Aaron Meurer

https://github.com/sympy/sympy-bot/pull/81

(continuing a discussion that was started elsewhere) I don't actually see what the issue with scalars is. They duck type as arrays, at least for all the purposes of the...

And just to be clear, the standard doesn't require all arrays from a given library to have the same type. This is even discussed explicitly https://data-apis.org/array-api/latest/design_topics/static_typing.html > Also note that...

Were there plans to fix this for NumPy 2.0? I think the fix according to the array API / NEP 56 amounts to just ```diff diff --git a/numpy/linalg/_linalg.py b/numpy/linalg/_linalg.py index...

This looks like it fixes the problem, although I'm not sure if the broadcast_to is necessary. Should we add tests?

I also completely agree that scalars are terrible for a dozen different reasons, and wish numpy just had 0-d arrays. But I guess this was too ambitious for NumPy 2.0....

It would be good to get a list of things that will definitely not work for any sparse implementation, so that we can add appropriate notes to the standard. There...

> Don't get me wrong I'm not trying to torpedo the attempt. But I also got burned enough times adopting a standard prematurely and spending unnecessary amount of work trying...

I admit I haven't followed it too closely, but I was a little surprised by your comment that scipy.sparse would be unlikely to support n-D or fill values. I had...

Monkeypatch or no, my suggestion would still be to just call it `asarray` (or `_asarray` if it really bothers you that it's a different signature than the usual one).