Justus Magin

the last round of commits should fix the final issue. I'm not at all sure my implementation is the most elegant it can be, but it at least it works...

> Looks like a typo: `assert xp_arr.nbytes == xp_arr.data.nbytes` This wouldn't work, because `xp_arr.data` does not define `nbytes`. So as far as I can tell, `numpy`'s answer is considered the...

but then this should be enough, no? ```python with pytest.raises(AttributeError): # xp arrays doesn't support nbytes, should fail: xp_arr.data.nbytes ``` We know `xp_arr.nbytes` doesn't raise since we just called it...

I opened data-apis/array-api#789 to discuss adding `nbytes` (in some form) to the array API.

> My idea was that `assert xp_arr.nbytes == xp_arr.data.nbytes` is what we really want to check. I've thought about this some more, and I agree that in general that would...

@shoyer, could I ask for another review? I _think_ I changed the code to what we discussed yesterday, but I might have missed something.

the remaining two errors on the upstream-dev CI are due to a bug in `numpy.isdtype` (numpy/numpy#26441). This appears to only have been an issue on `numpy`'s `main`, not their `2.0`...

looks like this should finally be ready for merging?

it does, but only parts of it (other fixes it deems unsafe so they require manual intervention). I don't have time to go through each of them this week, but...

we have a numpy 2 + python 3.9 compatible release now, so once the merge conflicts are fixed this should be good to go