Joel software test

This reflects a joel software test misunderstanding of what testers are supposed to do, which is joel evaluate new code, find the good things, software find the bad things, and give positive and negative reinforcement to the developers. What do you think about the points?