Mailman 3 April 2019 - Python-Dev

PEP 589 discussion (TypedDict) happening at typing-sig@
by Jukka Lehtosalo April 15, 2019

April 15, 2019

Hi everyone, I submitted PEP 589 (TypedDict: Type Hints for Dictionaries with a Fixed Set of Keys) for discussion to typing-sig [1]. Here's an excerpt from the abstract of the PEP: PEP 484 defines the type Dict[K, V] for uniform dictionaries, where each value has the same type, and arbitrary key values are supported. It doesn't properly support the common pattern where the type of a dictionary value depends on the string value of the key. This PEP proposes a type constructor typing.TypedDict to support the use case where a dictionary object has a specific set of string keys, each with a value of a specific type. Jukka Lehtosalo [1] https://mail.python.org/mailman3/lists/typing-sig.python.org/

2 1

Cannot find documented API in PEP-376 (Database of Installed Python Distributions)
by Stefano Borini April 15, 2019

April 15, 2019

Hello, I am on a PEP scouting effort to check the current status of python packaging and its historical context, mostly for learning purposes. I noted that the PEP defines some functions for pkgutil (e.g. get_distributions), but I cannot find them. I tried to do some searching on the mailing list history, but I came up with pretty much nothing of value. It appears that the topic was last considered in 2009 (the year of the PEP). dist-info was then implemented, but I cannot find any information about the missing API, nor any additional PEP, except for a brief reference in PEP-427. Does anyone have some context for this? I understand it was 10 years ago, so it's mostly a curiosity. Thanks. -- Kind regards, Stefano Borini

2 1

No longer enable Py_TRACE_REFS by default in debug build
by Victor Stinner April 15, 2019

April 15, 2019

Hi, When Python is built in debug mode, PyObject gets 2 new fields: _ob_prev and _ob_next. These fields change the offset of following fields in the PyObject structure and so breaks the ABI. I propose to modify the debug build (Py_DEBUG) to no longer imply Py_TRACE_REFS. Antoine Pitrou proposed this idea when the C API was discussed to get a stable ABI. https://bugs.python.org/issue36465 https://github.com/python/cpython/pull/12615 This change makes the debug build ABI closer to the release build ABI, but I am not sure how to compare these two ABI. Technically, C extensions still need to be recompiled. What do you think? -- I also wrote a PR to remove all code related to Py_TRACE_REFS: https://github.com/python/cpython/pull/12614 I don't think that it's right approach. I prefer to keep this special build around to see if anyone needs it, and wait one or two Python releases to decide what to do with it. Victor -- Night gathers, and now my watch begins. It shall not end until my death.

10 33

Season of Docs
by Senthil Kumaran April 15, 2019

April 15, 2019

Hello Python Developers, Google is running a program called Season of Docs ( https://developers.google.com/season-of-docs/) to encourage technical writers to improve the documentation of Open Source Projects. As Python-Dev, and Python Software Foundation, do you think: a) We should participate? b) If yes to a), are you willing to be a mentor and identify project ideas? If you are willing to mentor and have project ideas, please let us know and we can think about the next steps. The deadline for org application is April 23, 2019. This discussion started here https://discuss.python.org/t/will-python-apply-for-season-of-docs-and-allow… Thank you, Senthil

2 1

Collaboration on a set of Python snaps
by Alan Pope April 15, 2019

April 15, 2019

Hi Python devs, I work on the Snapcraft [0] team at Canonical. I'm looking for a Python contributor to collaborate with us on making snaps of supported releases of Python available in the Snap Store [1]. Travis CI and Canonical are looking for someone (preferably North-America based) to participate in an in-person Snapcraft Summit in downtown Montreal, Canada from 11th to 13th June. We're sponsoring a number of software vendors, device manufacturers people from the robotics sector to come. We'd love someone from the Python project to join us. We’ve published this blog post to explain the event in more detail: https://snapcraft.io/blog/snapcraft-summit-montreal The goal would be to create snaps of the major supported releases of Python, and authoritatively publish them in the Snap Store. This would enable users of many different Linux distributions to easily obtain up to date supported versions of Python directly from the Python project. It also enables providers of CI systems (such as Travis) to the latest builds of Python are easily available to developers who use their services. We've done this previously with NodeJS [2] and Ruby [3] - among others. It would be great to have Python available via this method too. All the best, Al. [0] - https://snapcraft.io/ [1] - https://snapcraft.io/store [2] - https://snapcraft.io/node [3] - https://snapcraft.io/ruby -- Alan Pope Community Advocate Canonical - Ubuntu Engineering and Services +44 (0) 7973 620 164 alan.pope(a)canonical.com http://ubuntu.com/

1 0

checking "errno" for math operaton is safe to determine the error status?
by Xin, Peixing April 14, 2019

April 14, 2019

Hi, Math experts: Looking at the codes below, for many math operations, CPython is checking errno to determine the error status even though the math function returns normal value back. Is it a safe solution? From the description here http://man7.org/linux/man-pages/man3/errno.3.html and https://wiki.sei.cmu.edu/confluence/pages/viewpage.action?pageId=87152351, it looks apis probably set the errno when normal result is returned. Or being a side effect by calling other APIs in the implementation. In this situation, CPython's math operation might raise exceptions however in fact the result is correct. https://github.com/python/cpython/blob/master/Modules/mathmodule.c#L956 https://github.com/python/cpython/blob/master/Modules/mathmodule.c#L864 Thanks, Peixing

3 4

Fixing the ctypes implementation of the PEP3118 buffer interface
by Eric Wieser April 14, 2019

April 14, 2019

I've recently been adding better support to Numpy 1.16 for interoperability with ctypes. In doing so, I came across two bugs in the implementation of the PEP3118 buffer interface within ctypes, affecting `Structure`s and arrays. Rather than repeating the issue summaries here, I've linked their tracker issues below, and the patches I filed to fix them. * https://bugs.python.org/issue32782 (patch: https://github.com/python/cpython/pull/5576) * https://bugs.python.org/issue32780 (patch: https://github.com/python/cpython/pull/5561) I've seen little to no response on either the bug tracker or the github PRs regarding these, so at the recommendation of the "Lifecycle of a Pull Request" am emailing this list. Without these fixes, numpy has no choice but to ignore the broken buffer interface that ctypes provides, and instead try to parse the ctypes types manually. The sooner this makes a CPython release, the sooner numpy can remove those workarounds. Thanks, Eric

2 1

duplicate method names in tests
by Xavier de Gaye April 13, 2019

April 13, 2019

The last post [1] in issue bpo-16079 lists the methods in Lib/test that have duplicate names and that should be fixed. Xavier [1] https://bugs.python.org/issue16079#msg340168

1 0

Removing PID check from signal handler
by Jeroen Demeyer April 12, 2019

April 12, 2019

The signal handler (that receives signals from the OS) in Python starts with a check if (getpid() == main_pid) Looking at the comments, the intent was to do a check for the main *thread* but this is checking the *process* id. So this condition is basically always true. Therefore, I suggest to remove it in https://bugs.python.org/issue36601 If you have any objections or comments, I suggest to post them to that bpo. Jeroen.

3 2

Update PEP 394: Distributions can choose what does python command mean
by Miro Hrončok April 12, 2019

April 12, 2019

Hello. Based on discussions in [1], Petr Viktorin and me have drafted a new update [2] to the PEP 394 (The "python" Command on Unix-Like Systems). The update gives distributors the opportunity to decide where does the "python" command lead to, whether it is present etc. Please, see the PR [2] for the suggested changes. [1]: https://mail.python.org/pipermail/python-dev/2019-February/156272.html [2]: https://github.com/python/peps/pull/989 Thanks, -- Miro Hrončok -- Phone: +420777974800 IRC: mhroncok

4 3