Talley Lambert

yeah, I'm seeing this too: ```pytb cls = line = 'D:\\tmp\\.tox\\py310-windows-pyqt5\\lib\\site-packages\\pydantic\\typing.py:42:5: error: Cannot assign to a type [misc]' .... m = cls.OUTPUT_RE.match(line) if not m: > raise ValueError("Not a valid...

I'm also curious to hear what the "official" loguru stance is on this. From https://github.com/Delgan/loguru/issues/553, it looks like the basic sentiment is that: > Packages shouldn't touch the logger handlers,...

same here, https://github.com/tlambert03/pyrepo-cookiecutter/runs/6818457778?check_suite_focus=true

Yes.... but, I'm not sure I'm quite following your thought: are you contrasting that with the previous setup?

looks like there's a line length linting error error, but I matched the declarations below exactly (which has the same line length). If i fix it here, i think it...

> @tlambert03 how does this solution compare to [napari's .pyi files](https://github.com/napari/napari/blob/4b809a1ad8d5497081690824efb4ba69708b77b9/napari/__init__.pyi)? Which would you favour more generally? I'm struggling to come up with a strong opinion :joy:. I guess the...

> For me tbh I find the if TYPE_CHECKING block itself very distracting and ugly, which is why I'd rather have .pyi files, see alternative approach building the dynamic imports...

> You think that until you actually use .pyi files. Then the inevitable litany of desynchronization woes, DRY violations, and caching quandaries plague your house with a vicious cycle of...

yep thanks. it's exactly the same issue, but applied to `CO_GENERATOR` instead of `CO_COROUTINE`. I'll have a look around that issue and linked issues and see what can be done/make...

Just came to request this feature. Great to see that it's well underway! I had taken some screenshots of my sphinx apidocs to show, and perhaps they're still of use....