I’ve been reviewing a unit test plan today. I should really be a simple document. Use this test framework. Write tests before writing code. Execute tests continually to verify good code.
Rinse and repeat until done. All tests will pass and after running a code coverage tool, justify any code not covered with tests.
But that apparently doesn’t match the approach this organization wants to use. They don’t want to run tests until code is ready for checkin. And they want all manner of metrics kept.
I find that this is a good group to work with, but the bureaucracy is toxic to me, at least.
Fix spelling. Add Technorati links.