Are you saying that it is irresponsible to test for regressions and to not document the ones you didn't find or that you think it is reasonable to expect regression tests for every possible regression?