Here's a not-very hypothetical situation for you: there are widespread reports of two problems that are somewhat related. One problem is detected and fixed, without attempting to detect the second problem.
At some later time, attempts are made to detect the second problem in order to fix it. However, no sign of the problem can be found.
Which of the two following explanations is more likely?
- The second problem is caused by some mysterious factor that is not present in your detection system, but which is otherwise widespread enough to generate so many reports.
- The two problems had the same cause, so fixing the first problem also fixed the second problem.
3 comments:
I find #2 to be true somewhat more often than #1, but I've seen plenty of both.
And you know what else I've seen plenty of?
#2 is more likely, up until an observed case of the problem showing up again. ;) Sometimes it's that the original fix makes the 2nd issue a lot rarer or harder to spot, but still present.
Heh, yeah. The extra information in this case was that it was an old bug that had been hard to reproduce in the first place, had never been explicitly fixed, but that hadn't been observed recently.
So #1 was somewhat more likely than normal - it had happened before! Sigh.
Post a Comment