You know, running tests which run tests so the first tests can parse the second test's output and the first tests can then spit out their own output which is then parsed by the test framework you're testing in the first place is a pain in the tush. I'm just sayin'.
By writings tests for the tests of the tests?
