Nejc Habjan

Thanks @mark-e-kibbe, this sounds familiar, I think I've seen the same before somewhere else, so we might need to just add dict there as one of the possible types.

@TimKnight-DWP if it only affects an individual test feel free to also skip it and we deal with it separately. GitLab often has subtle breaking changes in the API that...

@TimKnight-DWP also if we are religiously asserting on something that is maybe not so relevant, we could change the tests a bit to really only check what matters. Especially with...

> I think it may also. be worth validating the response from Delete APIs such as: https://docs.gitlab.com/ee/api/users.html#user-deletion > > If GL returns a 204, we don't need to check that...

> Showing as failing because -0.04% tests 😿 That's ok @TimKnight-DWP I can override that (and fix the coverage in the code that is the root cause :sweat_smile:). I was...

> Left some comments and some nits. > > I'm not sure I like the changes to `wait_for_sidekiq` and also how it isn't called anymore in a lot of places...

Thanks @TimKnight-DWP! The new smoke test failures seem unrelated as they also appear in https://github.com/python-gitlab/python-gitlab/pull/2833, might be something new in the build dependencies that change how the package is built,...

> Thanks @TimKnight-DWP! The new smoke test failures seem unrelated as they also appear in #2833, might be something new in the build dependencies that change how the package is...

@1ace they're different objects because they're created from different endpoints, and don't necessarily have the same amount of details returned (or even the same attribute structure, as we've found out...

@1ace with #2784 merged, does this cover your use case along with the docs above?