This may be a stretch, but do you think there is any chance that the decline of formal bug tracking may be related to the rise of automated unit and integration

This may be a stretch, but do you think there is any chance that the decline of formal bug tracking may be related to the rise of automated unit and integration testing? Obviously, they are different things - I don't think that very good test coverage eliminates the usefulness of a bug database... but maybe it does make it easier to get away without one. Suppose you have a discussion forum, and someone sees a potential bug. In the past, they might add a bug report. Now, they go to their integration testing suite and write "URL shortener should provide valid menu links at navigation depth greater than three" (just to come up with a completely contrived example that never happens in real life, heh). This still tracks the bug, and is probably more useful, because now you can't ignore it because there's a red bar in everyone's dev environment. I do notice a lot more unit and integration testing these days, so it is possible that some of this is standing in for what used to be a bug database. I hope nobody interprets this as a claim that it replaces everything you get from a bug database, I'm just seeing it as something that might reduce the impact of the absence of a bug database.