Stubtest failed on Sat Mar 05 2022

[github-actions]

Stubtest runs are listed here: https://github.com/python/typeshed/actions/workflows/stubtest.yml

[AlexWaygood]

Many, many failures.

[hauntsaninja]

I'm reverting checking positional-only differences on dunders in python/mypy#12293, as previously discussed in #7344 (comment)
There are other failures, but these are the majority and the busy-work-est of them.

This way we keep typeshed and stubtest master in a mostly reconcilable state and we can then ship improvements at our leisure.

[hauntsaninja]

I misdiagnosed a little bit, looks like some of the positional-only differences was something stubtest always reported, but interacts a little unfortunately with python/mypy#12203.

python/mypy#12294 changes things to ignore those as well.

Once typeshed CI is green, we can look into potentially reverting python/mypy#12293 and python/mypy#12294. In particular, I think one way forward is to not check for positional-only differences in the case where stub inherits the relevant dunder.

[AlexWaygood]

In particular, I think one way forward is to not check for positional-only differences in the case where stub inherits the relevant dunder.

I like this idea a lot :)

[hauntsaninja]

Green again! https://github.com/python/typeshed/actions/workflows/daily.yml