Friday, August 17, 2012

Stop Testing! (but know when)

On a rare occasion, a feature that I'm scheduled to test is very different than what I expected. Generally, this happens when either:

  1. The feature was not documented fully because 'the team doesn't do that sort of thing'

  2. The feature was not documented because 'it is so minor, it's not worth documenting'

  3. The feature was implemented by someone that needs (ahem) oversight.

In any case, I have found myself in each of these conditions and am faced with a challenge. What to do?  
  1. First, Stop Running Your Planned Tests. Really, you aren't going to get anywhere.
  2. Start exploring what was implemented. You may find that it's your misunderstanding and not the developers.
  3. Keep good notes.
  4. Raise a Red Flag. Either you need to re-write your tests or the software needs to be re-written. Either way, this is going to affect the schedule.
  5. Re-test once this is cleared up.
All too often, I found that I have misunderstood the documentation, so taking the time to dig in will save lots of grief with false bug reports. You will also learn a bit about how you can make mistakes and can avoid them in the future.

Good Luck!