Joachim Jablon

> First types.JSONValue is kinda misleading, as it has dict[str, t.Any] and list[t.Any], not dict[str, JSONValue] and list[JSONValue]. Not sure if it's intentional that JSON-ness is only guaranteed at the...

This somewhat perfectly captures my opinion of 2025 so far :) https://github.com/user-attachments/assets/cd354f60-2c78-4e70-88bd-9d16fe0a82c7 Joke apart, let me know if you need any help.

I understand you have to fix pressing issues, and I respect that. That said, I still think it makes sense, either for you or someone else, in due time, to...

Ok so: I've started working on this to get a better hang on the remaining issues: - Recursive JSON has been implemented, though indeed, accurate types mean we'll have to...

> If a custom encoder is set, just shorten the resulting args string as we can't know what the resulting format is (this must then be documented). I'm not sure...

This might be a solid lead.

There's so much changing on the way we handle everything with v3 that it will be worth checking if the issue still happens once we start releasing v3 beta.

closing in favor of https://github.com/procrastinate-org/procrastinate/issues/1198

Done by #1185

before https://github.com/py-cov-action/python-coverage-comment-action/pull/573, we used to generate the diff locally. But then we changed that because it was much more complex and it's often not trivial to get an accurate diff...