Jason R. Coombs

> we rely on pinned OpenPGP certificates If there were multiple certificates used to sign Setuptools releases, would that satisfy Arch, or does it require there to be a single...

> The expected output should **not** include... The issue I see with this expectation is that under other circumstances, a subfolder in a package is data. Consider if all you...

> 1. I don't know if it is accidental but after PEP 420 a folder ... will be "importable" It's intentional. During the design, we considered requiring a namespace package...

> I find importlib really hard to make sense of - the different types of finders & loaders, optional methods, deprecated methods... it's too much to keep in my head...

> Have you been able to confirm whether it actually fixes things for setuptools' test suite? Or can that only be verified once there's a release on PyPI? We should...

> That happens because the vendored version of `importlib_metadata` in setuptools relies on APIs that are implemented recently, isn't it? Yes. Probably. It probably implies that this project may need...

> It would appear #199 returns a stdlib (`importlib.metadata`) distribution which setuptools then attempts to operate on using the backport (`importlib_metadata`) and the two are incompatible on Python 3.8 and...

Oh, do we understand why this issue only affects Python 3.9 and earlier? If the underlying problem is a distribution with no `Name`, I'd expect that also to cause problems....

> @jaraco, I suspect that this parish aspect is an issue in setuptools because it uses entry-points to generate the `.egg-info` directory itself. So setuptools create an empty `.egg-info` folder,...

I unfortunately have paged out the rich context, but I agree that it sounds as if this narrow and surgical fix may be a very solid improvement. Let's go with...