Mailman 3 June 2009 - Python-Dev

Re: [Python-Dev] Status of 2.7 and 3.2
by Eric Pruitt June 13, 2009

June 13, 2009

I am in the process of implementing a number of often requested features and proposed patches in the subprocess module for my Google Summer of Code 2009 project. For information on my progress, check out my blog located at * http://subdev.blogspot.com/* <http://subdev.blogspot.com/>. Any comments and suggestions are greatly appreciated. > Just my 0.02 cents, but struggling with all warts of 2.5 subprocessing > > in Windows I would vote for more time for stabilizating things - not > > adding new features. Long awaited subprocess as replacement for > > os.popen() AFAIK is still incapable to asynchronously communicate with > > spawned processes on Windows. I would call this feature as critical > > even on 2.6 As a release testcase - try porting pyexpect module to > > this platform. Absence of native curses/console/readline module also > > makes Python one-way unix shell language while many users expect it to > > be crossplatform. > > I am not quite sure whether you are for new features or not. Your > first sentence ("vote for ... not adding new features") seems to > suggest that you would not like to see new features, and your last > sentence ("absence of native curses/console/readline module") > suggests that you *do* want to see new features (namely, a native > curses module, and a native readline module). > > Which one is it? > > Regards, > Martin >

2 1

Status of 2.7 and 3.2
by Terry Reedy June 13, 2009

June 13, 2009

2.7 Once upon a time, the plan was to come out with 2.6 and 3.0, and then after the usual interval, 2.7 and 3.1. As it turned out, 3.0 came out 3 months after 2.6, but, as it typical of x.0 releases, had some flaws leading to 3.1 now just 6 month later. I have thought that 2.7 was now to come out instead with 3.2 and would include backported 3.2 new features. Others expect 2.7 to come out soon after 3.1 and to only contain new 3.1 features. So Guido or someone, please clarify: is 2.7 to be the counterpart of 3.1 or 3.2? Data: On the tracker, new features are being assigned to 2.7 and 3.2. PEP 373 Python 2.7 Release Schedule says zilch: Release Schedule: Not yet finalized Possible features for 2.7: Nothing here 3.2 At some point, 3.x will become the "trunk" branch. It seems to me that this should be done with 3.2 as part of the transition to Mercurial. A. As long as 2.x is 'trunk', some people will view 3.x as 'experimental'. That was true for 3.0, but (much?) less so for 3.1. Is there any known reason why 3.2 should not soon be considered and treated as the main development version, to become the main production version? B. All new features will go into 3.2. Only some will be backported to 2.x. So it seems that the flow should be to develop for 3.2 and maybe backport thereafter. 2.final Is there any thought of making 2.7 be 2.final? A. To my mind, the main reason to add features to 2.x is to make transition to 3.x easier, rather than to discourage transition to 3.x. B. Do we really want to encourage library developers to put their 'upgrade to a new version' energy into 2.x to 2.x+1 upgrades, rather than a 2.x to 3.y upgrades? C. Some people are sticking with some version of 2.x because they want a stable version with minimal disturbance. Such people might have preferred, for instance, getting 2.5.5 in April instead of the latest 2.6 release. Instead people 2.5 fixes are being told "Sorry. 2.5 is out of Maintenance phase and into SecurityFix only phase." Once 2.x is put in feature freeze, micro bugfix releases can appear for years, as long as bugs are found and patches submitted and committed. It should gradually become truly rock solid. Terry Jan Reedy

12 21

Summary of Python tracker Issues
by Python tracker June 12, 2009

June 12, 2009

ACTIVITY SUMMARY (06/05/09 - 06/12/09) Python tracker at http://bugs.python.org/ To view or respond to any of the issues listed below, click on the issue number. Do NOT respond to this message. 2225 open (+39) / 15869 closed (+26) / 18094 total (+65) Open issues with patches: 882 Average duration of open issues: 652 days. Median duration of open issues: 402 days. Open Issues Breakdown open 2197 (+39) pending 28 ( +0) Issues Created Or Reopened (68) _______________________________ Issue with RotatingFileHandler logging handler on Windows 06/08/09 CLOSED http://bugs.python.org/issue4749 reopened r.david.murray add disable_nagle_algorithm to SocketServer.TCPServer 06/08/09 http://bugs.python.org/issue6192 reopened pitrou patch, patch, easy test_float fails on Windows 06/05/09 CLOSED http://bugs.python.org/issue6198 reopened eric.smith path separator output ignores shell's path separator: / instead 06/05/09 http://bugs.python.org/issue6208 created ThurnerRupert compilation error in std. lib. module shutil (Python 3.1rc1, pla 06/05/09 CLOSED http://bugs.python.org/issue6209 created dpodbori Exception Chaining missing method for suppressing context 06/05/09 http://bugs.python.org/issue6210 created poke [Tutorial] Section 4.7.2 has a wrong description of an example 06/05/09 CLOSED http://bugs.python.org/issue6211 created cofi patch piped input 06/05/09 CLOSED http://bugs.python.org/issue6212 created rtmq Incremental encoder incompatibility between 2.x and py3k 06/05/09 http://bugs.python.org/issue6213 created pitrou test__locale broken on trunk 06/05/09 CLOSED http://bugs.python.org/issue6214 created pitrou Backport the IO lib to trunk 06/05/09 http://bugs.python.org/issue6215 created pitrou patch Raise Unicode KEEPALIVE_SIZE_LIMIT from 9 to 32? 06/06/09 http://bugs.python.org/issue6216 created tjreedy Add _io._TextIOWrapper.errors 06/06/09 CLOSED http://bugs.python.org/issue6217 created pjenvey patch Make io.BytesIO and io.StringIO picklable. 06/06/09 http://bugs.python.org/issue6218 created alexandre.vassalotti patch, patch nested list value change 06/06/09 CLOSED http://bugs.python.org/issue6219 created pushkarparanjpe typo: opne in "doanddont" 06/06/09 CLOSED http://bugs.python.org/issue6220 created cate Windows buildbot failure in test_winreg 06/06/09 CLOSED http://bugs.python.org/issue6221 created pitrou 2to3 except fixer failed in certain case 06/06/09 CLOSED http://bugs.python.org/issue6222 created bhy Make _PyUnicode_AsString as public API 06/06/09 CLOSED http://bugs.python.org/issue6223 created bhy References to JPython 06/06/09 CLOSED http://bugs.python.org/issue6224 created thijs Fixing several minor bugs in Tkinter.Canvas and one in Misc._con 06/06/09 http://bugs.python.org/issue6225 created gpolo patch Inconsistent 'file' vs 'stream' kwarg in pprint, other stdlibs 06/06/09 http://bugs.python.org/issue6226 created pjenvey doctest_aliases doesn't test duplicate removal 06/07/09 http://bugs.python.org/issue6227 created abbeyj patch round() error 06/07/09 CLOSED http://bugs.python.org/issue6228 created steve21 Installation python on mac 06/07/09 CLOSED http://bugs.python.org/issue6229 created kostonstyle ElementTree.Element and cElementTree.Element have slightly diffe 06/07/09 http://bugs.python.org/issue6230 created Neil Muller patch ElementInclude may drop text 06/07/09 http://bugs.python.org/issue6231 created Neil Muller patch Improve test coverage of ElementTree and cElementTree 06/07/09 http://bugs.python.org/issue6232 created Neil Muller ElementTree (py3k) doesn't properly encode characters that can't 06/07/09 http://bugs.python.org/issue6233 created Neil Muller patch cgi.FieldStorage is broken when given POST data 06/07/09 http://bugs.python.org/issue6234 created efosmark \d missing from effects of re.ASCII flag 06/08/09 CLOSED http://bugs.python.org/issue6235 created MLModel os.popen causes illegal seek on AIX in Python 3.1rc 06/08/09 http://bugs.python.org/issue6236 created nestor Build errors when using LDFLAGS="-Wl,--no-undefined" 06/08/09 http://bugs.python.org/issue6237 created fundawang string.ljust documentation is missing optional fillchar descript 06/08/09 CLOSED http://bugs.python.org/issue6238 created jaraco c_char_p return value returns string, not bytes 06/08/09 http://bugs.python.org/issue6239 created georg.brandl API to get source encoding as defined by PEP 263 06/08/09 CLOSED http://bugs.python.org/issue6240 created srid Better type checking for the arguments of io.StringIO 06/08/09 http://bugs.python.org/issue6241 created alexandre.vassalotti patch Fix reference leak in io.StringIO 06/08/09 http://bugs.python.org/issue6242 created alexandre.vassalotti patch getkey() can segfault in combination with curses.ungetch() 06/08/09 http://bugs.python.org/issue6243 created Trundle patch Support for tcl 8.6 06/08/09 http://bugs.python.org/issue6244 created misc patch Add "intel" universal architecture on OSX 06/08/09 http://bugs.python.org/issue6245 created ronaldoussoren patch, needs review Python debugger ignores exception if instructed to return from g 06/08/09 http://bugs.python.org/issue6246 created inducer should we include argparse 06/09/09 CLOSED http://bugs.python.org/issue6247 reopened rickysarraf TCP Sockets not closed by TCPServer and StreamRequestHandler 06/09/09 http://bugs.python.org/issue6248 created krisvale patch, patch, needs review Error Prompt 06/10/09 CLOSED http://bugs.python.org/issue6249 created SonMarvin Python compiles dead code 06/10/09 http://bugs.python.org/issue6250 created abbeyj patch c++ extension module implementation guide/example in extending/e 06/10/09 http://bugs.python.org/issue6251 created subgeometer logging midnight rotation 06/10/09 CLOSED http://bugs.python.org/issue6252 created SanityIO optparse.OptionParser.get_usage uses wrong formatter 06/10/09 http://bugs.python.org/issue6253 created ccx patch tarfile unnecessarily requires seekable files 06/10/09 CLOSED http://bugs.python.org/issue6254 created johnsonm PyInt_FromSize_t is undocumented. 06/10/09 http://bugs.python.org/issue6255 created naoki Wrong stacklevel in warning for contextlib.nested 06/10/09 CLOSED http://bugs.python.org/issue6256 created hagen patch Idle terminates on source save while debugging 06/10/09 http://bugs.python.org/issue6257 created andyharrington distributions built with bdist_msi on 64-bit Windows fail to ins 06/10/09 http://bugs.python.org/issue6258 created jaraco ctypes pointer arithmetic 06/10/09 http://bugs.python.org/issue6259 created theller patch os.utime should allow None values for ATIME or MTIME 06/10/09 CLOSED http://bugs.python.org/issue6260 created purpleidea Clarify behaviour of random.uniform 06/10/09 CLOSED http://bugs.python.org/issue6261 created marketdickinson VS 2008 binaries 06/11/09 CLOSED http://bugs.python.org/issue6262 created bellenot syntax error in get_msvcr 06/11/09 CLOSED http://bugs.python.org/issue6263 created tarek Misleading instructions for building extensions with mingw 06/11/09 CLOSED http://bugs.python.org/issue6264 created bdew cElementTree & ElementTree use different exceptions for XML Erro 06/11/09 http://bugs.python.org/issue6265 created Neil Muller cElementTree.iterparse & ElementTree.iterparse return differentl 06/11/09 http://bugs.python.org/issue6266 created Neil Muller Cumulative patch to http and xmlrpc 06/11/09 http://bugs.python.org/issue6267 created krisvale Seeking to the beginning of a text file a second time will retur 06/11/09 http://bugs.python.org/issue6268 created eggy threading documentation makes no mention of the GIL 06/11/09 http://bugs.python.org/issue6269 created segfaulthunter patch Menu deletecommand fails if command is already deleted 06/11/09 http://bugs.python.org/issue6270 created gregcouch patch mmap: don't close file description if fd=-1 06/12/09 http://bugs.python.org/issue6271 created haypo patch Upgrading xml.etree to ElementTree 1.3 06/12/09 http://bugs.python.org/issue6272 created orsenthil Issues Now Closed (52) ______________________ Improve the hackish runtime_library_dirs support for gcc 653 days http://bugs.python.org/issue1032 alexandre.vassalotti patch Builtin round function is sometimes inaccurate for floats 50 days http://bugs.python.org/issue1869 marketdickinson patch Set up "supported"-only buildbot waterfall view 452 days http://bugs.python.org/issue2376 exarkun "No windows home dir" doc error 384 days http://bugs.python.org/issue2922 r.david.murray subprocess (Replacing popen) - add a warning / hint 382 days http://bugs.python.org/issue2947 r.david.murray patch, easy __newobj__ pickle feature is not documented 272 days http://bugs.python.org/issue3816 alexandre.vassalotti easy ctypes documentation 208 days http://bugs.python.org/issue4309 mnewman Issue with RotatingFileHandler logging handler on Windows 3 days http://bugs.python.org/issue4749 vsajip Using LDFLAGS='-rpath=\$$LIB:/some/other/path' ./configure break 121 days http://bugs.python.org/issue5201 tarek patch test_fdopen fails with vs2005, release build on Windows 2000 70 days http://bugs.python.org/issue5623 krisvale patch "No such file or directory" with framework build under MacOS 10. 48 days http://bugs.python.org/issue5809 ronaldoussoren When setting complete PYTHONPATH on Python 3.x, paths in the PYT 36 days http://bugs.python.org/issue5924 amaury.forgeotdarc patch Make pickle generated by Python 3.x compatible with 2.x and vice 12 days http://bugs.python.org/issue6137 rhettinger patch Python 3.1 rc1 will not build on OS X 10.5.7 with macports libin 8 days http://bugs.python.org/issue6154 brett.cannon patch fcntl footnote about O_SHLOCK and O_EXLOCK is misleading 4 days http://bugs.python.org/issue6194 georg.brandl easy Serious regression in doctest in Py3.1rc1 8 days http://bugs.python.org/issue6195 r.david.murray patch tarfile.extractall(readaccess=True) 1 days http://bugs.python.org/issue6196 lars.gustaebel test_float fails on Windows 4 days http://bugs.python.org/issue6198 eric.smith test_winreg fails 5 days http://bugs.python.org/issue6201 amaury.forgeotdarc Obsolete default file encoding "mac-roman" on OS X, not influenc 2 days http://bugs.python.org/issue6202 nad patch Missing reference in section 4.6 to chapter on classes 1 days http://bugs.python.org/issue6204 georg.brandl patch Correct a trivial typo introduced by r73238. 1 days http://bugs.python.org/issue6206 georg.brandl patch Simple For-Loops 0 days http://bugs.python.org/issue6207 pitrou compilation error in std. lib. module shutil (Python 3.1rc1, pla 0 days http://bugs.python.org/issue6209 benjamin.peterson [Tutorial] Section 4.7.2 has a wrong description of an example 1 days http://bugs.python.org/issue6211 georg.brandl patch piped input 0 days http://bugs.python.org/issue6212 georg.brandl test__locale broken on trunk 1 days http://bugs.python.org/issue6214 ezio.melotti Add _io._TextIOWrapper.errors 1 days http://bugs.python.org/issue6217 benjamin.peterson patch nested list value change 0 days http://bugs.python.org/issue6219 rhettinger typo: opne in "doanddont" 0 days http://bugs.python.org/issue6220 ezio.melotti Windows buildbot failure in test_winreg 1 days http://bugs.python.org/issue6221 benjamin.peterson 2to3 except fixer failed in certain case 0 days http://bugs.python.org/issue6222 benjamin.peterson Make _PyUnicode_AsString as public API 0 days http://bugs.python.org/issue6223 benjamin.peterson References to JPython 0 days http://bugs.python.org/issue6224 georg.brandl round() error 0 days http://bugs.python.org/issue6228 marketdickinson Installation python on mac 0 days http://bugs.python.org/issue6229 benjamin.peterson \d missing from effects of re.ASCII flag 0 days http://bugs.python.org/issue6235 georg.brandl string.ljust documentation is missing optional fillchar descript 0 days http://bugs.python.org/issue6238 georg.brandl API to get source encoding as defined by PEP 263 0 days http://bugs.python.org/issue6240 benjamin.peterson should we include argparse 3 days http://bugs.python.org/issue6247 bethard Error Prompt 0 days http://bugs.python.org/issue6249 georg.brandl logging midnight rotation 1 days http://bugs.python.org/issue6252 vsajip tarfile unnecessarily requires seekable files 2 days http://bugs.python.org/issue6254 lars.gustaebel Wrong stacklevel in warning for contextlib.nested 0 days http://bugs.python.org/issue6256 rhettinger patch os.utime should allow None values for ATIME or MTIME 2 days http://bugs.python.org/issue6260 purpleidea Clarify behaviour of random.uniform 1 days http://bugs.python.org/issue6261 rhettinger VS 2008 binaries 1 days http://bugs.python.org/issue6262 loewis syntax error in get_msvcr 0 days http://bugs.python.org/issue6263 tarek Misleading instructions for building extensions with mingw 1 days http://bugs.python.org/issue6264 bdew IDLE / PyOS_InputHook 2108 days http://bugs.python.org/issue798058 gpolo import on Windows: please call SetErrorMode first 1664 days http://bugs.python.org/issue1069410 srid logging module / wrong bytecode? 876 days http://bugs.python.org/issue1633605 annacoder Top Issues Most Discussed (10) ______________________________ 31 Issue with RotatingFileHandler logging handler on Windows 3 days closed http://bugs.python.org/issue4749 12 should we include argparse 3 days closed http://bugs.python.org/issue6247 10 3.x locale does not default to C, contrary to the documentation 7 days open http://bugs.python.org/issue6203 8 Make pickle generated by Python 3.x compatible with 2.x and vic 12 days closed http://bugs.python.org/issue6137 8 Make logging configuration files easier to use 15 days open http://bugs.python.org/issue6136 7 Serious regression in doctest in Py3.1rc1 8 days closed http://bugs.python.org/issue6195 7 GCC detection for runtime_library_dirs when ccache is used 1403 days open http://bugs.python.org/issue1254718 5 tarfile unnecessarily requires seekable files 2 days closed http://bugs.python.org/issue6254 5 Add "intel" universal architecture on OSX 4 days open http://bugs.python.org/issue6245 5 typo: opne in "doanddont" 0 days closed http://bugs.python.org/issue6220

1 0

Adding key and reverse args to bisect
by David A. Barrett June 11, 2009

June 11, 2009

I propose adding the key parameter to the bisect.bisect routines. This would allow it to be used on lists with an ordering other than the one "natural" to the contained objects. (and anywhere else it makes sense in the bisect module). Would this be easy enough to do? It looks like the main difficulty may have to do with the C implementation. I've also noticed that sort appears far faster; is the C implementation working as expected? It may be nice to have the reverse parameter as well, but I'm not sure how that could be implemented right off the bat. David A. Barrett - - - - Here's an example I've coded up for my own use (reverse hasn't been implemented here): def bisect_right(a, x, lo=0, hi=None, key=lambda x: x, reverse=False): """A keyed version of bisect.bisect_bisect_right: return i: for all e in a[:i] e <= x < f, for all f in a[i:] """ if hi is None: hi = len(a) if reverse: while lo < hi: mid = (lo+hi)//2 if key(a[mid]) < key(x): hi = mid else: lo = mid+1 else: while lo < hi: mid = (lo+hi)//2 if key(x) < key(a[mid]): hi = mid else: lo = mid+1 return lo def bisect_left(a, x, lo=0, hi=None, key=lambda x: x, reverse=False): """A keyed version of bisect.bisect_bisect_left: return i: for all e in a[:i] e < x <= f, for all f in a[i:] """ if hi is None: hi = len(a) if reverse: while lo < hi: mid = (lo+hi)//2 if key(x) < key(a[mid]): lo = mid+1 else: hi = mid else: while lo < hi: mid = (lo+hi)//2 if key(a[mid]) < key(x): lo = mid+1 else: hi = mid return lo

3 2

Rewrite os.popen using subprocess.Popen?
by Nestor June 10, 2009

June 10, 2009

pydoc.py uses os.popen once on line 1359. According to the documentation that function is deprecated since Python 2.6. Does it make sense to rewrite that line using the newer subprocess instead? I am asking because os.popen stopped working for me in Python 3.1 but I don't know if it is worth investing time in a module that is deprecated when the newer one does work. http://bugs.python.org/issue6236

1 0

Tags from closed branch heads?
by Dirkjan Ochtman June 9, 2009

June 9, 2009

So ca8d05e1f1d1 changed the default for repo.heads() to closed=False, so that calls to heads() by default will not return closed heads. Unfortunately, this also means that any tags from those heads will not be considered anymore. That seems inadvertent at best, and may be should be reverted. Conceptually, it seems wrong that deleting a branch would also delete all the tags from it. I'd like to propose this change: diff --git a/mercurial/localrepo.py b/mercurial/localrepo.py --- a/mercurial/localrepo.py +++ b/mercurial/localrepo.py @@ -320,7 +320,7 @@ def _hgtagsnodes(self): last = {} ret = [] - for node in reversed(self.heads()): + for node in reversed(self.heads(closed=True)): c = self[node] rev = c.rev() try: Is that something we can agree on? Cheers, Dirkjan

1 1

xmlrpc and http/1.1
by Kristján Valur Jónsson June 9, 2009

June 9, 2009

Hello there. I've been doing some work on xmlrpc for CCP in the last weeks. I'm trying to communicate the results of this back to the python trunk. I've had the following issues open for a while now: http://bugs.python.org/issue6096 http://bugs.python.org/issue6099 I would appreciate some comments on this. Moving forward, I'd like to add support for the "gzip" content encoding for XMLRPC requests and responses. K

1 0

Cannot set PYTHONPATH with big paths with Python 3.0 and 3.1
by Fabio Zadrozny June 9, 2009

June 9, 2009

Hi all, I've reported bug http://bugs.python.org/issue5924 some time ago and I think it's a release blocker -- it seems easy to fix, but I don't have time to actually submit a patch, so, I'd like to draw attention to it, especially as a release candidate is already out. Cheers, Fabio

4 3

Draft PEP 385: Migrating from svn to Mercurial
by Dirkjan Ochtman June 8, 2009

June 8, 2009

So, a while ago Martin von Löwis asked who would champion the migration to Mercurial, and I volunteered. He asked me to produce a PEP outlining which steps to take, which I've been working on. The PEP is numbered 385, and is just about ready for your review. I'd really welcome any sort of feedback or questions about Mercurial and surrounding tools relevant to the migration project (I'm happy to answer other questions as well, but maybe python-dev is not the appropriate forum for that -- instead, try mercurial(a)selenic.com and/or #mercurial on freenode). http://www.python.org/dev/peps/pep-0385/ In particular, you may want to review (a) branches that you care about and whether they would be migrated under the current proposal, (b) non-release tags that you think are useful and (c) the author map (check if it has a correct email address for you and people you know). Thanks to Antoine, Benjamin, Georg and Martin for early feedback. Cheers, Dirkjan

7 12

Roundup keywords for bug tracking
by anatoly techtonik June 8, 2009

June 8, 2009

It is impossible to edit roundup keywords and this takes away the flexibility in selecting bugs related to a module/function/test or some other aspect of development. For example, I need to gather all subprocess bugs in one query and things that won't be fixed in deprecated os.popen() into another. In Trac I would use "subprocess" and "os.popen" keywords. On ohloh I would add similar tags (if bugs were software) without, but I can't do anything about Python roundup. Is there any reason for such restriction? --anatoly t.

4 4