-
-
Notifications
You must be signed in to change notification settings - Fork 2.9k
mypy: remove subprocess samples #8838
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Conversation
They depend heavily on _subprocess, which is no longer a module that exists: python/typeshed#4022 (comment)
Seems like we should add in some newer samples. Current subprocess.py and test_subprocess.py don't type check off the bat, but I can twist them until they do |
I'm not sure that's necessary. I think these samples were added as a way to run mypy on a largish codebase when the project was getting started, but now mypy is run on huge codebases that are more useful to find regressions (like Dropbox's codebases). @JukkaL what do you think? |
I don't think that the stdlib samples are very useful any more. We can keep those that can be maintained without much effort, but non-trivial investments into them don't seem worth it. |
This is no longer valid code on Python 3.9, so py39 CI fails. We've talked about not investing in keeping these samples up to date: python#8838 (comment)
This is no longer valid code on Python 3.9, so py39 CI fails. We've talked about not investing in keeping these samples up to date: #8838 (comment) Co-authored-by: hauntsaninja <>
I don't know that these have caught any real issues in the last couple years, however, the standard library ones have been a bit of a pain, most recently in python#11905. Jukka was okay removing the stdlib samples two years ago, see python#8838 (comment) We now have tooling like mypy primer that runs mypy over a much larger and more varied corpus of code on every PR. The other samples haven't caused me much bother, but except for crawl2, they're all very simple. I don't think they provide us much value, unless they have sentimental or historical value.
I don't know that these have caught any real issues in the last couple years, however, the standard library ones have been a bit of a pain, most recently in #11905. Jukka was okay removing the stdlib samples two years ago, see #8838 (comment) We now have tooling like mypy primer that runs mypy over a much larger and more varied corpus of code on every PR. The other samples haven't caused me much bother (except for crawl, which I removed in #9375), but except for crawl2, they're all very simple. I don't think they provide us much value, unless they have sentimental or historical value.
I don't know that these have caught any real issues in the last couple years, however, the standard library ones have been a bit of a pain, most recently in python#11905. Jukka was okay removing the stdlib samples two years ago, see python#8838 (comment) We now have tooling like mypy primer that runs mypy over a much larger and more varied corpus of code on every PR. The other samples haven't caused me much bother (except for crawl, which I removed in python#9375), but except for crawl2, they're all very simple. I don't think they provide us much value, unless they have sentimental or historical value.
They depend heavily on _subprocess, which is no longer a module that
exists: python/typeshed#4022 (comment)