Personally, I really don’t see distinctive differences between “ad-hoc (directed) and “Bug Bash” that would warrent me tracking these as separate data points in a defect tracking system. I bet if we asked testers what the difference between them is we would get a bunch of hand-waving and fanciful words thrown about by a few, a handful of others would stare at us like deer in headlights (thinking it is some sort of trick question perhaps), and (hopefully) most tester’s would say…”you know…that’s a really good question; I’ve never really thought about it.”

With the negative connotation associated with ad-hoc, personally I would probably pull that from the flavor list and suggest using Bug Bash or some other descriptive word/phrase to identify defects that are exposed outside the scope of more structured (although I am not sure structured is the right word here) testing approaches.

I think the how found category in defect tracking databases is often over-loaded and quite abused. I once saw a database in which anyone could edit the How Found category and there were three separate entries for Vulcan mind meld (Vulcan mind meld, Vulcan Mind-meld, and Vulcan Mindmeld), and of course before these entries was an entry for “a little birdie told me.”

Ultimately, the data points tracked in the defect database should provide meaningful information. If I was not going to use the information obtained in the How Found category to adust my processes or strategy, then perhaps I might just have 2 choices for that category: How Found == Testing and How Found == Guessing; and I might even consider removing one of those choices!

Wednesday, October 31, 2007 2:26 AM by I.M.Testy

Adam

Wednesday, October 31, 2007 1:33 AM by adamu