Re: PR/60496 CVS commit: src/tests/lib/libc/locale

[Taylor R Campbell <riastradh%NetBSD.org@localhost>]

> Date: Wed, 28 Oct 2026 17:35:20 +0000
> From: "Robert Elz" <kre%netbsd.org@localhost>
> 
> These new tests are currently expected to fail - but are not going to
> be marked as expected fails, I want to see them actually fail.

Please mark the tests as xfail.

Verifying that they actually fail is _exactly the point_ of marking
them as xfail:

0. The tests pass but are missing anything to trigger a certain bug.

1. You add a new test marked xfail.
   => The tests should still pass, verifying the bug is there.  (The
      new test will be counted among the expected failures rather than
      among the successes.)

2. You fix the bug and remove the xfail marker in a single changeset.
   => The tests should still pass, verifying that the fix changed the
      test from failing to passing.

If test detects the bug and the change fixes the bug as you intended,
the whole test suite will pass at every step of the way.

If you got anything wrong -- if the test failed to detect the bug, if
the changeset failed to fix the bug -- only _then_ will the test suite
fail.  That should be cause for alarm, and shouldn't be used for
routine verification of whether your change worked.