <7LDL4.534$v85.2563@news-server.bigpond.net.au> <sfu6d657l6e184@corp.supernews.com>
Message-ID: <CTFL4.41$a6.62889@paloalto-snr1.gtei.net>

You should also maybe give wxPython a go.  Run the demo, it's selection of
widgets is very impressive!
It's also cross-platform, it's fast and it uses native widgets where it can.
http://alldunn.com/wxPython/
http://www.wxwindows.org/

"Michael A. Crawford" <michaelc at flashcom.net> wrote in message
news:sfu6d657l6e184 at corp.supernews.com...
> I'm curious.  Why did you not steer him in the Tkinter direction for his
GUI
> needs?  Is MFC better?  I learned (and forgot MFC) some time ago, years in
> fact.  Same with X-Windows programming.  Now I'm working my way throught
> "Python and Tkinter Programming" by Grayson.  I thought Tkinter was the
> preferred method for Pythong GUI programming.  Am I mistaken?
>
> -Michael
>
> "Mark Hammond" <mhammond at skippinet.com.au> wrote in message
> news:7LDL4.534$v85.2563 at news-server.bigpond.net.au...
> > "Don Tuttle" <tuttledon at hotmail.com> wrote in message
> > news:chuL4.8483$O7.17315 at typhoon.southeast.rr.com...
> > > [Mark]
> > > > You are really going to need to go through an MFC tutorial or book
> > before
> > > > you can make heads or tails of Pythonwin from that level...
> > >
> > > Any recommendations?
> >
> > Afraid not.  I already new MFC when I started Pythonwin.  I learnt it
from
> > the MFC "scribble" tutorial.  I find I tend to learn best with a
tutorial
> > and then playing - that is what I did with Python (back then there was
no
> > other alternative, but even now that is my recommended strategy for
> > competant existing programmers learning Python...)
> >
> > So assuming some C++ knowledge, my suggestion for MFC would still be
> > scribble tutorial, then invent an app to write in MFC C++, then port
that
> > app to Python.  By then you will be a wiz :-)
> >
> > Mark.
> >
> >
> >
>
>




From psl at mitre.org  Tue Apr 18 10:49:24 2000
From: psl at mitre.org (Pete)
Date: Tue, 18 Apr 2000 10:49:24 -0400
Subject: focus with PMW
Message-ID: <38FC75F4.D4E01F19@mitre.org>

	I am trying to determine which widget,  the focus is currently on in my
code.  When I make a call to get_focus() it gives me an instance of a
TKinter.Entry,  but I am only creating PMW widgets (which are of course
extended from TKinter).  Since I am not creating TKinter widgets
explicitly in my own code,  how can I determine which PMW widget that
specific TKinter.Entry instance belongs to?


From dln at jump.net  Tue Apr 25 13:12:11 2000
From: dln at jump.net (Dick Norton)
Date: Tue, 25 Apr 2000 12:12:11 -0500
Subject: Python binary for win 3.1 with threads, where?
Message-ID: <3905D1EB.3BEABA69@jump.net>

Hello,

Is there a python for windows 3.1 that already has
threads turned on?  I've been unsuccessful in cross building
(NT to w 3.1) to enable threads.

Alternatively, is there a non-blocking keyboard reader for
w 3.1.  I've given up on kbhit at this point.

Thanks,
Dick Norton


From meh9 at cornell.edu  Mon Apr  3 15:16:03 2000
From: meh9 at cornell.edu (Matthew Hirsch)
Date: Mon, 03 Apr 2000 15:16:03 -0400
Subject: Creating variables on the fly...
Message-ID: <meh9-25D0C5.15160303042000@news.cit.cornell.edu>

Hi All,

Let's say I have four variables: 

a=1
b=2
c=3
d=4

And I have a function that adds these variables together:

def add(a,b,c,d):
   return a+b+c+d

But now let's say I have twenty variables that I want to add together.  
This function will no longer work.  I would have to rewrite it as 
a+b+c+d+e+f+g+...+(20th letter).  Is there a way to dynamically create a 
variable name?  So that my add function can automatically determine how 
many variable names to create and then add the values together.  In 
other words, I'd ideally like something like:


def add(number_of_variables):
   return a+b+c+...+(letter corresponding to number_of_variables)

Thanks for your help,

Matt


From godzilla at netmeg.net  Tue Apr 11 20:40:28 2000
From: godzilla at netmeg.net (Les Schaffer)
Date: Wed, 12 Apr 2000 00:40:28 GMT
Subject: issubclass funny business
References: <m2snwsd5ah.fsf@netmeg.net> <ekII4.1212$Za1.21055@newsc.telia.net>
Message-ID: <m2purwcfjf.fsf@netmeg.net>

</F> said:

> how about:
>        if not isinstance(std, StandardArrayFileScanner):

i tried that first, but same problem.

> when you run FileScanner as a script, it's evaluated in the __main__
> namespace.  when you import it, it's evaluated again, in the
> FileScanner namespace.

groan.... its not intuitive (to me). because i can make CIE_Standards
class inherit from FileScanner classes in the udder module no problem
at runtime, so its a surprise to see it won't trace the tree properly
at runtime for isinstance issubclass.

> solution: never import the script you use as the main program.

i am only at the testing stage. neither of these modules will be main
at "usage" time.

les schaffer


From sholden at bellatlantic.net  Sun Apr  2 14:51:32 2000
From: sholden at bellatlantic.net (Steve Holden)
Date: Sun, 02 Apr 2000 18:51:32 GMT
Subject: Why should I switch to Python?
References: <000201bf9c14$0a79d9e0$182d153f@tim>
Message-ID: <38E7967E.DE18EFA5@bellatlantic.net>

Tim Peters wrote:
> 
> [Aaron Turner]
> > ...
> > Basically if you know both languages [Perl & Python], and were planning
> > a mid-sized project, why would you choose Python?
> 
> When I was very young, I wondered why my dad's electric razor didn't rip his
> face to shreds.  As a budding empirical scientist, I tested my theories by
> trying it first on my cheek, and then on my tongue.  This was very
> educational -- although it's still painful to talk about <wink>.
> 
Gives a whole new meaning the the phrase "smooth talker"!

> In other words, "voice of experience".  There aren't a lot of bilingual P
> programmers out there, but I expect you'll find the majority of them hanging
> out on c.l.py rather than c.l.p.m.  That should tell you the conclusion
> you're likely to reach if you decide to become one of them too.  Nobody is
> going to argue you into it, though, and Python is easy enough to learn that
> there's no significant barrier against just giving it a try.
> 
Seems that generally Pythonistas, at least from the evidence of newsgroups,
currently tend to have significantly better social skills than are found
on "other newsgroups".  My own blunders have been met with patient nudging
which gave me a chance to see the error of my ways.

> From what I've seen-- which matches my own experience --people generally
> switch over gradually.  The prime motivation seems to be that, time after
> time, when tracking down a bug in a Perl program, you increasingly realize
> "hmm -- if I had written this in Python, this wouldn't have happened".
> Whether it's the 200th time you've been shredded by an array-vs-scalar
> context surprise, or the 12th time an implicit coercion guessed the wrong
> way, they're traps you *can't* fall into in Python.
> 
Again speaking personally, I gave up Perl long before I took up Python, and
this was one of the major reasons.  I had inherited a collection of Perl-
scripted web sites, and maintenance was a real pain.  Boy, was I glas when I
left that job.

Perl's avowed intention that everything which *might* mean something in
the language *should* mean something makes the odds much higher that an
incorrect program won't be caught until execution.  This means that for
a learner much more time has to be spent testing code to find out just
how Perl interprets what *has* to be regarded as a pretty grotty syntax.

As for the OO extension, maybe I'm just a purist, but I found Perl's
exposure of the whole mechanism just a little too rich for my taste.

> Python's syntax is simpler, its semantics more regular, and its philosophy
> of raising exceptions is "gripe whenever there's a reasonable doubt" rather
> than "assume the programmer meant whatever they wrote".  These conspire to
> make larger Python systems less work over the long haul, and especially in a
> team project (TMTOWTDI loses its cowboy charm when you have to decipher 6
> other peoples' code every day too).
> 
Agreed.  The exception mechanism is a major win for Python, and it encourages
much more extensive error checking in the average Python program.  It's much
cleaner, and that makes programs shorter, safer and more comprehensible.

> > It's hard to take my friend seriously in his evaluation when he says
> > he'd rather do this project in C++ than in Perl (especially when he
> > doesn't know Perl).
> 
> Depends on the project.  I make my living these days on a project that's
> about a half million lines of C++ and assembler -- Perl and Python are both
> too fat and slow to have much use in it apart from automating testing.  Perl
> gets used a lot for the latter, but other groups where I work, with large
> projects that can afford the fat of either P language, all use Python now,
> or Java + JPython.  There's no company policy on this, it's a group-by-group
> decision.  And despite that we now have many more lines of Python than Perl,
> I still spend most of my "help others debug their scripts" time staring at
> Perl.
> 
> a-day-of-experience-is-worth-a-decade-of-usenet<wink>-ly y'rs  - tim

regards
 Steve
--
"Bulding information systems just because it's fun."
Steve Holden  sholden at bellatlantic.net  703 716 7275


From raffael at mediaone.net  Fri Apr 28 00:33:49 2000
From: raffael at mediaone.net (Raffael Cavallaro)
Date: Fri, 28 Apr 2000 00:33:49 -0400
Subject: The Simple Economics of Open Source
In-Reply-To: <1255228265-1994958@hypernet.com>
Message-ID: <B52E8CEC.5222%raffael@mediaone.net>

> So has bashing people over the head and taking all their
> money.

Yes, but doing the moral equivalent in software (closed source vendor lock
in) is legal and even more lucrative.

>> You're conflating separate markets. Unique artworks (paintings,
>> for example) have no real equivalent in the software world,
>> because software sales are based on essentially unlimited
>> identical copies.
> 
> 'Twas your analogy.

No, it wasn't, as I explain here. Did you even bother to read it?

> 
>> My example referred to IP in the art world - secret working
>> methods that allowed artists to produce works their
>> contemporaries could not - methods of bronze casting, for
>> example, or formulas for oil painting media. These IPs are like
>> closed source in that, without them, competitors are shut out of
>> a lucrative market.
> 
> There's very little of that in code. The question "how do they
> do that?" is rarely unanswered for longer than a few months.

Oh really? Then why has no competitor yet figured out a *reliable* way to
read and edit MS Word format documents? Just because the IP involved is
intentional obfuscation, doesn't mean that "there's very little of that in
code." In fact, there's a great deal of that in closed source code,
precisely because that's what needs to be there to prevent open source
alternatives from competing.

 
> You really think MS's "monopoly" is based on IP?

Yes! Remember IP doesn't mean "great innovations in science." It just means
"intellectual property," which ranges all the way from the polio vaccine to
ren and stimpy. MS IP is rather lame, but it's sufficintlly dificult to
reverse engineer that it keeps consumers locked into their product
offerings.



>> I think part of the open source community's problem is this naive
>> belief that just because a certain model (open source) leads to
>> products that benefit consumers more, markets will allow
>> consumers to choose this production model. This belief is naive
>> because it fails to consider that the producers' interests are in
>> no way congruent with the consumer's.
> 
> An indicator that the model is diseased.

>From a moral standpoint, no question. From an economic standpoint, the
closed source model is a big winner. It is the anemic open source model,
which must rely on service revenues for any profit that looks sick when it
comes to profitability.

> They are not "monopolies". They are large companies with a
> lot of money, salesman, distribution channels, lawyers and
> programmers. The competition has quite a few programmer,
> and just started to get a little money.

If they produce software that:

1. cannot be easily reverse engineered

2. achievs sufficient market penetration that interoperability with others
becomes a significant feature

then yes, they are effective monopolies. This is *precisely* how MS got
where they are.

>> BTW, all software is *not* support. This is another open source
>> community bias which comes from living in the server room. Lots
>> of software is *client* software, which requires little or no
>> support from the vendor.
> 
> And client software is often free.

Only as a loss leader, or a monopolist's attempt to undercut potential
competition (Netscape for the former, MSIE for the latter)

>  Where software has client
> and server, the money's all made from the server. If the
> vendor's smart, they'll open up the client side so they won't
> have to waste time developing it.


> Bzzt. A 10,000 seat license complains that X is broken, X will
> get fixed. (And when enterprise learns that Y requires only
> 50% the support costs, they will switch.)

Really. This actually has me laughing out loud. Then why are MS Office, and
Windows98 still so badly broken? Lots of enterprises have 10,000 or more
seat licences. Why haven't they been fixed for them?
Answer: for the same reason that your second point is false - because even
big customers can make no credible threat of looking elsewhere, because of
the need to interoperate with others. Enterprises will *never* drop MS
office, no matter how buggy, no matter how much cheaper alternative Y is to
support, unless Y is seamlessly interoperable with everyone else, who use MS
Office.


> Congress does not understand software. Lawyers and judges
> do not.

Judge Jackson understands software very well. You should read his findings
some time. You might understand how MS got where they are, and why closed
source will be with us for a long, long time.

> Most businesses do not. Most people do not.

Oh, but you do, right?


> They're commanding position has nothing to do with IP. Most
> of the maybe-almost-original things they've done (ODBC and
> COM) are pretty thoroughly documented. Their success is due
> to business practices that are common in, say, the food
> industry or the mega-store segment, but are (rightfully)
> considered evil in the software market.

Again, you confuse IP with great scientific innovation. It's not. It's just
intellectual property, which can be something as silly and trivial as a
company logo, or the intentionally obfuscated MS Word document format.


> 
> The open source model(s) will almost certainly require
> adjustment - they are mostly a reflection of the fact that the
> current model is warped. But the big boys are dinosaurs.

Billionaire dinosaurs, who will be with us for as long as capitalism favors
the capitalist over the consumer. I don't see that political balance
changing anytime soon.

You seem to think that this is a technical issue of the "new" economy. It's
not. It's the same old bias that gave us unsafe, gas guzzling cars and toxic
waste dumps. The political system is biased toward the interests of capital,
because of capital's campaign spending power. The result is an economic
system that is heavily biased toward the capitalist at the expense of the
consumer. Software is just another market where this bias manifests itself.
Specifically, why are we even *debating* UCITA, instead of software quality
waranties with some real teeth? Becuase the consumer can't throw big wads of
cash at candidates, or hire full time lobbyists. I don't think that open
source is going to shift this political balance any time soon.

Ralph

-- 

Raffael Cavallaro, Ph.D.
raffael at mediaone.net






From effbot at telia.com  Thu Apr 20 14:49:46 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Thu, 20 Apr 2000 18:49:46 GMT
Subject: pythonware site is down??
References: <8dnhcr$104$1@nnrp1.deja.com>
Message-ID: <ejIL4.3259$Za1.48132@newsc.telia.net>

> But the pythonware site seems t be down.

http://hem.passagen.se/eff/url.htm
has the full story, including pointers to a mirror
of the pythonware.com download site.

</F>




From tjg at avalongroup.net  Fri Apr 21 14:19:45 2000
From: tjg at avalongroup.net (Timothy Grant)
Date: Fri, 21 Apr 2000 11:19:45 -0700
Subject: The Simple Economics of Open Source
References: <1255786824-35980039@hypernet.com>
Message-ID: <39009BC1.91F0713C@exceptionalminds.com>

Gordon McMillan wrote:
> 
> Why does everyone miss the main point?
> 
> I (and you) are literally 100s of times more productive because
> of open source. Imagine having to either invent it all yourself,
> or duct tape together expensive half-assed closed "solutions".
> Shudder.

This is phenomenally true. I've been using Python much longer than I
have had the "Open Source Religion," but one of the reasons I converted
was because I can now do more, experience more, and assist my clients
more for far less money than I ever could when I was paying licensing
fees to Microsoft or Novell.

> Win-win situations are rarely profitably analyzed by
> competitive models.

A phenomenally good book on this very subject is called "Coopetition" by
Brandenburger and Nalebuff. They come to conclusion that it is in
everyone's best interest to find ways to make relationships between
consumers and producers win-win. This is a great read even for nerds.
Highly recommended.

-- 
Stand Fast,
    tjg.

Timothy Grant                         tjg at exceptionalminds.com
Chief Technology Officer              www.exceptionalminds.com
Red Hat Certified Engineer                      (503) 246-3630
Avalon Technology Group, Inc.              fax  (503) 246-3124
>>>>>>>>>>>>Linux...Because rebooting isn't normal<<<<<<<<<<<<



From aahz at netcom.com  Sat Apr 15 18:00:00 2000
From: aahz at netcom.com (Aahz Maruch)
Date: 15 Apr 2000 22:00:00 GMT
Subject: Python 1.6a2 Unicode bug (was Re: comparing strings and ints)
References: <jr%I4.643$rc9.190628352@newsb.telia.net> <1256565470-46720619@hypernet.com> <6M1J4.662$rc9.209708544@newsb.telia.net>
Message-ID: <8daop0$8fk$1@slb6.atl.mindspring.net>

In article <6M1J4.662$rc9.209708544 at newsb.telia.net>,
Fredrik Lundh <effbot at telia.com> wrote:
>
>otoh, 1.6a2 cannot compare strings safely:
>
>Python 1.6a2 (#0, Apr  5 2000, 23:56:55)
>Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>>> u"abc" < "???"
>Traceback (most recent call last):
>  File "<stdin>", line 1, in ?
>ValueError: UTF-8 decoding error: unexpected code byte

That looks like a bug to me.
--
                      --- Aahz (Copyright 2000 by aahz at netcom.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

Why doesn't "Just Say NO" include caffeine, nicotine, alcohol, Prozac,
and Ritalin?  --Aahz


From Florian.Weimer at RUS.Uni-Stuttgart.DE  Tue Apr 25 13:15:52 2000
From: Florian.Weimer at RUS.Uni-Stuttgart.DE (Florian Weimer)
Date: 25 Apr 2000 19:15:52 +0200
Subject: [UNIX] Detaching from parent process
Message-ID: <tgn1minltj.fsf@mercury.rus.uni-stuttgart.de>

I'm trying to implement a CVS loginfo script with Python.  The script
has to fork to the background, because the parent process has to
release the repository lock.  But it seems as if Python doesn't detach
completely from the parent process, because CVS exists only after the
lengthy operation has completed.

Currently I use the following code:

    if os.fork() == 0:
        os.setsid()
        if os.fork() == 0:
            os.close(0)
            os.close(1)
            os.close(2)
            os.open("/dev/null", os.O_RDWR)
            os.dup(0)
            os.dup(0)
            # Lengthy operation claiming the repository lock follows.

-- 
Florian Weimer 	                  Florian.Weimer at RUS.Uni-Stuttgart.DE
University of Stuttgart           http://cert.uni-stuttgart.de/
RUS-CERT                          +49-711-685-5973/fax +49-711-685-5898
http://ca.uni-stuttgart.de:11371/pks/lookup?op=get&search=0xC06EC3B5


From hamish_lawson at yahoo.co.uk  Sat Apr  8 10:28:15 2000
From: hamish_lawson at yahoo.co.uk (Hamish Lawson)
Date: Sat, 08 Apr 2000 07:28:15 -0700
Subject: Cacheing sys.path rather than building it each time?
References: <2e070214.0e300ac8@usw-ex0108-063.remarq.com> <38EF1B53.D21A7514@bellatlantic.net>
Message-ID: <0502b954.18a27ec8@usw-ex0103-023.remarq.com>

Steve Holden wrote:

> Probably not an awfully good idea, given that "installation" of
> a new module often consists of just dropping a ".py" file into
> a library directory.  Without an common "install" process the
> work of checking whether the cahced path is up to date would
> be almost as much as that of building the path from scratch
> as is now done.

Had you thought I was proposing that such a check be done
whenever Python was started? I'd agree that the work of checking
whether the cached path is up to date would be almost as much
work as building the path. However that wasn't part of my
proposal.

Instead when a module is installed, the pathbuilder tool would be
run either by the module's install program (perhaps via the
distutils) or manually by the user. (You could even run it at
scheduled intervals if you thought that necessary to catch
omissions.) My goal is to avoid having the path built every time
Python starts.

A command-line option for the Python interpreter could be used to
determine whether it uses the cached path or builds the path from
scratch; the decision as to which should be the default I'll
leave to others.

Hamish


* Sent from RemarQ http://www.remarq.com The Internet's Discussion Network *
The fastest and easiest way to search and participate in Usenet - Free!



From mwh21 at cam.ac.uk  Mon Apr 24 15:51:18 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 24 Apr 2000 20:51:18 +0100
Subject: code not true?
References: <8ds3ku$q0l$1@nnrp1.deja.com> <3901aa0f.0@news.proweb.co.uk> <M3iM4.518$wJ1.17829@ptah.visi.com> <NhiM4.1704$v85.10013@news-server.bigpond.net.au>
Message-ID: <m3wvlnffbd.fsf@atrus.jesus.cam.ac.uk>

"Mark Hammond" <mhammond at skippinet.com.au> writes:

> "Grant Edwards" <nobody at nowhere.nohow> wrote in message
> news:M3iM4.518$wJ1.17829 at ptah.visi.com...
> 
> > This seems to be something that we were all told as freshmen
> > but stumble over later anyway.  One wonders if a safe language
> > intended for use by beginners should even have an equality
> > operator for floating point objects.
> 
> Of all the solutions to this problem, this does strike me as one of the
> best - simply remove equality testing al together for floating point:
> 

This is what the latest revision of ML actually does; it doesn't put
real in the "Eq" typeclass (or whatever the ML terminology is).

It's a bit of a broad brush solution, but a reasonable one, IMHO, for
ML.  I don't think Python could get away with it.  Similar to the int
division problem (I'd be quite happy if that blew up).

M.

-- 
58. Fools ignore complexity. Pragmatists suffer it. Some can avoid
    it. Geniuses remove it.
     -- Alan Perlis, http://www.cs.yale.edu/~perlis-alan/quotes.html


From tiddlerdeja at my-deja.com  Fri Apr 28 07:58:30 2000
From: tiddlerdeja at my-deja.com (tiddlerdeja at my-deja.com)
Date: Fri, 28 Apr 2000 11:58:30 GMT
Subject: soaplib???
Message-ID: <8ebud1$jk4$1@nnrp1.deja.com>

Is anyone working on a soaplib?

Just wondering.


Sent via Deja.com http://www.deja.com/
Before you buy.


From rupole at compaq.net  Wed Apr 19 19:36:40 2000
From: rupole at compaq.net (Roger Upole)
Date: Wed, 19 Apr 2000 23:36:40 GMT
Subject: Where do I send a contribution for the Starship recovery ?
Message-ID: <iUqL4.142056$ra.3216773@news.easynews.com>

I personally would consider it money well spent.
(unlike most of the other things I spend money on)
                    Roger Upole




From tim_one at email.msn.com  Sun Apr  2 01:18:34 2000
From: tim_one at email.msn.com (Tim Peters)
Date: Sun, 2 Apr 2000 01:18:34 -0500
Subject: regular expressions questions
In-Reply-To: <3d66u8y0rw.fsf@amarok.cnri.reston.va.us>
Message-ID: <000e01bf9c6b$4656f180$752d153f@tim>

[Andrew M. Kuchling]
> ...
> I'm doubtful regular expressions are better for CP4E's audience than
> a fairly simple parser generator such as, say, John Aycock's SPARK.
> Consider how many people are confused by the fact that <p>.*</p> does
> a greedy match and matches all of '<p>para1</p><p>para2</p>'.  And
> didn't SNOBOL provide string processing features that were powerful
> but still usable by non-programmers?  (But my memory's fuzzy.)

SNOBOL and Icon both have pattern-matching systems much more powerful than
regexps, and easier to use, and their audiences are much more from the
humanities than from either comp sci or cgi sci <wink>.  You still have to
be a programmer, though -- specifying a rigorous pattern *is* programming!

it's-cp4e-not-cp4idiots<wink>-ly y'rs  - tim





From phd at phd.russ.ru  Fri Apr 14 06:47:15 2000
From: phd at phd.russ.ru (Oleg Broytmann)
Date: Fri, 14 Apr 2000 10:47:15 +0000 (GMT)
Subject: simple httplib and urllib timeout question
In-Reply-To: <8d6rjk$tdd$1@nnrp1.deja.com>
Message-ID: <Pine.LNX.4.21.0004141043150.26190-100000@fep132.fep.ru>

On Fri, 14 Apr 2000, Georg Mischler wrote:
> phd at phd.russ.ru wrote:
> > Georg Mischler wrote:
> > > What exactly is your problem with signal.alarm() ?
> >
> >    Python delivers alarm only when in Python VM. If the script is
> > blocked in C (even worse - in system call), the signal will NOT be
> > delivered.
> >    Of course, URL checker, while running through a big list of URLs,
> > certainly spend most of the time in connect() and recv() system calls.
> > This makes signals unusable.
> 
> This surprises me, because the snipped I posted is in everyday
> use in something very similar to a link checker, and it works

   I wrote such link checker years ago, and first time I stucked with
timeout problem - yes, I tried to use SIGALRM, sure. That didn't work, and
Guido - many thanks - explained, that there is big problem with signal
handling in Python.

> perfectly well there. Reading throught the Python docs on

   How much URLs are you checking at once? My database is about 3000 URLs,
and I run robot over it every few days.

> signals again, I guess that the different behaviour might
> be caused by the underlying OS implementation.

   May be, but again, may be not.

> I use it on Linux (going to check on FreeBSD soon), and it looks
> like Linux will interrupt system level IO calls when a signal
> arrives for the process. What platform does it fail on for you?

   Linux. I am almost alway use Debian distro. But I tried on Sparc Solaris
with the same effect.

Oleg.
---- 
    Oleg Broytmann    http://members.xoom.com/phd2.1/    phd2 at earthling.net
           Programmers don't die, they just GOSUB without RETURN.




From aek at mail1.stofanet.dk  Tue Apr  4 19:18:30 2000
From: aek at mail1.stofanet.dk (Anders Eggers-Krag)
Date: Wed, 5 Apr 2000 01:18:30 +0200
Subject: Two questions
References: <38EA5FE8.853E9B75@callware.com> <38EA6508.90BF630C@roguewave.com> <38EA6E87.40EDB542@callware.com> <38EA6422.61054B11@exceptionalminds.com> <38EA7452.9BFDF37E@callware.com>
Message-ID: <8cdt7c$cl1$1@news101.telia.com>

an alternative solution is to call an ADO (ActiveX Data Object) object and
use it through
there which is quite simple.

Anders Eggers - Krag

"Ivan Van Laningham" <ivanlan at callware.com> wrote in message
news:38EA7452.9BFDF37E at callware.com...
> Hi All, Tim--
>
> Timothy Grant wrote:
> >
> [snip]
> > Ivan,
> >
> > mxODBC works beautifully, but I've never used it to access and access
> > database, only SQL database. Perchance do you know how Windows
> > advertises an Access database to an ODBC driver?
> >
>
> Nope, not yet.  However, I'll have to find out, won't I?
>
> <you'll-be-the-first-tim-to-know>-ly y'rs,
> Ivan
> ----------------------------------------------
> Ivan Van Laningham
> Callware Technologies, Inc.
> http://www.pauahtun.org
> http://www.foretec.com/python/workshops/1998-11/proceedings.html
> Army Signal Corps:  Cu Chi, Class of '70
> Author:  Teach Yourself Python in 24 Hours
>




From scarblac-spamtrap at pino.selwerd.nl  Wed Apr 19 06:52:15 2000
From: scarblac-spamtrap at pino.selwerd.nl (Remco Gerlich)
Date: 19 Apr 2000 10:52:15 GMT
Subject: testing true condition
References: <Pine.BSF.4.10.10004191138390.292-100000@hercule.perceval.be>
Message-ID: <slrn8fr775.dvd.scarblac-spamtrap@flits104-37.flits.rug.nl>

Gregoire Welraeds wrote in comp.lang.python:
> I know that I can't do something like the following:
> 
> >>> while (line= f.readline()):
> 	...
> 
> as i could in C.  Is there any replacement tricks ?

The standard Python idiom is:

while 1:
   line = f.readline()
   if not line:
      break

There are also wrapper classes around file input, so you can use it in a for
loop, but it doesn't read in the whole file at once (same difference as with
range and xrange). Don't remember its name, but it's not hard to roll your
own.

-- 
Remco Gerlich,  scarblac at pino.selwerd.nl
  Murphy's Rules, "Try a firecracker":
   In TSR's Top Secret, it usually takes three grenade explosions to kill
   the average player-character-agent.


From ivanlan at callware.com  Wed Apr 12 15:42:40 2000
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Wed, 12 Apr 2000 13:42:40 -0600
Subject: Off-topic
References: <38F4BE12.A0C73D1D@callware.com> <m3zoqzgois.fsf@atrus.jesus.cam.ac.uk> <38F4C52C.450F611F@callware.com> <20000412153637.A9859@better.net>
Message-ID: <38F4D1B0.151A534D@callware.com>

Hi All--

William Park wrote:
> 
> On Wed, Apr 12, 2000 at 12:49:16PM -0600, Ivan Van Laningham wrote:
> > Sh scripts do work, as long as I remember to do ./script ;-).  I'm
> > running as root (but the same thing happens when I run as myself, too).
> > However, this could be explained by the default behavour of exec(),
> > which states that if the kernel can't run a program, it execs sh and
> > hands it the program name in argv.
> 
> What is your PATH?  Is your problem related to that fact that '.' is not
> in part of PATH environment variable?
> 

Oh, PATH was OK.  See my other post for how I screwed up.

I just wanted to thank everyone for their assistance in helping me
figure out which end my head was attached to.

<not-a-word-out-of-you-gordon>-ly y'rs,
Ivan;-)
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
http://www.pauahtun.org 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
Author:  Teach Yourself Python in 24 Hours



From gward at cnri.reston.va.us  Mon Apr  3 16:56:21 2000
From: gward at cnri.reston.va.us (Greg Ward)
Date: Mon, 3 Apr 2000 16:56:21 -0400
Subject: [Python-Dev] New Features in Python 1.6
In-Reply-To: <200004031604.MAA05283@eric.cnri.reston.va.us>; from guido@python.org on Mon, Apr 03, 2000 at 12:04:02PM -0400
References: <200004011740.MAA04675@eric.cnri.reston.va.us> <200004031604.MAA05283@eric.cnri.reston.va.us>
Message-ID: <20000403165621.A9955@cnri.reston.va.us>

On 03 April 2000, Guido van Rossum said:
> Not only was it an April fool's joke, but it wasn't mine!  It was
> forged by an insider.  I know by who, but won't tell, because it was
> so good.  It shows that I can trust to delegate way more to the Python
> community than I think I can! :-)
> 
> BTW, the biggest give-away that it wasn't mine was the absence of my
> standard sign-off line:
> 
> --Guido van Rossum (home page: http://www.python.org/~guido/)

D'ohhh!!!

Hasn't anyone noticed that the largest amount of text in the joke
feature list was devoted to the Distutils?  I thought *that* would give
it away "fer shure".  You people are *so* gullible!  ;-)

And for my next trick... *poof*!

        Greg



From moshez at math.huji.ac.il  Sun Apr 23 01:00:21 2000
From: moshez at math.huji.ac.il (Moshe Zadka)
Date: Sun, 23 Apr 2000 08:00:21 +0300 (IDT)
Subject: code not true?
In-Reply-To: <jwbnews-6EFA9D.13383522042000@news.olympus.net>
Message-ID: <Pine.GSO.4.10.10004230758200.15546-100000@sundial>

On Sat, 22 Apr 2000, John W. Baxter wrote:

> > > >>> print 7.0 == 7
> > > 1
> > > >>> print (.07 * 100) == 7
> > > 0
> > 
> > Because computers represent floats with too little bits.
> > Never ever compare two floats for equality!
> > 
> Not quite.  No number of bits is sufficient to represent .07 exactly in 
> the typical (and IEEE standard) representation.  

Yes, but enough bits could make sure that 0.07*100 is indistinguishable
from 7.0

But I agree with the point -- floats are evil. Very very evil. In fact,
I wouldn't mind at all if there was some specific command-line flag to the
interpreter to enable floats.

--
Moshe Zadka <mzadka at geocities.com>. 
http://www.oreilly.com/news/prescod_0300.html
http://www.linux.org.il -- we put the penguin in .com




From milgram at cgpp.com  Sat Apr 15 23:16:15 2000
From: milgram at cgpp.com (milgram at cgpp.com)
Date: 15 Apr 2000 23:16:15 -0400
Subject: cgi handling of embedded quotes
Message-ID: <m266tivips.fsf_-_@tcraft.cgpp.com>


Hi,

have a cgi script written in python, use the cgi module and begin
something like this:

import cgi
form = cgi.FieldStorage()        

etc. 

The form uses post method.

I recently discovered that form entries containing double-quotes cause
the data entry to be truncated, that is, if the user enters:

101 "I" St.

in a text field, the variable, after dereferencing form, gets
truncated right at the first double-quote.

Surely there's an easy way to deal with this, but I can't figure out
what it is. Any pointers to docs, or other suggestions?

thanks in advance,

Judah

-- 
Judah Milgram        milgram at cgpp.com
P.O. Box 8376, Langley Park, MD 20787
(301) 422-4626 (-3047 fax)


From ke at gnu.franken.de  Thu Apr 27 00:38:23 2000
From: ke at gnu.franken.de (Karl EICHWALDER)
Date: 27 Apr 2000 06:38:23 +0200
Subject: Scripting and Gnome and KDE
References: <Pine.GSO.4.10.10004150741520.11261-100000@sundial> <3dk8hzb90u.fsf@newcnri.cnri.reston.va.us>
Message-ID: <shk8hk2m68.fsf@tux.gnu.franken.de>

Andrew Kuchling <akuchlin at mems-exchange.org> writes:

> Then reality sets in: GNOME uses patched versions of all sorts of
> things such as automake, but this is never clearly documented
> anywhere.

libtool might be an exception.  But normally it's enough to use recent
versions of all involved tools.

> There's a large pile of libraries, and it's not clear which ones are
> needed and which ones are obsolete;

autoconf resp. ./configure is considered to do all needed checks for
you.  If it fails for some reason it's a bug; please, report it.  We'll
try to improve the check and add a note to the README file.

The problem is, it works nicely for the devolopers and package makers
(like me) who are familiar with the progress GNOME makes.

-- 
work : ke at suse.de                          |
     : http://www.suse.de/~ke/             |          ------    ,__o
home : ke at gnu.franken.de                   |         ------   _-\_<,
     : http://www.franken.de/users/gnu/ke/ |        ------   (*)/'(*)


From boud at rempt.xs4all.nl  Fri Apr 21 01:38:06 2000
From: boud at rempt.xs4all.nl (Boudewijn Rempt)
Date: 21 Apr 2000 05:38:06 GMT
Subject: Using python on the web
References: <EDFD2A95EE7DD31187350090279C6767754BA8@THRESHER>
Message-ID: <8dopfu$nrl$1@news1.xs4all.nl>

Brent Fulgham <brent.fulgham at xpsystems.com> wrote:

> Of course, the Postgres Python bindings may be just as bad from your
> perspective.  :-)

Well, the Postgres Python bindings don't even conform to the
DB-API II specs - that kind of rules them out for a lot of
work. I myself am quite satisfied with the MySQL bindings, but
like the original poster, I had to write my own object layer
around the primitive functions. Maybe it would be a good idea
to have another standard layer on top of DB-API II, for those
who want a more sophisticated interface?

-- 

Boudewijn Rempt  | http://www.valdyas.org


From crystalc at interlog.com  Sat Apr 29 16:09:30 2000
From: crystalc at interlog.com (Brian Alexander)
Date: Sat, 29 Apr 2000 16:09:30 -0400
Subject: Linux Processes from Python HOW-TO?
References: <8ecpch$j73$1@nnrp1.deja.com> <20000428234608.L11481@xs4all.nl>
Message-ID: <54HO4.2549$Ip.71417@cac1.rdr.news.psi.ca>

Thomas;

You would have gotten 100% on an exam I just wrote: Systems Tools and
Programming for UNIX.

Brian :)


"Thomas Wouters" <thomas at xs4all.net> wrote in message
news:20000428234608.L11481 at xs4all.nl...
> On Fri, Apr 28, 2000 at 07:39:00PM +0000, chris_barker at my-deja.com wrote:
>
> > Has anyone written a HOW-TO or tutorial that explains how to start and
> > manipulate other processes with Python. I imagine it's trivial if you
> > are familiar with the Linux/Unix process model, but I get very confused
> > about when to use:
>
> > system
> > popen and it's variants
> > exec and it's variants
> > fork
> > etc.
>
> Well, it's detailed in most books about UNIX, and in most books about
> operating systems in general. (At least on the programming level.) I dont
> know any HOWTO-level explanation of it, but I'm happy to supply you with
> some short descriptions. I'll do it in the reverse order of what you
listed,
> though, because it's easier that way ;)
>
> First, the basics: under UNIX, and most current operating systems, a
running
> program consists of one or more processes. A single program, or a single
> task, is usually a single process, but it's possible for several processes
> to communicate with eachother, to share workload. Each process has its own
> memoryspace, invisible to other processes, and it can only access its own
> memoryspace.
>
> To create a new process, you use the system call fork(). It duplicates the
> current process, copies all of its memory, instruction pointers, function
> stack, everything. The only difference between the two processes is the
> process-id (pid), and the return code of fork(): In the new process, the
> 'child', fork returns 0. In the original process, the 'parent', fork()
> returns the pid of the newly created child process. (If fork() fails, it
> wont create a child process, of course, and the return code will signify
an
> error.)
>
> Because fork() creates an exact duplicate of the old process, you can't
use
> it to start a new program. You use one of the 'exec*()' calls for that.
This
> system call *replaces* the calling process by a new process, running a new
> program as indicated by the first argument to the exec() call. If exec()
> returns anything, it's always an error, because if it succeeds, the
process
> that called it will be gone. The new process isn't really a new process,
> though, because it will have the pid of the old one. It'll just be running
a
> different program ;)
>
> The difference between the various exec*() calls is fairly subtle, and a
bit
> strange from a Python point of view: they are mostly about how to pass the
> arguments: the execl*() calls take a variable number of arguments,
> whereas the execv*() calls take a pointer to a list of arguments. The
> exec*p() calls search the environment variable PATH for the name of the
> executable, whereas the other exec*() functions expect the name of the
> executable to be a full path to the program. And lastly, the exec*e()
> calls also take an argument that list the environment variables that
> should be passed to the new program, whereas the other calls pass the same
> environment as the current process has.
>
> These two system calls are all that is necessary to create as many and as
> varying processes as you wish. None the less, the Standard C Library
> provides a few funky and helpful functions that save you a lot of time in
> the common cases:
>
> system() is a library call that takes a single string describing the
program it
> should start, and the arguments it should get. It forks() and exec()s for
> you, and then waits for the child to exit (using one of the wait() calls)
> and returns the exit code. You dont have particularly much control over
how
> the string is executed, other than via the standard 'shell' commands:
system
> always starts /bin/sh with your string as argument.
>
> popen() and it's friends are similar functions, but instead of running the
> string you pass and returning the exitcode, it starts the new process with
> its standard input or standard output (or both, in the case of popen2)
> connected to the calling process (using a UNIX 'pipe', hence the first p
in
> popen()). This way, the calling process can pass data to the program just
as
> a user would type it on a computer, or read data that the new program
writes
> to the terminal.
>
> So, basically, if you simply want to run a command and dont care much
about
> runtime influence, you usually use system(). If you need to pass it some
> data for it to work on, for example 'gzip' or some other program that
works
> on the input stream, you use popen(). And if you want more control over
> things, or if you simply want your python process to do two tasks at once,
> or you wish to start a lot of programs in a specific order, or other
> complicated things, you use fork() and exec() manually.
>
> > I'm really hoping someopne has written this up!
>
> Well, I just did, didn't I ? ;) There are a lot more details I didn't
> mention, by the way: threads (processes that share their memory space);
> redirecting stdin, stdout and stderr somewhere specific, or to another
> process; creating pipes and sockets yourself; how to avoid deadlocks in
> popen2; shared memory; controling ttys, how to see the difference between
a
> simple pipe and a real user, and how to fool programs that try to see that
> difference; UNIX session ids; setuid() and other ways to change the
current
> process's permissions; And probably a lot more ;P Nevertheless, I hope
this
> helps a bit :-)
>
> If you need more info than this, either read a UNIX manualpage for one of
> those functions, or if you prefer readable material, buy a good book on
UNIX
> programming. They usually start with a good explanation of processes, the
> memory layout, pipes, sockets and networking. Some even cover shared
memory.
>
> Drained-ly y'rs,
> --
> Thomas Wouters <thomas at xs4all.net>
>
> Hi! I'm a .signature virus! copy me into your .signature file to help me
spread!
>




From kcazabon at home.com  Sun Apr  9 17:30:24 2000
From: kcazabon at home.com (Kevin Cazabon)
Date: Sun, 09 Apr 2000 21:30:24 GMT
Subject: Tkinter:  centering windows on screen
References: <ak1I4.121418$Dv1.1306619@news1.rdc1.bc.home.com> <38F0C6A7.10878609@dkfz-heidelberg.de> <38F0C95D.1F7E853@dkfz-heidelberg.de> <xR4I4.274$rc9.188784640@newsb.telia.net>
Message-ID: <QD6I4.121510$Dv1.1316301@news1.rdc1.bc.home.com>

Ahhhh... kinda painful until I call .update_idletasks() before reading the
geometry... It's a toplevel widget, and geometry() returns 1x1 as the size
if you dont.

Kevin.


"Fredrik Lundh" <effbot at telia.com> wrote in message
news:xR4I4.274$rc9.188784640 at newsb.telia.net...
> Kilian Lorenz <kilian.lorenz at dkfz-heidelberg.de> wrote:
> > >> that works but doesn't help me much because I'm interested in the
> > >> return value. So I would really like to know how to call Tk commands
> > >> from within Python because .... blah blah
> >
> > Now I also found out how this works ... pretty easy with
> > <widget>.tk.call(<command>, <parameter>, <next parameter> ...)
> >
> > You just have to browse Tkinter.py and get the answers to all your
> > questions !  :)
>
> if you browsed the source, how come you didn't find the
> corresponding Tkinter methods?
>
> windows geometry:
>
http://w1.132.telia.com/~u13208596/tkintrobook/x9867-window-geometry-methods
> .htm
>
> file dialogues:
> http://w1.132.telia.com/~u13208596/tkintrobook/x1164-data-entry.htm
>
> (direct pointers to the current draft; official version at
> www.pythonware.com will be synchronized next week).
>
> </F>
>
>




From JoeSmith at bogusaddress.com  Fri Apr  7 06:52:25 2000
From: JoeSmith at bogusaddress.com (Joe Smith)
Date: Fri, 07 Apr 2000 10:52:25 GMT
Subject: favorable media attention :-)
References: <jlitxusyen.fsf@endor.bbb.caltech.edu>
Message-ID: <38EDBEB8.BEF502AB@bogusaddress.com>

This might be a better link:
    http://metalab.unc.edu/Dave/Dr-Fun/df200004/df20000406.jpg.

Michael Vanier wrote:

> Hey everyone, check out the April 6th Doctor Fun cartoon, currently at:
>
> http://metalab.unc.edu/Dave/Dr-Fun/latest.jpg
>
> You'll be glad you did ;-)
>
> BTW IMO Doctor Fun is the funniest cartoon around these days; it reminds me
> of the Far Side at its peak.
>
> Mike
>
> --------------------------------------------------------------
> Mike Vanier     mvanier at bbb.caltech.edu
> Department of Computation and Neural Systems, Caltech 216-76
> GNU/Linux: We can't lose; we're on a mission from God.



From moshez at math.huji.ac.il  Sat Apr  1 18:03:05 2000
From: moshez at math.huji.ac.il (Moshe Zadka)
Date: Sun, 2 Apr 2000 01:03:05 +0200 (IST)
Subject: Why should I switch to Python?
In-Reply-To: <000201bf9c14$0a79d9e0$182d153f@tim>
Message-ID: <Pine.GSO.4.10.10004020056440.20943-100000@sundial>

On Sat, 1 Apr 2000, Tim Peters wrote:

> There aren't a lot of bilingual P
> programmers out there, but I expect you'll find the majority of them hanging
> out on c.l.py rather than c.l.p.m.

I wonder what made you say that? Actually, I know quite a few bi-lingual
P'ers (my ex-girlfriend, her ex-boyfriend, and her current boyfriend.
Shows you what kind of crowd I mix in) and no highly-proficienct one-P 
programmers (sure, some Perl dabblers, but nothing serious).

So from my prespective, most people know both P languages. And most prefer
Python, though some still spend some time writing in Perl. (Thanks to
Zope, at least those in web-development switch to Python).

Of course, it possible that those who prefer Perl are simply afraid to
tell me.

(Actually, the only Perler but not Pythonista I know is now thinking of
learning Zope...so we might yet conquer the world).

you'll-never-know-if-this-is-just-april's-fool-ly y'rs, Z.
--
Moshe Zadka <mzadka at geocities.com>. 
http://www.oreilly.com/news/prescod_0300.html
http://www.linux.org.il -- we put the penguin in .com




From nobody at nowhere.nohow  Sat Apr 29 18:45:34 2000
From: nobody at nowhere.nohow (Grant Edwards)
Date: Sat, 29 Apr 2000 22:45:34 GMT
Subject: How can I get the name of an object???
References: <8e712k$65c$1@nnrp1.deja.com> <etd8zy03m44.fsf@w20-575-108.mit.edu>
Message-ID: <slrn8gmpge.81p.nobody@isis.visi.com>

In article <etd8zy03m44.fsf at w20-575-108.mit.edu>, Alex wrote:
>
>An object could be referred to by any number of names:
>
>dd = zz ()
>ee = dd
>
>Which name for the object do you wish to know?  ee or dd?

There are also cases where an object has no name at all:

x = []
x.append(zz())

-- 
Grant Edwards                   grante             Yow!  Maybe we could paint
                                  at               GOLDIE HAWN a rich PRUSSIAN
                               visi.com            BLUE --


From cfelling at iae.nl  Tue Apr 25 19:24:03 2000
From: cfelling at iae.nl (Carel Fellinger)
Date: 26 Apr 2000 01:24:03 +0200
Subject: Question about exception-handling mechanism
References: <8e46mt$4nh$1@nnrp1.deja.com>
Message-ID: <8e59ej$j77$1@vvs.superst.iae.nl>

willfg at my-deja.com wrote:
> Hello all,

> colleague asked why in an exception handling mechanism you'd want the
> ELSE block to be executed if you don't throw an exception as opposed to
> a FINALLY block. Anyone used this feature in practice? Thanks in
> advance for your input, -- Will

others have given a more abstract anwser, but being in a talkative mood
I'll try to add a real example:)

>>> open_connection_to_other_computer_via_the_phone()
>>> try:
>>>     do_something_worthwile()
>>> finally:
>>>     #to avoid extreme bills we have to hang-up no matter what
>>>     close_connection_and_hang_up_the_phone()

the finally clause here is the obvious way to guarantee that the phone
is put on the hook, even if some unexpected exception occurred and the
program is really dying. Ofcourse you can approach the same with a cach-all

>>> except:
>>>     close_connection_and_hang_up_the_phone()
>>>     raise   #reraise the current exception

But somehow this is deemed inferior and less clear then the above.
(Funny if you think of the down-out horrible
but widely defended while 1:...break idiom:)

>>> dict = {1: "the time is 1:30", 2: "the time is 2:45"}
>>> try:
>>>     time_string = dict[1]
>>> except KeyError:
>>>     t = some_default_time_value_not_needing_convertion_from_string
>>> else:
>>>     t = extract_time_from_string(time_string)

Again this can easily be achieved in different ways in this example.


ceterum censeo
    while: ... break
should have been part of the language to begin with
-- 
groetjes, carel


From mhammond at skippinet.com.au  Tue Apr  4 20:05:01 2000
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Wed, 05 Apr 2000 00:05:01 GMT
Subject: Map a Network Drive and Printers in NT?
References: <38ea3a47.7163227@news.flashnet.it>
Message-ID: <NqvG4.5063$1b4.24398@news-server.bigpond.net.au>

Get the win32all extensions, and lookup the documentation on the
win32wnet module - everything you need is there...

Mark.

"Fra" <fra at bastardi.org> wrote in message
news:38ea3a47.7163227 at news.flashnet.it...
> Hi all
> How can I map a network drive and a printer in Win NT and in
Win 95?
> Can I iport the dos command line and type the command "Net
use...." or
> is there a different way?
> And could you also tell me how to ping with a Python script?
And maybe
> how to let this script analyse the result.
>
> I couldn't find it on the Tutorial, neither on the FAQ. Am I
wrong?
> Thank you,
>
> Greetings,
>    Francesca




From senn at maya.com  Wed Apr  5 09:15:19 2000
From: senn at maya.com (Jeff Senn)
Date: 05 Apr 2000 09:15:19 -0400
Subject: Tkinter:  "Get root for a widget" is private?
In-Reply-To: Randall Hopper's message of "Wed, 05 Apr 2000 07:35:26 -0400"
References: <20000331112048.A1257432@vislab.epa.gov> <QdEF4.892$n68.190327808@newsb.telia.net> <m3snx3n1x0.fsf@devel.no.rwd.dinet> <20000331112048.A1257432@vislab.epa.gov> <QdEF4.892$n68.190327808@newsb.telia.net> <20000403064335.A1560771@vislab.epa.gov> <2Q_F4.1092$n68.187097600@newsb.telia.net> <20000405073526.A1760238@vislab.epa.gov>
Message-ID: <wk8zysd6zc.fsf@SNIPE.maya.com>

Randall Hopper <aa8vb at yahoo.com> writes:
> I have a timer class built around Tk timers.  I realize I could just pass
> any_old_widget into this routine, and it would store it off into state data
> and use it periodically to access the 'after' and 'after_cancel' methods.
> 
> However, this timer has nothing to do with this widget in particular, so
> why should it have a reference to it (versus the button next to it, or the
> list widget below it)?  It makes more sense for the timer class to accept
> and bookmark the root object.  This root is associated with the global Tk
> state, and it makes sense for a timer to bookmark this to use global Tk
> state commands.

Unless you're doing something wacky (creating multiple roots??), you
could just do what Tkinter itself (generally) does and use
Tkinter._default_root to get the instance of the root object....

-- 
-Jas




From seh at ffi.no  Tue Apr 18 09:36:26 2000
From: seh at ffi.no (seh at ffi.no)
Date: Tue, 18 Apr 2000 15:36:26 +0200
Subject: Running shell programs from Python
Message-ID: <38FC64DA.7D297563@ffi.no>

I have just started with Python and have some questions.. I have some
compiled c programs that I normally run from the shell. I can use | to
pipe the output from one as input to the next. I would like to call the
programs from Python. Could someone please indicate how this is done?
Any references?

Regards
Svein-Erik Hamran



From rkhale at earthlink.net  Sun Apr 23 09:09:27 2000
From: rkhale at earthlink.net (Richard Hale)
Date: Sun, 23 Apr 2000 13:09:27 GMT
Subject: The Simple Economics of Open Source
References: <B527E7A5.51AA%raffael@mediaone.net>
Message-ID: <MPG.136cc01af4ab703b989709@news.earthlink.net>

In article <B527E7A5.51AA%raffael at mediaone.net>, raffael at mediaone.net 
says...
> I see a parallael to the software industry here as well. Those with a
> lucrative cash cow (e.g., MS) are reluctant to share code, but those who
> just make a living programming are happy to share source. It would seem that
> sharing is inversely proportional to the value of that which is shared.
> 
> 
> 
_Or_ it would seem that sharing is inversely proportional to the 
_perceived_ value of that which is shared.

The Open Source software that I have used has been very valuable to me.  
Did those who produced it not see as much value in it?

I would answer the previous question no and observe that the parallel is 
not valid.  The models and motivations are different and not parallel.

Richard


From youngd at digiship.com  Thu Apr 27 10:16:50 2000
From: youngd at digiship.com (Darren Young)
Date: Thu, 27 Apr 2000 09:16:50 -0500
Subject: DCOracle
Message-ID: <NEBBIEPHGLOHPEDKJLGJAEBICAAA.youngd@digiship.com>

Anyone have any basic code that is 'known' to work against an Oracle (8i)
database using the DCOracle module? I want to verify that the module is
working properly. I got it to compile cleanly and leave no unresolved symbol
errors. The Connect method seems to be working, however there are no clear
examples to actually get some rows.

Any pointers on examples of working code out there?

Thanks in advance...

---------------------
Darren Young
Digiship
312-527-9440
---------------------




From tismer at tismer.com  Sun Apr 16 12:47:58 2000
From: tismer at tismer.com (Christian Tismer)
Date: Sun, 16 Apr 2000 18:47:58 +0200
Subject: [ANNOUNCE] Garbage collection for Python
References: <20000407221750.A29279@acs.ucalgary.ca> <m3r9cagx3c.fsf@atrus.jesus.cam.ac.uk> <PRKJ4.1078$rc9.191689728@newsb.telia.net> <38F7894D.9997636A@gmx.de>
Message-ID: <38F9EEBE.F48A664A@tismer.com>


Michael Scharf wrote:
> 
> Fredrik Lundh wrote:
> >
> > Michael Hudson <mwh21 at cam.ac.uk> wrote:
> > > It would be, in my opinion, be a really really really good thing; for
> > > one thing it would help in language wars...
> >
> > sorry, but stackless python is the only really really really good
> > thing that hasn't yet been added to the core.  this is just really
> > really good.
> 
> And what are the reasons for not adding the really really really good
> stackless python??

Since it is all a bit ahead of time. Actually, I had stolen
Guido's time machine, and I was not allowed to take this
stuff home, but I did. Guilty, should ave known better!
Now I have the hard task to grow enough people to understand
this stuff, and to realize that there is no stackful future
after they saw that it *can* be done.

need-to-go-back-to-the-future-and-kidnap-that-Py3k-Guido 
  - ly y'rs - chris

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaunstr. 26                  :    *Starship* http://starship.python.net
14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     where do you want to jump today?   http://www.stackless.com



From rcunning at acm.org  Mon Apr 17 15:01:42 2000
From: rcunning at acm.org (Robert W. Cunningham)
Date: Mon, 17 Apr 2000 19:01:42 GMT
Subject: Filtering web proxy
References: <Pine.LNX.4.21.0004171406250.30999-100000@fep132.fep.ru>
Message-ID: <38FB5F90.DFC5000F@acm.org>

Oleg Broytmann wrote:

>    I want a filtering web proxy. I can write one myself, but if there is a
> thing already... well, I don't want to reinvent the wheel. If there is such
> thing (free and opensourse, 'course), I'll extend it for my needs.
> <snip>
>    I wrote dozen HTML parsers in Python, so I can write one more, and turn
> it into a proxy, but may be I can start with some already debugged code?

If you want a porting project, there is the FilterProxy program, written in
Perl.  It is a true daemon, and implements several nifty features, such as a
modular architecture that lets users create their own filters.  Present filters
include a banner ad blocker, and an cool HTTP compressor (a remote FilterProxy
instance on the far end of a modem link gzips all HTTP traffic for a 5x size
reduction, which Netscape decompresses on the fly).  There is also a template
to simplify writing new modules, and if I ever get around to it I'm going to
add simple stream encryption (just for fun).

FilterProxy is still alpha, but it is already fairly potent.  Future
improvements may include support for chaining proxies, so that local and/or
remote instances may be used in combination with an ISP's HTTP caching proxy
server.

If you have a broadband connection (cable modem, DSL, LAN), FilterProxy
introduces a noticeable delay, but not objectionable.  It costs only a little
more than an IPchains firewall in terms of CPU bandwidth.  And the Blocker
module is getting very good at removing some of the more complex ad formats
(including some Java and Javascript devils).

It would be an excellent comparison to have Python and Perl versions
side-by-side, so the power of each can be more easily observed.

I don't recall the URL, but it is on Freshmeat.


-BobC




From rehakm1 at cti.ecp.fr  Fri Apr 28 16:49:25 2000
From: rehakm1 at cti.ecp.fr (Martin Rehak)
Date: Fri, 28 Apr 2000 22:49:25 +0200
Subject: Newbie - COM parameters problem
Message-ID: <3909F954.9B51E117@cti.ecp.fr>

Hi
Being a student of an engineering school, I've started using Python as a
scripting language for a program that's used for an analyze of the
images taken by an electron microscope. As an application interface, we
are using COM (OLE), and as we use the interfaces herited from
IDispatch, we are limited to the use of the VARIANT type compliant
parameters. Each image is represented by one COM object and the whole
application is running as an in-process server.
In one method , we have to pass the reference to the object as an [in]
parameter.
And my question is, how can we actually do it, when the only VARIANT
"sub-type" fit for this is an IDispatch * and in the whole Python
documentation, that I found on the net,  is no mention of some pointer
type or its substitution ?
Thank for your help or any piece of information on the subject.

Martin Rehak
P.S. Python is COOL, anyway!

--




From tismer at tismer.com  Sun Apr 16 14:27:06 2000
From: tismer at tismer.com (Christian Tismer)
Date: Sun, 16 Apr 2000 20:27:06 +0200
Subject: [ANNOUNCE] Garbage collection for Python
References: <20000407221750.A29279@acs.ucalgary.ca> <m3r9cagx3c.fsf@atrus.jesus.cam.ac.uk> <l03102809b51d41895a80@[193.78.237.135]> <38F9F01B.57273052@tismer.com> <m33domez7n.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <38FA05FA.B0526CAB@tismer.com>


Michael Hudson wrote:
> 
> Christian Tismer <tismer at tismer.com> writes:
> 
> > It will become simpler, simpler and simpler.
> 
> I cam across a great quote the other day; I think it was attributed to
> Alan Perlis, and it was:
> 
>     Simplicity follows complexity, not the other way round.

Thank you a lot, it was exactly the quote I was trying
to look up, but I didn't find it again. It is very true.

> I think (and in fact have always thought, even before I read the
> quote) Stackless Python is an almost perfect example of this.

1 + 1 == 42 for-sufficiently-large-values-of-1 - ly y'rs - chris

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaunstr. 26                  :    *Starship* http://starship.python.net
14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     where do you want to jump today?   http://www.stackless.com
        Simplicity follows complexity, not the other way round.



From rupole at compaq.net  Wed Apr 19 19:36:39 2000
From: rupole at compaq.net (Roger Upole)
Date: Wed, 19 Apr 2000 23:36:39 GMT
Subject: Where do I send a contribution for the Starship recovery ?
Message-ID: <gNqL4.142052$ra.3209287@news.easynews.com>

I personally would consider it money well spent.
(unlike most of the other things I spend money on)
                    Roger Upole




From genehilton at my-deja.com  Fri Apr 28 15:11:32 2000
From: genehilton at my-deja.com (genehilton at my-deja.com)
Date: Fri, 28 Apr 2000 19:11:32 GMT
Subject: accessing BLT vectors from C
References: <8dq957$tc8$1@nnrp1.deja.com> <20000427220821.63091@nms.otc.telstra.com.au>
Message-ID: <8ecnol$h6g$1@nnrp1.deja.com>

In article <20000427220821.63091 at nms.otc.telstra.com.au>,
Greg

Thanks for the input.  I had figured out the PYVEC stuff - where I am
right now is that it is either a linking or a threading problem.  If you
dig about 5 layers down in the source for blt (mostly in bltVector.c and
bltUtil.c) I find that the reason my code doesn't find PY_VEC0 is that
it can't find the Uid (unique identifier not User ID) - that is
Blt_FindUid returns false.  On further examination I see that there are
two different instances of Blt_FindUid() linked (or threaded?), one by
_tkinter and one by my module - so Blt_FindUid() sees the static
variable uidInitialized=0 so a new Uid hash table is built.  Furthermore
if I add a print statment to Blt_FindUid() to print the address
Blt_FindUid() I get a different address if it is called by _tkinter (via
PMW) or if it is called by my module.  I think its all a matter of
understanding how  my code gets loaded vs _tkinter.  I think I should be
able to straighten this out when I find a few hours.  Any bright ideas
are certainly welcome however.

Gene

>
> What is the value of bltname?  The names that Pmw give to Blt vectors
> are PY_VEC0, PY_VEC1, PY_VEC2, etc.  You can see what Tcl calls are
> being made by calling Pmw.tracetk().  For example, this Pmw:
>     v = Pmw.Blt.Vector(1)


Sent via Deja.com http://www.deja.com/
Before you buy.


From arnold at dstc.edu.au  Tue Apr  4 08:45:57 2000
From: arnold at dstc.edu.au (David Arnold)
Date: Tue, 04 Apr 2000 22:45:57 +1000
Subject: Python 1.6 alpha 1 released 
In-Reply-To: Your message of "Sun, 02 Apr 2000 22:32:14 -0400."
             <8c8vfj$5n6$1@news.udel.edu> 
Message-ID: <200004041245.WAA17086@piglet.dstc.edu.au>

-->"Terry" == Terry Reedy <tjreedy at udel.edu> writes:

  Terry> Without countervailing philosophical justification, this
  Terry> code-breaking change strikes me as contrary to both the
  Terry> spirit and practice of python.

the justification is (and always was) that connect() does not always
take a host and port.  it *does* always take a single address
parameter.  the format of that address parameter depends on the
address/protocol family with which the socket was created.

  Terry> If the connect() code currently also accepts the new form,
  Terry> then the API could be tightened just as well by eliminating
  Terry> the one-tuple variant.  How much code would that break?

(just to be pedantic, it's not the new form. it's the form from as
long as i can remember -- 1.1? 1.3 at least)

maintaining the current hack to support two arguments is probably the
best solution, given the storm of protest this has caused.  i don't
know why Guido proposed its removal, but perhaps it's not worth it?


d



From nospam at nowhere.net  Fri Apr 21 00:26:54 2000
From: nospam at nowhere.net (Ray&Maria)
Date: Thu, 20 Apr 2000 22:26:54 -0600
Subject: Tkinter Canvas -- help?
References: <8ddtdm$b2s$1@nnrp1.deja.com> <dRyK4.2542$Za1.37532@newsc.telia.net> <XrQL4.6240$lw2.184908@news.corecomm.net>
Message-ID: <aMQL4.6243$lw2.185769@news.corecomm.net>

Never mind.  I've found the answer from another Fredrik's reply.

http://hem.passagen.se/eff/url.htm

"Ray&Maria" <nospam at nowhere.net> wrote in message
news:XrQL4.6240$lw2.184908 at news.corecomm.net...
> Whatever happen to your excellant (and only) Tkinter manual site Fredrik?
> The pages are gone for the last 2 days, and python.org does not respond to
> my e-mail except the thank you robot.  It appears that pythonware no
longer
> holds any content for Tkinter.  Does anyone know if I can download a copy
> from a mirror site?
>
> Jueiming Chen
> NOSPAMjchen at NOSPAMquark.com
>
>
> "Fredrik Lundh" <effbot at telia.com> wrote in message
> news:dRyK4.2542$Za1.37532 at newsc.telia.net...
> > <lbear11 at my-deja.com> wrote:
> > > I'm just starting to learn Python/tkinter, and I have found that there
> > > is virtually _no_ documentation for the tkinter canvas. Could anybody
> > > help? I'm specifically looking for stuff like creating an image by a
> > > grid in canvas on "mouseDown", and how to read the items in the
canvas,
> > > and write them to a text file, in their proper positions. I would
> > > _really_ appreciate it if anybody could help me with this.
> >
> > how to capture mouse down:
> >
>
http://www.pythonware.com/library/tkinter/introduction/events-and-bindings.h
> > tm
> >
> > how to create image items:
> > http://www.pythonware.com/library/tkinter/introduction/canvas-image.htm
> >
> > how to read items in the canvas:
> > http://www.pythonware.com/library/tkinter/introduction/canvas.htm
> > (look under methods, for find_all, coords, etc)
> >
> > > P.S. The only tkinter canvas documentation that I have is the
> > > "Introduction to Tkinter" from pythonware.com
> >
> > </F>
> >
> >
>
>




From raffael at mediaone.net  Sun Apr 23 23:40:29 2000
From: raffael at mediaone.net (Raffael Cavallaro)
Date: Mon, 24 Apr 2000 03:40:29 GMT
Subject: The Simple Economics of Open Source
References: <B527E7A5.51AA%raffael@mediaone.net> <3903316C.5D6F1611@tcac.net>
Message-ID: <raffael-34658F.23400223042000@news.ne.mediaone.net>

>Talent is not a "secret" that can
>be hoarded or shared.  It is visible in the end product, yet no amount
>of sharing how it was done will allow the process to be duplicated.

Oh, so that's why museums have been repeatedly fooled to the tune of 
millions of dollars over the years by forgeries of the works of great 
artists.

Fact is, knowing how great art is made allows reasonably skilled 
individuals to make effectively indistinguishable copies, or fake 
originals. This simple fact was well known to artists in the past, and 
it's why they guarded their secrets well.

It is only in an era when many artists have little or no technical 
talent that such an obviously false statement as Mr. Wagner's could even 
be thought to be true.

Ralph

-- 

Raffael Cavallaro, Ph.D.
raffael at mediaone.net


From order1 at zxmail.com  Sun Apr  9 23:36:58 2000
From: order1 at zxmail.com (order1 at zxmail.com)
Date: Mon, 10 Apr 2000 03:36:58 GMT
Subject: FREE MONEY MAKING SYSTEM
Message-ID: <u%bI4.34549$m54.67242@typhoon2.san.rr.com>

Check it out for yourself http://www.geocities.com/freemoneymakingsystem/index.html


From kragen at dnaco.net  Sun Apr 23 01:39:05 2000
From: kragen at dnaco.net (Kragen Sitaker)
Date: Sun, 23 Apr 2000 05:39:05 GMT
Subject: code not true?
References: <Pine.BSF.4.10.10004230013310.29461-100000@chi.pair.com>
Message-ID: <Z%vM4.1491$RM6.3518433@news-east.usenetserver.com>

In article <Pine.BSF.4.10.10004230013310.29461-100000 at chi.pair.com>,
Michal Wallace (sabren) <sabren at manifestation.com> wrote:
>> Floating point number on a binary computer are strange. They're not that
>> accurate. 0.07 * 100 != 7.0.
>
>Why is this the case? I mean.. I understand why decimals are hard to
>show in binary... But why can't we use variably long strings, or
>something? Is it just a speed issue?

In base 3, one third is 0.1.  Two thirds is 0.2.

If we convert these base 3 numbers into finite-length decimal
fractions, we get something like 0.3333333 and 0.6666667.  If we
multiply these numbers by three, we get 0.9999999 and 2.0000001.  You
can use more digits of precision if you want, but no finite number of
digits of precision will be enough to get exactly 1 and exactly 2
(which are mathematically the right answer to three times one third and
three times two thirds.)

Analogously, you can't take fractions like 0.2 and convert them to
binary and expect to get exactly 0.2, no matter how many bits you use.
0.5 is fine; 0.2, not.

That said, it *is* just a speed issue --- there are representations
other than floating point, like the ordered-pair rational-number
representation, that give the right answer but are slower to work
with.
-- 
<kragen at pobox.com>       Kragen Sitaker     <http://www.pobox.com/~kragen/>
The Internet stock bubble didn't burst on 1999-11-08.  Hurrah!
<URL:http://www.pobox.com/~kragen/bubble.html>
The power didn't go out on 2000-01-01 either.  :)


From hoel at GermanLloyd.org  Thu Apr 27 10:49:46 2000
From: hoel at GermanLloyd.org (Berthold Hoellmann)
Date: Thu, 27 Apr 2000 16:49:46 +0200
Subject: mail forwarding not working on the starship?
Message-ID: <3908538A.E74456A5@GermanLloyd.org>

Hello,

Sorry for posting this to the news group, but I'm not shure whom to talk
to with my problem.

Now that my .starship_info is set up to forward my mail, any test mail
send to my starship account still seems to be lost somwhere without any
error message. My starship address is "bhoel at starship.python.net" and
mail should be forwarded to "se6y095 at public.uni-hamburg.de". 

Is anyone out there who can help me?

Thanks

Berthold
-- 
email: hoel at GermanLloyd.org
   )
  (
C[_]  These opinions might be mine, but never those of my employer.


From tseaver at starbase.neosoft.com  Mon Apr  3 00:35:28 2000
From: tseaver at starbase.neosoft.com (Tres Seaver)
Date: 2 Apr 2000 23:35:28 -0500
Subject: New Features in Python 1.6
References: <200004011740.MAA04675@eric.cnri.reston.va.us> <8c84n9$pf0$3@newshost.accu.uu.nl>
Message-ID: <C84330F9D2351C56.FB77D2842BAF1633.D056A992DD04BE59@lp.airnews.net>

In article <8c84n9$pf0$3 at newshost.accu.uu.nl>,
Martijn Faassen <m.faassen at vet.uu.nl> wrote:
>Guido van Rossum <guido at python.org> wrote:
>> New Features in Python 1.6
>> ==========================
>
>> With the recent release of Python 1.6 alpha 1, a lot of people have
>> been wondering what's new.  This short note aims to explain the major
>> changes in Python 1.6.
>
>You almost got me! I didn't read the rest of the thread, and I was already
>thinking about why all these would be bad ideas, and then I realized it
>was posted on April 1. Good one!

Got me, too!  I actually have a page in my notebook titled, "Perlish crap,"
summarizing the first four points (it took me that long to get it).

Honi-soit-qui-mal-y-pense'ly,

Tres.
-- 
---------------------------------------------------------------
Tres Seaver        tseaver at digicool.com     http://www.zope.org
Digital Creations


From darius at connect.com.au  Mon Apr 10 09:16:40 2000
From: darius at connect.com.au (darius at connect.com.au)
Date: 10 Apr 2000 13:16:40 GMT
Subject: Tuples -- who needs 'em
References: <AF223B12F1FDD2119253006097C4FCEC0AD357@cats.pdri.xerox.com> <8chol6$6em$1@newshost.accu.uu.nl>
Message-ID: <8csk7o$b54$1@perki.connect.com.au>

Martijn Faassen <m.faassen at vet.uu.nl> wrote:
: People tend to use classes for this purpose. You can give a class a public
: interface (enforced or not) with only get_() methods, for instance. That
: gets you just about the same thing as an immutable dictionary would give
: you. Again, a dictionary tends to contain a larger number of homogenous
: objects, while an object tends to contain a smaller number of heterogenous
: objects.

Incidentally, I'd give my right arm for a non-mutable dictionary.  Why?
Because certain methods (in the new module, plus the exec keyword, as classic
examples) won't let me pass anything but a "true" dictionary in as namespace
- and I want to run methods in a namespace they cannot modify.

Would be even better if you could use user classes in those places, but I'd
happily settle for non-mutable dictionaries ;)

KevinL


From anthony at dstc.edu.au  Fri Apr 28 04:56:53 2000
From: anthony at dstc.edu.au (Anthony J Wilkinson)
Date: Fri, 28 Apr 2000 18:56:53 +1000 (EST)
Subject: Concatening string and integer
In-Reply-To: <39087D99.3CAEF5AA@visionart.com>
Message-ID: <Pine.OSF.4.20.0004281848300.3262-100000@azure.dstc.edu.au>

On Thu, 27 Apr 2000, Pete Shinners wrote:

> you can also use the back ticks as a shortcut for the str()
> builtin. you can do it like this;
> 
> strTemp = "I make a test number: " + `iMyNumber` + " once again"

Actually the back ticks appear to be a shortcut for repr() not str() 
which in the above case give the same result but won't for all objects,
for example:

>>> class aclass:
...     def __str__(self):
...         return 'spam'
...
>>> a = aclass()
>>> `a`
'<__main__.aclass instance at f93a8>'
>>> str(a)
'spam'
>>> repr(a)
'<__main__.aclass instance at f93a8>'

Cheers,
Anthony
_____________________________________________________________________
Anthony J Wilkinson                               anthony at dstc.edu.au
Software Engineer                                http://www.fnorb.com
DSTC Pty Ltd                                     Ph:  +61 7 3365 4310





From akuchlin at mems-exchange.org  Sat Apr 22 22:55:06 2000
From: akuchlin at mems-exchange.org (Andrew Kuchling)
Date: 22 Apr 2000 22:55:06 -0400
Subject: SSL w/ urllib
References: <3901D800.DB50D968@flashmail.com>
Message-ID: <3ditx9v851.fsf@newcnri.cnri.reston.va.us>

Jp Calderone <exarkun at flashmail.com> writes:
>  f = urllib.urlopen('https://www.mudprovider.com/cgi-bin/printvars.cgi',
> params)
> TypeError: too many arguments; expected 2, got 3

You've found a bug.  I've checked a fix into the CVS tree, so you can
either update, or just apply the patch below.  Thanks for the bug
report!

--amk

Index: urllib.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/urllib.py,v
retrieving revision 1.91
diff -C2 -r1.91 urllib.py
*** urllib.py	2000/02/04 15:28:41	1.91
--- urllib.py	2000/04/23 02:48:11
***************
*** 300,304 ****
  
      if hasattr(socket, "ssl"):
!         def open_https(self, url):
              """Use HTTPS protocol."""
              import httplib
--- 300,304 ----
  
      if hasattr(socket, "ssl"):
!         def open_https(self, url, data=None):
              """Use HTTPS protocol."""
              import httplib
***************
*** 324,328 ****
                                key_file=self.key_file,
                                cert_file=self.cert_file)
!             h.putrequest('GET', selector)
              if auth: h.putheader('Authorization: Basic %s' % auth)
              for args in self.addheaders: apply(h.putheader, args)
--- 324,334 ----
                                key_file=self.key_file,
                                cert_file=self.cert_file)
!             if data is not None:
!                 h.putrequest('POST', selector)
!                 h.putheader('Content-type',
!                             'application/x-www-form-urlencoded')
!                 h.putheader('Content-length', '%d' % len(data))
!             else:
!                 h.putrequest('GET', selector)
              if auth: h.putheader('Authorization: Basic %s' % auth)
              for args in self.addheaders: apply(h.putheader, args)





From rs59922 at swt.edu  Mon Apr 10 22:30:45 2000
From: rs59922 at swt.edu (rakesh)
Date: Mon, 10 Apr 2000 21:30:45 -0500
Subject: Installation Problem in Windows
Message-ID: <38F28E55.67280329@swt.edu>

Folks,

I am new in python. I have installed python 1.5 installer for Windows at
home. I ran the setup. I need to use teh program for both, the
compilation of python programs and to use Tk widgets. I wrote one
test.py and tried to compile it inthe python.exe window. It gave an
error "stdin". Can anybody help this newbie.




-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20000410/c6416089/attachment.html>

From grant at nowhere.  Tue Apr 18 11:28:04 2000
From: grant at nowhere. (Grant Edwards)
Date: Tue, 18 Apr 2000 15:28:04 GMT
Subject: Integer solutions to linear equation?
Message-ID: <slrn8fovo2.nq.grant@grante.comtrol.com>

This isn't really a Python question, but my example is in
Python, and there seem to be plenty of people who know a fair
bit of math here....

A friend of mine ran across a brain-teaser involving a bunch of
flowers, some magic bowls and some other camoflage text.  What
you end up with is having to solve the equation

     64x = 41y + 1
     
Where x and y are both integers.  After scratching our heads
for a while, we used the brute force approach:

for x in range(1,100):
    y = ((64*x)-1)/41
    if 64*x == 41*y+1:
        print (x,y)
                
The results:

  (25, 39)
  (66, 103)

25 was the expected solution, so we got both the equation and
the Python snippet correct.  Is there a non-iterative way to
solve a linear equation when the results are contrained to be
integers?  I don't remember anything from any of my math
classes that seems relevent, but I didn't take any anything
beyond what is required for all undergrad engineers.

-- 
Grant Edwards                   grante             Yow!  I didn't order
                                  at               any WOO-WOO... Maybe a
                               visi.com            YUBBA... But no WOO-WOO!


From mhammond at skippinet.com.au  Wed Apr 19 22:14:35 2000
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Thu, 20 Apr 2000 02:14:35 GMT
Subject: (Fwd) Re: Formatting Excel cells with PythonCOM
References: <1E141210786@nds-3.azg.nl>
Message-ID: <fKtL4.17$v85.83@news-server.bigpond.net.au>

"P.J.W.S. Vrijlandt" <P.J.W.S.VRIJLANDT at INT.azg.nl> wrote in message
news:1E141210786 at nds-3.azg.nl...
> On  6 Apr 00 at 23:45, Mark Hammond wrote:
>

> I recently found a problem with this (I didn't run the test suite
> though). On my computer (Win95, Excel2000), the NumberFormat propery
> only accepts setting the time in my locale's language (i.e. Dutch).
> This is rather unfortunate, because the macrorecorder writes in

Hrm.  Looks like we are in a no-win situation here.  Many international
users want COM and makepy setup to use the default locale - which it is -
so some of the names etc get localised.  However, doing this obviously
changes the behaviour of certain functions.

Im not sure what the correct answer is, but if there is a consensus from
non US versions of COM that we should specify US English, it could be
done.

Mark.





From tjreedy at udel.edu  Wed Apr 26 00:37:47 2000
From: tjreedy at udel.edu (Terry Reedy)
Date: Wed, 26 Apr 2000 00:37:47 -0400
Subject: Metaclasses?
References: <3905F4A0.2B39C705@darmstadt.gmd.de>
Message-ID: <8e5rlu$e2m$1@news.udel.edu>

"Robb Shecter" <shecter at darmstadt.gmd.de> wrote in message
news:3905F4A0.2B39C705 at darmstadt.gmd.de...
> Hi,
>
> I read an old post about metaclasses (included on the cd-rom with the
> book "Mit Python programmieren"), and got very interested, because I
> imagine that by making a new metaclass, I could maybe fix the one thing
> about Python that annoys me: the obligatory use of "self" in method
> declarations and variable access.  (Python's syntax is so clean and neat
> in all other areas, I feel like it's a shame that there's excess
> verbosity in this one place.)

The literal string 'self' is not obligatory.  's' or 'me' work just as
well, as long as you are consistent.





From wtanksle at dolphin.openprojects.net  Thu Apr  6 13:54:41 2000
From: wtanksle at dolphin.openprojects.net (William Tanksley)
Date: Thu, 06 Apr 2000 17:54:41 GMT
Subject: Tuples -- who needs 'em
References: <AF223B12F1FDD2119253006097C4FCEC0AD357@cats.pdri.xerox.com> <8chol6$6em$1@newshost.accu.uu.nl> <026e01bf9fe7$9554fd20$74eb0b18@stcla1.sfba.home.com>
Message-ID: <slrn8epjr1.mne.wtanksle@dolphin.openprojects.net>

On Thu, 6 Apr 2000 09:45:57 -0700, Bob Alexander wrote:
>As you say, the functional programming paradigm can be nicely expressed in
>Python, and is not dependent on tuples.

Yes.  AFAIK, the only thing dependant on the immutability of tuples is
dictionaries, and only when you use a tuple as a key to a dictionary.  If
you're willing to get rid of being able to key dictionaries with tuples,
you could get rid of tuples.

But tuples aren't the only immutable sequences; so are strings.  That's
why you can key dictionaries with them.  I don't think you'd want to get
rid of that ability.

>Bob

-- 
-William "Billy" Tanksley


From ivanlan at callware.com  Tue Apr  4 17:34:32 2000
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Tue, 04 Apr 2000 15:34:32 -0600
Subject: Two questions
Message-ID: <38EA5FE8.853E9B75@callware.com>

Hi All--
I'm finally getting a chance to use Python at work.  I have two
questions:

1)  Is there a way to run silent (i.e., we determine where things go and
the customer never sees any wizards) installs of Python, Tcl/Tk and
Win32all?

2)  What stuff do I need to use msaccess with Python?  Do I need to use
Win32all, or can I use Tkinter for my (minimal) GUI, with msaccess?

<there-that-didn't-hurt-did-it>-ly y'rs,
Ivan
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
http://www.pauahtun.org 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
Author:  Teach Yourself Python in 24 Hours



From skodela at my-deja.com  Thu Apr 27 20:04:18 2000
From: skodela at my-deja.com (Sindh)
Date: Fri, 28 Apr 2000 00:04:18 GMT
Subject: Python does not let me format time.
Message-ID: <8eakhq$797$1@nnrp1.deja.com>

Hi folks

I have a file full of date, value pairs. Date is like MM/DD/YY. My
problem is I want to format this file since it contains a few thousand
lines of data , using python. I tried using string and time modules. I
want to format the time value to a float which is returned by
time.gmtime normally. I tried to split it and to use parts of it to make
a tuple and to use time.mktime but it says illegal operation for
builtin.

Summary:

data is:
	03/11/98 abcd
	03/12/98 efgh

I split it to
	x=string.split(n[0],' ')
	which gave me x='03/11/98'
Now I split it again by
	x=string.split(x,'/')
	which gave a tuple of '03','11','98'
Now I tried  to make time as
time.mktime(x[2],x[0],x[1],'','','','','','')
The problem is it says illegal operation for builtin. I tried to use the
string val and int val as the arguments for mktime.

Anyone with any helpful ideas!!

If there is already a moduls to do the same , much helpful.

Thanks
sreekant

--
A man needs to sleep for 36 hours a day atleast.


Sent via Deja.com http://www.deja.com/
Before you buy.


From jason at jedi.net  Thu Apr 13 15:51:32 2000
From: jason at jedi.net (Jason Crickmer)
Date: Thu, 13 Apr 2000 19:51:32 GMT
Subject: odd exception in JPython execution
Message-ID: <8d58ft$5jv$1@nnrp1.deja.com>

I am using JPython as an evalator for Java.  It works in some cases, and
not others.  But, the failures are consistent.  Basicly, given a
PythonInterpreter, I set several variables into it and then call eval()
(I have also tried exec(), but it has the same problems).

Here is an example of what happens:

  PythonInterpreter interp = new PythonInterpreter();

  ...

  interp.set("scale", "0.5");

  ...


  String result = interp.eval("scale").toString();


At this point, an exception in thrown:

Traceback (innermost last):
  (no code object) at line 0
java.lang.NoClassDefFoundError: org/python/core/PyFunctionTable
 at java.lang.ClassLoader.defineClass0(Native Method)
 at java.lang.ClassLoader.defineClass(ClassLoader.java:442)
 at java.lang.ClassLoader.defineClass(ClassLoader.java:392)
 at
org.python.core.BytecodeLoader.loadClassFromBytes(BytecodeLoader.java:76)

 at org.python.core.BytecodeLoader.loadBytes(BytecodeLoader.java:92)  at
org.python.core.BytecodeLoader.makeCode(BytecodeLoader.java:103)  at
org.python.core.Py.compile(Py.java:1341)  at
org.python.core.Py.compile(Py.java:1354)  at
org.python.core.__builtin__.compile(__builtin__.java:188)  at
org.python.core.Py.exec(Py.java:946)  at
org.python.util.PythonInterpreter.exec(PythonInterpreter.java:111)  at
com.hire.services.export.PythonEntityResolver.resolveEscriptEntity(PythonEnti
tyResolver.java:551)

 at
com.hire.services.export.PythonEntityResolver.resolveEntity(PythonEntityResol
ver.java:471)

 at
com.hire.er.EREntityResolver.resolveEntity(EREntityResolver.java:319)
 ...


I am using Sun's Java 1.2.2-RC2 JRE with native theads and no JIT on
Linux.  I am also using JPython 1.1.  Anyone have any ideas?

Thanks,
Jason


Sent via Deja.com http://www.deja.com/
Before you buy.


From mwh21 at cam.ac.uk  Wed Apr 26 08:01:03 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 26 Apr 2000 13:01:03 +0100
Subject: ungreedy regular expressions
References: <slrn8gdlaj.k0s.ckotso@theseas.softlab.ece.ntua.gr>
Message-ID: <m3n1mhoyv4.fsf@atrus.jesus.cam.ac.uk>

ckotso at cs.ntua.gr (Constantinos A. Kotsokalis) writes:

> I recently added two lines of code to the python distribution, in order
> to support ungreedy regular expressions (which are supported by pcre
> but I could find no way to use them in python code). If there is some
> way already that I am not aware of, please e-mail me :-) If you are
> interested in having it available, edit Modules/pcremodule.c and add
> the line:

Aren't the *?,+? patterns "ungreedy"?

>>> re.match("(a+)(.*)","aaabbb").groups()
('aaa', 'bbb')
>>> re.match("(a+?)(.*)","aaabbb").groups()
('a', 'aabbb')

HTH,
Michael

-- 
  it's not that perl programmers are idiots, it's that the language
  rewards idiotic behavior in a  way that no other language or tool 
  has ever done                        -- Erik Naggum, comp.lang.lisp


From thor at localhost.localdomain  Thu Apr 13 15:40:38 2000
From: thor at localhost.localdomain (Manuel Gutierrez Algaba)
Date: 13 Apr 2000 19:40:38 GMT
Subject: Crosswords for Linux Backtracking recursive crossword generator
Message-ID: <slrn8fcfm8.223.thor@localhost.localdomain>

Organization: 4th Invasion army, Alpha Centurian Armada
Reply-To: irmina at ctv.es
Followup-To: 
Keywords: 
Summary: 

-------
Dear pythonistas, 

I'm the current maintainer of Crosswords For Linux
http://www.ctv.es/USERS/irmina/cruo/cruolinux.htm

And I'm the process of releasing all the generators and filters.
I've finished the translation of the generator, which may be
interesting for many of you , it's only 400 lines code and 
fast enough : 

http://www.ctv.es/USERS/irmina/cruo/genaut.py 

It's runnable, it doesn't need threads, nor other special libraries
nor data, just download and : python genaut.py 

Comments/improvements are welcome!

MGA




From jorn at spekter.no  Thu Apr 27 08:39:43 2000
From: jorn at spekter.no (=?iso-8859-1?Q?J=F8rn?= Eilertsen)
Date: Thu, 27 Apr 2000 14:39:43 +0200
Subject: python + linux + cron = problem
References: <39081703.48B64D1A@spekter.no> <3908265E.A3C21409@home.com>
Message-ID: <3908350F.922D7B4E@spekter.no>

here's the whole script, well the part that gives me grief at least
:
#!/usr/bin/python
import commands

koblinger = commands.getoutput('df | grep eggum | cut -d 1 -f 1 | cut -d / -f
4')

I've tried to put the shell command into a shell script and run
commands.getoutput on the script but with the same result. Here's the cron
entry:
0 1 * * *   /usr/local/bin/sjekkmounts
and I've tried sh /usr/local/bin/sjekkmounts

Jorn

Doug Hellmann wrote:

> J?rn Eilertsen wrote:
> >
> > Hi,
> >
> > I've posted this earlier but noone replied so I'm trying again
> >
> > I have this small program which includes the following line og code:
> >  mounts = commands.getoutput('df | grep eggum | cut -d 1 -f 1 | cut -d /
> > -f 4')
> >
> > this runs fine from the commandline but when run as a cronjob I get the
> > following error:
> >
> > File "/usr/lib/python1.5/commands.py", line 42, in getoutput
> >     return getstatusoutput(cmd)[1]
> >   File "/usr/lib/python1.5/commands.py", line 53, in getstatusoutput
> >     sts = pipe.close()
> > IOError: (10, 'No child processes')
> >
> > The use of pipes seems to be the trouble here.
> >
> > Any help on how the get around this would be greatly appreciated.
> >
> >
>
> If the problem is really having pipes in the command, why not put your
> shell commands in a shell script, then run commands.getoutput() on that?
>
> If the program runs from the command line but not cron, the problem is
> more likely to be related to PATH problems within the cron environment.
> How do you set up the cron job?
>
> Doug



From mhammond at skippinet.com.au  Fri Apr  7 22:54:36 2000
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Sat, 08 Apr 2000 02:54:36 GMT
Subject: Actually Running a function from a DLL
References: <38EE87A3.4080802@draper.com>
Message-ID: <MbxH4.1025$4v.5306@news-server.bigpond.net.au>

The only way to do this is to use the "calldll" or "dynawin" or
whatever it is called today extensions by Sam Rushing.  Can be
obtained from www.nightmare.com.  (A quick look indicates that
you need the dynawin stuff, even though it wont look like it :-)

Mark.

"Peter Olsen" <pcolsen at draper.com> wrote in message
news:38EE87A3.4080802 at draper.com...
I'm writing ask for a quick explanation of how to load and use
dll files
from Python running under MS Windows.

I have installed PythonWin and win32all, but I seem to be
completely
unable to run a function from a dll. I'm loading the dll with
win32api.LoadLibrary, but I can't figure out how to access the
functions. I've made some statements about how fast we can learn
Python
and the start development, but if I don't find an answer quickly,
I may
have to eat my words. If the best answer is "RTFM," please give
me a
hint as to where in TFM I should R.

I would greatly appreciate receiving answers by email because my
company
makes it hard to read news.

Thanks!

Peter






From claird at starbase.neosoft.com  Thu Apr 20 09:57:08 2000
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 20 Apr 2000 08:57:08 -0500
Subject: python templating system
References: <Pine.BSF.4.10.10004191555250.14584-100000@chi.pair.com>
Message-ID: <CDA59A302BEB26AC.DE30D79F63ED12C2.5C8017889EA62DAB@lp.airnews.net>

In article <Pine.BSF.4.10.10004191555250.14584-100000 at chi.pair.com>,
Michal Wallace (sabren) <sabren at manifestation.com> wrote:
>On Wed, 19 Apr 2000 pozman at my-deja.com wrote:
>
>> Does anyone know of a templating system (other than zope) that is
>> implemented with python [ comparisons would be embperl with perl]
> 
>http://zebra.sourceforge.net/
			.
			.
			.
According to the text currently at that page, Zebra does *not* now
interpret Python.  However, I collect systems which do at <URL:http://
starbase.neosoft.com/~claird/comp.lang.python/web_python.html>.
-- 

Cameron Laird <claird at NeoSoft.com>
Business:  http://www.Phaseit.net
Personal:  http://starbase.neosoft.com/~claird/home.html


From ullrich at math.okstate.edu  Sun Apr  9 16:41:04 2000
From: ullrich at math.okstate.edu (David C. Ullrich)
Date: 9 Apr 2000 20:41:04 GMT
Subject: Newbie: urllib.urlretrieve()
References: <221ee89a.3bf4ca4b@usw-ex0102-016.remarq.com>
Message-ID: <01bfa265$5fcf9720$973f8ad1@daves-dell>


ez4gl <ez4glNOezSPAM at hotmail.com.invalid> wrote in article
<221ee89a.3bf4ca4b at usw-ex0102-016.remarq.com>...
> I'm trying to clone some Intranet pages for a contract
> assignment and thought Python would be a good tool because of
> it's urllib module.  However, when trying to use
> urllib.urlretrieve(), I have found that if you specify the
> second argument, the file for writing to has to already exist.

	Are you certain this is the problem? I don't have any
problem using urlretrieve(url, newfilename). 


From ivanlan at callware.com  Tue Apr  4 17:56:39 2000
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Tue, 04 Apr 2000 15:56:39 -0600
Subject: Two questions
References: <38EA5FE8.853E9B75@callware.com> <01f801bf9e80$5a522a80$29dc9f26@unitedelectric>
Message-ID: <38EA6517.978953E4@callware.com>

Hi All--

David Fisher wrote:
> 
> ----- Original Message -----
> From: Ivan Van Laningham <ivanlan at callware.com>
> To: Python Mailing List <python-list at python.org>
> Sent: Tuesday, April 04, 2000 4:34 PM
> Subject: Two questions
> 
> > 2)  What stuff do I need to use msaccess with Python?  Do I need to use
> > Win32all, or can I use Tkinter for my (minimal) GUI, with msaccess?
> 
> What do you want to do with ms access?
> 
> dnf
> 

Query for information, using a not-very-complicated schema.  Read a file
and, after massaging the data, stuff the information into the db.

As reading the file progresses and the information gets transferred to
the db, update a progress indicator (that's the extent of the GUI;-)

Metta,
Ivan
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
http://www.pauahtun.org 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
Author:  Teach Yourself Python in 24 Hours



From roger at netconnect.no  Tue Apr 11 08:34:08 2000
From: roger at netconnect.no (Roger Baklund)
Date: Tue, 11 Apr 2000 14:34:08 +0200
Subject: Huh? func_defaults, default values in function calls
Message-ID: <8cv5qj$35p$1@zinc.intern.netconnect.no>

>>> def a(x='',y='',z={}):
...     if x:
...       z['x'] = x
...     if y:
...       z['y'] = y
...     print str(z)
...
>>> a('a')
{'x': 'a'}
>>> a('b')
{'x': 'b'}
>>> a('b','c')
{'x': 'b', 'y': 'c'}
>>> a('a')
{'x': 'a', 'y': 'c'}
>>>

I don't get it. Why isn't the dictionary reset to {'x':'a'} when i call
the function the last time?

>>> a.func_defaults
('', '', {'x': 'a', 'y': 'c'})

Why is not {} the default for the z parameter?

--
Roger




From bill at libc.org  Tue Apr 25 01:36:51 2000
From: bill at libc.org (Bill Anderson)
Date: Mon, 24 Apr 2000 23:36:51 -0600
Subject: The Simple Economics of Open Source
References: <14591.15630.192472.440375@goon.cnri.reston.va.us> <slrn8fv2rd.1kt.neelk@brick.cswv.com> <3900611F.3B525BDA@cfdrc.com>  <FtDG4z.774@world.std.com> <slrn8g26tu.p7o.neelk@brick.cswv.com> <39011EEF.4BAEC3C9@tcac.net> <22042000.3@sanctum.jae.ddns.org>
Message-ID: <39052EF3.EEF864AD@libc.org>

"Juergen A. Erhard" wrote in part:
...
> [1] The human-accessible part of this solar system is finite, right?
> So there simple cannot be an infinite number of pies, even if you
> could bake a pie out of one atom... ;-)


Only when considered in a finite amount of time.
There is always tomorrow, and if there isn't, the number of pies
available today is irrelevant. :^)

-- 
In flying I have learned that carelessness and overconfidence are 
usually far more dangerous than deliberately accepted risks. 
          -- Wilbur Wright in a letter to his father, September 1900


From nospam.newton at gmx.li  Tue Apr  4 15:37:36 2000
From: nospam.newton at gmx.li (Philip 'Yes, that's my address' Newton)
Date: Tue, 04 Apr 2000 19:37:36 GMT
Subject: New Features in Python 1.6
References: <200004011740.MAA04675@eric.cnri.reston.va.us> <_pyF4.25562$E85.640966@news1.rdc1.md.home.com> <slrn8edfmk.185.wtanksle@dolphin.openprojects.net> <ltvh21ywfs.fsf@asfast.com> <oq66txol0w.fsf@titan.progiciels-bpi.ca> <GOpG4.5050$74.81782@newsc.telia.net>
Message-ID: <38ea40a1.47458347@news.nikoma.de>

On Tue, 04 Apr 2000 17:40:54 GMT, "Fredrik Lundh" <effbot at telia.com>
wrote:

>why not just use two different compression algorithms repeatedly,
>until you've squeezed the file down to a suitable size?  it takes a
>little work, but you can design them so that one of them can always
                  ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
>strip away at least one bit from the input stream.
 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

Yes? I challenge you to provide any algorithm that will strip away one
bit from the four sequences 11, 10, 01, and 00, while still being able
to restore the previous state. (Unless you were talking about lossy
compression, in which case you might as well say "my compression
algorithm just drops off the leftmost bit" or similar, which will
*always* compress any non-empty file.)

Cheers,
Philip
-- 
Philip Newton <nospam.newton at gmx.li>


From just at letterror.com  Wed Apr 26 10:13:13 2000
From: just at letterror.com (Just van Rossum)
Date: Wed, 26 Apr 2000 15:13:13 +0100
Subject: Python 1.6a2 Unicode bug (was Re: comparing strings and ints)
Message-ID: <l03102805b52ca7830b18@[193.78.237.154]>

I wrote:
>A utf-8-encoded 8-bit string in Python is *not* a string, but a "ByteArray".

Another way of putting this is:
- utf-8 in an 8-bit string is to a unicode string what a pickle is to an
object.
- defaulting to utf-8 upon coercing is like implicitly trying to unpickle
an 8-bit string when comparing it to an instance. Bad idea.

Defaulting to Latin-1 is the only logical choice, no matter how
western-culture-centric this may seem.

Just





From javanet at dynacap.com  Thu Apr 27 15:24:17 2000
From: javanet at dynacap.com (Arnaldo Riquelme)
Date: Thu, 27 Apr 2000 11:24:17 -0800
Subject: Newbie Question
Message-ID: <sgh4qdmbqtj130@corp.supernews.com>

Suppose I have a file name tick.tkr that looks like this

SUNW.
MFST.
FOO.
SPAM.
MOO.
End_Of_File

I want to read the file into a list

myfile = open("tick.tkr", "r")
mylist = myfile.readlines()

Now I have mylist with the contents of tick.tkr. "mylist" looks like this
['SUNW.\012', 'MSFT.\012', ..........'End_of_File/012']

How do I read the file into a list without getting the newline character, so
I can accomplish this

for tic in mylist:
    #create file tic.prn

I know the other P language has something like 'chop' or something like
that, but I'm not going there.

Thanks


Arnaldo,






From ngps at madcap.  Fri Apr  7 11:55:51 2000
From: ngps at madcap. (Ng Pheng Siong)
Date: 7 Apr 2000 15:55:51 GMT
Subject: Socket Programming Question
References: <aZdH4.95109$AT6.106753@dfw-read.news.verio.net>
Message-ID: <8cl0e7$ju8$1@violet.singnet.com.sg>

According to Ray Van Dolson <rayvd at nospam.firetail.org>:
> I'm trying to write a simple chat server w/ Python sockets that supports 
> multiple connections and am having a little trouble... 

Try using SocketServer.ThreadingTCPServer.

Here's an excerpt of the docu:

    The SocketServer module simplifies the task of writing network 
    servers. 

    There are four basic server classes: ...

    These four classes process requests synchronously; each request 
    must be completed before the next request can be started. This 
    isn't suitable if each request takes a long time to complete, 
    because it requires a lot of computation, or because it returns 
    a lot of data which the client is slow to process.

    The solution is to create a separate process or thread to handle 
    each request; the ForkingMixIn and ThreadingMixIn mix-in classes 
    can be used to support asynchronous behaviour.


-- 
Ng Pheng Siong <ngps at post1.com> * http://www.post1.com/home/ngps



From scherbi at bam.com  Fri Apr  7 08:34:56 2000
From: scherbi at bam.com (Bill Scherer)
Date: Fri, 07 Apr 2000 08:34:56 -0400
Subject: Newbie: How to do an exit ?
References: <38EDC1DA.62CBFCF3@sxb.bsf.alcatel.fr>
Message-ID: <38EDD5F0.B1D5042@bam.com>

import sys
sys.exit(0)


 - or -

raise SystemExit

Pascal Fremaux wrote:

> How to exit from a program ?
> I don't find a command such as exit(0) in C
>
> --
>  Pascal Fremaux, SSII Alten
>  Study Engineer at Alcatel Telecom
>  R&D, Illkirch, France
>
> --
> http://www.python.org/mailman/listinfo/python-list

--
William K. Scherer
Sr. Member of Applications Staff
Verizon Wireless






From mikael at isy.liu.se  Mon Apr 17 11:31:30 2000
From: mikael at isy.liu.se (Mikael Olofsson)
Date: Mon, 17 Apr 2000 17:31:30 +0200 (MET DST)
Subject: Another question
In-Reply-To: <dVFK4.41$vr2.5817@news.uswest.net>
Message-ID: <XFMail.000417173130.mikael@isy.liu.se>

On 17-Apr-00 Jeff Massung wrote:
 >  Okay, another quick question. If it is quicker to do:
 >  
 >  
 >  import random
 >  from random import randint
 >  for i in range(1,1000000):
 >      print randint(1,10)
 >  
 >  
 >  than this:
 >  
 >  
 >  import random
 >  for i in range(1,1000000):
 >      print random.randint(1,10)

I'm not the expert here, but I think it's just as quick. Someone may
correct me, though.

 >  Why would I not do "from <module> import *" with every module I load?

Because you may redefine functions if you do

  from <module> import *

If <module> defines its own max(), you loose the original max(). But if
you instead do 

  import <module>

you still have the original max(), and the max() in <module> is called
by 

  <module>.max()

/Mikael

-----------------------------------------------------------------------
E-Mail:  Mikael Olofsson <mikael at isy.liu.se>
WWW:     http://www.dtr.isy.liu.se/dtr/staff/mikael
Phone:   +46 - (0)13 - 28 1343
Telefax: +46 - (0)13 - 28 1339
Date:    17-Apr-00
Time:    17:26:14

This message was sent by XF-Mail.
-----------------------------------------------------------------------


From thomas at xs4all.net  Mon Apr 17 10:51:02 2000
From: thomas at xs4all.net (Thomas Wouters)
Date: Mon, 17 Apr 2000 16:51:02 +0200
Subject: read/write file problem
In-Reply-To: <38FB1375.2FF65B96@sec.noaa.gov>; from jvickroy@sec.noaa.gov on Mon, Apr 17, 2000 at 07:36:53AM -0600
References: <38FB1375.2FF65B96@sec.noaa.gov>
Message-ID: <20000417165102.G15664@xs4all.nl>

On Mon, Apr 17, 2000 at 07:36:53AM -0600, j vickroy wrote:

> I am attempting to read and, subsequently, write a jpeg file.
> The following interactive session illustrates the problem I am having
> (i.e., if the source file is read byte-by-byte into memory and
> subsequently
> rewritten using the file.write() method, only the initial portion of the
> file appears to be written).
> How should I be doing this in Python 1.5.2.


> >>> file = open (file_name, 'w')
> >>> for i in xrange (0,source_byte_cnt):
> ...  file .write (buf[i])

[...]

> >>> file = open (file_name, 'w')

This overwrites the previously written portion, and since you write the same
amount of bytes, it seems unchanged. Change this to:

file = open (file_name, 'a')

and it will work as expected ('a' is append, which means writing to the file
from the end onward.)

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From drs at labs.agilent.com  Fri Apr 14 14:12:13 2000
From: drs at labs.agilent.com (David Smith)
Date: Fri, 14 Apr 2000 11:12:13 -0700
Subject: No strides? (was Re: List mapping?)
References: <38f5d68e@news.xtml.co.uk>
Message-ID: <38F75F7D.1C73D177@labs.agilent.com>

Why aren't strides implemented for slices of lists and tuples?  I would
have liked to suggest

	map( None, a[0::2], a[1::2] )

			David Smith


Nick Trout wrote:
> 
> Is there a nice way of pairing up the members to form tuple pairs in another
> list? ie. like you might to processing input args.
> 
> ie. changing ['/a','1','/b','2','/c','3'] into [ ('/a','1'), ('/b','2'),
> ('/c','3') ]
> 
> eg:
> 
> args = '/a 1 /b 2 /c 3'
> listargs= string.split(args)
> tuplist = []
> i=0
> while i < len(listargs):
>     tuplist.append( (listargs[i],listargs[i+1]) )
>     i = i+2
> for i in tuplist:
>     processargs(i)
> 
> or:
> 
> args = '/a 1 /b 2 /c 3'
> for i in string.split(args):
>     i = argflag
>     # i want next arg now as a parameter!!!
> 
> Can I (you!) get rid of the long winded while loop cleverly?!!  :-)
> 
> Thanks in advance,
> Nick.


From eviltofu at rocketmail.com  Thu Apr 20 12:33:44 2000
From: eviltofu at rocketmail.com (Jerome Chan)
Date: Thu, 20 Apr 2000 16:33:44 GMT
Subject: Creating Classes Dynamically?
Message-ID: <eviltofu-783F16.12334920042000@news.earthlink.net>

Is there a way to create an object of a class from a string?


From banderson at boi.hp.com  Fri Apr 14 16:36:44 2000
From: banderson at boi.hp.com (Bill Anderson)
Date: Fri, 14 Apr 2000 14:36:44 -0600
Subject: Python and VTK and TK
References: <38F51AB1.29FB4C1D@earthlink.net> <38F528C8.2C18364F@gte.net>
Message-ID: <38F7815C.57EC8A83@boi.hp.com>

sue wrote:
> 
> Scott -
> 
> Unfortunately, it depends on your definition of moderate.
> Can you elaborate a bit?
> 
> You might want to start a dirty prototype in TK, to check
> your overall design, and go to VTK/FLTK when time (speed or
> due date :) becomes an issue.
> 
> You might want to check pyFLTK as well.  I've compiled all
> three without any major problems, but if I remember
> correctly, VTK ends up wanting several other small libs.

I've been playing with pyFLTK lately. Nice toolkit/bindings, though
documentation was found severly lacking.


--
Bill Anderson 			Linux/Unix Administrator, Security Analyst
ESBU (ARC)			bill_anderson at boi.hp.com
My opinions are just that; _my_ opinions.


From psl at mitre.org  Tue Apr 25 15:40:34 2000
From: psl at mitre.org (Pete)
Date: Tue, 25 Apr 2000 15:40:34 -0400
Subject: question about calldll
Message-ID: <3905F4B2.60ED44DA@mitre.org>

I am trying to use calldll to load a dll and call a function from that
dll in python.  I am having a problem when I use the get_proc_address. 
I have tried using several different strings as the procedure_name,  but
I always get a return value of 0,  meaning that there was an error in
trying to find the address (probably no procedure exits with that
name).  I tried using the name that is shown under the name column in
the output from dumpbin,  but I got the same error.  Can anyone tell me
what I'm doing wrong?

Also,  why is there both a in_args_format_string and an argument_tuple, 
what is the difference between these two?  And is the
result_format_string just anyold string?  It doesn't need to be defined
to be anything specific?


From effbot at telia.com  Mon Apr 24 20:36:43 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Tue, 25 Apr 2000 00:36:43 GMT
Subject: Simple list.append() question
References: <meh9-5ED8CD.11192424042000@news.cit.cornell.edu> <ZHZM4.360$qg7.2949@news.uswest.net> <39046FA2.603EAFE5@roguewave.com> <IL3N4.4283$L3.47222@typhoon.midsouth.rr.com>
Message-ID: <vM5N4.1522$wYl.188802560@newsb.telia.net>

Bill Wilkinson <bwilk_97 at yahoo.com> wrote:
> l = [[][:]] * 3

same problem as the original code:

>>> l = [[][:]] * 3
>>> l
[[], [], []]
>>> l[0].append(1)
>>> l
[[1], [1], [1]]

</F>




From ivanlan at callware.com  Tue Apr  4 21:54:28 2000
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Tue, 04 Apr 2000 19:54:28 -0600
Subject: What is python's language level?
References: <38EA1B8F.A797A130@lmco.com> <38E9DDD3.3313FF6@bioreason.com> <slrn8ekddq.gm3.zawrotny@gecko.sb.fsu.edu> <roy-D857DA.21212304042000@netnews.nyu.edu>
Message-ID: <38EA9CD4.3473B910@callware.com>

Hi All--

Roy Smith wrote:
> 
> zawrotny at gecko.sb.fsu.edu (Michael Zawrotny) wrote:
> > Lets start with the fact that nroff and html are markup
> > languages not programming languages.
> 
> HTML certainly is markup, but I don't know if I'd call nroff markup.
> It's much more like a programing language.
> 
> Nroff has arithmetic expressions and operators, conditional execution,
> several different forms of data structures (number registers, text
> diversions), user-definable functions with argument passing, traps, etc.
> I don't remember if it supports recursion or not, but I think so.
> 

I believe it does, but it's been at least 8 years since I did serious
work in n/troff.

> I will admit, the syntax leaves something to be desired (an obfuscated
> nroff contest would be pointless because it'd be too hard to pick one
> winner). 

Yes indeed.  Any n/troff source file could qualify.  Macro files, by
definition, transcend obfuscated.  I wonder what's next up that ladder?

I once wrote a macro to draw a padlock character.  I really was a very
nice padlock, too.  If anyone's interested, I could post it (or write to
me b/c).

> I'll leave it to the theory wonks to figure out if it's Turing
> complete or what not, but it's very much more than just a markup
> language.
> 

It's most certainly Turing-complete.  It ain't what you call convenient
or readable, but once you understand what you're doing you can write
anything you want in it.  Including calculators and prime-number sieves.

<it's-still-moose-turd-pie-though>-ly y'rs,
Ivan;-)
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
http://www.pauahtun.org and
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
Author:  Teach Yourself Python in 24 Hours



From richard_chamberlain at ntlworld.com  Thu Apr 20 03:30:22 2000
From: richard_chamberlain at ntlworld.com (Richard Chamberlain)
Date: Thu, 20 Apr 2000 08:30:22 +0100
Subject: win32all-128.exe availability?
References: <38fd01b2$0$10982@senator-bedfellow.mit.edu>
Message-ID: <8dmbm3$asu$1@gxsn.com>

Sorry folks,

I've got 128 installed but I haven't got the exe - only 125. I must have
trashed it after installing.

Richard

"Robert" <soundhack at hotmail.com> wrote in message
news:38fd01b2$0$10982 at senator-bedfellow.mit.edu...
> Now that starship has crashed, I was wondering if it was possible to
> get the 128 build of Pythonwin extensions off the net? By an unfortunate
> coincidence, my own hard drive chose to crash around the time starship
> did.
>
> Thanks
> Robert
>
>




From jmassung at magpiesystems.com  Thu Apr 20 11:08:33 2000
From: jmassung at magpiesystems.com (Jeff Massung)
Date: Thu, 20 Apr 2000 09:08:33 -0600
Subject: Mac Python Questions
References: <IRjL4.271$CP3.3183@news.uswest.net> <38FE8032.DE233C29@cosc.canterbury.ac.nz>
Message-ID: <T_EL4.46$RL.9579@news.uswest.net>

On the MacPython SIG I got this response, and it works:

Add $(PYTHON):Lib:lib-tk and $(PYTHON):Extensions:Image:PIL to the paths and
that did it. The installer puts some files in the wrong places and forgets
to add the lib-tk path.

Of course, I was also notified that not all of Tkinter works on the Mac - so
be careful and test often I guess ;)

Now if we just had a color IDE... ;)

Jeff

Greg Ewing wrote in message <38FE8032.DE233C29 at cosc.canterbury.ac.nz>...
>Jeff Massung wrote:
>>
>> b) No matter what I do, I can't get Tkinter to work :(.
>
>Neither can I, and from what I've seen here, neither
>can anyone else. Tkinter just seems to be broken on
>the Mac at present.
>
>--
>Greg Ewing, Computer Science Dept,
>+--------------------------------------+
>University of Canterbury,    | A citizen of NewZealandCorp, a   |
>Christchurch, New Zealand    | wholly-owned subsidiary of USA Inc.  |
>greg at cosc.canterbury.ac.nz    +--------------------------------------+




From darrell at dorb.com  Sun Apr  9 01:31:29 2000
From: darrell at dorb.com (Darrell)
Date: Sun, 09 Apr 2000 05:31:29 GMT
Subject: Algorithm: combinations of (k) taken from (n) values
References: <meh9-ACCD3B.16041406042000@news.cit.cornell.edu> <38ED5285.719F26D7@udel.edu> <xmUH4.543$4w1.13051@typhoon.nyroc.rr.com>
Message-ID: <RAUH4.544$4w1.13271@typhoon.nyroc.rr.com>

I wrote:
> Is it possible to figure out a combination at some huge offset without
> figuring out all previous combinations?
Now I see how, but it will take some work.
And I don't know if such a thing would be useful.

> print 'getslice:', choices[40:50]

Yelp that's out of range.

--Darrell






From johann at physics.berkeley.edu  Tue Apr 25 12:22:33 2000
From: johann at physics.berkeley.edu (Johann Hibschman)
Date: 25 Apr 2000 09:22:33 -0700
Subject: Efficient coding advice?
References: <8e2sjn$te4l$1@swen.emba.uvm.edu> <mtvh16qi1m.fsf@astron.berkeley.edu>
Message-ID: <mtr9buqhfa.fsf@astron.berkeley.edu>

Johann Hibschman writes:

Ooops.  Badness.  Left out a line.

> Try:

> ords = map(ord, data)            # take the ord of all the chars

ords = Numeric.array(ords)

> ords.shape = (nframes/2, 2)      # reshape the array in place

...otherwise this will fail.

> left  = ords[:,0]                # pick out individual columns
> right = ords[:,1]

-- 
Johann Hibschman                           johann at physics.berkeley.edu


From dalke at acm.org  Sat Apr  8 14:31:31 2000
From: dalke at acm.org (Andrew Dalke)
Date: Sat, 8 Apr 2000 12:31:31 -0600
Subject: Refeing to substring in list of strings
References: <MPG.1359171bf0a65199989683@nntp.server.uni-frankfurt.de> <00c301bfa145$dff11f80$0201a8c0@spkydomain>
Message-ID: <8cntqq$upi$2@slb0.atl.mindspring.net>

David Fisher
> If you're trying to assign to the substring, like s[1:2] = 'e',
> you can't, strings are immutable.


So the usual solution is to do things like:

  s = s[:lborder] + new_substr + s[:rborder]

You might also want to play around with array.array

s = "This is a string"
import array
a = array.array("c", s)
a[5:7] = array.array("c", "was")
print a.tostring()

                    Andrew
                    dalke at acm.org





From kc5tja at garnet.armored.net  Wed Apr 12 06:59:20 2000
From: kc5tja at garnet.armored.net (Samuel A. Falvo II)
Date: 12 Apr 2000 10:59:20 GMT
Subject: Python 4 DOS
References: <sf7hsua14nk34@corp.supernews.com> <38F48C6D.FD671082@interet.com> <sf99mtkv4nk80@corp.supernews.com> <Pine.SO4.4.02.10004121254400.29583-100000@ub.d.umn.edu>
Message-ID: <slrn8f9h6r.7op.kc5tja@garnet.armored.net>

In article <Pine.SO4.4.02.10004121254400.29583-100000 at ub.d.umn.edu>, Aaron J
Reichow wrote:

>Did you try running it *with* the DOS extender?  I have no experience
>using Python on DOS with the extender, but I've used other prorgams which
>required it (DooM, DJGPP [I think]), and they won't work without it.
>GO32.exe, or something like that?

Remember -- he has specified DOS 6.x -- NO LONG FILENAMES, which Python
requires for some of its supported libraries.

--
KC5TJA/6, DM13, QRP-L #1447
Samuel A. Falvo II
Oceanside, CA


From alberanid at bigfoot.togli.questo.com  Sat Apr  8 11:14:03 2000
From: alberanid at bigfoot.togli.questo.com (Davide Alberani)
Date: Sat, 08 Apr 2000 15:14:03 GMT
Subject: Socket Programming Question
References: <aZdH4.95109$AT6.106753@dfw-read.news.verio.net>
Message-ID: <jggnc8.ov1.ln@snoopy.mio>

rayvd at nospam.firetail.org (Ray Van Dolson) wrote:
>
> However, this code only allows one connection at a time... the
> other can connect to the socket but is simply queued.  I would
> think the main_server_thread() therad would create a new thread
> for each connection?

I'm not an expert, but I think you can use the SocketServer module.

-- 
Homevideo: e` in edicola Class of Nuke'em High: grande       _/Davide\_
capolavoro trash della Troma.                                \Alberani/


From junkster at nospam.rochester.rr.com  Fri Apr 28 21:53:17 2000
From: junkster at nospam.rochester.rr.com (Benjamin Schollnick)
Date: Sat, 29 Apr 2000 01:53:17 GMT
Subject: ODBC / DBI help (via Win32 Extensions)
Message-ID: <rNHvjEdhm5Pp-pn2-RytgtQHbLg9d@d185d18f2.rochester.rr.com>

====================================================
          (Remove "NoSpam" to Email me)
====================================================
Hi All,

	I'm doing some MS Access/SQL work via the ODBC & DBI 
modules... and it's working great...

	I've got the read, and most of the "write" functions working
fine, except when it comes to utilizing the date values...
I can read and convert the date information into a python readable
value.... But I can't seem to do an UPDATE or INSERT INTO command
with a date value inside it.

	Can someone possibly post a example, or a fragment of a 
working  UPDATE / INSERT INTO command?  Or point me to some
python resources that might assist?

	I've checked the ODBC how-to's on the python web site, and
they don't even touch on the subject of WRITING a date... The examples
on reading were very informative, but obviously I'm doing something 
wrong.

		- Benjamin

====================================================
          (Remove "NoSpam" to Email me)
====================================================
Please feel free to copy any and or all of this sig.
A little something for spam bots:

root at localhost postmaster at localhost admin at localhost
abuse at localhost postmaster at 127.0.0.1

Chairman William Kennard: bkennard at fcc.gov 
Commissioner Susan Ness: sness at fcc.gov
Commissioner Harold Furchtgott-Roth: hfurchtg at fcc.gov
Commissioner Michael Powell: mpowell at fcc.gov
Commissioner Gloria Tristani: gtristan at fcc.gov
consumerline at ftc.gov
fccinfo at fcc.gov
ssegal at fcc.gov



From gchiaramonte at my-deja.com  Wed Apr 26 19:06:05 2000
From: gchiaramonte at my-deja.com (gchiaramonte at my-deja.com)
Date: Wed, 26 Apr 2000 23:06:05 GMT
Subject: Python compiler ? (was Re: Python for drivers?)
References: <3900EDBE.282BC85@dsuper.net> <8dr2pn$ec8$1@newshost.accu.uu.nl> <3906f804@nntp.server.uni-frankfurt.de> <015b01bfaf8b$9cd03930$0500a8c0@secret.pythonware.com> <39076316@nntp.server.uni-frankfurt.de>
Message-ID: <8e7sop$5io$1@nnrp1.deja.com>

>
> > if you have $60,000 to spare:
> > http://home.t-online.de/home/aotto/compiler.html
>
> Hmm... $60.000 - this is a funny guy... however,
> do we have 60.000 python users ? If everyone was
> to donate 1$... :))

How about $100 each? Then we only need 600 users. If his compiler
really works, count me in.

Gene


Sent via Deja.com http://www.deja.com/
Before you buy.


From ullrich at math.okstate.edu  Thu Apr  6 11:52:26 2000
From: ullrich at math.okstate.edu (David C. Ullrich)
Date: Thu, 06 Apr 2000 10:52:26 -0500
Subject: Why should I switch to Python? - Infinity of Primes
References: <38EAC8C4.ADC25640@cosc.canterbury.ac.nz> <000d01bf9ec5$3fb90820$3e2d153f@tim> <a9snesk870r1djmo4u9qh9bo86qp9eb2bv@4ax.com>
Message-ID: <38ECB2BA.B6737DAE@math.okstate.edu>

Ben Wolfson wrote:

> On Wed, 5 Apr 2000 02:07:38 -0400, "Tim Peters"
> <tim_one at email.msn.com> wrote:
>
> >not-that-archie-Pu-will-be-fooled-ly y'rs  - tim
>
> Doesn't he grep?

    AP is the guy who invented the phrase "search engine bombing".
That's what you're doing when you mention his name in a sig (in
a post that's not actually about him).
    One of the many useful things you can learn on sci.math.

DU

>
> --
> Barnabas T. Rumjuggler
>
> Who is general Pao, and why are we eating his chicken?
>  -- possibly Legion, in one o' dem wacky-ass Wildcard books



From pinard at iro.umontreal.ca  Tue Apr  4 18:08:30 2000
From: pinard at iro.umontreal.ca (=?ISO-8859-1?Q?Fran=E7ois_Pinard?=)
Date: 04 Apr 2000 18:08:30 -0400
Subject: blessed be Guido, for "5 <= X <= 10" does what it should...
In-Reply-To: kc5tja@garnet.armored.net's message of "4 Apr 2000 13:09:37 GMT"
References: <3TpG4.3349$HG1.98231@nnrp1.uunet.ca> <slrn8eklr4.9d7.kc5tja@garnet.armored.net>
Message-ID: <oqitxxmsdd.fsf@titan.progiciels-bpi.ca>

kc5tja at garnet.armored.net (Samuel A. Falvo II) writes:

> Astonishingly enough, it is.  Or so I've been told.  Not sure where or how
> to verify this, though, as I've never used APL.  It's part of their 32-bit
> character set.

Unicode is more waving around 16 bits, aiming 20 maybe, and having 31 as
ultimate maximum.  They never had 32 bits in their plan, so far that I know.

-- 
Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard





From tseaver at starbase.neosoft.com  Mon Apr 17 22:11:18 2000
From: tseaver at starbase.neosoft.com (Tres Seaver)
Date: 17 Apr 2000 21:11:18 -0500
Subject: ODMG 3.0 and Python
References: <QyGK4.8376$HG1.232986@nnrp1.uunet.ca>
Message-ID: <50D7A6195EC5C31F.B5F86491A9C3822A.B21BE0D5E6CDA8F8@lp.airnews.net>

In article <QyGK4.8376$HG1.232986 at nnrp1.uunet.ca>,
Warren Postma <embed at geocities.com> wrote:
>I was just reading at the ODMG's home page (www.odmg.org) that they have
>released ODMG 3.0, and it represents a not-so-subtle-shift in what they
>consider the ODMG standard to involve. I quote from the disclaimer added
>above the existing online documentation which covers only the "2.0"
>standard.
>
>"Note: This text is for ODMG 2.0.  The new release, ODMG 3.0, was published
>in January
>2000.   The following text does not reflect the ODMG's charter change.  The
>ODMG is now
>an Object Storage API standard that can work with any DBMS or tool.  We are
>in the process of changing all references to Object Databases (ODBMS) to
>Object Data. "
>
>Isn't it neat when a standards body actually ADMITs it is time to
>refactor/rename their
>standard?
>
>Anyways... Is anybody interested in an ODMG-3.0 compliant database for
>Python?
>That would be a LARGE undertaking, but perhaps we could Borgle
>(borrow+burgle) bits of gadfly to help out, such as the excellent kjbuckets,
>parser, and other similar bits. As for the underlying data store, persisting
>Python objects is already trivial, so adding a variety of persistence
>techniques (b+tree of marshalled or pickled objects, etc) would not exactly
>be Difficult.  A certain amount of C++ and Smalltalk Bias would have to be
>overcome, but that's not insurmountable.

You should probably also check out Jim Fulton's talk on the Zope Object
Database, presented at IPC8:

  http://www.python.org/workshops/2000-01/proceedings/papers/fulton/zodb3.html

Sycophantic'ly,

Tres.
-- 
---------------------------------------------------------------
Tres Seaver        tseaver at digicool.com     http://www.zope.org
Digital Creations


From mhammond at skippinet.com.au  Mon Apr 10 08:19:47 2000
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Mon, 10 Apr 2000 12:19:47 GMT
Subject: freeze not working (win)
References: <memH4.4438$HG1.127494@nnrp1.uunet.ca> <Y_tH4.903$4v.4613@news-server.bigpond.net.au> <1g13fs0mkfjv57quaahj6sg6qi5g25arni@4ax.com>
Message-ID: <DFjI4.627$UB.3586@news-server.bigpond.net.au>

"Toby Dickenson" <mbel44 at dial.pipex.net> wrote in message
news:1g13fs0mkfjv57quaahj6sg6qi5g25arni at 4ax.com...
> (posted and cced to Mark)
>
> "Mark Hammond" <mhammond at skippinet.com.au> wrote:
>
> >Note that you must also provide pywintypes1x.dll - this can not be
frozen.
>
> I'm freezing pywintypes15.dll, and havent had any problems _yet_.
> Should I expect bad things?

This is fine as long as you never import a .pyd that referenced
pywintypes.  If you do, then you will get 2 copies of everything - the
frozen one, and the real pywintypes1x.dll that will be loaded via the
.pyd.

Basically, this means if you freeze pywintypes1x.dll, you can never import
a win32* module that is not also frozen into the app.

Mark.





From hdemers at ASTRO.UMontreal.CA  Thu Apr  6 15:50:43 2000
From: hdemers at ASTRO.UMontreal.CA (Hugues Demers)
Date: Thu, 6 Apr 2000 15:50:43 -0400
Subject: Beginning to be disapointed
In-Reply-To: <38ECE311.6B64E4C@tismer.com>
References: <Aw4H4.8078$pR.68745@carnaval.risq.qc.ca> <mt3dozt6sh.fsf@astron.berkeley.edu> <38ECE311.6B64E4C@tismer.com>
Message-ID: <200004061950.PAA30610@polar.ASTRO.UMontreal.CA>

In comp.lang.python, you wrote:
>
>7. How much memory does your machine have?

Plenty (128M), top shows no heavy use of memory. 

>
>It sounds like memory becomes very fragmented, and Python is
>swapping all the time to get some space between scattered
>objects.
>
>ciao - chris
>
>-- 
>Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
>Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
>Kaunstr. 26                  :    *Starship* http://starship.python.net
>14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
>PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
>     where do you want to jump today?   http://www.stackless.com
>



From wware at world.std.com  Mon Apr 24 10:45:22 2000
From: wware at world.std.com (Will Ware)
Date: Mon, 24 Apr 2000 14:45:22 GMT
Subject: The Simple Economics of Open Source
References: <B527E7A5.51AA%raffael@mediaone.net> <3903316C.5D6F1611@tcac.net> <LHEoAJAXWFB5EwF6@jessikat.demon.co.uk>
Message-ID: <FtIznM.EH1@world.std.com>

In article <390451C0.4B5D3196 at tcac.net>, Albert Wagner
<alwagner at tcac.net> writes:
>You STILL don't understand.  Artifice can only copy.  It cannot invent.
>An end product can be copied, even a style can be copied, but not the
>creative process itself.

Robin Becker (robin at jessikat.demon.co.uk) wrote:
> This position is like the weak AI position; you assert there's something
> magical about creation. I think A Turing was right; any human behaviour
> can be simulated and the best simulators (people) are around already.

Then perhaps, like Turing, it would be useful here to design a test to
decide the question. Can somebody do a good enough job spoofing van
Gogh that no art expert can rule out the possibility that the paintings
are previously undiscovered works? This actually isn't that great a test;
because of van Gogh's fame, the a-priori probability that any van-Gogh-
like paining is an undiscovered original rather than a later spoof is
very small. Are there other talented artists with distinctive styles
for whom it wouldn't be very unreasonable for there to be undiscovered
works appearing?

ObQuasiPython: If one wrote a van Gogh spoofer in Python, it could be
called Goghbot, and it would kinda rhyme, but I don't know if it would
be offensively ungrammatical in Dutch without the "van".
-- 
 - - - - - - - - - - - - - - - - - - - - - - - -
Resistance is futile. Capacitance is efficacious.
Will Ware	email:    wware @ world.std.com


From jschmitt at vmlabs.com  Tue Apr 25 20:04:51 2000
From: jschmitt at vmlabs.com (jschmitt at vmlabs.com)
Date: Wed, 26 Apr 2000 00:04:51 GMT
Subject: os.path.walk() question.
References: <8e55m0$pvl$1@hammer.msfc.nasa.gov>
Message-ID: <8e5bqo$c6o$1@nnrp1.deja.com>

I came up with this.  Critique welcome!
##--FindFile.py
import os
import sys
import fnmatch


#-----------------------------------------------------------------------
--------------------------
class FoundFileException:
	def __init__( self, fullname ):
		self.fullname = fullname


#-----------------------------------------------------------------------
--------------------------
def callbackfunc( filename, dir, names ):
	for name in names:
		if fnmatch.fnmatch( name, filename ):
			fullname = os.path.abspath( os.path.join( dir,
name ) )
			raise FoundFileException( fullname )
			# - raising an exception forces the search to
stop as soon as any match
			#   is found


#-----------------------------------------------------------------------
--------------------------
def FindFile( filename, dir=os.getcwd() ):
	foundfile = None
	try:
		os.path.walk( dir, callbackfunc, filename )
	except FoundFileException, filefound:
		foundfile = filefound.fullname
	return foundfile


#-----------------------------------------------------------------------
--------------------------
if __name__ == '__main__':
	# if no directory is passed in, assume the current directory
	if ( len(sys.argv) < 3 ):
		directory = os.getcwd()
	else:
		directory = sys.argv[2]
	fullname = FindFile( sys.argv[1], directory )
	if fullname != None:
		print fullname
	else:
		print sys.argv[1], 'is not in', os.path.abspath(
directory )
#-end of file

John

PS Hopefully deja.com doesn't mess up or wrap my posting.

In article <8e55m0$pvl$1 at hammer.msfc.nasa.gov>,
  "Andrew McDowell" <drew.mcdowell at msfc.nasa.gov> wrote:
> I didn't see this question mentioned in any older posts so forgive me
if I'm
> dragging up and old topic...but:
>
> I need to find out if a file exists under a directory hierarchy.
> The hierarchy could be quite large so I don't want to continue
recursing
> through the hierarchy if the file is found one or two directories
down.   Is
> there a way that you can tell os.path.walk() to stop searching?
>
> I suppose I could customize the code, but my first notion was that
someone
> on the group had probably addressed this before.
>
> Any advice would be much appreciated :)
>
> -Drew
>
>


Sent via Deja.com http://www.deja.com/
Before you buy.


From tjdukes at my-deja.com  Mon Apr  3 19:56:15 2000
From: tjdukes at my-deja.com (tjdukes at my-deja.com)
Date: Mon, 03 Apr 2000 23:56:15 GMT
Subject: error running Demo/extend
Message-ID: <8cbb2o$833$1@nnrp1.deja.com>

I have Python-1.5.2, built on Solaris 2.6 using gcc 2.91.66.
I tried to run the Demo on extending python with c.
I typed
./make_shared
and a bunch of text went by. Then I typed
$ python
Python 1.5.2 (#1, Oct 20 1999, 13:48:38) [GCC egcs-2.91.66 19990314
(egcs-1.1.2 on sunos5
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> import xx
Traceback (innermost last):
File "<stdin>", line 1, in ?
ImportError: ld.so.1: python: fatal: relocation error: file
./xxmodule.so: symbol _PyObject_New: referenced symbol not found
>>>
What is causing the unfound referenced symbol?
Here is the line from my config.status showing how
python was configured.
# ./configure --prefix=/nfs/ibmoto/tools/tlv6/gnome/gnome --with-thread
--with-threads
Any help is appreciated.
thanks,
Todd.


Sent via Deja.com http://www.deja.com/
Before you buy.


From ullrich at math.okstate.edu  Wed Apr  5 10:44:15 2000
From: ullrich at math.okstate.edu (David C. Ullrich)
Date: 5 Apr 2000 14:44:15 GMT
Subject: Why should I switch to Python? - Infinity of Primes
References: <NDBBKEGCNONMNKGDINPFGEJDDFAA.infonuovo@email.com> <8ccoug$f4o$1@pegasus.csx.cam.ac.uk> <7F51E902E09F3F60.8F612607812ED16B.82CB8B41435E14FD@lp.airnews.net> <38EA369C.26A1C25@math.okstate.edu> <38EAC8C4.ADC25640@cosc.canterbury.ac.nz>
Message-ID: <01bf9f0e$df164380$09d68ad1@daves-dell>


Greg Ewing <greg at cosc.canterbury.ac.nz> wrote in article
<38EAC8C4.ADC25640 at cosc.canterbury.ac.nz>...
> "David C. Ullrich" wrote:
> > 
> >     There's nothing non-constructive about the traditional
> > proof of the infinitude of the sequence of primes - given
> > a sequence of primes it _constructs_ a prime not on the
> > list.
> 
> Um, no it doesn't - it constructs a number which is
> *either* prime *or* divisible by some other prime bigger
> than the one you started with.

	For heaven's sake. Here's a constructive way to find
whether a number is prime: To check whether n is prime
test whether it's divisible by j, for 2 <= j < n.
	Here's a constructive way to find a prime factor
of n: For 2<= k <= n test whether k is prime and then test
whether n is divisible by k.

	I mean really, I didn't mean that the way the proof
is usually _presented_ actually gives that prime. But the
proof contains all the cleverness needed - once we have
that number which is either a larger prime or divisible
by a larger prime then finding the larger prime itself is
trivial. In a perfectly constructive way.
	(Um, actually you should note that the word
"larger" is irrelevant - you inserted the word "bigger".
Given any finite set of primes the construction gives
a prime not in the set.

> If someone actually came up with a formula for constructing
> primes, it would be rather large news -- isn't that one of
> the Big Unsolved Problems?

	I guess you're serious. Finding a "formula" for the
n-th prime is not so easy, but finding an algorithm to
calculate the n-th prime is trivial. (Finding an
_efficient_ algorithm is a different question.)

> -- 
> Greg Ewing, Computer Science Dept,
> +--------------------------------------+
> University of Canterbury,	   | A citizen of NewZealandCorp, a	  |
> Christchurch, New Zealand	   | wholly-owned subsidiary of USA Inc.  |
> greg at cosc.canterbury.ac.nz	   +--------------------------------------+
> 


From jj66ca at netscape.net  Wed Apr 19 21:48:47 2000
From: jj66ca at netscape.net (jj)
Date: Wed, 19 Apr 2000 18:48:47 -0700
Subject: Port Python for Vxworks with MPC860
References: <sH0K4.13$Ny3.1761@newreader.ukcore.bt.net>
Message-ID: <38FE61FF.DEB7E9CF@netscape.net>


Hi, every one

I tried to port Python to Vxworks for MPC860T and the following error message
occurred when I tried to link python.o into the vxWorks image.

vxWorks.tmp: In function `my_fgets':
vxWorks.tmp(.text+0xfe04): undefined reference to `PyOS_InterruptOccurred'
vxWorks.tmp(.text+0xfe18): undefined reference to `PyOS_InterruptOccurred'
vxWorks.tmp: In function `Py_Finalize':
vxWorks.tmp(.text+0x463b0): undefined reference to `PyOS_FiniInterrupts'
vxWorks.tmp: In function `initsigs':
vxWorks.tmp(.text+0x47fec): undefined reference to `PyOS_InitInterrupts'
vxWorks.tmp: In function `PyErr_CheckSignals':
vxWorks.tmp(.text+0x480fc): undefined reference to `PyOS_InterruptOccurred'
make: *** [vxWorks] Error 0x1

Has anyone been successful in terms of porting Python for MPC860? Please share
the experience, any help is greatly appreciated.

Thank you.

Lei



From alwagner at tcac.net  Fri Apr 21 12:56:03 2000
From: alwagner at tcac.net (Albert Wagner)
Date: Fri, 21 Apr 2000 11:56:03 -0500
Subject: The Simple Economics of Open Source
References: <14591.15630.192472.440375@goon.cnri.reston.va.us>
Message-ID: <39008823.A3B67E8@tcac.net>

I think that the most interesting part of the article was how "startled"
economists were to find a different motivation than greed functioning. 
In Douglas Hoffstadter's book "Metamagical Themas", in the chapter "The
Prisoner's Dilemma and the Evolution of Cooperation", he describes what
I believe is happening with Open Source.  Especially, interesting is the
computer tournament wherein the TIT-FOR-TAT strategy consistently wins. 
Also:

"The Evolution of Cooperation" by Robert Axelrod: "Mutual cooperation
can emerge in a world of egoists without central authority, by starting
with a cluster of individuals who rely on reciprocity."

"The Selfish Gene" by Richard Dawkins

"Small is Beautiful" by E.F.Schumacher

An elegant algorithm can be as beautiful as a painting or a song.  The
truely creative desperately need to share their creations.  The
craftsman wants to sign his work.

-- 
Small is Beautiful


From rcc at nospamthanks_jennic.com  Thu Apr 20 11:55:27 2000
From: rcc at nospamthanks_jennic.com (Robert Cragie)
Date: Thu, 20 Apr 2000 15:55:27 GMT
Subject: Using popen in windows
References: <newscache$vzabtf$ia6$1@jenpc07.jennic.co.uk>
Message-ID: <newscache$6tnbtf$h54$1@jenpc07.jennic.co.uk>

I have had an e-mail suggesting I install win32all-128.exe. I did this, and
everything is now OK, so thanks for this advice.

Robert

Robert Cragie <rcc at nospamthanks_jennic.com> wrote in message
news:newscache$vzabtf$ia6$1 at jenpc07.jennic.co.uk...
| I tried the following in Windows 98 as an experiment:
|
| from win32pipe import popen
| p = popen('dir','r')
| str = p.read()
| print str
|
| It just hangs when trying to do the read, and I have to kill the shell
off.
| I have tried IDLE 0.5 and PythonWin and the 'DOS' command line, both do
the
| same. I am using Python 1.5.2, and the win32pipe.pyd DLL is dated 3rd May
| 1999.
|
| I have done the equivalent on Linux and it works fine:
|
| from os import popen
| p = popen('ls','r')
| str = p.read()
| print str
|
| Anyone got any ideas?
|
| TIA
|
| Robert Cragie
|
|




From thomas.heller at ion-tof.com  Thu Apr 20 05:54:25 2000
From: thomas.heller at ion-tof.com (Thomas Heller)
Date: Thu, 20 Apr 2000 11:54:25 +0200
Subject: win32all-128.exe availability?
References: <38fd01b2$0$10982@senator-bedfellow.mit.edu> <8dmbm3$asu$1@gxsn.com>
Message-ID: <001e01bfaaae$69c67b30$4500a8c0@thomasnb>

I've uploaded my saved copy of win32all-128.exe
to the following URL:

http://lightning.prohosting.com/~theller/win32all-128.exe

Thomas Heller





From pecora at anvil.nrl.navy.mil  Sun Apr 30 10:33:27 2000
From: pecora at anvil.nrl.navy.mil (Louis M. Pecora)
Date: Sun, 30 Apr 2000 14:33:27 GMT
Subject: Reading in strings -> numbers ??
Message-ID: <300420001033289484%pecora@anvil.nrl.navy.mil>

Bruce wrote:
>
The syntax is not equivalent in the same sense as printf vs. scanf, but
the
re module will you to parse out substrings from whatever text file
format
you are trying to read, and then eval() can be used on the components
of the
result, to turn the "fields" that are numeric strings into Python
numbers.


Bruce,

Checked out the re.py module and for me it's not transparent (I've been
at Python for on ly 3 weeks, sorry).  Would you or anyone have any
examples of doing what you suggest to read in numbers?  Thanks.


From bruce_dodson.nospam at bigfoot.com  Fri Apr 28 21:15:21 2000
From: bruce_dodson.nospam at bigfoot.com (Bruce Dodson)
Date: Fri, 28 Apr 2000 22:15:21 -0300
Subject: Reading in strings -> numbers ??
References: <280420001700308281%pecora@anvil.nrl.navy.mil>
Message-ID: <GIqO4.3939$d73.50248@newscontent-01.sprint.ca>

The syntax is not equivalent in the same sense as printf vs. scanf, but the
re module will you to parse out substrings from whatever text file format
you are trying to read, and then eval() can be used on the components of the
result, to turn the "fields" that are numeric strings into Python numbers.

Bruce


Louis M. Pecora wrote in message
<280420001700308281%pecora at anvil.nrl.navy.mil>...
>Is there a Python "read-partner" for the write(string) function, e.g.,
>
>fileobject.write(" %e %e %d" % (1.2, -3.5, 10))




From pointal at lure.u-psud.fr  Fri Apr 28 08:25:28 2000
From: pointal at lure.u-psud.fr (Laurent POINTAL)
Date: Fri, 28 Apr 2000 12:25:28 GMT
Subject: Do I always have to write "self." ?
References: <270420001706386648%pecora@anvil.nrl.navy.mil> <8eacdf$nnm$1@newshost.accu.uu.nl> <280420000751353947%pecora@anvil.nrl.navy.mil>
Message-ID: <3909817b.2156432776@news.u-psud.fr>

On Fri, 28 Apr 2000 11:51:34 GMT, "Louis M. Pecora"
<pecora at anvil.nrl.navy.mil> wrote:
 
>: Martijn,
>: 
>: Thanks for the very nice tutorial.  Your example brings up some typical
>: stumbling blocks for us beginners.  Namely, references to mutable and
>: immutable objects.
>: 
>: In the above I would describe things as follows (am I right or wrong?):
>: 
>: (1) data and self.data point to the same object, a list, so when a
>: method is called it is the method that works on that instance of a list
>: and either data or self.data will append 'foo' to the same list object.
>: 
>: (2) when you do more=more+1 an new object (the more+1 value) is created
>: and more then points to it and no longer to the old value of self.more. 
>: The latter still exists and points to the original 10 object.
>: 
>: (3) Similar story here for interesting where a new object
>: interesting[10:] is created and interesting is "switched" to point to
>: it.  self.interesting still points tohe "foo" object.
>: 
>: I hope I got that right, because they have tripped me up a few times.
>: 
>: Thanks, again.

*** Note: taking references copies of the self members is the same
problem as importing * from a module: its only reference copying... 

*** Note2: there may be a hack (not tested), with copying
automatically members into locals:
	def copyself(self,loc) :
		for m in self.__dict__.keys():
			loc[m] = eval("self."+m)
	def mymethod(self) :
		self.copyself(locals())
		# Then you can use directly all members.

But this is still only a reference copying... if you wants to change
of referenced object in a member, you must use self.xxxxxx. 

A+

Laurent.

---
Laurent POINTAL - CNRS/LURE - Service Informatique Experiences
Tel/fax: 01 64 46 82 80 / 01 64 46 41 48
email  : pointal at lure.u-psud.fr  ou  lpointal at planete.net 


From olipt at mayo.edu  Wed Apr 19 22:14:09 2000
From: olipt at mayo.edu (Travis Oliphant)
Date: Wed, 19 Apr 2000 21:14:09 -0500
Subject: Python tools for PDE and Linear algebra
In-Reply-To: <ya0og79vsem.fsf@sessrymnir.ifi.uio.no>
References: <ya0vh1kzsaf.fsf@grindaflet.ifi.uio.no> <38F75987.26AD3FEA@oreilly.com> <ya0og79vsem.fsf@sessrymnir.ifi.uio.no>
Message-ID: <Pine.LNX.4.10.10004192110450.13408-100000@us2.mayo.edu>

I don't know of wrapped libraries.  However, I have an alpha Sparse matrix
implementation that would be useful for interfacing with such routines.

With this Sparse Matrix implementation,  finite difference tools are
simple to implement (and can be done in Python for the most part).

Finite element routines are a different story. I've used it to make an
Assembly routine for assembling the matrix for an Inhomogeneous Helmholtz
problem.  I would like to continue work in this area as soon as I graduate
(which is taking my attention at the moment). If anybody would like to
discuss this with me, I am at least somewhat aware of what's out there
and what needs to be done.

Thanks

Travis




From cgw at alum.mit.edu  Tue Apr  4 19:56:09 2000
From: cgw at alum.mit.edu (Charles G. Waldman)
Date: Tue, 04 Apr 2000 23:56:09 GMT
Subject: Why should I switch to Python? - Infinity of Primes
References: <NDBBKEGCNONMNKGDINPFGEJDDFAA.infonuovo@email.com>
Message-ID: <tivG4.1408$g35.73564@newsread1.prod.itd.earthlink.net>

In article <NDBBKEGCNONMNKGDINPFGEJDDFAA.infonuovo at email.com>, "Dennis E.
Hamilton" <infonuovo at email.com> wrote:

> 
> It goes something like this (matching the one that was already posted
> for you):
> 
> 1.	Assume P is the largest prime.
> 
> 2.	Let Q be the product of all of the primes up to P: 2*3*5* ... *P
> 
> 3.	The number Q+1 is not divisible by any of the primes that divide Q
> (they
> each leave a remainder of 1).  Therefore Q+1 is a prime or there is some
> other prime, larger than P (i.e., not among 2, 3, 5, ..., P), that
> divides Q+1.  So P is not the largest prime.  QED.
> 
> This all stands on the fundamental theorem of arithmetic: that every
> integer number is a product of primes in exactly one way.  
> That is Exercise 1.2.4-21 in "The Art of Computer Programming, v.1

I'm pretty sure that this last statement is not correct. 
You don't need unique factorization here.  Euclid's proof does not require
a hypothesis of unique factorization into primes, just a well-ordering of 
primes, and existence of division-with-remainder.

You just need the fact that when you divide A into (A*B)+1, you get a 
remainder of 1.  That's enough to show that none of the primes
2..P divide Q=1+(2*3*5*7..*P).

Since none of these primes divide Q, either Q is prime, or some other prime not 
in the list 2..P divides Q, either of which options contradicts the hypothesis that
Q is the largest prime.  Where do you need to make use of the fact that prime
factorizations are unique?   Euclid's proof is self-contained and referencing 
Knuth for a proof of unique factorization just muddies the waters.

Interestingly, there are some quadratic number fields (discrete subsets of the 
complex numbers) which don't have unique factorization, but still have infinitely
many prime elements.



From tzman at bhmi.com  Wed Apr  5 09:52:13 2000
From: tzman at bhmi.com (T Zimmerman)
Date: Wed, 5 Apr 2000 08:52:13 -0500
Subject: Tkinter Meta Keys
Message-ID: <38eb450e$0$6938@wodc7nh6.news.uu.net>

Can anyone get the following script to work on MS NT. I have been unable to
get the use of "alt-*" keys on this platform. Any suggestions would be
greatly appreciated.

Thanks,
tzman at bhmi.com

*************************************************************
import Tkinter

root = Tkinter.Tk()

button = Tkinter.Menubutton(text = 'File', underline = 0)
button.pack()
menu = Tkinter.Menu(button)
button.configure(menu = menu)
menu.add_command(label = 'Hello World', underline = 0)

root.mainloop()
**********************************************************************




From mhammond at skippinet.com.au  Fri Apr 14 22:26:28 2000
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Sat, 15 Apr 2000 02:26:28 GMT
Subject: Comparing PythonWin and IDLE?
References: <NDBBLJJLLFOJMLCMJPLOGEEADJAA.charlie@intelligenesis.net> <38F774E5.E34B0365@gmx.de>
Message-ID: <orQJ4.1385$1D.12897@news-server.bigpond.net.au>

"Michael Scharf" <Michael.Scharf at gmx.de> wrote in message
news:38F774E5.E34B0365 at gmx.de...

> I know it is possible to change the code of existing functions.
> So, with some ticks it should be possible to replace the code
> of existing functions. The same must be done for all mutable objects
> (list, dicts etc.). Only not mutable direct imported variable (int,
> string, tuple etc.) cannot be changed that way.
>
> I just wonder if anybody has written such a clever reimport...

_Years_ ago, Scott Hassan posted one to the newsgroup.  Deja should find
it.  This code predated packages, but still an excellent starting point.
I very nearly adapted it for use in Pythonwin, but never got motivated
enough...

Mark.





From javanet at dynacap.com  Wed Apr 12 12:42:40 2000
From: javanet at dynacap.com (Arnaldo Riquelme)
Date: Wed, 12 Apr 2000 08:42:40 -0800
Subject: Python 4 DOS
References: <sf7hsua14nk34@corp.supernews.com> <38F48C6D.FD671082@interet.com>
Message-ID: <sf99mtkv4nk80@corp.supernews.com>

James C. Ahlstrom <jim at interet.com> wrote in message
news:38F48C6D.FD671082 at interet.com...
> Arnaldo Riquelme wrote:
> >
> > Can anyone tell me if I can get python binaries for DOS (16-bit).
> > I heard of  Hans Nowak 's  Python-DX but the links to his website is
broken?
> > Where are you Hans?
>
> www.python.org has a DOS Python running with a DOS extender.
>
> Jim

Thanks Jim

I want to run Python in just plain dos (6.x), I don't think the version in
http://www.python.org/ftp/python/wpy/dos.html, can run
in just DOS. I tried  and it doesn't work.


Thanks





From effbot at telia.com  Sun Apr  2 11:35:01 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Sun, 02 Apr 2000 15:35:01 GMT
Subject: Is www.python.org down?
References: <eviltofu-C63EA3.11234102042000@news.earthlink.net>
Message-ID: <FMJF4.959$n68.188429824@newsb.telia.net>

Jerome Chan <eviltofu at rocketmail.com> wrote:
> Have not been able to connect to the web server although the machine
> responds to pings. Is it just me?

like everything else, it's being moved to belgium :-)

while you're waiting for the real thing to get back on
the net, here's a list of mirror sites:

http://www.cwi.nl/www.python.org/Mirrors.html

</F>




From moshez at math.huji.ac.il  Fri Apr 28 22:03:44 2000
From: moshez at math.huji.ac.il (Moshe Zadka)
Date: Sat, 29 Apr 2000 05:03:44 +0300 (IDT)
Subject: Metaclasses?
In-Reply-To: <wk66t2gy2n.fsf@turangalila.harmonixmusic.com>
Message-ID: <Pine.GSO.4.10.10004290503010.22564-100000@sundial>

On 28 Apr 2000, Dan Schmidt wrote:

> I just do this:
> 
> def method(_, arg1, arg2):
> 	_.spam = "spam"
> 	_.eggs = arg1 + arg2
> 
> Will this practice get me kicked out of any Python Users Groups?

I doubt it, but selfnanny might give you a sound beating <wink>

nannies-can-be-nice-too-ly y'rs, Z.
--
Moshe Zadka <mzadka at geocities.com>. 
http://www.oreilly.com/news/prescod_0300.html
http://www.linux.org.il -- we put the penguin in .com




From boud at rempt.xs4all.nl  Sun Apr 16 07:50:12 2000
From: boud at rempt.xs4all.nl (Boudewijn Rempt)
Date: 16 Apr 2000 11:50:12 GMT
Subject: Event triggering and weak references
References: <8dbif3$hm0$1@metro.ucc.usyd.edu.au>
Message-ID: <8dc9dk$1fp$1@news1.xs4all.nl>

Ricardo Nogueira <rnog2438 at mail.usyd.edu.au> wrote:

> I am Smalltalk addicted : )  and this means that I can't do with any other
> language, ex. C, C++, Java,  : (
> This was true until I found Python 2 weeks ago at Sydney Uni. I loved its
> modularity, pluggin behaviour and freedom.
> But I am missing Smalltalk event handling... Smalltalk (like Python) support
> the call-Back mechanism, but there is a more powerful inter-object interest
> registering that I haven't found in Python (so far).

If I understand you correctly, then you need to take a look at PyQt. It
offers a signals and slot concept, whereby you have a widget emitting
signals, which you can connect to slots in other widgets.

To translate your example into actual code:


class QCar(Qobject):

  def __init__(self, *args):
	  QObject.__init__,(self, ) + args)
		self.speed=120

  def SpeedChanged(self):
	  # whenever we call self.speedChanged, all other objects
		# that have registered with the signal "speedChanged"
		# will get notified.
	  self.emit(PYSIGNAL("speedChanged", (self.speed,))
		
class QCarSpeedoMeter(QObject):

  def setSpeed(speed):
	  ...
	
class QCarTooFastFlashLight(QObject):

  def checkSpeed(speed):
	  if speed > 100:
		  ...
			
class QCarApp(QObject):
  
	def __init__(self, *args)
	  QObject.__init__,(self, ) + args)
		#
		# Create the various objects
		#
		self.car=QCar()
		self.speedometer=QCarSpeedoMeter()
		self.flashlight=QCarTooFastFlashLight()
	  #
		# Connect the signals to the slots.
		#
		self.connect( self.car
		            , PYSIGNAL("speedChanged(int)")
								, self.speedometer.setSpeed
								)
    self.connect( self.car
		            , PYSIGNAL("speedChanged(int)")
								, self.flashlight.checkSpeed
								)

Using this concept is even easier and more natural in Python than
in C++ - in C++ apps you need a precompiler to make it work, in Python
it just works.

You can find the bindings at: 
   http://www.river-bank.demon.co.uk/software/ 
	 
Qt is at: http://www.troll.no

A tutorial and example programs at my website:
   http://www.valdyas.org/python/index.html



-- 

Boudewijn Rempt  | http://www.valdyas.org


From ehagemann at home.com  Sat Apr  1 21:39:54 2000
From: ehagemann at home.com (Eric Hagemann)
Date: Sun, 02 Apr 2000 02:39:54 GMT
Subject: New Features in Python 1.6
References: <200004011740.MAA04675@eric.cnri.reston.va.us>
Message-ID: <_pyF4.25562$E85.640966@news1.rdc1.md.home.com>

What are the options for users wanting 1.6 that may not have access to
modules.python.org
on a regular basis or at all.  I'd hate to be disconnected from the net and
find I need that
'one' function that used to be in the standard library ......... e.g. can I
down load the
entire library and forgo the net check ?  What happens if there is no net --
will the import function
get unhappy ?

Also when you import like this does the code check the net for a newer
version or only
when you have no version of the particular sub-module ?






"Guido van Rossum" <guido at python.org> wrote in message
news:200004011740.MAA04675 at eric.cnri.reston.va.us...
(uid)
>
> 2. Imputils
>
> Complementary to the Distutils are the Imputils, or Import Utilities.
> Python's import mechanism has been reworked to make it easy for Python
> programmers to put "hooks" into the code that finds and loads modules.
> The default import mechanism now includes hooks, written in Python, to
> load modules via HTTP from a known URL.
>
> This has allowed us to drop most of the standard library from the
> distribution.  Now, for example, when you import a less-commonly-needed
> module from the standard library, Python fetches the code for you.  For
> example, if you say
>
>     import tokenize
>
> then Python -- via the Imputils -- will fetch
> http://modules.python.org/lib/tokenize.py for you and install it on your
> system for future use.  (This is why the Python interpreter is now
> installed as a setuid binary under Unix -- if you turn off this bit, you
> will be unable to load modules from the standard library!)
>
> If you try to import a module that's not part of the standard library,
> then the Imputils will find out -- again from modules.python.org --
> where it can find this module.  It then downloads the entire relevant
> module distribution, and uses the Distutils to build and install it on
> your system.  It then loads the module you requested.  Simplicity
> itself!
>
>




From andrer at stud.cs.uit.no  Sat Apr  1 17:13:46 2000
From: andrer at stud.cs.uit.no (=?iso-8859-1?Q?Andr=E9?= Larsen Risnes)
Date: Sun, 02 Apr 2000 00:13:46 +0200
Subject: Paid to work in Python? (was: Re: Choice Of Language)
References: <Uqvw4.2261$yL2.28868@typhoon.southeast.rr.com> <8ajmib$elo$1@nnrp1.deja.com> <8c3ats$b5f$1@nntp.Stanford.EDU> <slrn8eact2.sq7.nobody@bambi.visi.com>
Message-ID: <38E6749A.97D33BF8@stud.cs.uit.no>

Grant Edwards wrote:
> 
> Pointless tanget: Is it just me, or is Google really the
> last usable search engine?
> 

Google's great. You could also try http://www.alltheweb.com


-- 
Andr? Larsen Risnes


From godzilla at netmeg.net  Mon Apr 24 16:46:56 2000
From: godzilla at netmeg.net (Les Schaffer)
Date: Mon, 24 Apr 2000 20:46:56 GMT
Subject: looking for library function browser to run inside (x)emacs
Message-ID: <m24s8rmd9e.fsf@netmeg.net>

I have been using the OOBrowser inside XEmacs for a couple years and
its useful for browsing class hierarchies.

but i find it does not allow for browsing function hierarchies, such
as is found in the [...]/python/site-packages and subdirs.

does anyone know of a package which runs inside emacs allowing to
browse function hierarchies?

by hierarchies, i mean something like all the standalone functions
which reside in modules like so (below site-packages):

scientific/ArrayIO.py           
scientific/Color.py             
scientific/Derivatives.py       
....
Numerical/ArrayPrinter.py   
Numerical/FFT.py           
Numerical/LinearAlgebra.py 
etc...

for example:

>>>
(gustav)/usr/lib/python1.5/site-packages/: cd scientific
(gustav)/usr/lib/python1.5/site-packages/scientific/: grep "^def " ArrayIO.py
def readArray(filename):
def writeArray(a, filename):
def writeDataSets(datasets, filename, separator = ''):
<<<

the problem with OO browser is it seems to only like to catalogue
functions which exit in the so-called system tree -- that is -- the
files one is creating locally for a project. it doesnt catalog modules
(like those listed above) containing class-less functions which exist
in the library hierarchy inside scientific, Numerical, etc.

many thanks for any pointers...

i'm tempted to wrote something myself which grabs __doc__ strings
from the lib hierarchy and arranges/formats them inside an xemacs
buffer. (assuming all functions are suitably doc'ed ;-) but the more
interesting part would be to make the buffer clickable, so if i see a
function i like, i can read it into xemacs and see if it does what i
need and learn how to use it...

i wrote a little script to grep for 'def XXX' strings hard left
in each py file in the site-packages hierarchy:

(gustav)/usr/lib/python1.5/site-packages/: findpydef | wc -l
   3550

thatsalottafunctions!!!

interestingly, not a lot of them are marked module private 

e.g.

def _ummagumma():
   """Careful with that ax, eugene"""

though i am betting a lot __could be__.....

les schaffer

======= how many def's you got??? ====

#!/bin/sh

cd /usr/lib/python1.5/site-packages

PYFILES=`find . -name "*.py"`

for file in $PYFILES; do

  echo "Checking $file ..."
  grep "^def " $file
done  


From billtut at microsoft.com  Thu Apr 20 19:44:02 2000
From: billtut at microsoft.com (Bill Tutt)
Date: Thu, 20 Apr 2000 16:44:02 -0700
Subject: Why MFC instead of Tkinter -- was Re: Pythonwin?
Message-ID: <4D0A23B3F74DD111ACCD00805F31D8101D8BCF91@RED-MSG-50>


> From: Don Tuttle [mailto:tuttledon at hotmail.com]

> Which brings up the question near and dear to a lazy man's heart like
> myself..  Which is the eaiser Windows GUI builder for 
> Python--VB or VC++ ?
> 

Dunno, never written an MFC UI. :)

> With VB,  I've only seen Python mentioned as a COM object to 
> be used by VB.
> Are there other ways?
> 

Yes but its more convoluted.
e.g.: Write a VB ActiceX UI Control, and then host it in some Python code
that uses MFC. :) 

What you mentioned is certainly the most straight forward approach.


Bill



From tim.lavoie at mts.net  Thu Apr 27 12:04:31 2000
From: tim.lavoie at mts.net (Tim Lavoie)
Date: Thu, 27 Apr 2000 16:04:31 GMT
Subject: Best database to use with Python
References: <1255280638-66426487@hypernet.com>
Message-ID: <slrn8ggos5.4un.tim.lavoie@goibniu.beyondtv.net>

In article <1255280638-66426487 at hypernet.com>, Gordon McMillan wrote:

>Sometime ago linuxplanet did a comparison (sorry I can't 
>check the link - my ISP is hosed right now):
>http://www.linuxplanet.com/linuxplanet/tutorials/1224/1/
>
>As I recall, Postgres was the most complete DB; MySQL the 
>least complete. MySQL was very fast at some operations 
>(probably related to lack of transaction support), but the most 
>startlingly fast one was bulk load, which does you no good in 
>a CGI situation. Again, from memory, MySQL was fast at very 
>simple queries, but had a very limited SQL dialect and fell 
>apart with fancier queries.

Probably the simpler answer here is for the original poster to have a quick
go at both, he'll form his own opinions soon enough.  :-)

I did a work project involving a search engine of sorts, with a crawler
creeping through the data at night, and a CGI front-end to look for one or
more words. The first iteration used Postgres, but was slow.... The crawler
took forever on inserts, and since the data changed often, there were a lot
of them. Queries for single words were acceptable, but multi-word 'OR'
searches were painfully slow. In this case, MySQL fit the bill very well;
data was added very quickly, and queries were extremely fast. Also, we
didn't need transactions or anything of the sort, since at the very worst
we'd crawl the whole works again.

P.S. This .sig was a coincidence, but I thought it fits anyway...

-- 
Give a small boy a hammer and he will find that everything he encounters
needs pounding.


From loewis at informatik.hu-berlin.de  Sat Apr  8 14:56:56 2000
From: loewis at informatik.hu-berlin.de (Martin von Loewis)
Date: 08 Apr 2000 20:56:56 +0200
Subject: New Features in Python 1.6
References: <200004011740.MAA04675@eric.cnri.reston.va.us> <_pyF4.25562$E85.640966@news1.rdc1.md.home.com> <slrn8edfmk.185.wtanksle@dolphin.openprojects.net> <ltvh21ywfs.fsf@asfast.com> <oq66txol0w.fsf@titan.progiciels-bpi.ca> <GOpG4.5050$74.81782@newsc.telia.net> <38ea40a1.47458347@news.nikoma.de> <vWrG4.5095$74.82533@newsc.telia.net>
Message-ID: <p6qzor4ju9z.fsf@informatik.hu-berlin.de>

"Fredrik Lundh" <effbot at telia.com> writes:

> okay, okay, I shouldn't have posted that several days after
> april 1st...  but hey, this group is (indirectly) named after a
> infamous comedy troupe, so what did you expect?

That's hard to say ... since nobody expects the Spanish Inquisition.

Regards,
Martin


From dwhite2 at seas.upenn.edu  Fri Apr  7 18:04:17 2000
From: dwhite2 at seas.upenn.edu (dave white)
Date: Fri, 07 Apr 2000 18:04:17 -0400
Subject: MySQL-module
References: <55jH4.145$rD3.2312@news1.online.no>
Message-ID: <38EE5B37.5D0C2D9C@seas.upenn.edu>

I use MySQLdb.  It works very nicely and utilizes the python db-api.  It
uses some c that has to be compiled, but you don't have to recompile
python itself.

david

Thomas Weholt wrote:
> 
> Hi,
> 
> I`m going to try MySQL and want to access stuff using Python. At Python.org
> I found to links to two different modules, one written in python, the other
> C. Is there a special reason for choosing one over the other?
> 
> I`d like to use all python, not compile anything cuz I`m scared python won`t
> work anymore. Had this problem earlier with re-compilation of python. zlib
> etc. didn`t work anymore.
> 
> Any tips? I can create the actual databases/tables manually, but need to
> acces/insert/manipulate the data from a python-script.
> 
> Thomas


From Michael.Scharf at gmx.de  Thu Apr 13 19:15:13 2000
From: Michael.Scharf at gmx.de (Michael Scharf)
Date: Fri, 14 Apr 2000 01:15:13 +0200
Subject: [ANNOUNCE] Garbage collection for Python
References: <20000407221750.A29279@acs.ucalgary.ca> <m3r9cagx3c.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <38F65501.665243F6@gmx.de>

Michael Hudson wrote:
> 
> "Neil Schemenauer" <nascheme at enme.ucalgary.ca> writes:
> 
> > Garbage collection for Python:
> >
> >     http://www.enme.ucalgary.ca/~nascheme/python/gc.html
> > The patch is against the current CVS version of Python.  If
> > enough people ask for a patch against the stable version of
> > Python I will make one.  Please send comments and bug reports.
> 
> To someone who might know: what hope is there for this to go into 1.6?
> 
> It would be, in my opinion, be a really really really good thing; for
> one thing it would help in language wars...

I would *love* to see this in 1.6!!!!


Michael
-- 
     ''''\     Michael Scharf
    ` c-@@     TakeFive Software GmbH, a Wind River Company
    `    >     http://www.WindRiver.com
     \_ V      mailto:Michael.Scharf at gmx.de


From ns56645 at swt.edu  Sun Apr 23 00:48:22 2000
From: ns56645 at swt.edu (ns56645)
Date: Sat, 22 Apr 2000 23:48:22 -0500
Subject: Radio-button in Tkinter
Message-ID: <39028096.24B972B4@swt.edu>

In the following code the radiobuttons that have been set are not seen
as default selected in the output.
For e.g         v1.set(3) # initialize should set the radiobutton with
text 'BOTH' as default.


What could be the wrong with the code?
Please help?

Shah


#######        Frame 3  Radio Buttons
##########################################################

        # Create the "Location" contents of the page.
        group = Pmw.Group(frame3, tag_text = 'Location')
        group.pack(fill = 'both', expand = 1, padx = 15, pady = 10)

        v1 = IntVar()
        v1.set(3) # initialize
        rb1 = Radiobutton(group.interior(), text = 'SWT', variable=v1,
value=1)
        rb1.grid(row = 0, column = 0)
        rb2 = Radiobutton(group.interior(), text = 'WHS', variable=v1,
value=2)
        rb2.grid(row = 0, column = 1)
        rb3 = Radiobutton(group.interior(), text = 'BOTH', variable=v1,
value=3)
        rb3.grid(row = 0, column = 2)


        # Create the "Level" contents of the page.
        group = Pmw.Group(frame3, tag_text = 'Level')
        group.pack(fill = 'both', expand = 1, padx = 15, pady = 10)

        v2 = IntVar()
        v2.set("13") # initialize
        rb11 = Radiobutton(group.interior(), text = 'Grad', variable=v2,
value=11)
        rb11.grid(row = 0, column = 0)
        rb12 = Radiobutton(group.interior(), text = 'UnderGrad',
variable=v2, value=12)
        rb12.grid(row = 0, column = 1)
        rb13 = Radiobutton(group.interior(), text = 'BOTH', variable=v2,
value=13)
        rb13.grid(row = 0, column = 2)


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20000422/ec19e573/attachment.html>

From pecora at anvil.nrl.navy.mil  Wed Apr  5 09:14:58 2000
From: pecora at anvil.nrl.navy.mil (Louis M. Pecora)
Date: Wed, 05 Apr 2000 13:14:58 GMT
Subject: Why should I switch to Python?
References: <000701bf9c5a$b806b420$752d153f@tim> <oqity066km.fsf@titan.progiciels-bpi.ca> <slrn8ej2gc.b3t.warlock@raven.myth> <m3itxyz744.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <050420000914595917%pecora@anvil.nrl.navy.mil>

In article <m3itxyz744.fsf at atrus.jesus.cam.ac.uk>, Michael Hudson
<mwh21 at cam.ac.uk> wrote:

> warlock at eskimo.com (Jim Richardson) writes:
> 
> > Out of curiousity, how does one (mathematically) prove that?
> > (if this turns out to be one of those 12' blackboards full of greek
> > letters and funny squiggles, Ok, I cry uncle...)
> 
> It's not all that hard.  Write the prime numbers as p1,p2,p3,etc.
> Suppose that this series stops at pn.  Then let P=p1*p2*...*pn+1.
> This number is not a multiple of any of the pm, so either it is prime,
> or it has a prime factor that is not in the list p1,p2,...,pn.  Either
> way, pn cannot have been the highest prime number.

Nicely done!

> It's also possible to show that if a and n have no common factor, then
> there are an infinte number of primes in the artihmetic progression
> a+dn as d=0,1,2,3,4,..., but the proof is, umm, somewhat harder (ie. I
> don't know it; I think it involves slinging Riemann zeta functions).
                                     ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

There's a good phrase.


From kent at tiamat.goathill.org  Fri Apr 21 16:23:59 2000
From: kent at tiamat.goathill.org (Kent Polk)
Date: 21 Apr 2000 20:23:59 GMT
Subject: Mac Python Questions
References: <IRjL4.271$CP3.3183@news.uswest.net> <38FE8032.DE233C29@cosc.canterbury.ac.nz> <956247497.216189@fezzik.endicor.com> <Pine.SO4.4.02.10004210015280.25675-100000@ub.d.umn.edu>
Message-ID: <956347062.17103@fezzik.endicor.com>

On Fri, 21 Apr 2000 00:16:48 -0500, Aaron J Reichow wrote:
>
>On 20 Apr 2000, Kent Polk wrote:
>
>> The standard Python installation program installed the wrong version
>> of the _tkinter shared library and(/or) didn't set the path to the
>> correct (ppc) version, so Python couldn't load _tkinter. If that
>> is the problem, there are several ways to solve it.  If that isn't
>> the problem, I'm not sure what to tell you as it's working on I
>> think 3 or 4 different mac platforms here.
>
>That's my problem.  I'm using OS 9, and I even added a path to where
>Tkinter appears to be, but it still cannot find _tkinter. <shrug>  This
>problem has been known for a while (I've heard of it since at least 1.5.1
>as being just a screw up with the path), so why hasn't it been fixed?

I have no idea where things are actually *supposed* to be, but the
installer placed _tkinter.CFM68k.slb in :Mac:PlugIns: and did not
install the _tkinter.ppc.slb there also. From what I saw, you can
either copy _tkinter.ppc.slb into :Mac:PlugIns: or add its location
to the Path in order to be able to import _tkinter. We are copying
to :Mac:PlugIns: as it appears that was the installer's intent.



From jeff_islay at my-deja.com  Sat Apr 22 07:49:52 2000
From: jeff_islay at my-deja.com (jeff_islay at my-deja.com)
Date: Sat, 22 Apr 2000 11:49:52 GMT
Subject: code not true?
Message-ID: <8ds3ku$q0l$1@nnrp1.deja.com>

Why does the second example return false?

>>> print 7.0 == 7
1
>>> print (.07 * 100) == 7
0


Sent via Deja.com http://www.deja.com/
Before you buy.


From robin at alldunn.com  Mon Apr 17 13:48:44 2000
From: robin at alldunn.com (Robin Dunn)
Date: Mon, 17 Apr 2000 10:48:44 -0700
Subject: New wxPython list
Message-ID: <007401bfa895$2e33e690$3225d2d1@ARES>

Since the Starship has flamed out I have created a new wxPython mail list.
I had planned on moving the list anyway so this seems like a good time to do
it.  Unfortunately since starship is dead I can't get the current subscriber
list (or the archives!) so you'll have to resubscribe.

You can subscribe at http://wxwindows.org/mailman/listinfo/wxpython-users or
send mail to wxpython-users-request at wxwindows.org


--
Robin Dunn
Software Craftsman
robin at AllDunn.com
http://AllDunn.com/robin/
http://AllDunn.com/wxPython/  Check it out!





From gmcm at hypernet.com  Mon Apr 24 12:43:45 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Mon, 24 Apr 2000 12:43:45 -0400
Subject: questions about imports and classes
In-Reply-To: <3904681D.FE39B41A@hologic.com>
Message-ID: <1255527869-51555877@hypernet.com>

Tim Iskander wrote:

> 2 - has there been any/much discussion on extending the import
> command to allow
>         renaming the module at import time,
>         ie: import myLongWindedModuleName as myShortName
> 
>         using the (new) keyword "as" to rename the module. We are
> dealing with a multi
>        developer, multitask application using lots of python
>        modules and
> we would like to
>         give the files non-ambiguous names without having unruly
>         module
> names in our code.

Until (if) we get "as", my advice is keep it simple.

import ModuleWithAReallyStupidLongName
SLN = ModuleWithAReallyStupidLongName
SLN.somethingorother()

> p.s. I am not very vi - literate and cannot get it to use spaces
> instead of tabs for autoindent... I will go and read the FAQ
> etc... but if its a simple answer that someone can send me I'd be
> really happy :-)
> 
A minimal .vimrc file would be

set expandtab
set sts=4
set shiftwidth=4

Now hitting the tab key will give you 4 spaces.

- Gordon



From mwh21 at cam.ac.uk  Thu Apr 20 16:11:50 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 20 Apr 2000 21:11:50 +0100
Subject: Good style for multi-valued returns
References: <889A6EB6AE399393.619B0539BFCAAE0F.5C6D110C3DA908F7@lp.airnews.net>
Message-ID: <m3hfcwh6rd.fsf@atrus.jesus.cam.ac.uk>

claird at starbase.neosoft.com (Cameron Laird) writes:

> Let me guess:  folks generally pass tuples for small and
> determinate-in-number multi-values, and dictionaries other-
> wise, 'cept some do it all with tuples.  Accurate?

I've always used tuples; I wouldn't have thought of something that
returns a dictionary as return multiple values.

Any reason for asking?

Cheers,
M.

-- 
59. In English every word can be verbed. Would that it were so in 
    our programming languages.
     -- Alan Perlis, http://www.cs.yale.edu/~perlis-alan/quotes.html


From jon at dgs.monash.edu.au  Fri Apr 14 19:51:43 2000
From: jon at dgs.monash.edu.au (Jonathan Giddy)
Date: Sat, 15 Apr 2000 09:51:43 +1000 (EST)
Subject: else clauses in while and for loops
In-Reply-To: <38F79B93.B03F0A6B@mindspring.com> from "Chuck Esterbrook" at Apr 14, 2000 06:28:35 PM
Message-ID: <200004142351.JAA02824@nexus.csse.monash.edu.au>

Chuck Esterbrook declared:
>
>So did anyone present, in this thread, an example that shows a real need?

The difficulty is what will you accept as a "real need"?  It sounds like you 
want an example that cannot be done any other way.  Even those who like else 
clauses in for and while will admit they're not essential.

Just like there's no "real need" for tuples (use lists), for loops (use
while), or elif (use else: if). 

All of these concepts are redundant in the sense that they can be mimicked
using an alternative sequence of code.  But they're nice because they
express higher-level concepts about the algorithm.

OTOH, a quick look at my code shows that I regularly use tuples, for
loops, and elif, but I rarely use else with anything but if and
try..except clauses.

The only example I found creates a list of usable temporary directories on
Unix, checking the environment first.  Certainly no real need here.

_tmpdirs = ['/tmp', '/var/tmp']

for envname in ['TMPDIR', 'TMP', 'TEMP']:
    try:
	paths = [os.environ[envname]]
    except KeyError:
	pass
    else:
	basepaths = validate_paths(paths)
	if basepaths:
	    break
else:
    basepaths = validate_paths(_tmpdirs)






From johngrayson at home.com  Sat Apr 22 11:47:41 2000
From: johngrayson at home.com (johngrayson at home.com)
Date: Sat, 22 Apr 2000 15:47:41 GMT
Subject: Tkinter's OptionMenu
References: <ZiYL4.110745$8S2.1383832@news1.rdc1.tn.home.com>
Message-ID: <8dship$83j$1@nnrp1.deja.com>

In article <ZiYL4.110745$8S2.1383832 at news1.rdc1.tn.home.com>,
  "Mike Callahan" <mcalla at home.com> wrote:
> How does one change an OptionMenu widget's list? I know that you can
do this
> easily using the PMW widget, but I would like to stay with basic
Tkinter
> widgets to keep the code as small as possible. I could not find any
methods
> in the code and I could not find any examples in either the online
> documentation or the new book on Tkinter.
>
>

I didn't expand on Tkinter's OptionMenu class too much in _Python and
Tkinter Programming_, since using Pmw's version is usually more
convenient. The problem is that you really have to 'roll your own'
to do much with the vanilla widget.

However, if you insist, here is some sample code to manipulate
Tkinter's OptionMenu:

from Tkinter import *
root = Tk()
root.title('OptionMenu')

class setit:
	def __init__(self, var, value):
		self.__value = value
		self.__var = var
	def __call__(self, *args):
		self.__var.set(self.__value)

class OptionM:
    def __init__(self,parent):

        self.var = StringVar()
        self.var.set('BRM')

        self.opt_menu = OptionMenu(parent, self.var, 'Stockbroker',
                                   'Quantity Surveyor', 'Church Warden',
                                   'BRM' )
        self.opt_menu.pack(anchor=W, padx=20, pady=30)

        Button(parent, text='Add item', command=self.addItem).pack()

        Button(parent, text='Change items',
               command=self.changeItems).pack()

    def addItem(self):
        self.opt_menu["menu"].add_command(label='Dinsdale',
                                  command=setit(self.var, 'Dinsdale'))

    def changeItems(self):
        self.opt_menu["menu"].delete(0, END)
        for item in ['Passion fruit','Loganberries','Mangoes in syrup',
                      'Oranges','Apples','Grapefruit']:
            self.opt_menu["menu"].add_command(label=item,
                                      command=setit(self.var, item))
        self.var.set('Loganberries')

om = OptionM(root)
root.mainloop()

A bit crude, but it's a start. Use Pmw if you can!

    John


Sent via Deja.com http://www.deja.com/
Before you buy.


From skip at mojam.com  Thu Apr  6 16:31:02 2000
From: skip at mojam.com (Skip Montanaro)
Date: Thu, 6 Apr 2000 15:31:02 -0500 (CDT)
Subject: [Python-Dev] Round Bug in Python 1.6?
In-Reply-To: <38ED0016.E1C4A26C@tismer.com>
References: <38ED0016.E1C4A26C@tismer.com>
Message-ID: <14572.62470.804145.677372@beluga.mojam.com>

    Chris> I happened to observe the following rounding bug.  It happens in
    Chris> Stackless Python, which is built against the pre-unicode CVS
    Chris> branch.

    Chris> Is this changed for 1.6, or might it be my bug?

I doubt it's your problem.  I see it too with 1.6a2 (no stackless):

    % ./python
    Python 1.6a2 (#2, Apr  6 2000, 15:27:22)  [GCC pgcc-2.91.66 19990314 (egcs-1.1.2 release)] on linux2
    Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
    >>> round(3.1415926585, 4)
    3.1415999999999999

Same behavior whether compiled with -O2 or -g.

-- 
Skip Montanaro | http://www.mojam.com/
skip at mojam.com | http://www.musi-cal.com/



From bozon at natur.cuni.cz  Tue Apr 11 11:23:06 2000
From: bozon at natur.cuni.cz (Michal Bozon)
Date: Tue, 11 Apr 2000 17:23:06 +0200
Subject: starship.python.net
Message-ID: <Pine.OSF.4.10.10004111721150.9584-100000@prfdec.natur.cuni.cz>

What the hell is with the starship.python.net server? Today and yesterday
I cannont connect (from Czech Republic).
Someone knows?

Michal Bozon



From maxm at normik.dk  Tue Apr 25 04:11:40 2000
From: maxm at normik.dk (=?iso-8859-1?Q?Max_M=F8ller_Rasmussen?=)
Date: Tue, 25 Apr 2000 10:11:40 +0200
Subject: SV: editor for win32
Message-ID: <7BD10B680501D411B9DF009027E06F32011A4E@Server.Exchange>

Fra: Scott Hathaway [mailto:slhath at flash.net]

> Can anyone recommend a python editor for Windows?

My favourite editor of 'em all is ultraedit. www.ultraedit.com It's
inexpensive and well worth the money.

It is pretty easy to edit the the wordfile to give it syntax highlighting
for python, and you can also make it run your scripts from the editor.

Everyone i know who has ever used it, has never used another editor on
windows...

Regards Max M



From richard_chamberlain at ntlworld.com  Mon Apr 10 06:51:50 2000
From: richard_chamberlain at ntlworld.com (Richard Chamberlain)
Date: Mon, 10 Apr 2000 11:51:50 +0100
Subject: check EOF
References: <8cs6g4$rje$1@porthos.nl.uu.net>
Message-ID: <8csblt$fv$1@gxsn.com>

Python returns an empty string as EOF so something like this is probably
easiest:

file=open("myfile","r")
for line in file.readlines():
    # do something with line

you could do something with while loops as well but this is probably
shorter.


"Soundwave" <maybe at lucky.com> wrote in message
news:8cs6g4$rje$1 at porthos.nl.uu.net...
> how can I check EOF in python ???
>
>




From julschae at usa.net  Sat Apr  8 05:07:57 2000
From: julschae at usa.net (Peter Kovalsky)
Date: Sat, 8 Apr 2000 11:07:57 +0200
Subject: Refering to substring of list containing strings
Message-ID: <MPG.13591514a982d0a4989682@nntp.server.uni-frankfurt.de>

Hi,

I am new to Python and lately experienced a little problem. I have a list 
with several strings and have to modify these strings. Therefore I used a
for-construct (for i in range(0, len(mylist)))- to operate every lists 
element. But how do I adress a substring in such a list? I tried a
	mylist[i][lborder:rorder]
- just like one would do with a simple string-var - wich won' t work. Any
help would be greatly appreciated,

thanx in advance,

Peter Kovalsky


From junkster at nospam.rochester.rr.com  Sat Apr 29 15:05:19 2000
From: junkster at nospam.rochester.rr.com (Benjamin Schollnick)
Date: Sat, 29 Apr 2000 19:05:19 GMT
Subject: ODBC / DBI help (via Win32 Extensions)
References: <rNHvjEdhm5Pp-pn2-RytgtQHbLg9d@d185d18f2.rochester.rr.com> <8ef01r$3c0$1@bob.news.rcn.net>
Message-ID: <rNHvjEdhm5Pp-pn2-tMI0tTCvJH2g@d185d18f2.rochester.rr.com>

On Sun, 29 Apr 3900 11:46:01, "Andy Smith" <andy at ictibroker.com> 
wrote:

> Calling the str method on a dbiDate will return a human-readable date.
> Also, calling int(myDbiDate) will return a 'seconds since the epoch'
> integer, which you can then easily work with using the functions in the
> 'time' module. I confess there may be methods of a dbiDate that can achieve
> the same, but I'm happy enough to just convert them to integers ;)
> 
> Here's an operation on dbiDates that should help you get an access-friendly
> format - (but remember to wrap in #'s cos that's crucial too!)
> 
> def dbToReuters(x):
>     a = time.localtime(int(x))
>     return time.strftime("%m/%d/%Y", time.localtime(int(x)))

Thanks, I'll try this on monday, and see what happens.... 

All of the development gets done at work, Thank God, or my
Girlfriend/Financee would kill me >g<  or worse (!!) pull the power
on the computers.... >g<

		- Benjamin


From mtanzer at bigfoot.com  Sun Apr 30 18:09:31 2000
From: mtanzer at bigfoot.com (mtanzer at bigfoot.com)
Date: Sun, 30 Apr 2000 22:09:31 +0000
Subject: FSBO Affiliate Program
Message-ID: <5F5DZPEU.T6TO6082@bigfoot.com>

Would you like to increase the income and hits on your website...at no cost to you?
You are invited to join our FSBO affiliate program. You were personally selected because, looking at your website, we see that you are active in working with FSBOs.  Please take a minute and visit http://www.home-buyer-seller-tips.com/affiliates.htm and discover how this program works and how it can bring you a sizable profit each month...and it's FREE!
*******************************************************
If you are not involved in FSBOs and wish to be removed from our mailing list, please type "remove" in the subject box of this message and return it to us.
*******************************************************
Milt Tanzer, Email: mtanzer at bigfoot.com
Real estate broker and counselor since 1973




From rgh at inmet.com  Tue Apr 25 19:00:01 2000
From: rgh at inmet.com (Randy Hudson)
Date: Tue, 25 Apr 2000 19:00:01 -0400
Subject: OOP books that compare - what about Python?
References: <vb5N4.247204$Hq3.6239116@news2.rdc1.on.home.com>
Message-ID: <39062371.6F30E31D@inmet.com>


Tom wrote:
> 
> I've been browsing two books that compare and/or teach OO languages: An
> Intro to OOP by Timothy Budd, and Objects Unencapsulated by Ian Joyner.
> Both seem good, both cover C++ and Java, but neither consider Python.

I don't know the Joyner book, but I started learning OOP from an earlier
edition of Budd's book, and I highly recommend it. What I like about his
approach is that he tries to explain the underlying concepts of OOP, 
and the variations on the themes, then
illustrate the concepts with the languages he's chosen to focus on. If
you work your way through the book, you'll be in a better position to
figure out Python yourself, how it's like and unlike other OO languages.

-- Randy Hudson


From mak at imakhno.freeserve.co.uk  Thu Apr  6 12:09:34 2000
From: mak at imakhno.freeserve.co.uk (Makhno)
Date: Thu, 6 Apr 2000 17:09:34 +0100
Subject: Some suggestion for a GUI kit evolved on Perl
References: <8cg7oo$m10$1@newsg4.svr.pol.co.uk> <8cgbj5$n1s$1@nntp6.u.washington.edu>
Message-ID: <8cijoc$gfs$1@newsg3.svr.pol.co.uk>

>label = OrcLabel
>NewOrcWindow(
>    title = "Test"
>).Add(
>    OrcVGroup(
>        OrcButton,
>        label
>    )
>).Show()

Yes, this appears to be the best way. I thought that Python statements could
only be continued onto a new line if a '\' was supplied, but perhaps I was
wrong.

I don't like the way the Label has to be pre-made (not created in the
constructor) in order to keep a reference. There's doesn't seem to be much I
can do about this though, unless I make it take a ref to the ref in it's
constructor perhaps?






From embed at geocities.com  Thu Apr 13 09:49:54 2000
From: embed at geocities.com (Warren Postma)
Date: Thu, 13 Apr 2000 09:49:54 -0400
Subject: python-embedded -"Reset?"
Message-ID: <cgkJ4.6719$HG1.191904@nnrp1.uunet.ca>

I have an embedded single process system, where I cannot have "short lived
processes". In fact, I can only have a single process running at all.  So,
one of the odd little things I need to do in this process is have my python
interpreter hard-linked into the program, and then run scripts. The problem
is that after a script completes it leaves a lot of junk around. Suppose I
want to clean that cruft out, or better still, prevent it in the first
place, do I have to do anything more/less than pass in a new local and
global namespace? Ie, is there anything else BESIDES the namespaces that
have to be cleaned up after some Python code executes?






From alex at somewhere.round.here  Wed Apr 12 12:50:08 2000
From: alex at somewhere.round.here (Alex)
Date: 12 Apr 2000 12:50:08 -0400
Subject: Python/Perl Popularity (Re: A Mountain of Perl...)
References: <200004102111.HAA03380@envy.fulcrum.com.au> <8d0pcq$7ti$1@nnrp1.deja.com> <dpg0sr37fw.fsf@mraz.iskon.hr>
Message-ID: <etdn1mz8drz.fsf@w20-575-109.mit.edu>

> >  -- How can I browse the Python documentation without firing
> >     up a web browser?
> 
> By using Info.  That's what I do, all the time.  Info has the added
> nicety of being able to search through the whole documentation for a
> keyword.

I love to use info, too.  Unfortunately, I can't get the latest version
of the documentation in that format, so I don't know whether I'm going
to be able to use it for much longer.

Alex.


From jstok at bluedog.apana.org.au  Tue Apr 18 12:41:56 2000
From: jstok at bluedog.apana.org.au (Jason Stokes)
Date: Wed, 19 Apr 2000 02:41:56 +1000
Subject: Lists
References: <A63AFB20111ED311AC5500A0C96B7BF5015F0F2C@orsmsx36.jf.intel.com>
Message-ID: <Hi0L4.7913$5D.17329@ozemail.com.au>

Daley, MarkX wrote in message ...
>This feels like a real newbie question, but here is the code that is
causing
>my question:
>
># Learning exceptions
>
>def test():
>    a = [-5, -4, -3, -2, -1, 0, 1, 2, 3, 4, 5]
>    print a
>    b = 5
>    for item in a:
>       try:
>           print b / a[item]
>       except ZeroDivisionError:
>           pass


Firstly, please retain correct indenting when posting code snippets.
Indenting is significant in Python.  I've corrected it for you.

>Why is the list being processed in reverse?  The exception works fine, but
>the math is backwards.  I don't recall this happening before.


Iteration one binds to the name item the first element of a, which happens
to be -5.

You then try to divide b by a[item], which is a indexed by -5.  Negative
indexes are legal in python and the effect is to return a reference to the
fifth element from the *right*.

In short, you've misapplied "for item in a" construct, which is very
different from the C construct.  In Python it repeatedly binds the name
"item" to the next element of list "a", does the sub-block, then repeats
until there's no longer any elements left in a.   You don't need to index a.
To do that, you'd do:

for i in range(0...len(a)):
   try:
       print b / a[i]
   except ZeroDivisionError:
      pass

But this will mark you with the stigma of a programmer hankering for his old
C days.... what you needed to write was merely:

for item in a:
   try:
       print b / item
   except ZeroDivisionError:
      pass

Cheers,
Jason Stokes: jstok at bluedog.apana.org.au




From arthur.haas at westgeo.com  Thu Apr 27 11:48:16 2000
From: arthur.haas at westgeo.com (Art Haas)
Date: 27 Apr 2000 10:48:16 -0500
Subject: Schwartzian Transform in Python?
Message-ID: <lr3do7tuin.fsf@yoda.wg.waii.com>

I'm looking for the Python way to do this ...

opendir(DIR,".") || die "Can't open '.'! $!\n"
@sorted =
	map { $_->[0] }
	sort { $a->[1] <=> $b->[1] }
	map { [ $_ , -M ]} readdir(DIR)
closedir(DIR);
print "Files sorted by date ...\n";
foreach (@times) {
	print "$_\n";
}

(Yes, its that other 'P' language).

I'm getting lost in the map/sort/map part for python. My python psuedo
code looks like (don't laugh too much)

import os
sorted = map(???,(map(os.stat???,listdir(".").sort())))
print "Files sorted by date ..."
for file in sorted:
	print file

I've been working with that language that dare not speak its name here
for about 5 years, but my python coding has been much much less. A
search on starship provided the nice p***/python page dealing with
lists, hashes/dictionaries, and strings, but I haven't found much
dealing with what is done above. The python documentation doesn't make
clear to me how to do things like "sort a list of lists by the second
field in each sublist", which is what the Schwartzian transform is
based on.

Any help/references would be welcomed.

-- 
###############################
# Art Haas
# (713) 689-2417
###############################


From gmcm at hypernet.com  Sun Apr 23 16:48:40 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Sun, 23 Apr 2000 16:48:40 -0400
Subject: partial parsing?
In-Reply-To: <8dvilp$88s$1@slb6.atl.mindspring.net>
Message-ID: <1255599573-47242942@hypernet.com>

Andrew Dalke wrote:

>   I'm interested in writing a parser generator which is somewhat
> different than yacc/flex, SPARK, Plex, etc.  Those parsers generate
> code which verifies every byte in the file and expects that I'm
> interested in most of the data.
> 
>   In my case, I have a lot of data (>100MB) in a known good format
> of which I'm only interested in a few items, but the specific items
> can change.

Can't answer you're direct question, Andrew, but I can tell you 
that I've used SPARK without a full grammar (although not on 
100MB files). I used an "outer scanner" and an "inner 
scanner". The outer just looked for interesting bits. When it 
found something, it passed it to the inner for full mastication.

If you need more sophistication, you probably need one where 
the parser can reach back into the scanner and change the 
rules on the fly. I think Aaron's stuff allows that, but I don't 
know for sure.


- Gordon



From kbd at iti-oh.com  Fri Apr 14 08:52:33 2000
From: kbd at iti-oh.com (Keith Davidson)
Date: Fri, 14 Apr 2000 12:52:33 GMT
Subject: Future compatibility of the Python C API
Message-ID: <lwEJ4.7980$Ib7.101092@typhoon2.kc.rr.com>

1) Has there been any discussion about the upwards compatibility of the C
API in future versions of Python?  Historically our libraries have had a
7-10+ year lifespan where we provide 95+% backwards compatibility (i.e. we
don't want our C API users to have to make more than a day or so worth of
changes between releases).  My big concern is Python 3000- what are the odds
that its C API will be backwards compatibility with Python 1.x?

2) We are considering using the Python C API heavily in a set of libraries
that will be used by both C applications and as Python extension modules.
Has anyone else done this?  How did it work out?

Thanks,
Keith




From mjackson at wc.eso.mc.xerox.com  Fri Apr 21 14:53:38 2000
From: mjackson at wc.eso.mc.xerox.com (Mark Jackson)
Date: 21 Apr 2000 18:53:38 GMT
Subject: docstrings
References: <6v61gs8b1op8esg3o7hf0hii97jlvma01r@4ax.com>
Message-ID: <8dq83i$1ea$1@news.wrc.xerox.com>

Ben Wolfson <rumjuggler at home.com> writes:
> On 20 Apr 2000 16:22:49 GMT, mjackson at wc.eso.mc.xerox.com (Mark
> Jackson) wrote:
> 
> >Ben Wolfson <rumjuggler at home.com> writes:
> >> is there a way to use %c and %s within a docstring, or to add strings
> >> together in a docstring, so that if some strings are used in all
> >> docstrings, but might be changed, I can simply change one or two
> >> variables?
> >
> >This might do what you want, although it's ugly (in the sense that it
> >serves the code-documentation - as opposed to user-documentation -
> >function less well):
> >
> >    cat > spam.py
> >    """I am a docstring"""
> >    
> >    sharedoc = """ and I am a widely-reused docstring clause"""
> >    
> >    __doc__ = __doc__ + sharedoc
> >    
> >    del sharedoc
> >    ^D
> >    yngvi> python
> >    Python 1.5.2 (#7, May  6 1999, 14:39:45)  [GCC 2.8.1] on sunos5
> >    Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
> >    >>> import spam
> >    >>> print spam.__doc__
> >    I am a docstring and I am a widely-reused docstring clause
> >    >>> 
> 
> This approach doesn't work for function docstrings, unless I tried it
> wrong.

Did I say my previous method was ugly?  On reflection I misspoke;
*this* is ugly:

<spam.py>

"""I am the main module docstring"""

sharedoc = """ but I am a widely-reused docstring clause"""

__doc__ = __doc__ + sharedoc

def eggs():
	"""I am eggs"""

	return 'yolk'

def bacon():
	"""I am bacon"""

	return 'grease'

def and_spam():
	"""I am and_spam"""

	return 'various parts'

for i in eggs, bacon, and_spam:
		i.__doc__ = i.__doc__ + sharedoc

<example>

>>> import spam
>>> spam.__doc__
'I am the main module docstring but I am a widely-reused docstring clause'
>>> spam.eggs.__doc__
'I am eggs but I am a widely-reused docstring clause'
>>> spam.bacon.__doc__
'I am bacon but I am a widely-reused docstring clause'
>>> spam.and_spam.__doc__
'I am and_spam but I am a widely-reused docstring clause'

-- 
Mark Jackson - http://www.alumni.caltech.edu/~mjackson
	An advertiser will happily make you feel bad about
	yourself if that will make you buy, say, a Bic pen.
				- George Meyer




From embed at geocities.com  Thu Apr 27 15:52:17 2000
From: embed at geocities.com (Warren Postma)
Date: Thu, 27 Apr 2000 15:52:17 -0400
Subject: Python 2.0 - case sensitivity can be a good thing, but don't abuse it.
References: <3907B52D.B01D5748@myself.com> <8e9c8t$hi6$1@newshost.accu.uu.nl> <vndln1zv72y.fsf@camelot-new.ccs.neu.edu>
Message-ID: <5X0O4.12719$HG1.340674@nnrp1.uunet.ca>

> Yow.  I would personally consider that one a disaster, but my vote
> doesn't count for much.  I very much like the idiom of an instance
> being a lowercased version of its class name, in some cases.

To me, any code that works like that would be a disaster:

class X:
    pass
x = X()

Yuck!  I guess it's a matter of taste though.  I tend to steer the course of
"if this works in Python but would be awful in C or Pascal, then I don't do
it that way". This is similar to avoiding having "regional dialect" stuff
creep into the professional writing I do. Case sensitivity or insensitivity,
or relying on either as part of my code style, therefore does not 'pass
muster'.

I am fine with Python staying case sensitive, because it forces neater
coding conventions, much like the tabification forces neater indentation,
and in the end, even beginners appreciate this.  :-)

Warren




From effbot at telia.com  Tue Apr 18 17:51:01 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Tue, 18 Apr 2000 21:51:01 GMT
Subject: Using python on the web
References: <MPG.1365adcfa8470eef9896aa@news> <8di2bk$l1s$1@newshost.accu.uu.nl> <MPG.136631472f892a119896ac@news> <_W2L4.986$GK5.20898@news2-win.server.ntlworld.com> <MPG.13665ef93eedbda29896ad@news>
Message-ID: <9N4L4.499$wYl.191816704@newsb.telia.net>

Charley Horse <someone at somewhere.com> wrote:
> It was just an unexpected surprise. I'm fairly new to all of this and
> always heard of CGI in the context of "you don't have to user CGI
> anymore because of the fancy new blah blah technology we offer!"

in Python land, that fancy technology is called
Zope (www.zope.org).

well worth checking out, but be careful -- your
brain might explode.

</F>




From danzap_python at yahoo.com.br  Thu Apr 13 12:40:13 2000
From: danzap_python at yahoo.com.br (=?iso-8859-1?q?Daniel=20Zapata?=)
Date: Thu, 13 Apr 2000 13:40:13 -0300 (ART)
Subject: How do I call a function from the WinApi ???
Message-ID: <20000413164013.28593.qmail@web4303.mail.yahoo.com>

I?d like to call a function in version.dll called
GetFileVersionInfo.

How do I do that ?

Thanks

____________________________________________________________________________
YAHOO! BATE-PAPO
Conhe?a gente nova, fa?a amigos e divirta-se! - http://www.yahoo.com.br



From dan at cgsoftware.com  Sun Apr 23 01:26:49 2000
From: dan at cgsoftware.com (Daniel Berlin+list.python)
Date: 23 Apr 2000 01:26:49 -0400
Subject: Python for drivers?
In-Reply-To: Moshe Zadka's message of "Sun, 23 Apr 2000 08:09:17 +0300 (IDT)"
References: <Pine.GSO.4.10.10004230806560.15546-100000@sundial>
Message-ID: <hfcttmjq.fsf@dan.resnet.rochester.edu>

Moshe Zadka <moshez at math.huji.ac.il> writes:

Also, considering you have to write interrupt handlers, you'd end up
with some very wacky code to prevent screwing yourself over.


> On Sun, 23 Apr 2000, jepler epler wrote:
> 
> > If you can use a tool like SWIG (or a hand written C module) to provide
> > access to the necessary low-level features, you may be able to do this.
> > 
> > For instance, if the card needs the idea of 'store {8,16,32}-bit integer X at
> > index L of the video card's address space' and 'read {8,16,32}-bit
> > integer at index L of the video card's address space', then you'd just
> > need to write those following functions in a C module for Python to call.
> 
> Um...unless you're talking about something like Hurd-like userspace
> drivers, it'd still be pretty close to impossible. In general, you avoid
> memory allocation and deallocation inside the kernel like the flu,
> especially few-bytes allocation. On the other hand, the Python interpreter
> usually allocates and dellocates like crazy. This would probably kill
> a Linux kernel.
> 



From scarblac-spamtrap at pino.selwerd.nl  Fri Apr 21 11:11:47 2000
From: scarblac-spamtrap at pino.selwerd.nl (Remco Gerlich)
Date: 21 Apr 2000 15:11:47 GMT
Subject: printing a list of doc strings
References: <8dnvf7$h0r$1@nnrp1.deja.com>
Message-ID: <slrn8g0v5c.ite.scarblac-spamtrap@flits104-37.flits.rug.nl>

schmitthead at my-deja.com wrote in comp.lang.python:
> What I had in mind was something like this:
> 
> for i in dir(os):
>     print eval( string.join( str(
> i ), "__doc__" ) )
> 
> Any advice?  I just want a list of all the items
> in os with their doc
> strings.

Well, it's something like that alright. How about:

for i in dir(os):
  print eval("os.%s.__doc__" % i)
  
Of course, this gives an exception for object that have no __doc__, like int
constants. 

Probably the best thing is to use __dict__:

for (name, object) in os.__dict__.items():
  if hasattr(object, '__doc__'):
    print "Docstring of object os.%s:" % name
    print object.__doc__
  else:
    print "Object os.%s has no docstring." % name

-- 
Remco Gerlich,  scarblac at pino.selwerd.nl
-rwxr-xr-x  1 root  5.89824e37 Oct 22  1990 /usr/bin/emacs [STR]


From nascheme at enme.ucalgary.ca  Tue Apr 18 15:43:34 2000
From: nascheme at enme.ucalgary.ca (Neil Schemenauer)
Date: Tue, 18 Apr 2000 19:43:34 GMT
Subject: __del__ problem (destructor called After members of a class are deleted?)
References: <aw2L4.8907$HG1.245743@nnrp1.uunet.ca>
Message-ID: <slrn8fpeob.r3h.nascheme@cranky.arctrix.com>

Warren Postma <embed at geocities.com> wrote:
>Why is __del__ being called with a self = None ?
...
>Oversimplified Code Snippet:
...

Yes, oversimplified.  There is nothing wrong with the code you
posted.  Try adding a "print repr(self)" to the __del__ method to
debug this.

    Neil

-- 
"If you're a great programmer, you make all the routines depend on each
other, so little mistakes can really hurt you." -- Bill Gates, ca. 1985.


From mskott at image.dk  Mon Apr 24 12:03:12 2000
From: mskott at image.dk (Martin Skøtt)
Date: Mon, 24 Apr 2000 16:03:12 GMT
Subject: Trouble compiling PyGreSQL 2.4 on Linux
References: <Pine.LNX.4.21.0004241211380.6019-100000@fep132.fep.ru>
Message-ID: <g0sbh8jd.fsf@jumbo.skott.dk>

Oleg Broytmann <phd at phd.russ.ru> writes:

>    I think even on RH you may do it "standard way" - install python-dev
> RPM, and you've got Makefile.pre.in and Setup.in with them. Copy these
> files to you work directory, modify Setup.in to your needs (to PyGreSQL
> needs this time :) and make -f Makefile.pre.in && make.

I know it works, but having a Linux README which didn't work bothered
me immensly.

> 
> > As I mention in my first posting the readme.linux file describes how
> > to do an red hat install (make redhat) but it doesn't seem to work. An
> > address is also mentioned for an RPM but the file doesn't exist on the server.
> 
>    D'Arcy would gladly accept any patch.

Seems like a great time for me to learn how to write a make file. I
will post it to him one of the following days.

-- 
Martin Sk?tt
mskott at image.dk

'In a certain sense, all decent programming languages are equally powerfull'
Dr. Alan Turing


From m.faassen at vet.uu.nl  Thu Apr 27 14:22:02 2000
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 27 Apr 2000 18:22:02 GMT
Subject: Python 2.0
References: <3907B52D.B01D5748@myself.com> <8e9c8t$hi6$1@newshost.accu.uu.nl> <vndln1zv72y.fsf@camelot-new.ccs.neu.edu>
Message-ID: <8ea0ga$kik$1@newshost.accu.uu.nl>

Justin Sheehy <dworkin at ccs.neu.edu> wrote:
> m.faassen at vet.uu.nl (Martijn Faassen) writes:

>>       * case insensitivity. foo = Foo() would break
>>  
>>     Opinions on them vary; they may reek a bit like Do-What-I-Mean,
>>     but nobody thinks either change would be a disaster.

> Yow.  I would personally consider that one a disaster, but my vote
> doesn't count for much.  I very much like the idiom of an instance
> being a lowercased version of its class name, in some cases.

> I guess it depends on one's definition of "distaster"...

Oh, I'm definitely with you. I like the idiom too. Let's not get
into a war on case sensitivity here though. Status report (AFAIK):

  * It's still unclear if it'll be in p3k. Might be in the editor/environment
    instead, as I and several people argued; for instance in Idle. It's
    catering to absolute newbies, after all. I saw one article that claimed
    that Guido said it would be one of the changes in p3k, but if that
    was right and if Guido changes his mind before the year 3000 is of course
    unclear. :)

  * There's the problem with "what does case insensitivity *mean*
    internationally speaking; do we teach the Python interpreter that, say,
    Egyptian hieroglyphs don't have cases? (if they don't and they're in
    Unicode :).

> There will be plenty of time to argue over this stuff before it
> becomes really relevant, in any case.

Right. Let's cancel this thread right now. I'm sorry I brought it up
in the first case. Just thinking about this "p3k will or won't break code"
thing after reading some claims here that it wouldn't break a lot.

Regards,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From kdahlhaus at yahoo.com  Sat Apr  8 15:00:33 2000
From: kdahlhaus at yahoo.com (Kevin Dahlhausen)
Date: Sat, 08 Apr 2000 15:00:33 -0400
Subject: PPython
References: <38ede548.353776233@news.oh.verio.com> <38EDF1ED.D102B7C9@searchy.net>
Message-ID: <38EF81D0.F9F28577@yahoo.com>

Penpal International wrote:

> I bet also your web hoster has support for python. It comes along with
> almost every ditribution of Linux. (95 % of all webservers are working
> with linux).
> Why you don't try it out. Make a little script, put it in the cgi-bin
> directory (Don't forget to chmod it 0755). you also can login with
> telnet and type: python  . If you get a sort of shell you know for sure
> you can execute python scripts.

That's a great idea, only the site is on an NT machine (GRrrrrrrrrrrrrrrr).
Files ending in .cgi or .pl are run as perl scripts.  I haven't talked to them yet
about installing Python.    I thought one option might be to write a perl wrapper
that spawns the CGI python interpreter.   I'd rather talk to them first though than
be sneaky like that.




From nmm1 at cus.cam.ac.uk  Mon Apr 10 05:30:55 2000
From: nmm1 at cus.cam.ac.uk (Nick Maclaren)
Date: 10 Apr 2000 09:30:55 GMT
Subject: Python paradigms
References: <Pine.LNX.4.21.0004100757580.19783-100000@fep132.fep.ru>
Message-ID: <8cs70f$ojf$1@pegasus.csx.cam.ac.uk>

In article <Pine.LNX.4.21.0004100757580.19783-100000 at fep132.fep.ru>, Oleg Broytmann <phd at phd.russ.ru> writes:
|> On Mon, 10 Apr 2000, Robert W. Cunningham wrote:
|> 
|> > Oleg Broytmann wrote:
|> > > > python's "and" and "or" operators doesn't evaluate the
|> > > > second argument unless they really have to...
|> > >
|> > >    Yes, I missed that, sorry.
|> > 
|> > So, it works just fine, right?
|> 
|>    No. Code is intended not only to beeing run on processor, but also (and
|> often more important) to beeing read by human. Code is speech, it is the
|> way to interchange ideas.
|>    In that sense ugly code is not working even if you add three books of
|> comments. Code should be readable. Period.

Yes, precisely.  In the real world, programs often have long lifetimes,
are worked on by many programmers (in parallel and sequentially), are
sometimes transliterated between languages, and it is common to need
to modify code which was written 20 years ago by some long-retired
programmer.

At best, confusing code wastes effort.  At worst, it causes mistakes
to be introduced.  I have seen the latter effect many times.

That is why I was talking about the benefits of those paradigms for
code clarification, and neither efficiency nor compactness.  Python
has some very useful facilities, but is a bit tedious for some kinds
of scripting.


Regards,
Nick Maclaren,
University of Cambridge Computing Service,
New Museums Site, Pembroke Street, Cambridge CB2 3QG, England.
Email:  nmm1 at cam.ac.uk
Tel.:  +44 1223 334761    Fax:  +44 1223 334679


From roy at popmail.med.nyu.edu  Wed Apr  5 09:31:15 2000
From: roy at popmail.med.nyu.edu (Roy Smith)
Date: Wed, 05 Apr 2000 09:31:15 -0400
Subject: What is python's language level?
References: <38EA1B8F.A797A130@lmco.com> <roy-D857DA.21212304042000@netnews.nyu.edu> <38EA9CD4.3473B910@callware.com> <roy-CF57ED.23102504042000@netnews.nyu.edu> <ED6060DB649A6FED.125EA090C5A03FDE.AA052F5C80E50E94@lp.airnews.net>
Message-ID: <roy-F7834A.09311505042000@netnews.nyu.edu>

claird at starbase.neosoft.com (Cameron Laird) wrote:
>> Have you ever looked at the output of a bib | tbl | eqn pipeline?
> Are you kids still using that as a pick-up line?

I'm a pragmatist.  As long as it keeps working, I'll keep using it :-)

-- 
Roy Smith <roy at popmail.med.nyu.edu>
New York University School of Medicine



From mjackson at wc.eso.mc.xerox.com  Tue Apr  4 14:31:15 2000
From: mjackson at wc.eso.mc.xerox.com (Mark Jackson)
Date: 4 Apr 2000 18:31:15 GMT
Subject: What is python's language level?
References: <38EA1B8F.A797A130@lmco.com>
Message-ID: <8cdcdj$rbm$1@news.wrc.xerox.com>

Chuck Meyers <chuck.meyers at lmco.com> writes:
> Does anyone know python's language level? I did not find it in the list
> at:
> 
> http://www.spr.com/library/0langtbl.htm
> 
> I would guess that it would be somewhere between 15 and 30. This is an
> important point, even if language levels are imprecise they can make it
> easier to convince people to look at or try a language like python.

Python's language level is 42.  At least that's the way it's written,
but it's pronounced 'Throatwobbler Mangrove.'  Thanks for asking.

-- 
Mark Jackson - http://www.alumni.caltech.edu/~mjackson
	An advertiser will happily make you feel bad about
	yourself if that will make you buy, say, a Bic pen.
				- George Meyer




From danzap_python at yahoo.com.br  Thu Apr 13 12:36:29 2000
From: danzap_python at yahoo.com.br (=?iso-8859-1?q?Daniel=20Zapata?=)
Date: Thu, 13 Apr 2000 13:36:29 -0300 (ART)
Subject: Getting file version information
Message-ID: <20000413163629.28006.qmail@web4303.mail.yahoo.com>

I?d like to know how to get the file version
information in Windows. I?m using pythonwin and
couldn?t find the GetFileVersionInfo API on the
win32api module.

Thanks

Daniel Zapata

____________________________________________________________________________
YAHOO! BATE-PAPO
Conhe?a gente nova, fa?a amigos e divirta-se! - http://www.yahoo.com.br



From pete at visionart.com  Mon Apr 24 19:00:44 2000
From: pete at visionart.com (Pete Shinners)
Date: Mon, 24 Apr 2000 16:00:44 -0700
Subject: new linereading standard?
Message-ID: <3904D21B.333DD930@visionart.com>

i've come to the point where i can't get by even the simplest python
project without dumping this code into my files. it is a simple class
that handles the "while 1: .... break" carnival for reading files.

i've seen a couple arguments back and forth of the cleanliness of this
syntax, and i never payed strong attention, but i must admit i find it
both ugly and ackward.

i'm sure someone has posted a class like the before, but i'm hoping it
can one day make it in as a python standard someday. (6.0, there's still
time!)

class filelines():
    def __init__(self, file):
        self.file = file
    def __getitem__(self, index):
        line = self.file.readline()
        if not line: raise IndexError
        return line

now i can code in a preferred style

myfile = open('myfile.txt', 'r')
for line in filelines(myfile):
    print line
myfile.close()

as opposed to

myfile = open('myfile.txt', 'r')
while 1:
    line = myfile.readline()
    if not line: break
    print line
myfile.close()

as a person trying to introduce his friends to python, i repeatedly find
this is the ugliest wart while doing basic stuff with python.

anyways, i now use this code ALWAYS. unfortunately, i haven't found a
good way to make this part of my standard environment while working on
many different machines.

if someone wants, i'm close to creating a much improved version of this
class that can optionally handle trickier situations (like continuing
lines ending with \). perhaps also allowing input of a filename string
and perform the needed open and close requests as well.

anyways, what are peoples thoughts if someone was to create a more
robust class like this, would or could it ever be grafted into a later
python release?

(i'm also anxious for some += style operators, and a := operator that
performs assignment and returns the assigned value (as mentioned in the
FAQ))

another simple option would be to have the filehandle objects evaluate
to boolean 0,FALSE when you cannot read from the file

--
_____Pete Shinners______
   pete at visionart.com





From mwh21 at cam.ac.uk  Tue Apr 18 06:03:57 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 18 Apr 2000 11:03:57 +0100
Subject: Here is some starship.python.net news
References: <38FA9A30.33DD1994@concentric.net> <38FB90B4.35E9CA1B@bullfrog-tech.com> <38FBE4DB.67289F42@mindspring.com> <20000418102817.J15664@xs4all.nl>
Message-ID: <m3g0sjd8tu.fsf@atrus.jesus.cam.ac.uk>

Thomas Wouters <thomas at xs4all.net> writes:

> Official mirrors ? I'd be willing to lobby for a starship (and
> python.org ?)  mirror at XS4ALL (dutch ISP, where I work, so it
> shouldn't be too hard to arange) but it kind of depends on the needs
> for the starship website -- having different platforms it runs on
> might screw up the scripts that run there. Would having a
> data/webpage only mirror be sufficient ?

Wouldn't help those of use who use python.net email addresses, but a
web mirror sounds like a good idea.  Aren't there issues to do with
making sure all the people with content there assent to having their
material mirrored?

Cheers,
M.

-- 
39. Re graphics:  A picture is worth 10K  words - but only those 
    to describe the picture. Hardly any sets of 10K words can be 
    adequately described with pictures.
     -- Alan Perlis, http://www.cs.yale.edu/~perlis-alan/quotes.html


From mrisney at us.oracle.com  Fri Apr 14 14:33:36 2000
From: mrisney at us.oracle.com (Marc Risney)
Date: Fri, 14 Apr 2000 11:33:36 -0700
Subject: problems with DCOracle
Message-ID: <8d7o6h$7vt$1@inet16.us.oracle.com>

has anyone experienced the same problem that
I experiecne with the DCOracle module.

    when attempting to connect to an  Oracle SID, my function works fine
while using python interactively, but as a script, it fails.

very odd.

Marc Risney
Advanced Technologies
Oracle Redwood shores
mrisney at us.oracle.com




From PA0215481 at ntu.edu.sg  Tue Apr 25 02:02:52 2000
From: PA0215481 at ntu.edu.sg (#VEDALA NARASIMHA KUMAR#)
Date: Tue, 25 Apr 2000 14:02:52 +0800
Subject: Simple list.append() question
References: <Pine.GSO.4.10.10004250813240.20737-100000@sundial>
Message-ID: <0CF260C495FED111A6610000F866308D0D893596@mail3.ntu.edu.sg>

l=[]
i=0
while 1:
	if i <3:
		l.append([])
	else:break
	i=i+1
l[0].append(3)

 crude..but works 


On Mon, 24 Apr 2000, Bill Wilkinson wrote:

> l = [[][:]] * 3

No, this won't work.




From greg at cosc.canterbury.ac.nz  Wed Apr 19 22:22:23 2000
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Thu, 20 Apr 2000 14:22:23 +1200
Subject: dir(..)   vs.   PyList_GetItem
References: <213CDEBFDDA7D3118081009027E40C56071159@SHEF-MS1>
Message-ID: <38FE69DF.593EDAC4@cosc.canterbury.ac.nz>

Mark Pope wrote:
> 
> The code I posted was just a snippet.
> I have already successfully imported the module. I now need to get a list of
> the functions it exposes.

Are you *sure* you have imported the module properly?
Because the result you posted looked an awful lot like
you were looking at an empty module. Note that even if
you imported the module previously, PyModule_New will
*still* create a new module object. 

Can you post the code you actually ran?

If you've got the right module object, I can't imagine
why looking at the keys of the module dict wouldn't
work -- particularly since it works in Python, too:

>>> import string
>>> dir(string)
['__builtins__', '__doc__', '__file__', '__name__', '_idmap', '_idmapL',
'_lower', '_re', '_safe_env', '_swapcase', '_upper', 'atof',
'atof_error', 'atoi', 'atoi_error', 'atol', 'atol_error', 'capitalize',
'capwords', 'center', 'count', 'digits', 'expandtabs', 'find',
'hexdigits', 'index', 'index_error', 'join', 'joinfields', 'letters',
'ljust', 'lower', 'lowercase', 'lstrip', 'maketrans', 'octdigits',
'replace', 'rfind', 'rindex', 'rjust', 'rstrip', 'split', 'splitfields',
'strip', 'swapcase', 'translate', 'upper', 'uppercase', 'whitespace',
'zfill']
>>> string.__dict__.keys()
['_safe_env', '_idmap', '_swapcase', 'capitalize', 'atoi_error',
'octdigits', 'rfind', 'index', 'expandtabs', 'center', 'join', 'lower',
'digits', 'atof', '__name__', 'hexdigits', 'atol', 'zfill', 'count',
'atoi', 'atof_error', '_re', 'swapcase', '__file__', '__builtins__',
'joinfields', 'lowercase', '__doc__', 'translate', 'rstrip', 'split',
'splitfields', '_lower', 'atol_error', 'rindex', 'whitespace',
'_idmapL', '_upper', 'rjust', 'find', 'ljust', 'strip', 'index_error',
'capwords', 'replace', 'maketrans', 'uppercase', 'lstrip', 'letters',
'upper']
>>>

The only difference seems to be that dir() sorts the
list.

-- 
Greg Ewing, Computer Science Dept,
+--------------------------------------+
University of Canterbury,	   | A citizen of NewZealandCorp, a	  |
Christchurch, New Zealand	   | wholly-owned subsidiary of USA Inc.  |
greg at cosc.canterbury.ac.nz	   +--------------------------------------+


From quinn at ngwee.ugcs.caltech.edu  Mon Apr 24 00:26:16 2000
From: quinn at ngwee.ugcs.caltech.edu (Quinn Dunkan)
Date: 24 Apr 2000 04:26:16 GMT
Subject: Sorting a list of classes
References: <OmnqwHar$GA.233@cpmsnbbsa03>
Message-ID: <slrn8g7jbo.omn.quinn@ngwee.ugcs.caltech.edu>

On Mon, 24 Apr 2000 00:08:49 -0400, lexberezhny <lexberezhny at email.msn.com>
wrote:
>If i have a python list of classes, can i do something like classList.sort()
>and how can i control this behavior within the classes. ie
>i have a address book, and a have a class called 'Person', and it has an
>attribute Person.Name, so i have a list of 'Person's how can i sort the
>list, and maybe sort it by name? do i have to write my own function or can i
>use list.sort() some how?

See the docs for sort():

addr_book.sort(lambda a, b: cmp(a.name, b.name))

Alternately, since you wanted the logic in the class (see the lang ref under
"special method names"):

class Person:
    ... methods ...
    def __cmp__(self, other):
        return cmp(self.name, other.name)


From jae at ilk.de  Fri Apr 21 22:07:25 2000
From: jae at ilk.de (Juergen A. Erhard)
Date: Sat, 22 Apr 2000 04:07:25 +0200
Subject: Pythonware Website???? Gone on vacation with Starship? ;-))
In-Reply-To: <8dqp6u$plm$1@newshost.accu.uu.nl> (m.faassen@vet.uu.nl)
References: <1255771354-36910521@hypernet.com> <8dqp6u$plm$1@newshost.accu.uu.nl>
Message-ID: <22042000.1@sanctum.jae.ddns.org>

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

>>>>> "Martijn" == Martijn Faassen <m.faassen at vet.uu.nl> writes:

    Martijn> Gordon McMillan <gmcm at hypernet.com> wrote:
    >> Martijn Faassen wrote:

[snipped that dangerous knowlegde... at least they can't say I was
spreading it further...]
 
    Martijn> Do-you-have-to-kill-me-now?-ly yours,

No, they probably won't kill you... but whatever they do to you,
you've doomed us all... I'll go into hiding now, I know to much.

Do-you-know-a-good-(and-expendable)-cosmetic-surgeon?-ly y'rs, J

PS: Of course any trace of your account of what happened at Roswell
will have to be erased... so, the starship crash was just the
beginning.

PPS: The phone rang just once, then stopped... I better hurry.

- -- 
J?rgen A. Erhard           eMail: erased           phone: disconnected
                       My WebHome: gone forever
           "Outside of a dog, a man's best friend is a book;
           inside of a dog, it's very dark." --  Groucho Marx
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.1 (GNU/Linux)
Comment: Use Mailcrypt and GnuPG <http://www.gnupg.org/>

iEYEARECAAYFAjkBCVoACgkQN0B+CS56qs1d1QCgpYo5AqcSSQsCP1P+yDcJ4M80
DmsAn0aZt8aqceuMNtURVNLIhToZn2X8
=3hWp
-----END PGP SIGNATURE-----



From gward at mems-exchange.org  Fri Apr  7 08:49:15 2000
From: gward at mems-exchange.org (Greg Ward)
Date: Fri, 7 Apr 2000 08:49:15 -0400
Subject: [Python-Dev] Round Bug in Python 1.6?
In-Reply-To: <200004062130.RAA26273@eric.cnri.reston.va.us>; from guido@python.org on Thu, Apr 06, 2000 at 05:30:21PM -0400
References: <38ED0016.E1C4A26C@tismer.com> <200004062130.RAA26273@eric.cnri.reston.va.us>
Message-ID: <20000407084914.A13606@mems-exchange.org>

On 06 April 2000, Guido van Rossum said:
> This is because repr() now uses full precision for floating point
> numbers.  round() does what it can, but 3.1416 just can't be
> represented exactly, and "%.17g" gives 3.1415999999999999.
> 
> This is definitely the right thing to do for repr() -- ask Tim.
> 
> However, it may be time to switch so that "immediate expression"
> values are printed as str() instead of as repr()...

+1 on this: it's easier to change "foo" to "`foo`" than to "str(foo)" or
"print foo".  It just makes more sense to use str().

Oh, joy! oh happiness! someday soon, I may be able to type
"blah.__doc__" at the interactive prompt and get a readable result!

        Greg



From effbot at telia.com  Thu Apr 20 11:54:55 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Thu, 20 Apr 2000 15:54:55 GMT
Subject: closures
References: <38FF2069.1F25@marshall.edu>
Message-ID: <jLFL4.3233$Za1.47715@newsc.telia.net>

Mark Hathaway <hathawa2 at marshall.edu> wrote:
> I've been discussing OO with Smalltalkers on comp.object. The
> issue/topic of blocks (as used in Smalltalk) comes up frequently.
> I am wondering how Python handles 'if's and 'for's which is where
> Smalltalk commonly uses it's blocks. How does Python do it?
>
> if <condition>:
>     <statements>
>
> for each in list:
>     <statements>

the statement blocks are inlined.

</F>




From mdefreitas at sikorsky.com  Thu Apr 27 13:00:36 2000
From: mdefreitas at sikorsky.com (mdefreitas at sikorsky.com)
Date: Thu, 27 Apr 2000 17:00:36 GMT
Subject: Newbie Question: Is there a Python equiv to Perl's "do" statement?
Message-ID: <8e9rna$ajh$1@nnrp1.deja.com>

I'm trying to call a python script from another python script (given
the filename).

Thanks.

Marco


Sent via Deja.com http://www.deja.com/
Before you buy.


From nmm1 at cus.cam.ac.uk  Mon Apr 10 16:07:18 2000
From: nmm1 at cus.cam.ac.uk (Nick Maclaren)
Date: 10 Apr 2000 20:07:18 GMT
Subject: Python paradigms
References: <8cmr70$b0e$1@pegasus.csx.cam.ac.uk> <3H4I4.709$4w1.17557@typhoon.nyroc.rr.com> <8cs3ho$l06$1@pegasus.csx.cam.ac.uk> <slrn8f4be7.1t0.bwinton@tor.dhs.org>
Message-ID: <8ctc9m$sp2$1@pegasus.csx.cam.ac.uk>

In article <slrn8f4be7.1t0.bwinton at tor.dhs.org>,
Blake Winton <bwinton at iname.com> wrote:
>>
>>No.  Firstly, it relies on an artifact of my example (i.e. that it is
>>in an assignment, rather than a test condition or function call) and,
>>secondly, the more lines that code takes the harder it is to see the
>>function in its entirety.
>
>Now perhaps it's just because I've been forced to switch brace styles at
>work, but this comment kind of rubs me the wrong way.  Even on the linux
>console (possibly the least configurable text mode I know), you can get
>80 lines on the screen at a time.  It's my (not so) humble opinion that
>if your functions are 80 lines long, then you're probably doing
>something wrong.  Even if you have to break your single line up into
>four or five, surely it's not causing your functions to run off the end
>of the screen, or if it is, perhaps it's time to break your functions up
>a little.

It really is quite amusing to see these same old myths coming back
every few decades :-)

That particular dogma was very popular back in the late 1960s and
early 1970s, when practical programmers pointed out that it was a
virtually impossibility for a normal person to keep track of 50
trivial functions in their head, but it was quite easy to look
at 60 lines of sequential code.  That didn't get listened to
then, either!

See Les Hatton "Safer C" for some evidence of the disadvantages
of "mincing" code.

My current infliction is a accounting file format, and the the
function to check the input (you DO check your input, don't you?)
needs to perform 100 tests in series.  Yes, I am inventing
trivial functions, but it would be clearer to put more of the
tests inline.

Incidentally, the smallest number of lines that I have ever had
to program through was 6, and the maximum number on the early
IBM PCs was 20.  BBC Micros did better at 24 - one of their many
advantages.


Regards,
Nick Maclaren,
University of Cambridge Computing Service,
New Museums Site, Pembroke Street, Cambridge CB2 3QG, England.
Email:  nmm1 at cam.ac.uk
Tel.:  +44 1223 334761    Fax:  +44 1223 334679


From fredrik at pythonware.com  Wed Apr 26 08:02:10 2000
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 26 Apr 2000 14:02:10 +0200
Subject: Python 1.6a2 Unicode bug (was Re: comparing strings and ints)
References: <jr%I4.643$rc9.190628352@newsb.telia.net><1256565470-46720619@hypernet.com><6M1J4.662$rc9.209708544@newsb.telia.net><8daop0$8fk$1@slb6.atl.mindspring.net><d3HK4.2612$Za1.38971@newsc.telia.net> <8e6bsl$f1a$1@nnrp1.deja.com> <87ln212k8e.fsf@psyche.evansnet>
Message-ID: <00b501bfaf77$421f2690$0500a8c0@secret.pythonware.com>

Carey Evans wrote:
> In the latest versions of Perl and Tcl, aren't all strings Unicode
> strings anyway?

yes.  or no, depending on how you view things.

the important issue is that in both languages, a string is always
a sequence of characters.  the length of a string is always the
number of characters, independent of the internal encoding.
indexing and slicing works as expected.  etc.

or in short, whatever internal encoding a string happens to
use, it's usually not visible for the programmer.

that's not true for python 1.6a2.

> It seems to me you're still in for _less_ of a surprise in Python 1.6.

have you played with the current 1.6 implementation?

</F>




From noreply at antispam.com  Mon Apr  3 14:42:25 2000
From: noreply at antispam.com (Brian Schmidt)
Date: Mon, 03 Apr 2000 18:42:25 GMT
Subject: lightweight embedded database table object
References: <sS4G4.2894$HG1.86777@nnrp1.uunet.ca>
Message-ID: <lC5G4.102$Ta6.86472@paloalto-snr1.gtei.net>

> I am trying to find or build a small python database for a deeply embedded
> Python application, that runs on an embedded piece of hardware we're
> developing. The hardware is
> basically PC compatible, and runs an embedded realtime operating system.
> Python runs great on it, but I have to have ALL C modules statically
linked
> in.

Hi Warren,
Have you looked at MetaKit? http://www.equi4.com/metakit/python.html

"Warren Postma" <embed at geocities.com> wrote in message
news:sS4G4.2894$HG1.86777 at nnrp1.uunet.ca...
>
> I am trying to find or build a small python database for a deeply embedded
> Python application, that runs on an embedded piece of hardware we're
> developing. The hardware is
> basically PC compatible, and runs an embedded realtime operating system.
> Python runs great on it, but I have to have ALL C modules statically
linked
> in.
>
> I have attached some sample code to show the sorts of things I'm trying to
> do.  The performance of the pure-python isn't great, I'd rather use a
> compiled C python module
> that did similar things. Has anyone written one.
>
> I've looked at Gadfly, and since Gadfly can consume loads of RAM, it's not
> exactly ideal for me. I don't really need an SQL grammar, but if it
weren't
> for the RAM issue, I'd consider it.
>
> The lwdb module I've attached uses BSDDB v2 as the disk-based
> store-and-retrieve method.
>
> It adds ascii-delimited (CSV) import and export features (not well tested
> yet), and a trivial assortment of types (say, float, string and integer
for
> now, but I'll be adding others like datetime fields, etc).
>
> I'd like an ultra-compact binary storage method and maybe even a very-fast
> compression method, and I'm thinking of adding support for the "ROTOR"
> builtin Python module, so that the database store can be somewhat
"opaque".
> I will be storing user names and passwords in there, so I'll probably want
> to add some encryption.
>
> Perhaps all those things I can do with Python plus a little marshalling or
> pickling are worth the overhead of a BSDDB file, which seems to use about
> 200K of disk space for every 100K of data in the variable-length-record
> field.
>
> I've considered things like dBase files, but they are such a crusty old
> standard, really.  If I had to go that way, I would buy Codebase from
> Sequiter Software, and whip up a C to Python interface to them. Anyone
else
> done any Python+Codebase work?
>
> I don't really need wicked-fast lookup on disk. I'm wondering if I should
> just invent my own C module to do this job.  Then again, I could be very
> foolish, as any "database" code you write ends up always being more
> complicated than you originally thought.
>
> Warren Postma
>
> ----
>
> attached sample code:
>      lwdb.py - lightweight database table object (stores in a bsd
database)
>      lwdbrec.py - lightweight database record object (primitive
> typed-database-'tuple')
>
> [I have no idea of attachments bounce from the mailing list or newsgroup,
or
> if they considered In Bad Taste, but as they are small, I'm going ahead
> anyways....]
>
>
>
>
>
>




From greg at cosc.canterbury.ac.nz  Wed Apr 12 21:40:42 2000
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Thu, 13 Apr 2000 13:40:42 +1200
Subject: [ANNOUNCE] Garbage collection for Python
References: <000a01bfa3cf$159c8440$7f2d153f@tim>
Message-ID: <38F5259A.B51EFDE2@cosc.canterbury.ac.nz>

Tim Peters wrote:
> 
> See Py_NewInterpreter() in the
> "Initialization, Finalization, and Threads" section of the Python/C API
> manual.  There isn't a Python-level interface to this yet;

That's an intriguing idea! There's been a notion in the back
of my mind that a Python IDE, such as IDLE, should be able to
run the program under development in a separate Python
environment, with its own global module list, etc., so that
you can cleanly throw it all away and restart without any
hassles over reloading changed modules, etc.

This could be quite easy to do if there were a Python
function for creating a new Python interpreter!

One thing that might get a bit hairy is that, for debugging,
code running in the first-level interpreter would need to
be able to access objects belonging to the second-level
interpreter. Would this sharing of objects between
interpreters lead to anything bad happening?

The other thing is what to do about extension modules.
Ideally it should be possible to load multiple "instances"
of an extension module into different interpreters, and
there should be a way of finalising one of these instances
when the interpreter it belongs to goes away. I don't think
that's possible at the moment, because extension modules
don't have any finalisation entry point.

-- 
Greg Ewing, Computer Science Dept,
+--------------------------------------+
University of Canterbury,	   | A citizen of NewZealandCorp, a	  |
Christchurch, New Zealand	   | wholly-owned subsidiary of USA Inc.  |
greg at cosc.canterbury.ac.nz	   +--------------------------------------+


From mhammond at skippinet.com.au  Mon Apr  3 19:45:04 2000
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Mon, 03 Apr 2000 23:45:04 GMT
Subject: embedding python question
References: <38e922ef$0$27307@wodc7nh7.news.uu.net>
Message-ID: <42aG4.4037$1b4.19463@news-server.bigpond.net.au>

"Dave Barrett" <dave at humancode.com> wrote in message
news:38e922ef$0$27307 at wodc7nh7.news.uu.net...

> have to be exact, just close.  How would I go about doing this
with the
> embedded api ?  Is it possible ?  If not, are they any similiar
things I can
> try ?

Python doesnt (yet?) provide anything to make this easier than
most other languages.

So you need to solve this problem without thinking Python.
Imagine you wanted the exact same thing, but the little bits of
code you wanted to call were also written in C.  There are a
number of "solutions", none of which are particularly elegant,
but whatever design come up with will also work with Python.  But
unfrotunately Python doesnt provide anything to make it simpler
than it sounds...

In a nut-shell, what you want is not trivial :-)

Mark.




From effbot at telia.com  Sun Apr  2 05:16:31 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Sun, 02 Apr 2000 09:16:31 GMT
Subject: tcl 8.2 OK?
References: <38E4BE16.497FB1B@tcac.net>
Message-ID: <PdEF4.891$n68.190327808@newsb.telia.net>

Albert Wagner wrote:
> For use by tkinter I installed tcl/tk8.2 (I couldn't make 8.0.4
> compile).  Is this OK? Or can I expect trouble later?

if you can get a simple program to work (like the test code in
Tkinter.py), it's probably okay.

there are some initialization issues with Tk 8.1, but they may be
fixed in 8.2.  if you get messages like "cannot find usable tk.tcl in
the following directories" followed by a number of empty lines,
you need to patch the _tkinter module.

(don't have a pointer handy; if you need the patch and cannot
find it, followup to this post).

</F>




From JamesL at Lugoj.Com  Thu Apr  6 00:03:04 2000
From: JamesL at Lugoj.Com (James Logajan)
Date: Wed, 05 Apr 2000 21:03:04 -0700
Subject: Imaginative solutions to connect. Was: Python 1.6 alpha 1 released
References: <Pine.GSO.4.10.10004040035420.21084-100000@sundial> <5E877FD7637AB46E.DABBED9CF3C1D64B.CDDDC05AE7374A78@lp.airnews.net> <5l7leern5v.fsf@eric.cnri.reston.va.us>
Message-ID: <38EC0C78.C6170660@Lugoj.Com>

Guido van Rossum wrote:
> 
> Repeat after me: SOCKETS ARE NOT JUST FOR IP CONNECTIONS.
> 
> But I will fix the code -- it's the only practical thing to do.

Yes, I think a number of us understood sockets are address family neutral
but had pragmatic concerns. And as one of the first to probably fill your
mailbox (thought I would be in a minority!) it is gratifying to see a
pragmatist like yourself accede to the angry and discontented rabble before
anything unpleasant took place. <wink>

But that still leaves a problem about how to clean things up. Not sure that
there are a lot of alternatives, but here are some for a start (feel free to
add to this list!):

1) Force the issue only with Py3k but not before.
2) Create an additional class, maybe called socket1 or Socket (or TLI
<wink>) with the correction.
3) Create additional methods in socket, say Connect() or connect1() (or
correct() <wink>) [same for Bind too of course].
4) Kill all the programmers who used the wrong arguments and burn all their
code (I was thinking of changing careers anyway).

For alternatives 2 and 3 you need some incentive for people to change; cash
might work but could get expensive. Rigging keyboards to deliver electrical
shocks might work but might be too time consuming. Making the old interface
much slower might help. Perhaps someone can think of a good incentive...
(get some new and improved methods only if you use the new class, or
something?)


From PClaerhout at CREO.BE  Thu Apr 13 04:29:34 2000
From: PClaerhout at CREO.BE (Pieter Claerhout)
Date: Thu, 13 Apr 2000 10:29:34 +0200
Subject: Problems with PythonWin
Message-ID: <2B1262E83448D211AE4B00A0C9D61B03BA72B0@MSGEURO1>

The exact error message that I get is:

Registration of the AXScript Engine COM server failed.
Installation will continue, but this server will require
manual registration before it will function.

pywintypes.com_error: (-2147467262, 'No such interface supported', None, None)

Does anyone know how to do this manual registration?

Kind regards,


Pieter Claerhout
Pieter_Claerhout at creoscitex.com 

-----Original Message-----
From: Pieter Claerhout 
Sent: Thursday, April 13, 2000 10:22 AM
To: python-list at python.org
Subject: Problems with PythonWin


Hello,

I tried to reinstall PythonWin build 128 today, but I got errors
when he was trying to register the Python AXScript engine.
Any idea what could be wrong? Any help would be appreciated.

Kind regards,


Pieter
Pieter_Claerhout at creoscitex.com

-- 
http://www.python.org/mailman/listinfo/python-list



From thomas at xs4all.net  Thu Apr  6 12:28:02 2000
From: thomas at xs4all.net (Thomas Wouters)
Date: Thu, 6 Apr 2000 18:28:02 +0200
Subject: __builtin__.open on windows NT
In-Reply-To: <8cicd6$1s0$1@nnrp1.deja.com>; from wright7@my-deja.com on Thu, Apr 06, 2000 at 04:01:54PM +0000
References: <8cicd6$1s0$1@nnrp1.deja.com>
Message-ID: <20000406182802.C13830@xs4all.nl>

On Thu, Apr 06, 2000 at 04:01:54PM +0000, Rob Wright wrote:

> i'm new to python, so be prepared...i'm having trouble with open() on
> windows NT.  my script uses the os library and it appears that open()
> in my script is conflicting with an open() function in the os lib.  so
> i tried to use __builtin__.open(), but i get this error "NameError:
> __builtin__"

Well, if you have that conflict, you are doing 'from os import *'.
The easy solution is 'dont do that then'. 'from <module> import *' can be
very bad for your health, for exactly the reason you specify. My advice (and
probably that of most people here ;) would be to change the import into just
'import os' and fix the places where you actually use os.

If you *really* want to use __builtins__, you can: you just have to import
it first ! :-) Remember to do that with 'import __builtins__', or it might
overwrite some of os's variables...

From-armageddon-import-*-ly y'rs,
-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From akira.kiyomiya at autodesk.com  Wed Apr 26 14:44:45 2000
From: akira.kiyomiya at autodesk.com (Akira Kiyomiya)
Date: Wed, 26 Apr 2000 11:44:45 -0700
Subject: Simple Function Question
Message-ID: <8eaah7$26@autodesk.autodesk.com>

Hi,  Could you explain this in dummy's language?

This function prints out '4' and I have a feeling that x=1 and n=3, so
add1(3) prints '4'.

However, I have a difficlut time trying to see how these two arguments
assigned to both x (especially x) and n.

Thanks

Akira


# Return a function that returns its argument incremented by 'n'
def make_incrementer(n):
    def increment(x, n=n):
        return x+n
    return increment

add1 = make_incrementer(1)
print add1(3)  # This prints '4'







From kragen at dnaco.net  Sat Apr 22 19:45:30 2000
From: kragen at dnaco.net (Kragen Sitaker)
Date: Sat, 22 Apr 2000 23:45:30 GMT
Subject: LAmbda calculus evaluator (interactive)
References: <8dro55$eq6$1@nnrp1.deja.com> <vndk8hq3t1z.fsf@camelot-new.ccs.neu.edu> <8dt0cq$nte$1@nnrp1.deja.com> <vnd8zy54vr8.fsf@camelot-new.ccs.neu.edu>
Message-ID: <uQqM4.461$RM6.88244@news-east.usenetserver.com>

In article <vnd8zy54vr8.fsf at camelot-new.ccs.neu.edu>,
Justin Sheehy  <dworkin at ccs.neu.edu> wrote:
>Andrew Cooke <andrew at andrewcooke.free-online.co.uk> writes:
>> ...but if not, i thought the (unquoted) context made clear that i was
>> talking about programs that manipulated expressions in lambda calculus
>> (so a "not interactive" program simply takes an expression and follows
>> some evaluation strategy as far as possible; an "interactive"
>> program allows me to step reduction by reduction, do abstractions, etc).
>> in other words, manipulate the appropriate ast.
>
>Ah.  Well, that's not at all what most people that I know mean by the
>distinction between "interactive" and "non-interactive" when referring to
>a language implementation.
>
>If I refer to an interpreter (or compiler, whatever) as "interactive",
>I mean simply that it has an read-eval-print loop that one can
>interact with.
>
>With regard to the lambda calculus debugger you seem to be looking
>for, I'm not sure exactly what you'd want it to look like.  Even if I
>did, I don't know if I'd be able to direct you to one.

You can think of the lambda-calculus as a programming language, but it
might be more accurate to think of it as a set of formulas, a set of
ways of modifying those formulas so that they remain equivalent in some
sense, and a set of ways of combining formulas to make new (not
necessarily equivalent) formulas.

I'd like to play with the thing Andrew is proposing too.  Many years
ago, Colby Kraybill suggested to me that it would be nice to have
something similar for mathematical expressions --- something that would
do nothing on its own, but would do the grunt-work involved in an
action like "take the reciprocal of both sides" or "divide both the top
and bottom by (x-1)".

>Many other languages (C++, Perl, many others) typically do not have an
>interactive interpreter.  One can refer to their implementations as
>non-interactive.

perl -d  :)


-- 
<kragen at pobox.com>       Kragen Sitaker     <http://www.pobox.com/~kragen/>
The Internet stock bubble didn't burst on 1999-11-08.  Hurrah!
<URL:http://www.pobox.com/~kragen/bubble.html>
The power didn't go out on 2000-01-01 either.  :)


From gmcm at hypernet.com  Wed Apr  5 10:41:21 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Wed, 5 Apr 2000 10:41:21 -0400
Subject: Class type coercion
In-Reply-To: <newscache$j8mjsf$cqh$1@jenpc07.jennic.co.uk>
Message-ID: <1257176811-9948833@hypernet.com>

Robert Cragie

> (Apologies if this has been covered before, I'm a bit of a Python newbie)
> 
> I posted a question to wxWindows list, and got an answer from Robin Dunn,
> but it raised a more general question...
> 
> In C++, it is easy to use a pointer to a object declared as a super class of
> that object's class, and then coerce it to be the same class type as the
> object, e.g.
> 
> // Declare instance of class 'MyClass', which is derived from class
> 'MySuperClass'
> MyClass MyClassObj;
> 
> // Declare pointer to class 'MySuperClass'
> MySuperClass* pMySuperClass;
> 
> pMySuperClass = &MyClassObj; // OK, class 'MyClass' is derived from class
> 'MySuperClass'
> 
> // Coerce pointer to get to method which is only in class 'MyClass'
> ((MyClass*)pMySuperClass)->MyMethodOnlyInMyClass();
> 
> Can you do the equivalent in Python? I feel there must be some
> straightforward way to do this, I just haven't figured it out yet. Or is
> there...? As there is no type declaration in Python maybe this just isn't
> possible.

Python is completely dynamically bound, so the question 
doesn't arise.
a.MyMethodOnlyInMyClass() 
will either succeed or fail.

To find out if it will work, read up on isinstance().
 
> There is a real-world need for this e.g. parent Window object contains
> various child control objects. All the control classes are subclassed from
> the Window class. All control objects have an ID. There often exists a
> Window method to get a child window by ID. However, it can only return a
> type of Window class. Therefore, to get to the actual control, this return
> type needs to be subclassed.



- Gordon



From jvickroy at sec.noaa.gov  Tue Apr 25 13:53:42 2000
From: jvickroy at sec.noaa.gov (j vickroy)
Date: Tue, 25 Apr 2000 11:53:42 -0600
Subject: passing Python list to C extension function
Message-ID: <3905DBA6.60AA7A48@sec.noaa.gov>

Hello all,

I am trying to pass a Python list to a C extension module function and
retrieve a result.  Using Python 1.5.2 on Win NT 4.0, the C function
generates a memory error when trying to parse () the Python list
parameter.

Any assistence would be appreciated.

Thanks for your time,
    - jv

Here is the Python code:

import c_extender
list = []
for i in xrange (10):
    list .append ( float (i) )
 print 'total:', c_extender .total_it (list)


... and here is the C function.

#include "Python.h" // Python-C interface definitions


static PyObject *c_extender_total_it (parameter)
    PyObject *parameter;
     /*
     */
    {
    int i;
    float result;
    double *array;
    double c_double;
    PyObject *python_list;

    printf ("\t <executes this statement> \n");
    // interpret parameter as a Python list
        if (! PyArg_ParseTuple (parameter, "O!", &PyList_Type,
&python_list) ) return NULL;
    // *** memory exception reported here ***
    printf ("\t <does NOT execute this statement> \n");

    // perform calculation here

    free (array);
    return Py_BuildValue ("total", result);
    }




From arta at x-stream.nl  Mon Apr 17 07:21:43 2000
From: arta at x-stream.nl (A[r]TA)
Date: Mon, 17 Apr 2000 13:21:43 +0200
Subject: input in loop
Message-ID: <etNSv#Jq$GA.202@net025s>

Heyz, I want to write a loop
in which there's something like raw_input but which
doesn't stop the loop until the enterbutton is hitten.
How do I do that???

A[r]TA


--
We shall not cease from exploration, and the end of all our exploring will
be to arrive where we started and know the place for the first time.
- T. S. Eliot





From nmm1 at cus.cam.ac.uk  Fri Apr  7 07:35:09 2000
From: nmm1 at cus.cam.ac.uk (Nick Maclaren)
Date: 7 Apr 2000 11:35:09 GMT
Subject: Was Re: need wrapper for ssh/scp
References: <8civ14$rm7$1@hammer.msfc.nasa.gov> <20000407130942.E13830@xs4all.nl>
Message-ID: <8ckh5d$bj2$1@pegasus.csx.cam.ac.uk>

In article <20000407130942.E13830 at xs4all.nl>, Thomas Wouters <thomas at xs4all.net> writes:
|> On Thu, Apr 06, 2000 at 09:19:32PM +0000, Andrew McDowell wrote:
|> 
|> > I also working with ssh to execute some remote commands on different
|> > servers.  I'd previously been hobbling together this strange idea of
|> > generating an Expect script in python then running it. (*bleh*)  So once I
|> > saw was clued onto the pty module I started playing with it.  Well...I'm
|> > either severely confused, or something isn't working properly.  Could anyone
|> > offer some help?
|> 
|> Help is easiest if you also explain what the problem is ;) I already see one
|> problem though: you read the password with 'string.strip(raw_input())',
|> which means the terminating newline gets stripped off the password, but you
|> dont supply it with os.write(), which means ssh thinks there is more
|> password to come, and patiently sits waiting for it, untill you brutally
|> kill it ;)
|> 
|> In your real app you probably want to use select(), by the way, before
|> attempting to read something. Doing a read() when there's no data to be read
|> makes a blocking read, which can lead to deadlocks. See telnetlib for some
|> examples on how to use it.

If you are trying to write a reliable application, be aware that this
area is a minefield.  select is not in POSIX, and therefore has no
specification - it does subtly different things on different Unices.
The same is regrettably true for the detailed semantics of read and
write, such as what happens on failures and how (non)blocking modes
work on network connexions (such as sockets.)


Regards,
Nick Maclaren,
University of Cambridge Computing Service,
New Museums Site, Pembroke Street, Cambridge CB2 3QG, England.
Email:  nmm1 at cam.ac.uk
Tel.:  +44 1223 334761    Fax:  +44 1223 334679


From rcunning at acm.org  Wed Apr  5 22:15:44 2000
From: rcunning at acm.org (Robert W. Cunningham)
Date: Thu, 06 Apr 2000 02:15:44 GMT
Subject: Pass by reference ?
References: <1257157531-11108477@hypernet.com> <38EBCD7D.7030602@acm.org> <vnd66twdmfo.fsf@camelot-new.ccs.neu.edu>
Message-ID: <38EBF2AC.B472F267@acm.org>

Justin Sheehy wrote:

> The line:
>
>   w=None
>
> simply makes the local variable `w' no longer be bound to the argument
> passed, but rather to None.  Reassigning the name `w' is simply that:
> reassigning a name.  Whatever was previously referred to by that name
> is irrelevant.

Doh!  That's what I get for trying to play with a new language at
midnight.  I *knew* the effect of assigning None.  Maybe I should have used
"del" instead?

Evidently I'm at the stage where I know enough to be truly dangerous.  I'll
take a moment to become a bit more fluent and have another go at this
later.  Of all the examples that could have dropped out of my mind, that's
the one I had to type in.  <sigh>

I have better examples at hand, but I wanted to write my own.  Hoping, of
course, that I'd learn something in the process.  As the Chinese would
say:  "Be careful what you ask for..."

-BobC




From alan.gauld at gssec.bt.co.uk  Fri Apr  7 10:54:23 2000
From: alan.gauld at gssec.bt.co.uk (Alan Gauld)
Date: Fri, 07 Apr 2000 15:54:23 +0100
Subject: CGI Help Needed
References: <8ck9uo$ga08@imsp212.netvigator.com>
Message-ID: <38EDF69F.8FD477DC@gssec.bt.co.uk>

Paul Gresham wrote:
> What is the normal and respected method of User authentication on a web
> site, how should I present a login box when the user is not authenticated

Normally this is a function of the web server and you can specify
directories or modules which require authentication.

> without using cookies, allowing them to continue surfing 

Tricky, the 'normal way' in my experience is to use cookies! 
Usually set to time out after the session ends.

Alan G.

-- 
=================================================
This post represents the views of the author 
and does not necessarily accurately represent 
the views of BT.


From dalke at acm.org  Mon Apr 24 05:29:29 2000
From: dalke at acm.org (Andrew Dalke)
Date: Mon, 24 Apr 2000 03:29:29 -0600
Subject: The Simple Economics of Open Source
References: <B527E7A5.51AA%raffael@mediaone.net> <MPG.136cc01af4ab703b989709@news.earthlink.net> <raffael-4899DC.23562823042000@news.ne.mediaone.net>
Message-ID: <8e140f$cco$1@slb0.atl.mindspring.net>

Raffael Cavallaro wrote:
>Fact is, only commodity items, that are known to many programmers, are
>open sourced. If there were sufficient demand to support a closed source
>business, then it would have happened that way. The proof of this is the
>BSDs, which would allow a closed source, proprietary fork at any time.
>Hasn't happened 'cause there's no demand for such a product.


Not quite fact.  There are counter-examples.  What's the major
competitor to TeX?  From my understanding, when it was first released
there was nothing like it, proprietary or otherwise, so it wasn't
a commodity.

The first few web servers and browsers weren't commodities, but they
were open source.

But I do agree that most open-source projects seem to be years behind
their closed-source equivalents.

                    Andrew Dalke
                    dalke at acm.org





From ajung at sz-sb.de  Mon Apr  3 07:17:20 2000
From: ajung at sz-sb.de (Andreas Jung)
Date: Mon, 3 Apr 2000 13:17:20 +0200
Subject: New Features in Python 1.6
In-Reply-To: <200004011740.MAA04675@eric.cnri.reston.va.us>; from guido@python.org on Sat, Apr 01, 2000 at 12:00:00PM -0500
References: <200004011740.MAA04675@eric.cnri.reston.va.us>
Message-ID: <20000403131720.A10313@sz-sb.de>

On Sat, Apr 01, 2000 at 12:00:00PM -0500, Guido van Rossum wrote:
> 
> Python strings can now be stored as Unicode strings.  To make it easier
> to type Unicode strings, the single-quote character defaults to creating
> a Unicode string, while the double-quote character defaults to ASCII
> strings.  If you need to create a Unicode string with double quotes,
> just preface it with the letter "u"; likewise, an ASCII string can be
> created by prefacing single quotes with the letter "a".  For example:
> 
>    foo = 'hello'                # Unicode
>    foo = "hello"                # ASCII

Is single-quoting for creating unicode clever ? I think there might be a problem
with old code when the operations on unicode strings are not 100% compatible to
the standard string operations. I don't know if this is a real problem - it's
just a point for discussion.

Cheers,
Andreas 



From nmm1 at cus.cam.ac.uk  Wed Apr 12 13:27:11 2000
From: nmm1 at cus.cam.ac.uk (Nick Maclaren)
Date: 12 Apr 2000 17:27:11 GMT
Subject: Why should I switch to Python?
References: <seas5ia89177@corp.supernews.com> <slrn8f4d0j.1t0.bwinton@tor.dhs.org> <8d2aa8$n58$1@newshost.accu.uu.nl>
Message-ID: <8d2blf$b1n$1@pegasus.csx.cam.ac.uk>

In article <8d2aa8$n58$1 at newshost.accu.uu.nl>, m.faassen at vet.uu.nl (Martijn Faassen) writes:
|> Blake Winton <bwinton at tor.dhs.org> wrote:
|> > On Fri, 31 Mar 2000 19:43:58 -0800, Aaron Turner wrote:
|> >>I've been coding Perl for over 3 years and really like it.  (Especially
|> >>it's syntax and ability to grok my programming style.)
|> 
|> > Now, just as a side note, why is it that the sets of The Perl Hackers I
|> > Know, and The People I Know Who Use The Word "Grok" are the same set?
|> 
|> It's in the hacker's dictionary:
|> 
|> http://www.tuxedo.org/~esr/jargon/html/entry/grok.html

According to something I read, incorrectly attributed!  Apparently
Heinlein got it from something he read, and forgot the fact - it
was then pointed out to him and he said something equivalent to
"Oops".  That doesn't stop the attribution to him circulating :-)

The same thing has happened with the expression "To damn with
faint praise", incidentally - it is attributed to Pope, but is
actually Wycherley.


Regards,
Nick Maclaren,
University of Cambridge Computing Service,
New Museums Site, Pembroke Street, Cambridge CB2 3QG, England.
Email:  nmm1 at cam.ac.uk
Tel.:  +44 1223 334761    Fax:  +44 1223 334679


From breiter at usf.Uni-Osnabrueck.DE  Tue Apr 11 05:27:44 2000
From: breiter at usf.Uni-Osnabrueck.DE (Bernhard Reiter)
Date: 11 Apr 2000 09:27:44 GMT
Subject: Python/Perl Popularity (Re: A Mountain of Perl...)
References: <200004102244.IAA03513@envy.fulcrum.com.au>
Message-ID: <8cur6g$snq$1@newsserver.rrzn.uni-hannover.de>

In article <200004102244.IAA03513 at envy.fulcrum.com.au>,
	Richard.Jones at fulcrum.com.au (Richard Jones) writes:

>        Richard who is tired of the endless "vs" crap in the newsgroup :)

But these are among the most interesting questions.
If the python community can point at very good FAQs or writeups of
the arguments, I think that traffic and high temperature threads
are less likly to occur.

Well *head scratch* the last two postings in this thread might
as well increase chances.

not-willing-to-invent-interesting-timbot-like-style-closings-ly, yours
	Bernhard
-- 
Professional Service around Free Software                (intevation.net)  
The FreeGIS Project				            (freegis.org)
Association for a Free Informational Infrastructure            (ffii.org)


From richard_chamberlain at ntlworld.com  Thu Apr 20 03:12:40 2000
From: richard_chamberlain at ntlworld.com (Richard Chamberlain)
Date: Thu, 20 Apr 2000 08:12:40 +0100
Subject: How do I get this working?
References: <8dm932$q13$1@porthos.nl.uu.net>
Message-ID: <8dmand$7hj$1@gxsn.com>

You were very close...

import uerrors
import sys

    def CheckParams():
                if (len(sys.argv)>5) or (len(sys.argv<2):
                            return uerrors.ERR_WRONGPARAMS
                return uerrors.ERR_NOERROR

You had an 'and' in your version, sys.argv is never going to be above 5 and
below 2 therefore it was just returning ERR_NOERROR which was 0.

Richard

"Martijn van der Kooij" <DeleteThisDev at stabiplan.nl> wrote in message
news:8dm932$q13$1 at porthos.nl.uu.net...
> Hi,
>
> I'm very new to Python and just started to port a application to Python,
but
> i have a problem.
>
> I've declared a unit with some constants this way:
>
> ERR_NOERROR = 0
> ERR_WRONGPARAMS = 1
> ERR_NOREPORTFILE = 2
> ERR_NOOPTIONSFILE = 3
> ERR_NOMAINPLAYERFOUND = 4
> ERR_PARSINGREPORT = 5
> ERR_READINGREPORTFILE = 6
>
> this is imported into another module:
>
> import uerrors
> import Sys
>
> def CheckParams():
>     if (len(sys.argv) > 5) and (len(sys.argv) < 2):
>         return uerrors.ERR_WRONGPARAMS
>     return uerrors.ERR_NOERROR
>
> def DomainBot():
>     res = CheckParams()
>     if ret:
>         return res
>     return uerrors.ERR_NOERROR
>
> which is used in the main program.
>
> The problem is that it looks like that res = CheckParams() returns the
> function instead of one of the error codes.
>
> Probably i'm docing something terribly wrong here. But can someone help me
> getting this working?
>
> Martijn
>
>
>
>




From mwh21 at cam.ac.uk  Mon Apr  3 06:28:31 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 03 Apr 2000 11:28:31 +0100
Subject: Python 1.6 alpha 1 released
References: <200004010159.LAA00403@piglet.dstc.edu.au> <8c9nq3$ghk$4@newsserver.rrzn.uni-hannover.de> <m3bt3rcweu.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <m38zyvcwc0.fsf@atrus.jesus.cam.ac.uk>

Michael Hudson <mwh21 at cam.ac.uk> writes:

> And, to answer you're question, about as
> complicated as:
> 
> http://x42.deja.com/[ST_rn=ps]/getdoc.xp?AN=591243549
> 
> after `sed -e s/append/connect/', which doesn't get the bound method
> case, but I am having trouble imagining circumstances in which you'd
> reasonably do that.

Oops, horrible run-on sentence.  Sorry.

M.

-- 
well, take it from an old hand: the only reason it would be easier
to program in C is that you can't easily express complex  problems
in C, so you don't.                 -- Erik Naggum, comp.lang.lisp


From brennanw at pond.com  Sun Apr 30 13:43:12 2000
From: brennanw at pond.com (William C. Brennan)
Date: Sun, 30 Apr 2000 17:43:12 GMT
Subject: problem building Python in MachTen Unix
Message-ID: <01HW.B531E90F0013E3A1037B464C@news.pond.com>

Please, someone save me from Perl!

I'm trying to build Python from sources to work under Tenon's CodeBuilder
Unix (MachTen Lite), which is a BSD variety which runs under the MacOS.  
(Yes, Unix _under_ the MacOS).  I want to use Python because I'm
frustrated by my inability to understand even my own Perl scripts, let alone
other folks'.  Python looks like it has everything I want.

I've downloaded and unpacked the Python source, and run the configure
script without error.  Among the many lines of information, it displays:

   checking for --with-next-framework... no
   checking for --with-dyld... required for framework build

When I proceed to performing the "make", I get an error while making the
object files in tho Python sub-directory:

   gcc -g -O2 -I./../Include -I.. -DHAVE_CONFIG_H   -c getversion.c -o getversion.o
   gcc -g -O2 -I./../Include -I.. -DHAVE_CONFIG_H   -c graminit.c -o graminit.o
   gcc -g -O2 -I./../Include -I.. -DHAVE_CONFIG_H   -c import.c -o import.o
   gcc -c -g -O2 -I./../Include -I.. -DHAVE_CONFIG_H -I/ ./importdl.c
   ./importdl.c:269: mach-o/dyld.h: No such file or directory
   make[1]: *** [importdl.o] Error 1
   make[1]: Leaving directory `/my/Python-1.5.2/Python'

which implies to me that the configure script thinks my "with-dyld" or 
"with-next-framework" settings are set.  It looks like I need to tell the 
configure script that I'm running without frameworks and without dyld, 
but my attempts to communicate this to the configure script have failed.  
I've tried:

  ./configure --without-next-framework --without-dyld

which has worked even less well.  I've also tried manually editing the configure 
script (which is almost totally unintelligible to me) by changing the defaults
at the top from:

   ac_help="$ac_help
   --with-next-framework     Build (OpenStep|Rhapsody|MacOS10) framework"
   ac_help="$ac_help
   --with-dyld               Use (OpenStep|Rhapsody|MacOS10) dynamic linker"
to
   ac_help="$ac_help
   --without-next-framework     Build (OpenStep|Rhapsody|MacOS10) framework"
   ac_help="$ac_help
   --without-dyld               Use (OpenStep|Rhapsody|MacOS10) dynamic linker"

which hasn't worked either.  I've looked through the configure script to
see if I could figure out any other flags which may indirectly cause
"with-next-framwork" or "with-dyld" to be set, but I can't find anything.

Could someone suggest something to this naive Python builder to get his
"make" script to work?

Thanks!

-- bill


-- 
Bill Brennan
brennanw at pond.com



From moshez at math.huji.ac.il  Mon Apr  3 18:21:21 2000
From: moshez at math.huji.ac.il (Moshe Zadka)
Date: Tue, 4 Apr 2000 00:21:21 +0200 (IST)
Subject: Help on Big Python Program
In-Reply-To: <38E8DB50.A79CB88@nmt.edu>
Message-ID: <Pine.GSO.4.10.10004040020330.21084-100000@sundial>

On Mon, 3 Apr 2000, Enrique Herrera wrote:

> 	I have been working on a python program that is getting big. When its
> all finished it will have several modules that have been written for it. 
> 
> 	I was wandering if there is a way to "compile" the python source into
> one executable program?

If you just need some way to distribute the thing, go for Gordon
McMillan's installer (found at his starship page)
--
Moshe Zadka <mzadka at geocities.com>. 
http://www.oreilly.com/news/prescod_0300.html
http://www.linux.org.il -- we put the penguin in .com




From sblakey at freei.com  Wed Apr 12 13:56:31 2000
From: sblakey at freei.com (Sean Blakey)
Date: Wed, 12 Apr 2000 10:56:31 -0700 (PDT)
Subject: global variables - how??
In-Reply-To: <20000412134115.A9193@better.net>
Message-ID: <20000412175218.784841CE49@dinsdale.python.org>

On 12 Apr, William Park wrote:
> On Wed, Apr 12, 2000 at 07:28:37PM +0200, Anders Eggers-Krag wrote:
>> how do I define a propper global variable that works in all functions in all
>> modules?
> 
> As far as I know, you can't.  You have to use 'global ...' inside all
> your functions.
> 
> --William
> 
>> 
>> I have tried simply
>> defining one outside of any function in the main module,
>> I have tried using the global keyword under the same cicumstances,
>> but they are not even avialible inside of functions within the main
>> module...
>> 
>> 
>> --
>> Anders Eggers - Krag
>> 
>> 
>> -- 
>> http://www.python.org/mailman/listinfo/python-list
> 

Short answer: Python discourages globals.  This is probably intentional.

Long answer:  You can do some work with globals, but it is painful and
ugly.  You also need to keep track of namespaces (globals from your
main module will be in the __main__ namespace).

For example:
-----file one.py-----
TEST_VALUE = 'spam'

def printTest():
    print TEST_VALUE

def changeTest():
    global TEST_VALUE			# Without this, the next line
    					# will simply assign 'spamspam'
    					# to the local variable
    					# TEST_VALUE
    TEST_VALUE = TEST_VALUE + TEST_VALUE

if __name__ == '__main__':
    import two
    printTest()
    two.printTest()
    changeTest()
    printTest()
    two.printTest()
    
-----file two.py-----
import __main__

def printTest():
    print __main__.TEST_VALUE


Now, from the shell:
seanb at seanb:~$ python one.py
spam
spam
spamspam
spamspam


-- 

Sean Blakey
FreeInternet.com
sblakey at freei.com
(253)796-6500x1025




From aek at mail1.stofanet.dk  Tue Apr  4 18:58:35 2000
From: aek at mail1.stofanet.dk (Anders Eggers-Krag)
Date: Wed, 5 Apr 2000 00:58:35 +0200
Subject: Tkinter, <select> style menu
References: <8cdqom$aqp$1@news101.telia.com> <38EA636C.EB62A9BF@exceptionalminds.com>
Message-ID: <00d501bf9e89$56f22080$2c6ba8c0@frodo>

----- Original Message ----- 
From: "Timothy Grant" <tjg at avalongroup.net>
To: "Anders Eggers-Krag" <aek at mail1.stofanet.dk>
Cc: <python-list at python.org>
Sent: Tuesday, April 04, 2000 11:49 PM
Subject: Re: Tkinter, <select> style menu


> Anders Eggers-Krag wrote:
> > 
> > Is there really no reasoanble way to make a box similar to that of the HTML
> > <SELECT>
> > in tkinter?
> > 
> > the nearest I have been able to find is from Pmw, and it doesn't really look
> > like windows...
> 
> Hmm, I thought that the Pmw.ComboBox() when used in drop down mode
> looked almost just like the Windows Drop Down box.
> 
> What is your problem with it?
well take a look at this picture and tell me if you think they look alike
http://www.aek.dk/python.gif

left is windows in case there is any doubt :-)

my personal oppinion is the difference is too great for my user...

Anders Eggers - Krag





From offer at sgi.com  Mon Apr 24 10:40:25 2000
From: offer at sgi.com (Nhi Vanye)
Date: 24 Apr 2000 14:40:25 GMT
Subject: Using python on the web
References: <EDFD2A95EE7DD31187350090279C6767754BA8@THRESHER> <FtEE3B.Ar2.0.bloor@torfree.net> <tShM4.42074$h6.182368@typhoon.jacksonville.mediaone.net>
Message-ID: <8e1mcp$3bhdu$1@fido.engr.sgi.com>

$ from hank at prosysplus.com -#87740 | sed "1,$s/^/> /"
>
>
>Hi Ken,
>
>An article on Python Server Pages appeared in two parts recently in DDJ.
>The web site associated with the article is at
>http://www.ciobriefings.com/psp/.

I was looking forward to being able to use PSP instead of PHP in my web
pages, but:

  o Its written in JPython, and when what I really need is mod_psp. 
  o The license requires me to submit all changes I make back to AEI.
  o The license requires that I state in a prominent way that I have
    made modifications to the software and the specific changes made
  o I'm not allowed to make any changes that may be "harmful" to other
    users/systems. So no site specific customising.

I'll be staying with PHP for the time being :-(

>
>Hank Fay
>


richard.
-- 
Richard Offer           Widget FAQ --> http://reality.sgi.com/widgetFAQ
MTS-Core Design (Motif)
___________________________________________http://reality.sgi.com/offer



From chambert at borel2.institut.math.jussieu.fr  Fri Apr 21 16:36:51 2000
From: chambert at borel2.institut.math.jussieu.fr (Antoine Chambert-Loir)
Date: 21 Apr 2000 22:36:51 +0200
Subject: Mac Python Questions
References: <IRjL4.271$CP3.3183@news.uswest.net> <38FE8032.DE233C29@cosc.canterbury.ac.nz> <956247497.216189@fezzik.endicor.com>
Message-ID: <8dqe53$2i6$1@borel2.institut.math.jussieu.fr>

In article <956247497.216189 at fezzik.endicor.com>,
Kent Polk <kent at goatnospamhill.org> wrote:
>The standard Python installation program installed the wrong version
>of the _tkinter shared library and(/or) didn't set the path to the
>correct (ppc) version, so Python couldn't load _tkinter. If that
>is the problem, there are several ways to solve it.

I would be very happy to see tkinter at work on a mac.
So can you be more precise ?

Regards,

Antoine Chambert-Loir

--
Institut de Math?matiques de Jussieu ? Chevaleret
Universit? Pierre et Marie Curie
Boite 247, 4 place Jussieu, F-75252 Paris Cedex 05
Tel: 01 44 27 75 20       Fax: 01 44 27 48 44
http://www.math.jussieu.fr/~chambert
-- 
Antoine Chambert-Loir  o-o  Institut de Math?matiques de Jussieu (Chevaleret)
Tel: 01 44 27 75 20     |   Universit? Pierre et Marie Curie
Fax: 01 44 27 48 44     -   Boite 247, 4 place Jussieu, F-75252 Paris Cedex 05
http://www.math.jussieu.fr/~chambert


From gregm at iname.com  Fri Apr  7 08:19:26 2000
From: gregm at iname.com (Greg McFarlane)
Date: Fri, 7 Apr 2000 22:19:26 +1000
Subject: Widget for plotting graphs?
In-Reply-To: <38EC23E6.9975FD7D@pacbell.net>; from Bruce Wolk on 05 Apr 2000 at 10:43:02PM
References: <38EC23E6.9975FD7D@pacbell.net>
Message-ID: <20000407221926.57509@nms.otc.telstra.com.au>

Have you looked at the Blt graph widget?  It is a highly configurable
plotting library for 2D graphs.  There is a python interface to it in
the Pmw package (http://www.dscpl.com.au/pmw/).  There is some
recently released documentation of the Pmw Blt interface at
http://www.ifi.uio.no/~hpl/Pmw.Blt/doc/ which is worth looking at. 
There are lots of examples, demos and pictures so you can see what it
is before downloading anything.

On 5 Apr, Bruce Wolk wrote:
> I am exploring Tkinter and Pmw and wondering whether anyone has created
> a widget for graphing points and functions.  Also, is there a way to
> place one pixel (i.e., a point) on a Canvas widget?  Thanks.
> 
> Bruce
> -- 
> http://www.python.org/mailman/listinfo/python-list
> 

-- 
Greg McFarlane     INMS Telstra Australia     gregm at iname.com



From boud at rempt.xs4all.nl  Mon Apr 17 16:54:17 2000
From: boud at rempt.xs4all.nl (Boudewijn Rempt)
Date: 17 Apr 2000 20:54:17 GMT
Subject: Scripting and Gnome and KDE
References: <Pine.GSO.4.10.10004150741520.11261-100000@sundial> <8d95m0$9j6$1@news1.xs4all.nl> <8damat$4ja$1@slb2.atl.mindspring.net> <8dbsgc$35o$1@news1.xs4all.nl> <38FB4336.2BC6C235@bioreason.com>
Message-ID: <8dftlp$3s1$1@news1.xs4all.nl>

Mitch Chapman <chapman at bioreason.com> wrote:

> In fact, I'd considered trying to write a similar article about PyQt
> and PyKDE.  But that article would have been even bleaker, because
> the process of discovery w. PyQt is still difficult for me.  I've
> had a few recurring problems.  Maybe this would be a good place to
> ask for enlightenment :)  I'll start w. the two biggies:
> 	o The Qt C++ interfaces include many overloaded functions.
> 	  Is there a consistent pattern for determining which of
> 	  these is implemented by the Python binding?  Or must
> 	  you look at the source code for the Python bindings?
> 	o In many cases, if I make a wrong guess as to which 
> 	  overloaded method is implemented by the binding, I get
> 	  a segmentation fault or (slightly better) a traceback 
> 	  indicating an incorrect argument type.  Is there an easy 
> 	  way to work backwards from a segmentation fault to
> 	  the correct usage for a PyQT widget?

Well, I've been working with PyQt and PyKDE for about year now, and I must
say I've never had that problem. Most overloaded functions differ in the
number of arguments, making it simple to determine which one is chosen;
as for the functions where the type of arguments is the only difference,
the bindings check the type of the arguments and choose the right
function:

The docs:

The QListViewItem class implements a list view item. More... 

Public Members

  * QListViewItem ( QListView * parent, const char *, const char * = 0,
const char * = 0, const char * = 0, const char * = 0, const char * = 0,
const char * = 0, const char * = 0 )  
  * QListViewItem ( QListViewItem * parent, const char *, const char * = 0,
const char * = 0, const char * = 0, const char * = 0, const char * = 0,
const char * = 0, const char * = 0 )  

Python code:
  listview=QListView(parent...)
  item1=QListViewItem(listview, "aaa")
	item2=QListViewItem(item1, "bbb")

Complete working example at:

  http://www.valdyas.org/python/overload.py

So the overloading problem doesn't exist at all ;-). No wonder I
hadn't noticed it. I wonder, how does wxPython handle this?

I haven't had many segfaults - mostly when trying out new features
in beta releases. 

>> Absolutely, but I don't know that an essay on the problems is exactly
>> the right way of introducing the toolkit.

> That's probably true.  Given the goal of showing how to do non-obvious
> things with a toolkit, what changes would you recommend?

Well, I'm not sure I saw any really non-obvious things - a treeview
isn't that spectacular nowadays, and neither is layout management.
(Although I must admit that I'd never thought that I would have to 
manually add scrolling to a treeview - that was very non-obvious to
me.)

> If I understand your comments, it would have been better to simply 
> say, "Here's how you do X," rather than to explain "Here's *why* you 
> do it this way".  I have one reservation with such an approach:

No, not exactly: what I look for in an introductory article is the
advantages of an approach. For instance, 'this toolkit has a really
easy and unique canvas', or 'writing a webbrowser in this toolkit takes
just 100 lines with toolbar and all', or 'this toolkit supports unicode
through-and-through' - showing off the really unique and interesting
things.

> it tells the reader how to solve a specific problem, but
> doesn't provide a general method for discovering how to use other
> parts of the toolkit.

I agree that that's useful. I spend a lot of time helping people
starting out with PyQt and PyKDE, showing them how to translate
the C++ documentation to Python. It's not difficult, just a mental
search-and-replace, but it's something one has to learn.

Besides, toolkits are different enough that migrating from one to
another often demands a complete new way of thinking: that's something
the article in Dr Dobbs did show: tkInter is one way of working,
PyQt another, but PyGTK is completely different once again. That's
an important insight, I feel.

-- 

Boudewijn Rempt  | http://www.valdyas.org


From dworkin at ccs.neu.edu  Wed Apr  5 21:53:47 2000
From: dworkin at ccs.neu.edu (Justin Sheehy)
Date: 05 Apr 2000 21:53:47 -0400
Subject: Pass by reference ?
References: <1257157531-11108477@hypernet.com> <38EBCD7D.7030602@acm.org>
Message-ID: <vnd66twdmfo.fsf@camelot-new.ccs.neu.edu>

"Robert W. Cunningham" <rcunning at acm.org> writes:

[snip long explanation of dubious relevance to the following code]

> Let's pick some representations:
> 
>    a=5
>    b=(5)
>    c=[5]
>    d={"five":5}
> 
> And let's create some functions to manipulate them.  In this case I 
> chose an extremely generic Python function, though I SHOULD choose a 
> separate function for each representation to illustrate that the 
> function doesn't matter:  The passing does.
> 
>    def Nuke(w):
>      print w,
>      w=None
>      print w
> 
> Let's call each of these functions:
> 
>    Nuke(a)
>    Nuke(b)
>    Nuke(c)
>    Nuke(d)
> 
> Look at the output:
> 
>    5 None
>    5 None
>    [5] None
>    {"five":5} None
> 
> And then look at the original values:
> 
>    print a,b,c,d
> 
>    5 5 [5] {"five":5}
> 
> None changed!  Therefore, representations the functions modified were 
> DIFFERENT from those passed, though it is clear that the correct VALUE 
> was passed.

No, unless by VALUE you mean the value of the reference that was passed.

Your assertion here is not really related to "pass by reference"
vs "pass by value".  All that your example does is demonstrate the
semantics of assignment in Python.  The two are certainly related
topics, but you have not shown anything interesting about argument
passing here.

The line:

  w=None  

simply makes the local variable `w' no longer be bound to the argument 
passed, but rather to None.  Reassigning the name `w' is simply that:
reassigning a name.  Whatever was previously referred to by that name
is irrelevant.

To show that Python does not do what you claim, I offer the following
slight change to your example:

>>> def NukeToo(w):
...   print w,
...   w[1] = None
...   print w
... 
>>> a = [1, 2, 3]
>>> NukeToo(a)
[1, 2, 3] [1, None, 3]
>>> a
[1, None, 3]

Does that still meet your old definition of pass-by-value?

-Justin

 



From warlock at eskimo.com  Tue Apr  4 02:28:13 2000
From: warlock at eskimo.com (Jim Richardson)
Date: Tue, 04 Apr 2000 06:28:13 GMT
Subject: Advice requested: GUI project beginning
References: <m12avmV-000wcEC@swing.co.at> <slrn8eiuu1.b3t.warlock@raven.myth>  <38E97E29.2687864D@acm.org>
Message-ID: <slrn8ej39h.h48.warlock@raven.myth>

On Tue, 04 Apr 2000 05:31:14 GMT, 
 Robert W. Cunningham, in the persona of <rcunning at acm.org>,
 brought forth the following words...:

>Jim Richardson wrote:
>
>> On Fri, 31 Mar 2000 09:23:47 +0200,
>>  Christian Tanzer, in the persona of <tanzer at swing.co.at>,
>>  brought forth the following words...:
>>
>> >
>> >"Russell E. Owen" <owen at astroNOJNK.washington.edu.invalid> wrote:
>> >
>> >> I am starting work on a telescope remote control user interface. It's a
>> >> big program with lots of windows and controls and display graphics
>> >> supporting the telescope and numerous instruments.
>>
>> Not discourage you from writing the app yourself, but there is a
>> linux Telescope control app at http://206.169.225.102/eots.html
>> certainly worth a look at least. I know nothing of the app itself, don't
>> know if it would fit the bill for you.
>
>Programming?  Telescope control?  Doesn't this thread belong in a Forth NG?
>
>I mean, wasn't Forth CREATED to control a telescope?
>

hm, I thought that was Algol?

-- 
Jim Richardson
	Anarchist, pagan and proud of it
WWW.eskimo.com/~warlock
	Linux, because life's too short for a buggy OS.



From andy at petdance.com  Wed Apr 26 13:29:29 2000
From: andy at petdance.com (Andy Lester)
Date: Wed, 26 Apr 2000 12:29:29 -0500 (CST)
Subject: <nifty hieroglyphics here>
In-Reply-To: <BWEN4.12235$HG1.329193@nnrp1.uunet.ca>
References: <8e4cc2$d5t@netnews.hinet.net> <BWEN4.12235$HG1.329193@nnrp1.uunet.ca>
Message-ID: <Pine.WNT.4.21.0004261229180.246-100000@nt_alester>

> The official language of usenet newsgroups is English,

Oh?  And where's THAT defined?

--
Andy Lester, andy at petdance.com, http://www.petdance.com
    "Inflammable material is planted in my head
     It's a suspect device that's left two thousand dead."
    "Hey, is this the new Green Day?"




From scarblac-spamtrap at pino.selwerd.nl  Wed Apr 26 04:41:00 2000
From: scarblac-spamtrap at pino.selwerd.nl (Remco Gerlich)
Date: 26 Apr 2000 08:41:00 GMT
Subject: execl problem on linux
References: <8e63b2$6v2$0@216.39.162.232>
Message-ID: <slrn8gdas4.bp5.scarblac-spamtrap@flits104-37.flits.rug.nl>

Michael Esveldt wrote in comp.lang.python:
> This may be bordering on off topic but the following is happening with 
> all the exec() methods...
> 
> Python 1.5.2 (#0, Apr  3 2000, 14:46:48)  [GCC 2.95.2 20000313 (Debian 
> GNU/Linux
> )] on linux2
> Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
> >>> import os
> >>> os.execl("/usr/bin/vim")
> Internal error: lalloc(0, )
> Internal error: lalloc(0, )
> bash-2.03$
> 
> I'm at a loss, any ideas?

'man execl' seems to say that it needs at least one argument string as well,
and indeed, os.execl("/usr/bin/vim","") works.

I don't know what's causing the error, maybe it's vim crashing on an undefined
argument list or something.

>>> import os
>>> os.execl("/usr/bin/emacs")
*** Error in XEmacs initialization
(wrong-type-argument stringp nil)
etc

Yes, that seems to be it :)

-- 
Remco Gerlich,  scarblac at pino.selwerd.nl
Hi! I'm a .sig virus! Join the fun and copy me into yours!


From tim_one at email.msn.com  Sun Apr 30 21:48:05 2000
From: tim_one at email.msn.com (Tim Peters)
Date: Sun, 30 Apr 2000 21:48:05 -0400
Subject: Discrepency between JPython and CPython
In-Reply-To: <m3k8hfjlex.fsf@boss.ksu.edu>
Message-ID: <000001bfb30f$4bebf720$94a2143f@tim>

[Venkatesh Prasad Ranganath]
> Consider the expression str(1.14 * 200).
> In CPython it yields 228.0
> In JPython it yields 227.99999999999997
>
> Shouldn't this result in consistent values across variuos
> implementation?

The values computed are identical; it's only the choice of display format
that differs.  Under 1.6a2 CPython,

>>> repr(1.14 * 200)
'227.99999999999997'
>>> str(1.14 * 200)
'228.0'
>>>

> If so, which of the implementation is buggy?
> I think it is JPython, may be I am wrong.

It's fuzzy, really.  Java rigidly defines what float->string conversion
does, and JPython appears to play along with Java's rules.  In 1.5.2
CPython, repr(float) and str(float) are identical, and don't follow any
comprehensible rules (they inherit random platform libc decisions).  In 1.6
CPython, str(float) is the same as in 1.5.2, but repr(float) produces enough
digits so that float(repr(x)) == x for all finite x on any IEEE-754
conformant platform (but not necessarily the *same* string across all 754
platforms).

If you're starting to get the feeling this is more involved than you think,
you're underestimating your confusion <wink>.  Over time, I expect the
JPython rules will gain favor, as they're the only ones on the table that
actually define the result precisely across platforms.

if-you-want-exactness-stick-to-integers-ly y'rs  - tim





From rupole at compaq.net  Fri Apr 21 09:51:14 2000
From: rupole at compaq.net (Roger Upole)
Date: Fri, 21 Apr 2000 13:51:14 GMT
Subject: 'Unexected Failure' from COM object
Message-ID: <m1ZL4.150916$ra.3345893@news.easynews.com>

Testing the win32 131 build with 1.6a2, I get this error:

>>> import win32com.client
>>> apssess = win32com.client.Dispatch('Awmonocx.Awmonocxctrl.1')
>>> apssess.ConnectToAPS('username','servername','passwd')
Traceback (innermost last):
  File "<interactive input>", line 1, in ?
  File "<COMObject Awmonocx.Awmonocxctrl.1>", line 2, in ConnectToAPS
com_error: (-2147418113, 'Unexpected failure', None, None)

note: With 1.52 and build 128, I occasionally get a 'catastrophic failure'
error msg
         from the same method, but this happens consistently with 131.
I've tried out several other COM objects, and this one seems to be the only
one
thats broken.
                   Roger Upole





From cfelling at iae.nl  Tue Apr 25 18:45:46 2000
From: cfelling at iae.nl (Carel Fellinger)
Date: 26 Apr 2000 00:45:46 +0200
Subject: Question about exception-handling mechanism
References: <8e46mt$4nh$1@nnrp1.deja.com> <eviltofu-A9A4DD.10052525042000@news.earthlink.net>
Message-ID: <8e576q$j65$1@vvs.superst.iae.nl>

Jerome Chan <eviltofu at rocketmail.com> wrote:

...
> Does this mean I can't do
> try:
>     x = MyClass("s")
>     # some other stuff that might raise an exception
> except Exception,message:
>     print message
> else:
>     print "something went wrong but I don't know what it is!"
> finally:
>     del x # or some other function

No, it's either a single finally clause or the except/else bit.
-- 
groetjes, carel


From gonter at zechine.wu-wien.ac.at  Thu Apr 27 12:51:11 2000
From: gonter at zechine.wu-wien.ac.at (Gerhard Gonter)
Date: Thu, 27 Apr 2000 18:51:11 +0200
Subject: Any Web-Mail Gateway (written in Python)?
References: <6D8A17398E28D3119F860090274DD7DB4986A3@pces.cadlab.it> <200004181250.IAA11781@sprawl.sis.yorku.ca>
Message-ID: <39086FFF.83358E00@zechine.wu-wien.ac.at>

Ed F. McCurdy wrote:
> You should look at mailman; http://www.list.org/

I think, the original poster had a web based e-mail system in mind,
something like hotmail, yahoo, etc.  There is in fact a web based
e-mail solution called "mailman" somewhere (I din't find the url
handy, but last time I looked, this wasn't something I would recommend
anyway).

Most web based mail solutions are written in PHP, however, there
is at least one such thing written in Python and runs in Zope:

http://demo.worldpilot.com:8080/site/

Hope that helps

+gg


From thomas.heller at ion-tof.com  Tue Apr 18 13:37:19 2000
From: thomas.heller at ion-tof.com (Thomas Heller)
Date: Tue, 18 Apr 2000 19:37:19 +0200
Subject: [Announce] autoreload.py
References: <001601bfa8a0$c6d86f50$4500a8c0@thomasnb> <9p4pfsch59ovo5atgl9l4usb91gl6qs5vs@4ax.com>
Message-ID: <010701bfa95c$bf119870$4500a8c0@thomasnb>

> My newsreader seriously munged this.  Can you repost the code, not as
> an attachment?
Sure. Probably my fault...

-------- cut here --------
#
# autoreload.py - automatically reload changed source
# code into a running program
#
# You may want to place the following two lines
# into your sitecustomize.py:
#
# import autoreload
# autoreload.run()
#
# Thomas Heller, 2000-04-17
#
# python-style license
#

import sys, imp, __builtin__, string
import time, os
import threading
import types


#####################################################################
#
# Most of this file is copied (and hacked) from standard module knee
#
# The purpose is the replace the __import__ function with
# a new one that keeps track of imported modules.
#
#####################################################################

# Replacement for __import__()
def import_hook(name, globals=None, locals=None, fromlist=None):
    parent = determine_parent(globals)
    q, tail = find_head_package(parent, name)
    m = load_tail(q, tail)
    if not fromlist:
        return q
    if hasattr(m, "__path__"):
        ensure_fromlist(m, fromlist)
    return m

def determine_parent(globals):
    if not globals or  not globals.has_key("__name__"):
        return None
    pname = globals['__name__']
    if globals.has_key("__path__"):
        parent = sys.modules[pname]
        assert globals is parent.__dict__
        return parent
    if '.' in pname:
        i = string.rfind(pname, '.')
        pname = pname[:i]
        parent = sys.modules[pname]
        assert parent.__name__ == pname
        return parent
    return None

def find_head_package(parent, name):
    if '.' in name:
        i = string.find(name, '.')
        head = name[:i]
        tail = name[i+1:]
    else:
        head = name
        tail = ""
    if parent:
        qname = "%s.%s" % (parent.__name__, head)
    else:
        qname = head
    q = import_module(head, qname, parent)
    if q: return q, tail
    if parent:
        qname = head
        parent = None
        q = import_module(head, qname, parent)
        if q: return q, tail
    raise ImportError, "No module named " + qname

def load_tail(q, tail):
    m = q
    while tail:
        i = string.find(tail, '.')
        if i < 0: i = len(tail)
        head, tail = tail[:i], tail[i+1:]
        mname = "%s.%s" % (m.__name__, head)
        m = import_module(head, mname, m)
        if not m:
            raise ImportError, "No module named " + mname
    return m

def ensure_fromlist(m, fromlist, recursive=0):
    for sub in fromlist:
        if sub == "*":
            if not recursive:
                try:
                    all = m.__all__
                except AttributeError:
                    pass
                else:
                    ensure_fromlist(m, all, 1)
            continue
        if sub != "*" and not hasattr(m, sub):
            subname = "%s.%s" % (m.__name__, sub)
            submod = import_module(sub, subname, m)
            if not submod:
                raise ImportError, "No module named " + subname

_watched_modules = {}

def import_module(partname, fqname, parent, force_reload=0):
    if not force_reload:
        try:
            return sys.modules[fqname]
        except KeyError:
            pass
    try:
        fp, pathname, stuff = imp.find_module(partname,
                                              parent and parent.__path__)
    except ImportError:
        return None
    try:
        m = imp.load_module(fqname, fp, pathname, stuff)
        if stuff[2] == imp.PY_SOURCE:
            pathname = os.path.abspath (pathname)
            _watched_modules[pathname] = (os.stat(pathname), m)
    finally:
        if fp: fp.close()
    if parent:
        setattr(parent, partname, m)
    return m

# Replacement for reload()
def reload_hook(module):
    name = module.__name__
    if '.' not in name:
        return import_module(name, name, None, force_reload=1)
    i = string.rfind(name, '.')
    pname = name[:i]
    parent = sys.modules[pname]
    return import_module(name[i+1:], name, parent, force_reload=1)

# Save the original hooks
original_import = __builtin__.__import__
original_reload = __builtin__.reload

# Now install our hooks
__builtin__.__import__ = import_hook
__builtin__.reload = reload_hook

#####################################################################
#
# Now the real new stuff
#
#####################################################################


def _monitor ():
    while 1:
        time.sleep (1)
        start = time.time()
        for path, (status, module) in _watched_modules.items():
            if os.stat(path) != status:
                superreload (module)

_old_classes = {}

def superreload(module, reload=reload):
    """superreload (module) -> module

    Enhanced version of the builtin reload function.
    superreload replaces the class dictionary of every top-level
    class in the module with the new one automatically.
    """
    for name, oldclass in module.__dict__.items():
        if type(oldclass) == types.ClassType:
     _old_classes[name] = _old_classes.get (name, []) + [oldclass]
            
    result = reload(module)
            
    for name, oldclasses in _old_classes.items():
 newclass = getattr(module, name)
 for oldclass in oldclasses:
     oldclass.__dict__ = newclass.__dict__.copy()
    return result

_worker = None

def run():
    global _worker
    _worker = threading.Thread (target=_monitor)
    _worker.setDaemon(1)
    _worker.start()
--------cut here--------




From tismer at tismer.com  Thu Apr 27 13:23:57 2000
From: tismer at tismer.com (Christian Tismer)
Date: Thu, 27 Apr 2000 19:23:57 +0200
Subject: Py_BuildValue
References: <390873A5.E94E5376@mitre.org>
Message-ID: <390877AD.EC6241C0@tismer.com>


Pete wrote:
> 
> Does anyone know where Py_BuildValue lives,  or what arguments it takes?

Yes. I know. :-)

ciao - chris

p.s.:
It is in Python/modsupport.c, line 384.
It uses a format string and a variable number or arguments.
For the description of its usage, see chapter 1.9 of
"Extending and Embedding the Python Interpreter".
This document is part of the standard distribution.

again ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaunstr. 26                  :    *Starship* http://starship.python.net
14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     where do you want to jump today?   http://www.stackless.com



From python at rose164.wuh.wustl.edu  Sat Apr 29 16:52:38 2000
From: python at rose164.wuh.wustl.edu (David Fisher)
Date: Sat, 29 Apr 2000 15:52:38 -0500
Subject: Problem running NNTP example from docs.
References: <87hfckk6fv.fsf@tea.thpoon.com>
Message-ID: <008001bfb21c$dca08900$61e9093f@spkydomain>

----- Original Message -----
From: "Arcady Genkin" <a.genkin at utoronto.ca>
Newsgroups: comp.lang.python
To: <python-list at python.org>
Sent: Saturday, April 29, 2000 3:17 PM
Subject: Problem running NNTP example from docs.


> I'm very new to Python, and was trying to run the example from
> documentation to nntplib library. I get an error when I try to
> instantiate NNTP object.
>
> ,----
> | >>> import nntplib
> | >>> s = NNTP('news')

need to be:

s = nntplib.NNTP('news')
      ^^^^^^^

You get a name error because there is no name 'NNTP', but there is a name
'nntplib' with attribute (well, method I guess) 'NNTP'.

dir() will get you a list of names in the global name dictionary.
dir('nntplib') will get you a list of names (attributes) of object
'nntplib'.

dnf


> | Traceback (innermost last):
> |   File "<stdin>", line 1, in ?
> | NameError: NNTP
> `----
>
> Could somebody point out what I am missing?
>
> Thanks!
> --
> Arcady Genkin                                 http://www.thpoon.com
> Nostalgia isn't what it used to be.
> --
> http://www.python.org/mailman/listinfo/python-list




From Michael.Scharf at gmx.de  Thu Apr 13 20:46:06 2000
From: Michael.Scharf at gmx.de (Michael Scharf)
Date: Fri, 14 Apr 2000 02:46:06 +0200
Subject: Comparing PythonWin and IDLE?
References: <IOjJ4.6669$HG1.192116@nnrp1.uunet.ca> <38F5F313.832B4CE6@labs.agilent.com>
Message-ID: <38F66A4E.92E3BDCA@gmx.de>

David Smith wrote:
> 
> Warren Postma wrote:
> 
> >  I find once a module has
> > been imported once, it often will not import again, even if I have changed
> > the source (.py) file, the .pyc file does not appear to be regenerated, and
> > any errors in the .py file can only be fixed by editing and saving the .py
> > file, quitting the IDE and restarting. This appears to happen for both IDLE
> > and Python. Anybody else seen this?
> 
> When you edit a file under Idle, Save it and then hit Control-F5.  Then
> you can re-import the file.

But a re-import should also change existing instances of 
functions, classes and methods defined in the module. OK, 
there are some cases, where a new version is incompatible 
with the old data members, but sometimes it would be 
helpful....


Michael
-- 
     ''''\     Michael Scharf
    ` c-@@     TakeFive Software GmbH, a Wind River Company
    `    >     http://www.WindRiver.com
     \_ V      mailto:Michael.Scharf at gmx.de


From david.jeschke at pobox.com  Thu Apr 20 12:51:25 2000
From: david.jeschke at pobox.com (David Jeschke)
Date: Thu, 20 Apr 2000 16:51:25 GMT
Subject: Grand debugging of python in Emacs
Message-ID: <hAGL4.174753$_G1.1397916@news1.sttls1.wa.home.com>

When I try to run the python debugger, pdb, through the grand unified
debugger interface in Emacs I do not see any output from the debugger in the
shell window until the debugger has exited.

I'm running Emacs 20.6 with python 1.5.2 on NT4 (SP6).  I'm not using a
UNIX-type shell.  I have a short python program which is a series of three
print statements.  I launch the debugger using M-x pdb and all I see is the
gud shell window with a line showing the current directory.  If I hit C-c
C-s several times to step through the program nothing appears in the shell
window until I have pressed it a half a dozen times.  Then all at once I get
the output from the debugger showing up in the shell window:

Current directory is c:/progra~1/python/lib/
> <string>(0)?()
(Pdb) > <string>(1)?()
(Pdb) > e:\temp\test.py(0)?()
(Pdb) > e:\temp\test.py(2)?()
-> print "hello, world!"
(Pdb) hello, world!
> e:\temp\test.py(3)?()
-> print "yes, sir!"
(Pdb) yes, sir!
> e:\temp\test.py(4)?()
-> print "goodbye."
(Pdb) goodbye.
--Return--
> e:\temp\test.py(4)?()->None
-> print "goodbye."
(Pdb) --Return--
> <string>(1)?()->None
(Pdb)
Debugger finished

Has anyone else run into this problem before and figured out a solution?
Thanks in advance for any assistance.

David Jeschke





From neelk at brick.cswv.com  Fri Apr 28 21:39:37 2000
From: neelk at brick.cswv.com (Neel Krishnaswami)
Date: 29 Apr 2000 01:39:37 GMT
Subject: Python compiler ? (was Re: Python for drivers?)
References: <3900EDBE.282BC85@dsuper.net> <8dr2pn$ec8$1@newshost.accu.uu.nl> <3906f804@nntp.server.uni-frankfurt.de> <015b01bfaf8b$9cd03930$0500a8c0@secret.pythonware.com> <39076316@nntp.server.uni-frankfurt.de> <8e7sop$5io$1@nnrp1.deja.com> <FtnGu1.Dso@world.std.com>  <slrn8gf5im.vjb.nascheme@cranky.arctrix.com>
Message-ID: <slrn8gkhdq.3f3.neelk@brick.cswv.com>

Neil Schemenauer <nascheme at enme.ucalgary.ca> wrote:
> 
> I don't think this guy is making native code compilers.  The
> "compiled" TkDiff demo is still implemented using Tcl.  Unlike my
> grandfather, I can't read German so I can't figure out what his
> "compiler" is supposed to do from the "whitepaper".  Babelfish
> does not seem to help.

This sounds more like freeze or GM's installer than a compiler.

I'd guess that places to start if you wanted to try optimizing Python
would be, ordered by increasing sexiness/wizardry, are:

o The Oaklisp implementation. Oaklisp is a fully OO Scheme whose
  object system is basically identical to Python's. Its compiler
  and runtime would make a good starting point if you wanted 
  something we mortals could understand. The Squeak Smalltalk
  implementation would also be worth looking at (it seems to be
  about 3-4 times faster than Python). 

  http://www.cs.unm.edu/~bap/oaklisp/
  http://www.squeak.org

o Similix, a partial evaluation program specialization framework
  written by a gang of crazy Danes. If you wrote a Python interpreter
  in Scheme, you could use Similix to automatically generate
  a compiler from it.

  http://www.diku.dk/research-groups/topps/activities/similix.html

o Stalin, Jeffrey Mark Siskind's hyper-optimizing whole-program
  Scheme compiler. Basically, you would need to study it, and then
  use its techniques to implement a Python compiler. The Self project
  would probably also have much of use. 

  http://www.neci.nj.nec.com/homepages/qobi/software.html


Neel


From danzap_python at yahoo.com.br  Thu Apr 13 12:38:18 2000
From: danzap_python at yahoo.com.br (=?iso-8859-1?q?Daniel=20Zapata?=)
Date: Thu, 13 Apr 2000 13:38:18 -0300 (ART)
Subject: How to execute a program ?
Message-ID: <20000413163818.28277.qmail@web4303.mail.yahoo.com>

How do I execute a program or a batch file (under
Windows) using python and get the return information
???


Thanks,

Daniel

____________________________________________________________________________
YAHOO! BATE-PAPO
Conhe?a gente nova, fa?a amigos e divirta-se! - http://www.yahoo.com.br



From rminsk at my-deja.com  Tue Apr 11 14:29:18 2000
From: rminsk at my-deja.com (rminsk at my-deja.com)
Date: Tue, 11 Apr 2000 18:29:18 GMT
Subject: Porting Modules to Python
Message-ID: <8cvqti$698$1@nnrp1.deja.com>

I am porting a package called Modules (www.modules.org) to Python.
It is unfortunate that the name conflicts.  I am by no means a Python
expert and would like to ask some help in porting a initialization
script from perl to Python.  Thank you for your help...

The perl script follows...

if (! defined $ENV{MODULE_VERSION} ) {
        $ENV{MODULE_VERSION_STACK}="@VERSION@";
        $ENV{MODULE_VERSION}="@VERSION@";
} else {
        $ENV{MODULE_VERSION_STACK}="$MODULE_VERSION";
}

sub module {
        local ($exec_prefix);
        $exec_prefix = "@BASEPREFIX@".$ENV{MODULE_VERSION};

        eval `$exec_prefix/bin/modulecmd perl @_`;
}

if (! defined $ENV{MODULEPATH} ) {
  $ENV{MODULEPATH} = `sed 's/#.*$//' ${MODULESHOME}/init/.modulespath |
awk 'NF==1{printf("%s:",$1)}'`
}

if (! defined $ENV{LOADEDMODULES} ) {
  $ENV{LOADEDMODULES} = "";
}

$ENV{MODULESHOME} = "@prefix@";


Sent via Deja.com http://www.deja.com/
Before you buy.


From mcalla at home.com  Fri Apr 21 09:01:45 2000
From: mcalla at home.com (Mike Callahan)
Date: Fri, 21 Apr 2000 13:01:45 GMT
Subject: Tkinter's OptionMenu
Message-ID: <ZiYL4.110745$8S2.1383832@news1.rdc1.tn.home.com>

How does one change an OptionMenu widget's list? I know that you can do this
easily using the PMW widget, but I would like to stay with basic Tkinter
widgets to keep the code as small as possible. I could not find any methods
in the code and I could not find any examples in either the online
documentation or the new book on Tkinter.




From mjhand at concentric.net  Mon Apr 17 12:23:57 2000
From: mjhand at concentric.net (Manus Hand)
Date: Mon, 17 Apr 2000 10:23:57 -0600
Subject: Here is some starship.python.net news
References: <38FA9A30.33DD1994@concentric.net>
Message-ID: <38FB3A9D.27CCF793@concentric.net>

Here is some more news.  Everyone cross your fingers for the starship's
disk.

Manus

Reply-To: "Paul Wolotsky" <pdw at mcsp.com>
Date: Mon, 17 Apr 2000 05:04:22 -0400
Organization: Medical Computer Systems

I actually have been out of town, and will return to DC tomorrow.  I am
not
sure off hand who the person is, but I have had my secretary contact
him.  I
am not sure what the schedule is for him to come in.  My understanding
from
my NOC technicians is that the primary disk has crashed in that machine -
so
perhaps there is a delay in someone coming in to fix it in terms of them
getting a replacement disk, etc.  I will let you know tomorrow when I
return
to the office - if there is something that you can do I will advise at
that
time!!!  Thanks for the offer.



From jmassung at magpiesystems.com  Mon Apr 24 14:04:25 2000
From: jmassung at magpiesystems.com (Jeff Massung)
Date: Mon, 24 Apr 2000 12:04:25 -0600
Subject: Curses
Message-ID: <IW%M4.420$r21.19263@news.uswest.net>

Is the curses module only for linux/unix? or are there corresponding modules
for Win32 and Mac?

Jeff




From jtschirr at engineering.uiowa.edu  Sat Apr 15 10:58:06 2000
From: jtschirr at engineering.uiowa.edu (Juerg Tschirren)
Date: Sat, 15 Apr 2000 14:58:06 GMT
Subject: Imaging Library & Numerical Python
References: <38F7F865.9C86399C@engineering.uiowa.edu> <PjTJ4.137181$Dv1.1488334@news1.rdc1.bc.home.com>
Message-ID: <38F883B6.A8ADF538@engineering.uiowa.edu>

To be honest, I didn't try yet. I am at the very begining of Python
programming (my experience is restricted to some two or three scripts
with 20 lines or so each). I just try to evaluate and see if these
packages fit my needs. That's why I asked this "lazy" question. But
after your answer now, it looks very promising!

Thanks

Juerg


Kevin Cazabon wrote:
> 
> I use them all the time together... are you experiencing any problems?
> 
> "Juerg Tschirren" <jtschirr at engineering.uiowa.edu> wrote in message
> news:38F7F865.9C86399C at engineering.uiowa.edu...
> >
> > Does anybody have experience with getting the Imaging Library and
> > Numerical Python working together? What I would like to do is importing
> > an imaging with the Imaging Library and then using functions from both
> > -- the Imaging Library and Numerical Python -- on this image.
> >
> > Juerg

-- 
+----------------------------------------------------------+
| Juerg Tschirren                   The University of Iowa |
| 2035 9th St., #6                  205 CC                 |
| Coralville, IA 52241              Iowa City, 52242       |
| USA                                                      |
|                                                          |
| jtschirr at engineering.uiowa.edu    home (319)351-4737     |
|                                   office (319)384-0660   |
+----------------------------------------------------------+


From mal at lemburg.com  Tue Apr  4 05:47:10 2000
From: mal at lemburg.com (M.-A. Lemburg)
Date: Tue, 04 Apr 2000 11:47:10 +0200
Subject: Text editing/replacing
References: <8cam4g$gdj$1@nnrp1.deja.com>
Message-ID: <38E9BA1E.F875915E@lemburg.com>

bragib at my-deja.com wrote:
> 
> Are there any good algorithms available for text editing/replacing
> of words in a huge text file.

Try mxTextTools available on my Python Pages.
 
> If anyone knows of any sample codes, books, algorithms etc.
> I would appreciate it.

Here are some pointers:

	http://www.cs.purdue.edu/homes/stelo/pattern.html
	http://www.dir.univ-rouen.fr/~charras/string/string.html

-- 
Marc-Andre Lemburg
______________________________________________________________________
Business:                                      http://www.lemburg.com/
Python Pages:                           http://www.lemburg.com/python/





From just at letterror.com  Tue Apr  4 06:32:29 2000
From: just at letterror.com (Just van Rossum)
Date: Tue, 4 Apr 2000 11:32:29 +0100
Subject: Python 1.6 alpha 1 released
In-Reply-To: <8cca93$9oh$1@nnrp1.deja.com>
References: <Pine.GSO.4.10.10004040035420.21084-100000@sundial>
 <5E877FD7637AB46E.DABBED9CF3C1D64B.CDDDC05AE7374A78@lp.airnews.net>
Message-ID: <l03102802b50f72e09280@[193.78.237.148]>

At 8:48 AM +0000 04-04-2000, fredrik at pythonware.com wrote:
>> Consider that *every one* of the protocol modules which offers
>> a connect() method uses a two-argument signature for it:  ftplib,
>> gopherlib, httplib, imaplib, nntplib, poplib, smtplib, telnetlib,
>> urllib.  And they all *use* the two-argument connect when invoking
>> their underlying socket!
>
>not true.  look again (doesn't matter if you look in the 1.5.2
>library or the 1.6 library -- what you just said is not correct
>in either case).

Here's the result from a quick search through the Python 1.5.2 lib:

ftplib.py, line 109: 	def connect(self, host = '', port = 0):
ftplib.py, line 267: 			conn.connect(host, port)
gopherlib.py, line 69: 	s.connect(host, port)
httplib.py, line 51:         if host: self.connect(host, port)
httplib.py, line 62:     def connect(self, host, port = 0):
httplib.py, line 79:         self.sock.connect(host, port)
imaplib.py, line 185: 		self.sock.connect(self.host, self.port)
nntplib.py, line 70: 		self.sock.connect(self.host, self.port)
poplib.py, line 80: 		self.sock.connect(self.host, self.port)
smtplib.py, line 182:             (code, msg) = self.connect(host, port)
smtplib.py, line 195:     def connect(self, host='localhost', port = 0):
smtplib.py, line 216:         self.sock.connect(host, port)
test_socket.py, line 136:             s.connect(hostname, PORT)
urllib.py, line 571:         self.ftp.connect(self.host, self.port)

>(bogus facts or double standards doesn't count).

Lies don't help much either, Fredrik!

I honestly think Tres Seaver made some *excellent* points. Whether you like
it or not, people have been using the demos *and* the library as
documentation. Even if the official documentation says it should be a
tuple: one line of example code is worth more than 10 lines of
documentation, and in this case we're talking about tens of line of
examample code versus *2* lines of documentation.

Just





From owen at astroNOJNK.washington.edu.invalid  Thu Apr  6 19:21:50 2000
From: owen at astroNOJNK.washington.edu.invalid (Russell E. Owen)
Date: Thu, 06 Apr 2000 16:21:50 -0700
Subject: Tuples -- who needs 'em
References: <AF223B12F1FDD2119253006097C4FCEC0AD357@cats.pdri.xerox.com> <8chol6$6em$1@newshost.accu.uu.nl> <026e01bf9fe7$9554fd20$74eb0b18@stcla1.sfba.home.com>
Message-ID: <8cj66e$1aca$1@nntp6.u.washington.edu>

In article <026e01bf9fe7$9554fd20$74eb0b18 at stcla1.sfba.home.com>, "Bob 
Alexander" <bobalex at home.com> wrote:

>There are other possible motivations besides fear of criticism, such as 
>the
>desire to observe the spirit of the language. In doing that, it has 
>occurred
>to me that tuples might not provide enough benefit to warrant the 
>additional
>cognitive load. So far there have been zero responses in favor of my point
>of view, so I'm in a very tiny minority  :-)

Well, some of us have just been quiet. I agree with you. I don't see 
that tuples add anything to the elegance or expressiveness of the 
language.

As far as I can see, the only significant advantage tuples have over 
lists is that they can be used as keys in dictionaries (if the tuple 
contains no mutable elements). And that seems to be for purposes of 
efficiency in the implementation of dictionaries.

If that is so, fine. Dictionaries are important and so is performance. 
And every language has its quirks.

Some folks have argued that tuples are good because one sometimes wants 
lists as constants. That brings up a different point. Does Python have 
any way to create named constants (like C++'s "const")? I'm hoping, but 
haven't found it yet.

Thanks for the advice to avoid tuples if a list will do as well, 
especially during the early phases of development. Sounds wise.

Regards,

-- Russell


From don.limpio at gmx.net  Fri Apr 28 10:16:38 2000
From: don.limpio at gmx.net (Georg Hammerbacher)
Date: Fri, 28 Apr 2000 16:16:38 +0200
Subject: newbie   dbase dataaccess
References: <ms19e8.uv5.ln@dorian.georks.de> <mtiigskn8uumhk94rklnr2fdealunasoot@4ax.com>
Message-ID: <rd6ce8.ss8.ln@dorian.georks.de>

Hello Anders,

thanks for your guess.

The thing inside my head is to write an adressmanager
- which is the *central* adressbook
- who works under Windows and Linux
- has an GUI Interface
this are the features for the beginning.

my dream is (under the GPL licence), to expand this piece of software and
integrate the good Features of MS-Outlook and Lotus Organizer. (comments are
welcome)

I?ve found "gadfly" for databasing. Is it really so good as it seems??

Thanks
Georg


Anders M Eriksson <anders.eriksson at morateknikutveckling.se> schrieb in im
Newsbeitrag: mtiigskn8uumhk94rklnr2fdealunasoot at 4ax.com...
> On Thu, 27 Apr 2000 11:40:46 +0200, "Georg Hammerbacher"
> <don.limpio at gmx.net> wrote:
>
> >Hello everybody,
> >
>
> Hello Georg!
>
> >I?m just new in programming Python and now I want to make my own personal
> >adressmanager with a simple dBase-file.
> >How can I get access to it or where can I find more info?s about using
> >dbase-files within Python?
> >
> First: Why do you want to use dBASE as the database format?
>
> I haven't found any good dBASE interface for Python. So if you want to
> use the dBASE format you need to use ODBC and then you hopefully are
> on a computer that runs unders Windows NT/98.
>
> If you don't NEED to use dBASE then look at
> http://www.python.org/topics/database/ here you will find alot of info
> about how to use Python for accessing databases....
>
> // Anders
>




From tjreedy at udel.edu  Tue Apr  4 15:01:17 2000
From: tjreedy at udel.edu (Terry Reedy)
Date: Tue, 4 Apr 2000 15:01:17 -0400
Subject: Python 1.6 alpha 1 released
References: <200003312130.QAA04361@eric.cnri.reston.va.us> <38E5205F.DE811F61@roguewave.com> <Wf9F4.4423$74.70231@newsc.telia.net> <38E58D2D.384F1546@Lugoj.Com> <S9kF4.4472$74.71414@newsc.telia.net> <38E6B929.28816494@roguewave.com> <auGF4.914$n68.186931712@newsb.telia.net> <38E7BAC4.3B7AD0FF@uswest.net> <Z_PF4.4787$74.76606@newsc.telia.net>
Message-ID: <8cddpk$4mf$1@news.udel.edu>

"Fredrik Lundh" <effbot at telia.com> wrote in message
news:Z_PF4.4787$74.76606 at newsc.telia.net...

> so in other words, either this change remains, and the example
> should be fixed.

(Including the examples in the PP and IPwP books.)

>  or someone adds backwards compatibility to the
> socket module, in which case the method description should be
> fixed:
>
>     connect(address[, port])

I agree that the conflict/inconsistency should be fixed.  I prefer this
remedy.  Good suggestion.

TJR





From nascheme at enme.ucalgary.ca  Thu Apr 20 02:11:53 2000
From: nascheme at enme.ucalgary.ca (Neil Schemenauer)
Date: Thu, 20 Apr 2000 06:11:53 GMT
Subject: __del__ problem - would adopting Garbage Collection fix this?
References: <aw2L4.8907$HG1.245743@nnrp1.uunet.ca> <7ojL4.9424$HG1.254543@nnrp1.uunet.ca> <Z%oL4.9643$HG1.257533@nnrp1.uunet.ca> <m31z41iyhg.fsf@atrus.jesus.cam.ac.uk> <38FE7F1E.168E914A@cosc.canterbury.ac.nz> <l03102803b52456fb3df8@[193.78.237.146]>
Message-ID: <slrn8ft7um.u9n.nascheme@cranky.arctrix.com>

Just van Rossum <just at letterror.com> wrote:
>Hm, wouldn't it be better to collect cyclic objects with __del__
>anyway, and just not call __del__? Possibly printing something
>to stderr, similarly to when an exception occurs within a
>__del__ method.

I'm not sure if the is any better.  By default, a message is
printed to sys.stderr when uncollectable objects are found.  The
programmer should be able to redesign things to fix this.

    Neil

-- 
'Slashdot, with its uncontrolled content and participants' poor
impulse control, remains Internet culture's answer to "Lord of
the Flies."' - Salon


From akuchlin at mems-exchange.org  Sat Apr 15 12:53:37 2000
From: akuchlin at mems-exchange.org (Andrew Kuchling)
Date: 15 Apr 2000 12:53:37 -0400
Subject: Scripting and Gnome and KDE
References: <Pine.GSO.4.10.10004150741520.11261-100000@sundial>
Message-ID: <3dk8hzb90u.fsf@newcnri.cnri.reston.va.us>

Moshe Zadka <moshez at math.huji.ac.il> writes:
> Personally I chose Gtk+/GNOME, since (among other things), the Gtk+ type
> system is specifically targetted to support bindings to other languages.

I've played a bit (only a bit, though) with both PyGTk and PyQt.
While I think I marginally prefer PyGTk, because James Henstridge has
done a really nice job of writing the bindings, I use KDE on my
desktop at home and at work, and therefore the PyQt/KDE bindings are
of greater practical interest to me.  They're pretty good, too, though
the extension module is utterly huge, around 4Mb last time I looked.

Every so often I go to some Linux conference and come within the range
of Miguel de Icaza's reality distortion field, and I get all excited
over some new GNOME feature, such as Bonobo's embedding, and try to
compile GNOME when I get home.  Then reality sets in: GNOME uses
patched versions of all sorts of things such as automake, but this is
never clearly documented anywhere.  There's a large pile of libraries,
and it's not clear which ones are needed and which ones are obsolete;
oh, you need imlib, but wait, it's been replaced by gdk-pixbuf, and
esound has been replaced by something else... Usually I never manage
to get a fully working GNOME setup, and the reason for the failure is
usually not apparent.

KDE, on the other hand, is pretty easy to compile, even from the CVS
tree; compile Qt, compile kdesupport and kdelibs, and that's it --
you're all set to compile chosen applications.  Much better
structuring of the development environment -- KDE's setup shares some
of Python's elegance.  (Maybe it's a European thing... :) ) So it's
odd; while I prefer PyGTk, it's tied to a much less well-structured
environment and hence loses out through no fault of its own.

--amk


From John.Grayson at GD-CS.COM  Mon Apr 17 07:12:26 2000
From: John.Grayson at GD-CS.COM (Grayson, John)
Date: Mon, 17 Apr 2000 07:12:26 -0400
Subject: Tkinter Canvas -- help?
Message-ID: <3774EF539472D211B98F0008C7F468A1033B681A@TNTNEX01.tntn.gtegsc.com>

 jkior writes:

> I'm just starting to learn Python/Tkinter, and I have found that there
> is virtually _no_ documentation for the Tkinter canvas. Could anybody
> help? I'm specifically looking for stuff like creating an image by a
> grid in canvas on "mouseDown", and how to read the items in the canvas,
> and write them to a text file, in their proper positions. I would
> _really_ appreciate it if anybody could help me with this.

> P.S. The only Tkinter canvas documentation that I have is the
> "Introduction to Tkinter" from pythonware.com

> Thanks,
> jkior

If you want some *free* clues, take a look at:

    http://www.manning.com/Grayson/chapt11.pdf

Chapter 10 contains lots of canvas info...

If you decide to buy the book, use this link:

    http://www.python.org/psa/bookstore/

since the PSA will benefit.


John Grayson

General Dynamics
Communication Systems
400 John Quincy Adams
Taunton, MA 02780




From ppetru at coltronix.com  Wed Apr 12 20:17:52 2000
From: ppetru at coltronix.com (Petru Paler)
Date: Thu, 13 Apr 2000 03:17:52 +0300
Subject: How to escape escape?
In-Reply-To: <8d32kl$sjg$1@nnrp02.primenet.com>; from larrygg@primenet.com on Wed, Apr 12, 2000 at 04:59:01PM -0700
References: <8d32kl$sjg$1@nnrp02.primenet.com>
Message-ID: <20000413031751.F10842@coltronix.com>

On Wed, Apr 12, 2000 at 04:59:01PM -0700, Larry Gantz wrote:
> I'm new to the snake.
> I think I would like to be able to turn off and on the escape character
> processing so the snake would perform escape processing when I want it.
> 
> It seems that Python always has escape processing.

   Instead of 

mystring = 'asd'
use
mystring = r'asd'

   In the "r" form only the closing quote is a metacharacter (you can't even
escape it).

-Petru



From dworkin at ccs.neu.edu  Thu Apr 27 18:11:30 2000
From: dworkin at ccs.neu.edu (Justin Sheehy)
Date: 27 Apr 2000 18:11:30 -0400
Subject: Simple Modules Question
In-Reply-To: "Akira Kiyomiya"'s message of "Thu, 27 Apr 2000 14:20:06 -0700"
References: <8eaaul$16p@autodesk.autodesk.com>
Message-ID: <vnd66t3urcd.fsf@camelot-new.ccs.neu.edu>

"Akira Kiyomiya" <akira.kiyomiya at autodesk.com> writes:

> from string import atoi    # imports 'string'
>                                      # places atoi in current namespace

That comment is incorrect.  This statement gets atoi from string into
the current namespace, but does not really import string.

Unless you are using the interactive interpreter or are using one of
the very few modules designed for 'from module import *', you
shouldn't use that.  Do 'import module' and use full names of module
members instead.

-Justin

 



From tuttledon at hotmail.com  Wed Apr 19 22:51:52 2000
From: tuttledon at hotmail.com (Don Tuttle)
Date: Thu, 20 Apr 2000 02:51:52 GMT
Subject: Pythonwin?
References: <eMz2Nmiq$GA.73@net025s> <4QtL4.24$v85.41@news-server.bigpond.net.au>
Message-ID: <chuL4.8483$O7.17315@typhoon.southeast.rr.com>

[Mark]
> You are really going to need to go through an MFC tutorial or book before
> you can make heads or tails of Pythonwin from that level...

Any recommendations?

Don




From aahz at netcom.com  Wed Apr  5 19:59:51 2000
From: aahz at netcom.com (Aahz Maruch)
Date: 5 Apr 2000 23:59:51 GMT
Subject: Zope consultant?  (Mtn. View)
Message-ID: <8cgk1n$gdo$1@nntp9.atl.mindspring.net>

My bosses haven't had much luck getting a response from Digital
Creations, so they asked me to check around and see if there are any
Zope consultants near Mountain View, California.  Any takers?
--
                      --- Aahz (Copyright 2000 by aahz at netcom.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

"It's 106 miles to Chicago.  We have a full tank of gas, a half-pack of
cigarettes, it's dark, and we're wearing sunglasses."  "Hit it."


From moshez at math.huji.ac.il  Wed Apr 19 00:48:24 2000
From: moshez at math.huji.ac.il (Moshe Zadka)
Date: Wed, 19 Apr 2000 07:48:24 +0300 (IDT)
Subject: can someone tell me about...
In-Reply-To: <38FCBB16.BF50EB03@stsci.edu>
Message-ID: <Pine.GSO.4.10.10004190747190.3061-100000@sundial>

On Tue, 18 Apr 2000, Kia Arab wrote:

> variable names that start an underscore "_".
> 
> Interactively, python will recognize a variable name that begins with
> one underscore (_myVar). The variable naming rules even say that is
> acceptable.
> 
> However, when I try to use it in a module, it bombs. 

Please give an example of the "bombinh". I suspect you do 
"from module import *". 

Solutions:

a) don't do it
b) read the language ref. about from module import *, and names that begin
with underscore.
--
Moshe Zadka <mzadka at geocities.com>. 
http://www.oreilly.com/news/prescod_0300.html
http://www.linux.org.il -- we put the penguin in .com




From walter at data.franken.de  Thu Apr 27 04:56:52 2000
From: walter at data.franken.de (Walter Doerwald)
Date: Thu, 27 Apr 2000 10:56:52 +0200
Subject: [Python-Dev] Re: Python 1.6a2 Unicode bug (was Re: comparing strings and ints)
References: <l03102805b52ca7830b18@[193.78.237.154]> <39072F0C.5214E339@lemburg.com> <BlHN4.1719$wYl.195075584@newsb.telia.net> <oqvh14fikb.fsf@titan.progiciels-bpi.ca>
Message-ID: <45919986@data.franken.de>

On 26 Apr 2000 21:17:56 -0400 Fran?ois Pinard wrote:

> "Fredrik Lundh" <effbot at telia.com> writes:
> 
> > M.-A. Lemburg <mal at lemburg.com> wrote:
> 
> > does this mean that the 8-bit string type is deprecated ???
> 
> I did not follow the implementation discussions, but from the outside, it
> looks like 8-bit strings are kept and will be speedy.  People should
> either:
> 
> 1) use them for ASCII without bothering or thinking much,
> 
> 2) write Python source files in UTF-8, relying on the default encoding,

Does this only affect the strings variables in the source, or the comments
too?

Would it be possible to write the source code in UTF-16 (which is much
easier to to under WinNT with NotePad, I know no text editor under that
writes UTF-8)

Bye...
	Walter Doerwald



From effbot at telia.com  Sat Apr  1 14:38:15 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Sat, 01 Apr 2000 19:38:15 GMT
Subject: Python 1.6 alpha 1 released
References: <200003312130.QAA04361@eric.cnri.reston.va.us> <38E5205F.DE811F61@roguewave.com> <Wf9F4.4423$74.70231@newsc.telia.net> <38E58D2D.384F1546@Lugoj.Com> <S9kF4.4472$74.71414@newsc.telia.net> <38E646B6.12011A4E@Lugoj.Com>
Message-ID: <HesF4.4556$74.72888@newsc.telia.net>

James Logajan wrote:
> Any reason Guido doesn't post to this newsgroup like he used to?

well, what do *you* think is the reason?

> The pragmatist in me says I should look for a language that abides
> by a standard that can not be easily changed.

ahem.  GvR changed the CPython implementation to adhere
to the standard.  *you* want him to change the standard.

...

but sure, if you know of a language which is strictly specified,
never changes, and for which all implementations are truly strict
(that is, they shouldn't just accept all valid constructs, but also
flag all possible invalid constructs -- and be completely bug free,
of course), let us know.

</F>




From arta at x-stream.nl  Wed Apr 19 12:37:11 2000
From: arta at x-stream.nl (A[r]TA)
Date: Wed, 19 Apr 2000 18:37:11 +0200
Subject: Pythonwin?
Message-ID: <eMz2Nmiq$GA.73@net025s>

Hello,

I've just started with Pythonwin and I want to create a main application
with many dialogs
and stuff like that. But how do I create a main Window? With the example
down here,
it doesn't shows a Window or something and I don't get an error.

    import win32ui
    import win32con
    import win32api
    from pywin.mfc import window, object

    window.Wnd()
    object.Object()


Plzzz, help this newbee!!
Thnx in advance, A[r]TA


--
We shall not cease from exploration, and the end of all our exploring will
be to arrive where we started and know the place for the first time.
- T. S. Eliot









From tjung at interport.net  Mon Apr 17 11:07:34 2000
From: tjung at interport.net (Tony Jung)
Date: Mon, 17 Apr 2000 15:07:34 GMT
Subject: How do you Kill a run away process in Oracle?
Message-ID: <38fb2862.4140403@news.supernews.com>

Hello,


I was wondering if anyone has been able to automate a procedure to
kill a run away process in Oracle with Python?


Thanks,

- Tony


From rob at hooft.net  Mon Apr  3 02:17:47 2000
From: rob at hooft.net (Rob W. W. Hooft)
Date: 03 Apr 2000 08:17:47 +0200
Subject: Tkinter:  "Get root for a widget" is private?
References: <20000331112048.A1257432@vislab.epa.gov> <QdEF4.892$n68.190327808@newsb.telia.net>
Message-ID: <m3snx3n1x0.fsf@devel.no.rwd.dinet>

>>>>> "FL" == Fredrik Lundh <effbot at telia.com> writes:

 FL> Randall Hopper wrote:
 >> I have a widget and want to retrieve the root.  I see the _root()
 >> method, but by Python convention, isn't this method private?
 >> 
 >> Should users of Tkinter be calling this?

 FL> as long as nobody calls the Tkinter police, you're safe.

 FL> however, using the *toplevel* rather than the actual root object
 FL> is usually sufficient:

 FL>     root = widget.winfo_toplevel()

 FL> to get from the toplevel to the real root, just add:

 FL>     if root.master: root = root.master # real root

Hm. Except that my Toplevel windows are often nested deeply, and
certainly not restricted to children of the root! I think that your
approach is distinctly different from _root() except if you severely
restrict your toplevel window creation.

Rob

-- 
=====   rob at hooft.net          http://www.xs4all.nl/~hooft/rob/  =====
=====   R&D, Nonius BV, Delft  http://www.nonius.nl/             =====
===== PGPid 0xFA19277D ========================== Use Linux! =========


From jnw22 at cam.ac.uk  Tue Apr 11 09:20:41 2000
From: jnw22 at cam.ac.uk (James "Wez" Weatherall)
Date: Tue, 11 Apr 2000 14:20:41 +0100
Subject: Huh? func_defaults, default values in function calls
References: <8cv5qj$35p$1@zinc.intern.netconnect.no>
Message-ID: <8cv8ra$hio$1@pegasus.csx.cam.ac.uk>

"Roger Baklund" <roger at netconnect.no> wrote in message
news:8cv5qj$35p$1 at zinc.intern.netconnect.no...
[snip of some code]
>
> I don't get it. Why isn't the dictionary reset to {'x':'a'} when i call
> the function the last time?
>
> >>> a.func_defaults
> ('', '', {'x': 'a', 'y': 'c'})
>
> Why is not {} the default for the z parameter?

According to my understanding of the Python docs, defaults are only
evaluated at function definition time, not at function invokation time.

As a result, if you set a mutable object (list, dictionary, etc) as a
default, it'll be set when the function is defined and then if the function
changes it, it'll stay changed for subsequent invokations.  The function
effectively has a "side-effect", of modifying it's own default!

You won't get this behaviour with something like:

def myfun(x, a = 5):
    a = x

though, I don't think, because "5" isn't mutable.

Cheers,

--
James "Wez" Weatherall





From moshez at math.huji.ac.il  Mon Apr  3 18:15:53 2000
From: moshez at math.huji.ac.il (Moshe Zadka)
Date: Tue, 4 Apr 2000 00:15:53 +0200 (IST)
Subject: Unicode program representation
In-Reply-To: <%o3G4.1111$n68.191610880@newsb.telia.net>
Message-ID: <Pine.GSO.4.10.10004040011450.21084-100000@sundial>

On Mon, 3 Apr 2000, Fredrik Lundh wrote:

> hmm.  wouldn't that mean that we end up using different encodings
> in different parts of the script?  feels a little scary, to say the least...

FWIW, I believe that Python's (official) parser should be strictly UTF-8.
It could be done in stages:
 -- stage 1: warn about non-ASCII characters in scripts (warning framework
     TBD)
 -- stage 2: don't accept non-ASCII characters in scripts at all
 -- stage 3: assume all scripts are UTF-8

This isn't a "socket.connect"-like "I've been using this feature for
years" issue: it's trivial to write a script to convert between anything
and UTF-8 (in Python, of course <wink>).

However, there are some non-trivial issues here: should Python identifiers
be able to include all characters Unicode defines as letters?
--
Moshe Zadka <mzadka at geocities.com>. 
http://www.oreilly.com/news/prescod_0300.html
http://www.linux.org.il -- we put the penguin in .com




From helka.folch at edf.fr  Tue Apr  4 04:18:15 2000
From: helka.folch at edf.fr (Helka FOLCH)
Date: Tue, 04 Apr 2000 10:18:15 +0200
Subject: Which DOM implementation?
Message-ID: <38E9A547.97CE0D0C@edf.fr>

There seems to be at least 2 python DOM implementations (4DOM and the
one in the standard xml distribution).
Can you give me advice as to which one to choose  for a project I'm
undertaking? (its quite a long-term project so I wouldn't like to have
compatability problems half way through)

Many Thanks,

Helka Folch



From eviltofu at rocketmail.com  Tue Apr 25 10:05:25 2000
From: eviltofu at rocketmail.com (Jerome Chan)
Date: Tue, 25 Apr 2000 14:05:25 GMT
Subject: Question about exception-handling mechanism
References: <8e46mt$4nh$1@nnrp1.deja.com>
Message-ID: <eviltofu-A9A4DD.10052525042000@news.earthlink.net>

In article <8e46mt$4nh$1 at nnrp1.deja.com>, willfg at my-deja.com wrote:

> Hello all,
> 
>    I'm new to Python, but find its exception handling mechanism much
> more powerful than comparable languages; but a question was put to me
> that being new I can't come up with a very articulate answer. A
> colleague asked why in an exception handling mechanism you'd want the
> ELSE block to be executed if you don't throw an exception as opposed to
> a FINALLY block. Anyone used this feature in practice? Thanks in
> advance for your input, -- Will
> 
> 
> Sent via Deja.com http://www.deja.com/
> Before you buy.

According to the Python Pocket Reference

else: Run if no exceptions raised
finally: Always run on the way out

but the question I have is the following definations for the exception 
handling routines.

try: suite
[except [name [,data]]: suite]*
[else: suite]

try: suite
finally: suite

Does this mean I can't do
try:
    x = MyClass("s")
    # some other stuff that might raise an exception
except Exception,message:
    print message
else:
    print "something went wrong but I don't know what it is!"
finally:
    del x # or some other function

Also, is there a command line switch that would do what Java does when 
checking for exception handling? That the source code would be checked 
to see if any exceptions thrown are not caught? Or maybe a 3rd party 
utility?


From mhammond at skippinet.com.au  Wed Apr 19 22:24:45 2000
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Thu, 20 Apr 2000 02:24:45 GMT
Subject: ADO Question
References: <20000419203624.53393.qmail@hotmail.com>
Message-ID: <NTtL4.28$v85.219@news-server.bigpond.net.au>

"Matheus Rodrigues" <matheusrodrigues at hotmail.com> wrote in message
news:20000419203624.53393.qmail at hotmail.com...

> Traceback (innermost last):
>   File "c:\python\win32com\client\dynamic.py", line 242, in
_make_method_
>     codeObject = compile(methodCode, "<COMObject %s>" %
> self._username_,"exec")
>   File "<string>", line 1
>      def Open(self, ConnectionString=L'', UserID=L'', Password=L'',
> Options=-1):
>
>                                        ^
> SyntaxError: invalid syntax

What win32all build is this?  The syntax error is the L'', but I can't
work out who generated this funky syntax...

Mark.





From thomas at xs4all.net  Tue Apr 18 17:49:32 2000
From: thomas at xs4all.net (Thomas Wouters)
Date: Tue, 18 Apr 2000 23:49:32 +0200
Subject: Another question
In-Reply-To: <p53L4.1217$zk4.22638@news.uswest.net>; from jmassung@magpiesystems.com on Tue, Apr 18, 2000 at 02:01:32PM -0600
References: <dVFK4.41$vr2.5817@news.uswest.net> <38FB2EEB.A32AF88C@bam.com> <38FB339F.604364AD@bam.com> <200004171605.MAA04929@zippy.init.com> <38FB7610.80EAF965@ichips.intel.com> <200004172120.RAA11102@zippy.init.com> <jWNK4.432$vr2.32822@news.uswest.net> <m3ln2cckg6.fsf@atrus.jesus.cam.ac.uk> <p53L4.1217$zk4.22638@news.uswest.net>
Message-ID: <20000418234932.L15664@xs4all.nl>

On Tue, Apr 18, 2000 at 02:01:32PM -0600, Jeff Massung wrote:

> 2. There are many different ways to do something (ala Perl) - which has
> drawbacks and advantages.

You've failed to spot the obvious difference between the many different ways
of Python and the many different ways of, eg., Perl, though ;)

In Python, everything is based off a few very simple concepts. For instance,
the namespace tricks you saw in this thread all boiled down to 2 simple
things:

- Assignment does not copy an object, it only binds a local (or, when you
use the 'global' keyword on the name first, a global) name to that object.

- Everything is an object, and you can any object anywhere.

In comparison, Perl idioms often deal with special cases (or so it seems
from my point of view) -- usually (but absolutely no way near always)
confusing perl constructs can be boiled down to some form of logical rule,
but the overall quantity of 'logical rules' is immense, compared to Python.

(I could post some perlish examples if necessary, but I'm sure noone here
wants to see them ;)

Of course not all Python code can be boiled down to these two things (they
aren't really 'rules', by the way, merely descriptions of the Python view of
the world) but understanding these subtle differences can really make you
grok Python ;) Tim Peters once wrote down the Python Philosophy (or zen):

"""
Beautiful is better than ugly.
Explicit is better than implicit.
Simple is better than complex.
Complex is better than complicated.
Flat is better than nested.
Sparse is better than dense.
Readability counts.
Special cases aren't special enough to break the rules.
Although practicality beats purity.
Errors should never pass silently.
Unless explicitly silenced.
In the face of ambiguity, refuse the temptation to guess.
There should be one-- and preferably only one --obvious way to do it.
Although that way may not be obvious at first unless you're Dutch.
Now is better than never.
Although never is often better than *right* now.
If the implementation is hard to explain, it's a bad idea.
If the implementation is easy to explain, it may be a good idea.
Namespaces are one honking great idea -- let's do more of those!

dict-is-short-for-dictator-cuz-gui-was-already-in-use-ly y'rs  - tim
"""

And I must say, thinking about the entries I didn't understand right away
brought me lots closer to the Truth, and rethinking those I thought I'd
figured out right away even more ;)

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From effbot at telia.com  Wed Apr 12 11:00:00 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Wed, 12 Apr 2000 15:00:00 GMT
Subject: XML-RPC and None objects
References: <m12fNMH-000CnCC@artcom0.artcom-gmbh.de>
Message-ID: <Qb0J4.649$rc9.190983680@newsb.telia.net>

Peter Funk wrote:
> I stumbled into the same problem and since I currently use XML-RPC only
> between my own server and clients I decided to ignore the official
> specification and added a 'None'-handler to my version of Fredrik Lundhs
> library.  The needed patch is indeed very very simple:  Add the following
> lines to the class 'Marshaller' in 'xmlrpclib.py' you are done:
>
>     def dump_None(self, none):
>         self.write("<value><null></null></value>\n")
>     dispatch[NoneType] = dump_None
>
> Since the class 'Unmarshaller' already returns 'None' for all unknown
values
> a modification there is not required.  This works fine for me.  YMMV.

footnote: if you change "null" to "nil", you'll be compatible
with several other XML-RPC implementations.

(iirc, the current xmlrpclib codebase has a flag to enable "nil"
encoding support.  it also fixes the accidental "return None
for unknown values" feature...).

otoh, you will soon be able to use "soaplib.py" instead of
"xmlrpclib.py", but that's another story.

</F>




From mwh21 at cam.ac.uk  Tue Apr 11 08:14:57 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 11 Apr 2000 13:14:57 +0100
Subject: [ANNOUNCE] Garbage collection for Python
References: <20000407221750.A29279@acs.ucalgary.ca> <38F13FB9.D1D16AF@mindspring.com> <slrn8f2if1.dop.nascheme@cranky.arctrix.com> <87r9cdhdb0.fsf@deneb.cygnus.argh.org> <m3bt3hkk76.fsf@atrus.jesus.cam.ac.uk> <87bt3geszz.fsf@deneb.cygnus.argh.org>
Message-ID: <m3u2h8kf5q.fsf@atrus.jesus.cam.ac.uk>

Florian Weimer <fw at deneb.cygnus.argh.org> writes:

> Michael Hudson <mwh21 at cam.ac.uk> writes:
> 
> > Florian Weimer <fw at deneb.cygnus.argh.org> writes:
> > 
> > > nascheme at enme.ucalgary.ca (Neil Schemenauer) writes:
> > > 
> > > > I would rather get the bugs stomped out of my patch and get it
> > > > included in 1.6 or 1.7.  We can then remove the FAQ entry.  :)
> > > 
> > > Is it possible to extend your GC in a way which would permit the
> > > execution of several bytecode interpreters in multiple threads?
> > 
> > Eh?
> 
> With the reference counting scheme, you can't do that, because you
> have to synchronize access to the reference count, which would result
> in a major performance penalty.

Oh, right.  But:

1) Neil's scheme does not replace reference counting.
2) There are more things between the status quo and a free threaded
   interpreter than reference counting issues.

So the answer to your question is: "no".

Cheers,
Michael

-- 
  well, take it from an old hand: the only reason it would be easier
  to program in C is that you can't easily express complex  problems
  in C, so you don't.                   -- Erik Naggum, comp.lang.lisp


From me at nospam.edu  Tue Apr 25 13:27:28 2000
From: me at nospam.edu (Me)
Date: Tue, 25 Apr 2000 17:27:28 GMT
Subject: VIM NT +python
References: <8e4j27$hgq$1@nnrp1.deja.com>
Message-ID: <4AkN4.9$kM.3048@news>

So would I!

<tiddlerdeja at my-deja.com> wrote in message
news:8e4j27$hgq$1 at nnrp1.deja.com...
> Does anyone here have a winNT build of the VIM editor that has python
> support switched on?
>
> I'd love to get my hands on a copy.
>
>
> Sent via Deja.com http://www.deja.com/
> Before you buy.




From vertega at yahoo.com  Thu Apr 13 20:06:56 2000
From: vertega at yahoo.com (vertega at yahoo.com)
Date: Fri, 14 Apr 2000 00:06:56 GMT
Subject: FREE DVD PLAYER!!
Message-ID: <AitJ4.5617$9F6.6669@typhoon.tampabay.rr.com>

Here is a new online site offering a free DVD player as part of a promotion.  Can't hurt to try I guess.  
 
http://www.ikena.com/downloads/?refCID=dbakriwa
 
 
Here is another site that is offering a Palm computer.  Why not?
 
http://www.jackpot.com/win.asp?ref=105-4980



From aahz at netcom.com  Mon Apr 17 20:57:43 2000
From: aahz at netcom.com (Aahz Maruch)
Date: 18 Apr 2000 00:57:43 GMT
Subject: Removing circular references (was Re: Editors and books)
References: <XBFK4.16$Uq2.1831@news.uswest.net> <20000417184833.H15664@xs4all.nl>
Message-ID: <8dgbu7$k10$1@slb6.atl.mindspring.net>

In article <20000417184833.H15664 at xs4all.nl>,
Thomas Wouters  <thomas at xs4all.net> wrote:
>
>The problem with RC is that you can create circular references:
>
>a = [some large list]
>b = [some other large list]
>a.append(b)
>b.append(a)
>
>del a
>del b

Note that you can break the circular reference if you do this between
the append and the del:

del a[-1]
del b[-1]

Similar tricks work in most other cases where circular references get
created.
--
                      --- Aahz (Copyright 2000 by aahz at netcom.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

"First one back from the bathroom gets to be the bottom."


From amitp at Xenon.Stanford.EDU  Tue Apr  4 22:42:45 2000
From: amitp at Xenon.Stanford.EDU (Amit Patel)
Date: 5 Apr 2000 02:42:45 GMT
Subject: New Features in Python 1.6
References: <200004011740.MAA04675@eric.cnri.reston.va.us> <oq66txol0w.fsf@titan.progiciels-bpi.ca> <GOpG4.5050$74.81782@newsc.telia.net> <38ea40a1.47458347@news.nikoma.de>
Message-ID: <8ce975$hda$1@nntp.Stanford.EDU>

 Philip 'Yes, that's my address' Newton <nospam.newton at gmx.li> wrote:
| 
| Yes? I challenge you to provide any algorithm that will strip away one
| bit from the four sequences 11, 10, 01, and 00, while still being able
| to restore the previous state. (Unless you were talking about lossy
| compression, in which case you might as well say "my compression
| algorithm just drops off the leftmost bit" or similar, which will
| *always* compress any non-empty file.)
| 

Okay:

  def compress(s):
    if s=='11':    return '11'
    elif s=='10':  return '1'
    elif s=='01':  return '0'
    elif s=='00':  return ''

  def uncompress(s):
    if s=='11':    return '11'
    elif s=='1':   return '10'
    elif s=='0':   return '01'
    elif s=='':    return '00'

Oh, maybe that doesn't count.  <0.5wink>

   -implicit-lengths-are-fun-ly y'rs, Amit



From cnetzer at stanford.edu  Thu Apr 20 16:11:11 2000
From: cnetzer at stanford.edu (Chad Netzer)
Date: Thu, 20 Apr 2000 13:11:11 -0700
Subject: Integer solutions to linear equation?
References: <slrn8fovo2.nq.grant@grante.comtrol.com>
Message-ID: <38FF645F.A7095EE8@stanford.edu>

Grant Edwards wrote:

> Is there a non-iterative way to
> solve a linear equation when the results are contrained to be
> integers?  I don't remember anything from any of my math
> classes that seems relevent, but I didn't take any anything
> beyond what is required for all undergrad engineers.

Besides the Euclid method described by others, you could also look into Groebner
Bases, which are a kind of generalization of Euclid's method (along with
Gaussian Elimination and Linear Programming).  I'm no expert, but here is a link
to a Grobner Bases info page,to get you started.  I doubt there is anything
written for Python, although there are good algorithms for computing these bases
that COULD be done in Python.  But there are other algebra systems that already
do it (including Mathematica and Maple).

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20000420/933c58aa/attachment.html>

From darrell at dorb.com  Tue Apr 18 22:33:28 2000
From: darrell at dorb.com (Darrell)
Date: Wed, 19 Apr 2000 02:33:28 GMT
Subject: Implementing a Python process when PC is idle?
References: <38fc9694@wwwproxy3.westgroup.com>
Message-ID: <YV8L4.6619$4w1.86918@typhoon.nyroc.rr.com>

"MikeZ"
>     I would like to execute a Python process that gets executed when my
> Windows PC is in screensaver mode (so my PC can work when it is idle).
>
Hi Mike !

If you set the priority on your task low then it won't be noticed except on
the performance monitor.

But if you really want to just run while the screen save is running...
I have only these ideas to offer.

HKEY_CURRENT_USER\CONTROL PANEL\SCREENSAVERNAME

Search for *.scr in WINNT, these are .exe's in disguise.
So the registry key points at a .scr that when run is the screen saver.
You could have the .scr file spawn off the real screen saver then do your
background work. When the screen saver returns stop the background work also
and exit.

This might all work in Python by creating a Python exe and renaming it to
.scr.


--Darrell Gallion






From jvickroy at sec.noaa.gov  Tue Apr 25 16:59:33 2000
From: jvickroy at sec.noaa.gov (j vickroy)
Date: Tue, 25 Apr 2000 14:59:33 -0600
Subject: passing Python list to C extension function
References: <3905DBA6.60AA7A48@sec.noaa.gov>
Message-ID: <39060734.3A06E7CE@sec.noaa.gov>

I finally discovered my error after carefully re-reading the tutorial on
extending Python; the C function should be:

static PyObject *c_extender_total_it (self, parameter)
    PyObject *self;
    PyObject *parameter;

j vickroy wrote:

> Hello all,
>
> I am trying to pass a Python list to a C extension module function and
> retrieve a result.  Using Python 1.5.2 on Win NT 4.0, the C function
> generates a memory error when trying to parse () the Python list
> parameter.
>
> Any assistence would be appreciated.
>
> Thanks for your time,
>     - jv
>
> Here is the Python code:
>
> import c_extender
> list = []
> for i in xrange (10):
>     list .append ( float (i) )
>  print 'total:', c_extender .total_it (list)
>
> ... and here is the C function.
>
> #include "Python.h" // Python-C interface definitions
>
> static PyObject *c_extender_total_it (parameter)
>     PyObject *parameter;
>      /*
>      */
>     {
>     int i;
>     float result;
>     double *array;
>     double c_double;
>     PyObject *python_list;
>
>     printf ("\t <executes this statement> \n");
>     // interpret parameter as a Python list
>         if (! PyArg_ParseTuple (parameter, "O!", &PyList_Type,
> &python_list) ) return NULL;
>     // *** memory exception reported here ***
>     printf ("\t <does NOT execute this statement> \n");
>
>     // perform calculation here
>
>     free (array);
>     return Py_BuildValue ("total", result);
>     }



From dante at oz.net  Wed Apr 26 02:45:54 2000
From: dante at oz.net (Michael Esveldt)
Date: 26 Apr 2000 06:45:54 GMT
Subject: execl problem on linux
Message-ID: <8e63b2$6v2$0@216.39.162.232>

This may be bordering on off topic but the following is happening with 
all the exec() methods...

Python 1.5.2 (#0, Apr  3 2000, 14:46:48)  [GCC 2.95.2 20000313 (Debian 
GNU/Linux
)] on linux2
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> import os
>>> os.execl("/usr/bin/vim")
Internal error: lalloc(0, )
Internal error: lalloc(0, )
bash-2.03$

I'm at a loss, any ideas?
_______________________________________________________________
dante at oz.net - Michael Esveldt - #FightThePower on Openprojects


From m.faassen at vet.uu.nl  Wed Apr 26 14:07:44 2000
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 26 Apr 2000 18:07:44 GMT
Subject: Simple del[] list question
References: <meh9-9D8292.13251626042000@news.cit.cornell.edu>
Message-ID: <8e7b9g$r2s$1@newshost.accu.uu.nl>

Matthew Hirsch <meh9 at cornell.edu> wrote:
[snip problem description]
> Why did a change?  Why suddenly when you go to two dimensions does the 
> behavior change?

It's because [:] only does a shallow copy; that is, it copies the list,
the individual elements are *not* copied; only new references are made.
So when you delete an element of a sublist, both lists will now have 
a reference to the same sublist which now misses an element.

Python is really very consistent about this, but tricky sometimes, indeed.

In order to make a deep copy of the list, use the deepcopy() function from
the standard copy module.

Note that another strategy which tends to avoid these troubles is to
construct a new list if you modify one. That is, usually if you're going
to alter some list, you're going to alter multiple elements, or perhaps
not include multiple elements, and not just a single one. In that case it's
easier and possibly even faster to do something like this:

mylist = [...] 
newlist = []
for element in mylist:
    if some_criterion(element):
        newlist.append(element)
mylist = newlist

By constructing a new list, you avoid the potential problems of some other
variable still pointing to your modified list.

Regards,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From mhammond at skippinet.com.au  Thu Apr  6 20:02:14 2000
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Fri, 07 Apr 2000 00:02:14 GMT
Subject: Problem calling COM Object written in Python
References: <8cin8n$dhk$1@nnrp1.deja.com>
Message-ID: <aA9H4.35$4v.286@news-server.bigpond.net.au>

<rgb122 at my-deja.com> wrote in message
news:8cin8n$dhk$1 at nnrp1.deja.com...
> site = Publisher.loadSite(str(filename))
> return site

What is "site" here?  It sounds like site is a tuple, so the COM
framework is attempting to convert your return value (site) into
a safearray.  You are probably expecting site to be a string or
other simple type?

> The PythonWin Trace Collector gives this friendly output:
>
> Object with win32trace dispatcher created (object=None)
> in _GetDispID_ with LoadSite 8
> in _InvokeEx_ with 1000 1033L 3
> (L'D:\\rbender\\Desktop\\pyTemplate\\brands.xml',) None
> <PyIServiceProvider at 0x210ebfc with obj at 0x202f130>

The fact that no traceback is printed indicates that the error
was not raised by Python code, byt by the PythonCOM framework on
the way out of the function - ie, your function has correctly
returned to the PythonCOM framework, but the error occurs in
re-packing the COM objects from the Python return value.

> Is there some sort of problem passing the string as a
parameter? I'm
> running this under Win2000, so perhaps there's a lovely new
"feature"
> that's cropped up.

Ive been on Win2000 for some time, so its not related to that...

Mark.





From george_planansky at harvard.edu  Wed Apr  5 13:41:59 2000
From: george_planansky at harvard.edu (George Planansky)
Date: 05 Apr 2000 13:41:59 -0400
Subject: ataman rsh windows_pc hello.py fails
Message-ID: <w3yem8k4f88.fsf@gigue.eas.harvard.edu>

Using ataman telnet and rsh to connect to an
nt 4.0 sp5 pc from solaris,  I can get back 
the output from a windows "dir" command, but 
not from a "hello.py" script.

How do I fix this?

If I login in to the pc via "solaris% telnet ataman_pc",
hello.py works fine:

  d:\home\atadmin>hello.py
  hello

And rsh from the sun, invoking "dir"  works fine:
  
   solaris% rsh ataman_pc dir 
  
   Volume in drive D has no label.
   Volume Serial Number is 04B8-D727

   Directory of d:\home\atadmin
   ....

But not rsh with hello.py"

   solaris% rsh ataman_pc hello.py

   <I get nothing displayed back from the pc.>

However, the python scripts do in fact execute via rsh, 
since e.g., with hello2.py

   import sys
   mylogfile = open('d:\\home\\atadmin\\hello2.txt', 'w')
   sys.stdout = mylogfile
   print 'hello2'

The invocation:

   solaris% rsh atamin_pc hello2.py

does indeed create a hello2.txt on the pc,
with contents "hello2".

Ataman support suspects:

  The difference between telnet and rsh (besides the obvious 
  ones) is that rsh just connects pipes to stdin and stdout. 
  Some programs detect that and behave somewhat differently.

Anyone know how to make the python script output
act like dir's output?

george_planansky at harvard.edu


From fw at deneb.cygnus.argh.org  Tue Apr 11 13:47:49 2000
From: fw at deneb.cygnus.argh.org (Florian Weimer)
Date: 11 Apr 2000 19:47:49 +0200
Subject: [ANNOUNCE] Garbage collection for Python
References: <20000407221750.A29279@acs.ucalgary.ca> <38F13FB9.D1D16AF@mindspring.com> <slrn8f2if1.dop.nascheme@cranky.arctrix.com> <87r9cdhdb0.fsf@deneb.cygnus.argh.org> <FsunGI.BtB@world.std.com> <dphfd87s1h.fsf@mraz.iskon.hr> <Fsuuw7.Inu@world.std.com> <dpk8i4646r.fsf@mraz.iskon.hr>
Message-ID: <87em8ccywq.fsf@deneb.cygnus.argh.org>

Hrvoje Niksic <hniksic at iskon.hr> writes:

> > Are we talking about multiple Python instances running in a single
> > thread?
> 
> No, we're talking (I think) about multiple Python instances running in
> a single process, each in its own thread.

Yes.  AFAIK, Python doesn't perform well in a SMP environment, because
the interpreter only runs on one CPU at a given time, even if you
create multiple Python threads.


From effbot at telia.com  Tue Apr  4 18:27:27 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Tue, 04 Apr 2000 22:27:27 GMT
Subject: ANN: argumentclinic.py, version 1
Message-ID: <j%tG4.1350$n68.194631680@newsb.telia.net>

#! /usr/bin/env python

# Released as 'checkappend.py' to the public domain, by Tim Peters, 28
# February 2000.

# With a little luck, the eff-bot managed to turn this into something
# that looks for a few more deprecated calls...

"""argumentclinic.py -- search for deprecated multi-argument calls.

Usage:  specify one or more file or directory paths:
    argumentclinic [-v] [-m method] file_or_dir [file_or_dir] ...

Each file_or_dir is checked for commonly used but deprecated multi-
argument method calls, such as append and connect.  When a directory,
all .py files in the directory, and recursively in its subdirectories,
are checked.

Use -v for status msgs.  Use -vv for more status msgs.

Use -m to look for just one given method.  If omitted, the program
looks for append, bind, connect, connect_ex, count, has_key, and
remove.  You can use -m multiple times, to look for a bunch of methods
of your own choice.

In the absence of -v, the only output is pairs of the form

    filename(linenumber):
    line containing the suspicious method

Note that this finds multi-argument methods calls regardless of
whether they're attached to the right kind of objects.  If a module
defines a class with e.g. an connect method that takes more than one
argument, calls to that method will be listed.

Note that this will not find e.g. multi-argument list.append calls
made via a bound method object.  For example, this is not caught:

    somelist = []
    push = somelist.append
    push(1, 2, 3)
"""

__version__ = 1, 0, 0

import os
import sys
import string
import getopt
import tokenize

verbose = 0

methods = (
    "append", "count", "remove", # lists
    "has_key", # dictionaries
    "bind", "connect", "connect_ex" # sockets
    )

def errprint(*args):
    msg = string.join(args)
    sys.stderr.write(msg)
    sys.stderr.write("\n")

def main():
    args = sys.argv[1:]
    global verbose, methods
    try:
        opts, args = getopt.getopt(sys.argv[1:], "vm:")
    except getopt.error, msg:
        errprint(msg + "\n\n" + __doc__)
        return
    for opt, optarg in opts:
        if opt == '-v':
            verbose = verbose + 1
        elif opt == '-m':
            try:
                methods.append(optarg)
            except AttributeError:
                methods = [optarg]
    if not args:
        errprint(__doc__)
        return
    for arg in args:
        check(arg)

def check(file):
    if os.path.isdir(file) and not os.path.islink(file):
        if verbose:
            print "%s: listing directory" % `file`
        names = os.listdir(file)
        for name in names:
            fullname = os.path.join(file, name)
            if ((os.path.isdir(fullname) and
                 not os.path.islink(fullname))
                or os.path.normcase(name[-3:]) == ".py"):
                check(fullname)
        return

    try:
        f = open(file)
    except IOError, msg:
        errprint("%s: I/O Error: %s" % (`file`, str(msg)))
        return

    if verbose > 1:
        print "checking", `file`, "..."

    ok = ArgumentChecker(file, f).run()
    if verbose and ok:
        print "%s: Clean bill of health." % `file`

[FIND_DOT,
 FIND_METHOD,
 FIND_LPAREN,
 FIND_COMMA,
 FIND_STMT]   = range(5)

class ArgumentChecker:
    def __init__(self, fname, file):
        self.fname = fname
        self.file = file
        self.state = FIND_DOT
        self.nerrors = 0

    def run(self):
        try:
            tokenize.tokenize(self.file.readline, self.tokeneater)
        except tokenize.TokenError, msg:
            errprint("%s: Token Error: %s" % (`self.fname`, str(msg)))
            self.nerrors = self.nerrors + 1
        return self.nerrors == 0

    def tokeneater(self, type, token, start, end, line,
                NEWLINE=tokenize.NEWLINE,
                JUNK=(tokenize.COMMENT, tokenize.NL),
                OP=tokenize.OP,
                NAME=tokenize.NAME):

        state = self.state

        if type in JUNK:
            pass

        elif state is FIND_DOT:
            if type is OP and token == ".":
                state = FIND_METHOD

        elif state is FIND_METHOD:
            if type is NAME and token in methods:
                self.line = line
                self.lineno = start[0]
                state = FIND_LPAREN
            else:
                state = FIND_DOT

        elif state is FIND_LPAREN:
            if type is OP and token == "(":
                self.level = 1
                state = FIND_COMMA
            else:
                state = FIND_DOT

        elif state is FIND_COMMA:
            if type is OP:
                if token in ("(", "{", "["):
                    self.level = self.level + 1
                elif token in (")", "}", "]"):
                    self.level = self.level - 1
                    if self.level == 0:
                        state = FIND_DOT
                elif token == "," and self.level == 1:
                    self.nerrors = self.nerrors + 1
                    print "%s(%d):\n%s" % (self.fname, self.lineno,
                                           self.line)
                    # don't gripe about this stmt again
                    state = FIND_STMT

        elif state is FIND_STMT:
            if type is NEWLINE:
                state = FIND_DOT

        else:
            raise SystemError("unknown internal state '%s'" % `state`)

        self.state = state

if __name__ == '__main__':
    main()

# --- end ---




From nolan at ethereal.dhis.org  Sun Apr 30 19:43:16 2000
From: nolan at ethereal.dhis.org (Nolan Darilek)
Date: 30 Apr 2000 18:43:16 -0500
Subject: ZODB troubles
Message-ID: <87pur7jgtn.fsf@ethereal.dyndns.org>

For hack value, I've semi-revived some old code which I wrote and have
been attempting to rewrite it to work with ZODB.

I'm trying to create an environment in which python code is stored
alongside objects. So, you'd have a database of objects which could be
used, and you'd be able to 'dump' their code to text. But, I've been
having some ZODB issues and was hoping that someone could help.

I'm writing small test scripts to ensure that my code is behaving the
way it should. So far I have three tests:

1.  Open the ZODB and create a test object, closing it again.
2.  Re-open the ZODB, create a persistent function' and run it to
prove that it works.
3.  Open the ZODB, access the object, run the persistent function and
print the code.

Test 1 works flawlessly. Test 2 runs, but in order to save the
newly-added function, I have to delete the object from the database
and re-insert it; it can't be modified live it seems. Test 3 works
depending on which route is taken for 2; the long "delete and re-add
the object" way works, but the shorter "add an instance attribute to
the object and commit" doesn't.

I tried duplicating this problem on ints and other simple types, but I
couldn't. It only seems to appear under my circumstances. So, in order
to help you help me, I've flattened my hierarchy of scripts into a
single Python script which requires the ZODB modules. I have three
functions entitled one()/three(). Since two() determines whether this
script fails or succeeds, I've added a few instructions on how to
switch between the success and failure cases. To run this, 'python
test.py 1', or 2-3 to run the other tests. They should be run in
order, since they lay groundwork for the next test.

Another question; is it all right to embed
get_transaction().commit()'s (or commit(1)'s) deep within my code, in
places where fundamental changes are made? (See Function.compile() for
an example of what I mean.

I'm inserting the script below. If anyone could help with this, I'd
greatly appreciate it. Thanks.

--------------------------------------------------------------------------------
#!/usr/bin/env python

import Persistence
import re
import rexec
import sys
from ZODB import *
from ZODB.FileStorage import *

class Database:

    """Simple wrapper around ZODB.

    This class provides a simple wrapper around the lower-level ZODB functions.
    Some functionality is not yet implemented.
    """

    def __init__(self, file='Data.fs'):
        """ Create a database, connect to it and get the root
        object out of it.  """
        self.db = DB(FileStorage(file))
        self.connection = self.db.open()
        self.root = self.connection.root()

    def addObject(self, object):
        """Add an object to the database."""
        dbref = `len(self.root.keys())`
        while self.root.has_key(dbref):
            dbref = `int(dbref)+1`
        self.root[dbref] = object
        get_transaction().commit()
        return dbref

    def dump(self):
        """Commit all transactions."""
        get_transaction().commit()

    def __getitem__(self, item):
        return self.root[item]
    

class Function(Persistence.Persistent, rexec.RExec):

    """This is a 'persistent function.'

    A persistent function acts like a regular function, but code exists
    alongside compiled code. This also neatly handles recompiling functions,
    since the function objects don't play nice with ZODB/Pickle.
    """

    def __init__(self, code = ""):
        """Initialize a new instance.
        
        If `code' is specified, compile it.
        """
        self.code = code
        if code != "":
            self.compile(code)

    def __call__(self, *args):
        """Execute the function code.

        If specified, `args' are passed directly to the function code.
        """
        # setstate() can't seem to find self.code.
        # Since the actual function code objects don't persist, the uncompiled
        # code is kept until the function is initially called, then we compile.
        try:
            self._v_func
        except AttributeError:
            self.compile(self.code)
        apply(self._v_func, args)

    def compile(self, code):
        """Compiles the function.

        `code' is a list of strings consisting of this function's Python code.
        """
        self.code = code
        exec(code)
        counter = 0
        match = re.match("def\s+(.*?)\s*\(.*\):", self.code)
        if not match:
            return
        fname = match.string[match.regs[1][0]:match.regs[1][1]]
        exec("self._v_func = %s" % fname)
        # Will this commit the modified function?
        # Is it a good idea to nestle commits within places like this where
        # changes are made?
        get_transaction().commit()

    def __repr__(self):
        return self.code

class Object(Persistence.Persistent):

    """A persistent object with some changes.

    This acts like a normal object with a notable exception: all attributes
    are stored in 'Attribute' objects, with setattr/getattr handling access.
    This will allow certain attributes to be public, private, etc.
    """

    def __init__(self):
        self.__dict__["__attributes__"] = {}

    def __setattr__(self, name, value):
        self.__attributes__[name] = value
        get_transaction().commit()

    def __getattr__(self, name):
        return self.__attributes__[name]
    
def one():
    print "Test 1: Initializing a new database and adding test object 0."
    db = Database()
    t = Object()
    db.addObject(t)
    print "Done!"

def two():
    print "Test 2: Retrieving test object 0 and adding a function."
    db = Database()
    # t = db.root['0']
    # t.hw = Function("def hw():\n print 'Hello, world.'\n")
    # del db.root['0']
    # db.addObject(t)
    # Uncomment the previous 4 lines and comment out the next to fix this.
    db.root['0'].hw = Function("def hw():\n print 'Hello, world.'\n")
    # As you can see, the previous line is much more aesthetically pleasing.
    # How can I get it to do what the above 4 lines do?
    db.root['0'].hw()
    db.dump()
    print "Done!"

def three():
    print "Test 3: Retrieving test object and function."
    db = Database()
    db.root['0'].hw()
    print db.root['0'].hw
    print "Done!"

if sys.argv[1] == '1':
    one()
if sys.argv[1] == '2':
    two()
if sys.argv[1] == '3':
    three()
--------------------------------------------------------------------------------


From jbauer at rubic.com  Mon Apr 24 15:31:21 2000
From: jbauer at rubic.com (Jeff Bauer)
Date: Mon, 24 Apr 2000 14:31:21 -0500
Subject: A Mountain of Perl Books + Python Advocacy
References: <20000405053920.20579.qmail@web2102.mail.yahoo.com> <38ED40B0.A48EC4CC@udel.edu> <tNvI4.18$_N4.1751@vic.nntp.telstra.net> <8e24m5$2vn$1@nnrp1.deja.com>
Message-ID: <3904A109.31848817@rubic.com>

sp00fD wrote:
> I used perl formerly and found it to be
> sooo easy (partially because I'm a unix 
> admin and perl is very unix-like), there 
> were wonderful examples everywhere on how 
> to do things.

Why did you quit using Perl, then?  (This isn't
meant to be sarcasm. It's just that there are
plenty of opportunities available to continue 
coding in Perl.)

> Python, on the other hand, has __extremely__ 
> poor documentation (as do most languages), 
> and I believe this is a large part of why 
> perl has become so popular.  Most people 
> can't discuss theory and get coding ideas 
> from it, nor can they see a method and 
> parameters and know exactly how to use the 
> module correctly.  Frankly, the stock python
> documentation is worthless to me, and often 
> times I find myself perusing modules, without 
> luck because unfortunately most of them are
> not very well commented either.

Can you provide an example of a problem you were
unable to solve due to the poor documentation?

For the record, I used Perl before there was
any decent documentation (before the first
O'Reilly book).  Due to its similarity to
C, it was reasonably easy to figure stuff
out from the Perl man page.  However, that's
because I was already programming in C.

Even back in 1995, I found Python documentation 
to be well-written and more than adequate to 
for the tasks at hand, which included embedding 
and extending the interpreter, writing large-scale 
applications, interfacing with 3rd-party database 
engines, etc.  Also, in my experience, it has been 
easier to train non-C/Unix-geek types to program 
Python rather than Perl, though neither have 
particularly high barriers (IMO).

If you require additional assistance beyond
the stock documentation in the distribution, 
there are plenty of sources worth checking out:

http://www.python.org/psa/bookstore/

Hope this helps!

Jeff Bauer
Rubicon Research



From martink at cix.co.uk  Tue Apr  4 07:55:00 2000
From: martink at cix.co.uk (Martin Keefe)
Date: Tue, 4 Apr 2000 12:55 +0100 (BST)
Subject: Passing array of doubles to COM object?
Message-ID: <memo.20000404125544.208A@martink.compulink.co.uk>

I am trying to use an Automation server that has the following method (among 
others):

  VARIANT_BOOL LoadFeature(
                            BSTR code, 
                            BSTR label, 
                            long size, 
                            VARIANT xArray, 
                            VARIANT yArray);

Its documentation assumes that the client will be written in VB and says that 
xArray and yArray are arrays of Doubles. And true enough, passing arrays of 
Doubles from VB works fine.

In Python, after creating the object with:

 mx = win32com.client.Dispatch("Maplex.Application")

I've tried the following:

 xa = [100000.0, 600000.0, 600000.0, 100000.0]
 ya = [100000.0, 100000.0, 900000.0, 900000.0]
 mx.LoadFeature( 'AREA', 'Simple', 4, xa, ya )

and:

 xa = buffer( array('d',[100000.0, 600000.0, 600000.0, 100000.0]) )
 ya = buffer( array('d',[100000.0, 100000.0, 900000.0, 900000.0]) )
 mx.LoadFeature( 'AREA', 'Simple', 4, xa, ya )

without success. Both ran without errors or exceptions but the data didn't 
appear to get to the server intact. I've tried both dynamic and static dispatch 
with the same results. Can anybody shed some light on this please?

The system is NT4 Server SP3, Python 1.5.2, win32all-128. Thanks,

  ]\/[artin Keefe ][ Dorking,Surrey,UK ][ martink at cix.co.uk



From gmcm at hypernet.com  Wed Apr  5 10:41:21 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Wed, 5 Apr 2000 10:41:21 -0400
Subject: Exceptions, when and who?
In-Reply-To: <8cffin$pl0$1@nnrp1.deja.com>
Message-ID: <1257176800-9949457@hypernet.com>

Arint?  wrote:

> When and why would python could throw an exception in an embedded app?

For any of the reasons that non-embedded Python might 
throw an exception.

> I am not getting the exception on the c side but in the python code and
> I can't figure out why.
> 
> If I set the Error String in the c code when and who should clear that
> error?  How would you clear it in python, in c I know you use
> PyErr_Clear (right?).

Catch the exception.

- Gordon



From tim_one at email.msn.com  Sat Apr  1 00:29:05 2000
From: tim_one at email.msn.com (Tim Peters)
Date: Sat, 1 Apr 2000 00:29:05 -0500
Subject: Does Python have a long floating data type please?
In-Reply-To: <38E4D4F7.DDD56F00@math.okstate.edu>
Message-ID: <000201bf9b9b$32560e80$2f2d153f@tim>

[David C. Ullrich, notes some examples of dubious intervals computed
 by Jurjen N.E. Bos's real.py module]

Yes, and I've posted others in the past (including cases that aren't near
0).

Jurjen used a very clever storage-efficient representation, but didn't
record correspondingly clever correctness arguments <wink>.  Sometimes it's
easy to see a bug, sometimes I have no idea what he was thinking.

This isn't unusual for these kinds of pkgs, though!  Correctness proofs can
be quite involved.  Val?rie M?nissier-Morain has submitted a 59-page article
giving correctness proofs for her version of this quest, available somewhere
more-or-less obvious under

    http://pauillac.inria.fr/~menissier/


> ... - at this point I decided to give up on real.py and make my own.

Which is why there are a dozen packages like this out in the world each with
its own set of unique subtle bugs <wink>.  Since this stuff runs too slow to
use for truly large problems anyway, I never understood why pkgs try to
conserve storage at the cost of vastly subtler reasoning.  Guess it's seen
as a challenge.

personally-more-excited-by-things-that-are-obviously-right-than-by-
    things-that-aren't-obviously-wrong-ly y'rs  - tim





From kcazabon at home.com  Tue Apr 25 01:57:03 2000
From: kcazabon at home.com (Kevin Cazabon)
Date: Tue, 25 Apr 2000 05:57:03 GMT
Subject: Simple list.append() question
References: <meh9-5ED8CD.11192424042000@news.cit.cornell.edu> <ZHZM4.360$qg7.2949@news.uswest.net> <39046FA2.603EAFE5@roguewave.com>
Message-ID: <PsaN4.174166$Dv1.2059732@news1.rdc1.bc.home.com>

well, if not prettier, why not obfuscate it?  q:]  Although this isn't
pretty, I'm sure we can make it a little less intelligable with some work.

entire_list = [];for i in range(3):entire_list.append([])


Kevin.

"Bjorn Pettersen" <bjorn at roguewave.com> wrote in message
news:39046FA2.603EAFE5 at roguewave.com...
> Jeff Massung wrote:
> >
> > I'm just learning, but my guess is this:
> >
> > >>> l=[[]]*3
> > >>> l
> > [[], [], []]
> > >>> l[0].append(3)
> > >>> l
> > [[3], [3], [3]]
> > >>> id(l[0])
> > 8539088
> > >>> id(l[1])
> > 8539088
> > >>> id(l[2])
> > 8539088
> > >>>
> >
> > The *3 just made three lists pointing to the same place. I don't know
how to
> > get what you want tho ;) someone else can enlighten me, too :).
>
> One way is:
>
> x = map(lambda x:[], [[]]*3)
>
> although there has to be a prettier way...
>
> -- bjorn
>




From bwilk_97 at yahoo.com  Mon Apr 24 18:19:20 2000
From: bwilk_97 at yahoo.com (Bill Wilkinson)
Date: Mon, 24 Apr 2000 22:19:20 GMT
Subject: Simple list.append() question
References: <meh9-5ED8CD.11192424042000@news.cit.cornell.edu> <ZHZM4.360$qg7.2949@news.uswest.net> <39046FA2.603EAFE5@roguewave.com>
Message-ID: <IL3N4.4283$L3.47222@typhoon.midsouth.rr.com>

| One way is:
|
| x = map(lambda x:[], [[]]*3)
|
| although there has to be a prettier way...
|
| -- bjorn
|

l = [[][:]] * 3

:)






From gmcm at hypernet.com  Fri Apr  7 14:05:44 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Fri, 7 Apr 2000 14:05:44 -0400
Subject: Member function names
In-Reply-To: <87vh1t6foz.fsf@womble.dur.ac.uk>
Message-ID: <1256991731-21081482@hypernet.com>

Andrew Stribblehill wrote:

> For some things, such as XML parsers, based on xmllib, I have to do
> stuff like this:
> 
> class foo(XMLParser):
>     start_h1:
>         print "start_h1",
>     end_h1:
>         print "end_h1",
> 
> (etc)
> 
> Is there any way I can do something like this?
> 
> class foo(XMLParser):
>     generic:
>         print __member_function_name__,
>     start_h1=generic
>     end_h1=generic
> 

When everything about your example code is broken, it's 
difficult to know what broke on your real attempts. In general, 
this technique works fine.

>>> class A:
...  def a(self):
...   print "hi"
...  b = a
...
>>> a = A()
>>> a.b()
hi
>>>

- Gordon



From claird at starbase.neosoft.com  Thu Apr 20 09:49:19 2000
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 20 Apr 2000 08:49:19 -0500
Subject: Building python on HPUX-10.2
References: <38FE1392.D1C01D64@wag.caltech.edu> <38fe34a4.1103181328@news1.on.sympatico.ca>
Message-ID: <53EFCD0D35A4C170.C1B237DE5B166A8A.ECBDC4EFB1E450CA@lp.airnews.net>

In article <38fe34a4.1103181328 at news1.on.sympatico.ca>,
Robert Roy <rjroy at takingcontrol.com> wrote:
>I have built python successfully on several HPUX 10.2 boxes using GCC.
>
>If you are using the basic cc compiler, you might have problems. If
>you have the full development version, I don't know. On one machine I
>had to install GCC, (hmmm that might have been because of one of the
>libraries I was linking now that I think if it.) GNU make is also a
>good idea...
>
>Another gotcha is that once compiled, it might fail the fcntl tests.
>see
>http://x22.deja.com/[ST_rn=ps]/getdoc.xp?AN=517003463&CONTEXT=956184259.663814174
>for how to fix this
			.
			.
			.
It *is* possible to generate a correct Python with the bundled
HP-UX compiler.  There are a lot of problems with that compiler,
though, and majority sentiment is that it's easier just to install
GCC.  If there's sufficient interest, I can unpack my notes on the
subject.  I generally wait for provocation before doing so, as I
don't have convenient access to HP-UX these days.
-- 

Cameron Laird <claird at NeoSoft.com>
Business:  http://www.Phaseit.net
Personal:  http://starbase.neosoft.com/~claird/home.html


From arnold at dstc.edu.au  Thu Apr  6 20:58:45 2000
From: arnold at dstc.edu.au (David Arnold)
Date: Fri, 07 Apr 2000 10:58:45 +1000
Subject: C's ?: expression 
In-Reply-To: Your message of "Thu, 06 Apr 2000 20:29:15 -0400."
             <000801bfa028$4eaa8fc0$34a0143f@tim> 
Message-ID: <200004070058.KAA30275@piglet.dstc.edu.au>

-->"Tim" == Tim Peters <tim_one at email.msn.com> writes:

  Tim> However, I have never used it in real code, and have been known
  Tim> to kill people who do.

i find it useful with the functional stuff (map, filter, reduce), but
then your comment about killing people who use this in real code can
often apply there too ...



d



From aek at mail1.stofanet.dk  Tue Apr  4 19:49:56 2000
From: aek at mail1.stofanet.dk (Anders Eggers-Krag)
Date: Wed, 5 Apr 2000 01:49:56 +0200
Subject: Two questions
References: <38EA5FE8.853E9B75@callware.com> <38EA6508.90BF630C@roguewave.com> <38EA6E87.40EDB542@callware.com> <38EA6422.61054B11@exceptionalminds.com> <38EA7452.9BFDF37E@callware.com> <8cdt7c$cl1$1@news101.telia.com> <38EA7D9C.7024CF0B@roguewave.com> <0 <38EA8076.D8385E18@roguewave.com>
Message-ID: <01b601bf9e90$7dfb1770$2c6ba8c0@frodo>

> > > > an alternative solution is to call an ADO (ActiveX Data Object) object and
> > > > use it through
> > > > there which is quite simple.
> > >
> > > If you do this, make sure you get Mark's programming win32 book.  It's
> > > an excellent book for many reasons, but it goes into special detail on
> > > COM programming (and is so far the only resource I've seen of how COM
> > > can be simple and useful -- instead of the normal
> > > add-on-later-because-we-need-the-buzzword I normally associate with COM
> > > <wink>).
> > >
> > it is not like ADO is a complex system, it is extremely easy to use, I use it alle time
> > writing database systems for ASP.
> 
> Oh, I completely agree... However, if you don't know COM, or only know
> it through C++, I would still recomend getting the win32 book.
> 

while we are at it - how do I instantiate an Activex/COM object in python,
I use them plentily in ASP and HTA where it is quite easy, but I have only 
yesterday or so started working with Tkinter, so now I no longer have the easy
methods suchs as new ActiveXObject("..."), or Server.CreateObject("..."),
I looked at it a while ago in Python and decided to wait until I needed it :-)

Anders Eggers - Krag




From matthias_w at my-deja.com  Fri Apr 28 12:45:25 2000
From: matthias_w at my-deja.com (Matthias)
Date: Fri, 28 Apr 2000 16:45:25 GMT
Subject: win32all-131 at python.org
References: <1UDL4.539$v85.2477@news-server.bigpond.net.au>
Message-ID: <8ecf70$766$1@nnrp1.deja.com>

Hy Mark :-)
Are the sources anywhere ??
(You told me to include the new startup-code in my project)

BTW
I have the following problems:
(1) I want to use python-win for the interface in my programm, but it
is most important to hide the interactive windows for most of the users
(they should not know that they are powered by python), and
(2) I have a conflict with my main windows. In the InitInstance() from
my application the pyglue-code creates a new window and now visualc
does not know which of them is the main window. (e.g. I open
splittst.py in my window, have to choose debug (go is disabled ??) from
the python-window and the splitter - test window is a mdi-child in my
own main window again (little confuseing, isn't it ;-) - the menus are
quite the same, but in my window all python-related menus are diasbled)

Thank you in advance
Matthias :->

In article <1UDL4.539$v85.2477 at news-server.bigpond.net.au>,
  "Mark Hammond" <mhammond at skippinet.com.au> wrote:
> Just a quick update.  win32all-131 for Python 1.6a2 can be found at
> http://www.python.org/1.6/
>
> Note that this version has 2 known (and semi-serious) bugs:
> * The Pythonwin debugger simply doesnt work (but its such a simple
bug!)
> * A Unicode issue with Scintilla prevents calltips and multi-line
commands
> in the Pythonwin interactive window(!!) working correctly on NT/2000.
>
> I expect to have 132 available soon with these fixes plus a few other
> tweaks.  If starship is still not available, it will also be at
> http://www.python.org/1.6/
>
> Finally, once this win32all set stabilizes, I will also be releasing
> another Python 1.5 build.  Depending on what happens, this may or may
not
> be the last 1.5 build, but most new builds are likely to target 1.6
> (although "popular demand" is likely to have some influence ;-)
>
> Mark.
>
>


Sent via Deja.com http://www.deja.com/
Before you buy.


From deepak_marda at my-deja.com  Sat Apr 29 03:25:02 2000
From: deepak_marda at my-deja.com (deepak_marda at my-deja.com)
Date: Sat, 29 Apr 2000 07:25:02 GMT
Subject: COMObject / Zope Question
Message-ID: <8ee2o8$vr7$1@nnrp1.deja.com>

Hi,

I have been able to create and connect a COMObject on Zope
with id = excel and dispatch = Excel.Application

How can I use this object to change the cells?
The following code works...
<dtml-in excel().RecentFiles>
<dtml-var Name>
</dtml-in>

How do I access the COMObject (excel) from a python
External Method? I tried to access it as self.excel
(where self is the folder object and it is same for
the external method). However, if I tried to access
self.excel.RecentFiles it raises an exception for invalid
attribute "RecentFiles"

How can I access the methods and properties for the
COM object. An example for modifying a cell value would
be useful...

Thanks,
Deepak



Sent via Deja.com http://www.deja.com/
Before you buy.


From arcege at shore.net  Wed Apr 19 08:09:38 2000
From: arcege at shore.net (Michael P. Reilly)
Date: Wed, 19 Apr 2000 12:09:38 GMT
Subject: testing true condition
References: <Pine.BSF.4.10.10004191138390.292-100000@hercule.perceval.be> <slrn8fr775.dvd.scarblac-spamtrap@flits104-37.flits.rug.nl>
Message-ID: <6mhL4.579$lL2.69807@news.shore.net>

Remco Gerlich <scarblac-spamtrap at pino.selwerd.nl> wrote:
: Gregoire Welraeds wrote in comp.lang.python:
:> I know that I can't do something like the following:
:> 
:> >>> while (line= f.readline()):
:> 	...
:> 
:> as i could in C.  Is there any replacement tricks ?

: The standard Python idiom is:

: while 1:
:    line = f.readline()
:    if not line:
:       break

This is only one of the two, simple, "standard" idioms, the other is:

>>> line = f.readline()
>>> while line:
...   ...
...   line = f.readline()
...

(Many ppl do not like, with reasons, the `if not line: break'
construct, but all that is in past debates on this NG available
somewhere on dejanews.com.)

Since the original poster also had a fear of reading in the entire
file, the readlines() method can also be used with improved performance
in a slightly more complex fashion:

>>> for list_of_lines in f.readlines(16384):  # read in 16k at a time
...   for line in list_of_lines:
...     ...
...

This is more efficient when you have a large text file to read, but do
not want to read the entire file.  Sets of pages are read in at a time
and split into the lines.  The size passed to readlines may need to be
adjusted on the end machine.  The fileinput module will not be as
efficient (at least in 1.5.2, not sure about 1.6a).

  -Arcege



From raoul at ez-ways.com  Thu Apr 20 16:00:50 2000
From: raoul at ez-ways.com (Raoul-Sam Daruwala)
Date: Thu, 20 Apr 2000 20:00:50 GMT
Subject: Python 1.6 and Python 3000's new features?
References: <6D8A17398E28D3119F860090274DD7DB498538@pces.cadlab.it> <1Ivx4.153$PI5.135469056@newsb.telia.net>
Message-ID: <38FF64F4.71892A75@ez-ways.com>

will python3000 have proper scoping? ie. will code like

y = 5
f = lambda x: x * y
t = f(3)

work? or will we still have to play inane games and use

f = lamdba x,y=y:x*y

which, if I'd known about it in time, would have sent me fleeing from python
altogether.

r-s

Fredrik Lundh wrote:

> Alessandro Bottoni wrote:
> > [Python 3000] The contents of this version are still under discussion, but
> > will almost certainly be backward-incompatible.
>
> > [Python 1.6] In addition, Guido will include some existing bug fixes and
> > probably Fredrik Lundh's new regular expressions engine
> >
> > Why Python 3000 should be backward-incompatible? Is it possible to have a
> > "big-picture" of this major version?
>
> spring cleaning.  get rid of most warts, add typing, clean
> up the (rather messy, by now) standard library.  etc.
>
> > Why new regular expressions? What was wrong with the "old" re module?
>
> it doesn't support unicode.
>
> </F>



From m.faassen at vet.uu.nl  Thu Apr 27 17:45:19 2000
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 27 Apr 2000 21:45:19 GMT
Subject: Do I always have to write "self." ?
References: <270420001706386648%pecora@anvil.nrl.navy.mil>
Message-ID: <8eacdf$nnm$1@newshost.accu.uu.nl>

Louis M. Pecora <pecora at anvil.nrl.navy.mil> wrote:
> I am two weeks plus into learning Python and have started on a first
> real project.  Amazing language.  BUT, now that I've gotten going with
> classes, I'm finding I hate seeing all those  "self."  prefixes on the
> class variables.  Python is generally very clean, but the classes I
> write use a lot of "per-instance" variables, hence seem to require
> self.  As a result the expressions can become cluttered and harder to
> read than C++ in this particular case -- imagine that!  Is there any
> relief here?  Any (simple) way around writing that self. every time?

There's a way, but stuff can also become more cluttered. In some
cases it'll also be a bit faster, as local variable access is
faster than access through self. In lots of other cases it can be slower.
I'll demonstrate:

class Foo:
    def __init__(self):
        self.data = []
        self.more = 10
        self.interesting = "foo"
	
    def involved_method(self, something):
        data, more, interesting = self.data, self.more, self.interesting

        if more > 15:
            data.append(something)
            data.append(interesting)

But note this!

    def uhoh_method(self):
        data, more, interesting = self.data, self.more, self.interesting
        # this changes self.data, neat
        data.append('foo') 
        # this won't change self.more!
        more = more + 1
        # and this won't change self.interesting
        interesting = interesting[10:]

So, beware when you want to change attributes that refer to immutable
objects. You do have to use self in that case. I think that you'll
find that eventually you'll get used to 'self', and you may even
be grateful for its presence. At least you can see instantly where
your variable are coming from. The local variable trick, used in moderation,
can help some as well.

Regards,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From greg at cosc.canterbury.ac.nz  Thu Apr 27 21:18:36 2000
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Fri, 28 Apr 2000 13:18:36 +1200
Subject: Simple list.append() question
References: <meh9-5ED8CD.11192424042000@news.cit.cornell.edu> <ZHZM4.360$qg7.2949@news.uswest.net> <39046FA2.603EAFE5@roguewave.com> <PsaN4.174166$Dv1.2059732@news1.rdc1.bc.home.com> <8e4akc$p7o$1@newshost.accu.uu.nl> <m366t6ozqx.fsf@greebo.nodomain.de>
Message-ID: <3908E6EC.B9E6B2E8@cosc.canterbury.ac.nz>

Bernhard Herzog wrote:
> 
> It would be a good idea IMO to find a
> good syntax for this case and ISTM that list comprehensions might be
> useful here:
> 
>   multi_list = [[] for i in range(10)]
> 
> or whatever the precise syntax was. Hmm, the unused variable in there is
> a bit ugly though.

Maybe the LC syntax should include something like

   [[] times 10]

-- 
Greg Ewing, Computer Science Dept,
+--------------------------------------+
University of Canterbury,	   | A citizen of NewZealandCorp, a	  |
Christchurch, New Zealand	   | wholly-owned subsidiary of USA Inc.  |
greg at cosc.canterbury.ac.nz	   +--------------------------------------+


From bjorn at roguewave.com  Tue Apr  4 19:34:48 2000
From: bjorn at roguewave.com (Bjorn Pettersen)
Date: Tue, 04 Apr 2000 17:34:48 -0600
Subject: Two questions
References: <38EA5FE8.853E9B75@callware.com> <38EA6508.90BF630C@roguewave.com> <38EA6E87.40EDB542@callware.com> <38EA6422.61054B11@exceptionalminds.com>
Message-ID: <38EA7C18.8F3150D1@roguewave.com>

Timothy Grant wrote:
> 
> Ivan Van Laningham wrote:
> > Well, that's two recommendations.  And judging from previous Marc-Andre
> > packages I've used, it'll be a snap.
> >
> > Steve Holden, in another post, said that Tkinter should work fine along
> > with mxODBC.  That's a relief, but does anyone have direct personal
> > experience with doing so?
> 
> Ivan,
> 
> mxODBC works beautifully, but I've never used it to access and access
> database, only SQL database. Perchance do you know how Windows
> advertises an Access database to an ODBC driver?
[snip]

In the control panel there should be a Odbc 32 control (should be
installed with ms access).  It has a simple point-and-click interface
where you have to choose the ms access driver, give the filename for the
database, and choose a DSN (Data Source Name?).  The DSN is what you use
to connect with.

I'll-disavow-knowing-anything-about-databases'ly y'rs
-- bjorn



From gklix at hassler.de  Sun Apr 30 16:12:03 2000
From: gklix at hassler.de (Gerald Klix)
Date: Sun, 30 Apr 2000 22:12:03 +0200
Subject: Event triggering and weak references
References: <8dbif3$hm0$1@metro.ucc.usyd.edu.au>
Message-ID: <390C9393.96622374@hassler.de>

Ricardo Nogueira wrote:

> First thanks for this beautiful language : )
>
> I am Smalltalk addicted : )  and this means that I can't do with any other
> language, ex. C, C++, Java,  : (
> This was true until I found Python 2 weeks ago at Sydney Uni. I loved its
> modularity, pluggin behaviour and freedom.
> But I am missing Smalltalk event handling... Smalltalk (like Python) support
> the call-Back mechanism, but there is a more powerful inter-object interest
> registering that I haven't found in Python (so far).
>
> This is what I would like to do:
> Imagine I could have two or more Widgets open displaying a car speed value,
> say a numerical and a graphic display, if this car speed changes (ex. some
> one pushes the break), I would like to have any open widget on this value
> updated (on opening each window would register interest with this car
> speed).
>
> Smalltalk implements this with:
>
> The widget would register interest in one aspect of the car  (on opening
> over aCar):
>   aCar when:#speedChanged send: #updateSpeedDisplay to: self
>
> And the car would just notify who is interested, if any (when its speed
> changes):
>   self trigger: #speedChanged
>
> But if any of this windows is not referenced anymore or crashes I would like
> the garbage collector clear it (the car would not be notified and his
> pointer to the window shouldn't keep the garbage collector form freeing the
> window memory).
>
> Again, Smalltalk implements this with weak references, that doesn't count
> for the garbage collector.
>
> Is there support for this in Python?
>
> Any suggestions?

Having the same background and facing the same problem, I wrote a small module
called "subject" with one simple class called "Subject".
Here it is:
--- snip --
##############################################################################
#
"""This module implements stuff for observed python objects, called
subjects."""
#
# $Header: /cvsroot/python/pss/subject.py,v 1.3 2000/02/23 14:45:59 bear Exp $
#
##############################################################################

__version__ = "1.1"
__revision__ = '$Revision: 1.3 $'[11:-2]

class Subject:
 """This class represents an observed subject."""

 def __init__( self ):
  """Initialize the subject."""
  self.observers = []

 def attach( self, observer ):
  """Add an observer to the subject.
  Observers are callable objects."""
  self.observers.append( observer )

 def detach( self, observer ):
  """Remove an observer."""
  self.observers.remove( observer )

 def notify( self, facet = None ):
  """Notifiy all observers of the subjects
  change in facet."""
  for observer in self.observers:
      observer( self, facet )

 def detachAll( self ):
  """Detach all observers."""
  self.observers = []


##############################################################################
#
# $Log: subject.py,v $
# Revision 1.3  2000/02/23 14:45:59  bear
# Added a detach all method.
#
# Revision 1.2  1998/10/22 21:33:03  bear
# Changed or added revsion and version info.
# Added module documentation strings.
# Started wrapping proxy.
#
# Revision 1.1  1998/10/20 18:57:35  bear
# Finished test code. Started real work.
#
##############################################################################
--- snip --
Gerald Klix



From rumjuggler at home.com  Thu Apr 20 12:05:38 2000
From: rumjuggler at home.com (Ben Wolfson)
Date: Thu, 20 Apr 2000 16:05:38 GMT
Subject: docstrings
Message-ID: <00bufsonasp7bgkf86lfb6sv5np5v8mqlj@4ax.com>

is there a way to use %c and %s within a docstring, or to add strings
together in a docstring, so that if some strings are used in all
docstrings, but might be changed, I can simply change one or two
variables?

thanks.

-- 
Barnabas T. Rumjuggler

Oh, oh, 'courtesan', oh aren't we grand.  Harlot's not good enough for
us eh?  Paramour, concubine, fille de joie.  That's what we are not.
Well listen to me my fine fellow, you are a bit of tail, that's what
you are.
 -- Monty Python, "Biggles Dictates a Letter"


From bragib at my-deja.com  Mon Apr  3 19:52:54 2000
From: bragib at my-deja.com (bragib at my-deja.com)
Date: Mon, 03 Apr 2000 23:52:54 GMT
Subject: Creating variables on the fly...
References: <meh9-25D0C5.15160303042000@news.cit.cornell.edu> <38E8EFF0.D71D0415@gmd.de> <meh9-32FDF8.16171703042000@news.cit.cornell.edu>
Message-ID: <8cbash$7p5$1@nnrp1.deja.com>

Using Joachims approach you would

use apply(add, (1,2,3,4,....,n))

or if you were for example adding these variables to a list
or a tuple you could have.
list = [1,2,3,4]
tuple = (1,2,3,4)
apply(add,list)   --> returns 10
apply(add, tuple) --> returns 10


Bragi



In article <meh9-32FDF8.16171703042000 at news.cit.cornell.edu>,
Matthew Hirsch <meh9 at cornell.edu> wrote:
> I like this solution. How do I use the apply function in the body to
> call add()? (I won't know ahead of time how many arguments are going
in
> the function).
>
> Thanks,
> Matt
>
> In article <38E8EFF0.D71D0415 at gmd.de>, Joachim Kaeber <kaeber at gmd.de>
> wrote:
>
> > Hi,
> >
> > what about this:
> >
> > def add(*args):
> > sum=0
> > for arg in args:
> > sum=sum+arg
> > return sum
> >
> > add(1,2,3,4)
> > => 10
> >
> > add (1,2,3,4,5,6,7)
> > => 28
> >
> >
> > Matthew Hirsch wrote:
> > >
> > > Hi All,
> > >
> > > Let's say I have four variables:
> > >
> > > a=1
> > > b=2
> > > c=3
> > > d=4
> > >
> > > And I have a function that adds these variables together:
> > >
> > > def add(a,b,c,d):
> > > return a+b+c+d
> > >
> > > But now let's say I have twenty variables that I want to add
together.
> > > This function will no longer work. I would have to rewrite it as
> > > a+b+c+d+e+f+g+...+(20th letter). Is there a way to dynamically
create a
> > > variable name? So that my add function can automatically determine
how
> > > many variable names to create and then add the values together. In
> > > other words, I'd ideally like something like:
> > >
> > > def add(number_of_variables):
> > > return a+b+c+...+(letter corresponding to number_of_variables)
> > >
> > > Thanks for your help,
> > >
> > > Matt
>


Sent via Deja.com http://www.deja.com/
Before you buy.


From pecora at anvil.nrl.navy.mil  Wed Apr 12 07:14:40 2000
From: pecora at anvil.nrl.navy.mil (Louis M. Pecora)
Date: Wed, 12 Apr 2000 11:14:40 GMT
Subject: Need full example of C extension on Macintosh (CW)
Message-ID: <120420000714412887%pecora@anvil.nrl.navy.mil>

I am homing in on writing a simple C extension on a Macintosh using
CodeWarrior, but, as usual, the nitty gritty details are confounding
me.  I could use a FULL example.  I mean a project, .exp file, AND the
C source files.  In the Python download there are Build examples in the
Mac folder, but they are missing the C source.  Does anyone have the C
source or can point to it on the web for one of these examples, like
AE.prj and AE.prj.exp which are already there, but the source code
(AE.c and AE.h ?) are missing?  Thanks for any help.

BTW, I will be able to provide a detailed example for the FAQ if I ever
get far enough.


From m.faassen at vet.uu.nl  Sat Apr  8 19:59:14 2000
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 8 Apr 2000 23:59:14 GMT
Subject: Why should I switch to Python? - Infinity of Primes
References: <Pine.GSO.4.10.10004060914420.3428-100000@sundial>
Message-ID: <8coh4i$hnu$1@newshost.accu.uu.nl>

Summary of the thread:

Why should I switch to Python from Perl?
[usenet lore knows that huge flamewar may be about to explode]
Some discussion on Python versus Perl.
Under what circumstances Python could be better.
Several people say: stick to Perl if you're already comfortable with it.
[usenet lore scratches its head]
Tim Peters joins in.
thread segues into math.
[usenet lore looks flabbergasted] 
long discussion on math, formal proof of the infinity of primes ensues
Constructivist math, Brouwer, half of Pythondom joins into the discussion
[usenet lore panics!]
Discussion mentiones Haskell and Scheme, as they always do.
[usenet lore goes mad and concludes original poster should switch to
Haskell, *now*]

I just think this thread is very funny. It should count as a reason to
use Python. Although the place has become a bit hectic at times in this
newsgroups, we seem to have various antibodies against flamewars here. :)

One-of-which-is-called-the-timbot-ly yours,

Martijn



From michael_peck at my-deja.com  Thu Apr 13 13:20:52 2000
From: michael_peck at my-deja.com (Michael Peck)
Date: Thu, 13 Apr 2000 17:20:52 GMT
Subject: Embedded interp --> call back to embeddor?
Message-ID: <8d4vkv$r9p$1@nnrp1.deja.com>

Hi,

I can't tell from any of the references I've read whether my objective
is possible. What I want to do is provide a C program that embeds a
Python interpreter in such a way that the executed Python code can call
back to C functions within my program.

Is this possible? Can someone provide information on the procedure?

Help is much appreciated.

MJP


Sent via Deja.com http://www.deja.com/
Before you buy.


From musingattheruins at my-deja.com  Fri Apr  7 14:56:59 2000
From: musingattheruins at my-deja.com (musingattheruins at my-deja.com)
Date: Fri, 07 Apr 2000 18:56:59 GMT
Subject: anyone else wrap gd-1.8?
References: <m2wvn9pijt.fsf@netmeg.net>
Message-ID: <8clb1j$8ul$1@nnrp1.deja.com>

Yeah, I wrapped it but did not use SWIG so it took longer than 1.25
hours.  It's available with source for Win32 at
http://musingattheruins.homepage.com/PythonGD.html

There's a compiled version or instructions for creating the module
yourself.  Uses the latest of everything (zlib, libpng, libjpg, and
gd). Also supports PNG and JPEG.

The compiled version (GD.pyd) is uncompressed and can just be dumped
into the \Program Files\Python\DLLs folder.  It does not create any
ancillary files or registry entries so to uninstall it just delete it.

Although I do not use keywords in parameters (maybe in a future
version) it does allow for tuples and __doc__ strings.

Also, image file pointers have been wrapped and access is through the
the file name and/or a python TYPE, which seems a little easier to read.

Regards,
Bill :-)


In article <m2wvn9pijt.fsf at netmeg.net>,
  Les Schaffer <godzilla at netmeg.net> wrote:
> I just wrapped the latest gd module -- version 1.8 which now handles
> both png and jpeg file formats -- into python. It was quite easy with
> SWIG.
>
> I noticed the gdmodule link on
>
> http://www.python.org/topics/scicomp/plotting.html
>
> is dead. has anyone else wrapped gd up lately? otherwise i am happy to
> distribute (all 1.25 hours of my accumulated time :-)
>
> les schaffer
>
> this gives you an idea what you can do:
>
> #!/usr/bin/python -O
>
> from gd import *
>
> # this demo taken __straight__ from the gd sample demo on boutell's GD
> # home page: http://www.boutell.com/gd and pythonized
>
> # Allocate the image: 64 pixels across by 64 pixels tall
> im = gdImageCreate(64, 64)
>
> # Allocate the color black (red, green and blue all minimum).
> # Since this is the first color in a new image, it will
> # be the background color.
> black = gdImageColorAllocate(im, 0, 0, 0)
>
> # Allocate the color white (red, green and blue all maximum).
> white = gdImageColorAllocate(im, 255, 255, 255)
>
> # Draw a line from the upper left to the lower right,
> # using white color index.
> gdImageLine(im, 0, 0, 63, 63, white)
>
> # Open a file for writing. "wb" means "write binary"
> pngout = fopen("test.png", "wb")
>
> # Do the same for a JPEG-format file.
> jpegout = fopen("test.jpg", "wb")
>
> # Output the same image in JPEG format, using the default
> # JPEG quality setting.
> gdImageJpeg(im, jpegout, -1)
>
> # Output the image to the disk file in PNG format.
> gdImagePng(im, pngout)
>
> # Close the files.
> fclose(pngout)
> fclose(jpegout)
>
> # Destroy the image in memory.
> gdImageDestroy(im)
>


Sent via Deja.com http://www.deja.com/
Before you buy.


From mjhand at concentric.net  Tue Apr 18 10:49:22 2000
From: mjhand at concentric.net (Manus Hand)
Date: Tue, 18 Apr 2000 08:49:22 -0600
Subject: More starship.python.net news
Message-ID: <38FC75F2.ADD6AC72@concentric.net>

The mscp folks informed me this morning (Tuesday) that apparently Mike
McLay
physically removed the starship machine from their data center
yesterday.

Hopefully to repair it, not bury, it??

Can someone here on the newsgroup possibly get in touch with Mike so we
could get an update?

Manus



From dgoodger at bigfoot.com  Sun Apr 23 00:26:05 2000
From: dgoodger at bigfoot.com (David Goodger)
Date: Sun, 23 Apr 2000 00:26:05 -0400
Subject: Distribution
In-Reply-To: <39013f48@flexnet239.uunt.net>
References: <39013f48@flexnet239.uunt.net>
Message-ID: <B527F39C.24A7%dgoodger@bigfoot.com>

on 2000-04-22 01:57, Highdesertman (HighdesertmanREMOVE at yahoo.com) wrote:
> If I use python to develop an application for the windows 9x environment,
> I have already downloaded the appropriate libraries and version) what form
> does the finished program take?

Python is an interpreted language, as far as the programmer & user are
concerned. A Python program is a text file, the "source code".

> How does it get distributed?

As a text file, perhaps as a .zip or .tar.gz archive of several files &
directories. Also, some parts (extensions) may be written in C (for speed),
and these may be in either binary or source (text) form.

The user must have Python installed. But since it's absolutely free with no
strings attached whatsoever, and multi-platform to boot, that's not a
problem.

> But since Python
> remains relatively unknown as yet, the average user isn't likely to have a
> python interpreter handy.

Python is becoming more popular with each passing day. Anyone with Linux has
it. Anyone else can get it, free.

> In addition, one of the real advantages of the
> development of the compiler was that it gave your code a level of protection
> against being tampered with and made it easily useable by a wide audience.

One of the greatest advantages of interpreted languages in general, and
Python in particular, is that their programs can be run under multiple
operating systems (if they are written correctly, of course). You are
running Win95. I run MacOS. Bruce runs Linux. We can all run the same Python
program. Result: even wider audience!

If someone really wants to get into (decompile, reverse engineer) a compiled
program (.exe), they can. It takes some work to do this, but compiling is
not secure encryption.

If you release your code to the public, slap a copyright on it. Make it open
source, and you may be surprised; if someone else finds it useful, they may
just fix/extend/modify it and return the improved version to you. See
http://gotools.sourceforge.net/links.html for some links to sources of
further information.

-- 
David Goodger    dgoodger at bigfoot.com    Open-source projects:
 - The Go Tools Project: http://gotools.sourceforge.net
 (more to come!)



From greene at hctc.com  Thu Apr 13 23:57:01 2000
From: greene at hctc.com (Collin Greene)
Date: Thu, 13 Apr 2000 20:57:01 -0700
Subject: *total* newbie help
Message-ID: <sfd5gp9ealn159@corp.supernews.com>

I am new to to programming but it would seem to em that you could white some
code and then run it in a different action, does python ingrate these into
one action? I am just a little confused

thanks for your time

Collin Greene




From mwh21 at cam.ac.uk  Tue Apr 25 03:32:51 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 25 Apr 2000 08:32:51 +0100
Subject: Sorting a list of classes
References: <OmnqwHar$GA.233@cpmsnbbsa03>
Message-ID: <m3snwafxek.fsf@atrus.jesus.cam.ac.uk>

"lexberezhny" <lexberezhny at email.msn.com> writes:

> If i have a python list of classes, can i do something like
> classList.sort() and how can i control this behavior within the
> classes.

Do you mean "class"?  In which case, no.  If you mean "instance", then
see the other replies...

Cheers,
M.

-- 
41. Some programming languages manage to absorb change, but 
    withstand progress.
     -- Alan Perlis, http://www.cs.yale.edu/~perlis-alan/quotes.html


From bjorn at roguewave.com  Tue Apr  4 14:11:58 2000
From: bjorn at roguewave.com (Bjorn Pettersen)
Date: Tue, 04 Apr 2000 12:11:58 -0600
Subject: Python 1.6 alpha 1 released
References: <Pine.GSO.4.10.10004040035420.21084-100000@sundial> <5E877FD7637AB46E.DABBED9CF3C1D64B.CDDDC05AE7374A78@lp.airnews.net> <8cca93$9oh$1@nnrp1.deja.com> <B964E942A897FF5F.529887C74CF2FC7E.977238F9475EAC23@lp.airnews.net> <FdoG4.5023$74.81126@news
Message-ID: <38EA306E.B8991085@roguewave.com>

Fredrik Lundh wrote:
> 
> Tres Seaver wrote:
> > >not true.  the documented interface explicitly says that
> > >there is one single argument.  it does not mention optional
> > >arguments.
> >
> > Your insistence that the "documented interface" consists only of the
> > description in the module reference, and not of the example on the next
> > page or of the way the API is actually used in the standard libraries
> > baffles me.
> 
> I thought I explained that in the next sentence:
> 
> > >(the example was broken, but so were other examples in earlier
> > >releases of the documentation.  see other messages in this
> > >thread for details)

Your definition of broken seems very peculiar.  Correct me if I'm wrong,
but it seems to go something like:

  - you don't know if examples are binding documentation
  - pair in the description of connect _must_ mean tuple
    and _not_ a pair of arguments because ??? (I would 
    have thought it would have said tuple if it meant 
    tuple -- tuples being Python concept as opposed to pairs)
  - the example in question is broken because, even though
    it works, and seems to follow the description from the
    previous page, ...
  - ... other examples have been broken in the past...

I must say I'm confused...

The good news is that Guido realized how much code this would break, and
will fix the code.

it's-the-only-practical-thing-to-do'ly y'rs
-b



From bjorn at roguewave.com  Tue Apr 11 15:56:26 2000
From: bjorn at roguewave.com (Bjorn Pettersen)
Date: Tue, 11 Apr 2000 13:56:26 -0600
Subject: [ANNOUNCE] Garbage collection for Python
References: <20000407221750.A29279@acs.ucalgary.ca> <38F13FB9.D1D16AF@mindspring.com> <slrn8f2if1.dop.nascheme@cranky.arctrix.com> <87r9cdhdb0.fsf@deneb.cygnus.argh.org> <FsunGI.BtB@world.std.com> <38F319A4.89621192@mdlink.de> <38F351B0.45E137AA@roguewave
Message-ID: <38F3836A.5F9A7C6F@roguewave.com>

Helge Hess wrote:
> 
> Bjorn Pettersen wrote:
> > Helge Hess wrote:
> > > I think what he wants to do (me too ;-) is to run several copies of the
> > > interpreter in *one* process, possibly in multiple threads.
> > > But I'm not sure whether this is a GC issue though (RC isn't thread-safe
> > > anyway, right ?).
> >
> > Well, it actually can be -- at least on Windows :-)
> 
> The question is, what overhead is involved here. If these Windows
> functions wrap the RC ops in locks you have gained nothing (wrapping RCs
> in locks breaks performance in a big way). Note that std-RC is often
> only one CPU op (incr,decr).
> 
> Also note that not only the ++ or -- needs to be protected, but the
> '==0' check as well. Eg:
> 
>    [task1]            [task2]
>    dec(obj);          inc(obj);

Note that both task 1 and task two must have a reference to the object
for this to be possible, so the ref count must at least be two. (unless
you're thinking of sharing names between tasks without locking -- that
would make rc much more resource intensive...)

> could be
>    [task1]            [task2]
>    obj->rc--;
>    if (obj->rc==0)
>      <task-switch>    obj->rc++;
>      free(obj);       <task-switch>
>    object-freed       assumes to have a valid obj

If the reference count was one, adding mutexes wouldn't help:

    aquire(obj->mutex)
    obj->rc--;
    if (obj->rc==0)
      free(obj);   
    release(obj->mutex)

                         aquire(obj->mutex)  // does it crash here
                         obj->rc++           // or here?
                         ...

> [note that checking for 0 in task2 doesn't solve the problem, since the
> check wouldn't be atomic again, but maybe I'm missing something].
> 
> A (m&s) GC is somewhat in advantage here, since it is run only at
> specified (safe) points, so there is no locking required at such high
> frequency.

No argument from me there :-)

The api calls are direct wrappers around the processors
increment-and-test or test-and-set instructions, and are extremely fast
compared to a mutex solution (which makes it interesting to see other
people on this thread claim that it would be _way_ to slow without even
testing it...).  Our tests showed an order of magnitude difference on
some tests...

-- bjorn



From effbot at telia.com  Mon Apr 17 12:44:12 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Mon, 17 Apr 2000 16:44:12 GMT
Subject: newbie question
References: <8dfdfk$uum$1@nnrp1.deja.com>
Message-ID: <wbHK4.2614$Za1.38989@newsc.telia.net>

Arint? <jamarijr at hotmail.com> wrote:
> If i PyErr_SetString(PyExc_TypeError,"Invalid command value for IOCTL");
> in my c/c++ program, how would python script get that error string?

return NULL to the interpreter.  do *not* return
a python value.

also see:
http://www.python.org/doc/current/ext/errors.html

</F>





From zorro at zipzap.ch  Thu Apr 20 03:22:31 2000
From: zorro at zipzap.ch (Boris Borcic)
Date: Thu, 20 Apr 2000 09:22:31 +0200
Subject: Starship update: April 18, #2
References: <3daeirht58.fsf@amarok.cnri.reston.va.us>
Message-ID: <38FEB037.A2A306A6@zipzap.ch>

Andrew M. Kuchling wrote:

> * We have found a backup of /home dating from February 1, 2000.

Irony is a temptation, although I am in no position
to claim anything like moral high ground on the matter.

Your statement *does* suggest a serendipitous archeological
discovery...

Regards,

BB
--
"Aesthetic sense is to soul as eyes are to viewpoint"


From tseaver at starbase.neosoft.com  Fri Apr 28 00:00:48 2000
From: tseaver at starbase.neosoft.com (Tres Seaver)
Date: 27 Apr 2000 23:00:48 -0500
Subject: web site test framework? in python?
References: <Pine.BSF.4.10.10004271745230.28211-100000@chi.pair.com>
Message-ID: <37AC13C2BCECB1CD.9E56BFDEBBCEBE89.2AC3EC74BF6CB660@lp.airnews.net>

In article <Pine.BSF.4.10.10004271745230.28211-100000 at chi.pair.com>,
Michal Wallace (sabren) <sabren at manifestation.com> wrote:
>On Thu, 27 Apr 2000 tiddlerdeja at my-deja.com wrote:
>
>> I'd also like to code the testing framework in Python.
>> 
>> Again, has anyone any experience in this testing websites arena?
>
>Well, there's http://pyunit.sourceforge.net/ .. I't s a pretty good
>test suite manager.
>
>
>> I was thinking of having JSP or ASP pages check for a testing variable
>> (registry or environment). If pages were called in test mode they would
>> return XML to be parsed which would can then be used as input data to
>> call the page again (this way you could store the test data with the
>> page you are testing). The XML would also contain the desired correct
>> data to check against to ensure the page is working correctly.
>
>Perhaps you should stick all this code in an object so you don't have
>to complicate your tests by communicating via http. You can set it
>up so you just pass objects around..

PyUnit is fine for *unit* testing, but tiddler is asking for what XP
would call *functional* testing -- XP requires that *developers* write
unit tests, but *users* write functional tests, in order to verify
that the system is delivering the promised externally-visible behavior.

I for one would be extremely interested in a means of running
automated functional tests through the web.  My own crude attempts at
this use urllib and diff(1), and break on a number of my pages.

Manual-testing-is-for-infinite-monkeys'ly

Tres.
-- 
---------------------------------------------------------------
Tres Seaver        tseaver at digicool.com     http://www.zope.org
Digital Creations


From phd at phd.russ.ru  Sun Apr  9 11:12:30 2000
From: phd at phd.russ.ru (Oleg Broytmann)
Date: Sun, 9 Apr 2000 15:12:30 +0000 (GMT)
Subject: Python paradigms
In-Reply-To: <38F00890.3C98EBDB@acm.org>
Message-ID: <Pine.LNX.4.21.0004091506110.13883-100000@fep132.fep.ru>

On Sun, 9 Apr 2000, Robert W. Cunningham wrote:
>  blah = (((test and [true]) or [false])[0])             # blah = (test ?
> true : false)
> 
> Works for me!  Only if I have a good reason to suspect a problem in this
> particular line of code will I have to look past the comment.

   There is no problem in this particular line, but you'll have plenty of
them if you'd try to use this paradigm:

   blah = ((foo() and [poo()]) or [boo()])[0]

   In C, "foo() ? poo() : boo()" will not call boo() if foo() returns true;
but in Python, you'll have a side-effect of calling boo() even if foo()
returns true - what is wrong, usually.

> Are there ANY Python programmers who don't also know at least this much C?

   I am. I came to Python after 10 years of Borland Pascal and no C. Firest
I tryied Perl, but was very unsuccessful writing my programs; then tryied
Python - and voila! - I use Python for over 4 years and don't see a need to
change anything! :)
   I learned C after Python ('cause I switched from DOS to UNIX the same
time I turned from BPascal to Perl or Python). And still don't know much C.

Oleg.            (All opinions are mine and not of my employer)
---- 
    Oleg Broytmann      Foundation for Effective Policies      phd at phd.russ.ru
           Programmers don't die, they just GOSUB without RETURN.




From jmassung at magpiesystems.com  Mon Apr 17 11:00:56 2000
From: jmassung at magpiesystems.com (Jeff Massung)
Date: Mon, 17 Apr 2000 09:00:56 -0600
Subject: Editors and books
Message-ID: <XBFK4.16$Uq2.1831@news.uswest.net>

First, I'd like to give my undying praise to PYTHON: Essential Reference, by
David M. Beazley. I'm an accomplished C/C++ programmer who wanted to learn
Python. This book was IT! Anyone whose reading this that is in my shoes -
get this book! It's some of the best $35 I've spent!

Next, a question for those who've been doing this: is there a color IDE
(just a nicer thing that IDLE) for the Mac? I've heard that Alpha (which I
just downloaded) has a Python mode, but I can't seem to find it. Is there
something else, either? can MPW do it? BBEdit?

Thanks, guys.

Also, on another NG there's talk about Python, and how it does (and still
doesn't) do garbage collecting. Could someone here expand on it so I know
what it can and can't do? Memory is a big issue in my project and I need to
know how to structure my program so that garbage is collected. Thanks
again!!

-- Jeff Massung (jmassung at magpiesystems.com)
-- Lead programmer for Magpie Systems
-- Bringing smart pigs to the pipeline industry!




From chapman at bioreason.com  Mon Apr 17 13:00:38 2000
From: chapman at bioreason.com (Mitch Chapman)
Date: Mon, 17 Apr 2000 11:00:38 -0600
Subject: Scripting and Gnome and KDE
References: <Pine.GSO.4.10.10004150741520.11261-100000@sundial> <8d95m0$9j6$1@news1.xs4all.nl> <8damat$4ja$1@slb2.atl.mindspring.net> <8dbsgc$35o$1@news1.xs4all.nl>
Message-ID: <38FB4336.2BC6C235@bioreason.com>


Boudewijn Rempt wrote:
> 
> Andrew Dalke <dalke at acm.org> wrote:
> 
> > Even though it says "The result is a great environment for rapid
> > development of GUI-based applications." ?
> 
> Yes - the article didn't make clear what was so great, but it did point
> out a lot of problems (lack of documentation, inconsistent API's, bugs).
> 
> > I used to work at Bioreason, though I wasn't involved in the GUI
> > development.  They are pretty happy about PyGTK.  The idea of the
> > DDJ article was to point out how to take advantage of the toolkit,
> > based on finding out what doesn't work.

That's exactly right.  It's fairly easy to get started with pygtk,
but some things about the toolkit are difficult to discover.  We wanted
to show how to do the difficult (i.e. obscure) things, in order to 
make them easy.  The article introduction should have made that more
clear.

> 
> Very honest, but it doesn't make me all that eager ;-).
> 
> > You could write a similarly bleak picture of any toolkit.

In fact, I'd considered trying to write a similar article about PyQt
and PyKDE.  But that article would have been even bleaker, because
the process of discovery w. PyQt is still difficult for me.  I've
had a few recurring problems.  Maybe this would be a good place to
ask for enlightenment :)  I'll start w. the two biggies:
	o The Qt C++ interfaces include many overloaded functions.
	  Is there a consistent pattern for determining which of
	  these is implemented by the Python binding?  Or must
	  you look at the source code for the Python bindings?
	o In many cases, if I make a wrong guess as to which 
	  overloaded method is implemented by the binding, I get
	  a segmentation fault or (slightly better) a traceback 
	  indicating an incorrect argument type.  Is there an easy 
	  way to work backwards from a segmentation fault to
	  the correct usage for a PyQT widget?

> Absolutely, but I don't know that an essay on the problems is exactly
> the right way of introducing the toolkit.

That's probably true.  Given the goal of showing how to do non-obvious
things with a toolkit, what changes would you recommend?

If I understand your comments, it would have been better to simply 
say, "Here's how you do X," rather than to explain "Here's *why* you 
do it this way".  I have one reservation with such an approach:
it tells the reader how to solve a specific problem, but
doesn't provide a general method for discovering how to use other
parts of the toolkit.

Hoping that made some sense...

--
Mitch Chapman
chapman at bioreason.com


From dgoodger at bigfoot.com  Fri Apr 21 09:30:40 2000
From: dgoodger at bigfoot.com (David Goodger)
Date: Fri, 21 Apr 2000 09:30:40 -0400
Subject: suggestion on an operator
In-Reply-To: <Pine.LNX.4.21.0004202210480.9838-100000@saturn.ispol.com>
References: <Pine.LNX.4.21.0004202210480.9838-100000@saturn.ispol.com>
Message-ID: <B525D03D.2419%dgoodger@bigfoot.com>

on 2000-04-20 22:27, Gregory Trubetskoy (grisha at ispol.com) wrote:
> If I want to configure apache to use that handler, I would put something
> like this in the .htaccess config file:
> 
> PythonAuthenHandle  mypackage.mymodule::authenitcate

Why not just a space (or whitespace, r'\s+')? Seems natural, and
string.split() is a no-brainer.

-- 
David Goodger    dgoodger at bigfoot.com    Open-source projects:
 - The Go Tools Project: http://gotools.sourceforge.net
 (more to come!)



From dante at oz.net  Wed Apr 26 22:43:19 2000
From: dante at oz.net (Michael Esveldt)
Date: 27 Apr 2000 02:43:19 GMT
Subject: regsub backreference question
Message-ID: <8e89g7$dfg$0@216.39.162.232>

I'm trying to figure out how back referencing works in python's regsub 
module. I thought the following code would simply double the string 
"foo" but it's just returning "foo".

regsub.sub(r"(.*)", r"\1\1", "foo")

What am I doing wrong here?

thanks in advance,
Michael
_______________________________________________________________
dante at oz.net - Michael Esveldt - #FightThePower on Openprojects


From ben at co.and.co  Fri Apr 14 12:05:22 2000
From: ben at co.and.co (ben at co.and.co)
Date: Fri, 14 Apr 2000 16:05:22 GMT
Subject: Question about wave.py
Message-ID: <6lHJ4.84496$l35.172139@afrodite.telenet-ops.be>

It looks like I'm missing the point, so if somebody could enlighten me?

Read all frames of a *.wav file:
------------------------------------
import wave

w = wave.open('test.wav', rb)
frames = w.readframes(w.getnframes())

# now, we have all frames in ``frames''
# but how are we supposed to handle the individual frames?
# perhaps like this, but...

import array

frames_2 = array.array('h')          # assuming 16bit mono
frames_2 = array.fromstring(frames)

print "Frame[100]: %d" % (frames_2[100],)

# ... this works, but now we have 2 copies of all frames :-(

Thanks in advance,
-- 
ben . de . rydt at pandora . be ------------------ your comments
http://users.pandora.be/bdr/ ------- inl. IPv6, Linux en Pandora



From alander at nortelnetworks.com  Tue Apr 11 18:00:49 2000
From: alander at nortelnetworks.com (Anthony Lander)
Date: 11 Apr 2000 22:00:49 GMT
Subject: Trouble with Tkinter under IDLE
References: <8F139460Canthonylander@47.80.11.122> <38F37BF9.DCAB3570@exceptionalminds.com>
Message-ID: <8F13B1259anthonylander@47.80.11.122>

tjg at avalongroup.net (Timothy Grant) wrote in 
<38F37BF9.DCAB3570 at exceptionalminds.com>:

>I believe that the solution is *not* to call mainloop() from your
>programme as it is conflicting with the mainloop() of IDLE. Though I
>don't use IDLE, I do believe that the mainloop in IDLE will allow your
>code to run correctly without you calling it.

Indeed. That seems to be the case.


Looking through the Tkinter.py source, it seems that there is a variable 
called _default_root which wants to be set to the default root window. Is 
the normal procedure to check this variable and use it as the root window 
if it's defined?

 thanks,

 anthony


From kilian.lorenz at dkfz-heidelberg.de  Sun Apr  9 14:18:05 2000
From: kilian.lorenz at dkfz-heidelberg.de (Kilian Lorenz)
Date: Sun, 09 Apr 2000 20:18:05 +0200
Subject: Tkinter:  centering windows on screen
References: <ak1I4.121418$Dv1.1306619@news1.rdc1.bc.home.com> <38F0C6A7.10878609@dkfz-heidelberg.de>
Message-ID: <38F0C95D.1F7E853@dkfz-heidelberg.de>

>> that works but doesn't help me much because I'm interested in the
>> return value. So I would really like to know how to call Tk commands
>> from within Python because .... blah blah

Now I also found out how this works ... pretty easy with
<widget>.tk.call(<command>, <parameter>, <next parameter> ...)

You just have to browse Tkinter.py and get the answers to all your
questions !  :)


From PClaerhout at CREO.BE  Wed Apr 12 06:27:51 2000
From: PClaerhout at CREO.BE (Pieter Claerhout)
Date: Wed, 12 Apr 2000 12:27:51 +0200
Subject: Getting an overview if all available driveletters on NT
Message-ID: <2B1262E83448D211AE4B00A0C9D61B03BA729F@MSGEURO1>

Ok,

here's the complete story. I used win32api.GetLogicalDriveStrings() to get a list of
all available drives, and then I used win32file.GetDriveType(i) to find out the type
of the drives. This is what I found:

- 2 -> floppy
- 3 -> fixed hard drive or subst'd drive
- 4 -> mapped network drive
- 5 -> cd-rom drive

Kind regards,



Pieter
Pieter_Claerhout at creoscitex.com



-----Original Message-----
From: fredrik at pythonware.com [mailto:fredrik at pythonware.com]
Sent: Wednesday, April 12, 2000 11:58 AM
To: python-list at python.org
Subject: Re: Getting an overview if all available driveletters on NT


Ulf Engstrom wrote:
> As the effbot told you the win32api is probably the one you wanna
use, but
> the function is GetLogicalDriveStrings() This won't gives you all the
the
> driveletters, but not their labels, there might be something like that
> there, just never have had to use it. Someone else might now that
part?
> >>> win32api.GetLogicalDriveStrings()
> 'A:\\\000C:\\\000D:\\\000E:\\\000'

umm.  didn't the original poster ask for a way to
tell ordinary drives from network drives?  that's
exactly what GetDriveType does...

</F>


Sent via Deja.com http://www.deja.com/
Before you buy.
-- 
http://www.python.org/mailman/listinfo/python-list



From alberanid at bigfoot.togli.questo.com  Thu Apr  6 13:34:37 2000
From: alberanid at bigfoot.togli.questo.com (Davide Alberani)
Date: Thu, 06 Apr 2000 17:34:37 GMT
Subject: encoding type in a2ps to pretty-print
References: <8chtbq$ohg$1@news.net.uni-c.dk>
Message-ID: <1phic8.537.ln@snoopy.mio>

"Henrik Berg Nielsen" <n99hbn at nat.sdu.dk> wrote:
>
> When I pretty-print my code via a2ps all places where I have a '<='
> are turned into the similar mathematical sign, that only takes up
> one space on the line,

Probably you're running a2ps with the -g (or --highligh-level=heavy)
option.

Also look at /etc/a2ps.cfg .

-- 
Homevideo: e` in edicola Class of Nuke'em High: grande       _/Davide\_
capolavoro trash della Troma.                                \Alberani/


From matthias_w at my-deja.com  Fri Apr 14 04:10:51 2000
From: matthias_w at my-deja.com (Matthias)
Date: Fri, 14 Apr 2000 08:10:51 GMT
Subject: Embedding Python in Visual C++ 6.0
References: <8d1he4$26d$1@nnrp1.deja.com> <N2YI4.2030$UB.13624@news-server.bigpond.net.au> <8d1od2$9fl$1@nnrp1.deja.com> <Sj%I4.2225$UB.14895@news-server.bigpond.net.au> <8d3qvu$hud$1@nnrp1.deja.com> <dnvJ4.799$1D.4997@news-server.bigpond.net.au>
Message-ID: <8d6jq2$lh7$1@nnrp1.deja.com>

Your are right!!!
Thank you! I was blind! |-)
I worked 3 days on this (no recommendation for me). I tried to download
the new source, but got no response from the server - is the starship
down?

Thank you anyway
Matthias

In article <dnvJ4.799$1D.4997 at news-server.bigpond.net.au>,
  "Mark Hammond" <mhammond at skippinet.com.au> wrote:
> It appears that 2 win32ui.pyd files are being loaded.  This is almost
> certainly my fault.  It is possible that in the registry you already
have
> a "Modules" subkey that specifies the location of win32ui.pyd?
>
> Also, FYI, in the CVS version of Pythonwin, this implementation has
> changed.  It now doesnt attempt to find registered modules or any of
the
> other funky stuff that it does now.  The new code is simpler and more
> robust - but does insist that win32ui.pyd is either on your
PYTHONPATH, or
> in the same directory as pythonwin.exe - no a burden IMO!  The new
code
> also magically works with either 1.5 or 1.6.
>
> Anyway, whatever the reason, the clues are:
>
> ...
> > Loaded symbols for 'C:
> > \Programme\Python\Pythonwin\src\work\Pythonwin\Build\win32ui_d.pyd'
>
> That is fine:
> ...
>
> > Extension module C:
> > \Programme\Python\Pythonwin\src\work\Pythonwin\Build\win32ui_d.pyd
> > initialising.
>
> Still looking good!
>
> > LDR: Dll win32ui_d.pyd base 1e400000 relocated due to collision
with C:
> > \Programme\Python\Pythonwin\src\work\Pythonwin\Build\win32ui_d.pyd
> > Preloaded symbols may not match 'C:
> > \Programme\Python\Pythonwin\win32ui_d.pyd'.
>
> Did you cut any messages?  This message appears to imply another
> win32ui_d.pyd has been loaded, and its symbols have been relocated
due to
> a clash with another win32ui_d.pyd.
>
> Can you ensure you system has no others win32ui*.pyd.
>
> Also, I highly recommend getting the CVS version of Pythonwin - just
for
> the new startup code if nothing else.
>
> Mark.
>
>


Sent via Deja.com http://www.deja.com/
Before you buy.


From m.faassen at vet.uu.nl  Tue Apr 25 12:51:31 2000
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 25 Apr 2000 16:51:31 GMT
Subject: how to wrap c++ software to python?
References: <8e3ggq$2r1d$1@news.cz.js.cn> <8e4g20$eeq$1@nnrp1.deja.com>
Message-ID: <8e4iej$eoc$1@newshost.accu.uu.nl>

portaldata at my-deja.com wrote:
> HOW DO I INSTANCE A SHARED OBJECT IN PYTHON.

I'm not quite sure what you mean. You can make an instance of a class
like this:

class MyClass:
    ...

myinstance = MyClass()

> ALSO, HOW DO I POST A NEW MESSAGE IN THIS GROUP ?

You seem to have been successful. I'd recommend lowercase text instead
of ALL CAPITAL LETTERS, as the latter is often interpreted as shouting
on the net.

Or do you mean, how do I start a new thread? Currently this message
seems attached to some other irrelevant thread. It depends on your
newsreader software (or on Deja.com if you're using that to post).

Regards,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From python at rose164.wuh.wustl.edu  Wed Apr 26 08:52:59 2000
From: python at rose164.wuh.wustl.edu (David Fisher)
Date: Wed, 26 Apr 2000 07:52:59 -0500
Subject: Closing Excel 97 with Python
References: <8e6lhb$8h93@esiami.tsai.es>
Message-ID: <000701bfaf7e$59ae52c0$0201a8c0@spkydomain>

----- Original Message -----
From: "AMR" <alvaro at ocse.infonegocio.com>
Newsgroups: comp.lang.python
To: <python-list at python.org>
Sent: Wednesday, April 26, 2000 7:03 AM
Subject: Closing Excel 97 with Python


> Hello,
>
> I'm using COM to exchange information with Excel and my problem is the
next.
>
> This is my way to open Excel:
>
> self.o = win32com.client.dynamic.Dispatch("Excel.Application")
>
> self.o.Visible = Visible
>
> self.o.Workbooks.Open(self.MyFile)
>
> And this is my way to close Excel:
>
> self.o.Visible = 0
>
> self.o.Workbooks(1).Close(0)
>
> self.o.Workbooks.Close()
>
> self.o.Quit ## For try something

try calling:

self.o.Quit()
                ^^

instead
dnf




From fernstrom at multimania.com  Sat Apr  1 16:03:35 2000
From: fernstrom at multimania.com (Christer Fernstrom)
Date: Sat, 1 Apr 2000 23:03:35 +0200
Subject: controlling Excel2000 with COM
Message-ID: <8c5o77$cg4$1@vega.worldonline.fr>

I tried the most obvious thing with COM with Excel 20000 in Windows 2000 (p.
67 in "Programming on 32")

>> import win32com.client
>> xl = win32com.client.Dispatch("Excel.Application")
  -- This launches Excel (as verified with the windows task manager)
>> xl.visible =1 (I also tried to capitalise Visible)
Result:
Traceback (innermost last):
  File "<pyshell#2>", line 1, in ?
    xl.visible =1
  File "D:\tools\Python\win32com\client\dynamic.py", line 453, in
__setattr__
    raise AttributeError, "Property '%s.%s' can not be set." %
(self._username_, attr)
AttributeError: Property 'Excel.Application.visible' can not be set.

Anybody got a clue? Is this a problem related to Office2000?

    christer





From fredrik at pythonware.com  Wed Apr 12 05:57:44 2000
From: fredrik at pythonware.com (fredrik at pythonware.com)
Date: Wed, 12 Apr 2000 09:57:44 GMT
Subject: Getting an overview if all available driveletters on NT
References: <2B1262E83448D211AE4B00A0C9D61B03BA7294@MSGEURO1> <_FGI4.1194$Za1.20497@newsc.telia.net> <00ed01bfa442$53c66150$858647c1@DEVELOPMENT>
Message-ID: <8d1hak$25p$1@nnrp1.deja.com>

Ulf Engstrom wrote:
> As the effbot told you the win32api is probably the one you wanna
use, but
> the function is GetLogicalDriveStrings() This won't gives you all the
the
> driveletters, but not their labels, there might be something like that
> there, just never have had to use it. Someone else might now that
part?
> >>> win32api.GetLogicalDriveStrings()
> 'A:\\\000C:\\\000D:\\\000E:\\\000'

umm.  didn't the original poster ask for a way to
tell ordinary drives from network drives?  that's
exactly what GetDriveType does...

</F>


Sent via Deja.com http://www.deja.com/
Before you buy.


From smalleys at gte.net  Sun Apr 30 17:38:19 2000
From: smalleys at gte.net (sue)
Date: Sun, 30 Apr 2000 21:38:19 GMT
Subject: converting word documents to html
References: <390B501C.1891CB80@uiuc.edu> <B530EDCA.5754%abrahams@mediaone.net>
Message-ID: <390CA7E5.F378A910@gte.net>

I've used tidy for this for two years.

http://www.w3.org/People/Raggett/tidy/

My only complaint is that it doesn't collapse styles.  But
they're easy enough to spot.

Sue.


From pinard at iro.umontreal.ca  Wed Apr 26 09:28:26 2000
From: pinard at iro.umontreal.ca (=?ISO-8859-1?Q?Fran=E7ois_Pinard?=)
Date: 26 Apr 2000 09:28:26 -0400
Subject: Sourceforge and Python
In-Reply-To: "Neil Hodgson"'s message of "Wed, 26 Apr 2000 10:18:28 +1000"
References: <oq8zy1zqto.fsf@titan.progiciels-bpi.ca> <007b01bfaf14$f33114b0$3cdd3fcb@neil>
Message-ID: <oqzoqhxa85.fsf@titan.progiciels-bpi.ca>

"Neil Hodgson" <neilh at hare.net.au> writes:

> [...] some aspects of SourceForge are more difficult than they should be.

kc5tja at garnet.armored.net (Samuel A. Falvo II) writes:

> [...] What am I missing with respect to CVS with SourceForge?

I thank you for these warnings.  This is helpful information to me.

-- 
Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard





From eviltofu at rocketmail.com  Mon Apr 17 21:59:38 2000
From: eviltofu at rocketmail.com (Jerome Chan)
Date: Tue, 18 Apr 2000 01:59:38 GMT
Subject: Forward References?
Message-ID: <eviltofu-4C5688.21593817042000@news.earthlink.net>

Just a silly question but how do I do forward references?


class a:
   def x(self):
      y(self)
   def y(self):
      print "Hello"

c = a()
c.x()
c.y()

gives a NameError: y


From effbot at telia.com  Mon Apr  3 05:51:23 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Mon, 03 Apr 2000 09:51:23 GMT
Subject: List and Dicts as default args
References: <326536345498D311B3BC00105A39802A074654@newsintern.dspace.de>
Message-ID: <vQZF4.1086$n68.187094016@newsb.telia.net>

Stefan Migowsky <smigowsky at dspace.de> wrote:
> I was just wondering how to handle lists and dictionaries in
> a simple way as default arguments to functions. Since following
> strange behaviour occured :

when exposed to "strange" behaviour, please check the FAQ
http://www.python.org/doc/FAQ.html#6.25

or the fine manual
http://www.python.org/doc/current/ref/function.html

before posting.

> >>> def f(Index,List = [], Dict = {}):
> ...     List.append(1)
> ...     List.append(2)
> ...     print List
> ...     Dict[Index] = Index
> ...     print Dict
> >>> f(1)
> [1, 2]
> {1: 1}
> >>> f(2)
> [1, 2, 1, 2]
> {2: 2, 1: 1}
> >>> f(3)
> [1, 2, 1, 2, 1, 2]
> {3: 3, 2: 2, 1: 1}
>
> This behaviour only occurs with dictionaries and list. All other
> types are "well" behaved.

no, they're not.  default arguments are created once, when
the function object itself is created.  this means that *all*
mutable default values behave the same way -- there is a
single default value, used for all calls.  if you modify that in
place, you loose.

(to understand why, you need to dig a little deeper into what
the 'def' statement actually does.  see the language reference
for details).

> I could try the following but that doesn't look so nice:
>
> def f(Index,List = None, Dict = None):
>     if not List: List = []
>     List.append(1)
>     List.append(2)
>     print List
>     if not Dict: Dict = {}
>     Dict[Index] = Index
>     print Dict

nice or not, that's the right way to do it.

(note that using Caps for variable names isn't good python
style, but that's another story).

</F>




From aahz at netcom.com  Tue Apr  4 16:14:59 2000
From: aahz at netcom.com (Aahz Maruch)
Date: 4 Apr 2000 20:14:59 GMT
Subject: Python 1.6 alpha 1 released
References: <Pine.GSO.4.10.10004040035420.21084-100000@sundial> <5E877FD7637AB46E.DABBED9CF3C1D64B.CDDDC05AE7374A78@lp.airnews.net> <5l7leern5v.fsf@eric.cnri.reston.va.us> <8cdg92$j0m$1@nntp6.u.washington.edu>
Message-ID: <8cdig3$s7f$1@slb6.atl.mindspring.net>

In article <8cdg92$j0m$1 at nntp6.u.washington.edu>,
Russell E. Owen <owen at astroNOJNK.washington.edu.invalid> wrote:
>
>But tuples seem a less than wonderful choice to me. Protocol flexibility 
>presumably requires each protocol to have its own tuple with its own 
>numer of fields in its own order. That's a lot to remember and try not 
>to get mixed up.

I second this.  I'd be happy with either dict or class, with a *slight*
preference for a class.
--
                      --- Aahz (Copyright 2000 by aahz at netcom.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

"It's 106 miles to Chicago.  We have a full tank of gas, a half-pack of
cigarettes, it's dark, and we're wearing sunglasses."  "Hit it."


From m.faassen at vet.uu.nl  Fri Apr 28 14:25:13 2000
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 28 Apr 2000 18:25:13 GMT
Subject: Do I always have to write "self." ?
References: <270420001706386648%pecora@anvil.nrl.navy.mil> <8eacdf$nnm$1@newshost.accu.uu.nl> <280420000751353947%pecora@anvil.nrl.navy.mil>
Message-ID: <8ecl29$euk$1@newshost.accu.uu.nl>

Louis M. Pecora <pecora at anvil.nrl.navy.mil> wrote:
[snip]
> Thanks for the very nice tutorial.  Your example brings up some typical
> stumbling blocks for us beginners.  Namely, references to mutable and
> immutable objects.

Right, that's a common question on this newsgroup. One that I keep getting
involved in, too. :)

> In the above I would describe things as follows (am I right or wrong?):

> (1) data and self.data point to the same object, a list, so when a
> method is called it is the method that works on that instance of a list
> and either data or self.data will append 'foo' to the same list object.

Correct.

> (2) when you do more=more+1 an new object (the more+1 value) is created
> and more then points to it and no longer to the old value of self.more. 
> The latter still exists and points to the original 10 object.

That's right.

> (3) Similar story here for interesting where a new object
> interesting[10:] is created and interesting is "switched" to point to
> it.  self.interesting still points tohe "foo" object.

Yup.

> I hope I got that right, because they have tripped me up a few times.

Yes, you seem to have gotten it all right. The trick is in remembering 
which objects are mutable and which are not. You can't really change an
immutable object, but you *can* change which immutable object a particular
variable is pointing at. The bad thing about this whole system is that
it sometimes trips up newbies (or even non-newbies -- it tripped me up a
while back when I was multiplying a list). The good thing about it is that it's
very consistent, and easy to reason about once you understand it.

Regards,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From breiter at usf.Uni-Osnabrueck.DE  Wed Apr  5 12:40:39 2000
From: breiter at usf.Uni-Osnabrueck.DE (Bernhard Reiter)
Date: 5 Apr 2000 16:40:39 GMT
Subject: Teaching the art of programming, in python
References: <20000405053920.20579.qmail@web2102.mail.yahoo.com>
    <8cfe42$o02$1@nnrp1.deja.com> <dpJG4.3744$HG1.106895@nnrp1.uunet.ca>
    <38EB68A3.B1B5809F@callware.com>
Message-ID: <8cfqa7$4ml$3@newsserver.rrzn.uni-hannover.de>

In article <38EB68A3.B1B5809F at callware.com>,
	Ivan Van Laningham <ivanlan at callware.com> writes:

> Please forgive the BSP (Blatant Self-Promotion) here.

> On April 17, according to Macmillan, my book _Teach Yourself Python in
> 24 Hours_ will finally be out.  Its mission is to provide exactly what
> you're looking for, a book that teaches programming, art, attitude and
> philosophy, using Python.

> determine.  The book should really have been called _Python Mind,
> Beginner's Mind_, because it draws heavily on Shunryu Suzuki's _Zen
> Mind, Beginner's Mind_.  It is not a _Python for Dummies_ book, however,
> because I don't believe that people who want to learn how to program
> should either be called or treated like dummies.  I try to convey how
> much fun programming is, and try to show some of the really neat things
> you can do with Python.

And all in 24 hours?
This is as misleading as the dummy attribute.

Ranting-about-dummy-booktitles-ly-yours,
	Bernhard

-- 
Professional Service around Free Software                (intevation.net)  
The FreeGIS Project				            (freegis.org)
Association for a Free Informational Infrastructure            (ffii.org)


From thomas at xs4all.net  Wed Apr 19 11:53:10 2000
From: thomas at xs4all.net (Thomas Wouters)
Date: Wed, 19 Apr 2000 17:53:10 +0200
Subject: Very Horrible Question About Goto's
In-Reply-To: <C5A7720BDEB7D3118A5200E018C200A037EA43@EXCHANGE>; from corg@copernic.com on Wed, Apr 19, 2000 at 11:17:21AM -0400
References: <C5A7720BDEB7D3118A5200E018C200A037EA43@EXCHANGE>
Message-ID: <20000419175310.N15664@xs4all.nl>

On Wed, Apr 19, 2000 at 11:17:21AM -0400, Gaetan Corneau wrote:

> I was surprised to learn that Python doesn't have a goto statement.

Really ? I never really wondered about it. I use goto in C sometimes, mostly
as a performance hack, but performance issues are moot in Python. I haven't
seen a single bit of code that needed a goto. When it needs a goto, it's
usually misshapen(*) in other ways as well.

(* my own, purely objective <1.1 wink> opinion, of course.)

> Would it be difficult to add?

Hopefully. I hope whoever implements it has to do so using a burning
lava-chair, on a keyboard with diamond spikes instead of keys... and i would
like to see his face when Guido rejects the patch :-)

Continuations, however, are something else entirely... you can possibly do
what you want with continuations, and continuations, continuations are very
very very very cool ;)

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From bjorn at roguewave.com  Mon Apr 10 15:05:23 2000
From: bjorn at roguewave.com (Bjorn Pettersen)
Date: Mon, 10 Apr 2000 13:05:23 -0600
Subject: Refering to substring of list containing strings
References: <MPG.13591514a982d0a4989682@nntp.server.uni-frankfurt.de>
Message-ID: <38F225F3.31B187D5@roguewave.com>

That should work (except you can't assign to it -- strings are
immutable).  A little more context might clarify exactly what you're
trying to do...

-- bjorn

Peter Kovalsky wrote:
> 
> Hi,
> 
> I am new to Python and lately experienced a little problem. I have a list
> with several strings and have to modify these strings. Therefore I used a
> for-construct (for i in range(0, len(mylist)))- to operate every lists
> element. But how do I adress a substring in such a list? I tried a
>         mylist[i][lborder:rorder]
> - just like one would do with a simple string-var - wich won' t work. Any
> help would be greatly appreciated,
> 
> thanx in advance,
> 
> Peter Kovalsky
> --
> http://www.python.org/mailman/listinfo/python-list



From effbot at telia.com  Sat Apr 15 07:27:38 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Sat, 15 Apr 2000 11:27:38 GMT
Subject: Nasty Zope & Python Imaging Library name collision
References: <slrn8fg5pq.td6.miles@caddr.com>
Message-ID: <KmYJ4.1126$rc9.190032896@newsb.telia.net>

miles <miles at caddr.com> wrote:
> I've finally traced a mysterious failure in some Zope external methods
> I've been working on to a module name collision between Zope.  In a
> nutshell, they both contain an 'ImageFile' module and the wrong one
> gets loaded.
>
> This is essentially what my external method does:
>
> import Image
> i = Image.open(filename.jpg)
> i = i.resize((x,y))
> i.save(newfilename.jpg)
>
> This eventually results in the construction of a Jpeg image object which
> is supposed to derive from ImageFile.ImageFile from the Imaging library.
> Sadly, the __init__ method from Zope's ImageFile module is called instead,
> which promptly goes berserk.
>
> Is there anything I can do to fix this?  How do people do deal with these
> kinds of problems in general?  It seems to be very easy to inadvertently
> collide with another module.  Perl avoids this by allowing arbitrarily
nested
> modules.  For example, in this case I'd have Zope::ImageFile and
> Imaging::ImageFile and everything would be roses.

assuming you have PIL 1.0, just make sure you've installed
all the PIL files in a directory named PIL, and then do:

    from PIL import Image

instead of "import Image".

if you have an older version, you may have to add an
__init__.py file to the PIL directory, and explicitly import
format handlers for the file formats you want to use (e.g.
import PIL.JpegImagePlugin, PIL.GifImagePlugin)

</F>





From mwh21 at cam.ac.uk  Thu Apr 13 17:16:53 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 13 Apr 2000 22:16:53 +0100
Subject: Crosswords for Linux Backtracking recursive crossword generator
References: <8d59hk$6to$1@nnrp1.deja.com>
Message-ID: <m3g0sphfay.fsf@atrus.jesus.cam.ac.uk>

Manuel Gutierrez Algaba <algaba at my-deja.com> writes:

> Comments/improvements are welcome!

I'm just looking at it, but in playing around I got this output:

- - - - - - - - - - - - - - -
- - - - - - - - - - c - - c -
- - - - - - - - - s o l f a -
- - - - - - - - - - n - - r -
- - - - - - - - - - f - - a -
- - - - - - - - - - u - - n -
- - - - - - - - - - s - - t -
- - - z a l a m e r i a - o -
- - - - - - - - - - o - m ? -
- - - - c o h e r e n c i a -
- - - - a - l - - - - - m - -
- - - - p - i n s u l t o - -
- - - m a n o - - - - - - - -
- - - - - - - - - - - - - - -
- - - - - - - - - - - - - - -

That's not right is it?  The `m' and `?' being next to each other.

Cheers,
M.

-- 
  ... but I guess there are some things that are so gross you just have
  to forget,  or it'll destroy something within you.  perl is the first
  such thing I have known.                 -- Erik Naggum, comp.lang.lisp


From cut_me_out at hotmail.com  Thu Apr 13 09:37:32 2000
From: cut_me_out at hotmail.com (Alex)
Date: 13 Apr 2000 09:37:32 -0400
Subject: Python/Perl Popularity (Re: A Mountain of Perl...)
References: <200004102111.HAA03380@envy.fulcrum.com.au> <8d0pcq$7ti$1@nnrp1.deja.com> <dpg0sr37fw.fsf@mraz.iskon.hr> <etdn1mz8drz.fsf@w20-575-109.mit.edu> <dp66tmv3cp.fsf@mraz.iskon.hr>
Message-ID: <etdln2it943.fsf@w20-575-109.mit.edu>

>> I love to use info, too.  Unfortunately, I can't get the latest
>> version of the documentation in that format, so I don't know whether
>> I'm going to be able to use it for much longer.
>
>Well, I don't know.  I'm using the Debian package `python-doc' which
>contains info, so I guess it's possible to create it.

That's true, but that's probably not for the latest version.  If you
download the current distribution and try to make info files from it,
you get errors from the html2info script.  At least, I did.

Alex.


From maybe at lucky.com  Tue Apr 11 11:04:02 2000
From: maybe at lucky.com (Soundwave)
Date: Tue, 11 Apr 2000 17:04:02 +0200
Subject: AF_INET error
Message-ID: <8cvero$g8k$1@porthos.nl.uu.net>

I want to learne some things about socket programming, so I started the
socet programming tutorial on www.python.org

but when I copy te code from some sample programm's and run it, it doesn't
work, it don't recognize the AF_INET thing, is this because I am using
Windows ??

can somebody please explain this to me ?

Or what am I doeing wrong ?>

thanx




From rcunning at acm.org  Wed Apr  5 19:37:06 2000
From: rcunning at acm.org (Robert W. Cunningham)
Date: Wed, 05 Apr 2000 23:37:06 GMT
Subject: Pass by reference ?
References: <1257157531-11108477@hypernet.com>
Message-ID: <38EBCD7D.7030602@acm.org>

Gordon McMillan wrote:

> Robert W. Cunningham wrote:
> 
> > The CS models of "Pass By Reference" and "Pass By Value" are fairly well understood,
> > but they do not seem to map simply and directly to Python.  
> 
> Grrr. Pascal causes brain damage. Python is exactly like Java 
> (with the exception of having no primitive types).
>  
> These notions should state *what* is being passed. In Pascal, 
> that's "value of a slot" or "reference to a slot". To be precise, 
> in Python it is *always* "value of a reference", (think "pointer 
> value", not "value of that which is pointed to").

<sigh>  Let's go back to CS101 and start with at least a MINIMAL set of 
common words, definitions and concepts.

1. Computers contain and operate upon REPRESENTATIONS only.

That is to say, there is no such thing as the "integer number five" in a 
computer, only that which we CHOOSE to REPRESENT the VALUE of the purely 
math-theoretical notion of "5".

In this case, any pattern of bits of which the lower three have the 
value "101" (base 2) with all higher bits being zero is commonly 
accepted as a proper representation of the "integer number five" within 
a computer.

2. REPRESENTATIONS exist in various forms.  (Or, more simply, there is 
more than one way to skin a cat.)

If a RAM location or a CPU register or data on a hard disk contains the 
bit pattern representing the integer number five (hereafter referred to 
by the symbol "5"), we commonly refer to the VALUE of the specific 
REPRESENTATION as being 5.

3. Various implementations of various programming languages variously 
manipulate various VALUES and their various REPRESENTATIONS in various 
ways.  (Except when they vary too much.)

In our particular context (the storage and communication of information 
within the Python language), some non-trivial representations are 
employed to store and communicate values.

Within this context, the PHYSICAL representations themselves are almost 
immaterial:  They could be changed at will and we'd still have Python. 
No matter the word width, no matter the endian-ness, no matter what. 
What is more relevant, and more to the point, are the rules used to 
manipulate the various representations, and the results of applying 
those rules.

In common usage, "Pass By Value" implies that the original value is 
copied from it's identified representation within the caller's "space" 
into another (possibly similar, but physically different) representation 
when passed to a function.  The actual "re-representation" is performed 
before the called function accesses the item, but the actual action may 
be performed within the code space of either the caller or the callee, 
and is determined as a language implementation decision.  The called 
function, however, can still only operate upon the re-representation, 
and has no access to the caller's original representation.

"Pass By Reference", as it is normally used, instead implies that the 
original representation is not "re-represented" in any way, and that 
direct access to that representation is provided to the called function. 
  This access may be direct or indirect, so long as the final result is 
that any change made to the item by the called function is visible to 
the caller after the function returns.  There are many implementation 
strategies for this, including "Copy & Copy Back", "Pointers", etc.  The 
implementation is often selected to meet other needs within the 
language, such as having the ability to pass by reference to functions 
across thread or machine boundaries.

Now, what does Python do?  First, we must acknowledge that Python 
REPRESENTATIONS are significantly more complex than those of C and 
Pascal, and have different properties than those of Java or Perl.  Which 
is to be expected:  Python is a different language.

How are specific Python representations passed to functions?

[I am a newbie and this is probably not even correct Python.  So correct 
my details, but please try to follow my logic.]

Let's pick some representations:

   a=5
   b=(5)
   c=[5]
   d={"five":5}

And let's create some functions to manipulate them.  In this case I 
chose an extremely generic Python function, though I SHOULD choose a 
separate function for each representation to illustrate that the 
function doesn't matter:  The passing does.

   def Nuke(w):
     print w,
     w=None
     print w

Let's call each of these functions:

   Nuke(a)
   Nuke(b)
   Nuke(c)
   Nuke(d)

Look at the output:

   5 None
   5 None
   [5] None
   {"five":5} None

And then look at the original values:

   print a,b,c,d

   5 5 [5] {"five":5}

None changed!  Therefore, representations the functions modified were 
DIFFERENT from those passed, though it is clear that the correct VALUE 
was passed.

This behavior meets the classical definition of "Pass By Value".

What other tests have I omitted, and what do they yield?


-BobC



From smalleys at gte.net  Wed Apr 19 23:00:04 2000
From: smalleys at gte.net (sue)
Date: Thu, 20 Apr 2000 03:00:04 GMT
Subject: Python tools for PDE and Linear algebra
References: <ya0vh1kzsaf.fsf@grindaflet.ifi.uio.no> <38F75987.26AD3FEA@oreilly.com> <ya0og79vsem.fsf@sessrymnir.ifi.uio.no> <Pine.LNX.4.10.10004192110450.13408-100000@us2.mayo.edu>
Message-ID: <38FE72C8.934FB45D@gte.net>

Have you checked

http://www.python.org/topics/scicomp/numbercrunching.html

And SWIG can wrap well-designed libraries quickly.

Sue.


From senn at maya.com  Wed Apr 12 11:14:41 2000
From: senn at maya.com (Jeff Senn)
Date: 12 Apr 2000 11:14:41 -0400
Subject: Huh? func_defaults, default values in function calls
In-Reply-To: "Fredrik Lundh"'s message of "Wed, 12 Apr 2000 13:48:04 GMT"
References: <8cv5qj$35p$1@zinc.intern.netconnect.no> <8cv8ra$hio$1@pegasus.csx.cam.ac.uk> <nMGI4.42394$U4.182659@news1.rdc1.az.home. <38F3EBBF.7B23@marshall.edu> <o8%I4.642$rc9.189485056@newsb.telia.net>
Message-ID: <wkr9cbuza6.fsf@SNIPE.maya.com>

"Fredrik Lundh" <effbot at telia.com> writes:
> Mark Hathaway <hathawa2 at marshall.edu> wrote:
> > >> As a result, if you set a mutable object (list, dictionary, etc) as a
> > >> default, it'll be set when the function is defined and then if the
> > >> function changes it, it'll stay changed for subsequent invokations.
> > >> The function effectively has a "side-effect", of modifying it's own
> > >> default!
> >
> > This way you'll never really know what the default is going to be.
> 
> uhuh?  so you mean evaluating the default value every time the
> function is called would automatically make it easier to know what
> the default is going to be?  what about non-trivial default values?
> what about scoping?  what about side effects? what about per-
> formance?  etc.
> 
> > Ha ha ha ha. that's pathetic.
> 
> really?  I'm pretty sure Guido designed things this way on purpose,
...

Yep. Note (also) that you can use a non-mutable value as the
default if you don't want this behavior. Oh....I mean... if there *are*
such in things in the language...  maybe I should go wade into the
do-we-really-need-immutable-aggregates-in-Python argument....

...maybe not...;-) 

mutably-winkingly-
-Jas





From gbreed at cix.compulink.co.uk  Tue Apr 11 06:05:42 2000
From: gbreed at cix.compulink.co.uk (gbreed at cix.compulink.co.uk)
Date: 11 Apr 2000 10:05:42 GMT
Subject: Using ADO in IIS
References: <yBsI4.1142$uE2.458712@news.pacbell.net>
Message-ID: <8cutdm$3mh$1@plutonium.compulink.co.uk>

> Can someone please help me use ADO from IIS.  I just purchased Mark's 
> new
> book on Win32 but his example on page #254 is for Pythonwin not for
> ASP/Python.  The ASP context requires different object creation syntax.
> I've been asking this question for over six months and no one seems to 
> know
> or care about finding a solution.

I need it to work as well.  Here's a basic script to read and return a 
table:

<%@ LANGUAGE = Python %>
<%
import win32com.client

oconn=win32com.client.Dispatch("ADODB.connection")
# add the correct fields for your database in the next line
oconn.Open ("database=*;DSN=*;ID=*;Password=*;")
objRecords, thing = oconn.Execute (
  "SELECT name, colour FROM fruit ORDER BY name")

Response.Write("<TABLE><TR>")
for objField in objRecords.Fields:
  Response.Write("<TH>"+objField.Name+"</TH>")
Response.Write("</TR>")
while not objRecords.EOF :
  Response.Write("<TR>")
  for objField in objRecords.Fields:
    Response.Write("<TD>"+objField.Value+"</TD>")
  objRecords.MoveNext()
  Response.Write("</TR>")
Response.Write("</TABLE>")

oconn.close
oconn=None
%>

Obviously you'll need this particular table in your database for it to 
work, but you should be able to adapt it.  The only gotcha I've found is 
that Execute returns a tuple.  Otherwise, it's a straight conversion from 
VBScr*pt.

> I've seen many others ask this same question over and over with no 
> resolve.
> Can we work together to finally find an ASP/ADO/Python solution.

Sure.  What problems are you having?

   Graham


From ahagan at speedynet.net  Thu Apr 27 13:19:24 2000
From: ahagan at speedynet.net (Aaron Hagan)
Date: Thu, 27 Apr 2000 13:19:24 -0400
Subject: win32net setting a account to expire
Message-ID: <8e9sub$v7h$1@tracy.nacs.net>

Is there a function to set a user account to expire ..
similar to the command NET USER <USERNAME> /EXPIRE:xx/xx/xx /DOMAIN ?
thanks




From echuck at mindspring.com  Fri Apr 14 18:48:27 2000
From: echuck at mindspring.com (Chuck Esterbrook)
Date: Fri, 14 Apr 2000 18:48:27 -0400
Subject: for item in some_class_object...
References: <4j6ffs0b6ins12qq559q07u28pc6o0mc9l@4ax.com>
Message-ID: <38F7A03B.9FCEFFA9@mindspring.com>

Ben Wolfson wrote:

> What special methods should I define for a class if I want to support
> a for item in ... syntax for it?
>
> --
> Barnabas T. Rumjuggler al-Mumit
>
> Further-more, I have long since put the finishing touches on my own
> official obituary, a 350-page tale of heart-rending loss which
> details, among other things: my meteoric rise from Onion stringer to
> editor to publisher; my assassination at the hands of lady-actress
> Ethel Barrymore's jealous suitors; the thousand days of mourning which
> followed my death, during which a cold rain fell constantly, no bird
> sang, and children were born antlered and hooved; ...
>  -- T. Herman Zweibel

I believe that would be __getitem__ which needs to raise an IndexError when the buck stops short.

class X:
    def __getitem__(self, index):
        if index>=0 and index<10:
            return index
        else:
            raise IndexError


-Chuck



From tismer at tismer.com  Wed Apr 19 08:49:10 2000
From: tismer at tismer.com (Christian Tismer)
Date: Wed, 19 Apr 2000 14:49:10 +0200
Subject: Very Horrible Question About Goto's
References: <6325E90C8B6ED21189080001FA7E2ADA0D4051@IBA1>
Message-ID: <38FDAB46.5469FA9C@tismer.com>


> Richard Jones wrote:
> 
> I'm trying to create a converter from our own internal language (horrid) into
> python.  My only problem so far is that the original language uses *LOTS* of
> goto's.  Is there any way that this can be replicated within python ?

Sure. Use Stackless python and tis continuation module,
and you have a way to express goto's. :-)

Well, to be honest, there is currently no good way
to collect the "labels" of a program as continuations,
but I'm considering how to implement this.
Maybe you can help me with a layout proposal? I would like
to implement multiple-entrypoint-functions in SLP 1.3.

ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaunstr. 26                  :    *Starship* http://starship.python.net
14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     where do you want to jump today?   http://www.stackless.com



From darrell at dorb.com  Sat Apr  1 16:20:17 2000
From: darrell at dorb.com (Darrell)
Date: Sat, 01 Apr 2000 21:20:17 GMT
Subject: Confusion about re lookahead assertions
References: <8c5n76$9le$1@nnrp1.deja.com>
Message-ID: <lKtF4.13255$mV5.117956@typhoon.nyroc.rr.com>

Looks like a negative lookahead assertion doesn't consume anything.
The doc for ?= says this but ?! doesn't.
So the (?![.]) matches the "0". I didn't know that you could use "." instead
of "\."

>From doc:
(?=...)
Matches if ... matches next, but doesn't consume any of the string

--Darrell

"Skip Montanaro"
> I'm trying to use re's lookahead assertion constructs for the first time
> and am more than a bit confused. If I understand them correctly, the
> regular expression r'(?![.])([0-9]+)\s+' should only match digits if they
> are not preceeded by a dot, yet the example below clearly contradicts
> that.
> >>> pat = re.compile(r'(?![.])([0-9]+)\s+')
> >>> print pat.search("$5.00 mem")
> <re.MatchObject instance at 1ff99a0>
> >>> print pat.search("$5.00 mem").group(1)
> 00
> >>> mat = pat.search("$5.00 mem")
> >>> print mat
> <re.MatchObject instance at 1ff93e0>
> >>> print mat.group(1)
> 00
> >>> pat = re.compile(r'(?![-A-Za-z0-9:._])([0-9]+)\s+')
> >>> mat = pat.search("$5.00 mem")
> >>> print mat
> None
> I think mat should be None or mat.group(1) should be '0' in both cases.
> What am I missing?





From robin at jessikat.demon.co.uk  Sat Apr 22 04:35:37 2000
From: robin at jessikat.demon.co.uk (Robin Becker)
Date: Sat, 22 Apr 2000 09:35:37 +0100
Subject: The Simple Economics of Open Source
References: <3900611F.3B525BDA@cfdrc.com> <1255786824-35980039@hypernet.com>
 <f4JdfuAkLKA5EwcQ@jessikat.demon.co.uk>
 <raffael-63EB69.23345321042000@news.ne.mediaone.net>
Message-ID: <M02ZsJAZRWA5EwfD@jessikat.demon.co.uk>

In article <raffael-63EB69.23345321042000 at news.ne.mediaone.net>, Raffael
Cavallaro <raffael at mediaone.net> writes
>In article <f4JdfuAkLKA5EwcQ at jessikat.demon.co.uk>, Robin Becker 
><robin at jessikat.demon.co.uk> wrote:
>
>>Painters
>>exchange ideas, concepts and criticism freely, but compete for
>>customers.
>
>Speaking as a painter, I can tell you that many painters do not share 
>ideas and concepts, but are quite secretive about them. In fact, there 
>is a long history of artists keeping their working methods secret from 
>each other, for precisely the same reason that any other set of business 
>competitors do.
>
>Ralph
>
Clearly not all do; same with hackers. You must know a different set of
artists and painters. The ones I know seem to have a sort of verbal
diarrhoea about how wonderful they are and how they achieved exactly
which effect etc etc etc.

Grinding the paint etc corresponds to real infrastructure like getting
the chips to work better. The problems and solutions of perspective,
style, lighting and the like are discoverable in the product by the keen
observer. It might be harder to discover the secret of 'true' vermilion
or exactly how such accurate perspective effects were achieved.

Artists are forced into a dialogue if not directly with each other at
least with their clients and hence with each other. 
-- 
Robin Becker


From dalke at acm.org  Thu Apr 27 04:28:58 2000
From: dalke at acm.org (Andrew Dalke)
Date: Thu, 27 Apr 2000 02:28:58 -0600
Subject: Import File Madness
References: <8e8ik5$t4v$1@nnrp1.deja.com>
Message-ID: <8e8tht$bbv$1@slb6.atl.mindspring.net>

jsolbrig at my-deja.com asked several questions/comments on the import
statement
>1: Once a python application is large and various classes have static
member
>objects of different types, import statements seem to have a similar logic
to
>C include files. I.E., the question of what order to do the include becomes
>non-trivial. Is there anyway around this or ways to escape this?


That's in the FAQ, at http://www.python.org/doc/FAQ.html#4.37

It basically says, don't do it.  Usually there are ways around it, though
it's harder with class variables.

A tricky way, if you really want to play nasty, is to override
__getattr__, as in:

 class Spam:
   def __getattr__(self, key):
     if name == "class_variable":
       Spam.class_variable = other_package.property
       return other_package.property
     raise AttributeError, key


Can't really help with the other 3 questions.  Haven't seen those
problems.

                    Andrew
                    dalke at acm.org





From mhammond at skippinet.com.au  Tue Apr 11 20:19:20 2000
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Wed, 12 Apr 2000 00:19:20 GMT
Subject: Deadlocking IDLE under Win32
References: <8cuudo$8pj$1@pegasus.csx.cam.ac.uk>
Message-ID: <ciPI4.1680$UB.11667@news-server.bigpond.net.au>

My guess is that the "print" under IDLE is not thread-safe.  Change the
program to not print (or to append items to a queue, which the main thread
then prints) and see if this is the problem.

If so, consider submitting a patch to IDLE.  In Pythonwin, the
sys.stdout.write function checks if it is the main thread, and if not
simply queues the output (module Queue) for the main thread to print.

[This may also explain why I see the threading test hang when running the
Python test suite under IDLE]

Mark.

"James "Wez" Weatherall" <jnw22 at cam.ac.uk> wrote in message
news:8cuudo$8pj$1 at pegasus.csx.cam.ac.uk...
> I've noticed a bug with which a simple multithreaded test program such
as:
>
> import threading
>
> class MyThread(threading.Thread):
>     def run(self):
>         while self.a < 1000:
>             print self.a,
>             self.a=self.a+2
>
> thread1=MyThread()
> thread2=MyThread()
> thread1.a, thread2.a = 0, 1
> thread1.start()
> thread2.start()
>
> This program will print out about twenty numbers and then hang the whole
> interpreter if you run it from IDLE but it runs fine from the
command-line
> version, suggesting some locking problem in the Tk-based print command.
>
> Does anyone have any ideas about this?
>
> Cheers,
>
> --
> James "Wez" Weatherall
>
>
>




From thomas.heller at ion-tof.com  Wed Apr 26 05:44:39 2000
From: thomas.heller at ion-tof.com (Thomas Heller)
Date: Wed, 26 Apr 2000 11:44:39 +0200
Subject: Unsigned integer arithmetic
References: <newscache$51sktf$rzc$1@jenpc07.jennic.co.uk>
Message-ID: <01e701bfaf64$0ae7d860$4500a8c0@thomasnb>

> Hi all,
>
> I'm having a lot of trouble with arithmetic manipulation of numbers above
> 0x7fffffff. Whilst an integer is represented in a 32 bit word, as you'd
> expect, 0x80000000 - 0xffffffff are treated as negative numbers. However,
I
> want to treat a 32 bit number as an unsigned value, and be able to do i =
i
> + 1 for an integer number above 0x7fffffff without getting an overflow.
>

What about writing an Uint32 class? You could implement __add__, __hex__,
and so on...

Thomas




From mhammond at skippinet.com.au  Fri Apr 21 06:47:50 2000
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Fri, 21 Apr 2000 10:47:50 GMT
Subject: finding real $0
References: <OWeCyjAOWCA5Ew86@jessikat.demon.co.uk>
Message-ID: <qlWL4.1038$v85.6158@news-server.bigpond.net.au>

Isnt sys.executable what you are after?

Mark.

"Robin Becker" <robin at jessikat.demon.co.uk> wrote in message
news:OWeCyjAOWCA5Ew86 at jessikat.demon.co.uk...
> Is there any pythonic way to locate the original argument zero of the
> process? I mean the location of python itself not the script. This would
> allow me to launch subscripts without having to rely on the path being
> set up correctly.
> --
> Robin Becker




From mhammond at skippinet.com.au  Sun Apr  2 18:33:08 2000
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Sun, 02 Apr 2000 22:33:08 GMT
Subject: Some bugs in Pythonwin b.128
References: <Pine.OSF.4.10.10003281852310.6437-100000@prfdec.natur.cuni.cz> <_NaE4.6119$V56.24967@news-server.bigpond.net.au> <8F0AAC6BEm69ph83qstudentssuse@127.0.0.1>
Message-ID: <EUPF4.2880$1b4.14419@news-server.bigpond.net.au>

"Per Hogberg" <m69ph83q at students.su.se> wrote in message
news:8F0AAC6BEm69ph83qstudentssuse at 127.0.0.1...
> mhammond at skippinet.com.au (Mark Hammond) wrote in
> <_NaE4.6119$V56.24967 at news-server.bigpond.net.au>:
>
> Actually, if you're one of those cursed few with a
non-US/Brit./Australian
> keyboard you may have to tinker a bit. The events for
parenthesis in the
> config file are wired to the Shift-9 & 0 keys. My keyboard uses
Shift-8 &
> 9.

The Oz one shouldnt make a difference - the only thing we have
over the US keyboards is a beer-holder.

> Try changing these settings in your config file:
>     Shift-9           = <<paren-open>>
>     Shift-0           = <<paren-close>>
>

That is excellent to know.  Does anyone know how I can determine
the keyboard layout?  ie, how can I programatically determine
where the paren keys are?

> to the default file. Not even idle.cfg in the distribution
worked this way.
> I ended up copying the whole default.cfg file. I didn't try
hard so this
> just might work for everyone else :-)

Onm what way didnt this work?  This would be an option I
suppose - provide specifical configurations for specific keyboard
layouts?

Mark.





From culliton at clark.net  Thu Apr  6 18:52:57 2000
From: culliton at clark.net (Tom Culliton)
Date: Thu, 06 Apr 2000 22:52:57 GMT
Subject: More about variables
References: <meh9-5C17A0.10445406042000@news.cit.cornell.edu> <3X3H4.16435$0o4.108310@iad-read.news.verio.net> <meh9-158D1A.14240006042000@news.cit.cornell.edu>
Message-ID: <dz8H4.16465$0o4.108593@iad-read.news.verio.net>

In article <meh9-158D1A.14240006042000 at news.cit.cornell.edu>,
Matthew Hirsch  <meh9 at cornell.edu> wrote:
>Why create named variables?
>
>I always thought that having 20 lists saved more space than having one 
>list of 20 lists.  Maybe I'm wrong.

You are, but it's an honest mistake since Python is different from
many common languages here.  Every Python "variable" is really a
dictionary entry, which means a string for the name, the data, and a
pair of references to them.  So a list of lists saves you the string,
one of the references, and lets you deal with the data more uniformly
making it a win all the way around.

You should probably dig through some of the online docs and tutorials
to get a better handle on how Python represents things internally and
how "variable" names and assignment really work.  Rather than names
being tightly fixed to a particular area of memory they are loosely
and indirectly associated with data objects.  An object can have many
different names during it's lifetime (sequentially or in parallel),
and a name can (sequentially) refer to many different objects.  There
are really no such thing as variables in python only objects and
their (more or less temporary) names.

For example: (the id() function essentially returns the address of the
object refered too.)

Python 1.5.2 (#1, Sep 17 1999, 20:15:36)  [GCC egcs-2.91.66 19990314/Linux (egcs- on linux-i386
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> x = 1
>>> y = x
>>> z = y
>>> id(x)
134882664
>>> id(y)
134882664
>>> id(z)
134882664
>>> y=2
>>> z=3
>>> id(x)
134882664
>>> id(y)
134882616
>>> id(z)
134882604
>>>



From effbot at telia.com  Tue Apr  4 17:33:07 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Tue, 04 Apr 2000 21:33:07 GMT
Subject: Warning switches (was: Python 1.6 alpha 1 released)
References: <Pine.GSO.4.10.10004040035420.21084-100000@sundial> <5E877FD7637AB46E.DABBED9CF3C1D64B.CDDDC05AE7374A78@lp.airnews.net> <5l7leern5v.fsf@eric.cnri.reston.va.us> <38ea1de0$0$17688@personalnews.de.uu.net>
Message-ID: <nctG4.5133$74.83090@newsc.telia.net>

Johannes Stezenbach <yawyi at gmx.de> wrote:
> >However, I'll do so for purely backwards compatible reasons, and I
> >will continue to press for the tuple address as the correct form.
>
> Will python then support a -w switch to support people in writing
> correct programs (i.e. compatible with likely furure language/library
> changes), just like our-favourite-four-letter-P-language?

how about -v<version> ?

there has been talks about compatibility "modes" (probably an option,
but probably also a way to set options programmatically), as a way to
help people prepare for "Python 3000".  I think we might expect some
of it in 1.7, and more of it in 1.8 (if that ever materializes).

but hopefully, Python 3000 will be completely free from modes.

> Seriously: Either accept the 2-arg form of connect() as *correct* for
> all future versions of python or keep the change to make it illegal
> *now*.

guido tried to do the latter.  guess he didn't expect that everyone was
using connect in every single script they'd written ;-)

> Just postponing the change is evil since it will only hurt more people
> in the future, as the number of python users is growing. (CP4E?)

yeah, but at the moment, guido and the rest of the 1.6 team have
more important things to do than to deal with the spanish inquisition.
so I'm afraid it'll have to wait until that -v1.6 switch is in place.

> If it's not broken then don't fix it.

    "I try to avoid breaking user programs, but if something
    needs to be done it really needs to be done."
    -- linus torvalds

</F>




From drew.csillag at starmedia.net  Wed Apr 26 14:33:51 2000
From: drew.csillag at starmedia.net (Andrew Csillag)
Date: Wed, 26 Apr 2000 14:33:51 -0400
Subject: C API - Creating Python Classes
References: <213CDEBFDDA7D3118081009027E40C5607116E@SHEF-MS1>
Message-ID: <3907368F.A4957DF3@starmedia.net>

Mark Pope wrote:
> 
> Hi,
> 
> Using the python C API I can, for example, create an Integer Python object
> with the call:
> 
>         PyObject *pMyInt = PyInt_FromLong( 123 );
> 
> I have a python class which I need to create an instance of. How is this
> done?
> 
> I assume I need something like:
> 
>         PyObject *pModule = PyImport_ImportModule( "MyClassScript" );
> 
>         PyObject *pInstance = NEW_INSTANCE( pModule, args );
> 
Close, but not quite...

PyObject *pModule  = PyImport_ImportModule( "MyClassScript" );
PyObject *classObj = PyObject_GetAttrString( pModule, "MyClassName" );
PyObject *args     = Py_BuildValue("(<argument tuple)", arg1, arg2, arg3
...);
PyObject *instance = PyObject_CallObject( classObj, args );

Obviously, you'd need to make <argument tuple> correspond to the
arguments to the MyClassName constructor.  Not to mention the
appropriate Py_DECREF calls when you're done.

Cheers,
Drew

-- 
print(lambda(q,p):'pmt:$%0.2f'%(q*p/(p-1)))((lambda(a,r,n),t:(a*r/
t,pow(1+r/t,n*12)))(map(input,('amt:','%rate:','years:')),1200.0))



From embed at geocities.com  Wed Apr 19 10:55:27 2000
From: embed at geocities.com (Warren Postma)
Date: Wed, 19 Apr 2000 10:55:27 -0400
Subject: Pythonware Website???? Gone on vacation with Starship? ;-))
References: <20000419154202.B480@owl.rhein-zeitung.de>
Message-ID: <AQjL4.9437$HG1.254772@nnrp1.uunet.ca>

I'm getting an error 403:

http://www.algonet.se/error_403.html

Forbidden
... snip...




From kcazabon at home.com  Sat Apr 15 01:43:11 2000
From: kcazabon at home.com (Kevin Cazabon)
Date: Sat, 15 Apr 2000 05:43:11 GMT
Subject: Imaging Library & Numerical Python
References: <38F7F865.9C86399C@engineering.uiowa.edu>
Message-ID: <PjTJ4.137181$Dv1.1488334@news1.rdc1.bc.home.com>

I use them all the time together... are you experiencing any problems?



"Juerg Tschirren" <jtschirr at engineering.uiowa.edu> wrote in message
news:38F7F865.9C86399C at engineering.uiowa.edu...
>
> Does anybody have experience with getting the Imaging Library and
> Numerical Python working together? What I would like to do is importing
> an imaging with the Imaging Library and then using functions from both
> -- the Imaging Library and Numerical Python -- on this image.
>
> Juerg




From wright7 at my-deja.com  Thu Apr  6 12:01:54 2000
From: wright7 at my-deja.com (Rob Wright)
Date: Thu, 06 Apr 2000 16:01:54 GMT
Subject: __builtin__.open on windows NT
Message-ID: <8cicd6$1s0$1@nnrp1.deja.com>

i'm new to python, so be prepared...i'm having trouble with open() on
windows NT.  my script uses the os library and it appears that open()
in my script is conflicting with an open() function in the os lib.  so
i tried to use __builtin__.open(), but i get this error "NameError:
__builtin__"

any ideas,

rob


Sent via Deja.com http://www.deja.com/
Before you buy.


From PClaerhout at CREO.BE  Tue Apr 11 05:37:31 2000
From: PClaerhout at CREO.BE (Pieter Claerhout)
Date: Tue, 11 Apr 2000 11:37:31 +0200
Subject: Saving a copy of a remote NT event log
Message-ID: <2B1262E83448D211AE4B00A0C9D61B03BA7295@MSGEURO1>

Hello,

does anyone know if it's possible to save a copy of the NT-event log
from a remote machine using Python? Normally, you have to do a
save as from that machine, and then copy over the logfile, but I want
to automate this using Python. Any ideas?

Kind regards,


Pieter

[ Pieter Claerhout :: CreoScitex Response Centre :: Applications support ]
[ Pieter_Claerhout at creoscitex.com    ::    http://www.creoscitex.com/ ]
[ CreoScitex Europe  ::   Excelsiorlaan 21   ::  1930 Zaventem  ::   Belgium ]




From hellan at acm.org  Thu Apr 13 03:36:19 2000
From: hellan at acm.org (Jon K Hellan)
Date: 13 Apr 2000 09:36:19 +0200
Subject: Using python with Corba
References: <eieJ4.8408$H7.704991@brie.direct.ca>
Message-ID: <871z4awiz0.fsf@parus.parus.no>

"Fanny Yeung" <fannyy at idmail.com> writes:

> Hi everybody,
> 
> I am new python and I am particular interest in use python with corba. After
> some research, I find out omniOrb provide interface 'omniOrbpy'. Any one
> know where I can find it ?

http://www.uk.research.att.com/omniORB/omniORB.html

> Also, besides omniOrb2, is there any other Corba that Python can use ?

Fnorb is almost pure Python.

ILU does a lot of languages.

ORBit is extremely fast.

> Anyone know TAO can be used in Python ?

Don't think so. But a Python ORB can interoperate with TAO.

Regards

Jon K?re


From tismer at tismer.com  Thu Apr  6 15:18:41 2000
From: tismer at tismer.com (Christian Tismer)
Date: Thu, 06 Apr 2000 21:18:41 +0200
Subject: Beginning to be disapointed
References: <Aw4H4.8078$pR.68745@carnaval.risq.qc.ca> <mt3dozt6sh.fsf@astron.berkeley.edu>
Message-ID: <38ECE311.6B64E4C@tismer.com>


Johann Hibschman wrote:
> 
> Hugues Demers writes:
> 
> > Can somebody help me on this? I'm using python to read big arrays
> > (100x2000) of float. I manipulate those arrays, I do small
> > operations on them. That is ok. I do others smalls calculations not
> > involving the big arrays and then the interpreter hang ! The cpu
> > percentage used by python goes up and there is nothing else to be
> > done except to kill the interpreter. I have no idea what's going
> > on. Do I need to free memory in some way ? Am I doing something
> > wrong ? There is nothing special about the calculations involving
> > the big arrays. The only thing special is that they are big.
> 
> Are you using the Numeric package?  I've never seen behaviour like that.
> 
> 1. are you using Numeric or array?
> 2. how are you reading them in?
> 3. how are you calculating them?  Numeric math ops or explicit loops over
>    raw arrays?
> 4. could something be keeping a reference to the arrays?
> 5. what OS are you on?
> 6. version of python?

7. How much memory does your machine have?

It sounds like memory becomes very fragmented, and Python is
swapping all the time to get some space between scattered
objects.

ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaunstr. 26                  :    *Starship* http://starship.python.net
14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     where do you want to jump today?   http://www.stackless.com



From greg at quokka.demon.co.uk  Wed Apr 26 18:40:46 2000
From: greg at quokka.demon.co.uk (greg at quokka.demon.co.uk)
Date: Wed, 26 Apr 2000 22:40:46 GMT
Subject: plugins for browsers?
References: <3904ccb0.32134281@news.demon.co.uk> <8e43dt$ods$1@newshost.accu.uu.nl>
Message-ID: <3907706c.14550359@news.demon.co.uk>

On 25 Apr 2000 12:35:09 GMT, m.faassen at vet.uu.nl (Martijn Faassen)
wrote:


>> 	Are there any plugins or other means of running Python in Web
>> browsers? ...
>
>JPython:
>
>http://www.jpython.org
>

Actually I know about and use JPython - it's very good - but I
wondered if it's possible to run CPython in browsers. I've just been
playing with the Squeak plugin, and think that having Squeak run in a
web browser is rather neat.



From pumilia at my-deja.com  Sat Apr 29 05:20:20 2000
From: pumilia at my-deja.com (pumilia at my-deja.com)
Date: Sat, 29 Apr 2000 09:20:20 GMT
Subject: MySQLdb adn GroupLounge
Message-ID: <8ee9gf$6ig$1@nnrp1.deja.com>

Hi, I am trying to instal GroupLounge on my linux/debian box
but such program has a python script that still uses the old mysql
python modules.
I am completely new to python and mysql; reading the python docs,
I was able to change mySQL to MySQLdb (the first line in the included
cutting from my script file) but i do not know how to change the
remaining lines.


Here is my GroupLounge script:
  #!/usr/bin/env python

  import MySQLdb
  from string import *

  import config
  import engine.wslib
   .....

My relevant python setup is:

ii python-base     1.5.2-6    An interactive object-oriented scripting
ii python-mysqldb  0.1.1-1  A Python interface for MySQL


your help will be apprecitiated

thank you
Pol


Sent via Deja.com http://www.deja.com/
Before you buy.


From pinard at iro.umontreal.ca  Sun Apr  2 23:44:58 2000
From: pinard at iro.umontreal.ca (=?ISO-8859-1?Q?Fran=E7ois_Pinard?=)
Date: 02 Apr 2000 23:44:58 -0400
Subject: Unicode program representation
In-Reply-To: "Neil Hodgson"'s message of "Mon, 3 Apr 2000 11:25:14 +1000"
References: <015c01bf9d0b$76f9b2a0$59dd3fcb@neil>
Message-ID: <oqem8n4zlx.fsf@titan.progiciels-bpi.ca>

"Neil Hodgson" <neilh at hare.net.au> writes:

> su = u"hi ?"

> I think this should be changed to interpreting the literal as a UTF-8
> literal.  The advantage here is that non-roman string literals become
> a natural part of the language.

It might not be convenient expecting all sources to be expressed using
UTF-8, which would be a consequence of your suggestion.  A lot of people
use other representations, and Python 1.6 is doing well, in my opinion.
Despite it _does_ favour UTF-8 for a representation, it does not force it.

Instead of writing:

   u"hi ?"

which is indeed a way to write Unicode strings using 7-bits, or even 8-bits
in a Latin-1 environment, you might write:

   unicode("hi ?")

(without the `u' prefix) to trigger an UTF-8 to Unicode conversion.

> Saving the scripts as UCS-2 shows that the interpreter is unable to
> deal with UCS-2 scripts, which is what I expected. I think very few
> people will be creating script files in UCS-2, instead preferring to
> keep source code in UTF-8 files.

We might guess right now that a writing style will soon develop in which
Python sources could be recoded between charsets (or encodings, to use
the Python terminology), with the sole need of changing a single variable
in each module.  The `unicode' function, for example, accepts a second
argument stating the encoding, when not UTF-8, and if that argument is
always used, and holds the same module-wide variable all over, this could
be easily achieved.

-- 
Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard





From aahz at netcom.com  Fri Apr 21 13:02:57 2000
From: aahz at netcom.com (Aahz Maruch)
Date: 21 Apr 2000 17:02:57 GMT
Subject: Good style for multi-valued returns
References: <889A6EB6AE399393.619B0539BFCAAE0F.5C6D110C3DA908F7@lp.airnews.net> <m3hfcwh6rd.fsf@atrus.jesus.cam.ac.uk> <43E1059BD6C613C1.19161A8DC1C439F1.3E710DAA95C4F8D8@lp.airnews.net>
Message-ID: <8dq1k1$6na$1@slb2.atl.mindspring.net>

In article <43E1059BD6C613C1.19161A8DC1C439F1.3E710DAA95C4F8D8 at lp.airnews.net>,
Cameron Laird <claird at starbase.neosoft.com> wrote:
>
>Particular thanks to the three people who've already gently written to
>remind me that what I really want in the complex-and-big case is to
>return an *object* (but why do you all call it a class?  To my ear, it
>truly sounds like a confusion to say, "class", rather than, "instance
>of a class").

Well, if I wanted to be technically accurate, I'd say "class object" or
"class instance", but I was being lazy and only wrote the first five
characters.  ;-)

>I think I still have special cases where I want to move dictionaries
>around, but I'm learning that their proper domain is much smaller than
>I used to believe.

Well, yeah.  For passing information around, dictionaries are
appropriate when the key/value linkage is external rather than an
artifact of your program logic -- basically using a dict as an in-memory
database.  In addition, I find dictionaries extremely useful for
managing % strings.

I think the use of dicts as a substitute for a class is a bit of a
Perlism.  Classes should be used any time you'd create a C struct.
--
                      --- Aahz (Copyright 2000 by aahz at netcom.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

Why is this newsgroup different from all other newsgroups?  --Aahz

Return-Path: <sabren at manifestation.com>
Delivered-To: python-list at python.org
Received: from chi.pair.com (chi.pair.com [209.68.1.37])
	by dinsdale.python.org (Postfix) with ESMTP id 418881CD1C
	for <python-list at python.org>; Fri, 21 Apr 2000 03:26:42 -0400 (EDT)
Received: from localhost (sabren at localhost) by chi.pair.com (8.9.1/8.6.12) with ESMTP id DAA13285; Fri, 21 Apr 2000 03:26:44 -0400 (EDT)
X-Envelope-To: python-list at python.org
Date: Fri, 21 Apr 2000 03:26:43 -0400 (EDT)
From: "Michal Wallace (sabren)" <sabren at manifestation.com>
X-Sender: sabren at chi.pair.com
To: Undisclosed.recipients: ;
Subject: help wanted: linkwatcher.com
Message-ID: <Pine.BSF.4.10.10004210259150.13031-100000 at chi.pair.com>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Sender: python-list-admin at python.org
Errors-To: python-list-admin at python.org
X-BeenThere: python-list at python.org
X-Mailman-Version: 2.0beta3
Precedence: bulk
List-Id: General discussion list for the Python programming language <python-list.python.org>

Hey all,

I'm looking for some volunteers to help with a www.linkwatcher.com, a
python-based website which monitors various weblogs (constantly
updated personal pages that link to sites on the net) and newsfeeds
and updates users when they've changed.

Linkwatcher is growing. What we want to do now is to build it into a
powerful reporting and monitoring tool for the weblog community. Later
on, it could even grow into a distributed system capable of indexing
and monitoring many thousands of pages and newsfeeds, and making all
of this data available in real time.

The system is free for everyone. There's no ads on the site, though we
may sell some premium services in the future.

Along the way, we're building some fairly powerful, open source
tools for python developers. Here's an outline of the challenges
we're looking to tackle:


-- core technologies --

* Generic web library for python - stuff like sessions,
  authentication, etc. I've been working on a port of the 
  PHP base classes to python, and have it working with a 
  MySQL database. It's about 80% usable.

* User, Node, and Content objects - these are objects that
  control user management and globs of text. User works with
  the authentication system above to let users sign up and becom
  members of the system .Node is for tree-like structures (eg, for
  categorizing blogs).. Content may or may not apply to linkwatcher
  proper, but I'm bundling it in with the others because I've
  been planning to wrap them in a set of objects which more or less
  competes with Manila. (codename 'zikebase' but I'm looking for
  a better name)

* linkwatcher objects - this is what I was talking about in
  the post the other day. the objects are (at least) Url, Blog,
  Channel, and History. They provide a high level interface to
  records in the MySQL database.

-- Once those are mostly working, the really fun stuff can start: --

* Favorites - allow users to pick the blogs and RDF channels
  they want to see.

* Collaborative Filtering - given the above, it should be possible
  to say "people who like the blogs you like also like..."

* Registration system - submit a blog to the list

* Administration system - approve and categorize blogs, administer channels, etc

* Better search engine - keep track of links on a page, so that we can
  see, for example "new links to http://myblog.com" (vs permanent links
  to myblog.com).. So, if Joe Blogger has a permanent link to MyBlog.com, but
  sees something interesting on it and adds another link inside HIS blog,
  the new link will show up in the search engine.

* Custom channels - if you monitor keywords / links to your blog, why
  not just have the search results show up on your custom page?

* Scoop index - since we track all the links, we can do fun reports off 
  of them. what's the hottest link, who linked it first, etc, etc..

* Bloggerbot - blogger publishes an XML feed of all its pages, and when
  they were updated. This data ought to go into linkwatcher.

* Pitas / Groksoup / edithispage - work with these services so we can
  get updates for their pages, too... We'll only actually fetch pages
  for indexing if they are both updated AND on the list of blogs..

* Distribute the bot - this opens up a LOT of possibilities for making
  linkwatcher a more powerful search engine.

* Distribute the server ??????????? Hey, if napster and freenet can do
  this.. maybe there's some way we can do it too... I don't know if
  this is even possible.

* Blogger profiles - articles/pictures/etc about the people who run
  the blogs..

* Promote the site - get people using it, etc..

* Make it look pretty (and make it usable!) - UI, graphics, etc..

* channelbot - need a parser for RDF/RSS in python.. (there may already be
  one from the python xml-sig, but I don't think so)

* channelMAKERbot - turn html into rdf for people who don't have 
  xml generators but want headlines

The source code for everything we have so far is online at
http://www.zike.net/zike/dev/ with cvs access available from
sourceforge.

Anyway, we're looking for people. There's quite a few interested
people. Right now, most of that interest is from the weblog community
itself. We've got quite a few developers, but not many who know
python. I've made the offer to help any of them learn python, and
I'll extend the same offer to any python newbies around here who are
interested in helping out.

Anyone who's interested can email me (sabren at manifestation.com) or
join the project's egroup: http://www.egroups.com/group/weblog-monitor


Thanks,


- Michal
-------------------------------------------------------------------------
http://www.manifestation.com/         http://www.linkwatcher.com/metalog/
-------------------------------------------------------------------------





From n99hbn at nat.sdu.dk  Thu Apr  6 07:38:17 2000
From: n99hbn at nat.sdu.dk (Henrik Berg Nielsen)
Date: Thu, 6 Apr 2000 13:38:17 +0200
Subject: encoding type in a2ps to pretty-print
Message-ID: <8chtbq$ohg$1@news.net.uni-c.dk>

Hi, I am complete newbie and just starting to learn Python. I have just
installed a2ps for dos so I can pretty-print all my small programs. I have
one problem though. When I pretty-print my code via a2ps all places where I
have a '<=' are turned into the similar mathematical sign, that only takes
up one space on the line, likewise all '==' are converted to the
mathematical "by definition" sign (that is 3 lines on top of eachother). Why
is this so? It is really annoying and I would really like a2ps to
pretty-print my code the way it looks on my monitor. I think it has
something todowith the encoding that a2ps uses,but I'm not sure.

Thanks for any help

-Henrik Berg Nielsen





From donn at u.washington.edu  Wed Apr 19 17:24:22 2000
From: donn at u.washington.edu (Donn Cave)
Date: 19 Apr 2000 21:24:22 GMT
Subject: modifying os.environ
References: <slrn8frsu1.3bu.quinn@euro.ugcs.caltech.edu> <p6qhfczo50y.fsf@informatik.hu-berlin.de> <slrn8f9mn1.ven.quinn@zloty.ugcs.caltech.edu> <8di2il$10oq$1@nntp6.u.washington.edu>
Message-ID: <8dl866$rr2$1@nntp6.u.washington.edu>

Quoth quinn at euro.ugcs.caltech.edu (Quinn Dunkan):
| On 18 Apr 2000 16:30:13 GMT, Donn Cave <donn at u.washington.edu> wrote:
...
|> - Raise exception on exit status, complete with error output as
|>   exception value.  A bit of a pain to implement, and again not
|>   not always the right thing to do, but usually it is the right
|>   thing.
|
| I don't understand... are you saying the close() method should throw the
| exception?  What's wrong with just checking it's return value normally?  How
| is a child you started exiting exceptional?  And if the child has been started
| asynchronously, do you mean throw an exception from the SIGCHLD handler, or
| waitpid()?

The exception, I should have said, is on exceptional exit status,
i.e. non-zero.  In theory this means the program failed, terminated
abnormally, etc.  Of course exit(0) raises no exception.

In C, we'd check the exit status (I hope), just as we'd check the
return (I hope) from a system function like unlink() or whatever.
In Python, we get exceptions instead and don't have to insert
``I hope'' into our discussions about how things work.  I find
this as useful with external commands as with internal functions.

In my version of this, I will raise the exception after finding
the exit status via waitpid().  I never trap SIGCHLD, don't think
it's a very good idea, so luckily I don't have to think about what
would happen if I had to raise an exception from a signal handler.

	Donn Cave, University Computing Services, University of Washington
	donn at u.washington.edu


From nobody at nowhere.nohow  Sat Apr 22 09:22:35 2000
From: nobody at nowhere.nohow (Grant Edwards)
Date: Sat, 22 Apr 2000 13:22:35 GMT
Subject: How2 use curses for data input
References: <38FF6DE2.5751423A@sprint.ca> <slrn8fusjd.99a.grant@grante.comtrol.com> <38FF8F9C.9D533E74@sprint.ca>
Message-ID: <vIhM4.515$wJ1.16679@ptah.visi.com>

Miles Thompson <milesthompson at sprint.ca> wrote:

>I located it on the RH system here in the office at home. The
>target machine is Debian, and I think I can install RPM on it
>so that if need be, I can get the RPM. Debian doesn't seem to
>have a snack package.

I've put the newest sources (the ones that come with RH6.2) at

  <ftp://ftp.visi.com/users/grante/stuff/newt-0.50.8.tar.gz>

It looks like the standard untar/configure/make procedure
should work for anybody with python stuff in /usr/lib/python-1.5. 
If your python installation is somewhere else, you might want
to modify Makefile.in before doing the "configure".

-- 
Grant Edwards                   grante             Yow!  I FORGOT to do the
                                  at               DISHES!!
                               visi.com            


From darrell at dorb.com  Mon Apr  3 01:37:01 2000
From: darrell at dorb.com (Darrell)
Date: Sun, 2 Apr 2000 22:37:01 -0700
Subject: NameError when assigning dictionary values
References: <20000402231757.6604.qmail@web2105.mail.yahoo.com>
Message-ID: <004301bf9d2e$a34a5f80$6401a8c0@dorb>

"lewst"
>
> >>> one, two = [], []
>
> >>> mydict = { ".": one, "+": two }
>
> This seems error prone if the the number of variables grows very large
> as it will for me.
>
So you have a large number of things to add to a dictionary an you don't
like the available syntax.

No problem.
Write in what ever syntax you like, then parse it into a syntax Python can
handle.

In a text file:
Key:value
Key:value

Then something like this can read the file into the dictionary.

dic={}
lines=open(theFile).readlines()
for l in lines:
    key, value = string.split(l,':')
    dic[key]=value

For large data sets I've generated the python source instead of this kind of
loop. Then import the resulting file to generate a .pyc file. Then you have
a huge dictionary that is ready to go.

--Darrell







From mfletch at tpresence.com  Mon Apr 24 02:01:38 2000
From: mfletch at tpresence.com (Mike Fletcher)
Date: Mon, 24 Apr 2000 02:01:38 -0400
Subject: Slashdot review of Python Programming On Win32
Message-ID: <CC59868A0A76D311B1F50090277B321C1F32A5@VRTMAIL>

http://slashdot.org/article.pl?sid=00/04/17/1147228&mode=nocomment

Generally positive review, got me thinking I should buy the book.

Enjoy all,
Mike
__________________________________
 Mike C. Fletcher
 Designer, VR Plumber
 http://members.home.com/mcfletch



From cut_me_out at hotmail.com  Fri Apr 28 12:04:16 2000
From: cut_me_out at hotmail.com (Alex)
Date: 28 Apr 2000 12:04:16 -0400
Subject: How can I get the name of an object???
References: <8e712k$65c$1@nnrp1.deja.com> <etd8zy03m44.fsf@w20-575-108.mit.edu> <8e9090$c4d$1@nnrp1.deja.com> <slrn8gglsv.ep7.scarblac-spamtrap@flits104-37.flits.rug.nl> <8eca76$19c$1@nnrp1.deja.com>
Message-ID: <etdwvlixldr.fsf@w20-575-109.mit.edu>

> I have an instance of the navigation-class, but for the "exec"code I
> need the name of the *instance*.

If you have some way of referring to the instance in the main body of
the code, why not just give it a name of your own choosing that you can
use in the exec command?

e.g.

instance = get_the_relevant_instance ()

exec ("do_something (%s)" % 'instance')

> >>> from shadow import * #IMPORTANT for me

Just out of curiosity, how come?

Alex.


From jmg at ecs.soton.ac.uk  Wed Apr  5 04:27:04 2000
From: jmg at ecs.soton.ac.uk (Jacek Generowicz)
Date: Wed, 05 Apr 2000 09:27:04 +0100
Subject: Pass by reference ?
References: <38E864C9.E1EE94F7@ecs.soton.ac.uk>
Message-ID: <38EAF8D8.4B7C43B4@ecs.soton.ac.uk>

Jacek Generowicz wrote:

> Hi,
>
> I'm just trying to familiarize myself with Python.
> In Magnus Lie Hatland's Instant Python, he
> mentions that `all parameters in Python are passed
> by reference'.

[snip]

Thank you for your answers.

I would welcome a critique of the following summary:

Python variables are implemented as references to
values.

Assignment (potentially) alters the reference rather
than the value to which it refers.

Objects can be classified into two sets: mutalble and
immutable.

Mutation of a mautable passed into a function will be
reflected in the mutation of the object which was
passed into the function; achieving similar behaviour
for immutables is more complicated.

Jacek




From gduzan at gte.com  Thu Apr 13 15:00:10 2000
From: gduzan at gte.com (Gary D. Duzan)
Date: Thu, 13 Apr 2000 15:00:10 -0400
Subject: 1st non-trivial program - gentle criticism requested
References: <1e90r16.1sd2ts41e7ur6kN@gershwin.bio.ic.ac.uk> <8d4tgu$11qa$1@nntp6.u.washington.edu>
Message-ID: <38F6193A.66284778@gte.com>

Donn Cave wrote:
> 
> There's a builtin max() that I think can do the work of your findMax.
> 
> I have plenty of these in my programs:
> 
>    print "The alphabet is '" + gAlphabet + "'"
> 
> but today I prefer
> 
>    print "The alphabet is '%s'" % gAlphabet
> 
> because of the greater tolerance for objects that are not really
> strings but have string representations.  (OK, it's a string, but
> what will happen if you later change this object to something else?)

   Well, if you change it to a tuple, you're going to have problems,
anyway. If you really want to be more change-proof, you'll want:

    print "The alphabet is '%s'" % (gAlphabet,)

or:

    print "The alphabet is '%s'" % str(gAlphabet)

or:

    print "The alphabet is '" + str(gAlphabet) + "'"

This potential stumbling block makes me wonder if allowing the
singleton case was really wise.

					Gary Duzan
					GTE Laboratories


From rumjuggler at home.com  Wed Apr  5 22:02:37 2000
From: rumjuggler at home.com (Ben Wolfson)
Date: Thu, 06 Apr 2000 02:02:37 GMT
Subject: Why should I switch to Python? - Infinity of Primes
References: <38EAC8C4.ADC25640@cosc.canterbury.ac.nz> <000d01bf9ec5$3fb90820$3e2d153f@tim>
Message-ID: <a9snesk870r1djmo4u9qh9bo86qp9eb2bv@4ax.com>

On Wed, 5 Apr 2000 02:07:38 -0400, "Tim Peters"
<tim_one at email.msn.com> wrote:

>not-that-archie-Pu-will-be-fooled-ly y'rs  - tim

Doesn't he grep?

-- 
Barnabas T. Rumjuggler

Who is general Pao, and why are we eating his chicken?
 -- possibly Legion, in one o' dem wacky-ass Wildcard books


From just at letterror.com  Sat Apr 15 06:38:03 2000
From: just at letterror.com (Just van Rossum)
Date: Sat, 15 Apr 2000 11:38:03 +0100
Subject: [ANNOUNCE] Garbage collection for Python
In-Reply-To: <38F826CF.DBDC71EF@zipzap.ch>
References: <20000407221750.A29279@acs.ucalgary.ca>	
 <m3r9cagx3c.fsf@atrus.jesus.cam.ac.uk>
 <l03102809b51d41895a80@[193.78.237.135]>
Message-ID: <l03102801b51df49c16a0@[193.78.237.135]>

At 10:22 AM +0200 15-04-2000, Boris Borcic wrote:
>Just van Rossum wrote:
>>
>> - Will Ware and I have worked hard on a complete rewrite of the
>>   uthread module. It is a *vast* improvement over the old one.
>>   It offers a very nice thread object and several handy threading
>>   widgets. It already can do stuff threading.py can only *dream*
>>   about, like pausing (blocking) a thread from another thread or
>>   posting an exception from another thread (!).
>
>Simple examples of useful new idioms that this permits ?

Not sure about general new idioms, but I use it in an experimental version
of my MacPython IDE like so:
When the user starts a script, it gets executed in a new thread, allowing
the main GUI thread to continue to do it's work, as well as the user. The
cool thing is, the user can *pause* a running script at any given time, eg.
to inspect how the program is doing so far. Posting exceptions is useful in
a similar way: if user want to interrupt a running program, it just pushes
the "stop" button, and a KeyboardInterrupt (or SystemExit, I still don't
know what would be better) exception is posted from the "main" thread to
the running thread, making it stop gracefully. The builtin
KeyboardInterrupt mechanism always causes the KeyboardInterrupt to appear
in a random thread, which is not good.

Just





From moshez at math.huji.ac.il  Fri Apr 28 01:17:27 2000
From: moshez at math.huji.ac.il (Moshe Zadka)
Date: Fri, 28 Apr 2000 08:17:27 +0300 (IDT)
Subject: run detached
In-Reply-To: <3908B8CB.97A0755D@inka.de>
Message-ID: <Pine.GSO.4.10.10004280817140.8383-100000@sundial>

On Fri, 28 Apr 2000, Michael [iso-8859-1] Str?der wrote:

> HI!
> 
> How can I automatically detach from console after starting up with
> SocketServer? (E.g. web server with access and error logging to
> files.)

Fork and exit from the parent.
--
Moshe Zadka <mzadka at geocities.com>. 
http://www.oreilly.com/news/prescod_0300.html
http://www.linux.org.il -- we put the penguin in .com




From alwagner at tcac.net  Wed Apr 12 19:53:33 2000
From: alwagner at tcac.net (Albert Wagner)
Date: Wed, 12 Apr 2000 18:53:33 -0500
Subject: redirecting stdout to printer
Message-ID: <38F50C7D.1E7ED48E@tcac.net>

How do I specify my printer as stdout?  Know of a module that has basic
prettyprinting with headings and pagenumbers?

-- 
Small is Beautiful


From grisha at ispol.com  Thu Apr 20 22:27:10 2000
From: grisha at ispol.com (Gregory Trubetskoy)
Date: Thu, 20 Apr 2000 22:27:10 -0400
Subject: suggestion on an operator
Message-ID: <Pine.LNX.4.21.0004202210480.9838-100000@saturn.ispol.com>

Hello there - 

I'm looking for suggestions for the best operator to use in deliniating
modules from what's inside of them in a configuration file for the next
version of mod_python (a.k.a httpdapy).

Basically, I'd like to give the user an option to specify a handler for a
particular Apache event in Apache configuration. A handler will be a
Python function (perhaps a class, but that's not relevant) in a Python
module.

E.g. mypackage.mymodule will have an authenticate() function.

If I want to configure apache to use that handler, I would put something
like this in the .htaccess config file:

PythonAuthenHandle  mypackage.mymodule::authenitcate

The question is - is "::" a good choice of syntax, or would there be any
problem with it? What I like about it is that it is somewhat intuitive for
someone who's familiar with C++. What I don't like about it is that it
isn't exactly a "scope resolution operator" in this context (and also that
it looks too much like perl)

Other alternatives: ":" this is simpler, but could be confused with a
Python function/class declaration. "/" - this would be consistent with
ZPublisher, but it makes it look like it's a file. "->" ???

Grisha



From scherbi at bam.com  Mon Apr 17 11:54:07 2000
From: scherbi at bam.com (Bill Scherer)
Date: Mon, 17 Apr 2000 11:54:07 -0400
Subject: Another question
References: <dVFK4.41$vr2.5817@news.uswest.net> <38FB2EEB.A32AF88C@bam.com>
Message-ID: <38FB339F.604364AD@bam.com>

I just did a little test and found this to be substantially fatser than the
other three:

import random
randint = random.randint

r = []

for i in xrange(1,10000):
    r.append(randint(1,10))

for i in r: print i
#


note that I dropped the second arg to xrange down by 2 orders.  I couldn't
wait for a million to scroll by ;-) , but the technique should scale.

I tested each algorithm 3 times and average the times.  My first attempt
was actually the slowest of the first three at 3.59.  Jeff's first (the
first listed) came in at 3.43, and his second at 3.48.  The above came in
at 2.49, ~44% faster than my first, and ~37% faster than Jeff's 3.43.

Bill Scherer wrote:

> Jeff -
>
> The main reason (AFAIK) you don't want to do 'from <mod> import *' is to
> avoid namespace clashing.
> A better way is this:
>
> #untested
> import random
> randint = random.randint
>
> for i in xrange(1,1000000):
>     print randint(1,10)
>
> #
>
> This method allows you to control exactly which names you import into
> your local namespace form the module.  The above should also be a little
> faster due to the use of xrange.  Se the docs for why.
>
> -Bill
>
> Jeff Massung wrote:
>
> > Okay, another quick question. If it is quicker to do:
> >
> > import random
> > from random import randint
> > for i in range(1,1000000):
> >     print randint(1,10)
> >
> > than this:
> >
> > import random
> > for i in range(1,1000000):
> >     print random.randint(1,10)
> >
> > Why would I not do "from <module> import *" with every module I load?
> >
> > Jeff
> >
> > --
> > http://www.python.org/mailman/listinfo/python-list
>
> --
> http://www.python.org/mailman/listinfo/python-list




From nmm1 at cus.cam.ac.uk  Fri Apr  7 10:04:45 2000
From: nmm1 at cus.cam.ac.uk (Nick Maclaren)
Date: 7 Apr 2000 14:04:45 GMT
Subject: PPython
References: <38ede548.353776233@news.oh.verio.com>
Message-ID: <8ckptt$j8s$1@pegasus.csx.cam.ac.uk>

In article <38ede548.353776233 at news.oh.verio.com>, morse at harborcom.net (Kevin Dahlhausen) writes:
|> So many web hosting sites only allow Perl scripting, I got to thinking.  That's
|> often a dangerous thing.   Would it be possible to implement the Python virtual
|> machine in Perl?   OK, anything is possible,  but I guess the real question is
|> does anyone have a feel for the scope of such a project?  Say using a native
|> interpreter to pre-compile the Python code to pyc files and the perl cgi vm to
|> run them?      Feasible, or just a silly idea?

If you want a really silly idea, it probably wouldn't be hard to
convert Python's C source to Perl, because the former looks pretty
clean and therefore could be translated semi-automatically without
needing a full-blown C to Perl compiler.  That's the easy bit.

When you put 130,000 lines of translated source into Perl, you
would discover things about Perl that you didn't want to know and
even its author had not suspected :-)

And I suspect that those Web sites would disembowel you after their
systems had melted down ....



Regards,
Nick Maclaren,
University of Cambridge Computing Service,
New Museums Site, Pembroke Street, Cambridge CB2 3QG, England.
Email:  nmm1 at cam.ac.uk
Tel.:  +44 1223 334761    Fax:  +44 1223 334679


From mwh21 at cam.ac.uk  Thu Apr 13 18:42:56 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 13 Apr 2000 23:42:56 +0100
Subject: How do I undo an import??
References: <7QrJ4.185421$Hq3.4353581@news2.rdc1.on.home.com>
Message-ID: <m3d7nthbbj.fsf@atrus.jesus.cam.ac.uk>

"Tom" <tom-main at REMOVEME.home.com> writes:

> I'm a C++ programmer, new to Python, working with v1.5.2 on Windows.
> 
> I run the command-line interpreter, then I type:
> 
> >>> from NetCfg import *
> 
> to import my extension.  Now I want to unload my NetCfg DLL without exiting
> the interpreter.  How do I do this?

You don't, in general.

del sys.modules['NetCfg']

goes some of the way, but it's unlikely it goes far enough to let the
dll be unloaded from memory.
 
> Also, I type the above import command every time I start the interpreter.
> Is there some way to get have this command executed automatically?

Set the environment variable "PYTHONSTARTUP" to point to a file
containing commands you want executed.

At least, that's what I do on Linux.

Cheers,
M.

-- 
  ... but I guess there are some things that are so gross you just have
  to forget,  or it'll destroy something within you.  perl is the first
  such thing I have known.                 -- Erik Naggum, comp.lang.lisp


From mhammond at skippinet.com.au  Fri Apr 21 23:38:09 2000
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Sat, 22 Apr 2000 03:38:09 GMT
Subject: Pythonwin com question
References: <8dqver$l4v$1@nnrp1.deja.com>
Message-ID: <B89M4.1434$v85.8046@news-server.bigpond.net.au>

You could try passing different clsctx values to
win32com.client.Dispatch - my guess is that Python's default isnt quite
the same as VBs...

Let me know if changing this works...

Mark.

"Sam Schulenburg" <samschul at pacbell.net> wrote in message
news:8dqver$l4v$1 at nnrp1.deja.com...
> I have a com object that I can excercise with Visual Basic as such:
>
> Dim WApp as New Winx32Lib.Winx32App    ' Get an instance of Winx32App
> WApp.ShowDemoBox("This is a demo box") ' Call the client ShowDemoBox()
>
> From Python I try the following which starts the client application and
> then timesout with the following error msg:
>
> >>> import win32com.client
> >>> MyCLSID = '{1A762221-D8BA-11CF-AFC2-508201C10000}'
> >>> WApp = win32com.client.Dispatch(MyCLSID)
> Traceback (innermost last):
>   File "<interactive input>", line 1, in ?
>   File "D:\Python\win32com\client\__init__.py", line 94, in Dispatch
>     dispatch, userName = dynamic._GetGoodDispatchAndUserName
> (dispatch,userName,clsctx)
>   File "D:\Python\win32com\client\dynamic.py", line 66, in
> _GetGoodDispatchAndUserName
>     return (_GetGoodDispatch(IDispatch, clsctx), userName)
>   File "D:\Python\win32com\client\dynamic.py", line 57, in
> _GetGoodDispatch
>     IDispatch = pythoncom.CoCreateInstance(IDispatch, None, clsctx,
> pythoncom.IID_IDispatch)
> com_error: (-2146959355, 'Server execution failed', None, None)
> >>>
>
> Any help would be appreciated
> Samuel W. Schulenburg
>
>
> Sent via Deja.com http://www.deja.com/
> Before you buy.




From effbot at telia.com  Sun Apr  2 07:50:30 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Sun, 02 Apr 2000 11:50:30 GMT
Subject: Python 1.6 alpha 1 released
References: <200003312130.QAA04361@eric.cnri.reston.va.us> <38E5205F.DE811F61@roguewave.com> <Wf9F4.4423$74.70231@newsc.telia.net> <38E58D2D.384F1546@Lugoj.Com> <S9kF4.4472$74.71414@newsc.telia.net> <38E6B929.28816494@roguewave.com>
Message-ID: <auGF4.914$n68.186931712@newsb.telia.net>

Bjorn Pettersen wrote:
> I haven't seen anyone claim that examples in the manual are not
> documentation...

in earlier releases of the library reference, some examples
didn't even work.  you still think they're formally binding?

> It would be very nice if "someone" could do that, but based on the
> rethoric I've seen from you on this issue, and the absence of comment
> from Guido, I certainly don't feel confident that it would be accepted.

you don't know that until you've tried it...

you may have to work on your presentation, though. thinking
about the following may help you come up with a stronger case:

-- how does typical socket.connect code look?

-- what exception do you get after this change?  how
   does the error message look?  can anyone else figure
   out what's wrong?

-- is there any chance that people might be expecting
   the resulting exception from socket.connect (or the
   code invoking it)?  or in other words, is it likely that
   there will be a try/except clause around the call, that
   will misbehave (thus hiding the actual error message)?

-- can anyone but the original author fix the problem
   without breaking the program?

-- is socket.connect used in all kinds of scripts and
   applications, or mostly in support libraries?  is the
   number of affected programs equal to the number
   of code changes that have to be made?  (if your
   answer is "yes", please tell us why).

-- the user base is growing exponentially.  so is the
   number of python implementations.  given that,
   when is the right time to tighten up the core inter-
   faces?  now, later, or never?

when you've done this, replace "socket.connect" with
"list.append", and try again.

if you still think that this change is a disaster, please submit
a patch, and a rationale for accepting it.  all you have to do
is to look for NO_STRICT_LIST_APPEND in the 1.6a1 sources,
and copy that code to the socket module.

shouldn't take you more than 10 minutes, or so.

(btw, don't forget updating the documentation).

</F>




From bjorn at roguewave.com  Wed Apr  5 13:13:54 2000
From: bjorn at roguewave.com (Bjorn Pettersen)
Date: Wed, 05 Apr 2000 11:13:54 -0600
Subject: Extending Python with C-Libs
References: <8cfpl8$9b2$1@newnews.ilk.de>
Message-ID: <38EB7452.D3BD8A38@roguewave.com>

Check out http://www.swig.org

-b

Frank Rotermund wrote:
> 
> Hello
> 
> I have a Library written in C with its Headerfiles but without the
> Sourcecode.
> Does anybody know an easy way to extend Python with this Library?
> 
> Thanks
> Frank Rotermund
> 
> --
> http://www.python.org/mailman/listinfo/python-list



From gandreas at delver._REMOVE_TO_REPLY_.com  Fri Apr  7 11:57:30 2000
From: gandreas at delver._REMOVE_TO_REPLY_.com (Glenn Andreas)
Date: Fri, 07 Apr 2000 10:57:30 -0500
Subject: Tuples -- who needs 'em
References: <00e101bf9d80$a8384d20$74eb0b18@stcla1.sfba.home.com> <va3G4.1110$n68.190454272@newsb.telia.net> <037f01bf9e55$6d328760$74eb0b18@stcla1.sfba.home.com> <8cfa75$fho$1@newshost.accu.uu.nl> <38ED4292.2EE8D3CA@cosc.canterbury.ac.nz> <38ED8EF2.B0835BE1@acm.org>
Message-ID: <gandreas-AEF36E.10573007042000@news.visi.com>

In article <38ED8EF2.B0835BE1 at acm.org>, "Robert W. Cunningham" 
<rcunning at acm.org> wrote:

> Greg Ewing wrote:
> 
> > Unfortunately, Python *does* need 1-tuples. If
> > only ASCII had more kinds of brackets, Guido might
> > have chosen differently. Maybe if Python 3000 uses
> > Unicode for source files...
> 
> Well, if they can be syntactically disambiguated from Python relational
> comparison operators, "<" and ">" can be used as "brokets".
> 

Funny, I was thinking the exact same thing.  Other than needing some 
extra spaces, it should work (after all, C++ uses them as well).  
However, depending on the details of the parser (and it's been way too 
long since I've brushed up on parser technology), you might have 
problems with:

   x = < a, a > b >  # (a,a>b) in old syntax

At the very least, this visually is jarring and you need to mentally 
"reparse" when you get to the second ">".  Another alternative is to 
switch to multi-character brackets, perhaps:

   x = (* a, b *)

or:

   x = <| a, b |>

With two characters, there could be hundreds of possible "bracket" 
pairs.  Unfortunately, that makes your code look like it has been 
infested with "smileys".  I seem to remember that some other languages 
using them though (besides old Pascal style comments, that is).


From cut_me_out at hotmail.com  Sun Apr 23 21:40:43 2000
From: cut_me_out at hotmail.com (Alex)
Date: 23 Apr 2000 21:40:43 -0400
Subject: Why doesn't O'Reilly use parrots on its python books?
Message-ID: <etdvh186ztw.fsf@w20-575-115.mit.edu>

Given their venerable standing in Python lore, it surprises me that none
of O'reilly's python books have a parrot on the cover.  There must be
lots of types of parrots to go around, even if some O'reilly books
already have them.  And the python book could be distinguished by the
fact that its parrot is lying on its back with its feet curled up...

A really agressive looking rabbit would work, too.

Alex.


From kcazabon at home.com  Tue Apr 18 00:12:54 2000
From: kcazabon at home.com (Kevin Cazabon)
Date: Tue, 18 Apr 2000 04:12:54 GMT
Subject: Any Web-Mail Gateway (written in Python)?
References: <6D8A17398E28D3119F860090274DD7DB4986A3@pces.cadlab.it>
Message-ID: <ahRK4.147245$Dv1.1585605@news1.rdc1.bc.home.com>

Are you looking for something like a fetchmail utility?

There's a good one called "getmail" written in Python that's available at:

http://www.qcc.sk.ca/~charlesc/software



Kevin.


"Alessandro Bottoni" <Alessandro.Bottoni at think3.com> wrote in message
news:6D8A17398E28D3119F860090274DD7DB4986A3 at pces.cadlab.it...
> I'm looking for a web-mail gateway (written in Python, if possible) for
> making accessible the Postfix mail boxes of my users through my PyApache
web
> server (both running under Linux, of course).
> Any suggestion?
> TIA
> ----------------------------------------------------------------
> Alessandro Bottoni (Alessandro.Bottoni at Think3.com)
> Web Programmer @ Think3 inc. (www.think3.com)
> I do not speak for think3 and they return the favour
>
>




From warlock at eskimo.com  Tue Apr  4 21:59:11 2000
From: warlock at eskimo.com (Jim Richardson)
Date: Tue, 04 Apr 2000 18:59:11 -0700 (PDT)
Subject: Why should I switch to Python? - Infinity of Primes
In-Reply-To: <NDBBKEGCNONMNKGDINPFGEJDDFAA.infonuovo@email.com>
Message-ID: <XFMail.20000404185911.warlock@eskimo.com>

On 04-Apr-2000 Dennis E. Hamilton wrote:
> The standard approach is a proof by contradiction starting from the
> assumption that there is a largest prime.
> 
> It's usually one of the earliest proofs in books on Number Theory.  Euclid
> provided a proof and a version is given in section 4.2 of "Concrete
> Mathematics: A Foundation for Computer Science" by Ronald L. Graham, Donald
> E. Knuth, and Oren Patashnik.  Addison-Wesley (Reading, MA: 1989).   ISBN
> 0-201-14236-8.
> 



Thanks for the great responce! I learned math as well as Python bits on the 
python NG, cool... :)



Jim Richardson
        Anarchist, pagan and proud of it
WWW.eskimo.com/~warlock
        Linux, because life's too short for a buggy OS.




From grant at nowhere.  Tue Apr 18 12:52:18 2000
From: grant at nowhere. (Grant Edwards)
Date: Tue, 18 Apr 2000 16:52:18 GMT
Subject: Integer solutions to linear equation?
References: <slrn8fovo2.nq.grant@grante.comtrol.com> <38FC8CCF.B8B786CB@pacbell.net>
Message-ID: <slrn8fp4m0.16t.grant@grante.comtrol.com>

In article <38FC8CCF.B8B786CB at pacbell.net>, Bruce Wolk wrote:

>ax + by = c, where a, b, and c are integers, and the solution
>(x,y) must be an integer, is one of the simplest Diophantine
>equations. There is an algorithm, but at the moment I can't
>remember it. I do recall that if the greatest common divisor of
>a and b divides c, then there are infinitely many solutions,
>otherwise there are none.  Thus, if c = 1, there are always
>infintely many solutions if a and b are relatively prime, as in
>your example.
>
>You probably can find the algorithm in any basic book on number
>theory.

I had a feeling that this required more number theory than I
know (an amount close enough to 0 that it can be safely
ignored).  :)

-- 
Grant Edwards                   grante             Yow!  I've got a COUSIN
                                  at               who works in the GARMENT
                               visi.com            DISTRICT...


From lvw at lfpt.rwth-aachen.de  Wed Apr 19 07:20:49 2000
From: lvw at lfpt.rwth-aachen.de (Lars von Wedel)
Date: Wed, 19 Apr 2000 13:20:49 +0200 (MET DST)
Subject: testing true condition
Message-ID: <200004191120.NAA04238@gromit.rwth-aachen.de>


Hi,

You can do something like this if you want an easier way to iterate
over a number of lines of a number of files

import fileinput

for line in fileinput.input():
    print line,

This statement iterates over all lines of all files submitted via the
command line. Look into the documentation for module fileinput, it
offers a number of other interesting stuff. You could e.g. call

 test.py *.txt > concat.txt

to generate a concatenated list of all .txt files with the above.

Don't forget the ',' at the end of the print statement. Otherwise
each line will have a blank line after it when you read text files.

Lars



Remco Gerlich writes:
 > Gregoire Welraeds wrote in comp.lang.python:
 > > I know that I can't do something like the following:
 > > 
 > > >>> while (line= f.readline()):
 > > 	...
 > > 
 > > as i could in C.  Is there any replacement tricks ?
 > 
 > The standard Python idiom is:
 > 
 > while 1:
 >    line = f.readline()
 >    if not line:
 >       break
 > 
 > There are also wrapper classes around file input, so you can use it in a for
 > loop, but it doesn't read in the whole file at once (same difference as with
 > range and xrange). Don't remember its name, but it's not hard to roll your
 > own.
 > 
 > -- 
 > Remco Gerlich,  scarblac at pino.selwerd.nl
 >   Murphy's Rules, "Try a firecracker":
 >    In TSR's Top Secret, it usually takes three grenade explosions to kill
 >    the average player-character-agent.
 > -- 
 > http://www.python.org/mailman/listinfo/python-list
 > 




From grant at nowhere.  Mon Apr 24 13:58:32 2000
From: grant at nowhere. (Grant Edwards)
Date: Mon, 24 Apr 2000 17:58:32 GMT
Subject: Running shell programs from Python
References: <38FC64DA.7D297563@ffi.no> <8di47s$mkg$1@nntp6.u.washington.edu> <390485A8.B5ADAAE9@sage.att.com>
Message-ID: <slrn8g92q5.5d6.grant@grante.comtrol.com>

In article <390485A8.B5ADAAE9 at sage.att.com>, Garry Hodgson wrote:
>Donn Cave wrote:
>
>> The references are mainly UNIX man pages pipe(2), fork(2), fcntl(2)
>> (look for dup2), exec(2) and wait(2).
>> 
>> Basically, the sequence is repeat ( open pipes, fork, close pipes )
>> After a fork, the parent and child both have both ends of each pipe
>> open, and it's important to close the unused "write" ends so the
>> pipe will generate an EOF when its real writer exits.
>
>i've always been curious why this wasn't encapsulated as a single
>command in the os module.  i use it infrequently enough that i always
>need to go look the idiom, or skip it for simpler things.  seems like
>a single command that returned a process id, and a pair of file
>descriptors would have lots of uses.

You mean like popen2?

-- 
Grant Edwards                   grante             Yow!  I like your SNOOPY
                                  at               POSTER!!
                               visi.com            


From larrygg at primenet.com  Wed Apr 12 19:59:01 2000
From: larrygg at primenet.com (Larry Gantz)
Date: Wed, 12 Apr 2000 16:59:01 -0700
Subject: How to escape escape?
Message-ID: <8d32kl$sjg$1@nnrp02.primenet.com>

I'm new to the snake.
I think I would like to be able to turn off and on the escape character
processing so the snake would perform escape processing when I want it.

It seems that Python always has escape processing.

Thanks for any help.




From oli at rz-online.net  Fri Apr 14 13:03:08 2000
From: oli at rz-online.net (Oliver Andrich)
Date: Fri, 14 Apr 2000 19:03:08 +0200
Subject: Python 1.6a2 crashes in test_fork1 on my Mandrake Linux 7 machine
In-Reply-To: <20000414154718.A15664@xs4all.nl>; from thomas@xs4all.net on Fri, Apr 14, 2000 at 03:47:18PM +0200
References: <20000413010344.A13931@owl.rhein-zeitung.de> <38F658C6.66AD9CA5@clark.net> <20000414154718.A15664@xs4all.nl>
Message-ID: <20000414190308.A30502@owl.rhein-zeitung.de>

Hi,

On Fri, Apr 14, 2000 at 03:47:18PM +0200, Thomas Wouters wrote:
> I dont know enough about threads to say anything useful about this, but it
> is kind of strange that you dont see the same problem under Python 1.5.2. My
> redhat boxes pass the test, though the BSDI boxes dont (but threading is
> screwed in more than one way under most BSDI versions.) I haven't tried
> compiling 1.5.2 under BSDI 4.1 (the newest, which seems to work reasonably
> ok with threading) though, so it might be a change in python's
> thread-handling...

this is the interesting thing. The system is the same where I run python 1.5.2
and 1.6a2 on. Threading code seems to be the same in the source distribution.
As far as I checked it. It is getting stranger and stranger. 

> Wasn't Mandrake based on Redhat ? Does it use a different/strange libc or
> something ? i dont quite see why there should be a difference in rh 6.1 and
> mandrake, in this context.

Well, they use the same glibc and so should behave equivalent. Only possible
source of errors is the different gccs they use. In python 1.5.2 the gcc of
Mandrake 6 killed the long integer code when optimizations were turned on.
Without it worked. May be we have a again a code generation problem on
Mandrake 7.

Bye, Oliver



From mwh21 at cam.ac.uk  Wed Apr 19 11:43:54 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 19 Apr 2000 16:43:54 +0100
Subject: Very Horrible Question About Goto's
References: <C5A7720BDEB7D3118A5200E018C200A037EA43@EXCHANGE>
Message-ID: <m3g0sihz9h.fsf@atrus.jesus.cam.ac.uk>

Gaetan Corneau <corg at copernic.com> writes:

> I was surprised to learn that Python doesn't have a goto statement.
> Would it be difficult to add?

No. The necessary opcodes are there (well, the one you need:
JUMP_ABSOLUTE).  You could use bytecodehacks to implement it, if you
liked...

I doubt it's a good idea though.

Cheers,
M.

-- 
  "declare"?  my bogometer indicates that you're really programming
  in some other language and trying  to force Common Lisp into your
  mindset.  this won't work.           -- Erik Naggum, comp.lang.lisp


From emile at fenx.com  Sun Apr  9 12:55:48 2000
From: emile at fenx.com (Emile van Sebille)
Date: Sun, 9 Apr 2000 09:55:48 -0700
Subject: Python paradigms
References: <8cmr70$b0e$1@pegasus.csx.cam.ac.uk> <8cntqp$upi$1@slb0.atl.mindspring.net> <8cocgb$iuk$1@pegasus.csx.cam.ac.uk> <8cokt5$opj$1@slb1.atl.mindspring.net> <38EFE45A.2E866C58@mindspring.com>
Message-ID: <0a9301bfa244$75afe200$01ffffc0@worldnet.att.net>

This construct has been occasionally suggested
as one option:

>>> print ['falseresponse','truresponse'][1==2]
falseresponse
>>> print ['falseresponse','truresponse'][1==1]
truresponse

HTH,

Emile van Sebille
emile at fenx.com
-------------------


----- Original Message -----
From: Chuck Esterbrook <echuck at mindspring.com>
Newsgroups: comp.lang.python
To: <python-list at python.org>
Sent: Saturday, April 08, 2000 7:00 PM
Subject: Re: Python paradigms


>
> I miss ?: and in fact, find myself writing extra if-else statements
because of it. I think ?: is a pretty natural construct.
>
> Maybe Python 3K....
>
> -Chuck
>
>
> --
> http://www.python.org/mailman/listinfo/python-list
>





From Robert at nospam.edu  Sun Apr 30 21:36:11 2000
From: Robert at nospam.edu (Robert)
Date: Mon, 01 May 2000 01:36:11 GMT
Subject: Discrepency between JPython and CPython
References: <m3k8hfjlex.fsf@boss.ksu.edu>
Message-ID: <fc5P4.22$kM.6868@news>

You should go to the JPython site (http://www.jpython.org/) and read up on
the differences between CPython and JPython.

Bob

"Venkatesh Prasad Ranganath" <rvprasad at cis.ksu.edu> wrote in message
news:m3k8hfjlex.fsf at boss.ksu.edu...
> Hi,
>
> Consider the expression str(1.14 * 200).
> In CPython it yields 228.0
> In JPython it yields 227.99999999999997
>
> Shouldn't this result in consistent values across variuos implementation?
> If so, which of the implementation is buggy?
> I think it is JPython, may be I am wrong.
>
> --
> Venkatesh Prasad Ranganath
>
> e-mail: rvprasad at cis.ksu.edu, rvprasad at ksu.edu
>
> Address: Department of Computing and Information Science,
>          234 Nichols Hall,
>          Kansas State University,
>          Manhattan, KS 66506
>
> Phone: (785)-532-6350 Extn: 48
>




From phd at phd.russ.ru  Wed Apr  5 06:56:55 2000
From: phd at phd.russ.ru (Oleg Broytmann)
Date: Wed, 5 Apr 2000 10:56:55 +0000 (GMT)
Subject: access form from included cgi
Message-ID: <Pine.LNX.4.21.0004051048140.28309-100000@fep132.fep.ru>

Hello!

   I have a need to include a CGI in an HTML (Apache):
<!--#include vitrual="/cgi-bin/search.py"--> (for example; and name the
HTML search.html).

   I found that I cannot access form data (QUERY_STRING) when I call the
HTML with parameters: http://my.site/search.html?date=21-12-1967.
   Deeper investigation showd that QUERY_STRING is always empty, but Apache
passed form data in QUERY_STRING_UNESCAPED.

   Found this, I put the following code:
if not os.environ.get("QUERY_STRING", ""):
   qs = os.environ.get("QUERY_STRING_UNESCAPED", "") # method == INCLUDED
   if qs:
      os.environ["QUERY_STRING"] = qs

before form = cgi.FieldStorage().

   I think I can put corresponding patch for cgi.py, but I want to discuss
it here before commiting the patch...

Oleg.            (All opinions are mine and not of my employer)
---- 
    Oleg Broytmann      Foundation for Effective Policies      phd at phd.russ.ru
           Programmers don't die, they just GOSUB without RETURN.




From ronjeffries at acm.org  Thu Apr  6 05:19:40 2000
From: ronjeffries at acm.org (Ronald E Jeffries)
Date: Thu, 06 Apr 2000 05:19:40 -0400
Subject: What/where is HTMLBuilder?
Message-ID: <CB8AAF150D049DFB.1B47358A250E7184.84110B3E40F5C2FB@lp.airnews.net>

I've searched my docs and python.org for HTMLBuilder with no luck.
Where, please?

Thanks!

Ron Jeffries
http://www.XProgramming.com


From P.J.W.S.VRIJLANDT at INT.azg.nl  Wed Apr 19 05:02:47 2000
From: P.J.W.S.VRIJLANDT at INT.azg.nl (P.J.W.S. Vrijlandt)
Date: Wed, 19 Apr 2000 10:02:47 +0100
Subject: (Fwd) Re: Formatting Excel cells with PythonCOM
Message-ID: <1E141210786@nds-3.azg.nl>

On  6 Apr 00 at 23:45, Mark Hammond wrote:

> This code is from the win32com test suite (in testMSOffice.py)
> 
>  # test dates out with Excel
>  xl.Cells(5,1).Value = "Excel time"
>  xl.Cells(5,2).Formula = "=Now()"
> 
>  import time
>  xl.Cells(6,1).Value = "Python time"
>  xl.Cells(6,2).Value = pythoncom.MakeTime(time.time())
>  xl.Cells(6,2).NumberFormat = "d/mm/yy h:mm"
>  xl.Columns("A:B").EntireColumn.AutoFit()
> 

I recently found a problem with this (I didn't run the test suite
though). On my computer (Win95, Excel2000), the NumberFormat propery
only accepts setting the time in my locale's language (i.e. Dutch).
This is rather unfortunate, because the macrorecorder writes in
English.
Patrick Vrijlandt



From arnold at dstc.edu.au  Tue Apr  4 07:53:27 2000
From: arnold at dstc.edu.au (David Arnold)
Date: Tue, 04 Apr 2000 21:53:27 +1000
Subject: opening a serial port at a given baudrate. How? 
In-Reply-To: Your message of "Tue, 04 Apr 2000 06:10:40 GMT."
             <slrn8ej283.b3t.warlock@raven.myth> 
Message-ID: <200004041153.VAA15869@piglet.dstc.edu.au>

-->"Jim" == Jim Richardson <warlock at eskimo.com> writes:

  Jim> I can't seem to find any info on opening a serial port in
  Jim> python!. (I am using linux for developement, but would like to
  Jim> be able to use the app on any python system, it is text only,
  Jim> no gui, at least for now. )

there is no portable way to open a serial port.  this is not least
because there is no portable way to name your serial ports, but goes
further than that.

on unix (including linux), just open the device special file
corresponding to the serial port of interest (ie. /dev/ttyS0) and you
can then control it using fcntl.ioctl() function and otherwise read
from and write to it like a normal file object.

on win32 there is a serial I/O module (serialmodule?) that a quick
deja search will locate for you.

i'd recommend the POSIX serial programming guide (any search engine
should find it) as a good primer for Unix serial programming ...



d



From dworkin at ccs.neu.edu  Sun Apr 16 20:52:01 2000
From: dworkin at ccs.neu.edu (Justin Sheehy)
Date: 16 Apr 2000 20:52:01 -0400
Subject: pattern-based string expansion
In-Reply-To: sue's message of "Thu, 13 Apr 2000 03:09:19 GMT"
References: <38F52C61.A5D72F89@gte.net>
Message-ID: <vndog79ft1q.fsf@camelot-new.ccs.neu.edu>

sue <smalleys at gte.net> writes:

> I need to generate a (largish) set of strings from a
> pattern, sort of like glob with csh-style expansion.
> 
> Given "foo{,bar,baz/[aced]{,x,z}}", generate:
> 
> foo
> foobar
> foobaz/a
> foobaz/c
> foobaz/e
> foobaz/d
> foobaz/ax
> foobaz/az
> 
> etc.
> 
> The manual doesn't seem to have this glob(3) option.

Python has glob, but neither Python's glob nor the glob(3) UNIX
library function does what you describe.  Both of them produce all
pathnames that match the given glob, not all possible strings that
could match the given glob.

In fact, if you retain the expressivity of glob, the function you
describe is not practically possible.  What should it return if given
the glob "*"?

-Justin

 



From jeremy at cnri.reston.va.us  Fri Apr 28 13:08:51 2000
From: jeremy at cnri.reston.va.us (jeremy at cnri.reston.va.us)
Date: Fri, 28 Apr 2000 17:08:51 GMT
Subject: parser and tokenize modules: sample code?
References: <3909B42D.2E564923@uab.edu>
Message-ID: <8ecgim$8q2$1@nnrp1.deja.com>

In article <3909B42D.2E564923 at uab.edu>,
  Shae Erisson <shapr at uab.edu> wrote:
> Can someone point me to sample code for the parser and tokenize
modules?
> As Moshe Zadka says, the parser module doesn't seem easy to use. I've
> read through the example.py for the parser module, but I'd like to see
> more code if someone has some.

You can find an elaborate example in the CVS repository
under nondist/src/Compiler.  The module compiler.transformer
uses the parser module to generator a more-easily-used
parse tree.

http://cvs.python.org/cgi-bin/viewcvs.cgi/python/nondist/src/Compiler/

-- Jeremy Hylton <http://www.python.org/~jeremy/>


Jeremy


Sent via Deja.com http://www.deja.com/
Before you buy.


From geek+ at cmu.edu  Wed Apr 19 16:45:27 2000
From: geek+ at cmu.edu (Brian E Gallew)
Date: Wed, 19 Apr 2000 16:45:27 -0400 (EDT)
Subject: Very Horrible Question About Goto's
In-Reply-To: <20000419175310.N15664@xs4all.nl>
References: <C5A7720BDEB7D3118A5200E018C200A037EA43@EXCHANGE> <20000419175310.N15664@xs4all.nl>
Message-ID: <200004192045.QAA09539@smtp3.andrew.cmu.edu>

> Delivered-To: python-list at python.org
On Wed, Apr 19, 2000 at 11:17:21AM -0400, Gaetan Corneau wrote:
> I was surprised to learn that Python doesn't have a goto statement.

I've found that if you really want goto functionality, use
exceptions.  Either that, or figure out what the code is doing, and
then write it correctly.  YMMV.

-- 
=====================================================================
| JAVA must have been developed in the wilds of West Virginia.      |
| After all, why else would it support only single inheritance??    |
=====================================================================
| Finger geek at cmu.edu for my public key.                            |
=====================================================================
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 266 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20000419/ae1fcd7d/attachment.sig>

From mhammond at skippinet.com.au  Wed Apr 26 20:52:33 2000
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Thu, 27 Apr 2000 00:52:33 GMT
Subject: Non SSH accounts on starship?
Message-ID: <lbMN4.5092$v85.31026@news-server.bigpond.net.au>

Starship seems to have been running quite well for a while now, and a few
days ago we hear the disk is unlikely to be recovered, so we should start
rebuilding!

But I can't log on.  How can I determine what my account has been reset
to?  Im getting _lots_ of enquiries about my starship page, so Im keen to
get it back online ASAP...

Thanks,

Mark.




From cut_me_out at hotmail.com  Tue Apr 25 09:22:01 2000
From: cut_me_out at hotmail.com (Alex)
Date: 25 Apr 2000 09:22:01 -0400
Subject: Efficient coding advice?
References: <8e2sjn$te4l$1@swen.emba.uvm.edu>
Message-ID: <etd3doafh8m.fsf@w20-575-109.mit.edu>

> 1.  Is there an easy way to split a large string so that every-other
> element gets assigned to an array?

Not really easy, but it might be fast:
from operator import getitem

def odds_and_evens (l):
    length = len (l)
    even_indices = range (0, length, 2)
    odd_indices = range (1, length, 2)
    even_elements = map (getitem, len (even_indices) * [l], even_indices)
    odd_elements = map (getitem, len (odd_indices) * [l], odd_indices)
    return even_elements, odd_elements

> 2.  Is there an easy way to turn an entire array of characters into
> their ord value?

map (ord, l)

Alex.



From raoul at ez-ways.com  Fri Apr  7 15:19:45 2000
From: raoul at ez-ways.com (Raoul-Sam Daruwala)
Date: Fri, 07 Apr 2000 19:19:45 GMT
Subject: Regular Expression Help for Python Newbie.
Message-ID: <38EE3799.DA0A9638@ez-ways.com>

I have a problem. I wrote a python program to parse HTML files using the
HTMLParser and all that I need to do with the files can be done very
easily using this wonderful class. Kudos to the authors!

My problem is that one of the sets of files that I'm trying to parse has
badly formatted tables. Now when I do a view source on the files I can
see the problem clearly. It's quite simple, the tables in these files
starts out properly formatted but after a standard header the script
than generates them leaves out the <TR> tag. This is incredible to me
because both Netscape and IE read can view the tables properly. But the
HTMLParser dies on them.

Let me be a little more explicit. The problem files, and there are over
300 of them, have the following structure to their tables.

<TABLE>
<TR>
<TD> foo</TD> <TD>bar</TD>
</TR>
<TD> 1</TD><TD>2</TD></TR>
<TD>3</TD><TD>4</TD></TR>
</TABLE>

Incredibly, this will display correctly (or rather incorrectly as is
shouldn't display at all) in most browsers.

What I need to do to fix this is run a quick pre-processor and using the
re module replace all occurences of

    </TR> junk </TR> with
    </TR><TR> junk </TR>
where junk does not contain the tag <TR>

Can anyone tell me what the re for this is? I can't seem to get anything
to work. right now.

Regards,

Raoul-Sam



From P.J.W.S.VRIJLANDT at INT.azg.nl  Tue Apr 18 12:32:04 2000
From: P.J.W.S.VRIJLANDT at INT.azg.nl (P.J.W.S. Vrijlandt)
Date: Tue, 18 Apr 2000 17:32:04 +0100
Subject: telnetlib and NT vs. AIX
In-Reply-To: <8dhnr5$glm$1@nnrp1.deja.com>
Message-ID: <1D0BE221345@nds-3.azg.nl>

On 18 Apr 00 at 13:27, dbeagle at my-deja.com wrote:

> I am trying to figure out what the difference is
> between telnet to Windows NT and AIX.  As you can
> see below the calls made to each telnet server
> are the same but for some reason (unknown to me)
> after I enter username in the NT telnet session
> it never returns asking for the password.
> Testing on the AIX system worked as expected. The
> printout from both sessions is shown below.
> 
> Thank you for your help
> db
> 

[WinNT]
> Telnet(WinNT4SVR_1,23): IAC DO 253
> Telnet(WinNT4SVR_1,23): IAC WILL 251

I suppose the problem is here. telnetlib does not attempt to do any 
protocol negotiation. In this point, telnetlib does not adhere to the 
relevant rfc's, because some requests must be anwered.

I have not found time to implement an adequate option negotiation 
yet. That's partly because I don't need it: Answering any DO with 
WONT and WILL with DONT was sufficient for my problem.

Note that standard telnet does not ECHO and requires GO-AHEAD. You 
might want to enable these options on the server side.

Any chr(0) in the input stream gets lost in telnetlib and this is not 
correct if you recieve binary files.

I will mail you my solution (not ready for publication) seperately 
when I'm at home.


Patrick Vrijlandt



From gslee111 at daisy.kwangwoon.ac.kr  Sun Apr  2 20:57:00 2000
From: gslee111 at daisy.kwangwoon.ac.kr (Gang Seong Lee)
Date: Mon, 3 Apr 2000 09:57:00 +0900
Subject: getting a key without an echo and Enter
Message-ID: <Q0SF4.8$tu.536@news.hananet.net>

Hi,

How can I get a key input without having it echoed.
The following statement echoes and waits until I press Enter key.

ch = sys.stdin.read(1)

Can I get the key as soon as I press it without an echo. (No Enter key)

Thanks in advance

Gang





From ljz at asfast.com  Tue Apr  4 18:35:44 2000
From: ljz at asfast.com (Lloyd Zusman)
Date: 04 Apr 2000 18:35:44 -0400
Subject: New Features in Python 1.6
References: <200004011740.MAA04675@eric.cnri.reston.va.us> <gHrF4.16072$0o4.101062@iad-read.news.verio.net> <8cdbvg$3vi$1@news.udel.edu> <38ea3fda.47259642@news.nikoma.de>
Message-ID: <ltd7o5cx4v.fsf@asfast.com>

nospam.newton at gmx.li (Philip 'Yes, that's my address' Newton) writes:

> On Tue, 4 Apr 2000 14:30:06 -0400, "Terry Reedy" <tjreedy at udel.edu>
> wrote:
> 
> >> Python strings can now be stored as Unicode strings.  To make it easier
> >> to type Unicode strings, the single-quote character defaults to creating
> >> a Unicode string, while the double-quote character defaults to ASCII
> >> strings.
> >
> >' = 1 byte/char, " = 2 bytes/ char is more straightforward.
> 
> Yes, but this depends on what you mean by "Unicode". You have to
> represent those Unicode characters in bytes, somehow, and how many bytes
> you need depends on the format. Also, full Unicode is 32-bit (not just
> the 16 bits of the Basic Multilingual Plane). UTF-16 is two bytes (but
> can be four for surrogates), UTF-8 is a *variable* number of bytes (1 to
> 6, I believe), etc. It's not always 2 bytes per Unicode character.

Well, then, a good solution would be to preceed each character by the
right combination of single and double quotes to signal how many bytes
it takes up, as summarized in this handy table:

  # of bytes    Preceding character sequence(s)

      1         '
      2         ''  "
      3         '"  "'
      4         ''''  "''  '"'  ''"
      5         '''''  "'''  '"''  ''"'  '''"  ""'  "'"  '""
      6         ''''''  "''''  '"'''  ''"''  '''"'  ''''"
                ""''  "'"'  "''"  '"'"  ''""  '""'  """

This could result in fairly long string representations, but that
problem could be solved by using my new algorithm to compess all of
these UTF-whatever representations as one bit (see my earlier post for
details).  Then, using . (dot) and - (dash) for a zero-bit and a
one-bit, respectively, all UTF-whatever strings could then be easily
represented as either . or - in our Python programs.  For example:

  # old program
  print ""''H"'e'"'''l'l""'o'' ''""w"""o'"r'''''l'"'d"!

  # new program
  print -


-- 
 Lloyd Zusman
 ljz at asfast.com


From meh9 at cornell.edu  Fri Apr  7 16:12:24 2000
From: meh9 at cornell.edu (Matthew Hirsch)
Date: Fri, 07 Apr 2000 16:12:24 -0400
Subject: More about variables
References: <meh9-5C17A0.10445406042000@news.cit.cornell.edu> <3X3H4.16435$0o4.108310@iad-read.news.verio.net> <meh9-158D1A.14240006042000@news.cit.cornell.edu>
Message-ID: <meh9-E07DDD.16122407042000@news.cit.cornell.edu>

Thanks for the advice.  It makes a lot of sense.

Matt







In article <meh9-158D1A.14240006042000 at news.cit.cornell.edu>, Matthew 
Hirsch <meh9 at cornell.edu> wrote:

> Why create named variables?
> 
> I always thought that having 20 lists saved more space than having one 
> list of 20 lists.  Maybe I'm wrong.
> 
> Thanks for your help,
> Matt
> 
> 
> 
> 
> 
> 
> In article <3X3H4.16435$0o4.108310 at iad-read.news.verio.net>, 
> culliton at clark.net (Tom Culliton) wrote:
> 
> > What (or maybe what do you think) you are trying to do?  Most likely
> > you would be better off with just a list or tuple if you're dealing
> > with and unknown number of data items.  Not everything needs a name,
> > and there isn't much difference between:
> > 
> > v0, v1, v2, v3, v4, v5
> > 
> > and
> > 
> > v[0], v[1], v[2], v[3], v[4], v[5]
> > 
> > Why do you think this variable length collection of data needs
> > individual names?
> > 
> > Now, if after consideration there is still some compelling reason to
> > generate names, you can use something like this with a dictionary
> > (possibly the one returned by globals()):
> > 
> > v = {}
> > for i in range(number_of_variables):
> >         name = "variable%d" % i
> >         v[name] = []
> > 
> > But again, Why?
> > 
> > In article <meh9-5C17A0.10445406042000 at news.cit.cornell.edu>,
> > Matthew Hirsch  <meh9 at cornell.edu> wrote:
> > >Hi All,
> > >
> > >Let's say I had something like:
> > >
> > >number_of_variables=int(raw_input('Enter number of variables: '))
> > >
> > >if number_of_variables==1:
> > >   variable1=[]
> > >elif number_of_variables==2:
> > >   variable1=[]
> > >   variable2=[]
> > >elif number_of_variables==3:
> > >   variable1=[]
> > >   variable2=[]
> > >   variable3=[]
> > >elif number_of_variables==4:
> > >   variable1=[]
> > >   variable2=[]
> > >   variable3=[]
> > >   variable4=[]
> > >elif number_of_variables==5:
> > >   variable1=[]
> > >   variable2=[]
> > >   variable3=[]
> > >   variable4=[]
> > >   variable5=[]
> > >else:
> > >   print 'else'
> > >
> > >Is there a more efficient way of doing this? What if I wanted to 
> > >create 
> > >100 variables (lists).  It doesn't make sense to write an if statement 
> > >for that many conditions.
> > >
> > >Thanks for your help,
> > >Matt


From effbot at telia.com  Sat Apr  8 04:52:33 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Sat, 08 Apr 2000 08:52:33 GMT
Subject: Regular Expression Help for Python Newbie.
References: <38EE3799.DA0A9638@ez-ways.com>
Message-ID: <lrCH4.80$rc9.181720576@newsb.telia.net>

Raoul-Sam Daruwala wrote:
> I have a problem. I wrote a python program to parse HTML files using the
> HTMLParser and all that I need to do with the files can be done very
> easily using this wonderful class. Kudos to the authors!
>
> My problem is that one of the sets of files that I'm trying to parse has
> badly formatted tables. Now when I do a view source on the files I can
> see the problem clearly. It's quite simple, the tables in these files
> starts out properly formatted but after a standard header the script
> than generates them leaves out the <TR> tag. This is incredible to me
> because both Netscape and IE read can view the tables properly.

<TR>'s are optional -- if the browser stumbles upon <TD>
in a <TABLE> context, it should insert <TR>'s all by itself.

> What I need to do to fix this is run a quick pre-processor and using the
> re module replace all occurences of
>
>     </TR> junk </TR> with
>     </TR><TR> junk </TR>
> where junk does not contain the tag <TR>
>
> Can anyone tell me what the re for this is? I can't seem to get anything
> to work. right now.

if you cannot get it to work, how come you're so sure you
need to use the re module? ;-)

try using the sgmllib parser instead; here's an example from the
eff-bot guide (see below).  it shouldn't be that hard to tweak the
filter class to insert TR tags if it sees a TD following directly after
a TABLE.

# sgmllib-example-4.py

import sgmllib
import cgi, string, sys

class SGMLFilter(sgmllib.SGMLParser):
    # sgml filter.  override start/end to manipulate
    # document elements

    def __init__(self, outfile=None, infile=None):
        sgmllib.SGMLParser.__init__(self)
        if not outfile:
            outfile = sys.stdout
        self.write = outfile.write
        if infile:
            self.load(infile)

    def load(self, file):
        while 1:
            s = file.read(8192)
            if not s:
                break
            self.feed(s)
        self.close()

    def handle_entityref(self, name):
        self.write("&%s;" % name)

    def handle_data(self, data):
        self.write(cgi.escape(data))

    def unknown_starttag(self, tag, attrs):
        tag, attrs = self.start(tag, attrs)
        if tag:
            if not attrs:
                self.write("<%s>" % tag)
            else:
                self.write("<%s" % tag)
                for k, v in attrs:
                    self.write(" %s=%s" % (k, repr(v)))
                self.write(">")

    def unknown_endtag(self, tag):
        tag = self.end(tag)
        if tag:
            self.write("</%s>" % tag)

    def start(self, tag, attrs):
        return tag, attrs # override

    def end(self, tag):
        return tag # override

class Filter(SGMLFilter):

    def fixtag(self, tag):
        if tag == "em":
            tag = "i"
        if tag == "string":
            tag = "b"
        return string.upper(tag)

    def start(self, tag, attrs):
        return self.fixtag(tag), attrs

    def end(self, tag):
        return self.fixtag(tag)

c = Filter()
c.load(open("samples/sample.htm"))

</F>

<!-- (the eff-bot guide to) the standard python library:
http://www.pythonware.com/people/fredrik/librarybook.htm
-->




From bkhunter at best.com  Wed Apr 26 20:15:18 2000
From: bkhunter at best.com (Bill Hunter)
Date: Wed, 26 Apr 2000 17:15:18 -0700
Subject: BAD starship.python.net news!
References: <38FC852A.9B11F458@concentric.net> <38FCBD77.A5925D65@darwin.in-berlin.de> <38FD8FC1.EC5DDCD7@tismer.com>
Message-ID: <6oQHOYGsk3CUxVh00LNjwelELGrD@4ax.com>

Many ISP's, at least if they are over some threshold of
professionalism, run backup tapes on a schedule and save archival
copies at intervals.  While they might be unsympathetic to putting in
a lot of work to recover something for an individual, they might be
persuaded that this is important to an entire global community - which
it is.

My two cents worth.

(It is far from being a sure thing that a disk recovery will work, by
the way, no matter how much you pay the specialist recovery firm.)

On Wed, 19 Apr 2000 12:51:45 +0200, Christian Tismer
<tismer at tismer.com> wrote:

>Ironically, yes it could have been used. My last copy wasn't older
>than a month. But I forgot about that when I gave up
>starship.skyport, and I asked my ISP about it:
>They wiped the machine last Friday.




From dalke at acm.org  Sat Apr  8 14:24:36 2000
From: dalke at acm.org (Andrew Dalke)
Date: Sat, 8 Apr 2000 12:24:36 -0600
Subject: Python paradigms
References: <8cmr70$b0e$1@pegasus.csx.cam.ac.uk>
Message-ID: <8cntqp$upi$1@slb0.atl.mindspring.net>

Nick Maclaren wrote
>    x = (a != NULL ? a[i]->weeble : 0) + (b != NULL ? b[i]->wombat : 0)
> [...]
>What paradigms are there for replacing such constructions by?


How about

  x = getattr(a, "weeble", 0) + getattr(b, "wombat", 0)

  ?

                    Andrew
                    dalke at acm.org





From youngd at digiship.com  Thu Apr 27 15:54:00 2000
From: youngd at digiship.com (Darren Young)
Date: Thu, 27 Apr 2000 14:54:00 -0500
Subject: DCOracle
Message-ID: <NEBBIEPHGLOHPEDKJLGJEEBLCAAA.youngd@digiship.com>

Question for all,

I have a module with the following text:

import SICKvars
import sys
sys.path.append('/usr/lib/python1.5/DCOracle')

from DCOracle import Connect

# Check for the database
#
def checkdb(dbuser, dbpass):
     try:
          connectString = dbuser + '/' + dbpass + '@' +
SICKvars.dbinfo["server"]
          dbc = Connect(connectString)
          con = dbc.cursor()
          dbc.execute("SELECT table_name FROM user_tables where table_name =
'SICKAUTH'")
          check = dbc.fetchall()
          dbc.close()
          return len(check)
     except:
          print "Caught exception: ", sys.exc_type, sys.exc_value

I have another script tha#!/usr/bin/env python
#
import os
import sys
import glob
import time
import string
import shutil
import oradbinstall


dbuser = ""
dbpass = ""
#
#  If the password is blank, we'll assume that there isn't one..(untested)
#  NOTE: Oracle might not allow that
#
while not dbuser:
     dbuser=raw_input("What is your database admin username? [root]: ")
     if not dbuser:
          dbuser = "root"
     os.system("stty -echo")
     dbpass = raw_input("What is the admin's password?: ")
     os.system("stty echo")

redo=""

try:
     isinstalled = oradbinstall.checkdb(dbuser, dbpass)
except:
     isinstalled = 0

if isinstalled:
     print "\n\nIt appears that the database is already there..."
     redo=raw_input("Would you like me to remove and re-install it? (y/n) ")
else:
     print "\n\nIt's not installed."t looks like this:

When I run it I get this:
[root at linsrv01 SICKnotes-0.6c]# ./orainstall
What is your database admin username? [root]: cm
What is the admin's password?:

It's not installed.
Segmentation fault (core dumped)
[root at linsrv01 SICKnotes-0.6c]#

No matter what I try, it cores out on me...



---------------------
Darren Young
Digiship
312-527-9440
---------------------




From osuchw at ecn.ab.ca  Mon Apr 24 09:16:11 2000
From: osuchw at ecn.ab.ca (WaldekO)
Date: Mon, 24 Apr 2000 07:16:11 -0600
Subject: Using DAO and Python
References: <bMHM4.23114$fF5.684678@news1.rdc1.il.home.com>
Message-ID: <8e1hit$171$1@iceman.tac.net>

dsavitsk <dsavitsk at e-coli.net> wrote in message
news:bMHM4.23114$fF5.684678 at news1.rdc1.il.home.com...

> here is what i have tried (esentially following the example on pg 252 of
> python programming on win32)
> 1. ran makepy and selected microsoft dao 3.6 from the list
> 2. in PythonWin typed
> >>> import win32com.client
> >>> daoengine = win32com.client.Dispatch('DAO.DBEngine')

You did not provide full name of DBEngine.  Try:

daoengine = win32com.client.Dispatch('DAO.DBEngine.36')

For a sample of using MSAccess refer to:
http://starship.python.net/crew/bwilk/access.html






From gmcm at hypernet.com  Tue Apr 11 10:29:11 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Tue, 11 Apr 2000 10:29:11 -0400
Subject: Getting an overview if all available driveletters on NT
In-Reply-To: <2B1262E83448D211AE4B00A0C9D61B03BA7294@MSGEURO1>
Message-ID: <1256659135-41087041@hypernet.com>

Pieter Claerhout wrote:

> is there a module in Python, which can list all available driveletters, together with
> their label. I'm using Windows NT? I already thought of iterating over all possible
> driveletters, and see if the path exists or not, but I also want to separate the
> network volumes from the other volumes. Any ideas?

In Mark's Win32 extensions the module win32api has 
GetLogicalDriveStrings() and GetVolumeInformation(...).


- Gordon



From doughellmann at home.com  Sun Apr 30 08:23:07 2000
From: doughellmann at home.com (Doug Hellmann)
Date: Sun, 30 Apr 2000 12:23:07 GMT
Subject: python + linux + cron = problem
References: <39081703.48B64D1A@spekter.no> <3908265E.A3C21409@home.com> <3908350F.922D7B4E@spekter.no>
Message-ID: <390C26A3.EB8263DD@home.com>

What OS are you on?  I'm not getting any output from that df command
(even if I remove the "grep eggum") under Linux.

After registering the python script directly as a cron job, I had no
problem getting it to run.

The commands module actually runs a different command than what you send
in:

    pipe = os.popen('{ ' + cmd + '; } 2>&1', 'r')

Have you tried running that shell command yourself from the command line
to see if you get a better error message?

Doug

J?rn Eilertsen wrote:
> 
> here's the whole script, well the part that gives me grief at least
> :
> #!/usr/bin/python
> import commands
> 
> koblinger = commands.getoutput('df | grep eggum | cut -d 1 -f 1 | cut -d / -f
> 4')
> 
> I've tried to put the shell command into a shell script and run
> commands.getoutput on the script but with the same result. Here's the cron
> entry:
> 0 1 * * *   /usr/local/bin/sjekkmounts
> and I've tried sh /usr/local/bin/sjekkmounts
> 
> Jorn
> 
> Doug Hellmann wrote:
> 
> > J?rn Eilertsen wrote:
> > >
> > > Hi,
> > >
> > > I've posted this earlier but noone replied so I'm trying again
> > >
> > > I have this small program which includes the following line og code:
> > >  mounts = commands.getoutput('df | grep eggum | cut -d 1 -f 1 | cut -d /
> > > -f 4')
> > >
> > > this runs fine from the commandline but when run as a cronjob I get the
> > > following error:
> > >
> > > File "/usr/lib/python1.5/commands.py", line 42, in getoutput
> > >     return getstatusoutput(cmd)[1]
> > >   File "/usr/lib/python1.5/commands.py", line 53, in getstatusoutput
> > >     sts = pipe.close()
> > > IOError: (10, 'No child processes')
> > >
> > > The use of pipes seems to be the trouble here.
> > >
> > > Any help on how the get around this would be greatly appreciated.
> > >
> > >
> >
> > If the problem is really having pipes in the command, why not put your
> > shell commands in a shell script, then run commands.getoutput() on that?
> >
> > If the program runs from the command line but not cron, the problem is
> > more likely to be related to PATH problems within the cron environment.
> > How do you set up the cron job?
> >
> > Doug


From cut_me_out at hotmail.com  Thu Apr 13 10:53:13 2000
From: cut_me_out at hotmail.com (Alex)
Date: 13 Apr 2000 10:53:13 -0400
Subject: List mapping?
References: <CC59868A0A76D311B1F50090277B321C1F326A@VRTMAIL>
Message-ID: <etditxmt5ly.fsf@w20-575-109.mit.edu>

from operator import getitem

def pair (l):
    assert len (l) % 2 == 0
    evens = range (0, len (l), 2)
    evens = map (getitem, len (evens) * [l], evens)
    odds = range (1, len (l), 2)
    odds = map (getitem, len (odds) * [l], odds)
    return map (None, evens, odds)

if __name__ == '__main__':
    print pair (range (10))

Alex.


From icurhere2 at nowhere.here  Tue Apr 11 15:34:33 2000
From: icurhere2 at nowhere.here (icurhere2)
Date: Tue, 11 Apr 2000 19:34:33 GMT
Subject: Python and Zope help needed for test project
Message-ID: <d7LI4.601$0n5.9723@c01read04.service.talkway.com>

Hello,
 
My name is Mike Becraft, and I am one of the Test Development Managers
at http://www.brainbench.com , an online, independent certification
company.  I am looking into the possibility of having a test on Python
and Zope knowledge developed for use on our web site, and I was
wondering if there was anyone interested in assisting.  If this is a
possibility, I would be more than happy to speak with anyone with
relevant programming experience.

I also encourage you to visit our site (free for individuals) to check
out some of the tests and certifications we currently offer.

Thanks greatly,
Michael Becraft

Michael Bryan Becraft
Assessment Development Manager - Brainbench
http://www.brainbench.com
mailto:michael.becraft@(please-remove-this-area)brainbench.com
--
Free audio & video emails, greeting cards and forums
Talkway - http://www.talkway.com - Talk more ways (sm)



From effbot at telia.com  Sat Apr  8 04:43:54 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Sat, 08 Apr 2000 08:43:54 GMT
Subject: Tkinter install?
References: <081c4ffb.329993cd@usw-ex0102-014.remarq.com>
Message-ID: <ejCH4.78$rc9.171149312@newsb.telia.net>

John Schmitt wrote:
> -------------------
> TclError: Can't find a usable init.tcl in the following
> directories:
>     {} ./lib/tcl8.0 D:/pyprojects/grayson/tcl8.0/library
> D:/pyprojects/grayson/Examples/library
>
> This probably means that Tcl wasn't installed properly.
> -------------------

this might help:
http://www.pythonware.com/people/fredrik/fyi/fyi02.htm

</F>




From banderson at boi.hp.com  Fri Apr 14 16:48:06 2000
From: banderson at boi.hp.com (Bill Anderson)
Date: Fri, 14 Apr 2000 14:48:06 -0600
Subject: ANN: Python training in Boulder
References: <000101bfa571$e31a43a0$71d95da6@lutz>
Message-ID: <38F78406.2032D7AB@boi.hp.com>

Mark Lutz wrote:
> 
> This is to announce that the second Python training session
> to be held at Tom Christiansen's site in Boulder, Colorado,
> will be held May 24-26.
> 
> These classes are equivalent to the on-site Python training I
> offer, and are open to individuals on a first-come first-served
> basis.  I am the instructor, but all registration details are
> handled by Tom's company.  For more information, please visit
> these pages:
> 
> http://home.rmi.net/~lutz/boulder-classes.html   (my page)
> http://training.perl.com/open-courses.html       (tom's page)
                  ^^^^^^^^^

Maybe it's me, but something seems fundamentally wrong here. ;^)



-- 
Bill Anderson 			Linux/Unix Administrator, Security Analyst
ESBU (ARC)			bill_anderson at boi.hp.com
My opinions are just that; _my_ opinions.


From embed at geocities.com  Mon Apr  3 16:29:34 2000
From: embed at geocities.com (Warren Postma)
Date: Mon, 3 Apr 2000 15:29:34 -0500
Subject: Creating variables on the fly...
References: <meh9-25D0C5.15160303042000@news.cit.cornell.edu>
Message-ID: <gl6G4.2926$HG1.87612@nnrp1.uunet.ca>

You're using the Wrong Tool for the Job.

You want a Dictionary, or an Array. If you really like naming these objects
a,b, and c, use a dictionary:


V = {} # create a dictionary just for this bunch of variables

def AddVars( a, count, initial_value):
    global V
    c=ord(a[0])
    for n in range(c,c+count):
        V[chr(n)] = initial_value

def Sum(dict):
    n = 0
    for i in dict.keys():
        n = n + dict[i]
    return n

def Dump(dict):
    k = dict.keys()
    k.sort()
    for i in k:
        print i, "= ",dict[i]

AddVars("a",25,0)
V["a"] = 5
V["b"] = 10
print Sum(V)
Dump(V)







From someone at somewhere.com  Tue Apr 18 03:23:04 2000
From: someone at somewhere.com (Charley Horse)
Date: Tue, 18 Apr 2000 07:23:04 GMT
Subject: Using python on the web
Message-ID: <MPG.1365adcfa8470eef9896aa@news>

I've just started fooling with python. I've looked around the python 
site and a number of the other sites referenced there and am a bit 
puzzled. Umm... how does one go about using python for web applications, 
especially database driven apps? Almost everything deals with python 
just as a language, as C++ would be. Not much about the web 
specifically. I was a bit shocked to find the only direct refs on about 
this at python.org are entwined with CGI (unless I missed it). I have no 
exposure to CGI, but rather to php, CF, and ASP, so python has gone from 
looking very attractive to looking... well, like a puzzle. I know about 
Zope, but am interested in how one uses python on the web just by 
itself.

Thanks


From tom_hines at my-deja.com  Fri Apr 28 16:59:10 2000
From: tom_hines at my-deja.com (Tom Hines)
Date: Fri, 28 Apr 2000 20:59:10 GMT
Subject: win32com: _NewEnum being cached
References: <8ecta7$nn5$1@nnrp1.deja.com>
Message-ID: <8ecu2l$ol1$1@nnrp1.deja.com>

BTW, I'm using Python 1.5.2 and win32com build 125.

In article <8ecta7$nn5$1 at nnrp1.deja.com>,
  Tom Hines <tom_hines at my-deja.com> wrote:
> Hello.  As you know, to implement "for x in my_list" functionality in
> COM, you must code a get__NewEnum() function, which I have done and it
> works fine under Visual Basic.  But when called from Python, the call
> seems to get cached, as subsequent Python "for" blocks don't result in
> calls to get__NewEnum().
>
> I have a COM class that implements a dynamic collection, so I
> need "for" blocks to result in calls to get__NewEnum() every time.
> Does anybody know how to force this?
>
> Thanks.
>
> --
> Tom Hines
> Rockville, MD
>
> Sent via Deja.com http://www.deja.com/
> Before you buy.
>


Sent via Deja.com http://www.deja.com/
Before you buy.


From sabren at manifestation.com  Sun Apr 23 06:32:53 2000
From: sabren at manifestation.com (Michal Wallace (sabren))
Date: Sun, 23 Apr 2000 06:32:53 -0400 (EDT)
Subject: python-only crypt()?
Message-ID: <Pine.BSF.4.10.10004230627580.5642-100000@chi.pair.com>

Hey all,

  Is there a python-only implementation of the crypt() module?  I'm
looking for users who don't have access to rebuild python with crypt
enabled...

  I could only find C modules... I'm taking a crack at porting
Crypt::UnixCrypt from perl, but I can't really tell what the code is
doing once it gets down into the DES algorithm, so it's a little
tedious.

Cheers,

- Michal
-------------------------------------------------------------------------
http://www.manifestation.com/         http://www.linkwatcher.com/metalog/
-------------------------------------------------------------------------




From gresham at mediavisual.com  Wed Apr 12 04:03:41 2000
From: gresham at mediavisual.com (Paul Gresham)
Date: Wed, 12 Apr 2000 16:03:41 +0800
Subject: Classes, Inheritance - Stupid lazy question
Message-ID: <8d1cfh$hme7@imsp212.netvigator.com>

Hi,
This is stupid and lazy of me, I have looked through the docs, but I think
my brain is fried today, I just couldn't see it ... when I have a sub-class
with an overridden function/method how do I call the same function/method in
the parent ?

i.e.

class lazy():
    __init__(self, name):
        self.name = name

    fired(self):
        print 'You are fired %s' % self.name

class paul(lazy):
    __init__(self, name):
        self.name = name  # <---- ?????
        self.april_pay = 0

    fired(self):
        print 'Here's a box, collect the things from your desk'
        lazy.fired(self) # <--- ??????

The thing I'm just not clear about is the __init__ in my subclass, I want to
do an additional initialisation, without having to the whole init from the
parent class. I think what I am doing is kindof close ... but just want to
know the real way to do this !!

Sorry I've just really lost the plot today!

Thanks guys
Paul




From robin at jessikat.demon.co.uk  Thu Apr 27 14:33:08 2000
From: robin at jessikat.demon.co.uk (Robin Becker)
Date: Thu, 27 Apr 2000 19:33:08 +0100
Subject: module case bug in 1.6a2 Win9x
References: <UGQuwCAnnFC5Ew+e@jessikat.demon.co.uk>
 <390873A2.BD0EF60F@endea.demon.nl>
Message-ID: <K8Wj4VAkfIC5EwuO@jessikat.demon.co.uk>

In article <390873A2.BD0EF60F at endea.demon.nl>, Niels Diepeveen
<niels at endea.demon.nl> writes
>
>
>Robin Becker schreef:
>> 
>> I believe there's a module case bug with long named modules under win98
>> at least. I believe the bug relates to the length of the filename
>> somehow.
>> 
>
>Yes, that looks like a bug in import.c:allcaps8x3(). It seems to return
>true instead of false for long filenames. As far as I can tell, this
>probably happens on all Win32 platforms at least since 1.5.2.
>
It seems a bit odd that allcaps83 returns 1 if the first chunk is longer
than 8 or the extension is longer than 3 or there are more than one '.'
in the name doesn't it. I guess all of these return 1's should be return
0's except the last.
-- 
Robin Becker


From moshez at math.huji.ac.il  Tue Apr 25 13:47:55 2000
From: moshez at math.huji.ac.il (Moshe Zadka)
Date: Tue, 25 Apr 2000 20:47:55 +0300 (IDT)
Subject: Question about exception-handling mechanism
In-Reply-To: <8e46mt$4nh$1@nnrp1.deja.com>
Message-ID: <Pine.GSO.4.10.10004252046040.28957-100000@sundial>

On Tue, 25 Apr 2000 willfg at my-deja.com wrote:

> Hello all,
> 
>    I'm new to Python, but find its exception handling mechanism much
> more powerful than comparable languages; but a question was put to me
> that being new I can't come up with a very articulate answer. A
> colleague asked why in an exception handling mechanism you'd want the
> ELSE block to be executed if you don't throw an exception as opposed to
> a FINALLY block. Anyone used this feature in practice? Thanks in
> advance for your input, -- Will

Lots of time.

Consider some sort of

def f(file):
	try:
		fp = open(file)
	except IOError:
		return 0 # it doesn't exist
	else:
		return fp.read(2)=='PK'

To check if a certain file is both readable and is a zip-file. You don't
want the except to cover the "read" -- an error in read is unexpected.
--
Moshe Zadka <mzadka at geocities.com>. 
http://www.oreilly.com/news/prescod_0300.html
http://www.linux.org.il -- we put the penguin in .com




From dalcolmo at mediaone.net  Sat Apr 22 13:31:49 2000
From: dalcolmo at mediaone.net (Josef Dalcolmo)
Date: Sat, 22 Apr 2000 17:31:49 GMT
Subject: Python for drivers?
References: <3900EDBE.282BC85@dsuper.net>
Message-ID: <9mlM4.11346$QX1.481771@typhoon.ne.mediaone.net>

Dave <redsand at dsuper.net> wrote in message
news:3900EDBE.282BC85 at dsuper.net...
> Can Python be used for writing device drivers? I just got BeOS 5 and it
> doesn't support some video cards. I'd like to write one for the Cirrus
> Logic card and I want a language that's clear and concise- Python. I've
> never written device drivers. Any suggestions?
>

Just my five cents here:
for writing a device driver you want a language that can compile to native
code, in order to gain best machine access and best speed.
C or Assembly are suitable, but so are some versions of Pascal, Oberon and
Forth (amongst others). Note however that writing device drivers is
something VERY close to the hardware, and unless a language has built-in
support for it, forget it and use C, which can be used for about anything,
even though it is not an elegant language for high level programming.

- Josef




From phd at phd.russ.ru  Mon Apr 10 04:51:58 2000
From: phd at phd.russ.ru (Oleg Broytmann)
Date: Mon, 10 Apr 2000 08:51:58 +0000 (GMT)
Subject: win32all: set process title
In-Reply-To: <rwRH4.1668$4v.9405@news-server.bigpond.net.au>
Message-ID: <Pine.LNX.4.21.0004100844040.19783-100000@fep132.fep.ru>

Hello!

   I am working an a module to set process name (title) in the output of
"ps" command. It is UNIX-only module, of course.
   I want to provide some portability, if at all possible, but do not know
anything about modern Windows and Python for Win (I departed from
DOS/Windows 3.1 world many yeras ago :)

   Can you help me? Could you please send me a snippet of code and a list
of platforms - how I can set process name (some comments on how user can
look at process name will be good accessories :)
   Something like this (I rememeber seeing such code in c.l.py some time
ago):

import sys
if sys.platform == "posix": # Oops, wrong test, DJGPP-compiled DOS port is "posix", too :/
   from setproctitle import setproctitle
elif sys.platform in ("nt", "md95"):
   from win32all import CurProc, SetProcessName
   def setproctitle(title):
      SetProcessName(CurProc(), title)
else:
   raise NotImplementedError("Your platform is not supported")

Oleg.
---- 
    Oleg Broytmann    http://members.xoom.com/phd2.1/    phd2 at earthling.net
           Programmers don't die, they just GOSUB without RETURN.




From jvickroy at sec.noaa.gov  Thu Apr 20 15:28:22 2000
From: jvickroy at sec.noaa.gov (j vickroy)
Date: Thu, 20 Apr 2000 13:28:22 -0600
Subject: can't find SWIG for Python
Message-ID: <38FF5A56.F76794D8@sec.noaa.gov>

Hello all,

I have been unable to locate SWIG for Python.  My references list
www.swig.org but the server can not be found.

Could someone point me to the site or provide some other Win 32 options
for easing the process of extending Python to use existing C libraries.

Thanks for your time.



From morse at harborcom.net  Fri Apr  7 09:44:43 2000
From: morse at harborcom.net (Kevin Dahlhausen)
Date: Fri, 07 Apr 2000 13:44:43 GMT
Subject: PPython
Message-ID: <38ede548.353776233@news.oh.verio.com>

So many web hosting sites only allow Perl scripting, I got to thinking.  That's
often a dangerous thing.   Would it be possible to implement the Python virtual
machine in Perl?   OK, anything is possible,  but I guess the real question is
does anyone have a feel for the scope of such a project?  Say using a native
interpreter to pre-compile the Python code to pyc files and the perl cgi vm to
run them?      Feasible, or just a silly idea?




From ullrich at math.okstate.edu  Wed Apr 26 14:57:23 2000
From: ullrich at math.okstate.edu (David C. Ullrich)
Date: Wed, 26 Apr 2000 18:57:23 GMT
Subject: Simple Function Question
References: <8e7ddt$18n@autodesk.autodesk.com>
Message-ID: <39073aef.12145997@news>

On Wed, 26 Apr 2000 11:44:10 -0700, "Akira Kiyomiya"
<akira.kiyomiya at autodesk.com> wrote:

>Hi,  Could you explain this in dummy's language?
>
>This function prints out '4' and I have a feeling that x=1 and n=3, so
>add1(3) prints '4'.
>
>However, I have a difficlut time trying to see how these two arguments
>assigned to both x (especially x) and n.

	Not sure exactly what the question is, but here's what
happens:

># Return a function that returns its argument incremented by 'n'
>def make_incrementer(n):
>    def increment(x, n=n):
>        return x+n
>    return increment
>
>add1 = make_incrementer(1)

	Think about what just happened when you said
make_incrementer(1). This actually executed the following
function definition:

def increment(x, n=1):
        return x+n

The "n=1" means that n is an optional
parameter - if you don't include n then
it is set to 1; so calling increment(3)
is the same as saying increment(3, 1),
which should return 4.

	If you'd said add2 = make_incrementer(2)
that would execute the definition

def increment(x, n=2):
        return x+n

and set add2 to the function so defined - then
add2(x) is the same as add2(x,2).

>print add1(3)  # This prints '4'
>
>
>



From python-list at teleo.net  Sun Apr  9 18:31:37 2000
From: python-list at teleo.net (Patrick Phalen)
Date: Sun, 9 Apr 2000 15:31:37 -0700
Subject: new Scripting Language Survey at scripting.com
Message-ID: <00040915402004.00554@quadra.teleo.net>

Yet Another What is Your Favorite Scripting Language survey just got
going at http://www.scripting.com.

Not surprisingly (considering that is Frontier country), Frontier is in
the lead. Surprisingly, Python is placing a strong second, well ahead
of AppleScript, Javascript, Perl, PHP, Tcl, VBScript and Other.

Actually, Other is doing rather well ...

vote-early-and-often-ly y'rs  - patrick



From effbot at telia.com  Tue Apr 25 14:12:39 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Tue, 25 Apr 2000 18:12:39 GMT
Subject: mxODBC error trapping
References: <A63AFB20111ED311AC5500A0C96B7BF5015F0F4B@orsmsx36.jf.intel.com>
Message-ID: <relN4.4106$Za1.60270@newsc.telia.net>

Daley, MarkX <markx.daley at intel.com> wrote:
> I am accessing databases using the notes by John Dell'Aquila for ODBC
using
> Python Database API.  It works fine, but now I am trying to set up error
> trapping in the event that a connection fails to be made.  Here is the
basic
> code structure:
>
> def collect():
>     "Routine to query all servers for all names for time period defined
>     in timer."
>    # Define some variables
>    for index in servername:
>        try:
>            # Big routine to collect data and write to file
>        except dbi.operation-error:
>            servername.append(index)

"operation-error" is not a valid name (it's interpreted
as "operation" minus "error").

> I guess I don't know enough about exceptions to trap them properly.  Any
> hints?

according to the DB specification [1], OperationalError is the
correct name.  you still have to qualify it with the module name,
though:

    try:
        ...
    except dbi.OperationalError:
        ...

if that doesn't work, try printing the contents of the "dbi" module:

print dir(dbi)

</F>

1) http://www.python.org/topics/database/DatabaseAPI-2.0.html




From Richard.Jones at fulcrum.com.au  Mon Apr 10 17:11:27 2000
From: Richard.Jones at fulcrum.com.au (Richard Jones)
Date: Tue, 11 Apr 2000 07:11:27 +1000
Subject: Python/Perl Popularity (Re: A Mountain of Perl...) 
In-Reply-To: Message from tom__98@my-deja.com of 2000-Apr-10 20:8:31,
            <8ctcar$ml9$1@nnrp1.deja.com> 
Message-ID: <200004102111.HAA03380@envy.fulcrum.com.au>

[tom__98 at my-deja.com]
> In article <20000405053920.20579.qmail at web2102.mail.yahoo.com>,
>   lewst <lewst at yahoo.com> wrote:
> > What is it about Perl that makes it so much more popular and have such
> > a huge grassroots swell?  I personally find Perl an abomination and
> > Python a breath of fresh air.  Perl has that first mover advantage I
> > suppose, but should that really make such a hugh difference?
> 
> I don't know about others, but I frequently end up using
> Perl instead of Python, even though I greatly prefer Python
> as a language.

   Tom you little troll-meister. Half those arguments you cite are bollocks if 
you _really_ knew and "greatly prefer" Python. Try again.


        Richard





From dirck at pacbell.net  Tue Apr 18 15:45:48 2000
From: dirck at pacbell.net (Dirck Blaskey)
Date: Tue, 18 Apr 2000 12:45:48 -0700
Subject: module __call__
Message-ID: <LX2L4.1275$B43.102174@news.pacbell.net>

Just a little GWIBNI....

...Gee-Wouldn't-It-Be-Nice-If...

modules could def __call__():

so you could call a module, i.e.

    import StringIO
    f = StringIO()

instead of
    import StringIO
    f = StringIO.StringIO()

or
    from StringIO import StringIO
    f = StringIO()

Yes, it's a very little gwibni...

(Not-hip-enough-to-sign-with-an-adverb)

d





From effbot at telia.com  Wed Apr 12 12:46:58 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Wed, 12 Apr 2000 16:46:58 GMT
Subject: comparing strings and ints
References: <jr%I4.643$rc9.190628352@newsb.telia.net> <1256565470-46720619@hypernet.com>
Message-ID: <6M1J4.662$rc9.209708544@newsb.telia.net>

Gordon McMillan <gmcm at hypernet.com> wrote:
> Certainly nothing in Doc/lib/built-in-funcs.html mentions any
> danger in comparing arbitrary objects, nor the discussion in
> 2.1.5.2 ( Lib | Built-in Types | Mutable Sequence Types | sort).

otoh, 1.6a2 cannot compare strings safely:

Python 1.6a2 (#0, Apr  5 2000, 23:56:55)
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> u"abc" < "???"
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
ValueError: UTF-8 decoding error: unexpected code byte

</F>




From vlaw at my-deja.com  Sat Apr  8 18:06:38 2000
From: vlaw at my-deja.com (vlaw at my-deja.com)
Date: Sat, 08 Apr 2000 22:06:38 GMT
Subject: Merchant Empires - python CGI based game up on SourceForge
Message-ID: <8coah6$buv$1@nnrp1.deja.com>


I've got a SourceForge (www.sourceforge.net) site set up for Merchant
Empires, my python CGI web-based strategy game here:

http://merchempires.sourceforge.net

Anyone wanting to learn CGI based python might take a look at this.
Anyone who knows a lot about it might also and tell me if I am doing
anything wrong.  I'm using cookies for authentication.  I also use
PostgreSQL, PyGreSQL, and HTMLgen.

I couldn't find any complete Python (CGI) based web application of the
web to use for examples.  I can only assume I'm doing things right with
the use of Cookie.py.

I'm looking for people to assist with this project.

Oh yeah I just learned Python last week so some of the code is probably
pretty rough.


Sent via Deja.com http://www.deja.com/
Before you buy.


From just at letterror.com  Sun Apr 30 12:43:10 2000
From: just at letterror.com (Just van Rossum)
Date: Sun, 30 Apr 2000 17:43:10 +0100
Subject: [Python-Dev] Re: Python 1.6a2 Unicode bug (was Re: comparing
 strings and ints)
In-Reply-To: <45919991@data.franken.de>
References: <l03102805b52ca7830b18@[193.78.237.154]>
 <39072F0C.5214E339@lemburg.com>
 <BlHN4.1719$wYl.195075584@newsb.telia.net>
 <oqvh14fikb.fsf@titan.progiciels-bpi.ca> <45919986@data.franken.de>
 <048501bfb0b4$c793d760$4add3fcb@neil>
Message-ID: <l03102801b53212ef9f1a@[193.78.237.136]>

At 9:56 AM +0200 28-04-2000, Walter Doerwald wrote:
>But I find UCS-2 much cleaner, because it's easily distinguished from
>ASCII/UTF-8/ISO8859-1/..., distinguishing between UTF-8 and ISO8859-1 has
>to be done by "external" methods. How does Python do it?

It doesn't; it'll just assume utf-8...

Just





From embed at geocities.com  Mon Apr  3 15:50:31 2000
From: embed at geocities.com (Warren Postma)
Date: Mon, 3 Apr 2000 14:50:31 -0500
Subject: Tuples -- who needs 'em
References: <00e101bf9d80$a8384d20$74eb0b18@stcla1.sfba.home.com>
Message-ID: <HM5G4.2915$HG1.87302@nnrp1.uunet.ca>

> How does having both types significantly improve life for Python
> programmers?

I don't want to reopen the can of worms, but here's how I think of it,
conceptually:

First, every time you invoke a function:

    f(x,y,z)

- Python looks in it's namespace(s) for the function f, this is a python
object
- Python creates a Tuple of the arguments you passed, ie, args=(x,y,z)
- Python invokes the function internally like this: Apply(f,args)

So, Tuples are an unavoidable concept inside the construction of Python. The
only
remaining question is, what value is there in exposing them to the user.



Well, the first value to me, is that Python by nature allows introspection
to an unprecedented degree. That is the most powerful thing, to me, about
Python programming. I can separate a function name from the parameters,
store them some place, and combine them, chemistry-set style, using map and
apply.



Nice Uses of Tuples-"The Point":

A point consists of X,Y point pairs. This is a natural use for a Tuple. If
the file format was 3D, you'd have an X,Y,Z tuple.  As a whole, the tuple
represents a single point in space. It is no longer two unrelated integers.
It's a literal-value with a presence in two-space or three-space, if you
like to think of things that way.  Of course, you can't change part of a
Tuples, any more than you can change Part of the number 3.  It's either 3 or
it's not, and it's a point at (5,10) or it's not.  If you still don't get
the Immutable thing, you need to meditate some more on the Way of the
Pythonista.  Read the Manuals, while chanting the Newbie Mantra: "Hmmm".


Lists + Tuples: "Sequence of Points"

Imagine putting the tuples (point pairs) together into a List of points. A
list is the appropriate type, because you can insert and delete items from a
list, something which would be a Bad Idea in Tuples, because they are
Immutable.

The list of tuples gives you the best of both worlds. Variable length in the
list, so you can have any number of Points in the list, but you can't go
adding a third item to a Tuple once created, nor delete the second item.
You can pass this tuple around the system by reference and it will never get
Messed With either.  If someone changes their own local copy, they get a new
tuple just for them.

So without the distinction, Python's "object reference" scheme would be
horribly broken.


Dictionaries + Tuples: "Point to Name Map"

Suppose now you are mapping Points on a Map to City Names. A map could be
large, and it might be a bad idea to create a 2 dimensional array.   You
can't use a List as the Key to a dictionary, and here's an example of where
a Tuple is just the ticket:

CityName = {} # empty dictionary
CityName[(5,24)] = 'London'
CityName[(7,23)] = 'Paris'
CityName[(9,24)] = 'Berlin'

A dictionary using a Tuple as the key is a nice way to do some kinds of
Sparse arrays.


Tuples in Tuples: "Gettin' Jiggy with Recursion"

Tuples nested inside Tuples can be useful as well.  You can use them to do
certain recursive types of things nicely:

from types import *
b = ( 9, "*", (7, "-", 2)) # infix sequence of operations as a tuple
def f(x):
  if (type(x)!=TupleType):
    return x
  a,b,c = x
  v = str(f(a))+b+str(f(c))
  print v
  return eval( v )

print f(b)
#output:
#
# 7-2
# 9*5
# 45


Warren








From gregm at localhost.localdomain  Thu Apr  6 00:16:14 2000
From: gregm at localhost.localdomain (Greg McFarlane)
Date: 6 Apr 2000 14:16:14 +1000
Subject: Tkinter, <select> style menu
References: <8cdqom$aqp$1@news101.telia.com> <38EA636C.EB62A9BF@exceptionalminds.com> <00d501bf9e89$56f22080$2c6ba8c0@frodo>
Message-ID: <8ch32e$cld$1@localhost.localdomain>

In article <00d501bf9e89$56f22080$2c6ba8c0 at frodo>,
Anders Eggers-Krag <aek at mail1.stofanet.dk> wrote:
>> Anders Eggers-Krag wrote:
>> >=20
>> > Is there really no reasoanble way to make a box similar to that of =
>the HTML
>> > <SELECT>
>> > in tkinter?
>> >=20
>> > the nearest I have been able to find is from Pmw, and it doesn't =
>really look
>> > like windows...
>>=20
>> Hmm, I thought that the Pmw.ComboBox() when used in drop down mode
>> looked almost just like the Windows Drop Down box.
>>=20
>> What is your problem with it?
>well take a look at this picture and tell me if you think they look =
>alike
>http://www.aek.dk/python.gif
>
>left is windows in case there is any doubt :-)
>
>my personal oppinion is the difference is too great for my user...

Pmw is highly configurable.  I am sure that it could be configured
to look just like Mr Gates' dropdown box.  If you tell me what you
want to change, I will be able to tell you the Pmw options to set.

Greg McFarlane     Open Telecommunications     gregm at ot.com.au
-- 
Greg McFarlane     Open Telecommunications     gregm at ot.com.au


From paul at qc-ns.com  Mon Apr  3 12:21:00 2000
From: paul at qc-ns.com (Paul MAJGIER)
Date: Mon, 03 Apr 2000 18:21:00 +0200
Subject: SubmitWolf
Message-ID: <38E8C4EC.ECF4794A@qc-ns.com>

Hello everyone,

Do you happen to know of a program or part of a program which can submit
a site to a thousand search engines at the same time ? To
give you a more precise idea of what I am looking for, I can give you
the example of "submitwolf" which does something similar under windows.

I could obviously come up with such a program myself but there are many
types of forms to enter and if such a program exists already, it must
certainly have a rich form database and that would be very helpful to
me.

Thanks in advance for your help.





From nospam at nowhere.net  Wed Apr 26 00:56:36 2000
From: nospam at nowhere.net (Ray&Maria)
Date: Tue, 25 Apr 2000 22:56:36 -0600
Subject: Sorting a list of classes
References: <OmnqwHar$GA.233@cpmsnbbsa03> <m3snwafxek.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <eGuN4.1277$sf5.15587@news.corecomm.net>

No, you can sort classes, too.  It is very useful if you want to sort a list
of plug-ins in a pull-down menu.

class A:
    className = "A"
class B:
    className = "B"
class C:
    className = "C"

classList = [B, C, A]
print "before sorting:", classList
classList.sort(lambda a, b: cmp(a.className, b.className))
print "after sorting:", classList

Ray
NOSPAMjchen at NOSPAMquark.com


"Michael Hudson" <mwh21 at cam.ac.uk> wrote in message
news:m3snwafxek.fsf at atrus.jesus.cam.ac.uk...
> "lexberezhny" <lexberezhny at email.msn.com> writes:
>
> > If i have a python list of classes, can i do something like
> > classList.sort() and how can i control this behavior within the
> > classes.
>
> Do you mean "class"?  In which case, no.  If you mean "instance", then
> see the other replies...
>
> Cheers,
> M.
>
> --
> 41. Some programming languages manage to absorb change, but
>     withstand progress.
>      -- Alan Perlis, http://www.cs.yale.edu/~perlis-alan/quotes.html




From m.faassen at vet.uu.nl  Fri Apr 21 15:20:19 2000
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 21 Apr 2000 19:20:19 GMT
Subject: Using python on the web
References: <EDFD2A95EE7DD31187350090279C6767754BA8@THRESHER>
Message-ID: <8dq9lj$pjg$2@newshost.accu.uu.nl>

Brent Fulgham <brent.fulgham at xpsystems.com> wrote:
> A couple of questions:

>> We build medium-busy, CGI-intensive sites in Python. We were briefly
>> quite excited about Zope, but once we realized it locked us into Yet
>> Another Programming Language, we lost interest. DHTML is particularly
>> problematic,
> 1.  I'm not familiar with Zope.  Is DHTML specific to Zope?  It seems
> to be proudly advertised on a lot of "webware" brochures I see these
> days.

Acronym mixup. DHTML (Dynamic HTML) and DTML (Document Template Markup Language)
have little in common besides that they both are used with HTML. DHTML is
client side (in the browser), DTML server side (in Zope). DHTML is to do
fancy graphics effects, DTML is a reporting language that talks to Zope
(read Python) objects and usually generates HTML.

Regards,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From news at bachi.ch  Tue Apr 25 05:54:44 2000
From: news at bachi.ch (Martin Bachmann)
Date: 25 Apr 2000 09:54:44 GMT
Subject: Python/Zope or better PHP4
Message-ID: <8F2172931newsbachich77@192.168.1.1>

We're currently developping our web applications in PHP for use with 
Apache/mod_php. This works great and PHP4 seems to bring lots of 
improvements. 

However, we're looking for an integrated, object-oriented development frame 
with some security concepts, so Zope came to our view.

My question now, does anyone here have experience with both PHP and Zope? 
Would it be worth take the effort to change to Zope? What about 
advantages/disadvantages? How well does Python work as a web programming 
language compared to PHP (which was designed right from the start for this 
purpose) ?

Thanks for any comments and views,
Martin


From greg at cosc.canterbury.ac.nz  Tue Apr  4 23:20:35 2000
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Wed, 05 Apr 2000 15:20:35 +1200
Subject: Pass by reference ?
References: <38E864C9.E1EE94F7@ecs.soton.ac.uk> <38E8D449.FBC1E2AD@roguewave.com> <8caohp$j4v$1@pegasus.csx.cam.ac.uk>
Message-ID: <38EAB103.E76E74C2@cosc.canterbury.ac.nz>

Nick Maclaren wrote:
> 
> Unless I am thoroughly confused, Python uses bog-standard call by
> reference, with the wrinkle that assignment is a rebinding rather
> than a copying operation.  AM I confused about that?

The trouble with describing Python parameter passing as
"call by reference" is that that term implies something
like a Pascal VAR parameter, which Python most certainly
does *not* have. That's why Mr. Generowicz got confused -
he thought it meant that a reference to the *variable*
x was being passed.

If I had to pick one of the standard terms, I would describe 
Python parameter passing as call-by-value, with the wrinkle 
that the value being passed is always a reference to an object.

But I think it's better to avoid using any of the terms
developed for describing parameter passing in Algol-like
languages when explaining Python to a newcomer, because
none of them really fit properly, and further explanation
is needed to avoid confusion. It's better to just give the
full explanation in the first place.

-- 
Greg Ewing, Computer Science Dept,
+--------------------------------------+
University of Canterbury,	   | A citizen of NewZealandCorp, a	  |
Christchurch, New Zealand	   | wholly-owned subsidiary of USA Inc.  |
greg at cosc.canterbury.ac.nz	   +--------------------------------------+


From samschul at pacbell.net  Mon Apr 24 23:07:28 2000
From: samschul at pacbell.net (Sam Schulenburg)
Date: Tue, 25 Apr 2000 03:07:28 GMT
Subject: Pythonwin COM Question
Message-ID: <8e3259$i5$1@nnrp1.deja.com>

In Visual Basic I have the following:
'===============================================
Dim theFrame As Varient
Dim dx as DocFile
Dim ex as New ExpSetup
Dim bRes as Integer

if (ex.Start(dx)) then
   While ex.GetParam(EXP_RUNNING,bRes)
       DoEvents 'Wait for detector to finish
   Wend
   dx.GetFrame 1, theFrame 'Retreive one image
end if
'===============================================

In Python I have the equivalent:

from PrincetonInst import *  # Get the renamed Makepy file for the
                             # COM object

theFrame = []                # Set up the Varient
dx = DocFile()               # Get the DocFile class
ex = win32com.client.Dispatch(ExpSetup.CLSID)
bRes = 0

if (ex.Start(dx)) then       # Start detector scanning  for data
   While ex.GetParam(EXP_RUNNING,bRes):  # Detector is scanning
      ## DoEvents  # Wait for detector to finish, This has no
                   # equivalent in Pythonwin
         s = 1     # This gives the while loop something to do
   #### This is where everything falls apart
   #### theFrame is not the correct datatype
   dx.GetFrame(1,theFrame)  # Retreive one image

How do I define the correct Varient type in Python?

Sam Schulenburg


Sent via Deja.com http://www.deja.com/
Before you buy.


From pecora at anvil.nrl.navy.mil  Sun Apr 30 10:00:06 2000
From: pecora at anvil.nrl.navy.mil (Louis M. Pecora)
Date: Sun, 30 Apr 2000 14:00:06 GMT
Subject: Do I always have to write "self." ?
References: <270420001706386648%pecora@anvil.nrl.navy.mil> <slrn8ghc6t.f01.scarblac-spamtrap@flits104-37.flits.rug.nl> <280420000746597329%pecora@anvil.nrl.navy.mil> <8ec67h$b14$1@ites.inria.fr>
Message-ID: <300420001000073453%pecora@anvil.nrl.navy.mil>

[[ This message was both posted and mailed: see
   the "To," "Cc," and "Newsgroups" headers for details. ]]

In article <8ec67h$b14$1 at ites.inria.fr>, Cedric Adjih
<adjih at crepuscule.com> wrote:

> or the other suggestions for read-only attributes:
> 
> 
> class Polynom2:
>         def delta(self):
>                 a,b,c=self.a,self.b,self.c
>                 return b*b-4*a*c

Yes, read only. Thanks.


From jmassung at magpiesystems.com  Tue Apr 18 15:47:50 2000
From: jmassung at magpiesystems.com (Jeff Massung)
Date: Tue, 18 Apr 2000 13:47:50 -0600
Subject: else clauses in while and for loops
References: <%TNK4.430$vr2.32615@news.uswest.net> <2JKK4.325$vr2.19414@news.uswest.net> <Zv_C4.6415$4N.169440@news.tpnet.pl> <38DD2EE3.C7F3CB54@webamused.com> <8boe2k$gmh$1@inputplus.demon.co.uk> <v0RD4.13$n68.186989568@newsb.telia.net> <8c7b29$37v$1@inputplus.demon.co.uk> <38F79B93.B03F0A6B@mindspring.com> <8dg14u$f08$1@nntp6.u.washington.edu> <8dh0ss$4ia$0@216.39.151.169>
Message-ID: <jU2L4.1208$zk4.22542@news.uswest.net>

Donn Cave wrote in message <8dh0ss$4ia$0 at 216.39.151.169>...
>Now it's simpler to say "else executes if break doesn't".

Got it now ;). Thanks.

Jeff




From thomas at cintra.no  Sun Apr 30 12:39:15 2000
From: thomas at cintra.no (thomas at cintra.no)
Date: Sun, 30 Apr 2000 16:39:15 GMT
Subject: Comments/Help on Index-module
Message-ID: <390c564b.102144@news.c2i.net>

Hi,

I`m trying to make a index-module. Words are mapped to a tuple
containg info using a dbhash-based dictionary. The tuple contains a
file-position and how many bytes to read. A second file contain a
serie of integers making a serie of ids. Three integers are used as a
id. 

I`ve managed to write integers to a single file, but I need to merge
several files created like above into one huge one. The problem is
that somewhere something got screwed up, and the integers I`ve put
into the scripts are not the same as the ones coming out when I try
lookup words.

I put in:
>>> x.add('eggs',(33,2,42))
and this comes out :
>>> looking for eggs: [(536870912, 16777216, -1610612641)]

Other values seem to get thru ok, but it looks like they are located
first in the file with integers. I therefore suspect that the error is
located  somewhere in the seeking in the binary-file.

If somebody understands what I`m trying to do can point to where I
screw up, or has another faster, better or perhaps easier way of doing
this, I`d be grateful

I`ve included my script below. Sorry if this is lots of text and take
up lots of bandwidth, but I`ve spent hours on this and don`t seem to
come any closer to a solution.

thanks.

Thomas 

#!/usr/bin/python

import dbhash
import shelve
import struct
import os

StillOpen = "Indexer is still open. Close before searching."
Nothoing2Merge = "Nothin to merge. Add more entries."

def opendb(file):
    return shelve.Shelf(dbhash.open(file, "crw"))

class Storage:

    def __init__(self):

        self.items = {}
        self.indexed_items = 0
        if not os.path.exists('main.dat'):
            self.outputfile1 = 'main.idx'
            self.outputfile2 = 'main.dat'
        else:
            self.outputfile1 = 'idx_cd.idx'
            self.outputfile2 = 'idx_cd.dat'
        self.dat = opendb(self.outputfile2)
        self.out = open(self.outputfile1, "wb")
        self.open = 1

    def close(self):
        self.dat.close()
        self.out.close()
        self.open = 0
        
    def flush(self):

        where_ = 0
        fmt = 'lll'
        sz = struct.calcsize(fmt)

        for key in self.items.keys():

            self.dat[key] = (where_, len(self.items[key]))
            where_ = where_ + len(self.items[key])

            for entry in self.items[key]:
                self.out.write(struct.pack(fmt, entry[0], entry[1],
entry[2]))

        self.items = {}
        self.indexed_items = 0
        self.dat.close()
        self.out.close()
        self.open = 0
        
    def add(self, word, id):

        if not self.items.has_key(word):
            self.items[word] = []

        self.items[word].append(id)
            
    def merge(self):

        if not os.path.exists('idx_cd.dat'):
            raise Nothoing2Merge
        else:

            processed = []
            fmt = 'lll'
            sz = struct.calcsize(fmt)
            
            os.rename('main.dat', '_main.dat')
            os.rename('main.idx', '_main.idx')

            dOld = opendb('_main.dat')
            iOld = open('_main.idx','rb')
            dNew = opendb('idx_cd.dat')
            iNew = open('idx_cd.idx','rb')

            d = opendb('main.dat')
            i = open('main.idx','wb')
            where = 0
            
            for key in dOld.keys():

                processed.append(key)
                whereOld, noOld = dOld[key]
                seek2 = whereOld
                iOld.seek(seek2)

                for j in range(0, noOld):
                    data = iOld.read(sz)
                    ds = struct.unpack(fmt, data)
                    i.write(struct.pack(fmt, ds[0], ds[1], ds[2]))

                if dNew.has_key(key):

                    whereNew, noNew = dNew[key]
                    seek2 = whereNew
                    iNew.seek(seek2)
                    noOld =  noOld + noNew

                    for j in range(0, noNew):
                        data = iNew.read(sz)
                        ds = struct.unpack(fmt, data)
                        i.write(struct.pack(fmt, ds[0], ds[1], ds[2]))

                d[key] = (where, noOld)
                where = i.tell()

            for key in dNew.keys():

                if not key in processed:
                    where_, no = dNew[key]
                    seek2 = where_
                    iNew.seek(seek2)

                    for j in range(0, no):
                        data = iNew.read(sz)
                        ds = struct.unpack(fmt, data)
                        i.write(struct.pack(fmt, ds[0], ds[1], ds[2]))

                    d[key] = (where, no)
                    where = i.tell()
                           
            d.close()
            i.close()
            dOld.close()
            iOld.close()
            dNew.close()
            iNew.close()
            
            if os.path.exists('_main.dat'):
                os.remove('_main.dat')
                os.remove('_main.idx')
            if os.path.exists('idx_cd.dat'):
                os.remove('idx_cd.dat')
                os.remove('idx_cd.idx')

class Locator:

    def __init__(self, words, intersection = 1):

        self.hits = {}
        self.result = []
        
        f = open('main.idx','rb')        
        d = opendb('main.dat')

        fmt = 'iii'
        sz = struct.calcsize(fmt)

        for word in words:

            self.hits[word] = []
            
            where, no = d[word]
            seek2 = where
            f.seek(seek2)

            for j in range(0,no):
                data = f.read(sz)
                ds = struct.unpack(fmt, data)
                self.hits[word].append(ds)

            workOn = []
            for key in self.hits.keys():
                workOn.append(self.hits[key])
            
        if intersection: 
            res = []
            if workOn: 
                    for x in workOn[0]:
                            for other in workOn[1:]:
                                    if not x in other:break
                            else:
                                res.append(x)
            for list in res:
                self.result.append(list)
                    
        else: # union => match all words
            res_ = {}
            for list in workOn:
                for item in list:
                    res_[item] = None
            self.result = res_.keys()
    
if __name__ == '__main__':

    x = Storage()
    x.add('spam',(21,2,32))
    x.add('spam',(24321,3232,3232))
    x.add('spam',(321,32,432))
    x.add('eggs',(33,2,42))
    x.add('spam',(33,2,42))
    x.add('python',(312,3232,44322))
    x.add('spam',(3323,12,443222))
    x.flush()

    x = Storage()
    x.add('python',(3323,12,142))
    x.add('monty',(1133,22,4662))
    x.add('cheese',(33213,121,1142))
    x.add('biscuit',(3323,132,41322))
    x.add('spam',(2333,32,432))
    x.flush()
    x.merge()

    y = Locator(['eggs'])
    print "looking for eggs:", y.result



From jeremy at cnri.reston.va.us  Thu Apr 20 13:23:26 2000
From: jeremy at cnri.reston.va.us (Jeremy Hylton)
Date: Thu, 20 Apr 2000 13:23:26 -0400 (EDT)
Subject: The Simple Economics of Open Source
Message-ID: <14591.15630.192472.440375@goon.cnri.reston.va.us>

A brief article in today's NY Times discusses a new (Feb 2000) paper
explores the economics of open source software:
http://www.nytimes.com/library/financial/columns/042000econ-scene.html
(The article requires user registration.)

The paper itself is available from one of the authors --
http://www.people.hbs.edu/jlerner/simple.pdf

The Simple Economics of Open Source

Abstract: "There has been a recent surge of interest in open
source software development, which involves developers at many
different localtions and organizations sharing code to develop and
refine programs.  To an economist, the behavior of individual
programmers and commercial companies engaged in open source projects
is initially startling.  We highlight the extent to which labor
economics, especially the literature on 'career concerns,' can explain
many of these projects' features.  Aspects of the future of open
source development process, however, remain somewhat difficult to
predict with 'off-the-self' economic models."

The authors are Josh Lerner -- 
    http://www.people.hbs.edu/jlerner/
-- and Jean Tirole -- 
    http://www.elsevier.nl/hes/homepage/cvs/tirole.htm

I haven't read the paper yet, but the abstract and the discussion in
the Times make it sound like a worthwhile read.  Thought I'd share the
pointer to the paper and see if anyone else has read the paper or has
comments on it.

Jeremy






From markx.daley at intel.com  Thu Apr 27 17:10:50 2000
From: markx.daley at intel.com (Daley, MarkX)
Date: Thu, 27 Apr 2000 14:10:50 -0700
Subject: ODBC error trap resolved!  Now...
Message-ID: <A63AFB20111ED311AC5500A0C96B7BF5015F0F50@orsmsx36.jf.intel.com>

OK, so the error trapping is good, but now I run into an append problem.  I
have a list defined in the beginning of the function, like so:

servername = ['ServerA', 'ServerB', 'ServerC']
for index in servername:
	try:
		#Big routine to collect and write data to a file.
	except dbi.opError:
		servername.append(index)

So, if one of the servers doesn't make the connection, it gets added to the
end of the list until it does connect.  This worked great in my test
scripts, but when put to practice, I get this:

Traceback (innermost last):
  File "<pyshell#1>", line 1, in ?
    collect.collect()
  File "C:\PROGRA~1\Python\collect.py", line 51, in collect
    servername.append(index)
AttributeError: 'string' object has no attribute 'append'

I didn't see this when I tested my append function.  What do I not
understand about lists in this context?

Once again, thank you for all your help in advance.

- Mark




From rupole at compaq.net  Wed Apr 19 19:36:40 2000
From: rupole at compaq.net (Roger Upole)
Date: Wed, 19 Apr 2000 23:36:40 GMT
Subject: Where do I send a contribution for the Starship recovery ?
Message-ID: <ETqL4.142055$ra.3215092@news.easynews.com>

I personally would consider it money well spent.
(unlike most of the other things I spend money on)
                    Roger Upole




From ljz at asfast.com  Sun Apr 23 12:01:19 2000
From: ljz at asfast.com (Lloyd Zusman)
Date: 23 Apr 2000 12:01:19 -0400
Subject: code not true?
References: <Z%vM4.1491$RM6.3518433@news-east.usenetserver.com> <1255627015-45592331@hypernet.com>
Message-ID: <ltpurgyfg0.fsf@asfast.com>

"Gordon McMillan" <gmcm at hypernet.com> writes:

> [ ... ]
> 
> The theoretically correct alternative is rationals. But they're 
> slow and use unpredictable amounts of memory. I think 
> Python should have BCD as an alternative. While slow, they're 
> predictable in memory use, and the mathematical 
> shortcomings they have are exactly those that a naive user 
> expects.

Python already has unlimited-precision long integers as part of the
language, and the `mpz' module that comes with the standard Python
release is another multiple-precesion option (built on top of GNU MP).
See sections 2.1.4 ("Numeric Types") and 15.2 ("mpz") of the Python
Library Reference manual.

It should be fairly straightforward to use the unlimited-precision
long integers or `mpz' to implement a BCD module that handles decimal
arithmetic and not just integer math.

Ages ago I wrote an unlimited-precision BCD math library in C ... I'm
sure I could do it again (and much more easily and quickly) in Python.

Then again, I'd hate to re-invent the wheel ... I'd be surprised if
someone hasn't already created such a thing in Python ... ???


> [ ... ]
> 
> - Gordon

-- 
 Lloyd Zusman
 ljz at asfast.com


From mhammond at skippinet.com.au  Fri Apr 14 09:31:02 2000
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Fri, 14 Apr 2000 13:31:02 GMT
Subject: Future compatibility of the Python C API
References: <lwEJ4.7980$Ib7.101092@typhoon2.kc.rr.com>
Message-ID: <q4FJ4.1198$1D.10037@news-server.bigpond.net.au>

No one can speak for Guido, so here are my opinions and guesses :-)

"Keith Davidson" <kbd at iti-oh.com> wrote in message
news:lwEJ4.7980$Ib7.101092 at typhoon2.kc.rr.com...
> 1) Has there been any discussion about the upwards compatibility of the
C
> API in future versions of Python?

Yes.

>  Historically our libraries have had a
> 7-10+ year lifespan where we provide 95+% backwards compatibility (i.e.
we
> don't want our C API users to have to make more than a day or so worth
of
> changes between releases).

And this is the goal with the 1.x family.

> My big concern is Python 3000- what are the odds
> that its C API will be backwards compatibility with Python 1.x?

Not very good.  If it is a real problem, simply dont upgrade.  I would
however expect the changes to cause compiler errors that are simple to
fix - ie, very similar APIs, and unlikely to silently change behaviour.

> 2) We are considering using the Python C API heavily in a set of
libraries
> that will be used by both C applications and as Python extension
modules.
> Has anyone else done this?  How did it work out?

Sure.  Great.  Really, there is no difference between a set of libraries
and an application that embeds Python - just the entry points change!

Mark.




From schorsch at schorsch.com  Wed Apr 19 06:34:36 2000
From: schorsch at schorsch.com (Georg Mischler)
Date: Wed, 19 Apr 2000 10:34:36 GMT
Subject: urllib redirect
References: <4.1.20000418125557.030592d0@202.95.1.37> <Pine.OSF.4.20.0004181610230.20942-100000@azure.dstc.edu.au>
Message-ID: <8dk23t$4c2$1@nnrp1.deja.com>

Anthony J Wilkinson wrote:
> On Tue, 18 Apr 2000, Thomas wrote:
>
> > I want one of my functions to redirect the URL.
> > (ie I call the script and off I go to www.spam.com)
> >
> > So, here's what I get, and have not overcome yet:
> >
> > import urllib
> > def simpleredirect() :
> >     ############
> >      print "Content-type: text/html"
> >      print urllib.urlopen('http://www.spam.com').read()
> >
> > simpleredirect()
>
> You really aren't redirecting but attempting to get the data yourself.
> If you just print out the following you can use the browser's own
> redirection facility.
>
> Content-type: text/html
> <html>
> ...
> </html>

This can be done in a much simpler way, whitout sending any
acutal page data to the client:

  your_new_path = 'http://www.spam.com/'

  print 'Status: 302 Redirected'
  print 'Location: %s' your_new_path

As soon as the browser (works with *every* browser!) sees the
"Location:" header, it will stop and try fetching the new page.

Unless the original poster didn't really want redirection
as claimed in the text, but rather a proxy service as hinted
at in the code. I think the latter is only recommendable in
very special circumstances.


Have fun!

-schorsch

--
Georg Mischler  --  simulations developer  --  schorsch at schorsch.com
+schorsch.com+  --  lighting design tools  --  http://www.schorsch.com/


Sent via Deja.com http://www.deja.com/
Before you buy.


From jae at ilk.de  Sat Apr 22 00:29:56 2000
From: jae at ilk.de (Juergen A. Erhard)
Date: Sat, 22 Apr 2000 06:29:56 +0200
Subject: How2 use curses for data input
In-Reply-To: <38FF8F9C.9D533E74@sprint.ca> (message from Miles Thompson on
	Thu, 20 Apr 2000 20:15:40 -0300)
References: <38FF6DE2.5751423A@sprint.ca> <slrn8fusjd.99a.grant@grante.comtrol.com> <38FF8F9C.9D533E74@sprint.ca>
Message-ID: <22042000.4@sanctum.jae.ddns.org>

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

>>>>> "Miles" == Miles Thompson <milesthompson at sprint.ca> writes:

    Miles> I located it on the RH system here in the office at
    Miles> home. The target machine is Debian, and I think I can
    Miles> install RPM on it so that if need be, I can get the
    Miles> RPM. Debian doesn't seem to have a snack package.

My Debian system has a snack package... `snack' and `snack-dev'.

But that is potato (the upcoming release)... so it might be that slink
didn't.

If you need it for slink... well, you could try to build it for slink.
Might even work ;-) (*Building* a Debian source package is really
simple... lemme know if you try it and need assistance...)

HTH(S),

Bye, J

- -- 
J?rgen A. Erhard      eMail: jae at ilk.de      phone: (GERMANY) 0721 27326
     MARS: http://members.tripod.com/Juergen_Erhard/mars_index.html
   GNU Privacy Guard (http://www.d.shuttle.de/isil/crypt/gnupg.html)
  "Windows NT" is an acronym for "Windows? No thanks." -- Russ McManus
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.1 (GNU/Linux)
Comment: Use Mailcrypt and GnuPG <http://www.gnupg.org/>

iEYEARECAAYFAjkBKsAACgkQN0B+CS56qs3EdACeIlyz6xB7/RZTQ1ZowMnPPqpN
kHsAoI3KNP27chgRgC0ItaV5/aPVPlOe
=ByvG
-----END PGP SIGNATURE-----



From rhicks at nospam.rma.edu  Mon Apr 17 11:10:34 2000
From: rhicks at nospam.rma.edu (Robert Hicks)
Date: Mon, 17 Apr 2000 15:10:34 GMT
Subject: Editors and books
References: <XBFK4.16$Uq2.1831@news.uswest.net>
Message-ID: <KPFK4.10$Pa.2204@news>

try this... http://www.python.org/download/download_mac.html

I am not a Mac person...but it might get you there.

Bob

"Jeff Massung" <jmassung at magpiesystems.com> wrote in message
news:XBFK4.16$Uq2.1831 at news.uswest.net...
> First, I'd like to give my undying praise to PYTHON: Essential Reference,
by
> David M. Beazley. I'm an accomplished C/C++ programmer who wanted to learn
> Python. This book was IT! Anyone whose reading this that is in my shoes -
> get this book! It's some of the best $35 I've spent!
>
> Next, a question for those who've been doing this: is there a color IDE
> (just a nicer thing that IDLE) for the Mac? I've heard that Alpha (which I
> just downloaded) has a Python mode, but I can't seem to find it. Is there
> something else, either? can MPW do it? BBEdit?
>
> Thanks, guys.
>
> Also, on another NG there's talk about Python, and how it does (and still
> doesn't) do garbage collecting. Could someone here expand on it so I know
> what it can and can't do? Memory is a big issue in my project and I need
to
> know how to structure my program so that garbage is collected. Thanks
> again!!
>
> -- Jeff Massung (jmassung at magpiesystems.com)
> -- Lead programmer for Magpie Systems
> -- Bringing smart pigs to the pipeline industry!
>
>




From Michael.Scharf at gmx.de  Fri Apr 14 15:43:33 2000
From: Michael.Scharf at gmx.de (Michael Scharf)
Date: Fri, 14 Apr 2000 21:43:33 +0200
Subject: Comparing PythonWin and IDLE?
References: <NDBBLJJLLFOJMLCMJPLOGEEADJAA.charlie@intelligenesis.net>
Message-ID: <38F774E5.E34B0365@gmx.de>

Charlie Derr wrote:
> 
>         instead of:
> 
> import myModule
> 
>         try this:
> 
> reload(myModule)
> 
>         and you should be able to get your recently modified changes to take effect
> without restarting python

Well, not quite:

if you wrire (in another module)

  from myModule import MyFunction,MyClass
  ...
  MyFunction()
  my_callback=MyFunction
  my_instance=MyClass()

reload will *not* change the function nor the myinstance.
If you would write:

  import myModule
  ...
  myModule.MyFunction()
  my_callback=myModule.MyFunction
  my_instance=myModule.MyClass()

the new version of MyFunction an MyClass will be called, but
existing instances of MyClass will not be changed to the new version.
Also the old version of my_callback would be called.

I know it is possible to change the code of existing functions.
So, with some ticks it should be possible to replace the code
of existing functions. The same must be done for all mutable objects
(list, dicts etc.). Only not mutable direct imported variable (int,
string, tuple etc.) cannot be changed that way.

I just wonder if anybody has written such a clever reimport...

Michael
-- 
     ''''\     Michael Scharf
    ` c-@@     TakeFive Software GmbH, a Wind River Company
    `    >     http://www.WindRiver.com
     \_ V      mailto:Michael.Scharf at gmx.de


From tom-main at REMOVEME.home.com  Fri Apr 14 15:29:11 2000
From: tom-main at REMOVEME.home.com (Tom)
Date: Fri, 14 Apr 2000 19:29:11 GMT
Subject: How do I undo an import??
References: <7QrJ4.185421$Hq3.4353581@news2.rdc1.on.home.com> <m3d7nthbbj.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <bkKJ4.190780$Hq3.4442202@news2.rdc1.on.home.com>

After importing NetCfg I type:

>>> del sys.modules['NetCfg']

Python responds as follows:

Traceback (innermost last):
  File "<stdin>", line 1, in ?
NameError: sys

Any ideas??

Thanks,

Tom.

"Michael Hudson" <mwh21 at cam.ac.uk> wrote in message
news:m3d7nthbbj.fsf at atrus.jesus.cam.ac.uk...
> "Tom" <tom-main at REMOVEME.home.com> writes:
>
> > I'm a C++ programmer, new to Python, working with v1.5.2 on Windows.
> >
> > I run the command-line interpreter, then I type:
> >
> > >>> from NetCfg import *
> >
> > to import my extension.  Now I want to unload my NetCfg DLL without
exiting
> > the interpreter.  How do I do this?
>
> You don't, in general.
>
> del sys.modules['NetCfg']
>
> goes some of the way, but it's unlikely it goes far enough to let the
> dll be unloaded from memory.
>
> > Also, I type the above import command every time I start the
interpreter.
> > Is there some way to get have this command executed automatically?
>
> Set the environment variable "PYTHONSTARTUP" to point to a file
> containing commands you want executed.
>
> At least, that's what I do on Linux.
>
> Cheers,
> M.
>
> --
>   ... but I guess there are some things that are so gross you just have
>   to forget,  or it'll destroy something within you.  perl is the first
>   such thing I have known.                 -- Erik Naggum, comp.lang.lisp




From greg at cosc.canterbury.ac.nz  Thu Apr  6 21:54:35 2000
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Fri, 07 Apr 2000 13:54:35 +1200
Subject: Why should I switch to Python? - Infinity of Primes
References: <1257244378-5884705@hypernet.com>
Message-ID: <38ED3FDB.D4954AB8@cosc.canterbury.ac.nz>

Gordon McMillan wrote:
> 
> David C. Ullrich wrote:
> 
> There are at least 3 well known constructive
> proofs that all odd numbers are prime:

And, of course, you are all familiar with the proof
that every natural number has an interesting property:

The number 1 has the interesting property of being
the first natural number. Now suppose that all natural 
numbers up to n have interesting properties, but
n+1 does not. That would make n+1 the first natural
number without an interesting property -- and that
is an interesting property!

This-is-my-last-post-on-this-topic-really-I-promise,

-- 
Greg Ewing, Computer Science Dept,
+--------------------------------------+
University of Canterbury,	   | A citizen of NewZealandCorp, a	  |
Christchurch, New Zealand	   | wholly-owned subsidiary of USA Inc.  |
greg at cosc.canterbury.ac.nz	   +--------------------------------------+


From dalke at acm.org  Sat Apr  1 06:28:36 2000
From: dalke at acm.org (Andrew Dalke)
Date: Sat, 1 Apr 2000 04:28:36 -0700
Subject: Function override?
References: <8c0dtu$lr3$1@nnrp1.deja.com> <38E4D7A4.4C83CFF7@math.okstate.edu>
Message-ID: <8c4mds$5p7$2@slb6.atl.mindspring.net>

David C. Ullrich wrote:
>    This is certainly one of the things that _I_ find keen,
>amyway - I hope the people who are talking about adding
>stricter typing to Python make it optional somehow. There
>are obvious advantages to stricter typing but you lose
>something as well. (For example, I get a big kick out of
>the fact that exactly the  same euclidean-algorithm code
>works to find the GCD of two integers or the GCD of
>two polynomials. Etc.)


I went to a talk by Stepanov about the Standard Template
Library in C++.  One of his examples was a geneic GCD
algorithm which works on integers and polynomials.  Thus,
stricter typing and generic programming don't have to
be the same thing.

In Python, this could be done by a syntax requiring not
that "i" and "j" are ints, but that type(i) == type(j).

                    Andrew
                    dalke at acm.org





From thomas at xs4all.net  Tue Apr 18 16:31:05 2000
From: thomas at xs4all.net (Thomas Wouters)
Date: Tue, 18 Apr 2000 22:31:05 +0200
Subject: Pointers
In-Reply-To: <8dhu2j$1gm$1@proxy2.fe.internet.bosch.de>; from Klaus-Gerd.Meyer@de.bosch.com on Tue, Apr 18, 2000 at 05:13:21PM +0200
References: <38D00DCE.9740380F@be-research.ucsd.edu> <8bfsqo$kje$1@tribune.oar.net> <dp4s9wl9df.fsf@mraz.iskon.hr> <uPYC4.55946$3b6.224410@ozemail.com.au> <4k5D4.7566$0o4.54442@iad-read.news.verio.net> <38FB98F1.239EC0FB@whetstonelogic.com> <slrn8foe80.bd3.scarblac-spamtrap@flits104-37.flits.rug.nl> <8dhu2j$1gm$1@proxy2.fe.internet.bosch.de>
Message-ID: <20000418223105.K15664@xs4all.nl>

On Tue, Apr 18, 2000 at 05:13:21PM +0200, Klaus-Gerd Meyer wrote:

> Integers and strings with same value are identical, but tuples and list with
> same value are not identical...

Not all integers and strings with the same value are identical, though!

> >>> x = 1
> >>> y = 1
> >>> id(x),id(y)
> (7879224, 7879224)
> >>> x is y            #x and y: same value and identical
> 1

>>> x = 100000
>>> y = 100000
>>> id(x),id(y)   
(135113376, 135230104)
>>> x is y
0
>>> x == y
1

> >>>
> >>> x = "a"
> >>> y = "a"
> >>> id(x),id(y)
> (8041104, 8041104)
> >>> x is y            #x and y: same value and identical
> 1

>>> x = "Ham, spam, eggs, spam, bacon, spam, pate foie gras, spam, saussages and spam"
>>> y = "Ham, spam, eggs, spam, bacon, spam, pate foie gras, spam, saussages and spam"
>>> id(x),id(y)   
(135233304, 135240648)
>>> x is y
0
>>> x == y
1

For strings, you can 'fix' this by using 'intern()':

>>> y = intern("Ham, spam, eggs, spam, bacon, spam, pate foie gras, spam, saussages and spam")
>>> x = intern("Ham, spam, eggs, spam, bacon, spam, pate foie gras, spam, saussages and spam")
>>> id(x),id(y)
(135241592, 135241592)
>>> y == x
1
>>> y is x
1

But you have to force *both* strings this way, interning one isn't enough.
It also doesn't work for integers (i dont think there is a way to do it with
ints) so the best solution is just to not rely on this behaviour.

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From lars.lundstedt at telia.com  Wed Apr 26 00:30:35 2000
From: lars.lundstedt at telia.com (Lars Lundstedt)
Date: Wed, 26 Apr 2000 04:30:35 GMT
Subject: SV: Don't understand error message
References: <lukN4.1574$wYl.191480320@newsb.telia.net> <etd1z3uf5ua.fsf@w20-575-109.mit.edu>
Message-ID: <LhuN4.4195$Za1.61781@newsc.telia.net>

Ahhhh...that did the trick!

Thanks!

Alex <cut_me_out at hotmail.com> skrev i
diskussionsgruppsmeddelandet:etd1z3uf5ua.fsf at w20-575-109.mit.edu...
>
> > It goes like this:
> >
> > import sys, glob, operator
> > print sys.argv[1:]
> > sys.argv = reduce(operator.add, map(glob.glob, sys.argv))
> > print sys.argv[1:]
> >
> > When called with "python glob.py *.py" on Win98 I'm supposed to get a
list
> > of all .py files in the current directory but instead I get:
> >
> > Traceback (innermost last):
> >   File "glob.py", line 1, in ?
> >     import sys, glob, operator
> >   File "glob.py", line 3, in ?
> >     list = reduce(operator.add, map(glob.glob, '*.py'))
> > TypeError: call of non-function (type module)
> >
> > Anyone know why?
>
> Yeah, the problem is, your file is called glob.py, and that is loaded
> when you import glob.  Instead, you should be loading the glob from the
> standard python library.  To get around this, one thing you could do is
> rename your glob.py file to my_glob.py, and remove your old glob.py and
> glob.pyc.  Then 'python my_glob.py *.py' should work fine.  At least,
> that solved the problem for me.
>
> Alex.




From dirck at pacbell.net  Wed Apr  5 13:57:54 2000
From: dirck at pacbell.net (Dirck Blaskey)
Date: Wed, 5 Apr 2000 10:57:54 -0700
Subject: Extending Python with C-Libs
References: <8cfpl8$9b2$1@newnews.ilk.de>
Message-ID: <F6LG4.10$Od3.7950@news.pacbell.net>

Frank Rotermund <frankr at first-mind.de> wrote in message
news:8cfpl8$9b2$1 at newnews.ilk.de...
> Hello
>
> I have a Library written in C with its Headerfiles but without the
> Sourcecode.
> Does anybody know an easy way to extend Python with this Library?

Don't miss http://www.swig.org
it will make the process easier.




From just at letterror.com  Sat Apr  8 13:23:48 2000
From: just at letterror.com (Just van Rossum)
Date: Sat, 8 Apr 2000 18:23:48 +0100
Subject: No threads in MacPython 1.5.2?
In-Reply-To: <8cl7rh$kvm$3@flotsam.uits.indiana.edu>
Message-ID: <l0310280ab5151a93baca@[193.78.237.146]>

At 6:02 PM +0000 07-04-2000, Gregory L. Hansen wrote:
>My Python can find Tkinter now, thanks for the help.  But now it can't
>find the module named "thread".  And yet there it is in the Lib folder,
>thread.py

Erm, there is no such module as thread.py.

>and threading.py, and the search path $(PYTHON):Lib is set for
>it.
>
>Now what's wrong?

Nothing: MacPython is currently not built with thread support. That's not
so much because the OS like Christian suggested, but rather that there was
no posix-compatible thread implementation available, making Python
integration hard. However, this has changed with GUSI2 (a posixey I/O
emulation library), and the upcoming MacPython 1.6a1 *does* have threads. I
haven't seen it yet, so I don't know how well it works. Subscribe to the
pythonmac-sig, and you'll soon find out!

(In the meantime I have microthreads running like a charm, thanks to
Stackless Python ;-)

Just





From grisha at ispol.com  Fri Apr 21 09:44:47 2000
From: grisha at ispol.com (Gregory Trubetskoy)
Date: Fri, 21 Apr 2000 09:44:47 -0400
Subject: suggestion on an operator
In-Reply-To: <3900043E.78539877@nsw.bigpond.net.au>
References: <Pine.LNX.4.21.0004202210480.9838-100000@saturn.ispol.com> <3900043E.78539877@nsw.bigpond.net.au>
Message-ID: <Pine.LNX.4.21.0004210938400.12562-100000@saturn.ispol.com>

On Fri, 21 Apr 2000, John Lehmann wrote:

> Gregory Trubetskoy wrote:
> 
> What is wrong with using the period again?
> 
> eg.
> 
> PythonAuthenHandle mypackage.mymodule.authenticate


With a period I don't know where the module ends and stuff in it begins. I
have to import the module to gain access to what's in it.

For example -

a.b.c.d could be interpreted as package.module.object.function, or
package.subpackage.subsubpackage.function.

If Python allowd me to access string.split without doing "import string"
first, then the period would work, but such is not the case.

David Gooder suggested a space - a space won't work because I would like
to keep the option of allowing multiple handler separated by space, a la
mod_perl.

Grisha





From pj at sgi.com  Tue Apr 18 11:15:06 2000
From: pj at sgi.com (Paul Jackson)
Date: 18 Apr 2000 15:15:06 GMT
Subject: else clauses in while and for loops
References: <2JKK4.325$vr2.19414@news.uswest.net> <8dg14u$f08$1@nntp6.u.washington.edu> <%TNK4.430$vr2.32615@news.uswest.net> <2VTK4.421$wYl.187705344@newsb.telia.net>
Message-ID: <8dhu5q$14anj$1@fido.engr.sgi.com>

Perhaps I need some more coffee, but I suspect that the
example in the previous message is indented wrong, and
should be:

        s = list_of_strings

        for i in range(len(s)):
            if s[i] == something:
                break
        else:
            print "didn't find anything"
            raise ValueError

        print "index is", i
-- 

=======================================================================
I won't rest till it's the best ...	   Software Production Engineer
Paul Jackson (pj at sgi.com; pj at usa.net) 3x1373 http://sam.engr.sgi.com/pj


From owen at astroNOJNK.washington.edu.invalid  Thu Apr 20 14:47:27 2000
From: owen at astroNOJNK.washington.edu.invalid (Russell E. Owen)
Date: Thu, 20 Apr 2000 11:47:27 -0700
Subject: Mac Python Questions
References: <IRjL4.271$CP3.3183@news.uswest.net> <38FE8032.DE233C29@cosc.canterbury.ac.nz>
Message-ID: <8dnjbt$hcs$1@nntp6.u.washington.edu>

In article <38FE8032.DE233C29 at cosc.canterbury.ac.nz>, Greg Ewing 
<greg at cosc.canterbury.ac.nz> wrote:

>Jeff Massung wrote:
>> 
>> b) No matter what I do, I can't get Tkinter to work :(.
>
>Neither can I, and from what I've seen here, neither
>can anyone else. Tkinter just seems to be broken on
>the Mac at present.

It works fine.

Tricks:
- Make sure the necessary paths are added (as another poster detailed)

- Make sure PythonInterpreter is configured to NOT go into interactive 
mode after the script completes. (Drop PythonInterpreter onto 
EditPythonPrefs, then push the button that lets you configure startup 
defaults; it's the first checkbox). I have not seen this mentioned 
before, but I find it crucial for Tk scripts to exit correctly!

- Put your Tk scripts in files and drop them onto PythonInterpreter. 
Don't use the IDE!!! (Typing scripts into the interpreter may also work; 
I've not tried it.)

-- Russell


From darcy at vex.net  Mon Apr 24 07:50:53 2000
From: darcy at vex.net (D'Arcy J.M. Cain)
Date: 24 Apr 2000 11:50:53 GMT
Subject: Trouble compiling PyGreSQL 2.4 on Linux
References: <Pine.LNX.4.21.0004241126280.6019-100000@fep132.fep.ru>
Message-ID: <8e1cet$kmd$1@news.tht.net>

Oleg Broytmann <phd at phd.russ.ru> wrote:
>> How stable is the DB-API version of the PostgreSQL module? Can it
>> be used for production work (or semi serious)?

>    It was released few days ago, so it just hadn't been tested...

In fact it hasn't been released yet but you can pick up the beta for testing
from ftp://ftp.druid.net/pub/distrib/PyGreSQL-beta.tgz.

-- 
D'Arcy J.M. Cain <darcy at vex.net>           |  Democracy is three wolves
http://www.vex.net/                        |  and a sheep voting on         
+1 416 425 1212     (DoD#0082)    (eNTP)   |  what's for dinner.


From dwhite2 at seas.upenn.edu  Fri Apr  7 18:00:10 2000
From: dwhite2 at seas.upenn.edu (dave white)
Date: Fri, 07 Apr 2000 18:00:10 -0400
Subject: Why should I switch to Python? - Infinity of Primes
References: <000601bf9f8e$bd31e780$ea2d153f@tim> <pcoln2q3bet.fsf@math.ntnu.no>
Message-ID: <38EE5A40.4557F4EA@seas.upenn.edu>

Since we're on the subject of primes...  I've heard that there's a nice
probabilistic algorithm for computing large primes using Fermat's Little
Theorem.  Python seems like it would be well suited for this since
bignums are built into the language.  Does anyone know this algorithm?  

Fermat's Little Theorem:
Let p be a prime which does not divide the integer a, then a^(n-1) =
1(mod p)

cheers,
david


From thomas at xs4all.net  Tue Apr 18 04:28:17 2000
From: thomas at xs4all.net (Thomas Wouters)
Date: Tue, 18 Apr 2000 10:28:17 +0200
Subject: Here is some starship.python.net news
In-Reply-To: <38FBE4DB.67289F42@mindspring.com>; from echuck@mindspring.com on Tue, Apr 18, 2000 at 12:30:19AM -0400
References: <38FA9A30.33DD1994@concentric.net> <38FB90B4.35E9CA1B@bullfrog-tech.com> <38FBE4DB.67289F42@mindspring.com>
Message-ID: <20000418102817.J15664@xs4all.nl>

On Tue, Apr 18, 2000 at 12:30:19AM -0400, Chuck Esterbrook wrote:

> The situation with the Starship seems to reveal that:

> [a] It's highly relied upon by the Python community (consider all the
> modules people find at V of P that are really hosted on the Starship for
> downloading).

> [b] It's not well monitored and/or taken care of

> [c] There is no suitable backup/contingency machine to switch over to.

> How can the PSA change the situation to address [b] and [c]? I'd certainly
> be willing to pay higher dues if that's what it took, but there may be
> other ideas.

Official mirrors ? I'd be willing to lobby for a starship (and python.org ?)
mirror at XS4ALL (dutch ISP, where I work, so it shouldn't be too hard to
arange) but it kind of depends on the needs for the starship website --
having different platforms it runs on might screw up the scripts that run
there. Would having a data/webpage only mirror be sufficient ?

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From btang at pacific.jpl.nasa.gov  Mon Apr 17 19:23:33 2000
From: btang at pacific.jpl.nasa.gov (Benyang Tang)
Date: Mon, 17 Apr 2000 16:23:33 -0700
Subject: RE question
Message-ID: <38FB9CF5.444457CF@pacific.jpl.nasa.gov>


>>> m = re.match("([abc])+", "abc"); m.group(1)
'c'

Why is it 'c', instead of 'a'?


From aa8vb at yahoo.com  Wed Apr 12 11:32:47 2000
From: aa8vb at yahoo.com (Randall Hopper)
Date: Wed, 12 Apr 2000 11:32:47 -0400
Subject: Trouble with Tkinter under IDLE
In-Reply-To: <8F13B1259anthonylander@47.80.11.122>
References: <8F139460Canthonylander@47.80.11.122>
 <38F37BF9.DCAB3570@exceptionalminds.com> <8F13B1259anthonylander@47.80.11.122>
Message-ID: <20000412113247.B2446939@vislab.epa.gov>

Anthony Lander:
 |Looking through the Tkinter.py source, it seems that there is a variable 
 |called _default_root which wants to be set to the default root window. Is 
 |the normal procedure to check this variable and use it as the root window 
 |if it's defined?

No.  By convention any module identifier that begins with an _ is private.
Tkinter sets this internally.

I believe idIf you create a widget and don't specify a parent, Tkinter
takes it to be default_root.

-- 
Randall Hopper
aa8vb at yahoo.com



From fredh at ix.netcom.com  Wed Apr  5 02:25:59 2000
From: fredh at ix.netcom.com (Fred W. Helenius)
Date: Wed, 05 Apr 2000 02:25:59 -0400
Subject: Why should I switch to Python? - Infinity of Primes
References: <NDBBKEGCNONMNKGDINPFGEJDDFAA.infonuovo@email.com> <8ccoug$f4o$1@pegasus.csx.cam.ac.uk> <7F51E902E09F3F60.8F612607812ED16B.82CB8B41435E14FD@lp.airnews.net> <38EA369C.26A1C25@math.okstate.edu> <38EAC8C4.ADC25640@cosc.canterbury.ac.nz>
Message-ID: <gklles093dk38k8s7h3r11okr4uftmpfus@4ax.com>

Greg Ewing <greg at cosc.canterbury.ac.nz> wrote:

>If someone actually came up with a formula for constructing
>primes, it would be rather large news -- isn't that one of
>the Big Unsolved Problems?

Actually, many formulas for primes are known; the catch is that
they are too inefficient for practical use, as the evaluation of
the formulas covertly includes all the work to test many numbers
for primality, or depends upon the use of a real constant that
encodes all the primes.  A few such formulas are exhibited at
http://mathworld.wolfram.com/PrimeFormulas.html .

The sci.math FAQ includes many additional references to formulas
for primes, see
http://www.cs.unb.ca/~alopez-o/math-faq/node36.html .

-- 
Fred W. Helenius	<fredh at ix.netcom.com>



From tismer at tismer.com  Sun Apr  2 10:52:12 2000
From: tismer at tismer.com (Christian Tismer)
Date: Sun, 02 Apr 2000 16:52:12 +0200
Subject: Debugging confusion -- too many stacks!
References: <000a01bf9c6b$3f80c2a0$752d153f@tim> <38E739A1.45F4B199@tismer.com> <m37legk3kx.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <38E75E9C.7501756C@tismer.com>

Hey, we forgot about a stack :-))

Michael Hudson wrote:
> 
> Christian Tismer <tismer at tismer.com> writes:
> 
> > This VM stack could also be replaced by a set of registers, since
> > the maximum stack size is always known at compile time.
> 
> My gut tells me that making this work with exceptions would be a dog.

I don't think so. Exception handlers have their own little stack
in each frame. It is currently fixed to 20 items, wasting quite
a lot of space most of the time (and this will vanish in
Stackless "Heresy" 1.2), but it won't cause problems with a
register machine.

> It's the sort of thing that might be fun to play with if the compiler
> and eval loop were written in Python, but if I have to hack C to play
> with these things, it just becomes tedious.  Still, maybe if I have a
> month or two with nothing else to do...

A while ago, I considered this, too. Meanwhile I don't, since
it doesn't bear a very reasonable speed gain. The win would be
to have less of the tiny operations occouring all the time,
and to have less of stack pointer operations. But when looking
closer to it, you will find that the most slowdown lies in
the outmost interpreter loop, which checks for periodic things
for every opcode, regardless wether it is very fast or a slow
function call.
In Stackless Python, I have stolen these cycles already. The eval
loop has some internal shortcuts to avoid this many checks. The fast
opcodes are all running as fast as possible, avoiding these
checks. One could still save some jumps and some assignments,
but this is not enough to justify the lots of changes that a
register implementation would impose to other modules.

It can still make sense to provide a hand-full of combined
opcodes which have high frequency. jump_if_true combined
with pop_top for instance, or some load/store combinations.
If you have a proposal, and if you will support them through
the bytecodehacks, I'd be happy to add them to Stackless Heresy.

Thanks for reminding me of that extra stack :-)

ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaunstr. 26                  :    *Starship* http://starship.python.net
14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     we're tired of banana software - shipped green, ripens at home



From moshez at math.huji.ac.il  Fri Apr 28 00:56:08 2000
From: moshez at math.huji.ac.il (Moshe Zadka)
Date: Fri, 28 Apr 2000 07:56:08 +0300 (IDT)
Subject: Newbie Question
In-Reply-To: <sgh4qdmbqtj130@corp.supernews.com>
Message-ID: <Pine.GSO.4.10.10004280755340.8383-100000@sundial>

On Thu, 27 Apr 2000, Arnaldo Riquelme wrote:

> Suppose I have a file name tick.tkr that looks like this
> 
> SUNW.
> MFST.
> FOO.
> SPAM.
> MOO.
> End_Of_File
> 
> I want to read the file into a list
> 
> myfile = open("tick.tkr", "r")
> mylist = myfile.readlines()
> 
> Now I have mylist with the contents of tick.tkr. "mylist" looks like this
> ['SUNW.\012', 'MSFT.\012', ..........'End_of_File/012']
> 
> How do I read the file into a list without getting the newline character, so
> I can accomplish this
> 
> for tic in mylist:
>     #create file tic.prn
> 
> I know the other P language has something like 'chop' or something like
> that, but I'm not going there.

Try "string.strip". In your case, it should do fine.
--
Moshe Zadka <mzadka at geocities.com>. 
http://www.oreilly.com/news/prescod_0300.html
http://www.linux.org.il -- we put the penguin in .com




From tim_one at email.msn.com  Thu Apr  6 02:09:59 2000
From: tim_one at email.msn.com (Tim Peters)
Date: Thu, 6 Apr 2000 02:09:59 -0400
Subject: Why should I switch to Python? - Infinity of Primes
In-Reply-To: <pcoog7najs8.fsf@math.ntnu.no>
Message-ID: <000601bf9f8e$bd31e780$ea2d153f@tim>

[Harald Hanche-Olsen]
> ...
> Since this is a posted to a computer language group, after all, I
> include my own simple implementation of the Sieve in Haskell:
>
> primes :: [Integer]
> primes = 2 : ([3..] `except` composites)
>
> composites :: [Integer]
> composites  = flatten (map (\ p -> [p*p, p*(p+1)..]) primes)
>
> union :: [Integer] -> [Integer] -> [Integer]
> xl@(x:xs) `union` yl@(y:ys) | x<y       = x : xs `union` yl
>                             | x>y       = y : xl `union` ys
>                             | otherwise = x : xs `union` ys
>
> except :: [Integer] -> [Integer] -> [Integer]
> xl@(x:xs) `except` yl@(y:ys) | x<y       = x : xs `except` yl
>                              | x==y      = xs `except` ys
>                              | otherwise = xl `except` ys
>
> flatten :: [[Integer]] -> [Integer]
> flatten ((x:xs):zz) = x : xs `union` (flatten zz)
>
> I wonder if someone can come up with an equally immediate solution
> (i.e., a straight translation from the mathematical formulation) in
> python?

Hmm.  Try out this one-liner Haskell defn, which is so pretty it *should*
bring tears of joy to everyone's eyes <blink>:

primes = sieve [2..]
    where sieve (p:rest) = p : sieve [n | n <- rest, mod n p /= 0]

Within the last month or two I posted a Python version of that under subject
"Fun w/ Lazy Streams".  Given general machinery for supporting lazy streams
(see the thread), the sieve example ended up looking like:

def removemults(s, n):   # s but without any multiples of n
    def notdivisible(i, n=n):
        return i % n
    return sfilter(s, notdivisible)

def sieve(s):
    return Stream(s.head(),
                  lambda s=s:
                      sieve(removemults(s.tail(), s.head())))

primes = sieve(srange(2))

That's approximately as irritating as it is in Scheme, were laziness also
needs to be faked.

> Indeed, I would be tempted to say that Haskell is better suited for
> learning (some kinds of) mathematics than python.
>
> but-I-wouldn't-dare-say-so-in-comp.lang.python-ly y'rs,

I would -- I've often recommended Haskell here.  Learning to think in a
purely functional way is really helpful, right up until it get
counterproductive <wink>.

if-god-intended-us-to-use-functional-languages-he-wouldn't-have-
    created-guido-ly y'rs  - tim





From dworkin at ccs.neu.edu  Wed Apr  5 18:56:22 2000
From: dworkin at ccs.neu.edu (Justin Sheehy)
Date: 05 Apr 2000 18:56:22 -0400
Subject: was Re: Why should I switch to Python?
In-Reply-To: m.faassen@vet.uu.nl's message of "2 Apr 2000 18:16:14 GMT"
References: <seas5ia89177@corp.supernews.com> <1EeF4.43$CA.2422@news> <oqzord6h5i.fsf@titan.progiciels-bpi.ca> <38E618D7.A15F9D9B@webwitches.com> <8c82pe$pf0$2@newshost.accu.uu.nl>
Message-ID: <vndd7o4dund.fsf@camelot-new.ccs.neu.edu>

m.faassen at vet.uu.nl (Martijn Faassen) writes:

> > Emacs allows for multiple windows on multiple displays.

> How do you do this? I wanna know!

I'm not sure about FSF Emacs, but under XEmacs the function
`make-frame-on-display' may get you going in the right direction.

-Justin

 



From thomas at xs4all.net  Mon Apr 17 03:49:59 2000
From: thomas at xs4all.net (Thomas Wouters)
Date: Mon, 17 Apr 2000 09:49:59 +0200
Subject: Python w/ SSL
In-Reply-To: <38FA3450.1DD845B2@flashmail.com>; from exarkun@flashmail.com on Sun, Apr 16, 2000 at 09:44:48PM +0000
References: <38FA3450.1DD845B2@flashmail.com>
Message-ID: <20000417094959.F15664@xs4all.nl>

On Sun, Apr 16, 2000 at 09:44:48PM +0000, Jp Calderone wrote:

>  Can anyone point me to an *up to date* Python/SSL interface?  I've
> found about 4 w/ search engines, but the most recent is from 1996 and
> has a link to Python 1.3 in case you are without an interpreter ;)
> (It also doesn't help that starship is still down - again :( )

Python 1.6's socket library has direct support for SSL (if you enable this,
during compilation. It requires OpenSSL) and modules like urllib and httplib
use this to provide SSL connections, too. I'm not sure if there's a seperate
patch for 1.5.2, though, but perhaps you can give 1.6a2 a shot ? ;-)

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From aahz at netcom.com  Tue Apr  4 16:42:45 2000
From: aahz at netcom.com (Aahz Maruch)
Date: 4 Apr 2000 20:42:45 GMT
Subject: Advice requested: GUI project beginning
References: <m12avmV-000wcEC@swing.co.at> <slrn8eiuu1.b3t.warlock@raven.myth> <38E97E29.2687864D@acm.org> <slrn8ej39h.h48.warlock@raven.myth>
Message-ID: <8cdk45$ts9$1@nntp9.atl.mindspring.net>

In article <slrn8ej39h.h48.warlock at raven.myth>,
Jim Richardson <warlock at eskimo.com> wrote:
>On Tue, 04 Apr 2000 05:31:14 GMT, 
> Robert W. Cunningham, in the persona of <rcunning at acm.org>,
> brought forth the following words...:
>>
>>Programming?  Telescope control?  Doesn't this thread belong in a Forth NG?
>>I mean, wasn't Forth CREATED to control a telescope?
>
>hm, I thought that was Algol?

Nope, Forth.
--
                      --- Aahz (Copyright 2000 by aahz at netcom.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

"It's 106 miles to Chicago.  We have a full tank of gas, a half-pack of
cigarettes, it's dark, and we're wearing sunglasses."  "Hit it."


From dalke at acm.org  Sun Apr  9 00:12:33 2000
From: dalke at acm.org (Andrew Dalke)
Date: Sat, 8 Apr 2000 22:12:33 -0600
Subject: Python paradigms
References: <8cmr70$b0e$1@pegasus.csx.cam.ac.uk> <8cntqp$upi$1@slb0.atl.mindspring.net>
Message-ID: <8covpn$4jq$1@slb6.atl.mindspring.net>

I wrote:
>Nick Maclaren wrote
>>    x = (a != NULL ? a[i]->weeble : 0) + (b != NULL ? b[i]->wombat : 0)
>> [...]


>How about
>  x = getattr(a, "weeble", 0) + getattr(b, "wombat", 0)


My code, BTW, is wrong.  I didn't see the "[i]" part of the original
expression.  To my view of things, this means the expression is
too complex.

You asked about other idioms.  You could have a class wrapper
for accessing elements, like:

class SpecialLookup:
  def __init__(self, container, attr, default = 0):
    self.container = container
    self.attr = attr
    self.default = default
  def __getitem__(self, i):
    if self.container is None:
      return self.default
    return getattr(self.container[i], attr)

new_a = SpecialLookup(a, "weeble")
new_b = SpecialLookup(b, "wombat")
x = new_a[i] + new_b[i]

For slightly better performance, you could even create a special __getitem__
for empty containers so the check for a None container
is only done ones.

                    Andrew
                    dalke at acm.org





From claird at starbase.neosoft.com  Sat Apr  1 11:07:16 2000
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 1 Apr 2000 10:07:16 -0600
Subject: was Re: Why should I switch to Python?
References: <seas5ia89177@corp.supernews.com> <1EeF4.43$CA.2422@news> <oqzord6h5i.fsf@titan.progiciels-bpi.ca> <38E618D7.A15F9D9B@webwitches.com>
Message-ID: <05AB7F747D3FC9FA.FF1C6A9D99274EAE.D8581C4B2696A7E4@lp.airnews.net>

In article <38E618D7.A15F9D9B at webwitches.com>,
Shae Erisson  <shae at webwitches.com> wrote:
			.
			.
			.
>maybe a newsgroup bot that finds common text patterns and posts FAQtual
>replies?
			.
			.
			.
Comp.lang.tcl does this, by the way; a bot emits
<URL:http://www.westend.com/~kupries/c.l.t.welcome.html>
to first-time posters.  There is discussion now about
refinements involving multi-posting, posts that mention
"newbie", automated answers to common questions, ...
We're thinking as seriously as possible about how to
send it out *before* they first ask for it ...
-- 

Cameron Laird <claird at NeoSoft.com>
Business:  http://www.Phaseit.net
Personal:  http://starbase.neosoft.com/~claird/home.html


From ng at hardlight.couk.com  Sat Apr 22 09:27:57 2000
From: ng at hardlight.couk.com (.:|:.)
Date: Sat, 22 Apr 2000 14:27:57 +0100
Subject: code not true?
References: <8ds3ku$q0l$1@nnrp1.deja.com>
Message-ID: <3901aa0f.0@news.proweb.co.uk>

interestingly enough on both Win32 and Linux

>>> a = 0.07 * 100

>>> a
7.0

>>> a == 7.0
0

>>> int(a) == 7
1

>>> type(a)
<type 'float'>

>>> type(7.0)
<type 'float'>

>>> (0.7 * 10) == (0.07 * 100)
0

>>>  (0.07 * 10) == (0.007 * 100)
1

>>> (0.07 * 100) == (0.007 * 1000)
0

>>> (0.07 * 1000.0) == (0.007 * 10000.0)
1







From mskott at image.dk  Tue Apr 25 14:36:32 2000
From: mskott at image.dk (Martin Skøtt)
Date: Tue, 25 Apr 2000 18:36:32 GMT
Subject: Working with data returned by PyGreSQL
Message-ID: <itx6ni3w.fsf@jumbo.skott.dk>

Hi
I have finally installed the PyGreSQL module and it works fine (I can
do connects and querys). My problem is how do I get to anything other
than print my query results? What datatype does the query return? I
would prefer if it came out in either a list or a dictionary (for my
current project I only need strings but a dictionary might be needed
later).

I know the DB-API compliant Postgres module is in beta but has any one
tried and to which extent does it work?

-- 
Martin Sk?tt
mskott at image.dk

'In a certain sense, all decent programming languages are equally powerfull'
Dr. Alan Turing


From tismer at tismer.com  Sun Apr  2 08:14:25 2000
From: tismer at tismer.com (Christian Tismer)
Date: Sun, 02 Apr 2000 14:14:25 +0200
Subject: Debugging confusion -- too many stacks!
References: <000a01bf9c6b$3f80c2a0$752d153f@tim>
Message-ID: <38E739A1.45F4B199@tismer.com>


Tim Peters wrote:
> 
> [Jason Stokes]
> > I'm confused about the difference between the C stack, the Frame
> > stack, and the stack *inside* each execution frame.
> 
> The difference between the first two is an artifact of the current
> implementation; indeed, getting rid of the distinction is the major point of
> Christian Tismer's "Stackless" Python variant (where "less" refers to
> getting rid of the C-stack component).

Yes, at first "less" referred to getting rid of the C stack.
Later, I realized that the resulting tree (see below) is
no stack at all, since it can now be modified at any point.

> > From reading the source, I believe that each new code object gets a new
> > frame allocated on the frame stack,
> 
> Each *invocation* of a code object, yes; and, e.g., if a function calls
> itself recursively, there's only one code object but a separate frame for
> each call level.

Meanwhile, functions feel to me a little like classes, and frames
are their instances.

> Note that in CT's Stackless Python, the frames actually form a tree
> (although at any leaf, the path back to the root is unique and can be viewed
> as a stack).

My current view is like chains, paths, with a common root.
These paths are the default sequence when an explicit or
implicit return is executed. "returning" from a function
instance is to move one step towards the common chain root.
"calling" a function is to extend the current chain by appending
a new function instance.
The other important role of these paths are the responsibility
for exceptions. This hierarchy is upside down: The closest
function instance is consulted first about exception handlers.
This has a bit of similarity to inheritance in classes, but just
a little: We have a dynamic chain of instances which propagate
exceptions, but the behavior is only like single inheritance.

> > within which is stored all kinds of useful information about the
> > context the code is executing in -- the global and local environment,
> > a tuple of constants, a tuple of arguments etc.  *Within* each frame
> > is *another* stack, upon which the Python virtual machine loads and
> > manipulates intermediate values.
> 
> This is another artifact of the current implementation, and is best viewed
> as an internal detail of no visible consequence.

This VM stack could also be replaced by a set of registers, since
the maximum stack size is always known at compile time.

> > However, I'm not sure what the terminology is to refer to this
> > third stack.
> 
> "The frame's eval stack" works for me <wink>.
> 
> > Understanding the distinction seems to be vital to using pdb, though.
> 
> I don't use pdb much, but don't see why this would be true.  The frame's
> eval stack is invisible, and the C and frame stacks happen to be intertwined
> one-for-one today because a Python call happens to invoke the interpreter
> recursively (at the C level) today.  Conceptually, there's only one stack
> involved in calls.
> 
> Perhaps you could be specific about what in pdb is confusing you, and
> someone who uses it could straighten it out.

Personally, I prefer to use the Visual Studio debugger. Debugging
is still not easy with any debugger, since you always have a
hard time to se what's in an object. Almost all source code uses
the general object interface, and especially the eval loop does
so. To figure out what type an object actually is and what it
contains, is quite difficult. In Visual Studio, you can set up some
watches with typecasts. That helps for frames (but not their stack),
integers, and seeing the size of lists and tuples. To see the contents
of variable sized objects like strings, you need to add watches like
((char*)v+20) or do proper casts like (PyStringObject*)v and then
open the structure (which takes more time). Inspection of other
dynamic objects is much harder, since there is no built-in support
for them like with strings.
One very useful watch for eval_loop is this: Cast the current frame
in a way that you can see the name of the running Python function.

The following watches are always active when I'm debuging:
(are there similar ways to do this with PDB?)

(char*)(f->f_code->co_name)+20       # current function name
(char*)(f->f_code->co_filename)+20   # current file name
f->f_lineno                          # line number

In order to see which frame is calling you, do the same, but
replace "f" by "f->f_back".

Well, as a last hint, how do I find a certain place in my
Python code?
The trick is this: Use a seldom used opcode and set a breakpoint.
I prefer to use "BINARY_LSHIFT". Put a line like "1 << 42" into
your Python script, right before the place you want to debug.
You will find yourself at the BINARY_LSHIFT breakpoint, just
a few opcodes apart from your place, and it is easy to single
step until you are there.

What I really wished were a way to extend a given debugger
with your own support code, to make PyObjects more visible.

cheers - chrs

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaunstr. 26                  :    *Starship* http://starship.python.net
14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     we're tired of banana software - shipped green, ripens at home



From robin at alldunn.com  Tue Apr 18 16:47:53 2000
From: robin at alldunn.com (Robin Dunn)
Date: Tue, 18 Apr 2000 13:47:53 -0700
Subject: can someone tell me about...
References: <38FCBB16.BF50EB03@stsci.edu>
Message-ID: <BT3L4.64$Lwf1.8585530@news.randori.com>

> Interactively, python will recognize a variable name that begins with
> one underscore (_myVar). The variable naming rules even say that is
> acceptable.
>
> However, when I try to use it in a module, it bombs.
>

When you say "use it in a module" do you mean like this:

    from the_module import *
    print _myVar

If so then that is normal as names begining with an underscore are not
imported with the *, (it's a way to help prevent namespace pollution.)  On
the other hand, this should work:

    import the_module
    print the_module._myVar


--
Robin Dunn
Software Craftsman
robin at AllDunn.com
http://AllDunn.com/robin/
http://AllDunn.com/wxPython/  Check it out!





From m.faassen at vet.uu.nl  Sat Apr 22 10:11:33 2000
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 22 Apr 2000 14:11:33 GMT
Subject: The Simple Economics of Open Source
References: <14591.15630.192472.440375@goon.cnri.reston.va.us> <slrn8fv2rd.1kt.neelk@brick.cswv.com> <3900611F.3B525BDA@cfdrc.com>  <FtDG4z.774@world.std.com> <slrn8g26tu.p7o.neelk@brick.cswv.com> <22042000.2@sanctum.jae.ddns.org>
Message-ID: <8dsbul$n1o$1@newshost.accu.uu.nl>

Juergen A. Erhard <jae at ilk.de> wrote:
>     Neel> Seriously, the rejection of altruism as a possible motive strikes me
>     Neel> as a very reasonable decision. Altruism is generally not sufficient
>     Neel> motive in other aspects of life -- [...]

> Same goes for egoism... see below.

Not as much. See below.

>     Neel> Second, economists assume that everyone is basically self interested,
>     Neel> including themselves -- it's a standard analytic assumption, not
>     Neel> snobbery.

> But it is not just a `basic analytic assumption' in most cases... most
> people in the field, as you yourself have stated above, have
> `internalized its assumptions'.  Now, you could argue that you didn't
> mean it like *that*... but my impression of people in that field of
> science[1] is that the *did* indeed internalize this... which means
> that *everything*, ever single point of data that goes into an
> analysis, into a study, is colored and filtered by this worldview.

But the assumption that people are selfish makes quite a bit of biological
sense. The whole theory of evolution, selfish genes, limited resources,
and so on. Then of course kin selection and even more importantly,
game theory, come and throw lots and lots of mud in the water. Biological
conclusion: our genetic core is selfish, but selfish behavior, especially
in social, intelligent beings like humans, may entail so much cooperation
and altruism that it's indistinguishable from the real thing (if there
*is* any more real form of cooperation than that for selfish reasons --
if you're cooperating out of altruistic reasons, i.e. you're losing by
cooperating, is that cooperation or slavery?)

And economics is a field that thinks about game theory, like theoretical
biology. Theoretical economics and theoretical biology have lots in common,
though theoretical biology does focus more on the creative aspects of the
process more than economics; economics are interested in the process of
itself, optimizing that, while biologists in the end are interested in the
outcome of the process (things like elephants, humans and trees), and
why they happened.
 
[snip]
> No, I think (and this thought just occurred to me, so it's pretty
> fresh) that there's a third thing... it's the answer to the old
> question asked of a mountaineers: "Why did you climb that mountain?".
> "Well... because it was there."  Or, in other words, "because I had
> to".[4]

That's an interesting distinction; I think that maps to what theoretical
biologists are interested in; complexity and life in the world is
driven by natural selection, but that's not the only thing. There has
to be some mathematical structure to the whole process that allows
the complexity to happen. Though perhaps that's an anthropic point
of view.

> Egoism is often equated with the profit motive these days (little
> wonder in our money-based society...[2])

> Now this is a nice explanation for many things... but not for all.
> I'm certain that the greatest thinkers, the greatest musicians,
> artists... the greatest innovators were *not* motivated by that...

But they were hardly being so interested in their field of interest
because they hated it. See my finishing note.

[why did Einstein do it?]

> Well, I think because he *had* *to*.

[snip]

> I think this can be compared to egoism and altruism like this:

>   Egoism is "I get something".

>   Altruism is "`They' get something".

>   Xyzzy (for lack of a name[3]) is "I get *rid* of something".  Or "I
>   get something of my chest".

No, xyzzy is "I am." The creative individual does this because that individual
is *defined* by what he does. "This is ME." "Hello world."

Interesting analysis.

This belongs in the field of psychology at least for a large part, though.

[snip]
> [4] I'm sure someone has found some weird explanation that makes this
> something egoistical...

They have. According to some, this expression of selfhood is the most
egoistic thing one can possibly do. Also the most valuable, as value to
the individual is measured by that individual, in the end. But of course
there is quite a bit of societal value derived from it.

The world is a complicated, messy thing, and any attempt 
at explanation is by necessity partial. So biology, economy, psychology
and so on do need to work with some basic assumptions, otherwise they'll
never get any abstractions off the ground. So let's forgive the
economists in this case. They discounted some important things a bit
easily (the shared commons that improves faster than the partitioned
one). The economics of informational products does differ significantly
from that of physical goods. But they had some interesting things to 
say, and an interesting toolkit they brought for their analysis.

An off topic thread that I can contribute lots of deep sounding stuff to! 
I didn't know enough about mathematics to contribute to that infinity of
primes one. Ah, but philosophy, ah! :)

Regards,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From kaeber at gmd.de  Mon Apr  3 15:24:32 2000
From: kaeber at gmd.de (Joachim Kaeber)
Date: Mon, 03 Apr 2000 21:24:32 +0200
Subject: Creating variables on the fly...
References: <meh9-25D0C5.15160303042000@news.cit.cornell.edu>
Message-ID: <38E8EFF0.D71D0415@gmd.de>

Hi,

what about this:

def add(*args):
   sum=0
   for arg in args:
     sum=sum+arg
   return sum

add(1,2,3,4)
=> 10

add (1,2,3,4,5,6,7)
=> 28


Matthew Hirsch wrote:
> 
> Hi All,
> 
> Let's say I have four variables:
> 
> a=1
> b=2
> c=3
> d=4
> 
> And I have a function that adds these variables together:
> 
> def add(a,b,c,d):
>    return a+b+c+d
> 
> But now let's say I have twenty variables that I want to add together.
> This function will no longer work.  I would have to rewrite it as
> a+b+c+d+e+f+g+...+(20th letter).  Is there a way to dynamically create a
> variable name?  So that my add function can automatically determine how
> many variable names to create and then add the values together.  In
> other words, I'd ideally like something like:
> 
> def add(number_of_variables):
>    return a+b+c+...+(letter corresponding to number_of_variables)
> 
> Thanks for your help,
> 
> Matt

-- 
Joachim Kaeber (kaeber at gmd.de) (http://imk.gmd.de/~kaeber)
GMD - Forschungszentrum Informationstechnik GmbH
    - German National Research Center for Information Technology
Schloss Birlinghoven                    email: kaeber at gmd.de
D-53757 Sankt Augustin 1, Germany       phone: (+49 2241) 14-2546


From zessin at my-deja.com  Fri Apr  7 04:51:20 2000
From: zessin at my-deja.com (Uwe Zessin)
Date: Fri, 07 Apr 2000 08:51:20 GMT
Subject: Odp: Python on Alpha OpenVMS
References: <z4_C4.6214$4N.168038@news.tpnet.pl> <hDlBpKAS7T34Ew1e@gol.com> <_KCD4.2797$a01.54385@news.tpnet.pl> <ZRhvtDAzBR74EwzH@gol.com>
Message-ID: <8ck7i4$20p$1@nnrp1.deja.com>

In article <ZRhvtDAzBR74EwzH at gol.com>,
  Ian Parker <parker at gol.com> wrote:
> In article <_KCD4.2797$a01.54385 at news.tpnet.pl>, Tomek Lisowski
> <Lisowski.Tomasz at sssa.nospam.pl> writes
[...]
> >I am just not sure, for which platform were these modules compiled,
> >and for which version of OpenVMS system

Tomek,
I thought that is clearly described on the page that Ian has
mentioned...

> Tomasz
>
> I'm glad to say that the VMS objects worked perfectly on both Alpha
> and VAX VMS systems.  They're both V7.x - I didn't bother to check
> specific version numbers.

I develope on OpenVMS VAX V6.1 - there are currently no features beyond
that version. Given OpenVMS?s good upward compatibility I am not that
much surprised ;-)

[...]
> I'm really impressed with the OpenVMS port (Uwe, please keep up the
> good work!).

Thanks. That?s encouraging...

> Only one qualification, at the moment I think the objects don't
> support threading, but haven't really investigated yet.

That is correct. I haven?t done much in this area yet, however I have
received very value feedback for the next release (sorry, no date, yet).
I am sure you won?t be disappointed, though...

Please remember that you need to recompile everything for thread support
and you might have to do some changes for OpenVMS V7 and up.

--
Uwe Zessin


Sent via Deja.com http://www.deja.com/
Before you buy.


From zorro at zipzap.ch  Wed Apr 26 07:25:54 2000
From: zorro at zipzap.ch (Boris Borcic)
Date: Wed, 26 Apr 2000 13:25:54 +0200
Subject: The Simple Economics of Open Source
References: <14591.15630.192472.440375@goon.cnri.reston.va.us>
	 <slrn8fv2rd.1kt.neelk@brick.cswv.com> <3900611F.3B525BDA@cfdrc.com>
	 <FtDG4z.774@world.std.com> <uFaBgiAr5HA5EwJt@jessikat.demon.co.uk>
Message-ID: <3906D242.EB68BFED@zipzap.ch>

Robin Becker wrote:
> 
> I'm afraid your desire for a better world reduces your altruism
> quotient.
> Doing it because it helps others is altruism, doing it to help your self
> and or relatives isn't really.

This comment can be made irrefutable, provided you *define* "self
and relatives" as the beneficiaries of ("pseudo") altruism.

Reciprocally, altruism may be defined as the action of paradoxical
definitions of self. Paradoxical - means it conflicts with the
choices that follow from other, more common, definitions of self,
that make the term "selfish" determinate.

Language that postulates a unique applicable definition of self
from the start, is thus inadequate to an enlightened discussion
of altruism, imho.

BB
--
"I think as I think, therefore I am who I am"


From greg at perceval.be  Wed Apr 26 03:32:02 2000
From: greg at perceval.be (Gregoire Welraeds)
Date: Wed, 26 Apr 2000 09:32:02 +0200 (CEST)
Subject: HTML type form?
In-Reply-To: <003301bfaec2$c4529c60$2801000a@iel.ie>
Message-ID: <Pine.BSF.4.10.10004260930570.24044-100000@hercule.perceval.be>

In reply to the message of Shaun sent on Apr 25 (see below) :

hummm, it is absolutly unsave to record Credit card numbers... :)

--
Life is not fair
But the root password helps
--

Gregoire Welraeds
greg at perceval.be
Perceval Development team
-------------------------------------------------------------------------------
Perceval Technologies sa/nv	Tel: +32-2-6409194		
Rue Tenbosch, 9			Fax: +32-2-6403154		
B-1000 Brussels			general information:   info at perceval.net
BELGIUM				technical information: helpdesk at perceval.net
URL: http://www.perceval.be/
-------------------------------------------------------------------------------

On Tue, 25 Apr 2000, Shaun wrote:

> Date: Tue, 25 Apr 2000 15:30:12 +0100
> From: Shaun <shogan at iel.ie>
> To: python <python-list at cwi.nl>
> Subject: HTML type form?
> 
> 
> how would i create a HTML type user interface using python which would take
> in say, -name-address-credit card number- and record it with the intention
> of struct.pack-ing the values entered,
> thanks
> Shaun
> =====================
> Shaun Hogan
> Interactive Enterprise Ltd.
> alt. E-mail : shaun_hogan at yahoo.com
> +353 86 8342529
> 
> 
> 
> 
> -- 
> http://www.python.org/mailman/listinfo/python-list
> 
> 




From scarblac-spamtrap at pino.selwerd.nl  Wed Apr 12 13:52:30 2000
From: scarblac-spamtrap at pino.selwerd.nl (Remco Gerlich)
Date: 12 Apr 2000 17:52:30 GMT
Subject: global variables - how??
References: <8d2bne$qn7$1@news101.telia.com>
Message-ID: <slrn8f9h8a.ufe.scarblac-spamtrap@flits104-37.flits.rug.nl>

Anders Eggers-Krag wrote in comp.lang.python:
> how do I define a propper global variable that works in all functions in all
> modules?

You can't. A global variable can be accessed everywhere in its module, but
not in all modules. You can import the module with the variable and use it
that way in other modules though.


-- 
Remco Gerlich,  scarblac at pino.selwerd.nl
  Murphy's Rules, "Triple time! Triple time!":
   In Avalon Hill's Squad Leader, an infantryman carrying a flamethrower
   can run at 25 mph along a road, providing a sergeant or other leader
   is encouraging him.


From smalleys at gte.net  Wed Apr 19 22:51:50 2000
From: smalleys at gte.net (sue)
Date: Thu, 20 Apr 2000 02:51:50 GMT
Subject: pattern-based string expansion
References: <38F52C61.A5D72F89@gte.net> <vndog79ft1q.fsf@camelot-new.ccs.neu.edu> <8del87$5md$1@slb6.atl.mindspring.net>
Message-ID: <38FE70DB.29A4B1A1@gte.net>

I'm not looking for all possible strings with a wildcard,
just a pseudo-RE-based expansion.  The RE would not support
open-ended arguments (* or +), but would have ?,[] and so
on.
The expansion needs to be ordered.

I've thought about hacking the glob(3) code to work on an
input RE, but I was hoping this already existed somewhere.

I have csh scripts which generate a set of strings, and then
operate on the string sets. (The strings are NOT filenames,
csh doesn't actually care.)  I'd like to move them over to
Python, but until I can get expansions, I'm stuck.

I didn't see an expansion utility in "Understanding REs".
Everything I've seen so far is focused on matching or
reducing strings.

Sue.


From maybe at lucky.com  Mon Apr 10 05:23:08 2000
From: maybe at lucky.com (Soundwave)
Date: Mon, 10 Apr 2000 11:23:08 +0200
Subject: check EOF
Message-ID: <8cs6g4$rje$1@porthos.nl.uu.net>

how can I check EOF in python ???




From ivanlan at callware.com  Mon Apr 24 22:59:54 2000
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Mon, 24 Apr 2000 20:59:54 -0600
Subject: Tkinter
References: <A63AFB20111ED311AC5500A0C96B7BF5015F0F44@orsmsx36.jf.intel.com> <vD5N4.1521$wYl.189935104@newsb.telia.net>
Message-ID: <39050A2A.D71E9EFD@callware.com>

Hi All--

Fredrik Lundh wrote:
> 
> Daley, MarkX <markx.daley at intel.com> wrote:
> > I am trying to use Tkinter for a GUI interface, but I am having a hard
> time
> > finding manuals for it.  Is there documentation comparable to Python's?
> 
> http://www.python.org/topics/tkinter/doc.html
> 

While /F's documentation is invaluable, online and free, there are two
other books which may help clarify the situation.

1)  If you're at a beginning level, then I recommend the final 8
chapters of my _Teach Yourself Python in 24 Hours_.  These are the eight
chapters I wish had been available when I started using Tkinter.

2)  If you're at a more advanced level, where the /F documentation
doesn't cover enough, or not in enough depth, take a look at John
Grayson's _Python and Tkinter Programming_.  John's writing style is
appealing, he covers the topics in admirable depth, and he gives clever
and entertaining examples.

<no-funny-stuff-today,-just-bsp>-ly y'rs,
Ivan
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
http://www.pauahtun.org and
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
Author:  Teach Yourself Python in 24 Hours



From effbot at telia.com  Tue Apr 18 11:53:12 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Tue, 18 Apr 2000 15:53:12 GMT
Subject: Tkinter items question
References: <38fc7a7d@news2.sireco.de>
Message-ID: <Ix%K4.462$wYl.187476480@newsb.telia.net>

Weil <weil at sireconnect.de> wrote:
> def DeleteLastItem(self,event):
>     items = self.find_all()
>     if items:
>         self.delete(items[-1])
>
> But, I was wondering, if there is a more direct way to
> access the last item other than retrieving all items first.
> I could of course keep a list of all created items myself,
> but that seems a bit ugly, especially as my app couldn't
> use this list for anything else.

you don't want to keep track of the items yourself, and
you don't want to use Tk's list of items?  hmm...

okay, how about storing a serial number as a tag.  when
you create a new object, do something like:

    self.serial = self.serial + 1
    tag = "item%d" % self.serial
    self.create_item(..., tags=tag)

and

    tag = "item%d" % self.serial
    self.delete(self.serial)
    serial.serial = serial.serial - 1

you'd probably consume less memory by storing a list of
int(handle)'s, but that's another story.

</F>




From aquarius at kryogenix.albatross.co.uk  Tue Apr 18 20:53:43 2000
From: aquarius at kryogenix.albatross.co.uk (Aquarius)
Date: 19 Apr 2000 01:53:43 +0100
Subject: Telnet client using curses
Message-ID: <8dj02n$61k$1@kryogenix.yi.org>

Having used Python for a few brief scripts (the sort of thing I used to
use Perl for :), I thought I'd try something a bit more ambitious, to
teach myself the language. What I decided on, because I need one, is a
telnet client. The key thing is that I'd like to use curses to give me
one input line into which text it typed, while data returned from the
connection goes into the window above. Something like TinyFugue.

Looking at telnetlib, it uses select to poll stdin. However, I'll be
getting user input using curses input functions, not input() or
raw_input(), so the user input doesn't go into stdin. Is the best way to
handle this:
1. get user input with the curses function and then somehow slip this
input into stdin, so it looks like it came from raw_input() or whatever?
Obviously, raw_input() gets its data into stdin, but it's a built-in
function, so I don't know how it does it :-)
2. write a file object which I feed data *into* from the curses input
function, and which then feeds that data back out when requested to do
so, and poll *that* instead of stdin?
3. A better solution?

If your answer is "3", some kind of suggestion as to the Right Way to do
it would be appreciated :-)

Aquarius

-- 
Well *done*, Aquarius -- a phrase we love typing because it makes us feel
like the boss-character from a 1980s Glen A Larson action series...
  NTK, http://www.ntk.net/index.cgi?back=archive99/now0618.txt&line=292#l



From rwiserREMOVEthis at mindspring.com  Wed Apr 12 18:55:14 2000
From: rwiserREMOVEthis at mindspring.com (Randy Wiser)
Date: Wed, 12 Apr 2000 22:55:14 GMT
Subject: Calling os.waitpid() From Secondary Thread Under Linux 2.2.12-20?
References: <200004120650.QAA22353@nexus.csse.monash.edu.au>
Message-ID: <38f4fc63.3036292948@news>

On Wed, 12 Apr 2000 16:50:52 +1000 (EST), Jonathan Giddy
<jon at dgs.monash.edu.au> wrote:

>Anyway, my (non-)solution was to redesign my code to wait from the same 
>thread as the fork().  If you can't move the waitpid() into the main thread,
>maybe you can move the fork() into the second thread.
>
>Jon.
>

Thanks! I tried your suggestion and it seems to work fine on Linux for
now (should work on Solaris as well).

I heard a rumor that this Linux feature will be changed to be more
like other UNIXs in the future?

- Randy


From gmunsey at adobe.com  Sun Apr 23 22:02:21 2000
From: gmunsey at adobe.com (Grant Munsey)
Date: Sun, 23 Apr 2000 19:02:21 -0700
Subject: Anyone generating SWF content with Python?
Message-ID: <FVNM4.664$nb6.400276@news.pacbell.net>

Is anyone out there generating Shochwave Flash (swf) format files from
Python?




From hniksic at iskon.hr  Wed Apr 12 07:05:55 2000
From: hniksic at iskon.hr (Hrvoje Niksic)
Date: 12 Apr 2000 13:05:55 +0200
Subject: Python/Perl Popularity (Re: A Mountain of Perl...)
References: <200004102111.HAA03380@envy.fulcrum.com.au> <8d0pcq$7ti$1@nnrp1.deja.com>
Message-ID: <dpg0sr37fw.fsf@mraz.iskon.hr>

tom__98 at my-deja.com writes:

>  -- How can I browse the Python documentation without firing
>     up a web browser?

By using Info.  That's what I do, all the time.  Info has the added
nicety of being able to search through the whole documentation for a
keyword.

>     function or class quickly?  Is there any equivalent of
>     "perldoc -f split" or "perldoc Net::POP3"?

For the latter, I use `info python-lib' followed by `g poplib RET'.


From jmassung at magpiesystems.com  Wed Apr 26 16:02:15 2000
From: jmassung at magpiesystems.com (Jeff Massung)
Date: Wed, 26 Apr 2000 14:02:15 -0600
Subject: global variables or inheritence
References: <28p8zy0irgc.fsf@lada.kom.auc.dk>
Message-ID: <eRHN4.1113$3K4.19980@news.uswest.net>

Sorry if I get this wrong... but I thought you'd just import your other
file.

Example:

# File1.py
myVar = 2


# File2.py
import File1
print File1.myVar #should print 2

Jeff

Thomas Rasmussen wrote in message <28p8zy0irgc.fsf at lada.kom.auc.dk>...
>Hi
>
>I have quite a big problem. I have two different files, where I in one
>file want to declare a variable and then in the other file be able to
>access and use this variable (no need to change it), but I just can't
>seem to get it to work... Is there any nice way of doing this easily?
>I have tried the global var but this doesn't seem to work.
>
>Anyone got a good idea?
>
>Thomas
>--
>KOM Network student helper
>
>"To alcohol! The cause of - and solution to - all of life's problems!"
>-- Homer Simpson




From python-list at teleo.net  Tue Apr  4 13:10:05 2000
From: python-list at teleo.net (Patrick Phalen)
Date: Tue, 4 Apr 2000 10:10:05 -0700
Subject: Starving for an Advanced Python Book
In-Reply-To: <urlG4.2557$d13.25939@typhoon.southeast.rr.com>
References: <20000404040941.7766.qmail@web2106.mail.yahoo.com> <urlG4.2557$d13.25939@typhoon.southeast.rr.com>
Message-ID: <00040410183100.05408@quadra.teleo.net>

[Don Tuttle, on Tue, 04 Apr 2000]
:: "lewst"
:: >ORA's "Programming Python" is a natural choice, but I
:: > can't bear buying this outdated book when the 2nd edition is due out
:: > in a few months (I'm dying here Lutz, hurry it up already!).
:: 
:: I went ahead and bought it because Oreilly's cheif editor says, "We're
:: timing the new edition of Programming Python to follow shortly after the
:: release of Python 2.0."(aka Python 3K).  If that turns out to be the case,
:: it won't published until sometime in 2001.  See his complete (and somewhat
:: dated) remarks at  http://www.oreilly.com/frank/python.html

I'd also suggest going ahead with a purchase of Programming Python. If
you're starving, this food isn't exactly rancid. It's an excellent book
still, despite its age. People have criticized its organization, which
is probably due to the fact that it is really more a textbook than a
reference book. IOW, if you start at the beginning and "do" the whole
text, you'll be exposed to nearly the full range of Python
functionality, but it can be frustrating to try to zero in on something
pertinent to your immediate problem domain. Do, however, check out the
errata on the Web site and mark them in the book.

I haven't seen anyone mention The Quick Python Book in this thread.
Also worth having on your bookshelf.



From richard_chamberlain at ntlworld.com  Sat Apr 29 04:20:32 2000
From: richard_chamberlain at ntlworld.com (Richard Chamberlain)
Date: Sat, 29 Apr 2000 09:20:32 +0100
Subject: ODBC / DBI help (via Win32 Extensions)
References: <rNHvjEdhm5Pp-pn2-RytgtQHbLg9d@d185d18f2.rochester.rr.com> <RRtO4.26899$R4.157537@news1.rdc1.nj.home.com> <rNHvjEdhm5Pp-pn2-dzBINcOCBoWi@d185d18f2.rochester.rr.com>
Message-ID: <YTwO4.2136$Px3.51907@news2-win.server.ntlworld.com>

Benjamin,

I've set up a simple table with one field a date field and inserted into it
okay.

On your INSERT statement you had speech marks around the date so:
'#12/15/00#'

mycursor.execute('INSERT INTO Table1 (myDate) VALUES (#12/15/2000#)')

I know that doesn't answer your question about dbiDates but you may not
require that depending on where you're getting your date from.

Richard

You don't actually need those so it should be just #12/15/00#.

My statement was
Benjamin Schollnick <junkster at nospam.rochester.rr.com> wrote in message
news:rNHvjEdhm5Pp-pn2-dzBINcOCBoWi at d185d18f2.rochester.rr.com...
> On Sun, 29 Apr 3900 04:49:53, "S. Hoon Yoon" <cpu at bigfoot.com> wrote:
>
> > I think that's because you should use format #date# on update. Unless, I
did
> > not understand your problems.
> > That's really more of SQL and access convention rather than Python.
> > Pls look at the docs for Access and SQL.
> >
> > UPDATE [BT Dates] SET [BT Dates].[date] = #12/12/2000#;
> >
> > This is what I get if I use Query wizard and you should follow it.
> > Hope it helps,
>
> The real problem is that I can't find any documentation on the DBI
> module...
>
> The pound convention I found out about...
> Unfornately, I don't have my code with me... But...(From memory)
>
> sql_cmd = """
> INSERT INTO logfile (datetime, loglevel,, logdata)
> VALUES ('#12/15/2000#', 1, 'test')
> """
> # I've tried 12/15/00, 09/09/99, 09/09/1999, and a few other
> # date variants....
>
> # And as far as I can tell the ODBC execute command
> # appends a semicolon?  Or at least everything works
> # without having me add one...
>
> If I attempt to execute that command, I get a SQL error from the ODBC
> driver.
>
> If I remove the datetime field and just submit using the loglevel,
> logdata
> it works fine... (But that sort of defeats the purpose of a log
> file....)
>
> The datetime field is setup in Access as a standard "Date/Time"
> field, so I figure I might need to convert the datetime field into a
> dbiDate
> record, but as I mentioned I can't find any instructions on the fields
> &
> methods for the dbiDate field(s).... (I checked the Win 32
> documentation
> without any luck).
>
> - Benjamin




From johann at physics.berkeley.edu  Tue Apr 25 14:23:23 2000
From: johann at physics.berkeley.edu (Johann Hibschman)
Date: 25 Apr 2000 11:23:23 -0700
Subject: new linereading standard?
References: <3904D21B.333DD930@visionart.com> <390510C9.A744D22E@yahoo.com> <R2g+MIAFMUB5EwUl@jessikat.demon.co.uk>
Message-ID: <mt1z3uyr8k.fsf@astron.berkeley.edu>

Robin Becker writes:

> am I wrong for writing

> for line in open('myfile.txt').readlines():
>     print line

Of course not!  Wrongness is such a relative thing that no one,
especially inhuman compilers, clearly cannot judge it.  If it seems
right to you, that is all that matters.

<wink>

But seriously, I don't think there's anything wrong with that.
It depends on what you're doing.  You can get away with a lot of
syntax crunching if what you're doing is basically simple.  That
code fits the basically simple criteria in my brain.

In short, looks fine to me, and I do it a lot, when all I need from a
file is a list of lines in order.


ObComplexity:

To go off on a wild tangent, this reminds me of the whole map/lambda
functional programming fracas.  See, I agree with most people that I
don't really want to see code like

  map(lambda s: string.split(string.strip(s), ',')[0],
      open('data.dat').readlines())

I much prefer

  out = []
  for s in open('data.dat').readlines():
     fields = string.split(string.strip(s), ',')
     out.append(fields[0])

But, if I'm doing this at the interactive prompt, what I end up typing
is most often something like:

>>> lines = open('data.dat').readlines()
>>> lines = map(string.strip, lines)
>>> lines = map(lambda l: string.split(l, ','), lines)
>>> out   = map(lambda fs: fs[0], lines)

So, basically, I find the functional stuff much more useful from
the command line than I do from actual programs.  I think of the
right transform, then I apply it.  Interesting.

Heh.  I'm free-associating here, please bear with me. Anyone else have
this experience?

--Johann

-- 
Johann Hibschman                           johann at physics.berkeley.edu


From pecora at anvil.nrl.navy.mil  Sun Apr 30 10:14:01 2000
From: pecora at anvil.nrl.navy.mil (Louis M. Pecora)
Date: Sun, 30 Apr 2000 14:14:01 GMT
Subject: Mac Python Questions
References: <IRjL4.271$CP3.3183@news.uswest.net>
Message-ID: <300420001014013701%pecora@anvil.nrl.navy.mil>

In article <IRjL4.271$CP3.3183 at news.uswest.net>, Jeff Massung
<jmassung at magpiesystems.com> wrote:

> Okay, I've been looking all over the net and the documentation that came
> with the MacPython distribution, but can't seem to find the answers to these
> questions. Hopefully someone else out there on a Mac can help me out?
> 
> a) There is a file In $(PYTHON):Mac:Contrib:... called PyColorEdit.py. Good
> color editor. I would like to use that instead of PyEdit.py. Is there a way
> to install the color one?

Don't know about that one.
 
> b) No matter what I do, I can't get Tkinter to work :(. Lately I'm led to
> believe that the Mac version of Python doesn't use it, but uses the Mac
> stuff instead. Is this the case? If not, how can I get Tkinter to work?
> Whenever I run something that imports it, it dies on the line with the
> comment "If this dies your computer might not be set up for tkinter". How
> can I set up my machine for it? Thanks :)

Here are two "answers" re: MacPython & Tk I culled from previous
messages.  Maybe they'll help.  Also, you should tune into the
PythonMacSIG (see: http://www.python.org/pipermail/pythonmac-sig/ and
http://www.python.org/mailman/listinfo/pythonmac-sig.  Good people hang
out there who can give lots of answers.

---------------------------------------------------

(1) Tk and setting the right Mac path

Thanks to Ben Coburn and Richard Gordon I was able to get Tk to work on
my mac.  I fixed the problem by placing the following path's in the
Python Preferences:

$(PYTHON):Lib:lib-tk $(PYTHON):Extensions:Imaging:PIL

Is there a reason that these two paths are not automatically set during
the installation.  I was a bit frustrated when I couldn't get through
the example code in Chapter 1 of Learning Python (pg 23 has a Tk
sample).

(2) How to: Tk on Mac

Programming graphical user interfaces (GUI's) with the Tk-toolkit and
MacPython is easy and                        makes fun. But
unfortunately if you try the simplest demo script with a fresh
installed MacPython                        you will get the following
error messages:

Traceback (innermost last):
  File "MacHD:Python
1.5.2c1:Demo:tkinter:matt:animation-w-velocity-ctrl.py", line 1, in ?
from Tkinter import *
File "MacHD:Python 1.5.2c1:Lib:lib-tk:Tkinter.py", line 8, in ?

import _tkinter # If this fails your Python may not be configured for                       
Tk 
ImportError: No module named _tkinter

The reason therefore is a bug in the installer of MacPython1.5.2c1: It
puts the shared libs ?_tkinter.CFM68K.slb? as well as
?_tkinter.ppc.slb? in the folder ?TK? which is in the folder ?Imaging?
in the folder ?Extensions?. You have to put one of them (or both ) in
the folder ?PlugIns? which is in the folder ?Mac?.

And also the fresh standard installation has no path to
?$(PYTHON):Lib:lib-tk?, you have to add that manually with the program
?EditPythonPrefs?.

Unfortunately there is also a bug within the interpreter. Calling
?Tkinter? from the interpreter interactivetely freezes them. You can
only run Tk-based Python scripts by typing them in your editor and
running them by dropping over the interpreter or saving them as Python
files and starting them by doubleclick.

And you never can run Tkinter based scripts from the IDE. But this
behaviour is documented.


From weil at sireconnect.de  Thu Apr 20 11:01:29 2000
From: weil at sireconnect.de (Weil)
Date: Thu, 20 Apr 2000 17:01:29 +0200
Subject: Tkinter items question
References: <38fc7a7d@news2.sireco.de> <Ix%K4.462$wYl.187476480@newsb.telia.net>
Message-ID: <38ff1c2b@news2.sireco.de>

Fredrik Lundh <effbot at telia.com> schrieb in im Newsbeitrag:
Ix%K4.462$wYl.187476480 at newsb.telia.net...
[snip]

> okay, how about storing a serial number as a tag.  when
> you create a new object, do something like:
>
>     self.serial = self.serial + 1
>     tag = "item%d" % self.serial
>     self.create_item(..., tags=tag)
>
> and
>
>     tag = "item%d" % self.serial
>     self.delete(self.serial)

Shouldn't this read: self.delete(tag) ? Or am I getting something wrong?

>     serial.serial = serial.serial - 1
>
> you'd probably consume less memory by storing a list of
> int(handle)'s, but that's another story.

That's right, I think.

> </F>

Thanks a lot for your help.
At least now I know it wasn't my fault I didn't find any
'last'-tag, automatically handled by Tk as for example 'current'.






From parkw at better.net  Tue Apr  4 12:37:12 2000
From: parkw at better.net (William Park)
Date: Tue, 4 Apr 2000 12:37:12 -0400
Subject: Clear screen?
In-Reply-To: <38eb6a36@207.71.36.3>; from artswan@ktc.com on Wed, Apr 05, 2000 at 11:33:03AM -0500
References: <38eb6a36@207.71.36.3>
Message-ID: <20000404123712.A934@better.net>

On Wed, Apr 05, 2000 at 11:33:03AM -0500, Art Swan wrote:
> In Perl I can clear the screen using system("cls"). What is the
> equivalent in Python?

os.system("cls")	:-)



From bsb at winnegan.de  Sun Apr  9 02:16:40 2000
From: bsb at winnegan.de (Siggy Brentrup)
Date: 09 Apr 2000 08:16:40 +0200
Subject: tabular data
In-Reply-To: poupaerte@my-deja.com's message of "Sat, 08 Apr 2000 20:26:20 GMT"
References: <8co4l2$5r2$1@nnrp1.deja.com>
Message-ID: <873dovhk8n.fsf@orge.winnegan.de>

poupaerte at my-deja.com writes:

> Hi
> 

[...]

> For example, say that I have the following tabular data: [(2,1,"REC02"),
> (1,3,"REC03"),(3,1,"REC01")], and I want to obtain: [(1,3,"REC03"),
> (2,1,"REC02"),(3,1,"REC01")], that is, the tabular data should be
> sorted on the first and then the second column.

> How can I have Python perform an in-place sort on a combination of
> several columns?

Simply use the list's sort() method, it does a lex sort:

Python 1.5.2 (#0, Apr  3 2000, 14:46:48)  [GCC 2.95.2 20000313 (Debian GNU/Linux)] on linux2
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> l = [(2,1,"REC02"),(1,3,"REC03"),(3,1,"REC01")]
>>> l.sort()
>>> l
[(1, 3, 'REC03'), (2, 1, 'REC02'), (3, 1, 'REC01')]
>>> l.append((1,2,'REC4'))
>>> l.sort()
>>> l
[(1, 2, 'REC4'), (1, 3, 'REC03'), (2, 1, 'REC02'), (3, 1, 'REC01')]
>>> # You may even specify a comparision function
>>> l.sort(lambda a,b: cmp((a[1],a[0]),(b[1],b[0])))
>>> l
[(2, 1, 'REC02'), (3, 1, 'REC01'), (1, 2, 'REC4'), (1, 3, 'REC03')]

> By the way, should I represent rows by tuples or by embedded lists?

See the thread "Tuples -- who needs 'en"
[Gordon McMillan]
| Lists are collections. The competing choice to a list is a
| dictionary. Tuples are toy structs; the competing choice is a class.

[...]

Regards
  Siggy

-- 
Siggy Brentrup - bsb at winnegan.de - http://www.winnegan.de/
                  bsb at north.de - http://www.north.de/~bsb/
****** ceterum censeo javascriptum esse restrictam *******



From cut_me_out at hotmail.com  Tue Apr 18 16:19:09 2000
From: cut_me_out at hotmail.com (Alex)
Date: 18 Apr 2000 16:19:09 -0400
Subject: Split
References: <38FCC20D.624AB2F6@searchy.net>
Message-ID: <etdsnwjupqa.fsf@w20-575-109.mit.edu>

> split = regsub.split(line,'\s+')
> 
> how can I print only the 5th for example?

print split[4] will probably do what you want.

Alex.


From embed at geocities.com  Thu Apr 27 15:57:55 2000
From: embed at geocities.com (Warren Postma)
Date: Thu, 27 Apr 2000 15:57:55 -0400
Subject: Converting = to ==
Message-ID: <n01O4.12725$HG1.340768@nnrp1.uunet.ca>

I have an "end user query tool" that uses Python expressions, but I want
comparisons to be able to be written either as A=B or A==B then I want to
compare them using Python's Eval function.

Problem is if you search and replace all = with == then users have to use
the single-equals convention, and some of them will be python programmers,
and will want to use the "non idiot mode" of the query tool.

So i did this, but is there a better way?

    >>> string.replace( string.replace("a=b a==b", "=", "=="), "====","==")
    a==b a==b

Warren







From arcanedragonNOSPAM at NOSPAMhome.com  Thu Apr  6 08:15:54 2000
From: arcanedragonNOSPAM at NOSPAMhome.com (Chris Herborth)
Date: Thu, 06 Apr 2000 12:15:54 GMT
Subject: [OT] Dr. Fun vs Guido
Message-ID: <_d%G4.17632$9g4.526936@news5.giganews.com>

Just in case nobody's pointed this out already, today's Dr. Fun cartoon
(http://metalab.unc.edu/Dave/Dr-Fun/df200004/df20000406.jpg) is Python-
related. :-)

-- 
----------========================================================----------
Chris Herborth, DNRC Holder of Past Knowledge     Arcane Dragon -==(UDIC)==-
BeOS hacker, Catharon Productions, Inc.             http://www.catharon.com/
CTO, Next Generation Entertainment                     http://www.ngent.com/


From andrew at andrewcooke.free-online.co.uk  Thu Apr 27 06:02:36 2000
From: andrew at andrewcooke.free-online.co.uk (Andrew Cooke)
Date: Thu, 27 Apr 2000 10:02:36 GMT
Subject: Metaclasses?
References: <3905F4A0.2B39C705@darmstadt.gmd.de> <m3em7tg8ei.fsf@atrus.jesus.cam.ac.uk> <3906542D.62B94F17@darmstadt.gmd.de>
Message-ID: <8e937n$f4p$1@nnrp1.deja.com>

In article <3906542D.62B94F17 at darmstadt.gmd.de>,
  Robb Shecter <shecter at darmstadt.gmd.de> wrote:
> Michael Hudson wrote:
[...]
> Choice A:  Have a smalltalk-like instance variable declaration.
[...]
> to say that I like "Choice A".  I come from Java, and it did bug me a
> bit that there was no one easy place to see what instance variables a
> class has.  It made understanding the XML parser source code harder,
for
> example.

this doesn't really save typing - instead of typing self you have to
type the local declaration.  worse, this introduces extra syntax (after
all, you still need ".").

but more importantly, having "self" is good for the soul.  in most other
languages objects function in some "magical" manner.  having self makes
the mechanism clear - which is helpful both to newbies understanding OOP
for the first time, and to fools like me that get easily confused when
things happen behind their back...

andrew

ps it doesn't have to be "self", afaik - you could use just "s" to save
key presses.... :-)


Sent via Deja.com http://www.deja.com/
Before you buy.


From grant at nowhere.  Fri Apr  7 14:43:24 2000
From: grant at nowhere. (Grant Edwards)
Date: Fri, 07 Apr 2000 18:43:24 GMT
Subject: More about variables
References: <meh9-5C17A0.10445406042000@news.cit.cornell.edu> <3X3H4.16435$0o4.108310@iad-read.news.verio.net> <meh9-158D1A.14240006042000@news.cit.cornell.edu> <38ED017E.1E374737@bellatlantic.net>
Message-ID: <slrn8esb2b.jj1.grant@grante.comtrol.com>

In article <38ED017E.1E374737 at bellatlantic.net>, Steve Holden wrote:
>
>> Why create named variables?
>
>> I always thought that having 20 lists saved more space than having one
>> list of 20 lists.  Maybe I'm wrong.
>
>First make it work.  *Then* make it work better!

That's an _extremely_ valuable thing to remember.

And it seems to be something that many CSci graduates escape
from college without learning.  _Some_ of the engineering
graduates have learned it, but not many.

In practice you'll find that

1) Half of the time you'll discover what you thought you wanted
   wasn't really what you needed, so having something that does
   it smaller or faster is equally useless as something that
   does it bigger and slower.  The sooner you get to that point
   the better off you are.

2) The other half of the time, you'll discover than something
   slightly bigger or slower is good enough, and time spent
   making it work better is best spend on something else.

Taking these observations into account, a second good rule to
remember is:

   First make it work.  Then make something else work.

;)

-- 
Grant Edwards                   grante             Yow!  I'm into SOFTWARE!
                                  at               
                               visi.com            


From artymiak at home.linuxapprentice.com  Sun Apr  9 04:38:27 2000
From: artymiak at home.linuxapprentice.com (Jacek Artymiak)
Date: Sun, 09 Apr 2000 08:38:27 GMT
Subject: LinuxJournal Python Article
Message-ID: <Pine.LNX.4.20.0004090954320.508-100000@home.linuxapprentice.com>

Hi All,

This is just to let you know that the links to code samples in my
Python article published by LinuxJournal are working now.

http://www2.linuxjournal.com/lj-issues/issue73/3946.html

Enjoy!

Jacek Artymiak

-------------------------------------------------------------------
 See how many people have registered at the Python Users Counter!
http://www.wszechnica.safenet.pl/cgi-bin/checkpythonuserscounter.py
-------------------------------------------------------------------
    Autor/Dziennikarz/Konsultant - Author/Journalist/Consultant
       artymiak at safenet.pl, http://www.wszechnica.safenet.pl
co-author:  StarOffice for Linux Bible (IDG Books Worldwide, Inc.)
http://www.amazon.com/exec/obidos/ASIN/0764533630/polskawszechnica
-------------------------------------------------------------------



From bruce_dodson.nospam at bigfoot.com  Sun Apr  9 18:15:37 2000
From: bruce_dodson.nospam at bigfoot.com (Bruce Dodson)
Date: Sun, 9 Apr 2000 19:15:37 -0300
Subject: Python Language Support for ArcView GIS
References: <q0MH4.2398$Bq2.38636@newscontent-01.sprint.ca> <38EFAC08.2A10932D@verecomm.com>
Message-ID: <bl7I4.3764$Bq2.59574@newscontent-01.sprint.ca>

A good thought, but as Reini pointed out, it doesn't yet qualify to go
there.  I hope the current license allows enough freedom that you won't feel
too constrained in what you can do with it.

Peter Kropf wrote in message <38EFAC08.2A10932D at verecomm.com>...
>Just an FYI: You might want to consider moving AVPython to
>sourceforge.net. The entire environment is designed for
>various levels of open source development...




From smalleys at gte.net  Mon Apr 24 19:30:44 2000
From: smalleys at gte.net (sue)
Date: Mon, 24 Apr 2000 23:30:44 GMT
Subject: partial parsing?
References: <8dvilp$88s$1@slb6.atl.mindspring.net> <8e10n7$rkl$1@slb3.atl.mindspring.net> <8e1ng0$4sao$1@uvaix7e1.comp.UVic.CA>
Message-ID: <3904D93B.E47BBC89@gte.net>

I think I remember a followup article in DDJ about a year
later in their annual "Parsing" issue.

Sue.


From mwh21 at cam.ac.uk  Wed Apr 26 09:45:16 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 26 Apr 2000 14:45:16 +0100
Subject: Metaclasses?
References: <3905F4A0.2B39C705@darmstadt.gmd.de> <slrn8gc6qj.p7e.quinn@hedono.ugcs.caltech.edu> <200004261201.IAA05678@zippy.init.com>
Message-ID: <m3k8hlou1f.fsf@atrus.jesus.cam.ac.uk>

Donald Beaudry <donb at init.com> writes:

> quinn at hedono.ugcs.caltech.edu (Quinn Dunkan) wrote,
> > Anyway, even if you could figure out the ambiguity problem, I don't
> > think metaclasses would help you, bytecodehacks might.
> 
> *If* bytecodehacks could help, metaclasses could be used to make
> calling them neat and clean.  The idea is that during class
> instantiation the metaclass would, using bytecodehacks, rewrite all of
> the class's methods.  This might confuse the poor guy who comes along
> and tries to derive a new class from one of these 'self less'.

What, like this? :-)

This has recieved light testing only, so be suitably careful.

deadline-on-Tuesday?-no-I-must-have-imagined-it-ly y'rs
Michael

from bytecodehacks.code_editor import Function
from bytecodehacks.ops import *
import types

def rewrite_method(meth,attrs):
    f = Function(meth)
    code = f.func_code
    cs = code.co_code

    code.co_varnames.insert(0,"self")
    code.co_names.insert(0,"self")
    
    i = 0
    while i < len(cs):
        op = cs[i]
        if op.__class__ in [LOAD_FAST,STORE_FAST,DELETE_FAST]:
            op.arg = op.arg + 1
        if  (op.has_name_or_local() and op.name in attrs
             and op.__class__ not in [LOAD_ATTR,STORE_ATTR]):
            if op.arg < code.co_argcount:
                raise ValueError, "parameter also instance member!"
            if op.__class__ in [LOAD_FAST, LOAD_NAME, LOAD_GLOBAL]:
                attrop = LOAD_ATTR
            else:
                attrop = STORE_ATTR
            cs[i:i+1] = [LOAD_FAST(0),attrop(op.name)]
            i = i + 2
        else:
            i = i + 1
    code.co_argcount = code.co_argcount + 1
    return f.make_function()

class SelflessClass:
    def __init__(self,name,bases,ns):
        self.__name__ = name
        self.__bases__ = bases
        self.__namespace__ = ns
        if ns.has_key('__attrs__'):
            attrs = ns['__attrs__']
            for k,v in ns.items():
                if type(v) is types.FunctionType:
                    ns[k] = rewrite_method(v,attrs)
    def __call__(self):
        return SelflessInstance(self)

class MetaMethodWrapper:
    def __init__(self, func, inst):
        self.func = func
        self.inst = inst
        self.__name__ = self.func.__name__

    def __call__(self, *args, **kw):
        return apply(self.func, (self.inst,) + args, kw)

class SelflessInstance:
    def __init__(self,klass):
        self.__klass__ = klass
        try:
            init = self.__klass__.__namespace__["__init__"]
        except:
            pass
        else:
            init(self)
    def __getattr__(self, name):
        try:
            value = self.__klass__.__namespace__[name]
            if type(value) is types.FunctionType:
                return MetaMethodWrapper(value,self)
            else:
                return value
        except KeyError:
            raise AttributeError, name

Selfless = SelflessClass("Selfless",(),{})

class MyClass(Selfless):
    __attrs__ = ["a"]
    def __init__():
        a = 1
    def get_a():
        return a
    def set_a(newa):
        a = newa

-- 
  If i don't understand lisp,  it would be wise to not bray about
  how lisp is stupid or otherwise criticize, because my stupidity 
  would be archived and open for all in the know to see.
                                               -- Xah, comp.lang.lisp


From breiter at usf.Uni-Osnabrueck.DE  Thu Apr 20 09:30:59 2000
From: breiter at usf.Uni-Osnabrueck.DE (Bernhard Reiter)
Date: 20 Apr 2000 13:30:59 GMT
Subject: Mac Python Questions
References: <IRjL4.271$CP3.3183@news.uswest.net>
    <38FE8032.DE233C29@cosc.canterbury.ac.nz>
Message-ID: <8dn0qj$8gt$1@newsserver.rrzn.uni-hannover.de>

In article <38FE8032.DE233C29 at cosc.canterbury.ac.nz>,
	Greg Ewing <greg at cosc.canterbury.ac.nz> writes:
> Jeff Massung wrote:
>> 
>> b) No matter what I do, I can't get Tkinter to work :(.
> 
> Neither can I, and from what I've seen here, neither
> can anyone else. Tkinter just seems to be broken on
> the Mac at present.

Well it works for me here. Little tests and the demo run
fine with the distribution I got for the Mac. (It tell you python
1.5.2c, but it is supposed to be a 1.5.2)

Here is some exerpt of my correspondence with Jack Jansen:

| > Tk was not installed correclty, import Tkinter after import _tkinter
| > failed. I had to add Lib:lib-tk to the Python Configuration.
| 
| This is a known bug in this distribution. And unfortunately this
| distribution has suffered quite some delays, so that by the time this 
| problem surfaced it had become pretty much impossible to do a new one,
| as my sources had progressed quite a bit beyond 1.5.2. The fix is to
| either move _tkinter from Extensions:img to Plugins, or to add the
| folder where _tkinter lives to sys.path with EditPythonPrefs.


Bernhard

-- 
Professional Service around Free Software                (intevation.net)  
The FreeGIS Project				            (freegis.org)
Association for a Free Informational Infrastructure            (ffii.org)


From michael.stroeder at inka.de  Thu Apr 20 07:30:12 2000
From: michael.stroeder at inka.de (Michael =?iso-8859-1?Q?Str=F6der?=)
Date: Thu, 20 Apr 2000 13:30:12 +0200
Subject: socket.SO_REUSEADDR
Message-ID: <38FEEA44.6C05E2B0@inka.de>

HI!

For avoiding problems with errors like

  socket.error: (98, 'Address already in use.')

during immediate restart of my web server I'm following the hint to
set the socket option SO_REUSEADDR. But I did not succeed. The
server starts and works. When I stop it and try to restart
immediately the above message appears again. I had a short look at
man 7 socket and man setsockopt but this was a little bit too much
for me...

Ciao, Michael.

My code looks like this:

------------------------ bite here -------------------------
class MyHTTPServer(SocketServer.TCPServer):

  def server_bind(self):
    """Override server_bind to set socket options."""
    self.socket.setsockopt(
      socket.getprotobyname('tcp'),
      socket.SO_REUSEADDR,1
    )
    SocketServer.TCPServer.server_bind(self)
------------------------------------------------------------


From danzap_python at bol.com.br  Sat Apr  8 11:12:15 2000
From: danzap_python at bol.com.br (danzap_python)
Date: Sat,  8 Apr 2000 12:12:15 -0300
Subject: Getting file version information 
Message-ID: <FSPE8F$IjZKhrIOjoVVjGCvzttMZuqS9S8MZjgaBc5tgLYmP@bol.com.br>

I'd like to know how to get the file version information
in Windows.

I've looked up at the win32api module, but there isn't a 
GetFileVersionInfo function.

So how to access it ???

Thanks.

Daniel


______________________________________________
EconoMiner
Tudo sobre Economia? O Miner acha para voc?.
http://miner.bol.com.br/econominer.html 
 




From markmcmahon at my-deja.com  Mon Apr 10 12:28:50 2000
From: markmcmahon at my-deja.com (Mark)
Date: Mon, 10 Apr 2000 16:28:50 GMT
Subject: Redirecting stdin/stdout on Windows with '>'and '<'
Message-ID: <8csvff$7pn$1@nnrp1.deja.com>

Hi,

I seem to have a problem and can't find it in the archives or FAQ.

I would like to redirect output of my script with '>' or '<'

If I try to redirect out then the file is created but has no contents.

If I try to redirect in eg...
"test.py < autoexec.bat"

I get the following traceback

Traceback (most recent call last):
  File "C:\test.py", line 4, in ?
    s = sys.stdin.read()
IOError: [Errno 9] Bad file descriptor

Does anybody have an idea!!!

(Windows nt 4 sp 6, python 1.52)

Thanks
  Mark


Sent via Deja.com http://www.deja.com/
Before you buy.


From BobAlex at uppercase.xerox.com  Wed Apr  5 15:22:28 2000
From: BobAlex at uppercase.xerox.com (Alexander, Bob)
Date: Wed, 5 Apr 2000 12:22:28 -0700 
Subject: Tuples -- who needs 'em
Message-ID: <AF223B12F1FDD2119253006097C4FCEC0AD357@cats.pdri.xerox.com>


m.faassen at vet.uu.nl [mailto:m.faassen at vet.uu.nl] wrote:
> 
> Bob Alexander <bobalex at home.com> wrote:
> > I agree with this, too (noting Lisp as possibly the worst example --
> > although a fine language with clear favorable influences on 
> Python). But
> > my issue is not about syntax.
> 
> > However, too bad about that nasty syntax irregularity that 
> using parens for
> > tuples presents  (1,)  (Couldn't resist  ;-)

BTW, a couple of folks pointer out my error in associating the parentheses
with tuple syntax. The parens aren't really part of the syntax, although
they are often used for grouping and visual distinction. However the
dangling comma, as we well know, is always need for one-element tuples.

> But tuples are nice in particular for their *syntax*. With tuples, you
> can do:
> 
> def foo():
>     return 1, 2, 3, 4
> 
> first, second, third, fourth = foo()
> 
> I think this is nice syntax. Without tuples it'd not be as 
> nice. It's not
> so much about efficiency. This pattern occurs all over Python.

I like that syntax too. If tuples and lists were unified, that bracket-free
syntax could also be used for the list/tuple datatype, and nice syntax you
mention would not change.

> [snip]
> > Yet the concept of array is pretty fundamental to 
> programming. Arrays are
> > usually thought of as mutable. Anyone having trouble with 
> that concept is
> > pre-beginner  :-)
> 
> Anyone having trouble with the concept of immutable lists 
> must have never
> heard of functional programming?
> 
> [snip]
> > I'm sure tuples are not an acciental feature. And my issue 
> comes from
> > understanding, not lack of understanding.
> 
> Given the above two items, I don't think you understand tuples yet. I
> consider tuples to be mostly nice syntax (with one nit, the 
> single element
> tuple). You could almost see the immutability semantics being there to
> support the syntax, not the other way around. It might be 
> possible to preserve
> the nice syntax and still get rid of tuples, if you still allow this:
> 
> foo, bar, baz = [1, 2, 3]
> 
> But this is less symmetrical, and since lists are mutable, 
> this may go wrong
> far more easily.

I am wondering where the serious concern about mutable lists comes from, as
it is a pattern in the responses I've gotten on this topic. For me, the
immutable list is a relatively new concept, since I can't think of many
other languages that have it. I'm a programming-language-design kind of
person, and I never really saw mutable lists as a problem that needed
solving. I've used many Python-type languages that have mutable lists but
not immutable ones, and if I don't want the elements to change they don't
change -- I simply avoid the x[i] = y construct :-) And I still haven't
heard from anyone about why there isn't similar concern about immutable
dictionaries.

Anyway, it seems to me that most of the tuple advantages we enjoy could and
should still exist with a merged tuple-list type, and the only possible
drawback is the loss of an immutable sequence type. The gain, though, is one
less decision that we must frequently make (between tuple and list) while
programming.

-- Bob



From thomas at cintra.no  Fri Apr  7 06:39:31 2000
From: thomas at cintra.no (Thomas Weholt)
Date: Fri, 7 Apr 2000 12:39:31 +0200
Subject: MySQL-module
Message-ID: <55jH4.145$rD3.2312@news1.online.no>

Hi,

I`m going to try MySQL and want to access stuff using Python. At Python.org
I found to links to two different modules, one written in python, the other
C. Is there a special reason for choosing one over the other?

I`d like to use all python, not compile anything cuz I`m scared python won`t
work anymore. Had this problem earlier with re-compilation of python. zlib
etc. didn`t work anymore.

Any tips? I can create the actual databases/tables manually, but need to
acces/insert/manipulate the data from a python-script.

Thomas




From grant at nowhere.  Fri Apr 21 13:23:09 2000
From: grant at nowhere. (Grant Edwards)
Date: Fri, 21 Apr 2000 17:23:09 GMT
Subject: How2 use curses for data input
References: <38FF6DE2.5751423A@sprint.ca> <slrn8fusjd.99a.grant@grante.comtrol.com> <38FF8F9C.9D533E74@sprint.ca>
Message-ID: <slrn8g13jq.e00.grant@grante.comtrol.com>

In article <38FF8F9C.9D533E74 at sprint.ca>, Miles Thompson wrote:

>I located it on the RH system here in the office at home. The
>target machine is Debian, and I think I can install RPM on it
>so that if need be, I can get the RPM. Debian doesn't seem to
>have a snack package.

It's something that RH did for use in their installer and
system admin tools, so I don't think it's part of the standard
Python dist.  Which is a shame.

>Thanks for the lead, much appreciated. I've searched google,
>www.python.org, and www.debian.org for information on how to
>use it. 

Didn't find much, eh?  (Been there.)

>All that's on my system is snack.py, .pyc and _snackmodule.so.
>I *hoped* there might be some clues in __builtin__ or in
>snakc.py, but I drew a blank.

Here's a little demo program:

======================================================================
import snack
import time

screen = snack.SnackScreen()
screen.drawRootText(1,1,"background text")
screen.refresh()
time.sleep(1)
x = snack.ButtonChoiceWindow(screen,
                         "test window", 
                         "This is a test of the button choice window")
y = snack.EntryWindow(screen, 
                "EntryWindow Title",
                "This is some text",
                ["prompt1", "prompt2"])
z = snack.ListboxChoiceWindow(screen,
                              "ListboxChoiceWindow Title",
                              "This is a bunch of text for thelist box choice window",
                              ['item1','item2','item3'])
time.sleep(1)
screen.finish()
print x
print y
print z
======================================================================

The above program uses some pre-defined window layouts.  The
newt source package contains some other more complex examples
showing how to layout your own windows using a grid placement
manager.  The nested checkbox-tree widget is pretty cool.

I'll put the newt source package in my ftp directory:

   <ftp://ftp.visi.com/users/grante/stuff/newt-0.50.tar.gz>

Newt/snack is really a handy little package. Too bad it's so
little known.  The tutorial contained in the source package I
have is a bit out of date and doesn't explain grid widgets, so
you've got to figure them out based on the source code and the
example programs.  I should look around to see if there's a
newer version of newt available...

-- 
Grant Edwards                   grante             Yow!  I OWN six pink
                                  at               HIPPOS!!
                               visi.com            


From quinn at euro.ugcs.caltech.edu  Wed Apr 19 13:56:03 2000
From: quinn at euro.ugcs.caltech.edu (Quinn Dunkan)
Date: 19 Apr 2000 17:56:03 GMT
Subject: modifying os.environ
References: <p6qhfczo50y.fsf@informatik.hu-berlin.de> <slrn8f9mn1.ven.quinn@zloty.ugcs.caltech.edu> <8di2il$10oq$1@nntp6.u.washington.edu>
Message-ID: <slrn8frsu1.3bu.quinn@euro.ugcs.caltech.edu>

On 18 Apr 2000 16:30:13 GMT, Donn Cave <donn at u.washington.edu> wrote:
>Quoth Martin von Loewis <loewis at informatik.hu-berlin.de>:
>| quinn at zloty.ugcs.caltech.edu (Quinn Dunkan) writes:
>...
>| > Or perhaps there's a better way to do the (common) thing I'm trying
>| > to do: clean the environment to insure some minimum sane values.  I
>| > can't make a copy of the environ and then pass to execve() since
>| > that doesn't help much for popen and system.
>|
>| You could implement popen and system yourself, starting from popen2
>| (for example).
>
>That may sound like the re-invention of the wheel, but there are
>other improvements that can be made at the same time:
>
> - eliminate system()'s sh -c 'shell command', which can be a
>   security liability in some situations.  The shell can do
>   some really useful things, but if you don't need those things
>   you should be able to pass in an argv sequence and skip sh.

I do need the shell's useful things, and the system() commands are constants,
so things should be pretty bulletproof as long as the environment is also
constant, right?  But it *would* be useful to have a spawn() or run() or
something builtin which does a proper vfork/exec with no subshell involved.  I
often find myself writing one of those.

> - Raise exception on exit status, complete with error output as
>   exception value.  A bit of a pain to implement, and again not
>   not always the right thing to do, but usually it is the right
>   thing.

I don't understand... are you saying the close() method should throw the
exception?  What's wrong with just checking it's return value normally?  How
is a child you started exiting exceptional?  And if the child has been started
asynchronously, do you mean throw an exception from the SIGCHLD handler, or
waitpid()?

I wound up just writing unsetenv(), and then setenviron() which sets the
environ to a given dict.


From trossignol at my-deja.com  Tue Apr 25 15:50:34 2000
From: trossignol at my-deja.com (trossignol at my-deja.com)
Date: Tue, 25 Apr 2000 19:50:34 GMT
Subject: threads on Linux freezing?
References: <20000413010344.A13931@owl.rhein-zeitung.de> <38F658C6.66AD9CA5@clark.net> <20000414154718.A15664@xs4all.nl> <20000414190308.A30502@owl.rhein-zeitung.de> <38F91D1B.3D663B0C@stanford.edu> <38FD72B1.DBB2F731@stanford.edu> <slrn8ft95u.ubt.nascheme@cranky.arctrix.com> <slrn8ftbco.1t1.nascheme@cranky.arctrix.com>
Message-ID: <8e4su3$rmv$1@nnrp1.deja.com>

I've duplicated your results on a couple of dual
PIII servers and a workstation all running glibc
2.1.2 or 2.1.3.  I was able to run the python test
script on a machine running glibc 2.0.7 and the
problem did not show up.  This machine just
continued to print stars.

I'm running Zope (http://zope.org), which is
written in Python, and on occation I've been
getting mysterious lockups.  This only has been
occurring on the machines with the newer version
of glibc.  Could this be related?

Does anyone know how to change the scheduling
policy for python?  In the FAQ previously quoted
it was indicated that SCHED_RR or SCHED_FIFO might
correct this condition.


Sent via Deja.com http://www.deja.com/
Before you buy.


From JoeSmith at bogusaddress.com  Sat Apr  1 22:41:55 2000
From: JoeSmith at bogusaddress.com (Joe Smith)
Date: Sun, 02 Apr 2000 03:41:55 GMT
Subject: New Features in Python 1.6
References: <200004011740.MAA04675@eric.cnri.reston.va.us> <_pyF4.25562$E85.640966@news1.rdc1.md.home.com> <slrn8edfmk.185.wtanksle@dolphin.openprojects.net>
Message-ID: <38E6C25C.8D45A388@bogusaddress.com>

I think that there are security issues with the auto download.  I feel
comfortable being able to download one big thing and being able to examine it
source code wise and with anti-virus programs and so forth.  I get kind of
worried about somebody some how injecting malicous code in transit or fooling my
machine into downloading something from a hacker's box because they have done
something to a DNS server or something.  There should be a way to disable this
feature on a per machine basis.

William Tanksley wrote:

> On Sun, 02 Apr 2000 02:39:54 GMT, Eric Hagemann wrote:
> >What are the options for users wanting 1.6 that may not have access to
> >modules.python.org
> >on a regular basis or at all.  I'd hate to be disconnected from the net and
> >find I need that
> >'one' function that used to be in the standard library ......... e.g. can I
> >down load the
> >entire library and forgo the net check ?  What happens if there is no net --
> >will the import function
> >get unhappy ?
>
> It's actually not a problem -- all modern OSes support TCP/IP even when
> you're disconnected from the net, and one particular address, 127.0.0.1,
> is always live.  This happens to be the site from which Python loads its
> modules (we have an agreement with the owner of that site, which also
> hosts many other products of the same nature, including some marketing
> apps).
>
> >Also when you import like this does the code check the net for a newer
> >version or only when you have no version of the particular sub-module ?
>
> Only when you have no version.  That's not a bad idea, though.  I suspect
> it would be best as a single, user-initiated command, though.
>
> >"Guido van Rossum" <guido at python.org> wrote in message
>
> >> 2. Imputils
>
> >> Complementary to the Distutils are the Imputils, or Import Utilities.
> >> Python's import mechanism has been reworked to make it easy for Python
> >> programmers to put "hooks" into the code that finds and loads modules.
> >> The default import mechanism now includes hooks, written in Python, to
> >> load modules via HTTP from a known URL.
> >>
> >> This has allowed us to drop most of the standard library from the
> >> distribution.  Now, for example, when you import a less-commonly-needed
> >> module from the standard library, Python fetches the code for you.  For
> >> example, if you say
> >>
> >>     import tokenize
> >>
> >> then Python -- via the Imputils -- will fetch
> >> http://modules.python.org/lib/tokenize.py for you and install it on your
> >> system for future use.  (This is why the Python interpreter is now
> >> installed as a setuid binary under Unix -- if you turn off this bit, you
> >> will be unable to load modules from the standard library!)
> >>
> >> If you try to import a module that's not part of the standard library,
> >> then the Imputils will find out -- again from modules.python.org --
> >> where it can find this module.  It then downloads the entire relevant
> >> module distribution, and uses the Distutils to build and install it on
> >> your system.  It then loads the module you requested.  Simplicity
> >> itself!
>
> --
> -William "Billy" Tanksley



From wware at world.std.com  Tue Apr 11 10:00:06 2000
From: wware at world.std.com (Will Ware)
Date: Tue, 11 Apr 2000 14:00:06 GMT
Subject: [ANNOUNCE] Garbage collection for Python
References: <20000407221750.A29279@acs.ucalgary.ca> <38F13FB9.D1D16AF@mindspring.com> <slrn8f2if1.dop.nascheme@cranky.arctrix.com> <87r9cdhdb0.fsf@deneb.cygnus.argh.org> <FsunGI.BtB@world.std.com> <dphfd87s1h.fsf@mraz.iskon.hr>
Message-ID: <Fsuuw7.Inu@world.std.com>

Hrvoje Niksic (hniksic at iskon.hr) wrote:
> This is still different from being able to embed multiple Python
> instances in threads.

Are we talking about multiple Python instances running in a single
thread? That sounds like a very big challenge. The Python VM
would need to be rewritten to be reentrant so that multiple VMs
could simultaneously coexist, with distinct copies of the entire
computing environment (globals, locals, which modules had been
imported, etc). Then you'd need to arrange context switches; this
might be borrowed from the context switcher in Chris Tismer's
Stackless Python.

Making the PyVM reentrant would be a ridiculously large amount of
work, for which no other demand exists, as far as I'm aware. Maybe
the most helpful thing to do would be to point him to some other
language that already has a reentrant VM. Does Lua or Pike fit this
description?
-- 
 - - - - - - - - - - - - - - - - - - - - - - - -
Resistance is futile. Capacitance is efficacious.
Will Ware	email:    wware @ world.std.com


From rupole at compaq.net  Sun Apr  9 19:48:06 2000
From: rupole at compaq.net (Roger Upole)
Date: Sun, 09 Apr 2000 23:48:06 GMT
Subject: win32com and Reflection2
Message-ID: <WE8I4.100257$ra.2387838@news.easynews.com>

   I'm scripting Reflection2 sessions with win32com.client.Dispatch.
However, there are some constants (error codes, transmission modes,
etc) in a typelib that I can't seem to get at from the Dispatch object.
Anybody know of a way to access these from Python ?
   If not, is there any way to get to them from C++ ?
I have very little experience with C, but if it is simple I can
probably blunder my way through it.
             Roger Upole





From janssen at parc.xerox.com  Tue Apr 11 22:17:38 2000
From: janssen at parc.xerox.com (Bill Janssen)
Date: Tue, 11 Apr 2000 19:17:38 PDT
Subject: ANNOUNCE: Python 1.6 alpha 2 
In-Reply-To: Your message of "Tue, 11 Apr 2000 13:29:51 PDT."
             <200004112029.QAA09762@eric.cnri.reston.va.us> 
Message-ID: <00Apr11.191729pdt."3438"@watson.parc.xerox.com>

ILU seems to work fine with it.

Bill



From gregm at iname.com  Thu Apr 20 10:20:52 2000
From: gregm at iname.com (Greg McFarlane)
Date: Fri, 21 Apr 2000 00:20:52 +1000
Subject: Tkinter/Pmw/Python 1.6a#: xsync?
In-Reply-To: <m2zoqqcxg6.fsf@hooft.net>; from Rob W. W. Hooft on 19 Apr 2000 at 10:22:01AM
References: <m2zoqqcxg6.fsf@hooft.net>
Message-ID: <20000421002052.34558@nms.otc.telstra.com.au>

Why not try to narrow it down a bit?  First, check if a Tkinter-only
program exhibits the same behviour.  If not, then it is probably Pmw
that is causing the problem (unlikely :).  Otherwise, try a Tk gui (no
python), then an X only gui, then change your desktop - use a really
simple windown manager, such as twm (the first and still the best). 
You could also change your X library or X server.  Eventually, you
will change or eliminate something and the problem will go away - and
then you will have found the culprit.

But you already know all this.

On 19 Apr, Rob W. W. Hooft wrote:
> I upgraded too many things at the same time: Computer, OS revision,
> Desktop environment, and Python version....
> 
> For the python software that I make for my work, I also create a
> package including the python interpreter and all extensions required
> by my python code.
> 
> Since I upgraded to Python-in-CVS (from 1.5.2b1), Pmw0.8.3 (from 0.8)
> and started running on RedHat Linux 6.2 in Gnome, I have the
> impression that there are sometimes a bunch of X commands that stick
> in the pipeline: the buildup of the GUI stalls every time for about 1
> second just when the application starts a calculation or becomes
> idle waiting for input. Dialog windows do pop up immediately, but
> they stay blank for a while.
> 
>  - Is there anyone that has seen a similar effect? 
>  - Is there a way I can avoid this?
> 
> Regards,
> 
> Rob Hooft
> -- 
> =====   rob at hooft.net          http://www.xs4all.nl/~hooft/rob/  =====
> =====   R&D, Nonius BV, Delft  http://www.nonius.nl/             =====
> ===== PGPid 0xFA19277D ========================== Use Linux! =========
> -- 
> http://www.python.org/mailman/listinfo/python-list
> 

-- 
Greg McFarlane     INMS Telstra Australia     gregm at iname.com



From bawolk at pacbell.net  Wed Apr 12 01:33:07 2000
From: bawolk at pacbell.net (Bruce Wolk)
Date: Tue, 11 Apr 2000 22:33:07 -0700
Subject: How to know what re.sub took?
References: <38F387AE.1010704@gmx.net>
Message-ID: <38F40A93.9051A4A4@pacbell.net>

Try this:

>>> rex.findall(st)
['hans']

There would be more elements in the list if there were more matches
made.

>>> st = "max hans bill fred"
>>> rex.findall(st)
['hans', 'fred']

Cheers, Bruce

Per Kistler wrote:
> 
> Hello
> 
> How to know what has been substituted away, if one does something like:
> 
>  >>> import re
>  >>> rex = re.compile(r"(hans|fred|simon)")
>  >>> st = "max hans bill"
>  >>> st2 = rex.sub("",st)
>  >>> st2
> 'max  bill'
> 
> It took "hans" away, but how can I learn this automatically?
> 
> Thanks, Per.


From tom at parlant.com  Mon Apr  3 18:08:38 2000
From: tom at parlant.com (Thomas Lane)
Date: Mon, 03 Apr 2000 16:08:38 -0600
Subject: tkinter - changing the mouse pointer?
Message-ID: <38E91666.D3B87B4C@parlant.com>

Does anyone out there know how to change the mouse pointer in tkinter? I'd like
to change the mouse pointer to an hourglass (or the default 'wait' icon for the
os) and back to normal after a short operation. How do I go about doing this?
I've hunted through the Grayson book without success.

Thomas Lane
Parlant Technology
tom at parlant.com


From mwh21 at cam.ac.uk  Wed Apr 19 05:12:08 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 19 Apr 2000 10:12:08 +0100
Subject: Pointers
References: <38D00DCE.9740380F@be-research.ucsd.edu> <8bfsqo$kje$1@tribune.oar.net> <dp4s9wl9df.fsf@mraz.iskon.hr> <uPYC4.55946$3b6.224410@ozemail.com.au> <4k5D4.7566$0o4.54442@iad-read.news.verio.net> <38FB98F1.239EC0FB@whetstonelogic.com> <slrn8foe80.bd3.scarblac-spamtrap@flits104-37.flits.rug.nl> <8dhu2j$1gm$1@proxy2.fe.internet.bosch.de> <20000418223105.K15664@xs4all.nl>
Message-ID: <m3zoqqihef.fsf@atrus.jesus.cam.ac.uk>

Thomas Wouters <thomas at xs4all.net> writes:

> But you have to force *both* strings this way, interning one isn't
> enough.  It also doesn't work for integers (i dont think there is a
> way to do it with ints) so the best solution is just to not rely on
> this behaviour.

Compiling the code they turn up in is one way:

>>> def f():
...    x = 100000
...    y = 100000
...    return x is y
... 
>>> f()   
1

But, yes.

M.

-- 
  well, take it from an old hand: the only reason it would be easier
  to program in C is that you can't easily express complex  problems
  in C, so you don't.                   -- Erik Naggum, comp.lang.lisp


From someone at somewhere.com  Tue Apr 18 20:30:35 2000
From: someone at somewhere.com (Charley Horse)
Date: Wed, 19 Apr 2000 00:30:35 GMT
Subject: Using python on the web
References: <MPG.1365adcfa8470eef9896aa@news> <8di2bk$l1s$1@newshost.accu.uu.nl> <MPG.136631472f892a119896ac@news> <_W2L4.986$GK5.20898@news2-win.server.ntlworld.com> <MPG.13665ef93eedbda29896ad@news> <9N4L4.499$wYl.191816704@newsb.telia.net>
Message-ID: <MPG.13669ea3c1565e649896ae@news>

OK, thanks for the warning <g>. I already admire Zope, I'm just trying 
to learn some python to make sure I like the framework language.

In article <9N4L4.499$wYl.191816704 at newsb.telia.net>, effbot at telia.com 
says...
> Charley Horse <someone at somewhere.com> wrote:
> > It was just an unexpected surprise. I'm fairly new to all of this and
> > always heard of CGI in the context of "you don't have to user CGI
> > anymore because of the fancy new blah blah technology we offer!"
> 
> in Python land, that fancy technology is called
> Zope (www.zope.org).
> 
> well worth checking out, but be careful -- your
> brain might explode.
> 
> </F>
> 


From mwh21 at cam.ac.uk  Wed Apr 19 06:05:33 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 19 Apr 2000 11:05:33 +0100
Subject: testing true condition
References: <Pine.BSF.4.10.10004191138390.292-100000@hercule.perceval.be>
Message-ID: <m3snwiiexe.fsf@atrus.jesus.cam.ac.uk>

Gregoire Welraeds <helpdesk at perceval.be> writes:

> Hi,
> 
> I know that I can't do something like the following:
> 
> >>> while (line= f.readline()):
> 	...
> 
> as i could in C.  Is there any replacement tricks ?
> I know that I could use something with f.read() but i don't want to have
> the file completly loaded in memory

Tried the standard library module fileinput?

Cheers,
M.

-- 
112. Computer Science is embarrassed by the computer.
     -- Alan Perlis, http://www.cs.yale.edu/~perlis-alan/quotes.html


From chris at hazelnut.millersv.edu  Sat Apr 29 19:31:36 2000
From: chris at hazelnut.millersv.edu (Christopher Medalis)
Date: Sat, 29 Apr 2000 19:31:36 -0400 (EDT)
Subject: converting word documents to html
In-Reply-To: <390B501C.1891CB80@uiuc.edu>
Message-ID: <Pine.LNX.4.20.0004291930590.8040-100000@hazelnut.millersv.edu>

There is a package that does the job available here:

http://www.csn.ul.ie/~caolan/docs/MSWordView.html

Hope it helps! 
=) 
cheers
-clm

On Sat, 29 Apr 2000, Joe Saltiel wrote:

#:Date: Sat, 29 Apr 2000 16:11:56 -0500
#:From: Joe Saltiel <jsaltiel at uiuc.edu>
#:To: python-list at python.org
#:Newsgroups: comp.lang.python
#:Subject: converting word documents to html
#:
#:Hey,
#:
#:    I was wondering if anybody had done any work on converting
#:word processors type documents to html.  (word or word perfect,
#:pdf or anything else).  I am looking for a nice way to post
#:documents online that are laden with footnotes/endnotes.
#:Word2000 does an allright job, but adds a bunch of
#:junk that makes the file 3-4 larger then it has to be.
#:I was thinking of writing something to do that but just want
#:to check before I reinvent the wheel.....
#:
#:thanks
#:
#:Joe
#:
#:
#:-- 
#:http://www.python.org/mailman/listinfo/python-list
#:

============================================================
Christopher Medalis
chris at hazelnut.millersv.edu
chris at millersv.edu (work)




From robin at jessikat.demon.co.uk  Sat Apr 22 03:44:29 2000
From: robin at jessikat.demon.co.uk (Robin Becker)
Date: Sat, 22 Apr 2000 08:44:29 +0100
Subject: The Simple Economics of Open Source
References: <1255771356-36910442@hypernet.com>
 <3dya67xga4.fsf@amarok.cnri.reston.va.us>
Message-ID: <833l8JAdhVA5Ewp$@jessikat.demon.co.uk>

In article <3dya67xga4.fsf at amarok.cnri.reston.va.us>, Andrew M. Kuchling
<akuchlin at mems-exchange.org> writes
....
>Also, ever had to debug a problem in a binary-only library through
>trial-and-error?  You can burn lots of time that way.  <core dump>
>"Hmm... maybe that parameter can't be NULL."  <core dump> "Maybe that
>struct field should be 1." <core dump>.  I like the fact that, if I
>start seeing mysterious errors, I can dive down into the language
>interpreter, farther down into the C library, and even into the kernel
>if needed.  Though I've only followed a problem all the way into the
>Linux kernel once (to discover it was my error), it was reassuring to
>be able to do so.
>
Well I did a bit of reverse engineering in my time, but I think that
along with many others I'm living in the past. Software doesn't cost
nearly so much as it used to. In real terms the average application
doesn't cost nearly so much as my perceived $10k=car. The computer I
have at home is about the same power as a CDC 8600 of 20 years ago. It
doesn't need 50 people to run it and the software cost me around 300$.
I don't think I need to worry about cost (even though it still seems to
hurt). I worry more about the control of the software development
process more than the economics.

As for diving into the mechanics, would you do the same to your car? How
about fixing the video recorder or the mobile phone? Similar
intellectual requirements, but not so interesting to the 'nerds'.
-- 
Robin Becker


From schorsch at schorsch.com  Fri Apr 14 06:23:59 2000
From: schorsch at schorsch.com (Georg Mischler)
Date: Fri, 14 Apr 2000 10:23:59 GMT
Subject: simple httplib and urllib timeout question
References: <Pine.LNX.4.21.0004130904470.16470-100000@fep132.fep.ru>
Message-ID: <8d6rjk$tdd$1@nnrp1.deja.com>

phd at phd.russ.ru wrote:
> Georg Mischler wrote:
> > What exactly is your problem with signal.alarm() ?
>
>    Python delivers alarm only when in Python VM. If the script is
> blocked in C (even worse - in system call), the signal will NOT be
> delivered.
>    Of course, URL checker, while running through a big list of URLs,
> certainly spend most of the time in connect() and recv() system calls.
> This makes signals unusable.

This surprises me, because the snipped I posted is in everyday
use in something very similar to a link checker, and it works
perfectly well there. Reading throught the Python docs on
signals again, I guess that the different behaviour might
be caused by the underlying OS implementation.

I use it on Linux (going to check on FreeBSD soon), and it looks
like Linux will interrupt system level IO calls when a signal
arrives for the process. What platform does it fail on for you?


-schorsch

--
Georg Mischler  --  simulations developer  --  schorsch at schorsch.com
+schorsch.com+  --  lighting design tools  --  http://www.schorsch.com/


Sent via Deja.com http://www.deja.com/
Before you buy.


From gregholmes at my-deja.com  Tue Apr 18 08:21:51 2000
From: gregholmes at my-deja.com (gregholmes at my-deja.com)
Date: Tue, 18 Apr 2000 12:21:51 GMT
Subject: TKinter-destroying a window
Message-ID: <8dhk0j$cj5$1@nnrp1.deja.com>

Sorry, I'm reading TFM but haven't figured this
out yet. If I am done with a secondary (or
tertiary, or whatever) window (in TKinter), can I
just destroy the window or do I need to get rid
of any class instantiation that is
attached/associated to it? i.e.

....
theWindow=Toplevel()
solveThis=SolveThis(theWindow)
....
theWindow.destroy()
??

Will the class instantiation and its variables
and everything still be laying around?


Sent via Deja.com http://www.deja.com/
Before you buy.


From rzantow at my-deja.com  Wed Apr  5 09:12:47 2000
From: rzantow at my-deja.com (rzantow at my-deja.com)
Date: Wed, 05 Apr 2000 13:12:47 GMT
Subject: A Mountain of Perl Books + Python Advocacy
References: <20000405053920.20579.qmail@web2102.mail.yahoo.com>
Message-ID: <8cfe42$o02$1@nnrp1.deja.com>

In article <20000405053920.20579.qmail at web2102.mail.yahoo.com>,
  lewst <lewst at yahoo.com> wrote:
...
> What is it about Perl that makes it so much more popular and have such
> a huge grassroots swell?  I personally find Perl an abomination and
> Python a breath of fresh air.  Perl has that first mover advantage I
> suppose, but should that really make such a hugh difference?
...
Here's my theory, which is my own: It all depends on syntax. Perl's is
so nasty that its newbies *need* the books to make any use of it.
Python's is so clean that the occasional online nudge is sufficient.
Publishers see the market in Perl books, so they keep cranking them
out, and Perl users keep searching for the elusive key to understanding
the whole language, and buying more books. I'm really surprised there
are as many as 13 Python books on the market. I'd have thought there
was a need for maybe three, along with about two pamphlets and a poster.


Sent via Deja.com http://www.deja.com/
Before you buy.


From aa8vb at yahoo.com  Sat Apr  1 08:19:08 2000
From: aa8vb at yahoo.com (Randall Hopper)
Date: Sat, 01 Apr 2000 08:19:08 -0500
Subject: Tkinter dying a Flailing Death?
In-Reply-To: <3.0.6.32.20000331090918.007e4480@njdc.com>
References: <3.0.6.32.20000331090918.007e4480@njdc.com>
Message-ID: <20000401081908.A1353928@vislab.epa.gov>

anders schneiderman:
 |I've been working for a while on a Tkinter program, and almost every time I
 |quit it, it twice complains that "This program has performed an illegal
 |operation and will be shut down."  
 |
 |Since I'm new to Python, I figured it must be something I was doing.  So, I
 |tried running a demo from the Tkinter library, Dialog.py.  Surprise!  It
 |also dies twice, bitterly complaining about illegal operations.  The code
 |below is the Tkinter demo.  Any idea why it's so unhappy when it quits?

Sounds like an installation problem.  Someone with Python-on-MSWindows
experience is going to need to help you.

FWIW, your code works fine here on SGI IRIX (Python 1.5.2, Tcl/Tk 8.0.4).

-- 
Randall Hopper
aa8vb at yahoo.com



From rcunning at acm.org  Wed Apr  5 11:10:13 2000
From: rcunning at acm.org (Robert W. Cunningham)
Date: Wed, 05 Apr 2000 15:10:13 GMT
Subject: Pass by reference ?
References: <38E864C9.E1EE94F7@ecs.soton.ac.uk> <38EAF8D8.4B7C43B4@ecs.soton.ac.uk>
Message-ID: <38EB56AE.F938CE06@acm.org>

Jacek Generowicz wrote:

> Jacek Generowicz wrote:
>
> > Hi,
> >
> > I'm just trying to familiarize myself with Python.
> > In Magnus Lie Hatland's Instant Python, he
> > mentions that `all parameters in Python are passed
> > by reference'.
>
> [snip]
>
> Thank you for your answers.
>
> I would welcome a critique of the following summary:
>
> Python variables are implemented as references to
> values.
>
> Assignment (potentially) alters the reference rather
> than the value to which it refers.
>
> Objects can be classified into two sets: mutalble and
> immutable.
>
> Mutation of a mautable passed into a function will be
> reflected in the mutation of the object which was
> passed into the function; achieving similar behaviour
> for immutables is more complicated.
>
> Jacek

As a Python newbie, this sounds about right to me.  In my own mind, to
escape the ambiguity of how "Pass By Reference" and "Pass By Value" are
discussed in Computer Science, and how those same terms are (mis)used
when discussing Python, I've been using the idea of "Pass By Clone".

In essence, it seems to me Python and traditional CS meanings of these
phrases are cleanly separated by one level of abstraction.  And, given
that the two types of passing are themselves separated by one level of
abstraction (pass the data vs. pass a pointer to the data), confusion
seems inevitable.

Python can either 1) Pass a reference to the original data, or 2) Pass a
reference to a copy of the original data.  At no time does Python "pass
the data" directly (as in the "C" sense of, say, a register variable),
since everything is treated as though it were (in some sense) an object
(even if not "really" an object of the First Water).

Which, when mapped to conventional CS terminology (as extended by Moi),
becomes "Pass By Reference" and "Pass By Clone".

Clearly, "Pass By Clone" makes the original data item completely
inaccessible within the called function (ignoring bytehacks for now),
and thus immutable.  For me, this helps to clarify things, since Python
literally passes only references.

But, of course, that depends:  Is this a clarification, an obfuscation,
or Just Plain Wrong?


Finished-the-Python-tutorial-yesterday-and-the-FAQ-today-ly yours,

-BobC




From greg at cosc.canterbury.ac.nz  Wed Apr 12 22:36:34 2000
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Thu, 13 Apr 2000 14:36:34 +1200
Subject: global variables - how??
References: <20000412175218.784841CE49@dinsdale.python.org> <8d2jsh$3a9$1@news101.telia.com>
Message-ID: <38F532B2.C1518977@cosc.canterbury.ac.nz>

Anders Eggers-Krag wrote:
> 
> in my case I could otherwise send them along with every function, creating a
> huge
> overhead passing a million pointers around

If you find yourself doing that, your program is
badly in need of redesign. Think about grouping
related global variables together and making them
attributes of a class instance rather than globals. 
Then pass a reference to the instance instead of 
passing all the variables separately. Even better, 
make the functions which operate on those variables 
methods of the class whenever you can.

> but this really anoys me, and I *was* starting to like python...

Don't give up! It's always possible to redesign
any program which uses a lot of global variables
so that it uses very few, or even none. And in
my experience, the new design is *always* better -
clearer, easier to understand, easier to modify
and extend.

And once you get over global-variable-withdrawal,
you'll just keep liking Python more and more!

-- 
Greg Ewing, Computer Science Dept,
+--------------------------------------+
University of Canterbury,	   | A citizen of NewZealandCorp, a	  |
Christchurch, New Zealand	   | wholly-owned subsidiary of USA Inc.  |
greg at cosc.canterbury.ac.nz	   +--------------------------------------+


From nascheme at enme.ucalgary.ca  Sat Apr 15 17:54:21 2000
From: nascheme at enme.ucalgary.ca (Neil Schemenauer)
Date: Sat, 15 Apr 2000 21:54:21 GMT
Subject: [ANNOUNCE] Garbage collection for Python
References: <20000407221750.A29279@acs.ucalgary.ca> <m3r9cagx3c.fsf@atrus.jesus.cam.ac.uk> <38F65501.665243F6@gmx.de> <38F77397.45F26301@boi.hp.com>
Message-ID: <slrn8fhp95.1jv.nascheme@cranky.arctrix.com>

Bill Anderson <banderson at boi.hp.com> wrote:
>I would love to see a patch for 1.5.2!

Okay:

    http://www.enme.ucalgary.ca/~nascheme/python/gc-cycle-152.diff
    http://www.enme.ucalgary.ca/~nascheme/python/py152_gc1.zip

The zip file contains a Python 1.5.2 DLL for people on compiler
challenged machines.

    Neil

-- 
Real programmers don't make mistrakes


From avv at quasar.ipa.nw.ru  Tue Apr 18 18:39:06 2000
From: avv at quasar.ipa.nw.ru (Alexander V. Voinov)
Date: Tue, 18 Apr 2000 15:39:06 -0700
Subject: Playing wave files on Windows
References: <38FB6ED7.4C72773E@callware.com> <38FB7624.943A8C35@rubic.com> <38FB7862.7B15633@callware.com>
Message-ID: <38FCE40A.1D882566@quasar.ipa.nw.ru>

Hi,

Ivan Van Laningham wrote:
> > Then I can't interest you in my Python interface to
> > the Diamond Rio MP3 player?  <0.5 wink>
> >
> 
> Not unless everyone that I'll be distributing my application to will
> have one plugged in and ready to go;-)

But indeed, is it inevitable to have a specific module for one specific
MP3 player? I have no ideas about MP3, but it's strange that there are
no abstraction layers.

Seriously I'd be interested in crossplatform solutions. Does this FAQ
have an entry for ALSA/OSS? MIDI?

Regards

Alexander



From warlock at eskimo.com  Tue Apr  4 02:15:05 2000
From: warlock at eskimo.com (Jim Richardson)
Date: Tue, 04 Apr 2000 06:15:05 GMT
Subject: Why should I switch to Python?
References: <000701bf9c5a$b806b420$752d153f@tim>  <oqity066km.fsf@titan.progiciels-bpi.ca>
Message-ID: <slrn8ej2gc.b3t.warlock@raven.myth>

On 02 Apr 2000 08:16:57 -0400, 
 Fran?ois Pinard, in the persona of <pinard at iro.umontreal.ca>,
 brought forth the following words...:

>"Tim Peters" <tim_one at email.msn.com> writes:
>
>> When in doubt, there's no substitute for exhaustive enumeration.
>
>Hmph!  Hope you do not doubt there is an infinite number of primes! :-)

Out of curiousity, how does one (mathematically) prove that?
(if this turns out to be one of those 12' blackboards full of greek
letters and funny squiggles, Ok, I cry uncle...)
-- 
Jim Richardson
	Anarchist, pagan and proud of it
WWW.eskimo.com/~warlock
	Linux, because life's too short for a buggy OS.



From joacim at home.se  Thu Apr 27 09:46:26 2000
From: joacim at home.se (JJ)
Date: Thu, 27 Apr 2000 15:46:26 +0200
Subject: Concatening string and integer
Message-ID: <8e9gco$jlt$1@vg170.it.volvo.se>

Hi!

I want to concatenate a string containing both strings and integers like:

strTemp = 'I make test number: ' + iMyNumber + ' once again'

It doesn't work, how should I do this?

BR
Joacim




From donn at oz.net  Tue Apr 18 02:55:24 2000
From: donn at oz.net (Donn Cave)
Date: 18 Apr 2000 06:55:24 GMT
Subject: else clauses in while and for loops
References: <%TNK4.430$vr2.32615@news.uswest.net> <2JKK4.325$vr2.19414@news.uswest.net> <Zv_C4.6415$4N.169440@news.tpnet.pl> <38DD2EE3.C7F3CB54@webamused.com> <8boe2k$gmh$1@inputplus.demon.co.uk> <v0RD4.13$n68.186989568@newsb.telia.net> <8c7b29$37v$1@inputplus.demon.co.uk> <38F79B93.B03F0A6B@mindspring.com> <8dg14u$f08$1@nntp6.u.washington.edu>
Message-ID: <8dh0ss$4ia$0@216.39.151.169>

Quoth "Jeff Massung" <jmassung at magpiesystems.com>:
| Donn Cave wrote in message <8dg14u$f08$1 at nntp6.u.washington.edu>...
|> Yes, but if you look at the examples, they also use "break".  After a
|> loop, "else" basically means "if no iteration of the loop met the
|> conditions to break out."  It's a common usage, where in C we'd compare
|> the final loop index with the maximum (or whatever) loop index to see
|> if the loop ran all the way through.
|
| I don't think that I'm following you, give me an example where the loop
| ends, but the else wouldn't be executed and I think I'll better understand.

OK, let's bring back the example from a few posts back:

for envname in ['TMPDIR', 'TMP', 'TEMP']:
    if os.environ.has_key(envname):
        path = [os.environ[envname]]
        basepaths = validate_paths(path)
        if basepaths:
            break
else:
    basepaths = validate_paths(_tmpdirs)

Here the else branch executes if none of the envnames turns up
in environ.  The conditions to break out are 1) envname is there,
and 2) validate_paths() can use it;  if those conditions are never
met, the else clause executes.

Now it's simpler to say "else executes if break doesn't".  I'm
just trying to complicate matters with a formula that tries to
make sense in terms of the common usage of "else" in other familiar
contexts.

	Donn Cave, donn at oz.net


From sabren at manifestation.com  Sun Apr 16 20:39:00 2000
From: sabren at manifestation.com (Michal Wallace (sabren))
Date: Sun, 16 Apr 2000 20:39:00 -0400 (EDT)
Subject: module naming?
Message-ID: <Pine.BSF.4.10.10004162018190.25069-100000@chi.pair.com>

Hey All,

 I've got a package I'm getting ready to share named zdc, and it has
some submodules: zdc.Record, zdc.Field, zdc.IdxDict.. Each one of
those contains a class with the corresponding name: zdc.Record.Record,
zdc.Field.Field, zdc.IdxDict.IdxDict.

  I've read the essay on built-in package support,
( http://www.python.org/doc/essays/packages.html ), but I just don't like
the notion of having to say:

   import zdc.Record
   rec = zdc.Record.Record()


when this looks so much nicer:

   import zdc
   rec = zdc.Record()


Now, I know I can do this by putting all my classes in zdc.py, but I
think it's a lot cleaner to keep my classes in their own files.

I also don't want to do "from zdc import *" because people who use my
modules will probably only want one of the classes at a time. Or at
least I will. :)

The other approach I'm considering is changing my module names to
lowercase, and doing something like this in zdc/__init__.py :


def Record(dbc, table):
   import zdc.record, zdc
   zdc.Record = zdc.record.Record
   return zdc.record.Record(dbc, table)


.. That way, the first time you call zdc.Record() it imports
zdc.record, and every time after that, it just initializes a
zdc.record.Record object... 

Is there a better way to do this?

Cheers,

- Michal
-------------------------------------------------------------------------
http://www.manifestation.com/         http://www.linkwatcher.com/metalog/
-------------------------------------------------------------------------




From effbot at telia.com  Tue Apr  4 13:09:14 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Tue, 04 Apr 2000 17:09:14 GMT
Subject: Tuples -- who needs 'em
References: <00e101bf9d80$a8384d20$74eb0b18@stcla1.sfba.home.com> <va3G4.1110$n68.190454272@newsb.telia.net> <dpn1nauiek.fsf@mraz.iskon.hr>
Message-ID: <_kpG4.5040$74.81544@newsc.telia.net>

Hrvoje Niksic <hniksic at iskon.hr> wrote:
> "Fredrik Lundh" <effbot at telia.com> writes:
>
> > -- tuples can be used as dictionary keys.  mutable collections can
> >    not.
>
> This is wrong.

this is why it usually takes more than one iteration to come up
with a watertight specification.

yeah, you're right -- python allows you to pretend that a mutable
collection class isn't really that mutable...

...just like tuples are basically lists that pretend they're not really
mutable (okay, CPython stores tuples in a slightly more efficient
ways, but that's an optimization)

> > -- good designers know that hypergeneralization tends to be a bad
> >    idea.
>
> ...and this is arm-waiving.

oh, I don't know.  I've designed stuff for some 20 years, and I've
seen way too many failed attempts to design overly general stuff
(even when the designers succeed, it can take ages for ordinary
users to catch up -- just witness SGML vs. XML vs. SML vs. good
old text files...).

but maybe it's just me.

</F>




From alwagner at tcac.net  Mon Apr 24 21:10:10 2000
From: alwagner at tcac.net (Albert Wagner)
Date: Mon, 24 Apr 2000 20:10:10 -0500
Subject: Tkinter
References: <A63AFB20111ED311AC5500A0C96B7BF5015F0F44@orsmsx36.jf.intel.com>
Message-ID: <3904F072.DAD6D48D@tcac.net>

http://www.python.org/topics/tkinter/doc.html

"Daley, MarkX" wrote:
> 
> I am trying to use Tkinter for a GUI interface, but I am having a hard time
> finding manuals for it.  Is there documentation comparable to Python's?
> Also, is there a demo script floating around anywhere that gives generic
> examples for common GUI functions or widgets?
> 
> TIA
> 
> - Mark

-- 
Small is Beautiful


From sp00fD at yahoo.com  Mon Apr 24 14:44:24 2000
From: sp00fD at yahoo.com (sp00fD)
Date: Mon, 24 Apr 2000 18:44:24 GMT
Subject: A Mountain of Perl Books + Python Advocacy
References: <20000405053920.20579.qmail@web2102.mail.yahoo.com> <38ED40B0.A48EC4CC@udel.edu> <tNvI4.18$_N4.1751@vic.nntp.telstra.net>
Message-ID: <8e24m5$2vn$1@nnrp1.deja.com>

This is probably the only thing that I hate about python, the lack of
good freakin' documentation.  I used perl formerly and found it to be
sooo easy (partially because I'm a unix admin and perl is very unix-
like), there were wonderful examples everywhere on how to do things.
Python, on the other hand, has __extremely__ poor documentation (as do
most languages), and I believe this is a large part of why perl has
become so popular.  Most people can't discuss theory and get coding
ideas from it, nor can they see a method and parameters and know
exactly how to use the module correctly.  Frankly, the stock python
documentation is worthless to me, and often times I find myself
perusing modules, without luck because unfortunately most of them are
not very well commented either.  I think I'm going to have to go club
Tchrist over the head and reprogram him into a pythoneer, then maybe
we'd have some good documentation as well.

Mike

In article <tNvI4.18$_N4.1751 at vic.nntp.telstra.net>,
  i_luv_fudge at I_HATE_SPAM.hotmail.com (.,.) wrote:
>
>
> >> Now something I can't quite figure out is: why are there so many
more
> >> books on Perl out there than on Python?
>
> >6 months ago, I discovered Python and it immediately clicked.
> >Learned most of Python before buying my first Python book.
>
> I don't know about any of the rest of you guys,  but I figure this is
where I
> should speak up.  I have just discovered Python in the last week.  I
was
> trying to manipulate about 400 smallish text files with identical
datafields
> and differing values into a format that I could display on the web.
After
> spending about 80 hours cutting, pasting deleting and hten realizing
that it
> would be good to start again,  so I could include that peice of info
afterall,
>  I decided there had to be a better way.
>
>  I have done one semester in C++ about 2 years back, and nothing
before or
> since.  I have a very great friend who is helping me out with my
script and
> what I want to do, but this is because they have also only just
discovered
> Python and have a direct need to use it for a project over the next
week.
> This is very lucky for me, cause I am so close to forgetting
absolutely
> everything I learnt in that semester, except the basic tenets of form
and
> structure.
>
> Trying to get information out of the python website is virtually
impossible
> for the layman trying to learn Python.   I am finding it very very
hard to
> expand on the stuff my mate has shown me already,  and I need to
expand so
> much!  I am looking for the Learning Python book now,  but I doubt I
will find
> it on any shelves in my town, as I live in a backwater country
downunder.
>
> ======
>
> Whuddaya know?   I just rang TechBooks and they have 4 in stock  :-)
>
> I am off to buy one.
>
> It still doesn't change the fact that this is a VERY hard language to
learn as
> a layman without a guide or some sort of coding TLA translator.  When
and if I
> get proficient I might write a page for people with that in mind.
>
> :-)
>
> I have no doubt you will be hearing from me again as I post newbie
> questions....
>
> *sigh*
>
> Nemir
>


Sent via Deja.com http://www.deja.com/
Before you buy.


From rpm at wag.caltech.edu  Wed Apr 19 16:14:10 2000
From: rpm at wag.caltech.edu (Richard P. Muller)
Date: Wed, 19 Apr 2000 13:14:10 -0700
Subject: Building python on HPUX-10.2
Message-ID: <38FE1392.D1C01D64@wag.caltech.edu>

I have successfully built python on a number of diffent platforms, but
I'm having serious problems on HPUX 10.2. I go through the normal
configure/make procedure, but the make crashes right away, when it is
trying to link the first object (pgen), because it can't find the
correct dl library (?).

Has anyone successfully compiled Python on HPUX-10.*? If so, what
configure string did you use?

Rick



From moshez at math.huji.ac.il  Sat Apr 15 01:26:57 2000
From: moshez at math.huji.ac.il (Moshe Zadka)
Date: Sat, 15 Apr 2000 07:26:57 +0200 (IST)
Subject: Complaint Dept: My shoes!
In-Reply-To: <8d7rf5$odi$1@newshost.accu.uu.nl>
Message-ID: <Pine.GSO.4.10.10004150725240.11261-100000@sundial>

On 14 Apr 2000, Martijn Faassen wrote:

> And now I'm going to tell you how to make len(None) return 0. This
> is an implementation detail. It's a *bug* that ought to be fixed.
> 
> Put this in the module where you want len(None) to be 0:
> 
> None = ""

Even better, anywhere in the application put

import __builtin__
__builtin.None = ""
del __builtin__

making-python-scarier-for-everybody-ly y'rs, Z.

--
Moshe Zadka <mzadka at geocities.com>. 
http://www.oreilly.com/news/prescod_0300.html
http://www.linux.org.il -- we put the penguin in .com




From effbot at telia.com  Thu Apr 13 05:58:40 2000
From: effbot at telia.com (Fredrik Lundh)
Date: 13 Apr 2000 04:58:40 -0500
Subject: Dr. Dobb's Python-URL! - weekly Python news and links (Apr 13)
Message-ID: <75FC993A455AAF55.E13EAC3AEC7F76DE.AD552F20A3413B64@lp.airnews.net>

Sightings:
    Python 1.6 alpha 2 is out!  Lots of bug fixes, changes to
    the windows installer, and more.  Expect a third alpha
    within two weeks or so.
        http://www.python.org/1.6/

    On a related note: argumentclinic.py
        http://www.deja.com/=dnc/getdoc.xp?AN=606780882

    Python Wins Jolt Productivity Award:
        http://www.sdmagazine.com/sdonline/jolts/joltopen.html

    Linux Journal: Python Programming for Beginners
        http://www2.linuxjournal.com/lj-issues/issue73/3946.html

    Remember the slightly misleading 1.6 release note we pointed
    to last week?  Turned out that Guido didn't do it -- or at least
    that's what he says:
        http://www.deja.com/=dnc/getdoc.xp?AN=605330122
        http://www.deja.com/=dnc/getdoc.xp?AN=606107926

    In fact, Greg Ward claims he did it.  And you know what?  He
    might be right -- after all, who else would use the word "frob"
    sixteen times in a single post?
        http://www.deja.com/=dnc/getdoc.xp?AN=606237127

    On the other hand, we keep hearing rumors that Greg is just
    a scapegoat, and that the real story is that Barry has found
    a portal into Guido's head:
        http://www.lfw.org/jminc/Guido%20Van%20Rossum/http://www.python.org/

    Or whatever...
        http://metalab.unc.edu/Dave/Dr-Fun/df200004/df20000406.jpg

Stuff:
    Neil Schemenauer: Garbage collection for Python
        http://www.deja.com/=dnc/getdoc.xp?AN=608256399

    Kevin Cazabon: PIL compatible fonts
        http://members.home.com/lenticular/

    Bruce Dodson: Python Language Support for ArcView GIS
        http://www.deja.com/=dnc/getdoc.xp?AN=608466145

    Mark Hammond: win32all 130 for Python 1.6 available
        http://www.deja.com/=dnc/getdoc.xp?AN=605169788

    Ken Seehof: Implementing generators in Python
        http://starship.python.net/crew/seehof/Generator.html

    Thomas Bellman: Why your network code may stop working
    when switching from FDDI to Ethernet
        http://www.deja.com/=dnc/getdoc.xp?AN=606778326

    Just van Rossum: Stackless Python for MacOS
        http://www.deja.com/=dnc/getdoc.xp?AN=606779928

    Michael Hudson: bytecodehacks, April 2000
        http://www.deja.com/=dnc/getdoc.xp?AN=605263124


========================================================================

Everything you want is probably one or two clicks away in these pages:

      Python.org's Python Language Website is the center of Pythonia
          http://www.python.org

      eff-bot's complements this digest with his daily python url
          http://hem.passagen.se/eff/url.htm

      Much of Python's real work takes place on Special-Interest Group
          mailing lists http://www.python.org/sigs/

      Python Consortium emerges as an independent nexus of activity
          http://www.python.org/consortium

      The Vaults of Parnassus ambitiously collects Python resources
          http://www.vex.net/~x/parnassus/   

      Python To-Do List anticipates some of Python's future direction
          http://www.python.org/cgi-bin/todo.py

      Python Journal is at work on its second issue
          http://www.pythonjournal.com

      Links2Go is a new semi-automated link collection; it's impressive
      what AI can generate
          http://www.links2go.com/search?search=python

      Archive probing trick of the trade:
          http://www.dejanews.com/dnquery.xp?QRY=&DBS=2&ST=PS&defaultOp=AND&LNG=ALL&format=threaded&showsort=date&maxhits=100&groups=comp.lang.python


Previous - (U)se the (R)esource, (L)uke! - messages are listed here:
  http://purl.org/thecliff/python/url.html
or
  http://www.dejanews.com/dnquery.xp?QRY=~g%20comp.lang.python%20Python-URL%21


Suggestions/corrections for next week's posting are always welcome.
  http://www.egroups.com/list/python-url-leads/

To receive a new issue of this posting in e-mail each Monday morning,
ask <claird at neosoft.com> to subscribe.  Mention "Python-URL!".


-- The Python-URL! Team--

Dr. Dobb's Journal (http://www.ddj.com) is pleased to participate in and
sponsor the "Python-URL!" project.
-- 

Cameron Laird <claird at NeoSoft.com>
Business:  http://www.Phaseit.net
Personal:  http://starbase.neosoft.com/~claird/home.html


From burnout at heathers.stdio.com  Fri Apr 14 14:24:32 2000
From: burnout at heathers.stdio.com (Eric H.)
Date: Fri, 14 Apr 2000 14:24:32 -0400
Subject: Looking for IRIX Python programmers for help
Message-ID: <38F76260.16603101@heathers.stdio.com>

Hi,

While building Python 1.5.2 for IRIX 6.5, I was editing Modules/Setup
and discovered there is a section that mentions that you need a
"multimedia development option(I think)" to build the "al" module and a
couple others. Can someone clarify this? Is this really needed and if
so, where do I get it? Thanx for your wonderful response and L8a


P.S. A response via email would be much welcome since I don't get to
check newsgroups every day. 


-- 
Parts of this message, including the email address, ? 2000-Eric Haddix. 
Use implies agreement to the terms and conditions made available on my 
homepage. Keep the Internet free, refuse to support the providers that
copyright YOUR email and messages. All rights reserved. U.C.C. 1-207

visit my homepage @ http://www.bounce.to/EyesCremeProductions


From neelk at brick.cswv.com  Thu Apr 27 22:09:43 2000
From: neelk at brick.cswv.com (Neel Krishnaswami)
Date: 28 Apr 2000 02:09:43 GMT
Subject: The Simple Economics of Open Source
References: <B529350E.51B9%raffael@mediaone.net> <390534BB.3D6CEBDF@libc.org> <raffael-01CD89.20575826042000@news.ne.mediaone.net>  <3908DB34.5549085F@libc.org>
Message-ID: <slrn8ghuqg.2q1.neelk@brick.cswv.com>

Bill Anderson <bill at libc.org> wrote:
> 
> *All* software is under copyright or patent. By your argument, no
> software is commodity.

Minor correction: it's possible to place software, like any other IP,
in the public domain. Some quite significant open source projects,
such as the CMUCL Common Lisp compiler, are in the public domain. j


Neel



From dworkin at ccs.neu.edu  Tue Apr  4 17:26:28 2000
From: dworkin at ccs.neu.edu (Justin Sheehy)
Date: 04 Apr 2000 17:26:28 -0400
Subject: ldap support?
In-Reply-To: oliver's message of "Tue, 4 Apr 2000 16:06:06 -0400"
References: <Pine.LNX.4.10.10004041604050.10324-100000@worf.mcnc.org>
Message-ID: <vndg0t1k16j.fsf@camelot-new.ccs.neu.edu>

oliver <fwang2 at yahoo.com> writes:

> I wonder is there any standard LDAP support module in python? I looked up
> SIGs, doesn't show anything there. I appreciate your help.

Nothing standard at this time, but there is Python-LDAP.

http://python-ldap.sourceforge.net/

-Justin

 



From quinn at riyal.ugcs.caltech.edu  Tue Apr  4 20:27:28 2000
From: quinn at riyal.ugcs.caltech.edu (Quinn Dunkan)
Date: 5 Apr 2000 00:27:28 GMT
Subject: Getting the argument names passed to a function ?
References: <20000404225810.A12156@sz-sb.de> <HRsG4.5120$74.82867@newsc.telia.net>
Message-ID: <slrn8el27m.3h1.quinn@riyal.ugcs.caltech.edu>

On Tue, 04 Apr 2000 21:08:55 GMT, Fredrik Lundh <effbot at telia.com> wrote:
>Andreas Jung <ajung at sz-sb.de> wrote:
>>
>> def myfunc(arg1,arg2):
>> ...
>>
>>
>> a=1;b=2
>> myfunc(a,b)
>>
>> Is it possible to get the name of variables that  were
>> used as arguments to call myfunc(). myfunc() should be
>> able to determine the names 'a' and 'b'. Don't ask
>> me why I need this - it's for a worse Python hack :-)
>
>short answer: no
>
>longer answer: maybe, using the incredibly clever and
>totally evil bytecodehacks stuff:
>
>    http://bytecodehacks.sourceforge.net

Hmm, could you say:

def arghack(arg):
    for k, v in globals().items():
        if arg is v:
            return k
    else:
        return None

Seems to work ok...


From scarblac-spamtrap at pino.selwerd.nl  Thu Apr 27 11:07:55 2000
From: scarblac-spamtrap at pino.selwerd.nl (Remco Gerlich)
Date: 27 Apr 2000 15:07:55 GMT
Subject: How can I get the name of an object???
References: <8e712k$65c$1@nnrp1.deja.com> <etd8zy03m44.fsf@w20-575-108.mit.edu> <8e9090$c4d$1@nnrp1.deja.com>
Message-ID: <slrn8gglsv.ep7.scarblac-spamtrap@flits104-37.flits.rug.nl>

spex66 at my-deja.com wrote in comp.lang.python:
> In article <etd8zy03m44.fsf at w20-575-108.mit.edu>,
>   Alex <cut_me_out at hotmail.com> wrote:
> >
> > An object could be referred to by any number of names:
> >
> > dd = zz ()
> > ee = dd
> >
> > Which name for the object do you wish to know?  ee or dd?
>
> dd or in other words, the name of the FIRST made instance...

What would happen in the following cases:
1)
  dd = zz()
  ee = dd
  del dd
  print ee.__name__
2) 
  dd = [xx(), yy(), zz()]
  print dd[2].__name__
3)
  print zz().__name__


Classes have a name (the one used in their class: statement). 
Instances have no name, period.

But why would you need it? In most cases where people try to use something
like this, dictionaries are a solution.

-- 
Remco Gerlich,  scarblac at pino.selwerd.nl
"This gubblick contains many nonsklarkish English flutzpahs, but the
 overall pluggandisp can be glorked from context"  (David Moser)


From scarblac-spamtrap at pino.selwerd.nl  Fri Apr 14 10:12:08 2000
From: scarblac-spamtrap at pino.selwerd.nl (Remco Gerlich)
Date: 14 Apr 2000 14:12:08 GMT
Subject: Future compatibility of the Python C API
References: <lwEJ4.7980$Ib7.101092@typhoon2.kc.rr.com> <q4FJ4.1198$1D.10037@news-server.bigpond.net.au> <8d76t2$pig@newton.cc.rl.ac.uk>
Message-ID: <slrn8fed2o.4fe.scarblac-spamtrap@flits104-37.flits.rug.nl>

Richard Brodie wrote in comp.lang.python:
> 
> "Mark Hammond" <mhammond at skippinet.com.au> wrote in message
> news:q4FJ4.1198$1D.10037 at news-server.bigpond.net.au...
> 
> > > My big concern is Python 3000- what are the odds
> > > that its C API will be backwards compatibility with Python 1.x?
> >
> > Not very good.  If it is a real problem, simply dont upgrade.  I would
> > however expect the changes to cause compiler errors that are simple to
> > fix - ie, very similar APIs, and unlikely to silently change behaviour.
> 
> I suppose that someone out there is working on persuading SWIG to
> generate suitable wrapper code?

Since noone knows what Python 3000 is going to look like yet, that would be
very hard, wouldn't it? :)


-- 
Remco Gerlich,  scarblac at pino.selwerd.nl
  5:08pm  up 39 days,  4:28,  8 users,  load average: 2.27, 2.10, 2.09


From phd at phd.russ.ru  Tue Apr 11 07:18:11 2000
From: phd at phd.russ.ru (Oleg Broytmann)
Date: Tue, 11 Apr 2000 11:18:11 +0000 (GMT)
Subject: Redirecting the output from sys.stdout
In-Reply-To: <XFMail.000411131415.mikael@isy.liu.se>
Message-ID: <Pine.LNX.4.21.0004111116560.30946-100000@fep132.fep.ru>

On Tue, 11 Apr 2000, Mikael Olofsson wrote:
>  >  while 1:
> 
> Why not the following?

   It was just an example.
   Usually I do parsing line-by-line, instead of slurping the whole file in
- I often work with huge files...

Oleg.
---- 
    Oleg Broytmann    http://members.xoom.com/phd2.1/    phd2 at earthling.net
           Programmers don't die, they just GOSUB without RETURN.




From jstok at bluedog.apana.org.au  Sat Apr  1 04:20:19 2000
From: jstok at bluedog.apana.org.au (Jason Stokes)
Date: Sat, 1 Apr 2000 19:20:19 +1000
Subject: Debugging confusion -- too many stacks!
Message-ID: <7cjF4.60441$3b6.236709@ozemail.com.au>

I'm confused about the difference between the C stack, the Frame stack, and
the stack *inside* each execution frame.

>From reading the source, I believe that each new code object gets a new
frame allocated on the frame stack, within which is stored all kinds of
useful information about the context the code is executing in -- the global
and local environment, a tuple of constants, a tuple of arguments etc.
*Within* each frame is *another* stack, upon which the Python virtual
machine loads and manipulates intermediate values.

However, I'm not sure what the terminology is to refer to this third stack.
Understanding the distinction seems to be vital to using pdb, though.




From kc5tja at garnet.armored.net  Sat Apr 29 14:11:27 2000
From: kc5tja at garnet.armored.net (Samuel A. Falvo II)
Date: 29 Apr 2000 18:11:27 GMT
Subject: What does "first class" mean?
References: <3909F2F5.A6F91256@spacenet.tn.cornell.edu> <slrn8gkalp.3bk.neelk@brick.cswv.com>
Message-ID: <slrn8gn54f.4ih.kc5tja@garnet.armored.net>

In article <slrn8gkalp.3bk.neelk at brick.cswv.com>, Neel Krishnaswami wrote:
>In Pascal or C, for example, you can't write a procedure that creates
>a function and returns it to its callers, whereas in Scheme or Python
>this is trivial.

This isn't *strictly* true.  The following code, for x86 processors, returns
a function that does nothing:

typedef void NoOp();

NoOp *GenerateNoOpFunction( void )
{
   char *buf;
   
   buf = malloc( 2 );
   if( buf != NULL )
   {
      buf[0] = 0x90;	/* NOP instruction; for demonstration */
      buf[1] = 0xC3;	/* RET instruction */
   }
   
   return buf;
}

Sorry, I couldn't resist.  :-)

Seriously, though, this technique has been used in various graphics drivers,
most often with software-only blit engines (where speed is critical).
They'd basically act as dedicated compilers.  Windows 95's stock VGA device
driver does this, if memory serves me correctly.

-- 
KC5TJA/6, DM13, QRP-L #1447
Samuel A. Falvo II
Oceanside, CA


From akuchlin at mems-exchange.org  Tue Apr 11 16:13:38 2000
From: akuchlin at mems-exchange.org (Andrew M. Kuchling)
Date: 11 Apr 2000 16:13:38 -0400
Subject: [ANNOUNCE] Garbage collection for Python
References: <20000407221750.A29279@acs.ucalgary.ca> <38F13FB9.D1D16AF@mindspring.com> <slrn8f2if1.dop.nascheme@cranky.arctrix.com> <87r9cdhdb0.fsf@deneb.cygnus.argh.org> <FsunGI.BtB@world.std.com> <38F319A4.89621192@mdlink.de> <38F351B0.45E137AA@roguewave <38F3836A.5F9A7C6F@roguewave.com>
Message-ID: <3d7le4gzv1.fsf@amarok.cnri.reston.va.us>

Synchronicity again ... someone just submitted some patches that seem
to make this change and use interlocked functions for refcounting; see
http://www.python.org/pipermail/patches/2000-April/000413.html .
We'll have to wait and see if they get in.

--amk


From wilkes at memphisonline.com  Fri Apr 21 11:36:53 2000
From: wilkes at memphisonline.com (wilkes at memphisonline.com)
Date: Fri, 21 Apr 2000 15:36:53 GMT
Subject: win32all build 125 debug builds?
Message-ID: <8dpsik$fsu$1@nnrp1.deja.com>

Does anyone have the debug builds for win32all build 125, or the source
so that I may build them.

Thanks,
Wilkes Joiner
wilkes at electracash.com


Sent via Deja.com http://www.deja.com/
Before you buy.


From greg at cosc.canterbury.ac.nz  Wed Apr 19 23:31:12 2000
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Thu, 20 Apr 2000 15:31:12 +1200
Subject: __del__ problem - more details provided. still stumped though.
References: <aw2L4.8907$HG1.245743@nnrp1.uunet.ca> <7ojL4.9424$HG1.254543@nnrp1.uunet.ca>
Message-ID: <38FE7A00.102FB5A1@cosc.canterbury.ac.nz>

Warren Postma wrote:
> 
> class DBPtr :
>     def __init__(self,this):
>         self.this = this
>         self.thisown = 0
>     def __del__(self):
>         if self.thisown == 1 :
>             dbc.delete_DB(self.this)  # appears some exceptions being raised
> here.

That's a SWIG-generated shadow class! I had the very
same kind of problem when I used SWIG to wrap some
Xlib functions. What is boils down to is that:

*** SWIG generates buggy __del__ methods! ***

They don't protect themselves against the loss of
module-level names at finalisation time. In the
example above, if 'dbc' in the current module, or
'delete_DB' in the dbc module, has been clobbered
by Py_Finalize before the __del__ method is called,
trouble will result.

My solution was to write my own shadow classes rather
than relying on SWIG (I had other reasons to do this
in any case). In your case, you could try editing the
__del__ method to say

      def __del__(self, delete_DB = dbc.delete_DB):
         if self.thisown == 1 :
             delete_DB(self.this)

and similarly in any other __del__ methods that
are lying about.

-- 
Greg Ewing, Computer Science Dept,
+--------------------------------------+
University of Canterbury,	   | A citizen of NewZealandCorp, a	  |
Christchurch, New Zealand	   | wholly-owned subsidiary of USA Inc.  |
greg at cosc.canterbury.ac.nz	   +--------------------------------------+


From psl at mitre.org  Thu Apr 27 10:50:33 2000
From: psl at mitre.org (Pete)
Date: Thu, 27 Apr 2000 10:50:33 -0400
Subject: getting return values from calldll
Message-ID: <390853B8.87AEBF0C@mitre.org>

How can I check the return value from calldll?  I called a function with
a boolean return,  and tried printing both the return value argument and
a variable I set equal to the function call,  but they both printed out
nothing.



From ulf.engstrom at b2b-link.com  Wed Apr  5 04:36:35 2000
From: ulf.engstrom at b2b-link.com (=?iso-8859-1?Q?Ulf_Engstr=F8m?=)
Date: Wed, 5 Apr 2000 10:36:35 +0200
Subject: alphanumerics in mails
Message-ID: <029501bf9eda$0dd8e2c0$858647c1@DEVELOPMENT>

I've made a small mail-client which is working very good, except for when I
get mails with alphanumerics (which happens quite often since I'm in
scandinavia;).
If I send a mail with subject:
'Ett ?'
The subject line in my mailclient will say:
'=?iso-8859-1?Q?Ett_=E5?='
Is there some kind of parser for this, has someone made anything like this
or do I have to make one?
Handling my mail uses poplib,cgi,multifile and mimetools.
Regards
Ulf




From tim_one at email.msn.com  Sun Apr  2 01:18:25 2000
From: tim_one at email.msn.com (Tim Peters)
Date: Sun, 2 Apr 2000 01:18:25 -0500
Subject: multiple interpreter sessions
In-Reply-To: <8c2lkc$1di$1@cantaloupe.srv.cs.cmu.edu>
Message-ID: <000b01bf9c6b$413cca80$752d153f@tim>

[Helge Hess]
> I would like to have separate Python runtimes in a single server, so
> that they don't affect each other (eg for security and state reasons).
> ...

See the description of Py_NewInterpreter() in the "Initialization,
Finalization, and Threads" section of the Python/C API manual.  You can
create sub-interpreters from a C extension this way, but there's no
Python-level interface to this at present.  Haven't used it myself, and,
indeed, have yet to see any use of it.

you-could-be-the-first-on-your-block!-ly y'rs  - tim





From markx.daley at intel.com  Tue Apr 18 12:26:34 2000
From: markx.daley at intel.com (Daley, MarkX)
Date: Tue, 18 Apr 2000 09:26:34 -0700
Subject: Lists
Message-ID: <A63AFB20111ED311AC5500A0C96B7BF5015F0F2C@orsmsx36.jf.intel.com>

This feels like a real newbie question, but here is the code that is causing
my question:

# Learning exceptions

def test():
	a = [-5, -4, -3, -2, -1, 0, 1, 2, 3, 4, 5]
	print a
	b = 5
	for item in a:
		try:
			print b / a[item]
		except ZeroDivisionError:
			pass

Here is the output of this code under IDLE:

[-5, -4, -3, -2, -1, 0, 1, 2, 3, 4, 5]
5
2
1
1
1
-1
-2
-2
-3
-5

Why is the list being processed in reverse?  The exception works fine, but
the math is backwards.  I don't recall this happening before.

TIA

- Mark




From aek at mail1.stofanet.dk  Wed Apr 12 13:28:37 2000
From: aek at mail1.stofanet.dk (Anders Eggers-Krag)
Date: Wed, 12 Apr 2000 19:28:37 +0200
Subject: global variables - how??
Message-ID: <8d2bne$qn7$1@news101.telia.com>

how do I define a propper global variable that works in all functions in all
modules?

I have tried simply
defining one outside of any function in the main module,
I have tried using the global keyword under the same cicumstances,
but they are not even avialible inside of functions within the main
module...


--
Anders Eggers - Krag




From mjackson at wc.eso.mc.xerox.com  Thu Apr 20 12:22:49 2000
From: mjackson at wc.eso.mc.xerox.com (Mark Jackson)
Date: 20 Apr 2000 16:22:49 GMT
Subject: docstrings
References: <00bufsonasp7bgkf86lfb6sv5np5v8mqlj@4ax.com>
Message-ID: <8dnasp$iep$1@news.wrc.xerox.com>

Ben Wolfson <rumjuggler at home.com> writes:
> is there a way to use %c and %s within a docstring, or to add strings
> together in a docstring, so that if some strings are used in all
> docstrings, but might be changed, I can simply change one or two
> variables?

This might do what you want, although it's ugly (in the sense that it
serves the code-documentation - as opposed to user-documentation -
function less well):

    cat > spam.py
    """I am a docstring"""
    
    sharedoc = """ and I am a widely-reused docstring clause"""
    
    __doc__ = __doc__ + sharedoc
    
    del sharedoc
    ^D
    yngvi> python
    Python 1.5.2 (#7, May  6 1999, 14:39:45)  [GCC 2.8.1] on sunos5
    Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
    >>> import spam
    >>> print spam.__doc__
    I am a docstring and I am a widely-reused docstring clause
    >>> 

-- 
Mark Jackson - http://www.alumni.caltech.edu/~mjackson
	An advertiser will happily make you feel bad about
	yourself if that will make you buy, say, a Bic pen.
				- George Meyer




From wware at world.std.com  Sun Apr  9 22:29:19 2000
From: wware at world.std.com (Will Ware)
Date: Mon, 10 Apr 2000 02:29:19 GMT
Subject: A Mountain of Perl Books + Python Advocacy
References: <000f01bf9eca$2eec8c60$3e2d153f@tim> <38ECF4E6.B85C62D5@sightreader.com> <8cr4an$j40$2@newshost.accu.uu.nl>
Message-ID: <Fss48w.Aqs@world.std.com>

Martijn Faassen (m.faassen at vet.uu.nl) wrote:
> WRONGO! ALL UTHER LANGUZEZ SUCKZ!! PHYTHON IS THE BESTEST!!!!!1 FOR 
> EVERYTHING AND I HATEZ M$ TUU!! NOBUDDY ECSPEXTS THE SPANNISH ENQIUSITOAN
> HAHAHAHH! PHYTHON JOKE SPAM SPAM! <.75 WINK WINK NUDGE>!!!!!!!1

I had been searching for the perfect way to express my feelings for
Python. Your eloquence has brought an ineffable peace to my soul.
But I'd probably want to shorten it before I stick it in my .sig.
-- 
 - - - - - - - - - - - - - - - - - - - - - - - -
Resistance is futile. Capacitance is efficacious.
Will Ware	email:    wware @ world.std.com


From gmcm at hypernet.com  Thu Apr 27 09:23:52 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Thu, 27 Apr 2000 09:23:52 -0400
Subject: Import File Madness
In-Reply-To: <8e8ik5$t4v$1@nnrp1.deja.com>
Message-ID: <1255280657-66425781@hypernet.com>

jsolbrig at my-deja.com wrote:

> Several Questions/Comment on the python import statement
> 
> 1: Once a python application is large and various classes have
> static member objects of different types, import statements seem
> to have a similar logic to C include files. I.E., the question of
> what order to do the include becomes non-trivial. Is there anyway
> around this or ways to escape this?

I'm not understanding your first sentence.

import is easier than #include. For one thing, the first import of 
a module does all the work and installs it in sys.modules. 
Subsequent imports just get a reference from sys.modules.

For another, unless you're using "from ... import ..." you're 
effectively creating mount points, and changes wrought by 
other modules will be visible.
 
> 2: The import order/mechanism seems to be slightly different in
> different variations of python - WinNT vs. Unix

The search path (sys.path) is built differently, but the import 
mechanism is the same.
 
> 3: It would be nice to some equivalent of "forward definitions" -
> statements would let you construct a static element of type X
> before you have imported the file with X in it.

Ah, you are doing "from ... import ...". Don't do that.
 
> 4: There seem to simply be bugs in the import mechanism which can
> cause modules to be None after being imported intermitantly. Is
> this known, is that this some "feature"? Is this my imagination?
> It seems to happen in large applications - 200K source code.

Never seen it, never heard of it, unless you're using "from".

from Module import Variable

causes the current Module.Variable to be bound to the name 
Variable in your namespace. Subsequent assignments to 
Module.Variable outside your namespace won't be visible 
inside your namespace. They would, if you'd used "import 
Module" and referenced "Module.Variable".

- Gordon



From algaba at my-deja.com  Thu Apr 13 16:09:37 2000
From: algaba at my-deja.com (Manuel Gutierrez Algaba)
Date: Thu, 13 Apr 2000 20:09:37 GMT
Subject: Crosswords for Linux Backtracking recursive crossword generator
Message-ID: <8d59hk$6to$1@nnrp1.deja.com>

 Crosswords for Linux Backtracking recursive crossword generator
-------
Dear pythonistas,

I'm the current maintainer of Crosswords For Linux
http://www.ctv.es/USERS/irmina/cruo/cruolinux.htm

And I'm the process of releasing all the generators and filters.
I've finished the translation of the generator, which may be
interesting for many of you , it's only 400 lines code and
fast enough :

http://www.ctv.es/USERS/irmina/cruo/genaut.py

It's runnable, it doesn't need threads, nor other special libraries
nor data, just download and : python genaut.py

Comments/improvements are welcome!

--
-------------
Manolo


Sent via Deja.com http://www.deja.com/
Before you buy.


From ben at co.and.co  Mon Apr 17 22:22:41 2000
From: ben at co.and.co (ben at co.and.co)
Date: Tue, 18 Apr 2000 02:22:41 GMT
Subject: Scripting and Gnome and KDE
References: <Pine.GSO.4.10.10004150741520.11261-100000@sundial> <8d95m0$9j6$1@news1.xs4all.nl> <8damat$4ja$1@slb2.atl.mindspring.net> <8dbsgc$35o$1@news1.xs4all.nl> <38FB4336.2BC6C235@bioreason.com> <8dftlp$3s1$1@news1.xs4all.nl>
Message-ID: <RFPK4.106065$l35.216877@afrodite.telenet-ops.be>

Boudewijn Rempt <boud at rempt.xs4all.nl> wrote:
>>> Absolutely, but I don't know that an essay on the problems is exactly
>>> the right way of introducing the toolkit.
> 
>> That's probably true.  Given the goal of showing how to do non-obvious
>> things with a toolkit, what changes would you recommend?
> 
> Well, I'm not sure I saw any really non-obvious things - a treeview
> isn't that spectacular nowadays, and neither is layout management.

Layout managment was a relief after some MFC experience, and when I made
the mistake of choosing a 22pt default font in licq (Qt plugin), I
started seeing the merits of it. But then again, I'm not a seasoned Qt
developper.

>> it tells the reader how to solve a specific problem, but
>> doesn't provide a general method for discovering how to use other
>> parts of the toolkit.

After reading the tutorial on http://wwww.valdyas.org (hope I get it
right), I wondered how much PyQt and PyGtk resembled each other.
IMVVHO, a general purpose signal/slot-mechanism would be nice addition
to pyton.
 
> Besides, toolkits are different enough that migrating from one to
> another often demands a complete new way of thinking: that's something
> the article in Dr Dobbs did show: tkInter is one way of working,
> PyQt another, but PyGTK is completely different once again. That's
> an important insight, I feel.

There's still a huge problem with the various licences, as they don't
really match Python's license.

Greetings,
-- 
ben . de . rydt at pandora . be ------------------ your comments
http://users.pandora.be/bdr/ ------- inl. IPv6, Linux en Pandora



From grant at nowhere.  Tue Apr 18 12:57:24 2000
From: grant at nowhere. (Grant Edwards)
Date: Tue, 18 Apr 2000 16:57:24 GMT
Subject: Integer solutions to linear equation?
References: <slrn8fovo2.nq.grant@grante.comtrol.com> <m33dojcs60.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <slrn8fp4vj.16t.grant@grante.comtrol.com>

In article <m33dojcs60.fsf at atrus.jesus.cam.ac.uk>, Michael Hudson wrote:

>>      64x = 41y + 1
>>      
>> Where x and y are both integers.  After scratching our heads
>> for a while, we used the brute force approach:
>> 
>> for x in range(1,100):
>>     y = ((64*x)-1)/41
>>     if 64*x == 41*y+1:
>>         print (x,y)

>Use Euclid's algorithm; here's some code that does that:

You know, that looks vaguely familiar. I took numerical
analysis 20 years ago and haven't used it since.  I'll have to
do some reading on this (I've got a 20 hour plane trip coming
up soon, so I should have plenty of time).

-- 
Grant Edwards                   grante             Yow!  I request a weekend
                                  at               in Havana with Phil
                               visi.com            Silvers!


From dgoodger at bigfoot.com  Sun Apr 23 23:56:44 2000
From: dgoodger at bigfoot.com (David Goodger)
Date: Sun, 23 Apr 2000 23:56:44 -0400
Subject: Why doesn't O'Reilly use parrots on its python books?
References: <etdvh186ztw.fsf@w20-575-115.mit.edu>
Message-ID: <B5293E3C.255C%dgoodger@bigfoot.com>

How about half a bee?

If they ever run out of fauna and need to go flora, the only choice would
be: the larch.



From rcc at nospamthanks_jennic.com  Wed Apr 26 04:31:14 2000
From: rcc at nospamthanks_jennic.com (Robert Cragie)
Date: Wed, 26 Apr 2000 08:31:14 GMT
Subject: Unsigned integer arithmetic
References: <newscache$51sktf$rzc$1@jenpc07.jennic.co.uk> <PxvN4.1335$sf5.16905@news.corecomm.net>
Message-ID: <newscache$3s7mtf$gbg$1@jenpc07.jennic.co.uk>

Ray&Maria <nospam at nowhere.net> wrote in message
news:PxvN4.1335$sf5.16905 at news.corecomm.net...
| What's wrong with this:
|
| myInt = 0x7fffffff
| myLong = long(myInt)  #coerce to long
| myInt2 = int(myLong)  #coerce to int (if it fits)
|
| def printU32Hex(n):
|     if n > 0xffffffffL:
|         raise ValueError
|     print "0x%01x%07x" % (int(n >> 28), int(n & 0xfffffffL))
|
| printU32Hex(0x7fffffffL)
| printU32Hex(0xffffffffL)
| printU32Hex(0x100000000L)

Nothing at all in principle, but it would be nice to have a neater solution.
I guess I'm showing my C background here, but it would be nicer to have a
notion of unsigned numbers and also to be able to print longs.

I don't like this, as it implies that 'i' is a 32 bit signed integer, but
you don't have any control. It blows up here because it won't extend the
type to a long - as it's typeless, I don't see why it should be able to do
this automatically.

>>> i = 0x7fffffff
>>> i = i + 1
Traceback (innermost last):
  File "<pyshell#1>", line 1, in ?
    i = i + 1
OverflowError: integer addition

Typeless languages are all very well, but there are times when I really
would like to say 'this is a 32 bit unsigned integer - treat it as such'
instead of implied typing. The warning bells sounded a bit on p33 of
'Learning Python' when it suggested that "as a rule of thumb, if you find
yourself wanting to flip bits in Python, you should think long and hard
about which language you're really using". But there are more compelling
reasons why I am using Python than C for my particular application. I don't
particularly want to have to start writing C companion libraries if I can
avoid it.

Robert




From gbreed at cix.compulink.co.uk  Fri Apr 28 06:47:12 2000
From: gbreed at cix.compulink.co.uk (gbreed at cix.compulink.co.uk)
Date: 28 Apr 2000 10:47:12 GMT
Subject: Converting = to ==
References: <n01O4.12725$HG1.340768@nnrp1.uunet.ca>
Message-ID: <8ebq7g$lfr$1@plutonium.compulink.co.uk>

On Thu, 27 Apr 2000, Warren Postma wrote:

> I have an "end user query tool" that uses Python expressions, but I want
> comparisons to be able to be written either as A=B or A==B then I want 
> to
> compare them using Python's Eval function.
> 
> Problem is if you search and replace all = with == then users have to 
> use
> the single-equals convention, and some of them will be python 
> programmers,
> and will want to use the "non idiot mode" of the query tool.

That's not the only problem.  You'll also find != gets converted to !==, 
<= to <== and >= to >==.  This would not be the expected behaviour in "non 
 idiot mode".

> So i did this, but is there a better way?
> 
>     >>> string.replace( string.replace("a=b a==b", "=", "=="), 
> > "====","==")
>     a==b a==b

If you want to catch != etc as well, you'll need

s = string.replace("a=b a==b b!=c", "=", "==")
s = string.replace(s,"====","==")
s = string.replace(s,"!==","!=")
s = string.replace(s,"<==","<=")
s = string.replace(s,">==",">=")

Or you could use a regular expression.  I think this works:

s = re.sub("([^<>!=])=([^=])",r"\1==\2","a=b a==b b!=c")

Or the equivalent:

expr = re.compile("""
  ([^<>!=])  # check for no preceding <,>,! or =
  =          # matches = but not ==,<=, >= or !=
  ([^=])     # catch both parts of ==
  """, re.VERBOSE)

s = expr.sub(r"\1==\2","a=b a==b b!=c")

Which is clearer than a one-line regular expression, but longer than doing 
all the substitutions.  It may be more efficient if the expression is only 
compiled once, but executed many times.


From drew.mcdowell at msfc.nasa.gov  Thu Apr  6 17:19:32 2000
From: drew.mcdowell at msfc.nasa.gov (Andrew McDowell)
Date: 6 Apr 2000 21:19:32 GMT
Subject: Was Re: need wrapper for ssh/scp
Message-ID: <8civ14$rm7$1@hammer.msfc.nasa.gov>

Thomas Wouters <thomas at xs4all.net> wrote:
>ssh/scp (and indeed most programs that ask for a password) talk directly to
>the tty when getting sensitive information (like the password or the
>passphrase.) You can catch this by using the 'pty' module. You probably
want
>to pty.fork(), and then redirect stdin/stdout, and fiddle with the terminal
>on both ends to easily pass the passphrase... You face the usual deadlock
>problems though, when both ends of the pipe are waiting for the other to
>send data.

Umm....Help? :)

I also working with ssh to execute some remote commands on different
servers.  I'd previously been hobbling together this strange idea of
generating an Expect script in python then running it. (*bleh*)  So once I
saw was clued onto the pty module I started playing with it.  Well...I'm
either severely confused, or something isn't working properly.  Could anyone
offer some help?
Here's a sample of how I'm attempting it now:

#!/usr/local/bin/python

import os, string, pty, time, signal

def remote_ex(hostname, password):
  pid, fd = pty.fork()
  if pid == 0:
    os.execv("/usr/local/bin/ssh",[hostname])
  else:
    time.sleep(2)
    print "Child says %s" % string.strip(os.read(fd,1024))
    print "Child took %d password bytes." % os.write(fd, password)
    time.sleep(4) #being over cautious
    print "Child took %d command bytes" % os.write(fd, "echo $HOST >
ptyhost.txt")
    print "Child took %d exit bytes." % os.write(fd, "exit")
    os.kill(pid, signal.SIGKILL)
  return

#=-=-=-=-=-Main=-=-=-=-=-=-=#
def main():
  myhost = string.strip(raw_input("Enter a host:"))
  mypass = string.strip(raw_input("Enter your Password:"))
  remote_ex(myhost, mypass)

##########################

if __name__ == "__main__":
  main()




From quinn at mono.ugcs.caltech.edu  Mon Apr 10 20:44:34 2000
From: quinn at mono.ugcs.caltech.edu (Quinn Dunkan)
Date: 11 Apr 2000 00:44:34 GMT
Subject: Why should I switch to Python?
References: <seas5ia89177@corp.supernews.com> <slrn8f4d0j.1t0.bwinton@tor.dhs.org>
Message-ID: <slrn8f4tfr.6hh.quinn@mono.ugcs.caltech.edu>

On Mon, 10 Apr 2000 20:02:40 GMT, Blake Winton <bwinton at tor.dhs.org> wrote:
>On Fri, 31 Mar 2000 19:43:58 -0800, Aaron Turner wrote:
>>I've been coding Perl for over 3 years and really like it.  (Especially
>>it's syntax and ability to grok my programming style.)
>
>Now, just as a side note, why is it that the sets of The Perl Hackers I
>Know, and The People I Know Who Use The Word "Grok" are the same set?

I always suspected perl programmers were from Mars.  Now if we could figure
out where python programmers are from, someone could write a book and get
rich.

*runs and ducks*


From nospam.newton at gmx.li  Thu Apr  6 01:43:47 2000
From: nospam.newton at gmx.li (Philip 'Yes, that's my address' Newton)
Date: Thu, 06 Apr 2000 05:43:47 GMT
Subject: blessed be Guido, for "5 <= X <= 10" does what it should...
References: <3TpG4.3349$HG1.98231@nnrp1.uunet.ca>
Message-ID: <38ec201b.80500674@news.nikoma.de>

On Tue, 4 Apr 2000 13:42:41 -0500, "Warren Postma" <embed at geocities.com>
wrote:

>Hey, is the APL symbol set anywhere in Unicode!?

Of course :-)

In the "Miscellaneous Technical" Block, there's a subsection "APL" from
U+2336 to U+237A. HTH.

Cheers,
Philip
-- 
Philip Newton <nospam.newton at gmx.li>


From spamfranke at bigfoot.de  Wed Apr 26 19:26:21 2000
From: spamfranke at bigfoot.de (Stefan Franke)
Date: Wed, 26 Apr 2000 23:26:21 GMT
Subject: SPARK / Earley Parser question
Message-ID: <3907774d.32199072@news.btx.dtag.de>

I'm designing a grammar with John Aycock's SPARK toolkit. 

Earley's algorithm is said to be n^3 time bound in general, n^2 for
unambiguous grammars and linear for most practical programming
language grammars. 

Are there some rules of thumbs which kind of rules should be avoided
to retain linear complexity? 

In particular, I wonder if using empty right-hand sides for rules has
any disadvantages.

Online references would be appreciated.

Stefan



From effbot at telia.com  Thu Apr 13 13:02:42 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Thu, 13 Apr 2000 17:02:42 GMT
Subject: Complaint Dept: My shoes!
References: <8d4s4d$mvr$1@nnrp1.deja.com>
Message-ID: <S4nJ4.889$rc9.190032896@newsb.telia.net>

musingattheruins at my-deja.com wrote:
> if len(shoe)>0: print shoe, hit_on_head()

real python programmers use:

    if shoe:
        print shoe, hit_on_head()

for more information, read the fine manual; section 5.10 in
the language reference describes what's true and what's
false:

    http://www.python.org/doc/current/ref/lambda.html

</F>




From donb at init.com  Wed Apr  5 13:02:36 2000
From: donb at init.com (Donald Beaudry)
Date: Wed, 05 Apr 2000 13:02:36 -0400
Subject: Teaching the art of programming, in python 
References: <20000405053920.20579.qmail@web2102.mail.yahoo.com> <8cfe42$o02$1@nnrp1.deja.com> <dpJG4.3744$HG1.106895@nnrp1.uunet.ca> <8cfqf7$4ml$4@newsserver.rrzn.uni-hannover.de>
Message-ID: <200004051702.NAA04361@zippy.init.com>

In article <dpJG4.3744$HG1.106895 at nnrp1.uunet.ca>,
	"Warren Postma" <embed at geocities.com> writes:

> Speaking of Dummies, was there/is there a LISP for Dummies, or
> Smalltalk for Dummies?  Can Python for Dummies be far off?

You can find the answer to those and similar questions in "The
Complete Idiot's Guide to the Dummy Series"

--
Donald Beaudry                                     Ab Initio Software Corp.
                                                   201 Spring Street
donb at init.com                                      Lexington, MA 02421
                      ...Will hack for sushi...




From hamish_lawson at yahoo.co.uk  Sun Apr  9 12:19:26 2000
From: hamish_lawson at yahoo.co.uk (Hamish Lawson)
Date: Sun, 09 Apr 2000 09:19:26 -0700
Subject: Migrating Tcl system to Python
References: <095cc48a.7e9b36dd@usw-ex0108-063.remarq.com> <000C48B993BF96DA.95DEBE4215A79896.7F63D15FB99EE88D@lp.airnews.net>
Message-ID: <248819a8.00d5228e@usw-ex0106-045.remarq.com>

Cameron Laird wrote:

[Stuff about communicating between Tcl and Python.]

> Minotaur. CORBA. ILU. COM, maybe. Pipes. Named pipes.

The systems will be on Linux so COM is out but the rest are
possibilities. As far as I can see, one of the main issues in
communicating between Tcl and Python will be how to handle
passing lists and dictionaries (arrays in Tcl). XML-RPC manages
this automatically and although I don't have any experience with
either CORBA or ILU, I imagine they do too. If I were to use a
pipe approach, there would presumably be a need for a layer to
translate between Tcl's and Python's representations of data
structures; though I think this might be reasonably simple, are
there solutions already available, or issues I'd need to be
aware of?

Hamish Lawson


* Sent from RemarQ http://www.remarq.com The Internet's Discussion Network *
The fastest and easiest way to search and participate in Usenet - Free!



From gregm at iname.com  Thu Apr 27 08:25:13 2000
From: gregm at iname.com (Greg McFarlane)
Date: Thu, 27 Apr 2000 22:25:13 +1000
Subject: Radio-button in Tkinter
In-Reply-To: <39028096.24B972B4@swt.edu>; from ns56645 on 22 Apr 2000 at 11:48:22PM
References: <39028096.24B972B4@swt.edu>
Message-ID: <20000427222513.04343@nms.otc.telstra.com.au>

Shah,

This is a well-known feature of Tkinter.Variable - you must keep a
reference or the Tcl variable is destroyed.

In your case you will need to keep a reference to v1 and and v2.  If
the code snippet you supplied is from a method in a class, use self.v1
and self.v2 instead of just v1 and v2.  If the code is in a function,
you will have to make v1 and v2 globals.

Aside:  Why didn't you supply us with a snippet that we could run,
rather than just cut out a bit from your code?  There would be two
benefits to doing this:  First, myself and others would not have to
work out how to recreate the bits that you have left out (such as
frame3).  Secondly, if you had done what I did (and John Grayson, it
would appear) and run the code at the top level, rather than in a
function, you would have noticed that the problem went away - you
probably would have then realise what the cause was and would not have
had to post to the news group in the first place.

Moral - if you have code that does not behave as you expect, create
the absolute smallest runnable program that demonstrates the problem. 
If that does not lead you to the solution, post it.

Greg

On 22 Apr, ns56645 wrote:
> In the following code the radiobuttons that have been set are not seen
> as default selected in the output.
> For e.g         v1.set(3) # initialize should set the radiobutton with
> text 'BOTH' as default.
> 
> 
> What could be the wrong with the code?
> Please help?
> 
> Shah
> 
> 
> #######        Frame 3  Radio Buttons
> ##########################################################
> 
>         # Create the "Location" contents of the page.
>         group = Pmw.Group(frame3, tag_text = 'Location')
>         group.pack(fill = 'both', expand = 1, padx = 15, pady = 10)
> 
>         v1 = IntVar()
>         v1.set(3) # initialize
>         rb1 = Radiobutton(group.interior(), text = 'SWT', variable=v1,
> value=1)
>         rb1.grid(row = 0, column = 0)
>         rb2 = Radiobutton(group.interior(), text = 'WHS', variable=v1,
> value=2)
>         rb2.grid(row = 0, column = 1)
>         rb3 = Radiobutton(group.interior(), text = 'BOTH', variable=v1,
> value=3)
>         rb3.grid(row = 0, column = 2)
> 
> 
>         # Create the "Level" contents of the page.
>         group = Pmw.Group(frame3, tag_text = 'Level')
>         group.pack(fill = 'both', expand = 1, padx = 15, pady = 10)
> 
>         v2 = IntVar()
>         v2.set("13") # initialize
>         rb11 = Radiobutton(group.interior(), text = 'Grad', variable=v2,
> value=11)
>         rb11.grid(row = 0, column = 0)
>         rb12 = Radiobutton(group.interior(), text = 'UnderGrad',
> variable=v2, value=12)
>         rb12.grid(row = 0, column = 1)
>         rb13 = Radiobutton(group.interior(), text = 'BOTH', variable=v2,
> value=13)
>         rb13.grid(row = 0, column = 2)
> 
> 

-- 
Greg McFarlane     INMS Telstra Australia     gregm at iname.com



From ivanlan at callware.com  Thu Apr 13 00:30:59 2000
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Wed, 12 Apr 2000 22:30:59 -0600
Subject: need help just begining
References: <11822ca0.dd34a400@usw-ex0109-068.remarq.com> <38F41190.C0BB507B@sympatico.ca> <38F53E5A.977913FE@cosc.canterbury.ac.nz>
Message-ID: <38F54D83.B0F620A8@callware.com>

Hi All--

Greg Ewing wrote:
> 
> R E Heath wrote:
> 
> > SheaMon wrote:
> > >
> > > can
> > > I just leave the numbers and stuff like that and just edit
> > > the rest? I get so lost in the number part of the
> > > programming.
> 
[snip]

> I don't think that's necessarily good advice. If
> numbers turn you off, traditional BASIC is going to
> turn you off even more. You can do quite a lot of
> useful things in Python without ever going near a
> number -- something that can't be said about
> traditional BASIC!
> 
[snip]

> There are introductory books to Python around. I
> haven't had personal experience of any of them, but
> many people seem to think that "Learning Python"
> is reasonably good. Whether it's any less biased
> towards number-intensive examples I don't know.
> I'm sure others will offer advice on the matter.
> 

I'm holding in my hands the very first copy of _Teach Yourself Python in
24 Hours_.  I would like to recommend it to you, SheaMon, and I will,
with the caveat that there is a bit of number stuff in there.  However,
what's included isn't that hard, and I give pointers to websites where
you can do a little catchup.  The hardest "number stuff" in the book is
the Mandelbrot set, and Python makes the math *easy*.

I encourage you to take a look at it; there's plenty of info in there
for the compleat newbie to programming,  so don't dismiss it out of
hand.

Check the PSA Bookstore, or your local B&N or Borders if you want to see
it before you buy it.

<only-a-little-BSP>-ly y'rs,
Ivan
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
http://www.pauahtun.org and
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
Author:  Teach Yourself Python in 24 Hours



From aahz at netcom.com  Sun Apr 16 14:04:55 2000
From: aahz at netcom.com (Aahz Maruch)
Date: 16 Apr 2000 18:04:55 GMT
Subject: Python paradigms
References: <8cmr70$b0e$1@pegasus.csx.cam.ac.uk> <8dbdj3$dgt$1@slb2.atl.mindspring.net> <8dcpom$fbc$1@slb3.atl.mindspring.net> <slrn8fk2kh.965.scarblac-spamtrap@flits104-37.flits.rug.nl>
Message-ID: <8dcvc7$au2$1@nntp9.atl.mindspring.net>

In article <slrn8fk2kh.965.scarblac-spamtrap at flits104-37.flits.rug.nl>,
Remco Gerlich <scarblac-rt at pino.selwerd.nl> wrote:
>
>Pass functions?
>
>def exp_if(test, true, false):
>  if test: return true()
>  else: return false()
>  
>x = exp_if(a is not None, lambda a=a,i=i: a[i].weeble, lambda: 0)
>
>Getting silly, though :)

It's not so silly if true() and false() are coroutines or generators.
That way you can get rid of the lambda.  ;-)
--
                      --- Aahz (Copyright 2000 by aahz at netcom.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

"First one back from the bathroom gets to be the bottom."


From sholden at bellatlantic.net  Wed Apr  5 18:25:37 2000
From: sholden at bellatlantic.net (Steve Holden)
Date: Wed, 05 Apr 2000 22:25:37 GMT
Subject: Two questions
References: <38EA5FE8.853E9B75@callware.com> <38EA6508.90BF630C@roguewave.com> <38EA6E87.40EDB542@callware.com> <38EA6422.61054B11@exceptionalminds.com> <38EA7C18.8F3150D1@roguewave.com> <38EA87DD.3F24BEBF@exceptionalminds.com>
Message-ID: <38EBBC4A.C2F94F8F@bellatlantic.net>

Timothy Grant wrote:
> 
[ODBC DSN creation details snipped]
>
> OK, So I've created a system level DSN named TestDB and pointed it at an
> MDB file my server.
> 
> When I do the following in the interactive window...
> 
> import ODBC.Windows
> 
> db = ODBC.Windows.Connect('TestDB')
> 
> ...I get the following error...
> 
> NotSupportedError: ('IM001', 0, '[Microsoft][ODBC Driver Manager] Driver
> does not support this function', 4226)
> 
> What am I doing wrong?
> 
> --
> Stand Fast,
>     tjg.
> 
Here's an extract from a program I'm using to index email in an Access
database:

class SubIndex:
    """Implements indexing of messages by subject and various other keys."""

    def __init__(self):
	self.db = ODBC.Windows.connect("Mail")
	self.CR = self.db.cursor()

This seems to work fine.  Thought you might have a case error in your
code, but it also seems to work if I use

	self.db = ODBC.Windows.Connect("Mail")

so I'm not sure why you are getting this error.  UNLESS you have the database
open in Access as you work?  Might be a weird locking problem?  Can't manage
to reproduce it here, though.

So far I'm working with LOCAL FILES as databases.  There may be a problem
with accessing a file on a network share?  Hold on ... (copies 57 MB file
to server and reconfigures ODBC source) ... dammit, that works too.

Out of bright ideas.  Let me know what you find.  I'm using Windows 98,
with the 3.50.342800 driver.  I could try an NT client if that would
help.

regards
 Steve

--
"Bulding information systems just because it's fun."
Steve Holden  sholden at bellatlantic.net  703 716 7275


From aa8vb at yahoo.com  Fri Apr 14 13:39:55 2000
From: aa8vb at yahoo.com (Randall Hopper)
Date: Fri, 14 Apr 2000 13:39:55 -0400
Subject: Doesn't anybody write to paper anymore?
In-Reply-To: <etdg0spubme.fsf@w20-575-109.mit.edu>
References: <38F5FDF3.EE2FBA01@tcac.net> <etdg0spubme.fsf@w20-575-109.mit.edu>
Message-ID: <20000414133955.A5376@vislab.epa.gov>

Alex:
 |> I would like to pass it into the queue.  IO redirection in the shell
 |> is ugly and awkward.
 |
 |I don't know whether this counts as ugly and awkward IO redirection, but
 |perhaps you could do something like this:
 |
 |from os import popen
 |printer = popen ('lpr')
 |printer.write("page 1\n\r\f")
 |printer.write("page 2\n\r\f")
 |printer.close ()

If it doesn't work, try lp instead of lpr.
Try a test from the shell (command window) and see what works:

     echo "This is an lpr test" | lpr
     echo "This is an lp  test" | lp

Note that both assume you or someone has configured the printing service on
this box (possibly not the case for a home FreeBSD or Linux box).  If this
hasn't been done, search your OS documentation (printcap, lpd, and lpsched
are useful search strings).

-- 
Randall Hopper
aa8vb at yahoo.com



From glyph at twistedmatrix.com  Wed Apr 19 13:20:00 2000
From: glyph at twistedmatrix.com (Glyph Lefkowitz)
Date: 19 Apr 2000 12:20:00 -0500
Subject: Pickling Robustly
Message-ID: <jak8huf1of.fsf@helix.twistedmatrix.com>

Please forgive me if I've missed something obvious, and don't flame
too hard -- I've read the FAQs, and in general tried to do my
homework.  This is my first post on USENET.

(Apologies again if this shows up multiple times; my ISP doesn't provide netnews, and the server I managed to snag doesn't allow posting :-( )

I am currently writing a client-server application in Python.  The
server has significant state, in the form of a directed graph with
many cyclic and shared references.

I was OVERJOYED to discover Pickle.  The persistence format that I was
using before (in the slower, buggier, more obfuscated Java version of
this software) was rather naive, performed poorly, and imposed several
arbitrary and inconvenient restrictions on the type of data that was
allowable.  (Funny, I worked with Java for 2 years, and python for
only a few months, but I find myself saying things like "inconvenient
and arbitrary" about Java all the time now ... as well as things like
"slower, buggier, more obfuscated")

Pickle is everything I wanted, and more.  The copy_reg module makes it
moreso.  I am so thrilled I can barely contain myself.

Except...

It appears that Pickle is not terribly robust.  One of the major
necessary features of the software I'm writing is the ability of naive
programmers to write all sorts of arbitrary code: the more naive and
the more arbitrary the better.  Python makes this part so easy it
feels like I'm not even doing anything.  However, it also means I may
have to put all this code into restricted execution mode (and
therefore reduce its random arbitrarity (is that a word?)) to prevent
it from munging the whole of my persistent data.  I am also terrified
that if I try this approach I will miss some special case that will be
pointed out to me when I lose a day's worth of work because someone
assigned a data member to some "bad" type.

Due to the extremely self-referential nature of the data structures
I'm using, it is much easier to keep everything in one pickle (and
that effectively happens anyway, even if I don't try to, thanks to the
dependency resolution that pickle does). The problem is that pickle
(as far as I can tell, anyway) does not allow one to register error
handlers to keep track of when a problem occurrs.  If an unpicklable
datatype is attempted to pickle, EVERYTHING previously pickled will
suddenly vanish and no indication will be given about where the
offending data was.  If you've got a large data structure like I
anticipate, this means writing and loading code on-the-fly to attempt
a traversal of that structure to determine where the unpicklable type
is.

I am using the following code to make some things more amenable to
pickling (see below) but it's not nearly enough; I can't hope to catch
every datatype this way.  (Or can I?  Is it possible to define some
sort of catch-all type?)  If there is no way to do this with the
existing pickle, would it be possible to hack it in somewhere easily,
or shall I have to do my own persistance algorythm?

Note -- it is perfectly acceptible to me for the solution to this
problem to merely ignore those types which it does not persist.  All I
want to do is be able to explicitly say 'these attributes are
unpicklable', and have them fail to appear when the pickle is
re-loaded.  The more types persist correctly, the better, of course,
but that's not my main concern.

---cut here: pickleplus.py---


import copy_reg, types, new
from cPickle import *

def method_pickle(method):
	
	return method_unpickle, (method.im_func.__name__,
							 method.im_self,
							 method.im_class)

def method_unpickle(im_name,
					im_self,
					im_class):
	
	unbound=getattr(im_class,im_name)
	if im_self is None:
		return unbound
	bound=new.instancemethod(unbound.im_func,
							 im_self,
							 im_class)
	return bound

def function_pickle(function):
	
	function_name=function.__name__
	module_name=function.func_globals['__name__']
	return function_unpickle, (module_name,
							   function_name)


def function_unpickle(module_name,
					  function_name):

	module=__import__(module_name,
					  # none none because there is no context
					  # for this import, 1 because we want the
					  # subpackage (if this is one)
					  None,None,1)
	return getattr(module, function_name)


copy_reg.pickle(types.MethodType,
				method_pickle,
				method_unpickle)


---cut here---

Thanks very much for any help!

-- 
                  __________________________________________
                 |    ______      __   __  _____  _     _   |
                 |   |  ____ |      \_/   |_____] |_____|   |
                 |   |_____| |_____  |    |       |     |   |
                 |   @ t w i s t e d m a t r i x  . c o m   |
                 |   http://www.twistedmatrix.com/~glyph/   |
                 `__________________________________________'



From thomas at xs4all.net  Thu Apr  6 04:05:30 2000
From: thomas at xs4all.net (Thomas Wouters)
Date: Thu, 6 Apr 2000 10:05:30 +0200
Subject: CLI over Sockets?
In-Reply-To: <20000406002732.D1027@shell2.xcom.net>; from cdaniel@xcom.net on Thu, Apr 06, 2000 at 12:27:32AM -0400
References: <20000406002732.D1027@shell2.xcom.net>
Message-ID: <20000406100529.A13830@xs4all.nl>

On Thu, Apr 06, 2000 at 12:27:32AM -0400, Cliff Daniel wrote:

> Any such module exist that might help in implementing a
> CLI over sockets?  One such beast would preferrably operate in
> character mode and would would need to implement vt100  to provide
> the command line editing features (arrow keys, insert at cursor, etc).
> I'm know I'm asking for a lot but it's worth a shot in case someone
> has implemented this already.

I wrote a little thing like that once, to provide a front-end for our
Lucent (formerly Livingston) Portmasters. I needed a telnet that could log,
and log timestamps. (Portmasters have extraordinary debug information... all
the hexdumps of PPP packets and ISDN frames you want, but no timestamps.) I
shortly thought about adding command history & editing as well (which the
PMs lack too) but the tool had served my purpose and i haven't had use for
it since.

It's dead easy to write such a tool though: look at telnetlib for some good
examples. I ended up overriding the interact() method to add my logging, but
you still get all the protocol stuff for free. (telnet is more than just a
TCP socket, you see.) I'd guess that it isn't too much trouble to add
readline and your own completer function (if necessary, that last bit) but
i'm not sure if readline works properly with the select() method interact
uses... you might want to use mt_interact or something similar instead.

(Then again, i haven't looked at readline at all, partly because of it's
lack of documentation last time i looked, so it might be a lot harder than
that.)

I have my logging telnet somewhere on my laptop still, i could send it if
you want an example.

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From chafik at nevrax.com  Fri Apr  7 03:40:35 2000
From: chafik at nevrax.com (Sameh chafik pro)
Date: Fri, 7 Apr 2000 09:40:35 +0200
Subject: Some problem with python 1.6
Message-ID: <021b01bfa064$8fab3880$1001a8c0@nevrax.net>

I work with Visual C++ (6.0). When i build a project who have a python support i have the following error:
1) The library Python16_d.lid dos not  existe.
        I renamed in the config.h the pragama comment (lib,"python16_d.lib") to pragama comment (lib,"python16.lib"). But
        now i have the follow link error:
2) Linking...
        Creating library Debug/PYTHONLIB.lib and object Debug/PYTHONLIB.exp
    PyLib.obj : error LNK2001: unresolved external symbol __imp___Py_Dealloc
    PyLib.obj : error LNK2001: unresolved external symbol __imp___Py_RefTotal
    PyLib.obj : error LNK2001: unresolved external symbol __imp__Py_InitModule4TraceRefs
    PythonMethode.obj : error LNK2001: unresolved external symbol __imp__Py_InitModule4TraceRefs
    
    This link error happen only with the debug version, not that the difference between the debug and release comme from the         _DEBUG preprocessor.
    
.....................Thanks for answer.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20000407/a9095e8d/attachment.html>

From m.faassen at vet.uu.nl  Fri Apr 21 15:22:23 2000
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 21 Apr 2000 19:22:23 GMT
Subject: Using python on the web
References: <EDFD2A95EE7DD31187350090279C6767754BA8@THRESHER> <8dq6k3$qmb$1@nnrp1.deja.com>
Message-ID: <8dq9pf$pjg$3@newshost.accu.uu.nl>

spenrose at my-deja.com wrote:
> In article <EDFD2A95EE7DD31187350090279C6767754BA8 at THRESHER>,
>   Brent Fulgham <brent.fulgham at xpsystems.com> wrote:
>> A couple of questions:
>>
>> > We build medium-busy, CGI-intensive sites in Python. We were briefly
>> > quite excited about Zope, but once we realized it locked us into Yet
>> > Another Programming Language, we lost interest. DHTML is
>> > particularly
>> > problematic,
>> 1.  I'm not familiar with Zope.  Is DHTML specific to Zope?  It seems
>> to be proudly advertised on a lot of "webware" brochures I see these
>> days.

> My limited understanding of Zope is that to run Zope is to use DHTML and
> vice-versa. I welcome corrections on this point.

No, that's *DTML*, not DHTML. Confusing naming! DTML stands for 
"Document Template Markup Language" and is quite a different beast than
DHTML. See my other post.

Regards,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From m.faassen at vet.uu.nl  Tue Apr  4 07:07:44 2000
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 4 Apr 2000 11:07:44 GMT
Subject: Python 1.6 alpha 1 released
References: <Pine.GSO.4.10.10004040035420.21084-100000@sundial> <5E877FD7637AB46E.DABBED9CF3C1D64B.CDDDC05AE7374A78@lp.airnews.net> <l03102802b50f72e09280@[193.78.237.148]>
Message-ID: <8ccie0$7fs$1@newshost.accu.uu.nl>

Just van Rossum <just at letterror.com> wrote:
> At 8:48 AM +0000 04-04-2000, fredrik at pythonware.com wrote:
>>> Consider that *every one* of the protocol modules which offers
>>> a connect() method uses a two-argument signature for it:  ftplib,
>>> gopherlib, httplib, imaplib, nntplib, poplib, smtplib, telnetlib,
>>> urllib.  And they all *use* the two-argument connect when invoking
>>> their underlying socket!
>>
>>not true.  look again (doesn't matter if you look in the 1.5.2
>>library or the 1.6 library -- what you just said is not correct
>>in either case).

> Here's the result from a quick search through the Python 1.5.2 lib:

[snip internals using host, port arguments]

And what Tres also mentioned was the external interfaces of ftplib,
etc. A quick glance at the 1.6 alpha shows that they also have separate
host/port arguments. It would seem inconsistent to change 'connect' to use
tuples in one case, but 'connect' in ftplib a host and port argument..

Regards,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From kbd at iti-oh.com  Thu Apr 13 08:41:08 2000
From: kbd at iti-oh.com (tribune.oar.net)
Date: Thu, 13 Apr 2000 08:41:08 -0400
Subject: [Announce] CAD data access from Python
Message-ID: <8d4fec$7bm$1@tribune.oar.net>

After becoming hooked on Python 2 years ago, we have been developing code to
allow Python to directly access CAD data (with other related types in the
works).
We have just released our first product, CADScript, based on this
technology.  We are also releasing a new version of our CAD model quality
checker, CAD/IQ, that is written completely in CADScript.  The full press
release is below.

Note that we will be portions of this freely available to developers, more
details will be announced over the next several weeks.

I'd like to say thanks to Guido and everyone else in the Python community
who has made Python what it is today.  Thanks to Python, we are allowing a
*much* larger group of people to access data that, until now, has been only
accessible to people with a full blown C/C++ development environment.  We've
also significantly cut our internal development time/costs by embracing
Python and encourage others to take a serious look at Python.

Keith Davidson
Chief Technology Director
International TechneGroup Inc.

----------------------------------------------------------------------------
----------
International TechneGroup Incorporated (ITI) announces the release of
CADScript, a Python based development tool that significantly reduces the
time and effort spent developing and maintaining multi-CAD applications.

CADScript is based on the Python scripting language and provides modules
that allow access to native CAD data.  CADScript programs can be written
once and then run against any CAD API. The software is extremely valuable
for a vendor of downstream applications whose software complements multiple
CAD systems. CADScript serves as an enhancement for native CAD APIs and is
not intended to replace them. ITI's CADScript provides a consistent method
of accessing multiple CAD APIs for data and functionality. By utilizing the
Python language, CAD developers immediately have access to the powerful OS,
internet, and other modules included with Python.

ITI will be announcing an online community for CADScript developers in the
near future.

For more information visit http://www.iti-oh.com/pdi or contact Robert
Farrell at 800-783-9199, 513-576-3900, or info at iti-oh.com.







From aa8vb at yahoo.com  Mon Apr 17 14:02:33 2000
From: aa8vb at yahoo.com (Randall Hopper)
Date: Mon, 17 Apr 2000 14:02:33 -0400
Subject: comparing strings and ints
In-Reply-To: <20000412112935.A2446939@vislab.epa.gov>
References: <2YmI4.366$rc9.188014080@newsb.telia.net>
 <20000411134358.A2352304@vislab.epa.gov> <1256641282-42160637@hypernet.com>
 <20000412075812.A2449004@vislab.epa.gov>
 <jr%I4.643$rc9.190628352@newsb.telia.net>
 <20000412112935.A2446939@vislab.epa.gov>
Message-ID: <20000417140233.A319729@vislab.epa.gov>

Arg.  This just bit again (me this time), but with the max() function:

>>> max(3,'2')
'2'


I wish there'd have been a type exception here.  

For now, I guess just be very careful when parsing datafiles that you
convert all your numeric strings to some other type; you'll get no warning
if you don't.

-- 
Randall Hopper
aa8vb at yahoo.com



From mhammond at skippinet.com.au  Sun Apr  9 21:30:00 2000
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Mon, 10 Apr 2000 01:30:00 GMT
Subject: win32com and Reflection2
References: <WE8I4.100257$ra.2387838@news.easynews.com>
Message-ID: <s8aI4.48$UB.261@news-server.bigpond.net.au>

You should be able to run "makepy" over the reflection object (makepy will
present a list of objects).  You can find "makepy" on the Pythonwin
"tools" menu.

Once you have done that and created the Reflection object, the constants
should be available in win32com.client.constants

eg,

win32com.client.constants.SomeConstantName

Note that the Reflection constants wont appear until you have created the
reflection object using win32com.client.Dispatch()

Mark.

"Roger Upole" <rupole at compaq.net> wrote in message
news:WE8I4.100257$ra.2387838 at news.easynews.com...
>    I'm scripting Reflection2 sessions with win32com.client.Dispatch.
> However, there are some constants (error codes, transmission modes,
> etc) in a typelib that I can't seem to get at from the Dispatch object.
> Anybody know of a way to access these from Python ?
>    If not, is there any way to get to them from C++ ?
> I have very little experience with C, but if it is simple I can
> probably blunder my way through it.
>              Roger Upole
>
>
>




From jsolbrig at my-deja.com  Thu Apr 27 01:19:10 2000
From: jsolbrig at my-deja.com (jsolbrig at my-deja.com)
Date: Thu, 27 Apr 2000 05:19:10 GMT
Subject: Import File Madness
Message-ID: <8e8ik5$t4v$1@nnrp1.deja.com>

Several Questions/Comment on the python import statement

1: Once a python application is large and various classes have static member
objects of different types, import statements seem to have a similar logic to
C include files. I.E., the question of what order to do the include becomes
non-trivial. Is there anyway around this or ways to escape this?

2: The import order/mechanism seems to be slightly different in different
variations of python - WinNT vs. Unix

3: It would be nice to some equivalent of "forward definitions" - statements
would let you construct a static element of type X before you have imported
the file with X in it.

4: There seem to simply be bugs in the import mechanism which can cause
modules to be None after being imported intermitantly. Is this known, is that
this some "feature"? Is this my imagination? It seems to happen in large
applications - 200K source code.

Joe Solbrig
jsolbrig at webcom.com



Sent via Deja.com http://www.deja.com/
Before you buy.


From schmertz at wam.umd.edu  Tue Apr 18 11:41:59 2000
From: schmertz at wam.umd.edu (Bob Schmertz)
Date: Tue, 18 Apr 2000 11:41:59 -0400
Subject: Project idea - python
In-Reply-To: <38FC0E72.9AB025DB@webwitches.com>
References: <BxGK4.811$9B.103178@news.bc.tac.net> <38FC0E72.9AB025DB@webwitches.com>
Message-ID: <Pine.GSO.4.21.0004181113110.22094-100000@sun21epsl.wam.umd.edu>

On Tue, 18 Apr 2000, Shae Erisson wrote:

>Parag Shah wrote:
>> 
>> hi,
>>    i am planning to do a project for my bachelors where i wish to write a translator for a language (undecided at
>> the moment ) which generates java bytecode. how do i go about doing it???
>> thanx in advance
>> cheers

This is almost exactly what I'm doing for a class, only we're actually
using Java.

>
>read up on jpython, it uses JavaCC.
>
>http://www.jpython.org/docs/compile.html

You may also want to use the following:

Plex - essentially a Python version of Lex, the most popular lexical
analyzer (http://www.cosc.canterbury.ac.nz/~greg/python/Plex/)

SPARK - sounds like a Python implementation of Yacc/Bison, though from
what I can tell it doesn't handle complex grammars such as LR
(http://www.csr.UVic.CA/~aycock/python/)

JavaClass - OK, I haven't had a chance to look at this, and since it's in
and about Java, I'm not sure if this is actually going to help.  Anyway,
this seems to be related to generating or manipulating Java bytecode.

Is this the sort of help you were looking for?

--
Cheers,
Bob Schmertz






From dan at cgsoftware.com  Wed Apr 12 19:11:31 2000
From: dan at cgsoftware.com (Daniel Berlin+list.python)
Date: 12 Apr 2000 19:11:31 -0400
Subject: Python/Perl Popularity (Re: A Mountain of Perl...)
In-Reply-To: pj@sgi.com's message of "12 Apr 2000 22:50:26 GMT"
References: <200004102111.HAA03380@envy.fulcrum.com.au> <38F4996F.E3BD2626@acm.org> <8d2p3q$2vj9o$1@fido.engr.sgi.com> <snwrf1ls.fsf@dan.resnet.rochester.edu> <8d2uji$301cu$1@fido.engr.sgi.com>
Message-ID: <d7nugbj0.fsf@dan.resnet.rochester.edu>

pj at sgi.com (Paul Jackson) writes:


http://artax.karlin.mff.cuni.cz/~mikulas/vyplody/links/

> |> Nope, both lynx, and links, exist.
> 
> oh - could you offer a link to "links" -- it's
> a hard name to search for -- too many false hits.
> 



From helge.hess at mdlink.de  Tue Apr 11 13:14:25 2000
From: helge.hess at mdlink.de (Helge Hess)
Date: Tue, 11 Apr 2000 19:14:25 +0200
Subject: [ANNOUNCE] Garbage collection for Python
References: <20000407221750.A29279@acs.ucalgary.ca> <38F13FB9.D1D16AF@mindspring.com> <slrn8f2if1.dop.nascheme@cranky.arctrix.com> <87r9cdhdb0.fsf@deneb.cygnus.argh.org> <FsunGI.BtB@world.std.com> <38F319A4.89621192@mdlink.de> <38F351B0.45E137AA@roguewave.com>
Message-ID: <38F35D71.4DD20040@mdlink.de>

Bjorn Pettersen wrote:
> Helge Hess wrote:
> > I think what he wants to do (me too ;-) is to run several copies of the
> > interpreter in *one* process, possibly in multiple threads.
> > But I'm not sure whether this is a GC issue though (RC isn't thread-safe
> > anyway, right ?).
> 
> Well, it actually can be -- at least on Windows :-)

The question is, what overhead is involved here. If these Windows
functions wrap the RC ops in locks you have gained nothing (wrapping RCs
in locks breaks performance in a big way). Note that std-RC is often
only one CPU op (incr,decr).

Also note that not only the ++ or -- needs to be protected, but the
'==0' check as well. Eg:

   [task1]            [task2]
   dec(obj);          inc(obj);
could be
   [task1]	      [task2]
   obj->rc--;	      
   if (obj->rc==0)
     <task-switch>    obj->rc++;
     free(obj);       <task-switch>
   object-freed	      assumes to have a valid obj

[note that checking for 0 in task2 doesn't solve the problem, since the
check wouldn't be atomic again, but maybe I'm missing something].

A (m&s) GC is somewhat in advantage here, since it is run only at
specified (safe) points, so there is no locking required at such high
frequency.

Helge



From a.eyre at optichrome.com  Tue Apr  4 09:40:33 2000
From: a.eyre at optichrome.com (Adrian Eyre)
Date: Tue, 4 Apr 2000 14:40:33 +0100
Subject: Converting a hex string to an integer
In-Reply-To: <UsRE4.4303$74.68121@newsc.telia.net>
Message-ID: <000c01bf9e3b$596e1970$3acbd9c2@optichrome.com>

> >>> import string
> >>> string.atoi("0xab", 0)
> 171
> 
> or in 1.6:
> 
> >>> int("0xab", 0)
> 171

or (probably available since v1.0):

>>> eval("0xab")
171

#include <UsualEvalWarnings.h>

-----------------------------------------------------------------
Adrian Eyre <a.eyre at optichrome.com> - http://www.optichrome.com 




From effbot at telia.com  Sat Apr  1 11:58:30 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Sat, 01 Apr 2000 16:58:30 GMT
Subject: Paid to work in Python? (was: Re: Choice Of Language)
References: <Uqvw4.2261$yL2.28868@typhoon.southeast.rr.com> <8ajmib$elo$1@nnrp1.deja.com> <8c3ats$b5f$1@nntp.Stanford.EDU> <slrn8eact2.sq7.nobody@bambi.visi.com>
Message-ID: <WUpF4.4527$74.72573@newsc.telia.net>

Grant Edwards wrote:
>
> >I should post Google's Python-related job openings to
> >          http://www.python.org/Jobs.html
>
> Pointless tanget: Is it just me, or is Google really the last
> usable search engine?

I just love their new MentalPlex search feature!

(doesn't work in eff-bot mode, though -- I keep
getting "error 008: remove aluminum foil")

</F>




From kragen at dnaco.net  Sun Apr 23 01:48:58 2000
From: kragen at dnaco.net (Kragen Sitaker)
Date: Sun, 23 Apr 2000 05:48:58 GMT
Subject: Distribution
References: <39013f48@flexnet239.uunt.net> <B527F39C.24A7%dgoodger@bigfoot.com>
Message-ID: <e9wM4.1506$RM6.3645144@news-east.usenetserver.com>

In article <B527F39C.24A7%dgoodger at bigfoot.com>,
David Goodger  <dgoodger at bigfoot.com> wrote:
>on 2000-04-22 01:57, Highdesertman (HighdesertmanREMOVE at yahoo.com) wrote:
>><stuff>
>The user must have Python installed. But since it's absolutely free with no
>strings attached whatsoever, and multi-platform to boot, that's not a
>problem.

Maybe you should go back and read the Tanenbaum vs. Torvalds debate
about why allowing redistribution of patches makes Minix unviable.
It's the same debate.

I just built Python on NeXTSTEP 3.3 a few weeks ago.  It took me less
than an hour to get it installed and working properly (as far as it's
possible for it to work properly.)  An hour is a lot of work to install
a program.

>> In addition, one of the real advantages of the
>> development of the compiler was that it gave your code a level of protection
>> against being tampered with and made it easily useable by a wide audience.
>
>One of the greatest advantages of interpreted languages in general, and
>Python in particular, is that their programs can be run under multiple
>operating systems (if they are written correctly, of course). You are
>running Win95. I run MacOS. Bruce runs Linux. We can all run the same Python
>program. Result: even wider audience!

Well, platform-independent bytecode is platform-independent, too.
Python does have .pyc code, which can be run without the .py file it
was compiled from.  I don't know how platform-independent it is; my
guess is "very".

That said, one of the greatest advantages of interpreted languages is
that the user can "tamper" with the code easily.

I'd be nervous about trying to run code from someone who thought they
had the right to "protect" the code telling *my* machine what to do *on
my behalf* from me knowing what it did or changing it to suit my
fancy.

>If someone really wants to get into (decompile, reverse engineer) a compiled
>program (.exe), they can. It takes some work to do this, but compiling is
>not secure encryption.

You can't have secure encryption if you're going to run your code on a
machine under someone else's control.  If it were securely encrypted,
their machine wouldn't be able to read it in order to run it.  :)
-- 
<kragen at pobox.com>       Kragen Sitaker     <http://www.pobox.com/~kragen/>
The Internet stock bubble didn't burst on 1999-11-08.  Hurrah!
<URL:http://www.pobox.com/~kragen/bubble.html>
The power didn't go out on 2000-01-01 either.  :)


From amused at webamused.com  Fri Apr 28 04:43:03 2000
From: amused at webamused.com (Joshua Macy)
Date: Fri, 28 Apr 2000 08:43:03 GMT
Subject: file open() problem
References: <8ebcm3$lo$1@nnrp1.deja.com>
Message-ID: <390941CF.BFB93167@webamused.com>

 That's the right syntax, and it works for me (admittedly, on Linux,
Win95, and WinNT, not Win98)


Python 1.5.2 (#1, Sep 17 1999, 20:15:36)  [GCC egcs-2.91.66
19990314/Linux (egcs- on linux-i386
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> f = open('file.txt', 'r')
Traceback (innermost last):
  File "<stdin>", line 1, in ?
IOError: [Errno 2] No such file or directory: 'file.txt'
>>> f = open('file.txt', 'w')
>>> f.write('Spam, Eggs, and Spam')
>>> f.close()
>>> f = open('file.txt', 'r')
>>> r = f.readlines()
>>> print r
['Spam, Eggs, and Spam']
>>> 


 Have you done anything that might have hidden the builtin "open" with a
different one?  Is this the first thing that you're doing at the >>>
prompt, or have you entered some "from blah import *" lines first?


  Joshua

ekw1 at my-deja.com wrote:
> 
> I am using python1.5.2 on windows98se.  all of the following gives me
> an error.
> 
> f = open('file.txt', 'r')
> f = open("file.txt", "r")
> f = open('c:/dir/file.txt', 'r')
> f = open('c:\dir\file.txt', 'r')
> f = open("c:/dir/file.txt", "r")
> f = open("c:\dir\file.txt", "r")
> f = open("c:\\dir\\file.txt", "r")
> 
> The error from the python shell for any of the above is:
> 
> Traceback (innermost last):
>   File "<pyshell#17>", line 1, in ?
>     f = open('file.txt', 'r')
> TypeError: illegal argument type for built-in operation
> 
> Can anyone help me?  Thanks.
> 
> ekw
> 
> Sent via Deja.com http://www.deja.com/
> Before you buy.


From ivanlan at callware.com  Mon Apr 17 19:05:01 2000
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Mon, 17 Apr 2000 17:05:01 -0600
Subject: TKMessageBox, and TKINTER console?
References: <rNHvjEdhm5Pp-pn2-Y8HWP45V5PG0@d185d18f2.rochester.rr.com>
Message-ID: <38FB989C.C8AECBD8@callware.com>

Hi All--

Benjamin Schollnick wrote:
> 
> Folks,
> 

[bobbitry]

>         I've tried a variety of different things to get rid of it...For
> example, calling
> junk= TK.tk() to a variable, and passing it as "master=junk", or as
> "parent=junk"
> in the code...
> 
>         I'm not sure if there is anyway to get rid of that darn console.
> 
>         Anyone got some suggestions for a really confused first time
> Tkinter'er?
> 

root=Tk()
...blah,blah,blah
...build your popups here....
root.withdraw()
root.mainloop()


<or-you-could-put-something-useful-in-that-blank-space>-ly y'rs,
Ivan
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
http://www.pauahtun.org 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
Author:  Teach Yourself Python in 24 Hours



From gmcm at hypernet.com  Fri Apr 28 00:02:58 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Fri, 28 Apr 2000 00:02:58 -0400
Subject: The Simple Economics of Open Source
In-Reply-To: <8eavgs$ker$1@news.udel.edu>
Message-ID: <1255227918-2015166@hypernet.com>

Terry Reedy wrote:

> 
> "Gordon McMillan" <gmcm at hypernet.com> wrote in message
> news:1255272357-
> 
> > One of those open source delusions is that being open source
> > somehow makes a difference to the end consumer. Bunk.
> 
> Whether most consumers yet know it or not, there is all the
> difference in the world. 

Of course there is; all your points are correct. But no end user 
will read the source and discover the bug / trojan horse / 
whatever. It's another developer that will.

My point was that to the end user, those files are a waste of 
disk space. They get value from the fact that it's open 
sourced, not from the source.

- Gordon



From ivanlan at callware.com  Mon Apr 17 16:47:30 2000
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Mon, 17 Apr 2000 14:47:30 -0600
Subject: Playing wave files on Windows
References: <38FB6ED7.4C72773E@callware.com> <38FB7624.943A8C35@rubic.com>
Message-ID: <38FB7862.7B15633@callware.com>

Hi All--

Jeff Bauer wrote:
> 
> > Ivan Van Laningham wrote:
> > How can I play a sound file on Windows.
> 
> Ivan, use the built-in winsound module, e.g.:
> 
> >>> import winsound
> >>> winsound.PlaySound(r'C:\WINNT\Media\tada.wav', winsound.SND_FILENAME)
> 

Yup, about 2 minutes after I posted the query, I found the documentation
for that.  I still think it should be a FAQ entry. ...

> > I don't want to do anything else, just play a
> > wave file. ...
> 
> Then I can't interest you in my Python interface to
> the Diamond Rio MP3 player?  <0.5 wink>
>

Not unless everyone that I'll be distributing my application to will
have one plugged in and ready to go;-)

<since-i-don't-own-one-the-maximal-set-is-everyone-minus-one>-ly y'rs,
Ivan;-)
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
http://www.pauahtun.org 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
Author:  Teach Yourself Python in 24 Hours



From ppetru at coltronix.com  Thu Apr 13 15:08:29 2000
From: ppetru at coltronix.com (Petru Paler)
Date: Thu, 13 Apr 2000 22:08:29 +0300
Subject: Simple formatting string question
In-Reply-To: <meh9-637743.14563813042000@news.cit.cornell.edu>; from meh9@cornell.edu on Thu, Apr 13, 2000 at 02:56:38PM -0400
References: <meh9-637743.14563813042000@news.cit.cornell.edu>
Message-ID: <20000413220829.C7824@coltronix.com>

On Thu, Apr 13, 2000 at 02:56:38PM -0400, Matthew Hirsch wrote:
> I'm trying to print something like this:
> 
> >>> print 6*'%2i' '%1i' % tuple([5,5,5,5,5,5,1])
> 
> but I'm getting an error.
> 
> Do you know what the correct syntax should be?

print (6*'%2i'+'%1i') % tuple([5,5,5,5,5,5,1])

-Petru



From slhath at flash.net  Thu Apr 20 13:48:29 2000
From: slhath at flash.net (Scott Hathaway)
Date: Thu, 20 Apr 2000 17:48:29 GMT
Subject: editor for win32
Message-ID: <NpHL4.779$fK2.113609@news.flash.net>

Can anyone recommend a python editor for Windows?

If there is an IDE for windows (other than IDLE), even better.

Thanks,
Scott




From effbot at telia.com  Wed Apr 12 09:48:04 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Wed, 12 Apr 2000 13:48:04 GMT
Subject: Huh? func_defaults, default values in function calls
References: <8cv5qj$35p$1@zinc.intern.netconnect.no><8cv8ra$hio$1@pegasus.csx.cam.ac.uk> <nMGI4.42394$U4.182659@news1.rdc1.az.home. <38F3EBBF.7B23@marshall.edu>
Message-ID: <o8%I4.642$rc9.189485056@newsb.telia.net>

Mark Hathaway <hathawa2 at marshall.edu> wrote:
> >> As a result, if you set a mutable object (list, dictionary, etc) as a
> >> default, it'll be set when the function is defined and then if the
> >> function changes it, it'll stay changed for subsequent invokations.
> >> The function effectively has a "side-effect", of modifying it's own
> >> default!
>
> This way you'll never really know what the default is going to be.

uhuh?  so you mean evaluating the default value every time the
function is called would automatically make it easier to know what
the default is going to be?  what about non-trivial default values?
what about scoping?  what about side effects? what about per-
formance?  etc.

> Ha ha ha ha. that's pathetic.

really?  I'm pretty sure Guido designed things this way on purpose,
but maybe you can come up with a better way to do it?

</F>




From claird at starbase.neosoft.com  Tue Apr  4 15:46:35 2000
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 4 Apr 2000 14:46:35 -0500
Subject: Why should I switch to Python? - Infinity of Primes
References: <NDBBKEGCNONMNKGDINPFGEJDDFAA.infonuovo@email.com> <8ccoug$f4o$1@pegasus.csx.cam.ac.uk> <7F51E902E09F3F60.8F612607812ED16B.82CB8B41435E14FD@lp.airnews.net> <38EA369C.26A1C25@math.okstate.edu>
Message-ID: <467ED74C3A44BDFC.B66525D52208472C.27972C76E272D139@lp.airnews.net>

In article <38EA369C.26A1C25 at math.okstate.edu>,
David C. Ullrich <ullrich at math.okstate.edu> wrote:
			.
			.
			.
>    There's nothing non-constructive about the traditional
>proof of the infinitude of the sequence of primes - given
>a sequence of primes it _constructs_ a prime not on the
>list. (Of course it doesn't actually return infinitely many
			.
			.
			.
The constructivist sect most familiar to me says, sure,
"it constructs a prime not on the list", but that's sub-
ordinated in the proof to a *reductio ad absurdum*, which
is inherently non-constructive.  That's part of the pro-
gram:  rewriting such formulations in purely
constructive language.  Or am I myself speaking entirely
in anachronisms?  Has modern constructivism somehow come
to rely on the Excluded Middle (something like the Roman
Church and monotheletism)?
-- 

Cameron Laird <claird at NeoSoft.com>
Business:  http://www.Phaseit.net
Personal:  http://starbase.neosoft.com/~claird/home.html


From jeremy at cnri.reston.va.us  Thu Apr 27 19:41:53 2000
From: jeremy at cnri.reston.va.us (Jeremy Hylton)
Date: Thu, 27 Apr 2000 19:41:53 -0400 (EDT)
Subject: [ANNOUNCE] Pisces 1.0a1 release
Message-ID: <14600.53313.621205.524349@goon.cnri.reston.va.us>

I'm pleased to announce the first alpha release of Pisces, a Python
implementation of the SPKI/SDSI Certificate standard.

http://www.cnri.reston.va.us/software/pisces/

This release, 1.0 alpha 1, contains a mostly complete implementation
of the SPKI/SDSI protocols.  It is missing a few features, like
threshold subjects.  I hope to include these in the final 1.0 release.

It also contains:

- spkitool.py, a command-line utility for managing keys, certificates,
and signatures in the style of PGP or GPG,

- pyarrowd.py, a Yarrow-160 random number generator for Linux and
Solaris, 

and

- two demonstration applications, one implementing access controls on
method calls of a Python instance and one implementing a client-server
secure socket protocol in the style of TLS.

I plan to issue a few more alpha and beta releases next month.  Any
comments, suggestions, or questions are welcome.

-- Jeremy Hylton <http://www.python.org/~jeremy/>




From spex66 at my-deja.com  Wed Apr 26 10:54:41 2000
From: spex66 at my-deja.com (spex66 at my-deja.com)
Date: Wed, 26 Apr 2000 14:54:41 GMT
Subject: Namespace Problem with global declaration in module
Message-ID: <8e6vv9$4n8$1@nnrp1.deja.com>

Hi,
it seemed very simple but don't worked like expected:

#FILE: test.py
dd = None #INIT a shadow instance for automated access
class zz:
    #ONLY one class will be instantiated
    def __init__(self):
        self.name = 'hallo'
        global dd
        dd = self
#END OF FILE

>>> from test import *
>>> yy = zz()
>>> dd.name
    ...
    AttributeError
    ...

++++++++++++++
the same code interactive at the python console works perfect!
>>> ...same as above...
>>> dd.name
    'hallo'

any hints?
in use: 1.5.2 / idle0.5
thanks

Peter
(=PA=)



Sent via Deja.com http://www.deja.com/
Before you buy.


From cfelling at iae.nl  Fri Apr 14 20:29:59 2000
From: cfelling at iae.nl (Carel Fellinger)
Date: 15 Apr 2000 02:29:59 +0200
Subject: reduce versus recursion? what am I missing...
References: <YznH4.4508$HG1.128287@nnrp1.uunet.ca>
Message-ID: <8d8d67$ii$1@vvs.superst.iae.nl>

Warren Postma <embed at geocities.com> wrote:
> # I am trying to figure out why function d (recursive) and e (reduce)
> # DON'T do the SAME thing:

> def d(*tup):
>     "why doesn't this work?"
>     if len(tup)>1:
>         return d(tup[1:])+abs(tup[0]-tup[1])

Wel to start with, why not use a loop?

def m(*tup):
    r = 0 # mind you, you 'seeded' the reduce with 0!
    for t in tup:
	r = abs(r-t)
    return r

You see fast, clear and crisp and no surprises here:)
But if you insist on recursion...

def d(*tup):
    # don't forget the 'seed' value for the recursive 'reduce'
    print "why, it works!",
    return d(0, tup)

def dd(r, tup):
    if len(tup)>0:
	# return d(tup[1:])+abs(tup[0]-tup[1])
	return dd(abs(r-tup[0]), tup[1:])
    else:
        return r

-- 
groetjes, carel


From felixt at dicksonstreet.com  Fri Apr 28 00:39:48 2000
From: felixt at dicksonstreet.com (Felix Thibault)
Date: Thu, 27 Apr 2000 23:39:48 -0500
Subject: The Simple Economics of Open Source
In-Reply-To: <24042000.3@sanctum.jae.ddns.org>
References: <LHEoAJAXWFB5EwF6@jessikat.demon.co.uk>
 <B527E7A5.51AA%raffael@mediaone.net>
 <3903316C.5D6F1611@tcac.net>
 <raffael-34658F.23400223042000@news.ne.mediaone.net>
 <390451C0.4B5D3196@tcac.net>
 <LHEoAJAXWFB5EwF6@jessikat.demon.co.uk>
Message-ID: <3.0.5.32.20000427233948.0093f530@mail.dicksonstreet.com>

At 23:57 4/24/00 +0200, Juergen A. Erhard wrote:
>-----BEGIN PGP SIGNED MESSAGE-----
>Hash: SHA1
>
>>>>>> "Robin" == Robin Becker <robin at jessikat.demon.co.uk> writes:
>
>    Robin> In article <390451C0.4B5D3196 at tcac.net>, Albert Wagner
>    Robin> <alwagner at tcac.net> writes
>    >> You STILL don't understand.  Artifice can only copy.  It cannot
invent.
>    >> An end product can be copied, even a style can be copied, but not the
>    >> creative process itself.
>    >> 
>    >> Raffael Cavallaro wrote:
>    Robin> ...
>
>    Robin> This position is like the weak AI position; you assert
>    Robin> there's something magical about creation. I think A Turing
>    Robin> was right; any human behaviour can be simulated and the
>    Robin> best simulators (people) are around already.
>
>Fundamental question is: Are we just a complex arrangement of
>bio-chemical (and thus ultimately physical) processes... or is there
>more to being `human beings'?
>
>If we are just a collection of bio-chemical processes, it's simple to
>`simulate' or replicate us in machine form[1].

<rant mode='ax-grinding'>

No it's not. Even when we're looking at (what seem to us to be) simple
processes like a solution of a pure compound in a pure solvent, a 
lot of the data we're interested in (solubility, acidity) has to be
empirically determined.

Let me underline this point:
       a. we _know_ (as much as we know anything) that chemicals
          are just part of the material world, and the domain of
          chemistry (atoms, molecules, solutions) is built of the
          domain of physics (dynamics, masses, waves)

       b. still, we can't reduce the behavior of a solution to a
          set of fudge-factorless equations.

So I don't see that there's any reason to believe that our being all
biochemical implies we are simple to simulate.

</rant>

Just-because-we-may-not-be-what-we-think-we-are-doesn't-mean-we-
aren't-what-we've-always-been-ly yrs,

Felix

>
>If we are more... than it will not be possible.
>
>Since I'm a christian, I guess it's obvious where I'll put my bets...
>
>Bye, J
>
>[1] I'm just reading Bill Joy's Wired essay `Why the future doesn't
>need us'...
>
>PS: For the record, I'm 100% with Albert (though maybe he's not 100%
>with me ;-)
>
>- -- 
>J?rgen A. Erhard      eMail: jae at ilk.de      phone: (GERMANY) 0721 27326
>     MARS: http://members.tripod.com/Juergen_Erhard/mars_index.html
>                    "Ever wonder why the SAME PEOPLE
>      make up ALL the conspiracy theories?" -- Michael K. Johnson
>-----BEGIN PGP SIGNATURE-----
>Version: GnuPG v1.0.1 (GNU/Linux)
>Comment: Use Mailcrypt and GnuPG <http://www.gnupg.org/>
>
>iEYEARECAAYFAjkEw1oACgkQN0B+CS56qs09zQCfTMfxscQZVhcLpqgIiSVYShoO
>vjwAn0G8CdZImNdfPk1s5mvgOjR7GxVu
>=W73r
>-----END PGP SIGNATURE-----
>
>-- 
>http://www.python.org/mailman/listinfo/python-list
>
>




From lefebvre at medz.mcgill.ca  Thu Apr 20 10:07:09 2000
From: lefebvre at medz.mcgill.ca (Luc Lefebvre)
Date: Thu, 20 Apr 2000 14:07:09 GMT
Subject: Mac Python Questions
References: <IRjL4.271$CP3.3183@news.uswest.net> <38FE8032.DE233C29@cosc.canterbury.ac.nz>
Message-ID: <lefebvre-2004001006310001@landru.medcor.mcgill.ca>

In article <38FE8032.DE233C29 at cosc.canterbury.ac.nz>, Greg Ewing
<greg at cosc.canterbury.ac.nz> wrote:

> Jeff Massung wrote:
> > 
> > b) No matter what I do, I can't get Tkinter to work :(.
> 
> Neither can I, and from what I've seen here, neither
> can anyone else. Tkinter just seems to be broken on
> the Mac at present.

I managed to have a look at this (demos and some test code) and it seems
to work for me on a superficial level.  I have MacPython 1.5.2 with
MacOS8.1 on a beige PowerMac G3.

Rather a python newbie though (more experience with Tcl/Tk).

-- 
***     Luc Lefebvre
NOTE:   if your reply via email
        please remove the z from my return address
***


From jayFreem at direcpc.com  Fri Apr 14 19:00:25 2000
From: jayFreem at direcpc.com (Jay Freeman)
Date: Fri, 14 Apr 2000 18:00:25 -0500
Subject: Doesn't anybody write to paper anymore?
References: <38F5FDF3.EE2FBA01@tcac.net> <38F73C80.E1432809@sage.att.com>
Message-ID: <38F7A309.F1242F00@direcpc.com>

Let's see:

C:\WINDOWS>ping www.fcc.gov

Pinging infoserver.fcc.gov [192.104.54.12] with 32 bytes of data:

Reply from 192.104.54.12: bytes=32 time=510ms TTL=246
Reply from 192.104.54.12: bytes=32 time=511ms TTL=246
Reply from 192.104.54.12: bytes=32 time=510ms TTL=246
Reply from 192.104.54.12: bytes=32 time=511ms TTL=246

Ping statistics for 192.104.54.12:
    Packets: Sent = 4, Received = 4, Lost = 0 (0% loss),
Approximate round trip times in milli-seconds:
    Minimum = 510ms, Maximum =  511ms, Average =  510ms


Yep, still have a government, so we must still have paper too, QED :).

Jay

Garry Hodgson wrote:

> "Daddy, what's "paper"?"
>
> well, maybe not yet...
>
> --
> Garry Hodgson                   Every night
> garry at sage.att.com                a child is born
> Software Innovation Services        is a Holy Night.
> AT&T Labs                         - Sophia Lyon Fahs

--
*------------------------------------------*
* Jay Freeman - WT9S                 ARRL  *
* G-QRP 10319 QRP-ARCI 9981 ARS 562        *
* SASS #18700                    NRA Life  *
*------------------------------------------*





From mhammond at skippinet.com.au  Wed Apr 19 22:08:31 2000
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Thu, 20 Apr 2000 02:08:31 GMT
Subject: Can COM-objects be used in a thread?
References: <dn1g0sj1vxw.fsf@kikp0524.chembio.ntnu.no>
Message-ID: <zEtL4.8$v85.116@news-server.bigpond.net.au>

Python COM fully supports all COM threading models.  However, you are
correct in that COM (not Python) does impose certain restrictions on COM
threads - however, they arent as simple to explain as "must have a message
loop" (although in some cases that is true).

I dont have time to explain in detail again.  I believe I DejaNews should
find a post from me with some detail in this area, or Andy and my book
goes into this in depth, including sample code demonstrating these
rules...

Mark.

"Olaf Trygve Berglihn" <olafb at pvv.org> wrote in message
news:dn1g0sj1vxw.fsf at kikp0524.chembio.ntnu.no...
> I'd like to make a thread work on a COM-object created by the main
> thread. Is this possible? Can the COM-object be passed to the thread
> and function properly there?
>
> I vaguely remember that somebody posted a message about this, saying
> it is not possible for a thread to host a COM-object as the main
> thread is not dispatching Windows messages to it.
> --
> * Olaf Trygve Berglihn <olafb at pvv.org>




From mfletch at tpresence.com  Tue Apr  4 09:43:14 2000
From: mfletch at tpresence.com (Mike Fletcher)
Date: Tue, 4 Apr 2000 09:43:14 -0400 
Subject: Tuples -- who needs 'em
Message-ID: <CC59868A0A76D311B1F50090277B321C1F324F@VRTMAIL>

Certainly it raises a TypeError, currently the implementation prevents you
from making the mistake of using a mutable sequence as a dictionary key. If
you prefer, the question is:
	in this sequence, what should be the value of the final line, if we
allow mutable sequences as dictionary keys.

The question of what a "meaningful" hash function would be for a mutable
sequence is the crux of the issue, and the thrust of my attempted
clarification.  Without a fixed value, the "meaning" of a hash function for
a mutable sequence is currently undefined.  Define what that meaning should
be (and do so convincingly), and the implementers might be able to implement
it :) .

Of course, since you were just playing with the words, I'll let the matter
rest (matters are often tired out by all this word-play, crotchety old
things, should never let them out of the rest home...) :) 
*poof*

-----Original Message-----
From: Hrvoje Niksic [mailto:hniksic at iskon.hr]
Sent: Tuesday, April 04, 2000 9:43 AM
To: Mike Fletcher
Cc: python-list at python.org
Subject: Re: Tuples -- who needs 'em


Mike  Fletcher <mfletch at tpresence.com> writes:

> I think the point that is being made is this:
> 	The _value_ of a mutable object cannot (easily) be used as a
> dictionary key...
> 	i.e. in this sequence, what is the value of the final line...
> 
> s = [ 1,2,3 ]
> p = { s: 4, [2,3,4]: 5 }
> s[:] = [2,3,4]
> p[ [2,3,4] ]

The second line raises a TypeError, therefore the last line does not
have a meaningful value.

I only noted that Fredrik's sentence, as it was formulated, was wrong.
Mutable collections can be used as dictionary keys, as long as they
provide a meaningful hash function.

-- 
http://www.python.org/mailman/listinfo/python-list



From effbot at telia.com  Sun Apr  2 18:56:57 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Sun, 02 Apr 2000 22:56:57 GMT
Subject: Web hosting python CGI and or Zope
References: <8c8hie$5lo$1@nnrp1.deja.com>
Message-ID: <ZeQF4.4792$74.76556@newsc.telia.net>

zsoltman at my-deja.com wrote:
> Question:  Does anyone know of Web Hosting companies that support
> python or Zope?

this might help:
http://www.zope.org/Resources/ZSP

</F>




From pinard at iro.umontreal.ca  Wed Apr 26 10:26:02 2000
From: pinard at iro.umontreal.ca (=?ISO-8859-1?Q?Fran=E7ois_Pinard?=)
Date: 26 Apr 2000 10:26:02 -0400
Subject: Python - Next Release Questions
In-Reply-To: Mark Atwood's message of "23 Apr 2000 20:20:54 -0700"
References: <NDBBKEGCNONMNKGDINPFEEJCDEAA.infonuovo@email.com> <m3ya64nq09.fsf@flash.localdomain>
Message-ID: <oqr9btx7k5.fsf@titan.progiciels-bpi.ca>

Mark Atwood <mra at pobox.com> writes:

> I.E., the various cultural groups that dislike the Unicode Han unification
> because they wish to continue their racial illusion that C, J, & V
> characters are completely different and seperate sets, and that it's
> somehow corrupting to stir them together.

I would rather let people judge what is good for them, than impose my
opinion or judgemental remarks to them, while not even understanding or
sharing the culture.  Such statements might be good incentive for Asian
people to escape Americanish colonialism, and maybe Unicode in particular.
The nicest would be for me to shut up, when things are none of my business.

I got a few people haughty enough to preach me, in English, about my needs
for French.  You really have to live the experience to know the feeling! :-)

-- 
Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard





From scarblac-spamtrap at pino.selwerd.nl  Sat Apr 29 16:50:27 2000
From: scarblac-spamtrap at pino.selwerd.nl (Remco Gerlich)
Date: 29 Apr 2000 20:50:27 GMT
Subject: Problem running NNTP example from docs.
References: <87hfckk6fv.fsf@tea.thpoon.com>
Message-ID: <slrn8gmims.i5g.scarblac-spamtrap@flits104-37.flits.rug.nl>

Arcady Genkin wrote in comp.lang.python:
> I'm very new to Python, and was trying to run the example from
> documentation to nntplib library. I get an error when I try to
> instantiate NNTP object.
> 
> ,----
> | >>> import nntplib
> | >>> s = NNTP('news')
> | Traceback (innermost last):
> |   File "<stdin>", line 1, in ?
> | NameError: NNTP
> `----
> 
> Could somebody point out what I am missing?

You have imported the module nttplib; to use the object NNTP within it, you
write nntlib.NNTP. Ie,

s = nntplib.NNTP('news')

(have never used that lib myself, but this is probably it)
-- 
Remco Gerlich,  scarblac at pino.selwerd.nl
Hi! I'm a .sig virus! Join the fun and copy me into yours!


From genius at idirect.com  Sat Apr 29 16:03:18 2000
From: genius at idirect.com (Snoopy :-)))
Date: Sat, 29 Apr 2000 20:03:18 GMT
Subject: Help Needed-IDLE (Linux)Upgrading
References: <3909CC26.D684AD92@idirect.com> <QmmO4.21$kM.5530@news>
Message-ID: <390B3F75.404DD416@idirect.com>

Hi Bob :-)
Thanks for the reply. Your sugestion helped me to solve the problem.

  Actually I was able to figure it out and solve it last night. I just
didn't have the chance to get back until now.

What I did was: to go into GnoRpm and did a "Query" on Python and was
able to trace idle. Then I just followed the downloaded README by
copying the files into the old directory.

So what do you think?  Is there any hope for me in becoming a genius one
day?? :-))
Best regards
Charles


Me wrote:
> 
> It is there! I did the same thing...either do a search for "python" or
> "idle" or look in (forgive me I am not on my box at the moment) /usr/share
> ????  IT IS THERE THOUGH! I would move the old (just in case) and then copy
> it in. I am new too.
> 
> Bob
> 
> "Snoopy :-))" <genius at idirect.com> wrote in message
> news:3909CC26.D684AD92 at idirect.com...
> > No Flames Please. I am a NEWBIE :-))
> >
> > Using Linux (RH-6.1) I have downloaded the Idle-0.5 Upgrade "tgz".
> > package.But I don't know where and how to install the package.
> >
> > After unzipping the package I've read the README
> > file which doesn't seem togive clear instructions regarding installation
> > of the upgrade. On my system Idle is in /usr/bin
> > but this is only the executable file without subdirectories where I
> > would be able to put the untared and uzippedfiles.
> >
> > I would appreciate a helping hand
> > Best regards
> > Charles Takacs


From cut_me_out at hotmail.com  Wed Apr 26 18:56:02 2000
From: cut_me_out at hotmail.com (Alex)
Date: 26 Apr 2000 18:56:02 -0400
Subject: first non-trivial python program
References: <fbsegsgbqq6803j5n54bvo2d3gcna93sl8@4ax.com>
Message-ID: <etd66t4320t.fsf@w20-575-97.mit.edu>

> would it be considered rude to post it here for helpful advice?

Well, frankly, you're showing quite a hide by even asking, but we'll let
you post it this time. :)

People are unlikely to read such a large program in its entirety for
altruistic reasons, but you might get a couple of helpful ideas.

Alex.


From phd at phd.russ.ru  Fri Apr 21 11:52:08 2000
From: phd at phd.russ.ru (Oleg Broytmann)
Date: Fri, 21 Apr 2000 15:52:08 +0000 (GMT)
Subject: PyApache memory leak?
In-Reply-To: <00042107521900.02973@quadra.teleo.net>
Message-ID: <Pine.LNX.4.21.0004211549250.4798-100000@fep132.fep.ru>

On Fri, 21 Apr 2000, Patrick Phalen wrote:
> [Oleg Broytmann, on Fri, 21 Apr 2000]
> :: I don't see how httpdapy could help spawn less interpreters instances -
> :: it is Apache job to fork and reap off children.
> 
> A clarification, from the Httpdapy README:
> 
> """
> 2. PyApache creates a sub-interpreter (Py_NewInterpreter()) for every
> request and destroys it when done. This means that if your script
> begins with "import HTMLgen", HTMLgen is imported (bytecode read from
> file) for every hit.
> 
> Httpdapy keeps the interpreter around from the first hit and until the
> process dies. So only the first hit will actually read HTMLgen.py(c),
> all the subsequent won't.
> """

   This don't help saving memory. Other way round - PyApache destroys
subinterpreter after processing a request, thus saving memory; but httpdapy
keeps the interpreter with all (unneccessary and superfluos) imported
modules, thus leaking memory.
   At least this is how I understand the situation. Not that I'm saying
PyApache is "best" and httpdaoy is "worst" - both are great products, and I
am happy they are there.

Oleg.            (All opinions are mine and not of my employer)
---- 
    Oleg Broytmann      Foundation for Effective Policies      phd at phd.russ.ru
           Programmers don't die, they just GOSUB without RETURN.




From kevinmcd at NOSPAM.linuxsolutions.demon.co.uk  Wed Apr 26 11:39:51 2000
From: kevinmcd at NOSPAM.linuxsolutions.demon.co.uk (Kevin McDermott)
Date: Wed, 26 Apr 2000 16:39:51 +0100
Subject: Best database to use with Python
References: <3906ea27.698463777@news.online.no>
Message-ID: <39070DC7.37F8089D@NOSPAM.linuxsolutions.demon.co.uk>

Hi Thomas,

> 
> What database-system is easiest, fastest etc. to use with Python? I`ve
> looked at MySQL, but it seems a little overkill just to get a little
> more functionality than Berkley DB-functionality. What about Postgres?
> mSQL?
> 
> Postgres comes with my Linux distro and seem smaller than MySQL. How
> does it compare to MySQL when it comes to administration and speed,
> scalability etc. ?
> 
> I`m doing alot of index-lookups for a
> Yahoo/Altavista/WhatEverSearchEngine-like web-app.
> 
> <entering_demand_mode>
> It have to be fast, capable of HUGE amounts of data, free and easy to
> install and administer.
> </entering_demand_mode>
> 

MySQL is very fast, capable of LARGE amounts of data, depends on what
you term HUGE I suppose.

It's very stable, but it lacks is Transactions and some advanced SQL
techniques.

Depends on what you want to use it for...if it's to connect CGI programs
to, then I'd recommend MySQL.

PostgreSQL is different, the support for BLOBs is not quite as good, but
the V7.0 release expected RSN should be a good deal faster.

mSQL is not as good as either of the above once you get above a certain
level, there is no facility for GROUPing  data or SUM functions.

It's reasonably fast and very lightweight, but for storing HUGE amounts
of data, I'd not recommend it.

Kevin


From kilian.lorenz at dkfz-heidelberg.de  Sun Apr  9 14:06:31 2000
From: kilian.lorenz at dkfz-heidelberg.de (Kilian Lorenz)
Date: Sun, 09 Apr 2000 20:06:31 +0200
Subject: Tkinter:  centering windows on screen
References: <ak1I4.121418$Dv1.1306619@news1.rdc1.bc.home.com>
Message-ID: <38F0C6A7.10878609@dkfz-heidelberg.de>

Kevin Cazabon wrote:
> 
> I'm having a bit of a problem controlling where on screen a Toplevel() or
> Tk() widget appears on the screen...
> 
> I have some large widgets that sometimes appear partially off-screen, and I
> want to make sure that (at least) they're centered.
>

I have a similar problem ... I would like to call the tk_getOpenFile
Standard Dialog that is provided by Tk but I don't know how to do it.
If I use it as an option for a widget like this

self.filemenu.add_command(label='Open', command='tk_getOpenFile')

that works but doesn't help me much because I'm interested in the
return value. So I would really like to know how to call Tk commands
from
within Python because that would solve both our problems ...you could
send a Tk command like 'wm geometry <kevins enormous widget>' to receive
an return value like this: '200x200+400+300' and you could change the
position just by sending 'wm geometry <kevins enormous widget> +20+30'
to move it to somewhere near the left, upper corner. 

In the meantime I found out that exactly this can be done with Tkinter
with the wm_geometry() method.

Regards, Kilian


From glyph at twistedmatrix.com  Mon Apr 24 15:50:53 2000
From: glyph at twistedmatrix.com (Glyph Lefkowitz)
Date: 24 Apr 2000 14:50:53 -0500
Subject: Pickling Robustly
References: <jak8huf1of.fsf@helix.twistedmatrix.com> <8dpk1i$7de$1@newshost.accu.uu.nl> <jazoqn2s0b.fsf@helix.twistedmatrix.com> <slrn8g4fs9.33l.jepler.lnk@potty.housenet>
Message-ID: <ja3dobqnvm.fsf@helix.twistedmatrix.com>

jepler.lnk at lnk.ispi.net (jepler epler) writes:

> On 21 Apr 2000 14:06:28 -0500, Glyph Lefkowitz
>  <glyph at twistedmatrix.com> wrote:
> >Perhaps I should ask this question in a different way.  How does
> >Pickle tell if an object is pickleable or not?  Is there some
> >pickle.ispicklable method I could refer to?
> 
> If you have control over the object hierarchy, why not use this
> 'unpicklable mixin' and convenience functions I just tossed together:

[ code snipped ]

Your suggestion didn't solve the problem (thanks for the code though,
useful nonetheless); however, I believe it may have elucidated to me
that what I want is fundamentally impossible.  If I boil down my
request, it's for control over the object heirarchy without control
over the object heirarchy...

With the additions that I've stuck in, the only unpicklable things are
native code... I can achieve the control that I want by overriding the
interfaces that the pickle module describes, I think ...

It would still be nice if there were an option to register an error
handler to just ignore attributes which couldn't be pickled, rather
than failing.  At this point though, I think I'd need to rewrite the
module to get that.

-- 
                  __________________________________________
                 |    ______      __   __  _____  _     _   |
                 |   |  ____ |      \_/   |_____] |_____|   |
                 |   |_____| |_____  |    |       |     |   |
                 |   @ t w i s t e d m a t r i x  . c o m   |
                 |   http://www.twistedmatrix.com/~glyph/   |
                 `__________________________________________'



From ulf.engstrom at b2b-link.com  Fri Apr 28 05:16:34 2000
From: ulf.engstrom at b2b-link.com (=?iso-8859-1?Q?Ulf_Engstr=F8m?=)
Date: Fri, 28 Apr 2000 11:16:34 +0200
Subject: Concatening string and integer
References: <Pine.OSF.4.20.0004281848300.3262-100000@azure.dstc.edu.au>
Message-ID: <018e01bfb0f2$73fda480$858647c1@DEVELOPMENT>

Also important to remember when using ` with strings is this:
>>> text = 'text'
>>> `text`
"'text'"
>>> num = 5
>>> `num`
'5'

Adding an unknown variable to a textstring, the variable being a number is
fine, being a string already gives you
'text "variable" and more text', not very pretty.

(Other way to deal with that is 'text %s and more text' %
(unknownvariabletype), or checking type() before using it)
Regards
Ulf

> Actually the back ticks appear to be a shortcut for repr() not str()
> which in the above case give the same result but won't for all objects,
> for example:
>
> >>> class aclass:
> ...     def __str__(self):
> ...         return 'spam'
> ...
> >>> a = aclass()
> >>> `a`
> '<__main__.aclass instance at f93a8>'
> >>> str(a)
> 'spam'
> >>> repr(a)
> '<__main__.aclass instance at f93a8>'
>
> Cheers,
> Anthony
> _____________________________________________________________________
> Anthony J Wilkinson                               anthony at dstc.edu.au
> Software Engineer                                http://www.fnorb.com
> DSTC Pty Ltd                                     Ph:  +61 7 3365 4310
>
>
>
> --
> http://www.python.org/mailman/listinfo/python-list




From gmcm at hypernet.com  Sat Apr 22 09:10:07 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Sat, 22 Apr 2000 09:10:07 -0400
Subject: The Simple Economics of Open Source
In-Reply-To: <3901247A.E87B5D0A@rubic.com>
Message-ID: <1255713484-40391375@hypernet.com>

Jeff Bauer <wink>ed:

> If I didn't have my part-time performance art
> income to help pay the bills, I could never
> afford to support my programming lifestyle.

Mimes for food and MIMEs for fun...

- Gordon



From effbot at telia.com  Sat Apr 15 12:45:49 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Sat, 15 Apr 2000 16:45:49 GMT
Subject: BSDi and PIL
References: <sH0K4.13$Ny3.1761@newreader.ukcore.bt.net>
Message-ID: <111K4.40$wYl.186672128@newsb.telia.net>

David Reid <dreid at jetnet.co.uk> wrote:
> Has anyone had any joy in getting Python 1.5.2 and PIL 1.0 to build
together
> on BSDi 4.0.1?

$ cd libImaging
$ ./configure
$ make
$ cd ..
$ make -f Makefile.pre.in boot
$ make
$ ls -l _imaging.so

how far did you get?

</F>




From donn at u.washington.edu  Tue Apr 18 12:30:13 2000
From: donn at u.washington.edu (Donn Cave)
Date: 18 Apr 2000 16:30:13 GMT
Subject: modifying os.environ
References: <p6qhfczo50y.fsf@informatik.hu-berlin.de> <slrn8f9mn1.ven.quinn@zloty.ugcs.caltech.edu>
Message-ID: <8di2il$10oq$1@nntp6.u.washington.edu>

Quoth Martin von Loewis <loewis at informatik.hu-berlin.de>:
| quinn at zloty.ugcs.caltech.edu (Quinn Dunkan) writes:
...
| > Or perhaps there's a better way to do the (common) thing I'm trying
| > to do: clean the environment to insure some minimum sane values.  I
| > can't make a copy of the environ and then pass to execve() since
| > that doesn't help much for popen and system.
|
| You could implement popen and system yourself, starting from popen2
| (for example).

That may sound like the re-invention of the wheel, but there are
other improvements that can be made at the same time:

 - eliminate system()'s sh -c 'shell command', which can be a
   security liability in some situations.  The shell can do
   some really useful things, but if you don't need those things
   you should be able to pass in an argv sequence and skip sh.
 - Raise exception on exit status, complete with error output as
   exception value.  A bit of a pain to implement, and again not
   not always the right thing to do, but usually it is the right
   thing.

	Donn Cave, University Computing Services, University of Washington
	donn at u.washington.edu


From aaron_watters at my-deja.com  Wed Apr 26 10:39:58 2000
From: aaron_watters at my-deja.com (aaron_watters at my-deja.com)
Date: Wed, 26 Apr 2000 14:39:58 GMT
Subject: Sourceforge and Python
References: <oq8zy1zqto.fsf@titan.progiciels-bpi.ca> <007b01bfaf14$f33114b0$3cdd3fcb@neil> <slrn8gcgvp.u19.kc5tja@garnet.armored.net>
Message-ID: <8e6v3r$3ss$1@nnrp1.deja.com>

In article <slrn8gcgvp.u19.kc5tja at garnet.armored.net>,
  kc5tja at garnet.armored.net (Samuel A. Falvo II) wrote:
> How the *hell* did you get your CVS repo to work on SourceForge?  :)
>
> Whenever I try to upload to the repo I've set up for my own project,
it
> tells me I don't have write permissions after processing about four
or five
> files (I forget the exact error phrase, but that's basically what it
> reports).  Now, I've used CVS before -- I have a repo set up on my
box at
> home.  What am I missing with respect to CVS with SourceForge?

I don't know if this will help, but I had trouble with this
too and with the help of Andy R. I came up with the following
command line which always works.

cvs -z6
-d :ext:aaron_watters at cvs.reportlab.sourceforge.net:/cvsroot/reportlab
commit pdfdoc.py

It then prompts me for a password etc.  This is using the
ssh and the cvs versions recommended by sourceforge
and carefully setting up the needed directories and environment
variables, vis, in autoexec.bat


set HOME=c:\users\arw
set path=%path%;c:\usr\local\bin;c:\python;"C:\Program Files\GNU\WinCvs"
set CVS_RSH=ssh
SET
CVSROOT=aaron_watters at cvs.reportlab.sourceforge.net:/cvsroot/reportlab
SET USERNAME=aaron_watters
SET HOMEDRIVE=c:
SET HOME=\users\arw
SET HOMEPATH=\users\arw
SET CVS_RSH=ssh

and in c:\etc\passwd

arw:x:1:10:Watters, Aaron R.:/users/arw/:/windows/cmd.exe

Why this particular black magic does the trick, I've no clue,
neither do I know whether it's all really needed,
but I don't waste cycles thinking too hard about it.

In my youth I used to waste immense amounts of emotional
energy getting mad about this kind of nonsense.  Now I'm
just trying to have a little fun before I die, as the muse
says...

   -- Aaron Watters

===
and the colored girls go
doot
tadoot
doot
doottadoot...   -- Lou Reed



Sent via Deja.com http://www.deja.com/
Before you buy.


From mwh21 at cam.ac.uk  Thu Apr 27 19:14:05 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 28 Apr 2000 00:14:05 +0100
Subject: mail forwarding not working on the starship?
References: <3908538A.E74456A5@GermanLloyd.org> <8eagr4$30h$1@nnrp1.deja.com>
Message-ID: <m33do7p26a.fsf@atrus.jesus.cam.ac.uk>

timehorse at unforgettable.com writes:

> Anyway, one thing that everyone on Starship should know is that we
> are no longer using SENDMAIL as our transport agent.  We have been
> moved to qmail as it is LEAPS AND BOUNDS easier to maintain.

starship's been on qmail for as long as I've been there (about 18
months; long before the crash).  So it's not that...

Cheers,
M.

-- 
41. Some programming languages manage to absorb change, but 
    withstand progress.
     -- Alan Perlis, http://www.cs.yale.edu/~perlis-alan/quotes.html


From aa8vb at yahoo.com  Tue Apr 11 13:43:58 2000
From: aa8vb at yahoo.com (Randall Hopper)
Date: Tue, 11 Apr 2000 13:43:58 -0400
Subject: comparing strings and ints
In-Reply-To: <2YmI4.366$rc9.188014080@newsb.telia.net>
References: <8css5t$tti$1@merrimack.Dartmouth.EDU>
 <2YmI4.366$rc9.188014080@newsb.telia.net>
Message-ID: <20000411134358.A2352304@vislab.epa.gov>

Fredrik Lundh:
 |Mark H.H. Montague wrote:
 |> Does anyone know why it is not a type error to compare a string and an
 |> int with '>' ?  For example, I just spent over an hour debugging code
 |> that essentially did this:
 |>
 |> >>> 45 > '0'
 |> 0
 |>
 |> Of course, the '0' was a variable read in from a file, which I forgot
 |> to convert to a string.  Bad old perl habits.  But I'm having a hard
 |> time imagining a situation where someone would be intentionally and
 |> meaningfully comparing a string and an int.
 |>
 |> Any ideas?
 |
 |why not read the fine documentation:
 |http://www.python.org/doc/current/ref/comparisons.html

I suspect many folks consult it first, but (though excellent on most
topics) occasionally it's difficult to find what you want.  And sometimes
(rarely) if and when you find the topic, the behavior doesn't "seem" to
make much sense at first glance.  Thus prompting a question to the group.

(An example is the % operator.  It's an operator so it's in the language
ref manual, right?  No, not exactly.  If you don't think "string method"
you'll never find it.  Compare to +/-.  Why is it special?  Beats me.)

 |which explains that:
 |
 |    Objects of different types always compare unequal,
 |    and are ordered consistently but arbitrarily.

Reminds me of qsort() core dumps because the compiled C library compared
NaNs and Infs arbitrarily.

I can see 45 == '45' being false, and 45 != '45' being true.  But it seems
to me that not throwing an exception for attempts at an ordered comparison
with logical ordering operators (<, <=, etc.) on objects of different
primitive types (e.g. 45 < '45', 45 > '45') only lets bugs pass through.

Is there a case where this could be useful (with its undefined behavior)?
I believe that is the root of the question.
   
 |    This unusual definition of comparison was used to
 |    simplify the definition of operations like sorting and
 |    the "in" and "not in" operators.

== and != seems to cover "in"/"not in".  In what context is sorting done
on instances that have different primitive types?

 |and goes on to say:
 |
 |    In the future, the comparison rules for objects of
 |    different types are likely to change.

Possibly making ording operations illegal in this case (?)

-- 
Randall Hopper
aa8vb at yahoo.com



From parkw at better.net  Sun Apr  2 16:12:29 2000
From: parkw at better.net (William Park)
Date: Sun, 2 Apr 2000 16:12:29 -0400
Subject: Creating variables on the fly...
In-Reply-To: <meh9-25D0C5.15160303042000@news.cit.cornell.edu>; from meh9@cornell.edu on Mon, Apr 03, 2000 at 03:16:03PM -0400
References: <meh9-25D0C5.15160303042000@news.cit.cornell.edu>
Message-ID: <20000402161229.A1954@better.net>

On Mon, Apr 03, 2000 at 03:16:03PM -0400, Matthew Hirsch wrote:
> Hi All,
> 
> Let's say I have four variables: 
> 
> a=1
> b=2
> c=3
> d=4
> 
> And I have a function that adds these variables together:
> 
> def add(a,b,c,d):
>    return a+b+c+d
> 
> But now let's say I have twenty variables that I want to add together.  
> This function will no longer work.  I would have to rewrite it as 
> a+b+c+d+e+f+g+...+(20th letter).  Is there a way to dynamically create a 
> variable name?  So that my add function can automatically determine how 
> many variable names to create and then add the values together.  In 
> other words, I'd ideally like something like:
> 
> 
> def add(number_of_variables):
>    return a+b+c+...+(letter corresponding to number_of_variables)
> 
> Thanks for your help,
> 
> Matt

Straightforward approach is 
    def add(*argv):
	<Add each element of tuple 'argv' in a loop>

    sum = add(a, b, c, ...)

Or, you can use dictionary,
    def add(**argv):
	<Add each element of dictionary 'argv' in a loop>

    sum = add(a=a, b=b, ...)
    
--William



From aahz at netcom.com  Sat Apr 15 23:55:15 2000
From: aahz at netcom.com (Aahz Maruch)
Date: 16 Apr 2000 03:55:15 GMT
Subject: Python paradigms
References: <8cmr70$b0e$1@pegasus.csx.cam.ac.uk>
Message-ID: <8dbdj3$dgt$1@slb2.atl.mindspring.net>

In article <8cmr70$b0e$1 at pegasus.csx.cam.ac.uk>,
Nick Maclaren <nmm1 at cus.cam.ac.uk> wrote:
>
>    x = (a != NULL ? a[i]->weeble : 0) + (b != NULL ? b[i]->wombat : 0)

Here's a solution I haven't seen before (refactored slightly for the
most likely equivalent Python idiom):

def expression_if( test, true, false ):
  if test: return true
  else: return false

x = expression_if( a is not None, 
   a[i].weeble, 0) + expression_if( b is not None, b[i].wombat, 0)
--
                      --- Aahz (Copyright 2000 by aahz at netcom.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

"First one back from the bathroom gets to be the bottom."


From markx.daley at intel.com  Wed Apr 12 14:10:45 2000
From: markx.daley at intel.com (Daley, MarkX)
Date: Wed, 12 Apr 2000 11:10:45 -0700
Subject: global variables - how??
Message-ID: <A63AFB20111ED311AC5500A0C96B7BF5015F0F16@orsmsx36.jf.intel.com>

You can also assign them as lists instead of standard variables.  Then, when
you reference them in your modules, just use the first entry in the list.

e.g.

a = [Vorpal]
b = [Bunny!]

def eek():
	print a[0]
	print b[0]

I find this list usage quite useful.  Hope this helps!

Mark

-----Original Message-----
From: William Park [mailto:parkw at better.net]
Sent: Wednesday, April 12, 2000 10:41 AM
To: python-list at python.org
Subject: Re: global variables - how??


On Wed, Apr 12, 2000 at 07:28:37PM +0200, Anders Eggers-Krag wrote:
> how do I define a propper global variable that works in all functions in
all
> modules?

As far as I know, you can't.  You have to use 'global ...' inside all
your functions.

--William

> 
> I have tried simply
> defining one outside of any function in the main module,
> I have tried using the global keyword under the same cicumstances,
> but they are not even avialible inside of functions within the main
> module...
> 
> 
> --
> Anders Eggers - Krag
> 
> 
> -- 
> http://www.python.org/mailman/listinfo/python-list





From jvickroy at sec.noaa.gov  Mon Apr 24 09:59:41 2000
From: jvickroy at sec.noaa.gov (j vickroy)
Date: Mon, 24 Apr 2000 07:59:41 -0600
Subject: Slashdot review of Python Programming On Win32
References: <CC59868A0A76D311B1F50090277B321C1F32A5@VRTMAIL>
Message-ID: <3904534D.2D324F7B@sec.noaa.gov>

I recently purchased a copy of this book and have only skimmed most
chapters to date.  I have, however, read Chapter 22 more carefully and
found it essential for developing a dll extension library for Python.
Being unfamiliar with Win 32 programming, I found the chapter clear,
complete, and accurate.

Mike Fletcher wrote:

> http://slashdot.org/article.pl?sid=00/04/17/1147228&mode=nocomment
>
> Generally positive review, got me thinking I should buy the book.
>
> Enjoy all,
> Mike
> __________________________________
>  Mike C. Fletcher
>  Designer, VR Plumber
>  http://members.home.com/mcfletch
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20000424/9134b93d/attachment.html>

From quynhna at hotmail.com  Thu Apr 13 23:36:17 2000
From: quynhna at hotmail.com (Nguyen Anh Quynh)
Date: Thu, 13 Apr 2000 20:36:17 -0700
Subject: Change NT password with Python
Message-ID: <38F69231.EE1D9B6D@hotmail.com>

Dear my friends,
I am looking for the way to write on Python stuff helping me to change
NT password, since I would like to change the password of WinNT domain
via Web page. (That Python stuff will be the CGI)
Could you point me the way?
Thank you very much,




From thomas at xs4all.net  Fri Apr 14 09:47:18 2000
From: thomas at xs4all.net (Thomas Wouters)
Date: Fri, 14 Apr 2000 15:47:18 +0200
Subject: Python 1.6a2 crashes in test_fork1 on my Mandrake Linux 7 machine
In-Reply-To: <38F658C6.66AD9CA5@clark.net>; from jrl@clark.net on Thu, Apr 13, 2000 at 06:31:18PM -0500
References: <20000413010344.A13931@owl.rhein-zeitung.de> <38F658C6.66AD9CA5@clark.net>
Message-ID: <20000414154718.A15664@xs4all.nl>

On Thu, Apr 13, 2000 at 06:31:18PM -0500, John Lloyd wrote:
> Oliver Andrich wrote:

[ about test_fork1 failing on a Mandrake 7 system ]

I dont know enough about threads to say anything useful about this, but it
is kind of strange that you dont see the same problem under Python 1.5.2. My
redhat boxes pass the test, though the BSDI boxes dont (but threading is
screwed in more than one way under most BSDI versions.) I haven't tried
compiling 1.5.2 under BSDI 4.1 (the newest, which seems to work reasonably
ok with threading) though, so it might be a change in python's
thread-handling...

Wasn't Mandrake based on Redhat ? Does it use a different/strange libc or
something ? i dont quite see why there should be a difference in rh 6.1 and
mandrake, in this context.

> I had the same fault but when I compiled without thread support it works
> OK. 

This is not very suprising, since test_fork1 tests for the combination of
threading and fork. (wether threads survive after a fork(); apparently they
shouldn't, but they do, on many systems.)

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From tismer at tismer.com  Thu Apr  6 17:22:30 2000
From: tismer at tismer.com (Christian Tismer)
Date: Thu, 06 Apr 2000 23:22:30 +0200
Subject: Round Bug in Python 1.6?
Message-ID: <38ED0016.E1C4A26C@tismer.com>

Hi,

asa side effect, I happened to observe the following rounding bug.
It happens in Stackless Python, which is built against the
pre-unicode CVS branch.

Is this changed for 1.6, or might it be my bug?

D:\python\spc>python
Python 1.5.42+ (#0, Mar 29 2000, 20:23:26) [MSC 32 bit (Intel)] on win32
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> round(3.1415926585, 4)
3.1415999999999999
>>> ^Z

D:\python>python
Python 1.5.2 (#0, Apr 13 1999, 10:51:12) [MSC 32 bit (Intel)] on win32
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> round(3.1415926585, 4)
3.1416
>>> ^Z

ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaunstr. 26                  :    *Starship* http://starship.python.net
14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     where do you want to jump today?   http://www.stackless.com



From gmcm at hypernet.com  Mon Apr 10 13:32:46 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Mon, 10 Apr 2000 13:32:46 -0400
Subject: Redirecting stdin/stdout on Windows with '>'and '<'
In-Reply-To: <8csvff$7pn$1@nnrp1.deja.com>
Message-ID: <1256734529-36551909@hypernet.com>

Mark <markmcmahon at my-deja.com> wrote:

> I seem to have a problem and can't find it in the archives or FAQ.

Not sure why not. It's come up enough times.
 
> I would like to redirect output of my script with '>' or '<'
> 
> If I try to redirect out then the file is created but has no contents.
> 
> If I try to redirect in eg...
> "test.py < autoexec.bat"
> 
> I get the following traceback
> 
> Traceback (most recent call last):
>   File "C:\test.py", line 4, in ?
>     s = sys.stdin.read()
> IOError: [Errno 9] Bad file descriptor
> 
> Does anybody have an idea!!!
> 
> (Windows nt 4 sp 6, python 1.52)

Apparently the bug is fixed in Win2K, but in NT and Win9x, 
redirection is busted unless you spell it out:
 python test.py >test.out


- Gordon



From sabren at manifestation.com  Sun Apr 23 06:34:10 2000
From: sabren at manifestation.com (Michal Wallace (sabren))
Date: Sun, 23 Apr 2000 06:34:10 -0400 (EDT)
Subject: code not true?
In-Reply-To: <JHzM4.1116$wYl.189856768@newsb.telia.net>
Message-ID: <Pine.BSF.4.10.10004230633180.5642-100000@chi.pair.com>

On Sun, 23 Apr 2000, Fredrik Lundh wrote:

> Michal Wallace (sabren) <sabren at manifestation.com> wrote:
> > Isn't unicode much slower to work with than regular strings?
> 
> nope.

A teeny bit slower? :)

Cheers,

- Michal
-------------------------------------------------------------------------
http://www.manifestation.com/         http://www.linkwatcher.com/metalog/
-------------------------------------------------------------------------




From lutz at rmi.net  Thu Apr 13 13:58:34 2000
From: lutz at rmi.net (Mark Lutz)
Date: Thu, 13 Apr 2000 11:58:34 -0600
Subject: ANN: Python training in Boulder
Message-ID: <000101bfa571$e31a43a0$71d95da6@lutz>

This is to announce that the second Python training session
to be held at Tom Christiansen's site in Boulder, Colorado, 
will be held May 24-26.

These classes are equivalent to the on-site Python training I
offer, and are open to individuals on a first-come first-served
basis.  I am the instructor, but all registration details are
handled by Tom's company.  For more information, please visit 
these pages:

http://home.rmi.net/~lutz/boulder-classes.html   (my page)
http://training.perl.com/open-courses.html       (tom's page)    

--Mark Lutz  (http://rmi.net/~lutz)




From jae at ilk.de  Sat Apr 22 00:08:42 2000
From: jae at ilk.de (Juergen A. Erhard)
Date: Sat, 22 Apr 2000 06:08:42 +0200
Subject: The Simple Economics of Open Source
In-Reply-To: <39011EEF.4BAEC3C9@tcac.net> (message from Albert Wagner on Fri,
	21 Apr 2000 22:39:27 -0500)
References: <14591.15630.192472.440375@goon.cnri.reston.va.us> <slrn8fv2rd.1kt.neelk@brick.cswv.com> <3900611F.3B525BDA@cfdrc.com>  <FtDG4z.774@world.std.com> <slrn8g26tu.p7o.neelk@brick.cswv.com> <39011EEF.4BAEC3C9@tcac.net>
Message-ID: <22042000.3@sanctum.jae.ddns.org>

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

>>>>> "Albert" == Albert Wagner <alwagner at tcac.net> writes:

    Albert> [...]  All that is required is that a "critical mass" of
    Albert> participants realize that the shared rewards of long term
    Albert> cooperation exceed the rewards of short term reward at the
    Albert> expense of others.

Right... but the more important the stock exchanges of this world
become (in the public eye), the bigger this problem gets... there seem
to be (at least to the eyes of a layman (like me ;-)) a lot of
companies which *only* look at their stock price... and thus only at
their quarterly balance report (or whatever that was called).  And not
at any long-term effects of what they're doing...

    Albert> Those with uncooperative strategies view the pie as finite
    Albert> and feel that they must optimize their slice by
    Albert> diminishing the share of others.

When money's all you see (and most of those people do...), it *is* a
zero-sum game.  So their strategies are mostly correctly adapted to
the way they see the world around them.

But when you look at the bigger picture (`Spaceship Earth' comes to
mind[2])...

    Albert> Those with cooperative strategies have discovered that
    Albert> there are an infinite number of pies.

Wrong.[1] They simply discovered (or, rather, understood) that there
are enough pies so that *everyone* can be more than satiated ("Jeder
kann sich bis zum Erbrechen den Bauch vollschlagen" I'd say in
german).

In short: They found that there's enough for everyone...

And the long-term benefits are immeasurable (simply because you can't
measure the opposite... you can't know what *would* have happened,
just what *has* :-).

Bye, J

[1] The human-accessible part of this solar system is finite, right?
So there simple cannot be an infinite number of pies, even if you
could bake a pie out of one atom... ;-)

- -- 
J?rgen A. Erhard      eMail: jae at ilk.de      phone: (GERMANY) 0721 27326
     MARS: http://members.tripod.com/Juergen_Erhard/mars_index.html
                  The GNU Project (http://www.gnu.org)
         Win32 has many known work arounds. For example, Linux.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.1 (GNU/Linux)
Comment: Use Mailcrypt and GnuPG <http://www.gnupg.org/>

iEYEARECAAYFAjkBJcUACgkQN0B+CS56qs1g4gCfUTKfxxZMQO/EOwr3FDT+TnZc
hRYAn14uzj1MeewijEOjo2uGeVPVYJb5
=X+0+
-----END PGP SIGNATURE-----



From alwagner at tcac.net  Tue Apr 18 07:15:43 2000
From: alwagner at tcac.net (Albert Wagner)
Date: Tue, 18 Apr 2000 06:15:43 -0500
Subject: How to write to printer in Windows?
Message-ID: <38FC43DF.78A38D5D@tcac.net>

How can I write(aLine) to a printer in Windows?  I tried this:

from win32pipe import popen
:
self.printer = popen('lpt1' 'w')
:
self.printer.write(aLine)

but get a general protection fault.  Any help will be greatly
appreciated.

-- 
Small is Beautiful


From ke at gnu.franken.de  Thu Apr 27 12:51:03 2000
From: ke at gnu.franken.de (Karl EICHWALDER)
Date: 27 Apr 2000 18:51:03 +0200
Subject: Scripting and Gnome and KDE
References: <Pine.GSO.4.10.10004150741520.11261-100000@sundial> <3dk8hzb90u.fsf@newcnri.cnri.reston.va.us> <shk8hk2m68.fsf@tux.gnu.franken.de> <wkln1zll9r.fsf@mac.com>
Message-ID: <shem7r32tk.fsf@tux.gnu.franken.de>

Jody Winston <josephwinston at mac.com> writes:

> I have to agree with Andrew because I have had these problems and
> more.  I don't work on a Linux machine, but an SGI.

This is a completely different story.  Andrew said GNOME needs patched
tools; this isn't the truth.

> It was very difficult to get the libraries that GNOME needs to work to
> run.

If time permits and good bug reports are submitted the GNOME developers
are very friendly to change the code etc.  Since the developers normally
don't have access to all these systems you might guess how difficult it
is to support them.

> It looks like to me that we have gone back to the model that all the
> world is a Vax and if it compiles and runs on the Vax there isn't any
> problem with the software.  All the world isn't Linux and the only
> compiler isn't gcc.

Sure, but sooner or later, Linux will run on all architecure and you'll
be free to choose the OS that fits your needs best :)

-- 
work : ke at suse.de                          |
     : http://www.suse.de/~ke/             |          ------    ,__o
home : ke at gnu.franken.de                   |         ------   _-\_<,
     : http://www.franken.de/users/gnu/ke/ |        ------   (*)/'(*)


From effbot at telia.com  Tue Apr  4 17:08:55 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Tue, 04 Apr 2000 21:08:55 GMT
Subject: Getting the argument names passed to a function ?
References: <20000404225810.A12156@sz-sb.de>
Message-ID: <HRsG4.5120$74.82867@newsc.telia.net>

Andreas Jung <ajung at sz-sb.de> wrote:
>
> def myfunc(arg1,arg2):
> ...
>
>
> a=1;b=2
> myfunc(a,b)
>
> Is it possible to get the name of variables that  were
> used as arguments to call myfunc(). myfunc() should be
> able to determine the names 'a' and 'b'. Don't ask
> me why I need this - it's for a worse Python hack :-)

short answer: no

longer answer: maybe, using the incredibly clever and
totally evil bytecodehacks stuff:

    http://bytecodehacks.sourceforge.net

</F>

<!-- (the eff-bot guide to) the standard python library:
http://www.pythonware.com/people/fredrik/librarybook.htm
-->




From fra at bastardi.org  Tue Apr  4 15:07:17 2000
From: fra at bastardi.org (Fra)
Date: Tue, 04 Apr 2000 19:07:17 GMT
Subject: Map a Network Drive and Printers in NT?
Message-ID: <38ea3a47.7163227@news.flashnet.it>

Hi all
How can I map a network drive and a printer in Win NT and in Win 95?
Can I iport the dos command line and type the command "Net use...." or
is there a different way?
And could you also tell me how to ping with a Python script? And maybe
how to let this script analyse the result.

I couldn't find it on the Tutorial, neither on the FAQ. Am I wrong?
Thank you,

Greetings,
	   Francesca


From python-list at teleo.net  Wed Apr 12 12:16:50 2000
From: python-list at teleo.net (Patrick Phalen)
Date: Wed, 12 Apr 2000 09:16:50 -0700
Subject: Looking for BoboHTTPServer.py
In-Reply-To: <8d28r6$9bs$2@news.cineca.it>
References: <8d28r6$9bs$2@news.cineca.it>
Message-ID: <00041209195100.01669@quadra.teleo.net>

[Alessandro Muzzetta, on Wed, 12 Apr 2000]
:: Hello,
:: 
:: I'm looking for BoboHTTPServer.py but cannot find it anywhere.
:: Can someone tell me where I can find it, or some other server
:: that integrates Bobo (not Medusa).

Bobo has grown up; now it's called Zope.

http://www.zope.org

Look at ZServer (and ZPublisher, ZODB)



From fw at deneb.cygnus.argh.org  Tue Apr 11 13:35:56 2000
From: fw at deneb.cygnus.argh.org (Florian Weimer)
Date: 11 Apr 2000 19:35:56 +0200
Subject: [ANNOUNCE] Garbage collection for Python
References: <20000407221750.A29279@acs.ucalgary.ca> <38F13FB9.D1D16AF@mindspring.com> <slrn8f2if1.dop.nascheme@cranky.arctrix.com> <87r9cdhdb0.fsf@deneb.cygnus.argh.org> <FsunGI.BtB@world.std.com> <38F319A4.89621192@mdlink.de> <38F351B0.45E137AA@roguewave.com>
Message-ID: <87n1n0czgj.fsf@deneb.cygnus.argh.org>

Bjorn Pettersen <bjorn at roguewave.com> writes:

> Windows have system calls for atomic increment/decrement
> (InterlockedIncrement/InterlockedExchangeAdd), something that is sorely
> lacking from Unix (*sigh*)

A system call is much too slow for that purpose.  If you've got
hardware support, you can use it, otherwise, it becomes quite
expensive.


From thomas at cintra.no  Thu Apr 27 13:22:44 2000
From: thomas at cintra.no (thomas at cintra.no)
Date: Thu, 27 Apr 2000 17:22:44 GMT
Subject: Using struct writing integers
Message-ID: <390873b2.726725@news.c2i.net>

Hi,

I got a list of tuples containing integers, like :

stuff = [ (21,323,32), (432,43,3234), (432,565,532), (434,43,987) ]

Each tuple in the list represent an id for a entry in a database, the
three integers form a key. The list above contains four ids. I want to
write them to a binary file using the struct module. Then I want to
read them pack, unpack them into the same kind of list.

Any hints, examples etc?

Thanks.

Thomas




From nmm1 at cus.cam.ac.uk  Wed Apr  5 04:56:30 2000
From: nmm1 at cus.cam.ac.uk (Nick Maclaren)
Date: 5 Apr 2000 08:56:30 GMT
Subject: blessed be Guido, for "5 <= X <= 10" does what it should...
References: <3TpG4.3349$HG1.98231@nnrp1.uunet.ca> <slrn8el1su.3h1.quinn@riyal.ugcs.caltech.edu>
Message-ID: <8cev3u$bm5$1@pegasus.csx.cam.ac.uk>

In article <slrn8el1su.3h1.quinn at riyal.ugcs.caltech.edu>,
Quinn Dunkan <quinn at riyal.ugcs.caltech.edu> wrote:
>On Tue, 4 Apr 2000 13:42:41 -0500, Warren Postma <embed at geocities.com> wrote:
>><X-DISCLAIMER: I'm in a weird mood.>
>>
>>I was just playing around today,and it occurred to me to test the
>>'transitive' properties of comparison operators. Ie, writing on a
>>chalkboard, you might say X is in the range 5..9 this way:
>>
>>        5 <= X <= 9
>
>Yes, this has always struck me as one of the weirder corners of python.
>There's a bit of magic for comparison operators:
>
>From the lang ref 5.9:
>
>Formally, if a, b, c, ..., y, z are expressions and opa, opb, ..., opy are
>comparison operators, then a opa b opb c ...y opy z is equivalent to a opa b
>and b opb c and ... y opy z, except that each expression is evaluated at most
>once. 
>
>It's kinda cool, I guess.  Maybe ABC did it first?

Dunno.  If I remember correctly, the three operand form was there in
an early version of SPSS (late 1960s.)


Regards,
Nick Maclaren,
University of Cambridge Computing Service,
New Museums Site, Pembroke Street, Cambridge CB2 3QG, England.
Email:  nmm1 at cam.ac.uk
Tel.:  +44 1223 334761    Fax:  +44 1223 334679


From cmfinlay at SPAMmagnet.com.au  Sun Apr 30 08:15:38 2000
From: cmfinlay at SPAMmagnet.com.au (cmfinlay at SPAMmagnet.com.au)
Date: Sun, 30 Apr 2000 22:15:38 +1000
Subject: My Question has now been answered by a member of this Usenet group. Thanks
References: <t=0LOZtIsnos1t9tIo=000hy2MF1@4ax.com>
Message-ID: <ViMMOVjJqxrg5jHL7rl2hSYRzjfm@4ax.com>

My Question has now been answered by a member of this Usenet group.
Thank you for providing such a vital service and community


From rvprasad at cis.ksu.edu  Sun Apr 30 18:04:06 2000
From: rvprasad at cis.ksu.edu (Venkatesh Prasad Ranganath)
Date: 30 Apr 2000 17:04:06 -0500
Subject: Discrepency between JPython and CPython
Message-ID: <m3k8hfjlex.fsf@boss.ksu.edu>

Hi,

Consider the expression str(1.14 * 200).
In CPython it yields 228.0
In JPython it yields 227.99999999999997

Shouldn't this result in consistent values across variuos implementation?  
If so, which of the implementation is buggy?  
I think it is JPython, may be I am wrong.

-- 
Venkatesh Prasad Ranganath

e-mail: rvprasad at cis.ksu.edu, rvprasad at ksu.edu

Address: Department of Computing and Information Science,
         234 Nichols Hall, 
         Kansas State University,
         Manhattan, KS 66506

Phone: (785)-532-6350 Extn: 48



From johann at physics.berkeley.edu  Tue Apr 18 13:00:51 2000
From: johann at physics.berkeley.edu (Johann Hibschman)
Date: 18 Apr 2000 10:00:51 -0700
Subject: Lists
References: <A63AFB20111ED311AC5500A0C96B7BF5015F0F2C@orsmsx36.jf.intel.com>
Message-ID: <mtsnwjs5rw.fsf@astron.berkeley.edu>

Daley, MarkX writes:

> This feels like a real newbie question, but here is the code that is causing
> my question:

> # Learning exceptions

> def test():
> 	a = [-5, -4, -3, -2, -1, 0, 1, 2, 3, 4, 5]
> 	print a
> 	b = 5
> 	for item in a:
> 		try:
> 			print b / a[item]
> 		except ZeroDivisionError:
> 			pass

No, it's iterating in order.
On the first step,
     item = -5, a[item] = 1, and 5/1 = 1
then item = -4, a[item] = 2, and 5/2 = 2
then item = -3, a[item] = 3, and 5/3 = 1

I'm guessing that you either:
  1. Really want "b/item" not "b/a[item]"
or
  2. Don't understand negative indexing.  A negative index to a list
     counts from the back of the list front.


-- 
Johann Hibschman                           johann at physics.berkeley.edu


From dante at oz.net  Wed Apr 26 21:31:13 2000
From: dante at oz.net (Michael Esveldt)
Date: 27 Apr 2000 01:31:13 GMT
Subject: xmlrpclib problem
Message-ID: <8e8591$2ro$0@216.39.162.232>

I'm working with xmlrpclib to talk to Userland's Manila application and 
I'm having a problems sending html via xmlrpc. Here's the code I'm 
working with

mysite = xmlrpclib.Server("http://mysite.editthispage.com") 
mysite.manila.message.set(username, password, sitename, msgNum, "my 
subject", "body containing html", "text/x-outline-tabbed", {}, 
{"name":"pikeRenderer", "flRenderOnEntry":"true"})

I've tried changing "text/x-outline-tabbed" to "text/plain" and 
"text/html", none of these have any effect. What's happening is that my 
"body containing html" is coming out on the other side with &lt; and 
&gt; instead of actual tags. Obviously this results in displaying html 
instead of rendering it.

Is this a problem with xmlrpclib, how I'm using it, or with the server 
decoding my xmlrpc messages?

many thanks,
Michael
_______________________________________________________________
dante at oz.net - Michael Esveldt - #FightThePower on Openprojects


From phd at phd.russ.ru  Tue Apr 11 06:15:38 2000
From: phd at phd.russ.ru (Oleg Broytmann)
Date: Tue, 11 Apr 2000 10:15:38 +0000 (GMT)
Subject: Redirecting the output from sys.stdout
In-Reply-To: <2B1262E83448D211AE4B00A0C9D61B03BA7297@MSGEURO1>
Message-ID: <Pine.LNX.4.21.0004111012310.29384-100000@fep132.fep.ru>

On Tue, 11 Apr 2000, Pieter Claerhout wrote:
> Any examples on how to use the popen of the os module? Never used it before.

import os, string
pipe = os.popen("dir /w", 'r') # Open pipe for reading

lines = []
while 1:
   line = pipe.readline()
   if not line: break # EOF

   lines.append(line)

print string.join(lines, '') # Every line already contains newline, not need it here

Oleg.
---- 
    Oleg Broytmann    http://members.xoom.com/phd2.1/    phd2 at earthling.net
           Programmers don't die, they just GOSUB without RETURN.




From jbauer at rubic.com  Fri Apr 14 10:24:52 2000
From: jbauer at rubic.com (Jeff Bauer)
Date: Fri, 14 Apr 2000 09:24:52 -0500
Subject: Complaint Dept: My shoes!
References: <8d4s4d$mvr$1@nnrp1.deja.com> <38F72CE2.CB7B7C5D@att.net>
Message-ID: <38F72A33.69B244DA@rubic.com>

aflinsch wrote:
> musingattheruins at my-deja.com wrote:
> >
> > If shoe is None then len(shoe) raises an exception.
> > Would be nice if len(None) was 0 and NOT raise an 
> > exeption.
> >
> I think I would prefer it if len(None) returned None. 
> Makes a bit more sense to me having an unsized object 
> return None, rather than zero.

It's no more appropriate for len() to return
a value for None than for it to return a
value for the length of an integer, a function,
or an open socket connection.   In all cases
the appropriate action is to raise a TypeError.

This isn't an argument in favor of purity of
design (or whatever).  It's of practical consequence 
for the interpreter to pass useful information back
to the user (in the form of an exception), rather 
than for the interpreter to assume it understands 
the programmer's intentions.

Jeff Bauer
Rubicon Research



From nick at spam_me_notvideosystem.co.uk  Thu Apr 13 11:58:02 2000
From: nick at spam_me_notvideosystem.co.uk (Nick Trout)
Date: Thu, 13 Apr 2000 16:58:02 +0100
Subject: List mapping?
References: <CC59868A0A76D311B1F50090277B321C1F326B@VRTMAIL>
Message-ID: <38f5efe6@news.xtml.co.uk>

"Mike Fletcher" <mfletch at tpresence.com> wrote in message
news:CC59868A0A76D311B1F50090277B321C1F326B at VRTMAIL...
> Oh, how would I actually do the work of parsing command lines?  Probably use
> the getopt module :)

Oops. Didnt know about that.

Doesnt work with DOS style /'s though  :-/
And my flags have multiple arguments!!
And it seems to stop parsing options once it meets an argument (ie. not a -
option).





From dante at oz.net  Sun Apr 16 17:36:57 2000
From: dante at oz.net (Michael Esveldt)
Date: 16 Apr 2000 21:36:57 GMT
Subject: trapping dict alterations in a class
Message-ID: <8ddbpp$7dp$0@216.39.162.232>

I want to be able to do for a dictionary in a given class what 
__setattr__ does for variable binding.

class foo:
    def __setattr__(self, name, value):
        print name, value
        self.__dict__[name] = value
bar = foo()
bar.quux = 1
bar.baz = {"key":"value"}

The above obviously prints "quux 1" and then "baz {'key':'value'}". But 
what if I want to trap further alterations to the foo.baz dictionary? Is 
this possible?

thanks,
Michael
_______________________________________________________________
dante at oz.net - Michael Esveldt - #FightThePower on Openprojects


From jkahanpa at pcu.helsinki.fi.invalid  Fri Apr  7 08:36:43 2000
From: jkahanpa at pcu.helsinki.fi.invalid (Jere Kahanpaa)
Date: 7 Apr 2000 12:36:43 GMT
Subject: Newbie: How to do an exit ?
References: <38EDC1DA.62CBFCF3@sxb.bsf.alcatel.fr>
Message-ID: <8ckkor$jrt$1@oravannahka.helsinki.fi>

Pascal Fremaux <pf at sxb.bsf.alcatel.fr> wrote:
: How to exit from a program ?
: I don't find a command such as exit(0) in C

The 'sys' module has everything you need
Try

   import sys
   sys.exit()

or

  from sys import exit
  exit()


BTW: This is explained in the tutorial in chapter 2.1 (at least in version 1.5.2). Unfortunately 
the standard Python tutorial is not very user-friendly: it doesn't have any index or searching
tools. Personally used to do a lot or grepping around until I found the clues I needed. 

Happy hacking
	Jere Kahanp??



From mwh21 at cam.ac.uk  Mon Apr 10 15:49:27 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 10 Apr 2000 20:49:27 +0100
Subject: Problem running Python script as cron job
References: <wrd-0AEDDA.12282507042000@news.awenet.com> <38EE41F4.8050708@gmx.net> <wrd-70F5CD.12423810042000@news.awenet.com>
Message-ID: <m3k8i5los8.fsf@atrus.jesus.cam.ac.uk>

"William R. Dickson" <wrd at awenet.com> writes:

> In article <38EE41F4.8050708 at gmx.net>, Per Kistler <kistler at gmx.net> 
> wrote:
> 
> > Hi Bill
> > 
> > Try stderr?
> > /path/prog > /path/log 2>&1
> > 
> > Maybe it spits some errors?
> 
> Tried that, but I still wind up with an empty log.  Anyway to force it 
> to write immediately rather than buffering?

[mwh21 at atrus mwh21]$ python --bleargh
Unknown option: --
usage: python [option] ... [-c cmd | file | -] [arg] ...
Options and arguments (and corresponding environment variables):
-d     : debug output from parser (also PYTHONDEBUG=x)
-i     : inspect interactively after running script, (also PYTHONINSPECT=x)
         and force prompts, even if stdin does not appear to be a terminal
-O     : optimize generated bytecode (a tad; also PYTHONOPTIMIZE=x)
-OO    : remove doc-strings in addition to the -O optimizations
-S     : don't imply 'import site' on initialization
-t     : issue warnings about inconsistent tab usage (-tt: issue errors)
-u     : unbuffered binary stdout and stderr (also PYTHONUNBUFFERED=x)
-v     : verbose (trace import statements) (also PYTHONVERBOSE=x)
-x     : skip first line of source, allowing use of non-Unix forms of #!cmd
-X     : disable class based built-in exceptions
-c cmd : program passed in as string (terminates option list)
file   : program read from script file
-      : program read from stdin (default; interactive mode if a tty)
arg ...: arguments passed to program in sys.argv[1:]
Other environment variables:
PYTHONSTARTUP: file executed on interactive startup (no default)
PYTHONPATH   : ':'-separated list of directories prefixed to the
               default module search path.  The result is sys.path.
PYTHONHOME   : alternate <prefix> directory (or <prefix>:<exec_prefix>).
               The default module search path uses <prefix>/python1.5.

so that'll be `python -u' then ...

Cheers,
M.

-- 
  ... but I guess there are some things that are so gross you just have
  to forget,  or it'll destroy something within you.  perl is the first
  such thing I have known.                 -- Erik Naggum, comp.lang.lisp


From dworkin at ccs.neu.edu  Mon Apr 17 21:36:07 2000
From: dworkin at ccs.neu.edu (Justin Sheehy)
Date: 17 Apr 2000 21:36:07 -0400
Subject: Beginner: string to long
In-Reply-To: hudson_melville@my-deja.com's message of "Tue, 18 Apr 2000 00:35:39 GMT"
References: <8dgakr$2i$1@nnrp1.deja.com>
Message-ID: <vndzoqsdwc8.fsf@camelot-new.ccs.neu.edu>

hudson_melville at my-deja.com writes:

> The outputs are similar to '3564-752+617' How do I convert it so it
> can evaluate into a long integer? Is there a built-in that I don't
> know about that would do this?

>>> eval('3564-752+617')
3429

-Justin

 



From ivanlan at callware.com  Wed Apr 19 18:56:24 2000
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Wed, 19 Apr 2000 16:56:24 -0600
Subject: Browse dialog in PythonWin
References: <38FE32AB.520C1736@callware.com> <38fe3297.36998571@news-server.socal.rr.com>
Message-ID: <38FE3998.1ED7FB24@callware.com>

Hi All, Roger--

Roger Burnham wrote:
> 
[snippery]
> 
> Here's a dialog I adapted from something I found in the PythonWin
> distribution.
> 

[more ruthless snippery]

That will do *very* nicely!  Thank you!

<now-how-do-i-get-these-chompy-places-out-of-my-keyboard?>-ly y'rs,
Ivan;-)
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
http://www.pauahtun.org 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
Author:  Teach Yourself Python in 24 Hours



From josht at iname.com  Sun Apr 30 22:03:18 2000
From: josht at iname.com (Josh Tompkins)
Date: Mon, 01 May 2000 02:03:18 GMT
Subject: Python Compiler
Message-ID: <8F26DEAB4joshtinamecom@209.208.0.7>

This is maybe a dumb question, but here goes: (donning asbestos underwear)

What would be the difficulty of writing a real python compiler?  Or maybe a 
compiler/interpreter similar to the one in VB?  Leaving platform indepenacy 
and other similar concerns aside, what are the technical hurdles?  Is it 
hard to write a compiler?  Given proper pointers to tutorials and 
references I'd give it a shot, if it wasn't above my (low) level of 
programming skill...

Anyway, thanks for your time.

Josue

-- 
________________________________________________________________

"Destined For Great Things -- but pacing myself."
- From a t-shirt.

E-Mail:  josht at crosswinds.net
ICQ:  21219667
AIM:  JosueTheGreat
Web:  http://www.crosswinds.net/~josht
_________________________________________________________________


From tismer at tismer.com  Tue Apr 11 14:31:16 2000
From: tismer at tismer.com (Christian Tismer)
Date: Tue, 11 Apr 2000 20:31:16 +0200
Subject: Modifying "traceback" objects
References: <20000406075124.A1848828@vislab.epa.gov> <m33doslw6f.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <38F36F74.A14349DE@tismer.com>


Michael Hudson wrote:
> 
> Randall Hopper <aa8vb at yahoo.com> writes:
> 
> >      Is it possible to add frames to the current traceback object from
> > inside a C extension?
> 
> I'd say it would be difficult.  What would you put in the "f_code" field?

Just have a look at the builtins "map" code in Stackless Python.
It provides a name, a code object and so forth, and it appears
nicely inside tracebacks.
I think it was very similar to what Michael proposes.

ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaunstr. 26                  :    *Starship* http://starship.python.net
14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     where do you want to jump today?   http://www.stackless.com



From pecora at anvil.nrl.navy.mil  Mon Apr 10 18:22:59 2000
From: pecora at anvil.nrl.navy.mil (Louis M. Pecora)
Date: Mon, 10 Apr 2000 22:22:59 GMT
Subject: Why should I switch to Python?
References: <seas5ia89177@corp.supernews.com> <slrn8f4d0j.1t0.bwinton@tor.dhs.org>
Message-ID: <100420001822595187%pecora@anvil.nrl.navy.mil>

In article <slrn8f4d0j.1t0.bwinton at tor.dhs.org>, Blake Winton
<bwinton at tor.dhs.org> wrote:

> On Fri, 31 Mar 2000 19:43:58 -0800, Aaron Turner wrote:
> >I've been coding Perl for over 3 years and really like it.  (Especially
> >it's syntax and ability to grok my programming style.)
> 
> Now, just as a side note, why is it that the sets of The Perl Hackers I
> Know, and The People I Know Who Use The Word "Grok" are the same set?

Sorry. I don't know Perl. I guess the sets just have some overlap.  I'm
just a physicist formerly of the hippy generation who read Heinelin's
Stranger in a Strange Land which I grokked. :-)


From tim_one at email.msn.com  Thu Apr  6 20:29:15 2000
From: tim_one at email.msn.com (Tim Peters)
Date: Thu, 6 Apr 2000 20:29:15 -0400
Subject: C's ?: expression
In-Reply-To: <38ED28D6.2C5519ED@bioeng.ucsd.edu>
Message-ID: <000801bfa028$4eaa8fc0$34a0143f@tim>

[Curtis Jensen]
> I while back I posted about how to simulate C's ?: expression in
> Python.  There were quite a few responces to it.  Recently, I found on
> pp135-136 of Mark Lutz's "Programmin Python" book that there is a way.
> If anyone is interested:
>
> ((TEST and [trueAction]) or [falseAction])[0]
>
> where TEST is the boolean expression and tueAction is what to do if the
> expression is true and falseAction is what to do if the expression is
> false.

Yes, I invented that, and you could have found it more easily in the Python
FAQ (note that the inner parens aren't needed -- "and" binds tighter than
"or").  However, I have never used it in real code, and have been known to
kill people who do.  You really want to risk that, just to avoid an obvious
if/else <wink>?

better-programming-thru-intimidation-ly y'rs  - tim





From aa8vb at yahoo.com  Sat Apr  1 07:57:17 2000
From: aa8vb at yahoo.com (Randall Hopper)
Date: Sat, 01 Apr 2000 07:57:17 -0500
Subject: Single-line functions
In-Reply-To: <8c2gu2$j2i$1@kopp.stud.ntnu.no>
References: <20000331094943.A1251675@vislab.epa.gov>
 <8c2gu2$j2i$1@kopp.stud.ntnu.no>
Message-ID: <20000401075717.A1359257@vislab.epa.gov>

Magnus Lie Hetland:
 |Seems you can't begin more than one block on one line...
 |
 |For instance,
 |
 |  if 1: if 1: print "spam"
 |
 |doesn't work, although
 |
 |  if 1: print "spam"
 |
 |does.

Thanks.  I figured it was something like that.  Just didn't remember ever
reading that in the Python docs.  Digging further, I see that it is in Mark
Lutz's _Programming Python_.

-- 
Randall Hopper
aa8vb at yahoo.com



From pup at pobox.com  Mon Apr 24 15:31:24 2000
From: pup at pobox.com (Joel Burton)
Date: Mon, 24 Apr 2000 15:31:24 -0400
Subject: Gzip -> FTP w/o using a temporary file
Message-ID: <8e279b$2rf0$1@nnrp-corp.news.cais.net>

I'm writing a small program to help users move a large file. Since I can't
predict what FTP command-line programs they will have, it's being written as
part of a python script.

I have a simple working solution using gzip and ftplib. I create a new file
and gzip it w/Python's gzip lib, then send it using ftplib's ftp. However,
in doing this, I end up creating a real gzipped file.

Since ftplib has me sending a file using a file descriptor, it would appear
that I could gzip -> fd -> FTP, w/o the overhead and space of creating a
real file. However, I'm stuck trying to get this to work.

I tried using os.pipe() to create an r/w pipe, but the 'w' part of the pipe
isn't writable, according to gzip.GzipFile()

Has anyone a clue?

Thanks in advance.





From aahz at netcom.com  Tue Apr  4 16:27:26 2000
From: aahz at netcom.com (Aahz Maruch)
Date: 4 Apr 2000 20:27:26 GMT
Subject: Creating variables on the fly...
References: <meh9-25D0C5.15160303042000@news.cit.cornell.edu> <gl6G4.2926$HG1.87612@nnrp1.uunet.ca>
Message-ID: <8cdj7e$tqr$1@nntp9.atl.mindspring.net>

In article <gl6G4.2926$HG1.87612 at nnrp1.uunet.ca>,
Warren Postma <embed at geocities.com> wrote:
>
>def Sum(dict):
>    n = 0
>    for i in dict.keys():
>        n = n + dict[i]
>    return n

I think you mean:

def Sum(dict):
    n = 0
    for i in dict.values():
        n = n + i
    return n

This should be faster.
--
                      --- Aahz (Copyright 2000 by aahz at netcom.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

"It's 106 miles to Chicago.  We have a full tank of gas, a half-pack of
cigarettes, it's dark, and we're wearing sunglasses."  "Hit it."


From smalleys at gte.net  Sun Apr 30 17:33:21 2000
From: smalleys at gte.net (sue)
Date: Sun, 30 Apr 2000 21:33:21 GMT
Subject: Environmental variable $PYTHONPATH
References: <77363B9E386.AAA2589@smtp1.polyu.edu.hk>
Message-ID: <390CA6BB.37CF0618@gte.net>

I think any app installed under PYTHONHOME is automatically
visible. (I set python up two years ago and haven't had to
touch it. :-)

These both default to your python install dir.
I also put an apps.pth file in the python install directory.

Then I filter the output from dir into it, with one of two
conventions for the filter:

1. that every subdirectory is an app.
2. Or I have a "site_pkg" directory to match the UNIX
structure and apps are under that.

Either works.  And the filter is in Python.  Actually, I
think it's glob.glob() now, instead of dir.

You can read the gory details in site.py in the standard
library.

On the other hand, at least you can read it, unlike
registry-based stuff.

Sue.


From hnowak at cuci.nl  Wed Apr 12 01:12:57 2000
From: hnowak at cuci.nl (Hans Nowak)
Date: Wed, 12 Apr 2000 07:12:57 +0200
Subject: Python 4 DOS
Message-ID: <200004120513.e3C5Dqf26509@hera.cuci.nl>

On 11 Apr 00, at 16:50, Arnaldo Riquelme wrote:

> Can anyone tell me if I can get python binaries for DOS (16-bit).
> I heard of  Hans Nowak 's  Python-DX but the links to his website is broken?
> Where are you Hans?

I'm right here! :) And the URL has changed. Try:

http://www.hvision.nl/~ivnowa/newsite/Python/Python-DX/python-dx.html

Note that Python-DX is unmaintained, and I will probably not upgrade to 
1.6.

Blessed be,

--Hans Nowak (zephyrfalcon at hvision.nl)
Homepage: http://www.hvision.nl/~ivnowa/newsite/
Python questions? See http://tor.dhs.org/~zephyrfalcon/snippets
You call me a masterless man. You are wrong. I am my own master.



From rob at hooft.net  Wed Apr 19 04:22:01 2000
From: rob at hooft.net (Rob W. W. Hooft)
Date: 19 Apr 2000 10:22:01 +0200
Subject: Tkinter/Pmw/Python 1.6a#: xsync?
Message-ID: <m2zoqqcxg6.fsf@hooft.net>

I upgraded too many things at the same time: Computer, OS revision,
Desktop environment, and Python version....

For the python software that I make for my work, I also create a
package including the python interpreter and all extensions required
by my python code.

Since I upgraded to Python-in-CVS (from 1.5.2b1), Pmw0.8.3 (from 0.8)
and started running on RedHat Linux 6.2 in Gnome, I have the
impression that there are sometimes a bunch of X commands that stick
in the pipeline: the buildup of the GUI stalls every time for about 1
second just when the application starts a calculation or becomes
idle waiting for input. Dialog windows do pop up immediately, but
they stay blank for a while.

 - Is there anyone that has seen a similar effect? 
 - Is there a way I can avoid this?

Regards,

Rob Hooft
-- 
=====   rob at hooft.net          http://www.xs4all.nl/~hooft/rob/  =====
=====   R&D, Nonius BV, Delft  http://www.nonius.nl/             =====
===== PGPid 0xFA19277D ========================== Use Linux! =========


From bwilk_97 at yahoo.com  Wed Apr 26 04:52:50 2000
From: bwilk_97 at yahoo.com (Bill Wilkinson)
Date: Wed, 26 Apr 2000 08:52:50 GMT
Subject: Simple list.append() question
References: <meh9-5ED8CD.11192424042000@news.cit.cornell.edu> <ZHZM4.360$qg7.2949@news.uswest.net> <39046FA2.603EAFE5@roguewave.com> <IL3N4.4283$L3.47222@typhoon.midsouth.rr.com> <vM5N4.1522$wYl.188802560@newsb.telia.net>
Message-ID: <C7yN4.5719$L3.54577@typhoon.midsouth.rr.com>

oh. I see.

Even though it is making a copy of the original list. It is a reference to
the same copy three times.

Thanks

:)


"Fredrik Lundh" <effbot at telia.com> wrote in message
news:vM5N4.1522$wYl.188802560 at newsb.telia.net...
| Bill Wilkinson <bwilk_97 at yahoo.com> wrote:
| > l = [[][:]] * 3
|
| same problem as the original code:
|
| >>> l = [[][:]] * 3
| >>> l
| [[], [], []]
| >>> l[0].append(1)
| >>> l
| [[1], [1], [1]]
|
| </F>
|
|




From gmcm at hypernet.com  Thu Apr 27 14:22:01 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Thu, 27 Apr 2000 14:22:01 -0400
Subject: Using struct writing integers
In-Reply-To: <390873b2.726725@news.c2i.net>
Message-ID: <1255262771-67501474@hypernet.com>

thomas at cintra.no wrote:

> Hi,
> 
> I got a list of tuples containing integers, like :
> 
> stuff = [ (21,323,32), (432,43,3234), (432,565,532), (434,43,987)
> ]
> 
> Each tuple in the list represent an id for a entry in a database,
> the three integers form a key. The list above contains four ids.
> I want to write them to a binary file using the struct module.
> Then I want to read them pack, unpack them into the same kind of
> list.
> 
> Any hints, examples etc?

>>> stuff = [ (21,323,32), (432,43,3234), (432,565,532), 
(434,43,987) ]
>>> import struct
>>> f = open('test.out','wb')
>>> fmt = 'lll'
>>> sz = struct.calcsize(fmt)
>>> for x,y,z in stuff:
...   f.write(struct.pack(fmt, x, y, z))
...
>>> f.close()
>>> stuff2 = []
>>> f = open('test.out', 'rb')
>>> while 1:
...  s = f.read(sz)
...  if not s:
...    break
...  tpl = struct.unpack(fmt, s)
...  stuff2.append(tpl)
...
>>> stuff2
[(21, 323, 32), (432, 43, 3234), (432, 565, 532), (434, 43, 987)]

- Gordon



From python-list at teleo.net  Thu Apr 20 13:58:27 2000
From: python-list at teleo.net (Patrick Phalen)
Date: Thu, 20 Apr 2000 10:58:27 -0700
Subject: PyApache memory leak?
In-Reply-To: <8dni3s$1o1$1@nnrp1.deja.com>
References: <8dimg4$kii$1@nnrp1.deja.com> <8dni3s$1o1$1@nnrp1.deja.com>
Message-ID: <00042011042800.02559@quadra.teleo.net>

[spenrose at my-deja.com, on Thu, 20 Apr 2000]
:: We commented out PyApache in httpd.conf. Two days and six hours
:: later...the server did not freeze. We've been up for 11 days and
:: counting and that was the only change we made. I freely acknowledge that
:: this is not definitive proof of a showstopper in PyApache, but I am not
:: up to reading thousands of lines of C to pursue other theories.

I haven't used PyApache, but am under the impression that it spawns
multiple interpreter instances, which could certainly cause memory
problems on a busy, CGI-intensive site.

Have you looked at Httpdapy (soon to be renamed mod_python)?
http://www.ispol.com/home/grisha/httpdapy/




From aa8vb at yahoo.com  Wed Apr 12 11:40:01 2000
From: aa8vb at yahoo.com (Randall Hopper)
Date: Wed, 12 Apr 2000 11:40:01 -0400
Subject: Python idiom: Multiple search-and-replace
In-Reply-To: <871z4bqtbg.fsf@orge.winnegan.de>
References: <20000412100816.A2431031@vislab.epa.gov>
 <871z4bqtbg.fsf@orge.winnegan.de>
Message-ID: <20000412114001.C2446939@vislab.epa.gov>

Siggy Brentrup:
 |Randall Hopper:
 |> I want to do search-and-replace of multiple symbols on each line of a file.
 |> But the simple-minded code below takes a while.
...
 |> --------------------------------------------------------------------------
 |> 
 |>   symbol_map = { 'oldsym1' : 'newsym1', oldsym2' : 'newsym2', ... }
 |>   fp = open( net_path, "r" )
 |> 
 |>   while 1:
 |>     line = fp.readline()
 |>     if not line: break
 |> 
 |>     for old_sym in symbol_map.keys():
 |>       line = string.replace( line, old_sym, symbol_map[ old_sym ] )
 |> 
 |>     out_fp.write( line )
 |> 
 |> --------------------------------------------------------------------------
 |
 |symbol_items = symbol_map.items()
 |
 |while 1:
 |    line = fp.readline()
 |    if not line: break
 |
 |    for old, new in symbol_items:
 |        line = string.replace(line, old, new)
 |
 |    out_fp.write(line)


Actually I'd tried this.  About the same speed (a few seconds slower).  The
dictionary lookups were not slowing the original algorithm.  But thanks.

-- 
Randall Hopper
aa8vb at yahoo.com



From cbbrowne at news.hex.net  Mon Apr  3 20:41:08 2000
From: cbbrowne at news.hex.net (Christopher Browne)
Date: Tue, 04 Apr 2000 00:41:08 GMT
Subject: Smalltalk on the small end (was: Advice requested: GUI project beginning)
References: <m12b4OW-000wcEC@swing.co.at> <8D79BC9E4960ED70.7B6609342D6EB723.7625773D14AD9140@lp.airnews.net>
Message-ID: <ESaG4.9536$9g4.372176@news5.giganews.com>

Centuries ago, Nostradamus foresaw a time when Cameron Laird would say:
>In article <m12b4OW-000wcEC at swing.co.at>,
>Christian Tanzer <tanzer at swing.co.at> wrote:
>			.
>			.
>			.
>>Recently, I find myself writing Python scripts of 5 to 20 lines for
>>which I normally would have used the unix shell. I wouldn't use
>>Smalltalk for this purpose.
>But others do (particularly with GNU Smalltalk),
>so this isn't conclusive.
>
>Actually, there's another thread here asking to
>be launched.  Python (and Tcl, too, in a somewhat
>different fashion is (are) really quite extraor-
>dinary for shell-level scripting and operation.
>I find myself wondering more and more how people
>get along without such (*particularly* on the
>conventional desktops--Win* and MacOS).  I do
>think Python is clearly superior to Smalltalk
>for those one-page scripts that most people do
>with /bin/sh; still, I'm willing to consider that
>Smalltalk is more capable in this regard than I
>yet realize, and, in any case, I can well imagine
>Smalltalk proponents favoring it over /bin/sh.

The "power" of Smalltalk comes not when you're writing one page shell
scripts, but when you have a bunch of "objects that do useful things"
that can all run in a single Smalltalk environment.

The point is that there's little point to "scripting" in Smalltalk
when it doesn't really get powerful 'til you build up a rich set of
functionality in a "Smalltalk Environment."  And *that* is something
you don't want to respawn repeatedly...

Similar could be said about scripting in Common Lisp.  There are CL
implementations (notably CLISP) that can load fairly quickly, but the
merits of writing script-like applications in CL would come not in
having a bunch of little scripts invoked via
   #!/usr/bin/clisp
but rather by building a somewhat larger environment that expresses
those "little scripts" as Lisp functions that can all run within a
single CL instance.

The same is probably also true for attempts to build applications in
Java; you don't want to keep spawning more JVMs, but rather to invoke
a bunch of applications from within the same one.
-- 
Developmental Psychology
"Schoolyard behavior resembles adult primate behavior because "Ontogeny
Recapitulates Phylogeny" doesn't stop at birth."
-- Mark Miller
cbbrowne at hex.net- <http://www.ntlug.org/~cbbrowne/lsf.html>


From n8gray at earthlink.net  Wed Apr 12 14:50:47 2000
From: n8gray at earthlink.net (Nathaniel Gray)
Date: Wed, 12 Apr 2000 18:50:47 GMT
Subject: CGI for voting
Message-ID: <38F4C64C.8AD133CA@earthlink.net>

Can somebody point me to a CGI script in Python for a voting form that
can enforce "one user, one vote"?  I'm a newbie to CGI scripting and
Python, so simplicity is more important to me than functionality.

Thanks!
-- 
Nathaniel A. Gray
--
"But the sun is going down!"
"No, no, you're all confused.  The horizon is moving up."
     -The Firesign Theatre
--
PGP Key:
http://certserver.pgp.com:11371/pks/lookup?op=get&search=0x95345747
For PGP: http://www.pgpi.com/


From sjmachin at lexicon.net  Thu Apr 13 18:46:13 2000
From: sjmachin at lexicon.net (sjmachin at lexicon.net)
Date: Thu, 13 Apr 2000 22:46:13 GMT
Subject: an intern-like memory saver
Message-ID: <8d5ing$heb$1@nnrp1.deja.com>

Problem:
I have an application that works with words and can have millions of them
in memory at one time. Apart from the main data structure, a dictionary is
used to maintain frequencies. As the words are loaded from files, multiple
instances of the same word don't share memory. The memory savings
could be huge --- see how frequent "the" is in English text, or "Smith" in
an Anglo telephone directory.

Trial solution:
(1) Clone dictobject.c. Make it into an extension module for a type called
"mydict". Add a method called "key_ref".with one argument:
adict.key_ref(obj). If adict.has_key(obj) is true, this returns a reference to
the key value inside the dictionary; else it returns "obj".
(2) Make simple changes to the application:
(a) Change
   freq_dict = {}
to
   freq_dict = mydict.mydict()
(b) assuming for purposes of exposition that words are stored simply in a
list, after
   freq_dict[w] = freq_dict.get(w, 0) + 1
change
   word_list.append(w)
to
   word_list.append(freq_dict.key_ref(w))

Results:
Gratifying. An exercise that was running out of real memory (384 MB) and
taking a day now takes an hour or so.

Questions:
(1) Would this be sufficiently generally useful to make it a method of the
standard dictionary object in Python?
(2) As methods seem to be found by sequential search, wouldn't it be a
good idea to move "get" a bit higher up the method_def list in
dictobject.c? At the end, after "update" and "copy", doesn't seem like a
good idea.
(3) Has anyone had any success in compiling Python on WinNT 4.0 with
gcc 2.95.2? It's just fine for making extension modules; I haven't tried
compiling the whole Python yet.
(4) Has anyone any better ideas for gauging Python memory usage than
sitting watching the graphical display in WinNT's Task Manager? Might an
instrumented or instrumentable malloc/free package (like Doug Lea's) that
permitted implementation of a Python builtin memused() be the way to go,
or is there a policy of using the standard malloc from the C library on each
platform?


Sent via Deja.com http://www.deja.com/
Before you buy.


From scarblac-spamtrap at pino.selwerd.nl  Thu Apr 13 13:32:40 2000
From: scarblac-spamtrap at pino.selwerd.nl (Remco Gerlich)
Date: 13 Apr 2000 17:32:40 GMT
Subject: List mapping?
References: <Pine.LNX.3.96.1000413174501.3264B-100000@valdez.netg.se> <newscache$autysf$ye8$1@jenpc07.jennic.co.uk>
Message-ID: <slrn8fc4eu.1d0.scarblac-spamtrap@flits104-37.flits.rug.nl>

Robert Cragie wrote in comp.lang.python:
> > Howcome you use reduce at all?
> >
> > Works fine with just doing
> >
> > tuplist = map(lambda x: (mylist[x], mylist[x+1]), range(0, len(mylist),
> 2))
> 
> Yep, that's neater. My excuse is that I've only been learning Python for a
> couple of weeks...

In that case, forget you ever read something about map, lambda and reduce.
Save it for later and code nice for loops now. 95% of the time, those things
only obfuscate code anyway.

-- 
Remco Gerlich,  scarblac at pino.selwerd.nl

   This is no way to be
     Man ought to be free      -- Ted Bundy
       That man should be me


From vsuvorov at cccglobal.com  Wed Apr 12 13:15:48 2000
From: vsuvorov at cccglobal.com (Vadim Suvorov)
Date: Wed, 12 Apr 2000 12:15:48 -0500
Subject: static attributes & __getattr__ override
Message-ID: <38f4b2e0@news.raccoon.com>

Hello,

I am trying to redefine class attributes access, to include static
attributes definitions from class (shared by all class instances). It is ok
for instance to shade class attribute. So I wrote

class MyClass:
    var1 = 1
    var2 = 2
    def __init__(self, x3):
        self.var3 = x3

    def __getattr__(self, name):
        return self.__class__.__dict__.get(name, None)

MyObj = MyClass(3)
print MyObj.var1, MyObj.var2, MyObj.var3, MyObj.var4

(expected result) >>> 1 2 3 None

It works fine for attributes. However, I can not access standard methods of
MyClass. I resorted to define __str__ and __repr__ (to be able to 'print
MyObj'), and even then I could not use MyObj <> None which really hurts me.

Can you advice me on best way to proceed, and explain briefly what is going
on? I do not think it is a bug, I probably missing something in the way
attributes/methods work.

Thanks,

Vadim




From sabren at manifestation.com  Wed Apr 26 06:57:21 2000
From: sabren at manifestation.com (Michal Wallace (sabren))
Date: Wed, 26 Apr 2000 06:57:21 -0400 (EDT)
Subject: [ANNOUNCE] weblib
Message-ID: <Pine.BSF.4.10.10004260651430.23888-100000@chi.pair.com>

Weblib is a library for Python programmers who want to build
interactive websites. It offers objects that are similar in feel to
those found in Active Server Pages and the PHP base libraries. More
specifically, weblib provides the following objects:

   * request : manages cookies, forms, and query strings coming from the browser. 
   * response : manages your script's response. 
   * sess : manages persistent sessions. 
   * auth : provides an authentication framework. 
   * perm : provides permission handling. 

In addition, weblib provides a wrapper module that handles most of the
CGI work and provides the above services transparently to your
scripts.

----------

The project is still pre-Alpha, but there's documentation and CVS access
over at sourceforge:

http://weblib.sourceforge.net/

Any feedback would be much appreciated.


Cheers,

- Michal
-------------------------------------------------------------------------
http://www.manifestation.com/         http://www.linkwatcher.com/metalog/
-------------------------------------------------------------------------




From fw at deneb.cygnus.argh.org  Tue Apr 11 13:46:06 2000
From: fw at deneb.cygnus.argh.org (Florian Weimer)
Date: 11 Apr 2000 19:46:06 +0200
Subject: [ANNOUNCE] Garbage collection for Python
References: <20000407221750.A29279@acs.ucalgary.ca> <38F13FB9.D1D16AF@mindspring.com> <slrn8f2if1.dop.nascheme@cranky.arctrix.com> <87r9cdhdb0.fsf@deneb.cygnus.argh.org> <m3bt3hkk76.fsf@atrus.jesus.cam.ac.uk> <87bt3geszz.fsf@deneb.cygnus.argh.org> <m3u2h8kf5q.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <87itxocyzl.fsf@deneb.cygnus.argh.org>

Michael Hudson <mwh21 at cam.ac.uk> writes:

> 1) Neil's scheme does not replace reference counting.

Oh.

> 2) There are more things between the status quo and a free threaded
>    interpreter than reference counting issues.

Yes, of course.  But the reference counting appears to be quite a
challange (at least if you haven't got an atomic load-increment-store
operation).

> So the answer to your question is: "no".

I see.


From gavrilov at iname.com  Mon Apr 10 13:19:23 2000
From: gavrilov at iname.com (Alexander Gavrilov)
Date: Mon, 10 Apr 2000 17:19:23 GMT
Subject: The gzip module... neverending story
References: <8cr7os$du3$1@lacerta.tiscalinet.it>
Message-ID: <v2oI4.15946$k5.306494@news1.frmt1.sfba.home.com>

Look at zipfile.py in ftp://ftp.interet.com/pub/pylib.html

"FrancescO" <francesco.contini at tiscalinet.it> wrote in message
news:8cr7os$du3$1 at lacerta.tiscalinet.it...
> Hello!,
> I have, another questions!!
> How can I zip i list of files? (i have done it, but... the files can't be
> readed from winzip and gunzip...)
>
> Wich is the correct usage of the fileobj (4th parameters in the
> gzip.GzipFile.. explain..:
> GzipFile ([filename[, mode[, compresslevel[, fileobj]]]]) )?
>
> Is the name of the zipfile? what kind of object is it?
> i'have try in a lot of ways to call this object, but I'm not "good" to
work
> it with the 4th parameter please...
>
> HELP ME...!!!!
>
>
>
>




From rnog2438 at mail.usyd.edu.au  Sun Apr 16 11:40:44 2000
From: rnog2438 at mail.usyd.edu.au (Ricardo Nogueira)
Date: Mon, 17 Apr 2000 01:40:44 +1000
Subject: Event triggering and weak references
References: <8dbif3$hm0$1@metro.ucc.usyd.edu.au> <MPG.1363441c166d9ca98969e@news-server>
Message-ID: <8dcmv0$7ub$1@metro.ucc.usyd.edu.au>

Robert Getter <rgetter at usa.net> wrote in message
news:MPG.1363441c166d9ca98969e at news-server...
> In article <8dbif3$hm0$1 at metro.ucc.usyd.edu.au>,
> rnog2438 at mail.usyd.edu.au says...
> > First thanks for this beautiful language : )
> >
> > I am Smalltalk addicted : )  and this means that I can't do with any
other
> > language, ex. C, C++, Java,  : (
> > This was true until I found Python 2 weeks ago at Sydney Uni. I loved
its
> > modularity, pluggin behaviour and freedom.
> > But I am missing Smalltalk event handling... Smalltalk (like Python)
support
> > the call-Back mechanism, but there is a more powerful inter-object
interest
> > registering that I haven't found in Python (so far).
> >
> > This is what I would like to do:
> > Imagine I could have two or more Widgets open displaying a car speed
value,
> > say a numerical and a graphic display, if this car speed changes (ex.
some
> > one pushes the break), I would like to have any open widget on this
value
> > updated (on opening each window would register interest with this car
> > speed).
> >
> > Smalltalk implements this with:
> >
> > The widget would register interest in one aspect of the car  (on opening
> > over aCar):
> >   aCar when:#speedChanged send: #updateSpeedDisplay to: self
> >
> > And the car would just notify who is interested, if any (when its speed
> > changes):
> >   self trigger: #speedChanged
> >
> > But if any of this windows is not referenced anymore or crashes I would
like
> > the garbage collector clear it (the car would not be notified and his
> > pointer to the window shouldn't keep the garbage collector form freeing
the
> > window memory).
> >
> > Again, Smalltalk implements this with weak references, that doesn't
count
> > for the garbage collector.
> >
> > Is there support for this in Python?
> >
> > Any suggestions?
> >
> >
> >
>
> Here's some code which does some of what you want:
> =========
> # simple test of a scheme for registering interest in an event.
> # exception safety is left as an exercise for the reader
>
> class watcher:
> def __init__(self, watchee, voyeur=None):
> self.watchee = watchee
> if voyeur == None:
> self.list = []
> else:
> self.list = [ voyeur ]
>
> def __call__(self, *args, **kwargs):
> rc = apply(self.watchee, args, kwargs)
> for voyeur in self.list:
> apply(voyeur, args, kwargs)
> return rc
>
> def register(self, voyeur=None):
> if not voyeur: return
> self.list.append(voyeur)
>
> def unregister(self, voyeur=None):
> if voyeur: self.list.remove(voyeur)
> return len(self.list)
>
>
> def register(object, method, voyeur=None):
> # try to add this voyeur to the watcher object
> try:
> getattr(object, method).register(voyeur)
> # if this fails, there is no watcher object. Add one.
> except:
> try:
> setattr(object, method,
> watcher(getattr(object, method), voyeur))
> # if this fails, there is no method of that name. return
> None.
> except:
> return None
> # finally, return the newly installed watcher method
> return getattr(object, method)
>
> def unregister(object, method, voyeur=None):
> try:
> if not getattr(object, method).unregister(voyeur):
> # list is empty. remove watcher
> setattr(object, method,
> getattr(object, method).watchee)
> except: pass
>
> class test:
> def x(self): print "x()"
>
> def watch1():
> print "watch1()"
>
> def watch2():
> print "watch2()"
>
> def main():
> a = test()
> a.x()
> print
> register(a, "x", watch1)
> a.x()
> print
> register(a, "x", watch2)
> a.x()
> print
> unregister(a, "x", watch1)
> a.x()
> print
> unregister(a, "x", watch2)
> a.x()
>
> if __name__ == "__main__": main()
> ========
>
> What this code does is install a wrapper around the method you want to
> watch. This wrapper maintains a list of interested parties. There are a
> few problems with this method:
>
> 1. No weak references. I think there is a weak reference extension
> somewhere which could be added without too much trouble.
>
> 2. In Smalltalk, AFAIK, a method can only be referenced through its
> object. In Python, each method is a callable object which can be
> manipulated separately. In other words, object methods in Python are
> really instance variables which just happen to be callable. As a result,
> anywhere that objects pass bound methods around won't work right. If a
> bound method is passed before the wrapper is installed, calls to the
> method will not be sent to the watchers. This means that watchers can't
> be reliably watched, since they work by getting references to themselves
> put on a list in the watcher class.
>
> The only real solution I can think of for problem 2 is a generic
> mechanism for attaching watchers to any callable object. This is similar
> to what a debugger does. If you look into the debugging API, you may find
> something that is suitable or can be modified.
>
> I never much thought about it before, but this could be a really useful
> capability to have.
>
> Another way that this could be done which would also allow other nice
> things to happen would be the ability to globally change all references
> to one object into references to a different object.
> --
> Rob Getter

This is amazing! If I were this object "a" in your main method, I would feel
like a child that have his "coca-cola" swapped to a "cuba"! I do really have
lots to learn in this "do-what-ever-you-want-to-someones-object" filosofy. I
will keep this wrapper around the method in mind.
I implemented this observer pattern with a class that register interest from
any boundedMethod to be called when a named event occurs. Then the object
that inherits from this class has the option of triggering these events when
it wants. The advantage is more encapsulation, with colaboration between
objects. The disadvantage is that the trigger has to predict what will be
wanted from him in advance.
But both versions work with boundedMethods for the observer. If you change
your main() to register(a, "x" , objY.msg) it will keep objY from being
garbageCollected if we clear any other references to it. It is necessary to
always call unregister before the "last" pointer to objY is cleared. Not
very nice.

How can we read objY reference count?
If we come to know this number, compare to the number of boundedMessages we
have for it, we can conclude if it is garbage and clear its references
instead of calling its methods.

Still, weak references would do the job beautifully... : )

-----
RiNo




From maybe at lucky.com  Tue Apr 11 03:56:18 2000
From: maybe at lucky.com (Soundwave)
Date: Tue, 11 Apr 2000 09:56:18 +0200
Subject: readline skips on line
Message-ID: <8culp8$23a$1@porthos.nl.uu.net>

I got a part of a programm looking like this

f=open('test.txt', 'r')
while f.readline()!='':
    print f.readline()
print 'ready'

but when it is runned, it wil print out the lines in the text file, but it
skips all the the odd line, it only displays line 2,4,6,8,10 and further

How can I solve this problem ?




From arnold at dstc.edu.au  Sat Apr  1 19:48:57 2000
From: arnold at dstc.edu.au (David Arnold)
Date: Sun, 02 Apr 2000 10:48:57 +1000
Subject: New Features in Python 1.6 
In-Reply-To: Your message of "Sat, 01 Apr 2000 22:00:59 GMT."
             <vkuF4.4584$74.73299@newsc.telia.net> 
Message-ID: <200004020048.KAA31167@piglet.dstc.edu.au>

-->"Fredrik" == Fredrik Lundh <effbot at telia.com> writes:

  Fredrik> I think things are worse than this -- someone just sent me
  Fredrik> something that looks like a draft press release, announcing
  Fredrik> that O'Reilly had taken over Python development, with Jon
  Fredrik> Orwant as CTO, and Larry Wall as the Python 1.6 release
  Fredrik> manager.

and the date on the release was ???



d



From sabren at manifestation.com  Tue Apr 18 23:39:12 2000
From: sabren at manifestation.com (Michal Wallace (sabren))
Date: Tue, 18 Apr 2000 23:39:12 -0400 (EDT)
Subject: module naming?
In-Reply-To: <38FD1A1D.D59B594B@kraken.itc.gu.edu.au>
Message-ID: <Pine.BSF.4.10.10004182334460.8776-100000@chi.pair.com>

On Wed, 19 Apr 2000, Joal Heagney wrote:

 
> I'm new to the pythoning game, but is there a way to check to see if a file
> is loaded. I seem to remember that the from statement doesn't support that,
> but maybe you could make a call to a function that does a check on if the
> class already exists in program space, and if it doesn't imports it using
> the from ... import statement?

here's what I wound up doing (in weblib/__init__.py)

#################

def Request():
    import Request, weblib
    weblib.Request = Request.Request
    return Request.Request()

##################

So, the first time I call it I'm really calling a function that calls
a constructor and returns the new object.. and the second time, I call
the constructor directly.

Cheers,

- Michal
-------------------------------------------------------------------------
http://www.manifestation.com/         http://www.linkwatcher.com/metalog/
-------------------------------------------------------------------------




From effbot at telia.com  Sat Apr  1 09:23:24 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Sat, 01 Apr 2000 14:23:24 GMT
Subject: why did string.py change in 1.6?
References: <8c4mdr$5p7$1@slb6.atl.mindspring.net>
Message-ID: <wDnF4.4506$74.72019@newsc.telia.net>

Andrew Dalke <dalke at acm.org> wrote:
> In the CVS version (1.6a1) I see that string.py has been
> The function strop.join still exists (so there is duplicated
> code between that and the string module).  Given the extra
> method call overhead, can anyone tell me why this change was
> made?

python 1.6 has two string types.  the 'strop' module
only supports one of them.

(the 'strop' module is obsolete.  the 'string' module should be
considered as "mostly deprecated", at least for anything that
is better done via string methods :-)

</F>




From wpostma at earthlink.net  Thu Apr  6 19:51:07 2000
From: wpostma at earthlink.net (Warren Postma)
Date: Thu, 06 Apr 2000 23:51:07 GMT
Subject: "tostring" as a string method?
References: <8c45ue$b7u$1@slb1.atl.mindspring.net>
Message-ID: <Lp9H4.3526$y4.52453@newsread1.prod.itd.earthlink.net>

>   Here's an example, which is used to translate DNA to
> protein:
>
> def translate(dna, table = Translation.standard_table):
>   s = dna    # or "dna.tostring()" for array.array
>   x = []
>   for i in range(0, len(s), 3):
>     x.append(table[s[i:i+3]])
>   return string.join(x, "")
>

Why not write str(dna) and overload __str__ where necessary?


# polymorphism exhibited for all types, conversion to string:

class myClass:
 def __init__(self,value='something'):
  self.value ='something'
 def __str__(self):
  return self.value
myObject = myClass()

a = [ 5, 5.5, myObject, 'test']

for i in a:
 print str(i)


Warren




From meh9 at cornell.edu  Thu Apr 13 14:56:38 2000
From: meh9 at cornell.edu (Matthew Hirsch)
Date: Thu, 13 Apr 2000 14:56:38 -0400
Subject: Simple formatting string question
Message-ID: <meh9-637743.14563813042000@news.cit.cornell.edu>

Hi All,

I'm trying to print something like this:

>>> print 6*'%2i' '%1i' % tuple([5,5,5,5,5,5,1])

but I'm getting an error.

Do you know what the correct syntax should be?

Thanks,
Matt


From tismer at tismer.com  Fri Apr 28 09:21:45 2000
From: tismer at tismer.com (Christian Tismer)
Date: Fri, 28 Apr 2000 15:21:45 +0200
Subject: Simple list.append() question
References: <meh9-5ED8CD.11192424042000@news.cit.cornell.edu> <ZHZM4.360$qg7.2949@news.uswest.net> <39046FA2.603EAFE5@roguewave.com> <PsaN4.174166$Dv1.2059732@news1.rdc1.bc.home.com> <8e4akc$p7o$1@newshost.accu.uu.nl> <m366t6ozqx.fsf@greebo.nodomain.de> <3908E6EC.B9E6B2E8@cosc.canterbury.ac.nz> <m3zoqenzt6.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <39099069.7BFD8387@tismer.com>


Michael Hudson wrote:
> 
> Greg Ewing <greg at cosc.canterbury.ac.nz> writes:
> 
> > Bernhard Herzog wrote:
> > >
> > > It would be a good idea IMO to find a
> > > good syntax for this case and ISTM that list comprehensions might be
> > > useful here:
> > >
> > >   multi_list = [[] for i in range(10)]
> > >
> > > or whatever the precise syntax was. Hmm, the unused variable in there is
> > > a bit ugly though.
> >
> > Maybe the LC syntax should include something like
> >
> >    [[] times 10]
> 
> But wouldn't that make "times" a keyword?  Not sure I'd want that...

Currently, yes, but it is absolutely not necessary to be so.
It requires to change the grammar to treat operators
differently. A similar change would allow me to use

def foo(from=1, to=42):
   ...

since there is no reason to recognise from as a keyword in
parameter position. Similarily: in

class foo:
   def impossible(self, a, b, c):
      self.pass = a
      self.continue = b
      self.return = c

could quite easily be allowed.

But I'm tired to fight such a proposal again for no result...

ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaunstr. 26                  :    *Starship* http://starship.python.net
14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     where do you want to jump today?   http://www.stackless.com



From dalke at acm.org  Sat Apr 15 17:11:01 2000
From: dalke at acm.org (Andrew Dalke)
Date: Sat, 15 Apr 2000 15:11:01 -0600
Subject: Scripting and Gnome and KDE
References: <Pine.GSO.4.10.10004150741520.11261-100000@sundial> <8d95m0$9j6$1@news1.xs4all.nl>
Message-ID: <8damat$4ja$1@slb2.atl.mindspring.net>

Boudewijn Rempt said:
> when I read the article on PyGTK
> in Dr Dobbs I can only be glad I choose for PyQt and PyKDE. It painted
> a very bleak picture...


Even though it says "The result is a great environment for rapid
development of GUI-based applications." ?

I used to work at Bioreason, though I wasn't involved in the GUI
development.  They are pretty happy about PyGTK.  The idea of the
DDJ article was to point out how to take advantage of the toolkit,
based on finding out what doesn't work.

You could write a similarly bleak picture of any toolkit.

Take the last section, "Troubleshooting."  It says the method
and parameters names aren't always sufficient, so you need a
copy of the documentation.  But the docs don't keep up with the
code so you need the code as well.  The same text can be applied
to Python.

                    Andrew
                    dalke at acm.org





From robin at jessikat.demon.co.uk  Fri Apr 14 15:51:44 2000
From: robin at jessikat.demon.co.uk (Robin Becker)
Date: Fri, 14 Apr 2000 20:51:44 +0100
Subject: Python idiom: Multiple search-and-replace
References: <20000412100816.A2431031@vislab.epa.gov>
 <5C%I4.645$rc9.190288896@newsb.telia.net>
Message-ID: <GFZLQKAQb394Ewgs@jessikat.demon.co.uk>

In article <5C%I4.645$rc9.190288896 at newsb.telia.net>, Fredrik Lundh
<effbot at telia.com> writes
>Randall Hopper <aa8vb at yahoo.com> wrote:
>> Is there a Python feature or standard library API that will get me less
>> Python code spinning inside this loop?   re.multisub or equivalent? :-)
>
>haven't benchmarked it, but I suspect that this approach
>is more efficient:
>
>...
>
># based on re-example-5.py
>
>import re
>import string
>
>symbol_map = { "foo": "FOO", "bar": "BAR" }
>
>def symbol_replace(match, get=symbol_map.get):
>    return get(match.group(1), "")
>
>symbol_pattern = re.compile(
>    "(" + string.join(map(re.escape, symbol_map.keys()), "|") + ")"
>    )
>
>print symbol_pattern.sub(symbol_replace, "foobarfiebarfoo")
>
>...
>
></F>
>
...
I'm trying to use the above idiom for the strings '(' ')' '\\' but seem
to have a speed problem ie

import re
import string
from time import time

symbol_map = { '(': '\\(', ')': '\\)', '\\': '\\\\' }
def symbol_replace(match, get=symbol_map.get):
    return get(match.group(1), "")
PAT = "(" + string.join(map(re.escape, symbol_map.keys()), "|") + ")"
print PAT
symbol_pattern = re.compile( PAT )

def doit(str,n):
        N = xrange(n)
        t0 = time()
        for n in N:
                s=string.replace(str,'\\','\\\\')
                s=string.replace(s,'(','\(')
                s=string.replace(s,')','\)')
        t1 = time()
        print 'string %s %.4f' % (s,t1-t0)
        t0 = time()
        for n in N:
                t = symbol_pattern.sub(symbol_replace, str )
        t1 = time()
        print 're     %s %.4f' % (t,t1-t0)


doit("a b (c) \\t",10000)

produces 
(\\|\(|\))
string a b \(c\) \\t 0.4900
re     a b \(c\) \\t 5.3900

ie for this simple case string is best. Could I do better?
-- 
Robin Becker


From zhuravle at us.ibm.com  Mon Apr 17 16:30:15 2000
From: zhuravle at us.ibm.com (zhuravle at us.ibm.com)
Date: 17 Apr 2000 16:30:15 EDT
Subject: Resume
Message-ID: <44311C77F56CD31187F40090279C687A08366D@RUSSELL>


Thank you and be careful -- there is a severe weather alert in Washington
area

Polina Zhuravlev
IBM Global Services
I/T Specialist

Phone:  (614)659-7347   T/L: 657-7347
Fax:  (614)659-7300
Pager: (800)759-8888 PIN 1364734
E-mail:  zhuravle at us.ibm.com




From kern at caltech.edu  Sat Apr  1 15:27:46 2000
From: kern at caltech.edu (Robert Kern)
Date: 1 Apr 2000 20:27:46 GMT
Subject: Python/GTK on Windows?
References: <#XSVXB5m$GA.240@cpmsnbbsa03>
Message-ID: <8c5m42$1dv@gap.cco.caltech.edu>

In article <#XSVXB5m$GA.240 at cpmsnbbsa03>,
	"lexberezhny" <lexberezhny at email.msn.com> writes:
> Hi,
>   I was wondering if anyone was able to get PyGTK to work on windows, and if
> its possible at all. I am trying to port a Linux app to windows that uses
> PyGTK. If you know anything about this or can point me in the right
> direction, I would really appreciate it. Thanks.

http://theopenlab.uml.edu/pygtkwin/

>  - Lex Berezhny

-- 
Robert Kern
kern at caltech.edu

"In the fields of hell where the grass grows high
 Are the graves of dreams allowed to die."
  -- Richard Harter


From milesthompson at sprint.ca  Thu Apr 20 19:15:40 2000
From: milesthompson at sprint.ca (Miles Thompson)
Date: Thu, 20 Apr 2000 20:15:40 -0300
Subject: How2 use curses for data input
References: <38FF6DE2.5751423A@sprint.ca> <slrn8fusjd.99a.grant@grante.comtrol.com>
Message-ID: <38FF8F9C.9D533E74@sprint.ca>

Grant,

I located it on the RH system here in the office at home. The target machine
is Debian, and I think I can install RPM on it so that if need be, I can get
the RPM. Debian doesn't seem to have a snack package.

Thanks for the lead, much appreciated. I've searched google, www.python.org,
and www.debian.org for information on how to use it. All that's on my system
is snack.py, .pyc and _snackmodule.so. I *hoped* there might be some clues
in __builtin__ or in snakc.py, but I drew a blank.

Do you have any suggestions where I might dig something up? (I had really
hoped that RH might have something, but nope!)

Regards - Miles

Grant Edwards wrote:

> In article <38FF6DE2.5751423A at sprint.ca>, Miles Thompson wrote:
>
> >I 've read the curses part of the manual, and the HOWTO, but I'm not
> >much wiser. I need to capture some user input (name of a new directory
> >for archiving some web files), and I understand how to define a window,
> >etc. What I don't know to do is how to edit the lines, and loop from
> >last to first.
>
> You might want to take a look at the "snack" module, which is a
> wrapper for the newt library (which is based on slang rather
> than curses).  It impliments a set of text-mode widgets such as
> buttons, scroll lists, data entry fields, checkboxes, etc.  You
> stick a bunch of them on a "form" and pop it up.  When the user
> is done, you get the values back somehow (I forget exactly).
>
> If all you need to do is prompt the user for some data, it
> should do the trick.
>
> Redhat's text-mode installer is (or at least was) written in
> Python using snack/newt.
>
> --
> Grant Edwards                   grante             Yow!  Why was I BORN?
>                                   at
>                                visi.com



From andre at beta.telenordia.se  Sun Apr  2 07:59:43 2000
From: andre at beta.telenordia.se (André Dahlqvist)
Date: Sun, 02 Apr 2000 12:59:43 +0100
Subject: Emacs Python mode
References: <yayk8iii0ex.fsf@ganglot.ifi.uio.no><m3zorendyk.fsf@atrus.jesus.cam.ac.uk><yayzorcj1ox.fsf@ganglot.ifi.uio.no>
Message-ID: <8c794n$mjq$1@zingo.tninet.se>

> all in all, *emacs* has lots of shortcuts for handy things. I wish I
> knew more about it.. :)

While we're on the subject of Emacs Python mode, why is it that I can't 
indent a line under a line which has been commanted out? Emacs 
refuses to let me tab in there, but if I remove the comment it will happily 
let me do it.

// Andr?




From tiddlerdeja at my-deja.com  Fri Apr 28 08:24:57 2000
From: tiddlerdeja at my-deja.com (tiddlerdeja at my-deja.com)
Date: Fri, 28 Apr 2000 12:24:57 GMT
Subject: sleep function?
Message-ID: <8ebvud$l6h$1@nnrp1.deja.com>

Does python have a sleep function? e.g. sleep(60) to wait 60 seconds?

I realise this is a lame question, but I have no python documentation
around just now.

Thanks.


Sent via Deja.com http://www.deja.com/
Before you buy.


From dirck at pacbell.net  Tue Apr 11 10:54:04 2000
From: dirck at pacbell.net (Dirck Blaskey)
Date: Tue, 11 Apr 2000 07:54:04 -0700
Subject: WinNT and TCL/TK
References: <eviltofu-9BCDF3.08165411042000@news.earthlink.net> <8cv5vo$f0v$1@pegasus.csx.cam.ac.uk>
Message-ID: <8_GI4.1349$uE2.530668@news.pacbell.net>

James "Wez" Weatherall <jnw22 at cam.ac.uk> wrote in message
news:8cv5vo$f0v$1 at pegasus.csx.cam.ac.uk...
> "Jerome Chan" <eviltofu at rocketmail.com> wrote in message
> news:eviltofu-9BCDF3.08165411042000 at news.earthlink.net...
> > Have installed Python 1.5.2 on a WinNT Server but it says it can't find
> > a TCL lib when I start it up. Do I have to add something?
>
> This is a path-setting bug, as far as I can tell.  Under the My Computer
> Properties tab in Windows, change the system's PATH to include the
directory
> containing the Tcl library (usually c:\program files\tcl\bin).
>
> The other option is to copy the tcl80.dll (?) file from that directory
into
> the system32 directory.  That way you needn't alter the system PATH but
you
> will have a rogue file in your system directory!

Extra copies of the tcl/tk dlls can cause the 'init.tcl not found' error
instead of the dll not found error, so copying the dlls into system32 isn't
recommended.  Change the PATH or use a modified FixTk.py:
http://www.danbala.com/python/fixtk

More information is also available here:
http://www.pythonware.com/people/fredrik/fyi/fyi02.htm

Good luck

d
===========================
Dirck Blaskey
Danbala Software
dirck at danbala.com
http://www.danbala.com







From phd at phd.russ.ru  Mon Apr 24 11:46:21 2000
From: phd at phd.russ.ru (Oleg Broytmann)
Date: Mon, 24 Apr 2000 15:46:21 +0000 (GMT)
Subject: questions about imports and classes
In-Reply-To: <3904681D.FE39B41A@hologic.com>
Message-ID: <Pine.LNX.4.21.0004241545100.7343-100000@fep132.fep.ru>

On Mon, 24 Apr 2000, Tim Iskander x7483 wrote:
> 1 - is it possisble to get the instance name of an object (similar to
> the class.__name__
>         attribute)

   It is impossible. Consider
a = A()
b = a
   Now both a and b points to the same instance; what is the name, a or b? :)

> 2 - has there been any/much discussion on extending the import command
> to allow

   You need __import__ builtin function.

Oleg.            (All opinions are mine and not of my employer)
---- 
    Oleg Broytmann      Foundation for Effective Policies      phd at phd.russ.ru
           Programmers don't die, they just GOSUB without RETURN.




From neelk at brick.cswv.com  Wed Apr  5 19:18:58 2000
From: neelk at brick.cswv.com (Neel Krishnaswami)
Date: 5 Apr 2000 23:18:58 GMT
Subject: Why should I switch to Python? - Infinity of Primes
References: <Pine.GSO.4.10.10004050749030.24181-100000@sundial>
Message-ID: <slrn8enkhp.456.neelk@brick.cswv.com>

Moshe Zadka <moshez at math.huji.ac.il> wrote:
>
> [the axiom of choice]
> 
> Thus, the question was taken out of mathematics and into philosophy,
> and hence cannot be answered, so the only answer I can give is my
> own: I refuse to use the axiom of choice, but I do use the
> ultra-filter axiom.  This relates in a deep way to how I view the
> future of mathematics.

I tend towards Chaitin's view: "Axioms are very cool -- make some 
more up! After all, mathematics is basically arbitrary anyway!" I 
guess this shows I am really a frustrated Hilbertist at heart. :)

Seriously, I've developed an interest in constructive mathematics
since running across the whole style of syntax-directed soundness
proofs in type theory. It has (finally) dawned on me that getting
acquainted with a more constructive style of analysis will likely
make me a better programmer. Can you suggest a good textbook?


Neel


From dalke at acm.org  Thu Apr 20 01:48:22 2000
From: dalke at acm.org (Andrew Dalke)
Date: Wed, 19 Apr 2000 23:48:22 -0600
Subject: pattern-based string expansion
References: <38F52C61.A5D72F89@gte.net> <vndog79ft1q.fsf@camelot-new.ccs.neu.edu> <8del87$5md$1@slb6.atl.mindspring.net> <38FE70DB.29A4B1A1@gte.net> <KsvL4.1559$Ez2.89485@typhoon1.gnilink.net>
Message-ID: <8dndbe$dcu$1@slb6.atl.mindspring.net>

Nick Mathewson wrote:
>It's a very very simplified version of (a subset of (a more general
>regex-manipulation module for (a project that I'm working on))).  I
>didn't need an expansion function for my project, but it was pretty
>easy to hack one up.


You'ld probably be interested to know that Python 1.6a has an
sre_parse module which parses regular expressions into a Python
data structure.  But technically the distribution says you can
only use that module for alpha/beta testing.

                    Andrew





From jmg at ecs.soton.ac.uk  Mon Apr 17 13:42:34 2000
From: jmg at ecs.soton.ac.uk (Jacek Generowicz)
Date: Mon, 17 Apr 2000 18:42:34 +0100
Subject: Handling errors in open
Message-ID: <38FB4D0A.489FF50F@ecs.soton.ac.uk>

Hi,

Is there a way to prevent a Python script from
terminating when `open' fails, and have it handle
the error more helpfully . . . I was thinking
along these lines:


from sys import argv

try:
    file = open(argv[1],"r")
except IOerror:
    print "You lied about that file."


Jacek




From akira.kiyomiya at autodesk.com  Thu Apr 27 17:20:06 2000
From: akira.kiyomiya at autodesk.com (Akira Kiyomiya)
Date: Thu, 27 Apr 2000 14:20:06 -0700
Subject: Simple Modules Question
Message-ID: <8eaaul$16p@autodesk.autodesk.com>

I see NameError when I try to do string.atoi("45").

Is it because 'string' is at a different location??

Thanks

Akira


from string import atoi    # imports 'string'
                                     # places atoi in current namespace
print atoi("12345")         # invokes 'atoi' without module name

string.atoi("45")             # NameError: string




From boncelet at udel.edu  Thu Apr  6 22:53:06 2000
From: boncelet at udel.edu (Charles Boncelet)
Date: Fri, 07 Apr 2000 12:53:06 +1000
Subject: NumPy modelling of Belief Tables?
References: <8ciem1$scb$1@kopp.stud.ntnu.no>
Message-ID: <38ED4D92.6A256DB2@udel.edu>

Magnus Lie Hetland wrote:

>
>
> If, for instance, the table t1 has the dimensions X and Y, while t2 has the
> dimensions X and Z, then for each x, y, z in X, Y, X,
>
>         (t1 * t2)[x,y,z] = t1[x,y] * t2[x,z].
>
> This may be simple to do in NumPy (it is certainly easy to do
> algorithmically) - I just can't see how...

Two ideas:
1. loop over the columns (rows) of one of the matrices and let
numpy broadcasting work for you.  Then convert back to a single
numpy object.

2.try multiply.outer(t1,t2) or multiply.outer(t2,t1) followed
by some diagonal magic:
    tt = multiply.outer(t2,t1)
    t = diagonal(tt,axis1=1,axis2=3)
(You may need to rearrange things to get the output in the order
you want.  Bear in mind that tt is a 4d object, so this may be
slow.)

>
>
> Division is about the same thing, except that 0/0 = 0... For tables with the
> same dimensions, I guess one could use something like
>
>         t1 / where(equal(t2,0),1,t2)
>
> where 1 is just an arbitrary value. (I'm not sure how this solution
> generalizes when the dimensions differ...)

This should work.  (Note to others: he is assuming that t2=0 implies t1=0.)

>
>
> The other operation needed is marginalization of the tables. When a table
> is marginalized over some of its dimensions, the elements having equal
> coordinates in the remaining dimensions (i.e. the ones not being
> marginalized
> over) are added. For instance, if you have a two-dimensional table, t, over
> the variables X and Y, and you want to marginalize out Y, then for each
> x in X,
>
>         m[x] = sum with x in Y { t[x,y] },
>
> where m is the marginalized table.
>

"sum" is your friend here. If a=array([0,1],[2,3]), the sum(a)=array([2.4])
and sum(a,1)=array([1,5]).  Just specify the axis over which you want to
sum.

If you produce anything interesting, please publish it if you can.

    Charlie Boncelet

------
Charles Boncelet, University of Delaware,
On sabbatical at ADFA, Canberra Australia,
Home Page: http://www.ece.udel.edu/~boncelet/





From pyython-list at teleo.net  Mon Apr 17 10:38:00 2000
From: pyython-list at teleo.net (Patrick Phalen)
Date: Mon, 17 Apr 2000 07:38:00 -0700
Subject: Filtering web proxy
In-Reply-To: <Pine.LNX.4.21.0004171406250.30999-100000@fep132.fep.ru>
References: <Pine.LNX.4.21.0004171406250.30999-100000@fep132.fep.ru>
Message-ID: <00041707431801.00880@quadra.teleo.net>

[Oleg Broytmann, on Mon, 17 Apr 2000]
::    I want a filtering web proxy. I can write one myself, but if there is a
:: thing already... well, I don't want to reinvent the wheel. If there is such
:: thing (free and opensourse, 'course), I'll extend it for my needs.

Try CTC (Cut the Crap). It's written in Python and is GNU'd.

http://softlab.ntua.gr/~ckotso/CTC/



From bwinton at tor.dhs.org  Mon Apr 10 15:25:40 2000
From: bwinton at tor.dhs.org (Blake Winton)
Date: Mon, 10 Apr 2000 19:25:40 GMT
Subject: What is python's language level?
References: <38EA1B8F.A797A130@lmco.com> <roy-D857DA.21212304042000@netnews.nyu.edu> <38EA9CD4.3473B910@callware.com> <roy-CF57ED.23102504042000@netnews.nyu.edu> <ED6060DB649A6FED.125EA090C5A03FDE.AA052F5C80E50E94@lp.airnews.net>
Message-ID: <slrn8f4ar7.1t0.bwinton@tor.dhs.org>

On 4 Apr 2000 23:04:01 -0500, Cameron Laird wrote:
>Roy Smith  <roy at popmail.med.nyu.edu> wrote:
>>Have you ever looked at the output of a bib | tbl | eqn pipeline?
>Are you kids still using that as a pick-up line?

Well, you've got to admit that it's better than "Shocking bad hat".

Kids these days...
Blake.
-- 
2:54pm up 12 days, 15:28, 1 user, load average: 1.08, 1.02, 1.01


From pecora at anvil.nrl.navy.mil  Fri Apr 28 07:46:58 2000
From: pecora at anvil.nrl.navy.mil (Louis M. Pecora)
Date: Fri, 28 Apr 2000 11:46:58 GMT
Subject: Do I always have to write "self." ?
References: <270420001706386648%pecora@anvil.nrl.navy.mil> <slrn8ghc6t.f01.scarblac-spamtrap@flits104-37.flits.rug.nl>
Message-ID: <280420000746597329%pecora@anvil.nrl.navy.mil>

In article <slrn8ghc6t.f01.scarblac-spamtrap at flits104-37.flits.rug.nl>,
Remco Gerlich <scarblac-spamtrap at pino.selwerd.nl> wrote:

> No. However, this seems to make code *more* readable. In fact, many C++
> programmers use 'this.x' for all their instance variables, to be able to
> see at a glance which are just local variables of the function and which are
> instance variables. It's just more work to write, but that doesn't matter
> much.

I understand the potential for readability, but I do a lot of
mathematical programming so the expressions can get pretty cluttered
with variables and operators.  More self's only makes it worse for me. 
Maybe that's my problem, but then that's my complaint.

> Python needs it to know which vars are instance variables, and which are
> local variables.

I'm sure some way of declaring variables 'local' could be done in the
class defintion, but I doubt Guido would do that now.

> Funny how people say that every aspect of Python they don't like make the
> language less "clean". It is extremely consistent though.

Funny, I guess.  Harder to read for me.

> You *can* use another word for 'self' though, 's' or 'me' or something. But
> it makes your code harder to read, of course...

You mean by doing something like

s=self

at the beginning of each class method?

Thanks.


From hgg9140 at george.org  Sat Apr  1 18:48:44 2000
From: hgg9140 at george.org (Harry George)
Date: 01 Apr 2000 15:48:44 -0800
Subject: Why should I switch to Python?
References: <seas5ia89177@corp.supernews.com>
Message-ID: <m38zyx9ycj.fsf@wilma.george.org>

Of course, read the language comparisons.  But what you seem to be
seeking is personal experiences, so here is another one:

For 3 years I've written perl code.  1 LOC, 10 LOC, 100 LOC, 1000 LOC,
and larger.  It revolutionized our group's productivity and I got
several other people interested.  I made a point of timing projects
with a stopwatch (27 minutes from notion through rqmts, spec, code,
alpha, beta, and implementation is my personal best).  We also did
code reviews with at least 1 experienced perl-ist in the review --
this was valuable in improving community coding skills.

I noticed that people who wrote in it all the time did ok, but if you
go away (e.g., code in lisp, prolog, C, C++, Java), and come back, you
forget the rules for stuff like references.  I was spending a lot of
time helping people maintain code they originally wrote but could no
longer understand.  Personally, I write as maintainably as I can
("-wt", "use strict", subset of available syntax, etc.), but even I
was getting lost on references.

Then I did a multi-module, OO project with several thousand lines.  It
was heavily into references and overrides and hashes inside of arrays
inside of hashes.  I survived it, but was ready for another approach.
So I redid it in python.  It took a couple weeks to catch on to some
of the idioms, but the result was fewer files, fewer lines of code
total, and vastly more maintainable.  Among other things, I could see
the forest for the trees, anc changed some algorithms.

HOWEVER: It ran slower than the equivalent perl code.  Is this
important?  Not for my task, where the difference is roughly 2 seconds
vs 1 second.  Profiling showed a lot of that was in reg expr tasks, so
I'll see what 1.6 does.
 
I found python didn't support perl-ish idioms, some of which are very
succinct and just plain cool.  So I wrote pyperl and cgipm.  With
those in hand, python gets the perl idioms but still has python's
robust syntax.

Next, I started porting java apps (e.g., XML and database stuff) to
python.  Again, dramatic improvements in LOC, development time, and
maintainability.

These days, I still write perl and java where required by work
conditions, but use python where permitted.  Unfortunately, I've
managed to "sell" perl to several managers (who have invested in
sending their people to classes), and they aren't happy that I'm now
pushing something else.


Aaron Turner <aturner_ at _pobox.com> writes:

> A friend of mine is trying to convince me we should 
> do a project using ORBit in Python.  I've been 
> coding Perl for over 3 years and really like it.  
> (Especially it's syntax and ability to grok my 
> programming style.)  Now, I understand the Perl
> vs. Python war and everyone has their preferred
> taste in the matter.  
> 
> What I'd like to hear from people is what technical 
> resons I should switch to Python.  Are there any 
> features that Python has (perhaps in it's OO 
> offering) that Perl lacks?  From what I can tell
> people tend to prefer Python over Perl mostly
> for its syntax, which I find very likable.
> 
> Thanks.

-- 
Harry George
hgg9140 at seanet.com


From effbot at telia.com  Mon Apr  3 03:14:07 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Mon, 03 Apr 2000 07:14:07 GMT
Subject: Unicode program representation
References: <015c01bf9d0b$76f9b2a0$59dd3fcb@neil>
Message-ID: <3xXF4.1060$n68.186533888@newsb.telia.net>

Neil Hodgson <neilh at hare.net.au> wrote:
>    I think this should be changed to interpreting the literal as a UTF-8
> literal. The advantage here is that non-roman string literals become a
> natural part of the language.

however, perfectly valid ISO Latin 1 strings (or any other standard
8-bit encoding) are invalid UTF-8 strings.   changing the default to
UTF-8 this is likely to break many existing scripts.

maybe we can change the default in Python 3000, but to fix this be-
fore that, we'll need to introduce an "encoding pragma" of some kind,
and leave the default as "pass through".

</F>




From robin at jessikat.demon.co.uk  Sat Apr 22 08:19:03 2000
From: robin at jessikat.demon.co.uk (Robin Becker)
Date: Sat, 22 Apr 2000 13:19:03 +0100
Subject: code not true?
References: <8ds3ku$q0l$1@nnrp1.deja.com>
Message-ID: <4b9kXXA3iZA5EwLn@jessikat.demon.co.uk>

In article <8ds3ku$q0l$1 at nnrp1.deja.com>, jeff_islay at my-deja.com writes
>Why does the second example return false?
>
>>>> print 7.0 == 7
>1
>>>> print (.07 * 100) == 7
>0
>
>
>Sent via Deja.com http://www.deja.com/
>Before you buy.

.07 isn't exactly representable in your (probably IEEE) binary floating
point representation.

print .07*100 -7
8.8817841970012523e-016 
-- 
Robin Becker


From javanet at dynacap.com  Thu Apr 27 18:00:27 2000
From: javanet at dynacap.com (Arnaldo Riquelme)
Date: Thu, 27 Apr 2000 14:00:27 -0800
Subject: Newbie Question
Message-ID: <sghdv57tqtj79@corp.supernews.com>

Suppose I have a file name tick.tkr that looks like this

SUNW.
MFST.
FOO.
SPAM.
MOO.
End_Of_File

I want to read the file into a list

myfile = open("tick.tkr", "r")
mylist = myfile.readlines()

Now I have mylist with the contents of tick.tkr. "mylist" looks like this
['SUNW.\012', 'MSFT.\012', ..........'End_of_File/012']

How do I read the file into a list without getting the newline character, so
I can accomplish this

for tic in mylist:
    #create file tic.prn

I know the other P language has something like 'chop' or something like
that, but I'm not going there.

Thanks


Arnaldo,








From embed at geocities.com  Fri Apr  7 11:57:37 2000
From: embed at geocities.com (Warren Postma)
Date: Fri, 7 Apr 2000 11:57:37 -0400
Subject: reduce versus recursion? what am I missing...
Message-ID: <YznH4.4508$HG1.128287@nnrp1.uunet.ca>

# I am trying to figure out why function d (recursive) and e (reduce)
# DON'T do the SAME thing:

def d(*tup):
    "why doesn't this work?"
    if len(tup)>1:
        return d(tup[1:])+abs(tup[0]-tup[1])
    else:
        return 0

l = (1,3,5,7,9)

def e(*tup):
    "it works"
    return reduce(lambda x,y:abs(x-y),tup,0)

print "--- test:"
print "recursion: ",apply(d,l)
print "reduce:    ",apply(e,l)

#---
#Warren
#
# output:
#--- test:
#recursion:  2     ( appears not to actually be recursing more than 1
level!!!???)
#reduce:     5





From effbot at telia.com  Sat Apr  1 05:26:58 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Sat, 01 Apr 2000 10:26:58 GMT
Subject: Python 1.6 alpha 1 released
References: <200003312130.QAA04361@eric.cnri.reston.va.us> <38E5205F.DE811F61@roguewave.com> <Wf9F4.4423$74.70231@newsc.telia.net> <38E58D2D.384F1546@Lugoj.Com>
Message-ID: <S9kF4.4472$74.71414@newsc.telia.net>

James Logajan <JamesL at Lugoj.Com> wrote:
> Also, how good are you at math? You seem to feel that 5 minutes of "Guido
> the god's" time is of much more value than that of 1000 knuckle-dragging
> Neanderthal programmers like myself spending a mere 10 minutes each to
make
> sure their code doesn't break. Of course, if we happen to take more than
10
> minutes its our own damn fault I'm sure you'll agree.

the strange thing here is that there are changes listed on
that page that are far more difficult to spot and fix, but for
some reason, neither you or bjorn seem to care.

instead, bjorn announced that he'd never upgrade to 1.6
based on a single change in 1.6a1, despite the fact that
there will be a 1.6a2, a3, b1, b2, rc1, etc. before it's time
to really make that decision.

others have politely asked for backwards compatibility on
this one (which would mean *adding* extra code to handle
multiple calling syntaxes).  someone will probably come up
with a patch well before the next alpha, but probably not
because you guys are flaming away on this newsgroup.

besides, the notion that 10 minutes spent on fixing this will
outweigh all others costs involved in upgrading to a new
release of the core interpreter strikes me as rather odd...

(heck, both of you have probably spent far more than ten
minutes each writing your posts...)

> Now how did I get so pissed over this exchange of posts?

now how come I find your "guido the god" argument to be
directly offensive?

</F>




From BLAISE.BONNEVILLE at wanadoo.fr  Fri Apr 21 02:49:01 2000
From: BLAISE.BONNEVILLE at wanadoo.fr (BLAISE BONNEVILLE)
Date: Fri, 21 Apr 2000 08:49:01 +0200
Subject: print a pixel
References: <8dmcvb$r6n$1@wanadoo.fr> <20000420080534.A623423@vislab.epa.gov>
Message-ID: <000201bfab5e$c8e3a1a0$a688fcc1@wanadoo.fr>

By "printing a pixel", I mean to open a new window on the screen with
nothing in it. Then I want to get a pixel in this window when I enter its
coordinates (x ; y) and I would like to be able to chose its color as well.
I hope you can now understand my question.
I send the program in which I want to use this "print a pixel" function.
Maybe it will help you understand exactly what I want to do. It should draw
Mandelbrot's fractal : to do so, it tests every pixel in a blank window to
dertermine if this particular pixel checks a property (what I called the
"test" function"). If so, this pixel must appear in the window.
Thank you.


-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: mandel.py
URL: <http://mail.python.org/pipermail/python-list/attachments/20000421/1caa6580/attachment.ksh>

From kc5tja at garnet.armored.net  Sun Apr 30 19:55:27 2000
From: kc5tja at garnet.armored.net (Samuel A. Falvo II)
Date: 30 Apr 2000 23:55:27 GMT
Subject: Do I always have to write "self." ?
References: <270420001706386648%pecora@anvil.nrl.navy.mil> <slrn8ghc6t.f01.scarblac-spamtrap@flits104-37.flits.rug.nl> <8ecsgj$6lp$1@slb0.atl.mindspring.net> <m3bt2tonyg.fsf@atrus.jesus.cam.ac.uk> <300420001008012011%pecora@anvil.nrl.navy.mil> <slrn8gp6c4.ilf.kc5tja@garnet.armored.net> <bz1P4.75789$uF1.160076@afrodite.telenet-ops.be>
Message-ID: <slrn8gqdlg.k65.kc5tja@garnet.armored.net>

>OK, let's bite :-)
>
>    int WriteData(void *buffer, char *str, int len)
>
>I fail to see how this is somehow hard to read :-)

It's not, in this context.  But it was a decidedly simple example.  How
about this:

	int Evaluate( char *str1, char *param1, char *param2 );

Note the `param1' and `param2' argument names.  Hate to tell you this, but
that's Hungarian notation -- you're identifying those arguments as
"parameters".  But what the heck would str1 represent?  The command that
acts on those parameters?  You don't know -- you CAN'T know from this
definition.  It's cases like this that I choose to use hungarian notation:

	uint32 Evaluate( char *cmd, char *param1, char *param2 );

-- 
KC5TJA/6, DM13, QRP-L #1447
Samuel A. Falvo II
Oceanside, CA


From gmcm at hypernet.com  Tue Apr 11 15:45:49 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Tue, 11 Apr 2000 15:45:49 -0400
Subject: Functions and objects
In-Reply-To: <meh9-C7461E.15255011042000@news.cit.cornell.edu>
Message-ID: <1256640143-42229404@hypernet.com>

Matthew Hirsch wrote:

> >>> a
> [5]
> >>> for x in range(5):     # Case A
> ...     temp=a
> ...     print a
> ...     temp.append(1)
> ... 
> [5]
> [5, 1]
> [5, 1, 1]
> [5, 1, 1, 1]
> [5, 1, 1, 1, 1]
> 
> >>> a=f(5)
> >>> for x in range(5):     # Case B
> ...     temp=a[:]
> ...     print a
> ...     temp.append(1)
> ... 
> [5]
> [5]
> [5]
> [5]
> [5]
> >>> 
> 
> In case A, why doesn't temp reset itself to the value of a, [5], that 
> was predetermined before it entered the loop?  Why do you need to copy 
> the list to get the behavior I'm looking for in Case B?  Doesn't a 
> always equal [5]?

In case A, temp is another name for a. In B, temp is a copy of 
a. The existence of "f" has nothing to do with it.

- Gordon



From m.faassen at vet.uu.nl  Wed Apr 12 10:24:29 2000
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 12 Apr 2000 14:24:29 GMT
Subject: Python/Perl Popularity (Re: A Mountain of Perl...)
References: <200004102111.HAA03380@envy.fulcrum.com.au> <8d0pcq$7ti$1@nnrp1.deja.com> <g0ssm03o.fsf@dan.resnet.rochester.edu> <8d1lgu$ga0$1@newshost.accu.uu.nl> <zoqzlazr.fsf@dan.resnet.rochester.edu>
Message-ID: <8d20ut$gbe$1@newshost.accu.uu.nl>

Daniel Berlin+list.python <dan at cgsoftware.com> wrote:

[I mention that people do care about distutils]

> I didn't say nobody cares, i was basically saying i don't see all that
> many complaints about this on the python lists.

That's probably because people are patiently waiting for the distutils.
The issue has been raised quite a few times.

> Maybe i'm just missing them. I realize their is a whole SIG about the issue,
> but it's not a problem for me.

That's good for you. You came across a bit hostile, though, with your
'who gives a crap?' response.

[snip]
> Sure, i'm all for it too, I just don't see why it should be all that
> high of a priority.

Apparently Tom thinks it's nice; I didn't think he was telling people
to fix it now. He just mentioned some features of Perl that he
considers nice and absent in Python, right? 

> I also don't see why Tom praises the CPAN module so much.
[problem description with CPAN snipped]

I haven't used Perl and CPAN, but I've heard from several people on
this list that CPAN is pretty nice. Since I haven't used it myself,
I don't know about your problem.

[Daniel]
>>> Why do i get the feeling you'll never be satisfied?

[me]
>> Why attacking Tom so much? 

[Daniel]
> I've not attacked him at all.
> I simply asked a question.

Okay, then let me inform you that your style of posting and asking
questions does come across a trifle hostile once in a while to me. Perhaps
others disagree with me, but that's how it feels to me.

>> What has Tom done to you in
>> the past? Why does he get the bad reception he does?

> Because he comes in here with both barrels firing.
> What reception did he expect?

Um, I didn't see him do that at all. And even if someone comes in with
both barrels firing, you're not supposed to fire back. You're supposed to
discuss mathematical proofs of an infinity of primes. :) [not that the
thread I'm referring to had someone coming in with both barrels firing
either!]

Regards,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From jhauser at ifm.uni-kiel.de  Wed Apr 12 04:52:29 2000
From: jhauser at ifm.uni-kiel.de (Janko Hauser)
Date: 12 Apr 2000 10:52:29 +0200
Subject: Classes, Inheritance - Stupid lazy question
References: <8d1cfh$hme7@imsp212.netvigator.com>
Message-ID: <874s9766r6.fsf@ifm.uni-kiel.de>

I think this will do it

class lazy:
    def __init__(self, name):
        self.name = name

    def fired(self):
        print 'You are fired %s' % self.name

class paul(lazy):
    def __init__(self, name):
        self.name = name  # <---- ?????
        self.april_pay = 0

    def fired(self):
        print "Here's a box, collect the things from your desk"
        self.__class__.__bases__[0].fired(self) 

te = paul('myname')
te.fired()

HTH
__Janko

-- 
  Institut fuer Meereskunde             phone: 49-431-597 3989
  Dept. Theoretical Oceanography        fax  : 49-431-565876
  Duesternbrooker Weg 20                email: jhauser at ifm.uni-kiel.de
  24105 Kiel, Germany


From mhammond at skippinet.com.au  Mon Apr 10 20:39:07 2000
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Tue, 11 Apr 2000 00:39:07 GMT
Subject: Python 1.6 -- Any hope for new threading core?
References: <EDFD2A95EE7DD31187350090279C6767706F14@THRESHER>
Message-ID: <LuuI4.960$UB.6779@news-server.bigpond.net.au>

"Brent Fulgham" <brent.fulgham at xpsystems.com> wrote in message
news:EDFD2A95EE7DD31187350090279C6767706F14 at THRESHER...

> Just curious if any of this behavior is changed
> with 1.6?

Nope.

Mark.





From embed at geocities.com  Wed Apr 19 10:53:57 2000
From: embed at geocities.com (Warren Postma)
Date: Wed, 19 Apr 2000 10:53:57 -0400
Subject: Very Horrible Question About Goto's
References: <6325E90C8B6ED21189080001FA7E2ADA0D4051@IBA1>
Message-ID: <dPjL4.9434$HG1.254770@nnrp1.uunet.ca>

>I'm trying to create a converter from our own internal language (horrid)
into python.
> My only problem so far is that the original language uses *LOTS* of
goto's.
>Is there any way that this can be replicated within python ?

Sounds like you're trying to avoid dealing with the Accidental Complexity in
your system. You'd probably find that it's going to kill you if you don't
deal with it.  I'm saying this only because I've done that myself. <grin>

However, I find programmers like myself are always loath to wander into
irreducibly complex problems without first trying to take a stab into the
heart of whatever complexity they've run up against. Given that, you might
get Creative and try to write a code-analyzer and translator in Python to
parse all your Horrid 1.0 Sources and output the sorts of "meta-data" that
might help you in constructing your new system, or even help you automate
your conversion process.

People on this newsgroup are always up for a challenge, so why not post a
sample snippet of Horrid 1.0 and say what you'd like to do with it to
Pythonize it.

It's also possible for you to consider a "state based" or "event model"
based system for your existing code. You could create a dictionary of state
objects, and a "directed graph" model of showing the connections from state
to state. If you want to be really slick, make a visual model of your
state-charts. :-)

We are working on a state-based model of programming for Python, probably
going to use C.T's stackless-python eventually, but for now we are just
playing with some trivial python scripts to get a prototype of our event
model working. I am also considering the "graph" capabilities of the little
kjBuckets extension (see the GadFly web page for this little C extension)
which provides some extremely Nifty and Fast set/graph/dictionary extensions
that are helpful in this sort of case.

Warren






From alwagner at tcac.net  Mon Apr 17 09:03:50 2000
From: alwagner at tcac.net (Albert Wagner)
Date: Mon, 17 Apr 2000 08:03:50 -0500
Subject: Printer & redirection follow-up
Message-ID: <38FB0BB6.F8174FF8@tcac.net>

To all who replied to my posts, thanks again. Below is the result of my
hacking and your advice.  For you Windows guys:  This works under Linux;
will it work under Windows?

#
import sys
from DisplayFileList import *
try:
    queue = 'lpt'
    from win32pipe import popen
except ImportError:
    queue = 'lpr'
    from os import popen
        

class Paginator:

    def __init__(self):
        self.lineCount = 0
        self.linesMax = 50
        self.pageCount = 1
        
    def printList(self, rootDir, rootName):
        self.rootName = rootName
        self.displayer = DisplayFileList()
        sys.stdout = self       # Output of displayer redirected here
        self.printer = popen(queue, 'w')
        self.displayer.displayTree(rootDir) #prints to sys.stdout
        self.printer.close()
        sys.stdout = sys.__stdout__
        
    def write(self, aLine):     #Used by sys.stdout
        if self.lineCount == 0: #If first page then...
            self.pageHeader()   #print 1st page header 
        self.printer.write(aLine)
        if len(aLine) > 1:      #only count real lines, not \n only
lines
            self.lineCount = self.lineCount + 1
        if self.lineCount > self.linesMax:  #If end of page then...
            self.printer.write("\f")        #eject and print page header
            self.pageHeader()

    def pageHeader(self):
	<snip>

-- 
Small is Beautiful


From gmcm at hypernet.com  Tue Apr  4 15:55:16 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Tue, 4 Apr 2000 15:55:16 -0400
Subject: Why should I switch to Python? - Infinity of Primes
In-Reply-To: <38EA369C.26A1C25@math.okstate.edu>
Message-ID: <1257244378-5884705@hypernet.com>

David C. Ullrich wrote:

> ... Could be that there's a proof of the existence
> of infinitely many primes using FTA that has some property
> not shared by Euclid's proof, but "constructive" isn't it.

Oh, come on. There are at least 3 well known constructive 
proofs that all odd numbers are prime:

The mathematician's:
 3 is prime,
 5 is prime,
 7 is prime
 as so on, by induction.

The physicist's:
 3 is prime,
 5 is prime,
 7 is prime,
 9 is prime <experimental error>
 11 is prime...

The computer scientist's:
 3 is prime,
 3 is prime,
 3 is prime ....


- Gordon



From python-list at teleo.net  Sun Apr  2 19:43:36 2000
From: python-list at teleo.net (Patrick Phalen)
Date: Sun, 2 Apr 2000 16:43:36 -0700
Subject: Just doing local file testing
In-Reply-To: <38E7D6DE.EA09C891@regnoc.com>
References: <38E796BC.23191B2@regnoc.com> <00040214141502.04535@quadra.teleo.net> <38E7D6DE.EA09C891@regnoc.com>
Message-ID: <00040216504403.04535@quadra.teleo.net>

[Jim Conger, on Sun, 02 Apr 2000]
:: <!doctype html public "-//w3c//dtd html 4.0 transitional//en">
:: <html>

Yeeks. You forgot to turn off HTML formatting for your message.

:: I'm not even to the point of using Apache yet - I'm just to launch the
:: script locally from within the browser by having the form element point
:: to a script on disk.&nbsp; In other words, this is just a development environment
:: using only local files.
:: <p>Any other clues?

My original advice still holds, but now I'll add that you probably need
to do some additional study on client/server computing.

IOW, your browser is the client. Where's the server?






From vetler at ifi.uio.no  Thu Apr 13 06:27:38 2000
From: vetler at ifi.uio.no (Vetle Roeim)
Date: 13 Apr 2000 12:27:38 +0200
Subject: Python and VTK and TK
References: <38F51AB1.29FB4C1D@earthlink.net>
Message-ID: <yayln2is3c5.fsf@jotunheim.ifi.uio.no>

* Scott Harris
> Has anyone gotten VTK to work with the TkRenderWidget?
> I've got a simple code that draws two buttons and a vtkTkRenderWidget in
> a frame.
> It works under win98, but crashes my XServer on Linux.
>
> I'm using RH6.1 with XFree864.0 and a Voodoo3 card (so I have
> accelerated GL rendering)
> I've got VTK 3.1

Yes, I've gotten VTK to work with vtkTkRenderWidget without any
problems.

Do other OpenGL applications work properly?

 
-- 
Vetle Roeim 
Graduate Student 
Departement of Informatics, University of Oslo, Norway


From garabik at center.fmph.uniba.spam.sk  Tue Apr 25 12:06:02 2000
From: garabik at center.fmph.uniba.spam.sk (Radovan Garabik)
Date: 25 Apr 2000 16:06:02 GMT
Subject: invisible cursor in ncurses5
References: <87wvlnjlxd.fsf@lemon.tky.hut.fi>
Message-ID: <956678736.128659@cdwork.cvt.stuba.sk>

Ulf Betlehem <flu at iki.fi> wrote:
>         Using curses modules included with most python distributions,
>         there seems to be no way to turn off the cursor if the module
>         is linked against ncurses5. (earlier versions are ok)

try http://pyncurses.sourceforge.net/
it is a (written from scratch) curses binding for python,
and leaveok() and curs_set() work as expected.

-- 
 ----------------------------------------------------------
| Radovan Garabik       http://www.fmph.uniba.sk/~garabik/ |
| __..--^^^--..__       garabik @ center.fmph.uniba.sk     |
 ----------------------------------------------------------


From d98aron at dtek.chalmers.se  Wed Apr 12 14:17:32 2000
From: d98aron at dtek.chalmers.se (Fredrik Aronsson)
Date: 12 Apr 2000 18:17:32 GMT
Subject: Python paradigms
References: <000401bfa1dd$930fe500$172d153f@tim>
Message-ID: <8d2ejs$jfr$1@nyheter.chalmers.se>

In article <000401bfa1dd$930fe500$172d153f at tim>,
	"Tim Peters" <tim_one at email.msn.com> writes:
>    (a() and [b()] or [c()])[0]
> 

Just a litte thougth, can't I just use (a() and b()) or c() ?

/Fredrik


From mhammond at skippinet.com.au  Tue Apr  4 20:04:27 2000
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Wed, 05 Apr 2000 00:04:27 GMT
Subject: win32all-130 for Python 1.6 available
References: <_ziF4.1488$1b4.6926@news-server.bigpond.net.au> <d5rjeskov6c6drsok4n7vvcvvi74155lna@4ax.com>
Message-ID: <fqvG4.5062$1b4.24253@news-server.bigpond.net.au>

"Wolfgang Strobl" <Wolfgang.Strobl at gmd.de> wrote in message
news:d5rjeskov6c6drsok4n7vvcvvi74155lna at 4ax.com...
> "Mark Hammond" <mhammond at skippinet.com.au> wrote:
>
> While using the debugger, I've managed to move the floating
breakpoint
> list window to a position where the title bar is outside
(above) the
> desktop. Now these detachable windows can be moved by dragging
them by

OK - I will fix this (by ensuring that when a toolbar is
re-displayed, it is completely on screen.

> their title bar, only, so I'm stuck. I've found the registry
entry in
> question (or so I thougth), and fiddled with Software\Python
1.6\Python
> for Win32\ToolbarDebugging-Bar7, but in vain.

Delete _all_ "Toolbar*" subkeys - probably 20 or so in total!

Mark.





From akuchlin at mems-exchange.org  Tue Apr 11 11:05:45 2000
From: akuchlin at mems-exchange.org (Andrew M. Kuchling)
Date: 11 Apr 2000 11:05:45 -0400
Subject: The gzip module... neverending story
References: <8cr7os$du3$1@lacerta.tiscalinet.it>
Message-ID: <3daej0he46.fsf@amarok.cnri.reston.va.us>

"FrancescO" <francesco.contini at tiscalinet.it> writes:
> I have, another questions!!
> How can I zip i list of files? (i have done it, but... the files can't be
> readed from winzip and gunzip...)

First, if you're trying to generate PKZIP-style .zip archives, the
gzip module is the wrong one.  Jim Ahlstrom's zipfile.py, which
someone else posted a pointer to, is the right module to use in that
case.  Since you mention gunzip, I assume that you really do want
gzip-style files.

> Wich is the correct usage of the fileobj (4th parameters in the
> gzip.GzipFile.. explain..:
> GzipFile ([filename[, mode[, compresslevel[, fileobj]]]]) )?

Did you look at the documentation for the module, at
http://www.python.org/doc/lib/module-gzip.html ?  It explains:

	The new class instance is based on fileobj, which can be a
	regular file, a StringIO object, or any other object which
	simulates a file. It defaults to None, in which case filename
	is opened to provide a file object.

It's not clear what you're trying to do: compress a list of files into
individual compressed files, or compress a bunch of files into one big
huge file.  To do the former, the code would be something like (untested):

import gzip
filenames = ['a.txt', 'b.txt', 'c.exe']
for f in filenames:
    input = open(f, 'rb')
    output = gzip.open(f + '.gz', 'wb')

    # Copy the contents of the input file
    while 1:
        chunk = input.read(1024)
        if chunk == "": break
        output.write( chunk )

    input.close() ; output.close()

gzip doesn't support storing multiple files, though you can
concatenate several gzipped files; the decompressed data would then be
the contents of the original files in one big file.

-- 
A.M. Kuchling			http://starship.python.net/crew/amk/
Desire was right. Also untrustworthy, acerbic, dangerous, and cruel. But
right. You would have been better off leaving well enough alone.
  -- Destruction, in SANDMAN #48: "Brief Lives:8"

 


From moshez at math.huji.ac.il  Fri Apr 21 23:32:22 2000
From: moshez at math.huji.ac.il (Moshe Zadka)
Date: Sat, 22 Apr 2000 06:32:22 +0300 (IDT)
Subject: finding real $0
In-Reply-To: <OWeCyjAOWCA5Ew86@jessikat.demon.co.uk>
Message-ID: <Pine.GSO.4.10.10004220631490.5248-100000@sundial>

On Fri, 21 Apr 2000, Robin Becker wrote:

> Is there any pythonic way to locate the original argument zero of the
> process? I mean the location of python itself not the script. This would
> allow me to launch subscripts without having to rely on the path being
> set up correctly.

No.

I could give you a dozen algorithms that half work, but that'd be wasting
your time and mine.
--
Moshe Zadka <mzadka at geocities.com>. 
http://www.oreilly.com/news/prescod_0300.html
http://www.linux.org.il -- we put the penguin in .com




From jp at NOSPAMdemonseed.net  Thu Apr 27 18:28:24 2000
From: jp at NOSPAMdemonseed.net (Jason Petrone)
Date: Thu, 27 Apr 2000 17:28:24 -0500
Subject: PyOpenGL
References: <8e9fed$lm4$1@news.rz.uni-karlsruhe.de>
Message-ID: <8ueae8.0vj.ln@demonseed.net>

Bjoern Giesler <un4e at rzstud1.rz.uni-karlsruhe.de> wrote:
> Hi,

> is there an archive that carries a newer version of PyOpenGL? The newest I
> have is from the contrib-09-Dec-99 directory on ftp.python.org, and that
> seems to require tcl7.6, which I don't want to use anymore.

http://starship.python.net/crew/da/pythondists/

or cvs 

http://www.google.com/search?q=cache:starship.python.net:9673/crew/da/Code
/PyOpenGL/cvs.html

(sorry, the ugly url is because that page no longer exists on starship)

-jlp


From echuck at mindspring.com  Thu Apr 20 22:50:55 2000
From: echuck at mindspring.com (Chuck Esterbrook)
Date: Thu, 20 Apr 2000 22:50:55 -0400
Subject: [ANNOUNCE] Webware for Python 0.1
References: <5650A1190E4FD111BC7E0000F8034D26A0F40D@huina.oceanic.com> <38FEF62B.2A4402CD@mindspring.com> <T6DL4.25197$cZ.34941@typhoon.southeast.rr.com> <2cd13488.77055138@usw-ex0108-063.remarq.com> <FLLL4.15441$O7.22779@typhoon.southeast.rr.com>
Message-ID: <38FFC20F.A372174F@mindspring.com>

Don,

I sent you an e-mail today with the same solution. Was wondering if you got it?

-Chuck


Don Tuttle wrote:
> 
> Thanks Hamish!  You were righton the money. (and the code ain't bad either
> ;-)
> 
> Don
> 
> > The file is actually a gzipped tar file, but your browser is
> > probably curtailing the extension to .tar, thus miseading WinZip.
> > You can get round this by manually giving the file a .tar.gz
> > extension when you save it from your browser. The saved file
> > should now open OK in WinZip.
> >
> > Alternatively you could do it The Python Way (TM) :-)
> >
> > >>> import urllib
> > >>> urllib.urlretrieve(
> > 'http://writewithme.com/Webware/download.cgi', 'webware.tar.gz')
> >
> >
> > Hamish Lawson
> >
> >
> >
> >
> > * Sent from RemarQ http://www.remarq.com The Internet's Discussion Network
> *
> > The fastest and easiest way to search and participate in Usenet - Free!
> >


From ulf.engstrom at b2b-link.com  Thu Apr 13 07:32:37 2000
From: ulf.engstrom at b2b-link.com (=?iso-8859-1?Q?Ulf_Engstr=F8m?=)
Date: Thu, 13 Apr 2000 13:32:37 +0200
Subject: Authenticate with smtp
Message-ID: <018701bfa53b$f831f3e0$858647c1@DEVELOPMENT>

How can I log into an smtp server that requires authentication and send a
mail with smtplib, or is there any other package that helps? (Yes, I'm
authorized)
Regards
Ulf
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Ulf Engstr?m.vcf
Type: text/x-vcard
Size: 831 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20000413/c15b7d16/attachment.vcf>

From darcy at vex.net  Fri Apr 21 06:44:36 2000
From: darcy at vex.net (D'Arcy J.M. Cain)
Date: 21 Apr 2000 10:44:36 GMT
Subject: Using python on the web
References: <EDFD2A95EE7DD31187350090279C6767754BA8@THRESHER> <8dopfu$nrl$1@news1.xs4all.nl>
Message-ID: <8dpbek$nc3$1@news.tht.net>

Boudewijn Rempt <boud at rempt.xs4all.nl> wrote:
> Brent Fulgham <brent.fulgham at xpsystems.com> wrote:

>> Of course, the Postgres Python bindings may be just as bad from your
>> perspective.  :-)

> Well, the Postgres Python bindings don't even conform to the
> DB-API II specs - that kind of rules them out for a lot of

The next release will have DB-API support.  I (the PyGreSQL maintainer)
don't know much about the spec so I invite any and all criticism as long
as it comes with code patches on our mailing list so that others on the
team can review them before they become part of the package.  I expected
to have a release by this weekend but there appears to be a few small
problems to clear up yet.

> work. I myself am quite satisfied with the MySQL bindings, but
> like the original poster, I had to write my own object layer
> around the primitive functions. Maybe it would be a good idea
> to have another standard layer on top of DB-API II, for those
> who want a more sophisticated interface?

Actually, I kind of like the original PyGreSQL interface, especially with
the dictresult() method.  I am thinking of adding a few more convenience
methods like that.  For example, a method that takes a select and returns
a dictionary of dictionaries rather than a list of dictionaries.

http://www.druid.net/pygresql/ for details.

-- 
D'Arcy J.M. Cain <darcy at vex.net>           |  Democracy is three wolves
http://www.vex.net/                        |  and a sheep voting on         
+1 416 425 1212     (DoD#0082)    (eNTP)   |  what's for dinner.


From R.Brodie at rl.ac.uk  Fri Apr 14 10:43:28 2000
From: R.Brodie at rl.ac.uk (Richard Brodie)
Date: Fri, 14 Apr 2000 15:43:28 +0100
Subject: Future compatibility of the Python C API
References: <lwEJ4.7980$Ib7.101092@typhoon2.kc.rr.com> <q4FJ4.1198$1D.10037@news-server.bigpond.net.au> <8d76t2$pig@newton.cc.rl.ac.uk> <slrn8fed2o.4fe.scarblac-spamtrap@flits104-37.flits.rug.nl>
Message-ID: <8d7aqh$19qo@newton.cc.rl.ac.uk>

"Remco Gerlich" <scarblac-spamtrap at pino.selwerd.nl> wrote in message
news:slrn8fed2o.4fe.scarblac-spamtrap at flits104-37.flits.rug.nl...

> > I suppose that someone out there is working on persuading SWIG to
> > generate suitable wrapper code?
>
> Since noone knows what Python 3000 is going to look like yet, that would be
> very hard, wouldn't it? :)

Well possibly. However, I assumed that someone has an idea how it should
look. I presumed that one would not start by assuming that there wouldn't be
backwards compatibility, until far enough down the design to see that it
wouldn't be practical. Unless you work for MS, lack of compatibility isn't
usually a design goal.





From mwh21 at cam.ac.uk  Sun Apr 16 18:24:42 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 16 Apr 2000 23:24:42 +0100
Subject: trapping dict alterations in a class
References: <8ddbpp$7dp$0@216.39.162.232>
Message-ID: <m3d7npelat.fsf@atrus.jesus.cam.ac.uk>

Michael Esveldt <dante at oz.net> writes:

> I want to be able to do for a dictionary in a given class what 
> __setattr__ does for variable binding.
> 
> class foo:
>     def __setattr__(self, name, value):
>         print name, value
>         self.__dict__[name] = value
> bar = foo()
> bar.quux = 1
> bar.baz = {"key":"value"}
> 
> The above obviously prints "quux 1" and then "baz {'key':'value'}". But 
> what if I want to trap further alterations to the foo.baz dictionary? Is 
> this possible?

Well, you could implement __getattr__ and then when asked for "baz"
return a wrapper to the dictionary.

Eg:

import UserDict

class Wrapper(UserDict.UserDict):
    def __setitem__(self,item,value):
        print item, value
        UserDict.UserDict.__setitem__(self,item,value)

class Foo:
    def __init__(self):
        self.__dict__['magic'] = {}
    def __setattr__(self,attr,value):
        print attr, value
        self.magic[attr] = value
    def __getattr__(self,attr):
        return Wrapper(self.magic[attr])

This is far from being complete (writing __{get,set}attr__ methods is
a tricky and subtle business), but it's a start.  Watch out for
cycles, too.

HTH,
Michael

-- 
81. In computing, turning the obvious into the useful is a living
    definition of the word "frustration".
     -- Alan Perlis, http://www.cs.yale.edu/~perlis-alan/quotes.html


From mwh21 at cam.ac.uk  Fri Apr 28 18:33:27 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 28 Apr 2000 23:33:27 +0100
Subject: Do I always have to write "self." ?
References: <270420001706386648%pecora@anvil.nrl.navy.mil> <slrn8ghc6t.f01.scarblac-spamtrap@flits104-37.flits.rug.nl> <8ecsgj$6lp$1@slb0.atl.mindspring.net>
Message-ID: <m3bt2tonyg.fsf@atrus.jesus.cam.ac.uk>

dkuhlman at netcom.com (G. David Kuhlman) writes:
> > No. However, this seems to make code *more* readable. In fact, many C++
> > programmers use 'this.x' for all their instance variables, to be able to
> > see at a glance which are just local variables of the function and which 
> > are instance variables. It's just more work to write, but that doesn't 
> > matter much.
> 
> Thinking about readability, I wrote the C++ coding standard here
> where I work.  When we programmers reviewed it, a co-worker
> insisted that part of our standard be to prefix the names of all
> member variables in any class with "m_" so that we could
> distinguish them from other variables.  I believe that "m_" is a
> Microsoft-ism and in my humble opinion is just another way to spell
> "self.".  I think this shows that even C++ programmers are capable
> of seeing the wisdom of the Pythonic way.

I have unpleasant memories of trying to understand C++ code that named
a *local* variable m_something.  Bleargh.  Unambiguous is good.

Cheers,
M.

-- 
81. In computing, turning the obvious into the useful is a living
    definition of the word "frustration".
     -- Alan Perlis, http://www.cs.yale.edu/~perlis-alan/quotes.html


From wicarrol at unity.ncsu.edu  Wed Apr 12 11:46:48 2000
From: wicarrol at unity.ncsu.edu (William Isaac Carroll)
Date: 12 Apr 2000 15:46:48 GMT
Subject: Python and Curses under MS Windows
Message-ID: <8d25p8$7u9$1@uni00nw.unity.ncsu.edu>

I'm interested in finding a curses package that works under Windows. I've
searched the FAQ and archives with no results. Does such a thing exist? And
does it come in a pre-compiled form? (I have no decent Windows compiler
available.)

Thank you.
-- 
W Isaac Carroll (icarroll at pobox.com)

"People do not eat at once for all time, even when they eat a good deal."
        -- Planchet


From JGRAVES3 at austin.rr.com  Fri Apr  7 08:30:35 2000
From: JGRAVES3 at austin.rr.com (Jay Graves)
Date: Fri, 07 Apr 2000 12:30:35 GMT
Subject: Any news of ActiveState's Python developments?
References: <13caa058.14a1c69c@usw-ex0108-063.remarq.com>
Message-ID: <LxkH4.182$0p1.44921@typhoon.austin.rr.com>

Hamish Lawson wrote in message
<13caa058.14a1c69c at usw-ex0108-063.remarq.com>...
>Is there any news of Python developments at ActiveState since
>they adopted the language? The only thing I've seen is their
>announcement of plans to port Python (and Perl) to the Intel
>Itanium.


I traded some email with David Ascher about this a couple of weeks ago.  He
indicated that the will be announcing some stuff in the 1.6 timeline.  This
was _before_ the 1.6 beta test started, so I don't know if they knew it was
coming up so quickly.  The only really concrete thing that he said was that
ActiveState's Python offering will be less Windows-centric than their Perl
offerings.

HTH
Jay Graves




From marc at oscar.noc.cv.net  Tue Apr 25 10:32:02 2000
From: marc at oscar.noc.cv.net (Marc Spitzer)
Date: Tue, 25 Apr 2000 14:32:02 GMT
Subject: The Simple Economics of Open Source
References: <B527E7A5.51AA%raffael@mediaone.net> <MPG.136cc01af4ab703b989709@news.earthlink.net> <raffael-4899DC.23562823042000@news.ne.mediaone.net> <vndya62ykbi.fsf@camelot-new.ccs.neu.edu>
Message-ID: <slrn8gbaj2.43p.marc@oscar.noc.cv.net>

In article <vndya62ykbi.fsf at camelot-new.ccs.neu.edu>, Justin Sheehy wrote:
>Raffael Cavallaro <raffael at mediaone.net> writes:
>
>> Fact is, only commodity items, that are known to many programmers, are 
>> open sourced. If there were sufficient demand to support a closed source 
>> business, then it would have happened that way. The proof of this is the 
>> BSDs, which would allow a closed source, proprietary fork at any time. 
>> Hasn't happened 'cause there's no demand for such a product.
>
>You mean like BSD/OS, from BSDI?  
>
>It happened 9 years ago, and they've been selling it ever since.
>
>Interesting "proof" you have there.
>
>-Justin
>
this sounds like a religious argument, be carful when you 
show evadence that contradicts the 1 true faith. 

marc

> 
>


From moshez at math.huji.ac.il  Sun Apr 23 01:09:17 2000
From: moshez at math.huji.ac.il (Moshe Zadka)
Date: Sun, 23 Apr 2000 08:09:17 +0300 (IDT)
Subject: Python for drivers?
In-Reply-To: <slrn8g4gbn.33l.jepler.lnk@potty.housenet>
Message-ID: <Pine.GSO.4.10.10004230806560.15546-100000@sundial>

On Sun, 23 Apr 2000, jepler epler wrote:

> If you can use a tool like SWIG (or a hand written C module) to provide
> access to the necessary low-level features, you may be able to do this.
> 
> For instance, if the card needs the idea of 'store {8,16,32}-bit integer X at
> index L of the video card's address space' and 'read {8,16,32}-bit
> integer at index L of the video card's address space', then you'd just
> need to write those following functions in a C module for Python to call.

Um...unless you're talking about something like Hurd-like userspace
drivers, it'd still be pretty close to impossible. In general, you avoid
memory allocation and deallocation inside the kernel like the flu,
especially few-bytes allocation. On the other hand, the Python interpreter
usually allocates and dellocates like crazy. This would probably kill
a Linux kernel.

--
Moshe Zadka <mzadka at geocities.com>. 
http://www.oreilly.com/news/prescod_0300.html
http://www.linux.org.il -- we put the penguin in .com




From teemu.keskinarkaus at ramk.fi  Sun Apr 30 07:36:25 2000
From: teemu.keskinarkaus at ramk.fi (Teemu Keskinarkaus)
Date: Sun, 30 Apr 2000 11:36:25 GMT
Subject: Reading user input
References: <39096d88.3112113532@news.ramk.fi> <01d101bfb108$88862290$858647c1@DEVELOPMENT>
Message-ID: <390c1a3e.3287338993@news.ramk.fi>

On Fri, 28 Apr 2000 13:54:39 +0200, =?iso-8859-1?Q?Ulf_Engstr=F8m?=
<ulf.engstrom at b2b-link.com> wrote:

>I guess what you're looking for is raw_input(). Works like this:
>>>> text = raw_input('Enter text here->')
>Enter text here->Anything I want
>>>> text
>'Anything I want'
>Hope it helps :)
That's it! That I was looking for. I've coded with C so many years
that I didn't noticed that. I was looking for getch or readline (or
something like that) function. :) 

Thanks for answers.

TK


From mwh21 at cam.ac.uk  Wed Apr 12 14:54:00 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 12 Apr 2000 19:54:00 +0100
Subject: Off-topic
References: <38F4BE12.A0C73D1D@callware.com> <m3zoqzgois.fsf@atrus.jesus.cam.ac.uk> <38F4C52C.450F611F@callware.com>
Message-ID: <m3wvm3gng7.fsf@atrus.jesus.cam.ac.uk>

Ivan Van Laningham <ivanlan at callware.com> writes:

> Hi All--
> 
> Michael Hudson wrote:
[schnipp]
> > Someone's taken /usr/bin/python away?  Tried strace-ing it?
> > 
> 
> Perl and Python both work if I do 'perl h2n' or 'python hello.py'. 
> Interactive mode works fine.

But are python & perl in /usr/bin?  `which python`?
 
> Sh scripts do work, as long as I remember to do ./script ;-).  I'm
> running as root (but the same thing happens when I run as myself, too). 
> However, this could be explained by the default behavour of exec(),
> which states that if the kernel can't run a program, it execs sh and
> hands it the program name in argv.

Really?  Here (rh 6.1) "man execve" says:

       execve() executes the  program  pointed  to  by  filename.
       filename  must  be either a binary executable, or a script
       starting with a line of the form "#!  interpreter  [arg]".
       In  the latter case, the interpreter must be a valid path?
       name for an executable which is not itself a script, which
       will be invoked as interpreter [arg] filename.

gotta-love-computers-ly y'rs
M.

-- 
  languages shape the way we think, or don't.
                                      -- Erik Naggum, comp.lang.lisp


From tim_one at email.msn.com  Sat Apr  1 22:48:25 2000
From: tim_one at email.msn.com (Tim Peters)
Date: Sat, 1 Apr 2000 22:48:25 -0500
Subject: New Features in Python 1.6
In-Reply-To: <vkuF4.4584$74.73299@newsc.telia.net>
Message-ID: <000501bf9c56$4ce41600$752d153f@tim>

[mailed, not posted]
> now posting from Belgium, Tim Peters wrote:
>> What we've seen here is simply Guido *finally* embracing a
>> policy of giving people exactly what they ask for.

[Fredrik Lundh]
> I think things are worse than this -- someone just sent me
> something that looks like a draft press release, announcing
> that O'Reilly had taken over Python development, with Jon
> Orwant as CTO, and Larry Wall as the Python 1.6 release
> manager.
>
> (it wasn't clear if it was intended for public consumption
> at this time, so please don't spread the word around just
> yet.  if I get permission, I'll post more info to the list).

And so on.  You're *almost* right, but I couldn't say so on the list.  Guido
and I have been fighting this behind the scenes for two solid weeks, which
is why we've both been nearly invisible on c.l.py for the duration.  But,
like they say, the show ain't over 'til the python swallows the rat -- can't
say more than that now, but watch c.l.py for Really Big News from Guido
Himself over the next couple of weeks.  Even you'll be surprised <wink>.

worst-part-of-moving-to-belgium-is-learning-to-belch-ly y'rs  - tim





From moshez at math.huji.ac.il  Thu Apr 27 02:01:42 2000
From: moshez at math.huji.ac.il (Moshe Zadka)
Date: Thu, 27 Apr 2000 09:01:42 +0300 (IDT)
Subject: condense whitespace to single space?
In-Reply-To: <36NN4.31937$Xk2.119087@tor-nn1.netcom.ca>
Message-ID: <Pine.GSO.4.10.10004270900520.11812-100000@sundial>

On Wed, 26 Apr 2000, k wrote:

> How do I condense the sections of spaces in this string:
> 
> "abc              def         ghi"
> 
> into single spaces?
> 

string.join(string.split(s), ' ')

--
Moshe Zadka <mzadka at geocities.com>. 
http://www.oreilly.com/news/prescod_0300.html
http://www.linux.org.il -- we put the penguin in .com




From tseaver at starbase.neosoft.com  Tue Apr  4 09:08:12 2000
From: tseaver at starbase.neosoft.com (Tres Seaver)
Date: 4 Apr 2000 08:08:12 -0500
Subject: Python 1.6 alpha 1 released
References: <Pine.GSO.4.10.10004040035420.21084-100000@sundial> <5E877FD7637AB46E.DABBED9CF3C1D64B.CDDDC05AE7374A78@lp.airnews.net> <8cca93$9oh$1@nnrp1.deja.com>
Message-ID: <B964E942A897FF5F.529887C74CF2FC7E.977238F9475EAC23@lp.airnews.net>

In article <8cca93$9oh$1 at nnrp1.deja.com>,  <fredrik at pythonware.com> wrote:
>Tres Seaver wrote:
>> > it *breaks existing applications coded to the published interface*,
>
>not true.  the documented interface explicitly says that
>there is one single argument.  it does not mention optional
>arguments.

Your insistence that the "documented interface" consists only of the
description in the module reference, and not of the example on the next
page or of the way the API is actually used in the standard libraries
baffles me.

>
>(the example was broken, but so were other examples in earlier
>releases of the documentation.  see other messages in this
>thread for details)
>
>> Consider that *every one* of the protocol modules which offers
>> a connect() method uses a two-argument signature for it:  ftplib,
>> gopherlib, httplib, imaplib, nntplib, poplib, smtplib, telnetlib,
>> urllib.  And they all *use* the two-argument connect when invoking
>> their underlying socket!
>
>not true.  look again (doesn't matter if you look in the 1.5.2
>library or the 1.6 library -- what you just said is not correct
>in either case).

I just did, see below.  The only error I can see is that I falsely lumped
imaplib with the others as *exposing* a two-argument connect(), when in fact
it merely *uses* two-argument connect().

>
>> No Python module in the standard library uses the single-argument
>> connect, period.
>
>*all* python modules in the 1.6 standard library use
>single-argument socket connect.

Invalid argument -- nobody depends on the 1.6 library yet.

Here is a check from my (unfiddled) Python distro -- its the same on Windows,
too:

[/home/tres] $ python
Python 1.5.2 (#1, Sep 17 1999, 20:15:36)  [GCC egcs-2.91.66 19990314/Linux (egcs
- on linux-i386
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> ^D
[/home/tres] $ cd /usr/lib/python1.5
[/usr/lib/python1.5] $ grep "connect(" *.py
asyncore.py:            self.handle_connect()
asyncore.py:                    self.handle_connect()
asyncore.py:                    self.handle_connect()
ftplib.py:      Then use self.connect() with optional host and port argument.
ftplib.py:      # Optional arguments are host (for connect()),
ftplib.py:                      resp = self.connect(host)
ftplib.py:      def connect(self, host = '', port = 0):
ftplib.py:              self.sock.connect(self.host, self.port)
ftplib.py:              (this is read and squirreled away by connect())'''
ftplib.py:                      conn.connect(host, port)
gopherlib.py:   s.connect(host, port)
httplib.py:        if host: self.connect(host, port)
httplib.py:    def connect(self, host, port = 0):
httplib.py:        self.sock.connect(host, port)
httplib.py:    h.connect(host)
imaplib.py:             self.sock.connect(self.host, self.port)
nntplib.py:             self.sock.connect(self.host, self.port)
poplib.py:              self.sock.connect(self.host, self.port)
smtplib.py:            (code, msg) = self.connect(host, port)
smtplib.py:    def connect(self, host='localhost', port = 0):
smtplib.py:        self.sock.connect(host, port)
smtplib.py:            raise SMTPServerDisconnected('please run connect() first'
)
telnetlib.py:        self.sock.connect((self.host, self.port))
urllib.py:        self.ftp.connect(self.host, self.port)


The CVS log comment for the checkin you just made *six days ago* to
smtplib says:

  This fixes a bunch of socket.connect(host, post) calls.  Note that I
  haven't tested all modules -- I don't have enough servers here...

So don't tell me that the 1.5.2 library used one-argument connect().

>
>> This is *not* the same case as the list.append() one, where the
>> published interface was itself consistent.
>
>why not?  after all, there were many more uses of multi-argument
>append in the 1.5.2 standard library than uses of multi-argument
>socket calls.
>
>see my other post in this thread for the list of questions
>I ask myself for each change in Python.  if you use other
>criteria, which make it obvious to you that "append" is a
>good change, and "connect" is a bad one, please tell me what
>they are.
>
>(bogus facts or double standards doesn't count).
>
>after all, with tons of code all over the Python universe,
>I'm affected by *all* backwards incompatible changes in
>Python 1.6.  but for some reason, the "append" change is
>the only one that makes me worried.  what am I missing?

The append() change is hard to find, but fixes something that shouldn't work.
The connect() change could be equally hard to find (although I grant that the
call-through-a-differently-named-bound-pointer bit is much less likely)
but fixes something which *isn't* broken, *for no reason*.

In fact, I think one could make a reasonable case that the real bug here
is in the API doc:  single-argument connect() is unintuitive for AF_INET
sockets, and exists primarily to shoehorn orthagonality with AF_UNIX ones
(where a default port argument would do just as well.)

Still-haven't-seen-a-reasonable-defense-of-one-argument-connect()'ly,

Tres.
-- 
---------------------------------------------------------------
Tres Seaver        tseaver at digicool.com     http://www.zope.org
Digital Creations


From shecter at darmstadt.gmd.de  Tue Apr 25 15:40:16 2000
From: shecter at darmstadt.gmd.de (Robb Shecter)
Date: Tue, 25 Apr 2000 21:40:16 +0200
Subject: Metaclasses?
Message-ID: <3905F4A0.2B39C705@darmstadt.gmd.de>

Hi,

I read an old post about metaclasses (included on the cd-rom with the
book "Mit Python programmieren"), and got very interested, because I
imagine that by making a new metaclass, I could maybe fix the one thing
about Python that annoys me: the obligatory use of "self" in method
declarations and variable access.  (Python's syntax is so clean and neat
in all other areas, I feel like it's a shame that there's excess
verbosity in this one place.)

So - are metaclasses just a random idea, or is there something in the
works?

Thanks,
Robb


From amitp at Xenon.Stanford.EDU  Sat Apr  8 20:37:26 2000
From: amitp at Xenon.Stanford.EDU (Amit Patel)
Date: 9 Apr 2000 00:37:26 GMT
Subject: Problem with gzip module
References: <8cdojr$69n$1@lacerta.tiscalinet.it>
Message-ID: <8cojc6$li9$1@nntp.Stanford.EDU>


 FrancescO <francesco.contini at tiscalinet.it> wrote:
| Hi!
| I can't understand how is the right way to use the gzip mudule to perform a
| compression of a list of files, that i must read with gunzip or winzip
| ecc...
| 
| I have read some examples from starship.python.com but...
| 
| I must write a cron script to zip log files from a directory, now i used the
| zlib module, but for decompress I must use the uncompress method of the
| zlib... the program do it very well, but the compressed files must be readed
| from winzip from a Winnt on the net...
| does anyone explain me, maybe with an example?
| 


I believe the "gzip" format is used for a single file, unlike the
"pkzip" (winzip) format, which is used for multiple files.  In Unix,
the "tar" program is used to turn multiple files into one, and the
resulting "tarball" is compressed using gzip.  Winzip can
read gzipped tar files.  If you need to compress more than one file,
you need to somehow tar them into one.  (Sorry, I don't know the tar
format.)



	- Amit



From sverker.is at home.se  Sun Apr  2 18:50:13 2000
From: sverker.is at home.se (Sverker Nilsson)
Date: Mon, 03 Apr 2000 00:50:13 +0200
Subject: popen2 made >7 times faster!
Message-ID: <38E7CEA5.4CECEE0D@home.se>

Hi,

I was using the popen2 module to call out to a program. I found it
suspiciously slow to run a small quick program so I began to look for
the reason. Turned out that there was a loop in __init__ in the Popen3
class that closes files numbered 3 to 255 before it exec's the new
program. This took a very long time compared to the other code and
also compared to the time for the OS's fork and exec functions.

When I rewrote this loop in C it became much faster:

The time to call out 100 times to a hello world program dropped from
15 seconds to 2 seconds (*).  The time for 100 calls to the actual
program I wanted to run (PGP) dropped from 18 seconds to 6 seconds.

I guess there are a number of ways to fix this, I made just a quick
hack to close the files 3 to 255 from C. Should be better to keep track
of the files that are actually open to avoid the system calls
altogether - probably not too significant if there could only be 256
files open but would become a problem if the OS allows more
files. Also, popen2 has an hard-coded assumption that there can only
be 256 files open, regardless of what the OS actually allows.

Anyway, maybe somebody is working on this already...? 

'yours

Sverker Nilsson

(*) Pentium 90 Linux 2.0


From mikael at isy.liu.se  Tue Apr 25 05:21:29 2000
From: mikael at isy.liu.se (Mikael Olofsson)
Date: Tue, 25 Apr 2000 11:21:29 +0200 (MET DST)
Subject: Strange translation in GET
Message-ID: <XFMail.000425112129.mikael@isy.liu.se>

Hi!

This question is certainly not related to python, but I haven't the 
slightest idea where else to post it. Since I have great confidence in 
the participants of this group, I dare to post it here. Here goes:


I have a python script that is called as

  http://wherever/whatever?a=A&b=B

The argument ends up as

  'a=A&b=B'

which is perfectly OK. When I call this script in a link on a web page
using the HTML code

  <a href="http://wherever/whatever?a=A&b=B">spam</a>

it usually works OK.


Now to the problem: 

According to my logs, this link was recently clicked by someone in Japan, 
and my script failed because the argument had somehow been translated to 

  a+AD0-A+ACY-b+AD0-B

That is, each '=' was translated to '+AD0-', and each '&' was translated
to '+ACY-'. This probably has to do with the character set used by the
caller. 


My questions are:

  Does anyone know what might have happened here?
  Shall I suspect this to happen again?
  Should I consider it worth while fixing this?

I guess this all boils down to:

  Is this _the_ Japanese standard, or only _a_ Japanese standard?

Fixing this particular peculiarity is easy, but if I can expect several 
different variations on this theme, it seems uninterresting unless there 
is a way to cover all these variations easily.

I wrote the script early, when I had little experience in Python, so I
did not use any cgi-whatever module. So I should perhaps also ask:

  Is this (and other) translation(s) covered by some snazzy module?


someone-will-probably-tell-me-to-read-some-friendly-manual-ly y'rs

/Mikael

-----------------------------------------------------------------------
E-Mail:  Mikael Olofsson <mikael at isy.liu.se>
WWW:     http://www.dtr.isy.liu.se/dtr/staff/mikael
Phone:   +46 - (0)13 - 28 1343
Telefax: +46 - (0)13 - 28 1339
Date:    25-Apr-00
Time:    11:12:26

This message was sent by XF-Mail.
-----------------------------------------------------------------------


From effbot at telia.com  Thu Apr 20 04:40:59 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Thu, 20 Apr 2000 08:40:59 GMT
Subject: Testing a module's return
References: <38FE95E3.1A10D5B0@kraken.itc.gu.edu.au>
Message-ID: <vozL4.3190$Za1.47013@newsc.telia.net>

Joal Heagney <sctheagn at kraken.itc.gu.edu.au> wrote:
> I've found that I have to define several of my modules such that they
> return a string if they fail, rather than 0. (Let's face it, there are a
> lot more things that can go wrong than right). Now I originally wanted
> to have these modules return the values I was after. eg.
>
> def function_name(arg1, arg2, arg3, .... argN):
>     do something with the args, return value
>
> But as a beginner, *grins* I've found returning error strings a little
> more important
>
> def function_name(arg1, arg2, arg3, .... argN):
>     check for error condition 1:
>         return 'Whoops'
>     check for error condition 2:
>         return 'What are you doing, Dave?'
>     check for error condition 3:
>         return 'That ain't right, or humanly decent either'
>     do something with the args, return something.
>
> I can't sort out, should I return 0 for a success, and give the answer
> as a set attribute (Eg. a class variable)? Or is there a much neater way
> of doing things? There'd have to be.

if you had read the tutorial, you'd seen the answer: exceptions!

    def function_name(arg1, arg2, arg3, .... argN):
        assert not error condition 1, "Whoops!"
        assert not error condition 2, 'What are you doing, Dave?'
        ...
        do something with the args
        return something

for details, see chapter 8 in the tutorial:
http://www.python.org/doc/current/tut/node10.html

more info on the "assert" statement is found here:
http://www.python.org/doc/current/ref/assert.html

</F>




From karl at ulbrich.org  Wed Apr 26 18:22:00 2000
From: karl at ulbrich.org (Karl G. Ulbrich)
Date: Wed, 26 Apr 2000 18:22:00 -0400
Subject: PIL compile/make bombs out at step #7 -- help?
In-Reply-To: <asIN4.4302$Za1.64185@newsc.telia.net>; from effbot@telia.com on Wed, Apr 26, 2000 at 08:37:26PM +0000
References: <20000426135724.A26663@exeter.exeter.org> <asIN4.4302$Za1.64185@newsc.telia.net>
Message-ID: <20000426182200.B27107@exeter.exeter.org>

Quoting Fredrik Lundh (effbot at telia.com):
[...]
> from the look of it, you either have a broken /bin/sh,
> or you messed something up when editing the Setup
> file...

/bin/sh might really be /bin/bash under the hood, as
Linux seems most-fond of bash, but no other shellscripts
I've seen die on it.

> can you perhaps post the Setup file ?

Here's my setup file -- I disabled the TK part as I neglected
to install that package with Linux, but I have the python .h
files, jpeg and libz stuff.

# Setup file for the Imaging module
# =================================
#
# INSTRUCTIONS:
#
# 1) Edit the configuration settings as described below.
#
# 2) Go to the libImaging directory and build the support library.
#
# 3) Run "make -f Makefile.pre.in boot"
#
# 4) Run "make"
#
# This file has the same format as <pythonroot>/Modules/Setup.in.
#

# --------------------------------------------------------------------
# Remove (or comment) the following line if you wish to build a
# statically linked python interpreter.

*shared*

# --------------------------------------------------------------------
# Edit or remove the library specifications to match your configura-
# tion.  As they stand, they assume that you have the IJG JPEG library
# 6.0a and zlib 1.0.4 (or newer versions of these libraries) in a
# standard location.  You can comment out the Tcl/Tk stuff if you
# don't need ImageTk, or if you're using a custom _tkinter instead
# (like in earlier versions of PIL).

_imaging  _imaging.c decode.c encode.c map.c display.c outline.c path.c \
	-IlibImaging libImaging/libImaging.a \
#
# *** IJG JPEG library (libjpeg) location
#	-I/usr/local/include -L/usr/local/lib -ljpeg \
	-I/usr/include -L/usr/lib -ljpeg \
#
# *** ZLIB (libz) location
#	-I/usr/local/include -L/usr/local/lib -lz \
	-I/usr/include -L/usr/lib -lz \
#
# *** Tcl/Tk support and library location
#        Tk/tkImaging.c -DWITH_TKINTER -I/usr/local/include -L/usr/local/lib -ltcl8.0 -ltk8.0

# --------------------------------------------------------------------
# uncomment to enable the SANE driver (see Sane/README for more info)

# _sane Sane/_sanemodule.c -IlibImaging -L/usr/local/lib/sane/ -lsane-dll



From sabren at manifestation.com  Thu Apr 13 21:49:41 2000
From: sabren at manifestation.com (Michal Wallace (sabren))
Date: Thu, 13 Apr 2000 21:49:41 -0400 (EDT)
Subject: how do I know to what module an object belongs?
In-Reply-To: <Pine.BSF.4.10.10004132122420.4951-100000@chi.pair.com>
Message-ID: <Pine.BSF.4.10.10004132146270.4951-100000@chi.pair.com>

On Thu, 13 Apr 2000, Michal Wallace (sabren) wrote:

>    I know I can just as easily have them pass a reference to the
> module along with it, but I'd be happier if I could just look at the
> connection object and know which module it came from. Can I do this?

Um.. Okay....

for just about every module I've tried, I can do:

module = eval(dbc.__class__.__module__)

... EXCEPT for ODBC.Windows..

>>> dbc
<MS ODBC Manager Connection object at 76c3b0>
>>> dbc.__class__
Traceback (innermost last):
  File "<stdin>", line 1, in ?
AttributeError: __class__
>>>

what gives?

Cheers,

- Michal
-------------------------------------------------------------------------
http://www.manifestation.com/         http://www.linkwatcher.com/metalog/
-------------------------------------------------------------------------




From KingArthur at roundtable.dyndns.org  Thu Apr 20 19:37:53 2000
From: KingArthur at roundtable.dyndns.org (news-server)
Date: Thu, 20 Apr 2000 23:37:53 GMT
Subject: Medusa and POSTing...
Message-ID: <lxML4.7211$4w1.101767@typhoon.nyroc.rr.com>

I'm writing an MP3 Server/Player in Python using the Medusa HTTP server and
my own handlers written in Python... Alls well & good until I try to read
POSTed data from the client. I came up with a sure-fire way to test it (an
HTML form with a query button and hidden entry fields that submits its data
via posting).

Does anyone know how to get the form data? I've tried reading stdin and the
CGI module. I've also tried browsing through the medusa code to find
something but haven't had much success.

I appreciate the help,

John Brainard

KingArthur at roundtable.dyndns.org_remove_me





From spamfranke at bigfoot.de  Thu Apr 13 18:54:17 2000
From: spamfranke at bigfoot.de (Stefan Franke)
Date: Thu, 13 Apr 2000 22:54:17 GMT
Subject: DCOM configuration tool for Win98/2
Message-ID: <38f64ecb.1417276@news.btx.dtag.de>

This may be of interest for anyone who tries to follow the DCOM
examples in chapter 12 of "Python Programming in Win32":

Obviously Microsoft forgot to include the configuration tool
(dcomcnfg.exe) on the Windows 98 2nd Edition CD. 

The article 

    http://support.microsoft.com/support/kb/articles/Q253/3/11.ASP

contains a link to the download site.


Stefan




From darrell at dorb.com  Sun Apr  9 01:16:13 2000
From: darrell at dorb.com (Darrell)
Date: Sun, 09 Apr 2000 05:16:13 GMT
Subject: Algorithm: combinations of (k) taken from (n) values
References: <meh9-ACCD3B.16041406042000@news.cit.cornell.edu> <38ED5285.719F26D7@udel.edu>
Message-ID: <xmUH4.543$4w1.13051@typhoon.nyroc.rr.com>

Pasted the Charles algorithm into a class using a coroutine to provide
iterators.
The idea being to make the algorithm lazy.
Is it possible to figure out a combination at some huge offset without
figuring out all previous combinations? Seems like it might be, in which
case the isn't much use.

import coroutine
class Choices:
    def __init__(self,n,k=None) :
        """
        Prepares to return elements from a list consisting of all k-element
(ordered)
        subsets of {0 ... n-1}.
        """
        self._combinations=[]
        self._skip=1

        def combs(n=n,k=k, self=self):
            """returns sorted list of k items taken from n."""
            l=self._combinations
            if k==None:
                k=n
            for c in range(n):
                l.append([c])
            for i in range(k-1):
                li = []
                for c in l:
                    for j in range(c[-1]+1,n):
                        li.append(c+[j])
                            // Bet Christian can show me a better way to do
this.
                        if self._skip==0:
                            self._combinations=li
                            coroutine.resume_caller (li[-1])
                        self._skip=self._skip-1

                l = li
            return l

        self._coroutine=coroutine.coroutine(combs)

    def next(self):
        self._skip=1
        return self._coroutine.resume()

    def __getitem__(self, i):
        if i < len(self._combinations):
            return self._combinations[i]
        cnt=i-len(self._combinations)+1
        self._skip=cnt
        self._coroutine.resume()
        return self._combinations[i]

    def __getslice__(self, i, j):
        self[j]
        return self._combinations[i:j]

choices=Choices(30,2)
for x in range(3):
    print choices.next()

print 'getitem:',choices[7]
print 'getslice:', choices[40:50]






From bingdo19 at earthlink.net  Thu Apr 20 10:41:57 2000
From: bingdo19 at earthlink.net (Bob Langdon)
Date: Thu, 20 Apr 2000 14:41:57 GMT
Subject: Test_time Fails? / Installing on Virtual DOmain
Message-ID: <bingdo19-5535DF.10511520042000@news-server.twcny.rr.com>

Hi,

Upon installing Python in the home directory of a virtual domain (after 
making the necessary configure file changes; i.e.: replacing all 
instances of "/usr/" with: "/home/username/python/usr/")

"make test" results in 1 error:

>>1 test failed: test_time
19 tests skipped: test_al test_audioop test_bsddb test_cd test_cl 
test_crypt test_dbm test_dl test_gdbm test_gl test_gzip test_imageop 
test_imgfile test_nis test_rgbimg test_sunaudiodev test_thread 
test_timing test_zlib<<

Running Lib/test/test_time.py results in:
>>Lib/test/test_time.py: import: command not found
Lib/test/test_time.py: time.altzone: command not found
Lib/test/test_time.py: line 5: syntax error near unexpected token `t'
Lib/test/test_time.py: line 5: `t = time.time()'<<

Is this something I've failed to install, or something missing from the 
system?


From claird at starbase.neosoft.com  Tue Apr  4 13:42:01 2000
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 4 Apr 2000 12:42:01 -0500
Subject: Why should I switch to Python? - Infinity of Primes
References: <NDBBKEGCNONMNKGDINPFGEJDDFAA.infonuovo@email.com> <8ccoug$f4o$1@pegasus.csx.cam.ac.uk>
Message-ID: <7F51E902E09F3F60.8F612607812ED16B.82CB8B41435E14FD@lp.airnews.net>

In article <8ccoug$f4o$1 at pegasus.csx.cam.ac.uk>,
Nick Maclaren <nmm1 at cus.cam.ac.uk> wrote:
>
>In article <NDBBKEGCNONMNKGDINPFGEJDDFAA.infonuovo at email.com>, "Dennis E. Hamilton" <infonuovo at email.com> writes:
>|> The standard approach is a proof by contradiction starting from the
>|> assumption that there is a largest prime.
>
>There is also a constructive proof based on the Fundamental
>Theorem of Arithmetic, that is little more complex.
			.
			.
			.
To what extent are constructive proofs supplanting *... ad
absurdum* as standards?  At about the time I left academe
I was working on my own arguments that our conventional
view of the former as "more complex" is, as Brouwer et al.
taught, essentially only convention.  Tilting at construc-
tivist windmills is one of the half-dozen careers I'm
considering for my retirement; I recognize, though, that
I've quite lost touch with current practice.
-- 

Cameron Laird <claird at NeoSoft.com>
Business:  http://www.Phaseit.net
Personal:  http://starbase.neosoft.com/~claird/home.html


From nospam at nowhere.net  Wed Apr 26 02:15:33 2000
From: nospam at nowhere.net (Ray&Maria)
Date: Wed, 26 Apr 2000 00:15:33 -0600
Subject: Unsigned integer arithmetic
References: <newscache$51sktf$rzc$1@jenpc07.jennic.co.uk> <PxvN4.1335$sf5.16905@news.corecomm.net>
Message-ID: <gQvN4.1349$sf5.17376@news.corecomm.net>

> def printU32Hex(n):
>     if n > 0xffffffffL:
>         raise ValueError
>     print "0x%01x%07x" % (int(n >> 28), int(n & 0xfffffffL))

just in case someone send in int by mistake...

def printULHex(n):
    if n > 0xffffffffL or n < 0:
        raise ValueError
    print "0x%01x%07x" % (int(n >> 28), int(n & 0xfffffffL))





From effbot at telia.com  Sun Apr  2 05:16:32 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Sun, 02 Apr 2000 09:16:32 GMT
Subject: Tkinter:  "Get root for a widget" is private?
References: <20000331112048.A1257432@vislab.epa.gov>
Message-ID: <QdEF4.892$n68.190327808@newsb.telia.net>

Randall Hopper wrote:
> I have a widget and want to retrieve the root.  I see the _root() method,
> but by Python convention, isn't this method private?
>
> Should users of Tkinter be calling this?

as long as nobody calls the Tkinter police, you're safe.

however, using the *toplevel* rather than the actual root
object is usually sufficient:

    root = widget.winfo_toplevel()

to get from the toplevel to the real root, just add:

    if root.master:
        root = root.master # real root

</F>




From stevee at redac.co.uk  Fri Apr  7 09:33:59 2000
From: stevee at redac.co.uk (stevee at redac.co.uk)
Date: Fri, 07 Apr 2000 13:33:59 GMT
Subject: Cascading menus
Message-ID: <8cko46$j2i$1@nnrp1.deja.com>

Hello,

I am attempting to convert a program using an old in house GUI system
to use Python and Tkinter. The in house system uses an interactive
stack based language rather like Forth, called EPL.

Our current interface is a horrible! It contains masses of cascaded
menus. We wish to move a more modern style using with much fewer
menus.

However for the first phase we want to end up with the same menu set
as we currently have but implemented in Python. We can then improve
this using Python/Tkinter.

I have produced a translator that converts one of our old style menu
definitions into Python which works OK.

However I have problems with cascading menus due to a major difference
between the way Python and EPL do this. EPL behaves like this:

   A menu button with an associate sub menu knows that it is to
   display a right arrow to the right of the menu and that when
   selected the menu cascade has to remain on screen. It also has some
   EPL code associated with it that gets run whenever the menu button
   is selected. Somewhere inside that EPL code it will ask to display
   a sub menu. That sub menu will be posted in the correct location
   relative to the original menu. This mechanism allows a lot of
   flexibility in what happens before and after the sub menu is
   displayed. One thing that does happen is that a choice between
   several different sub menus can happen.

In python of course sub menus are directly connected to their parent
menu buttons using add_cascade() and python/Tkinter copes with
displaying the sub menu immediately when the user selects the options.

Now I have two possible solutions to this. I either:

 1. translate EPL into a more Python style. This is my preferred
    option as it is a better step towards having a Python based GUI.

or:

 2. Produce a modified Python/Tkinter with a new form of add_cascade
    (with a different name, so we can still use add_cascade) that:

    a) Displays an arrow on the menu button.
    b) On selection it leaves the current menus displayed.
    c) It can have a Python command associated with that we can use to
       run some arbitrary EPL code.
    d) No sub menu is automatically displayed.

    Then add a version of the menu post command that displays the menu
    at the correct location relative to the currently displayed menu.

    The obvious down side of this approach is that we will have our
    own version of Python/Tkinter/Tk that will need to be re applied
    if we wish to move to newer versions of the software in future.
    (This of course assumes that our changes will not be considered
    sufficiently general to go into the real source code.)

    The plus point is that this solution should be much faster to
    implement and I am very short of time (as usual!).

So my questions (finally) about the modified Python/Thinter approach
are:

1. What do people think of it?
2. Is it feasible?
3. Are there any gotchas I haven't considered?

Thanks for any help you may provide.

Steve


Sent via Deja.com http://www.deja.com/
Before you buy.


From rgb122 at my-deja.com  Tue Apr  4 11:00:17 2000
From: rgb122 at my-deja.com (rgb122 at my-deja.com)
Date: Tue, 04 Apr 2000 15:00:17 GMT
Subject: Loading image from URL with PIL
References: <8cau6t$pph$1@nnrp1.deja.com> <8ccagj$a1m$1@nnrp1.deja.com>
Message-ID: <8cd01o$13k$1@nnrp1.deja.com>

> fp = urllib.urlopen(myurl)
> im = Image.open(fp)

That gives me an exception:

Traceback (innermost last):
  File "<interactive input>", line 1, in ?
  File "D:\Program Files\Python\Lib\PIL\Image.py", line 878, in open
    fp.seek(0)
AttributeError: seek

-Rob


Sent via Deja.com http://www.deja.com/
Before you buy.


From bluearms at nownuri.net  Thu Apr 20 16:37:33 2000
From: bluearms at nownuri.net (H.C. Jin)
Date: Fri, 21 Apr 2000 05:37:33 +0900
Subject: can't find SWIG for Python
References: <38FF5A56.F76794D8@sec.noaa.gov>
Message-ID: <39004b90.0@news.nownuri.net>

"j vickroy" <jvickroy at sec.noaa.gov> wrote in message
news:38FF5A56.F76794D8 at sec.noaa.gov...
> Hello all,
>
> I have been unable to locate SWIG for Python.  My references list
> www.swig.org but the server can not be found.
>
> Could someone point me to the site or provide some other Win 32 options
> for easing the process of extending Python to use existing C libraries.
>
> Thanks for your time.
>

I found SWIG at the sourceforge.net with sourceforge search engine.






From jmassung at magpiesystems.com  Mon Apr 24 11:31:31 2000
From: jmassung at magpiesystems.com (Jeff Massung)
Date: Mon, 24 Apr 2000 09:31:31 -0600
Subject: Simple list.append() question
References: <meh9-5ED8CD.11192424042000@news.cit.cornell.edu>
Message-ID: <ZHZM4.360$qg7.2949@news.uswest.net>

I'm just learning, but my guess is this:

>>> l=[[]]*3
>>> l
[[], [], []]
>>> l[0].append(3)
>>> l
[[3], [3], [3]]
>>> id(l[0])
8539088
>>> id(l[1])
8539088
>>> id(l[2])
8539088
>>>

The *3 just made three lists pointing to the same place. I don't know how to
get what you want tho ;) someone else can enlighten me, too :).

- Jeff Massung (jmassung at magpiesystems.com)
- Lead software engineer for Magpie Systems (www.magpiesystems.com)
- Bringing smart pigs and data analysis to the pipeline industry!





From kistler at gmx.net  Tue Apr 11 16:14:38 2000
From: kistler at gmx.net (Per Kistler)
Date: Tue, 11 Apr 2000 22:14:38 +0200
Subject: How to know what re.sub took?
Message-ID: <38F387AE.1010704@gmx.net>

Hello

How to know what has been substituted away, if one does something like:

 >>> import re
 >>> rex = re.compile(r"(hans|fred|simon)")
 >>> st = "max hans bill"
 >>> st2 = rex.sub("",st)
 >>> st2
'max? bill'

It took "hans" away, but how can I learn this automatically?

Thanks, Per.



From johann at physics.berkeley.edu  Tue Apr 25 12:09:09 2000
From: johann at physics.berkeley.edu (Johann Hibschman)
Date: 25 Apr 2000 09:09:09 -0700
Subject: Efficient coding advice?
References: <8e2sjn$te4l$1@swen.emba.uvm.edu>
Message-ID: <mtvh16qi1m.fsf@astron.berkeley.edu>

Kelvin Chu writes:

> data = fp.readframes(nframes)
> left = Numeric.zeros(nframes)
> right = Numeric.zeros(nframes)
> for i in range(0,nframes):
>     left[i] = ord(data[i*2])
>     right[i] = ord(data[i*2+1])

Try:

ords = map(ord, data)            # take the ord of all the chars
ords.shape = (nframes/2, 2)      # reshape the array in place
left  = ords[:,0]                # pick out individual columns
right = ords[:,1]

That should be durn fast, as long as "readframes" isn't the bottleneck.

-- 
Johann Hibschman                           johann at physics.berkeley.edu


From bsb at winnegan.de  Thu Apr 13 16:10:04 2000
From: bsb at winnegan.de (Siggy Brentrup)
Date: 13 Apr 2000 22:10:04 +0200
Subject: threading question
In-Reply-To: "Thomas Heller"'s message of "Thu, 13 Apr 2000 19:00:42 +0200"
References: <018e01bfa569$cda0cd80$4500a8c0@thomasnotebook>
Message-ID: <87r9c9kbj7.fsf@orge.winnegan.de>

"Thomas Heller" <thomas.heller at ion-tof.com> writes:

> I want a thread to run in parallel with the interactive
> interpreter.
> Simple:
> 
> def do_work():
>     import time
>     while 1:
>         time.sleep (1)
>         ...
> 
> import threading

change this
> threading.Thread (target=do_work()).start()
to
  worker = threading.Thread (target=do_work)
  worker.setDaemon(1)
  worker.start()


> But when I hit EOF, python waits until this thread is finished.
> How can I avoid this? I want python to terminate the thread
> and exit in this case.

The main thread is special in installing an exitfunc that waits for
all non-daemonic threads to complete.

Siggy




From ivanlan at callware.com  Wed Apr 12 13:35:36 2000
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Wed, 12 Apr 2000 11:35:36 -0600
Subject: Why should I switch to Python?
References: <seas5ia89177@corp.supernews.com> <slrn8f4d0j.1t0.bwinton@tor.dhs.org> <8d2aa8$n58$1@newshost.accu.uu.nl> <8d2blf$b1n$1@pegasus.csx.cam.ac.uk>
Message-ID: <38F4B3E8.637D6EB5@callware.com>

Hi All--

Nick Maclaren wrote:
> 

[snip]

> According to something I read, incorrectly attributed!  Apparently
> Heinlein got it from something he read, and forgot the fact - it
> was then pointed out to him and he said something equivalent to
> "Oops".  That doesn't stop the attribution to him circulating :-)
> 

I've never heard this story before; I also never encountered the term
before 1961.  To the best of my knowledge, Heinlein is the source of the
term "grok."

Can you provide more information re the attribution story?

<relentless-in-pursuit-of-infinitesimal-detail>-ly y'rs,
Ivan  
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
http://www.pauahtun.org 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
Author:  Teach Yourself Python in 24 Hours



From cjc26 at nospam.cornell.edu  Mon Apr  3 12:06:02 2000
From: cjc26 at nospam.cornell.edu (Cliff Crawford)
Date: Mon, 03 Apr 2000 16:06:02 GMT
Subject: Tuples -- who needs 'em
References: <00e101bf9d80$a8384d20$74eb0b18@stcla1.sfba.home.com>
Message-ID: <slrn8ehggv.h19.cjc26@ith1-491.twcny.rr.com>

* Bob Alexander <bobalex at home.com> menulis:
| This is not an [intentional] flame bait, nor an April Fool's message, but
| think about it:

We have.  Search DejaNews. :)


| Python has two "sequence" data types, lists and tuples,
| which from the Python programmer's perspective are only slightly different.
| How does having both types significantly improve life for Python
| programmers?

Mutable sequences aren't too usable as dictionary keys.


-- 
cliff crawford    -><-    http://www.people.cornell.edu/pages/cjc26/
                                                       icq 68165166
"IS I yes wardrobe yield [the] evaluation."     member of A.H.M.A.D.


From tim_one at email.msn.com  Thu Apr 20 03:41:54 2000
From: tim_one at email.msn.com (Tim Peters)
Date: Thu, 20 Apr 2000 03:41:54 -0400
Subject: __del__ problem - would adopting Garbage Collection fix this?
In-Reply-To: <l03102805b52461abc14d@[193.78.237.146]>
Message-ID: <000701bfaa9b$e729eac0$d62d153f@tim>

[Neil Schemenauer]
> With the default debugging options the collector prints a message to
> sys.stderr when it finds uncollectable objects.  Redesigning the
> class structure should fix the problem.

[Just van Rossum]
> Maybe you're right, but to me it feels this is in the exact same category
> as a __del__ exception, so ignoring & collecting seems more logical to me
> as a default.

Except this isn't Fortran <0.5 wink>:  "in the face of ambiguity, refuse the
temptation to guess" is the Python Way.  That exceptions in __del__ are
diseased is no excuse for designing illness into more areas.  If you don't
want the msg, don't write sick code!

even-five-year-olds-understand-they're-supposed-to-clean-up-their-
    own-mess<wink>-ly y'rs  - tim





From g_will at cyberus.ca  Mon Apr 10 12:46:10 2000
From: g_will at cyberus.ca (Gordon Williams)
Date: Mon, 10 Apr 2000 16:46:10 GMT
Subject: Installer Producing Page Fault
Message-ID: <01bfa30b$e0707f60$2040c3d1@Cyberus>

Hi All,

I am using Gordon's installer program to make a standalone program that
self installs.  After running the install_BS_main.exe created, everything
is put into the proper directory (I think).  Then running BS_main.exe I
immediately get a crash with the following:



BS_MAIN caused an invalid page fault in
module MULTIARRAY.PYD at 0137:01523569.
Registers:
EAX=00000000 CS=0137 EIP=01523569 EFLGS=00010202
EBX=1e123550 SS=013f ESP=0063ebbc EBP=78024838
ECX=18865dbb DS=013f ESI=009673e0 FS=4fff
EDX=0000000b ES=013f EDI=00967990 GS=0000
Bytes at CS:EIP:
8b 08 51 68 74 5c 52 01 57 ff d3 8b 35 84 40 52 
Stack dump:
0063f034 0063f180 78024838 780120cc 1e12e3f5 01520000 1e12e42d 00000003
00000000 1e170238 0063f180 00000000 00000000 0063ec45 bff713e2 00000137 


My code is using the numerical package if that matters. (It appears to be
causing the fault as multiarray is part of the package)

Any ideas?

thanks,

Gordon Williams


From moshez at math.huji.ac.il  Thu Apr 20 15:40:12 2000
From: moshez at math.huji.ac.il (Moshe Zadka)
Date: Thu, 20 Apr 2000 22:40:12 +0300 (IDT)
Subject: docstrings
In-Reply-To: <00bufsonasp7bgkf86lfb6sv5np5v8mqlj@4ax.com>
Message-ID: <Pine.GSO.4.10.10004202239000.21090-100000@sundial>

On Thu, 20 Apr 2000, Ben Wolfson wrote:

> is there a way to use %c and %s within a docstring, or to add strings
> together in a docstring, so that if some strings are used in all
> docstrings, but might be changed, I can simply change one or two
> variables?

Short answer: no

Long answer: of course, but it'll take some doing.

class a:

	__doc__ = 'hello '*3+'world'

Works fine...

--
Moshe Zadka <mzadka at geocities.com>. 
http://www.oreilly.com/news/prescod_0300.html
http://www.linux.org.il -- we put the penguin in .com




From nick at spam_me_notvideosystem.co.uk  Thu Apr 13 10:09:52 2000
From: nick at spam_me_notvideosystem.co.uk (Nick Trout)
Date: Thu, 13 Apr 2000 15:09:52 +0100
Subject: List mapping?
Message-ID: <38f5d68e@news.xtml.co.uk>

Is there a nice way of pairing up the members to form tuple pairs in another
list? ie. like you might to processing input args.

ie. changing ['/a','1','/b','2','/c','3'] into [ ('/a','1'), ('/b','2'),
('/c','3') ]

eg:

args = '/a 1 /b 2 /c 3'
listargs= string.split(args)
tuplist = []
i=0
while i < len(listargs):
    tuplist.append( (listargs[i],listargs[i+1]) )
    i = i+2
for i in tuplist:
    processargs(i)

or:

args = '/a 1 /b 2 /c 3'
for i in string.split(args):
    i = argflag
    # i want next arg now as a parameter!!!

Can I (you!) get rid of the long winded while loop cleverly?!!  :-)

Thanks in advance,
Nick.







From jstok at bluedog.apana.org.au  Wed Apr  5 02:04:10 2000
From: jstok at bluedog.apana.org.au (Jason Stokes)
Date: Wed, 5 Apr 2000 16:04:10 +1000
Subject: complete novice question
References: <38EA5995.F7B9815D@newbridge.com>
Message-ID: <FIAG4.62953$3b6.243291@ozemail.com.au>

Paul Madsen wrote in message <38EA5995.F7B9815D at newbridge.com>...
>Downloaded Python-1.5.2.
>
>I've run 'configure' and it seems to have worked. Now, according to the
>Python package's README, I should be able to type 'make' for the rest of
>the install. If I don't have a 'make' file, did the configure portion
>not work?


Look for Makefile or makefile in the top level directory.

Also, it's usually easier just to install a prepackaged, compiled Python.
For example, in Debian Linux using the dpkg utility.




From drew.mcdowell at msfc.nasa.gov  Fri Apr  7 11:21:02 2000
From: drew.mcdowell at msfc.nasa.gov (Andrew McDowell)
Date: 7 Apr 2000 15:21:02 GMT
Subject: Was Re: need wrapper for ssh/scp
References: <8civ14$rm7$1@hammer.msfc.nasa.gov> <20000407130942.E13830@xs4all.nl> <8cksnu$in2$1@hammer.msfc.nasa.gov>
Message-ID: <8ckucu$jka$1@hammer.msfc.nasa.gov>

Ahhhh Hah!
Found the problem...the /etc/issue file was larger than I was telling
os.read() to expect.
(I was trying to pull some fields out of it for error checking)

Thanks for all the help!
This python group is wonderful.

-Drew

Andrew McDowell <drew.mcdowell at msfc.nasa.gov> wrote
> Thomas Wouters <thomas at xs4all.net> wrote
> > Help is easiest if you also explain what the problem is ;)





From anthony at dstc.edu.au  Tue Apr 18 02:18:06 2000
From: anthony at dstc.edu.au (Anthony J Wilkinson)
Date: Tue, 18 Apr 2000 16:18:06 +1000 (EST)
Subject: urllib redirect
In-Reply-To: <4.1.20000418125557.030592d0@202.95.1.37>
Message-ID: <Pine.OSF.4.20.0004181610230.20942-100000@azure.dstc.edu.au>

On Tue, 18 Apr 2000, Thomas wrote:

> I've got a pretty tame question.
> 
> I want one of my functions to redirect the URL.  
> (ie I call the script and off I go to www.spam.com)
> 
> So, here's what I get, and have not overcome yet:
> 
> import urllib
> def simpleredirect() :
>     ############
>      print "Content-type: text/html"
>      print urllib.urlopen('http://www.spam.com').read()
> 
> simpleredirect()

You really aren't redirecting but attempting to get the data yourself. If
you just print out the following you can use the browser's own redirection
facility.

Content-type: text/html
<html>
    <head>
        <meta HTTP-EQUIV="REFRESH" CONTENT="0; URL=http://www.spam.com/">
        <title> redirect page</title>
    </head>
    <body>
        If your browser does not support redirection, click
        <a HREF="http://www.spam.com/">here</a> to continue.
    </body>
</html>

Cheers,
Anthony
_____________________________________________________________________
Anthony J Wilkinson                                  anthony at dstc.com
Software Engineer                                     http://dstc.com
DSTC Pty Ltd                                     Ph:  +61 7 3365 4310




From dalke at acm.org  Mon Apr  3 01:52:49 2000
From: dalke at acm.org (Andrew Dalke)
Date: Sun, 2 Apr 2000 23:52:49 -0600
Subject: arraymodule questions
References: <8c4iqv$2na$1@nntp9.atl.mindspring.net> <vnlF4.1712$1b4.7736@news-server.bigpond.net.au>
Message-ID: <8c9bee$641$1@slb0.atl.mindspring.net>

Mark Hammond said:
>How about if we better expose a "buffer" object.

>It is basically a fixed array("c"), and supports all slicing and
>assignment operations.  The difference from an array object is
>that its size is fixed, but that can sometimes be a bonus.


> From you description, this would be an alternative solution?

I need a character array which is resizeable, so this won't work.
Some of the final array sizes are not known a priori, and I
need a data structure which can be used, for instance, in a
GUI sequence editor.

                    Andrew
                    dalke at acm.org





From kragen at dnaco.net  Sun Apr 23 02:21:34 2000
From: kragen at dnaco.net (Kragen Sitaker)
Date: Sun, 23 Apr 2000 06:21:34 GMT
Subject: code not true?
References: <Pine.GSO.4.10.10004230758200.15546-100000@sundial>
Message-ID: <ODwM4.1545$RM6.4069477@news-east.usenetserver.com>

In article <Pine.GSO.4.10.10004230758200.15546-100000 at sundial>,
Moshe Zadka  <mzadka at geocities.com> wrote:
>> Not quite.  No number of bits is sufficient to represent .07 exactly in 
>> the typical (and IEEE standard) representation.  
>
>Yes, but enough bits could make sure that 0.07*100 is indistinguishable
>from 7.0

How many decimal digits are needed to make sure that 3 * (1/3) is
indistinguishable from 1?

>But I agree with the point -- floats are evil. Very very evil. In fact,
>I wouldn't mind at all if there was some specific command-line flag to the
>interpreter to enable floats.

That would be kind of nice.
-- 
<kragen at pobox.com>       Kragen Sitaker     <http://www.pobox.com/~kragen/>
The Internet stock bubble didn't burst on 1999-11-08.  Hurrah!
<URL:http://www.pobox.com/~kragen/bubble.html>
The power didn't go out on 2000-01-01 either.  :)


From grant at nowhere.  Tue Apr 18 12:12:53 2000
From: grant at nowhere. (Grant Edwards)
Date: Tue, 18 Apr 2000 16:12:53 GMT
Subject: Running shell programs from Python
References: <38FC64DA.7D297563@ffi.no>
Message-ID: <slrn8fp2c3.13a.grant@grante.comtrol.com>

In article <38FC64DA.7D297563 at ffi.no>, seh at ffi.no wrote:
>I have just started with Python and have some questions.. I have some
>compiled c programs that I normally run from the shell. I can use | to
>pipe the output from one as input to the next. I would like to call the
>programs from Python. Could someone please indicate how this is done?
>Any references?

If you just want to run the programs you can use:

  os.system("foo | bar")

If you want to capture the output of the pipeline:

  p = os.popen("foo | bar","r")
  outputString = p.read()

If you want to write data to the pipeline:

  p = os.popen("foo | bar","w")
  p.write("data to be sent to foo\n")
  p.write("more data to foo\n")
  p.close()

(Remember to do an "import os" if you want to do any of the above)

If you want to both write data to the pipeline and capture it's
output, use the popen2 module, examples of which can be found
in the on-line docs

   <http://www.python.org/doc/current/lib/lib.html>

-- 
Grant Edwards                   grante             Yow!  I'm protected by
                                  at               a ROLL-ON I rented from
                               visi.com            AVIS...


From tim_one at email.msn.com  Sat Apr  1 15:24:37 2000
From: tim_one at email.msn.com (Tim Peters)
Date: Sat, 1 Apr 2000 15:24:37 -0500
Subject: Does Python have a long floating data type please?
In-Reply-To: <38E648F7.D4AB7F7F@math.okstate.edu>
Message-ID: <000601bf9c18$4d147e00$182d153f@tim>

[David C. Ullrich]
> ...
>     Exactly what convinced me that a real should be a pair of
> longRationals (giving a lower bound and an upper bound for the
> "real" value). At least the basic arithmetic operations seem
> simple enough that a person could get them right, and they'd be
> obviously right.
> ...
> class real:
>   def __init__(self, l, u):
>     self.l = l
>     self.u = u
>   def __add__(self, other):
>     return real(self.l + other.l, self.u + other.u)
>
> Pretty exciting stuff. (Well, the correctness of __mul__
> might not be obvious at one glance like __add__ above,
> because of what happens if l < 0 and u > 0, but it should
> nonetheless be extremely trivial to verify.)

Yes, this is what I'd do too -- you'll get something you can trust much
faster this way.  Jurjen's real.py can be viewed as space optimizations of
this approach, first representing the interval by a midpoint and a delta,
then fixing the delta at 1 by introducing a power-of-2 scale factor instead.
The details are tricky, though!

Something all "interval" approaches have problems with is systematic
pessimism.  The classic example is

    y = x - x

y is in fact exactly 0 no matter how fuzzy x is, but straightforward
interval subtraction yields a result centered at 0 and twice as fuzzy as x.

In your original post you had a milder example, x*x.  This has got to be >=
0 in actuality, but if x is e.g. [-2, 2] then straightforward interval
multiplication yields [-4, 4] instead of [0, 4].

Because of this pessimism, some simple series approaches to building
functions (like exp and sin etc) out of the arithmetic primitives can fail
to converge; this is where proofs start to get hard despite (or perhaps
because of <wink>) the simplicity of the primitives.

doable-in-the-end-but-it-doesn't-stay-trivial-ly y'rs  - tim





From andy at petdance.com  Thu Apr 27 09:55:10 2000
From: andy at petdance.com (Andy Lester)
Date: Thu, 27 Apr 2000 08:55:10 -0500 (CST)
Subject: Concatening string and integer
In-Reply-To: <8e9gco$jlt$1@vg170.it.volvo.se>
References: <8e9gco$jlt$1@vg170.it.volvo.se>
Message-ID: <Pine.WNT.4.21.0004270854480.225-100000@nt_alester>

> 
> I want to concatenate a string containing both strings and integers like:
> 
> strTemp = 'I make test number: ' + iMyNumber + ' once again'
> 
> It doesn't work, how should I do this?

strTemp = 'I make test number: %d once again' % iMyNumber

xoxo,
Andy

--
Andy Lester, andy at petdance.com, http://www.petdance.com
    "Inflammable material is planted in my head
     It's a suspect device that's left two thousand dead."
    "Hey, is this the new Green Day?"




From alwagner at tcac.net  Sun Apr  2 13:05:45 2000
From: alwagner at tcac.net (Albert Wagner)
Date: Sun, 02 Apr 2000 12:05:45 -0500
Subject: tcl 8.2 OK?
References: <38E4BE16.497FB1B@tcac.net> <PdEF4.891$n68.190327808@newsb.telia.net>
Message-ID: <38E77DE9.E1BED71C@tcac.net>

Thank you, Fredrik.  I tried a little example from Learning Python:
>>> from Tkinter import *
>>> w = Button(text="Hello", command='exit')
>>> w.pack()
>>> w.mainloop()
It seems to work.

Fredrik Lundh wrote:
> 
> Albert Wagner wrote:
> > For use by tkinter I installed tcl/tk8.2 (I couldn't make 8.0.4
> > compile).  Is this OK? Or can I expect trouble later?
> 
> if you can get a simple program to work (like the test code in
> Tkinter.py), it's probably okay.
> 
> there are some initialization issues with Tk 8.1, but they may be
> fixed in 8.2.  if you get messages like "cannot find usable tk.tcl in
> the following directories" followed by a number of empty lines,
> you need to patch the _tkinter module.
> 
> (don't have a pointer handy; if you need the patch and cannot
> find it, followup to this post).
> 
> </F>

-- 
Small is Beautiful


From tim_one at email.msn.com  Sat Apr  1 15:01:59 2000
From: tim_one at email.msn.com (Tim Peters)
Date: Sat, 1 Apr 2000 15:01:59 -0500
Subject: New Features in Python 1.6
In-Reply-To: <gHrF4.16072$0o4.101062@iad-read.news.verio.net>
Message-ID: <000301bf9c15$23a18f20$182d153f@tim>

[Tom Culliton]
> OK, it's April 1st so maybe I'm just not getting the joke but...
> ...
> I'm not one of the folks who howl at tightening up the behaviour of
> list.append, socket.connect and the like but these changes, especially
> viloating immutability, strike me as too bizarre and dangerous for
> words.  Please tell us it's just and April Fools joke!

It's very well known that Guido has no sense of humor.  What we've seen here
is simply Guido *finally* embracing a policy of giving people exactly what
they ask for <wink>.

i-especially-love-that-"split"-guesses-which-arg-is-the-splittee-ly y'rs  -
tim





From edcjones at erols.com  Mon Apr  3 14:04:43 2000
From: edcjones at erols.com (Edward C. Jones)
Date: Mon, 03 Apr 2000 14:04:43 -0400
Subject: List and Dicts as default args
References: <326536345498D311B3BC00105A39802A074654@newsintern.dspace.de> <vQZF4.1086$n68.187094016@newsb.telia.net>
Message-ID: <38E8DD3B.AC8B0AE1@erols.com>

Fredrik Lundh wrote:

> Stefan Migowsky <smigowsky at dspace.de> wrote:
> > I was just wondering how to handle lists and dictionaries in
> > a simple way as default arguments to functions. Since following
> > strange behaviour occured :
>
> when exposed to "strange" behaviour, please check the FAQ
> http://www.python.org/doc/FAQ.html#6.25
>
> or the fine manual
> http://www.python.org/doc/current/ref/function.html
>
> before posting.

Migowsky asked an advanced newbie question. People who are
inexperienced with a language have not learned enough facts yet to
easily use the documentation (even good documentation, like
Python's). When I have questions like this, I try to guess the "magic
words" (here "default arguments") Then I run a big grep

    grep 'default argument' /usr/doc/python-docs-1.5.2p2/*/*.html

A good context for this is provided by "Words and Rules: The
Ingredients of Language" by Stephen Pinker (Basic Books, 1999).



From nadavh at envision.co.il  Mon Apr  3 13:02:35 2000
From: nadavh at envision.co.il (Nadav Horesh)
Date: Mon, 03 Apr 2000 19:02:35 +0200
Subject: IDLE 0.6 Bug
Message-ID: <38E8CEAB.18BDC87E@envision.co.il>

Someone reported that python 1.6a1 crashes on closing edit window. It
crashes also when closing the help window --- could be the same problem.

    Nadav.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: nadavh.vcf
Type: text/x-vcard
Size: 360 bytes
Desc: Card for Nadav Horesh
URL: <http://mail.python.org/pipermail/python-list/attachments/20000403/76cd90ce/attachment.vcf>

From aek at mail1.stofanet.dk  Tue Apr  4 19:35:59 2000
From: aek at mail1.stofanet.dk (Anders Eggers-Krag)
Date: Wed, 5 Apr 2000 01:35:59 +0200
Subject: Two questions
References: <38EA5FE8.853E9B75@callware.com> <38EA6508.90BF630C@roguewave.com> <38EA6E87.40EDB542@callware.com> <38EA6422.61054B11@exceptionalminds.com> <38EA7452.9BFDF37E@callware.com> <8cdt7c$cl1$1@news101.telia.com> <38EA7D9C.7024CF0B@roguewave.com>
Message-ID: <014801bf9e8e$8a72fec0$2c6ba8c0@frodo>

> > 
> > an alternative solution is to call an ADO (ActiveX Data Object) object and
> > use it through
> > there which is quite simple.
> 
> If you do this, make sure you get Mark's programming win32 book.  It's
> an excellent book for many reasons, but it goes into special detail on
> COM programming (and is so far the only resource I've seen of how COM
> can be simple and useful -- instead of the normal
> add-on-later-because-we-need-the-buzzword I normally associate with COM
> <wink>).
>
it is not like ADO is a complex system, it is extremely easy to use, I use it alle time
writing database systems for ASP.

Anders Eggers - Krag




From bhahn at spam-spam.g0-away.com  Mon Apr  3 19:53:05 2000
From: bhahn at spam-spam.g0-away.com (Brendan Hahn)
Date: Mon, 03 Apr 2000 16:53:05 -0700
Subject: Creating variables on the fly...
References: <meh9-25D0C5.15160303042000@news.cit.cornell.edu> <38E8EFF0.D71D0415@gmd.de> <meh9-32FDF8.16171703042000@news.cit.cornell.edu>
Message-ID: <bhahn-ya02408000R0304001653050001@news.corp.hp.com>

Matthew Hirsch <meh9 at cornell.edu> wrote:
>I like this solution.  How do I use the apply function in the body to 
>call add()?  (I won't know ahead of time how many arguments are going in 
>the function).

You would "apply(add, list-or-tuple-full-of-arguments)".  You don't really
need a separate add-up-a-list function, though, you can use reduce:
"reduce(operator.add, list-or-tuple-full-of-arguments)".

bhahn at transoft.mmangle.net   <-- unmangle address to reply


From kcazabon at home.com  Sun Apr  9 11:28:06 2000
From: kcazabon at home.com (Kevin Cazabon)
Date: Sun, 09 Apr 2000 15:28:06 GMT
Subject: Tkinter:  centering windows on screen
Message-ID: <ak1I4.121418$Dv1.1306619@news1.rdc1.bc.home.com>

I'm having a bit of a problem controlling where on screen a Toplevel() or
Tk() widget appears on the screen...

I have some large widgets that sometimes appear partially off-screen, and I
want to make sure that (at least) they're centered.

Any pointers to the best solution would be appreciated... I've searched
through Fred's Tkinter manual online, but no luck.

Kevin.




From gmcm at hypernet.com  Wed Apr  5 10:41:21 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Wed, 5 Apr 2000 10:41:21 -0400
Subject: Debug builds on WIndows (was Re: (no subject) )
In-Reply-To: <006b01bf9ed9$dd9b5660$1001a8c0@nevrax.net>
Message-ID: <1257176809-9949178@hypernet.com>

Sameh chafik pro wrote:

> 
> Are there any way to have the python16_d.lib and .dll, or the python15_d.dll ?????
> 

You need to build Python from source to get these.

You will need to have python_d.exe, python16_d.dll and the 
debug build of your extension module all in the same dir to 
debug.

You also need to be sure that the python dll and your 
extension dll use exactly the same C runtime library. For 
debug that's "Debug Multithreaded DLL" and for release 
"Multithreaded DLL".


- Gordon



From lbear11 at my-deja.com  Sun Apr 16 22:37:29 2000
From: lbear11 at my-deja.com (lbear11 at my-deja.com)
Date: Mon, 17 Apr 2000 02:37:29 GMT
Subject: Tkinter Canvas -- help?
Message-ID: <8ddtd7$b2k$1@nnrp1.deja.com>

I'm just starting to learn Python/tkinter, and I have found that there
is virtually _no_ documentation for the tkinter canvas. Could anybody
help? I'm specifically looking for stuff like creating an image by a
grid in canvas on "mouseDown", and how to read the items in the canvas,
and write them to a text file, in their proper positions. I would
_really_ appreciate it if anybody could help me with this.


P.S. The only tkinter canvas documentation that I have is the
"Introduction to Tkinter" from pythonware.com

Thanks,
jkior


Sent via Deja.com http://www.deja.com/
Before you buy.


From thomas at xs4all.net  Fri Apr 14 18:49:20 2000
From: thomas at xs4all.net (Thomas Wouters)
Date: Sat, 15 Apr 2000 00:49:20 +0200
Subject: TCP/IP protocols
In-Reply-To: <ZSMJ4.1099$Y14.27933@news.uswest.net>; from jmassung@magpiesystems.com on Fri, Apr 14, 2000 at 04:28:03PM -0600
References: <ZSMJ4.1099$Y14.27933@news.uswest.net>
Message-ID: <20000415004919.B15664@xs4all.nl>

On Fri, Apr 14, 2000 at 04:28:03PM -0600, Jeff Massung wrote:

> I'm new to Python programming, but am very experienced in C/C++. Could
> someone point me to a good reference for TCP/IP programming with Python
> (pdf, book, anything :]) thanks.

The standard lib documentation on the 'socket' library should be what you
need. It's basically a wrapper around the systems' socket() call, which
returns an object which wraps the systems' accept(), bind(), connect(),
send(), write(), read(), recv(), etc. For async network I/O, there's
asyncore.

If you're looking for some higher-level programming, see libraries like
httplib, ftplib, telnetlib, urllib, SocketServer, *HTTPServer, etc.

Documentation for all these modules comes with the python source (in the
Doc/lib/ directory) in TeX format, with most binary distributions in some
form or another, or can be downloaded or browsed online at
http://www.python.org/doc/lib/ (There's an entire chapter, 11, dedicated to
'Internet Protocols and Support'.)

(If you insist, i'm sure we can find some book you can buy that lists the
same thing... but a lot of good docs can be found online, and you can
convert to pdf if you insist ;)

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From jlehmann at nsw.bigpond.net.au  Fri Apr 21 03:55:17 2000
From: jlehmann at nsw.bigpond.net.au (John Lehmann)
Date: Fri, 21 Apr 2000 07:55:17 GMT
Subject: "n choose m" module
Message-ID: <39000874.27D185B3@nsw.bigpond.net.au>


One day, in a fit of boredom when my newscientist subscription was a
week late and I was reduced to solving the enigma puzzle, I wrote a
class to generate the permutations from a list.  I assume I was wasting
my time and that would have been something in the standard libraries,
but in case anyone finds it useful...
-------------- next part --------------
"""

a class for generating permutations from an array

# choose 2 letters from 'abcd'
c = Chooser(['a', 'b', 'c', 'd'], 2)
while c.hasMoreElements():
	print c.next()

"""
import time, copy

class NoMoreChoices(StandardError):
	pass

class Chooser:
	def __init__(self, l, n):
		" generate permutations of n choices of l "
		if n > len(l):
			raise StandardError("cannot make %d choices from a list %d items long"%(n, len(l)))
		self.l = l
		self.n = n
		self.c = range(n)
		self.dirty = 0
		
	def hasMoreElements(self):
		" determine whether there more elements to generate "
		if self.c == None:
			return false
		if self.dirty:
			self.generatenext()
		return self.c <> None

	def next(self):
		" get the next element "
		if self.dirty:
			self.generatenext()
		self.dirty = 1
		return self.getmap(self.c)

	def getmap(self, c):
		m = []
		for i in c:
			m.append(self.l[i])
		return m
	
	def generatenext(self):
		# from the current choice
		self.dirty = 0
		c = self.c
		n = self.n
		ll = len(self.l)
		done = 0
		i = n - 1
		try:
			while not done:
				# take the last element and increment by 1		
				c[i] = c[i] + 1
				# check for modulo overflow
				while c[i] == ll:
					# carry the 1
					c[i] = 0
					i = i - 1
					c[i] = c[i] + 1
					if i == -1:
						raise NoMoreChoices()
				i = n - 1
				if not self.containsduplicates(c):
					done = 1
			self.c = c
		except NoMoreChoices, e:
			self.c = None

	def containsduplicates(self, c):
		" check for duplicate entries "
		for i in range(self.n):
			if c[i] in c[i + 1:]:
				return 1
		return 0

From nospam.newton at gmx.li  Tue Apr  4 15:37:35 2000
From: nospam.newton at gmx.li (Philip 'Yes, that's my address' Newton)
Date: Tue, 04 Apr 2000 19:37:35 GMT
Subject: New Features in Python 1.6
References: <200004011740.MAA04675@eric.cnri.reston.va.us> <gHrF4.16072$0o4.101062@iad-read.news.verio.net> <8cdbvg$3vi$1@news.udel.edu>
Message-ID: <38ea3fda.47259642@news.nikoma.de>

On Tue, 4 Apr 2000 14:30:06 -0400, "Terry Reedy" <tjreedy at udel.edu>
wrote:

>> Python strings can now be stored as Unicode strings.  To make it easier
>> to type Unicode strings, the single-quote character defaults to creating
>> a Unicode string, while the double-quote character defaults to ASCII
>> strings.
>
>' = 1 byte/char, " = 2 bytes/ char is more straightforward.

Yes, but this depends on what you mean by "Unicode". You have to
represent those Unicode characters in bytes, somehow, and how many bytes
you need depends on the format. Also, full Unicode is 32-bit (not just
the 16 bits of the Basic Multilingual Plane). UTF-16 is two bytes (but
can be four for surrogates), UTF-8 is a *variable* number of bytes (1 to
6, I believe), etc. It's not always 2 bytes per Unicode character.

Cheers,
Philip
-- 
Philip Newton <nospam.newton at gmx.li>


From mwh21 at cam.ac.uk  Mon Apr  3 16:17:22 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 03 Apr 2000 21:17:22 +0100
Subject: Python 1.6 alpha 1 released
References: <200004010159.LAA00403@piglet.dstc.edu.au> <8c9nq3$ghk$4@newsserver.rrzn.uni-hannover.de> <m3bt3rcweu.fsf@atrus.jesus.cam.ac.uk> <8cas94$po5$1@newsserver.rrzn.uni-hannover.de>
Message-ID: <m3r9cnym5p.fsf@atrus.jesus.cam.ac.uk>

breiter at usf.Uni-Osnabrueck.DE (Bernhard Reiter) writes:

> True the whole discussion is kind of pointless.
> Nontheless it is not an easy taks to just rewrite your code with
> a simple regular expression replacement. ;-) (And that was my point.)

Downright impossible (the old "can't match nested parens" would rear
it's head, would it not?).  I'd have thought it possible with a real
parser, though.

i-know-lets-flog-this-thread-a-bit-more-ly y'rs
Michael

-- 
well, take it from an old hand: the only reason it would be easier
to program in C is that you can't easily express complex  problems
in C, so you don't.                 -- Erik Naggum, comp.lang.lisp


From m.faassen at vet.uu.nl  Wed Apr  5 14:04:31 2000
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 5 Apr 2000 18:04:31 GMT
Subject: Two questions
References: <38EA5FE8.853E9B75@callware.com> <38EA6508.90BF630C@roguewave.com> <38EA6E87.40EDB542@callware.com> <38EA6422.61054B11@exceptionalminds.com>
Message-ID: <8cfv7f$l0e$2@newshost.accu.uu.nl>

Timothy Grant <tjg at avalongroup.net> wrote:
> mxODBC works beautifully, but I've never used it to access and access
> database, only SQL database. Perchance do you know how Windows
> advertises an Access database to an ODBC driver?

Windows does this automagically. If you have the ODBC manager in the
control panel, you can create a system DSN that points to the Access
database file. It'll allow you to send SQL to your db. You don't even
have to have MS Access installed, just the ODBC stuff.

Of course you can only access it from a Windows box, though there are
ODBC bridging solutions so you can connect to a Windows box with the
database on it over the network and transparently use ODBC from your
Unix machine.

Regards,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From andrew at andrewcooke.free-online.co.uk  Sat Apr 22 16:00:32 2000
From: andrew at andrewcooke.free-online.co.uk (Andrew Cooke)
Date: Sat, 22 Apr 2000 20:00:32 GMT
Subject: LAmbda calculus evaluator (interactive)
References: <8dro55$eq6$1@nnrp1.deja.com> <vndk8hq3t1z.fsf@camelot-new.ccs.neu.edu>
Message-ID: <8dt0cq$nte$1@nnrp1.deja.com>


ok, is this one of those pedantic geek jokes?  the worst problem with
never having studied computer science (apart from having to learn about
lambda calculus now) is that my sense of humour was never warped to the
point where i could detect this kind of thing for sure...

...but if not, i thought the (unquoted) context made clear that i was
talking about programs that manipulated expressions in lambda calculus
(so a "not interactive" program simply takes an expression and follows
some evaluation strategy as far as possible; an "interactive"
program allows me to step reduction by reduction, do abstractions, etc).
in other words, manipulate the appropriate ast.

anyway, i'm writing my own using the "little languages in python"
framework (now called spark) - but would still appreciate pointers to
other work.

cheers,
andrew


In article <vndk8hq3t1z.fsf at camelot-new.ccs.neu.edu>,
  Justin Sheehy <dworkin at ccs.neu.edu> wrote:
> Andrew Cooke <andrew at andrewcooke.free-online.co.uk> writes:
>
> > I know where I can get various (Oca)ML programs, but I'm not sure
> > they're interactive
>
> % ocaml
>         Objective Caml version 2.04
>
> # 1 + 1;;
> - : int = 2
> #
>
> -Justin
>
>


Sent via Deja.com http://www.deja.com/
Before you buy.


From wrd at awenet.com  Mon Apr 10 15:42:38 2000
From: wrd at awenet.com (William R. Dickson)
Date: Mon, 10 Apr 2000 12:42:38 -0700
Subject: Problem running Python script as cron job
References: <wrd-0AEDDA.12282507042000@news.awenet.com> <38EE41F4.8050708@gmx.net>
Message-ID: <wrd-70F5CD.12423810042000@news.awenet.com>

In article <38EE41F4.8050708 at gmx.net>, Per Kistler <kistler at gmx.net> 
wrote:

> Hi Bill
> 
> Try stderr?
> /path/prog > /path/log 2>&1
> 
> Maybe it spits some errors?

Tried that, but I still wind up with an empty log.  Anyway to force it 
to write immediately rather than buffering?

Thanks,

-Bill


From mrfx at 263.net  Wed Apr 19 03:37:12 2000
From: mrfx at 263.net (fmr)
Date: Wed, 19 Apr 2000 15:37:12 +0800
Subject: How to compile wxpython in win98 using vc6.0 for wxWindows2.1.15!
Message-ID: <8djn4c$22ch$1@news.cz.js.cn>

when i compile wxpython in win98 using visual c++ 6.0
for wxWindows2.1.15, i got following error:

wx.cpp:
wchar.h(700) : error C2733: second C linkage of overloaded function
'wmemchr' not allowed
wchar.h(699) : see declaration of 'wmemchr'
...

How can i compile it?
please help me!




From m.faassen at vet.uu.nl  Tue Apr 11 10:11:44 2000
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 11 Apr 2000 14:11:44 GMT
Subject: Pass by reference ?
References: <1257157531-11108477@hypernet.com> <38EBCD7D.7030602@acm.org>
Message-ID: <8cvbr0$q2i$2@newshost.accu.uu.nl>

Robert W. Cunningham <rcunning at acm.org> wrote:
[snip]
> None changed!  Therefore, representations the functions modified were 
> DIFFERENT from those passed, though it is clear that the correct VALUE 
> was passed.

Reference semantics is indistinguishable from value semantics if the
objects manipulated are all immutable. 

There was no modification in your example; there was only assignment.
Assignment does *not* modify an object, unless it is changing a reference
inside a mutable object (such as a[0] = 'foo').

> This behavior meets the classical definition of "Pass By Value".

Reference semantics is indistinguishable from value semantics if the
objects involved are all immutable. Some objects in Python such
as integers, string and tuples, are immutable, others can be mutated.

Reference semantics is indistinguishable from value semantics if you
use a mutable object in an immutable way (this takes lots of discipline,
though).

Assignment only changes what variables (or list elements, or dictionary
elements, or object attributes) refer to, but this indistinguishable from
value semantics if you use immutable objects or treat objects immutably.
Function or method arguments have the exact same semantics as Python
assignment.

I don't think there's any dictionary if you understand this. Python works with
references only. As long as you deal with immutable objects, or treat
objects as immutables, this is identical to value semantics, though.
It's very easy to think about. Easier than C's pointers for instance, in my 
experience.

Regards,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From mwh21 at cam.ac.uk  Fri Apr 28 09:21:02 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 28 Apr 2000 14:21:02 +0100
Subject: Simple list.append() question
References: <meh9-5ED8CD.11192424042000@news.cit.cornell.edu> <ZHZM4.360$qg7.2949@news.uswest.net> <39046FA2.603EAFE5@roguewave.com> <PsaN4.174166$Dv1.2059732@news1.rdc1.bc.home.com> <8e4akc$p7o$1@newshost.accu.uu.nl> <m366t6ozqx.fsf@greebo.nodomain.de> <3908E6EC.B9E6B2E8@cosc.canterbury.ac.nz> <m3zoqenzt6.fsf@atrus.jesus.cam.ac.uk> <39099069.7BFD8387@tismer.com>
Message-ID: <m3puranyyp.fsf@atrus.jesus.cam.ac.uk>

Christian Tismer <tismer at tismer.com> writes:
> > Greg Ewing <greg at cosc.canterbury.ac.nz> writes:
> > > Maybe the LC syntax should include something like
> > >
> > >    [[] times 10]
> > 
> > But wouldn't that make "times" a keyword?  Not sure I'd want that...
> 
> Currently, yes, but it is absolutely not necessary to be so.
> It requires to change the grammar to treat operators
> differently. 

But then what happens to things like

  times = 10
  return [[] times times]

?  I'm not sure context senstive keywords are a good idea (in any
language, ever).  Currently a naked identifier in Python is either a
variable reference or a keyword, and you can tell which just by
looking at *it*, not at its surroundings - and I think I like that.

Though it would be occasionally useful to be able to call a variable
`class'.

Cheers,
M.

-- 
  nonono,  while we're making wild  conjectures about the behavior 
  of completely  irrelevant tasks,  we must not  also make serious 
  mistakes, or the data might suddenly become statistically valid.
                                      -- Erik Naggum, comp.lang.lisp


From dkuhlman at netcom.com  Fri Apr 28 16:32:19 2000
From: dkuhlman at netcom.com (G. David Kuhlman)
Date: 28 Apr 2000 20:32:19 GMT
Subject: Do I always have to write "self." ?
References: <270420001706386648%pecora@anvil.nrl.navy.mil> <slrn8ghc6t.f01.scarblac-spamtrap@flits104-37.flits.rug.nl>
Message-ID: <8ecsgj$6lp$1@slb0.atl.mindspring.net>

> 
> No. However, this seems to make code *more* readable. In fact, many C++
> programmers use 'this.x' for all their instance variables, to be able to
> see at a glance which are just local variables of the function and which are
> instance variables. It's just more work to write, but that doesn't matter
> much.
> 

Thinking about readability, I wrote the C++ coding standard here
where I work.  When we programmers reviewed it, a co-worker
insisted that part of our standard be to prefix the names of all
member variables in any class with "m_" so that we could
distinguish them from other variables.  I believe that "m_" is a
Microsoft-ism and in my humble opinion is just another way to spell
"self.".  I think this shows that even C++ programmers are capable
of seeing the wisdom of the Pythonic way.

  - Dave



From neilh at hare.net.au  Thu Apr 20 21:21:39 2000
From: neilh at hare.net.au (Neil Hodgson)
Date: Fri, 21 Apr 2000 11:21:39 +1000
Subject: Using python on the web
References: <MPG.1365adcfa8470eef9896aa@news> <8di2bk$l1s$1@newshost.accu.uu.nl> <MPG.136631472f892a119896ac@news> <_W2L4.986$GK5.20898@news2-win.server.ntlworld.com> <MPG.13665ef93eedbda29896ad@news> <9N4L4.499$wYl.191816704@newsb.telia.net> <8do2te$kpr$1@nnrp1.deja.com>
Message-ID: <07ae01bfab2f$f2ea09a0$5ddd3fcb@neil>

> DHTML is particularly
> problematic, AFAI can tell, because it doesn't appear to allow the
> segregation of HTML development from coding the way Python string
> substitution does. We hire very skilled HTML people to do HTML and tell
> them to leave our sustitution tags alone, rather than taking the time
> and money to create novice DHTML people to do the same work.

   I think you are confusing Zope's DTML (Document Template Markup Language)
with DHTML, Dynamic HTML. I was about to launch into a stinging attack on
the belittlement of such a wonderful technology as DHTML but then realised
it is probably just an acronym mixup.

   Neil




From grant at nowhere.  Thu Apr 20 17:11:11 2000
From: grant at nowhere. (Grant Edwards)
Date: Thu, 20 Apr 2000 21:11:11 GMT
Subject: How2 use curses for data input
References: <38FF6DE2.5751423A@sprint.ca>
Message-ID: <slrn8fusjd.99a.grant@grante.comtrol.com>

In article <38FF6DE2.5751423A at sprint.ca>, Miles Thompson wrote:

>I 've read the curses part of the manual, and the HOWTO, but I'm not
>much wiser. I need to capture some user input (name of a new directory
>for archiving some web files), and I understand how to define a window,
>etc. What I don't know to do is how to edit the lines, and loop from
>last to first.

You might want to take a look at the "snack" module, which is a
wrapper for the newt library (which is based on slang rather
than curses).  It impliments a set of text-mode widgets such as
buttons, scroll lists, data entry fields, checkboxes, etc.  You
stick a bunch of them on a "form" and pop it up.  When the user
is done, you get the values back somehow (I forget exactly).

If all you need to do is prompt the user for some data, it
should do the trick.

Redhat's text-mode installer is (or at least was) written in
Python using snack/newt.

-- 
Grant Edwards                   grante             Yow!  Why was I BORN?
                                  at               
                               visi.com            


From zsoltman at my-deja.com  Sun Apr  2 19:07:04 2000
From: zsoltman at my-deja.com (zsoltman at my-deja.com)
Date: Sun, 02 Apr 2000 23:07:04 GMT
Subject: NameError when assigning dictionary values
References: <20000402220351.9903.qmail@web2103.mail.yahoo.com>
Message-ID: <8c8jqf$851$1@nnrp1.deja.com>

If I am not mistaken, you are trying create two new list variables on
the fly (If I'm wrong, you can ignore the rest).  If this is what you
are trying to do, you need to use the exec and eval functions.  Below
is an example.
Let me warn though, you must be vigilant with your error trapping.
These functions can end up causing you a major headache otherwise.
(Notice I threw the invalid variable name '^' to show how confusing
this can get)

print
listVariables=['one','^','two','three']
for eachVariable in listVariables:
  execString=eachVariable+'= []'
  try:
    print 'Running statement '+execString
    exec(execString)
  except:
    print 'Something went wrong trying to execute the
statement "'+execString+'"'

print 'one   = ',one
print 'two   = ',two
print 'three = ',three

# to dynamically get these values
print 'Now getting dynamically'
for eachVariable in listVariables:
  try:
   print eachVariable+'   =',eval(eachVariable)
  except:
    print 'Something went wrong trying to evaluate variable
named "'+eachVariable+'"'





In article <20000402220351.9903.qmail at web2103.mail.yahoo.com>,
lewst <lewst at yahoo.com> wrote:
> Gordon McMillan <gmcm at hypernet.com> wrote:
>
> > > # add each key-value pair to the dictionary and
> > > # then for each value, create an empty list.
> > > mydict = {}
> > > for key,value in [ (".", "one"), ("+", "two") ]:
> > > mydict[key] = value
> > > value = []
> > > print mydict
> > > print one,two
> > >
> > > Whe I run this I get "NameError: one". Why aren't my 2 empty lists
> > > named "one" and "two" being created?
> >
> > You're creating two empty lists, both named "value".The first
> > is thown away by creating the second. Nowhere are "one" and
> > "two" variable names, hence the NameError.
>
> Ok. So now how do I get the program to do what I want?
>
> __________________________________________________
> Do You Yahoo!?
> Talk to your friends online with Yahoo! Messenger.
> http://im.yahoo.com
>
>


Sent via Deja.com http://www.deja.com/
Before you buy.


From effbot at telia.com  Tue Apr 11 02:50:38 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Tue, 11 Apr 2000 06:50:38 GMT
Subject: Python paradigms
References: <8cmr70$b0e$1@pegasus.csx.cam.ac.uk> <8covpn$4jq$1@slb6.atl.mindspring.net> <8cqk7u$dmc$1@pegasus.csx.cam.ac.uk> <3H4I4.709$4w1.17557@typhoon.nyroc.rr.com> <8cs3ho$l06$1@pegasus.csx.cam.ac.uk>
Message-ID: <2XzI4.1123$Za1.20108@newsc.telia.net>

Nick Maclaren <nmm1 at cus.cam.ac.uk> writes:
> in what I am doing, they arise naturally out of the problem.

> The pity is that the expression that I am trying to evaluate
> is EXACTLY what the logic of the program requires, but it
> cannot be written in Python (or Fortran) without artificial
> variables.

> No.  Firstly, it relies on an artifact of my example (i.e. that it is
> in an assignment, rather than a test condition or function call)

how about telling us *exactly* what you're doing?

as long as you don't tell us, and just keep replying "no,
that won't work in *my* case" to anyone trying to help
you, you're just wasting everybody's time.

</F>




From pete at visionart.com  Wed Apr 19 22:41:30 2000
From: pete at visionart.com (Pete J Shinners)
Date: Wed, 19 Apr 2000 19:41:30 -0700
Subject: help learning "re"
Message-ID: <38FE6E5A.E7673A83@visionart.com>

i'm trying to get the hang of these regular expression

i'm trying to write a function that takes two filenames
and copies one to the other. while it is copying i want
it to scan for any environment variable names and substitute
the expanded value.

sounded like a fabulous time to learn the regexp.
sadly it doesn't work. the code runs, but none of the
variables are found or substituted

here's what i expected to work
----------------------------------------------
import os, re


def subvar(regfind):
    "callback for regex substitution"
    var = regfind.group(0)[1:]
    print 'REGEX found var:', var #temp debugging. never gets here
    return os.environ.get(var, '$'+var)


def smartcopy(source, dest):
    "copies files. substites env.variables with values"
    srcfile = open(source, 'r')
    dstfile = open(dest, 'w')
    regexp = re.compile(r"\$[0-9A-Z_]+")
    while 1:
        line = srcfile.readline()
        if not line: break
        regexp.sub(subvar, line)
        dstfile.write(line)

-----------------------------------------------
surely this can be done correctly, but how?


From claird at starbase.neosoft.com  Thu Apr 13 07:07:39 2000
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 13 Apr 2000 06:07:39 -0500
Subject: Rewards of celebrity (was: Dr. Dobb's Python-URL! - weekly Python news and links (Apr  3))
References: <42FC2C22B9F94665.58B059EC6AF1AB7B.326AADF624B2C732@lp.airnews.net>
Message-ID: <D12D4DEB60A7C46F.C3C621F660A5B69F.D79F776B10A930F9@lp.airnews.net>

In article <42FC2C22B9F94665.58B059EC6AF1AB7B.326AADF624B2C732 at lp.airnews.net>,
Fredrik Lundh <effbot at telia.com> wrote:
>The Software Development conference in San Jose turned out to be a major
>Python event.
			.
			.
			.
>    Some media coverage.  Like it or not, but Python is hot! :-)
			.
			.
			.
I've had occasion lately to scan a lot of comp.lang.perl
content.  I came across a few surprises that might interest
those with a stake in Python's popularity and prospects,
and perhaps suggestive of the merits of comp.lang.python*
reorganization.

comp.lang.perl.misc has a reputation for nastiness.  The
past few months have been less combative than it was in
1999; still, there's simply no mistaking c.l.p.m for c.l.p.  

The biggest surprise for me has been how thin the c.l.perl*
newsgroups are.  While I don't expect c.l.python's academi-
cism or even its aesthetic excesses, my current impression
is that comp.lang.perl.{misc,modules,moderated,tk} together
don't have as many serious and interesting posts as c.l.python.
It's not for lack of possibilities; Perlites have plenty left
to talk about on such subjects as OO style, new data-management
APIs, Win32 utility, GUI bindings, Unicode, thread capabilities,
numeric extensions, extensibility, how to teach reference use,
...  However, comp.lang.perl.misc is absolutely FULL of lowbrow
trespassing--Web site "developers" whose "shopping carts need
e-mail validation, and they'd rather do it in HTML anyway than
CGI, so could someone please give them either the perlscript
or netscape to make it work, even if it's unix?"  While that's
my paraphrase, it's absolutely not an unfair one; the true
postings are often worse.  Jeff Zucker went to the trouble of
beginning a collection of such drivel for the elevated purpose
<URL:http://www.deja.com/=dnc/getdoc.xp?AN=592780682> of
investigating patterns of learning.

In any case, Pythonians need not fret that they're missing a
whirlwind of production somewhere over in c.l.perl*.  Only a
sliver of what's happening there is other than forgettable.
Most of the good stuff appears to be happening in mailing
lists, which are, again, more pugnacious and posturing than
Python's, but otherwise comparable.
-- 

Cameron Laird <claird at NeoSoft.com>
Business:  http://www.Phaseit.net
Personal:  http://starbase.neosoft.com/~claird/home.html


From davehurt at flash.net  Fri Apr 28 16:25:43 2000
From: davehurt at flash.net (David Hurt)
Date: Fri, 28 Apr 2000 20:25:43 GMT
Subject: newbie   dbase dataaccess
References: <ms19e8.uv5.ln@dorian.georks.de> <mtiigskn8uumhk94rklnr2fdealunasoot@4ax.com> <rd6ce8.ss8.ln@dorian.georks.de>
Message-ID: <btmO4.1519$wb7.140836@news.flash.net>

Take a loot at MetaKit  http://www.equi4.com, I believe

In article <rd6ce8.ss8.ln at dorian.georks.de>, "Georg Hammerbacher" 
<don.limpio at gmx.net> wrote:
>Hello Anders,
>
>thanks for your guess.
>
>The thing inside my head is to write an adressmanager
>- which is the *central* adressbook
>- who works under Windows and Linux
>- has an GUI Interface
>this are the features for the beginning.
>
>my dream is (under the GPL licence), to expand this piece of software and
>integrate the good Features of MS-Outlook and Lotus Organizer. (comments are
>welcome)
>
>I?ve found "gadfly" for databasing. Is it really so good as it seems??
>
>Thanks
>Georg
>
>
>Anders M Eriksson <anders.eriksson at morateknikutveckling.se> schrieb in im
>Newsbeitrag: mtiigskn8uumhk94rklnr2fdealunasoot at 4ax.com...
>> On Thu, 27 Apr 2000 11:40:46 +0200, "Georg Hammerbacher"
>> <don.limpio at gmx.net> wrote:
>>
>> >Hello everybody,
>> >
>>
>> Hello Georg!
>>
>> >I?m just new in programming Python and now I want to make my own personal
>> >adressmanager with a simple dBase-file.
>> >How can I get access to it or where can I find more info?s about using
>> >dbase-files within Python?
>> >
>> First: Why do you want to use dBASE as the database format?
>>
>> I haven't found any good dBASE interface for Python. So if you want to
>> use the dBASE format you need to use ODBC and then you hopefully are
>> on a computer that runs unders Windows NT/98.
>>
>> If you don't NEED to use dBASE then look at
>> http://www.python.org/topics/database/ here you will find alot of info
>> about how to use Python for accessing databases....
>>
>> // Anders
>>
>
>


From pinard at iro.umontreal.ca  Wed Apr 26 09:45:11 2000
From: pinard at iro.umontreal.ca (=?ISO-8859-1?Q?Fran=E7ois_Pinard?=)
Date: 26 Apr 2000 09:45:11 -0400
Subject: new linereading standard?
In-Reply-To: Johann Hibschman's message of "25 Apr 2000 11:23:23 -0700"
References: <3904D21B.333DD930@visionart.com> <390510C9.A744D22E@yahoo.com> <R2g+MIAFMUB5EwUl@jessikat.demon.co.uk> <mt1z3uyr8k.fsf@astron.berkeley.edu>
Message-ID: <oqvh15x9g8.fsf@titan.progiciels-bpi.ca>

Johann Hibschman <johann at physics.berkeley.edu> writes:
> Robin Becker writes:

> > am I wrong for writing

> > for line in open('myfile.txt').readlines():
> >     print line

Under JPython (which I do not know), your file might stay opened until
the next garbage collection.  But I find the above writing so neat that
I prefer to rely on CPython reference counting, and consciously postpone
a lot of tedious rewriting if I ever have to use JPython.

All the clear writings which are allowed because reference counting are now,
for me, a strong source of Python addiction.  I feel ready to rely on it,
after I've been told that Python is going to stick with reference counting.

More to the specifics:

   for line in open('myfile.txt').readlines():
       print line

will give you a double spaced copy of your file, as `print' will add its own
newline after `line', which already has one.   You could write any of:

   for line in open('myfile.txt').readlines():
       print line,

   for line in open('myfile.txt').readlines():
       print line[:-1]

   for line in open('myfile.txt').readlines():
       print line[:-len(os.linesep)]

depending on taste, the system you are using, and the fact that the last
line is terminated or not.  Getting simpler and faster, you might try:

   for line in open('myfile.txt').readlines():
       sys.stdout.write(line)

   sys.stdout.writelines(open('myfile.txt').readlines())

   sys.stdout.write(open('myfile.txt').read())

-- 
Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard





From rcc at nospamthanks_jennic.com  Thu Apr 20 07:07:10 2000
From: rcc at nospamthanks_jennic.com (Robert Cragie)
Date: Thu, 20 Apr 2000 11:07:10 GMT
Subject: Using popen in windows
Message-ID: <newscache$vzabtf$ia6$1@jenpc07.jennic.co.uk>

I tried the following in Windows 98 as an experiment:

from win32pipe import popen
p = popen('dir','r')
str = p.read()
print str

It just hangs when trying to do the read, and I have to kill the shell off.
I have tried IDLE 0.5 and PythonWin and the 'DOS' command line, both do the
same. I am using Python 1.5.2, and the win32pipe.pyd DLL is dated 3rd May
1999.

I have done the equivalent on Linux and it works fine:

from os import popen
p = popen('ls','r')
str = p.read()
print str

Anyone got any ideas?

TIA

Robert Cragie




From jbauer at rubic.com  Sat Apr 22 00:03:06 2000
From: jbauer at rubic.com (Jeff Bauer)
Date: Fri, 21 Apr 2000 23:03:06 -0500
Subject: The Simple Economics of Open Source
References: <3900611F.3B525BDA@cfdrc.com> <1255786824-35980039@hypernet.com> <f4JdfuAkLKA5EwcQ@jessikat.demon.co.uk> <raffael-63EB69.23345321042000@news.ne.mediaone.net>
Message-ID: <3901247A.E87B5D0A@rubic.com>

>> Robin wrote:
>> Painters exchange ideas, concepts and criticism freely,
>> but compete for customers.

> Raffael responded:
> Speaking as a painter, I can tell you that many painters 
> do not share ideas and concepts, but are quite secretive 
> about them. In fact, there is a long history of artists 
> keeping their working methods secret from each other, 
> for precisely the same reason that any other set of 
> business competitors do.

And the success of that approach has made painting
a profitable and lucrative career choice, in contrast 
to the thousands of struggling software engineers who 
can't even qualify for an NEA grant. <wink>

If I didn't have my part-time performance art
income to help pay the bills, I could never
afford to support my programming lifestyle.

-Jeff



From gherman at darwin.in-berlin.de  Mon Apr 10 13:54:47 2000
From: gherman at darwin.in-berlin.de (Dinu C. Gherman)
Date: Mon, 10 Apr 2000 19:54:47 +0200
Subject: [reportlab-users] ReportLab PDF Library v0.92 - print code and 
 presentations to PDF!
References: <PGECLPOBGNBNKHNAGIJHCEPMCAAA.andy@reportlab.com>
Message-ID: <38F21567.C9BC701@darwin.in-berlin.de>

Andy Robinson wrote:
> 
> [announcing RL 0.92]
> 
> New features in this release include:
> - A preview of Dinu Gherman's "py2df", which prints Python
>   code with correct syntax colorizing and a host of options

I've just uploaded the entire final version 0.4 of py2pdf
into the egroups.com ReportLab-users vault:

  http://www.egroups.com/files/reportlab-users/dinu

Please note that there are many sample PDF docs that are
not included with the ReportLab distribution (for good rea-
sons). Also not that this is the final version 0.4 compared
to the current pre-release in RL 0.92.

Unfortunately, I was not able to remove the previous 0.3 ZIP
file from the vault which is most likely because of some 
changes to the Egroups site, but anyway... From here it will
also migrate to my starship cabin today or tomorrow...

Regards,

Dinu

-- 
Dinu C. Gherman
................................................................
"The thing about Linux or open software in general is that 
it actually tries to move software from being witchcraft to 
being a science," [...] "A lot of the programs you see today 
are actually put together by shamans, and you just take it and 
if the computer crashes you walk around it three times... and 
maybe it's OK." (Linus Thorvalds, LinuxWorld 2000, NYC)



From r2d2 at shaka.acc.umu.se  Fri Apr  7 03:27:44 2000
From: r2d2 at shaka.acc.umu.se (Niklas Frykholm)
Date: 7 Apr 2000 07:27:44 GMT
Subject: Teaching the art of programming, in python
References: <20000405053920.20579.qmail@web2102.mail.yahoo.com>    <8cfe42$o02$1@nnrp1.deja.com> <dpJG4.3744$HG1.106895@nnrp1.uunet.ca>    <38EB68A3.B1B5809F@callware.com>    <8cfqa7$4ml$3@newsserver.rrzn.uni-hannover.de>    <38EB73B6.6C0913A7@callware.com> <8chsij$k5r$1@newsserver.rrzn.uni-hannover.de>
Message-ID: <slrn8er3fd.fcd.r2d2@shaka.acc.umu.se>

>>> And all in 24 hours?
>>> This is as misleading as the dummy attribute.
>
>> I don't think so.  You can go at your own speed.  The book is broken up
>> into 24 chapters, each taking more or less than an hour to read.  
>
>Well I guess that a good speed would be 14 days, then. :)
>With two hour lections plus time to exercise.

As an interesting side note, a publisher here in Sweden has
recently started a 10-minute series, with titles such as
"Teach your self Internet in 10 minutes". Seems there is
inflation (or perhaps deflation) in these things.

And it _is_ a silly practice. But then again, corporate
silliness doesn't surprise me that much anymore.

// Niklas


From josht at iname.com  Sun Apr 30 22:00:09 2000
From: josht at iname.com (Josh Tompkins)
Date: Mon, 01 May 2000 02:00:09 GMT
Subject: Circular Dependancies
Message-ID: <8F26D44DDjoshtinamecom@209.208.0.7>

I've read in a couple of places that circular dependancies are not a good 
thing to have in a program.  I'm actually kind of curious:  why is this?  
Is it a style thing, or something that can actually confuse the 
interpreter/compiler?

Thanks.

josue

-- 
________________________________________________________________

"Destined For Great Things -- but pacing myself."
- From a t-shirt.

E-Mail:  josht at crosswinds.net
ICQ:  21219667
AIM:  JosueTheGreat
Web:  http://www.crosswinds.net/~josht
_________________________________________________________________


From kelley at bioreason.com  Tue Apr  4 08:19:31 2000
From: kelley at bioreason.com (Brian Kelley)
Date: Tue, 04 Apr 2000 13:19:31 +0100
Subject: What is python's language level?
References: <38EA1B8F.A797A130@lmco.com>
Message-ID: <38E9DDD3.3313FF6@bioreason.com>

Chuck Meyers wrote:

> Does anyone know python's language level? I did not find it in the list
> at:
>
> http://www.spr.com/library/0langtbl.htm
>
> I would guess that it would be somewhere between 15 and 30. This is an
> important point, even if language levels are imprecise they can make it
> easier to convince people to look at or try a language like python.
>
> -Thanx
>

Okay, I'll bite.  This seems kind of arbitrary to me.  Let me put it this
way, looking at the lists of "language levels", special purpose languages
rank highly:  Mathcad, excel, lotus, etc.

I use python to create libraries for manipulation chemistry and related
feature vectors.  These packages becomes available to our chemists and AI
researches to work with.  From their point of view, the "language level"
for these constructs are incredibly high.  From my point of view the
"language level" is much lower.  So who's right?  They get a big win
because I put time and effort converting python from a general language to
a specific language.  I get a big win because I don't have to teach them
make, gcc compiler flags, etc...

Of course every time somebody creates an HTML/XML/whatever package the
"language level" increases from my point of view.  Batteries included
indeed.

So, who gets to judge the language level, eh?

>
> chuck.meyers at LMCO.COM

--
Brian Kelley          (505) 982-7884 ext 207
Bioreason, Inc        kelley at bioreason.com
105 Washington #303, Santa Fe NM, 87501




From javanet at dynacap.com  Tue Apr 18 13:19:51 2000
From: javanet at dynacap.com (Arnaldo Riquelme)
Date: Tue, 18 Apr 2000 09:19:51 -0800
Subject: BAD starship.python.net news!
References: <38FC852A.9B11F458@concentric.net>
Message-ID: <sfp64err7l573@corp.supernews.com>

I'm willing to donate some money. Just tell me who to talk to or when to
send it.

Manus Hand <mjhand at concentric.net> wrote in message
news:38FC852A.9B11F458 at concentric.net...
> This from Mike McLay.  Everyone, let's pony up what it takes to recover
> our starship!
>
> Manus (who will backup more and more often than he has until now)
>
> > From: Mike McLay (mclay at nist.gov)
> > Subject: Re: starship.python.net??
> >
> > The prognosis isn't good.  The old drive is toast and backup software
> > was never installed.  I've located a data recovery company that will
> > provide a quote on the chances of recovery for $125.  The cost of
> > recovery, if it is possible, will be between $300-$1000.   They can
> > turn around the recovery in a few days time.  Someone will need
> > to fund the recovery.  So far I have a $200 pledge.
> >
> > I bought a new drive and installed it in the server last night. The
> > system will be back online tomorrow by noon.
>




From jp at NOSPAMdemonseed.net  Tue Apr 18 09:30:41 2000
From: jp at NOSPAMdemonseed.net (jason petrone)
Date: Tue, 18 Apr 2000 08:30:41 -0500
Subject: gaming and python
References: <000c01bfa91c$f421d6a0$0f8d0b3f@oemcomputer>
Message-ID: <12ohd8.4rs.ln@demonseed.net>

mrnatch <mrnatch at gateway.net> wrote:
> I told a friend of mine that I am learning to program using Python (I chose
> Python as my first language on the advice of a Llama) and he asked me an
> interesting question: Can Python be used for programming video games? And
> does it work with Windows API? Just curious.

Sure Python can be used for video games!

Though it would be difficult to write a complex game without dipping into any
C programming, there are modules for sound and 3d graphics available.  Check
out PyOpengl.

Later on, when you find enough things you can't do in Python, you can learn
enough C to write some extension modules.

-jp


From mhammond at skippinet.com.au  Sat Apr  1 03:38:18 2000
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Sat, 01 Apr 2000 08:38:18 GMT
Subject: win32all-130 for Python 1.6 available
Message-ID: <_ziF4.1488$1b4.6926@news-server.bigpond.net.au>

Hi all,
    To help encourage testing of Python 1.6, I have made a
version of win32all available.  This should also be considered
alpha!

It is available via
http://starship.python.net/crew/mhammond/win32/Downloads.html -
from that page:

For Python 1.6 only.  If you want a version for Python 1.5.2, see
win32all-128.  Like Python 1.6 itself, this should be considered
alpha. This co-exists fine with Python 1.5 and any of the 1.5
win32all builds.  All my testing shows no problems with 1.5 and
1.6 installed concurrently.

NOTE: The installer will copy Python16.dll from the Python
directory to your system32 directory.  I couldn't even make the
installer load Python16.dll from the Python directory, let alone
Pythonwin or COM objects!

This build has dropped the local Unicode object defined in
pywintypes in favour of the built-in Python Unicode object.
Win32com code no longer converts Unicode objects to strings.  The
COM test suite works fine, which indicates the long awaited
string-and-unicode integration is complete!

The version of Pythonwin in this build has some significant new
features, mainly thanks for a new Scintilla.  Source-code folding
is very cool, and each editor window now has a horizontal
splitter for a second view of the file!  A number of debugger
bugs have been fixed.

Please let me know of any problems...

Thanks,

Mark.




From quinn at hedono.ugcs.caltech.edu  Tue Apr 25 18:23:01 2000
From: quinn at hedono.ugcs.caltech.edu (Quinn Dunkan)
Date: 25 Apr 2000 22:23:01 GMT
Subject: Metaclasses?
References: <3905F4A0.2B39C705@darmstadt.gmd.de>
Message-ID: <slrn8gc6qj.p7e.quinn@hedono.ugcs.caltech.edu>

On Tue, 25 Apr 2000 21:40:16 +0200, Robb Shecter <shecter at darmstadt.gmd.de>
wrote:
>Hi,
>
>I read an old post about metaclasses (included on the cd-rom with the
>book "Mit Python programmieren"), and got very interested, because I
>imagine that by making a new metaclass, I could maybe fix the one thing
>about Python that annoys me: the obligatory use of "self" in method
>declarations and variable access.  (Python's syntax is so clean and neat
>in all other areas, I feel like it's a shame that there's excess
>verbosity in this one place.)
>
>So - are metaclasses just a random idea, or is there something in the
>works?

Metaclasses have been implemented for a long time, see
src/Python-x.y.z/Demo/metaclasses.  I don't think many people use them,
presumably because they like neat and clean, and metaclasses, while clever,
are not (shall we say) conducive to clarity of understanding.

Python's syntax has "excess" verbosity in *many* places (that's a feature),
but self.foo is not one of them.  In python, you *always* access the attribute
of an object with obj.attr  Some people don't like this, and there have been
threads in the past about it, but other people think "more verbosity and less
magic" is more pythonic than "less verbosity and more magic".

Anyway, even if you could figure out the ambiguity problem, I don't think
metaclasses would help you, bytecodehacks might.  Or you could just grit your
teeth, type five extra characters, and as a bonus get something that can be
read by other people.  You could shorten those five chars further by writing a
macro or abbr for your editor.


From BLAISE.BONNEVILLE at wanadoo.fr  Thu Apr 20 03:44:06 2000
From: BLAISE.BONNEVILLE at wanadoo.fr (BLAISE BONNEVILLE)
Date: Thu, 20 Apr 2000 09:44:06 +0200
Subject: print a pixel
Message-ID: <8dmcvb$r6n$1@wanadoo.fr>

I am a beginner in Python and I would like to know how to print a pixel in
Python with Windows 98, given its coordinates.
Thank you!




From cgw at alum.mit.edu  Sat Apr  1 13:46:48 2000
From: cgw at alum.mit.edu (Charles G Waldman)
Date: Sat, 01 Apr 2000 12:46:48 -0600
Subject: new method for string objects
References: <8c44e8$5hj$1@slb6.atl.mindspring.net> <xDnF4.4507$74.72019@newsc.telia.net>
Message-ID: <8c5g6o$prr$1@info3.fnal.gov>

In article <xDnF4.4507$74.72019 at newsc.telia.net>, "Fredrik Lundh"
<effbot at telia.com> wrote: 

> (if numeric python doesn't support the buffer interface,
> it should be fixed). 
> 
> </F>
> 

It doesn't, yet, so I decided to sit down and add this support.  

However I haven't been able to dig up documentation
on the buffer interface - is there a document somewhere
describing this, or do I just need to look at code such as
Objects/stringobject.c and figure out the buffer interface
by reading that?

It's a little more complex in the case of Numeric Python since
Numeric supports non-contiguous arrays.  I see that there is a
bf_getsegcount method which seems like it would be appropriate
in this case, but without a little more information I'm not sure
that I'm using it correctly.





From jamarijr at hotmail.com  Mon Apr 17 12:18:13 2000
From: jamarijr at hotmail.com (Arinté)
Date: Mon, 17 Apr 2000 16:18:13 GMT
Subject: newbie question
Message-ID: <8dfdfk$uum$1@nnrp1.deja.com>

If i PyErr_SetString(PyExc_TypeError,"Invalid command value for IOCTL");
in my c/c++ program, how would python script get that error string?


Thanx


Sent via Deja.com http://www.deja.com/
Before you buy.


From punck at SpamIsBadPenguinPowered.com  Wed Apr 26 21:42:17 2000
From: punck at SpamIsBadPenguinPowered.com (Chris Armstrong)
Date: Wed, 26 Apr 2000 20:42:17 -0500
Subject: Several Questions
Message-ID: <DXLN4.85$M5.512@newsfeed.slurp.net>

Hey all. Got a few questions here, they're all about
a telnet-able game I'm writing, in the spirit of LORD.

1> I'm kinda worried about how much memory each python
process will take up. Is there anything I can do about this?
make it single-threaded maybe? (like the boa httpd) Are
my worries even rational?

2> This one's derived from the last one. Does importing a module
multiple times within a program take up x amount of memory
for each time it's loaded? For instance if I import a module in
my main program and one of the Plugins written for my game
import the same plugin.

3> Speaking of modules, I'm writing one for my battle system.
Is there any reason to make it a class, other than being able
to create instances of it? (I don't need multiple instances.)

4> Currently I run my game with telnetd and the -L parameter
to it, so it can handle different terminals and the like to display
my ncurses display. This seems pretty "ghetto" to me, and I'd like
to know if there are any modules or ways to handle the displaying
to different terms myself.

-- 
Chris Armstrong
http://penguinmints.cx

Geek code version 3.12
GCS d? s:-- a--- C+++ UL++++>++++ P++ L+++ E--- W++ N+ o? K- w-- 
O-- !M V PS+(+++) PE? Y+(--) PGP(--) t+ 5 X+ R tv+ b+>++ DI+ D+++ 
G e- h!(--) r- y?
Go to www.geekcode.com to find out what on earth this is.



From effbot at telia.com  Tue Apr  4 12:26:35 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Tue, 04 Apr 2000 16:26:35 GMT
Subject: tkinter - changing the mouse pointer?
References: <38E91666.D3B87B4C@parlant.com>
Message-ID: <%IoG4.5028$74.81199@newsc.telia.net>

Thomas Lane <tom at parlant.com> wrote:
> Does anyone out there know how to change the mouse pointer in tkinter? I'd
like
> to change the mouse pointer to an hourglass (or the default 'wait' icon
for the
> os) and back to normal after a short operation. How do I go about doing
this?
> I've hunted through the Grayson book without success.

this might help:
http://www.deja.com/[=dnc]/getdoc.xp?AN=392027946

</F>




From ng at hardlight.couk.com  Fri Apr 14 05:54:43 2000
From: ng at hardlight.couk.com (.:|:.)
Date: Fri, 14 Apr 2000 10:54:43 +0100
Subject: exception problem
Message-ID: <38f6ebc3.0@news.proweb.co.uk>

  File "/pr0n/getGroup.py", line 58, in ?
    ArticleToFile(s, `i`)
  File "/pr0n/getGroup.py", line 26, in ArticleToFile
    decode(uufn, binfh)
  File "/usr/lib/python1.5/uu.py", line 95, in decode
    raise Error, 'No valid begin line found in input file'
uu.Error: No valid begin line found in input file

from uu import *
decode(infile, outfile)

raise Error, 'No valid begin line found in input file'
uu.Error: No valid begin line found in input file


how do i trap it

i've tried

try:
    decode(infile, outfile)
except uu.error:
    pass

gives me

NameError uu :


tia

Matt




From musingattheruins at my-deja.com  Thu Apr 13 17:39:15 2000
From: musingattheruins at my-deja.com (musingattheruins at my-deja.com)
Date: Thu, 13 Apr 2000 21:39:15 GMT
Subject: Complaint Dept: My shoes!
References: <8d4s4d$mvr$1@nnrp1.deja.com> <S4nJ4.889$rc9.190032896@newsb.telia.net>
Message-ID: <8d5eq1$d38$1@nnrp1.deja.com>

>     if shoe:
>         print shoe, hit_on_head()

good point, thanks.


Sent via Deja.com http://www.deja.com/
Before you buy.


From sjmachin at lexicon.net  Fri Apr 14 21:08:23 2000
From: sjmachin at lexicon.net (sjmachin at lexicon.net)
Date: Sat, 15 Apr 2000 01:08:23 GMT
Subject: an intern-like memory saver
References: <8d5ing$heb$1@nnrp1.deja.com> <38F7318F.AE4F0EE5@endea.demon.nl>
Message-ID: <8d8fe0$nr5$1@nnrp1.deja.com>

In article <38F7318F.AE4F0EE5 at endea.demon.nl>,
  Niels Diepeveen <niels at endea.demon.nl> wrote:
>
>
> sjmachin at lexicon.net schreef:
> >
> > Problem:
> > I have an application that works with words and can have millions
of them
> > in memory at one time. Apart from the main data structure, a
dictionary is
> > used to maintain frequencies. As the words are loaded from files,
multiple
> > instances of the same word don't share memory. The memory savings
> > could be huge --- see how frequent "the" is in English text,
or "Smith" in
> > an Anglo telephone directory.
> >
> > Trial solution:
> > (1) Clone dictobject.c. Make it into an extension module for a type
called
> > "mydict". Add a method called "key_ref".with one argument:
> > adict.key_ref(obj). If adict.has_key(obj) is true, this returns a
reference to
> > the key value inside the dictionary; else it returns "obj".
> > (2) Make simple changes to the application:
> > (a) Change
> >    freq_dict = {}
> > to
> >    freq_dict = mydict.mydict()
> > (b) assuming for purposes of exposition that words are stored
simply in a
> > list, after
> >    freq_dict[w] = freq_dict.get(w, 0) + 1
> > change
> >    word_list.append(w)
> > to
> >    word_list.append(freq_dict.key_ref(w))
> >
> > Results:
> > Gratifying. An exercise that was running out of real memory (384
MB) and
> > taking a day now takes an hour or so.
> >
> > Questions:
> > (1) Would this be sufficiently generally useful to make it a method
of the
> > standard dictionary object in Python?
>
> It seems to make sense. I don't know if it would be widely used
though.
> I can think of a number of ways to do very much the same thing in
> Python. I don't know whether you've tried, but I think simply using
>     w = intern(w)
> might be even faster than what you describe.

I tried "intern" and read the source code for it. Speed is not a
consideration until I run out of real memory and start swapping to
disk. My objective is to reduce memory usage so that that doesn't
happen. "intern" has its *own* hidden dictionary; if I use it, it may
even *increase* my memory usage. I already have my own dictionary with
all those strings as keys, so I save the overhead part of the intern
dictionary. For what it's worth my key_ref method works with any
objects that can be dictionary keys, not just strings. What are the
other ways you say one can do very much the same thing in Python?

>
> > (3) Has anyone had any success in compiling Python on WinNT 4.0 with
> > gcc 2.95.2? It's just fine for making extension modules; I haven't
tried
> > compiling the whole Python yet.
>
> I'm working on 1.6a2 now (for Win95, should work for NT too). I've got
> the
> core to compile and run, but it takes some patches. I haven't got
> Tkinter working yet (problems linking the Tcl/Tk libaries, any
> suggestions other than building those from source welcome).
> If you like, I can send you what I have so far.

Hmmm ... thanks but no thanks, I don't have any spare time to work on
that at the moment.

>
> > (4) Has anyone any better ideas for gauging Python memory usage than
> > sitting watching the graphical display in WinNT's Task Manager?
Might an
> > instrumented or instrumentable malloc/free package (like Doug
Lea's) that
> > permitted implementation of a Python builtin memused() be the way
to go,
> > or is there a policy of using the standard malloc from the C
library on each
> > platform?
>
> As far as I can tell, you can use any malloc package you like; there
are
> no assumptions other than the C standard.

I was happy enough with the idea of having an instrumented malloc in my
own Python (when I get around to compiling it with gcc), but was
wondering about the possibility of having an instrumented malloc as
part of the *standard* Python distribution.

>
> --
> Niels Diepeveen
> Endea automatisering

Thanks for your interest & comments,
John Machin


Sent via Deja.com http://www.deja.com/
Before you buy.


From mhammond at skippinet.com.au  Mon Apr  3 19:26:09 2000
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Mon, 03 Apr 2000 23:26:09 GMT
Subject: Python and COM without Automation (Dispatch)
References: <vphd7o7b1vc.fsf@hiirihaukka.cs.tut.fi>
Message-ID: <lM9G4.4024$1b4.19451@news-server.bigpond.net.au>

"Vespe Savikko" <vespe at cs.tut.fi> wrote in message
news:vphd7o7b1vc.fsf at hiirihaukka.cs.tut.fi...
>
> From the book _Python Programming on Win32_ I gather that I
would have
> to create a PythonCOM extension DLL and that might even be
quite a
> simple and possibly automated process (with the help of SWIG
and other
> tools), but I haven't been able to find any detailed
instructions on
> the subject.

Yes, and unfortunately there are no detailed instructions.  These
will never become "general tools" - they were written to assist
with the work already done, and are obviously free to be used by
anyone - but they will never become polished, documented tools.

> I haven't walked through the win32com source code yet so if the
> necessary information is located somewhere among the source
files, you
> can ignore this message.

I suggest you use SWIG, as it sounds like you only need to call
interfaces, not implement them.  In this case I would look at the
way MAPI is done, and just clone that.  You will need the SWIG
you should be able to find buried at my starship page...

Mark.





From ng at hardlight.couk.com  Tue Apr 11 07:40:39 2000
From: ng at hardlight.couk.com (.:|:.)
Date: Tue, 11 Apr 2000 12:40:39 +0100
Subject: readline skips on line
References: <8culp8$23a$1@porthos.nl.uu.net> <38F2EB27.751EFFDD@gmd.de>
Message-ID: <38f310b2.0@news.proweb.co.uk>

> try this:
>
> f=open('test.txt', 'r')
>
> while 1:
>   line=f.readline()
>   if not line:
>     break
>   print line
>
> print "ready"

while this is functionallty correct I prefer :

f=open('text.txt', 'r')
line = f.readline()
while line:
    print line
    line = f.readline()

print 'ready'


which is actually a few cycles quicker however
1. the print will append and extra carriage return on the end putting blank
lines in
2. the data will be lost to the porogram
3. repeated disk reads though buffered will slow your inner loop
4. it looks ugly

i'd go for this :

f = open('text.txt', 'r')
lines = f.readlines()

for line in lines:
    print line[:-1]

print 'ready'


> > f=open('test.txt', 'r')
> > while f.readline()!='':
> >     print f.readline()
> > print 'ready'
> >
> > but when it is runned, it wil print out the lines in the text file, but
it
> > skips all the the odd line, it only displays line 2,4,6,8,10 and further





From mwh21 at cam.ac.uk  Tue Apr 25 13:22:03 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 25 Apr 2000 18:22:03 +0100
Subject: Don't understand error message
References: <lukN4.1574$wYl.191480320@newsb.telia.net>
Message-ID: <m3k8hmf64k.fsf@atrus.jesus.cam.ac.uk>

"Lars Lundstedt" <lars.lundstedt at telia.com> writes:

> I'm halfway through "Learning Python" and I'm stuck on the example on page
> 255.
> 
> It goes like this:
> 
> import sys, glob, operator
> print sys.argv[1:]
> sys.argv = reduce(operator.add, map(glob.glob, sys.argv))
> print sys.argv[1:]
> 
> When called with "python glob.py *.py" on Win98 I'm supposed to get a list
> of all .py files in the current directory but instead I get:
> 
> Traceback (innermost last):
>   File "glob.py", line 1, in ?
>     import sys, glob, operator
>   File "glob.py", line 3, in ?
>     list = reduce(operator.add, map(glob.glob, '*.py'))
> TypeError: call of non-function (type module)
> 
> Anyone know why?

"Works here".  Is that *really* your *entire* script?  You haven't
snuck a module called "reduce" in anywhere?

Mind you, I'm on Linux - but that's shouldn't really affect this.

Cheers,
M.

-- 
  "declare"?  my bogometer indicates that you're really programming
  in some other language and trying  to force Common Lisp into your
  mindset.  this won't work.           -- Erik Naggum, comp.lang.lisp


From gmcm at hypernet.com  Tue Apr 11 15:26:52 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Tue, 11 Apr 2000 15:26:52 -0400
Subject: comparing strings and ints
In-Reply-To: <20000411134358.A2352304@vislab.epa.gov>
References: <2YmI4.366$rc9.188014080@newsb.telia.net>
Message-ID: <1256641282-42160637@hypernet.com>

Randall Hopper wrote:

> Fredrik Lundh:
>  |which explains that:
>  |
>  |    Objects of different types always compare unequal,
>  |    and are ordered consistently but arbitrarily.
> 
> Reminds me of qsort() core dumps because the compiled C library compared
> NaNs and Infs arbitrarily.
> 
> I can see 45 == '45' being false, and 45 != '45' being true.  But it seems
> to me that not throwing an exception for attempts at an ordered comparison
> with logical ordering operators (<, <=, etc.) on objects of different
> primitive types (e.g. 45 < '45', 45 > '45') only lets bugs pass through.
> 
> Is there a case where this could be useful (with its undefined behavior)?
> I believe that is the root of the question.

I'd happily bet your wife and firstborn child that there's code that
relies on it <wink>, if only because after sorting a list, all the ints
will end up in a clump, and all the strings in another, etc. 


- Gordon



From rms2000 at bellatlantic.net  Tue Apr  4 09:17:19 2000
From: rms2000 at bellatlantic.net (Richard M. Smith)
Date: Tue, 04 Apr 2000 13:17:19 GMT
Subject: Python and .HTA files in IE5
Message-ID: <38E9EBA2.CF62323A@bellatlantic.net>

Hello,

I am attempting to use Python as a scripting language in a .HTA
(HTML Application) file.  .HTA files are executed by IE5 just
like .HTM files except that there are no secrutiy restrictions on
what script code can do.  In a .HTA files, scripts can run programs,
read and write files, access the Windows registry, etc.  They are usefull
for developing Windows applications that use IE5 for their user
interface.

I am having trouble executing the following simple Python script
to read the contents of AUTOEXEC.BAT and output it on
the .HTA page:

<script language=python>
myfile = open("c:\\autoexec.bat")
document.write(str(myfile.readlines()))
myfile.close()
</script>

I am getting an AttributeError in line 400 of
the file dynamic.py.  It looks like the __getattr__
method is failing when the readlines method is
called by my script code.

Anyone know what is going on here and how to
fix the proble?.

A couple more data points here.  If I rename the .HTA
file to have the .HTM extension, I get the identical error
message.  This leads me to believe that the Python runtime
library isn't distinguishing between .HTA and .HTM files.

OTOH the equivalent .ASP file works fine:

<script language=python runat=server>
myfile = open("c:\\autoexec.bat")
Response.write(str(myfile.readlines()))
myfile.close()
</script>

Richard




From jnw22 at cam.ac.uk  Tue Apr 11 10:39:08 2000
From: jnw22 at cam.ac.uk (James "Wez" Weatherall)
Date: Tue, 11 Apr 2000 15:39:08 +0100
Subject: Huh? func_defaults, default values in function calls
References: <8cv5qj$35p$1@zinc.intern.netconnect.no> <8cv8ra$hio$1@pegasus.csx.cam.ac.uk> <ZFGI4.1192$Za1.20497@newsc.telia.net>
Message-ID: <8cvdec$lb8$1@pegasus.csx.cam.ac.uk>

> > You won't get this behaviour with something like:
> >
> > def myfun(x, a = 5):
> >     a = x
> >
> > though, I don't think, because "5" isn't mutable.
>
> footnote: that has nothing to do with mutability.  in
> the following example, the list isn't modified either:

True.  A better example might be

def myfunc(x, a=[]):
    print x, a
    a.append(x)

def myfunc2(x, a=0):
    print x, a
    a = a + x

I suppose. :)

--
James "Wez" Weatherall





From mwh21 at cam.ac.uk  Wed Apr  5 07:42:48 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 05 Apr 2000 12:42:48 +0100
Subject: Getting the argument names passed to a function ?
References: <slrn8el5t7.40k.quinn@riyal.ugcs.caltech.edu> <8ce6dm$4vp$1@news.wrc.xerox.com> <slrn8elpu4.kjk.quinn@zloty.ugcs.caltech.edu>
Message-ID: <m3r9ckn58n.fsf@atrus.jesus.cam.ac.uk>

quinn at zloty.ugcs.caltech.edu (Quinn Dunkan) writes:

> On 5 Apr 2000 01:55:02 GMT, Mark Jackson <mjackson at wc.eso.mc.xerox.com> wrote:
> >But you forget about:
> >
> >>>> a=1
> >>>> b=a
> >>>> arghack(a)
> >'a'
> >>>> arghack(b)
> >'a'
> >>>> 
> >
> >Your move. . . :-)
> 
> Hmm.  How about:
> 
[snip]

Wow.

Have you downloaded bytecodehacks yet?

Cheers,
M.

-- 
well, take it from an old hand: the only reason it would be easier
to program in C is that you can't easily express complex  problems
in C, so you don't.                 -- Erik Naggum, comp.lang.lisp


From phd at phd.russ.ru  Sun Apr  9 12:23:34 2000
From: phd at phd.russ.ru (Oleg Broytmann)
Date: Sun, 9 Apr 2000 16:23:34 +0000 (GMT)
Subject: tabular data
In-Reply-To: <8cq9io$ca0$1@nnrp1.deja.com>
Message-ID: <Pine.LNX.4.21.0004091620010.13883-100000@fep132.fep.ru>

On Sun, 9 Apr 2000 poupaerte at my-deja.com wrote:
> >    nlist = map( lambda x,indices=indices:
> 
> The expression "indices=indices" seems absolutely strange to me. I
> don't have a clue of what it could be doing, except for being true at
> all times.

   Seems you are a novice, just doing your first steps in Python! Welcome! :)

   "indices=indices" is not an expression at all - it is an assignment. The
"lambda" constructs a function in the global namespace, and the function
doesn't have access to local var "indices". The assignment give such
access.

> Is there some kind of lengthier tutorial on the lambda construct; it
> seems quite an advanced construct not easily to be found elsewhere in
> other programming languages, a bit like regexes ...

   It is from LISP, well-known programming language.

Oleg.            (All opinions are mine and not of my employer)
---- 
    Oleg Broytmann      Foundation for Effective Policies      phd at phd.russ.ru
           Programmers don't die, they just GOSUB without RETURN.




From mjackson at wc.eso.mc.xerox.com  Tue Apr  4 21:55:02 2000
From: mjackson at wc.eso.mc.xerox.com (Mark Jackson)
Date: 5 Apr 2000 01:55:02 GMT
Subject: Getting the argument names passed to a function ?
References: <slrn8el5t7.40k.quinn@riyal.ugcs.caltech.edu>
Message-ID: <8ce6dm$4vp$1@news.wrc.xerox.com>

quinn at riyal.ugcs.caltech.edu (Quinn Dunkan) writes:
> > > Hmm, could you say:
> > > 
> > > def arghack(arg):
> > >     for k, v in globals().items():
> > >         if arg is v:
> > >             return k
> > >     else:
> > >         return None
> > > 
> > > Seems to work ok...
> > This only works if the argument passed in is a global.  For example:
> > >>> def hackSpoiler():
> > ...     a = 2
> > ...     print arghack(a)
> > ...
> > >>> hackSpoiler()
> > None
> 
> Ok, Mr. Spoilsport :)
> 
> import sys
> def arghack(arg):
>     try:
>         1/0
>     except:
>         frame = sys.exc_info()[2].tb_frame.f_back
>     while frame:
>         for k, v in frame.f_locals.items():
>             if arg is v:
>                 return k
>         frame = frame.f_back
>     else:
>         return None

But you forget about:

>>> a=1
>>> b=a
>>> arghack(a)
'a'
>>> arghack(b)
'a'
>>> 

Your move. . . :-)

-- 
Mark Jackson - http://www.alumni.caltech.edu/~mjackson
	An advertiser will happily make you feel bad about
	yourself if that will make you buy, say, a Bic pen.
				- George Meyer




From pj at sgi.com  Thu Apr  6 16:40:08 2000
From: pj at sgi.com (Paul Jackson)
Date: 6 Apr 2000 20:40:08 GMT
Subject: A Mountain of Perl Books + Python Advocacy
References: <000f01bf9eca$2eec8c60$3e2d153f@tim> <38ECF4E6.B85C62D5@sightreader.com>
Message-ID: <8cisn8$ogi7$1@fido.engr.sgi.com>

Ken Seehof wrote:
|> Indeed.  I've noticed that while just about everyone
|> who I've known who has tried python ends up liking
|> it alot, there aren't many fanatics. ...
|> 
|> Perhaps this is partly because python borrows all
|> the best stuff from many other languages rather than
|> being built around some wild, revolutionary idea.

I think it's because Python is too easy.  Fanaticism
is founded on fear of failure - the extra energy
required in the struggle to master (or even get by
with) a tool is sometimes expressed as fanatic devotion.

-- 

=======================================================================
I won't rest till it's the best ...	   Software Production Engineer
Paul Jackson (pj at sgi.com; pj at usa.net) 3x1373 http://sam.engr.sgi.com/pj


From richard.jones at ibahealthcare.com  Wed Apr 19 03:09:01 2000
From: richard.jones at ibahealthcare.com (Richard Jones)
Date: Wed, 19 Apr 2000 08:09:01 +0100
Subject: Very Horrible Question About Goto's
Message-ID: <6325E90C8B6ED21189080001FA7E2ADA0D4051@IBA1>

I'm trying to create a converter from our own internal language (horrid)
into python.? My only problem so far is that the original language uses
*LOTS* of goto's.? Is there any way that this can be replicated within
python ?
?
Thanks for any help.
?
Richard Jones
?
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20000419/3165884a/attachment.html>

From python-list at teleo.net  Mon Apr 17 10:57:38 2000
From: python-list at teleo.net (Patrick Phalen)
Date: Mon, 17 Apr 2000 07:57:38 -0700
Subject: How do I make a constant global variable in Python Please
In-Reply-To: <vCj7OA2LfMfb=znR7yTHOvmiZAkd@4ax.com>
References: <vCj7OA2LfMfb=znR7yTHOvmiZAkd@4ax.com>
Message-ID: <00041708033202.00880@quadra.teleo.net>

[cmfinlay at magnet.com.au, on Mon, 17 Apr 2000]
:: How do I make a constant global variable in Python Please.
:: or just a constant.
:: or just a global.
:: 
:: I bought a Python Book. however I can't find this info in it.

Which book did you buy? A quick look at all the Python books on my
shelf shows that this information is available by looking up "global"
in the index. Did you try that?

I see also that a Google search for "python global" is productive.




From tim.hochberg at ieee.org  Tue Apr 11 01:43:41 2000
From: tim.hochberg at ieee.org (Tim Hochberg)
Date: Tue, 11 Apr 2000 05:43:41 GMT
Subject: Filter
References: <955427129.619305@neptune.uniserve.ca>
Message-ID: <hYyI4.41337$U4.177391@news1.rdc1.az.home.com>

"Matthew" <shayman at uniserve.com> wrote in message
news:955427129.619305 at neptune.uniserve.ca...
> I need to write a filter that inserts tabs between Log values .  So that
> lines 1 to 3 come out looking something like lines 4 to 6.
> I can then open it in Excel as a dat file with the values in different
> cells, where i can then use formulas and macros.
> The first 9 lines are constants in the header ,but there could be serveral
> headers with data  in the whole file .
>
> I'm totaly at square one here.
> I have to figure out how to open the file for modification - as a string ?
> write a for loop for each line beginning with L11 ?
> And a function to insert the white space's between the values.
>
>
> I imagine this to be an ideal task for a small script , but untill i get
> there any advice or help appreciated.

Maybe the following (untested, use at your own risk, etc) code will get you
started and help you refine your question a bit. You will probably wish to
improve it at least some -- for instance I think a blank line (bare newline)
would cause this to hang.

import string

def tabbify(infile, outfile):
    # Read and write the version line
    outfile.write(infile.readline())
    # Read first line
    line = infile.readline()
    # Loop till we run out of data
    while line != "": # No more data to be read (blank lines are "\n"):
        # Write 8 header lines unchanged
        for i in range(8):
            outfile.write(infile.readline())
        # Now read lines, converting spaces to tabs till more header lines
are reached
        while 1: # loop forever
            line = infile.readline()
            if line == "" or line[0] == "H": # File's empty or we have a new
set of headers
                break
            line = string.replace(line, " ", "\t") # replace spaces with
tabs
            outfile.write(line)


infile = open("somefilename")
outfile = open("someotherfilename", 'w')

tabbify(infile, outfile)



Enjoy,

-tim



> VER1.5A
> "H0:1,45"
> "H1:9085,*.*   "
> "H2:71G,NORTHVIEW DLS, ,MITSUI"
> "H3:VANCOUVER,CHILLIWACK"
> "H4:00100S,100,SB4,01, , , "
> "H5:000330, AH00-93-,SPRUCE, ,N"
> "H6:45,126825"
> H7:220000
> L11SSSP 503438G 1627                                      # 1
> L11SSSP 563338G 1826
> L11SSSP 573036G 1924
> L11    SS    SP     55    36    40    G     18    28      # 4
> L11    SS    SP     51    40    46    G     17    31
> L11    SS    SP     57    30    37    H     19    24
> etcXXXXXXXXXXXXXXXXXXXXXXX
> "H9:1,45"
>
>
>




From jae at ilk.de  Mon Apr 24 17:49:58 2000
From: jae at ilk.de (Juergen A. Erhard)
Date: Mon, 24 Apr 2000 23:49:58 +0200
Subject: The Simple Economics of Open Source
In-Reply-To: <8du1in$hth$1@slb7.atl.mindspring.net> (dalke@acm.org)
References: <14591.15630.192472.440375@goon.cnri.reston.va.us><3900611F.3B525BDA@cfdrc.com> <FtDG4z.774@world.std.com><slrn8g26tu.p7o.neelk@brick.cswv.com><14837CB4E98A72FE.8FD547BA6E928594.5A103CFE68327EB7@lp.airnews.net> <VLvUljAKNfA5EwPM@jessikat.demon.co.uk> <8du1in$hth$1@slb7.atl.mindspring.net>
Message-ID: <24042000.2@sanctum.jae.ddns.org>

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

>>>>> "Andrew" == Andrew Dalke <dalke at acm.org> writes:

[...]

    Andrew>   I'm walking down the street and there is a hole in my
    Andrew> front pocket.  I keep money in that pocket.  I have a hand
    Andrew> in that pocket because it's cold, and I'm jostling the
    Andrew> money.  As I pass a beggar, some money falls out through
    Andrew> the hole.  It is altruistic or an accident?  The only way
    Andrew> to find out is to ask me, and I can lie, so there's no
    Andrew> observational way to tell the difference.  Thus, I
    Andrew> considered the definition of "altruistic", when used in
    Andrew> this extreme case, to be useless.

*You* still know what it was... and that is ultimately what counts.

If a tree falls in the forrest and no-one is there to witness it, did
it still fall?  Of course it did.

And of course you know whether it was an accident or intentional.

We're talking motivation here...

Bye, J

- -- 
J?rgen A. Erhard      eMail: jae at ilk.de      phone: (GERMANY) 0721 27326
          My WebHome: http://members.tripod.com/Juergen_Erhard
           "Outside of a dog, a man's best friend is a book;
           inside of a dog, it's very dark." --  Groucho Marx
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.1 (GNU/Linux)
Comment: Use Mailcrypt and GnuPG <http://www.gnupg.org/>

iEYEARECAAYFAjkEwYMACgkQN0B+CS56qs2lLQCggwBbErYOqv/rDoKw7BmlkLXt
XIUAnjUZMaYjyd6yVv7k9e6hNJ5aMAmM
=tCAq
-----END PGP SIGNATURE-----



From mbel44 at dial.pipex.net  Fri Apr 28 04:05:05 2000
From: mbel44 at dial.pipex.net (Toby Dickenson)
Date: Fri, 28 Apr 2000 09:05:05 +0100
Subject: Need help with COMObject for Zope
References: <8e7jeo$s8c$1@nnrp1.deja.com> <T8MN4.5091$v85.31034@news-server.bigpond.net.au> <0EA44467D272A9D6.326DFBDD03E0FA2F.F1D433246D3DFD31@lp.airnews.net>
Message-ID: <kugigscp5r9lplksho2pc66b4h1mh4cb2o@4ax.com>

tseaver at starbase.neosoft.com (Tres Seaver) wrote:

>>Add the 2 lines:
>>import pythoncom
>>pythoncom.CoInitialize()

NOOOoooooooo.......

>>before your code, and it will almost certainly work.
>>
>>The Zope guys should document this fairly well.  Note that Zope is almost
>>certainly initializing COM on a different thread than your code is being
>>called on (hence you see this message) - but there are some serious
>>implications with COM and threading.  

Yes, this is definitely a problem. Zope products can't always play by
the apartment threading rules, therefore you _must_ call
CoInitializeEx(COINIT_MULTITHREADED).

>I dont believe you will hit them here, but someone will sometime :-)
>
>We have a bug in the Collector on the COMObject project for this issue:
>
>  http://classic.zope.org:8080/Collector/842/view
>
>The bug page has three distinct workarounds:  one is essentially what
>Mark posted (it has some ugly issues, but will work, slowly, for most code);

^^^^^^^^^
>one requires hooking the thread creation to perform the CoInitialize() then;

....and works for all code ;-)



Toby Dickenson
tdickenson at geminidataloggers.com


From philipp.jocham at salomon.at  Fri Apr 21 05:19:37 2000
From: philipp.jocham at salomon.at (philipp.jocham at salomon.at)
Date: Fri, 21 Apr 2000 09:19:37 GMT
Subject: Python 1.5.2 on HP-UX 11.00
Message-ID: <J2VL4.465$zm5.447@nreader2.kpnqwest.net>

If tried to compile the pyton interpreter on HP-UX 11.00
with the bundled c-compiler. Boldly I added both
readline(Modules/Setup) and thread support. 

	./configure --with-thread
	./make
	
Everything worked without problems, but when I tried
to run some tests, the executable couldn't be started
because of undefined symbols ( mutex something).

Further investigation showed that the configure process
found the 
	pthread_create()
function in the 'cma' library. Looking in pthread.h
revealed that 
	pthread_create() 
was defined there as a static inline function just calling
	__pthread_create_system()

Now I tricked configure, by editing configure.cache
(just change the yes/no flags at the appropriate tests)
rerun the configure/make process and the interpreter conducts
all tests without error.

Sorry that I don't supply a patch for configure.in (with a test
for __pthread_create_system()), but I'm not sure if this the
correct way to handle this problem.

Bye, Philipp
-- 
Philipp Jocham                            SALOMON AUTOMATION GmbH
                                                Friesacherstr. 15   
mailto: philipp.jocham at salomon.at    A-8114 Friesach bei Stuebing


From boud at rempt.xs4all.nl  Tue Apr 18 01:45:42 2000
From: boud at rempt.xs4all.nl (Boudewijn Rempt)
Date: 18 Apr 2000 05:45:42 GMT
Subject: Scripting and Gnome and KDE
References: <Pine.GSO.4.10.10004150741520.11261-100000@sundial> <8d95m0$9j6$1@news1.xs4all.nl> <8damat$4ja$1@slb2.atl.mindspring.net> <8dbsgc$35o$1@news1.xs4all.nl> <38FB4336.2BC6C235@bioreason.com> <8dftlp$3s1$1@news1.xs4all.nl> <RFPK4.106065$l35.216877@afrodite.telenet-ops.be>
Message-ID: <8dgsq6$a21$1@news1.xs4all.nl>

ben at co.and.co wrote:

> Layout managment was a relief after some MFC experience, and when I made
> the mistake of choosing a 22pt default font in licq (Qt plugin), I
> started seeing the merits of it. But then again, I'm not a seasoned Qt
> developper.

I can imagine with that background ;-). I've spent some time on layout
management in Visual Basic, until I discovered the various third party
controls that automate it, so I feel with you. Of course, Qt offers a
lot of possibilities in layout management. There's even a graphical
gui designer, qtArchitect, that lets you design dialogs with layout
management.

> After reading the tutorial on http://wwww.valdyas.org (hope I get it
> right), I wondered how much PyQt and PyGtk resembled each other.
> IMVVHO, a general purpose signal/slot-mechanism would be nice addition
> to pyton.

The resemblance isn't even skin-deep, I feel, not even when you
discount signals/slots. At least, I feel that knowing my way about
in PyQt doesn't really help me reading PyGTK (or tkInter) code.
wxPython looks more like PyQt, I think.

> There's still a huge problem with the various licences, as they don't
> really match Python's license.

Yes... We asked Troll for a free copy of Qt for Windows, in order to
produce a binary distribution of PyQt (it's an app - not every developer
needs the developers licence of Qt...), but they firmly remained
silent. And that's only the most obvious of the problems ;-(. GPL,
and even the LGLP, aren't as free as Python either.

-- 

Boudewijn Rempt  | http://www.valdyas.org
-- 

Boudewijn Rempt  | http://www.valdyas.org


From dalke at acm.org  Thu Apr  6 07:47:13 2000
From: dalke at acm.org (Andrew Dalke)
Date: Thu, 6 Apr 2000 05:47:13 -0600
Subject: UserDict's has_key() and get() should use __getitem__()
References: <0535db98.ec878f67@usw-ex0101-005.remarq.com> <m3u2hgn5jg.fsf@atrus.jesus.cam.ac.uk> <29328840.06b5b87b@usw-ex0101-005.remarq.com> <8cgdp3$nm5$1@slb6.atl.mindspring.net> <0aa264d3.021266df@usw-ex0101-005.remarq.com>
Message-ID: <8chta6$aam$1@slb6.atl.mindspring.net>

> * Do we know how much less efficient?

Take a look at Barry Warsaw's paper at

 http://www.python.org/~bwarsaw/ecshort.pdf .

The answer is a question - will the lookup mostly fail or
mostly succeed?

] This indicates that the has_key() idiom is usually the best
] one to choose, both because it is usually faster than the
] exception idiom, and because its costs are less variable.
] However, it does indicate that a common pattern that, for
] example, uses dictionaries to count events may be better off
] using the exception idiom, since it is likely that the key
] will be missing from the dictionary only once.

For what you're proposing, there's another method lookup
(since has_key() turns around to call self.__getitem__(key) )
which makes your proposal even slower than those numbers.

                    Andrew
                    dalke at acm.org





From abrahams at mediaone.net  Sun Apr 30 14:00:03 2000
From: abrahams at mediaone.net (Dave Abrahams)
Date: Sun, 30 Apr 2000 18:00:03 GMT
Subject: Do I always have to write "self." ?
References: <270420001706386648%pecora@anvil.nrl.navy.mil> <slrn8ghc6t.f01.scarblac-spamtrap@flits104-37.flits.rug.nl> <8ecsgj$6lp$1@slb0.atl.mindspring.net> <m3bt2tonyg.fsf@atrus.jesus.cam.ac.uk> <300420001008012011%pecora@anvil.nrl.navy.mil><20000430170952.A502@datapro.co.za>
Message-ID: <B531EC9B.57C1%abrahams@mediaone.net>

in article 20000430170952.A502 at datapro.co.za, Steve Crane at
craniac at datapro.co.za wrote on 4/30/00 10:09 AM:

> However if
> you're writing a program yourself and don't think that anyone else will
> ever have to read the code you can probably steer clear of Hungarian
> Notation in either form.

Alternatively, if you're writing a large program with many abstractions and
think lots of other people will have to read it steer clear of Hungarian
Notation in either form ;)

Hungarian breaks abstraction by tying the interface of an object to its
implementation, and seems utterly antithetical to the generic nature of
coding in Python. Who cares if it's a dict or a UserDict, or some other
object which models the appropriate mapping semantics?

flame-baiting-ly y'rs,
dave



From banderson at boi.hp.com  Fri Apr 14 15:47:59 2000
From: banderson at boi.hp.com (Bill Anderson)
Date: Fri, 14 Apr 2000 13:47:59 -0600
Subject: Python/Perl Popularity (Re: A Mountain of Perl...)
References: <200004102111.HAA03380@envy.fulcrum.com.au> <8d0pcq$7ti$1@nnrp1.deja.com>
Message-ID: <38F775EF.3256705B@boi.hp.com>

tom__98 at my-deja.com wrote:
> 
> In article <200004102111.HAA03380 at envy.fulcrum.com.au>,
>   Richard.Jones at fulcrum.com.au wrote:
> > Tom you little troll-meister. Half those arguments
> > you cite are bollocks if you _really_ knew and
> > "greatly prefer" Python. Try again.
> 
> Maybe you can contribute more constructively:
> 
>  -- How can I browse the Python documentation without firing
>     up a web browser?  How do I get information on a single Python
>     function or class quickly?  Is there any equivalent of
>     "perldoc -f split" or "perldoc Net::POP3"?  What do you
>     do in that situation?


A) The docs are available in more than html (see the site)
B) load the interpreter, and print the doc string:
>>> import os
>>> print os.link.__doc__
link(src, dst) -> None
Create a hard link to a file.
>>> 

 
>  -- Is there a Python equivalent of CPAN and the Perl CPAN
>     module?  The closest I know of is Parnassus.  But Parnassus
>     is merely a collection of links, not an archive, and it
>     doesn't have any facilities (AFAIK) for automatic installation.

CPAN is IME, highly overrated.

> 
>  -- Several extensions I have wanted to install have required access
>     to the full Python source tree, not just the header files and
>     the library.  Is that going to change?  Otherwise, how can

I've had perl extensions like that. Ask the author of the particular
extension.

>     I install Python extensions that want access to the source
>     tree on, say, a standard RedHat system?  Having the standard
>     RPM-based installation in parallel with a /usr/local installation
>     from source is confusing, and removing the RPM-based installation
>     risks breaking things.

For the compiling modules: ./configure --prefix=/usr



-- 
Bill Anderson 			Linux/Unix Administrator, Security Analyst
ESBU (ARC)			bill_anderson at boi.hp.com
My opinions are just that; _my_ opinions.


From shapr at uab.edu  Fri Apr 28 09:15:28 2000
From: shapr at uab.edu (Shae Erisson)
Date: Fri, 28 Apr 2000 13:15:28 GMT
Subject: soaplib???
References: <8ebud1$jk4$1@nnrp1.deja.com>
Message-ID: <39099193.C8A6068D@uab.edu>

tiddlerdeja at my-deja.com wrote:
> 
> Is anyone working on a soaplib?

yes:

http://casbah.org/Scarab/

The actual (incomplete?) code can be seen/acquired from their cvsweb:

http://casbah.org/cgi-bin/cvsweb.cgi/Scarab/python/
-- 
sHae mAtijs eRisson (sHae at wEbwitchEs.coM) gEnius fOr hIre
Heard about the Indian Tailor? He's named Dances With Gloves.


From quinn at zloty.ugcs.caltech.edu  Wed Apr 12 16:19:34 2000
From: quinn at zloty.ugcs.caltech.edu (Quinn Dunkan)
Date: 12 Apr 2000 20:19:34 GMT
Subject: modifying os.environ
Message-ID: <slrn8f9mn1.ven.quinn@zloty.ugcs.caltech.edu>

Is there any way to delete things from the environment in python?  A look at
os.py reveals that _Environ only overrides __setitem__ to putenv.  So how do
you *delete* environment values?  perl does it, and a look at the source
seems to indicate that it doodles with the C environ directly.  So perhaps
python needs a delenv _Environ can use or something, or implement os.environ
in C?

Or perhaps there's a better way to do the (common) thing I'm trying to do:
clean the environment to insure some minimum sane values.  I can't make a copy
of the environ and then pass to execve() since that doesn't help much for
popen and system.  I can set values to '', but that doesn't seem as safe as
just removing them entirely (IFS, LD_LIBRARY_PATH, etc.?)


From embed at geocities.com  Wed Apr  5 12:56:11 2000
From: embed at geocities.com (Warren Postma)
Date: Wed, 5 Apr 2000 11:56:11 -0500
Subject: Teaching the art of programming, in python
References: <20000405053920.20579.qmail@web2102.mail.yahoo.com> <8cfe42$o02$1@nnrp1.deja.com>
Message-ID: <dpJG4.3744$HG1.106895@nnrp1.uunet.ca>

> Here's my theory, which is my own: It all depends on syntax. Perl's is
> so nasty that its newbies *need* the books to make any use of it.
> Python's is so clean that the occasional online nudge is sufficient.

As per Occam's Razor, I offer you a Simpler Explanation: Perl has merely
been around longer.


More books doesn't mean anything. One LISP-lover wrote something to this
effect in a magazine editorial a few years back. I can only broadly
paraphrase what he said:

<<     C++ is too complicated. It's badly oversold and poorly designed, and
it
    requires reams and reams of books to explain to someone how to create
something
    useful with it, without falling into the thousands of pitfalls that come
up along
    the way.  People then see the reams of C++ documentation and say "Gee,
that's a
    really popular and well documented language" and so they decide to use
it.
>>

I disagree with *some* the above bracketed text, primarily, because it
Smacks of Bitterness, but I think that most people admit that a language
requires different kinds of documentation. I think the provided Python
documents are useful, practical, and minimal.  Exactly the sort that should
come with Python.

What I would like to find is an introduction to the Art of Programming, in
Python.

No language, Python, Smalltalk, etc, is so easy to understand that it
requires no documentation.    You see, Python will rise, and one day fall,
but the skill of programming, the ideas that went into the design of Python
are themselves timeless. These are the things which must be placed into a
Sacred Codex and never forgotten!

At least Perl programmers seem to admit that Perl grew from a rich and
quirky set of Unix command line tools.  If you know bash, awk, grep, etc,
then you can see where Perl is
coming from. In Larry Wall's own words, it's 'Postmodern'. I would say
instead, "chaotic and byzantine".  No apologies are offered.

Perhaps I will attempt to write my own version of the Art of Programming,
Python Style.

You see, I have this idea of 'elegance' in my head which Python satisfies,
but I haven't found a book that really captures that. I have read a few (4),
and none really do justice to the Art side.

Remember the famous 'Smalltalk-80 - The Language and Implementation'
purple-book?

The Python texts I've read are good References to Getting Stuff Done, but
they Don't Teach the Art. Speaking of Dummies, was there/is there a LISP for
Dummies, or Smalltalk for Dummies?  Can Python for Dummies be far off?
<sigh.>

Warren




From mvanier at endor.bbb.caltech.edu  Thu Apr  6 17:38:24 2000
From: mvanier at endor.bbb.caltech.edu (Michael Vanier)
Date: 06 Apr 2000 14:38:24 -0700
Subject: favorable media attention :-)
Message-ID: <jlitxusyen.fsf@endor.bbb.caltech.edu>

Hey everyone, check out the April 6th Doctor Fun cartoon, currently at:

http://metalab.unc.edu/Dave/Dr-Fun/latest.jpg

You'll be glad you did ;-)

BTW IMO Doctor Fun is the funniest cartoon around these days; it reminds me
of the Far Side at its peak.

Mike


--------------------------------------------------------------
Mike Vanier	mvanier at bbb.caltech.edu
Department of Computation and Neural Systems, Caltech 216-76
GNU/Linux: We can't lose; we're on a mission from God.


From scarblac-spamtrap at pino.selwerd.nl  Thu Apr 27 17:28:42 2000
From: scarblac-spamtrap at pino.selwerd.nl (Remco Gerlich)
Date: 27 Apr 2000 21:28:42 GMT
Subject: Do I always have to write "self." ?
References: <270420001706386648%pecora@anvil.nrl.navy.mil>
Message-ID: <slrn8ghc6t.f01.scarblac-spamtrap@flits104-37.flits.rug.nl>

Louis M. Pecora wrote in comp.lang.python:
> I am two weeks plus into learning Python and have started on a first
> real project.  Amazing language.  BUT, now that I've gotten going with
> classes, I'm finding I hate seeing all those  "self."  prefixes on the
> class variables.  Python is generally very clean, but the classes I
> write use a lot of "per-instance" variables, hence seem to require
> self.  As a result the expressions can become cluttered and harder to
> read than C++ in this particular case -- imagine that!  Is there any
> relief here?  Any (simple) way around writing that self. every time?

No. However, this seems to make code *more* readable. In fact, many C++
programmers use 'this.x' for all their instance variables, to be able to
see at a glance which are just local variables of the function and which are
instance variables. It's just more work to write, but that doesn't matter
much.

Python needs it to know which vars are instance variables, and which are
local variables.

Funny how people say that every aspect of Python they don't like make the
language less "clean". It is extremely consistent though.

You *can* use another word for 'self' though, 's' or 'me' or something. But
it makes your code harder to read, of course...


-- 
Remco Gerlich,  scarblac at pino.selwerd.nl
Hi! I'm a .sig virus! Join the fun and copy me into yours!


From dalke at acm.org  Fri Apr  7 04:55:12 2000
From: dalke at acm.org (Andrew Dalke)
Date: Fri, 7 Apr 2000 02:55:12 -0600
Subject: "tostring" as a string method?
References: <8c45ue$b7u$1@slb1.atl.mindspring.net> <Lp9H4.3526$y4.52453@newsread1.prod.itd.earthlink.net>
Message-ID: <8ck7kq$1cu$1@slb3.atl.mindspring.net>

Warren Postma said:
> Why not write str(dna) and overload __str__ where necessary?

> class myClass:
>  def __init__(self,value='something'):
>    self.value ='something'
>  def __str__(self):
>    return self.value
>
> myObject = myClass()

What I am looking for is a way to have the "value" be a string
or an array.array or even a Numeric.array.  The solution you have
only works for strings.  Adding a "tostring()" method to strings
would let me have a

  def __str__(self): return self.value.tostring()

/F pointed out the buffer interface, which is the route I'm
going to look at - I'm only using strings now because it turns
out building up a [] and string.join'ing them is faster than
using an array.array.  (Though I have ways to make arrays be
faster.)

Meanwhile, on the biopthon.org list, Jeff Chang pointed out that
__str__ is in general an ill-defined quantity.  The usual
interpretation is that __str__ returns a human-representation while
__repr__ returns a computer representation.  So using __str__
to return the character stream is not the right interface.  We're
now using "tostring()", which is more obvious.

                    Andrew
                    dalke at acm.org





From paul.magwene at yale.edu  Wed Apr 12 17:57:54 2000
From: paul.magwene at yale.edu (Paul Magwene)
Date: Wed, 12 Apr 2000 17:57:54 -0400
Subject: scrolledListBox problem
References: <38F4CE3C.3D6E7A68@mitre.org>
Message-ID: <38F4F162.B2B033FC@yale.edu>

Pete wrote:
> 
> I am using makeing a GUI that contains 2 scrolledListBoxes,  and I have
> come across something unfortunate.  When I select an item in one list,
> the currently selected item in the other list becomes deselected,  so
> only one list can have an item selected at any point in time.  Is this a
> bug in PMW?  Any ideas as to how to fix this?

This is a Tkinter thing.  By default the selection in a Listbox widget
is also the X selection, and selecting something outside the listbox
deselects it.  You need to set the exportselection option to zero to get
the behavior you want.

For example,

----------------------
Pure Tkinter solution:
----------------------

from Tkinter import *
root = Tk()

sl1 = Listbox(root,exportselection=0)
sl1.insert(END,'bob')
sl1.insert(END,'fred')

sl2 = Listbox(root,exportselection=0)
sl2.insert(END,'mary')
sl2.insert(END,'jane')

sl1.pack()
sl2.pack()

root.mainloop()




-----------------------
Pmw solution
-----------------------

from Tkinter import *
import Pmw

root = Pmw.initialise()

sl1 = Pmw.ScrolledListBox(root,items=('bob','fred','mary'),
                          listbox_exportselection=0)
sl2 = Pmw.ScrolledListBox(root,items=('jim','john','bob'),
                          listbox_exportselection=0)

sl1.pack()
sl2.pack()

root.mainloop()


-- 

Paul Magwene
paul.magwene at yale.edu


From mwh21 at cam.ac.uk  Wed Apr 12 05:30:57 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 12 Apr 2000 10:30:57 +0100
Subject: Classes, Inheritance - Stupid lazy question
References: <8d1cfh$hme7@imsp212.netvigator.com> <874s9766r6.fsf@ifm.uni-kiel.de>
Message-ID: <m3itxnis32.fsf@atrus.jesus.cam.ac.uk>

Janko Hauser <jhauser at ifm.uni-kiel.de> writes:

> I think this will do it
> 
> class lazy:
>     def __init__(self, name):
>         self.name = name
> 
>     def fired(self):
>         print 'You are fired %s' % self.name
> 
> class paul(lazy):
>     def __init__(self, name):
>         self.name = name  # <---- ?????
>         self.april_pay = 0
> 
>     def fired(self):
>         print "Here's a box, collect the things from your desk"
>         self.__class__.__bases__[0].fired(self) 

Nononono; think about what happens if you do this:

class bob(paul):
   pass

bob("adam").fired()

What *I* think you want is:

class paul(lazy):
    def __init__(self, name):
        lazy.__init__(self,name)
        self.april_pay = 0

    def fired(self):
        print "Here's a box, collect the things from your desk"
        lazy.fired(self)

After all, if you're going to inherit from a class, you must have
access to it, right?

another option is:

class paul(lazy):
    super = lazy
    def __init__(self, name):
        self.super.__init__(self,name)
        self.april_pay = 0

    def fired(self):
        print "Here's a box, collect the things from your desk"
        self.super.fired(self)

Though this has problems with repeated & multiple inheritance.

Cheers,
M.

-- 
  "declare"?  my bogometer indicates that you're really programming
  in some other language and trying  to force Common Lisp into your
  mindset.  this won't work.           -- Erik Naggum, comp.lang.lisp


From embed at geocities.com  Wed Apr 19 10:25:05 2000
From: embed at geocities.com (Warren Postma)
Date: Wed, 19 Apr 2000 10:25:05 -0400
Subject: __del__ problem - more details provided. still stumped though.
References: <aw2L4.8907$HG1.245743@nnrp1.uunet.ca>
Message-ID: <7ojL4.9424$HG1.254543@nnrp1.uunet.ca>

Okay, more details, since i've obviously oversimplified the previous attempt
at explaining the problem.

I have four modules here, all used in an embedded Python application on
Windows (using static linking, not the python15.dll, because this is a
deeply embedded system with no DLLs on the target CPU).

ztrdb.py - this is my "application specifics" module, containing a basic
"data definition" of the
            tables used in my database layer.

lwdb.py - lightweight database class (lwdb_base) and some inherited classes
that implement various
            data storage techniques, from a simple "pickle rows to a text
file" to various
            techniques that rely on the BSDDB 2.0's hash and btree file
storage.


Basically, I wanted to separate the Instances of Tables that were
application specific, from my
generic and hopefully reusable data-handling-classes.  I also wanted lwdb.py
to track ALL instances of tables created from ANY class defined in lwdb.py.
In other words, I wanted to be able to do operations on all tables (such as
closing all tables at once, or flushing changes to disk
on all tables at once).

So, in lwdb.py:

lwdb_tables = {}

class lwdb_base:
    def __init__(self,tablename,description, row_class_obj, filename)
         global lwdb_tables
         # rest of constructor snipped because it's not important here
         lwdb_tables[tablename] = self


Next, for reasons which are still not clear even to me, I built ztrdb.py,
which was my own instances. It did things like this:

table1 = lwdb.lwdb_base("tablename","a description", table1rowclassobj,
"tablename.txt")

Now because the Object reference was stored to both lwdb_tables (a strong
reference) and table1 also a strong reference) and there is no easy way to
make a weak-reference, there were now TWO reference counts to each table
object.

The problem I was having was that at shutdown I was getting this exception,
emitted somewhere from the bowels of the BSD Database layer 2.0 module
(written by Robin Dunn, and as far as I know it has been reliable, yet it is
complex enough that if it *was* having a problem, I wouldn't have the
foggiest idea how to fix it.). Here is the text of the exception, which
occurs when the module lwdb is being destroyed, and the lwdb_tables
dictionary, and the table objects it contains  are finally going out of
scope:


Exception except.AttributeError:
"'None' object has no attribute 'time'" in
<method lwdb_base.__del__ of lwdb_hash instance at 5de410>
ignored

Exception except.AttributeError:
"'None' object has no attribute 'delete_DB'" in
 <method DBPtr.__del__ of DB instance at 5dfd00>
ignored


Exception except.AttributeError:
"'None' object has no attribute 'time'" in
<method lwdb_base.__del__ of lwdb_base instance at 5de6d0>
ignored


In this case I have figured out the following:

1. this message indicates that errors during __del__ are masked, so that the
destructor can finish cleaning up.
2. there is one lwdb_base object being destroyed and two lwdb_hash objects
(a subclass of lwdb_base that uses a BSDDB hash file for storage).

I have figured out a workaround, but can't figure out why it works:

1. Don't keep the reference count of the table objects at 2, so you may only
have ONE reference to it. Here's how I did it:

# in ztrdb.py

# the lwdb.lwdb_tables dictionary now has a refcount of 2
#but the contents are still at refcount of 1
tables = lwdb.lwdb_tables

# at the beginning of shutting the system down, before anything
# else shuts down, we delete the table objects:
def call_me_at_shutdown():
    del lwdb.lwdb_tables
    tables=None


Also, in the lwdb_base.__del__ method I inserted a print statement so I
could see when __del__ got called. If All table Objects __del__ methods are
called before my C program reaches the Py_Finalize() call in my C program,
then everything is fine. If however the modules are left to clean up things
in their own preferred orders inside Py_Finalize, I get those errors above.

I found where those errors are originating from, too, but I'm still no
closer to understanding this.  The file db.py, which is included with Robin
Dunn's BSDDB 2.0 package for python, includes this definition:


class DBPtr :
    def __init__(self,this):
        self.this = this
        self.thisown = 0
    def __del__(self):
        if self.thisown == 1 :
            dbc.delete_DB(self.this)  # appears some exceptions being raised
here.
    # rest of class DBPtr snipped


That still leaves the two messages regarding unknown attribute named "time".
the only reference to time in my modules is calls to time.time() in a
function that could be called in the destructor of the table objects, while
updating the table headers one last time before closing the file. Is it
possible that the stuff added to our namespace by the "import time" function
has already been deleted before the objects in the module are deleted?
Shouldn't all instances of objects be deleted before the global namespace of
a module gets trashed? What gives?

Any help sorting these mysteries out would be greatly appreciated.

The code itself is about 10k, perhaps too large to attach here, but
certainly small enough to send to anyone. If it's possibly a fault with the
bsddb 2.0 I'd like to help find it and fix it if I can, and give Robin Dunn
a patch. Unfortunately, I'm stumped!

Warren




From dalke at acm.org  Fri Apr  7 09:32:54 2000
From: dalke at acm.org (Andrew Dalke)
Date: Fri, 7 Apr 2000 07:32:54 -0600
Subject: [Python-Dev] Round Bug in Python 1.6?
References: <38ED0016.E1C4A26C@tismer.com> <200004062130.RAA26273@eric.cnri.reston.va.us> <20000407084914.A13606@mems-exchange.org>
Message-ID: <8cknsa$1r4$1@slb6.atl.mindspring.net>

Greg Ward exclaimed:
>Oh, joy! oh happiness! someday soon, I may be able to type
>"blah.__doc__" at the interactive prompt and get a readable result!


I agree.  I write __str__ to be human readable, while __repr__
is more mean to be used in evals.  I have code which looks like:

    def __repr__(self):
        return "%s(%s, %s)" % (self.__class__.__name__,
                               repr(self.data),
                               repr(self.alphabet))
    def __str__(self):
        if len(self.data) > 60:
            s = repr(self.data[:60] + " ...")
        else:
            s = repr(self.data)
        return "%s(%s, %s)" % (self.__class__.__name__, s,
                               repr(self.alphabet))

The size of self.data could be megabytes, and I don't want that
scrolling line after line up the screen when in interactive mode.
I would much rather have the __str__ version which only shows
the first 60 characters.

                    Andrew
                    dalke at acm.org







From kcazabon at home.com  Sat Apr 15 01:48:31 2000
From: kcazabon at home.com (Kevin Cazabon)
Date: Sat, 15 Apr 2000 05:48:31 GMT
Subject: problem with os.system calls under win32
References: <1256392132-57146878@hypernet.com> <8d7o4d$c8h1@imsp212.netvigator.com>
Message-ID: <PoTJ4.137194$Dv1.1488699@news1.rdc1.bc.home.com>

If you want it to be even more bullet-proof, use the Win32 call for
GetShortPathName, and don't worry about it.

>>> import win32api
>>> print win32api.GetShortPathName('c:\\f o o\\bar.exe')
c:\FOO~1\BAR.EXE

q:]


Kevin.

"Gresh" <gresham at mediavisual.com> wrote in message
news:8d7o4d$c8h1 at imsp212.netvigator.com...
> Actually I was trying a similar thing, and decided to be smart and use
> os.path.join
> but this is what happened:
>
> >>> import os
> >>> os.system(os.path.join("D:","My Documents","test.bat"))
> 'D:\My' is not recognized as an internal or external command,
> operable program or batch file.
> 1
>
> However:
> >>> os.path.exists(os.path.join("D:","My Documents","test.bat"))
> 1
>
> That's a bit confusing !
>
> Of course from Gordon's example:
> >>> os.system('"%s"' % (os.path.join("D:","My Documents","test.bat")))
> 0
>
> works
>
> Regards,
> Paul
>
>
> "Gordon McMillan" <gmcm at hypernet.com> wrote in message
> news:1256392132-57146878 at hypernet.com...
> > william_93402 at my-deja.com wrote:
> >
> > > I am attempting to do a system call in a directory with spaces in the
> > > name. For example:
> > >
> > > >>>import os
> > > >>>os.system('c:\\f o o\\bar.exe')
> > > 1
> > >
> > > when I place the executable in a dir such as c:\\foo I can
successfully
> > > execute the system call. Is there something special I have to do when
> > > directory names contain spaces? Just for notes I am running by example
> > > in build 125 of pythonwin.
> >
> > Follow MS's rules:
> > os.system('"c:\\f o o\\bar.exe"')
> >
> > (must be double-quotes).
> >
> > - Gordon
> >
>
>




From tseaver at starbase.neosoft.com  Wed Apr 26 23:08:15 2000
From: tseaver at starbase.neosoft.com (Tres Seaver)
Date: 26 Apr 2000 22:08:15 -0500
Subject: Need help with COMObject for Zope
References: <8e7jeo$s8c$1@nnrp1.deja.com> <T8MN4.5091$v85.31034@news-server.bigpond.net.au>
Message-ID: <0EA44467D272A9D6.326DFBDD03E0FA2F.F1D433246D3DFD31@lp.airnews.net>

In article <T8MN4.5091$v85.31034 at news-server.bigpond.net.au>,
Mark Hammond <mhammond at skippinet.com.au> wrote:
><deepak_marda at my-deja.com> wrote in message
>news:8e7jeo$s8c$1 at nnrp1.deja.com...
>
>> I created a COMObject with id = word and Dispatch
>> = "Word.Application". WHen I try to connect this
>> COM object, it raises the following exception
>> Error Type: com_error
>> Error Value: (-2147221008, 'CoInitialize has not
>> been called.', None, None)
>
>Add the 2 lines:
>import pythoncom
>pythoncom.CoInitialize()
>
>before your code, and it will almost certainly work.
>
>The Zope guys should document this fairly well.  Note that Zope is almost
>certainly initializing COM on a different thread than your code is being
>called on (hence you see this message) - but there are some serious
>implications with COM and threading.  I dont believe you will hit them
>here, but someone will sometime :-)

We have a bug in the Collector on the COMObject project for this issue:

  http://classic.zope.org:8080/Collector/842/view

The bug page has three distinct workarounds:  one is essentially what
Mark posted (it has some ugly issues, but will work, slowly, for most code);
one requires hooking the thread creation to perform the CoInitialize() then;
the third makes use of a "pseudo-thread-local" table to track whether
CoInitialize() has been called for the current thread yet.  I favor the
third, myself, as it can be done without modifying Zope to support an
unsupported product.

Take-the-COMObject-product....PLEASE!'ly

Tres.
-- 
---------------------------------------------------------------
Tres Seaver        tseaver at digicool.com     http://www.zope.org
Digital Creations


From infonuovo at email.com  Wed Apr  5 00:17:06 2000
From: infonuovo at email.com (Dennis E. Hamilton)
Date: Tue, 4 Apr 2000 21:17:06 -0700
Subject: APL Characters in Unicode
In-Reply-To: <oqitxxmsdd.fsf@titan.progiciels-bpi.ca>
Message-ID: <NDBBKEGCNONMNKGDINPFMELFDFAA.infonuovo@email.com>

Character codes for the APL characters are in Unicode Standard Version 3.0.
They are on the U+2300 - U+23FF code page.  Their names are such things as
APL FUNCTIONAL SYMBOL IOTA, APL FUNCTIONAL SYMBOL ZILDE, etc.  They don't
have all of APL glyphs as unique Unicode APL characters.  There is APL
FUNCTIONAL SYMBOL EPSILON UNDERBAR, but it is apparently not necessary to
have APL FUNCTIONAL SYMBOL EPSILON separate from some other code such as
SMALL ELEMENT OF, which is closer to the APL epsilon than the GREEK SMALL
LETTER EPSILON in terms of the glyphs used in the book. (Messy, messy.)

I found this in a Borders Books & Music all by itself in the midst of the
computer books.

  The Unicode Consortium.  "The Unicode Standard: Version 3.0"
  Addison-Wesley Longman (Reading, MA: 1991-2000).  ISBN 0-201-61633-5.

There are over 49,000 assigned character codes.  The assigned characters are
assigned to 16-bit codes.   There are extensions devices for using paired
code values in UTF-16, but no assignments have been made for any of those.
I always thought quine corners were Quine corners, but I guess not.  (This
is like getting lost in a dictionary.  What does quincuncial mean anyhow?)

-- Dennis

-----Original Message-----
From: python-list-admin at python.org
[mailto:python-list-admin at python.org]On Behalf Of Fran?ois Pinard
Sent: Tuesday, April 04, 2000 15:09
To: Samuel A. Falvo II
Cc: python-list at python.org; python-list at python.org
Subject: Re: blessed be Guido, for "5 <= X <= 10" does what it should...


kc5tja at garnet.armored.net (Samuel A. Falvo II) writes:

> Astonishingly enough, it is.  Or so I've been told.  Not sure where or how
> to verify this, though, as I've never used APL.  It's part of their 32-bit
> character set.

Unicode is more waving around 16 bits, aiming 20 maybe, and having 31 as
ultimate maximum.  They never had 32 bits in their plan, so far that I know.

--
Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard



--
http://www.python.org/mailman/listinfo/python-list




From ivanlan at callware.com  Wed Apr 19 18:26:51 2000
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Wed, 19 Apr 2000 16:26:51 -0600
Subject: Browse dialog in PythonWin
Message-ID: <38FE32AB.520C1736@callware.com>

Hi All--
OK, I've been able to get my dialogs up and even install handlers for
button clicks, but now I need to start up a browse dialog that lets
users  choose directories.

And is there any news on Mark Hammond's Starship page?

<not-that-i'm-frustrated-but-my-keyboard-has-teeth-marks-now>-ly y'rs,
Ivan
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
http://www.pauahtun.org 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
Author:  Teach Yourself Python in 24 Hours



From dworkin at ccs.neu.edu  Tue Apr 25 19:47:37 2000
From: dworkin at ccs.neu.edu (Justin Sheehy)
Date: 25 Apr 2000 19:47:37 -0400
Subject: The Simple Economics of Open Source
In-Reply-To: marc@oscar.noc.cv.net's message of "Tue, 25 Apr 2000 14:32:02 GMT"
References: <B527E7A5.51AA%raffael@mediaone.net> <MPG.136cc01af4ab703b989709@news.earthlink.net> <raffael-4899DC.23562823042000@news.ne.mediaone.net> <vndya62ykbi.fsf@camelot-new.ccs.neu.edu> <slrn8gbaj2.43p.marc@oscar.noc.cv.net>
Message-ID: <vndk8hlyc86.fsf@camelot-new.ccs.neu.edu>

marc at oscar.noc.cv.net (Marc Spitzer) writes:

> this sounds like a religious argument, be carful when you 
> show evadence that contradicts the 1 true faith. 

Hm?  I was merely pointing out an obvious factual error.

-Justin

 



From PClaerhout at CREO.BE  Wed Apr 26 10:11:21 2000
From: PClaerhout at CREO.BE (Pieter Claerhout)
Date: Wed, 26 Apr 2000 16:11:21 +0200
Subject: VIM NT +python
Message-ID: <2B1262E83448D211AE4B00A0C9D61B03011152BA@MSGEURO1>

If you download VIM 5.6, unzipped it and ran install.exe, you can
choose to turn on syntax highlighting. Once you did this, the _vimrc
file is created with (python) syntax highlighting turned on.

Pieter

-----Original Message-----
From: Mike Steed [mailto:MSteed at altiris.com]
Sent: Tuesday, April 25, 2000 11:01 AM
To: python-list at python.org
Subject: RE: VIM NT +python


Hi,

   I just built the latest version (5.6.69) with the default feature set,
plus the Python interface.  I have not tested it beyond

   :python print 'hi'

which passed with flying colors.

http://athani.pair.com/msteed/vim/gvim56py.zip

M.

> From: Me [mailto:me at nospam.edu]
> Sent: Tuesday, April 25, 2000 11:27 AM
> To: python-list at python.org
> Subject: Re: VIM NT +python
> 
> 
> So would I!
> 
> <tiddlerdeja at my-deja.com> wrote in message
> news:8e4j27$hgq$1 at nnrp1.deja.com...
> > Does anyone here have a winNT build of the VIM editor that 
> has python
> > support switched on?
> >
> > I'd love to get my hands on a copy.
> >
> >
> > Sent via Deja.com http://www.deja.com/
> > Before you buy.

-- 
http://www.python.org/mailman/listinfo/python-list



From markus.oberhumer at jk.uni-linz.ac.at  Tue Apr 18 14:18:44 2000
From: markus.oberhumer at jk.uni-linz.ac.at (Markus F.X.J. Oberhumer)
Date: Tue, 18 Apr 2000 20:18:44 +0200
Subject: [ANNOUNCE] PySol 4.10 - a solitaire game collection
Message-ID: <20000418201844.A9959@laetitia.oberhumer.com>

-----BEGIN PGP SIGNED MESSAGE-----


                    PySol - a Solitaire Game Collection
                              Version 4.10

                          http://pysol.tsx.org

           Copyright (C) 1998, 1999, 2000 Markus F.X.J. Oberhumer
                   <markus.oberhumer at jk.uni-linz.ac.at>


What is PySol ?
===============

PySol is an exciting collection of 273 solitaire games.

Among the supported games are classics like Aces Up, Baker's Game,
Canfield, FreeCell, Forty Thieves, Golf, Klondike, Monte Carlo,
Osmosis, Pyramid, Scorpion, Spider, Yukon, and many more...

PySol may be freely distributed under the terms of the GNU GPL.

PySol aims to be Commercial Quality Freeware.


Why yet another solitaire game ?
================================

Here are some highlights of PySol:

  - freely available
  - currently supports 273 (!) distinct solitaire variants
  - lots of classic games like Forty Thieves, FreeCell, Klondike and Spider
  - special games like Hanafuda, Mahjongg, Poker and Tarot type games
  - based upon an extensible solitaire engine
  - very nice look and feel including multiple cardsets
    and background table tiles
  - sound samples and background music
  - unlimited undo & redo
  - load & save games
  - player statistics and log files
  - hint system and demo games
  - support for user written plug-ins - add your own solitaire variants
  - integrated HTML help browser
  - lots of documentation
  - fully portable across Unix/X11, Windows 95/98/2000/NT and MacOS
  - distributed under the GNU GPL with full source code
  - written in 100% pure Python
  - just run it - no need to compile anything


Nice. And what's new ?
======================

  * completely new Mahjongg game driver

  * fixed a bad memory leak (when will there finally be a real
    garbage collector in Python ???)


Cool. Where can I get it ?
==========================

Point your browser to http://pysol.tsx.org
The PySol Gallery is awaiting your visit as well.


What do I need to start playing ?
=================================

PySol requires Python 1.5.2 and Tcl/Tk 8.0.5. Both packages are
freely available for Unix, Windows 95/98/NT and Macintosh.

BTW, there is no need to compile anything since the whole program is just
a Python script. Just run it, and that's all.

The Windows version ships as a completely self-contained setup file.


License terms
=============

PySol is Copyright (C) 1998, 1999, 2000 Markus Franz Xaver Johannes Oberhumer

PySol is distributed under the terms of the GNU General Public License (GPL).
See the file COPYING.


Have fun,
Markus

http://pysol.tsx.org



-----BEGIN PGP SIGNATURE-----
Version: 2.6.3ia
Charset: noconv

iQCVAwUBOPyg7W10fyLu8beJAQHN9wP/XpBKnGuQzQO9zxYP1zkeCb6Nwxj8lf51
1UnpB8yNgYzstc0j45s0Ay8GrTVnGqwQvdale4cMfbMqTK7IWu1kV5KgBdeT5rne
TF6yCAEx7Yl9YEL3HctUR3floUyOPJXsPoEgKrVsp+AFUVwtypvtXIWwSGgDDgsb
w4OX2mXQPiw=
=IDJ9
-----END PGP SIGNATURE-----



From maybe at lucky.com  Wed Apr 12 09:35:21 2000
From: maybe at lucky.com (Soundwave)
Date: Wed, 12 Apr 2000 15:35:21 +0200
Subject: making a .exe
Message-ID: <8d1u0u$ikb$1@porthos.nl.uu.net>

How can I make an executable file from the *.py files ?

thanx




From rupole at compaq.net  Sat Apr 15 10:29:13 2000
From: rupole at compaq.net (Roger Upole)
Date: Sat, 15 Apr 2000 14:29:13 GMT
Subject: Change NT password with Python
References: <38F69231.EE1D9B6D@hotmail.com>
Message-ID: <Z0%J4.123290$ra.2854851@news.easynews.com>

You can use cgi.FieldStorage to get the old passwd, new passwd and passwd
confirmation off the web page, and use win32net.NetUserChangePassword to
perform the required action.
         Roger Upole

"Nguyen Anh Quynh" <quynhna at hotmail.com> wrote in message
news:38F69231.EE1D9B6D at hotmail.com...
> Dear my friends,
> I am looking for the way to write on Python stuff helping me to change
> NT password, since I would like to change the password of WinNT domain
> via Web page. (That Python stuff will be the CGI)
> Could you point me the way?
> Thank you very much,
>
>




From ivanlan at callware.com  Wed Apr  5 12:24:03 2000
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Wed, 05 Apr 2000 10:24:03 -0600
Subject: Teaching the art of programming, in python
References: <20000405053920.20579.qmail@web2102.mail.yahoo.com> <8cfe42$o02$1@nnrp1.deja.com> <dpJG4.3744$HG1.106895@nnrp1.uunet.ca>
Message-ID: <38EB68A3.B1B5809F@callware.com>

Hola, Listeros--

Warren Postma wrote:
> 

[snip]

> What I would like to find is an introduction to the Art of Programming, in
> Python.
> 

[snip]*2

> The Python texts I've read are good References to Getting Stuff Done, but
> they Don't Teach the Art. Speaking of Dummies, was there/is there a LISP for
> Dummies, or Smalltalk for Dummies?  Can Python for Dummies be far off?
> <sigh.>
> 

Please forgive the BSP (Blatant Self-Promotion) here.

On April 17, according to Macmillan, my book _Teach Yourself Python in
24 Hours_ will finally be out.  Its mission is to provide exactly what
you're looking for, a book that teaches programming, art, attitude and
philosophy, using Python.

I don't know how well I succeed.  That's left to the critics to
determine.  The book should really have been called _Python Mind,
Beginner's Mind_, because it draws heavily on Shunryu Suzuki's _Zen
Mind, Beginner's Mind_.  It is not a _Python for Dummies_ book, however,
because I don't believe that people who want to learn how to program
should either be called or treated like dummies.  I try to convey how
much fun programming is, and try to show some of the really neat things
you can do with Python.

There are even Mayan glyphs in the book.  You can visit the web page
(which includes the "advance praise" page;-) at

	http://www.pauahtun.org/TYPython/

<that's-a-feature,-gordon>-ly y'rs,
Ivan;-)
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
http://www.pauahtun.org 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
Author:  Teach Yourself Python in 24 Hours



From aahz at netcom.com  Sun Apr 16 13:15:32 2000
From: aahz at netcom.com (Aahz Maruch)
Date: 16 Apr 2000 17:15:32 GMT
Subject: Event triggering and weak references
References: <8dbif3$hm0$1@metro.ucc.usyd.edu.au> <MPG.1363441c166d9ca98969e@news-server> <8dcmv0$7ub$1@metro.ucc.usyd.edu.au>
Message-ID: <8dcsfk$h$1@nntp9.atl.mindspring.net>

In article <8dcmv0$7ub$1 at metro.ucc.usyd.edu.au>,
Ricardo Nogueira <rnog2438 at mail.usyd.edu.au> wrote:
>
>But both versions work with boundedMethods for the observer. If you change
>your main() to register(a, "x" , objY.msg) it will keep objY from being
>garbageCollected if we clear any other references to it. It is necessary to
>always call unregister before the "last" pointer to objY is cleared. Not
>very nice.
>
>How can we read objY reference count?
>If we come to know this number, compare to the number of boundedMessages we
>have for it, we can conclude if it is garbage and clear its references
>instead of calling its methods.

With the partial exception of the various garbage collection patches,
the Python model is that the programmer is responsible for managing
references.  If there are no references pointing at an object, the
object gets automatically cleaned up.  The Python scope model makes this
fairly easy, as long as you don't have any pure circular references.
--
                      --- Aahz (Copyright 2000 by aahz at netcom.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

"First one back from the bathroom gets to be the bottom."


From effbot at telia.com  Tue Apr 18 17:51:00 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Tue, 18 Apr 2000 21:51:00 GMT
Subject: can someone tell me about...
References: <38FCBB16.BF50EB03@stsci.edu>
Message-ID: <8N4L4.498$wYl.191816704@newsb.telia.net>

Kia Arab <karab at stsci.edu> wrote:
> variable names that start an underscore "_".
>
> Interactively, python will recognize a variable name that begins with
> one underscore (_myVar). The variable naming rules even say that is
> acceptable.
>
> However, when I try to use it in a module, it bombs.

"bombs"?

> What's the deal?

lousy subject line, no code sample, no error message,
vague description of the actual problem.

but alright, I'll guess: when you read this page:

    http://www.python.org/doc/current/ref/import.html

you missed the following sentence:

    If the list of identifiers is replaced by a star ("*"), all
    names defined in the module are bound, except those
    beginning with an underscore ("_").

</F>




From moshez at math.huji.ac.il  Mon Apr  3 18:33:20 2000
From: moshez at math.huji.ac.il (Moshe Zadka)
Date: Tue, 4 Apr 2000 00:33:20 +0200 (IST)
Subject: Creating variables on the fly...
In-Reply-To: <meh9-25D0C5.15160303042000@news.cit.cornell.edu>
Message-ID: <Pine.GSO.4.10.10004040032230.21084-100000@sundial>

On Mon, 3 Apr 2000, Matthew Hirsch wrote:

> But now let's say I have twenty variables that I want to add together.  
> This function will no longer work.  I would have to rewrite it as 
> a+b+c+d+e+f+g+...+(20th letter).  Is there a way to dynamically create a 
> variable name?  So that my add function can automatically determine how 
> many variable names to create and then add the values together.  In 
> other words, I'd ideally like something like:

Try:

def sum(*args):
	res = 0
	for arg in args:
		res = res+arg
	return res

--
Moshe Zadka <mzadka at geocities.com>. 
http://www.oreilly.com/news/prescod_0300.html
http://www.linux.org.il -- we put the penguin in .com




From shae at webwitches.com  Sat Apr  1 10:39:41 2000
From: shae at webwitches.com (Shae Erisson)
Date: Sat, 01 Apr 2000 15:39:41 GMT
Subject: was Re: Why should I switch to Python?
References: <seas5ia89177@corp.supernews.com> <1EeF4.43$CA.2422@news> <oqzord6h5i.fsf@titan.progiciels-bpi.ca>
Message-ID: <38E618D7.A15F9D9B@webwitches.com>

Fran?ois Pinard wrote:
<snip>

> We should not see the potential for a war every time Perl is discussed,
> and be patient instead.  It looks all pretty bearable so far.  And if it
> was becoming too heavy, we could prepare pre-canned replies for the most
> questions, and even suggest a pointer to this FAQ right into the message
> footer generated for this list.  Welcoming people in person is better!
> 
> --
> Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard

maybe a newsgroup bot that finds common text patterns and posts FAQtual
replies?

Seriously though, I joined the python community because perlers treated
me like a worthless newbie, and pythonistas patiently answered my
questions, no matter how clueless they were. I think that attitude is
what makes the python community so wonderful.

I wonder if the Sapir-Whorf hypothesis works within Perl and Python
also?
To me at least, Python seems written to be simple and open, Perl seems
to want to be mystical, magical, and arcane. Their communities seem to
think that way at least, so I may be overly influenced by that.

I wonder if the next step to CP4E would be online 'classes?' Emacs
allows for multiple windows on multiple displays. A friend and I have
used that to try out multi-user editing of a single document across
800km, and it works. Does anyone have other suggestions?
-- 
sHae mAtijs eRisson (sHae at wEbwitchEs.coM) gEnius fOr hIre
   bRing mE fIve sQuirrels aNd nO oNe wIll gEt hUrt


From bhoel at starship.python.net  Sun Apr  2 16:36:44 2000
From: bhoel at starship.python.net (Berthold Höllmann)
Date: 02 Apr 2000 22:36:44 +0200
Subject: GTK Referenc
References: <38E7A868.A1C9BD7C@searchy.net>
Message-ID: <m2snx4utnn.fsf@pchoel.privat.uni-hamburg.de>

Penpal International <ppi at searchy.net> writes:

> I couldn't find anywhere a reference for the GTK module for Python. Does
> anyone know one?
> On gtk's site I could only find some for C(++).

I've send a little script to the PyGTK mailing list (see
http://www.mail-archive.com/pygtk at daa.com.au/msg01225.html). It uses
pythondoc t generate a class and functuon reference. Due to the lack
of doc strings in the whole module, it is of only limited use, but it
helps me getting an idea of the gtk module classes. Besides that, you
have to use the GTK+ documentation.

Greetings

Berthold
-- 
bhoel at starship.python.net / http://starship.python.net/crew/bhoel/
        It is unlawful to use this email address for unsolicited ads
        (USC Title 47 Sec.227). I will assess a US$500 charge for
        reviewing and deleting each unsolicited ad.


From ljoramo at nickads.com  Mon Apr 17 11:55:17 2000
From: ljoramo at nickads.com (Lee Joramo)
Date: Mon, 17 Apr 2000 09:55:17 -0600
Subject: Editors and books
In-Reply-To: <XBFK4.16$Uq2.1831@news.uswest.net>
References: <XBFK4.16$Uq2.1831@news.uswest.net>
Message-ID: <v04210100b520e03fa38e@[192.168.1.2]>

>Next, a question for those who've been doing this: is there a color IDE
>(just a nicer thing that IDLE) for the Mac? I've heard that Alpha (which I
>just downloaded) has a Python mode, but I can't seem to find it. Is there
>something else, either? can MPW do it? BBEdit?

You have several choices:

Alpha: I believe that two python modes have been developed independently.
Here is a link to one of them: <http://www.python.org/emacs/python-mode/>
toward the bottom of the page you should find info about an Alpha mode.
Modes are written in TCL, so you can roll your own.

BBEdit: while I find that Alpha is a very powerful, I tend to use BBEdit
for most of my Python work. BBEdit is powerful enough and has a great user
interface. In BBEdit's language preferences I have created a file extension
of .py for python and set it to use the highlighting for TCL. The resulting
color highlighting very closely matches the needs of python. Note:
BareBones is considering adding direct support for Python in the next
release of BBedit. If you use BBEdit, I suggest you contact BBEdit and add
your vote for Python.

Mac Python IDE: I believe that the next release of MacPython is scheduled
to include syntax highlighting. Maybe this summer.

--
Lee A. Joramo                      ljoramo at nickads.com
The Nickel Want Ads                www.nickads.com
Internet Manager                   970-242-5555




From rumjuggler at home.com  Tue Apr 18 12:48:58 2000
From: rumjuggler at home.com (Ben Wolfson)
Date: Tue, 18 Apr 2000 16:48:58 GMT
Subject: [Announce] autoreload.py
References: <001601bfa8a0$c6d86f50$4500a8c0@thomasnb>
Message-ID: <9p4pfsch59ovo5atgl9l4usb91gl6qs5vs@4ax.com>

On Mon, 17 Apr 2000 21:11:46 +0200, "Thomas Heller"
<thomas.heller at ion-tof.com> wrote:

My newsreader seriously munged this.  Can you repost the code, not as
an attachment?

Thanks

 -- 
Barnabas T. Rumjuggler

Ibi nullus timet mortem
Sed pro Baccho mittunt sortem
 -- Carmina Burana


From matsaleh at my-deja.com  Sat Apr 22 16:50:18 2000
From: matsaleh at my-deja.com (matsaleh at my-deja.com)
Date: Sat, 22 Apr 2000 20:50:18 GMT
Subject: pickle.py and cPickle.c - persistent_id is always called - why?
Message-ID: <8dt3a5$r03$1@nnrp1.deja.com>

Hello fellow .py types...

I am doing a bit of work using pickle/cpickle and
am trying to optimize. I found that my user-
defined persistent_id() function is being called
for (just about) every attribute (name and value)
in my object that is being pickled.

My specific (somewhat ad-hoc) test case pickles
~100 objects in a containment hierarchy, using
persistent_id to break the containment
relationship and replace the references with a
proprietary object id. Although I have only ~100
object references to resolve, persistent_id() is
called ~6400 times. In tracing the code, it
appears to be called for every attribute and
value in my objects.

This appears to be caused by the fact that
pickle.py:save() is called with the default
pers_save flag of 0 in all cases except for when
it is called by save_pers(). This appears to
indicate that all types: tuples, dicts,
sequences, longs, strings, etc, cause my
persistent_id() to be called, when all I want is
for it to be called for object instances.

I modified pickle.py to change the default of the
pers_save flag to 1 in the save() method, and
then call it with a 0 only from within the
save_inst() method. This amounts to invoking my
persistent_id() function only when a reference to
an object instance is encountered as my objects
are being pickled.

My pickled objects do not appear to be adversely
affected by this change, and the number of calls
to persistent_id() was reduced from ~6400 to
~500, reducing the time spent in this method by
an order of magnitude (0.37 sec to 0.03 sec).

I have not yet tested this change in cPickle.c,
but in my original tests, persistent_id() was a
much more significant factor in my runs using
cPickle, because all the pickling code is in C
and is relatively much faster than my
persistent_id() method, which is in Python. I
expect the performance boost my making the change
in cPickle to be even greater, relatively
speaking.

My question is, why is persistent_id() being
invoked so often? I do not see the reason for
calling it for basic python types and structures
such as tuples, dicts, and the like. Is this a
reasonable change to make to the python source,
or is it not a safe change for the general
pickling cases that the pickle/cPickle modules
have to handle? I'm sure this code has been
scruitinized by many folks much more experienced
than I - I would be grateful for your insights.

Regards.



Sent via Deja.com http://www.deja.com/
Before you buy.


From paul at fizzylab.com  Fri Apr  7 15:12:46 2000
From: paul at fizzylab.com (paul beard)
Date: Fri, 07 Apr 2000 12:12:46 -0700
Subject: calendar/scheduling app
Message-ID: <38EE332E.2954EA49@fizzylab.com>

I am interesting in having a meeting scheduler/calendar app that does
what something like Exchange purports to do without requiring me to use
a client that doesn't exist. In other words, it should be
platform-agnostic and use webforms/cookies to handle transactions and
display options. 

Possible in Zope? How hard? How long to do, given a baseline knowledge
store of zero?


--
Paul Beard / 206 728 4282 x138
Fizzylab / 2025 First Ave, Suite 1150
Seattle WA 98121


From brockn at deskmedia.com  Mon Apr 10 23:21:13 2000
From: brockn at deskmedia.com (Brock Noland)
Date: Mon, 10 Apr 2000 22:21:13 -0500
Subject: Jpython??
References: <955394930.758181@newdm.deskmedia.com> <W7rI4.431$rc9.188052480@newsb.telia.net>
Message-ID: <955423157.1114@newdm.deskmedia.com>

Thanks


"Fredrik Lundh" <effbot at telia.com> wrote in message
news:W7rI4.431$rc9.188052480 at newsb.telia.net...
> Brock Noland <brockn at deskmedia.com> wrote:
> > Hello   I am a web developer.. Well part time.  I can 'program' in HTML
> and
> > am interested in learning Java but I was told that I should learn Python
> > first because it is a fairly newbie friendly language.  To the point I
saw
> a
> > post about Jpython yesterday what is it??
>
> www.jpython.org
>
> </F>
>
>




From aa8vb at yahoo.com  Thu Apr 20 20:36:31 2000
From: aa8vb at yahoo.com (Randall Hopper)
Date: Thu, 20 Apr 2000 20:36:31 -0400
Subject: Python Parsers
In-Reply-To: <CC59868A0A76D311B1F50090277B321C1F328F@VRTMAIL>
References: <CC59868A0A76D311B1F50090277B321C1F328F@VRTMAIL>
Message-ID: <20000420203631.A608349@vislab.epa.gov>

Mike Fletcher:
 |Here's part of it :)
 |
 |from simpleparse import generator
 |from TextTools import TextTools

Kudos to you an Mark L on an excellent tool!  I had it parsing and
reformatting my syntax in no time.

I think this would make a great addition to the core distribution,
...particular after having seen the wealth of "regexes won't work for
recursive grammar" posts in the archives.

Also using this package is "easy" (thanks to your and Mark's efforts) and
the syntax grammar is very readable -- more so than re expressions IMO,
even with the verbose option on.

This is a good tool to have up your sleeve. ;-)

Thanks again,

Randall


(For the Dejanews and the list archive, I've attached my short parser.  The
grammer is briefly described in the class docstring, and the output of the
test stub is listed below:

==============================================================================

RAW STRING TO PARSE       PARSED PYTHON
-------------------       -----------------------------------

1                     ::  [1]
 2.3                  ::  [2.2999999999999998]
3.4e20                ::  [3.4e+20]
56 78                 ::  [56, 78]
"string"              ::  ['string']
 identifier           ::  ['@@ identifier @@']
[]                    ::  [()]
[1]                   ::  [(1,)]
 [ 1.0 2.0 ]          ::  [(1.0, 2.0)]
[[1,2,3],[3,4,5 ] ]   ::  [((1, 2, 3), (3, 4, 5))]
{}                    ::  [[]]
 {{ "str" }}          ::  [[['str']]]
{ 1, [2,3], "str" }   ::  [[1, (2, 3), 'str']]
{ ident1, ident2 }    ::  [['@@ ident1 @@', '@@ ident2 @@']]
{[]{}}                ::  [[(), []]]
{ [ 1 2 ][ 4 5 ] }    ::  [[(1, 2), (4, 5)]]

==============================================================================


-- 
Randall Hopper
aa8vb at yahoo.com
-------------- next part --------------
#!/usr/bin/env python
#
#  grammar_parse.py  -  Parse a recursive grammar into basic Python data types.
#

from simpleparse import generator
from TextTools import TextTools
import types, string

class Parser:
  """
    GRAMMAR                 CONVERTED TO PYTHON AS
    ------------------      ----------------------
    { ... } is a list     - List
    ( ... ) is a vector   - Tuple    (vector, matrix, tensor, etc.)
    "..."   is a string   - String   ("string")
    ident   is an ID str  - String   ("@@ ident @@")
    1       is an integer - Integer
    2.3     is a float    - Float
  """

  PARSERDECLARATION = r'''
  root           := ts, element*
  element        := (('"',string,'"')/vector/list/number/identifier), ts
  list           := '{', ts, element*, '}'
  vector         := '[', ts, vect_element*, ']'
  vect_element   := (vector/number), ts
  number         := [-+]*, ( ('0',[xX],[0-9]+) / ([0-9.]+,([eE],[-+0-9.]+)?))
  string         :=  (char/escapedchar)*
  char           :=  -[\134"]+
  escapedchar    :=  '\134"' / '\134\134'
  identifier     :=  [a-zA-Z_@], [a-zA-Z0-9_'@]*
  <ts>           :=  ( [ \011-\015,]+ / ('#',-'\012'*,'\n')+ )*
  '''

  #'"    #<-- Keep python.el mode happy with quoting :-(

  PARSERTABLE = generator.buildParser( PARSERDECLARATION )
  ROOTITEMPARSER = PARSERTABLE.parserbyname( "root" )
  
  def __init__( self, string ):
    self.data = string

  def Parse( self, parseonly = 0 ):
    success, tags, next = TextTools.tag( self.data, Parser.ROOTITEMPARSER )
    if not success:
      raise ValueError, "TextTools.tag failed"
    if next != len( self.data ):
      raise ValueError, "Failed to parse entire expression"
    #import pprint
    #pprint.pprint( tags )
    if parseonly: return tags
    else:         return self.ConvertToPython( tags )

  def ConvertToPython( self, tags ):
    return map( self._dispatch, tags )

  def _dispatch( self, (tag, start, stop, children) ):
    return getattr( self, tag)( (tag, start, stop, children) )

  # tag handlers
  def element( self, (tag, start, stop, (child,) ) ):
    '''Always has a single child: either string, vector, list
       number, or identifier '''
    val = self._dispatch( child )
    return val
  def vect_element( self, (tag, start, stop, (child,)) ):
    '''Always has a single child: either vector or number'''
    return self._dispatch( child )
  def vector( self, (tag, start, stop, elements) ):
    '''Some arbitrary number of elements, return as list'''
    val = map( self._dispatch, elements )
    if type(val) == types.ListType: 
      return tuple(val)
    else:
      return val
  def list( self, (tag, start, stop, elements) ):
    '''Some arbitrary values returned as list'''
    return  map( self._dispatch, elements )
    
  def string( self, (tag, start, stop, children) ):
    '''A quoted string'''
    return self.data[ start:stop ]

  def identifier( self, (tag, start, stop, children) ):
    '''An identifier'''
    return "@@ %s @@" % self.data[ start:stop ]

  def number( self, (tag, start, stop, children) ):
    '''Could be either int or float, check first, then second'''
    try:
      return string.atoi( self.data[ start:stop ], 0 )
    except ValueError:
      return string.atof( self.data[ start:stop ] )

if __name__ == "__main__":

  test_expr = (
    '1', ' 2.3', '3.4e20 ', '56 78', '"string"', ' identifier',
    '[]', '[1]', ' [ 1.0 2.0 ] ', '[[1,2,3],[3,4,5 ] ]',
    '{}', ' {{ "str" }}', '{ 1, [2,3], "str" } ', '{ ident1, ident2 }',
    '{[]{}}', '{ [ 1 2 ][ 4 5 ] }'
    )

  for expr in test_expr:
    #print Parser( expr ).Parse( parseonly = 1 )
    print '%-20s  ::  %s' % ( expr, repr( Parser( expr ).Parse() ) )



From meh9 at cornell.edu  Mon Apr  3 15:52:17 2000
From: meh9 at cornell.edu (Matthew Hirsch)
Date: Mon, 03 Apr 2000 15:52:17 -0400
Subject: Creating variables on the fly...
References: <meh9-25D0C5.15160303042000@news.cit.cornell.edu>
Message-ID: <meh9-C7E402.15521703042000@news.cit.cornell.edu>

I've been trying to figure this out for some time now.  I didn't realize 
I could use *args.  

Thanks again,
Matt




In article <meh9-25D0C5.15160303042000 at news.cit.cornell.edu>, Matthew 
Hirsch <meh9 at cornell.edu> wrote:

> Hi All,
> 
> Let's say I have four variables: 
> 
> a=1
> b=2
> c=3
> d=4
> 
> And I have a function that adds these variables together:
> 
> def add(a,b,c,d):
>    return a+b+c+d
> 
> But now let's say I have twenty variables that I want to add together.  
> This function will no longer work.  I would have to rewrite it as 
> a+b+c+d+e+f+g+...+(20th letter).  Is there a way to dynamically create a 
> variable name?  So that my add function can automatically determine how 
> many variable names to create and then add the values together.  In 
> other words, I'd ideally like something like:
> 
> 
> def add(number_of_variables):
>    return a+b+c+...+(letter corresponding to number_of_variables)
> 
> Thanks for your help,
> 
> Matt


From effbot at telia.com  Mon Apr 10 11:51:17 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Mon, 10 Apr 2000 15:51:17 GMT
Subject: Extending built-in language (Win 95)
References: <213CDEBFDDA7D3118081009027E40C56071118@SHEF-MS1>
Message-ID: <VLmI4.365$rc9.188792320@newsb.telia.net>

Mark Pope <M_Pope at shef.rage.co.uk> wrote:
> I've added xxmodule to the lib and recompiled. Entering:
> a = xx.new()
> just gives a name error on xx.
>
> Have I missed something?

1) the Modules/Setup file is used to add modules
to the standard library.

2) you forgot to *import* the xx module.

</F>




From johngrayson at home.com  Sun Apr  9 18:29:47 2000
From: johngrayson at home.com (John Grayson)
Date: Sun, 09 Apr 2000 22:29:47 GMT
Subject: Two questions
References: <38EA5FE8.853E9B75@callware.com>
Message-ID: <38F1045C.FA3CEDD2@home.com>

Well, you can't use Win32all with Tkinter -- but you
know that already :-)   

I got a good ODBC interface working with Tkinter for
an airline application. I can send you the extension module,
if it will help...

    John

Ivan Van Laningham wrote:
> 
> Hi All--
> I'm finally getting a chance to use Python at work.  I have two
> questions:
> 
> 1)  Is there a way to run silent (i.e., we determine where things go and
> the customer never sees any wizards) installs of Python, Tcl/Tk and
> Win32all?
> 
> 2)  What stuff do I need to use msaccess with Python?  Do I need to use
> Win32all, or can I use Tkinter for my (minimal) GUI, with msaccess?
> 
> <there-that-didn't-hurt-did-it>-ly y'rs,
> Ivan
> ----------------------------------------------
> Ivan Van Laningham
> Callware Technologies, Inc.
> http://www.pauahtun.org
> http://www.foretec.com/python/workshops/1998-11/proceedings.html
> Army Signal Corps:  Cu Chi, Class of '70
> Author:  Teach Yourself Python in 24 Hours


From aa8vb at yahoo.com  Thu Apr  6 07:51:24 2000
From: aa8vb at yahoo.com (Randall Hopper)
Date: Thu, 06 Apr 2000 07:51:24 -0400
Subject: Modifying "traceback" objects
Message-ID: <20000406075124.A1848828@vislab.epa.gov>

     Is it possible to add frames to the current traceback object from
inside a C extension?



     I have Python calling C, which makes callbacks to Python, and I'm
missing the middle C layers in my tracebacks.

Tim Peters summarizes it nicely:

   | If A calls B tails calls C tail calls D blows up, I damn sure don't
   | want a traceback claiming that A called D directly.  I want to get a
   | traceback object with the full chain intact

Thanks,

-- 
Randall Hopper
aa8vb at yahoo.com



From shae at webwitches.com  Wed Apr 26 17:40:16 2000
From: shae at webwitches.com (Shae Erisson)
Date: Wed, 26 Apr 2000 21:40:16 GMT
Subject: global variables or inheritence
References: <28p8zy0irgc.fsf@lada.kom.auc.dk> <eRHN4.1113$3K4.19980@news.uswest.net> <28p4s8oipzh.fsf@lada.kom.auc.dk> <j6IN4.1168$3K4.20204@news.uswest.net> <28pzoqghaoh.fsf@lada.kom.auc.dk>
Message-ID: <390764A7.E986CEF8@webwitches.com>

Thomas Rasmussen wrote:

> Well I still want to be able to do this....
> 
>   >>  testprog2.py: import os, sys, string, testprog1 def second():
>   >> print testprog1.myvar if __name__ == '__main__': print 'nope'
>   >>

> Well i'm programming a small admin tool, which must use the users
> $EDITOR for editing files. This works fine if the EDITOR sysenv is
> set, if not, then i want to ask the user which editor he want to use,
> and then pass this variable to the module that handles
> filemanipulation. This modules is also in another file if it
> matters... I must add that this is my first real program made in
> python...

this is untested code, but it might give you some ideas:

import os
import usefulmodule

def ask():
    usereditor = raw_input("What's your choice of editor?")
    return usereditor

try:
    usereditor = os.environ['EDITOR']
except KeyError:
    usereditor = askforeditor()

usefulmodule.dostuff(usereditor)

-- 
sHae mAtijs eRisson (sHae at wEbwitchEs.coM) gEnius fOr hIre
   bRing mE fIve sQuirrels aNd nO oNe wIll gEt hUrt
13:00pm up 1 season, 1 squirrel, load average: 1+1j 0.5+2j 0.2+0.5j


From ullrich at math.okstate.edu  Tue Apr  4 19:34:39 2000
From: ullrich at math.okstate.edu (David C. Ullrich)
Date: 4 Apr 2000 23:34:39 GMT
Subject: Learning in Stereo:  Math + Python
References: <ua1fes81hmb12dh1klbv0p7d0373qhleps@4ax.com> <ur3fessapua3bi1k4e5g89i3ndbuejhabo@4ax.com> <38EA329F.4D407309@math.okstate.edu>
Message-ID: <01bf9e8f$d054e600$6296cdcf@daves-dell>


David C. Ullrich <ullrich at math.okstate.edu> wrote in article
<38EA329F.4D407309 at math.okstate.edu>...
> Kirby Urner wrote:
> 
> > [...]
> > And here's another fast converger [to pi] using every other
> > Fibonacci number:
> >
> > def pifib(n):
> >    # return approximation for pi using
> >    # arctan(1) = 4 * SIGMA [ arctan(1.0/Fib(2i+1))]

Or arctan(1) = SIGMA [ arctan(1.0/Fib(2i+1))] or something.

> >    sum = 0
> >    for i in range(1,n+1):
> >       sum = sum + math.atan(1.0/fibo(2*i+1))
> >    return 4*sum
> 
>     I didn't know that arctan(1) was equal to that sum, 

	But it is (heh-heh): you show by induction that the 
difference between arctan(1) and the sum of the first n 
terms is exactly arctan(1/fibo(2*n+2)) (or 2*n+1 or 
something, the notation's different in my notes...)

	(This is on-topic because Python got used in finding
the proof: You need to show that x_n -> 0, where the
error is arctan(x_n). A little trigonometry gives a purely
algebraic recurrence involving the numbers x_n. Now you 
look at the first few using your favorite rational.py and they 
look like 1/3, 1/8, ... then you know what to prove.)

DU




From fredrik at pythonware.com  Wed Apr  5 06:45:39 2000
From: fredrik at pythonware.com (fredrik at pythonware.com)
Date: Wed, 05 Apr 2000 10:45:39 GMT
Subject: Starving for an Advanced Python Book
References: <38EA9A5D.F35B3E74@callware.com>
Message-ID: <8cf5gh$eil$1@nnrp1.deja.com>

Ivan Van Laningham wrote:
> My only quibble with the printed version--sucky index.

didn't know there was an index :-)

there's a script/module index, but no traditional
index.  this will be fixed in the printed edition.

</F>


Sent via Deja.com http://www.deja.com/
Before you buy.


From dave at dogwood.com  Mon Apr 10 15:10:17 2000
From: dave at dogwood.com (Dave Cornejo)
Date: Mon, 10 Apr 2000 19:10:17 GMT
Subject: 2nd CFV: comp.lang.ruby
References: <954456262.23656@isc.org>
Message-ID: <955393816.29413@isc.org>

                      LAST CALL FOR VOTES (of 2)
                   unmoderated group comp.lang.ruby

Newsgroups line:
comp.lang.ruby		The Ruby dynamic OO programming language.

Votes must be received by 23:59:59 UTC, 20 Apr 2000.

This vote is being conducted by a neutral third party.  Questions
about the proposed group should be directed to the proponent.

Proponent: Conrad Schneiker <schneiker at jump.net>
Votetaker: Dave Cornejo <dave at dogwood.com>

RATIONALE: comp.lang.ruby

Ruby is a very-high-level, dynamic object oriented programming language.
Ruby provides many of the best-liked power and convenience features of Perl
and Python on one hand, with an elegantly simple and powerful syntax that
was partly inspired by Eiffel and Ada on the other hand.  Ruby is a more
thoroughly objected oriented language than Python, and some Ruby features
were inspired by Smalltalk. Ruby is an open source language.

There are over 1000 members of 4 Japanese Ruby mail lists (general,
developers, math, extension writers), with a peak volume topping 100
messages per day. At present, there is one English Ruby mail list with
about 175 members, with a recent volume often exceeding 10 messages per
day and growing.  Many users of the Japanese Ruby mail lists have some
understanding of English and are also prospective comp.lang.ruby users
as well.

Much of 1999's newsgroup discussion of Ruby took place on comp.lang.python,
which is not optimal, since comp.lang.python is dedicated to the Python
language. For obvious reasons, Ruby users do not post there frequently. In
2000, most Ruby discussion has occurred in comp.lang.misc. This again is not
ideal since other comp.lang.misc regulars are typically more interested in
issues that transcend Ruby, and have said as much.

In Japan, Ruby has overtaken Python in terms of popular usage. During 1999
and into early 2000, Ruby has been featured in several articles in the
English language software trade press, indicating it is at the takeoff
threshold that Perl and Python attained years ago.  (Please see
http://www-4.ibm.com/software/developer/library/ruby.html for the latest
example.) The English language Ruby home page is
http://www.ruby-lang.org/en/, which provides links to documentation and
download pages for Ruby.

CHARTER: comp.lang.ruby

The comp.lang.ruby newsgroup is devoted to discussions of the Ruby
programming language and related issues.

Examples of relevant postings include, but are not be limited to,
the following subjects:

- Bug reports
- Announcements of software written with Ruby
- Examples of Ruby code
- Suggestions for Ruby developers
- Requests for help from new Ruby programmers

The newsgroup is not moderated.  Binaries are prohibited (except the
small PGP type). Advertising is prohibited (except for announcements
of new Ruby-related products).

END CHARTER.

HOW TO VOTE:

Follow these instructions *exactly*!  Votes are counted by computer.
You should send E-MAIL (posts to a newsgroup are invalid) to:

    vote at dogwood.com

Please do not assume that just replying to this message will work.
Check the address before you mail your vote.  Your mail message
should contain one and only one of the following vote statements:

    I vote YES on comp.lang.ruby
    I vote NO on comp.lang.ruby

    Voter name:

If your mail software does not indicate your real name (for example,
AOL does not), include _exactly_ the statement above on a _separate_
line and add your name after the colon.  Having your name in your
signature line is NOT enough!  Do NOT join the lines together or
remove the words "Voter name"!

You may also vote ABSTAIN (which records an empty vote) or CANCEL
(which removes any earlier votes).  ABSTAIN does not affect the final
vote count in any way but is listed, whereas CANCEL is not.

IMPORTANT VOTING PROCEDURE NOTES:

Standard Guidelines for voting apply.  One vote per person, one
account per voter.  Votes must be mailed directly from the voter to
the votetaker.  Anonymous, forwarded or proxy votes are not valid.
Votes mailed by WWW/HTML/CGI forms are considered to be anonymous
votes.  Votes from non-existent email addresses are not valid.

Vote counting is automated.  Failure to follow these directions may
mean that your vote does not get counted.  If you do not receive an
acknowledgment of your vote within three days contact the votetaker
about the problem.  It's your responsibility to make sure your vote is
registered correctly.  Duplicate votes are resolved in favor of the
most recent valid vote.  Addresses, names and votes of all voters will
be published in the final voting results post.

The purpose of a Usenet vote is to determine the genuine interest of
persons who would read a proposed newsgroup.  Soliciting votes from
disinterested parties defeats this purpose.  Please do not distribute
this CFV.  If you must, direct people to the official CFV as posted to
news.announce.newgroups.  Distributing pre-marked or otherwise edited
copies of this CFV is generally considered to be vote fraud.  When in
doubt, ask the votetaker.

-- 
Voting question & problems: Dave Cornejo <dave at dogwood.com>
Voting address: vote at dogwood.com


From amitp at Xenon.Stanford.EDU  Mon Apr 17 22:49:54 2000
From: amitp at Xenon.Stanford.EDU (Amit Patel)
Date: 18 Apr 2000 02:49:54 GMT
Subject: Filtering web proxy
References: <Pine.LNX.4.21.0004171406250.30999-100000@fep132.fep.ru> <ku66tgh0cm.fsf@lasipalatsi.fi> <slrn8fn3de.gvj.nascheme@cranky.arctrix.com>
Message-ID: <8dgigi$89e$1@nntp.Stanford.EDU>

 Neil Schemenauer <nascheme at enme.ucalgary.ca> wrote:
| 
| Yes and if your connection is fast enough that you don't need
| incremental loading you probably don't care too much about ads.
| In my experience, filtering ads greatly enhances your experience
| if your browsing on a slow connection.

Even with a fast connection, there are still things like filtering out
pop-ups, blocking cookies from certain sites, forcing pages to be
cachable (by modifying Cache-Control headers .. evil evil!) that can
be useful.  One really handy thing I wrote was to highlight keywords
you searched for, so for example, you search for "big dog" and visit a
page, and it'll highlight "big" and "dog" on that page.  However,
Google just added this feature (when you use its cached link) so I
don't need it so much in a proxy.

| Is the situation with XML the same as HTML?  Are XML documents
| forced to adhere to the standard or are parsers supposed to try
| to do something intelligent with whatever crap they get fed?

I believe XML and XHTML parsers are required to reject bad stuff.

| Saying that is parses HTML is a bit of a stretch however.  It
| just uses a couple of regexs.  I'm sure Tim Peters would love
| it. :)

Regexps can "parse" HTML tags.  There was something called REX.py that
was posted here somewhere.  It's based on REX for Perl, which is based
on Robert D. Cameron "REX: XML Shallow Parsing with Regular
Expressions", Technical Report TR 1998-17, School of Computing
Science, Simon Fraser University, November, 1998.

The idea is that REX is a humongous hairy regexp (sorry Timbot!) that
will match one tag or non-tag at a time.  You just keep feeding it
data  and it can keep giving you tags/non-tags.  With that, I hope to
build a filtering proxy that tokenizes all the HTML and does evil
transformations to it, incrementally.


	 - Amit



P.S.  Here's the regexp, just for Tim Peters:

'[^<]+|<(?:!(?:--(?:[^-]*-(?:[^-][^-]*-)*->?)?|\\[CDATA\\[(?:[^\\]]*](?:[^\\]]+])*]+(?:[^\\]>][^\\]]*](?:[^\\]]+])*]+)*>)?|DOCTYPE(?:[\\n\\t\\r]+(?:[A-Za-z_:]|[^\\x00-\\x7F])(?:[A-Za-z0-9_:.-]|[^\\x00-\\x7F])*(?:[\\n\\t\\r]+(?:(?:[A-Za-z_:]|[^\\x00-\\x7F])(?:[A-Za-z0-9_:.-]|[^\\x00-\\x7F])*|"[^"]*"|\'[^\']*\'))*(?:[\\n\\t\\r]+)?(?:\\[(?:<(?:!(?:--[^-]*-(?:[^-][^-]*-)*->|[^-](?:[^\\]"\'><]+|"[^"]*"|\'[^\']*\')*>)|\\?(?:[A-Za-z_:]|[^\\x00-\\x7F])(?:[A-Za-z0-9_:.-]|[^\\x00-\\x7F])*(?:\\?>|[\\n\\r\\t][^?]*\\?+
(?:[^>?][^?]*\\?+)*>))|%(?:[A-Za-z_:]|[^\\x00-\\x7F])(?:[A-Za-z0-9_:.-]|[^\\x00-\\x7F])*;|[\\n\\t\\r]+)*](?:[\\n\\t\\r]+)?)?>?)?)?|\\?(?:(?:[A-Za-z_:]|[^\\x00-\\x7F])(?:[A-Za-z0-9_:.-]|[^\\x00-\\x7F])*(?:\\?>|[\\n\\r\\t][^?]*\\?+(?:[^>?][^?]*\\?+)*>)?)?|/(?:(?:[A-Za-z_:]|[^\\x00-\\x7F])(?:[A-Za-z0-9_:.-]|[^\\x00-\\x7F])*(?:[\\n\\t\\r]+)?>?)?|(?:(?:[A-Za-z_:]|[^\\x00-\\x7F])(?:[A-Za-z0-9_:.-]|[^\\x00-\\x7F])*(?:[\\n\\t\\r]+(?:[A-Za-z_:]|[^\\x00-\\x7F])(?:[A-Za-z0-9_:.-]|[^\\x00-\\x7F])*(?:[\\n\\t\\r]+)?=?(?
:[ \\n\\t\\r]+)?(?:"[^<"]*"|\'[^<\']*\'|\\w+))*(?:[ \\n\\t\\r]+)?/?>?)?)'

--
Amit J Patel, Computer Science Department, Stanford University
http://www-cs-students.stanford.edu/~amitp/


-- 
--
Amit J Patel, Computer Science Department, Stanford University
http://www-cs-students.stanford.edu/~amitp/


From corg at copernic.com  Wed Apr 19 11:17:21 2000
From: corg at copernic.com (Gaetan Corneau)
Date: Wed, 19 Apr 2000 11:17:21 -0400
Subject: Very Horrible Question About Goto's
Message-ID: <C5A7720BDEB7D3118A5200E018C200A037EA43@EXCHANGE>

Hi,

> 1) the only valid use of goto's in C I've seen looks like this:
> 
> 	if (bogus)
> 		goto cleanup
> 	do stuff;
> 	return(okay);
> cleanup:
> 	close file handles;
> 	do cleanup
> 	return(error)

There are, at least, two other valid uses:
1) the most important: goto's are used a lot by code generators (it
simplifies writing the generator alot in some cases)

2) in extreme cases, to improve performances.


IMHO, Goto's *are* evil in languages where you can jump around from one
function to another, but they can be useful in some cases. 

My own rules are: 
- don't use goto's

if you HAVE to use goto's:
- don't jump from one function to another (if your language supports it)
- don't jump backward

I was surprised to learn that Python doesn't have a goto statement.
Would it be difficult to add?
______________________________________________________
   Gaetan Corneau
   Software Developer 
   Copernic.com 
   http://www.copernic.com
   E-mail: corg at copernic.com                         
______________________________________________________
"Tu l'as trop ?cras?, C?sar, ce Port Salut"




From mskott at image.dk  Tue Apr 25 16:56:31 2000
From: mskott at image.dk (Martin Skøtt)
Date: Tue, 25 Apr 2000 20:56:31 GMT
Subject: Working with data returned by PyGreSQL
References: <itx6ni3w.fsf@jumbo.skott.dk>
Message-ID: <em7toq6i.fsf@jumbo.skott.dk>

I apoligize for asking this question without checking all the
available documentation. After reading the source in pgmodule.c
everything became much more clear to me. 
I have now subscribed to the PyGreSQL mailing list and plan to start
testing the new DB-API module one of the following days.
-- 
Martin Sk?tt
mskott at image.dk

'In a certain sense, all decent programming languages are equally powerfull'
Dr. Alan Turing


From erno at iki.fi  Tue Apr  4 15:20:25 2000
From: erno at iki.fi (Erno Kuusela)
Date: 04 Apr 2000 22:20:25 +0300
Subject: opening a serial port at a given baudrate. How?
References: <200004041153.VAA15869@piglet.dstc.edu.au>
Message-ID: <kubt3pisg6.fsf@lasipalatsi.fi>

>>>>> "davida" == David Arnold <arnold at dstc.edu.au> writes:
    davida> on unix (including linux), just open the device special
    davida> file corresponding to the serial port of interest
    davida> (ie. /dev/ttyS0) and you can then control it using
    davida> fcntl.ioctl() function and otherwise read from and write
    davida> to it like a normal file object.

i believe the termios module is for ftwiddling with serial
ports on unix. see for example ftp://ftp.tummy.com/pub/tummy/cid/
for example code.

   -- erno


From hamish_lawson at yahoo.co.uk  Thu Apr 20 10:40:28 2000
From: hamish_lawson at yahoo.co.uk (Hamish Lawson)
Date: Thu, 20 Apr 2000 07:40:28 -0700
Subject: [ANNOUNCE] Webware for Python 0.1
References: <5650A1190E4FD111BC7E0000F8034D26A0F40D@huina.oceanic.com> <38FEF62B.2A4402CD@mindspring.com> <T6DL4.25197$cZ.34941@typhoon.southeast.rr.com>
Message-ID: <2cd13488.77055138@usw-ex0108-063.remarq.com>

Don Tuttle wrote:
> WinZip is having problems with your tar.  It's giving me the
> error "Error reading header after processing 0 times."

The file is actually a gzipped tar file, but your browser is
probably curtailing the extension to .tar, thus miseading WinZip.
You can get round this by manually giving the file a .tar.gz
extension when you save it from your browser. The saved file
should now open OK in WinZip.

Alternatively you could do it The Python Way (TM) :-)

>>> import urllib
>>> urllib.urlretrieve(
'http://writewithme.com/Webware/download.cgi', 'webware.tar.gz')


Hamish Lawson




* Sent from RemarQ http://www.remarq.com The Internet's Discussion Network *
The fastest and easiest way to search and participate in Usenet - Free!



From pinard at iro.umontreal.ca  Sat Apr  1 09:16:09 2000
From: pinard at iro.umontreal.ca (=?ISO-8859-1?Q?Fran=E7ois_Pinard?=)
Date: 01 Apr 2000 09:16:09 -0500
Subject: Why should I switch to Python?
In-Reply-To: "Chris Ryland"'s message of "Sat, 01 Apr 2000 04:09:33 GMT"
References: <seas5ia89177@corp.supernews.com> <1EeF4.43$CA.2422@news>
Message-ID: <oqzord6h5i.fsf@titan.progiciels-bpi.ca>

"Chris Ryland" <cpr at emsoftware.com> writes:

> Oh, no, not another Perl vs. Python war! ;-)

That remembers me of a little story that made me smile, years ago.  This was
the comment of a discouraged nurse at a recruiting centre for the army,
at the entrance of the medical examination room.  "I told them for _years_
to remove their shirt, and they still do not know they should."

There will be newcomers to Python all the time, and they are likely to ask
questions which I asked, and which you asked, when you initially came here.
I quite remember I got in, not so long ago, with this brilliant, brand new,
clever idea that we should write a Perl to Python automatic converter.
You people received me with an exquisite politeness and patience, and my
liking of Python deepened instantly as it got kind of fuzzily intermixed,
in my fragile mind, with the feeling of good human values, as well.

We should not see the potential for a war every time Perl is discussed,
and be patient instead.  It looks all pretty bearable so far.  And if it
was becoming too heavy, we could prepare pre-canned replies for the most
questions, and even suggest a pointer to this FAQ right into the message
footer generated for this list.  Welcoming people in person is better!

-- 
Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard





From mfletch at tpresence.com  Thu Apr 13 10:33:25 2000
From: mfletch at tpresence.com (Mike Fletcher)
Date: Thu, 13 Apr 2000 10:33:25 -0400
Subject: List mapping?
Message-ID: <CC59868A0A76D311B1F50090277B321C1F326A@VRTMAIL>

not elegant, but here's another way...

original = string.split( x, ' ')
result = []
while original:
	result.append( tuple(original[:2]) )
	del original[:2]

and another...

original = string.split( x, ' ')
result = []
for x in range(0, len(original), 2 ):
	result.append( tuple( original[x:x+2]))

HTH,
Mike

-----Original Message-----
From: Nick Trout [mailto:nick at spam_me_notvideosystem.co.uk]
Sent: Thursday, April 13, 2000 10:10 AM
To: python-list at python.org
Subject: List mapping?


Is there a nice way of pairing up the members to form tuple pairs in another
list? ie. like you might to processing input args.

ie. changing ['/a','1','/b','2','/c','3'] into [ ('/a','1'), ('/b','2'),
('/c','3') ]

eg:

args = '/a 1 /b 2 /c 3'
listargs= string.split(args)
tuplist = []
i=0
while i < len(listargs):
    tuplist.append( (listargs[i],listargs[i+1]) )
    i = i+2
for i in tuplist:
    processargs(i)

or:

args = '/a 1 /b 2 /c 3'
for i in string.split(args):
    i = argflag
    # i want next arg now as a parameter!!!

Can I (you!) get rid of the long winded while loop cleverly?!!  :-)

Thanks in advance,
Nick.





-- 
http://www.python.org/mailman/listinfo/python-list



From leursst at aol.com  Sat Apr  8 16:14:23 2000
From: leursst at aol.com (Leursst)
Date: 08 Apr 2000 20:14:23 GMT
Subject: starting python
References: <8cnkfi$at7$1@porthos.nl.uu.net>
Message-ID: <20000408161423.19747.00000485@ng-ba1.aol.com>

Hi!
You can find the Python langage on the web site www.python.org (official).
You can download the version 1.5.2 for windows 95,98,NT or other systems.

You've got, on this site, all the documents you need to begin Python.

Good luck...


From paul.magwene at yale.edu  Tue Apr 11 16:06:41 2000
From: paul.magwene at yale.edu (Paul Magwene)
Date: Tue, 11 Apr 2000 16:06:41 -0400
Subject: Functions and objects
References: <meh9-C7461E.15255011042000@news.cit.cornell.edu>
Message-ID: <38F385D1.9ED3FACC@yale.edu>

Matthew Hirsch wrote:
> 
> Hi All,
> 
> I'm confused.  Take a look at the following:
> 
> >>> def f(x):
> ...     return [x]
> ...
> >>> a=f(5)
> >>> a
> [5]
> >>> for x in range(5):     # Case A
> ...     temp=a
> ...     print a
> ...     temp.append(1)
> ...
> [5]
> [5, 1]
> [5, 1, 1]
> [5, 1, 1, 1]
> [5, 1, 1, 1, 1]
> 
> >>> a=f(5)
> >>> for x in range(5):     # Case B
> ...     temp=a[:]
> ...     print a
> ...     temp.append(1)
> ...
> [5]
> [5]
> [5]
> [5]
> [5]
> >>>
> 
> In case A, why doesn't temp reset itself to the value of a, [5], that
> was predetermined before it entered the loop?  Why do you need to copy
> the list to get the behavior I'm looking for in Case B?  Doesn't a
> always equal [5]?
> 
> Thanks for your help,
> Matt

Matthew,

Many of your questions to the group are addressed in the tutorial and/or
FAQ.  Obviously people in this newsgroup are glad to offer help and
information, but it's always a good idea to try and track down your own
sources of info first.

You question in this case is addressed in the Python tutorial, sections
4.6 and 4.7 (and footnote 4.1).

In a Python function, what gets passed to the function is an object
reference.  

Assignment always stores a reference to an object, it DOES NOT make a
copy of the object. 

The object you are passing in both cases is a list, therefore mutable. 
In case A, the assignment temp = a, is equivalent to sticking a new
"label" on your mutable list object.  Then your appending stuff to this
object (called either "a" or "temp").  In case B, your are generating a
new copy of a list object, and calling this new object "temp". 
Therefore, modifying the "temp" object has no bearing on the "a" object.

If you are going to do a significant amount of Python programming it
might be useful to acquire one of the several available python texts. 
Learning Python, by Lutz and Ascher is a good one to start.

Best,

Paul


From dyoo at uclink4.berkeley.edu  Tue Apr 25 13:03:55 2000
From: dyoo at uclink4.berkeley.edu (dyoo)
Date: Tue, 25 Apr 2000 17:03:55 GMT
Subject: new linereading standard?
References: <3904D21B.333DD930@visionart.com>
Message-ID: <%dkN4.33801$k5.960576@news1.frmt1.sfba.home.com>

In article <3904D21B.333DD930 at visionart.com>, 
Pete Shinners <pete at visionart.com> wrote:
> i've come to the point where i can't get by even the simplest python
> project without dumping this code into my files. it is a simple class
> that handles the "while 1: .... break" carnival for reading files.
> now i can code in a preferred style

> myfile = open('myfile.txt', 'r')
> for line in filelines(myfile):
>     print line
> myfile.close()

> as a person trying to introduce his friends to python, i repeatedly find
> this is the ugliest wart while doing basic stuff with python.

Python does allow you to read the whole file at a time.  
readlines() will give you a list of all the lines in the file, so 
your code can look like:

  myfile = open('myfile.txt', 'r')
  for line in myfile.readlines():
    print line
  myfile.close()



From arnold at dstc.edu.au  Tue Apr  4 23:19:31 2000
From: arnold at dstc.edu.au (David Arnold)
Date: Wed, 05 Apr 2000 13:19:31 +1000
Subject: opening a serial port at a given baudrate. How? 
In-Reply-To: Your message of "Tue, 04 Apr 2000 19:08:39 MST."
             <XFMail.20000404190839.warlock@eskimo.com> 
Message-ID: <200004050319.NAA08789@piglet.dstc.edu.au>

-->"Jim" == Jim Richardson <warlock at eskimo.com> writes:

  Jim> I have been using open('/dev/ttyS0','r+') but need to control
  Jim> baud rate. fcntl.ioctl() will do this?

as someone else pointed out (sorry, lost the mail), the termios module
is the go for setting baud rates, parity, etc.

some old code:  not sure if it works, but might be useful ...


	#-- open serial port to device
	self.ir_file = open(str_port, "w+")
	ir_fh = self.ir_file.fileno()

	#-- get current settings
	[c_iflag, c_oflag, c_cflag, c_lflag, ispeed, ospeed, ccs] = \
		  termios.tcgetattr(ir_fh)

	#-- input flags
	off = 0
	off = off | TERMIOS.IGNBRK    # don't ignore break
	off = off | TERMIOS.BRKINT    # no interrupt on break
	off = off | TERMIOS.IGNPAR    # don't throw away parity errors
	off = off | TERMIOS.PARMRK    # don't mark parity errors
	off = off | TERMIOS.ISTRIP    # don't strip 8bit
	off = off | TERMIOS.INLCR     # no nl->cr xlate
	off = off | TERMIOS.IGNCR     # don't ignore CR
	off = off | TERMIOS.ICRNL     # no cr->nl xlate
	off = off | TERMIOS.IXON      # no output flow control
	off = off | TERMIOS.IXOFF     # no input flow control
	on  = 0
        on  = on  | TERMIOS.INPCK     # disable parity checking
	c_iflag = (c_iflag & ~off) | on

	#-- output flags
	off = 0
	off = off | TERMIOS.OPOST     # no ouput processing
	off = off | TERMIOS.OLCUC     # no lower to upper xlate
	off = off | TERMIOS.OCRNL     # no cr->nl xlate
	off = off | TERMIOS.ONOCR     # cr at 0 ok
	off = off | TERMIOS.ONLRET    # nl does not imply cr
	off = off | TERMIOS.OFILL     # no fill chars generated
        on  = 0
	c_oflag = (c_oflag & ~off) | on

	#-- control flags (see p10, CE-IR2 manual)
	off = 0
	off = off | TERMIOS.CSIZE     # zero out the char-size bits
	off = off | TERMIOS.CSTOPB    # one stop bit
	off = off | TERMIOS.HUPCL     # don't hangup on close
	on  = 0
	on  = on  | TERMIOS.CS8       # 8 bit chars
	on  = on  | TERMIOS.CREAD     # allow receives
	on  = on  | TERMIOS.PARENB    # enable parity generation
	on  = on  | TERMIOS.PARODD    # use odd parity
	on  = on  | TERMIOS.CLOCAL    # ignore modem status lines
        c_cflag = (c_cflag & ~off) | on

	#-- local flags
	off = 0
	off = off | TERMIOS.ECHO      # no local echo of output
	off = off | TERMIOS.ISIG      # no signals from special chars
	off = off | TERMIOS.ICANON    # disable canonical processing
	off = off | TERMIOS.IEXTEN    # disable other special chars
	off = off | TERMIOS.NOFLSH    # no flush on signal
	off = off | TERMIOS.TOSTOP    # no stop for background output
        on  = 0
	c_lflag = (c_lflag & ~off) | on

	#-- set VTIME and VMIN both to zero (ie. poll and return what's there)
	ccs[TERMIOS.VMIN] = 0     # read a minimum of 0 chars
	ccs[TERMIOS.VTIME] = 0    # wait for 0/10 sec for data to arrive

	termios.tcsetattr(ir_fh, TERMIOS.TCSANOW, \
			  [c_iflag,c_oflag,c_cflag,c_lflag,9600,9600,ccs])




From andy at reportlab.com  Mon Apr 10 07:15:06 2000
From: andy at reportlab.com (Andy Robinson)
Date: Mon, 10 Apr 2000 12:15:06 +0100
Subject: ReportLab PDF Library v0.92 - print code and presentations to PDF!
Message-ID: <PGECLPOBGNBNKHNAGIJHCEPMCAAA.andy@reportlab.com>

ReportLab version 0.92 is now available.

ReportLab is an Open Source library for creating PDF
documents in pure Python.  ReportLab puts Python and 
PDF together to create the ultimate reporting tool - 
acquire data from anywhere, format it any way you want,
deliver it any way you want.

New features in this release include:
- A preview of Dinu Gherman's "py2df", which prints Python
code with correct syntax colorizing and a host of options
- PythonPoint, an app to make presentation slides from XML
markup.  
- experimental support for outlines and hyperlinks
- enhanced support for colors - use HTML color names

Check out the gallery and demos at http://www.reportlab.com/ 
to see what our users have created.

Download (about 340kb):
ftp://ftp.reportlab.com/reportlab.tgz or
ftp://ftp.reportlab.com/reportlab.zip 

Support and discussion:  
Send a message with 'subscribe' in the subject 
line to "reportlab-users-subscribe at egroups.com"


Andy Robinson
CEO and Chief Architect, ReportLab Inc.



From bill at libc.org  Thu Apr 27 20:55:15 2000
From: bill at libc.org (Bill Anderson)
Date: Thu, 27 Apr 2000 18:55:15 -0600
Subject: The Simple Economics of Open Source
References: <B529350E.51B9%raffael@mediaone.net> <390534BB.3D6CEBDF@libc.org> <raffael-01CD89.20575826042000@news.ne.mediaone.net>
Message-ID: <3908E173.5E2DFD72@libc.org>

Raffael Cavallaro wrote:
> 
> In article <390534BB.3D6CEBDF at libc.org>, Bill Anderson <bill at libc.org>
> wrote:
> 
> >Unless you are redefining
> >'commodity' (possibly by your 'known to many' remark) to fit whatever is
> >provided as counter evidence.
> 
> A commodity is that which is available in large quantities, from many
> vendors offering essentially interchangeable product. This makes
> commodities inexpensive, hence the term "at commodity prices" meaning,
> "inexpensive."
> 
> In intellectual property then, a commodity is that which is known to
> many and not under copyright or patent.

All software is under copyright and/or patent. You are confusing
license/use/ubiquitousness with patent/copyright.

> Because of the near zero cost of
> copying, the large quantity availability is a non-issue with software.
> What matters is whether or not it is known to many, because you then
> have many potential suppliers. If it is only known to a few, then the
> number of potential suppliers is few, and prices are not like those of a
> commodity market.
> 
> Examples: The JPEG standard is known to many, and not under copyright or
> patent. Therfore, software that compresses/decompresses JPEGs is
> available from many suppliers (again, the ease of copying electronic
> media moots the quantity issue). Because of the fact that JPEG
> compression/decompression is known to many, software to perform these
> functions is plentiful and cheap (i.e., often free).

Incorrect. 
Although the "baseline" variety of JPEG is *believed* patent-free, there
are many patents associated with some optional features of JPEG, namely
arithmetic coding and hierarchical storage. 

GIF is at least as prevalent as JPEG, yet it is under restrictions (note
again, this is different than copyright)
 
> The Microsoft Word (.doc) format is not known to many (really, only to
> MS programmers) and binaries to read/display/edit data in this format
> are under copyright. Although others have tried to reverse engineer this
> ever changing format, they have not been 100% successful. The result is
> that if you want to be certain that you can read/display/edit a document
> in this format, you have only one supplier, namely, Microsoft. Because
> of this, software to perform these functions is not a commodity item,
> and is quite expensive.

The MSWord format is not under patent or copyright. Software to read and
write MSWord format is freely available, and free. You may want to pick
a better example set.
 
> In short, I'm not "redifining" the word "commodity." I'm using it as it
> has always been used. I think that many open source advocates have
> failed to apply simple logic to the economics of open source, believing
> that somehow, the internet, or software, is immune from simple laws of
> supply and demand. They are not.

Yes, you are most certainly redefining commodity. At a minimum you are
misinterpreting what 'commodities' are. Perhaps the below will help:
"""
1 : an economic good: as a : a product of agriculture or mining b :
an article of commerce especially when delivered for shipment
<commodities futures> c : a mass-produced unspecialized
product <commodity chemicals> <commodity memory chips>

2 a : something useful or valued <that valuable commodity
patience> b : CONVENIENCE, ADVANTAGE

3 obsolete : QUANTITY, LOT

4 : one that is subject to ready exchange or exploitation within a
market <the sensitive female singer-songwriter as a viable pop
commodity -- Elysa Gardner> 
"""

Absolutely nothing here is in accordance with your use of commodity. I
have _never_ heard commodity used in the way you are using or defining
it.


-- 
In flying I have learned that carelessness and overconfidence are 
usually far more dangerous than deliberately accepted risks. 
          -- Wilbur Wright in a letter to his father, September 1900


From Michael.Scharf at gmx.de  Wed Apr 12 21:36:05 2000
From: Michael.Scharf at gmx.de (Michael Scharf)
Date: Thu, 13 Apr 2000 03:36:05 +0200
Subject: Classes, Inheritance - Stupid lazy question
References: <8d1cfh$hme7@imsp212.netvigator.com> <874s9766r6.fsf@ifm.uni-kiel.de> <m3itxnis32.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <38F52485.1599B01C@gmx.de>

Michael Hudson wrote:
> another option is:
> 
> class paul(lazy):
>     super = lazy
>     def __init__(self, name):
>         self.super.__init__(self,name)
>         self.april_pay = 0
> 
>     def fired(self):
>         print "Here's a box, collect the things from your desk"
>         self.super.fired(self)
> 
> Though this has problems with repeated & multiple inheritance.

What happens if you want to inherit from paul? 
self.super would be overwritten!

Make super private (__super) and it will work fine. 

We use a similar trick in C++ (a private <typedef MySuperClas super>),
so changing the superclass requires only two changes in the header...

#---snip---

class lazy:
    def __init__(self, name):
        self.name = name

    def fired(self):
        print 'You are fired %s!' % self.name

class paul(lazy):
    __super=lazy
    def __init__(self, name):
        self.__super.__init__(self,name)
        self.april_pay = 0

    def fired(self):
        print "Here's a box, collect the things from your desk."
        self.__super.fired(self)

class michael(paul):
    __super=paul
    def fired(self):
        print self.name, "will fire paul!"
        self.__super.fired(self)

lazy("lazy").fired()
print
paul("paul").fired()
print
michael("michael").fired()

#---snap---

Michael
-- 
     ''''\     Michael Scharf
    ` c-@@     TakeFive Software GmbH, a Wind River Company
    `    >     http://www.WindRiver.com
     \_ V      mailto:Michael.Scharf at gmx.de


From niels at endea.demon.nl  Thu Apr 27 13:06:42 2000
From: niels at endea.demon.nl (Niels Diepeveen)
Date: Thu, 27 Apr 2000 19:06:42 +0200
Subject: module case bug in 1.6a2 Win9x
References: <UGQuwCAnnFC5Ew+e@jessikat.demon.co.uk>
Message-ID: <390873A2.BD0EF60F@endea.demon.nl>


Robin Becker schreef:
> 
> I believe there's a module case bug with long named modules under win98
> at least. I believe the bug relates to the length of the filename
> somehow.
> 

Yes, that looks like a bug in import.c:allcaps8x3(). It seems to return
true instead of false for long filenames. As far as I can tell, this
probably happens on all Win32 platforms at least since 1.5.2.

-- 
Niels Diepeveen
Endea automatisering


From mlh at swl.msd.ray.com  Thu Apr 27 09:58:05 2000
From: mlh at swl.msd.ray.com (Milton Hankins)
Date: Thu, 27 Apr 2000 13:58:05 GMT
Subject: plugins for browsers?
References: <3904ccb0.32134281@news.demon.co.uk>
Message-ID: <8e9h15$tm0$1@nnrp1.deja.com>

In article <3904ccb0.32134281 at news.demon.co.uk>,
  greg at quokka.demon.co.uk wrote:
> 	Are there any plugins or other means of running Python in Web
> browsers? I seem to recall that there were a while ago ...

I think WPY has some support for doing that.  There was some work on a
plug-in by Brook and Smirl, but I think it has since vanished into
the great abyss.

I worked on a Tkinter plug-in for Netscape at my company, but there is
quite a bit of red tape involved in releasing any info on that.  I could
act as a technical reference for anyone wishing to start such a new
endeavor on their own.


Sent via Deja.com http://www.deja.com/
Before you buy.


From nmm1 at cus.cam.ac.uk  Tue Apr  4 08:14:09 2000
From: nmm1 at cus.cam.ac.uk (Nick Maclaren)
Date: 4 Apr 2000 12:14:09 GMT
Subject: Bugs in Python 1.6 alpha 1
References: <Pine.LNX.4.10.10003311651590.22919-100000@starship.python.net> <8c8vfj$5n6$1@news.udel.edu>
Message-ID: <8ccmah$chj$1@pegasus.csx.cam.ac.uk>

Here are some problems with building and running the tests of the
1.6 Alpha release:

    1) The Modules/sre* files have control-M at the end.

    2) The 'unsigned char *' on lines 596 and 631 of mmapmodule.c should
be 'char *'.  There is also a pointless check '(where >= 0)' for an
unsigned value on line 408.

    3) _sre.c lines 222 and 225 and mmapmodule.c lines 715 compare
pointers to object types and pointers to void - this is not allowed,
but I don't think that it is particularly unsafe.  But it is an
error in a few compilers, so should be fixed.

    4) If the call to mmap in mmapmodule.c fails, the diagnostic is
"Invalid argument" - which can be downright wrong (as it was in this
case.)

    5) There is something very odd with the test harness, though I can't
see the problem in regrtest.py.  'make test' says test_pyexpat crashed,
but running the single test says that the module is not configured.  This
happens even under Irix.

    6) test_unicode fails on Hitachi systems -- Writing: '*', expected:
'T'.  But only if -v is NOT specified!  Again, I failed to tie it down,
and it looks like a test harness problem.

    7) configure is still buggy and gets things wrong :-(

    8) A few other failures weren't Python's fault, and I have
reported them to the vendors.


Regards,
Nick Maclaren,
University of Cambridge Computing Service,
New Museums Site, Pembroke Street, Cambridge CB2 3QG, England.
Email:  nmm1 at cam.ac.uk
Tel.:  +44 1223 334761    Fax:  +44 1223 334679


From musingattheruins at my-deja.com  Thu Apr 13 12:20:47 2000
From: musingattheruins at my-deja.com (musingattheruins at my-deja.com)
Date: Thu, 13 Apr 2000 16:20:47 GMT
Subject: Complaint Dept: My shoes!
Message-ID: <8d4s4d$mvr$1@nnrp1.deja.com>

If shoe is None then len(shoe) raises an exception.  Would be nice if
len(None) was 0 and NOT raise an exeption.

Example: (Python 1.5.2)

def show_shoe_wear_to_complaint_dept(name):
   if name == "Idle":
      return None
   else:
      return "pick up hammer..."

def hit_on_head(): return "POW!!!"

shoe = ""
shoe = show_shoe_wear_to_complaint_dept("Idle")
if len(shoe)>0: print shoe, hit_on_head()

If len(shoe) returned 0 then the example would work without error, the
example above raises an exception whenever Idle shows his shoe to the
complaint department.

The work around for the above is....
shoe = show_shoe_wear_to_complaint_dept("Idle") or ""
which gets rid of the problem but feels weird.

On the other hand, if len(shoe) returned None then we would have to
discuss why the following prints "oww!"

if None>1: print "oww!"

and maybe we don't want to go there.

L8R :-)


Sent via Deja.com http://www.deja.com/
Before you buy.


From jae at ilk.de  Thu Apr 27 18:17:16 2000
From: jae at ilk.de (Juergen A. Erhard)
Date: Fri, 28 Apr 2000 00:17:16 +0200
Subject: Scripting and Gnome and KDE
In-Reply-To: <wkln1zll9r.fsf@mac.com> (message from Jody Winston on 27 Apr
	2000 08:35:12 -0500)
References: <Pine.GSO.4.10.10004150741520.11261-100000@sundial> <3dk8hzb90u.fsf@newcnri.cnri.reston.va.us> <shk8hk2m68.fsf@tux.gnu.franken.de> <wkln1zll9r.fsf@mac.com>
Message-ID: <28042000.1@sanctum.jae.ddns.org>

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

>>>>> "Jody" == Jody Winston <josephwinston at mac.com> writes:

[...]

    Jody> It looks like to me that we have gone back to the model that
    Jody> all the world is a Vax and if it compiles and runs on the
    Jody> Vax there isn't any problem with the software.  All the
    Jody> world isn't Linux and the only compiler isn't gcc.

I have to say that for *me*, all the world *I know* *is* Linux...  I
do not have access to anything else, so I don't even *know* what works
or what doesn't (I've worked on a DECStation 5100 a *loong* time ago,
and I've had the `pleasure' of working in VMS Pascal (on a Vax of
course)).

This is one thing I found lacking in autoconf (which is generally a
great tool for these things): A guide to things that should be
autoconfiscated (beyond the obvious like word order etc[1]).

Hmmm, someone should really start something like this...

Generally, most Free Software developers are dependent on user reports
(and patches).  So, if something doesn't work on your box, let the
developers know...

Bye, J

[1] `Obvious' is, of course, very subjective.  Word order is obvious
for *me* because I once did all my (personal) development on an
Amiga... which is big-endian in contrast to Intel chips.

- -- 
J?rgen A. Erhard      eMail: jae at ilk.de      phone: (GERMANY) 0721 27326
          My WebHome: http://members.tripod.com/Juergen_Erhard
              Linux - Free PC Unix (http://www.linux.org)
             pros do it for money -- amateurs out of love.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.1 (GNU/Linux)
Comment: Use Mailcrypt and GnuPG <http://www.gnupg.org/>

iEYEARECAAYFAjkIvGsACgkQN0B+CS56qs0ZegCgmDIN8gSngm4E2QIKdPhFv6w1
6mAAn0U2siR4xtQ9aoaMNOiIMRGeN4fn
=0dHj
-----END PGP SIGNATURE-----



From gmcm at hypernet.com  Tue Apr 11 16:49:37 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Tue, 11 Apr 2000 16:49:37 -0400
Subject: Thread-safe atomic operation
In-Reply-To: <8d016r$jbf$1@nntp9.atl.mindspring.net>
Message-ID: <1256636312-42459974@hypernet.com>

Aahz Maruch wrote:

> In article <1256659126-41087547 at hypernet.com>,
> Gordon McMillan <gmcm at hypernet.com> wrote:
> >Dong-gweon Oh writes:
> >>
> >> Are these statements thread-safe?
> >> 1. import xxx
> >> 2. xxx = __import__('xxx')
> >> 3. reload(xxx)
> >> 4. dict['xxx'] = sys.modules['xxx']
> >> 5. dict.has_key('xxx')
> >
> >Numbers 2 and 4 could be interrupted between evaluation of 
> >the RHS and assigment to LHS. For #2 that would be of 
> >concern only if xxx was a global.
> 
> Are you absolutely certain about item 1?  That implies that Python as a
> whole could block on importing a module that did something like
> 
> for i in range(100):
>   sleep(1)

I didn't say it was atomic, I agreed it was thread safe. There's 
special code in there dealing with competing imports between 
threads, but the global lock is not held for the duration of the 
import (in particular, the top level Python code in the imported 
module executes according to the normal rules).

- Gordon



From chafik at nevrax.com  Wed Apr 12 08:33:01 2000
From: chafik at nevrax.com (Sameh chafik pro)
Date: Wed, 12 Apr 2000 14:33:01 +0200
Subject: Are there any one know how can i do to run a node table ?
Message-ID: <8d1mhg$aa2$1@gw.localdomain>

Hello evrybody,
I find in the Python C API the folwing function:
struct _node* PyParser_SimpleParseString (char *str, int start);
struct _node* PyParser_SimpleParseFile (FILE *fp, char *filename, int
start);
but i do'nt fin a function like
PyObject* PyRun_String (struct _node *str, int start);
Are there any one, now how can i do to run a node table ?






From dalke at acm.org  Sat Apr  8 21:06:17 2000
From: dalke at acm.org (Andrew Dalke)
Date: Sat, 8 Apr 2000 19:06:17 -0600
Subject: Python paradigms
References: <8cmr70$b0e$1@pegasus.csx.cam.ac.uk> <8cntqp$upi$1@slb0.atl.mindspring.net> <8cocgb$iuk$1@pegasus.csx.cam.ac.uk>
Message-ID: <8cokt5$opj$1@slb1.atl.mindspring.net>

Me:
>>How about
>>  x = getattr(a, "weeble", 0) + getattr(b, "wombat", 0)


Nick Maclaren:
>Thank you, yes, that does deal with that specific example.  But
>it does not extend to the general case.


Ahh, I thought the general case was "test if something exists
and, if so, get data from it, otherwise use a given value."

What is the general case you are interested in - that of C's
ternary ?: operator?  I've rarely missed it, because most of
the subcases are handled by things like getattr.

                    Andrew
                    dalke at acm.org





From nickb at earth.ox.ac.uk  Tue Apr 18 08:42:44 2000
From: nickb at earth.ox.ac.uk (Nick Belshaw)
Date: Tue, 18 Apr 2000 13:42:44 +0100
Subject: May 27th PARTY!
References: <006201bfa756$896e5a80$da00a8c0@austin.rr.com>
Message-ID: <38FC5844.D7DF90EB@earth.ox.ac.uk>

Gees - I know they want everyone to go but -

do you think we could just send Guido as our representative? - I am busy
on the 27th.

/N






From aa8vb at yahoo.com  Fri Apr 21 07:40:28 2000
From: aa8vb at yahoo.com (Randall Hopper)
Date: Fri, 21 Apr 2000 07:40:28 -0400
Subject: how do i pack and unpack messages in python?
In-Reply-To: <00ed01bfaad2$abcf3200$2801000a@iel.ie>
References: <008301bfaacb$42fe2c60$2801000a@iel.ie>
 <20000420100706.A624577@vislab.epa.gov> <00ed01bfaad2$abcf3200$2801000a@iel.ie>
Message-ID: <20000421074028.A734152@vislab.epa.gov>

Shaun:
 |Randall:
 |> Shaun:
 |>  |id           : 10
 |>  |subid        : 13
 |>  |NV sequence  : {14, "shaun"-string
 |>  |               {20, ["Anytown, Anywhere"]-list
 |>  |               {34, 1234567890-int
 |>  |
 |>  |i want to pack the message to look like this:
 |>  |
 |>  |10 fs 13 fs 14 fs s shaun fs 20 fs l s anytown fs s anywhere fs i
 |>  |1234567890 gs gs
 |>
 |> I was going to mention the struct, array, or xdrlib modules to you, but
 |> from your desired output format (all ASCII it appears), it looks like
 |> a simple string.join would do the job:
 |
 |actually randall, i kinda simplified the output, it was more representative
 |than actual, so if you have time to mention the struct, array, or xdrlib
 |modules, i would be gratefull.

struct gives you packing and unpacking of basic Python types into binary
data structures (like C structures).  array does this for long homogeneous
sequences of integer and floating point types (you can use struct, but
array saves looping in Python).  And xdrlib gives you the capability to
serialize to a binary RFC 1014 stream such as that used by RPC.

So if you want to write a binary stream of the above data, you may want to
check out struct or xdrlib.

>>> import struct
>>> struct.pack( 'iiii5s', 10, 13, 14, len('shaun'), 'shaun' )
'\000\000\000\012\000\000\000\015\000\000\000\016\000\000\000\005shaun'

>>> import xdrlib
>>> packer = xdrlib.Packer()
>>> packer.pack_int( 10 )
>>> packer.pack_int( 13 )
>>> packer.pack_int( 14 )
>>> packer.pack_string( 'shaun' )
>>> packer.get_buffer()
'\000\000\000\012\000\000\000\015\000\000\000\016\000\000\000\005shaun\000\000\000'

-- 
Randall Hopper
aa8vb at yahoo.com



From mwh21 at cam.ac.uk  Fri Apr 14 04:27:03 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 14 Apr 2000 09:27:03 +0100
Subject: how do I know to what module an object belongs?
References: <Pine.BSF.4.10.10004132146270.4951-100000@chi.pair.com>
Message-ID: <m3zoqxf5pk.fsf@atrus.jesus.cam.ac.uk>

"Michal Wallace (sabren)" <sabren at manifestation.com> writes:

> On Thu, 13 Apr 2000, Michal Wallace (sabren) wrote:
> 
> >    I know I can just as easily have them pass a reference to the
> > module along with it, but I'd be happier if I could just look at the
> > connection object and know which module it came from. Can I do this?
> 
> Um.. Okay....
> 
> for just about every module I've tried, I can do:
> 
> module = eval(dbc.__class__.__module__)
> 
> ... EXCEPT for ODBC.Windows..
> 
> >>> dbc
> <MS ODBC Manager Connection object at 76c3b0>
> >>> dbc.__class__
> Traceback (innermost last):
>   File "<stdin>", line 1, in ?
> AttributeError: __class__
> >>>
> 
> what gives?

Presumably dbc is not a Python class, but an extension type.

See what type(dbc).__name__ says; maybe you can special case for that?

Otherwise, I think you're hosed.  I'm not sure it's a good idea;
it sounds rather fragile to me.

Cheers,
M.

-- 
  ... but I guess there are some things that are so gross you just have
  to forget,  or it'll destroy something within you.  perl is the first
  such thing I have known.                 -- Erik Naggum, comp.lang.lisp


From vetler at ifi.uio.no  Sat Apr  1 08:59:25 2000
From: vetler at ifi.uio.no (Vetle Roeim)
Date: Sat, 1 Apr 2000 15:59:25 +0200 (MET DST)
Subject: Emacs Python mode
In-Reply-To: <20000401082513.B1353928@vislab.epa.gov>
Message-ID: <Pine.SOL.4.21.0004011557570.26669-100000@ganglot.ifi.uio.no>

on 2000-04-01, Randall Hopper wrote:

>  |I just had an idea. Using the Emacs python-mode, would it be possible
>  |to somehow hide the function-bodies?
>  |
>  |Sometimes it would be useful to get an overview of the functions
>  |provided by a class.
> 
> Your prayers have been answered.  See Ken Manheimer's outdent.el for Emacs
> (http://www.python.org/emacs/).  Collapsable Python blocks, block skips,
> etc.

sounds great! thanks :)
 

vr




From frankr at first-mind.de  Thu Apr 13 12:16:05 2000
From: frankr at first-mind.de (Frank Rotermund)
Date: Thu, 13 Apr 2000 18:16:05 +0200
Subject: Call by Reference / Value
Message-ID: <8d4rr7$t8k$1@newnews.ilk.de>

Hello

I'm building an interface between Python and a library written in C by using
Swig.
My problem is, that many methods of the library are changing the parameters
that
are given to them.
Is there a way, to tell Python that it should work on with the changed
values after
a method-call.

Thanks in advance!




From markx.daley at intel.com  Tue Apr 25 13:45:38 2000
From: markx.daley at intel.com (Daley, MarkX)
Date: Tue, 25 Apr 2000 10:45:38 -0700
Subject: mxODBC error trapping
Message-ID: <A63AFB20111ED311AC5500A0C96B7BF5015F0F4B@orsmsx36.jf.intel.com>

I am accessing databases using the notes by John Dell'Aquila for ODBC using
Python Database API.  It works fine, but now I am trying to set up error
trapping in the event that a connection fails to be made.  Here is the basic
code structure:

def collect():
	"Routine to query all servers for all names for time period defined
in timer."
	# Define some variables
	for index in servername:
		try:
			# Big routine to collect data and write to file
		except dbi.operation-error:
			servername.append(index)

Here is the error message I get:

Traceback (innermost last):
  File "<pyshell#2>", line 1, in ?
    collect.collect()
  File "C:\PROGRA~1\Python\collect.py", line 50, in collect
    except dbi.operation-error:
AttributeError: operation

If I change the exception to just OperationError, I get this:

Traceback (innermost last):
  File "<pyshell#1>", line 1, in ?
    collect.collect()
  File "C:\PROGRA~1\Python\collect.py", line 50, in collect
    except OperationalError:
NameError: OperationalError

I guess I don't know enough about exceptions to trap them properly.  Any
hints?

- Mark




From martin.franklin at waii.com  Thu Apr 13 06:24:53 2000
From: martin.franklin at waii.com (Martin Franklin)
Date: Thu, 13 Apr 2000 11:24:53 +0100
Subject: Why is if a in dict.keys() so slow.....
Message-ID: <38F5A075.CDFA32D9@bedford.waii.com>

Hi,

Just thought i'd throw this out there....

This is very slow:-


if split[0] in map_dict.keys():
    print 'found it!', map_dict[split[0]]
else:
    print 'not found'

Compared to:-


try:
    print 'found it!',map_dict[split[0]] 
 
except:
    print 'Not found'
        
The dictionary had about 40000 key:value pairs and I needed to 
look through it 250000 times.....

Question is why?


Thanks,
Martin.



From mwh21 at cam.ac.uk  Sat Apr  1 08:33:07 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 01 Apr 2000 14:33:07 +0100
Subject: Emacs Python mode
References: <yayk8iii0ex.fsf@ganglot.ifi.uio.no>
Message-ID: <m3zorendyk.fsf@atrus.jesus.cam.ac.uk>

Vetle Roeim <vetler at ifi.uio.no> writes:

> I just had an idea. Using the Emacs python-mode, would it be possible
> to somehow hide the function-bodies?

Try "M-8 C-x $", which isn't the easiest thing to type, but it works
reasonably well.

It's worth readong through the extended help on python-mode (type "C-c
C-h m" or "C-c ?"), which I did yesterday and realised that there were
lots of shortcuts for handy things I never knew existed.  Of course I
now can't remember what the key combinations for any of these cool
things were.

Cheers,
M.

-- 
well, take it from an old hand: the only reason it would be easier
to program in C is that you can't easily express complex  problems
in C, so you don't.                 -- Erik Naggum, comp.lang.lisp


From ivanlan at callware.com  Tue Apr  4 21:43:57 2000
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Tue, 04 Apr 2000 19:43:57 -0600
Subject: Starving for an Advanced Python Book
Message-ID: <38EA9A5D.F35B3E74@callware.com>

Hi All--

Bjorn Pettersen wrote:
> 

[snip]

> At $12.00 I would say the book is a steal and contains a wealth of
> information no serious Python programmer should be without :-)  And if
> you're like me, you'll print it out, bind it at Kinko's, and place it
> half way between your Linux and win2k boxes <wink>.
> 

I agree.  Even without Linux support and with the stupid licensing (it
ought to be per person, not per computer--that means I can't legally use
it at home), it's a steal.  I paid $16 to have it double-sided and bound
at Kinko's, and consider it a small price to pay for an excellent
resource.  My only quibble with the printed version--sucky index.

<running-and-hiding-from-/F>-ly y'rs,
Ivan;-) 
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
http://www.pauahtun.org and
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
Author:  Teach Yourself Python in 24 Hours



From moshez at math.huji.ac.il  Sat Apr 29 10:24:59 2000
From: moshez at math.huji.ac.il (Moshe Zadka)
Date: Sat, 29 Apr 2000 17:24:59 +0300 (IDT)
Subject: GUI developping with Python (Any tutorial )
In-Reply-To: <Pine.A41.3.96.1000429092757.42286A-100000@aix2.uottawa.ca>
Message-ID: <Pine.GSO.4.10.10004291723130.26644-100000@sundial>

On Sat, 29 Apr 2000 abouf066 at aix2.uottawa.ca wrote:

> Greetings, 
> 
> 	I wonder if anybody can point me to a site where I could find the
> starting steps to develop a GUI using python. 

The first step would be to decided what GUI toolkit to use, which in turn
depends on your platform (UNIX? Windows? Mac? Multi-platform), available
tools (C compilers, C++ compilers) and other things. If you indicate what
you need to develop GUI for, we might be able to help you choose a GUI
toolkit, and find a tutorial for it.
--
Moshe Zadka <mzadka at geocities.com>. 
http://www.oreilly.com/news/prescod_0300.html
http://www.linux.org.il -- we put the penguin in .com




From tjreedy at udel.edu  Thu Apr 27 23:16:58 2000
From: tjreedy at udel.edu (Terry Reedy)
Date: Thu, 27 Apr 2000 23:16:58 -0400
Subject: The Simple Economics of Open Source
References: <1255272357-66924550@hypernet.com>
Message-ID: <8eavgs$ker$1@news.udel.edu>

"Gordon McMillan" <gmcm at hypernet.com> wrote in message news:1255272357-

> One of those open source delusions is that being open source
> somehow makes a difference to the end consumer. Bunk.

Whether most consumers yet know it or not, there is all the difference in
the world.
The closed source model gives us spyware (Trojon-horse ware) like
RealJukebox and many others
(and, if rumor were true, Windows -- the possibility is always there).
 On ad-related spy engine has apparently been used in hundreds of 'free'
(closed source) programs
(I believe www.grc.com is URL with info on this).
Fom a privacy/security viewpoint, putting closed-source software on a
networked computer with anything inportant on it is a bit insane.
 I predict that in a few years, more consumers will understand this.

> End  users don't read code, and never will.

Does not matter for privacy/security checking see above and below.

> Opening the source
> benefits other developers, who in turn will almost certainly
> benefit you by making the code better, or integrating with
> other software - thus indirectly benefitting users.

The other benefit is that responsible people can check for funny business.
This is much easier with open source than with having to reverse engineer
the behavior of closed-source stuff.

Terry J. Reedy





From raffael at mediaone.net  Wed Apr 26 20:57:57 2000
From: raffael at mediaone.net (Raffael Cavallaro)
Date: Thu, 27 Apr 2000 00:57:57 GMT
Subject: The Simple Economics of Open Source
References: <B529350E.51B9%raffael@mediaone.net> <390534BB.3D6CEBDF@libc.org>
Message-ID: <raffael-01CD89.20575826042000@news.ne.mediaone.net>

In article <390534BB.3D6CEBDF at libc.org>, Bill Anderson <bill at libc.org> 
wrote:

>Unless you are redefining
>'commodity' (possibly by your 'known to many' remark) to fit whatever is
>provided as counter evidence.

A commodity is that which is available in large quantities, from many 
vendors offering essentially interchangeable product. This makes 
commodities inexpensive, hence the term "at commodity prices" meaning, 
"inexpensive."

In intellectual property then, a commodity is that which is known to 
many and not under copyright or patent. Because of the near zero cost of 
copying, the large quantity availability is a non-issue with software. 
What matters is whether or not it is known to many, because you then 
have many potential suppliers. If it is only known to a few, then the 
number of potential suppliers is few, and prices are not like those of a 
commodity market.

Examples: The JPEG standard is known to many, and not under copyright or 
patent. Therfore, software that compresses/decompresses JPEGs is 
available from many suppliers (again, the ease of copying electronic 
media moots the quantity issue). Because of the fact that JPEG 
compression/decompression is known to many, software to perform these 
functions is plentiful and cheap (i.e., often free).

The Microsoft Word (.doc) format is not known to many (really, only to 
MS programmers) and binaries to read/display/edit data in this format 
are under copyright. Although others have tried to reverse engineer this 
ever changing format, they have not been 100% successful. The result is 
that if you want to be certain that you can read/display/edit a document 
in this format, you have only one supplier, namely, Microsoft. Because 
of this, software to perform these functions is not a commodity item, 
and is quite expensive.

In short, I'm not "redifining" the word "commodity." I'm using it as it 
has always been used. I think that many open source advocates have 
failed to apply simple logic to the economics of open source, believing 
that somehow, the internet, or software, is immune from simple laws of 
supply and demand. They are not.

Ralph

-- 

Raffael Cavallaro, Ph.D.
raffael at mediaone.net


From hniksic at iskon.hr  Thu Apr 13 03:59:02 2000
From: hniksic at iskon.hr (Hrvoje Niksic)
Date: 13 Apr 2000 09:59:02 +0200
Subject: Python/Perl Popularity (Re: A Mountain of Perl...)
References: <200004102111.HAA03380@envy.fulcrum.com.au> <8d0pcq$7ti$1@nnrp1.deja.com> <dpg0sr37fw.fsf@mraz.iskon.hr> <etdn1mz8drz.fsf@w20-575-109.mit.edu>
Message-ID: <dp66tmv3cp.fsf@mraz.iskon.hr>

Alex <alex at somewhere.round.here> writes:

> > >  -- How can I browse the Python documentation without firing
> > >     up a web browser?
> > 
> > By using Info.  That's what I do, all the time.  Info has the added
> > nicety of being able to search through the whole documentation for a
> > keyword.
> 
> I love to use info, too.  Unfortunately, I can't get the latest
> version of the documentation in that format, so I don't know whether
> I'm going to be able to use it for much longer.

Well, I don't know.  I'm using the Debian package `python-doc' which
contains info, so I guess it's possible to create it.


From bjorn at roguewave.com  Mon Apr 24 12:00:34 2000
From: bjorn at roguewave.com (Bjorn Pettersen)
Date: Mon, 24 Apr 2000 10:00:34 -0600
Subject: Simple list.append() question
References: <meh9-5ED8CD.11192424042000@news.cit.cornell.edu> <ZHZM4.360$qg7.2949@news.uswest.net>
Message-ID: <39046FA2.603EAFE5@roguewave.com>

Jeff Massung wrote:
> 
> I'm just learning, but my guess is this:
> 
> >>> l=[[]]*3
> >>> l
> [[], [], []]
> >>> l[0].append(3)
> >>> l
> [[3], [3], [3]]
> >>> id(l[0])
> 8539088
> >>> id(l[1])
> 8539088
> >>> id(l[2])
> 8539088
> >>>
> 
> The *3 just made three lists pointing to the same place. I don't know how to
> get what you want tho ;) someone else can enlighten me, too :).

One way is:

	x = map(lambda x:[], [[]]*3)

although there has to be a prettier way...

-- bjorn



From jtschirr at engineering.uiowa.edu  Sat Apr 15 01:03:43 2000
From: jtschirr at engineering.uiowa.edu (Juerg Tschirren)
Date: Sat, 15 Apr 2000 05:03:43 GMT
Subject: Imaging Library & Numerical Python
Message-ID: <38F7F865.9C86399C@engineering.uiowa.edu>

Does anybody have experience with getting the Imaging Library and
Numerical Python working together? What I would like to do is importing
an imaging with the Imaging Library and then using functions from both
-- the Imaging Library and Numerical Python -- on this image.

Juerg


From effbot at telia.com  Mon Apr 10 12:04:14 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Mon, 10 Apr 2000 16:04:14 GMT
Subject: comparing strings and ints
References: <8css5t$tti$1@merrimack.Dartmouth.EDU>
Message-ID: <2YmI4.366$rc9.188014080@newsb.telia.net>

Mark H.H. Montague wrote:
> Does anyone know why it is not a type error to compare a string and an
> int with '>' ?  For example, I just spent over an hour debugging code
> that essentially did this:
>
> >>> 45 > '0'
> 0
>
> Of course, the '0' was a variable read in from a file, which I forgot
> to convert to a string.  Bad old perl habits.  But I'm having a hard
> time imagining a situation where someone would be intentionally and
> meaningfully comparing a string and an int.
>
> Any ideas?

why not read the fine documentation:
http://www.python.org/doc/current/ref/comparisons.html

which explains that:

    Objects of different types always compare unequal,
    and are ordered consistently but arbitrarily.

    This unusual definition of comparison was used to
    simplify the definition of operations like sorting and
    the "in" and "not in" operators.

and goes on to say:

    In the future, the comparison rules for objects of
    different types are likely to change.

</F>




From niels at endea.demon.nl  Mon Apr 17 09:09:22 2000
From: niels at endea.demon.nl (Niels Diepeveen)
Date: Mon, 17 Apr 2000 15:09:22 +0200
Subject: an intern-like memory saver
References: <8d5ing$heb$1@nnrp1.deja.com> <38F7318F.AE4F0EE5@endea.demon.nl> <8d8fe0$nr5$1@nnrp1.deja.com>
Message-ID: <38FB0D02.49B6F3CD@endea.demon.nl>


sjmachin at lexicon.net schreef:
> 
> I tried "intern" and read the source code for it. Speed is not a
> consideration until I run out of real memory and start swapping to
> disk. My objective is to reduce memory usage so that that doesn't
> happen. "intern" has its *own* hidden dictionary; if I use it, it may
> even *increase* my memory usage. 

If the number of distinct strings weren't large, that would not be a
problem, but if it is, it is.

> I already have my own dictionary with
> all those strings as keys, so I save the overhead part of the intern
> dictionary. For what it's worth my key_ref method works with any
> objects that can be dictionary keys, not just strings. What are the
> other ways you say one can do very much the same thing in Python?

Variations on the same theme. All of them will use more memory than what
you were proposing.

-- 
Niels Diepeveen
Endea automatisering



From m.faassen at vet.uu.nl  Tue Apr 18 12:26:28 2000
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 18 Apr 2000 16:26:28 GMT
Subject: Using python on the web
References: <MPG.1365adcfa8470eef9896aa@news>
Message-ID: <8di2bk$l1s$1@newshost.accu.uu.nl>

Charley Horse <someone at somewhere.com> wrote:
> I've just started fooling with python. I've looked around the python 
> site and a number of the other sites referenced there and am a bit 
> puzzled. Umm... how does one go about using python for web applications, 
> especially database driven apps? Almost everything deals with python 
> just as a language, as C++ would be.

That's because Python is a language. Python isn't a web-specific language,
after all. 

> Not much about the web specifically.
> I was a bit shocked to find the only direct refs on about 
> this at python.org are entwined with CGI (unless I missed it).

Why is CGI so shocking? It's used quite a bit, after all? You can write
a database driven web application using Python with CGI. CGI from Python
isn't hard either.

> I have no exposure to CGI, but rather to php, CF, and ASP, so python has
> gone from looking very attractive to looking... well, like a puzzle.

I'm not sure why your exposure to PHP, ColdFusion or ASP instead of CGI
would turn Python into a puzzle.

There _is_ ASP support for Python, enabling you to use Python as a
ASP language. Others will have to explain to you how and where to look
for information, as I've never used it.

> I know about Zope, but am interested in how one uses python on the web just
> by itself.

Zope would be the equivalent of ColdFusion for Python, as it's a web
application server.

Besides using Python with CGI or ASP, you can use Zope's ZPublisher
(formerly Bobo) independently to publish arbitrary Python objects to the
web. I've never used this, and the documentation may be scarce, but it
looks quite neat. You may want to look that up.

Database integration in Python isn't bad either, look at the database
topic guide:

http://www.python.org/topics/database/

And look at the web topic guide if you haven't seen it:

http://www.python.org/topics/web/

The vaults of Parnassus are also a good resource, here is for
instance a link to its web category:

http://www.vex.net/~x/parnassus/apyllo.py/812237977.127386987

Also look at the 'info' category of the Vaults.

Good luck!

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From scherbi at bam.com  Mon Apr 24 12:11:25 2000
From: scherbi at bam.com (Bill Scherer)
Date: Mon, 24 Apr 2000 12:11:25 -0400
Subject: ftplib sendcmd
Message-ID: <3904722D.12EE9927@bam.com>

Howdy -

I'm having trouble with ftplib's sendcmd method (Python 1.5.2 on RH
linux 6.1):

>>> ftp = ftplib.FTP(host, user, password)
>>> ftp.sendcmd('umask 002')
Traceback (innermost last):
  File "<stdin>", line 1, in ?
  File "/usr/local/lib/python1.5/ftplib.py", line 228, in sendcmd
    return self.getresp()
  File "/usr/local/lib/python1.5/ftplib.py", line 201, in getresp
    raise error_perm, resp
ftplib.error_perm: 500 'UMASK 002': command not understood.
>>>

The umask command does, of course, work using a standard ftp client to
the same host.

Other commands, chmod, etc, suffer equal fate.

Is this not what sendcmd is for? What am I doing wrong?  Thanks!

--
William K. Scherer
Sr. Member of Applications Staff
Verizon Wireless






From mbel44 at dial.pipex.net  Wed Apr  5 07:12:15 2000
From: mbel44 at dial.pipex.net (Toby Dickenson)
Date: Wed, 05 Apr 2000 12:12:15 +0100
Subject: Python 1.6 alpha 1 released
References: <Pine.GSO.4.10.10004040035420.21084-100000@sundial> <5E877FD7637AB46E.DABBED9CF3C1D64B.CDDDC05AE7374A78@lp.airnews.net> <5l7leern5v.fsf@eric.cnri.reston.va.us> <8cdg92$j0m$1@nntp6.u.washington.edu> <8cdig3$s7f$1@slb6.atl.mindspring.net>
Message-ID: <fp7messqdb29oirr9j5uktgi22ic6od5ab@4ax.com>

aahz at netcom.com (Aahz Maruch) wrote:

>In article <8cdg92$j0m$1 at nntp6.u.washington.edu>,
>Russell E. Owen <owen at astroNOJNK.washington.edu.invalid> wrote:
>>
>>But tuples seem a less than wonderful choice to me. Protocol flexibility 
>>presumably requires each protocol to have its own tuple with its own 
>>numer of fields in its own order. That's a lot to remember and try not 
>>to get mixed up.
>
>I second this.  I'd be happy with either dict or class, with a *slight*
>preference for a class.

Does the single-argument connect require a 2-tuple, or just a two
element sequence?

If a sequence will do, then you can use an address class today.


Toby Dickenson
tdickenson at geminidataloggers.com


From embed at geocities.com  Thu Apr 20 11:24:20 2000
From: embed at geocities.com (Warren Postma)
Date: Thu, 20 Apr 2000 11:24:20 -0400
Subject: Python Parsers Comparison
References: <20000420093945.A624723@vislab.epa.gov>
Message-ID: <FlFL4.9921$HG1.265980@nnrp1.uunet.ca>

> For example, a comma-separated list of nested vectors would be a good
example:
>
>     [ 1, 2 ], [ 3, 4 ]
>     [ [ 1,2 ],[ 3,4 ] ], [ 5.6 ]
>     etc.
>

Wherever I am able to do so, I use eval() and design my grammar to be
basically PYTHON.  Very nice scheme, and you can even use 'restricted
execution' if safety is a concern.

In the above cases:

safedict={"__builtins__":{},"NULL":None} # your own symbols, and constants
str = '[1,2,],[1,2],[[1,2],[1,2]],[NULL,NULL]'
x = eval("["+str+"]",safedict,safedict)
print x

If you are in control of the grammar, and you are inventing it, why not
stick with python, if on the other hand, the grammar is something you're
stuck with 'as-is', and it's not a proper subset of some Python construct,
then all bets are off.

I certainly WOULD like to see some tutorial material on constructing such
parsers, because I don't know the first thing about them. We studied this
stuff in undergrad C.S., but it was trivial stuff (BNF grammars, some simple
parsing code techniques in C).  We never did learn the Nifty Hardcode Stuff.

Warren




From echuck at mindspring.com  Thu Apr 20 00:23:28 2000
From: echuck at mindspring.com (Chuck Esterbrook)
Date: Thu, 20 Apr 2000 00:23:28 -0400
Subject: [ANNOUNCE] Webware for Python 0.1
References: <5650A1190E4FD111BC7E0000F8034D26A0F40D@huina.oceanic.com>
Message-ID: <38FE863F.BBCC8399@mindspring.com>


Thanks for the report. I tested before hand, but apparently not well enough.

It should be fixed now.

-Chuck


Doug Stanfield wrote:

> The following intrigues me but its not probably the best endorsement.  When
> I hit your published link I get:
>
> ERROR
>
> Traceback (innermost last):
>   File "MainPage.py", line 55, in main
>     _main()
>   File "MainPage.py", line 49, in _main
>     inc_counter()
>   File "MainPage.py", line 23, in inc_counter
>     f.write(repr(counter+1))
> NameError: counter
>
> But I guess you're at V0.1 so proving the error reporting works is a good
> thing. :-)
>
> This was from a redirect to http://writewithme.com/Webware/ by the way, so
> maybe therin lies the problem.  I am still interested in looking at the
> package.
>
> -Doug-
>
> > -----Original Message-----
> > From: Chuck Esterbrook [mailto:echuck at mindspring.com]
> > Sent: Wednesday, April 19, 2000 5:52 PM
> > To: python-list at python.org
> > Subject: [ANNOUNCE] Webware for Python 0.1
> >
> >
> >
> > Webware for Python is a definitive suite of Python software
> > components for use in web development. Each component has a
> > focused purpose and can often be used on it's own. At the
> > same time, the components are designed to work together where
> > appropriate. As a user of Webware, you can bite off as much
> > or as little as you want to chew.
> >
> > You can find more and download at:
> >
> >       http://www.writers-webshop.com/Webware
> >
> >
> > -Chuck
> > --
> > http://www.python.org/mailman/listinfo/python-list
> >

From chafik at nevrax.com  Wed Apr  5 04:48:03 2000
From: chafik at nevrax.com (Sameh chafik pro)
Date: Wed, 5 Apr 2000 10:48:03 +0200
Subject: python16_d.lib, python16_d.dll python15_d.dll
Message-ID: <013101bf9edb$a76984c0$1001a8c0@nevrax.net>


----- Original Message ----- 
From: Sameh chafik pro 
To: Python mail list 
Sent: Wednesday, April 05, 2000 10:35 AM
Subject: (no subject)


Hello evry body,

Are there any way to have the python16_d.lib and .dll, or the python15_d.dll ?????
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20000405/10744e85/attachment.html>

From mwh21 at cam.ac.uk  Sun Apr  2 21:12:28 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 03 Apr 2000 02:12:28 +0100
Subject: getting a key without an echo and Enter
References: <Q0SF4.8$tu.536@news.hananet.net>
Message-ID: <m3vh20hts3.fsf@atrus.jesus.cam.ac.uk>

"Gang Seong Lee" <gslee111 at daisy.kwangwoon.ac.kr> writes:

> Hi,
> 
> How can I get a key input without having it echoed.
> The following statement echoes and waits until I press Enter key.
> 
> ch = sys.stdin.read(1)
> 
> Can I get the key as soon as I press it without an echo. (No Enter key)
> 
> Thanks in advance

Heh.  I was trying to do this the other day; I came up with the rather
hairy & unix specific:

import termios,TERMIOS,os,select,string,sys

def read1():
    oldattr = termios.tcgetattr(0)
    try:
        attr = termios.tcgetattr(0)
        attr[2] = (attr[2] & ~TERMIOS.NLDLY) | TERMIOS.NL0
        attr[3] = attr[3] & ~(TERMIOS.ICANON|TERMIOS.ECHO)
        termios.tcsetattr(0,TERMIOS.TCSANOW,attr)
        return os.read(0,1)
    finally:
        termios.tcsetattr(0,TERMIOS.TCSANOW,oldattr)

On windows, I think some variant of

import msvcrt
msvcrt.getch()

do the job, but I can't test that here.

HTH,
M.

-- 
well, take it from an old hand: the only reason it would be easier
to program in C is that you can't easily express complex  problems
in C, so you don't.                 -- Erik Naggum, comp.lang.lisp


From mal at lemburg.com  Tue Apr  4 05:43:46 2000
From: mal at lemburg.com (M.-A. Lemburg)
Date: Tue, 04 Apr 2000 11:43:46 +0200
Subject: Starving for an Advanced Python Book
References: <20000404040941.7766.qmail@web2106.mail.yahoo.com>
Message-ID: <38E9B952.A56C3877@lemburg.com>

lewst wrote:
> 
> Having finished "Learning Python", I'm now looking for an advanced
> Python text.  ORA's "Programming Python" is a natural choice, but I
> can't bear buying this outdated book when the 2nd edition is due out
> in a few months (I'm dying here Lutz, hurry it up already!).  However,
> I really need something in the interim period.  The library reference
> manual is fine but usage is difficult for me to figure out sometimes
> because there are very few examples included.
> 
> Does anyone have any suggestions?  I suppose I'm looking for a text
> that has examples of how to solve various kinds of programming
> problems in Python and also examples of code showing how the library
> and various modules actually work.  How is the "Python Essential
> Reference" on examples?  Is it just a rebinding of the python.org
> library reference

Fredrik Lundh has compiled such a cookbook. I don't have the URL
handy, but I'm sure you'll find it on their www.pythonware.com
site.

-- 
Marc-Andre Lemburg
______________________________________________________________________
Business:                                      http://www.lemburg.com/
Python Pages:                           http://www.lemburg.com/python/





From warlock at eskimo.com  Mon Apr  3 00:36:02 2000
From: warlock at eskimo.com (Jim Richardson)
Date: Mon, 03 Apr 2000 04:36:02 GMT
Subject: Just doing local file testing
References: <38E796BC.23191B2@regnoc.com> <00040214141502.04535@quadra.teleo.net>  <38E7D6DE.EA09C891@regnoc.com>
Message-ID: <slrn8eg8ad.fj2.warlock@raven.myth>

On Sun, 02 Apr 2000 23:23:59 GMT, 
 Jim Conger, in the persona of <jimc at regnoc.com>,
 brought forth the following words...:

><!doctype html public "-//w3c//dtd html 4.0 transitional//en">
><html>
>I'm not even to the point of using Apache yet - I'm just to launch the
>script locally from within the browser by having the form element point
>to a script on disk.&nbsp; In other words, this is just a development environment
>using only local files.


This won't work, as it is the webserver that decides that the .py is a script
and passes it to python for execution. In local mode, your browser simply
treats the script as another text file, and an improperly formatted (in html 
terms) on at that.

-- 
Jim Richardson
	Anarchist, pagan and proud of it
WWW.eskimo.com/~warlock
	Linux, because life's too short for a buggy OS.



From jbauer at rubic.com  Tue Apr 18 18:09:50 2000
From: jbauer at rubic.com (Jeff Bauer)
Date: Tue, 18 Apr 2000 17:09:50 -0500
Subject: Date class?
References: <8dikk2$p4s@flatland.dimensional.com> <etdr9c3ul5q.fsf@w20-575-109.mit.edu>
Message-ID: <38FCDD2E.A1C2CF6E@rubic.com>

Alex,

Thanks for the kind words.  I emailed Jim a copy
of normalDate.py, since the Starship is still down.

Jeff Bauer
Rubicon Research

Alex wrote:
> You could try normalDate.  I have found that fairly useful, 
> once or twice.



From guido at cnri.reston.va.us  Tue Apr  4 09:49:48 2000
From: guido at cnri.reston.va.us (Guido van Rossum)
Date: 04 Apr 2000 09:49:48 -0400
Subject: Python 1.6 alpha 1 released
References: <Pine.GSO.4.10.10004040035420.21084-100000@sundial> <5E877FD7637AB46E.DABBED9CF3C1D64B.CDDDC05AE7374A78@lp.airnews.net>
Message-ID: <5l7leern5v.fsf@eric.cnri.reston.va.us>

tseaver at starbase.neosoft.com (Tres Seaver) writes:

> The point of this debate is that classifying it as a *bug* is, well,
> debateable.  Years of socket programming experience makes:
> 
>   s.connect( host, port )
> 
> seem like a perfectly reasonable, sane interface;  disallowing it
> may preserve "orthagonality," or some other academic purity, but
> it *breaks existing applications coded to the published interface*,
> with no appreciable real-world benefit (yes, I realize that the published
> interface is inconsistent;  I do not accept the deprecation of the
> examples as a reasonable argument.)  Consider that *every one* of the
> protocol modules which offers a connect() method uses a two-argument
> signature for it:  ftplib, gopherlib, httplib, imaplib, nntplib, poplib,
> smtplib, telnetlib, urllib.  And they all *use* the two-argument connect
> when invoking their underlying socket!  No Python module in the standard
> library uses the single-argument connect, period.
> 
> This is *not* the same case as the list.append() one, where the
> published interface was itself consistent -- I was quite surprised to
> learn just this spring that multi-argument append was even possible.
> But the socket module is one of the first I learned in Python, and I
> have consistenly used the two-argument connect.  Constructing a spurious
> "address" object (which has no behavior, and exists only to be torn
> apart inside the implementation) seems a foolish consistency, beyond
> doubt.
> 
> Rough-consensus-and-running-code-wins-every-time'ly

Before more people mail this to me, let me apologize.  I didn't
realize that there were so many examples of the 2-argument form.  I
don't want to break that much code, so I'll revert the decision in the
next alpha release.

However, I'll do so for purely backwards compatible reasons, and I
will continue to press for the tuple address as the correct form.
This has nothing to do with academic purity.  It is because sockets
are, by their very nature, address family neutral, and not all address
families use a host and a port to identify a socket.  For example,
Unix sockets use just a filename.

The argument that gopherlib c.s. define a 2-arg connect() doesn't mean
that socket should, to: those are all Internet specific protocols,
where host+port is the norm; the lower-level socket API must support
other address families.

The argument that the address object is spurious doesn't hold water.
There are many places where it is convenient to hold an address in a
single variable, and a tuple is the obvious way to represent this.
(The alternative, a "host:port" string, seems too Perlish to me.)
Note also that recvfrom() returns the address as a tuple (if it is an
IP address) and sendto() requires the address to be passed as a tuple
-- the canonical form of an address is clearly a tuple.

Repeat after me: SOCKETS ARE NOT JUST FOR IP CONNECTIONS.

But I will fix the code -- it's the only practical thing to do.

--Guido van Rossum (home page: http://www.python.org/~guido/)


From mhammond at skippinet.com.au  Mon Apr 24 22:31:25 2000
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Tue, 25 Apr 2000 02:31:25 GMT
Subject: Pythonwin com question
References: <8dqver$l4v$1@nnrp1.deja.com> <8e2sfb$r6e$1@nnrp1.deja.com>
Message-ID: <1s7N4.3219$v85.20999@news-server.bigpond.net.au>

"Sam Schulenburg" <samschul at pacbell.net> wrote in message
news:8e2sfb$r6e$1 at nnrp1.deja.com...

>       ## DoEvents  # Wait for detector to finish, This has no
>                    # equivalent in Pythonwin

Actually, you _may_ need to call pythoncom.PumpWaitingMessages() here (and
it can never really hurt :-)

>          s = 1     # This gives the while loop something to do
>    #### This is where everything falls apart
>    #### theFrame is not the correct datatype
>    dx.GetFrame(1,theFrame)  # Retreive one image
>
> How do I define the correct Varient type in Python?

You shouldnt need to.

Try saying:
    theFrame = dx.GetFrame(1)

[Or try passing None as the second param]

And magically it should all work.  If it doesnt, look into running
"makepy" over your object and try again.  If it still doesnt work, post
the code makepy generated for the GetFrame() method and the specific
traceback you see...

Mark.




From ian_maurer at my-deja.com  Wed Apr 12 11:18:27 2000
From: ian_maurer at my-deja.com (Ian Maurer)
Date: Wed, 12 Apr 2000 15:18:27 GMT
Subject: making a .exe
References: <8d1u0u$ikb$1@porthos.nl.uu.net>
Message-ID: <8d243j$mro$1@nnrp1.deja.com>

In article <8d1u0u$ikb$1 at porthos.nl.uu.net>,
  "Soundwave" <maybe at lucky.com> wrote:
> How can I make an executable file from the *.py files ?
>
> thanx
>
>

http://www.mcmillan-inc.com/install1.html

--
Ian Maurer




Sent via Deja.com http://www.deja.com/
Before you buy.


From jhauser at ifm.uni-kiel.de  Fri Apr 14 04:22:28 2000
From: jhauser at ifm.uni-kiel.de (Janko Hauser)
Date: 14 Apr 2000 10:22:28 +0200
Subject: Questions on 1.6a2's string methods
References: <38F6C904.76B538E5@concentric.net>
Message-ID: <87snwp3xdn.fsf@ifm.uni-kiel.de>

Manus Hand <mjhand at concentric.net> writes:

> 3.  Are the classlike standard types (list, dictionary, and now
>     string) equipped with a __dict__ attribute?  I can see the names
>     of all functions supported by a user-defined class by saying
>     className.__dict__.keys(), but I cannot see the list of methods
>     for the string type (at least not in the same way).  Thus my
>     need to ask silly questions like #1 above (maybe capwords is
>     there by some other name??)
> 
 I haven't tried it with the new string type but you can look up the
 methods bye the __methods__ attribute.

>>> a=[1,2]
>>> a.__methods__
['append', 'count', 'extend', 'index', 'insert', 'pop', 'remove',
'reverse', 'sort']
>>> b={1:''}
>>> b.__methods__
['clear', 'copy', 'get', 'has_key', 'items', 'keys', 'update', 'values']


HTH
__Janko
-- 
  Institut fuer Meereskunde             phone: 49-431-597 3989
  Dept. Theoretical Oceanography        fax  : 49-431-565876
  Duesternbrooker Weg 20                email: jhauser at ifm.uni-kiel.de
  24105 Kiel, Germany


From zorro at zipzap.ch  Fri Apr 14 08:41:11 2000
From: zorro at zipzap.ch (Boris Borcic)
Date: Fri, 14 Apr 2000 14:41:11 +0200
Subject: Python Starship Grounded??
References: <38F6B2DF.940E7C23@concentric.net>
Message-ID: <38F711E7.3FA0CA58@zipzap.ch>

Manus Hand wrote:
> 
> Anyone here able to do anything about the illness that seems to have
> come over starship.python.net?

I second the question.

Boris Borcic
--
"Money was invented to enslave arithmetics"


From fredrik at pythonware.com  Thu Apr 20 08:23:23 2000
From: fredrik at pythonware.com (fredrik at pythonware.com)
Date: Thu, 20 Apr 2000 12:23:23 GMT
Subject: socket.SO_REUSEADDR
References: <38FEEA44.6C05E2B0@inka.de>
Message-ID: <8dmsrd$8ta$1@nnrp1.deja.com>

Michael Str?der wrote:
> For avoiding problems with errors like
>
>   socket.error: (98, 'Address already in use.')
>
> during immediate restart of my web server I'm following the hint to
> set the socket option SO_REUSEADDR. But I did not succeed. The
> server starts and works. When I stop it and try to restart
> immediately the above message appears again. I had a short look at
> man 7 socket and man setsockopt but this was a little bit too much
> for me...

that getproto stuff looks strange.  try this instead:

   self.socket.setsockopt(socket.SOL_SOCKET, socket.SO_REUSEADDR, 1)

</F>


Sent via Deja.com http://www.deja.com/
Before you buy.


From nick at webthing.com  Thu Apr 20 17:30:04 2000
From: nick at webthing.com (Nick Kew)
Date: Thu, 20 Apr 2000 21:30:04 +0000
Subject: Using python on the web
References: <MPG.1365adcfa8470eef9896aa@news>
    <MPG.136631472f892a119896ac@news>
    <_W2L4.986$GK5.20898@news2-win.server.ntlworld.com>
    <MPG.13665ef93eedbda29896ad@news>
    <BC4B61746187C623.3FA0E7D58DF4EA7E.ECC1D97F577DEDE5@lp.airnews.net>
Message-ID: <sssnd8.b91.ln@sleipnir.webthing.com>

In article <BC4B61746187C623.3FA0E7D58DF4EA7E.ECC1D97F577DEDE5 at lp.airnews.net>,
	claird at starbase.neosoft.com (Cameron Laird) writes:

> CGI has been dead for at least five years.

:-)

> I recognize no one here is disputing the proposition.  It's
> just struck me recently how pervasive the propaganda on this
> score is.  CGI has an *abyssmal* popular reputation.

Several reasons.  Like the number of non-programmers who dabble with it,
and produce something horrible that gives the impression of working.

But I think the main reason is, it's the standard every bandwagon and
marketing department has to knock.  M$ of course have worked hard to
undermine it, with severely broken implementations and "alternatives".
The JAVA folks push servlets.  And nobody in the open source community
has made any serious effort to stand up for it: indeed, rather the
opposite - with things like mod_perl stressing that it is Not CGI[1].

I have to confess to some measure of guilt.  I started work on an 
Apache module to implement non-forking CGI, eliminating the so-called
CGI overhead.  In test conditions it benchmarked very nicely against
standard CGI, static HTML pages, and mod_perl, but I never completed
it to an operational standard.  Well, maybe I'll resurrect it as a
getting-to-know-apache-2.0 exercise.

[1] Actually mod_perl _is_ CGI, in that it meets the CGI spec.  Though
of course it does a whole lot more too.

-- 
Nick Kew


From andres at corrada.com  Wed Apr 26 22:08:00 2000
From: andres at corrada.com (Andres Corrada)
Date: Wed, 26 Apr 2000 22:08:00 -0400
Subject: condense whitespace to single space?
References: <36NN4.31937$Xk2.119087@tor-nn1.netcom.ca>
Message-ID: <3907A100.4C8522E1@corrada.com>

k wrote:
> 
> How do I condense the sections of spaces in this string:
> 
> "abc              def         ghi"
> 
> into single spaces?
> 
> "abc def ghi"

One way to do this that avoids using regular expressions is:

stringParts = string.split( stringToSplit )
reconstitutedString = string.join( stringParts )
-- 
------------------------------------------------------
Andres Corrada-Emmanuel   Email: andres at corrada.com
------------------------------------------------------



From smalleys at gte.net  Mon Apr 24 17:09:58 2000
From: smalleys at gte.net (sue)
Date: Mon, 24 Apr 2000 21:09:58 GMT
Subject: pattern-based string expansion
References: <38F52C61.A5D72F89@gte.net> <vndog79ft1q.fsf@camelot-new.ccs.neu.edu> <8del87$5md$1@slb6.atl.mindspring.net> <38FE70DB.29A4B1A1@gte.net> <KsvL4.1559$Ez2.89485@typhoon1.gnilink.net> <8dndbe$dcu$1@slb6.atl.mindspring.net>
Message-ID: <3904B835.327E27AB@gte.net>

I've hacked RE engines before, starting with HPUX 7, using
regex.h.

Nick - Thank you for the source.  I'll try and work on it
this week.

Andrew - Python access to an RE engine.  My Ada parser will
love it.  Thanks!

Sue.


From bsb at winnegan.de  Fri Apr  7 13:06:09 2000
From: bsb at winnegan.de (Siggy Brentrup)
Date: 07 Apr 2000 19:06:09 +0200
Subject: Paid to work in Python? (was: Re: Choice Of Language)
In-Reply-To: ngps@madcap.dyndns.org's message of "7 Apr 2000 16:34:34 GMT"
References: <Uqvw4.2261$yL2.28868@typhoon.southeast.rr.com> <8c3ats$b5f$1@nntp.Stanford.EDU> <slrn8eact2.sq7.nobody@bambi.visi.com> <38e5f8ed.29659873@news.nikoma.de> <8cl2mq$971$1@mawar.singnet.com.sg>
Message-ID: <877le9suwu.fsf@orge.winnegan.de>

ngps at madcap.dyndns.org (Ng Pheng Siong) writes:

> According to Philip 'Yes, that's my address' Newton <nospam.newton at gmx.li>:
> > (The German term "eierlegende Wollmilchsau"[1] comes to mind.)
> >
> > [1] literally, "egg-laying wool-milk-sow [female pig]", i. e., something
> > that tries to do too much at once.
> 
> I once saw a German term that was translated to "improve-worse-ment."
> 
> Can some one provide the German word, please?

I guess you mean Ver-schlimm-besserung (remove dashes), but thats a worsening
correction. You won't find it in dictionaries though.

Siggy

-- 
Siggy Brentrup - bsb at winnegan.de - http://www.winnegan.de/
****** ceterum censeo javascriptum esse restrictam *******



From kcazabon at home.com  Mon Apr 17 23:44:52 2000
From: kcazabon at home.com (Kevin Cazabon)
Date: Tue, 18 Apr 2000 03:44:52 GMT
Subject: TKMessageBox, and TKINTER console?
References: <rNHvjEdhm5Pp-pn2-Y8HWP45V5PG0@d185d18f2.rochester.rr.com>
Message-ID: <USQK4.147145$Dv1.1584032@news1.rdc1.bc.home.com>

Hey Benjamin;

I think your problem is that the TKMESSAGEBOX is not meant to be a main
window, but a child of a parent window.  So, when you invoke the "Frame" for
it, it creates a Toplevel widget to go along with it.

Normally, your 'root' application is a Tkinter.Tk() widget, which you pack
all your other things into.  The "popup" would be a child of that window,
and both would be displayed on the screen.

If you create a non-toplevel style widget on it's own, such as a "Frame", it
will automatically create the Toplevel as a second window.  This is your
problem.

If you don't want the "parent" to be displayed, create it, but 'withdraw' it
before the mainloop() is called.  do this by calling x.withdraw()

However, if you're using the "Message" widget, just pack it in the parent Tk
window, as follows, instead of having it as a separate Frame.

########################
import Tkinter

a = Tkinter.Tk()

b = Tkinter.Message(a, text = "Hello, this is a Message Widget")
        # the first argument above specifies the 'parent' to pack it in.
b.pack()

a.mainloop()

########################
"Benjamin Schollnick" <junkster at nospam.rochester.rr.com> wrote in message
news:rNHvjEdhm5Pp-pn2-Y8HWP45V5PG0 at d185d18f2.rochester.rr.com...
> Folks,
>
> Crazy Question..... I'm starting to mess around with Tkinter, and
> for now at least, just trying to display "pop-up" dialogs.  I found
> the
> TKMESSAGEBOX unit, in the lib-tk directory, and that works great for
> my purposes.
>
> Except, for a small "appearance" problem.  I call the TKmessagebox
> routines, and their are TWO windows.  One is the pop-up I've called,
> and the
> other is the Tkinter "Console" (??).  Or at least that's the nickname,
> I've given it.
>
> I've tried a variety of different things to get rid of it...For
> example, calling
> junk= TK.tk() to a variable, and passing it as "master=junk", or as
> "parent=junk"
> in the code...
>
> I'm not sure if there is anyway to get rid of that darn console.
>
> Anyone got some suggestions for a really confused first time
> Tkinter'er?
>
> - Benjamin
> ====================================================
>           (Remove "NoSpam" to Email me)
> ====================================================
> Please feel free to copy any and or all of this sig.
> A little something for spam bots:
>
> root at localhost postmaster at localhost admin at localhost
> abuse at localhost postmaster at 127.0.0.1
>
> Chairman William Kennard: bkennard at fcc.gov
> Commissioner Susan Ness: sness at fcc.gov
> Commissioner Harold Furchtgott-Roth: hfurchtg at fcc.gov
> Commissioner Michael Powell: mpowell at fcc.gov
> Commissioner Gloria Tristani: gtristan at fcc.gov
> consumerline at ftc.gov
> fccinfo at fcc.gov
> ssegal at fcc.gov
>




From noreply at antispam.com  Sun Apr  2 12:35:29 2000
From: noreply at antispam.com (Brian Schmidt)
Date: Sun, 02 Apr 2000 16:35:29 GMT
Subject: www.python.org
Message-ID: <lFKF4.120$wo4.48486@paloalto-snr1.gtei.net>

Anyone else having problems getting into www.python.org the last couple of
days?




From m.faassen at vet.uu.nl  Sun Apr  2 18:47:18 2000
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 2 Apr 2000 22:47:18 GMT
Subject: New Features in Python 1.6
References: <200004011740.MAA04675@eric.cnri.reston.va.us> <gHrF4.16072$0o4.101062@iad-read.news.verio.net>
Message-ID: <8c8ilm$7fs$1@newshost.accu.uu.nl>

Tom Culliton <culliton at clark.net> wrote:
>>4. assignment to None now works

> Huhn?  What exactly changed here?  You can assign to None in current
> versions just like you can to any other name.  I thought the push was
> to _prevent_ this.

Wild, I never knew that the current behavior allowed assigning to None.
Quite confusing if you do it:

>>> [None, None]
[None, None]

(okay..)

>>> None = 3
>>> [None, None]
[3, 3]
>>> None
3

Never caused any bugs for me, but I could imagine that could be confusing...

It's still in the 1.6a too.

Regards,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From thomas at xs4all.net  Fri Apr  7 07:09:42 2000
From: thomas at xs4all.net (Thomas Wouters)
Date: Fri, 7 Apr 2000 13:09:42 +0200
Subject: Was Re: need wrapper for ssh/scp
In-Reply-To: <8civ14$rm7$1@hammer.msfc.nasa.gov>; from drew.mcdowell@msfc.nasa.gov on Thu, Apr 06, 2000 at 09:19:32PM +0000
References: <8civ14$rm7$1@hammer.msfc.nasa.gov>
Message-ID: <20000407130942.E13830@xs4all.nl>

On Thu, Apr 06, 2000 at 09:19:32PM +0000, Andrew McDowell wrote:

> I also working with ssh to execute some remote commands on different
> servers.  I'd previously been hobbling together this strange idea of
> generating an Expect script in python then running it. (*bleh*)  So once I
> saw was clued onto the pty module I started playing with it.  Well...I'm
> either severely confused, or something isn't working properly.  Could anyone
> offer some help?

Help is easiest if you also explain what the problem is ;) I already see one
problem though: you read the password with 'string.strip(raw_input())',
which means the terminating newline gets stripped off the password, but you
dont supply it with os.write(), which means ssh thinks there is more
password to come, and patiently sits waiting for it, untill you brutally
kill it ;)

In your real app you probably want to use select(), by the way, before
attempting to read something. Doing a read() when there's no data to be read
makes a blocking read, which can lead to deadlocks. See telnetlib for some
examples on how to use it.

> Here's a sample of how I'm attempting it now:

> def remote_ex(hostname, password):

[..]

>   pid, fd = pty.fork()
>   if pid == 0:
>     os.execv("/usr/local/bin/ssh",[hostname])
>   else:
>     time.sleep(2)
>     print "Child says %s" % string.strip(os.read(fd,1024))
>     print "Child took %d password bytes." % os.write(fd, password)

[..]

> def main():
>   myhost = string.strip(raw_input("Enter a host:"))
>   mypass = string.strip(raw_input("Enter your Password:"))
>   remote_ex(myhost, mypass)

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From bjorn at roguewave.com  Fri Apr 14 20:10:23 2000
From: bjorn at roguewave.com (Bjorn Pettersen)
Date: Fri, 14 Apr 2000 18:10:23 -0600
Subject: The Opposite of list(a,b,c)?
References: <38F7A07F.9B0F309C@bigfoot.com>
Message-ID: <38F7B36F.C7E8BD0@roguewave.com>


"Andrew P. Jewell" wrote:
> 
> Can someone please tell me the opposite of list("A", "b", "string")?
>  Or maybe there's a smarter way to do what I'm trying to do.  I have a
> dictionary that looks like this:
> 
>     dtnA["key1"] = [val1, val2]
>     dtnA["key2"] = [val3, val4]
>     dtnA["key3"] = [val5, val6]
> 
> I want to do something like this:
> 
> def func(SearchTerm):
>   <...>
>     retval = []
>     for keyval in (dtnA.keys()):
>         mtch = re.search(patSrchTrm, keyval)
>         if mtch:
>             retval.append(UNLIST(dtnA[keyval]))

retval = reval + UNLIST(dtnA[keyval])

-bjorn

> 
> I want to end up with just a list of matching values from the
> dictionary, ala:
>     ['val1, val2', 'val3, val4']
> 
> But I can only get a list of lists: [[val1,val2],[val3,val4]] - which
> makes perfect sense based on how I'm doing it - but I don't WANT a
> list of lists!  Is there anyway to do this?  I don't really want to
> tuple-ize my dictionary since it's attached to its list values in
> other ways.  It seems ugly to have to iterate my dictionary value list
> to UN-list them manually - and the repr deal is REALLY ugly. What am I
> missing??  Thanks for any help!
> 
> Andy
> 
>   -- http://www.python.org/mailman/listinfo/python-list



From tseaver at starbase.neosoft.com  Wed Apr 26 11:00:29 2000
From: tseaver at starbase.neosoft.com (Tres Seaver)
Date: 26 Apr 2000 10:00:29 -0500
Subject: VIM NT +python
References: <2B1262E83448D211AE4B00A0C9D61B03011152BA@MSGEURO1>
Message-ID: <56ABD32374048A4C.D7ACAC2B7E495CA1.2AA90F35BCFCAA8E@lp.airnews.net>

In article <2B1262E83448D211AE4B00A0C9D61B03011152BA at MSGEURO1>,
Pieter Claerhout  <PClaerhout at CREO.BE> wrote:

>If you download VIM 5.6, unzipped it and ran install.exe, you can
>choose to turn on syntax highlighting. Once you did this, the _vimrc
>file is created with (python) syntax highlighting turned on.

Just to clarify -- this thread is about finding a Vim for Win32 with the
embedded Python interpreter compiled in -- *not* about python syntax
highlighting with Vim.  Mike very generously compiled and made available
for download such an executable, which is a hard thing to do if you don't
have access to M$ VC--.

>
>Pieter
>
>-----Original Message-----
>From: Mike Steed [mailto:MSteed at altiris.com]
>Sent: Tuesday, April 25, 2000 11:01 AM
>To: python-list at python.org
>Subject: RE: VIM NT +python
>
>
>Hi,
>
>   I just built the latest version (5.6.69) with the default feature set,
>plus the Python interface.  I have not tested it beyond
>
>   :python print 'hi'
>
>which passed with flying colors.
>
>http://athani.pair.com/msteed/vim/gvim56py.zip
>
>M.
>
>> From: Me [mailto:me at nospam.edu]
>> Sent: Tuesday, April 25, 2000 11:27 AM
>> To: python-list at python.org
>> Subject: Re: VIM NT +python
>> 
>> 
>> So would I!
>> 
>> <tiddlerdeja at my-deja.com> wrote in message
>> news:8e4j27$hgq$1 at nnrp1.deja.com...
>> > Does anyone here have a winNT build of the VIM editor that 
>> has python
>> > support switched on?
>> >
>> > I'd love to get my hands on a copy.

Now-to-figure-out-how-to-use-ftplib/urllib-from-within-Vim'ly,

Tres.
-- 
---------------------------------------------------------------
Tres Seaver        tseaver at digicool.com     http://www.zope.org
Digital Creations


From bjorn at roguewave.com  Tue Apr  4 19:41:16 2000
From: bjorn at roguewave.com (Bjorn Pettersen)
Date: Tue, 04 Apr 2000 17:41:16 -0600
Subject: Two questions
References: <38EA5FE8.853E9B75@callware.com> <38EA6508.90BF630C@roguewave.com> <38EA6E87.40EDB542@callware.com> <38EA6422.61054B11@exceptionalminds.com> <38EA7452.9BFDF37E@callware.com> <8cdt7c$cl1$1@news101.telia.com>
Message-ID: <38EA7D9C.7024CF0B@roguewave.com>

Anders Eggers-Krag wrote:
> 
> an alternative solution is to call an ADO (ActiveX Data Object) object and
> use it through
> there which is quite simple.

If you do this, make sure you get Mark's programming win32 book.  It's
an excellent book for many reasons, but it goes into special detail on
COM programming (and is so far the only resource I've seen of how COM
can be simple and useful -- instead of the normal
add-on-later-because-we-need-the-buzzword I normally associate with COM
<wink>).

-b



From dalke at acm.org  Sat Apr 22 02:54:28 2000
From: dalke at acm.org (Andrew Dalke)
Date: Sat, 22 Apr 2000 00:54:28 -0600
Subject: getting win32clipboard data
Message-ID: <8drih8$d54$1@slb7.atl.mindspring.net>

Hello,

  I've been trying to figure out how to get data from the clipboard
using win32clipboard.  I'm having a problem turning the integer
handle returned from GetClipboardData into something I can extract
data from.

  I've done a workaround using Sam Rushing's calldll and DynWin.windll,
based on hints from Wesley Phoa's "clipboard.py" at
http://www.deja.com/[ST_rn=ps]/getdoc.xp?AN=486991372&fmt=text

 = = = =
kernel32 = windll.module('Kernel32')

def get_data_by_format_id(id):
    win32clipboard.OpenClipboard(0)
    h = win32clipboard.GetClipboardData(id)
    if not h:
        win32clipboard.CloseClipboard()
        return ""

    kernel32.GlobalLock(h)
    try:
        size = kernel32.GlobalSize(h)
        text = []
        for j in range(size):
            t = calldll.read_byte(h+j)
            # read_byte returns a signed character
            if t < 0:
                t = t + 256
            text.append(t)
    finally:
        kernel32.GlobalUnlock(h)
        win32clipboard.CloseClipboard()

    # convert to the actual byte string
    text = string.join(map(chr, text), "")
    return text
 = = = =
(Note: this isn't the original function - I moved in the Open/Close
Clipboard calls from the calling function, but haven't tested it.)

I would like to replace the kernel32.{GlobalLock, GlobalUnlock,
GlobalSize} and the calldll.read_byte calls with something part of
the standard win32 distribution.  I figure that if win32clipboard
is a standard module, there has to be a way to use it.

Also, I'm not reading text data, which explains the read_byte
loop as compared to Wesley Phoa's use of read_string.  It seems
to work.  It would be nice to have a "read_data" :)

                    Andrew
                    dalke at acm.org





From rburnham at cri-inc.com  Wed Apr 19 18:38:07 2000
From: rburnham at cri-inc.com (Roger Burnham)
Date: Wed, 19 Apr 2000 22:38:07 GMT
Subject: Browse dialog in PythonWin
References: <38FE32AB.520C1736@callware.com>
Message-ID: <38fe3297.36998571@news-server.socal.rr.com>

On Wed, 19 Apr 2000 16:26:51 -0600, Ivan Van Laningham <ivanlan at callware.com> wrote:

>Hi All--
>OK, I've been able to get my dialogs up and even install handlers for
>button clicks, but now I need to start up a browse dialog that lets
>users  choose directories.
>
>And is there any news on Mark Hammond's Starship page?
>
><not-that-i'm-frustrated-but-my-keyboard-has-teeth-marks-now>-ly y'rs,
>Ivan
>----------------------------------------------
>Ivan Van Laningham
>Callware Technologies, Inc.
>http://www.pauahtun.org 
>http://www.foretec.com/python/workshops/1998-11/proceedings.html
>Army Signal Corps:  Cu Chi, Class of '70
>Author:  Teach Yourself Python in 24 Hours
>
Ivan,

Here's a dialog I adapted from something I found in the PythonWin
distribution.

-------------------------------------------------------------------

import sys, os
import win32ui, win32api, win32con, win32file
import string
from pywin.tools import hierlist
import commctrl


def dirParts(path):
    parts = []
    drive, path = os.path.splitdrive(path)
    pp = os.path.split(path)
    parts.append(pp[-1])
    while pp[0] != path:
        path = pp[0]
        pp = os.path.split(path)
        if pp[-1]:
            parts.append(pp[-1])
    parts.append(drive+'\\')
    parts.reverse()
    return parts

class DirHierList(hierlist.HierList):
    
    def __init__(self, root, listBoxID=win32ui.IDC_LIST1, defaultDir=None):
        self.defDir = defaultDir
        hierlist.HierList.__init__(self, root,
                                   win32ui.IDB_HIERFOLDERS, listBoxID)
        self.choice = None
        
    def HierInit(self, parent, listControl=None ):
        hierlist.HierList.HierInit(self, parent, listControl)
        if self.defDir is not None:
            parts = dirParts(self.defDir)
            path = ''
            for part in parts:
                path = os.path.join(path, part)
                items = self.itemHandleMap.items()
                for key,val in items:
                    if string.lower(val) == string.lower(path):
                        self.list.SelectItem(key)
                        if self.IsExpandable(val):
                            try:
                                self.list.Expand(key, commctrl.TVE_EXPAND)
                            except:
                                pass
                        
    def GetText(self, item):
        ret = os.path.basename(item)
        if not ret:
            ret = item
        return ret
    
    def GetSubList(self, item):
        if os.path.isdir(item):
            try:
                ret = map(lambda path, base=item:
                          os.path.join(base, path), os.listdir(item))
                ret = filter(os.path.isdir, ret)
            except os.error:
                ret = None
        elif item == 'My Computer':
            possible = string.splitfields(win32api.GetLogicalDriveStrings(),
                                          '\000')
            ret = []
            for drive in possible:
                dType = win32file.GetDriveType(drive)
                if (drive and
                    dType == win32con.DRIVE_FIXED or
                    dType == win32con.DRIVE_REMOTE):
                    try:
                        os.listdir(drive)
                        ret.append(string.upper(drive))
                    except os.error:
                        pass
        else:
            ret = None
        if ret is None:
            return ret
        else:
            ret.sort()
            return ret

    def IsExpandable(self, item):
        return os.path.isdir(item)
    
    def GetSelectedBitmapColumn(self, item):
        return self.GetBitmapColumn(item)+6
    
    def TakeDefaultAction(self, item):
        return PerformItemSelected(item)

    def PerformItemSelected(self, item):
        self.choice = item
        w = len(self.title)
        self.dlg.SetWindowText(self.title+' ...%s' % item[-(49-w):])


def DirectoryDialog(title, topDir, defDir):
    dirList = DirHierList('My Computer', defaultDir=defDir)
    dirList.title = title
    dlg = hierlist.HierDialog(title, dirList)
    dirList.dlg = dlg
    if dlg.DoModal() == win32con.IDOK:
        return dirList.choice
    else:
        return None



if __name__== '__main__':
    direc = DirectoryDialog('Test of directory selection',
                            'My Computer',
                            'C:\\CRI\\Images')
    win32ui.MessageBox('You chose: %s' % direc)

-------------------------------------------------------------------

Cheers,

Roger Burnham
rburnham at cri-inc.com


From effbot at telia.com  Mon Apr 17 12:44:10 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Mon, 17 Apr 2000 16:44:10 GMT
Subject: Python flak???
References: <8dfdmq$v1o$1@nnrp1.deja.com>
Message-ID: <ubHK4.2613$Za1.38989@newsc.telia.net>

Arint? <jamarijr at hotmail.com> wrote:
> What are the reprecussions <sp?> of doing a PyErr_Clear() when there is
> no error?

same as setting a variable to NULL if it already
contains NULL.

(well, to be precise, PyErr_Clear actually clears
three internal variables...)

</F>




From tom__98 at my-deja.com  Mon Apr 10 16:08:31 2000
From: tom__98 at my-deja.com (tom__98 at my-deja.com)
Date: Mon, 10 Apr 2000 20:08:31 GMT
Subject: Python/Perl Popularity (Re: A Mountain of Perl...)
References: <20000405053920.20579.qmail@web2102.mail.yahoo.com>
Message-ID: <8ctcar$ml9$1@nnrp1.deja.com>

In article <20000405053920.20579.qmail at web2102.mail.yahoo.com>,
  lewst <lewst at yahoo.com> wrote:
> What is it about Perl that makes it so much more popular and have such
> a huge grassroots swell?  I personally find Perl an abomination and
> Python a breath of fresh air.  Perl has that first mover advantage I
> suppose, but should that really make such a hugh difference?

I don't know about others, but I frequently end up using
Perl instead of Python, even though I greatly prefer Python
as a language.

The reasons?  Several:

 * perldoc makes it really easy to find documentation
   for modules

 * CPAN and the CPAN module make it really easy
   to find and install modules: they are all in a single
   place, they are well indexed, they are searchable,
   and they can be downloaded and installed with one
   or two commands.

 * Even when installing by hand, I find Perl modules
   easier to install than Python modules.

 * There seem to be a lot more Perl modules out there
   than Python modules.

 * Some modules that are important to me (HTTP, HTML,
   database, etc.) seem to be more up-to-date for Perl
   than for Python.

 * Perl actually has a few handy language features,
   including lexical closures and automatic caching
   of compiled regular expressions, which makes it feel
   not quite as clunky as it is.  Also, its syntax makes
   writing short scripts and one liners disproportionately
   easy.

 * Perl now comes preinstalled on a lot of operating systems,
   so I can pretty much use it in place of the shell for
   scripting (and it's a lot nicer than the shell for anything
   non-trivial).

Again, I greatly prefer Python.  But if the choice is between
writing a 100 line Perl script based on a module that takes a
minute to download and install or trying for an hour to find
and install some Python extension, Python loses out.  Of course,
the bigger the job, the more important Python's linguistic
advantages become.  So, I end up writing lots of smaller Perl
scripts (some of which, unfortunately, grow to be quite big),
and the occasional larger Python project.

Tom.




Sent via Deja.com http://www.deja.com/
Before you buy.


From effbot at telia.com  Wed Apr 12 10:19:45 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Wed, 12 Apr 2000 14:19:45 GMT
Subject: Python idiom: Multiple search-and-replace
References: <20000412100816.A2431031@vislab.epa.gov>
Message-ID: <5C%I4.645$rc9.190288896@newsb.telia.net>

Randall Hopper <aa8vb at yahoo.com> wrote:
> Is there a Python feature or standard library API that will get me less
> Python code spinning inside this loop?   re.multisub or equivalent? :-)

haven't benchmarked it, but I suspect that this approach
is more efficient:

...

# based on re-example-5.py

import re
import string

symbol_map = { "foo": "FOO", "bar": "BAR" }

def symbol_replace(match, get=symbol_map.get):
    return get(match.group(1), "")

symbol_pattern = re.compile(
    "(" + string.join(map(re.escape, symbol_map.keys()), "|") + ")"
    )

print symbol_pattern.sub(symbol_replace, "foobarfiebarfoo")

...

</F>

<!-- (the eff-bot guide to) the standard python library:
http://www.pythonware.com/people/fredrik/librarybook.htm
-->




From tim_one at email.msn.com  Tue Apr 25 01:43:59 2000
From: tim_one at email.msn.com (Tim Peters)
Date: Tue, 25 Apr 2000 01:43:59 -0400
Subject: List comprehensions?
In-Reply-To: <slrn8g19cn.mpr.scarblac-spamtrap@flits104-37.flits.rug.nl>
Message-ID: <000601bfae79$41485b00$152d153f@tim>

[Remco Gerlich]
> I thought list comprehensions had been in CVS for quite a while, so I
> expected them to be in 1.6. ...

Greg Ewing's patch (against 1.5.2) for list comprehensions is all there is,
and you'll have to get that from Greg (see DejaNews for his announcement,
and his URL).

a-false-premise-may-yet-lead-to-a-true-conclusion-ly y'rs  - tim





From echuck at mindspring.com  Tue Apr 18 00:30:19 2000
From: echuck at mindspring.com (Chuck Esterbrook)
Date: Tue, 18 Apr 2000 00:30:19 -0400
Subject: Here is some starship.python.net news
References: <38FA9A30.33DD1994@concentric.net> <38FB90B4.35E9CA1B@bullfrog-tech.com>
Message-ID: <38FBE4DB.67289F42@mindspring.com>

icarr at bullfrog-tech.com wrote:
> 
> I heard once more from the hosts of starship.python.net
> 
> Apparently, the "responsible (for fixing it) person" is our very own
> Mike McLay, who is theoretically on-site with the machine today
> (Monday) and I got a "hopefully, it will be back up today" from the
> mscp guys.
> 
> I say theoretically and hopefully because according to the news on
> my pager, today is the worst day in the history of Washington, D.C.
> traffic.  So who knows if Mike McLay even made it to the machine.
> 
> If he did or once he does, let us hope that he can put it back into
> working order.
> 
> Manus Hand
> posting from God-knows-whose address, so please don't reply


The situation with the Starship seems to reveal that:

[a] It's highly relied upon by the Python community (consider all the modules people find at V of P that are really hosted on the Starship for downloading).
[b] It's not well monitored and/or taken care of
[c] There is no suitable backup/contingency machine to switch over to.


How can the PSA change the situation to address [b] and [c]? I'd certainly be willing to pay higher dues if that's what it took, but there may be other ideas.


-Chuck


From chuyser at flash.net  Sun Apr 23 00:28:22 2000
From: chuyser at flash.net (Cindy Huyser)
Date: Sat, 22 Apr 2000 23:28:22 -0500
Subject: Tkinter bind of class -- help
Message-ID: <39027BE6.8DF89328@flash.net>

I've created a class that contains a number of Canvas widgets inside
frames.  When I import this class and instantiate it in my main program
I put it inside a frame.  

What I would like to do is to get x,y coordinate information on events
(either <Button-1> or <Motion>) that occur in the area occupied by the
class instance so that I can display detail information in another area
of the main window.  At first, I tried to use bind() on the class
instance, but received an error.  I have tried callbacks from the frame
in which the object resides, but find I only get event information for
that area unoccupied by the object; I have tried looking at information
stored in a member variable of the class, but have found the response
unacceptably slow.

Any suggestions for solving this problem would be greatly appreciated by
this newbie Tkinter programmer.


From akuchlin at mems-exchange.org  Sun Apr 23 10:11:42 2000
From: akuchlin at mems-exchange.org (Andrew Kuchling)
Date: 23 Apr 2000 10:11:42 -0400
Subject: SSL w/ urllib
References: <3901D800.DB50D968@flashmail.com> <3ditx9v851.fsf@newcnri.cnri.reston.va.us> <390249B5.F5D545F7@flashmail.com>
Message-ID: <3dhfcsnbz5.fsf@newcnri.cnri.reston.va.us>

Jp Calderone <exarkun at flashmail.com> writes:
>  If I switch it over to a GET request, it works fine, but I
> really need POST to work.  Is this another bug or am I doing
> something wrong?

Oops!  I wrote the patch too quickly and forgot one more piece; it
never sends the body of the POST.  <bullwinkle>This time for
sure!</bullwinkle>

--amk

Index: urllib.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/urllib.py,v
retrieving revision 1.91
diff -C2 -r1.91 urllib.py
*** urllib.py	2000/02/04 15:28:41	1.91
--- urllib.py	2000/04/23 14:09:44
***************
*** 328,331 ****
--- 328,333 ----
              for args in self.addheaders: apply(h.putheader, args)
              h.endheaders()
+             if data is not None:
+                 h.send(data + '\r\n')
              errcode, errmsg, headers = h.getreply()
              fp = h.getfile()


From hgg9140 at seanet.com  Mon Apr 17 20:47:25 2000
From: hgg9140 at seanet.com (Harry George)
Date: 17 Apr 2000 17:47:25 -0700
Subject: Python CGI
References: <38F852D9.E2144F17@flashmail.com>
Message-ID: <m3aeisxmjm.fsf@wilma.george.org>

I run python cgi's on RS6000 (AIX 4.3.2), using Netscape Enterprise
Server.  No problem.  Can you post the code, or at least a skeleton
which exhibits the behavior (leaving out sensitive content)?

Jp Calderone <exarkun at flashmail.com> writes:

>  I recently converted one of my perl CGI's to Python, and came across an
> bizarre and extremely undesirable side affect - Whenever a form that uses
> the Python script as its ACTION, the python scripts seems to get executed
> about 8 times.  At first I thought there was some logic error in my program
> making it octuple all the output, but I can't find anything that might do
> this.  Has anyone seen this before, or know how to fix it?
> 
>  Thanks in advance,
>   Jp
> 
> -- 
> Where a calculator on the ENIAC is equipped with 18,000 vacuum tubes and
> weighs 30 tons, computers in the future may have only 1,000 vacuum tubes and
> weigh only 1.5 tons.    -- Popular Mechanics, March 1949
> --
>  11:27am up 17 days, 16:34, 1 user, load average: 0.23, 0.11, 0.04

-- 
Harry George
hgg9140 at seanet.com


From boncelet at udel.edu  Thu Apr  6 23:14:14 2000
From: boncelet at udel.edu (Charles Boncelet)
Date: Fri, 07 Apr 2000 13:14:14 +1000
Subject: Algorithm: combinations of (k) taken from (n) values
References: <meh9-ACCD3B.16041406042000@news.cit.cornell.edu>
Message-ID: <38ED5285.719F26D7@udel.edu>

Matthew Hirsch wrote:

> Hi All,
>
> Onto the next question...
>
> Can anyone think of an algorithm to store as lists all possible
> combinations of k numbers taken from n possible numbers.  For example,
> given 5 values, I want to choose 2.  The number of possible combinations
> is given by 5!/(3!2!)=10.  They are:
>

Funny you should ask, since I just wrote one two days ago:

def combs(n,k=None):
    """returns sorted list of k items taken from n."""
    if k==None:
        k=n
    l = []
    for c in range(n):
        l.append([c])
    for i in range(k-1):
        li = []
        for c in l:
            for j in range(c[-1]+1,n):
                li.append(c+[j])
        l = li
    return l

    Charlie Boncelet
------
Charles Boncelet, University of Delaware,
On sabbatical at ADFA, Canberra Australia,
Home Page: http://www.ece.udel.edu/~boncelet/




From boud at rempt.xs4all.nl  Sat Apr 29 16:02:21 2000
From: boud at rempt.xs4all.nl (Boudewijn Rempt)
Date: 29 Apr 2000 20:02:21 GMT
Subject: GUI developping with Python (Any tutorial )
References: <Pine.A41.3.96.1000429092757.42286A-100000@aix2.uottawa.ca>
Message-ID: <8eff4d$ftl$1@news1.xs4all.nl>

abouf066 at aix2.uottawa.ca wrote:
> Greetings, 

> 	I wonder if anybody can point me to a site where I could find the
> starting steps to develop a GUI using python. 

One starting place is http://www.valdyas.org/python/tutorial.html - but
that's only for PyQt and PyKDE. There's such a variety of gui's
available, and most of them wildly different from each other, that you
really need to look around and try a few.

-- 

Boudewijn Rempt  | http://www.valdyas.org


From m.faassen at vet.uu.nl  Fri Apr 21 15:29:28 2000
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 21 Apr 2000 19:29:28 GMT
Subject: Pythonware Website???? Gone on vacation with Starship? ;-))
References: <20000419154202.B480@owl.rhein-zeitung.de> <JCjL4.540$wYl.187815424@newsb.telia.net> <3d4s8yotiy.fsf@amarok.cnri.reston.va.us>
Message-ID: <8dqa6o$pjg$4@newshost.accu.uu.nl>

Andrew M. Kuchling <akuchlin at mems-exchange.org> wrote:
> "Fredrik Lundh" <effbot at telia.com> writes:
>>     www.pythonware.com is down
>>     the reason appears to be hardware problems on
>>     the ISP site. ...

> First Zope.org (down because an Ethernet cable got pulled out), then
> Starship (disk crash), now pythonware.com.  Has someone released evil
> whitespace-eating nanobots that also attack computer hardware?

It also happened to www.happypenguin.org, but that was all a plot by
sourceforge to take it over. Since the PSU is bigger than the sourceforge
conspiracy, that couldn't happen to us. Also, since the PSU has access to
Guido's time machine and didn't prevent this, the PSU must be in on this.
Perhaps they're testing those samples of whitespace-eating nanoviruses that
they got from that UFO crash near Roswell back in the 50s.

Paranoid-ly yours,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From tuttledon at hotmail.com  Thu Apr 20 19:36:30 2000
From: tuttledon at hotmail.com (Don Tuttle)
Date: Thu, 20 Apr 2000 23:36:30 GMT
Subject: Why MFC instead of Tkinter -- was Re: Pythonwin?
References: <eMz2Nmiq$GA.73@net025s> <4QtL4.24$v85.41@news-server.bigpond.net.au> <chuL4.8483$O7.17315@typhoon.southeast.rr.com> <7LDL4.534$v85.2563@news-server.bigpond.net.au> <sfu6d657l6e184@corp.supernews.com> <38ff21b8$0$10972@senator-bedfellow.mit.edu> <38FF281E.5B2DCC71@tismer.com> <38ff823d$0$10971@senator-bedfellow.mit.edu>
Message-ID: <2wML4.15897$O7.23504@typhoon.southeast.rr.com>

"Robert"
> I was actually referring to using the dialog
> editor vs. programmatically adding your widgets using a layout manager,
and
> not specfically VC++. Visual Basic
> (and I assume) Delphi has similar and probably even easier editors...

Which brings up the question near and dear to a lazy man's heart like
myself..  Which is the eaiser Windows GUI builder for Python--VB or VC++ ?

With VB,  I've only seen Python mentioned as a COM object to be used by VB.
Are there other ways?

With VC++,  I gather you use the dialog editor to generate the MFC code that
you then plug into Python?  Correct?

Don







From p.agapow at ic.ac.uk  Wed Apr 19 10:22:50 2000
From: p.agapow at ic.ac.uk (Paul-Michael Agapow)
Date: Wed, 19 Apr 2000 15:22:50 +0100
Subject: 1st non-trivial program - gentle criticism requested
References: <1e90r16.1sd2ts41e7ur6kN@gershwin.bio.ic.ac.uk> <38F59E2E.308BD8C4@acm.org>
Message-ID: <1e9322v.axv3rdi1jtwaN@gershwin.bio.ic.ac.uk>

Thanks to all for your suggestions - getting the mindset of a new
language can be tricky, but this exercise has defintely been worthwhile.
Oh and ...

Robert W. Cunningham <rcunning at acm.org> wrote:
> I couldn't get it to run under Linux until I trivially initialized the
> following variables:
> 
> gTarget = ""
> gNumGenerations = gMutationRate = gStepSize = None

... which didn't present any problem on the Macintosh. Interesting.

-- 
Paul-Michael Agapow (p.agapow at ic.ac.uk), Biology, Silwood Park
"The world is made of fire."


From effbot at telia.com  Wed Apr 19 10:43:53 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Wed, 19 Apr 2000 14:43:53 GMT
Subject: Pythonware Website???? Gone on vacation with Starship? ;-))
References: <20000419154202.B480@owl.rhein-zeitung.de>
Message-ID: <JCjL4.540$wYl.187815424@newsb.telia.net>

Oliver Andrich <oli at rz-online.net> wrote:
> I can't access the www.pythonware.com webpages since yesterday
> evening. Does someelse also have the problems or is it just me?

quoting the daily python URL (http://hem.passagen.se/eff/url.htm):

    www.pythonware.com is down

    the reason appears to be hardware problems on
    the ISP site. this also affects mail traffic to (but
    not from) the pythonware.com and secretlabs.com
    domains. our ISP is working on this, and expect to
    have the site back online within a few hours.

urgent mail can be sent to this address.

</F>




From mlauer at trollinger-fe.rz.uni-frankfurt.de  Mon Apr  3 06:11:55 2000
From: mlauer at trollinger-fe.rz.uni-frankfurt.de (mlauer at trollinger-fe.rz.uni-frankfurt.de)
Date: 3 Apr 2000 12:11:55 +0200
Subject: GTK Referenc
References: <38E7A868.A1C9BD7C@searchy.net>
Message-ID: <38e86e6b@nntp.server.uni-frankfurt.de>

Penpal International (ppi at searchy.net) wrote:
> I couldn't find anywhere a reference for the GTK module for Python. Does
> anyone know one?
> On gtk's site I could only find some for C(++).

Check out http://theopenlab.uml.edu/pygtools

--
--
    Regards & Gruesse from Mickey @ http://www.Vanille.de
  ---------------------------------------------------------
  How could anyone know me - when I don't even know myself ?




From phd at phd.russ.ru  Fri Apr 21 03:44:30 2000
From: phd at phd.russ.ru (Oleg Broytmann)
Date: Fri, 21 Apr 2000 07:44:30 +0000 (GMT)
Subject: PyApache memory leak?
In-Reply-To: <00042011042800.02559@quadra.teleo.net>
Message-ID: <Pine.LNX.4.21.0004210743260.2712-100000@fep132.fep.ru>

On Thu, 20 Apr 2000, Patrick Phalen wrote:
> I haven't used PyApache, but am under the impression that it spawns
> multiple interpreter instances, which could certainly cause memory
> problems on a busy, CGI-intensive site.
> 
> Have you looked at Httpdapy (soon to be renamed mod_python)?
> http://www.ispol.com/home/grisha/httpdapy/

   I don't see how httpdapy could help spawn less interpreters instances -
it is Apache job to fork and reap off children.

Oleg.            (All opinions are mine and not of my employer)
---- 
    Oleg Broytmann      Foundation for Effective Policies      phd at phd.russ.ru
           Programmers don't die, they just GOSUB without RETURN.




From don.limpio at gmx.net  Thu Apr 27 05:40:46 2000
From: don.limpio at gmx.net (Georg Hammerbacher)
Date: Thu, 27 Apr 2000 11:40:46 +0200
Subject: newbie   dbase dataaccess
Message-ID: <ms19e8.uv5.ln@dorian.georks.de>

Hello everybody,

I?m just new in programming Python and now I want to make my own personal
adressmanager with a simple dBase-file.
How can I get access to it or where can I find more info?s about using
dbase-files within Python?

Thanks

Georg





From charlie at intelligenesis.net  Mon Apr 10 11:53:49 2000
From: charlie at intelligenesis.net (Charlie Derr)
Date: Mon, 10 Apr 2000 11:53:49 -0400
Subject: Extending built-in language (Win 95)
In-Reply-To: <213CDEBFDDA7D3118081009027E40C56071118@SHEF-MS1>
Message-ID: <LOBBJCAMDNLNCGCCHGEICEDIDPAA.charlie@intelligenesis.net>

~ > ? No, you use open() instead, which just returns an instance of a file
~ > object. In the xxmodule, xx_new is used to make new instances, which
~ > can be called from Python as xx.new().
~ 
~ I've added xxmodule to the lib and recompiled. Entering:
~ a = xx.new()
~ 
~ just gives a name error on xx.
~ 
~ Have I missed something?
~ 

I think you need to import your module:

import xxmodule

	or

from xxmodule import *

~ 
~ (thanks for helping)
~ 

you're welcome (if it's the right answer :-] )

~ mark.
~ 
~ -- 
~ http://www.python.org/mailman/listinfo/python-list
~ 



From rgetter at usa.net  Sun Apr 16 04:31:01 2000
From: rgetter at usa.net (Robert Getter)
Date: Sun, 16 Apr 2000 08:31:01 GMT
Subject: Event triggering and weak references
References: <8dbif3$hm0$1@metro.ucc.usyd.edu.au>
Message-ID: <MPG.1363441c166d9ca98969e@news-server>

In article <8dbif3$hm0$1 at metro.ucc.usyd.edu.au>, 
rnog2438 at mail.usyd.edu.au says...
> First thanks for this beautiful language : )
> 
> I am Smalltalk addicted : )  and this means that I can't do with any other
> language, ex. C, C++, Java,  : (
> This was true until I found Python 2 weeks ago at Sydney Uni. I loved its
> modularity, pluggin behaviour and freedom.
> But I am missing Smalltalk event handling... Smalltalk (like Python) support
> the call-Back mechanism, but there is a more powerful inter-object interest
> registering that I haven't found in Python (so far).
> 
> This is what I would like to do:
> Imagine I could have two or more Widgets open displaying a car speed value,
> say a numerical and a graphic display, if this car speed changes (ex. some
> one pushes the break), I would like to have any open widget on this value
> updated (on opening each window would register interest with this car
> speed).
> 
> Smalltalk implements this with:
> 
> The widget would register interest in one aspect of the car  (on opening
> over aCar):
>   aCar when:#speedChanged send: #updateSpeedDisplay to: self
> 
> And the car would just notify who is interested, if any (when its speed
> changes):
>   self trigger: #speedChanged
> 
> But if any of this windows is not referenced anymore or crashes I would like
> the garbage collector clear it (the car would not be notified and his
> pointer to the window shouldn't keep the garbage collector form freeing the
> window memory).
> 
> Again, Smalltalk implements this with weak references, that doesn't count
> for the garbage collector.
> 
> Is there support for this in Python?
> 
> Any suggestions?
> 
> 
> 

Here's some code which does some of what you want:
=========
# simple test of a scheme for registering interest in an event.
# exception safety is left as an exercise for the reader

class watcher:
	def __init__(self, watchee, voyeur=None):
		self.watchee = watchee
		if voyeur == None:
			self.list = []
		else:
			self.list = [ voyeur ]

	def __call__(self, *args, **kwargs):
		rc = apply(self.watchee, args, kwargs)
		for voyeur in self.list:
			apply(voyeur, args, kwargs)
		return rc

	def register(self, voyeur=None):
		if not voyeur: return
		self.list.append(voyeur)

	def unregister(self, voyeur=None):
		if voyeur: self.list.remove(voyeur)
		return len(self.list)


def register(object, method, voyeur=None):
	# try to add this voyeur to the watcher object
	try:
		getattr(object, method).register(voyeur)
	# if this fails, there is no watcher object. Add one.
	except:
		try:
			setattr(object, method,
				watcher(getattr(object, method), voyeur))
		# if this fails, there is no method of that name. return 
None.
		except:
			return None
	# finally, return the newly installed watcher method
	return getattr(object, method)

def unregister(object, method, voyeur=None):
	try:
		if not getattr(object, method).unregister(voyeur):
			# list is empty. remove watcher
			setattr(object, method,
				getattr(object, method).watchee)
	except: pass

class test:
	def x(self): print "x()"

def watch1():
	print "watch1()"

def watch2():
	print "watch2()"

def main():
	a = test()
	a.x()
	print
	register(a, "x", watch1)
	a.x()
	print
	register(a, "x", watch2)
	a.x()
	print
	unregister(a, "x", watch1)
	a.x()
	print
	unregister(a, "x", watch2)
	a.x()

if __name__ == "__main__": main()
========

What this code does is install a wrapper around the method you want to 
watch. This wrapper maintains a list of interested parties. There are a 
few problems with this method:

1. No weak references. I think there is a weak reference extension 
somewhere which could be added without too much trouble.

2. In Smalltalk, AFAIK, a method can only be referenced through its 
object. In Python, each method is a callable object which can be 
manipulated separately. In other words, object methods in Python are 
really instance variables which just happen to be callable. As a result, 
anywhere that objects pass bound methods around won't work right. If a 
bound method is passed before the wrapper is installed, calls to the 
method will not be sent to the watchers. This means that watchers can't 
be reliably watched, since they work by getting references to themselves 
put on a list in the watcher class.

The only real solution I can think of for problem 2 is a generic 
mechanism for attaching watchers to any callable object. This is similar 
to what a debugger does. If you look into the debugging API, you may find 
something that is suitable or can be modified.

I never much thought about it before, but this could be a really useful 
capability to have.

Another way that this could be done which would also allow other nice 
things to happen would be the ability to globally change all references 
to one object into references to a different object.
-- 
Rob Getter


From bawolk at pacbell.net  Thu Apr 20 10:00:32 2000
From: bawolk at pacbell.net (Bruce Wolk)
Date: Thu, 20 Apr 2000 07:00:32 -0700
Subject: Tkinter aspect question
Message-ID: <38FF0D80.7B6EBA8C@pacbell.net>

I tried to post this earlier but it never appeared.  Apologies if both
show up.

How can I keep the aspect ratio of a window fixed?  I want my window to
stay square, no matter what its size is.  The aspect method seemed
promising, but when I tried

win = Toplevel(master)
win.geometry('400x400+100+100')
win.aspect(1,1,1,1)

it didn't have any effect.  The aspect could still be changed. Thanks.

Bruce


From m.faassen at vet.uu.nl  Thu Apr 27 06:17:41 2000
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 27 Apr 2000 10:17:41 GMT
Subject: Simple del[] list question
References: <Pine.BSF.4.10.10004271004170.27597-100000@hercule.perceval.be>
Message-ID: <8e9445$f22$1@newshost.accu.uu.nl>

Gregoire Welraeds <greg at perceval.be> wrote:
> In reply to the message of Martijn Faassen sent on Apr 26 (see below) :

>> Note that another strategy which tends to avoid these troubles is to
>> construct a new list if you modify one. That is, usually if you're going

> Isn't it time and memory consuming when working on big list (eg with 1000
> or above elements) ?

Yes, probably, depending on what you do. If you're going to delete 
a lot of elements in a 1000 element list, the list will internally be
copied several times as well, each time it's resized, if I understand it
well. This happens in fast C code, though. Still, this is going to add up.
At some point it may start to be faster to construct a new list (using
the cheap 'append()' operation) and just leave those elements out. I don't
know when that's the case, however. It depends on the cost of each 
for loop iteration, the length of the list, the amount of elements that
are deleted. If it's really speed critical, just try out both strategies,
I suppose. I think there's no difference in memory consumption; the
copy the list strategy will generally be cheaper, unless you're only going
to delete a single element, and in that case it won't matter, I think.

I think for shorter lists with multiple deletes the 'construct a new list'
strategy will often be clearer and faster. Note that in the deletion case
the program needs to find out which elements to delete; this often involves
going through the lists already.

All this said, I think it's far more important in general that the 
'construct a new list' idiom produces code that's easier to understand;
efficiency is a secondary concern.

Regards,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From pinard at iro.umontreal.ca  Sun Apr  2 08:16:57 2000
From: pinard at iro.umontreal.ca (=?ISO-8859-1?Q?Fran=E7ois_Pinard?=)
Date: 02 Apr 2000 08:16:57 -0400
Subject: Why should I switch to Python?
In-Reply-To: "Tim Peters"'s message of "Sat, 1 Apr 2000 23:20:03 -0500"
References: <000701bf9c5a$b806b420$752d153f@tim>
Message-ID: <oqity066km.fsf@titan.progiciels-bpi.ca>

"Tim Peters" <tim_one at email.msn.com> writes:

> When in doubt, there's no substitute for exhaustive enumeration.

Hmph!  Hope you do not doubt there is an infinite number of primes! :-)

-- 
Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard





From firefly64 at my-deja.com  Sun Apr 16 21:16:24 2000
From: firefly64 at my-deja.com (firefly64 at my-deja.com)
Date: Mon, 17 Apr 2000 01:16:24 GMT
Subject: building on linux w/o threads?
Message-ID: <8ddokt$65t$1@nnrp1.deja.com>

Hi,
I'm trying to build python 1.5.2 on a linux box
but without thread support. I'm using gcc 2.9.5,
on Mandrak Linux 6.1 kernel version 2.2.14

Even though my system has threads, I don't want
any thread support in python (no linkage to libpthreads).
I just can't seem to convince the build process
to proceed without threads.

I've done everything I can think of, including
./configure --without-thread and editing the
Makefiles to include -DWITH_THREAD=0. The build
process still always either produces a python
linked to libpthreads or dies complaining about
not being able to find various thread functions.
(Yes, I've done make clean).

The reason I'm doing this is because I'm trying to
get httpdapy to work. The author suggested that I
try building a python without threads.


Thanks a lot everyone,
mike salib


Sent via Deja.com http://www.deja.com/
Before you buy.


From fig at oreilly.com  Tue Apr 18 19:41:34 2000
From: fig at oreilly.com (Stephen R. Figgins)
Date: Tue, 18 Apr 2000 23:41:34 GMT
Subject: o'rielly python devcenter
References: <Pine.BSF.4.10.10004181241280.15729-100000@chi.pair.com>
Message-ID: <38FCF25E.4D945D81@oreilly.com>

It isn't old news, it just went public last night(Monday) Sort of a soft
launch, as I don't have much in the way of content up yet.  But I am
working on that problem!  I will have some original articles in a couple
of weeks.  

The Network is actually a subsidiary of O'Reilly & Associates, and I am
plenty happy to plug other's books.  (Actually I think the only book I
have linked right now is the venerable Programming Python, and that is
just because I mentioned it in me introduction.) 

The links I am building under topics are not links to other sites, but
rather to documents on other sites.  I am looking for tutorials,
articles, documentation, anything good and juicy with information.  I am
open to suggestions.  

And I like criticism, particularly the constructive kind.  So send me
plenty.  

And if any of you want to line up for writing some good articles for me,
I am paying!

Stephen R. Figgins
Editor, O'Reilly Network


From pinard at iro.umontreal.ca  Tue Apr  4 15:22:52 2000
From: pinard at iro.umontreal.ca (=?ISO-8859-1?Q?Fran=E7ois_Pinard?=)
Date: 04 Apr 2000 15:22:52 -0400
Subject: New Features in Python 1.6
In-Reply-To: "Terry Reedy"'s message of "Tue, 4 Apr 2000 14:30:06 -0400"
References: <200004011740.MAA04675@eric.cnri.reston.va.us> <gHrF4.16072$0o4.101062@iad-read.news.verio.net> <8cdbvg$3vi$1@news.udel.edu>
Message-ID: <oqwvmdn01f.fsf@titan.progiciels-bpi.ca>

"Terry Reedy" <tjreedy at udel.edu> writes:

> ' = 1 byte/char, " = 2 bytes/ char is more straightforward.

It would be much more logical, indeed.

Of course, we known that ASCII is ideally 7/8 of an apostrophe, but we
could use a whole apostrophe as an acceptable compromise.  For Unicode,
which is internally represented as UCS-2, a double quote is very natural.

When Python will later be adapted to process UCS-4 internally, we could use
quadruple apostrophes, conveniently represented by two double quotes, like
in ""This is an UCS-4 string"".  We could also use '''', "'', '"' and ''"
to discriminate between the four usual kind of endianness and byte swapping.

P.S. - Sorry! :-)  Just letting some of the steam out...  Back to work! :-)

-- 
Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard





From aa8vb at yahoo.com  Thu Apr 20 13:39:52 2000
From: aa8vb at yahoo.com (Randall Hopper)
Date: Thu, 20 Apr 2000 13:39:52 -0400
Subject: Python Parsers Comparison
In-Reply-To: <FlFL4.9921$HG1.265980@nnrp1.uunet.ca>
References: <20000420093945.A624723@vislab.epa.gov>
 <FlFL4.9921$HG1.265980@nnrp1.uunet.ca>
Message-ID: <20000420133952.A626243@vislab.epa.gov>

Warren Postma:
 |> For example, a comma-separated list of nested vectors would be a good
 |example:
 |>
 |>     [ 1, 2 ], [ 3, 4 ]
 |>     [ [ 1,2 ],[ 3,4 ] ], [ 5.6 ]
 |>     etc.
 |>
 |
 |Wherever I am able to do so, I use eval() and design my grammar to be
 |basically PYTHON.
...
 |If you are in control of the grammar, and you are inventing it, why not
 |stick with python, if on the other hand, the grammar is something you're
 |stuck with 'as-is', and it's not a proper subset of some Python construct,
 |then all bets are off.


You've hit it on the head.  The grammer is from a third-party tool which
I'm picking up and using as-is.  For simply expressions I already re.sub it
into Python and eval.  But more complex expressions don't play so nice with
this approach.


 |I certainly WOULD like to see some tutorial material on constructing such
 |parsers, because I don't know the first thing about them. We studied this
 |stuff in undergrad C.S., but it was trivial stuff (BNF grammars, some simple
 |parsing code techniques in C).  We never did learn the Nifty Hardcode Stuff.


Exactly.  We covered them as well and I have no doubt that with sufficient
time and effort I (or most folks certainly) could learn them all and pick
the best one.  Though, a "Python Parsers FAQ" written from the 30,000 foot
view (feature comparison, limitations, examples, still-maintained status,
etc.) would really be a timesaver for folks looking for the best parser for
their task.


-- 
Randall Hopper
aa8vb at yahoo.com



From gmcm at hypernet.com  Mon Apr 10 13:23:29 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Mon, 10 Apr 2000 13:23:29 -0400
Subject: Installer Producing Page Fault
In-Reply-To: <01bfa30b$e0707f60$2040c3d1@Cyberus>
Message-ID: <1256735085-36518465@hypernet.com>

Gordon Williams wrote:

> I am using Gordon's installer program to make a standalone program that
> self installs.  After running the install_BS_main.exe created, everything
> is put into the proper directory (I think).  Then running BS_main.exe I
> immediately get a crash with the following:
> 
> 
> 
> BS_MAIN caused an invalid page fault in
> module MULTIARRAY.PYD at 0137:01523569.
[snip]
> My code is using the numerical package if that matters. (It appears to be
> causing the fault as multiarray is part of the package)
> 
> Any ideas?

Never tried it myself, but I got this note once from Gary Herron:

> ** Numeric Python.  The 'import Numeric' statment imports _numpy.pyd,
> but this is not caught by builder.py.  Add
>   misc= C:\Program Files\Python\Numeric\_numpy.pyd
> in the appropriate target of the cfg file.

That was a couple releases ago, so I don't know if it still 
applies.

- Gordon



From nospam at nowhere.net  Wed Apr 26 01:55:53 2000
From: nospam at nowhere.net (Ray&Maria)
Date: Tue, 25 Apr 2000 23:55:53 -0600
Subject: Unsigned integer arithmetic
References: <newscache$51sktf$rzc$1@jenpc07.jennic.co.uk>
Message-ID: <PxvN4.1335$sf5.16905@news.corecomm.net>

What's wrong with this:

myInt = 0x7fffffff
myLong = long(myInt)  #coerce to long
myInt2 = int(myLong)  #coerce to int (if it fits)

def printU32Hex(n):
    if n > 0xffffffffL:
        raise ValueError
    print "0x%01x%07x" % (int(n >> 28), int(n & 0xfffffffL))

printU32Hex(0x7fffffffL)
printU32Hex(0xffffffffL)
printU32Hex(0x100000000L)

Ray
NOSPAMjchen at NOSPAMquark.com

"Robert Cragie" <rcc at nospamthanks_jennic.com> wrote in message
news:newscache$51sktf$rzc$1 at jenpc07.jennic.co.uk...
> Hi all,
>
> I'm having a lot of trouble with arithmetic manipulation of numbers above
> 0x7fffffff. Whilst an integer is represented in a 32 bit word, as you'd
> expect, 0x80000000 - 0xffffffff are treated as negative numbers. However,
I
> want to treat a 32 bit number as an unsigned value, and be able to do i =
i
> + 1 for an integer number above 0x7fffffff without getting an overflow.
>
> I'm finding I have to use the inefficient longs for the arithmetic, and do
> clumsy conversions using the struct module, e.g.
>
> import struct
>
> # integer to unsigned long
> # in the strings, it's an upper case 'eye' and a lower case 'ell' in that
> order
> myUlong = struct.unpack('I', struct.pack('l',myInt))[0]
>
> # unsigned long to integer
> # in the strings, it's a lower case 'ell' and an upper case 'eye' in that
> order
> myInt = struct.unpack('l', struct.pack('I',myUlong))[0]
>
> Also, I can't seem to print a long using
>
> print '0x%08x' % myLong
>
> if 'myLong' is greater than 0x7fffffff - I get the rather obscure 'long
int
> too long to convert'.
>
> Is there a better way to do all this?
>
> TIA
>
> Robert Cragie
>
>
>




From mwh21 at cam.ac.uk  Tue Apr 25 17:47:33 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 25 Apr 2000 22:47:33 +0100
Subject: Metaclasses?
References: <3905F4A0.2B39C705@darmstadt.gmd.de>
Message-ID: <m3em7tg8ei.fsf@atrus.jesus.cam.ac.uk>

Robb Shecter <shecter at darmstadt.gmd.de> writes:

> I read an old post about metaclasses (included on the cd-rom with
> the book "Mit Python programmieren"), and got very interested,
> because I imagine that by making a new metaclass, I could maybe fix
> the one thing about Python that annoys me: the obligatory use of
> "self" in method declarations and variable access.  (Python's syntax
> is so clean and neat in all other areas, I feel like it's a shame
> that there's excess verbosity in this one place.)

a) No, metaclasses are unlikely to "help".
b) How, when you've "fixed" Python do you propose, in method bodies,
   to distinguish between local and instance variables?  I might drop
   my longstanding habit of complaining about people complaining about
   the necessity of "self" if you can answer this question...

> So - are metaclasses just a random idea, or is there something in
> the works?

Metaclasses are real, but have this tendency to make one's head
explode...

Cheers,
M.

-- 
  If i don't understand lisp,  it would be wise to not bray about
  how lisp is stupid or otherwise criticize, because my stupidity 
  would be archived and open for all in the know to see.
                                               -- Xah, comp.lang.lisp


From lewst at yahoo.com  Wed Apr  5 01:39:20 2000
From: lewst at yahoo.com (lewst)
Date: Tue, 4 Apr 2000 22:39:20 -0700 (PDT)
Subject: A Mountain of Perl Books + Python Advocacy
Message-ID: <20000405053920.20579.qmail@web2102.mail.yahoo.com>

Before I launch into another question and gripe, I'd like to thank
everyone who offered comments and suggestions on the advanced Python
books that are out there.

Now something I can't quite figure out is: why are there so many more
books on Perl out there than on Python?  

Searching through Fatbrain.COM (which is where I order my books from),
I found 68 books, 4 training manuals, and 2 eMatter documents on Perl.
Compare this to Python's 13 books and 2 eMatter documents.

What is it about Perl that makes it so much more popular and have such
a huge grassroots swell?  I personally find Perl an abomination and
Python a breath of fresh air.  Perl has that first mover advantage I
suppose, but should that really make such a hugh difference?

I'll admit that Perl is what led me to Python in the first place.
After hearing about how great Perl was several years ago from the
local sysadmin, I learned it and started using it for my scripting
needs.  If there is one thing that sums up my Perl experience, it is
that it always took me too long write the programs that I needed to
write.  The syntax was always no unnatural for me that I could never
get my head out of the reference manual, and errors were always so
tough to track down.  In the end was frustration albeit a working
result.  This frustration led me to look into Python and I'll never
touch Perl again.  Python was so natural for me I often found myself
correctly "guessing" at the syntax as I learned it.  I rewrote all my
Perl scripts in one weekend and most of them worked on the first try.

At this point I wondered if my brain was just different than all those
Perl junkies out there.  But now I really don't think so; I think it's
a question of awareness.  Perl is very publicized and well-known while
the better language is sitting here a dark corner unnoticed.  Sure
there will always be some religious fanatics that won't even give
Python a try, but I think Python's popularity could be vastly improved
with some serious advocacy work.  CNRI and/or PSA should seriously
look into funding a Python "marketing" campaign of sorts.  I think the
result would be allot of converts and more understanding and respect
for Python.

With this in mind, let me include one of my favorite pro-Python quotes
of all time.  This is from a message to the fetchmail-announce mailing
list by Eric S. Raymond <http://tuxedo.org/~esr>, the famous
open-source advocate and author of many popular software programs.  He
is discussing "fetchmailconf" which is the Tkinter GUI for his
fetchmail POP/IMAP mail client.  I think it demonstrates the point I
make above perfectly.

  "A note about fetchmailconf. It took me approximately six days to
  write this elaborate multi-paneled GUI -- that's counting the four
  days it took me to learn the implementation language in the process.
  This could easily have been a two-month project in C (with six weeks
  of that spent debugging and bugs still left).  Or a week-long
  project in Perl, with working but ugly and unmaintainable results."

  "The verdict: Python is *waaaay* cool!  I'm sold.  It's clean, it's
  elegant, it's easy, and it's astonishingly powerful.  I'm not going
  to program anything longer than one screenful of script in Perl
  anymore.  I love Larry Wall dearly, but Guido van Rossum is the
  better designer -- I haven't had this much fun with a language since
  the glory days of LISP.  Eric sez check it out."

Regards.


__________________________________________________
Do You Yahoo!?
Talk to your friends online with Yahoo! Messenger.
http://im.yahoo.com



From richard_chamberlain at ntlworld.com  Wed Apr 26 14:19:34 2000
From: richard_chamberlain at ntlworld.com (Richard Chamberlain)
Date: Wed, 26 Apr 2000 19:19:34 +0100
Subject: Python Time classes
References: <slrn8ge5b0.1gu.philh@vision25.demon.co.uk>
Message-ID: <RnGN4.4133$0d2.88602@news2-win.server.ntlworld.com>

You could use mxDateTime which are at:

http://starship.python.net/crew/lemburg/mxDateTime.html

Richard


phil hunt <philh at vision25.demon.co.uk> wrote in message
news:slrn8ge5b0.1gu.philh at vision25.demon.co.uk...
> Is there a high-level Time class for Python?
>
> I'm aware of the 'time' module in the standard library. But I'm
> looking for something more high-level (I can't find anything
> relevant on Parnassus or Freshmeat). What I want is roughly the
> ability to say things like:
>
>    t1 = Time.now()
>    # (t1) is an instance of Time containing the time now
>
>    t2 = Time(2000, 4, 26, 17, 4, 0)
>    # (t2) is 2000-Apr-26 17:04:00
>
>    sec = t1 - t2
>    # (sec) is the number of seconds between (t1) and (t2)
>
>    str = t1.format("%H:%M:%S")
>    # (str) is a string containing the hours minutes and seconds of (t1)
>    # in a form like 23:59:06
>
> It'd also be nice if there was a similar Date class which stored
> a date (this would behave like a Time except that (i) the hh:mm:ss
> would always be 00:00:00, and (ii) subtracting two dates would return
> the number of days difference). Obviously times and dates would be
> easily convertable to each other:
>
>    aTime = Time(aDate)
>    aDate = Date(aTime)
>
> (or similar).
>
>
> Does this exist, or shall I write one?
>
> --
> ***** Phil Hunt ***** send email to phil at comuno.com *****
> Moore's Law: hardware speed doubles every 18 months
> Gates' Law: software speed halves every 18 months




From adjih at crepuscule.com  Sun Apr 30 12:45:48 2000
From: adjih at crepuscule.com (Cedric Adjih)
Date: 30 Apr 2000 16:45:48 GMT
Subject: MORE INFO
References: <8ea0ao$fro$1@nnrp1.deja.com> <8ea9uu$rh0$1@nnrp1.deja.com> <8ecad6$bun$1@ites.inria.fr> <8ecref$lck$1@nnrp1.deja.com>
Message-ID: <8ehnvs$m0h$1@ites.inria.fr>

mdefreitas at sikorsky.com wrote:


>> Do you want to run several python scripts (and several ui.commands
>> C code) concurently (how would you synchronize?) ?

> I do not want to run concurrently, I just want to run in one thread of
> execution, but recursively.

Ok.

>> Note that if all you want is to have C code that calls Python
>> code that calls C code (through a module) that calls Python code
>> without threads, PyRun_XXX functions are re-entrant.

> Thanks! That's what I needed to know. I now modified (and simplified)
> my code as follows:

> void interp(char *script) {
>    static int nest_level = 0;
>    FILE *fd = fopen(script, "r");
>    if (nest_level == 0) {
>       Py_Initialize();
>       PyRun_SimpleString("import ui\n"); // import my ui
>    }
>    nest_level++;
>    PyRun_SimpleFile(fd, script);
>    nest_level--;
>    if (nest_level == 0) Py_Finalize();
> }

> The only problem is that the nested script that is run "remembers" the
> variable settings from it's caller. Is there a way to give each nested
> python script a fresh, clean environment? The nested script should not
> effect the environment of it's caller either. Is that what the global
> and local dictionaries for the API function PyRun_File are for? If so,
> is there any examples on their usage. The C/API doc isn't all that
> explicit.

  Indeed.
  PyRun_File is actually implementing most of the "execfile(...)"
code (see http://www.python.org/doc/current/lib/built-in-funcs.html).
Python source code for execfile is also an example of use for 
PyRun_File.
  Passing empty dictionaries would be a way to have a "clean"
environment except if a script imports a Python module and fiddles
with its internals. Because modules will be shared among scripts.
But fiddling with internals is very bad style.

-- Cedric



From dalke at acm.org  Mon Apr 24 22:20:04 2000
From: dalke at acm.org (Andrew Dalke)
Date: Mon, 24 Apr 2000 20:20:04 -0600
Subject: new linereading standard?
References: <3904D21B.333DD930@visionart.com>
Message-ID: <8e2v7j$cod$1@slb1.atl.mindspring.net>

Pete Shinners wrote:
> it is a simple class
> that handles the "while 1: .... break" carnival for reading files.
[...]
> i'm sure someone has posted a class like the before, but i'm hoping it
> can one day make it in as a python standard someday. (6.0, there's still
> time!)


Try looking at the standard 'fileinput' module.

> now i can code in a preferred style
>
> myfile = open('myfile.txt', 'r')
> for line in filelines(myfile):
>     print line
> myfile.close()

import fileinput
myfile = fileinput.FileInput('myfile.txt')
for line in myfile:
   print line
myfile.close()

Not quite the same, since it only deals with filenames and yours
uses any file-like object.

So the module for what you want is present.  You'll find that nearly
no one uses it.  Part of the reason is the standard idiom (your
'carnival') is used for a lot more than reading files, so people get
used to seeing it.  Another reason is the performance suffers some
since you have the extra method call overhead.

Indeed, I've only used fileinput in examples for non-Python people
where I don't want to have to explain the while 1:/test loop.  They
more easily follow "this is a black box thingy which returns input
lines" compared to having to explain "yes, I know it's a couple
extra lines compared to C, but you get used to it and it prevents
some hard to find errors."

                    Andrew
                    dalke at acm.org





From gmcm at hypernet.com  Wed Apr 12 17:15:01 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Wed, 12 Apr 2000 17:15:01 -0400
Subject: global variables - how??
In-Reply-To: <8d2jsh$3a9$1@news101.telia.com>
Message-ID: <1256548392-47748201@hypernet.com>

Anders Eggers-Krag wrote:

> it is obvious not to use global variables except when the use of them is
> obvious,
> in my case I could otherwise send them along with every function, creating a
> huge
> overhead passing a million pointers around, using exactly the same amount of
> memory,
> or in fact more, than I would using globals.

Oh, it uses more memory all right (the code objects are larger, 
and more stack gets used), but it's also significantly faster 
(fewer name lookups).
 
> I did figure out how to do it myself after a lot of trying, but as you said
> it is ugly,
> but this really anoys me, and I *was* starting to like python...

There are a number of ways to do it (most of which will lose 
the speed advantage). I don't find anything ugly about 
everyone importing a common module, for example (or 
passing around a single context object).

- Gordon



From rjroy at takingcontrol.com  Wed Apr 19 18:47:35 2000
From: rjroy at takingcontrol.com (Robert Roy)
Date: Wed, 19 Apr 2000 22:47:35 GMT
Subject: Building python on HPUX-10.2
References: <38FE1392.D1C01D64@wag.caltech.edu>
Message-ID: <38fe34a4.1103181328@news1.on.sympatico.ca>

I have built python successfully on several HPUX 10.2 boxes using GCC.

If you are using the basic cc compiler, you might have problems. If
you have the full development version, I don't know. On one machine I
had to install GCC, (hmmm that might have been because of one of the
libraries I was linking now that I think if it.) GNU make is also a
good idea...

Another gotcha is that once compiled, it might fail the fcntl tests.
see
http://x22.deja.com/[ST_rn=ps]/getdoc.xp?AN=517003463&CONTEXT=956184259.663814174
for how to fix this

On Wed, 19 Apr 2000 13:14:10 -0700, "Richard P. Muller"
<rpm at wag.caltech.edu> wrote:

>I have successfully built python on a number of diffent platforms, but
>I'm having serious problems on HPUX 10.2. I go through the normal
>configure/make procedure, but the make crashes right away, when it is
>trying to link the first object (pgen), because it can't find the
>correct dl library (?).
>
>Has anyone successfully compiled Python on HPUX-10.*? If so, what
>configure string did you use?
>
>Rick
>



From hamish_lawson at yahoo.co.uk  Sun Apr  2 10:43:33 2000
From: hamish_lawson at yahoo.co.uk (Hamish Lawson)
Date: Sun, 02 Apr 2000 07:43:33 -0700
Subject: New Features in Python 1.6
References: <200004011740.MAA04675@eric.cnri.reston.va.us> <8c6mkg$8i$1@slb2.atl.mindspring.net>
Message-ID: <0a509c7e.1fbf25d6@usw-ex0101-006.remarq.com>

As I read Guido's announcement an increasing sense of unease
came over me. Then I saw the posting date. Ah, it must be an
April Fool joke. No, really ... It MUST be an April Fool Joke!

However (assuming it was a joke) some people seem to have taken
the bait and are quite concerned. Might it be a good idea now to
admit that it was a joke before more Fear, Uncertainty and Doubt
is spread?

Hamish Lawson


* Sent from RemarQ http://www.remarq.com The Internet's Discussion Network *
The fastest and easiest way to search and participate in Usenet - Free!



From drs at labs.agilent.com  Thu Apr 13 12:17:23 2000
From: drs at labs.agilent.com (David Smith)
Date: Thu, 13 Apr 2000 09:17:23 -0700
Subject: Comparing PythonWin and IDLE?
References: <IOjJ4.6669$HG1.192116@nnrp1.uunet.ca>
Message-ID: <38F5F313.832B4CE6@labs.agilent.com>

Warren Postma wrote:

>  I find once a module has
> been imported once, it often will not import again, even if I have changed
> the source (.py) file, the .pyc file does not appear to be regenerated, and
> any errors in the .py file can only be fixed by editing and saving the .py
> file, quitting the IDE and restarting. This appears to happen for both IDLE
> and Python. Anybody else seen this?

When you edit a file under Idle, Save it and then hit Control-F5.  Then
you can re-import the file.

	David Smith


From lewst at yahoo.com  Tue Apr  4 19:13:58 2000
From: lewst at yahoo.com (lewst)
Date: Tue, 4 Apr 2000 16:13:58 -0700 (PDT)
Subject: Starving for an Advanced Python Book
Message-ID: <20000404231358.17447.qmail@web2105.mail.yahoo.com>

"M.-A. Lemburg" <mal at lemburg.com> wrote:

> Fredrik Lundh has compiled such a cookbook. I don't have the URL
> handy, but I'm sure you'll find it on their www.pythonware.com
> site.

Yes, this is Fredrik Lundh's eff-bot guide to The Standard Python
Library which is an on-line book in e-Matter format.  This looks like
it would be a great book for what I need, but there are a few
problems.  First, eMatter is not currently available for Unix/Linux.
While I do have access to Microsoft Windows, I don't own such a
machine.  Because the eMatter license is tied to the particular
computer you install the book on, this pretty much rules it out for me
as a possibility which is unfortunate.  Second, at $12.00, the book
seems a bit expensive for what it is.  I can understand paying that
much for an eMatter version of a published book which has gone through
a rigorous and professional editing process, but not otherwise.
Nonetheless, I'd still buy it if it supported Linux.                 


__________________________________________________
Do You Yahoo!?
Talk to your friends online with Yahoo! Messenger.
http://im.yahoo.com



From pecora at anvil.nrl.navy.mil  Thu Apr 27 17:06:38 2000
From: pecora at anvil.nrl.navy.mil (Louis M. Pecora)
Date: Thu, 27 Apr 2000 17:06:38 -0400
Subject: Do I always have to write "self." ?
Message-ID: <270420001706386648%pecora@anvil.nrl.navy.mil>

I am two weeks plus into learning Python and have started on a first
real project.  Amazing language.  BUT, now that I've gotten going with
classes, I'm finding I hate seeing all those  "self."  prefixes on the
class variables.  Python is generally very clean, but the classes I
write use a lot of "per-instance" variables, hence seem to require
self.  As a result the expressions can become cluttered and harder to
read than C++ in this particular case -- imagine that!  Is there any
relief here?  Any (simple) way around writing that self. every time?

Thanks.  Love Python so far otherwise.


From gmcm at hypernet.com  Wed Apr 26 15:48:25 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Wed, 26 Apr 2000 15:48:25 -0400
Subject: How can I get the name of an object???
In-Reply-To: <eviltofu-66EBF2.14590526042000@news.earthlink.net>
Message-ID: <1255343973-62617861@hypernet.com>

Jerome Chan wrote:

> object.__name__ returns a tuple of all vars referencings the
> object?

Not at all. Not all objects have an attribute "__name__", and 
Python does not attempt to point back to referencing objects. 
Those objects having a __name__ attribute (eg, class objects, 
function objects) tend to use it for the name given when the 
object was defined.

class A:
  pass

Now A.__name__ is "A".

B = A
And so is B.__name__.


- Gordon



From Richard.Jones at fulcrum.com.au  Mon Apr 10 18:44:23 2000
From: Richard.Jones at fulcrum.com.au (Richard Jones)
Date: Tue, 11 Apr 2000 08:44:23 +1000
Subject: Python/Perl Popularity (Re: A Mountain of Perl...) 
In-Reply-To: Message from Martijn Faassen of 2000-Apr-10 22:37:18,
            <8ctl2u$bko$3@newshost.accu.uu.nl> 
Message-ID: <200004102244.IAA03513@envy.fulcrum.com.au>

[Martijn Faassen]
> Richard Jones <Richard.Jones at fulcrum.com.au> wrote:
> > [tom__98 at my-deja.com]
> >> I don't know about others, but I frequently end up using
> >> Perl instead of Python, even though I greatly prefer Python
> >> as a language.
> 
> >    Tom you little troll-meister. Half those arguments you cite are bollocks
> > if you _really_ knew and "greatly prefer" Python. Try again.
> 
> It's considered good Usenet style to include at least an attempt at
> refuting arguments, instead of just saying they're bollocks. :)
> 
> Is this an example of a flaming Python advocacy as talked about earlier
> in the thread? :)

   bah!


> PHYTON-RULZEZ!!!-ly yours,

   You forgot PERL SUKZ!!!!


       Richard who is tired of the endless "vs" crap in the newsgroup :)





From neilh at hare.net.au  Tue Apr 25 20:18:28 2000
From: neilh at hare.net.au (Neil Hodgson)
Date: Wed, 26 Apr 2000 10:18:28 +1000
Subject: Sourceforge and Python
References: <oq8zy1zqto.fsf@titan.progiciels-bpi.ca>
Message-ID: <007b01bfaf14$f33114b0$3cdd3fcb@neil>

> So, merely beginning to explore the possibility, I'm asking for advice
> on this list, as to if there is environments offering a set of services
> similar to `sourceforge', but that would be especially Python-using,
> or Python-aware.  Of course, and also given some recent saga reported on
> this list, stability and availability are parts of the requirements :-).

   I don't know about a more Python specific site, but I moved part of
Scintilla to SourceForge a couple of months ago. Most of it is OK, but some
aspects of SourceForge are more difficult than they should be.

   The worst to me is file uploading where you upload each file over FTP to
a staging area (shared with all the other projects) then use a slow set of
web pages to move them file by file (with no remembering of common input
between files) into your download area. If your distribution is only one or
two files then it will be OK, but for a large set of downloadables, it is
too hard.

   Aspects which are good at SourceForge include having MailMan to set up
mailing lists with archives and no advertising. Your site may be found by
people searching or browsing SourceForge. CVS access is normally good
although there are problems on Windows with SourceForge's choice of SSH
version _1_ for write access.

   Neil




From donn at u.washington.edu  Wed Apr  5 17:35:33 2000
From: donn at u.washington.edu (Donn Cave)
Date: 5 Apr 2000 21:35:33 GMT
Subject: Some suggestion for a GUI kit evolved on Perl
References: <8cg7oo$m10$1@newsg4.svr.pol.co.uk>
Message-ID: <8cgbj5$n1s$1@nntp6.u.washington.edu>

Quoth "Makhno" <mak at imakhno.freeserve.co.uk>:
| I have written a GUI kit called Orchard. It is essentially a code library
| with c-like functions which when called correctly will create windows,
| buttons and other controls. I have had great success writing an interface
| for this with Perl, and now I'm about to attempt a Python module.
| One of the main points of Orchard is to keep the GUI-creation code as quick
| and as neat as possible. Because Perl ignores whitespace, I have designed
| the module in such a way so that the user can take advantage of this and use
| whitespace indicate GUI structure. For example, the following code creates a
| window with a button and a label in it:
|
| Orchard::NewOrcWindow
| (
|     title=>"Test"
| )->Add
| (
|     OrcVGroup
|     (
|         OrcButton,
|         $label=OrcLabel
|     )
| )->Show;
|
| As you can see, the window consists of a button and a label arranged
| horizontally. A ref is kept to the label so it can be changed in later code,
| and the window's Add() method returns a ref to the window so that Show() can
| be called on it.

(Looks to me like the button and label are arranged vertically?)

| The problem I'm having is that Python has reserved whitespace (to indicate
| program scope). Having looked at the Tk modules, I see it doing the
| equivalent of:
|
| window=Orchard.NewOrcWindow(title="Test")
| button=OrcButton
| label=Label
| vgroup=OrcVGroup
| window.add(vgroup)
| vgroup.add(button)
| vgroup.add(label)
| window.Show
|
| The problem is that this code stores references to everything (There's often
| no need to keep references to buttons, it's looked after internally), it's
| not as easy to read, and is easily complicated by GUI additions.
|
| An alternative would be to use the style I've used with Perl; creating GUI
| elements in the contructors of other GUI elements.
|
| Orchard.NewOrcWindow(title=Text).Add(OrcVGroup(...
|
| But without whitespace to help separate the elements I end up with a long
| line that looks even worse than before. Separating the long line to several
| smaller lines doesn't help readablity either (it just ends up all wrapped
| round rather than structured). And the line has to be split up when I want
| to keep a reference to the element anyway.
|
| This complication is precisely the sort of thing I wished to avoid when
| creating Orchard!
| Does anybody have any advice as to how I could 'structure' Python code to
| make neat, extendable GUI code? Essentially it comes down to what I can use
| to indicate GUI structure. I can't use whitespace, so what can I use?

I hope I'm not missing the point here, but I'm not sure you can't
use white space.  For example, following is legal Python:

label = OrcLabel
NewOrcWindow(
    title = "Test"
).Add(
    OrcVGroup(
        OrcButton,
        label
    )
).Show()

The "label" variable has to be assigned outside the function call,
that's just the way it is.  Other than that, we're looking at a
couple of changes around the parentheses.  I changed "Show" to "Show()",
assuming you don't mean to return the bound function but rather the
result of its execution.

If it does the same thing as the Perl example, it's not all that different
in appearance.

	Donn Cave, donn at oz.net


From fig at oreilly.com  Mon Apr  3 12:39:25 2000
From: fig at oreilly.com (Stephen R. Figgins)
Date: Mon, 03 Apr 2000 16:39:25 GMT
Subject: Python Topics
Message-ID: <38E8C989.3BDE6D00@oreilly.com>

I am working on setting up a Python Developer's Center for the O'Reilly
Network.  One of my tasks is to select topics to organize existing
resources on the web.   I only want 10 of these, and I want them to be
useful for people looking for particular resources.  Here is what I have
so far: 

Topic			Keywords
Systems Programming	Network, Sockets, Signals, Sys Admin
Development Tools	Idle, IDE, Debugging, SWIG
Education		CP4E, Python in the Classroom
User Interfaces		GUI, Tk, GTK, Gnome 
Database		XML, ODBC, Python Database
Internet		CGI, ZOPE, Applets
Jpython			Java, Servlets, Applets
Graphic			3D, PyOpenGL
Scientific		Numerical Python, Astronomy, Genetics
Distributed		COM, DCOM, CORBA


What do you think?  Are there resources out there that really don't fit
in any of these?  Are these useful divisions?


From paulb at infercor.no  Wed Apr 26 06:31:17 2000
From: paulb at infercor.no (Paul Boddie)
Date: Wed, 26 Apr 2000 12:31:17 +0200
Subject: mxODBC error trapping
References: <A63AFB20111ED311AC5500A0C96B7BF5015F0F4D@orsmsx36.jf.intel.com> <390640A8.F1392B0F@webamused.com>
Message-ID: <3906C575.8B933429@infercor.no>

Joshua Macy wrote:
> 
> "Daley, MarkX" wrote:
> >
> > Traceback (innermost last):
> >   File "<pyshell#1>", line 1, in ?
> >     collect.collect()
> >   File "C:\PROGRA~1\Python\collect.py", line 50, in collect
> >     except dbi.OperationalError:
> > AttributeError: OperationalError
> >
> > Here is the printout of the dbi module:
> >
> > ['DATE', 'NUMBER', 'RAW', 'ROWID', 'STRING', 'TYPES', '__doc__', '__file__',
> > '__name__', 'dataError', 'dbDate', 'dbRaw', 'dbiDate', 'dbiRaw',
> > 'integrityError', 'internalError', 'noError', 'opError', 'progError']
> 
>   According to the dbi module dictionary (which you've printed), it
> looks like it should be opError, not OperationalError.  Or you could
> just use an unqualified except, and be sure of catching any error that
> the dbi can raise.

You should be looking in the module where the 'connect' function was defined.
So, if you used 'ODBC.iODBC.connect' to connect to the database, you will find
that 'OperationalError' is defined in 'ODBC.iODBC' too:

>>> dir(ODBC.iODBC)
['BIGINT', 'BINARY', 'BIND_USING_PYTHONTYPE', 'BIND_USING_SQLTYPE', 'BIT',
'CHAR', 'Connect', 'DATE', 'DATETIME_DATETIMEFORMAT', 'DECIMAL', 'DOUBLE',
'DataError', 'DatabaseError', 'DriverConnect', 'Error', 'FLOAT', 'INTEGER',
'IntegrityError', 'InterfaceError', 'InternalError', 'LONGVARBINARY',
'LONGVARCHAR', 'NUMERIC', 'NotSupportedError', 'OperationalError',
'ProgrammingError', 'REAL', 'SMALLINT', 'SQL', 'STRING_DATETIMEFORMAT', 'TIME',
'TIMESTAMP', 'TIMEVALUE_DATETIMEFORMAT', 'TINYINT', 'TUPLE_DATETIMEFORMAT',
'VARBINARY', 'VARCHAR', 'Warning', '__builtins__', '__doc__', '__file__',
'__name__', '__path__', '__version__', 'apilevel', 'connect', 'error',
'errorclass', 'mxODBC', 'mxODBCType', 'mxODBCursorType', 'paramstyle',
'sqltype', 'threadlevel']

It would be "expected" that errors and exceptions were defined in the same
module as the classes and functions, but I suppose the structure of mxODBC makes
this assumption a little uncertain. Still, 'dir' used liberally often tells you
what you need to know.

Paul


From shogan at iel.ie  Wed Apr 26 06:34:49 2000
From: shogan at iel.ie (Shaun)
Date: Wed, 26 Apr 2000 11:34:49 +0100
Subject: how would i make a HTML type UI?
Message-ID: <000f01bfaf6b$0c9c6200$2801000a@iel.ie>

how would i create a HTML type user interface using python which would take
in say, -name-address-date of birth- and record it with the intention
of struct.pack-ing the values entered,
thanks
Shaun
=====================
Shaun Hogan
Interactive Enterprise Ltd.
alt. E-mail : shaun_hogan at yahoo.com
+353 86 8342529






From falknor at worldspy.net  Sun Apr  2 12:51:46 2000
From: falknor at worldspy.net (Falknor)
Date: 2 Apr 2000 09:51:46 -0700
Subject: Hmm, little Tk question.
Message-ID: <20000402165146.20535.cpmta@c000.sfo.cp.net>

An embedded and charset-unspecified text was scrubbed...
Name: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20000402/c32889e5/attachment.ksh>

From ivanlan at callware.com  Wed Apr 12 14:18:58 2000
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Wed, 12 Apr 2000 12:18:58 -0600
Subject: Off-topic
Message-ID: <38F4BE12.A0C73D1D@callware.com>

Hi All--
This is not precisely about Python, but it does affect my *usage* of
Python on redhat 6.1.

Why don't #! scripts seem to work anymore?  #!/usr/bin/python and
#!/usr/bin/perl scripts don't work.  Error is

bash:  ./h2n:  No such file or directory

but "perl h2n" gives:

Must specify one -d and at least one -n.
I give up ... sorry.

Which is the correct behaviour.

h2n starts with
#!/usr/bin/perl
blah ....

There are no funny things going on with line-endings, no control
characters, no extra spaces.

This all works just fine on redhats 5.2 and 6.0, but it fails
consistently with 6.1.

<in-a-heightened-state-of-pisstivity>-ly y'rs,
Ivan
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
http://www.pauahtun.org 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
Author:  Teach Yourself Python in 24 Hours



From btang at pacific.jpl.nasa.gov  Tue Apr 18 12:44:21 2000
From: btang at pacific.jpl.nasa.gov (Benyang Tang)
Date: Tue, 18 Apr 2000 09:44:21 -0700
Subject: Lists
References: <A63AFB20111ED311AC5500A0C96B7BF5015F0F2C@orsmsx36.jf.intel.com>
Message-ID: <38FC90E5.455DC023@pacific.jpl.nasa.gov>

Is following what you want to do?


a = [-5, -4, -3, -2, -1, 0, 1, 2, 3, 4, 5]
print a
b = 5
for item in range(len(a)):
        try:
                print b / a[item]
        except ZeroDivisionError:
                pass



"Daley, MarkX" wrote:
> 
> This feels like a real newbie question, but here is the code that is causing
> my question:
> 
> # Learning exceptions
> 
> def test():
>         a = [-5, -4, -3, -2, -1, 0, 1, 2, 3, 4, 5]
>         print a
>         b = 5
>         for item in a:
>                 try:
>                         print b / a[item]
>                 except ZeroDivisionError:
>                         pass
> 
> Here is the output of this code under IDLE:
> 
> [-5, -4, -3, -2, -1, 0, 1, 2, 3, 4, 5]
> 5
> 2
> 1
> 1
> 1
> -1
> -2
> -2
> -3
> -5
> 
> Why is the list being processed in reverse?  The exception works fine, but
> the math is backwards.  I don't recall this happening before.
> 
> TIA
> 
> - Mark

-- 
<> Benyang Tang <>


From shapr at uab.edu  Fri Apr 28 11:43:03 2000
From: shapr at uab.edu (Shae Erisson)
Date: Fri, 28 Apr 2000 15:43:03 GMT
Subject: parser and tokenize modules: sample code?
Message-ID: <3909B42D.2E564923@uab.edu>

Can someone point me to sample code for the parser and tokenize modules?
As Moshe Zadka says, the parser module doesn't seem easy to use. I've
read through the example.py for the parser module, but I'd like to see
more code if someone has some.

thanks,
-- 
sHae mAtijs eRisson (sHae at wEbwitchEs.coM) gEnius On fIre
  nossirEaehS?ikiw/nib-igc/moc.2c://ptth


From roy at popmail.med.nyu.edu  Tue Apr  4 23:10:25 2000
From: roy at popmail.med.nyu.edu (Roy Smith)
Date: Tue, 04 Apr 2000 23:10:25 -0400
Subject: What is python's language level?
References: <38EA1B8F.A797A130@lmco.com> <38E9DDD3.3313FF6@bioreason.com> <slrn8ekddq.gm3.zawrotny@gecko.sb.fsu.edu> <roy-D857DA.21212304042000@netnews.nyu.edu> <38EA9CD4.3473B910@callware.com>
Message-ID: <roy-CF57ED.23102504042000@netnews.nyu.edu>

Ivan Van Laningham <ivanlan at callware.com> wrote:
> Yes indeed.  Any n/troff source file could qualify.  Macro files, by
> definition, transcend obfuscated.  I wonder what's next up that ladder?

Have you ever looked at the output of a bib | tbl | eqn pipeline?


From al_amerin at yahoo.com  Mon Apr 24 23:28:09 2000
From: al_amerin at yahoo.com (Al-Amerrho H. Amerin)
Date: Mon, 24 Apr 2000 22:28:09 -0500
Subject: new linereading standard?
References: <3904D21B.333DD930@visionart.com>
Message-ID: <390510C9.A744D22E@yahoo.com>

Pete Shinners wrote:

>
> now i can code in a preferred style
>
> myfile = open('myfile.txt', 'r')
> for line in filelines(myfile):
>     print line
> myfile.close()
>
> as opposed to
>
> myfile = open('myfile.txt', 'r')
> while 1:
>     line = myfile.readline()
>     if not line: break
>     print line
> myfile.close()
>
>
> _____Pete Shinners______
>    pete at visionart.com

A beginner myself, I would write it like this:

inputlines = open('myfile.txt').readlines()
for line in inputlines:
    print line

Instead of the 'while 1: ....break' thingy.

Al-



From aa8vb at yahoo.com  Wed Apr 12 12:07:51 2000
From: aa8vb at yahoo.com (Randall Hopper)
Date: Wed, 12 Apr 2000 12:07:51 -0400
Subject: Python idiom: Multiple search-and-replace
In-Reply-To: <20000412144609.7026.qmail@web2006.mail.yahoo.com>
References: <20000412144609.7026.qmail@web2006.mail.yahoo.com>
Message-ID: <20000412120751.E2446939@vislab.epa.gov>

Darrell Gallion:
 |My solution is more work than Fredrik's, unless you
 |plan on making a lot of changes to the buffer.
 |Like many hundreds of patterns on a large buffer.
 |
 |Keep a list of changes to be performed.
 |(start, seqNumber, text, end)
 |
 |seqNumber stabilizes the sort of this list.
 |
 |If the start==end then delete a section.
 |
 |Sort the list and perform all operations at once.
 |This is like a transaction, which allows for some
 |other cool features. Such as detecting conflicting or
 |overlapping changes. 
 |
 |I have code if you want it.

For the dataset and change sizes I'm working with, Fredrick's re
replacement function approach gets the running time down sufficiently.
Thanks for the reply though.  I'll keep your technique in mind if I'm faced
with much larger datasets.

-- 
Randall Hopper
aa8vb at yahoo.com



From embed at geocities.com  Fri Apr  7 10:24:28 2000
From: embed at geocities.com (Warren Postma)
Date: Fri, 7 Apr 2000 10:24:28 -0400
Subject: freeze not working (win)
Message-ID: <memH4.4438$HG1.127494@nnrp1.uunet.ca>

I read the README in c:\python\tools\freeze, and it said to try this as a
starting point for Freeze:

>freeze -p c:\Python hello.py
(Using Python source directory)

[--snipped about 50 lines--]
freezing string ...
freezing tempfile ...
freezing types ...
generating table of frozen modules
C:\Python\win32\win32api.dsp: [Errno 2] The system cannot find the file
specified: 'C:\\Python\\win32\\win32api.dsp'
No definition of module win32api in any specified map file.
Warning: unknown modules remain: win32api


What do I do?  If i had to guess I'd say that the *.dsp is a Visual C++
Project file.
What does Freeze want with the source code to the Win32 extensions? Do I
have to get this file from somewhere? I looked on Mark Hammonds page and my
only option is to grab source from CVS. {Haven't done that before. don't
know how.} What do I do!?

Warren




From senn at maya.com  Wed Apr 19 13:23:17 2000
From: senn at maya.com (Jeff Senn)
Date: 19 Apr 2000 13:23:17 -0400
Subject: __del__ problem - more details provided. still stumped though.
In-Reply-To: "Gordon McMillan"'s message of "Wed, 19 Apr 2000 10:55:05 -0400"
References: <1255966388-25179470@hypernet.com>
Message-ID: <d7nmqa2i.fsf@SNIPE.maya.com>

"Gordon McMillan" <gmcm at hypernet.com> writes:
> Warren Postma wrote:
...
> >     def __del__(self):
...
> The problem is that the Py_Finalize logic has a real hard time 
> knowing in what order to zap things. By the time this __del__ 
> runs, the dbc module has already been zapped. Squirrelling 
> away a reference in the method definition will ensure it's 
> existence.  In general, a __del__ method (or anything called 
> from a __del__ method) cannot rely on any references outside 
> the object unless it uses this trick.

Yes -- a number of folks who come from (non-Python) "OO" languages
seem to fall into this particular trap.  Variables assigned in the
module look a lot like things that would be called "static (class)
variables" in other languages -- the fact is, that instances of classes
in Python can outlive the module in which the class was defined. So
this pattern of using "module variables" as "(class) statics" is a
dubious thing in Python -- it tends to work only because people don't
tend to write code that cleans up modules... it will show up other
places then __del__ if you try that (anyone tried calling reload()
aggressively? e.g. in a dynamic development env...)

If this is not in docs somewhere with an example it probably should
be...

-- 
-Jas




From spenrose at my-deja.com  Thu Apr 20 19:12:58 2000
From: spenrose at my-deja.com (spenrose at my-deja.com)
Date: Thu, 20 Apr 2000 23:12:58 GMT
Subject: Using python on the web
References: <MPG.1365adcfa8470eef9896aa@news> <8di2bk$l1s$1@newshost.accu.uu.nl> <MPG.136631472f892a119896ac@news> <_W2L4.986$GK5.20898@news2-win.server.ntlworld.com> <MPG.13665ef93eedbda29896ad@news> <9N4L4.499$wYl.191816704@newsb.telia.net>
Message-ID: <8do2te$kpr$1@nnrp1.deja.com>

We build medium-busy, CGI-intensive sites in Python. We were briefly
quite excited about Zope, but once we realized it locked us into Yet
Another Programming Language, we lost interest. DHTML is particularly
problematic, AFAI can tell, because it doesn't appear to allow the
segregation of HTML development from coding the way Python string
substitution does. We hire very skilled HTML people to do HTML and tell
them to leave our sustitution tags alone, rather than taking the time
and money to create novice DHTML people to do the same work. If we were
going to learn another language for production purposes, Perl, Java, and
C++ all have vastly larger infrastructures (code bases, labor pools,
documentation, etc.) than DHTML ever will. There are enough hassles in
using Python, which I love, because of its relatively small installed
base. No way am I going to something which is an order of magnitude less
supported.

The biggest hassle for us has been the weak MySQL support in Python. The
spec keeps changing and the interface is raw. (That should not be read
as a complaint directed at the people who have worked on the current
modules. I am tremendously grateful for their existence; I just wish the
community had more resources to throw at the problem.) We've written a
dictionary-based interface to sit on top of the modules, but it's
remains the source of much wailing and gnashing of teeth.

In article <9N4L4.499$wYl.191816704 at newsb.telia.net>,
  "Fredrik Lundh" <effbot at telia.com> wrote:
> Charley Horse <someone at somewhere.com> wrote:
> > It was just an unexpected surprise. I'm fairly new to all of this
and
> > always heard of CGI in the context of "you don't have to user CGI
> > anymore because of the fancy new blah blah technology we offer!"
>
> in Python land, that fancy technology is called
> Zope (www.zope.org).
>
> well worth checking out, but be careful -- your
> brain might explode.
>
> </F>
>
>


Sent via Deja.com http://www.deja.com/
Before you buy.


From mfletch at tpresence.com  Thu Apr 20 00:11:39 2000
From: mfletch at tpresence.com (Mike Fletcher)
Date: Thu, 20 Apr 2000 00:11:39 -0400
Subject: [Announce] mcf.vrml release 2.0b
Message-ID: <CC59868A0A76D311B1F50090277B321C1F328B@VRTMAIL>

mcf.vrml version 2.0 beta is now available at

http://members.home.com/mcfletch/programming/mcf_vrml.htm


>From the home page:

The mcf.vrml libraries are an extensive set of programming tools for
manipulating VRML 97 sceneGraphs. Their primary purpose is to allow the
creation of design-time tools to programmatically process and generate
complex VRML 97 content. Lately they have been used for generating VRML 97
sceneGraphs in response to Zope queries.

*	Full parser, supports all VRML 97 constructs.  Written on top of the
simpleparse parsing engine which is, in turn, built on top of the
mxTextTools library. Parses/loads at approximately 160,000 cps (with
considerable potential optimisation untapped). 
*	All sceneGraph/Node/Route/etceteras objects can be
pickled/unpickled.  This allows for storage in Zope databases for
server-side generation tasks. 
*	Full mcf.walker processing/rewriting engine sub-classes allow for
creating advanced processing applications (Tpresence's commercial
SkeletonBuilder H|Anim avatar converter is based on mcf.vrml and mcf.walker)

*	Node-creation syntax in Python closely mimics VRML syntax while
still standard Python 
*	Robust linearisation/output including prototype dependency
negotiation, USE negotiation, etcetera. Allows application developers to
largely ignore the question of sceneGraph ordering. 
*	Written in Python, which has a supportive developer community. 
*	Includes useful tools including material stripper for 3DSMax
pre-lighting support, and transform zeroing (convert coordinates to
world-space) 
*	Open Source (Python/BSD) license (see distribution for details)

>From the release notes:

2000/04/19 -- version 2.0 beta
This represents the first public release after re-writing the core node
structures (prototype, node, etc.) and revising all scripts to use the new
(simplified) interfaces (surprising minor revisions, giving the level of the
change).

The re-writing was triggered by switching to the SimpleParse parser
generator from the older mcf.pars parser generator (it was simpler to write
a new builder/parser using the SimpleParse generator's more reliable parse
tree than to retro-fit the original fragile converter).  In doing so, the
memory overhead for parsing long files with many root-level nodes has been
reduced.  Overall speed appears somewhat better, but not dramatically so.

Enjoy all,
Mike
__________________________________
 Mike C. Fletcher
 Designer, VR Plumber
 http://members.home.com/mcfletch



From breiter at usf.Uni-Osnabrueck.DE  Wed Apr 12 10:56:16 2000
From: breiter at usf.Uni-Osnabrueck.DE (Bernhard Reiter)
Date: 12 Apr 2000 14:56:16 GMT
Subject: Python/Perl Popularity (Re: A Mountain of Perl...)
References: <200004102111.HAA03380@envy.fulcrum.com.au>
    <8d0pcq$7ti$1@nnrp1.deja.com> <g0ssm03o.fsf@dan.resnet.rochester.edu>
    <8d149t$jo6$1@nnrp1.deja.com>
Message-ID: <8d22qg$oai$1@newsserver.rrzn.uni-hannover.de>

In article <8d149t$jo6$1 at nnrp1.deja.com>,
	tom__98 at my-deja.com writes:

>>     tom>  -- How can I browse the Python documentation without firing
>>     tom> up a web browser?
>>
>> Convert it to text format. It's latex, you can make it into anything
>> you want, and do anything you want with it.
> 
> Well, what I'm suggesting is that it might be a good idea
> to make a command-line accessible version of the documentation,
> together with a command line tool, available as part of the
> standard distribution.

lynx is a command line tool and will work fine and fast.
	Bernhard

-- 
Professional Service around Free Software                (intevation.net)  
The FreeGIS Project				            (freegis.org)
Association for a Free Informational Infrastructure            (ffii.org)


From greg at cosc.canterbury.ac.nz  Wed Apr 19 23:53:02 2000
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Thu, 20 Apr 2000 15:53:02 +1200
Subject: __del__ problem - would adopting Garbage Collection fix this?
References: <aw2L4.8907$HG1.245743@nnrp1.uunet.ca> <7ojL4.9424$HG1.254543@nnrp1.uunet.ca> <Z%oL4.9643$HG1.257533@nnrp1.uunet.ca> <m31z41iyhg.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <38FE7F1E.168E914A@cosc.canterbury.ac.nz>

Michael Hudson wrote:
> 
> "Warren Postma" <embed at geocities.com> writes:
> 
> > 2. Is this convincing proof, or not, that Garbage Collection would
> > be a Good Thing.
> 
> I don't think so.  You still have to destroy things; true gc might
> help I suppose, but I'm fairly sure it doesn't solve the problem
> completely.

Actually, it might. The reason that Py_Finalize
explicitly destroys module namespaces is to break
the circular references that occur between the
module dictionary and the functions defined therein.
With true GC, breaking references in an arbitrary 
order like that would no longer be necessary.

Although another set of problems would then crop
up, concerning in what order to finalise objects
that reference each other. Last time this was
discussed, I think we decided that the simplest
option was simply not to call the __del__ methods
of unreachable objects participating in cycles.
The __del__ method would only be called if the
object was collected as a result of its refcount
dropping to 0.

The destructors of C objects would still have
to be called, so that resources external to the
Python system (files, windows, etc.) would be
reclaimed. Such C destructors would have to be
careful about making any callbacks to Python
code (preferably not do it at all).

Another option would be to call all __del__
methods of all the objects about to be reclaimed
first, then free the memory (after checking that
none of them have been resurrected).

Either way should reduce the chance of having
something disappear out from under you unexpectedly
in a __del__ method.

By the way, does anyone know what the recently
announced GC patch does about __del__ methods?

-- 
Greg Ewing, Computer Science Dept,
+--------------------------------------+
University of Canterbury,	   | A citizen of NewZealandCorp, a	  |
Christchurch, New Zealand	   | wholly-owned subsidiary of USA Inc.  |
greg at cosc.canterbury.ac.nz	   +--------------------------------------+


From sigma_seitse at hotmail.com  Mon Apr 17 13:43:49 2000
From: sigma_seitse at hotmail.com (sigma seitse)
Date: Mon, 17 Apr 2000 17:43:49 GMT
Subject: guestbook
Message-ID: <20000417174349.71830.qmail@hotmail.com>

hi.

i'm trying to write a guestbook in python. well,
it should consist additional feature like this:
it should enable the owner to change and response to visitors' comments and 
show it out.
have you got anything like this or maybe you know an address or smth, where 
I could find something
like this?

a newbie..:)....
______________________________________________________
Get Your Private, Free Email at http://www.hotmail.com




From gmcm at hypernet.com  Sun Apr  2 17:43:33 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Sun, 2 Apr 2000 17:43:33 -0400
Subject: NameError when assigning dictionary values
In-Reply-To: <20000402203338.21701.qmail@web2105.mail.yahoo.com>
Message-ID: <1257410679-53490486@hypernet.com>

lewst writes:

> This simple problem is driving my crazy.  I can't figure out why I get
> a NameError with this program.
> 
>   # add each key-value pair to the dictionary and
>   # then for each value, create an empty list.
>   mydict = {}                         
>   for key,value in [ (".", "one"), ("+", "two") ]:
>       mydict[key] = value
>       value = []
>   print mydict
>   print one,two
> 
> Whe I run this I get "NameError: one".  Why aren't my 2 empty lists
> named "one" and "two" being created?

You're creating two empty lists, both named "value".The first 
is thown away by creating the second. Nowhere are "one" and 
"two" variable names, hence the NameError. 

- Gordon



From culliton at clark.net  Sat Apr  1 15:40:15 2000
From: culliton at clark.net (Tom Culliton)
Date: Sat, 01 Apr 2000 20:40:15 GMT
Subject: New Features in Python 1.6
References: <000301bf9c15$23a18f20$182d153f@tim>
Message-ID: <P8tF4.16091$0o4.101397@iad-read.news.verio.net>

In article <000301bf9c15$23a18f20$182d153f at tim>,
Tim Peters <tim_one at email.msn.com> wrote:
>[Tom Culliton]
>> OK, it's April 1st so maybe I'm just not getting the joke but...

>i-especially-love-that-"split"-guesses-which-arg-is-the-splittee-ly y'rs  -

That scared me too...


From kevinmcd at NOSPAM.linuxsolutions.demon.co.uk  Wed Apr 26 11:47:08 2000
From: kevinmcd at NOSPAM.linuxsolutions.demon.co.uk (Kevin McDermott)
Date: Wed, 26 Apr 2000 16:47:08 +0100
Subject: Perl Documentation SHOCK!
Message-ID: <39070F7C.8B6AA986@NOSPAM.linuxsolutions.demon.co.uk>

Hmmm,

Doesn't look like anybody's posted this before:

>From the CPAN available file perlpalmdoc-doc-500503-1.tar

README

"Core Perl Documentation Formatted for your PalmPilot!

'Finally, you can have perl on your Palm!'

'An indispensable reference!'

created with:
    pod2html 
    and
    htmltodoc from the palm-doctoolkit
    pyrite (some Python scripts for dealing with PalmOS 'stuff') 

Converted by:
    Robert Spier <rspier+cpan at speed.seas.upenn.edu>
    
PalmPerlDoc comes in two formats:
    'doc'  -> standard Palm doc format
    'teal' -> fancier tealdoc format with links and fonts

The entire core documentation requires about 1.2 MB of disk space.  
Space in your palm will vary.

ChangeLog:
    Aug 06 1999 -> first release to CPAN"

"Hmmm, gotta get a man to do a man's job?"

Kevin


From moshez at math.huji.ac.il  Fri Apr  7 14:45:41 2000
From: moshez at math.huji.ac.il (Moshe Zadka)
Date: Fri, 7 Apr 2000 20:45:41 +0200 (IST)
Subject: [Python-Dev] Round Bug in Python 1.6?
In-Reply-To: <200004062130.RAA26273@eric.cnri.reston.va.us>
Message-ID: <Pine.GSO.4.10.10004072044240.25661-100000@sundial>

On Thu, 6 Apr 2000, Guido van Rossum wrote:

> However, it may be time to switch so that "immediate expression"
> values are printed as str() instead of as repr()...

Just checking my newly bought "Guido Channeling" kit -- you mean str()
but special case the snot out of strings(TM), don't you

Trademark probably belong to Tim Peters.
--
Moshe Zadka <mzadka at geocities.com>. 
http://www.oreilly.com/news/prescod_0300.html
http://www.linux.org.il -- we put the penguin in .com




From mwh21 at cam.ac.uk  Wed Apr 12 14:30:51 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 12 Apr 2000 19:30:51 +0100
Subject: Off-topic
References: <38F4BE12.A0C73D1D@callware.com>
Message-ID: <m3zoqzgois.fsf@atrus.jesus.cam.ac.uk>

Ivan Van Laningham <ivanlan at callware.com> writes:

> Hi All--
> This is not precisely about Python, but it does affect my *usage* of
> Python on redhat 6.1.
> 
> Why don't #! scripts seem to work anymore?  #!/usr/bin/python and
> #!/usr/bin/perl scripts don't work.  Error is

Someone's taken /usr/bin/python away?  Tried strace-ing it?

I presume /bin/sh scripts are still working (wouldn't like to think
what happened if they went away).

HTH,
Michael

-- 
  incidentally, asking why things are "left out of the language" is
  a good sign that the asker is fairly clueless.
                                      -- Erik Naggum, comp.lang.lisp


From mwh21 at cam.ac.uk  Sun Apr 16 14:06:31 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 16 Apr 2000 19:06:31 +0100
Subject: Python paradigms
References: <8cmr70$b0e$1@pegasus.csx.cam.ac.uk> <8dbdj3$dgt$1@slb2.atl.mindspring.net> <8dcpom$fbc$1@slb3.atl.mindspring.net> <slrn8fk2kh.965.scarblac-spamtrap@flits104-37.flits.rug.nl>
Message-ID: <m3u2h1ex94.fsf@atrus.jesus.cam.ac.uk>

scarblac-spamtrap at pino.selwerd.nl (Remco Gerlich) writes:

> Pass functions?
> 
> def exp_if(test, true, false):
>   if test: return true()
>   else: return false()
>   
> x = exp_if(a is not None, lambda a=a,i=i: a[i].weeble, lambda: 0)
> 
> Getting silly, though :)

Watch out, you'll be writing code in cps (continuation passing style)
next...

Cheers,
M.

-- 
  well, take it from an old hand: the only reason it would be easier
  to program in C is that you can't easily express complex  problems
  in C, so you don't.                   -- Erik Naggum, comp.lang.lisp


From dkuhlman at netcom.com  Fri Apr 28 15:30:38 2000
From: dkuhlman at netcom.com (G. David Kuhlman)
Date: 28 Apr 2000 19:30:38 GMT
Subject: Remote debugging background processes
Message-ID: <8ecosu$t1e$1@slb2.atl.mindspring.net>

I would like to be able to re-route the output from and input to
the Python debugger (pdb) to another window, possibly on another
machine.  We have background, server-side scripts that we run from
within our Web server add-on that we want to be able to debug.

I searched for this in DejaNews and found something about
MrCreosote.  That seems hopeful.

I also looked briefly at the asynchat stuff in Medusa/ZServer.  I'm
a believer in Sam Rushing, so that looks very cool to me even
though I don't understand it yet.

Has someone already done something like this?

Is there support for this kind of thing in mod_python or some other
support for python in Apache?  Where should I look?

If anyone has some help, please let me know.  Otherwise, I think I
have a fun weekend project for myself.

  - Dave


From phd at phd.russ.ru  Mon Apr 24 07:07:33 2000
From: phd at phd.russ.ru (Oleg Broytmann)
Date: Mon, 24 Apr 2000 11:07:33 +0000 (GMT)
Subject: compress_web.py
In-Reply-To: <3900A524.4A32674D@binary.net>
Message-ID: <Pine.LNX.4.21.0004241105310.6019-100000@fep132.fep.ru>

On Fri, 21 Apr 2000, Mike Hostetler wrote:

>		for i in newfiles:
>			tar = tar+ " " + i

   This is just

tar = string.join(newfiles, " ")

   Simpler and faster.

Oleg.            (All opinions are mine and not of my employer)
---- 
    Oleg Broytmann      Foundation for Effective Policies      phd at phd.russ.ru
           Programmers don't die, they just GOSUB without RETURN.





From tzman at bhmi.com  Tue Apr 11 08:36:52 2000
From: tzman at bhmi.com (T Zimmerman)
Date: Tue, 11 Apr 2000 07:36:52 -0500
Subject: Tcl vs Tkinter
Message-ID: <38f31c65$0$27527@wodc7nh0.news.uu.net>

    I again am looking for help on the use of 'alt-*' to access the menus in
any application written with Tkinter. I have tried the tcl examples that
were included with the download of Python and they work as expected. So, I
am confused as to why it does not work for Tkinter widgets.  The following
script should work and should open the 'File' menu when 'alt-f' is pressed.
On NT this does nothing.


==================================================================
import Tkinter

root = Tkinter.Tk()

button = Tkinter.Menubutton(text = 'File', underline = 0)
button.pack()
menu = Tkinter.Menu(button)
button.configure(menu = menu)
menu.add_command(label = 'Hello World', underline = 0)

root.mainloop()
==================================================================

Regards,
Travis Zimmerman
tzman at bhmi.com




From mjhand at concentric.net  Fri Apr 14 03:32:12 2000
From: mjhand at concentric.net (Manus Hand)
Date: 14 Apr 2000 03:32:12 EDT
Subject: Questions on 1.6a2's string methods
Message-ID: <38F6C904.76B538E5@concentric.net>

1.  I see that string objects now support (as methods) most of the
    functions from the string module.  Among these are:
        upper, lower, split, strip, find
    and surely others.  My question is, why is capwords (which seems
    to be in the same genre as upper and lower) not a method?

2.  Along these same lines, since split() became a method of the
    string type, wouldn't it make sense to make join() a method of
    the list type?

3.  Are the classlike standard types (list, dictionary, and now
    string) equipped with a __dict__ attribute?  I can see the names
    of all functions supported by a user-defined class by saying
    className.__dict__.keys(), but I cannot see the list of methods
    for the string type (at least not in the same way).  Thus my
    need to ask silly questions like #1 above (maybe capwords is
    there by some other name??)

Please copy me on any response.  Thanks!

Manus


From loewis at informatik.hu-berlin.de  Tue Apr 18 10:30:37 2000
From: loewis at informatik.hu-berlin.de (Martin von Loewis)
Date: 18 Apr 2000 16:30:37 +0200
Subject: modifying os.environ
References: <slrn8f9mn1.ven.quinn@zloty.ugcs.caltech.edu>
Message-ID: <p6qhfczo50y.fsf@informatik.hu-berlin.de>

quinn at zloty.ugcs.caltech.edu (Quinn Dunkan) writes:

> Is there any way to delete things from the environment in python?  A
> look at os.py reveals that _Environ only overrides __setitem__ to
> putenv.  So how do you *delete* environment values?  perl does it,
> and a look at the source seems to indicate that it doodles with the
> C environ directly.  So perhaps python needs a delenv _Environ can
> use or something, or implement os.environ in C?

Some systems provide unsetenv(3), but that is not generally available.

> Or perhaps there's a better way to do the (common) thing I'm trying
> to do: clean the environment to insure some minimum sane values.  I
> can't make a copy of the environ and then pass to execve() since
> that doesn't help much for popen and system.

You could implement popen and system yourself, starting from popen2
(for example).

Regards,
Martin


From mjhand at concentric.net  Mon Apr 17 01:01:37 2000
From: mjhand at concentric.net (Manus Hand)
Date: 17 Apr 2000 01:01:37 EDT
Subject: Here is some starship.python.net news
Message-ID: <38FA9A30.33DD1994@concentric.net>

Not good news, but news.  This from the people at mcsp.com, the
hosts of the ailing starship python.  Does anyone know who the
"responsible person" spoken of below might be and how the Python
community can help bring the problem to his attention??

Manus (who has this fear that the "responsible person" has a
python.net e-mail address, so he's not getting the messages being
left for him :-)

Forwarded message:

From: "Paul Wolotsky" <pdw at mcsp.com>
Date: Sun, 16 Apr 2000 06:33:51 -0400
Organization: Medical Computer Systems

Yes - we are aware that the machine is down - we have left messages
for our contact of the responsible person for the machine - but have
had no reply - we finally shut the machine down the other day after
several days of repetitive crashes and disk errors.  There is nothing
that we can do until the responsible parties come in and service the
machine!

-----------------------------------------------------------
Paul Wolotsky
President, Medical Computer Systems
1729 21st Street, NW
Washington, DC 20009-1101


From simpsons at kom.auc.dk  Wed Apr 26 15:37:40 2000
From: simpsons at kom.auc.dk (Thomas Rasmussen)
Date: Wed, 26 Apr 2000 19:37:40 GMT
Subject: global variables or inheritence
Message-ID: <28p8zy0irgc.fsf@lada.kom.auc.dk>

Hi

I have quite a big problem. I have two different files, where I in one
file want to declare a variable and then in the other file be able to
access and use this variable (no need to change it), but I just can't
seem to get it to work... Is there any nice way of doing this easily?
I have tried the global var but this doesn't seem to work.

Anyone got a good idea?

Thomas
-- 
KOM Network student helper

"To alcohol! The cause of - and solution to - all of life's problems!"
-- Homer Simpson


From ajung at sz-sb.de  Sat Apr 29 10:31:58 2000
From: ajung at sz-sb.de (Andreas Jung)
Date: Sat, 29 Apr 2000 16:31:58 +0200
Subject: How to get the python object from a variable name ?
Message-ID: <20000429163158.A4091@yetix.sz-sb.de>

Inside a C program I need to get a pointer to the python object
for a object or variable by its variable name in the current
namespace of the interpreter. What I want to do ? My programs
gets names of variables and I need to get the str() representation
for given variable/object names as a C (char *). Maybe a stupid
question but I've no idea :)

Merci,
Andreas



From embed at geocities.com  Wed Apr 19 16:48:55 2000
From: embed at geocities.com (Warren Postma)
Date: Wed, 19 Apr 2000 16:48:55 -0400
Subject: __del__ problem - would adopting Garbage Collection fix this?
References: <aw2L4.8907$HG1.245743@nnrp1.uunet.ca> <7ojL4.9424$HG1.254543@nnrp1.uunet.ca>
Message-ID: <Z%oL4.9643$HG1.257533@nnrp1.uunet.ca>

I have received a helpful not explaing that the C function Py_Finalize does
not
always correctly figure out the right order in which to destroy things.

The usual hack to work around this is like this:

        def myclass:
            def __del__(self,foo=foo): #ensure symbol foo is not freed
                foo.bar() # call foo method in destructor without worry

Another similar hack I am now using:

        def myclass:
            _foo=foo
            def __del__(self):
                self._foo.bar()  # class variable _foo (1 per class)

My questions:

1. Shouldn't freeing modules in Py_Finalize be in the reverse order of the
way in which the modules were created?  Could a module-destruction order be
explicitly maintained by the system?

2. Is this convincing proof, or not, that Garbage Collection would be a Good
Thing. Is JPython immune to these bizarre behaviours?

3. What about adding a __finalize__ method to objects.  Python could then
finalize all objects first, and then delete them, and the rule is you ONLY
EVER USE __del__ when you Absolutely Have to, and with some known issues
(such as this). In other words, all objects are finalized in Py_Finalize,
then destroyed. When an object is deleted with del(obj), it is finalized
first, then freed.

4. Until such point as there is a __finalize__, I am actually going to
explicitly call a finalization method, then delete objects. Side effects on
Destructors can arguably be a Bad Thing.

Any other thoughts?

Warren




From dgoodger at bigfoot.com  Mon Apr 10 00:16:26 2000
From: dgoodger at bigfoot.com (David Goodger)
Date: Mon, 10 Apr 2000 00:16:26 -0400
Subject: indexed dictionaries?
In-Reply-To: <Pine.BSF.4.10.10004092225040.1361-100000@chi.pair.com>
References: <Pine.BSF.4.10.10004092225040.1361-100000@chi.pair.com>
Message-ID: <B516CDD9.17E6%dgoodger@bigfoot.com>

on 2000-04-09 22:30, Michal Wallace (sabren) (sabren at manifestation.com)
wrote:
>  I'm looking for a cross between a dictionary and a list,
> such that keys() always returns keys in the order in which they were
> added, and that instance[0] refers to the first key...
...
> Anyone got something like this?

I have implemented a list/dictionary hybrid in my SGF Parser Library (class
Node in http://gotools.sourceforge.net/sgflib/ ). It doesn't do ordered
keys(), but it does do indexing by integer (note that in this library, no
numeric dictionary keys will ever be used). It may give you a few ideas.

Please send me a copy of your final implementation of indexed dictionaries;
I'd be interested to see what you come up with.

-- 
David Goodger    dgoodger at bigfoot.com    Open-source projects:
 - The Go Tools Project: http://gotools.sourceforge.net
 (more to come!)



From akira.kiyomiya at autodesk.com  Wed Apr 26 14:44:10 2000
From: akira.kiyomiya at autodesk.com (Akira Kiyomiya)
Date: Wed, 26 Apr 2000 11:44:10 -0700
Subject: Simple Function Question
Message-ID: <8e7ddt$18n@autodesk.autodesk.com>

Hi,  Could you explain this in dummy's language?

This function prints out '4' and I have a feeling that x=1 and n=3, so
add1(3) prints '4'.

However, I have a difficlut time trying to see how these two arguments
assigned to both x (especially x) and n.

Thanks

Akira


# Return a function that returns its argument incremented by 'n'
def make_incrementer(n):
    def increment(x, n=n):
        return x+n
    return increment

add1 = make_incrementer(1)
print add1(3)  # This prints '4'





From johnm at magnet.com  Wed Apr 19 10:44:53 2000
From: johnm at magnet.com (John Mitchell)
Date: Wed, 19 Apr 2000 10:44:53 -0400
Subject: Very Horrible Question About Goto's
In-Reply-To: <A63AFB20111ED311AC5500A0C96B7BF5015F0F33@orsmsx36.jf.intel.com>
Message-ID: <Pine.SGI.4.05.10004191040010.6784-100000@lemur.magnet.com>

On Wed, 19 Apr 2000, Daley, MarkX wrote:

> It depends on what the GOTOs are doing.  If they are providing a GOSUB-like
> feature, you can do the same thing in Python by defining the task and then
> invoking it in other functions.  If they are providing flow control, it may
> be a little trickier. 

> -----Original Message-----
> From: Richard Jones [mailto:richard.jones at ibahealthcare.com]
> Sent: Wednesday, April 19, 2000 12:09 AM
> 
> I'm trying to create a converter from our own internal language (horrid)
> into python.  My only problem so far is that the original language uses
> *LOTS* of goto's.  Is there any way that this can be replicated within
> python ?

I have two suggestions:

1) the only valid use of goto's in C I've seen looks like this:

	if (bogus)
		goto cleanup
	do stuff;
	return(okay);
cleanup:
	close file handles;
	do cleanup
	return(error)

One way to do this in Python is using Exceptions, which in practice is
much cleaner:

	status = error
	try:
		do stuff
		status = okay
	except Bogus:
		cleanup

	return status

Also checkout try/finally for this type of thing.


2) Often, when porting a program from a procedural language (C, Perl) to
Python, you'd be better off *rewriting* the entire program.  Yes, I know
it's a pain; but for some reason whenever I do this the code *shrinks* by
about 30-50%, and is more flexible and easier to read.

I used to write procedural code, then get sick of it after ~150 lines and
refactor it into Objects and Methods (ie: cut-and-paste).  Resulting
program was shorter and much, much easier...


just my 2c


- j




From kcazabon at home.com  Sun Apr  9 16:39:32 2000
From: kcazabon at home.com (Kevin Cazabon)
Date: Sun, 09 Apr 2000 20:39:32 GMT
Subject: Tkinter:  centering windows on screen
References: <ak1I4.121418$Dv1.1306619@news1.rdc1.bc.home.com> <38F0C6A7.10878609@dkfz-heidelberg.de> <38F0C95D.1F7E853@dkfz-heidelberg.de> <xR4I4.274$rc9.188784640@newsb.telia.net>
Message-ID: <8U5I4.121484$Dv1.1314709@news1.rdc1.bc.home.com>

Thanks Fred... It'll work, but that means that I'll actually have to do some
work to get it positioned right!  q:]

I guessed it wouldn't be as easy as widget.place(CENTER).  Oh well...

Is there a Python or Tkinter call to return the screen resolution?  I know I
can probably get it through Win32, but I'd rather make it cross-platform...
I'll do some digging.

Kevin.

"Fredrik Lundh" <effbot at telia.com> wrote in message
news:xR4I4.274$rc9.188784640 at newsb.telia.net...
> Kilian Lorenz <kilian.lorenz at dkfz-heidelberg.de> wrote:
> > >> that works but doesn't help me much because I'm interested in the
> > >> return value. So I would really like to know how to call Tk commands
> > >> from within Python because .... blah blah
> >
> > Now I also found out how this works ... pretty easy with
> > <widget>.tk.call(<command>, <parameter>, <next parameter> ...)
> >
> > You just have to browse Tkinter.py and get the answers to all your
> > questions !  :)
>
> if you browsed the source, how come you didn't find the
> corresponding Tkinter methods?
>
> windows geometry:
>
http://w1.132.telia.com/~u13208596/tkintrobook/x9867-window-geometry-methods
> .htm
>
> file dialogues:
> http://w1.132.telia.com/~u13208596/tkintrobook/x1164-data-entry.htm
>
> (direct pointers to the current draft; official version at
> www.pythonware.com will be synchronized next week).
>
> </F>
>
>




From jmassung at magpiesystems.com  Mon Apr 17 12:34:15 2000
From: jmassung at magpiesystems.com (Jeff Massung)
Date: Mon, 17 Apr 2000 10:34:15 -0600
Subject: How do I make a constant global variable in Python Please
References: <vCj7OA2LfMfb=znR7yTHOvmiZAkd@4ax.com>
Message-ID: <rZGK4.4613$Uq2.3181@news.uswest.net>

Along the same lines... is there a way to say a variable must be declared
before use? I know Perl has something like this. I just want to avoid
something like this following:

ate = 1
while ate:
    # do something here
    ...
    are = 1

This will loop infinitely just because of a typo (that if the above were
added I would get an error saying that "are" wasn't declared).

Jeff

cmfinlay at magnet.com.au wrote in message ...
>How do I make a constant global variable in Python Please.
>or just a constant.
>or just a global.
>
>I bought a Python Book. however I can't find this info in it.
>
>Please e-mail me at cmfinlay at magnet.com.au




From francesco.contini at tiscalinet.it  Tue Apr  4 18:05:23 2000
From: francesco.contini at tiscalinet.it (FrancescO)
Date: Wed, 5 Apr 2000 00:05:23 +0200
Subject: Problem with gzip module
Message-ID: <8cdojr$69n$1@lacerta.tiscalinet.it>

Hi!
I can't understand how is the right way to use the gzip mudule to perform a
compression of a list of files, that i must read with gunzip or winzip
ecc...

I have read some examples from starship.python.com but...

I must write a cron script to zip log files from a directory, now i used the
zlib module, but for decompress I must use the uncompress method of the
zlib... the program do it very well, but the compressed files must be readed
from winzip from a Winnt on the net...
does anyone explain me, maybe with an example?

Thanks!





From sabren at manifestation.com  Fri Apr 14 06:54:59 2000
From: sabren at manifestation.com (Michal Wallace (sabren))
Date: Fri, 14 Apr 2000 06:54:59 -0400 (EDT)
Subject: how do I know to what module an object belongs?
In-Reply-To: <m3zoqxf5pk.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <Pine.BSF.4.10.10004140652030.15615-100000@chi.pair.com>

On 14 Apr 2000, Michael Hudson wrote:


> Presumably dbc is not a Python class, but an extension type.

yup.

 
> See what type(dbc).__name__ says; maybe you can special case for that?
> 
> Otherwise, I think you're hosed.  I'm not sure it's a good idea;
> it sounds rather fragile to me.

it gives "mxODBC Connection". Huh. :/

I wound up just writing a wrapper module around ODBC.Windows, and
it seemed to do the job. Not the best answer, but it worked. 


Thanks,

- Michal
-------------------------------------------------------------------------
http://www.manifestation.com/         http://www.linkwatcher.com/metalog/
-------------------------------------------------------------------------




From cdh at mail.ala.net  Wed Apr  5 22:41:51 2000
From: cdh at mail.ala.net (cdh)
Date: Wed, 5 Apr 2000 21:41:51 -0500
Subject: special attributes (and greetings from a new list user)
Message-ID: <002201bf9f71$ad333b80$4f186bce@ala.net>

Hi,
My name is Chris Hickman and though I've been using python for a while now,
I'm new to the list. I've been a member on another prog-lang list before and
found it very fun and interesting. 

Below is a module I'd like to submit for comments. It make's attribute handling
a bit easier as it takes care of __getattr__ and __setattr__ for you. The document
string should explain it fairly well.

Happy python programming,
Christoper D. Hickman

###############################################################################

# Author: Christopher D. Hickman
# Date: 4/3/2000

"""
  Module: spec_attr.py
    Simplifies adding special get and set handlers for attributes.

  Exports:
    class spec_attr:
      Defines the methods __getattr__ and __setattr__ implement magic handlers for special attributes.
      The special attribute must be defined with an _ (underscore) prepended.
      The methods are defined as getattr<special attr> and setattr<special attr>.
      The special attributes will then automatically call their handlers whenever accessed without
      the leading underscore.
      You may also implement "virtual attributes" which are handlers only with no special attribute 
      supporting them.
      See example below for usage.
      
    class spec_attr_declarative:
      Same as the spec_attr class except it disallows implicit attribute creation.

    class SpecialAttributeError(AttributeError):
      Error class raised by spec_attr_declarative.__setattr__() when attempting to implicitly 
      create a new attribute by assigning a value to a non-existant attribute. 

  Example:
    >>> from spec_attr import spec_attr
    >>> class foo(spec_attr):
    ...  _x = 0 # special attribute x
    ...  _y = 0
    ...  _z = 0
    ...  count = 5
    ...  def getattr_x(self): # special get method for x
    ...    print "x attribute of %s was read" % repr(self) # verbalize all accesses to x
    ...    return self._x # caveat: you must use the underscore prepended form for all "gets"
    ...                   #         of a special attribute when defining it's get method.
    ...                   #         otherwise, infinite recursion will ensue.
    ...  def setattr_x(self, value):
    ...    print "x attribute of %s was changed from %s to %s" % (self, self._x, value)
    ...    self._x = value # same caveat applies for sets in set methods as for gets in get methods.
    ...  def setattr_y(self, value):
    ...    if value % 2:
    ...      raise "y must be even!"
    ...    self._y = value
    ...  def getattr_xyz(self): # a virtual get method (there's no xyz attribute)
    ...    return self.x, self.y, self.z
    ...  def setattr_xyz(self, value): # a virtual set method
    ...    self.x, self.y, self.z = value
    ...  def getattr_decr_count(self):
    ...    val = self.count
    ...    self.count = self.count - 1
    ...    return val
    ...  
    >>> a = foo()
    >>> a.x
    x attribute of <__main__.foo instance at 10e2bf0> was read
    0
    >>> a.x = 12
    x attribute of <__main__.foo instance at 10e2bf0> was changed from 0 to 12
    >>> a.xyz = (1,2,3)
    x attribute of <__main__.foo instance at 10e2bf0> was changed from 12 to 1
    >>> a.xyz
    x attribute of <__main__.foo instance at 10e2bf0> was read
    (1, 2, 3)
    >>> while a.decr_count:
    ...  print a.count
    ...  
    4
    3
    2
    1
    0
    >>> a.xyz = (6, 7, 8)
    x attribute of <__main__.foo instance at 10e2bf0> was changed from 1 to 6
    Traceback (innermost last):
      File "<interactive input>", line 1, in ?
      .
      .
      .
      File "<interactive input>", line 16, in setattr_y
    y must be even!
    >>> 

"""

class SpecialAttributeError(AttributeError):
  pass

class spec_attr_declarative:

  # important things to remember about __getattr__: 
  #   1. __getattr__ is *only* called when a "real" attr is not present
  #   2. hasattr() can cause recursion to __getattr__,  and traps any exceptions generated
  def __getattr__(self, attr):
    # see if there is a method defined to handle gets for attr
    if attr[:8] == 'getattr_':
      # we've either recursed because getattr_<attr> doesn't exist, (in which case the exception
      # generated here is trapped by the hasattr() call from the caller), or we tried to call
      # the non-existant getattr_<attr> directly.
      raise AttributeError, attr
    if hasattr(self, 'getattr_' + attr):
      return eval ('self.getattr_' + attr + '()') 
    # if no method is present, try to return the value of the special attribute behind the one
    # we're trying to access
    if attr[:2] == '__':
      # we're probably checking for ('_' + _<attr>) due to recursive calls, a private attribute, 
      # or an undefined instance of one of python's "magic" attributes (like __repr__), 
      # so give up.
      raise AttributeError, attr
    if hasattr(self, '_' + attr):
      return eval ('self._' + attr)
    # otherwise, it simply doesn't exist
    raise AttributeError, attr
 
  # unlike __getattr__, __setattr__ is always called
  def __setattr__(self, attr, value):
    # if it's in the __dict__ it's not special, so process it quickly
    if self.__dict__.has_key(attr):
      self.__dict__[attr] = value
      return
    # see if there is a method defined to handle sets for attr
    if hasattr(self, 'setattr_' + attr):
      exec ('self.setattr_' + attr + '(value)')
      return
    # if no method is present, try to set the value of the special attribute behind the one
    # we're trying to set
    if hasattr(self, '_' + attr):
      exec ('self._' + attr + ' = value')
      return
    # otherwise attr could be present in the <instance>.__class__.__dict__ (or somewhere), in which 
    # case hasattr() will return true even though the above tests failed. Create a real attribute 
    # in the instance's __dict__. (this mirrors the default behavior in this case when no __setattr__ 
    # method is defined)
    if hasattr(self, attr):
      self.__dict__[attr] = value
      return
    # for the declarative version, don't allow de-facto creation of a new attribute. (explicitly use 
    # __dict__ to circumvent this)
    raise SpecialAttributeError, "Attempted to create a new attribute %s for the declarative class instance %s" % (attr, repr(self))

class spec_attr(spec_attr_declarative):
  def __setattr__(self, attr, value):
    try:
      spec_attr_declarative.__setattr__(self, attr, value)
    except SpecialAttributeError:
      # allow creation of new normal attributes (without explicitly using __dict__) in the 
      # non-declaritive version.
      self.__dict__[attr] = value 


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20000405/d8abbf92/attachment.html>

From R.Brodie at rl.ac.uk  Thu Apr 13 06:45:02 2000
From: R.Brodie at rl.ac.uk (Richard Brodie)
Date: Thu, 13 Apr 2000 11:45:02 +0100
Subject: Why is if a in dict.keys() so slow.....
References: <38F5A075.CDFA32D9@bedford.waii.com>
Message-ID: <8d48ff$ih0@newton.cc.rl.ac.uk>

"Martin Franklin" <martin.franklin at waii.com> wrote in message
news:38F5A075.CDFA32D9 at bedford.waii.com...
> Just thought i'd throw this out there....
>
> This is very slow:-
>
> if split[0] in map_dict.keys():
>
> Compared to:-
>
> try:
>     print 'found it!',map_dict[split[0]]
>
> The dictionary had about 40000 key:value pairs and I needed to
> look through it 250000 times.....
>
> Question is why?

I suppose creating a 40000 element list 250000 times might have
something to do with it. Have you tried map_dict.has_key(split[0]) ?





From jae at ilk.de  Tue Apr 25 20:26:14 2000
From: jae at ilk.de (Juergen A. Erhard)
Date: Wed, 26 Apr 2000 02:26:14 +0200
Subject: new linereading standard?
In-Reply-To: <mt1z3uyr8k.fsf@astron.berkeley.edu> (message from Johann
	Hibschman on 25 Apr 2000 11:23:23 -0700)
References: <3904D21B.333DD930@visionart.com> <390510C9.A744D22E@yahoo.com> <R2g+MIAFMUB5EwUl@jessikat.demon.co.uk> <mt1z3uyr8k.fsf@astron.berkeley.edu>
Message-ID: <26042000.2@sanctum.jae.ddns.org>

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

>>>>> "Johann" == Johann Hibschman <johann at physics.berkeley.edu> writes:

[J W B]

    Johann> To go off on a wild tangent, this reminds me of the whole
    Johann> map/lambda functional programming fracas.  See, I agree
    Johann> with most people that I don't really want to see code like

    Johann>   map(lambda s: string.split(string.strip(s), ',')[0],
    Johann>       open('data.dat').readlines())

    Johann> I much prefer

    Johann>   out = []
    Johann>   for s in open('data.dat').readlines():
    Johann>      fields = string.split(string.strip(s), ',')
    Johann>      out.append(fields[0])

    Johann> But, if I'm doing this at the interactive prompt, what I end up typing
    Johann> is most often something like:

    >>>> lines = open('data.dat').readlines()
    >>>> lines = map(string.strip, lines)
    >>>> lines = map(lambda l: string.split(l, ','), lines)
    >>>> out   = map(lambda fs: fs[0], lines)

    Johann> So, basically, I find the functional stuff much more
    Johann> useful from the command line than I do from actual
    Johann> programs.  I think of the right transform, then I apply
    Johann> it.  Interesting.

Actually, I find the third example more readable than the other two...

   map(string.strip, lines)

is, IMHO, more readable than a for loop doing the same.

Conciseness *can* improve readability.

Though it also depends on what you're used to... (maps etc are *more*
readable in Lisp... than in Python)

    Johann> Heh.  I'm free-associating here, please bear with
    Johann> me. Anyone else have this experience?

Free-associating?  Yes, I know that experience... ;-)

Bye, J

PS: /me learned lisp a pretty long time ago from a four-part article
in `Scientific American' by Douglas Hofstadter... :-) (Actually, it
was a German translation in `Spektrum der Wissenschaft')

- -- 
J?rgen A. Erhard      eMail: jae at ilk.de      phone: (GERMANY) 0721 27326
          My WebHome: http://members.tripod.com/Juergen_Erhard
           "Outside of a dog, a man's best friend is a book;
           inside of a dog, it's very dark." --  Groucho Marx
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.1 (GNU/Linux)
Comment: Use Mailcrypt and GnuPG <http://www.gnupg.org/>

iEYEARECAAYFAjkGN6UACgkQN0B+CS56qs1+mwCfa9+imyRXQaLqwLXzb0uCkc0W
XbYAoJAlRQENPIdp6dUZqvi8oIwCuaeL
=7ZPG
-----END PGP SIGNATURE-----



From rkiendl at gmx.net  Fri Apr 14 07:15:36 2000
From: rkiendl at gmx.net (Robert Kiendl)
Date: Fri, 14 Apr 2000 13:15:36 +0200
Subject: what is pythonwin?
Message-ID: <38F6FDD8.DA2CC785@gmx.net>

i ve got embedded the normal pyhton (python15.dll) into a windows
mfc-app.

now i want to use some of the win32ui stuff, but cannot load that
mysterious *.pyd files, whereas normal python.exe from the winall
distribution does it. are these dll's? is pythonwin not based on the
normal python? do i have to handle the pythonwin source-distribution?
link another dll?
i do not get the right track from the standard docs.

can anyone point me to the correct coherences?

robert


From tismer at tismer.com  Wed Apr 19 14:15:21 2000
From: tismer at tismer.com (Christian Tismer)
Date: Wed, 19 Apr 2000 20:15:21 +0200
Subject: Very Horrible Question About Goto's
References: <C5A7720BDEB7D3118A5200E018C200A037EA43@EXCHANGE> <m3g0sihz9h.fsf@atrus.jesus.cam.ac.uk> <m3d7nmhz4z.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <38FDF7B9.9781CFDA@tismer.com>


Michael Hudson wrote:
> 
> Michael Hudson <mwh21 at cam.ac.uk> writes:
> 
> > Gaetan Corneau <corg at copernic.com> writes:
> >
> > > I was surprised to learn that Python doesn't have a goto statement.
> > > Would it be difficult to add?
> >
> > No. The necessary opcodes are there (well, the one you need:
> > JUMP_ABSOLUTE).  You could use bytecodehacks to implement it, if you
> > liked...
> 
> Oh, hang on: the block stack would cause problems; how would you get
> 
> def f(x):
>     label: bob
>     for i in x:
>         if x:
>             goto: bob
> 
> to work? (by declaring it to be an error, I suppose).

This situation is exactly where continuations would
be superior than just a goto. The goto would jump
at bob's continuation, which simply does not have
this block.
But jumping from outside the block into the block
would also not work with continuations, unless
the block had a chance to start up before.

ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaunstr. 26                  :    *Starship* http://starship.python.net
14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     where do you want to jump today?   http://www.stackless.com



From wtopa at dmcom.net  Sat Apr  8 12:12:47 2000
From: wtopa at dmcom.net (Wayne Topa)
Date: Sat, 8 Apr 2000 12:12:47 -0400
Subject: starting python
In-Reply-To: <8cnkfi$at7$1@porthos.nl.uu.net>; from maybe@lucky.com on Sat, Apr 08, 2000 at 05:51:11PM +0200
References: <8cnkfi$at7$1@porthos.nl.uu.net>
Message-ID: <20000408121247.A18876@dmcom.net>

	Subject: starting python
	Date: Sat, Apr 08, 2000 at 05:51:11PM +0200

In reply to:Soundwave

Quoting Soundwave(maybe at lucky.com):
Soundwave> Hi, I want to learn a bit python, I know a bit C and PASCAL, but I am
Soundwave> intrested in other languages.
Soundwave> 
Soundwave> where can I start ????
Soundwave> 
Soundwave> Do you naad a compiler or something for python, and where can I find one ?
Soundwave> 
Soundwave> also, if anyone got a good tutor, please, give me the URL
Soundwave> 

You will find all of the above at
http://www.python.org

HTH
-- 
"As part of the conversion, computer specialists rewrote 1,500
programs; a process that traditionally requires some debugging.
-- USA Today, referring to the IRS switchover to a new computer system.  
_______________________________________________________



From jmassung at magpiesystems.com  Fri Apr 14 18:28:03 2000
From: jmassung at magpiesystems.com (Jeff Massung)
Date: Fri, 14 Apr 2000 16:28:03 -0600
Subject: TCP/IP protocols
Message-ID: <ZSMJ4.1099$Y14.27933@news.uswest.net>

I'm new to Python programming, but am very experienced in C/C++. Could
someone point me to a good reference for TCP/IP programming with Python
(pdf, book, anything :]) thanks.

Jeff




From tismer at tismer.com  Fri Apr 21 15:29:26 2000
From: tismer at tismer.com (Christian Tismer)
Date: Fri, 21 Apr 2000 21:29:26 +0200
Subject: compress_web.py
References: <3900A524.4A32674D@binary.net>
Message-ID: <3900AC16.11E4B6D5@tismer.com>


Mike Hostetler wrote:
...
> Email me if you have questions, comments, or sympathies.

Just a few comments.

...
> def compress_web(tar,days,webdir,bad_files):
> 
>         os.chdir(webdir)
> 
>         newfiles = []
> 
>         os.path.walk(webdir,find_time,newfiles)
> 
>         if (newfiles != []):

It is a common pattern just to ask
          if newfiles:

Also, you don't need the braces (probably from using C).

For the next loop, I'd sugest to name the variable
as what it is: a filename, for sure not an "i".

                  for fname in newfiles:

>                 for i in newfiles:
>                         tar = tar+ " " + i
>                 os.system(tar)
>         else:
>                 print "\n### no files found that needed archived\n"
> 
> #####################################
> # this function is called by the os.path.walk for each file in webdir
> def find_time(newfiles, dirname, names):
> 
>         strlen = len(webdir)+1
>         time_limit = time.time() - (86400 * days)
> 
>         for file in names:
>                 filename= dirname+"/"+file
> 
> # Wow, this line sucks.  But it works.

Yes, it sucks, since it is recommended practice to use

                  filename = os.path.join(dirname, file)

which handles all cases of slashes for you, for all OSes.

Congrats for using Python now, with such success in the first place!

ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaunstr. 26                  :    *Starship* http://starship.python.net
14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     where do you want to jump today?   http://www.stackless.com



From bawolk at pacbell.net  Sat Apr  8 19:46:39 2000
From: bawolk at pacbell.net (Bruce Wolk)
Date: Sat, 08 Apr 2000 16:46:39 -0700
Subject: Python Megawidgets - Pmw.Blt problem
Message-ID: <38EFC4DF.76F3C698@pacbell.net>

Has anyone been able to use the Pmw.Blt megawidget on a Win98 system? 
Pmw generally seems to work ok, but when I try the examples at

http://www.ifi.uio.no/~hpl/Pmw.Blt/doc/

they don't run.  I get errors like

Traceback (innermost last):
  File "C:\PROGRA~1\PYTHON\TOOLS\IDLE\ScriptBinding.py", line 131, in
run_module_event
    execfile(filename, mod.__dict__)
  File "C:\Program Files\Python\Bruce\HelloWorld1.py", line 5, in ?
    g = Pmw.Blt.Graph(master)       # make a new graph area
  File "C:\PROGRA~1\PYTHON\Pmw\Pmw_0_~1\lib\PmwBlt.py", line 260, in
__init__
    Tkinter.Widget.__init__(self, master, _graphCommand, cnf, kw)
  File "C:\Program Files\Python\Lib\lib-tk\Tkinter.py", line 1084, in
__init__
    self.tk.call(
TclError: invalid command name "::blt::graph"

Any ideas?

Bruce


From lewst at yahoo.com  Wed Apr  5 00:47:32 2000
From: lewst at yahoo.com (lewst)
Date: Tue, 4 Apr 2000 21:47:32 -0700 (PDT)
Subject: Starving for an Advanced Python Book
Message-ID: <20000405044732.10810.qmail@web2103.mail.yahoo.com>

Patrick Phalen <python-list at teleo.net> wrote:

> I haven't seen anyone mention The Quick Python Book in this thread.
> Also worth having on your bookshelf.

I originally thought this book was purely introductory so I ignored
it.  After checking it out this evening I actually decided to buy it.
It really looks like it will make a nice intro-to-intermediate Python
text while I wait for the next edition of Programming Python.


__________________________________________________
Do You Yahoo!?
Talk to your friends online with Yahoo! Messenger.
http://im.yahoo.com



From mhammond at skippinet.com.au  Mon Apr  3 06:09:18 2000
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Mon, 03 Apr 2000 10:09:18 GMT
Subject: Some bugs in Pythonwin b.128
References: <Pine.OSF.4.10.10003281852310.6437-100000@prfdec.natur.cuni.cz> <_NaE4.6119$V56.24967@news-server.bigpond.net.au> <8F0AAC6BEm69ph83qstudentssuse@127.0.0.1> <EUPF4.2880$1b4.14419@news-server.bigpond.net.au> <4cQF4.4791$74.76270@newsc.telia.net>
Message-ID: <i5_F4.3421$1b4.16633@news-server.bigpond.net.au>

"Fredrik Lundh" <effbot at telia.com> wrote in message
news:4cQF4.4791$74.76270 at newsc.telia.net...
> Mark Hammond wrote:
> > > Try changing these settings in your config file:
> > >     Shift-9           = <<paren-open>>
> > >     Shift-0           = <<paren-close>>
> >
> > That is excellent to know.  Does anyone know how I can
determine
> > the keyboard layout?  ie, how can I programatically determine
> > where the paren keys are?
>
> umm.  why should you?  shouldn't you either use
TranslateMessage
> and look for "(" or ")" WM_CHAR events, or use
ToAscii/ToUnicode to
> do the translation "inline"?

Ahhh - of course - thanks!  Pythonwin processes these on
WM_KEYDOWN, so does indeed need this translation.

> (in Tkinter, all you need to do is to bind to <Shift-(> and
<Shift-)>,
> but I supports you're not using that one here :-)

Well, we can't have that, can we :-)  <few hours later> OK - the
next version of Pythonwin will use these exact same binding names
in its .cfg file, and will be processed after the necessary
translation (ie, any builds later than 130 should have this
fixed!)

Thanks,

Mark.





From willfg at my-deja.com  Tue Apr 25 09:31:09 2000
From: willfg at my-deja.com (willfg at my-deja.com)
Date: Tue, 25 Apr 2000 13:31:09 GMT
Subject: Question about exception-handling mechanism
Message-ID: <8e46mt$4nh$1@nnrp1.deja.com>

Hello all,

   I'm new to Python, but find its exception handling mechanism much
more powerful than comparable languages; but a question was put to me
that being new I can't come up with a very articulate answer. A
colleague asked why in an exception handling mechanism you'd want the
ELSE block to be executed if you don't throw an exception as opposed to
a FINALLY block. Anyone used this feature in practice? Thanks in
advance for your input, -- Will


Sent via Deja.com http://www.deja.com/
Before you buy.


From bjorn at roguewave.com  Sun Apr 23 16:55:28 2000
From: bjorn at roguewave.com (Bjorn Pettersen)
Date: Sun, 23 Apr 2000 14:55:28 -0600
Subject: 'database' search
Message-ID: <39036340.BE82BEAD@roguewave.com>

I haven't done this before, so I thought I'd ask if anyone has any
experience...

I have a relatively stable database of between 1500 and 5000 items (new
insertions are probably going to be less than 100 during the lifetime of
the database).  The items are keyed on a text string.

I'm creating a gui where I want to implement field completion as the
user is typing, i.e. if the user types 'aa', all entries (or at least a
subset) with that prefix will be (unobtrusively) presented so the user
can use an alternate selection method (similar to the file open dialog
on win2k, although I'm not tied to any specific implementation).

I'm interested in experiences anyone might have with respect to both
which datastructure to use, and if any gui toolkit is better suited for
this.

The smallest datastructure I've found so far, is a 'prefix dictionary',
so if e.g. "aardwark" was in the database, the 'prefix dictionary' would
contain:

 pfdict = {
  'a':        ['aardwark', ...],
  'aa':       ['aardwark', ...],
  'aar':      ['aardwark'],
  'aard':     ['aardwark'],
  'aardw':    ['aardwark'],
  'aardwa':   ['aardwark'],
  'aardwar':  ['aardwark'],
  'aardwark': ['aardwark']
 }

The maximum size for this would be n * avg len of item, with the real
size being much(?) smaller with sufficient prefix sharing...

-- bjorn



From cjensen at bioeng.ucsd.edu  Thu Apr  6 20:16:22 2000
From: cjensen at bioeng.ucsd.edu (Curtis Jensen)
Date: Thu, 06 Apr 2000 17:16:22 -0700
Subject: C's ?: expression
Message-ID: <38ED28D6.2C5519ED@bioeng.ucsd.edu>

I while back I posted about how to simulate C's ?: expression in
Python.  There were quite a few responces to it.  Recently, I found on
pp135-136 of Mark Lutz's "Programmin Python" book that there is a way. 
If anyone is interested:

((TEST and [trueAction]) or [falseAction])[0]

where TEST is the boolean expression and tueAction is what to do if the
expression is true and falseAction is what to do if the expression is
false.

-- 
Curtis Jensen
cjensen at bioeng.ucsd.edu
http://www-bioeng.ucsd.edu/~cjensen/
FAX (425) 740-1451


From rcc at nospamthanks_jennic.com  Wed Apr 26 04:49:15 2000
From: rcc at nospamthanks_jennic.com (Robert Cragie)
Date: Wed, 26 Apr 2000 08:49:15 GMT
Subject: execl problem on linux
References: <8e63b2$6v2$0@216.39.162.232>
Message-ID: <newscache$3m8mtf$ewg$1@jenpc07.jennic.co.uk>

Michael Esveldt <dante at oz.net> wrote in message
news:8e63b2$6v2$0 at 216.39.162.232...
| This may be bordering on off topic but the following is happening with
| all the exec() methods...
|
| Python 1.5.2 (#0, Apr  3 2000, 14:46:48)  [GCC 2.95.2 20000313 (Debian
| GNU/Linux
| )] on linux2
| Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
| >>> import os
| >>> os.execl("/usr/bin/vim")
| Internal error: lalloc(0, )
| Internal error: lalloc(0, )
| bash-2.03$
|
| I'm at a loss, any ideas?

I think you might be better off using system(), or popen() (or popen2()) if
you need to pipe stuff into (and out of) the process. I found the following
works:

>>> import os
>>> os.system("/usr/bin/vim")

I'm not sure why execve() doesn't work, but I presume it might be something
to do with the command taking over the process which calls execve(), whereas
I presume system() etc. spawn a separate process. Maybe the real experts can
help here :-)

Robert




From portaldata at my-deja.com  Tue Apr 25 12:10:50 2000
From: portaldata at my-deja.com (portaldata at my-deja.com)
Date: Tue, 25 Apr 2000 16:10:50 GMT
Subject: how to wrap c++ software to python?
References: <8e3ggq$2r1d$1@news.cz.js.cn>
Message-ID: <8e4g20$eeq$1@nnrp1.deja.com>

HOW DO I INSTANCE A SHARED OBJECT IN PYTHON.
ALSO, HOW DO I POST A NEW MESSAGE IN THIS GROUP ?


In article <8e3ggq$2r1d$1 at news.cz.js.cn>,
  "fmr" <mrfx at 263.net> wrote:
> I will call c++ routine in python,how to wrap
> c++ routine for python.
>
> Is there any utils to work except SWIG.
>
>



Sent via Deja.com http://www.deja.com/
Before you buy.


From moshez at math.huji.ac.il  Sat Apr 29 04:15:35 2000
From: moshez at math.huji.ac.il (Moshe Zadka)
Date: Sat, 29 Apr 2000 11:15:35 +0300 (IDT)
Subject: Do I always have to write "self." ?
In-Reply-To: <8edc3l$7sj$1@nnrp1.deja.com>
Message-ID: <Pine.GSO.4.10.10004291114410.24197-100000@sundial>

On Sat, 29 Apr 2000 jsolbrig at my-deja.com wrote:

> Actually,
> What I'd like to see would for "self" to an official keyword rather than
> simply a convention. The more things that are "conventional" but not
> required, the more potential there is for occasional bugs.
> And occasional bugs are the WORST.

Here's a solution:
Search the deja-news archives for "selfnanny", and run it over your code.
It will warn you about all methods not beginning with "self".

--
Moshe Zadka <mzadka at geocities.com>. 
http://www.oreilly.com/news/prescod_0300.html
http://www.linux.org.il -- we put the penguin in .com




From eviltofu at rocketmail.com  Sat Apr 22 20:14:02 2000
From: eviltofu at rocketmail.com (Jerome Chan)
Date: Sun, 23 Apr 2000 00:14:02 GMT
Subject: Overloading the = operator?
Message-ID: <eviltofu-5E48DE.20140022042000@news.earthlink.net>

Is this possible?

class base:
   def __init__(self,q):
      self.list = q
   def __add__(self,other):
      # some type checking code
      return self.list + other.list

class another(base):
   def opr(self):
      # do something
      pass

class other(base):
   # do something too
   def work(self):
   pass

x = base([1,2])
y = another([2,4,3])

z = other([])

z = x + y
z.work()

where z now contains x.list + y.list???

I'm confused...


From zorro at zipzap.ch  Thu Apr 27 05:53:05 2000
From: zorro at zipzap.ch (Boris Borcic)
Date: Thu, 27 Apr 2000 11:53:05 +0200
Subject: The Simple Economics of Open Source
References: <14591.15630.192472.440375@goon.cnri.reston.va.us>
	 <3900611F.3B525BDA@cfdrc.com> <FtDG4z.774@world.std.com>
	 <slrn8g26tu.p7o.neelk@brick.cswv.com>
	 <14837CB4E98A72FE.8FD547BA6E928594.5A103CFE68327EB7@lp.airnews.net> <VLvUljAKNfA5EwPM@jessikat.demon.co.uk>
Message-ID: <39080E01.5DC5B4E@zipzap.ch>

Robin Becker wrote:
> 
> I don't think the pessimists amongst whom I count myself reject altruism
> totally. A purely altruistic action has benefit for others and none for
> the actor.

Makes me wonder. It is relatively obvious to argue that a non-existent
is the natural representation of the existents that resemble it most.
As long as the actor survives, there must exist ways to view aspects
of his/her further life as a benefit of the action. This leads to the
totem-and-taboo case of Jesus Himself. One might say that Jesus redeems
your argumentation by providing a admitted proof that perfect altruism
is in principle possible - thus forbidding to take perfect altruism in
your sense, as an adequate metaphor for the imperfect altruism that you
exclude. But I gather that this was not the intention of the Scripture.

I would satisfy myself with an action that is consistent with being
purely motivated by the benefit for others. Altruism makes sense as
a choice to behave while aiming at the greater common good *assuming
by fiat* a society of like, mutually altruistic agents, thus making
such a society *possible*, and in any case, such a society is
inconsistent with the "goal" implied by your overstrict
definition of altruism : depriving agents of any benefit, when they
are viewed one by one.

Regards,

BB
--
"Hope achieves the square root of the impossible"


From effbot at telia.com  Tue Apr 11 10:30:17 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Tue, 11 Apr 2000 14:30:17 GMT
Subject: Huh? func_defaults, default values in function calls
References: <8cv5qj$35p$1@zinc.intern.netconnect.no> <8cv8ra$hio$1@pegasus.csx.cam.ac.uk>
Message-ID: <ZFGI4.1192$Za1.20497@newsc.telia.net>

James "Wez" Weatherall wrote:
> You won't get this behaviour with something like:
>
> def myfun(x, a = 5):
>     a = x
>
> though, I don't think, because "5" isn't mutable.

footnote: that has nothing to do with mutability.  in
the following example, the list isn't modified either:

    def myfun(x, a = []):
        a = x

</F>




From niels at endea.demon.nl  Fri Apr 14 11:25:52 2000
From: niels at endea.demon.nl (Niels Diepeveen)
Date: Fri, 14 Apr 2000 17:25:52 +0200
Subject: IndexError and for x in list
References: <Pine.BSF.4.10.10004132041530.4951-100000@chi.pair.com>
Message-ID: <38F73880.885E6DCE@endea.demon.nl>


"Michal Wallace (sabren)" schreef:
>     def _toStringKey(self, key):
>         """Convert numeric keys into string keys. Leaves string keys as is"""
>         # handle numeric keys:
>         if _isNum(key):
>             if not (0 <= key < len(self.idx)):
>                 ## oddly enough, it is this IndexError here
>                 ## that allows you to do "for x in myIdxDict:"
>                 raise IndexError, `key` + " is out of bounds."
>             # convert it to a string key
>             key = self.idx[key]
> 
>         return key

Or, a bit simpler:
    def _toStringKey(self, key):
        """Convert numeric keys into string keys. Leaves string keys as
is"""
        # note that sequence indexes must be integers
        if type(key) is type(0):   # note that sequence indexes must be
integers
            return self.idx[key]   # this will raise the IndexError when
needed
        else:
            return key

> Basically, I don't want to rely on an undocumented feature that could
> go away sometime. Is IndexError the right way to simulate a list?

If you take the above approach, that problem will go away, since you
will simply pass on the error handling of a real list, whatever that may
be. To be on the safe side, you might also implement __len__ if that
isn't already done. That should take care of any alternative
implementation of 'for' that I can imagine. If you really want to do
everything a list can do, read 'Emulating sequence and mapping types' in
the language reference.

-- 
Niels Diepeveen
Endea automatisering



From vetler at ifi.uio.no  Mon Apr  3 07:28:54 2000
From: vetler at ifi.uio.no (Vetle Roeim)
Date: 03 Apr 2000 13:28:54 +0200
Subject: www.python.org
References: <lFKF4.120$wo4.48486@paloalto-snr1.gtei.net>
Message-ID: <yayln2va0eh.fsf@ganglot.ifi.uio.no>

* Brian Schmidt
> Anyone else having problems getting into www.python.org the last couple of
> days?

there was a posting here today or yesterda from the python.org
webmaster. iirc, the httpd had stopped for some reason. 

anyway, it's working now.


vr


From josephwinston at mac.com  Thu Apr 27 09:35:12 2000
From: josephwinston at mac.com (Jody Winston)
Date: 27 Apr 2000 08:35:12 -0500
Subject: Scripting and Gnome and KDE
References: <Pine.GSO.4.10.10004150741520.11261-100000@sundial> <3dk8hzb90u.fsf@newcnri.cnri.reston.va.us> <shk8hk2m68.fsf@tux.gnu.franken.de>
Message-ID: <wkln1zll9r.fsf@mac.com>

Karl EICHWALDER <ke at gnu.franken.de> writes:

> Andrew Kuchling <akuchlin at mems-exchange.org> writes:
> 
> > Then reality sets in: GNOME uses patched versions of all sorts of
> > things such as automake, but this is never clearly documented
> > anywhere.
> libtool might be an exception.  But normally it's enough to use recent
> versions of all involved tools.
> 
> > There's a large pile of libraries, and it's not clear which ones are
> > needed and which ones are obsolete;
> 
> autoconf resp. ./configure is considered to do all needed checks for
> you.  If it fails for some reason it's a bug; please, report it.  We'll
> try to improve the check and add a note to the README file.
> 
> The problem is, it works nicely for the devolopers and package makers
> (like me) who are familiar with the progress GNOME makes.
> 

I have to agree with Andrew because I have had these problems and
more.  I don't work on a Linux machine, but an SGI.  It was very
difficult to get the libraries that GNOME needs to work to run.  I
finally gave up on GNOME.  These problems are not just with GNOME and
its assorted libraries but with other Open Software tools such as
gimp.

I've been building and using Open Software since 1982 on all sorts of
platforms.  I've been using and devloping on Linux for over 5 years.
This includes several commercial applications (Visual Shell, Digital
Analysis for Pipeline corrosion and deformation) and custom device
drivers (stepper motors, ccd camera controller).

It looks like to me that we have gone back to the model that all the
world is a Vax and if it compiles and runs on the Vax there isn't any
problem with the software.  All the world isn't Linux and the only
compiler isn't gcc.

-- 
Jody Winston


From R.Brodie at rl.ac.uk  Fri Apr 14 09:36:33 2000
From: R.Brodie at rl.ac.uk (Richard Brodie)
Date: Fri, 14 Apr 2000 14:36:33 +0100
Subject: Future compatibility of the Python C API
References: <lwEJ4.7980$Ib7.101092@typhoon2.kc.rr.com> <q4FJ4.1198$1D.10037@news-server.bigpond.net.au>
Message-ID: <8d76t2$pig@newton.cc.rl.ac.uk>

"Mark Hammond" <mhammond at skippinet.com.au> wrote in message
news:q4FJ4.1198$1D.10037 at news-server.bigpond.net.au...

> > My big concern is Python 3000- what are the odds
> > that its C API will be backwards compatibility with Python 1.x?
>
> Not very good.  If it is a real problem, simply dont upgrade.  I would
> however expect the changes to cause compiler errors that are simple to
> fix - ie, very similar APIs, and unlikely to silently change behaviour.

I suppose that someone out there is working on persuading SWIG to
generate suitable wrapper code? It would be a pain if there were a
long transitional period when you had to go back to handcrafting
wrapper code. No doubt doing so for the whole of win32 would keep
you out of trouble a while ;)




From phd at phd.russ.ru  Fri Apr 21 03:43:23 2000
From: phd at phd.russ.ru (Oleg Broytmann)
Date: Fri, 21 Apr 2000 07:43:23 +0000 (GMT)
Subject: PyApache memory leak?
In-Reply-To: <8dni3s$1o1$1@nnrp1.deja.com>
Message-ID: <Pine.LNX.4.21.0004210741580.2712-100000@fep132.fep.ru>

On Thu, 20 Apr 2000 spenrose at my-deja.com wrote:
> We rolled in PyApache; the guy who did it is gone but the tar ball is
> labelled 4.19 so I assume that's the version.

   It is the latest released version. But I have never saw any memory
problem beeing running PyApache for lon periods. Are you sure it is
PyApache and not your programs/libraries/extensions?

Oleg.            (All opinions are mine and not of my employer)
---- 
    Oleg Broytmann      Foundation for Effective Policies      phd at phd.russ.ru
           Programmers don't die, they just GOSUB without RETURN.




From mwh21 at cam.ac.uk  Thu Apr 20 07:50:10 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 20 Apr 2000 12:50:10 +0100
Subject: Starship update: April 18
References: <3du2gzmiki.fsf@amarok.cnri.reston.va.us>
Message-ID: <m3r9c1gff1.fsf@atrus.jesus.cam.ac.uk>

akuchlin at mems-exchange.org (Andrew M. Kuchling) writes:
[mixed news]

Do you know how long it's going to be before python.net mail is
working again, ie. should I go running around and unsubscribe from all
these mailing lists I've subscribed to as mwh at python.net and
resubscribe as something that works?

Cheers,
M.

-- 
3. Syntactic sugar causes cancer of the semicolon.
     -- Alan Perlis, http://www.cs.yale.edu/~perlis-alan/quotes.html


From lewst at yahoo.com  Wed Apr  5 17:34:38 2000
From: lewst at yahoo.com (lewst)
Date: Wed, 5 Apr 2000 14:34:38 -0700 (PDT)
Subject: Starving for an Advanced Python Book
Message-ID: <20000405213438.11603.qmail@web2106.mail.yahoo.com>

fredrik at pythonware.com wrote:

> (first you dis my eBook using the usual publisher
> arguments, and now you have inside information on
> when PP2E will be out ;-)

Actually, I saw the error in my ways and ended up buying 
your eBook last night.  :-)

Regards.


__________________________________________________
Do You Yahoo!?
Talk to your friends online with Yahoo! Messenger.
http://im.yahoo.com



From btribble at blueshiftinc.com  Wed Apr  5 18:14:29 2000
From: btribble at blueshiftinc.com (Brett Tribble)
Date: Wed, 5 Apr 2000 15:14:29 -0700
Subject: os.utime limitation on Windows?
Message-ID: <senek66krfg118@corp.supernews.com>

I have the following routine:

def synctime(self, source, target):
    try:
        syncstat = os.stat(source)
    except os.error:
        self.statuswindow.insert('end', source + ': cannot get stats\n')
    try:
        os.utime(target, (syncstat[ST_ATIME], syncstat[ST_MTIME]))
    except os.error:
        self.statuswindow.insert('end', target + ': cannot change time\n')
    return

This works fine under NT on an NTFS drive, but doesn't seem to do anything
on a 95 box on a FAT16 partition. Is this a limitation with 95, FAT16,
Python, or a combination therein?


Thanks in advance,

Brett






From alex at somewhere.round.here  Thu Apr  6 10:55:05 2000
From: alex at somewhere.round.here (Alex)
Date: Thu, 06 Apr 2000 14:55:05 GMT
Subject: More about variables
References: <meh9-816CC5.10434906042000@news.cit.cornell.edu>
Message-ID: <m1wvmbxorf.fsf@h002078e00408.ne.mediaone.net>

> Let's say I had something like:
> 
> number_of_variables=raw_input('Enter number of variables: ')
> 
> if number_of_variables==1:
>    variable1=[]
> elif number_of_variables==2:
>    variable1=[]
>    variable2=[]

There is probably a more appropriate way of solving whatever problem is
behind this question.  How do you plan on accessing the variables once
they are created?

On the face of it, I think you might have more luck trying this sort of
arrangement:

number_of_variables=raw_input('Enter number of variables: ')
variables = []
for i in number_of_variables * [None]:
    variables.append ([])

However, if you are bent on using a separate name for each list, you
could do something like this:

for i in range (1, number_of_variables + 1):
    exec 'variable%i = []' % i
                
Alex.


From mjackson at wc.eso.mc.xerox.com  Thu Apr 13 13:05:46 2000
From: mjackson at wc.eso.mc.xerox.com (Mark Jackson)
Date: 13 Apr 2000 17:05:46 GMT
Subject: Complaint Dept: My shoes!
References: <20000413184853.D14644@xs4all.nl>
Message-ID: <8d4upa$45i$1@news.wrc.xerox.com>

Thomas Wouters <thomas at xs4all.net> writes:

> It entirely depends on what
> else you wish to do with your shoe. (Dont get me started on how Shoe should
> actually be a class with it's 'put_on' and 'throw' methods... ;)

Don't forget the ever-popular 'drop' method, which acts on the *other*
instance of the Shoe class. . .

-- 
Mark Jackson - http://www.alumni.caltech.edu/~mjackson
	An advertiser will happily make you feel bad about
	yourself if that will make you buy, say, a Bic pen.
				- George Meyer




From phd at phd.russ.ru  Fri Apr 28 08:40:34 2000
From: phd at phd.russ.ru (Oleg Broytmann)
Date: Fri, 28 Apr 2000 12:40:34 +0000 (GMT)
Subject: sleep function?
In-Reply-To: <8ebvud$l6h$1@nnrp1.deja.com>
Message-ID: <Pine.LNX.4.21.0004281240160.18766-100000@fep132.fep.ru>

On Fri, 28 Apr 2000 tiddlerdeja at my-deja.com wrote:
> Does python have a sleep function? e.g. sleep(60) to wait 60 seconds?

   from time import sleep

Oleg.            (All opinions are mine and not of my employer)
---- 
    Oleg Broytmann      Foundation for Effective Policies      phd at phd.russ.ru
           Programmers don't die, they just GOSUB without RETURN.




From hniksic at iskon.hr  Tue Apr  4 09:42:40 2000
From: hniksic at iskon.hr (Hrvoje Niksic)
Date: 04 Apr 2000 15:42:40 +0200
Subject: Tuples -- who needs 'em
In-Reply-To: Mike  Fletcher's message of "Tue, 4 Apr 2000 09:13:15 -0400"
References: <CC59868A0A76D311B1F50090277B321C1F324E@VRTMAIL>
Message-ID: <dpem8mugmn.fsf@mraz.iskon.hr>

Mike  Fletcher <mfletch at tpresence.com> writes:

> I think the point that is being made is this:
> 	The _value_ of a mutable object cannot (easily) be used as a
> dictionary key...
> 	i.e. in this sequence, what is the value of the final line...
> 
> s = [ 1,2,3 ]
> p = { s: 4, [2,3,4]: 5 }
> s[:] = [2,3,4]
> p[ [2,3,4] ]

The second line raises a TypeError, therefore the last line does not
have a meaningful value.

I only noted that Fredrik's sentence, as it was formulated, was wrong.
Mutable collections can be used as dictionary keys, as long as they
provide a meaningful hash function.



From pecora at anvil.nrl.navy.mil  Sun Apr 30 10:22:35 2000
From: pecora at anvil.nrl.navy.mil (Louis M. Pecora)
Date: Sun, 30 Apr 2000 14:22:35 GMT
Subject: Reading in strings -> numbers ??
References: <390A2D23.7029E557@fenx.com> <8ede36$9r6$1@nnrp1.deja.com>
Message-ID: <300420001022354625%pecora@anvil.nrl.navy.mil>

In article <8ede36$9r6$1 at nnrp1.deja.com>, <chris_barker at my-deja.com>
wrote:

> Well, the partner to file.write is file.read. What you are really
> looking for is a partner to '%f %f'%(3,4), which is essentially an
> sprintf(). So, what you want is an sscanf() or fscanf(). I've been
> looking for that for a while. What I would like is a version like the
> one in MATLAB, that gives you a whole matrix in one swipe.
> 
> Unfortunatly, it doesn't exist for python. Someone did make a Python
> sscanf module, but I never got around to compiling it. there is also a
> tablio (or something like that) module around. In the mean time, you can
> always use something like:
> 
> 
> list = map(float,string.split(file.readline)))
> 
> or, to use your example:
> 
> >>> map(float,string.split((" %e %e %d" % (1.2, -3.5, 10))))
> [1.2, -3.5, 10.0]
> 
> it doesn't work if you want mixed types, but it gets you by.

Chris,

Thanks. I'll try that.  

After three weeks of learning Python have I actually found a real wart? 
A common requirment in programming (especially for numerical stuff) is
to read in data that is often generated by other programs and other
people.  The common form is a "table" structure:

data11(white space)data12(white space)...data1m(return/newline)
data21(white space)data22(white space)...data2m(return/newline)
...
datan1(white space)datan2(white space)...datanm(return/newline/EOF)
EOF

So you're saying that reading in something as basic as this is a
"work-around?"  Sigh.

Thanks for your help.


From loredo at spacenet.tn.cornell.edu  Fri Apr 14 13:54:13 2000
From: loredo at spacenet.tn.cornell.edu (Tom Loredo)
Date: Fri, 14 Apr 2000 13:54:13 -0400
Subject: Need full example of C extension on Macintosh (CW)
References: <120420000714412887%pecora@anvil.nrl.navy.mil> <38F62C97.FD5073B@spacenet.tn.cornell.edu> <140420000717194110%pecora@anvil.nrl.navy.mil>
Message-ID: <38F75B45.B2B56BEC@spacenet.tn.cornell.edu>

Hi folks-

For those of you following this on c.l.p, this discussion is continuing
on the Pythonmac-SIG, where I've posted some CodeWarrior Pro 5 project
stationery for building extensions implementing simple math functions
and accessing NumPy arrays.  A web site with more documentation of
these examples is forthcoming.

-Tom Loredo


From sbarron at ansic.net  Thu Apr  6 11:39:09 2000
From: sbarron at ansic.net (Scott)
Date: Thu, 6 Apr 2000 15:39:09 GMT
Subject: Formatting Excel cells with PythonCOM
Message-ID: <38ECAF9D.ECA74709@ansic.net>

Hi,

I am looking for information regarding how to set the Date/Time format
for Excel cells.  When I enter my dates/times I use the DateTime module
and the COMDate member of my objects.  When I open Excel it displays
them as ###### until I manually set the date/time format.  How can I do
this from within my Python program?

Thanks,
Scott


From PClaerhout at CREO.BE  Thu Apr 13 04:04:33 2000
From: PClaerhout at CREO.BE (Pieter Claerhout)
Date: Thu, 13 Apr 2000 10:04:33 +0200
Subject: Python CGI running commands under NT
Message-ID: <2B1262E83448D211AE4B00A0C9D61B03BA72AC@MSGEURO1>

Hello,

I made a Python CGI-script that executes a DOS-command and
should return the result in a HTML page. The script looks
as follows:

  import os

  pipe = os.popen("ipconfig", 'r')
  text = pipe.read()
  del pipe

  print "Content-type: text/html"
  print
  print "IPConfig on AV2700"
  print "<br>"

  print '<pre>'
  print text
  print '</pre>'

If I run this script from the command line, it works fine, but
when I open it using Internet Information Server, all is shown
except the contents of the variable text. Any ideas what could
be wrong? Is it maybe a setting in IIS or are these options
not allowed in CGi-scripts?

Kind regards,


Pieter
Pieter_Claerhout at creoscitex.com



From fredrik at pythonware.com  Wed Apr 26 05:26:11 2000
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 26 Apr 2000 11:26:11 +0200
Subject: execl problem on linux
References: <8e63b2$6v2$0@216.39.162.232>
Message-ID: <00e201bfaf61$85bd56d0$0500a8c0@secret.pythonware.com>

Michael Esveldt wrote:
> This may be bordering on off topic but the following is happening with 
> all the exec() methods...
> 
> Python 1.5.2 (#0, Apr  3 2000, 14:46:48)  [GCC 2.95.2 20000313 (Debian 
> GNU/Linux
> )] on linux2
> Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
> >>> import os
> >>> os.execl("/usr/bin/vim")

in addition to the name of the executable, execl also expects
you to give it some arguments to put into sys.argv.

since argv[0] is supposed to contain the name of the program,
this should work:

    os.execl("/usr/bin/vim", "vim")

</F>




From embed at geocities.com  Mon Apr 17 12:06:50 2000
From: embed at geocities.com (Warren Postma)
Date: Mon, 17 Apr 2000 12:06:50 -0400
Subject: How do I make a constant global variable in Python Please
References: <vCj7OA2LfMfb=znR7yTHOvmiZAkd@4ax.com>
Message-ID: <wHGK4.8403$HG1.233034@nnrp1.uunet.ca>

<cmfinlay at magnet.com.au> wrote in message
news:vCj7OA2LfMfb=znR7yTHOvmiZAkd at 4ax.com...
> How do I make a constant global variable in Python Please.
> or just a constant.
> or just a global.
>
> I bought a Python Book. however I can't find this info in it.
>
> Please e-mail me at cmfinlay at magnet.com.au

A=1

Then don't change it!

Warren




From jhauser at ifm.uni-kiel.de  Wed Apr  5 12:32:48 2000
From: jhauser at ifm.uni-kiel.de (Janko Hauser)
Date: 05 Apr 2000 18:32:48 +0200
Subject: Clear screen?
References: <38eb6a36@207.71.36.3>
Message-ID: <87n1n87bkf.fsf@ifm.uni-kiel.de>

os.system('cls') # :-)

HTH
__Janko
-- 
  Institut fuer Meereskunde             phone: 49-431-597 3989
  Dept. Theoretical Oceanography        fax  : 49-431-565876
  Duesternbrooker Weg 20                email: jhauser at ifm.uni-kiel.de
  24105 Kiel, Germany


From shae at webwitches.com  Mon Apr 17 06:53:59 2000
From: shae at webwitches.com (Shae Erisson)
Date: Mon, 17 Apr 2000 10:53:59 GMT
Subject: Was Re: [ANNOUNCE] Garbage collection for Python, now Stackless comments
References: <20000407221750.A29279@acs.ucalgary.ca>  <m3r9cagx3c.fsf@atrus.jesus.cam.ac.uk> <l03102809b51d41895a80@[193.78.237.135]> <38F9F01B.57273052@tismer.com> <slrn8fjtao.2d4.quinn@drachma.ugcs.caltech.edu> <38FA05C9.442868E4@tismer.com>
Message-ID: <38FAEEC9.B5CED499@webwitches.com>

Christian Tismer wrote:
> Right. One should perhaps not advertize continuations in the first
> place, until people know them from school.
> That's why we try to build really useful stuff from them, familiar
> and understandable for many people, like threads.
> This also means, after some stabilization phase and iterations,
> most of these applications on top of co's will get optimized
> back into C, dealing with the co stuff from there.
> This doesn't mean to remove the co's Python interface,
> but to improve and simplify it further.

So when do we get to Pickle continuation objects and send them out
across the network, or even better, to another VM running on a second
processor? :)

Personally, I think continuations and coroutines are beautiful, and
easier and simpler than threads. Yay Christian! (and Yay Guido for
writing Python!)

PS. Can someone point me to example code for building smalltalk style
closures easily with continuations?
-- 
sHae mAtijs eRisson (sHae at wEbwitchEs.coM) gEnius fOr hIre
   bRing mE fIve sQuirrels aNd nO oNe wIll gEt hUrt
13:00pm up 1 season, 1 squirrel, load average: 1+1j 0.5+2j 0.2+0.5j


From mwh21 at cam.ac.uk  Sun Apr  2 19:24:01 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 03 Apr 2000 00:24:01 +0100
Subject: NameError when assigning dictionary values
References: <20000402231757.6604.qmail@web2105.mail.yahoo.com>
Message-ID: <m3zorchysu.fsf@atrus.jesus.cam.ac.uk>

lewst <lewst at yahoo.com> writes:

> Fredrik Lundh <effbot at telia.com> wrote:
> 
> > if you want to create variables in the local namespace
> > on the fly, you're obviously trying to emulate some other
> > programming language (Tcl?) in Python.  that's usually a 
> > rather bad idea.
> 
> I was trying to build a dictionary and create lists corresponding to
> the value names in the dictionary in one loop.

I'm afraid that doesn't help (me, at least).  It still sounds like
you're trying to write tcl or perl in Python.  What problem are you
trying to solve (at a higher level)?  It's likely that there is a
reasonable elegant way of doing it in Python, but it's unlikely to
look much like what a solution would look like in another langauge.

> 
> Do I really have to type "one" and "two" twice?  Once to make it a
> variable and a second time to add to my dictionary?  I.e,
> 
> >>> one, two = [], []
> 
> >>> mydict = { ".": one, "+": two }
> 
> This seems error prone if the the number of variables grows very large
> as it will for me.

I'd say so.  Here you could type:

mydict = { ".":[], "+":[] }

but I'd guess that's of no help.

Hoping to help,
M.

-- 
well, take it from an old hand: the only reason it would be easier
to program in C is that you can't easily express complex  problems
in C, so you don't.                 -- Erik Naggum, comp.lang.lisp


From mhammond at skippinet.com.au  Sat Apr  8 22:01:59 2000
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Sun, 09 Apr 2000 02:01:59 GMT
Subject: win32all-130 for Python 1.6 available
References: <_ziF4.1488$1b4.6926@news-server.bigpond.net.au> <Q4JH4.605$hh6.22805@news2-win.server.ntlworld.com>
Message-ID: <rwRH4.1668$4v.9405@news-server.bigpond.net.au>

"Adrian Eyre" <adrian at simspon.dyndns.org> wrote in message
news:Q4JH4.605$hh6.22805 at news2-win.server.ntlworld.com...

> 3. I don't suppose it would be possible to make debugged scripts
> run in a separate thread, to stop it blocking the program's event
> loop?

Ive thought about this.  The only problem occurs when the script being
debugged is itself a GUI script.  Then the script is likely to screw
up big time when running under the debugger, as we wont be following
the win32 GUI threading rules.

A better and more general solution would be for Python to grow "remote
debugging" capabilities that would allow out-of-process (even across
machine) debugging.  Pythonwin could then have the ability to debug a
script in a complete new process....

Mark.




From ns566645 at swt.edu  Sun Apr 23 23:37:01 2000
From: ns566645 at swt.edu (ABC)
Date: Sun, 23 Apr 2000 22:37:01 -0500
Subject: Can there be scrollbar attached to group widget
Message-ID: <3903C15D.E33126@swt.edu>

Hi ,

I have a notebook and the group is embedded in the notebook. can the
scrollbar can be attached to the group.

Thanks



From makarki.NOSPAM at news.hit.fi  Wed Apr  5 03:50:00 2000
From: makarki.NOSPAM at news.hit.fi (Matti Johannes Kärki)
Date: 5 Apr 2000 07:50:00 GMT
Subject: Platf. indep. console lib.
References: <8c03u1$igk$1@news.hit.fi> <38E8AF93.70F4CCF7@ansic.net>
Message-ID: <8cer78$2cv$1@news.hit.fi>

Scott (sbarron at ansic.net) wrote:
: There is an slang module which wraps the slang library.  It's available
: for Unix and Win32 though I haven't had the time to actually use it
: yet.  Here is an ftp site for it: 
: ftp://quasar.ipa.nw.ru/pub/avv/Python/pyslang .  If you give it a try,
: let me know how it turns out!

Well, I checked the wrapper. I didn't test it. I don't have slang under
Windows and I don't want to compile/install one (I have Watcom compiler
and Slang seems to require VC). Besides, based on documents, the library
is a bit too complicated to be actually practical. I think I'll just write
my own library. Now I have to learn how to write my own modules and native
libraries under Python...



Sincerely,

    Matti


From peter at verecomm.com  Sat Apr  8 18:00:40 2000
From: peter at verecomm.com (Peter Kropf)
Date: Sat, 08 Apr 2000 15:00:40 -0700
Subject: Python Language Support for ArcView GIS
References: <q0MH4.2398$Bq2.38636@newscontent-01.sprint.ca>
Message-ID: <38EFAC08.2A10932D@verecomm.com>

Just an FYI: You might want to consider moving AVPython to
sourceforge.net. The entire environment is designed for 
various levels of open source development... 


Bruce Dodson wrote:
> 
> Python Language Support for ArcView GIS (aka AVPython) is now available
> online as a free download.  I promised this several months ago but never
> managed to find a home for it (too busy) so finally I decided to just slap
> together a page for it on GeoCities:
> www.geocities.com/brucedodson.rm/avpython.htm
> 
> If you're an ArcView developer and want to be able to control COM-enabled
> programs like Excel or ArcInfo, or if you want to be able to interoperate
> with distributed systems through standards like XML and HTTP, you might want
> to give it a try.  The same goes for Python programmers who happen to be
> working with ArcView GIS (hey, there must be a few!), and want to be able to
> apply their Python skills (and use a better language) when developing
> ArcView applications.
> 
> Python is free and open source.  I hope AVPython will eventually be the
> same, but we haven't yet settled on an acceptable open source license, so
> for now the best I can do is provide a binary that is freely
> redistributable.
> 
> If you use AVPython on any interesting projects, or if you have any positive
> feedback or things that you'd like to change, feed my hubris: tell me about
> it!  Just take nospam out of my address, or use the one listed in the
> documentation.
> 
> Sincerely,
> Bruce Dodson

-- 
Peter Kropf
Codesmith
VEReCOMM Corp.
2475 Augustine Drive, Suite 102
Santa Clara, CA 95054
408-919-3107; Fax 408-969-0688
peter at verecomm.com


From phil at geog.ubc.ca  Sat Apr 29 12:33:04 2000
From: phil at geog.ubc.ca (Phil Austin)
Date: 29 Apr 2000 09:33:04 -0700
Subject: Looking for example of embedded Python
References: <8ecpi4$j9q$1@nnrp1.deja.com>
Message-ID: <m3ya5wq33z.fsf@brant.geog.ubc.ca>

chris_barker at my-deja.com writes:

> Hi,
> 
> I'm trying to persuade my co-workers that it would be extememly usefull
> to imbed Python in some of our apps. I'm looking for example
> applications wioth Python embedded. Ideally something fairly small with
> source code available.
> 

In addition to Demo/embed/demo.c, a nice example that tests Numpy
array classes using CXX-4.2 (http://CXX.sourceforge.net) has been
written by Paul Dubois.  It works for me using KCC under solaris and
linux, and should also compile with g++ 2.95.2. (I got segfaults
under linux until I  remember that the correct  linker  sequence
needed -Xlinker -export-dynamic,  e.g:

gcc  -Xlinker -export-dynamic demo.o -L/home/phil/lib/pythonmain/linux/lib/python1.5/config \
   -lpython1.5   -lieee -ldl  -lm  -o mydemo


Here's the output,  followed by the source:

<peacock ~/CXX/CXX-4.2> arraytest
Array facility test
[ 0.   0.5  1.   1.5  2.   2.5  3.   3.5  4.   4.5  5.   5.5]
Rank is 1
Species is 7
dimension(1) = 12
Pointers 125188, 1200b8, 125188
A list: [9.0, 4.0, 1.0]
toArray: [ 9.  4.  1.]
<type 'array'>
list+list: [ 18.   8.   2.]
add 2: [ 11.   6.   3.]
list[1]: 4.0
sorted: [ 1.  4.  9.]


/* Minimal main program -- everything is loaded from the library */

#include "Python.h"
#include <iostream>
#include "CXX_Array.h"
#include <algorithm>

using namespace Py;

extern "C" int Py_Main(int argc, char** argv);
extern "C" void Py_Initialize();
extern "C" void Py_Finalize();
//extern "C" void PyImport_ImportModule(char *name);

int
main(int argc, char** argv)
{
    PyObject *ptype;
    PyObject *pvalue;
    PyObject *ptrace;

	std::cout << "Array facility test" << std::endl;
	Py_Initialize();
	import_array();
	PyImport_ImportModule("Numeric");

    try {
        Array t(12);
        for(int i=0; i < t.length(); ++i) t[i] = Float(i / 2.0);
        std::cout << t << std::endl;
        std::cout << "Rank is " << t.rank() << std::endl;
        std::cout << "Species is " << t.species() << std::endl;
        std::cout << "dimension(1) = " << t.dimension(1) << std::endl;
        std::cout << "Pointers " << t.ptr() << ", " << (void *) t.to_C() << 
            ", " << t.as_contiguous().ptr() << std::endl;
        List w(3);
        w[0] = Float(9.0);
        w[1] = Float(4.0);
        w[2] = Float(1.0);
        Array from_list = toArray(w);
        std::cout << "A list: " << w << std::endl;
        std::cout << "toArray: " << from_list << std::endl;
        std::cout << from_list.type() << std::endl;
	//return Py_Main(argc, argv);
        std::cout << "list+list: " << from_list + from_list << std::endl;
        std::cout << "add 2: " << from_list + Int(2) << std::endl;
		std::cout << "list[1]: " << from_list[1] << std::endl;
		std::sort(from_list.begin(), from_list.end());
		std::cout << "sorted: " << from_list << std::endl;
    }
    catch(Exception&) {
        std::cout << "Test failed via Exception" << std::endl;
        PyErr_Fetch (&ptype, &pvalue, &ptrace);
        FromAPI p(pvalue);
        Object v(p);
        std::cout << v << std::endl;
    }

    Py_Finalize();
	return 0;
}


From jepler.lnk at lnk.ispi.net  Sat Apr 22 22:31:20 2000
From: jepler.lnk at lnk.ispi.net (jepler epler)
Date: Sun, 23 Apr 2000 02:31:20 GMT
Subject: Pickling Robustly
References: <jak8huf1of.fsf@helix.twistedmatrix.com> <8dpk1i$7de$1@newshost.accu.uu.nl> <jazoqn2s0b.fsf@helix.twistedmatrix.com>
Message-ID: <slrn8g4fs9.33l.jepler.lnk@potty.housenet>

On 21 Apr 2000 14:06:28 -0500, Glyph Lefkowitz
 <glyph at twistedmatrix.com> wrote:
>Perhaps I should ask this question in a different way.  How does
>Pickle tell if an object is pickleable or not?  Is there some
>pickle.ispicklable method I could refer to?

If you have control over the object hierarchy, why not use this
'unpicklable mixin' and convenience functions I just tossed together:

import pickle, UserList

class Unpicklable:
	def __getinitargs__(self):
		raise Unpicklable
		
def ispicklable(a):
	"""This tests if a can be pickled (by actually trying to do so)"""
	try:
		pickle.dumps(a)
	except Unpicklable:
		return 0
	return 1

def ispicklableinstance(a):
	"""Note that this function does not tell whether a ultimately
contains an unpicklable instance. But at least it's a simple test.""
	return isinstance(a, Unpicklable)

def ispicklableclass(A):
	"""Note that this function does not tell whether an instance
might ultimately be unpicklable due to something nested inside it."""
	return issubclass(a, Unpicklable)

if __name__=='__main__':
	class UserListU(UserList.UserList, Unpicklable): pass

	x = UserList.UserList([1,2,3])
	print pickle.dumps(x)

	y = UserListU([1,2,3])
	print pickle.dumps(y)  # This will cause an exception

Jeff


From pete at visionart.com  Tue Apr 25 21:50:44 2000
From: pete at visionart.com (Pete Shinners)
Date: Tue, 25 Apr 2000 18:50:44 -0700
Subject: new linereading standard (ThanksAll!)
References: <3904D21B.333DD930@visionart.com> <390510C9.A744D22E@yahoo.com> <R2g+MIAFMUB5EwUl@jessikat.demon.co.uk> <mt1z3uyr8k.fsf@astron.berkeley.edu> <26042000.2@sanctum.jae.ddns.org>
Message-ID: <39064B74.287DD3DE@visionart.com>

i've seen all manner of input on the linereading stuff. i'm
still in the learning phase and it was great to see all the
alternatives.

i think i'll still be writing my own line reading class that
handles much more special cases. (comments, continued line with
backslash, stripping off newlines, empty lines, etc)

if anyone cares i'll post my module when it's ready.


as nifty as that 'map' is, i still find a for loop is
most asthetically pleasing.

-- 
_____Pete Shinners______
   pete at visionart.com


From tismer at tismer.com  Tue Apr  4 06:42:52 2000
From: tismer at tismer.com (Christian Tismer)
Date: Tue, 04 Apr 2000 12:42:52 +0200
Subject: Stackless Python on O'Reilly.de Homepage
Message-ID: <38E9C72C.E9CD8AB2@tismer.com>

Hi all,

For those who can read German a little.
O'Reilly (German department) takes notice of
the Python of the future :-))  *) and microthreads.

http://www.oreilly.de/

ciao - chris

*) at least the future of the computation [ahem]
-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaunstr. 26                  :    *Starship* http://starship.python.net
14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     we're tired of banana software - shipped green, ripens at home



From daniels at mindspring.com  Sat Apr 22 21:17:17 2000
From: daniels at mindspring.com (Alan Daniels)
Date: 23 Apr 2000 01:17:17 GMT
Subject: code not true?
References: <8ds3ku$q0l$1@nnrp1.deja.com> <3901aa0f.0@news.proweb.co.uk> <M3iM4.518$wJ1.17829@ptah.visi.com> <NhiM4.1704$v85.10013@news-server.bigpond.net.au>
Message-ID: <slrn8g45lj.oa.daniels@localhost.localdomain>

On Sat, 22 Apr 2000 14:02:21 GMT, Mark Hammond
<mhammond at skippinet.com.au> wrote:

[snip...]
>Of all the solutions to this problem, this does strike me as one of
>the best - simply remove equality testing al together for floating
>point.

Are you sure of this? It would be awful confusing for a programming
newbie to do:

x = 5.0
y = 5.0 # Later on down in the code.
x == y  # And even further...

And get an exception thrown. All the "why doesn't this work?" messages
we've seen about floating point have to do with doing manipulation of
the numbers and *then* having them not work.

For 0.07 * 100 to not be the same as 7.0 is understandable, because
that's just how floating point works on CPUs today: Not all numbers
are representable in IEEE format. But to outlaw asking if 7.0 is equal
to 7.0 definitely goes against the principle of least suprise. And
learning that floating-point math sometimes leads to slight numerical
errors is an important lesson that newbie programmers shouldn't be
shielded from.

I like Python 1.6's handling of it better: Showing floating point
values in all their slightly-inaccurate glory seems like the better
plan.

-- 
============================================================
Alan Daniels
daniels at mindspring.com


From ppi at searchy.net  Sun Apr  2 16:07:04 2000
From: ppi at searchy.net (Penpal International)
Date: Sun, 02 Apr 2000 22:07:04 +0200
Subject: GTK Referenc
Message-ID: <38E7A868.A1C9BD7C@searchy.net>

I couldn't find anywhere a reference for the GTK module for Python. Does
anyone know one?
On gtk's site I could only find some for C(++).




-- 
Penpal International
http://ppi.searchy.net/
ppi at searchy.net


From hanche at math.ntnu.no  Fri Apr  7 23:26:11 2000
From: hanche at math.ntnu.no (Harald Hanche-Olsen)
Date: 07 Apr 2000 23:26:11 -0400
Subject: Why should I switch to Python? - Infinity of Primes
References: <000601bf9f8e$bd31e780$ea2d153f@tim> <pcoln2q3bet.fsf@math.ntnu.no> <38EE5A40.4557F4EA@seas.upenn.edu>
Message-ID: <pco1z4hclyk.fsf@math.ntnu.no>

+ dave white <dwhite2 at seas.upenn.edu>:

| Since we're on the subject of primes...  I've heard that there's a
| nice probabilistic algorithm for computing large primes using
| Fermat's Little Theorem.  Python seems like it would be well suited
| for this since bignums are built into the language.  Does anyone
| know this algorithm?

You're thinking of using FLT for primality *testing*, not for
*computing* primes.  The algorithm is trivial to explain:

Assume p is an integer, suspected to be a prime.  Pick an integer a
such that 2 <= a < p and compute a**(p-1) mod p.  If you get 1, then p
*may* be a prime, but again it may not.  Repeat to gain more
confidence.  On the other hand, if you don't get 1, p is definitely
not a prime.

To find large probable primes using this algorithm, pick large odd
numbers at random and subject them to the test until you find one that
passes the test for sufficiently many different a to give you that
warm fuzzy feeling that surely, this must be a prime.

Unfortunately, there are some composite numbers that will always pass
this test.  These are called pseudoprimes.

On the other hand, there are other probabilistic primality tests that
don't have this weakness.  Perhaps the best known is the Miller-Rabin
test, which just like the FLT test is applied with numbers a between 2
and p, and which always succeeds when p is a prime.  Moreover, the
Miller-Rabin test has the pleasant property that it fails for at least
3/4 of the available values a when p is composite, so that, e.g., 20
successful iterations of the test gives you roughly a billion to one
confidence in the primality of p.

I'd post some code, but all I have is some five year old Scheme code,
and that is probably not all that appropriate in this group.  The
comments to that old code has two literature references, though:

;;; Monier: Evaluation and comparison of two efficient
;;; probabilistic primality testing algorithms,
;;; Theoretical Computer Science 12 (1980) 97-108

;;; A.O.L. Atkin and R.G. Larson: On a primality test of Soloway 
;;; and Strassen, SIAM J. Comput. 11 (1984), 789-791

If you go surfing the web you'll probably find lots of implementations
in every language known to man, however.

And-for-dessert-there-is-primality-*proving*-using-elliptical-curves-ly y'rs,

-- 
* Harald Hanche-Olsen     <URL:http://www.math.ntnu.no/~hanche/>
- "There arises from a bad and unapt formation of words
   a wonderful obstruction to the mind."  - Francis Bacon


From chuck.meyers at lmco.com  Tue Apr  4 12:42:55 2000
From: chuck.meyers at lmco.com (Chuck Meyers)
Date: Tue, 04 Apr 2000 09:42:55 -0700
Subject: What is python's language level?
Message-ID: <38EA1B8F.A797A130@lmco.com>

Does anyone know python's language level? I did not find it in the list
at:

http://www.spr.com/library/0langtbl.htm

I would guess that it would be somewhere between 15 and 30. This is an
important point, even if language levels are imprecise they can make it
easier to convince people to look at or try a language like python.


-Thanx

chuck.meyers at LMCO.COM






From thomas.heller at ion-tof.com  Mon Apr 17 15:11:46 2000
From: thomas.heller at ion-tof.com (Thomas Heller)
Date: Mon, 17 Apr 2000 21:11:46 +0200
Subject: [Announce] autoreload.py
Message-ID: <001601bfa8a0$c6d86f50$4500a8c0@thomasnb>

This is a module which automatically takes care of reloading
source modules into the running python interpreter from a background
thread when the source has changed.

An additional benefit is that top-level classes from the reloaded modules
are automatically updated to contain the new class dictionary.

The net-effect is that you can run a program in pythonwin (tested) or
IDLE (untested), change the instance methods, save the file
and automatically the running program uses the new methods.

Warning: This is only sighly tested, but it may be usefull.

Enjoy!

Thomas Heller
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: autoreload.py
URL: <http://mail.python.org/pipermail/python-list/attachments/20000417/9c5307e7/attachment.ksh>

From gmcm at hypernet.com  Thu Apr 27 14:22:01 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Thu, 27 Apr 2000 14:22:01 -0400
Subject: Schwartzian Transform in Python?
In-Reply-To: <lr3do7tuin.fsf@yoda.wg.waii.com>
Message-ID: <1255262762-67501940@hypernet.com>

Art Haas wrote:

> 
> I'm looking for the Python way to do this ...
> 
> opendir(DIR,".") || die "Can't open '.'! $!\n"
> @sorted =
>  map { $_->[0] }
>  sort { $a->[1] <=> $b->[1] }
>  map { [ $_ , -M ]} readdir(DIR)
> closedir(DIR);
> print "Files sorted by date ...\n";
> foreach (@times) {
>  print "$_\n";
> }
> 
> (Yes, its that other 'P' language).
> 
> I'm getting lost in the map/sort/map part for python. My python
> psuedo code looks like (don't laugh too much)
> 
> import os
> sorted = map(???,(map(os.stat???,listdir(".").sort())))
> print "Files sorted by date ..."
> for file in sorted:
>  print file

>>> import os, stat
>>> l = map(lambda f: (os.stat(f)[stat.ST_MTIME], f), 
os.listdir('.'))
>>> l.sort()
>>> for (t,f) in l:
...  print f

Don't speak that other language, so I'm taking what you said 
you want at face value...


- Gordon



From gbreed at cix.compulink.co.uk  Tue Apr 18 05:07:47 2000
From: gbreed at cix.compulink.co.uk (gbreed at cix.compulink.co.uk)
Date: 18 Apr 2000 09:07:47 GMT
Subject: Microsoft ActiveX Scripting
References: <7TNK4.3520$1D.25703@news-server.bigpond.net.au>
Message-ID: <8dh8l3$c2p$1@plutonium.compulink.co.uk>

In article <7TNK4.3520$1D.25703 at news-server.bigpond.net.au>, 
mhammond at skippinet.com.au (Mark Hammond) wrote:

> All you need to do then is say 'LANGUAGE="Python"' :-)

You also need to run something like 
win32comext\axscript\client\pyscript.py, and note that it isn't quite 
where the docs say it is.


From jstok at bluedog.apana.org.au  Tue Apr 18 12:46:45 2000
From: jstok at bluedog.apana.org.au (Jason Stokes)
Date: Wed, 19 Apr 2000 02:46:45 +1000
Subject: Lists
References: <A63AFB20111ED311AC5500A0C96B7BF5015F0F2C@orsmsx36.jf.intel.com> <Hi0L4.7913$5D.17329@ozemail.com.au>
Message-ID: <bn0L4.7914$5D.17489@ozemail.com.au>

Jason Stokes wrote in message ...

>for i in range(0...len(a)):


I don't know what came over me (actually, I do, some work in Haskell
earlier.  Anyone else get language leak through?)

for i in range(0, len(a)):

>




From dave.rose at wcom.com  Tue Apr 25 17:19:30 2000
From: dave.rose at wcom.com (Dave Rose)
Date: Tue, 25 Apr 2000 21:19:30 GMT
Subject: Memory Problems on AIX 4.2
Message-ID: <CZnN4.55$7r2.2891@pm02news.wcom.com>

I am using a C function to perform some shared memory tasks and every time
the C function is called PS is showing the 4K is being reserved from the
free list and the executible image is increasing 4K as well.  Has anyone
else seen this type of memory leak with the AIX Shared object calls.

Any help would be greatly appreciated.  Been searching for the leak for a
couple of weeks now and can not find it.

If anyone is interested, I can include the code

Thanks in advance

Dave Rose





From aahz at netcom.com  Tue Apr 11 16:16:27 2000
From: aahz at netcom.com (Aahz Maruch)
Date: 11 Apr 2000 20:16:27 GMT
Subject: Thread-safe atomic operation
References: <1256659126-41087547@hypernet.com>
Message-ID: <8d016r$jbf$1@nntp9.atl.mindspring.net>

In article <1256659126-41087547 at hypernet.com>,
Gordon McMillan <gmcm at hypernet.com> wrote:
>Dong-gweon Oh writes:
>>
>> Are these statements thread-safe?
>> 1. import xxx
>> 2. xxx = __import__('xxx')
>> 3. reload(xxx)
>> 4. dict['xxx'] = sys.modules['xxx']
>> 5. dict.has_key('xxx')
>
>Numbers 2 and 4 could be interrupted between evaluation of 
>the RHS and assigment to LHS. For #2 that would be of 
>concern only if xxx was a global.

Are you absolutely certain about item 1?  That implies that Python as a
whole could block on importing a module that did something like

for i in range(100):
  sleep(1)
--
                      --- Aahz (Copyright 2000 by aahz at netcom.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

Why doesn't "Just Say NO" include caffeine, nicotine, alcohol, Prozac,
and Ritalin?  --Aahz


From alwagner at tcac.net  Tue Apr 18 08:48:44 2000
From: alwagner at tcac.net (Albert Wagner)
Date: Tue, 18 Apr 2000 07:48:44 -0500
Subject: How to write to printer in Windows?
Message-ID: <38FC59AC.4B71492A@tcac.net>

How can I write(aLine) to a printer in Windows?  I tried this:

from win32pipe import popen
:
self.printer = popen('lpt1' 'w')
:
self.printer.write(aLine)

but get a general protection fault.  Any help will be greatly
appreciated.

-- 
Small is Beautiful


From gmcm at hypernet.com  Sat Apr  8 20:24:14 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Sat, 8 Apr 2000 20:24:14 -0400
Subject: Tuples -- who needs 'em
In-Reply-To: <8cofcf$hns$1@newshost.accu.uu.nl>
Message-ID: <1256882636-27643467@hypernet.com>

Martijn Faassen wrote:

[discussing the need for tuples with Bob Alexander]

> Okay, but it's generally practiced in idiom. 
[snip]
> ... I think it's more
> than just the dictionary key argument.

Very true. There are 2 technical arguments for tuples 
(dictionary keys and efficiency); both are weak. The main 
difference is idiomatic.

[Bob]
> > There have been cases where I used tuples, to find out
> > later that I wanted mutability, so went back and changed.
> 
> Hm, I rarely if ever seem to run into this. If I need mutability of
> a set of heterogenous objects I tend to use a class. There may be a 
> switch from tuples to classes, but not often from tuples to lists, I 
> think.

That's because you understand the idiom, Martijn. Lists are 
collections. The competing choice to a list is a dictionary. 
Tuples are toy structs; the competing choice is a class.

Experience tells me that trying to use a list where a tuple is 
needed will yield either unsafe or safe but bloated code. No 
technical reason; it's just that lists (as collections) will 
eventually be mutated in a way that's natural for collections, 
but disastrous where the purpose is struct-like.
 
> > This sort of thing
> > happens in exploratory programming, so when in that mode one might want to
> > use lists unless one is really sure in advance that tuples are the right
> > thing.

By all means, use lists whenever you're unsure. You'll soon 
come to recognize when a tuple is more appropriate. It won't 
take long (it couldn't, Martijn's too young to have taken long to 
learn anything <wink>).
 


- Gordon



From fredrik at pythonware.com  Tue Apr  4 04:52:39 2000
From: fredrik at pythonware.com (fredrik at pythonware.com)
Date: Tue, 04 Apr 2000 08:52:39 GMT
Subject: Loading image from URL with PIL
References: <8cau6t$pph$1@nnrp1.deja.com>
Message-ID: <8ccagj$a1m$1@nnrp1.deja.com>

rgb122 at my-deja.com wrote:
> Is it possible to load an image (GIF or JPEG)
> straight from a URL without saving it to a file
> first. I tried this:
>
> u = urllib.urlopen
> ("http://localhost/cache/test.jpg")
> data = u.read()
> s = array.array("b", data).tostring()

fp = urllib.urlopen(myurl)
im = Image.open(fp)

> Also, how can I convert a GIF to a JPEG using PIL?

see the documentation for the 'save' method.

to do the conversion in memory, save to a StringIO
object, and make sure you pass a second argument to
'save'.

</F>


Sent via Deja.com http://www.deja.com/
Before you buy.


From effbot at telia.com  Tue Apr 11 10:30:18 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Tue, 11 Apr 2000 14:30:18 GMT
Subject: Getting an overview if all available driveletters on NT
References: <2B1262E83448D211AE4B00A0C9D61B03BA7294@MSGEURO1>
Message-ID: <_FGI4.1194$Za1.20497@newsc.telia.net>

Pieter Claerhout <PClaerhout at CREO.BE> wrote:
> is there a module in Python, which can list all available driveletters,
together with
> their label. I'm using Windows NT? I already thought of iterating over all
possible
> driveletters, and see if the path exists or not, but I also want to
separate the
> network volumes from the other volumes. Any ideas?

just guessing here (don't have win32api on this box),
but this might do the trick:

    type = win32api.GetDriveType("%s:\\" % drive)

where drive is 'A' through 'Z'.

</F>




From donn at u.washington.edu  Fri Apr 28 13:56:39 2000
From: donn at u.washington.edu (Donn Cave)
Date: 28 Apr 2000 17:56:39 GMT
Subject: ?sys.argv values more literally?
References: <3909B1AF.32FC8BED@email.sps.mot.com>
Message-ID: <8ecjcn$19ui$1@nntp6.u.washington.edu>

Quoth Norman Shelley <Norman_Shelley-RRDN60 at email.sps.mot.com>:
| I need to write a frontend program that will call one of two other
| programs (both of which are NOT binaries but are tcl programs therefore
| they HAVE to be invoke with os.system() or "/bin/sh" has to be the first
| argument to a os.execv() call).
|
| THe frontend program must detect a "-ver someversionname" options and
| remove the two words and then call the appropriate tcl program with the
| EXACT arguments given on the command line (less those two words).
| Problem is words in sys.argv
| that were quoted on the command line look like words that were not
| quoted on the command, e.g.
|
| %frontend -ver 1.0 word1 word2 "word three" "word(three)"
| argv looks like
| ["frontend", "-ver", "1.0", "word1", "word2",  "word three",
| "word(three)"]
|
| but this fails when os.system or os.execv is called (with frontend,
| -ver, and 1.0 replaced with the tcl program full pathname (/bin/sh added
| at front  for execv) as "word three" and "word(three)" are missing there
| protective quotes (the /bin/sh needs to see those as one word not two
| words or a word  with parens).
|
| I need   argv to look like this:
| ["frontend", "-ver", "1.0", "word1", "word2",  '"word three"',
| '"word(three)"']
| where words that were quoted on the command line are strings with
| beginning and ending quotes in the argv.
|
| Any way to do this or get around this problem?

As another followup has noted, it's OK to quote all of them.  The
python interpreter never gets the quotes you want, they're for the
benefit of the shell that invokes python.

However, I think it solves the problem the wrong way, and I'm not
convinced the problem is real either.  What happens if an original
argument includes a quoted quote?  This extra run through the shell
interpreter is not healthful.

When we call system() to execute a command, as you surmised we're
executing /bin/sh, specifically in effect it's like this:
  execv('/bin/sh', ('sh', '-c', command_line))

This shell parses the command line, locates its argv0 file and
in turn execs it.  We don't need to know how that file is implemented,
because the shell can run any command.  Not because it has any magic
powers, but because UNIX can run any well-formed executable file
including scripts.  Python can do the same as the shell, and if you
can figure out how to get that working, you can exec your command
directly and omit the re-quoting.

I can't say why Tcl scripts might need to be executed by /bin/sh,
so I can't be much help there.  If they were really shell scripts,
I'd say they might need an appropriate #! line, the line at the
top of the script file that specifies its interpreter.  E.e.g.,
#!/bin/sh                 or
#!/usr/bin/env python     or
#!/bin/awk -f             etc.

In the absence of this line, the UNIX exec fails and the shell will
usually try to execute the script itself.  If you want to have a
rule like that when you invoke programs, you can make the default
whatever you want for your implementation - the shell, Tcl, etc. -
but it's much better to fix the file so the UNIX exec will work as
intended.

	Donn Cave, University Computing Services, University of Washington
	donn at u.washington.edu

PS.  The HTML and V-card didn't help much.


From loewis at informatik.hu-berlin.de  Mon Apr 10 09:33:02 2000
From: loewis at informatik.hu-berlin.de (Martin von Loewis)
Date: 10 Apr 2000 15:33:02 +0200
Subject: Extending built-in language (Win 95)
References: <213CDEBFDDA7D3118081009027E40C56071113@SHEF-MS1>
Message-ID: <p6qzor2cc8h.fsf@informatik.hu-berlin.de>

Mark Pope <M_Pope at shef.rage.co.uk> writes:

> I would like to extend the objects available in python, however It is not
> appropriate to build normal extension dll's.

Why is it not appropriate? It works perfectly well to add new types in
an extension module. Many extensions do.

> Appears to be a template for defining new built in objects which will be
> compiled into the Python static lib. However there must be other steps to
> making the new object reachable from python script.

The new type becomes only available by means of creation of
values. Look at the file type. Do you write

f = file('foo.txt','w')

? No, you use open() instead, which just returns an instance of a file
object. In the xxmodule, xx_new is used to make new instances, which
can be called from Python as xx.new().

Hope this helps,
Martin


From dworkin at ccs.neu.edu  Mon Apr 24 22:40:33 2000
From: dworkin at ccs.neu.edu (Justin Sheehy)
Date: 24 Apr 2000 22:40:33 -0400
Subject: The Simple Economics of Open Source
In-Reply-To: Raffael Cavallaro's message of "Mon, 24 Apr 2000 03:57:00 GMT"
References: <B527E7A5.51AA%raffael@mediaone.net> <MPG.136cc01af4ab703b989709@news.earthlink.net> <raffael-4899DC.23562823042000@news.ne.mediaone.net>
Message-ID: <vndya62ykbi.fsf@camelot-new.ccs.neu.edu>

Raffael Cavallaro <raffael at mediaone.net> writes:

> Fact is, only commodity items, that are known to many programmers, are 
> open sourced. If there were sufficient demand to support a closed source 
> business, then it would have happened that way. The proof of this is the 
> BSDs, which would allow a closed source, proprietary fork at any time. 
> Hasn't happened 'cause there's no demand for such a product.

You mean like BSD/OS, from BSDI?  

It happened 9 years ago, and they've been selling it ever since.

Interesting "proof" you have there.

-Justin

 



From godzilla at netmeg.net  Tue Apr 11 11:24:20 2000
From: godzilla at netmeg.net (Les Schaffer)
Date: Tue, 11 Apr 2000 15:24:20 GMT
Subject: issubclass funny business
Message-ID: <m2snwsd5ah.fsf@netmeg.net>

I have the following class hierarchy:

------ in module FileScanner -----

class StandardArrayFileScanner(ArrayFileScanner):

--------- in module cie ----------

from FileScanner import StandardArrayFileScanner

class CIE_Standards(StandardArrayFileScanner):
...    

class Photopic(CIE_Standards):
...

class CIE(CIE_Standards):
...


but when i need to test whther an object is an instance of a subclass
of StandardArrayFileScanner:

        # make sure std is of the correct type
        if not issubclass(std.__class__, StandardArrayFileScanner):
            import cie
            print std.__class__, cie.CIE.__bases__
	    raise NotStandardArrayError


but it doesnt work, even though std is of type cie.CIE :

(gustav)~/Engineering/dspring/stoplite/matlab/Jue-Data/: python FileScanner.py

cie.CIE (<class cie.CIE_Standards at 813b320>,)  <=== result of print stmt

huh??????

les schaffer


From scottrharris at earthlink.net  Wed Apr 12 20:59:23 2000
From: scottrharris at earthlink.net (Scott Harris)
Date: Thu, 13 Apr 2000 00:59:23 GMT
Subject: Python and VTK and TK 
Message-ID: <38F51AB1.29FB4C1D@earthlink.net>

Has anyone gotten VTK to work with the TkRenderWidget?
I've got a simple code that draws two buttons and a vtkTkRenderWidget in
a frame.
It works under win98, but crashes my XServer on Linux.

I'm using RH6.1 with XFree864.0 and a Voodoo3 card (so I have
accelerated GL rendering)
I've got VTK 3.1

Anyhow, I'm going to try wxPython and it's VTK widget in the next few
days.

If you had to start a new moderately sized project would you use
wxPython or Tk?

-Scott Harris
scottrharris at earthlink.net



From johngrayson at home.com  Sun Apr  9 18:25:31 2000
From: johngrayson at home.com (John Grayson)
Date: Sun, 09 Apr 2000 22:25:31 GMT
Subject: Migrating Tcl system to Python
References: <095cc48a.7e9b36dd@usw-ex0108-063.remarq.com>
Message-ID: <38F1035B.9A893FC3@home.com>

Hamish,

I converted several of the examples in the Widgets tour 
for Tk when I was writing Python and Tkinter Programming.
I tried to keep the basic structure of the original scripts
but I could see that there were more compact ways of getting
the same effect in Python. I'm sure that the same would be true
for pure Tcl scripts.

   John Grayson

Hamish Lawson wrote:
> 
> I'd be interested in anyone's experience of migrating a (several-
> thousand line) system written in Tcl to Python. And since we'll
> have to deploy new functionality while we're in the process of
> migrating, I'd particularly like to hear of experiences with
> having Tcl code using Python code (and vice versa). I'm
> considering XML-RPC/SOAP as a possible way to do that, but what
> other approaches might there be?
> 
> Hamish Lawson
> 
> * Sent from RemarQ http://www.remarq.com The Internet's Discussion Network *
> The fastest and easiest way to search and participate in Usenet - Free!


From gmcm at hypernet.com  Thu Apr 27 11:42:16 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Thu, 27 Apr 2000 11:42:16 -0400
Subject: The Simple Economics of Open Source
In-Reply-To: <raffael-21A8E7.10481427042000@news.ne.mediaone.net>
Message-ID: <1255272357-66924550@hypernet.com>

Raffael Cavallaro wrote:

> In article <vndya62ykbi.fsf at camelot-new.ccs.neu.edu>, Justin
> Sheehy <dworkin at ccs.neu.edu> wrote:
> 
> >You mean like BSD/OS, from BSDI?  
> >
> >It happened 9 years ago, and they've been selling it ever since.
> 
> Yeah, and BSDI is just tearing up the software world. BSDI sells
> *support* not software, because the software they sell is
> essentially interchangable with the free BSD variants. BSDI have
> *themselves* recognized this fact by merging with Walnut Creek.
> In future, they'll be merging the code of FreeBSD and BSD/OS.

I guess you missed ESR's talk at IPC7. Software company 
sells package for $500, street price is $400. Software 
company goes broke. What's the street price now? If software 
were a product, it would be, say $300. But the street price is 
$50. Software *is* a service, the price is a reflection of 
expected future support.

Other counter-examples: you can pay thousands for your web 
server, or get it for free. You can pay hundreds for your office 
suite, or get it for free. You can pay hundreds to thousands for 
your compiler or get it for free. (You're stuck with Python - that 
only comes for free). Even where the closed source package 
has technical advantages, it does not support the price 
differential.

Closed source protecting an original piece of intellectual 
property is actually quite rare (and where it exists, it's quite 
short-lived). Normally it "protects" an investment of money and 
effort. And actually it reflects what consumers are willing to 
pay for utility and (quite probably deluded notions of) future 
support, upgrades and porting.

The closed source world is based on a diseased model 
(pretending software is a product). The open source world is 
currently in large part shaped by being a reaction to the 
closed source model. Thus, it too is deformed.

One of those open source delusions is that being open source 
somehow makes a difference to the end consumer. Bunk. End 
users don't read code, and never will. Opening the source 
benefits other developers, who in turn will almost certainly 
benefit you by making the code better, or integrating with 
other software - thus indirectly benefitting users.

Consumers benefit directly because the "product" price is 
based on the marginal costs of production (near $0), and 
support costs are based on reality.

Back to your art world analogy - how much would still be true 
if you considered only the sub-market of unlimited production 
prints with trivial reproduction costs?

Scarcity of an intellectual secret may make a software 
package containing that secret more valuable, but scarcity of 
a software package is likely to make it less valuable (since 
support will be hard to find, and upgrades unlikely). OTOH it's 
scarcity of a piece of artwork that makes it valuable.



- Gordon



From eviltofu at rocketmail.com  Thu Apr 27 20:52:54 2000
From: eviltofu at rocketmail.com (Jerome Chan)
Date: Fri, 28 Apr 2000 00:52:54 GMT
Subject: Python does not let me format time.
References: <8eakhq$797$1@nnrp1.deja.com>
Message-ID: <eviltofu-76BEBF.20525427042000@news.earthlink.net>

In article <8eakhq$797$1 at nnrp1.deja.com>, Sindh <skodela at my-deja.com> 
wrote:

> Hi folks
> 
> I have a file full of date, value pairs. Date is like MM/DD/YY. My
> problem is I want to format this file since it contains a few thousand
> lines of data , using python. I tried using string and time modules. I
> want to format the time value to a float which is returned by
> time.gmtime normally. I tried to split it and to use parts of it to make
> a tuple and to use time.mktime but it says illegal operation for
> builtin.
> 
> Summary:
> 
> data is:
> 	03/11/98 abcd
> 	03/12/98 efgh
> 
> I split it to
> 	x=string.split(n[0],' ')
> 	which gave me x='03/11/98'
> Now I split it again by
> 	x=string.split(x,'/')
> 	which gave a tuple of '03','11','98'
> Now I tried  to make time as
> time.mktime(x[2],x[0],x[1],'','','','','','')

use 0,0,0,0,0,0 for the rest of the tuple

> The problem is it says illegal operation for builtin. I tried to use the
> string val and int val as the arguments for mktime.
> 
> Anyone with any helpful ideas!!
> 
> If there is already a moduls to do the same , much helpful.
> 
> Thanks
> sreekant
> 
> --
> A man needs to sleep for 36 hours a day atleast.
> 
> 
> Sent via Deja.com http://www.deja.com/
> Before you buy.


From chafik at nevrax.com  Mon Apr  3 12:56:59 2000
From: chafik at nevrax.com (Sameh chafik pro)
Date: Mon, 3 Apr 2000 18:56:59 +0200
Subject: Mail List
Message-ID: <000a01bf9d8d$a05c0d00$1001a8c0@nevrax.net>

Please, detach me from the mail list.
Thanks
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20000403/7f814629/attachment.html>

From andrer at stud.cs.uit.no  Sun Apr  2 19:05:14 2000
From: andrer at stud.cs.uit.no (=?iso-8859-1?Q?Andr=E9?= Larsen Risnes)
Date: Mon, 03 Apr 2000 01:05:14 +0200
Subject: Advice requested: GUI project beginning
References: <m12b4OW-000wcEC@swing.co.at>
Message-ID: <38E7D22A.C81EA4@stud.cs.uit.no>

> What is Squeak?
> 

http://www.squeak.org/

"Squeak is an open, highly-portable Smalltalk-80 implementation 
whose virtual machine is written entirely in Smalltalk, ..."

--
Andr? Larsen Risnes


From scarblac-spamtrap at pino.selwerd.nl  Mon Apr 10 18:54:58 2000
From: scarblac-spamtrap at pino.selwerd.nl (Remco Gerlich)
Date: 10 Apr 2000 22:54:58 GMT
Subject: Tuples -- who needs 'em
References: <00e101bf9d80$a8384d20$74eb0b18@stcla1.sfba.home.com> <va3G4.1110$n68.190454272@newsb.telia.net> <037f01bf9e55$6d328760$74eb0b18@stcla1.sfba.home.com> <8cfa75$fho$1@newshost.accu.uu.nl> <38ED4292.2EE8D3CA@cosc.canterbury.ac.nz> <38ED8EF2.B0835BE1@acm.org> <gandreas-AEF36E.10573007042000@news.visi.com>
Message-ID: <slrn8f4q7l.n2c.scarblac-spamtrap@flits104-37.flits.rug.nl>

Glenn Andreas wrote in comp.lang.python:
> At the very least, this visually is jarring and you need to mentally 
> "reparse" when you get to the second ">".  Another alternative is to 
> switch to multi-character brackets, perhaps:
> 
>    x = (* a, b *)
> 
> or:
> 
>    x = <| a, b |>
> 
> With two characters, there could be hundreds of possible "bracket" 
> pairs.  Unfortunately, that makes your code look like it has been 
> infested with "smileys".  I seem to remember that some other languages 
> using them though (besides old Pascal style comments, that is).

LPC uses ({ "spam": "eggs" }) for dictionaries (called "mappings" in that
world). I like that. But that gives problems with the current syntax, of
course.


-- 
Remco Gerlich,  scarblac at pino.selwerd.nl
  Murphy's Rules, "Seat belt? What seat belt?":
   In Avalon Hill's Speed Circuit, a car which rear-ends another may spin
   out and crash, but the rammed vehicle is unaffected.


From mwh21 at cam.ac.uk  Fri Apr 28 09:02:45 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 28 Apr 2000 14:02:45 +0100
Subject: Simple list.append() question
References: <meh9-5ED8CD.11192424042000@news.cit.cornell.edu> <ZHZM4.360$qg7.2949@news.uswest.net> <39046FA2.603EAFE5@roguewave.com> <PsaN4.174166$Dv1.2059732@news1.rdc1.bc.home.com> <8e4akc$p7o$1@newshost.accu.uu.nl> <m366t6ozqx.fsf@greebo.nodomain.de> <3908E6EC.B9E6B2E8@cosc.canterbury.ac.nz>
Message-ID: <m3zoqenzt6.fsf@atrus.jesus.cam.ac.uk>

Greg Ewing <greg at cosc.canterbury.ac.nz> writes:

> Bernhard Herzog wrote:
> > 
> > It would be a good idea IMO to find a
> > good syntax for this case and ISTM that list comprehensions might be
> > useful here:
> > 
> >   multi_list = [[] for i in range(10)]
> > 
> > or whatever the precise syntax was. Hmm, the unused variable in there is
> > a bit ugly though.
> 
> Maybe the LC syntax should include something like
> 
>    [[] times 10]

But wouldn't that make "times" a keyword?  Not sure I'd want that...

Cheers,
M.

-- 
  If i don't understand lisp,  it would be wise to not bray about
  how lisp is stupid or otherwise criticize, because my stupidity 
  would be archived and open for all in the know to see.
                                               -- Xah, comp.lang.lisp


From rnog2438 at mail.usyd.edu.au  Sun Apr 16 11:13:35 2000
From: rnog2438 at mail.usyd.edu.au (Ricardo Nogueira)
Date: Mon, 17 Apr 2000 01:13:35 +1000
Subject: Event triggering and weak references
Message-ID: <8dbif3$hm0$1@metro.ucc.usyd.edu.au>

First thanks for this beautiful language : )

I am Smalltalk addicted : )  and this means that I can't do with any other
language, ex. C, C++, Java,  : (
This was true until I found Python 2 weeks ago at Sydney Uni. I loved its
modularity, pluggin behaviour and freedom.
But I am missing Smalltalk event handling... Smalltalk (like Python) support
the call-Back mechanism, but there is a more powerful inter-object interest
registering that I haven't found in Python (so far).

This is what I would like to do:
Imagine I could have two or more Widgets open displaying a car speed value,
say a numerical and a graphic display, if this car speed changes (ex. some
one pushes the break), I would like to have any open widget on this value
updated (on opening each window would register interest with this car
speed).

Smalltalk implements this with:

The widget would register interest in one aspect of the car  (on opening
over aCar):
  aCar when:#speedChanged send: #updateSpeedDisplay to: self

And the car would just notify who is interested, if any (when its speed
changes):
  self trigger: #speedChanged

But if any of this windows is not referenced anymore or crashes I would like
the garbage collector clear it (the car would not be notified and his
pointer to the window shouldn't keep the garbage collector form freeing the
window memory).

Again, Smalltalk implements this with weak references, that doesn't count
for the garbage collector.

Is there support for this in Python?

Any suggestions?




From mmiller3 at iupui.edu  Thu Apr  6 18:03:46 2000
From: mmiller3 at iupui.edu (Michael A. Miller)
Date: 06 Apr 2000 17:03:46 -0500
Subject: Beginning to be disapointed
References: <Aw4H4.8078$pR.68745@carnaval.risq.qc.ca> <mt3dozt6sh.fsf@astron.berkeley.edu> <mtu2hfrnl7.fsf@astron.berkeley.edu>
Message-ID: <87snwydgzh.fsf@lumen.med.iupui.edu>

>>>>> "Johann" == Johann Hibschman <johann at physics.berkeley.edu> writes:

    > [Yes, I know.  I'm fouling up the reference chain.  But
    > there's weirdness afoot, and I can't do any better.]

    > Hugues Demers writes:

    >>> 2. how are you reading them in?

    >> I use a module written by M. Miller called TableIO to read
    >> my data because it's fast. It read text file into Numeric
    >> array

    > Aha.

    > This might be it.  IIRC, some of the innards of the
    > reference counting changed at some point in the NumPy
    > distribution.  If you're using a different version of NumPy
    > than TableIO was designed for, there might be a memory
    > allocation problem.

The innards of _tableio doesn't use NumPy at all.  All it does is
tokenize a file, run atof on it and return a list.  The list is
stuffed into a dictionary and there certainly could be some
weirdness in my reference counting there, but I've never
experienced the sort of problem that Demers reports.  NumPy is
used in a python wrapper to convert the list to an array.

Maybe Hugues Demers would post the offending code for us to look
at?

Regards, Mike

-- 
Michael A. Miller                      mmiller3 at iupui.edu
  Krannert Institute of Cardiology, IU School of Medicine


From mwh21 at cam.ac.uk  Fri Apr 14 10:17:35 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 14 Apr 2000 15:17:35 +0100
Subject: starship.python.net
References: <Pine.OSF.4.10.10004111721150.9584-100000@prfdec.natur.cuni.cz> <Pine.OSF.4.10.10004111740110.11294-100000@prfdec.natur.cuni.cz>
Message-ID: <m3itxkg41s.fsf@atrus.jesus.cam.ac.uk>

Michal Bozon <bozon at natur.cuni.cz> writes:

> > What the hell is with the starship.python.net server? Today and yesterday
> > I cannont connect (from Czech Republic).
> > Someone knows?
> > 
> > Michal Bozon
> > 
> 
> Just after I posted this message I tried connection again, and it is OK.

Really?  Still seems to be wedged from here.  I can ping it, but can't
get it to play on any other port.

Cheers,
M.

-- 
  incidentally, asking why things are "left out of the language" is
  a good sign that the asker is fairly clueless.
                                      -- Erik Naggum, comp.lang.lisp


From rcunning at acm.org  Wed Apr 12 11:45:13 2000
From: rcunning at acm.org (Robert W. Cunningham)
Date: Wed, 12 Apr 2000 15:45:13 GMT
Subject: Python/Perl Popularity (Re: A Mountain of Perl...)
References: <200004102111.HAA03380@envy.fulcrum.com.au> <8d0pcq$7ti$1@nnrp1.deja.com> <8d1inu$nsq$1@newshost.accu.uu.nl>
Message-ID: <38F4996F.E3BD2626@acm.org>

Martijn Faassen wrote:

> tom__98 at my-deja.com wrote:
> >  -- How can I browse the Python documentation without firing
> >     up a web browser?
>
> A text mode web browser can be useful for this. I've found w3m on
> Linux to be very nice. I can just do:
>
> w3m file:///usr/doc/python-docs-1.5.2/html/lib/modindex.html
>
> And the documentation page is there instantly.

The text-mode browser named "links" (look for it on Freshmeat) is also
very fast, and supports (IMHO) a more complete HTML renderer.


-BobC




From hgg9140 at seanet.com  Fri Apr 14 00:45:05 2000
From: hgg9140 at seanet.com (Harry George)
Date: 13 Apr 2000 21:45:05 -0700
Subject: Ann: pdx, cgipm, pyperl now use distutils
Message-ID: <m3ln2h9tpq.fsf@wilma.george.org>

Also, mkpythonproj generates the basic setup.py

See:
http://www.seanet.com/~hgg9140/comp/index.html

-- 
Harry George
hgg9140 at seanet.com


From shogan at iel.ie  Tue Apr 25 08:56:44 2000
From: shogan at iel.ie (Shaun)
Date: Tue, 25 Apr 2000 13:56:44 +0100
Subject: what does XDR stand for? and what is it?
Message-ID: <002101bfaeb5$b57f5dc0$2801000a@iel.ie>

im here looking and trying to figure out the xdrlib module, 
it says here - http://www.python.org/doc/current/lib/module-xdrlib.html
"The xdrlib module defines two classes, 
one for packing variables into XDR representation, 
and another for unpacking from XDR representation. 
There are also two exception classes."
my question is what is XDR? what does it stand for?
thanks
Shaun
=====================
Shaun Hogan
Interactive Enterprise Ltd.
alt. E-mail : shaun_hogan at yahoo.com
+353 86 8342529






From ivanlan at callware.com  Wed Apr  5 13:11:18 2000
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Wed, 05 Apr 2000 11:11:18 -0600
Subject: Teaching the art of programming, in python
References: <20000405053920.20579.qmail@web2102.mail.yahoo.com>
	    <8cfe42$o02$1@nnrp1.deja.com> <dpJG4.3744$HG1.106895@nnrp1.uunet.ca>
	    <38EB68A3.B1B5809F@callware.com> <8cfqa7$4ml$3@newsserver.rrzn.uni-hannover.de>
Message-ID: <38EB73B6.6C0913A7@callware.com>

Hi All--

Bernhard Reiter wrote:
> 

[snip]

> And all in 24 hours?
> This is as misleading as the dummy attribute.
> 

I don't think so.  You can go at your own speed.  The book is broken up
into 24 chapters, each taking more or less than an hour to read.  If you
pay attention and apply yourself, you could, I think, arrive at some
degree of understanding in the overall 24-hour time frame--give or take
a few hours.

I don't think you could do this with Perl, unless you already had a
programming language or two under your belt, or at least were already
familiar with many of the Unix tools it's based upon.  But Python is
clear, elegant, easy-to-read and understand, and there are very, very
few gotchas and surprises.  And I'm assuming, as you yourself said in a
separate post, "people who do Python are no dummies."

I picked up the basics of Python in perhaps four hours.  That's a
tribute to the language, not me.  Given an interested, attentive newbie,
I think it's entirely possible to teach programming and expose the
newbie to some of the finer things in life in 24 hours or so--with
Python.

> Ranting-about-dummy-booktitles-ly-yours,
>  

<ranting-is-misleading>-ly y'rs,
Ivan
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
http://www.pauahtun.org 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
Author:  Teach Yourself Python in 24 Hours



From brockn at deskmedia.com  Mon Apr 10 15:30:47 2000
From: brockn at deskmedia.com (Brock Noland)
Date: Mon, 10 Apr 2000 14:30:47 -0500
Subject: Jpython??
Message-ID: <955394930.758181@newdm.deskmedia.com>


Hello   I am a web developer.. Well part time.  I can 'program' in HTML and
am interested in learning Java but I was told that I should learn Python
first because it is a fairly newbie friendly language.  To the point I saw a
post about Jpython yesterday what is it?? I am completely in the dark can
someone enlighten me?? Thanks

Brock




From tismer at tismer.com  Fri Apr 21 15:20:08 2000
From: tismer at tismer.com (Christian Tismer)
Date: Fri, 21 Apr 2000 21:20:08 +0200
Subject: docstrings
References: <00bufsonasp7bgkf86lfb6sv5np5v8mqlj@4ax.com> <8dnasp$iep$1@news.wrc.xerox.com> <6v61gs8b1op8esg3o7hf0hii97jlvma01r@4ax.com>
Message-ID: <3900A9E8.3D16413F@tismer.com>


Ben Wolfson wrote:
> 
> On 20 Apr 2000 16:22:49 GMT, mjackson at wc.eso.mc.xerox.com (Mark
> Jackson) wrote:
> 
> >Ben Wolfson <rumjuggler at home.com> writes:
> >> is there a way to use %c and %s within a docstring, or to add strings
> >> together in a docstring, so that if some strings are used in all
> >> docstrings, but might be changed, I can simply change one or two
> >> variables?
> >
> >This might do what you want, although it's ugly (in the sense that it
> >serves the code-documentation - as opposed to user-documentation -
> >function less well):
> >
> >    cat > spam.py
> >    """I am a docstring"""
> >
> >    sharedoc = """ and I am a widely-reused docstring clause"""
> >
> >    __doc__ = __doc__ + sharedoc
> >
> >    del sharedoc
> >    ^D
> >    yngvi> python
> >    Python 1.5.2 (#7, May  6 1999, 14:39:45)  [GCC 2.8.1] on sunos5
> >    Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
> >    >>> import spam
> >    >>> print spam.__doc__
> >    I am a docstring and I am a widely-reused docstring clause
> >    >>>
> 
> This approach doesn't work for function docstrings, unless I tried it
> wrong.

You cannot do this inside the function.
But the __doc__ attribute is writable, so in
your module you can do something like this

sharedoc = """ (imported from module %s)""" % __name__

def somefunc():
    """I am the useless function somefunc"""
    pass

somefunc.__doc__ = somefunc.__doc__ + sharedoc

ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaunstr. 26                  :    *Starship* http://starship.python.net
14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     where do you want to jump today?   http://www.stackless.com



From rupole at compaq.net  Thu Apr 13 17:45:00 2000
From: rupole at compaq.net (Roger Upole)
Date: Thu, 13 Apr 2000 21:45:00 GMT
Subject: windows registry
References: <era0iGYp$GA.273@net003s>
Message-ID: <udrJ4.116036$ra.2711022@news.easynews.com>

Use the win32 extensions, win32api.Reg...
          Roger Upole

"A[r]TA" <arta at x-stream.nl> wrote in message news:era0iGYp$GA.273 at net003s...
> Heyz..
>
> How do I edit the windows registry with Python?
>
> Thnx, A[r]TA
>
>
> --
> We shall not cease from exploration, and the end of all our exploring will
> be to arrive where we started and know the place for the first time.
> - T. S. Eliot
>
>
>




From richard_chamberlain at ntlworld.com  Sun Apr 30 15:37:34 2000
From: richard_chamberlain at ntlworld.com (Richard Chamberlain)
Date: Sun, 30 Apr 2000 20:37:34 +0100
Subject: xitami configuration?
References: <wPZO4.13426$r6.69041@stones>
Message-ID: <wU%O4.3706$E44.82923@news6-win.server.ntlworld.com>

Steven,

I don't really do anything to set Xitami up but just drop:

#!c:\progra~1\python\python

on to the top of my script.

Richard

Steven <steven at nospam.trance.org> wrote in message
news:wPZO4.13426$r6.69041 at stones...
> I'm just starting out with python after programming php3 for the past 8
> months and i'm trying to set it up with xitami.
>
> could an xitami/python user please tell me what line to add to my filters
> section in the setup file? I tried:
>
> .py=D:\progra~2\python\python.exe
>
> which will envoke python and run a simple script fine, but won't pass any
> arguments (unless i'm doing it wrong). If this is correct, how do i pass
> arguments to a script?
>
>




From ivanlan at callware.com  Fri Apr 21 22:17:57 2000
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Fri, 21 Apr 2000 20:17:57 -0600
Subject: Pythonware Website???? Gone on vacation with Starship? ;-))
References: <1255771354-36910521@hypernet.com> <8dqp6u$plm$1@newshost.accu.uu.nl>
Message-ID: <39010BD5.3360C362@callware.com>

Hola, Listeros--

Martijn Faassen wrote:
> 

[snip]

> Gordon McMillan stood in the desert, looking sadly at the remains of Guido's
> time machine.
> 

[more rabid snippery]

> Decades passed while Gordon and the timbot did their work. The timbot sent
> off some surreptitious messages to various people. John McCarthy got a
> few hints on recursion. Bill Gates was funded by an anonymous investor.
> (so, they needed the money! what?!) Larry Wall received a pamphlet
> on postmodern linguistic analysis. And one christmas, Guido van Rossum
> received a thick envelope containing the complete plans for a working time
> machine.
> 
> Do-you-have-to-kill-me-now?-ly yours,
> 
> Martijn
> 

*Only* if you give away the story of where the Mayans got their X
terminals!

<god-9-is-watching-you-very-carefully,-martijn>-ly y'rs,
Ivan;-)
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
http://www.pauahtun.org and
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
Author:  Teach Yourself Python in 24 Hours



From jmassung at magpiesystems.com  Mon Apr 17 16:50:19 2000
From: jmassung at magpiesystems.com (Jeff Massung)
Date: Mon, 17 Apr 2000 14:50:19 -0600
Subject: else clauses in while and for loops
References: <Zv_C4.6415$4N.169440@news.tpnet.pl> <38DD2EE3.C7F3CB54@webamused.com> <8boe2k$gmh$1@inputplus.demon.co.uk> <v0RD4.13$n68.186989568@newsb.telia.net> <8c7b29$37v$1@inputplus.demon.co.uk> <38F79B93.B03F0A6B@mindspring.com>
Message-ID: <2JKK4.325$vr2.19414@news.uswest.net>

I'm new to Python, so perhaps I'm miss reading something in these posts, but
it appears as though the "else" statement (especially in while loops -
unless infinite) always get executed regardless.

i = 0
while i<=5:
    i = i + 1
    print i
else:
    print 10

should spit out "1 2 3 4 5 10" right?

Now a use that I could definitely see for the "else" would be if the loop
were never entered at all! Example:

i = 6
while i<=5:
    i = i+1
    print i
else:
    print 10

would print "10", and if 'i' started at 0, would print "1 2 3 4 5".

That could definitely be useful, because it would eliminate an "if"
statement right before the "while" statement. Extremely useful for debugging
purposes.

Jeff

Chuck Esterbrook wrote in message <38F79B93.B03F0A6B at mindspring.com>...
>So did anyone present, in this thread, an example that shows a real need?





From amphgobb at bjord.org  Sat Apr 22 00:44:43 2000
From: amphgobb at bjord.org (amphgobb)
Date: Sat, 22 Apr 2000 04:44:43 GMT
Subject: Using python on the web
References: <MPG.1365adcfa8470eef9896aa@news> <MPG.136631472f892a119896ac@news> <_W2L4.986$GK5.20898@news2-win.server.ntlworld.com> <MPG.13665ef93eedbda29896ad@news> <BC4B61746187C623.3FA0E7D58DF4EA7E.ECC1D97F577DEDE5@lp.airnews.net>
Message-ID: <slrn8g28ah.cc.amphgobb@eccie.blackflag.net>

>We were wrong.
>than CGI-based alternatives), let alone compete with CGI's
>portability, ...
>I recognize no one here is disputing the proposition.  It's
>just struck me recently how pervasive the propaganda on this
>score is.  CGI has an *abyssmal* popular reputation.

There is a reason for that and I question whether the CGI naysayers
were ever wrong. The arrival of mod_perl and mod_php provide two
distinct development platforms which do not need to fork a new 
process for each instance of the script execution.

I am not aware of many sysadmins who allow universal CGI access without
mod_perl, and a cgi-wrapper. 

For my money, I will stick with these 'next-generation' web technologies. 


From embed at geocities.com  Thu Apr 13 11:02:57 2000
From: embed at geocities.com (Warren Postma)
Date: Thu, 13 Apr 2000 11:02:57 -0400
Subject: How to Copy a Dictionary from a C Extension?
Message-ID: <JklJ4.6736$HG1.192758@nnrp1.uunet.ca>

I am embedding Python. I need to make a copy of the Globals dictionary when
I start a new session in my embedded system, then delete it, and recreate
it, so I can "drop" any changes to the globals after a remote session is
completed.

Problem: If you want to copy the existing Globals used by the interpreter,
what is the way to do this?  I found pydict_copy in dictobject.c, but it's
static, so I can't invoke it directly.

Warren




From grisha at ispol.com  Thu Apr 27 08:48:12 2000
From: grisha at ispol.com (Gregory Trubetskoy)
Date: Thu, 27 Apr 2000 08:48:12 -0400
Subject: PyApache memory leak?
In-Reply-To: <Pine.LNX.4.21.0004211549250.4798-100000@fep132.fep.ru>
References: <Pine.LNX.4.21.0004211549250.4798-100000@fep132.fep.ru>
Message-ID: <Pine.LNX.4.21.0004270834490.27709-100000@saturn.ispol.com>

On Fri, 21 Apr 2000, Oleg Broytmann wrote:

>    This don't help saving memory. Other way round - PyApache destroys
> subinterpreter after processing a request, thus saving memory; but httpdapy
> keeps the interpreter with all (unneccessary and superfluos) imported
> modules, thus leaking memory.

I wouldn't call this "leaking". "Leaking" means forgetting to free memory
where you should be freeing it (like overlooking a Py_DECREF), loosing a
pointer to memory, etc. To the best of my knowledge, httpdapy 1.7b does
not leak memory. I've done a lot of work to make sure of this.

Another note to consider - I am not 100% sure that Py_EndInterpreter()
actually frees all memory. In fact I'm almost sure it doesn't, and you
still need to watch your reference counts if you create object via the
C-API. If that is the case, then PyApache has a higher likelyhood of
actually leaking memory.

BTW, if I remember it correctly, the decision by Lele (the author of
PyApache) to recreate the interpreter was based not on memory usage, but
because the CGI wouldn't work as expected otherwise and for security
reasons.

Grisha



From mhammond at skippinet.com.au  Mon Apr  3 19:37:07 2000
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Mon, 03 Apr 2000 23:37:07 GMT
Subject: Visual Interdev & Python
References: <sehlnm24eop139@corp.supernews.com>
Message-ID: <DW9G4.4031$1b4.19487@news-server.bigpond.net.au>

"Arnaldo Riquelme" <javanet at dynacap.com> wrote in message
news:sehlnm24eop139 at corp.supernews.com...

> Is there any way to integrate Visual Interdev with python.
Visual Interdev
> seems to be a very popular tool
> for ASPs and I think it could be a great way to attract Windows
Developers
> (VB) to use python.

Sure - it should all work.  Interdev has no specific knowledge
about .py files, but that is OK.  You can use Python's active
scripting, including debugging it, using interdev.

Install the latest win32all, and look for the Active Scripting
documentation...

Mark.





From jmassung at magpiesystems.com  Wed Apr 19 11:04:56 2000
From: jmassung at magpiesystems.com (Jeff Massung)
Date: Wed, 19 Apr 2000 09:04:56 -0600
Subject: Mac Python Questions
Message-ID: <IRjL4.271$CP3.3183@news.uswest.net>

Okay, I've been looking all over the net and the documentation that came
with the MacPython distribution, but can't seem to find the answers to these
questions. Hopefully someone else out there on a Mac can help me out?

a) There is a file In $(PYTHON):Mac:Contrib:... called PyColorEdit.py. Good
color editor. I would like to use that instead of PyEdit.py. Is there a way
to install the color one?

b) No matter what I do, I can't get Tkinter to work :(. Lately I'm led to
believe that the Mac version of Python doesn't use it, but uses the Mac
stuff instead. Is this the case? If not, how can I get Tkinter to work?
Whenever I run something that imports it, it dies on the line with the
comment "If this dies your computer might not be set up for tkinter". How
can I set up my machine for it? Thanks :)

c) Back to PyColorEdit.py: when I run it in Python IDE, it works perfect.
But if I use BuildApplication it dies during compile with errors in
macmodulefind.py. My question is just why this would happen? And it seems to
happen on other files I've tried to compile, too.

Thanks, everyone - you've been great! Good NG.

Jeff




From urner at alumni.princeton.edu  Sun Apr  2 14:10:44 2000
From: urner at alumni.princeton.edu (Kirby Urner)
Date: Sun, 02 Apr 2000 11:10:44 -0700
Subject: Learning in Stereo:  Math + Python
Message-ID: <ua1fes81hmb12dh1klbv0p7d0373qhleps@4ax.com>

Re: "Learning in Stereo" by K. Urner (April 2, 2000)

VHLLs (very high level languages) are human-readable
-- not just computer-readable.  

VHLL programs = notations suitable for expressing 
high precision ops -- such as we find in math books.

Yes, of course, we still need to learn the time-honored 
symbols, like SIGMA (<- capital greek letter goes here), 
accepted internationally and accessible to all with 
the proper training (and the right typesetting equipment)

But why not "hit two targets with one throw" and learn 
a VHLL in tandem?  Everyone should know some computer
language or other, n'est pas?

By learning a VHLL in tandem with math, you'll be able 
to translate back and forth between math symbols and 
a computer language -- using each to interpret the 
other (call it "learning in stereo").

Enter Python, a VHLL eminently suitable to this task:

  >>> def sigma(n,func): 
        sum = 0
        for i in range(1,n+1):
           sum = sum + func(i)
        return sum

  >>> def f(x):
        return 1.0/(x*x)

  >>> def pi(n):
        return (6 * sigma(n,f))**0.5

  >>> pi(100)
  3.04936163598

  >>> pi(10000)
  3.14149716395

Yes, we're converging to PI ( = 3.14159...) albiet very 
slowly.  In other words (in more conventional math 
notation):

  PI^2
  ---- = 1 + 1/4 + 1/9 + 1/25 + 1/36 ...
    6
          
       = SIGMA [ 1/i^2]
         i = 1

  i.e. PI = [ 6 * SIGMA [1/i^2] ]^(1/2)
                  i = 1

At my Oregon Curriculum network website, you'll find a
4-part essay entitled 'Numeracy + Computer Literacy' 
series.  This will give you a clear idea of my approach.

See: http://www.inetarena.com/~pdx4d/ocn/cp4e.html

Is this a new idea?  Not really -- people have used 
BASIC for the same purpose.  But Python has a more up 
to date design, is the "new BASIC" for people just 
starting out today.  Your kids get to start with 
something better than you had as a kid.

Of course what you learn from Python you can adapt to 
other languages, such as Java, C/C++ and Perl.  

In the mean time, why not teach and/or learn object 
oriented programming and spatial geometry at the same 
time?  The math might come in handy as well.

This what we're cooking up at the Oregon Curriculum 
network website.

Come on by and give us a try!

Kirby
Curriculum writer




From aahz at netcom.com  Mon Apr 10 00:52:58 2000
From: aahz at netcom.com (Aahz Maruch)
Date: 10 Apr 2000 04:52:58 GMT
Subject: Paid to work in Python? (was: Re: Choice Of Language)
References: <NDBBJPNCJLKKIOBLDOMJMEMECDAA.DavidA@ActiveState.com> <8c7tt5$7dl$1@slb3.atl.mindspring.net> <38EBD8C0.86FD484D@pacbell.net>
Message-ID: <8crmna$hj6$1@slb2.atl.mindspring.net>

In article <38EBD8C0.86FD484D at pacbell.net>,
Michael Naumann  <mnaumann at pacbell.net> wrote:
>Aahz Maruch wrote:
>> In article <NDBBJPNCJLKKIOBLDOMJMEMECDAA.DavidA at ActiveState.com>,
>> David Ascher <DavidA at ActiveState.com> wrote:
>>>Grant Edwards
>>>>
>>>> Pointless tanget: Is it just me, or is Google really the last usable
>>>> search engine?
>>>
>>>And it's _all_ due to their use of Python!  Cool, eh? =)
>>
>> That doesn't explain why Infoseek is no longer particularly usable.
>
>Hmmmmm...  It's actually a breeze to work with -- I have no idea
>what problems you may be having...

I find that the Infoseek web site (now reborn as infoseek.go.com) just
doesn't return useful results the way it used to.  For a while, when
Google was in its infancy and hadn't indexed much of the web, I'd use
Infoseek first and then switch to Google if I didn't get useful results
out of Infoseek.  As time passed, I reached the point where only half of
my Infoseek searches returned useful results and I dropped Infoseek like
a hot rock.
--
                      --- Aahz (Copyright 2000 by aahz at netcom.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

Why doesn't "Just Say NO" include caffeine, nicotine, alcohol, Prozac,
and Ritalin?  --Aahz


From Wolfgang.Strobl at gmd.de  Tue Apr  4 10:02:15 2000
From: Wolfgang.Strobl at gmd.de (Wolfgang Strobl)
Date: Tue, 04 Apr 2000 16:02:15 +0200
Subject: win32all-130 for Python 1.6 available
References: <_ziF4.1488$1b4.6926@news-server.bigpond.net.au>
Message-ID: <d5rjeskov6c6drsok4n7vvcvvi74155lna@4ax.com>

"Mark Hammond" <mhammond at skippinet.com.au> wrote:

>Please let me know of any problems...

Its a minor problem, and not related to Python 1.6, but anyway.

While using the debugger, I've managed to move the floating breakpoint
list window to a position where the title bar is outside (above) the
desktop. Now these detachable windows can be moved by dragging them by
their title bar, only, so I'm stuck. I've found the registry entry in
question (or so I thougth), and fiddled with Software\Python 1.6\Python
for Win32\ToolbarDebugging-Bar7, but in vain.

I'm using Win2000 prof./US btw.

-- 
      o      (    Wolfgang.Strobl at gmd.de          (+49 2241) 14-2394
     /\        *  GMD Forschungszentrum Informationstechnik GmbH
   _`\ `_<===     Schloss Birlinghoven,           | #include
__(_)/_(_)___.-._ D-53754 Sankt Augustin, Germany | <std.disclaimer>


From echuck at mindspring.com  Sat Apr  8 22:00:58 2000
From: echuck at mindspring.com (Chuck Esterbrook)
Date: Sat, 08 Apr 2000 22:00:58 -0400
Subject: Python paradigms
References: <8cmr70$b0e$1@pegasus.csx.cam.ac.uk> <8cntqp$upi$1@slb0.atl.mindspring.net> <8cocgb$iuk$1@pegasus.csx.cam.ac.uk> <8cokt5$opj$1@slb1.atl.mindspring.net>
Message-ID: <38EFE45A.2E866C58@mindspring.com>

Andrew Dalke wrote:

> Me:
> >>How about
> >>  x = getattr(a, "weeble", 0) + getattr(b, "wombat", 0)
>
> Nick Maclaren:
> >Thank you, yes, that does deal with that specific example.  But
> >it does not extend to the general case.
>
> Ahh, I thought the general case was "test if something exists
> and, if so, get data from it, otherwise use a given value."
>
> What is the general case you are interested in - that of C's
> ternary ?: operator?  I've rarely missed it, because most of
> the subcases are handled by things like getattr.
>
>                     Andrew
>                     dalke at acm.org

I miss ?: and in fact, find myself writing extra if-else statements because of it. I think ?: is a pretty natural construct.

Maybe Python 3K....

-Chuck




From phd at phd.russ.ru  Tue Apr 25 04:15:07 2000
From: phd at phd.russ.ru (Oleg Broytmann)
Date: Tue, 25 Apr 2000 08:15:07 +0000 (GMT)
Subject: Trouble compiling PyGreSQL 2.4 on Linux
In-Reply-To: <7ldngy6z.fsf@jumbo.skott.dk>
Message-ID: <Pine.LNX.4.21.0004250814050.13869-100000@fep132.fep.ru>

On Mon, 24 Apr 2000, Martin Sk?tt wrote:
> >    Better, learn to write Distutil, and write one for PyGreSQL - that's a
> > Real Challenge!
> 
> I have allready thought about it and that seems a lot smarter, but
> that should propably be left over to the more experienced Python installers. 

   Today I saw Distutils-based solution in PyGreSQL mail list!
Check it and report results.

Oleg.            (All opinions are mine and not of my employer)
---- 
    Oleg Broytmann      Foundation for Effective Policies      phd at phd.russ.ru
           Programmers don't die, they just GOSUB without RETURN.




From webmaster at nxinteractive.com  Thu Apr 20 14:44:22 2000
From: webmaster at nxinteractive.com (webmaster at nxinteractive.com)
Date: 20 Apr 2000 18:44:22 GMT
Subject: Investing
Message-ID: <38ff5006.0@news.cyberway.com.sg>

Affiliate and associate programs have quietly swept Internet marketing.
Discover how to set up you online sales channel now.
http://www.nxinteractive.com/ActiveMarketPlace.htm

You're about to discover some of the best insider tactics, techniques and
strategies for building and promoting a high-income business on the
Internet...
http://www.nxinteractive.com/AISMedia1.htm

Discover the pros' proven, "insider" strategies that reveal how you can
quickly build your own high-income Internet business that you can operate in
your spare time from any computer in the world.
http://www.nxinteractive.com/AISMedia2.htm

If you're interested in complete IPO information, you've come to the right
place! With 3,846 IPO filings and 1,582 underwriters we have the most
comprehensive database on the web. Utilize the free Alert-My-IPO! feature to
keep track of up to twenty (20) IPOs from a single web page.
http://www.nxinteractive.com/Alert-IPO.htm

Why having your own associate program can explode your traffic and sales...
One of the hottest ways of promoting your business online. Because you don't
spend any money UNTIL you make money... the best way to do business.
http://www.nxinteractive.com/AssocTRAC.htm

Business For You-A fully diversified direct marketing company specializing
in a wide range of products and services to enable individuals to start
their Own Business. With this wide array to choose from, BusinessForYou.com
is the ultimate opportunity for the beginning entrepreneur.
http://www.nxinteractive.com/Business4U.htm

Clicksure is the world's most comprehensive and reliable quality e-commerce
certification program. We rigorously verify online merchants against the
Quality Standard for Electronic Commerce before awarding our Quality Seal.
http://www.nxinteractive.com/ClickSure.htm

CreditNow is an industry leader in providing online consumer credit
services. There are over 45 million people in the USA who have imperfect
credit. These people are very receptive, and motivated to receive our
services.
http://www.nxinteractive.com/CreditNow.htm

QuoteXpress - track the stocks you want to follow...  alerting you with
real-time information;  SplitXpress - real-time notification of breaking
market news; RelayXpress - financial markets, news, weather and lifestyle
information; CompanyNewsX - alerts you of breaking news about the companies
you want to track or invest in.
http://www.nxinteractive.com/Datalink.htm

The dot com directory is a revolutionary service that enables users to stop
'searching' and start finding business sites and data on the Web. In
addition, it provides the latest stock information, revenue statistics, news
reports...
http://www.nxinteractive.com/DotCom.htm

Exchange-it is a free banner exchange program enabling you to promote your
service or product by advertising across a network of web sites.
http://www.nxinteractive.com/ExchangeIt.htm

Marketing Tips strategies for internet marketing, online advertising &
website promotion increase profits for small offices & home based
businesses. Entrepreneurs or home offices access free information &
marketing tips to promote products & increase profits.
http://www.nxinteractive.com/IMC.htm

"How You Can Successfully Market Anything on the Internet With No Previous
Experience and a $150 Budget" A step-by-step guide to building a profitable
Internet business using free and low cost marketing concepts anyone can do
from home in their spare time.
http://www.nxinteractive.com/InternetMarketing.htm

LendingTree is the online loan center that quickly connects you to a network
of lenders who compete for your business.  Find a mortgage, home equity
loan, credit card, personal loan, auto loan or debt consolidation loan NOW.
http://www.nxinteractive.com/LendingTree.htm

MoneyCentral is the best source for all your personal finance needs. Get
quotes, stock research, portfolio tracking, news, and online bill payment.
Keep informed and plan for retirement, college, taxes, insurance, real
estate, savings and more.
http://www.nxinteractive.com/MoneyCentral.htm

What separates you from a professional investor ?  One little click between
you and awardwinning MoneyMaximizer software.
http://www.nxinteractive.com/MoneyMaximizer.htm

Stock and Option investing advisory published daily.  Free two week trial.
For aggressive online traders, investors and daytraders.
http://www.nxinteractive.com/NetPicks.htm

Investing site targeting individual investors interested in learning more
about trading stocks online.  Infostock features stock quotes, and a
fee-based newsletter with stock recommendations with a special focus on
technical analysis.
http://www.nxinteractive.com/Singles&Double.htm

Systematic daily trading signals on the S&P 500 stock index for individual
investors asset management and financial firms.
http://www.nxinteractive.com/spTrading.htm

Kevin C. Myers, best-selling author, presents real estate investment
techniques using quick flip strategies and insider secrets to acquire,
rehab, sell, fixer-uppers, handyman specials, foreclosures, tax lien
certificate properties...
http://www.nxinteractive.com/StreetSmart.htm

Stock quotes to your pager, PCS phone. Todays current Business and Financial
News. Don't miss an investment opportunity!  Let Timely Quote alert you to
your stocks every day! Searchable financial, investment and business site
directory and search engine!
http://www.nxinteractive.com/TimelyQuote.htm

****************************************************************************
 A FREE offline portal application providing FREE contents, such as
 World Time, PIM, Articles, News, Realtime Quotes, Horoscopes,
 Free/Sharewares, Internet Access/Call, Fax, Weather, Sweepstakes...
 http://www.nxinteractive.com
****************************************************************************



From wware at world.std.com  Fri Apr 21 10:55:47 2000
From: wware at world.std.com (Will Ware)
Date: Fri, 21 Apr 2000 14:55:47 GMT
Subject: The Simple Economics of Open Source
References: <14591.15630.192472.440375@goon.cnri.reston.va.us> <slrn8fv2rd.1kt.neelk@brick.cswv.com> <3900611F.3B525BDA@cfdrc.com>
Message-ID: <FtDG4z.774@world.std.com>

Ed (elb at cfdrc.com) wrote:
> ... I didn't think much of the article...
> ...they gave the impression that programmers behave
> somewhat like monkeys, motivated primarily by concerns
> about group status and dominance...

They also mentioned, but immediately belittled, the possibility
that altruism might be a real motivation. They may have been
projecting their own thinking upon programmers in doing so. I'm
not an open source purist, but I can easily see that a future world
with a lot of free software will be a lot more pleasant for me to
live in than one from which free software is absent. Benefits to
the rest of society and to myself are not in conflict here. Maybe
that's not really altruism, maybe it's just long-term thinking, a
motivation they didn't mention at all.
-- 
 - - - - - - - - - - - - - - - - - - - - - - - -
Resistance is futile. Capacitance is efficacious.
Will Ware	email:    wware @ world.std.com


From sabren at manifestation.com  Thu Apr 27 04:28:36 2000
From: sabren at manifestation.com (Michal Wallace (sabren))
Date: Thu, 27 Apr 2000 04:28:36 -0400 (EDT)
Subject: mysqldb at sourceforge?
Message-ID: <Pine.BSF.4.10.10004270424570.12033-100000@chi.pair.com>

Hey all,

  I've got a MySQL database set up at sourceforge, and I'm trying to
set up MySQLdb.. I've done this on several other unix accounts, but I
can't seem to find the mysql libraries on sourceforge, though the
mysql is on the machine... If anyone else has gotten this working, I'd
love to hear about it.

Thanks,

- Michal
-------------------------------------------------------------------------
http://www.manifestation.com/         http://www.linkwatcher.com/metalog/
-------------------------------------------------------------------------




From pinard at iro.umontreal.ca  Tue Apr  4 08:46:05 2000
From: pinard at iro.umontreal.ca (=?ISO-8859-1?Q?Fran=E7ois_Pinard?=)
Date: 04 Apr 2000 08:46:05 -0400
Subject: Unicode program representation
In-Reply-To: fredrik@pythonware.com's message of "Tue, 04 Apr 2000 09:32:38 GMT"
References: <015c01bf9d0b$76f9b2a0$59dd3fcb@neil> <3xXF4.1060$n68.186533888@newsb.telia.net> <001201bf9d62$3e1ac1c0$36dd3fcb@neil> <%o3G4.1111$n68.191610880@newsb.telia.net> <oqn1nb11l3.fsf@titan.progiciels-bpi.ca> <8cccrm$cc3$1@nnrp1.deja.com>
Message-ID: <oqsnx2nieq.fsf@titan.progiciels-bpi.ca>

fredrik at pythonware.com writes:

> what scares me is if we end up with multiple encodings in the
> same source file.

Oh, I agree with you that is not going to be usual, at least at first.

I foresee it might happen quite naturally in Python scripts generating
multilingual announcements, say.  Or more generally, in Python scripts
holding translation correspondences between languages, or anything in which
we have substantial amount of editing to do in one language followed by a
substantial amount in another language.  Emacs Mule/MULE is likely able to do
this, without resorting to UTF-8.  Unicode on-the-fly combining is not well
supported at the font rendering level (despite it is increasingly needed),
and Unicode raises other disputes, like for the Han unification it imposed.

As for the "simpler" cases like of a Vietnamese sentence quoting a German
proverb, it is likely in the long run to see a common encoding like UTF-8.

> (and no, it's not because we're developing a python development environment
> -- supporting this in our python editor would be fairly trivial ;-)

It would be?  You should address the matter then.  It is a worth goal.

-- 
Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard





From moshez at math.huji.ac.il  Fri Apr 28 03:08:09 2000
From: moshez at math.huji.ac.il (Moshe Zadka)
Date: Fri, 28 Apr 2000 10:08:09 +0300 (IDT)
Subject: Using BaseHTTPServer
In-Reply-To: <200004280649.IAA00894@kazul.rwth-aachen.de>
Message-ID: <Pine.GSO.4.10.10004281007030.10249-100000@sundial>

On Fri, 28 Apr 2000, Lars von Wedel wrote:

> my CGI project suffers from a lengthy delay for starting up
> python and loading all required modules. I'm thinking about
> setting up a small HTTP server in python itself so that the
> interpreter and all the required modules have to be loaded only
> once.
> 
> >From the module documentation it is not clear to me, what I actually
> have to do to catch the GET calls arriving at the server.
> 
> I assume that I can use BaseHTTPServer as it is. I just need
> to implement a working request handler, right?
> 
> Now comes the more fuzzy part:
> 
> Since I do not want to execute external files but internal methods,
> I think I could modify CGIHTTPServer by subclassing it and overriding
> its do_GET() method to decode the request directly? Or would I rather
> use SimpleHTTPServer, here?
> 
> Within any of the do_GET methods, how do I access data that has been
> sent by the client (...&data1=val1?data2=val2?...)?

Use the CGI module <wink>. Just copy the code from CGIHTTPServer to create
the dictionary, (not as os.environ), and use the cgi module to parse
that dictionary.

--
Moshe Zadka <mzadka at geocities.com>. 
http://www.oreilly.com/news/prescod_0300.html
http://www.linux.org.il -- we put the penguin in .com




From poupaerte at my-deja.com  Sun Apr  9 12:02:39 2000
From: poupaerte at my-deja.com (poupaerte at my-deja.com)
Date: Sun, 09 Apr 2000 16:02:39 GMT
Subject: tabular data
References: <8co4l2$5r2$1@nnrp1.deja.com> <O8OH4.46089$Uz3.487916@typhoon.midsouth.rr.com>
Message-ID: <8cq9io$ca0$1@nnrp1.deja.com>

Hi

> def sortedby(list,*indices):

I didn't know Python worked with pointers, or else what does *indices
mean? The fine manuals do not explain the *var construct ...

>    nlist = map( lambda x,indices=indices:

The expression "indices=indices" seems absolutely strange to me. I
don't have a clue of what it could be doing, except for being true at
all times.

Is there some kind of lengthier tutorial on the lambda construct; it
seems quite an advanced construct not easily to be found elsewhere in
other programming languages, a bit like regexes ...

Greetings
Erik


Sent via Deja.com http://www.deja.com/
Before you buy.


From scarblac-spamtrap at pino.selwerd.nl  Fri Apr 14 05:30:04 2000
From: scarblac-spamtrap at pino.selwerd.nl (Remco Gerlich)
Date: 14 Apr 2000 09:30:04 GMT
Subject: Need an enum type extension
References: <38F52A7A.D8841724@gte.net> <JdeJ4.1552$Za1.25652@newsc.telia.net> <38F68AF4.46DDA670@gte.net>
Message-ID: <slrn8fdshr.3qc.scarblac-spamtrap@flits104-37.flits.rug.nl>

sue wrote in comp.lang.python:
> In Ada, I can do things like:
> 
> type MyEnum is ( A, B, C );
> for MyEnum use ( A : 3; B : 7; C : 13 );
> 
> Ada uses "'string" to refer to attributes.
> 
> MyEnum'len == 3.	(number of items in list)
> A'next == B		(next item in list)
> A'value == 3		(its low-level representation)
> B'index = 1		(its position in the list)
> 
> But A == 3 fails.  only A == A passes.  A'value == 3 is
> true.

This simple try seems to do that, at least. It needs some work (EnumVar
needs to be immutable, I suppose), but maybe you can base it on this?

class EnumVar:
  def __init__(self, value, index, enumlist):
    self.value = value
    self.index = index
    self.__enumlist = enumlist
  def next(self):
    return self.__enumlist.get_enum(self.index+1)
	
class EnumList:
  def __init__(self, values):
    self.__list = []
	self.len = len(values)
    for i in range(self.len):
      self.__list.append(EnumVar(values[i], i, self)
  def get_enum(self, i):
    return self.__list[i]
  def get_list(self):
    return self.__list
	
Now use this like:

>>> MyEnum = EnumList([3, 7, 13])
>>> A, B, C = MyEnum.get_list()
>>> MyEnum.len
3
>>> A.next() is B
1
>>> A.value
3
>>> B.index
1
>>> A == 3
0
>>> A == A
1



-- 
Remco Gerlich,  scarblac at pino.selwerd.nl
 12:26pm  up 38 days, 23:46,  8 users,  load average: 1.49, 1.23, 1.18


From PClaerhout at CREO.BE  Wed Apr 12 07:44:03 2000
From: PClaerhout at CREO.BE (Pieter Claerhout)
Date: Wed, 12 Apr 2000 13:44:03 +0200
Subject: Getting version information of .dll and .sys files under Windows 
	NT
Message-ID: <2B1262E83448D211AE4B00A0C9D61B03BA72A2@MSGEURO1>

Hello,

is there a way to access the version information that is included in .sys
and .dll files from Python? I want an way to check for example driver
versions unders Windows NT. Any idea?

Kind regards,


Pieter
Pieter_Claerhout at creoscitex.com



From adjih at crepuscule.com  Sun Apr 30 13:43:05 2000
From: adjih at crepuscule.com (Cedric Adjih)
Date: 30 Apr 2000 17:43:05 GMT
Subject: Do I always have to write "self." ?
References: <270420001706386648%pecora@anvil.nrl.navy.mil> <slrn8ghc6t.f01.scarblac-spamtrap@flits104-37.flits.rug.nl> <8ecsgj$6lp$1@slb0.atl.mindspring.net> <m3bt2tonyg.fsf@atrus.jesus.cam.ac.uk> <300420001008012011%pecora@anvil.nrl.navy.mil> <20000430170952.A502@datapro.co.za>
Message-ID: <8ehrb9$mm5$1@ites.inria.fr>

Steve Crane <craniac at datapro.co.za> wrote:
> On Sun, Apr 30, 2000 at 02:08:00PM +0000, Louis M. Pecora wrote:

>> Isn't this called something like "Hungarian" variable naming?

> Hungarian Notation was introduced by Charles Simonyi (who happened to
> work for Microsoft).  In it's initial form it used a prefix to indicate
> the data type of the variable but not the scope.  Later people started
> to add a scope prefix but I don't think that Microsoft uses this as a
> standard.

> While it may seem to make code harder to read it is in fact very useful
> when working on large projects involving may programmers. 

  Actually this is debatable.

  If you can read, you should be able to read the definition of 
any variable, and there are serious drawbacks. A flame of this
notation is for instance in "Developing Windows NT Device Drivers":
"The so-called `Hungarian Notation' (developped by a Hungarian
programmer at Microsoft) is one of the worst ideas to have hit
software development in many years"
  The main problem, is that when you change the type of a structure
you _cannot_ change the name (because other people are using it),
so the type specification becomes plain wrong. The 'wParam'
is reported to be such an example (now unsigned 32 bits), there
are other examples in Microsoft API (should be worse with Win64).
  Also each time you change the type of one variable/parameter, 
you have change all the variable names of all the code under
your control.
  And finally, for multi-platform code, it makes less sense:
what happens if you don't have 2-bytes integers and you
have code with wXXXX variables (unsigned int 16 bits) ?

-- Cedric



From jmassung at magpiesystems.com  Mon Apr 17 20:26:49 2000
From: jmassung at magpiesystems.com (Jeff Massung)
Date: Mon, 17 Apr 2000 18:26:49 -0600
Subject: else clauses in while and for loops
References: <2JKK4.325$vr2.19414@news.uswest.net> <Zv_C4.6415$4N.169440@news.tpnet.pl> <38DD2EE3.C7F3CB54@webamused.com> <8boe2k$gmh$1@inputplus.demon.co.uk> <v0RD4.13$n68.186989568@newsb.telia.net> <8c7b29$37v$1@inputplus.demon.co.uk> <38F79B93.B03F0A6B@mindspring.com> <8dg14u$f08$1@nntp6.u.washington.edu>
Message-ID: <%TNK4.430$vr2.32615@news.uswest.net>

Donn Cave wrote in message <8dg14u$f08$1 at nntp6.u.washington.edu>...
>Yes, but if you look at the examples, they also use "break".  After a
>loop, "else" basically means "if no iteration of the loop met the
>conditions to break out."  It's a common usage, where in C we'd compare
>the final loop index with the maximum (or whatever) loop index to see
>if the loop ran all the way through.

I don't think that I'm following you, give me an example where the loop
ends, but the else wouldn't be executed and I think I'll better understand.
:)

>| That could definitely be useful, because it would eliminate an "if"
>| statement right before the "while" statement. <SNIP> useful for debugging
>| purposes.

>
>I think I would go along with that if you took out "Extremely" - and then

Consider it gone ;)

>your paragraph would justify better on the right margin, too!  Seriously,
>"else" does invariably execute if the loop was never run once, but as
>mentioned above it also executes in other circumstances.

I think it would be better suited if it didn't execute under "other
circumstances". It would be nice to know _exactly_ when it would execute,
and only under 1 or 2 conditions (maybe that is the way it is now, I just
don't see it).

Jeff




From chapman at bioreason.com  Mon Apr 10 16:11:52 2000
From: chapman at bioreason.com (Mitch Chapman)
Date: Mon, 10 Apr 2000 14:11:52 -0600
Subject: Problem running Python script as cron job
References: <wrd-0AEDDA.12282507042000@news.awenet.com> <0004071555090M.01234@quadra.teleo.net> <wrd-1E3688.12433910042000@news.awenet.com>
Message-ID: <38F23588.5C1C9E7@bioreason.com>

There are a couple of ways to force your output to be immediately 
flushed.

An ugly solution is a sys.stdout.flush() after every print statement.

A less ugly solution is to reopen sys.stdout as soon as
your script starts up, and to turn off output buffering:
  sys.stdout = os.fdopen(sys.stdout.fileno(), 'w', 0)

I thought it used to be possible to do a setbuf on the file object,
provided you hadn't already written to it.  But this appears not to
be the case.

Another option is to change your cron job so it sets the 
PYTHONUNBUFFERED environment variable before starting your program.
Something like this might work:

  /usr/bin/env PYTHONUNBUFFERED=1 /path/to/statsthing -r >/usr/local/tmp/.log

For that matter you could write a shell script wrapper which
set the environment variable, then invoke it from cron:

  #!/bin/sh
  PYTHONUNBUFFERED=1
  export PYTHONUNBUFFERED
  /path/to/statsthing -r >/usr/local/tmp/.log

Here's hoping this is more help than hindrance...

--
Mitch Chapman
chapman at bioreason.com

"William R. Dickson" wrote:
> 
> In article <0004071555090M.01234 at quadra.teleo.net>, Patrick Phalen
> <python-list at teleo.net> wrote:
> 
> > [William R. Dickson, on Fri, 07 Apr 2000]
> > :: Anybody know of any issues with running python scripts from Cron?
> >
> > No. I've had multiple Python scripts running every one, two, five and
> > ten minutes around the clock for years on multiple servers without
> > problems. [...]
> 
> I think so too, but I can't find it, because the script runs flawlessly
> every time I run it by hand, and the buffer never flushes the output to
> the log when it's run by Cron.


From lexberezhny at email.msn.com  Tue Apr  4 20:00:21 2000
From: lexberezhny at email.msn.com (lexberezhny)
Date: Tue, 4 Apr 2000 20:00:21 -0400
Subject: lightweight embedded database table object
References: <1257258860-5013801@hypernet.com>
Message-ID: <uX6DuEpn$GA.239@cpmsnbbsa03>

gadfly? http://www.chordate.com/gadfly.html

 - Lex


Gordon McMillan <gmcm at hypernet.com> wrote in message
news:1257258860-5013801 at hypernet.com...
> Warren Postma wrote:
> >
> > > Hi Warren,
> > > Have you looked at MetaKit? http://www.equi4.com/metakit/python.html
> >
> > It didn't look 'ready for prime time' -- too many disclaimers of the
kind I
> > usually put on something when I expect it's still quite buggy or
unscalable.
> > I suppose I should give it more of a try though, eh!? Is anyone using
this
> > in a 'production environment'?
>
> Yes. Metakit, extensively. Probably less so for Mk4py.
>
> The problems you'll find aren't stability problems. Some of the
> advanced stuff (the relational algebra) can be tricky to get
> working - it's not documented well, and you may find, for
> example, that some operation returns a read-only view where
> updates silently fail. It should all show up during development,
> though.
>
> - Gordon
>




From dgreen at kryten.kitenet.net  Wed Apr  5 00:51:52 2000
From: dgreen at kryten.kitenet.net (Dan Green)
Date: Wed, 05 Apr 2000 04:51:52 GMT
Subject: ASP web-type stuff
References: <RZyG4.3061$9o.5812531@news.magma.ca>
Message-ID: <IDzG4.3063$9o.5826909@news.magma.ca>

My mistake ... slight error in my slrn setup. If you want to contact me about
this, I can be gotten at dgreen at rewindtech.com

In article <RZyG4.3061$9o.5812531 at news.magma.ca>, Dan Green wrote:
>Ahoy, folks.
>	I was looking over things like Embperl and Apache::ASP and liked the
>idea. I would like the idea more if I could do with python. So I had a
>couple of questions:
>
>Is there such a thing?
>Is there interest in such a thing?
>Does anbody want to hear about my 75 lines of code hack that basically does
>the trick, but needs a lot of work?
>
>If nobody cares at all, I'll just shut up and go home. :)
>
>For those of you who have no idea what I'm talking about, it's basically an
>alternative to CGI that allows you to embed the code right into the HTML.
>
>something like this:
>
><HTML>
>	<BODY>
><%  for i in range(5): %>
>	Loop item <%= i %><BR>
>	</BODY>
></HTML>
>
>would look like this when called from a browser:
>
><HTML>
>	<BODY>
>	Loop item 1<BR>
>	Loop item 2<BR>
>	Loop item 3<BR>
>	Loop item 4<BR>
>	Loop item 5<BR>
>	</BODY>
></HTML>
>
>Let me know. If you're interested, please e-mail me.
>
>--
>Dan Green           | "So What? It's your problem to learn to live with,
>Developer           |  destroy us, or make us saints."
>Rewind Technologies |  -- Ministry, "So What?"


-- 
Dan Green           | "So What? It's your problem to learn to live with,
Developer           |  destroy us, or make us saints."
Rewind Technologies |  -- Ministry, "So What?"


From echuck at mindspring.com  Sun Apr  9 22:29:53 2000
From: echuck at mindspring.com (Chuck Esterbrook)
Date: Sun, 09 Apr 2000 22:29:53 -0400
Subject: Python paradigms
References: <8cmr70$b0e$1@pegasus.csx.cam.ac.uk> <8cntqp$upi$1@slb0.atl.mindspring.net> <8covpn$4jq$1@slb6.atl.mindspring.net> <8cqk7u$dmc$1@pegasus.csx.cam.ac.uk>
Message-ID: <38F13CA1.1A91D85@mindspring.com>

Nick Maclaren wrote:

> To the several people who suggested:
>
>     (a() and [b()] or [c()])[0]
>
> Thank you for the suggestion.  I have occasionally used perversities
> like that (such as '*(x ? &a : &b) = y;' in C to mimic the Algol 68
> facility), but I don't regard them as clarifying the code, either!

OK, so what does the Algol 68 code look like for that C code and any other interesting cases?


-Chuck




From frankr at first-mind.de  Wed Apr  5 03:53:17 2000
From: frankr at first-mind.de (Frank Rotermund)
Date: Wed, 5 Apr 2000 09:53:17 +0200
Subject: building Extensions with compile.py
Message-ID: <8cercu$7bb$1@newnews.ilk.de>

Hello everybody,

I have a problem building C extensions for Python. I tried it the way
the Python Manuals describe it ( Using compile.py to get VC++ 6.0 started
... )
but VC gives my a linkage error which tells me, that VC got an unsolved
extern
symbol initxxx ( xxx is my source name ). Trying to solve this error I found
 '/libpath:"C:\programme\python\libs" /export:initxxx' in the
project-properties
of VC. I removed the export part. This resolved in linking without error,
but
the resolting pyd could not be imported in Python.

Has anybody a solution for this problem?





From dan at cgsoftware.com  Thu Apr  6 00:20:21 2000
From: dan at cgsoftware.com (Daniel Berlin+list.python)
Date: 06 Apr 2000 00:20:21 -0400
Subject: Imaginative solutions to connect. Was: Python 1.6 alpha 1 released
In-Reply-To: James Logajan's message of "Wed, 05 Apr 2000 21:03:04 -0700"
References: <Pine.GSO.4.10.10004040035420.21084-100000@sundial> <5E877FD7637AB46E.DABBED9CF3C1D64B.CDDDC05AE7374A78@lp.airnews.net> <5l7leern5v.fsf@eric.cnri.reston.va.us> <38EC0C78.C6170660@Lugoj.Com>
Message-ID: <ya6r3loa.fsf@dan.resnet.rochester.edu>

>>>>> "James" == James Logajan <JamesL at Lugoj.Com> writes:


    James> Guido van Rossum wrote:
    >>  Repeat after me: SOCKETS ARE NOT JUST FOR IP CONNECTIONS.
    >> 
    >> But I will fix the code -- it's the only practical thing to do.


    James> For alternatives 2 and 3 you need some incentive for people
    James> to change; cash might work but could get expensive. Rigging
    James> keyboards to deliver electrical shocks might work but might
    James> be too time consuming. Making the old interface much slower
    James> might help. Perhaps someone can think of a good
    James> incentive...  (get some new and improved methods only if
    James> you use the new class, or something?)  --

Why don't we introduce an official deprecation mechanism so that
python will say "Warning, connect(arg1,arg2) is deprecated, use
connect(arg1) instead" the first time it sees it or something.
Simple lookup table for what's deprecated, and what to say about it.
So code doesn't just break with no programmer visible warning from the
interpreter.
Hey, we could be really evil, and store the number of times we've
warned you, and have the interpreter get more and more annoyed at the
user, finally just refusing to interpret the code at all.
Then again, i'm for the electrical shocks.

--Dan



From ljoramo.list at nickads.com  Tue Apr  4 10:36:31 2000
From: ljoramo.list at nickads.com (Lee Joramo)
Date: Tue, 4 Apr 2000 08:36:31 -0600
Subject: Starving for an Advanced Python Book
In-Reply-To: <20000404040941.7766.qmail@web2106.mail.yahoo.com>
References: <20000404040941.7766.qmail@web2106.mail.yahoo.com>
Message-ID: <v04210102b50faa345450@[192.168.1.2]>

Besides 'Learning Python', I have the following books:

'Python Essential Reference'; David Beazley; New Riders
An excellent reworking of the online documentation. This is a reference
book with out tutorial, but I have learned much from it. HIGHLY  RECOMMENDED

'Python Annotated Archives'; Martin Brown; Osborne/McGraw Hill
Looking for lots of sample code, this book has it. Following the code of
each program is a detailed explanation of how the code works. I find this
type of book very useful. The authors commentary on the code provides
pointers that is often hard to get by reading code alone.

'Python Pocket Reference'; Mark Lutz; O'Reilly
75 pages, nice quick reference book. Only useful for looking up information
with which you are already generally familiar. If you like the small size,
get it. Otherwise, 'Python Essential Reference' will fill all your
reference needs.


--
Lee A. Joramo                      ljoramo at nickads.com
The Nickel Want Ads                www.nickads.com
Internet Manager                   970-242-5555



>Having finished "Learning Python", I'm now looking for an advanced
>Python text.  ORA's "Programming Python" is a natural choice, but I
>can't bear buying this outdated book when the 2nd edition is due out
>in a few months (I'm dying here Lutz, hurry it up already!).  However,
>I really need something in the interim period.  The library reference
>manual is fine but usage is difficult for me to figure out sometimes
>because there are very few examples included.
>
>Does anyone have any suggestions?  I suppose I'm looking for a text
>that has examples of how to solve various kinds of programming
>problems in Python and also examples of code showing how the library
>and various modules actually work.  How is the "Python Essential
>Reference" on examples?  Is it just a rebinding of the python.org
>library reference




From donn at u.washington.edu  Wed Apr 19 12:01:59 2000
From: donn at u.washington.edu (Donn Cave)
Date: 19 Apr 2000 16:01:59 GMT
Subject: Very Horrible Question About Goto's
References: <C5A7720BDEB7D3118A5200E018C200A037EA43@EXCHANGE>
Message-ID: <8dkl9n$f3a$1@nntp6.u.washington.edu>

Quoth Gaetan Corneau <corg at copernic.com>:
[quoting John Mitchell]
|> 1) the only valid use of goto's in C I've seen looks like this:
|>=20
|> 	if (bogus)
|> 		goto cleanup
|> 	do stuff;
|> 	return(okay);
|> cleanup:
|> 	close file handles;
|> 	do cleanup
|> 	return(error)
|
| There are, at least, two other valid uses:
| 1) the most important: goto's are used a lot by code generators (it
| simplifies writing the generator alot in some cases)
|
| 2) in extreme cases, to improve performances.

Now that we have a list of situations where goto is valid, we are on
the road to realizing that it's just an inherently useful tool.
Which it is.  The cleanup idiom shown above is a common case, but
there are an infinite variety of similar flow of control problems
where goto works out well.

| My own rules are:
| - don't use goto's
|
| if you HAVE to use goto's:
| - don't jump from one function to another (if your language supports =
| it)
| - don't jump backward
|
| I was surprised to learn that Python doesn't have a goto statement.
| Would it be difficult to add?

I wouldn't.  In the abstract, goto is a good thing, but it would be
out of place in Python, whose block constructs, exceptions and so
forth already provide plenty of higher level flow control options.

Here's an idea for a rewrite in Python (for the original poster):
Take your routine and implement it as a class, with the routine local
variables stored in the class instance so you can branch around to
various methods without losing them.  Each class method stores a
final state before it returns, except one method which loops over
the final state value and calls whatever method is appropriate for
that state.  This will be terribly verbose if you have a big mass
of little goto branches that do trivial things, so it will help if
you can cook that down a little first with if blocks etc., and
also identify some block-local variables.

	Donn Cave, donn at u.washington.edu


From nascheme at enme.ucalgary.ca  Tue Apr 18 18:16:19 2000
From: nascheme at enme.ucalgary.ca (Neil Schemenauer)
Date: 18 Apr 2000 22:16:19 -0000
Subject: Distributing A Python Application
In-Reply-To: <384449751.956092588323.JavaMail.root@web135-mc.mail.com>
References: <384449751.956092588323.JavaMail.root@web135-mc.mail.com>
Message-ID: <20000418221619.30185.qmail@cranky.arctrix.com>

In comp.lang.python, you wrote:
>I want to distribute a windows GUI application developed in Python/Tkinter.

  http://www.enme.ucalgary.ca/~nascheme/python/standalone.html

Let me know how it goes.

    Neil

-- 
"With a PC, I always felt limited by the software available. On
Unix, I am limited only by my knowledge."
        --Peter J. Schoenster <pschon at baste.magibox.net>



From rob at hooft.net  Tue Apr 11 03:30:35 2000
From: rob at hooft.net (Rob W. W. Hooft)
Date: 11 Apr 2000 09:30:35 +0200
Subject: sys.platform details...
Message-ID: <m3r9cdm6w4.fsf@devel.no.rwd.dinet>

schol[51]:ccd>python
Python 1.5.2b1 (#4, Jan 14 1999, 12:05:48)  [GCC egcs-2.90.21 971202 (egc on irix5
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> import sys
>>> sys.platform
'irix5'
>>> ^D
schol[52]:ccd>uname -a
IRIX schol 6.3 12161207 IP32

This python was compiled on an irix5 box, but is running on an irix6 box.
Hm. I'm off looking for a better way to find out what kind of box my program
is running on....

Rob

-- 
=====   rob at hooft.net          http://www.xs4all.nl/~hooft/rob/  =====
=====   R&D, Nonius BV, Delft  http://www.nonius.nl/             =====
===== PGPid 0xFA19277D ========================== Use Linux! =========


From mhammond at skippinet.com.au  Mon Apr  3 08:22:17 2000
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Mon, 3 Apr 2000 22:22:17 +1000
Subject: [Python-Dev] Re: New Features in Python 1.6
In-Reply-To: <38E884F5.8F2FB271@lemburg.com>
Message-ID: <ECEPKNMJLHAPFFJHDOJBIEOPCHAA.mhammond@skippinet.com.au>

>
> > On Sat, Apr 01, 2000 at 12:00:00PM -0500, Guido van
> Rossum wrote:
>
> The above line has all the answers ;-) ...

That was pretty sneaky tho! <grin>  Had the added twist of being
half-true...

Mark.





From musingattheruins at my-deja.com  Fri Apr  7 15:45:10 2000
From: musingattheruins at my-deja.com (musingattheruins at my-deja.com)
Date: Fri, 07 Apr 2000 19:45:10 GMT
Subject: Palm OS
Message-ID: <8clds2$c25$1@nnrp1.deja.com>

Anyone know of a Python release for Palm?


Sent via Deja.com http://www.deja.com/
Before you buy.


From mhammond at skippinet.com.au  Wed Apr 26 20:46:35 2000
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Thu, 27 Apr 2000 00:46:35 GMT
Subject: Python COM troubles
References: <390a2f04.149999918@news.onlineisp.com>
Message-ID: <L5MN4.5088$v85.30985@news-server.bigpond.net.au>

"William Wicker" <remove.me.wwicker at spectratechnologies.com> wrote in
message news:390a2f04.149999918 at news.onlineisp.com...

> Is there a way around this appearance of typelessness? There is a type
> library, and I've run make-py successfully. I just can't see an easy
> way into it.

You can try running "makepy.py -i", and select your type library.  This
will spit out some code, something like:

gencache.EnsureModule(...)

change this to:

mod = gencache.EnsureModule(...)

And you should be able to create any of the class objects in the makepy
module - eg:

ob = mod.TheClassName()

If you search deja news, there should be lots of info about this - but
generally relating to "Lotus Notes", which also has the same problem.

[BTW - drop them a line, and tell them they should fix their objects :-]

> COM dispatch methods that are documented to return SAFEARRAYS of
> dispatch pointers fail with pythonwin v125 and 128 (the only ones I've
> felt safe enough to try). The traceback is something like this:

> SystemError: NULL result without error in call_object
> >>>

>From a CVS checkin message on 11th of Feb:
Can you believe that we have never been able to recieve arrays of
IDispatch or IUnknown objects?!?!?

If you tried, you would have got "SystemError: NULL result without error
in call_object"

:-)  This was fixed in 129.  But 129 hasnt made it public yet - I will do
this as soon as I can log back in to starship!

Mark.





From dalke at acm.org  Mon Apr 24 04:58:55 2000
From: dalke at acm.org (Andrew Dalke)
Date: Mon, 24 Apr 2000 02:58:55 -0600
Subject: partial parsing?
References: <8dvilp$88s$1@slb6.atl.mindspring.net> <B528E531.4EF9%abrahams@mediaone.net>
Message-ID: <8e1265$imv$1@slb6.atl.mindspring.net>

Dave Abrahams:
>You might think about looking at chart parsing technology, sometimes
>used in natural language processing.


What I have about chart parsing is from Grune&Jacobs' "Parsing
Techniques".  They say that it's the same as CYK, but then elaborate
that variations of CYK are identical to variations of chart parsing.
Now if I only recall how CYK works.  Well, that's why I printed
out that document.  :)

>You can use a chart parser to extract all parseable sub-regions from a
>sequence of tokens. Just turn off the grammar rules you're not interested
in
>and you should be all set (provided the grammar rules you care about don't
>generate false positives due to the lack of contextual information that
>would be neccessary for a complete parse).

I don't quite follow that.  I think there's too much "self similarity"
in what I have because it's a large number of records in a row.  Again,
I've got to read more.

>Just a caveat: in typical program parsers much more time is spent in
lexical
>analysis than in actual parsing. This tends to indicate that using YACC is
>probably about as well as you'll do for speed and trying to optimize away
>parsing time is misguided.


I know in perl that my full parser code was about an order of magnitude
slower than the parse-only-exactly-what-I-want parser.

For a record like:
ID   BPT1_BOVIN     STANDARD;      PRT;   100 AA.
AC   P00974;
[...]
SQ   SEQUENCE   100 AA;  10903 MW;  6A778A4AD763FB19 CRC64;
     MKMSRLCLSV ALLVLLGTLA ASTPGCDTSN QAKAQRPDFC LEPPYTGPCK ARIIRYFYNA
     KAGLCQTFVY GGCRAKRNNF KSAEDCMRTC GGAIGPWENL
//

where I'm interested in the first word of the ID line and the sequence
lines, I should be able to have a (fake mxTools-like) implementation
like:

while not eof:
 skip 5 characters
 read until ' '
 read until 'S' and send read text to callback "id"
 skip 46 characters (the start of the '[')
 while text != "    " and text != "//":
   read up to '\n'
 while text != "//":
   while text == ' ':
     skip the ' '
     read up to ' ' or '\n' and send read text to callback "sq"
 skip 3 characters (the // followed by the \n)

So it seems what I want is better performance in the lexer part, which
agrees with your point.

                    Andrew
                    dalke at acm.org





From effbot at telia.com  Tue Apr 25 12:28:34 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Tue, 25 Apr 2000 16:28:34 GMT
Subject: Newbie Needs Help -- Name and Key errors
References: <bingdo19-4B8F57.12162225042000@news-server.twcny.rr.com>
Message-ID: <SIjN4.1570$wYl.193603072@newsb.telia.net>

Bob Langdon <bingdo19 at earthlink.net> wrote:
> New to Python, and moving a site for a client. The backend sas developed
> by another developer in Python.  I've i9nstalled Python1.5.2 in my home
> direcpory and tested it. The interpreter seems to functioning properly.
>
> However, some paths seem to be defined incorrectly.
>
> When running a script "getresume.py" for example,
>
> >>try:
>    import sys
>    sys.stderr = sys.stdout
>    sys.path.insert(0,'.')
>    sys.path.append('/usr/local/lib/python1.5')
>    sys.path.append('/usr/local/lib/python1.5/site-packages')
>    import os
>    TEMPLATES=os.environ['DOCUMENT_ROOT']
>    from calendar import month_name
>    from types import *
>    from formhandler import FormHandler
>    from string import strip, split, join, atol, atof, upper, lower,
> capitalize
>    from refer import getReferer, addGetResume
>    from time import asctime, localtime, time, strftime
>    from urllib import quote_plus, unquote_plus
>    from fcsetup import LOCALHOSTNAME
>    from fcsetup import *
>    import Mail
> except:
>    print '<pre>\n'
>    import traceback
>    traceback.print_exc()
>    print sys.path
>
> resumeDIR = DEPOT

just guessing here, but it looks like

1) you need to the DOCUMENT_ROOT environment variable
before running the script

2) the fcsetup.py module should define a variable called DEPOT,
but doesn't.

> I think DEPOT might be a reference to the MySQL database, using the
> Berkley dbmodule (which I can't find). I certainly can't find a
> directory named DEPOT anywhere on the old host's server.

DEPT is a variable name, not a directory name.  the "from
fcsetup import *" line imports all variables from the fcsetup
module.

> Neither can I fathom where "DOCUMENT_ROOT" is defined, or find a file
> called "os.environ."

"os.environ" refers to a variable in the standard "os" module,
not a file.  see the library reference for more info.

</F>




From claird at starbase.neosoft.com  Thu Apr 20 09:37:15 2000
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 20 Apr 2000 08:37:15 -0500
Subject: Dynamic web pages! not Dynamic HTML
References: <200004192223.IAA10606@envy.fulcrum.com.au>
Message-ID: <3E0EB84BC85217A9.75297C30B305575F.DF96EDF011E52E7D@lp.airnews.net>

In article <200004192223.IAA10606 at envy.fulcrum.com.au>,
Richard Jones <Richard.Jones at fulcrum.com.au> wrote:
			.
			.
			.
>   Phil has spent quite a lot of time developing that site (Tcl over a 
>relational database) - and you should expect to spend some time on it also. 
>That's assuming he won't just give you his code :)
As Brent Fulghum has already mentioned, PhilG
gives it all away.  All the software is freely
available.

Phil certainly has invested a LOT of time in
photo.net.  On the other hand, I suspect Mr.
Hadap will be able to implement in Python the
aspects that interest him quite quickly, and
he might feel more comfortable with the result.
>
>   I would suggest that you develop a Zope site - for three reasons:
>   . it'll make using a "real" database much simpler thanks to Zope's 
>easy-as-pie SQL methods. Remember, Zope comes with gadfly bundled in the box.
>   . you'll automatically get access to a bunch of public-contributioned prior 
>art: guestbooks, news pages, FAQ handlers and so on. See the member area on 
>www.zope.org for a taste.
>   . once you're done, you can make the Zope products you generate available to 
>others so they can put up their photos too!
>
>   What you describe with the @photo tag in a "sorta HTML" page is actually 
>going to be quite simple to do in Zope. Ask around on the zope mailing lists if 
>you're stuck - they're all quite helpful people over there :)
All true, and worth repeating.
			.
			.
			.
-- 

Cameron Laird <claird at NeoSoft.com>
Business:  http://www.Phaseit.net
Personal:  http://starbase.neosoft.com/~claird/home.html


From scarblac-spamtrap at pino.selwerd.nl  Tue Apr 11 20:14:26 2000
From: scarblac-spamtrap at pino.selwerd.nl (Remco Gerlich)
Date: 12 Apr 2000 00:14:26 GMT
Subject: I need Help
References: <38F3AD34.E6C88698@home.com>
Message-ID: <slrn8f7j8e.sfl.scarblac-spamtrap@flits104-37.flits.rug.nl>

HC2k wrote in comp.lang.python:
> Hello I am new to programming and someone said start with Python it's a
> great learning language and can be very useful what excactly does python
> program? And how long did it take anyone else to learn thanks !

Python can program a lot of things. It's very good at controlling Windows,
it works on a lot of other platforms, and it's a good language for the Web.
But that shouldn't matter to a beginner. Python has a very straight forward
design, and a consistent structure, so it's good to learn to program with.

There are several tutorials for beginner programmers on the web, see
http://www.python.org/doc/Intros.html (especially the What is Python? link
and all the links after 'Introductions to Python for non-programmers').

Try to work through those tutorials. If you have problems, just ask on the
newsgroup, or mail me personally, or subscribe to the python-tutors mailing
list, a list especially for people learning Python as their first language -
see http://www.python.org/mailman/listinfo/tutor .

-- 
Remco Gerlich,  scarblac at pino.selwerd.nl
  3:10am  up 36 days, 14:30,  8 users,  load average: 3.27, 2.75, 1.98


From amitp at Xenon.Stanford.EDU  Thu Apr 13 17:11:14 2000
From: amitp at Xenon.Stanford.EDU (Amit Patel)
Date: 13 Apr 2000 21:11:14 GMT
Subject: simple httplib and urllib timeout question
References: <CgJI4.14892$eh.1040983@news.bc.tac.net>
Message-ID: <8d5d5i$95k$1@nntp.Stanford.EDU>


  <darrell at rubina.funkomite.com> wrote:

| Just a quick question, im writing a small webbot for a python
| learning experience and it comes across some sites that will make it
| hang for a while, if the site is down or extremely slow etc, is
| there a way to set a faster timeout for httplib . Or is there a
| better way to handle it ?


You can set an alarm using signal.alarm, and then you can throw an
exception in the signal handler.  I used to do this, but then I
discovered asyncore.  If you want to crawl multiple pages at once, you
might want to switch to asyncore.  I believe the asyncore documentation page
has an example HTTP client.

    - Amit

--
Amit J Patel, Computer Science Department, Stanford University
http://www-cs-students.stanford.edu/~amitp/



From robin at jessikat.demon.co.uk  Mon Apr  3 03:04:56 2000
From: robin at jessikat.demon.co.uk (Robin Becker)
Date: Mon, 3 Apr 2000 08:04:56 +0100
Subject: Python 1.6 alpha 1 released
References: <200003312130.QAA04361@eric.cnri.reston.va.us>
Message-ID: <dwhmLcAYKE64EwJw@jessikat.demon.co.uk>

In article <200003312130.QAA04361 at eric.cnri.reston.va.us>, Guido van
Rossum <guido at python.org> writes
>I've just released a source tarball and a Windows installer for Python
>1.6 alpha 1 to the Python website:
>
>  http://www.python.org/1.6/
>
>Probably the biggest news (if you hadn't heard the rumors) is Unicode
>support.  More news on the above webpage.
>
>Note: this is an alpha release.  Some of the code is very rough!
>Please give it a try with your favorite Python application, but don't
>trust it for production use yet.  I plan to release several more alpha
>and beta releases over the next two months, culminating in an 1.6
>final release around June first.
>
>We need your help to make the final 1.6 release as robust as possible
>-- please test this alpha release!!!
>
>--Guido van Rossum (home page: http://www.python.org/~guido/)
>
I see from the download info that python now seems to require
installation of it's own Tkinter. Can I inhibit this at install time and
continue to use my own version? I'm quite prepared for some problems of
getting it to work, but the whole point of having Tcl/Tk is that it is
extensible and I doubt very much whether an arbitrary version will work
with all the Tcl extensions I currently have. Blt for example is still
issued by the author in version specific form.
-- 
Robin Becker


From moshez at math.huji.ac.il  Sat Apr 22 00:48:19 2000
From: moshez at math.huji.ac.il (Moshe Zadka)
Date: Sat, 22 Apr 2000 07:48:19 +0300 (IDT)
Subject: The Simple Economics of Open Source
In-Reply-To: <3dya67xga4.fsf@amarok.cnri.reston.va.us>
Message-ID: <Pine.GSO.4.10.10004220745540.5248-100000@sundial>

On 21 Apr 2000, Andrew M. Kuchling wrote:

> Also, ever had to debug a problem in a binary-only library through
> trial-and-error?  You can burn lots of time that way.  <core dump>
> "Hmm... maybe that parameter can't be NULL."  <core dump> "Maybe that
> struct field should be 1." <core dump>.  I like the fact that, if I
> start seeing mysterious errors, I can dive down into the language
> interpreter, farther down into the C library, and even into the kernel
> if needed.  Though I've only followed a problem all the way into the
> Linux kernel once (to discover it was my error), it was reassuring to
> be able to do so.

Reminds me of the time I had to delve into Xt source to look for a way
to get out of the main loop without exiting the program. After 15 minutes
of looking at the docs to no use, it took 5 minutes of grepping through
the source, only to return to the docs and find out it was there -- 
only well hidden with an unlikely name and a surprising section.

x-sources-are-strange-and-wonderful-ly y'rs, Z.
--
Moshe Zadka <mzadka at geocities.com>. 
http://www.oreilly.com/news/prescod_0300.html
http://www.linux.org.il -- we put the penguin in .com




From amitp at Xenon.Stanford.EDU  Tue Apr  4 22:55:08 2000
From: amitp at Xenon.Stanford.EDU (Amit Patel)
Date: 5 Apr 2000 02:55:08 GMT
Subject: [Bug?] Time utils lose DST information
Message-ID: <8ce9uc$hoi$1@nntp.Stanford.EDU>

First, my program:

from time import *

print strftime("%a %b %d %H:%M:%S %Y", 
               localtime(mktime(strptime("Mon Apr 3 05:06:07 2000"))))

I'd like this to return "Mon Apr 3 05:06:07 2000".

However, it doesn't.  It returns "Mon Apr 3 06:06:07 2000".

I found this odd, so I looked at the Python 1.6a1 source... strptime
leaves the daylight savings (DST) flag as 0 (known to NOT be DST)
instead of -1 (unknown status of DST).  (strptime (in timemodule.c)
fills in the structure with 0's, and C's strptime doesn't set this
flag unless you use a %Z in the time format.)

Given that strptime doesn't know the DST status, I think it should
initialize the `tm' structure to contain -1 in the DST field,
indicating "unknown", so that converting to and back from Unix time
doesn't give the wrong number.

Any thoughts on this?


   -i'd-rather-absolish-daylight-savings-completely-y'rs, Amit


From pinard at iro.umontreal.ca  Mon Apr  3 20:30:28 2000
From: pinard at iro.umontreal.ca (=?ISO-8859-1?Q?Fran=E7ois_Pinard?=)
Date: 03 Apr 2000 20:30:28 -0400
Subject: Unicode program representation
In-Reply-To: Moshe Zadka's message of "Tue, 4 Apr 2000 00:15:53 +0200 (IST)"
References: <Pine.GSO.4.10.10004040011450.21084-100000@sundial>
Message-ID: <oq1z4m1zdn.fsf@titan.progiciels-bpi.ca>

Moshe Zadka <moshez at math.huji.ac.il> writes:

> FWIW, I believe that Python's (official) parser should be strictly UTF-8.

I would be very unhappy if I was loosing the capability of editing my files
directly in Latin-1.  I might come to use UTF-8 all over one of these days,
or maybe not.  Let it be my choice.

> it's trivial to write a script to convert between anything and UTF-8
> (in Python, of course <wink>).

Granted.  But I do not feel like being forced into using such a script.

It is good that Python is able to process Unicode.  It would be wrong if it
was starting to impose Unicode on users, in so sharing the fanatic attitude
some people have about Unicode.  I like what Python 1.6 does currently, that
is, leaving me the choice about where is my own comfort while using Python.

-- 
Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard





From culliton at clark.net  Sat Apr  1 14:00:28 2000
From: culliton at clark.net (Tom Culliton)
Date: Sat, 01 Apr 2000 19:00:28 GMT
Subject: New Features in Python 1.6
References: <200004011740.MAA04675@eric.cnri.reston.va.us>
Message-ID: <gHrF4.16072$0o4.101062@iad-read.news.verio.net>

OK, it's April 1st so maybe I'm just not getting the joke but...

In article <200004011740.MAA04675 at eric.cnri.reston.va.us>,
Guido van Rossum  <guido at python.org> wrote:
>New Features in Python 1.6
>==========================

<snip>

>Note that use of any string method on a particular string renders it
>mutable.  This is for consistency with lists, which are mutable and have
>methods like 'append()' and 'sort()' that modify the list.  Thus,
>"foo.strip()" modifies the string 'foo' in-place.  "strip(foo)" retains
>its old behavior of returning a modified copy of 'foo'.

Violating immutablity like this seem like a misfeature and potential
source of nasty bugs unless there is some kind of copy on write going
on.  For example, imagine A is set to a string value, which is used as
a dictioanry key, and then passed to a fucntion, which passes it to
another function, which modifies it in place.  Suddenly the dictionary
keys aren't unique since the same value was shared by all.  Just the
thought scares me snotless.

>4. assignment to None now works

Huhn?  What exactly changed here?  You can assign to None in current
versions just like you can to any other name.  I thought the push was
to _prevent_ this.

Python 1.5.2 (#1, Sep 17 1999, 20:15:36)  [GCC egcs-2.91.66 19990314/Linux (egcs- on linux-i386
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> def xyz(a):
...     global None
...     None = a
...
>>> def abc(a):
...     if a == None:
...             print "None-sense!"
...
>>> xyz(42)
>>> abc(42)
None-sense!

>In previous version of Python, values assigned to None were lost.  For
>example, this code:
>
>    (username,None,None,None,realname,homedir,None) = getpwuid(uid)
>
>would only preserve the user name, real name, and home directory fields
>from a password file entry -- everything else of interest was lost.
>
>In Python 1.6, you can meaningfully assign to None.  In the above
>example, None would be replaced by a tuple containing the four values of
>interest.

This is either nasty unpythonic special case magic or a much broader
change.  The current behaviour is the same whether you use the name
"None", "X", "y", "junk", ...  The name ends up refering to the last
value in the list.  The whole point of this type of assignment to None
(or junk, or ...) is usually to discard elements you don't want.  As a
general change it could be useful but may break existing code in a way
thats VERY hard to spot.

>You can also use the variable argument list syntax here, for example:
>
>    (username,password,uid,uid,*None) = getpwuid(uid)
>
>would set None to a tuple containing the last three elements of the
>tuple returned by getpwuid.

I'm not one of the folks who howl at tightening up the behaviour of
list.append, socket.connect and the like but these changes, especially
viloating immutability, strike me as too bizarre and dangerous for
words.  Please tell us it's just and April Fools joke!



From lewst at yahoo.com  Tue Apr  4 00:09:41 2000
From: lewst at yahoo.com (lewst)
Date: Mon, 3 Apr 2000 21:09:41 -0700 (PDT)
Subject: Starving for an Advanced Python Book
Message-ID: <20000404040941.7766.qmail@web2106.mail.yahoo.com>

Having finished "Learning Python", I'm now looking for an advanced
Python text.  ORA's "Programming Python" is a natural choice, but I
can't bear buying this outdated book when the 2nd edition is due out
in a few months (I'm dying here Lutz, hurry it up already!).  However,
I really need something in the interim period.  The library reference
manual is fine but usage is difficult for me to figure out sometimes
because there are very few examples included.

Does anyone have any suggestions?  I suppose I'm looking for a text
that has examples of how to solve various kinds of programming
problems in Python and also examples of code showing how the library
and various modules actually work.  How is the "Python Essential
Reference" on examples?  Is it just a rebinding of the python.org
library reference

Thank-you.


__________________________________________________
Do You Yahoo!?
Talk to your friends online with Yahoo! Messenger.
http://im.yahoo.com



From ullrich at math.okstate.edu  Sat Apr  1 13:39:50 2000
From: ullrich at math.okstate.edu (David C. Ullrich)
Date: Sat, 01 Apr 2000 12:39:50 -0600
Subject: Function override?
References: <000101bf9b9b$312f8fe0$2f2d153f@tim>
Message-ID: <38E64275.CF8F97DC@math.okstate.edu>

Tim Peters wrote:

> [David C. Ullrich]
> > ...
> > This is certainly one of the things that _I_ find keen,
> > amyway - I hope the people who are talking about adding
> > stricter typing to Python make it optional somehow.
>
> Yup, not to worry.  All proposals to date make static typing declarations
> optional, and by default not needed.

    Fine, I can sleep tonight.

    Or I could if I hadn't just read the bit about how in 1.6 strings
have a split method, and if s1 and s2 are strings then

s1.split(s2)

either splits s1 using s2 as the delimiter or the other way around,
depending on which occurs more frequently in the other. I mean
surely it will happen that we want to split a bunch of strings
using 'delimiter' as a delimiter, and then it just happens that
the routine that does that gets passed 'lim' as an argument and
does the opposite of what we want?

> [...]
>
> > There are obvious advantages to stricter typing but you lose
> > something as well. (For example, I get a big kick out of
> > the fact that exactly the  same euclidean-algorithm code
> > works to find the GCD of two integers or the GCD of
> > two polynomials. Etc.)
>
> This isn't a problem with static typing per se, provided the system is rich
> enough to let you define a suitable EuclideanDomain type.  The system is not
> rich enough today,

    Not all that important, but I'm missing something here. Why can't
we make a EuclideanDomain type? I've made a VectorSpace type,
for example.
    Oh. You meant a EuclideanDomain type instances of which
would be euclidean domains? In the sort of thing I've done
a _subclass_ of EuclideanDomain would be a euclidean domain,
instances of which are the elements.

> but people throughout this industry hold a charming faith
> that tomorrow's solutions won't suffer today's problems <wink>.
>
> just-a-fan-of-the-illusion-of-progress-ly y'rs  - tim



From scarblac-spamtrap at pino.selwerd.nl  Thu Apr 27 21:14:46 2000
From: scarblac-spamtrap at pino.selwerd.nl (Remco Gerlich)
Date: 28 Apr 2000 01:14:46 GMT
Subject: Python does not let me format time.
References: <8eakhq$797$1@nnrp1.deja.com>
Message-ID: <slrn8ghpeo.f01.scarblac-spamtrap@flits104-37.flits.rug.nl>

Sindh wrote in comp.lang.python:
> If there is already a moduls to do the same , much helpful.

>>> import time
>>> time.strptime("00/04/28","%y/%m/%d")
(2000,04,17,0,0,0,6,1,0)

strptime is the inverse of strftime, in a way.

-- 
Remco Gerlich,  scarblac at pino.selwerd.nl

   This is no way to be
     Man ought to be free      -- Ted Bundy
       That man should be me


From michael.stroeder at inka.de  Tue Apr 18 02:39:05 2000
From: michael.stroeder at inka.de (Michael =?iso-8859-1?Q?Str=F6der?=)
Date: Tue, 18 Apr 2000 08:39:05 +0200
Subject: May 27th PARTY!
References: <006201bfa756$896e5a80$da00a8c0@austin.rr.com>
Message-ID: <38FC0309.972C4781@inka.de>

> Rebecca Stidolph wrote:
> 
> ATTENTION FRIENDS AND FAMILY AROUND THE NATION!!
> 
> We (David and Rebecca Stidolph) are celebrating our 17th wedding
> anniversary

Wow! Why going to a Python conference when we have a chance to meet
at a wedding anniversary.

Ciao, Michael.


From eblosch at earthlink.net  Sun Apr 30 22:50:34 2000
From: eblosch at earthlink.net (Ed Blosch)
Date: Mon, 01 May 2000 02:50:34 GMT
Subject: CGI File Size Limit?
Message-ID: <_h6P4.14007$g4.347236@newsread2.prod.itd.earthlink.net>

I'm having trouble uploading an 18K file via CGI and saving it to a file from my Python script.  The file gets saved, but the file size is only 1k.  I'm assuming that the cgi module is only reading the first 1k bytes.  I'd like to know if this sounds reasonable and how to solve the problem....and if not reasonable, where else I might look for the problem.

Thanks....
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20000501/9f82d615/attachment.html>

From scarblac-spamtrap at pino.selwerd.nl  Thu Apr 27 09:53:03 2000
From: scarblac-spamtrap at pino.selwerd.nl (Remco Gerlich)
Date: 27 Apr 2000 13:53:03 GMT
Subject: Concatening string and integer
References: <8e9gco$jlt$1@vg170.it.volvo.se>
Message-ID: <slrn8gghgk.el7.scarblac-spamtrap@flits104-37.flits.rug.nl>

JJ wrote in comp.lang.python:
> Hi!
> 
> I want to concatenate a string containing both strings and integers like:
> 
> strTemp = 'I make test number: ' + iMyNumber + ' once again'
> 
> It doesn't work, how should I do this?

1) You make a string of the integer:

strTemp = 'I make test number: ' + str(iMyNumber) + ' once again'

2) You use string formatting:

strTemp = 'I make test number: %d once again' % iMyNumber

-- 
Remco Gerlich,  scarblac at pino.selwerd.nl
Hi! I'm a .sig virus! Join the fun and copy me into yours!


From adjih at crepuscule.com  Fri Apr 28 11:23:18 2000
From: adjih at crepuscule.com (Cedric Adjih)
Date: 28 Apr 2000 15:23:18 GMT
Subject: MORE INFO
References: <8ea0ao$fro$1@nnrp1.deja.com> <8ea9uu$rh0$1@nnrp1.deja.com>
Message-ID: <8ecad6$bun$1@ites.inria.fr>

mdefreitas at sikorsky.com wrote:
> By the way... I finally tested the code I wrote and it evidently is
> incorrect. It works fine on the first nest level, but my application
> aborts on the second nest level.

> The error message is:
> Fatal Python Error: PyThreadState_Get: no current thread

> OK... I'm lost. Any ideas? Anybody? Buelher?

  Hmm... I can't answer because I'm not a Python threading expert,
but I didn't understand your thread model: do you want to run
several python scripts (and several ui.commands C code)
concurently (how would you synchronize?) ?
  Note that if all you want is to have C code that calls Python
code that calls C code (through a module) that calls Python code
without threads, PyRun_XXX functions are re-entrant
(these functions are even called reentrantly by the Python interpreter
itself for the "eval", "execfile" functions, check Python/bltinmodule.c).

If you want to pass arguments to your script, you could use the
other PyRun_File function with a dictionnary with args for locals, see
  http://www.python.org/doc/current/api/veryhigh.html
(the 'int start' should be Py_file_input)


Note that a more flexible way to implement a 'python'
command is:
- to initialize the python interpreter
- to load a fixed module ("pythonexec.py" for instance)
(PyImport_ImportModule)
- to call a function with a given name from this module
(PyModule_GetDict + PyDict_GetItemString + PyObject_CallFunction 
should do, maybe there is something shorter), with as
argument the whole command string.

I.e. do the equivalent of: 
import pythonexec
pythonexec.doEval(<the python command string>)

Then you can start with 'pythonexec.py' being:
import string
def doEval(cmdLine):
  arg=string.split(cmdLine)
  execfile(arg[0],{'argv': cmdLine[1:]})


And then later change doEval to do more elaborate
parsing (for instance if an argument begins with '$',
first eval the argument as an ui command, or if
the filename is 'eval' just evaluate the string,
or don't do execfile, but an import+apply, ...)

-- Cedric



From thomas at xs4all.net  Thu Apr  6 12:35:09 2000
From: thomas at xs4all.net (Thomas Wouters)
Date: Thu, 6 Apr 2000 18:35:09 +0200
Subject: __builtin__.open on windows NT
In-Reply-To: <20000406182802.C13830@xs4all.nl>; from thomas@xs4all.net on Thu, Apr 06, 2000 at 06:28:02PM +0200
References: <8cicd6$1s0$1@nnrp1.deja.com> <20000406182802.C13830@xs4all.nl>
Message-ID: <20000406183509.D13830@xs4all.nl>

On Thu, Apr 06, 2000 at 06:28:02PM +0200, Thomas Wouters wrote:
> If you *really* want to use __builtins__, you can: you just have to import
> it first ! :-) Remember to do that with 'import __builtins__', or it might

Err, it's 'import __builtin__'. Sorry for that (consistent) typo. (Cant
claim it's a thinko because I dont recall thinking about it ;)

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From yawyi at gmx.de  Tue Apr  4 12:37:34 2000
From: yawyi at gmx.de (Johannes Stezenbach)
Date: Tue, 4 Apr 2000 18:37:34 +0200
Subject: Python 1.6 alpha 1 released
References: <Pine.GSO.4.10.10004040035420.21084-100000@sundial> <5E877FD7637AB46E.DABBED9CF3C1D64B.CDDDC05AE7374A78@lp.airnews.net> <5l7leern5v.fsf@eric.cnri.reston.va.us>
Message-ID: <38ea1de0$0$17688@personalnews.de.uu.net>

Guido van Rossum <guido at cnri.reston.va.us> wrote:
>I don't want to break that much code, so I'll revert the decision in
>the
>next alpha release.
>
>However, I'll do so for purely backwards compatible reasons, and I
>will continue to press for the tuple address as the correct form.

Will python then support a -w switch to support people in writing
correct programs (i.e. compatible with likely furure language/library
changes), just like our-favourite-four-letter-P-language?

Seriously: Either accept the 2-arg form of connect() as *correct* for
all future versions of python or keep the change to make it illegal
*now*.
Just postponing the change is evil since it will only hurt more people
in the future, as the number of python users is growing. (CP4E?)

>But I will fix the code -- it's the only practical thing to do.

If it's not broken then don't fix it.



From mikael at isy.liu.se  Wed Apr 26 05:32:07 2000
From: mikael at isy.liu.se (Mikael Olofsson)
Date: Wed, 26 Apr 2000 11:32:07 +0200 (MET DST)
Subject: huh? Python says that AttributeError: 'string' object has no
In-Reply-To: <8e62qm$511$1@nnrp1.deja.com>
Message-ID: <XFMail.000426113207.mikael@isy.liu.se>

On 26-Apr-00 schmitthead at my-deja.com wrote:
 >  I received a whole folder full of files with MacOS type text files, with
 >  different newline characters.  I thought, a perfect job for python.
 >  Below is my attempt, but it generates the following error message:
 >  
 > >mac2dos tt_prvt.h
 >  Traceback (innermost last):
 >    File "C:\bin\mac2dos.py", line 33, in ?
 >      mac2dos( directory )
 >    File "C:\bin\mac2dos.py", line 24, in mac2dos
 >      os.path.walk( os.getcwd(), callbackfunc, '' )
 >    File "d:\python\Lib\ntpath.py", line 265, in walk
 >      func(arg, top, names)
 >    File "C:\bin\mac2dos.py", line 18, in callbackfunc
 >      mac2dosfile( file )
 >    File "C:\bin\mac2dos.py", line 9, in mac2dosfile
 >      lines = myline.split( myline, chr(10) )
 >  AttributeError: 'string' object has no attribute 'split'

The line should be

    lines = string.split( myline, chr(10) )
            ^^^^^^

split is a function in the module string.

/Mikael

-----------------------------------------------------------------------
E-Mail:  Mikael Olofsson <mikael at isy.liu.se>
WWW:     http://www.dtr.isy.liu.se/dtr/staff/mikael
Phone:   +46 - (0)13 - 28 1343
Telefax: +46 - (0)13 - 28 1339
Date:    26-Apr-00
Time:    11:29:45

This message was sent by XF-Mail.
-----------------------------------------------------------------------


From zorro at zipzap.ch  Wed Apr  5 05:45:52 2000
From: zorro at zipzap.ch (Boris Borcic)
Date: Wed, 05 Apr 2000 11:45:52 +0200
Subject: ANN: Stackless Python for MacOS
References: <l03102815b5101e19e027@[193.78.237.148]>
Message-ID: <38EB0B50.880ED056@zipzap.ch>

Just van Rossum wrote:
> 
> Even though I still hardly understand the implications of it all yet,
> Stackless is IHMO the most exciting new Python development at the time,
> together with Unicode support in 1.6. Stackless will most likely not be
> incorporated into Python 1.6, but *maybe* if we make enough noise we can
> convince Guido it should go into 1.7... More stackless killer apps please!

I'd argue for "Python 1.618" - the golden ratio.

BB
--
"C++ is a contradiction in terms" -- Einstein


From kaeber at gmd.de  Tue Apr 11 05:06:47 2000
From: kaeber at gmd.de (Joachim Kaeber)
Date: Tue, 11 Apr 2000 11:06:47 +0200
Subject: readline skips on line
References: <8culp8$23a$1@porthos.nl.uu.net>
Message-ID: <38F2EB27.751EFFDD@gmd.de>

Hi,

try this:

f=open('test.txt', 'r')

while 1:
  line=f.readline()
  if not line:
    break
  print line

print "ready"

(you program uses odd lines for the test, even lines for output...)

Soundwave wrote:
> 
> I got a part of a programm looking like this
> 
> f=open('test.txt', 'r')
> while f.readline()!='':
>     print f.readline()
> print 'ready'
> 
> but when it is runned, it wil print out the lines in the text file, but it
> skips all the the odd line, it only displays line 2,4,6,8,10 and further
> 
> How can I solve this problem ?

-- 
Joachim Kaeber (kaeber at gmd.de) (http://imk.gmd.de/~kaeber)
GMD - Forschungszentrum Informationstechnik GmbH
    - German National Research Center for Information Technology
Schloss Birlinghoven                    email: kaeber at gmd.de
D-53757 Sankt Augustin 1, Germany       phone: (+49 2241) 14-2546


From cut_me_out at hotmail.com  Thu Apr 20 19:27:41 2000
From: cut_me_out at hotmail.com (Alex)
Date: 20 Apr 2000 19:27:41 -0400
Subject: Where do I send a contribution for the Starship recovery ?
References: <gNqL4.142052$ra.3209287@news.easynews.com> <tlLL4.146289$ra.3293534@news.easynews.com>
Message-ID: <etdr9c0jqtu.fsf@nerd-xing.mit.edu>

> Sorry about the repeats. 

No worries.  I was just trying to be funny.

Alex.


From jbenderl at olf.com  Fri Apr  7 10:14:27 2000
From: jbenderl at olf.com (Jason Benderly)
Date: Fri, 07 Apr 2000 10:14:27 -0400
Subject: PPython
References: <38ede548.353776233@news.oh.verio.com> <8ckptt$j8s$1@pegasus.csx.cam.ac.uk>
Message-ID: <38EDED43.78A97864@olf.com>

Hi Nick,

Could yo be more specific about those things that you wouldn't want to know.
I want to know...

Jason

Nick Maclaren wrote:

> If you want a really silly idea, it probably wouldn't be hard to
> convert Python's C source to Perl, because the former looks pretty
> clean and therefore could be translated semi-automatically without
> needing a full-blown C to Perl compiler.  That's the easy bit.
>
> When you put 130,000 lines of translated source into Perl, you
> would discover things about Perl that you didn't want to know and
> even its author had not suspected :-)
>
> And I suspect that those Web sites would disembowel you after their
> systems had melted down ....
>
> Regards,
> Nick Maclaren,
> University of Cambridge Computing Service,
> New Museums Site, Pembroke Street, Cambridge CB2 3QG, England.
> Email:  nmm1 at cam.ac.uk
> Tel.:  +44 1223 334761    Fax:  +44 1223 334679



From ivanlan at callware.com  Tue Apr 18 17:34:51 2000
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Tue, 18 Apr 2000 15:34:51 -0600
Subject: PythonWin and Dialogs in DLLs
Message-ID: <38FCD4FB.6E07A632@callware.com>

Hi All--
OK, I followed the instructions in previous messages to this list to
build a resource-only DLL from which I wish to import custom dialogs. 
The build works, but using "import migrate1", where the dll is named
"migrate1.dll" gives me the error message

	dynamic module does not define init function (initmigrate1)

Digging around for that error message seems to indicate that it's
thinking I want an extension module with code in it, when all I want to
do is just have *resources only*.

What am I missing?  What's staring me in the face if only I would open
my eyes?

<too-crabby-for-funny-tag-lines-today>-ly y'rs,
Ivan
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
http://www.pauahtun.org 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
Author:  Teach Yourself Python in 24 Hours



From mhammond at skippinet.com.au  Fri Apr 28 19:31:44 2000
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Fri, 28 Apr 2000 23:31:44 GMT
Subject: win32com: _NewEnum being cached
References: <8ecta7$nn5$1@nnrp1.deja.com>
Message-ID: <AbpO4.6824$v85.40107@news-server.bigpond.net.au>

"Tom Hines" <tom_hines at my-deja.com> wrote in message
news:8ecta7$nn5$1 at nnrp1.deja.com...
> Hello.  As you know, to implement "for x in my_list" functionality in
> COM, you must code a get__NewEnum() function, which I have done and it
> works fine under Visual Basic.  But when called from Python, the call
> seems to get cached, as subsequent Python "for" blocks don't result in
> calls to get__NewEnum().

You should be able to say:

enum = ob._NewEnum()
for i in enum:
  ...

Mark.





From rhicks at nospam.rma.edu  Tue Apr 18 20:59:18 2000
From: rhicks at nospam.rma.edu (Robert Hicks)
Date: Wed, 19 Apr 2000 00:59:18 GMT
Subject: o'rielly python devcenter
References: <Pine.BSF.4.10.10004181241280.15729-100000@chi.pair.com> <38FCF25E.4D945D81@oreilly.com>
Message-ID: <Gx7L4.176$Pa.3485@news>

Do I write an article and then just email it to you?

Bob

"Stephen R. Figgins" <fig at oreilly.com> wrote in message
news:38FCF25E.4D945D81 at oreilly.com...
> It isn't old news, it just went public last night(Monday) Sort of a soft
> launch, as I don't have much in the way of content up yet.  But I am
> working on that problem!  I will have some original articles in a couple
> of weeks.
>
> The Network is actually a subsidiary of O'Reilly & Associates, and I am
> plenty happy to plug other's books.  (Actually I think the only book I
> have linked right now is the venerable Programming Python, and that is
> just because I mentioned it in me introduction.)
>
> The links I am building under topics are not links to other sites, but
> rather to documents on other sites.  I am looking for tutorials,
> articles, documentation, anything good and juicy with information.  I am
> open to suggestions.
>
> And I like criticism, particularly the constructive kind.  So send me
> plenty.
>
> And if any of you want to line up for writing some good articles for me,
> I am paying!
>
> Stephen R. Figgins
> Editor, O'Reilly Network




From lars.lundstedt at telia.com  Tue Apr 25 13:21:21 2000
From: lars.lundstedt at telia.com (Lars Lundstedt)
Date: Tue, 25 Apr 2000 17:21:21 GMT
Subject: Don't understand error message
Message-ID: <lukN4.1574$wYl.191480320@newsb.telia.net>

I'm halfway through "Learning Python" and I'm stuck on the example on page
255.

It goes like this:

import sys, glob, operator
print sys.argv[1:]
sys.argv = reduce(operator.add, map(glob.glob, sys.argv))
print sys.argv[1:]

When called with "python glob.py *.py" on Win98 I'm supposed to get a list
of all .py files in the current directory but instead I get:

Traceback (innermost last):
  File "glob.py", line 1, in ?
    import sys, glob, operator
  File "glob.py", line 3, in ?
    list = reduce(operator.add, map(glob.glob, '*.py'))
TypeError: call of non-function (type module)

Anyone know why?

/LasseLu





From talon at coursen.ifu.net  Sat Apr 15 12:55:13 2000
From: talon at coursen.ifu.net (Stephen Coursen)
Date: Sat, 15 Apr 2000 16:55:13 GMT
Subject: (no subject)
References: <001f01bfa6f8$b05dc9e0$69ee85ce@default>
Message-ID: <slrn8fh7pb.9cf.talon@coursen.ifu.net>

On Sat, 15 Apr 2000 09:35:41 -0700, Pepper-Emporium <jcmllc at earthlink.net> 
wrote:
>This is a multi-part message in MIME format.
>
>------=_NextPart_000_001C_01BFA6BD.F6F75FA0
>Content-Type: text/plain;
>	charset="iso-8859-1"
>Content-Transfer-Encoding: quoted-printable
>
>PLEASE HELP:
>
>I'm getting irate e-mails from people on my list indicating that when =
>ever they send a message is bounces back to them instead of me. My list =
>is called "hotlist" and I have about 27,000 members.
>
>I want to REMOVE ALL the members and start with a clean slate.=20
>
>Whenever I try to login to MAILMAN I get the following error message:
>Bug in Mailman version 2.0beta2
>
>We're sorry, we hit a bug!
>If you would like to help us identify the problem, please email a copy =
>of this page to the webmaster for this site with a description of what =
>happened. Thanks!=20
>
>Traceback:
>
>Traceback (innermost last):
>  File "/usr/local/mailman/scripts/driver", line 89, in run_main
>    main()
>  File "/usr/local/mailman/Mailman/Cgi/admin.py", line 68, in main
>    mlist =3D MailList.MailList(listname)
>  File "/usr/local/mailman/Mailman/MailList.py", line 69, in __init__
>    self.Load()
>  File "/usr/local/mailman/Mailman/MailList.py", line 858, in Load
>    self.Lock()
>  File "/usr/local/mailman/Mailman/MailList.py", line 1300, in Lock
>    self.__lock.lock()
>  File "/usr/local/mailman/Mailman/LockFile.py", line 209, in lock
>    os.link(self.__lockfile, self.__tmpfname)
>OSError: [Errno 28] No space left on device
>
>
I'm gonna go out on a limb here and say that there's no space left on the
device (see the error message says so :) ).  Most likey, what every drive
mailman is using is filled up and you need to make some room (or move the
mailman installation to a different drive).

That would be the first thing to address and is probably the root of the
problem.

HTH,
Steve

[the rest snipped]


From moshez at math.huji.ac.il  Thu Apr 20 09:34:07 2000
From: moshez at math.huji.ac.il (Moshe Zadka)
Date: Thu, 20 Apr 2000 16:34:07 +0300 (IDT)
Subject: Confused by List behavior
In-Reply-To: <8dmqsd$6m0$1@nnrp1.deja.com>
Message-ID: <Pine.GSO.4.10.10004201633140.18417-100000@sundial>

On Thu, 20 Apr 2000 ejbaker_seekonk at my-deja.com wrote:

> As a python newbie, I entered this script to explore list processing
> behaviours.
> 
> yy = dir()
> for element in yy:
>   print element
> 
> for element in yy:
>   print dir(element)
> 
> The first 'for' loop works as expected. It returns,
> ['__builtins__', '__doc__', '__name__'].
> 
> The second loop, however, returns three empty lists. For the life of
> me, I can't understand that. Becauese, if I type 'dir(__builtins__)',
> in a script, I get all of the elements of __builtins__.
> 
> Can anyone explain this?

Hint: what happens if you type dir('__builtins__')?

thank-god-this-isn't-tcl-ly y'rs, Z.
--
Moshe Zadka <mzadka at geocities.com>. 
http://www.oreilly.com/news/prescod_0300.html
http://www.linux.org.il -- we put the penguin in .com




From johngrayson at home.com  Mon Apr 24 06:35:09 2000
From: johngrayson at home.com (johngrayson at home.com)
Date: Mon, 24 Apr 2000 10:35:09 GMT
Subject: Tkinter.Button error
References: <3903E8C1.39B3E3D7@swt.edu>
Message-ID: <8e180t$46d$1@nnrp1.deja.com>

In article <3903E8C1.39B3E3D7 at swt.edu>,
  ABC <ns566645 at swt.edu> wrote:
> I have two frames
> top frame and bottom frame.
> In the top frame I implemented a group.
> Now, when I try to add Tkinter.Buttomn widget in the group in the top
> frame it gives error . Can anyone help me and tell me what the error
is?
>
> Here is the code
>
> frame1=Frame(master, relief=sunken', borderwidth='2')
> frame2=Frame(master, relief='sunken', borderwidth='1')
>
> frame1.place(relx=0, rely=0, relwidth=1, relheight=0.50)
> frame2.place(relx=0. rely=0.50, relwidth=1, relheight=0.50)
>
> group1=Pmw.Group(frame1, tag_text="Hello")
> group1.pack(fill='both')
>
> button=Tkinter.Button(group1.interior(), text='8.00')
> button.pack()
>
> The above program gives an error in TKinter in the second last line.
> I have no clue of what the error is? Can anyone help me?
> Need futher information, please let me know.
>
>

You probably have:

from Tkinter import *

Since your Frame constructors do not use Tkinter.Frame,
I'd guess:

button=Button(group1.interior(), text='8.00')

will work.

    John


Sent via Deja.com http://www.deja.com/
Before you buy.


From M_Pope at shef.rage.co.uk  Mon Apr 10 11:19:42 2000
From: M_Pope at shef.rage.co.uk (Mark Pope)
Date: Mon, 10 Apr 2000 16:19:42 +0100
Subject: Extending built-in language (Win 95)
Message-ID: <213CDEBFDDA7D3118081009027E40C56071118@SHEF-MS1>

> From: Martin von Loewis [mailto:loewis at informatik.hu-berlin.de]
> 
> Mark Pope <M_Pope at shef.rage.co.uk> writes:
> 
> > I would like to extend the objects available in python, 
> however It is not
> > appropriate to build normal extension dll's.
> 
> Why is it not appropriate? It works perfectly well to add new types in
> an extension module. Many extensions do.

Ah, didn't realise that modules built into the LIB. (I'm new to python)

> The new type becomes only available by means of creation of
> values. Look at the file type. Do you write
> 
> f = file('foo.txt','w')
> 
> ? No, you use open() instead, which just returns an instance of a file
> object. In the xxmodule, xx_new is used to make new instances, which
> can be called from Python as xx.new().

I've added xxmodule to the lib and recompiled. Entering:
a = xx.new()

just gives a name error on xx.

Have I missed something?


(thanks for helping)

mark.



From jbauer at rubic.com  Sat Apr  8 19:53:51 2000
From: jbauer at rubic.com (Jeff Bauer)
Date: Sat, 08 Apr 2000 18:53:51 -0500
Subject: ANN: Rio MP3 player interface
Message-ID: <38EFC68F.48B2449B@rubic.com>

I've written a Python interface to the Diamond Rio
MP3 player.  It works on Windows 95, 98, NT and
probably 2000.  The interface only works with the
parallel port versions of the Rio at present, but
will (probably) soon support the USB versions.

ftp://starship.python.net/pub/crew/jbauer/py_rio_latest.zip

I'll put up some web pages soon.  Any feedback 
is welcome.

The code is based on the work of the Snowblind
Alliance: http://www.world.co.uk/sba/rio.htm

If you own a portable MP3 player, you might want
to check them out, if for no other reason than
they appear to be irritating the Diamond droids:

http://www.wired.com/news/culture/0,1284,17529,00.html

"We're not happy," said Ken Wirt, vice president of
corporate marketing for Diamond Multimedia Systems.


Jeff Bauer
Rubicon Research



From gresham at mediavisual.com  Wed Apr 12 10:42:01 2000
From: gresham at mediavisual.com (Paul Gresham)
Date: Wed, 12 Apr 2000 22:42:01 +0800
Subject: XML-RPC and None objects
References: <m12fNMH-000CnCC@artcom0.artcom-gmbh.de>
Message-ID: <8d21vt$hut1@imsp212.netvigator.com>

Excuse my ignorance Peter, but is this the package I should look at if I
want
to serve a Wap application?
Regards

"Peter Funk" <pf at artcom-gmbh.de> wrote in message
news:m12fNMH-000CnCC at artcom0.artcom-gmbh.de...
> Hi!
>
> XML-RPC <http://www.xmlrpc.com> is a very nice vehicle to build
> distributed server/client systems.  Fredrik Lundhs xmlrpc package
> available at <ttp://www.pythonware.com/products/xmlrpc/> provides
> everything you need to build clients and servers in Python.
>
> But the current XML-RPC specification has a serious problem, which
> was pointed out over a year ago by Skip Montanaro in the XML-RPC
> discussion forum:
>
> > Some time ago the subject of supporting discriminated null data in
> > XML-RPC was brought up and apparently dropped. I just checked the
> > spec and didn't see anything new. I scanned back through a couple
> > pages of topics but didn't see anything there either.
>
> > I'm busy using XML-RPC to communicate between front-end Python
> > and Perl stuff (CGI scripts, modules published through
> > ZopeHTTPServer or Mason components) and back-end Python stuff
> > (databases). The biggest problem I've encountered when converting
> > existing code to use XML-RPC is identifying and converting cases
> > where I use Python's None object. I have to figure out how it's used in
> > the caller and callee, then figure out a reasonable replacement. The end
> > result is usually that my code isn't as clear as it used to be and that
I
> > wind up adding the occasional extra bug to my code (as if I don't have
> > enough bugs to squash already).
>
> > <whine> I would *really* like to see a <nil></nil> or <nil/> data type
> > added to XML-RPC. </whine>
>
> I stumbled into the same problem and since I currently use XML-RPC only
> between my own server and clients I decided to ignore the official
> specification and added a 'None'-handler to my version of Fredrik Lundhs
> library.  The needed patch is indeed very very simple:  Add the following
> lines to the class 'Marshaller' in 'xmlrpclib.py' you are done:
>
>     def dump_None(self, none):
>         self.write("<value><null></null></value>\n")
>     dispatch[NoneType] = dump_None
>
> Since the class 'Unmarshaller' already returns 'None' for all unknown
values
> a modification there is not required.  This works fine for me.  YMMV.
>
> Regards, Peter
> --
> Peter Funk, Oldenburger Str.86, D-27777 Ganderkesee, Germany, Fax:+49
4222950260
> office: +49 421 20419-0 (ArtCom GmbH, Grazer Str.8, D-28359 Bremen)
>




From Richard.Jones at fulcrum.com.au  Wed Apr 19 18:23:19 2000
From: Richard.Jones at fulcrum.com.au (Richard Jones)
Date: Thu, 20 Apr 2000 08:23:19 +1000
Subject: Dynamic web pages! not Dynamic HTML 
In-Reply-To: Message from Sunil Hadap of 2000-Apr-19 20:52:2,
            <38FE0052.4BE78518@cui.unige.ch> 
Message-ID: <200004192223.IAA10606@envy.fulcrum.com.au>

[Sunil Hadap]
> I am really dumb in databases and perl. I am good at python and it's
> my scripting language for whatever I do in graphics. I want to start
> doing my homepage which is inspired from
> 
> http://www.photo.net/photo/
> 
[snip] 
> 
> I want to do this in Python. Is zope useful for this, I don't want to
> learn a big system as I only want few specific functions only. Which
> free database?

   What system are you running? If it's RedHat Linux (or similar) then you'll 
already have a relational database at your fingertips. You will need to figure 
out the basics of using it. If you have a reasonable number of photos (say < 
10,000) then you could even use gadfly, the Python RDBMS!

   Phil has spent quite a lot of time developing that site (Tcl over a 
relational database) - and you should expect to spend some time on it also. 
That's assuming he won't just give you his code :)

   I would suggest that you develop a Zope site - for three reasons:
   . it'll make using a "real" database much simpler thanks to Zope's 
easy-as-pie SQL methods. Remember, Zope comes with gadfly bundled in the box.
   . you'll automatically get access to a bunch of public-contributioned prior 
art: guestbooks, news pages, FAQ handlers and so on. See the member area on 
www.zope.org for a taste.
   . once you're done, you can make the Zope products you generate available to 
others so they can put up their photos too!

   What you describe with the @photo tag in a "sorta HTML" page is actually 
going to be quite simple to do in Zope. Ask around on the zope mailing lists if 
you're stuck - they're all quite helpful people over there :)


       Richard





From nmm1 at cus.cam.ac.uk  Tue Apr  4 15:25:57 2000
From: nmm1 at cus.cam.ac.uk (Nick Maclaren)
Date: 4 Apr 2000 19:25:57 GMT
Subject: What is python's language level?
References: <38EA1B8F.A797A130@lmco.com> <seqG4.3372$HG1.98334@nnrp1.uunet.ca>
Message-ID: <8cdfk5$58s$1@pegasus.csx.cam.ac.uk>

In article <seqG4.3372$HG1.98334 at nnrp1.uunet.ca>,
Warren Postma <embed at geocities.com> wrote:
>
>"Chuck Meyers" <chuck.meyers at lmco.com> wrote in message
>news:38EA1B8F.A797A130 at lmco.com...
>> Does anyone know python's language level? I did not find it in the list
>> at:
>
>Cheap guess; Perl is a 15, so we'll take the Perl number.

Well, considering that the whole thing smacks of a marketing
exercise, the level that you get may be what you pay for :-)

>Two things that make me laugh about this:

Only two?  I noticed three major factual errors and a host of known
lunatic classifications in just a quick glance.  Compare Algol 68,
Fortran 66, Basic and so on for a good giggle.

There was some serious work done in this area in the 1970s, and
the conclusions included the obvious one that a single level didn't
make a lot of sense.  If I remember correctly, other results
included:

    Higher is better if it doesn't stop you doing what you need to,
and is suited to the application domain.

    Very strong checking slows down time to the first demonstration
and speeds up time to the first stable version.

    Any language is bad if it is incomprehensible or too full of
"gotchas".

The expressions "motherhood and apple pie" and "so what else is
new?" spring to mind - though point 2 is very much out of fashion
at present, especially in the C and Java community.


Regards,
Nick Maclaren,
University of Cambridge Computing Service,
New Museums Site, Pembroke Street, Cambridge CB2 3QG, England.
Email:  nmm1 at cam.ac.uk
Tel.:  +44 1223 334761    Fax:  +44 1223 334679


From ns56645 at swt.edu  Mon Apr 24 11:56:00 2000
From: ns56645 at swt.edu (ABC)
Date: Mon, 24 Apr 2000 10:56:00 -0500
Subject: Tkinter.Button error
References: <3903E8C1.39B3E3D7@swt.edu> <8e180t$46d$1@nnrp1.deja.com>
Message-ID: <39046E90.4D4F89EF@swt.edu>

Yep,

That worked



From effbot at telia.com  Fri Apr 14 14:38:08 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Fri, 14 Apr 2000 18:38:08 GMT
Subject: No strides? (was Re: List mapping?)
References: <38f5d68e@news.xtml.co.uk> <38F75F7D.1C73D177@labs.agilent.com>
Message-ID: <kAJJ4.1036$rc9.189763584@newsb.telia.net>

David Smith <drs at labs.agilent.com> wrote:
> Why aren't strides implemented for slices of lists and tuples?  I would
> have liked to suggest
>
> map( None, a[0::2], a[1::2] )

someone has to write the code...

feel free to send your favourite implementation to
patches at python.org (see python.org for more info
on how to submit patches).

</F>




From mwh21 at cam.ac.uk  Wed Apr 12 14:26:10 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 12 Apr 2000 19:26:10 +0100
Subject: Python paradigms
References: <000401bfa1dd$930fe500$172d153f@tim> <8d2ejs$jfr$1@nyheter.chalmers.se>
Message-ID: <m33dori3b1.fsf@atrus.jesus.cam.ac.uk>

d98aron at dtek.chalmers.se (Fredrik Aronsson) writes:

> In article <000401bfa1dd$930fe500$172d153f at tim>,
> 	"Tim Peters" <tim_one at email.msn.com> writes:
> >    (a() and [b()] or [c()])[0]
> > 
> 
> Just a litte thougth, can't I just use (a() and b()) or c() ?

Not if b() returns false.

"(1 and 0 or 2)==2"-ly y'rs
M.

-- 
  On the other hand,  the following areas are subject to boycott
  in reaction to the rampant impurity of design or execution, as
  determined after a period of study, in no particular order: 
    ...                            http://www.naggum.no/profile.html


From alex at somewhere.round.here  Tue Apr 11 10:21:16 2000
From: alex at somewhere.round.here (Alex)
Date: 11 Apr 2000 10:21:16 -0400
Subject: Why should I switch to Python?
References: <seas5ia89177@corp.supernews.com> <slrn8f4d0j.1t0.bwinton@tor.dhs.org> <slrn8f4tfr.6hh.quinn@mono.ugcs.caltech.edu> <IvFI4.5759$HG1.168717@nnrp1.uunet.ca>
Message-ID: <etdu2h8btwj.fsf@w20-575-109.mit.edu>

>> I always suspected perl programmers were from Mars.  Now if we could
>> figure out where python programmers are from, someone could write a
>> book and get rich.
> 
> Yorkshire.

"Perl Programmers are from Mars, Python Programmers are from Yorkshire?"
Catchy title.

Alex.


From nick at spam_me_notvideosystem.co.uk  Thu Apr 13 14:00:59 2000
From: nick at spam_me_notvideosystem.co.uk (Nick Trout)
Date: Thu, 13 Apr 2000 19:00:59 +0100
Subject: How do I execute a program ?
References: <20000413163828.28296.qmail@web4303.mail.yahoo.com>
Message-ID: <38f60cb6@news.xtml.co.uk>

import os
exit_status = os.system('blah.exe')


"Daniel Zapata" <danzap_python at yahoo.com.br> wrote in message
news:20000413163828.28296.qmail at web4303.mail.yahoo.com...
> How do I execute a program or a batch file (under
> Windows) using python and get the return information
> ???
>
>
> Thanks,
>
> Daniel
>
> ____________________________________________________________________________
> YAHOO! BATE-PAPO
> Conhe?a gente nova, fa?a amigos e divirta-se! - http://www.yahoo.com.br
>




From soundhack at hotmail.com  Thu Apr 20 21:29:53 2000
From: soundhack at hotmail.com (Robert)
Date: Thu, 20 Apr 2000 21:29:53 -0400
Subject: Why MFC instead of Tkinter -- was Re: Pythonwin?
References: <eMz2Nmiq$GA.73@net025s> <4QtL4.24$v85.41@news-server.bigpond.net.au> <chuL4.8483$O7.17315@typhoon.southeast.rr.com> <7LDL4.534$v85.2563@news-server.bigpond.net.au> <sfu6d657l6e184@corp.supernews.com> <38ff21b8$0$10972@senator-bedfellow.mit.edu> <38FF281E.5B2DCC71@tismer.com> <38ff823d$0$10971@senator-bedfellow.mit.edu> <dNLL4.680$wYl.189480448@newsb.telia.net>
Message-ID: <38ffafc0$0$10979@senator-bedfellow.mit.edu>

Fredrik Lundh wrote in message ...
>Robert <soundhack at hotmail.com> wrote:
>> (and I assume) Delphi has similar and probably even easier editors, but
>the
>> fact remains that (to me) wysiwyg editing is better than layout managers.
>
>just like refrigerators are better than car stereos?


since I have the former and not the latter, yes :)

>
>not all wysiwyg tools restrict you to drawing things on a
>coordinate grid, you know...


yup, I havent seen any (tho i havent been looking too hard) but sure, there
is nothing about layout managers that precludes a graphical tool from
generating code.
If you know of any that are as polished as VC/VB's then I'ld be happy to try
them out.

On a more design issue, how can you guarantee that a UI looks exactly the
way you want to without using coordinates? I have some UIs that require lots
of different sized controls that do not fit a neat NSEW or table/matrix
spatial organization.

Robert




From mhammond at skippinet.com.au  Wed Apr 19 22:18:35 2000
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Thu, 20 Apr 2000 02:18:35 GMT
Subject: COM and returning dates
References: <8dkj9b$lvr$1@gxsn.com>
Message-ID: <%NtL4.21$v85.112@news-server.bigpond.net.au>

Look up the documentation for pythoncom.MakeTime() - this will create a
time object from a time represented by the standard Python "time" module.
Create and pass one of these objects to COM as a date/time value.

mxDateTime objects can also be converted to COM date/times, so this is
another alternative.

Mark.

"Richard Chamberlain" <richard_chamberlain at ntlworld.com> wrote in message
news:8dkj9b$lvr$1 at gxsn.com...
> Hi,
>
> I'm trying to use Python as a COM object (called from VB) basically
> returning values from a database, some of which are dates.
>
> The problem I'm getting is that the dbi date object obviously doesn't
> translate.
>
> So is there a way a can simply do this?
>
> I was thinking of running through the list and converting the dbi date
> objects into string representations and passing those. But I'm not sure
what
> I do with them on the VB side.
>
> Any better ideas?
>
> Thanks
>
> Richard
>
>




From mwh21 at cam.ac.uk  Wed Apr  5 12:18:46 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 05 Apr 2000 17:18:46 +0100
Subject: reload imported module
References: <05fd01bf9f18$6e5a33e0$01ffffc0@worldnet.att.net>
Message-ID: <m3vh1wldw9.fsf@atrus.jesus.cam.ac.uk>

"Emile van Sebille" <emile at fenx.com> writes:

> I'm testing a new class in its own module from within
> idle.  Is there a way to reload y after having done:
> 
>     from x import y

reload(sys.modules['x']); from x import y

I don't think there's any feasible way to automate this.
 
> or is it easiest to just import x and reload it?

Yes.  This works too:

import module_with_really_long_name

mwrln = module_with_really_long_name

... use mwrln ...

reload(mwrln)

... use updated version ...

Cheers,
M.

-- 
  well, take it from an old hand: the only reason it would be easier
  to program in C is that you can't easily express complex  problems
  in C, so you don't.                   -- Erik Naggum, comp.lang.lisp


From jorn at spekter.no  Thu Apr 27 06:31:31 2000
From: jorn at spekter.no (=?iso-8859-1?Q?J=F8rn?= Eilertsen)
Date: Thu, 27 Apr 2000 12:31:31 +0200
Subject: python + linux + cron = problem
Message-ID: <39081703.48B64D1A@spekter.no>

Hi,

I've posted this earlier but noone replied so I'm trying again

I have this small program which includes the following line og code:
 mounts = commands.getoutput('df | grep eggum | cut -d 1 -f 1 | cut -d /
-f 4')

this runs fine from the commandline but when run as a cronjob I get the
following error:

File "/usr/lib/python1.5/commands.py", line 42, in getoutput
    return getstatusoutput(cmd)[1]
  File "/usr/lib/python1.5/commands.py", line 53, in getstatusoutput
    sts = pipe.close()
IOError: (10, 'No child processes')

The use of pipes seems to be the trouble here.

Any help on how the get around this would be greatly appreciated.

Thanks,
Jorn





From kcazabon at home.com  Wed Apr 26 00:24:41 2000
From: kcazabon at home.com (Kevin Cazabon)
Date: Wed, 26 Apr 2000 04:24:41 GMT
Subject: Simple list.append() question
References: <meh9-5ED8CD.11192424042000@news.cit.cornell.edu> <ZHZM4.360$qg7.2949@news.uswest.net> <39046FA2.603EAFE5@roguewave.com> <PsaN4.174166$Dv1.2059732@news1.rdc1.bc.home.com> <8e4akc$p7o$1@newshost.accu.uu.nl>
Message-ID: <dcuN4.177124$Dv1.2114909@news1.rdc1.bc.home.com>

Aw, comon, I was having fun... q:]


"Martijn Faassen" <m.faassen at vet.uu.nl> wrote in message
news:8e4akc$p7o$1 at newshost.accu.uu.nl...
> Kevin Cazabon <kcazabon at home.com> wrote:
> > well, if not prettier, why not obfuscate it?  q:]  Although this isn't
> > pretty, I'm sure we can make it a little less intelligable with some
work.
>
> > entire_list = [];for i in range(3):entire_list.append([])
>
> If you write it the normal way:
>
> entire_list = []
> for i in range(3):
>     entire_list.append([])
>
> It's pretty, readable, flexible, and it works. The only disadvantage is
> that it's less short. But I'm willing to pay that price in this situation.
>
> Generally it's a good idea to avoid * on sequences, unless that sequence
is
> immutable and contains immutable things; i.e. strings. I don't recall
> using * on tuples ever, but you can do so safely if the tuple contains
> immutable things only.
>
> Is there any useful way to use * on lists that I missed? Usually we want
> the copy semantics here, not the reference semantics. Perhaps it's a good
> idea to completely forbid * on lists in p3k? Tuples too for all I care.
> Then again I may be missing important uses, so enlighten me.
>
> Regards,
>
> Martijn
> --
> History of the 20th Century: WW1, WW2, WW3?
> No, WWW -- Could we be going in the right direction?




From gansevle at cs.utwente.nl  Thu Apr 13 10:23:19 2000
From: gansevle at cs.utwente.nl (Fred Gansevles)
Date: Thu, 13 Apr 2000 16:23:19 +0200
Subject: Redirecting the output from sys.stdout
References: <2B1262E83448D211AE4B00A0C9D61B03BA7296@MSGEURO1>
Message-ID: <38F5D857.F11B0B4A@cs.utwente.nl>

Pieter Claerhout wrote:

> Hello,
>
> I want to run a DOS command from Python, and I want to redirect the output of
> that command to a variable. Anyone a suggestion on how to accomplish this? The
> output of the command should not be shown on the screen.
>
> Kind regards,
>
> Pieter
>
> [ Pieter Claerhout :: CreoScitex Response Centre :: Applications support ]
> [ Pieter_Claerhout at creoscitex.com    ::    http://www.creoscitex.com/ ]
> [ CreoScitex Europe  ::   Excelsiorlaan 21   ::  1930 Zaventem  ::   Belgium ]

My favorite:

    def run (*args, **kw):
        import win32pipe
        command = string.join (args)
        i, o, e = win32pipe.popen3 (command)
        if kw.has_key ('send'):
            i.write (kw['send'])
        i.close ()
        out = o.read ()
        err = e.read ()
        return out    # discard stderr output

    ...

    bat_files = run ("dir /s *.bat")

___________________________________________________________________________
 Fred Gansevles <mailto:Fred.Gansevles at cs.utwente.nl> Phone: +31 53 489 4613
        >>> Your one-stop-shop for Linux/WinNT/NetWare <<<
 Org.: Twente University, Fac. of CS, Box 217, 7500 AE Enschede, Netherlands
            "Bill needs more time to learn Linux" - Steve B.



From rumjuggler at home.com  Fri Apr 14 18:19:26 2000
From: rumjuggler at home.com (Ben Wolfson)
Date: Fri, 14 Apr 2000 22:19:26 GMT
Subject: for item in some_class_object...
Message-ID: <4j6ffs0b6ins12qq559q07u28pc6o0mc9l@4ax.com>

What special methods should I define for a class if I want to support
a for item in ... syntax for it?

-- 
Barnabas T. Rumjuggler al-Mumit

Further-more, I have long since put the finishing touches on my own
official obituary, a 350-page tale of heart-rending loss which
details, among other things: my meteoric rise from Onion stringer to
editor to publisher; my assassination at the hands of lady-actress
Ethel Barrymore's jealous suitors; the thousand days of mourning which
followed my death, during which a cold rain fell constantly, no bird
sang, and children were born antlered and hooved; ...
 -- T. Herman Zweibel


From DOUGS at oceanic.com  Sat Apr 29 12:58:10 2000
From: DOUGS at oceanic.com (Doug Stanfield)
Date: Sat, 29 Apr 2000 06:58:10 -1000
Subject: newbie   dbase dataaccess
Message-ID: <8457258D741DD411BD3D0050DA62365907A12D@huina.oceanic.com>

You're starting on an ambitious project.  I find what works for me is to
simplify all the elements of such a project until they are done trivially.
If I've successfully modularized things I can then improve each piece after
I've got a working prototype.  For the data store the progression of
improvement might be a set of flat files, a db file using the 'shelve'
module, and then maybe start to learn Gadfly as a database backend.
Learning to use SQL in a new database package at the same time working out
program logic and learning Python would be enough to squelch my
motivation:-)

speaking-from-current-experience'ly yr's -Doug-

> -----Original Message-----
> From: Georg Hammerbacher [mailto:don.limpio at gmx.net]
> Sent: Friday, April 28, 2000 4:17 AM
> To: python-list at python.org
> Subject: Re: newbie dbase dataaccess
> 
> 
> Hello Anders,
> 
> thanks for your guess.
> 
> The thing inside my head is to write an adressmanager
> - which is the *central* adressbook
> - who works under Windows and Linux
> - has an GUI Interface
> this are the features for the beginning.
> 
> my dream is (under the GPL licence), to expand this piece of 
> software and
> integrate the good Features of MS-Outlook and Lotus 
> Organizer. (comments are
> welcome)
> 
> I?ve found "gadfly" for databasing. Is it really so good as it seems??
> 
> Thanks
> Georg
> 
> 
> Anders M Eriksson <anders.eriksson at morateknikutveckling.se> 
> schrieb in im
> Newsbeitrag: mtiigskn8uumhk94rklnr2fdealunasoot at 4ax.com...
> > On Thu, 27 Apr 2000 11:40:46 +0200, "Georg Hammerbacher"
> > <don.limpio at gmx.net> wrote:
> >
> > >Hello everybody,
> > >
> >
> > Hello Georg!
> >
> > >I?m just new in programming Python and now I want to make 
> my own personal
> > >adressmanager with a simple dBase-file.
> > >How can I get access to it or where can I find more info?s 
> about using
> > >dbase-files within Python?
> > >
> > First: Why do you want to use dBASE as the database format?
> >
> > I haven't found any good dBASE interface for Python. So if 
> you want to
> > use the dBASE format you need to use ODBC and then you hopefully are
> > on a computer that runs unders Windows NT/98.
> >
> > If you don't NEED to use dBASE then look at
> > http://www.python.org/topics/database/ here you will find 
> alot of info
> > about how to use Python for accessing databases....
> >
> > // Anders
> >
> 
> 
> -- 
> http://www.python.org/mailman/listinfo/python-list
> 



From mwh21 at cam.ac.uk  Tue Apr 11 06:26:05 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 11 Apr 2000 11:26:05 +0100
Subject: [ANNOUNCE] Garbage collection for Python
References: <20000407221750.A29279@acs.ucalgary.ca> <38F13FB9.D1D16AF@mindspring.com> <slrn8f2if1.dop.nascheme@cranky.arctrix.com> <87r9cdhdb0.fsf@deneb.cygnus.argh.org>
Message-ID: <m3bt3hkk76.fsf@atrus.jesus.cam.ac.uk>

Florian Weimer <fw at deneb.cygnus.argh.org> writes:

> nascheme at enme.ucalgary.ca (Neil Schemenauer) writes:
> 
> > I would rather get the bugs stomped out of my patch and get it
> > included in 1.6 or 1.7.  We can then remove the FAQ entry.  :)
> 
> Is it possible to extend your GC in a way which would permit the
> execution of several bytecode interpreters in multiple threads?

Eh?

-- 
  please realize that the Common  Lisp community is more than 40 
  years old.  collectively, the community has already been where 
  every clueless newbie  will be going for the next three years.  
  so relax, please.                 -- Erik Naggum, comp.lang.lisp


From francesco.contini at tiscalinet.it  Sun Apr  9 20:44:04 2000
From: francesco.contini at tiscalinet.it (FrancescO)
Date: Mon, 10 Apr 2000 02:44:04 +0200
Subject: The gzip module... neverending story
Message-ID: <8cr7os$du3$1@lacerta.tiscalinet.it>

Hello!,
I have, another questions!!
How can I zip i list of files? (i have done it, but... the files can't be
readed from winzip and gunzip...)

Wich is the correct usage of the fileobj (4th parameters in the
gzip.GzipFile.. explain..:
GzipFile ([filename[, mode[, compresslevel[, fileobj]]]]) )?

Is the name of the zipfile? what kind of object is it?
i'have try in a lot of ways to call this object, but I'm not "good" to work
it with the 4th parameter please...

HELP ME...!!!!






From phd at phd.russ.ru  Mon Apr 10 09:12:41 2000
From: phd at phd.russ.ru (Oleg Broytmann)
Date: Mon, 10 Apr 2000 13:12:41 +0000 (GMT)
Subject: Python paradigms
In-Reply-To: <8csbab$1422@newton.cc.rl.ac.uk>
Message-ID: <Pine.LNX.4.21.0004101310560.21798-100000@fep132.fep.ru>

On Mon, 10 Apr 2000, Richard Brodie wrote:
> If this were always the case, why does Python have an interactive mode?

   This is for beginners - to try out few simple things :)

> The Fast
> Fourier Transform is one of the classic algorithmic optimisations but
> produces obscure code. Likewise most of the standard fast sorting
> algorithms.

   Yes. And you cannot comment every line of code - your code in that case
should be acomplished with math details (not programming details).

Oleg.
---- 
    Oleg Broytmann    http://members.xoom.com/phd2.1/    phd2 at earthling.net
           Programmers don't die, they just GOSUB without RETURN.




From greg at cosc.canterbury.ac.nz  Thu Apr 27 00:50:03 2000
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Thu, 27 Apr 2000 16:50:03 +1200
Subject: List comprehensions?
References: <000601bfae79$41485b00$152d153f@tim> <slrn8gbcrj.a24.scarblac-spamtrap@flits104-37.flits.rug.nl>
Message-ID: <3907C6FB.A418A32D@cosc.canterbury.ac.nz>

Remco Gerlich wrote:
> 
> Tim Peters wrote in comp.lang.python:
> > [Remco Gerlich]
> > > I thought list comprehensions had been in CVS for quite a while, so I
> > > expected them to be in 1.6. ...
> >
> > Greg Ewing's patch (against 1.5.2) for list comprehensions is all there is,
> > and you'll have to get that from Greg (see DejaNews for his announcement,
> > and his URL).

Skip Montario (skip at mojam.com) has recently informed me that he has
updated
my patch to work with 1.6a2, and is going to try to get it included
officially.

-- 
Greg Ewing, Computer Science Dept,
+--------------------------------------+
University of Canterbury,	   | A citizen of NewZealandCorp, a	  |
Christchurch, New Zealand	   | wholly-owned subsidiary of USA Inc.  |
greg at cosc.canterbury.ac.nz	   +--------------------------------------+


From neelk at brick.cswv.com  Tue Apr 18 19:37:06 2000
From: neelk at brick.cswv.com (Neel Krishnaswami)
Date: 18 Apr 2000 23:37:06 GMT
Subject: Project idea - python
References: <BxGK4.811$9B.103178@news.bc.tac.net> <38FC0E72.9AB025DB@webwitches.com>  <Pine.GSO.4.21.0004181113110.22094-100000@sun21epsl.wam.umd.edu>
Message-ID: <slrn8fpudf.5fl.neelk@brick.cswv.com>

Bob Schmertz <schmertz at wam.umd.edu> wrote:
> 
> SPARK - sounds like a Python implementation of Yacc/Bison, though from
> what I can tell it doesn't handle complex grammars such as LR
> (http://www.csr.UVic.CA/~aycock/python/)
 
SPARK uses the Earley parsing algorithm, which can parse *any* context
free grammar, even if the grammar is ambiguous or otherwise wonky.

IIRC, Earley parsers have the neat property of being "adaptive" --
their performance is O(n) on LL/LR grammars, rising up to O(n**3) for
pathologically ambiguous grammars. This generality doesn't come for
free, though -- the extra overhead in data structures tends to reduce
their speed by roughly a factor of 10 versus the usual algorithms for
simple grammars.

But for this modest price you can specify the grammar that most
naturally generates your language, rather than going through weird
contortions needed to satisfy Yacc. :)


Neel


From tim.hochberg at ieee.org  Mon Apr 24 18:13:41 2000
From: tim.hochberg at ieee.org (Tim Hochberg)
Date: Mon, 24 Apr 2000 22:13:41 GMT
Subject: sys.exit
References: <3904C35C.B16BDE75@mindspring.com>
Message-ID: <pG3N4.72826$U4.436278@news1.rdc1.az.home.com>

> The long story:
>
> I have a try..except like so:
>
> try:
>     execfile(filename, namespace)
> except:
>     # handle problems


Try something like

try:
   # whatever
except SystemExit, err:
    if err.args in  [(0,), ()]: # sys.exit called as sys.exit(0) or
sys.exit()
        # exit cleanly
    else:
        # log error
except:
    # handle any remaining errors.


Enjoy,

-tim






From effbot at telia.com  Wed Apr 12 12:25:32 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Wed, 12 Apr 2000 16:25:32 GMT
Subject: Python idiom: Multiple search-and-replace
References: <20000412100816.A2431031@vislab.epa.gov><5C%I4.645$rc9.190288896@newsb.telia.net> <20000412120012.D2446939@vislab.epa.gov>
Message-ID: <0s1J4.655$rc9.190631424@newsb.telia.net>

Randall Hopper <aa8vb at yahoo.com> wrote:
> Thanks!  It's much more efficient.  The 140 seconds original running time
> was reduced to 11.6 seconds.  I can certainly live with that.

thought so ;-)

while you're at it, try replacing the original readline loop with:

    while 1:
        lines = fp.readlines(BUFFERSIZE)
        if not lines:
            break
        lines = string.join(lines, "")
        lines = re.sub(...)
        out_fp.write(lines)

where BUFFERSIZE is 1000000 or so...

</F>




From aa8vb at yahoo.com  Tue Apr 18 07:20:54 2000
From: aa8vb at yahoo.com (Randall Hopper)
Date: Tue, 18 Apr 2000 07:20:54 -0400
Subject: Starship dead (again)?!
In-Reply-To: <38FC39F1.21DD2D75@darwin.in-berlin.de>
References: <38FC39F1.21DD2D75@darwin.in-berlin.de>
Message-ID: <20000418072054.A416925@vislab.epa.gov>

Dinu C. Gherman:
 |I can't get anything from http://starship.python.net or
 |http://starship.python.net:8080 for more than a day, now.
 |Is that some known problem?
 |
 |This seems to happen ever more frequently. Does somebody
 |simply pull the plug, or is there another, better, reason?

Reposted message:
==============================================================================

Subject: Re: Here is some starship.python.net news
Author:  Manus Hand <mjhand at concentric.net>
Date:    2000/04/17
Forum:   comp.lang.python

Here is some more news.  Everyone cross your fingers for the starship's
disk.

Manus

Reply-To: "Paul Wolotsky" <pdw at mcsp.com>
Date: Mon, 17 Apr 2000 05:04:22 -0400
Organization: Medical Computer Systems

...My understanding from my NOC technicians is that the primary disk has
crashed in that machine - so perhaps there is a delay in someone coming in
to fix it in terms of them getting a replacement disk, etc.  I will let you
know tomorrow when I return to the office - if there is something that you
can do I will advise at that time!!!  Thanks for the offer.

==============================================================================

--
Randall Hopper
aa8vb at yahoo.com



From tuttledon at hotmail.com  Fri Apr 14 15:49:46 2000
From: tuttledon at hotmail.com (Don Tuttle)
Date: Fri, 14 Apr 2000 19:49:46 GMT
Subject: Getting file version information
References: <20000413163629.28006.qmail@web4303.mail.yahoo.com> <81wJ4.811$1D.5094@news-server.bigpond.net.au>
Message-ID: <uDKJ4.20564$042.80272@typhoon.southeast.rr.com>

> You may be forced to using calldll (www.nightmare.com) to call the API
> functions directly.

Speaking of Nightmare.com, I've been trying for 4 days to grab calldll but
there's no response.  Anyone know what happened to it?

Don




From michaelc at flashcom.net  Fri Apr 21 02:35:53 2000
From: michaelc at flashcom.net (Michael A. Crawford)
Date: Thu, 20 Apr 2000 23:35:53 -0700
Subject: Why MFC instead of Tkinter -- was Re: Pythonwin?
References: <eMz2Nmiq$GA.73@net025s> <4QtL4.24$v85.41@news-server.bigpond.net.au> <chuL4.8483$O7.17315@typhoon.southeast.rr.com> <7LDL4.534$v85.2563@news-server.bigpond.net.au> <sfu6d657l6e184@corp.supernews.com> <38ff21b8$0$10972@senator-bedfellow.mit.edu>
Message-ID: <sfvtfmj7l6e171@corp.supernews.com>

I'm sorry I asked.

-Michael

"Robert" <soundhack at hotmail.com> wrote in message
news:38ff21b8$0$10972 at senator-bedfellow.mit.edu...
> To each his own I guess. I work mostly on Windows platforms, and for me
> MFC is a lot better than Tk for a couple of reasons.
> 1) I can port MFC/C++ programs a lot easier
> 2) I may be mistaken, but MFC based guis are faster than Tk
> 3) I really do not like using layout managers, and using Visual C++'s
dialog
> editor
> is the best thing since sliced bread IMHO.
>
> One suggestion for learning MFC fresh: don't read books that basically
teach
> you how to use Visual C++'s AppWizard. AppWizard is really nice and easy
to
> use but it basically automates so much of creating MFC programs that it
was
> hard (at least for me) to make the transition to a non-wizard MFC
> environment in python.
>
> Robert
>
> Michael A. Crawford wrote in message ...
> >I'm curious.  Why did you not steer him in the Tkinter direction for his
> GUI
> >needs?  Is MFC better?  I learned (and forgot MFC) some time ago, years
in
> >fact.  Same with X-Windows programming.  Now I'm working my way throught
> >"Python and Tkinter Programming" by Grayson.  I thought Tkinter was the
> >preferred method for Pythong GUI programming.  Am I mistaken?
> >
> >-Michael
>
>
>




From mwh21 at cam.ac.uk  Fri Apr 14 10:13:58 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 14 Apr 2000 15:13:58 +0100
Subject: Complaint Dept: My shoes!
References: <8d4s4d$mvr$1@nnrp1.deja.com> <38F72CE2.CB7B7C5D@att.net>
Message-ID: <m3ln2gg47t.fsf@atrus.jesus.cam.ac.uk>

aflinsch <avflinsch at att.net> writes:

> musingattheruins at my-deja.com wrote:
> > 
> > If shoe is None then len(shoe) raises an exception.  Would be nice if
> > len(None) was 0 and NOT raise an exeption.
> > 
> 
> I think I would prefer it if len(None) returned None. Makes a bit more
> sense to me having an unsized object return None, rather than zero.

I would prefer that len(ob) raised an exception whenever it was passed
an `ob' that didn't have a well defined length (like None).

Second guessing what would be most helpful to the programmer smacks of
another p-language.

Is "ob and len(ob) or 0" too hard to spell?

always-willing-to-back-the-status-quo-in-cases-like-these-ly y'rs
m.

-- 
  On the other hand,  the following areas are subject to boycott
  in reaction to the rampant impurity of design or execution, as
  determined after a period of study, in no particular order: 
    ...                            http://www.naggum.no/profile.html


From c.evans at clear.net.nz  Wed Apr 12 06:31:18 2000
From: c.evans at clear.net.nz (Carey Evans)
Date: Wed, 12 Apr 2000 10:31:18 GMT
Subject: Processing BCD and Signed Integer Fields with Python
References: <38f334ea@wwwproxy3.westgroup.com>
	<8cvdel$lb9$1@pegasus.csx.cam.ac.uk>
	<38f33dca@wwwproxy3.westgroup.com>
Message-ID: <878zyjpq4p.fsf@psyche.evansnet>

"Chris Lada" <chris.lada at westgroup.com> writes:

>  Being a newbie in python, I have to ask these questions: what functions can
> I use to process these fields ?  What can I use to separate the number from
> the sign ?   Is there a conversion for this type of data to strings ? What
> data types would they be ?

Here's some code that works with the kind of data I can get from our
IBM AS/400.  It could probably be more efficient, and your mainframe
might not be generating the same data formats, but it should give you
the right idea.

For example:

>>> packed2num('\x12\x34\x5f')
12345L
>>> zoned2num('\xf1\xf2\xd3')
-123L

--------------------
import array

def zoned2num(z):
    a = array.array('B', z)
    v = 0L

    for i in a:
        v = (v * 10) + (i & 0xf)

    if (a[-1] & 0xf0) == 0xd0:
        v = -v

    return v

def packed2num(p):
    a = array.array('B', p)
    v = 0L

    for i in a[:-1]:
        v = (v * 100) + (((i & 0xf0) >> 4) * 10) + (i & 0xf)

    i = a[-1]
    v = (v * 10) + ((i & 0xf0) >> 4)
    if (i & 0xf) == 0xd:
        v = -v

    return v
--------------------

-- 
	 Carey Evans  http://home.clear.net.nz/pages/c.evans/

"Validate me!  Give me eternal digital life!  Quote me in your .sigs!"
                                                             - djc in asr


From python at rose164.wuh.wustl.edu  Tue Apr  4 17:54:21 2000
From: python at rose164.wuh.wustl.edu (David Fisher)
Date: Tue, 4 Apr 2000 16:54:21 -0500
Subject: Two questions
References: <38EA5FE8.853E9B75@callware.com>
Message-ID: <01f801bf9e80$5a522a80$29dc9f26@unitedelectric>

----- Original Message ----- 
From: Ivan Van Laningham <ivanlan at callware.com>
To: Python Mailing List <python-list at python.org>
Sent: Tuesday, April 04, 2000 4:34 PM
Subject: Two questions

> 2)  What stuff do I need to use msaccess with Python?  Do I need to use
> Win32all, or can I use Tkinter for my (minimal) GUI, with msaccess?

What do you want to do with ms access?

dnf




From effbot at telia.com  Mon Apr 10 16:49:58 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Mon, 10 Apr 2000 20:49:58 GMT
Subject: Jpython??
References: <955394930.758181@newdm.deskmedia.com>
Message-ID: <W7rI4.431$rc9.188052480@newsb.telia.net>

Brock Noland <brockn at deskmedia.com> wrote:
> Hello   I am a web developer.. Well part time.  I can 'program' in HTML
and
> am interested in learning Java but I was told that I should learn Python
> first because it is a fairly newbie friendly language.  To the point I saw
a
> post about Jpython yesterday what is it??

www.jpython.org

</F>




From jvickroy at sec.noaa.gov  Wed Apr 19 09:46:15 2000
From: jvickroy at sec.noaa.gov (j vickroy)
Date: Wed, 19 Apr 2000 07:46:15 -0600
Subject: Python in the future
Message-ID: <38FDB8A7.E701F81F@sec.noaa.gov>

Hello all,

I'm relatively new to Python and am trying to gain support for it at my
work place.  To that end, I'm giving a brief (overview) presentation.
One of the items I have been asked to consider is where will Python be
5-10 years from now.  I certainly don't know where C++ or Java will be
in that time frame!  Any ideas on where to research the future plans for
Python would be appreciated.

Thanks for your time,

- jv



From ns56645 at swt.edu  Sat Apr 29 13:19:09 2000
From: ns56645 at swt.edu (ns56645)
Date: Sat, 29 Apr 2000 12:19:09 -0500
Subject: problem in python function --Name-error
Message-ID: <390B198D.7AC26774@swt.edu>

Hi,

I am sending you the whole code of my program as an attachment.  But the
error lies in def drawCanvas( ): I have emphasized the error. The error
says nameError frame1. The output of my program looks similar to
http://www.swt.edu/~ns56645.
Could you please help me with the error.






-------------- next part --------------
import sys
sys.path[:0] = ['../']



from Tkinter import *
import Pmw

class Schedule:

       def __init__(self,master):
	global mycourses
	mycourses = [1000, 1203, 3224, 1244, 3453, 3342, 2232, 2131, 2342, 2342, 2222, 5544, 5232]
	time = [8.00, 9.30, 11.00, 12.30, 2.00, 3.30, 5.00, 6.30]

#def create_frames(master):
    # relief and borderwidth arguments make frames visible

##############   Create four frames ##############################################

   	frame1 = Frame(master, relief='sunken', borderwidth=2)
    	frame2 = Frame(master, relief='sunken', borderwidth=1)
    	frame3 = Frame(master, relief='sunken', borderwidth=1)
    	frame4 = Frame(master, relief='sunken', borderwidth=1)

    	frame1.place(relx=0, rely=0, relwidth=1, relheight=0.70)
	frame2.place(relx=0, rely=0.70, relwidth=0.40, relheight=0.30)
    	frame3.place(relx=0.40, rely=0.70, relwidth=0.30, relheight=0.30)
    	frame4.place(relx=0.70, rely=0.70, relwidth=0.30, relheight=0.30)
       	
############### Canvas for Days in Frame 1 ##########################################

	self.days=Canvas(frame1,width=50, height=430, bg='white')
	self.drawDays()
	self.days.grid(row=0, column=0, rowspan=7, sticky=W+N)
	

################# Canvas for Time in frame 1 #########################################

	self.time = Canvas(frame1, width=800, height=30, bg='white')
	self.time = Canvas(width=600, height=30, bg='white')
	self.drawTime()
	self.time.grid(row=0, column=1)

############## DRAW CANVASES IN FRAME 1  ###############################################

	self.drawCanvas()

###################   COMBOBOX FOR COURSES IN FRAME 2    ################################################


        group = Pmw.Group(frame2, tag_text = 'Selection Criteria')
        group.pack(fill = 'both', expand = 1, padx = 15, pady = 10)

	# Create and pack the dropdown ComboBox.
        courses = [' ',4310, 5389, 4328,3358, 5310, 7000]
	dropdown1 = Pmw.ComboBox(group.interior(),
                label_text = 'Courses      ',
	        labelpos = 'w',
		listheight = 150,
#		buttonaspect = 2.0,
                #selectioncommand = self.changeColour,
		scrolledlist_items = courses,
        )
#	dropdown1.grid(row=1)
        dropdown1.pack(pady=5)

	# Display the first colour.
	first = courses[0]
	dropdown1.selectitem(first)
	#self.changeText(first)


###############  COMBOBOX FOR ROOMS  IN FRAME 2      #######################################################


	# Create and pack the dropdown ComboBox.
        rooms = [' ','DH241', 'DH113', 'DH240','DH213']
	dropdown2 = Pmw.ComboBox(group.interior(),
                label_text = 'Rooms       ',
	        labelpos = 'w',
                #selectioncommand = self.changeColour,
		scrolledlist_items = rooms,
        )
#	dropdown2.grid(row=2)
	dropdown2.pack()
	# Display the first colour.
	first = rooms[0]
	dropdown2.selectitem(first)
	#self.changeColour(first)

#################       COMBOBOX FOR PROFESSORS IN FRAME2        ################################


	# Create and pack the dropdown ComboBox.
        instruc = [' ','Davis', 'Mccabe', 'Kaikhah', 'Peng', 'Hall']
	dropdown3 = Pmw.ComboBox(group.interior(),
                label_text = 'Instructors ',
	        labelpos = 'w',
		hull_width=5,
                #selectioncommand = self.changeColour,
		scrolledlist_items = instruc,
        )
#	dropdown2.grid(row=2)
	dropdown3.pack()
	# Display the first colour.
	first = instruc[0]
	dropdown3.selectitem(first)
	#self.changeColour(first)


######################COMBOBOX FOR LABS IN FRAME 2 ################################################


	# Create and pack the dropdown ComboBox.
        labs = [' ','DH231', 'DH235', 'DH231C', 'DH231B']
	dropdown4 = Pmw.ComboBox(group.interior(),
                label_text = 'Labs          ',
	        labelpos = 'w',
                #selectioncommand = self.changeColour,
		scrolledlist_items = labs,        )
#	dropdown2.grid(row=2)
	dropdown4.pack()
	# Display the first colour.
	first = labs[0]
	dropdown4.selectitem(first)
	#self.changeColour(first)



####         Frame 3  Radio Buttons        #####################################################

        # Create the "Location" contents of the page.
        group = Pmw.Group(frame3, tag_text = 'Location')
        group.pack(fill = 'both', expand = 1, padx = 15, pady = 10)

	v1 = IntVar()
	v1.set(3) # initialize
        rb1 = Radiobutton(group.interior(), text = 'SWT', variable=v1, value=1)
        rb1.grid(row = 0, column = 0)
        rb2 = Radiobutton(group.interior(), text = 'WHS', variable=v1, value=2)
        rb2.grid(row = 0, column = 1)
        rb3 = Radiobutton(group.interior(), text = 'BOTH', variable=v1, value=3)
        rb3.grid(row = 0, column = 2)
	

        # Create the "Level" contents of the page.
        group = Pmw.Group(frame3, tag_text = 'Level')
        group.pack(fill = 'both', expand = 1, padx = 15, pady = 10)

	v2 = IntVar()
	v2.set("13") # initialize
        rb11 = Radiobutton(group.interior(), text = 'Grad', variable=v2, value=11)
        rb11.grid(row = 0, column = 0)
        rb12 = Radiobutton(group.interior(), text = 'UnderGrad', variable=v2, value=12)
        rb12.grid(row = 0, column = 1)
        rb13 = Radiobutton(group.interior(), text = 'BOTH', variable=v2, value=13)
        rb13.grid(row = 0, column = 2)


############      Frame 4     ##########################################################


#################        Add Combobox         ##################


        add = [' ', 'Course', 'Instructor', 'Room', 'Time', 'Record']
	dropdown10 = Pmw.ComboBox(frame4,
                label_text = 'ADD    ',
	        labelpos = 'w',
		entry_width=4,
		listheight = 150,
#		buttonaspect = 2.0,
                #selectioncommand = self.changeColour,
		scrolledlist_items = add,
        )
#	dropdown1.grid(row=1)
        dropdown10.pack(pady=5)

	# Display the first colour.
	first = add[0]
	dropdown10.selectitem(first)
	#self.changeText(first)


################     Modify ComboBox In frame 4     ####################


	# Create and pack the dropdown ComboBox.
        modify = [' ', 'Course', 'Instructor', 'Room', 'Time', 'Record']
	dropdown11 = Pmw.ComboBox(frame4,
                label_text = 'MODIFY',
	        labelpos = 'w',
                #selectioncommand = self.changeColour,
		scrolledlist_items = modify,
        )
#	dropdown2.grid(row=2)
	dropdown11.pack(pady=5)
	# Display the first colour.
	first = modify[0]
	dropdown11.selectitem(first)
	#self.changeColour(first)


################      Delete  Combobox  in frame 4  ######################


	# Create and pack the dropdown ComboBox.
        delete = [' ', 'Course', 'Instructor', 'Room', 'Time', 'Record']
	dropdown12 = Pmw.ComboBox(frame4,
                label_text = 'DELETE',
	        labelpos = 'w',
                #selectioncommand = self.changeColour,
		scrolledlist_items = delete,
        )
	#dropdown12.component(delete).configure(width = 4)
#	dropdown2.grid(row=2)
	dropdown12.pack()
	# Display the first colour.
	first = delete[0]
	dropdown12.selectitem(first)
	#self.changeColour(first)


####################    Ok,help,quit  buttons  in farme 4   #######################

	# Create and pack the ButtonBox.
	self.buttonBox = Pmw.ButtonBox(frame4,
		labelpos = 'E',
#		label_text = 'ButtonBox:',
                frame_borderwidth = 2,
                frame_relief = 'groove')
	self.buttonBox.pack(fill = 'both', expand = 1, padx = 15, pady = 15)

	# Add some buttons to the ButtonBox.
	self.buttonBox.add('OK')#, command = self.ok)
	self.buttonBox.add('HELP')#, command = self.apply)
	self.buttonBox.add('QUIT')#, command = self.cancel)

	# Set the default button (the one executed when <Return> is hit).
	self.buttonBox.setdefault('OK')
#	frame4.bind('<Return>', self._processReturnKey)
	frame4.focus_set()

	# Make all the buttons the same width.
	self.buttonBox.alignbuttons()



################## writing days on canvas  in frame1    ########################

       def drawDays(self):
     # Make a list of days of the week

        enumDays=['Mon', 'Tue', 'Wed', 'Thu', 'Fri']

     # Draw text for each of the days

	xpos=20
	ypos=65
	ydiff=80
	for day in enumDays:
	        self.days.create_text(xpos, ypos, text=day, font=("arial", 12, "bold"))
	        ypos=ypos+ydiff


###################       Drawing Time on canvas in frame 1   ##########################################


       def drawTime(self):
     # Make a list of days of the week
        enumTime=['8.00a', '9.00a', '10.00a', '11.00a', '12.00n', '1.00p', '2.00p', '3.00p','4.00p','5.00p','6.00p','7.00p','8.00p']

     	xpos=68
	ypos=15
	xdiff=54
	for time in enumTime:
	        self.time.create_text(xpos, ypos, text=time, font=("arial", 12, "bold"))
	        xpos=xpos+xdiff

	self.time.grid(row=0, column=1)



##############    List of other canvases in frame 1    ###########################################


     # Make a list of days of the week
       def drawCanvas(self):
        CanvasNameList=['MonCanvas', 'TueCanvas', 'WedCanvas', 'ThuCanvas', 'FriCanvas']
     	xpos=68
	ypos=15
	xdiff=54
	for canvas in CanvasNameList:
		self.canvas=Pmw.ScrolledCanvas(frame1,
		borderframe = 1,
		labelpos = 'wn',
		usehullsize = 5,
		hull_width = 800,
		hull_height = 80,
        	)
		i=1
		self.canvas.grid(frame1,row=1, column=i)
		i=i+1
        	self.canvas.resizescrollregion()
 		self.canvas.configure(hscrollmode = 'dynamic')
		self.canvas.configure(vscrollmode = 'static')	                	


#########################################################################################



      	
root = Tk()
app=Schedule(root)
root.mainloop()






From martinol at algonet.se  Sat Apr  1 08:55:18 2000
From: martinol at algonet.se (Martin Olsson)
Date: 1 Apr 2000 13:55:18 GMT
Subject: HP Bug
Message-ID: <8c4v46$1v5$1@zingo.tninet.se>

About a year ago I heard about a bug only on Windows CE HP devices.
Does someone know if it is fixed yet?

/Martin Olsson



From breiter at usf.Uni-Osnabrueck.DE  Tue Apr 11 05:32:09 2000
From: breiter at usf.Uni-Osnabrueck.DE (Bernhard Reiter)
Date: 11 Apr 2000 09:32:09 GMT
Subject: Installation Problem in Windows
References: <38F28E55.67280329@swt.edu>
Message-ID: <8curep$snq$2@newsserver.rrzn.uni-hannover.de>

Rakesh,
it is hard to tell what your problem are with just the
information you gave us.

There is not really a need to "compile" python program.
Try the following to run your python program:

	c:\programs\python\python.exe test.py
and send us the results.
(Of course you have to change the paths above to the paths matching
your setup.)

Cheers,
	Bernhard
ps.: Please stop do double postings in text and html. Switch of the
HTML or chances are high that most people will ignore you on
usenet.

Ie article <38F28E55.67280329 at swt.edu>,
	rakesh <rs59922 at swt.edu> writes:

> I am new in python. I have installed python 1.5 installer for Windows at
> home. I ran the setup. I need to use teh program for both, the
> compilation of python programs and to use Tk widgets. I wrote one
> test.py and tried to compile it inthe python.exe window. It gave an
> error "stdin". Can anybody help this newbie.

> --------------14F405F271273C11DA4321F5
> Content-Type: text/html; charset=us-ascii
> Content-Transfer-Encoding: 7bit
> 
> <!doctype html public "-//w3c//dtd html 4.0 transitional//en">
> <html>
> <font face="Arial,Helvetica"><font size=-1>Folks,</font></font><font face="Arial,Helvetica"><font size=-1></font></font>

-- 
Professional Service around Free Software                (intevation.net)  
The FreeGIS Project				            (freegis.org)
Association for a Free Informational Infrastructure            (ffii.org)


From akuchlin at mems-exchange.org  Tue Apr 18 19:43:47 2000
From: akuchlin at mems-exchange.org (Andrew M. Kuchling)
Date: 18 Apr 2000 19:43:47 -0400
Subject: Starship update: April 18, #2
Message-ID: <3daeirht58.fsf@amarok.cnri.reston.va.us>

The current status of the Starship restoration effort is:

* A new disk is in the machine, and Red Hat 6.2 has been reinstalled
  from scratch.

* A temporary Web server has been put in place that explains what has
  happened.

* No e-mail server is running on the machine.

* We have found a backup of /home dating from February 1, 2000.  
  tar -ztvf reported no errors on the backup file, so it's probably 
  OK.  So, the worst case might now be that we've only lost 
  changes since February 1.  

* We haven't recreated user accounts or restored from the backup file
  yet, because we want to know the outcome of the data recovery effort
  first.  We'll decide what to do next once we know how much data can
  be recovered.  

* If you were a volunteer administrator on starship, please let me
  know; I'll add you to starship-fix at egroups.com, a mailing list for
  coordinating the resuscitation process.  Also, please send me your
  SSH public key.

Many thanks to Michael McLay for getting the new disk, reinstalling
Linux, and getting the system up again.

-- 
A.M. Kuchling			http://starship.python.net/crew/amk/
in-any-case-the-best-christmas-present-i-got-today!-ly y'rs - tim
  -- Tim Peters, 29 Dec 91 [First occurrence of Tim Peters's long-phrase-ly
       idiom.]



From tismer at trixie.triqs.com  Sun Apr 23 17:22:43 2000
From: tismer at trixie.triqs.com (Christian Tismer)
Date: Sun, 23 Apr 2000 23:22:43 +0200
Subject: 'database' search
References: <39036340.BE82BEAD@roguewave.com>
Message-ID: <390369A3.99971CA2@trixie.triqs.com>


Bjorn Pettersen wrote:
[database with 1500 to 5000 items]

> The smallest datastructure I've found so far, is a 'prefix dictionary',
> so if e.g. "aardwark" was in the database, the 'prefix dictionary' would
> contain:
> 
>  pfdict = {
>   'a':        ['aardwark', ...],
>   'aa':       ['aardwark', ...],
>   'aar':      ['aardwark'],
>   'aard':     ['aardwark'],
>   'aardw':    ['aardwark'],
>   'aardwa':   ['aardwark'],
>   'aardwar':  ['aardwark'],
>   'aardwark': ['aardwark']
>  }
> 
> The maximum size for this would be n * avg len of item, with the real
> size being much(?) smaller with sufficient prefix sharing...

Well, your database is small enough to stand this number
of generated dict items.
In the general (larger) case I'd suggest to use a "trie"
structure. This is dividing your key set into maximum
unique subkeys. This may be coded with nested dictionaries
quite nicely.

Example:
aardwark      1
aaron         2
advertizing   3
advantage     4
python        5

gives you the following structure:

trie = {
  "a" : {
          "a":  {
                  "rdwark" : 1
                  "ron"    : 2
                }
          "dv": {
                  "ertizing": 3
                  "antage"  : 4
                }
        }
  "p" : {
          "ython": 5
        }
}

This should tend to give you quite small dicts, and relatively
fast access. Whenever you end up in a unique subkey, you may
do the completion. Using keys() and then sorting out isn't
too slow. Preofix sharing is there, too.

ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaunstr. 26                  :    *Starship* http://starship.python.net
14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     where do you want to jump today?   http://www.stackless.com



From zorro at zipzap.ch  Fri Apr 28 04:50:51 2000
From: zorro at zipzap.ch (Boris Borcic)
Date: Fri, 28 Apr 2000 10:50:51 +0200
Subject: The Simple Economics of Open Source
References: <B52E8CEC.5222%raffael@mediaone.net>
Message-ID: <390950EB.8213862E@zipzap.ch>

Raffael Cavallaro wrote:
> 

> 
> If they produce software that:
> 
> 1. cannot be easily reverse engineered
> 
> 2. achievs sufficient market penetration that interoperability with others
> becomes a significant feature
> 
> then yes, they are effective monopolies. This is *precisely* how MS got
> where they are.

billing gates.

BB


From mdefreitas at sikorsky.com  Fri Apr 28 16:14:22 2000
From: mdefreitas at sikorsky.com (mdefreitas at sikorsky.com)
Date: Fri, 28 Apr 2000 20:14:22 GMT
Subject: MORE INFO
References: <8ea0ao$fro$1@nnrp1.deja.com> <8ea9uu$rh0$1@nnrp1.deja.com> <8ecad6$bun$1@ites.inria.fr>
Message-ID: <8ecref$lck$1@nnrp1.deja.com>


> Do you want to run several python scripts (and several ui.commands
> C code) concurently (how would you synchronize?) ?

I do not want to run concurrently, I just want to run in one thread of
execution, but recursively.

> Note that if all you want is to have C code that calls Python
> code that calls C code (through a module) that calls Python code
> without threads, PyRun_XXX functions are re-entrant.

Thanks! That's what I needed to know. I now modified (and simplified)
my code as follows:

void interp(char *script) {
   static int nest_level = 0;
   FILE *fd = fopen(script, "r");
   if (nest_level == 0) {
      Py_Initialize();
      PyRun_SimpleString("import ui\n"); // import my ui
   }
   nest_level++;
   PyRun_SimpleFile(fd, script);
   nest_level--;
   if (nest_level == 0) Py_Finalize();
}

The only problem is that the nested script that is run "remembers" the
variable settings from it's caller. Is there a way to give each nested
python script a fresh, clean environment? The nested script should not
effect the environment of it's caller either. Is that what the global
and local dictionaries for the API function PyRun_File are for? If so,
is there any examples on their usage. The C/API doc isn't all that
explicit.

Thanks for the help.



Sent via Deja.com http://www.deja.com/
Before you buy.


From chafik at nevrax.com  Tue Apr  4 04:12:08 2000
From: chafik at nevrax.com (Sameh chafik pro)
Date: Tue, 4 Apr 2000 10:12:08 +0200
Subject: Tr: builtding library .pyd file format
Message-ID: <010f01bf9e0d$78af96c0$1001a8c0@nevrax.net>

Hello evry body,
I want to know how to do to built a .pyd library file format form the windows .dll file format
thanks



-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20000404/bdd11787/attachment.html>

From emile at fenx.com  Fri Apr 28 20:30:27 2000
From: emile at fenx.com (emile at fenx.com)
Date: Fri, 28 Apr 2000 17:30:27 -0700
Subject: Reading in strings -> numbers ??
Message-ID: <390A2D23.7029E557@fenx.com>

If you're trying to put the info on disk to read it
in again later, look at eval(repr()), eg:

>>> a,b,c = eval(repr((1.2, -3.5, 10)))
>>> a,b,c
(1.2, -3.5, 10)
>>> 

Emile van Sebille

Louis M. Pecora <pecora at anvil.nrl.navy.mil> wrote in message
news:<280420001700308281%pecora at anvil.nrl.navy.mil>...
> Is there a Python "read-partner" for the write(string) function, e.g., 
> 
> fileobject.write(" %e %e %d" % (1.2, -3.5, 10))  
> 
> in the sense that the C functions fprintf and fscanf are partners.
> 
> I've looked in 4 different Python books and in the pdf references, but
> I see nothing in read or readlines or strings that seems to make for an
> easy conversion from a string to a list or tuple of objects (numbers). 
> In other words, I want the opposite of the write above so I can read
> those numbers back into my Python code.  
> 
> Thanks for any help.
> -- 
> http://www.python.org/mailman/listinfo/python-list
>



From mjackson at wc.eso.mc.xerox.com  Wed Apr 26 14:39:34 2000
From: mjackson at wc.eso.mc.xerox.com (Mark Jackson)
Date: 26 Apr 2000 18:39:34 GMT
Subject: How can I get the name of an object???
References: <390731A2.12788170@freei.com>
Message-ID: <8e7d56$t01$1@news.wrc.xerox.com>

Sean Blakey <sblakey at freei.com> writes:
> Peter,
> Python objects do not have any way of finding out what name they are
> referenced by.  Names refer to objects, but objects do not have name.
> 
> This is reasonable, since it is simple for an object to have 0, 1, or
> more names.  Imagine what an imaginary __name__ attribute would be in
> the following cases:
> >>>class A:
> ....  pass
> ....
> >>>A().__name__		# What should this be? ''?
> >>>a = A()
> >>>b = a		# b and a are now two names for the same object
> >>>b.__name__		# Should this be 'a', 'b', or ['a', 'b']?
> 
> There are some clever hacks to attempt this (like iterating through all
> the names in globals() looking for a match with the object), but I have
> not seen any that are completely effective.

Right.  Consider also the following:

>>> a=999
>>> b=999
>>> c=b
>>> a is b
0
>>> b is c
1

All fine; c and b are the same, a just "happens" to be equal.  But

>>> c=1
>>> d=1
>>> c is d
1

Because of implementation details (driven by considerations of
efficiency), there will be instances of "accidental" identity.  Unless
the objective is tolerant of this, any scheme to find *the* name of an
object would seem to be doomed.

-- 
Mark Jackson - http://www.alumni.caltech.edu/~mjackson
	An advertiser will happily make you feel bad about
	yourself if that will make you buy, say, a Bic pen.
				- George Meyer




From HighdesertmanREMOVE at yahoo.com  Sat Apr 22 01:57:20 2000
From: HighdesertmanREMOVE at yahoo.com (Highdesertman)
Date: Fri, 21 Apr 2000 22:57:20 -0700
Subject: Distribution
Message-ID: <39013f48@flexnet239.uunt.net>

Hi all

I am a newbie to python and haven't done any programming for some time. The
last time I did any amateur programming was back when MSDOS was (more or
less) king and Quick C was a popular version of that language.

Well things have changed and I am older. I have decided that after almost
ten years of programming inactivity, I would try my hand at it again. I
decided that I would be careful about the language I would learn (I would
have to relearn everything even if I used a language I used to be proficient
at such as compiled basic or C) and I also decided that I simply wasn't
interested (at this juncture) in getting back in the C saddle. For a variety
of reasons.

I wanted a language that was simple to use, powerful, portable and fast. I
also was looking to minimize the learning curve. I stumbled over Python
quite by accident (it was recommended in a rather obscure section of the
ZDTV website) and it seems to meet these initial requirements.

Now to my question:

If I use python to develop an application for the windows 9x environment,
 I have already downloaded the appropriate libraries and version) what form
does the finished program take? How does it get distributed? I remember in
the early days of BASIC programming, (before compiled basic) your program
was just raw code that could be (often accidentally) altered or otherwise
corrupted. In order for your basic program to be used, the user had to have
a basic compiler on board and they loaded your raw source from a text file
and executed it. Is this how Python is used? Forgive me for asking what must
be a terribly green question, but I am really curious about the final
product of a python program. In what way does the end user utilize your
code? If you should be writing to an unknown audience as in the case of
making a freeware utility (you discover that that little ditty you wrote to
solve a headache really is very useful to others) Is there any way to turn
your python code into an exe that can be easily utilized by others? If
Python is dependent on it's interpreter to run, and I am only programming
for my own needs, then that dosen't represent a weakness. But since Python
remains relatively unknown as yet, the average user isn't likely to have a
python interpreter handy. In addition, one of the real advantages of the
development of the compiler was that it gave your code a level of protection
against being tampered with and made it easily useable by a wide audience.
Python looks very elegant and seems to meet my needs with the exception of
this unknown. If anyone could help me answer this question, I would be very
much in thier debt. I am sure this need has been addressed, and I would just
like to know what the solution is.


Cheers,

Mathew




From kc5tja at garnet.armored.net  Sun Apr 30 09:03:12 2000
From: kc5tja at garnet.armored.net (Samuel A. Falvo II)
Date: 30 Apr 2000 13:03:12 GMT
Subject: Do I always have to write "self." ?
References: <270420001706386648%pecora@anvil.nrl.navy.mil> <slrn8ghc6t.f01.scarblac-spamtrap@flits104-37.flits.rug.nl> <8ecsgj$6lp$1@slb0.atl.mindspring.net> <m3bt2tonyg.fsf@atrus.jesus.cam.ac.uk> <300420001008012011%pecora@anvil.nrl.navy.mil><20000430170952.A502@datapro.co.za> <B531EC9B.57C1%abrahams@mediaone.net>
Message-ID: <slrn8gp7ei.ilf.kc5tja@garnet.armored.net>

In article <B531EC9B.57C1%abrahams at mediaone.net>, Dave Abrahams wrote:
>Alternatively, if you're writing a large program with many abstractions and
>think lots of other people will have to read it steer clear of Hungarian
>Notation in either form ;)

I disagree.

>coding in Python. Who cares if it's a dict or a UserDict, or some other
>object which models the appropriate mapping semantics?

If it's a dict, or if it's a UserDict, or if it's anything else that exports
the dict interface, then it's perfectly valid to name a parameter
dictWhatever.  I fail to see what's wrong with this concept.

-- 
KC5TJA/6, DM13, QRP-L #1447
Samuel A. Falvo II
Oceanside, CA


From mal at lemburg.com  Tue Apr  4 05:42:08 2000
From: mal at lemburg.com (M.-A. Lemburg)
Date: Tue, 04 Apr 2000 11:42:08 +0200
Subject: VMware/Win4Lin/Wine + Linux... and Python?
References: <38E96834.B500B786@acm.org>
Message-ID: <38E9B8F0.6799A70F@lemburg.com>

"Robert W. Cunningham" wrote:
> 
> With at least three ways to run Windows applications simultaneously with
> Linux applications, it seems to me there may be a need and/or
> opportunity to improve the connectivity and interface between the two
> worlds.  And I suspect Python may be just the tool to do it.
> 
> ...
>
> But is there a way to get instances of Python on both platforms talking
> to each other?  Of course, a socket could be opened and data shuttled
> back and forth, but that will be fairly slow and demand a fair amount of
> overhead.  Has there been any work done on a higher-level
> Python-to-Python interface?

You could use shared memory for the task. What methods are
implementable really depends on the architecture used by the
different WinXX-emulators. I would suspect Unix domain sockets
to be a fairly portable and fast solution, but then WinXX only
supports TCP/IP sockets. Pipes might also work (I think OS/2
used these for the communication  OS/2<->DOS and OS/2<->Win3.1).

> ...
>
> I have already dangled the idea before Trelos, the makers of Win4Lin,
> and they are definitely excited by the possibilities.  Plus, with
> Win4Lin being a commercial product, an inter-OS Python interface would
> allow Trelos to expose interfaces without having to give away any of
> their code (or revenue).  And this is one group that sure seems to be
> earning their money, even given their unfortunate lack of support for
> Open Source.  Python may well be the wedge needed to open that door.

This idea seems to be getting common around commercial
vendors: they wrap their internal C API using SWIG or
something similar and then publish the Python API for others
to hook into. As a nice side-effect this also has the potential
of making their application scriptable.

In any case: the idea is well worth looking into.

-- 
Marc-Andre Lemburg
______________________________________________________________________
Business:                                      http://www.lemburg.com/
Python Pages:                           http://www.lemburg.com/python/





From mhammond at skippinet.com.au  Wed Apr 19 22:20:48 2000
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Thu, 20 Apr 2000 02:20:48 GMT
Subject: Pythonwin?
References: <eMz2Nmiq$GA.73@net025s>
Message-ID: <4QtL4.24$v85.41@news-server.bigpond.net.au>

"A[r]TA" <arta at x-stream.nl> wrote in message
news:eMz2Nmiq$GA.73 at net025s...
> Hello,
> down here,
> it doesn't shows a Window or something and I don't get an error.
>
>     window.Wnd()

You have created a Window object, but not the Window.  You need something
like:

w = window.Wnd()
w.CreateWindow(...)

It is simpler to start with a dialog - d=win32ui.CreateFileDialog();
d.DoModal()

You are really going to need to go through an MFC tutorial or book before
you can make heads or tails of Pythonwin from that level...

Mark.





From bjorn at roguewave.com  Wed Apr  5 12:19:36 2000
From: bjorn at roguewave.com (Bjorn Pettersen)
Date: Wed, 05 Apr 2000 10:19:36 -0600
Subject: Pass by reference ?
References: <38E864C9.E1EE94F7@ecs.soton.ac.uk> <38E8D449.FBC1E2AD@roguewave.com> <8caohp$j4v$1@pegasus.csx.cam.ac.uk>
Message-ID: <38EB6797.F03EEFB3@roguewave.com>

Nick Maclaren wrote:
> 
> In article <38E8D449.FBC1E2AD at roguewave.com>,
> Bjorn Pettersen  <bjorn at roguewave.com> wrote:
> >Jacek Generowicz wrote:
> >>
> >> I'm just trying to familiarize myself with Python.
> >> In Magnus Lie Hatland's Instant Python, he
> >> mentions that `all parameters in Python are passed
> >> by reference'.
> >
> >The technically correct (although much less known) term is
> >pass-by-object-reference.  All this means is that you can change an
> >argument to a function by calling methods on the object, but you can't
> >change the identity of the object (ie. simply calling a function can't
> >change any name-bindings in your current scope).
> 
> Hang on.  Why is that technically correct?  Call by reference has
> never implied changing any bindings in the caller's scope in any
> work that I have seen - it certainly didn't back in the 1960s.
> Unless I am thoroughly confused, Python uses bog-standard call by
> reference, with the wrinkle that assignment is a rebinding rather
> than a copying operation.  AM I confused about that?

Call by reference is (at least now) generally used to refer to Pascal
VAR arguments or C++ & arguments, both of which has similar semantics
which allow rebinding in the caller's scope.  This is in contrast to
many OO languages where arguments can be mutated but not rebound...  I'm
not sure if call-by-reference has changed meaning since the 60s, but
pass/call-by-object-reference is the term consistently used when I was
doing my type theory work in 85-87...  

> The form of argument passing that DOES involve rebinding the
> caller's names certainly exists, but I have never seen an agreed
> name for it.  It was reckoned to be a bad idea (just like Fortran's
> equally deceptive argument passing of scalars) by 1970, but kept
> coming back in special-purpose languages.

Just because something is a bad idea doesn't mean you can't get a
research grant for it <wink>

--bjorn



From pinard at iro.umontreal.ca  Wed Apr  5 12:35:42 2000
From: pinard at iro.umontreal.ca (=?ISO-8859-1?Q?Fran=E7ois_Pinard?=)
Date: 05 Apr 2000 12:35:42 -0400
Subject: Pass by reference ?
In-Reply-To: "Robert W. Cunningham"'s message of "Wed, 05 Apr 2000 15:10:13 GMT"
References: <38E864C9.E1EE94F7@ecs.soton.ac.uk> <38EAF8D8.4B7C43B4@ecs.soton.ac.uk> <38EB56AE.F938CE06@acm.org>
Message-ID: <oqbt3oh5ep.fsf@titan.progiciels-bpi.ca>

"Robert W. Cunningham" <rcunning at acm.org> writes:

> In my own mind, to escape the ambiguity of how "Pass By Reference" and
> "Pass By Value" are discussed in Computer Science, and how those same
> terms are (mis)used when discussing Python, I've been using the idea of
> "Pass By Clone".

Just a little story, popping into my mind while reading this.

Niklaus Wirth, whose name is attached to Pascal, was asked by a conference
attendee how his name should be pronounced.  He replied: "If you call me by
name, you should say Veert.  But if you call me by value, rather say Worth."

-- 
Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard





From mikael at isy.liu.se  Wed Apr  5 03:00:08 2000
From: mikael at isy.liu.se (Mikael Olofsson)
Date: Wed, 05 Apr 2000 09:00:08 +0200 (MET DST)
Subject: Why should I switch to Python? - Infinity of Primes
In-Reply-To: <1257244378-5884705@hypernet.com>
Message-ID: <XFMail.000405090008.mikael@isy.liu.se>

On 04-Apr-00 Gordon McMillan wrote:
 >  Oh, come on. There are at least 3 well known constructive 
 >  proofs that all odd numbers are prime:
 >  
 >  The mathematician's:
 >   3 is prime,
 >   5 is prime,
 >   7 is prime
 >   as so on, by induction.
 >  
 >  The physicist's:
 >   3 is prime,
 >   5 is prime,
 >   7 is prime,
 >   9 is prime <experimental error>
 >   11 is prime...
 >  
 >  The computer scientist's:
 >   3 is prime,
 >   3 is prime,
 >   3 is prime ....

The last one was new to me. But I miss the following one:

The engineer's:
 3 is prime,
 5 is prime,
 7 is prime,
 9 is prime ...

For-yers-ugo-I-coultnt-evun-spel-injunjeer-an-now-I-are-one-ly y'rs

/Mikael

-----------------------------------------------------------------------
E-Mail:  Mikael Olofsson <mikael at isy.liu.se>
WWW:     http://www.dtr.isy.liu.se/dtr/staff/mikael
Phone:   +46 - (0)13 - 28 1343
Telefax: +46 - (0)13 - 28 1339
Date:    05-Apr-00
Time:    08:55:41

This message was sent by XF-Mail.
-----------------------------------------------------------------------



From effbot at telia.com  Mon Apr 17 15:05:46 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Mon, 17 Apr 2000 19:05:46 GMT
Subject: import disappointment
References: <002b01bfa89d$834c49d0$4500a8c0@thomasnb>
Message-ID: <egJK4.376$wYl.186843136@newsb.telia.net>

Thomas Heller <thomas.heller at ion-tof.com> wrote:
> Using imputil is out of question because most programs
> including pythonwin are only expecting strings in sys.path.

I suggest taking another look at imputil -- unless it
has changed a lot since I last downloaded it, it's per-
fectly possible to use it without stuffing non-strings
in sys.path...

</F>




From kens at sightreader.com  Thu Apr  6 16:34:46 2000
From: kens at sightreader.com (Ken Seehof)
Date: Thu, 06 Apr 2000 13:34:46 -0700
Subject: A Mountain of Perl Books + Python Advocacy
References: <000f01bf9eca$2eec8c60$3e2d153f@tim>
Message-ID: <38ECF4E6.B85C62D5@sightreader.com>

Indeed.  I've noticed that while just about everyone who I've known who
has tried python ends up liking it alot, there aren't many fanatics.  I define
a fanatic as someone who is absolutely certain that their language is by far
the best language for every possible application, and that other languages
don't even deserve to exist.  Python programmers tend to acknowledge
the positive qualities of other languages.

Perhaps this is partly because python borrows all the best stuff from many
other languages rather than being built around some wild, revolutionary idea.


Tim Peters wrote:

> [snip...]

> Python *is* astonishing for the tiny number of evangelists it's attracted!
> We sucked in Eric Raymond, for but for some reason he doesn't seem to
> consider promoting Python to be his Primary Mission from God -- the best
> hope for a PSA marketing campaign may be to hire a deprogrammer to
> straighten out Tom Christiansen's brain while there's still something left
> <wink>.
>
> not-mentioning-that-perl-is-a-better-language-for-some-tasks-and-
>     that-tells-you-how-effective-an-advocate-i-am-ly y'rs  - tim






From python-list at teleo.net  Fri Apr  7 18:45:29 2000
From: python-list at teleo.net (Patrick Phalen)
Date: Fri, 7 Apr 2000 15:45:29 -0700
Subject: calendar/scheduling app
In-Reply-To: <38EE332E.2954EA49@fizzylab.com>
References: <38EE332E.2954EA49@fizzylab.com>
Message-ID: <0004071547040L.01234@quadra.teleo.net>

[paul beard, on Fri, 07 Apr 2000]
:: I am interesting in having a meeting scheduler/calendar app that does
:: what something like Exchange purports to do without requiring me to use
:: a client that doesn't exist. In other words, it should be
:: platform-agnostic and use webforms/cookies to handle transactions and
:: display options. 
:: 
:: Possible in Zope? How hard? How long to do, given a baseline knowledge
:: store of zero?


Try http://www.worldpilot.com



From jepler.lnk at lnk.ispi.net  Sat Apr 22 22:31:23 2000
From: jepler.lnk at lnk.ispi.net (jepler epler)
Date: Sun, 23 Apr 2000 02:31:23 GMT
Subject: Python for drivers?
References: <3900EDBE.282BC85@dsuper.net>
Message-ID: <slrn8g4gbn.33l.jepler.lnk@potty.housenet>

On Sat, 22 Apr 2000 01:10:04 GMT, Dave
 <redsand at dsuper.net> wrote:
>Can Python be used for writing device drivers? I just got BeOS 5 and it
>doesn't support some video cards. I'd like to write one for the Cirrus
>Logic card and I want a language that's clear and concise- Python. I've
>never written device drivers. Any suggestions?
>THanks

If you can use a tool like SWIG (or a hand written C module) to provide
access to the necessary low-level features, you may be able to do this.

For instance, if the card needs the idea of 'store {8,16,32}-bit integer X at
index L of the video card's address space' and 'read {8,16,32}-bit
integer at index L of the video card's address space', then you'd just
need to write those following functions in a C module for Python to call.

It may be slow, as other posters suggest.

Jeff


From timehorse at unforgettable.com  Sat Apr 15 21:43:42 2000
From: timehorse at unforgettable.com (timehorse at unforgettable.com)
Date: Sun, 16 Apr 2000 01:43:42 GMT
Subject: Python Starship Grounded??
References: <38F6B2DF.940E7C23@concentric.net> <OqQJ4.8866$fV.715336@bgtnsc05-news.ops.worldnet.att.net>
Message-ID: <8db5sb$gsc$1@nnrp1.deja.com>

In article <OqQJ4.8866$fV.715336 at bgtnsc05-news.ops.worldnet.att.net>,
  "Michael Appelmans" <mappel at 123ibm.net> wrote:
> Yes, I was hoping to get hold of win32all build 131 for Python 1.6a2
but to
> no avail. I fear it is lost in some freakish nebulae.

Not that I'm complaining as I only wish I was doing more to promote the
Starship's upkeep but it is rather annoying this as I route all my
personal e-mail through Starship because it's currently my only shell
a/c and thus the only place I can have a true, bona fide .forward which
will send a copy of my mail to @home and a copy to my work which runs
on a Micro$oft network.  I don't store my mail on Starship, only route
through it, but obviously with Starship down, I haven't received any e-
mail in days -- not necessarily a bad thing what with the advent of
Spam.  :)  Still, I average about 10 or at MOST 20 e-mails a day and
since I am only re-routing them, I seriously doubt I'm degrading the
system greatly (we could always install ORBS if it'll cut down on the
spam I receive -- no complaints from me about that at all! :)  :)  :)
Of course, Zope being down is a bit of a bug!

Anyway, the long and the sort of it is, I'm not asking for it's timely
recovery -- it can recover when it recovers -- but it would be nice if
we could receive some kind of status report on the situation, non?

Be Seeing You,

Jeffrey.


Sent via Deja.com http://www.deja.com/
Before you buy.


From wrd at awenet.com  Mon Apr 10 15:43:39 2000
From: wrd at awenet.com (William R. Dickson)
Date: Mon, 10 Apr 2000 12:43:39 -0700
Subject: Problem running Python script as cron job
References: <wrd-0AEDDA.12282507042000@news.awenet.com> <0004071555090M.01234@quadra.teleo.net>
Message-ID: <wrd-1E3688.12433910042000@news.awenet.com>

In article <0004071555090M.01234 at quadra.teleo.net>, Patrick Phalen 
<python-list at teleo.net> wrote:

> [William R. Dickson, on Fri, 07 Apr 2000]
> :: Anybody know of any issues with running python scripts from Cron?
> 
> No. I've had multiple Python scripts running every one, two, five and
> ten minutes around the clock for years on multiple servers without
> problems.
> 
> I'd suspect you might have a bug lurking.

I think so too, but I can't find it, because the script runs flawlessly 
every time I run it by hand, and the buffer never flushes the output to 
the log when it's run by Cron.

Blorg.

-Bill


From javanet at dynacap.com  Fri Apr 28 13:23:38 2000
From: javanet at dynacap.com (Arnaldo Riquelme)
Date: Fri, 28 Apr 2000 09:23:38 -0800
Subject: Great!!
Message-ID: <sgji3u4i7qo109@corp.supernews.com>

Thanks a lot for all your help guys!!!

Python makes my life a lot easier


Arnaldo


later




From richard_chamberlain at ntlworld.com  Wed Apr  5 06:13:00 2000
From: richard_chamberlain at ntlworld.com (Richard Chamberlain)
Date: Wed, 5 Apr 2000 11:13:00 +0100
Subject: Remotely running a python script
Message-ID: <8cf452$pdv$1@gxsn.com>

Hi,

Does anyone know a way I can remotely trigger a python script on a Win95
machine?

What I want to do is trigger an event from a vb client and run a python
script on another PC in the business.

They are able to see each other via IP and the client has a drive mapped
that is a shared folder on the python side.

Because I'm using Win95 I'm not sure how good the DCOM support is.

Thanks

Richard






From mhammond at skippinet.com.au  Thu Apr 20 09:47:41 2000
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Thu, 20 Apr 2000 13:47:41 GMT
Subject: win32all-131 at python.org
Message-ID: <1UDL4.539$v85.2477@news-server.bigpond.net.au>

Just a quick update.  win32all-131 for Python 1.6a2 can be found at
http://www.python.org/1.6/

Note that this version has 2 known (and semi-serious) bugs:
* The Pythonwin debugger simply doesnt work (but its such a simple bug!)
* A Unicode issue with Scintilla prevents calltips and multi-line commands
in the Pythonwin interactive window(!!) working correctly on NT/2000.

I expect to have 132 available soon with these fixes plus a few other
tweaks.  If starship is still not available, it will also be at
http://www.python.org/1.6/

Finally, once this win32all set stabilizes, I will also be releasing
another Python 1.5 build.  Depending on what happens, this may or may not
be the last 1.5 build, but most new builds are likely to target 1.6
(although "popular demand" is likely to have some influence ;-)

Mark.




From mwh21 at cam.ac.uk  Thu Apr 27 09:03:57 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 27 Apr 2000 14:03:57 +0100
Subject: when to use intern()
References: <Pine.LNX.4.21.0004270820120.27672-100000@saturn.ispol.com>
Message-ID: <m37ldjpuf6.fsf@atrus.jesus.cam.ac.uk>

Gregory Trubetskoy <grisha at ispol.com> writes:

> It says in the manual that some things in Python get interned
> automatically. I'm not sure where intern() is necessary - 

Almost never.
 
> if I have a statement like:
> 
> if self.config.has_key("PythonOption"):
>    ...
> 
> that gets executed a lot - should I consider interning "PythonOption" or

Well, you can, but...

> is it done automatically?

Yes.

> 
> what about
> 
> PYTHON_OPTION = "PythonOption"
> if self.config.has_key(PYTHON_OPTION):
>    ...
> 
> is this interned?

Yes.

See Python/compile.c in the source distribution round about line 240
for the gory details...

HTH,
Michael

-- 
48. The best book on programming for the layman is "Alice in
    Wonderland";  but  that's because it's  the best book on 
    anything for the layman.
     -- Alan Perlis, http://www.cs.yale.edu/~perlis-alan/quotes.html


From muzzetta at liotro.dipmat.unict.it  Wed Apr 12 12:39:02 2000
From: muzzetta at liotro.dipmat.unict.it (Alessandro Muzzetta)
Date: 12 Apr 2000 16:39:02 GMT
Subject: Looking for BoboHTTPServer.py
Message-ID: <8d28r6$9bs$2@news.cineca.it>

Hello,

I'm looking for BoboHTTPServer.py but cannot find it anywhere.
Can someone tell me where I can find it, or some other server
that integrates Bobo (not Medusa).

Thanks



From smalleys at gte.net  Sat Apr 15 14:38:05 2000
From: smalleys at gte.net (sue)
Date: Sat, 15 Apr 2000 18:38:05 GMT
Subject: need help just begining
References: <11822ca0.dd34a400@usw-ex0109-068.remarq.com> <38F41190.C0BB507B@sympatico.ca> <38F53E5A.977913FE@cosc.canterbury.ac.nz> <38F54D83.B0F620A8@callware.com> <38F69621.3B8F5B29@gte.net>
Message-ID: <38F8A912.B05575D4@gte.net>

My local Borders said "two weeks" but it is orderable.
Sue.


From jbauer at rubic.com  Mon Apr 17 16:37:56 2000
From: jbauer at rubic.com (Jeff Bauer)
Date: Mon, 17 Apr 2000 15:37:56 -0500
Subject: Playing wave files on Windows
References: <38FB6ED7.4C72773E@callware.com>
Message-ID: <38FB7624.943A8C35@rubic.com>

> Ivan Van Laningham wrote:
> How can I play a sound file on Windows. 

Ivan, use the built-in winsound module, e.g.:

>>> import winsound
>>> winsound.PlaySound(r'C:\WINNT\Media\tada.wav', winsound.SND_FILENAME)

> I don't want to do anything else, just play a 
> wave file. ...

Then I can't interest you in my Python interface to
the Diamond Rio MP3 player?  <0.5 wink>

Jeff Bauer
Rubicon Research



From pinard at iro.umontreal.ca  Fri Apr 28 16:37:12 2000
From: pinard at iro.umontreal.ca (=?ISO-8859-1?Q?Fran=E7ois_Pinard?=)
Date: 28 Apr 2000 16:37:12 -0400
Subject: urllib redirect
In-Reply-To: Georg Mischler's message of "Wed, 19 Apr 2000 10:34:36 GMT"
References: <4.1.20000418125557.030592d0@202.95.1.37>
	<Pine.OSF.4.20.0004181610230.20942-100000@azure.dstc.edu.au>
	<8dk23t$4c2$1@nnrp1.deja.com>
Message-ID: <oq66t27yiv.fsf@titan.progiciels-bpi.ca>

Georg Mischler <schorsch at schorsch.com> writes:

> Anthony J Wilkinson wrote:

> > If you just print out the following you can use the browser's own
> > redirection facility.
> >
> > <html>
> >     <head>
> >         <meta HTTP-EQUIV="REFRESH" CONTENT="0; URL=http://www.spam.com/">
> >         <title> redirect page</title>
> >     </head>
> >     <body>
> >         If your browser does not support redirection, click
> >         <a HREF="http://www.spam.com/">here</a> to continue.
> >     </body>
> > </html>

> This can be done in a much simpler way, whitout sending any
> actual page data to the client:

>   your_new_path = 'http://www.spam.com/'

>   print 'Status: 302 Redirected'
>   print 'Location: %s' your_new_path

> As soon as the browser (works with *every* browser!) sees the
> "Location:" header, it will stop and try fetching the new page.

The advantage of the HTML code is that it is mere HTML, so usable wherever
an HTML page was used.  Even if the second solution is simpler, it is only
usable through a CGI script, that is, from a location allowing execution
of CGI scripts.  It happened to me that this is exactly _because_ the
original is at a place forbidding CGI scripts, that I need page relocation.
So the original HTML solution is worth being noted, at least for some cases.

-- 
Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard




From gslee111 at daisy.kwangwoon.ac.kr  Wed Apr 26 20:59:33 2000
From: gslee111 at daisy.kwangwoon.ac.kr (Gang Seong Lee)
Date: Thu, 27 Apr 2000 09:59:33 +0900
Subject: defining __getitem__
Message-ID: <WiMN4.1$7C.101@news.hananet.net>

I'd like to have a method __getitem__ in a class which acts like other
sequence type.
For example,
------------------------
class data:
    def __init__(self, n):
        self.maxsize = n

    def __getitem__(self, k):
        if k >= self.maxsize:
              # I want to stop here
              # ????
        return get_data_from_some_where_else(k)

d = data(10)
for el in d:
    print el
------------------------
I want to make it stop at some point automatically, like range() does. I
don't want to use try: exception clause for this. Is there any way for this?

Thanks

Gang Seong





From hnowak at cuci.nl  Mon Apr 24 17:21:20 2000
From: hnowak at cuci.nl (Hans Nowak)
Date: Mon, 24 Apr 2000 23:21:20 +0200
Subject: Installing Python on NT
In-Reply-To: <8e2b92$9n6$1@nnrp1.deja.com>
Message-ID: <200004242122.e3OLMIG24743@hera.cuci.nl>


On 24 Apr 00, at 20:36, Lee Malatesta wrote:

> I'm interested in installing Python on a properly battened down NT
> workstation for which I do not have administrative priveledges.  This is
> my workstation at work.  Mostly I just want to play around with
> SGMLtools which is largely written in Python.  The installer told me
> that I have to install logged in as administrator.  This is the first
> program I've installed where this was necessary.  Unfortuntely for me to
> install this as admin, I'd have to put in a request to PC support, they
> would have to review to see if there is a need for me to have Python
> installed and then send one of the PC support personelle out to install
> for me.
> 
> Anyone know of anyway I can circumvent the need to be root to install?

I didn't test this, but methinx you should be able to just take the files 
(exe, standard library) and copy them somewhere on your computer. I don't 
know if everything will work this way, esp. the Tkinter stuff. If all else 
fails, you could install a DOS version, which lacks some stuff (GUI, 
sockets and such) but which is fully functional otherwise.

Blessed be,

--Hans Nowak (zephyrfalcon at hvision.nl)
Homepage: http://www.hvision.nl/~ivnowa/newsite/
Python questions? See http://tor.dhs.org/~zephyrfalcon/snippets
You call me a masterless man. You are wrong. I am my own master.



From gmcm at hypernet.com  Sat Apr 29 14:30:11 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Sat, 29 Apr 2000 14:30:11 -0400
Subject: The Simple Economics of Open Source
In-Reply-To: <MPG.1374c41fa6c09b7198970f@news.earthlink.net>
Message-ID: <1255089462-10343171@hypernet.com>

Richard Hale wrote:

> In article <1255108931-9171946 at hypernet.com>, gmcm at hypernet.com
> says... > If that were the model, we'd all be using Apples. MS
> got where > they are by outmanuvering IBM and DR. If Gary Kildall
> had > been willing to play hardball with a bunch of IBM suits,
> the > world would be quite different. 

> You must mean "play
> ball".

As Jeff Bauer forwarded to me:
http://www.forbes.com/forbes/97/0707/6001336a.htm

The point being (in this rather deranged thread) that it was all 
business, and IP had nothing to do with it.

> CP/M was the OS "monopoly" of its day.  A perfect illustration of
> how success breeds failure and how every trend has within it the
> seeds of its own destruction.
> 
> It must be Microsoft's turn now!?

Hmmm. Despite being fat, MS remains surprisingly hungry.

- Gordon



From just at letterror.com  Wed Apr 12 07:44:06 2000
From: just at letterror.com (Just van Rossum)
Date: Wed, 12 Apr 2000 12:44:06 +0100
Subject: Classes, Inheritance - Stupid lazy question
In-Reply-To: <8d1ihl$q244@imsp212.netvigator.com>
References: <8d1cfh$hme7@imsp212.netvigator.com>
 <874s9766r6.fsf@ifm.uni-kiel.de> <m3itxnis32.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <l03102803b51a10e9fb9e@[193.78.237.130]>

"Michael Hudson" <mwh21 at cam.ac.uk> wrote in message
> another option is:
>
> class paul(lazy):
>     super = lazy
>     def __init__(self, name):
>         self.super.__init__(self,name)
>         self.april_pay = 0
>
>     def fired(self):
>         print "Here's a box, collect the things from your desk"
>         self.super.fired(self)

At 6:18 PM +0800 12-04-2000, Paul Gresham wrote:
>I like this solution a lot, it makes a lot of sense. Thanks Michael.
>I guess somewhere in Python is a defined route back to the super class, on a
>better day I may just dig through the code.

Actually, the above is flawed, too:

class bob(paul):
    super = paul  # possibly needed for *other* method overrides...

b = bob()
b.fired()  # <-- kaboom!

It's really best to call the overridden method explicitly.

Just





From sabren at manifestation.com  Sun Apr 16 20:59:26 2000
From: sabren at manifestation.com (Michal Wallace (sabren))
Date: Sun, 16 Apr 2000 20:59:26 -0400 (EDT)
Subject: module naming? 
In-Reply-To: <200004170054.KAA14097@piglet.dstc.edu.au>
Message-ID: <Pine.BSF.4.10.10004162056040.25069-100000@chi.pair.com>

On Mon, 17 Apr 2000, David Arnold wrote:

> -->"Michal" == sabren  <Michal> writes:
> 
>   Michal> Is there a better way to do this?
> 
> what about something like
> 
>    from Record import Record
>    from Field import Field
>    from IdxDict import IdxDict
> 
> in your zdc/__init__.py ?
> 

Hey David,

  That's certainly a lot cleaner! But is there a way to do that and
not have it load each one of those files every time? There's actually
more than 3, and even though they're all fairly small, I'd rather not
have the overhead of loading them until they're needed.

Cheers,

- Michal
-------------------------------------------------------------------------
http://www.manifestation.com/         http://www.linkwatcher.com/metalog/
-------------------------------------------------------------------------




From drew.mcdowell at msfc.nasa.gov  Tue Apr 25 18:19:44 2000
From: drew.mcdowell at msfc.nasa.gov (Andrew McDowell)
Date: 25 Apr 2000 22:19:44 GMT
Subject: os.path.walk() question.
Message-ID: <8e55m0$pvl$1@hammer.msfc.nasa.gov>

I didn't see this question mentioned in any older posts so forgive me if I'm
dragging up and old topic...but:

I need to find out if a file exists under a directory hierarchy.
The hierarchy could be quite large so I don't want to continue recursing
through the hierarchy if the file is found one or two directories down.   Is
there a way that you can tell os.path.walk() to stop searching?

I suppose I could customize the code, but my first notion was that someone
on the group had probably addressed this before.

Any advice would be much appreciated :)

-Drew






From dgoodger at bigfoot.com  Sun Apr  9 17:42:25 2000
From: dgoodger at bigfoot.com (David Goodger)
Date: Sun, 09 Apr 2000 17:42:25 -0400
Subject: Python paradigms
References: <8cmr70$b0e$1@pegasus.csx.cam.ac.uk> <8cntqp$upi$1@slb0.atl.mindspring.net> <8cocgb$iuk$1@pegasus.csx.cam.ac.uk> <8cokt5$opj$1@slb1.atl.mindspring.net> <38EFE45A.2E866C58@mindspring.com><0a9301bfa244$75afe200$01ffffc0@worldnet.att.net>
Message-ID: <B516717B.17AD%dgoodger@bigfoot.com>

talking about C's "test ? true : false" operator,

on 2000-04-09 12:55, Emile van Sebille (emile at fenx.com) wrote:
> This construct has been occasionally suggested
> as one option:
> 
>>>> print ['falseresponse','truresponse'][1==2]
> falseresponse
>>>> print ['falseresponse','truresponse'][1==1]
> truresponse

There is a hidden pitfall here. You have to be careful with the test (used
as list index) at the end ("[1==2]" & "[1==1]" here). Whereas the comparison
operators (==, <, >=, etc.) will always return either 0 (false) or 1 (true),
more complex boolean operations involving "and" and/or "or" may not. For
example:

    print ['falseresponse','truresponse'][a and b]

The result depends on the values of a & b. If a is 0, fine, 'falseresponse'.
If a is an empty string, an empty list, etc. (all false-equivalent values),
you'll get an error. If a is non-zero or true-equivalent, the index will be
b's value, whatever that is. If b isn't either 0 or 1, you'll get an error
or unpredictable results.

You can of course avoid this by reversing the list and adding a "not" around
the boolean:

    print ['truresponse','falseresponse'][not(a and b)]

"not" always returns 0 or 1.

See section 5.10 of the Python Reference Manual, "Boolean operations".

-- 
David Goodger    dgoodger at bigfoot.com    Open-source projects:
 - The Go Tools Project: http://gotools.sourceforge.net
 (more to come!)



From bgue at my-deja.com  Thu Apr 20 22:24:13 2000
From: bgue at my-deja.com (bgue at my-deja.com)
Date: Fri, 21 Apr 2000 02:24:13 GMT
Subject: serial port and event loop question
Message-ID: <8doe40$h8$1@nnrp1.deja.com>

Hi everyone. This is mostly a general programming question, but since
I'm using Python for this project, I thought I'd ask here... I wrote
database written with Python and Gadfly, and I've got a nice little
Tkinter GUI for it. I need to make it monitor the serial port for
incoming data (from a barcode scanner). I'm using the Serial module,
but I don't know how to have my program watch for new data
constantly...any help?

Thanks in advance,

Brian Gue
bgue at my-deja.com


Sent via Deja.com http://www.deja.com/
Before you buy.


From soundhack at hotmail.com  Tue Apr 18 20:49:08 2000
From: soundhack at hotmail.com (Robert)
Date: Tue, 18 Apr 2000 20:49:08 -0400
Subject: win32all-128.exe availability?
Message-ID: <38fd01b2$0$10982@senator-bedfellow.mit.edu>

Now that starship has crashed, I was wondering if it was possible to
get the 128 build of Pythonwin extensions off the net? By an unfortunate
coincidence, my own hard drive chose to crash around the time starship
did.

Thanks
Robert




From tiddlerdeja at my-deja.com  Fri Apr 28 06:37:25 2000
From: tiddlerdeja at my-deja.com (tiddlerdeja at my-deja.com)
Date: Fri, 28 Apr 2000 10:37:25 GMT
Subject: command pattern/framework in python?
Message-ID: <8ebpl4$eq9$1@nnrp1.deja.com>

Anyone got an example of a Command pattern/framework that they have
used in python?

Any information appreciated.


Sent via Deja.com http://www.deja.com/
Before you buy.


From jon at dgs.monash.edu.au  Wed Apr  5 21:37:07 2000
From: jon at dgs.monash.edu.au (Jonathan Giddy)
Date: Thu, 6 Apr 2000 11:37:07 +1000 (EST)
Subject: Tuples -- who needs 'em
In-Reply-To: <AF223B12F1FDD2119253006097C4FCEC0AD357@cats.pdri.xerox.com> from "Alexander, Bob" at Apr 05, 2000 12:22:28 PM
Message-ID: <200004060137.LAA27607@nexus.csse.monash.edu.au>

I nearly fell out of my tree when Alexander, Bob declared:

>I am wondering where the serious concern about mutable lists comes from, as
>it is a pattern in the responses I've gotten on this topic. For me, the
>immutable list is a relatively new concept, since I can't think of many
>other languages that have it. I'm a programming-language-design kind of
>person, and I never really saw mutable lists as a problem that needed
>solving. 

How about C?  Surely 'const' is an attempt to mark arrays as immutable.
Apparently, the ANSI C committee saw mutable arrays as a problem that
needed solving.

>I've used many Python-type languages that have mutable lists but
>not immutable ones, and if I don't want the elements to change they don't
>change -- I simply avoid the x[i] = y construct :-) 

Avoiding the x[i] = y construct is much harder in non-trivial programs, 
where assumptions about mutability need to be specified as part of the call
interface.  Hence the proliferation of const in the C string library.

>And I still haven't
>heard from anyone about why there isn't similar concern about immutable
>dictionaries.

I could express some concern, but dictionaries are used much less than
sequences, and of all the places in my code where I make possibly needless
copies of a dictionary, it is usually to modify the dictionary with a clear
conscience, not something that immutability would solve.

Jon.



From cgw at alum.mit.edu  Tue Apr 25 17:51:13 2000
From: cgw at alum.mit.edu (Charles G Waldman)
Date: Tue, 25 Apr 2000 15:51:13 -0600
Subject: SV: editor for win32
References: <7BD10B680501D411B9DF009027E06F32011A4E@Server.Exchange>
Message-ID: <8e50g2$i01$1@info3.fnal.gov>

> Fra: Scott Hathaway [mailto:slhath at flash.net]
> 
>> Can anyone recommend a python editor for Windows?

XEmacs is ported to Windows and has the best Python
support of any editor going (as well as being, IMO, one
of the finest pieces of software ever written).   See 

 http://www.xemacs.org



From pinard at iro.umontreal.ca  Tue Apr  4 13:04:15 2000
From: pinard at iro.umontreal.ca (=?ISO-8859-1?Q?Fran=E7ois_Pinard?=)
Date: 04 Apr 2000 13:04:15 -0400
Subject: New Features in Python 1.6
In-Reply-To: Lloyd Zusman's message of "01 Apr 2000 23:10:47 -0500"
References: <200004011740.MAA04675@eric.cnri.reston.va.us> <_pyF4.25562$E85.640966@news1.rdc1.md.home.com> <slrn8edfmk.185.wtanksle@dolphin.openprojects.net> <ltvh21ywfs.fsf@asfast.com>
Message-ID: <oq66txol0w.fsf@titan.progiciels-bpi.ca>

Lloyd Zusman <ljz at asfast.com> writes:

> I'll use the new compression algorithm I recently invented.  It's amazing!
> .. it compresses *anything* down to one bit.

Beware.  It might be patented already...  Let's see.  Does it work this way?

* If the file is empty, you're done.
* If the file holds a single bit, you're done.
* Split the whole file into pairs of sequences, the first part of the pair
  being an eager, but possibly empty sequence of consecutive zero-bits,
  the second part being an eager, but possibly empty sequence of consecutive
  one-bits.  Notice that there is only one way to do this split.
* Replace each sequence above by the count of bits it contains, expressed
  directly as a binary number for the first count of the pair, and as the
  one-complements of the binary number for the second count of the pair.
  Notice that given these counts, it is trivial to rebuild the contents
  of the previous step.
* Iterate.

-- 
Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard





From robin at jessikat.demon.co.uk  Fri Apr 21 07:05:56 2000
From: robin at jessikat.demon.co.uk (Robin Becker)
Date: Fri, 21 Apr 2000 12:05:56 +0100
Subject: finding real $0
In-Reply-To: <20000421124817.C11481@xs4all.nl>
References: <OWeCyjAOWCA5Ew86@jessikat.demon.co.uk>
 <20000421124817.C11481@xs4all.nl>
Message-ID: <K3YTaKAUYDA5EwuT@jessikat.demon.co.uk>

In message <20000421124817.C11481 at xs4all.nl>, Thomas Wouters
<thomas at xs4all.net> writes

That's what I need :)
>On Fri, Apr 21, 2000 at 10:55:26AM +0100, Robin Becker wrote:
>
>> Is there any pythonic way to locate the original argument zero of the
>> process? I mean the location of python itself not the script. This would
>> allow me to launch subscripts without having to rely on the path being
>> set up correctly.
>
>Use 'sys.executable':
>
>centurion:~ > python
>>>> import sys 
>>>> sys.executable
>'/usr/bin/python'
>
>centurion:~ > ln -s /usr/bin/python penguin-on-the-telly
>centurion:~ > ./penguin-on-the-telly 
>>>> import sys
>>>> sys.executable
>'./penguin-on-the-telly'
>
>It behaves the same way in a script (it returns the python executable
>started, not the script.) Dont forget about relative paths though, if you
>are going to do stuff based on sys.executable. You probably want to get the
>cwd at the earliest possible moment, and translate sys.executable into an
>absolute path.
>

-- 
Robin Becker



From hamish_lawson at yahoo.co.uk  Thu Apr  6 05:28:24 2000
From: hamish_lawson at yahoo.co.uk (Hamish Lawson)
Date: Thu, 06 Apr 2000 02:28:24 -0700
Subject: UserDict's has_key() and get() should use __getitem__()
References: <0535db98.ec878f67@usw-ex0101-005.remarq.com> <m3u2hgn5jg.fsf@atrus.jesus.cam.ac.uk> <29328840.06b5b87b@usw-ex0101-005.remarq.com> <8cgdp3$nm5$1@slb6.atl.mindspring.net>
Message-ID: <0aa264d3.021266df@usw-ex0101-005.remarq.com>

Andrew Dalke wrote:

>I don't believe your estimate, that UserDicts are usually
>derived "for the purpose of providing new definitions of
>setting and getting items", is correct.

It may well not be. When I reread what I'd written I see that
I've made more of a claim that I meant to regarding the reasons
why UserDict is subclassed, when I was really just venturing a
guess on that.

>In addition, in a good OO design, you should be able to use
>an instance of a derived class anywhere you use an instance
>of a base class.  Your example (a caseless Dict), changes
>how/when lookup failures occur.

An instance of CaselessDict could still be used wherever an
instance of UserDict could be, in the sense that it supports the
same operations. However if it's considered part of UserDict's
contract that unique strings should be able to be used as
separate keys, then it's true that CaselessDict violates this.
Perhaps this wasn't my best example.

>Then Python's UserDict is just an implementation of the
>AbstractUserDict base class. ... And because this is Python,
>it's easy to implement.

I see where you and Michael are coming from. And if UserDict
*is* considered an implementation of a (possibly notional)
abstract class, then I suppose it is reasonable for it to use an
implementation that it is more efficient. I suppose what I was
after was to have UserDict still provide a useful default
implementation (albeit a bit less efficient one*), but make it
more abstract (semi-abstract?) to make it more easily subclassed
and to let aspects of the default implementation be more readily
replaced.

* Do we know how much less efficient?

But so far my proposal doesn't seem to have won favour. Ah
well :-)

Thanks for your thoughts.

Hamish



* Sent from RemarQ http://www.remarq.com The Internet's Discussion Network *
The fastest and easiest way to search and participate in Usenet - Free!



From lenny at squiggie.com  Mon Apr 10 16:43:56 2000
From: lenny at squiggie.com (Lenny Self)
Date: Mon, 10 Apr 2000 13:43:56 -0700
Subject: using httpdlib with a proxy server
References: <ChoI4.844$g54.17363@news.uswest.net>
Message-ID: <K2rI4.1243$g54.27608@news.uswest.net>

Ok.  I can use urllib to send a get request through a proxy server.  It will
return the page as well as the headers.  I am not, however, able to pull out
the error code (ie 200,404).  I need these codes as part of my logging .
Does anyone have any suggestions? Am I just not seeing where to pull them
from in the urllib module?

Thanks.

    -- Lenny Self
        lenny at squiggie.com


"Lenny Self" <lenny at squiggie.com> wrote in message
news:ChoI4.844$g54.17363 at news.uswest.net...> I am attempting to write a
small script that sends a list of URLs though a
> proxy server and then records a bit of information on the respose.  I
though
> I could use the httplib module to help with this, however, I don't see
where
> it can send a request though a proxy server. Does anyone know of a module
> that will allow me to do this?
>
>
>     -- Lenny Self
>         lenny at squiggie.com
>
>




From tjreedy at udel.edu  Tue Apr  4 14:30:06 2000
From: tjreedy at udel.edu (Terry Reedy)
Date: Tue, 4 Apr 2000 14:30:06 -0400
Subject: New Features in Python 1.6
References: <200004011740.MAA04675@eric.cnri.reston.va.us> <gHrF4.16072$0o4.101062@iad-read.news.verio.net>
Message-ID: <8cdbvg$3vi$1@news.udel.edu>

> Python strings can now be stored as Unicode strings.  To make it easier
> to type Unicode strings, the single-quote character defaults to creating
> a Unicode string, while the double-quote character defaults to ASCII
> strings.

' = 1 byte/char, " = 2 bytes/ char is more straightforward.

El Gato Sonreyendo







From ullrich at math.okstate.edu  Sat Apr  1 13:45:49 2000
From: ullrich at math.okstate.edu (David C. Ullrich)
Date: Sat, 01 Apr 2000 12:45:49 -0600
Subject: Function override?
References: <8c0dtu$lr3$1@nnrp1.deja.com> <38E4D7A4.4C83CFF7@math.okstate.edu> <8c4mds$5p7$2@slb6.atl.mindspring.net>
Message-ID: <38E643DD.76C31FAE@math.okstate.edu>

Andrew Dalke wrote:

> David C. Ullrich wrote:
> >    This is certainly one of the things that _I_ find keen,
> >amyway - I hope the people who are talking about adding
> >stricter typing to Python make it optional somehow. There
> >are obvious advantages to stricter typing but you lose
> >something as well. (For example, I get a big kick out of
> >the fact that exactly the  same euclidean-algorithm code
> >works to find the GCD of two integers or the GCD of
> >two polynomials. Etc.)
>
> I went to a talk by Stepanov about the Standard Template
> Library in C++.  One of his examples was a geneic GCD
> algorithm which works on integers and polynomials.  Thus,
> stricter typing and generic programming don't have to
> be the same thing.

    Either I have the meaning of some of the terms backwards
or there was a typo in that last sentence. Did you really mean
something like "Thus, stricter typing and generic programming
need not be incompatible" or something?
    If you meant what you wrote I've got the meaning of at
least one of the words backwards. If you meant more or less
what I conjecture you meant: I didn't mean to say anything
about C++, I know nothing about it. Just meant to point
out that a person could do "generic" programming in
Python very easily.

>
> In Python, this could be done by a syntax requiring not
> that "i" and "j" are ints, but that type(i) == type(j).
>
>                     Andrew
>                     dalke at acm.org



From remove.me.wwicker at spectratechnologies.com  Fri Apr 28 13:53:53 2000
From: remove.me.wwicker at spectratechnologies.com (William Wicker)
Date: Fri, 28 Apr 2000 17:53:53 GMT
Subject: Python COM troubles
References: <390a2f04.149999918@news.onlineisp.com> <L5MN4.5088$v85.30985@news-server.bigpond.net.au>
Message-ID: <390acf5e.62946872@news.onlineisp.com>

On Thu, 27 Apr 2000 00:46:35 GMT, "Mark Hammond"
<mhammond at skippinet.com.au> wrote:

>"William Wicker" <remove.me.wwicker at spectratechnologies.com> wrote in
>message news:390a2f04.149999918 at news.onlineisp.com...
>
>> Is there a way around this appearance of typelessness? There is a type
>> library, and I've run make-py successfully. I just can't see an easy
>> way into it.
>
>You can try running "makepy.py -i", and select your type library.  This
>will spit out some code, something like:
>
>gencache.EnsureModule(...)
>
>change this to:
>
>mod = gencache.EnsureModule(...)
>
>And you should be able to create any of the class objects in the makepy
>module - eg:
>
>ob = mod.TheClassName()
>
>If you search deja news, there should be lots of info about this - but
>generally relating to "Lotus Notes", which also has the same problem.
>
>[BTW - drop them a line, and tell them they should fix their objects :-]
>

Can you fill me in with what in particular I should complain about?
Comments that RTTI was lacking brought the response that
QueryInterface was supported.

Also, (did I already ask this?) I'm guessing that using the
EnsureModule approach will make DispatchWithEvents unavailable.
Reading dynamic.py suggests that there is an alternate (more complex?)
way.

(I hate knowing what I want to do, but not how to do it.)

	William.



From thomas at cintra.no  Wed Apr 26 09:24:55 2000
From: thomas at cintra.no (‘5ÛHH575-UAZWKVVP-7H2H48V3)
Date: Wed, 26 Apr 2000 13:24:55 GMT
Subject: Best database to use with Python
Message-ID: <3906ea27.698463777@news.online.no>

Hi,

What database-system is easiest, fastest etc. to use with Python? I`ve
looked at MySQL, but it seems a little overkill just to get a little
more functionality than Berkley DB-functionality. What about Postgres?
mSQL? 

Postgres comes with my Linux distro and seem smaller than MySQL. How
does it compare to MySQL when it comes to administration and speed,
scalability etc. ?

I`m doing alot of index-lookups for a
Yahoo/Altavista/WhatEverSearchEngine-like web-app.

<entering_demand_mode>
It have to be fast, capable of HUGE amounts of data, free and easy to
install and administer.
</entering_demand_mode>


Hints???!

Thomas




From gmcm at hypernet.com  Tue Apr 11 10:54:48 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Tue, 11 Apr 2000 10:54:48 -0400
Subject: Processing BCD and Signed Integer Fields with Python
In-Reply-To: <38f334ea@wwwproxy3.westgroup.com>
Message-ID: <1256657606-41178742@hypernet.com>

Chris Lada wrote:

> We have been receiving record/field-based information from a data provider.
> Some of the fields are BCD (Binary Coded Decimal), and some are signed
> integer. The data originates from a mainframe/Cobol(?) system. The record
> layouts have definitions like "PIC S99999" (signed integer), and "PIC S99999
> Comp-3" (BCD). Has anyone had to process this sort of data with Python ?
> Can Python handle this sort of data type ?  Anyone have suggestions on how
> to approach this problem ?

You want the struct module.

- Gordon



From alex_c at mit.edu  Wed Apr 19 21:12:26 2000
From: alex_c at mit.edu (Alexander S Coventry)
Date: 19 Apr 2000 21:12:26 -0400
Subject: Where do I send a contribution for the Starship recovery ?
References: <kXqL4.142058$ra.3216975@news.easynews.com>
Message-ID: <etdbt35vamd.fsf@mint-square.mit.edu>

> I personally would consider it money well spent.
> (unlike most of the other things I spend money on)
> 		      Roger Upole

Please also send your donation four times. :)

Alex.


From michael.stroeder at inka.de  Sun Apr 16 08:03:08 2000
From: michael.stroeder at inka.de (Michael =?iso-8859-1?Q?Str=F6der?=)
Date: Sun, 16 Apr 2000 14:03:08 +0200
Subject: Python CGI
References: <38F852D9.E2144F17@flashmail.com>
Message-ID: <38F9ABFC.EDF7E70B@inka.de>

Jp Calderone wrote:
> 
> Whenever a form that uses
> the Python script as its ACTION, the python scripts seems to get 
> executed about 8 times.

<IMG SRC="[python script]"> ?

Ciao, Michael.


From mwh21 at cam.ac.uk  Wed Apr  5 13:12:36 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 05 Apr 2000 18:12:36 +0100
Subject: Why should I switch to Python?
References: <000701bf9c5a$b806b420$752d153f@tim> <oqity066km.fsf@titan.progiciels-bpi.ca> <slrn8ej2gc.b3t.warlock@raven.myth> <m3itxyz744.fsf@atrus.jesus.cam.ac.uk> <050420000914595917%pecora@anvil.nrl.navy.mil> <mt1z4k8oh1.fsf@astron.berkeley.edu>
Message-ID: <m3pus4lbej.fsf@atrus.jesus.cam.ac.uk>

Johann Hibschman <johann at physics.berkeley.edu> writes:

> Louis M Pecora writes:
> 
> > In article <m3itxyz744.fsf at atrus.jesus.cam.ac.uk>, Michael Hudson
> > <mwh21 at cam.ac.uk> wrote:
> 
> >> don't know it; I think it involves slinging Riemann zeta functions).
> >                                      ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
> 
> > There's a good phrase.
> 
> Holy slinging Riemann zeta functions, Batman!
> 

<grin> - mind you "Riemann zeta function" is a pretty cool name for
any function.  I also like "Weierstrass p-function", but mainly
because the p is usually drawn in an entertainly deranged way.

what-was-this-ng-about-again?-ly y'rs
Michael

-- 
  nonono,  while we're making wild  conjectures about the behavior 
  of completely  irrelevant tasks,  we must not  also make serious 
  mistakes, or the data might suddenly become statistically valid.
                                      -- Erik Naggum, comp.lang.lisp


From phd at phd.russ.ru  Mon Apr 24 07:26:09 2000
From: phd at phd.russ.ru (Oleg Broytmann)
Date: Mon, 24 Apr 2000 11:26:09 +0000 (GMT)
Subject: Installing PyApache
In-Reply-To: <3901CF0C.E9A722F6@myself.com>
Message-ID: <Pine.LNX.4.21.0004241125010.6019-100000@fep132.fep.ru>

On Sat, 22 Apr 2000, Wayne Izatt wrote:
> I tried to build Apache with PyApache. I already have Python 1.5.2
> installed (with tkinter).
> 
> The build dies when linking, complaining that a whole bunch of tk...
> references can't be resolved.
> 
> Now, since PyApache probably won't be needing tkinter ;-) is there any
> simple way to prevent this situation.
> 
> (I suspect that rebuilding Python without tkinter support might do the
> trick, but that seems like a long way round).

   I guess that you've built your Python static, or at least built  it with
static tkinter? Than there is no way to avoid rebuilding...

Oleg.            (All opinions are mine and not of my employer)
---- 
    Oleg Broytmann      Foundation for Effective Policies      phd at phd.russ.ru
           Programmers don't die, they just GOSUB without RETURN.




From alwagner at tcac.net  Sun Apr 23 13:22:52 2000
From: alwagner at tcac.net (Albert Wagner)
Date: Sun, 23 Apr 2000 12:22:52 -0500
Subject: The Simple Economics of Open Source
References: <B527E7A5.51AA%raffael@mediaone.net>
Message-ID: <3903316C.5D6F1611@tcac.net>

Raffael Cavallaro wrote:
> 
> 
> Sarcasm aside, I think you miss the point. Those artists least willing to
> share, most jealous of their "secrets" have always been those who were most
> successful. Unsucessful young artists are, by and large, quite willing to
> share ideas, working techniques, etc. It's actually the more established,
> financially succesful ones, who are not.
> 

I think you confuse artifice and art.  Talent is not a "secret" that can
be hoarded or shared.  It is visible in the end product, yet no amount
of sharing how it was done will allow the process to be duplicated.  I
think you also are confused about what "successful" means to a talented
person.  The "commodification" (sorry, Marx) of anything and everything,
assigning it a dollar value, is at the very root of the mindset that is
destroying the world. 
-- 
Small is Beautiful


From aycock at csc.uvic.ca  Thu Apr 27 12:24:16 2000
From: aycock at csc.uvic.ca (John Aycock)
Date: 27 Apr 2000 16:24:16 GMT
Subject: [ANNOUNCE] SPARK 0.6.1
Message-ID: <8e9pjg$4cjm$1@uvaix7e1.comp.UVic.CA>

The latest release of SPARK, the Scanning, Parsing, and Rewriting Kit, is
available at

	http://csr.uvic.ca/~aycock/python

Version 0.6.1 fixes a bug in the parser; sorry if it inconvenienced anyone!

Also, if you'd be interested in testing SPARK pre-releases, please send
me some mail and I'll add you to the list.  (Well, first I'll *make* a
list, then I'll add you to it :-)

John


From slhath at flash.net  Fri Apr 21 09:10:08 2000
From: slhath at flash.net (Scott Hathaway)
Date: Fri, 21 Apr 2000 13:10:08 GMT
Subject: editor for win32
References: <NpHL4.779$fK2.113609@news.flash.net>
Message-ID: <QqYL4.1341$fK2.161623@news.flash.net>

Thanks for all the good input!

I am new to python and excited about giving it a whirl.

Scott

"Scott Hathaway" <slhath at flash.net> wrote in message
news:NpHL4.779$fK2.113609 at news.flash.net...
> Can anyone recommend a python editor for Windows?
>
> If there is an IDE for windows (other than IDLE), even better.
>
> Thanks,
> Scott
>
>




From phd at phd.russ.ru  Thu Apr  6 03:58:34 2000
From: phd at phd.russ.ru (Oleg Broytmann)
Date: Thu, 6 Apr 2000 07:58:34 +0000 (GMT)
Subject: looking for LL(1), LR(1), or LALR(1) syntaxic analysis in python.
In-Reply-To: <00bc01bf9f9e$240c53a0$1001a8c0@nevrax.net>
Message-ID: <Pine.LNX.4.21.0004060758040.4220-100000@fep132.fep.ru>

On Thu, 6 Apr 2000, Sameh chafik pro wrote:
> At present i develop an open source in agent IA, i write for you because
> i'm looking for any LL(1), LR(1), or LALR(1) syntaxic analysis in python
> implementation.

   http://csr.uvic.ca/~aycock/python/
   http://theory.stanford.edu/~amitp/Yapps/

Oleg.            (All opinions are mine and not of my employer)
---- 
    Oleg Broytmann      Foundation for Effective Policies      phd at phd.russ.ru
           Programmers don't die, they just GOSUB without RETURN.




From tim.hochberg at ieee.org  Wed Apr 26 13:46:25 2000
From: tim.hochberg at ieee.org (Tim Hochberg)
Date: Wed, 26 Apr 2000 17:46:25 GMT
Subject: Simple del[] list question
References: <meh9-9D8292.13251626042000@news.cit.cornell.edu>
Message-ID: <RXFN4.78502$U4.727175@news1.rdc1.az.home.com>

"Matthew Hirsch" <meh9 at cornell.edu> wrote:
> Hi All,
>
> What is going on here?
>
> This leaves a intact.
>
> >>> a=[[1,2,3],[4,5,6]]
> >>> b=a[:]

This makes a _shallow_ copy of a. It's probably more obivous what's going on
if we name the sublists in a"

>>> x = [1,2,3]
>>> y = [4,5,6]
>>> a = [x,y]  # This is equivalent to a=[[1,2,3],[4,5,6]]
>>> b = a[:]

Since this makes a shallow copy, b is a new list containing x and y. In
other words, in this case, b = a[:] is equivalent to b = [x, y].

> >>> del b[0]
> >>> b
> [[4, 5, 6]]

This removes x from b, but since a is a copy of b it has no effect on a.

> >>> a
> [[1, 2, 3], [4, 5, 6]]
>
>
> This also changes a.
>
> >>> a=[[1,2,3],[4,5,6]]
> >>> b=a[:]
> >>> del b[0][0]

Here del b[0][0] is equivalent to del x[0]. Since a and b share x, this
effects both arrays.

> >>> b
> [[2, 3], [4, 5, 6]]
> >>> a
> [[2, 3], [4, 5, 6]]
>
> Why did a change?  Why suddenly when you go to two dimensions does the
> behavior change?

Hope the helps,

-tim





From nascheme at enme.ucalgary.ca  Mon Apr 17 18:17:50 2000
From: nascheme at enme.ucalgary.ca (Neil Schemenauer)
Date: Mon, 17 Apr 2000 22:17:50 GMT
Subject: Filtering web proxy
References: <Pine.LNX.4.21.0004171406250.30999-100000@fep132.fep.ru> <ku66tgh0cm.fsf@lasipalatsi.fi>
Message-ID: <slrn8fn3de.gvj.nascheme@cranky.arctrix.com>

Erno Kuusela <erno at iki.fi> wrote:
>a html parser would need to work incrementally, unless you want to
>wait for the whole document to be transferred over the network before
>seeing any of it rendered.

Yes and if your connection is fast enough that you don't need
incremental loading you probably don't care too much about ads.
In my experience, filtering ads greatly enhances your experience
if your browsing on a slow connection.

>i guess you could do it incrementally with sgmllib (iirc you feed it a
>file object?), but you run into the fact that a big part of
>the html documents on the web are malformed and rely on the
>error correcting heuristics of the major browsers to function...

Right and there seems to be a lot of bad HTML code out there.
Unfortunately, I don't think you can easily make sgmllib parse
incrementally.  Someone please correct me if I'm wrong.

Is the situation with XML the same as HTML?  Are XML documents
forced to adhere to the standard or are parsers supposed to try
to do something intelligent with whatever crap they get fed?

>one starting point could be the "gray proxy" (i forget what it was
>really called). that was written on top of medusa, i think there was
>an announcement here? probably a year or so ago.  it parsed the html
>and changed all the colors to grayscale, and did the same for
>images. medusa isn't free though.. (except the version in zope?)

You can try my "munchy" proxy.  Its at:

    http://www.enme.ucalgary.ca/~nascheme/python/

Saying that is parses HTML is a bit of a stretch however.  It
just uses a couple of regexs.  I'm sure Tim Peters would love
it. :)

In my experience, filtering ads at the HTML level is more
effective than filtering at the request level (like junkbuster).
My list of blocked URLs is very short but still catches ads on
almost all the sites I visit.  Also, when ads are filtered I
usually cannot tell by looking at the page.  Of course, YMMV.


    Neil

-- 
HTML needs a rant tag. --Alan Cox


From ullrich at math.okstate.edu  Tue Apr  4 14:38:21 2000
From: ullrich at math.okstate.edu (David C. Ullrich)
Date: Tue, 04 Apr 2000 13:38:21 -0500
Subject: Why should I switch to Python? - Infinity of Primes
References: <NDBBKEGCNONMNKGDINPFGEJDDFAA.infonuovo@email.com> <8ccoug$f4o$1@pegasus.csx.cam.ac.uk> <7F51E902E09F3F60.8F612607812ED16B.82CB8B41435E14FD@lp.airnews.net>
Message-ID: <38EA369C.26A1C25@math.okstate.edu>

    Seeing your name here for a second I was confused
whether I'd hit the sci.math button...

    There's nothing non-constructive about the traditional
proof of the infinitude of the sequence of primes - given
a sequence of primes it _constructs_ a prime not on the
list. (Of course it doesn't actually return infinitely many
primes in one calculation, but neither does any other
algorithm.) Could be that there's a proof of the existence
of infinitely many primes using FTA that has some property
not shared by Euclid's proof, but "constructive" isn't it.

Cameron Laird wrote:

> In article <8ccoug$f4o$1 at pegasus.csx.cam.ac.uk>,
> Nick Maclaren <nmm1 at cus.cam.ac.uk> wrote:
> >
> >In article <NDBBKEGCNONMNKGDINPFGEJDDFAA.infonuovo at email.com>, "Dennis E. Hamilton" <infonuovo at email.com> writes:
> >|> The standard approach is a proof by contradiction starting from the
> >|> assumption that there is a largest prime.
> >
> >There is also a constructive proof based on the Fundamental
> >Theorem of Arithmetic, that is little more complex.
>                         .
>                         .
>                         .
> To what extent are constructive proofs supplanting *... ad
> absurdum* as standards?  At about the time I left academe
> I was working on my own arguments that our conventional
> view of the former as "more complex" is, as Brouwer et al.
> taught, essentially only convention.  Tilting at construc-
> tivist windmills is one of the half-dozen careers I'm
> considering for my retirement; I recognize, though, that
> I've quite lost touch with current practice.
> --
>
> Cameron Laird <claird at NeoSoft.com>
> Business:  http://www.Phaseit.net
> Personal:  http://starbase.neosoft.com/~claird/home.html



From mal at lemburg.com  Mon Apr  3 09:12:58 2000
From: mal at lemburg.com (M.-A. Lemburg)
Date: Mon, 03 Apr 2000 15:12:58 +0200
Subject: [Python-Dev] Re: New Features in Python 1.6
References: <ECEPKNMJLHAPFFJHDOJBIEOPCHAA.mhammond@skippinet.com.au>
Message-ID: <38E898DA.B69D7ED6@lemburg.com>

Mark Hammond wrote:
> 
> >
> > > On Sat, Apr 01, 2000 at 12:00:00PM -0500, Guido van
> > Rossum wrote:
> >
> > The above line has all the answers ;-) ...
> 
> That was pretty sneaky tho! <grin>  Had the added twist of being
> half-true...

... and on time like a CRON-job ;-)

-- 
Marc-Andre Lemburg
______________________________________________________________________
Business:                                      http://www.lemburg.com/
Python Pages:                           http://www.lemburg.com/python/





From ngps at madcap.dyndns.org  Sat Apr  8 11:58:49 2000
From: ngps at madcap.dyndns.org (Ng Pheng Siong)
Date: 8 Apr 2000 15:58:49 GMT
Subject: Paid to work in Python? (was: Re: Choice Of Language)
References: <Uqvw4.2261$yL2.28868@typhoon.southeast.rr.com> <38e5f8ed.29659873@news.nikoma.de> <8cl2mq$971$1@mawar.singnet.com.sg> <38ee674b.1229749@news.main-echo.net>
Message-ID: <8cnkvp$t89$1@violet.singnet.com.sg>

According to Arnold Weis <A.Weis at t-online.de>:
> ngps at madcap.dyndns.org (Ng Pheng Siong) wrote:
> 
> >I once saw a German term that was translated to "improve-worse-ment."
> >
> >Can some one provide the German word, please?
> 
> Verschlimmbesserung?
> (Trying to make something better, but making it worse instead)

Thanks. That must be it, since so far all posters from .de say so. ;-)

I like that word very much. ;-)

Cheers.


-- 
Ng Pheng Siong <ngps at post1.com> * http://www.post1.com/home/ngps



From embed at geocities.com  Tue Apr 11 09:10:58 2000
From: embed at geocities.com (Warren Postma)
Date: Tue, 11 Apr 2000 09:10:58 -0400
Subject: Why should I switch to Python?
References: <seas5ia89177@corp.supernews.com> <slrn8f4d0j.1t0.bwinton@tor.dhs.org> <slrn8f4tfr.6hh.quinn@mono.ugcs.caltech.edu>
Message-ID: <IvFI4.5759$HG1.168717@nnrp1.uunet.ca>

> I always suspected perl programmers were from Mars.  Now if we could
figure
> out where python programmers are from, someone could write a book and get
> rich.

Yorkshire.







From mskott at image.dk  Sun Apr 23 15:18:30 2000
From: mskott at image.dk (Martin Skøtt)
Date: Sun, 23 Apr 2000 19:18:30 GMT
Subject: Trouble compiling PyGreSQL 2.4 on Linux
Message-ID: <k8hoeidh.fsf@jumbo.skott.dk>

Hi 
I'm trying to compile the PyGreSQL 2.4 module on my Red Hat Linux 6.1
box. I use the following command:

gcc -fpic -shared -o _pg.so -I/usr/include/python1.5
-I/usr/local/pgsql/include pgmodule.c -lpq

With the following result:
pgmodule.c: In function `pgconnect':
pgmodule.c:694: warning: passing arg 4 of `PyArg_ParseTupleAndKeywords' from incompatible pointer type
/usr/bin/ld: cannot open -lpq: No such file or directory

I can't use the make command mentioned in the readme.linux file
because the simply isn't any make file in the archive. Can you help me
getting it compiled?

And a little addon queston:
How stable is the DB-API version of the PostgreSQL module? Can it
be used for production work (or semi serious)?

-- 
Martin Sk?tt
mskott at image.dk

'In a certain sense, all decent programming languages are equally powerfull'
Dr. Alan Turing


From gresham at mediavisual.com  Fri Apr  7 12:27:43 2000
From: gresham at mediavisual.com (Paul Gresham)
Date: Sat, 8 Apr 2000 00:27:43 +0800
Subject: CGI Help Needed
References: <8ck9uo$ga08@imsp212.netvigator.com> <38EDDA7B.47FC3FDE@mach.uni-karlsruhe.de>
Message-ID: <8cl2a2$kcn4@imsp212.netvigator.com>

Hi Andreas,
Thankyou for your suggestion, (and thanks to everyone else who responded).
Actually I have a book on PHP programming but never looked this up in there,
dispite spending hours searching the web for answers.

For anyone who is interested, Prentice Hall, Core PHP Programming, Atkinson,
ISBN 0-13-020787-X Pages 475-479 has an entire article on PHP session
tracking with a recommended approach and sample code. Exactly what I needed
for my Python project, very simple (as expected) but just one of those
things that looks like magic until you see the trick.

Regards
Paul Gresham

"Andreas Cardeneo" <Andreas.Cardeneo at mach.uni-karlsruhe.de> wrote in message
news:38EDDA7B.47FC3FDE at mach.uni-karlsruhe.de...
>
> >
> > What is the normal and respected method of User authentication on a web
> > site, how should I present a login box when the user is not
authenticated
> > already, and how do I maintain a session with an authenticated user,
without
> > using cookies, allowing them to continue surfing around the web site ?
>
> Hi,
>
> well. I think you're right that this is not exactly Python specific, but
> you might
> use Python in a CGI-script to look up session-keys. If you have a look
> at PHP, then
> you will see that there is a function to generate a session key (some
> large number
> that can not be guessed easily and has some checksum in it) that you
> pass to your pages
> as a URL paramter in the form "?session=4242424242424242&nextpar=....".
> You then evaluate
> the key in your script looking it up in a database or file and checking
> if it is still
> valid.
> I would certainly not use JavaScipt to authenticate the user because the
> implementations
> have some security problems. If I remember correctly Microsoft had
> problems on their
> Hotmail-system with Javascript. One possibility would be to use the
> .htaccess file of
> a directory and let the browser/server check authorization. But that's
> probably a bad idea
> for larger sites. You could better use a form within a frame on your
> page so the URL is
> not being displayed in the browser location line (but in the HTML code
> it is).
>
> Hope this helps,
>
> Andreas




From rurban at x-ray.at  Sun Apr  9 11:21:11 2000
From: rurban at x-ray.at (Reini Urban)
Date: Sun, 09 Apr 2000 15:21:11 GMT
Subject: win32all-130 for Python 1.6 available
References: <_ziF4.1488$1b4.6926@news-server.bigpond.net.au> <d5rjeskov6c6drsok4n7vvcvvi74155lna@4ax.com> <fqvG4.5062$1b4.24253@news-server.bigpond.net.au> <ph5mes8vkdpgk05lgneus29fi3apbvtofp@4ax.com>
Message-ID: <38f0a0c1.21524280@judy>

Wolfgang Strobl wrote:
>"Mark Hammond" <mhammond at skippinet.com.au> wrote:
>May I take the opportunity to suggest another feature? The Zope sources,
>like many others packages, have each and every file starting with a
>lengthy, boring copyright message, license, or something similar, which
>doesn't aid in understanding a programs structure in any way.  Wouldn't
>it be nice if "Toggle top-level" and "Collapse all" would collapse
>comment blocks, too?

you may want to look at the hide-copyleft.el emacs module.
A very recent version is at my site, for the string regexp's to fold at.

http://xarch.tu-graz.ac.at/autocad/lsp_tools/ntemacs/hide-copyleft.el
--
Reini Urban
http://xarch.tu-graz.ac.at/autocad/news/faq/autolisp.html


From weil at sireconnect.de  Tue Apr 18 11:07:34 2000
From: weil at sireconnect.de (Weil)
Date: Tue, 18 Apr 2000 17:07:34 +0200
Subject: Tkinter items question
Message-ID: <38fc7a7d@news2.sireco.de>

Hello,
I'm making a little drawing tool using Tkinter, where
the user can delete the last item by right-clicking with
the mouse (and by continuing to click clean the whole
canvas). The following method (of my canvas class)
does just what I want:

def DeleteLastItem(self,event):
    items = self.find_all()
    if items:
        self.delete(items[-1])

But, I was wondering, if there is a more direct way to
access the last item other than retrieving all items first.
I could of course keep a list of all created items myself,
but that seems a bit ugly, especially as my app couldn't
use this list for anything else.
It is also not possible to just hold the last item's  id and
just decrease it when deleting. There is no guarantee,
that the ids are consecutive integers, at least when
mixing creation and and deletion of items. (I know, I've tried).

So, did I just miss the right spot in the fine manuals (even tried
that of Tk) ?
I'd appreciate any suggestions that might help me cure my
uneasiness with the above code.
Thanks.




From boud at rempt.xs4all.nl  Sat Apr 15 13:44:32 2000
From: boud at rempt.xs4all.nl (Boudewijn Rempt)
Date: 15 Apr 2000 17:44:32 GMT
Subject: Scripting and Gnome and KDE
References: <Pine.GSO.4.10.10004150741520.11261-100000@sundial> <3dk8hzb90u.fsf@newcnri.cnri.reston.va.us>
Message-ID: <8da9q0$j0t$1@news1.xs4all.nl>

Andrew Kuchling <akuchlin at mems-exchange.org> wrote:
> Moshe Zadka <moshez at math.huji.ac.il> writes:
>> Personally I chose Gtk+/GNOME, since (among other things), the Gtk+ type
>> system is specifically targetted to support bindings to other languages.

> I've played a bit (only a bit, though) with both PyGTk and PyQt.
> While I think I marginally prefer PyGTk, because James Henstridge has
> done a really nice job of writing the bindings, I use KDE on my
> desktop at home and at work, and therefore the PyQt/KDE bindings are
> of greater practical interest to me.  They're pretty good, too, though
> the extension module is utterly huge, around 4Mb last time I looked.

Depending on whether you also want KDE support, that's a bit better
now. The qt extension library is just 2Mb now. Of course, adding the
various KDE libraries adds a bit of weight, too - but the most
important, kdecore is just 300Kb. The days when doing an import qt
added 12Mb to you program are long gone!

> Every so often I go to some Linux conference and come within the range
> of Miguel de Icaza's reality distortion field, and I get all excited
> over some new GNOME feature, such as Bonobo's embedding, and try to
> compile GNOME when I get home.  Then reality sets in: GNOME uses
> patched versions of all sorts of things such as automake, but this is
> never clearly documented anywhere.  There's a large pile of libraries,
> and it's not clear which ones are needed and which ones are obsolete;
> oh, you need imlib, but wait, it's been replaced by gdk-pixbuf, and
> esound has been replaced by something else... Usually I never manage
> to get a fully working GNOME setup, and the reason for the failure is
> usually not apparent.

Just wait until PyKDE supports dCop and kParts... And for other fun things
I'm going to try and build a Unicode text editor with Python 1.6 and
PyQt 0.11 this week - should be possible, since both support Unicode now.
(That's one thing I thought really impressive in Idle - paste a bit
of unicode in the editor, run it, and it prints the correct Unicode
characters!)

> KDE, on the other hand, is pretty easy to compile, even from the CVS
> tree; compile Qt, compile kdesupport and kdelibs, and that's it --
> you're all set to compile chosen applications.  Much better
> structuring of the development environment -- KDE's setup shares some
> of Python's elegance.  (Maybe it's a European thing... :) ) So it's
> odd; while I prefer PyGTk, it's tied to a much less well-structured
> environment and hence loses out through no fault of its own.

I'd say that the general design of Qt is cleaner, too. At least, I'd
say that if I had spent enough time looking that the gtk design. Qt
_is_ very clean, though.

-- 

Boudewijn Rempt  | http://www.valdyas.org


From cut_me_out at hotmail.com  Tue Apr 18 17:57:53 2000
From: cut_me_out at hotmail.com (Alex)
Date: 18 Apr 2000 17:57:53 -0400
Subject: Date class?
References: <8dikk2$p4s@flatland.dimensional.com>
Message-ID: <etdr9c3ul5q.fsf@w20-575-109.mit.edu>

You could try normalDate.  I have found that fairly useful, once or
twice.

Alex.


From mwh21 at cam.ac.uk  Sun Apr 16 13:31:52 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 16 Apr 2000 18:31:52 +0100
Subject: Python paradigms
References: <8cmr70$b0e$1@pegasus.csx.cam.ac.uk> <8dbdj3$dgt$1@slb2.atl.mindspring.net> <8dcpom$fbc$1@slb3.atl.mindspring.net>
Message-ID: <m3zoqudkaf.fsf@atrus.jesus.cam.ac.uk>

"Andrew Dalke" <dalke at acm.org> writes:

> def expression_if(test, true_string, false_string):
>   # some nasty code to get the locals of the calling function
>   if test:
>     return eval(true_string with the given locals)
>   else:
>     return eval(false_string with the given locals
> 
> x = expression_if(a is not None, "a[i].weeble", "0") + ...
> 
> but that's just too ugly.

bytecodehacks can do this (except that the relavent bits seem to have
suffered bit-rot; I'll do something about that presently).

I also can't remebmer how that bit of bytecodehacks works (! - just as
well I wrote docs <wink>).

Cheers,
M.

-- 
  "declare"?  my bogometer indicates that you're really programming
  in some other language and trying  to force Common Lisp into your
  mindset.  this won't work.           -- Erik Naggum, comp.lang.lisp


From MSteed at altiris.com  Thu Apr 20 15:35:49 2000
From: MSteed at altiris.com (Mike Steed)
Date: Thu, 20 Apr 2000 13:35:49 -0600
Subject: editor for win32
Message-ID: <65118AEEFF5AD3118E8300508B1248774CB2AC@ALTNET>

I'm sure lots of people will recommend the Pythonwin IDE by Mark Hammond.
You can find links to it from the python.org site.

As a Vim addict, I have to mention it as well ( http://www.vim.org/ ).  It
comes with Python syntax highlighting.  There are suggestions for useful
mappings, etc., in the newsgroup archives.

Mike

> From: Scott Hathaway [mailto:slhath at flash.net]
> Sent: Thursday, April 20, 2000 11:48 AM
> To: python-list at python.org
> Subject: editor for win32
> 
> 
> Can anyone recommend a python editor for Windows?
> 
> If there is an IDE for windows (other than IDLE), even better.
> 
> Thanks,
> Scott



From mwh21 at cam.ac.uk  Tue Apr 25 13:48:15 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 25 Apr 2000 18:48:15 +0100
Subject: Don't understand error message
References: <65118AEEFF5AD3118E8300508B1248774CB2B4@ALTNET>
Message-ID: <m3hfcqf4ww.fsf@atrus.jesus.cam.ac.uk>

Mike Steed <MSteed at altiris.com> writes:

> > Cheers,
> > M.
>   ^^
> (I thought that was my .sig)
> M.

Well, my other signoff is:

Cheers,
Michael

which isn't necessarily any more distinctive.

oh,-now-what-do-I-do?-ly y'rs
m.

-- 
39. Re graphics:  A picture is worth 10K  words - but only those 
    to describe the picture. Hardly any sets of 10K words can be 
    adequately described with pictures.
     -- Alan Perlis, http://www.cs.yale.edu/~perlis-alan/quotes.html


From dworkin at ccs.neu.edu  Tue Apr 18 18:53:14 2000
From: dworkin at ccs.neu.edu (Justin Sheehy)
Date: 18 Apr 2000 18:53:14 -0400
Subject: Forward References?
In-Reply-To: "Jeff Massung"'s message of "Tue, 18 Apr 2000 14:12:13 -0600"
References: <eviltofu-4C5688.21593817042000@news.earthlink.net> <oh3L4.1227$zk4.23905@news.uswest.net>
Message-ID: <vndog77c97p.fsf@camelot-new.ccs.neu.edu>

"Jeff Massung" <jmassung at magpiesystems.com> writes:

> >class a:
> >   def x(self):
> >      y(self)

> Should this be a.y(self)?

No, it should be self.y()

It could be a.y(self), and that would work for many cases, but that
isn't really what you want.

> Also, someone mentioned above having: self.y() (or self.y(self)), my
> question is what is the difference between a.y and self.y?

It's easy to find out in the interactive interpreter.

>>> class a:
...  def y(self):
...   print 1
... 
>>> b = a()
>>> a.y
<unbound method a.y>
>>> b.y
<method a.y of a instance at 80a67e8>
>>> b.y()
1
>>> a.y()
Traceback (innermost last):
  File "<stdin>", line 1, in ?
TypeError: unbound method must be called with class instance 1st
argument
>>> a.y(b)
1

Or, of course, one could read the documentation or tutorial.  This is
pretty well explained, iirc.

-Justin

 



From fredrik at pythonware.com  Wed Apr  5 07:54:15 2000
From: fredrik at pythonware.com (fredrik at pythonware.com)
Date: Wed, 05 Apr 2000 11:54:15 GMT
Subject: Python 1.6 alpha 1 released
References: <Pine.GSO.4.10.10004040035420.21084-100000@sundial> <5E877FD7637AB46E.DABBED9CF3C1D64B.CDDDC05AE7374A78@lp.airnews.net> <5l7leern5v.fsf@eric.cnri.reston.va.us> <8cdg92$j0m$1@nntp6.u.washington.edu> <8cdig3$s7f$1@slb6.atl.mindspring.net> <fp7messqdb29oirr9j5uktgi22ic6od5ab@4ax.com>
Message-ID: <8cf9h2$ipa$1@nnrp1.deja.com>

Toby Dickenson wrote:
> Does the single-argument connect require a 2-tuple, or just a two
> element sequence?

sure.

class WebServer:
    def __init__(self, host, port=80):
        self.address = host, port
    def __len__(self):
        return len(self.address)
    def __getitem__(self, index):
        return self.address[index]

import socket

host = WebServer("www.pythonware.com")

s = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
s.connect(host)

# ...

> If a sequence will do, then you can use an address class today.

exactly.

</F>


Sent via Deja.com http://www.deja.com/
Before you buy.


From effbot at telia.com  Sun Apr  2 17:40:15 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Sun, 02 Apr 2000 21:40:15 GMT
Subject: NameError when assigning dictionary values
References: <20000402203338.21701.qmail@web2105.mail.yahoo.com>
Message-ID: <37PF4.4774$74.76551@newsc.telia.net>

lewst <lewst at yahoo.com> wrote:
> This simple problem is driving my crazy.  I can't figure out why I get
> a NameError with this program.
>
>   # add each key-value pair to the dictionary and
>   # then for each value, create an empty list.
>   mydict = {}
>   for key,value in [ (".", "one"), ("+", "two") ]:
>       mydict[key] = value
>       value = []
>   print mydict
>   print one,two
>
> Whe I run this I get "NameError: one".  Why aren't my 2 empty lists
> named "one" and "two" being created?

I haven't the foggiest idea of what you expect Python
to do here (or why you expect it to work in that way),
but let's step through your code by hand:

>   mydict = {}

this creates an empty dictionary, named 'mydict'.

>   for key,value in [ (".", "one"), ("+", "two") ]:

this loops over a list, containing two 2-tuples.
the first time, it sets 'key' to ".", and 'value' to
"one".  the second time, it sets 'key' to "+" and
'value' to "two".

>       mydict[key] = value

this adds a member to the dictionary, with the
key "." and value "one"

>       value = []

this creates a new empty list, and changes 'value'
to point to the list instead of the original value.

not that it matters much -- the second time through
the loop, 'value' will be set to "two" by the for-in state-
ment.

>   print mydict

this prints the contents of the dictionary (which, in
this case, is something like {"+": "two", ".": "one"} )

>   print one,two

this attempts to print the variables 'one' and 'two.

there are no such variables -- you haven't assigned
them.  that's why

there are two *keys* in the dictionary, but that's a
completely different thing.  dictionaries are container
objects.  putting things is a dictionary doesn't make
it visible in the variable namespace.

I suggest reading the tutorial over again, and perhaps
signing up to the tutor mailing list.  you'll find pointers
at www.python.org.

</F>




From cut_me_out at hotmail.com  Mon Apr 17 17:54:13 2000
From: cut_me_out at hotmail.com (Alex)
Date: 17 Apr 2000 17:54:13 -0400
Subject: funny:  Re: May 27th PARTY!
References: <OF25503401.CB160CEC-ON882568C4.00680F11@peoplesoft.com>
Message-ID: <etdsnwkmm0q.fsf@w20-575-120.mit.edu>

> Sounds like a blast...What should I bring?

A time machine would be a good start.

Alex.


From robin at jessikat.demon.co.uk  Sat Apr 15 05:33:31 2000
From: robin at jessikat.demon.co.uk (Robin Becker)
Date: Sat, 15 Apr 2000 10:33:31 +0100
Subject: [ANNOUNCE] Garbage collection for Python
References: <20000407221750.A29279@acs.ucalgary.ca>
 <m3r9cagx3c.fsf@atrus.jesus.cam.ac.uk>
 <l03102809b51d41895a80@[193.78.237.135]> <38F826CF.DBDC71EF@zipzap.ch>
Message-ID: <AB0KBJArdD+4Ew5t@jessikat.demon.co.uk>

In article <38F826CF.DBDC71EF at zipzap.ch>, Boris Borcic <zorro at zipzap.ch>
writes
>Just van Rossum wrote:
>> 
>> - Will Ware and I have worked hard on a complete rewrite of the
>>   uthread module. It is a *vast* improvement over the old one.
>>   It offers a very nice thread object and several handy threading
>>   widgets. It already can do stuff threading.py can only *dream*
>>   about, like pausing (blocking) a thread from another thread or
>>   posting an exception from another thread (!).
>
>Simple examples of useful new idioms that this permits ?
>
>BTW, I remember there exists a package for lisp, called
>"screamer", that turns it in a backtracking non-deterministic
>language in the style of prolog, through a device called
>CPS (continuation-passing style) transformation. While
>python lacks S-expressions and lisp-style macros that
>allow this for lisp, continuations are clearly the key
>component of this gimmick. It might be worth its while
>to have a better look.
>
>There is also Oz. http://www.mozart-oz.org/ - I've long
>dreamt of a language with some of the semantics of Oz
>and Python syntax and libs. continuations and uthreads
>seem to bring python somewhat closer to Oz.
>
...
I second this, while the ideas may be complex continuations provide some
really useful idioms eg iterators etc
-- 
Robin Becker


From tim.hochberg at ieee.org  Fri Apr  7 12:35:31 2000
From: tim.hochberg at ieee.org (Tim Hochberg)
Date: Fri, 07 Apr 2000 16:35:31 GMT
Subject: reduce versus recursion? what am I missing...
References: <YznH4.4508$HG1.128287@nnrp1.uunet.ca>
Message-ID: <n7oH4.31787$U4.139420@news1.rdc1.az.home.com>

> # I am trying to figure out why function d (recursive) and e (reduce)
> # DON'T do the SAME thing:
>
> def d(*tup):
>     "why doesn't this work?"
>     if len(tup)>1:
>         return d(tup[1:])+abs(tup[0]-tup[1])
>     else:
>         return 0

There are two problems here. First, you need to call apply(d, x) here
instead of d(x), on the second call, the function sees a singleton tuple
that looks like (tup[1:],) where tup is the original tuple.

Second, the recursive problem is not doing the same thing as the reduce: the
first element of the tuple passed down the chain needs to be made equal to
the difference. I believe the following does what you want, although it's a
little akward.

def d(*tup):
    if len(tup)>1:
        l = list(tup[1:])
        l[0] = abs(tup[0]-tup[1])
        return apply(d,l)
    else:
        return tup[0]

> l = (1,3,5,7,9)
>
> def e(*tup):
>     "it works"
>     return reduce(lambda x,y:abs(x-y),tup,0)
>
> print "--- test:"
> print "recursion: ",apply(d,l)
> print "reduce:    ",apply(e,l)
>
> #---
> #Warren
> #
> # output:
> #--- test:
> #recursion:  2     ( appears not to actually be recursing more than 1
> level!!!???)
> #reduce:     5
>
>
>




From garabik at center.fmph.uniba.spam.sk  Tue Apr 25 07:05:20 2000
From: garabik at center.fmph.uniba.spam.sk (Radovan Garabik)
Date: 25 Apr 2000 11:05:20 GMT
Subject: new linereading standard?
References: <3904D21B.333DD930@visionart.com> <390510C9.A744D22E@yahoo.com>
Message-ID: <956660695.1945@cdwork.cvt.stuba.sk>

Al-Amerrho H. Amerin <al_amerin at yahoo.com> wrote:
> Pete Shinners wrote:

>>
>> now i can code in a preferred style
>>
>> myfile = open('myfile.txt', 'r')
>> for line in filelines(myfile):
>>     print line
>> myfile.close()
>>
>> as opposed to
>>
>> myfile = open('myfile.txt', 'r')
>> while 1:
>>     line = myfile.readline()
>>     if not line: break
>>     print line
>> myfile.close()
>>
>>
>> _____Pete Shinners______
>>    pete at visionart.com

> A beginner myself, I would write it like this:

> inputlines = open('myfile.txt').readlines()
> for line in inputlines:
>     print line

> Instead of the 'while 1: ....break' thingy.

...with the disadvantage that this will read the whole 
file into memory, which can be a bit difficult to handle when it 
is several MB long...


-- 
 ----------------------------------------------------------
| Radovan Garabik       http://www.fmph.uniba.sk/~garabik/ |
| __..--^^^--..__       garabik at center.fmph.uniba.sk       |
 ----------------------------------------------------------


From dalke at acm.org  Sat Apr  1 15:43:20 2000
From: dalke at acm.org (Andrew Dalke)
Date: Sat, 1 Apr 2000 13:43:20 -0700
Subject: why did string.py change in 1.6?
References: <8c4mdr$5p7$1@slb6.atl.mindspring.net> <wDnF4.4506$74.72019@newsc.telia.net>
Message-ID: <8c5mvv$bn4$1@slb6.atl.mindspring.net>

Fredrik Lundh wrote:
> python 1.6 has two string types.  the 'strop' module
> only supports one of them.

Ah, I get it.  The new string library allows both string
and unicode list elements and seperator; returning string
only if both terms are string.

My concern about performance disappears because I can get
the join method of the seperator, as in this hypothetical
example:

  join = "\t".join
  for i in range(data.shape[0]):
    print join(data[i])

Thanks!

                        Andrew
                        dalke at acm.org





From zorro at zipzap.ch  Tue Apr 18 03:58:06 2000
From: zorro at zipzap.ch (Boris Borcic)
Date: Tue, 18 Apr 2000 09:58:06 +0200
Subject: [ANNOUNCE] Garbage collection for Python
References: <20000407221750.A29279@acs.ucalgary.ca> <m3r9cagx3c.fsf@atrus.jesus.cam.ac.uk> <l03102809b51d41895a80@[193.78.237.135]> <38F9F01B.57273052@tismer.com> <m33domez7n.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <38FC158E.9AC3AD79@zipzap.ch>

Michael Hudson wrote:
> 
>     Simplicity follows complexity, not the other way round.

A good illustration being given by the 10 years or so it took
*after* man's first step on the moon to devise personal delta
wings (not to speak of parachute wings).

BB


From phd at phd.russ.ru  Wed Apr 12 04:27:33 2000
From: phd at phd.russ.ru (Oleg Broytmann)
Date: Wed, 12 Apr 2000 08:27:33 +0000 (GMT)
Subject: simple httplib and urllib timeout question
In-Reply-To: <Pine.BSF.4.10.10004111337520.29635-100000@chi.pair.com>
Message-ID: <Pine.LNX.4.21.0004120814500.6832-100000@fep132.fep.ru>

Hi, there!

On Tue, 11 Apr 2000, Michal Wallace (sabren) wrote:
> meanwhile... HEY OLEG! -> have you done anything w/ this?

   I am about to release the third version of my "Bookmarks database and
Internet robot" project. It is 98% finished, I just need a bit more
testing. If you are interested, I can publish it in current beta state.

   The third version is a complete rewrite. All things had been revamped,
and rewriteen in OO manner, and made more modular and pluggable. I don't
value OO by itself, but this time the OO approach helped me to make the
code simpler, smaller, more readable and reusable.

   After releasing it, I'll continue to work on new modules and plugins.
There are thre areas for plugins - database managers, so called writers
(data dumpers that could dump data to different text formats) and robots.
   Currently I (re)implemented 2 robots - simple and forking, Forking robot
is fine, but slow - it checks one URL at a time.

   In the field of your interest - yes, I want to write more robots, at
least two - I want a threading robot, but much more than that I want an
asyncore-based robot. I haven't done anything yet; it seems I need to
rewrite the entire urllib to use asyncore, and it is hard work. Anyway I
need the work to be done, so if no one would help me - I will do it myself.
At least will TRY to do it :)

Oleg.
---- 
    Oleg Broytmann    http://members.xoom.com/phd2.1/    phd2 at earthling.net
           Programmers don't die, they just GOSUB without RETURN.




From breiter at usf.Uni-Osnabrueck.DE  Wed Apr  5 12:43:19 2000
From: breiter at usf.Uni-Osnabrueck.DE (Bernhard Reiter)
Date: 5 Apr 2000 16:43:19 GMT
Subject: Teaching the art of programming, in python
References: <20000405053920.20579.qmail@web2102.mail.yahoo.com>
    <8cfe42$o02$1@nnrp1.deja.com> <dpJG4.3744$HG1.106895@nnrp1.uunet.ca>
Message-ID: <8cfqf7$4ml$4@newsserver.rrzn.uni-hannover.de>

In article <dpJG4.3744$HG1.106895 at nnrp1.uunet.ca>,
	"Warren Postma" <embed at geocities.com> writes:

> The Python texts I've read are good References to Getting Stuff Done, but
> they Don't Teach the Art. Speaking of Dummies, was there/is there a LISP for
> Dummies, or Smalltalk for Dummies?  Can Python for Dummies be far off?

Yes, because people who do Python are no dummies.
At least I hope so.

Ranting-about-dummy-book-titles-ly-yours,
	Bernhard
-- 
Professional Service around Free Software                (intevation.net)  
The FreeGIS Project				            (freegis.org)
Association for a Free Informational Infrastructure            (ffii.org)


From deepak_marda at my-deja.com  Wed Apr 26 16:34:06 2000
From: deepak_marda at my-deja.com (deepak_marda at my-deja.com)
Date: Wed, 26 Apr 2000 20:34:06 GMT
Subject: Need help with COMObject for Zope
Message-ID: <8e7jrs$sk6$1@nnrp1.deja.com>

Hi,
I am trying to use COMObject with Zope. I have the latest versions of
Python and Zope installed on my Windows NT machine and have the Python
COM extensions installed as well. The following works fine from a
vanilla Python environment: -

python
>>>import win32com.client
>>>o=win32com.client.Dispatch('Word.Application')
>>>o.Visible = 1
>>>...
and so on...

However, when I created a COMObject in Zope with id = word and Dispatch
= Word.Application it raises the following exception: -
Error Type: com_error
Error Value: (-2147221008, 'CoInitialize has not been called.', None,
None)

I have replaced the Python installation that comes with Zope to the
latest version of Python with the COM extensions and I have also tried
restarting Zope.

Please help,
Deepak





Sent via Deja.com http://www.deja.com/
Before you buy.


From gmcm at hypernet.com  Sat Apr 22 22:04:28 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Sat, 22 Apr 2000 22:04:28 -0400
Subject: Overloading the = operator?
In-Reply-To: <vnd4s8t4mk4.fsf@camelot-new.ccs.neu.edu>
References: Jerome Chan's message of "Sun, 23 Apr 2000 00:14:02 GMT"
Message-ID: <1255667027-43185679@hypernet.com>

Justin Sheehy wrote:

> Nope.  `=' rebinds the name on the left hand side, and never even
> looks to see what that name was previously bound to.  

Except to decref it.

- Gordon



From neilh at hare.net.au  Mon Apr  3 08:12:13 2000
From: neilh at hare.net.au (Neil Hodgson)
Date: Mon, 3 Apr 2000 22:12:13 +1000
Subject: Unicode program representation
References: <015c01bf9d0b$76f9b2a0$59dd3fcb@neil> <oqem8n4zlx.fsf@titan.progiciels-bpi.ca>
Message-ID: <00ac01bf9d65$d988c460$36dd3fcb@neil>

    Fran?ois writes,

> Instead of writing:
>
>    u"hi ?"
>
> which is indeed a way to write Unicode strings using 7-bits, or even
8-bits
> in a Latin-1 environment, you might write:
>
>    unicode("hi ?")

   OK. That's not too painful.

   But then, what role do you see the u" form having?

   Neil




From tiddlerdeja at my-deja.com  Tue Apr 25 13:02:06 2000
From: tiddlerdeja at my-deja.com (tiddlerdeja at my-deja.com)
Date: Tue, 25 Apr 2000 17:02:06 GMT
Subject: VIM NT +python
Message-ID: <8e4j27$hgq$1@nnrp1.deja.com>

Does anyone here have a winNT build of the VIM editor that has python
support switched on?

I'd love to get my hands on a copy.


Sent via Deja.com http://www.deja.com/
Before you buy.


From redsand at dsuper.net  Fri Apr 21 21:10:04 2000
From: redsand at dsuper.net (Dave)
Date: Sat, 22 Apr 2000 01:10:04 GMT
Subject: Python for drivers?
Message-ID: <3900EDBE.282BC85@dsuper.net>

Can Python be used for writing device drivers? I just got BeOS 5 and it
doesn't support some video cards. I'd like to write one for the Cirrus
Logic card and I want a language that's clear and concise- Python. I've
never written device drivers. Any suggestions?
THanks



From shecter at darmstadt.gmd.de  Tue Apr 25 15:44:08 2000
From: shecter at darmstadt.gmd.de (Robb Shecter)
Date: Tue, 25 Apr 2000 21:44:08 +0200
Subject: XML Schemas and Python?
Message-ID: <3905F588.124AF24E@darmstadt.gmd.de>

Hi,

Has anyone done anything with schemas?  I have a research project
originally written in Java, but the more I work on it, the more it
mutates into a Python project... :)

Anyhow, if there's any sample code or articles out there, I'd like to
hear about it.

Thanks,
Robb


From parkw at better.net  Thu Apr 27 17:16:14 2000
From: parkw at better.net (William Park)
Date: Thu, 27 Apr 2000 17:16:14 -0400
Subject: Converting = to ==
In-Reply-To: <n01O4.12725$HG1.340768@nnrp1.uunet.ca>; from embed@geocities.com on Thu, Apr 27, 2000 at 03:57:55PM -0400
References: <n01O4.12725$HG1.340768@nnrp1.uunet.ca>
Message-ID: <20000427171614.B1788@better.net>

On Thu, Apr 27, 2000 at 03:57:55PM -0400, Warren Postma wrote:
> I have an "end user query tool" that uses Python expressions, but I want
> comparisons to be able to be written either as A=B or A==B then I want to
> compare them using Python's Eval function.
> 
> Problem is if you search and replace all = with == then users have to use
> the single-equals convention, and some of them will be python programmers,
> and will want to use the "non idiot mode" of the query tool.
> 
> So i did this, but is there a better way?
> 
>     >>> string.replace( string.replace("a=b a==b", "=", "=="), "====","==")
>     a==b a==b
> 
> Warren

You can use regular expression '([^=])=([^=])', but I think your
solution is cleaner.

--William



From apjewell at bigfoot.com  Fri Apr 14 19:50:54 2000
From: apjewell at bigfoot.com (Andrew P. Jewell)
Date: Fri, 14 Apr 2000 23:50:54 GMT
Subject: The Opposite of list(a,b,c)?
Message-ID: <38F7A07F.9B0F309C@bigfoot.com>

An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20000414/ec27df7f/attachment.html>

From cut_me_out at hotmail.com  Fri Apr 14 15:42:53 2000
From: cut_me_out at hotmail.com (Alex)
Date: 14 Apr 2000 15:42:53 -0400
Subject: Attributes (+Numerical Python)
References: <meh9-986AAB.15155814042000@news.cit.cornell.edu>
Message-ID: <etd66tk31vm.fsf@w20-575-109.mit.edu>

> if the .imag attribute exists:
>    do something
> else
>    do something else
> 
> How should this be written?

if hasattr (object, 'imag'):
    do something
else:
    do something else

Alex.


From tjs at cs.monash.edu.au  Tue Apr 25 07:52:21 2000
From: tjs at cs.monash.edu.au (Toby J Sargeant)
Date: Tue, 25 Apr 2000 21:52:21 +1000
Subject: how to wrap c++ software to python?
In-Reply-To: <8e3ggq$2r1d$1@news.cz.js.cn>; from mrfx@263.net on Tue, Apr 25, 2000 at 03:09:17PM +0800
References: <8e3ggq$2r1d$1@news.cz.js.cn>
Message-ID: <20000425215220.A29950@fangorn.csse.monash.edu.au>

On Tue, Apr 25, 2000 at 03:09:17PM +0800, fmr wrote:
> I will call c++ routine in python,how to wrap
> c++ routine for python.
> 
> Is there any utils to work except SWIG.

sip, (http://www.river-bank.demon.co.uk/software) is very good. it will take
a c++ header (with a few modifications), and it will spit out a python module
that will provide a python type that mimics the c++ class. it also handles
(single) inheritance, which is a big win.

It's used as the basis of the PyQT bindings, which are also of a very high
standard.

Toby.



From gmcm at hypernet.com  Wed Apr 26 17:17:56 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Wed, 26 Apr 2000 17:17:56 -0400
Subject: global variables or inheritence
In-Reply-To: <28p8zy0irgc.fsf@lada.kom.auc.dk>
Message-ID: <1255338616-62939142@hypernet.com>

Thomas Rasmussen wrote:

> I have quite a big problem. I have two different files, where I
> in one file want to declare a variable and then in the other file
> be able to access and use this variable (no need to change it),
> but I just can't seem to get it to work... Is there any nice way
> of doing this easily? I have tried the global var but this
> doesn't seem to work.
> 
> Anyone got a good idea?

  import other
  print other.something

works just fine. Bet you used "from other import *", and then 
tried to change the value in other.

- Gordon



From donn at u.washington.edu  Tue Apr 18 12:58:36 2000
From: donn at u.washington.edu (Donn Cave)
Date: 18 Apr 2000 16:58:36 GMT
Subject: Running shell programs from Python
References: <38FC64DA.7D297563@ffi.no>
Message-ID: <8di47s$mkg$1@nntp6.u.washington.edu>

Quoth seh at ffi.no:
| I have just started with Python and have some questions.. I have some
| compiled c programs that I normally run from the shell. I can use | to
| pipe the output from one as input to the next. I would like to call the
| programs from Python. Could someone please indicate how this is done?
| Any references?

The references are mainly UNIX man pages pipe(2), fork(2), fcntl(2)
(look for dup2), exec(2) and wait(2).

Basically, the sequence is repeat ( open pipes, fork, close pipes )
After a fork, the parent and child both have both ends of each pipe
open, and it's important to close the unused "write" ends so the 
pipe will generate an EOF when its real writer exits.

Each of the forks will call execve() or a related function to invoke
the program, after forking if that's necessary.  Use dup2() to open
units 0 and 1 on the read and write pipes respectively, and close the
original units.  Trap any possible exceptions in the fork, so control
doesn't return to the caller in the fork.

While I do this with single commands, I personally don't ever do it
with a complex pipeline, because it's kind of hard to get right and
the UNIX shell just does a much better job.  If you just want to run
the shell from Python, as someone else already has mentioned, use
os.system(), os.popen() and various cool things in the popen2 module.

	Donn Cave, University Computing Services, University of Washington
	donn at u.washington.edu


From banderson at boi.hp.com  Fri Apr 14 15:37:59 2000
From: banderson at boi.hp.com (Bill Anderson)
Date: Fri, 14 Apr 2000 13:37:59 -0600
Subject: [ANNOUNCE] Garbage collection for Python
References: <20000407221750.A29279@acs.ucalgary.ca> <m3r9cagx3c.fsf@atrus.jesus.cam.ac.uk> <38F65501.665243F6@gmx.de>
Message-ID: <38F77397.45F26301@boi.hp.com>

Michael Scharf wrote:
> 
> Michael Hudson wrote:
> >
> > "Neil Schemenauer" <nascheme at enme.ucalgary.ca> writes:
> >
> > > Garbage collection for Python:
> > >
> > >     http://www.enme.ucalgary.ca/~nascheme/python/gc.html
> > > The patch is against the current CVS version of Python.  If
> > > enough people ask for a patch against the stable version of
> > > Python I will make one.  Please send comments and bug reports.
> >
> > To someone who might know: what hope is there for this to go into 1.6?
> >
> > It would be, in my opinion, be a really really really good thing; for
> > one thing it would help in language wars...
> 
> I would *love* to see this in 1.6!!!!


I would love to see a patch for 1.5.2!
:)

-- 
Bill Anderson 			Linux/Unix Administrator, Security Analyst
ESBU (ARC)			bill_anderson at boi.hp.com
My opinions are just that; _my_ opinions.


From m.faassen at vet.uu.nl  Tue Apr 25 11:04:23 2000
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 25 Apr 2000 15:04:23 GMT
Subject: Distribution
References: <39013f48@flexnet239.uunt.net>
Message-ID: <8e4c5n$p7q$1@newshost.accu.uu.nl>

Highdesertman <HighdesertmanREMOVE at yahoo.com> wrote:
[snip]
> If I use python to develop an application for the windows 9x environment,
>  I have already downloaded the appropriate libraries and version) what form
> does the finished program take? How does it get distributed?

Interesting to you may be Gordon McMillan's installer. See this Parnassus
link:

http://www.vex.net/parnassus/apyllo.py?i=22208368

Unfortunately the link referred to at Parnassus is currently down, as 
it's on the Python starship, which has crashed and its currently being
repeared (the warp^H^H^H^H hard drive gave up).

Also interesting may be the Python distutils:

http://www.python.org/sigs/distutils-sig/

> I remember in
> the early days of BASIC programming, (before compiled basic) your program
> was just raw code that could be (often accidentally) altered or otherwise
> corrupted.

Must be the alien whitespace eating nanoviruses accidentally altering and
corrupting the raw code. If you are in an environment where code can be
accidentally corrupted, choosing for source or binary code won't matter;
binary code can be corrupted as easily and is even harder to repair.

> In order for your basic program to be used, the user had to have
> a basic compiler on board and they loaded your raw source from a text file
> and executed it. Is this how Python is used?

Python is often used in a similar way. There is a Python interpreter on the
system, and you just execute some .py file by typing: python foo.py, or
on Windows, double-clicking on foo.py

Of course then the Python interpreter needs to be installed. There are
various installer solutions which install their own interpreters, Gordon's
being one of them. Perhaps others can add to this thread; I haven't worked
with any.

[snip]
> If Python is dependent on it's interpreter to run, and I am only programming
> for my own needs, then that dosen't represent a weakness.

It also doesn't represent a weakness if you're writing code for other Python
developers to use; for instance a reusable module that does something
interesting. It's also not a big deal if you're using Python to code CGI
scripts for a website; Python does need to be installed on the webserver
then, of course, but that's fairly easy to control.

Note that one idea behind open source is that easy access to the source
makes it easy for other developers to check out and improve your code. Please
keep that in mind when thinking of installation solutions; hiding the source
will make it harder for potentially helpful developers. Also such developers
may not 'trust' closed-source code as much.

> But since Python
> remains relatively unknown as yet, the average user isn't likely to have a
> python interpreter handy.

That's right. That's even a problem for Java, a well-known and well-hyped
language. I've been told that once a Java environment is installed, you can
run all the Java code in the world with it. Now only to install a 
proper Java. I heard there were various versions.. Oh, and on Linux, too! :)

> In addition, one of the real advantages of the
> development of the compiler was that it gave your code a level of protection
> against being tampered with and made it easily useable by a wide audience.

I don't see why you're so afraid people will be tampering with code. It's
been shown that allowing other people to tamper with your code gives you
the possibility to create rather important software -- Linux, Apache, 
Python, and so on. It doesn't have to happen, but it won't happen at all
if you don't open your source. Of course you may make a distinction between
a developer's edition and a user's edition of your program, but why make it
harder for developers to get at your code when you don't have to?

That said, the other argument; more easy to use for normal users, may
be true. It's a large step for the average user to install a Python
interpreter, while they usually don't mind running some installer to
get access to some program. So bundling Python with the program, as
various packaging tools do, may be a good idea in that case. Disadvantage
of this approach is that a user may end up with 20 versions of the Python
interpreter on his system. :)

It's interesting how philosophies about source code changed since the
days of MSDOS; I remember thinking about compilers in the same way.
"A way to make my code seem 'real'!" 

Regards and good luck!

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From darcy at vex.net  Mon Apr 24 22:17:20 2000
From: darcy at vex.net (D'Arcy J.M. Cain)
Date: 25 Apr 2000 02:17:20 GMT
Subject: Trouble compiling PyGreSQL 2.4 on Linux
References: <Pine.LNX.4.21.0004241211380.6019-100000@fep132.fep.ru> <g0sbh8jd.fsf@jumbo.skott.dk>
Message-ID: <8e2v7g$1so4$1@news.tht.net>

Martin Sk?tt <mskott at image.dk> wrote:
> I know it works, but having a Linux README which didn't work bothered
> me immensly.

I have always depended on the kindness of strangers.

Context diffs to correct it always welcome.  I don't use Linux myself
so please send to the mailing list for peer review.

-- 
D'Arcy J.M. Cain <darcy at vex.net>           |  Democracy is three wolves
http://www.vex.net/                        |  and a sheep voting on         
+1 416 425 1212     (DoD#0082)    (eNTP)   |  what's for dinner.


From parkw at better.net  Tue Apr 25 16:50:50 2000
From: parkw at better.net (William Park)
Date: Tue, 25 Apr 2000 16:50:50 -0400
Subject: Need advice performing a search
In-Reply-To: <3905F795.A536CFA7@stsci.edu>; from karab@stsci.edu on Tue, Apr 25, 2000 at 07:58:38PM +0000
References: <3905F795.A536CFA7@stsci.edu>
Message-ID: <20000425165050.A222@better.net>

On Tue, Apr 25, 2000 at 07:58:38PM +0000, Kia Arab wrote:
> 
> I am trying to search a file for two strings. If *either* string is
> found, I need the line in which the string was found to be returned  (or
> somehow extract the line). It should keep searching and keep returning
> the lines where it finds either string, until the end of file is
> reached.
> 
> Any suggestions?

On Unix, 'grep' is what you want.  But, I guess you can read line by
line, and test for the 2 strings.

--William



From embed at geocities.com  Tue Apr  4 14:42:41 2000
From: embed at geocities.com (Warren Postma)
Date: Tue, 4 Apr 2000 13:42:41 -0500
Subject: blessed be Guido, for "5 <= X <= 10" does what it should...
Message-ID: <3TpG4.3349$HG1.98231@nnrp1.uunet.ca>

<X-DISCLAIMER: I'm in a weird mood.>

I was just playing around today,and it occurred to me to test the
'transitive' properties of comparison operators. Ie, writing on a
chalkboard, you might say X is in the range 5..9 this way:

        5 <= X <= 9

(Actually I'd write the <= as '?' (Unicode U+2264) but using such things in
our code would seriously cramp portability.  Hey, is the APL symbol set
anywhere in Unicode!?)

In some programming languages, which reduce each comparison to a boolean
True or False, using the precedence rules, the result is:

        n = (5<=X)
        n <= 9

Since 'n' is probably 0 or 1 (commodore 64 basic used -1 for true, remember
that?) anywyas, 'n' is always <= 9, which means the whole expression is
always false.

If precedence rules were right to left:
        n = (X<=9)
        5 <= n

In that case, the statement is always false since n is always less than 5.

So, Us Really Smart C Programmers Who Have Been Burned By This Before
(hereafter, 'URSCPWHBBBTB') have long since taken to writing such things as:

    (5 <= X) && (X <= 9)

Anyways, that was C, so what about Python?

I decided to test if (a) Python could handle 'transitive' ways of writing
comparisons, and (b) if it emulated what most other languages do when you
explicity bracket the comparisons, and force sub-expression evaluation.

Here's a sample bit of code: It prints out the comparison, and the results


# test i<=j<=k behaviour, compare to (i<=j)<=k and i<=(j<=k)
for i in range(0,4):
 for j in range(0,4):
   for k in range(0,4):
     n = i <= j <= k # n = 1 if j between i and k
     x = ( i <= j) <= k # bug 1 ?
     y =  i <= (j <= k) # bug 2 ?
     print i,"<=",j,"<=",k," : ",n,"  ",x,y


Here's a Hack for the Mighty Pythonista to consider: Can you construct a
program that you can call like this:

    test_transitive( 0,100, '<=', 5 ) # generate a<=b<=c<=d<=e for range
0..100

Anyways, blessed be Guido, for near as I can figure it, he hath got it
right! :-)

Warren




From bryanNObrSPAM at dunsinane.net.invalid  Mon Apr  3 11:36:20 2000
From: bryanNObrSPAM at dunsinane.net.invalid (BryanBr)
Date: Mon, 03 Apr 2000 08:36:20 -0700
Subject: Python CGI based web game
Message-ID: <101a63d2.8b02be5a@usw-ex0109-069.remarq.com>

I am writing a python CGI based web game based on:

http://merchant.shareplay.com.

I am using a simple CGI framework and HTMLgen.

Anyone interested in helping out, email me.  After two days
of coding I have a workable game model.

I have a question: How do I redirect http requests with
python cgi?

Sending "Status 302" works from a directly requested python
cgi file.  However, I call all of my pages from a cgi
framework that actually calls my pages with the function
execfile.  When the page that has been called with execfile
output "Status 302" no redirect occurs.


* Sent from AltaVista http://www.altavista.com Where you can also find related Web Pages, Images, Audios, Videos, News, and Shopping.  Smart is Beautiful


From just at letterror.com  Sat Apr 29 14:12:04 2000
From: just at letterror.com (Just van Rossum)
Date: Sat, 29 Apr 2000 19:12:04 +0100
Subject: Where can I find NumPy
In-Reply-To: <8ef4pn$jge$1@kopp.stud.ntnu.no>
Message-ID: <l03102802b530d65f195d@[193.78.237.122]>

At 5:05 PM +0000 29-04-2000, Jan Erik Breimo wrote:
>The official site (ftp-icf.llnl.gov) hasn't been responding for more than
>a week, so are there any other sites where I can find the source
>files for NumPy?

numpy.sourceforge.net.

Just





From mikael at isy.liu.se  Wed Apr  5 03:05:31 2000
From: mikael at isy.liu.se (Mikael Olofsson)
Date: Wed, 05 Apr 2000 09:05:31 +0200 (MET DST)
Subject: Why should I switch to Python? - Infinity of Primes
In-Reply-To: <000d01bf9ec5$3fb90820$3e2d153f@tim>
Message-ID: <XFMail.000405090531.mikael@isy.liu.se>

On 05-Apr-00 Tim Peters wrote:
 >  [Greg Ewing]
 > > ...
 > > If someone actually came up with a formula for constructing
 > > primes, it would be rather large news -- isn't that one of
 > > the Big Unsolved Problems?
 >  
 >  Several such formulas were developed in the 70's, although they're
 >  strained.
 >  Paulo Ribenboim's accessible "The Little Book of Big Primes"
 >  (Springer-Verlag, 1991) gives a bunch of 'em in chapter 3, along with 6
 >  other proofs of the infinitude of the primes.

Yeah, like the one taking two integers as arguments, p and q, and then
generating p-1 if that happens to be a prime (for some, but not for all q), 
and generating 2 in all other cases...

It certainly is a function that generates primes for all inputs.

/Mikael

-----------------------------------------------------------------------
E-Mail:  Mikael Olofsson <mikael at isy.liu.se>
WWW:     http://www.dtr.isy.liu.se/dtr/staff/mikael
Phone:   +46 - (0)13 - 28 1343
Telefax: +46 - (0)13 - 28 1339
Date:    05-Apr-00
Time:    09:01:45

This message was sent by XF-Mail.
-----------------------------------------------------------------------



From nobody at nowhere.nohow  Sat Apr 22 09:47:24 2000
From: nobody at nowhere.nohow (Grant Edwards)
Date: Sat, 22 Apr 2000 13:47:24 GMT
Subject: code not true?
References: <8ds3ku$q0l$1@nnrp1.deja.com> <3901aa0f.0@news.proweb.co.uk>
Message-ID: <M3iM4.518$wJ1.17829@ptah.visi.com>

On Sat, 22 Apr 2000 14:27:57 +0100, ng at hardlight.couk.com <ng at hardlight.couk.com> wrote:
>interestingly enough on both Win32 and Linux

>>>> (0.7 * 10) == (0.07 * 100)
>0
>
>>>>  (0.07 * 10) == (0.007 * 100)
>1
>
>>>> (0.07 * 100) == (0.007 * 1000)
>0
>
>>>> (0.07 * 1000.0) == (0.007 * 10000.0)
>1

You've proven once again that testing floating point numbers
for equality is not a particularly useful operation.  ;)

This seems to be something that we were all told as freshmen
but stumble over later anyway.  One wonders if a safe language
intended for use by beginners should even have an equality
operator for floating point objects.

-- 
Grant Edwards                   grante             Yow!  Now KEN is having
                                  at               a MENTAL CRISIS beacuse
                               visi.com            his "R.V." PAYMENTS are
                                                   OVER-DUE!!


From jwbnews at scandaroon.com  Sun Apr  2 18:09:48 2000
From: jwbnews at scandaroon.com (John W. Baxter)
Date: Sun, 02 Apr 2000 15:09:48 -0700
Subject: New Features in Python 1.6
References: <200004011740.MAA04675@eric.cnri.reston.va.us> <8c84n9$pf0$3@newshost.accu.uu.nl>
Message-ID: <jwbnews-372AE6.15094802042000@news.olympus.net>

In article <8c84n9$pf0$3 at newshost.accu.uu.nl>, m.faassen at vet.uu.nl 
(Martijn Faassen) wrote:

> You almost got me! I didn't read the rest of the thread, and I was already
> thinking about why all these would be bad ideas, and then I realized it
> was posted on April 1. Good one!

He did get me, but I saw the post when there was no "rest of the thread" 
yet on my news server.

Well done, Guido!

   --John

-- 
John W. Baxter   Port Ludlow, WA USA  jwbnews at scandaroon.com


From gregholmes at my-deja.com  Wed Apr 19 06:15:48 2000
From: gregholmes at my-deja.com (gregholmes at my-deja.com)
Date: Wed, 19 Apr 2000 10:15:48 GMT
Subject: How to write to printer in Windows?
References: <38FC43DF.78A38D5D@tcac.net>
Message-ID: <8dk10c$3e0$1@nnrp1.deja.com>

  Albert Wagner <alwagner at tcac.net> wrote:
> How can I write(aLine) to a printer in Windows?  I tried this:
>
> from win32pipe import popen
> :
> self.printer = popen('lpt1' 'w')
> :
> self.printer.write(aLine)
>
> but get a general protection fault.  Any help will be greatly
> appreciated.
>
> --
> Small is Beautiful
>
Here is a snippet of how I did it:

import win32ui
import win32con
self.dc = win32ui.CreateDC()
self.dc.CreatePrinterDC()   # ties it to your default printer
horzres=self.dc.GetDeviceCaps(win32con.HORZRES)
vertres=self.dc.GetDeviceCaps(win32con.VERTRES)
self.printxscale=(horzres/300)*0.8
self.printyscale=(vertres/300)*0.8
self.redpen = win32ui.CreatePen(win32con.PS_SOLID, 5, self.red)
self.blackpen = win32ui.CreatePen(win32con.PS_SOLID, 5, self.black)
self.dc.StartDoc('Graph')
#self.dc.SetMapMode(win32con.MM_LOENGLISH)
self.dc.StartPage()
self.dc.SelectObject(self.blackpen)
self.dc.SetTextAlign(win32con.TA_CENTER|win32con.TA_BASELINE)
self.dc.MoveTo((0+self.xoffset)*self.printxscale,10*self.printyscale)
self.dc.LineTo((0+self.xoffset)*self.printxscale,210*self.printyscale)
self.dc.MoveTo((0+self.xoffset)*self.printxscale,210*self.printyscale)
self.dc.LineTo((200+self.xoffset)*self.printxscale,210*self.printyscale)
self.dc.MoveTo((100+self.xoffset)*self.printxscale,210*self.printyscale)
self.dc.LineTo((100+self.xoffset)*self.printxscale,220*self.printyscale)
self.dc.MoveTo((50+self.xoffset)*self.printxscale,210*self.printyscale)
self.dc.LineTo((50+self.xoffset)*self.printxscale,220*self.printyscale)
self.dc.MoveTo((150+self.xoffset)*self.printxscale,210*self.printyscale)
self.dc.LineTo((150+self.xoffset)*self.printxscale,220*self.printyscale)
self.dc.MoveTo((0+self.xoffset)*self.printxscale,110*self.printyscale)
self.dc.LineTo((0+self.xoffset-10*self.printxscale,110*self.printyscale)
self.dc.MoveTo((0+self.xoffset)*self.printxscale,10*self.printyscale)
self.dc.LineTo((0+self.xoffset-10)*self.printxscale,10*self.printyscale)
self.dc.MoveTo((200+self.xoffset)*self.printxscale,210*self.printyscale)
self.dc.LineTo((200+self.xoffset)*self.printxscale,220*self.printyscale)
self.dc.MoveTo((0+self.xoffset)*self.printxscale,160*self.printyscale)
self.dc.LineTo((0+self.xoffset-10*self.printxscale,160*self.printyscale)
self.dc.MoveTo((0+self.xoffset)*self.printxscale,60*self.printyscale)
self.dc.LineTo((0+self.xoffset-10)*self.printxscale,60*self.printyscale)
-------
That prints some lines (the axes of a graph).

To do text you can do:
--------------
self.dc.TextOut
((100+self.xoffset*self.printxscale,230*self.printyscale, "some text")
---------------

To end the page, the document, and clean up:
--------------
self.dc.EndPage()
self.dc.EndDoc()
print 'sent to printer'
del self.dc
-----------------

Hope this helps.


Sent via Deja.com http://www.deja.com/
Before you buy.


From vkolosov at unitedmedia.com  Mon Apr  3 09:48:10 2000
From: vkolosov at unitedmedia.com (Kolosov, Victor)
Date: Mon, 3 Apr 2000 09:48:10 -0400 
Subject: Python 1.6 alpha 1 released
Message-ID: <461182F0B9B7D111B4000008C7282ADC22B0FC@nynt04.umny.scripps.com>

	He became a god, gods do not talk to people - it is one way
relationship.
	:-)


> James Logajan wrote:
> > Any reason Guido doesn't post to this newsgroup like he used to?
> 
> well, what do *you* think is the reason?
> 



From jmg at ecs.soton.ac.uk  Wed Apr  5 12:49:53 2000
From: jmg at ecs.soton.ac.uk (Jacek Generowicz)
Date: Wed, 05 Apr 2000 17:49:53 +0100
Subject: Pass by reference ?
References: <38E864C9.E1EE94F7@ecs.soton.ac.uk> <38EAF8D8.4B7C43B4@ecs.soton.ac.uk> <38EB56AE.F938CE06@acm.org> <m33dp0muk9.fsf@atrus.jesus.cam.ac.uk> <38EB63BC.FE03BA31@ecs.soton.ac.uk> <m3ya6sle0j.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <38EB6EB1.9759397C@ecs.soton.ac.uk>

Michael Hudson wrote:

> The code you posted only confuses if you have a flawed understanding of
> assignment in Python

That is exactly the point. My original confusion arose precisely because I had
made an incorrect assumption about assignment in Python (an assumption which, I
suspect, many newcomers to the language will make until they learn the truth).
Correcting this error is far more constructive than discussing what name should be
given to the semantics of argument passing.

Jacek






From andy at petdance.com  Tue Apr 25 09:08:51 2000
From: andy at petdance.com (Andy Lester)
Date: Tue, 25 Apr 2000 08:08:51 -0500 (CST)
Subject: Programming Python 2nd Edition?- Mark Lutz
In-Reply-To: <20000331113449.A2053@sirius.ftsm.ukm.my>
Message-ID: <Pine.WNT.4.21.0004250808350.240-100000@nt_alester>

> If true, probably I could wait a while longer to spend my hard earned cash
> to buy ONE book to learn python proper.

Pick up "The Quick Python Book".  It's quite excellent.

xoxo,
andy

--
Andy Lester, andy at petdance.com, http://www.petdance.com
    "Inflammable material is planted in my head
     It's a suspect device that's left two thousand dead."
    "Hey, is this the new Green Day?"




From ngps at madcap.dyndns.org  Fri Apr  7 12:34:34 2000
From: ngps at madcap.dyndns.org (Ng Pheng Siong)
Date: 7 Apr 2000 16:34:34 GMT
Subject: Paid to work in Python? (was: Re: Choice Of Language)
References: <Uqvw4.2261$yL2.28868@typhoon.southeast.rr.com> <8c3ats$b5f$1@nntp.Stanford.EDU> <slrn8eact2.sq7.nobody@bambi.visi.com> <38e5f8ed.29659873@news.nikoma.de>
Message-ID: <8cl2mq$971$1@mawar.singnet.com.sg>

According to Philip 'Yes, that's my address' Newton <nospam.newton at gmx.li>:
> (The German term "eierlegende Wollmilchsau"[1] comes to mind.)
>
> [1] literally, "egg-laying wool-milk-sow [female pig]", i. e., something
> that tries to do too much at once.

I once saw a German term that was translated to "improve-worse-ment."

Can some one provide the German word, please?

TIA. Cheers.

-- 
Ng Pheng Siong <ngps at post1.com> * http://www.post1.com/home/ngps



From mhammond at skippinet.com.au  Wed Apr 26 20:38:12 2000
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Thu, 27 Apr 2000 00:38:12 GMT
Subject: [Win32] Retrieve icon for file type?
References: <CC59868A0A76D311B1F50090277B321C1F32B6@VRTMAIL>
Message-ID: <UZLN4.5084$v85.31003@news-server.bigpond.net.au>

"Mike Fletcher" <mfletch at tpresence.com> wrote in message
news:CC59868A0A76D311B1F50090277B321C1F32B6 at VRTMAIL...

> Incidentally, the IShellFolder/IShellIcon stuff looks like the "right"
way
> to do this, I just don't see any way to touch it from Python (apparently
you
> need the functions "SHGetDesktopFolder" etceteras in shell32.dll to get
> anywhere with it.

These are in win32com.shell.shell.

Note that ExtractIcon/DestroyIcon will be in the next version of
win32gui...

Mark.





From rcc at nospamthanks_jennic.com  Tue Apr 25 09:53:28 2000
From: rcc at nospamthanks_jennic.com (Robert Cragie)
Date: Tue, 25 Apr 2000 13:53:28 GMT
Subject: Unsigned integer arithmetic
Message-ID: <newscache$51sktf$rzc$1@jenpc07.jennic.co.uk>

Hi all,

I'm having a lot of trouble with arithmetic manipulation of numbers above
0x7fffffff. Whilst an integer is represented in a 32 bit word, as you'd
expect, 0x80000000 - 0xffffffff are treated as negative numbers. However, I
want to treat a 32 bit number as an unsigned value, and be able to do i = i
+ 1 for an integer number above 0x7fffffff without getting an overflow.

I'm finding I have to use the inefficient longs for the arithmetic, and do
clumsy conversions using the struct module, e.g.

import struct

# integer to unsigned long
# in the strings, it's an upper case 'eye' and a lower case 'ell' in that
order
myUlong = struct.unpack('I', struct.pack('l',myInt))[0]

# unsigned long to integer
# in the strings, it's a lower case 'ell' and an upper case 'eye' in that
order
myInt = struct.unpack('l', struct.pack('I',myUlong))[0]

Also, I can't seem to print a long using

print '0x%08x' % myLong

if 'myLong' is greater than 0x7fffffff - I get the rather obscure 'long int
too long to convert'.

Is there a better way to do all this?

TIA

Robert Cragie





From claird at starbase.neosoft.com  Wed Apr  5 00:04:01 2000
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 4 Apr 2000 23:04:01 -0500
Subject: What is python's language level?
References: <38EA1B8F.A797A130@lmco.com> <roy-D857DA.21212304042000@netnews.nyu.edu> <38EA9CD4.3473B910@callware.com> <roy-CF57ED.23102504042000@netnews.nyu.edu>
Message-ID: <ED6060DB649A6FED.125EA090C5A03FDE.AA052F5C80E50E94@lp.airnews.net>

In article <roy-CF57ED.23102504042000 at netnews.nyu.edu>,
Roy Smith  <roy at popmail.med.nyu.edu> wrote:
			.
			.
			.
>Have you ever looked at the output of a bib | tbl | eqn pipeline?

Are you kids still using that as a pick-up line?
-- 

Cameron Laird <claird at NeoSoft.com>
Business:  http://www.Phaseit.net
Personal:  http://starbase.neosoft.com/~claird/home.html


From cmfinlay at magnet.com.au  Mon Apr 17 11:39:47 2000
From: cmfinlay at magnet.com.au (cmfinlay at magnet.com.au)
Date: Tue, 18 Apr 2000 01:39:47 +1000
Subject: How do I make a constant global variable in Python Please
Message-ID: <vCj7OA2LfMfb=znR7yTHOvmiZAkd@4ax.com>

How do I make a constant global variable in Python Please.
or just a constant.
or just a global.

I bought a Python Book. however I can't find this info in it.

Please e-mail me at cmfinlay at magnet.com.au


From mwh21 at cam.ac.uk  Wed Apr 19 05:10:15 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 19 Apr 2000 10:10:15 +0100
Subject: Forward References?
References: <eviltofu-4C5688.21593817042000@news.earthlink.net> <oh3L4.1227$zk4.23905@news.uswest.net>
Message-ID: <m33doijw20.fsf@atrus.jesus.cam.ac.uk>

"Jeff Massung" <jmassung at magpiesystems.com> writes:

> Also, someone mentioned above having: self.y() (or self.y(self)), my
> question is what is the difference between a.y and self.y?

a.y(self) almost certainly doesn't do what you want in the presence of
inheritance.

Cheers,
M.

-- 
81. In computing, turning the obvious into the useful is a living
    definition of the word "frustration".
     -- Alan Perlis, http://www.cs.yale.edu/~perlis-alan/quotes.html


From moshez at math.huji.ac.il  Wed Apr 26 04:21:07 2000
From: moshez at math.huji.ac.il (Moshe Zadka)
Date: Wed, 26 Apr 2000 11:21:07 +0300 (IDT)
Subject: os.path.walk() question.
In-Reply-To: <8e55m0$pvl$1@hammer.msfc.nasa.gov>
Message-ID: <Pine.GSO.4.10.10004261119590.3499-100000@sundial>

On 25 Apr 2000, Andrew McDowell wrote:

> I need to find out if a file exists under a directory hierarchy.
> The hierarchy could be quite large so I don't want to continue recursing
> through the hierarchy if the file is found one or two directories down.   Is
> there a way that you can tell os.path.walk() to stop searching?

Here's a quick-n-dirty hack: throw an exception, containing the name
of the file you found.

--
Moshe Zadka <mzadka at geocities.com>. 
http://www.oreilly.com/news/prescod_0300.html
http://www.linux.org.il -- we put the penguin in .com




From gmcm at hypernet.com  Fri Apr 21 12:47:50 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Fri, 21 Apr 2000 12:47:50 -0400
Subject: The Simple Economics of Open Source
In-Reply-To: <3900611F.3B525BDA@cfdrc.com>
Message-ID: <1255786824-35980039@hypernet.com>

Why does everyone miss the main point?

I (and you) are literally 100s of times more productive because 
of open source. Imagine having to either invent it all yourself, 
or duct tape together expensive half-assed closed "solutions". 
Shudder.

Developers who share code are *enormously* more productive 
than those who don't. So the range of problems that can be 
attacked increases geometrically, so computers penetrate 
new areas, so geeks stand a better chance of making a living.

Win-win situations are rarely profitably analyzed by 
competitive models.


- Gordon



From phd at phd.russ.ru  Tue Apr 18 05:16:03 2000
From: phd at phd.russ.ru (Oleg Broytmann)
Date: Tue, 18 Apr 2000 09:16:03 +0000 (GMT)
Subject: Filtering web proxy
In-Reply-To: <20000417181037.B8191@bigskytel.com>
Message-ID: <Pine.LNX.4.21.0004180914080.5479-100000@fep132.fep.ru>

Hello!

   Thanks to all who replied. I'll look into all the code. Right now proxy3
(MURI) looks best, with "munchy" follows.

On Mon, 17 Apr 2000, David Porter wrote:
> If the junkbuster version I mentioned is not satisfactory to you I would be
> highly surprised, give it a go!

   It doesn't block Javascript.

Oleg.            (All opinions are mine and not of my employer)
---- 
    Oleg Broytmann      Foundation for Effective Policies      phd at phd.russ.ru
           Programmers don't die, they just GOSUB without RETURN.





From gmcm at hypernet.com  Wed Apr  5 21:45:42 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Wed, 5 Apr 2000 21:45:42 -0400
Subject: Pass by reference ?
In-Reply-To: <38EBCD7D.7030602@acm.org>
Message-ID: <1257136951-12346320@hypernet.com>

Robert W. Cunningham wrote:

> Gordon McMillan wrote:
> 
> > Robert W. Cunningham wrote:
> > 
> > > The CS models of "Pass By Reference" and "Pass By Value" are fairly well understood,
> > > but they do not seem to map simply and directly to Python.  
> > 
> > Grrr. Pascal causes brain damage. Python is exactly like Java 
> > (with the exception of having no primitive types).
> >  
> > These notions should state *what* is being passed. In Pascal, 
> > that's "value of a slot" or "reference to a slot". To be precise, 
> > in Python it is *always* "value of a reference", (think "pointer 
> > value", not "value of that which is pointed to").
> 
> <sigh>  Let's go back to CS101 and start with at least a MINIMAL set of 
> common words, definitions and concepts.
> 
> 1. Computers contain and operate upon REPRESENTATIONS only.

Nah. They operate on voltage levels represented by 1s and 0s. 
The fact that higher levels of representation are built on top of 
that is a useful fantasy.
 
> That is to say, there is no such thing as the "integer number five" in a 
> computer, only that which we CHOOSE to REPRESENT the VALUE of the purely 
> math-theoretical notion of "5".

See how you take the fantasy for reality? No way does any 
computer ever manipulate or represent an integer. The closest 
you'll come is an element of a finite algebraic group. But 
computer languages let us believe they are integers, for awhile 
at least.
 

>    def Nuke(w):
>      print w,
>      w=None
>      print w
>
> Let's call each of these functions:
> 
>    Nuke(a)
>    Nuke(b)
>    Nuke(c)
>    Nuke(d)
> 
> Look at the output:
> 
>    5 None
>    5 None
>    [5] None
>    {"five":5} None
> 
> And then look at the original values:
> 
>    print a,b,c,d
> 
>    5 5 [5] {"five":5}
> 
> None changed!  Therefore, representations the functions modified were 
> DIFFERENT from those passed, though it is clear that the correct VALUE 
> was passed.

And in each case the VALUE OF THE REFERENCE was 
nuked. You didn't modify any referent. You reassigned.
 
> This behavior meets the classical definition of "Pass By Value".

Yes, the value of the reference. When you say "a=5" in 
Python, a holds a reference to an object. The object holds a 
representation of something we take to be the number 5. 
Passing a to a function is done by copying the value of a (not 
the object a references). Therefor assigning to a in the called 
function has no effect on the caller. If the object is mutable, 
changes to the object will be visible in the caller.

This is NOT unique to Python. Technically, this is "call by 
value", but it's "value of the reference". However, because 
changes to mutable objects are visible to the caller, people 
often call it "call by reference", which, as you've observed, it's 
not.

The real problem is that "call by value" and "call by reference" 
have an unspoken assumption that variables are slots.
 
> What other tests have I omitted, and what do they yield?

You've missed a level of indirection. The label is not the object.

- Gordon



From donn at u.washington.edu  Wed Apr 26 13:43:31 2000
From: donn at u.washington.edu (Donn Cave)
Date: 26 Apr 2000 17:43:31 GMT
Subject: Unsigned integer arithmetic
References: <newscache$3s7mtf$gbg$1@jenpc07.jennic.co.uk> <newscache$51sktf$rzc$1@jenpc07.jennic.co.uk> <PxvN4.1335$sf5.16905@news.corecomm.net>
Message-ID: <8e79s3$vv2$1@nntp6.u.washington.edu>

Quoth "Robert Cragie" <rcc at nospamthanks_jennic.com>:
...
| I guess I'm showing my C background here, but it would be nicer to have a
| notion of unsigned numbers and also to be able to print longs.
|
| I don't like this, as it implies that 'i' is a 32 bit signed integer, but
| you don't have any control. It blows up here because it won't extend the
| type to a long - as it's typeless, I don't see why it should be able to do
| this automatically.
|
| >>> i = 0x7fffffff
| >>> i = i + 1
| Traceback (innermost last):
|   File "<pyshell#1>", line 1, in ?
|     i = i + 1
| OverflowError: integer addition
|
| Typeless languages are all very well, but there are times when I really
| would like to say 'this is a 32 bit unsigned integer - treat it as such'
| instead of implied typing. The warning bells sounded a bit on p33 of
| 'Learning Python' when it suggested that "as a rule of thumb, if you find
| yourself wanting to flip bits in Python, you should think long and hard
| about which language you're really using". But there are more compelling
| reasons why I am using Python than C for my particular application. I don't
| particularly want to have to start writing C companion libraries if I can
| avoid it.

I guess there's a tension here between these somewhat exotic
applications and more general aspirations for Python, where
even the present reliance on machine numerical types is arguably
inappropriate.  I personally feel that Python's position on the
Very High Level scale here is just typical of its practicality,
but with the unsigned integer, in C we have an example to be
avoided.  I mean, its rules for operations on unsigned integers
make some sense insofar as I can see, but the ANSI library
standards are clearly on a different page about it with things
like strlen() returning unsigned integers.  If the arbiters of
ANSI C can't even keep it straight, how can Python users be
expected to sort these things out?

Anyway, rant over, it's not entirely clear which way you want to
go, but I think you want 0x7fffffff to roll over to 0x80000000
regardless of hardware integer size, without the overflow exception.
 Would it be reasonable, considering that this is an exotic and
not generally useful behavior, to do that kind of arithmetic with
a functional notation like c = uadd(a, b, 32), c = umul(a, b, 32)?

Ironically our high level language looks a little like assembly
language here, but then neither one has any notion of type declaration
so perhaps it's inevitable.

	Donn Cave, University Computing Services, University of Washington
	donn at u.washington.edu


From mak at mikroplan.com.pl  Wed Apr 12 12:43:21 2000
From: mak at mikroplan.com.pl (Grzegorz Makarewicz)
Date: Wed, 12 Apr 2000 18:43:21 +0200
Subject: Python and Curses under MS Windows
In-Reply-To: <8d25p8$7u9$1@uni00nw.unity.ncsu.edu>
Message-ID: <NDBBIKNLJKPLOLAJJJAPMEFLDCAA.mak@mikroplan.com.pl>


> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of William Isaac Carroll
> Sent: Wednesday, April 12, 2000 5:47 PM
> To: python-list at python.org
> Subject: Python and Curses under MS Windows
> 
> 
> I'm interested in finding a curses package that works under Windows. I've
> searched the FAQ and archives with no results. Does such a thing 
> exist? And
> does it come in a pre-compiled form? (I have no decent Windows compiler
> available.)
> 

Curses module (compiled and source) for python15 you can find
on http://www.mikroplan.com.pl/~mak/python

Grzegorz Makarewicz




From phd at phd.russ.ru  Fri Apr 28 04:18:08 2000
From: phd at phd.russ.ru (Oleg Broytmann)
Date: Fri, 28 Apr 2000 08:18:08 +0000 (GMT)
Subject: HTML parser example, anybody?
In-Reply-To: <slrn8gh2a9.oj.haering@sunhalle8.informatik.tu-muenchen.de>
Message-ID: <Pine.LNX.4.21.0004280814400.18766-100000@fep132.fep.ru>

On 27 Apr 2000, Gerhard Haering wrote:
> I want to transform my own HTML templates (with a few template tags,
> special comments, and so on) into final HTML for the browser. I would like

   I'd recommend you to use DocumentTemplate (ZTemplates from www.zope.org)
instead.

> to use Pyhton's htmllib, but from the docs I can get no clue on how to get
> it done. I seem to have to use a writer and a formatter class ... 

   Do not use HTMLLib, it is designed for formatting text representation of
HTML. Use SGMLLib instead.

> I would greatly appreciate an example of HTML -> HTML using
> htmllib/sgmllib. Can anybody provide pointers/code snippets?

   Cannot show HTML->HTML, but can show HTML parsing with SGMLLib. Look
into my project "Bookmarks database and Internet robot". There is a parser
for bookmarks.html - one of the parsers I wrote so many times.

   Master site: http://sun.med.ru/~phd/Software/Python/#bookmarks_db

   Faster mirrors: http://skyscraper.fortunecity.com/unix/797/Software/Python/#bookmarks_db
   http://members.xoom.com/_XMCM/phd2.1/Software/Python/index.html#bookmarks_db

Oleg.            (All opinions are mine and not of my employer)
---- 
    Oleg Broytmann      Foundation for Effective Policies      phd at phd.russ.ru
           Programmers don't die, they just GOSUB without RETURN.




From ulf.engstrom at b2b-link.com  Fri Apr 28 07:54:39 2000
From: ulf.engstrom at b2b-link.com (=?iso-8859-1?Q?Ulf_Engstr=F8m?=)
Date: Fri, 28 Apr 2000 13:54:39 +0200
Subject: Reading user input
References: <39096d88.3112113532@news.ramk.fi>
Message-ID: <01d101bfb108$88862290$858647c1@DEVELOPMENT>

I guess what you're looking for is raw_input(). Works like this:
>>> text = raw_input('Enter text here->')
Enter text here->Anything I want
>>> text
'Anything I want'

Hope it helps :)
Regards
Ulf

> Hi.
> 
> I'm doing some studing about python and MySQL. I've done little test
> program where I use database from python. Now I need get user input to
> do some queries(and other stuff) to database. I've read Python manual
> but haven't found simple solution to do that. I can read something
> using curses but I don't like it. Is there other(and simpler) way to
> do that??
> 
> I've done only CGI programs with python and there haven't been any
> need to read chars from user input. Only stuff from stdin that httpd
> have put there.
> 
> At this case 'user input' is not arguments given when program starts
> but variables that are given when program is running. 
> 
> Of course I could do html-page and do my test program as CGI and use
> it that way but it's too much trouble when I'm only trying to learn
> databases and how to use them from python.
> 
> TK
> -- 
> http://www.python.org/mailman/listinfo/python-list




From fredrik at pythonware.com  Tue Apr  4 04:48:38 2000
From: fredrik at pythonware.com (fredrik at pythonware.com)
Date: Tue, 04 Apr 2000 08:48:38 GMT
Subject: Python 1.6 alpha 1 released
References: <Pine.GSO.4.10.10004040035420.21084-100000@sundial> <5E877FD7637AB46E.DABBED9CF3C1D64B.CDDDC05AE7374A78@lp.airnews.net>
Message-ID: <8cca93$9oh$1@nnrp1.deja.com>

Tres Seaver wrote:
> > it *breaks existing applications coded to the published interface*,

not true.  the documented interface explicitly says that
there is one single argument.  it does not mention optional
arguments.

(the example was broken, but so were other examples in earlier
releases of the documentation.  see other messages in this
thread for details)

> Consider that *every one* of the protocol modules which offers
> a connect() method uses a two-argument signature for it:  ftplib,
> gopherlib, httplib, imaplib, nntplib, poplib, smtplib, telnetlib,
> urllib.  And they all *use* the two-argument connect when invoking
> their underlying socket!

not true.  look again (doesn't matter if you look in the 1.5.2
library or the 1.6 library -- what you just said is not correct
in either case).

> No Python module in the standard library uses the single-argument
> connect, period.

*all* python modules in the 1.6 standard library use
single-argument socket connect.

> This is *not* the same case as the list.append() one, where the
> published interface was itself consistent.

why not?  after all, there were many more uses of multi-argument
append in the 1.5.2 standard library than uses of multi-argument
socket calls.

see my other post in this thread for the list of questions
I ask myself for each change in Python.  if you use other
criteria, which make it obvious to you that "append" is a
good change, and "connect" is a bad one, please tell me what
they are.

(bogus facts or double standards doesn't count).

after all, with tons of code all over the Python universe,
I'm affected by *all* backwards incompatible changes in
Python 1.6.  but for some reason, the "append" change is
the only one that makes me worried.  what am I missing?

</F>


Sent via Deja.com http://www.deja.com/
Before you buy.


From petersc at stud.ntnu.no  Thu Apr 27 09:57:54 2000
From: petersc at stud.ntnu.no (Peter Schneider-Kamp)
Date: Thu, 27 Apr 2000 15:57:54 +0200
Subject: Concatening string and integer
References: <8e9gco$jlt$1@vg170.it.volvo.se>
Message-ID: <39084762.C074814E@stud.ntnu.no>

JJ wrote:
> 
> Hi!
> 
> I want to concatenate a string containing both strings and integers like:
> 
> strTemp = 'I make test number: ' + iMyNumber + ' once again'

Adapted from the FAQ:

strTemp = 'I make test number: ' + repr(iMyNumber) + ' once again'

bye
Peter
--
Peter Schneider-Kamp          ++47-7388-7331
Herman Krags veg 51-11        mailto:peter at schneider-kamp.de
N-7050 Trondheim              http://schneider-kamp.de



From akuchlin at mems-exchange.org  Mon Apr 24 17:15:44 2000
From: akuchlin at mems-exchange.org (Andrew M. Kuchling)
Date: 24 Apr 2000 17:15:44 -0400
Subject: Gzip -> FTP w/o using a temporary file
References: <8e279b$2rf0$1@nnrp-corp.news.cais.net>
Message-ID: <3du2grxksf.fsf@amarok.cnri.reston.va.us>

"Joel Burton" <pup at pobox.com> writes:
> I tried using os.pipe() to create an r/w pipe, but the 'w' part of the pipe
> isn't writable, according to gzip.GzipFile()

os.pipe() returns file descriptors, which are small integers, while
GzipFile expects a Python file object.  Wrap the file descriptor in
a file object using os.fdopen(), something like
output = os.fdopen( w, 'wb')

-- 
A.M. Kuchling			http://starship.python.net/crew/amk/
Sending a newgroup message without permission of Leader Kibo: Poster is forced
to adopt twelve wacky sitcom children.
  -- Kibo, in the Happynet Manifesto


From godzilla at netmeg.net  Mon Apr 24 10:39:28 2000
From: godzilla at netmeg.net (Les Schaffer)
Date: Mon, 24 Apr 2000 14:39:28 GMT
Subject: ANN: NJ Piggies welcome Andy Robinson
References: <39044E57.58467ED@reportlab.com>
Message-ID: <m2r9bvv9oj.fsf@netmeg.net>

> Please join us as the New Jersey Python interest group (NJ Piggies)
> welcome Andy Robinson, Chief Architect and CEO of ReportLab,
> visiting from London.

what is planned for this occasion? lecture? topics? ...

les schaffer


From tseaver at starbase.neosoft.com  Wed Apr 12 22:49:53 2000
From: tseaver at starbase.neosoft.com (Tres Seaver)
Date: 12 Apr 2000 21:49:53 -0500
Subject: Why should I switch to Python?
References: <seas5ia89177@corp.supernews.com> <slrn8f4d0j.1t0.bwinton@tor.dhs.org> <8d2aa8$n58$1@newshost.accu.uu.nl> <8d2blf$b1n$1@pegasus.csx.cam.ac.uk>
Message-ID: <0F8FA87D4E1A9FC3.175617597B7C89BA.B3FB1819189281D7@lp.airnews.net>

In article <8d2blf$b1n$1 at pegasus.csx.cam.ac.uk>,
Nick Maclaren <nmm1 at cus.cam.ac.uk> wrote:
>
>In article <8d2aa8$n58$1 at newshost.accu.uu.nl>, m.faassen at vet.uu.nl
>  (Martijn Faassen) writes:
>|> Blake Winton <bwinton at tor.dhs.org> wrote:
>|> > On Fri, 31 Mar 2000 19:43:58 -0800, Aaron Turner wrote:
>|> >>I've been coding Perl for over 3 years and really like it.  (Especially
>|> >>it's syntax and ability to grok my programming style.)
>|> 
>|> > Now, just as a side note, why is it that the sets of The Perl Hackers I
>|> > Know, and The People I Know Who Use The Word "Grok" are the same set?
>|> 
>|> It's in the hacker's dictionary:
>|> 
>|> http://www.tuxedo.org/~esr/jargon/html/entry/grok.html
>
>According to something I read, incorrectly attributed!  Apparently
>Heinlein got it from something he read, and forgot the fact - it
>was then pointed out to him and he said something equivalent to
>"Oops".  That doesn't stop the attribution to him circulating :-)

I don't buy "got it from something he read" unless the Marsport Times
was being delivered to Colorado Springs c. 1960 using Guido's TimeMachine.
You need *better* attribution to contest priority of existing, valid
attribution.

See RAH's _Grumbles from the Grave_ for more history on SIASL.

>The same thing has happened with the expression "To damn with
>faint praise", incidentally - it is attributed to Pope, but is
>actually Wycherley.

Hmm, maybe you'll argue that LeGuin's _Left Hand of Darkness_ should be
attributed there, too :)

Tres.
-- 
---------------------------------------------------------------
Tres Seaver        tseaver at digicool.com     http://www.zope.org
Digital Creations


From cut_me_out at hotmail.com  Sat Apr 29 13:24:15 2000
From: cut_me_out at hotmail.com (Alex)
Date: 29 Apr 2000 13:24:15 -0400
Subject: problem in python function --Name-error
References: <390B198D.7AC26774@swt.edu>
Message-ID: <etdvh10yg5c.fsf@w20-575-109.mit.edu>

> The error says nameError frame1.

Perhaps you want to call this self.frame1 in the __init__ method.  Then
you will still be able to refer to it in the drawCanvas method by the
same name.  At the moment, you are getting an error because there is no
variable called frame1 in the namespace of the drawCanvas method.

Alex.


From mhammond at skippinet.com.au  Sat Apr  8 22:04:33 2000
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Sun, 09 Apr 2000 02:04:33 GMT
Subject: Python on Google?
References: <8F0ED22EBjoshtinamecom@209.208.0.7> <8col53$m3o$1@nntp.Stanford.EDU>
Message-ID: <RyRH4.1669$4v.9486@news-server.bigpond.net.au>

"Amit Patel" <amitp at Xenon.Stanford.EDU> wrote in message
news:8col53$m3o$1 at nntp.Stanford.EDU...
> Josh Tompkins <josht at iname.com> wrote:

> Python has been used at Google since the very beginning, and it
> continues to be used, although some of the more performance critical
> stuff is in C++.  New code _is_ being written in Python (in fact, I
do
> all my work in Python) and it's being used exclusively to deal with
> some large sets of data.  Email me if you want more details.

It would be very cool if you could knock up a few paras of information
about this that we could put on the Python web site in the relevant
places.  Advocacy from high-profile Python users is always good.

Mark.





From effbot at telia.com  Mon Apr  3 12:11:39 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Mon, 03 Apr 2000 16:11:39 GMT
Subject: Unicode program representation
References: <015c01bf9d0b$76f9b2a0$59dd3fcb@neil> <3xXF4.1060$n68.186533888@newsb.telia.net> <001201bf9d62$3e1ac1c0$36dd3fcb@neil>
Message-ID: <%o3G4.1111$n68.191610880@newsb.telia.net>

Neil Hodgson <neilh at hare.net.au> wrote:
> > however, perfectly valid ISO Latin 1 strings (or any other standard
> > 8-bit encoding) are invalid UTF-8 strings.   changing the default to
> > UTF-8 this is likely to break many existing scripts.
>
>    But there are relatively few scripts currently using the u" form since
it
> has yet to appear in a release version of Python.

hmm.  wouldn't that mean that we end up using different encodings
in different parts of the script?  feels a little scary, to say the least...

</F>




From arnold at dstc.edu.au  Sun Apr 16 20:54:29 2000
From: arnold at dstc.edu.au (David Arnold)
Date: Mon, 17 Apr 2000 10:54:29 +1000
Subject: module naming? 
In-Reply-To: Your message of "Sun, 16 Apr 2000 20:39:00 -0400."
             <Pine.BSF.4.10.10004162018190.25069-100000@chi.pair.com> 
Message-ID: <200004170054.KAA14097@piglet.dstc.edu.au>

-->"Michal" == sabren  <Michal> writes:

  Michal> Is there a better way to do this?

what about something like

   from Record import Record
   from Field import Field
   from IdxDict import IdxDict

in your zdc/__init__.py ?



d



From alex_thomas at cableinet.spam.co.uk  Fri Apr 14 16:46:13 2000
From: alex_thomas at cableinet.spam.co.uk (Alex Thomas)
Date: Fri, 14 Apr 2000 20:46:13 GMT
Subject: Scripting and Gnome and KDE
Message-ID: <7t0ffs8ncrrh3irgc52rbsb68ue48rfovf@4ax.com>

This might seem an odd way to choose a GUI, but if you had the choice
between scripting Gnome/Gtk or KDE/Qt-based components using Python, which
would you go for? 

Thanks peeps
Alex


From echuck at mindspring.com  Sun Apr 16 21:00:18 2000
From: echuck at mindspring.com (Chuck Esterbrook)
Date: Sun, 16 Apr 2000 21:00:18 -0400
Subject: else clauses in while and for loops
References: <200004170028.KAA07736@nexus.csse.monash.edu.au>
Message-ID: <38FA6222.E8BFE8B5@mindspring.com>

Jonathan Giddy wrote:

> Chuck Esterbrook declared:
> >
> >On another topic: With regards to the "try...except" technique above
> >(which I've seen quite a bit in Python) does the following code do the
> >same thing and if so, which do you find more readable?
>
> It does do the same thing with the extra line included below.  Now that
> I've looked at this code in more detail than usual, I would probably choose
> your form over my original, mainly on the grounds that it's quite likely
> for the test to fail, so it's not really "exceptional" behaviour.
>
> However, for code that is executed once over a limited number of
> iterations, I'm not losing sleep either way.

Agreed. I just wanted to know about the two styles in general.

Thanks,

-Chuck

From a.genkin at utoronto.ca  Sat Apr 29 16:17:40 2000
From: a.genkin at utoronto.ca (Arcady Genkin)
Date: Sat, 29 Apr 2000 20:17:40 GMT
Subject: Problem running NNTP example from docs.
Message-ID: <87hfckk6fv.fsf@tea.thpoon.com>

I'm very new to Python, and was trying to run the example from
documentation to nntplib library. I get an error when I try to
instantiate NNTP object.

,----
| >>> import nntplib
| >>> s = NNTP('news')
| Traceback (innermost last):
|   File "<stdin>", line 1, in ?
| NameError: NNTP
`----

Could somebody point out what I am missing?

Thanks!
-- 
Arcady Genkin                                 http://www.thpoon.com
Nostalgia isn't what it used to be.


From smigowsky at dspace.de  Thu Apr 13 06:09:52 2000
From: smigowsky at dspace.de (Stefan Migowsky)
Date: Thu, 13 Apr 2000 11:09:52 +0100
Subject: Problems with PythonWin
Message-ID: <326536345498D311B3BC00105A39802A074686@newsintern.dspace.de>

>
>pywintypes.com_error: (-2147467262, 'No such interface 
>supported', None, None)
>

This error occurs if you have a special version of the comcat.dll
which resides in the system32 directory. Microsoft has updated 
this dll without supporting older interfaces. But now they solved 
the error.

To register the ActiveX Script Engine you have to have either version
4.71 or the new Version 5.0 from the Internet Explorer 5.0.

The faulty version is 5.0 (!) from Internet Explorer 4.01. 

Sometime I don't know what they are doing in Redmond.

Bye 

     Stefan



From lang at zid.tu-graz.ac.at  Thu Apr  6 07:46:08 2000
From: lang at zid.tu-graz.ac.at (Mario Lang)
Date: 06 Apr 2000 13:46:08 +0200
Subject: Python and Audio-input/output/processing/mixing
Message-ID: <8766tvqwov.fsf@fzidpc73.tu-graz.ac.at>

Hello.

I am quite new to python but I am very impressed by the
way it is designed.
I am planning to write a music-studio in python based
on a completely object oriented design.

As I searched for libs, I couldnt really find something useful to base
my work on.
Are there any major modules for python and audio processing?

-- 
Regards,
      Mario Lang

Technical University Graz           E-Mail: lang at zid.tu-graz.ac.at
Department Computing Services          ICQ: 69372257
Phone: +43-316-873-8508
  Fax: +43-316-873-7699



From nobody at nowhere.nohow  Sun Apr 23 22:58:57 2000
From: nobody at nowhere.nohow (Grant Edwards)
Date: Mon, 24 Apr 2000 02:58:57 GMT
Subject: Why doesn't O'Reilly use parrots on its python books?
References: <etdvh186ztw.fsf@w20-575-115.mit.edu>
Message-ID: <RLOM4.1158$wJ1.23728@ptah.visi.com>

On 23 Apr 2000 21:40:43 -0400, Alex <cut_me_out at hotmail.com> wrote:

>Given their venerable standing in Python lore, it surprises me that none
>of O'reilly's python books have a parrot on the cover.  There must be
>lots of types of parrots to go around, even if some O'reilly books
>already have them.

IIRC, the Python parrot would need to be a Norwegian Blue.
Lovely plumage...

>And the python book could be distinguished by the fact that its
>parrot is lying on its back with its feet curled up...
>
>A really agressive looking rabbit would work, too.

With sharp, pointy teeth?

-- 
Grant Edwards                   grante             Yow!  .. If I cover this
                                  at               entire WALL with MAZOLA,
                               visi.com            wdo I have to give my AGENT
                                                   ten per cent??


From Gareth.McCaughan at pobox.com  Wed Apr  5 15:05:44 2000
From: Gareth.McCaughan at pobox.com (Gareth McCaughan)
Date: 05 Apr 2000 20:05:44 +0100
Subject: Why should I switch to Python?
References: <000701bf9c5a$b806b420$752d153f@tim>
	<oqity066km.fsf@titan.progiciels-bpi.ca>
	<slrn8ej2gc.b3t.warlock@raven.myth>
	<m3itxyz744.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <864s9g74hj.fsf@g.local>

Michael Hudson wrote:

> It's also possible to show that if a and n have no common factor, then
> there are an infinte number of primes in the artihmetic progression
> a+dn as d=0,1,2,3,4,..., but the proof is, umm, somewhat harder (ie. I
> don't know it; I think it involves slinging Riemann zeta functions).

Dirichlet L-series, actually. Closely related to the Riemann
zeta function, but not quite the same thing. You look at
sum{1..oo} of a(k)/k^s where a is a "character mod n" (which
just means that a(k) depends only on k mod n and a(pq) = a(p)a(q)),
and the crux of the proof is showing that this thing is well
behaved at s=1 if a isn't the obvious thing that's always
equal to 1. Once you've got that the rest of the proof isn't
much harder than the following witty proof that there are
infinitely many primes: "sum{1..oo} of n^-s = product{primes}
of 1/(1-p^-s). As s->1, the former goes to infinity, so the
latter does too, therefore there must be infinitely many
factors in the product. QED."

Therefore you should switch to Python. QED. :-)

-- 
Gareth McCaughan  Gareth.McCaughan at pobox.com
sig under construction


From gherman at darwin.in-berlin.de  Tue Apr 11 04:06:11 2000
From: gherman at darwin.in-berlin.de (Dinu C. Gherman)
Date: Tue, 11 Apr 2000 10:06:11 +0200
Subject: Starship Python is down!?
Message-ID: <38F2DCF3.BA257022@darwin.in-berlin.de>

I can't get anything from http://starship.python.net or
http://starship.python.net:8080 for more than a day, now.
Is that some known problem?

Regards,

Dinu

-- 
Dinu C. Gherman
................................................................
"The thing about Linux or open software in general is that 
it actually tries to move software from being witchcraft to 
being a science," [...] "A lot of the programs you see today 
are actually put together by shamans, and you just take it and 
if the computer crashes you walk around it three times... and 
maybe it's OK." (Linus Thorvalds, LinuxWorld 2000, NYC)



From ivanlan at callware.com  Tue Apr  4 22:08:02 2000
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Tue, 04 Apr 2000 20:08:02 -0600
Subject: Two questions
References: <1257223201-7158461@hypernet.com>
Message-ID: <38EAA002.D82AFAB8@callware.com>

Hi All--

Gordon McMillan wrote:
> 
[snippery]
> It's not widely known, but it blows chunks if you have Mayan
> glyphs on your machine. That's a feature.
> 

Ta' e' ku' yu'uch'ul.  (That's Yucatec Mayan, for those of you who need
a hint.)

> If you get creative (about a dozen lines of code) you can install
> a fully functional Python in 2 files. You can put Python and
> IDLE on a floppy with room left over.
> 

Yes, but will it fit under the seat?


> At the moment, you'll have a bit of trouble with my stuff and
> the fact that Marc-Andre puts his .pyds inside the packages.
> There are workarounds, and I have a patch from Thomas Heller
> which I haven't finished reconciling to some of my other
> changes.
> 

How nasty are the workarounds?  Do they make life annoying for the guy
who writes the installs?  (Acceptable--possibly even desirable.) 
Annoying for me?  (Irritating but acceptable.)  Or infuriating for our
customers?  (Unacceptable.)

Of course, if there is a way to make life unbearable for the salesmen
(and they are all men), that would be ideal.


> > <floccinausinihilipilification>-ly y'rs,
> > Ivan
> 
> Are you describing your hairline, your nervous twitch, or the
> effect of the latter on your former?
> 

C'mon, Gordon.  Get out your OED!  I _know_ you've got one.

<it's-a-magic-word-that-makes-long-haired-pipe-smoking-leftover-hippies-lose-their-cachet>-ly
y'rs,
Ivan;-)
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
http://www.pauahtun.org and
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
Author:  Teach Yourself Python in 24 Hours



From akuchlin at mems-exchange.org  Fri Apr 21 18:04:03 2000
From: akuchlin at mems-exchange.org (Andrew M. Kuchling)
Date: 21 Apr 2000 18:04:03 -0400
Subject: The Simple Economics of Open Source
References: <1255771356-36910442@hypernet.com>
Message-ID: <3dya67xga4.fsf@amarok.cnri.reston.va.us>

> Robin Becker wrote:
> > I prefer not to worry about the exact electronics/chemistry behind my
> > computer. So in that sense my 'open sourceness' has limits. I guess this

That may be only partly true.  Python is the only free software I've
done serious hacking on.  But I've contributed minor patches to a
number of projects -- mod_java, PCRE, SANE.  They might be only fixes
for NULL pointer derefs or documentation typos, but those other
projects have gained a teensy bit more polish from the fact that I
used them.  Of such incremental improvements is great stability made.

Similarly, people who write software on top of Python, or Zope, or
Linux, can trust that *someone* will take care of the maintenance of
that substrate for them, and, in the unlikely event all the
maintainers go away, you could still do it yourself if needed.

"Gordon McMillan" <gmcm at hypernet.com> writes:
> internet technologies)? Remember when IBM truly believed the 
> world would communicate through SNA? Remember when 
> damn near any real app cost at least $10K?

Also, ever had to debug a problem in a binary-only library through
trial-and-error?  You can burn lots of time that way.  <core dump>
"Hmm... maybe that parameter can't be NULL."  <core dump> "Maybe that
struct field should be 1." <core dump>.  I like the fact that, if I
start seeing mysterious errors, I can dive down into the language
interpreter, farther down into the C library, and even into the kernel
if needed.  Though I've only followed a problem all the way into the
Linux kernel once (to discover it was my error), it was reassuring to
be able to do so.

-- 
A.M. Kuchling			http://starship.python.net/crew/amk/
He spoke to the scar-dancers, to the straw-dust-women, to the old man with a
swan's arm who tends the back stairs, to the three children of the autopsy, to
the painters and the scriveners and the walls.
  -- From a week in Dream's life, in SANDMAN #64: "The Kindly Ones:8"



From mark at chem.uwa.edu.au  Sat Apr 22 20:29:09 2000
From: mark at chem.uwa.edu.au (Mark C Favas)
Date: 23 Apr 00 00:29:09 GMT
Subject: executable "wrapper" gui -- advice sought
References: <xhnM4.1153$PP5.17158@news.swbell.net>
Message-ID: <mark.956449749@declan>

"Grant Goodyear" <g2boojum at hotmail.com> writes:

>I'm in the process of writing an exceedingly simple gui that will take an
>already existing binary executable and input text file (I don't want to
>break the
>working system!) and allow the user to modify the input arguments,
>if desired, and then start the executable.  That part is straightforward.
>What I'm not sure how to do is to capture the executable's std out as
>it's generated.  My goal is to echo that output to a seperate window, and
>also to use it to generate a periodically updated html "status report" so
>that the program's progress can be monitored remotely.  Working
>through the gui aspects isn't all that difficult, but I have _no_ idea how
>to grab the executable's std out _while it is running_ and then have the
>python gui do something (anything!) with that data.  Any advice?

Have a look at the on-line module documentation for "commands" (Unix only),
and "popen2" (also Unix, though Mark Hammond's win32 package contains a working
pipe implementation amongst other goodies). If your executable writes to
/dev/tty instead of stdout you'll need to play with the pseudo tty module,
"pty" (doco says available for IRIX & Linux, but it's pretty easy to get it
working for Solaris and Tru64 Unix (nee Digital Unix)). Watch out for buffering
of the output, also.

--
Email  - mark at chem.uwa.edu.au      ,-_|\                           Mark C Favas
Phone  - +61 9 380 3482           /     \               Department of Chemistry
Fax    - +61 9 380 1005      ---> *_,-._/   The University of Western Australia
                                       v                               Nedlands
Loc    - 31.97 S, 115.81 E                               Western Australia 6009


From adjih at crepuscule.com  Fri Apr 28 10:12:01 2000
From: adjih at crepuscule.com (Cedric Adjih)
Date: 28 Apr 2000 14:12:01 GMT
Subject: Do I always have to write "self." ?
References: <270420001706386648%pecora@anvil.nrl.navy.mil> <slrn8ghc6t.f01.scarblac-spamtrap@flits104-37.flits.rug.nl> <280420000746597329%pecora@anvil.nrl.navy.mil>
Message-ID: <8ec67h$b14$1@ites.inria.fr>

Louis M. Pecora <pecora at anvil.nrl.navy.mil> wrote:
> In article <slrn8ghc6t.f01.scarblac-spamtrap at flits104-37.flits.rug.nl>,
> Remco Gerlich <scarblac-spamtrap at pino.selwerd.nl> wrote:

>> No. However, this seems to make code *more* readable. In fact, many C++
>> programmers use 'this.x' for all their instance variables, to be able to
>> see at a glance which are just local variables of the function and which are
>> instance variables. It's just more work to write, but that doesn't matter
>> much.

> I understand the potential for readability, but I do a lot of
> mathematical programming so the expressions can get pretty cluttered
> with variables and operators.  More self's only makes it worse for me. 
> Maybe that's my problem, but then that's my complaint.

Alternatively you could use 1-liners, if it is justified

class Polynom2:
	def __init__(self,a,b,c):
		self.a,self.b,self.c=a,b,c
	def delta(self):
		return self.b*self.b-4*self.a*self.c

Often, 1-liners will remove the need for one comment.

>> Python needs it to know which vars are instance variables, and which are
>> local variables.

> I'm sure some way of declaring variables 'local' could be done in the
> class defintion, but I doubt Guido would do that now.

>> Funny how people say that every aspect of Python they don't like make the
>> language less "clean". It is extremely consistent though.

> Funny, I guess.  Harder to read for me.

>> You *can* use another word for 'self' though, 's' or 'me' or something. But
>> it makes your code harder to read, of course...

> You mean by doing something like

> s=self
> at the beginning of each class method?

The name 'self' is just a convention.
But since some parsing software will rely on it, "s=self"
might be wiser.

class Polynom2:
	def delta(s): return s.b*s.b-4*s.a*s.c

or

class Polynom2:
	def delta(self):
		s=self
		return s*b*s.b-4*s.a*s.c

or the other suggestions for read-only attributes:


class Polynom2:
        def delta(self):
                a,b,c=self.a,self.b,self.c
                return b*b-4*a*c


-- Cedric



From wayne.izatt at myself.com  Wed Apr 26 23:34:25 2000
From: wayne.izatt at myself.com (Wayne Izatt)
Date: Thu, 27 Apr 2000 03:34:25 GMT
Subject: Python 2.0
Message-ID: <3907B52D.B01D5748@myself.com>

Hi all,

I'm evaluating the viability of developing a number of largish
applications over the next few months. However, I'm concerned about them
not being conformant with Python 2.0. I know that the interpreter could
simply be left at the 1.5.2 or 1.6 level, but this will probably not
satisfy my customer. I could go with that other scripting language, but
would rather not.

I guess what I'm wanting to know is how *much* the core language will
change. I've seen the overview slides describing the differences, but
they are very high-level. Given that my apps will span a number of areas
- text processing, networking, maybe even win32 (yuck) - I don't weant
to find myself with a huge maintenance task when 2.0 does come out.
(Although I admit anything written in that other language becomes a huge
maintenance task after a couple of weeks ;->

So, can anyone give me any advice? Is there any cause for alarm?

ta




From tismer at tismer.com  Tue Apr 25 12:38:38 2000
From: tismer at tismer.com (Christian Tismer)
Date: Tue, 25 Apr 2000 18:38:38 +0200
Subject: Princeton Instruments Python interface
References: <8e2i25$gou$1@nnrp1.deja.com>
Message-ID: <3905CA0E.E1BDF217@tismer.com>


Sam Schulenburg wrote:
> 
> I am developing a python library to access Princeton Instruments
> Spectroscopic WinX/32 COM objects. If anyone else has had done this I
> would like to get in touch with you to share ideas.
> 
> Sam Schulenburg

Do you know what PIPI means in Germany? :-))

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaunstr. 26                  :    *Starship* http://starship.python.net
14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     where do you want to jump today?   http://www.stackless.com



From samschul at pacbell.net  Mon Apr 24 21:30:19 2000
From: samschul at pacbell.net (Sam Schulenburg)
Date: Tue, 25 Apr 2000 01:30:19 GMT
Subject: Pythonwin com question
References: <8dqver$l4v$1@nnrp1.deja.com>
Message-ID: <8e2sfb$r6e$1@nnrp1.deja.com>

In Visual Basic I have the following:
'===============================================
Dim theFrame As Varient
Dim dx as DocFile
Dim ex as New ExpSetup
Dim bRes as Integer

if (ex.Start(dx)) then
   While ex.GetParam(EXP_RUNNING,bRes)
       DoEvents 'Wait for detector to finish
   Wend
   dx.GetFrame 1, theFrame 'Retreive one image
end if
'===============================================

In Python I have the equivalent:

from PrincetonInst import *  # Get the renamed Makepy file for the
                             # COM object

theFrame = []                # Set up the Varient
dx = DocFile()               # Get the DocFile class
ex = win32com.client.Dispatch(ExpSetup.CLSID)
bRes = 0

if (ex.Start(dx)) then       # Start detector scanning  for data
   While ex.GetParam(EXP_RUNNING,bRes):  # Detector is scanning
      ## DoEvents  # Wait for detector to finish, This has no
                   # equivalent in Pythonwin
         s = 1     # This gives the while loop something to do
   #### This is where everything falls apart
   #### theFrame is not the correct datatype
   dx.GetFrame(1,theFrame)  # Retreive one image

How do I define the correct Varient type in Python?

Sam Schulenburg





Sent via Deja.com http://www.deja.com/
Before you buy.


From ivanlan at callware.com  Wed Apr 19 23:18:11 2000
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Wed, 19 Apr 2000 21:18:11 -0600
Subject: Browse dialog in PythonWin
References: <38FE32AB.520C1736@callware.com> <38fe3297.36998571@news-server.socal.rr.com> <38FE3998.1ED7FB24@callware.com> <lXtL4.32$v85.230@news-server.bigpond.net.au>
Message-ID: <38FE76F3.4E352DA6@callware.com>

Hi All--

Mark Hammond wrote:
> 

[bobbitry]

> An alternative to get the real Windows browser dialog:
> 
> >>> import win32com.shell.shell
> >>> win32com.shell.shell.SHBrowseForFolder(0, None, "Select it", 0, None,
> 0)
> ('\024\000\037P\340O\320 \352:i\020\242\330\010\000+00\235\024\000.w
> \354!\352:i\020\242\335\010\000+00\235\000\000', 'Control Panel', 21)
> >>>
> 
> I selected "Control Panel" - You will need to read the docs to see how to
> use that funky string (the PIDL)...
> 

Oh, Mark, this makes my eyes cross and my brain hurt.  And I'm bound to
try it, too.  I guess I really am a masochist.

> No news on starship yet - if its not available for me to re-upload in 24
> hours I will do it somewhere else...
> 

Not to put too fine a point on it, but good.

<just-don't-do-it-in-the-road>-ly y'rs,
Ivan;-)
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
http://www.pauahtun.org and
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
Author:  Teach Yourself Python in 24 Hours



From parkw at better.net  Fri Apr 14 15:49:49 2000
From: parkw at better.net (William Park)
Date: Fri, 14 Apr 2000 15:49:49 -0400
Subject: Attributes (+Numerical Python)
In-Reply-To: <meh9-986AAB.15155814042000@news.cit.cornell.edu>; from meh9@cornell.edu on Fri, Apr 14, 2000 at 03:15:58PM -0400
References: <meh9-986AAB.15155814042000@news.cit.cornell.edu>
Message-ID: <20000414154949.A2643@better.net>

On Fri, Apr 14, 2000 at 03:15:58PM -0400, Matthew Hirsch wrote:
> Hi All,
> 
> In Numerical Python, you can say something like vector.imag and it will 
> return arrays of the imaginary parts of the elements in the array, 
> vector.  But this attribute only exists when the array is complex.  I'd 
> like to do something like the following, but I have no idea what the 
> syntax should be:
> 
> if the .imag attribute exists:
>    do something
> else
>    do something else
> 
> How should this be written?
> 
> Thanks for your help,
> Matt

How about
    try:
	<do something with 'vector.imag'>
    except:	    # or except AttributeError:
	<do something else>

--William



From arta at x-stream.nl  Fri Apr 14 10:05:53 2000
From: arta at x-stream.nl (A[r]TA)
Date: Fri, 14 Apr 2000 16:05:53 +0200
Subject: raw_input in loop
Message-ID: <e$crmckp$GA.196@net025s>

Heyz, I want to write a loop
in which there's something like raw_input but which
doesn't stop the loop until the enterbutton is hitten.
How do I do that???

A[r]TA

--
We shall not cease from exploration, and the end of all our exploring will
be to arrive where we started and know the place for the first time.
- T. S. Eliot





From helge.hess at mdlink.de  Tue Apr 11 08:25:08 2000
From: helge.hess at mdlink.de (Helge Hess)
Date: Tue, 11 Apr 2000 14:25:08 +0200
Subject: [ANNOUNCE] Garbage collection for Python
References: <20000407221750.A29279@acs.ucalgary.ca> <38F13FB9.D1D16AF@mindspring.com> <slrn8f2if1.dop.nascheme@cranky.arctrix.com> <87r9cdhdb0.fsf@deneb.cygnus.argh.org> <FsunGI.BtB@world.std.com>
Message-ID: <38F319A4.89621192@mdlink.de>

Will Ware wrote:
> Florian Weimer (fw at deneb.cygnus.argh.org) wrote:
> > Is it possible to extend your GC in a way which would permit the
> > execution of several bytecode interpreters in multiple threads?
> 
> It sounds like you might want to simply run several copies of the
> Python interpreter in separate processes, which isn't a GC issue.

I think what he wants to do (me too ;-) is to run several copies of the
interpreter in *one* process, possibly in multiple threads.
But I'm not sure whether this is a GC issue though (RC isn't thread-safe
anyway, right ?).

Helge
-- 
SKYRIX-OS  Web Operating System       - http://www.skyrix.com
SKYRIX-IAS IntraNet Application Suite - http://www.skyrix.com


From parkw at freewwweb.com  Wed Apr 26 22:25:43 2000
From: parkw at freewwweb.com (William Park)
Date: Wed, 26 Apr 2000 22:25:43 -0400
Subject: condense whitespace to single space?
In-Reply-To: <36NN4.31937$Xk2.119087@tor-nn1.netcom.ca>; from none@none.ca on Wed, Apr 26, 2000 at 10:57:15PM -0300
References: <36NN4.31937$Xk2.119087@tor-nn1.netcom.ca>
Message-ID: <20000426222543.A1377@freewwweb.com>

On Wed, Apr 26, 2000 at 10:57:15PM -0300, k wrote:
> How do I condense the sections of spaces in this string:
> 
> "abc              def         ghi"
> 
> into single spaces?
> 
> "abc def ghi"

1.  string.join(string.split("abc              def         ghi"))

2.  re.sub(' +', ' ', "abc              def         ghi")

--William



From tim.hochberg at ieee.org  Sat Apr 29 13:20:18 2000
From: tim.hochberg at ieee.org (Tim Hochberg)
Date: Sat, 29 Apr 2000 17:20:18 GMT
Subject: Where can I find NumPy
References: <8ef4pn$jge$1@kopp.stud.ntnu.no>
Message-ID: <mREO4.86254$U4.764472@news1.rdc1.az.home.com>



"Jan Erik Breimo" <janeb at stud.ntnu.no> wrote in message
news:8ef4pn$jge$1 at kopp.stud.ntnu.no...
> The official site (ftp-icf.llnl.gov) hasn't been responding for more than
> a week, so are there any other sites where I can find the source
> files for NumPy?

Yeah, go to sourceforge (www.sourceforge.com) and do a search for numpy --
you should be able to find it.

-tim


>
> --
> Jan Erik Breimo
> email: janeb at stud.ntnu.no




From python-list at teleo.net  Fri Apr 21 10:46:02 2000
From: python-list at teleo.net (Patrick Phalen)
Date: Fri, 21 Apr 2000 07:46:02 -0700
Subject: PyApache memory leak?
In-Reply-To: <Pine.LNX.4.21.0004210743260.2712-100000@fep132.fep.ru>
References: <Pine.LNX.4.21.0004210743260.2712-100000@fep132.fep.ru>
Message-ID: <00042107521900.02973@quadra.teleo.net>

[Oleg Broytmann, on Fri, 21 Apr 2000]
:: I don't see how httpdapy could help spawn less interpreters instances -
:: it is Apache job to fork and reap off children.

A clarification, from the Httpdapy README:

"""
2. PyApache creates a sub-interpreter (Py_NewInterpreter()) for every
request and destroys it when done. This means that if your script
begins with "import HTMLgen", HTMLgen is imported (bytecode read from
file) for every hit.

Httpdapy keeps the interpreter around from the first hit and until the
process dies. So only the first hit will actually read HTMLgen.py(c),
all the subsequent won't.
"""



From hamish_lawson at yahoo.co.uk  Mon Apr  3 18:29:23 2000
From: hamish_lawson at yahoo.co.uk (Hamish Lawson)
Date: Mon, 03 Apr 2000 15:29:23 -0700
Subject: For 1.6 make UserDict use __setitem__ and __getitem__ in other methods?
Message-ID: <07168ec0.10e4da64@usw-ex0101-006.remarq.com>

Consider the code below which defines a dictionary class whose
keys are case-insensitive.


    import UserDict, string

    class CaselessDict(UserDict):

        def __setitem__(self, key, value):
            self.data[string.lower(key)] = value

        def __getitem__(self, key):
            return self.data[string.lower(key)]

    d = CaselessDict()
    d['Joe'] = "Rhubarb"
    print d['JOE']                  # prints "Rhubarb"
    print d.has_key('JOE')          # prints 0 i.e.
    print d.get('JOE', None)        # prints None


An element set as 'Joe' can be fetched as 'JOE'. However as it
stands the semantics of has_key('JOE') and get('JOE') are
broken, since those methods don't think that a 'JOE' element
exists. This arises because in the parent UserDict class, these
methods access the self.data attribute.

Of course the author of CaselessDict could provide definitions
for has_key() and get(), but somehow this seems a bit inelegant
to me; I feel that the behaviour of has_key() and get() (and
maybe other methods) should follow on from whatever behaviour
has been defined for __getitem__ in the derived class. I propose
that for Python 1.6, UserDict's methods are rewritten to do
this. Is there any reason why this would be a bad idea? Below
I've provided rewritten versions of has_key() and get() to get
started.

    def has_key(self, key):
        try:
            self[key]
            return 1
        except KeyError:
            return 0

    def get(self, key, failobj=None):
        try:
            return self[key]
        except KeyError:
            return failobj


Hamish Lawson


* Sent from RemarQ http://www.remarq.com The Internet's Discussion Network *
The fastest and easiest way to search and participate in Usenet - Free!



From gmcm at hypernet.com  Fri Apr 14 08:18:15 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Fri, 14 Apr 2000 08:18:15 -0400
Subject: what is pythonwin?
In-Reply-To: <38F6FDD8.DA2CC785@gmx.net>
Message-ID: <1256407800-56204710@hypernet.com>

Robert Kiendl wrote:

> i ve got embedded the normal pyhton (python15.dll) into a windows
> mfc-app.
> 
> now i want to use some of the win32ui stuff, but cannot load that
> mysterious *.pyd files, whereas normal python.exe from the winall
> distribution does it. are these dll's? 

Yup.

> is pythonwin not based on the
> normal python? 

Normal python.

> do i have to handle the pythonwin source-distribution?
> link another dll?
> i do not get the right track from the standard docs.

Are you by chance using a debug build? Then all the .pyd's 
need to built for debug too. Otherwise it's probably something 
silly, like your PYTHONPATH.


- Gordon



From robin at jessikat.demon.co.uk  Fri Apr  7 18:45:45 2000
From: robin at jessikat.demon.co.uk (Robin Becker)
Date: Fri, 7 Apr 2000 23:45:45 +0100
Subject: anyone else wrap gd-1.8?
References: <65118AEEFF5AD3118E8300508B1248774CB282@ALTNET>
Message-ID: <BCgBXUAZUm74Ewhm@jessikat.demon.co.uk>

In article <65118AEEFF5AD3118E8300508B1248774CB282 at ALTNET>, Mike Steed
<MSteed at altiris.com> writes
>On a semi-related note, I have wrapped (without SWIG) gdchart, a library for
>making charts and graphs, which uses gd.  I have a new version (to be posted
>Real Soon Now) that uses gd-1.8, with support via the Python interface for
>PNG and JPEG output.  There is an entry in the Vaults of Parnassus.
>
>Mike
>msteed at altiris.com
...
I have wrapped using the old gdmodule gd-1.8.1. Unfortunately it's a
hacked version of gd-1.8.1. I sent a patch to T. Boutell maybe it'll get
in sometime.
-- 
Robin Becker


From effbot at telia.com  Thu Apr 20 04:59:40 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Thu, 20 Apr 2000 08:59:40 GMT
Subject: Pythonware Website? (Update, Apr 20)
References: <20000419154202.B480@owl.rhein-zeitung.de> <JCjL4.540$wYl.187815424@newsb.telia.net>
Message-ID: <0GzL4.3191$Za1.47063@newsc.telia.net>

quoting the daily python URL (http://hem.passagen.se/eff/url.htm):

    our ISP's file servers are down. this affects the web site as
    well as mail traffic to (but not from) the pythonware.com and
    secretlabs.com domains.

    our ISP is working on it, but the best they can say at this
    moment is "we'll get back to you when we know more"...

if the problem persists, we'll mirror the most important stuff on other
sites.  see the daily URL page for pointers and more status updates.

and as before, urgent mail can be sent to effbot at telia.com

</F>




From greg at cosc.canterbury.ac.nz  Thu Apr 20 01:28:53 2000
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Thu, 20 Apr 2000 17:28:53 +1200
Subject: Pliant?
References: <8di0lk$4h0$1@coco.singnet.com.sg>
Message-ID: <38FE9595.7E80B5D9@cosc.canterbury.ac.nz>

Ng Pheng Siong wrote:
> 
> What are your views on Pliant, as a Python programmer?
> 
> Pliant is here: http://pliant.cams.ehess.fr

I've had a brief look. Some observations, in no
particular order.

Personal opinion codes:
   :-) I like it
   :-( I don't like it
   :-! Not sure whether to like it or not
   :-] I think that's rather funny

* Indentation for structure: I approve, obviously! :-)
  Spaces are mandated, tabs forbidden - probably a
  wise move, judging by the Python community's painfully
  gained experience. :-)

* Not so sure about all the different syntactic sugars
  for grouping, though - Too Many Ways To Do It for my
  liking. :-(

* Different operators for integer and floating point division :-)

* Whether "-" is unary or binary depends on surrounding whitespace :-(

* Two different assignment operators, for value and reference
  assignment. :-( (Assignment seems to be quaintly referred to as
  "affectation" :-])

* Pliant introduces Yet Another Way To Spell Else-If: "eif". :-(

* There is a notion of types that can be compared for equality
  but not order (something lacking in Python at the moment). :-)

* There is a large and confusing set of argument passing and
  access modes, cryptically named. :-(

* The type system seems to include unsafe pointers. Refcounting is
  available, but you have to explicitly ask for it, so memory
  leaks and dangling pointers seem to be possible. :-(

* Not sure what to make of the OO stuff. It's based on the
  "generic function" idea, so you can add methods to an existing
  type. The docs don't say whether single or multiple dispatch
  is used. Also they say there is no "heritage" (which presumably
  means inheritance) but then go on to describe what looks a lot
  like an inheritance mechanism. So, I'm confused. :-(

  The author also tries to make a rather unconvincing argument
  that the generic-function approach makes programs easier to
  read than the class-oriented approach, but as far as I can
  see it only makes the problem worse!

* Globals live in modules, as in Python. Modules can have
  explicit "public" and "private" parts. :-)

* Something akin to Scheme macros can be written that operate
  on the parse tree, although it seems that they can go right
  down to the machine code level if desired. :-!

That's all I have time for now,

-- 
Greg Ewing, Computer Science Dept,
+--------------------------------------+
University of Canterbury,	   | A citizen of NewZealandCorp, a	  |
Christchurch, New Zealand	   | wholly-owned subsidiary of USA Inc.  |
greg at cosc.canterbury.ac.nz	   +--------------------------------------+


From rupole at compaq.net  Thu Apr 27 17:57:40 2000
From: rupole at compaq.net (Roger Upole)
Date: Thu, 27 Apr 2000 21:57:40 GMT
Subject: win32net setting a account to expire
References: <8e9sub$v7h$1@tracy.nacs.net>
Message-ID: <oJ2O4.18314$g8.1845193@news.easynews.com>

win32net.NetUserSetInfo(srvr,user,lvl,dict)
The dictionary contains varying info depending on the level.
You might have to play around with NetUserGetInfo to figure
out which level has the parameters you want to set by the
names of the keys in the dict returned.  There's more info in
the help file in the python\win32\help directory
       Roger Upole

"Aaron Hagan" <ahagan at speedynet.net> wrote in message
news:8e9sub$v7h$1 at tracy.nacs.net...
> Is there a function to set a user account to expire ..
> similar to the command NET USER <USERNAME> /EXPIRE:xx/xx/xx /DOMAIN ?
> thanks
>
>




From cnetzer at stanford.edu  Wed Apr 19 04:47:45 2000
From: cnetzer at stanford.edu (Chad Netzer)
Date: Wed, 19 Apr 2000 01:47:45 -0700
Subject: Python 1.6a2 crashes in test_fork1 on my Mandrake Linux 7 machine
References: <20000413010344.A13931@owl.rhein-zeitung.de> <38F658C6.66AD9CA5@clark.net> <20000414154718.A15664@xs4all.nl> <20000414190308.A30502@owl.rhein-zeitung.de> <38F91D1B.3D663B0C@stanford.edu>
Message-ID: <38FD72B1.DBB2F731@stanford.edu>

Well, as a followup to my last message, I recompiled my kernel for single-CPU
operation (I have a dual CPU machine), and was able to pass the "test_fork1.py" test
each time, so there appears to be a problem w/ threads on dual CPUs.  Whether it is
specific to Linux, Glibc, or Python's threadmodule, I have no idea.  But anyone else
with a dual CPU Linux (or other) setup would be a big help if they reported their
results. (Note, I tested this under Linux kernel 2.3.99pre5 and 2.2.15, with the
same results.  I'm using Glibc 2.1.3)

Neil Schemenauer sent me the following code, which demonstrates a problem quite
clearly.  It locks up on me in dual-CPU mode, but works fine in single-CPU mode.

I'll investigate more if I can, but learning about PDEs is keeping me very busy. :)

Chad Netzer
cnetzer at stanford.edu

--------------- [ cut here ] ----------------

import threading
import os, sys

class MyThread(threading.Thread):
    def start(self):
        threading.Thread.start(self)

    def run(self):
        while 1:
            if os.fork() == 0:
                print os.getpid()
                os._exit(0)
            os.wait()

MyThread().start()
MyThread().start()
MyThread().start()




From pup at pobox.com  Mon Apr 24 15:28:04 2000
From: pup at pobox.com (Joel Burton)
Date: Mon, 24 Apr 2000 15:28:04 -0400
Subject: Gzip -> FTP w/out using a temporary file?
Message-ID: <8e2733$2re8$1@nnrp-corp.news.cais.net>





From simpsons at kom.auc.dk  Wed Apr 26 16:09:22 2000
From: simpsons at kom.auc.dk (Thomas Rasmussen)
Date: Wed, 26 Apr 2000 20:09:22 GMT
Subject: global variables or inheritence
References: <28p8zy0irgc.fsf@lada.kom.auc.dk> <eRHN4.1113$3K4.19980@news.uswest.net>
Message-ID: <28p4s8oipzh.fsf@lada.kom.auc.dk>

>>>>> "Jeff" == Jeff Massung <jmassung at magpiesystems.com> writes:

  Jeff> Sorry if I get this wrong... but I thought you'd just import
  Jeff> your other file.

  Jeff> Example:

  Jeff> # File1.py myVar = 2


  Jeff> # File2.py import File1 print File1.myVar #should print 2

Well that was also what I thought, but the following program wont
execute:
testprog1.py:
import os, sys, string, testprog2
def main():
    print 'hello'
    myvar = 2
    testprog2.second()
if __name__ == '__main__':
    main()

testprog2.py:
import os, sys, string, testprog1
def second():
    print testprog1.myvar
if __name__ == '__main__':
    print 'nope'

    
and then I execute python testprog1.py and get the following error:

22:03 simpsons at lada% python testprog1.py                               ~/python
hejsa
Traceback (innermost last):
  File "testprog1.py", line 7, in ?
    main()
  File "testprog1.py", line 5, in main
    testprog2.second()
  File "testprog2.py", line 3, in second
    print testprog1.myvar
AttributeError: myvar

:-(

Thomas

-- 
KOM Network student helper

"To alcohol! The cause of - and solution to - all of life's problems!"
-- Homer Simpson


From ns566645 at swt.edu  Sun Apr 23 23:06:47 2000
From: ns566645 at swt.edu (Shah, Navneet)
Date: Sun, 23 Apr 2000 22:06:47 -0500
Subject: Scrollbar needed to get attached to Canvas
Message-ID: <3903BA47.A7CABC48@swt.edu>

I am using Pmw Mega widget. I  already  have canvas with buttons in it.
How can I attached scrollbar to the canvas. Any help will be
appreciated.

Shah

My code goes as follows:

class Demo:
    def __int__(self,parent)
        #Create the canvas
        self.sc=Pmw.ScrolledCanvas(parent,
                     borderframe=1,
                     labelpos ='n',
                     usehullsize = 1,
                     hull_width = 400,
                     hull_height = 300,
           )

          self.sc.pack()
          button=Tkinter.Button(self.sc.interior(), text='Hello')
          button.pack(side='left')






From erno at iki.fi  Mon Apr 17 17:41:13 2000
From: erno at iki.fi (Erno Kuusela)
Date: 18 Apr 2000 00:41:13 +0300
Subject: Filtering web proxy
References: <Pine.LNX.4.21.0004171406250.30999-100000@fep132.fep.ru>
Message-ID: <ku66tgh0cm.fsf@lasipalatsi.fi>

>>>>> "Oleg" == Oleg Broytmann <phd at phd.russ.ru> writes:

    Oleg> Hello!  I want a filtering web proxy. I can write one
    Oleg> myself, but if there is a thing already... well, I don't
    Oleg> want to reinvent the wheel. If there is such thing (free and
    Oleg> opensourse, 'course), I'll extend it for my needs.

i am using junkbuster. it's simple, and works well.

it only does url blocking (so banners), but i never have javascript
turned on anyway. it can block cookies selectively (i have everything
but slashdot blocked), hide/spoof user-agent, and use other http
proxies.

it doesn't do html parsing, and it's not written in python though.
but i haven't missed anything from it so it being written
in c hasn't bothered me much. with javascript/java turned
off there's not so much need for html parsing...

    Oleg>    I wrote dozen HTML parsers in Python, so I can write one
    Oleg> more, and turn it into a proxy, but may be I can start with
    Oleg> some already debugged code?

a html parser would need to work incrementally, unless you want to
wait for the whole document to be transferred over the network before
seeing any of it rendered.

i guess you could do it incrementally with sgmllib (iirc you feed it a
file object?), but you run into the fact that a big part of
the html documents on the web are malformed and rely on the
error correcting heuristics of the major browsers to function...

one starting point could be the "gray proxy" (i forget what it was
really called). that was written on top of medusa, i think there was
an announcement here? probably a year or so ago.  it parsed the html
and changed all the colors to grayscale, and did the same for
images. medusa isn't free though.. (except the version in zope?)

mozilla allows you to block/allow javascript by domain, iirc.

  -- erno







From bjorn at roguewave.com  Mon Apr  3 13:26:33 2000
From: bjorn at roguewave.com (Bjorn Pettersen)
Date: Mon, 03 Apr 2000 11:26:33 -0600
Subject: Pass by reference ?
References: <38E864C9.E1EE94F7@ecs.soton.ac.uk>
Message-ID: <38E8D449.FBC1E2AD@roguewave.com>

Jacek Generowicz wrote:
> 
> Hi,
> 
> I'm just trying to familiarize myself with Python.
> In Magnus Lie Hatland's Instant Python, he
> mentions that `all parameters in Python are passed
> by reference'.

The technically correct (although much less known) term is
pass-by-object-reference.  All this means is that you can change an
argument to a function by calling methods on the object, but you can't
change the identity of the object (ie. simply calling a function can't
change any name-bindings in your current scope).

Std disclaimer that anything can happen if you are willing to mess with
frame objects apply...

-- bjorn



From thehaas at my-deja.com  Wed Apr  5 16:22:38 2000
From: thehaas at my-deja.com (thehaas at my-deja.com)
Date: Wed, 05 Apr 2000 20:22:38 GMT
Subject: A Mountain of Perl Books + Python Advocacy
References: <20000405053920.20579.qmail@web2102.mail.yahoo.com>
Message-ID: <8cg79s$luf$1@nnrp1.deja.com>

In article <20000405053920.20579.qmail at web2102.mail.yahoo.com>,
  lewst <lewst at yahoo.com> wrote:
[snip]
> What is it about Perl that makes it so much more popular and have such
> a huge grassroots swell?  I personally find Perl an abomination and
> Python a breath of fresh air.  Perl has that first mover advantage I
> suppose, but should that really make such a hugh difference?

If you use Unix, Perl is easy to learn and you can write fairly powerful
scripts in a short amount of time.  Remember that Perl was basically
designed for system administration tasks, not necessarily for an
all-purpose language.  Sys Admins generally don't care how the
script/programs work, they just need something simple and need it
written fast.

One of my good friends is happens to be one of the biggest
language-bigots in the world.  I had been telling him for a long time
that he needed to try Python out, that he would like it (He was using
PHP and C++ before).  He refused, saying that he didn't want to be tied
down by the whitespace and that he didn't approve of how Python did
object-oriented stuff (I don't know why).  At this time, I wasn't that
serious about Python, but knew that he would like it (since he considers
Perl a freak of nature that should be avoided).  Well, he finally got
tired of PHP and C++ (I think the three days he spend trying to put in a
tree into PHP and failing did it) and bought Learning Python.  Now he
loves Python - I have never heard him say this many good things about
any programming language.  However, for the record, the whitespace
bothers him and he thinks how Python does object-oriented things is
weird, but he loves it and will not use anything else.  This got me more
serious about Python, so I bought Learning Python and am hooked (but I
haven't rewritten a lot of my Perl stuff yet, since most of the my Perl
things work well and I hate looking at Perl code).

The moral of the story: even if you are a language bigot, you will still
love Python.  So what isn't it more popular?  Who says it isn't popular?
 Perl has been out longer, but we are also seeing places move from Perl
to Python (or something else) because they don't want to maintain Perl
programs that are quickly becoming way too convolted.  Python is much,
much cleaner and is very easy to write programs in.  I predict that we
will see an expontential increase in Python use/awareness in the coming
months.  IMHO, it will be because of Perl and it's weaknesses.

- mikeh



Sent via Deja.com http://www.deja.com/
Before you buy.


From michael.stroeder at inka.de  Thu Apr 20 15:22:29 2000
From: michael.stroeder at inka.de (Michael =?iso-8859-1?Q?Str=F6der?=)
Date: Thu, 20 Apr 2000 21:22:29 +0200
Subject: socket.SO_REUSEADDR
References: <38FEEA44.6C05E2B0@inka.de> <8dn9o5$jsp$1@mawar.singnet.com.sg>
Message-ID: <38FF58F5.E1464322@inka.de>

Ng Pheng Siong wrote:
> 
> According to Michael Str?der  <michael.stroeder at inka.de>:
> >     self.socket.setsockopt(
> >       socket.getprotobyname('tcp'),
> >       socket.SO_REUSEADDR,1
> >     )
> 
> Michael,
> 
>     self.socket.setsockopt(socket.SOL_SOCKET, socket.SO_REUSEADDR, 1)

And with M2Crypto? (AttributeError: setsockopt)

Ciao, Michael.


From ns566645 at swt.edu  Mon Apr 24 02:25:05 2000
From: ns566645 at swt.edu (ABC)
Date: Mon, 24 Apr 2000 01:25:05 -0500
Subject: Tkinter.Button error
Message-ID: <3903E8C1.39B3E3D7@swt.edu>

I have two frames
top frame and bottom frame.
In the top frame I implemented a group.
Now, when I try to add Tkinter.Buttomn widget in the group in the top
frame it gives error . Can anyone help me and tell me what the error is?

Here is the code

frame1=Frame(master, relief=sunken', borderwidth='2')
frame2=Frame(master, relief='sunken', borderwidth='1')

frame1.place(relx=0, rely=0, relwidth=1, relheight=0.50)
frame2.place(relx=0. rely=0.50, relwidth=1, relheight=0.50)

group1=Pmw.Group(frame1, tag_text="Hello")
group1.pack(fill='both')

button=Tkinter.Button(group1.interior(), text='8.00')
button.pack()

The above program gives an error in TKinter in the second last line.
I have no clue of what the error is? Can anyone help me?
Need futher information, please let me know.









From ivanlan at callware.com  Tue Apr  4 19:01:38 2000
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Tue, 04 Apr 2000 17:01:38 -0600
Subject: Two questions
References: <38EA5FE8.853E9B75@callware.com> <38EA6508.90BF630C@roguewave.com> <38EA6E87.40EDB542@callware.com> <38EA6422.61054B11@exceptionalminds.com>
Message-ID: <38EA7452.9BFDF37E@callware.com>

Hi All, Tim--

Timothy Grant wrote:
> 
[snip]
> Ivan,
> 
> mxODBC works beautifully, but I've never used it to access and access
> database, only SQL database. Perchance do you know how Windows
> advertises an Access database to an ODBC driver?
> 

Nope, not yet.  However, I'll have to find out, won't I?

<you'll-be-the-first-tim-to-know>-ly y'rs,
Ivan
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
http://www.pauahtun.org 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
Author:  Teach Yourself Python in 24 Hours



From banderson at boi.hp.com  Wed Apr 19 12:57:35 2000
From: banderson at boi.hp.com (Bill Anderson)
Date: Wed, 19 Apr 2000 10:57:35 -0600
Subject: Python paradigms
References: <8cmr70$b0e$1@pegasus.csx.cam.ac.uk> <B515729E.1745%dgoodger@bigfoot.com> <38F00890.3C98EBDB@acm.org>
Message-ID: <38FDE57F.65FF34CF@boi.hp.com>

"Robert W. Cunningham" wrote:
...
>  blah = (((test and [true]) or [false])[0])             # blah = (test ?
> true : false)
> 
> Works for me!  Only if I have a good reason to suspect a problem in this
> particular line of code will I have to look past the comment.  Which, I
> believe, is the entire purpose of in-line comments in the first place.  The
> best way to get someone to read every single line of your code is to fail to
> properly comment it.
> 
> Are there ANY Python programmers who don't also know at least this much C?

Given that Python is being used to teach programming to people who don't
know it, yes. And you can expect that number to increase.

-- 
Bill Anderson 			Linux/Unix Administrator, Security Analyst
ESBU (ARC)			bill_anderson at boi.hp.com
My opinions are just that; _my_ opinions.


From mfletch at tpresence.com  Sun Apr 16 13:03:36 2000
From: mfletch at tpresence.com (Mike Fletcher)
Date: Sun, 16 Apr 2000 13:03:36 -0400
Subject: [ANNOUNCE] Garbage collection for Python
Message-ID: <CC59868A0A76D311B1F50090277B321C1F3277@VRTMAIL>

Hmm...

Stackless
Continuations (exposed to Python programmers! please :) )
Garbage collection
List comprehensions

together in Python 1.7/1.6.x would get you approximately an 8-really rating
(1 point for the lc's, 2 bonus for having all working together).  Oh, and
heal that type-class split while you're at it ;) .  If you get any extra
time :o) , add optional static typing and native compilation...  might get
yourself a full ten-really rating... and we all know what that means!

The first step on the path to Guido-hood is a path unto itself.
The second is a lifetime wandering in the desert of code.
At the third step, your true journey begins.
Mike

-----Original Message-----
From: Christian Tismer [mailto:tismer at tismer.com]
Sent: Sunday, April 16, 2000 12:54 PM
To: Just van Rossum
Cc: Fredrik Lundh; python-list at python.org
Subject: Re: [ANNOUNCE] Garbage collection for Python




Just van Rossum wrote:
> 
> Michael Hudson <mwh21 at cam.ac.uk> wrote:
> > To someone who might know: what hope is there for this to go
> > into 1.6?
> [ ... ]
> > It would be, in my opinion, be a really really really good thing; for
> > one thing it would help in language wars...
> 
> Fredrik Lundh wrote:
> > sorry, but stackless python is the only really really really good
> > thing that hasn't yet been added to the core.  this is just really
> > really good.

Hey! If Stackless goes with GC, can we get 5 times "really"? :-))

...



From jtc at dimensional.com  Tue Apr 18 17:38:10 2000
From: jtc at dimensional.com (Jim Cochrane)
Date: 18 Apr 2000 15:38:10 -0600
Subject: Date class?
Message-ID: <8dikk2$p4s@flatland.dimensional.com>

I just checked the python books I have and did a search of the standard
library and was suprised to find that Python has no standard date class.
Does anyone know if there is a good date class available?  I did come
across a module called mxDateTime in a web search that has a DateTime class
and it looks like this may be a defacto standard.  Is that the case?

Concidentally, I just read here that the starship server just died; and
that appears to be the only place where mxDateTime is available.  Does
anyone know of another site that has the mxDateTime package?

Thanks!
-- 
Jim Cochrane
jtc at dimensional.com


From vetler at ifi.uio.no  Sat Apr  1 05:24:38 2000
From: vetler at ifi.uio.no (Vetle Roeim)
Date: 01 Apr 2000 12:24:38 +0200
Subject: Emacs Python mode
Message-ID: <yayk8iii0ex.fsf@ganglot.ifi.uio.no>

I just had an idea. Using the Emacs python-mode, would it be possible
to somehow hide the function-bodies?

Sometimes it would be useful to get an overview of the functions
provided by a class.


vr


From philh at vision25.demon.co.uk  Wed Apr 26 12:12:17 2000
From: philh at vision25.demon.co.uk (phil hunt)
Date: Wed, 26 Apr 2000 17:12:17 +0100
Subject: Python Time classes
Message-ID: <slrn8ge5b0.1gu.philh@vision25.demon.co.uk>

Is there a high-level Time class for Python?

I'm aware of the 'time' module in the standard library. But I'm
looking for something more high-level (I can't find anything
relevant on Parnassus or Freshmeat). What I want is roughly the
ability to say things like:

   t1 = Time.now()   
   # (t1) is an instance of Time containing the time now

   t2 = Time(2000, 4, 26, 17, 4, 0) 
   # (t2) is 2000-Apr-26 17:04:00

   sec = t1 - t2   
   # (sec) is the number of seconds between (t1) and (t2)

   str = t1.format("%H:%M:%S")
   # (str) is a string containing the hours minutes and seconds of (t1)
   # in a form like 23:59:06

It'd also be nice if there was a similar Date class which stored
a date (this would behave like a Time except that (i) the hh:mm:ss
would always be 00:00:00, and (ii) subtracting two dates would return
the number of days difference). Obviously times and dates would be
easily convertable to each other:

   aTime = Time(aDate)
   aDate = Date(aTime)

(or similar).


Does this exist, or shall I write one?

-- 
***** Phil Hunt ***** send email to phil at comuno.com *****
Moore's Law: hardware speed doubles every 18 months
Gates' Law: software speed halves every 18 months 


From tom__98 at my-deja.com  Wed Apr 12 02:15:34 2000
From: tom__98 at my-deja.com (tom__98 at my-deja.com)
Date: Wed, 12 Apr 2000 06:15:34 GMT
Subject: Python/Perl Popularity (Re: A Mountain of Perl...)
References: <200004102111.HAA03380@envy.fulcrum.com.au> <8d0pcq$7ti$1@nnrp1.deja.com> <g0ssm03o.fsf@dan.resnet.rochester.edu>
Message-ID: <8d149t$jo6$1@nnrp1.deja.com>

>     tom>  -- How can I browse the Python documentation without firing
>     tom> up a web browser?
>
> Convert it to text format. It's latex, you can make it into anything
> you want, and do anything you want with it.

Well, what I'm suggesting is that it might be a good idea
to make a command-line accessible version of the documentation,
together with a command line tool, available as part of the
standard distribution.

>     tom>  -- Is there a Python equivalent of CPAN and the Perl CPAN
>     tom> module?  The closest I know of is Parnassus.  But Parnassus
>     tom> is merely a collection of links, not an archive, and it
>     tom> doesn't have any facilities (AFAIK) for automatic
>     tom> installation.
>
> So, who gives a crap?
> I try not to be that lazy.

So, you are saying that you basically agree: compiling and
installing modules is more work on Python, but that really
shouldn't matter because expecting to install modules quickly
is just laziness.  Well, I admit it, I'm lazy.

perldoc and the Perl CPAN module are really nice and well done,
and as a user, I miss that functionality when I use Python.
Why not learn from Perl where they did something good? This
antagonism between the two camps, expressed in your and other
posts, just seems really silly to me as a user of both languages.

Tom.



Sent via Deja.com http://www.deja.com/
Before you buy.


From roy at popmail.med.nyu.edu  Wed Apr  5 09:32:47 2000
From: roy at popmail.med.nyu.edu (Roy Smith)
Date: Wed, 05 Apr 2000 09:32:47 -0400
Subject: What is python's language level?
References: <38EA1B8F.A797A130@lmco.com> <38E9DDD3.3313FF6@bioreason.com> <slrn8ekddq.gm3.zawrotny@gecko.sb.fsu.edu> <roy-D857DA.21212304042000@netnews.nyu.edu> <38EA9CD4.3473B910@callware.com> <bt3ptdy4.fsf@dan.resnet.rochester.edu>
Message-ID: <roy-7D22BC.09324705042000@netnews.nyu.edu>

dan at cgsoftware.com wrote:
> Reminds me of a neat little raytracer written in postscript someone
> posted many years ago.
> 
> I was just playing with it a few minutes ago.
> About 7k, raytraces 5 spheres by default.
> you have to set the max ray depth to about 3 or 4 or it'll take
> forever.

I remember running that on our original Apple LaserWriter many years ago.  I 
have no idea how long it took; I gave up waiting and went home for the night.  
The next morning, it had printed the image.

-- 
Roy Smith <roy at popmail.med.nyu.edu>
New York University School of Medicine



From thomas at xs4all.net  Fri Apr 28 16:57:28 2000
From: thomas at xs4all.net (Thomas Wouters)
Date: Fri, 28 Apr 2000 22:57:28 +0200
Subject: Python does not let me format time.
In-Reply-To: <slrn8ghpeo.f01.scarblac-spamtrap@flits104-37.flits.rug.nl>; from scarblac-spamtrap@pino.selwerd.nl on Fri, Apr 28, 2000 at 01:14:46AM +0000
References: <8eakhq$797$1@nnrp1.deja.com> <slrn8ghpeo.f01.scarblac-spamtrap@flits104-37.flits.rug.nl>
Message-ID: <20000428225727.K11481@xs4all.nl>

On Fri, Apr 28, 2000 at 01:14:46AM +0000, Remco Gerlich wrote:
> Sindh wrote in comp.lang.python:
> > If there is already a moduls to do the same , much helpful.

> >>> import time
> >>> time.strptime("00/04/28","%y/%m/%d")
> (2000,04,17,0,0,0,6,1,0)

> strptime is the inverse of strftime, in a way.

But unfortunately, not all operating systems supply it, and if they do not,
then the time module doesn't, either. If you expect your app to be portable,
dont rely on time.strptime ;) (Or write a replacement strptime, of course,
and get it included in the standard time module. I'd like that a lot, but
haven't the time to figure out all the gory strptime details.)

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From taashlo at sandia.gov  Thu Apr  6 08:54:46 2000
From: taashlo at sandia.gov (taashlo at sandia.gov)
Date: 06 Apr 2000 06:54:46 -0600
Subject: What/where is HTMLBuilder?
References: <CB8AAF150D049DFB.1B47358A250E7184.84110B3E40F5C2FB@lp.airnews.net>
Message-ID: <u3doz75k9.fsf@sandia.gov>

Ronald E Jeffries <ronjeffries at acm.org> writes:

> I've searched my docs and python.org for HTMLBuilder with no luck.
> Where, please?
> 
> Thanks!
> 
> Ron Jeffries
> http://www.XProgramming.com


Ron,

I found it by going to
<http://www.python.org/topics/xml/download.html> and following the
link to "Current working files for the XML-SIG" at the bottom of the
page and downloading PyXML-0.5.3.tar.gz.

Tad


From wtanksle at dolphin.openprojects.net  Sat Apr  1 22:30:28 2000
From: wtanksle at dolphin.openprojects.net (William Tanksley)
Date: Sun, 02 Apr 2000 03:30:28 GMT
Subject: New Features in Python 1.6
References: <200004011740.MAA04675@eric.cnri.reston.va.us> <_pyF4.25562$E85.640966@news1.rdc1.md.home.com>
Message-ID: <slrn8edfmk.185.wtanksle@dolphin.openprojects.net>

On Sun, 02 Apr 2000 02:39:54 GMT, Eric Hagemann wrote:
>What are the options for users wanting 1.6 that may not have access to
>modules.python.org
>on a regular basis or at all.  I'd hate to be disconnected from the net and
>find I need that
>'one' function that used to be in the standard library ......... e.g. can I
>down load the
>entire library and forgo the net check ?  What happens if there is no net --
>will the import function
>get unhappy ?

It's actually not a problem -- all modern OSes support TCP/IP even when
you're disconnected from the net, and one particular address, 127.0.0.1,
is always live.  This happens to be the site from which Python loads its
modules (we have an agreement with the owner of that site, which also
hosts many other products of the same nature, including some marketing
apps).

>Also when you import like this does the code check the net for a newer
>version or only when you have no version of the particular sub-module ?

Only when you have no version.  That's not a bad idea, though.  I suspect
it would be best as a single, user-initiated command, though.

>"Guido van Rossum" <guido at python.org> wrote in message

>> 2. Imputils

>> Complementary to the Distutils are the Imputils, or Import Utilities.
>> Python's import mechanism has been reworked to make it easy for Python
>> programmers to put "hooks" into the code that finds and loads modules.
>> The default import mechanism now includes hooks, written in Python, to
>> load modules via HTTP from a known URL.
>>
>> This has allowed us to drop most of the standard library from the
>> distribution.  Now, for example, when you import a less-commonly-needed
>> module from the standard library, Python fetches the code for you.  For
>> example, if you say
>>
>>     import tokenize
>>
>> then Python -- via the Imputils -- will fetch
>> http://modules.python.org/lib/tokenize.py for you and install it on your
>> system for future use.  (This is why the Python interpreter is now
>> installed as a setuid binary under Unix -- if you turn off this bit, you
>> will be unable to load modules from the standard library!)
>>
>> If you try to import a module that's not part of the standard library,
>> then the Imputils will find out -- again from modules.python.org --
>> where it can find this module.  It then downloads the entire relevant
>> module distribution, and uses the Distutils to build and install it on
>> your system.  It then loads the module you requested.  Simplicity
>> itself!

-- 
-William "Billy" Tanksley


From bobalex at home.com  Tue Apr  4 12:47:12 2000
From: bobalex at home.com (Bob Alexander)
Date: Tue, 4 Apr 2000 09:47:12 -0700
Subject: Tuples -- who needs 'em
References: <00e101bf9d80$a8384d20$74eb0b18@stcla1.sfba.home.com> <va3G4.1110$n68.190454272@newsb.telia.net>
Message-ID: <037f01bf9e55$6d328760$74eb0b18@stcla1.sfba.home.com>

"Fredrik Lundh" <effbot at telia.com> wrote:

> -- http://www.python.org/doc/FAQ.html#6.15

Thanks for the pointer, but I'd already seen it. My post was not about lack
of understanding of tuples, it was simply intended to provoke some
interesting discussion. I realize tuples are probably a religious issue in
the Python community, so please don't take it the wrong way.

Let me repeat that I really like Python, I've written a lot of code with it,
I don't have any trouble understanding the concept of tuples vs. lists, and
I use both constructs at appropriate times. But I'm exploring the idea that
programming life might actually be somewhat better with lists and tuples
unified to a single datatype. Then there would be no need for FAQ # 6.15 :-)

I'm aware that this issue is not some great flash of insight. It is so
obvious that the Python in-group certainly have given it plenty of thought.
But that doesn't mean it's not worth talking about,

> -- things should be as simple as possible, but no simpler.

(Cool expression -- did you make that up?)

I completely agree, but several very good computer languages are
existence-proofs that lack of immutable lists is not "simpler than
possible".

> -- from a human perspective, using different syntax for
>    different things is good.

I agree with this, too (noting Lisp as possibly the worst example --
although a fine language with clear favorable influences on Python). But
my issue is not about syntax.

However, too bad about that nasty syntax irregularity that using parens for
tuples presents  (1,)  (Couldn't resist  ;-)

>    tuples are simple structures.
>    lists are collections of homogenous (in one sense or
>    another) values.  from a program design perspective,
>    that are two radically different things.

Well, this is a half-empty vs. half-full thing -- I view them as pretty
similar. As a programmer I can choose to not modify a list and voila, it's
basically a tuple. It's not really necessary to have the language enforce
this (although it can be helpful in a large project).

BTW, both tuples and lists are ordered collections of non-homogeneous values
(i.e. the elements can be if different types).

> -- tuples can be used as dictionary keys.  mutable
>    collections can not.

This is the most compelling justification I've seen so far.

> -- python's core types fall into two categories: the first
>    is simple types, like numbers, strings, and tuples. they
>    are all immutable.
>
>    the other category is container (or collection) objects.
>    they're far more powerful, and harder to use for beginners.

Yet the concept of array is pretty fundamental to programming. Arrays are
usually thought of as mutable. Anyone having trouble with that concept is
pre-beginner  :-)

Without tuples, there would be no need to explain to
beginners the difference between tuples and lists. The fact that this is a
recurring topic on this list says something about that.

> -- this has been discussed over and over again.  have you
>    studied earlier threads on this topic?

No.

> -- good designers know that hypergeneralization tends to be
>    a bad idea.

Yes -- there is a proper balance for everything. I've happily used languages
that don't have immutable lists, though, and was
never aware that they were lacking something.

By the way, if immutable lists are necessary, why are immutable dictionaries
not necessary? Seems that all the same arguments would apply.

> -- what would happen if we took them away?
>
> -- do we really need yet another "I don't fully understand this,
>    so it should go away" thread?  do you really think that tuples
>    are just an accidental feature?  (just asking...)

I'm sure tuples are not an acciental feature. And my issue comes from
understanding, not lack of understanding. I harbor a suspicion that the
motivation for tuples might have been more for implementation concerns
than for users' convenience. More on this below.

> > Suppose Python had only one sequence data type: list, and that we could
> use
> > it in all places where tuples are needed now. I would never have to
think
> > about whether to use tuple or list. I would never have to convert from
> tuple
> > to list or vice versa. Just one more thing I don't have to bother with.
>
> if you find that you end up doing this all the time, you're not
> following the "fixed structures vs. homogenous collections"
> design rule.
>
> (or you're using extension modules that haven't been upgraded
> to use the abstract sequence API.  most standard extensions do,
> these days).
>
> if you don't do this all the time, what's the big deal?

Well, okay, I don't do it all the time, and it's not a big deal. I'm just
exploring the opportunity for a small reduction in the "cognitive load" of
Python programming. In my opinion, it's the difference in cognitive load
that mainly differentiates ease of programming in languages like Python from
languages like C++. Thankfully we don't have to decide among 10 or so
different collection types. But even small choices to make when choices are
not necessary cause the mind to leave the problem that is being solved to
deal with a programming "bookkeeping" issue.

> > I suspect there will be several answers offered as to why we need both
> > types, but I also suspect most of the benefits are for the Python
> > implementor(s), not programmers.
>
> umm.  I thought having two similar types meant more work for
> the python implementors, not less?

I wish there was something like the C++ ARM for Python, where choices in
language design are discussed and justified. I don't particularly like C++,
but I like that book because of the insight it provides. If there is such a
book or paper on Python, I'd appreciate a pointer to it.

There is a design pattern (that is not in the book!) where some program
features exist more for for benefit of the program implementor(s) than for
the program user(s). Is there some element of this in the existence of
tuples as a source language construct? I'm aware the the feature has been
well-sold as a benefit to programmers.

My suspicion (lacking any real knowledge about this, and with deepest
respect for the designer(s) of Python) is that there was
already a need for an internal "tuple" datatype in the language's
implementation, and it was efficient for
things like function argument frames, and maybe for
multiple function return values, etc. Given that, maybe apply() could be
implemented
 more efficiently if tuples were exposed at the source language level and
the
user passed tuples to apply(). (Note that apply() will convert sequences to
tuples if they are not already, thus efficient use of apply requires that
tuples be passed, not lists.) Given this sort of justfication for tuples,
many more
reasons were then noticed why tuples are a Good Thing.

I'm all for efficient implementation, but Python usually opts for
programming ease over raw efficiency. I agree with that philosophy -- design
for ease of use and eventually techniques will be figured out for
making the underlying implementation more efficient.

-- Bob







From effbot at telia.com  Tue Apr 18 14:50:49 2000
From: effbot at telia.com (Fredrik Lundh)
Date: 18 Apr 2000 13:50:49 -0500
Subject: Dr. Dobb's Python-URL! - weekly Python news and links (Apr 18)
Message-ID: <C68CF56240C3E4C6.8487B526202CCFE5.1FB6D07DA8BB3257@lp.airnews.net>

>From the minor disaster department:
    starship.python.net is down.  It will be up again soon,
    but most of the contents have been lost in a major disk
    crash.  If you, for some reason, have mirrored portions
    of it, contact the maintainers.  More info here:
        http://www.deja.com/=dnc/getdoc.xp?AN=612763208
        http://www.deja.com/=dnc/getdoc.xp?AN=612100048

In other news:
    Software Carpentry: The Design Competition Entries
    are in:
        http://software-carpentry.codesourcery.com/entries/index.html

    Ng Pheng Siong: M2Crypto 0.04
        http://mars.post1.com/home/ngps/m2/
        http://www.deja.com/=dnc/getdoc.xp?AN=612311988

    John Aycock: SPARK 0.6
        http://csr.uvic.ca/~aycock/python
        http://www.deja.com/=dnc/getdoc.xp?AN=610996110

    CADScript: CAD data access from Python
        http://www.deja.com/=dnc/getdoc.xp?AN=610644248

    Mark Lutz: Python Training in Boulder, Colorado
        http://www.deja.com/=dnc/getdoc.xp?AN=610763192

    Cristian Echeverria: Theme support for Tkinter
        http://www.geocities.com/cristian_echeverria/Chevegui.html

    PyUnit - a unit testing framework for Python
        http://pyunit.sourceforge.net/


========================================================================

Everything you want is probably one or two clicks away in these pages:

      Python.org's Python Language Website is the center of Pythonia
          http://www.python.org

      eff-bot's complements this digest with his daily python url
          http://hem.passagen.se/eff/url.htm

      Much of Python's real work takes place on Special-Interest Group
          mailing lists http://www.python.org/sigs/

      Python Consortium emerges as an independent nexus of activity
          http://www.python.org/consortium

      The Vaults of Parnassus ambitiously collects Python resources
          http://www.vex.net/~x/parnassus/   

      Python To-Do List anticipates some of Python's future direction
          http://www.python.org/cgi-bin/todo.py

      Python Journal is at work on its second issue
          http://www.pythonjournal.com

      Links2Go is a new semi-automated link collection; it's impressive
      what AI can generate
          http://www.links2go.com/search?search=python

      Archive probing trick of the trade:
          http://www.dejanews.com/dnquery.xp?QRY=&DBS=2&ST=PS&defaultOp=AND&LNG=ALL&format=threaded&showsort=date&maxhits=100&groups=comp.lang.python


Previous - (U)se the (R)esource, (L)uke! - messages are listed here:
  http://purl.org/thecliff/python/url.html
or
  http://www.dejanews.com/dnquery.xp?QRY=~g%20comp.lang.python%20Python-URL%21


Suggestions/corrections for next week's posting are always welcome.
  http://www.egroups.com/list/python-url-leads/

To receive a new issue of this posting in e-mail each Monday morning,
ask <claird at neosoft.com> to subscribe.  Mention "Python-URL!".


-- The Python-URL! Team--

Dr. Dobb's Journal (http://www.ddj.com) is pleased to participate in and
sponsor the "Python-URL!" project.
-- 

Cameron Laird <claird at NeoSoft.com>
Business:  http://www.Phaseit.net
Personal:  http://starbase.neosoft.com/~claird/home.html


From thehaas at binary.net  Fri Apr 21 14:59:48 2000
From: thehaas at binary.net (Mike Hostetler)
Date: Fri, 21 Apr 2000 13:59:48 -0500
Subject: compress_web.py
Message-ID: <3900A524.4A32674D@binary.net>

I have a web page that doesn't have many large pieces, but it has many
large ones.  When I added a page, I would have to update the index, etc.
It was hard to remember which pages I updated, so I just tar'ed them all
up.  This got old, because I had to upload a sort-of large tarball
(~100k gziped) when I only created one html file less then 3k.  I
thought, "You know, it would be really cool if I wrote a program that
would check the date of the files, then just compress the new files." 
So I did.  The original was written in Perl, and it sucked.  Then I
discovered Python.  The Python version was not only cleaner, but it's
even faster.

So here it is, submitted for your approval.  I'm including it here in
case it's useful for someone, either to use the script yourself or to
learn a bit more about Python (I know I learned a lot of Python writing
this).  The license is GPL'ed.  I also signed it with gpg, for those
interested/worried about such things.  My gpg signature is available at
http://www.binary.net/thehaas/mikeh.gpg 

Email me if you have questions, comments, or sympathies.

- mikeh

-- 
Mike Hostetler          
thehaas at binary.spam_must_die.net 
http://www.binary.net/thehaas 
GnuPG key: http://www.binary.net/thehaas/mikeh.gpg
-------------- next part --------------
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

#!/usr/local/bin/python
######################################
# compress_web.py
# (C) 2000  by Mike Hostetler
######################################
#    This program is free software; you can redistribute it and/or
#    modify
#    it under the terms of the GNU General Public License as published
#    by
#    the Free Software Foundation; either version 2 of the License, or
#    (at your option) any later version.
#
#    This program is distributed in the hope that it will be useful,
#    but WITHOUT ANY WARRANTY; without even the implied warranty of
#    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
#    GNU General Public License for more details.
#
#    You should have received a copy of the GNU General Public License
#    along with this program; if not, write to the Free Software
#    Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
#
######################################

import os,stat,time,string

#####################################
def compress_web(tar,days,webdir,bad_files):

	os.chdir(webdir)

	newfiles = []

	os.path.walk(webdir,find_time,newfiles)
	
	if (newfiles != []):
		for i in newfiles:
			tar = tar+ " " + i
		os.system(tar)
	else: 
		print "\n### no files found that needed archived\n"

#####################################
# this function is called by the os.path.walk for each file in webdir 
def find_time(newfiles, dirname, names):

	strlen = len(webdir)+1
	time_limit = time.time() - (86400 * days)

	for file in names:	
		filename= dirname+"/"+file

# Wow, this line sucks.  But it works.
# What it does:
#        1st condition - is it in list of files we don't want?
#        2nd condition - is it a regular file (i.e. not a directory)?
		if not(file in bad_files) and (stat.S_ISREG(os.stat(filename)[stat.ST_MODE])):

# This checks to see if the file meets the time requirement
			if os.stat(filename)[stat.ST_MTIME] > time_limit:

# It is important to strip "filename" by the strlen - this makes sure we
# grab the directory name of the file as well.  If we just used "file",
# then we wouldn't grab the directory.  That would be bad.
				newfiles.append(filename[strlen:])
	return newfiles
#####################################

if __name__ == '__main__':

	webdir = "/home/mikeh/public_html"
	bad_files = ("web.tar.gz","dan_o.dat")
	days = 2
	tar_command = "tar cvzf web.tar.gz"

	compress_web(tar_command,days,webdir,bad_files)

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.0 (GNU/Linux)
Comment: For info see http://www.gnupg.org

iD8DBQE5AKFuaP33v4T41CURAkLwAKCmZEptU0rtPt05dT1qKhM25qDJSQCgi9ms
naDnMjhIo3Op0e4JAzKgw8w=
=qj24
-----END PGP SIGNATURE-----

From tim_one at email.msn.com  Sat Apr  1 23:20:03 2000
From: tim_one at email.msn.com (Tim Peters)
Date: Sat, 1 Apr 2000 23:20:03 -0500
Subject: Why should I switch to Python?
In-Reply-To: <Pine.GSO.4.10.10004020056440.20943-100000@sundial>
Message-ID: <000701bf9c5a$b806b420$752d153f@tim>

[Tim]
> There aren't a lot of bilingual P programmers out there ...

[Moshe Zadka]
> I wonder what made you say that?

Why, measurement, of course!

> Actually, I know quite a few bi-lingual P'ers (my ex-girlfriend, her
> ex-boyfriend, and her current boyfriend.  Shows you what kind of crowd
> I mix in) and no highly-proficienct one-P programmers (sure, some Perl
> dabblers, but nothing serious).

See -- that's just anecdotal.  When in doubt, there's no substitute for
exhaustive enumeration.

> So from my prespective, most people know both P languages.

While from mine, people are fascinated by floating-point endcases, think
"Twin Peaks" is the greatest television show of all time, believe "Alien" is
the greatest horror movie all time, "Aliens" the greatest monster movie of
all time, and that the world would be a much better place if professional
wrestling were taught at the JFK School of Government instead of diplomacy.

no-arguments-about-the-rest-ly y'rs  - tim





From tiddlerdeja at my-deja.com  Thu Apr 27 13:52:35 2000
From: tiddlerdeja at my-deja.com (tiddlerdeja at my-deja.com)
Date: Thu, 27 Apr 2000 17:52:35 GMT
Subject: web site test framework? in python?
Message-ID: <8e9uor$eaa$1@nnrp1.deja.com>

I'm looking to build a test framework to use for websites that I may
build.

I'm looking for more than broken link checks. I'd like to post form
submissions to pages and check for correct/desired results to be
returned.

I know I could write specific scripts for each site, but I'd like a
generic solution. I was wondering if anyone else had been down this
path.

I was thinking about solutions along the lines of...

I was thinking of having JSP or ASP pages check for a testing variable
(registry or environment). If pages were called in test mode they would
return XML to be parsed which would can then be used as input data to
call the page again (this way you could store the test data with the
page you are testing). The XML would also contain the desired correct
data to check against to ensure the page is working correctly.

You could then have a script crawling a site, calling a page once to
get the test case, and calling it again to see if it has worked
correctly.

There are problems with this solution in that you can have dependencies
on pages. I.e. one page must be called before another if it is to work
correctly.

I'd also like to code the testing framework in Python.

Again, has anyone any experience in this testing websites arena?

Perhaps I'm trying to reinvent the wheel?

Any info greatly appreciated.


Sent via Deja.com http://www.deja.com/
Before you buy.


From scarblac-spamtrap at pino.selwerd.nl  Sun Apr 16 13:50:53 2000
From: scarblac-spamtrap at pino.selwerd.nl (Remco Gerlich)
Date: 16 Apr 2000 17:50:53 GMT
Subject: Python paradigms
References: <8cmr70$b0e$1@pegasus.csx.cam.ac.uk> <8dbdj3$dgt$1@slb2.atl.mindspring.net> <8dcpom$fbc$1@slb3.atl.mindspring.net>
Message-ID: <slrn8fk2kh.965.scarblac-spamtrap@flits104-37.flits.rug.nl>

Andrew Dalke wrote in comp.lang.python:
> Aahz Maruch wrote:
> >def expression_if( test, true, false ):
> >  if test: return true
> >  else: return false
> >
> >x = expression_if( a is not None,
> >   a[i].weeble, 0) + expression_if( b is not None, b[i].wombat, 0)
> 
> 
> Suppose a is None.  The a[i].weeble will give an exception.

Pass functions?

def exp_if(test, true, false):
  if test: return true()
  else: return false()
  
x = exp_if(a is not None, lambda a=a,i=i: a[i].weeble, lambda: 0)

Getting silly, though :)

-- 
Remco Gerlich,  scarblac at pino.selwerd.nl
  8:46pm  up 41 days,  8:07,  6 users,  load average: 1.19, 1.08, 1.07


From gherman at darwin.in-berlin.de  Tue Apr 18 06:33:21 2000
From: gherman at darwin.in-berlin.de (Dinu C. Gherman)
Date: Tue, 18 Apr 2000 12:33:21 +0200
Subject: Starship dead (again)?!
Message-ID: <38FC39F1.21DD2D75@darwin.in-berlin.de>

I can't get anything from http://starship.python.net or
http://starship.python.net:8080 for more than a day, now.
Is that some known problem?

This seems to happen ever more frequently. Does somebody
simply pull the plug, or is there another, better, reason?

Regards,

Dinu

-- 
Dinu C. Gherman
................................................................
"The thing about Linux or open software in general is that 
it actually tries to move software from being witchcraft to 
being a science," [...] "A lot of the programs you see today 
are actually put together by shamans, and you just take it and 
if the computer crashes you walk around it three times... and 
maybe it's OK." (Linus Thorvalds, LinuxWorld 2000, NYC)


From paul.magwene at yale.edu  Thu Apr 27 09:55:59 2000
From: paul.magwene at yale.edu (Paul Magwene)
Date: Thu, 27 Apr 2000 09:55:59 -0400
Subject: Concatening string and integer
References: <8e9gco$jlt$1@vg170.it.volvo.se>
Message-ID: <390846EF.79A67E02@yale.edu>

JJ wrote:
> 
> Hi!
> 
> I want to concatenate a string containing both strings and integers like:
> 
> strTemp = 'I make test number: ' + iMyNumber + ' once again'
> 
> It doesn't work, how should I do this?
> 
> BR
> Joacim


Use the str() function (one of the builtins), like so:

strTemp = "I make a test number: " + str(iMyNumber) + " once again"


--Paul


From z3penguin at penguinpowered.com  Sun Apr 23 19:43:19 2000
From: z3penguin at penguinpowered.com (Z 3 Penguin)
Date: Sun, 23 Apr 2000 19:43:19 -0400
Subject: Posting CGI
References: <8dv4j8$2bs$1@bob.news.rcn.net> <JzFM4.3672$Za1.53909@newsc.telia.net>
Message-ID: <8e0234$rs9$1@bob.news.rcn.net>

F-
Got it, thanks
-Z




From mfletch at tpresence.com  Mon Apr  3 15:20:43 2000
From: mfletch at tpresence.com (Mike Fletcher)
Date: Mon, 3 Apr 2000 15:20:43 -0400 
Subject: Creating variables on the fly...
Message-ID: <CC59868A0A76D311B1F50090277B321C1F324B@VRTMAIL>

Here's a few ways...  Note the *args construct... also note that you could
use **namedargs to indicate arguments which are passed explicitly by name.
*args gets a tuple of values in the order passed, **namedargs would get a
dictionary mapping name:value.

PythonWin 1.5.2 (#0, Apr 13 1999, 10:51:12) [MSC 32 bit (Intel)] on win32
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
Portions Copyright 1994-1999 Mark Hammond (MHammond at skippinet.com.au)
>>> def add( *args ):
... 	return reduce( lambda x,y: x+y, args )
... 
>>> add( 2,3,4,5,6,7)
27
>>> import operator
>>> def add2( *args ):
... 	return reduce( operator.add, args )
... 
>>> add2( 2,3,4,5,6,7)
27
>>> import Numeric
>>> Numeric.sum( (2,3,4,5,6) )
20
>>> 

HTH,
Mike



From mwh21 at cam.ac.uk  Tue Apr 18 12:03:51 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 18 Apr 2000 17:03:51 +0100
Subject: Integer solutions to linear equation?
References: <slrn8fovo2.nq.grant@grante.comtrol.com>
Message-ID: <m33dojcs60.fsf@atrus.jesus.cam.ac.uk>

grant at nowhere. (Grant Edwards) writes:

> This isn't really a Python question, but my example is in
> Python, and there seem to be plenty of people who know a fair
> bit of math here....

I hope I count ...

> A friend of mine ran across a brain-teaser involving a bunch of
> flowers, some magic bowls and some other camoflage text.  What
> you end up with is having to solve the equation
> 
>      64x = 41y + 1
>      
> Where x and y are both integers.  After scratching our heads
> for a while, we used the brute force approach:
> 
> for x in range(1,100):
>     y = ((64*x)-1)/41
>     if 64*x == 41*y+1:
>         print (x,y)
>                 
> The results:
> 
>   (25, 39)
>   (66, 103)
> 
> 25 was the expected solution, so we got both the equation and
> the Python snippet correct.  Is there a non-iterative way to
> solve a linear equation when the results are contrained to be
> integers?  I don't remember anything from any of my math
> classes that seems relevent, but I didn't take any anything
> beyond what is required for all undergrad engineers.

Use Euclid's algorithm; here's some code that does that:

def find_hcf_as_lc(x0,y0):
    """ return h,q,r st h==q*x0+r*y0 """
    x=abs(x0); y=abs(y0)
    qn1=rn=1
    rn1=qn=0
    while y:
        (q,y),x=divmod(x,y),y
        qn,qn1=qn1-q*qn,qn
        rn,rn1=rn1-q*rn,rn
    if x0<0: qn1=-qn1
    if y0<0: rn1=-rn1
    return x,qn1,rn1

def solve(m,n,a):
    """ solve(m,n,a:Integer) -> (Integer,Integer)

find integers (x,y,r,s) such that (m+u*r)*x+(n+u*s)*y==a 
for all integers u."""
    h,x,y = find_hcf_as_lc(m,n)
    if a%h <> 0:
        raise "no solution"
    else:
        return (x*a/h,y*a/h,n/h,-m/h)

Then:

>>> solve(64,-41,1)
(-16, -25, -41, -64)

And your solution is:

(-16 - -41)*64 + (-25 - -64)*-41

I feel I must have neater code knocking about somewhere to do this,
but I can't find it right now.

Cheers,
M.

-- 
  it's not that perl programmers are idiots, it's that the language
  rewards idiotic behavior in a  way that no other language or tool 
  has ever done                        -- Erik Naggum, comp.lang.lisp


From bjorn at roguewave.com  Sat Apr  1 22:06:17 2000
From: bjorn at roguewave.com (Bjorn Pettersen)
Date: Sat, 01 Apr 2000 20:06:17 -0700
Subject: Python 1.6 alpha 1 released
References: <200003312130.QAA04361@eric.cnri.reston.va.us> <38E5205F.DE811F61@roguewave.com> <Wf9F4.4423$74.70231@newsc.telia.net> <38E58D2D.384F1546@Lugoj.Com> <S9kF4.4472$74.71414@newsc.telia.net>
Message-ID: <38E6B929.28816494@roguewave.com>

Fredrik Lundh wrote:
> 
> James Logajan <JamesL at Lugoj.Com> wrote:
> > Also, how good are you at math? You seem to feel that 5 minutes of "Guido
> > the god's" time is of much more value than that of 1000 knuckle-dragging
> > Neanderthal programmers like myself spending a mere 10 minutes each to
> make
> > sure their code doesn't break. Of course, if we happen to take more than
> 10
> > minutes its our own damn fault I'm sure you'll agree.
> 
> the strange thing here is that there are changes listed on
> that page that are far more difficult to spot and fix, but for
> some reason, neither you or bjorn seem to care.

Surely you jest.  The crux of the matter here though, is that a
documented interface is being changed in backwards incompatible ways (I
haven't seen anyone claim that examples in the manual are not
documentation...)  Since there is almost no chance of changing Guido's
mind, it would just be futile to argue about e.g. multiargument append
where the manual says one thing and the implementation does another
(besides I was lucky enough not to get bit by that -- probably because I
read the manual...)

> instead, bjorn announced that he'd never upgrade to 1.6
> based on a single change in 1.6a1, despite the fact that
> there will be a 1.6a2, a3, b1, b2, rc1, etc. before it's time
> to really make that decision.

Well, to be correct, I said that this change in documented behavior
would force me to put upgrade plans on hold.  If this change remains, I
would still have to make the same decision no matter how many alpha and
beta releases we have in the next two months.

> others have politely asked for backwards compatibility on
> this one (which would mean *adding* extra code to handle
> multiple calling syntaxes).  someone will probably come up
> with a patch well before the next alpha, but probably not
> because you guys are flaming away on this newsgroup.

It would be very nice if "someone" could do that, but based on the
rethoric I've seen from you on this issue, and the absence of comment
from Guido, I certainly don't feel confident that it would be accepted. 
As far as flamage is going, you should probably go back and read through
your own posts one more time.  You might even want to read through some
of your other posts, if I remember correctly you were quite hefty in
your flamage of the poor JPython newbie from Volvo.

-b



From effbot at telia.com  Mon Apr 24 12:35:05 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Mon, 24 Apr 2000 16:35:05 GMT
Subject: ftplib sendcmd
References: <3904722D.12EE9927@bam.com>
Message-ID: <ZI_M4.1472$wYl.192602112@newsb.telia.net>

Bill Scherer <scherbi at bam.com> wrote:
> I'm having trouble with ftplib's sendcmd method (Python 1.5.2 on RH
> linux 6.1):
>
> >>> ftp = ftplib.FTP(host, user, password)
> >>> ftp.sendcmd('umask 002')
> Traceback (innermost last):
>   File "<stdin>", line 1, in ?
>   File "/usr/local/lib/python1.5/ftplib.py", line 228, in sendcmd
>     return self.getresp()
>   File "/usr/local/lib/python1.5/ftplib.py", line 201, in getresp
>     raise error_perm, resp
> ftplib.error_perm: 500 'UMASK 002': command not understood.
> >>>
>
> The umask command does, of course, work using a standard ftp client to
> the same host.
>
> Other commands, chmod, etc, suffer equal fate.
>
> Is this not what sendcmd is for? What am I doing wrong?  Thanks!

sendcmd is used to send FTP protocol commands, which
is not the same as the commands you type into your ftp
client...

to see what your client sends, try running the client in
debug mode (-d on my box.  ymmv).  or you can look it
up in the FTP specification (RFC 959).

(iirc, there is no umask command in the standard FTP
protocol, but I might be wrong...)

</F>




From effbot at telia.com  Tue Apr  4 15:15:15 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Tue, 04 Apr 2000 19:15:15 GMT
Subject: Python 1.6 alpha 1 released
References: <Pine.GSO.4.10.10004040035420.21084-100000@sundial> <5E877FD7637AB46E.DABBED9CF3C1D64B.CDDDC05AE7374A78@lp.airnews.net> <8cca93$9oh$1@nnrp1.deja.com> <B964E942A897FF5F.529887C74CF2FC7E.977238F9475EAC23@lp.airnews.net> <FdoG4.5023$74.81126@news <38EA306E.B8991085@roguewave.com>
Message-ID: <7brG4.5071$74.82114@newsc.telia.net>

Bjorn Pettersen <bjorn at roguewave.com> wrote:
> Your definition of broken seems very peculiar.  Correct me if I'm wrong,
> but it seems to go something like:

I've explained all this to you already.  lousy newsfeed? :-)

>   - you don't know if examples are binding documentation
>   - pair in the description of connect _must_ mean tuple
>     and _not_ a pair of arguments because ???

I've explained that in another post -- the library reference use
a special syntax for optional arguments.  the description of the
connect method doesn't use that syntax.  other method
descriptions do (this includes descriptions on the same page).

besides, "pair" is used quite extensively in the library reference,
usually with exactly the same meaning as here: a 2-tuple.  see,
for example, the description of "socket.error", "accept", and
"recvfrom".

>     (I would have thought it would have said tuple if it meant
>     tuple -- tuples being Python concept as opposed to pairs)

I've explained that in another post -- it tells you that the
address format is (host, port) -- which is perfectly valid tuple
syntax.

but alright, let's change "pair" to "2-tuple" everywhere.  will
fixing this make you accept the change?

>   - the example in question is broken because, even though
>     it works, and seems to follow the description from the
>     previous page, ...

I've explained that in another post -- it worked because the
connect implementation didn't handle its argument properly,
not because it was designed that way.  it stopped working
when Guido fixed the argument handling bug.

and yes, this is exactly the same reason why append, insert,
etc. did work in 1.5.2, and why they don't work in 1.6.

yet, you think there is a difference.

>   - ... other examples have been broken in the past...

I've explained that in another post -- broken examples are
obviously not formally binding.  You replied that broken
examples should be fixed.  Of course it should.  So should
all broken demo code and the broken examples, just like
all code that uses "append" etc must be fixed.

but will fixing this make you accept the change?

>
> I must say I'm confused...

yes, I think you're confused.

you would probably be less confused if you read my replies
to your own posts...

> The good news is that Guido realized how much code this would break, and
> will fix the code.

yeah, and he'll probably break it again in 1.7.

just like he'll break append and a few other things in 1.6,
and just like he broke stuff in 1.5.  and in 1.4.  etc.

you can keep on arguing until the end of c.l.py, but there's
nothing special about "connect".  except, maybe, that you
*know* that you're affected by this change, and you think
you're not affected by any of the others.

if so, that's not only a very egocentric view, it's also rather
naive.

(no, don't reply.  just read my other posts in this thread. or
if you think I don't have a clue, don't.  I don't really care)

</F>




From sverker.is at home.se  Sun Apr  2 20:05:38 2000
From: sverker.is at home.se (Sverker Nilsson)
Date: Mon, 03 Apr 2000 02:05:38 +0200
Subject: popen2 5-7 times faster (Was: Re: popen2 made >7 times faster!)
References: <38E7CEA5.4CECEE0D@home.se>
Message-ID: <38E7E051.637AC894@home.se>

Sverker Nilsson (me)wrote:
[snip]
> The time to call out 100 times to a hello world program dropped from
> 15 seconds to 2 seconds (*). 

To clarify, this was by removing some other stuff in popen2 that wasnt
needed for the simple hello world. With all that stuff in the time
became closer to 3 seconds after the fix. I think it should be
possible to get down to 2 seconds though by rewriting the rest of the
code for popen2 in C.

> The time for 100 calls to the actual
> program I wanted to run (PGP) dropped from 18 seconds to 6 seconds.

That is, 3 seconds more than the comparable hello world.

S.


From donn at oz.net  Fri Apr 14 00:01:16 2000
From: donn at oz.net (Donn Cave)
Date: 14 Apr 2000 04:01:16 GMT
Subject: 1st non-trivial program - gentle criticism requested
References: <t3pJ4.913$rc9.192648192@newsb.telia.net> <1e90r16.1sd2ts41e7ur6kN@gershwin.bio.ic.ac.uk> <8d4tgu$11qa$1@nntp6.u.washington.edu> <38F6193A.66284778@gte.com>
Message-ID: <8d656c$282$0@216.39.151.169>

Quoth "Fredrik Lundh" <effbot at telia.com>:
| Gary D. Duzan <gduzan at gte.com> wrote:
|> This potential stumbling block makes me wonder if allowing the
|> singleton case was really wise.
|
| stumbling block?
|
| what kind of programmer changes his code without thinking,
| reading what he just typed, or running the resulting program?
|
| if you want to optimize python for that kind of programmer,
| you have to change a whole lot more than the string formatting
| operator...

That sounds like me, if I understand you right.  I'm not sure
how ``reading what he just typed'' would save anyone here, but
``without thinking'' certainly seems to describe some of my work,
and ``running the resulting program'' is a frequently omitted
step if we require that the particular code in question must execute.

I don't have any suggestions for solving my problems, just wanted
to volunteer as an example of that kind of programmer.  I believe
I'm exceptional, though.

	Donn Cave, donn at u.washington.edu


From effbot at telia.com  Sun Apr  2 18:39:53 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Sun, 02 Apr 2000 22:39:53 GMT
Subject: Python 1.6 alpha 1 released
References: <200003312130.QAA04361@eric.cnri.reston.va.us> <38E5205F.DE811F61@roguewave.com> <Wf9F4.4423$74.70231@newsc.telia.net> <38E58D2D.384F1546@Lugoj.Com> <S9kF4.4472$74.71414@newsc.telia.net> <38E6B929.28816494@roguewave.com> <auGF4.914$n68.186931712@newsb.telia.net> <38E7BAC4.3B7AD0FF@uswest.net>
Message-ID: <Z_PF4.4787$74.76606@newsc.telia.net>

Bjorn Pettersen <bpetterson at uswest.net> wrote:
> > in earlier releases of the library reference, some examples
> > didn't even work.  you still think they're formally binding?
>
> Are you saying none of the examples are binding, formally or otherwise?

I have no idea -- I'm just pointing out that the library reference
has contained broken examples.  the problems I've reported have
all been fixed, but I haven't looked at all example...

> (Even in the C++ standard where examples are _explicitly_ not normative,
> they're usually fixed if conflicting with normative text.)

oh, I'm sure this will be fixed in the 1.6 documentation, in one way
or another.  currently, the example do conflict with the "normative"
text -- the method description talks about one argument:

    connect(address)

the address field is defined (in an earlier section) as either a string
or a (host, port) pair, depending on the protocol.

now, "pair" isn't formally defined ("tuple" would have been clearer),
but if you look at methods that are designed to take optional args,
you'll find that the library reference always spells that out.  as an
example, look at getsockopt a bit further down the page:

    getsockopt(level, optname[, buflen])

so in other words, either this change remains, and the example
should be fixed.  or someone adds backwards compatibility to the
socket module, in which case the method description should be
fixed:

    connect(address[, port])

/.../ If a port number is present, the first argument should be a
string containing the host name.  This is equivalent... /.../

> Your other arguments for a patch are good, but seems to reinforce the
feeling
> that changing this behavior was ill though out in the first place.

does it?  I didn't come to that conclusion for "connect" (still not
sure about "append", but that's another story).

and more importantly, I'm pretty sure the powers that be went
through the same list, decided that this is a real bug, and that
it would be far cheaper to fix this now, than at a time when
Python has twice as many users (and implementations...).

so what did we miss?

> (and remember that not all companies have enough Python expertise that a
> traceback leads to a fix ten minutes later)

well, if this is a mission critical system, you should use a staging
server, right?

if breaking the scripts would just cause a minor annoyance, most
organizations I've worked with can deal with a release note con-
taining something like:

    "if you stumble upon problems, let us know, and we'll
    fix it as soon as possible.  if you really need the old
    version, just type 'python1.5' instead of 'python'"

or you can do it the other way around:

    "we've installed a new version of the core environment.
    you can help us test this out by typing 'python1.6' instead
    of 'python'.  if we don't hear anything by the end of next
    week, we'll change the default environment"

if you don't expose the interpreter to the users, you can
rename the scripts ("just append -old"), provide an alternate
setup file, or whatever.  nothing new here.

</F>




From effbot at telia.com  Sat Apr  1 09:23:25 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Sat, 01 Apr 2000 14:23:25 GMT
Subject: new method for string objects
References: <8c44e8$5hj$1@slb6.atl.mindspring.net>
Message-ID: <xDnF4.4507$74.72019@newsc.telia.net>

Andrew Dalke <dalke at acm.org> wrote:
>   I know with 1.6 strings will have methods.  If it isn't
> already there, I would like to propose an additional method,
> "tostring()", which returns itself.
>
>   I'm working on a project which may use array.array of
> characters, Numeric.array, or a string, and I would like
> to have a way that's guaranteed to return the string
> value for the object.

assuming that "string value" means the "raw contents of the
sequence object", you should use a buffer objects instead.
you can use the 'buffer' builtin to turn most sequence objects
into buffers:

>>> s = u"123"
>>> b = buffer(s)
>>> b
<read-only buffer for 7969b0, ptr 7969d0, size 6 at 796530>
>>> len(b)
6
>>> b[0]
'1'
>>> str(b)
'1\0002\0003\000'
>>> fp.write(b) # writes raw contents to file without extra copy

(if numeric python doesn't support the buffer interface,
it should be fixed).

</F>




From echuck at mindspring.com  Thu Apr 20 08:20:59 2000
From: echuck at mindspring.com (Chuck Esterbrook)
Date: Thu, 20 Apr 2000 08:20:59 -0400
Subject: [ANNOUNCE] Webware for Python 0.1
References: <5650A1190E4FD111BC7E0000F8034D26A0F40D@huina.oceanic.com>
Message-ID: <38FEF62B.2A4402CD@mindspring.com>

Thanks for the quick report. I did some testing before the announcement, but apparently
not enough. I quickly fixed the problem last night and according to my log there have
been no exceptions since I went to bed, despite dozens of visits and downloads. 

-Chuck 


Doug Stanfield wrote:
> 
> The following intrigues me but its not probably the best endorsement.  When
> I hit your published link I get:
> 
> ERROR
> 
> Traceback (innermost last):
>   File "MainPage.py", line 55, in main
>     _main()
>   File "MainPage.py", line 49, in _main
>     inc_counter()
>   File "MainPage.py", line 23, in inc_counter
>     f.write(repr(counter+1))
> NameError: counter
> 
> But I guess you're at V0.1 so proving the error reporting works is a good
> thing. :-)
> 
> This was from a redirect to http://writewithme.com/Webware/ by the way, so
> maybe therin lies the problem.  I am still interested in looking at the
> package.
> 
> -Doug-


From alwagner at tcac.net  Wed Apr 26 23:00:34 2000
From: alwagner at tcac.net (Albert Wagner)
Date: Wed, 26 Apr 2000 22:00:34 -0500
Subject: The Simple Economics of Open Source
References: <B529350E.51B9%raffael@mediaone.net> <390534BB.3D6CEBDF@libc.org> <raffael-01CD89.20575826042000@news.ne.mediaone.net>
Message-ID: <3907AD52.FC0630FD@tcac.net>

Raffael Cavallaro wrote:
> 
<snip>
> I think that many open source advocates have
> failed to apply simple logic to the economics of open source, believing
> that somehow, the internet, or software, is immune from simple laws of
> supply and demand. They are not.
> 
I am going to have to go back and reread all of your previous posts. 
You seem to be rebutting a point that was never made.  Of course, Open
Source software is not immune to the laws of supply and demand.  Who
said it was?  It is in response to those laws that it is free.  But free
only in the world of economics where dollar worth is all that matters.
It is this very world, where the only motivation taken into account is
money profit, that Open Source has escaped.  And the escape has been
dazzling.  
---
"But beware!  The time for all this is not yet.  For a least another
hundred years we must pretend that fair is foul and foul is fair;  for
foul is useful and fair is not.  Avarice and usury and precaution must
be our gods for a little longer still.  For only they can lead us out of
the tunnel of economic necessity into daylight."
-- Lord Keynes, 1930


From dworkin at ccs.neu.edu  Sat Apr 22 21:45:47 2000
From: dworkin at ccs.neu.edu (Justin Sheehy)
Date: 22 Apr 2000 21:45:47 -0400
Subject: Overloading the = operator?
In-Reply-To: Jerome Chan's message of "Sun, 23 Apr 2000 00:14:02 GMT"
References: <eviltofu-5E48DE.20140022042000@news.earthlink.net>
Message-ID: <vnd4s8t4mk4.fsf@camelot-new.ccs.neu.edu>

Jerome Chan <eviltofu at rocketmail.com> writes:

> Is this possible?

> x = base([1,2])
> y = another([2,4,3])
> 
> z = other([])
> 
> z = x + y
> z.work()
> 
> where z now contains x.list + y.list???

Nope.  `=' rebinds the name on the left hand side, and never even
looks to see what that name was previously bound to.  

-Justin

 





From dalke at acm.org  Sun Apr 16 06:17:22 2000
From: dalke at acm.org (Andrew Dalke)
Date: Sun, 16 Apr 2000 04:17:22 -0600
Subject: Python paradigms
References: <8cmr70$b0e$1@pegasus.csx.cam.ac.uk> <8dbdj3$dgt$1@slb2.atl.mindspring.net>
Message-ID: <8dcpom$fbc$1@slb3.atl.mindspring.net>

Aahz Maruch wrote:
>def expression_if( test, true, false ):
>  if test: return true
>  else: return false
>
>x = expression_if( a is not None,
>   a[i].weeble, 0) + expression_if( b is not None, b[i].wombat, 0)


Suppose a is None.  The a[i].weeble will give an exception.

You can imagine:

def expression_if(test, true_string, false_string):
  # some nasty code to get the locals of the calling function
  if test:
    return eval(true_string with the given locals)
  else:
    return eval(false_string with the given locals

x = expression_if(a is not None, "a[i].weeble", "0") + ...

but that's just too ugly.

                    Andrew
                    dalke at acm.org





From mwh21 at cam.ac.uk  Wed Apr  5 04:25:20 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 05 Apr 2000 09:25:20 +0100
Subject: Starving for an Advanced Python Book
References: <20000405045855.26682.qmail@web2101.mail.yahoo.com>
Message-ID: <m366txnedr.fsf@atrus.jesus.cam.ac.uk>

lewst <lewst at yahoo.com> writes:

> This is off-topic, but what is with this funky way
> of signing articles?  I've now seen many people do 
> this on the Python list but am not quite getting it.

It's all Tim Peters' fault.

The first example of its use is preserved for future generations in
the Python quotations file:

  http://starship.python.net/crew/amk/quotations/python-quotes.html

which you should read anyway, it's a fine advert for this newsgroup.

Cheers,
M.

-- 
well, take it from an old hand: the only reason it would be easier
to program in C is that you can't easily express complex  problems
in C, so you don't.                 -- Erik Naggum, comp.lang.lisp


From moshez at math.huji.ac.il  Sat Apr 15 01:42:48 2000
From: moshez at math.huji.ac.il (Moshe Zadka)
Date: Sat, 15 Apr 2000 07:42:48 +0200 (IST)
Subject: Scripting and Gnome and KDE
In-Reply-To: <7t0ffs8ncrrh3irgc52rbsb68ue48rfovf@4ax.com>
Message-ID: <Pine.GSO.4.10.10004150741520.11261-100000@sundial>

On Fri, 14 Apr 2000, Alex Thomas wrote:

> This might seem an odd way to choose a GUI, but if you had the choice
> between scripting Gnome/Gtk or KDE/Qt-based components using Python, which
> would you go for? 

Personally I chose Gtk+/GNOME, since (among other things), the Gtk+ type
system is specifically targetted to support bindings to other languages.

--
Moshe Zadka <mzadka at geocities.com>. 
http://www.oreilly.com/news/prescod_0300.html
http://www.linux.org.il -- we put the penguin in .com




From mhammond at skippinet.com.au  Tue Apr  4 20:02:45 2000
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Wed, 05 Apr 2000 00:02:45 GMT
Subject: Passing array of doubles to COM object?
References: <memo.20000404125544.208A@martink.compulink.co.uk>
Message-ID: <FovG4.5061$1b4.24316@news-server.bigpond.net.au>

This definately _should_ work.  VB _may_ pass it as an array of
byref doubles tho?

Can you run makepy over the object?  If so, it definately should
work from there.

In the meantime, I will add arrays of doubles to the COM test
suite (specifically, to the VB test interfaces)

Mark.

"Martin Keefe" <martink at cix.co.uk> wrote in message
news:memo.20000404125544.208A at martink.compulink.co.uk...
>
> I am trying to use an Automation server that has the following
method (among
> others):
>
>   VARIANT_BOOL LoadFeature(
>                             BSTR code,
>                             BSTR label,
>                             long size,
>                             VARIANT xArray,
>                             VARIANT yArray);
>
> Its documentation assumes that the client will be written in VB
and says that
> xArray and yArray are arrays of Doubles. And true enough,
passing arrays of
> Doubles from VB works fine.
>
> In Python, after creating the object with:
>
>  mx = win32com.client.Dispatch("Maplex.Application")
>
> I've tried the following:
>
>  xa = [100000.0, 600000.0, 600000.0, 100000.0]
>  ya = [100000.0, 100000.0, 900000.0, 900000.0]
>  mx.LoadFeature( 'AREA', 'Simple', 4, xa, ya )
>
> and:
>
>  xa = buffer( array('d',[100000.0, 600000.0, 600000.0,
100000.0]) )
>  ya = buffer( array('d',[100000.0, 100000.0, 900000.0,
900000.0]) )
>  mx.LoadFeature( 'AREA', 'Simple', 4, xa, ya )
>
> without success. Both ran without errors or exceptions but the
data didn't
> appear to get to the server intact. I've tried both dynamic and
static dispatch
> with the same results. Can anybody shed some light on this
please?
>
> The system is NT4 Server SP3, Python 1.5.2, win32all-128.
Thanks,
>
>   ]\/[artin Keefe ][ Dorking,Surrey,UK ][ martink at cix.co.uk
>




From pinard at iro.umontreal.ca  Mon Apr  3 14:28:08 2000
From: pinard at iro.umontreal.ca (=?ISO-8859-1?Q?Fran=E7ois_Pinard?=)
Date: 03 Apr 2000 14:28:08 -0400
Subject: Unicode program representation
In-Reply-To: "Fredrik Lundh"'s message of "Mon, 03 Apr 2000 16:11:39 GMT"
References: <015c01bf9d0b$76f9b2a0$59dd3fcb@neil> <3xXF4.1060$n68.186533888@newsb.telia.net> <001201bf9d62$3e1ac1c0$36dd3fcb@neil> <%o3G4.1111$n68.191610880@newsb.telia.net>
Message-ID: <oqn1nb11l3.fsf@titan.progiciels-bpi.ca>

"Fredrik Lundh" <effbot at telia.com> writes:

> hmm.  wouldn't that mean that we end up using different encodings in
> different parts of the script?  feels a little scary, to say the least...

This is very likely to happen, but this should not be seen as scary.

Modules handling linguistic or cultural matters might come from various
sources, which sources do not necessarily use the same encodings, but yet
be used simultaneously in a program building on them all.  It is fairly
evident to me that editing strings on a screen, while seeing them nicely
displayed, far overweight the little advantages of limiting sources to
7 bits.  It is nice that Python offers a 7-bits common denominator.  Yet,
making a rule of 7-bits, or just abusing of it, might be counter-productive.

-- 
Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard





From akuchlin at mems-exchange.org  Wed Apr  5 10:12:47 2000
From: akuchlin at mems-exchange.org (Andrew M. Kuchling)
Date: 05 Apr 2000 10:12:47 -0400
Subject: Q: Status of SOAP (Simple Object Access Protocol)?
References: <38EA7EBD.BE75C460@gmx.de>
Message-ID: <3d66tw1vs0.fsf@amarok.cnri.reston.va.us>

Michael Scharf <Michael.Scharf at gmx.de> writes:
> I am wondering, what the current python-SOAP status is?
> (about SOAP http://www.develop.com/soap/)

Ken McLeod announced a minimal implementation last week; it might be
downloadable from http://casbah.org/Scarab/ by now, or you can get it
through anonymous CVS: http://casbah.org/cgi-bin/cvsweb.cgi/Scarab/python/

-- 
A.M. Kuchling			http://starship.python.net/crew/amk/
Tourist, Rincewind decided, meant "idiot".
  -- Terry Pratchett, _The Colour of Magic_



From kc5tja at garnet.armored.net  Wed Apr 26 07:57:30 2000
From: kc5tja at garnet.armored.net (Samuel A. Falvo II)
Date: 26 Apr 2000 11:57:30 GMT
Subject: <nifty hieroglyphics here>
References: <8e4cc2$d5t@netnews.hinet.net> <BWEN4.12235$HG1.329193@nnrp1.uunet.ca> <Pine.WNT.4.21.0004261229180.246-100000@nt_alester>
Message-ID: <slrn8gei3a.2um.kc5tja@garnet.armored.net>

In article <Pine.WNT.4.21.0004261229180.246-100000 at nt_alester>, Andy Lester
wrote:

>> The official language of usenet newsgroups is English,
>
>Oh?  And where's THAT defined?

If memory serves me correctly, I believe that it is a *recommendation* (not
necessarily a strict standard) that English be the `official' language of
Usenet newsgroups, in the RFCs.  I recall seeing a statement to the effect
of (paraphrased), "Newsgroups which are not in English should append their
two-character ISO country identifier to the newsgroup name." Thus, the
French-speaking comp.lang.python would appear as comp.lang.python.fr;
likewise, an Italian alt.rec.radio.amateur would be formed as
alt.rec.radio.amateur.it.

-- 
KC5TJA/6, DM13, QRP-L #1447
Samuel A. Falvo II
Oceanside, CA


From kelley at bioreason.com  Tue Apr  4 10:40:33 2000
From: kelley at bioreason.com (Brian Kelley)
Date: Tue, 04 Apr 2000 15:40:33 +0100
Subject: blessed be Guido, for "5 <= X <= 10" does what it should...
References: <3TpG4.3349$HG1.98231@nnrp1.uunet.ca>
Message-ID: <38E9FEE1.B7DB83A7@bioreason.com>

> Here's a Hack for the Mighty Pythonista to consider: Can you construct a
> program that you can call like this:
>
>     test_transitive( 0,100, '<=', 5 ) # generate a<=b<=c<=d<=e for range
> 0..100
>

A fun one here's the short version( it doesn't use the 5 limit):

import string
def test_transitive(start, end, compareString):
  return eval( string.join( map(str, range(start,end)), compareString ) )

A little obtuse, eh?

range(0,5) == [0,1,2,3,4]
map(str, range(0,5)) == ['0', '1', '2', '3', '4']
string.join(map(str,range(0,5)), ">") == "0>1>2>3>4"
eval(string.join(map(str,range(0,5)), ">")) == 1

I like python...

>
> Anyways, blessed be Guido, for near as I can figure it, he hath got it
> right! :-)
>
> Warren

--
Brian Kelley          (505) 982-7884 ext 207
Bioreason, Inc        kelley at bioreason.com
105 Washington #303, Santa Fe NM, 87501




From quinn at lira.ugcs.caltech.edu  Tue Apr 25 13:40:59 2000
From: quinn at lira.ugcs.caltech.edu (Quinn Dunkan)
Date: 25 Apr 2000 17:40:59 GMT
Subject: Don't understand error message
References: <lukN4.1574$wYl.191480320@newsb.telia.net>
Message-ID: <slrn8gbm9s.ruj.quinn@lira.ugcs.caltech.edu>

On Tue, 25 Apr 2000 17:21:21 GMT, Lars Lundstedt <lars.lundstedt at telia.com>
wrote:
>I'm halfway through "Learning Python" and I'm stuck on the example on page
>255.
>
>It goes like this:
>
>import sys, glob, operator
>print sys.argv[1:]
>sys.argv = reduce(operator.add, map(glob.glob, sys.argv))
>print sys.argv[1:]
>
>When called with "python glob.py *.py" on Win98 I'm supposed to get a list
>of all .py files in the current directory but instead I get:
>
>Traceback (innermost last):
>  File "glob.py", line 1, in ?
>    import sys, glob, operator
>  File "glob.py", line 3, in ?
>    list = reduce(operator.add, map(glob.glob, '*.py'))
>TypeError: call of non-function (type module)
>
>Anyone know why?

Eh, this is a cute one.  You named your module glob.py, so when you import
glob, it imports itself, not the glob.py from the stdlib.  So glob.glob *is* a
module: itself (so is glob.glob.glob.glob.glob.glob...).  Try renaming your
module :)


From effbot at telia.com  Sat Apr  1 14:02:04 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Sat, 01 Apr 2000 19:02:04 GMT
Subject: Why should I switch to Python?
References: <seas5ia89177@corp.supernews.com><00033120055300.03687@quadra.teleo.net> <secfidq8921@corp.supernews.com>
Message-ID: <MIrF4.4539$74.72631@newsc.telia.net>

Aaron Turner wrote:
> Honestly the more I reasearch, the more it seems that it's purely
> a matter of taste.

exactly.

> If you like Perl, there's nothing inherient about Python that makes
> it better.

works both ways, you know...

> Basically if you know both languages, and were planning a mid-sized
> project, why would you choose Python?  It's hard to take my friend
> seriously in his evaluation when he says he'd rather do this project
> in C++  than in Perl (especially when he doesn't know Perl).

well, I know several highly experienced C++ programmers who've
tried to learn Perl, and failed (quoting them would be trolling, so I
won't do that here :-).

haven't heard any such stories about Python.

so assuming that you're be doing this as a team effort, and that
the team's overall throughput matters, maybe you should respect
his opinion, and see if *you* can adapt to Python's way of doing
things?

</F>




From tom__98 at my-deja.com  Wed Apr 12 18:36:14 2000
From: tom__98 at my-deja.com (tom__98 at my-deja.com)
Date: Wed, 12 Apr 2000 22:36:14 GMT
Subject: Python/Perl Popularity (Re: A Mountain of Perl...)
References: <200004102111.HAA03380@envy.fulcrum.com.au> <8d0pcq$7ti$1@nnrp1.deja.com> <87og7fopvs.fsf@ifm.uni-kiel.de>
Message-ID: <8d2tot$k9s$1@nnrp1.deja.com>

In article <87og7fopvs.fsf at ifm.uni-kiel.de>,
  Janko Hauser <jhauser at ifm.uni-kiel.de> wrote:
> Use docindex.py!
>
> This module is definetly no substitute for perldoc but a start that
> answers your question. It gets the relevant information from the html
> tree which needs to be installed on your system.

Thanks, that's very helpful.  I'd suggest bundling something like
this with the distribution and making sure that it gets installed
(together with the documentation) by default.

I think it's really important for people to have
a simple, well-known way of getting to documentation no matter
where they are using Python or where it is installed.

> A benefit over docstrings is, that one does not need to import a
> module. This is a prototype, but shows that it is possible. Not that
> this will change your mind about Python :-)

Well, I don't think my mind needs to be changed about Python; I
have been using it since 1996 and already like the language.

Tom.


Sent via Deja.com http://www.deja.com/
Before you buy.


From tim_one at email.msn.com  Sun Apr 23 23:29:52 2000
From: tim_one at email.msn.com (Tim Peters)
Date: Sun, 23 Apr 2000 23:29:52 -0400
Subject: 'database' search
In-Reply-To: <39036340.BE82BEAD@roguewave.com>
Message-ID: <000001bfad9d$5a129060$4a2d153f@tim>

[Bjorn Pettersen, wants to do word completion]
> ...
> The smallest datastructure I've found so far, is a 'prefix dictionary',
> so if e.g. "aardwark" was in the database, the 'prefix dictionary' would
> contain:
>
>  pfdict = {
>   'a':        ['aardwark', ...],
>   'aa':       ['aardwark', ...],
>   'aar':      ['aardwark'],
>   'aard':     ['aardwark'],
>   'aardw':    ['aardwark'],
>   'aardwa':   ['aardwark'],
>   'aardwar':  ['aardwark'],
>   'aardwark': ['aardwark']
>  }
>
> The maximum size for this would be n * avg len of item, with the real
> size being much(?) smaller with sufficient prefix sharing...

Before going nuts with variations on prefix trees, try the attached.  This
uses a sorted list of all words, and does a binary search to find the ones
starting with a given prefix -- dirt simple.  The advantage to that is that
it works <wink>.

BTW, the amount of compression you can get from a tree (or trie) approach
depends on the language and the set of words, but when coding in Python
you're almost certainly going to consume more storage with all the
supporting hair (ancillary dicts and lists) than you'll save by sharing a
few initial bytes in the words.  Trees are used for speed reasons instead,
in intense searching applications (which yours is not).

> I have a relatively stable database of between 1500 and 5000 items

So even if they average 10 letters/item, this won't consume more than 50Kb.

> new insertions are probably going to be less than 100 during the
> lifetime of the database)

In 1.5.2 and 1.6, appending a new word to a sorted list and then sorting
again takes time linear in the # of words -- it's recognized as a special
case (of another special case <wink>), and doesn't do a full sort.  So the
method below is more efficient than you might think.

when-your-database-grows-100x-larger-ask-again-ly y'rs  - tim


from bisect import bisect

class Completer:
    def __init__(self, wordlist):
        """wordlist -> a Completer for the list of words."""
        wordlist.sort()
        self.words = wordlist

    def complete(self, prefix):
        """prefix -> alphabetical list of all words starting w/ prefix."""
        words = self.words
        # Find i, j s.t. all in words[i:j] start w/ prefix.
        i = bisect(words, prefix)
        # Subtle:  if prefix is in the list, i points *after* it.
        # Fiddle i to point before it.  The bisect module should
        # really have more functions to control what happens here.
        while i > 0 and words[i-1] == prefix:
            i = i-1
        j = i
        n = len(words)
        # note that .startswith() is a 1.6 feature
        while j < n and words[j].startswith(prefix):
            j = j+1
        return words[i:j]

    def addword(self, word):
        """Add a word to the list of words."""
        self.words.append(word)
        self.words.sort()





From r-stadhe at online.no  Mon Apr 17 07:55:00 2000
From: r-stadhe at online.no (RCS)
Date: Mon, 17 Apr 2000 13:55:00 +0200
Subject: Tkinter Look
References: <8d4qkq$l7s$1@nnrp1.deja.com>
Message-ID: <N8DK4.4400$By1.92248@news1.online.no>

Fantastic!
You have just saved me hours (days) of work!
If I come up with something that could add to this package, I will let you
know.

RCS

<cer at tutopia.com> wrote in message news:8d4qkq$l7s$1 at nnrp1.deja.com...
> I want to share this work related with a Look&Feel package
> for Tkinter.
>
> For info and download see :
> http://www.geocities.com/cristian_echeverria/Chevegui.html
>
> This has been very usefull to my and maybe somebody can do
> a better job with it.
>
> Bye.......Cristian Echeverria
>
>
> Sent via Deja.com http://www.deja.com/
> Before you buy.




From sadwolf at chollian.net  Tue Apr 11 02:02:28 2000
From: sadwolf at chollian.net (Dong-gweon Oh)
Date: Tue, 11 Apr 2000 15:02:28 +0900
Subject: Thread-safe atomic operation
Message-ID: <8cuf0o$p9a$1@news1.kornet.net>

Is there any documents which describe Python's atomic operation in
multithreaded environment. I guess Python virtual instruction is atomic.
But, without definitive documentation in high level Python statements, I
tend to over-lock my programs.

Are these statements thread-safe?
1. import xxx
2. xxx = __import__('xxx')
3. reload(xxx)
4. dict['xxx'] = sys.modules['xxx']
5. dict.has_key('xxx')

Thanks,

Dong-gweon Oh






From mwh21 at cam.ac.uk  Fri Apr 28 12:48:37 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 28 Apr 2000 17:48:37 +0100
Subject: parser and tokenize modules: sample code?
References: <3909B42D.2E564923@uab.edu>
Message-ID: <m3n1menpcq.fsf@atrus.jesus.cam.ac.uk>

Shae Erisson <shapr at uab.edu> writes:

> Can someone point me to sample code for the parser and tokenize modules?
> As Moshe Zadka says, the parser module doesn't seem easy to use. I've
> read through the example.py for the parser module, but I'd like to see
> more code if someone has some.

The file "transformer.py" from p2c takes the output of parser and
produces something easier to work with (allegedly; I've never used
it).  Start with one of these links:

http://lima.mudlib.org/~rassilon/p2c/
http://www.pythonpros.com/cgi-bin/viewcvs.cgi/p2c/

HTH,
Michael

-- 
31. Simplicity does not precede complexity, but follows it. 
     -- Alan Perlis, http://www.cs.yale.edu/~perlis-alan/quotes.html


From richard_chamberlain at ntlworld.com  Fri Apr 28 11:57:11 2000
From: richard_chamberlain at ntlworld.com (Richard Chamberlain)
Date: Fri, 28 Apr 2000 16:57:11 +0100
Subject: file open() problem
References: <8ebcm3$lo$1@nnrp1.deja.com>
Message-ID: <8uiO4.1393$E44.27433@news6-win.server.ntlworld.com>

Just to let you know that I'm running win98se and it works fine.

Richard
<ekw1 at my-deja.com> wrote in message news:8ebcm3$lo$1 at nnrp1.deja.com...
> I am using python1.5.2 on windows98se.  all of the following gives me
> an error.
>
> f = open('file.txt', 'r')
> f = open("file.txt", "r")
> f = open('c:/dir/file.txt', 'r')
> f = open('c:\dir\file.txt', 'r')
> f = open("c:/dir/file.txt", "r")
> f = open("c:\dir\file.txt", "r")
> f = open("c:\\dir\\file.txt", "r")
>
> The error from the python shell for any of the above is:
>
> Traceback (innermost last):
>   File "<pyshell#17>", line 1, in ?
>     f = open('file.txt', 'r')
> TypeError: illegal argument type for built-in operation
>
> Can anyone help me?  Thanks.
>
> ekw
>
>
>
>
> Sent via Deja.com http://www.deja.com/
> Before you buy.




From Michael.Scharf at gmx.de  Fri Apr 14 17:10:37 2000
From: Michael.Scharf at gmx.de (Michael Scharf)
Date: Fri, 14 Apr 2000 23:10:37 +0200
Subject: [ANNOUNCE] Garbage collection for Python
References: <20000407221750.A29279@acs.ucalgary.ca> <m3r9cagx3c.fsf@atrus.jesus.cam.ac.uk> <PRKJ4.1078$rc9.191689728@newsb.telia.net>
Message-ID: <38F7894D.9997636A@gmx.de>

Fredrik Lundh wrote:
> 
> Michael Hudson <mwh21 at cam.ac.uk> wrote:
> > It would be, in my opinion, be a really really really good thing; for
> > one thing it would help in language wars...
> 
> sorry, but stackless python is the only really really really good
> thing that hasn't yet been added to the core.  this is just really
> really good.

And what are the reasons for not adding the really really really good
stackless python??


Michael
-- 
     ''''\     Michael Scharf
    ` c-@@     TakeFive Software GmbH, a Wind River Company
    `    >     http://www.WindRiver.com
     \_ V      mailto:Michael.Scharf at gmx.de


From mhammond at skippinet.com.au  Wed Apr  5 20:25:59 2000
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Thu, 06 Apr 2000 00:25:59 GMT
Subject: Remotely running a python script
References: <8cf452$pdv$1@gxsn.com> <7GOG4.2280$bh.22442@typhoon.nyroc.rr.com>
Message-ID: <rQQG4.6094$1b4.29122@news-server.bigpond.net.au>

"Darrell" <darrell at dorb.com> wrote in message
news:7GOG4.2280$bh.22442 at typhoon.nyroc.rr.com...

> If you have Mark Hammond's Win32 stuff installed try naming
your script with
> .pyw
> It'll run nicely out of site without a console.

Actually, you dont need the win32 extensions for this to work -
Python itself installs this magic quite independently of
win32all.

Mark.





From mjhand at concentric.net  Fri Apr 14 01:57:42 2000
From: mjhand at concentric.net (Manus Hand)
Date: 14 Apr 2000 01:57:42 EDT
Subject: Python Starship Grounded??
Message-ID: <38F6B2DF.940E7C23@concentric.net>

Anyone here able to do anything about the illness that seems to have
come over starship.python.net?

It was out of commission all day Tuesday, and now all day today
(Thursday) without a visible light at the end of the tunnel.

It occasionally answers ping (well, it did earlier today, but....)
and sometimes a traceroute will crawl through to it, but http and
ftp and ssh, etc., etc. have failed to answer all day.  (And were
sluggish at best yesterday, and non-existent the day before.)

Manus


From Klaus-Gerd.Meyer at de.bosch.com  Tue Apr 18 11:13:21 2000
From: Klaus-Gerd.Meyer at de.bosch.com (Klaus-Gerd Meyer)
Date: Tue, 18 Apr 2000 17:13:21 +0200
Subject: Pointers
References: <38D00DCE.9740380F@be-research.ucsd.edu> <8bfsqo$kje$1@tribune.oar.net> <dp4s9wl9df.fsf@mraz.iskon.hr> <uPYC4.55946$3b6.224410@ozemail.com.au> <4k5D4.7566$0o4.54442@iad-read.news.verio.net> <38FB98F1.239EC0FB@whetstonelogic.com> <slrn8foe80.bd3.scarblac-spamtrap@flits104-37.flits.rug.nl>
Message-ID: <8dhu2j$1gm$1@proxy2.fe.internet.bosch.de>

>Try playing around in the interpreter for a while, this is a key thing to
>understand about Python in my opinion. You can test objects to see if
>they're the same object with 'x is y', and test their values with 'x == y'.


A good idea. ;-)
Also nice: id(x), returns an objekt id (currently implemented as its
address).

Integers and strings with same value are identical, but tuples and list with
same value are not identical...

>>> x = 1
>>> y = 1
>>> id(x),id(y)
(7879224, 7879224)
>>> x is y            #x and y: same value and identical
1
>>>
>>> x = "a"
>>> y = "a"
>>> id(x),id(y)
(8041104, 8041104)
>>> x is y            #x and y: same value and identical
1
>>> x = ("a","b")
>>> y = ("a","b")
>>> id(x),id(y)
(8864080, 8919376)
>>> x is y            #x and y: same value and not identical
0
>>> x = ["a"]
>>> y = ["a"]
>>> id(x),id(y)
(9378096, 9394512)
>>> x is y            #x and y: same value and not identical
0






From chafik at nevrax.com  Wed Apr  5 04:31:12 2000
From: chafik at nevrax.com (Sameh chafik pro)
Date: Wed, 5 Apr 2000 10:31:12 +0200
Subject: Some problem with python 1.6
Message-ID: <001201bf9ed9$4cd6ca60$1001a8c0@nevrax.net>

I work with Visual C++ (6.0). When i build a project who have a python support i have the following error:
1) The library Python16_d.lid dos not  existe.
        I renamed in the config.h the pragama comment (lib,"python16_d.lib") to pragama comment (lib,"python16.lib"). But
        now i have the follow link error:
2) Linking...
        Creating library Debug/PYTHONLIB.lib and object Debug/PYTHONLIB.exp
    PyLib.obj : error LNK2001: unresolved external symbol __imp___Py_Dealloc
    PyLib.obj : error LNK2001: unresolved external symbol __imp___Py_RefTotal
    PyLib.obj : error LNK2001: unresolved external symbol __imp__Py_InitModule4TraceRefs
    PythonMethode.obj : error LNK2001: unresolved external symbol __imp__Py_InitModule4TraceRefs
    
    This link error happen only with the debug version, not that the difference between the debug and release comme from the         _DEBUG preprocessor.
    
.....................Thanks for answer.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20000405/4b131766/attachment.html>

From zorro at zipzap.ch  Wed Apr 12 09:24:36 2000
From: zorro at zipzap.ch (Boris Borcic)
Date: Wed, 12 Apr 2000 15:24:36 +0200
Subject: Python/Perl Popularity (Re: A Mountain of Perl...)
References: <200004102111.HAA03380@envy.fulcrum.com.au> <8d0pcq$7ti$1@nnrp1.deja.com> <g0ssm03o.fsf@dan.resnet.rochester.edu> <8d1lgu$ga0$1@newshost.accu.uu.nl>
Message-ID: <38F47914.5F13F5C5@zipzap.ch>

Martijn Faassen wrote:

> Quick-someone-start-a-math-discussion!-ly yours,

I can propose my april's fool contribution that
was refused by sci.physics.research's moderation

--

1. Monstrous Moonshine Over Giordano Bruno

Giordano Bruno was burnt at the stake at the
Piaza dei Fiori on February 17th, 1600, in Rome, for
heresy.

On the 400th anniversary of the event, a few weeks
ago, the catholic church acknowledged that its
method of defending "the Truth" had not been
up to the standards that Jesus promoted. It
nevertheless reasserted that Bruno was *wrong*,
an opinion that is in practical agreement with that
of many of Bruno's post-mortem defenders, starting
with Diderot, who presented Bruno in his encyclopaedia
as an example of Church victim, but depicted his work
as obscure and not really interesting.

Bruno's specialty was mnemotechnics. Here it will
be shown that mnemotechnical methods, such as those
employed by Bruno, are actually sufficient to defend
his memory without need for the help of half-hearted
opportunistic defenders.

1.1 Truly Visual Basic

The first step in the method, in today's words, is
called "Truly Visual Basic". Shortly put, the method
consists in representing letters, names, objects,
beings, facts, using arithmetical or mathematical
symbols that best mimic them in an iconic manner.

So we start with the stake to which Bruno was bound,
and represent it with the digit that best represents
it : a 1.

Then we represent Giordano Bruno himself, by his
initials, G and B, and in turn represent these using
the digits that best mimic them : 6 and 3, respectively.

This takes a bit of training to *see* : there is no
digit that mimics a capital G better than does 6.
And none that mimics a B, better than does 3. (it takes
a bit of effort, but the reader should be reassured, though,
that Truly Visual Basic (tm) is much easier to use
than is its namesake by Mr Bill Gates).

The stake and Giordano Bruno together give us : 163.

They were bound together, which we represent by
surrounding the above expression with parenthesis : (163).

Witnesses say that a priest presented Bruno with
a cross, while he was turning his head away : T(163).

This was taking place on a square, which we mimic
with the abreviation SQR. So we get to : SQRT(163).

The square, being a former horseracing stadium, is
and was actually round. This contradicts the standard
expectations stemming from the noun "square". So
we rectify this expectation by (a) marking the above
expression with an asterisk : SQRT(163)*, (b) giving
archimedes constant, the natural symbol for all that's
round, PI, as an explanation for the asterisk : *PI,
and (c) gluing both together for economy : SQRT(163)*PI.

This last expression is thus a representation for the
square and the public in it. These were spellbound to
the scene, just as GB was himself bound to the stake :
(SQRT(163)*PI).

Bruno was then burnt in *expiation* his purported
blasphemies, just as we today, are concerned with having
the church *expiate* for what it did to him. We represent
this by (a) reordering the expression in conformance
to mathematical standards : (PI*SQRT(163)), and
(b) prepending it with EXP, which gives us :

[formula I] EXP(PI*SQRT(163)) !

[...]

1.3 The Monster Group and its 196884-dimensional Space

The above is the title of chapter 29 of the famous
Book by Conway and Sloane, "Sphere packings, Lattices,
and Groups", which I cite as a shortcut to get back to
Giordano Bruno after the above digressions.

Recall formula [I] that we derived, using Truly Visual
Basic (tm), from the sorry fate of Giordano Bruno in the
hands of the catholic church. We shall show below that
the above cited telltale number of the famous Monster
Group's Moonshine, e.g. 196884, lurks behind that formula
(a fact that many know already, I am sure). We urge the
reader, first, to note that there is something moral to see
the Monster Group occur in the trace of a victim of
religious dogmatism such as Mr. Giordano Bruno. Indeed,
the Monster Group may rightfully be called the central
room of the castle of exceptional algebraic structures,
and religious dogmatism is nothing else than generalizing
rules above any measures, thus denying exceptions.

Computing formula [I] EXP(PI*SQRT(163)), using an
infinite precision package, gives us the strange value :

v := 262537412640768743.999999999999250072597198185688879353856...

Dogmatics, seeing such a value, will undoubtetly conclude
to a computer arithmetic error, and round the number to
262537412640768744. Actually, there is no error, and since
we despise dogmatics - if only because we are currently
encouraging the memory of one of their victims - we decide
to turn our attention on what they would discard : the
fractional part of the number v, most naturally expressed
as the difference to the rounded number.

w := ceil(v)-v = 000000000000749927402801814311120646143662663...

This isn't very handy, but groping around, we find out
that :

v*w := 196883.999999999918130677895382415018280732177224...

Which, having already proved that we are not dogmatic
about it, we feel no shame to round to the integer, (noting
also that it starts with 1968, a most undogmatic year) :

196884 - to discover the telltale sign of Monstrous
Moonshine over Giordano Bruno.

1.4 The Holy Trinity and Mary's Virgin Birth

Giordano Bruno's condamnation was motivated by his rejection
of a variety of dogmas, most notably the Holy Trinity
and Mary's Virgin Birth. (The unknowing reader should
be made aware that, contrary to a popular misconception,
the dogma of Mary's Virgin Birth refers not to Jesus,
but to the notion that Mary herself was born free of the
sequel of the original sin that Eve and Adam commited.

What proves the power of GB's mnemotechnics, is that if
we take the 196884 above, that we have obtained from the
picture of his torture, EXP(PI*SQRT(163)), multiply it
by 3 as a reference to the dogma of the Holy Trinity,
and substract the result, as a number of days, from the
date of Giordano Bruno's execution, we obtain a date
that is likely to be that of Mary's Birth, some 17
years before our era.

[...]


From aa8vb at yahoo.com  Thu Apr 20 10:07:06 2000
From: aa8vb at yahoo.com (Randall Hopper)
Date: Thu, 20 Apr 2000 10:07:06 -0400
Subject: how do i pack and unpack messages in python?
In-Reply-To: <008301bfaacb$42fe2c60$2801000a@iel.ie>
References: <008301bfaacb$42fe2c60$2801000a@iel.ie>
Message-ID: <20000420100706.A624577@vislab.epa.gov>

Shaun:
 |so i now have a message that looks like:
 |
 |id                       : 10
 |subid                  :13
 |NV sequence :   {14, "shaun"-string
 |                         {20, ["Anytown, Anywhere"]-list
 |                         {34, 1234567890-int
 |
 |
 |i want to pack the message to look like this:
 |
 |10 fs 13 fs 14 fs s shaun fs 20 fs l s anytown fs s anywhere fs i 1234567890
 |gs gs
 |
 |(NV=Name Value Pair, fs=field seperator, s=string, i=int, l=list, group
 |seperator)
 |
 |can anyone please tell me how id go about manipulating the strings so as i
 |can join them using field seperators, any ideas anyone, or even tell me
 |where i might find out how to do this problem.

I was going to mention the struct, array, or xdrlib modules to you, but
from your desired output format (all ASCII it appears), it looks like
a simple string.join would do the job:

   >>> import string
   >>> string.join( [ str(10), str(13), str(14), str("shaun") ], ";" )
   '10;13;14;shaun'

   etc.

-- 
Randall Hopper
aa8vb at yahoo.com



From mhammond at skippinet.com.au  Tue Apr 18 20:58:26 2000
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Wed, 19 Apr 2000 00:58:26 GMT
Subject: PythonWin and Dialogs in DLLs
References: <38FCD4FB.6E07A632@callware.com> <38fcde9b.12797461@news-server.socal.rr.com>
Message-ID: <Sw7L4.5283$1D.31197@news-server.bigpond.net.au>

"Roger Burnham" <rburnham at cri-inc.com> wrote in message
news:38fcde9b.12797461 at news-server.socal.rr.com...
> On Tue, 18 Apr 2000 15:34:51 -0600, Ivan Van Laningham
> <ivanlan at callware.com> wrote:
>
> In your pywin.mfc.dialog.Dialog derived class, call the parent init
> with the dll:
>
>            Dialog.__init__(self, dllid='pathToYourDll')

Or, in some module simply say:

dll = win32ui.LoadLibrary('pathToYourDll')

And magically your DLL should be searched for resources.  Just use the
existing dialog functions as normal, but you can specify resource IDs from
your DLL.

Mark.





From josht at iname.com  Thu Apr  6 21:58:51 2000
From: josht at iname.com (Josh Tompkins)
Date: Fri, 07 Apr 2000 01:58:51 GMT
Subject: Python on Google?
Message-ID: <8F0ED22EBjoshtinamecom@209.208.0.7>

Buenos Dias, everyone.

This is perhaps an off-topic question, but having just found google, I'm 
curious...

Does anyone know to what extent Python is used on www.google.com?  I 
noticed on thier "job offerings" (or whatever) page that they list python 
experience as "a plus".

Just wondering...

Josue

________________________________________________________________

"Destined For Great Things -- but pacing myself."
- From a t-shirt.

E-Mail:  josht at iname.com 
ICQ:  21219667
AIM:  JosueTheGreat
Web:  http://www.crosswinds.net/~josht
_________________________________________________________________


From rcc at nospamthanks_jennic.com  Thu Apr 20 10:16:24 2000
From: rcc at nospamthanks_jennic.com (Robert Cragie)
Date: Thu, 20 Apr 2000 14:16:24 GMT
Subject: Using popen in windows
References: <1255890154-29764960@hypernet.com>
Message-ID: <newscache$y7jbtf$od$1@jenpc07.jennic.co.uk>

Gordon McMillan <gmcm at hypernet.com> wrote in message
news:1255890154-29764960 at hypernet.com...
| Robert Cragie wrote:
|
| > I tried the following in Windows 98 as an experiment:
| >
| > from win32pipe import popen
| > p = popen('dir','r')
| > str = p.read()
| > print str
| >
| > It just hangs when trying to do the read, and I have to kill the shell
off.
| > I have tried IDLE 0.5 and PythonWin and the 'DOS' command line, both do
the
| > same. I am using Python 1.5.2, and the win32pipe.pyd DLL is dated 3rd
May
| > 1999.
| >
| > I have done the equivalent on Linux and it works fine:
| >
| > from os import popen
| > p = popen('ls','r')
| > str = p.read()
| > print str
| >
| > Anyone got any ideas?
|
| Ah, but it's not equivalent :-). 'dir' is a cmd.exe builtin, not an
| "external program". You can't (directly) popen any of the
| cmd.exe builtins (and the list changes with each OS release).

I tried this and it still hangs; cvs.exe is definitely not a built in:

from win32pipe import popen
p = popen('c:\\cygwin\\contrib\\bin\\cvs -v','r')
str = p.read()
print str

Again, the equivalent works fine on Linux. So how do I run a command in
Windows and capture the output? The following works, but is clumsy, and you
get the annoying DOS box while it runs - is there a better way?:

import os
os.system('dir >temp')
f = open('temp', 'r')
f.read()

TIA

Robert




From scarblac-spamtrap at pino.selwerd.nl  Wed Apr 26 09:17:35 2000
From: scarblac-spamtrap at pino.selwerd.nl (Remco Gerlich)
Date: 26 Apr 2000 13:17:35 GMT
Subject: Need advice performing a search
References: <3905F795.A536CFA7@stsci.edu> <3906944F.F718FE91@teleatlas.com> <slrn8gdbak.bp5.scarblac-spamtrap@flits104-37.flits.rug.nl> <oq8zy1ypqu.fsf@titan.progiciels-bpi.ca>
Message-ID: <slrn8gdr2l.c28.scarblac-spamtrap@flits104-37.flits.rug.nl>

Fran?ois Pinard wrote in comp.lang.python:
> scarblac-spamtrap at pino.selwerd.nl (Remco Gerlich) writes:
> 
> > from string import find
> > def testline(line):
> >   return find(line, "first_string") or find(line, "second_string")
> 
> Shouldn't we have:
> 
>    return find(line, "first_string") >= 0 or find(line, "second_string") >= 0

<sigh>. Yes.

-- 
Remco Gerlich,  scarblac at pino.selwerd.nl

   This is no way to be
     Man ought to be free      -- Ted Bundy
       That man should be me


From jimc at regnoc.com  Sun Apr  2 22:41:08 2000
From: jimc at regnoc.com (Jim Conger)
Date: Mon, 03 Apr 2000 02:41:08 GMT
Subject: Just doing local file testing
References: <38E796BC.23191B2@regnoc.com> <00040214141502.04535@quadra.teleo.net> <38E7D6DE.EA09C891@regnoc.com> <00040216504403.04535@quadra.teleo.net>
Message-ID: <38E7F616.48CF9284@regnoc.com>

An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20000403/dd5be326/attachment.html>

From jamarijr at hotmail.com  Tue Apr 18 09:05:49 2000
From: jamarijr at hotmail.com (Arinté)
Date: Tue, 18 Apr 2000 13:05:49 GMT
Subject: Still trying to learn
Message-ID: <8dhmj2$f9m$1@nnrp1.deja.com>

I have this code here

	def IOCTL(self, comm, args):
		command1 = long(0)
		try:
			if type(comm)==type(2):
				poss.alert("Totally")
		except:
			command1 = comm
			print sys.exc_type
			print sys.exc_value
			return -1
		if type(comm)==type(long(2)):
			print "IOCTL was int"
			command1 = long(comm)
			print "IOCTL Set to Long"
		else:
			if type(comm) == type("  "):
				print "IOCTL was a string"
				command1 = comm
			else:
				if type(comm)!=type(2L):
					print "IOCTL fail"
					return -1
		poss.IOCtl(self.devname,command1,args)
		return 0

The poss object is going to my c\c++ app and I have traced thru it and
it seems to pass back valid values everytime.  The problem I am having
is on the 2nd call to this function the try..except fails with this
error:

exceptions.TypeError
len() of unsized object

Here is how I make the call
for i in range(10000):
	res = nex.IOCTL("POS_SYS_ACQUIRE_DEVICE",0)
	if res <0:
		break
	else:
		poss.alert("OK")

Thanx and sorry for the long post.




Sent via Deja.com http://www.deja.com/
Before you buy.


From michael_peck at my-deja.com  Thu Apr 13 14:29:12 2000
From: michael_peck at my-deja.com (Michael Peck)
Date: Thu, 13 Apr 2000 18:29:12 GMT
Subject: Embedded interp --> call back to embeddor?
References: <1256474779-52175680@hypernet.com>
Message-ID: <8d53l1$45$1@nnrp1.deja.com>

In article <1256474779-52175680 at hypernet.com>,
  gmcm at hypernet.com wrote:
> Michael Peck  wrote:

> There's a (moronically) simple example at
> http://www.mcmillan-inc.com/embed2.html

Wow. That is exactly what I was looking for. My undying gratitude,
Gordon.

MJP


Sent via Deja.com http://www.deja.com/
Before you buy.


From johann at physics.berkeley.edu  Thu Apr  6 14:37:18 2000
From: johann at physics.berkeley.edu (Johann Hibschman)
Date: 06 Apr 2000 11:37:18 -0700
Subject: Beginning to be disapointed
References: <Aw4H4.8078$pR.68745@carnaval.risq.qc.ca>
Message-ID: <mt3dozt6sh.fsf@astron.berkeley.edu>

Hugues Demers writes:

> Can somebody help me on this? I'm using python to read big arrays
> (100x2000) of float. I manipulate those arrays, I do small
> operations on them. That is ok. I do others smalls calculations not
> involving the big arrays and then the interpreter hang ! The cpu
> percentage used by python goes up and there is nothing else to be
> done except to kill the interpreter. I have no idea what's going
> on. Do I need to free memory in some way ? Am I doing something
> wrong ? There is nothing special about the calculations involving
> the big arrays. The only thing special is that they are big.

Are you using the Numeric package?  I've never seen behaviour like that.

1. are you using Numeric or array?
2. how are you reading them in?
3. how are you calculating them?  Numeric math ops or explicit loops over
   raw arrays?
4. could something be keeping a reference to the arrays?
5. what OS are you on?
6. version of python?

Sorry, I think we all need a bit more information to answer you.

--Johann

-- 
Johann Hibschman                           johann at physics.berkeley.edu


From gmcm at hypernet.com  Thu Apr 13 13:41:54 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Thu, 13 Apr 2000 13:41:54 -0400
Subject: Embedded interp --> call back to embeddor?
In-Reply-To: <8d4vkv$r9p$1@nnrp1.deja.com>
Message-ID: <1256474779-52175680@hypernet.com>

Michael Peck  wrote:

> I can't tell from any of the references I've read whether my objective
> is possible. What I want to do is provide a C program that embeds a
> Python interpreter in such a way that the executed Python code can call
> back to C functions within my program.
> 
> Is this possible? Can someone provide information on the procedure?

There's a (moronically) simple example at
http://www.mcmillan-inc.com/embed2.html


- Gordon



From effbot at telia.com  Thu Apr 13 14:00:57 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Thu, 13 Apr 2000 18:00:57 GMT
Subject: Doesn't anybody write to paper anymore?
References: <38F5FDF3.EE2FBA01@tcac.net>
Message-ID: <tXnJ4.1780$Za1.27183@newsc.telia.net>

Albert Wagner <alwagner at tcac.net> wrote:
> I've searched the docs and my "Learning Python" book and I can't find
> anything on printing to a real printer.  Through trial and error I
> finally made the following work:
>
> printer = open('/dev/lp0', 'w')
> printer.write("page 1\n\r\f")
> printer.write("page 2\n\r\f")
> printer.close

oh, you're on unix...

this might work:

    printer = os.popen("lpr", "w")
    printer.write("page 1\n\f")
    printer.close()

for more information, read the OS documentation for
lpr and friends.

</F>




From thomas.heller at ion-tof.com  Thu Apr 13 13:00:42 2000
From: thomas.heller at ion-tof.com (Thomas Heller)
Date: Thu, 13 Apr 2000 19:00:42 +0200
Subject: threading question
Message-ID: <018e01bfa569$cda0cd80$4500a8c0@thomasnotebook>

I want a thread to run in parallel with the interactive
interpreter.
Simple:

def do_work():
    import time
    while 1:
        time.sleep (1)
        ...

import threading
threading.Thread (target=do_work()).start()

But when I hit EOF, python waits until this thread is finished.
How can I avoid this? I want python to terminate the thread
and exit in this case.

Thomas Heller





From phd at phd.russ.ru  Mon Apr 24 08:16:23 2000
From: phd at phd.russ.ru (Oleg Broytmann)
Date: Mon, 24 Apr 2000 12:16:23 +0000 (GMT)
Subject: Trouble compiling PyGreSQL 2.4 on Linux
In-Reply-To: <k8hnhfh0.fsf@jumbo.skott.dk>
Message-ID: <Pine.LNX.4.21.0004241211380.6019-100000@fep132.fep.ru>

On Mon, 24 Apr 2000, Martin Sk?tt wrote:
> >    There is no Makefile 'cause the module is intended to use standard
> > Python Makefile.pre.in/Setup.in scheme.
> 
> That is onlu if you chose the standard Python module installation. Red
> Hat has made something special where modules are stored in a directory
> called lib-dyn/ some where. To me it seems like the modules then are
> loaded automaticly.

   I think even on RH you may do it "standard way" - install python-dev
RPM, and you've got Makefile.pre.in and Setup.in with them. Copy these
files to you work directory, modify Setup.in to your needs (to PyGreSQL
needs this time :) and make -f Makefile.pre.in && make.
   I didn't test my advice - I run only Debian systems and one Solaris; and
in any case I installed Python from sources - I need some more modules
(locale, most notable) and some less (no tkinter, please :)

> As I mention in my first posting the readme.linux file describes how
> to do an red hat install (make redhat) but it doesn't seem to work. An
> address is also mentioned for an RPM but the file doesn't exist on the server.

   D'Arcy would gladly accept any patch.

Oleg.            (All opinions are mine and not of my employer)
---- 
    Oleg Broytmann      Foundation for Effective Policies      phd at phd.russ.ru
           Programmers don't die, they just GOSUB without RETURN.




From tuttledon at hotmail.com  Tue Apr  4 08:43:06 2000
From: tuttledon at hotmail.com (Don Tuttle)
Date: Tue, 04 Apr 2000 12:43:06 GMT
Subject: Starving for an Advanced Python Book
References: <20000404040941.7766.qmail@web2106.mail.yahoo.com>
Message-ID: <urlG4.2557$d13.25939@typhoon.southeast.rr.com>

"lewst"
>ORA's "Programming Python" is a natural choice, but I
> can't bear buying this outdated book when the 2nd edition is due out
> in a few months (I'm dying here Lutz, hurry it up already!).

I went ahead and bought it because Oreilly's cheif editor says, "We're
timing the new edition of Programming Python to follow shortly after the
release of Python 2.0."(aka Python 3K).  If that turns out to be the case,
it won't published until sometime in 2001.  See his complete (and somewhat
dated) remarks at  http://www.oreilly.com/frank/python.html

> Does anyone have any suggestions?  I suppose I'm looking for a text
> that has examples of how to solve various kinds of programming
> problems in Python and also examples of code showing how the library
> and various modules actually work.

Fredrick Lundh's "(the Eff-bot Guide To) The Standard Python Library"
 http://www1.mightywords.com/asp/bookinfo/bookinfo.asp?theisbn=EB00002582
has many short examples.

Also "The Python Annotated Achives" is pretty much just examples and
comments.
http://shop.barnesandnoble.com/booksearch/isbnInquiry.asp?userid=589TPW6Z1E&
mscssid=7T95UX3412S92K8S0017QRP4NV5DCFDD&srefer=&isbn=0072121041

>How is the "Python Essential
> Reference" on examples?  Is it just a rebinding of the python.org
> library reference

Excelent reference, but no examples.

Don





From just at letterror.com  Fri Apr 14 18:06:06 2000
From: just at letterror.com (Just van Rossum)
Date: Fri, 14 Apr 2000 23:06:06 +0100
Subject: [ANNOUNCE] Garbage collection for Python
In-Reply-To: <PRKJ4.1078$rc9.191689728@newsb.telia.net>
References: <20000407221750.A29279@acs.ucalgary.ca>
 <m3r9cagx3c.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <l03102809b51d41895a80@[193.78.237.135]>

Michael Hudson <mwh21 at cam.ac.uk> wrote:
> To someone who might know: what hope is there for this to go
> into 1.6?
[ ... ]
> It would be, in my opinion, be a really really really good thing; for
> one thing it would help in language wars...

Fredrik Lundh wrote:
> sorry, but stackless python is the only really really really good
> thing that hasn't yet been added to the core.  this is just really
> really good.

Oh how I agree... (Now lets hope the new Python GC isn't
incompatible with Stackless... Project anyone? ;-)

While were at it, here's a little Stackless news:
- Christian is redesigning the continuation module, it already has
  improved quite a bit over the last two weeks.
- Even so, Chris and I agree that continuations should eventually
  be completely hidden from the user, since it's so incredibly
  obscure, hardly anyone will understand it. Not good for CP4E ;-)
- However, it's currently needed so microthreads and coroutines
  can be largely implemented in Python itself, which in itself is
  extremely cool.
- We're pretty far in fleshing out the microthreading needs;
  the microthread kernel will at some point be implemented in C,
  making it pretty darn efficient. (Although the current Python
  version is actually not so bad itself...)
- Dunno about coroutines: it appears they are much harder to get
  right, and the people on the python-coro at egroups.com list only
  seem to make slow progress. (But then again, I've only been
  there for a week or so, so who am I to say that...)
- Will Ware and I have worked hard on a complete rewrite of the
  uthread module. It is a *vast* improvement over the old one.
  It offers a very nice thread object and several handy threading
  widgets. It already can do stuff threading.py can only *dream*
  about, like pausing (blocking) a thread from another thread or
  posting an exception from another thread (!).
- It seems our new version of uthread.py is close to being
  system-thread safe, allowing each system thread to have its
  own set of microthreads...
- I'm playing with a uthreaded socket wrapper which uses select
  and non-blocking sockets under the hood, without the client
  even knowing it. If it turns out this works well, uthreads will be
  a good alternative for asyncore.py, being easier to use and all.

Just





From tuttledon at hotmail.com  Sat Apr  1 22:32:50 2000
From: tuttledon at hotmail.com (Don Tuttle)
Date: Sun, 02 Apr 2000 03:32:50 GMT
Subject: New Features in Python 1.6
References: <000301bf9c15$23a18f20$182d153f@tim> <vkuF4.4584$74.73299@newsc.telia.net>
Message-ID: <CbzF4.9197$784.144089@typhoon.southeast.rr.com>

Ha Ha Ha!....shit fire and McDonalds Fries!....Oh, my side hurts!...Ha
Ha....oops, just wet myself....gotta go

"Fredrik Lundh"
> Tim,
>
> I think things are worse than this -- someone just sent me
> something that looks like a draft press release, announcing
> that O'Reilly had taken over Python development, with Jon
> Orwant as CTO, and Larry Wall as the Python 1.6 release
> manager.
>
> (it wasn't clear if it was intended for public consumption
> at this time, so please don't spread the word around just
> yet.  if I get permission, I'll post more info to the list).
>
> given this, Guido's post makes a lot more sense, in a slightly
> scary way -- the press release also talks about fusing Python
> 3000 and Perl's "Topaz" project, rewriting the whole thing in
> Scheme.
>
> note that this seems to have been in the work for quite a
> while; things like Perl companies hiring python talent, Tom
> Christiansen's Python courses [1], Guido schmoozing with
> Randal Schwartz on the SD2000 conference [2], the Casbah
> folks working on a unified Python/Perl engine, etc.
>
> still think it's kinda weird, though...
>
> </F>
>
> 1) http://training.perl.com/open-courses.html
> 2) http://www.ddj.com/articles/2000/0050/0050toc.htm
>
>




From karab at stsci.edu  Tue Apr 25 15:58:38 2000
From: karab at stsci.edu (Kia Arab)
Date: Tue, 25 Apr 2000 19:58:38 GMT
Subject: Need advice performing a search
Message-ID: <3905F795.A536CFA7@stsci.edu>

I am trying to search a file for two strings. If *either* string is
found, I need the line in which the string was found to be returned  (or
somehow extract the line). It should keep searching and keep returning
the lines where it finds either string, until the end of file is
reached.

Any suggestions?


From alwagner at tcac.net  Wed Apr 12 22:27:39 2000
From: alwagner at tcac.net (Albert Wagner)
Date: Wed, 12 Apr 2000 21:27:39 -0500
Subject: redirecting stdout to printer
Message-ID: <38F5309B.487B228B@tcac.net>

How do I specify my printer as stdout?  Know of a module that has basic
prettyprinting with headings and pagenumbers?

-- 
Small is Beautiful


From josht at iname.com  Thu Apr  6 22:05:26 2000
From: josht at iname.com (Josh Tompkins)
Date: Fri, 07 Apr 2000 02:05:26 GMT
Subject: Searching for Python Material...
Message-ID: <8F0ED18C1joshtinamecom@209.208.0.7>

Buenos Dias, everyone.

I'm fairly new to this whole python thing.  I think that I'm getting better 
at it, but I seem to be suffering from a lack of reference material.  None 
of the bookstores in my area carry any of the python books, and being 
woefully without a credit card, I must wait until the stores special order 
my requests...

So, anyway, I was wondering if anyone could point me to some websites that 
cover material that is more advanced than the stuff in most of the 
tutorials out there.  I'm actually looking for something along the list of:

to do (something cool in python) do this...

where something cool in python is a variable with values such as "Accessing 
webpages and/or e-mail", "Connecting to a database", etc, while still 
keeping the deep magic stuff to a relative minimum.

Thanks for any help you can give me!

Josue

________________________________________________________________

"Destined For Great Things -- but pacing myself."
- From a t-shirt.

E-Mail:  josht at iname.com 
ICQ:  21219667
AIM:  JosueTheGreat
Web:  http://www.crosswinds.net/~josht
_________________________________________________________________


From grant at nowhere.  Fri Apr 14 14:43:56 2000
From: grant at nowhere. (Grant Edwards)
Date: Fri, 14 Apr 2000 18:43:56 GMT
Subject: Authenticate with smtp
References: <018701bfa53b$f831f3e0$858647c1@DEVELOPMENT>
Message-ID: <slrn8fepnb.8ds.grant@grante.comtrol.com>

In article <018701bfa53b$f831f3e0$858647c1 at DEVELOPMENT>, Ulf Engstr?m wrote:

>How can I log into an smtp server that requires authentication and send a
>mail with smtplib, or is there any other package that helps? (Yes, I'm
>authorized)
>Regards
>Ulf

A friendly hint:  Your posting was 50+ lines long with 5 lines
of actual content.  Delete all the MIME and vcard stuff.

About SMTP authentication: what type of authentication is
required?  Some servers (MS-Exchange) do plaintext username and
password authentication.  Others require you to have an active
IMAP or POP connection.  Others have a more secure username
password scheme.

I've got C code that does the MS-Exchange username/password
authentication scheme (there may be other similar ones), and
you're welcome to that:   

 <ftp://ftp.visi.com/users/grante/stuff/ssmtp-2.33-auth.tar.gz>

-- 
Grant Edwards                   grante             Yow!  I'd like TRAINED
                                  at               SEALS and a CONVERTIBLE on
                               visi.com            my doorstep by NOON!!


From gregm at iname.com  Mon Apr 10 09:18:43 2000
From: gregm at iname.com (Greg McFarlane)
Date: Mon, 10 Apr 2000 23:18:43 +1000
Subject: Python Megawidgets - Pmw.Blt problem
In-Reply-To: <38EFC4DF.76F3C698@pacbell.net>; from Bruce Wolk on 08 Apr 2000 at 04:46:39PM
References: <38EFC4DF.76F3C698@pacbell.net>
Message-ID: <20000410231843.05124@nms.otc.telstra.com.au>

Yes.

You don't have Blt installed on your machine.  You will need to go to
the BLT home page at
    http://www.tcltk.com/blt/
and download it and install it into your Tcl.

On 8 Apr, Bruce Wolk wrote:
> Has anyone been able to use the Pmw.Blt megawidget on a Win98 system? 
> Pmw generally seems to work ok, but when I try the examples at
> 
> http://www.ifi.uio.no/~hpl/Pmw.Blt/doc/
> 
> they don't run.  I get errors like
> 
> Traceback (innermost last):
>   File "C:\PROGRA~1\PYTHON\TOOLS\IDLE\ScriptBinding.py", line 131, in
> run_module_event
>     execfile(filename, mod.__dict__)
>   File "C:\Program Files\Python\Bruce\HelloWorld1.py", line 5, in ?
>     g = Pmw.Blt.Graph(master)       # make a new graph area
>   File "C:\PROGRA~1\PYTHON\Pmw\Pmw_0_~1\lib\PmwBlt.py", line 260, in
> __init__
>     Tkinter.Widget.__init__(self, master, _graphCommand, cnf, kw)
>   File "C:\Program Files\Python\Lib\lib-tk\Tkinter.py", line 1084, in
> __init__
>     self.tk.call(
> TclError: invalid command name "::blt::graph"
> 
> Any ideas?
> 
> Bruce
> -- 
> http://www.python.org/mailman/listinfo/python-list
> 

-- 
Greg McFarlane     INMS Telstra Australia     gregm at iname.com



From fw at deneb.cygnus.argh.org  Tue Apr 11 08:12:32 2000
From: fw at deneb.cygnus.argh.org (Florian Weimer)
Date: 11 Apr 2000 14:12:32 +0200
Subject: [ANNOUNCE] Garbage collection for Python
References: <20000407221750.A29279@acs.ucalgary.ca> <38F13FB9.D1D16AF@mindspring.com> <slrn8f2if1.dop.nascheme@cranky.arctrix.com> <87r9cdhdb0.fsf@deneb.cygnus.argh.org> <m3bt3hkk76.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <87bt3geszz.fsf@deneb.cygnus.argh.org>

Michael Hudson <mwh21 at cam.ac.uk> writes:

> Florian Weimer <fw at deneb.cygnus.argh.org> writes:
> 
> > nascheme at enme.ucalgary.ca (Neil Schemenauer) writes:
> > 
> > > I would rather get the bugs stomped out of my patch and get it
> > > included in 1.6 or 1.7.  We can then remove the FAQ entry.  :)
> > 
> > Is it possible to extend your GC in a way which would permit the
> > execution of several bytecode interpreters in multiple threads?
> 
> Eh?

With the reference counting scheme, you can't do that, because you
have to synchronize access to the reference count, which would result
in a major performance penalty.


From christian.bau at isltd.insignia.com  Mon Apr 10 05:56:18 2000
From: christian.bau at isltd.insignia.com (Christian Bau)
Date: Mon, 10 Apr 2000 10:56:18 +0100
Subject: Why should I switch to Python? - Infinity of Primes
References: <NDBBKEGCNONMNKGDINPFGEJDDFAA.infonuovo@email.com> <8ccoug$f4o$1@pegasus.csx.cam.ac.uk> <7F51E902E09F3F60.8F612607812ED16B.82CB8B41435E14FD@lp.airnews.net> <38EA369C.26A1C25@math.okstate.edu> <38EAC8C4.ADC25640@cosc.canterbury.ac.nz>
Message-ID: <christian.bau-1004001056180001@christian-mac.isltd.insignia.com>

In article <38EAC8C4.ADC25640 at cosc.canterbury.ac.nz>, Greg Ewing
<greg at cosc.canterbury.ac.nz> wrote:

> "David C. Ullrich" wrote:
> > 
> >     There's nothing non-constructive about the traditional
> > proof of the infinitude of the sequence of primes - given
> > a sequence of primes it _constructs_ a prime not on the
> > list.
> 
> Um, no it doesn't - it constructs a number which is
> *either* prime *or* divisible by some other prime bigger
> than the one you started with.
> 
> If someone actually came up with a formula for constructing
> primes, it would be rather large news -- isn't that one of
> the Big Unsolved Problems?

Take any non-empty set S of prime numbers, calculate their product, add
one, and take the smallest factor that is not equal to one. That factor is
a prime number that is not in the set S.


From phil at geog.ubc.ca  Sat Apr 29 00:17:37 2000
From: phil at geog.ubc.ca (Phil Austin)
Date: 28 Apr 2000 21:17:37 -0700
Subject: Reading in strings -> numbers ??
References: <280420001700308281%pecora@anvil.nrl.navy.mil>
Message-ID: <m3g0s5v8v2.fsf@brant.geog.ubc.ca>

"Louis M. Pecora" <pecora at anvil.nrl.navy.mil> writes:

> Is there a Python "read-partner" for the write(string) function, e.g., 
> 
> fileobject.write(" %e %e %d" % (1.2, -3.5, 10))

You might want to check out Konrad Hinsen's  FortanFormat package
in ScientificPython:

http://starship.python.net/crew/hinsen/scientific.html

from FortranFormat import *
input=open('lemout1.dat','r')
format=FortranFormat('(f15.6,1x,f20.6)')
x=[]
c=[]

line=1
while(line):
    line=input.readline()
    getit=FortranLine(line,format)
    x.append(getit[0])
    c.append(getit[1])

Regards, Phil


From jim at interet.com  Wed Apr 12 10:47:09 2000
From: jim at interet.com (James C. Ahlstrom)
Date: Wed, 12 Apr 2000 10:47:09 -0400
Subject: Python 4 DOS
References: <sf7hsua14nk34@corp.supernews.com>
Message-ID: <38F48C6D.FD671082@interet.com>

Arnaldo Riquelme wrote:
> 
> Can anyone tell me if I can get python binaries for DOS (16-bit).
> I heard of  Hans Nowak 's  Python-DX but the links to his website is broken?
> Where are you Hans?

www.python.org has a DOS Python running with a DOS extender.

Jim


From grant at nowhere.  Fri Apr 14 13:53:30 2000
From: grant at nowhere. (Grant Edwards)
Date: Fri, 14 Apr 2000 17:53:30 GMT
Subject: comparing strings and ints
References: <2YmI4.366$rc9.188014080@newsb.telia.net> <20000411134358.A2352304@vislab.epa.gov> <1256641282-42160637@hypernet.com> <20000412075812.A2449004@vislab.epa.gov> <jr%I4.643$rc9.190628352@newsb.telia.net> <20000412112935.A2446939@vislab.epa.gov> <slrn8f9r9j.in3.grant@grante.comtrol.com> <m3u2h6hmcn.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <slrn8femop.89j.grant@grante.comtrol.com>

In article <m3u2h6hmcn.fsf at atrus.jesus.cam.ac.uk>, Michael Hudson wrote:
>grant at nowhere. (Grant Edwards) writes:
>
>> No, it said "_consistently_ but arbitrarily" [emph added].
>> That means that you don't which result you will get when a
>> string and an int are compared, but you know you will always
>> get the _same_ result.
>
>That's one reading; however I'd have thought a more likely one would
>be to specify that if x>y is true once for some x and y of differing
>types, then x>y is true each subsequent time it is evaluated.

That's a reasonable (and more conservative/safer)
interpretation.

>can-I-have-a-language-lawyer-brownie-point?-ly y'rs

Sure, take one out of stock.

-- 
Grant Edwards                   grante             Yow!  .. I'll make you
                                  at               an ASHTRAY!!
                               visi.com            


From me at nospam.edu  Fri Apr 28 16:18:56 2000
From: me at nospam.edu (Me)
Date: Fri, 28 Apr 2000 20:18:56 GMT
Subject: Help Needed-IDLE (Linux)Upgrading
References: <3909CC26.D684AD92@idirect.com>
Message-ID: <QmmO4.21$kM.5530@news>

It is there! I did the same thing...either do a search for "python" or
"idle" or look in (forgive me I am not on my box at the moment) /usr/share
????  IT IS THERE THOUGH! I would move the old (just in case) and then copy
it in. I am new too.

Bob

"Snoopy :-))" <genius at idirect.com> wrote in message
news:3909CC26.D684AD92 at idirect.com...
> No Flames Please. I am a NEWBIE :-))
>
> Using Linux (RH-6.1) I have downloaded the Idle-0.5 Upgrade "tgz".
> package.But I don't know where and how to install the package.
>
> After unzipping the package I've read the README
> file which doesn't seem togive clear instructions regarding installation
> of the upgrade. On my system Idle is in /usr/bin
> but this is only the executable file without subdirectories where I
> would be able to put the untared and uzippedfiles.
>
> I would appreciate a helping hand
> Best regards
> Charles Takacs




From burnout at heathers.stdio.com  Tue Apr 18 14:17:01 2000
From: burnout at heathers.stdio.com (Eric H.)
Date: Tue, 18 Apr 2000 14:17:01 -0400
Subject: Implementing a Python process when PC is idle?
References: <38fc9694@wwwproxy3.westgroup.com>
Message-ID: <38FCA69D.85CC27DA@heathers.stdio.com>

I don't know if this will help but a fellow who helps with the
setiathome project created a program that will run processes once cpu
load gets below a certain user set point. The binaries he has are for
SGI and Sun but he also has source that might help.. The page to find
this program is..

http://www.pattosoft.com.au/jason/idleize/

I use it on my Indigo2 and love it..I can control anything like this..I
think I can even use some of my raytracers with this..haven't tried yet
though...prolly will be next.. Hope this helps...and L8a


MikeZ wrote:
> 
> Hello,
> 
>     I would like to execute a Python process that gets executed when my
> Windows PC is in screensaver mode (so my PC can work when it is idle).
> 
>     Has anyone ever implemented this, or have an idea how this could be done
> ? I prefer to do this all in Python, including any Windows commands
> necessary to register the process as a screensaver.
> 
> Thanks in advance,
> -Mike Z.

-- 
Parts of this message, including the email address, ? 2000-Eric Haddix. 
Use implies agreement to the terms and conditions made available on my 
homepage. Keep the Internet free, refuse to support the providers that
copyright YOUR email and messages. All rights reserved. U.C.C. 1-207

visit my homepage @ http://www.bounce.to/EyesCremeProductions



From dan at cgsoftware.com  Wed Apr 12 00:09:31 2000
From: dan at cgsoftware.com (Daniel Berlin+list.python)
Date: 12 Apr 2000 00:09:31 -0400
Subject: Python/Perl Popularity (Re: A Mountain of Perl...)
In-Reply-To: tom__98@my-deja.com's message of "Wed, 12 Apr 2000 03:09:26 GMT"
References: <200004102111.HAA03380@envy.fulcrum.com.au> <8d0pcq$7ti$1@nnrp1.deja.com>
Message-ID: <g0ssm03o.fsf@dan.resnet.rochester.edu>

>>>>> "tom" == tom 98 <tom__98 at my-deja.com> writes:


    tom> In article <200004102111.HAA03380 at envy.fulcrum.com.au>,
    tom> Richard.Jones at fulcrum.com.au wrote:

    tom> Maybe you can contribute more constructively:

    tom>  -- How can I browse the Python documentation without firing
    tom> up a web browser?

Convert it to text format. It's latex, you can make it into anything
you want, and do anything you want with it.
If you tell me your environment for programming, i'll tell you a nice
way to be able to get at it that doesn't involve firing up a web browser
    tom> Or,   How do I get information on a single
    tom> Python function or class quickly?
Personally, i read the docstrings.
I have a little utility function in my python startup file.
SO all i do is thus:
import socket 
>>> info(socket.htonl) 
Documentation: 
------------------------------------------------------------------------

Object <built-in function htonl> : 
    htonl(integer) -> integer  

    Convert a 32-bit integer from host to network byte order.  
 
 

    tom> Is there any equivalent of
    tom> "perldoc -f split" or "perldoc Net::POP3"?  What do you do in
    tom> that situation?

You fire up the interperter, which you probably already have running,
and just look at the docstring.
If they didn't provide one, well, you yell at them.


    tom>  -- Is there a Python equivalent of CPAN and the Perl CPAN
    tom> module?  The closest I know of is Parnassus.  But Parnassus
    tom> is merely a collection of links, not an archive, and it
    tom> doesn't have any facilities (AFAIK) for automatic
    tom> installation.

So, who gives a crap?
I try not to be that lazy.

    tom>  -- Several extensions I have wanted to install have required
    tom> access to the full Python source tree, not just the header
    tom> files and the library. 
Such as?
    tom>  Is that going to change?  Otherwise,
    tom> how can I install Python extensions that want access to the
    tom> source tree on, say, a standard RedHat system?
I have yet to see a single one that required access to the full source
tree.
If it doesn't have a damn good reason, the package is broken.
    tom> Having the
    tom> standard RPM-based installation in parallel with a /usr/local
    tom> installation from source is confusing, and removing the
    tom> RPM-based installation risks breaking things.


Why do i get the feeling you'll never be satisfied?

    tom> Tom.




From cut_me_out at hotmail.com  Wed Apr 26 11:42:03 2000
From: cut_me_out at hotmail.com (Alex)
Date: 26 Apr 2000 11:42:03 -0400
Subject: How can I get the name of an object???
References: <8e712k$65c$1@nnrp1.deja.com>
Message-ID: <etd8zy03m44.fsf@w20-575-108.mit.edu>

An object could be referred to by any number of names:

dd = zz ()
ee = dd

Which name for the object do you wish to know?  ee or dd?

Alex.


From warlock at eskimo.com  Tue Apr  4 22:08:39 2000
From: warlock at eskimo.com (Jim Richardson)
Date: Tue, 04 Apr 2000 19:08:39 -0700 (PDT)
Subject: opening a serial port at a given baudrate. How?
In-Reply-To: <200004041153.VAA15869@piglet.dstc.edu.au>
Message-ID: <XFMail.20000404190839.warlock@eskimo.com>

On 04-Apr-2000 David Arnold wrote:
> -->"Jim" == Jim Richardson <warlock at eskimo.com> writes:
> 
>   Jim> I can't seem to find any info on opening a serial port in
>   Jim> python!. (I am using linux for developement, but would like to
>   Jim> be able to use the app on any python system, it is text only,
>   Jim> no gui, at least for now. )
> 
> there is no portable way to open a serial port.  this is not least
> because there is no portable way to name your serial ports, but goes
> further than that.
> 
> on unix (including linux), just open the device special file
> corresponding to the serial port of interest (ie. /dev/ttyS0) and you
> can then control it using fcntl.ioctl() function and otherwise read
> from and write to it like a normal file object.
>

I have been using open('/dev/ttyS0','r+')
but need to control baud rate. fcntl.ioctl() will do this? 
 I can't seem to find (on machine, haven't checked on-line yet
how to do this, but thanks for the pointer! Knowing where to look is at
least half the battle.



> on win32 there is a serial I/O module (serialmodule?) that a quick
> deja search will locate for you.
> 
> i'd recommend the POSIX serial programming guide (any search engine
> should find it) as a good primer for Unix serial programming ...
> 

OK, will do, thanks again.

Jim Richardson
        Anarchist, pagan and proud of it
WWW.eskimo.com/~warlock
        Linux, because life's too short for a buggy OS.




From jmassung at magpiesystems.com  Thu Apr 27 17:53:40 2000
From: jmassung at magpiesystems.com (Jeff Massung)
Date: Thu, 27 Apr 2000 15:53:40 -0600
Subject: Simple Modules Question
References: <8eaaul$16p@autodesk.autodesk.com>
Message-ID: <0A2O4.502$3j.5393@news.uswest.net>

Akira Kiyomiya wrote in message <8eaaul$16p at autodesk.autodesk.com>...

>from string import atoi    # imports 'string'
>                                     # places atoi in current namespace
>print atoi("12345")         # invokes 'atoi' without module name
>string.atoi("45")             # NameError: string

You didn't import "string", therefore "string" is not in namespace.

import string
print string.atoi("45") # prints 45
atoi("45") # NameError: atoi

A way around this:

import string
from string import *

Now you can use either string.atoi() or atoi() - but this is redundant.

Jeff




From bpetterson at uswest.net  Sun Apr  2 19:25:03 2000
From: bpetterson at uswest.net (Bjorn Pettersen)
Date: Sun, 02 Apr 2000 17:25:03 -0600
Subject: Python 1.6 alpha 1 released
References: <200003312130.QAA04361@eric.cnri.reston.va.us> <38E52716.A1C63B5D@roguewave.com> <rJ9F4.549$n68.187227648@newsb.telia.net> <38E52D76.1820FC32@roguewave.com> <8c7tfd$k80$1@nntp9.atl.mindspring.net> <5gLF4.970$n68.190385152@newsb.telia.net>
Message-ID: <38E7D6CF.C6723ECD@uswest.net>

Fredrik Lundh wrote:

> Aahz Maruch wrote:
> > I'm sympathetic.  We only upgraded our production system from 1.5.1 to
> > 1.5.2 last week, because we'd made a number of 1.5.1-specific hacks and
> > had to back them out correctly for 1.5.2.
>
> we upgraded a customer site from 1.3 to 1.5.2 in november
> (distributed, mission critical, many gigabytes per day, custom
> interpreters, lots of C and C++ extensions, etc).
>
> catching up with changes in Python and its standard libraries
> was a very small part of the total effort.  getting it all to work
> on a new version of the operating system was much harder...

Very good points -- to upgrade a system there is a certain amount of
planning, and you have to carefully implement the result of your planning to
minimize any unavoidable downtime.  For some organizations this might not be
required, but I can't think of a comercial one where it isn't.  This takes
real time and costs real money, and sooner or later someone will have to do a
cost/benefit analysis of whether the cost of upgrading to Python 1.6 is worth
the benefits.  In my situation the cost would currently outweigh the
benefits, so I will hold off on transitioning until that changes.  For anyone
working in a software company, this should hardly be surprising (?)

-b





From mikael at isy.liu.se  Wed Apr 12 03:49:30 2000
From: mikael at isy.liu.se (Mikael Olofsson)
Date: Wed, 12 Apr 2000 09:49:30 +0200 (MET DST)
Subject: Huh? func_defaults, default values in function calls
In-Reply-To: <38F3EBBF.7B23@marshall.edu>
Message-ID: <XFMail.000412094930.mikael@isy.liu.se>

On 12-Apr-00 Mark Hathaway wrote:
 >  This way you'll never really know what the default is going to be.
 >  Ha ha ha ha. that's pathetic.

Actually, it's a feature. It makes it possible to have a local variable
that survives from one call of the function to another.

/Mikael

-----------------------------------------------------------------------
E-Mail:  Mikael Olofsson <mikael at isy.liu.se>
WWW:     http://www.dtr.isy.liu.se/dtr/staff/mikael
Phone:   +46 - (0)13 - 28 1343
Telefax: +46 - (0)13 - 28 1339
Date:    12-Apr-00
Time:    09:46:57

This message was sent by XF-Mail.
-----------------------------------------------------------------------


From warlock at eskimo.com  Tue Apr  4 02:10:40 2000
From: warlock at eskimo.com (Jim Richardson)
Date: Tue, 04 Apr 2000 06:10:40 GMT
Subject: opening a serial port at a given baudrate. How?
Message-ID: <slrn8ej283.b3t.warlock@raven.myth>

I am looking at using python for a simple app to control an embedded system
for testing, the system is controlled via a standard 3 wire serial port, 
at 9600baud, 8-N-1. The problem is, I can't seem to find any info on opening
a serial port in python!. (I am using linux for developement, but would
like to be able to use the app on any python system, it is text only, no
gui, at least for now. ) Am I just being dense? I mean there has got to be 
info on this out there, so please, flame away if you must, but point me 
to where to find it please...

(crouches behind hopefully flame-proof shield... :)

-- 
Jim Richardson
	Anarchist, pagan and proud of it
WWW.eskimo.com/~warlock
	Linux, because life's too short for a buggy OS.



From genehilton at my-deja.com  Fri Apr 21 15:11:47 2000
From: genehilton at my-deja.com (genehilton at my-deja.com)
Date: Fri, 21 Apr 2000 19:11:47 GMT
Subject: accessing BLT vectors from C
Message-ID: <8dq957$tc8$1@nnrp1.deja.com>

Has anyone out there written anything that will enable me to access
the BLT C api.  I am using BLT as a part of a data acquisition
system.  It works great, but filling the vectors is really slow in
python.  I wrote a python module in C to use the Blt c api, but it can't
find the vectors I created with Pmw.Blt.Vector().  I can access (in the
c module) vectors that I create in the c-module.  This must be some sort
of TCL thing I don't understand.  Here is a snippet from the C module


        if (!PyArg_ParseTuple(args, "OO",&master,&bltobj)) return NULL;

/* get the name of the vector - PMW stores it as the repr. */
        bltobjname=PyObject_Str(bltobj);
        if (!bltobjname) return NULL;
        bltname=PyString_AsString(bltobjname);
        if (!bltname) return NULL;
        printf("%s\n",bltname);
/* get the interpreter from an argument (for now) */
         interp=((TkappObject *)master)->interp;

/* this doesn't work (returns a 0) */
	printf("%d\n",Blt_VectorExists(interp,bltname));
        printf("%d\n",Blt_CreateVector(interp,"myvec",1,&vec));
/* this works (returns a 1) */
        printf("%d\n",Blt_VectorExists(interp,"myvec"));

I call it with the following python code

import Tkinter
import Pmw
import mymod
v=Pmw.Blt.Vector(1)
mymod.myroutine(master.tk,v)
v=Pmw.Blt.Vector(1)

I get an output like
0     (couldn't find it - v created in python)
0     (created ok in c)
1     (found the one created in c)

Any pointers would be real helpful

Gene Hilton
hilton at boulder.nist.gov


Sent via Deja.com http://www.deja.com/
Before you buy.


From wrd at awenet.com  Thu Apr 27 19:17:41 2000
From: wrd at awenet.com (William R. Dickson)
Date: Thu, 27 Apr 2000 16:17:41 -0700
Subject: Still problems running script from cron
Message-ID: <wrd-500C44.16174127042000@news.awenet.com>

Hi all,

I posted a while ago about a script that runs fine from the command 
line, but hangs when run from cron.  I've made progress, but am not all 
the way there yet.  Many people suggested helpful ways to redirect 
stdout into files, but they had no effect.  I eventually managed to get 
past the point where it was hanging by changing:

os.system("%s -c %s -n %s -o %s %s" % (webalizer, conffile, self.server, 
outputdir, self.currlog))

to:

rv = os.system("%s -c %s -n %s -o %s %s" % (webalizer, conffile, 
self.server, outputdir, self.currlog))

It then stopped hanging there.  Now it hangs when my forks are supposed 
to close, here:


children = 1

for server in servers:
  child = os.fork()
  if child == 0:            # Fork returns 0 in the child; if this is now
    break                   # a child, we want out of the loop.
  else:
    children = children+1   # Otherwise, we check to see if we're already
    if children > C:        # at the max number of processes; if not, we
      rv = os.wait()        # go to the top of the loop and fork another.
      children = children-1 # Otherwise, we wait until a fork closes.



The end of the child process looks like this:


# When both server stats and user stats are completed, we can finally get
# rid of the daily log and close the fork.
    
os.remove(nicelog)
      
print ("+ %s  %s %s: No problems detected; closing fork."
       % (strftime(T, localtime(time())), procdate, id))

sys.exit()  # I've also tried using return here with no change


When run from cron, the message "No problems detected; closing fork." 
prints as expected, but the child process never closes, according to ps.  
When run from the command line, it closes as expected.

Any idea why this might happen?

Thanks,

-Bill


From haering at informatik.tu-muenchen.de  Thu Apr 27 20:38:26 2000
From: haering at informatik.tu-muenchen.de (haering at informatik.tu-muenchen.de)
Date: Fri, 28 Apr 2000 00:38:26 +0000
Subject: HTML parser example, anybody?
In-Reply-To: <3908909D.4022A846@roguewave.com>
References: <slrn8gh2a9.oj.haering@sunhalle8.informatik.tu-muenchen.de> <3908909D.4022A846@roguewave.com>
Message-ID: <20000427223825Z12160-466+9398@hpsystem14.informatik.tu-muenchen.de>

In comp.lang.python, you wrote:
>Why don't you just use Python's string syntax, e.g. if you have the
>following in a template file called foo.templ:
>
>   <html><head><title>%(title)s</title></head>
>   <body><h1>%(heading)s</h1></body></html>
>
>you can do something like:
>
>   htmlfile = open('foo.html', 'w')
>   replacements = {
>     'title':   'my title',
>     'heading': 'A Heading',
>   }
>   htmlfile.write( open(foo.templ).read() % replacements )
>   htmlfile.close()
>
>-- bjorn
>

Thanks for the info, but I'd rather transform HTML -> HTML. My current
project uses special comments of the form <!-- at INS keyword param1=value1
param2=value2 -->. I'd like to parse for comments and filter out + process these. The
prime reason for a real parser is that I need to rewrite all links in the
HTML source tree and do some other checks.

Gerhard



From jstok at bluedog.apana.org.au  Tue Apr 18 12:08:33 2000
From: jstok at bluedog.apana.org.au (Jason Stokes)
Date: Wed, 19 Apr 2000 02:08:33 +1000
Subject: Editors and books
References: <XBFK4.16$Uq2.1831@news.uswest.net>
Message-ID: <nP%K4.7905$5D.17539@ozemail.com.au>

Jeff Massung wrote in message ...
>First, I'd like to give my undying praise to PYTHON: Essential Reference,
by
>David M. Beazley. I'm an accomplished C/C++ programmer who wanted to learn
>Python. This book was IT! Anyone whose reading this that is in my shoes -
>get this book! It's some of the best $35 I've spent!


I like the book a lot too.  Sure, a lot of it is duplicated in the online
material, but it's concise, better presented, and perfect for the
accomplished programmer who can learn other languages fast.

>Also, on another NG there's talk about Python, and how it does (and still
>doesn't) do garbage collecting. Could someone here expand on it so I know
>what it can and can't do? Memory is a big issue in my project and I need to
>know how to structure my program so that garbage is collected. Thanks
>again!!


Python uses a reference counting scheme to collect unused objects.  Each
object stores a count of how many references refer to it.  When this drops
to zero, the object is deallocated.  This scheme has the usual drawbacks of
reference counting -- if there's a data structure whose object reference
graph involves cycles (eg a doubly linked list) the reference counts of
those objects will never reach zero because they refer to each other, even
if the base of the list becomes unreachable.

To deal with this you need to create specific destruction routines for your
data structures that run through and break these references, eg by setting
them to null.

This is annoying, but since you're used to C++ you're used to it (it's the
same scheme used by reference counting smart pointers.)  There's no
language-level reason why Python can't have true garbage collection, and in
fact true garbage collection for Python has recently been implemented and
with luck we'll see it become a standard part of Python implementations.


>-- Jeff Massung (jmassung at magpiesystems.com)
>-- Lead programmer for Magpie Systems
>-- Bringing smart pigs to the pipeline industry!

Oh wow!  I bet that's more interesting than working on inventory databases
or visual basic scripts....




From dgobbi at minime.irus.rri.on.ca  Thu Apr 20 19:49:31 2000
From: dgobbi at minime.irus.rri.on.ca (David Gobbi)
Date: 20 Apr 2000 23:49:31 GMT
Subject: Python and VTK and TK
References: <38F51AB1.29FB4C1D@earthlink.net>
Message-ID: <8do52b$ina$1@panther.uwo.ca>

Scott Harris <scottrharris at earthlink.net> wrote:
> Has anyone gotten VTK to work with the TkRenderWidget?
> I've got a simple code that draws two buttons and a vtkTkRenderWidget in
> a frame.
> It works under win98, but crashes my XServer on Linux.

Hi Scott,

I've used the vtkTkRenderWidget with XFree-3.9.18 and there weren't
any problems -- not with hardware acceleration though.  In order to
make it work, you have to either copy it into /usr/lib or setenv 
the VTK_TK_WIDGET_PATH environment variable to <wherever>/vtk/python.
Oh, and you must also configure VTK using --with-shared.

I have had problems with vtk and accelerated glx under XFree-3.3.*,
though... the glx doesn't define glIsTexture() and calls to this 
function will crash the server.  This happens whenever VTK tries
to release a texture.  The 'fix' is simple but inelegant -- comment
out the glIsTexture() line in vtkOpenGLTexture.cxx. 

If you get wxPython's widget working under Linux I'd like to hear
about it.  We're doing all our stuff in Tk, but we're certainly
not satisfied with Tk as a GUI toolkit.  We do 95% of our interaction
through picking and dragging, though, so the toolkit we use doesn't
matter all that much.

 - David

--
  David Gobbi, MSc                    dgobbi at irus.rri.on.ca
  Advanced Imaging Research Group
  Robarts Research Institute, University of Western Ontario

> I'm using RH6.1 with XFree864.0 and a Voodoo3 card (so I have
> accelerated GL rendering)
> I've got VTK 3.1

> Anyhow, I'm going to try wxPython and it's VTK widget in the next few
> days.

> If you had to start a new moderately sized project would you use
> wxPython or Tk?

> -Scott Harris
> scottrharris at earthlink.net



From effbot at telia.com  Tue Apr  4 16:59:26 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Tue, 04 Apr 2000 20:59:26 GMT
Subject: Debugging Python 1.3 -> 1.5.2 (was Re: Python 1.6 alpha 1 released)
References: <Pine.GSO.4.10.10004040035420.21084-100000@sundial> <8cca93$9oh$1@nnrp1.deja.com> <B964E942A897FF5F.529887C74CF2FC7E.977238F9475EAC23@lp.airnews.net> <FdoG4.5023$74.81126@newsc.telia.net> <8cdiaa$44i$1@slb7.atl.mindspring.net>
Message-ID: <OIsG4.5116$74.82310@newsc.telia.net>

Aahz Maruch wrote:
> >1) stumbled upon one today: it turned out that by changing from 1.3
> >to 1.5.2, large data transfers from a high-performance server to clients
> >on an 10 Mbit Ethernet were unexpectedly terminated.  the same trans-
> >fers worked just fine with high-performance workstations, on an FDDI
> >network.  ten guru points to anyone who can explain why.
>
> Not without a bit more info about the data transfer mechanism and the
> error reported by client & server.

okay, I'll trade you some info for three guru points:

non-blocking AF_INET sockets.  client read via socket.makefile, and
got an unexpected end of line.  server reported error in write callback.
all code had worked flawlessly under 1.2 and 1.3, for several years.

</F>




From phd at phd.russ.ru  Thu Apr 13 06:18:24 2000
From: phd at phd.russ.ru (Oleg Broytmann)
Date: Thu, 13 Apr 2000 10:18:24 +0000 (GMT)
Subject: simple httplib and urllib timeout question
In-Reply-To: <1WbJ4.86$9B.7202@news.bc.tac.net>
Message-ID: <Pine.LNX.4.21.0004131006220.16470-100000@fep132.fep.ru>

Hi!

On Wed, 12 Apr 2000, darrell wrote:
> Yeah I would love to:
> a) check out your Internet Robot

   Download it here: http://sun.med.ru/~phd/Software/Python/#bookmarks_db -
below the table, after the words "third, new, much improved version".
   Faster mirrors:
http://members.xoom.com/phd2.1/Software/Python/
http://skyscraper.fortunecity.com/unix/797/Software/Python/#bookmarks_db

> b) help out with any code

   Any help will be greatly appreciated! Start with looking into the code.
What do you think about it all? Can you understand it?

   I need more plugins, so choose the area you want to touch:
1) database (storage) managers
2) writers
3) robots.

Oleg.            (All opinions are mine and not of my employer)
---- 
    Oleg Broytmann      Foundation for Effective Policies      phd at phd.russ.ru
           Programmers don't die, they just GOSUB without RETURN.






From noreply at nospam.org  Sun Apr 30 19:05:23 2000
From: noreply at nospam.org (Brian)
Date: Sun, 30 Apr 2000 23:05:23 GMT
Subject: Discrepency between JPython and CPython
References: <m3k8hfjlex.fsf@boss.ksu.edu>
Message-ID: <T_2P4.370$cv3.157016@paloalto-snr1.gtei.net>

Here is a list of differences you might find useful:
http://www.jpython.org/docs/differences.html

"Venkatesh Prasad Ranganath" <rvprasad at cis.ksu.edu> wrote in message
news:m3k8hfjlex.fsf at boss.ksu.edu...
> Hi,
>
> Consider the expression str(1.14 * 200).
> In CPython it yields 228.0
> In JPython it yields 227.99999999999997
>
> Shouldn't this result in consistent values across variuos implementation?
> If so, which of the implementation is buggy?
> I think it is JPython, may be I am wrong.
>
> --
> Venkatesh Prasad Ranganath
>
> e-mail: rvprasad at cis.ksu.edu, rvprasad at ksu.edu
>
> Address: Department of Computing and Information Science,
>          234 Nichols Hall,
>          Kansas State University,
>          Manhattan, KS 66506
>
> Phone: (785)-532-6350 Extn: 48
>




From phd at phd.russ.ru  Mon Apr 17 10:20:02 2000
From: phd at phd.russ.ru (Oleg Broytmann)
Date: Mon, 17 Apr 2000 14:20:02 +0000 (GMT)
Subject: Filtering web proxy
Message-ID: <Pine.LNX.4.21.0004171406250.30999-100000@fep132.fep.ru>

Hello!

   I want a filtering web proxy. I can write one myself, but if there is a
thing already... well, I don't want to reinvent the wheel. If there is such
thing (free and opensourse, 'course), I'll extend it for my needs.

   Well, what are my needs? I am braindamaged idiot, who do not want to see
banners, pop-ups and all the crap, so I run Navigator with graphics/Java/
Javascript/cookies turned off 99% of the time. But for some sites I need
to go to preference dialog and turn it on manually. Then turn it back off
before visiting another site. (I need frames, tables and other features, so
I am not using lynx; yes, I know there are "links" and "w3m", and yes, I
use them from time to time; anyway I run Navigator).
   It is tiresome. I want to automate the task.

   So I want to turn graphics/script/cookies forever on, but filter the
crap in the proxy. For 99% sites I'll just remove the junk. For some sites
(a list of which I'll care manually) the proxy will pass HTML, graphics and
cookies unchanged (or a bit modified, 'cause in any case I don't want to
eat Doubleclick's cookies attached to their ads).
   Once I tried Junkbuster, but found it inadequate. My need is simple -
just a short list of "white" sites. Never tried adzapper...

   I wrote dozen HTML parsers in Python, so I can write one more, and turn
it into a proxy, but may be I can start with some already debugged code?

Oleg.
---- 
     Oleg Broytmann          http://phd.russ.ru/~phd/          phd2 at mail.com
           Programmers don't die, they just GOSUB without RETURN.




From lewst at yahoo.com  Sun Apr  2 19:17:57 2000
From: lewst at yahoo.com (lewst)
Date: Sun, 2 Apr 2000 16:17:57 -0700 (PDT)
Subject: NameError when assigning dictionary values
Message-ID: <20000402231757.6604.qmail@web2105.mail.yahoo.com>

Fredrik Lundh <effbot at telia.com> wrote:

> if you want to create variables in the local namespace
> on the fly, you're obviously trying to emulate some other
> programming language (Tcl?) in Python.  that's usually a 
> rather bad idea.

I was trying to build a dictionary and create lists corresponding to
the value names in the dictionary in one loop.

Do I really have to type "one" and "two" twice?  Once to make it a
variable and a second time to add to my dictionary?  I.e,

>>> one, two = [], []

>>> mydict = { ".": one, "+": two }

This seems error prone if the the number of variables grows very large
as it will for me.


__________________________________________________
Do You Yahoo!?
Talk to your friends online with Yahoo! Messenger.
http://im.yahoo.com



From mwh21 at cam.ac.uk  Wed Apr  5 04:43:42 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 05 Apr 2000 09:43:42 +0100
Subject: A Mountain of Perl Books + Python Advocacy
References: <20000405053920.20579.qmail@web2102.mail.yahoo.com> <38EAF7A0.E372EB2C@myself.com>
Message-ID: <m33dp1ndj5.fsf@atrus.jesus.cam.ac.uk>

Wayne Izatt <wayne.izatt at myself.com> writes:

> Books are one thing, but a regular printed journal would be especially
> useful. A journal is visible, creates a strong sense of community, and is
> an easy way to combine advanced and general Python topics in one tight
> package. I wonder what the minimum number of Pythoners it would take to
> keep something like that afloat?

Hmm.  There's 
 
    http://www.pythonjournal.com

but it's dormant; however the Python community seems to be growing
exponentially (at least to judge by the volume on the newsgroup), so
it's might be worth having another go.  Of course this means getting
people to write stuff, organising things, and general effort.  I could
spare a few hours to such a cause, but perhaps not in any useful way
(do you *really* want an article about bytecodehacks <wink>?).

> lewst wrote:
> 
> > Now something I can't quite figure out is: why are there so many more
> > books on Perl out there than on Python?

It's been more popular for longer.  As I understand it, perl was
created to do be the gaffer tape of a unix system, so people knew it
when the web took off; however it seems that people are realising that
it might be a good idea to use a real programming language.

> > Searching through Fatbrain.COM (which is where I order my books from),
> > I found 68 books, 4 training manuals, and 2 eMatter documents on Perl.
> > Compare this to Python's 13 books and 2 eMatter documents.

This is changing *very* fast - Learning Python is only, what, a year
old - and that was the third real Python book (I think; there's PP and
IPwP - I'm not counting the little reference book).

[schnipp]
> > At this point I wondered if my brain was just different than all those
> > Perl junkies out there.  But now I really don't think so; I think it's
> > a question of awareness.  Perl is very publicized and well-known while
> > the better language is sitting here a dark corner unnoticed.  Sure
> > there will always be some religious fanatics that won't even give
> > Python a try, but I think Python's popularity could be vastly improved
> > with some serious advocacy work.  CNRI and/or PSA should seriously
> > look into funding a Python "marketing" campaign of sorts.  I think the
> > result would be allot of converts and more understanding and respect
> > for Python.

A year ago, I might have agreed with you; now I'm more concerned how
the Python community is going to evolve as it grows.

> >   "The verdict: Python is *waaaay* cool!  I'm sold.  It's clean, it's
> >   elegant, it's easy, and it's astonishingly powerful.  I'm not going
> >   to program anything longer than one screenful of script in Perl
> >   anymore.  I love Larry Wall dearly, but Guido van Rossum is the
> >   better designer -- I haven't had this much fun with a language since
> >   the glory days of LISP.  Eric sez check it out."

Python rocks.  I think it took me about a day to learn - it just fit
so well with my way of thinking.  I've now since learnt scheme,
eiffel, haskell, common lisp and bits of sundry other languages, but I
keep coming back to Python.

Cheers,
M.

-- 
well, take it from an old hand: the only reason it would be easier
to program in C is that you can't easily express complex  problems
in C, so you don't.                 -- Erik Naggum, comp.lang.lisp


From mwh21 at cam.ac.uk  Thu Apr 20 09:26:26 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 20 Apr 2000 14:26:26 +0100
Subject: how do i pack and unpack messages in python?
References: <008301bfaacb$42fe2c60$2801000a@iel.ie>
Message-ID: <m3ln28hpj1.fsf@atrus.jesus.cam.ac.uk>

"Shaun" <shogan at iel.ie> writes:

> so i have 3 values that i want to pack into a message so as the info can be
> sent (but dont worry about th3 sending bit for the mo)

Hmm, what are you sending the message to?  Email?  Some kind of IPC?
An HTML form?  I guess I don't understand what you want.

> can anyone please tell me how id go about manipulating the strings so as i
> can join them using field seperators, any ideas anyone, or even tell me
> where i might find out how to do this problem.

Have you tried the struct module?

Cheers,
M.

-- 
  "declare"?  my bogometer indicates that you're really programming
  in some other language and trying  to force Common Lisp into your
  mindset.  this won't work.           -- Erik Naggum, comp.lang.lisp


From donb at init.com  Mon Apr 17 17:20:31 2000
From: donb at init.com (Donald Beaudry)
Date: Mon, 17 Apr 2000 17:20:31 -0400
Subject: Another question 
References: <dVFK4.41$vr2.5817@news.uswest.net> <38FB2EEB.A32AF88C@bam.com> <38FB339F.604364AD@bam.com> <200004171605.MAA04929@zippy.init.com> <38FB7610.80EAF965@ichips.intel.com>
Message-ID: <200004172120.RAA11102@zippy.init.com>

jimsp <jimsp at ichips.intel.com> wrote:
> How about
> 
> >>> import random
> >>> x = [0]*1000000
> >>> y = map(lambda x: random.randint(1,10), x)

Hmm.... believe it or not that one is much closer to what I meant to
type as my 'or better' example, than the rubbish that I did type ;)

But, this one beats it (and I tested it this time ;).

  import random
  y = map(lambda x, randint=random.randint: randint(1,10), xrange(100000))

and this one is better still (at least on my machine):

  def do_it():
      from random import randint
      r = [0,] * 100000
      for i in xrange(100000):
         r[i] = randint(1,10)
  do_it()

--
Donald Beaudry                                     Ab Initio Software Corp.
                                                   201 Spring Street
donb at init.com                                      Lexington, MA 02421
                      ...Will hack for sushi...



From ajung at sz-sb.de  Wed Apr  5 02:00:43 2000
From: ajung at sz-sb.de (Andreas Jung)
Date: Wed, 5 Apr 2000 08:00:43 +0200
Subject: Getting the argument names passed to a function ?
In-Reply-To: <slrn8el5t7.40k.quinn@riyal.ugcs.caltech.edu>; from quinn@riyal.ugcs.caltech.edu on Wed, Apr 05, 2000 at 01:30:10AM +0000
References: <20000404225810.A12156@sz-sb.de> <HRsG4.5120$74.82867@newsc.telia.net> <slrn8el27m.3h1.quinn@riyal.ugcs.caltech.edu> <slrn8el5t7.40k.quinn@riyal.ugcs.caltech.edu>
Message-ID: <20000405080043.A4282@sz-sb.de>

> import sys
> def arghack(arg):
>     try:
>         1/0
>     except:
>         frame = sys.exc_info()[2].tb_frame.f_back
>     while frame:
>         for k, v in frame.f_locals.items():
>             if arg is v:
>                 return k
>         frame = frame.f_back
>     else:
>         return None

That's the perfect solution I was looking for :-) Thanks !

Cheers,
Andreas



From embed at geocities.com  Thu Apr 27 13:46:55 2000
From: embed at geocities.com (Warren Postma)
Date: Thu, 27 Apr 2000 13:46:55 -0400
Subject: Python Time classes
References: <slrn8ge5b0.1gu.philh@vision25.demon.co.uk>
Message-ID: <qz%N4.12701$HG1.340144@nnrp1.uunet.ca>

"phil hunt" <philh at vision25.demon.co.uk> wrote in message
news:slrn8ge5b0.1gu.philh at vision25.demon.co.uk...
> Is there a high-level Time class for Python?
>
> I'm aware of the 'time' module in the standard library. But I'm
> looking for something more high-level (I can't find anything
> relevant on Parnassus or Freshmeat). What I want is roughly the
> ability to say things like:
>
>    t1 = Time.now()
>    # (t1) is an instance of Time containing the time now
>
>    t2 = Time(2000, 4, 26, 17, 4, 0)


mxDateTime does exactly this and more. It even comes with nifty code to
synchronize you with an atomic clock, and a variety of wrapper classes
designed for various purposes. It can handle historical dates (dates before
the Unix epoch, Jan 1, 1970) and can handle the milliseconds component of
Time values without problems, as well as handling day of week, leap-years,
and even includes code for algorithms for moveable feasts (When is Easter
Sunday in the year 2143, for instance.)

The core date time object is written in C, and all the extraneous stuff is
in Python modules (.py), which is pretty nice.

I hope they rename mxDateTime to DateTime and roll it right into the stock
Python distribution at some point. It just seems "right" to me that Python
should include a powerful date and time handling object as a native type,
along with lists, dictionaries, tuples, and other objects.

Warren




From hank at prosysplus.com  Sat Apr 22 09:33:13 2000
From: hank at prosysplus.com (Hank Fay)
Date: Sat, 22 Apr 2000 13:33:13 GMT
Subject: Using python on the web
References: <EDFD2A95EE7DD31187350090279C6767754BA8@THRESHER> <FtEE3B.Ar2.0.bloor@torfree.net>
Message-ID: <tShM4.42074$h6.182368@typhoon.jacksonville.mediaone.net>

Hi Ken,

An article on Python Server Pages appeared in two parts recently in DDJ.
The web site associated with the article is at
http://www.ciobriefings.com/psp/.

Chuck's announcement of WebWare in this group is just as exciting, since it
gets rid of the overhead of the servlet engine; putting together "server
pages" in Visual FoxPro on Wintel was a snap using a similar tool that only
(of course) works on WinTel.  I'm looking forward to doing the same with
Python, using WebKit as the tool.

Hank Fay

--
http://www.prosysplus.com
"Ken McCracken" <aa175 at torfree.net> wrote in message
news:FtEE3B.Ar2.0.bloor at torfree.net...
> Brent Fulgham (brent.fulgham at xpsystems.com) wrote:
> : A couple of questions:
>
> :
> : > We build medium-busy, CGI-intensive sites in Python. We were briefly
> : > quite excited about Zope, but once we realized it locked us into Yet
> : > Another Programming Language, we lost interest. DHTML is particularly
>
> I'm a bit of an armchair programmer, but it would seem that JSP nad
> Jpython and a Db could be put together to end up with a zopelike result,
> without language lockin. No?
>
> Ken McCracken
> Toronto




From sholden at bellatlantic.net  Sat Apr  8 07:52:39 2000
From: sholden at bellatlantic.net (Steve Holden)
Date: Sat, 08 Apr 2000 11:52:39 GMT
Subject: Cacheing sys.path rather than building it each time?
References: <2e070214.0e300ac8@usw-ex0108-063.remarq.com>
Message-ID: <38EF1B53.D21A7514@bellatlantic.net>

Hamish Lawson wrote:
> 
> Some overhead in Python's startup time is caused by building
> sys.path. Of course you could use the -S option, but then you
> don't get a proper sys.path! What about cacheing the constructed
> path and a command-line option to specify use of that rather
> than building it each time? Presumably you'd also need a tool to
> rebuild the path when necessary - perhaps it could be invoked
> automatically by the distribution tools whenever a new module
> gets installed.
> 
> Hamish Lawson
> 
> * Sent from RemarQ http://www.remarq.com The Internet's Discussion Network *
> The fastest and easiest way to search and participate in Usenet - Free!

Probably not an awfully good idea, given that "installation" of a new
module often consists of just dropping a ".py" file into a library
directory.  Without an common"install" process the work of checking
whether the cahced path is up to date would be almost as much as that
of building the path from scratch as is now done.

Unless, that is, someone else knows different...

regards
 Steve
--
"Bulding information systems just because it's fun."
Steve Holden  sholden at bellatlantic.net  703 716 7275


From herzog at online.de  Mon Apr  3 08:57:47 2000
From: herzog at online.de (Bernhard Herzog)
Date: 03 Apr 2000 14:57:47 +0200
Subject: List and Dicts as default args
References: <326536345498D311B3BC00105A39802A074654@newsintern.dspace.de> <vQZF4.1086$n68.187094016@newsb.telia.net>
Message-ID: <m3n1nbpcj8.fsf@greebo.nodomain.de>

"Fredrik Lundh" <effbot at telia.com> writes:

> Stefan Migowsky <smigowsky at dspace.de> wrote:
[mutable default parameters]
> > I could try the following but that doesn't look so nice:
> >
> > def f(Index,List = None, Dict = None):
> >     if not List: List = []

It's probably better to test whether List is None:

	if List is None: List = []

Otherwise, if you actually call f with an empty list as the second
parameter, you'll still create a new list object and won't modify the
list that was passed in.

> >     List.append(1)
> >     List.append(2)
> >     print List
> >     if not Dict: Dict = {}

The same applies here.

> >     Dict[Index] = Index
> >     print Dict
> 
> nice or not, that's the right way to do it.


-- 
Bernhard Herzog   | Sketch, a drawing program for Unix
herzog at online.de  | http://sketch.sourceforge.net/


From donn at u.washington.edu  Thu Apr 13 12:44:14 2000
From: donn at u.washington.edu (Donn Cave)
Date: 13 Apr 2000 16:44:14 GMT
Subject: 1st non-trivial program - gentle criticism requested
References: <1e90r16.1sd2ts41e7ur6kN@gershwin.bio.ic.ac.uk>
Message-ID: <8d4tgu$11qa$1@nntp6.u.washington.edu>

Quoth p.agapow at ic.ac.uk (Paul-Michael Agapow):
| I'm experimenting with Python for a number of purposes (an embedded
| language for simulations, a tool for exploring experimental data, a good
| first programming language) and so I've cobbled together my first
| non-trivial program. With some trepidation, I post it here in the hopes
| that experts will be able to coach me in "The Python Way". I realise it
| looks like C++ (my usual langauge), but criticisms of style would be
| useful too.

OK:  It looks too much like C++!  You don't need to end lines with ;.
You don't need to surround expressions with () in an if statement.  It's
not the worst of sins, but you might as well ditch this stuff.

As for the data structures, like "Alex" I would think about more classes
and fewer module globals.  This is not a critical flaw, assuming your
program works, but as a general rule the class is your most adaptable
programming tool and to benefit from it, you have to use it.  As an
exercise, rewrite so the only module level statements are "import",
"class" and "if" (i.e., if __main__...)

That tends towards the other extreme of usage, inasmuch as it is really
very common to use a few module level functions or global constants.
It's just an exercise, not a religious mandate.  In the course of the
exercise, you may find yourself in territory where good usage is
somewhat controversial.  For example, objects other than methods
are legal and fairly common in "class" scope, but there are some
pitfalls to this that worry some people.  Experiment, look for the
properties of an object defined in class scope when seen through
the class instance ("self"), vs. when accessed directly via the class.

There's a builtin max() that I think can do the work of your findMax.

I have plenty of these in my programs:

   print "The alphabet is '" + gAlphabet + "'"

but today I prefer

   print "The alphabet is '%s'" % gAlphabet

because of the greater tolerance for objects that are not really
strings but have string representations.  (OK, it's a string, but
what will happen if you later change this object to something else?)

	Donn Cave, donn at u.washington.edu


From rogerha at ifi.uio.no  Mon Apr 17 08:11:29 2000
From: rogerha at ifi.uio.no (Roger Hansen)
Date: 17 Apr 2000 14:11:29 +0200
Subject: Python tools for PDE and Linear algebra
References: <ya0vh1kzsaf.fsf@grindaflet.ifi.uio.no>
	<38F75987.26AD3FEA@oreilly.com>
Message-ID: <ya0og79vsem.fsf@sessrymnir.ifi.uio.no>

* Me
> > Does anyone know
> > some PDE tools or are working on such projects? Support for different
> > numerical methods as Finite Difference and Finite Element Methods
> > would be nice.

* Stephen R. Figgins
> I don't know of any in Python, Roger, but there are some free
> libraries available for C++. You can find them listed at:
> 
>   http://oonumerics.org/oon/
> 
> You might be able to adapt one of these, rewrite it in python or
> maybe just call it from python.
> 
> I expect these would be rather computationally expensive, and so you
> might be better off with having them written in C or C++.

Yes, there are several libraries for these things in C/C++. But since
I find Python a much better programming language than C++ I'd like to
use Python. I realize that I then might have to wrap an existing
library. But that may be a rather big job, especially wrapping a C++
library, so if anyone has done this already I'd be very happy. :-)


TTFN,

Roger


From mwh21 at cam.ac.uk  Thu Apr 20 16:19:08 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 20 Apr 2000 21:19:08 +0100
Subject: can't find SWIG for Python
References: <38FF5A56.F76794D8@sec.noaa.gov>
Message-ID: <m3em80h6f7.fsf@atrus.jesus.cam.ac.uk>

j vickroy <jvickroy at sec.noaa.gov> writes:

> Hello all,
> 
> I have been unable to locate SWIG for Python.  My references list
> www.swig.org but the server can not be found.

Egad, you're right.  This is getting a bit familiar...

> Could someone point me to the site or provide some other Win 32 options
> for easing the process of extending Python to use existing C libraries.

My local mirror (www.mirror.ac.uk) has about a bazillion different
copies of it; maybe you can get one of them to work?

Cheers,
M.

-- 
59. In English every word can be verbed. Would that it were so in 
    our programming languages.
     -- Alan Perlis, http://www.cs.yale.edu/~perlis-alan/quotes.html


From bwinton at tor.dhs.org  Mon Apr 10 16:02:40 2000
From: bwinton at tor.dhs.org (Blake Winton)
Date: Mon, 10 Apr 2000 20:02:40 GMT
Subject: Why should I switch to Python?
References: <seas5ia89177@corp.supernews.com>
Message-ID: <slrn8f4d0j.1t0.bwinton@tor.dhs.org>

On Fri, 31 Mar 2000 19:43:58 -0800, Aaron Turner wrote:
>I've been coding Perl for over 3 years and really like it.  (Especially
>it's syntax and ability to grok my programming style.)

Now, just as a side note, why is it that the sets of The Perl Hackers I
Know, and The People I Know Who Use The Word "Grok" are the same set?

Just wondering,
Blake.
-- 
2:54pm up 12 days, 15:28, 1 user, load average: 1.08, 1.02, 1.01


From nickm at mit.edu  Thu Apr 20 00:12:26 2000
From: nickm at mit.edu (Nick Mathewson)
Date: Thu, 20 Apr 2000 04:12:26 GMT
Subject: pattern-based string expansion
References: <38F52C61.A5D72F89@gte.net> <vndog79ft1q.fsf@camelot-new.ccs.neu.edu> <8del87$5md$1@slb6.atl.mindspring.net> <38FE70DB.29A4B1A1@gte.net>
Message-ID: <KsvL4.1559$Ez2.89485@typhoon1.gnilink.net>

On Thu, 20 Apr 2000 02:51:50 GMT, sue <smalleys at gte.net> wrote:
>I'm not looking for all possible strings with a wildcard,
>just a pseudo-RE-based expansion.  The RE would not support
>open-ended arguments (* or +), but would have ?,[] and so
>on.
>The expansion needs to be ordered.
>
>I've thought about hacking the glob(3) code to work on an
>input RE, but I was hoping this already existed somewhere.

Well, the code below will probably only be useful to a hacker, but if
you're prepared to hack glob(3), this should be a snap.  

It's a very very simplified version of (a subset of (a more general
regex-manipulation module for (a project that I'm working on))).  I
didn't need an expansion function for my project, but it was pretty
easy to hack one up.

============================================================
#!/usr/bin/python

import string

def expand_re(s):
    '''Given a regular expression, returns all a list of all possible 
       expansions.  It supports (), |, [], ?, \[dwsDWS], and '.'.
       The backslash character escapes any other character.  

       Example:
           expand_re('spam( spam( spam[!?]?)?)?|Bloody Vikings!')
              => [ 'spam spam spam!', 
                   'spam spam spam?', 
                   'spam spam spam',
	           'spam spam',
		   'spam',
		   'Bloody Vikings!' ]

       BUG: Only mildly tested.	      
       BUG: Doesn't catch bad regular expressions, or give good error
            messages.
       BUG: Isn't very fast or efficient.
       BUG: Can't handle character escapes within [].
       BUG: Can't handle [^...].
       BUG: Doesn't support unicode. 
       BUG: Doesn't support case-insensitive matching.
       BUG: Always assumes that . should match newline.
       BUG: Doesn't support +, *, {n,m}.  In order to do these, we'd need
            a function to only consider expansions up to a certain length,
            or to only return the first n expansions.
       '''

    re, pos = _parseRE(s,0)
    return re.get_expansions()

def _parseRE(s, pos):
    '''Does the dirty work of parsing a regular expression from string 's',
       starting at position 'pos'.  Returns the Seq/Alt/Opt/Charset object
       corresponding to s, along with the position immediately following
       
       returns element, end tuple'''

    # First, we find all the elements in s.  Elements are:
    #  (groups)
    #  [character sets]     (also '.', \d, and friends.)
    #  characters
    #  optional-elements?
    #  |
    elements = []
    while pos < len(s):
	ch = s[pos]
	if ch == '(':
	    group, pos = _parseRE(s, pos+1)
	    pos = pos-1
	    elements.append(group)
	elif ch == ')':
	    pos = pos + 1
	    break
	elif ch == '\\':
	    next = _special.get(s[pos+1], None)
	    if next:
		elements.append(next)
	    else:
		elements.append(s[pos+1])
	    pos = pos + 1
	elif ch == '[':
	    chrs = []
	    while 1:
		pos = pos + 1
		ch = s[pos]		
		if ch == ']':
		    break
		chrs.append(ch)
	    elements.append(CharSet(string.join(chrs,'')))
	elif ch == '?':
	    elements[-1] = Alt([elements[-1], ''])
	elif ch == '|':
	    elements.append(None) #HACKHACK
	elif ch == '.':
	    elements.append(_allChars)
	else:
	    elements.append(ch)
	pos = pos + 1

    # Now we compact all adjacent characters into strings.  This will help
    # our efficiency in the end, I think.  The revised list is stored in
    # 'els'.
    els = []
    cur_str = None
    for el in elements:
	if type(el) == type(''):
	    if cur_str is not None:
		cur_str = cur_str + el
	    else:
		cur_str = el
	else:
	    if cur_str is not None:
		els.append(cur_str)
		cur_str = None
	    els.append(el)
    if cur_str:
	els.append(cur_str)

    # Now we go through the list to look for vertical bars -- encoded as
    # None.  We find all alternatives within the current group.
    groups = []
    cur_group = []
    for el in els:
	if el is None:
	    groups.append(cur_group)
	    cur_group = []
	else:
	    cur_group.append(el)
    groups.append(cur_group)
    
    # If there's more than one alternative, return an Alt object.  Otherwise,
    # a Seq.
    if len(groups) == 1:
	return Seq(groups[0]), pos
    else:
	return Alt(map(Seq,groups)), pos

class Seq:
    '''This class represents a series of regular expression elements, one 
       following another'''
    def __init__(self, members):
	self.members = members

    def get_expansions(self):
	member_expansions = map(_expand, self.members)
	return _joinings(member_expansions)
    
class Alt:
    '''This class represents a choice between two or more elements.'''
    def __init__(self, alternatives):
	self.alternatives = alternatives

    def get_expansions(self):
	member_expansions = map(_expand, self.alternatives)
	ex = []
	for el in member_expansions:
	    # This is a hack; I'd rather say 'ex.extend(el)', but not
            # everyone is using Python >=1.5.2
	    ex[len(ex):len(ex)] = el
	return ex

class CharSet:
    '''This class represents a group of possible characters, 
       such as [abc] or \d' '''
    def __init__(self, characters):
	self.characters = list(characters)

    def get_expansions(self):
	return self.characters    

def _expand(element):
    '''Returns the expansion of an element.'''
    if type(element) == type(''):
	return [ element ]
    else:
	return element.get_expansions()

def _joinings(lst):
    ''' Takes a list of lists of strings, such as [ ['a1','a2'], ['b1, b2']].
        Returns a new list consisting of the joins of one member from list1,
        one from list2, and so forth. (Such as ['a1b1','a1b2','a2b1','a1b2'])
	'''
    prefixes = ['']
    for sublist in lst:
	next_prefixes = []
	if len(sublist) == 0:
	    return []
	for prefix in prefixes:
	    for el in sublist:
		next_prefixes.append(prefix + el)
	prefixes = next_prefixes 
    return prefixes

_allChars = CharSet(string.join(map(chr,range(0,255)),''))
_special = { 
             't' : '\t',
	     'n' : '\n',
	     'r' : '\r',
	     'f' : '\f',
	     'e' : '\e',
             's' : CharSet(' \t\n\f\r'),
	     'd' : CharSet(string.digits),
	     'w' : CharSet(string.letters + string.digits + "_"),
	     'W':  CharSet(map(chr, [32,96] + range(0,48) + range(58,65) +
			       range(91,95) + range(123,256))),
	     'S':  CharSet(map(chr, [65,11]+range(0,8)+range(14,32)+
			       range(33,92)+
			       range(93,256))),
	     'D':  CharSet(map(chr, [32]+range(0,48)+range(58,256)))
	     }

if __name__ == '__main__':
    import sys
    expr = sys.argv[1]
    for el in expand_re(expr):
	print el
============================================================

Was that about what you had in mind?

Delurkingly-yrs
-- 
Nick Mathewson     <nickm at mit.edu>     http://www.mit.edu/~nickm/


From m.faassen at vet.uu.nl  Sun Apr  9 19:22:00 2000
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 9 Apr 2000 23:22:00 GMT
Subject: Tuples -- who needs 'em
References: <1256882636-27643467@hypernet.com>
Message-ID: <8cr3ao$j40$1@newshost.accu.uu.nl>

[Bob]
>> > This sort of thing
>> > happens in exploratory programming, so when in that mode one might want to
>> > use lists unless one is really sure in advance that tuples are the right
>> > thing.

[Gordon]
> By all means, use lists whenever you're unsure. You'll soon 
> come to recognize when a tuple is more appropriate. It won't 
> take long (it couldn't, Martijn's too young to have taken long to 
> learn anything <wink>).

Heh. Hey, perhaps I'm just very bright? Nah, impossible. I didn't
understand half of the 'Perl versus Python' math thread. :)

Regards,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From thomas.heller at ion-tof.com  Mon Apr 17 14:48:25 2000
From: thomas.heller at ion-tof.com (Thomas Heller)
Date: Mon, 17 Apr 2000 20:48:25 +0200
Subject: import disappointment
Message-ID: <002b01bfa89d$834c49d0$4500a8c0@thomasnb>

After playing with 'import' and __import__ for quite some time
I discovered that neither ihooks nor knee works correctly.
Put

import ihooks
ihooks.install()

into your sitecustomize.py and start Pythonwin: it will fail.

There are several problems here:
1. reload does not what it should (I submitted a patch for ihooks
to correct this problem, the same technique must be used for knee).
2. ihooks does not import modules via the registry: there is a bug
report in the python bug list for this one, I will also submit a (partial)
patch for that.
3. Another yet unfound problem

Using imputil is out of question because most programs
including pythonwin are only expecting strings in sys.path.

Before I use more work on this: Has anybody a version
which really works correctly?

Thomas Heller





From kern at caltech.edu  Sat Apr 29 18:59:25 2000
From: kern at caltech.edu (Robert Kern)
Date: 29 Apr 2000 22:59:25 GMT
Subject: Where can I find NumPy
References: <8ef4pn$jge$1@kopp.stud.ntnu.no>
Message-ID: <8efpgd$c49@gap.cco.caltech.edu>

In article <8ef4pn$jge$1 at kopp.stud.ntnu.no>,
	Jan Erik Breimo <janeb at stud.ntnu.no> writes:
> The official site (ftp-icf.llnl.gov) hasn't been responding for more than
> a week, so are there any other sites where I can find the source
> files for NumPy?

http://numpy.sourceforge.net

> --
> Jan Erik Breimo
> email: janeb at stud.ntnu.no

-- 
Robert Kern
kern at caltech.edu

"In the fields of hell where the grass grows high
 Are the graves of dreams allowed to die."
  -- Richard Harter


From steven at nospam.trance.org  Sun Apr 30 13:16:41 2000
From: steven at nospam.trance.org (Steven)
Date: Sun, 30 Apr 2000 18:16:41 +0100
Subject: xitami configuration?
Message-ID: <wPZO4.13426$r6.69041@stones>

I'm just starting out with python after programming php3 for the past 8
months and i'm trying to set it up with xitami.

could an xitami/python user please tell me what line to add to my filters
section in the setup file? I tried:

.py=D:\progra~2\python\python.exe

which will envoke python and run a simple script fine, but won't pass any
arguments (unless i'm doing it wrong). If this is correct, how do i pass
arguments to a script?




From markx.daley at intel.com  Mon Apr 24 19:19:23 2000
From: markx.daley at intel.com (Daley, MarkX)
Date: Mon, 24 Apr 2000 16:19:23 -0700
Subject: Tkinter
Message-ID: <A63AFB20111ED311AC5500A0C96B7BF5015F0F44@orsmsx36.jf.intel.com>

I am trying to use Tkinter for a GUI interface, but I am having a hard time
finding manuals for it.  Is there documentation comparable to Python's?
Also, is there a demo script floating around anywhere that gives generic
examples for common GUI functions or widgets?

TIA

- Mark




From chaitin at watson.ibm.com  Sat Apr  8 10:33:50 2000
From: chaitin at watson.ibm.com (chaitin at watson.ibm.com)
Date: Sat, 08 Apr 2000 14:33:50 GMT
Subject: math vs physics
References: <Pine.GSO.4.10.10004050749030.24181-100000@sundial> <slrn8enkhp.456.neelk@brick.cswv.com>
Message-ID: <8cng0e$gcu$1@nnrp1.deja.com>

In article <slrn8enkhp.456.neelk at brick.cswv.com>,
  neelk at alum.mit.edu wrote:

>
> I tend towards Chaitin's view: "Axioms are very cool -- make some
> more up! After all, mathematics is basically arbitrary anyway!" I
> guess this shows I am really a frustrated Hilbertist at heart. :)
>

You might enjoy my latest talk.
You can find it at http://www.cs.umaine.edu/~chaitin/cmu.html
Rgds,
Greg Chaitin


Sent via Deja.com http://www.deja.com/
Before you buy.


From Michael.Scharf at gmx.de  Tue Apr  4 19:46:05 2000
From: Michael.Scharf at gmx.de (Michael Scharf)
Date: Wed, 05 Apr 2000 01:46:05 +0200
Subject: Q: Status of SOAP (Simple Object Access Protocol)?
Message-ID: <38EA7EBD.BE75C460@gmx.de>

I am wondering, what the current python-SOAP status is?
(about SOAP http://www.develop.com/soap/)

That's what I found on the net:
- Digital Creations announced SOAP support
  (couldn't figure out, what the status is)
- Fredrik Lundh said some time ago, he's working on it....

We are currently (happily) using XML-RPC (http://www.xmlrpc.com/),
but the future seems to be SOAP....

On the other hand, SOAP seems more complicated than XML-RPC.

Michael
-- 
     ''''\     Michael Scharf
    ` c-@@     TakeFive Software GmbH, a Wind River Company
    `    >     http://www.WindRiver.com
     \_ V      mailto:Michael.Scharf at gmx.de


From wware at world.std.com  Wed Apr 26 20:46:49 2000
From: wware at world.std.com (Will Ware)
Date: Thu, 27 Apr 2000 00:46:49 GMT
Subject: Python compiler ? (was Re: Python for drivers?)
References: <3900EDBE.282BC85@dsuper.net> <8dr2pn$ec8$1@newshost.accu.uu.nl> <3906f804@nntp.server.uni-frankfurt.de> <015b01bfaf8b$9cd03930$0500a8c0@secret.pythonware.com> <39076316@nntp.server.uni-frankfurt.de> <8e7sop$5io$1@nnrp1.deja.com>
Message-ID: <FtnGu1.Dso@world.std.com>

gchiaramonte at my-deja.com wrote:
> How about $100 each? Then we only need 600 users. If his compiler
> really works, count me in.

That's not a bad price per-person, but what are the chances that he'd
be willing to open-source it? He seems to regard this as a revenue
generator, so he'd likely consider the source code (or at least a
large fraction of it) a trade secret. I'd also be willing to kick in
$100 for an efficient Python compiler, but only if it were open source.
-- 
 - - - - - - - - - - - - - - - - - - - - - - - -
Resistance is futile. Capacitance is efficacious.
Will Ware	email:    wware @ world.std.com


From moshez at math.huji.ac.il  Sun Apr 23 02:47:00 2000
From: moshez at math.huji.ac.il (Moshe Zadka)
Date: Sun, 23 Apr 2000 09:47:00 +0300 (IDT)
Subject: code not true?
In-Reply-To: <ODwM4.1545$RM6.4069477@news-east.usenetserver.com>
Message-ID: <Pine.GSO.4.10.10004230946161.17585-100000@sundial>

On Sun, 23 Apr 2000, Kragen Sitaker wrote:

> How many decimal digits are needed to make sure that 3 * (1/3) is
> indistinguishable from 1?

moshe-bombei:/home/moshe>python
Python 1.5.2 (#1, Feb 21 2000, 14:52:33)  [GCC 2.95.2 19991024 (release)]
on sunos5
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> (1.0/3)*3 == 1
1


Less then 17





From sholden at bellatlantic.net  Thu Apr  6 17:34:41 2000
From: sholden at bellatlantic.net (Steve Holden)
Date: Thu, 06 Apr 2000 21:34:41 GMT
Subject: More about variables
References: <meh9-5C17A0.10445406042000@news.cit.cornell.edu> <3X3H4.16435$0o4.108310@iad-read.news.verio.net> <meh9-158D1A.14240006042000@news.cit.cornell.edu>
Message-ID: <38ED017E.1E374737@bellatlantic.net>

Matthew Hirsch wrote:
> 
> Why create named variables?
> 
> I always thought that having 20 lists saved more space than having one
> list of 20 lists.  Maybe I'm wrong.
> 
> Thanks for your help,
> Matt
> 

First make it work.  *Then* make it work better!

If you have to be this concerned about optimization so early in
your development you must be working in a VERY restricted environment...

Python is great at all kinds of stuff.  Let the language do what it's
good at, and optimize only when performance or resource usage gets the
better of you!

regards
 Steve

> In article <3X3H4.16435$0o4.108310 at iad-read.news.verio.net>,
> culliton at clark.net (Tom Culliton) wrote:
> 
> > What (or maybe what do you think) you are trying to do?  Most likely
> > you would be better off with just a list or tuple if you're dealing
> > with and unknown number of data items.  Not everything needs a name,
> > and there isn't much difference between:
> >
> > v0, v1, v2, v3, v4, v5
> >
> > and
> >
> > v[0], v[1], v[2], v[3], v[4], v[5]
> >
> > Why do you think this variable length collection of data needs
> > individual names?
> >
> > Now, if after consideration there is still some compelling reason to
> > generate names, you can use something like this with a dictionary
> > (possibly the one returned by globals()):
> >
> > v = {}
> > for i in range(number_of_variables):
> >         name = "variable%d" % i
> >         v[name] = []
> >
> > But again, Why?
> >
> > In article <meh9-5C17A0.10445406042000 at news.cit.cornell.edu>,
> > Matthew Hirsch  <meh9 at cornell.edu> wrote:
> > >Hi All,
> > >
> > >Let's say I had something like:
> > >
> > >number_of_variables=int(raw_input('Enter number of variables: '))
> > >
> > >if number_of_variables==1:
> > >   variable1=[]
> > >elif number_of_variables==2:
> > >   variable1=[]
> > >   variable2=[]
> > >elif number_of_variables==3:
> > >   variable1=[]
> > >   variable2=[]
> > >   variable3=[]
> > >elif number_of_variables==4:
> > >   variable1=[]
> > >   variable2=[]
> > >   variable3=[]
> > >   variable4=[]
> > >elif number_of_variables==5:
> > >   variable1=[]
> > >   variable2=[]
> > >   variable3=[]
> > >   variable4=[]
> > >   variable5=[]
> > >else:
> > >   print 'else'
> > >
> > >Is there a more efficient way of doing this? What if I wanted to create
> > >100 variables (lists).  It doesn't make sense to write an if statement
> > >for that many conditions.
> > >
> > >Thanks for your help,
> > >Matt

--
"Bulding information systems just because it's fun."
Steve Holden  sholden at bellatlantic.net  703 716 7275


From reic0024 at ub.d.umn.edu  Fri Apr 21 01:16:48 2000
From: reic0024 at ub.d.umn.edu (Aaron J Reichow)
Date: Fri, 21 Apr 2000 00:16:48 -0500
Subject: Mac Python Questions
In-Reply-To: <956247497.216189@fezzik.endicor.com>
References: <IRjL4.271$CP3.3183@news.uswest.net> <38FE8032.DE233C29@cosc.canterbury.ac.nz> <956247497.216189@fezzik.endicor.com>
Message-ID: <Pine.SO4.4.02.10004210015280.25675-100000@ub.d.umn.edu>

On 20 Apr 2000, Kent Polk wrote:

> The standard Python installation program installed the wrong version
> of the _tkinter shared library and(/or) didn't set the path to the
> correct (ppc) version, so Python couldn't load _tkinter. If that
> is the problem, there are several ways to solve it.  If that isn't
> the problem, I'm not sure what to tell you as it's working on I
> think 3 or 4 different mac platforms here.

That's my problem.  I'm using OS 9, and I even added a path to where
Tkinter appears to be, but it still cannot find _tkinter. <shrug>  This
problem has been known for a while (I've heard of it since at least 1.5.1
as being just a screw up with the path), so why hasn't it been fixed?

Aaron
-----
"The best way to predict the future is to invent it." -- Alan Kay




From wware at world.std.com  Wed Apr 26 15:28:42 2000
From: wware at world.std.com (Will Ware)
Date: Wed, 26 Apr 2000 19:28:42 GMT
Subject: Simple Function Question
References: <8e7dh3$1gc@autodesk.autodesk.com>
Message-ID: <Ftn23v.BuD@world.std.com>

Akira Kiyomiya (akira.kiyomiya at autodesk.com) wrote:
> This function prints out '4' and I have a feeling that x=1 and n=3, so
> add1(3) prints '4'.
> However, I have a difficlut time trying to see how these two arguments
> assigned to both x (especially x) and n.
> def make_incrementer(n):
>     def increment(x, n=n):
>         return x+n
>     return increment
> add1 = make_incrementer(1)
> print add1(3)  # This prints '4'

When you run make_incrementer(1), let's see what happens. You've given
it an argument of 1, which is given the name 'n' inside the body of
make_incrementer. Within that body, you come to 'def increment...'
which, as you know, defines a function. The notation 'n=n' means that
increment can be called with either one argument or with two arguments,
and if it's called with one argument, then its second argument will
assume a default value. That value is 'n' (taken from its value within
the body of make_incrementer), so when you later say 'make_incrementer(1)'
that default value becomes 1. The next thing that happens is that
make_incrementer returns this function; in Python, functions are
objects that can be passed or returned, just like numbers, strings,
lists, dictionaries, or class instances. So this function is given
the name 'add1'. When you call 'add1(3)', the function (originally
called incrementer, and with 1 as the default value for its second
argument) takes 3 as the first argument, adds its two arguments, and
returns the sum (4) which gets printed.

The first tricky idea here is that functions can be passed like any
other object. You'll be familiar with this if you've used Lisp or
Scheme, or Postscript, and probably other languages that escape me now.
In C, a function is pretty tightly connected to its original name; you
can pass pointers to functions but not the function itself. And the
only way to hide the name is to declare the function static, and then
it's only hidden outside that particular source file.

The second tricky idea is that functions' arguments can have default
values, which are set when the 'def' happens and don't change after
that. This allows you to hide all kinds of useful information in an
anonymous function, or create multiple instances of the function with
different default arguments. A default argument can be another function,
possibly with its own default arguments.

I hope this helps.
-- 
 - - - - - - - - - - - - - - - - - - - - - - - -
Resistance is futile. Capacitance is efficacious.
Will Ware	email:    wware @ world.std.com


From amused at webamused.com  Thu Apr 13 19:41:36 2000
From: amused at webamused.com (Joshua Macy)
Date: Thu, 13 Apr 2000 23:41:36 GMT
Subject: Help
References: <38eddc38$0$17341@wodc7nh6.news.uu.net>
Message-ID: <38F64DA3.ECB23626@webamused.com>

T Zimmerman wrote:
> 
> I have left several posts concerning Tkinter Meta keys, and yet to get a
> response from anyone. If this doesnt work for anyone else than it is a bug.
> Otherwise there is something wrong on my end. The following script should
> allow the use of 'alt-f' to access the 'File' menu. This does not work on
> any of the NT boxes in my office. If this does or does not work for someone
> please let me know.
> 
> Thanks,
> T Zimmerman
> tzman at bhmi.com
> 
> ==================================================================
> import Tkinter
> 
> root = Tkinter.Tk()
> 
> button = Tkinter.Menubutton(text = 'File', underline = 0)
> button.pack()
> menu = Tkinter.Menu(button)
> button.configure(menu = menu)
> menu.add_command(label = 'Hello World', underline = 0)
> 
> root.mainloop()
> ==================================================================


  It works for me on Linux.  Hitting alt-F selects the File menu. 
Hitting alt-H doesn't do anything visible, I presume because the command
isn't bound to anything.  I'll try it on my NT box at work tomorrow if
you'd like.

 Joshua


From donb at init.com  Thu Apr 20 10:26:49 2000
From: donb at init.com (Donald Beaudry)
Date: Thu, 20 Apr 2000 10:26:49 -0400
Subject: globals: a dirty little secret? 
References: <Pine.BSF.4.10.10004200321050.29174-100000@chi.pair.com>
Message-ID: <200004201426.KAA16104@zippy.init.com>

"Michal Wallace (sabren)" <sabren at manifestation.com> wrote,
> 
> 
> Globals have come up a couple times recently.. If someone posted the
> following approach, I missed it...
> 
> 
> ### step 1: globals.py #######################
> 
> import __builtin__
> 
> def setGlobal(key, value):
>    __builtin__.__dict__[key] = value
> 
> __builtin__.setGlobal = setGlobal # makes the above function global

*If* I found that functionality useful, I'd prefer

    def setGlobal(**d):
        __builtin__.__dict__.update(d)

so that it could be called like:

    setGlobal(x=5)

which fits better with my sense of aesthetics.

Sure it works, but what good does it do you?  It only provides
"convenient" read-only access to your globals.  To write a global, you
must use the setGlobal() function.  Instead, consider this module:

    #
    # globals.py
    #

    pass

Yup... that's it.

Now inside your program:

    import globals
    globals.x = 5

    print globals.x

With this approach your intentions become very clear.

--
Donald Beaudry                                     Ab Initio Software Corp.
                                                   201 Spring Street
donb at init.com                                      Lexington, MA 02421
                      ...Will hack for sushi...



From pickaxe22 at stny.rr.com  Sun Apr 23 14:33:14 2000
From: pickaxe22 at stny.rr.com (Andy Mullen)
Date: Sun, 23 Apr 2000 18:33:14 GMT
Subject: setting up Tkinter
Message-ID: <KlHM4.8299$4w1.119156@typhoon.nyroc.rr.com>

As i understand from the website (www.python.com), this program doesn't run
on windows 98. Since i was already thinking of putting linux on my machine
under a partition of the HD, i was wondering if someone could point me in
the right direction as to where to get linux online, how to go about
install, and how to run the 2 separate OS's on one machine.  any response
would be appreciated




From zorro at zipzap.ch  Sat Apr 15 04:22:39 2000
From: zorro at zipzap.ch (Boris Borcic)
Date: Sat, 15 Apr 2000 10:22:39 +0200
Subject: [ANNOUNCE] Garbage collection for Python
References: <20000407221750.A29279@acs.ucalgary.ca>
	 <m3r9cagx3c.fsf@atrus.jesus.cam.ac.uk> <l03102809b51d41895a80@[193.78.237.135]>
Message-ID: <38F826CF.DBDC71EF@zipzap.ch>

Just van Rossum wrote:
> 
> - Will Ware and I have worked hard on a complete rewrite of the
>   uthread module. It is a *vast* improvement over the old one.
>   It offers a very nice thread object and several handy threading
>   widgets. It already can do stuff threading.py can only *dream*
>   about, like pausing (blocking) a thread from another thread or
>   posting an exception from another thread (!).

Simple examples of useful new idioms that this permits ?

BTW, I remember there exists a package for lisp, called
"screamer", that turns it in a backtracking non-deterministic
language in the style of prolog, through a device called
CPS (continuation-passing style) transformation. While
python lacks S-expressions and lisp-style macros that
allow this for lisp, continuations are clearly the key
component of this gimmick. It might be worth its while
to have a better look.

There is also Oz. http://www.mozart-oz.org/ - I've long
dreamt of a language with some of the semantics of Oz
and Python syntax and libs. continuations and uthreads
seem to bring python somewhat closer to Oz.

BB
--
"Hope achieves the square root of the impossible"


From paulb at infercor.no  Mon Apr 10 08:28:22 2000
From: paulb at infercor.no (Paul Boddie)
Date: Mon, 10 Apr 2000 14:28:22 +0200
Subject: opening a serial port at a given baudrate. How?
References: <XFMail.20000404190839.warlock@eskimo.com>
Message-ID: <38F1C8E6.73E527DB@infercor.no>

Jim Richardson wrote:
> 

[...]

> I have been using open('/dev/ttyS0','r+')
> but need to control baud rate. fcntl.ioctl() will do this?
>  I can't seem to find (on machine, haven't checked on-line yet
> how to do this, but thanks for the pointer! Knowing where to look is at
> least half the battle.

Try the Linux Documentation Project at http://www.linuxdoc.org - the
Serial-HOWTO is probably what you want. You may have this document somewhere in
/usr/doc, depending on your Linux distribution.

Paul


From mikael at isy.liu.se  Fri Apr  7 08:57:38 2000
From: mikael at isy.liu.se (Mikael Olofsson)
Date: Fri, 07 Apr 2000 14:57:38 +0200 (MET DST)
Subject: [Python-Dev] Round Bug in Python 1.6?
In-Reply-To: <20000407084914.A13606@mems-exchange.org>
Message-ID: <XFMail.000407145738.mikael@isy.liu.se>

On 07-Apr-00 Greg Ward wrote:
 >  Oh, joy! oh happiness! someday soon, I may be able to type
 >  "blah.__doc__" at the interactive prompt and get a readable result!

Just i case... I hope you haven't missed "print blah.__doc__".

/Mikael

-----------------------------------------------------------------------
E-Mail:  Mikael Olofsson <mikael at isy.liu.se>
WWW:     http://www.dtr.isy.liu.se/dtr/staff/mikael
Phone:   +46 - (0)13 - 28 1343
Telefax: +46 - (0)13 - 28 1339
Date:    07-Apr-00
Time:    14:56:52

This message was sent by XF-Mail.
-----------------------------------------------------------------------



From jamarijr at hotmail.com  Wed Apr  5 09:37:28 2000
From: jamarijr at hotmail.com (Arinté)
Date: Wed, 05 Apr 2000 13:37:28 GMT
Subject: Exceptions, when and who?
Message-ID: <8cffin$pl0$1@nnrp1.deja.com>

When and why would python could throw an exception in an embedded app?
I am not getting the exception on the c side but in the python code and
I can't figure out why.

If I set the Error String in the c code when and who should clear that
error?  How would you clear it in python, in c I know you use
PyErr_Clear (right?).


Sent via Deja.com http://www.deja.com/
Before you buy.


From mwh21 at cam.ac.uk  Wed Apr 19 11:46:36 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 19 Apr 2000 16:46:36 +0100
Subject: Very Horrible Question About Goto's
References: <C5A7720BDEB7D3118A5200E018C200A037EA43@EXCHANGE> <m3g0sihz9h.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <m3d7nmhz4z.fsf@atrus.jesus.cam.ac.uk>

Michael Hudson <mwh21 at cam.ac.uk> writes:

> Gaetan Corneau <corg at copernic.com> writes:
> 
> > I was surprised to learn that Python doesn't have a goto statement.
> > Would it be difficult to add?
> 
> No. The necessary opcodes are there (well, the one you need:
> JUMP_ABSOLUTE).  You could use bytecodehacks to implement it, if you
> liked...

Oh, hang on: the block stack would cause problems; how would you get

def f(x):
    label: bob
    for i in x:
        if x: 
            goto: bob

to work? (by declaring it to be an error, I suppose).

Cheers,
M.

-- 
41. Some programming languages manage to absorb change, but 
    withstand progress.
     -- Alan Perlis, http://www.cs.yale.edu/~perlis-alan/quotes.html


From psl at mitre.org  Wed Apr 12 15:27:56 2000
From: psl at mitre.org (Pete)
Date: Wed, 12 Apr 2000 15:27:56 -0400
Subject: scrolledListBox problem
Message-ID: <38F4CE3C.3D6E7A68@mitre.org>

I am using makeing a GUI that contains 2 scrolledListBoxes,  and I have
come across something unfortunate.  When I select an item in one list, 
the currently selected item in the other list becomes deselected,  so
only one list can have an item selected at any point in time.  Is this a
bug in PMW?  Any ideas as to how to fix this?


From jsaltiel at uiuc.edu  Sat Apr 29 17:11:56 2000
From: jsaltiel at uiuc.edu (Joe Saltiel)
Date: Sat, 29 Apr 2000 16:11:56 -0500
Subject: converting word documents to html
Message-ID: <390B501C.1891CB80@uiuc.edu>

Hey,

    I was wondering if anybody had done any work on converting
word processors type documents to html.  (word or word perfect,
pdf or anything else).  I am looking for a nice way to post
documents online that are laden with footnotes/endnotes.
Word2000 does an allright job, but adds a bunch of
junk that makes the file 3-4 larger then it has to be.
I was thinking of writing something to do that but just want
to check before I reinvent the wheel.....

thanks

Joe




From mak at imakhno.freeserve.co.uk  Sun Apr 23 13:28:37 2000
From: mak at imakhno.freeserve.co.uk (Makhno)
Date: Sun, 23 Apr 2000 18:28:37 +0100
Subject: Calling Python from C code with subs as strings
Message-ID: <8dvc19$mun$1@news7.svr.pol.co.uk>

Can somebody suggest why the following C/Python code does not work? It's
supposed to call a C function with a string argument that describes what
Python function to call. That C function then calls that Python function.

C:

void CallPythonSub(char *subname)
{
   if (subname[0]==0) return; //do check this....
   PyObject *moduleref=PyImport_ImportModule("__main__");
   PyObject *thesub=PyObject_GetAttrString(moduleref,subname);
   if (thesub==NULL) {printf("Error\n");return;}
   PyObject_CallFunction(thesub);
}



Python:

def thissub():
   print 'im working!'

CallPythonSub('thissub')



No Nulls are returned from the Py... calls. It just does nothing.




From rich.glazerman at synchrologic.com  Wed Apr 12 15:24:51 2000
From: rich.glazerman at synchrologic.com (Rich Glazerman)
Date: Wed, 12 Apr 2000 15:24:51 -0400
Subject: Config Parser Options Case Sensitivity
Message-ID: <cX3J4.4121$Uz6.5404@newsfeed.slurp.net>

I've been having a problem with ConfigParser and option names. From what I
see, ConfigParser reads in optionnames from the file, and in an attempt to
keep optionnames case insensitive, sets them to lowercase before using them
as keys in a dictionary. The problem I have with this is that in the config
file I'm using, we use some OptionNames that are case sensitive in terms of
needing to remain in the same case as when they are read. What I've done to
fix this is as follows:

I've added self.__options dictionary to the ConfigParser Class.

In the __read() function, before converting optionname to lowercase, I
append the optionname to the list (self.__options[sectname]).

Now, self.options(section) returns self.__options[section] instead of the
keys inside of self.__sections[section]. This way, self.options(section)
returns the options in the same case in which they are found in the file,
while keeping case insensitivity for self.__get.

Does anybody have any other ideas on this? Is this something anybody else
has run into? Does this seem like a valid solution?

Thanks for the feedback.

Rich
--
Rich Glazerman
Test Engineer
Synchrologic, Inc.

rich.glazerman at synchrologic.com
770-754-5600x176




From sctheagn at kraken.itc.gu.edu.au  Tue Apr 18 22:29:49 2000
From: sctheagn at kraken.itc.gu.edu.au (Joal Heagney)
Date: Wed, 19 Apr 2000 12:29:49 +1000
Subject: module naming?
References: <Pine.BSF.4.10.10004162056040.25069-100000@chi.pair.com>
Message-ID: <38FD1A1D.D59B594B@kraken.itc.gu.edu.au>

"Michal Wallace (sabren)" wrote:

> On Mon, 17 Apr 2000, David Arnold wrote:
>
> > -->"Michal" == sabren  <Michal> writes:
> >
> >   Michal> Is there a better way to do this?
> >
> > what about something like
> >
> >    from Record import Record
> >    from Field import Field
> >    from IdxDict import IdxDict
> >
> > in your zdc/__init__.py ?
> >
>
> Hey David,
>
>   That's certainly a lot cleaner! But is there a way to do that and
> not have it load each one of those files every time? There's actually
> more than 3, and even though they're all fairly small, I'd rather not
> have the overhead of loading them until they're needed.
>
> Cheers,
>
> - Michal
> -------------------------------------------------------------------------
> http://www.manifestation.com/         http://www.linkwatcher.com/metalog/
> -------------------------------------------------------------------------

I'm new to the pythoning game, but is there a way to check to see if a file
is loaded. I seem to remember that the from statement doesn't support that,
but maybe you could make a call to a function that does a check on if the
class already exists in program space, and if it doesn't imports it using
the from ... import statement?

Joal Heagney/AncientHart



From python-list at teleo.net  Mon Apr 17 12:59:59 2000
From: python-list at teleo.net (Patrick Phalen)
Date: Mon, 17 Apr 2000 09:59:59 -0700
Subject: guestbook
In-Reply-To: <20000417174349.71830.qmail@hotmail.com>
References: <20000417174349.71830.qmail@hotmail.com>
Message-ID: <00041710080906.00880@quadra.teleo.net>

[sigma seitse, on Mon, 17 Apr 2000]
:: i'm trying to write a guestbook in python. well,
:: it should consist additional feature like this:
:: it should enable the owner to change and response to visitors' comments and 
:: show it out.
:: have you got anything like this or maybe you know an address or smth, where 
:: I could find something
:: like this?

An excellent tutorial on this can be found at:

http://classic.zope.org:8080/Documentation/Reference/ObjectPublishingIntro

The examples used there are based on integrating a Python guestbook
module with the Zope ORB, Templates and Object Database, which has
advantages over relying on low level CGI.



From markus.oberhumer at jk.uni-linz.ac.at  Thu Apr 27 20:46:05 2000
From: markus.oberhumer at jk.uni-linz.ac.at (Markus F.X.J. Oberhumer)
Date: Fri, 28 Apr 2000 02:46:05 +0200
Subject: [ANNOUNCE] PySol 4.20 - a solitaire game collection
Message-ID: <20000428024605.A31501@laetitia.oberhumer.com>

-----BEGIN PGP SIGNED MESSAGE-----


                    PySol - a Solitaire Game Collection
                              Version 4.20

                          http://pysol.tsx.org

           Copyright (C) 1998, 1999, 2000 Markus F.X.J. Oberhumer
                   <markus.oberhumer at jk.uni-linz.ac.at>


What is PySol ?
===============

PySol is an exciting collection of 287 solitaire games.

Among the supported games are classics like Aces Up, Baker's Game,
Canfield, FreeCell, Forty Thieves, Golf, Klondike, Monte Carlo,
Osmosis, Pyramid, Scorpion, Spider, Yukon, and many more...

PySol may be freely distributed under the terms of the GNU GPL.

PySol aims to be Commercial Quality Freeware.


Why yet another solitaire game ?
================================

Here are some highlights of PySol:

  - freely available
  - currently supports 287 (!) distinct solitaire variants
  - lots of classic games like Forty Thieves, FreeCell, Klondike and Spider
  - special games like Hanafuda, Mahjongg, Poker and Tarock type games
  - based upon an extensible solitaire engine
  - very nice look and feel including multiple cardsets
    and background table tiles
  - sound samples and background music
  - unlimited undo & redo
  - load & save games
  - player statistics and log files
  - hint system and demo games
  - support for user written plug-ins - add your own solitaire variants
  - integrated HTML help browser
  - lots of documentation
  - fully portable across Unix/X11, Windows 95/98/2000/NT and MacOS
  - distributed under the GNU GPL with full source code
  - written in 100% pure Python
  - just run it - no need to compile anything


Nice. And what's new ?
======================

  * 14 new games, including HexADeck and Memory type variants

  * some new great-looking cardsets

  * much improved statistics dialog

  * added a "Recent games" menubar entry

  * fixed the detection of Straights in Poker type games

  * fixed a problem when changing the card background


Cool. Where can I get it ?
==========================

Point your browser to http://pysol.tsx.org
The PySol Gallery is awaiting your visit as well.


What do I need to start playing ?
=================================

PySol requires Python 1.5.2 and Tcl/Tk 8.0.5. Both packages are
freely available for Unix, Windows 95/98/NT and Macintosh.

BTW, there is no need to compile anything since the whole program is just
a Python script. Just run it, and that's all.

The Windows version ships as a completely self-contained setup file.


License terms
=============

PySol is Copyright (C) 1998, 1999, 2000 Markus Franz Xaver Johannes Oberhumer

PySol is distributed under the terms of the GNU General Public License (GPL).
See the file COPYING.


Have fun,
Markus

http://pysol.tsx.org



-----BEGIN PGP SIGNATURE-----
Version: 2.6.3ia
Charset: noconv

iQCVAwUBOQjO5G10fyLu8beJAQGN5AP/fSUT/67iJFnk3nz0Y0IbmmIl+/qzk8AL
mIaGipsALcBaUsJZ5H0r+TR9YMEpSeIbx6FBNuKQyIuC29PoBKX8UJ8KH5bhe3Wp
EfJxIXYihCdZtQooWs3y/fJhYu9JvPpqPlnlCC9cBpnV7eKdq2GUjleS3TJhUHdp
T7I0DjFZQc0=
=OfHV
-----END PGP SIGNATURE-----



From jae at ilk.de  Mon Apr 24 17:57:47 2000
From: jae at ilk.de (Juergen A. Erhard)
Date: Mon, 24 Apr 2000 23:57:47 +0200
Subject: The Simple Economics of Open Source
In-Reply-To: <LHEoAJAXWFB5EwF6@jessikat.demon.co.uk> (message from Robin
	Becker on Mon, 24 Apr 2000 15:09:27 +0100)
References: <B527E7A5.51AA%raffael@mediaone.net> <3903316C.5D6F1611@tcac.net>
 <raffael-34658F.23400223042000@news.ne.mediaone.net>
 <390451C0.4B5D3196@tcac.net> <LHEoAJAXWFB5EwF6@jessikat.demon.co.uk>
Message-ID: <24042000.3@sanctum.jae.ddns.org>

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

>>>>> "Robin" == Robin Becker <robin at jessikat.demon.co.uk> writes:

    Robin> In article <390451C0.4B5D3196 at tcac.net>, Albert Wagner
    Robin> <alwagner at tcac.net> writes
    >> You STILL don't understand.  Artifice can only copy.  It cannot invent.
    >> An end product can be copied, even a style can be copied, but not the
    >> creative process itself.
    >> 
    >> Raffael Cavallaro wrote:
    Robin> ...

    Robin> This position is like the weak AI position; you assert
    Robin> there's something magical about creation. I think A Turing
    Robin> was right; any human behaviour can be simulated and the
    Robin> best simulators (people) are around already.

Fundamental question is: Are we just a complex arrangement of
bio-chemical (and thus ultimately physical) processes... or is there
more to being `human beings'?

If we are just a collection of bio-chemical processes, it's simple to
`simulate' or replicate us in machine form[1].

If we are more... than it will not be possible.

Since I'm a christian, I guess it's obvious where I'll put my bets...

Bye, J

[1] I'm just reading Bill Joy's Wired essay `Why the future doesn't
need us'...

PS: For the record, I'm 100% with Albert (though maybe he's not 100%
with me ;-)

- -- 
J?rgen A. Erhard      eMail: jae at ilk.de      phone: (GERMANY) 0721 27326
     MARS: http://members.tripod.com/Juergen_Erhard/mars_index.html
                    "Ever wonder why the SAME PEOPLE
      make up ALL the conspiracy theories?" -- Michael K. Johnson
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.1 (GNU/Linux)
Comment: Use Mailcrypt and GnuPG <http://www.gnupg.org/>

iEYEARECAAYFAjkEw1oACgkQN0B+CS56qs09zQCfTMfxscQZVhcLpqgIiSVYShoO
vjwAn0G8CdZImNdfPk1s5mvgOjR7GxVu
=W73r
-----END PGP SIGNATURE-----



From effbot at telia.com  Wed Apr 12 10:08:15 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Wed, 12 Apr 2000 14:08:15 GMT
Subject: comparing strings and ints
References: <2YmI4.366$rc9.188014080@newsb.telia.net><20000411134358.A2352304@vislab.epa.gov> <1256641282-42160637@hypernet.com> <20000412075812.A2449004@vislab.epa.gov>
Message-ID: <jr%I4.643$rc9.190628352@newsb.telia.net>

Randall Hopper <aa8vb at yahoo.com> wrote
> Gordon McMillan:
>  |Randall Hopper wrote:
>  |> Fredrik Lundh:
>  |>  |
>  |>  |    Objects of different types always compare unequal, and are
>  |>  |    ordered consistently but arbitrarily.
>  |>
>  |> I can see 45 == '45' being false, and 45 != '45' being true.  But it
seems
>  |> to me that not throwing an exception for attempts at an ordered
comparison
>  |> with logical ordering operators (<, <=, etc.) on objects of different
>  |> primitive types (e.g. 45 < '45', 45 > '45') only lets bugs pass
through.
>  |>
>  |> Is there a case where this could be useful (with its undefined
behavior)?
>  |> I believe that is the root of the question.
>  |
>  |I'd happily bet your wife and firstborn child that there's code that
>  |relies on it <wink>, if only because after sorting a list, all the ints
>  |will end up in a clump, and all the strings in another, etc.
>
>
> Oh, I also have __no doubt__ that someone is using this unofficial,
> undocumented behavior.  But they don't have a leg to stand on if the
> ordering breaks in a future Python version ;-)

interestingly enough, you included my documentation quote in
a post claiming that this is an undocumented feature...

(if you want more background on *why* this design was chosen,
consider that python has a single comparision primitive __cmp__,
and that until a few versions ago, the interpreter happily ignored
exceptions raised by that primitive.  for more info on this topic,
search the site/newsgroup for "rich comparisions").

</F>

<!-- (the eff-bot guide to) the standard python library:
http://www.pythonware.com/people/fredrik/librarybook.htm
-->




From rcunning at acm.org  Sun Apr  9 22:13:41 2000
From: rcunning at acm.org (Robert W. Cunningham)
Date: Mon, 10 Apr 2000 02:13:41 GMT
Subject: Python paradigms
References: <Pine.LNX.4.21.0004091636070.13883-100000@fep132.fep.ru>
Message-ID: <38F138CD.38983629@acm.org>

Oleg Broytmann wrote:

> On Sun, 9 Apr 2000, Fredrik Lundh wrote:
> > Oleg Broytmann wrote:
> > >    blah = ((foo() and [poo()]) or [boo()])[0]
> > >
> > >    In C, "foo() ? poo() : boo()" will not call boo() if foo() returns
> > true;
> > > but in Python, you'll have a side-effect of calling boo() even if foo()
> > > returns true - what is wrong, usually.
> >
> > did you try running the code?
> >
> > python's "and" and "or" operators doesn't evaluate the
> > second argument unless they really have to...
>
>    Yes, I missed that, sorry.

So, it works just fine, right?

>    Anyway, it is strange (not to say "ugly") construct; I prefer not to see
> it in a code...

I remember a professor who had a mantra he'd repeat just before accepting
programming assignments from his students:

  "All code without comments is UGLY code.  And ugly code gets an 'F', even if it
works perfectly."

Which had an implicit corollary:  "Nasty Code + Comments = Non-Ugly Code"

Not all problems have "nice" solutions, but even the messiest solution can be
made palatable with the addition of suitable comments.

The best example of this I can recall was in an extremely obfuscated MD5 hash
generator that fit in something like two or three lines of code.  The comments
required two pages (to explain both the MD5 algorithm and this implementation),
and did their job quite well.

The "self documenting" version of the code, on the other hand, required only one
page, but even with maximum compiler optimizations it ran 4 times slower than the
obfuscated version.

I'm not saying either of these were ideal pieces of code.  But when the emphasis
is on performance AND maintainability, then nasty code combined with sufficient
comments works extremely well.  When the emphasis is on "readable" or "self
documenting" code, it can be much harder to obtain suitable performance.  It has
also been my experience that performing structural optimizations to code is
somewhat easier when the code is more compact and well documented, than when it
is "self-documented" and spread over a greater area.

In Python, performance quickly slows to a crawl unless a handful of trusty
structural optimizations are used, such as assigning methods to variables before
calling them repeatedly within a loop.  This clearly makes the code less "self
documenting", since it imposes a structural optimization to circumvent a basic
behavior of the Python execution environment.  Surely, someday the Python
interpreter will be able to create such optimizations for us on the fly (just a
little loop invariant migration, right?), but in the meantime these manual
optimization constructs are so prevalent that they have almost become part of the
Python language.  Indeed, they are documented as "necessary" by GvR himself!

I see each and every one of them as a wart on the basic elegance of the Python
language, since they generally serve only to manually optimize (circumvent)
access to namespaces (or other Python "features"), to "break" the most basic
elements of OO implementation!

It is getting to the point that these optimizations and constructs aren't even
documented in most of the code I've looked at, and I'm still a Python newbie.  To
me, even some "guru" code is ugly, ugly, ugly.  And such code is present in the
documents that define the Python language and its use.

So, when I hear Python gurus arguing about the "ugliness" of various coding
constructs, it makes me laugh out loud.  The common use of Python *demands* hand
optimization (aka "uglification") of the code to gain decent performance.  As the
code becomes faster and more usable, it gets uglier and uglier.  Maybe that's too
strong a statement:  IMHO, the "prettiest" "self documenting" Python code is
often unusably slow, especially where loops are involved.

The argument in this case should be about the degree of commenting needed, not
the code itself!  AFAIK, the code is FINE (though possibly not "perfect").  Has
it been tested for performance and side effects against other forms?  Until we
know better, it may be the case that the code we've been discussing to create
something equivalent to C's "?:" construct is OPTIMAL!

Let's assume, for the moment, that this is indeed the case:  Now, obviously, the
only thing left to discuss is the comments!  Right?  Who would toss fast code
merely because it was "ugly"?  Certainly no Python programmer!

And what IS the best way to comment such an obtuse line of code?  I submit my
example that's quoted at the top.  If you understand the "?:" construct from C,
it does the job perfectly (if minimally).  No matter how "ugly" the related
Python code is.

Python is, for now, condemned to use "ugly" code constructs if it is to see
maximum use and performance.  Let's use them when we must, but also pay strict
attention to thoroughly documenting them whenever they are used.

Which brings me to a related thought:  Has anyone looked at post-processing .pyc
files to optimize them?  Or are there plans to include a bytecode optimizer in
the Python interpreter or runtime environment (as a JIT optimizer/compler)?
Without, of course, bringing the entire JVM or JRE into the game...  That might
be overkill.  Any such moves should serve to reduce or eliminate the *need* for
much of the present ugly Python code.


Oh-to-have-loop-invariant-migration-ly yours,

-BobC




From robin at jessikat.demon.co.uk  Sun Apr 23 13:43:12 2000
From: robin at jessikat.demon.co.uk (Robin Becker)
Date: Sun, 23 Apr 2000 18:43:12 +0100
Subject: The Simple Economics of Open Source
References: <B527E7A5.51AA%raffael@mediaone.net> <3903316C.5D6F1611@tcac.net>
Message-ID: <KHRTGIAwYzA5Ewtz@jessikat.demon.co.uk>

In article <3903316C.5D6F1611 at tcac.net>, Albert Wagner
<alwagner at tcac.net> writes
>Raffael Cavallaro wrote:
>> 
>> 
>> Sarcasm aside, I think you miss the point. Those artists least willing to
>> share, most jealous of their "secrets" have always been those who were most
>> successful. Unsucessful young artists are, by and large, quite willing to
>> share ideas, working techniques, etc. It's actually the more established,
>> financially succesful ones, who are not.
>> 
>
>I think you confuse artifice and art.  Talent is not a "secret" that can
>be hoarded or shared.  It is visible in the end product, yet no amount
>of sharing how it was done will allow the process to be duplicated.  I
>think you also are confused about what "successful" means to a talented
>person.  The "commodification" (sorry, Marx) of anything and everything,
>assigning it a dollar value, is at the very root of the mindset that is
>destroying the world. 

I concur wholeheartedly
-- 
Robin Becker


From tom_hines at my-deja.com  Fri Apr 28 16:46:04 2000
From: tom_hines at my-deja.com (Tom Hines)
Date: Fri, 28 Apr 2000 20:46:04 GMT
Subject: win32com: _NewEnum being cached
Message-ID: <8ecta7$nn5$1@nnrp1.deja.com>

Hello.  As you know, to implement "for x in my_list" functionality in
COM, you must code a get__NewEnum() function, which I have done and it
works fine under Visual Basic.  But when called from Python, the call
seems to get cached, as subsequent Python "for" blocks don't result in
calls to get__NewEnum().

I have a COM class that implements a dynamic collection, so I
need "for" blocks to result in calls to get__NewEnum() every time.
Does anybody know how to force this?

Thanks.

--
Tom Hines
Rockville, MD



Sent via Deja.com http://www.deja.com/
Before you buy.


From ivanlan at callware.com  Thu Apr  6 08:48:01 2000
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Thu, 06 Apr 2000 06:48:01 -0600
Subject: Teaching the art of programming, in python
References: <20000405053920.20579.qmail@web2102.mail.yahoo.com>
	    <8cfe42$o02$1@nnrp1.deja.com> <dpJG4.3744$HG1.106895@nnrp1.uunet.ca>
	    <38EB68A3.B1B5809F@callware.com>
	    <8cfqa7$4ml$3@newsserver.rrzn.uni-hannover.de>
	    <38EB73B6.6C0913A7@callware.com> <8chsij$k5r$1@newsserver.rrzn.uni-hannover.de>
Message-ID: <38EC8781.60AA68F9@callware.com>

Hi All--

Bernhard Reiter wrote:
> 
[snip]

> >> Ranting-about-dummy-booktitles-ly-yours,
> Well I am just ranting about book title giving wrong impressions.
> 24 times a one hours session can be quite a bit to teach people the
> basics, I agree with that. But for all the philosophy I guess that
> you need to spread the lectures about a larger time period. :)
> 

Unfortunately, I had absolutely no choice about the book title.  Given
my druthers, I would have used what the book is *really* named, _Python
Mind, Beginner's Mind_.  Think of it that way, and it's a lot easier to
take;-)

<the-python-that-can-be-spoken-is-not-the-true-python>-ly y'rs,
Ivan;-)
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
http://www.pauahtun.org and
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
Author:  Teach Yourself Python in 24 Hours




From nmm1 at cus.cam.ac.uk  Tue Apr  4 17:42:38 2000
From: nmm1 at cus.cam.ac.uk (Nick Maclaren)
Date: 4 Apr 2000 21:42:38 GMT
Subject: Warning switches (was: Python 1.6 alpha 1 released)
References: <Pine.GSO.4.10.10004040035420.21084-100000@sundial> <5l7leern5v.fsf@eric.cnri.reston.va.us> <38ea1de0$0$17688@personalnews.de.uu.net> <nctG4.5133$74.83090@newsc.telia.net>
Message-ID: <8cdnke$b05$1@pegasus.csx.cam.ac.uk>

In article <nctG4.5133$74.83090 at newsc.telia.net>,
Fredrik Lundh <effbot at telia.com> wrote:
>Johannes Stezenbach <yawyi at gmx.de> wrote:
>> >However, I'll do so for purely backwards compatible reasons, and I
>> >will continue to press for the tuple address as the correct form.
>>
>> Will python then support a -w switch to support people in writing
>> correct programs (i.e. compatible with likely furure language/library
>> changes), just like our-favourite-four-letter-P-language?
>
>how about -v<version> ?
>
>there has been talks about compatibility "modes" (probably an option,
>but probably also a way to set options programmatically), as a way to
>help people prepare for "Python 3000".  I think we might expect some
>of it in 1.7, and more of it in 1.8 (if that ever materializes).
>
>but hopefully, Python 3000 will be completely free from modes.

It is important to distinguish modes from warning levels.  The
former imply that there are several subtly different languages,
and are usually regarded as snares and delusions.  The latter
are best described by the following question:

    Do you want me to tell you when I think that you may be
making a mistake, when you are doing something that is asking
for trouble, or only when you have definitely got it wrong?"


Regards,
Nick Maclaren,
University of Cambridge Computing Service,
New Museums Site, Pembroke Street, Cambridge CB2 3QG, England.
Email:  nmm1 at cam.ac.uk
Tel.:  +44 1223 334761    Fax:  +44 1223 334679


From ns56645 at swt.edu  Sun Apr 30 13:50:35 2000
From: ns56645 at swt.edu (ns56645)
Date: Sun, 30 Apr 2000 12:50:35 -0500
Subject: ScrollCanvas Help Needed?
Message-ID: <390C726B.A01B6471@swt.edu>

Hi Folks,

I am trying to implement scrollcanvas widget of Pmw widget in the
frames. I can create those canvas using grid but the scrollbar doesn't
scroll the canvas instead it scrolls the button inside. I went through
the sample code and they have used pack function instead of grid. Does
that make any difference. Let me know how can I make the canvas scroll
and not the buttons inside. I hope you understand me. If I use pack
instead of grid the canvas widget doesn't show on the frame.

Please let me know if I need to send the code to you guys.

Shah
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20000430/ab8a4373/attachment.html>

From dalke at acm.org  Sun Apr 23 15:27:55 2000
From: dalke at acm.org (Andrew Dalke)
Date: Sun, 23 Apr 2000 13:27:55 -0600
Subject: partial parsing?
Message-ID: <8dvilp$88s$1@slb6.atl.mindspring.net>

Hello,

  I'm interested in writing a parser generator which is somewhat
different than yacc/flex, SPARK, Plex, etc.  Those parsers generate
code which verifies every byte in the file and expects that I'm
interested in most of the data.

  In my case, I have a lot of data (>100MB) in a known good format
of which I'm only interested in a few items, but the specific items
can change.

  What people usually do is hand write a parser which extracts only
the specified fields.  I've written parsers which parse every field
and use a callback event so clients can listen only to fields they
want, but the performance suffers.

  Instead, I want to start with the full grammer and specify both
which events I'm interested in getting and a stringency criterion.
A low stringency could potentially skip 200 bytes, or read 5 lines,
without actually checking to see if the data was correct.

  I've been looking for existing parser generators which work like
this, to use as a basis for my understanding, but I can't seem to
find anything this this.  All the systems I've seen assume you want
everything except whitespace.

  Can anyone here provide pointers to relevant programs or papers?

                    Andrew
                    dalke at acm.org





From greg at cosc.canterbury.ac.nz  Thu Apr  6 22:06:10 2000
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Fri, 07 Apr 2000 14:06:10 +1200
Subject: Tuples -- who needs 'em
References: <00e101bf9d80$a8384d20$74eb0b18@stcla1.sfba.home.com> <va3G4.1110$n68.190454272@newsb.telia.net> <037f01bf9e55$6d328760$74eb0b18@stcla1.sfba.home.com> <8cfa75$fho$1@newshost.accu.uu.nl>
Message-ID: <38ED4292.2EE8D3CA@cosc.canterbury.ac.nz>

Martijn Faassen wrote:
> 
> def foo():
>     return 1, 2, 3, 4
> 
> first, second, third, fourth = foo()
> 
> I think this is nice syntax.

It is very nice. Just a pity it loses its beauty in 
the case of 1-tuples.

Haskell avoids this problem by simply not having
any such thing as a 1-tuple. It works there because
you never need a 1-tuple in Haskell -- a bare value
always does just as well. 

Unfortunately, Python *does* need 1-tuples. If
only ASCII had more kinds of brackets, Guido might
have chosen differently. Maybe if Python 3000 uses
Unicode for source files...

-- 
Greg Ewing, Computer Science Dept,
+--------------------------------------+
University of Canterbury,	   | A citizen of NewZealandCorp, a	  |
Christchurch, New Zealand	   | wholly-owned subsidiary of USA Inc.  |
greg at cosc.canterbury.ac.nz	   +--------------------------------------+


From chalaoux at cybercable.fr  Wed Apr 26 16:57:07 2000
From: chalaoux at cybercable.fr (FR Chalaoux)
Date: Wed, 26 Apr 2000 22:57:07 +0200
Subject: Wanted Daniel Larsson
Message-ID: <39075823.F3511829@cybercable.fr>

Hi,

I looking for the right e-mail adress of "Daniel Larsson", the creator
of Pythondoc?
"Daniel.Larsson at vasteras.mail.telia.com" seems to be incorrect.

Bye, FR.







From cut_me_out at hotmail.com  Fri Apr 14 10:46:14 2000
From: cut_me_out at hotmail.com (Alex)
Date: 14 Apr 2000 10:46:14 -0400
Subject: Choral Music for Holy Week - Chanticleer_Mysteria- Gregorian Chants_10_Holy Sunday- Hymn- Vexilla Regis.mp3 (07/11), exception problem
References: <1pBJ4.29582$h6.134427@typhoon.jacksonville.mediaone.net>
Message-ID: <etd7le03fm1.fsf@w20-575-109.mit.edu>

> from uu import *
> decode(infile, outfile)
> 
> raise Error, 'No valid begin line found in input file'
> uu.Error: No valid begin line found in input file
> 
> 
> how do i trap it
> 
> i've tried
> 
> try:
>     decode(infile, outfile)
> except uu.error:
>     pass
> 
> gives me
> 
> NameError uu :

I'm not sure, but I think you've bought all of uu into your current
namespace.  If that's the case, you probably want

try:
    decode(infile, outfile)
except error:
    pass

...except that 'error' should probably be 'Error' judging by the
traceback.

On the face of it, it seems like you would be better off keeping uu in
its own namespace in this case:

import uu
try:
    uu.decode (infile, outfile)
except: uu.Error
    pass

Alex.


From jimsp at ichips.intel.com  Mon Apr 17 16:37:36 2000
From: jimsp at ichips.intel.com (jimsp)
Date: Mon, 17 Apr 2000 13:37:36 -0700
Subject: Another question
References: <dVFK4.41$vr2.5817@news.uswest.net> <38FB2EEB.A32AF88C@bam.com> <38FB339F.604364AD@bam.com> <200004171605.MAA04929@zippy.init.com>
Message-ID: <38FB7610.80EAF965@ichips.intel.com>

How about

>>> import random
>>> x = [0]*1000000
>>> y = map(lambda x: random.randint(1,10), x)
>>> len(x)
1000000
>>> len(y)
1000000
>>>
>>>
>>> x[1:100]
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0
, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0,
 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
>>> y[1:100]
[6, 5, 6, 8, 2, 4, 10, 7, 10, 5, 6, 8, 4, 3, 9, 10, 7, 9, 9, 5, 6, 7, 1, 3, 7,
2
, 7, 4, 7, 2, 6, 3, 9, 5, 4, 9, 2, 7, 2, 6, 5, 8, 7, 8, 1, 6, 10, 4, 3, 7, 3,
3,
 4, 8, 5, 1, 5, 7, 5, 9, 9, 5, 2, 3, 1, 8, 3, 5, 3, 4, 2, 9, 1, 6, 8, 8, 4, 6,
4
, 7, 2, 8, 5, 7, 2, 7, 9, 10, 7, 4, 8, 10, 6, 9, 4, 3, 3, 10, 8]
>>>

--james



Donald Beaudry wrote:

> Bill Scherer <scherbi at bam.com> wrote,
> > I just did a little test and found this to be substantially fatser than the
> > other three:
> >
> > import random
> > randint = random.randint
> >
> > r = []
> >
> > for i in xrange(1,10000):
> >     r.append(randint(1,10))
> >
> > for i in r: print i
> > #
>
> ...then this one should be really impressive ;)
>
>   def do_it():
>       from random import randint
>       r = []
>       append = r.append
>       for i in xrange(1,10000):
>          append(randint(i))
>       for i in r:
>          print i
>
>   do_it()
>
> or better yet:
>
>   def do_it():
>       from random import randint
>       r = []
>       append = r.append
>       map(append, xrange(1,10000))
>       for i in r:
>          print i
>
>   do_it()
>
> --
> Donald Beaudry                                     Ab Initio Software Corp.
>                                                    201 Spring Street
> donb at init.com                                      Lexington, MA 02421
>                   ...So much code, so little time...



From jvickroy at sec.noaa.gov  Tue Apr 11 11:40:14 2000
From: jvickroy at sec.noaa.gov (j vickroy)
Date: Tue, 11 Apr 2000 09:40:14 -0600
Subject: PythonCOM: com_error: (-2147418113, 'Catastrophic failure', None, None)
Message-ID: <38F3475E.C720D791@sec.noaa.gov>

Hello all,

I am attempting to drive the COM interface to RSI-IDL (Interactive Data
Language) from Python as follows:

>>> o = win32com.client.Dispatch("IDLDRAWX2.IDLDrawWidget.2")
>>> o.IdlPath = "c:\\rsi\\idl53\\bin\\bin.x86"
Traceback (innermost last):
  File "<interactive input>", line 1, in ?
  File "C:\Program Files\Python\win32com\client\dynamic.py", line 413,
in __setattr__
    self._oleobj_.Invoke(self._olerepr_.propMap[attr].dispid, 0,
pythoncom.DISPATCH_PROPERTYPUT, 0, value)
com_error: (-2147418113, 'Catastrophic failure', None, None)
>>>


The following variants were also tried with the same result.

o.IdlPath = "c:\rsi\idl53\bin\bin.x86"
o.IdlPath = "c:/rsi/idl53/bin/bin.x86"

I have been working with the vendor (RSI) but they do not have Python
expertise (and neither do I).

I would appreciate any suggestions.

Thanks.



From tom.mcdavid at boeing.com  Wed Apr  5 17:07:55 2000
From: tom.mcdavid at boeing.com (Tom McDavid)
Date: Wed, 5 Apr 2000 21:07:55 GMT
Subject: Python for QNX?
Message-ID: <38EBAB2B.BDA6D20F@boeing.com>

Has anyone ported python to qnx4?
Tom McDavid
Boeing - Phantom Works
Kent, WA


From akuchlin at mems-exchange.org  Tue Apr 18 13:21:01 2000
From: akuchlin at mems-exchange.org (Andrew M. Kuchling)
Date: 18 Apr 2000 13:21:01 -0400
Subject: Starship update: April 18
Message-ID: <3du2gzmiki.fsf@amarok.cnri.reston.va.us>

This is an update on the current status of starship.python.net; here
is bad news, very bad news, and a little bit of good news.

First, the bad news: The hard drive on starship is toast; it makes
nasty rattling noises when the power is turned on.

The very bad news is that there do not seem to be any backups, because
no one ever volunteered to set them up.  The most recent data we have
is the data that was on an older disk drive, which went out of service
back in 1998.

The good news: a new disk has been put into the machine, a clean
install of Linux done, and the machine should be online again
tomorrow.  Creating user accounts will take longer, of course.  These
newly created accounts will all have empty directories, since the data
hasn't been restored.  Setting up the Web server again will take
longer still.

A data recovery firm has been contacted to reconstruct the data from
the hard drive; their service costs around US$300-1000 if the data is
recovered successfully, and is free if nothing is recovered.  We'll
probably wind up asking for donations to cover the cost, if they
manage to recover the data.

In the meantime... 

   * If you can recover your starship data through other means
     (google.com's caching, your Web browser cache, copies on your own
     computer, etc.), please do so immediately, before the caches get
     expired.

   * If you had downloadable files on Starship (such as software)
     which you no longer have copies of, you might want to post to
     comp.lang.python or the appropriate mailing list and ask if
     anyone grabbed a copy recently.

   * If you downloaded someone else's software from Starship before
     the crash, DO NOT DELETE IT; your copy might be the only one
     left.  This is particularly true for packages that are no longer 
     under active development; you might consider e-mailing the author 
     to check if they still have a copy.

Further updates will be posted as events warrant.

-- 
A.M. Kuchling			http://starship.python.net/crew/amk/
Our sister died. But we never took revenge. We knew there'd be trouble one
day, her being mortal. It's not as if we never said anything. We still miss
her. We still mourn.
  -- Stheno, in SANDMAN #60: "The Kindly Ones:4"


From rcunning at acm.org  Tue Apr  4 11:40:33 2000
From: rcunning at acm.org (Robert W. Cunningham)
Date: Tue, 04 Apr 2000 15:40:33 GMT
Subject: From Python Newbie to Python Guru on the Fast Track:  HOW?
References: <38E9B48F.F1827543@acm.org> <38E9FD78.C740CF4E@gssec.bt.co.uk>
Message-ID: <38EA0C49.DD9BD052@acm.org>

Alan Gauld wrote:

> Finally consider a web based approach to what you want, that
> might be simpler albeit more retricted in the long term...
> It all looks mightily ambitious to me.

That's a fact!

My real goal is a bit less glorified than making Linux newbies more
comfortable with Windows apps on their Linux desktops, or grabbing some
of Windows' unique functionality for Linux.  No, what I really want to do
is see how much extra "stuff" I can get from the combination of Win4Lin
and Linux, with the hope of encouraging Trelos to include specialized
Python interfaces in Win4Lin.

I'd still like to do all the other stuff (one day), but I'm a realist:
There is no way a Python newbie such as myself can pull this off (this
year).  But I do feel I can produce some "proof of concept" examples that
may get a few of the *real* Python gurus interested in the project.  If
for no other reason than to clean up my code...


-BobC




From pj at sgi.com  Thu Apr  6 22:02:48 2000
From: pj at sgi.com (Paul Jackson)
Date: 7 Apr 2000 02:02:48 GMT
Subject: Python on Google?
References: <8F0ED22EBjoshtinamecom@209.208.0.7>
Message-ID: <8cjfk8$qtlo$1@fido.engr.sgi.com>

|> to what extent Python is used on www.google.com?

My vague impression is that quite a bit of the
original "interesting" code was done in Python, but
that most of the serious stuff now is being done
in C or C-like C++, due to the serious performance
challenges handling their current large and growing
load.

Perhaps some of the non-critical path stuff is still
in Python.
-- 

=======================================================================
I won't rest till it's the best ...	   Software Production Engineer
Paul Jackson (pj at sgi.com; pj at usa.net) 3x1373 http://sam.engr.sgi.com/pj


From gmcm at hypernet.com  Thu Apr  6 13:07:34 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Thu, 6 Apr 2000 13:07:34 -0400
Subject: Installer / mxDateTime problem
In-Reply-To: <1257094563-14896354@hypernet.com>
References: <8chm9o$qbv$1@gxsn.com>
Message-ID: <1257081637-15673599@hypernet.com>

I wrote:

> Richard Chamberlain wrote:
> 
> > I've got a simple script that uses mxDateTime that I want to run as an exe.
> > I run it through simple.py and I get the following error:
> > 
> > NameError: Case mismatch for module name feasts
> > (filename C:\Program Files\Python\DateTime\Feasts.py)
> 
> If I attempt to import DateTime.feasts interactively, I get the 
> same error. Do you have PYTHONCASEOK set? The exe isn't 
> going to do that for you.

And then I played with mxDateTime and I see what you're 
talking about. Aargh. You've run into a bug in Installer, 
(Windows case preserving case insensitive file names again), 
but after that you'll run into a perversion in mxDateTime (the 
fact the mxDateTime is both the name of a package and the 
name of a module, so the "from mxDateTime import *" 
succeeds even when it fails.) (Actually he does this with 
DateTime, too, but that one works).

All these "import *"s also include all the test code. I'm 
suprised that dependency tracking isn't fooled by LazyModule, 
considering what it does to "dir()".

- Gordon



From xaveer at dutetto.et.tudelft.nl  Wed Apr 26 15:44:07 2000
From: xaveer at dutetto.et.tudelft.nl (Xaveer Leijtens)
Date: 26 Apr 2000 19:44:07 GMT
Subject: Building python on HPUX-10.2
References: <38FE1392.D1C01D64@wag.caltech.edu>
Message-ID: <8e7gu7$ip$1@delphi.et.tudelft.nl>

You might want to try the HP-UX porting and archive center, which has
Python 1.5.2 prepackaged (and lots of other stuff):

  http://hpux.ced.tudelft.nl/hppd/hpux/Languages/python-1.5.2/

Good luck.
Xaveer

Richard P. Muller <rpm at wag.caltech.edu> wrote:
> I have successfully built python on a number of diffent platforms, but
> I'm having serious problems on HPUX 10.2. I go through the normal
> configure/make procedure, but the make crashes right away, when it is
> trying to link the first object (pgen), because it can't find the
> correct dl library (?).

> Has anyone successfully compiled Python on HPUX-10.*? If so, what
> configure string did you use?

> Rick



From pj at sgi.com  Wed Apr 12 17:16:42 2000
From: pj at sgi.com (Paul Jackson)
Date: 12 Apr 2000 21:16:42 GMT
Subject: Python/Perl Popularity (Re: A Mountain of Perl...)
References: <200004102111.HAA03380@envy.fulcrum.com.au> <8d0pcq$7ti$1@nnrp1.deja.com> <8d1inu$nsq$1@newshost.accu.uu.nl> <38F4996F.E3BD2626@acm.org>
Message-ID: <8d2p3q$2vj9o$1@fido.engr.sgi.com>

BobC wrote:
|> The text-mode browser named "links" 

I suspect that you meant "lynx" ??

-- 

=======================================================================
I won't rest till it's the best ...	   Software Production Engineer
Paul Jackson (pj at sgi.com; pj at usa.net) 3x1373 http://sam.engr.sgi.com/pj


From smalleys at gte.net  Fri Apr 14 00:05:27 2000
From: smalleys at gte.net (sue)
Date: Fri, 14 Apr 2000 04:05:27 GMT
Subject: Need an enum type extension
References: <38F52A7A.D8841724@gte.net> <JdeJ4.1552$Za1.25652@newsc.telia.net>
Message-ID: <38F68AF4.46DDA670@gte.net>

Will, that idiom does generate integers.

Frederic, I need some of the tuple/list functionality as
well.  I'm not sure what your code does.  I haven't done
lambdas except for number crunching.

In Ada, I can do things like:

type MyEnum is ( A, B, C );
for MyEnum use ( A : 3; B : 7; C : 13 );

Ada uses "'string" to refer to attributes.

MyEnum'len == 3.	(number of items in list)
A'next == B		(next item in list)
A'value == 3		(its low-level representation)
B'index = 1		(its position in the list)

But A == 3 fails.  only A == A passes.  A'value == 3 is
true.

Strings don't work directly because I need _some_ of the
list/tuple functionality.

I started with the mess below, but it can't put the EnumLits
in the current scope.  And I could see it getting ugly real
fast.

class EnumLit:
	count = 0
	def __init__( self, name ):
		self.name = name
		self.value = self.count
		self.repr = self.count
		self.count = self.count + 1
		self.lock = 0
	def set_repr( self, repr ):
		if not self.lock: self.repr = repr
	def finalize( self ):
		self.lock = 1
		self.count = 0
	def index( self ): return self.index	# etc

class Enum:
	def __init__( self, *args ):
		self.lock = 0
		self.names = args
		self.data = map( EnumLit, args )
		# or dict, or DictList?
	def set_repr( self, **args ):
		for (name, val) in args:
			if name in self.args:
				i = self.names.index(name)
				e = self.data[i]
				e.set_repr( val )
		self.finalize()
	def finalize( self ):
		self.lock = 1

and maybe some get/setattr stuff to finalize on any other
kind of access?

Sue.


From effbot at telia.com  Mon Apr  3 05:51:22 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Mon, 03 Apr 2000 09:51:22 GMT
Subject: Pass by reference ?
References: <38E864C9.E1EE94F7@ecs.soton.ac.uk>
Message-ID: <uQZF4.1085$n68.187094016@newsb.telia.net>

Jacek Generowicz <jmg at ecs.soton.ac.uk> wrote:
> def increment(x):
>     x = x + 1
>     return x
>
> a = 0
> print a
> print increment(a)
> print a
> ====================
>
> output:
>
> 0
> 1
> 0
>
> Given my understanding of passing by reference,
> the last number should be 1.
>
> What am I missing ?

two things, at least:

-- python variables are names that points to a value.

-- python's assignment statement doesn't modify values,
   it modify names.  "x = y" means "change 'x' so it points
   to the current value of 'y'", not "copy the value of 'y'
   into 'x'"

so in this case, "x = x + 1" means

    "calculate the sum of 'x' plus one, and change 'x' to point
    to that sum",

instead of

    "replace the value pointed to by 'x' with the sum"

the only way to modify a value in place is to call a method on it
(directly, or through syntactic sugar like member assignment)

</F>




From philh at vision25.demon.co.uk  Wed Apr 26 20:03:03 2000
From: philh at vision25.demon.co.uk (phil hunt)
Date: Thu, 27 Apr 2000 01:03:03 +0100
Subject: SPARK / Earley Parser question
References: <3907774d.32199072@news.btx.dtag.de>
Message-ID: <slrn8gf0tn.35o.philh@vision25.demon.co.uk>

On Wed, 26 Apr 2000 23:26:21 GMT, Stefan Franke <spamfranke at bigfoot.de> wrote:
>I'm designing a grammar with John Aycock's SPARK toolkit. 
>
>Earley's algorithm is said to be n^3 time bound in general, n^2 for
>unambiguous grammars and linear for most practical programming
>language grammars. 
>
>Are there some rules of thumbs which kind of rules should be avoided
>to retain linear complexity? 
>
>In particular, I wonder if using empty right-hand sides for rules has
>any disadvantages.
>
>Online references would be appreciated.

The rule of thumb I use is: code it the simplest way possible, and
only try to speed it up if it is too slow.

I use SPARK to build the parser for my Parrot program[*]. I find it is 
fast enough for my purposes -- compiles c. 150 lines of source per second 
on an AMD K6-2 running at 300MHz.

I'd suggest try it and see. If fast speed is really essential, you
might try using C/C++ and yacc instead.


(*) see <http://www.comuno.com/linux/parrot/intro.html> for details.

-- 
***** Phil Hunt ***** send email to phil at comuno.com *****
Moore's Law: hardware speed doubles every 18 months
Gates' Law: software speed halves every 18 months 


From ivanlan at callware.com  Tue Apr  4 23:28:18 2000
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Tue, 04 Apr 2000 21:28:18 -0600
Subject: Getting the argument names passed to a function ?
References: <slrn8el5t7.40k.quinn@riyal.ugcs.caltech.edu> <8ce6dm$4vp$1@news.wrc.xerox.com>
Message-ID: <38EAB2D2.401FE437@callware.com>

Hi All

Mark Jackson wrote:
> 

[wholesale snippery]

> But you forget about:
> 
> >>> a=1
> >>> b=a
> >>> arghack(a)
> 'a'
> >>> arghack(b)
> 'a'
> >>>
> 
> Your move. . . :-)
> 
>

You guys are sick.  And getting sicker.  You need professional help.

<time-share-avatar-for-a-slew-of-minor-deities.--inquire within.>-ly
y'rs,
Ivan;-)
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
http://www.pauahtun.org and
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
Author:  Teach Yourself Python in 24 Hours



From donn at u.washington.edu  Wed Apr 12 16:11:25 2000
From: donn at u.washington.edu (Donn Cave)
Date: 12 Apr 2000 20:11:25 GMT
Subject: global variables - how??
References: <8d2bne$qn7$1@news101.telia.com>
Message-ID: <8d2l9d$8jg$1@nntp6.u.washington.edu>

Quoth "Anders Eggers-Krag" <aek at mail1.stofanet.dk>:
| how do I define a propper global variable that works in all functions in all
| modules?

A proper global variable is a bigger job in Python than in some other
languages.  You still have to import its home module, in each module
that will use it, and if any but that home module will modify it, the
context must refer to some mutable object.

Here's one way to do it:

#  --- module main.py ---
class Core:
	def __init__(self):
		self.debug = 0

core = Core()

#  --- module spud.py ---
from main import core

def debug(on):
	if core.debug:
		print 'debug', on
	core.debug = on

You could eliminate the class instance and use the module directly,
changing "from main import core" to "import core".  But the class
gives you more flexibility that you may wish for later.

	Donn Cave, donn at u.washington.edu


From karab at stsci.edu  Tue Apr 18 18:11:54 2000
From: karab at stsci.edu (Kia Arab)
Date: Tue, 18 Apr 2000 22:11:54 GMT
Subject: can someone tell me about...
References: <38FCBB16.BF50EB03@stsci.edu>
Message-ID: <38FCDC5D.56E0AD8C@stsci.edu>

Robin,

Thanks for clearing that up.


From elb at cfdrc.com  Wed Apr 19 12:22:31 2000
From: elb at cfdrc.com (Ed)
Date: Wed, 19 Apr 2000 11:22:31 -0500
Subject: Python in the future
References: <38FDB8A7.E701F81F@sec.noaa.gov>
Message-ID: <38FDDD47.BAFDD041@cfdrc.com>

Python will be there for NOAA 5 and 10 years from now.  Why?  Here's
conventional answer --- it's free so you can get the source and maintain it
yourslef if you want.  The real scare for organizational commitment is not
existence, but rather support --- will the number of Python users and level
of user-to-user support be the same or bigger in 5 to 10 years?... I feel
comfortable saying yes because Python is easy and it fills so many roles.

j vickroy wrote:

> Hello all,
>
> I'm relatively new to Python and am trying to gain support for it at my
> work place.  To that end, I'm giving a brief (overview) presentation.
> One of the items I have been asked to consider is where will Python be
> 5-10 years from now.  I certainly don't know where C++ or Java will be
> in that time frame!  Any ideas on where to research the future plans for
> Python would be appreciated.
>
> Thanks for your time,
>
> - jv



From darrell at dorb.com  Sun Apr  2 12:27:46 2000
From: darrell at dorb.com (Darrell)
Date: Sun, 02 Apr 2000 16:27:46 GMT
Subject: Debugging?
References: <9JFz4.19$GJ3.1756@news7.onvoy.net>
Message-ID: <6yKF4.16811$mV5.120823@typhoon.nyroc.rr.com>

"chris davis"
> Ok, I just need to know the best way to debug my python scripts?  If this
is
> a stupid question, I apologize, but I currently use IDLE and it's
completely
> useless when it comes to debugging.  It locks up on both my window and
Linux
> machines and doesn't set break points!  Is there another way?  another
IDE?
> anything? If there isn't that's fines, I can wait until IDLE is fully
> developed, but I just don't want to waste my time searching.
>
Take a minute to look at pdb. At first the lack of a GUI seems a killer but
give it a chance, you'll be glad.

--Darrell




From sblakey at freei.com  Wed Apr 26 14:12:50 2000
From: sblakey at freei.com (Sean Blakey)
Date: Wed, 26 Apr 2000 18:12:50 +0000
Subject: How can I get the name of an object???
References: <8e712k$65c$1@nnrp1.deja.com>
Message-ID: <390731A2.12788170@freei.com>

Peter,
Python objects do not have any way of finding out what name they are
referenced by.  Names refer to objects, but objects do not have name.

This is reasonable, since it is simple for an object to have 0, 1, or
more names.  Imagine what an imaginary __name__ attribute would be in
the following cases:
>>>class A:
...  pass
...
>>>A().__name__		# What should this be? ''?
>>>a = A()
>>>b = a		# b and a are now two names for the same object
>>>b.__name__		# Should this be 'a', 'b', or ['a', 'b']?

There are some clever hacks to attempt this (like iterating through all
the names in globals() looking for a match with the object), but I have
not seen any that are completely effective.
def lookupName(object):
	'''
	Tries to find an object in the global dictionary.
	If the object is found, the name of that object is returned.
	Otherwise, the empty string is returned.
	'''
	for name, value in globals.items():
		if object is value:	# pointer comparison
			return name
	else:			# I know this "else" is unneeded, 
		return ''	#but it is what I mean

spex66 at my-deja.com wrote:
> 
> Hi again,
> 
> other trivial (looking) question:
> 
> >>> class zz: pass
> >>> dd = zz()
> 
> I cannot find the hook to get back the name of 'dd' as a string.
> Why? Cause I need the names for code generation :-)
> 
> like estimated
> >>> dd.__name__
> >>> 'dd' #DON'T WORK
> 
> any hacks available?
> 
> Peter
> (=PA=)
> 
> Sent via Deja.com http://www.deja.com/
> Before you buy.
> --
> http://www.python.org/mailman/listinfo/python-list

-- 
Sean Blakey
FreeInternet.com
sblakey at freei.com
(253)796-6500x1025



From meh9 at cornell.edu  Thu Apr  6 16:04:14 2000
From: meh9 at cornell.edu (Matthew Hirsch)
Date: Thu, 06 Apr 2000 16:04:14 -0400
Subject: Algorithm: combinations of (k) taken from (n) values
Message-ID: <meh9-ACCD3B.16041406042000@news.cit.cornell.edu>

Hi All,

Onto the next question...

Can anyone think of an algorithm to store as lists all possible 
combinations of k numbers taken from n possible numbers.  For example, 
given 5 values, I want to choose 2.  The number of possible combinations 
is given by 5!/(3!2!)=10.  They are:

[0,1], [1,2], [2,3], [3,4]
[0,2], [1,3], [2,4]
[0,3], [1,4]
[0,4]

I have something like:

n=5

for first_choice in range(n):
   for second_choice in range(first_choice+1,n):
      for third_choice in range(second_choice+1,n):
         print first_choice, second_choice, third_choice


The problem is that I'll be dealing with large combinations of numbers 
and it doesn't make sense to continue writing nested for loops.  I'm 
trying to figure out something that generates all combinations, just 
based on values of k and n.  Any ideas?

Thanks for your help,
Matt


From gmcm at hypernet.com  Tue Apr 11 12:23:22 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Tue, 11 Apr 2000 12:23:22 -0400
Subject: issubclass funny business
In-Reply-To: <m2snwsd5ah.fsf@netmeg.net>
Message-ID: <1256652290-41498655@hypernet.com>

Les Schaffer wrote:

> I have the following class hierarchy:
> 
> ------ in module FileScanner -----
> 
> class StandardArrayFileScanner(ArrayFileScanner):
> 
> --------- in module cie ----------
> 
> from FileScanner import StandardArrayFileScanner
> 
> class CIE_Standards(StandardArrayFileScanner):
> ...    
> 
> class Photopic(CIE_Standards):
> ...
> 
> class CIE(CIE_Standards):
> ...
> 
> 
> but when i need to test whther an object is an instance of a subclass
> of StandardArrayFileScanner:
> 
>         # make sure std is of the correct type
>         if not issubclass(std.__class__, StandardArrayFileScanner):
>             import cie
>             print std.__class__, cie.CIE.__bases__
> 	    raise NotStandardArrayError
> 
> 
> but it doesnt work, even though std is of type cie.CIE :
> 
> (gustav)~/Engineering/dspring/stoplite/matlab/Jue-Data/: python FileScanner.py

You are running FileScanner. Therefore, in FileScanner, class 
StandardArrayFileScanner is defined in __main__. If you 
trickle up to the top of __bases__ above, you'll find 
FileScanner.StandardArrayFileScanner, which is not the same 
object. At least, that's my guess at what's going on.
 
> cie.CIE (<class cie.CIE_Standards at 813b320>,)  <=== result of print stmt
> 
> huh??????


- Gordon



From scarblac-spamtrap at pino.selwerd.nl  Tue Apr 11 15:48:03 2000
From: scarblac-spamtrap at pino.selwerd.nl (Remco Gerlich)
Date: 11 Apr 2000 19:48:03 GMT
Subject: Functions and objects
References: <meh9-C7461E.15255011042000@news.cit.cornell.edu>
Message-ID: <slrn8f73l1.sfl.scarblac-spamtrap@flits104-37.flits.rug.nl>

Matthew Hirsch wrote in comp.lang.python:
> >>> a
> [5]

a is now a name for that list [5].

> >>> for x in range(5):     # Case A
> ...     temp=a

Now, temp is another name for the same object. a en temp are the same list.

> ...     print a
> ...     temp.append(1)

This appends to that list in place. a and temp are still just different
names for the same list, so they both point to the changed list.

> ... 
> [5]
> [5, 1]
> [5, 1, 1]
> [5, 1, 1, 1]
> [5, 1, 1, 1, 1]
> 
> >>> a=f(5)
> >>> for x in range(5):     # Case B
> ...     temp=a[:]

Here, temp is a copy of a, so now temp and a are two distinct lists, with
the same value ([5]). 

> ...     print a
> ...     temp.append(1)

Temp is changed, but since that was not the same object as a anyway, a isn't
affected.

> In case A, why doesn't temp reset itself to the value of a, [5], that 
> was predetermined before it entered the loop?  Why do you need to copy 
> the list to get the behavior I'm looking for in Case B?  Doesn't a 
> always equal [5]?

No. Assignment in Python always works by reference. This is one of the key
features of Python that you need to grok.

-- 
Remco Gerlich,  scarblac at pino.selwerd.nl
'Oook?'


From mlauer at trollinger-fe.rz.uni-frankfurt.de  Wed Apr 26 10:07:00 2000
From: mlauer at trollinger-fe.rz.uni-frankfurt.de (mlauer at trollinger-fe.rz.uni-frankfurt.de)
Date: 26 Apr 2000 16:07:00 +0200
Subject: Python compiler ? (was Re: Python for drivers?)
References: <3900EDBE.282BC85@dsuper.net> <8dr2pn$ec8$1@newshost.accu.uu.nl>
Message-ID: <3906f804@nntp.server.uni-frankfurt.de>

Martijn Faassen (m.faassen at vet.uu.nl) wrote:
> Dave <redsand at dsuper.net> wrote:
> > Can Python be used for writing device drivers? I just got BeOS 5 and it
> > doesn't support some video cards. I'd like to write one for the Cirrus
> > Logic card and I want a language that's clear and concise- Python. I've
> > never written device drivers. Any suggestions?

> Most probably: no. If you got it going, it'd be sloooooooooow.
> Python is clear and concise, but it doesn't execute very quickly compared
> to C. Python's also not as good as C at access to the bare metal
> (the video card); you'd have to get significantly funky to be able to do
> that (and you'd need some C in there). So in this case I think that even 
> prototyping in Python, then moving to C, isn't going to work. Sorry!
> Wish I could've given you more positive suggestions. It's just one of those
> cases where Python isn't fast or low-level enough. This is rare, but it
> does happen.

Speaking if this... there was a guy in this newsgroup claiming
that he could write a Python compiler if there would be enough
interest. What has become of this (guy | project) ?

--
    Regards & Gruesse from Mickey @ http://www.Vanille.de
  ---------------------------------------------------------
  How could anyone know me - when I don't even know myself ?




From jacobs at darwin.epbi.cwru.edu  Mon Apr 10 15:28:30 2000
From: jacobs at darwin.epbi.cwru.edu (Kevin Jacobs)
Date: 10 Apr 2000 19:28:30 GMT
Subject: indexed dictionaries?
References: <Pine.BSF.4.10.10004092225040.1361-100000@chi.pair.com> <B516CDD9.17E6%dgoodger@bigfoot.com>
Message-ID: <8cta0u$2bg$2@eeyore.INS.CWRU.Edu>

David Goodger <dgoodger at bigfoot.com> wrote:
> on 2000-04-09 22:30, Michal Wallace (sabren) (sabren at manifestation.com)
> wrote:
>>  I'm looking for a cross between a dictionary and a list,
>> such that keys() always returns keys in the order in which they were
>> added, and that instance[0] refers to the first key...
> ...
>> Anyone got something like this?

> I have implemented a list/dictionary hybrid in my SGF Parser Library (class
> Node in http://gotools.sourceforge.net/sgflib/ ). It doesn't do ordered
> keys(), but it does do indexing by integer (note that in this library, no
> numeric dictionary keys will ever be used). It may give you a few ideas.

> Please send me a copy of your final implementation of indexed dictionaries;
> I'd be interested to see what you come up with.

Here is my rough 3 minute attempt.  

~Kevin

# WARNING: Untested code
class SequenceDict(UserDict):
  def __init__(self):
    UserDict.__init__(self)
    self._seq = []

  def __setitem__(self,key,value):
    if not self.data.has_key(key):
      self._seq.append(key)
    self.data[key]=value

  def __delitem__(self,key):
    if self.data.has_key(key):
      self._seq.remove(key)
      del self.data[key]

  def seq_keys(self):
    return self._seq

  def seq_items(self):
    return map(None, self._seq, self.seq_values())

  def seq_values(self):
    return map(operator.getitem, [self]*len(self._seq), self._seq)



-- 
----------->  Kevin Jacobs  <-----------|------->  (216) 778-8487  <--------
S.A.G.E. Project Technical Coordinator  | Department of Epidemiology
  & System Administrator                |   & Biostatistics
Internet E-mail: jacobs at darwin.cwru.edu | Case Western Reserve University
----------------------------------------------------------------------------


From exarkun at flashmail.com  Sun Apr 23 02:23:33 2000
From: exarkun at flashmail.com (Jp Calderone)
Date: Sun, 23 Apr 2000 06:23:33 +0000
Subject: python-only crypt()?
References: <Pine.BSF.4.10.10004230627580.5642-100000@chi.pair.com>
Message-ID: <390296E5.D75BB2B8@flashmail.com>

"Michal Wallace (sabren)" wrote:
> 
> Hey all,
> 
>   Is there a python-only implementation of the crypt() module?  I'm
> looking for users who don't have access to rebuild python with crypt
> enabled...
> 
>   I could only find C modules... I'm taking a crack at porting
> Crypt::UnixCrypt from perl, but I can't really tell what the code is
> doing once it gets down into the DES algorithm, so it's a little
> tedious.

 I dunno about the first part, but there's a very good (IMHO) 
explanation of the DES algorithm at http://www.aci.net/kalliste/des.htm
if you choose to go that route.

  Jp

-- 
Where a calculator on the ENIAC is equipped with 18,000 vacuum tubes and
weighs 30 tons, computers in the future may have only 1,000 vacuum tubes and
weigh only 1.5 tons.    -- Popular Mechanics, March 1949
--
 6:22am up 25 days, 11:29, 0 users, load average: 0.07, 0.04, 0.01


From cdaniel at xcom.net  Thu Apr  6 00:27:32 2000
From: cdaniel at xcom.net (Cliff Daniel)
Date: Thu, 6 Apr 2000 00:27:32 -0400
Subject: CLI over Sockets?
Message-ID: <20000406002732.D1027@shell2.xcom.net>

Any such module exist that might help in implementing a
CLI over sockets?  One such beast would preferrably operate in
character mode and would would need to implement vt100  to provide
the command line editing features (arrow keys, insert at cursor, etc).
I'm know I'm asking for a lot but it's worth a shot in case someone
has implemented this already.

Thanks,
Cliff



From bparsia at email.unc.edu  Fri Apr  7 10:09:12 2000
From: bparsia at email.unc.edu (Bijan Parsia)
Date: Fri, 7 Apr 2000 10:09:12 -0400
Subject: Refactoring browser for Python?
References: <16d61288.5cc55c16@usw-ex0101-005.remarq.com> <19c2a6b4.30d6893e@usw-ex0103-019.remarq.com>
Message-ID: <1e8pqj7.1pc3krx1gpiqleN%bparsia@email.unc.edu>

wheineman <wheinemanNOwhSPAM at uconect.net.invalid> wrote:

> There seems to be some discussion of just this at WikiWiki. See
> 
> http://c2.com/cgi-bin/wiki?PythonRefactoringBrowser
> 
> and
> 
> http://c2.com/cgi-bin/wiki?BicycleRepairMan
> 
> BicycleRepairMan being the current name of the browser (highly
> commendable choice). There's only talk at the moment but they are
> looking for volunteers.

I've been interested in Python refactoring for a while, at least, as a
theoretical issue. It wouldn't be all that difficult to add automatic
refactorings to, say, IDLE. A tricky bit is coming up with a list of
suitible refactorings. But if one started with simple, intramodular
refactorings (fucntion->method, method->function, moving things up and
down a class heirarchy, fuction stored to local, etc.) I don't imagine
it would take much time automate them at all and they would be
immediately useful.

Hmm. I imagine various attribute settings and gettings are ripe for
refactorings, but I don't know, off hand, the details.

-- 
Bijan Parsia
http://monkeyfist.com/
...among many things.


From senn at maya.com  Wed Apr  5 13:52:25 2000
From: senn at maya.com (Jeff Senn)
Date: 05 Apr 2000 13:52:25 -0400
Subject: Tkinter:  "Get root for a widget" is private?
In-Reply-To: Randall Hopper's message of "Wed, 05 Apr 2000 12:42:58 -0400"
References: <20000331112048.A1257432@vislab.epa.gov> <QdEF4.892$n68.190327808@newsb.telia.net> <m3snx3n1x0.fsf@devel.no.rwd.dinet> <20000331112048.A1257432@vislab.epa.gov> <QdEF4.892$n68.190327808@newsb.telia.net> <20000403064335.A1560771@vislab.epa.gov> <2Q_F4.1092$n68.187097600@newsb.telia.net> <20000405073526.A1760238@vislab.epa.gov> <wk8zysd6zc.fsf@SNIPE.maya.com> <20000405124258.A1747369@vislab.epa.gov>
Message-ID: <wkwvmcbfl2.fsf@SNIPE.maya.com>

Randall Hopper <aa8vb at yahoo.com> writes:
> Jeff Senn:
>  |Unless you're doing something wacky (creating multiple roots??), you
>  |could just do what Tkinter itself (generally) does and use
>  |Tkinter._default_root to get the instance of the root object....
> 
> Right.  I'm using _root() now which works.  My original point wasn't that I
...
> If _root() was ripped out in the next version of Python, I wouldn't have a
> leg to stand on.  And by all rights it's protected so removing it should
> not affect me (or any other external code), unless I (or it) was deriving
> off this class.

[...deriving from one of the widget class... this is a common practice, no?
So, you're probably safe anyway...]

> As Fredrik mentioned I could just snitch global Tk commands off
> any_old_widget, but an event loop timer object that has nothing to do with
> "any" widget shouldn't be keeping a handle to any_old_widget.  That's not
> clean code.

Definitely.

...
> So my request was for _root() to be declared public (e.g. as .root())

Sorry -- I missed the very beginning of the thread.  Indeed a public
Tkinter.root() (or .default_root()) seems like the best solution.

For the short run you could duplicate _root() in your own code -- it
relies only on "public" members after all ;-)

Class Misc:
...
	def _root(self):
		w = self
		while w.master: w = w.master
		return w
-- 
-Jas




From lvw at lfpt.rwth-aachen.de  Wed Apr 19 04:28:46 2000
From: lvw at lfpt.rwth-aachen.de (Lars von Wedel)
Date: Wed, 19 Apr 2000 10:28:46 +0200 (MET DST)
Subject: Using CGI module with 'canned queries'
Message-ID: <200004190828.KAA03403@gromit.rwth-aachen.de>

Hello,

I want to use Python to implement some CGI scripts. I want these
scripts to be callable from any link (not only via a form),
e.g. the user shall be able to enter something like

http://.../test.py?name=Bill

into Netscape directly.

I expected that such a URL could be processed using

    form = cgi.FieldStorage()    print '<UL>'
    for k in form.keys:
	print '<LI>' % k % ':' % form[k] % '</LI>'

(after printing an html header of course).

However, the dictionary seems to be empty when the http daemon
starts my script.

How do I access parameters within the script if not started via a form?

Lars



From g2boojum at hotmail.com  Sat Apr 22 15:36:16 2000
From: g2boojum at hotmail.com (Grant Goodyear)
Date: Sat, 22 Apr 2000 14:36:16 -0500
Subject: executable "wrapper" gui -- advice sought
Message-ID: <xhnM4.1153$PP5.17158@news.swbell.net>

I'm in the process of writing an exceedingly simple gui that will take an
already existing binary executable and input text file (I don't want to
break the
working system!) and allow the user to modify the input arguments,
if desired, and then start the executable.  That part is straightforward.
What I'm not sure how to do is to capture the executable's std out as
it's generated.  My goal is to echo that output to a seperate window, and
also to use it to generate a periodically updated html "status report" so
that the program's progress can be monitored remotely.  Working
through the gui aspects isn't all that difficult, but I have _no_ idea how
to grab the executable's std out _while it is running_ and then have the
python gui do something (anything!) with that data.  Any advice?

Many thanks,
Grant Goodyear




From tismer at tismer.com  Thu Apr 20 11:54:06 2000
From: tismer at tismer.com (Christian Tismer)
Date: Thu, 20 Apr 2000 17:54:06 +0200
Subject: Why MFC instead of Tkinter -- was Re: Pythonwin?
References: <eMz2Nmiq$GA.73@net025s> <4QtL4.24$v85.41@news-server.bigpond.net.au> <chuL4.8483$O7.17315@typhoon.southeast.rr.com> <7LDL4.534$v85.2563@news-server.bigpond.net.au> <sfu6d657l6e184@corp.supernews.com> <38ff21b8$0$10972@senator-bedfellow.mit.edu>
Message-ID: <38FF281E.5B2DCC71@tismer.com>


Robert wrote:
> 
> To each his own I guess. I work mostly on Windows platforms, and for me
> MFC is a lot better than Tk for a couple of reasons.
> 1) I can port MFC/C++ programs a lot easier
> 2) I may be mistaken, but MFC based guis are faster than Tk
> 3) I really do not like using layout managers, and using Visual C++'s dialog
> editor
> is the best thing since sliced bread IMHO.

Really? Then you never used Delphi, I believe.
If you have a chance to try that, try that,
and you will most probably throw MFC far far away.
For me, it is an order of magnitude easier to use.
Unfortunately, I have no current info about the
PyDelphi project.

ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaunstr. 26                  :    *Starship* http://starship.python.net
14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     where do you want to jump today?   http://www.stackless.com



From pf at artcom-gmbh.de  Tue Apr 18 07:59:29 2000
From: pf at artcom-gmbh.de (Peter Funk)
Date: Tue, 18 Apr 2000 13:59:29 +0200 (MEST)
Subject: Lazy imports (was Re: module naming)
In-Reply-To: <Pine.BSF.4.10.10004162056040.25069-100000@chi.pair.com> from "Michal Wallace (sabren)" at "Apr 16, 2000  8:59:26 pm"
Message-ID: <m12hWfB-000CnvC@artcom0.artcom-gmbh.de>

Hi!

> On Mon, 17 Apr 2000, David Arnold wrote:
> > what about something like
> > 
> >    from Record import Record
> >    from Field import Field
> >    from IdxDict import IdxDict
> > 
> > in your zdc/__init__.py ?
[...]

Michal Wallace (sabren) asks:
>   That's certainly a lot cleaner! But is there a way to do that and
> not have it load each one of those files every time? There's actually
> more than 3, and even though they're all fairly small, I'd rather not
> have the overhead of loading them until they're needed.

You could try to adopt the lazy loader concept implemented in 
the Pmw package.  The idea is to use a small tool, that builds a 
dictionary of exported features (classes and more) by investigating
all module files in the package and places this in a file called
for example 'exports.def'.  Than you can use the following
__init__.py file to load features (classes, constants, functions,
or modules, what you want) on demand.  The idea is "stolen" from
Greg McFarlanes wonderful Pmw package:

Regards, Peter
---- 8< ---- 8< ---- cut here ---- 8< ---- schnipp ---- 8< ---- schnapp ----
#!/usr/bin/env python
## vim:ts=4:et:nowrap
"""__init__ : This file is executed when the package is imported.  It creates
a lazy importer/dynamic loader for the package and replaces the package 
module with it.  This is a very simplified version of the loader supplied 
with Pmw.  All the package version management has been stripped off."""

import sys, os, string, types

_EXP_DEF = 'exports.def'       # export definition file
_BASEMODULE = 'base'           # Name of Base module for the package

class _Loader:
    """An instance of this class will replace the module in sys.modules"""

    def __init__(self, path, package):
        self._path, self._package = path, package
        self._initialised = 0
        
    def _getmodule(self, modpath):
        __import__(modpath)
        mod = sys.modules[modpath]
        return mod

    def _initialise(self):
        # Create attributes for the Base classes and functions.
        basemodule = self._getmodule('_'+self._package+'.'+_BASEMODULE)
        for k,v in basemodule.__dict__.items():
            if k[0] is not '_' and type(v) != types.ModuleType:
                self.__dict__[k] = v
        # Set the package definitions from the exports.def file.
        dict = {
            '_features'     : {},
            '_modules'      : {},
        }
        for name in dict.keys():
            self.__dict__[name] = {}
            d = {}
            execfile(os.path.join(self._path, _EXP_DEF), d)
            for k,v in d.items():
                if dict.has_key(k):
                    if type(v) == types.TupleType:
                        for item in v:
                            ## modpath = self._package + item
                            modpath = item
                            dict[k][item] = modpath
                    elif type(v) == types.DictionaryType:
                        for k1, v1 in v.items():
                            ## modpath = '_'+self._package +'.'+ v1
                            modpath = v1
                            dict[k][k1] = modpath
        self.__dict__.update(dict)
        self._initialised = 1

    def __getattr__(self, name):
        """This will solve references to not yet used features"""
        if not self._initialised:
            self._initialise()
            # Beware: _initialise may have defined 'name'
            if self.__dict__.has_key(name):
                return self.__dict__[name]
        # The requested feature is not yet set. Look it up in the
        # tables set by exports.def, import the appropriate module and
        # set the attribute so that it will be found next time.
        if self._features.has_key(name):
            # The attribute is a feature from one of the modules.
            modname = self._features[name]
            mod  = self._getmodule('_'+self._package+'.'+modname)
            feature = getattr(mod, name)
            self.__dict__[name] = feature
            return feature
        elif self._modules.has_key(name):
            # The attribute is a module
            mod = self._getmodule('_'+self._package+'.'+name)
            self.__dict__[name] = mod
            return mod
        else:
            # The attribute is not known by the package, report an error.
            raise AttributeError, name

# Retrieve the name of the package:
_package = os.path.split(__path__[0])[1]
# Rename (hide) the original package for later perusual:
sys.modules['_'+_package] = sys.modules[_package]
# Create the dynamic loader and install it into sys.modules:
sys.modules[_package] = _Loader(__path__[0], _package)



From mwh21 at cam.ac.uk  Fri Apr 14 05:58:16 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 14 Apr 2000 10:58:16 +0100
Subject: exception problem
References: <38f6ebc3.0@news.proweb.co.uk>
Message-ID: <m3wvm1f1hj.fsf@atrus.jesus.cam.ac.uk>

".:|:." <ng at hardlight.couk.com> writes:

>   File "/pr0n/getGroup.py", line 58, in ?
>     ArticleToFile(s, `i`)
>   File "/pr0n/getGroup.py", line 26, in ArticleToFile
>     decode(uufn, binfh)
>   File "/usr/lib/python1.5/uu.py", line 95, in decode
>     raise Error, 'No valid begin line found in input file'
> uu.Error: No valid begin line found in input file
> 
> from uu import *
> decode(infile, outfile)
> 
> raise Error, 'No valid begin line found in input file'
> uu.Error: No valid begin line found in input file
> 
> 
> how do i trap it
> 
> i've tried
> 
> try:
>     decode(infile, outfile)
> except uu.error:
>     pass
> 
> gives me
> 
> NameError uu :

Well, that's 'cause you did

from uu import *

which does not put `uu' into the local namespace ...

Two options:

1) don't do "from ... import *" (this is the right one <wink>)
2) try this:

try:
    decode(infile, outfile)
except error:
    pass

but that's fragile (see option 1).

Cheers,
M.

-- 
  "declare"?  my bogometer indicates that you're really programming
  in some other language and trying  to force Common Lisp into your
  mindset.  this won't work.           -- Erik Naggum, comp.lang.lisp


From gmcm at hypernet.com  Sat Apr  8 13:20:00 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Sat, 8 Apr 2000 13:20:00 -0400
Subject: Cacheing sys.path rather than building it each time?
In-Reply-To: <0502b954.18a27ec8@usw-ex0103-023.remarq.com>
Message-ID: <1256908084-26112509@hypernet.com>

Hamish Lawson wrote:

> Steve Holden wrote:
> 
> > Probably not an awfully good idea, given that "installation" of
> > a new module often consists of just dropping a ".py" file into
> > a library directory.  Without an common "install" process the
> > work of checking whether the cahced path is up to date would
> > be almost as much as that of building the path from scratch
> > as is now done.
> 
> Had you thought I was proposing that such a check be done
> whenever Python was started? I'd agree that the work of checking
> whether the cached path is up to date would be almost as much
> work as building the path. However that wasn't part of my
> proposal.
> 
> Instead when a module is installed, the pathbuilder tool would be
> run either by the module's install program (perhaps via the
> distutils) or manually by the user. (You could even run it at
> scheduled intervals if you thought that necessary to catch
> omissions.) My goal is to avoid having the path built every time
> Python starts.
> 
> A command-line option for the Python interpreter could be used to
> determine whether it uses the cached path or builds the path from
> scratch; the decision as to which should be the default I'll
> leave to others.

There are 3 pieces to the puzzle.

1) How do you find those things that Python requires 
(bootstrap).
2) How do you build the list of places to look for Python 
modules.
3) What is the complete universe of importable Python 
modules.

You are talking about #3, but the original question seems to 
have been about #2.

Caching #3 would involve rewriting the entire import 
mechanism. It would yield enormous runtime speedups, since 
an import would take 1 I/O (vs something along the lines of 4 * 
len(sys.path)/2). This might be very cool in a CGI type 
situation, but it's very unlikely to ever become standard 
Python because it violates the "no surprises" rule. There are 
also other, simpler ways of speeding up imports (eg, archives).

But the startup overhead is mostly in #1. There's a lot of code 
in getpath.c dealing with things like developer builds, strange 
installations and other sys admin hacks. For a general 
purpose Python installation, this code needs to be there (well, 
it needs to be someplace, not necesarily in getpath.c); but for 
a special purpose Python, it's relatively easy to hack 
getpath.c to your needs.

Once you've done that, #2 is really a non-issue. Consolidate 
your .pth files and /or fine tune sitecustomize.py and sys.path 
will get filled out with minimal I/Os.

- Gordon



From aycock at csc.uvic.ca  Mon Apr 24 10:59:12 2000
From: aycock at csc.uvic.ca (John Aycock)
Date: 24 Apr 2000 14:59:12 GMT
Subject: partial parsing?
References: <8dvilp$88s$1@slb6.atl.mindspring.net> <8e10n7$rkl$1@slb3.atl.mindspring.net>
Message-ID: <8e1ng0$4sao$1@uvaix7e1.comp.UVic.CA>

Andrew Dalke <dalke at acm.org> wrote:
> I've been looking at existing parser generation systems, such as
> lex/yacc, SPARK and Plex.  They all seem to be designed to verify
> every character in the file, so I lose some of the performance
> given the guarantee of having a valid format.

Well, parsers operate on a token stream, which can be as abstract as
you like.

It sounds like what you want is "fuzzy parsing," a term coined in

	A Systematic Approach to Fuzzy Parsing
	RAINER KOPPLER
	Software: Practice and Experience
	Volume: 27, Issue: 6, Date: June 1997, Pages: 637-649

John


From Norman_Shelley-RRDN60 at email.sps.mot.com  Fri Apr 28 11:43:44 2000
From: Norman_Shelley-RRDN60 at email.sps.mot.com (Norman Shelley)
Date: Fri, 28 Apr 2000 08:43:44 -0700
Subject: ?sys.argv values more literally?
Message-ID: <3909B1AF.32FC8BED@email.sps.mot.com>

I need to write a frontend program that will call one of two other
programs (both of which are NOT binaries but are tcl programs therefore
they HAVE to be invoke with os.system() or "/bin/sh" has to be the first
argument to a os.execv() call).

THe frontend program must detect a "-ver someversionname" options and
remove the two words and then call the appropriate tcl program with the
EXACT arguments given on the command line (less those two words).
Problem is words in sys.argv
that were quoted on the command line look like words that were not
quoted on the command, e.g.

%frontend -ver 1.0 word1 word2 "word three" "word(three)"
argv looks like
["frontend", "-ver", "1.0", "word1", "word2",  "word three",
"word(three)"]

but this fails when os.system or os.execv is called (with frontend,
-ver, and 1.0 replaced with the tcl program full pathname (/bin/sh added
at front  for execv) as "word three" and "word(three)" are missing there
protective quotes (the /bin/sh needs to see those as one word not two
words or a word  with parens).

I need   argv to look like this:
["frontend", "-ver", "1.0", "word1", "word2",  '"word three"',
'"word(three)"']
where words that were quoted on the command line are strings with
beginning and ending quotes in the argv.

Any way to do this or get around this problem?

--
Best regards,


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20000428/f568567b/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Norman_Shelley-RRDN60.vcf
Type: text/x-vcard
Size: 281 bytes
Desc: Card for Norman Shelley
URL: <http://mail.python.org/pipermail/python-list/attachments/20000428/f568567b/attachment.vcf>

From mskott at image.dk  Mon Apr 24 14:18:23 2000
From: mskott at image.dk (Martin Skøtt)
Date: Mon, 24 Apr 2000 18:18:23 GMT
Subject: Trouble compiling PyGreSQL 2.4 on Linux
References: <Pine.LNX.4.21.0004241614020.7343-100000@fep132.fep.ru>
Message-ID: <7ldngy6z.fsf@jumbo.skott.dk>

Oleg Broytmann <phd at phd.russ.ru> writes:

> On Mon, 24 Apr 2000, Martin Sk?tt wrote:
> > Seems like a great time for me to learn how to write a make file. I
> > will post it to him one of the following days.
> 
>    You couldn't write a plain makefile - there are many different systems,
> and installations, etc.

As far as I know, this special situation only occurs on Red Hat
systemts and not other Linux distributions (maybe on mandrake which is
RH with some extra vanilla).
>    Better, learn to write Distutil, and write one for PyGreSQL - that's a
> Real Challenge!

I have allready thought about it and that seems a lot smarter, but
that should propably be left over to the more experienced Python installers. 

-- 
Martin Sk?tt
mskott at image.dk

'In a certain sense, all decent programming languages are equally powerfull'
Dr. Alan Turing


From mfletch at tpresence.com  Wed Apr 26 11:21:20 2000
From: mfletch at tpresence.com (Mike Fletcher)
Date: Wed, 26 Apr 2000 11:21:20 -0400
Subject: [Win32] Retrieve icon for file type?
Message-ID: <CC59868A0A76D311B1F50090277B321C1F32B4@VRTMAIL>

I'm attempting to use the Windows registry to retrieve (and convert to a
PNG/JPEG) the Windows icons for given files (so that a Windows-based server
can display file-type-specific icons).  I do not seem to be able to find the
function shellapi.ExtractIcon (
http://msdn.microsoft.com/library/psdk/winui/icons_7h2m.htm ) in the Win32
extensions.

Is there an easy way to get at the Shell icons from Python?

def query ( extension):
	''' get icon for the given extension '''
	try:
		path = win32api.RegQueryValue(
			HKEY_CLASSES_ROOT,
			extension,
		)
		filedata = win32api.RegQueryValue(
			HKEY_CLASSES_ROOT,
			'%s\\DefaultIcon'%(path),
		)
		# how do I load the icon data?
		# ExtractIcon?
	except win32api.error:
		return None

__________________________________
 Mike C. Fletcher
 Designer, VR Plumber
 http://members.home.com/mcfletch



From chafik at nevrax.com  Fri Apr  7 03:46:41 2000
From: chafik at nevrax.com (Sameh chafik pro)
Date: Fri, 7 Apr 2000 09:46:41 +0200
Subject: runin a node table
Message-ID: <022401bfa065$69b3df00$1001a8c0@nevrax.net>

Hello evrybody,
I find in the Python C API the folwing function:
struct _node* PyParser_SimpleParseString (char *str, int start);
struct _node* PyParser_SimpleParseFile (FILE *fp, char *filename, int start);
but i do'nt fin a function like 
PyObject* PyRun_String (struct _node *str, int start); 
Are there any one, or how can i do to run a node table ?
Thank for all.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20000407/bf8b8c6c/attachment.html>

From fredrik at pythonware.com  Wed Apr 26 05:11:57 2000
From: fredrik at pythonware.com (fredrik at pythonware.com)
Date: Wed, 26 Apr 2000 09:11:57 GMT
Subject: Python 1.6a2 Unicode bug (was Re: comparing strings and ints)
References: <jr%I4.643$rc9.190628352@newsb.telia.net> <1256565470-46720619@hypernet.com> <6M1J4.662$rc9.209708544@newsb.telia.net> <8daop0$8fk$1@slb6.atl.mindspring.net> <d3HK4.2612$Za1.38971@newsc.telia.net>
Message-ID: <8e6bsl$f1a$1@nnrp1.deja.com>

Fredrik Lundh wrote:
> Aahz Maruch <aahz at netcom.com> wrote:
> > >Python 1.6a2 (#0, Apr  5 2000, 23:56:55)
> > >Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
> > >>>> u"abc" < "???"
> > >Traceback (most recent call last):
> > >  File "<stdin>", line 1, in ?
> > >ValueError: UTF-8 decoding error: unexpected code byte
> >
> > That looks like a bug to me.
>
> as far as I can tell, it's supposed to be a feature.
>
> if you mix 8-bit strings with unicode strings, python 1.6a2
> attempts to interpret the 8-bit string as an utf-8 encoded
> unicode string.
>
> but yes, I also think it's a bug.  but this far, my attempts
> to get someone else to fix it has failed.  might have to do
> it myself... ;-)

postscript: the powers-that-be has decided that this is not
a bug.  if you thought that strings were just sequences of
characters, just as in Perl and Tcl, you're in for one big
surprise in Python 1.6...

</F>


Sent via Deja.com http://www.deja.com/
Before you buy.


From cmfinlay at SPAMmagnet.com.au  Sun Apr 30 05:42:59 2000
From: cmfinlay at SPAMmagnet.com.au (cmfinlay at SPAMmagnet.com.au)
Date: Sun, 30 Apr 2000 19:42:59 +1000
Subject: How do I make list global please?
Message-ID: <t=0LOZtIsnos1t9tIo=000hy2MF1@4ax.com>

How do I make list global please?

I have a list which comprise of mags[0] and mags[1]
I want to pass and return it from a subroutine.
my code is (in two text files)


from inittest import InitTest
mags = [333]
mags.remove(333)
mags.append(0) # (Big) array of magnets
mags.append(1) # (Big) array of magnets
print mags[0]
print mags[1]
InitTest()
print mags[0]
print mags[1]

# file :inittest.py
def InitTest():
   global mags
   print "I am in the Sub"
   Print mags[0]
   Print mags[1]
   print "Leaving Sub"
   return

I get the error message
QUOTE
Traceback (innermost last):
  File "inittest1.py", line 1, in ?
    from inittest import InitTest
  File "inittest.py", line 5
    Print mags[0]
             ^
SyntaxError: invalid syntax
UNQUOTE

my code works when it is all in the main routine.

Please post here or reply to
cmfinlay at SPAMmagnet.com.au


From mwh21 at cam.ac.uk  Wed Apr  5 12:16:12 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 05 Apr 2000 17:16:12 +0100
Subject: Pass by reference ?
References: <38E864C9.E1EE94F7@ecs.soton.ac.uk> <38EAF8D8.4B7C43B4@ecs.soton.ac.uk> <38EB56AE.F938CE06@acm.org> <m33dp0muk9.fsf@atrus.jesus.cam.ac.uk> <38EB63BC.FE03BA31@ecs.soton.ac.uk>
Message-ID: <m3ya6sle0j.fsf@atrus.jesus.cam.ac.uk>

Jacek Generowicz <jmg at ecs.soton.ac.uk> writes:

> Michael Hudson wrote:
> 
> > Python never ever copies anything when you pass it as an argument to a
> > function - it's just that sometimes the things you pass are immutable
> > - which is when it looks like call by value - and sometimes they're
> > mutable - when it looks a bit like call by reference.
> 
> Not necessarily; consider the following (and please excuse my Python-naive
> coding style):

[snip]
 
> So whether the call resembles by-reference or by-value depends on what you do
> with the mutable object.

which is why I said:

> Try to cast Python semantics in terms of what other langauges do is
> confusing - I've just successfully confused myself trying (as you
> could see if you could read my kill-ring) - and is hence to be
> considered bogus, because Python's semantics are actually extremely
> straightforward.

but it still doesn't copy the argument.  The code you posted only
confuses if you have a flawed understanding of assignment in Python
(which is the same problem, of course, as the semantics of argument
passing and variable assigment are the same - for a very good reason).

getting-bored-of-thie-yet-everyone?-ly y'rs
M.

-- 
  ... but I guess there are some things that are so gross you just have
  to forget,  or it'll destroy something within you.  perl is the first
  such thing I have known.                 -- Erik Naggum, comp.lang.lisp


From mike.zaharkin at westgroup.com  Tue Apr 18 13:08:33 2000
From: mike.zaharkin at westgroup.com (MikeZ)
Date: Tue, 18 Apr 2000 13:08:33 -0400
Subject: Implementing a Python process when PC is idle?
Message-ID: <38fc9694@wwwproxy3.westgroup.com>

Hello,

    I would like to execute a Python process that gets executed when my
Windows PC is in screensaver mode (so my PC can work when it is idle).

    Has anyone ever implemented this, or have an idea how this could be done
? I prefer to do this all in Python, including any Windows commands
necessary to register the process as a screensaver.

Thanks in advance,
-Mike Z.




From michaelc at flashcom.net  Thu Apr 20 10:55:52 2000
From: michaelc at flashcom.net (Michael A. Crawford)
Date: Thu, 20 Apr 2000 07:55:52 -0700
Subject: Why MFC instead of Tkinter -- was Re: Pythonwin?
References: <eMz2Nmiq$GA.73@net025s> <4QtL4.24$v85.41@news-server.bigpond.net.au> <chuL4.8483$O7.17315@typhoon.southeast.rr.com> <7LDL4.534$v85.2563@news-server.bigpond.net.au>
Message-ID: <sfu6d657l6e184@corp.supernews.com>

I'm curious.  Why did you not steer him in the Tkinter direction for his GUI
needs?  Is MFC better?  I learned (and forgot MFC) some time ago, years in
fact.  Same with X-Windows programming.  Now I'm working my way throught
"Python and Tkinter Programming" by Grayson.  I thought Tkinter was the
preferred method for Pythong GUI programming.  Am I mistaken?

-Michael

"Mark Hammond" <mhammond at skippinet.com.au> wrote in message
news:7LDL4.534$v85.2563 at news-server.bigpond.net.au...
> "Don Tuttle" <tuttledon at hotmail.com> wrote in message
> news:chuL4.8483$O7.17315 at typhoon.southeast.rr.com...
> > [Mark]
> > > You are really going to need to go through an MFC tutorial or book
> before
> > > you can make heads or tails of Pythonwin from that level...
> >
> > Any recommendations?
>
> Afraid not.  I already new MFC when I started Pythonwin.  I learnt it from
> the MFC "scribble" tutorial.  I find I tend to learn best with a tutorial
> and then playing - that is what I did with Python (back then there was no
> other alternative, but even now that is my recommended strategy for
> competant existing programmers learning Python...)
>
> So assuming some C++ knowledge, my suggestion for MFC would still be
> scribble tutorial, then invent an app to write in MFC C++, then port that
> app to Python.  By then you will be a wiz :-)
>
> Mark.
>
>
>




From thomas.heller at ion-tof.com  Tue Apr 18 14:05:34 2000
From: thomas.heller at ion-tof.com (Thomas Heller)
Date: Tue, 18 Apr 2000 20:05:34 +0200
Subject: [Announce] autoreload.py
References: <001601bfa8a0$c6d86f50$4500a8c0@thomasnb>
Message-ID: <012901bfa960$b1422580$4500a8c0@thomasnb>

I just discovered that it is easier to monitor sys.modules instead
of catching all imports.
Will post a new version shortly.

Thomas Heller




From darrell at dorb.com  Sun Apr 16 10:22:36 2000
From: darrell at dorb.com (Darrell)
Date: Sun, 16 Apr 2000 14:22:36 GMT
Subject: Event triggering and weak references
References: <8dbif3$hm0$1@metro.ucc.usyd.edu.au>
Message-ID: <M0kK4.7275$wM3.71361@typhoon.nyroc.rr.com>

"Ricardo Nogueira"
> This is what I would like to do:
> Imagine I could have two or more Widgets open displaying a car speed
value,
> say a numerical and a graphic display, if this car speed changes (ex. some
> one pushes the break), I would like to have any open widget on this value
> updated (on opening each window would register interest with this car
> speed).
>
Here's an approach that requires changes to the watched class.
Maybe some of Robert's method wrappers would be nice.
Being able to hook a method from the outside is a powerful way to decouple
code. But wouldn't it be harder to understand? Sometimes I think C++ is
difficult because only when you trace though it do you realize where it's
going. Having the watched thing provide an explicit signal seems easier to
understand.

Don't know if you can call this weak references but they can't cause cycles.
The use of strings to avoid cycles may not be required soon with the new GC
stuff.

uthreads.py might be cool for this sort of thing, somehow?

Thanks for the Sunday morning puzzle.
--Darrell Gallion


class Watches:
        # Only strings are stored
    _Obs={}

    def regWatch(self, obs, watch):
        """
        obs     : observer name string
        watch   : string name of thing to watch
        """
        Watches._Obs[watch].append(obs)

    def unregWatch(self, obs, watch):
        l=Watches._Obs[watch]
        x=l.index(obs)
        del l[x]

    def regWatchAble(self, watch):
        Watches._Obs[watch]=[]

    def unregWatchAble(self, obs, watch):
        del Watches._Obs[watch]

    def signal(self,watch):
        l=Watches._Obs[watch]
        w=eval(watch)
        for i in l:
            i=eval(i)
            i(w)

    def clean(self):
        for k in Watches._Obs.keys():
            try:
                w=eval(watch)
            except:
                del Watches._Obs[k]
                print 'Cleaned:',k

watches=Watches()

class W1:
    """Something to watch"""
    def __init__(self, watches,name):
        self._var1=0
        self._watches=watches
        self._name=name
        watches.regWatchAble(name)

    def setVar(self,v):
        self._var1=v
        self._watches.signal(self._name)

    # 'w1' would be more like 'moduleName.w1'
w1=W1(watches, 'w1')

def obs1(w):
    print "Signaled:",w

watches.regWatch('obs1','w1')

w1.setVar(6)
del w1
watches.clean()

############ Output
Signaled: <__main__.W1 instance at 895b48>
Cleaned: w1






From teemu.keskinarkaus at ramk.fi  Fri Apr 28 06:59:41 2000
From: teemu.keskinarkaus at ramk.fi (Teemu Keskinarkaus)
Date: Fri, 28 Apr 2000 10:59:41 GMT
Subject: Reading user input
Message-ID: <39096d88.3112113532@news.ramk.fi>

Hi.

I'm doing some studing about python and MySQL. I've done little test
program where I use database from python. Now I need get user input to
do some queries(and other stuff) to database. I've read Python manual
but haven't found simple solution to do that. I can read something
using curses but I don't like it. Is there other(and simpler) way to
do that??

I've done only CGI programs with python and there haven't been any
need to read chars from user input. Only stuff from stdin that httpd
have put there.

At this case 'user input' is not arguments given when program starts
but variables that are given when program is running. 

Of course I could do html-page and do my test program as CGI and use
it that way but it's too much trouble when I'm only trying to learn
databases and how to use them from python.

TK


From dworkin at ccs.neu.edu  Wed Apr  5 14:05:28 2000
From: dworkin at ccs.neu.edu (Justin Sheehy)
Date: 05 Apr 2000 14:05:28 -0400
Subject: Teaching the art of programming, in python
In-Reply-To: "Warren Postma"'s message of "Wed, 5 Apr 2000 11:56:11 -0500"
References: <20000405053920.20579.qmail@web2102.mail.yahoo.com> <8cfe42$o02$1@nnrp1.deja.com> <dpJG4.3744$HG1.106895@nnrp1.uunet.ca>
Message-ID: <vndem8kfmon.fsf@camelot-new.ccs.neu.edu>

"Warren Postma" <embed at geocities.com> writes:

> Speaking of Dummies, was there/is there a LISP for Dummies ... ?

Not exactly, but The Little Lisper is the closest thing I know of.

-Justin

 





From Marc.Poinot at onera.fr  Tue Apr 25 05:33:36 2000
From: Marc.Poinot at onera.fr (Marc Poinot)
Date: Tue, 25 Apr 2000 11:33:36 +0200
Subject: how to wrap c++ software to python?
References: <8e3ggq$2r1d$1@news.cz.js.cn>
Message-ID: <39056670.7DB4D09F@onera.fr>

fmr wrote:
> 
> I will call c++ routine in python,how to wrap
> c++ routine for python.
> 
> Is there any utils to work except SWIG.
> 
Have a try with:

http://sourceforge.net/project/?group_id=3180

Marcvs [alias But don't use add_keyword_method ;)  ]



From sp00fD at yahoo.com  Tue Apr 18 05:00:48 2000
From: sp00fD at yahoo.com (sp00fD)
Date: Tue, 18 Apr 2000 09:00:48 GMT
Subject: Using python on the web
References: <MPG.1365adcfa8470eef9896aa@news>
Message-ID: <8dh87r$3b$1@nnrp1.deja.com>

It's pretty easy really, to get the basics working at least.  In
general, this is about what you'd need

# For receiving data...
import cgi
query = cgi.FieldStorage()
# For each parameter we expect to be passed in we do
if query.has_key("ourparam"):
       result = query["ourparam"].value

# For printing pages, you could use the cgi module or just do it by hand
print "Content-type: text/html\n"
print "<HTML><HEAD><TITLE>Foobar</TITLE></HEAD>"
print "<BODY>"
import os
for k,v in os.environ.items():
        print "<B>%s => %s</B><BR>" % (k,v)
print "</BODY></HTML>"

For the database interaction, if you were using MySQL for instance, you
need to get the MySQLdb module
(http://starship.python.net/crew/adustman), build it, and then:

import MySQLdb
c = MySQLdb.connect(host="myhost", db="mydb", port=3316, user="me",
passwd="mypass")
con = c.cursor()
sql = "SELECT * FROM foo"
con.execute(sql)
rs = con.fetchall() # returns a tuple or list
c.close()


In article <MPG.1365adcfa8470eef9896aa at news>,
  Charley Horse <someone at somewhere.com> wrote:
> I've just started fooling with python. I've looked around the python
> site and a number of the other sites referenced there and am a bit
> puzzled. Umm... how does one go about using python for web
applications,
> especially database driven apps? Almost everything deals with python
> just as a language, as C++ would be. Not much about the web
> specifically. I was a bit shocked to find the only direct refs on
about
> this at python.org are entwined with CGI (unless I missed it). I have
no
> exposure to CGI, but rather to php, CF, and ASP, so python has gone
from
> looking very attractive to looking... well, like a puzzle. I know
about
> Zope, but am interested in how one uses python on the web just by
> itself.
>
> Thanks
>


Sent via Deja.com http://www.deja.com/
Before you buy.


From mwh21 at cam.ac.uk  Wed Apr 19 14:05:16 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 19 Apr 2000 19:05:16 +0100
Subject: modifying os.environ
References: <p6qhfczo50y.fsf@informatik.hu-berlin.de> <slrn8f9mn1.ven.quinn@zloty.ugcs.caltech.edu> <8di2il$10oq$1@nntp6.u.washington.edu> <slrn8frsu1.3bu.quinn@euro.ugcs.caltech.edu>
Message-ID: <m34s8yhspv.fsf@atrus.jesus.cam.ac.uk>

quinn at euro.ugcs.caltech.edu (Quinn Dunkan) writes:

> I do need the shell's useful things, and the system() commands are
> constants, so things should be pretty bulletproof as long as the
> environment is also constant, right?  But it *would* be useful to
> have a spawn() or run() or something builtin which does a proper
> vfork/exec with no subshell involved.  I often find myself writing
> one of those.

The time machine's got that one too:

os.spawnlp(os.P_WAIT,"ls","ls")

This will be 1.6.  No docstrings yet... mmm.

HTH,
Michael

-- 
42. You can measure a programmer's perspective by noting his 
    attitude on the continuing vitality of FORTRAN.
     -- Alan Perlis, http://www.cs.yale.edu/~perlis-alan/quotes.html


From basv at sara.nl  Tue Apr 11 08:26:21 2000
From: basv at sara.nl (Bas van der Vlies)
Date: 11 Apr 2000 12:26:21 GMT
Subject: sys.platform details...
References: <m3r9cdm6w4.fsf@devel.no.rwd.dinet>
Message-ID: <8cv5ld$8nm@mira.sara.nl>

In article <m3r9cdm6w4.fsf at devel.no.rwd.dinet>,
	rob at hooft.net (Rob W. W. Hooft) writes:
> 
> schol[51]:ccd>python
> Python 1.5.2b1 (#4, Jan 14 1999, 12:05:48)  [GCC egcs-2.90.21 971202 (egc on irix5
> Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>>> import sys
>>>> sys.platform
> 'irix5'
>>>> ^D
> schol[52]:ccd>uname -a
> IRIX schol 6.3 12161207 IP32

Rob,

 The following code works.

 import os

   platform = os.uname()
   print platform

		Hope This helps


-- 
********************************************************************
*                                                                  *
*  Bas van der Vlies                     e-mail: basv at sara.nl      *
*  SARA - Academic Computing Services    phone:  +31 20 592 8012   *
*  Kruislaan 415                         fax:    +31 20 6683167    *
*  1098 SJ Amsterdam                                               *
*                                                                  *
********************************************************************


From sabren at manifestation.com  Wed Apr 19 15:55:48 2000
From: sabren at manifestation.com (Michal Wallace (sabren))
Date: Wed, 19 Apr 2000 15:55:48 -0400 (EDT)
Subject: python templating system
In-Reply-To: <8dl0oc$7ct$1@nnrp1.deja.com>
Message-ID: <Pine.BSF.4.10.10004191555250.14584-100000@chi.pair.com>

On Wed, 19 Apr 2000 pozman at my-deja.com wrote:

> Does anyone know of a templating system (other than zope) that is
> implemented with python [ comparisons would be embperl with perl]
 
http://zebra.sourceforge.net/


Cheers,

- Michal
-------------------------------------------------------------------------
http://www.manifestation.com/         http://www.linkwatcher.com/metalog/
-------------------------------------------------------------------------




From thomas at xs4all.net  Mon Apr 17 12:48:33 2000
From: thomas at xs4all.net (Thomas Wouters)
Date: Mon, 17 Apr 2000 18:48:33 +0200
Subject: Editors and books
In-Reply-To: <XBFK4.16$Uq2.1831@news.uswest.net>; from jmassung@magpiesystems.com on Mon, Apr 17, 2000 at 09:00:56AM -0600
References: <XBFK4.16$Uq2.1831@news.uswest.net>
Message-ID: <20000417184833.H15664@xs4all.nl>

On Mon, Apr 17, 2000 at 09:00:56AM -0600, Jeff Massung wrote:

> Also, on another NG there's talk about Python, and how it does (and still
> doesn't) do garbage collecting. Could someone here expand on it so I know
> what it can and can't do? Memory is a big issue in my project and I need to
> know how to structure my program so that garbage is collected. Thanks
> again!!

Python doesn't do GC, it does RC, reference counting. That means that as
long as there is a reference to an object, it will not be collected. When
you delete a reference to an object, it's refcount goes done by one, and
when the count has dropped to 0, it's eligible for collection. (Note that
this doesn't mean it gets collected right away! It usually happens fairly
soon, though.)

The problem with RC is that you can create circular references:

a = [some large list]
b = [some other large list]
a.append(b)
b.append(a)

del a
del b

Now you can never refer to a and b and their contents again, but as they are
still referenced (by the other list) they can't be collected, either. So
avoid recursive constructs ;) It's fairly easy to fall into this trap when,
for instance, fiddling with tracebacks, though that isn't common practice
last time I looked.

There is work being done to cure this problem in a limited way -- not
all-encompassing, but enough to cure all but the most pathological case, I
gather. See the newsgroup archives ;)

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From phd at phd.russ.ru  Tue Apr 25 12:47:34 2000
From: phd at phd.russ.ru (Oleg Broytmann)
Date: Tue, 25 Apr 2000 16:47:34 +0000 (GMT)
Subject: Bookmarks database and Internet robot
Message-ID: <Pine.LNX.4.21.0004251644460.17533-100000@fep132.fep.ru>

Hello!

   I am happy to announce the third version of my project

                    BOOKMARKS database and internet robot!

WHAT IS IT
   This is a set of classes, libraries and programs I use to manipulate my
bookmarks.html. I like Netscape Navigator, but I need more features, so I
write these programs for my needs. I need to extend Navigator's "What's new"
feature (Navigator 4 names it "Update bookmarks").


WHAT'S NEW
   Complete rewrite from scratch. Created mechanism for pluggable storage
managers, writers and robots. This opens the possibility to invite more
programmers to work on different plugins. Some people already expressed
interest to use and extend these programs.
   All code had been rewritten in much more object-oriented fashion. I do not
value object orientation by itself, but for this project I found a point of
view where OO helped me to create much better code - simpler, smaller, a bit
faster and much more reusable.


WHERE TO GET
   Master site: http://sun.med.ru/~phd/Software/Python/#bookmarks_db

   Faster mirrors: http://skyscraper.fortunecity.com/unix/797/Software/Python/#bookmarks_db
   http://members.xoom.com/_XMCM/phd2.1/Software/Python/index.html#bookmarks_db


AUTHOR
   Oleg Broytmann <phd2 at earthling.net>

COPYRIGHT
   Copyright (C) 1997-2000 PhiloSoft Design

LICENSE
   GPL

STATUS
   Parser is Ok.
   Storage managers: pickle, FLAD (Flat ASCII Database).
   Writers: HTML, text, FLAD.
   Robots (URL checkers): simple, forking.

TODO
   More storage managers: shelve, SQL, MetaKit.
   More robots: threading, async.

Oleg.
---- 
  Oleg Broytmann http://www.fortunecity.com/skyscraper/unix/797/ phd2 at email.com
           Programmers don't die, they just GOSUB without RETURN.






From schorsch at schorsch.com  Thu Apr 13 04:23:01 2000
From: schorsch at schorsch.com (Georg Mischler)
Date: Thu, 13 Apr 2000 08:23:01 GMT
Subject: simple httplib and urllib timeout question
References: <Pine.BSF.4.10.10004111337520.29635-100000@chi.pair.com>
Message-ID: <8d404q$oee$1@nnrp1.deja.com>

Michal Wallace wrote:

> there's an example script called blogbot.py there, too.. it uses
> timeout.py and another hack with signal.alarm(), but neither really
> works all that well..

What exactly is your problem with signal.alarm() ?
It works very fine with me. Of course, I haven't tried this
on any M$ system, and it's not the right approach for
concurrent connections.
What I do is fairly simple:


def alarm_handler(signum, frame):
    '''what to do when an alarm signal reaches us'''
    raise socket.error, 'connection timed out'

USER_AGENT = 'DemoBot/0.0.7'
FETCH_TIMEOUT = 60 # in seconds

def fetch_request(req, host, path):
    '''fetch a page online with error handling and timeout,
       req is either 'GET' or 'POST' '''
    path = urllib.quote(path)
    if req == 'POST':
        path, form = string.split(path, '?')
    signal.signal(signal.SIGALRM, alarm_handler)
    signal.alarm(FETCH_TIMEOUT)
    try:
        connection = httplib.HTTP(host)
        connection.putrequest(req, path)
        connection.putheader('Host', host)
        connection.putheader(USER_AGENT)
        connection.putheader('Accept', 'text/html')
        if req == 'POST':
            connection.putheader('Content-Type',
                'application/x-www-form-urlencoded')
            connection.putheader('Content-Length', str(len(form)))
        connection.endheaders()
        if req == 'POST':
            connection.send(form)
        errcode, errmsg, headers = connection.getreply()
    finally: signal.alarm(0) # reset timeout
    return connection, errcode, errmsg, headers


The calling function will then read from the connection
and close it again. It will also catch any network related
exceptions, including the timeout.

All this assumes that there is only one connection in place
at any time, and you just don't want to wait overly long for
unresponsive hosts. If you need concurrent connections,
then asyncore is more powerful. You will have to maintain
a lot of state, but once you understand how all the different
parts of ansyncore work together it should be fairly easy to
get the individual timeouts to do the right thing.


Have fun!

-schorsch

--
Georg Mischler  --  simulations developer  --  schorsch at schorsch.com
+schorsch.com+  --  lighting design tools  --  http://www.schorsch.com/


Sent via Deja.com http://www.deja.com/
Before you buy.


From darcy at vex.net  Thu Apr 27 06:00:04 2000
From: darcy at vex.net (D'Arcy J.M. Cain)
Date: 27 Apr 2000 10:00:04 GMT
Subject: Best database to use with Python
References: <3906ea27.698463777@news.online.no>
Message-ID: <8e9334$qba$1@news.tht.net>

?5?HH575-UAZWKVVP-7H2H48V3 <thomas at cintra.no> wrote:
> What database-system is easiest, fastest etc. to use with Python? I`ve
> looked at MySQL, but it seems a little overkill just to get a little
> more functionality than Berkley DB-functionality. What about Postgres?
> mSQL? 

> Postgres comes with my Linux distro and seem smaller than MySQL. How

Actually I think you will find PostgreSQL larger than MySQL.

> does it compare to MySQL when it comes to administration and speed,
> scalability etc. ?

I don't use MySQL as I find that PostgreSQL has more features and works
fast enough for my needs.  Like most RDBMSs in its class the expensive
part is updates and inserts but the retrievals are quite snappy.  The
one time I tried to administer a MySQL system I was a little horrified.
PostgreSQL was easier to understand.  That could just be me, of course.

Doesn't mSQL have some funny licensing issues?

> I`m doing alot of index-lookups for a
> Yahoo/Altavista/WhatEverSearchEngine-like web-app.

> <entering_demand_mode>
> It have to be fast, capable of HUGE amounts of data, free and easy to
> install and administer.
> </entering_demand_mode>

Since PostgreSQL comes with your system I would stick with it unless you
find something in MySQL that you need that PostgreSQL doesn't have.

And don't rule out Berkley DB-functionality if your requirements are simple
enough.  It can be extremely fast, it's small and there are some nice
Python interfaces to it, particularly shelve.

-- 
D'Arcy J.M. Cain <darcy at vex.net>           |  Democracy is three wolves
http://www.vex.net/                        |  and a sheep voting on         
+1 416 425 1212     (DoD#0082)    (eNTP)   |  what's for dinner.


From alwagner at tcac.net  Thu Apr 13 07:40:32 2000
From: alwagner at tcac.net (Albert Wagner)
Date: Thu, 13 Apr 2000 06:40:32 -0500
Subject: redirecting stdout to printer
Message-ID: <38F5B230.2FEFB9BC@tcac.net>

How do I specify my printer as stdout?  Know of a module that has basic
prettyprinting with headings and pagenumbers?

-- 
Small is Beautiful


From pf at artcom-gmbh.de  Sun Apr  2 15:24:02 2000
From: pf at artcom-gmbh.de (Peter Funk)
Date: Sun, 2 Apr 2000 21:24:02 +0200 (MEST)
Subject: Who is this years winner?
Message-ID: <m12bpyc-000CnCC@artcom0.artcom-gmbh.de>

Today on April 2nd I began to wonder whether the PSA Steering Committee has already
decided on the winner of the third Pythonic Award.  Fortunately
www.python.org is back online now but unfortunately the page
	<http://www.python.org/PSAward.html>
seems to show up a file from two year old backup, and so does the 
Shrine of the of the winner of the first Python Award at
	<http://www.python.org/tim_one/>

May I kindly ask some member of the famous Steering Committee to publish
this years decision here?  

;-) Best regards, Peter.
-- 
Peter Funk, Oldenburger Str.86, D-27777 Ganderkesee, Germany, Fax:+49 4222950260
office: +49 421 20419-0 (ArtCom GmbH, Grazer Str.8, D-28359 Bremen)



From banderson at boi.hp.com  Mon Apr 17 15:07:50 2000
From: banderson at boi.hp.com (Bill Anderson)
Date: Mon, 17 Apr 2000 13:07:50 -0600
Subject: Scripting and Gnome and KDE
References: <7t0ffs8ncrrh3irgc52rbsb68ue48rfovf@4ax.com>
Message-ID: <38FB6106.7E3BFC1D@boi.hp.com>

Alex Thomas wrote:
> 
> This might seem an odd way to choose a GUI, but if you had the choice
> between scripting Gnome/Gtk or KDE/Qt-based components using Python, which
> would you go for?
> 
> Thanks peeps
> Alex

The biggest problem I have with the bindings I have seen (pyFLTK ..
currently using this one, pyGTK/pyGNOME,pyQt) is that the documentation
assumes you use the tookit's original documentation. Many pythoneers
don't use C or C+fuss. In order to understand the bindings, you have to
understand the original toolkit's native language. That alone has put
off many aspiring pythoneers, I can attest to.




-- 
Bill Anderson 			Linux/Unix Administrator, Security Analyst
ESBU (ARC)			bill_anderson at boi.hp.com
My opinions are just that; _my_ opinions.


From phd at phd.russ.ru  Wed Apr 26 04:21:46 2000
From: phd at phd.russ.ru (Oleg Broytmann)
Date: Wed, 26 Apr 2000 08:21:46 +0000 (GMT)
Subject: Working with data returned by PyGreSQL
In-Reply-To: <itx6ni3w.fsf@jumbo.skott.dk>
Message-ID: <Pine.LNX.4.21.0004260820560.23026-100000@fep132.fep.ru>

On Tue, 25 Apr 2000, Martin Sk?tt wrote:
> do connects and querys). My problem is how do I get to anything other
> than print my query results? What datatype does the query return? I

   http://sun.med.ru/~phd/Software/Python/#rgb_example

   There are source tarball and running demo. Welcome.

Oleg.            (All opinions are mine and not of my employer)
---- 
    Oleg Broytmann      Foundation for Effective Policies      phd at phd.russ.ru
           Programmers don't die, they just GOSUB without RETURN.




From neelk at brick.cswv.com  Mon Apr  3 22:08:18 2000
From: neelk at brick.cswv.com (Neel Krishnaswami)
Date: 4 Apr 2000 02:08:18 GMT
Subject: [Python-Dev] New Features in Python 1.6
References: <200004011740.MAA04675@eric.cnri.reston.va.us> <200004031604.MAA05283@eric.cnri.reston.va.us>  <20000403165621.A9955@cnri.reston.va.us>
Message-ID: <slrn8eilnl.jqh.neelk@brick.cswv.com>

Greg Ward <gward at cnri.reston.va.us> wrote:
> 
> Hasn't anyone noticed that the largest amount of text in the joke
> feature list was devoted to the Distutils?  I thought *that* would give
> it away "fer shure".  You people are *so* gullible!  ;-)

Hey! You actually fooled me for a good long while. Until I started
thinking, "M4? I didn't think distutils used a lot of M4." Then I
checked the web page, then I checked the date, and then I went,
"Oh. Duh."


Neel


From paulb at infercor.no  Wed Apr 19 06:36:38 2000
From: paulb at infercor.no (Paul Boddie)
Date: Wed, 19 Apr 2000 12:36:38 +0200
Subject: Using CGI module with 'canned queries'
References: <200004190828.KAA03403@gromit.rwth-aachen.de>
Message-ID: <38FD8C36.4242425F@infercor.no>

Lars von Wedel wrote:
> 
> Hello,
> 
> I want to use Python to implement some CGI scripts. I want these
> scripts to be callable from any link (not only via a form),
> e.g. the user shall be able to enter something like
> 
> http://.../test.py?name=Bill
> 
> into Netscape directly.

The parameter(s) should still be received correctly in the CGI script.

> I expected that such a URL could be processed using
> 
>     form = cgi.FieldStorage()    print '<UL>'
>     for k in form.keys:
>         print '<LI>' % k % ':' % form[k] % '</LI>'

Surely you mean...

          print '<LI>' + k + ':' + form[k].value + '</LI>'

> (after printing an html header of course).
> 
> However, the dictionary seems to be empty when the http daemon
> starts my script.

Are you sure that the problem doesn't lie in the fact that form[k] is not likely
to be a string, but a FieldStorage object instead? Thus, you must refer to the
value attribute of form[k].

> How do I access parameters within the script if not started via a form?

You should be able to do just that with the 'cgi' module.

Paul


From jddst19 at pitt.edu  Tue Apr  4 11:22:15 2000
From: jddst19 at pitt.edu (John Duncan)
Date: Tue, 4 Apr 2000 11:22:15 -0400
Subject: Smalltalk on the small end (was: Advice requested: GUI project beginning)
References: <m12b4OW-000wcEC@swing.co.at> <8D79BC9E4960ED70.7B6609342D6EB723.7625773D14AD9140@lp.airnews.net> <ESaG4.9536$9g4.372176@news5.giganews.com>
Message-ID: <8cd1c2$3ut$1@usenet01.srv.cis.pitt.edu>

> The point is that there's little point to "scripting" in Smalltalk
> when it doesn't really get powerful 'til you build up a rich set of
> functionality in a "Smalltalk Environment."  And *that* is something
> you don't want to respawn repeatedly...

Hmm...  This brings up an interesting conflict to the term "general-purpose
programming language."  I have seen this term used quite often, usually
meaning that a language is turing-complete and allows access to system
structures.  But when it comes down to it, there are no general-purpose
languages.  The market thinks so, but anyone who programs in Smalltalk knows
better.  As Squeak evolves, people often ask, "Can Squeak do this feature of
some other system?"  The Squeak team says, "No, but no-one's stopping you."
If the person goes ahead and does it, something like implementing native
widgets or scripting support for a certain OS, the response, even by the
implementor, is usually, "Why did I care in the first place?"  Simply put,
it's much easier to let environments that are good at one thing do that
thing, rather than trying to make a single environment do everything.  For
the same reason, I don't understand why people still write user interfaces
in C++.  Or even Java for that matter.  The tools are simply not advanced
enough to do it well.  But, you do it in ST, and you have something quick
and elegant.  Similarly, you have people trying to make Smalltalk do hard
realtime systems.  Why?  To me, this is kind of like implementing Linux on
an Atari 800.  It has geek value but there's still a much better tool for
the job.

-John




From drew.csillag at starmedia.net  Fri Apr 28 10:52:46 2000
From: drew.csillag at starmedia.net (Andrew Csillag)
Date: Fri, 28 Apr 2000 10:52:46 -0400
Subject: How to use DCOracle and LOBs in Oracle 8i ????
References: <20000428124808.A21389@sz-sb.de>
Message-ID: <3909A5BE.277160BC@starmedia.net>

Andreas Jung wrote:
> 
> After reading old postings concerning DCOracle and LOBs I
> can not solve the following problem:
> 
> I have  a table under Oracle 8i with a NCLOB field. The base
> character set for the complete database it UTF8. My first
> problem with DCOracle was an ORA-12704 character set mismatch error when
> I tried to insert an instance of DCOracle.dbi.dbiRaw() in to the database.
> 
> My current code looks like this (the table contains the NCLOB field 'content'):
> 
> lob = 'very long text .....'
> raw = DCOracle.dbi.dbiRaw(lob)
> 
> com = 'insert into tab (docnum,content) values(:x,translate(:y using nchar_cs))'
> curs.execute(com,x='1234',y=raw)
> 
> This works for LOBs with less than 4000 bytes. For longer LOBs I get the following error
> message: ORA-01461 can bind a LONG value only for insert into a LONG column.
> 
> Now...how can I store larger LOBs with more than 4000 bytes ? I have read something
> about a dbms_lob package but I have really no idea how to use it.
> 
> Has anyone a working solution or a code example ?!
> 

What you have to do on insert or update is something like this:
a table defined as such:
create table foo_table (pkoid NUMBER, blobval BLOB)

conn = DCOracle.connect('user/pw')
cur = conn.cursor()
cur.execute('insert into foo_table (pkoid, blobval) values (1,
EMPTY_BLOB())')
cur.execute('select blobval from foo_table where pkoid = 1 for update')
blob = cur.fetchone()[0] #can't use fetchall or is blobval is converted
to string!
blob.write(stringToPutIntoBLOB)
conn.commit()

The weird INSERT/SELECT dance is just the way Oracle works, and not a
problem so much with DCOracle.  To do an UPDATE, you update the lob row
with EMPTY_BLOB() and do the select for update dance again.

Unfortunately, there seems to be an issue with dealing with more than
one lob at a time in terms of read()ing and write()ing.  So if you're
only dealing with one, that's ok, but if you try to deal with more than
one at a time, you need to play with it a bit to see if it will work. 
This may not normally be a problem, but there is a hidden issue: 
cursor.fetchall() won't work if there is more than one lob in the result
set that isn't empty (it tries to stringify lobs on the result rows), so
you have to use cursor.fetchone() to cycle through the result rows
(which is probably what you'd do anyhow, so it's not such a big deal,
but one to be aware of in any case).

Hope this helps,
Drew
-- 
print(lambda(q,p):'pmt:$%0.2f'%(q*p/(p-1)))((lambda(a,r,n),t:(a*r/
t,pow(1+r/t,n*12)))(map(input,('amt:','%rate:','years:')),1200.0))



From dan at cgsoftware.com  Wed Apr 12 17:31:11 2000
From: dan at cgsoftware.com (Daniel Berlin+list.python)
Date: 12 Apr 2000 17:31:11 -0400
Subject: Python/Perl Popularity (Re: A Mountain of Perl...)
In-Reply-To: pj@sgi.com's message of "12 Apr 2000 21:16:42 GMT"
References: <200004102111.HAA03380@envy.fulcrum.com.au> <8d0pcq$7ti$1@nnrp1.deja.com> <8d1inu$nsq$1@newshost.accu.uu.nl> <38F4996F.E3BD2626@acm.org> <8d2p3q$2vj9o$1@fido.engr.sgi.com>
Message-ID: <snwrf1ls.fsf@dan.resnet.rochester.edu>

>>>>> "Paul" == Paul Jackson <pj at sgi.com> writes:

Nope, both lynx, and links, exist.


    Paul> BobC wrote: |> The text-mode browser named "links"

    Paul> I suspect that you meant "lynx" ??

    Paul> --




From alwagner at tcac.net  Thu Apr 27 09:20:11 2000
From: alwagner at tcac.net (Albert Wagner)
Date: Thu, 27 Apr 2000 08:20:11 -0500
Subject: The Simple Economics of Open Source
References: <B529350E.51B9%raffael@mediaone.net> <390534BB.3D6CEBDF@libc.org> <raffael-01CD89.20575826042000@news.ne.mediaone.net>
Message-ID: <39083E8B.25845D83@tcac.net>

Raffael Cavallaro wrote:
> 
<snip>
> I think that many open source advocates have
> failed to apply simple logic to the economics of open source, believing
> that somehow, the internet, or software, is immune from simple laws of
> supply and demand. They are not.
> 
I am going to have to go back and reread all of your previous posts. 
You seem to be rebutting a point that was never made.  Of course, Open
Source software is not immune to the laws of supply and demand.  Who
said it was?  It is in response to those laws that it is free.  But free
only in the world of economics where dollar worth is all that matters.
It is this very world, where the only motivation taken into account is
money profit, that Open Source has escaped.  And the escape has been
dazzling.  
---
"But beware!  The time for all this is not yet.  For a least another
hundred years we must pretend that fair is foul and foul is fair;  for
foul is useful and fair is not.  Avarice and usury and precaution must
be our gods for a little longer still.  For only they can lead us out of
the tunnel of economic necessity into daylight."
-- Lord Keynes, 1930


From kern at caltech.edu  Thu Apr 13 15:34:49 2000
From: kern at caltech.edu (Robert Kern)
Date: 13 Apr 2000 19:34:49 GMT
Subject: Simple formatting string question
References: <meh9-637743.14563813042000@news.cit.cornell.edu>
Message-ID: <8d57gp$1om@gap.cco.caltech.edu>

In article <meh9-637743.14563813042000 at news.cit.cornell.edu>,
	Matthew Hirsch <meh9 at cornell.edu> writes:
> Hi All,
> 
> I'm trying to print something like this:
> 
>>>> print 6*'%2i' '%1i' % tuple([5,5,5,5,5,5,1])
> 
> but I'm getting an error.

"""
TypeError: not enough arguments for format string
"""
specifically.

The problem is because of the following behavior:

>>> 6*'%2i' '%1i'
'%2i%1i%2i%1i%2i%1i%2i%1i%2i%1i%2i%1i'

The implicit string concatenation happens before the multiplication.  Add a
+ between the strings first, and the normal order of operations will do what
you want.

GuruQuestion: 
Should implicit string concatenation have a different precedence than
explicit string concatenation?  Perhaps only a mention in
http://www.python.org/doc/current/ref/string-catenation.html ?

> Do you know what the correct syntax should be?
> 
> Thanks,
> Matt

-- 
Robert Kern
kern at caltech.edu

"In the fields of hell where the grass grows high
 Are the graves of dreams allowed to die."
  -- Richard Harter


From hamish_lawson at yahoo.co.uk  Thu Apr 13 11:09:29 2000
From: hamish_lawson at yahoo.co.uk (Hamish Lawson)
Date: Thu, 13 Apr 2000 08:09:29 -0700
Subject: Building MySQLdb on Solaris gives PyTuple_New not found
Message-ID: <0ef708b6.3904ad2d@usw-ex0108-063.remarq.com>

When I try to build MySQLdb-0.2.0 on Solaris 2.6 I get the
following relocation error about PyTuple_New:

------------------------------------------------------------

gcc   -I/usr/local/mysql/include -g -O2 -
I/usr/local/include/python1.5 -I/usr/lo
cal/include/python1.5 -DHAVE_CONFIG_H -c ./_mysqlmodule.c

ld -G  _mysqlmodule.o  -L/usr/local/mysql/lib -lmysqlclient -o
_mysqlmodule.so

Whoa, couldn't import _mysql: ld.so.1: python: fatal: relocation
error: file ./_
mysqlmodule.so: symbol PyTuple_New: referenced symbol not found

------------------------------------------------------------

Any suggestions?

Hamish Lawson


* Sent from RemarQ http://www.remarq.com The Internet's Discussion Network *
The fastest and easiest way to search and participate in Usenet - Free!



From paul.magwene at yale.edu  Wed Apr 12 11:51:26 2000
From: paul.magwene at yale.edu (Paul Magwene)
Date: Wed, 12 Apr 2000 11:51:26 -0400
Subject: No method overloadin I suppose ?
References: <38F4938E.8F245B68@altern.org>
Message-ID: <38F49B7E.A68D91C1@yale.edu>

Thomas SMETS wrote:
> 
> ## Beginning of Python program ...
> class Lister:
>   def __init__(self):
>     print 'In constructor ...'
>     __init__(1)
> 
>   def __init__(self, lValue):
>     print 'In constructor with parameter'
>     self.l = lValue
> 
>   def __repr__(self):
>     print '\n******\n\tValue is : ', l, '\n********'
> 
> if __name__ == "__main__":
>   x = Lister()
>   y = Lister (2)
> ## End of python program
> 
> Trace of execution is :
> Traceback (innermost last):
>   File "Minheritence.py", line 14, in ?
>   x = Lister()
> TypeError: not enough arguments; expected 2, got 1
> End of trace ...
> 
> I guess I can't do methods overloadin', as shown... WHY is that ? I'm
> wrong ?

You can't overload methods as you would in C++, but the following
accomplishes the same thing.  You could also test for the types of
parameters using the function type() and the types module.

--Paul


class Lister:
  def __init__(self,lValue=None):
    if lvalue:
        print "Constructed with parameter"
        self.l = lValue
    else:
        print "Default construction"
        self.l = 1

  def __repr__(self):
    print '\n******\n\tValue is : ', l, '\n********'


From hdemers at venus.astro.umontreal.ca  Thu Apr  6 14:17:04 2000
From: hdemers at venus.astro.umontreal.ca (Hugues Demers)
Date: Thu, 06 Apr 2000 18:17:04 GMT
Subject: Beginning to be disapointed
Message-ID: <Aw4H4.8078$pR.68745@carnaval.risq.qc.ca>

Hi,

Can somebody help me on this? I'm using python to read big arrays (100x2000)
of float. I manipulate those arrays, I do small operations on them. That is
ok. I do others smalls calculations not involving the big arrays and then the
interpreter hang ! The cpu percentage used by python goes up and there is
nothing else to be done except to kill the interpreter. I have no idea
what's going on. Do I need to free memory in some way ? Am I doing
something wrong ? There is nothing special about the calculations involving
the big arrays. The only thing special is that they are big.


Thank you

Hugues


From effbot at telia.com  Mon Apr  3 11:56:11 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Mon, 03 Apr 2000 15:56:11 GMT
Subject: Tuples -- who needs 'em
References: <00e101bf9d80$a8384d20$74eb0b18@stcla1.sfba.home.com>
Message-ID: <va3G4.1110$n68.190454272@newsb.telia.net>

Bob Alexander wrote:
> This is not an [intentional] flame bait, nor an April Fool's message, but
> think about it: Python has two "sequence" data types, lists and tuples,
> which from the Python programmer's perspective are only slightly
different.
> How does having both types significantly improve life for Python
> programmers?

some random remarks (please consider them all before
responding):

-- http://www.python.org/doc/FAQ.html#6.15

-- things should be as simple as possible, but no simpler.

-- from a human perspective, using different syntax for
   different things is good.   tuples are simple structures.
   lists are collections of homogenous (in one sense or
   another) values.  from a program design perspective,
   that are two radically different things.

-- adding a comment won't make an object immutable.

-- tuples can be used as dictionary keys.  mutable
   collections can not.

-- python's core types fall into two categories: the first
   is simple types, like numbers, strings, and tuples. they
   are all immutable.

   the other category is container (or collection) objects.
   they're far more powerful, and harder to use for beginners.

-- this has been discussed over and over again.  have you
   studied earlier threads on this topic?

-- good designers know that hypergeneralization tends to be
   a bad idea.

-- what would happen if we took them away?

-- do we really need yet another "I don't fully understand this,
   so it should go away" thread?  do you really think that tuples
   are just an accidental feature?  (just asking...)

> Suppose Python had only one sequence data type: list, and that we could
use
> it in all places where tuples are needed now. I would never have to think
> about whether to use tuple or list. I would never have to convert from
tuple
> to list or vice versa. Just one more thing I don't have to bother with.

if you find that you end up doing this all the time, you're not
following the "fixed structures vs. homogenous collections"
design rule.

(or you're using extension modules that haven't been upgraded
to use the abstract sequence API.  most standard extensions do,
these days).

if you don't do this all the time, what's the big deal?

> I suspect there will be several answers offered as to why we need both
> types, but I also suspect most of the benefits are for the Python
> implementor(s), not programmers.

umm.  I thought having two similar types meant more work for
the python implementors, not less?

</F>




From phd at phd.russ.ru  Thu Apr 27 10:11:52 2000
From: phd at phd.russ.ru (Oleg Broytmann)
Date: Thu, 27 Apr 2000 14:11:52 +0000 (GMT)
Subject: About Zope`s FTP-server
In-Reply-To: <39084411.13519730@news.online.no>
Message-ID: <Pine.LNX.4.21.0004271411060.11789-100000@fep132.fep.ru>

On Thu, 27 Apr 2000, 5?HH575-UAZWKVVP-7H2H48V3 wrote:
> found. But my database inside Zope was bigger. This is fine and nice
> for small pics and html-stuff, but I want to upload pickled objects
> that can be several megabytes in size, then have a Zope-like or
> Zope-integrated method store them in some database, NOT gadfly, cuz

   Use external DB - BerkeleyDB or SQL.

Oleg.            (All opinions are mine and not of my employer)
---- 
    Oleg Broytmann      Foundation for Effective Policies      phd at phd.russ.ru
           Programmers don't die, they just GOSUB without RETURN.




From gmcm at hypernet.com  Wed Apr  5 16:02:33 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Wed, 5 Apr 2000 16:02:33 -0400
Subject: Pass by reference ?
In-Reply-To: <38EB838A.53122AF6@acm.org>
Message-ID: <1257157531-11108477@hypernet.com>

Robert W. Cunningham wrote:

> The CS models of "Pass By Reference" and "Pass By Value" are fairly well understood,
> but they do not seem to map simply and directly to Python.  

Grrr. Pascal causes brain damage. Python is exactly like Java 
(with the exception of having no primitive types).
 
These notions should state *what* is being passed. In Pascal, 
that's "value of a slot" or "reference to a slot". To be precise, 
in Python it is *always* "value of a reference", (think "pointer 
value", not "value of that which is pointed to").

For all of Herr Werth's genius, he bamboozled millions with his 
little sleight of hand.



- Gordon



From parkw at better.net  Wed Apr 12 13:41:15 2000
From: parkw at better.net (William Park)
Date: Wed, 12 Apr 2000 13:41:15 -0400
Subject: global variables - how??
In-Reply-To: <8d2bne$qn7$1@news101.telia.com>; from aek@mail1.stofanet.dk on Wed, Apr 12, 2000 at 07:28:37PM +0200
References: <8d2bne$qn7$1@news101.telia.com>
Message-ID: <20000412134115.A9193@better.net>

On Wed, Apr 12, 2000 at 07:28:37PM +0200, Anders Eggers-Krag wrote:
> how do I define a propper global variable that works in all functions in all
> modules?

As far as I know, you can't.  You have to use 'global ...' inside all
your functions.

--William

> 
> I have tried simply
> defining one outside of any function in the main module,
> I have tried using the global keyword under the same cicumstances,
> but they are not even avialible inside of functions within the main
> module...
> 
> 
> --
> Anders Eggers - Krag
> 
> 
> -- 
> http://www.python.org/mailman/listinfo/python-list



From Ted.Horst at wdr.com  Fri Apr  7 11:06:07 2000
From: Ted.Horst at wdr.com (Ted Horst)
Date: Fri,  7 Apr 2000 10:06:07 -0500
Subject: [Python-Dev] Round Bug in Python 1.6? (str vs. repr)
In-Reply-To: <8cknsa$1r4$1@slb6.atl.mindspring.net>
References: <38ED0016.E1C4A26C@tismer.com>
	<200004062130.RAA26273@eric.cnri.reston.va.us>
	<20000407084914.A13606@mems-exchange.org>
	<8cknsa$1r4$1@slb6.atl.mindspring.net>
Message-ID: <200004071506.AA18015@ch1d2833nwk>

On Fri, 7 Apr 2000, Greg Ward wrote:
> On 07 April 2000, Mikael Olofsson said:
> >
> > On 07-Apr-00 Greg Ward wrote:
> >  > Oh, joy! oh happiness! someday soon, I may be able to type
> >  > "blah.__doc__" at the interactive prompt and get a readable result!
> >
> > Just i case... I hope you haven't missed "print blah.__doc__".
>
> Yeah, I know: my usual mode of operation is this:
>
>   >>> blah.__doc__
> ...repr of docstring...
> ...sound of me cursing...
>   >>> print blah.__doc__
>
> The real reason for using str() at the interactive prompt is not to save
> me keystrokes, but because it just seems like the sensible thing to do.
> People who understand the str/repr difference, and really want the repr
> version, can slap backquotes around whatever they're printing.
>
> Greg
>
> --
> <http://www.python.org/mailman/listinfo/python-list

I share your frustration with __doc__, but I really prefer to see repr in  
the interactive interpreter.  The reason is that I like to know what kind of  
object I am looking at.  I have a tendency to call functions or methods and  
look at what they return before I decide to assign them to a variable (with  
x = _).  This technique would be less useful if I can't tell by looking at  
the return value what it is.  This is especially true for strings where you  
can see the quote.

Maybe its just me, though.

-----------------------------
Ted Horst <Ted.Horst at wdr.com>



From anthony at interlink.com.au  Sun Apr 16 23:49:49 2000
From: anthony at interlink.com.au (Anthony Baxter)
Date: Mon, 17 Apr 2000 13:49:49 +1000
Subject: Python w/ SSL 
In-Reply-To: Message from Jp Calderone <exarkun@flashmail.com> 
   of "Sun, 16 Apr 2000 21:44:48 GMT." <38FA3450.1DD845B2@flashmail.com> 
Message-ID: <200004170349.NAA03160@mbuna.arbhome.com.au>

Try http://mars.post1.com/home/ngps/m2/


>>> Jp Calderone wrote
> 
>  Can anyone point me to an *up to date* Python/SSL interface?  I've
> found about 4 w/ search engines, but the most recent is from 1996 and
> has a link to Python 1.3 in case you are without an interpreter ;)
> (It also doesn't help that starship is still down - again :( )
> 
>  Thanks in advance
>  Jp
> 
> 
> -- 
> If the automobile had followed the same development as the computer, a 
> Rolls-Royce would today cost $100, get a million miles per per gallon,
> and explode once a year killing everyone inside.
>                 -- Robert Cringely, InfoWorld
> --
>  9:43pm up 19 days, 2:50, 0 users, load average: 0.02, 0.07, 0.09
> -- 
> http://www.python.org/mailman/listinfo/python-list
> 

-- 
Anthony Baxter     <anthony at interlink.com.au>   
It's never too late to have a happy childhood.



From ivanlan at callware.com  Wed Apr 12 14:49:16 2000
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Wed, 12 Apr 2000 12:49:16 -0600
Subject: Off-topic
References: <38F4BE12.A0C73D1D@callware.com> <m3zoqzgois.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <38F4C52C.450F611F@callware.com>

Hi All--

Michael Hudson wrote:
> 
> Ivan Van Laningham <ivanlan at callware.com> writes:
> 
> > Hi All--
> > This is not precisely about Python, but it does affect my *usage* of
> > Python on redhat 6.1.
> >
> > Why don't #! scripts seem to work anymore?  #!/usr/bin/python and
> > #!/usr/bin/perl scripts don't work.  Error is
> 
> Someone's taken /usr/bin/python away?  Tried strace-ing it?
> 

Perl and Python both work if I do 'perl h2n' or 'python hello.py'. 
Interactive mode works fine.

> I presume /bin/sh scripts are still working (wouldn't like to think
> what happened if they went away).
> 

Sh scripts do work, as long as I remember to do ./script ;-).  I'm
running as root (but the same thing happens when I run as myself, too). 
However, this could be explained by the default behavour of exec(),
which states that if the kernel can't run a program, it execs sh and
hands it the program name in argv.


>   incidentally, asking why things are "left out of the language" is
>   a good sign that the asker is fairly clueless.

<but-what-about-things-left-out-of-the-kernel?>-ly y'rs,
Ivan
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
http://www.pauahtun.org 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
Author:  Teach Yourself Python in 24 Hours



From mwh21 at cam.ac.uk  Wed Apr 12 06:35:04 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 12 Apr 2000 11:35:04 +0100
Subject: Classes, Inheritance - Stupid lazy question
References: <8d1cfh$hme7@imsp212.netvigator.com> <874s9766r6.fsf@ifm.uni-kiel.de> <m3itxnis32.fsf@atrus.jesus.cam.ac.uk> <8d1ihl$q244@imsp212.netvigator.com>
Message-ID: <m3g0srip47.fsf@atrus.jesus.cam.ac.uk>

"Paul Gresham" <gresham at mediavisual.com> writes:

> I like this solution a lot, it makes a lot of sense. Thanks Michael.
> I guess somewhere in Python is a defined route back to the super
> class, on a better day I may just dig through the code.

For a variey of reasons, class heirachies in Python are usually pretty
shallow, so this isn't often too much of a problem.  I think the
median depth of class heriachies in code I right is probably 1...

Cheers,
M.

-- 
  On the other hand,  the following areas are subject to boycott
  in reaction to the rampant impurity of design or execution, as
  determined after a period of study, in no particular order: 
    ...                            http://www.naggum.no/profile.html


From gmcm at hypernet.com  Thu Apr 20 08:05:39 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Thu, 20 Apr 2000 08:05:39 -0400
Subject: socket.SO_REUSEADDR
In-Reply-To: <38FEEA44.6C05E2B0@inka.de>
Message-ID: <1255890155-29764878@hypernet.com>

Michael Str?der wrote:
 
> For avoiding problems with errors like
> 
>   socket.error: (98, 'Address already in use.')
> 
> during immediate restart of my web server I'm following the hint to
> set the socket option SO_REUSEADDR. But I did not succeed. The
> server starts and works. When I stop it and try to restart
> immediately the above message appears again. I had a short look at
> man 7 socket and man setsockopt but this was a little bit too much
> for me...

> My code looks like this:
> 
> ------------------------ bite here -------------------------
> class MyHTTPServer(SocketServer.TCPServer):
> 
>   def server_bind(self):
>     """Override server_bind to set socket options."""
>     self.socket.setsockopt(
>       socket.getprotobyname('tcp'),
>       socket.SO_REUSEADDR,1
>     )
>     SocketServer.TCPServer.server_bind(self)
> ------------------------------------------------------------

You need socket.SOL_SOCKET, not 
socket.getprotobyname(). But there's still no substitute for a 
try / finally that closes the socket explicitly.




- Gordon



From andreas.krueger at uumail.de  Tue Apr 18 13:34:03 2000
From: andreas.krueger at uumail.de (Andreas Krüger)
Date: Tue, 18 Apr 2000 19:34:03 +0200
Subject: need a script to set IP-Adress an machine name
Message-ID: <38fc9ce6$0$18302@personalnews.de.uu.net>

i'm looking for a script (or parts) to set up an nt-client nic-parameters.

it mus call for ip-adress and name, and set the network-parameter automatic
after re-image.
also it must set commandline-parameter and copy four files.
later i will call for the mac-adress and call the parameters from a database
(access or d-base)

does anybody know how to do this? i'm an absolute beginner for programming.

thanks, andreas

please repost a copy to andreas.ad.krueger at bku.db.de




From c.evans at clear.net.nz  Wed Apr 26 05:13:08 2000
From: c.evans at clear.net.nz (Carey Evans)
Date: Wed, 26 Apr 2000 09:13:08 GMT
Subject: python-only crypt()?
References: <Pine.BSF.4.10.10004230627580.5642-100000@chi.pair.com>
	<39065A08.F776CBE1@home.com>
Message-ID: <87d7nd444b.fsf@psyche.evansnet>

bo Vandenberg <famvandenberg at home.com> writes:

> "a pure Python implementation of the Unix crypt(3) algorithm"
> 
> http://home.clear.net.nz/pages/c.evans/sw/

Wow.  I finished writing that and uploaded it less than six hours
before your post.  Would you mind letting me know how you found it so
quickly?

> Post back if this is helpful

Given that I ported it because I couldn't get Michal's port of
UnixCrypt going, it probably is.

-- 
	 Carey Evans  http://home.clear.net.nz/pages/c.evans/

"For some of us, it's *always* Monday morning."  - Simon Cozens in asr


From gregm at iname.com  Thu Apr 20 10:34:59 2000
From: gregm at iname.com (Greg McFarlane)
Date: Fri, 21 Apr 2000 00:34:59 +1000
Subject: focus with PMW
In-Reply-To: <38FCCAC0.8868ED95@mitre.org>; from Pete on 18 Apr 2000 at 04:51:12PM
References: <38FC75F4.D4E01F19@mitre.org> <38FCCAC0.8868ED95@mitre.org>
Message-ID: <20000421003459.01994@nms.otc.telstra.com.au>

On 18 Apr, Pete wrote:
> I am trying to determine which widget,  the focus is currently on in my
> code.  When I make a call to get_focus() it gives me an instance of a
> TKinter.Entry,  but I am only creating PMW widgets (which are of course
> extended from TKinter).  Since I am not creating TKinter widgets
> explicitly in my own code,  how can I determine which PMW widget that
> specific TKinter.Entry instance belongs to?

You will need to maintain a list of all the Tkinter widgets in your
application that can get focus and compare them against the one
returned by focus.get().  You can get the component widgets of a
Pmw megawidget by using the component() method. For example
    entryfield = Pmw.EntryField()
    entry = entryfield.component('entry')
The Pmw documentation lists all the components of each megawidget,
along with their types.

However, would it not be better to use something like
bind('<FocusIn>') on all the widgets (or component widgets of
megawidgets) you are interested in, then your application can be
notified whenever focus changes, rather than having to ask for it.

-- 
Greg McFarlane     INMS Telstra Australia     gregm at iname.com



From m.faassen at vet.uu.nl  Wed Apr 12 07:09:18 2000
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 12 Apr 2000 11:09:18 GMT
Subject: Python/Perl Popularity (Re: A Mountain of Perl...)
References: <200004102111.HAA03380@envy.fulcrum.com.au> <8d0pcq$7ti$1@nnrp1.deja.com> <g0ssm03o.fsf@dan.resnet.rochester.edu>
Message-ID: <8d1lgu$ga0$1@newshost.accu.uu.nl>

Daniel Berlin+list.python <dan at cgsoftware.com> wrote:

>     tom>  -- Is there a Python equivalent of CPAN and the Perl CPAN
>     tom> module?  The closest I know of is Parnassus.  But Parnassus
>     tom> is merely a collection of links, not an archive, and it
>     tom> doesn't have any facilities (AFAIK) for automatic
>     tom> installation.

> So, who gives a crap?
> I try not to be that lazy.

I care, for instance. The distutils people care. Please don't assume
nobody cares when there's a whole SIG about this issue. 

It's not only about laziness (though that's a good trait in a programmer :),
it's about accessibility. Python modules should be very easy to get
and install. This makes Python more useable by people who don't know
Python, such as sysadmins, and users. It also is good for Computer Programming
for Everybody.

[snip]
> Why do i get the feeling you'll never be satisfied?

Why attacking Tom so much? What has Tom done to you in the past? Why
does he get the bad reception he does? He isn't spreading misinformation,
is he? He has reasonable requests, in my opinion. 

Quick-someone-start-a-math-discussion!-ly yours,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From effbot at telia.com  Thu Apr  6 12:59:21 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Thu, 06 Apr 2000 16:59:21 GMT
Subject: __builtin__.open on windows NT
References: <8cicd6$1s0$1@nnrp1.deja.com> <20000406182802.C13830@xs4all.nl>
Message-ID: <Jn3H4.254$Za1.4213@newsc.telia.net>

Thomas Wouters <thomas at xs4all.net> wrote:
> > i'm new to python, so be prepared...i'm having trouble with open() on
> > windows NT.  my script uses the os library and it appears that open()
> > in my script is conflicting with an open() function in the os lib.  so
> > i tried to use __builtin__.open(), but i get this error "NameError:
> > __builtin__"
>
> Well, if you have that conflict, you are doing 'from os import *'.
> The easy solution is 'dont do that then'. 'from <module> import *' can be
> very bad for your health, for exactly the reason you specify. My advice
(and
> probably that of most people here ;) would be to change the import into
just
> 'import os' and fix the places where you actually use os.
>
> If you *really* want to use __builtins__, you can: you just have to import
> it first ! :-) Remember to do that with 'import __builtins__', or it might
> overwrite some of os's variables...

ehem.  the module is called __builtin__, while __builtins__ is a
magic attribute which may or may not exist.

if you really need to refer to the module containing the standard
builtins, do:

    import __builtin__
    fp = __builtin__.open(...)

the best solution is to only use "from import" if you know exactly
if you're doing (or in other words, if you use "from import" and don't
understand what happens, don't use it).  also see:

    http://www.pythonware.com/people/fredrik/fyi/fyi06.htm

</F>

<!-- (the eff-bot guide to) the standard python library:
http://www.pythonware.com/people/fredrik/librarybook.htm
-->




From garry at sage.att.com  Mon Apr 24 13:27:03 2000
From: garry at sage.att.com (Garry Hodgson)
Date: Mon, 24 Apr 2000 17:27:03 GMT
Subject: gaming and python
References: <000c01bfa91c$f421d6a0$0f8d0b3f@oemcomputer>
Message-ID: <390483E7.109E2EE1@sage.att.com>

> mrnatch wrote:
> 
> I told a friend of mine that I am learning to program using Python (I
> chose Python as my first language on the advice of a Llama) and he
> asked me an interesting question: Can Python be used for programming
> video games? And does it work with Windows API? Just curious.

i wrote a simple pacman game for my kids in python.
piece o' cake.  it uses tkinter for graphics, so 
should be win-portable.

-- 
Garry Hodgson                   Every night 
garry at sage.att.com                a child is born
Software Innovation Services        is a Holy Night.
AT&T Labs                         - Sophia Lyon Fahs


From johnm at magnet.com  Fri Apr 14 19:01:09 2000
From: johnm at magnet.com (John Mitchell)
Date: Fri, 14 Apr 2000 19:01:09 -0400
Subject: TCP/IP protocols
In-Reply-To: <20000415004919.B15664@xs4all.nl>
Message-ID: <Pine.SGI.4.05.10004141855400.4276-100000@lemur.magnet.com>

On Sat, 15 Apr 2000, Thomas Wouters wrote:

> On Fri, Apr 14, 2000 at 04:28:03PM -0600, Jeff Massung wrote:
> 
> > I'm new to Python programming, but am very experienced in C/C++. Could
> > someone point me to a good reference for TCP/IP programming with Python
> > (pdf, book, anything :]) thanks.

> If you're looking for some higher-level programming, see libraries like
> httplib, ftplib, telnetlib, urllib, SocketServer, *HTTPServer, etc.

These are actually just two libraries: client-side (urllib handles FTP,
HTTP, local files), plus server-side (Socket/HTTPServer).

For example, to fetch a web page is this:

import urllib
data = urllib.urlopen('http://www.magnet.com/')

To get some information out, it's much more complicated: 3 lines:

import htmllib, formatter
p.feed(data)
p.title
# =>  'Magnet Interactive Communications'


Show this to your local Java weenie.  He'll whine like a stuck pig!


> Documentation for all these modules comes with the python source (in the
> Doc/lib/ directory) in TeX format, with most binary distributions in some
> form or another, or can be downloaded or browsed online at
> http://www.python.org/doc/lib/ (There's an entire chapter, 11, dedicated to
> 'Internet Protocols and Support'.)

oh yeah.  But the above is your 60-second introduction.


- j





From avflinsch at att.net  Fri Apr 14 10:36:18 2000
From: avflinsch at att.net (aflinsch)
Date: Fri, 14 Apr 2000 09:36:18 -0500
Subject: Complaint Dept: My shoes!
References: <8d4s4d$mvr$1@nnrp1.deja.com>
Message-ID: <38F72CE2.CB7B7C5D@att.net>

musingattheruins at my-deja.com wrote:
> 
> If shoe is None then len(shoe) raises an exception.  Would be nice if
> len(None) was 0 and NOT raise an exeption.
> 

I think I would prefer it if len(None) returned None. Makes a bit more
sense to me having an unsized object return None, rather than zero.


From psl at mitre.org  Tue Apr 18 16:51:12 2000
From: psl at mitre.org (Pete)
Date: Tue, 18 Apr 2000 16:51:12 -0400
Subject: focus with PMW
References: <38FC75F4.D4E01F19@mitre.org>
Message-ID: <38FCCAC0.8868ED95@mitre.org>


I am trying to determine which widget,  the focus is currently on in my
code.  When I make a call to get_focus() it gives me an instance of a
TKinter.Entry,  but I am only creating PMW widgets (which are of course
extended from TKinter).  Since I am not creating TKinter widgets
explicitly in my own code,  how can I determine which PMW widget that
specific TKinter.Entry instance belongs to?


From python at rose164.wuh.wustl.edu  Sat Apr  1 00:09:24 2000
From: python at rose164.wuh.wustl.edu (David Fisher)
Date: Fri, 31 Mar 2000 23:09:24 -0600
Subject: regular expressions questions
References: <11A17AA2B9EAD111BCEA00A0C9B4179305321B0A@molach.origin.ea.com> <slrn8e2ln4.cou.neelk@brick.cswv.com> <x3zE4.6232$jX3.92560@typhoon.nyroc.rr.com> <010001bf9acf$500cff00$d4c2043f@spkydomain> <%z8F4.15922$0o4.95443@iad-read.news.verio.net>
Message-ID: <00b601bf9b98$742cf1a0$cbe9093f@spkydomain>

Well, that'll teach me to jump into a thread without reading the previous
posts. ;) In context I get it now.  And it's nice to have my prejudices on
the subject confirmed.
Thanks,
David

----- Original Message -----
From: "Tom Culliton" <culliton at clark.net>
Newsgroups: comp.lang.python
To: <python-list at python.org>
Sent: Friday, March 31, 2000 3:15 PM
Subject: Re: regular expressions questions


> In article <010001bf9acf$500cff00$d4c2043f at spkydomain>,
> David Fisher <python at rose164.wuh.wustl.edu> wrote:
> > This statement suprised me.  I had gotten the impression from
> > lurking in the newsgroup that the functions in the string module
> > were as fast or faster for many operations.
>
> You misunderstood the statement.  The string module is mainly a front
> for the strop module written in C which is faster for the cases where
> you can use it.
>
> >Is re really that much faster that string?
>
> No, of course not, unless it lets you do something in one go that
> would take you many string operations and connecting logic in python.
>
> Doing string.find will generally be faster than doing re.search for
> simple strings, because re.search is a more complex operation.  On the
> other hand doing a complex re.sub will be faster than emulating it
> with several string operations.
>




From rcc at jennic.com  Wed Apr  5 08:17:19 2000
From: rcc at jennic.com (Robert Cragie)
Date: Wed, 05 Apr 2000 12:17:19 GMT
Subject: Class type coercion
Message-ID: <newscache$j8mjsf$cqh$1@jenpc07.jennic.co.uk>

(Apologies if this has been covered before, I'm a bit of a Python newbie)

I posted a question to wxWindows list, and got an answer from Robin Dunn,
but it raised a more general question...

In C++, it is easy to use a pointer to a object declared as a super class of
that object's class, and then coerce it to be the same class type as the
object, e.g.

// Declare instance of class 'MyClass', which is derived from class
'MySuperClass'
MyClass MyClassObj;

// Declare pointer to class 'MySuperClass'
MySuperClass* pMySuperClass;

pMySuperClass = &MyClassObj; // OK, class 'MyClass' is derived from class
'MySuperClass'

// Coerce pointer to get to method which is only in class 'MyClass'
((MyClass*)pMySuperClass)->MyMethodOnlyInMyClass();

Can you do the equivalent in Python? I feel there must be some
straightforward way to do this, I just haven't figured it out yet. Or is
there...? As there is no type declaration in Python maybe this just isn't
possible.

There is a real-world need for this e.g. parent Window object contains
various child control objects. All the control classes are subclassed from
the Window class. All control objects have an ID. There often exists a
Window method to get a child window by ID. However, it can only return a
type of Window class. Therefore, to get to the actual control, this return
type needs to be subclassed.






From gdeering_telstra at my-deja.com  Tue Apr 18 18:10:25 2000
From: gdeering_telstra at my-deja.com (gdeering_telstra at my-deja.com)
Date: Tue, 18 Apr 2000 22:10:25 GMT
Subject: Performance Issues When Using Python/ASP
Message-ID: <8dimg4$kii$1@nnrp1.deja.com>

What are the performance issues when using Python for ASP, as opposed
to JScript (and dare I say VBScript).

What are the other advantages/disadvantages



Sent via Deja.com http://www.deja.com/
Before you buy.


From rmuschall.fih at t-online.de  Tue Apr  4 16:18:24 2000
From: rmuschall.fih at t-online.de (Ralf Muschall)
Date: Tue, 04 Apr 2000 22:18:24 +0200
Subject: What is python's language level?
References: <38EA1B8F.A797A130@lmco.com> <38E9DDD3.3313FF6@bioreason.com> <slrn8ekddq.gm3.zawrotny@gecko.sb.fsu.edu>
Message-ID: <38EA4E10.4DC2A4FE@t-online.de>

Michael Zawrotny schrieb:

> Very arbitrary.  I liked the fact that in their table:

> common lisp = 5.00
...
> scheme = 6.00

They seem to have split Common Lisp into CLOS (level 15)
and a CLOS-deprived rest (and probably did not look at
#Z(series) at all).

Scheme > Lisp might be due to the slightly diffrerent dominant
style in the two communities (schemers seem to prefer a more
functional style).

OTOH, I'd guess that they did not look very hard at lispisch
languages -- the dynamically scoped mulisp got the same
level like the real (lexically scoped) lisps.

The high value for HTML is funny, particularly if one looks
at such nasty lines like

<font size="2"> (for the main text bodies above the level table),
...
<table border="0" width="565">
...
<td valign="top" width="565">

(the only purpose of which is to make the result look ugly
and the rendering slower) in their code.

Ralf


From alwagner at tcac.net  Thu Apr 20 20:22:43 2000
From: alwagner at tcac.net (Albert Wagner)
Date: Thu, 20 Apr 2000 19:22:43 -0500
Subject: os.listdir & os.stat fail on windows zipdrive
Message-ID: <38FF9F53.A4698EE6@tcac.net>

What is peculiar about zipdrives on windows that would cause
os.listdir(dir) and/or os.stat(dir) to fail?  The following method works
on Linux and on Windows with a harddrive dir, but fails on windows when
dir is a zipdrive(e.g. D:\), resulting in no valid files or directories
being found.  
-- 
Small is Beautiful


From effbot at telia.com  Tue Apr  4 18:45:52 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Tue, 04 Apr 2000 22:45:52 GMT
Subject: Warning switches (was: Python 1.6 alpha 1 released)
References: <Pine.GSO.4.10.10004040035420.21084-100000@sundial> <5l7leern5v.fsf@eric.cnri.reston.va.us> <38ea1de0$0$17688@personalnews.de.uu.net> <nctG4.5133$74.83090@newsc.telia.net> <8cdnke$b05$1@pegasus.csx.cam.ac.uk>
Message-ID: <AguG4.1352$n68.191768064@newsb.telia.net>

Nick Maclaren wrote:
> >but hopefully, Python 3000 will be completely free from modes.
>
> It is important to distinguish modes from warning levels.  The
> former imply that there are several subtly different languages,
> and are usually regarded as snares and delusions.

in this case, the goal is to help you migrate towards Python 3000,
not to allow you to pick your favourite python dialect [1].  dunno
if that's best done with warnings (and an option to switch them
off) or modes (where the same option switches between warnings
and exceptions), or some mixture of that...

> The latter are best described by the following question:
>
>     Do you want me to tell you when I think that you may be
> making a mistake, when you are doing something that is asking
> for trouble, or only when you have definitely got it wrong?"

I think the current approach is to leave such warnings to the
development environment (or to pluggable parsers/compilers).
see the CP4E paper for more ideas in this direction.

(after all, the interpreter's energy is better spent on other
things...)

</F>

1) unless guido reads this, of course:
http://www.geocities.com/SiliconValley/Lab/6888/langopts.htm




From darrell at dorb.com  Wed Apr 12 17:20:49 2000
From: darrell at dorb.com (Darrell)
Date: Wed, 12 Apr 2000 21:20:49 GMT
Subject: How to know what re.sub took?
References: <38F387AE.1010704@gmx.net>
Message-ID: <RM5J4.3822$wM3.42670@typhoon.nyroc.rr.com>

"Per Kistler"
> How to know what has been substituted away, if one does something like:
>
>  >>> import re
>  >>> rex = re.compile(r"(hans|fred|simon)")
>  >>> st = "max hans bill"
>  >>> st2 = rex.sub("",st)
>  >>> st2
> 'max bill'
>
> It took "hans" away, but how can I learn this automatically?
>
This won't work for stuff like \g<1> in the replacement string.

import re
def sub(pat, repl, string):
    areas=[]
    def detect(mo, repl=repl, areas=areas):
        areas.append(mo.groups())
        return repl

    return re.sub(pat, detect, string), areas

print sub(r"(hans|fred|simon)","","max hans bill")


('max  bill', [('hans',)])







From ivanlan at callware.com  Sat Apr 15 12:02:08 2000
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Sat, 15 Apr 2000 10:02:08 -0600
Subject: Python CGI
References: <38F852D9.E2144F17@flashmail.com>
Message-ID: <38F89280.58B4ABC0@callware.com>

Hi All--

Jp Calderone wrote:
> 
>  I recently converted one of my perl CGI's to Python, and came across an
> bizarre and extremely undesirable side affect - Whenever a form that uses
> the Python script as its ACTION, the python scripts seems to get executed
> about 8 times.  At first I thought there was some logic error in my program
> making it octuple all the output, but I can't find anything that might do
> this.  Has anyone seen this before, or know how to fix it?
> 

What's your webserver?  I've run Apache on NT and Linux, and IIS on
NT;-(, and have never ever seen this.  You should also post a sample
script, too.

<guessing-in-the-face-of-ambiguity>-ly y'rs,
Ivan
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
http://www.pauahtun.org and
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
Author:  Teach Yourself Python in 24 Hours



From doughellmann at home.com  Thu Apr 27 07:38:39 2000
From: doughellmann at home.com (Doug Hellmann)
Date: Thu, 27 Apr 2000 11:38:39 GMT
Subject: python + linux + cron = problem
References: <39081703.48B64D1A@spekter.no>
Message-ID: <3908265E.A3C21409@home.com>

J?rn Eilertsen wrote:
> 
> Hi,
> 
> I've posted this earlier but noone replied so I'm trying again
> 
> I have this small program which includes the following line og code:
>  mounts = commands.getoutput('df | grep eggum | cut -d 1 -f 1 | cut -d /
> -f 4')
> 
> this runs fine from the commandline but when run as a cronjob I get the
> following error:
> 
> File "/usr/lib/python1.5/commands.py", line 42, in getoutput
>     return getstatusoutput(cmd)[1]
>   File "/usr/lib/python1.5/commands.py", line 53, in getstatusoutput
>     sts = pipe.close()
> IOError: (10, 'No child processes')
> 
> The use of pipes seems to be the trouble here.
> 
> Any help on how the get around this would be greatly appreciated.
> 
> 

If the problem is really having pipes in the command, why not put your
shell commands in a shell script, then run commands.getoutput() on that?

If the program runs from the command line but not cron, the problem is
more likely to be related to PATH problems within the cron environment. 
How do you set up the cron job?

Doug


From scarblac-spamtrap at pino.selwerd.nl  Sat Apr 22 08:12:18 2000
From: scarblac-spamtrap at pino.selwerd.nl (Remco Gerlich)
Date: 22 Apr 2000 12:12:18 GMT
Subject: code not true?
References: <8ds3ku$q0l$1@nnrp1.deja.com>
Message-ID: <slrn8g35ot.1pb.scarblac-spamtrap@flits104-37.flits.rug.nl>

jeff_islay at my-deja.com wrote in comp.lang.python:
> Why does the second example return false?
> 
> >>> print 7.0 == 7
> 1
> >>> print (.07 * 100) == 7
> 0

Floating point number on a binary computer are strange. They're not that
accurate. 0.07 * 100 != 7.0.

Try
>>> 0.07*100 == 7.0
0
>>> 0.07*100 - 7.0
8.881784197e-16

In Python 1.6, the interpreter doesn't try to hide this:
>>> 0.07*100
7.0000000000000009

For way too much info, look for old posts on this group by Tim Peters with
the words "floating point".

-- 
Remco Gerlich,  scarblac at pino.selwerd.nl

   This is no way to be
     Man ought to be free      -- Ted Bundy
       That man should be me


From matthias_w at my-deja.com  Wed Apr 12 05:59:36 2000
From: matthias_w at my-deja.com (matthias_w at my-deja.com)
Date: Wed, 12 Apr 2000 09:59:36 GMT
Subject: Embedding Python in Visual C++ 6.0
Message-ID: <8d1he4$26d$1@nnrp1.deja.com>

Hy! :-)

Maybe someone (Mark Hammond??) can help me. I
tried to embedd the win32 extentions from Mark in
a visual c++ project, but everytime the
<b>Win32uiInitInstance()</b> is called (via
<b>glue.InitInstance()</b>) I get an unhandled
exception at 0xC0000005 Access Violation and the
debugger jumps to 1E401F69 - the address of the
Win32uiInitInstance() function. When i try create
a break-point in this function this break-point
is disabled %-( by the compiler. The same is true
if i try to debug the original pythonwin project.

Help please!

Matthias


Sent via Deja.com http://www.deja.com/
Before you buy.


From warlock at eskimo.com  Sun Apr 30 14:36:25 2000
From: warlock at eskimo.com (Jim Richardson)
Date: Sun, 30 Apr 2000 18:36:25 GMT
Subject: The Simple Economics of Open Source
References: <1255089462-10343171@hypernet.com>
Message-ID: <slrn8gmb6a.ln.warlock@troll.myth>

On Sat, 29 Apr 2000 14:30:11 -0400, 
 Gordon McMillan, in the persona of <gmcm at hypernet.com>,
 brought forth the following words...:

>Richard Hale wrote:
>
>> CP/M was the OS "monopoly" of its day.  A perfect illustration of
>> how success breeds failure and how every trend has within it the
>> seeds of its own destruction.
>> 
>> It must be Microsoft's turn now!?
>
>Hmmm. Despite being fat, MS remains surprisingly hungry.
>
Tapeworms will do that  :)

-- 
Jim Richardson
	Anarchist, pagan and proud of it
WWW.eskimo.com/~warlock
	Linux, because life's too short for a buggy OS.



From exarkun at flashmail.com  Sat Apr 22 20:54:13 2000
From: exarkun at flashmail.com (Jp Calderone)
Date: Sun, 23 Apr 2000 00:54:13 +0000
Subject: SSL w/ urllib
References: <3901D800.DB50D968@flashmail.com> <3ditx9v851.fsf@newcnri.cnri.reston.va.us>
Message-ID: <390249B5.F5D545F7@flashmail.com>

Andrew Kuchling wrote:
> 
> Jp Calderone <exarkun at flashmail.com> writes:
> >  f = urllib.urlopen('https://www.mudprovider.com/cgi-bin/printvars.cgi',
> > params)
> > TypeError: too many arguments; expected 2, got 3
> 
> You've found a bug.  I've checked a fix into the CVS tree, so you can
> either update, or just apply the patch below.  Thanks for the bug
> report!
> 
> --amk
> 
> [snip patch]

 Wow, cool :)  And like 5 hours from report to solution.  How I love
open source :)

Here's a follow up question.  I applied the patch and it doesn't
raise an exception anymore, but it sits there for about 3 or 4
minutes, then returns an "addinfourl" object that looks like
this:

>>> print f.headers
None
>>> print f.url
http://www.mudprovider.com/cgi-bin/printvars.cgi
>>> print f.read()

>>>

 If I switch it over to a GET request, it works fine, but I
really need POST to work.  Is this another bug or am I doing
something wrong?

 Also, if there's a page that explains all this somewhere, I'll
happily take just a link to it and work this all out by myself :)

 Thanks, Jp

-- 
"Pascal is Pascal is Pascal is dog meat."
                -- M. Devine and P. Larson, Computer Science 340
--
 12:35am up 25 days, 5:42, 0 users, load average: 0.02, 0.02, 0.00


From kc5tja at garnet.armored.net  Fri Apr 28 07:58:00 2000
From: kc5tja at garnet.armored.net (Samuel A. Falvo II)
Date: 28 Apr 2000 11:58:00 GMT
Subject: Do I always have to write "self." ?
References: <270420001706386648%pecora@anvil.nrl.navy.mil> <slrn8ghc6t.f01.scarblac-spamtrap@flits104-37.flits.rug.nl> <280420000746597329%pecora@anvil.nrl.navy.mil>
Message-ID: <slrn8gjqs8.ak0.kc5tja@garnet.armored.net>

In article <280420000746597329%pecora at anvil.nrl.navy.mil>, Louis M. Pecora
wrote:
>> You *can* use another word for 'self' though, 's' or 'me' or something. But
>> it makes your code harder to read, of course...
>
>You mean by doing something like

No.  He means something like this:

>>> class Foo:
...   def MyMethod( m, arg1, arg2 ):
...     m.field1 = arg1
...     m.field2 = arg2/arg1

`self' is just an argument, like any other on the list.  It's distinguished
*only* in that it's supplied by the Python environment automatically; what
you choose to call it is entirely up to you.

Use of the word `self' is entirely for conventional reasons.

-- 
KC5TJA/6, DM13, QRP-L #1447
Samuel A. Falvo II
Oceanside, CA


From haering at sunhalle8.informatik.tu-muenchen.de  Thu Apr 27 14:39:05 2000
From: haering at sunhalle8.informatik.tu-muenchen.de (Gerhard Haering)
Date: 27 Apr 2000 18:39:05 GMT
Subject: HTML parser example, anybody?
Message-ID: <slrn8gh2a9.oj.haering@sunhalle8.informatik.tu-muenchen.de>

Hi!

I want to transform my own HTML templates (with a few template tags,
special comments, and so on) into final HTML for the browser. I would like
to use Pyhton's htmllib, but from the docs I can get no clue on how to get
it done. I seem to have to use a writer and a formatter class ... 

I would greatly appreciate an example of HTML -> HTML using
htmllib/sgmllib. Can anybody provide pointers/code snippets?

Gerhard


From mhammond at skippinet.com.au  Thu Apr 20 22:22:19 2000
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Fri, 21 Apr 2000 02:22:19 GMT
Subject: Why MFC instead of Tkinter -- was Re: Pythonwin?
References: <eMz2Nmiq$GA.73@net025s> <4QtL4.24$v85.41@news-server.bigpond.net.au> <chuL4.8483$O7.17315@typhoon.southeast.rr.com> <7LDL4.534$v85.2563@news-server.bigpond.net.au> <sfu6d657l6e184@corp.supernews.com> <38ff21b8$0$10972@senator-bedfellow.mit.edu> <38FF281E.5B2DCC71@tismer.com> <38ff823d$0$10971@senator-bedfellow.mit.edu> <2wML4.15897$O7.23504@typhoon.southeast.rr.com>
Message-ID: <vXOL4.748$v85.4298@news-server.bigpond.net.au>

"Don Tuttle" <tuttledon at hotmail.com> wrote in message
news:2wML4.15897$O7.23504 at typhoon.southeast.rr.com...

> Which brings up the question near and dear to a lazy man's heart like
> myself..  Which is the eaiser Windows GUI builder for Python--VB or VC++
?
>
> With VB,  I've only seen Python mentioned as a COM object to be used by
VB.
> Are there other ways?

Probably, but this is the best.  Using this technique, you generally end
up with all the UI logic (ie, this is selected, so I need to disable these
controls, etc) written in VB.  Only your "business logic" is done in
Python.  It probably is possible to do some of the GUI logic in Python,
but that doesnt make as much sense given VBs IDE support for this.  You
can swap VB for Delphi in this sentence too...

> With VC++,  I gather you use the dialog editor to generate the MFC code
that
> you then plug into Python?  Correct?

Exactly.  With MFC/Pythonwin, you get to create a DLL that contains UI
resources, and use them.  "Use them" means load them, display them, and
hook all the messages and notifications for them - ie, your GUI logic ends
up completely in Python, without the nice, trendy
"double-click-on-an-object-to-edit-its-event-list" IDE that VB/Delphi give
you.

Mark.





From jon at tryme.com  Mon Apr 10 18:32:37 2000
From: jon at tryme.com (charlie)
Date: Mon, 10 Apr 2000 15:32:37 -0700
Subject: Using ADO in IIS
Message-ID: <yBsI4.1142$uE2.458712@news.pacbell.net>

Can someone please help me use ADO from IIS.  I just purchased Mark's new
book on Win32 but his example on page #254 is for Pythonwin not for
ASP/Python.  The ASP context requires different object creation syntax.
I've been asking this question for over six months and no one seems to know
or care about finding a solution.

The only reason I'm not using Python at work is over ADO:SQL accessibility.
It's cool to loop and call cool functions, but without database connectivity
Python is pretty much useless in this web world.  Everything requires
database connectivity this days. Please don't reply with: you can use ODBC.
I've been hearing that for months.  Besides, it's not supported.  I can't
sell Python to my boss if I can't use ADO.

I've seen many others ask this same question over and over with no resolve.
Can we work together to finally find an ASP/ADO/Python solution.

The way I see it, solving this challenge could significantly boost Python
adoption in the web developer realm.

ed




From mwh21 at cam.ac.uk  Sun Apr  2 09:57:50 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 02 Apr 2000 14:57:50 +0100
Subject: Debugging confusion -- too many stacks!
References: <000a01bf9c6b$3f80c2a0$752d153f@tim> <38E739A1.45F4B199@tismer.com>
Message-ID: <m37legk3kx.fsf@atrus.jesus.cam.ac.uk>

Christian Tismer <tismer at tismer.com> writes:

> This VM stack could also be replaced by a set of registers, since
> the maximum stack size is always known at compile time.

My gut tells me that making this work with exceptions would be a dog.
It's the sort of thing that might be fun to play with if the compiler
and eval loop were written in Python, but if I have to hack C to play
with these things, it just becomes tedious.  Still, maybe if I have a
month or two with nothing else to do...

Cheers,
M.

-- 
well, take it from an old hand: the only reason it would be easier
to program in C is that you can't easily express complex  problems
in C, so you don't.                 -- Erik Naggum, comp.lang.lisp


From echuck at mindspring.com  Wed Apr 19 23:52:06 2000
From: echuck at mindspring.com (Chuck Esterbrook)
Date: Wed, 19 Apr 2000 23:52:06 -0400
Subject: [ANNOUNCE] Webware for Python 0.1
Message-ID: <38FE7EE6.A3FDFEC9@mindspring.com>

Webware for Python is a definitive suite of Python software components for use in web development. Each component has a focused purpose and can often be used on it's own. At the same time, the components are designed to work together where appropriate. As a user of Webware, you can bite off as much or as little as you want to chew.

You can find more and download at:

      http://www.writers-webshop.com/Webware


-Chuck


From tony at lsl.co.uk  Mon Apr 10 05:32:43 2000
From: tony at lsl.co.uk (Tony J Ibbs (Tibs))
Date: Mon, 10 Apr 2000 10:32:43 +0100
Subject: Regular Expression Help for Python Newbie.
In-Reply-To: <lrCH4.80$rc9.181720576@newsb.telia.net>
Message-ID: <005f01bfa2cf$b99b0ea0$f0c809c0@lslp7o.lsl.co.uk>

Raoul-Sam Daruwala wrote:
> My problem is that one of the sets of files that I'm trying to parse has
> badly formatted tables. Now when I do a view source on the files I can
> see the problem clearly. It's quite simple, the tables in these files
> starts out properly formatted but after a standard header the script
> than generates them leaves out the <TR> tag. This is incredible to me
> because both Netscape and IE read can view the tables properly.

Fredrik Lundh replied:
> <TR>'s are optional -- if the browser stumbles upon <TD>
> in a <TABLE> context, it should insert <TR>'s all by itself.

Well, no...

OK. As far as I can see, <TR>'s are not optional - given an (even vaguely)
SGML based specification I don't understand how they could be. And looking
at the HTML 3.2 Reference Specification, they clearly aren't (although the
</TR> is). And that makes sense as the <TR> defines the start of the row
"element".

Of course, it's entirely possible to write a browser (or other HTML reader)
that can cope with missing <TR>'s *if* the </TR>'s are present. And
obviously that's what some browsers do. This way also lies madness - it's
going to be impossible to guess which "mistakes" the browser will
self-correct for, and which it won't (it'll have to be by some ad-hoc
mixture of emulating what other browsers appear to do, noticing what
mistakes one has actually seen, guessing which mistakes are likely to
happen, and fixing some because they're easy to fix, even if unlikely). I
always find it strange that otherwise intelligent people don't want their
compilers to work like this, but do want their browsers to.

As to the specific problem, it's clearly possible to write code that will
trigger on the </TR>'s instead of on the <TR>'s. This is left as an
excercise for the reader, especially if they want to make it harder by using
regular expressions...

Tibs
(damn - it's Monday and I've already disagreed with /F, not something I'd
normally do - let's hope the rest of the week settles down)
--
Tony J Ibbs (Tibs)      http://www.tibsnjoan.demon.co.uk/
Give a pedant an inch and they'll take 25.4mm
(once they've established you're talking a post-1959 inch, of course)
My views! Mine! Mine! (Unless Laser-Scan ask nicely to borrow them.)




From KingArthur at roundtable.dyndns.org  Thu Apr 20 21:37:51 2000
From: KingArthur at roundtable.dyndns.org (news-server)
Date: Fri, 21 Apr 2000 01:37:51 GMT
Subject: editor for win32
References: <NpHL4.779$fK2.113609@news.flash.net>
Message-ID: <PhOL4.9343$wM3.107684@typhoon.nyroc.rr.com>

Textpad is a great programming text-editor. It has Syntax highlighting for
all the popular languages and various other languages. Even has syntax
highlighting for SQL and IDL.

You can get it at www.textpad.com (you can also add your python interpreter
to the tools and run your programs from within textpad).


Scott Hathaway wrote in message ...
>Can anyone recommend a python editor for Windows?
>
>If there is an IDE for windows (other than IDLE), even better.
>
>Thanks,
>Scott
>
>




From mwalker at origin.ea.com  Mon Apr 24 11:55:59 2000
From: mwalker at origin.ea.com (Matthew Walker)
Date: Mon, 24 Apr 2000 10:55:59 -0500
Subject: pickle.py and cPickle.c - persistent_id() is always called - why?
Message-ID: <sg8rkffdcdo80@corp.supernews.com>

(Previously posted via Deja.com by matsaleh but I never saw the post on
usenet - forgive me if this shows up more than once.)

Hello fellow .py types...

I am doing a bit of work using pickle/cpickle and am trying to
optimize. I found that my user-defined persistent_id() function is being
called for (just about) every attribute (name and value) in my object that
is being pickled.

My specific (somewhat ad-hoc) test case pickles ~100 objects in a
containment hierarchy, using persistent_id to break the containment
relationship and replace the references with a proprietary object id.
Although I have only ~100 object references to resolve, persistent_id() is
called ~6400 times. In tracing the code, it appears to be called for every
attribute and value in my objects.

This appears to be caused by the fact that pickle.py:save() is called with
the default pers_save flag of 0 in all cases except for when it is called by
save_pers(). This appears to indicate that all types: tuples, dicts,
sequences, longs, strings, etc, cause my persistent_id() to be called, when
all I want is for it to be called for object instances.

I modified pickle.py to change the default of the pers_save flag to 1 in the
save() method, and then call it with a 0 only from within the save_inst()
method. This amounts to invoking my persistent_id()
function only when a reference to an object instance is encountered as my
objects are being pickled.

My pickled objects do not appear to be adversely  affected by this change,
and the number of calls to persistent_id() was reduced from ~6400 to ~500,
reducing the time spent in this method by an order of magnitude (0.37 sec to
0.03 sec).

I have not yet tested this change in cPickle.c, but in my original tests,
persistent_id() was a much more significant factor in my runs using cPickle,
because all the pickling code is in C and is relatively much faster than my
persistent_id() method, which is in Python. I expect the performance boost
my making the change in cPickle to be even greater, relatively speaking.

My question is, why is persistent_id() being invoked so often? I do not see
the reason for calling it for basic python types and structures such as
tuples, dicts, and the like. Is this a reasonable change to make to the
python source, or is it not a safe change for the general pickling cases
that the pickle/cPickle modules have to handle? I'm sure this code has been
scruitinized by many folks much more experienced than I - I would be
grateful for your insights.

All timings and function counts done by profile.py and pstats.py.

Regards.






From wware at world.std.com  Fri Apr 28 08:16:35 2000
From: wware at world.std.com (Will Ware)
Date: Fri, 28 Apr 2000 12:16:35 GMT
Subject: Do I always have to write "self." ?
References: <270420001706386648%pecora@anvil.nrl.navy.mil> <slrn8ghc6t.f01.scarblac-spamtrap@flits104-37.flits.rug.nl> <280420000746597329%pecora@anvil.nrl.navy.mil>
Message-ID: <Ftq7Fo.z7@world.std.com>

Louis M. Pecora (pecora at anvil.nrl.navy.mil) wrote:
> I'm sure some way of declaring variables 'local' could be done in the
> class defintion, but I doubt Guido would do that now.

It wouldn't be too hard to write a preprocessor that would do this
for you. You might put tags in your code such as
#Locals: foo bar xyzzy
and
#Endlocals
so that the translation process for foo, bar, and xyzzy would be limited
to the method where they are used. If you promise yourself never to use
a 'def' inside a method definition, you could omit the '#Endlocals' tag,
and just look for the next line with the regular expression 'def ' in it.

> You mean by doing something like
> s=self
> at the beginning of each class method?

The choice to use the word 'self' is not imposed by Python, it's just
a convention for readability. You can use 's' instead if you wish.

def quadratic(self):
  det = self.b ** 2 - 4 * self.a * self.c
  return ((-self.b + math.sqrt(det)) / (2 * self.a),
          (-self.b - math.sqrt(det)) / (2 * self.a))

versus

def quadratic(s):
  det = s.b ** 2 - 4 * s.a * s.c
  return ((-s.b + math.sqrt(det)) / (2 * s.a),
          (-s.b - math.sqrt(det)) / (2 * s.a))
-- 
 - - - - - - - - - - - - - - - - - - - - - - - -
Resistance is futile. Capacitance is efficacious.
Will Ware	email:    wware @ world.std.com


From mwh21 at cam.ac.uk  Tue Apr 11 06:35:20 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 11 Apr 2000 11:35:20 +0100
Subject: A Mountain of Perl Books + Python Advocacy
References: <20000405053920.20579.qmail@web2102.mail.yahoo.com> <38ED40B0.A48EC4CC@udel.edu> <tNvI4.18$_N4.1751@vic.nntp.telstra.net>
Message-ID: <m38zylkjrr.fsf@atrus.jesus.cam.ac.uk>

i_luv_fudge at I_HATE_SPAM.hotmail.com (.,.) writes:

> It still doesn't change the fact that this is a VERY hard language
> to learn as a layman without a guide or some sort of coding TLA
> translator.  

Compared to what?  perl?  C++?  Don't make us laugh...

> When and if I get proficient I might write a page for people with
> that in mind.

Tried these? (found on http://www.python.org/doc/Intros.html)

    http://www.idi.ntnu.no/~mlh/python/programming.html
    http://www.honors.montana.edu/~jjc/easytut/easytut/
    http://members.xoom.com/alan_gauld/tutor/tutindex.htm

Not to discourage you from your literary efforts, more would always be
good, but they may be of some use.

Cheers,
M.

-- 
  incidentally, asking why things are "left out of the language" is
  a good sign that the asker is fairly clueless.
                                      -- Erik Naggum, comp.lang.lisp


From p.agapow at ic.ac.uk  Thu Apr 13 04:10:16 2000
From: p.agapow at ic.ac.uk (Paul-Michael Agapow)
Date: Thu, 13 Apr 2000 09:10:16 +0100
Subject: 1st non-trivial program - gentle criticism requested
Message-ID: <1e90r16.1sd2ts41e7ur6kN@gershwin.bio.ic.ac.uk>

Hi y'all,

I'm experimenting with Python for a number of purposes (an embedded
language for simulations, a tool for exploring experimental data, a good
first programming language) and so I've cobbled together my first
non-trivial program. With some trepidation, I post it here in the hopes
that experts will be able to coach me in "The Python Way". I realise it
looks like C++ (my usual langauge), but criticisms of style would be
useful too.

Basically, it's a recreation of the "Methinks it is a weasel" simulation
detailed by Richard Dawkins in "The Blind Watchmaker". A population of
strings is compared vs. a target string and the "fittest" (the one that
most resembles the target) is used to form the next generation, with a
small chance. So it's a simple genetic algorithm.

thanks

p-m


--- program follows

"""
weasel.py

A simple program exploring the possibilities of using Python, esp. in
simulations, modelling the 'methinks it is a weasel' problem from
Richard Dawkins' "The Blind Watchmaker". The program flow is:

1. Init population of random individuals
2. Pick fittest individual
3. If the individual is the target:
   3a. Return
   3b. Else, breed new population from individual (mutation)
4. Goto 2. 

"""

#--- Includes

import whrandom
import string


#--- Module constants & variables

kDefTarget = "METHINKS IT IS LIKE A WEASEL"
kDefPopSize = 10;
kDefStepSize = 50;

gPopulation = [];                         # the weasels, as it were
gRng = whrandom.whrandom()                # random source
gTarget                                   # what fitness is measured
against
gNumGenerations                           # how many cycles did it take?
gAlphabet = "ABCDEFGHIJKLMNOPQRSTUVWXYZ " # all possible symbols
gMutationRate                             # chance of a symbol flipping
gStepSize                                 # how often to report stats


#--- Module functions -----------------------------------------------

#--- Run

# Setup the variables and run
def run (size=kDefPopSize, target=kDefTarget, step=kDefStepSize):
   global gTarget, gNumGenerations, gStepSize, gMutationRate
   # initialise
   gTarget = target
   gNumGenerations = 0;
   gStepSize = step
   gMutationRate = 1.0/ (len(gTarget) + 1)
   initPop (size);
   # tell user the settings
   print "WEASEL in action:"
   print "-----------------"
   printStartParams();
   print "-----------------"
   # evolve until the end is reached
   while (0 == evolvePop (gStepSize)):
      if (0 == (gNumGenerations % kDefStepSize)):
         print "At generation", gNumGenerations
   # print end results
   print "-----------------"
   print "Reached target after ", gNumGenerations, "generations"
   printPop();
   return


#--- Setup

def initPop (popsize = 10):
   # Initialise the contents of the population
   # Note: do this or gPop is local and temporary!
   global gPopulation
   gPopulation = [];
   for i in range (popsize):
      theNewMember = MakeNewMember (len(gTarget))
      gPopulation.append (theNewMember)
   return

def MakeNewMember (chromsize = len(gTarget)):
   # For initing the population, generate a random string of symbols
   # Note: seems inefficient creating strings this way
   theNewString = "";
   while (len(theNewString) < chromsize):
      theNewString = theNewString + GetRandomSymbol()
   return theNewString
   

#--- Mutation

def MutateIndividual (theTargetIndiv):
   """Iterates over symbols of individual changing their 'genes'"""
   # Must be a faster way of doing this
   theNewIndiv = ""
   for theSymbol in theTargetIndiv:
      if (gRng.random() < gMutationRate):
         theNewSymbol = GetRandomSymbol()
      else:
         theNewSymbol = theSymbol
      theNewIndiv = theNewIndiv + theNewSymbol
   return theNewIndiv


def GetRandomSymbol ():
   """For the purposes of mutation, this selects a random symbol from
the alphabet"""
   # The global declaration isn't necessary as we don't assign to these
   # vars, but I'm doing it as a reminder
   global gRng, gAlphabet
   theSymbolIndex = gRng.randint(0, len(gAlphabet) - 1)
   return gAlphabet[theSymbolIndex]
   
      
#--- Evolve

# Evolve for a given number of time steps
def evolvePop (iNumSteps):
   global gNumGenerations
   for i in range (iNumSteps):
      if (evolveStep()):
         return 1
      else:
         gNumGenerations = gNumGenerations + 1
   return 0
   

# Evolve a single step
def evolveStep ():
   global gPopulation
   if (gTarget in gPopulation):
      return 1
   else:
      # calculate fitness
      theFitness = []
      for theIndiv in gPopulation:
         theFitness.append (calcFitness (theIndiv))   
      # select winner
      theWinnerIndex = findMax (theFitness)
      theWinner = gPopulation[theWinnerIndex]
      assert (theWinnerIndex <> None)
      # breed from winner
      for i in range(len(gPopulation)):
         gPopulation[i] = MutateIndividual (theWinner)
      return 0


def calcFitness (anIndiv):
   theFitness = 0
   for i in range (len(gTarget)):
      if (gTarget[i] == anIndiv[i]):
         theFitness = theFitness + 1;
   return theFitness

def findMax (iCollection):
   # Returns the index of the largest member of collection
   # Just a little bullet-proofing
   if (len(iCollection) == 0):
      return
   # Setup initial sensible values
   theMaxIndex = 0
   theMaxVal = iCollection[0];
   # Iterate over remainder of container
   # Inefficient to look at the first element again, but it will do
   for i in range(len(iCollection)):
      if (theMaxVal < iCollection[i]):
         theMaxVal = iCollection[i]
         theMaxIndex = i
   return theMaxIndex
   

#--- Output

# Print out the contents of the population
def printStartParams ():
   print "The target is", gTarget
   print "The mutation rate is", gMutationRate
   print "The alphabet is '" + gAlphabet + "'"
   print "The population size is", len(gPopulation)
   printPop()
   return
   
   
# Print out the contents of the population
def printPop ():
   print "The population is currently:"
   gPopulation.sort()
   theCount = 0;
   for theIndividual in gPopulation:
      if (theIndividual == gTarget):
         thePrefix = "*"
      else:
         thePrefix = ""
      theCount = theCount + 1
      thePrefix = thePrefix + repr(theCount) + " :"
      print string.rjust(thePrefix,8), theIndividual


#--- if this is called as a standalone script   

if (__name__ == "__main__"):
   run()
   
   
-- 
Paul-Michael Agapow (p.agapow at ic.ac.uk), Biology, Silwood Park
"The world is made of fire."


From osuchw at my-deja.com  Mon Apr 17 22:21:15 2000
From: osuchw at my-deja.com (osuchw at my-deja.com)
Date: Tue, 18 Apr 2000 02:21:15 GMT
Subject: Python CGI running commands under NT
References: <2B1262E83448D211AE4B00A0C9D61B03BA72AC@MSGEURO1>
Message-ID: <8dggqe$709$1@nnrp1.deja.com>

In article
<2B1262E83448D211AE4B00A0C9D61B03BA72AC at MSGEURO1>,
  Pieter Claerhout <PClaerhout at CREO.BE> wrote:
> I made a Python CGI-script that executes a DOS-
command and
> should return the result in a HTML page. The
script looks
> as follows:
>
>   import os
>
>   pipe = os.popen("ipconfig", 'r')
>   text = pipe.read()
>   del pipe
>
>   print "Content-type: text/html"
>   print
>   print "IPConfig on AV2700"
>   print "<br>"
>
>   print '<pre>'
>   print text
>   print '</pre>'
>
> If I run this script from the command line, it
works fine, but
> when I open it using Internet Information
Server, all is shown
> except the contents of the variable text. Any
ideas what could
> be wrong? Is it maybe a setting in IIS or are
these options
> not allowed in CGi-scripts?

I used ASP and got results you wanted. Sample
included.

Waldemar Osuch

8<-------------------------------------
<%@ LANGUAGE = Python %>

<%
import win32pipe
from win32api import GetComputerName

pipe = win32pipe.popen("ipconfig", 'r')
txt = pipe.read()
del pipe

prn = Response.Write
%>

<HTML>
<head>
<title>IPconfig Results</title>
</head>
<body>
<h3>IPConfig on <%prn(GetComputerName())%></h3>
<pre><%prn(txt)%></pre>
</body>
</HTML>
8<----------------------------------------


Sent via Deja.com http://www.deja.com/
Before you buy.


From scarblac-spamtrap at pino.selwerd.nl  Tue Apr 18 05:33:42 2000
From: scarblac-spamtrap at pino.selwerd.nl (Remco Gerlich)
Date: 18 Apr 2000 09:33:42 GMT
Subject: Pointers
References: <38D00DCE.9740380F@be-research.ucsd.edu> <8bfsqo$kje$1@tribune.oar.net> <dp4s9wl9df.fsf@mraz.iskon.hr> <uPYC4.55946$3b6.224410@ozemail.com.au> <4k5D4.7566$0o4.54442@iad-read.news.verio.net> <38FB98F1.239EC0FB@whetstonelogic.com>
Message-ID: <slrn8foe80.bd3.scarblac-spamtrap@flits104-37.flits.rug.nl>

Peter Halliday wrote in comp.lang.python:
> I just started learning python.  I just want to make sure that I understand the
> object interaction you were discussion here.  To use your statement.  So x + 1
> is an object, x is an object, and y is an object.  Y and x happen to both be
> pointing to x + 1.  Is this right? 

Sort of. x can never point to x+1 :). Rather, x = x+1 computes the result of
the expression x+1, makes a new object with that value, and makes x point to
it.

> If you then did something to y would you be doing something to x + 1?

No, not if this is about integers anyway.

> I would say, no because it is a seperate object right!

Actually, that's not the reason. y and x are both exactly the same object.
Thing is, if it's an integer, then you can't change it! You can only make a
new object and assign it to y.

> >         x = x + 1
> >         y = x

Now y and x are the same object. But the only way to change y, say with y=y+1,
makes a new object (the result of y+1) and lets y refer to it.

It's different with mutable objects like lists, they can be changed in place:
>>> x = [1,2]
>>> y = x  # They're both the same list now
>>> x
[1, 2]
>>> y.append(3)
>>> y
[1, 2, 3]
>>> x
[1, 2, 3]
>>> y = y+[4] # This still creates a new object
>>> y
[1, 2, 3, 4]
>>> x
[1, 2, 3]

Assignment never changes an object, except when it's an assignment to some
internal of the object (like y[1] = 3).

Try playing around in the interpreter for a while, this is a key thing to
understand about Python in my opinion. You can test objects to see if
they're the same object with 'x is y', and test their values with 'x == y'.
-- 
Remco Gerlich,  scarblac at pino.selwerd.nl
Hi! I'm a .sig virus! Join the fun and copy me into yours!


From moshez at math.huji.ac.il  Sat Apr 29 10:53:04 2000
From: moshez at math.huji.ac.il (Moshe Zadka)
Date: Sat, 29 Apr 2000 17:53:04 +0300 (IDT)
Subject: How to get the python object from a variable name ?
In-Reply-To: <20000429163158.A4091@yetix.sz-sb.de>
Message-ID: <Pine.GSO.4.10.10004291752200.26991-100000@sundial>

On Sat, 29 Apr 2000, Andreas Jung wrote:

> Inside a C program I need to get a pointer to the python object
> for a object or variable by its variable name in the current
> namespace of the interpreter. What I want to do ? My programs
> gets names of variables and I need to get the str() representation
> for given variable/object names as a C (char *). Maybe a stupid
> question but I've no idea :)

Try passing the name of the variable to Py_Eval*() functions.

probably-some-nicer-solution-exists-ly y'rs, Z.
--
Moshe Zadka <mzadka at geocities.com>. 
http://www.oreilly.com/news/prescod_0300.html
http://www.linux.org.il -- we put the penguin in .com




From mwh21 at cam.ac.uk  Fri Apr 14 04:22:56 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 14 Apr 2000 09:22:56 +0100
Subject: Questions on 1.6a2's string methods
References: <38F6C904.76B538E5@concentric.net>
Message-ID: <m33dopgkgv.fsf@atrus.jesus.cam.ac.uk>

I think weve had all these discussions recently.

Manus Hand <mjhand at concentric.net> writes:

> 1.  I see that string objects now support (as methods) most of the
>     functions from the string module.  Among these are:
>         upper, lower, split, strip, find
>     and surely others.  My question is, why is capwords (which seems
>     to be in the same genre as upper and lower) not a method?

Start here:

http://x28.deja.com/[ST_rn=ps]/viewthread.xp?AN=586654400&recnum=%3cm3snysor7o.fsf at atrus.jesus.cam.ac.uk%3e%231/1

Conclusion: the "missing methods" might get implemented, if someone
gets round to it.  THe unicode string type has more methods (it has
center, ljust, rjust).

> 2.  Along these same lines, since split() became a method of the
>     string type, wouldn't it make sense to make join() a method of
>     the list type?

Start here:

http://x29.deja.com/[ST_rn=ps]/viewthread.xp?AN=588318989

Conclusion: nothing obvious, but the status quo seems the most likely
(& best - to me) winner.

> 3.  Are the classlike standard types (list, dictionary, and now
>     string) equipped with a __dict__ attribute?  I can see the names
>     of all functions supported by a user-defined class by saying
>     className.__dict__.keys(), but I cannot see the list of methods
>     for the string type (at least not in the same way).  Thus my
>     need to ask silly questions like #1 above (maybe capwords is
>     there by some other name??)

No, but they ways that always worked still work <wink>:

>>> dir('')
['capitalize', 'center', 'count', 'endswith', 'expandtabs', 'find',
'index', 'isdigit', 'islower', 'isspace', 'istitle', 'isupper',
'join', 'ljust', 'lower', 'lstrip', 'replace', 'rfind', 'rindex',
'rjust', 'rstrip', 'split', 'splitlines', 'startswith', 'strip',
'swapcase', 'title', 'translate', 'upper']
>>> dir(u'')
['capitalize', 'center', 'count', 'encode', 'endswith', 'expandtabs',
'find', 'index', 'isdecimal', 'isdigit', 'islower', 'isnumeric',
'isspace', 'istitle', 'isupper', 'join', 'ljust', 'lower', 'lstrip',
'replace', 'rfind', 'rindex', 'rjust', 'rstrip', 'split',
'splitlines', 'startswith', 'strip', 'swapcase', 'title', 'translate',
'upper']
>>> ''.__methods__
['capitalize', 'center', 'count', 'endswith', 'expandtabs', 'find',
'index', 'isdigit', 'islower', 'isspace', 'istitle', 'isupper',
'join', 'ljust', 'lower', 'lstrip', 'replace', 'rfind', 'rindex',
'rjust', 'rstrip', 'split', 'splitlines', 'startswith', 'strip',
'swapcase', 'title', 'translate', 'upper']

No, no capwords.

HTH,
M.

-- 
  languages shape the way we think, or don't.
                                      -- Erik Naggum, comp.lang.lisp


From raffael at mediaone.net  Sun Apr 23 23:17:35 2000
From: raffael at mediaone.net (Raffael Cavallaro)
Date: Sun, 23 Apr 2000 23:17:35 -0400
Subject: The Simple Economics of Open Source
In-Reply-To: <200004230436.AAA12178@kirk.dnaco.net>
Message-ID: <B529350E.51B9%raffael@mediaone.net>

> Correlation doesn't imply causality, and when it does, it's easy to get
> causality backwards if you're not careful.  You could get misled into
> spending lots of money to get rich if you just look at the
> correlations.

In the case of cash cows, we don't need a causal argument at all. It's
simply enough to note that people don't share them. Whether they became cash
cows because they weren't shared (possible, though not likely), or they are
not now shared because they have become cash cows doesn't really matter. The
lesson for open source is the same - expect only commodity items, known to
many, to be open sourced. Expect lucrative cash cows to remain closed.

Ralph

-- 

Raffael Cavallaro, Ph.D.
raffael at mediaone.net




From python-list at teleo.net  Sun Apr  2 17:05:10 2000
From: python-list at teleo.net (Patrick Phalen)
Date: Sun, 2 Apr 2000 14:05:10 -0700
Subject: HTML Form to Python Script Problem
In-Reply-To: <38E796BC.23191B2@regnoc.com>
References: <38E796BC.23191B2@regnoc.com>
Message-ID: <00040214141502.04535@quadra.teleo.net>

[Jim Conger, on Sun, 02 Apr 2000]
:: 
:: I'm having trouble getting a simple form to launch a python script on my
:: RH6.1 system.  Python runs fine alone, but I don't get any action when I
:: launch a script.  Here is the HTML form:

<snip

:: The python script is in the same directory as the HTML file and looks
:: like:

<snip>

Your form and CGI script are in the same directory? Your script isn't
in /cgi-bin?

:: The script is set to be executable by all and runs fine alone.  However,
:: nothing happens when I click the submit button on the HTML form.  Any
:: clues?

(Having a CGI script executable by all is not a good idea.)

Since this is RH6.1, presumably you're running Apache. Sounds like you
might want to visit http://www.apache.org and read the docs,
particularly the docs for ScriptAlias.



From richard_chamberlain at ntlworld.com  Wed Apr 19 11:27:42 2000
From: richard_chamberlain at ntlworld.com (Richard Chamberlain)
Date: Wed, 19 Apr 2000 16:27:42 +0100
Subject: COM and returning dates
Message-ID: <8dkj9b$lvr$1@gxsn.com>

Hi,

I'm trying to use Python as a COM object (called from VB) basically
returning values from a database, some of which are dates.

The problem I'm getting is that the dbi date object obviously doesn't
translate.

So is there a way a can simply do this?

I was thinking of running through the list and converting the dbi date
objects into string representations and passing those. But I'm not sure what
I do with them on the VB side.

Any better ideas?

Thanks

Richard




From guido at python.org  Fri Apr  7 15:18:40 2000
From: guido at python.org (Guido van Rossum)
Date: Fri, 07 Apr 2000 15:18:40 -0400
Subject: [Python-Dev] Round Bug in Python 1.6?
In-Reply-To: Your message of "Fri, 07 Apr 2000 20:45:41 +0200."
             <Pine.GSO.4.10.10004072044240.25661-100000@sundial> 
References: <Pine.GSO.4.10.10004072044240.25661-100000@sundial> 
Message-ID: <200004071918.PAA27474@eric.cnri.reston.va.us>

> Just checking my newly bought "Guido Channeling" kit -- you mean str()
> but special case the snot out of strings(TM), don't you

Except I'm not sure what kind of special-casing should be happening.

Put quotes around it without worrying if that makes it a valid string
literal is one thought that comes to mind.

Another approach might be what Tk's text widget does -- pass through
certain control characters (LF, TAB) and all (even non-ASCII) printing
characters, but display other control characters as \x.. escapes
rather than risk putting the terminal in a weird mode.  No quotes
though.  Hm, I kind of like this: when used as intended, it will just
display the text, with newlines and umlauts etc.; but when printing
binary gibberish, it will do something friendly.

There's also the issue of what to do with lists (or tuples, or dicts)
containing strings.  If we agree on this:

>>> "hello\nworld\n\347" # octal 347 is a cedilla
hello
world
?
>>>

Then what should ("hello\nworld", "\347") show?  I've got enough serious
complaints that I don't want to propose that it use repr():

>>> ("hello\nworld", "\347")
('hello\nworld', '\347')
>>>

Other possibilities:

>>> ("hello\nworld", "\347")
('hello
world', '?')
>>>

or maybe

>>> ("hello\nworld", "\347")
('''hello
world''', '?')
>>>

Of course there's also the Unicode issue -- the above all assumes
Latin-1 for stdout.

Still no closure, I think...

--Guido van Rossum (home page: http://www.python.org/~guido/)



From hamish_lawson at yahoo.co.uk  Thu Apr  6 11:16:06 2000
From: hamish_lawson at yahoo.co.uk (Hamish Lawson)
Date: Thu, 06 Apr 2000 08:16:06 -0700
Subject: Refactoring browser for Python?
Message-ID: <16d61288.5cc55c16@usw-ex0101-005.remarq.com>

Is anyone working on a refactoring browser for Python?

See http://st-www.cs.uiuc.edu/~droberts/tapos/TAPOS.htm for a
paper on the Smalltalk original.

Unfortunately I don't think they have any plans to produce a
Python (or any other language) version.

Hamish Lawson


* Sent from RemarQ http://www.remarq.com The Internet's Discussion Network *
The fastest and easiest way to search and participate in Usenet - Free!



From dsavitsk at e-coli.net  Sun Apr 23 15:01:27 2000
From: dsavitsk at e-coli.net (dsavitsk)
Date: Sun, 23 Apr 2000 19:01:27 GMT
Subject: Using DAO and Python
Message-ID: <bMHM4.23114$fF5.684678@news1.rdc1.il.home.com>

I am jumping in in the middle here...
I have a VB program that needs to be more dynamic than VB wants to be.  So I
use vb to generate a python script that the user can then modify.  The
primary purpose of the script is to use DAO or ADO to connect to a ms access
db (it also connects to an NCS OpScan OMR Scanner, but that is a different
story).
my problem thus far, and there will be more, is just getting python to use
dao correctly.
here is what i have tried (esentially following the example on pg 252 of
python programming on win32)
1. ran makepy and selected microsoft dao 3.6 from the list
2. in PythonWin typed
>>> import win32com.client
>>> daoengine = win32com.client.Dispatch('DAO.DBEngine')

at this point there is an error
esentially i know i oversimplified this way too much, but any suggestions as
to what i left out will be greatly appreciated.

---------------------------------------------------------------------
dsavitsk at e-coli.net
http://www.e-coli.net
---------------------------------------------------------------------





From pauldubois at home.com  Thu Apr 13 18:02:19 2000
From: pauldubois at home.com (Paul F. Dubois)
Date: Thu, 13 Apr 2000 22:02:19 GMT
Subject: Hey, the expression parser must be working!
Message-ID: <LtrJ4.19490$k5.394885@news1.frmt1.sfba.home.com>

Just in case anybody was nervous about whether Python arithmetic expressions
have been adequately tested, the following works. Cool, huh? (Composed using
Maple 6 via Fortran codegeneration and a text editor)

from math import sqrt
def cmplx (x, y):
    return x + 1.0j * y

def cubic_roots (a, b, c, d):
    "Roots of a*x**3+b*x**2+c*x+d"
    t0 = 1/a*(36*c*b*a-108*d*a**2-8*b**3+12*sqrt(3.0)*sqrt(4*c**3*a-c\
        **2*b**2-18*c*b*a*d+27*d**2*a**2+4*d*b**3)*a)**(1.0/3.0)/6-2.0/\
        3.0*(3*c*a-b**2)/a/(36*c*b*a-108*d*a**2-8*b**3+12*sqrt(3.0)*sqrt\

(4*c**3*a-c**2*b**2-18*c*b*a*d+27*d**2*a**2+4*d*b**3)*a)**(1.0/3.0)-b/a/3

    t1 = -1/a*(36*c*b*a-108*d*a**2-8*b**3+12*sqrt(3.0)*sqrt(4*c**3*a-\
        c**2*b**2-18*c*b*a*d+27*d**2*a**2+4*d*b**3)*a)**(1.0/3.0)/12+(3*\
        c*a-b**2)/a/(36*c*b*a-108*d*a**2-8*b**3+12*sqrt(3.0)*sqrt(4*c**3*\
        a-c**2*b**2-18*c*b*a*d+27*d**2*a**2+4*d*b**3)*a)**(1.0/3.0)/3-b/\
        a/3+cmplx(0.0,1.0/2.0)*sqrt(3.0)*(1/a*(36*c*b*a-108*d*a**2-8*b\
        **3+12*sqrt(3.0)*sqrt(4*c**3*a-c**2*b**2-18*c*b*a*d+27*d**2*a**2+\
        4*d*b**3)*a)**(1.0/3.0)/6+2.0/3.0*(3*c*a-b**2)/a/(36*c*b*a-108\
        *d*a**2-8*b**3+12*sqrt(3.0)*sqrt(4*c**3*a-c**2*b**2-18*c*b*a*d+27\
        *d**2*a**2+4*d*b**3)*a)**(1.0/3.0))

    t2 = -1/a*(36*c*b*a-108*d*a**2-8*b**3+12*sqrt(3.0)*sqrt(4*c**3*a-\
        c**2*b**2-18*c*b*a*d+27*d**2*a**2+4*d*b**3)*a)**(1.0/3.0)/12+(3*\
        c*a-b**2)/a/(36*c*b*a-108*d*a**2-8*b**3+12*sqrt(3.0)*sqrt(4*c**3*\
        a-c**2*b**2-18*c*b*a*d+27*d**2*a**2+4*d*b**3)*a)**(1.0/3.0)/3-b/\
        a/3+cmplx(0.0,-1.0/2.0)*sqrt(3.0)*(1/a*(36*c*b*a-108*d*a**2-8*\
        b**3+12*sqrt(3.0)*sqrt(4*c**3*a-c**2*b**2-18*c*b*a*d+27*d**2*a**2\
        +4*d*b**3)*a)**(1.0/3.0)/6+2.0/3.0*(3*c*a-b**2)/a/(36*c*b*a-108\
        *d*a**2-8*b**3+12*sqrt(3.0)*sqrt(4*c**3*a-c**2*b**2-18*c*b*a*d+27\
        *d**2*a**2+4*d*b**3)*a)**(1.0/3.0))

    return [t0, t1, t2]

if __name__ == "__main__":
    print cubic_roots(1.0, -1.0, 1.0, -1.0)






From m.faassen at vet.uu.nl  Tue Apr  4 11:40:07 2000
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 4 Apr 2000 15:40:07 GMT
Subject: Python 1.6 alpha 1 released
References: <Pine.GSO.4.10.10004040035420.21084-100000@sundial> <5E877FD7637AB46E.DABBED9CF3C1D64B.CDDDC05AE7374A78@lp.airnews.net> <5l7leern5v.fsf@eric.cnri.reston.va.us>
Message-ID: <8cd2cn$bg8$1@newshost.accu.uu.nl>

Guido van Rossum <guido at cnri.reston.va.us> wrote:
[snip]
> The argument that gopherlib c.s. define a 2-arg connect() doesn't mean
> that socket should, to: those are all Internet specific protocols,
> where host+port is the norm; the lower-level socket API must support
> other address families.

Okay, makes sense.

Regards,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From elb at cfdrc.com  Fri Apr 21 10:09:35 2000
From: elb at cfdrc.com (Ed)
Date: Fri, 21 Apr 2000 09:09:35 -0500
Subject: The Simple Economics of Open Source
References: <14591.15630.192472.440375@goon.cnri.reston.va.us> <slrn8fv2rd.1kt.neelk@brick.cswv.com>
Message-ID: <3900611F.3B525BDA@cfdrc.com>

On the other hand, I didn't think much of the article.  If felt the authors
were snobbish because they gave the impression that programmers behave
somewhat like monkeys, motivated primarily by concerns about group status
and dominance.  The abstract's mention of "career concerns" as a motivating
factor was not explored in any depth; it simply made the abstract more
pallatable.

I thought they left out two important motivations for support/involvement in
open-source software: (1) frustration (with pointy-haired bosses who don't
know what you're talking about until you show them; with dysfunctional
software tools; with commercial-software "documentation" that is really
advertisement); and (2) risk reduction (concern over the potential money pit
that a commitment to Microsoft and commercialized software might bring).

Of course, even though I didn't like it, I'm glad they wrote it.
Misinformation, delivered to economists and marketing agencies whose
interests are to profit from emerging markets by finding influence points,
is a good thing.

Neel Krishnaswami wrote:

> Jeremy Hylton <jeremy at cnri.reston.va.us> wrote:
> >
> > The Simple Economics of Open Source
> >
> > The paper itself is available from one of the authors --
> > http://www.people.hbs.edu/jlerner/simple.pdf
> >
> > [...] The authors are Josh Lerner --
> >     http://www.people.hbs.edu/jlerner/
> > and Jean Tirole --
> >     http://www.elsevier.nl/hes/homepage/cvs/tirole.htm
> >
> > I haven't read the paper yet, but the abstract and the discussion in
> > the Times make it sound like a worthwhile read.  Thought I'd share the
> > pointer to the paper and see if anyone else has read the paper or has
> > comments on it.
>
> The paper is quite good and very readable, as one might expect -- Jean
> Tirole literally wrote the book on market structure and imperfect
> competition[*].  There's nothing much in the way of new economics in
> the paper; it's more like a case study than a theoretical research
> paper.
>
> However, there are a number of excellent insights, among them an
> explanation (in a footnote!) of why the BSDs have been able to keep up
> with Linux technically, even though the Linux community is much
> larger. (Or for that matter, Python and Perl.)
>
> [*] _The Theory of Industrial Organization_: This is a wonderful book;
> it put in my mind the idea that the word "lucid" might have been
> invented just to describe it. Only a few other texts (notably SICP)
> have impressed me this much. If you are at all familiar with
> microeconomics, then you really owe it to yourself to dig this up and
> dig into it.
>
> Neel



From thomas at madeforchina.com  Sun Apr 23 10:06:57 2000
From: thomas at madeforchina.com (Thomas)
Date: Sun, 23 Apr 2000 22:06:57 +0800
Subject: getting the current URL
Message-ID: <4.1.20000423220537.02bd5850@202.95.1.37>

Hi everyone,

I think this is a real easy question, but I can't get it to work.

I have a form on a page.  After posting, I am trying to get the url 
of the page which it just posted from (the same form will be on many pages)

Ultimately, I want my page to reload itself after submitting the request
.... ie.

def reloadpage():
     url= * the page I just submitted from"
     print "Status: 302 Found"
     print "Uri: %s" % url


Thanks!,
Thomas 

Thomas Duterme - IT manager
Madeforchina.com 
Technology based marketing tools made for China. 
___________________________ 
Direct Line - (8610) 6417 2665 (ext) 224
Fax - (8610) 6415 1086 
Business email: thomas at madeforchina.com
Company URL: www.madeforchina.com
Get a life at www.51go.com





From m.faassen at vet.uu.nl  Wed Apr  5 08:05:57 2000
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 5 Apr 2000 12:05:57 GMT
Subject: Tuples -- who needs 'em
References: <00e101bf9d80$a8384d20$74eb0b18@stcla1.sfba.home.com> <va3G4.1110$n68.190454272@newsb.telia.net> <037f01bf9e55$6d328760$74eb0b18@stcla1.sfba.home.com>
Message-ID: <8cfa75$fho$1@newshost.accu.uu.nl>

Bob Alexander <bobalex at home.com> wrote:
> I agree with this, too (noting Lisp as possibly the worst example --
> although a fine language with clear favorable influences on Python). But
> my issue is not about syntax.

> However, too bad about that nasty syntax irregularity that using parens for
> tuples presents  (1,)  (Couldn't resist  ;-)

But tuples are nice in particular for their *syntax*. With tuples, you
can do:

def foo():
    return 1, 2, 3, 4

first, second, third, fourth = foo()

I think this is nice syntax. Without tuples it'd not be as nice. It's not
so much about efficiency. This pattern occurs all over Python.

[snip]
> Yet the concept of array is pretty fundamental to programming. Arrays are
> usually thought of as mutable. Anyone having trouble with that concept is
> pre-beginner  :-)

Anyone having trouble with the concept of immutable lists must have never
heard of functional programming?

[snip]
> I'm sure tuples are not an acciental feature. And my issue comes from
> understanding, not lack of understanding.

Given the above two items, I don't think you understand tuples yet. I
consider tuples to be mostly nice syntax (with one nit, the single element
tuple). You could almost see the immutability semantics being there to
support the syntax, not the other way around. It might be possible to preserve
the nice syntax and still get rid of tuples, if you still allow this:

foo, bar, baz = [1, 2, 3]

But this is less symmetrical, and since lists are mutable, this may go wrong
far more easily.

Regards,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From fig at oreilly.com  Thu Apr 20 10:28:37 2000
From: fig at oreilly.com (Stephen R. Figgins)
Date: Thu, 20 Apr 2000 14:28:37 GMT
Subject: Gtkinter documentation
References: <8F1C7387Elminutieidit@192.6.1.10>
Message-ID: <38FF13D3.BCB79F69@oreilly.com>

Luca wrote:
 
> Where can I find documentation about Gtkinter?
> 
> Luca.

With the PyGTK documentation.  You can find the PyGTK home page at: 

http://www.daa.com.au/~james/pygtk/

Gtkinter.py has been replaced with gtk.py (since about 11/98).  So
unless you are maintaining legacy code, you probably want to look at the
gtk.py documentation.

Stephen Figgins
http://www.oreillynet.com/python


From lminuti at eid.it  Thu Apr 20 05:36:28 2000
From: lminuti at eid.it (Luca)
Date: Thu, 20 Apr 2000 11:36:28 +0200
Subject: Gtkinter documentation
Message-ID: <8F1C7387Elminutieidit@192.6.1.10>

Where can I find documentation about Gtkinter?

Luca.


From mwh21 at cam.ac.uk  Sun Apr 16 13:24:12 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 16 Apr 2000 18:24:12 +0100
Subject: [ANNOUNCE] Garbage collection for Python
References: <20000407221750.A29279@acs.ucalgary.ca> <m3r9cagx3c.fsf@atrus.jesus.cam.ac.uk> <l03102809b51d41895a80@[193.78.237.135]> <38F9F01B.57273052@tismer.com>
Message-ID: <m33domez7n.fsf@atrus.jesus.cam.ac.uk>

Christian Tismer <tismer at tismer.com> writes:

> It will become simpler, simpler and simpler.

I cam across a great quote the other day; I think it was attributed to
Alan Perlis, and it was:

    Simplicity follows complexity, not the other way round.

I think (and in fact have always thought, even before I read the
quote) Stackless Python is an almost perfect example of this.

Cheers,
M.

-- 
  On the other hand,  the following areas are subject to boycott
  in reaction to the rampant impurity of design or execution, as
  determined after a period of study, in no particular order: 
    ...                            http://www.naggum.no/profile.html


From echuck at mindspring.com  Sat Apr 15 19:25:50 2000
From: echuck at mindspring.com (Chuck Esterbrook)
Date: Sat, 15 Apr 2000 19:25:50 -0400
Subject: CGI support question
References: <fUWE4.23$hc7.707@news2-win.server.ntlworld.com>
Message-ID: <38F8FA7E.8901E6C4@mindspring.com>

Richard Chamberlain wrote:

> Can anyone tell me of an ISP who supports python in the UK? Or even better
> would be a site that would give me free space who supports Python.
>
> Thanks
>
> Richard

Go to "Python-Friendly ISPs" and search for "UK".

http://www.corrt.com/info/pyisp-list.html


-Chuck


From m.faassen at vet.uu.nl  Fri Apr 21 22:29:11 2000
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 22 Apr 2000 02:29:11 GMT
Subject: Python for drivers?
References: <3900EDBE.282BC85@dsuper.net>
Message-ID: <8dr2pn$ec8$1@newshost.accu.uu.nl>

Dave <redsand at dsuper.net> wrote:
> Can Python be used for writing device drivers? I just got BeOS 5 and it
> doesn't support some video cards. I'd like to write one for the Cirrus
> Logic card and I want a language that's clear and concise- Python. I've
> never written device drivers. Any suggestions?

Most probably: no. If you got it going, it'd be sloooooooooow.
Python is clear and concise, but it doesn't execute very quickly compared
to C. Python's also not as good as C at access to the bare metal
(the video card); you'd have to get significantly funky to be able to do
that (and you'd need some C in there). So in this case I think that even 
prototyping in Python, then moving to C, isn't going to work. Sorry!
Wish I could've given you more positive suggestions. It's just one of those
cases where Python isn't fast or low-level enough. This is rare, but it
does happen.

Regards,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From breiter at usf.Uni-Osnabrueck.DE  Mon Apr  3 05:21:07 2000
From: breiter at usf.Uni-Osnabrueck.DE (Bernhard Reiter)
Date: 3 Apr 2000 09:21:07 GMT
Subject: Python 1.6 alpha 1 released
References: <200004010159.LAA00403@piglet.dstc.edu.au>
Message-ID: <8c9nq3$ghk$4@newsserver.rrzn.uni-hannover.de>

In article <200004010159.LAA00403 at piglet.dstc.edu.au>,
	David Arnold <arnold at dstc.edu.au> writes:
> -->"Bjorn" == Bjorn Pettersen <bjorn at roguewave.com> writes:
> 
>   Bjorn> The more time I spend on fixing
>   Bjorn> compatibility issues in Python scripts, the less time I spend
>   Bjorn> on "real" work.
> 
> you have been using incorrect arguments.  the fact that they worked
> was unfortunate because it leads to circumstances like this.  but the
> time you have spent on these emails would have been enough to run a
> simple script over your code to fix the problem ... ;-)

Hmm how simple would a script be to get all occurences of:

	socket.connect(what, \ 
			else \
			)

	sc=socket.connect

	sc(what,else)
	sc(what, \
		else)

just a thought.
	Bernhard

-- 
Professional Service around Free Software                (intevation.net)  
The FreeGIS Project				            (freegis.org)
Association for a Free Informational Infrastructure            (ffii.org)


From moshez at math.huji.ac.il  Wed Apr  5 05:23:42 2000
From: moshez at math.huji.ac.il (Moshe Zadka)
Date: Wed, 5 Apr 2000 11:23:42 +0200 (IST)
Subject: Why should I switch to Python? - Infinity of Primes
In-Reply-To: <38EAC8C4.ADC25640@cosc.canterbury.ac.nz>
Message-ID: <Pine.GSO.4.10.10004051110580.24181-100000@sundial>

On Wed, 5 Apr 2000, Greg Ewing wrote:

> Um, no it doesn't - it constructs a number which is
> *either* prime *or* divisible by some other prime bigger
> than the one you started with.

The second case includes the first one, and that is still constructive,
since there is an algorithm to find a prime factor of a number.
--
Moshe Zadka <mzadka at geocities.com>. 
http://www.oreilly.com/news/prescod_0300.html
http://www.linux.org.il -- we put the penguin in .com




From neilh at hare.net.au  Sat Apr  8 21:37:29 2000
From: neilh at hare.net.au (Neil Hodgson)
Date: Sun, 9 Apr 2000 11:37:29 +1000
Subject: win32all-130 for Python 1.6 available
References: <_ziF4.1488$1b4.6926@news-server.bigpond.net.au> <Q4JH4.605$hh6.22805@news2-win.server.ntlworld.com>
Message-ID: <00bc01bfa1c4$2b8a2940$46dd3fcb@neil>

> 1. The following does not fold correctly:
>
> def spam():
>     return """spam
> is
> good"""

   OK. The folding rules will require tweaking for a while to approach
correctness. This one looks fairly easy: Lines that start in triple quote
style may not be fold headers.

> 2. Similarly, the following should perhaps not fold:
>
> # A comment ending with a colon:
> # more comment here

   The colon isn't used in the folding rules so there is some more context
causing this.

   Neil




From tom-main at REMOVEME.home.com  Fri Apr 14 16:06:48 2000
From: tom-main at REMOVEME.home.com (Tom)
Date: Fri, 14 Apr 2000 20:06:48 GMT
Subject: How do I undo an import??
References: <7QrJ4.185421$Hq3.4353581@news2.rdc1.on.home.com> <m3d7nthbbj.fsf@atrus.jesus.cam.ac.uk> <bkKJ4.190780$Hq3.4442202@news2.rdc1.on.home.com> <slrn8festb.8ld.grant@grante.comtrol.com>
Message-ID: <sTKJ4.190986$Hq3.4443569@news2.rdc1.on.home.com>

Uh... no, I didn't.

Sorry, I'm quite new to Python.

Tom.

"Grant Edwards" <grant at nowhere.> wrote in message
news:slrn8festb.8ld.grant at grante.comtrol.com...
> In article <bkKJ4.190780$Hq3.4442202 at news2.rdc1.on.home.com>, Tom wrote:
>
> >After importing NetCfg I type:
> >
> >>>> del sys.modules['NetCfg']
> >
> >Python responds as follows:
> >
> >Traceback (innermost last):
> >  File "<stdin>", line 1, in ?
> >NameError: sys
> >
> >Any ideas??
>
> Did you do an "import sys"?
>
> --
> Grant Edwards                   grante             Yow!  You mean you
don't
>                                   at               want to watch WRESTLING
>                                visi.com            from ATLANTA?




From tom__98 at my-deja.com  Wed Apr 12 18:35:31 2000
From: tom__98 at my-deja.com (tom__98 at my-deja.com)
Date: Wed, 12 Apr 2000 22:35:31 GMT
Subject: Python/Perl Popularity (Re: A Mountain of Perl...)
References: <200004102111.HAA03380@envy.fulcrum.com.au> <8d0pcq$7ti$1@nnrp1.deja.com> <g0ssm03o.fsf@dan.resnet.rochester.edu> <8d1lgu$ga0$1@newshost.accu.uu.nl> <zoqzlazr.fsf@dan.resnet.rochester.edu>
Message-ID: <8d2tnj$k9m$1@nnrp1.deja.com>

In article <zoqzlazr.fsf at dan.resnet.rochester.edu>,
  dan at cgsoftware.com wrote:
> I also don't see why Tom praises the CPAN module so much.
> My main problem with it is
> "
> 203 installed modules have no parseable version number
> "
> Yet i haven't installed any modules for PERL on my own.

I have only used CPAN, not looked into how it is implemented,
but versioning seems to work pretty well: the CPAN tool
will actually check for updated modules and recommend upgrades.
BTW, modules also all come with documentation in a standard format,
as well as a standard set of regression tests.  There is also
some MD5 checksum test, although I'm not sure how much security
against Trojan horses that actually provides.  I recommend giving
it a try; it's an important, widely used example of a
software distribution system.

Just to reiterate: I think CPAN accounts significantly for
the popularity of Perl.  It works pretty well, it's useful, and
I'd like to have something similar for Python.  With something
like CPAN, it's much easier to use non-standard modules and
shared one's scripts and programs with others.  I realize that
there is a SIG and that some work has been done on this for
Python; I think that's great, and the sooner it comes out, the
better.

Tom.


Sent via Deja.com http://www.deja.com/
Before you buy.


From gsez020 at kryten.bedford.waii.com  Wed Apr  5 03:59:53 2000
From: gsez020 at kryten.bedford.waii.com (Pete Forman)
Date: 05 Apr 2000 08:59:53 +0100
Subject: blessed be Guido, for "5 <= X <= 10" does what it should...
References: <3TpG4.3349$HG1.98231@nnrp1.uunet.ca> <slrn8eklr4.9d7.kc5tja@garnet.armored.net>
Message-ID: <tu3dp19dvq.fsf@kryten.bedford.waii.com>

>>>>> "Samuel" == Samuel A Falvo <kc5tja at garnet.armored.net> writes:

    Samuel> In article <3TpG4.3349$HG1.98231 at nnrp1.uunet.ca>, Warren
    Samuel> Postma wrote:
    >> (Actually I'd write the <= as '?' (Unicode U+2264) but using
    >> such things in our code would seriously cramp portability.
    >> Hey, is the APL symbol set anywhere in Unicode!?)

    Samuel> Astonishingly enough, it is.  Or so I've been told.  Not
    Samuel> sure where or how to verify this, though, as I've never
    Samuel> used APL.  It's part of their 32-bit character set.

APL runs from U+2336 to U+237A.  It additionally reuses some other
characters.

BTW Unicode is 16 bit; it is aligned with the BMP subset of the 31 bit
ISO 10646.
-- 
Pete Forman              | Disclaimer: This posting is originated by
Western Geophysical      | myself and does not represent the opinion
pete.forman at westgeo.com  | of Baker Hughes or its divisions.


From ivanlan at callware.com  Tue Apr 18 13:15:57 2000
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Tue, 18 Apr 2000 11:15:57 -0600
Subject: o'reilly python devcenter
References: <Pine.BSF.4.10.10004181241280.15729-100000@chi.pair.com>
Message-ID: <38FC984D.59407BCB@callware.com>

Hi All--

"Michal Wallace (sabren)" wrote:
> 
> I didn't see this posted already, but sorry if its old news.
> O'Reilly started a new python site:
> 
>      http://www.oreillynet.com/python/
> 

Hmmm.  I looked it over, admittedly not in great depth, but:

	1)  There seems to be no mention of www.python.org
	2)  There seems to be no mention of the python mailing lists
	3)  The only books mentioned are O'Reilly books
	4)  There seem to be no links to other websites

Maybe I'm very wrong here, but I think this sort of attempt to control
the vertical and horizontal is a bit foreign to Python.  I like (make
that love) O'Reilly books, but I'll say the same thing I said back when
the Honda Civic made its debut:

	"There's no reason to apply hard-sell arm-twisting techniques here. 
The car sells itself, the engineering is that good."

Of course, that didn't prevent me from getting enough arm-twisting FUD
phone calls from the salesman I made the mistake of giving my name to
that I ended up not buying a Honda.

"Open Source" is a philosophy that applies to more than source.


<please-please-tell-me-i'm-wrong>-ly y'rs,
Ivan
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
http://www.pauahtun.org 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
Author:  Teach Yourself Python in 24 Hours



From Jhlee at Brooks.com  Mon Apr  3 15:26:11 2000
From: Jhlee at Brooks.com (Lee, Jaeho)
Date: Mon, 3 Apr 2000 15:26:11 -0400 
Subject: Creating variables on the fly...
Message-ID: <857F15D7E3D8D3118D290008C7CF058627070F@mail-naeast1.brooks.com>

I think that this is the place you should use list.

What about this.

def add(listVar)
   r = 0
   for something in listVar
      r = r + something
   return r

I'm still learning python. So please forgive any syntax or naming convetion
problem.

Jaeho

>Hi All,
>
>Let's say I have four variables: 
>
>a=1
>b=2
>c=3
>d=4
>
>And I have a function that adds these variables together:
>
>def add(a,b,c,d):
>   return a+b+c+d
>
>But now let's say I have twenty variables that I want to add 
>together.  
>This function will no longer work.  I would have to rewrite it as 
>a+b+c+d+e+f+g+...+(20th letter).  Is there a way to 
>dynamically create a 
>variable name?  So that my add function can automatically 
>determine how 
>many variable names to create and then add the values together.  In 
>other words, I'd ideally like something like:
>
>
>def add(number_of_variables):
>   return a+b+c+...+(letter corresponding to number_of_variables)
>
>Thanks for your help,
>
>Matt



From samschul at pacbell.net  Mon Apr 24 18:32:37 2000
From: samschul at pacbell.net (Sam Schulenburg)
Date: Mon, 24 Apr 2000 22:32:37 GMT
Subject: Princeton Instruments Python interface
Message-ID: <8e2i25$gou$1@nnrp1.deja.com>

I am developing a python library to access Princeton Instruments
Spectroscopic WinX/32 COM objects. If anyone else has had done this I
would like to get in touch with you to share ideas.

Sam Schulenburg


Sent via Deja.com http://www.deja.com/
Before you buy.


From mwh21 at cam.ac.uk  Wed Apr  5 07:36:19 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 05 Apr 2000 12:36:19 +0100
Subject: UserDict's has_key() and get() should use __getitem__()
References: <0535db98.ec878f67@usw-ex0101-005.remarq.com>
Message-ID: <m3u2hgn5jg.fsf@atrus.jesus.cam.ac.uk>

Hamish Lawson <hamish_lawson at yahoo.co.uk> writes:

> However, under the current implementation of UserDict, if you
> redefine __getitem__() in your own UserDict-derived class, you
> will probably also have to redefine has_key() and get() (and
> maybe others) in order to have them behave as expected. This
> wholescale redefining is caused by the fact that UserDict's
> has_key() and get() methods don't make use of the class's own
> __getitem__() method in determining whether a key exists, as
> shown in the extract below:
> 
>     def __getitem__(self, key):
>         return self.data[key]
>     def has_key(self, key):
>         return self.data.has_key(key)
>     def get(self, key, failobj=None):
>         return self.data.get(key, failobj)
> 
> My contention is that the behaviour we expect from get() and
> has_key() is intimately bound up with how __getitem__() behaves,
> and this should be reflected in get() and has_key() making use
> of __getitem__():
> 
>     def has_key(self, key):
>         try:
>             self[key]
>             return 1
>         except KeyError:
>             return 0
>     def get(self, key, failobj=None):
>         try:
>             return self[key]
>         except KeyError:
>             return failobj

Hmm.  These are likely to be rather less efficient than the current
implementation, so you want to penalise those who use UserDict for the
sake of those who don't subclass it thoroughly - do you really want to
do this?  Let's face it, you're unlikely to make this mistake again.

Cheers,
M.

-- 
well, take it from an old hand: the only reason it would be easier
to program in C is that you can't easily express complex  problems
in C, so you don't.                 -- Erik Naggum, comp.lang.lisp


From embed at geocities.com  Mon Apr  3 16:17:38 2000
From: embed at geocities.com (Warren Postma)
Date: Mon, 3 Apr 2000 15:17:38 -0500
Subject: lightweight embedded database table object
References: <sS4G4.2894$HG1.86777@nnrp1.uunet.ca> <lC5G4.102$Ta6.86472@paloalto-snr1.gtei.net>
Message-ID: <5a6G4.2924$HG1.87456@nnrp1.uunet.ca>

> Hi Warren,
> Have you looked at MetaKit? http://www.equi4.com/metakit/python.html

It didn't look 'ready for prime time' -- too many disclaimers of the kind I
usually put on something when I expect it's still quite buggy or unscalable.
I suppose I should give it more of a try though, eh!? Is anyone using this
in a 'production environment'?

Warren




From brent.fulgham at xpsystems.com  Mon Apr 10 13:15:57 2000
From: brent.fulgham at xpsystems.com (Brent Fulgham)
Date: Mon, 10 Apr 2000 10:15:57 -0700
Subject: Python 1.6 -- Any hope for new threading core?
Message-ID: <EDFD2A95EE7DD31187350090279C6767706F14@THRESHER>

With the 1.6 Alpha release, I was wondering if
any changes had been made to the implementation's
threading core.  While 1.5 is a definite improvement
over 1.4, it's still necessary to grab the global
interpreter lock, etc., when writing code for a
multi-threaded C extension.

Just curious if any of this behavior is changed 
with 1.6?

Thanks,

-Brent 



From wpeloqui at ismi.net  Tue Apr  4 13:01:29 2000
From: wpeloqui at ismi.net (Willie Peloquin)
Date: Tue, 04 Apr 2000 17:01:29 GMT
Subject: Starving for an Advanced Python Book
References: <20000404040941.7766.qmail@web2106.mail.yahoo.com>
Message-ID: <8cd74u$9fi$1@nnrp1.deja.com>

Try "Python and Tkinter Programming" by John E. Grayson. It is an
excellent book, full of useful information on using Python in the real
world.

Willie Peloquin


In article <20000404040941.7766.qmail at web2106.mail.yahoo.com>,
lewst <lewst at yahoo.com> wrote:
> Having finished "Learning Python", I'm now looking for an advanced
> Python text. ORA's "Programming Python" is a natural choice, but I
> can't bear buying this outdated book when the 2nd edition is due out
> in a few months (I'm dying here Lutz, hurry it up already!). However,
> I really need something in the interim period. The library reference
> manual is fine but usage is difficult for me to figure out sometimes
> because there are very few examples included.
>
> Does anyone have any suggestions? I suppose I'm looking for a text
> that has examples of how to solve various kinds of programming
> problems in Python and also examples of code showing how the library
> and various modules actually work. How is the "Python Essential
> Reference" on examples? Is it just a rebinding of the python.org
> library reference
>
> Thank-you.
>
> __________________________________________________
> Do You Yahoo!?
> Talk to your friends online with Yahoo! Messenger.
> http://im.yahoo.com
>
>
--
Willie Peloquin
Senior Software Engineer - Contract
http://www.ismi.net/~wpeloqui


Sent via Deja.com http://www.deja.com/
Before you buy.


From billtut at microsoft.com  Thu Apr 20 11:40:19 2000
From: billtut at microsoft.com (Bill Tutt)
Date: Thu, 20 Apr 2000 08:40:19 -0700
Subject: Using popen in windows
Message-ID: <4D0A23B3F74DD111ACCD00805F31D8101D8BCF7D@RED-MSG-50>


> From: Gordon McMillan [mailto:gmcm at hypernet.com]
> 
> Robert Cragie
> 
> > I tried this and it still hangs; cvs.exe is definitely not 
> a built in:
> > 
> > from win32pipe import popen
> > p = popen('c:\\cygwin\\contrib\\bin\\cvs -v','r')
> > str = p.read()
> > print str
> > 
> > Again, the equivalent works fine on Linux. 
> 
> Works fine on NT, too. Bill Tutt put out something that works 
> around the Win9x bug. Searching dejanews on popen and 
> Win95 should find it.

Indeed, the 128 build of the Win32 extensions includes the fix (hack) for
Win9x for this cheesy and lame bug. 

Bill



From aa8vb at yahoo.com  Wed Apr 12 07:58:12 2000
From: aa8vb at yahoo.com (Randall Hopper)
Date: Wed, 12 Apr 2000 07:58:12 -0400
Subject: comparing strings and ints
In-Reply-To: <1256641282-42160637@hypernet.com>
References: <2YmI4.366$rc9.188014080@newsb.telia.net>
 <20000411134358.A2352304@vislab.epa.gov> <1256641282-42160637@hypernet.com>
Message-ID: <20000412075812.A2449004@vislab.epa.gov>

Gordon McMillan:
 |Randall Hopper wrote:
 |> Fredrik Lundh:
 |>  |
 |>  |    Objects of different types always compare unequal, and are
 |>  |    ordered consistently but arbitrarily.
 |> 
 |> I can see 45 == '45' being false, and 45 != '45' being true.  But it seems
 |> to me that not throwing an exception for attempts at an ordered comparison
 |> with logical ordering operators (<, <=, etc.) on objects of different
 |> primitive types (e.g. 45 < '45', 45 > '45') only lets bugs pass through.
 |> 
 |> Is there a case where this could be useful (with its undefined behavior)?
 |> I believe that is the root of the question.
 |
 |I'd happily bet your wife and firstborn child that there's code that
 |relies on it <wink>, if only because after sorting a list, all the ints
 |will end up in a clump, and all the strings in another, etc. 


Oh, I also have __no doubt__ that someone is using this unofficial,
undocumented behavior.  But they don't have a leg to stand on if the
ordering breaks in a future Python version ;-)


Capitalize-on-Python's-undefined-behavior-while-you-can'-ly yours,

Randall

-- 
Randall Hopper
aa8vb at yahoo.com



From vespe at cs.tut.fi  Mon Apr  3 12:11:51 2000
From: vespe at cs.tut.fi (Vespe Savikko)
Date: 03 Apr 2000 19:11:51 +0300
Subject: Python and COM without Automation (Dispatch)
Message-ID: <vphd7o7b1vc.fsf@hiirihaukka.cs.tut.fi>

PythonCOM works very well with COM objects that implement IDispatch
interface. However, I have a number of COM objects that do not
implement IDispatch and I would like to use them from Python as well.

>From the book _Python Programming on Win32_ I gather that I would have
to create a PythonCOM extension DLL and that might even be quite a 
simple and possibly automated process (with the help of SWIG and other
tools), but I haven't been able to find any detailed instructions on
the subject.

I haven't walked through the win32com source code yet so if the
necessary information is located somewhere among the source files, you
can ignore this message. 

Any help is appreciated.

  ++Vespe

-- 
  ------------------------------------------------------------------
       Vespe Savikko     vespe at cs.tut.fi     - to doom de doomsday -


From pecora at anvil.nrl.navy.mil  Fri Apr 28 07:51:34 2000
From: pecora at anvil.nrl.navy.mil (Louis M. Pecora)
Date: Fri, 28 Apr 2000 11:51:34 GMT
Subject: Do I always have to write "self." ?
References: <270420001706386648%pecora@anvil.nrl.navy.mil> <8eacdf$nnm$1@newshost.accu.uu.nl>
Message-ID: <280420000751353947%pecora@anvil.nrl.navy.mil>

In article <8eacdf$nnm$1 at newshost.accu.uu.nl>, Martijn Faassen
<m.faassen at vet.uu.nl> wrote:

> There's a way, but stuff can also become more cluttered. In some
> cases it'll also be a bit faster, as local variable access is
> faster than access through self. In lots of other cases it can be slower.
> I'll demonstrate:
> 
> class Foo:
>     def __init__(self):
>         self.data = []
>         self.more = 10
>         self.interesting = "foo"
>         
>     def involved_method(self, something):
>         data, more, interesting = self.data, self.more, self.interesting
> 
>         if more > 15:
>             data.append(something)
>             data.append(interesting)
> 
> But note this!
> 
>     def uhoh_method(self):
>         data, more, interesting = self.data, self.more, self.interesting
>         # this changes self.data, neat
>         data.append('foo') 
>         # this won't change self.more!
>         more = more + 1
>         # and this won't change self.interesting
>         interesting = interesting[10:]
> 
> So, beware when you want to change attributes that refer to immutable
> objects. You do have to use self in that case. I think that you'll
> find that eventually you'll get used to 'self', and you may even
> be grateful for its presence. At least you can see instantly where
> your variable are coming from. The local variable trick, used in moderation,
> can help some as well.

Martijn,

Thanks for the very nice tutorial.  Your example brings up some typical
stumbling blocks for us beginners.  Namely, references to mutable and
immutable objects.

In the above I would describe things as follows (am I right or wrong?):

(1) data and self.data point to the same object, a list, so when a
method is called it is the method that works on that instance of a list
and either data or self.data will append 'foo' to the same list object.

(2) when you do more=more+1 an new object (the more+1 value) is created
and more then points to it and no longer to the old value of self.more. 
The latter still exists and points to the original 10 object.

(3) Similar story here for interesting where a new object
interesting[10:] is created and interesting is "switched" to point to
it.  self.interesting still points tohe "foo" object.

I hope I got that right, because they have tripped me up a few times.

Thanks, again.


From aaron_watters at my-deja.com  Mon Apr 24 11:48:02 2000
From: aaron_watters at my-deja.com (aaron_watters at my-deja.com)
Date: Mon, 24 Apr 2000 15:48:02 GMT
Subject: ANN: NJ Piggies welcome Andy Robinson
References: <39044E57.58467ED@reportlab.com> <m2r9bvv9oj.fsf@netmeg.net>
Message-ID: <8e1qbb$nit$1@nnrp1.deja.com>

In article <m2r9bvv9oj.fsf at netmeg.net>,
  Les Schaffer <godzilla at netmeg.net> wrote:
> > Please join us as the New Jersey Python interest group (NJ Piggies)
> > welcome Andy Robinson, Chief Architect and CEO of ReportLab,
> > visiting from London.
>
> what is planned for this occasion? lecture? topics?

This is strickly a social/networking event.
Modest consumption of alcoholic beverages will
be permitted. ;c)

FWIW: If anyone is interested a number of the
attendees may be in a hiring mood, although I promise
nothing.

  Aaron Watters (http://www.chordate.com/map.pdf)


Sent via Deja.com http://www.deja.com/
Before you buy.


From johann at physics.berkeley.edu  Tue Apr  4 14:21:34 2000
From: johann at physics.berkeley.edu (Johann Hibschman)
Date: 04 Apr 2000 11:21:34 -0700
Subject: Tuples -- who needs 'em
References: <00e101bf9d80$a8384d20$74eb0b18@stcla1.sfba.home.com> <va3G4.1110$n68.190454272@newsb.telia.net> <037f01bf9e55$6d328760$74eb0b18@stcla1.sfba.home.com>
Message-ID: <mtog7pybf5.fsf@astron.berkeley.edu>

Bob Alexander writes:

> I'm aware that this issue is not some great flash of insight. It is so
> obvious that the Python in-group certainly have given it plenty of thought.
> But that doesn't mean it's not worth talking about,

Like so many things...

In any case, I'll throw in my two cents.  Tuples and lists seem
conceptually much different to me, and I expect that they feel much
different to a lot of people, but I don't expect they feel that way to
everyone.  ;-)

Tuples are mathematical.  They have fixed lengths.  If python
variables are a space V, an N-tuple is V^N.  Tensor products, and all
that.

Lists are, well, lists.  I put things in them.  I take things out.
They've got nothing to do with tuples.

Tuples can be used to let functions return more than one value,
without the conceptual overhead of making a class.(1)  Lists are just
the wrong thing for that, they send the wrong message.  Tuples wrap up
elements from larger spaces, like points, (name, address) pairs, and
so on.

Ooh.  I think I hit my big reason.  Ignore the actual details of
typing, and so on.  I see lists as being mostly homogeneous,
regardless of their implementation.  They contain a list of names, or
numbers, or widgets.

I see tuples as combining different types into one conceptual whole.
(name, age) pairs.  (x, y) points.  (place, time).  They mentally bind
tighter than lists, turning other objects into new composite objects.
Lists just string them together.

Now python doesn't enforce this typing, but that's the way I think of
them.  And in that mindset, tuples are a completely natural thing.
Most applications could use lists instead of tuples, but there
wouldn't then be any way to write the more elementary concept of
"tuple-ness".  And I like having that concept around.

I like being able to write

  (x, y) = calculate_point(foo)
  a, b   = b, a
  name, sex, birthdate = unpack_info_string(str)

and so on.

  [x, y] = calculate_point(foo)
  [a, b] = [b, a]
  [name, sex, birthdate] = unpack_info_string(str)

just seem wrong.

Those aren't really technical arguments.  They're "expressiveness"
arguments.  And, to be honest, I have no idea if they convey the
point.  You may be right; it may be a religious issue.  (And I may be
crazy.  But it just might be a lunatic you're looking for.)  But, IMHO,
it's religion with style!


Cheers,

--Johann


1. By the way, C++ would be a much nicer language if it had a good way
   to return tuples without all the template cruft.  ML is your
   friend.

2. Also, the lisps make this distinction.  Common Lisp has values and
   multiple-value-bind, while R5RS Scheme now has values and, well,
   complicated macros to write multiple-value-bind.

-- 
Johann Hibschman                           johann at physics.berkeley.edu


From rcunning at acm.org  Wed Apr  5 14:21:36 2000
From: rcunning at acm.org (Robert W. Cunningham)
Date: Wed, 05 Apr 2000 18:21:36 GMT
Subject: Pass by reference ?
References: <38E864C9.E1EE94F7@ecs.soton.ac.uk> <38EAF8D8.4B7C43B4@ecs.soton.ac.uk> <38EB56AE.F938CE06@acm.org> <m33dp0muk9.fsf@atrus.jesus.cam.ac.uk> <38EB63BC.FE03BA31@ecs.soton.ac.uk> <m3ya6sle0j.fsf@atrus.jesus.cam.ac.uk> <38EB6EB1.9759397C@ecs.soton.ac.uk>
Message-ID: <38EB838A.53122AF6@acm.org>

Jacek Generowicz wrote:

> Michael Hudson wrote:
>
> > The code you posted only confuses if you have a flawed understanding of
> > assignment in Python
>
> That is exactly the point. My original confusion arose precisely because I had
> made an incorrect assumption about assignment in Python (an assumption which, I
> suspect, many newcomers to the language will make until they learn the truth).
> Correcting this error is far more constructive than discussing what name should be
> given to the semantics of argument passing.

IMHO, they are, ultimately, one and the same.  It is useful to be able to discuss
Python assignment and passing semantics and syntax with those who know nothing about,
or have no interest in, Python itself, or are so new to the language that the only
common frame of reference is either academic (formal, theoretical) or via analogies
to other languages (informal, practical).  The abstractions are important, and thus,
so is the nomenclature (if the abstractions are to be understood and shared).  The
notion of "side effects" in general, as related to passing and assignment, is a
fairly high-level issue with very low-level implications.

So clearing up the terminology gets us all on the same page, so to speak, and thus
allows us to discuss Python specifics in the context of theory AND in comparison to
other languages.

In particular, living under "an assumption which, I suspect, many newcomers to the
language will make until they learn the truth" is EXACTLY the attitude I wish to
avoid.  If it is true, it should not require learning the entire language to
UNDERSTAND it.  It does, however, require learning the language to USE it.  It seems
obvious to me that allowing understanding to PRECEDE use is not a bad way to go,
since it allows you to build on existing knowledge (all that theory from college),
rather than be forced to start from zero.

The CS models of "Pass By Reference" and "Pass By Value" are fairly well understood,
but they do not seem to map simply and directly to Python.  Rather than say "Give up
on theory, just learn Python", I'd rather EXTEND (or clarify) the theory to account
for Python!  Then, I'm certain, I will learn Python better and faster.

It is the ability to make generalizations that distinguishes those who learn quickly
("I get the general idea, now lets focus on the details") from those who can't ("Just
give me enough examples and a good interpreter and debugger, and I'll get it
eventually").  The abstractions matter, especially when they can be mapped to
accepted theory.  If for no other reason, theory is important simply because it
allows us to form generalizations that permit previously learned knowledge to be
employed (remapped) in the effort to learn a new language.

There is not the "World Of All Other Computing Languages" and "Python" off by
itself.  It is a continuum that has a large and powerful body of theory and practice
behind it.  It makes sense to show how Python is part of that continuum before just
saying something self-limiting, like "Python is Different - Get used it".  Or that
newcomers should wait "until they learn the truth".

Those sentiments may apply to Perl <g>, but certainly not to Python!


Looking-for-how-Python-fits-into-the-grand-scheme-of-things-ly Yours,

-BobC




From richard_chamberlain at ntlworld.com  Wed Apr 19 05:37:35 2000
From: richard_chamberlain at ntlworld.com (Richard Chamberlain)
Date: Wed, 19 Apr 2000 10:37:35 +0100
Subject: win32all-128.exe availability?
References: <38fd01b2$0$10982@senator-bedfellow.mit.edu>
Message-ID: <8djuos$5oh$1@gxsn.com>

I have a copy as well, so if you have difficulties mail me and I'll forward
it to you.

Richard
"Robert" <soundhack at hotmail.com> wrote in message
news:38fd01b2$0$10982 at senator-bedfellow.mit.edu...
> Now that starship has crashed, I was wondering if it was possible to
> get the 128 build of Pythonwin extensions off the net? By an unfortunate
> coincidence, my own hard drive chose to crash around the time starship
> did.
>
> Thanks
> Robert
>
>




From johngrayson at home.com  Sat Apr 15 20:09:13 2000
From: johngrayson at home.com (johngrayson at home.com)
Date: Sun, 16 Apr 2000 00:09:13 GMT
Subject: Combobox in Tkinter
References: <38F8F133.54FCC64B@swt.edu>
Message-ID: <8db0b1$b3o$1@nnrp1.deja.com>

In article <38F8F133.54FCC64B at swt.edu>,
  "Shah, Navneet" <ns566645 at swt.edu> wrote:
> How to decrease the width of the combobox in Tkinter? Do we have any
> option for that
>
> Shah
>
>

set the width of the entryfield component:

assuming the combobox id is 'xxx':

   xxx.component('entryfield').configure(width=4)


Sent via Deja.com http://www.deja.com/
Before you buy.


From junkster at nospam.rochester.rr.com  Mon Apr 17 18:47:47 2000
From: junkster at nospam.rochester.rr.com (Benjamin Schollnick)
Date: Mon, 17 Apr 2000 22:47:47 GMT
Subject: TKMessageBox, and TKINTER console?
Message-ID: <rNHvjEdhm5Pp-pn2-Y8HWP45V5PG0@d185d18f2.rochester.rr.com>

Folks,

	Crazy Question..... I'm starting to mess around with Tkinter, and
for now at least, just trying to display "pop-up" dialogs.  I found 
the 
TKMESSAGEBOX unit, in the lib-tk directory, and that works great for
my purposes.

	Except, for a small "appearance" problem.  I call the TKmessagebox
routines, and their are TWO windows.  One is the pop-up I've called, 
and the
other is the Tkinter "Console" (??).  Or at least that's the nickname,
I've given it.

	I've tried a variety of different things to get rid of it...For 
example, calling
junk= TK.tk() to a variable, and passing it as "master=junk", or as 
"parent=junk"
in the code...

	I'm not sure if there is anyway to get rid of that darn console.

	Anyone got some suggestions for a really confused first time 
Tkinter'er?

			- Benjamin
====================================================
          (Remove "NoSpam" to Email me)
====================================================
Please feel free to copy any and or all of this sig.
A little something for spam bots:

root at localhost postmaster at localhost admin at localhost
abuse at localhost postmaster at 127.0.0.1

Chairman William Kennard: bkennard at fcc.gov 
Commissioner Susan Ness: sness at fcc.gov
Commissioner Harold Furchtgott-Roth: hfurchtg at fcc.gov
Commissioner Michael Powell: mpowell at fcc.gov
Commissioner Gloria Tristani: gtristan at fcc.gov
consumerline at ftc.gov
fccinfo at fcc.gov
ssegal at fcc.gov



From bellman at lysator.liu.se  Tue Apr  4 18:18:42 2000
From: bellman at lysator.liu.se (Thomas Bellman)
Date: 4 Apr 2000 22:18:42 GMT
Subject: Python 1.6 alpha 1 released
References: <Pine.GSO.4.10.10004040035420.21084-100000@sundial> <5E877FD7637AB46E.DABBED9CF3C1D64B.CDDDC05AE7374A78@lp.airnews.net> <8cca93$9oh$1@nnrp1.deja.com> <B964E942A897FF5F.529887C74CF2FC7E.977238F9475EAC23@lp.airnews.net> <FdoG4.5023$74.81126@newsc.telia.net>
Message-ID: <8cdpo2$l99$1@newsy.ifm.liu.se>

"Fredrik Lundh" <effbot at telia.com> wrote:

> 1) stumbled upon one today: it turned out that by changing from 1.3
> to 1.5.2, large data transfers from a high-performance server to clients
> on an 10 Mbit Ethernet were unexpectedly terminated.  the same trans-
> fers worked just fine with high-performance workstations, on an FDDI
> network.  ten guru points to anyone who can explain why.

Sounds like problem with segmentation on the different frame sizes
of Ethernet and FDDI.  Probably you were using TCP, and switching
to Python 1.5.2 made you send data that didn't fit inside an
Ethernet frame (1500 octets, including IP and TCP headers).  The
sending TCP stack thus split the stream into several TCP segments.
The program that received the data wasn't prepared for the fact
that TCP can segment data any which way it wants to, and broke
when the socket didn't return the entire record in one read(2).
The data didn't grow to overflow the 4352 octet MTU for FDDI
frames, thus the TCP stack still used only a single segment then.

Classic bug.  I have seen it in several programs.  The latest
example is from the IP masquerading code in Linux kernels, at
least of 2.0 and 2.2 vintage.  (Of course, *I* have never made
that mistake.  That I know of, at least...)

Of course, that is only my first guess.  There could be other
reasons.  One possibility could be that your programs were timing
sensitive, and the lower speed of 10 Mbit/s Ethernet triggered the
bug.  And I could probably think up other explanations that fit
the information you have given us.  But I believe TCP segmentation
is the most probable.


soon-to-be-neighbours-again'ly yours,

-- 
Thomas Bellman,   Lysator Computer Club,   Link?ping University,  Sweden
"Beware of bugs in the above code; I have    !  bellman @ lysator.liu.se
 only proved it correct, not tried it."      !  Make Love -- Nicht Wahr!


From thomas.heller at ion-tof.com  Wed Apr 26 12:22:29 2000
From: thomas.heller at ion-tof.com (Thomas Heller)
Date: Wed, 26 Apr 2000 18:22:29 +0200
Subject: [Win32] Retrieve icon for file type?
References: <CC59868A0A76D311B1F50090277B321C1F32B4@VRTMAIL>
Message-ID: <004b01bfaf9b$9e3f1830$4500a8c0@thomasnb>

> I'm attempting to use the Windows registry to retrieve (and convert to a
> PNG/JPEG) the Windows icons for given files (so that a Windows-based
server
> can display file-type-specific icons).  I do not seem to be able to find
the
> function shellapi.ExtractIcon (
> http://msdn.microsoft.com/library/psdk/winui/icons_7h2m.htm ) in the Win32
> extensions.

You might want to look at
MEInc/Dist/icon.py in Gordon's win32 installer package.
This file uses LoadResource to extract icons from executable files.
(Note however that you get the raw data, and not an HICON).

Thomas




From andymac at bullseye.apana.org.au  Wed Apr  5 09:06:31 2000
From: andymac at bullseye.apana.org.au (Andrew MacIntyre)
Date: Wed, 5 Apr 2000 23:06:31 +1000 (EST)
Subject: Tkinter and C extensions
Message-ID: <Pine.OS2.3.95.1000405222214.1184B-100000@CENTRAL>

I am looking at writing a C extension for python which links to a C
library (which is commercial and I don't have access to source), with a
python wrapper to tidy up a couple of rough edges.

This library can make use of an X "canvas", the details of which are
passed through an API function in terms of its X[lib|t] Display pointer 
and Window handle.

As I would like to direct the library to use a Tkinter canvas widget for
its display, I'm looking for a way to get these details about the target
widget.

My research has turned up ways of determining these via Tk's API
(Tk_Display(), Tk_Window() in tk.h) provided I can extract the tkwin
handle from Tkinter's canvas widget - something I've yet to figure out how
to do from within a C extension.

I'm wondering whether its possible to get at this information via Tkinter,
which would simplify the interactions between the various components.

I've perused the Fredrik's Tkinter docs, but I didn't find anything that
appeared to get close to what I think I'm looking for :-(.

I have downloaded the "Life preserver" stuff, but have yet to make much
headway with it.

I went looking for the GD module, which I thought might possibly have
done something like this, but following the link from the Vaults of
Parnassus yielded nothing.

Any hints on additional sources of enlightment, such as other modules
which have similarities, would be appreciated.

--
Andrew I MacIntyre                     "These thoughts are mine alone..."
E-mail: andrew.macintyre at aba.gov.au    (work) | Snail: PO Box 370
        andymac at bullseye.apana.org.au  (play) |        Belconnen  ACT  2616
        andymac at pcug.org.au           (play2) |        Australia




From anders.eriksson at morateknikutveckling.se  Fri Apr 28 04:49:18 2000
From: anders.eriksson at morateknikutveckling.se (Anders M Eriksson)
Date: Fri, 28 Apr 2000 10:49:18 +0200
Subject: newbie   dbase dataaccess
References: <ms19e8.uv5.ln@dorian.georks.de>
Message-ID: <mtiigskn8uumhk94rklnr2fdealunasoot@4ax.com>

On Thu, 27 Apr 2000 11:40:46 +0200, "Georg Hammerbacher"
<don.limpio at gmx.net> wrote:

>Hello everybody,
>

Hello Georg!

>I?m just new in programming Python and now I want to make my own personal
>adressmanager with a simple dBase-file.
>How can I get access to it or where can I find more info?s about using
>dbase-files within Python?
>
First: Why do you want to use dBASE as the database format?

I haven't found any good dBASE interface for Python. So if you want to
use the dBASE format you need to use ODBC and then you hopefully are
on a computer that runs unders Windows NT/98.

If you don't NEED to use dBASE then look at
http://www.python.org/topics/database/ here you will find alot of info
about how to use Python for accessing databases....

// Anders



From mwh21 at cam.ac.uk  Wed Apr  5 04:13:12 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 05 Apr 2000 09:13:12 +0100
Subject: Getting the argument names passed to a function ?
References: <20000404225810.A12156@sz-sb.de> <HRsG4.5120$74.82867@newsc.telia.net> <slrn8el27m.3h1.quinn@riyal.ugcs.caltech.edu> <slrn8el5t7.40k.quinn@riyal.ugcs.caltech.edu>
Message-ID: <m38zytnexz.fsf@atrus.jesus.cam.ac.uk>

quinn at riyal.ugcs.caltech.edu (Quinn Dunkan) writes:

> import sys
> def arghack(arg):
>     try:
>         1/0
>     except:
>         frame = sys.exc_info()[2].tb_frame.f_back
>     while frame:
>         for k, v in frame.f_locals.items():
>             if arg is v:
>                 return k
>         frame = frame.f_back
>     else:
>         return None

You don't need to scan back through the frames do you?  I mean the
argument has to be visible just one up for it to be passed to the
function...

Anyway, it says in my contract that I have to answer questions like
this - so we'd better have:

import sys
from bytecodehacks.code_editor import EditableCode
from bytecodehacks.ops import *

def argHack(var):
    try:
        raise TypeError
    except TypeError:
        frame = sys.exc_traceback.tb_frame.f_back
        code = EditableCode(frame.f_code)
        op = code.co_code[code.co_code.byte2op[frame.f_lasti] - 1]
        if op.__class__ is LOAD_FAST:
            return code.co_varnames[op.arg]
        elif op.__class__ in [LOAD_NAME,LOAD_GLOBAL]:
            return code.co_names[op.arg]
        else:
            return None

There are various problems with extending this to being "useful",
mainly to do with working out the layout of the stack.  Tedious.

It seems to work, though:

>>> x=1
>>> def test(s): return argHack(s), argHack(x)
... 
>>> test(1)
('s', 'x')


Cheers,
M.

-- 
well, take it from an old hand: the only reason it would be easier
to program in C is that you can't easily express complex  problems
in C, so you don't.                 -- Erik Naggum, comp.lang.lisp


From wayne.izatt at myself.com  Sat Apr 22 12:10:56 2000
From: wayne.izatt at myself.com (Wayne Izatt)
Date: Sat, 22 Apr 2000 16:10:56 GMT
Subject: Installing PyApache
Message-ID: <3901CF0C.E9A722F6@myself.com>

Hello,

I tried to build Apache with PyApache. I already have Python 1.5.2
installed (with tkinter).

The build dies when linking, complaining that a whole bunch of tk...
references can't be resolved.

Now, since PyApache probably won't be needing tkinter ;-) is there any
simple way to prevent this situation.

(I suspect that rebuilding Python without tkinter support might do the
trick, but that seems like a long way round).

cheers



From webmaster at python.org  Sun Apr  2 15:22:28 2000
From: webmaster at python.org (Python.Org Webmaster)
Date: Sun, 2 Apr 2000 15:22:28 -0400 (EDT)
Subject: www.python.org back up
Message-ID: <14567.40436.305264.573689@anthem.cnri.reston.va.us>

Sorry folks, I'm not sure why www.python.org was down, but I suspect
it had something to do with the monthly rotation of log files.  I've
restarted the web server so everything should be hunky dory again.

-Barry



From mnaumann at pacbell.net  Wed Apr  5 20:22:25 2000
From: mnaumann at pacbell.net (Michael Naumann)
Date: Wed, 05 Apr 2000 17:22:25 -0700
Subject: Paid to work in Python? (was: Re: Choice Of Language)
References: <NDBBJPNCJLKKIOBLDOMJMEMECDAA.DavidA@ActiveState.com> <8c7tt5$7dl$1@slb3.atl.mindspring.net>
Message-ID: <38EBD8C0.86FD484D@pacbell.net>

Hmmmmm...  It's actually a breeze to work with -- I have no idea
what problems you may be having...


Aahz Maruch wrote:

> In article <NDBBJPNCJLKKIOBLDOMJMEMECDAA.DavidA at ActiveState.com>,
> David Ascher <DavidA at ActiveState.com> wrote:
> >Grant Edwards
> >>
> >> Pointless tanget: Is it just me, or is Google really the last usable
> >> search engine?
> >
> >And it's _all_ due to their use of Python!  Cool, eh? =)
>
> That doesn't explain why Infoseek is no longer particularly usable.
> --
>                       --- Aahz (Copyright 2000 by aahz at netcom.com)
>
> Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
> Hugs and backrubs -- I break Rule 6
>
> "It's 106 miles to Chicago.  We have a full tank of gas, a half-pack of
> cigarettes, it's dark, and we're wearing sunglasses."  "Hit it."



From rupole at compaq.net  Sun Apr  9 23:34:56 2000
From: rupole at compaq.net (Roger Upole)
Date: Mon, 10 Apr 2000 03:34:56 GMT
Subject: win32com and Reflection2
References: <WE8I4.100257$ra.2387838@news.easynews.com> <s8aI4.48$UB.261@news-server.bigpond.net.au>
Message-ID: <AZbI4.101099$ra.2401798@news.easynews.com>

I had tried that already, but the Reflection2 object did not show up on the
list when I ran the makepy utility.  I assumed at the time that this meant
that
this library was not in a format that makepy could map.  However, after
a look at the docs in makepy, I noticed that it would also take the name
of the .TLB file, so I passed it c:\rwin\r2510e.tlb instead of the library
name.
It worked like a charm !

Conjecture: Couldn't find it by the library name because the RWIN
directory is not on the system path ?
Now I can go back and replace all those hardcoded integer parms :)

If Guido was god, Mark Hammond would be the high saint of win32!
(or the devil's advocate, according to the Unix guys ...)

           Many Thanks,
               Roger Upole

"Mark Hammond" <mhammond at skippinet.com.au> wrote in message
news:s8aI4.48$UB.261 at news-server.bigpond.net.au...
> You should be able to run "makepy" over the reflection object (makepy will
> present a list of objects).  You can find "makepy" on the Pythonwin
> "tools" menu.
>
> Once you have done that and created the Reflection object, the constants
> should be available in win32com.client.constants
>
> eg,
>
> win32com.client.constants.SomeConstantName
>
> Note that the Reflection constants wont appear until you have created the
> reflection object using win32com.client.Dispatch()
>
> Mark.
>
> "Roger Upole" <rupole at compaq.net> wrote in message
> news:WE8I4.100257$ra.2387838 at news.easynews.com...
> >    I'm scripting Reflection2 sessions with win32com.client.Dispatch.
> > However, there are some constants (error codes, transmission modes,
> > etc) in a typelib that I can't seem to get at from the Dispatch object.
> > Anybody know of a way to access these from Python ?
> >    If not, is there any way to get to them from C++ ?
> > I have very little experience with C, but if it is simple I can
> > probably blunder my way through it.
> >              Roger Upole
> >
> >
> >
>
>




From ivanlan at callware.com  Wed Apr  5 01:31:31 2000
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Tue, 04 Apr 2000 23:31:31 -0600
Subject: APL Characters in Unicode
References: <NDBBKEGCNONMNKGDINPFMELFDFAA.infonuovo@email.com>
Message-ID: <38EACFB3.4C37E8D8@callware.com>

Hi All--

"Dennis E. Hamilton" wrote:
> 

[bobbitry]

>  (This
> is like getting lost in a dictionary.  What does quincuncial mean anyhow?)
> 

[use monospaced font]

 .   .
   .
 .   .

<see-gordon-i-told-you-glyphs-would-come-in-handy-sometime>-ly y'rs,
Ivan;-)
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
http://www.pauahtun.org and
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
Author:  Teach Yourself Python in 24 Hours



From mmiller3 at iupui.edu  Tue Apr 18 16:29:09 2000
From: mmiller3 at iupui.edu (Michael A. Miller)
Date: 18 Apr 2000 15:29:09 -0500
Subject: Split
References: <38FCC20D.624AB2F6@searchy.net>
Message-ID: <87k8hvywyy.fsf@lumen.med.iupui.edu>

>>>>> "Penpal" == Penpal International <ppi at searchy.net> writes:

    > I've already managed to split a string with this:

    > split = regsub.split(line,'\s+')
[...]

    > But how can I print only the 5th for example?

You can also use the string module to split:

>>> import string
>>> line = 'This is a test, short and sweet, of split.'
>>> string.split(line)
['This', 'is', 'a', 'test,', 'short', 'and', 'sweet,', 'of', 'split.']
>>> string.split(line)[4]
'short'
>>> string.split(line,'a test')
['This is ', ', short and sweet, of split.']


From aa8vb at yahoo.com  Wed Apr 19 06:47:29 2000
From: aa8vb at yahoo.com (Randall Hopper)
Date: Wed, 19 Apr 2000 06:47:29 -0400
Subject: How to write to printer in Windows?
In-Reply-To: <38FC43DF.78A38D5D@tcac.net>
References: <38FC43DF.78A38D5D@tcac.net>
Message-ID: <20000419064729.A517358@vislab.epa.gov>

Albert Wagner:
 |How can I write(aLine) to a printer in Windows?  I tried this:
 |
 |from win32pipe import popen
 |:
 |self.printer = popen('lpt1' 'w')
 |:
 |self.printer.write(aLine)
 |
 |but get a general protection fault.  Any help will be greatly
 |appreciated.

I see you have a pointer to the win32 printer modules in another reply.

The problem with your above code may be something simple though.  From my
DOS days, I seem to recall that 'lpt1:' was the correct name for the
printer device.  lpt1 may be trying to write to a file in a directory for
which you don't have permission.

In a DOS box, try 'copy c:\config.sys lpt1:'.  That'll make sure you're
printer is attached to that device.  If so, you should be able to run your
Python program there with (I would think) the same results.

-- 
Randall Hopper
aa8vb at yahoo.com



From echuck at mindspring.com  Mon Apr 24 17:57:48 2000
From: echuck at mindspring.com (Chuck Esterbrook)
Date: Mon, 24 Apr 2000 17:57:48 -0400
Subject: sys.exit
Message-ID: <3904C35C.B16BDE75@mindspring.com>

The short question:

If I catch a SystemExit exception (because of an invocation of sys.exit(code)), how do I examine the argument to sys.exit()?


The long story:

I have a try..except like so:

	try:
	    execfile(filename, namespace)
	except:
	    # handle problems

I'd like to treat the case of "sys.exit(0)" as a special case in the exception handler because I'm executing another script. If that script calls sys.exit(0), that's OK. If it throws any other kind of exception, I need to log it and report it.

My attempt in the "except:" is this:

	self._errorOccurred = 1
	excInfo = sys.exc_info()
	if excInfo[0]==SystemExit:
		data = excInfo[1]
		if data==0 or data==None:
			self._errorOccurred = 0

But "_errorOccured" is never set to 0, because 'data' isn't what I expected. Upon printing the "excInfo" tuple I get:

    (<class exceptions.SystemExit at 80afc70>, <exceptions.SystemExit instance at 8110f58>, <traceback object at 811f540>) 

And the attributes of the SystemExit instance are:

    ['__doc__', '__init__', '__module__'] 

I don't know how to get my hands on the exit code.


-Chuck


From M_Pope at shef.rage.co.uk  Tue Apr 11 06:13:42 2000
From: M_Pope at shef.rage.co.uk (Mark Pope)
Date: Tue, 11 Apr 2000 11:13:42 +0100
Subject: Solution: Extending built-in language (Win 95)
Message-ID: <213CDEBFDDA7D3118081009027E40C5607111D@SHEF-MS1>

> Mark Pope <M_Pope at shef.rage.co.uk> wrote:
> > I've added xxmodule to the lib and recompiled. Entering:
> > a = xx.new()
> > just gives a name error on xx.
> >
> > Have I missed something?
> 
> 1) the Modules/Setup file is used to add modules
> to the standard library.
> 
> 2) you forgot to *import* the xx module.


Thanks for the help all,

On Win95 there is no Modules/Setup file. Instead one needs to manually edit
config.sys

then recompile all and importing and running works...

thanks,
m.



From M_Pope at shef.rage.co.uk  Fri Apr 14 11:09:57 2000
From: M_Pope at shef.rage.co.uk (Mark Pope)
Date: Fri, 14 Apr 2000 16:09:57 +0100
Subject: Building Core Python
Message-ID: <213CDEBFDDA7D3118081009027E40C56071139@SHEF-MS1>

Hi,

Does anyone have a list of what all the files in the 'Objects' and 'Modules'
directories do?

I'm trying to reduce my (Win95) build to be the minimum 'core' python, to
reduce Exe size.

It's difficult to work out what they all do, and whether it is core, due to
a lack of comments at the top of source files.

Thanks,

mark.



From effbot at telia.com  Tue Apr 11 03:02:05 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Tue, 11 Apr 2000 07:02:05 GMT
Subject: Regular Expression Help for Python Newbie.
References: <005f01bfa2cf$b99b0ea0$f0c809c0@lslp7o.lsl.co.uk>
Message-ID: <N5AI4.1125$Za1.20080@newsc.telia.net>

Tony J Ibbs wrote:
> Fredrik Lundh replied:
> > <TR>'s are optional -- if the browser stumbles upon <TD>
> > in a <TABLE> context, it should insert <TR>'s all by itself.
>
> Well, no...

you're right -- but only in theory ;-)

> Of course, it's entirely possible to write a browser (or other HTML
reader)
> that can cope with missing <TR>'s *if* the </TR>'s are present.

it's easier than that, in fact -- according to the data model,
a <TABLE> contains a <TBODY> which contains one or more
<TR>'s, which in turn contains <TH>'s and/or <TD>'s.

so when the browser stumbles upon a <TABLE> followed by
a <TD>, it can easily figure the rest out itself, based on the
DTD.

> And obviously that's what some browsers do. This way also lies
> madness.

some?  I suspect 99.99% of browsers used for serious surfing
do things like this -- and *much* more...  (if you want to see
some real madness, go browse the original navigator sources ;-)

> I always find it strange that otherwise intelligent people don't want
> their compilers to work like this, but do want their browsers to.

well, if the early web had been restricted to people who can
read DTD's and map cryptic error messages to typos in their
sources, I doubt we'd have the rich web we have today...

    nsgmls:mypage.htm:38:3:E: start tag for "TR" omitted,
    but its declaration does not permit this

    huh?  there's no frigging TR in there?  and it looks good
    in my browser, so who cares about the "declaration"...

(the real design mistake is of course to do the fixes in the
browser, instead of doing them in the authoring system or
the server, but that's another story).

</F>




From johann at physics.berkeley.edu  Thu Apr  6 16:17:24 2000
From: johann at physics.berkeley.edu (Johann Hibschman)
Date: 06 Apr 2000 13:17:24 -0700
Subject: Beginning to be disapointed
References: <Aw4H4.8078$pR.68745@carnaval.risq.qc.ca> <mt3dozt6sh.fsf@astron.berkeley.edu>
Message-ID: <mtu2hfrnl7.fsf@astron.berkeley.edu>

[Yes, I know.  I'm fouling up the reference chain.  But there's weirdness
afoot, and I can't do any better.]

Hugues Demers writes:

>> 2. how are you reading them in?

> I use a module written by M. Miller called TableIO to read my data because
> it's fast. It read text file into Numeric array

Aha.

This might be it.  IIRC, some of the innards of the reference counting
changed at some point in the NumPy distribution.  If you're using a
different version of NumPy than TableIO was designed for, there might
be a memory allocation problem.

Double-check that, since (to me) the errors you get sound like what
happens to me when I try to free something that I shouldn't.  It runs
for a bit, after the mistake, then dies mysteriously.

I bet that's it.  I bet TableIO fails to set a bit like OWN_DATA or
some such on the inputted arrays, and then something bad happens.
I don't know for sure, though.

As an experiment, try using Konrad Hinsen's ArrayIO module to read the
arrays instead.  That's pure python, so it will be a bit slow, but it
might help you track down the problem. (It's what I use, and I don't
find it to be too terribly slow.)

Good luck,

--Johann

-- 
Johann Hibschman                           johann at physics.berkeley.edu


From effbot at telia.com  Tue Apr 11 10:30:18 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Tue, 11 Apr 2000 14:30:18 GMT
Subject: Huh? func_defaults, default values in function calls
References: <8cv5qj$35p$1@zinc.intern.netconnect.no>
Message-ID: <_FGI4.1193$Za1.20497@newsc.telia.net>

Roger Baklund <roger at netconnect.no> wrote:
> I don't get it. Why isn't the dictionary reset to {'x':'a'} when i call
> the function the last time?
>
> >>> a.func_defaults
> ('', '', {'x': 'a', 'y': 'c'})
>
> Why is not {} the default for the z parameter?

read the fine manual:

    http://www.python.org/doc/current/ref/function.html

    Default parameter values are evaluated when the function
    definition is executed. This means that the expression is
    evaluated once, when the function is defined, and that
    that same "pre-computed" value is used for each call.

(reading the entire page might help you understand why this
is by design, not by accident)

</F>




From gmcm at hypernet.com  Wed Apr 26 08:29:03 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Wed, 26 Apr 2000 08:29:03 -0400
Subject: Metaclasses? 
In-Reply-To: <200004261201.IAA05678@zippy.init.com>
Message-ID: <1255370352-61030244@hypernet.com>

Donald Beaudry wrote:

> quinn at hedono.ugcs.caltech.edu (Quinn Dunkan) wrote,
> > Anyway, even if you could figure out the ambiguity problem, I
> > don't think metaclasses would help you, bytecodehacks might.
> 
> *If* bytecodehacks could help, metaclasses could be used to make
> calling them neat and clean.  The idea is that during class
> instantiation the metaclass would, using bytecodehacks, rewrite
> all of the class's methods.  This might confuse the poor guy who
> comes along and tries to derive a new class from one of these
> 'self less'.

And then we could argue about whether being 'self less' 
makes them "altrustic" <wink> [cf long-lived, contentless 
splinter of the Economics of Open Source thread].

- Gordon



From snorri at ensim.com  Sun Apr  2 03:11:54 2000
From: snorri at ensim.com (Snorri Gylfason)
Date: Sat, 01 Apr 2000 23:11:54 -0800
Subject: Python or OS forking/threading problem?
References: <slrn8doffm.4jk.nascheme@cranky.arctrix.com> <001301bf9616$c2a4f260$752d153f@tim> <20000325131427.A13074@acs.ucalgary.ca> <87k8iqu2mt.fsf@deneb.cygnus.argh.org>
Message-ID: <38E6F2BA.E66CAC90@ensim.com>

Thanks everybody for looking into this problem. I was
working with Naris on this (Naris posted the original 
question).


I have a few comments to your discussion that might be helpful.

1) According to Posix when doing fork inside a thread only
   that thread is cloned in the new process. 

   a) Thats exactly why in the code below none of
      the threads are cloned (only the main thread).

   b) This is known problem for the child process. 
      If the cloned thread needs to acquire a lock that
      at the time of the fork some other thread had, it
      will deadlock (since in the child process there
      is no thread to release that lock).
      pthread_atfork is there to help the child process
      to handle this problem.

2) The strange thing in our case is that it is the parent
   thread that starts behaving badly. And as I will soon
   point out, it depends on what the child process is doing.

3) We ran into this problem because we were using popen in
   a thread. Popen is basically only doing fork and exec plus
   some filehandling stuff. The popen module is implemented
   in python. I replaced that module with one in C and the
   problem went away  (so I have a workaround:)
   I did some more experiments and it seems that if the child
   process leaves an atomic operation (enters Python again) 
   our problem may occur.
   I don't know the implementation of Python but everything
   I have seen indicates that the child process is still sharing 
   some lock with the parent process. 

Hope this will help

Snorri

Florian Weimer wrote:
> 
> "Neil Schemenauer" <nascheme at enme.ucalgary.ca> writes:
> 
> > Not for my problem on Linux.  My code didn't call sleep().  It
> > may be a bug with the pthreads in libc6 for Linux.  I can't
> > reproduce it with C code though.
> 
> I think I can explain what happens.  Look at the following code:
> 
> #include <stdio.h>
> #include <string.h>
> 
> #include <unistd.h>
> #include <pthread.h>
> 
> void *
> thread(void * v)
> {
>   for (;;) {
>     char buf[40];
>     int len = sprintf(buf, "%lu\n", (unsigned long) getpid());
>     write(1, buf, len);
>     sleep(1);
>   }
> }
> 
> main()
> {
>   pthread_t t;
>   pthread_create (&t, NULL, &thread, NULL);
>   fork();
>   sleep(3600);
> }
> 
> This clearly shows that the thread is not duplicated on fork().
> 
> Now look at the Python source code:
> 
> static PyObject *
> posix_fork(self, args)
>         PyObject *self;
>         PyObject *args;
> {
>         int pid;
>         if (!PyArg_ParseTuple(args, ":fork"))
>                 return NULL;
>         pid = fork();
>         if (pid == -1)
>                 return posix_error();
>         PyOS_AfterFork();
>         return PyInt_FromLong((long)pid);
> }
> 
> void
> PyOS_AfterFork()
> {
> #ifdef WITH_THREAD
>         main_thread = PyThread_get_thread_ident();
>         main_pid = getpid();
> #endif
> }
> 
> long PyThread_get_thread_ident _P0()
> {
>         volatile pthread_t threadid;
>         if (!initialized)
>                 PyThread_init_thread();
>         /* Jump through some hoops for Alpha OSF/1 */
>         threadid = pthread_self();
>         return (long) *(long *) &threadid;
> }
> 
> In the child process, all threads have disappeared, but the code doesn't
> seem to be prepared to handle this.


From yawyi at gmx.de  Thu Apr 27 09:04:43 2000
From: yawyi at gmx.de (Johannes Stezenbach)
Date: Thu, 27 Apr 2000 15:04:43 +0200
Subject: Best database to use with Python
References: <3906ea27.698463777@news.online.no> <8e9334$qba$1@news.tht.net>
Message-ID: <39084a70$0$22983@personalnews.de.uu.net>

D'Arcy J.M. Cain <darcy at vex.net> wrote:
>?5?HH575-UAZWKVVP-7H2H48V3 <thomas at cintra.no> wrote:
>> What database-system is easiest, fastest etc. to use with Python? I`ve
>> looked at MySQL, but it seems a little overkill just to get a little
>> more functionality than Berkley DB-functionality. What about Postgres?
>> mSQL? 
>
>> Postgres comes with my Linux distro and seem smaller than MySQL. How
>
>Actually I think you will find PostgreSQL larger than MySQL.
>
>> does it compare to MySQL when it comes to administration and speed,
>> scalability etc. ?
>
>I don't use MySQL as I find that PostgreSQL has more features and works
>fast enough for my needs.  Like most RDBMSs in its class the expensive
>part is updates and inserts but the retrievals are quite snappy.  The
>one time I tried to administer a MySQL system I was a little horrified.
>PostgreSQL was easier to understand.  That could just be me, of course.

The one fundamential difference between MySQL and PostgreSQL (apart
from the license) is that PostgreSQL has a focus on preserving
data integrity and dealing with concurrency issues (using transactions,
triggers, etc.) while MySQL omits these features in favor of raw speed.
The MySQL folks claim to be 10 times faster for UPDATEs.

Johannes



From nobody at nowhere.nohow  Sat Apr  8 12:16:48 2000
From: nobody at nowhere.nohow (Grant Edwards)
Date: Sat, 08 Apr 2000 16:16:48 GMT
Subject: read _and_ write from/to commandline program
References: <8cnjnt$k7o$1@nnrp1.deja.com>
Message-ID: <QXIH4.1023$9L.31066@ptah.visi.com>

On Sat, 08 Apr 2000 15:37:34 GMT, lamzak8642 at my-deja.com <lamzak8642 at my-deja.com> wrote:

>Is it possible to read from _and_ write to a commandline program (Linux
>+ Python)? Apparantly I can't do this with os.popen() because it is
>unidirectional.

yes.  The popen2 module will give you file objects for the
child process stdin, stdout, and stderr.

-- 
Grant Edwards                   grante             Yow!  Mr and Mrs PED, can
                                  at               I borrow 26.7
                               visi.com            


From m.faassen at vet.uu.nl  Wed Apr 12 06:21:50 2000
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 12 Apr 2000 10:21:50 GMT
Subject: Python/Perl Popularity (Re: A Mountain of Perl...)
References: <200004102111.HAA03380@envy.fulcrum.com.au> <8d0pcq$7ti$1@nnrp1.deja.com>
Message-ID: <8d1inu$nsq$1@newshost.accu.uu.nl>

tom__98 at my-deja.com wrote:
> In article <200004102111.HAA03380 at envy.fulcrum.com.au>,
>   Richard.Jones at fulcrum.com.au wrote:
>> Tom you little troll-meister. Half those arguments
>> you cite are bollocks if you _really_ knew and
>> "greatly prefer" Python. Try again.

> Maybe you can contribute more constructively:

I'm not him, but I'll be constructive. :)

>  -- How can I browse the Python documentation without firing
>     up a web browser?

A text mode web browser can be useful for this. I've found w3m on
Linux to be very nice. I can just do:

w3m file:///usr/doc/python-docs-1.5.2/html/lib/modindex.html

And the documentation page is there instantly.

>  How do I get information on a single Python
>     function or class quickly?

The docstring situation in the library is pretty good:

>>> import string
>>> dir(string)
>>> print string.split.__doc__

>>> import BaseHTTPServer
>>> print BaseHTTPServer.SocketServer.__doc__

>  Is there any equivalent of
>     "perldoc -f split" or "perldoc Net::POP3"?  What do you
>     do in that situation?

Perhaps there are more easy tools to query docstrings; I'm sure there
should be done out there. Since I'm a web developer I usually have
a browser going somewhere, so I use the HTML documentation. :)

>  -- Is there a Python equivalent of CPAN and the Perl CPAN
>     module?  The closest I know of is Parnassus.  But Parnassus
>     is merely a collection of links, not an archive, and it
>     doesn't have any facilities (AFAIK) for automatic installation.

Parnassus is way cool, but it can't do this yet. There have been talks
in the past between the Distutils people and the Parnassus folks (um,
individual). Once distutils matures it should provide for automatic
installation, and it'll probably be integrated with Parnassus.

>  -- Several extensions I have wanted to install have required access
>     to the full Python source tree, not just the header files and
>     the library.  Is that going to change?  Otherwise, how can
>     I install Python extensions that want access to the source
>     tree on, say, a standard RedHat system?  Having the standard
>     RPM-based installation in parallel with a /usr/local installation
>     from source is confusing, and removing the RPM-based installation
>     risks breaking things.

Hm, haven't seen these yet myself. That does sound tricky. We're just
all rooting for distutils to save the day. :)

Regards,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From darrell at dorb.com  Sun Apr  2 15:30:35 2000
From: darrell at dorb.com (Darrell)
Date: Sun, 2 Apr 2000 12:30:35 -0700
Subject: Debugging confusion -- too many stacks!
References: <000a01bf9c6b$3f80c2a0$752d153f@tim> <38E739A1.45F4B199@tismer.com>
Message-ID: <030301bf9cd9$eb6786d0$6401a8c0@dorb>

From: "Christian Tismer"
> The following watches are always active when I'm debuging:
> (are there similar ways to do this with PDB?)
>
> (char*)(f->f_code->co_name)+20       # current function name
> (char*)(f->f_code->co_filename)+20   # current file name
> f->f_lineno                          # line number
>
> In order to see which frame is calling you, do the same, but
> replace "f" by "f->f_back".

pdb works pretty good.
I use it all the time. You can get the call stack or what ever you want to
look at.

>
> What I really wished were a way to extend a given debugger
> with your own support code, to make PyObjects more visible.
>
On a really bad day you could use both pdb and a 'C' level debugger.
If you have a lot of bad days, I could imagine there must be ways to enhance
the interplay of these two debuggers.

--Darrell





From dan at cgsoftware.com  Tue Apr  4 23:38:11 2000
From: dan at cgsoftware.com (Daniel Berlin+list.python)
Date: 04 Apr 2000 23:38:11 -0400
Subject: What is python's language level?
In-Reply-To: Ivan Van Laningham's message of "Tue, 04 Apr 2000 19:54:28 -0600"
References: <38EA1B8F.A797A130@lmco.com> <38E9DDD3.3313FF6@bioreason.com> <slrn8ekddq.gm3.zawrotny@gecko.sb.fsu.edu> <roy-D857DA.21212304042000@netnews.nyu.edu> <38EA9CD4.3473B910@callware.com>
Message-ID: <bt3ptdy4.fsf@dan.resnet.rochester.edu>

>>>>> "Ivan" == Ivan Van Laningham <ivanlan at callware.com> writes:


    Ivan> It's most certainly Turing-complete.  It ain't what you call
    Ivan> convenient or readable, but once you understand what you're
    Ivan> doing you can write anything you want in it.  Including
    Ivan> calculators and prime-number sieves.


Reminds me of a neat little raytracer written in postscript someone
posted many years ago.

I was just playing with it a few minutes ago.
About 7k, raytraces 5 spheres by default.
you have to set the max ray depth to about 3 or 4 or it'll take
forever.
Ghostscript can use it to raytrace a 128x128 image in about 70
seconds.
sphere col get aload vec copy /VU exch store'ly y'rs,
Dan



From hnowak at cuci.nl  Sun Apr 30 15:21:40 2000
From: hnowak at cuci.nl (Hans Nowak)
Date: Sun, 30 Apr 2000 21:21:40 +0200
Subject: Python & clipboard
Message-ID: <200004301922.e3UJMtG25451@hera.cuci.nl>

Howdy y'all,

I am trying to find a function which gets clipboard data (on Win98). I have 
been looking at the win32clipboard module, but it seems that the best I can 
get is a handle to an object containing clipboard data. I also looked into 
Tkinter, and I found clipboard_append and clipboard_clear methods, but no 
such thing as clipboard_get. 

Does anyone know how to do this? 

Blessed be,

--Hans Nowak (zephyrfalcon at hvision.nl)
Homepage: http://www.hvision.nl/~ivnowa/newsite/
Python questions? See http://tor.dhs.org/~zephyrfalcon/snippets
You call me a masterless man. You are wrong. I am my own master.



From fredrik at pythonware.com  Wed Apr  5 06:40:49 2000
From: fredrik at pythonware.com (fredrik at pythonware.com)
Date: Wed, 05 Apr 2000 10:40:49 GMT
Subject: Starving for an Advanced Python Book
References: <20000405044154.10369.qmail@web2103.mail.yahoo.com>
Message-ID: <8cf57f$e96$1@nnrp1.deja.com>

"lewst" wrote:
> Those remarks are misleading and should probably be removed.  I won't
> disclose my source, but I'll you bet dollars to donuts that the 2nd
> edition of PP will be out this summer -- much closer to the release of
> Python 1.6 than P3K.

heh.  you really are Tim O'Reilly, aren't you?

(first you dis my eBook using the usual publisher
arguments, and now you have inside information on
when PP2E will be out ;-)

</F>


Sent via Deja.com http://www.deja.com/
Before you buy.


From mwh21 at cam.ac.uk  Mon Apr 17 11:26:39 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 17 Apr 2000 16:26:39 +0100
Subject: Another question
References: <dVFK4.41$vr2.5817@news.uswest.net>
Message-ID: <m3ya6cd9zk.fsf@atrus.jesus.cam.ac.uk>

"Jeff Massung" <jmassung at magpiesystems.com> writes:

> Why would I not do "from <module> import *" with every module I load?

Because being able to predict what name lookups resolve to is more
important to some people than micro-optimisation?

Cheers,
M.

-- 
  ... but I guess there are some things that are so gross you just have
  to forget,  or it'll destroy something within you.  perl is the first
  such thing I have known.                 -- Erik Naggum, comp.lang.lisp


From rcc at nospamthanks_jennic.com  Thu Apr 13 13:25:28 2000
From: rcc at nospamthanks_jennic.com (Robert Cragie)
Date: Thu, 13 Apr 2000 17:25:28 GMT
Subject: List mapping?
References: <Pine.LNX.3.96.1000413174501.3264B-100000@valdez.netg.se>
Message-ID: <newscache$autysf$ye8$1@jenpc07.jennic.co.uk>

> Howcome you use reduce at all?
>
> Works fine with just doing
>
> tuplist = map(lambda x: (mylist[x], mylist[x+1]), range(0, len(mylist),
2))

Yep, that's neater. My excuse is that I've only been learning Python for a
couple of weeks...

Robert Cragie




From gmcm at hypernet.com  Tue Apr  4 21:48:12 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Tue, 4 Apr 2000 21:48:12 -0400
Subject: Two questions
In-Reply-To: <38EA6E87.40EDB542@callware.com>
Message-ID: <1257223201-7158461@hypernet.com>

Ivan Van Laningham wrote:
> 
> Bjorn Pettersen wrote:
> > 
> [bobbit]
> > 
> > If you want to package up your program into an exe and execute it,
> > Installer from Gordon's starship pages can do this
> > (http://starship.python.net/crew/gmcm/distribute.html).  It's actually
> > _much_ easier than it looks ;-)
> > 

It's not widely known, but it blows chunks if you have Mayan 
glyphs on your machine. That's a feature.
 
> <g>  This sounds like the right course to me.  I'd like to force
> customers to install Python on their machines, but my boss doesn't think
> that's all that great an idea (dunno why).  My only concern is
> size--just how big are the executables produced by Gordon's
> flummery^H^H^H^H^H^H^H^Hmagic?

If you get creative (about a dozen lines of code) you can install 
a fully functional Python in 2 files. You can put Python and 
IDLE on a floppy with room left over.
 
[mxODBC]
> Well, that's two recommendations.  And judging from previous Marc-Andre
> packages I've used, it'll be a snap.

At the moment, you'll have a bit of trouble with my stuff and 
the fact that Marc-Andre puts his .pyds inside the packages. 
There are workarounds, and I have a patch from Thomas Heller 
which I haven't finished reconciling to some of my other 
changes.
 
> Steve Holden, in another post, said that Tkinter should work fine along
> with mxODBC.  That's a relief, but does anyone have direct personal
> experience with doing so?
> 
> <floccinausinihilipilification>-ly y'rs,
> Ivan

Are you describing your hairline, your nervous twitch, or the 
effect of the latter on your former?

- Gordon



From tim_one at email.msn.com  Mon Apr  3 01:30:25 2000
From: tim_one at email.msn.com (Tim Peters)
Date: Mon, 3 Apr 2000 01:30:25 -0400
Subject: Who is this years winner?
In-Reply-To: <m12bpyc-000CnCC@artcom0.artcom-gmbh.de>
Message-ID: <000101bf9d2d$b724e1c0$e02d153f@tim>

[Peter Funk, seeking justice]
> Today on April 2nd I began to wonder whether the PSA Steering
> Committee has already decided on the winner of the third Pythonic
> Award.

Yes, they have.  It's ME!!!  And nobody can prove it's not.

> Fortunately www.python.org is back online now but unfortunately
> the page
> 	<http://www.python.org/PSAward.html>
> seems to show up a file from two year old backup, and so does the
> Shrine of the of the winner of the first Python Award at
> 	<http://www.python.org/tim_one/>

Yes too, that's from the First Pythonic Award.  Which remains the most
recent.  On 1-Apr-99, I unilaterally changed MY title to The Only Living
Pythonic Award Winner, and announced this would be MY honorific for life --
by any means necessary.  Since then, potential winners other than ME have,
umm, gracefully withdrawn.  I'm sure it's just coincidence that the PSA
Steering Committee has never been heard from again.

> May I kindly ask some member of the famous Steering Committee to publish
> this years decision here?

Sure!  Ask all you like <snicker>.

don't-ever-mess-with-an-insecure-bot-ly y'rs  - tim





From gidi_avrahami at my-deja.com  Mon Apr  3 11:46:24 2000
From: gidi_avrahami at my-deja.com (gidi_avrahami at my-deja.com)
Date: Mon, 03 Apr 2000 15:46:24 GMT
Subject: errors in _del__ during python shutdown
Message-ID: <8caec9$6v9$1@nnrp1.deja.com>

I have a large Python system with wrappers to some SWIG'ed C code.
Some objects need to do some cleanup in their __del__ , e.g.:

from myBasics import Object
import myCToolkit
kit = myCToolkit

    class SomeObject(Object):
        def __init__(self, *args):
            Object.__init__(self, *args)
            self._vec = kit.float_array(6)
        def __del__(self):
            kit.float_destroy(self._vec)
            Object.__del__(self)

My problem is that when I shut python down (by typing EOF to the prompt), the
SomeObject.__del__ method gets called in a funny state, where 'kit' and even
'Object' are None!  It would appear that Python is "winding down" and starts
killing off modules, even though some objects that refer to these modules
are still alive.

Any ideas?


Sent via Deja.com http://www.deja.com/
Before you buy.


From lbear11 at my-deja.com  Sun Apr 16 22:37:44 2000
From: lbear11 at my-deja.com (lbear11 at my-deja.com)
Date: Mon, 17 Apr 2000 02:37:44 GMT
Subject: Tkinter Canvas -- help?
Message-ID: <8ddtdm$b2s$1@nnrp1.deja.com>

I'm just starting to learn Python/tkinter, and I have found that there
is virtually _no_ documentation for the tkinter canvas. Could anybody
help? I'm specifically looking for stuff like creating an image by a
grid in canvas on "mouseDown", and how to read the items in the canvas,
and write them to a text file, in their proper positions. I would
_really_ appreciate it if anybody could help me with this.


P.S. The only tkinter canvas documentation that I have is the
"Introduction to Tkinter" from pythonware.com

Thanks,
jkior


Sent via Deja.com http://www.deja.com/
Before you buy.


From cut_me_out at hotmail.com  Wed Apr 19 08:30:23 2000
From: cut_me_out at hotmail.com (Alex)
Date: 19 Apr 2000 08:30:23 -0400
Subject: Very Horrible Question About Goto's
References: <6325E90C8B6ED21189080001FA7E2ADA0D4051@IBA1>
Message-ID: <etdwvlu6zog.fsf@w20-575-109.mit.edu>

Probably not, is my guess.  It would help if you posted some examples of
your language, though. 

Alex.


From claird at starbase.neosoft.com  Sat Apr 22 00:43:48 2000
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 21 Apr 2000 23:43:48 -0500
Subject: The Simple Economics of Open Source
References: <14591.15630.192472.440375@goon.cnri.reston.va.us> <3900611F.3B525BDA@cfdrc.com> <FtDG4z.774@world.std.com> <slrn8g26tu.p7o.neelk@brick.cswv.com>
Message-ID: <BA667459FFE26B95.CB43EBDC57FAA144.95EDC7F1E07A7B76@lp.airnews.net>

In article <slrn8g26tu.p7o.neelk at brick.cswv.com>,
Neel Krishnaswami <neelk at alum.mit.edu> wrote:
			.
			.
			.
>One of the interesting things about this thread is that it has given
>me a bit of insight into what economics looks like to people who
>haven't internalized its assumptions. I'm afraid that I'm tainted
			.
			.
			.
>the effort. Most people don't even bother digging up evidence to
>support their prejudices, let alone try to come to a careful decision.
>And searching the Census and BLS websites is *easier* than hacking on
>the Python core.
My wife says it reinforces her prejudices about
economists (or "dart-throwers", as she calls us)
that we would think the "Census and BLS websites"
have anything to do with "a careful decision."
			.
			.
			.
>snobbery. I read an article last year where two economists developed a
>model predicting that a disproportionate fraction of groundbreaking
>academic papers should be wrong. Basically, they assumed that since
>funding stems from status, and status stems from writing ground
>breaking papers, there's an incentive to draw flashy, counterintuitive
>conclusions rather than confirming prior work. (They did observe that
>their work implied their paper was probably wrong, since it was
>advancing a flashy, counterintuitive proposition. :)
>
>
>Neel

This has been part of academic folklore for a long,
LONG time--oh!; that means I should describe the
"flashy, counterintuitive, but already widely-known-
without-being-well-marketed" category.

Wrestling this back toward the neighborhood of Python
looks like a challenge; all I can think to say is 
that economics has no monopoly on generation of a
comfortable explanation for open-source community
process.  Anthropology, sociology, rhetoric, theology,
architecture, ... all recognize familiar elements in
what we do.
-- 

Cameron Laird <claird at NeoSoft.com>
Business:  http://www.Phaseit.net
Personal:  http://starbase.neosoft.com/~claird/home.html


From invalid.address at 127.0.0.1  Wed Apr 19 14:31:52 2000
From: invalid.address at 127.0.0.1 (David)
Date: Wed, 19 Apr 2000 18:31:52 GMT
Subject: EricH-Copyright Messages
References: <38fc9694@wwwproxy3.westgroup.com> <38FCA69D.85CC27DA@heathers.stdio.com>
Message-ID: <38fdfa72.42533156@news.telus.net>

If an ISP copyrights postings that its users make, then isn't it also
completely liable for those postings?

That seems awful dangerous.

Also, is it even possible for them to copyright one's own postings?  AFAIK,
you possess copyright the instance you write something, and I find it
difficult to believe that ownership can be implicitly and irrevocably
transfered.

Care to enlighten me?


On Tue, 18 Apr 2000 14:17:01 -0400, "Eric H." <burnout at heathers.stdio.com>
wrote:
<snippage>
Parts of this message, including the email address, ? 2000-Eric Haddix. 
Use implies agreement to the terms and conditions made available on my 
homepage. Keep the Internet free, refuse to support the providers that
copyright YOUR email and messages. All rights reserved. U.C.C. 1-207

visit my homepage @ http://www.bounce.to/EyesCremeProductions



From effbot at telia.com  Tue Apr 11 15:06:07 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Tue, 11 Apr 2000 19:06:07 GMT
Subject: AF_INET error
References: <8cvero$g8k$1@porthos.nl.uu.net> <cHHI4.1202$Za1.20911@newsc.telia.net> <8cvnve$l40$1@porthos.nl.uu.net>
Message-ID: <zIKI4.1250$Za1.21976@newsc.telia.net>

Soundwave <maybe at lucky.com> wrote:
> import socket
> #create an INET, STREAMing socket
> s = socket.socket(
>     socket.AF_INET, socket.SOCK_STREAM)
> #now connect to the web server on port 80
> # - the normal http port
> s.connect("www.mcmillan-inc.com", 80)
>
> when I run this code, I get the following output
>
> Traceback (innermost last):
>   File "C:\PROGRA~1\PYTHON\TOOLS\IDLE\ScriptBinding.py", line 131, in
> run_module_event
>     execfile(filename, mod.__dict__)
>   File "C:/Program Files/Python/socket.py", line 3, in ?
>     s = socket.socket(
> AttributeError: AF_INET

hmm.  this example works just fine on my windows 95 box,
running it from the command line prompt.

maybe IDLE is interfering with your code in some strange
way?  here are a few things you might wish to try:

-- restart IDLE, and try running the script again

-- try running it from an MS-DOS window

-- try adding the following two statements just after
   the import statement:

        print socket
        print dir(socket)

this should print something like:

<module 'socket' from 'C:\program~1\python\lib\plat-win\socket.pyc'>
>>> print dir(socket)
['AF_INET', 'INADDR_ALLHOSTS_GROUP', 'INADDR_ANY', ...lots of
other stuff... 'htons', 'ntohl', 'ntohs', 'socket']

</F>




From PClaerhout at CREO.BE  Tue Apr 11 05:36:09 2000
From: PClaerhout at CREO.BE (Pieter Claerhout)
Date: Tue, 11 Apr 2000 11:36:09 +0200
Subject: Getting an overview if all available driveletters on NT
Message-ID: <2B1262E83448D211AE4B00A0C9D61B03BA7294@MSGEURO1>

Hello,

is there a module in Python, which can list all available driveletters, together with
their label. I'm using Windows NT? I already thought of iterating over all possible
driveletters, and see if the path exists or not, but I also want to separate the
network volumes from the other volumes. Any ideas?

Kind regards.


Pieter

[ Pieter Claerhout :: CreoScitex Response Centre :: Applications support ]
[ Pieter_Claerhout at creoscitex.com    ::    http://www.creoscitex.com/ ]
[ CreoScitex Europe  ::   Excelsiorlaan 21   ::  1930 Zaventem  ::   Belgium ]




From herzog at online.de  Mon Apr 17 21:19:01 2000
From: herzog at online.de (Bernhard Herzog)
Date: 18 Apr 2000 03:19:01 +0200
Subject: Another question
References: <dVFK4.41$vr2.5817@news.uswest.net> <38FB2EEB.A32AF88C@bam.com> <38FB339F.604364AD@bam.com> <200004171605.MAA04929@zippy.init.com> <38FB7610.80EAF965@ichips.intel.com> <200004172120.RAA11102@zippy.init.com> <jWNK4.432$vr2.32822@news.uswest.net> <m3ln2cckg6.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <m33dok8auy.fsf@greebo.nodomain.de>

Michael Hudson <mwh21 at cam.ac.uk> writes:

> For that last dreg of performance you might try:
> 
> import random
> def do_it(randint=random.randint,xrange=xrange):
>     r = [0,] * 100000
>     for i in xrange(100000):
>        r[i] = randint(1,10)
> do_it()
> 
> But really, this is getting silly; if performance is this much of an
> obsession you aren't going to have that much fun with Python.

A much better optimization would be to use choice with a constant list
instead of randint. The implementation of choice is much more efficient
that randints:

def do_it(length):
    from random import choice
    r = [0,] * length
    ints = range(1, 11)
    for i in xrange(length):
        r[i] = choice(ints)

even better would be to inline choice or a custom version of randint.

Of course, this takes us away from the question originally asked...

-- 
Bernhard Herzog   | Sketch, a drawing program for Unix
herzog at online.de  | http://sketch.sourceforge.net/


From mwh21 at cam.ac.uk  Mon Apr  3 12:25:47 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 03 Apr 2000 17:25:47 +0100
Subject: errors in _del__ during python shutdown
References: <8caec9$6v9$1@nnrp1.deja.com>
Message-ID: <m3ya6vywvo.fsf@atrus.jesus.cam.ac.uk>

gidi_avrahami at my-deja.com writes:

> I have a large Python system with wrappers to some SWIG'ed C code.
> Some objects need to do some cleanup in their __del__ , e.g.:
> 
> from myBasics import Object
> import myCToolkit
> kit = myCToolkit
> 
>     class SomeObject(Object):
>         def __init__(self, *args):
>             Object.__init__(self, *args)
>             self._vec = kit.float_array(6)
>         def __del__(self):
>             kit.float_destroy(self._vec)
>             Object.__del__(self)
> 
> My problem is that when I shut python down (by typing EOF to the prompt), the
> SomeObject.__del__ method gets called in a funny state, where 'kit' and even
> 'Object' are None!  It would appear that Python is "winding down" and starts
> killing off modules, even though some objects that refer to these modules
> are still alive.
> 
> Any ideas?

Do this:

        def __del__(self,kit=kit,Object=Object):
            kit.float_destroy(self._vec)
            Object.__del__(self)

must be a FAQ, but I can't be bothered to track the entry down just
now, sorry.

Cheers,
M.

-- 
well, take it from an old hand: the only reason it would be easier
to program in C is that you can't easily express complex  problems
in C, so you don't.                 -- Erik Naggum, comp.lang.lisp


From rupole at compaq.net  Wed Apr 19 19:36:41 2000
From: rupole at compaq.net (Roger Upole)
Date: Wed, 19 Apr 2000 23:36:41 GMT
Subject: Where do I send a contribution for the Starship recovery ?
Message-ID: <kXqL4.142058$ra.3216975@news.easynews.com>

I personally would consider it money well spent.
(unlike most of the other things I spend money on)
                    Roger Upole




From ns56645 at swt.edu  Sat Apr 29 13:24:00 2000
From: ns56645 at swt.edu (ns56645)
Date: Sat, 29 Apr 2000 12:24:00 -0500
Subject: here is the emphasization of teh error
Message-ID: <390B1AB0.330C904B@swt.edu>

##############    List of other canvases in frame 1
###########################################


     # Make a list of days of the week
       def drawCanvas(self):
        CanvasNameList=['MonCanvas', 'TueCanvas', 'WedCanvas',
'ThuCanvas', 'FriCanvas']
        xpos=68
        ypos=15
        xdiff=54
        for canvas in CanvasNameList:
                self.canvas=Pmw.ScrolledCanvas(frame1,
                borderframe = 1,
                labelpos = 'wn',
                usehullsize = 5,
                hull_width = 800,
                hull_height = 80,
                )
                i=1
                self.canvas.grid(frame1,row=1, column=i)
                i=i+1
                self.canvas.resizescrollregion()
                self.canvas.configure(hscrollmode = 'dynamic')
                self.canvas.configure(vscrollmode = 'static')


#########################################################################################



-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20000429/0241d9c8/attachment.html>

From rwiserREMOVEthis at mindspring.com  Tue Apr 11 13:48:11 2000
From: rwiserREMOVEthis at mindspring.com (Randy Wiser)
Date: Tue, 11 Apr 2000 17:48:11 GMT
Subject: Calling os.waitpid() From Secondary Thread Under Linux 2.2.12-20?
Message-ID: <38f363e5.2931715373@news>

(Apologies if this is a duplicate, it's been awhile since I've posted
anything :-)

I've written a small python program that spawns one child process from
its main thread, and then creates a second thread.  The sole purpose
of the second thread (at this point) is to call os.waitpid(processID,
0) to wait for the child that the main thread created.  This works
fine under Solaris when I call os.waitpid() from the second thread. 

But, under Linux 2.2.12-20 I get the following OSError exception:

[Errno 10] No child processes

However, if (under Linux still) I call waitpid() from the main thread
(as a test), the call does indeed wait until the child process
terminates.

After scouring some of the Linux newsgroups, I've gathered that
threads spawned under Linux may have SIGCHLD signals (and other
signals) blocked, and that this _may_ be the reason I get the error,
even though the processID I pass is for a valid process.

Unfortunately there seems to be no easy way to set the signal mask for
the second thread.  Has anyone else run into this situation, and if so
can you advise how you solved it?

Thanks in advance,

- Randy


From scherbi at bam.com  Mon Apr 17 11:56:43 2000
From: scherbi at bam.com (Bill Scherer)
Date: Mon, 17 Apr 2000 11:56:43 -0400
Subject: How do I make a constant global variable in Python Please
References: <vCj7OA2LfMfb=znR7yTHOvmiZAkd@4ax.com>
Message-ID: <38FB343B.90F0133E@bam.com>

declare the variable as global inside the function you wish to use the
global variable, else you just create a new local variable of the same
name. eg:

foo = 1

def spam():
    #use the global foo
    global foo
    ...

def eggs():
    #this function has it's own foo
    foo = 2
    ...

OK?

-Bill

cmfinlay at magnet.com.au wrote:

> How do I make a constant global variable in Python Please.
> or just a constant.
> or just a global.
>
> I bought a Python Book. however I can't find this info in it.
>
> Please e-mail me at cmfinlay at magnet.com.au
> --
> http://www.python.org/mailman/listinfo/python-list






From gmcm at hypernet.com  Fri Apr  7 14:05:44 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Fri, 7 Apr 2000 14:05:44 -0400
Subject: Embedding Python: How To Hardwire Module Search Path
In-Reply-To: <38EDDB40.FC715534@gmx.net>
Message-ID: <1256991739-21081117@hypernet.com>

Robert Kiendl wrote:

> 
> I've embedded python into a C++ - windows app by a
> PyImport_AppendInittab() +  Py_Initialize() method.
> 
> Py_Initialize() calculates a module search path from the
> windows-registry settings.
> 
> I want to prevent Python from doing that and want to hardwire the search
> path to a app. relative path ("<app>/pythonmod") which points to a
> minimal python bunch to avoid the need for installing python.
> 
> Does anybody know what to do prior to Py_Initialize(). Or am I
> completely wrong?
> PySys_SetPath() crashes when called before Py_Initialize(). 

I suggest you just rewrite getpath.c to do what you want. It 
runs during Py_Initialize. Otherwise, you can undo what 
getpath.c does after Py_Initialize, but there's a hole there.

- Gordon



From emile at fenx.com  Thu Apr 13 08:53:56 2000
From: emile at fenx.com (Emile van Sebille)
Date: Thu, 13 Apr 2000 05:53:56 -0700
Subject: 1st non-trivial program - gentle criticism requested
References: <1e90r16.1sd2ts41e7ur6kN@gershwin.bio.ic.ac.uk>
Message-ID: <005701bfa547$561d8720$01ffffc0@worldnet.att.net>

Paul,

Just a few observations I have along the lines of
'oh!, you mean this' as I read your code.


>>> a = [0,0,0,0,0,0]
>>> for i in xrange(10000):
 j = int(whrandom.choice('012345'))
 a[j]=a[j]+1


>>> print max(a)
1719
>>> a.index(max(a))
3
>>>

HTH,

Emile van Sebille
emile at fenx.com
-------------------


----- Original Message -----
From: Paul-Michael Agapow <p.agapow at ic.ac.uk>
Newsgroups: comp.lang.python
To: <python-list at python.org>
Sent: Thursday, April 13, 2000 1:10 AM
Subject: 1st non-trivial program - gentle criticism requested


>
> Hi y'all,
>
> I'm experimenting with Python for a number of purposes (an embedded
> language for simulations, a tool for exploring experimental data, a
good
> first programming language) and so I've cobbled together my first
> non-trivial program. With some trepidation, I post it here in the
hopes
> that experts will be able to coach me in "The Python Way". I realise
it
> looks like C++ (my usual langauge), but criticisms of style would be
> useful too.
>
> Basically, it's a recreation of the "Methinks it is a weasel"
simulation
> detailed by Richard Dawkins in "The Blind Watchmaker". A population of
> strings is compared vs. a target string and the "fittest" (the one
that
> most resembles the target) is used to form the next generation, with a
> small chance. So it's a simple genetic algorithm.
>
> thanks
>
> p-m
>
>
> --- program follows
<snipped>





From gmcm at hypernet.com  Fri Apr 21 17:05:37 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Fri, 21 Apr 2000 17:05:37 -0400
Subject: Pythonware Website???? Gone on vacation with Starship? ;-))
In-Reply-To: <8dqa6o$pjg$4@newshost.accu.uu.nl>
Message-ID: <1255771354-36910521@hypernet.com>

Martijn Faassen wrote:

> .... Also, since the PSU has access to
> Guido's time machine and didn't prevent this, the PSU must be in on this.
> Perhaps they're testing those samples of whitespace-eating nanoviruses that
> they got from that UFO crash near Roswell back in the 50s.

Oh you silly boy. They're not alien. They simply used the time 
machine to bring back samples from the future.

if-i-told-you-how-i-knew-i'd-have-to-kill-you-ly y'rs

- Gordon



From mrnatch at gateway.net  Tue Apr 18 06:00:39 2000
From: mrnatch at gateway.net (mrnatch)
Date: Tue, 18 Apr 2000 03:00:39 -0700
Subject: gaming and python
Message-ID: <000c01bfa91c$f421d6a0$0f8d0b3f@oemcomputer>

I told a friend of mine that I am learning to program using Python (I chose Python as my first language on the advice of a Llama) and he asked me an interesting question: Can Python be used for programming video games? And does it work with Windows API? Just curious.
Thanks, 
Josh (And friend Tony)
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20000418/f4b117f6/attachment.html>

From pecora at anvil.nrl.navy.mil  Sun Apr 30 10:17:30 2000
From: pecora at anvil.nrl.navy.mil (Louis M. Pecora)
Date: Sun, 30 Apr 2000 14:17:30 GMT
Subject: Reading in strings -> numbers ??
References: <390A2D23.7029E557@fenx.com>
Message-ID: <300420001017316294%pecora@anvil.nrl.navy.mil>

In article <390A2D23.7029E557 at fenx.com>, <emile at fenx.com> wrote:

> If you're trying to put the info on disk to read it
> in again later, look at eval(repr()), eg:
> 
> >>> a,b,c = eval(repr((1.2, -3.5, 10)))
> >>> a,b,c
> (1.2, -3.5, 10)

Problem is I'm often reading in data generated by another program or
another person, e.g. ANSII or binary float values in "table form".


From mwh21 at cam.ac.uk  Tue Apr  4 02:56:59 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 04 Apr 2000 07:56:59 +0100
Subject: Why should I switch to Python?
References: <000701bf9c5a$b806b420$752d153f@tim> <oqity066km.fsf@titan.progiciels-bpi.ca> <slrn8ej2gc.b3t.warlock@raven.myth>
Message-ID: <m3itxyz744.fsf@atrus.jesus.cam.ac.uk>

warlock at eskimo.com (Jim Richardson) writes:

> Out of curiousity, how does one (mathematically) prove that?
> (if this turns out to be one of those 12' blackboards full of greek
> letters and funny squiggles, Ok, I cry uncle...)

It's not all that hard.  Write the prime numbers as p1,p2,p3,etc.
Suppose that this series stops at pn.  Then let P=p1*p2*...*pn+1.
This number is not a multiple of any of the pm, so either it is prime,
or it has a prime factor that is not in the list p1,p2,...,pn.  Either
way, pn cannot have been the highest prime number.

It's also possible to show that if a and n have no common factor, then
there are an infinte number of primes in the artihmetic progression
a+dn as d=0,1,2,3,4,..., but the proof is, umm, somewhat harder (ie. I
don't know it; I think it involves slinging Riemann zeta functions).

Cheers,
M.

-- 
well, take it from an old hand: the only reason it would be easier
to program in C is that you can't easily express complex  problems
in C, so you don't.                 -- Erik Naggum, comp.lang.lisp


From alwagner at tcac.net  Wed Apr 12 11:13:27 2000
From: alwagner at tcac.net (Albert Wagner)
Date: Wed, 12 Apr 2000 10:13:27 -0500
Subject: confused about path
Message-ID: <38F49297.274ABBAE@tcac.net>

Assuming I am set up correctly, shouldn't I be able to invoke idle like
this:

$ python idle.py

from any directory?

My setup:
PYTHONPATH=.:/usr/local/lib/python1.5:/usr/local/lib/python1.5/site-packages

In /usr/local/lib/python1.5/site-packages the file idle.pth which
contains 1 line:

idle-0.5

In /usr/local/lib/python1.5/site-packages/idle-0.5/idle.py

In fact, I can only start idle like this:

$ cd /usr/local/lib/python1.5/site-packages/idle-0.5
$ python idle.py

OR

$ python /usr/local/lib/python1.5/site-packages/idle-0.5/idle.py

OR interactively

$ python
>>>import idle

What am I missing?
-- 
Small is Beautiful


From karab at stsci.edu  Tue Apr 18 15:49:56 2000
From: karab at stsci.edu (Kia Arab)
Date: Tue, 18 Apr 2000 19:49:56 GMT
Subject: can someone tell me about...
Message-ID: <38FCBB16.BF50EB03@stsci.edu>

variable names that start an underscore "_".

Interactively, python will recognize a variable name that begins with
one underscore (_myVar). The variable naming rules even say that is
acceptable.

However, when I try to use it in a module, it bombs. 

What's the deal?


From mcg at halcyon.com  Thu Apr  6 16:46:06 2000
From: mcg at halcyon.com (Michael C. Gilbert)
Date: Thu, 6 Apr 2000 13:46:06 -0700
Subject: Seattle area scripter needed
Message-ID: <v04210112b512a7e43b61@[216.39.167.47]>

We're offering both stock options AND social change.

I work for a company developing ASP services for nonprofit 
organizations. I'm looking for a talented web application scripter to 
join our team. Feel free to check out the company and its current 
offerings at our web site:

    http://www.socialecology.com

Send email ASAP to mailto:jobs at socialecology.com if you are 
interested! Give us some idea of your experience and past projects.

Other details:

  - rapid web application development experience
  - a desire to deploy applications to nonprofits
  - a proven ability to pick up new tools

  - short contract based trial followed by permanent (W-2) employment
  - modest salaries (we work with nonprofits), excellent stock options 
& benefits
  - located in Seattle (Capitol Hill) with supplemental telecommuting

Thanks!

-- Michael Gilbert



From darrell at dorb.com  Wed Apr  5 17:58:27 2000
From: darrell at dorb.com (Darrell)
Date: Wed, 05 Apr 2000 21:58:27 GMT
Subject: Remotely running a python script
References: <8cf452$pdv$1@gxsn.com>
Message-ID: <7GOG4.2280$bh.22442@typhoon.nyroc.rr.com>

None of this is your solution but should provide some ideas.

Win32 approach.
win32file.FindFirstChangeNotification
See http://www.dorb.com/darrell/win32WorkSvr/makeThumbsDG.py

There's lots of socket examples on deja.com.
Here's an example that might help you get started if sockets are your cup of
tea.
http://www.dorb.com/darrell/network/sproxy.py

It waits for a connection on one port then forwards it to another port.
In your case it would just launch a .py file and return the stdout.

If you have Mark Hammond's Win32 stuff installed try naming your script with
.pyw
It'll run nicely out of site without a console.

--Darrell


"Richard Chamberlain"
> Does anyone know a way I can remotely trigger a python script on a Win95
> machine?
>
> What I want to do is trigger an event from a vb client and run a python
> script on another PC in the business.
>
> They are able to see each other via IP and the client has a drive mapped
> that is a shared folder on the python side.
>





From PClaerhout at CREO.BE  Tue Apr 11 06:02:40 2000
From: PClaerhout at CREO.BE (Pieter Claerhout)
Date: Tue, 11 Apr 2000 12:02:40 +0200
Subject: Redirecting the output from sys.stdout
Message-ID: <2B1262E83448D211AE4B00A0C9D61B03BA7297@MSGEURO1>

Any examples on how to use the popen of the os module? Never used it before.

Kind regards,


Pieter

[ Pieter Claerhout :: CreoScitex Response Centre :: Applications support ]
[ Pieter_Claerhout at creoscitex.com    ::    http://www.creoscitex.com/ ]
[ CreoScitex Europe  ::   Excelsiorlaan 21   ::  1930 Zaventem  :: Belgium ]


-----Original Message-----
From: Oleg Broytmann [mailto:phd at phd.russ.ru]
Sent: Tuesday, April 11, 2000 11:54 AM
To: Pieter Claerhout
Cc: 'python-list at python.org'
Subject: Re: Redirecting the output from sys.stdout


On Tue, 11 Apr 2000, Pieter Claerhout wrote:
> I want to run a DOS command from Python, and I want to redirect the output of
> that command to a variable. Anyone a suggestion on how to accomplish this? The
> output of the command should not be shown on the screen.

   Long time ago I wrote a version of popen for DOS. If popen is not
available in your version of module os - download it here:
   http://sun.med.ru/~phd/Software/Python/
Look for dospopen. These patches are now obsolete, I don't know how well
it'd fitted you.

Faster mirrors: http://members.xoom.com/phd2.1/Software/Python/
http://www.fortunecity.com/skyscraper/unix/797/Software/Python/

Oleg.
---- 
    Oleg Broytmann    http://members.xoom.com/phd2.1/    phd2 at earthling.net
           Programmers don't die, they just GOSUB without RETURN.


-- 
http://www.python.org/mailman/listinfo/python-list



From nmm1 at cus.cam.ac.uk  Mon Apr  3 14:39:53 2000
From: nmm1 at cus.cam.ac.uk (Nick Maclaren)
Date: 3 Apr 2000 18:39:53 GMT
Subject: Pass by reference ?
References: <38E864C9.E1EE94F7@ecs.soton.ac.uk> <38E8D449.FBC1E2AD@roguewave.com>
Message-ID: <8caohp$j4v$1@pegasus.csx.cam.ac.uk>

In article <38E8D449.FBC1E2AD at roguewave.com>,
Bjorn Pettersen  <bjorn at roguewave.com> wrote:
>Jacek Generowicz wrote:
>> 
>> I'm just trying to familiarize myself with Python.
>> In Magnus Lie Hatland's Instant Python, he
>> mentions that `all parameters in Python are passed
>> by reference'.
>
>The technically correct (although much less known) term is
>pass-by-object-reference.  All this means is that you can change an
>argument to a function by calling methods on the object, but you can't
>change the identity of the object (ie. simply calling a function can't
>change any name-bindings in your current scope).

Hang on.  Why is that technically correct?  Call by reference has
never implied changing any bindings in the caller's scope in any
work that I have seen - it certainly didn't back in the 1960s.
Unless I am thoroughly confused, Python uses bog-standard call by
reference, with the wrinkle that assignment is a rebinding rather
than a copying operation.  AM I confused about that?

The form of argument passing that DOES involve rebinding the
caller's names certainly exists, but I have never seen an agreed
name for it.  It was reckoned to be a bad idea (just like Fortran's
equally deceptive argument passing of scalars) by 1970, but kept
coming back in special-purpose languages.


Regards,
Nick Maclaren,
University of Cambridge Computing Service,
New Museums Site, Pembroke Street, Cambridge CB2 3QG, England.
Email:  nmm1 at cam.ac.uk
Tel.:  +44 1223 334761    Fax:  +44 1223 334679


From kc5tja at garnet.armored.net  Sun Apr 30 20:23:29 2000
From: kc5tja at garnet.armored.net (Samuel A. Falvo II)
Date: 1 May 2000 00:23:29 GMT
Subject: What does "first class" mean?
References: <3909F2F5.A6F91256@spacenet.tn.cornell.edu> <slrn8gkalp.3bk.neelk@brick.cswv.com> <slrn8gn54f.4ih.kc5tja@garnet.armored.net> <390c5db6.17614800@news.nikoma.de>
Message-ID: <slrn8gqfa2.k65.kc5tja@garnet.armored.net>

>You can't return functions in C, or have arrays of functions, etc.
>Pointers to functions, yes.

But semantically, they are complete equivalent.  If I call:

	NoOp *pNOPFunction;
	
	pNOPFunction = Foo();
	pNOPFunction();	/* <-- legal ANSI C! */

-- 
KC5TJA/6, DM13, QRP-L #1447
Samuel A. Falvo II
Oceanside, CA


From donb at init.com  Mon Apr 17 12:05:25 2000
From: donb at init.com (Donald Beaudry)
Date: Mon, 17 Apr 2000 12:05:25 -0400
Subject: Another question 
References: <dVFK4.41$vr2.5817@news.uswest.net> <38FB2EEB.A32AF88C@bam.com> <38FB339F.604364AD@bam.com>
Message-ID: <200004171605.MAA04929@zippy.init.com>

Bill Scherer <scherbi at bam.com> wrote,
> I just did a little test and found this to be substantially fatser than the
> other three:
> 
> import random
> randint = random.randint
> 
> r = []
> 
> for i in xrange(1,10000):
>     r.append(randint(1,10))
> 
> for i in r: print i
> #

...then this one should be really impressive ;)

  def do_it():
      from random import randint
      r = []
      append = r.append
      for i in xrange(1,10000):
         append(randint(i))
      for i in r:
         print i

  do_it()

or better yet:

  def do_it():
      from random import randint
      r = []
      append = r.append
      map(append, xrange(1,10000))
      for i in r:
         print i

  do_it()

--
Donald Beaudry                                     Ab Initio Software Corp.
                                                   201 Spring Street
donb at init.com                                      Lexington, MA 02421
                  ...So much code, so little time...





From sp00fD at yahoo.com  Tue Apr 18 05:43:35 2000
From: sp00fD at yahoo.com (sp00fD)
Date: Tue, 18 Apr 2000 09:43:35 GMT
Subject: urllib redirect
References: <4.1.20000418125557.030592d0@202.95.1.37>
Message-ID: <8dhao4$2qk$1@nnrp1.deja.com>

You haven't really given a description of your problem, but I'll go
ahead and take a stab anyway...maybe you need another newline character
like so 'print "Content-type: text/html\n"'.  Otherwise, I think what
you have should work, although I haven't played with urllib too much
(just enough to verify that what you're doing will print out a web-
page).  Here's a tip, anytime you're having trouble, take it to the
interactive python prompt and try it there.

In article <4.1.20000418125557.030592d0 at 202.95.1.37>,
  Thomas <thomas at madeforchina.com> wrote:
> I've got a pretty tame question.
>
> I want one of my functions to redirect the URL.
> (ie I call the script and off I go to www.spam.com)
>
> So, here's what I get, and have not overcome yet:
>
> import urllib
> def simpleredirect() :
>     ############
>      print "Content-type: text/html"
>      print urllib.urlopen('http://www.spam.com').read()
>
> simpleredirect()
>
> TIA,
> Tom
>
>


Sent via Deja.com http://www.deja.com/
Before you buy.


From dworkin at ccs.neu.edu  Thu Apr 27 16:15:12 2000
From: dworkin at ccs.neu.edu (Justin Sheehy)
Date: 27 Apr 2000 16:15:12 -0400
Subject: Scripting and Gnome and KDE
In-Reply-To: Karl EICHWALDER's message of "27 Apr 2000 18:51:03 +0200"
References: <Pine.GSO.4.10.10004150741520.11261-100000@sundial> <3dk8hzb90u.fsf@newcnri.cnri.reston.va.us> <shk8hk2m68.fsf@tux.gnu.franken.de> <wkln1zll9r.fsf@mac.com> <shem7r32tk.fsf@tux.gnu.franken.de>
Message-ID: <vndg0s7uwq7.fsf@camelot-new.ccs.neu.edu>

Karl EICHWALDER <ke at gnu.franken.de> writes:

> Sure, but sooner or later, Linux will run on all architecure and you'll
> be free to choose the OS that fits your needs best :)

Quite a non-sequitur you have there, sir.  ;-)

-Justin

 



From thomas at madeforchina.com  Tue Apr 18 00:59:31 2000
From: thomas at madeforchina.com (Thomas)
Date: Tue, 18 Apr 2000 12:59:31 +0800
Subject: urllib redirect
Message-ID: <4.1.20000418125557.030592d0@202.95.1.37>

I've got a pretty tame question.

I want one of my functions to redirect the URL.  
(ie I call the script and off I go to www.spam.com)

So, here's what I get, and have not overcome yet:

import urllib
def simpleredirect() :
    ############
     print "Content-type: text/html"
     print urllib.urlopen('http://www.spam.com').read()

simpleredirect()

TIA,
Tom




From g2boojum at hotmail.com  Sun Apr 23 13:57:53 2000
From: g2boojum at hotmail.com (Grant Goodyear)
Date: Sun, 23 Apr 2000 12:57:53 -0500
Subject: executable "wrapper" gui -- advice sought
References: <xhnM4.1153$PP5.17158@news.swbell.net> <8dtsp7$e7h$0@216.39.151.169>
Message-ID: <Z1HM4.1820$PP5.28540@news.swbell.net>

My thanks to those who replied.  I'm using linux (mandrake 7 to be
specific), and this seemed to work reasonably well:

#!/usr/bin/env python
# htmlout.py

from popen2 import Popen3

cmd = "./a.out"
htmlfile = "/home/grant/public_html/test.html"
htmlout = open(htmlfile, 'w')
htmlout.write("<pre>\n") #hack so that I don't have to format the html

process = Popen3(cmd)
stdout = process.fromchild
print stdout
while (process.poll() == -1):  #while still running
    line = stdout.readline()
    if (line):
        htmlout.write(line)
        htmlout.flush()





From jwbnews at scandaroon.com  Sun Apr 30 21:50:18 2000
From: jwbnews at scandaroon.com (John W. Baxter)
Date: Sun, 30 Apr 2000 18:50:18 -0700
Subject: Discrepency between JPython and CPython
References: <m3k8hfjlex.fsf@boss.ksu.edu> <etdpur72pmx.fsf@w20-575-109.mit.edu>
Message-ID: <jwbnews-E7BE88.18501830042000@[207.126.101.100]>

In article <etdpur72pmx.fsf at w20-575-109.mit.edu>, Alex 
<cut_me_out at hotmail.com> wrote:

> > Consider the expression str(1.14 * 200).
> > In CPython it yields 228.0
> > In JPython it yields 227.99999999999997
> > 
> > Shouldn't this result in consistent values across variuos 
> > implementation?  
> > If so, which of the implementation is buggy?  
> > I think it is JPython, may be I am wrong.
> 
> Both implementations just reach down to a lower level routine -- for
> CPython, it's multiplication in C, for JPython, it's multiplication in
> Java.  So probably the discrepency is between the C compiler used to
> build your CPython (or maybe your CPU, if it does floating point
> operations.  I understand they do, these days.  [The only CPU I ever
> learnt anything about in detail was the 6809.] :) and the JVM your
> JPython is running on.
> 
> Alex

No...this is just yet another instance of CPython (less than 1.6) lying 
about the value of floats.  Since 1.14 is not precise, 1.14 * 200 can't 
be...yet the interpreter lies (by rounding off) and pretends that it is.

   --John

-- 
John W. Baxter   Port Ludlow, WA USA  jwbnews at scandaroon.com


From chris_barker at my-deja.com  Fri Apr 28 21:32:22 2000
From: chris_barker at my-deja.com (chris_barker at my-deja.com)
Date: Sat, 29 Apr 2000 01:32:22 GMT
Subject: Reading in strings -> numbers ??
References: <390A2D23.7029E557@fenx.com>
Message-ID: <8ede36$9r6$1@nnrp1.deja.com>

Well, the partner to file.write is file.read. What you are really
looking for is a partner to '%f %f'%(3,4), which is essentially an
sprintf(). So, what you want is an sscanf() or fscanf(). I've been
looking for that for a while. What I would like is a version like the
one in MATLAB, that gives you a whole matrix in one swipe.

Unfortunatly, it doesn't exist for python. Someone did make a Python
sscanf module, but I never got around to compiling it. there is also a
tablio (or something like that) module around. In the mean time, you can
always use something like:


list = map(float,string.split(file.readline)))

or, to use your example:

>>> map(float,string.split((" %e %e %d" % (1.2, -3.5, 10))))
[1.2, -3.5, 10.0]

it doesn't work if you want mixed types, but it gets you by.

-Chris

> Louis M. Pecora <pecora at anvil.nrl.navy.mil> wrote in message
> news:<280420001700308281%pecora at anvil.nrl.navy.mil>...
> > Is there a Python "read-partner" for the write(string) function,
e.g.,
> >
> > fileobject.write(" %e %e %d" % (1.2, -3.5, 10))
> >
> > in the sense that the C functions fprintf and fscanf are partners.
> >
> > I've looked in 4 different Python books and in the pdf references,
but
> > I see nothing in read or readlines or strings that seems to make for
an
> > easy conversion from a string to a list or tuple of objects
(numbers).
> > In other words, I want the opposite of the write above so I can read
> > those numbers back into my Python code.



Sent via Deja.com http://www.deja.com/
Before you buy.


From joacim at home.se  Fri Apr 28 02:11:32 2000
From: joacim at home.se (JJ)
Date: Fri, 28 Apr 2000 08:11:32 +0200
Subject: Wanted Daniel Larsson
References: <39075823.F3511829@cybercable.fr>
Message-ID: <8eba3r$km0$1@vg170.it.volvo.se>

According to a search in:
http://epostkatalog.telia.com/public/english/pmain.menu
Following information shows up:

E-mail            daniel.larsson at vasteras.mail.telia.com
Name              Daniel Larsson
Street Address    Knotav?gen 17
Postal Code       723 52
City              V?ster?s
Country           Sweden
Phone             021-84 00 87
Home Page         http://hem.passagen.se/dlarsson
Last modified     1997-04-11

I hope this would help.

BR
Joacim





"FR Chalaoux" <chalaoux at cybercable.fr> wrote in message
news:39075823.F3511829 at cybercable.fr...
> Hi,
>
> I looking for the right e-mail adress of "Daniel Larsson", the creator
> of Pythondoc?
> "Daniel.Larsson at vasteras.mail.telia.com" seems to be incorrect.
>
> Bye, FR.
>
>
>
>
>




From phd at phd.russ.ru  Fri Apr  7 08:23:34 2000
From: phd at phd.russ.ru (Oleg Broytmann)
Date: Fri, 7 Apr 2000 12:23:34 +0000 (GMT)
Subject: Newbie: How to do an exit ?
In-Reply-To: <38EDC1DA.62CBFCF3@sxb.bsf.alcatel.fr>
Message-ID: <Pine.LNX.4.21.0004071222320.13572-100000@fep132.fep.ru>

On Fri, 7 Apr 2000, Pascal Fremaux wrote:
> How to exit from a program ?
> I don't find a command such as exit(0) in C

import sys
sys.exit(0)

   Or

raise SystemExit, 0

   If you need _exit(1) it is

import os
os._exit(1)

Oleg.            (All opinions are mine and not of my employer)
---- 
    Oleg Broytmann      Foundation for Effective Policies      phd at phd.russ.ru
           Programmers don't die, they just GOSUB without RETURN.




From jbekesi at my-deja.com  Thu Apr 13 18:20:39 2000
From: jbekesi at my-deja.com (Janos BEKESI)
Date: Thu, 13 Apr 2000 22:20:39 GMT
Subject: windows registry
References: <era0iGYp$GA.273@net003s>
Message-ID: <8d5h78$fpq$1@nnrp1.deja.com>

The pythonwin package contains some interesting sources:

win32/scripts/regsetup.py

which uses (I think) win32/lib/regutil.py (the main module for handling
the registry). With a little effort one can figure out how things work.
Additionally there is some documentation included in the
win32/Help/Python Win32 Extensions.hlp-File, but looking at the sources
should show how all is put together. I used regutil.py for checking
whether there is python installed on a win32 system and if not, setting
the appropriate registry keys.

Hope it helps,

Janos.


> How do I edit the windows registry with Python?


Sent via Deja.com http://www.deja.com/
Before you buy.


From mfletch at tpresence.com  Tue Apr 25 12:55:57 2000
From: mfletch at tpresence.com (Mike Fletcher)
Date: Tue, 25 Apr 2000 12:55:57 -0400
Subject: new linereading standard?
Message-ID: <CC59868A0A76D311B1F50090277B321C1F32B0@VRTMAIL>

Or, if you prefer...

import fileinput # standard module
f = fileinput.FileInput( filename )
for line in f:
	print line

Enjoy,
Mike

-----Original Message-----
From: dyoo [mailto:dyoo at uclink4.berkeley.edu]
Sent: Tuesday, April 25, 2000 1:04 PM
To: python-list at python.org
Subject: Re: new linereading standard?


In article <3904D21B.333DD930 at visionart.com>, 
Pete Shinners <pete at visionart.com> wrote:
> i've come to the point where i can't get by even the simplest python
> project without dumping this code into my files. it is a simple class
> that handles the "while 1: .... break" carnival for reading files.
> now i can code in a preferred style

> myfile = open('myfile.txt', 'r')
> for line in filelines(myfile):
>     print line
> myfile.close()

> as a person trying to introduce his friends to python, i repeatedly find
> this is the ugliest wart while doing basic stuff with python.

Python does allow you to read the whole file at a time.  
readlines() will give you a list of all the lines in the file, so 
your code can look like:

  myfile = open('myfile.txt', 'r')
  for line in myfile.readlines():
    print line
  myfile.close()

-- 
http://www.python.org/mailman/listinfo/python-list



From effbot at telia.com  Sun Apr  9 15:28:29 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Sun, 09 Apr 2000 19:28:29 GMT
Subject: Tkinter:  centering windows on screen
References: <ak1I4.121418$Dv1.1306619@news1.rdc1.bc.home.com> <38F0C6A7.10878609@dkfz-heidelberg.de> <38F0C95D.1F7E853@dkfz-heidelberg.de>
Message-ID: <xR4I4.274$rc9.188784640@newsb.telia.net>

Kilian Lorenz <kilian.lorenz at dkfz-heidelberg.de> wrote:
> >> that works but doesn't help me much because I'm interested in the
> >> return value. So I would really like to know how to call Tk commands
> >> from within Python because .... blah blah
>
> Now I also found out how this works ... pretty easy with
> <widget>.tk.call(<command>, <parameter>, <next parameter> ...)
>
> You just have to browse Tkinter.py and get the answers to all your
> questions !  :)

if you browsed the source, how come you didn't find the
corresponding Tkinter methods?

windows geometry:
http://w1.132.telia.com/~u13208596/tkintrobook/x9867-window-geometry-methods
.htm

file dialogues:
http://w1.132.telia.com/~u13208596/tkintrobook/x1164-data-entry.htm

(direct pointers to the current draft; official version at
www.pythonware.com will be synchronized next week).

</F>




From PClaerhout at CREO.BE  Thu Apr 27 11:07:48 2000
From: PClaerhout at CREO.BE (Pieter Claerhout)
Date: Thu, 27 Apr 2000 17:07:48 +0200
Subject: pythoncom / Installer on Win32
Message-ID: <2B1262E83448D211AE4B00A0C9D61B030111537F@MSGEURO1>

Look into your windows system folder. You'll see that there's only a file
which is called pythoncom15.dll. Duplicate this file and rename it to 
pythoncom.dll. Should solve the problem.

Kind regards,


Pieter

Pieter Claerhout
Application Support - CreoScitex Europe


-----Original Message-----
From: Richard Chamberlain [mailto:richard_chamberlain at ntlworld.com]
Sent: Thursday, April 27, 2000 1:21 AM
To: python-list at python.org
Subject: pythoncom / Installer on Win32


i,

I am trying to create a distribution using Gordon's Standalone.

I'm asking a fair amount because I'm using COM, ODBC etc.

any help would be appreciated,

Richard

bindepend cannot analyze %s - file not found!
analyzing odbc
 found python15.dll
 found ODBC32.dll
 found dbi.dll
 lib not found: dbi.dll dependency of  found MSVCRT.dll
 found KERNEL32.dll
 added filter <TypeFilter {'d': 1}>
Creating COMServer.exe
Traceback (innermost last):
  File "Builder.py", line 502, in ?
    main(opts, args)
  File "Builder.py", line 484, in main
    target.build()
  File "Builder.py", line 87, in build
    self.assemble()
  File "Builder.py", line 300, in assemble
    arch = ArchiveTarget.assemble(self, pkgname)
  File "Builder.py", line 275, in assemble
    arch.build(pkgnm, toc)
  File "C:\WINDOWS\DESKTOP\INSTAL~1\MEInc\Dist\archive.py", line 146, in
build
  File "C:\WINDOWS\DESKTOP\INSTAL~1\MEInc\Dist\carchive.py", line 167, in
add
IOError: [Errno 2] No such file or directory:
'c:\\windows\\system\\pythoncom.dll'



-- 
http://www.python.org/mailman/listinfo/python-list



From tseaver at starbase.neosoft.com  Tue Apr  4 19:10:38 2000
From: tseaver at starbase.neosoft.com (Tres Seaver)
Date: 4 Apr 2000 18:10:38 -0500
Subject: Python 1.6 alpha 1 released
References: <Pine.GSO.4.10.10004040035420.21084-100000@sundial> <5l7leern5v.fsf@eric.cnri.reston.va.us> <8cdg92$j0m$1@nntp6.u.washington.edu> <8cdig3$s7f$1@slb6.atl.mindspring.net>
Message-ID: <C74E7E295E002791.623B538A22382124.E69ECD3BBA29FE56@lp.airnews.net>

In article <8cdig3$s7f$1 at slb6.atl.mindspring.net>,
Aahz Maruch <aahz at netcom.com> wrote:
>In article <8cdg92$j0m$1 at nntp6.u.washington.edu>,
>Russell E. Owen <owen at astroNOJNK.washington.edu.invalid> wrote:
>>
>>But tuples seem a less than wonderful choice to me. Protocol flexibility 
>>presumably requires each protocol to have its own tuple with its own 
>>numer of fields in its own order. That's a lot to remember and try not 
>>to get mixed up.
>
>I second this.  I'd be happy with either dict or class, with a *slight*
>preference for a class.

I'd be quite happy with a set of address classes -- they could *add value*
in lots of ways that the tuple can't.  For instance, consider IPv6 dual-
mode issues -- an IPAddress object could know how to render itself
appropriately in either mode.  

Hmm, new proposal (for the Py3K timeframe):

Make class socket.socket into an abstract interface, and socket.socket()
into a factory which produces an appropriate derived instance, based on
family, type, and proto.  The derived type would then automatically know
how to interpret the *args to connect(), bind(), etc., and Do the Right
Thing [TM].

Such a scheme would make adding new families/types/protocols a matter of
plugging callables into the lookup table used by socket.socket(), and could
be therefore done in entirely separate modules (SOCK_RAW, anyone? or
AF_INET6/AF_IRDA?)

Why-propose-a-small-incompatible-change-when-BIG-ones-abound?'ly

Tres.
-- 
---------------------------------------------------------------
Tres Seaver        tseaver at digicool.com     http://www.zope.org
Digital Creations


From mappel at 123ibm.net  Fri Apr 14 22:25:50 2000
From: mappel at 123ibm.net (Michael Appelmans)
Date: Sat, 15 Apr 2000 02:25:50 GMT
Subject: Python Starship Grounded??
References: <38F6B2DF.940E7C23@concentric.net>
Message-ID: <OqQJ4.8866$fV.715336@bgtnsc05-news.ops.worldnet.att.net>

Yes, I was hoping to get hold of win32all build 131 for Python 1.6a2 but to
no avail. I fear it is lost in some freakish nebulae.

--
Michael Appelmans
// remove 123 to mail
mappel at ibm123.net

"Manus Hand" <mjhand at concentric.net> wrote in message
news:38F6B2DF.940E7C23 at concentric.net...
Anyone here able to do anything about the illness that seems to have
come over starship.python.net?

It was out of commission all day Tuesday, and now all day today
(Thursday) without a visible light at the end of the tunnel.

It occasionally answers ping (well, it did earlier today, but....)
and sometimes a traceroute will crawl through to it, but http and
ftp and ssh, etc., etc. have failed to answer all day.  (And were
sluggish at best yesterday, and non-existent the day before.)

Manus




From effbot at pythonware.com  Thu Apr 13 06:28:43 2000
From: effbot at pythonware.com (effbot at pythonware.com)
Date: Thu, 13 Apr 2000 10:28:43 GMT
Subject: Compiling Python Imaging under LinuxPPC
References: <FU%I4.646$rc9.190630912@newsb.telia.net> <20000412144842.06030.00002524@ng-ch1.aol.com>
Message-ID: <8d47gh$vuo$1@nnrp1.deja.com>

William F. Adams wrote:
> When I attempt to compile the Sketch .src.rpm, the compilation
> fails because Imaging.h is not where it's supposed to be.
>
> I ddin't see much in the way of documentation with the
> .src.rpm for PIL, sorry.

RPM?  the official PIL distribution is a .tar.gz file, available
from www.pythonware.com.  last time I checked, it definitely came
with both a README and an Imaging.h file.  if you have a source
RPM that doesn't include those files, complain to whoever pre-
pared that file, and get the original distribution.

...

but it sounds like your real problem is with the Sketch application.
did you see the PIL-specific notes in the Sketch installation guide?
it says:

  Tne point not mentioned in the PIL installation instructions are the
  headerfiles (.h files). For easier configuration of Sketch I recommend
  to install the files the following files found the libImaging
  subdirectory into a directory under the Python include directory:

     ImConfig.h  ImPlatform.h  Imaging.h

  (e.g. if your Python include directory is /usr/include/python1.5, put
  them into /usr/include/python1.5/Extensions)

e.g.

  $ cd Imaging-1.0
  $ mkdir /usr/include/python1.5/Extensions
  $ cp libImaging/Im*.h /usr/include/python1.5/Extensions

it's likely that PIL 1.1 will do this for you, but as of 1.0,
you have to copy the files yourself.

</F>


Sent via Deja.com http://www.deja.com/
Before you buy.


From johngrayson at home.com  Sun Apr  9 18:35:14 2000
From: johngrayson at home.com (John Grayson)
Date: Sun, 09 Apr 2000 22:35:14 GMT
Subject: tkinter - changing the mouse pointer?
References: <38E91666.D3B87B4C@parlant.com>
Message-ID: <38F105A2.176B2C38@home.com>

Try page 369...

You might want to look at Pmw's BLT busy cursor 
for a full busy cursor implementation.

  Grayson

Thomas Lane wrote:
> 
> Does anyone out there know how to change the mouse pointer in tkinter? I'd like
> to change the mouse pointer to an hourglass (or the default 'wait' icon for the
> os) and back to normal after a short operation. How do I go about doing this?
> I've hunted through the Grayson book without success.
> 
> Thomas Lane
> Parlant Technology
> tom at parlant.com


From paul at fizzylab.com  Fri Apr  7 19:51:27 2000
From: paul at fizzylab.com (paul beard)
Date: Fri, 07 Apr 2000 16:51:27 -0700
Subject: calendar/scheduling app
References: <38EE332E.2954EA49@fizzylab.com> <0004071547040L.01234@quadra.teleo.net>
Message-ID: <38EE747F.3BD26F7@fizzylab.com>

Patrick Phalen wrote:
> 
> [paul beard, on Fri, 07 Apr 2000]
> :: I am interesting in having a meeting scheduler/calendar app that does
> :: what something like Exchange purports to do without requiring me to use
> :: a client that doesn't exist. In other words, it should be
> :: platform-agnostic and use webforms/cookies to handle transactions and
> :: display options.
> ::
> :: Possible in Zope? How hard? How long to do, given a baseline knowledge
> :: store of zero?
> 
> Try http://www.worldpilot.com

--
Paul Beard / 206 728 4282 x138
Fizzylab / 2025 First Ave, Suite 1150
Seattle WA 98121



From maxm at maxmcorp.dk  Mon Apr  3 16:25:02 2000
From: maxm at maxmcorp.dk (Max M)
Date: Mon, 03 Apr 2000 20:25:02 GMT
Subject: Building and Transvering multi-level dictionaries
References: <NDBBLCDEPKGPHBPDFALAKEJICDAA.maxm@normik.dk> <vnd1z4rtcdv.fsf@camelot-new.ccs.neu.edu>
Message-ID: <y67G4.167$xw1.8321@news.worldonline.dk>

"Justin Sheehy" <dworkin at ccs.neu.edu> wrote in message
news:vnd1z4rtcdv.fsf at camelot-new.ccs.neu.edu...
> "maxm" <maxm at normik.dk> writes:
>
> > I'm in recursion hell here and have been trying to wring out the keys
for
> > almost two days now and can no longer think straight.
>
> You could certainly write a recursive function to gather all of the
> keys, and I was about to do so... but I decided that it wasn't worth
> the work.  Instead, just make a couple of small changes to the class.

Hi Justin

Thanks for the code, but actually I think that I have found a much simpler
method for faking tree functionality in Python. I just use a tuple as a key
in a Dictionary:

#################################

theTree = {}

theTree[(1,)] = 'Max'
theTree[(2,)] = 'Gitte'
theTree[(2,4)] = 'Caroline'
theTree[(2,5)] = 'Clara'
theTree[(1,3)] = 'Magnus'
theTree[(1,3,6)] = 'unborn'

treeKeys= theTree.keys()
treeKeys.sort()

#print treeKeys

for key in treeKeys:
    print (len(key)-1)*3*' ',theTree[key]

#################################




From M_Pope at shef.rage.co.uk  Wed Apr 26 05:41:56 2000
From: M_Pope at shef.rage.co.uk (Mark Pope)
Date: Wed, 26 Apr 2000 10:41:56 +0100
Subject: C API - Creating Python Classes
Message-ID: <213CDEBFDDA7D3118081009027E40C5607116E@SHEF-MS1>

Hi,

Using the python C API I can, for example, create an Integer Python object
with the call:

	PyObject *pMyInt = PyInt_FromLong( 123 );

I have a python class which I need to create an instance of. How is this
done?

I assume I need something like:

	PyObject *pModule = PyImport_ImportModule( "MyClassScript" );

	PyObject *pInstance = NEW_INSTANCE( pModule, args );


thanks for any help,

mark.



From aaron at mancala.semo.net  Thu Apr 27 16:08:39 2000
From: aaron at mancala.semo.net (Aaron Malone)
Date: 27 Apr 2000 15:08:39 -0500
Subject: Newbie Question
References: <sgh4qdmbqtj130@corp.supernews.com> <39089928.47B55E1B@tismer.com>
Message-ID: <87n1mfi9x4.fsf@mancala.semo.net>

Christian Tismer <tismer at tismer.com> writes:

> But if you want to be sure only to loose \n, and also might
> want to solve the problem that the list line might have no \n,
> you are better off with a loop, like so: (after reading)
> 
> for i in range(len(mylist)):
>     if mylist[i][-1:] == "\n":
>         mylist = mylist[:-1]

Correction:

for i in range(len(mylist)):
     if mylist[i][-1:] == "\n":
         mylist[i] = mylist[i][:-1]

-- 
Aaron Malone (aaron at semo.net)
System Administrator                     I can bend minds with my spoon.
Poplar Bluff Internet, Inc.
http://www.semo.net


From jimserac at mindspring.com  Sun Apr  2 00:41:51 2000
From: jimserac at mindspring.com (James Pannozzi)
Date: Sun, 2 Apr 2000 00:41:51 -0500
Subject: New Features in Python 1.6
References: <200004011740.MAA04675@eric.cnri.reston.va.us>
Message-ID: <8c6mkg$8i$1@slb2.atl.mindspring.net>

... mouth watering description of new Python 1.6 features...

"Note that the Python installer for Windows now installs GNU m4, the bash
shell, and Autoconf, so that Windows users will be able to use the
Distutils just like on Unix."


Excellent, wonderful, yes, yes, ... ah whoops!!!  What's this about it
installing a bash shell, I've already got my NT Xemacs configured to use the
Cygwin 20 Bash shell.

(Oh Oh! reconfigure, incompatabilities... confusion, chaos...)

Also, I notice that links to Python2c have quietly become non-functional or
non-existent.

Thanks
Jim


"Guido van Rossum" <guido at python.org> wrote in message
news:200004011740.MAA04675 at eric.cnri.reston.va.us...
> New Features in Python 1.6
> ==========================
>
> With the recent release of Python 1.6 alpha 1, a lot of people have
> been wondering what's new.  This short note aims to explain the major
> changes in Python 1.6.
>
>
> Core Language
> -------------
>





From mwh21 at cam.ac.uk  Thu Apr 13 19:06:05 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 14 Apr 2000 00:06:05 +0100
Subject: Module Loading Issues
References: <EDFD2A95EE7DD31187350090279C67677079B4@THRESHER>
Message-ID: <m38zyhha8y.fsf@atrus.jesus.cam.ac.uk>

Brent Fulgham <brent.fulgham at xpsystems.com> writes:

> Is there any disadvantage to having many modules loaded
> at the same time?  I.e., is it possible for badly-written
> modules to write entries into the "__main__" namespace
> that might clash with other entries?

So long as you don't do to many "from ... import *"s you should be
fine.  I mean it's possible that a module will stomp on __main__, but
in practice they don't (if you find one that does, find the author and
slap him/her until he/she agrees it's a bad idea).
 
> Or is the only concern memory use?

I've never had any problem with that either, but then I've never used
Python in a memory tight situation.

Cheers,
M.

-- 
  nonono,  while we're making wild  conjectures about the behavior 
  of completely  irrelevant tasks,  we must not  also make serious 
  mistakes, or the data might suddenly become statistically valid.
                                      -- Erik Naggum, comp.lang.lisp


From effbot at telia.com  Tue Apr 18 17:50:59 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Tue, 18 Apr 2000 21:50:59 GMT
Subject: module __call__
References: <LX2L4.1275$B43.102174@news.pacbell.net>
Message-ID: <7N4L4.497$wYl.191816704@newsb.telia.net>

Dirck Blaskey <dirck at pacbell.net> wrote:
> Just a little GWIBNI....
>
> ...Gee-Wouldn't-It-Be-Nice-If...
>
> modules could def __call__():
>
> so you could call a module, i.e.
>
>     import StringIO
>     f = StringIO()

and

class myStringIO(StringIO):
    ...

would explode in your face.  lousy idea.

</F>




From parkw at better.net  Wed Apr 12 15:36:37 2000
From: parkw at better.net (William Park)
Date: Wed, 12 Apr 2000 15:36:37 -0400
Subject: Off-topic
In-Reply-To: <38F4C52C.450F611F@callware.com>; from ivanlan@callware.com on Wed, Apr 12, 2000 at 12:49:16PM -0600
References: <38F4BE12.A0C73D1D@callware.com> <m3zoqzgois.fsf@atrus.jesus.cam.ac.uk> <38F4C52C.450F611F@callware.com>
Message-ID: <20000412153637.A9859@better.net>

On Wed, Apr 12, 2000 at 12:49:16PM -0600, Ivan Van Laningham wrote:
> Sh scripts do work, as long as I remember to do ./script ;-).  I'm
> running as root (but the same thing happens when I run as myself, too). 
> However, this could be explained by the default behavour of exec(),
> which states that if the kernel can't run a program, it execs sh and
> hands it the program name in argv.

What is your PATH?  Is your problem related to that fact that '.' is not
in part of PATH environment variable?

--William



From bruce_dodson.nospam at bigfoot.com  Sat Apr  8 15:43:23 2000
From: bruce_dodson.nospam at bigfoot.com (Bruce Dodson)
Date: Sat, 8 Apr 2000 16:43:23 -0300
Subject: Python Language Support for ArcView GIS
Message-ID: <q0MH4.2398$Bq2.38636@newscontent-01.sprint.ca>

Python Language Support for ArcView GIS (aka AVPython) is now available
online as a free download.  I promised this several months ago but never
managed to find a home for it (too busy) so finally I decided to just slap
together a page for it on GeoCities:
www.geocities.com/brucedodson.rm/avpython.htm

If you're an ArcView developer and want to be able to control COM-enabled
programs like Excel or ArcInfo, or if you want to be able to interoperate
with distributed systems through standards like XML and HTTP, you might want
to give it a try.  The same goes for Python programmers who happen to be
working with ArcView GIS (hey, there must be a few!), and want to be able to
apply their Python skills (and use a better language) when developing
ArcView applications.

Python is free and open source.  I hope AVPython will eventually be the
same, but we haven't yet settled on an acceptable open source license, so
for now the best I can do is provide a binary that is freely
redistributable.

If you use AVPython on any interesting projects, or if you have any positive
feedback or things that you'd like to change, feed my hubris: tell me about
it!  Just take nospam out of my address, or use the one listed in the
documentation.

Sincerely,
Bruce Dodson



From glyph at twistedmatrix.com  Sat Apr 22 01:31:18 2000
From: glyph at twistedmatrix.com (Glyph Lefkowitz)
Date: 22 Apr 2000 00:31:18 -0500
Subject: The Simple Economics of Open Source
In-Reply-To: "Juergen A. Erhard"'s message of "Sat, 22 Apr 2000 05:43:13 +0200"
References: <14591.15630.192472.440375@goon.cnri.reston.va.us> <slrn8fv2rd.1kt.neelk@brick.cswv.com> <3900611F.3B525BDA@cfdrc.com> <FtDG4z.774@world.std.com> <slrn8g26tu.p7o.neelk@brick.cswv.com> <22042000.2@sanctum.jae.ddns.org>
Message-ID: <ja1z3yog61.fsf@helix.twistedmatrix.com>

I was too tempted by the philosophical sounding thread.  Forgive me
for jumping in =)

"Juergen A. Erhard" <jae at ilk.de> writes:

> Egoism is often equated with the profit motive these days (little
> wonder in our money-based society...[2])

Yes, and THIS is where the mistake is from ... there are a LARGE
number of reasons to do something for personal gain that decidedly do
NOT involve gaining money.  This is where economic models fall short,
IMHO -- the fact that they are so largely successful seems to me to be
a cynical statement on our society.  I suppose it comes from the fact
that the distinction between a corporation (whose only reason for
existance is money) and an individual (whose reason for existance I
will withhold judgement on for the time being) become somewhat
confused in that context.

For example, if I buy flowers for my lady-friend, and she is happy,
the fact that she is happy makes *me* happy, and does nothing to
improve my financial standing.  I lose money, the flowers are dead
after a month, and likely by then she's forgotten about them ... but
I'll buy them anyway.  Flowers are important.

If a corporation buys flowers for its employees, those have to be
"justified", as an "investment" in "human resources".  Happiness
doesn't matter, unless it can somehow increase productivity.

>   Xyzzy (for lack of a name[3]) is "I get *rid* of something".  Or "I
>   get something of my chest".

As I'm not doing anything altruistic when I buy flowers for a lady in
the previous example (her happiness makes me happy, for reasons having
to do with psychology too complicated to get into here) and so I'm
buying them to make myself happy, if I'm doing something to "get
something off my chest" (which I think is a horrible way of looking at
creative endeavors -- I prefer to think of them as a personal *gain*,
not a zero-sum necessity) then I'm doing it to get something off *my*
chest.  So you could express that as an egoistic motivation.

In fact, any altruistic (or xyzzyistic ^_^) motivation could be
reduced to an egoistic one -- obviously, you think that it stands to
gain you *something* by doing that thing, otherwise YOU wouldn't do
it.  Even the most self-avowedly altruistic person must be making
decisions as to what sort of altruism to engage in for themselves, and
therefore acting on some sort of self-based motivation.  The purer the
altruism, the less governance the altruist allowed over those
decisions.

The ultimate altruist is a mind-controlled zombie.

That said, the other interesting thing about open-source software is
that unlike other apparently altruistic endeavors, it requires very
little maintenance -- progress made in OSS is archived, and
effectively, anyone who wishes to follow the rules (I.E. adhere to the
license) can build upon it without wasting or duplicating any effort.

For example, if you run across some dusty two-year-old python code
that (f'rinstance) facilitates time-travel, and you hack on it for a
while, bring it up to date, but get bored or disinterested because of
more pressing concerns, like the upcoming alien invasion, you can put
your code out there and somebody else with a like mind might come
along and pick up where you left off.  Keeping the code archived in a
couple hundred places on the internet in case someone else becomes
interested is remarkably cheap.  The lower the "barrier to entry", the
more likely a random passerby is to contribute.

A soup-kitchen on the other hand, despite its more immediate (and IMHO
more deeply gratifying) effect of diminishing the horrors of poverty,
will collapse because of infrastructure expenses if the volunteers
gradually stop showing up, and won't be around for the next set of
volunteers to pick up at the point at which it left off, so it
requires more investment on their part, and more momentum to get going
again.  Real world brick-and-mortar type stuff requires significantly
more upkeep than bits, and is much harder to reproduce.

> Ranting-is-my-middle-name-(Chorus:-"Obviously!")-ly y'rs, J

Is signing with an adjective some sort of initiation rite on this list? ;-)

life-is-a-game-we-play-in-first-person-perspective-ly y'rs, Glyph

-- 
                  __________________________________________
                 |    ______      __   __  _____  _     _   |
                 |   |  ____ |      \_/   |_____] |_____|   |
                 |   |_____| |_____  |    |       |     |   |
                 |   @ t w i s t e d m a t r i x  . c o m   |
                 |   http://www.twistedmatrix.com/~glyph/   |
                 `__________________________________________'



From sbarron at ansic.net  Fri Apr  7 10:23:58 2000
From: sbarron at ansic.net (Scott)
Date: Fri, 7 Apr 2000 14:23:58 GMT
Subject: Formatting Excel cells with PythonCOM
References: <38ECAF9D.ECA74709@ansic.net> <1k9H4.28$4v.151@news-server.bigpond.net.au>
Message-ID: <38EDEF7E.E13977AC@ansic.net>

Hi Mark,

Wouldn't you know not 5 minutes after I posted the messag I figured it
out using the Excel VBA help.  Anyway, thanks for the information.  Btw,
I bought and read your PPW over my spring break and it helped me
tremendously, thanks!

Scott

Mark Hammond wrote:
> 
> This code is from the win32com test suite (in testMSOffice.py)
> 
>  # test dates out with Excel
>  xl.Cells(5,1).Value = "Excel time"
>  xl.Cells(5,2).Formula = "=Now()"
> 
>  import time
>  xl.Cells(6,1).Value = "Python time"
>  xl.Cells(6,2).Value = pythoncom.MakeTime(time.time())
>  xl.Cells(6,2).NumberFormat = "d/mm/yy h:mm"
>  xl.Columns("A:B").EntireColumn.AutoFit()
>


From moshez at math.huji.ac.il  Sat Apr 22 08:25:38 2000
From: moshez at math.huji.ac.il (Moshe Zadka)
Date: Sat, 22 Apr 2000 15:25:38 +0300 (IDT)
Subject: code not true?
In-Reply-To: <8ds3ku$q0l$1@nnrp1.deja.com>
Message-ID: <Pine.GSO.4.10.10004221525010.8110-100000@sundial>

On Sat, 22 Apr 2000 jeff_islay at my-deja.com wrote:

> Why does the second example return false?
> 
> >>> print 7.0 == 7
> 1
> >>> print (.07 * 100) == 7
> 0

Because computers represent floats with too little bits.
Never ever compare two floats for equality!

--
Moshe Zadka <mzadka at geocities.com>. 
http://www.oreilly.com/news/prescod_0300.html
http://www.linux.org.il -- we put the penguin in .com




From moshez at math.huji.ac.il  Tue Apr  4 11:40:09 2000
From: moshez at math.huji.ac.il (Moshe Zadka)
Date: Tue, 4 Apr 2000 17:40:09 +0200 (IST)
Subject: Python 1.6 alpha 1 released
In-Reply-To: <8ccie0$7fs$1@newshost.accu.uu.nl>
Message-ID: <Pine.GSO.4.10.10004041738200.15935-100000@sundial>

On 4 Apr 2000, Martijn Faassen wrote:

> And what Tres also mentioned was the external interfaces of ftplib,
> etc. A quick glance at the 1.6 alpha shows that they also have separate
> host/port arguments. It would seem inconsistent to change 'connect' to use
> tuples in one case, but 'connect' in ftplib a host and port argument..

This is a compelling argument, but a wrong one. You see, FTP is an
internet level protocol -- it always uses network sockets, and so can
expose a simpler interface. The socket module should support all kinds of
sockets (like UNIX domain sockets), so its interface must remain general.

--
Moshe Zadka <mzadka at geocities.com>. 
http://www.oreilly.com/news/prescod_0300.html
http://www.linux.org.il -- we put the penguin in .com




From helge.hess at mdlink.de  Mon Apr  3 05:48:55 2000
From: helge.hess at mdlink.de (Helge Hess)
Date: Mon, 03 Apr 2000 11:48:55 +0200
Subject: multiple interpreter sessions
References: <000b01bf9c6b$413cca80$752d153f@tim>
Message-ID: <38E86907.9C52EBD0@mdlink.de>

Tim Peters wrote:
> [Helge Hess]
> > I would like to have separate Python runtimes in a single server, so
> > that they don't affect each other (eg for security and state reasons).
> > ...
> 
> See the description of Py_NewInterpreter() in the "Initialization,
> Finalization, and Threads" section of the Python/C API manual.  You can
> create sub-interpreters from a C extension this way, but there's no
> Python-level interface to this at present.  Haven't used it myself, and,
> indeed, have yet to see any use of it.

Do you have another suggestion on how to implement 'multi-user Python'
in a single process ? I have some kind of application server which is
written mostly in C and I want to let the user write extensions in
Python. Since multiple users (and scripts) can be active in one process
I need to separate them from each other. I can't see how I can
accomplish that in another way if the user chooses to import something
(since the import [at least the module init] is global).

Besides that Py_NewInterpreter() seems to be what I want. I'm somewhat
confused that I get back a thread-state, do I switch between different
interpreters by doing some kind of manual context switch of Python
threads [that is, by calling save and restore thread] ?

thanks
  Helge


From darrell at dorb.com  Sun Apr  9 15:17:19 2000
From: darrell at dorb.com (Darrell)
Date: Sun, 09 Apr 2000 19:17:19 GMT
Subject: Python paradigms
References: <8cmr70$b0e$1@pegasus.csx.cam.ac.uk> <8cntqp$upi$1@slb0.atl.mindspring.net> <8covpn$4jq$1@slb6.atl.mindspring.net> <8cqk7u$dmc$1@pegasus.csx.cam.ac.uk>
Message-ID: <3H4I4.709$4w1.17557@typhoon.nyroc.rr.com>

> x = (a != NULL ? a[i]->weeble : 0) + (b != NULL ? b[i]->wombat : 0)

x=0
if a != NULL:
    x=a[i].weeble

if b != NULL:
    x=x+b[i].wombat

Isn't this more clear?
In fact isn't not much if any longer than the "C" version when you count the
parens used.

--Darrell







From neilh at hare.net.au  Mon Apr  3 07:46:24 2000
From: neilh at hare.net.au (Neil Hodgson)
Date: Mon, 3 Apr 2000 21:46:24 +1000
Subject: Unicode program representation
References: <015c01bf9d0b$76f9b2a0$59dd3fcb@neil> <3xXF4.1060$n68.186533888@newsb.telia.net>
Message-ID: <001201bf9d62$3e1ac1c0$36dd3fcb@neil>

> >    I think this should be changed to interpreting the literal as a UTF-8
> > literal. The advantage here is that non-roman string literals become a
> > natural part of the language.
>
> however, perfectly valid ISO Latin 1 strings (or any other standard
> 8-bit encoding) are invalid UTF-8 strings.   changing the default to
> UTF-8 this is likely to break many existing scripts.

   But there are relatively few scripts currently using the u" form since it
has yet to appear in a release version of Python.

   Neil





From dalke at acm.org  Mon Apr 17 05:27:31 2000
From: dalke at acm.org (Andrew Dalke)
Date: Mon, 17 Apr 2000 03:27:31 -0600
Subject: pattern-based string expansion
References: <38F52C61.A5D72F89@gte.net> <vndog79ft1q.fsf@camelot-new.ccs.neu.edu>
Message-ID: <8del87$5md$1@slb6.atl.mindspring.net>

Justin Sheehy wrote:
>In fact, if you retain the expressivity of glob, the function you
>describe is not practically possible.  What should it return if given
>the glob "*"?


A generator for all possible strings?

Here's an outline for a very inefficient solution, assuming the
input is an re instead of a glob.  And, um, if there are only
a finite number of strings, this will still never terminate
until the integer counts overflow or you run out of memory.  :)

class MatchGenerator:
  def __init__(self, pattern):
    self.pattern = pattern
    self.size = 0
    self.letters = []
    self.offset = 0
    self._n = 0
  def __getitem__(self, i):
    assert self._n == i, "forward iteration only"
    x = self.next()
    if x is None:
      raise IndexError, i
    return x
  def next(self):
    self._n = self._n + 1
    while 1:
      s = string.join(self.letters)
      if self.pattern.match(s):
        break
        # advance the value of the current offset letter one character
        # if it's the end, advance the offset character
        # if *that's* at the end, try the next larger size string
    return s

That was only in jest, but it's a reasonable result.  If the
glob has an infinite number of matches, __len__ could raise some
sort of exception.  For a finite number, the getitem might even
be random access.

Finally, as an exercise for the student, it should take an optional
list of finite strings in the constructor, for backward compatibility
to glob :)

                    Andrew
                    dalke at acm.org





From ullrich at math.okstate.edu  Tue Apr  4 19:40:37 2000
From: ullrich at math.okstate.edu (David C. Ullrich)
Date: 4 Apr 2000 23:40:37 GMT
Subject: Why should I switch to Python? - Infinity of Primes
References: <1257244378-5884705@hypernet.com>
Message-ID: <01bf9e90$a57d6140$6296cdcf@daves-dell>


Gordon McMillan <gmcm at hypernet.com> wrote in article
<1257244378-5884705 at hypernet.com>...
> David C. Ullrich wrote:
> 
> > ... Could be that there's a proof of the existence
> > of infinitely many primes using FTA that has some property
> > not shared by Euclid's proof, but "constructive" isn't it.
> 
> Oh, come on. There are at least 3 well known constructive 
> proofs that all odd numbers are prime:

	Of course there are. And it follows that there are infinitely
many primes - so this gives several _more_ constructive proofs
of that fact.

	(I didn't say the proof wasn't constructive - of course it is.
I said "constructive" was not a property that this proof did not
share with the hypothetical FTA proof.)

DU


From effbot at telia.com  Tue Apr  4 16:05:47 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Tue, 04 Apr 2000 20:05:47 GMT
Subject: New Features in Python 1.6
References: <200004011740.MAA04675@eric.cnri.reston.va.us> <_pyF4.25562$E85.640966@news1.rdc1.md.home.com> <slrn8edfmk.185.wtanksle@dolphin.openprojects.net> <ltvh21ywfs.fsf@asfast.com> <oq66txol0w.fsf@titan.progiciels-bpi.ca> <GOpG4.5050$74.81782@newsc.telia.net> <38ea40a1.47458347@news.nikoma.de>
Message-ID: <vWrG4.5095$74.82533@newsc.telia.net>

Philip Newton wrote:
> >why not just use two different compression algorithms repeatedly,
> >until you've squeezed the file down to a suitable size?  it takes a
> >little work, but you can design them so that one of them can always
>                   ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
> >strip away at least one bit from the input stream.
>  ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
>
> Yes? I challenge you to provide any algorithm that will strip away one
> bit from the four sequences 11, 10, 01, and 00, while still being able
> to restore the previous state.

okay, okay, I shouldn't have posted that several days after
april 1st...  but hey, this group is (indirectly) named after a
infamous comedy troupe, so what did you expect?

anyway, here's the code:

    n = number_of_bits
    while n > minsize:
        if skip_false_bit():
            print "used algorithm 1"
        elif skip_true_bit():
            print "used algorithm 2"
        else:
            raise IOError, "deranged data"
        n = n - 1

where skip_false_bit() and skip_true_bit() returns true if
the succeed to actually compress the data.

to decode, just call insert_false_bit() and insert_true_bit()
repeatedly, in the *right* order ;-)

</F>




From smalleys at gte.net  Fri Apr 14 00:29:48 2000
From: smalleys at gte.net (sue)
Date: Fri, 14 Apr 2000 04:29:48 GMT
Subject: Call by Reference / Value
References: <8d4rr7$t8k$1@newnews.ilk.de>
Message-ID: <38F690BF.735CA9AF@gte.net>

I haven't had this problem with SWIG/TCL.  Are you sure
python can see the real data from the wrapper, and not just
the data types?  I always put
	typedef { stuff } TYPE; extern TYPE data;
in the .h file I feed to SWIG (unless I use custom accessor
fns).

Both python and tcl will create a variable if they can't
find it, and silently continue on.  This can give you good
separately tested C and python code which fails when you
hook the two together.

Sue.


From tjreedy at udel.edu  Tue Apr  4 15:31:42 2000
From: tjreedy at udel.edu (Terry Reedy)
Date: Tue, 4 Apr 2000 15:31:42 -0400
Subject: Python 1.6 alpha 1 released
References: <Pine.GSO.4.10.10004040035420.21084-100000@sundial>
Message-ID: <8ce387$dfr$1@news.udel.edu>

"Moshe Zadka" <moshez at math.huji.ac.il> wrote in message
news:Pine.GSO.4.10.10004040035420.21084-100000 at sundial...
...
> I beg to differ: it's not highly relevant. What Guido did was correct
> a *bug* in the implementation.

But *where* is the bug?  That is the question!
OK, there is currently a discrepency between
A. one sentence in the documentation
B. the implementation, another sentence in the documentation, and the books
written
by gurus Lutz, Watters, Rossum, and Ahlstrom.

> That people depended on it is sad,

Please excuse me and others for believing the cumulative evidence of B.

> and just goes to show it should be corrected ASAP before more people
depend on it.

There are two possible remedies to the conflict, not just one.
A. add 7 chars, '[, port]' , to the one sentence and leave everything else
alone [1 minute?].
B. change the implementation and break bits of code all over [days to fix
and test]

Seems reasonable to me to at least consider option A instead of B.

Terry J. Reedy






From abrahams at mediaone.net  Sat Apr 29 19:52:47 2000
From: abrahams at mediaone.net (Dave Abrahams)
Date: Sat, 29 Apr 2000 23:52:47 GMT
Subject: converting word documents to html
References: <390B501C.1891CB80@uiuc.edu>
Message-ID: <B530EDCA.5754%abrahams@mediaone.net>

in article 390B501C.1891CB80 at uiuc.edu, Joe Saltiel at jsaltiel at uiuc.edu
wrote on 4/29/00 4:11 PM:

> Hey,
> 
> I was wondering if anybody had done any work on converting
> word processors type documents to html.

Check out this freeware:

http://www.xwebware.com/products/rtf2html/



From aa8vb at yahoo.com  Thu Apr 20 09:39:45 2000
From: aa8vb at yahoo.com (Randall Hopper)
Date: Thu, 20 Apr 2000 09:39:45 -0400
Subject: Python Parsers Comparison
Message-ID: <20000420093945.A624723@vislab.epa.gov>

I'm again faced with the "regexes not powerful enough for nested grammar"
problem, and need to chose an extension.

Does someone know of a web page or paper which compares the various Python
parser engines on a simple example grammar (or on features)?

For example, a comma-separated list of nested vectors would be a good example:

    [ 1, 2 ], [ 3, 4 ]
    [ [ 1,2 ],[ 3,4 ] ], [ 5.6 ]
    etc.

mxTextTools, metalanguage, SimpleParse, SPARK, YAPPS, PyLR, kwParsing,
PyBison.  A comparison of any subset would be useful.

Thanks,

Randall


-- 
Randall Hopper
aa8vb at yahoo.com



From sbarron at ansic.net  Fri Apr  7 10:25:13 2000
From: sbarron at ansic.net (Scott)
Date: Fri, 7 Apr 2000 14:25:13 GMT
Subject: Formatting Excel cells with PythonCOM
References: <38ECAF9D.ECA74709@ansic.net> <1k9H4.28$4v.151@news-server.bigpond.net.au>
Message-ID: <38EDEFC9.8222AEB2@ansic.net>

Hi Mark,

Wouldn't you know not 5 minutes after I posted the messag I figured it
out using the Excel VBA help.  Anyway, thanks for the information.  Btw,
I bought and read your PPW over my spring break and it helped me
tremendously, thanks!

Scott

Mark Hammond wrote:
> 
> This code is from the win32com test suite (in testMSOffice.py)
> 
>  # test dates out with Excel
>  xl.Cells(5,1).Value = "Excel time"
>  xl.Cells(5,2).Formula = "=Now()"
> 
>  import time
>  xl.Cells(6,1).Value = "Python time"
>  xl.Cells(6,2).Value = pythoncom.MakeTime(time.time())
>  xl.Cells(6,2).NumberFormat = "d/mm/yy h:mm"
>  xl.Columns("A:B").EntireColumn.AutoFit()
>


From PClaerhout at CREO.BE  Tue Apr 11 05:41:13 2000
From: PClaerhout at CREO.BE (Pieter Claerhout)
Date: Tue, 11 Apr 2000 11:41:13 +0200
Subject: Redirecting the output from sys.stdout
Message-ID: <2B1262E83448D211AE4B00A0C9D61B03BA7296@MSGEURO1>

Hello,

I want to run a DOS command from Python, and I want to redirect the output of
that command to a variable. Anyone a suggestion on how to accomplish this? The
output of the command should not be shown on the screen.

Kind regards,


Pieter

[ Pieter Claerhout :: CreoScitex Response Centre :: Applications support ]
[ Pieter_Claerhout at creoscitex.com    ::    http://www.creoscitex.com/ ]
[ CreoScitex Europe  ::   Excelsiorlaan 21   ::  1930 Zaventem  ::   Belgium ]




From m.faassen at vet.uu.nl  Tue Apr 18 15:49:11 2000
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 18 Apr 2000 19:49:11 GMT
Subject: Pliant?
References: <8di0lk$4h0$1@coco.singnet.com.sg>
Message-ID: <8die7n$p4m$2@newshost.accu.uu.nl>

Ng Pheng Siong <ngps at madcap.dyndns.org> wrote:
> Saw a reference to the language Pliant on Freshmeat today.
> Looks pretty interesting. What are your views on Pliant,
> as a Python programmer?

>From the little I've read about Pliant, cool idea, but it's just a 
bit too pliant for my software engineering taste. It'll have the same
problems Perl does -- too many ways to do it, so you can't easily
read someone else's code (or your own code some days or weeks later). 
>From that perspective something like Python's clear but fairly rigid
syntax is nice.

>From a designer's point of view, a language which can morphed into
various things is definitely cool, though.

Note that I'm talking through my hat mostly, I haven't studied Pliant
in any detail. :)

Regards,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From wheinemanNOwhSPAM at uconect.net.invalid  Thu Apr  6 11:25:33 2000
From: wheinemanNOwhSPAM at uconect.net.invalid (wheineman)
Date: Thu, 06 Apr 2000 08:25:33 -0700
Subject: Refactoring browser for Python?
References: <16d61288.5cc55c16@usw-ex0101-005.remarq.com>
Message-ID: <19c2a6b4.30d6893e@usw-ex0103-019.remarq.com>

There seems to be some discussion of just this at WikiWiki. See

http://c2.com/cgi-bin/wiki?PythonRefactoringBrowser

and

http://c2.com/cgi-bin/wiki?BicycleRepairMan

BicycleRepairMan being the current name of the browser (highly
commendable choice). There's only talk at the moment but they are
looking for volunteers.

* Sent from RemarQ http://www.remarq.com The Internet's Discussion Network *
The fastest and easiest way to search and participate in Usenet - Free!



From kc5tja at garnet.armored.net  Sun Apr 30 08:44:51 2000
From: kc5tja at garnet.armored.net (Samuel A. Falvo II)
Date: 30 Apr 2000 12:44:51 GMT
Subject: Do I always have to write "self." ?
References: <270420001706386648%pecora@anvil.nrl.navy.mil> <slrn8ghc6t.f01.scarblac-spamtrap@flits104-37.flits.rug.nl> <8ecsgj$6lp$1@slb0.atl.mindspring.net> <m3bt2tonyg.fsf@atrus.jesus.cam.ac.uk> <300420001008012011%pecora@anvil.nrl.navy.mil>
Message-ID: <slrn8gp6c4.ilf.kc5tja@garnet.armored.net>

In article <300420001008012011%pecora at anvil.nrl.navy.mil>, Louis M. Pecora
wrote:
>Isn't this called something like "Hungarian" variable naming?  It has

Hungarian Notation is what it's called.

>use g for global variables:  gChannel, gSocket1, etc.  I thought it was
>an aberration. Ugly!  A suffix would be much more readable:  Channelg,
>Socket1g etc.  But even then...

Suffixes are infinitely HARDER to read, because the type moniker merges into
the name of the object itself.  To use your example above, "Channelg."  Is
Channelg the English word "Channel" followed by a 'g', or is it a word in
some other language?  I can't tell.  Also, you wouldn't say things like
"Channel global" would you?  In most English speaking countries, people
would tend to say "Global Channel" instead.  Hence, gChannel.

It also defeats the whole purpose of Hungarian notation to begin with.  The
idea of Hungarian notation is to allow the programmer to determine the type
of object the symbol refers to immediately.  And since we tend to read from
left to right, it makes sense to place such information left-most (e.g., to
prepend).

The place where suffixes are best applied are in actual data types (versus
variable names).  For example, "size_t" in ANSI C.  It's the Size Type.
However, having a variable that represents the count of the bytes in an
object would be cbObject, not Object_cb.

I always use Hungarian notation when it makes sense to.  For example, if I
have a function in C that accepts a string and an explicit length argument,
I would write something like this:

	uint32 WriteData( void *buffer, char *pString, uint32 cbString )

This tells me that WriteData() takes as parameters a buffer, a pointer to a
string, and the number of bytes in the said string.  I fail to see how this
is somehow hard to read.  :)

-- 
KC5TJA/6, DM13, QRP-L #1447
Samuel A. Falvo II
Oceanside, CA


From jmg at ecs.soton.ac.uk  Mon Apr 17 13:55:23 2000
From: jmg at ecs.soton.ac.uk (Jacek Generowicz)
Date: Mon, 17 Apr 2000 18:55:23 +0100
Subject: Handling errors in open
References: <38FB4D0A.489FF50F@ecs.soton.ac.uk>
Message-ID: <38FB500B.E004558F@ecs.soton.ac.uk>

Jacek Generowicz wrote before checking his spelling:

> Hi,
>
> Is there a way to prevent a Python script from
> terminating when `open' fails, and have it handle
> the error more helpfully . . . I was thinking
> along these lines:
>
> from sys import argv
>
> try:
>     file = open(argv[1],"r")
> except IOerror:

IOError does the trick !

>
>     print "You lied about that file."
>
> Jacek



From robin at jessikat.demon.co.uk  Fri Apr 21 14:50:12 2000
From: robin at jessikat.demon.co.uk (Robin Becker)
Date: Fri, 21 Apr 2000 19:50:12 +0100
Subject: The Simple Economics of Open Source
References: <3900611F.3B525BDA@cfdrc.com> <1255786824-35980039@hypernet.com>
Message-ID: <f4JdfuAkLKA5EwcQ@jessikat.demon.co.uk>

In article <1255786824-35980039 at hypernet.com>, Gordon McMillan
<gmcm at hypernet.com> writes
>Why does everyone miss the main point?
>
>I (and you) are literally 100s of times more productive because 
>of open source. Imagine having to either invent it all yourself, 
>or duct tape together expensive half-assed closed "solutions". 
>Shudder.

Developers who share code may actually do so for other than friendly
reasons. If M$ released win2000 source into the world it might actually
slow down Linux development.

I prefer not to worry about the exact electronics/chemistry behind my
computer. So in that sense my 'open sourceness' has limits. I guess this
argument has to do with the relative value 'developers' put on the
tools. On win32 VC++ is valuable; while gcc/egcs is really useful on
unices.
 
>
>Developers who share code are *enormously* more productive 
>than those who don't. So the range of problems that can be 
>attacked increases geometrically, so computers penetrate 
>new areas, so geeks stand a better chance of making a living.

This is a competitive position to take. Geeks presumably compete with
non-geeks for resources.

In addition the inter geek co-operation may actually be a form of
competition for other things like 'mindset' 'noosphere' etc. Painters
exchange ideas, concepts and criticism freely, but compete for
customers.

>
>Win-win situations are rarely profitably analyzed by 
>competitive models.
>
>
>- Gordon
>

Wrong! Non-zero sum games are widely studied in many areas of
mathematics, economics and control theory. Just because we can both win
doesn't mean we're not in competition. Win-Win games tend not to be so
stable as minimal-loss minimal-loss (Nash) games and for that reason are
of less interest. It requires a constant flow of information (co-
operation) to achieve best-best so that is rarely achieved or sustained.

Win-Win for the group of geeks is a sub-game in the wider community. As
a group the geeks strive to turn their knowledge into something useful
to get income. It may be that inter geek co-operation aids the group as
a whole; certainly that cannot be simply asserted as there exists a
large number of geeks who work only for $/? and who co-operate with the
prince of darkness in Seattle. I don't believe any comparative study of
commercial vs opensourcerer geeks has yet been carried out to see which
group does better.
-not to be take too serious-ly yrs- 
Robin Becker


From tim_one at email.msn.com  Thu Apr  6 22:19:00 2000
From: tim_one at email.msn.com (Tim Peters)
Date: Thu, 6 Apr 2000 22:19:00 -0400
Subject: [Python-Dev] Round Bug in Python 1.6?
In-Reply-To: <38ED0016.E1C4A26C@tismer.com>
Message-ID: <000601bfa037$a2c18460$6c2d153f@tim>

[posted & mailed]

[Christian Tismer]
> as a side effect, I happened to observe the following rounding bug.
> It happens in Stackless Python, which is built against the
> pre-unicode CVS branch.
>
> Is this changed for 1.6, or might it be my bug?

It's a 1.6 thing, and is not a bug.

> D:\python\spc>python
> Python 1.5.42+ (#0, Mar 29 2000, 20:23:26) [MSC 32 bit (Intel)] on win32
> Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
> >>> round(3.1415926585, 4)
> 3.1415999999999999
> >>> ^Z

The best possible IEEE-754 double approximation to 3.1416 is (exactly)

3.141599999999999948130380289512686431407928466796875

so the output you got is correctly rounded to 17 significant digits.  IOW,
it's a feature.

1.6 boosted the number of decimal digits repr(float) produces so that

    eval(repr(x)) == x

for every finite float on every platform with an IEEE-754-conforming libc.
It was actually rare for that equality to hold pre-1.6.  repr() cannot
produce fewer digits than this without allowing the equality to fail in some
cases.

The 1.6 str() still produces the *illusion* that the result is 3.1416 (as
repr() also did pre-1.6).

IMO it would be better if Python stopped using repr() (at least by default)
for formatting expressions at the interactive prompt (for much more on this,
see DejaNews).

the-two-things-you-can-do-about-it-are-nothing-and-love-it<wink>-ly
    y'rs  - tim





From sblakey at freei.com  Fri Apr 14 18:51:39 2000
From: sblakey at freei.com (Sean Blakey)
Date: Fri, 14 Apr 2000 15:51:39 -0700 (PDT)
Subject: for item in some_class_object...
In-Reply-To: <4j6ffs0b6ins12qq559q07u28pc6o0mc9l@4ax.com>
Message-ID: <20000414224718.D42301CD5F@dinsdale.python.org>

On 14 Apr, Ben Wolfson wrote:
> 
> What special methods should I define for a class if I want to support
> a for item in ... syntax for it?
> 
All you need to do is implement __getitem__.

>>>class Seq:
...    def __getitem__(self, i):
...        if i < 10:
...            return i
...        else:
...            raise IndexError
...
>>>s = Seq()
>>>for i in s:
>>>    print i

-- 

Sean Blakey
FreeInternet.com
sblakey at freei.com
(253)796-6500x1025




From gmcm at hypernet.com  Sun Apr  9 09:14:41 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Sun, 9 Apr 2000 09:14:41 -0400
Subject: Windows quotes (was Re: Installer Problem Finding Modules)
In-Reply-To: <8cp5d5$qcu$1@slb6.atl.mindspring.net>
Message-ID: <1256836409-30423792@hypernet.com>

Aahz Maruch wrote:

> In article <1257176798-9949680 at hypernet.com>,
> Gordon McMillan <gmcm at hypernet.com> wrote:
> >
> >Grrr. The fact that Windows allows spaces in directory names doesn't
> >mean the "feature" should be used.
> >
> >A better fix, I think, is to quote all the %s's.
> 
> I re-discovered recently that you need to use "-style quotes on Windows;
> '-style quotes don't work.

When I went to implement my proposed fix, I found it's even 
worse. Windows complains (when using os.system) if you 
quote an arg that *doesn't* contain a space. Using, of course, 
a suitably obtuse error message.

>>> os.system('"D:\\Programs\\Python\\python.exe" 
"d:\\programs\\python\\test.py"')
The filename, directory name, or volume label syntax is 
incorrect.

(Actually, it will accept quotes around the arg, but not around 
the program, unless there's a space in it.)


- Gordon



From JGRAVES3 at austin.rr.com  Fri Apr  7 08:38:44 2000
From: JGRAVES3 at austin.rr.com (Jay Graves)
Date: Fri, 07 Apr 2000 12:38:44 GMT
Subject: Searching for Python Material...
References: <8F0ED18C1joshtinamecom@209.208.0.7>
Message-ID: <oFkH4.203$0p1.45386@typhoon.austin.rr.com>

Josh Tompkins wrote in message <8F0ED18C1joshtinamecom at 209.208.0.7>...
>Buenos Dias, everyone.

Hello...

>I'm fairly new to this whole python thing.  I think that I'm getting better

If you already know some programming, Learning Python is a great book.  I
read it almost straight through, switching over to a Python interepreter
when I wanted to try something out.  Good introductory book, IMO.

>So, anyway, I was wondering if anyone could point me to some websites that
>cover material that is more advanced than the stuff in most of the
>tutorials out there.  I'm actually looking for something along the list of:
>
>to do (something cool in python) do this...
>
>where something cool in python is a variable with values such as "Accessing
>webpages and/or e-mail", "Connecting to a database", etc, while still
>keeping the deep magic stuff to a relative minimum.

Maybe the Vaults of Parnassus would be of some help.

http://www.vex.net/parnassus/

Welcome to Python...
Jay




From m.faassen at vet.uu.nl  Mon Apr 10 18:37:18 2000
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 10 Apr 2000 22:37:18 GMT
Subject: Python/Perl Popularity (Re: A Mountain of Perl...)
References: <200004102111.HAA03380@envy.fulcrum.com.au>
Message-ID: <8ctl2u$bko$3@newshost.accu.uu.nl>

Richard Jones <Richard.Jones at fulcrum.com.au> wrote:

> [tom__98 at my-deja.com]
>> I don't know about others, but I frequently end up using
>> Perl instead of Python, even though I greatly prefer Python
>> as a language.

>    Tom you little troll-meister. Half those arguments you cite are bollocks
> if you _really_ knew and "greatly prefer" Python. Try again.

It's considered good Usenet style to include at least an attempt at
refuting arguments, instead of just saying they're bollocks. :)

Is this an example of a flaming Python advocacy as talked about earlier
in the thread? :)

PHYTON-RULZEZ!!!-ly yours,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From michael.stroeder at inka.de  Mon Apr 17 02:22:06 2000
From: michael.stroeder at inka.de (Michael =?iso-8859-1?Q?Str=F6der?=)
Date: Mon, 17 Apr 2000 08:22:06 +0200
Subject: Python w/ SSL
References: <38FA3450.1DD845B2@flashmail.com>
Message-ID: <38FAAD8E.A84D6601@inka.de>

Jp Calderone wrote:
> 
>  Can anyone point me to an *up to date* Python/SSL interface?

M2Crypto on http://www.post1.com/home/ngps/m2/

Ciao, Michael.


From sholden at bellatlantic.net  Tue Apr  4 17:47:03 2000
From: sholden at bellatlantic.net (Steve Holden)
Date: Tue, 04 Apr 2000 21:47:03 GMT
Subject: complete novice question
References: <38EA5995.F7B9815D@newbridge.com>
Message-ID: <38EA6230.C6DFD7B2@bellatlantic.net>

Paul Madsen wrote:
> 
> Downloaded Python-1.5.2.
> 
> I've run 'configure' and it seems to have worked. Now, according to the
> Python package's README, I should be able to type 'make' for the rest of
> the install. If I don't have a 'make' file, did the configure portion
> not work?
> 
> thanks

Hopefully, "make" is in /usr/bin or some other path component.  It uses
"Makefile" or "makefile" in the local directory to decide what and how to
build.  Just hold your breath and type "make" at the comnmand line!

Good luck.

regards
 Steve
--
"Bulding information systems just because it's fun."
Steve Holden  sholden at bellatlantic.net  703 716 7275


From paulb at infercor.no  Mon Apr  3 08:28:15 2000
From: paulb at infercor.no (Paul Boddie)
Date: Mon, 03 Apr 2000 14:28:15 +0200
Subject: Python 1.6 alpha 1 released
References: <200003312130.QAA04361@eric.cnri.reston.va.us> <38E5205F.DE811F61@roguewave.com> <Wf9F4.4423$74.70231@newsc.telia.net> <38E52716.A1C63B5D@roguewave.com>
Message-ID: <38E88E5F.324CFB99@infercor.no>

Bjorn Pettersen wrote:
> 
> Fredrik Lundh wrote:
> >
> > Bjorn Pettersen <bjorn at roguewave.com> wrote:
> > > Just read the announcement page, and found that socket.connect() no
> > > longer takes two arguments as was previously documented.
> >
> > note that the actual reference page says that connect takes
> > just one argument, and it has done so since at least 1.4.
> 
> The reference page says connect takes address (in italics), and goes
> onto explain what it _can_ be.  I took the example to be part of the
> documentation, and specifically explaining how to use connect with the
> example's socket type.

We work in an industry where documentation often lags behind the code -
something which frequently irritates me, but then I probably do my own bit to
uphold this dubious tradition. So, when documentation is available, as it is in
this case, the developer will often breathe a sigh of relief that they will not
be obliged to read the source code, and then rely on such documentation more
than they should (as it turns out).

Socket programming, especially for anyone who tried it in C, can be daunting to
those new to the subject and it is only natural that people will make the most
out of the examples provided in order to get started. Given that the example in
the library reference (even as recently as the one provided for 1.5.2)
contradicts the supposedly correct way of calling 'connect', we can seemingly
infer that either the intention was always to support this usage of the method
in question, or that the documentation is not to be relied upon. In my
experience, systems with inadequate or inaccurate documentation suffer from a
lack of acceptance as developers fight with the "real" behaviour of the code. I
can hardly believe, though, that the purpose of the documentation is to trick
people into using the Python libraries incorrectly.

Whether or not anyone believes that examples constitute definitive
documentation, the reactions of other contributors in this thread give a useful
insight into the way people (including myself in this case) pick up assumptions
about the way things work. Personally, I tend to test my assumptions in the
interpreter and then, if they behave as anticipated, I believe my assumptions to
be valid. Perhaps this approach is quite common in the Python community, but due
to a reliance on the behaviour of possibly only one implementation of the Python
"standard" it can also be dangerous. However, it is interesting to note that
JPython 1.0.3 seems to permit multiple argument 'connect' calls as well,
although I know that JPython 1.0.3 is not exactly "current".

Paul


From raffael at mediaone.net  Thu Apr 27 22:22:06 2000
From: raffael at mediaone.net (Raffael Cavallaro)
Date: Thu, 27 Apr 2000 22:22:06 -0400
Subject: The Simple Economics of Open Source
In-Reply-To: <1255272357-66924550@hypernet.com>
Message-ID: <B52E6E0E.521E%raffael@mediaone.net>

> The closed source world is based on a diseased model
> (pretending software is a product).

This "diseased model" has made billions for a number of firms, so, from an
economic standpoint, your argument rings rather hollow, especially when
compared to numerous open source ventures which have yet to post a cent of
profit.


> Scarcity of an intellectual secret may make a software
> package containing that secret more valuable, but scarcity of
> a software package is likely to make it less valuable (since
> support will be hard to find, and upgrades unlikely). OTOH it's
> scarcity of a piece of artwork that makes it valuable.

You're conflating separate markets. Unique artworks (paintings, for example)
have no real equivalent in the software world, because software sales are
based on essentially unlimited identical copies.

My example referred to IP in the art world - secret working methods that
allowed artists to produce works their contemporaries could not - methods of
bronze casting, for example, or formulas for oil painting media. These IPs
are like closed source in that, without them, competitors are shut out of a
lucrative market.

My point was, and is, that it makes no economic sense to open source an IP
that could be used to create a monopoly. MS understands this, and works
ceaslessly to create such monopolies. Note that this has *nothing
whatsoever* to do with what is best for consumers. Obviously, the consumer
is better off in an open source world for two main reasons:

1. Open source leads to commodity pricing, and hence, lower costs for
consumers. Closed source leads to monopoly pricing, and higher costs for
consumers.

2. Open source leads to more robust code, and hence more reliable products,
benefiting consumers. Closed source leads to slipshod QA due to the desire
to be first to market in order to establish the monopoly. This poor QA
obviously does not benefit the consumer.

I think part of the open source community's problem is this naive belief
that just because a certain model (open source) leads to products that
benefit consumers more, markets will allow consumers to choose this
production model. This belief is naive because it fails to consider that the
producers' interests are in no way congruent with the consumer's. The
producer has every reason to go closed source and attempt to secure a quick
monopoly. We live in this world now, and there's no reason to believe that
it's going to go away any time soon. I think open source advocates grossly
underestimate the power of monopolies, and the unwillingness of consumers to
break ranks even though it *may* provide long term benefits, because
breaking ranks entails large short term costs, especially in network utility
- i.e., the loss of interoperability with the rest of the monopolist's
customers.

BTW, all software is *not* support. This is another open source community
bias which comes from living in the server room. Lots of software is
*client* software, which requires little or no support from the vendor.
Instead, these support services are provided by full time IT people, who may
be certified by the vendor, but who are paid by the enterprise. In other
words, MS makes their money on enterprise licences for Office, not for
troubleshooting the enterprise's problems with Office. For the most
lucrative software businesses, software *is* a product, not a service.

Ralph
-- 

Raffael Cavallaro, Ph.D.
raffael at mediaone.net





From gherman at darwin.in-berlin.de  Tue Apr 18 07:01:07 2000
From: gherman at darwin.in-berlin.de (Dinu C. Gherman)
Date: Tue, 18 Apr 2000 13:01:07 +0200
Subject: Beginner: string to long
References: <8dgakr$2i$1@nnrp1.deja.com>
Message-ID: <38FC4073.B272F20F@darwin.in-berlin.de>

hudson_melville at my-deja.com wrote:
> 
> I'm a very beginner, and I have a program with
> nested loops that put numeric operators (+,-) in
> a string of numbers ('3564752617') The outputs
> are similar to '3564-752+617' How do I convert it
> so it can evaluate into a long integer? Is there
> a built-in that I don't know about that would do
> this?

eval('long(3564-752+617)')
3429L

Is this what you mean?

Dinu

-- 
Dinu C. Gherman
................................................................
"The thing about Linux or open software in general is that 
it actually tries to move software from being witchcraft to 
being a science," [...] "A lot of the programs you see today 
are actually put together by shamans, and you just take it and 
if the computer crashes you walk around it three times... and 
maybe it's OK." (Linus Thorvalds, LinuxWorld 2000, NYC)


From gmcm at hypernet.com  Tue Apr 11 16:10:02 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Tue, 11 Apr 2000 16:10:02 -0400
Subject: [ANNOUNCE] Garbage collection for Python
In-Reply-To: <38F3836A.5F9A7C6F@roguewave.com>
Message-ID: <1256638692-42316420@hypernet.com>

Bjorn Pettersen wrote:

> Helge Hess wrote:

> > Also note that not only the ++ or -- needs to be protected, but the
> > '==0' check as well. Eg:
> > 
> >    [task1]            [task2]
> >    dec(obj);          inc(obj);
> 
> Note that both task 1 and task two must have a reference to the object
> for this to be possible, so the ref count must at least be two. (unless
> you're thinking of sharing names between tasks without locking -- that
> would make rc much more resource intensive...)
> 
> > could be
> >    [task1]            [task2]
> >    obj->rc--;
> >    if (obj->rc==0)
> >      <task-switch>    obj->rc++;
> >      free(obj);       <task-switch>
> >    object-freed       assumes to have a valid obj
> 
> If the reference count was one, adding mutexes wouldn't help:
> 
>     aquire(obj->mutex)
>     obj->rc--;
>     if (obj->rc==0)
>       free(obj);   
>     release(obj->mutex)
> 
>                          aquire(obj->mutex)  // does it crash here
>                          obj->rc++           // or here?

The release would, except the entire scenario is hopelessly 
flawed (how does task2 get ahold of "obj" without either 
holding the lock or having the ref count already increased?).

[interlocked increment & friends]
> The api calls are direct wrappers around the processors
> increment-and-test or test-and-set instructions, and are extremely fast
> compared to a mutex solution (which makes it interesting to see other
> people on this thread claim that it would be _way_ to slow without even
> testing it...).  Our tests showed an order of magnitude difference on
> some tests...

The lightest weight synchronization primitives known to man. 
But a still a whole lot bulkier than a simple increment.

- Gordon



From arcege at shore.net  Fri Apr 21 08:31:17 2000
From: arcege at shore.net (Michael P. Reilly)
Date: Fri, 21 Apr 2000 12:31:17 GMT
Subject: How to convert GIF file(s) to MIME to be mail?
References: <38FF812F.2F34ACA1@nrlssc.navy.mil>
Message-ID: <pSXL4.781$lL2.92452@news.shore.net>

Paul Martinolich <martinol at nrlssc.navy.mil> wrote:
: I need to run a shell script that will convert a single or multiple
: GIF files into something I can mail to another user.

: How can python do it?

Yup, pretty easily too.  There are some nice standard modules to help
you with this.  I do it by hand often enough.

>>> import MimeWriter, base64
>>> infile = open('/tmp/world.gif', 'rb')
>>> outfile = open('/tmp/world.gif.b64', 'w')
>>> base64.encode(infile, outfile)
>>> outfile.close()
>>> infile.close()
>>> imagesize = os.path.getsize('/tmp/world.gif.b64')
>>> outmime = open('/tmp/outmail.txt', 'w')
>>> mw = MimeWriter.MimeWriter(outmime)
>>> mw.addheader('content-length', imagesize)
>>> f = mw.startbody('image/gif', [('name', 'world.gif')])
>>> outfile = open('/tmp/world.gif.b64', 'r')
>>> f.write(outfile.read())
>>> outmime.close()
>>>

Myself, I always thought the MimeWriter module was overly complicated
(especially linking it with the mimetools module) and I wrote another
module for my own uses called mimecntl.

>>> import mimecntl
>>> imagefname = '/tmp/world.gif'
>>> imagedata = open(imagefname, 'rb')
>>> md = mimecntl.MIME_recoder(image, ctype='image/jpeg')
>>> encoded = md.encode('base64')
>>> encoded['content-length'] = len(encoded) # encoded data size
>>> outmime = open('/tmp/outmail.txt', 'w')
>>> outmime.write( str(encoded) )
>>> outmime.close()
>>>

Making multipart MIME messages instead is as easy as making a list of
them:
>>> second_encoded = mimecntl.MIME_recoder(
...   image2, ctype='image/jpeg').encode('base64')
>>> second_encoded['content-length'] = len(second_encoded)
>>> msg = mimecntl.MIME_document( [encoded, second_encoded] )
>>> outmime = open('/tmp/outmime.txt', 'w')
>>> outmime.write( str(msg) )
>>> outmime.close()

You can find the mimecntl module at
<URL: http://www.shore.net/~arcege/python/mimecntl.py>.

  -Arcege



From johngrayson at home.com  Mon Apr 24 11:02:59 2000
From: johngrayson at home.com (johngrayson at home.com)
Date: Mon, 24 Apr 2000 15:02:59 GMT
Subject: Tkinter bind of class -- help
References: <39027BE6.8DF89328@flash.net>
Message-ID: <8e1nmq$ki2$1@nnrp1.deja.com>

In article <39027BE6.8DF89328 at flash.net>,
  Cindy Huyser <chuyser at flash.net> wrote:
> I've created a class that contains a number of Canvas widgets inside
> frames.  When I import this class and instantiate it in my main
program
> I put it inside a frame.
>
> What I would like to do is to get x,y coordinate information on events
> (either <Button-1> or <Motion>) that occur in the area occupied by the
> class instance so that I can display detail information in another
area
> of the main window.  At first, I tried to use bind() on the class
> instance, but received an error.  I have tried callbacks from the
frame
> in which the object resides, but find I only get event information for
> that area unoccupied by the object; I have tried looking at
information
> stored in a member variable of the class, but have found the response
> unacceptably slow.
>
> Any suggestions for solving this problem would be greatly appreciated
by
> this newbie Tkinter programmer.
>

Without seeing your code, it is a little difficult to
see what you're trying to achieve. However, here is
some sample code that does something along the lines
of your explanation:

from Tkinter import *

root = Tk()
root.title('Canvas Info')

class MakeWindows:
    def __init__(self, parent):

	f1 = Frame(root)
	Label(f1, text='x,y: ').pack(side=LEFT, padx=10)
	self.coord = StringVar()
	Entry(f1, textvariable=self.coord).pack(side=LEFT, anchor=W)
	f1.pack(padx=10, pady=10)

	self.subWin = Toplevel(root, width=500, height=500)
	self.grid = Frame(self.subWin)

	for row in range(5):
	    for col in range(5):
		f = Frame(self.grid, borderwidth=2, relief=GROOVE)
		item = Canvas(f, width=90, height=90)
		item.bind('<Motion>', lambda e, s=self, i=item:
                           s.info(e, i))
		item.pack(fill=BOTH, expand=YES)
		f.grid(row=row, col=col, sticky=NSEW)
        self.grid.pack()

    def info(self, event, item):
	print 'Canvas ID =', item
	self.coord.set('%d:%d' % (event.x, event.y))

c=MakeWindows(root)

root.mainloop()


Sent via Deja.com http://www.deja.com/
Before you buy.


From rcc at nospamthanks_jennic.com  Fri Apr 14 12:07:46 2000
From: rcc at nospamthanks_jennic.com (Robert Cragie)
Date: Fri, 14 Apr 2000 16:07:46 GMT
Subject: List mapping?
References: <Pine.LNX.3.96.1000413174501.3264B-100000@valdez.netg.se> <newscache$autysf$ye8$1@jenpc07.jennic.co.uk> <slrn8fc4eu.1d0.scarblac-spamtrap@flits104-37.flits.rug.nl>
Message-ID: <newscache$swk0tf$4lh$1@jenpc07.jennic.co.uk>

Remco Gerlich <scarblac-spamtrap at pino.selwerd.nl> wrote in message
news:slrn8fc4eu.1d0.scarblac-spamtrap at flits104-37.flits.rug.nl...
| Robert Cragie wrote in comp.lang.python:
| > > Howcome you use reduce at all?
| > >
| > > Works fine with just doing
| > >
| > > tuplist = map(lambda x: (mylist[x], mylist[x+1]), range(0,
len(mylist),
| > 2))
| >
| > Yep, that's neater. My excuse is that I've only been learning Python for
a
| > couple of weeks...
|
| In that case, forget you ever read something about map, lambda and reduce.
| Save it for later and code nice for loops now. 95% of the time, those
things
| only obfuscate code anyway.

'map' and 'reduce', perhaps. I think 'map' is syntactically neat for some
things, and not too confusing.
'lambda' is very useful, but, I think, quite confusing in its use, which is
partly why I did not see the simpler solution.

Robert Cragie





From tim_one at email.msn.com  Sun Apr  2 01:18:21 2000
From: tim_one at email.msn.com (Tim Peters)
Date: Sun, 2 Apr 2000 01:18:21 -0500
Subject: Debugging confusion -- too many stacks!
In-Reply-To: <7cjF4.60441$3b6.236709@ozemail.com.au>
Message-ID: <000a01bf9c6b$3f80c2a0$752d153f@tim>

[Jason Stokes]
> I'm confused about the difference between the C stack, the Frame
> stack, and the stack *inside* each execution frame.

The difference between the first two is an artifact of the current
implementation; indeed, getting rid of the distinction is the major point of
Christian Tismer's "Stackless" Python variant (where "less" refers to
getting rid of the C-stack component).

> From reading the source, I believe that each new code object gets a new
> frame allocated on the frame stack,

Each *invocation* of a code object, yes; and, e.g., if a function calls
itself recursively, there's only one code object but a separate frame for
each call level.

Note that in CT's Stackless Python, the frames actually form a tree
(although at any leaf, the path back to the root is unique and can be viewed
as a stack).

> within which is stored all kinds of useful information about the
> context the code is executing in -- the global and local environment,
> a tuple of constants, a tuple of arguments etc.  *Within* each frame
> is *another* stack, upon which the Python virtual machine loads and
> manipulates intermediate values.

This is another artifact of the current implementation, and is best viewed
as an internal detail of no visible consequence.

> However, I'm not sure what the terminology is to refer to this
> third stack.

"The frame's eval stack" works for me <wink>.

> Understanding the distinction seems to be vital to using pdb, though.

I don't use pdb much, but don't see why this would be true.  The frame's
eval stack is invisible, and the C and frame stacks happen to be intertwined
one-for-one today because a Python call happens to invoke the interpreter
recursively (at the C level) today.  Conceptually, there's only one stack
involved in calls.

Perhaps you could be specific about what in pdb is confusing you, and
someone who uses it could straighten it out.

or-maybe-not-ly y'rs  - tim





From jvickroy at sec.noaa.gov  Mon Apr 17 09:36:53 2000
From: jvickroy at sec.noaa.gov (j vickroy)
Date: Mon, 17 Apr 2000 07:36:53 -0600
Subject: read/write file problem
Message-ID: <38FB1375.2FF65B96@sec.noaa.gov>

Hello all,

I am attempting to read and, subsequently, write a jpeg file.
The following interactive session illustrates the problem I am having
(i.e., if the source file is read byte-by-byte into memory and
subsequently
rewritten using the file.write() method, only the initial portion of the

file appears to be written).
How should I be doing this in Python 1.5.2.

Thanks for your time.


>>> import os.path
>>> file_name = 'd:/p1010001.jpg'
>>> source_byte_cnt = os.path.getsize (file_name)
>>> print source_byte_cnt
381965
>>> buf = []
>>> file = open (file_name, 'r')
>>> for i in xrange (0,source_byte_cnt):
...  buf .append (file .read(1))
...
>>> file .close()
>>> print len (buf) == source_byte_cnt
1
>>> file_name = 'd:/test.jpg'
>>> file = open (file_name, 'w')
>>> for i in xrange (0,source_byte_cnt):
...  file .write (buf[i])
...
>>> file .close()
>>> target_byte_cnt = os.path.getsize (file_name)
>>> print target_byte_cnt
70
>>> file = open (file_name, 'w')
>>> for i in xrange (0,source_byte_cnt):
...  file .write (str(buf[i]))
...
>>> file .close()
>>> target_byte_cnt = os.path.getsize (file_name)
>>> print target_byte_cnt
70
>>>




From paul.magwene at yale.edu  Mon Apr 17 16:23:12 2000
From: paul.magwene at yale.edu (Paul Magwene)
Date: Mon, 17 Apr 2000 16:23:12 -0400
Subject: Handling errors in open
References: <38FB4D0A.489FF50F@ecs.soton.ac.uk>
Message-ID: <38FB72B0.D929B73F@yale.edu>

Jacek Generowicz wrote:
> 
> Hi,
> 
> Is there a way to prevent a Python script from
> terminating when `open' fails, and have it handle
> the error more helpfully . . . I was thinking
> along these lines:
> 
> from sys import argv
> 
> try:
>     file = open(argv[1],"r")
> except IOerror:
>     print "You lied about that file."
> 
> Jacek


Yes, that works...? (slightly confused because you seem to have found
your own answer, yet you still posted the question.  Are you looking for
alternative ways to do this?)

--Paul


From amused at webamused.com  Tue Apr 25 22:00:08 2000
From: amused at webamused.com (Joshua Macy)
Date: Wed, 26 Apr 2000 02:00:08 GMT
Subject: mxODBC error trapping
References: <A63AFB20111ED311AC5500A0C96B7BF5015F0F4D@orsmsx36.jf.intel.com>
Message-ID: <390640A8.F1392B0F@webamused.com>

"Daley, MarkX" wrote:
> 
>
> Traceback (innermost last):
>   File "<pyshell#1>", line 1, in ?
>     collect.collect()
>   File "C:\PROGRA~1\Python\collect.py", line 50, in collect
>     except dbi.OperationalError:
> AttributeError: OperationalError
> 
> Here is the printout of the dbi module:
> 
> ['DATE', 'NUMBER', 'RAW', 'ROWID', 'STRING', 'TYPES', '__doc__', '__file__',
> '__name__', 'dataError', 'dbDate', 'dbRaw', 'dbiDate', 'dbiRaw',
> 'integrityError', 'internalError', 'noError', 'opError', 'progError']
> 
>


  According to the dbi module dictionary (which you've printed), it
looks like it should be opError, not OperationalError.  Or you could
just use an unqualified except, and be sure of catching any error that
the dbi can raise.

 Joshua


From lf11 at deletethistomailme.andonet.com  Tue Apr  4 16:39:41 2000
From: lf11 at deletethistomailme.andonet.com (lf11 at deletethistomailme.andonet.com)
Date: Tue, 04 Apr 2000 15:39:41 -0500
Subject: GTK Referenc
References: <38E7A868.A1C9BD7C@searchy.net><m2snx4utnn.fsf@pchoel.privat.uni-hamburg.de>
Message-ID: <8cdjcb$g8h$1@news.kersur.net>

In article <m2snx4utnn.fsf at pchoel.privat.uni-hamburg.de>, 
bhoel at starship.python.net (Berthold H?llmann) wrote:
> Penpal International <ppi at searchy.net> writes:
> 
>> I couldn't find anywhere a reference for the GTK module for Python. Does
>> anyone know one?
>> On gtk's site I could only find some for C(++).
> 
I don't know how experienced you are with Python.  I started playing 
with PyGTK, having absolutely no previous experience with Python.  
That method of learned is Not Recommended.

If you 'import gtk' instead of 'from gtk import *', you will import the 
complete function names.  The syntax of this is _very_similar_ to C GTK.  
In fact, it's almost identical.

Example with 'import gtk' (looks like C)
button = gtk_button_new_with_label('Button')

Example with 'from gtk import *' (faster to type, more Python-ish)
button = GtkButton('Button')

Both do the same thing.  I would recommend using 'import gtk' until you 
know more about it.  Since the syntax is so similar between Python and C, 
I simply use the C GTK reference from www.gtk.org.

A bunch of examples come with PyGTK, notably "testgtk.py".  I learned 
a lot (and am learning a lot) from that script.

Hope this helps,

 -lf


From sblakey at freei.com  Thu Apr 27 19:03:24 2000
From: sblakey at freei.com (Sean Blakey)
Date: Thu, 27 Apr 2000 23:03:24 +0000
Subject: run detached
References: <3908B8CB.97A0755D@inka.de>
Message-ID: <3908C73C.420BF3B7@freei.com>

Michael Str?der wrote:
> 
> HI!
> 
> How can I automatically detach from console after starting up with
> SocketServer? (E.g. web server with access and error logging to
> files.)
> 
> Ciao, Michael.
> --
> http://www.python.org/mailman/listinfo/python-list
I think what you want is a daemon process (I assume you are looking for
the Unix answer since you are posting from a linux email client).

Starting a daemon requires the same steps in any language:
1. Fork from the parent tty.
2. Exit the parent process (all further steps occur in the child
process).
3. Call setsid() to put your child process in a new process group.
4. Redirect stdin, stdout, and stderr to /dev/null (or a file of your
choice)
5. chdir to / (not strictly necessarry, but polite).
In python , it looks like this (assuming main() is your code):

import os
import sys
if os.fork():
	sys.exit(0)	# In parent process
else:
	os.setsid()
	sys.stdin = open('/dev/null')
	sys.stdout = open('/dev/null', 'w')
	sys.stderr = open('/dev/null', 'w')
	os.chdir('/')
	main()

I have been playing with creating an object wrapper for daemon
processes, and this may be useful to you (it's still pretty raw).  If
you want, you can check it out at http://members.home.com/seanb/python/
-- 
Sean Blakey
FreeInternet.com
sblakey at freei.com
(253)796-6500x1025



From infotechsys at pivot.net  Sun Apr 23 11:00:51 2000
From: infotechsys at pivot.net (Wayne)
Date: Sun, 23 Apr 2000 11:00:51 -0400
Subject: Off topic - But would like tour advice.
Message-ID: <39031023.B12ABF56@pivot.net>

Hi,
I'm using Python with Tkinter  to gather information pertaining to a
invoice.
I learnt groff/troff with tbl, pic and eqn to print said invoice. I will
be going to XML
and dsssl using jade but time is against me right now. My plan for
printing
the invoice is to take the groff file, which is a template with lots of
tags in the invoice ,
 and pass it thru re, substuting  the values I gather from the GUI
interface
into the template replacing the tags. Then use the os module to print
the invoice. Is this
a method that would be used to print the invoice? Or is there a standard
way of doing
something like this?
TIA
Wayne

P.S.   I'm using pic to design my invoice.




From MSteed at altiris.com  Fri Apr  7 15:14:45 2000
From: MSteed at altiris.com (Mike Steed)
Date: Fri, 7 Apr 2000 13:14:45 -0600 
Subject: anyone else wrap gd-1.8?
Message-ID: <65118AEEFF5AD3118E8300508B1248774CB282@ALTNET>

On a semi-related note, I have wrapped (without SWIG) gdchart, a library for
making charts and graphs, which uses gd.  I have a new version (to be posted
Real Soon Now) that uses gd-1.8, with support via the Python interface for
PNG and JPEG output.  There is an entry in the Vaults of Parnassus.

Mike
msteed at altiris.com

> From: musingattheruins at my-deja.com 
> [mailto:musingattheruins at my-deja.com]
> Sent: Friday, April 07, 2000 12:57 PM
> To: python-list at python.org
> Subject: Re: anyone else wrap gd-1.8?
> 
> 
> Yeah, I wrapped it but did not use SWIG so it took longer than 1.25
> hours.  It's available with source for Win32 at
> http://musingattheruins.homepage.com/PythonGD.html
> 
[...]
> 
> 
> In article <m2wvn9pijt.fsf at netmeg.net>,
>   Les Schaffer <godzilla at netmeg.net> wrote:
> > I just wrapped the latest gd module -- version 1.8 which now handles
> > both png and jpeg file formats -- into python. It was quite 
> easy with
> > SWIG.
> >
> > I noticed the gdmodule link on
> >
> > http://www.python.org/topics/scicomp/plotting.html
> >
> > is dead. has anyone else wrapped gd up lately? otherwise i 
> am happy to
> > distribute (all 1.25 hours of my accumulated time :-)
[...]



From gmcm at hypernet.com  Sun Apr 23 09:11:20 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Sun, 23 Apr 2000 09:11:20 -0400
Subject: code not true?
In-Reply-To: <Pine.BSF.4.10.10004230244460.5642-100000@chi.pair.com>
References: <Z%vM4.1491$RM6.3518433@news-east.usenetserver.com>
Message-ID: <1255627015-45592331@hypernet.com>

Michal Wallace wrote:

> On Sun, 23 Apr 2000, Kragen Sitaker wrote:
> > That said, it *is* just a speed issue --- there are representations
> > other than floating point, like the ordered-pair rational-number
> > representation, that give the right answer but are slower to work
> > with.
> 
> But... Python is a scripting language.. It's okay to be slow.  

It's not just a scripting language. Because numbers are 
objects (not primitives), Python is already slow in doing math, 
and the cumulative complaints about that are voluminous.

> From
> what I'm hearing, many people on this list shy away from floats simply
> because of the innacuracy.. Why don't we use the way that gets the
> right answer?

Some tech magazine columnist observed that there are 3 
types of float users: 
 1) those who think it's easy, 'cause they're just like ints
 2) those who think it's hard, 'cause they're nothing like ints
 3) those who really know.

We've got exactly one 3 (and he's busy - having just bought a 
CDROM with 660M of one-time pad <wink>). The transition 
from 1 to 2 is hard (as witnessed by this thread). But once 
you're a 2, it's not that hard to learn to use floats sanely.

The theoretically correct alternative is rationals. But they're 
slow and use unpredictable amounts of memory. I think 
Python should have BCD as an alternative. While slow, they're 
predictable in memory use, and the mathematical 
shortcomings they have are exactly those that a naive user 
expects.

But I, despite being in category 2 for close to 20 years, 
definitely don't want to be without floats.



- Gordon



From mwh21 at cam.ac.uk  Tue Apr 11 07:22:00 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 11 Apr 2000 12:22:00 +0100
Subject: Modifying "traceback" objects
References: <20000406075124.A1848828@vislab.epa.gov>
Message-ID: <m33doslw6f.fsf@atrus.jesus.cam.ac.uk>

Randall Hopper <aa8vb at yahoo.com> writes:

>      Is it possible to add frames to the current traceback object from
> inside a C extension?

I'd say it would be difficult.  What would you put in the "f_code" field?

Would

PyThreadState *tstate = PyThreadState_GET();
PyFrameObject* f = PyFrame_New(
		tstate,			/*back*/
		co,			/*code*/
		globals,		/*globals*/
		locals);		/*locals*/
if (f == NULL)
	return NULL;

tstate->frame = f;

be sufficient? (If you can cook up values for co, globals & locals).

Maybe:

int
PushFrame(char* func_name,char* filename,int lineno)
{
    PyThreadState *tstate = PyThreadState_GET();
    PyFrameObject* f;
    PyObject *emptytuple;
    PyObject *emptydict;
    PyObject *co;

    emptydict = PyDict_New();
    emptytuple = PyTuple_New(0);

    co = PyCode_New(0,0,0,0,"",
                    emptytuple,emptytuple,emptytuple,
                    filename,func_name,lineno,"");

    if (co == NULL) return 0;

    f = PyFrame_New(
                tstate,                 /*back*/
                co,                     /*code*/
                emptydict,              /*globals*/
                emptydict);             /*locals*/

    Py_DECREF(emptydict);
    Py_DECREF(emptytuple);
    Py_DECREF(emptytuple);

    if (f == NULL) return 0;

    tstate->frame = f;

    return 0;
}

to be used like:

    ...
    if (!PushFrame("frob",__FILE__,__LINE__)) return NULL;
    ...

though the above is monumentally untested (does it compile? I don't
know!).

Cheers,
M.

-- 
  On the other hand,  the following areas are subject to boycott
  in reaction to the rampant impurity of design or execution, as
  determined after a period of study, in no particular order: 
    ...                            http://www.naggum.no/profile.html


From effbot at telia.com  Wed Apr 26 13:42:41 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Wed, 26 Apr 2000 17:42:41 GMT
Subject: Simple del[] list question
References: <meh9-9D8292.13251626042000@news.cit.cornell.edu>
Message-ID: <lUFN4.1711$wYl.196641792@newsb.telia.net>

Matthew Hirsch <meh9 at cornell.edu> wrote:
> What is going on here?
>
> This leaves a intact.
>
> >>> a=[[1,2,3],[4,5,6]]
> >>> b=a[:]
> >>> del b[0]
> >>> b
> [[4, 5, 6]]
> >>> a
> [[1, 2, 3], [4, 5, 6]]
>
>
> This also changes a.
>
> >>> a=[[1,2,3],[4,5,6]]
> >>> b=a[:]
> >>> del b[0][0]
> >>> b
> [[2, 3], [4, 5, 6]]
> >>> a
> [[2, 3], [4, 5, 6]]
>
> Why did a change?  Why suddenly when you go to two dimensions does the
> behavior change?

[:] makes a *shallow* copy.  it doesn't copy the sublists.

</F>




From aa8vb at yahoo.com  Fri Apr 14 08:40:23 2000
From: aa8vb at yahoo.com (Randall Hopper)
Date: Fri, 14 Apr 2000 08:40:23 -0400
Subject: Python idiom: Multiple search-and-replace
In-Reply-To: <0s1J4.655$rc9.190631424@newsb.telia.net>
References: <20000412100816.A2431031@vislab.epa.gov>
 <5C%I4.645$rc9.190288896@newsb.telia.net>
 <20000412120012.D2446939@vislab.epa.gov>
 <0s1J4.655$rc9.190631424@newsb.telia.net>
Message-ID: <20000414084023.A2012@vislab.epa.gov>

Fredrik Lundh:
 |Randall Hopper <aa8vb at yahoo.com> wrote:
 |> Thanks!  It's much more efficient.  The 140 seconds original running time
 |> was reduced to 11.6 seconds.  I can certainly live with that.
 |
 |thought so ;-)
 |
 |while you're at it, try replacing the original readline loop with:
 |
 |    while 1:
 |        lines = fp.readlines(BUFFERSIZE)
 |        if not lines:
 |            break
 |        lines = string.join(lines, "")
 |        lines = re.sub(...)
 |        out_fp.write(lines)
 |
 |where BUFFERSIZE is 1000000 or so...

Elapsed: 2.7118 sec.  76% speedup.  98% speedup overall.

Thanks again Fredrik!

It's hard to believe we're burning 8.9 sec out of 11.6 seconds (77%) just
in the simple loop overhead of iterating over lines.  Where's that JIT
Python interpreter when you need it ;-)  

Seriously, it would be pretty spiffy if Python could compile the inner loop
to machine code on the Nth iteration (N is some bite-the-bullet threshold),
and reuse it for all subsequent iterations.  I know the dynamic nature of
the language complicates the analysis, but if it detects no strange
rebinding (string.join being redefined, etc.), seems like it'd be a real
win.

-- 
Randall Hopper
aa8vb at yahoo.com



From jbauer at rubic.com  Tue Apr  4 20:16:39 2000
From: jbauer at rubic.com (Jeff Bauer)
Date: Tue, 04 Apr 2000 19:16:39 -0500
Subject: Starving for an Advanced Python Book
References: <20000404231358.17447.qmail@web2105.mail.yahoo.com> <38EA7EE1.457A5533@roguewave.com>
Message-ID: <38EA85E7.38C7BACE@rubic.com>

Bjorn writes a glowing recommendation of eff-bot's book:
> At $12.00 I would say the book is a steal and 
> contains a wealth of information no serious Python 
> programmer should be without :-)  And if you're 
> like me, you'll print it out, bind it at Kinko's,
> and place it half way between your Linux and win2k
> boxes <wink>.

I would second Bjorn's recommendation.  It's a useful
book for anyone who writes Python code and wants quick 
access to library examples.

That said, the electronic distribution mechanism
will try your patience.  When you get it to work, 
the first thing you'll want to do is follow Bjorn's
advice and print it.

purchasing-the-book-is-like-taking-a-test-but-it's-
  all-downhill-from-there-ly y'rs,

Jeff Bauer
Rubicon



From wware at world.std.com  Sat Apr 15 08:32:17 2000
From: wware at world.std.com (Will Ware)
Date: Sat, 15 Apr 2000 12:32:17 GMT
Subject: [ANNOUNCE] Garbage collection for Python
References: <20000407221750.A29279@acs.ucalgary.ca> <38F826CF.DBDC71EF@zipzap.ch>
Message-ID: <Ft25Hu.MG4@world.std.com>

Boris Borcic (zorro at zipzap.ch) wrote:
> Just van Rossum wrote:
> > 
> > - Will Ware and I have worked hard on a complete rewrite of the
> >   uthread module. It is a *vast* improvement over the old one.
> >   It offers a very nice thread object and several handy threading
> >   widgets. It already can do stuff threading.py can only *dream*
> >   about, like pausing (blocking) a thread from another thread or
> >   posting an exception from another thread (!).

> Simple examples of useful new idioms that this permits ?

As far as programming idioms, you can get a sense of some of them
from the manual for the previous version of microthreads:
http://world.std.com/~wware/uthread.html

The examples in the manual run with almost no changes in the new
version. The timer has changed a bit, it now counts real milliseconds
rather than the fairly arbitrary context switches. There are also a
few new widgets that aren't in the manual yet.

As far as real applications, Just's MacPython IDE is the only real
one thusfar. I've been thinking about doing a Pacman game (one
microthread for each moving monster, and one for the Pacman) but I
haven't found the time yet. My original interest in microthreads was
to support simulations with large numbers of simultaneously interacting
agents (without having to write each agent's behavior in some ugly
form with explicit context switching).
-- 
 - - - - - - - - - - - - - - - - - - - - - - - -
Resistance is futile. Capacitance is efficacious.
Will Ware	email:    wware @ world.std.com


From rumjuggler at home.com  Wed Apr 26 18:39:10 2000
From: rumjuggler at home.com (Ben Wolfson)
Date: Wed, 26 Apr 2000 22:39:10 GMT
Subject: first non-trivial python program
Message-ID: <fbsegsgbqq6803j5n54bvo2d3gcna93sl8@4ax.com>

Is a translation and expansion of a C++ program I wrote to learn about
the STL and generally expand my C++-ish knowledge.  Anyway, the Python
version is 700+ lines long; would it be considered rude to post it
here for helpful advice?

-- 
Barnabas T. Rumjuggler

A cucumber should be well sliced, and dressed with pepper and vinegar,
and then thrown out, as good for nothing.
 -- Samuel Johnson


From aahz at netcom.com  Sat Apr 15 17:55:53 2000
From: aahz at netcom.com (Aahz Maruch)
Date: 15 Apr 2000 21:55:53 GMT
Subject: [ANNOUNCE] Garbage collection for Python
References: <20000407221750.A29279@acs.ucalgary.ca> <87em8ccywq.fsf@deneb.cygnus.argh.org> <8d01iu$ca0$1@slb2.atl.mindspring.net> <kur9cah4m6.fsf@lasipalatsi.fi>
Message-ID: <8daoh9$kt8$1@slb7.atl.mindspring.net>

In article <kur9cah4m6.fsf at lasipalatsi.fi>, Erno Kuusela  <erno at iki.fi> wrote:
>>>>>> "Aahz" == Aahz Maruch <aahz at netcom.com> writes:
>
>    >>  Yes.  AFAIK, Python doesn't perform well in a SMP environment,
>    >> because the interpreter only runs on one CPU at a given time,
>    >> even if you create multiple Python threads.
>
>    Aahz> That is true if you stick to pure Python code and extensions
>    Aahz> that don't release the global interpreter lock.  For
>    Aahz> example, if you're using HTTP, the socket library releases
>    Aahz> the global lock and threads are extremely effective.
>
>this won't help much if you would like your code to run on multiple
>cpu:s at once, since socket and file io are not cpu intensive.
>
>i bet switching to another cpu for doing the socket call while letting
>other python threads execute pyth code would decrease performance,
>as moving the thread to another cpu probably costs more than a
>socket syscall.

You don't "move" the thread to a different CPU; the OS handles the
thread scheduling.  And while what you say is true for a single socket
call, it's much less true for a group of fifty or a hundred threads each
doing intensive socket calls.

Also, while I'm talking about the easiest way to get useful threading
out of Python, do keep in mind that *any* extension that releases the
global lock can process in parallel.
--
                      --- Aahz (Copyright 2000 by aahz at netcom.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

Why doesn't "Just Say NO" include caffeine, nicotine, alcohol, Prozac,
and Ritalin?  --Aahz


From greg at cosc.canterbury.ac.nz  Wed Apr 19 23:57:38 2000
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Thu, 20 Apr 2000 15:57:38 +1200
Subject: Mac Python Questions
References: <IRjL4.271$CP3.3183@news.uswest.net>
Message-ID: <38FE8032.DE233C29@cosc.canterbury.ac.nz>

Jeff Massung wrote:
> 
> b) No matter what I do, I can't get Tkinter to work :(.

Neither can I, and from what I've seen here, neither
can anyone else. Tkinter just seems to be broken on
the Mac at present.

-- 
Greg Ewing, Computer Science Dept,
+--------------------------------------+
University of Canterbury,	   | A citizen of NewZealandCorp, a	  |
Christchurch, New Zealand	   | wholly-owned subsidiary of USA Inc.  |
greg at cosc.canterbury.ac.nz	   +--------------------------------------+


From quinn at groat.ugcs.caltech.edu  Mon Apr  3 06:02:06 2000
From: quinn at groat.ugcs.caltech.edu (Quinn Dunkan)
Date: 3 Apr 2000 10:02:06 GMT
Subject: Pass by reference ?
References: <38E864C9.E1EE94F7@ecs.soton.ac.uk>
Message-ID: <slrn8egr51.pe.quinn@groat.ugcs.caltech.edu>

On Mon, 03 Apr 2000 10:30:49 +0100, Jacek Generowicz <jmg at ecs.soton.ac.uk>
wrote:
>Hi,
>
>I'm just trying to familiarize myself with Python.
>In Magnus Lie Hatland's Instant Python, he
>mentions that `all parameters in Python are passed
>by reference'.
>
>So I tried the following code:
>
>#!/usr/bin/python
>
>def increment(x):
>    x = x + 1
>    return x

This makes a *new* local x, which disappears when it goes out of scope.
Whenever you write "<var> = <blah>" python *always* makes a new local binding
of blah to var, unless you've got a "global <var>" up there somewhere.

You can't modify numbers in python anyway (they're immutable, just like
strings).  Try it with a mutable object like a list to see 'pass by reference'
behaviour:

>>> def addelt(lst, x):
...     lst.append(x)
...
>>> a = [0, 1]
>>> addelt(a, 5)
>>> a
[0, 1, 5]


From gmcm at hypernet.com  Wed Apr 26 15:27:23 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Wed, 26 Apr 2000 15:27:23 -0400
Subject: Simple Function Question
In-Reply-To: <8e7dh3$1gc@autodesk.autodesk.com>
Message-ID: <1255345248-62540224@hypernet.com>

Akira Kiyomiya wrote:

> Hi,  Could you explain this in dummy's language?
[snip]
> # Return a function that returns its argument incremented by 'n'
> def make_incrementer(n):
>     def increment(x, n=n):
>         return x+n
>     return increment
> 
> add1 = make_incrementer(1)
> print add1(3)  # This prints '4'

When you call make_increment, the inner "def" is executed; 
(the "def" for make_increment was executed once, when the 
module / script was first loaded). The inner "def" uses a 
default argument (evaluated when make_increment is called) 
to store the value of n in the newly created function object.

If you execute add2 = make_increment(2), you'll find that 
add1.__name__ and add2.__name__ are both "increment", 
but they are different function objects, and the name 
"increment" is actually a local name to make_increment, (it 
doesn't make it to the global namespace).

- Gordon



From garabik at melkor.dnp.fmph.uniba.sk.spam  Mon Apr  3 07:43:44 2000
From: garabik at melkor.dnp.fmph.uniba.sk.spam (Radovan Garabik)
Date: 3 Apr 2000 11:43:44 GMT
Subject: Python 1.6 alpha 1 released
References: <200003312130.QAA04361@eric.cnri.reston.va.us> <38E5205F.DE811F61@roguewave.com> <Wf9F4.4423$74.70231@newsc.telia.net> <38E58D2D.384F1546@Lugoj.Com> <S9kF4.4472$74.71414@newsc.telia.net> <38E6B929.28816494@roguewave.com> <auGF4.914$n68.186931712@newsb.telia.net>
Message-ID: <954762216.922508@cdwork.cvt.stuba.sk>

Fredrik Lundh <effbot at telia.com> wrote:
 : Bjorn Pettersen wrote:
 :> I haven't seen anyone claim that examples in the manual are not
 :> documentation...

 : in earlier releases of the library reference, some examples
 : didn't even work.  you still think they're formally binding?

 :> It would be very nice if "someone" could do that, but based on the
 :> rethoric I've seen from you on this issue, and the absence of comment
 :> from Guido, I certainly don't feel confident that it would be accepted.

 : you don't know that until you've tried it...

 : you may have to work on your presentation, though. thinking
 : about the following may help you come up with a stronger case:

 : -- how does typical socket.connect code look?

s.connect(addr, port)

 : -- what exception do you get after this change?  how
 :    does the error message look?  can anyone else figure
 :    out what's wrong?

 : -- is there any chance that people might be expecting
 :    the resulting exception from socket.connect (or the
 :    code invoking it)?  or in other words, is it likely that
 :    there will be a try/except clause around the call, that
 :    will misbehave (thus hiding the actual error message)?

well, I am bitten by this....
I use try/except around socket.connect, and this change would break some of
my scripts (but fixin them takes about 2 minutes, so I am not complaining
much)


 : -- can anyone but the original author fix the problem
 :    without breaking the program?

write a wrapper function around socket.socket.... not too difficult

 : when you've done this, replace "socket.connect" with
 : "list.append", and try again.

I never used multiple arguments in list.append...



-- 
 -----------------------------------------------------------
| Radovan Garabik http://melkor.dnp.fmph.uniba.sk/~garabik/ |
| __..--^^^--..__    garabik @ melkor.dnp.fmph.uniba.sk     |
 -----------------------------------------------------------
Antivirus alert: file .signature infected by signature virus.
Hi! I'm a signature virus! Copy me into your signature file to help me spread!


From sheacouture at hotmail.com  Mon Apr 10 00:17:02 2000
From: sheacouture at hotmail.com (SheaMon)
Date: Sun, 09 Apr 2000 21:17:02 -0700
Subject: need help just begining
Message-ID: <11822ca0.dd34a400@usw-ex0109-068.remarq.com>

Hi everyone,
My name is Shea and I am just starting out on programming.
I know not one thing about it, and a person told me to try
Python programming. But I have not idea what to do to start
making a program.

I main question is this. If i got a copy of this tutor can
I just leave the numbers and stuff like that and just edit
the rest? I get so lost in the number part of the
programming. If anyone here has some time can you please
help me and email me at my address here.

thanks.
Shea
SheaCouture at hotmail.com


* Sent from AltaVista http://www.altavista.com Where you can also find related Web Pages, Images, Audios, Videos, News, and Shopping.  Smart is Beautiful


From scarblac-spamtrap at pino.selwerd.nl  Fri Apr 28 07:58:56 2000
From: scarblac-spamtrap at pino.selwerd.nl (Remco Gerlich)
Date: 28 Apr 2000 11:58:56 GMT
Subject: Do I always have to write "self." ?
References: <270420001706386648%pecora@anvil.nrl.navy.mil> <slrn8ghc6t.f01.scarblac-spamtrap@flits104-37.flits.rug.nl> <280420000746597329%pecora@anvil.nrl.navy.mil>
Message-ID: <slrn8giv6f.fl9.scarblac-spamtrap@flits104-37.flits.rug.nl>

Louis M. Pecora wrote in comp.lang.python:
> > You *can* use another word for 'self' though, 's' or 'me' or something. But
> > it makes your code harder to read, of course...
> 
> You mean by doing something like
> 
> s=self

No, you can just declare methods like

class Foo:
   def Method(s,x):
      s.y = x

The name 'self' is just a convention.

-- 
Remco Gerlich,  scarblac at pino.selwerd.nl
This space intentionally left blank.


From cbwles at hotmail.com  Mon Apr 17 13:36:22 2000
From: cbwles at hotmail.com (cbwles at hotmail.com)
Date: Mon, 17 Apr 2000 17:36:22 GMT
Subject: BLUEMOUNTAIN SUCKS!!!!  7474
Message-ID: <qYHK4.14006$iO6.4836@typhoon1.gnilink.net>

Bluemountains cards are all the same.  Boring.  I like www.cardgirl.com.   I think the cards are much better.  I dont get how cardgirl.com keeps it totally free.  They dont even have those annoying banners.  
   If anyone knows any other cool greeting card sites, please list them in a reply.  I will start the list with <a href="http://www.cardgirl.com">www.cardgirl.com</a>
btlfdtvdfewplvydvytlqgbgourrqwzmczkiqkffvtnfvwzqjzwgue



From sabren at manifestation.com  Wed Apr 12 12:48:40 2000
From: sabren at manifestation.com (Michal Wallace (sabren))
Date: Wed, 12 Apr 2000 12:48:40 -0400 (EDT)
Subject: ANN: botlib 0.1 !
In-Reply-To: <Pine.LNX.4.21.0004120814500.6832-100000@fep132.fep.ru>
Message-ID: <Pine.BSF.4.10.10004121232070.5514-100000@chi.pair.com>

On Wed, 12 Apr 2000, Oleg Broytmann wrote:

> least two - I want a threading robot, but much more than that I want an
> asyncore-based robot. I haven't done anything yet; it seems I need to
> rewrite the entire urllib to use asyncore, and it is hard work. Anyway I

Yes, I came to the same conclusion.. You have to rewrite just about
everything.. Though I'm thinking urllib2 can make the job it easier.
Anyway, I decided to take a stab at it, and used the example in the
asyncore docs to create something that does a good chunk of the job:

-----------------------------
announcing botlib 0.1!!!!!!!!
-----------------------------

The botlib module contains two classes: UniBot and MultiBot.

UniBot fetches a single URL.  It's http only, no authentication, no
redirects, etc, BUT: it allows you to specify a timeout down to
fractions of a second.

MultiBot fetches multiple URLS at once. You tell it which urls to
get, and which functions to call for each one once the content is
fetched.

Here's two quick examples:

    #####################

    import botlib

    bot = botlib.UniBot()
    content = bot.fetch('http://www.python.org/')

    #####################


and:

    #########################

    import botlib

    def callback(url, contents, extras):
        print url, 'is this big:', len(contents)

    queen = botlib.MultiBot()
    queen.fetch('http://www.python.org/', callback)
    queen.fetch('http://www.linkwatcher.com/', callback)
    queen.fetch('http://www.sabren.com/', callback)
    queen.loop()
    
    #########################


Anyway, it's far from perfect... The methods and stuff might change
completely, especially once more of the urllib2 features get worked
in... but it is a start.

check it out at: http://www.linkwatcher.com/src/botlib.py

Feedback welcome!

Cheers,

- Michal
-------------------------------------------------------------------------
http://www.manifestation.com/         http://www.linkwatcher.com/metalog/
-------------------------------------------------------------------------






From cut_me_out at hotmail.com  Wed Apr 26 18:53:12 2000
From: cut_me_out at hotmail.com (Alex)
Date: 26 Apr 2000 18:53:12 -0400
Subject: plugins for browsers?
References: <3904ccb0.32134281@news.demon.co.uk>
Message-ID: <etd7ldk325j.fsf@w20-575-97.mit.edu>

> Are there any plugins or other means of running Python in Web
> browsers?

I guess if you're using Netscape, you can specify the command yourself.
Go to Edit -> Preferences, then in the dialogue that brings up, go to
Navigator -> Applications.  Then go to New, and in the dialogue that
brings up, enter .py as the suffix, and python %s as the command.

Don't know if that helps at all.

Alex.


From moshez at math.huji.ac.il  Mon Apr 10 01:11:54 2000
From: moshez at math.huji.ac.il (Moshe Zadka)
Date: Mon, 10 Apr 2000 07:11:54 +0200 (IST)
Subject: Migrating Tcl system to Python
In-Reply-To: <248819a8.00d5228e@usw-ex0106-045.remarq.com>
Message-ID: <Pine.GSO.4.10.10004100709320.26767-100000@sundial>

On Sun, 9 Apr 2000, Hamish Lawson wrote:

> The systems will be on Linux so COM is out but the rest are
> possibilities. As far as I can see, one of the main issues in
> communicating between Tcl and Python will be how to handle
> passing lists and dictionaries (arrays in Tcl). XML-RPC manages
> this automatically and although I don't have any experience with
> either CORBA or ILU, I imagine they do too. If I were to use a
> pipe approach, there would presumably be a need for a layer to
> translate between Tcl's and Python's representations of data
> structures; though I think this might be reasonably simple, are
> there solutions already available, or issues I'd need to be
> aware of?

Tcl has exactly two types: strings, and arrays mapping strings to
strings. The problem is that those are often used to create
data-structures in strange and unpredictable ways. Hence, no good
general solution will probably be available. For simple enough cases,
two Tcl commands "marshal_string" and "marshal_array" which will
create Python expression for strings and dicts will suffice. Just send
these over a socket, and have a Python "eval" at the other side.
--
Moshe Zadka <mzadka at geocities.com>. 
http://www.oreilly.com/news/prescod_0300.html
http://www.linux.org.il -- we put the penguin in .com




From neelk at brick.cswv.com  Thu Apr 20 18:51:10 2000
From: neelk at brick.cswv.com (Neel Krishnaswami)
Date: 20 Apr 2000 22:51:10 GMT
Subject: __del__ problem - would adopting Garbage Collection fix this?
References: <aw2L4.8907$HG1.245743@nnrp1.uunet.ca> <7ojL4.9424$HG1.254543@nnrp1.uunet.ca> <Z%oL4.9643$HG1.257533@nnrp1.uunet.ca>  <l03102802b5245023a247@[193.78.237.147]>
Message-ID: <slrn8fv4eo.1kt.neelk@brick.cswv.com>

Just van Rossum <just at letterror.com> wrote:
> 
> I'm not sure if I understood it correctly, but I think the newly
> proposed garbage collector does *not* call __del__ if the object in
> question is part of a cycle. It seems indeed impossible to do so
> correctly. I think the site explained quite well why this is, but I
> can't seem to repeat it off the top of my head.

The natural order of calling finalizers -- the one that reference
countin automatically induces -- is the topological sort of the
references graph, and there can only be a well-defined topological
sort if there are no cycles in the reference graph. So if there are
object cycles there is no obvious order of calling the finalizers
within each cycle (which algorithm junkies will note are the
strongly-connected components of the reference graph). Note also that
reference-counting *never* calls the finalizers in a cycle.

So not calling the finalizers and giving the programmer access to the
list to break the cycle manually seems like a sensible idea. An even
more sensible idea would be to drop __del__ methods altogether, since
IMO you deserve to lose if you try to use the same deallocator for
arbitrary resources that you use for memory, but that idea is probably
not going to fly. ;)


Neel


From boud at rempt.xs4all.nl  Thu Apr 27 14:14:29 2000
From: boud at rempt.xs4all.nl (Boudewijn Rempt)
Date: 27 Apr 2000 18:14:29 GMT
Subject: Best database to use with Python
References: <3906ea27.698463777@news.online.no> <8e9334$qba$1@news.tht.net> <39084a70$0$22983@personalnews.de.uu.net>
Message-ID: <8ea025$s2c$1@news1.xs4all.nl>

Johannes Stezenbach <yawyi at gmx.de> wrote:

> The one fundamential difference between MySQL and PostgreSQL (apart
> from the license) is that PostgreSQL has a focus on preserving
> data integrity and dealing with concurrency issues (using transactions,
> triggers, etc.) while MySQL omits these features in favor of raw speed.
> The MySQL folks claim to be 10 times faster for UPDATEs.

Last time I looked (five minutes ago, that is), there still wasn't
a DB-API II compliant Python interface to PostgreSQL, which means
that you've got to muck about with ODBC if you want to keep
your code reasonably database-independent.

-- 

Boudewijn Rempt  | http://www.valdyas.org


From ppi at searchy.net  Mon Apr  3 14:38:32 2000
From: ppi at searchy.net (Penpal International)
Date: Mon, 03 Apr 2000 20:38:32 +0200
Subject: Upload
Message-ID: <38E8E528.8C757DA4@searchy.net>

Does anyone has a script to upload files from a form?



-- 
Penpal International
http://ppi.searchy.net/
ppi at searchy.net


From ez4glNOezSPAM at hotmail.com.invalid  Thu Apr  6 14:54:56 2000
From: ez4glNOezSPAM at hotmail.com.invalid (ez4gl)
Date: Thu, 06 Apr 2000 11:54:56 -0700
Subject: Newbie: urllib.urlretrieve()
Message-ID: <221ee89a.3bf4ca4b@usw-ex0102-016.remarq.com>

I'm trying to clone some Intranet pages for a contract
assignment and thought Python would be a good tool because of
it's urllib module.  However, when trying to use
urllib.urlretrieve(), I have found that if you specify the
second argument, the file for writing to has to already exist.
What I want to do instead is create the name of the file on the
fly, based on the name of the URL I'm retrieving.  I guess what
I need to do is to create an empty file ahead of time?!  But how
would I go about doing this under Windows? (Unix I could handle)

Thanks in advance

George



* Sent from RemarQ http://www.remarq.com The Internet's Discussion Network *
The fastest and easiest way to search and participate in Usenet - Free!



From pecora at anvil.nrl.navy.mil  Fri Apr  7 08:40:10 2000
From: pecora at anvil.nrl.navy.mil (Louis M. Pecora)
Date: Fri, 07 Apr 2000 12:40:10 GMT
Subject: Why should I switch to Python?
References: <000701bf9c5a$b806b420$752d153f@tim> <oqity066km.fsf@titan.progiciels-bpi.ca> <slrn8ej2gc.b3t.warlock@raven.myth> <m3itxyz744.fsf@atrus.jesus.cam.ac.uk> <050420000914595917%pecora@anvil.nrl.navy.mil> <mt1z4k8oh1.fsf@astron.berkeley.edu> <m3pus4lbej.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <070420000840091838%pecora@anvil.nrl.navy.mil>

In article <m3pus4lbej.fsf at atrus.jesus.cam.ac.uk>, Michael Hudson
<mwh21 at cam.ac.uk> wrote:

> <grin> - mind you "Riemann zeta function" is a pretty cool name for
> any function.  I also like "Weierstrass p-function", but mainly
> because the p is usually drawn in an entertainly deranged way.

Well, you've got something there.  Let's get some good names for
functions.  My first thought:


Hypergeometric Functions

 
> what-was-this-ng-about-again?-ly y'rs

Ah, who cares? This is more fun.  Just kidding Python guys.  :-)


From prestonlanders at my-deja.com  Mon Apr 10 11:14:10 2000
From: prestonlanders at my-deja.com (Preston Landers)
Date: Mon, 10 Apr 2000 15:14:10 GMT
Subject: Newbie: White space
References: <8bc1oo$qjc$1@yeppa.connect.com.au>
Message-ID: <8csr3s$2ks$1@nnrp1.deja.com>

Unlike most programming languages, whitespace is crucial to Python.  You
will not understand anything in Python until you run and go read the
tutorial:

http://www.python.org/doc/current/tut/tut.html

It's an easy read!

---Preston

In article <8bc1oo$qjc$1 at yeppa.connect.com.au>,
  "Ryan Abrahams" <abrahams_nospam_ at wizard.net.au> wrote:
> I am new to python, I am actualy using it for rendering (via the
Lightflow
> Python Module, at www.lightflowtech.com (and no I don't work for
them)) and
> what I wanted to know was does Python check white space, because I had
some
> code like this:
>
> s.lightBegin()
>     s.lightOn( Sunlight )
>     Surface2 = s.newMaterial("standard", [ "ka", vector3( 0.5, 0.5,
0.0 ),
> "kc", vector3( 0.4, 0.2, 0.0 ), "kd", 0.5, "km", 0.1 ] )
> s.lightEnd()
>
> and Python chucks a spit at the s.lightOn( Sunlight ) line. Now I have
been
> trying to figure it out, maybe python does not support whitespace,
because
> changing the code to
>
> s.lightBegin()
> s.lightOn( Sunlight )
> Surface2 = s.newMaterial("standard", [ "ka", vector3( 0.5, 0.5, 0.0 ),
"kc",
> vector3( 0.4, 0.2, 0.0 ), "kd", 0.5, "km", 0.1 ] )
> s.lightEnd()
>
> and compiling it works perfectly. So is it Python that does not
support hte
> white space, or perhaps the module I am including.
>
> Thanks
> Ryan
>
>

--
|| Preston Landers <prestonlanders at my-deja.com> ||


Sent via Deja.com http://www.deja.com/
Before you buy.


From mjhand at concentric.net  Mon Apr 17 13:58:31 2000
From: mjhand at concentric.net (Manus Hand)
Date: Mon, 17 Apr 2000 11:58:31 -0600
Subject: Upgrading IIS to 1.6 problem....
Message-ID: <38FB50C7.C7E891E4@concentric.net>

I run CGI scripts on a real machine, but I also run some on IIS on an NT
machine.

To configure IIS to know that .cgi files should run Python, I (long ago)
followed Aaron Watters' instructions, and added a "ScriptMap" parameter
in the Windows registry:
    HKEY_LOCAL_MACHINE
        System
            CurrentControlSet
                W3SVC
                    Parameters
                        ScriptMap
                            .cgi -> C:\Program Files\Python\Python.exe
%s %s

Worked just fine for a good long time.

The other day, I downloaded 1.6a2 and installed it on the PC.  BUT...not
at the
same location -- I installed it in the directory D:\Python16

I figured that changing the above registry entry to
"D:\Python16\\python.exe %s %s"
would be all I needed to do and my Python .cgi code would run using 1.6.

Long story short(er), it didn't work.  My .cgi code still runs using the
1.5.2 version
that is in C:\Program Files\Python.  Honest.  I made the registry change
(I promise) and even rebooted.  Twice to be sure.

Anyone know why my registry change hasn't affected IIS yet?

(Oh, I also -- though this shouldn't matter -- made sure to update my
%PATH%
in my autoexec.bat -- now when I type Python at an MS-DOS prompt, I get
1.6a2.)

Let me know.  Thanks,
Manus



From eviltofu at rocketmail.com  Tue Apr 11 08:16:55 2000
From: eviltofu at rocketmail.com (Jerome Chan)
Date: Tue, 11 Apr 2000 12:16:55 GMT
Subject: WinNT and TCL/TK
Message-ID: <eviltofu-9BCDF3.08165411042000@news.earthlink.net>

Have installed Python 1.5.2 on a WinNT Server but it says it can't find 
a TCL lib when I start it up. Do I have to add something?

Thanks!


From mhammond at skippinet.com.au  Fri Apr  7 19:16:08 2000
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Fri, 07 Apr 2000 23:16:08 GMT
Subject: freeze not working (win)
References: <memH4.4438$HG1.127494@nnrp1.uunet.ca>
Message-ID: <Y_tH4.903$4v.4613@news-server.bigpond.net.au>

"Warren Postma" <embed at geocities.com> wrote in message
news:memH4.4438$HG1.127494 at nnrp1.uunet.ca...
> No definition of module win32api in any specified map file.
> Warning: unknown modules remain: win32api
>

This simply means that win32api was not frozen.  The resulting
program will still run if you provide win32api.pyd.  Note that
you must also provide pywintypes1x.dll - this can not be frozen.

> What do I do?  If i had to guess I'd say that the *.dsp is a
Visual C++
> Project file.
> What does Freeze want with the source code to the Win32
extensions?

Freeze is capable of freezing these modules into the same .exe -
its needs the source to do so.

It shouldnt surprise you that freeze needs C sources - most of
freeze's time is spent compiling C code!

>  Do I
> have to get this file from somewhere? I looked on Mark Hammonds
page and my
> only option is to grab source from CVS. {Haven't done that
before. don't
> know how.} What do I do!?

I'd suggest you ignore the error and simply ship win32api.pyd.

Mark.





From claird at starbase.neosoft.com  Sat Apr  8 23:27:07 2000
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 8 Apr 2000 22:27:07 -0500
Subject: Migrating Tcl system to Python
References: <095cc48a.7e9b36dd@usw-ex0108-063.remarq.com>
Message-ID: <000C48B993BF96DA.95DEBE4215A79896.7F63D15FB99EE88D@lp.airnews.net>

In article <095cc48a.7e9b36dd at usw-ex0108-063.remarq.com>,
Hamish Lawson  <hamish_lawson at yahoo.co.uk> wrote:
>I'd be interested in anyone's experience of migrating a (several-
>thousand line) system written in Tcl to Python. And since we'll
>have to deploy new functionality while we're in the process of
>migrating, I'd particularly like to hear of experiences with
>having Tcl code using Python code (and vice versa). I'm
>considering XML-RPC/SOAP as a possible way to do that, but what
>other approaches might there be?
			.
			.
			.
Minotaur.  CORBA.  ILU.  COM, maybe.  Pipes.  Named pipes.
'Depends on your situation--like many things.
-- 

Cameron Laird <claird at NeoSoft.com>
Business:  http://www.Phaseit.net
Personal:  http://starbase.neosoft.com/~claird/home.html


From meh9 at cornell.edu  Mon Apr 24 11:19:24 2000
From: meh9 at cornell.edu (Matthew Hirsch)
Date: Mon, 24 Apr 2000 11:19:24 -0400
Subject: Simple list.append() question
Message-ID: <meh9-5ED8CD.11192424042000@news.cit.cornell.edu>

Hi All,

I have this:
>>> entire_list=[[]]*3
>>> entire_list
[[], [], []]
>>> entire_list[0].append(1.23)
>>> entire_list
[[1.23], [1.23], [1.23]]

I only want to append 1.23 to entire_list[0].  Why did it also append it 
to entire_list[1] and entire_list[2]?

Thanks for your help,
Matt


From garabik at melkor.dnp.fmph.uniba.sk.spam  Thu Apr  6 13:22:58 2000
From: garabik at melkor.dnp.fmph.uniba.sk.spam (Radovan Garabik)
Date: 6 Apr 2000 17:22:58 GMT
Subject: encoding type in a2ps to pretty-print
References: <8chtbq$ohg$1@news.net.uni-c.dk>
Message-ID: <955041768.638594@cdwork.cvt.stuba.sk>

Henrik Berg Nielsen <n99hbn at nat.sdu.dk> wrote:
 : Hi, I am complete newbie and just starting to learn Python. I have just
 : installed a2ps for dos so I can pretty-print all my small programs. I have
 : one problem though. When I pretty-print my code via a2ps all places where I
 : have a '<=' are turned into the similar mathematical sign, that only takes
 : up one space on the line, likewise all '==' are converted to the
 : mathematical "by definition" sign (that is 3 lines on top of eachother). Why
 : is this so? It is really annoying and I would really like a2ps to
 : pretty-print my code the way it looks on my monitor. I think it has
 : something todowith the encoding that a2ps uses,but I'm not sure.

try --highlight-level=normal



-- 
 -----------------------------------------------------------
| Radovan Garabik http://melkor.dnp.fmph.uniba.sk/~garabik/ |
| __..--^^^--..__    garabik @ melkor.dnp.fmph.uniba.sk     |
 -----------------------------------------------------------
Antivirus alert: file .signature infected by signature virus.
Hi! I'm a signature virus! Copy me into your signature file to help me spread!


From nascheme at enme.ucalgary.ca  Tue Apr 25 13:55:56 2000
From: nascheme at enme.ucalgary.ca (Neil Schemenauer)
Date: Tue, 25 Apr 2000 17:55:56 GMT
Subject: VIM NT +python
References: <8e4j27$hgq$1@nnrp1.deja.com>
Message-ID: <slrn8gbn3k.rkh.nascheme@cranky.arctrix.com>

tiddlerdeja at my-deja.com <tiddlerdeja at my-deja.com> wrote:
>Does anyone here have a winNT build of the VIM editor that has python
>support switched on?

    http://www.vim.org/

Vim on Win32 has highlighted Python for ages.  Perhaps you don't
have syntax highlighting turned on.

    Neil

-- 
"God, root, what is difference?" - Pitr| nascheme at enme.ucalgary.ca
"God is more forgiving." - Dave Aronson| http://www.ucalgary.ca/~nascheme


From sylvieb at icestormfx.com  Thu Apr 20 09:11:19 2000
From: sylvieb at icestormfx.com (Sylvie Bruneau)
Date: Thu, 20 Apr 2000 09:11:19 -0400
Subject: A tkMessageBox question!
References: <38FDC335.B9FDCFE4@icestormfx.com> <38FE8AF6.2A4CDBDA@gte.net>
Message-ID: <38FF01F6.898DA84F@icestormfx.com>

Hi Sue.

    Thank you for giving a try, but it doesn't work.  The tkMessageBox must be used
in that way :

tkMessageBox.function( title, message [, options] )

    where function can be showinfo, showwarning, showerror, askquestion, askokcancel,
askyesno, or askretryignore and options can be default, icon, message, parent, title
and type.  And as soon as you call one of these functions, the dialog is popped up,
ie, it is too late to try anything on the look.

    I really think there is no way to do this !

Sly


sue wrote:

> I don't remember the exact syntax, but try:
>
> msg = TkMessageBox("text")
> msg.configure( font = "something_decent" ).
>
> You might have to check that the font exists.
>
> Sue.

--
/=-/=-/=-/=-/=-/=-/=-/=-/=-/=-/=-/=-/=-/=-/=-/=-/=-/=-/=-/=-/
Sylvie Bruneau ( sylvieb at icestormfx.com ) Tel: (514) 527-3963
Software Developer, IceStorm Digital.     Fax: (514) 527-5165
2595 Place Chasse, Montreal, Quebec, H1Y 2C3.

La vie est un cadeau parfois difficile a deballer. (P. Brassard)


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20000420/f9b88645/attachment.html>

From nmm1 at cus.cam.ac.uk  Wed Apr 12 11:08:34 2000
From: nmm1 at cus.cam.ac.uk (Nick Maclaren)
Date: 12 Apr 2000 15:08:34 GMT
Subject: Python/Perl Popularity (Re: A Mountain of Perl...)
References: <200004102111.HAA03380@envy.fulcrum.com.au>     <8d0pcq$7ti$1@nnrp1.deja.com> <g0ssm03o.fsf@dan.resnet.rochester.edu>     <8d149t$jo6$1@nnrp1.deja.com> <8d22qg$oai$1@newsserver.rrzn.uni-hannover.de>
Message-ID: <8d23hi$3p8$1@pegasus.csx.cam.ac.uk>

In article <8d22qg$oai$1 at newsserver.rrzn.uni-hannover.de>, breiter at usf.Uni-Osnabrueck.DE (Bernhard Reiter) writes:
|> In article <8d149t$jo6$1 at nnrp1.deja.com>,
|> 	tom__98 at my-deja.com writes:
|> 
|> >>     tom>  -- How can I browse the Python documentation without firing
|> >>     tom> up a web browser?
|> >>
|> >> Convert it to text format. It's latex, you can make it into anything
|> >> you want, and do anything you want with it.
|> > 
|> > Well, what I'm suggesting is that it might be a good idea
|> > to make a command-line accessible version of the documentation,
|> > together with a command line tool, available as part of the
|> > standard distribution.
|> 
|> lynx is a command line tool and will work fine and fast.

Not really.  It is a 'full-screen" browser.  In particular, it
doesn't work together with scrollback, and you can't easily use
it to produce text for automatic scan.  The original CERN www
browser IS a command-line browser, but it has a pretty revolting
user interface.  Well, so has lynx, for that matter.

However, I do find this whole debate bizarre ....


Regards,
Nick Maclaren,
University of Cambridge Computing Service,
New Museums Site, Pembroke Street, Cambridge CB2 3QG, England.
Email:  nmm1 at cam.ac.uk
Tel.:  +44 1223 334761    Fax:  +44 1223 334679


From mike.zaharkin at westgroup.com  Tue Apr 25 14:03:29 2000
From: mike.zaharkin at westgroup.com (MikeZ)
Date: Tue, 25 Apr 2000 14:03:29 -0400
Subject: Newbie Question
References: <sgat0detcdo94@corp.supernews.com>
Message-ID: <3905ddf2@wwwproxy3.westgroup.com>

Luke,

    Try:

    http://www.mcmillan-inc.com/install1.html

    It works quite well.

Regards,
-Mike Z.

Luke macken wrote in message ...
>Hello, i just started teaching myself Python the other day, and I have been
>getting very into it. I have both the Linux version of it and the Windows
>version. I just want to know how i make executable files with it.
>Thanks
>
>




From gmcm at hypernet.com  Wed Apr 26 15:48:25 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Wed, 26 Apr 2000 15:48:25 -0400
Subject: Namespace Problem with global declaration in module
In-Reply-To: <8e6vv9$4n8$1@nnrp1.deja.com>
Message-ID: <1255343982-62616670@hypernet.com>

spex66 at my-deja.com wrote:

> Hi,
> it seemed very simple but don't worked like expected:
> 
> #FILE: test.py
> dd = None #INIT a shadow instance for automated access
> class zz:
>     #ONLY one class will be instantiated
>     def __init__(self):
>         self.name = 'hallo'
>         global dd
>         dd = self
> #END OF FILE
> 
> >>> from test import *

There's your problem. "from test import *" copies names from 
test to your namespace. Now dd is another name for None. 
yy() affects the dd in test's namespace, not yours.

> >>> yy = zz()
> >>> dd.name
>     ...
>     AttributeError
>     ...


- Gordon



From quinn at riyal.ugcs.caltech.edu  Tue Apr  4 21:30:10 2000
From: quinn at riyal.ugcs.caltech.edu (Quinn Dunkan)
Date: 5 Apr 2000 01:30:10 GMT
Subject: Getting the argument names passed to a function ?
References: <20000404225810.A12156@sz-sb.de> <HRsG4.5120$74.82867@newsc.telia.net> <slrn8el27m.3h1.quinn@riyal.ugcs.caltech.edu>
Message-ID: <slrn8el5t7.40k.quinn@riyal.ugcs.caltech.edu>

> > Hmm, could you say:
> > 
> > def arghack(arg):
> >     for k, v in globals().items():
> >         if arg is v:
> >             return k
> >     else:
> >         return None
> > 
> > Seems to work ok...
> This only works if the argument passed in is a global.  For example:
> >>> def hackSpoiler():
> ...     a = 2
> ...     print arghack(a)
> ...
> >>> hackSpoiler()
> None

Ok, Mr. Spoilsport :)

import sys
def arghack(arg):
    try:
        1/0
    except:
        frame = sys.exc_info()[2].tb_frame.f_back
    while frame:
        for k, v in frame.f_locals.items():
            if arg is v:
                return k
        frame = frame.f_back
    else:
        return None


From bwinton at tor.dhs.org  Mon Apr 17 15:08:53 2000
From: bwinton at tor.dhs.org (Blake Winton)
Date: Mon, 17 Apr 2000 19:08:53 GMT
Subject: else clauses in while and for loops
References: <200004142351.JAA02824@nexus.csse.monash.edu.au> <38F7B350.5D6548E1@mindspring.com>
Message-ID: <slrn8fmogb.ggf.bwinton@tor.dhs.org>

On Fri, 14 Apr 2000 20:09:52 -0400, Chuck Esterbrook wrote:
>Jonathan Giddy wrote:
>> Just like there's no "real need" for tuples (use lists), for loops (use
>> while), or elif (use else: if).
>>
>> All of these concepts are redundant in the sense that they can be mimicked
>> using an alternative sequence of code.  But they're nice because they
>> express higher-level concepts about the algorithm.
>
>I'm aware of all of this. In fact, I think we could all get our work
>done with ADD, NEG, JMP and JMPZ.   :-) But no thank you.

Heck, didn't someone prove it was possible to write a single-instruction
computer?  Decrement and Jump If Not Zero, was the instruction, IIRC.

Later,
Blake.
-- 
2:51pm up 19 days, 15:25, 1 user, load average: 1.00, 1.00, 1.00


From cut_me_out at hotmail.com  Thu Apr 13 13:58:01 2000
From: cut_me_out at hotmail.com (Alex)
Date: 13 Apr 2000 13:58:01 -0400
Subject: Doesn't anybody write to paper anymore?
References: <38F5FDF3.EE2FBA01@tcac.net>
Message-ID: <etdg0spubme.fsf@w20-575-109.mit.edu>

> I would like to pass it into the queue.  IO redirection in the shell
> is ugly and awkward.

I don't know whether this counts as ugly and awkward IO redirection, but
perhaps you could do something like this:

from os import popen
printer = popen ('lpr')
printer.write("page 1\n\r\f")
printer.write("page 2\n\r\f")
printer.close ()

I actually tried "echo 'test' | lpr -" on a hunch before I read the man
page.  You might try that instead if 'lpr' by itself doesn't work.

Alex.


From olafb at pvv.org  Tue Apr 18 07:37:47 2000
From: olafb at pvv.org (Olaf Trygve Berglihn)
Date: 18 Apr 2000 13:37:47 +0200
Subject: Can COM-objects be used in a thread?
Message-ID: <dn1g0sj1vxw.fsf@kikp0524.chembio.ntnu.no>

I'd like to make a thread work on a COM-object created by the main
thread. Is this possible? Can the COM-object be passed to the thread
and function properly there?

I vaguely remember that somebody posted a message about this, saying
it is not possible for a thread to host a COM-object as the main
thread is not dispatching Windows messages to it.
-- 
* Olaf Trygve Berglihn <olafb at pvv.org>


From mikael at isy.liu.se  Wed Apr  5 05:15:31 2000
From: mikael at isy.liu.se (Mikael Olofsson)
Date: Wed, 05 Apr 2000 11:15:31 +0200 (MET DST)
Subject: map.yahoo.com uses Python?
In-Reply-To: <38eb00af.8905516@news.mel.aone.net.au>
Message-ID: <XFMail.000405111531.mikael@isy.liu.se>

On 05-Apr-00 Geoff Deering wrote:
 >  I've heard all of Yahoo is Python based.

"All of" is probably an exaggeration, but Python is listed on

  http://docs.yahoo.com/info/misc/contributors.html

See far down on the page under "Public Domain Software".

/Mikael

-----------------------------------------------------------------------
E-Mail:  Mikael Olofsson <mikael at isy.liu.se>
WWW:     http://www.dtr.isy.liu.se/dtr/staff/mikael
Phone:   +46 - (0)13 - 28 1343
Telefax: +46 - (0)13 - 28 1339
Date:    05-Apr-00
Time:    11:13:25

This message was sent by XF-Mail.
-----------------------------------------------------------------------


From walter at data.franken.de  Thu Apr 27 04:46:28 2000
From: walter at data.franken.de (Walter Doerwald)
Date: Thu, 27 Apr 2000 10:46:28 +0200
Subject: Sourceforge and Python
References: <oq8zy1zqto.fsf@titan.progiciels-bpi.ca> <007b01bfaf14$f33114b0$3cdd3fcb@neil> <slrn8gcgvp.u19.kc5tja@garnet.armored.net> <8e6v3r$3ss$1@nnrp1.deja.com>
Message-ID: <45919985@data.franken.de>

On Wed, 26 Apr 2000 14:39:58 GMT aaron_watters wrote:

> [...]
> cvs -z6
> -d :ext:aaron_watters at cvs.reportlab.sourceforge.net:/cvsroot/reportlab
> commit pdfdoc.py
>
> [...]
>
> SET
> CVSROOT=aaron_watters at cvs.reportlab.sourceforge.net:/cvsroot/reportlab

If you change this to

SET CVSROOT=:ext:aaron_watters at cvs.reportlab.sourceforge.net:/cvsroot/reportlab

you won't have to type the
"-d :ext:aaron_watters at cvs.reportlab.sourceforge.net:/cvsroot/reportlab"
part in the command above.

Bye...
	Walter Doerwald



From jnw22 at cam.ac.uk  Tue Apr 11 08:55:40 2000
From: jnw22 at cam.ac.uk (James "Wez" Weatherall)
Date: Tue, 11 Apr 2000 13:55:40 +0100
Subject: Python port to Psion
Message-ID: <8cv7cc$g3h$1@pegasus.csx.cam.ac.uk>

The Python port to the Psion 5(mx) is missing some features, in particular
threads and serial IO.  I need both features for a project I'm doing, so I'd
like to add them in to the Psion port.

Unfortunately, searches on Deja and on the web pages for the two flavours of
the port fail to throw up anywhere I can get the patches or patched source
code from.

Does anyone know where I could find it?

Cheers,

--
James "Wez" Weatherall





From aahz at netcom.com  Sun Apr  2 12:52:53 2000
From: aahz at netcom.com (Aahz Maruch)
Date: 2 Apr 2000 16:52:53 GMT
Subject: Paid to work in Python? (was: Re: Choice Of Language)
References: <NDBBJPNCJLKKIOBLDOMJMEMECDAA.DavidA@ActiveState.com>
Message-ID: <8c7tt5$7dl$1@slb3.atl.mindspring.net>

In article <NDBBJPNCJLKKIOBLDOMJMEMECDAA.DavidA at ActiveState.com>,
David Ascher <DavidA at ActiveState.com> wrote:
>Grant Edwards
>>
>> Pointless tanget: Is it just me, or is Google really the last usable
>> search engine?
>
>And it's _all_ due to their use of Python!  Cool, eh? =)

That doesn't explain why Infoseek is no longer particularly usable.
--
                      --- Aahz (Copyright 2000 by aahz at netcom.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

"It's 106 miles to Chicago.  We have a full tank of gas, a half-pack of
cigarettes, it's dark, and we're wearing sunglasses."  "Hit it."


From claird at starbase.neosoft.com  Thu Apr 20 10:18:19 2000
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 20 Apr 2000 09:18:19 -0500
Subject: Removing circular references (was Re: Editors and books)
References: <XBFK4.16$Uq2.1831@news.uswest.net> <20000417184833.H15664@xs4all.nl> <8dgbu7$k10$1@slb6.atl.mindspring.net>
Message-ID: <1DCCE06693F01F93.B5A70EEF90EE3111.B0E354CFF4C183AE@lp.airnews.net>

In article <8dgbu7$k10$1 at slb6.atl.mindspring.net>,
Aahz Maruch <aahz at netcom.com> wrote:
>In article <20000417184833.H15664 at xs4all.nl>,
>Thomas Wouters  <thomas at xs4all.net> wrote:
>>
>>The problem with RC is that you can create circular references:
>>
>>a = [some large list]
>>b = [some other large list]
>>a.append(b)
>>b.append(a)
>>
>>del a
>>del b
>
>Note that you can break the circular reference if you do this between
>the append and the del:
>
>del a[-1]
>del b[-1]
>
>Similar tricks work in most other cases where circular references get
>created.
			.
			.
			.
And, for completeness, I'll note that dangling circular
references are innocuous in many (most?) cases.  It's
only long-running processes with systematic memory leaks
which are generally worth concern.
-- 

Cameron Laird <claird at NeoSoft.com>
Business:  http://www.Phaseit.net
Personal:  http://starbase.neosoft.com/~claird/home.html


From ben at co.and.co  Sun Apr 30 17:27:35 2000
From: ben at co.and.co (ben at co.and.co)
Date: Sun, 30 Apr 2000 21:27:35 GMT
Subject: Do I always have to write "self." ?
References: <270420001706386648%pecora@anvil.nrl.navy.mil> <slrn8ghc6t.f01.scarblac-spamtrap@flits104-37.flits.rug.nl> <8ecsgj$6lp$1@slb0.atl.mindspring.net> <m3bt2tonyg.fsf@atrus.jesus.cam.ac.uk> <300420001008012011%pecora@anvil.nrl.navy.mil> <slrn8gp6c4.ilf.kc5tja@garnet.armored.net>
Message-ID: <bz1P4.75789$uF1.160076@afrodite.telenet-ops.be>

Samuel A. Falvo II <kc5tja at garnet.armored.net> wrote:
 
> I always use Hungarian notation when it makes sense to.  For example, if I
> have a function in C that accepts a string and an explicit length argument,
> I would write something like this:
> 
> 	uint32 WriteData( void *buffer, char *pString, uint32 cbString )
> 
> This tells me that WriteData() takes as parameters a buffer, a pointer to a
> string, and the number of bytes in the said string.  I fail to see how this
> is somehow hard to read.  :)

OK, let's bite :-)

    int WriteData(void *buffer, char *str, int len)

I fail to see how this is somehow hard to read :-)

Greetings,
-- 
ben . de . rydt at pandora . be ------------------ your comments
http://users.pandora.be/bdr/ ------- inl. IPv6, Linux en Pandora



From godzilla at netmeg.net  Sat Apr  8 07:21:09 2000
From: godzilla at netmeg.net (Les Schaffer)
Date: Sat, 08 Apr 2000 11:21:09 GMT
Subject: anyone else wrap gd-1.8?
References: <65118AEEFF5AD3118E8300508B1248774CB282@ALTNET> <BCgBXUAZUm74Ewhm@jessikat.demon.co.uk>
Message-ID: <m2itxsygt7.fsf@netmeg.net>

> I have wrapped using the old gdmodule gd-1.8.1. Unfortunately it's a
> hacked version of gd-1.8.1

i've managed to wrap it without any hacks. i have most of the demos
and examples working, but it still needs a bit more work.

les schaffer


From nmm1 at cus.cam.ac.uk  Tue Apr 11 05:24:50 2000
From: nmm1 at cus.cam.ac.uk (Nick Maclaren)
Date: 11 Apr 2000 09:24:50 GMT
Subject: Python paradigms
References: <8cmr70$b0e$1@pegasus.csx.cam.ac.uk> <slrn8f4be7.1t0.bwinton@tor.dhs.org> <8ctc9m$sp2$1@pegasus.csx.cam.ac.uk> <8ctkm5$bko$2@newshost.accu.uu.nl>
Message-ID: <8cur12$5pf$1@pegasus.csx.cam.ac.uk>

In article <8ctkm5$bko$2 at newshost.accu.uu.nl>,
Martijn Faassen <m.faassen at vet.uu.nl> wrote:
>Nick Maclaren <nmm1 at cus.cam.ac.uk> wrote:
>[if your function gets longer than a screenfull, it's time to start
>thinking about splitting it up]
>
>> It really is quite amusing to see these same old myths coming back
>> every few decades :-)
>
>You think splitting a function up is a myth? I find it helps 
>me and my code. It becomes more readable and writeable. I'm genuinely
>surprised by your assertion this is an 'old myth'.

The myth is that splitting functions up necessarily improves clarity.
Sometimes it does; sometimes it doesn't.

>> That particular dogma was very popular back in the late 1960s and
>> early 1970s, when practical programmers pointed out that it was a
>> virtually impossibility for a normal person to keep track of 50
>> trivial functions in their head, but it was quite easy to look
>> at 60 lines of sequential code.  That didn't get listened to
>> then, either!
>
>Of course if you go about randomly chopping up your function into
>multiple ones, that's not going to improve matters. The idea is
>to conceptually split your function into multiple smaller ones
>that make more sense, are more easy to adapt, are coherent by 
>themselves, and make the original big function more readable, as
>by writing all the smaller functions you've hopefully also given
>them names. It's called refactoring, of course. :)

The dogma is that there is a maximum size of function that should
ever be written, and that complex functions necessarily split up
cleanly into smaller sections.  Sometimes they do; sometimes they
don't.  I have seen codes where the median size of functions was
below 5 lines, and you had to keep the specification of over 50 in
your head just to read the simplest function!

>My guidelines are:
>
>If your function is long and complicated, too many concepts
>and abstractions are mingling in the same place. Do a better job
>of abstracting stuff, and you'll end up with shorter, clearer
>functions.

Sometimes.  I have very often done that and backed off, because the
complex function was clearer.  Another good rule is, if auxiliary
functions need to access more arguments and artificial global
variables than they have lines of code, consider whether they would
be better written inline.

>If your function is long and repetitive, you're not being a lazy
>enough programmer. Write some smaller functions that generate the
>stuff, read it from a file or database, whatever.

If the repetitiveness is not quite regular enough to make that easy,
you are making unnecessary work for yourself.  Sometimes that is the
right solution; sometimes it isn't.

>Are you saying these guidelines are based on a myth?

No - the myth is that they invariably improve things.  Life is not
that simple.

>> My current infliction is a accounting file format, and the the
>> function to check the input (you DO check your input, don't you?)
>> needs to perform 100 tests in series.  Yes, I am inventing
>> trivial functions, but it would be clearer to put more of the
>> tests inline.
>
>Why not split this up into a bunch of validation objects or 
>something, and then check the input that way? And aren't at least
>some of these tests similar enough to abstract into a function
>(such as tests for maximum various lengths of strings). Of course
>the accounting file format may be so baroque your function just
>has to be baroque, but that sort of thing *ought* to be rare. :)

That is, of course, what I am doing.  But designing in enough
commonality takes a lot longer (and is more complex) than simply
repeating the very simple tests.  After all, it gains you nothing
by calling an auxiliary that then has a massive switch statement
with special code for each of its calls - and some people do write
code like that!

In this particular case, there are 50-60 entries, which need
40-50 separate test conditions.  By thinking, I can common up
quite a lot of that code, but there is some which is not amenable
to that.


If you remember, I asked whether there were any equivalent to a
couple of very common paradigms, that have been used to clarify
code for over 30 years.  The answer appears to be "no", so I have
to use somewhat less satisfactory ones.  This is not a major
problem - been there, done that, in dozens of languages.

What isn't acceptable, however, is putting up with people saying
that it is heretical to admit that such requirements exist!  Fads
are one thing, but dogma is harmful.


Regards,
Nick Maclaren,
University of Cambridge Computing Service,
New Museums Site, Pembroke Street, Cambridge CB2 3QG, England.
Email:  nmm1 at cam.ac.uk
Tel.:  +44 1223 334761    Fax:  +44 1223 334679


From algaba at my-deja.com  Fri Apr 14 20:03:49 2000
From: algaba at my-deja.com (Manuel Gutierrez Algaba)
Date: Sat, 15 Apr 2000 00:03:49 GMT
Subject: Crosswords for Linux Backtracking recursive crossword generator
References: <8d59hk$6to$1@nnrp1.deja.com> <m3g0sphfay.fsf@atrus.jesus.cam.ac.uk> <8d7gnd$l4a$1@nnrp1.deja.com> <38F74B6C.966EB04F@codeit.com>
Message-ID: <8d8bku$jsl$1@nnrp1.deja.com>

In article <38F74B6C.966EB04F at codeit.com>,
  Grakka <jflores at codeit.com> wrote:
> What does "hlio" means in spanish also?

Being jflores ( Jose Flores, Juan Flores, Javier Flores), you'd
know that it's lio ( mess ) and not hlio, anyway, when you run the
script you got "lio", so it's clear.


Sent via Deja.com http://www.deja.com/
Before you buy.


From nascheme at enme.ucalgary.ca  Sun Apr  9 23:24:14 2000
From: nascheme at enme.ucalgary.ca (Neil Schemenauer)
Date: Mon, 10 Apr 2000 03:24:14 GMT
Subject: [ANNOUNCE] Garbage collection for Python
References: <20000407221750.A29279@acs.ucalgary.ca> <38F13FB9.D1D16AF@mindspring.com>
Message-ID: <slrn8f2if1.dop.nascheme@cranky.arctrix.com>

Chuck Esterbrook <echuck at mindspring.com> wrote:
>Looking at Python FAQ 6.14 on GC, I see another one as well:
>
>http://starship.python.net/crew/gandalf/gc-ss.html

I can't tell too much from that page.  I think it modifies Python
to use the Boehm-Demers garbage collector.  It also seems to be a
bit out of date.

The Boehm-Demers GC works fine when it works but it is not as
portable as Python.  It also requires modifying existing
extension modules.  The chance of Guido accepting such a patch
for Python is about zero.

If you really want to use the Boehm-Demers GC, I have a patch on
my page for Python 1.5.2.

>Also, you might want to get yours mentioned in the FAQ.

I would rather get the bugs stomped out of my patch and get it
included in 1.6 or 1.7.  We can then remove the FAQ entry.  :)


    Neil

-- 
Real programmers don't make mistrakes


From cut_me_out at hotmail.com  Sun Apr 30 18:24:54 2000
From: cut_me_out at hotmail.com (Alex)
Date: 30 Apr 2000 18:24:54 -0400
Subject: Discrepency between JPython and CPython
References: <m3k8hfjlex.fsf@boss.ksu.edu>
Message-ID: <etdpur72pmx.fsf@w20-575-109.mit.edu>

> Consider the expression str(1.14 * 200).
> In CPython it yields 228.0
> In JPython it yields 227.99999999999997
> 
> Shouldn't this result in consistent values across variuos implementation?  
> If so, which of the implementation is buggy?  
> I think it is JPython, may be I am wrong.

Both implementations just reach down to a lower level routine -- for
CPython, it's multiplication in C, for JPython, it's multiplication in
Java.  So probably the discrepency is between the C compiler used to
build your CPython (or maybe your CPU, if it does floating point
operations.  I understand they do, these days.  [The only CPU I ever
learnt anything about in detail was the 6809.] :) and the JVM your
JPython is running on.

Alex.


From mwh21 at cam.ac.uk  Fri Apr 28 09:14:38 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 28 Apr 2000 14:14:38 +0100
Subject: Another question
References: <dVFK4.41$vr2.5817@news.uswest.net> <38FB2EEB.A32AF88C@bam.com> <38FB339F.604364AD@bam.com> <200004171605.MAA04929@zippy.init.com> <38FB7610.80EAF965@ichips.intel.com> <200004172120.RAA11102@zippy.init.com> <jWNK4.432$vr2.32822@news.uswest.net> <m3ln2cckg6.fsf@atrus.jesus.cam.ac.uk> <p53L4.1217$zk4.22638@news.uswest.net> <20000418234932.L15664@xs4all.nl> <3909543E.C775CAD6@uab.edu>
Message-ID: <m3snw6nz9d.fsf@atrus.jesus.cam.ac.uk>

Shae Erisson <shapr at uab.edu> writes:

> Thomas Wouters wrote:
> 
> > Of course not all Python code can be boiled down to these two
> > things (they aren't really 'rules', by the way, merely descriptions
> > of the Python view of the world) but understanding these subtle
> > differences can really make you grok Python ;) Tim Peters once
> > wrote down the Python Philosophy (or zen):
> 
> Can you tell me where you found the python philosophy quote? I'd
> like to read more along these lines.  thanks,

http://starship.python.net/crew/amk/quotations/python-quotes.html

Good stuff.  One of my favourites:

 Python is a truly wonderful language. When somebody comes up with a
 good idea it takes about 1 minute and five lines to program something
 that almost does what you want. Then it takes only an hour to extend
 the script to 300 lines, after which it still does almost what you
 want.
                                              Jack Jansen, 8 Jul 1992 

Cheers,
M.

-- 
  it's not that perl programmers are idiots, it's that the language
  rewards idiotic behavior in a  way that no other language or tool 
  has ever done                        -- Erik Naggum, comp.lang.lisp


From pmadsen at newbridge.com  Tue Apr  4 17:07:33 2000
From: pmadsen at newbridge.com (Paul Madsen)
Date: Tue, 04 Apr 2000 17:07:33 -0400
Subject: complete novice question
Message-ID: <38EA5995.F7B9815D@newbridge.com>

Downloaded Python-1.5.2.

I've run 'configure' and it seems to have worked. Now, according to the
Python package's README, I should be able to type 'make' for the rest of
the install. If I don't have a 'make' file, did the configure portion
not work?

thanks






From effbot at telia.com  Mon Apr  3 10:40:07 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Mon, 03 Apr 2000 14:40:07 GMT
Subject: Unicode program representation
References: <015c01bf9d0b$76f9b2a0$59dd3fcb@neil> <oqem8n4zlx.fsf@titan.progiciels-bpi.ca> <00ac01bf9d65$d988c460$36dd3fcb@neil> <oqwvmf1g2g.fsf@titan.progiciels-bpi.ca>
Message-ID: <b32G4.1106$n68.187651072@newsb.telia.net>

Fran?ois Pinard <pinard at iro.umontreal.ca> wrote:
> Until `sre' definitely replaces `re', you'll have to specify which one you
> want, forcing you as a programmer to a better tracking of narrow vs.
> wide string types: but this is only temporary, the goal seems to be that
> the programmer should not worry about these things.

footnote: 'sre' handles both ordinary strings and unicode
strings.

the reason we're shipping it separately at this time is that's
still very rough -- but if your regular expressions happen to
work under 'sre', you can feed it ordinary strings or unicode
strings.  it's figures the rest out all by itself.

for more information on the current status of 'sre', see the
'sre' homepage:

    http://w1.132.telia.com/~u13208596/sre

</F>




From embed at geocities.com  Wed Apr  5 12:25:03 2000
From: embed at geocities.com (Warren Postma)
Date: Wed, 5 Apr 2000 11:25:03 -0500
Subject: python16_d.lib, python16_d.dll python15_d.dll
References: <013101bf9edb$a76984c0$1001a8c0@nevrax.net>
Message-ID: <1YIG4.3731$HG1.106994@nnrp1.uunet.ca>

>Are there any way to have the python16_d.lib and .dll, or the
python15_d.dll ?????

If you have Visual C++, download the source code and build them. If not, beg
someone else who does have Visual C++.







From jbauer at rubic.com  Thu Apr 27 15:40:19 2000
From: jbauer at rubic.com (Jeff Bauer)
Date: Thu, 27 Apr 2000 14:40:19 -0500
Subject: xmlrpc: remote methods on attributes
Message-ID: <390897A3.D0BF0717@rubic.com>

Hi all. I've been playing around with PythonWare's 
xmlrpclib module.

It's pretty neat, but I want to remotely invoke methods 
on attributes.  After diddling around and trying to
fool the library, I opted for a more straightforward
approach: a RemoteMethod mixin.

### WARNING: Code for discussion purposes only. ###
from types import FunctionType

class RemoteMethod:
    def __init__(self):
        self.__methods__ = {}
        for n, m in get_members(self.__class__).items():
            if n[:2] != '__' and \
               type(m) is FunctionType:
                self.__methods__[n] = m
    def __rcall__(self, method, params):
        if self.__methods__.has_key(method):
            return apply(self.__methods__[method], (self,)+params)
        try:
            pos = string.index(method, '.')
        except ValueError:
            pass
        else:
            head = method[:pos]
            tail = method[pos+1:]
            if hasattr(self, head):
                return self.__dict__[head].__rcall__(tail, params)
        return None

def get_members(_class, members=None):
    if members is None:
        members = {}
    for k in _class.__bases__:
        get_members(k, members)
    for n, m in vars(_class).items():
        if not members.has_key(n):
            members[n] = m
    return members

########################################################

An example of how an xmlrpc server could be implemented:

class Blarg(RemoteMethod):
    def __init__(self):
        apply(RemoteMethod.__init__, (self,))
    def blarg_method(self, x):
        return "blarg_method(%s)" % x

class Server(RemoteMethod, xmlrpc_handler):
    def __init__(self):
        apply(RemoteMethod.__init__, (self,))
        self.blarg = Blarg()

########################################################

In a separate module, the server is remotely invoked.
I map 'call' to '__rcall__' in the top level connection
object, derived from xmlrpclib.

    server = Server("http://yadda-yadda")
    print server.blarg.blarg_method("Aieee!")

########################################################

I'd be interested in anyone's comments about this
approach, especially from other xmlrpc users.  Thanks.

Jeff Bauer
Rubicon Research



From effbot at telia.com  Mon Apr  3 12:03:47 2000
From: effbot at telia.com (Fredrik Lundh)
Date: 3 Apr 2000 11:03:47 -0500
Subject: Dr. Dobb's Python-URL! - weekly Python news and links (Apr  3)
Message-ID: <42FC2C22B9F94665.58B059EC6AF1AB7B.326AADF624B2C732@lp.airnews.net>

The Software Development conference in San Jose turned out to be a major
Python event.
    The conference site:
        http://www.sdexpo.com/

    Guido's keynote covered the CP4E project.  The slides haven't
    been posted yet, but they were based on this classic talk:
        http://www.python.org/doc/essays/ppt/acm-cp4e/index.htm

    Some media coverage.  Like it or not, but Python is hot! :-)
        http://www.ddj.com/articles/2000/0050/0050toc.htm
        http://www.oreillynet.com/pub/a/network/2000/03/30/softwaredev.html

    Lots of great talks about Python.  Among others, David Ascher
    talked about JPython and popular tools and extensions:
        http://www.activestate.com/sd2000.htm

Activities:
    Colorado Linux Info Quest 2000: Python and Zope BOF Sessions
        http://www.thecliq.org/bofs.html

    UCSC Extension: Python Programming Course [corrected link]
        http://www.ucsc-extension.edu/knowledge_is_timeless/qd/softlist.taf?function=detail&X_Number=X444.3

Releases:
    Python 1.6 alpha 1 is here! (first alpha, very rough in places)
        http://www.python.org/1.6/

    Guido tells us what's in the new release, sort of.  Note the
    posting date (for the true story, see the above site):
        http://www.deja.com/=dnc/getdoc.xp?AN=605330122

    Christian Tismer: Stackless Python 1.1 and Continuations 0.8
        http://www.deja.com/=dnc/getdoc.xp?AN=603584578

    Fred Drake: Updated Python 1.5.2 Documentation
        http://www.python.org/doc/

    Andrew Kuchling: Perfect Minimal Hash Generator
        http://starship.python.net/crew/amk/python/code/perfect-hash.html

    Ka-Ping Yee: Roundup - a simple and effective bug-tracking system
         http://www.lfw.org/python/

    Phil Thompson: PyQt/PyKDE v0.11 (python bindings for Qt and KDE)
        http://www.deja.com/=dnc/getdoc.xp?AN=603477664

    William Park: Simple Numerical Recipes in Python
        http://www.python.org/topics/scicomp/recipes_in_python.html

    ReportLab Inc: ReportLab Release 0.90 (PDF generator library)
        http://www.deja.com/=dnc/getdoc.xp?AN=603415004

    Digital Creations: Zope 2.1.6
        http://www.zope.com/Products/Zope/2.1.6/Zope_216_release

Bits and pieces from the newsgroup:
    Mark Hammond: How to determine the Ethernet HW Address (on
    Windows, requires PythonWin 130):
        http://www.deja.com/=dnc/getdoc.xp?AN=603363272

    Michael Hudson: How to use continuations in Stackless
    Python:
        http://www.deja.com/=dnc/getdoc.xp?AN=600190793

    Gordon McMillan: On Python's startup performance:
        http://www.deja.com/=dnc/getdoc.xp?AN=599756252


========================================================================

Everything you want is probably one or two clicks away in these pages:

      Python.org's Python Language Website is the center of Pythonia
          http://www.python.org

      eff-bot's complements this digest with his daily python url
          http://hem.passagen.se/eff/url.htm

      Much of Python's real work takes place on Special-Interest Group
          mailing lists http://www.python.org/sigs/

      Python Consortium emerges as an independent nexus of activity
          http://www.python.org/consortium

      The Vaults of Parnassus ambitiously collects Python resources
          http://www.vex.net/~x/parnassus/   

      Python To-Do List anticipates some of Python's future direction
          http://www.python.org/cgi-bin/todo.py

      Python Journal is at work on its second issue
          http://www.pythonjournal.com

      Links2Go is a new semi-automated link collection; it's impressive
      what AI can generate
          http://www.links2go.com/search?search=python

      Archive probing trick of the trade:
          http://www.dejanews.com/dnquery.xp?QRY=&DBS=2&ST=PS&defaultOp=AND&LNG=ALL&format=threaded&showsort=date&maxhits=100&groups=comp.lang.python


Previous - (U)se the (R)esource, (L)uke! - messages are listed here:
  http://purl.org/thecliff/python/url.html
or
  http://www.dejanews.com/dnquery.xp?QRY=~g%20comp.lang.python%20Python-URL%21


Suggestions/corrections for next week's posting are always welcome.
  http://www.egroups.com/list/python-url-leads/

To receive a new issue of this posting in e-mail each Monday morning,
ask <claird at neosoft.com> to subscribe.  Mention "Python-URL!".


-- The Python-URL! Team--

Dr. Dobb's Journal (http://www.ddj.com) is pleased to participate in and
sponsor the "Python-URL!" project.
-- 

Cameron Laird <claird at NeoSoft.com>
Business:  http://www.Phaseit.net
Personal:  http://starbase.neosoft.com/~claird/home.html


From markx.daley at intel.com  Wed Apr 19 10:32:09 2000
From: markx.daley at intel.com (Daley, MarkX)
Date: Wed, 19 Apr 2000 07:32:09 -0700
Subject: Very Horrible Question About Goto's
Message-ID: <A63AFB20111ED311AC5500A0C96B7BF5015F0F33@orsmsx36.jf.intel.com>

It depends on what the GOTOs are doing.  If they are providing a GOSUB-like
feature, you can do the same thing in Python by defining the task and then
invoking it in other functions.  If they are providing flow control, it may
be a little trickier.  I'm not an expert, but you should be able to work
around just about anything you've done in a previous language if you know
the right Python commands.
 
-Mark

-----Original Message-----
From: Richard Jones [mailto:richard.jones at ibahealthcare.com]
Sent: Wednesday, April 19, 2000 12:09 AM
To: 'python-list at python.org'
Subject: Very Horrible Question About Goto's


I'm trying to create a converter from our own internal language (horrid)
into python.  My only problem so far is that the original language uses
*LOTS* of goto's.  Is there any way that this can be replicated within
python ?
 
Thanks for any help.
 
Richard Jones
 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20000419/eb9d3aaa/attachment.html>

From effbot at telia.com  Wed Apr 12 09:27:50 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Wed, 12 Apr 2000 13:27:50 GMT
Subject: How to know what re.sub took?
References: <38F387AE.1010704@gmx.net>
Message-ID: <qR_I4.640$rc9.190893056@newsb.telia.net>

Per Kistler <kistler at gmx.net> wrote:
> How to know what has been substituted away, if one does something like:
>
> >>> import re
> >>> rex = re.compile(r"(hans|fred|simon)")
> >>> st = "max hans bill"
> >>> st2 = rex.sub("",st)
> >>> st2
> 'max bill'
>
> It took "hans" away, but how can I learn this automatically?

>>> st2 = rex.subn("",st)
>>> st2
('max  bill', 1)

the second tuple value is the number of substitutions made.

</F>




From smalleys at gte.net  Sat Apr 15 13:57:45 2000
From: smalleys at gte.net (sue)
Date: Sat, 15 Apr 2000 17:57:45 GMT
Subject: Need an enum type extension
References: <38F52A7A.D8841724@gte.net> <JdeJ4.1552$Za1.25652@newsc.telia.net> <38F68AF4.46DDA670@gte.net> <slrn8fdshr.3qc.scarblac-spamtrap@flits104-37.flits.rug.nl>
Message-ID: <38F89F9D.6B4162DE@gte.net>

Aargh!  I never thought of a _separate_ get_list() function
to pass the literals back out.  Thank you, thank you.  That
should suffice for now.  I knew there had to be an easier
way...

Sue.


From brent.fulgham at xpsystems.com  Thu Apr 20 19:34:45 2000
From: brent.fulgham at xpsystems.com (Brent Fulgham)
Date: Thu, 20 Apr 2000 16:34:45 -0700
Subject: Using python on the web
Message-ID: <EDFD2A95EE7DD31187350090279C6767754BA8@THRESHER>

A couple of questions:

 
> We build medium-busy, CGI-intensive sites in Python. We were briefly
> quite excited about Zope, but once we realized it locked us into Yet
> Another Programming Language, we lost interest. DHTML is particularly
> problematic,
1.  I'm not familiar with Zope.  Is DHTML specific to Zope?  It seems
to be proudly advertised on a lot of "webware" brochures I see these
days.

 
> The biggest hassle for us has been the weak MySQL support in 
> Python. The spec keeps changing and the interface is raw.

2.  Have you tried using PostgreSQL?  MySQL may be fast, but it's
really little more than a filesystem represented as a database.
It has no transaction support whatsoever which tends to lead to
problems when you have a community-style site that has multiple
parties trying to add messages, etc.  Have you had much success
using it in a production environment?  I realize Slashdot and I
think Sourceforge do, but they have their share of troubles.

Of course, the Postgres Python bindings may be just as bad from your
perspective.  :-)

Thanks,

-Brent



From x_mst at propack-data.com  Fri Apr 28 11:50:05 2000
From: x_mst at propack-data.com (Michael =?iso-8859-1?Q?Str=F6der?=)
Date: Fri, 28 Apr 2000 17:50:05 +0200
Subject: Using BaseHTTPServer
References: <Pine.GSO.4.10.10004281007030.10249-100000@sundial>
Message-ID: <3909B32D.734E1F0F@propack-data.com>

Moshe Zadka wrote:
> 
> On Fri, 28 Apr 2000, Lars von Wedel wrote:
> 
> > I assume that I can use BaseHTTPServer as it is. I just need
> > to implement a working request handler, right?
> >
> > Now comes the more fuzzy part:
> >
> > Since I do not want to execute external files but internal methods,
> > I think I could modify CGIHTTPServer by subclassing it and overriding
> > its do_GET() method to decode the request directly? Or would I rather
> > use SimpleHTTPServer, here?

I also started with CGIHTTPServer which does forking. But I wanted to be
able to let it run under Windows => forking is not an option. I modified
all my code to be thread-safe, derived my own handler from
SimpleHTTPServer.SimpleHTTPRequestHandler and used the
SocketServer.ThreadingMixIn for running multi-threaded.

> > Within any of the do_GET methods, how do I access data that has been
> > sent by the client (...&data1=val1?data2=val2?...)?
> 
> Use the CGI module <wink>. Just copy the code from CGIHTTPServer to create
> the dictionary, (not as os.environ), and use the cgi module to parse
> that dictionary.

Note: IMHO module cgi is NOT thread-safe (reads sys.stdin and use
os.environ). I'm using my own module cgiforms.py anyway.

See http://web2ldap.de for all of this stuff, especially module
pylib/w2lserver.py for starting the server. Currently I'm working on
letting it run with SSL support with the help of the great M2Crypto
module. Let me know if you want to have a (already working) snapshot of
this...

Ciao, Michael.


From rcc at nospamthanks_jennic.com  Thu Apr 13 11:01:49 2000
From: rcc at nospamthanks_jennic.com (Robert Cragie)
Date: Thu, 13 Apr 2000 15:01:49 GMT
Subject: List mapping?
References: <38f5d68e@news.xtml.co.uk>
Message-ID: <newscache$u6nysf$1l1$1@jenpc07.jennic.co.uk>

Nick Trout <nick at spam_me_notvideosystem.co.uk> wrote in message
news:38f5d68e at news.xtml.co.uk...
> Is there a nice way of pairing up the members to form tuple pairs in
another
> list? ie. like you might to processing input args.
>
> ie. changing ['/a','1','/b','2','/c','3'] into [ ('/a','1'), ('/b','2'),
> ('/c','3') ]
>
> eg:
>
> args = '/a 1 /b 2 /c 3'
> listargs= string.split(args)
> tuplist = []
> i=0
> while i < len(listargs):
>     tuplist.append( (listargs[i],listargs[i+1]) )
>     i = i+2
> for i in tuplist:
>     processargs(i)
>
> or:
>
> args = '/a 1 /b 2 /c 3'
> for i in string.split(args):
>     i = argflag
>     # i want next arg now as a parameter!!!
>
> Can I (you!) get rid of the long winded while loop cleverly?!!  :-)

Try this:
mylist = ['/a','1','/b','2','/c','3']
map(lambda x: reduce(lambda a, b: (a,b), mylist[x:x+2]), range(0,
len(mylist), 2))

Robert Cragie




From effbot at telia.com  Wed Apr 26 15:22:09 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Wed, 26 Apr 2000 19:22:09 GMT
Subject: [Python-Dev] Re: Python 1.6a2 Unicode bug (was Re: comparing strings and ints)
References: <l03102805b52ca7830b18@[193.78.237.154]> <39072F0C.5214E339@lemburg.com>
Message-ID: <BlHN4.1719$wYl.195075584@newsb.telia.net>

M.-A. Lemburg <mal at lemburg.com> wrote:
> Please note that the support for mixing strings and Unicode
> objects is really only there to aid porting applications
> to Unicode.
>
> New code should use Unicode directly and apply all needed
> conversions explicitly using one of the many ways to
> encode or decode Unicode data. The auto-conversions are
> only there to help out and provide some convenience.

does this mean that the 8-bit string type is deprecated ???

</F>




From greg at cosc.canterbury.ac.nz  Wed Apr 12 21:25:20 2000
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Thu, 13 Apr 2000 13:25:20 +1200
Subject: issubclass funny business
References: <m2snwsd5ah.fsf@netmeg.net> <ekII4.1212$Za1.21055@newsc.telia.net> <m2purwcfjf.fsf@netmeg.net> <m3itxojga4.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <38F52200.A93110A7@cosc.canterbury.ac.nz>

Michael Hudson wrote:
> 
> It'd be nice if when you `reload'ed
> modules that contain classes you could "smash" extant instances of
> those classes so that they were instances of the new class, but I
> can't for the life of me see how you might do this.

I can: when reloading a module containing a class
definition, and there is an existing definition of the
class, replace the contents of the existing class
object instead of making a new one.

Also, it seems to me that the "importing the main
module" problem could be solved by implementing the
running of a script from a file by means of importing
it.

-- 
Greg Ewing, Computer Science Dept,
+--------------------------------------+
University of Canterbury,	   | A citizen of NewZealandCorp, a	  |
Christchurch, New Zealand	   | wholly-owned subsidiary of USA Inc.  |
greg at cosc.canterbury.ac.nz	   +--------------------------------------+


From tuttledon at hotmail.com  Thu Apr 20 22:48:26 2000
From: tuttledon at hotmail.com (Don Tuttle)
Date: Fri, 21 Apr 2000 02:48:26 GMT
Subject: Why MFC instead of Tkinter -- was Re: Pythonwin?
References: <eMz2Nmiq$GA.73@net025s> <4QtL4.24$v85.41@news-server.bigpond.net.au> <chuL4.8483$O7.17315@typhoon.southeast.rr.com> <7LDL4.534$v85.2563@news-server.bigpond.net.au> <sfu6d657l6e184@corp.supernews.com> <38ff21b8$0$10972@senator-bedfellow.mit.edu> <38FF281E.5B2DCC71@tismer.com> <38ff823d$0$10971@senator-bedfellow.mit.edu> <2wML4.15897$O7.23504@typhoon.southeast.rr.com> <vXOL4.748$v85.4298@news-server.bigpond.net.au>
Message-ID: <_jPL4.33312$cZ.42202@typhoon.southeast.rr.com>

Thanks again Mark.  I quess I need to learn to "scribble"  and then check
out the Prosise book Brett recommended.

Don


> > With VB,  I've only seen Python mentioned as a COM object to be used by
> VB.
> > Are there other ways?
>
> Probably, but this is the best.  Using this technique, you generally end
> up with all the UI logic (ie, this is selected, so I need to disable these
> controls, etc) written in VB.  Only your "business logic" is done in
> Python.  It probably is possible to do some of the GUI logic in Python,
> but that doesnt make as much sense given VBs IDE support for this.  You
> can swap VB for Delphi in this sentence too...
>
> > With VC++,  I gather you use the dialog editor to generate the MFC code
> that
> > you then plug into Python?  Correct?
>
> Exactly.  With MFC/Pythonwin, you get to create a DLL that contains UI
> resources, and use them.  "Use them" means load them, display them, and
> hook all the messages and notifications for them - ie, your GUI logic ends
> up completely in Python, without the nice, trendy
> "double-click-on-an-object-to-edit-its-event-list" IDE that VB/Delphi give
> you.
>
> Mark.
>
>
>




From rcunning at acm.org  Mon Apr  3 23:57:35 2000
From: rcunning at acm.org (Robert W. Cunningham)
Date: Tue, 04 Apr 2000 03:57:35 GMT
Subject: VMware/Win4Lin/Wine + Linux... and Python?
Message-ID: <38E96834.B500B786@acm.org>

With at least three ways to run Windows applications simultaneously with
Linux applications, it seems to me there may be a need and/or
opportunity to improve the connectivity and interface between the two
worlds.  And I suspect Python may be just the tool to do it.

1. Python has excellent ports on both Win9x and Linux.

2. Python is well able to handle "tricky" UI issues on both platforms,
such as Drag & Drop.

Taking Drag & Drop as an example, there is a need for interoperability
at that level, especially for the huge wave of "newbie" Linux users who
can't abandon their Windows investment all at once.  Making the Windows
and Linux worlds play together a bit better should certainly improve the
acceptance of Linux.

Another example:  Make the Windows Media player available to Linux apps.

But is there a way to get instances of Python on both platforms talking
to each other?  Of course, a socket could be opened and data shuttled
back and forth, but that will be fairly slow and demand a fair amount of
overhead.  Has there been any work done on a higher-level
Python-to-Python interface?

While it is common for Java applets to talk to one another on a
cross-platform basis, the performance is low and the cost is high,
especially when the goal is to expose services available on one system
to applets running on another.  Can Python do this (to quote NASA's Dan
Goldin) "Faster, Better, Cheaper"?

The commercial efforts put forth by VMware and Win4Lin are very
significant, but I feel that Wine will be the way to go in the long run
(I want source).  But that run is a *very* long run, and I feel the
commercial approaches have lots to offer today (especially at the $50
Win4Lin price) that can take some of the intense pressure off of the
Wine developers.

In fact, some form of scriptable interface between Windows and Linux
should have immediate benefits for the Wine community as well, since it
should allow them to displace Windows functionality one piece at a time,
without having to build all the pieces above and below at the same time
(a development and debugging nightmare).

At the bottom level, I feel it should be possible to establish a common
"Python space" between the two OS's, possibly running a single Python
engine, that delivers services to both.  Or, more simply, a Python
package for each OS that establishes an interface and provides services
to independent Python instances running under each OS.  Either way, it
seems to me there is great potential for some very fast gains to be
made.

On a related tangent:  Could such an interface be added to a remote
control product (VNC?) to better integrate remote Windows systems and
applications into the Linux desktop?  Is it just a matter of scale, or
more?  What about those curious Windows Terminal Servers and Citrix
products, and the applications they serve up?  While it would probably
require lots of network bandwidth, the underlying idea seems to be the
same, IMHO.

I have already dangled the idea before Trelos, the makers of Win4Lin,
and they are definitely excited by the possibilities.  Plus, with
Win4Lin being a commercial product, an inter-OS Python interface would
allow Trelos to expose interfaces without having to give away any of
their code (or revenue).  And this is one group that sure seems to be
earning their money, even given their unfortunate lack of support for
Open Source.  Python may well be the wedge needed to open that door.

What do you all think?  Insane?  Worthless?  Useless?  Or the greatest
thing since vi and edlin?

And, even if it is a good idea, can it be made to work?  If so, then
how?


Regards,

-BobC




From aycock at csc.uvic.ca  Mon Apr  3 16:04:28 2000
From: aycock at csc.uvic.ca (John Aycock)
Date: 3 Apr 2000 20:04:28 GMT
Subject: Python 1.6 alpha 1 released
References: <200003312130.QAA04361@eric.cnri.reston.va.us> <38E52716.A1C63B5D@roguewave.com> <rJ9F4.549$n68.187227648@newsb.telia.net> <38E52D76.1820FC32@roguewave.com> <8c7tfd$k80$1@nntp9.atl.mindspring.net>
Message-ID: <8catgc$5i84$1@uvaix7e1.comp.UVic.CA>

Aahz Maruch <aahz at netcom.com> wrote:
> 
> I'm sympathetic.  We only upgraded our production system from 1.5.1 to
> 1.5.2 last week, because we'd made a number of 1.5.1-specific hacks and
> had to back them out correctly for 1.5.2.  Now we'll have to do the same
> for 1.6, but as with the 1.5.2 upgrade, we don't have to follow Guido's
> timeline.

I'm not going to take sides here, but I coincidentally just finished
reading something highly relevant:

	`I'd like to insist that if they [the programming languages
	 people] add something to a language, or change something in
	 a language, they also provide us with the algorithm which will
	 make the transfer in our programs so that we can automatically
	 translate our programs...'

That's part of the keynote address made by (the late) Grace Murray Hopper
to the 1978 ACM History of Programming Languages Conference.  Excellent
reading.

Although I don't have time to volunteer, it wouldn't be too difficult
to write a program which would look through existing Python code and
warn of potential uses of these changed constructs.

John


From pecora at anvil.nrl.navy.mil  Sun Apr 30 09:53:34 2000
From: pecora at anvil.nrl.navy.mil (Louis M. Pecora)
Date: Sun, 30 Apr 2000 13:53:34 GMT
Subject: Do I always have to write "self." ?
References: <270420001706386648%pecora@anvil.nrl.navy.mil> <slrn8ghc6t.f01.scarblac-spamtrap@flits104-37.flits.rug.nl> <280420000746597329%pecora@anvil.nrl.navy.mil> <slrn8giv6f.fl9.scarblac-spamtrap@flits104-37.flits.rug.nl>
Message-ID: <300420000953359862%pecora@anvil.nrl.navy.mil>

In article <slrn8giv6f.fl9.scarblac-spamtrap at flits104-37.flits.rug.nl>,
Remco Gerlich <scarblac-spamtrap at pino.selwerd.nl> wrote:

> 
> No, you can just declare methods like
> 
> class Foo:
>    def Method(s,x):
>       s.y = x
> 
> The name 'self' is just a convention.

Ahhhh...gotcha.  I keep forgetting how type-lose Python is.  And how
flexible.  Thank you.


From spin at fastlink.com.au  Sat Apr 15 00:27:42 2000
From: spin at fastlink.com.au (Malcolm Valentine)
Date: Sat, 15 Apr 2000 04:27:42 GMT
Subject: Python paradigms
References: <8cmr70$b0e$1@pegasus.csx.cam.ac.uk> <slrn8f4be7.1t0.bwinton@tor.dhs.org> <8ctc9m$sp2$1@pegasus.csx.cam.ac.uk> <8ctkm5$bko$2@newshost.accu.uu.nl> <8cur12$5pf$1@pegasus.csx.cam.ac.uk> <8cv91s$q2i$1@newshost.accu.uu.nl>
Message-ID: <slrn8fdfiu.v9q.spin@Dizzy.vals>

m.faassen at vet.uu.nl aka Martijn Faassen posted:

>Nick Maclaren <nmm1 at cus.cam.ac.uk> wrote:
>> Sometimes they do; sometimes they
>> don't.  I have seen codes where the median size of functions was
>> below 5 lines, and you had to keep the specification of over 50 in
>> your head just to read the simplest function!
>
>Didn't the functions have readable names then? Of course it's silly
>to take all this to ridiculous extremes, but if the naming was
>done well you have a lot of contextual information when you're reading
>a function which you wouldn't have otherwise.

  I think the murkiness is introduced when one has to remember what
exactly each of the smaller functions returns, and what each of them
does in the way of handling erroneos input,etc.

>> If you remember, I asked whether there were any equivalent to a
>> couple of very common paradigms, that have been used to clarify
>> code for over 30 years.
>
>Obfuscate code at least as often as clarify.

  The ?: operator is quite neat and easily parsed by a human reader
when it is one a single line, or two. When someone uses them in a 
huge multi-line construct filled with ->*||&&"'\ you know there is a
publican and an optomotrist in their family.

<snipped>
>in your first post, but I disagree that your ?: construction makes the
>code more easy to read than than various alternatives. I can grok C's
>(?:), but its main advantage is that it makes code more writeable, not
>more readable, in my opinion. Likewise any replacement to enable the

I'd say it ranks the same as the following;

    a,b,c,e,f,g = tupple_ware

  Note, I'm not knocking the above, but it isn't as simple as it looks.

<snipped>
>In general, the pattern in Python is that it avoids several handy but
>easy to convolute constructs; for instance we have no C style 
>for(;;) loop either, while it's clearly the general case. Instead in
                        1: Is not really any different.
                        
>No 'heretical'. Just trade-offs. Python's idioms for writing Python. The
>strategy does have a pay-off: Python code tends to be fairly readable.

  The great advantage I think is that despites ones particular methods
of coding practise, everyone can read everyone else's Python with a lot
more ease then is common in a language like C.

Cheers,
  Malcolm Valentine.


From python at rose164.wuh.wustl.edu  Thu Apr 13 07:25:32 2000
From: python at rose164.wuh.wustl.edu (David Fisher)
Date: Thu, 13 Apr 2000 06:25:32 -0500
Subject: Deadlocking IDLE under Win32
References: <8cuudo$8pj$1@pegasus.csx.cam.ac.uk> <ciPI4.1680$UB.11667@news-server.bigpond.net.au>
Message-ID: <00aa01bfa53b$2fe73bc0$0201a8c0@spkydomain>

----- Original Message -----
From: "Mark Hammond" <mhammond at skippinet.com.au>


> My guess is that the "print" under IDLE is not thread-safe.  Change the
> program to not print (or to append items to a queue, which the main thread
> then prints) and see if this is the problem.

You are correct sir.  Of course, your guess is a bit more educated that most
<wink>.  I ran into the problem printing from multiple threads in IDLE a
while ago, and just assumed it was something that everybody new about
already.  I'm pretty new to python.

While I'm here, I want to say thanks for win32com.  It totally kicks booty.
I have never done anything with COM before and in the last three days I
wrote a COM server to read data from several meters on a testing maching and
load them real-time into an Excel workbook.  Now the users who can't program
can edit the spreadsheet to change formulas and change how the report looks
on their own.

I still don't understand what's going on when I register the server, or how
the VBA object calls get to python, but I works great.

And while I'm at it, thanks for the win32 threads too.  I needed to change
thread priority, and there the solution was.  Too bad I had already hacked
thread_nt.h to allow me to do the same thing before I thought to check the
newsgroup.  I use your version though.  In case I ever want someone else to
run my code <wink>.

I could go on but you get the point.
David





From just at letterror.com  Thu Apr 20 02:45:54 2000
From: just at letterror.com (Just van Rossum)
Date: Thu, 20 Apr 2000 07:45:54 +0100
Subject: __del__ problem - would adopting Garbage Collection fix this?
In-Reply-To: <slrn8ft35a.uqs.nascheme@cranky.arctrix.com>
References: <aw2L4.8907$HG1.245743@nnrp1.uunet.ca>
 <7ojL4.9424$HG1.254543@nnrp1.uunet.ca>
 <Z%oL4.9643$HG1.257533@nnrp1.uunet.ca>
 <m31z41iyhg.fsf@atrus.jesus.cam.ac.uk>
 <38FE7F1E.168E914A@cosc.canterbury.ac.nz>
Message-ID: <l03102803b52456fb3df8@[193.78.237.146]>

At 4:50 AM +0000 20-04-2000, Neil Schemenauer wrote:
>When garbage is found by the collector (ie. reference cycles not
>reachable from Python), all objects reachable from __del__
>methods are moved into a separate set.  This is what I call
>uncollectable garbage.  Uncollectable instances are currently
>added to a global list of uncollectable objects reachable from
>within Python.  If the programmer creates such garbage it is up
>to them to deal with it.
>
>The bottom line is don't add __del__ methods to objects that form
>cycles.  It is, however, okay to reference objects with __del__
>methods from objects involved in reference cycles.

Hm, wouldn't it be better to collect cyclic objects with __del__ anyway,
and just not call __del__? Possibly printing something to stderr, similarly
to when an exception occurs within a __del__ method.

Just





From sp00fD at yahoo.com  Tue Apr 18 07:55:27 2000
From: sp00fD at yahoo.com (sp00fD)
Date: Tue, 18 Apr 2000 11:55:27 GMT
Subject: gaming and python
References: <000c01bfa91c$f421d6a0$0f8d0b3f@oemcomputer>
Message-ID: <8dhifa$aom$1@nnrp1.deja.com>

Well, I'd have a hard time believing that you could program something
like quake, but you could probably do a decent job with tetris or the
like.  Check out PySDL http://pysdl.sourceforge.net/, I haven't used
it, but it's probably exactly what you need, SDL is apparently a cross
platform library for game programming.  Also, check out PySol
http://wildsau.idv.uni-linz.ac.at/mfx/pysol.html, an excellent and very
well done game of solitare + 156 (or so).

Mike

In article <000c01bfa91c$f421d6a0$0f8d0b3f at oemcomputer>,
  "mrnatch" <mrnatch at gateway.net> wrote:
> This is a multi-part message in MIME format.
>
> ------=_NextPart_000_0009_01BFA8E2.46AECF80
> Content-Type: text/plain;
> 	charset="iso-8859-1"
> Content-Transfer-Encoding: quoted-printable
>
> I told a friend of mine that I am learning to program using Python (I
=
> chose Python as my first language on the advice of a Llama) and he
asked =
> me an interesting question: Can Python be used for programming video =
> games? And does it work with Windows API? Just curious.
> Thanks,=20
> Josh (And friend Tony)
>
> ------=_NextPart_000_0009_01BFA8E2.46AECF80
> Content-Type: text/html;
> 	charset="iso-8859-1"
> Content-Transfer-Encoding: quoted-printable
>
> <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
> <HTML><HEAD>
> <META content=3D"text/html; charset=3Diso-8859-1" =
> http-equiv=3DContent-Type>
> <META content=3D"MSHTML 5.00.2722.1300" name=3DGENERATOR>
> <STYLE></STYLE>
> </HEAD>
> <BODY bgColor=3D#ffffff>
> <DIV><FONT face=3DArial size=3D2>I told a friend of mine that I am =
> learning to=20
> program using Python (I chose Python as my first language on the
advice =
> of a=20
> Llama) and he asked me an interesting question: Can Python be used
for=20
> programming video games? And does it work with Windows API? Just=20
> curious.</FONT></DIV>
> <DIV><FONT face=3DArial size=3D2>Thanks, </FONT></DIV>
> <DIV><FONT face=3DArial size=3D2>Josh (And friend =
> Tony)</FONT></DIV></BODY></HTML>
>
> ------=_NextPart_000_0009_01BFA8E2.46AECF80--
>
>


Sent via Deja.com http://www.deja.com/
Before you buy.


From mwh21 at cam.ac.uk  Sun Apr  2 09:42:38 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 02 Apr 2000 14:42:38 +0100
Subject: Emacs Python mode
References: <yayk8iii0ex.fsf@ganglot.ifi.uio.no> <m3zorendyk.fsf@atrus.jesus.cam.ac.uk> <yayzorcj1ox.fsf@ganglot.ifi.uio.no>
Message-ID: <m3r9coobzl.fsf@atrus.jesus.cam.ac.uk>

Vetle Roeim <vetler at ifi.uio.no> writes:

> ..once python.org gets back up, of course.

There are mirrors all over; my usual one is

http://sunsite.doc.ic.ac.uk/Mirrors/ftp.python.org/pub/www.python.org/

(but they don't carry the pipermail arhives, grr.)

M.

-- 
well, take it from an old hand: the only reason it would be easier
to program in C is that you can't easily express complex  problems
in C, so you don't.                 -- Erik Naggum, comp.lang.lisp


From mfletch at tpresence.com  Thu Apr 20 10:13:34 2000
From: mfletch at tpresence.com (Mike Fletcher)
Date: Thu, 20 Apr 2000 10:13:34 -0400
Subject: Python Parsers Comparison
Message-ID: <CC59868A0A76D311B1F50090277B321C1F328F@VRTMAIL>

Here's part of it :)

Note: for heavy-lifting applications you'd likely want to define a custom
mxTextTools tuple tree to just appending the interpreted numbers to the
parse tree (instead of the 4-tuples) (likely also want to make the grammar
distinguish between hex, float and int).

from simpleparse import generator
from TextTools import TextTools
import string
class Parser:
	def __init__( self, string ):
		self.data = string
		self.position = 0
	def parse( self, parseonly = 0 ):
		success, tags, next = TextTools.tag( self.data,
ROOTITEMPARSER)
		if parseonly:
			return success, tags, next
##		import pprint
##		pprint.pprint( tags )
		return success, map( self._dispatch, tags ), next
	def _dispatch( self, (tag, start, stop, children)):
		return getattr( self, tag)( (tag, start, stop, children) )

	# tag handlers
	def element( self, (tag, start, stop, (child,) )):
		'''Always has a single child, either vector or number'''
		return self._dispatch( child )
	def vector( self, (tag, start, stop, elements)):
		'''Some arbitrary number of elements, return as list'''
		return 	map( self._dispatch, elements )
	def number( self, (tag, start, stop, children)):
		'''Could be either int or float, check first, then second'''
		try:
			return string.atoi( self.data[start:stop], 0 )
		except ValueError:
			return string.atof( self.data[start:stop] )
	


PARSERDECLARATION = r'''root           := ts, element*
vector         := '[', ts, element*, ']'
element        := (vector/number), ts


number         := [-+]*, ( ('0',[xX],[0-9]+) / ([0-9.]+,([eE],[-+0-9.]+)?))
<ts>           :=  ( [ \011-\015,]+ / ('#',-'\012'*,'\n')+ )*
'''

PARSERTABLE = generator.buildParser( PARSERDECLARATION )
ROOTITEMPARSER = PARSERTABLE.parserbyname( "root" )
if __name__ == "__main__":
	for testString in [
		'''2''',
		'''[2]''',
		'''[ 2 ]''',
		'''[]''',
		'''[ 2 ]''',
		'''[ []]''',
		'''[ 1,2 ]''',
		'''[ [ 1,2 ],[ 3,4 ] ], [ 5.6 ]''',
		]:
		print Parser( testString).parse()
	speedtest = '''[ [ 1,2 ],[ 3,4 ] ], [ 5.6 ]'''* 5000
	import time
	print 'starting speed test, string length', len(speedtest)
	t = time.time()
	elements = len( Parser( speedtest).parse()[1])
	print elements, "elements parsed in", time.time()-t, "seconds"
	print 'starting parse-only speed test'
	t = time.time()
	elements = len( Parser( speedtest).parse(parseonly=1)[1])
	print elements, "elements parsed in", time.time()-t, "seconds"


-----Original Message-----
From: Randall Hopper [mailto:aa8vb at yahoo.com]
Sent: Thursday, April 20, 2000 9:40 AM
To: python-list at python.org
Subject: Python Parsers Comparison


I'm again faced with the "regexes not powerful enough for nested grammar"
problem, and need to chose an extension.

Does someone know of a web page or paper which compares the various Python
parser engines on a simple example grammar (or on features)?

For example, a comma-separated list of nested vectors would be a good
example:

    [ 1, 2 ], [ 3, 4 ]
    [ [ 1,2 ],[ 3,4 ] ], [ 5.6 ]
    etc.

mxTextTools, metalanguage, SimpleParse, SPARK, YAPPS, PyLR, kwParsing,
PyBison.  A comparison of any subset would be useful.

Thanks,

Randall


-- 
Randall Hopper
aa8vb at yahoo.com

-- 
http://www.python.org/mailman/listinfo/python-list



From Jhlee at Brooks.com  Wed Apr 12 17:06:31 2000
From: Jhlee at Brooks.com (Lee, Jaeho)
Date: Wed, 12 Apr 2000 17:06:31 -0400
Subject: running python script on ksh on NT
Message-ID: <857F15D7E3D8D3118D290008C7CF0586270761@mail-naeast1.brooks.com>

Hi all. I am using MKS ksh on NT 4.0.
I could run on NT cmd. But not on ksh. I tried several option such as the
first line #!D:/progra~1/python/python.
How can I run .py file on ksh. python directory is in my path.
Thanks,

Jaeho Lee
  



From bjorn at roguewave.com  Thu Apr 27 20:40:48 2000
From: bjorn at roguewave.com (Bjorn Pettersen)
Date: Thu, 27 Apr 2000 18:40:48 -0600
Subject: Python does not let me format time.
References: <8eakhq$797$1@nnrp1.deja.com>
Message-ID: <3908DE10.2BBAB862@roguewave.com>

Sindh wrote:
> 
> Hi folks
> 
> I have a file full of date, value pairs. Date is like MM/DD/YY. My
> problem is I want to format this file since it contains a few thousand
> lines of data , using python. I tried using string and time modules. I
> want to format the time value to a float which is returned by
> time.gmtime normally. I tried to split it and to use parts of it to make
> a tuple and to use time.mktime but it says illegal operation for
> builtin.
> 
> Summary:
> 
> data is:
>         03/11/98 abcd
>         03/12/98 efgh
> 
> I split it to
>         x=string.split(n[0],' ')
>         which gave me x='03/11/98'
> Now I split it again by
>         x=string.split(x,'/')
>         which gave a tuple of '03','11','98'
> Now I tried  to make time as
> time.mktime(x[2],x[0],x[1],'','','','','','')
> The problem is it says illegal operation for builtin. I tried to use the
> string val and int val as the arguments for mktime.
> 
> Anyone with any helpful ideas!!
> 
> If there is already a moduls to do the same , much helpful.

time.mktime() needs integer arguments, and the result of string.split is
a list of strings.  Changing it to:

  time.mktime(int(x[2]),int(x[0]),int(x[1]),0,0,0,0,0,0)

should probably work (I didn't check how many arguments it requires ;-)

-- bjorn



From gonter at zechine.wu-wien.ac.at  Thu Apr 27 13:12:16 2000
From: gonter at zechine.wu-wien.ac.at (Gerhard Gonter)
Date: Thu, 27 Apr 2000 19:12:16 +0200
Subject: Using python on the web
References: <MPG.1365adcfa8470eef9896aa@news> <MPG.136631472f892a119896ac@news> <_W2L4.986$GK5.20898@news2-win.server.ntlworld.com> <MPG.13665ef93eedbda29896ad@news> <BC4B61746187C623.3FA0E7D58DF4EA7E.ECC1D97F577DEDE5@lp.airnews.net> <slrn8g28ah.cc.amphgobb@eccie.blackflag.net> <8ds81f$qne$1@mozo.cc.purdue.edu>
Message-ID: <390874F0.E23C67FF@zechine.wu-wien.ac.at>

Kyler Laird wrote:
> Hmmm...cgi-wrapper doesn't work with mod_perl (or
> mod_php, or ...), right?
> 
> (This is why I built my own system to do the job.)

Could you expand on this issue, e.g. post an url?

+gg


From embed at geocities.com  Thu Apr 13 09:18:23 2000
From: embed at geocities.com (Warren Postma)
Date: Thu, 13 Apr 2000 09:18:23 -0400
Subject: Comparing PythonWin and IDLE?
Message-ID: <IOjJ4.6669$HG1.192116@nnrp1.uunet.ca>

On Win32 I have noticed the following two things about IDLE and PythonWin:

- When running a script with a lot of PRINT statements, IDLE is much nicer,
in that the output window is updated during the run of the program, whereas
the output window appears "frozen" and does not repaint in PythonWin until
program execution halts (program completes normally, or an exception
occurs). This makes IDLE more useable.

- PythonWin's Editor window is much faster and more crisp on my machine
(Pentium-II 400). I suspect this is because Tcl/Tk doesn't use native
widgets, and the Tcl/Tk editor widget is therefore somewhat more
'inefficient', in the name of greater portability.


Something I have noticed on BOTH IDLE and PythonWin, is that once a module
appears loaded into memory there is no "soft reset" feature in either one.
I used to use a language called CA-Realizer from Computer Associates. It was
an interpreter, and you could hit Ctrl-Z to restore the state of the
interpreter to what it was when it started up. That is, you could hit Ctrl-Z
instead of shutting down the IDE and restarting it. I find once a module has
been imported once, it often will not import again, even if I have changed
the source (.py) file, the .pyc file does not appear to be regenerated, and
any errors in the .py file can only be fixed by editing and saving the .py
file, quitting the IDE and restarting. This appears to happen for both IDLE
and Python. Anybody else seen this?


Warren Postma





From smalleys at gte.net  Fri Apr 14 00:52:54 2000
From: smalleys at gte.net (sue)
Date: Fri, 14 Apr 2000 04:52:54 GMT
Subject: need help just begining
References: <11822ca0.dd34a400@usw-ex0109-068.remarq.com> <38F41190.C0BB507B@sympatico.ca> <38F53E5A.977913FE@cosc.canterbury.ac.nz> <38F54D83.B0F620A8@callware.com>
Message-ID: <38F69621.3B8F5B29@gte.net>

Goody!  Another Python book!  I'll try to get one Saturday.
I think Learning Python is the best intro to a modern
programming language that I've seen to date.  I'll be
interested to see how 24 Hours compares.

I have recommended Learning Python to four people, including
a manager and a teenager, neither of whom had much
programming experience.  Three of the four now like and use
python.  (The fourth thinks TCL is great.  Oh well.)

When I showed them Perl, TCL, C, C++, Java, they said
"What's all this weird stuff?  Semi-colons, braces, types? 
I have to type pages of code to duplicate ten lines of
Python!"

You need to start them off right.  Guido's educational
project is dead on.  Try explaining why C++ implicit type
conversions are evil to a new programmer who thinks they're
"neat".

Sue.


From timehorse at unforgettable.com  Thu Apr 27 19:01:01 2000
From: timehorse at unforgettable.com (timehorse at unforgettable.com)
Date: Thu, 27 Apr 2000 23:01:01 GMT
Subject: mail forwarding not working on the starship?
References: <3908538A.E74456A5@GermanLloyd.org>
Message-ID: <8eagr4$30h$1@nnrp1.deja.com>

In article <3908538A.E74456A5 at GermanLloyd.org>,
  Berthold Hoellmann <hoel at GermanLloyd.org> wrote:
> Now that my .starship_info is set up to forward my mail, any test mail
> send to my starship account still seems to be lost somwhere without
any
> error message. My starship address is "bhoel at starship.python.net" and
> mail should be forwarded to "se6y095 at public.uni-hamburg.de".
>
> Is anyone out there who can help me?

Hi Berthold,

I have been doing a lot of research on this issue as Andrew installed
an SSH key for me yesterday and now I have access to my A/C [Thanks
again Andrew!  :) ].  Anyway, one thing that everyone on Starship
should know is that we are no longer using SENDMAIL as our transport
agent.  We have been moved to qmail as it is LEAPS AND BOUNDS easier to
maintain.  This is obvious simply from the REAMS of man pages
associated with QMail installed on every possible mail topic from where
to store your mailbox to how to detect, trace and prevent forged mail.
Now, one thing I have observed is that QMail no longer uses the
Sendmail .forward scripts directly, and thus your mail will not be
forwarded using the old method I used to use.  Instead, you have
a .qmail file in your home directory and that can act as your .forward
file.  The format is simple:

#<comment>

for a comment

|<prog>

for a pipe

[&]<address>

for a forwarding address, the Ampersand (&) is only required when the e-
mail address does not start with an alphanumeric character.  The
address must be a valid internet address with a fully-qualified domain
name.

/<path>/<file>
.<path>/<file>

Store the mail in <file> with path <path>.

/<path>/
.<path>/

Use the path specified as the mail storage folder.

Now, if you still want to use your old .forward file from before, you
can do this too.  Basically you put a line in your .qmail file
something like:

|dot-forward [-nN] .forward

-N: Read and forward the message [default]
-n: Parse the file and print out the instructions it contains, usually
used for debugging on the command line.

Again, please read the man page to check the exact syntax.

So, for you Bertold, I would suggest setting up a .qmail file with:

&se6y095 at public.uni-hamburg.de

As its only entry.  That SHOULD in THEORY get things to where they
ought to be.  I can verify that when I send mail on starship to my
starship account, the forwarding works.

HOWEVER, I still can't receive mail from outside the ship!  I have a
friend who told me there was an infinite hop count with one of his
mails, but that was before I set up the .qmail file and all the test
mails I send seem to disappear into oblivion like yours.  It is very
distressing as I have had to deal with the same thing when posting to
the mailing lists I maintain on eGroups using the Web interface.  As it
is, deja.com may be my only way of communicating with people.  :(
Anyway, that you and I are both experiencing the evil mail void, I
don't have any idea what could be wrong.  Originally I though it may be
that the .forwarding was set by a super-user account so I was searching
Starship for occurrences of my official e-mail address in case it may
be
forwarding to the wrong address, but so far I have found naught.  Does
anyone else know more about QMail to explain why e-mail is not being
delivered?  Berthold and I would appreciate any help greatly!  Thanks
and hope my little man-page summary will help others.
Remember: .qmail, NOT .forward.  :)

Be Seeing You,

Jeffrey.


Sent via Deja.com http://www.deja.com/
Before you buy.


From bpetterson at uswest.net  Sun Apr  2 17:25:24 2000
From: bpetterson at uswest.net (Bjorn Pettersen)
Date: Sun, 02 Apr 2000 15:25:24 -0600
Subject: Python 1.6 alpha 1 released
References: <200003312130.QAA04361@eric.cnri.reston.va.us> <38E5205F.DE811F61@roguewave.com> <Wf9F4.4423$74.70231@newsc.telia.net> <38E58D2D.384F1546@Lugoj.Com> <S9kF4.4472$74.71414@newsc.telia.net> <38E6B929.28816494@roguewave.com> <auGF4.914$n68.186931712@newsb.telia.net>
Message-ID: <38E7BAC4.3B7AD0FF@uswest.net>

Fredrik Lundh wrote:

> Bjorn Pettersen wrote:
> > I haven't seen anyone claim that examples in the manual are not
> > documentation...
>
> in earlier releases of the library reference, some examples
> didn't even work.  you still think they're formally binding?

Are you saying none of the examples are binding, formally or otherwise?  I
certainly expect that when code in the documentation works as presented that
I can rely on it -- I don't think that's a particularily surprising position
to have.  (Even in the C++ standard where examples are _explicitly_ not
normative, they're usually fixed if conflicting with normative text.)

> > It would be very nice if "someone" could do that, but based on the
> > rethoric I've seen from you on this issue, and the absence of comment
> > from Guido, I certainly don't feel confident that it would be accepted.
>
> you don't know that until you've tried it...

I usually try to limit my excercises in futility.

Your other arguments for a patch are good, but seems to reinforce the feeling
that changing this behavior was ill though out in the first place. (and
remember that not all companies have enough Python expertise that a traceback
leads to a fix ten minutes later)

-b



From alvaro at ocse.infonegocio.com  Wed Apr 26 08:03:16 2000
From: alvaro at ocse.infonegocio.com (AMR)
Date: Wed, 26 Apr 2000 14:03:16 +0200
Subject: Closing Excel 97 with Python
Message-ID: <8e6lhb$8h93@esiami.tsai.es>

Hello,

I'm using COM to exchange information with Excel and my problem is the next.

This is my way to open Excel:

self.o = win32com.client.dynamic.Dispatch("Excel.Application")

self.o.Visible = Visible

self.o.Workbooks.Open(self.MyFile)

And this is my way to close Excel:

self.o.Visible = 0

self.o.Workbooks(1).Close(0)

self.o.Workbooks.Close()

self.o.Quit ## For try something

After this Excel is still opened in the task administrator, and when I try
to open Excel again from python i only see the menu , status bar and the
tool bar. I have to open the full view of excel to see the worksheet.

If you have a clue to close Excel 97...

Thank you anyway,






From tom__98 at my-deja.com  Tue Apr 11 23:09:26 2000
From: tom__98 at my-deja.com (tom__98 at my-deja.com)
Date: Wed, 12 Apr 2000 03:09:26 GMT
Subject: Python/Perl Popularity (Re: A Mountain of Perl...)
References: <200004102111.HAA03380@envy.fulcrum.com.au>
Message-ID: <8d0pcq$7ti$1@nnrp1.deja.com>

In article <200004102111.HAA03380 at envy.fulcrum.com.au>,
  Richard.Jones at fulcrum.com.au wrote:
> Tom you little troll-meister. Half those arguments
> you cite are bollocks if you _really_ knew and
> "greatly prefer" Python. Try again.

Maybe you can contribute more constructively:

 -- How can I browse the Python documentation without firing
    up a web browser?  How do I get information on a single Python
    function or class quickly?  Is there any equivalent of
    "perldoc -f split" or "perldoc Net::POP3"?  What do you
    do in that situation?

 -- Is there a Python equivalent of CPAN and the Perl CPAN
    module?  The closest I know of is Parnassus.  But Parnassus
    is merely a collection of links, not an archive, and it
    doesn't have any facilities (AFAIK) for automatic installation.

 -- Several extensions I have wanted to install have required access
    to the full Python source tree, not just the header files and
    the library.  Is that going to change?  Otherwise, how can
    I install Python extensions that want access to the source
    tree on, say, a standard RedHat system?  Having the standard
    RPM-based installation in parallel with a /usr/local installation
    from source is confusing, and removing the RPM-based installation
    risks breaking things.

Tom.


Sent via Deja.com http://www.deja.com/
Before you buy.


From kilian.lorenz at dkfz-heidelberg.de  Sun Apr  9 15:52:34 2000
From: kilian.lorenz at dkfz-heidelberg.de (Kilian Lorenz)
Date: Sun, 09 Apr 2000 21:52:34 +0200
Subject: Cascading menus
References: <8cko46$j2i$1@nnrp1.deja.com>
Message-ID: <38F0DF82.EEAB208D@dkfz-heidelberg.de>


Perhaps you can solve your problem by using the postcommand option.

When the menu is posted this function/method is called and
probably you can change your sub menu there. I don't expect that anyone
estimates code to be run anytime a menu item is selected a very clever
idea since with an indecisive user this could happen quite frequently
and I can't think of a good reason why there should be done something
else in the second call than in the first call (except perhaps making a
statistic about menu usage).

But I don't speak for the whole Python/Tkinter/Tk-community and
who knows ... nobody expects the Spanish Inquisition either ...

Regards, Kilian


From effbot at telia.com  Thu Apr 13 02:58:42 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Thu, 13 Apr 2000 06:58:42 GMT
Subject: Python idiom: Multiple search-and-replace
References: <20000412100816.A2431031@vislab.epa.gov><5C%I4.645$rc9.190288896@newsb.telia.net> <20000412120012.D2446939@vislab.epa.gov><0s1J4.655$rc9.190631424@newsb.telia.net> <B51AB591.1BAC%dgoodger@bigfoot.com>
Message-ID: <CeeJ4.1553$Za1.26079@newsc.telia.net>

David Goodger <dgoodger at bigfoot.com> wrote:
> on 2000-04-12 12:25, Fredrik Lundh (effbot at telia.com) wrote:
> > while you're at it, try replacing the original readline loop with:
> >
> > while 1:
> > lines = fp.readlines(BUFFERSIZE)
> > if not lines:
> > break
> > lines = string.join(lines, "")
> > lines = re.sub(...)
> > out_fp.write(lines)
> >
> > where BUFFERSIZE is 1000000 or so...
>
> why not just,
>
>     while 1:
>         lines = fp.read(BUFFERSIZE)
>         if not lines:
>             break
>         lines = re.sub(...)
>         out_fp.write(lines)
>
> ? Saves the string.join() step. Or am I missing something? (I await
> enlightenment...)

    set BUFFERSIZE to 1
    try replacing "foo" with "bar"
    await enlightenment

(not that it cannot be fixed, of course.  just add code to strip off
the last incomplete line and prepend it to the next read.  make sure
you don't mess up if the last character in the file is not a newline)

</F>




From warlock at eskimo.com  Tue Apr  4 22:55:51 2000
From: warlock at eskimo.com (Jim Richardson)
Date: Wed, 05 Apr 2000 02:55:51 GMT
Subject: opening a serial port at a given baudrate. How?
References: <200004041153.VAA15869@piglet.dstc.edu.au>  <kubt3pisg6.fsf@lasipalatsi.fi>
Message-ID: <slrn8elb7f.1t1.warlock@raven.myth>

On 04 Apr 2000 22:20:25 +0300, 
 Erno Kuusela, in the persona of <erno at iki.fi>,
 brought forth the following words...:

>>>>>> "davida" == David Arnold <arnold at dstc.edu.au> writes:
>    davida> on unix (including linux), just open the device special
>    davida> file corresponding to the serial port of interest
>    davida> (ie. /dev/ttyS0) and you can then control it using
>    davida> fcntl.ioctl() function and otherwise read from and write
>    davida> to it like a normal file object.
>
>i believe the termios module is for ftwiddling with serial
>ports on unix. see for example ftp://ftp.tummy.com/pub/tummy/cid/
>for example code.

Thanks


-- 
Jim Richardson
	Anarchist, pagan and proud of it
WWW.eskimo.com/~warlock
	Linux, because life's too short for a buggy OS.



From robin at jessikat.demon.co.uk  Fri Apr 14 18:41:32 2000
From: robin at jessikat.demon.co.uk (Robin Becker)
Date: Fri, 14 Apr 2000 23:41:32 +0100
Subject: ANN: Python training in Boulder
References: <000101bfa571$e31a43a0$71d95da6@lutz>
 <38F78406.2032D7AB@boi.hp.com>
Message-ID: <nLKjjKAc6594EwkL@jessikat.demon.co.uk>

In article <38F78406.2032D7AB at boi.hp.com>, Bill Anderson
<banderson at boi.hp.com> writes
>Mark Lutz wrote:
>> 
>> This is to announce that the second Python training session
>> to be held at Tom Christiansen's site in Boulder, Colorado,
>> will be held May 24-26.
>> 
>> These classes are equivalent to the on-site Python training I
>> offer, and are open to individuals on a first-come first-served
>> basis.  I am the instructor, but all registration details are
>> handled by Tom's company.  For more information, please visit
>> these pages:
>> 
>> http://home.rmi.net/~lutz/boulder-classes.html   (my page)
>> http://training.perl.com/open-courses.html       (tom's page)
>                  ^^^^^^^^^
>
>Maybe it's me, but something seems fundamentally wrong here. ;^)
>
>
>
I already saw something a bit blien about the idea Of TC + Python, but
his comparison perl-->Python was very good.
-- 
Robin Becker


From chafik at nevrax.com  Wed Apr  5 04:35:15 2000
From: chafik at nevrax.com (Sameh chafik pro)
Date: Wed, 5 Apr 2000 10:35:15 +0200
Subject: (no subject)
Message-ID: <006b01bf9ed9$dd9b5660$1001a8c0@nevrax.net>

Hello evry body,

Are there any way to have the python16_d.lib and .dll, or the python15_d.dll ?????
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20000405/93319803/attachment.html>

From tjg at avalongroup.net  Tue Apr 11 15:24:41 2000
From: tjg at avalongroup.net (Timothy Grant)
Date: Tue, 11 Apr 2000 12:24:41 -0700
Subject: Trouble with Tkinter under IDLE
References: <8F139460Canthonylander@47.80.11.122>
Message-ID: <38F37BF9.DCAB3570@exceptionalminds.com>

Anthony Lander wrote:

> All of this works just fine from the command line, though, so I'm pretty
> convinced it's a problem related to running Idle.
> 
> I'm running Python 1.5.2, idle 0.6 and tcl/tk 8.0 all under NT.
> 
> Anyone have any guesses?
> 
> Thanks much,

I believe that the solution is *not* to call mainloop() from your
programme as it is conflicting with the mainloop() of IDLE. Though I
don't use IDLE, I do believe that the mainloop in IDLE will allow your
code to run correctly without you calling it.


-- 
Stand Fast,
    tjg.

Timothy Grant                         tjg at exceptionalminds.com
Chief Technology Officer              www.exceptionalminds.com
Red Hat Certified Engineer                      (503) 246-3630
Avalon Technology Group, Inc.              fax  (503) 246-3124
>>>>>>>>>>>>Linux...Because rebooting isn't normal<<<<<<<<<<<<



From sblakey at freei.com  Thu Apr 27 19:30:48 2000
From: sblakey at freei.com (Sean Blakey)
Date: Thu, 27 Apr 2000 23:30:48 +0000
Subject: Simple Modules Question
References: <8eaaul$16p@autodesk.autodesk.com>
Message-ID: <3908CDA8.25428F36@freei.com>

Akira Kiyomiya wrote:
> 
> I see NameError when I try to do string.atoi("45").
> 
> Is it because 'string' is at a different location??
> 
> Thanks
> 
> Akira
> 
> from string import atoi    # imports 'string'
>                                      # places atoi in current namespace
> print atoi("12345")         # invokes 'atoi' without module name
> 
> string.atoi("45")             # NameError: string
> 
> --
> http://www.python.org/mailman/listinfo/python-list

When you run "from string import atoi" python remembers the name "atoi",
but does not remember where it got atoi from (string).  Python only
keeps access to the string module long enough to get the names you
selected.  This confusing weirdness is one of the many reasons I prefer
to use a simple "import string".

You seem to expect "from string import atoi" to be essentially like the
following code:
>>>import string
>>>atoi = string.atoi

The more accurate equivalent of "from string import atoi" is
>>>import string
>>>atoi = string.atoi
>>>del string

I could be wrong about this (haven't looked at the source in a while),
but thinking about it this way has worked for me so far.
-- 
Sean Blakey
FreeInternet.com
sblakey at freei.com
(253)796-6500x1025



From markx.daley at intel.com  Tue Apr 25 17:02:46 2000
From: markx.daley at intel.com (Daley, MarkX)
Date: Tue, 25 Apr 2000 14:02:46 -0700
Subject: mxODBC error trapping
Message-ID: <A63AFB20111ED311AC5500A0C96B7BF5015F0F4D@orsmsx36.jf.intel.com>

>>Daley, MarkX <markx.daley at intel.com> wrote:
>> I am accessing databases using the notes by John Dell'Aquila for ODBC
using
>> Python Database API.  It works fine, but now I am trying to set up error
>> trapping in the event that a connection fails to be made.  Here is the
basic
>> code structure:
>>
>> def collect():
>>     "Routine to query all servers for all names for time period defined
>>     in timer."
>>    # Define some variables
>>    for index in servername:
>>        try:
>>            # Big routine to collect data and write to file
>>        except dbi.operation-error:
>>            servername.append(index)

>"operation-error" is not a valid name (it's interpreted
>as "operation" minus "error").

>> I guess I don't know enough about exceptions to trap them properly.  Any
>> hints?

>according to the DB specification [1], OperationalError is the
>correct name.  you still have to qualify it with the module name,
>though:

>    try:
>        ...
>    except dbi.OperationalError:
>        ...

>if that doesn't work, try printing the contents of the "dbi" module:

>print dir(dbi)

></F>

>1) http://www.python.org/topics/database/DatabaseAPI-2.0.html


When I use dbi.OperationalError, I get this result:

Traceback (innermost last):
  File "<pyshell#1>", line 1, in ?
    collect.collect()
  File "C:\PROGRA~1\Python\collect.py", line 50, in collect
    except dbi.OperationalError:
AttributeError: OperationalError

Here is the printout of the dbi module:

['DATE', 'NUMBER', 'RAW', 'ROWID', 'STRING', 'TYPES', '__doc__', '__file__',
'__name__', 'dataError', 'dbDate', 'dbRaw', 'dbiDate', 'dbiRaw',
'integrityError', 'internalError', 'noError', 'opError', 'progError']

I am way out of my depth here.  Any and all help is appreciated.

- Mark




From tuttledon at hotmail.com  Thu Apr 20 18:59:17 2000
From: tuttledon at hotmail.com (Don Tuttle)
Date: Thu, 20 Apr 2000 22:59:17 GMT
Subject: Why MFC instead of Tkinter -- was Re: Pythonwin?
References: <eMz2Nmiq$GA.73@net025s> <4QtL4.24$v85.41@news-server.bigpond.net.au> <chuL4.8483$O7.17315@typhoon.southeast.rr.com> <7LDL4.534$v85.2563@news-server.bigpond.net.au> <sfu6d657l6e184@corp.supernews.com>
Message-ID: <9ZLL4.15615$O7.23007@typhoon.southeast.rr.com>

"Michael A. Crawford"
> I'm curious.  Why did you not steer him in the Tkinter direction for his
GUI
> needs?

Er...maybe it was because A[r]TA's question was about MFC and not Tkinter!
;-)

Don




From tismer at tismer.com  Mon Apr 17 10:48:42 2000
From: tismer at tismer.com (Christian Tismer)
Date: Mon, 17 Apr 2000 16:48:42 +0200
Subject: Was Re: [ANNOUNCE] Garbage collection for Python, now Stackless 
 comments
References: <20000407221750.A29279@acs.ucalgary.ca>  <m3r9cagx3c.fsf@atrus.jesus.cam.ac.uk> <l03102809b51d41895a80@[193.78.237.135]> <38F9F01B.57273052@tismer.com> <slrn8fjtao.2d4.quinn@drachma.ugcs.caltech.edu> <38FA05C9.442868E4@tismer.com> <38FAEEC9.B5CED499@webwitches.com>
Message-ID: <38FB244A.77BB7196@tismer.com>


Shae Erisson wrote:
> 
> Christian Tismer wrote:

[stuff about continuations, whether to hide them or not]

> So when do we get to Pickle continuation objects and send them out
> across the network, or even better, to another VM running on a second
> processor? :)

This is on my task list for 1.3 or 1.4, but it isn't
trivial. It is also quite different, since its relation
to continuation is much less than you might think.
ontinuations and frames are quite simple to pickle, but
this doesn't solve the whole problem. I must figure out
what to pickle at all, and I have to catch the state
of the current interpreter and restore it. This is
hard, since I might be in an import or in a C extension
that calls an interpreter, and so on.
Unless Stackless isn't completely stackless, the pickling
problem is not any simpler than before.

> Personally, I think continuations and coroutines are beautiful, and
> easier and simpler than threads. Yay Christian! (and Yay Guido for
> writing Python!)

Please let me take 10000 clones from you :-)

> PS. Can someone point me to example code for building smalltalk style
> closures easily with continuations?

Can you point me to a paper about smalltalk closures?
I don't have a blank idea at all.

ciao - the clueless implementor

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaunstr. 26                  :    *Starship* http://starship.python.net
14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     where do you want to jump today?   http://www.stackless.com



From gward at mems-exchange.org  Fri Apr  7 09:15:46 2000
From: gward at mems-exchange.org (Greg Ward)
Date: Fri, 7 Apr 2000 09:15:46 -0400
Subject: [Python-Dev] Round Bug in Python 1.6?
In-Reply-To: <XFMail.000407145738.mikael@isy.liu.se>; from mikael@isy.liu.se on Fri, Apr 07, 2000 at 02:57:38PM +0200
References: <20000407084914.A13606@mems-exchange.org> <XFMail.000407145738.mikael@isy.liu.se>
Message-ID: <20000407091545.B13606@mems-exchange.org>

On 07 April 2000, Mikael Olofsson said:
> 
> On 07-Apr-00 Greg Ward wrote:
>  >  Oh, joy! oh happiness! someday soon, I may be able to type
>  >  "blah.__doc__" at the interactive prompt and get a readable result!
> 
> Just i case... I hope you haven't missed "print blah.__doc__".

Yeah, I know: my usual mode of operation is this:

  >>> blah.__doc__
  ...repr of docstring...
  ...sound of me cursing...
  >>> print blah.__doc__

The real reason for using str() at the interactive prompt is not to save
me keystrokes, but because it just seems like the sensible thing to do.
People who understand the str/repr difference, and really want the repr
version, can slap backquotes around whatever they're printing.

        Greg



From gchiaramonte at mail.com  Tue Apr 18 17:16:28 2000
From: gchiaramonte at mail.com (Gene Chiaramonte)
Date: Tue, 18 Apr 2000 17:16:28 -0400 (EDT)
Subject: Distributing A Python Application
Message-ID: <384449751.956092588323.JavaMail.root@web135-mc.mail.com>

I want to distribute a windows GUI application developed in Python/Tkinter.
How can I distribute the python enviromnent and my application into a
directory such as c:\program files\myapp\bin, ...\lib, ...\dlls etc. I want
my installation of python to be insulated from any versions of python the
user may have or install in the future. (ie. I don't want the user breaking
my app by upgrading to python 1.6 or something like that)

Thanks,

Gene

______________________________________________
FREE Personalized Email at Mail.com
Sign up at http://www.mail.com/?sr=signup




From none at none  Mon Apr 17 19:44:09 2000
From: none at none (Eric Jacobs)
Date: Mon, 17 Apr 2000 18:44:09 -0500
Subject: newbie question
References: <8dfdfk$uum$1@nnrp1.deja.com><wbHK4.2614$Za1.38989@newsc.telia.net>
Message-ID: <sfn4abj87l5147@corp.supernews.com>

In article <wbHK4.2614$Za1.38989 at newsc.telia.net>,
"Fredrik Lundh" <effbot at telia.com> wrote:
> Arint? <jamarijr at hotmail.com> wrote:
>> If i PyErr_SetString(PyExc_TypeError,"Invalid command value for IOCTL");
>> in my c/c++ program, how would python script get that error string?
> 
> return NULL to the interpreter.  do *not* return
> a python value.

That, and then do something like

try:
     ...
except TypeError, exc
     msg = exc.args[0]


in Python. I think that's what you were looking for?


From pcolsen at draper.com  Fri Apr  7 21:13:07 2000
From: pcolsen at draper.com (Peter Olsen)
Date: Fri, 07 Apr 2000 21:13:07 -0400
Subject: Actually Running a function from a DLL
Message-ID: <38EE87A3.4080802@draper.com>

I'm writing ask for a quick explanation of how to load and use dll files 
from Python running under MS Windows.

I have installed PythonWin and win32all, but I seem to be completely 
unable to run a function from a dll.? I'm loading the dll with 
win32api.LoadLibrary, but I can't figure out how to access the 
functions.? I've made some statements about how fast we can learn Python 
and the start development, but if I don't find an answer quickly,? I may 
have to eat my words.? If the best answer is "RTFM," please give me a 
hint as to where in TFM I should R.

I would greatly appreciate receiving answers by email because my company 
makes it hard to read news.

Thanks!

Peter




From m.faassen at vet.uu.nl  Tue Apr 25 10:38:04 2000
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 25 Apr 2000 14:38:04 GMT
Subject: Simple list.append() question
References: <meh9-5ED8CD.11192424042000@news.cit.cornell.edu> <ZHZM4.360$qg7.2949@news.uswest.net> <39046FA2.603EAFE5@roguewave.com> <PsaN4.174166$Dv1.2059732@news1.rdc1.bc.home.com>
Message-ID: <8e4akc$p7o$1@newshost.accu.uu.nl>

Kevin Cazabon <kcazabon at home.com> wrote:
> well, if not prettier, why not obfuscate it?  q:]  Although this isn't
> pretty, I'm sure we can make it a little less intelligable with some work.

> entire_list = [];for i in range(3):entire_list.append([])

If you write it the normal way:

entire_list = []
for i in range(3):
    entire_list.append([])

It's pretty, readable, flexible, and it works. The only disadvantage is
that it's less short. But I'm willing to pay that price in this situation.

Generally it's a good idea to avoid * on sequences, unless that sequence is
immutable and contains immutable things; i.e. strings. I don't recall
using * on tuples ever, but you can do so safely if the tuple contains
immutable things only.

Is there any useful way to use * on lists that I missed? Usually we want
the copy semantics here, not the reference semantics. Perhaps it's a good
idea to completely forbid * on lists in p3k? Tuples too for all I care.
Then again I may be missing important uses, so enlighten me.

Regards,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From mindlace at imeme.net  Wed Apr 26 18:09:38 2000
From: mindlace at imeme.net (mindlace)
Date: Wed, 26 Apr 2000 16:09:38 -0600
Subject: newbie regexp question
References: <39075648.CA622A7D@imeme.net> <DwIN4.4305$Za1.64176@newsc.telia.net>
Message-ID: <39076922.CF83C53D@imeme.net>

Fredrik Lundh wrote:
> 
> mindlace <mindlace at imeme.net> wrote:
> > I've got a re.compile('\W') that I can use to search for non-alphanumeric
> > characters in a variable.  However, if the variable has ( in it, for
> example,
> > the search breaks.
> 
> breaks?

Er, yes, like if there's a ( it gives me an error.
 
> have no idea what you're talking about here, but maybe
> re.escape does what you wants?

Yes, it does.  Thanks very much.  I should have looked at Module Contents in the documentation on
python.org

-- 
ethan mindlace fremen        mindlace at imeme.net
zope    -&-     imap email   -&-   mailing list
weave your web with the web at http://imeme.net



From mwh21 at cam.ac.uk  Sun Apr 16 21:40:12 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 17 Apr 2000 02:40:12 +0100
Subject: building on linux w/o threads?
References: <8ddokt$65t$1@nnrp1.deja.com>
Message-ID: <m38zydec8z.fsf@atrus.jesus.cam.ac.uk>

firefly64 at my-deja.com writes:

> Hi,
> I'm trying to build python 1.5.2 on a linux box
> but without thread support. I'm using gcc 2.9.5,
> on Mandrak Linux 6.1 kernel version 2.2.14
> 
> Even though my system has threads, I don't want
> any thread support in python (no linkage to libpthreads).
> I just can't seem to convince the build process
> to proceed without threads.
> 
> I've done everything I can think of, including
> ./configure --without-thread and editing the
> Makefiles to include -DWITH_THREAD=0. The build
> process still always either produces a python
> linked to libpthreads or dies complaining about
> not being able to find various thread functions.
> (Yes, I've done make clean).

That's odd.  Have you tried running configure with no arguments at
all?  Reading through config.log?

Not much use I know, but "it works here".  My system is RedHat 6.1,
but in other respects is the same.

HTH (but I doubt it),
Michael

-- 
41. Some programming languages manage to absorb change, but 
    withstand progress.
     -- Alan Perlis, http://www.cs.yale.edu/~perlis-alan/quotes.html


From tismer at tismer.com  Fri Apr  7 16:06:42 2000
From: tismer at tismer.com (Christian Tismer)
Date: Fri, 07 Apr 2000 22:06:42 +0200
Subject: Palm OS
References: <8clds2$c25$1@nnrp1.deja.com>
Message-ID: <38EE3FD2.4A3AE50F@tismer.com>


musingattheruins at my-deja.com wrote:
> 
> Anyone know of a Python release for Palm?

Please do a search on recent c.l.py. postings. It is in progress.

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaunstr. 26                  :    *Starship* http://starship.python.net
14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     where do you want to jump today?   http://www.stackless.com



From bjorn at roguewave.com  Tue Apr  4 19:53:26 2000
From: bjorn at roguewave.com (Bjorn Pettersen)
Date: Tue, 04 Apr 2000 17:53:26 -0600
Subject: Two questions
References: <38EA5FE8.853E9B75@callware.com> <38EA6508.90BF630C@roguewave.com> <38EA6E87.40EDB542@callware.com> <38EA6422.61054B11@exceptionalminds.com> <38EA7452.9BFDF37E@callware.com> <8cdt7c$cl1$1@news101.telia.com> <38EA7D9C.7024CF0B@roguewave.com> <0
Message-ID: <38EA8076.D8385E18@roguewave.com>

Anders Eggers-Krag wrote:
> 
> > >
> > > an alternative solution is to call an ADO (ActiveX Data Object) object and
> > > use it through
> > > there which is quite simple.
> >
> > If you do this, make sure you get Mark's programming win32 book.  It's
> > an excellent book for many reasons, but it goes into special detail on
> > COM programming (and is so far the only resource I've seen of how COM
> > can be simple and useful -- instead of the normal
> > add-on-later-because-we-need-the-buzzword I normally associate with COM
> > <wink>).
> >
> it is not like ADO is a complex system, it is extremely easy to use, I use it alle time
> writing database systems for ASP.

Oh, I completely agree... However, if you don't know COM, or only know
it through C++, I would still recomend getting the win32 book.

it-also-explains-all-the-GBFLAs-in-ms'-db-libraries'ly y'rs
-- bjorn



From scarblac-spamtrap at pino.selwerd.nl  Tue Apr 25 11:02:33 2000
From: scarblac-spamtrap at pino.selwerd.nl (Remco Gerlich)
Date: 25 Apr 2000 15:02:33 GMT
Subject: List comprehensions?
References: <000601bfae79$41485b00$152d153f@tim>
Message-ID: <slrn8gbcrj.a24.scarblac-spamtrap@flits104-37.flits.rug.nl>

Tim Peters wrote in comp.lang.python:
> [Remco Gerlich]
> > I thought list comprehensions had been in CVS for quite a while, so I
> > expected them to be in 1.6. ...
> 
> Greg Ewing's patch (against 1.5.2) for list comprehensions is all there is,
> and you'll have to get that from Greg (see DejaNews for his announcement,
> and his URL).

Boggle. Brain didn't return correctly from that alternate reality, then.

Or maybe I'm just confused.

-- 
Remco Gerlich,  scarblac at pino.selwerd.nl
'Oook?'


From mjhand at concentric.net  Tue Apr 18 11:54:18 2000
From: mjhand at concentric.net (Manus Hand)
Date: Tue, 18 Apr 2000 09:54:18 -0600
Subject: BAD starship.python.net news!
Message-ID: <38FC852A.9B11F458@concentric.net>

This from Mike McLay.  Everyone, let's pony up what it takes to recover
our starship!

Manus (who will backup more and more often than he has until now)

> From: Mike McLay (mclay at nist.gov)
> Subject: Re: starship.python.net??
>
> The prognosis isn't good.  The old drive is toast and backup software
> was never installed.  I've located a data recovery company that will
> provide a quote on the chances of recovery for $125.  The cost of
> recovery, if it is possible, will be between $300-$1000.   They can
> turn around the recovery in a few days time.  Someone will need
> to fund the recovery.  So far I have a $200 pledge.
>
> I bought a new drive and installed it in the server last night. The
> system will be back online tomorrow by noon.



From garry at sage.att.com  Fri Apr 14 11:42:56 2000
From: garry at sage.att.com (Garry Hodgson)
Date: Fri, 14 Apr 2000 15:42:56 GMT
Subject: Doesn't anybody write to paper anymore?
References: <38F5FDF3.EE2FBA01@tcac.net>
Message-ID: <38F73C80.E1432809@sage.att.com>

"Daddy, what's "paper"?"


well, maybe not yet...

-- 
Garry Hodgson                   Every night 
garry at sage.att.com                a child is born
Software Innovation Services        is a Holy Night.
AT&T Labs                         - Sophia Lyon Fahs


From richard_chamberlain at ntlworld.com  Thu Apr 27 04:21:08 2000
From: richard_chamberlain at ntlworld.com (Richard Chamberlain)
Date: Thu, 27 Apr 2000 09:21:08 +0100
Subject: pythoncom / Installer on Win32
Message-ID: <YKSN4.65$E44.1172@news6-win.server.ntlworld.com>

i,

I am trying to create a distribution using Gordon's Standalone.

I'm asking a fair amount because I'm using COM, ODBC etc.

any help would be appreciated,

Richard

bindepend cannot analyze %s - file not found!
analyzing odbc
 found python15.dll
 found ODBC32.dll
 found dbi.dll
 lib not found: dbi.dll dependency of  found MSVCRT.dll
 found KERNEL32.dll
 added filter <TypeFilter {'d': 1}>
Creating COMServer.exe
Traceback (innermost last):
  File "Builder.py", line 502, in ?
    main(opts, args)
  File "Builder.py", line 484, in main
    target.build()
  File "Builder.py", line 87, in build
    self.assemble()
  File "Builder.py", line 300, in assemble
    arch = ArchiveTarget.assemble(self, pkgname)
  File "Builder.py", line 275, in assemble
    arch.build(pkgnm, toc)
  File "C:\WINDOWS\DESKTOP\INSTAL~1\MEInc\Dist\archive.py", line 146, in
build
  File "C:\WINDOWS\DESKTOP\INSTAL~1\MEInc\Dist\carchive.py", line 167, in
add
IOError: [Errno 2] No such file or directory:
'c:\\windows\\system\\pythoncom.dll'





From mrfx at 263.net  Tue Apr 25 03:09:17 2000
From: mrfx at 263.net (fmr)
Date: Tue, 25 Apr 2000 15:09:17 +0800
Subject: how to wrap c++ software to python?
Message-ID: <8e3ggq$2r1d$1@news.cz.js.cn>

I will call c++ routine in python,how to wrap
c++ routine for python.

Is there any utils to work except SWIG.






From pf at artcom-gmbh.de  Thu Apr 20 13:04:48 2000
From: pf at artcom-gmbh.de (Peter Funk)
Date: Thu, 20 Apr 2000 19:04:48 +0200 (MEST)
Subject: Lazy imports (was Re: module naming) (now hopefully correct)
Message-ID: <m12iKNk-000CnvC@artcom0.artcom-gmbh.de>

Hi!

Sorry if you see this twice. :-(  
Two days ago I posted a followup to a question by Michal Wallace, 
which included some Python code, which was BUGGY!  But see below.

> Michal Wallace (sabren) asks:
> >   That's certainly a lot cleaner! But is there a way to do that and
> > not have it load each one of those files every time? There's actually
> > more than 3, and even though they're all fairly small, I'd rather not
> > have the overhead of loading them until they're needed.
> 
> You could try to adopt the lazy loader concept implemented in 
> the Pmw package.  The idea is to use a small tool, that builds a 
> dictionary of exported features (classes and more) by investigating
> all module files in the package and places this in a file called
> for example 'exports.def'.  Than you can use the following
> __init__.py file to load features (classes, constants, functions,
> or modules, what you want) on demand.  The idea is "stolen" from
> Greg McFarlanes wonderful Pmw package:

As Greg McFarlane pointed out to me, a block of my program got the
wrong indentation, so it will not work as I intended it to.  Here is
an updated (and hopefully more correct version).  Many thanks to
Greg McFarlane for his original work and the time he spent to review
my code.  This time I also included a little helper script used to
automatically generate the 'exports.def' file used by the lazy importer.

Regards, Peter
---- 8< ---- 8< ---- cut here ---- 8< ---- schnipp ---- 8< ---- schnapp ----
#!/usr/bin/env python
## vim:ts=4:et:nowrap
"""__init__ : This file is executed when the package is imported.  It creates
a lazy importer/dynamic loader for the package and replaces the package 
module with it.  This is a very simplified version of the loader supplied 
with Pmw.  All the package version management has been stripped off."""

import sys, os, string, types

_EXP_DEF = 'exports.def'       # export definition file
_BASEMODULE = 'base'           # Name of Base module for the package

class _Loader:
    """An instance of this class will replace the module in sys.modules"""

    def __init__(self, path, package):
        self._path, self._package = path, package
        self._initialised = 0
        
    def _getmodule(self, modpath):
        __import__(modpath)
        mod = sys.modules[modpath]
        return mod

    def _initialise(self):
        # Create attributes for the Base classes and functions.
        basemodule = self._getmodule('_'+self._package+'.'+_BASEMODULE)
        for k,v in basemodule.__dict__.items():
            if k[0] is not '_' and type(v) != types.ModuleType:
                self.__dict__[k] = v
        # Set the package definitions from the exports.def file.
        dict = {
            '_features'     : {},
            '_modules'      : {},
        }
        for name in dict.keys():
            self.__dict__[name] = {}
        d = {}
        execfile(os.path.join(self._path, _EXP_DEF), d)
        for k,v in d.items():
            if dict.has_key(k):
                if type(v) == types.TupleType:
                    for item in v:
                        ## modpath = self._package + item
                        modpath = item
                        dict[k][item] = modpath
                elif type(v) == types.DictionaryType:
                    for k1, v1 in v.items():
                        ## modpath = '_'+self._package +'.'+ v1
                        modpath = v1
                        dict[k][k1] = modpath
        self.__dict__.update(dict)
        self._initialised = 1

    def __getattr__(self, name):
        """This will solve references to not yet used features"""
        if not self._initialised:
            self._initialise()
            # Beware: _initialise may have defined 'name'
            if self.__dict__.has_key(name):
                return self.__dict__[name]
        # The requested feature is not yet set. Look it up in the
        # tables set by exports.def, import the appropriate module and
        # set the attribute so that it will be found next time.
        if self._features.has_key(name):
            # The attribute is a feature from one of the modules.
            modname = self._features[name]
            mod  = self._getmodule('_'+self._package+'.'+modname)
            feature = getattr(mod, name)
            self.__dict__[name] = feature
            return feature
        elif self._modules.has_key(name):
            # The attribute is a module
            mod = self._getmodule('_'+self._package+'.'+name)
            self.__dict__[name] = mod
            return mod
        else:
            # The attribute is not known by the package, report an error.
            raise AttributeError, name

# Retrieve the name of the package:
_package = os.path.split(__path__[0])[1]
# Rename (hide) the original package for later perusual:
sys.modules['_'+_package] = sys.modules[_package]
# Create the dynamic loader and install it into sys.modules:
sys.modules[_package] = _Loader(__path__[0], _package)
---- 8< ---- 8< ---- cut here ---- 8< ---- schnipp ---- 8< ---- schnapp ----
#!/usr/bin/env python
## vim:ts=4:et:nowrap                    
"""build_exports.py -- create 'exports.def' helper file for lazy importer

This is a hack to aid changing the substructure of a ggui
package quickly and painless.
"""
import sys, types, pprint

modules = []
features = {}

multiple_defined = {}

template = '''## vim:ts=4:et:nowrap                     
# [Emacs: -*- python -*-]
"""export.def --- This is an exports definition file ---

this was automatically created by %(prog)s

It is invoked by a dynamic import loader in __init__.

    features      : dictionary from feature names to modules names.
    modules       : tuple of module names 
"""
#

_features = %(features)s

_modules = %(modules)s

'''

def spewout(stream=sys.stdout, modules=(), features={}):
    pp = pprint.PrettyPrinter(indent=4)
    d = { 'prog': sys.argv[0], 
          'modules': pp.pformat(tuple(modules)), 
          'features': pp.pformat(features),
    }
    stream.write(template % d)

def inspect(modulename, modules, features, multiple_defined):
    if modulename[-3:] == ".py": modulename = modulename[:-3]
    __import__(modulename)
    mod = sys.modules[modulename]
    for symbol in dir(mod):
        if symbol[:1] != '_' or (symbol == '_' and modulename == 'base'):
            obj = mod.__dict__[symbol]
            if type(obj) == types.ModuleType or symbol == "Pmw":
                if not symbol in modules:
                    modules.append(symbol)
            else:
                if features.has_key(symbol):
                    if multiple_defined.has_key(symbol):
                        multiple_defined[symbol] = multiple_defined[symbol] + \
                            " " + features[symbol]
                    else:
                        multiple_defined[symbol] = features[symbol]
                features[symbol] = modulename

if __name__ == "__main__":
    sys.path.insert(0, '.')
    if len(sys.argv) > 1:
        for arg in sys.argv[1:]:
            inspect(arg, modules, features, multiple_defined)
        outfile = sys.stdout
    else:
        import glob
        l = glob.glob("[a-z]*.py")
        print l
        for module in l:
            inspect(module, modules, features, multiple_defined)
        if multiple_defined == {}:
            outfile = open("exports.def", "w")
    if multiple_defined == {}:
        spewout(outfile, modules,  features)
    else:
        for k, v in multiple_defined.items():
            print k, "has multiple definitions in:", v, features[k]


Gru?, Peter
-- 
Peter Funk, Oldenburger Str.86, 27777 Ganderkesee, Tel: 04222 9502 70, Fax: -60
Wer sich zu wichtig f?r kleine Arbeiten h?lt,
ist meist zu klein f?r wichtige Arbeiten.     --      Jacques Tati



From Sunil.Hadap at cui.unige.ch  Wed Apr 19 14:52:02 2000
From: Sunil.Hadap at cui.unige.ch (Sunil Hadap)
Date: Wed, 19 Apr 2000 20:52:02 +0200
Subject: Dynamic web pages! not Dynamic HTML
Message-ID: <38FE0052.4BE78518@cui.unige.ch>

Hello,

I am really dumb in databases and perl. I am good at python and it's
my scripting language for whatever I do in graphics. I want to start
doing my homepage which is inspired from

http://www.photo.net/photo/

I want to implement the following features
a) As I too do lots of photography, I want to put my photographs all
over the web pages, may be randomly or depending upon context.
b) I don't want to hardcode any of the hyperlinks, so that I can move
my pages and still keep the links intact. Idea is, the link will be
always to a single page with a tag, which will take you to right page.

c) User will be able to add his comment to any page like this (bottom
of the page)
http://www.photo.net/photo/nikon/n90s.html
d) There will be lots of tags associated with pages. I just refer to
the tag and I should get automatic hyperlink to the appropriate point.

In short, I will be writing a page, not exactly a html page but rather
a template with special tags say @photo(location=geneva,
choice=random, qulity=best, addkey=lake). It will search the database
to narrow down and finally select a random photo. Another tag will be
just @"good exposure" or @"particle system". From these template
files, the system will generate html document replacing all tags with
appropriate hyperlink or image etc, on the fly.

I want to do this in Python. Is zope useful for this, I don't want to
learn a big system as I only want few specific functions only. Which
free database?

Thanks a lot

Sunil




--
"Live as if you would die tomorrow,
learn as if you would live forever."
                    --Mahatma Gandhi




From vmarkwart at my-deja.com  Sun Apr  2 20:05:53 2000
From: vmarkwart at my-deja.com (vmarkwart at my-deja.com)
Date: Mon, 03 Apr 2000 00:05:53 GMT
Subject: pytools - console and other facilities
Message-ID: <8c8n8p$bq0$1@nnrp1.deja.com>

Hi,
I've written a python extension for WIN32. At the moment it's not
finished, nor is it particularly well documented, but it can be
examined at http://www.pcug.org.au/~markwart/python/index.html

Any assistance in progressing this project would be greatly
appreciated.
Cheers
Victor


Sent via Deja.com http://www.deja.com/
Before you buy.


From tim_one at email.msn.com  Tue Apr 11 12:00:40 2000
From: tim_one at email.msn.com (Tim Peters)
Date: Tue, 11 Apr 2000 12:00:40 -0400
Subject: [ANNOUNCE] Garbage collection for Python
In-Reply-To: <Fsuuw7.Inu@world.std.com>
Message-ID: <000a01bfa3cf$159c8440$7f2d153f@tim>

[Hrvoje Niksic]
> This is still different from being able to embed multiple Python
> instances in threads.

[Will Ware]
> Are we talking about multiple Python instances running in a single
> thread? That sounds like a very big challenge. ...

Perhaps, but it's already been done <wink>.  See Py_NewInterpreter() in the
"Initialization, Finalization, and Threads" section of the Python/C API
manual.  There isn't a Python-level interface to this yet; if you want
multiple interpreters in a single thread, for now you need to call
Py_NewInterpreter() from a C extension.

> The Python VM would need to be rewritten to be reentrant so that
> multiple VMs could simultaneously coexist, with distinct copies of
> the entire computing environment (globals, locals, which modules had
> been imported, etc). Then you'd need to arrange context switches; ...

Right, that's all there already, but (again) only from C.  See the manual
for a description of subtleties.

have-yet-to-see-anyone-use-it!-ly y'rs  - tim





From chapmanb at arches.uga.edu  Mon Apr  3 07:21:23 2000
From: chapmanb at arches.uga.edu (Brad Chapman)
Date: Mon, 03 Apr 2000 11:21:23 GMT
Subject: Methods for password storage
Message-ID: <38E87E13.A10A1168@arches.uga.edu>

Hello!
    I'm working on a program which stores passwords (for database
access) that a user enters in a GUI. The program is targetted for unix
machines. I'm not very experienced with security issues like this and am
trying to store them in as secure a manner as possible.
    Does anyone have any advice/examples for how to securely store
sensitive information like this using python? Are there other security
issues I should worry about when dealing with passwords? Thank in
advance much for any advice!

Brad



From kc5tja at garnet.armored.net  Tue Apr  4 09:09:37 2000
From: kc5tja at garnet.armored.net (Samuel A. Falvo II)
Date: 4 Apr 2000 13:09:37 GMT
Subject: blessed be Guido, for "5 <= X <= 10" does what it should...
References: <3TpG4.3349$HG1.98231@nnrp1.uunet.ca>
Message-ID: <slrn8eklr4.9d7.kc5tja@garnet.armored.net>

In article <3TpG4.3349$HG1.98231 at nnrp1.uunet.ca>, Warren Postma wrote:
>(Actually I'd write the <= as '?' (Unicode U+2264) but using such things in
>our code would seriously cramp portability.  Hey, is the APL symbol set
>anywhere in Unicode!?)

Astonishingly enough, it is.  Or so I've been told.  Not sure where or how
to verify this, though, as I've never used APL.  It's part of their 32-bit
character set.

-- 
KC5TJA/6, DM13, QRP-L #1447
Samuel A. Falvo II
Oceanside, CA


From aa8vb at yahoo.com  Sat Apr 15 11:11:05 2000
From: aa8vb at yahoo.com (Randall Hopper)
Date: Sat, 15 Apr 2000 11:11:05 -0400
Subject: Doesn't anybody write to paper anymore?
In-Reply-To: <38F7A309.F1242F00@direcpc.com>
References: <38F5FDF3.EE2FBA01@tcac.net> <38F73C80.E1432809@sage.att.com>
 <38F7A309.F1242F00@direcpc.com>
Message-ID: <20000415111105.A8802@vislab.epa.gov>

Jay Freeman:
 |Subject: Doesn't anybody write to paper anymore?  
 |
 |Garry Hodgson wrote:
 |
 |> "Daddy, what's "paper"?"
 |
 |Let's see:
 |
 |C:\WINDOWS>ping www.fcc.gov
 |Reply from 192.104.54.12: bytes=32 time=510ms TTL=246
 |
 |Yep, still have a government, so we must still have paper too, QED :).

You guys are cruel.  ;-)  Don't you remember being a newbie?

But seriously, looking back Albert did post related questions about this
before (several times) as "redirecting stdout to printer".  You gotta
admit; the subject "Doesn't anybody write to paper anymore?" is more likely
to catch someone's eye.  8^)  Mission accomplished.

-- 
Randall Hopper
aa8vb at yahoo.com



From python-list at teleo.net  Fri Apr  7 18:51:41 2000
From: python-list at teleo.net (Patrick Phalen)
Date: Fri, 7 Apr 2000 15:51:41 -0700
Subject: Problem running Python script as cron job
In-Reply-To: <wrd-0AEDDA.12282507042000@news.awenet.com>
References: <wrd-0AEDDA.12282507042000@news.awenet.com>
Message-ID: <0004071555090M.01234@quadra.teleo.net>

[William R. Dickson, on Fri, 07 Apr 2000]
:: Anybody know of any issues with running python scripts from Cron?

No. I've had multiple Python scripts running every one, two, five and
ten minutes around the clock for years on multiple servers without
problems.

I'd suspect you might have a bug lurking.



From zsoltman at my-deja.com  Sun Apr  2 18:28:44 2000
From: zsoltman at my-deja.com (zsoltman at my-deja.com)
Date: Sun, 02 Apr 2000 22:28:44 GMT
Subject: Web hosting python CGI and or Zope
Message-ID: <8c8hie$5lo$1@nnrp1.deja.com>

I am advising a volunteer organization that gets some funding.  I am
the only technical person they have right now.  They desire to not only
have a Web Site, but have in mind two applications that they will need
immediately.  I am volunteering to hammer out basic functionality for
these applications to get them up and running.

If I recommend they purchase an NT machine (don't yell at me, they're
non-technical)and go DSL, I could make their applications very quickly
with either Zope, or Python CGI, but I cannot ignore the fact that
there are Web Hosting companies that have fault tolerance and would be
much cheaper.

Question:  Does anyone know of Web Hosting companies that support
python or Zope?

Alternate Question:  Are there other ways to approach this?

Thanks for any feedback.


-Zsolt
zsoltman at hotmail.com


Sent via Deja.com http://www.deja.com/
Before you buy.


From yemartin at garage.co.jp  Wed Apr 12 22:44:31 2000
From: yemartin at garage.co.jp (Yves-Eric Martin)
Date: Thu, 13 Apr 2000 11:44:31 +0900
Subject: Seg fault in Python/C API: PyString_FromString(NULL)
In-Reply-To: <c1YI4.2029$UB.13736@news-server.bigpond.net.au>
References: <38F422C5293.A67BYEMARTIN@smtp.garage.co.jp> <c1YI4.2029$UB.13736@news-server.bigpond.net.au>
Message-ID: <38F5348F25.C638YEMARTIN@smtp.garage.co.jp>

On Wed, 12 Apr 2000 10:15:36 GMT
"Mark Hammond" <mhammond at skippinet.com.au> wrote:

> >     First, I am new to the Python/C API, so correct me if I am wrong:
> > shouldn't PyString_FromString(NULL) return Py_None? Or am I out to
> > lunch?
> 
> Nope - you must check for NULL before making the call.


    Ok, that makes sense now. Thanks.


-- 
Yves-Eric Martin
Digital Garage Inc.
yemartin at garage.co.jp



From jp at NOSPAMdemonseed.net  Fri Apr 21 15:17:47 2000
From: jp at NOSPAMdemonseed.net (jason petrone)
Date: Fri, 21 Apr 2000 14:17:47 -0500
Subject: can't find SWIG for Python
References: <38FF5A56.F76794D8@sec.noaa.gov>
Message-ID: <rg9qd8.gin.ln@demonseed.net>

j vickroy <jvickroy at sec.noaa.gov> wrote:
> I have been unable to locate SWIG for Python.  My references list
> www.swig.org but the server can not be found.

> Could someone point me to the site or provide some other Win 32 options
> for easing the process of extending Python to use existing C libraries.

~> grep ftp /usr/ports/devel/swig/Makefile 
MASTER_SITES=   ftp://ftp.cs.utah.edu/pub/beazley/SWIG/

-jlp


From dirck at pacbell.net  Mon Apr 10 13:57:41 2000
From: dirck at pacbell.net (Dirck Blaskey)
Date: Mon, 10 Apr 2000 10:57:41 -0700
Subject: Redirecting stdin/stdout on Windows with '>'and '<'
References: <8csvff$7pn$1@nnrp1.deja.com>
Message-ID: <iAoI4.1020$Sz2.504367@news.pacbell.net>

Mark <markmcmahon at my-deja.com> wrote in message
news:8csvff$7pn$1 at nnrp1.deja.com...
> ...
> I would like to redirect output of my script with '>' or '<'
>
> If I try to redirect out then the file is created but has no contents.
>
> If I try to redirect in eg...
> "test.py < autoexec.bat"
>
> I get the following traceback
>
> Traceback (most recent call last):
>   File "C:\test.py", line 4, in ?
>     s = sys.stdin.read()
> IOError: [Errno 9] Bad file descriptor
>
> (Windows nt 4 sp 6, python 1.52)

Windows cmd.exe has a bug in file redirection when handling scripts.
The 'simple' fix is to use python on the command line:

python test.py < infile > outfile

(requires python in the path)

Attached is a Python script that uses an alternative syntax
to fix the problem:

test.py @infile !outfile

The script is also available from:
http://www.danbala.com/python/cmdfix
(currently down??)

Good luck

d
===========================
Dirck Blaskey
Danbala Software
dirck at danbala.com
http://www.danbala.com



begin 666 cmdfix.py
M(B(B#0H@(" @8VUD9FEX+G!Y("T at 9FEX('!I<&EN9R!P<F]B;&5M<R!W:71H
M(&-M9"YE>&4@#0H-"F-O;G1A8W0Z(&1I<F-K0&1A;F)A;&$N8V]M#0H-"E=I
M=&@@5VEN,S(@8VUD+F5X92P@)W!Y=&AO;B!S;VUE+G!Y(#YT)R!W;W)K<R!B
M=70@)W-O;64N<'D@/G0G(&1O97-N)W0N#0IC;61F:7 at N<'D at 9V5T<R!A<F]U
M;F0@=&AI<R!P<F]B;&5M(&)Y('5S:6YG(&%L=&5R;F%T:79E('!I<&EN9R!S
M>6YT87 at Z#0H@(" @<V]M92YP>2! :6YP=70@(6]U='!U=" C97)R;W(-"F]R
M#0H@(" @<V]M92YP>2! :6YP=70@(2-B;W1H;W5T<'5T#0H-"D%D9&EN9R N
M<'DO+G!Y8R!T;R!T:&4 at 96YV:7)O;FUE;G0 at 4$%42$585#T at 8VAA;F=E<R!T
M:&ES('1O('-I;7!L>3H-"B @("!S;VUE($!I;G!U=" A(V)O=&AO=71P=70-
M"@T*5&\@:6YS=&%L;"P at 8VAA;F=E('1H92!R96=I<W1R>2!E;G1R>2!F;W(@
M4'ET:&]N+D9I;&4O<VAE;&PO;W!E;B]C;VUM86YD('1O. at T*(" @($0Z7'!Y
M=&AO;EQP>71H;VXN97AE($0Z7'!Y=&AO;EQC;61F:7 at N<'D at 1#I<<'ET:&]N
M7'!Y=&AO;BYE>&4@(B4Q(B E*@T*4VEM:6QA<B!C:&%N9V5S(&%R92!R97%U
M:7)E9"!F;W(@4'ET:&]N+D-O;7!I;&5D1FEL92 H+G!Y8RD at 87,@=V5L;"X-
M"DEF('EO=2!P<F5C;VUP:6QE(&-M9&9I>"YP>2!Y;W4 at 8V%N('5S92!C;61F
M:7 at N<'EC(&EN(&ET)W,@<&QA8V4 at 9F]R( T*;&EK96QY('5N;F]T:6-E86)L
M92!P97)F;W)M86YC92!I;7!R;W9E;65N=',N#0H-"E1H:7,@<V-R:7!T(&ES
M('=R:71T96X at 9V5N97)I8V%L;'DL(&YO="!A<W-U;6EN9R!T:&4 at 97AE8W5T
M960@<')O8V5S<R!I<PT*<'ET:&]N+"!S;R!I="!W;W5L9"!A;'-O('=O<FL@
M9F]R('!E<FP@;W(@;W1H97(@<V-R:7!T:6YG(&QA;F=U86=E<RX-"@T*5&AE
M("$C('-Y;G1A>"!I<R!A;'-O('5S969U;"!F;W(@)VYO<FUA;"<@8V]M;6%N
M9',L(&)E8V%U<V4 at 8VUD(&YO<FUA;&QY#0IW;VXG="!P<F]D=6-E("=I;G1E
M<FQE879E9"<@<W1D;W5T+W-T9&5R<BX@(%5N9F]R='5N871E;'DL('1H92 N
M<'D-"F-O;F9I9W5R871I;VX@=VEL;"!P<F5P87)S92!T:&4 at 0"$C(&EF('1R
M>6EN9R!T;R!U<V4 at 8VUD9FEX+G!Y(&1I<F5C=&QY+ at T*#0I!<F=U;65N=',@
M8V]N=&%I;FEN9R!S<&%C97, at 87)E(')E87-S96UB;&5D(&%P<')O<')I871E
M;'DN#0HB(B(-"@T*:6UP;W)T('-Y<PT*:6UP;W)T('-T<FEN9PT*#0II;7!O
M<G0@<'EW:6YT>7!E<PT*:6UP;W)T('=I;C,R87!I#0II;7!O<G0@=VEN,S)E
M=F5N= T*:6UP;W)T('=I;C,R9FEL90T*:6UP;W)T('=I;C,R<')O8V5S<PT*
M#0ID968@;6%I;BAA<F=C+"!A<F=V*3H-"B @(" -"B @(" C(&]O<', at +2!I
M;F-O<G)E8W0@=7-A9V4-"B @(" -"B @("!I9B!N;W0 at 87)G8SH-"B @(" @
M(" @<')I;G0 at 7U]D;V-?7PT*(" @(" @("!R971U<FX-"@T*(" @(",@:6YP
M=70 at 9FEL92!N86UE('1A9V=E9"!W:71H($ -"B @(" -"B @("!I1FEL94YA
M;64@/2!A<F<H87)G=BP@)T G*0T*(" @(&EF(&E&:6QE3F%M93H-"B @(" @
M(" @:2 ](&]P96Y?<F5A9"AI1FEL94YA;64I.PT*(" @(" @("!W:6XS,F%P
M:2Y39713=&1(86YD;&4H=VEN,S)A<&DN4U1$7TE.4%547TA!3D1,12P@:2D-
M"@T*(" @(",@;W5T<'5T(&9I;&4@;F%M92!T86=G960@=VET:" A#0H@(" @
M;T9I;&5.86UE(#T at 87)G*&%R9W8L("<A)RD-"B @("!I9B!O1FEL94YA;64Z
M#0H@(" @(" @(&)O=&@@/2 P#0H@(" @(" @(",@)R$C)R!G:79E<R!M:7AE
M9"!S=&%N9&%R9"!A;F0 at 97)R;W(@;W5T<'5T+" -"B @(" @(" @(R @(" @
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M92!T86=G960@=VET:" C#0H@(" @949I;&5.86UE(#T at 87)G*&%R9W8L("<C
M)RD-"B @("!I9B!E1FEL94YA;64Z#0H@(" @(" @(&4@/2!O<&5N7W=R:71E
M*&5&:6QE3F%M92D[#0H@(" @(" @('=I;C,R87!I+E-E=%-T9$AA;F1L92AW
M:6XS,F%P:2Y35$1?15)23U)?2$%.1$Q%+"!O*3L-"B @(" -"B @(" C('!U
M="!T:&4@<F5S="!O9B!T:&4 at 8V]M;6%N9"!B86-K('1O9V5T:&5R+"!A;F0@
M97AE8W5T92!I=" @(" -"B @("!C;60@/2!J;VEN*&%R9W8I#0H@(" @97AE
M8W5T92AC;60I(" @( T*#0HC(&%R9R M(')E='5R;B!A('1A9V=E9"!A<F=U
M;65N="!A;F0 at 9&5L971E(&ET(&9R;VT@=&AE(&QI<W0-"@T*9&5F(&%R9RAA
M<F=S+"!T86<I. at T*(" @(&D@/2 P#0H@(" @9F]R(&$@:6X at 87)G<SH-"B @
M(" @(" @:68 at 85LP72 ]/2!T86<Z#0H@(" @(" @(" @("!D96P at 87)G<UMI
M70T*(" @(" @(" @(" @<F5T=7)N(&%;,3I=#0H@(" @(" @(&D@/2!I("L@
M,0T*#0HC(&IO:6X@=&AE(&%R9W5M96YT<R!B86-K('1O9V5T:&5R+"!H86YD
M;&4 at 87)G=6UE;G1S('=I=&@@<W!A8V5S(&)Y('%U;W1I;F<-"@T*9&5F(&IO
M:6XH87)G<RDZ#0H@(" @:2 ](# -"B @("!F;W(@82!I;B!A<F=S. at T*(" @
M(" @("!I9B!S=')I;F<N9FEN9"AA+" G("<I(#X](# Z#0H@(" @(" @(" @
M("!A<F=S6VE=(#T@)R(E<R(G("4 at 80T*(" @(" @("!I(#T@:2 K(#$-"B @
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M9B!E>&5C=71E*&-O;6UA;F0I. at T*(" @('-T87)T=7!);F9O(#T@=VEN,S)P
M<F]C97-S+E-405)455!)3D9/*"D-"B @("!P:6YF;R ]('=I;C,R<')O8V5S
M<RY#<F5A=&50<F]C97-S*$YO;F4L(&-O;6UA;F0L($YO;F4 at +"!.;VYE+" Q
M+" P+" -"B @(" @(" @(" @(" @(" @(" @(" @(" @(" @(" @(" @(" @
M($YO;F4L($YO;F4L('-T87)T=7!);F9O*0T*(" @(')E='5R;B!W:6XS,F5V
M96YT+E=A:71&;W)3:6YG;&5/8FIE8W0H<&EN9F];,%TL('=I;C,R979E;G0N
M24Y&24Y)5$4I#0H-"B,@:6YV;VME(&UA:6X@:68 at 87!P<F]P<FEA=&4-"@T*
M:68 at 7U]N86UE7U\@/3T@(E]?;6%I;E]?(CH-"B @("!M86EN*&QE;BAS>7,N
:87)G=BDM,2P@<WES+F%R9W9;,3I=*0T*#0H`
`
end



From mhammond at skippinet.com.au  Thu Apr 13 22:28:25 2000
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Fri, 14 Apr 2000 02:28:25 GMT
Subject: Embedding Python in Visual C++ 6.0
References: <8d1he4$26d$1@nnrp1.deja.com> <N2YI4.2030$UB.13624@news-server.bigpond.net.au> <8d1od2$9fl$1@nnrp1.deja.com> <Sj%I4.2225$UB.14895@news-server.bigpond.net.au> <8d3qvu$hud$1@nnrp1.deja.com>
Message-ID: <dnvJ4.799$1D.4997@news-server.bigpond.net.au>

It appears that 2 win32ui.pyd files are being loaded.  This is almost
certainly my fault.  It is possible that in the registry you already have
a "Modules" subkey that specifies the location of win32ui.pyd?

Also, FYI, in the CVS version of Pythonwin, this implementation has
changed.  It now doesnt attempt to find registered modules or any of the
other funky stuff that it does now.  The new code is simpler and more
robust - but does insist that win32ui.pyd is either on your PYTHONPATH, or
in the same directory as pythonwin.exe - no a burden IMO!  The new code
also magically works with either 1.5 or 1.6.

Anyway, whatever the reason, the clues are:

...
> Loaded symbols for 'C:
> \Programme\Python\Pythonwin\src\work\Pythonwin\Build\win32ui_d.pyd'

That is fine:
...

> Extension module C:
> \Programme\Python\Pythonwin\src\work\Pythonwin\Build\win32ui_d.pyd
> initialising.

Still looking good!

> LDR: Dll win32ui_d.pyd base 1e400000 relocated due to collision with C:
> \Programme\Python\Pythonwin\src\work\Pythonwin\Build\win32ui_d.pyd
> Preloaded symbols may not match 'C:
> \Programme\Python\Pythonwin\win32ui_d.pyd'.

Did you cut any messages?  This message appears to imply another
win32ui_d.pyd has been loaded, and its symbols have been relocated due to
a clash with another win32ui_d.pyd.

Can you ensure you system has no others win32ui*.pyd.

Also, I highly recommend getting the CVS version of Pythonwin - just for
the new startup code if nothing else.

Mark.





From jrayson at mitre.org  Fri Apr  7 17:30:53 2000
From: jrayson at mitre.org (Jim Rayson)
Date: Fri, 07 Apr 2000 17:30:53 -0400
Subject: PyOpenGL install problem on windows
Message-ID: <38EE538D.68D829BA@mitre.org>

Hi,
I'm trying to get PyOpenGL running on Windows 98.  I have lots of
C/Unix experience and very little Python, Tk, or Windows experience. 
Running Python 1.5.2, Tcl/TK 8.0.5, and PyOpenGl (couldn't find a
version
number but copied from it's main site, as a windows binary w/ installer,
a few weeks ago).  I first installed disutils, numpy, and pil as
directed.  By playing with paths and things I got it to the point where
I
could at least import modules from each package.  I got at least one
demo
program to work from numpy and pil, but not all of them.

When I installed PyOpenGL I managed to get opengl\demo\da\dots.py to
run properly.  Other demos failed because I seemed to be missing Togl.

? Found out on
http://starship.python.net:9673/crew/da/Code/PyOpenGL/Togl that
sometimes installer doesn't do togl properly, and gave instructions
on downloading some files to fix it.  However, it made no mention of the
togl file (appears to be some kind of index) that already exists in
C:\Program Files\Tcl\lib\tk8.0   Then again, maybe the installer stuck
that there.

? Just to be safe, I renamed that file as togl.old

? Based on those instructions I created the directory
C:\ProgramFiles\Tdl\lib\tk8.0\Togl   then downloaded and copied  
Togl.dll andpkgIndex.tcl

? Then ran samples, with no real success.  Got the message:? TclError:
error while autoloading "togl": couldn't load
file"C:\PROGRA~1\TCL\lib\tk8.0\Togl\Togl.dll": invalid argument? But I
checked that path and that file was at that location.  Since it's
already
there, I'm not sure why it can't find it.  pkgIndex.tcl says something
about setting $dir but my experiments playing with that also failed. 
Any help would be appreciated.

Thanks,
Jim
jrayson at mitre.org


From chris_barker at my-deja.com  Fri Apr 28 15:42:00 2000
From: chris_barker at my-deja.com (chris_barker at my-deja.com)
Date: Fri, 28 Apr 2000 19:42:00 GMT
Subject: Looking for example of embedded Python
Message-ID: <8ecpi4$j9q$1@nnrp1.deja.com>

Hi,

I'm trying to persuade my co-workers that it would be extememly usefull
to imbed Python in some of our apps. I'm looking for example
applications wioth Python embedded. Ideally something fairly small with
source code available.

Please sent me links to any examples you might have.

Platforms (in order of preference):
Macintosh
Linux
Windows



-Thanks,

-Chris

cbarker at jps.net


Sent via Deja.com http://www.deja.com/
Before you buy.


From pecora at anvil.nrl.navy.mil  Fri Apr 28 17:00:30 2000
From: pecora at anvil.nrl.navy.mil (Louis M. Pecora)
Date: Fri, 28 Apr 2000 17:00:30 -0400
Subject: Reading in strings -> numbers ??
Message-ID: <280420001700308281%pecora@anvil.nrl.navy.mil>

Is there a Python "read-partner" for the write(string) function, e.g., 

fileobject.write(" %e %e %d" % (1.2, -3.5, 10))  

in the sense that the C functions fprintf and fscanf are partners.

I've looked in 4 different Python books and in the pdf references, but
I see nothing in read or readlines or strings that seems to make for an
easy conversion from a string to a list or tuple of objects (numbers). 
In other words, I want the opposite of the write above so I can read
those numbers back into my Python code.  

Thanks for any help.


From hanche at math.ntnu.no  Thu Apr  6 23:12:48 2000
From: hanche at math.ntnu.no (Harald Hanche-Olsen)
Date: 06 Apr 2000 23:12:48 -0400
Subject: time-module questions
References: <CxIG4.3719$HG1.106792@nnrp1.uunet.ca>
Message-ID: <pcoitxu38pb.fsf@math.ntnu.no>

+ "Warren Postma" <embed at geocities.com>:

| First, it appears that time.localtime returns fractions of a second
| in it's values, but that time.mktime does not add this millisecond
| component back in, nor does it appear to be part of the time-tuple
| returned by time.localtime, and time.gmtime:

The time module is little more than an interface to the standard C
library, which does not make provisions for fractions of a second.  If
you wish to preserve that information, you'll have to hack it
yourself, or perhaps Marc-Andr? Lemburg's mxDateTime module has the
answers.

In fact, ``use mxDateTime'' appears to be the suggested solution to
every time/date-related problem in python.  I downloaded it but still
haven't had the time to check it out carefully, so I can't say for
sure whether it solves your problem.

| Secondly, mktime accepts localtimes only, how to I take y/m/d/h/m/s
| in GMT instead of the current timezone, and convert back to a
| time.time() type value (float).

Yes, mktime is the inverse of localtime.  You need the inverse of
gmtime, which is not in the standard C library, and hence not in the
time module.  Again, you have to hack it yourself, with a *large*
probability of getting it wrong, or use the mxDateTime module, which
seems to have functions for converting between UTC and localtime.

I once looked at the code in the GNU date program to see how it coped
with the situation, and it's not pretty.  After staring at the code
for half an hour, I found myself unable to prove its correctness,
though it performed correctly on all times that I threw at it (using
date -d '<some date and time> UTC'), even around the times when DST
starts or stops.

Normally I am kind of a peaceful guy, but if I ever meet the designers
of the standard C time conversion functions, I am not sure what I'll
do.  Fortunately I am not very big and strong, and I don't carry any
weapons -- so I'll probably settle for some choice verbal abuse, which
is unlikely to land me in jail.  8-)
-- 
* Harald Hanche-Olsen     <URL:http://www.math.ntnu.no/~hanche/>
- "There arises from a bad and unapt formation of words
   a wonderful obstruction to the mind."  - Francis Bacon


From darrell at rubina.funkomite.com  Tue Apr 11 14:24:56 2000
From: darrell at rubina.funkomite.com (darrell at rubina.funkomite.com)
Date: Tue, 11 Apr 2000 18:24:56 GMT
Subject: simple httplib and urllib timeout question
References: <Pine.BSF.4.10.10004111337520.29635-100000@chi.pair.com>
Message-ID: <Y5KI4.14911$eh.1041575@news.bc.tac.net>

"Michal Wallace (sabren)" <sabren at manifestation.com> wrote:
> On Tue, 11 Apr 2000 darrell at rubina.funkomite.com wrote:

>> Just a quick question, im writing a small webbot for a python
>> learning experience and it comes across some sites that will make it
>> hang for a while, if the site is down or extremely slow etc, is
>> there a way to set a faster timeout for httplib . Or is there a
>> better way to handle it ?

> Yes. The better way is to use asyncore whenever httplib/urllib uses a
> normal socket. Unfortunately, this requires breaking urllib into
> smaller pieces, and nobody seems to have done it yet. I'm actually
> working on the exact same problem (and posting regular updates
> at http://www.linkwatcher.com/ ) 

> I sent an email to Jeremy Hylton, who's working on urllib2. He's
> swamped, but willing to accept a patch that would add this feature
> to urllib2.. I think I'll wind up taking a crack at it tonight.

> There's also a partial solution using a different method if you look
> at...  http://www.linkwatcher.com/src/blogbot/timeout.py

> there's an example script called blogbot.py there, too.. it uses
> timeout.py and another hack with signal.alarm(), but neither really
> works all that well..

> hope that helps.

> meanwhile... HEY OLEG! -> have you done anything w/ this?

> Cheers,

> - Michal
> -------------------------------------------------------------------------
> http://www.manifestation.com/         http://www.linkwatcher.com/metalog/
> -------------------------------------------------------------------------

Thanks Michal , this helps alot :>


From rich at anna.UUCP  Mon Apr  3 01:22:53 2000
From: rich at anna.UUCP (Rich James)
Date: Mon, 03 Apr 2000 05:22:53 GMT
Subject: Just doing local file testing
References: <38E796BC.23191B2@regnoc.com> <38E7D6DE.EA09C891@regnoc.com> <00040216504403.04535@quadra.teleo.net> <38E7F616.48CF9284@regnoc.com>
Message-ID: <NUVF4.16811$9m6.678422@newsread1.prod.itd.earthlink.net>

Client-side scripts, yes.  Server-side scripts (i.e. cgi-scripts), no.

According to Jim Conger  <jimc at regnoc.com>:
> You can initiate a local application from within your browser without having a "server" in place.  For example, you
> can test Perl scripts, Java applets, and Active X components locally without needing a server.   That is what I am
> trying to do with the Python scripts. 
-- 
War is the health of the State.  
-- Randolph Bourne (1917)
---
Rich James


From gregm at iname.com  Wed Apr 12 09:46:30 2000
From: gregm at iname.com (Greg McFarlane)
Date: Wed, 12 Apr 2000 23:46:30 +1000
Subject: Compiling Python Imaging under LinuxPPC
In-Reply-To: <20000411173904.18637.00002261@ng-cg1.aol.com>; from William F. Adams on 11 Apr 2000 at 09:39:04PM
References: <20000411173904.18637.00002261@ng-cg1.aol.com>
Message-ID: <20000412234630.21095@nms.otc.telstra.com.au>

On 11 Apr, William F. Adams wrote:
> I'm trying to do this, and not getting far.
> 
> Here's what I get:
> .0 -L/usr/X11R6/lib -lX11 -lXext -lXmu -ltiff -ljpeg -lz -lXpm -0 _tkinter.so
> /usr/bin/ld: cannot find -lBLT
> collect2: ld returned 1 exit status
> make: *** [_tkinter.so] Error 1
> Bad exit status from /var/tmp/rpm-tmp.39884 (%build)
> 
> What's BLT?

A graphics extension to Tk:
    http://www.tcltk.com/blt/

-- 
Greg McFarlane     INMS Telstra Australia     gregm at iname.com



From tjreedy at udel.edu  Sun Apr  2 22:32:14 2000
From: tjreedy at udel.edu (Terry Reedy)
Date: Sun, 2 Apr 2000 22:32:14 -0400
Subject: Python 1.6 alpha 1 released
References: <Pine.LNX.4.10.10003311651590.22919-100000@starship.python.net>
Message-ID: <8c8vfj$5n6$1@news.udel.edu>

Guido is quoted as writing:

> """where you used to write sock.connect(host, port) you must now write
> sock.connect((host, port))"""

Since the de facto standard has been to use two parameters, I find this
change baffling.

Evidence of claim: the first two Python books ...
1. Internet Programming with Python by Watters, *van Rossum*, and Ahlstrom.
    p335:    s.connect(machine, port) # api definition, middle of page
    p342:    sock.connect(self.machine, self.port) # connect() method of
class GETRequest
2. Programming Python, Mark Lutz.
    p728:    s.connect(HOST, PORT) # sockclient.py
(Note: these are the only three uses I found in the two books.  The
GETRequest class, with its now-declared-to-be-an-error, is used in several
examples in the following pages.)
3. my memory impression of every other usage of connect that I have read
(I've never had cause yet to use it)

Additional reasons for opinion.
1. Adding the extra pair of parens is extra work for no gain.
2. The result strikes me as visually ugly.
Without countervailing philosophical justification, this code-breaking
change strikes me as contrary to both the spirit and practice of python.

The confusion-elimination reason for the list.append(item) API does not
apply here.

Is there any arguement (other than whatever the reference states) for
tuple-bundling the params of connect() that would not apply to every other
library function?

If the connect() code currently also accepts the new form, then the API
could be tightened just as well by eliminating the one-tuple variant.   How
much code would that break?

Terry J. Reedy




From claird at starbase.neosoft.com  Sun Apr  9 17:26:25 2000
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 9 Apr 2000 16:26:25 -0500
Subject: Migrating Tcl system to Python
References: <095cc48a.7e9b36dd@usw-ex0108-063.remarq.com> <000C48B993BF96DA.95DEBE4215A79896.7F63D15FB99EE88D@lp.airnews.net> <248819a8.00d5228e@usw-ex0106-045.remarq.com>
Message-ID: <8EDD15EFCAC9A6DF.7A5207F9F5767B8C.D17A9FDF542DC3C4@lp.airnews.net>

In article <248819a8.00d5228e at usw-ex0106-045.remarq.com>,
Hamish Lawson  <hamish_lawson at yahoo.co.uk> wrote:
>Cameron Laird wrote:
>
>[Stuff about communicating between Tcl and Python.]
>
>> Minotaur. CORBA. ILU. COM, maybe. Pipes. Named pipes.
>
>The systems will be on Linux so COM is out but the rest are
>possibilities. As far as I can see, one of the main issues in
>communicating between Tcl and Python will be how to handle
>passing lists and dictionaries (arrays in Tcl). XML-RPC manages
>this automatically and although I don't have any experience with
>either CORBA or ILU, I imagine they do too. If I were to use a
>pipe approach, there would presumably be a need for a layer to
>translate between Tcl's and Python's representations of data
>structures; though I think this might be reasonably simple, are
>there solutions already available, or issues I'd need to be
>aware of?
			.
			.
			.
You're right in everything you say; XML-RPC, ILU,
and CORBA handle things nicely, but I can think of
no effective semantic translator specifically for
the case of exchanging lists and dictionaries through
pipes.
-- 

Cameron Laird <claird at NeoSoft.com>
Business:  http://www.Phaseit.net
Personal:  http://starbase.neosoft.com/~claird/home.html


From nascheme at enme.ucalgary.ca  Wed Apr 26 21:21:11 2000
From: nascheme at enme.ucalgary.ca (Neil Schemenauer)
Date: Thu, 27 Apr 2000 01:21:11 GMT
Subject: Python compiler ? (was Re: Python for drivers?)
References: <3900EDBE.282BC85@dsuper.net> <8dr2pn$ec8$1@newshost.accu.uu.nl> <3906f804@nntp.server.uni-frankfurt.de> <015b01bfaf8b$9cd03930$0500a8c0@secret.pythonware.com> <39076316@nntp.server.uni-frankfurt.de> <8e7sop$5io$1@nnrp1.deja.com> <FtnGu1.Dso@world.std.com>
Message-ID: <slrn8gf5im.vjb.nascheme@cranky.arctrix.com>

Will Ware <wware at world.std.com> wrote:
>I'd also be willing to kick in $100 for an efficient Python
>compiler, but only if it were open source.

I don't think this guy is making native code compilers.  The
"compiled" TkDiff demo is still implemented using Tcl.  Unlike my
grandfather, I can't read German so I can't figure out what his
"compiler" is supposed to do from the "whitepaper".  Babelfish
does not seem to help.

Making Python more efficient is not an easy task.  If you are
looking for an easy solution you are out of luck.  It is probably
more useful to put effort into something like into Gwydion Dylan.

    Neil

-- 
"Everyone can be taught to sculpt: Michelangelo would have had to
be taught how not to. So it is with the great programmers" -- Alan Perlis


From mal at lemburg.com  Mon Apr  3 09:28:13 2000
From: mal at lemburg.com (M.-A. Lemburg)
Date: Mon, 03 Apr 2000 15:28:13 +0200
Subject: Unicode program representation
References: <015c01bf9d0b$76f9b2a0$59dd3fcb@neil>
Message-ID: <38E89C6D.E515DE8@lemburg.com>

Neil Hodgson wrote:
> 
>    This leads to the question of what the use of the u" form is. The current
> answer is that u" takes the ASCII string and makes a Unicode string object
> by extending each byte with another zero byte. Because its a Unicode string
> object it behaves appropriately with Unicode aware functions.
> 
>    I think this should be changed to interpreting the literal as a UTF-8
> literal. The advantage here is that non-roman string literals become a
> natural part of the language.

There was some discussion about this when the Unicode integration
was designed late last year. We finally decided to use a fixed
internal default encoding and have the user apply all necessary
conversions at his/her own will. As you might have guessed,
the default encoding is UTF-8. However, since Python scripts
are intended to be writable in plain 7-bit ASCII, there had
to be a painless form for encoding Unicode in 7-bit ASCII.
This is what the 'unicode-escape' encoding does.

Here's a quote from the Misc/unicode.txt file which currently
is the main documentation source:

"""
Python should provide a built-in constructor for Unicode strings which
is available through __builtins__:

  u = unicode(encoded_string[,encoding=<default encoding>][,errors="strict"])

  u = u'<unicode-escape encoded Python string>'

  u = ur'<raw-unicode-escape encoded Python string>'

With the 'unicode-escape' encoding being defined as:

? all non-escape characters represent themselves as Unicode ordinal
  (e.g. 'a' -> U+0061).

? all existing defined Python escape sequences are interpreted as
  Unicode ordinals; note that \xXXXX can represent all Unicode
  ordinals, and \OOO (octal) can represent Unicode ordinals up to U+01FF.

? a new escape sequence, \uXXXX, represents U+XXXX; it is a syntax
  error to have fewer than 4 digits after \u.

For an explanation of possible values for errors see the Codec section
below.

Examples:

u'abc'          -> U+0061 U+0062 U+0063
u'\u1234'       -> U+1234
u'abc\u1234\n'  -> U+0061 U+0062 U+0063 U+1234 U+005c

The 'raw-unicode-escape' encoding is defined as follows:

? \uXXXX sequence represent the U+XXXX Unicode character if and
  only if the number of leading backslashes is odd

? all other characters represent themselves as Unicode ordinal
  (e.g. 'b' -> U+0062)


Note that you should provide some hint to the encoding you used to
write your programs as pragma line in one the first few comment lines
of the source file (e.g. '# source file encoding: latin-1'). If you
only use 7-bit ASCII then everything is fine and no such notice is
needed, but if you include Latin-1 characters not defined in ASCII, it
may well be worthwhile including a hint since people in other
countries will want to be able to read you source strings too.
"""

BTW, if you prefer a different "default" encoding, simply
write a wrapper for unicode() which uses the modified default:

def u(text,encoding='latin-1'):
    return unicode(text, encoding)

You can also have stdout write Latin-1 for Unicode strings,
by wrapping it using the StreamCodecs in codecs.py.

-- 
Marc-Andre Lemburg
______________________________________________________________________
Business:                                      http://www.lemburg.com/
Python Pages:                           http://www.lemburg.com/python/




From dworkin at ccs.neu.edu  Sat Apr 22 14:10:48 2000
From: dworkin at ccs.neu.edu (Justin Sheehy)
Date: 22 Apr 2000 14:10:48 -0400
Subject: LAmbda calculus evaluator (interactive)
In-Reply-To: Andrew Cooke's message of "Sat, 22 Apr 2000 08:33:42 GMT"
References: <8dro55$eq6$1@nnrp1.deja.com>
Message-ID: <vndk8hq3t1z.fsf@camelot-new.ccs.neu.edu>

Andrew Cooke <andrew at andrewcooke.free-online.co.uk> writes:

> I know where I can get various (Oca)ML programs, but I'm not sure
> they're interactive

% ocaml
        Objective Caml version 2.04

# 1 + 1;;
- : int = 2
# 

-Justin

 




From philh at vision25.demon.co.uk  Sun Apr 30 16:12:04 2000
From: philh at vision25.demon.co.uk (phil hunt)
Date: Sun, 30 Apr 2000 21:12:04 +0100
Subject: The Simple Economics of Open Source
References: <B527E7A5.51AA%raffael@mediaone.net> <MPG.136cc01af4ab703b989709@news.earthlink.net> <raffael-4899DC.23562823042000@news.ne.mediaone.net>  <8e140f$cco$1@slb0.atl.mindspring.net>
Message-ID: <slrn8gp4sk.eb1.philh@vision25.demon.co.uk>

On Mon, 24 Apr 2000 03:29:29 -0600, 
Andrew Dalke <dalke at acm.org> wrote:
>
>Raffael Cavallaro wrote:
>>Fact is, only commodity items, that are known to many programmers, are
>>open sourced. If there were sufficient demand to support a closed source
>>business, then it would have happened that way. The proof of this is the
>>BSDs, which would allow a closed source, proprietary fork at any time.
>>Hasn't happened 'cause there's no demand for such a product.
>
>
>Not quite fact.  There are counter-examples.  What's the major
>competitor to TeX?  From my understanding, when it was first released
>there was nothing like it, proprietary or otherwise, so it wasn't
>a commodity.
>
>The first few web servers and browsers weren't commodities, but they
>were open source.
>
>But I do agree that most open-source projects seem to be years behind
>their closed-source equivalents.

How much more advanced is the closed-source equivalent of Leafwa?

-- 
***** Phil Hunt ***** send email to phil at comuno.com *****
Moore's Law: hardware speed doubles every 18 months
Gates' Law: software speed halves every 18 months 


From robin at jessikat.demon.co.uk  Wed Apr 26 09:30:52 2000
From: robin at jessikat.demon.co.uk (Robin Becker)
Date: Wed, 26 Apr 2000 14:30:52 +0100
Subject: The Simple Economics of Open Source
References: <14591.15630.192472.440375@goon.cnri.reston.va.us>
 <slrn8fv2rd.1kt.neelk@brick.cswv.com> <3900611F.3B525BDA@cfdrc.com>
 <FtDG4z.774@world.std.com> <uFaBgiAr5HA5EwJt@jessikat.demon.co.uk>
 <3906D242.EB68BFED@zipzap.ch>
Message-ID: <NdJYmAAM+uB5Ew0L@jessikat.demon.co.uk>

In article <3906D242.EB68BFED at zipzap.ch>, Boris Borcic <zorro at zipzap.ch>
writes
>Robin Becker wrote:
>> 
>> I'm afraid your desire for a better world reduces your altruism
>> quotient.
>> Doing it because it helps others is altruism, doing it to help your self
>> and or relatives isn't really.
>
>This comment can be made irrefutable, provided you *define* "self
>and relatives" as the beneficiaries of ("pseudo") altruism.
>
>Reciprocally, altruism may be defined as the action of paradoxical
>definitions of self. Paradoxical - means it conflicts with the
>choices that follow from other, more common, definitions of self,
>that make the term "selfish" determinate.
>
>Language that postulates a unique applicable definition of self
>from the start, is thus inadequate to an enlightened discussion
>of altruism, imho.
>
>BB
>--
>"I think as I think, therefore I am who I am"

wow :)
-- 
Robin Becker


From gmcm at hypernet.com  Fri Apr 14 12:39:22 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Fri, 14 Apr 2000 12:39:22 -0400
Subject: problem with os.system calls under win32
In-Reply-To: <8d7c0k$fjr$1@nnrp1.deja.com>
Message-ID: <1256392132-57146878@hypernet.com>

william_93402 at my-deja.com wrote:

> I am attempting to do a system call in a directory with spaces in the
> name. For example:
> 
> >>>import os
> >>>os.system('c:\\f o o\\bar.exe')
> 1
> 
> when I place the executable in a dir such as c:\\foo I can successfully
> execute the system call. Is there something special I have to do when
> directory names contain spaces? Just for notes I am running by example
> in build 125 of pythonwin.

Follow MS's rules:
os.system('"c:\\f o o\\bar.exe"')

(must be double-quotes).

- Gordon



From stevep at inkontact.com.au  Mon Apr 17 20:24:12 2000
From: stevep at inkontact.com.au (Stephen Purcell)
Date: Tue, 18 Apr 2000 10:24:12 +1000
Subject: [Module] PyUnit 1.0.1 - Unit testing framework
Message-ID: <38FBAB2B.F41415AA@yahoo.com>

                             PyUnit 1.0.1                             
                             ------------                            

Unit testing framework for Python.

The Python unit testing framework, dubbed 'PyUnit' by convention, is a
Python language version of JUnit and XUnit, the de facto standard unit
testing frameworks for Java and Smalltalk respectively. 

PyUnit provides a standard, proven, simple and elegant way to write unit
tests for Python software.

PyUnit has been widely used for a number of months, and this new
packaging is therefore considered a stable release.

       URL:  http://pyunit.sourceforge.net/
  Download:  http://download.sourceforge.net/pyunit/pyunit-1.0.1.tar.gz

   Licence:  Python

  Categories:  Testing


Steve Purcell, Technical Director
Inkontact, http://www.inkontact.com/

--
<a href="http://pyunit.sourceforge.net/">PyUnit 1.0.1</a> -- Unit
testing framework for Python



From PClaerhout at CREO.BE  Wed Apr 12 07:38:41 2000
From: PClaerhout at CREO.BE (Pieter Claerhout)
Date: Wed, 12 Apr 2000 13:38:41 +0200
Subject: Using xmllib
Message-ID: <2B1262E83448D211AE4B00A0C9D61B03BA72A1@MSGEURO1>

Hello,

does anyone has an example on how to use the xmllib to parse an xmlfile?
I have an xml-file, which looks a bit like this:

<!-- a general comment -->
<COLLECTION VERSION="1">
    <GENERAL PATH="S:\" />
    <ITEM NAME="Item1">
        <DESCRIPTION>Item 1 description.</DESCRIPTION>
        <DATE>Thu Dec 16 16:35:05 1999</DATE>
    </ITEM>
    <ITEM NAME="Item2">
        <DESCRIPTION>Item 2 description.</DESCRIPTION>
        <DATE>Wed Mar 15 11:54:27 2000</DATE>
    </ITEM>
</COLLECTION>

I want to convert it to something like this:

Collection from path s:\

- Item 1 (Thu Dec 16 16:35:05 1999)
  Item 1 description.
- Item 2 (Wed Mar 15 11:54:27 2000)
  Item 2 description.

Anyone an idea on how to do that with XMLlib?

Kind regards,


Pieter
Pieter_Claerhout at creoscitex.com



From parker at gol.com  Thu Apr  6 18:32:36 2000
From: parker at gol.com (Ian Parker)
Date: Thu, 06 Apr 2000 22:32:36 GMT
Subject: Odp: Python on Alpha OpenVMS
References: <z4_C4.6214$4N.168038@news.tpnet.pl> <hDlBpKAS7T34Ew1e@gol.com>
 <_KCD4.2797$a01.54385@news.tpnet.pl>
Message-ID: <ZRhvtDAzBR74EwzH@gol.com>

In article <_KCD4.2797$a01.54385 at news.tpnet.pl>, Tomek Lisowski
<Lisowski.Tomasz at sssa.nospam.pl> writes
>U?ytkownik Ian Parker <parker at gol.com> w wiadomo?ci do grup dyskusyjnych
>napisa?:hDlBpKAS7T34Ew1e at gol.com...
>> In article <z4_C4.6214$4N.168038 at news.tpnet.pl>, Tomek Lisowski
>> <Lisowski.Tomasz at sssa.nospam.pl> writes
>> >Hi!
>> >
>> >I have tried to compile Python sources on my VAX OpenVMS system, but soon
>I
>> >realized, that instead of my trusty DEC C++ compiler I need a native C
>> >compiler. Unfortunately my C++ compiler cannot be reverted to a native C
>> >language mode. Has anyone succeeded to create VAX OpenVMS Python
>binaries?
>> >
>> >Note: Apart from VAX OpenVMS I have also an Alpha OpenVMS server, but
>here I
>> >don't have any C/C++ compiler available. If anyone has created Alpha
>OpenVMS
>> >Python binaries, please contact me too.
>> >
>> >Thanks in advance
>> >
>> >Tomasz Lisowski
>> >
>>
>> Tomasz
>>
>> If you aren't planning changes to the source, why not just download and
>> link the object modules from:
>> http://decus.decus.de/~zessin/python/obj.html
>>
>> I'm so inspired by this solution, I'm at last downloading them myself.
>
>I am just not sure, for which platform were these modules compiled, and for
>which version of OpenVMS system. I suppose, there could be some differences
>between the OpenVMS versions. At least you can see them in the source code,
>where there are some #ifdef and #if directives concerning the OpenVMS
>version string.
>
>Please report, if you succeeded in making your Python running properly,
>including the platform and OS info.
>
>Regards
>
>Tomasz Lisowski
>

Tomasz

I'm glad to say that the VMS objects worked perfectly on both Alpha and
VAX VMS systems.  They're both V7.x - I didn't bother to check specific
version numbers.  I'm planning to deploy to about 20 VMS Alpha and VAX
boxes.  I even built a backup saveset of the resultant directory tree
and a have a DCL script that does the VAX or Alpha assembly and linking
on the target box - this keeps it very simple.   I may even go all the
way and make a VMSINSTALL/PCSI kit.    

I'm really impressed with the OpenVMS port (Uwe, please keep up the good
work!).  Only one qualification, at the moment I think the objects don't
support threading, but haven't really investigated yet.

Regards


Ian Parker


From ralph at inputplus.demon.co.uk  Sun Apr  2 07:31:21 2000
From: ralph at inputplus.demon.co.uk (Ralph Corderoy)
Date: 2 Apr 2000 12:31:21 +0100
Subject: else clauses in while and for loops
References: <Zv_C4.6415$4N.169440@news.tpnet.pl> <38DD2EE3.C7F3CB54@webamused.com> <8boe2k$gmh$1@inputplus.demon.co.uk> <v0RD4.13$n68.186989568@newsb.telia.net>
Message-ID: <8c7b29$37v$1@inputplus.demon.co.uk>

Hi,

> > I don't think that is a real need.
> 
> so?  do you suggest that we should remove it, and force people to
> rewrite their perfectly working pro- grams, just to make you happy?
> or maybe we could use all that energy to work on something more
> important?

I suggest you go back and read my post properly.  In particular, the
original poster's request.

> > Please give some examples showing the real need for else clauses in
> > this context.

Someone replied giving a pointless use of else.  I said out that this
was a pointless use of else and therefore didn't match the original
poster's request of an example `showing the real need'.

I didn't say there wasn't a real need for else, I said `I don't think
that is a real need' immediately after quoting the pointless example.


Ralph.


-------- Original post follows

> > Please give some examples showing the real need for else clauses in
> > this context.
>
> def lookup(stringList, word):
>     """Finds the first instance of word in stringList. Returns -1 if not
> found"""
>     i = 0
>     while i < len(stringList):
>         if stringList[i] == word:
>             break
>         i = i + 1
>     else:
>       # is only executed if we never hit that break
>         return -1
>     return i

I don't think that is a real need.  Wouldn't it normally be written to
return ASAP rather than jump through hoops.

    def lookup(stringList, word):
        i = 0
        while i < len(stringList):
            if stringList[i] == word:
                return i
            i = i + 1
        return -1


Ralph.



From moshez at math.huji.ac.il  Wed Apr  5 18:00:07 2000
From: moshez at math.huji.ac.il (Moshe Zadka)
Date: Thu, 6 Apr 2000 00:00:07 +0200 (IST)
Subject: Extending Python with C-Libs
In-Reply-To: <8cfpl8$9b2$1@newnews.ilk.de>
Message-ID: <Pine.GSO.4.10.10004052359350.29608-100000@sundial>

On Wed, 5 Apr 2000, Frank Rotermund wrote:

> Hello
> 
> I have a Library written in C with its Headerfiles but without the
> Sourcecode.
> Does anybody know an easy way to extend Python with this Library?

Use SWIG (http://www.swig.org). It only needs headers anyway.
--
Moshe Zadka <mzadka at geocities.com>. 
http://www.oreilly.com/news/prescod_0300.html
http://www.linux.org.il -- we put the penguin in .com




From thomas at cintra.no  Thu Apr 27 09:57:01 2000
From: thomas at cintra.no (‘5ÛHH575-UAZWKVVP-7H2H48V3)
Date: Thu, 27 Apr 2000 13:57:01 GMT
Subject: About Zope`s FTP-server
Message-ID: <39084411.13519730@news.online.no>

Hi,

Just logged into the running ftp-server running with Zope. Am I inside
a database suddenly? I tried to locate the "control_panel" folder
listed in the ftp-client I used inside the Zope-installation dir, but
no. Then I tried to create a folder and look for it, nowhere to be
found. But my database inside Zope was bigger. This is fine and nice
for small pics and html-stuff, but I want to upload pickled objects
that can be several megabytes in size, then have a Zope-like or
Zope-integrated method store them in some database, NOT gadfly, cuz
the info is so huge. What then? If all I upload end up inside a
database my question is; can that database, I guess it`s gadfly,
handle it? If not, can I store uploaded stuff someother way, outside
the database? 

I`d like to have users connect to the zope ftp-server, logging into a
area set up especially for that user, similar to ordinary ftp-servers.
This area would have to be available to other python-methods, either
thru that darn database-thingy or some other way. The access-control
has to be maintained too. Each user has one user-area that user can
access and not, at least no write/delete etc. access, to other areas.

Am I making any sense??!

The bottom line is; can the built-in stuff handle huge amounts of
data, spew it out quickly and in case of disaster, can the data be
restored easily? 

Thanks. I`ll go take my medication now. ;->

Thomas


From nmm1 at cus.cam.ac.uk  Mon Apr 10 09:52:56 2000
From: nmm1 at cus.cam.ac.uk (Nick Maclaren)
Date: 10 Apr 2000 13:52:56 GMT
Subject: Python paradigms
References: <Pine.LNX.4.21.0004100757580.19783-100000@fep132.fep.ru> <8cs70f$ojf$1@pegasus.csx.cam.ac.uk> <38F1C600.3FD9878B@acm.org>
Message-ID: <8csmbo$99t$1@pegasus.csx.cam.ac.uk>

"Robert W. Cunningham" <rcunning at acm.org> wrote:
>
> > At best, confusing code wastes effort.  At worst, it causes mistakes
> > to be introduced.  I have seen the latter effect many times.
> >
> > That is why I was talking about the benefits of those paradigms for
> > code clarification, and neither efficiency nor compactness.  Python
> > has some very useful facilities, but is a bit tedious for some kinds
> > of scripting.
> 
> Again, comments in the code can cover all these bases.  A good
> "paradigm" is a good thing to have, but it is no Holy Grail.  It is but
> one step in the creation of a good program.  ...

Your second sentences are correct, but your first is most definitely
false.  Dangerously so.  I would be the last person to agree with that
nonsense that any languages or programming styles are "self commenting",
but the converse is equally wrong.  Using unclear paradigms (whether
out of 'cleverness' or because the language forces you to) has the
following major problems:

    1) It makes it very hard to see when your code does not match your
intent.  I have frequently had arguments with people who have taken
your viewpoint, when I have pointed out that the code that they wrote
only SEEMS to match their comments!  Typically, the error is that it
does so, except for a few, nasty boundary cases.

    2) Code rot.  In the days of punched cards, only a few programmers
maintained their comments.  Nowadays, most do, but a large number tend
to cut corners when adding a 'simple' modification to trap a previously
unthought-of boundary condition (see above.)  And have you every tried
to get a large vendor to accept a bug report on a COMMENT when the code
works correctly?

    3) It is very hard to express the sort of details that cause
trouble in English - comments are good at saying WHAT should be done
but not so good at saying EXACTLY HOW.  This is best done in a pseudo
mathematical language, such as a programming language - which is, after
all, precisely what ALGOL 60 was designed for!

> Have any of you ever had to port a product written in a "dead" computer
> language?  I have, more than once.  The comments were the ONLY thing to
> make the task possible, since even texts adequately documenting the old
> languages (especially the libraries) were extremely difficult to find.

Not often - probably only a few dozen times on sizeable codes.  In the
cases that I had too look at in detail, things broke down something like
the following (all figures being wildly incorrect):

    75%    comments and code matched, and either was clear enough
    15%    comments and code matched, but both were inadequate
     5%    comments and code didn't match, and the code was correct
   2.5%    comments and code didn't match, and the comments were
   2.5%    I gave up, and reimplemented from my own algorithm!

> Human language, written well, is far more expressive than computer
> languages.  That's why computer hardware and software is documented with
> books, and not in their "own" languages.  ...

Not so.  They are written that way 'for simplicity' (i.e. because it is
quite hard to get most people to read mathematics as a specification)
and because there are no standardised languages for expressing the
semi-mathematical concepts.  However, when specifying things like
library functions, they very often ARE specified in terms of an
'example implementation' - i.e. in their own languages.

> At best, sound paradigms can simplify code and reduce the level of
> commenting needed, but can in no way eliminate the need for thorough
> commenting and other documentation.

And conversely.

> And if you can't write good comments, how can you POSSIBLY believe you
> can write good code?

And, if you can't say what you mean in a semi-precise language like
a programming language, how are you going to do it in English?


Regards,
Nick Maclaren,
University of Cambridge Computing Service,
New Museums Site, Pembroke Street, Cambridge CB2 3QG, England.
Email:  nmm1 at cam.ac.uk
Tel.:  +44 1223 334761    Fax:  +44 1223 334679


From moshez at math.huji.ac.il  Wed Apr  5 02:57:53 2000
From: moshez at math.huji.ac.il (Moshe Zadka)
Date: Wed, 5 Apr 2000 08:57:53 +0200 (IST)
Subject: Why should I switch to Python? - Infinity of Primes
In-Reply-To: <1257244378-5884705@hypernet.com>
Message-ID: <Pine.GSO.4.10.10004050856340.24181-100000@sundial>

On Tue, 4 Apr 2000, Gordon McMillan wrote:

> Oh, come on. There are at least 3 well known constructive 
> proofs that all odd numbers are prime:
> 
> The mathematician's:
>  3 is prime,
>  5 is prime,
>  7 is prime
>  as so on, by induction.
> 
> The physicist's:
>  3 is prime,
>  5 is prime,
>  7 is prime,
>  9 is prime <experimental error>
>  11 is prime...
> 
> The computer scientist's:
>  3 is prime,
>  3 is prime,
>  3 is prime ....

the biologist's:
  3 is prime,
  5 is prime,
  7 is prime,
  9 is prime,
 11 is prime,
--
Moshe Zadka <mzadka at geocities.com>. 
http://www.oreilly.com/news/prescod_0300.html
http://www.linux.org.il -- we put the penguin in .com




From loredo at spacenet.tn.cornell.edu  Fri Apr 28 16:22:13 2000
From: loredo at spacenet.tn.cornell.edu (Tom Loredo)
Date: Fri, 28 Apr 2000 16:22:13 -0400
Subject: What does "first class" mean?
Message-ID: <3909F2F5.A6F91256@spacenet.tn.cornell.edu>

Hi folks-

I've often seen reference here (and on the Ruby page) to Python's
treatment of functions and methods as "first class objects."  What
does this mean?  I thought I had most of the OO lingo in hand after
a year of EE/CS courses at MIT in the early 80s (back then most it
fell under the rubric of "data abstraction").  But "first class"
in this context must have been in one of the lectures I missed.
(There were quite a few of those, actually!)

Thanks,
Tom Loredo


From effbot at telia.com  Sun Apr  2 19:44:23 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Sun, 02 Apr 2000 23:44:23 GMT
Subject: NameError when assigning dictionary values
References: <20000402231757.6604.qmail@web2105.mail.yahoo.com>
Message-ID: <rXQF4.4797$74.76625@newsc.telia.net>

lewst <lewst at yahoo.com> wrote:
> I was trying to build a dictionary and create lists corresponding to
> the value names in the dictionary in one loop.

unlike Tcl, you don't need to create *variables* to be
able to store *values* to your dictionary.

> Do I really have to type "one" and "two" twice?  Once to make it a
> variable and a second time to add to my dictionary?  I.e,

no.  reset your brain.  forget about variables.  just put
new *values* (aka objects) in the dictionary.

> This seems error prone if the the number of variables grows very large
> as it will for me.

let me guess: what you really want to do is to have a
number of lists, each of which are identified by a string?

    mydict = {}
    for k in ("one", "two"):
        mydict[k] = []

and then you want to update one of them:

    mydict["one"].append("1")

and then you want to print them all:

    for k, v in mydict.items():
        print k, v

did I successfully read your mind this time?

</F>

<!-- (the eff-bot guide to) the standard python library:
http://www.pythonware.com/people/fredrik/librarybook.htm
-->




From dan at cgsoftware.com  Wed Apr  5 19:48:27 2000
From: dan at cgsoftware.com (Daniel Berlin)
Date: Wed, 5 Apr 2000 16:48:27 -0700 (PDT)
Subject: Pass by reference ?
In-Reply-To: <38EBCD7D.7030602@acm.org>
Message-ID: <Pine.LNX.4.10.10004051644260.13009-100000@propylaea.anduin.com>

> 
>    def Nuke(w):
>      print w,
>      w=None
>      print w

Sorry, this function is not a good demonstration.
w is a reference to whatever is passed in.
Then you are making w a reference to None.
What you should try is w.append(5), or something

Try thusly:
def Appender(w):
	print w
	w.append(5)
	print w
Now, let's call this:

x=[]
Appender(x)

You'll see:
[]
[5]
Then, if you print x, you'll see:
[5]

Thus, we passed it by reference, because it modified the original.
HTH,
Dan
> This behavior meets the classical definition of "Pass By Value".

Bad example.

> 
> What other tests have I omitted, and what do they yield?
See above.

> 
> 
> -BobC
> 
> -- 
> http://www.python.org/mailman/listinfo/python-list
> 




From MSteed at altiris.com  Tue Apr 25 14:01:00 2000
From: MSteed at altiris.com (Mike Steed)
Date: Tue, 25 Apr 2000 12:01:00 -0600
Subject: VIM NT +python
Message-ID: <65118AEEFF5AD3118E8300508B1248774CB2B6@ALTNET>

Hi,

   I just built the latest version (5.6.69) with the default feature set,
plus the Python interface.  I have not tested it beyond

   :python print 'hi'

which passed with flying colors.

http://athani.pair.com/msteed/vim/gvim56py.zip

M.

> From: Me [mailto:me at nospam.edu]
> Sent: Tuesday, April 25, 2000 11:27 AM
> To: python-list at python.org
> Subject: Re: VIM NT +python
> 
> 
> So would I!
> 
> <tiddlerdeja at my-deja.com> wrote in message
> news:8e4j27$hgq$1 at nnrp1.deja.com...
> > Does anyone here have a winNT build of the VIM editor that 
> has python
> > support switched on?
> >
> > I'd love to get my hands on a copy.
> >
> >
> > Sent via Deja.com http://www.deja.com/
> > Before you buy.



From dalke at acm.org  Sat Apr  1 05:25:22 2000
From: dalke at acm.org (Andrew Dalke)
Date: Sat, 1 Apr 2000 03:25:22 -0700
Subject: arraymodule questions
Message-ID: <8c4iqv$2na$1@nntp9.atl.mindspring.net>

Hello,

  (the questions are at the end; but first, discussion)

  I'm working on a project to store and analyze protein and
nucleotide sequences with Python.  These are often stored
as character strings.  I'm trying to find out the best storage
container for the data.  I've been experimenting with
array.array of characters instead of string because of its
in-place mutability.

For example, suppose I use a string as my primary data type,
then a function might be written as (given in another recent
post of mine)

def translate(dna, table = Translation.standard_table):
  protein = []
  for i in range(0, len(dna), 3):
    protein.append(table[dna[i:i+3]])
  return string.join(x, "")

This stores the temporary data in an [], and I don't like
that.  I would rather not have the additional conversion
back to a string.  With a character array:

def translate(dna, table = Translation.standard_table):
  dna = dna.tostring()
  protein = array.array( ("c", "") )
  for i in range(0, len(dna), 3):
    protein.append(table[dna[i:i+3]])
  return protein

This also requires a conversion, but conversion to a string
uses about 1/4th the memory as using a list, and the string
conversion is trivial since the array stores its data as a
block of characters.  (It's only 1/4 since Python special
cases single letter strings to reuse the same object.)


However, the performance of array.array is about 30% slower
than appending to a list followed by conversion to string.

Tracking it down, about 20% of the time is spent because
ins1 in arraymodule.c calls c_setitem twice.  The first
checks if the input object is a character, but with an index
meaning "don't acually set the item."  The second time
occurs after the array was resized, to actually set the
item.

What I did was add special code for append rather than
calling the generic insert function.  This always increases
the size by 1 item - optimistically assuming the setitem
will work.  Then it calls setitem and, if that fails, decrement
the size count by 1 without resizing the array back down.

(My assumption is working code most likely won't fail, so the
amortized cost of the infrequent extra realloc and few bytes
will be low.)

I got another 20% performance by preallocating an additional
64 items when the array size is increased, which reduces
the number of realloc calls.

With these changes, the two versions (string and array) of
my timing test come out rougly the same.  (I know, 30%-20%-20%
doesn't come out to be 0% - the percents aren't all based on
the same initial timings.)

I started thinking seriously about using an array for the
base character storage, but then noticed that arraymodule
doesn't expose the array object C level API.  There structure
is only defined in arraymodule.c, and not in any header file.
There will be C extensions which need access to the character
data, so I can't use an array.

So my questions are about changes for Python 1.6:
  1) would it be okay to make the arrayobject C API available?
  2) could I add the .append optimizations mentioned above?
  3) should I also add the code to preallocated space?

Some of the Python methods are commented out, such as "count".
Why is this so?  There are comments implying that if someone
needs them, they can be added.  I could do this.

I notice "extend" doesn't exist.  I could also add that.

                    Andrew Dalke
                    dalke at acm.org





From cnetzer at stanford.edu  Sat Apr 15 21:53:31 2000
From: cnetzer at stanford.edu (Chad Netzer)
Date: Sat, 15 Apr 2000 18:53:31 -0700
Subject: Python 1.6a2 crashes in test_fork1 on my Mandrake Linux 7 machine
References: <20000413010344.A13931@owl.rhein-zeitung.de> <38F658C6.66AD9CA5@clark.net> <20000414154718.A15664@xs4all.nl> <20000414190308.A30502@owl.rhein-zeitung.de>
Message-ID: <38F91D1B.3D663B0C@stanford.edu>

Oliver Andrich wrote:

> On Fri, Apr 14, 2000 at 03:47:18PM +0200, Thomas Wouters wrote:
> > I dont know enough about threads to say anything useful about this, but it
> > is kind of strange that you dont see the same problem under Python 1.5.2.

Well, I've been fooling around trying to figure this one out.  It doesn't appear
to
be related to GCC.  I tried the exact same GCC's on both a Redhat 6.2 and my
Mandrake 7 (w/ cooker) and it worked fine on Redhat, but failed on Mandrake.
I also downloaded a recent (4/10/2000) GCC from cvs, and had the same results
on both systems.  Also, test_fork1.py fails for me on Mandrake even with Python
1.5.2.  I also tried it w/ Linux kernel 2.3.pre99-5 and 2.2.14; same results.
And finally,
I installed the Mandrake glibc on Redhat and *still* had the same results.
However, my Mandrake machine is 2 CPU SMP, while the redhat was one CPU,
which leads me to the next phase of testing...

I traced the error, using strategic printf's, to the posixmodule posix_waitpid()
function, and I think it is a bug in the  Py_END_ALLOW_THREADS or
Py_BEGIN_ALLOW_THREADS macros.  Basically, when the problem occurs,
which is not always (and the frequency of occurence seems to change depending
on how many print statements I use in test_fork1.py), the code seems to go into
an
infinite loop in either of these two macros (using "top" shows python at 99%
cpu).
Here is an example of the "top" output when the test_fork1.py program locks up.

54 processes: 49 sleeping, 4 running, 1 zombie, 0 stopped
CPU states: 54.5% user,  2.8% system,  0.0% nice, 42.5% idle
Mem:   127780K av,  124460K used,    3320K free,   80140K shrd,    2472K buff
Swap:  265032K av,     836K used,  264196K free                   44804K cached

  PID USER     PRI  NI  SIZE  RSS SHARE STAT  LIB %CPU %MEM   TIME COMMAND
14133 chad      18   0  1172 1172   804 R       0 98.2  0.9  11:15 python
  521 root       0   0 28712  28M  2260 S       0  7.7 22.4   3:36 X

This indicates to me (naive as I am about threads) that there is spinlock going
nuts
waiting forever for threads to block or unblock.   Try putting a fprintf() just
before
and after these macros in posixmodule.c:posix_waitpid().  The actual kernel
waitpid()
call itself, and thus the forking, seems to be fine.

> May be we have a again a code generation problem on
> Mandrake 7.

I hope not... Because of school, work, and taxes, I'm probably too busy to track
this one
down further.  I forgot to test my Mandrake compiled binary on a Redhat 6.2
system, so
I would suggest that as another test.  But it appears, from where I'm sitting, to
be a problem
with threading, and not necessarily w/ code generation (for once).

If I discover more, I'll post it here.

Chad Netzer
cnetzer at stanford.edu




From bgdarnel at my-deja.com  Sat Apr 15 21:44:37 2000
From: bgdarnel at my-deja.com (Ben Darnell)
Date: 16 Apr 2000 01:44:37 GMT
Subject: PyApache and mxDateTime
Message-ID: <slrn8fi6o6.8na.bgdarnel@Sull-04-227.rh.ncsu.edu>

When I use the following trivial script under PyApache, it fails:
#!/usr/bin/python

print "Content-type: text/plain"
print

try:
        import DateTime
except:
        import traceback
        import sys
        sys.stderr=sys.stdout
        traceback.print_exc()

Output:
Traceback (innermost last):
  File "/usr/lib/cgi-bin/foo.py", line 7, in ?
    import DateTime
  File "/usr/lib/python1.5/site-packages/DateTime/__init__.py", line 20, in ?
    from DateTime import *
  File "/usr/lib/python1.5/site-packages/DateTime/DateTime.py", line 817, in ?
    def add_century(year,
TypeError: call of non-function (type None)

Going without the DateTime module is not an option, as it is required for 
MySQLdb.  At this time, I can get by without PyApache, but I would like
to have the option of switching to a more efficient processing model
when the need arises.  Is there any workaround for this problem?

-Ben


-- 
Ben Darnell              bgdarnel at unity.nc su.edu
http://thoughtstream.org



From tjdukes at my-deja.com  Tue Apr  4 12:23:52 2000
From: tjdukes at my-deja.com (tjdukes at my-deja.com)
Date: Tue, 04 Apr 2000 16:23:52 GMT
Subject: bug in build on Solaris?
Message-ID: <8cd4u5$6tu$1@nnrp1.deja.com>

I posted a message yesterday about not being able to perform the demo in
Demo/extend. This morning I tried the following.
- I erased Python-1.5.2/python and typed 'make'.
- The command used to link python was
   gcc python.o \
     ../libpython1.5.a -lsocket -lnsl -ldl -lthread -lm -o python
- I erased python again and typed the following commands in by hand.
  cd Modules
  gcc -export-dynamic python.o ../libpython1.5.a \
    -lsocket -lnsl -ldl -lthread -lm -o python
  mv python ..
This makes the example work correctly. Is this a bug in the makefile
shipped with Python 1.5.2 or is there a configuration option for this?
thanks,
Todd.


Sent via Deja.com http://www.deja.com/
Before you buy.


From embed at geocities.com  Thu Apr 20 11:15:38 2000
From: embed at geocities.com (Warren Postma)
Date: Thu, 20 Apr 2000 11:15:38 -0400
Subject: __del__ problem - would adopting Garbage Collection fix this?
References: <000801bfaa9c$c9957aa0$d62d153f@tim>
Message-ID: <vdFL4.9899$HG1.265876@nnrp1.uunet.ca>

> BTW, this was all discussed at extreme length on Python-Dev -- check it
out!
> I doubt anyone wants to repeat all of that whlie struggling to pretend
it's
> new & interesting <wink>.

Is there an archive on the web somewhere of Python-Dev discussions?! Sorry
if that's a FAQ.

Warren




From richard_chamberlain at ntlworld.com  Wed Apr  5 09:18:04 2000
From: richard_chamberlain at ntlworld.com (Richard Chamberlain)
Date: Wed, 5 Apr 2000 14:18:04 +0100
Subject: ASP web-type stuff
References: <RZyG4.3061$9o.5812531@news.magma.ca>
Message-ID: <8cfea6$71$1@gxsn.com>

Hi Dan,

ASP is not language dependent, most of the stuff you'll see is written in
VBScript or JScript but you can actually configure it to use other
languages.

See http://starship.python.net/crew/pirx/asp/py_asp.html to show you how to
use Python with ASP.

Richard

"Dan Green" <dgreen at kryten.kitenet.net> wrote in message
news:RZyG4.3061$9o.5812531 at news.magma.ca...
> Ahoy, folks.
> I was looking over things like Embperl and Apache::ASP and liked the
> idea. I would like the idea more if I could do with python. So I had a
> couple of questions:
>
> Is there such a thing?
> Is there interest in such a thing?
> Does anbody want to hear about my 75 lines of code hack that basically
does
> the trick, but needs a lot of work?
>
> If nobody cares at all, I'll just shut up and go home. :)
>
> For those of you who have no idea what I'm talking about, it's basically
an
> alternative to CGI that allows you to embed the code right into the HTML.
>
> something like this:
>
> <HTML>
> <BODY>
> <%  for i in range(5): %>
> Loop item <%= i %><BR>
> </BODY>
> </HTML>
>
> would look like this when called from a browser:
>
> <HTML>
> <BODY>
> Loop item 1<BR>
> Loop item 2<BR>
> Loop item 3<BR>
> Loop item 4<BR>
> Loop item 5<BR>
> </BODY>
> </HTML>
>
> Let me know. If you're interested, please e-mail me.
>
> --
> Dan Green           | "So What? It's your problem to learn to live with,
> Developer           |  destroy us, or make us saints."
> Rewind Technologies |  -- Ministry, "So What?"




From rayvd at nospam.firetail.org  Fri Apr  7 01:01:58 2000
From: rayvd at nospam.firetail.org (Ray Van Dolson)
Date: Fri, 07 Apr 2000 05:01:58 GMT
Subject: Socket Programming Question
Message-ID: <aZdH4.95109$AT6.106753@dfw-read.news.verio.net>

I'm trying to write a simple chat server w/ Python sockets that supports 
multiple connections and am having a little trouble... here's my server code so 
far:

#!/usr/bin/python

from socket import *
from threading import *

serversocket = socket(AF_INET,SOCK_STREAM)
serversocket.setsockopt(SOL_SOCKET,SO_REUSEADDR,1)
serversocket.bind((gethostname(), 9001))
serversocket.listen(5)

def client_thread(clientsocket):
        clientsocket.send("Welcome to the chat server...\r\n")
        while 1:
                data = clientsocket.recv(1024)
                if data:
                        clientsocket.send(data)
                else:
                        pass

def main_server_thread():
        while 1:
                (clientsocket, address) = serversocket.accept()
                ct = Thread(target=client_thread,args=((clientsocket),))
                ct.run()

mainThread = Thread(target=main_server_thread,args=())
mainThread.start()
while 1:
        pass

However, this code only allows one connection at a time... the other can connect 
to the socket but is simply queued.  I would think the main_server_thread() 
therad would create a new thread for each connection?

What do I need to do to support multiple, simultaneous connections with this 
code?

Thanks,
Ray Van Dolson


From JamesL at Lugoj.Com  Tue Apr  4 00:28:08 2000
From: JamesL at Lugoj.Com (James Logajan)
Date: Mon, 03 Apr 2000 21:28:08 -0700
Subject: Python 1.6 alpha 1 released
References: <200003312130.QAA04361@eric.cnri.reston.va.us> <38E52716.A1C63B5D@roguewave.com> <rJ9F4.549$n68.187227648@newsb.telia.net> <38E52D76.1820FC32@roguewave.com> <8c7tfd$k80$1@nntp9.atl.mindspring.net> <8catgc$5i84$1@uvaix7e1.comp.UVic.CA> <9i8G4.1196$n68.192123904@newsb.telia.net>
Message-ID: <38E96F58.96FCF3C5@Lugoj.Com>

Fredrik Lundh wrote:
> 
> John Aycock wrote:
> > Although I don't have time to volunteer, it wouldn't be too difficult
> > to write a program which would look through existing Python code and
> > warn of potential uses of these changed constructs.
> 
> after reading your post, tim peters used barry's new "post
> via the time machine" facility:
> 
> http://www.deja.com/=dnc/getdoc.xp?AN=591243549
> 
> changing it to look for "connect" should be trivial (he's probably
> done that by now).
> 
> fwiw, while it may miss many real appends (see comments in
> script), I'm pretty the number of missed connect calls are very
> close to zero.
> 
> </F>

And just when you think you've nailed all the connect()s in your client code
(which one must admit will in general probably occur in only one or two
locations for most network applications) and hope you've blessed all
appends, you find your server code is throwing exceptions in the bind()
calls....

P.S. Still waiting for a new release of the Unix standard so that the
spelling of the creat() system call can be corrected. <wink>


From mikael at isy.liu.se  Tue Apr 11 07:14:15 2000
From: mikael at isy.liu.se (Mikael Olofsson)
Date: Tue, 11 Apr 2000 13:14:15 +0200 (MET DST)
Subject: Redirecting the output from sys.stdout
In-Reply-To: <Pine.LNX.4.21.0004111012310.29384-100000@fep132.fep.ru>
Message-ID: <XFMail.000411131415.mikael@isy.liu.se>

On 11-Apr-00 Oleg Broytmann wrote:
 >  import os, string
 >  pipe = os.popen("dir /w", 'r') # Open pipe for reading
 >  
 >  lines = []
 >  while 1:
 >     line = pipe.readline()
 >     if not line: break # EOF
 >  
 >     lines.append(line)

Why not the following?

import os, string
pipe = os.popen("dir /w", 'r') # Open pipe for reading
lines = pipe.readlines()

Or even the following to throw away the pipe when you are done with it.

import os, string
lines = os.popen("dir /w", 'r').readlines()

/Mikael

-----------------------------------------------------------------------
E-Mail:  Mikael Olofsson <mikael at isy.liu.se>
WWW:     http://www.dtr.isy.liu.se/dtr/staff/mikael
Phone:   +46 - (0)13 - 28 1343
Telefax: +46 - (0)13 - 28 1339
Date:    11-Apr-00
Time:    13:09:36

This message was sent by XF-Mail.
-----------------------------------------------------------------------



From nmm1 at cus.cam.ac.uk  Mon Apr 10 04:31:52 2000
From: nmm1 at cus.cam.ac.uk (Nick Maclaren)
Date: 10 Apr 2000 08:31:52 GMT
Subject: Python paradigms
References: <8cmr70$b0e$1@pegasus.csx.cam.ac.uk> <8covpn$4jq$1@slb6.atl.mindspring.net> <8cqk7u$dmc$1@pegasus.csx.cam.ac.uk> <3H4I4.709$4w1.17557@typhoon.nyroc.rr.com>
Message-ID: <8cs3ho$l06$1@pegasus.csx.cam.ac.uk>

In article <3H4I4.709$4w1.17557 at typhoon.nyroc.rr.com>,
Darrell <darrell at dorb.com> wrote:
>> x = (a != NULL ? a[i]->weeble : 0) + (b != NULL ? b[i]->wombat : 0)
>
>x=0
>if a != NULL:
>    x=a[i].weeble
>
>if b != NULL:
>    x=x+b[i].wombat
>
>Isn't this more clear?
>In fact isn't not much if any longer than the "C" version when you count the
>parens used.

No.  Firstly, it relies on an artifact of my example (i.e. that it is
in an assignment, rather than a test condition or function call) and,
secondly, the more lines that code takes the harder it is to see the
function in its entirety.


Regards,
Nick Maclaren,
University of Cambridge Computing Service,
New Museums Site, Pembroke Street, Cambridge CB2 3QG, England.
Email:  nmm1 at cam.ac.uk
Tel.:  +44 1223 334761    Fax:  +44 1223 334679


From jmassung at magpiesystems.com  Mon Apr 17 20:29:19 2000
From: jmassung at magpiesystems.com (Jeff Massung)
Date: Mon, 17 Apr 2000 18:29:19 -0600
Subject: Another question
References: <dVFK4.41$vr2.5817@news.uswest.net> <38FB2EEB.A32AF88C@bam.com> <38FB339F.604364AD@bam.com> <200004171605.MAA04929@zippy.init.com> <38FB7610.80EAF965@ichips.intel.com> <200004172120.RAA11102@zippy.init.com>
Message-ID: <jWNK4.432$vr2.32822@news.uswest.net>

Donald Beaudry wrote in message <200004172120.RAA11102 at zippy.init.com>...
>and this one is better still (at least on my machine):
>
>  def do_it():
>      from random import randint
>      r = [0,] * 100000
>      for i in xrange(100000):
>         r[i] = randint(1,10)
>  do_it()
>

With the "from/import" being local, does the namespace go away after the
function is done? How fast is importing (say I called do_it() 10000 times -
then how efficient is your function?)

Jeff




From godzilla at netmeg.net  Sat Apr  1 07:49:06 2000
From: godzilla at netmeg.net (Les Schaffer)
Date: Sat, 01 Apr 2000 12:49:06 GMT
Subject: Emacs Python mode
References: <yayk8iii0ex.fsf@ganglot.ifi.uio.no>
Message-ID: <m2r9cqasoc.fsf@netmeg.net>

> I just had an idea. Using the Emacs python-mode, would it be
> possible to somehow hide the function-bodies?

check out the emacs folding-mode. all you have to do is put the
function bodies inside fold marks and you can hide their definitions. 

les schaffer


From rcunning at acm.org  Sun Apr  9 00:38:22 2000
From: rcunning at acm.org (Robert W. Cunningham)
Date: Sun, 09 Apr 2000 04:38:22 GMT
Subject: Python paradigms
References: <8cmr70$b0e$1@pegasus.csx.cam.ac.uk> <B515729E.1745%dgoodger@bigfoot.com>
Message-ID: <38F00890.3C98EBDB@acm.org>

David Goodger wrote:

> C's ?: operator, ie:
>
>     test ? true : false
>
> can be done in Python, with:
>
>     ((test and [true]) or [false])[0]
>
> Note that I said *can* be done, not *should* be done! It is very ugly and
> counter-intuitive and naughty. But sometimes necessary. Use sparingly,
> especially if your code will ever be read (even by you!).

As another Python newbie, I just came across this usage in some random
Python code I downloaded (so I could get a look at some "real world" Python
programming).  All the author did was something like this:

 blah = (((test and [true]) or [false])[0])             # blah = (test ?
true : false)

Works for me!  Only if I have a good reason to suspect a problem in this
particular line of code will I have to look past the comment.  Which, I
believe, is the entire purpose of in-line comments in the first place.  The
best way to get someone to read every single line of your code is to fail to
properly comment it.

Are there ANY Python programmers who don't also know at least this much C?

You ALL use in-line comments, don't you?  Though, in this particular case, I
suspect an apology should be added as well.

But all this can be avoided, of course, with judicious use of m4 or sed.
After all, shouldn't Python have a pre-processor too?  A pre-interpreter
interpreter?  Or, maybe, I should just stop learning Python and use C2Py
instead.  If, of course, such a beast were to exist.

All for the love of a nice big juicy side effect...


Lacking-words-for-a-better-closing-ly y'rs,

-BobC




From embed at geocities.com  Tue Apr  4 15:07:37 2000
From: embed at geocities.com (Warren Postma)
Date: Tue, 4 Apr 2000 14:07:37 -0500
Subject: What is python's language level?
References: <38EA1B8F.A797A130@lmco.com>
Message-ID: <seqG4.3372$HG1.98334@nnrp1.uunet.ca>

"Chuck Meyers" <chuck.meyers at lmco.com> wrote in message
news:38EA1B8F.A797A130 at lmco.com...
> Does anyone know python's language level? I did not find it in the list
> at:
>
> http://www.spr.com/library/0langtbl.htm
>
> I would guess that it would be somewhere between 15 and 30. This is an
> important point, even if language levels are imprecise they can make it
> easier to convince people to look at or try a language like python.
>

Cheap guess; Perl is a 15, so we'll take the Perl number.

Two things that make me laugh about this:

1. One of the highest 'level' tools is Lotus 1-2-3 for DOS? So why isn't
Linux or Windows 2000 written entirely in "lotus 1-2-3 for DOS"? That's
rich.   Go ahead and try to convince people to use Lotus 1-2-3 for DOS for
their new development efforts!

2. What's a function point?  Who decides what a "function point" is, and
says that it applies across all the various kinds of people who write
software as part of their professional duties?  How many lines of COBOL does
it take to create a python-like Dictionary object? ;-)


Warren




From stevey at polgar.cnd.hp.com  Sat Apr 22 18:16:19 2000
From: stevey at polgar.cnd.hp.com (Steve Yonkatits)
Date: 22 Apr 2000 22:16:19 GMT
Subject: SWIG
Message-ID: <slrn8g495g.svu.stevey@polgar.cnd.hp.com>

The directory that I found for SWIG was:
   ftp://ftp.cs.utah.edu/pub/dept/OLD/pub/beazley/SWIG/


Steve


From pf at sxb.bsf.alcatel.fr  Fri Apr  7 07:09:15 2000
From: pf at sxb.bsf.alcatel.fr (Pascal Fremaux)
Date: Fri, 07 Apr 2000 13:09:15 +0200
Subject: Newbie: How to do an exit ?
Message-ID: <38EDC1DA.62CBFCF3@sxb.bsf.alcatel.fr>

How to exit from a program ?
I don't find a command such as exit(0) in C

--
 Pascal Fremaux, SSII Alten
 Study Engineer at Alcatel Telecom
 R&D, Illkirch, France





From dalke at acm.org  Mon Apr  3 02:00:00 2000
From: dalke at acm.org (Andrew Dalke)
Date: Mon, 3 Apr 2000 00:00:00 -0600
Subject: new method for string objects
References: <8c44e8$5hj$1@slb6.atl.mindspring.net> <xDnF4.4507$74.72019@newsc.telia.net>
Message-ID: <8c9bru$9ut$1@slb0.atl.mindspring.net>

Fredrik Lundh wrote:
>>>> s = u"123"
>>>> b = buffer(s)
>>>> b
><read-only buffer for 7969b0, ptr 7969d0, size 6 at 796530>
>>>> len(b)
>6
>>>> b[0]
>'1'
>>>> str(b)
>'1\0002\0003\000'
>>>> fp.write(b) # writes raw contents to file without extra copy


BTW, what I'm really surprised about is that the following works:

a = "GAATAAC"
b = buffer(a)
import string
string.count(b, "A")

I guess that's what it means when the string code wants
a "read-only character buffer".  There's something fundamental
I'm missing on how strings and buffers work in Python.  Looks
like it's time to look at source code.

                    Andrew
                    dalke at acm.org





From tim_one at email.msn.com  Sat Apr  1 16:49:41 2000
From: tim_one at email.msn.com (Tim Peters)
Date: Sat, 1 Apr 2000 16:49:41 -0500
Subject: Confusion about re lookahead assertions
In-Reply-To: <8c5n76$9le$1@nnrp1.deja.com>
Message-ID: <000201bf9c24$2f91c020$482d153f@tim>

[posted & mailed]

[Skip Montanaro]
> I'm trying to use re's lookahead assertion constructs for the first time
> and am more than a bit confused. If I understand them correctly, the
> regular expression r'(?![.])([0-9]+)\s+' should only match digits if they
> are not preceeded by a dot, yet the example below clearly contradicts
> that.

It's your understanding that needs adjustment here.  An "assertion" never
consumes a character.  So, e.g.,

    (?!E1)(E2)

matches at a point p iff E1 does not match at p and E2 does match at p.  In
your case, a digit is not a decimal point, so (?![.])\d+ will match any
string of (one or more) digits.  More generally, if the set of strings E1
and E2 can match are disjoint, sticking (?!E1) in front of E2 does nothing
for you.  Think of it as "all the strings that can match E2, *except for*
the strings that match E1".  In your case, the set of strings composed of
digits minus the set of strings that begin with a decimal point is simply
the set you started with (because they have nothing in common).

> >>> pat = re.compile(r'(?![.])([0-9]+)\s+')

By the above, this is effectively the same as

    pat = re.compile(r'([0-9]+)\s+')

and everything you saw follows from that.

What you really want is a negative look*behind* assertion, to say "match a
string of digits, but not if looking backward from the start I see a decimal
point".

Presumably you can get the effect you want without this assertion stuff, via
e.g.

    pat = re.compile(r'(?:^|[^.])(\d+)\s+')

> pat = re.compile(r'(?![-A-Za-z0-9:._])([0-9]+)\s+')

In this case, every string that can match E2 also matches E1, so E2-E1 is
empty:  this can't match anything.

An interface that expressed this stuff as set operations on regular
languages instead would make it all obvious; try to *think* of it that way
anyway <wink>.

easy-to-do-if-you-don't-need-to-ly y'rs  - tim





From a.d.stribblehill at dur.ac.uk  Fri Apr  7 12:25:48 2000
From: a.d.stribblehill at dur.ac.uk (Andrew Stribblehill)
Date: 07 Apr 2000 17:25:48 +0100
Subject: Member function names
Message-ID: <87vh1t6foz.fsf@womble.dur.ac.uk>

For some things, such as XML parsers, based on xmllib, I have to do
stuff like this:

class foo(XMLParser):
    start_h1:
        print "start_h1",
    end_h1:
        print "end_h1",

(etc)

Is there any way I can do something like this?

class foo(XMLParser):
    generic:
        print __member_function_name__,
    start_h1=generic
    end_h1=generic

(etc)


Note: this is an example which I've managed to solve neatly a
different way but I still want to know if such a trick is possible,
for other less tractible problems!

Thanks,

Andrew Stribblehill


From javanet at dynacap.com  Mon Apr  3 13:41:15 2000
From: javanet at dynacap.com (Arnaldo Riquelme)
Date: Mon, 3 Apr 2000 09:41:15 -0800
Subject: Visual Interdev & Python
Message-ID: <sehlnm24eop139@corp.supernews.com>

I'm getting ready to build a web based app using ASP.
Since I'm totally new in this arena, I was kindly provided with ASP books, a
VBScript reference book
and Visual Interdev by management.
Most of the staff I've done in past had to deal with in the unix world,
using perl, bash, php, some c , etc.
I was a little disgusted by the VB syntax and I rather not spend anytime
learning it if could use another language,
preferably python, since I have been using it for a couple month and I'm
totally in love with it.

Is there any way to integrate Visual Interdev with python. Visual Interdev
seems to be a very popular tool
for ASPs and I think it could be a great way to attract Windows Developers
(VB) to use python.

Has anyone attempted to do this before? Can anyone  help with some
suggestions or point me to the right direction?

Best regards

Arnaldo Riquelme







From godzilla at netmeg.net  Tue Apr 11 20:48:47 2000
From: godzilla at netmeg.net (Les Schaffer)
Date: Wed, 12 Apr 2000 00:48:47 GMT
Subject: issubclass funny business
References: <1256652290-41498655@hypernet.com>
Message-ID: <m2n1n0cf5n.fsf@netmeg.net>

GMcM said:

> You are running FileScanner. Therefore, in FileScanner, class
> StandardArrayFileScanner is defined in __main__. If you trickle up
> to the top of __bases__ above, you'll find
> FileScanner.StandardArrayFileScanner, which is not the same object.

sigh... yep:

>       if not isinstance(std, StandardArrayFileScanner):
>            import cie
>            print std.__class__, cie.CIE.__bases__, cie.CIE_Standards.__bases__

gustav)~/Engineering/dspring/stoplite/matlab/Jue-Data/: python FileScanner.py


prints:

cie.CIE (<class cie.CIE_Standards at 810d1b8>,) 
(<class FileScanner.StandardArrayFileScanner at 81259f8>,) 

same problem with issubclass.....

whilst:
            print self.__class__.__bases__

(self is a subclass of ArrayFileScanner) gives:

(<class __main__.ArrayFileScanner at 8124f88>,)

les schaffer



From amitp at Xenon.Stanford.EDU  Sat Apr  8 20:52:19 2000
From: amitp at Xenon.Stanford.EDU (Amit Patel)
Date: 9 Apr 2000 00:52:19 GMT
Subject: A Mountain of Perl Books + Python Advocacy
References: <20000405053920.20579.qmail@web2102.mail.yahoo.com> <3FAG4.62951$3b6.243372@ozemail.com.au> <vnditxwfny7.fsf@camelot-new.ccs.neu.edu>
Message-ID: <8cok83$lpr$1@nntp.Stanford.EDU>

 Justin Sheehy  <dworkin at ccs.neu.edu> wrote:
| "Jason Stokes" <jstok at bluedog.apana.org.au> writes:
| 
| > The market for Perl is 10-100 times bigger than Python.
| 
| The "market", as defined how?
| 

The market for P* books is all the people who think P* isn't
obvious.  In that context, the market for Perl is 10-100 times
bigger than for Python.  ;-)

  -I-learned-Python-without-any-stinkin-book-ly y'rs, 
   Amit



From carlg at world.uucp  Tue Apr  4 12:53:59 2000
From: carlg at world.uucp (Carl E Gundel)
Date: Tue, 4 Apr 2000 16:53:59 GMT
Subject: Smalltalk on the small end (was: Advice requested: GUI project beginning)
References: <m12b4OW-000wcEC@swing.co.at> <8D79BC9E4960ED70.7B6609342D6EB723.7625773D14AD9140@lp.airnews.net> <ESaG4.9536$9g4.372176@news5.giganews.com>
Message-ID: <FsI4A0.755@world.uucp>

Christopher Browne (cbbrowne at news.hex.net) wrote:
: Centuries ago, Nostradamus foresaw a time when Cameron Laird would say:
: >Actually, there's another thread here asking to
: >be launched.  Python (and Tcl, too, in a somewhat
: >different fashion is (are) really quite extraor-
: >dinary for shell-level scripting and operation.
: >I find myself wondering more and more how people
: >get along without such (*particularly* on the
: >conventional desktops--Win* and MacOS).  I do
: >think Python is clearly superior to Smalltalk
: >for those one-page scripts that most people do
: >with /bin/sh; still, I'm willing to consider that
: >Smalltalk is more capable in this regard than I
: >yet realize, and, in any case, I can well imagine
: >Smalltalk proponents favoring it over /bin/sh.

: The "power" of Smalltalk comes not when you're writing one page shell
: scripts, but when you have a bunch of "objects that do useful things"
: that can all run in a single Smalltalk environment.

Smalltalk was the operating system on the machines it was originally
created to run on, and in that case scripting made lots of sense.  :-)

-Carl
-- 
------------------------------------------------------------------
 Carl Gundel  carlg at world.std.com  Shoptalk Systems  508-872-5315
 author of Liberty BASIC, a 1996 PC Magazine Awards Finalist!
 http://world.std.com/~carlg/basic.html
------------------------------------------------------------------



From exarkun at flashmail.com  Mon Apr 17 17:14:09 2000
From: exarkun at flashmail.com (Jp Calderone)
Date: Mon, 17 Apr 2000 21:14:09 +0000
Subject: Python CGI
References: <38F852D9.E2144F17@flashmail.com> <m3aeisxmjm.fsf@wilma.george.org>
Message-ID: <38FB7EA1.17F251A2@flashmail.com>

Harry George wrote:
> 
> I run python cgi's on RS6000 (AIX 4.3.2), using Netscape Enterprise
> Server.  No problem.  Can you post the code, or at least a skeleton
> which exhibits the behavior (leaving out sensitive content)?
> 

 *Arg*  I found the problem.. I slipped back into the C mentality for
a bit there.  Worst part is, I saw the problem every time I tried to
fix the program, it just never registered as a problem ;)  The problem:
I screwed up indentation underneath a for loop ;P

 Jp

> > [snip]

-- 
A sad spectacle.  If they be inhabited, what a scope for misery 
and folly.  If they be not inhabited, what a waste of space.
                -- Thomas Carlyle, looking at the stars
--
 9:09pm up 20 days, 2:16, 2 users, load average: 0.32, 0.10, 0.03


From mwh21 at cam.ac.uk  Wed Apr 12 07:00:05 2000
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 12 Apr 2000 12:00:05 +0100
Subject: Classes, Inheritance - Stupid lazy question
References: <8d1cfh$hme7@imsp212.netvigator.com> <874s9766r6.fsf@ifm.uni-kiel.de> <m3itxnis32.fsf@atrus.jesus.cam.ac.uk> <8d1ihl$q244@imsp212.netvigator.com>
Message-ID: <m3d7nvinyi.fsf@atrus.jesus.cam.ac.uk>

"Paul Gresham" <gresham at mediavisual.com> writes:

> I like this solution a lot, it makes a lot of sense. Thanks Michael.
> I guess somewhere in Python is a defined route back to the super
> class, on a better day I may just dig through the code.

For a variey of reasons, class heirachies in Python are usually pretty
shallow, so this isn't often too much of a problem.  I think the
median depth of class heriachies in code I write is probably 1...

Cheers,
M.

-- 
  On the other hand,  the following areas are subject to boycott
  in reaction to the rampant impurity of design or execution, as
  determined after a period of study, in no particular order: 
    ...                            http://www.naggum.no/profile.html


From cut_me_out at hotmail.com  Sun Apr 30 18:19:37 2000
From: cut_me_out at hotmail.com (Alex)
Date: 30 Apr 2000 18:19:37 -0400
Subject: How to make compile work for strings that contain ^M's?
Message-ID: <etdr9bn2pvq.fsf@w20-575-109.mit.edu>

Hi.  I just noticed that the command 

compile(file, 'test', 'exec')

can fail if 'file' contains '^M^J' at the end of each line.
Specifically, if file is WDhandler.py, in the webdebug package
(http://www.cyberclip.com/webdebug/webdebug-1.15.tar.gz), then I get the
error 

compile(file, 'test', 'exec')
Traceback (innermost last):
  File "<stdin>", line 1, in ?
  File "<string>", line 29
    """
       ^
SyntaxError: invalid syntax

... and sys.exc_info()[1].offset == 1043.

Is there any way that I can get compile to work properly in this case?
Neither python nor emacs seem to have a problem with '^M^J' line-endings
otherwise, but I use the compile(file, 'test', 'exec') command in a
script called by emacs to systematically test my scripts for syntax
errors before I save them.

It just occurred to me to strip out the ^M's in the script that does the
syntax checking.  I guess that'll do, but I'd still be interested in
whether there's a nicer way.

Alex.


From m.faassen at vet.uu.nl  Sun Apr  9 19:39:03 2000
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 9 Apr 2000 23:39:03 GMT
Subject: A Mountain of Perl Books + Python Advocacy
References: <000f01bf9eca$2eec8c60$3e2d153f@tim> <38ECF4E6.B85C62D5@sightreader.com>
Message-ID: <8cr4an$j40$2@newshost.accu.uu.nl>

Ken Seehof <kens at sightreader.com> wrote:
> Indeed.  I've noticed that while just about everyone who I've known who
> has tried python ends up liking it alot, there aren't many fanatics.  I define
> a fanatic as someone who is absolutely certain that their language is by far
> the best language for every possible application, and that other languages
> don't even deserve to exist.  Python programmers tend to acknowledge
> the positive qualities of other languages.

WRONGO! ALL UTHER LANGUZEZ SUCKZ!! PHYTHON IS THE BESTEST!!!!!1 FOR 
EVERYTHING AND I HATEZ M$ TUU!! NOBUDDY ECSPEXTS THE SPANNISH ENQIUSITOAN
HAHAHAHH! PHYTHON JOKE SPAM SPAM! <.75 WINK WINK NUDGE>!!!!!!!1

:)

Anyway, yeah, that's what I noticed in this newsgroup. Whenever someone asks
if Python is a good language Tim Peters tells them to check out Haskell, or
someone mentions Smalltalk or Scheme or Lisp or whatnot. It's rare that
people are being sent to C or C++ (unless it's for Python extensions) or
Perl though (but it does always get the 'good for fast text processing'
mention). It's nice to see so many comparisons between languages here.

> Perhaps this is partly because python borrows all the best stuff from many
> other languages rather than being built around some wild, revolutionary idea.

Well, but Perl isn't based on a wild, revolutionary idea either. It seems
to have borrowed from other languages far more, or at least it didn't cover
it up so nicely. Python is good in the way that it borrows stuff
in a fairly orthogonal way, so that it's easy to reason about its 
syntax and semantics.

Then again, I don't know if there are that many raging Perl advocates.

AN-EVERYBODY-SHUD-SING-OF-LIKE-THIS-LLY YR'S!!!!

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From infonuovo at email.com  Tue Apr  4 08:49:32 2000
From: infonuovo at email.com (Dennis E. Hamilton)
Date: Tue, 4 Apr 2000 05:49:32 -0700
Subject: Why should I switch to Python? - Infinity of Primes
In-Reply-To: <slrn8ej2gc.b3t.warlock@raven.myth>
Message-ID: <NDBBKEGCNONMNKGDINPFGEJDDFAA.infonuovo@email.com>

The standard approach is a proof by contradiction starting from the
assumption that there is a largest prime.

It's usually one of the earliest proofs in books on Number Theory.  Euclid
provided a proof and a version is given in section 4.2 of "Concrete
Mathematics: A Foundation for Computer Science" by Ronald L. Graham, Donald
E. Knuth, and Oren Patashnik.  Addison-Wesley (Reading, MA: 1989).   ISBN
0-201-14236-8.

It goes something like this (matching the one that was already posted for
you):

1.	Assume P is the largest prime.

2.	Let Q be the product of all of the primes up to P: 2*3*5* ... *P

3.	The number Q+1 is not divisible by any of the primes that divide Q (they
each leave a remainder of 1).  Therefore Q+1 is a prime or there is some
other prime, larger than P (i.e., not among 2, 3, 5, ..., P), that divides
Q+1.  So P is not the largest prime.  QED.

This all stands on the fundamental theorem of arithmetic: that every integer
number is a product of primes in exactly one way.  That is, there is no
other factorization of Q into primes than the one shown in its definition,
and Q+1 must have a factorization into primes, even if only Q+1 itself and
1.  That is Exercise 1.2.4-21 in "The Art of Computer Programming, v.1:
Fundamental Algorithms," ed.3 by Donald E. Knuth.  Addison-Wesley (Reading,
MA: 1997).  ISBN 0-201-89683-4.

Perhaps an interesting, indirect quality of Python is that it is supported
in a community of interest where one can discuss topics like the one you
just raised.

-- Dennis


-----Original Message-----
From: python-list-admin at python.org
[mailto:python-list-admin at python.org]On Behalf Of Jim Richardson
Sent: Monday, April 03, 2000 23:15
To: python-list at python.org
Subject: Re: Why should I switch to Python?


On 02 Apr 2000 08:16:57 -0400,
 Fran?ois Pinard, in the persona of <pinard at iro.umontreal.ca>,
 brought forth the following words...:

>"Tim Peters" <tim_one at email.msn.com> writes:
>
>> When in doubt, there's no substitute for exhaustive enumeration.
>
>Hmph!  Hope you do not doubt there is an infinite number of primes! :-)

Out of curiousity, how does one (mathematically) prove that?
(if this turns out to be one of those 12' blackboards full of greek
letters and funny squiggles, Ok, I cry uncle...)
--
Jim Richardson
	Anarchist, pagan and proud of it
WWW.eskimo.com/~warlock
	Linux, because life's too short for a buggy OS.

--
http://www.python.org/mailman/listinfo/python-list




From dgoodger at bigfoot.com  Sun Apr 23 00:28:45 2000
From: dgoodger at bigfoot.com (David Goodger)
Date: Sun, 23 Apr 2000 00:28:45 -0400
Subject: code not true?
References: <Pine.BSF.4.10.10004230013310.29461-100000@chi.pair.com>
Message-ID: <B527F43C.24AD%dgoodger@bigfoot.com>

on 2000-04-23 00:14, Michal Wallace (sabren) (sabren at manifestation.com)
wrote:
>> Floating point number on a binary computer are strange. They're not that
>> accurate. 0.07 * 100 != 7.0.
> 
> Why is this the case? I mean.. I understand why decimals are hard to
> show in binary... But why can't we use variably long strings, or
> something? Is it just a speed issue?

Try writing 1/3 in decimal. How many digits do you need to represent it
*exactly*? Same issue for 0.07 in binary floating point.

-- 
David Goodger    dgoodger at bigfoot.com    Open-source projects:
 - The Go Tools Project: http://gotools.sourceforge.net
 (more to come!)



From none at none.ca  Wed Apr 26 21:57:15 2000
From: none at none.ca (k)
Date: Wed, 26 Apr 2000 22:57:15 -0300
Subject: condense whitespace to single space?
Message-ID: <36NN4.31937$Xk2.119087@tor-nn1.netcom.ca>

How do I condense the sections of spaces in this string:

"abc              def         ghi"

into single spaces?

"abc def ghi"


kh





From ivanlan at callware.com  Wed Apr 19 14:11:22 2000
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Wed, 19 Apr 2000 12:11:22 -0600
Subject: Pythonware Website???? Gone on vacation with Starship? ;-))
References: <20000419154202.B480@owl.rhein-zeitung.de> <JCjL4.540$wYl.187815424@newsb.telia.net> <3d4s8yotiy.fsf@amarok.cnri.reston.va.us>
Message-ID: <38FDF6CA.C9A7AF28@callware.com>

Hi All--

"Andrew M. Kuchling" wrote:
> 
> "Fredrik Lundh" <effbot at telia.com> writes:
> >     www.pythonware.com is down
> >     the reason appears to be hardware problems on
> >     the ISP site. ...
> 
> First Zope.org (down because an Ethernet cable got pulled out), then
> Starship (disk crash), now pythonware.com.  Has someone released evil
> whitespace-eating nanobots that also attack computer hardware?
> 

This version eats dots (".").  It makes nameservers forget things, like
my domain, pauahtun.org;-(

<time-to-take-out-alien-abduction-insurance>-ly y'rs,
Ivan
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
http://www.pauahtun.org 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
Author:  Teach Yourself Python in 24 Hours



From gmcm at hypernet.com  Thu Apr 27 09:23:51 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Thu, 27 Apr 2000 09:23:51 -0400
Subject: Best database to use with Python
In-Reply-To: <8e93j5$qba$2@news.tht.net>
Message-ID: <1255280638-66426487@hypernet.com>

D'Arcy J.M. Cain wrote:
> Kevin McDermott <kevinmcd at nospam.linuxsolutions.demon.co.uk>
> wrote: > Depends on what you want to use it for...if it's to
> connect CGI programs > to, then I'd recommend MySQL.
> 
> Please don't take this as argumentative but I am curious why you
> think that being used with CGI programs per se makes one database
> better than the other.  It seems to me that there are a lot more
> issues to consider such as what kind of data, what kind of data
> relationships, how the data needs to be accessed, etc.  I use
> PostgreSQL for all my database work, CGI accessed or not, except
> for where the data sets are simple enough to use shelve.

Sometime ago linuxplanet did a comparison (sorry I can't 
check the link - my ISP is hosed right now):
http://www.linuxplanet.com/linuxplanet/tutorials/1224/1/

As I recall, Postgres was the most complete DB; MySQL the 
least complete. MySQL was very fast at some operations 
(probably related to lack of transaction support), but the most 
startlingly fast one was bulk load, which does you no good in 
a CGI situation. Again, from memory, MySQL was fast at very 
simple queries, but had a very limited SQL dialect and fell 
apart with fancier queries.

- Gordon



From moshez at math.huji.ac.il  Wed Apr 19 00:27:34 2000
From: moshez at math.huji.ac.il (Moshe Zadka)
Date: Wed, 19 Apr 2000 07:27:34 +0300 (IDT)
Subject: dir(..)   vs.   PyList_GetItem
In-Reply-To: <213CDEBFDDA7D3118081009027E40C56071155@SHEF-MS1>
Message-ID: <Pine.GSO.4.10.10004190726450.2850-100000@sundial>

On Tue, 18 Apr 2000, Mark Pope wrote:

> 	pModule = PyModule_New( "myScript" );
                  ^^^^^^^^^^^^
That function does not import a module, it creates a new empty module.

--
Moshe Zadka <mzadka at geocities.com>. 
http://www.oreilly.com/news/prescod_0300.html
http://www.linux.org.il -- we put the penguin in .com




From rupole at compaq.net  Wed Apr 19 23:43:51 2000
From: rupole at compaq.net (Roger Upole)
Date: Thu, 20 Apr 2000 03:43:51 GMT
Subject: Where do I send a contribution for the Starship recovery ?
Message-ID: <X1vL4.143650$ra.3235366@news.easynews.com>

I personally would consider it money well spent.
(unlike most of the other things I spend money on)
                    Roger Upole




From akuchlin at mems-exchange.org  Wed Apr 26 17:10:21 2000
From: akuchlin at mems-exchange.org (Andrew M. Kuchling)
Date: 26 Apr 2000 17:10:21 -0400
Subject: ungreedy regular expressions
References: <slrn8gdlaj.k0s.ckotso@theseas.softlab.ece.ntua.gr> <m3n1mhoyv4.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <3d3do8efgi.fsf@amarok.cnri.reston.va.us>

Michael Hudson <mwh21 at cam.ac.uk> writes:
> ckotso at cs.ntua.gr (Constantinos A. Kotsokalis) writes:
> > I recently added two lines of code to the python distribution, in order
> > to support ungreedy regular expressions (which are supported by pcre
>
> Aren't the *?,+? patterns "ungreedy"?

Yes, but the PCRE_UNGREEDY flag turns the syntax around, so * is
ungreedy and *? is greedy.  It's useful for applications where a naive
user might be entering regexes, since naive expectations usually don't
match the default greedy behaviour.

Unfortunately, the PCRE module is probably not going to be used in
Python 1.6, so applying the patch isn't likely to be very useful.
I'll think about it a bit.

-- 
A.M. Kuchling			http://starship.python.net/crew/amk/
When one has stopped loving somebody, one feels that he has become someone
else, even though he is still the same person.
  -- Sei Shonagon, _The Pillow Book_


From tim_one at email.msn.com  Wed Apr  5 02:42:59 2000
From: tim_one at email.msn.com (Tim Peters)
Date: Wed, 5 Apr 2000 02:42:59 -0400
Subject: A Mountain of Perl Books + Python Advocacy
In-Reply-To: <20000405053920.20579.qmail@web2102.mail.yahoo.com>
Message-ID: <000f01bf9eca$2eec8c60$3e2d153f@tim>

[lewst]
> ...
> What is it about Perl that makes it so much more popular and have such
> a huge grassroots swell?

1. It was in the right place at the right time twice (first as an eagerly
   embraced alternative to mounds of half-assed incompatible Unix tools
   pushed beyond their limits; and second, where it got its major growth,
   from legions of new CGI programmers (or "script kiddies", as they're
   derisively called over on c.l.p.m.)).

2. Larry and Perl's early adopters are unarguably better salespeople
   than Guido and Python's early adopters.

> ...
> Perl is very publicized and well-known while the better language is
> sitting here a dark corner unnoticed.

I don't grant that Python is unnoticed; to the contrary, its growth curve is
still exponential.  To the extent that it is true, though, the same has been
true of the Lisp family for decades -- this isn't a new phenomenon in
computer languages.  But unlike Lisp, Python hasn't splintered, so its
chances are still good.  Four years ago I wondered whether it would survive,
but no longer -- it's reached critical mass since then.

> Sure there will always be some religious fanatics that won't even give
> Python a try, but I think Python's popularity could be vastly improved
> with some serious advocacy work.  CNRI and/or PSA should seriously
> look into funding a Python "marketing" campaign of sorts.  I think the
> result would be allot of converts and more understanding and respect
> for Python.

Python *is* astonishing for the tiny number of evangelists it's attracted!
We sucked in Eric Raymond, for but for some reason he doesn't seem to
consider promoting Python to be his Primary Mission from God -- the best
hope for a PSA marketing campaign may be to hire a deprogrammer to
straighten out Tom Christiansen's brain while there's still something left
<wink>.

not-mentioning-that-perl-is-a-better-language-for-some-tasks-and-
    that-tells-you-how-effective-an-advocate-i-am-ly y'rs  - tim





From richard_chamberlain at ntlworld.com  Wed Apr 26 18:09:50 2000
From: richard_chamberlain at ntlworld.com (Richard Chamberlain)
Date: Wed, 26 Apr 2000 23:09:50 +0100
Subject: Saving pieces of the registry
References: <2B1262E83448D211AE4B00A0C9D61B03011152B9@MSGEURO1>
Message-ID: <MLJN4.4555$0d2.97679@news2-win.server.ntlworld.com>

Hi Pieter

There are several things in the win32api which may do what you want:

win32api.RegSaveKey and win32api.RegRestoreKey

Richard

Pieter Claerhout <PClaerhout at CREO.BE> wrote in message
news:2B1262E83448D211AE4B00A0C9D61B03011152B9 at MSGEURO1...
> Hi,
>
> does anyone know is there is a command in Python to save a part from the
registry
> to a .reg file, like you can do with regedit.exe? I would like for example
like to save
> all keys under HKEY_LOCAL_MACHINE\Software to a file. Anyone know if this
is
> possible or not?
>
> Kind regards,
>
>
> Pieter
>
>




From tim_one at email.msn.com  Tue Apr  4 00:28:48 2000
From: tim_one at email.msn.com (Tim Peters)
Date: Tue, 4 Apr 2000 00:28:48 -0400
Subject: [Python-Dev] New Features in Python 1.6
In-Reply-To: <20000403165621.A9955@cnri.reston.va.us>
Message-ID: <000001bf9dee$45f318c0$162d153f@tim>

[Greg Ward, fesses up]
> Hasn't anyone noticed that the largest amount of text in the joke
> feature list was devoted to the Distutils?  I thought *that* would
> give it away "fer shure".  You people are *so* gullible!  ;-)

Me too!  My first suspect was me, but for the life of me, me couldn't
remember writing that.  You were only second on me list (it had to be one of
us, as nobody else could have described legitimate Python features as if
they had been implemented in Perl <0.9 wink>).

> And for my next trick... *poof*!

Nice try.  You're not only not invisible, I've posted your credit card info
to a hacker list.

crushing-guido's-enemies-cuz-he's-too-much-of-a-wuss-ly y'rs  - tim





From moshez at math.huji.ac.il  Wed Apr 26 12:49:14 2000
From: moshez at math.huji.ac.il (Moshe Zadka)
Date: Wed, 26 Apr 2000 19:49:14 +0300 (IDT)
Subject: Metaclasses?
In-Reply-To: <m3k8hlou1f.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <Pine.GSO.4.10.10004261946010.7232-100000@sundial>

On 26 Apr 2000, Michael Hudson wrote:

<wonderfully horrible implementation of a selfless class using nothing
but bytecodehacks and metaclasses, proving that Donald Beaudry was right>

I'm shocked! I can see Python in tears!
Ouch!

who-said-you-can't-obfuscate-python-ly y'rs, Z.

--
Moshe Zadka <mzadka at geocities.com>. 
http://www.oreilly.com/news/prescod_0300.html
http://www.linux.org.il -- we put the penguin in .com




From owen at astroNOJNK.washington.edu.invalid  Fri Apr 21 19:00:28 2000
From: owen at astroNOJNK.washington.edu.invalid (Russell E. Owen)
Date: Fri, 21 Apr 2000 16:00:28 -0700
Subject: Tkinter -- does Canvas.find_closest work?
Message-ID: <8dqmib$1ar0$1@nntp6.u.washington.edu>

I am using Mac Python 1.5.2c1 (current) with Tkinter (based on Tk 8.0). 
I have a canvas called "cnv" with the following objects on it:
- 10 stars: tiny filled ovals tagged 'star'
- 9 grid markings: large unfilled ovals that are not tagged at all

I am trying to use the canvas method "find_closest" to find the closest 
star, ignoring the grid markings. Supposedly all I have to do is specify 
the start keyword with the value of a tag and it will find the closest 
object with that tag. Here's a code snippen and some of the output:

----- code snippet -----
nearestStar = self.cnv.find_closest (x=event.x, y=event.y, halo=0, 
start='star')
print "found star ", nearestStar[0]
objList = self.cnv.find_withtag('star')
print "star list = ", objList

----- portion of resulting output ------
found star  23
star list =  (23, 24, 25, 26, 27, 28, 29, 30, 31, 32)
found star  20
star list =  (23, 24, 25, 26, 27, 28, 29, 30, 31, 32)

find_withtag is acting as if I never specified the tag at all. By moving 
the mouse pointer around I can get find_withtag to find objects 14 - 32. 
23-32 are clearly the stars. 14-22 are presumably the grid circles. The 
output above shows it finding one star and one grid circle.

I also tried supplying a numeric values (object IDs) to the "start" 
keyword. In this mode find_closest is supposed to find objects with IDs 
smaller than the value of "start". However, again it doesn't seem to 
have any effect. find_closest still will find any object on the canvas 
as I move the mouse pointer around.

Can anybody see an obvious mistake? Has anybody gotten the start keyword 
to work in Canvas method "find_closest"?

-- Russell


From gmcm at hypernet.com  Thu Apr 20 07:51:21 2000
From: gmcm at hypernet.com (Gordon McMillan)
Date: Thu, 20 Apr 2000 07:51:21 -0400
Subject: __del__ problem - would adopting Garbage Collection fix this?
In-Reply-To: <000701bfaa9b$e729eac0$d62d153f@tim>
References: <l03102805b52461abc14d@[193.78.237.146]>
Message-ID: <1255891014-29713226@hypernet.com>

I think Neil has the correct approach: gc.garbage is a 
development aid. If anything shows up in there, fix the 
classes. Silently ignoring __del__ methods is a great way to 
leak resources, which is worse than leaking memory.

[Tim]
> even-five-year-olds-understand-they're-supposed-to-clean-up-their-
>     own-mess<wink>-ly y'rs  - tim

<snort>-five-year-old-bots-maybe-ly y'rs

- Gordon



From g_will at cyberus.ca  Tue Apr  4 08:50:34 2000
From: g_will at cyberus.ca (Gordon Williams)
Date: Tue, 04 Apr 2000 12:50:34 GMT
Subject: What happened to Python.announce?
Message-ID: <01bf9e34$08050d20$4842c3d1@Cyberus>

Hello all,

Is it just my service provider, or is python.announce really dead?

Regards,

Gordon Williams


From emile at fenx.com  Thu Apr 13 09:54:05 2000
From: emile at fenx.com (Emile van Sebille)
Date: Thu, 13 Apr 2000 06:54:05 -0700
Subject: Authenticate with smtp
References: <018701bfa53b$f831f3e0$858647c1@DEVELOPMENT>
Message-ID: <006101bfa54f$bc6ca760$01ffffc0@worldnet.att.net>

Ulf,

Are you trying to use that second mail server as a relay host?
What error are you getting?  What happens when you try using
the relay addressing structure? (<@host:user at host>)  Here I get
'unable to relay' errors when I try to bounce off the NT, but
not when I bounce off the linux system.  As far as I can tell
from the RFC, authentication on the receiver-SMTP process
appears to be limited to ensuring delivery to the recipient.

Emile van Sebille
emile at fenx.com
-------------------


----- Original Message -----
From: Ulf Engstr?m <ulf.engstrom at b2b-link.com>
To: <python-list at python.org>
Sent: Thursday, April 13, 2000 4:32 AM
Subject: Authenticate with smtp


> How can I log into an smtp server that requires authentication and
send a
> mail with smtplib, or is there any other package that helps? (Yes, I'm
> authorized)
> Regards
> Ulf
>





From effbot at telia.com  Mon Apr 10 11:51:16 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Mon, 10 Apr 2000 15:51:16 GMT
Subject: check EOF
References: <8cs6g4$rje$1@porthos.nl.uu.net>
Message-ID: <ULmI4.364$rc9.188792320@newsb.telia.net>

Soundwave <maybe at lucky.com> wrote:
> how can I check EOF in python ???

read/readline returns an empty string when you reach
the end of the file.

for more info, read the fine manual:

http://www.python.org/doc/current/lib/typesother.html
=> file objects

</F>




From bill at libc.org  Thu Apr 27 20:20:35 2000
From: bill at libc.org (Bill Anderson)
Date: Thu, 27 Apr 2000 18:20:35 -0600
Subject: The Simple Economics of Open Source
References: <B527E7A5.51AA%raffael@mediaone.net> <MPG.136cc01af4ab703b989709@news.earthlink.net> <raffael-4899DC.23562823042000@news.ne.mediaone.net> <vndya62ykbi.fsf@camelot-new.ccs.neu.edu> <raffael-21A8E7.10481427042000@news.ne.mediaone.net>
Message-ID: <3908D953.29F8078@libc.org>

Raffael Cavallaro wrote:
> 
> In article <vndya62ykbi.fsf at camelot-new.ccs.neu.edu>, Justin Sheehy
> <dworkin at ccs.neu.edu> wrote:
> 
> >You mean like BSD/OS, from BSDI?
> >
> >It happened 9 years ago, and they've been selling it ever since.
> 
> Yeah, and BSDI is just tearing up the software world. BSDI sells
> *support* not software, because the software they sell is essentially
> interchangable with the free BSD variants. BSDI have *themselves*
> recognized this fact by merging with Walnut Creek. In future, they'll be
> merging the code of FreeBSD and BSD/OS.
> 
> In other words, the history of BSDI is a perfect example why you can't
> succeed at a closed source business when your product is a software
> commodity. BSDI tried, and have, essentially, failed. In a market where

Surely you don't expect us to believe that the fact that the product was
a 'commodity' (for some definition of commodity) was the cause of their
'failure'. Of course, that also presumes that the merger is indiciative
of a failure. Something else not shown.

I would expect that you would understand that correlation is not
causality, and that merger is not failure.

-- 
In flying I have learned that carelessness and overconfidence are 
usually far more dangerous than deliberately accepted risks. 
          -- Wilbur Wright in a letter to his father, September 1900


From tjg at avalongroup.net  Tue Apr  4 20:25:01 2000
From: tjg at avalongroup.net (Timothy Grant)
Date: Tue, 04 Apr 2000 17:25:01 -0700
Subject: Two questions
References: <38EA5FE8.853E9B75@callware.com> <38EA6508.90BF630C@roguewave.com> <38EA6E87.40EDB542@callware.com> <38EA6422.61054B11@exceptionalminds.com> <38EA7C18.8F3150D1@roguewave.com>
Message-ID: <38EA87DD.3F24BEBF@exceptionalminds.com>

Bjorn Pettersen wrote:

> In the control panel there should be a Odbc 32 control (should be
> installed with ms access).  It has a simple point-and-click interface
> where you have to choose the ms access driver, give the filename for the
> database, and choose a DSN (Data Source Name?).  The DSN is what you use
> to connect with.

OK, So I've created a system level DSN named TestDB and pointed it at an
MDB file my server.

When I do the following in the interactive window...

import ODBC.Windows

db = ODBC.Windows.Connect('TestDB')

...I get the following error...

NotSupportedError: ('IM001', 0, '[Microsoft][ODBC Driver Manager] Driver
does not support this function', 4226)

What am I doing wrong?

-- 
Stand Fast,
    tjg.

Timothy Grant                         tjg at exceptionalminds.com
Chief Technology Officer              www.exceptionalminds.com
Red Hat Certified Engineer                      (503) 246-3630
Avalon Technology Group, Inc.              fax  (503) 246-3124
>>>>>>>>>>>>Linux...Because rebooting isn't normal<<<<<<<<<<<<



From sabren at manifestation.com  Tue Apr 11 13:46:02 2000
From: sabren at manifestation.com (Michal Wallace (sabren))
Date: Tue, 11 Apr 2000 13:46:02 -0400 (EDT)
Subject: simple httplib and urllib timeout question
In-Reply-To: <CgJI4.14892$eh.1040983@news.bc.tac.net>
Message-ID: <Pine.BSF.4.10.10004111337520.29635-100000@chi.pair.com>

On Tue, 11 Apr 2000 darrell at rubina.funkomite.com wrote:

> Just a quick question, im writing a small webbot for a python
> learning experience and it comes across some sites that will make it
> hang for a while, if the site is down or extremely slow etc, is
> there a way to set a faster timeout for httplib . Or is there a
> better way to handle it ?

Yes. The better way is to use asyncore whenever httplib/urllib uses a
normal socket. Unfortunately, this requires breaking urllib into
smaller pieces, and nobody seems to have done it yet. I'm actually
working on the exact same problem (and posting regular updates
at http://www.linkwatcher.com/ ) 

I sent an email to Jeremy Hylton, who's working on urllib2. He's
swamped, but willing to accept a patch that would add this feature
to urllib2.. I think I'll wind up taking a crack at it tonight.

There's also a partial solution using a different method if you look
at...  http://www.linkwatcher.com/src/blogbot/timeout.py

there's an example script called blogbot.py there, too.. it uses
timeout.py and another hack with signal.alarm(), but neither really
works all that well..

hope that helps.

meanwhile... HEY OLEG! -> have you done anything w/ this?

Cheers,

- Michal
-------------------------------------------------------------------------
http://www.manifestation.com/         http://www.linkwatcher.com/metalog/
-------------------------------------------------------------------------




From spenrose at my-deja.com  Thu Apr 20 14:26:24 2000
From: spenrose at my-deja.com (spenrose at my-deja.com)
Date: Thu, 20 Apr 2000 18:26:24 GMT
Subject: PyApache memory leak?
References: <8dimg4$kii$1@nnrp1.deja.com>
Message-ID: <8dni3s$1o1$1@nnrp1.deja.com>

[[I tried sending the following to PyApache's author; the email bounced
with "too many hops." I tried subscribing to py-apache-devel, which has
not seen a post in four months, and never got a reply. I post here in
the hopes that someone will get it in the right hands.]]

We just went live with a large, Python-driven web site (RedHat 6.1,
Apache 1.3.9 Red-Hat-Secure/3.1 mod_ssl/2.4.10 OpenSSL/0.9.4). We don't
have our log analysis software up and running yet, but we are getting
thousands of cgi page views a day (maybe low tens of thousands, not
hundreds of thousands). Each page view in turn involves a suite of
Python modules, both standard and written by us. All of ours are plain
Python (no C or Java).

We rolled in PyApache; the guy who did it is gone but the tar ball is
labelled 4.19 so I assume that's the version. Our beta testing was very
light; orders of magnitude lighter than the hits the site has gotten
since going live.

About two days and six hours after going live, the server froze.
Behavior was consistent with memory filling up--it was fine, then
suddenly it was slow, then within seconds you couldn't even ping it. We
restarted the machine, apologized to our clients, and tried to figure
out what was happening. There was some garbage written to one of the
Apache custom logs (we're running three virtual domains that share a
suite of Python cgis), but we couldn't tell what the problem was. A
little more than two days later, it crashed again with the same
behavior: everything fine--hey, my term window is sluggish--frozen.
Still no clue. Two days and roughly six hours after that, it froze a
third time. Then we started looking more carefully at the logs. They
read:

[many normal requests]
Normal request
@@@@line of garbage@@@@
Normal request
Normal request
Normal request
Normal request
[blank line]
[restart]

Our theory is this: since only Apache writes to the Apache log, one
Apache process was generating garbage seconds before the crash. A few
other processes had time to make an entry, then the machine froze. Ergo,
it was probably the garbage-writing process that took the machine out.
Apache 1.3.9 is used by millions of computers and RedHat Secure server
is used by tens of thousands (?), so it probably wasn't a bug in them.
The only other thing we've done to the standard distribution is add
PyApache.

We commented out PyApache in httpd.conf. Two days and six hours
later...the server did not freeze. We've been up for 11 days and
counting and that was the only change we made. I freely acknowledge that
this is not definitive proof of a showstopper in PyApache, but I am not
up to reading thousands of lines of C to pursue other theories.

I apologize for not giving more details, but we didn't know what else to
look at and as this is a production server I can't exactly turn PyApache
back on so I can gather better bug tracking info. I can be reached at
sam at ddmweb dot com if anyone has questions.

Cheers,
Sam


Sent via Deja.com http://www.deja.com/
Before you buy.


From montague at maravu.cs.dartmouth.edu  Mon Apr 10 11:32:13 2000
From: montague at maravu.cs.dartmouth.edu (Mark H.H. Montague)
Date: 10 Apr 2000 15:32:13 GMT
Subject: comparing strings and ints
Message-ID: <8css5t$tti$1@merrimack.Dartmouth.EDU>

Does anyone know why it is not a type error to compare a string and an
int with '>' ?  For example, I just spent over an hour debugging code
that essentially did this:

>>> 45 > '0'
0

Of course, the '0' was a variable read in from a file, which I forgot
to convert to a string.  Bad old perl habits.  But I'm having a hard
time imagining a situation where someone would be intentionally and
meaningfully comparing a string and an int.

Any ideas?

-mm


From effbot at telia.com  Tue Apr  4 13:40:54 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Tue, 04 Apr 2000 17:40:54 GMT
Subject: New Features in Python 1.6
References: <200004011740.MAA04675@eric.cnri.reston.va.us> <_pyF4.25562$E85.640966@news1.rdc1.md.home.com> <slrn8edfmk.185.wtanksle@dolphin.openprojects.net> <ltvh21ywfs.fsf@asfast.com> <oq66txol0w.fsf@titan.progiciels-bpi.ca>
Message-ID: <GOpG4.5050$74.81782@newsc.telia.net>

Fran?ois Pinard wrote:
> Beware.  It might be patented already...  Let's see.  Does it work this
way?
>
> * If the file is empty, you're done.
> * If the file holds a single bit, you're done.
> * Split the whole file into pairs of sequences, the first part of the pair
>   being an eager, but possibly empty sequence of consecutive zero-bits,
>   the second part being an eager, but possibly empty sequence of
consecutive
>   one-bits.  Notice that there is only one way to do this split.
> * Replace each sequence above by the count of bits it contains, expressed
>   directly as a binary number for the first count of the pair, and as the
>   one-complements of the binary number for the second count of the pair.
>   Notice that given these counts, it is trivial to rebuild the contents
>   of the previous step.
> * Iterate.

why not just use two different compression algorithms repeatedly,
until you've squeezed the file down to a suitable size?  it takes a
little work, but you can design them so that one of them can always
strip away at least one bit from the input stream.

by reversing this process in the right order, you can easily reconstruct
the original data stream.

a much faster version, known as "blom compression" after it's inventor,
uses 256 algorithms...

</F>




From rcc at nospamthanks_jennic.com  Thu Apr 13 11:36:03 2000
From: rcc at nospamthanks_jennic.com (Robert Cragie)
Date: Thu, 13 Apr 2000 15:36:03 GMT
Subject: List mapping?
Message-ID: <newscache$wroysf$2s2$1@jenpc07.jennic.co.uk>

> Try this:
> mylist = ['/a','1','/b','2','/c','3']
> map(lambda x: reduce(lambda a, b: (a,b), mylist[x:x+2]), range(0,
> len(mylist), 2))

Sorry, forgot to assign...

mylist = ['/a','1','/b','2','/c','3']
tuplist = map(lambda x: reduce(lambda a, b: (a,b), mylist[x:x+2]), range(0,
len(mylist), 2))

Robert Cragie




From thomas at xs4all.net  Sat Apr 15 13:37:34 2000
From: thomas at xs4all.net (Thomas Wouters)
Date: Sat, 15 Apr 2000 19:37:34 +0200
Subject: BSDi and PIL
In-Reply-To: <sH0K4.13$Ny3.1761@newreader.ukcore.bt.net>; from dreid@jetnet.co.uk on Sat, Apr 15, 2000 at 04:22:48PM +0000
References: <sH0K4.13$Ny3.1761@newreader.ukcore.bt.net>
Message-ID: <20000415193734.D15664@xs4all.nl>

On Sat, Apr 15, 2000 at 04:22:48PM +0000, David Reid wrote:

> Has anyone had any joy in getting Python 1.5.2 and PIL 1.0 to build together
> on BSDi 4.0.1?

Is your problem python or PIL ? Compiling stock 1.5.2 on BSDI 4.x can be a
challenge: If you enable shared libraries (which is a good thing) you have
to convince configure that you are running on an ELF system (BSDI 4.x is
ELF, 3.x and lower isn't) or change the resulting Makefiles to reflect this.
You know when you hit this problem when you get errors about undefined
symbols in shared modules, and those undefined symbols are python-internal
symbols. (Like PyErr* or PyNone)

If you enable threads, you're in for more treats: threading on BSDI 4.x is
screwed in many subtle ways, resulting in for instance a pause() call that
never bothers to return, even if you do get a signal.

The linkflags problem is relatively easy to fix: After configure, edit the
toplevel Makefile, and Modules/Makefile.pre (not the Makefile itself, for it
will be overwritten after you change Modules/Setup)

Look for this block:

# Symbols used for using shared libraries
SO=             .so
LDSHARED=       ld
CCSHARED=       
LINKFORSHARED=  
DESTSHARED=     $(BINLIBDEST)/lib-dynload

And change the middle three entries so it reads like this:

# Symbols used for using shared libraries
SO=             .so
LDSHARED=       gcc -shared
CCSHARED=       -fpic
LINKFORSHARED=  -Xlinker -export-dynamic
DESTSHARED=     $(BINLIBDEST)/lib-dynload

(If you ever lose this mail and forget what it was, steal those flags from
any system using ELF, like linux ;) This problem has been fixed, by the way,
in the CVS tree, so 1.6 wont require this patching.

The threading business is harder... The bugs are subtle, but unfortunately
not restricted to python code that uses threads: when you enable threads,
pause() will always hang your process indefinately. (I started python under
gdb once, to see what was happening... In thread-enabled code, the libc
pause() is a special with-threads version, and on receiving the signal,
there is some internal rumbling underneath pause(), but it never delivers
the signal to the calling function.)

So, my advice would be 'do not enable threads'. If you *have* to have
threads, or you would really like to have them, upgrade to BSDI 4.1, which
does handle threads OK (at least, the test that i threw at it.) Besides,
most BSDI patches are freely available, last time I looked, and fairly easy
to install.

As for PIL, well, I haven't bothered with installing it on our BSDI servers,
just on my linux workstations, but if you're still having trouble, I could
give it a go.



-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From drew.mcdowell at msfc.nasa.gov  Fri Apr  7 10:52:46 2000
From: drew.mcdowell at msfc.nasa.gov (Andrew McDowell)
Date: 7 Apr 2000 14:52:46 GMT
Subject: Was Re: need wrapper for ssh/scp
References: <8civ14$rm7$1@hammer.msfc.nasa.gov> <20000407130942.E13830@xs4all.nl>
Message-ID: <8cksnu$in2$1@hammer.msfc.nasa.gov>

Thomas Wouters <thomas at xs4all.net> wrote
> Help is easiest if you also explain what the problem is ;)

I'm trying to implement a script where I can enter my passphrase once, and
fire up identical processes on multiple hosts.
It's not quite as easy as I thought...
Oh, and I'm not allowed to use ssh-agent either. :(

> I already see one
> problem though: you read the password with 'string.strip(raw_input())',
> which means the terminating newline gets stripped off the password, but
you
> dont supply it with os.write(),

*doh* how'd that get in there?  <wink>

Ok, I added the newline and sshd is showing that the script is logging
in...but no commands are being executed.
Any further advice?

>which means ssh thinks there is more
> password to come, and patiently sits waiting for it, untill you brutally
> kill it ;)

Hehe..I was getting a bit flustrated :)

> In your real app you probably want to use select(), by the way, before
> attempting to read something. Doing a read() when there's no data to be
read
> makes a blocking read, which can lead to deadlocks. See telnetlib for some
> examples on how to use it.

Thanks for the pointer...I'll look into it!




From algaba at my-deja.com  Fri Apr 14 12:24:32 2000
From: algaba at my-deja.com (Manuel Gutierrez Algaba)
Date: Fri, 14 Apr 2000 16:24:32 GMT
Subject: Crosswords for Linux Backtracking recursive crossword generator
References: <8d59hk$6to$1@nnrp1.deja.com> <m3g0sphfay.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <8d7gnd$l4a$1@nnrp1.deja.com>

In article <m3g0sphfay.fsf at atrus.jesus.cam.ac.uk>,
  Michael Hudson <mwh21 at cam.ac.uk> wrote:
> Manuel Gutierrez Algaba <algaba at my-deja.com> writes:
>
> > Comments/improvements are welcome!
>
> I'm just looking at it, but in playing around I got this output:
>
> - - - - - - - - - - - - - - -
> - - - - - - - - - - c - - c -
> - - - - - - - - - s o l f a -
> - - - - - - - - - - n - - r -
> - - - - - - - - - - f - - a -
> - - - - - - - - - - u - - n -
> - - - - - - - - - - s - - t -
> - - - z a l a m e r i a - o -
> - - - - - - - - - - o - m ? -
> - - - - c o h e r e n c i a -
> - - - - a - l - - - - - m - -
> - - - - p - i n s u l t o - -
> - - - m a n o - - - - - - - -
> - - - - - - - - - - - - - - -
> - - - - - - - - - - - - - - -
>
> That's not right is it?  The `m' and `?' being next to each other.

It's not a "bug". It's that way, of course, the algorithm may be
modified for avoiding such things, but then it's much slower.


--
-------------
Manolo


Sent via Deja.com http://www.deja.com/
Before you buy.


From breiter at usf.Uni-Osnabrueck.DE  Fri Apr  7 07:19:55 2000
From: breiter at usf.Uni-Osnabrueck.DE (Bernhard Reiter)
Date: 7 Apr 2000 11:19:55 GMT
Subject: Teaching the art of programming, in python
References: <20000405053920.20579.qmail@web2102.mail.yahoo.com>
    <8cfe42$o02$1@nnrp1.deja.com> <dpJG4.3744$HG1.106895@nnrp1.uunet.ca>
    <38EB68A3.B1B5809F@callware.com>
    <8cfqa7$4ml$3@newsserver.rrzn.uni-hannover.de>
    <38EB73B6.6C0913A7@callware.com>
    <8chsij$k5r$1@newsserver.rrzn.uni-hannover.de>
    <slrn8er3fd.fcd.r2d2@shaka.acc.umu.se>
Message-ID: <8ckg8r$a7j$1@newsserver.rrzn.uni-hannover.de>

In article <slrn8er3fd.fcd.r2d2 at shaka.acc.umu.se>,
	r2d2 at shaka.acc.umu.se (Niklas Frykholm) writes:
>>>> And all in 24 hours?
>>>> This is as misleading as the dummy attribute.
>>
>>> I don't think so.  You can go at your own speed.  The book is broken up
>>> into 24 chapters, each taking more or less than an hour to read.  
>>
>>Well I guess that a good speed would be 14 days, then. :)
>>With two hour lections plus time to exercise.
> 
> As an interesting side note, a publisher here in Sweden has
> recently started a 10-minute series, with titles such as
> "Teach your self Internet in 10 minutes". Seems there is
> inflation (or perhaps deflation) in these things.

Just discovered that Peter Novig wrote some nice chapters about it
	http://www.norvig.com/21-days.html
Excerpt:
| Why is everyone in such a rush?
| 
| Walk into any bookstore, and you'll see how to Teach Yourself Java
| in 7 Days alongside endless variations offering to teach Visual
| Basic, Windows, the Internet, and so on in a few days or hours.
| I did a search 
[..]
| The conclusion is that either people are in a big rush to learn
| about computers, or that computers are somehow fabulously easier to
| learn than anything else. There are no books on how to learn
| Beethoven, or Quantum Physics, or even Dog Grooming in a few days or
| hours. 

> And it _is_ a silly practice. But then again, corporate
> silliness doesn't surprise me that much anymore.

:)
Bernhard

-- 
Professional Service around Free Software                (intevation.net)  
The FreeGIS Project				            (freegis.org)
Association for a Free Informational Infrastructure            (ffii.org)


From pecora at anvil.nrl.navy.mil  Sun Apr  9 08:38:55 2000
From: pecora at anvil.nrl.navy.mil (Louis M. Pecora)
Date: Sun, 09 Apr 2000 12:38:55 GMT
Subject: I need MacPython screenshots!
References: <20000408162815.50658.qmail@hotmail.com>
Message-ID: <090420000838551912%pecora@anvil.nrl.navy.mil>

In article <20000408162815.50658.qmail at hotmail.com>, Matheus Rodrigues
<matheusrodrigues at hotmail.com> wrote:

> Hello everybody
> 
> I need some screenshots of a Macintosh running
> MacPython. If possible I would like to receive one
> screenshot of the MacPython Installation too. The
> faster you can provide some material to me the more I
> thank you.
> 
> Please send the files (zipped) to my email as we dont
> want to create huge traffic in the list. It's not
> necessary to crop the images.
> 
> btw, where is the Macintosh discussion list ?

Check on the python web site:  http://www.python.org

Here are some archives:  http://www.python.org/pipermail/pythonmac-sig/

Post a message by emailing:  pythonmac-sig at python.org

People are very quick to answer and very helpful.  I speak from
experience.


From deepak_marda at my-deja.com  Wed Apr 26 16:27:26 2000
From: deepak_marda at my-deja.com (deepak_marda at my-deja.com)
Date: Wed, 26 Apr 2000 20:27:26 GMT
Subject: Need help with COMObject for Zope
Message-ID: <8e7jeo$s8c$1@nnrp1.deja.com>

Hi,
I am trying to use the COMObject for Zope. I have
the latest versions of Python and Zope and from
what I see working, COMObject is installed in the
right place.

I created a COMObject with id = word and Dispatch
= "Word.Application". WHen I try to connect this
COM object, it raises the following exception
Error Type: com_error
Error Value: (-2147221008, 'CoInitialize has not
been called.', None, None)

Can you please help me how I can make it work?
I have Word (from Office 2000) installed and the
following works from a vanilla python code: -

>>>import win32com.client
>>>o = win32com.client.Dispatch
('Word.Application')
>>>o.Visible = 1
>>> ... and so on

Please help,
Deepak


Sent via Deja.com http://www.deja.com/
Before you buy.


From corg at copernic.com  Tue Apr 18 08:45:40 2000
From: corg at copernic.com (Gaetan Corneau)
Date: Tue, 18 Apr 2000 08:45:40 -0400
Subject: Microsoft ActiveX Scripting
Message-ID: <C5A7720BDEB7D3118A5200E018C200A027DB6C@EXCHANGE>

> [Actually, Ive never personally used the scripting control, 
> but I believe it works!]

It does. Last time I have used it, it worked very well. I only had a couple
of little problems, but maybe they were caused by my application.

______________________________________________________
   Gaetan Corneau
   Software Developer 
   Copernic.com 
   http://www.copernic.com
   E-mail: corg at copernic.com                         
______________________________________________________
"Tu l'as trop ?cras?, C?sar, ce Port Salut"
> 



From tismer at tismer.com  Mon Apr 17 10:31:31 2000
From: tismer at tismer.com (Christian Tismer)
Date: Mon, 17 Apr 2000 16:31:31 +0200
Subject: [ANNOUNCE] Garbage collection for Python
References: <20000407221750.A29279@acs.ucalgary.ca> <m3r9cagx3c.fsf@atrus.jesus.cam.ac.uk> <l03102809b51d41895a80@[193.78.237.135]> <38F9F01B.57273052@tismer.com> <m33domez7n.fsf@atrus.jesus.cam.ac.uk> <38FA05FA.B0526CAB@tismer.com> <m3hfd1es10.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <38FB2043.76F75B98@tismer.com>


Michael Hudson wrote:
> 
> Christian Tismer <tismer at tismer.com> writes:
> 
> > > I cam across a great quote the other day; I think it was attributed to
> > > Alan Perlis, and it was:
> > >
> > >     Simplicity follows complexity, not the other way round.
> >
> > Thank you a lot, it was exactly the quote I was trying
> > to look up, but I didn't find it again. It is very true.
> 
> Actually from
> 
>     http://www.cs.yale.edu/~perlis-alan/quotes.html
> 
> we find:
> 
> 31. Simplicity does not precede complexity, but follows it.
> 
> so I got the wording slightly wrong.  The intent's clear though.
> 
> now-wait-while-I-add-most-of-those-to-my-sig-file-ly y'rs
> Michael

Thanks a lot, this is a goldmine!

A-programming-language-is-low-level-when-its-programs-
   require-attention-to-the-irrelevant-ly y'rs - chris

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaunstr. 26                  :    *Starship* http://starship.python.net
14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     where do you want to jump today?   http://www.stackless.com



From aa8vb at yahoo.com  Thu Apr 20 08:05:34 2000
From: aa8vb at yahoo.com (Randall Hopper)
Date: Thu, 20 Apr 2000 08:05:34 -0400
Subject: print a pixel
In-Reply-To: <8dmcvb$r6n$1@wanadoo.fr>
References: <8dmcvb$r6n$1@wanadoo.fr>
Message-ID: <20000420080534.A623423@vislab.epa.gov>

BLAISE BONNEVILLE:
 |I am a beginner in Python and I would like to know how to print a pixel in
 |Python with Windows 98, given its coordinates.

I think you'll need to give more details.  For example "print a pixel".
Pixel in particular: a pixel in an image file?  or off of the frame buffer?
if the former, what type of image file?  where does this pixel come from?

And "print": print it how?  as R,G,B coordinatees?  Or do you mean print it
on a hardcopy printer?

More details.

-- 
Randall Hopper
aa8vb at yahoo.com



From ns56645 at swt.edu  Sun Apr 23 00:40:10 2000
From: ns56645 at swt.edu (ns56645)
Date: Sat, 22 Apr 2000 23:40:10 -0500
Subject: Combobox in Tkinter
References: <38F8F133.54FCC64B@swt.edu> <8db0b1$b3o$1@nnrp1.deja.com> <20000421005845.36765@nms.otc.telstra.com.au>
Message-ID: <39027EAA.E19AABE9@swt.edu>

Yep,

This worked fine for me.

Thanks
Shah,
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20000422/c4910ab5/attachment.html>

From amitp at Xenon.Stanford.EDU  Sat Apr  8 21:07:47 2000
From: amitp at Xenon.Stanford.EDU (Amit Patel)
Date: 9 Apr 2000 01:07:47 GMT
Subject: Python on Google?
References: <8F0ED22EBjoshtinamecom@209.208.0.7>
Message-ID: <8col53$m3o$1@nntp.Stanford.EDU>

 Josh Tompkins <josht at iname.com> wrote:
| Buenos Dias, everyone.
| 
| This is perhaps an off-topic question, but having just found google, I'm 
| curious...
| 
| Does anyone know to what extent Python is used on www.google.com?  I 
| noticed on thier "job offerings" (or whatever) page that they list python 
| experience as "a plus".
| 


Python has been used at Google since the very beginning, and it
continues to be used, although some of the more performance critical
stuff is in C++.  New code _is_ being written in Python (in fact, I do
all my work in Python) and it's being used exclusively to deal with
some large sets of data.  Email me if you want more details.

       - Amit,
         Google, Inc.






From smalleys at gte.net  Sun Apr 30 17:48:36 2000
From: smalleys at gte.net (sue)
Date: Sun, 30 Apr 2000 21:48:36 GMT
Subject: problem building Python in MachTen Unix
References: <01HW.B531E90F0013E3A1037B464C@news.pond.com>
Message-ID: <390CAA4D.6C554E23@gte.net>

I suspect you are not seeing all of the message from the
configure dyld step.

Does ../Include/mach-o/dyld.h or ../mach-o/dyld.h exist?  If
not, the include path is wrong.

I have found it's futile to argue with configure.  Edit the
makefile directly and save a copy elsewhere.  Get the binary
built, and then backtrack.  You may need to check config.h.

Sue.


From claird at starbase.neosoft.com  Thu Apr 20 15:37:37 2000
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 20 Apr 2000 14:37:37 -0500
Subject: Good style for multi-valued returns
Message-ID: <889A6EB6AE399393.619B0539BFCAAE0F.5C6D110C3DA908F7@lp.airnews.net>

Examples of multi-valued returns:  (head, tail); (quotient,
remainder); (left, right); (prologue, body, epilogue);
(data, code); (CAR, CDR); (count, size); (handle, status);
(decomposition; remainder); ...

Let me guess:  folks generally pass tuples for small and
determinate-in-number multi-values, and dictionaries other-
wise, 'cept some do it all with tuples.  Accurate?
-- 

Cameron Laird <claird at NeoSoft.com>
Business:  http://www.Phaseit.net
Personal:  http://starbase.neosoft.com/~claird/home.html


From richard_chamberlain at ntlworld.com  Tue Apr 18 15:47:15 2000
From: richard_chamberlain at ntlworld.com (Richard Chamberlain)
Date: Tue, 18 Apr 2000 20:47:15 +0100
Subject: Using python on the web
References: <MPG.1365adcfa8470eef9896aa@news> <8di2bk$l1s$1@newshost.accu.uu.nl> <MPG.136631472f892a119896ac@news>
Message-ID: <_W2L4.986$GK5.20898@news2-win.server.ntlworld.com>

I'm curious why you would be put off using cgi?

having used python and cgi, plus java servlets/jsp and asp at various
different times I cannot say I see your point.

In many ways my preferred method would be python because I can get something
working in such a short space of time, and there's no significant speed
difference between, for instance, java and python.

cgi has to be about the most common way of construction on the fly web
pages, and I don't think it's going away just yet.

Richard

Charley Horse <someone at somewhere.com> wrote in message
news:MPG.136631472f892a119896ac at news...
> Thanks to all of those pointers. Finding out that CGI plays an important
> role in the python world was a shock to me as I only started working
> with web stuff in the last couple of years. I've just never touched a
> CGI script, shockingly I suppose <g>. ASP/CF and PHP are what I've
> worked with and CGI was just the "old world" technology for
> interactivity that I read about but never had to work with.
>
> So coming at this as a newbie, and for years hearing about python's
> elegance and oo-ness, it just seemed a strange pairing, python and CGI.
> Puzzling in fact! I see why it makes sense, but it's just not what I
> expected.
>
> I'll fool with CGi and perhaps it won't be enough of a turn off to
> affect my interest.
>
> In article <8di2bk$l1s$1 at newshost.accu.uu.nl>, m.faassen at vet.uu.nl
> says...
> > Charley Horse <someone at somewhere.com> wrote:
> > > I've just started fooling with python. I've looked around the python
> > > site and a number of the other sites referenced there and am a bit
> > > puzzled. Umm... how does one go about using python for web
applications,
> > > especially database driven apps? Almost everything deals with python
> > > just as a language, as C++ would be.
> >
> > That's because Python is a language. Python isn't a web-specific
language,
> > after all.
> >
> > > Not much about the web specifically.
> > > I was a bit shocked to find the only direct refs on about
> > > this at python.org are entwined with CGI (unless I missed it).
> >
> > Why is CGI so shocking? It's used quite a bit, after all? You can write
> > a database driven web application using Python with CGI. CGI from Python
> > isn't hard either.
> >
> > > I have no exposure to CGI, but rather to php, CF, and ASP, so python
has
> > > gone from looking very attractive to looking... well, like a puzzle.
> >
> > I'm not sure why your exposure to PHP, ColdFusion or ASP instead of CGI
> > would turn Python into a puzzle.
> >
> > There _is_ ASP support for Python, enabling you to use Python as a
> > ASP language. Others will have to explain to you how and where to look
> > for information, as I've never used it.
> >
> > > I know about Zope, but am interested in how one uses python on the web
just
> > > by itself.
> >
> > Zope would be the equivalent of ColdFusion for Python, as it's a web
> > application server.
> >
> > Besides using Python with CGI or ASP, you can use Zope's ZPublisher
> > (formerly Bobo) independently to publish arbitrary Python objects to the
> > web. I've never used this, and the documentation may be scarce, but it
> > looks quite neat. You may want to look that up.
> >
> > Database integration in Python isn't bad either, look at the database
> > topic guide:
> >
> > http://www.python.org/topics/database/
> >
> > And look at the web topic guide if you haven't seen it:
> >
> > http://www.python.org/topics/web/
> >
> > The vaults of Parnassus are also a good resource, here is for
> > instance a link to its web category:




From shayman at uniserve.com  Tue Apr 11 00:25:26 2000
From: shayman at uniserve.com (Matthew)
Date: Mon, 10 Apr 2000 21:25:26 -0700
Subject: Filter
Message-ID: <955427129.619305@neptune.uniserve.ca>

I need to write a filter that inserts tabs between Log values .  So that
lines 1 to 3 come out looking something like lines 4 to 6.
I can then open it in Excel as a dat file with the values in different
cells, where i can then use formulas and macros.
The first 9 lines are constants in the header ,but there could be serveral
headers with data  in the whole file .

I'm totaly at square one here.
I have to figure out how to open the file for modification - as a string ?
write a for loop for each line beginning with L11 ?
And a function to insert the white space's between the values.


I imagine this to be an ideal task for a small script , but untill i get
there any advice or help appreciated.





VER1.5A
"H0:1,45"
"H1:9085,*.*   "
"H2:71G,NORTHVIEW DLS, ,MITSUI"
"H3:VANCOUVER,CHILLIWACK"
"H4:00100S,100,SB4,01, , , "
"H5:000330, AH00-93-,SPRUCE, ,N"
"H6:45,126825"
H7:220000
L11SSSP 503438G 1627                                      # 1
L11SSSP 563338G 1826
L11SSSP 573036G 1924
L11    SS    SP     55    36    40    G     18    28      # 4
L11    SS    SP     51    40    46    G     17    31
L11    SS    SP     57    30    37    H     19    24
etcXXXXXXXXXXXXXXXXXXXXXXX
"H9:1,45"





From gresham at mediavisual.com  Wed Apr 12 06:18:24 2000
From: gresham at mediavisual.com (Paul Gresham)
Date: Wed, 12 Apr 2000 18:18:24 +0800
Subject: Classes, Inheritance - Stupid lazy question
References: <8d1cfh$hme7@imsp212.netvigator.com> <874s9766r6.fsf@ifm.uni-kiel.de> <m3itxnis32.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <8d1ihl$q244@imsp212.netvigator.com>

"Michael Hudson" <mwh21 at cam.ac.uk> wrote in message
news:m3itxnis32.fsf at atrus.jesus.cam.ac.uk...
> Janko Hauser <jhauser at ifm.uni-kiel.de> writes:
>
> > I think this will do it
> >
> > class lazy:
> >     def __init__(self, name):
> >         self.name = name
> >
> >     def fired(self):
> >         print 'You are fired %s' % self.name
> >
> > class paul(lazy):
> >     def __init__(self, name):
> >         self.name = name  # <---- ?????
> >         self.april_pay = 0
> >
> >     def fired(self):
> >         print "Here's a box, collect the things from your desk"
> >         self.__class__.__bases__[0].fired(self)
>
> Nononono; think about what happens if you do this:
>
> class bob(paul):
>    pass
>
> bob("adam").fired()
>
> What *I* think you want is:
>
> class paul(lazy):
>     def __init__(self, name):
>         lazy.__init__(self,name)
>         self.april_pay = 0
>
>     def fired(self):
>         print "Here's a box, collect the things from your desk"
>         lazy.fired(self)
>
> After all, if you're going to inherit from a class, you must have
> access to it, right?
>
> another option is:
>
> class paul(lazy):
>     super = lazy
>     def __init__(self, name):
>         self.super.__init__(self,name)
>         self.april_pay = 0
>
>     def fired(self):
>         print "Here's a box, collect the things from your desk"
>         self.super.fired(self)
>

I like this solution a lot, it makes a lot of sense. Thanks Michael.
I guess somewhere in Python is a defined route back to the super class, on a
better day I may just dig through the code.

Thanks again
>
> Though this has problems with repeated & multiple inheritance.
>
> Cheers,
> M.
>
> --
>   "declare"?  my bogometer indicates that you're really programming
>   in some other language and trying  to force Common Lisp into your
>   mindset.  this won't work.           -- Erik Naggum, comp.lang.lisp




From godzilla at netmeg.net  Sat Apr 22 09:54:20 2000
From: godzilla at netmeg.net (Les Schaffer)
Date: Sat, 22 Apr 2000 13:54:20 GMT
Subject: using PyArg_ParseTuple with a list
References: <3900B1B0.1CFD748A@sec.noaa.gov>
Message-ID: <m2og72me04.fsf@netmeg.net>

> I wish to pass a Python list [] to a C extension module, but I am
> unsure of the proper format to specify in PyArg_ParseTuple

here is a bit of code i wrote to pass a list of strings in to a C
extension:

...
    char * tok;         /* delimiter tokens for strtok */
    int cols;           /* number of cols to parse, from the left */

    int numLines;       /* how many lines we passed for parsing */
    char * line;        /* pointer to the line as a string */
    char * token;       /* token parsed by strtok */

    PyObject * listObj; /* the list of strings */
    PyObject * strObj;  /* one string in the list */
    
    /* the O! parses for a Python object (listObj) checked
       to be of type PyList_Type */
    if (! PyArg_ParseTuple( args, "O!is", &PyList_Type, &listObj, 
			   &cols, &tok )) return NULL;
    
    /* get the number of lines passed to us */
    numLines = PyList_Size(listObj);

    /* should raise an error here. */
    if (numLines < 0)	return NULL; /* Not a list */

...

    /* iterate over items of the list, grabbing strings, and parsing
       for numbers */
    for (i=0; i<numLines; i++){

	/* grab the string object from the next element of the list */
	strObj = PyList_GetItem(listObj, i); /* Can't fail */

	/* make it a string */
	line = PyString_AsString( strObj );

	/* now do the parsing */


les schaffer


From effbot at telia.com  Tue Apr  4 18:33:10 2000
From: effbot at telia.com (Fredrik Lundh)
Date: Tue, 04 Apr 2000 22:33:10 GMT
Subject: complete novice question
References: <38EA5995.F7B9815D@newbridge.com>
Message-ID: <G4uG4.1351$n68.191769088@newsb.telia.net>

Paul Madsen <pmadsen at newbridge.com> wrote:
> Downloaded Python-1.5.2.
>
> I've run 'configure' and it seems to have worked. Now, according to the
> Python package's README, I should be able to type 'make' for the rest of
> the install. If I don't have a 'make' file, did the configure portion
> not work?

did you try typing "make" ?

("make" is a unix command that should be available on any
decent unix machine.  afaik, it doesn't work on win95, but
neither do configure, so I suppose you're not posting from
the machine you're building on ;-)

(and yes, the configure command should create a file called
"Makefile", which tells "make" how to build python for you).

</F>




From herzog at online.de  Sat Apr  1 17:25:43 2000
From: herzog at online.de (Bernhard Herzog)
Date: 02 Apr 2000 00:25:43 +0200
Subject: Function override?
References: <000101bf9b9b$312f8fe0$2f2d153f@tim> <38E64275.CF8F97DC@math.okstate.edu>
Message-ID: <m37leh4fx4.fsf@greebo.nodomain.de>

"David C. Ullrich" <ullrich at math.okstate.edu> writes:

>     Or I could if I hadn't just read the bit about how in 1.6 strings
> have a split method, and if s1 and s2 are strings then
> 
> s1.split(s2)
> 
> either splits s1 using s2 as the delimiter or the other way around,
> depending on which occurs more frequently in the other. I mean
> surely it will happen that we want to split a bunch of strings
> using 'delimiter' as a delimiter, and then it just happens that
> the routine that does that gets passed 'lim' as an argument and
> does the opposite of what we want?

Oh, that's easy to work around. I'm sure we'll all get used to the
following new idiom:

(text + delim).split(delim)[:-1]

Surely, this is not harder to learn than:

while 1:
	line = file.readline()
	if not line:
		break
	# do something

or

while 1:
	# loop body

	if some_condition:
		break


-- 
Bernhard Herzog   | Sketch, a drawing program for Unix
herzog at online.de  | http://sketch.sourceforge.net/


From chris.lada at westgroup.com  Tue Apr 11 10:24:08 2000
From: chris.lada at westgroup.com (Chris Lada)
Date: Tue, 11 Apr 2000 10:24:08 -0400
Subject: Processing BCD and Signed Integer Fields with Python
Message-ID: <38f334ea@wwwproxy3.westgroup.com>

We have been receiving record/field-based information from a data provider.
Some of the fields are BCD (Binary Coded Decimal), and some are signed
integer. The data originates from a mainframe/Cobol(?) system. The record
layouts have definitions like "PIC S99999" (signed integer), and "PIC S99999
Comp-3" (BCD). Has anyone had to process this sort of data with Python ?
Can Python handle this sort of data type ?  Anyone have suggestions on how
to approach this problem ?

Thanks,
Chris




From scherbi at bam.com  Mon Apr  3 07:46:04 2000
From: scherbi at bam.com (Bill Scherer)
Date: Mon, 03 Apr 2000 07:46:04 -0400
Subject: Methods for password storage
References: <38E87E13.A10A1168@arches.uga.edu>
Message-ID: <38E8847C.50A33083@bam.com>

Brad -

Don't store the password.  Use crypt (unix only) or maybe MD5, and store
the encrypted password.  This way no one can view the user's real pasword.
When authenticating the user, you must encrypt the password provided (in
the same manner, obviously), and compare it to the stored version.  If they
match, the user knows the password.  crypt and md5 are one way ciphers,
that is there is no way, other than brute force, to get the original text
out of the ciphertext.

The crypt algorithm is what many unix systems use to encrypt users
passwords.  The encrypted password can be viewed by most users in the file
/etc/password.  You may choose to make your store of encrypted passwords
unaccesible to the average user.  That would make brute force attackes more
difficult ;-).

If I've gotten any of this wrong, someone will point that out soon, but
this is basically the method I'm using.

-Bill  Scherer

Brad Chapman wrote:

> Hello!
>     I'm working on a program which stores passwords (for database
> access) that a user enters in a GUI. The program is targetted for unix
> machines. I'm not very experienced with security issues like this and am
> trying to store them in as secure a manner as possible.
>     Does anyone have any advice/examples for how to securely store
> sensitive information like this using python? Are there other security
> issues I should worry about when dealing with passwords? Thank in
> advance much for any advice!
>
> Brad
>
> --
> http://www.python.org/mailman/listinfo/python-list

--
William K. Scherer
Sr. Member of Applications Staff
Bell Atlantic Mobile
http://ampeg.corp.bam.com






From herzog at online.de  Tue Apr 25 13:29:42 2000
From: herzog at online.de (Bernhard Herzog)
Date: 25 Apr 2000 19:29:42 +0200
Subject: Simple list.append() question
References: <meh9-5ED8CD.11192424042000@news.cit.cornell.edu> <ZHZM4.360$qg7.2949@news.uswest.net> <39046FA2.603EAFE5@roguewave.com> <PsaN4.174166$Dv1.2059732@news1.rdc1.bc.home.com> <8e4akc$p7o$1@newshost.accu.uu.nl>
Message-ID: <m366t6ozqx.fsf@greebo.nodomain.de>

m.faassen at vet.uu.nl (Martijn Faassen) writes:

> Generally it's a good idea to avoid * on sequences, unless that sequence is
> immutable and contains immutable things; i.e. strings.

It seems to me that the important criterion is whether the elements are
immutable. The mutability of the sequence itself doesn't really matter
here.

> Is there any useful way to use * on lists that I missed?

Something like [None] * 1000 is very useful IMO.

> Usually we want the copy semantics here, not the reference semantics.

If the elements are mutable. But then the question is how deeply do you
copy them?

> Perhaps it's a good idea to completely forbid * on lists in p3k?
> Tuples too for all I care. Then again I may be missing important uses,
> so enlighten me.

AFAICT, I use * with lists most often to create a new list of n
identical immutable elements, most often None or '' or somesuch and I
find that really useful.

What seems to cause the most problems is that newbies try to use it to
create empty multi-dimensional arrays and that fails. I think they try
it because they see it more as a declaration similar to C's
'int[10][20]' than an expression. It would be a good idea IMO to find a
good syntax for this case and ISTM that list comprehensions might be
useful here:

  multi_list = [[] for i in range(10)]

or whatever the precise syntax was. Hmm, the unused variable in there is
a bit ugly though.


-- 
Bernhard Herzog   | Sketch, a drawing program for Unix
herzog at online.de  | http://sketch.sourceforge.net/


From embed at geocities.com  Wed Apr  5 11:56:51 2000
From: embed at geocities.com (Warren Postma)
Date: Wed, 5 Apr 2000 10:56:51 -0500
Subject: time-module questions
Message-ID: <CxIG4.3719$HG1.106792@nnrp1.uunet.ca>

I have a few questions on python time code (import time).

First, it appears that time.localtime returns fractions of a second in it's
values, but that time.mktime does not add this millisecond component back
in, nor does it appear to be part of the time-tuple returned by
time.localtime, and time.gmtime:

for i in range(0,10):
 x = time.time()
 tup=time.localtime(x)
 x2 = time.mktime(tup)
 print x2,' ',x
 time.sleep(0.1)

954949853.0   954949853.238
954949853.0   954949853.379
954949853.0   954949853.509
954949853.0   954949853.639
954949853.0   954949853.769
954949853.0   954949853.899
954949854.0   954949854.03
954949854.0   954949854.16
954949854.0   954949854.29
954949854.0   954949854.42

Secondly, mktime accepts localtimes only, how to I take y/m/d/h/m/s in GMT
instead of the current timezone, and convert back to a time.time() type
value (float).

What to do? Is Python 1.6 any more 'robust' in it's handling of times/dates?

Warren





From thomas at xs4all.net  Wed Apr 19 04:27:00 2000
From: thomas at xs4all.net (Thomas Wouters)
Date: Wed, 19 Apr 2000 10:27:00 +0200
Subject: module __call__
In-Reply-To: <286L4.1328$B43.118787@news.pacbell.net>; from dirck@pacbell.net on Tue, Apr 18, 2000 at 04:23:43PM -0700
References: <LX2L4.1275$B43.102174@news.pacbell.net> <7N4L4.497$wYl.191816704@newsb.telia.net> <286L4.1328$B43.118787@news.pacbell.net>
Message-ID: <20000419102700.M15664@xs4all.nl>

On Tue, Apr 18, 2000 at 04:23:43PM -0700, Dirck Blaskey wrote:
> Fredrik Lundh <effbot at telia.com> wrote in message
> news:7N4L4.497$wYl.191816704 at newsb.telia.net...
> > Dirck Blaskey <dirck at pacbell.net> wrote:
> > > ...Gee-Wouldn't-It-Be-Nice-If...
> > >
> > > modules could def __call__():
> > >
> > > so you could call a module, i.e.
> > >
> > >     import StringIO
> > >     f = StringIO()
> >
> > and
> >
> > class myStringIO(StringIO):
> >     ...
> >
> > would explode in your face.  lousy idea.
> 
> Hmm...
> 
>     class myStringIO(StringIO):
> 
> already doesn't work;
> you have to
> 
>     class myStringIO(StringIO.StringIO):
> 
> as currently implemented.

Yes, but 'f = StringIO()' doesn't work either, in that case. Fredrik's
point, though a bit blunt, is that it would be confusing and thus unwise to
functionally separate

f = StringIO()

and

class myStringIO(StringIO)

If you really dislike typing StringIO twice, do what (nearly) everyone else
does:

from StringIO import StringIO

And if you dont subclass StringIO (haven't seen that happen a lot) you can
even go nifty and do:

try:
    from cStringIO import StringIO
except ImportError:
    from StringIO import StringIO

giving you the power of cStringIO when it's available.

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From neelk at brick.cswv.com  Wed Apr  5 19:01:12 2000
From: neelk at brick.cswv.com (Neel Krishnaswami)
Date: 5 Apr 2000 23:01:12 GMT
Subject: Tuples -- who needs 'em
References: <00e101bf9d80$a8384d20$74eb0b18@stcla1.sfba.home.com> <va3G4.1110$n68.190454272@newsb.telia.net> <037f01bf9e55$6d328760$74eb0b18@stcla1.sfba.home.com>  <38EAC102.7D5DF504@cosc.canterbury.ac.nz>
Message-ID: <slrn8enjg9.456.neelk@brick.cswv.com>

Greg Ewing <greg at cosc.canterbury.ac.nz> wrote:
> 
> In Haskell, a statically-typed language which also has
> both tuples and lists, the two types play distinctly
> different roles, and both are needed. A tuple is a 
> fixed-size structure whose elements may be of different 
> types, whereas a list is a variable-size structure whose 
> elements must be of the *same* type.
> 
> This is the way I tend to use them in Python as well --
> lists for homogeneous collections, tuples for "records".
 
Heh. I don't mind having lists and tuples; it's classes and tuples
that are the "same thing", in that both of these are record types. I
think the only reason to get rid of tuples would be if be nicer to
have a simple destructuring/pattern-matching syntax for classes and
instances than currently.

It's hard to overstate how convenient being able to do pattern
matching on tuples is -- it's one of the things that makes Python so
expressive. 

> If static typing is introduced into Python, a more rigid
> distinction may have to be made, in which case we may be
> glad that we already have the two kinds of structure.
 
Indeed. And then I think people would complain about having tuples and
classes. :)


Neel


From thomas at cintra.no  Thu Apr 27 16:14:18 2000
From: thomas at cintra.no (thomas at cintra.no)
Date: Thu, 27 Apr 2000 20:14:18 GMT
Subject: Example search-engine/indexer
Message-ID: <39089992.10423684@news.c2i.net>

Hello,

I was wondering if anybody has an example on how to index lots of
files or/and documents for fast searching. I`m doing sort of a
Yahoo/Altavista-thing and need an example that`s fast and doesn`t rely
on an external database-system. Berkley DB is ok, and all stuff that
comes with a standard Python distro.

For now I got a function that extracts words from files/folders and
words from documents. For now I`ve done something like this :

1. recursive process all files and folders in a given folder and its
subfolders. Documents found will be opened and words extracted. Words
ends up in a dictionary, the word acting as key, pointing to a list of
ids, ex. (1, 23), where 1 is the folder id and 23 is the file-id.
Words are handled seperatly but in a similar manner.

2. What I end up with are a dictionary like this :

>>print words['python']
>>[ (1,23), (21.45), (432, 21) ]
>>print words['linux']
>>[ (12,32), (30.90), (4, 42) ]

Using shelve and db_hash to store the keys and lists of ids works fine
for small amounts of data, but when the list of ids, actually tuples
like ( 32, 43 ), get huge, into several thousands, the thing takes up
lots of space and gets slower.

Has anybody done anything similar? Dr. Dobbs Journal #295, Jan 1999
has an example using Perl. Could anybody translate the article, the
code is short and sweet, like most Perl-code, but not that easy to
understand, like most Perl-code.

Any hints on how to do something like this, a different method or
whatever would be appreciated, example code most of all.

Tudelidu!!

Thanks,
Thomas