Stop ignoring test outcome for Django 3.2

[adamchainz]

It has been released!

[carltongibson]

Looks good.

CI: 👀

[adamchainz]

There is still one failure from a warning raised in django-guardian. @smithdc1 already made a pre-emptive PR there back in October - django-guardian/django-guardian#721 - it's just not released yet.

We could ignore that one warning.

[carltongibson]

Ah, yes, the periodic Guardian is not updated shuffle™ — Yes, skip/ignore as needed.

[adamchainz]

It seems guardian is only installed in tests as an example of a permissions backend that supports object-level permissions. We could also rewrite the tests to not rely on it.

[carltongibson]

@adamchainz — I think if you have a burst of energy for that it would save this issue coming up… (it's not often but...)

[adamchainz]

I am done for the day here but will come back to this and evaluate my energy levels then.

[terencehonles]

I was wondering how Django 3.2 was being specified, and since you're touching this file does it make sense to also adjust the django32 specification from Django>=3.2a1,<4.0 to Django>=3.2,<4.0?

I would have commented directly on the line, but GitHub doesn't allow that.

[carltongibson]

@adamchainz It may not be worth the cycles to rewrite the tests here right now. For reference here's a previous iteration 524a28c (you can see it's not a frequent issue) Up to you. 🙂 🥇

[johnthagen]

@adamchainz @carltongibson A new django-guardian has been published: https://pypi.org/project/django-guardian/2.4.0/

[carltongibson]

Thanks @johnthagen

@adamchainz(-bot) rebase & merge 😀