Backport PR #12535: ENH: fix compatibility with numpy 1.23 - #12537

Oops, a mandatory check fails complaining this is opened against 5.0.x!

Gonna do the close/open trick. I applied a special label just for you.

No need to run the CI again... I'll just merge.

Oh wait, the original CI didn't run, did it. I'll just let it run and revisit tomorrow. 😅

Hmm, devdeps fails, but looking at it, it is some kind of spacing issue in docs, definitely unrelated to this PR.

p.s. For meeseeksdev, maybe add a note that one should add the skip-basebranch-check label? (You may already have done it...)

I did but it requires you to read the template carefully... I don't know where to draw the line between having the template be complete vs readable.

https://raw.githubusercontent.com/astropy/astropy/main/.github/PULL_REQUEST_TEMPLATE.md

<!-- READ THIS FOR MANUAL BACKPORT FROM A MAINTAINER:
Apply "skip-basebranch-check" label **before** you open the PR! -->

Yes, failures are unrelated. Merging. Thanks!

Not quite sure where exactly to suggest, but I'd do the following in the message posted where the backport fails (i.e.,
#12535 (comment)) -- if that is under your control!

    Create a PR against branch v5.0.x, I would have named this PR:

    "Backport PR #12535 on branch v5.0.x (ENH: fix compatibility with numpy 1.23)"

And apply the correct labels and milestones.

Here I would change the last sentence to

And apply the correct labels and milestones, including the `skip-basebranch-check label.

Alas, the bot isn't ours. The message is hardcoded at https://github.com/MeeseeksBox/MeeseeksDev/blob/a1c4a737d593ad776016a110b2c5ee655e30c9fd/meeseeksdev/meeseeksbox/commands.py#L756

Ah, that makes sense! Anyway, only a very small nitpick - I like the system quite a lot!