Mailman 3 January 2001 - Python-Dev

codecity.com
by Guido van Rossum 31 Jan '01

31 Jan '01

Should I spread this word, or is this a joke? The Python quiz category is laughable. --Guido van Rossum (home page: http://www.python.org/~guido/) ------- Forwarded Message Date: Sat, 27 Jan 2001 23:16:02 -0800 From: "Jeff Cordova" <jeffc(a)codecity.com> To: <guido(a)python.org> Subject: New, fun way to learn Python. Hi Guido, I wanted to let you know about www.codecity.com After several years of managing large software projects in Silicon Valley, I realized that I was spending a lot of time teaching jr. programmers how to write code. So, I created CodeCity to help me automate some of that. If you go to the site, you'll see that I've created a category for Python. There's not much depth to the Python content yet (the site is only a week old) but I'm expecting the Python community to add their wisdom over a period of time. If you could spread the word, it would be highly appreciated. Thankyou, Jeff C. ------- End of Forwarded Message

3 2

RE: [Patch #103203] PEP 205: weak references implementation
by Tim Peters 31 Jan '01

31 Jan '01

> Patch #103203 has been updated. > > Project: python > Category: core (C code) > Status: Open > Submitted by: fdrake > Assigned to : tim_one > Summary: PEP 205: weak references implementation Fred, just noticed the new "assigned to". If you don't think it's a disaster(*), check it in! That will force more eyeballs on it quickly, and the quicker the better. I'm simply not going to do a decent review quickly on something this large starting cold. More urgently, I've been working long hours every day for several weeks, and need a break so I don't screw up last-second crises tomorrow. has-12-hours-of-taped-professional-wrestling-to-catch-up-on-ly y'rs - tim (*) otoh, if you do think it's a disaster, withdraw it for 2.1.

1 0

weak refs and jython
by Samuele Pedroni 31 Jan '01

31 Jan '01

Hi. I have read weak ref PEP, maybe too late. I don't know if portability of code using weak refs between python and jython was a goal or could be one, and up to which extent actual impl. will correspond to the PEP. But about The callbacks registered with weak references must accept a single parameter, which will be the weak-ly referenced object itself. The object can be resurrected by creating some other reference to the object in the callback, in which case the weak reference generating the callback will still be cleared but no remaining weak references to the object will be cleared. AFAIK using java weak refs (which I think is a natural choice) I see no way (at least no worth-the-effort way) to implement this in jython. Java weak refs cannot be resurrected. regards, Samuele Pedroni. PS: Mr. X is a jython developer.

3 3

WARNING: Changed build process for zlib on Windows
by Mark Hammond 31 Jan '01

31 Jan '01

Hi all, In an attempt to solve "[ Bug #129293 ] zlib library used for binary win32 distribution can crash" (https://sourceforge.net/bugs/?func=detailbug&group_id=5470&bug_id=129293), Tim and I have decided that we should fix the build process of zlib.pyd on windows. The current process requires that the builder download _2_ zlib archives - a binary distribution for zlib.lib, and the source archive for the headers. We believe that slight differences between the 2 are causing the above bug. A particular warning-light is that the current process defines ZLIB_DLL even though we are _not_ currently using the DLL but the static lib. Removing this #define generates linker errors. The new process is very simple, but may break some peoples build. In theory it _should_ still work for everyone, but if it fails to build, please check your directory structure.

2 1

Re: from ... import * ([Python-checkins] CVS: python/dist/src/Python compile.c,2.153,2.154)
by M.-A. Lemburg 31 Jan '01

31 Jan '01

> Update of /cvsroot/python/python/dist/src/Python > In directory usw-pr-cvs1:/tmp/cvs-serv17061/Python > > Modified Files: > compile.c > Log Message: > Enforce two illegal import statements that were outlawed in the > reference manual but not checked: Names bound by import statemants may > not occur in global statements in the same scope. The from ... import * > form may only occur in a module scope. > > I guess these changes could break code, but the reference manual > warned about them. Jeremy, your code breaks all uses of "from package import submodule" inside packages. Try distutils for example or setup.py.... -- Marc-Andre Lemburg ______________________________________________________________________ Company: http://www.egenix.com/ Consulting: http://www.lemburg.com/ Python Pages: http://www.lemburg.com/python/

2 2

FW: Python programmer needed (addition to urllib2 and HTTPS support)
by Tim Peters 31 Jan '01

31 Jan '01

Get rich quick! -----Original Message----- From: python-list-admin(a)python.org [mailto:python-list-admin@python.org]On Behalf Of Albert Chin-A-Young Sent: Wednesday, January 31, 2001 2:31 AM To: python-list(a)python.org Subject: Python programmer needed (addition to urllib2 and HTTPS support) We're in need of a contract Python programmer for the following: 1. Allow connecting to a host with urlopen() which requires BASIC HTTP authentication with a proxy (via urllib2.py). This should address bug #125217: http://sourceforge.net/bugs/?func=detailbug&bug_id=125217&group_id=5470 2. Allow connecting to a host with urlopen() which requires BASIC HTTP authentication with a proxy that requires BASIC HTTP authentication (via urllib2.py). 3. Support for non-authenticated clients to connect to a HTTPS server 4. Support for a client to authenticate the HTTPS host (to verify that it's certificate is valid) What we might consider adding (depends on cost): 1. Support for authenticated clients to connect to a HTTPS server. Please note that solutions to the four items above must be rolled back into the main Python distribution (implies the "community" and the Python developers need to agree on the adopted solution). -- albert chin (china at thewrittenword dot com) -- http://mail.python.org/mailman/listinfo/python-list

1 0

Can't enter new Python bugs on SourceForge?
by Tim Peters 31 Jan '01

31 Jan '01

Reported this earlier. Still can't create a new bug. Guido either. Here's the SF Support request opened on this: http://sourceforge.net/support/ index.php?func=detailsupport&support_id=113100&group_id=1 The good(?) news is that Python isn't the only project to report this problem.

1 0

SSL socket read at EOF; SourceForge problem
by Tim Peters 30 Jan '01

30 Jan '01

I tried to open an SF bug for the following msg from c.l.py, but SF balked: ERROR ERROR getting bug_id Logged out, logged in, tried it again, same outcome. Intended bug report content: Good question from c.l.py, assigned to Guido cuz he's a Socket Guy: From: Clarence Gardner <clarence(a)netlojix.com> Subject: RE: Thread Safety Date: Mon, 29 Jan 2001 09:51:03 -0800 ... I'm going to repeat a question that I posted about a week ago that passed without comment on the newsgroup. The issue is the SSL support in the socket module, which raises an exception when the reading socket is at EOF, rather than returning an empty string. I'm hesitant to call it a "bug", but I wouldn't have implemented it this way. There are the names of two people mentioned at the top of socketmodule.c, but no contact information, so I'm suggesting here that it be changed to conform to normal file/socket practice. (SSL was actually added at 2.0, so I'm late to the party with this; mea culpa, mea culpa. I delayed trying Python2 because of the extension rebuilding.)

2 1

Let's release 2.1a2 Thursday night
by Guido van Rossum 30 Jan '01

30 Jan '01

Things look good for a release of 2.1a2 this week; we're aiming for Thursday night. I won't be in town (speaking to the press at LinuxWorld Expo in New York) but Jeremy will handle the release process and the other PythonLabs folks will assist him. Tomorrow Fred will check in his weak references after making some changes (mostly making it more Spartan :-) that I suggested in a code review. After that, I think we're good for the second (and last!) alpha release; and enough has changed (e.g. nested scopes, lots of setup.py changes, flat Makefile) to warrant going ahead now. Now is the time for those last-minute bugfixes that you're all so famous for! I propose a checkin freeze for non-PythonLabs folks Wednesday midnight US west coast time, to give Jeremy c.s. enough time to build the release and give it a good work-out. (An internal freeze is up to Jeremy to declare, but should probably take Tim's sleep cycle into account.) --Guido van Rossum (home page: http://www.python.org/~guido/) PS. I'll be out of reach from noon US east coast time tomorrow (Wednesday), traveling to New York by train. I probably won't check my email while out there; I'll be back Friday night.

1 0

Read-only function attributes
by Ka-Ping Yee 30 Jan '01

30 Jan '01

Hi there. I see that the function attribute feature specifically allows assignment to func_code and func_defaults, but no other special attributes. This seems really suspect to me. Why would we want to allow the reassignment of special attributes at all? Functions have always been immutable objects, and i can see some motivation for attaching mutable dictionaries to them, but it's a more serious move to make the functions mutable themselves. I don't recall any discussion about changing special attributes; i don't see a clear purpose to them; and i do see a danger in making it harder to be certain that a program is safe and predictable. (Yes, i did notice that function attributes can't be set in restricted mode, but the addition of extra features requiring extra security checks makes me uneasy.) -- ?!ng

4 3