Alex Waygood

This makes sense to me, and we've already been following this for some stubs packages. But we'll inevitably have conversations about what it means for a project to be "unmaintained"...

Maybe we could explicitly acknowledge that it's a subjective call, in that case? "The package appears, according to the best judgement of the typeshed maintainers, to be unmaintained"?

Synthesis: we should try to use a wording that retains flexibility for us but is also explicit about specific situations where we know we might often use this new criterion

Interesting -- the docs pretty clearly imply that `errno` and `strerror` should always be (respectively) `int` and `str` for `OSError`: https://docs.python.org/3/library/exceptions.html#OSError. And there's nothing in the docs for `TimeoutError` to...

pre-commit.ci autofix

Looks like the recent TypedDict changes may have broken pyanalyze — cc. @JelleZijlstra

We should check how many become unused with the mypy master branch, as the way the lint works at mypy has just been completely reworked in https://github.com/python/mypy/pull/17392.

So yes, these are mypy's new complaints regarding overload-overlap if you use mypy `master`, fiddle with mypy's `master` branch so that you get rid of the thing where it stops...

I'm a pre-commit user, and I can confirm: - That using actionlint via pre-commit by default does not enable the shellcheck integration - That the changes made in this PR...

I agree with @carljm. This isn't a bug, and the the original suggestion of wrapping the `AttributeError` with a `RuntimeError` would have broken a _lot_ of code. But the suggestion...