Alex Waygood

Yeah... if you look at e.g. https://github.com/pandas-dev/pandas/blob/59bb3f44a9bdd6d79704caab4d5c4e229945aefd/pandas/core/computation/ops.py#L522-L523 (one of the errors going away here), it's clear that mypy already knows that it has an object of type `function` there, and...

@hamdanal, might be of interest to you? :)

> hopefully with less type checker screaming than in previous attempts I think you'll get exactly the same amount of type-checker screaming. It's just that they'll be screaming about the...

> mypy 1.13 I suppose? [python/mypy#17476](https://github.com/python/mypy/pull/17476) has support and was already merged. I believe @hauntsaninja is planning on releasing 1.13 in the next few days, and it will contain _only_...

Lots of new type annotations were added in the latest release that we could possibly look at incorporating into typeshed. but it sounds like our stubs for this package will...

I'm also not sure this is an improvement if it requires cross-referencing loads of protocols from different files in order to figure out what members a subprotocol has

`runtests.py` is meant to be a convenience tool for contributors, but isn't meant to exhaustively run every check in our CI pipeline. While most of our tests can be run...

Hmm, that sounds strange. Can you be more specific? Which tests were complaining, and which type hint was it?

> On the other hand it seems that `FileIO.readinto` _can't_ return `None`: Doesn't it return `None` in this code path here? https://github.com/python/cpython/blob/2a5d1eb7073179a13159bce937afdbe240432e7d/Modules/_io/fileio.c#L678

It might be good to split this into two PRs: a first PR that makes changes like this, which are arguably no-brainers; we can merge this first PR quickly and...