Mailman 3 April 2001 - Python-Dev

Re: [Python-checkins] CVS: python/dist/src/Lib/plat-unixware7 FCNTL.py,NONE,1.1 IN.py,NONE,1.1 SOCKET.py,NONE,1.1 STROPTS.py,NONE,1.1 TERMIOS.py,NONE,1.1 regen,NONE,1.1
by Michael Hudson 11 Apr '01

11 Apr '01

Guido van Rossum <gvanrossum(a)users.sourceforge.net> writes: > Update of /cvsroot/python/python/dist/src/Lib/plat-unixware7 > In directory usw-pr-cvs1:/tmp/cvs-serv11682 > > Added Files: > FCNTL.py IN.py SOCKET.py STROPTS.py TERMIOS.py regen Ehh... I didn't think we did TERMIOS.py or SOCKET.py any more? Cheers, M.

1 0

RE: [Python-checkins] CVS: python/dist/src/Lib/test test_math.py,1.10,1.11 test_fork1.py,1.8,1.9 test_fcntl.py,1.16,1.17
by Tim Peters 11 Apr '01

11 Apr '01

> Update of /cvsroot/python/python/dist/src/Lib/test > In directory usw-pr-cvs1:/tmp/cvs-serv12386 > > Modified Files: > test_math.py test_fork1.py test_fcntl.py > Log Message: > Unixware 7 support by Billy G. Allie (SF patch 413011) > ... > *************** > *** 36,40 **** > testit('atan2(-1, 0)', math.atan2(-1, 0), -math.pi/2) > testit('atan2(-1, 1)', math.atan2(-1, 1), -math.pi/4) > ! testit('atan2(0, 1)', math.atan2(0, 1), 0) > testit('atan2(1, 1)', math.atan2(1, 1), math.pi/4) > testit('atan2(1, 0)', math.atan2(1, 0), math.pi/2) > --- 37,44 ---- > testit('atan2(-1, 0)', math.atan2(-1, 0), -math.pi/2) > testit('atan2(-1, 1)', math.atan2(-1, 1), -math.pi/4) > ! if sys.platform in ['unixware7']: > ! testit('atan2(0, 1)', math.atan2(0, 1), math.pi) > ! else: > ! testit('atan2(0, 1)', math.atan2(0, 1), 0) > testit('atan2(1, 1)', math.atan2(1, 1), math.pi/4) > testit('atan2(1, 0)', math.atan2(1, 0), math.pi/2) atan2(0, 1) should be 0 on *all* platforms. It's too bad if the original test fails on unixware7, but if it does it means their libm's atan2() is buggy. C99 spells this out in excruciating detail. C89 isn't as clear, but is clear enough: The atan2(y, x) function computes the principal value of the arc tangent of y/x, using the signs of both arguments to determine the quadrant of the return value. A domain error may occur if both arguments are 0. IOW, atan2 returns the angle with x-axis made by a unit vector from the origin to the point (x, y). The point (1, 0) lies on the x axis, pointing to the right, so is at an angle of 0. The only question is whether it should be +0 or -0, and while C99 spells that out too, Python's test doesn't distinguish those cases so we don't have to answer that here. So, if nobody leaps to defend unixware7, I'm going to revert that part of the patch.

1 0

SocketServer and UserDict patches
by Ka-Ping Yee 11 Apr '01

11 Apr '01

Hello all, I'd like to call your attention to two small patches that i would like to check in for the 2.1 RC. They're small, but they correct breakages that i think are worth fixing. 1. UserDict.get(), .update(), and .setdefault() https://sourceforge.net/tracker/?func=detail&aid=413171&group_id=5470&atid=… These three methods are currently implemented by calling the underlying object's .get(), .update(), or .setdefault() method. This is bad because a UserDict that overrides __getitem__ now will have an inconsistent or failing get() method. A glaring example of this is cgi.SvFormContentDict. For such an object x, x['spam'] returns a single item but x.get('spam') returns a list of one item! Instead, these three methods should be implemented in terms of the object's own __getitem__, __setitem__, and has_key methods. This patch makes this change. 2. SocketServer.StreamRequestHandler https://sourceforge.net/tracker/?func=detail&aid=415111&group_id=5470&atid=… The setup() method here duplicates the socket twice (once to make a read-only file, once to make a write-only file). That yields three descriptors, but finish() closes only two. This causes my browser to hang indefinitely waiting for the socket to close when SimpleHTTPServer is used to deliver a small page. This patch adds self.connection.close() to setup() so that there are just two descriptors to worry about. -- ?!ng "All models are wrong; some models are useful." -- George Box

4 7

SourceForge search-by-ID implemented
by Guido van Rossum 10 Apr '01

10 Apr '01

I received the email below from the friendly folks at SourceForge -- you can now search bugs and patches by number. Cool! --Guido van Rossum (home page: http://www.python.org/~guido/) ------- Forwarded Message Date: Tue, 10 Apr 2001 19:45:55 +0300 From: Paul Sokolovsky <pfalcon(a)sourceforge.net> To: Guido van Rossum <guido(a)python.org> Subject: Fwd: [alexandria - Feature Requests] ANN: search artifacts (bugs etc.) by # Hello Guido, I would like to notify that another of your feature requests for SourceForge has been done. Sorry that it took so much time - search is one of the most straining parts of the site, and there was a marathory for any changes to it... This is a forwarded message From: noreply(a)sourceforge.net <noreply(a)sourceforge.net> To: noreply(a)sourceforge.net <noreply(a)sourceforge.net> Subject: [alexandria - Feature Requests] ANN: search artifacts (bugs etc.) by # ===8<==============Original message text=============== Read and respond to this message at: http://sourceforge.net/forum/message.php?msg_id=137990 By: pfalcon There were many request to allow to display bugs, support requests, etc. by their number, having typed it into search box. Finally, it's here. By typing digit sequence, optionally preceeded by #, you'll get that item (in addition to items where that string present literally, of course). Enjoy! ______________________________________________________________________ You are receiving this email because you elected to monitor this forum. To stop monitoring this forum, login to SourceForge and visit: http://sourceforge.net/forum/monitor.php?forum_id=4 ===8<===========End of original message text=========== - -- Paul Sokolovsky, http://sourceforge.net/users/pfalcon SourceForge developer http://sourceforge.net ------- End of Forwarded Message

2 1

distutils.sys: None in sys.modules
by Ka-Ping Yee 10 Apr '01

10 Apr '01

When i import distutils.util, i get: >>> sys.modules.keys() ['os', 'distutils.sys', 'distutils.os', 'exceptions', 'posix', 'distutils.spawn', 're', 'sre_constants', 'distutils.errors', 'string', 'signal', 'sre', 'posixpath', 'sre_parse', '_sre', 'os.path', 'distutils.string', 'readline', 'distutils.util', 'distutils.re', '__main__', 'distutils.dep_util', 'types', 'sys', '__builtin__', 'site', 'UserDict', 'distutils', 'sre_compile', 'copy_reg', 'stat', 'distutils.distutils'] What are 'distutils.sys', 'distutils.os', 'distutils.string', 'distutils.re', 'distutils.distutils' doing in there? (The sys.modules dictionary maps all these keys to None.) -- ?!ng "All models are wrong; some models are useful." -- George Box

4 5

INSTALL_PROGRAM
by Thomas Wouters 10 Apr '01

10 Apr '01

I just noticed that INSTALL_PROGRAM is defined as just INSTALL (either the system "install" or the install-sh script, with possibly -c as argument) without a -m argument (to set the mode.) INSTALL_DATA does have a -m argument, to set the mode for all 'data' files to 644 explicitly. INSTALL_PROGRAM gets called not just for the python executable, but also for all files in Lib/ that have their executable bit set. Because INSTALL_PROGRAM does not set the mode, the files (potentially, depending on the install program/script in question) are subject to the umask and/or the original file mode. I've already screwed up my Python installation on a couple of BSDI boxes twice, before I realized what the problem was :) What about we set the mode for executables to 755 explicitly ? Distutils seems to do the right thing, right now, but I'm pretty sure it was screwed up before. What logic does distutils use to figure these things out ? (There is also INSTALL_SCRIPT, but that doesn't seem to be used anywhere.) -- Thomas Wouters <thomas(a)xs4all.net> Hi! I'm a .signature virus! copy me into your .signature file to help me spread!

1 0

Release schedule heads up
by Guido van Rossum 10 Apr '01

10 Apr '01

We're planning the release candidate for 2.1 early this Friday (the 13th :-). We need to have all changes ready early Thursday. Then we plan to release the final version Monday the 16th, complete with a press release and all! The final release should be identical to the release candidate if all goes well. Between now and Thursday, only the most important bugs should be fixed. For anything else, please hold off till after 2.1final is released! --Guido van Rossum (home page: http://www.python.org/~guido/)

1 0

Free register of online company's profile
by info＠webb2e.com 09 Apr '01

09 Apr '01

How much are you paying to advertise your business to the world? Expose Your service to the world with bold register of online business profile. Sign up today! Introducing WebB2E.com -- your direct link to global information; source of business, products, education/research, social/culture, entertainment and travel... Additionally you can BUY, SELL or PROMOTE your products and services At www.webb2e.com <http://webb2e.com/promo1.asp> you'll get: --Message center (open to the public). --Employment center. --Sponsorship center. --Bulletin board (business and service issue). --Flexible Online Office (Business Online Report). --Economic news. --View thousands of trade leads. --Post business propositions. --Merchandise marketing (Vast advertising at a low cost). --World shopping center. .. and much more. Please visit www.webb2e.com <http://webb2e.com/promo1.asp> If you do not want to recieve any more e-mails from WebB2E.com and wish to be removed from e-mail list please click here <http://webb2e.net/remove.asp?email=python-dev@python.org> .

1 0

RE: [Python-Dev] A kind of configuration question
by Moore, Paul 09 Apr '01

09 Apr '01

From: Guido van Rossum [mailto:guido@digicool.com] > > On a related note, has anything happened on my other bug > > report (406280)? At the very least, using tempfilepager > > instead of pipepager as a workaround would be sensible. > > Leaving things broken makes no real sense. This is a > > patch: > > What's broken? After "from pydoc import help" I can use help(...) > just fine, both in the command line version (where it invokes some > pager) and in IDLE (where it just scrolls off the window, but IDLE has > infinite scroll-back so that's no problem). This is on Win98se with > Python 2.1b2. It doesn't work in console version python.exe if you set PAGER in the environment. I have PAGER set to "less", a much better replacement for "more". This is on Win2000 SP1. It works if you leave PAGER unset. Please can this bug-fix be applied before 2.1 release? It makes it look like pydoc just "doesn't work", as things stand. And the link to having PAGER set is obscure, at best. Paul.

1 0

Progress report on PEP 242
by Paul F. Dubois 09 Apr '01

09 Apr '01

If I understand correctly I need to supply a reference implementation for PEP 242. I have made considerable progress on this: C:\Paul\Numerical\Packages\kinds>python Python 2.1b2 (#12, Mar 23 2001, 14:01:30) [MSC 32 bit (Intel)] on win32 Type "copyright", "credits" or "license" for more information. >>> import kinds >>> f=kinds.float_kind(5,100) >>> f.max 1.7976931348623157e+308 >>> f.min 2.2250738585072014e-308 >>> f.epsilon 2.2204460492503131e-016 >>> f.radix 0.3010299956639812 >>> f.epsilonbyradix 1.1102230246251565e-016 >>> 10.0**f.radix 2.0 # in case you were wondering... >>> f(7) 7.0 >>> These five attributes are the standard ones computed by routines such as d1mach. (See netlib.org, search for d1mach). These attributes I made up: f.name ('Float' in this case) f.typecode ('d' in this case, a typecode suitable for modules array or Numeric I haven't updated the PEP itself with the comments I got, but it essentially amounts to fixing the section on coercion, which I mainly realized I don't really have to deal with.

3 5