Add 1s sleep to flaky modification-time-change test to avoid racing
I think this test sometimes fails because the file-modification-time being written and reflected asynchronously by the filesystem races with it being read back by the change test.
GitLab outage for software maintenance planned for all day December 18, 2025.
I think this test sometimes fails because the file-modification-time being written and reflected asynchronously by the filesystem races with it being read back by the change test.