Having done loads of bug validations for Testuff 1.1, I wanted to say something about them. Oddly enough Mike Kelly has just published a post about validations too.

Bug validations can be considered quite a tedious task. Finding a bug is quite fun, a treasure hunt except you don’t get the luxury of a map with X marks the spot. Bug Validations however do include that X, but in most cases after digging the ground you won’t find a treasure, or in other words a bug. Yet, what if you took a few steps to the right and dug again? Looked underneath a rock? Shaken a tree? There might be treasure there.

I do agree that validations are somewhat important to make sure that bugs have actually been fixed, because from time to time the devs do mess up (see my previous post). However, instead of being bored out of your mind as an automatic validating machine, use validations as a jumping board to find new bugs:

  • The bug may have been fixed locally, but not globally. Some bugs may reproduce again in a different screen or slight different configuration. Ask yourself where else the bug may appear other than the specified screen and configuration mentioned in the bug report. Take a few minutes to test it.
  • Changing the code probably inserted new bugs into the system, unintended side effects. It is up to you, the validator, to find them. How? Keep your eyes open, especially around the bug. Do some exploratory testing in the area, make sure no additional unintended harm was done.
  • Invent and execute new tests. Use validations as an excuse to do additional testing, maybe even for a “stable” section of the software that hasn’t been tested for a while. There could be old bugs there, performance issues, overlooked security issues, silly spelling mistakes everyone has overlooked, or even room for enhancements. And all this without bearing direct relation to the fixed bug.

Happy bug hunting!