[issue39273] ncurses does not include BUTTON5_* constants

New submission from Michael Yoo : Hi, Recently I was working with ncurses, and when handling the mouse scroll events, I noticed that the curses library does not include the BUTTON5_* macros provided by ncurses. On my system, BUTTON5 corresponds to the mouse down event. Is there a reason for

[issue39271] Dead assignment in pattern_subx

Serhiy Storchaka added the comment: Thank you Alex. -- resolution: -> fixed stage: patch review -> resolved status: open -> closed ___ Python tracker ___

[issue39271] Dead assignment in pattern_subx

Serhiy Storchaka added the comment: New changeset 1a183faccbe5c32c367dbced721a25c1444dc5c1 by Serhiy Storchaka (Alex Henrie) in branch 'master': bpo-39271: Remove dead assignment from pattern_subx (GH-17915) https://github.com/python/cpython/commit/1a183faccbe5c32c367dbced721a25c1444dc5c1

[issue39272] Dead assignment in _ssl__SSLContext_load_verify_locations_impl

Change by Alex Henrie : -- keywords: +patch pull_requests: +17328 stage: -> patch review pull_request: https://github.com/python/cpython/pull/17916 ___ Python tracker ___

[issue39272] Dead assignment in _ssl__SSLContext_load_verify_locations_impl

New submission from Alex Henrie : The function _ssl__SSLContext_load_verify_locations_impl currently contains the following code: if (r != 1) { ok = 0; if (errno != 0) { ERR_clear_error(); PyErr_SetFromErrno(PyExc_OSError);

[issue39271] Dead assignment in pattern_subx

Change by Alex Henrie : -- keywords: +patch pull_requests: +17327 stage: -> patch review pull_request: https://github.com/python/cpython/pull/17915 ___ Python tracker ___

[issue39271] Dead assignment in pattern_subx

New submission from Alex Henrie : The function pattern_subx currently sets the variable b to charsize, but that variable is reset to STATE_OFFSET(, state.start) before it is ever used. -- components: Regular Expressions messages: 359653 nosy: alex.henrie, ezio.melotti, mrabarnett

[issue39270] Dead assignment in config_init_module_search_paths

Change by Alex Henrie : -- keywords: +patch pull_requests: +17326 stage: -> patch review pull_request: https://github.com/python/cpython/pull/17914 ___ Python tracker ___

[issue39270] Dead assignment in config_init_module_search_paths

New submission from Alex Henrie : config_init_module_search_paths currently has the following code: const wchar_t *p = sys_path; while (1) { p = wcschr(sys_path, delim); The first assignment to p is unnecessary because it is immediately overwritten. Victor Stinner suggested

[issue39262] Unused error message in _sharedexception_bind

miss-islington added the comment: New changeset 45e5750a013291c5729e0ebad2b9e340fdffbd36 by Miss Islington (bot) in branch '3.8': closes bpo-39262: Use specific out-of-memory message in _sharedexception_bind. (GH-17908)

[issue39150] See if PyToken_OneChar would be faster as a lookup table

Andy Lester added the comment: I tried out some experimenting with the lookup table vs. the switch statement. The relevant diff (not including the patches to the code generator) is: --- Parser/token.c +++ Parser/token.c @@ -77,31 +77,36 @@ int PyToken_OneChar(int c1) { -switch (c1) {

[issue39262] Unused error message in _sharedexception_bind

Benjamin Peterson added the comment: New changeset 5cae042f686cc174e00093944dc118914c874b7c by Benjamin Peterson (Alex Henrie) in branch 'master': closes bpo-39262: Use specific out-of-memory message in _sharedexception_bind. (GH-17908)

[issue39262] Unused error message in _sharedexception_bind

Change by miss-islington : -- pull_requests: +17325 pull_request: https://github.com/python/cpython/pull/17913 ___ Python tracker ___

[issue39261] Dead assignment in pyinit_config

Benjamin Peterson added the comment: New changeset 2c7ed417a4c758f1c3f97fcbca70a49f79e58c07 by Benjamin Peterson (Alex Henrie) in branch 'master': closes bpo-39261: Remove dead assignment from pyinit_config. (GH-17907)

[issue39260] distutils.spawn: find_executable() Fails To Find Many Executables on Windows

Thomas Passin added the comment: I came across it while working on some code running in the Leo editor. I had no idea it was not really public; I just assumed that someone knew something I didn't. Then I discovered that it couldn't find some files I thought it clearly should, and wrote

[issue39259] poplib.POP3/POP3_SSL should reject timeout = 0 (non-blocking mode)

Dong-hee Na added the comment: ftplib: https://github.com/python/cpython/blob/efa3b51fd060352cc6220b27a1026e4d4d5401bd/Lib/ftplib.py#L155 nntplib: https://github.com/python/cpython/blob/2e6a8efa837410327b593dc83c57492253b1201e/Lib/nntplib.py#L1049 smtplib:

[issue39269] Descriptor how-to guide wanting update for 3.6+ features

Raymond Hettinger added the comment: Thanks for the reminder. I'll have time soon to work on this. Since the other tracker issue cover this, will close this one as a duplicate. -- resolution: -> duplicate stage: -> resolved status: open -> closed

[issue39259] poplib.POP3/POP3_SSL should reject timeout = 0 (non-blocking mode)

Change by Dong-hee Na : -- keywords: +patch pull_requests: +17323 stage: -> patch review pull_request: https://github.com/python/cpython/pull/17912 ___ Python tracker ___

[issue39260] distutils.spawn: find_executable() Fails To Find Many Executables on Windows

Éric Araujo added the comment: The problem is that distutils.spawn.find_executable is an internal API, never meant for general usage. Only minimal change happens to distutils these days, partly because the codebase is full of dark corners and weird interactions (so a fix can easily break

[issue39269] Descriptor how-to guide wanting update for 3.6+ features

Karthikeyan Singaravelan added the comment: Seems like a duplicate of https://bugs.python.org/issue34394 -- nosy: +xtreak ___ Python tracker ___

[issue39269] Descriptor how-to guide wanting update for 3.6+ features

Change by Karthikeyan Singaravelan : -- nosy: +rhettinger ___ Python tracker ___ ___ Python-bugs-list mailing list Unsubscribe:

[issue39269] Descriptor how-to guide wanting update for 3.6+ features

New submission from wim glenn : https://docs.python.org/3/howto/descriptor.html Current descriptor how-to guide, above, has no mention about API features added since Python 3.6 (see __set_name__ in PEP 487) It's an important and useful piece of using descriptors effectively and the guide

[issue28617] Why isn't "in" called a comparison operation?

Change by wim glenn : -- resolution: -> fixed stage: patch review -> resolved status: open -> closed ___ Python tracker ___ ___

[issue36761] Extended slice assignment + iterable unpacking

Change by wim glenn : -- stage: -> resolved status: open -> closed ___ Python tracker ___ ___ Python-bugs-list mailing list

[issue39264] Fix UserDict.get to account for __missing__

Bar Harel added the comment: See also bpo-39267 -- ___ Python tracker ___ ___ Python-bugs-list mailing list Unsubscribe:

[issue39267] Fix dict's __missing__ documentation

Change by Bar Harel : -- keywords: +patch pull_requests: +17322 stage: -> patch review pull_request: https://github.com/python/cpython/pull/17911 ___ Python tracker ___

[issue39268] test_asyncio: test_create_server_ssl_verified() failed on AMD64 FreeBSD Non-Debug 3.x

New submission from STINNER Victor : AMD64 FreeBSD Non-Debug 3.x: https://buildbot.python.org/all/#/builders/214/builds/123 == ERROR: test_create_server_ssl_verified (test.test_asyncio.test_events.SelectEventLoopTests)

[issue39267] Fix dict's __missing__ documentation

New submission from Bar Harel : Continuing bpo-39264, and according to the mailing list discussion at Python-Dev. Fixing dict's __missing__ documentation. Clarify .get() does not call __missing__, and move __missing__ from the data model to dict's section as it's not a general object or ABC

[issue39266] [2.7] test_bsddb3 leaked [1, 1, 1] file descriptors on AMD64 RHEL7 Refleaks 2.7

New submission from STINNER Victor : AMD64 RHEL7 Refleaks 2.7: https://buildbot.python.org/all/#/builders/51/builds/13 test_bsddb3 leaked [1, 1, 1] file descriptors, sum=3 -- components: Tests messages: 359638 nosy: vstinner priority: normal severity: normal status: open title: [2.7]

[issue39264] Fix UserDict.get to account for __missing__

Change by Bar Harel : -- keywords: +patch pull_requests: +17321 stage: -> patch review pull_request: https://github.com/python/cpython/pull/17910 ___ Python tracker ___

[issue38338] [2.7] test_ssl fails on RHEL8

STINNER Victor added the comment: I marked bpo-39265 as a duplicate. -- ___ Python tracker ___ ___ Python-bugs-list mailing list

[issue38338] [2.7] test_ssl: test_protocol_sslv23() and test_protocol_tlsv1_1() fail on RHEL8

Change by STINNER Victor : -- title: [2.7] test_ssl fails on RHEL8 -> [2.7] test_ssl: test_protocol_sslv23() and test_protocol_tlsv1_1() fail on RHEL8 ___ Python tracker ___

[issue39265] test_ssl failed on AMD64 RHEL8 Refleaks 2.7

STINNER Victor added the comment: Ok, I even already reported this issue as bpo-38338 :-) -- resolution: -> duplicate stage: -> resolved status: open -> closed superseder: -> [2.7] test_ssl fails on RHEL8 ___ Python tracker

[issue39265] test_ssl failed on AMD64 RHEL8 Refleaks 2.7

STINNER Victor added the comment: Note: it's not a regression, AMD64 RHEL8 Refleaks 2.7 is failing since build #1. Logs of oldest builds are no longer available, but I can see that test_ssl was already failing at build #10 (22 days ago, commit 5f2c1345a79f205c680ed6e0a6ed44199546d79e).

[issue39265] test_ssl failed on AMD64 RHEL8 Refleaks 2.7

Change by STINNER Victor : -- nosy: +cstratak ___ Python tracker ___ ___ Python-bugs-list mailing list Unsubscribe:

[issue39265] test_ssl failed on AMD64 RHEL8 Refleaks 2.7

New submission from STINNER Victor : AMD64 RHEL8 Refleaks 2.7: https://buildbot.python.org/all/#/builders/102/builds/45 test_protocol_sslv23 (test.test_ssl.ThreadedTests) Connecting to an SSLv23 server with various client options ... Could not scan /etc/ssl/openssl.cnf for MinProtocol:

[issue39264] Fix UserDict.get to account for __missing__

New submission from Bar Harel : Unlike dict, UserDict.__missing__ is called on .get(). After a discussion on the Python-Dev mailing list, mimicking dict's behavior was chosen as a solution to the issue. -- components: Library (Lib) messages: 359633 nosy: bar.harel priority: normal

[issue16806] col_offset is -1 and lineno is wrong for multiline string expressions

Pablo Galindo Salgado added the comment: > commit 995d9b92979768125ced4da3a56f755bcdf80f6e introduced a regression: > bpo-37603: parsetok(): Assertion `(intptr_t)(int)(a - line_start) == (a - > line_start)' failed, when running get-pip.py. Fixed in https://bugs.python.org/issue39209

[issue16806] col_offset is -1 and lineno is wrong for multiline string expressions

Change by Lysandros Nikolaou : -- pull_requests: +17320 pull_request: https://github.com/python/cpython/pull/17582 ___ Python tracker ___

[issue39263] Windows Installer can't select TargetDir in UI?

Jameson Nash added the comment: Yes, I thought it was confusing that button is labeled "choose location and features", but it only let me change a few features and not location. -- ___ Python tracker

[issue39255] Windows and Unix run-time differences

Paul Moore added the comment: For me, I headed straight for "Sharing state between processes" and the "Shared memory" object. That's probably because I was reviewing someone else's code, rather than writing my own, but nevertheless when coding I do tend to dive straight for the section that

[issue39263] Windows Installer can't select TargetDir in UI?

Steve Dower added the comment: You may have already had a previous install that was being upgraded. In that case, you should uninstall the old one and then install it to a new location (or else you may end up with some subtle corruption). -- ___

[issue39263] Windows Installer can't select TargetDir in UI?

Steve Dower added the comment: You could have chosen the "Customize installation" button (see the screenshot at https://docs.python.org/3.8/using/windows.html#installation-steps). The description says "Choose location and features". -- ___ Python

[issue39263] Windows Installer can't select TargetDir in UI?

New submission from Jameson Nash : When running the installer on Windows, I wanted to put Python in an easily accessible path (C:\Python38 in my case), however, the GUI didn't seem to provide any way to change the path (from AppData). And additionally, the "install for all users" checkbox

[issue39251] outdated windows store links in WindowsApps folder

Steve Dower added the comment: For reference, I'm going to wait for protobuf, simplejson and Tornado to provide Python 3.8 wheels before switching the default version. Ideally I'd like markupsafe to have wheels as well, but they have a fallback so it's okay. Everything else in the top 100

[issue39129] Incorrect import of TimeoutError while creating happy eyeballs connection

Karthikeyan Singaravelan added the comment: Sorry, I never got around writing a proper test for this so I am moving it to test needed if someone wants to volunteer for it. trio has some test cases for their happy eyeball implementation if it helps :

[issue39229] library/functions.rst causes translated builds to fail

Rafael Fontenelle added the comment: Hey Julien! I already fixed once, now I fixed them again, but the message persists: make[1]: Entering directory '/home/rffontenelle/repos/python-docs-pt-br/venv/workdirs/cpython/Doc' mkdir -p build Building NEWS from Misc/NEWS.d with blurb

[issue38623] Python documentation should mention how to find site-packages

Brett Cannon added the comment: I think "Python learners deserve to know about "site-packages" and (optionally) "dist-packages" " is a bit strong of a statement. I don't think the tutorial covers how to install third-party package which is what goes into site-packages, so I don't know what

[issue35292] Make SimpleHTTPRequestHandler load mimetypes lazily

Steve Dower added the comment: Thanks for the patch! -- resolution: -> fixed stage: patch review -> resolved status: open -> closed versions: +Python 3.9 -Python 3.8 ___ Python tracker

[issue35292] Make SimpleHTTPRequestHandler load mimetypes lazily

Steve Dower added the comment: New changeset 5907e61a8d4da6d0f11bf1062d6d17484560a15e by Steve Dower (An Long) in branch 'master': bpo-35292: Avoid calling mimetypes.init when http.server is imported (GH-17822) https://github.com/python/cpython/commit/5907e61a8d4da6d0f11bf1062d6d17484560a15e

[issue39262] Unused error message in _sharedexception_bind

Change by Alex Henrie : -- keywords: +patch pull_requests: +17319 stage: -> patch review pull_request: https://github.com/python/cpython/pull/17908 ___ Python tracker ___

[issue39262] Unused error message in _sharedexception_bind

New submission from Alex Henrie : The function _sharedexception_bind currently has the following bit of code in two places: if (PyErr_ExceptionMatches(PyExc_MemoryError)) { failure = "out of memory copying exception type name"; } failure = "unable to encode and copy

[issue39261] Dead assignment in pyinit_config

Change by Alex Henrie : -- type: -> performance ___ Python tracker ___ ___ Python-bugs-list mailing list Unsubscribe:

[issue39261] Dead assignment in pyinit_config

Change by Alex Henrie : -- versions: +Python 3.9 ___ Python tracker ___ ___ Python-bugs-list mailing list Unsubscribe:

[issue39261] Dead assignment in pyinit_config

Change by Alex Henrie : -- keywords: +patch pull_requests: +17318 stage: -> patch review pull_request: https://github.com/python/cpython/pull/17907 ___ Python tracker ___

[issue39261] Dead assignment in pyinit_config

New submission from Alex Henrie : The function pyinit_config currently contains the following line: config = >interp->config; However, the config variable is not used after that point. Victor Stinner has confirmed that this assignment is unnecessary:

[issue39255] Windows and Unix run-time differences

Eryk Sun added the comment: > Agreed it's not a bug, but I will say it took me a while to work out > *why* it's not a bug (namely, that even though the OP is using shared > memory values, the code relies on fork semantics to share the two > Value objects that *reference* the shared memory).

[issue39260] distutils.spawn: find_executable() Fails To Find Many Executables on Windows

Thomas Passin added the comment: That's so, and with a bit more complexity might be worth doing. At least, it would be good to have *some* means to find more the the executable files one would expect to find. -- ___ Python tracker

[issue39237] Redundant call to round in delta_new

Alex Henrie added the comment: Thank you! -- ___ Python tracker ___ ___ Python-bugs-list mailing list Unsubscribe:

[issue39253] Running the test suite with --junit-xml and -R incorrectly reports refleaks

Pablo Galindo Salgado added the comment: > But I suggest to raise an error and tune the buildbot configuration to not > use it, rather than silently ignore the option. I already configured the buildbots to not use --junit when running with =R --

[issue39260] distutils.spawn: find_executable() Fails To Find Many Executables on Windows

Eryk Sun added the comment: > In that the best tactic would be to look in those top-level > directories for a directory with the same name as the executable. It may be a common pattern, but such a guess is not reliable. The needed executable may be in an unrelated directory at an arbitrary

[issue39255] Windows and Unix run-time differences

Paul Moore added the comment: Agreed it's not a bug, but I will say it took me a while to work out *why* it's not a bug (namely, that even though the OP is using shared memory values, the code relies on fork semantics to share the two Value objects that *reference* the shared memory). It

[issue39253] Running the test suite with --junit-xml and -R incorrectly reports refleaks

STINNER Victor added the comment: FYI test_io is skipped in Python 2.7 when using -R, because test_io is not reliable. Extract of Lib/test/test_io.py: if huntrleaks: # FIXME: bpo-31731: test_io hangs with --huntrleaks print("Warning: bpo-31731: test_io hangs with

[issue39253] Running the test suite with --junit-xml and -R incorrectly reports refleaks

Pablo Galindo Salgado added the comment: > We can stop using --junit-xml if we want. GitHub Actions doesn't support > displaying test results anyway, so once Azure Pipelines is switched off we'll > just be reading the console all the time anyway (unless Buildbot has a way of > rendering

[issue39253] Running the test suite with --junit-xml and -R incorrectly reports refleaks

Pablo Galindo Salgado added the comment: > deactivating the --junit I mean deactivating the --junit *if* is used together with -R -- ___ Python tracker ___

[issue39255] Windows and Unix run-time differences

Steve Dower added the comment: Agreed it's not a bug. The best we could do is display a warning that fork is not portable (won't work on macOS anymore either, IIRC) and you should at least verify that spawn behaves the same. -- ___ Python

[issue39253] Running the test suite with --junit-xml and -R incorrectly reports refleaks

Steve Dower added the comment: My guess would be that the reference tracking needs some kind of awareness for logging results. The results are collected as regular Python objects and formatted later, as we need to calculate totals before writing the XML file. I guess keeping those objects

[issue39243] CDLL __init__ no longer supports name being passed as None when the handle is not None

Steve Dower added the comment: > I think it would be cleaner to split out the platform-specific work into a > separate _load_library method SGTM -- ___ Python tracker ___

[issue39253] Running the test suite with --junit-xml and -R incorrectly reports refleaks

Steve Dower added the comment: We can stop using --junit-xml if we want. GitHub Actions doesn't support displaying test results anyway, so once Azure Pipelines is switched off we'll just be reading the console all the time anyway (unless Buildbot has a way of rendering test results that I

[issue39260] distutils.spawn: find_executable() Fails To Find Many Executables on Windows

Thomas Passin added the comment: Well, yes, I realize that I could add the directories. But random users of find_executible may not realize that. Actually, it's probably a bit more complex than just adding the top-level directories %ProgramFiles% and %ProgramFiles(x86)%, because

[issue28002] ast.unparse can't roundtrip some f-strings

Change by Batuhan : -- components: +Library (Lib) -Demos and Tools title: Some f-strings do not round trip through Tools/parser/test_unparse.py -> ast.unparse can't roundtrip some f-strings versions: +Python 3.9 -Python 3.7, Python 3.8 ___ Python

[issue28002] Some f-strings do not round trip through Tools/parser/test_unparse.py

Change by Batuhan : -- nosy: +BTaskaya ___ Python tracker ___ ___ Python-bugs-list mailing list Unsubscribe:

[issue39259] poplib.POP3/POP3_SSL should reject timeout = 0 (non-blocking mode)

Dong-hee Na added the comment: Sure, I will submit the PR by tomorrow :) -- ___ Python tracker ___ ___ Python-bugs-list mailing

[issue39260] find_executable() Fails To Find Many Executables on Windows

STINNER Victor added the comment: find_executable() rely on the PATH environment variable: def find_executable(executable, path=None): """Tries to find 'executable' in the directories listed in 'path'. A string listing directories separated by 'os.pathsep'; defaults to

[issue39260] distutils.spawn: find_executable() Fails To Find Many Executables on Windows

Change by STINNER Victor : -- title: find_executable() Fails To Find Many Executables on Windows -> distutils.spawn: find_executable() Fails To Find Many Executables on Windows ___ Python tracker

[issue39260] find_executable() Fails To Find Many Executables on Windows

New submission from Thomas Passin : On Windows, find_executable() in distutils.spawn may fail to find executables that it ought to. This is because the PATH environmental variable no longer includes %ProgramFiles% and %ProgramFiles(x86)%. At least, that is the case on my brand new Windows

[issue39242] Update news.gmane.org to news.gmane.io

miss-islington added the comment: New changeset b24e4fac03409c4f845758d7ed884c5a99368493 by Miss Islington (bot) in branch '3.8': bpo-39242: Updated the Gmane domain into news.gmane.io (GH-17903) https://github.com/python/cpython/commit/b24e4fac03409c4f845758d7ed884c5a99368493 --

[issue39242] Update news.gmane.org to news.gmane.io

miss-islington added the comment: New changeset 00ac28ac4d06a311fc2386ef921b2603735ffae6 by Miss Islington (bot) in branch '3.7': bpo-39242: Updated the Gmane domain into news.gmane.io (GH-17903) https://github.com/python/cpython/commit/00ac28ac4d06a311fc2386ef921b2603735ffae6 --

[issue39259] poplib.POP3/POP3_SSL should reject timeout = 0

STINNER Victor added the comment: I agree, I was the one who suggest to reject timeout=0 when you added the parameter to imaplib :-) Do you want to work on a PR? -- ___ Python tracker

[issue39259] poplib.POP3/POP3_SSL should reject timeout = 0 (non-blocking mode)

Change by STINNER Victor : -- title: poplib.POP3/POP3_SSL should reject timeout = 0 -> poplib.POP3/POP3_SSL should reject timeout = 0 (non-blocking mode) ___ Python tracker

[issue39242] Update news.gmane.org to news.gmane.io

STINNER Victor added the comment: Thanks Dong-hee Na for the update. The fix landed in master, backports to 3.7 and 3.8 will automatically follow once the CI test will pas. -- components: +Library (Lib), Tests resolution: -> fixed stage: patch review -> resolved status: open ->

[issue39242] Update news.gmane.org to news.gmane.io

STINNER Victor added the comment: New changeset 2e6a8efa837410327b593dc83c57492253b1201e by Victor Stinner (Dong-hee Na) in branch 'master': bpo-39242: Updated the Gmane domain into news.gmane.io (GH-17903) https://github.com/python/cpython/commit/2e6a8efa837410327b593dc83c57492253b1201e

[issue39242] Update news.gmane.org to news.gmane.io

Change by miss-islington : -- pull_requests: +17316 pull_request: https://github.com/python/cpython/pull/17904 ___ Python tracker ___

[issue39242] Update news.gmane.org to news.gmane.io

Change by miss-islington : -- pull_requests: +17317 pull_request: https://github.com/python/cpython/pull/17905 ___ Python tracker ___

[issue39259] poplib.POP3/POP3_SSL should reject timeout = 0

New submission from Dong-hee Na : Since poplib.POP3/POP3_SSL's implementation depends on socket.makefile, the client should reject if the timeout is zero. Because socket.makefile said that 'The socket must be in blocking mode' and if we set timeout to zero, the client does not operate as

[issue39242] Update news.gmane.org to news.gmane.io

Change by Dong-hee Na : -- keywords: +patch pull_requests: +17315 stage: -> patch review pull_request: https://github.com/python/cpython/pull/17903 ___ Python tracker ___

[issue34033] distutils is not reproducible

Petr Viktorin added the comment: > There is also one aspect where i586 builds end up with different .pyc files > than x86_64 builds. And then we randomly chose one of them for our "noarch" > python module packages and hope they work everywhere (including on arm and > s390 architectures).

[issue38780] SysLogHandler crash atexit

Alan Robertson added the comment: There are a variety of different reasons this can fail, not just on MacOS. You could give it a bad IP address of a server, etc. [That was my particular case]. The constructor should create an attribute 'socket' and initialize it to None early on. Then, the

[issue24554] GC should happen when a subinterpreter is destroyed

STINNER Victor added the comment: Py_EndInterpreter() now calls gc.collect() at least twice: at the end of _PyImport_Cleanup() and in finalize_interp_clear(). I now consider the issue as fixed and so I close it. The issue that I described in my previous comment can be enhanced/fixed later.

[issue6531] atexit_callfuncs() crashing within Py_Finalize() when using multiple interpreters.

Change by STINNER Victor : -- resolution: -> fixed ___ Python tracker ___ ___ Python-bugs-list mailing list Unsubscribe:

[issue39242] Update news.gmane.org to news.gmane.io

Dong-hee Na added the comment: @vstinner Oh! sure, I will submit the PR right a way :) -- ___ Python tracker ___ ___

[issue39233] glossary entry for parameter out-of-date for positional-only parameters

miss-islington added the comment: New changeset 9a669d58e8cb586fba38c84d5b631cd8a95d0c0c by Miss Islington (bot) (Pablo Galindo) in branch 'master': bpo-39233: Update positional-only section in the glossary (GH-17874)

[issue39233] glossary entry for parameter out-of-date for positional-only parameters

Change by miss-islington : -- pull_requests: +17314 pull_request: https://github.com/python/cpython/pull/17902 ___ Python tracker ___

[issue39231] Mistaken notion in tutorial

Mark Dickinson added the comment: [Robert] > It is not a "positional argument" but an "optional argument". I don't think I understand. Here the phrase "positional argument" in the docs is, I assume, referring to the parameter "ham: str" in the parameter list. So it's not optional; it's a

[issue39248] test_distutils fails on PPC64 Fedora 3.x: /tmp/subprocess.py

Change by STINNER Victor : -- title: test_distutils fails on PPC64 Fedora 3.x -> test_distutils fails on PPC64 Fedora 3.x: /tmp/subprocess.py ___ Python tracker ___

[issue39237] Redundant call to round in delta_new

STINNER Victor added the comment: New changeset 998c54948a29cf5bd8bfa49f973f1ce5855004a0 by Victor Stinner (Alex Henrie) in branch 'master': bpo-39237, datetime: Remove redundant call to round from delta_new (GH-17877)

[issue39237] Redundant call to round in delta_new

STINNER Victor added the comment: Thanks Alex Henrie, I merged your PR. I don't think that it's worth it to backport such cleanup change to stable branches. -- resolution: -> fixed stage: patch review -> resolved status: open -> closed ___ Python

[issue39241] Popen of python3.6 hangs on os.read(errpipe_read, 50000)

STINNER Victor added the comment: os.pipe() creates non-inheritable file descriptors by default, see PEP 446: $ python3.6 Python 3.6.9 (default, Jul 26 2019, 00:00:00) >>> import os >>> a,b=os.pipe() >>> os.get_inheritable(a) False >>> os.get_inheritable(b) False -- nosy: +vstinner

[issue39242] Update news.gmane.org to news.gmane.io

STINNER Victor added the comment: Correct: nntplib and test_nntplib should be updated: Doc/library/nntplib.rst:23: >>> s = nntplib.NNTP('news.gmane.org') Doc/library/nntplib.rst:47: >>> s = nntplib.NNTP('news.gmane.org') Doc/library/nntplib.rst:76:>>> with NNTP('news.gmane.org') as n:

[issue39256] Exception handler set by set_exception_handler is called only when I run coroutine by create_task

Andrew Svetlov added the comment: asyncio exception handler is supposed to catch *unhandled* exceptions only, not all raised ones. The first two cases from your example propagate raised exceptions to the caller, these exceptions are unwound in a regular manner. So, no need to call the

  1   2   >