Otherwise, like with db1b484d98, the `nix
flake check` checks fail.
Same story as in ccb779782c for devShells.
Though here `nix flake check` is how I noticed the issue.
Two different PRs, two different issues.
The first PR, chronologically-merged-in, #426687 introduced a new
module, and while reviewers “reviewed” the `meta.maintainers` entry,
they did not notice the maintainer was not in the maintainers list file.
- https://github.com/NixOS/nixpkgs/pull/426687#discussion_r2311053832
Then, in #437310, merged a few days later, another new module was added
with M0ustach3 as a maintainer. At the very least, an attempt was made
in de66c73a19 to add the maintainer to the
list, but an unfortunate typo (or two?) in the maintainers list entry
means the two intended usage of the new entry still don't work.
The issue here is not from the contributors. The reviewers and
committers could have done better, but it's not even truly their fault.
The issue is the meta information in NixOS modules was never checked. In
any way AFAICT. Otherwise this would have been caught immediately, and
not four months later.
Anyway, the intent clearly is that the entry should matche the GitHub
username.
`__structuredAttrs = true` now lets us build sub-expressions in disabledTests
(which was necessary for skipping only one test that uses /tmp).
We can make the test skipping more explicit now by including the class name, and excluding
test functions that reuse the same name.
In the case of the TestTemplates tests, we were disabling all of them, so just disable the class in general.
test_init_cmake_project_* are covered by test_init_cmake_project being disabled. I was unaware that disabledTests
only looks for substrings, not complete matches.
See comment, but TLDR this is for backwards-compatibility.
(See 479628, where it failed to dismiss after fixing.)
We don't bother with `prepare.js` because it always errors
(and so should never be dismissed).
I have simply added the needed comments to each of `check-target-branch.js`'s
pre-existing reviews, because there are so few.
