Mailman 3 October 2020 - Python-Dev

Thank you Larry Hastings!
by Barry Warsaw 06 Oct '20

06 Oct '20

They say being a Python Release Manager is a thankless job, so the Python Secret Underground (PSU), which emphatically does not exist, hereby officially doesn’t thank Larry for his years of diligent service as the Python 3.4 and 3.5 release manager. On the other hand, the Python Steering Council, Python Software Foundation, and worldwide Python community, all of which emphatically *do* exist, all extend our heartfelt thanks to Larry for his excellent stewardship of Python 3.4 and 3.5! Python 3.4 and 3.5 were both pivotal releases. While the features of these two releases are too numerous to mention here, they introduced such staples as: * asyncio * enum * pathlib * async and await keywords * matrix multiplication operators * typing and zipapp modules and so much more. For details, see: * https://docs.python.org/3/whatsnew/3.4.html * https://docs.python.org/3/whatsnew/3.5.html Larry’s first official release of 3.4.0a1 was on 2013-08-03 and his last Python 3.5.10 release was 2020-09-05. That’s 7 years of exemplary release managing! Larry, from all of us, and from me personally, thank you so much for your invaluable contributions to Python. Enjoy your retirement! Cheers, -Barry (on behalf of the PSC and PSF)

15 14

PR checks hang because travis does not report back to github
by Stefan Behnel 06 Oct '20

06 Oct '20

Hi devs, I have a trivial documentation PR https://github.com/python/cpython/pull/22464 for which travis (unsurprisingly) had a successful run, https://travis-ci.com/github/python/cpython/builds/187435578 but github lists the travis build as "created" instead of "passed". https://github.com/python/cpython/pull/22464/checks?check_run_id=1188595760 I already tried closing the PR and reopening it, and also triggering the build again on travis side, but github still fails to pick up the build status. I tried creating a new PR, but it seems that github (or travis) deduplicate the build requests and still refer to the original build, so that there is still no response from travis. I also cannot find a way to terminate the checks process in github, or otherwise make it stop waiting for Godot. Is this a known issue? Is there anything I can do about it? Thanks, Stefan

6 8

Where it is decided whether "lib" or "lib64" ends up in sys.path on a POSIX (non MacOS) system?
by Mikhail Golubev 05 Oct '20

05 Oct '20

Hi everyone! First of all, sorry for bringing up, perhaps, a trivial matter, but since it's about the sources I guess it's better to ask it here. I'm investigating a curious problem caused by the fact that in a user's virtualenv, created with the standard "venv" module, "lib64" directory, traditionally symlinked to "lib", gets included in sys.path instead of "lib" itself. ``` $ python -msite sys.path = [ '/current/working/directory' '/usr/lib64/python36.zip', '/usr/lib64/python3.6', '/usr/lib64/python3.6/lib-dynload', '/path/to/venv/lib64/python3.6/site-packages', ] ``` The user is on Gentoo, and, though the venv's layout is exactly the same, it's different from the typical result on a Debian-based distribution where it's the original "lib" in sys.path, not its "lib64" link. I'm wondering what's controlling this behavior. Modules/getpath.c doesn't mention architecture-dependent lib directories anywhere, at least at first glance. Could someone please give me a hint about where to look in the sources or, maybe, which build options are affecting this? Thanks a lot!

3 3

Core Dev Sprint: Collecting Discord usernames
by Kyle Stanley 03 Oct '20

03 Oct '20

Hey all, The upcoming 2020 Python Core Development Sprint is right around the corner, and we're currently towards the end of finalizing platform setup. For using Discord during the sprint on the public Python Discord server (as voted on in https://discuss.python.org/t/2020-core-dev-sprint-poll-discord-or-zulip/501…), we are collecting Discord usernames ahead of time and requesting that all participants join Python Discord well before the start of the sprint (Oct. 19th). This will allow us to grant all participants a role for joining the sprint channels (in order to be given the role, participants must be on the server first). To send your Discord username, simply reply to this email (without CCing the list to keep it private) with it included. It is in the format of "name#0123", and can be seen in the lower left corner of the Discord client after creating an account (or in "User Settings" > "My Account"). I'll verify replies against the email address used in the signup form, so please use the same one if possible. Prior to joining Python Discord, I recommend checking out the Discord setup guide that I recently finished: https://python-core-sprint-2020.readthedocs.io/communication.html#discord-s…. The part on the privacy settings is particularly important for anyone with a public following to avoid receiving mass DMs and friend requests. Alternatively, you can join directly at the following perma-link: https://discord.gg/Q87A9Y9. As a reminder, potential participants include Python core developers, triagers, and those in a core dev mentorship. If you haven't already signed up and are interested in attending, please do so at https://forms.gle/fhzJdpRHR4GtSRCk9. Regards, Kyle Stanley

1 1

Summary of Python tracker Issues
by Python tracker 02 Oct '20

02 Oct '20

ACTIVITY SUMMARY (2020-09-25 - 2020-10-02) Python tracker at https://bugs.python.org/ To view or respond to any of the issues listed below, click on the issue. Do NOT respond to this message. Issues counts and deltas: open 7688 ( +9) closed 45960 (+41) total 53648 (+50) Open issues with patches: 3138 Issues opened (36) ================== #41773: Clarify handling of infinity and nan in random.choices https://bugs.python.org/issue41773 reopened by rhettinger #41864: Compiler error in _zoneinfo.c:1227: error: #pragma GCC diagnos https://bugs.python.org/issue41864 opened by micwin #41866: Document error in chinese version of contextlib. https://bugs.python.org/issue41866 opened by siyuan #41867: Include options for timespec in docstrings of isoformat https://bugs.python.org/issue41867 opened by cool-RR #41868: SMTPLIB integrate or provide option to use "logging" https://bugs.python.org/issue41868 opened by hpkkumar007 #41869: Multiprocessing Manager Object Cleanup https://bugs.python.org/issue41869 opened by boom0192 #41875: __builtin_unreachable error in gcc 4.4.5 https://bugs.python.org/issue41875 opened by yota moteuchi #41876: Add __repr__ for Tkinter Font objects https://bugs.python.org/issue41876 opened by serhiy.storchaka #41877: Check against misspellings of assert etc. in mock https://bugs.python.org/issue41877 opened by vabr2 #41878: python3 fails to use custom mapping object as symbols in eval( https://bugs.python.org/issue41878 opened by Robert Haschke #41879: Outdated description of async iterables in documentation of as https://bugs.python.org/issue41879 opened by nickgaya #41880: Get Python include directories from sysconfigdata https://bugs.python.org/issue41880 opened by benwolsieffer #41882: CCompiler.has_function does not delete temporary files https://bugs.python.org/issue41882 opened by termim #41883: ctypes pointee goes out of scope, then pointer in struct dangl https://bugs.python.org/issue41883 opened by NankerPhelge #41884: tempfile.py TemporaryDirectory/mkdtemp defaults to mode 0o700, https://bugs.python.org/issue41884 opened by ray-sit #41886: PyType_Type is documented incorrectly https://bugs.python.org/issue41886 opened by da-woods #41887: ast.literal_eval does not accept strings with leading whitespa https://bugs.python.org/issue41887 opened by gousaiyang #41889: enum: Mixin and int base class regression in 3.8.6 https://bugs.python.org/issue41889 opened by puddly #41891: asyncio.wait_for does not wait for task/future to be completed https://bugs.python.org/issue41891 opened by rkojedzinszky #41892: use both "for in" and "ElementTree.remove" has a index bug https://bugs.python.org/issue41892 opened by WoodyWoo #41894: UnicodeDecodeError during load failure in non-UTF-8 locale https://bugs.python.org/issue41894 opened by kadler #41895: PyMethodDef does NOT have any fields contain context in embedd https://bugs.python.org/issue41895 opened by dexter #41896: Moving index with wordstart expression includes non-alphanumbe https://bugs.python.org/issue41896 opened by pyTama #41898: Any logging causes assertLogs to pass https://bugs.python.org/issue41898 opened by udalrich.schermer #41900: XML C14N serialisation fails with default namespace https://bugs.python.org/issue41900 opened by scoder #41902: Micro optimization for range.index if step is 1 https://bugs.python.org/issue41902 opened by corona10 #41903: PyNumber_InPlacePower ignores o3 if o1 implements __ipow__ https://bugs.python.org/issue41903 opened by hodgestar #41904: datetime.datetime.today makes no sense and should be removed https://bugs.python.org/issue41904 opened by yurzo #41905: add update_abstractmethods function to update an ABC's abstrac https://bugs.python.org/issue41905 opened by avrahami.ben #41906: logging.config.dictConfig does not work with callable filters https://bugs.python.org/issue41906 opened by raybb #41907: Regression in IntFlag behaviour in f-string https://bugs.python.org/issue41907 opened by rt121212121 #41908: Make IDLE Start Menu entry more descriptive https://bugs.python.org/issue41908 opened by terry.reedy #41909: Segfault on __getattr__ infinite recursion on certain attribut https://bugs.python.org/issue41909 opened by bluetech #41910: Document that object.__eq__ implements `a is b` https://bugs.python.org/issue41910 opened by brett.cannon #41911: Language reference incorrectly says comparison expressions ret https://bugs.python.org/issue41911 opened by brett.cannon #41912: Long generator chain causes segmentation fault https://bugs.python.org/issue41912 opened by karzes Most recent 15 issues with no replies (15) ========================================== #41912: Long generator chain causes segmentation fault https://bugs.python.org/issue41912 #41911: Language reference incorrectly says comparison expressions ret https://bugs.python.org/issue41911 #41910: Document that object.__eq__ implements `a is b` https://bugs.python.org/issue41910 #41909: Segfault on __getattr__ infinite recursion on certain attribut https://bugs.python.org/issue41909 #41900: XML C14N serialisation fails with default namespace https://bugs.python.org/issue41900 #41898: Any logging causes assertLogs to pass https://bugs.python.org/issue41898 #41894: UnicodeDecodeError during load failure in non-UTF-8 locale https://bugs.python.org/issue41894 #41891: asyncio.wait_for does not wait for task/future to be completed https://bugs.python.org/issue41891 #41886: PyType_Type is documented incorrectly https://bugs.python.org/issue41886 #41882: CCompiler.has_function does not delete temporary files https://bugs.python.org/issue41882 #41880: Get Python include directories from sysconfigdata https://bugs.python.org/issue41880 #41879: Outdated description of async iterables in documentation of as https://bugs.python.org/issue41879 #41869: Multiprocessing Manager Object Cleanup https://bugs.python.org/issue41869 #41868: SMTPLIB integrate or provide option to use "logging" https://bugs.python.org/issue41868 #41867: Include options for timespec in docstrings of isoformat https://bugs.python.org/issue41867 Most recent 15 issues waiting for review (15) ============================================= #41907: Regression in IntFlag behaviour in f-string https://bugs.python.org/issue41907 #41905: add update_abstractmethods function to update an ABC's abstrac https://bugs.python.org/issue41905 #41902: Micro optimization for range.index if step is 1 https://bugs.python.org/issue41902 #41900: XML C14N serialisation fails with default namespace https://bugs.python.org/issue41900 #41894: UnicodeDecodeError during load failure in non-UTF-8 locale https://bugs.python.org/issue41894 #41892: use both "for in" and "ElementTree.remove" has a index bug https://bugs.python.org/issue41892 #41891: asyncio.wait_for does not wait for task/future to be completed https://bugs.python.org/issue41891 #41889: enum: Mixin and int base class regression in 3.8.6 https://bugs.python.org/issue41889 #41887: ast.literal_eval does not accept strings with leading whitespa https://bugs.python.org/issue41887 #41886: PyType_Type is documented incorrectly https://bugs.python.org/issue41886 #41882: CCompiler.has_function does not delete temporary files https://bugs.python.org/issue41882 #41880: Get Python include directories from sysconfigdata https://bugs.python.org/issue41880 #41876: Add __repr__ for Tkinter Font objects https://bugs.python.org/issue41876 #41875: __builtin_unreachable error in gcc 4.4.5 https://bugs.python.org/issue41875 #41867: Include options for timespec in docstrings of isoformat https://bugs.python.org/issue41867 Top 10 most discussed issues (10) ================================= #41902: Micro optimization for range.index if step is 1 https://bugs.python.org/issue41902 15 msgs #41892: use both "for in" and "ElementTree.remove" has a index bug https://bugs.python.org/issue41892 13 msgs #41851: tkinter: add font equal methods https://bugs.python.org/issue41851 6 msgs #41774: While removing element from list using for and remove(), which https://bugs.python.org/issue41774 5 msgs #41875: __builtin_unreachable error in gcc 4.4.5 https://bugs.python.org/issue41875 5 msgs #41905: add update_abstractmethods function to update an ABC's abstrac https://bugs.python.org/issue41905 5 msgs #41859: Uncaught ValueError https://bugs.python.org/issue41859 4 msgs #41861: Convert sqlite3 to PEP 384 https://bugs.python.org/issue41861 4 msgs #41883: ctypes pointee goes out of scope, then pointer in struct dangl https://bugs.python.org/issue41883 4 msgs #41903: PyNumber_InPlacePower ignores o3 if o1 implements __ipow__ https://bugs.python.org/issue41903 4 msgs Issues closed (39) ================== #13790: In str.format an incorrect error message for list, tuple, dict https://bugs.python.org/issue13790 closed by eric.smith #17490: Improve ast.literal_eval test suite coverage https://bugs.python.org/issue17490 closed by serhiy.storchaka #29127: Incorrect reference names in asyncio.subprocess documentation https://bugs.python.org/issue29127 closed by aeros #29893: create_subprocess_exec doc doesn't match software https://bugs.python.org/issue29893 closed by vstinner #29940: Add follow_wrapped=True option to help() https://bugs.python.org/issue29940 closed by taleinat #30953: Fatal python error when jumping into except clause https://bugs.python.org/issue30953 closed by ppperry #31256: xml.etree.ElementTree: add support for doctype in tostring met https://bugs.python.org/issue31256 closed by scoder #33867: Module dicts are wiped on module garbage collection https://bugs.python.org/issue33867 closed by serhiy.storchaka #35103: format_exception() doesn't work with PyErr_Fetch https://bugs.python.org/issue35103 closed by pablogsal #36255: Provide a simple way to delete and edit python's welcome messa https://bugs.python.org/issue36255 closed by rhettinger #38989: pip install selects 32 bit wheels for 64 bit python if vcvarsa https://bugs.python.org/issue38989 closed by steve.dower #39862: Why are the union relationships not implemented by default for https://bugs.python.org/issue39862 closed by rhettinger #40105: Updating zip comment doesn't truncate the zip file https://bugs.python.org/issue40105 closed by taleinat #41114: "TypeError: unhashable type" could often be more clear https://bugs.python.org/issue41114 closed by rhettinger #41295: CPython 3.8.4 regression on __setattr__ in multiinheritance wi https://bugs.python.org/issue41295 closed by scoder #41356: Convert bool.__new__ to argument clinic https://bugs.python.org/issue41356 closed by Dennis Sweeney #41613: get_type_hints regression for 3.9 and 3.10 https://bugs.python.org/issue41613 closed by BTaskaya #41692: Deprecate immortal interned strings: PyUnicode_InternImmortal( https://bugs.python.org/issue41692 closed by vstinner #41738: MIME type for *.zip becomes application/x-zip-compressed on Wi https://bugs.python.org/issue41738 closed by steve.dower #41802: Missing documentation for 'PyDict_DelItem' behavior https://bugs.python.org/issue41802 closed by serhiy.storchaka #41834: Remove _Py_CheckRecursionLimit variable https://bugs.python.org/issue41834 closed by vstinner #41842: Add codecs.unregister() to unregister a codec search function https://bugs.python.org/issue41842 closed by vstinner #41848: PEG parser doesn't allow lambda in for_if_clause https://bugs.python.org/issue41848 closed by gvanrossum #41858: Fix incomplete line on optparse documentation https://bugs.python.org/issue41858 closed by taleinat #41863: IDLE shell not opening with 3.5 on Mac Mohave. https://bugs.python.org/issue41863 closed by terry.reedy #41865: doc search for super().__init__() returns no useful results https://bugs.python.org/issue41865 closed by rhettinger #41870: Use PEP 590 vectorcall to speed up calls to bool() https://bugs.python.org/issue41870 closed by corona10 #41871: Add PyList_Remove() in listobject.c https://bugs.python.org/issue41871 closed by shihai1991 #41872: get_type_hints fails to resolve forward references in nested f https://bugs.python.org/issue41872 closed by gvanrossum #41873: Add vectorcall for float() https://bugs.python.org/issue41873 closed by corona10 #41874: datetime.strptime raises error if date is before 1900 https://bugs.python.org/issue41874 closed by redrooster #41881: New Line escape issue in raw strings https://bugs.python.org/issue41881 closed by ardath.kevin #41885: Unexpected behavior re.sub() with raw f-strings https://bugs.python.org/issue41885 closed by eric.smith #41888: Duplicate Opcode value in master and 3.9rc2 https://bugs.python.org/issue41888 closed by anthonypjshaw #41890: crypt produces wrong hashes for passwords containing newline c https://bugs.python.org/issue41890 closed by amarahzm #41893: remove() method is not working as expected(Hard to explain) https://bugs.python.org/issue41893 closed by eric.smith #41897: ast.parse in Python 3.9 does not produce SyntaxError for assig https://bugs.python.org/issue41897 closed by gvanrossum #41899: Poor example for Element.remove() https://bugs.python.org/issue41899 closed by scoder #41901: Added some explaining to pickle errors. https://bugs.python.org/issue41901 closed by Marco Sulla

1 0

Farewell, Python 3.5
by Larry Hastings 01 Oct '20

01 Oct '20

At last! Python 3.5 has now officially reached its end-of-life. Since there have been no checkins or PRs since I tagged 3.5.10, 3.5.10 will stand as the final release in the 3.5 series. As with a similar announcement I wrote about eighteen months ago, I know we can all look back fondly on Python 3.5. 3.5 added many new asynchronous I/O programming features, the "typing" module, and even a new operator ("@"). Plus many and varied quality-of-life improvements for the Python programmer, in both the language, the library, the core implementation, and even the installers. Python 3.5.0 was the best version of the best language at the time, and since then it's gotten even better! My thanks to all the members of the Python 3.5 release team. In alphabetical order: Georg Brandl Julian Palard Ned Deily Steve Dower Terry Reedy My thanks also to the Python infrastructure team. The end of Python 3.5 support also ends my tenure as a Python Release Manager. Congratulations, you survived me and my frequent mistakes! (Special shouts out to Ned and Benjamin for running around behind the scenes quietly cleaning up my messes--and not even telling me most of the time.) Rest assured that I leave you in /much/ better hands with the current crop of RMs: Ned, Łukasz, and Pablo. One amusing note. During my tenure as a Python release manager, I had to deal with /three/ different revision control systems. Although we'd switched CPython itself to Mercurial by the time 3.4 alpha 0 was released, there were still many supporting repositories still on Subversion. (I remember having to do Subversion branch merges as part of my 3.4 release work... what a pain.) And of course these days we're on Git (-hub). This straddling of three different workflows certainly complicated the lives of us Release Managers. So, my friends, please... make up your minds! ;-) It's been my honor to serve you, //arry/ p.s. As of today, every supported version of Python supports f-strings. The only remaining excuse for "we can't use f-strings" is no longer viable!

4 3