> This is what happens when we rewrite test cases en-masse. Regression tests
> stop protecting us from regressions.
+1, and the tests that have been rewritten are some of the most carefully written to avoid sensitive problems. We haven't had a tearing regression in quite a while (12+ months), so changing so much core testing and swapping code makes me nervous, especially because tearing is a very public regression that everyone tends to notice.
> This is what happens when we rewrite test cases en-masse. Regression tests
> stop protecting us from regressions.
+1, and the tests that have been rewritten are some of the most carefully written to avoid sensitive problems. We haven't had a tearing regression in quite a while (12+ months), so changing so much core testing and swapping code makes me nervous, especially because tearing is a very public regression that everyone tends to notice.