On 09/12/12 16:47, D'Arcy Cain wrote: > On Wed, 12 Sep 2012 16:37:11 -0400 > Terry Reedy <tjre...@udel.edu> wrote: >> assures one that the test is being run. (I don't always test first, but >> I once discovered a test not being run when I modified it in a way that >> should have made it fail, but it didn't.) > > 1. Write the test > 2. Run the test - make sure it fails > 3. write code until test passes > 4. Stop writing code > > People underestimate the importance of steps 2 and 4. Number 2 is > important because a test that can't fail is worse than useless. Number > 4 is important because it prevents code that doesn't have a > corresponding test. If you feel the need to write more code then go > back to step 1. > > And run all of your tests every day. You will sleep better at night.
Though I usually try to do test-driven development, I confess a fondness for Titus Brown's "Stupidity driven testing"[1] :-) -tkc [1] http://ivory.idyll.org/blog/stupidity-driven-testing.html (one of my most memorable pearls from PyCon'07) -- http://mail.python.org/mailman/listinfo/python-list
