I been having a disagreement on a bug, which brings up a philosophical question: what does it mean for a bug to be fixed? I was wondering where we stand on this since, IMO, we should have clear guidelines.
Some possibilities:
1. If the bug contains steps to reproduce, the bug is fixed if and only if those steps no longer reproduce the bug.
2. If the bug contains a patch, the bug is fixed if and only if that patch has been merged.
3. The bug is fixed if and only if the person who filed it agrees that it is fixed.
4. If the majority of comments in a bug discuss some problem, the bug is fixed if and only if that problem has been addressed, even if that problem is different from the one originally reported.
I would suggest the following:
- The person who filed the bug has final say on whether or not it has been fixed.
- In the absence of a definitive statement from the bug requester, the bug is fixed if and only if the steps to reproduce still reproduce the behavior.
- The steps to reproduce are the steps included in comment 0 along with any clarifications made by the person who filed the bug.
- Comments made by people other than the one who filed the bug can't change what the bug is about nor the criteria for considering it fixed.
So I ask you: is bug 234744 fixed or not? Is it a dupe or not?
It's still reproducible from the original steps and the person who filed it says it's not fixed. However, subsequent comments by other people discussed different (prerequisite) topics, which were either addressed by a patch in the bug or are a dupe of another bug.
Can people other than the bug requester change the focus of a bug, or should this be considered bug hijacking?
- Stefan
