one common pattern i've noticed is bug dependency. In some areas we need
to fix a series of 2-3 increasingly less trivial bugs to get enough test
exposure, tester confidence and developer attention to trigger (and fix)
the _truly_ bad bugs.
That's why agressive regression elimination (and prevention) is so
important IMHO - trivial regressions can totally block testing of
certain areas of code, and with an agressive 90 days release schedule
every day counts.
Ingo
--
