Alex Waygood
Alex Waygood
We now have 3.10.6 on Ubuntu and MacOS, but still not Windows, so this is deferred until we have it on Windows as well.
> We now have 3.10.6 on Ubuntu and MacOS, but still not Windows, so this is deferred until we have it on Windows as well. Seems like sometimes `setup-python` gives...
Looks like we're now getting 3.10.6 consistently on all platforms; I think this is now mergeable!
Thanks! Do any of our stubs packages that currently have `requirements-stubtest.txt` files specify `extra` dependencies at runtime? If so, can you update the stubs for those packages so that they...
According to [this](https://www.jetbrains.com/pycharm/whatsnew/) and [this](https://youtrack.jetbrains.com/issue/PY-49935), it seems PyCharm now supports `ParamSpec` 🎉
> I think we can now use ParamSpec without limitations in typeshed. There's still a mypy bug I'm waiting on: https://github.com/python/typeshed/issues/7689#issuecomment-1111721328. I'll add it to the tracker.
Whatever the case, we should probably delete this orphaned comment, which was there to explain why `__delattr__` was present in the stub but `__setattr__` was not: https://github.com/python/typeshed/blob/d7a5a147a0740816b837614c5f8fd429d0b8756e/stdlib/_decimal.pyi#L201-L202
`_threading_local.local.__delattr__` was also deleted in #8483. Unlike `_decimal.Context.__delattr__`, this actually _didn't_ have the same signature as `object.__delattr__`: while the `name` argument is positional-only in `object.__delattr__`, it is positional-or-keyword on `_threading_local.local.__delattr__`....
@erictraut, if we added `decimal.Context.__delattr__` back to the stub and gave it the signature `def __delattr__(self, __name: str) -> NoReturn` (to indicate that trying to delete any attributes on instances...
Thanks @erictraut! Let's leave `__delattr__` off `decimal.Context`, in that case (but improve the comment about why it's not there), and add it back to `_threading_local.local`, with a test case.