Leo Fang
Thanks, Ralf, yes now I recall, I guess it's a rediscovery then 😅 Maybe I just need to add a note to this page for better visibility? https://data-apis.org/array-api/latest/design_topics/data_interchange.html
> As far as I am aware there is no versioning for `__dlpack__` or `__dlpack_device__` and no way to move API/ABI forward without a dance using some kind of `try/except`....
Should this be v2021 or v2022? I see that v2021 was already tagged.
Also in `__init__.py` we need to import all functions to provide a flat namespace (except for the extension modules).
@kgryte Is there a way we can "backport" the fix #430 to v2021? It only fixed the `draft` version but not the `latest` (or 2021.12) version. Try comparing the outputs...
> A couple points that came up were whether to have a state or stateless API and NumPy's old and new APIs. I'd like to quote @rgommers from there (https://github.com/scikit-learn/scikit-learn/issues/22352#issuecomment-1126243581)...
If I understand Robert's post above correctly, you're trying to convince us this is one of the rare cases where API design is tightly coupled to the underlying algorithm. I'd...
Would be nice if we can discuss #425 again this week. It seems this and #427 are blocked.
So this PR should be merged after resolving #425.
Would be nice if we can discuss #425 again this week. It seems this and #446 are blocked.