If no one has any further comments over the weekend and Monday, I'll post it as the final summary that evening/night. ============= Announcements ============= ========= Summaries ========= --------------- About SSL tests --------------- An open bug about missing SSL tests (#451607) was brought up to close or keep open. Jean-Paul Calderone mentioned an improved testing method by spawning a 'openssl s_server' for testing purposes. This sparked some talk about the terminating of subprocesses in a cross-platform manner (See `Cross-Platform Subprocess Termination`). Contributing thread: - `About SSL tests http://mail.python.org/pipermail/python-dev/2007-April/072443.html`__ ------------------------------------- Cross-Platform Subprocess Termination ------------------------------------- Because os.kill only works on *nix and OS X platforms, leaving Windows platforms uncovered. Although subprocess.TerminateProcess exists for Windows platforms, it requires the use of handles and additional overhead for use. Support was overall given for adding a UNIX-only signal() method and a cross- platform terminate() method to Popen instances. Nothing was said of actually incorporating these changes into the subprocess module, although code was given. Contributing thread: - (This was mixed in with the `About SSL tests` thread.) ------------------------ Extended buffer protocol ------------------------ Updates to the buffer protocol are discussed, along with the proposal pre-PEP documenting the updates for Python 3000, but brought to Python-Dev, due so the plans of backporting to 2.6 afterwards. (Note: I couldn't summarize this well enough to cover much of it do any real degree, but it is currently a Py3k issue, so maybe that is OK. If no one wants to add to the summary, it will stay short.) Contributing threads: - `Extended buffer PEP http://mail.python.org/pipermail/python-dev/2007-April/072620.html`__ - `PEP 3118: Extended buffer protocol (new version) http://mail.python.org/pipermail/python-dev/2007-April/072537.html`__ - `Extended Buffer Protocol - simple use examples http://mail.python.org/pipermail/python-dev/2007-April/072535.html`__ ----------------------------------------- function for counting items in a sequence ----------------------------------------- A patch was submitted by Steven Bethard (http://bugs.python.org/1696199), implemented a discussed collections.counts() function to provide a mapping between items in an iterable and the number of times they appear. There were suggested names, but none overthrew the original 'counts()' and a question of items not appearing being counted as 0 or raising a KeyError, with 0 winning due to a just-makes-sense factor. Contributing thread: - `function for counting items in a sequence http://mail.python.org/pipermail/python-dev/2007-April/072502.html`__ ---------------------------------------- context manager - generator interaction? ---------------------------------------- A problem was brought up with iterator context managers and iteration inside the with-block raising its StopIteration, but being caught by the context manager mechanics. It was also responded that the problem would not exist without the use of overly broad try blocks, and this lead to the addition of a formal note in PEP 8 about keeping narrow try blocks. Contributing thread: - `context manager - generator interaction? http://mail.python.org/pipermail/python-dev/2007-April/072484.html`__ ----------------------------- proposed which.py replacement ----------------------------- Suggestion of replacing the which.py script in the Tools directory migrated to its proposal for inclusion into the standard library. A patch and tests have yet to be provided. Contributing thread: - `proposed which.py replacement http://mail.python.org/pipermail/python-dev/2007-April/072424.html`_ ----------------------- minidom and DOM level 2 ----------------------- What is missing for DOM Level 2 support in minidom was highlighted and some work jumpstarted. Contributing thread: - `minidom and DOM level 2 http://mail.python.org/pipermail/python-dev/2007-April/072499.html`__ ---------------------------------------------- test_pty.py hangs in verbose mode on Mac OS X? ---------------------------------------------- Differing buffering behavior was causing test_pty to block only in verbose mode. Solutions may include reading to clear the buffer of child processes before a waitpid() call. Contributing thread: - `test_pty.py hangs in verbose mode on Mac OS X? http://mail.python.org/pipermail/python-dev/2007-April/072592.html`__ ------------------------- HTTP Responses and Errors ------------------------- In 2xx HTTP responses mean that the request was handled OK. The existing library was special-casing the most common responses and treating others as errors. After verifying that there wasn't a good reason for the old behavior, Facundo Batista fixed it. Contributing thread: - `HTTP responses and errors http://mail.python.org/pipermail/python-dev/2007-April/072547.html`__ ------------------------ Build Problem on Windows ------------------------ It is a generated file; the actual problem is in make_buildinfo Contributing thread: - `build problem on windows: unable to find getbuildinfo2.c http://mail.python.org/pipermail/python-dev/2007-April/072580.html`__ -------------------------- BaseException Pickle Issue -------------------------- Exceptions are now new-style classes; this caused some obscure problems with picking and unpickling. http://www.python.org/sf/1498571 and later http://www.python.org/sf/1692335 should resolve the issue. Contributing thread: - `BaseException pickle issue http://mail.python.org/pipermail/python-dev/2007-April/072416.html`__ --------------------------------- Deprecating BaseException.message --------------------------------- Plans changed; BaseException will still accept an args tuple, so don't bother with the .message attribute. Probably the shorted leaved Python feature ever. See PEP 352. Contributing thread: - `deprecating BaseException.message http://mail.python.org/pipermail/python-dev/2007-April/072542.html`__ --------------------- Changes to decimal.py --------------------- The external standard has been updated; python's implementation will be updated to match. Contributing threads: - `Changes to decimal.py http://mail.python.org/pipermail/python-dev/2007-April/072548.html`__ - `Fwd: Re: Changes to decimal.py http://mail.python.org/pipermail/python-dev/2007-April/072556.html`__ ---------------------------------------- Pydoc Rewrite Discussion at doc-sig list ---------------------------------------- An announcement that the (normally quiet) doc-sig mailing list would be discussing a rewrite of pydoc. http://mail.python.org/pipermail/doc-sig/ Contributing thread: - `Pydoc Rewrite Discussion at doc-sig list. http://mail.python.org/pipermail/python-dev/2007-April/072596.html`__ ------------------------------ Making builtins more efficient ------------------------------ Andrea Griffini posted a patch at sourceforge that makes builtin lookups almost as fast as locals. https://sourceforge.net/tracker/?func=detail&atid=305470&aid=1616125&group_id=5470 Contributing thread: - `Making builtins more efficient http://mail.python.org/pipermail/python-dev/2007-April/072610.html`__ ================ Deferred Threads ================ ================== Previous Summaries ================== =============== Skipped Threads =============== - `Checking PEP autobuild results http://mail.python.org/pipermail/python-dev/2007-April/072544.html`__ - `Python 2.5.1c1 pickle problem http://mail.python.org/pipermail/python-dev/2007-April/072565.html`__ - `Summary of Tracker Issues http://mail.python.org/pipermail/python-dev/2007-April/072417.html`__ - `possible urllib bug on Windows XP http://mail.python.org/pipermail/python-dev/2007-April/072445.html`__ - `Py2.5.1 release candidate http://mail.python.org/pipermail/python-dev/2007-April/072550.html`__ - `Some new additions to functools http://mail.python.org/pipermail/python-dev/2007-April/072614.html`__ - `Python+XUL http://mail.python.org/pipermail/python-dev/2007-April/072435.html`__ - `Distutils and -framework on MacOSX http://mail.python.org/pipermail/python-dev/2007-April/072451.html`__ - `[Python-checkins] svn dead? http://mail.python.org/pipermail/python-dev/2007-April/072559.html`__ - `Just saying hello http://mail.python.org/pipermail/python-dev/2007-April/072572.html`__ - `ok to just checkin minor obvious fixes? http://mail.python.org/pipermail/python-dev/2007-April/072600.html`__ - `__lltrace__ http://mail.python.org/pipermail/python-dev/2007-April/072608.html`__ - `new subscriber looking for grunt work http://mail.python.org/pipermail/python-dev/2007-April/072612.html`__ - `functools additions http://mail.python.org/pipermail/python-dev/2007-April/072615.html`__ - `Python Documentation Problem Example http://mail.python.org/pipermail/python-dev/2007-April/072427.html`__ - `Get 2.5 changes in now, branch will be frozen soon http://mail.python.org/pipermail/python-dev/2007-April/072429.html`__ - `Quoting netiquette reminder [Re: proposed which.py replacement] http://mail.python.org/pipermail/python-dev/2007-April/072440.html`__ - `branch is frozen for release of 2.5.1c1 (and 2.5.1) http://mail.python.org/pipermail/python-dev/2007-April/072483.html`__ - `More exposure for PEP8 (was: context manager - generator interaction?) http://mail.python.org/pipermail/python-dev/2007-April/072496.html`__ - `[Python-checkins] Python Regression Test Failures opt (1) http://mail.python.org/pipermail/python-dev/2007-April/072498.html`__ - `Weekly Python Patch/Bug Summary http://mail.python.org/pipermail/python-dev/2007-April/072534.html`__ - `USE_FAST code in stringobject.c http://mail.python.org/pipermail/python-dev/2007-April/072551.html`__ - `Fwd: Re: Py2.5.1 release candidate http://mail.python.org/pipermail/python-dev/2007-April/072553.html`__ - `svn.python.org http://mail.python.org/pipermail/python-dev/2007-April/072561.html`__ - `[PATCH] pep 0324 URL update http://mail.python.org/pipermail/python-dev/2007-April/072574.html`__ - `my 2.5 checkins http://mail.python.org/pipermail/python-dev/2007-April/072604.html`__ - `fdopen mode needs sanitizing http://mail.python.org/pipermail/python-dev/2007-April/072611.html`__ - `Py3: function signatures, type checking, and related crap http://mail.python.org/pipermail/python-dev/2007-April/072625.html`__ - `concerns regarding callable() method http://mail.python.org/pipermail/python-dev/2007-April/072514.html`__ - `A Survey on Defect Management Practices in Free/Open Source Software http://mail.python.org/pipermail/python-dev/2007-April/072470.html`__ - `RELEASED Python 2.5.1, release candidate 1 http://mail.python.org/pipermail/python-dev/2007-April/072558.html`__ - `Python 3000 PEP: Postfix type declarations http://mail.python.org/pipermail/python-dev/2007-April/072419.html`__ - `test_socketserver flakey? http://mail.python.org/pipermail/python-dev/2007-April/072441.html`__ -- Read my blog! I depend on your acceptance of my opinion! I am interesting! http://ironfroggy-code.blogspot.com/