On Thursday, 16 September 2021 at 14:35:08 UTC, Paul Backus wrote:


In my experience, the only severity settings most people actually use when filing issues on Bugzilla are "enhancement", "normal", and "regression". And when people do use the other settings, there's no consistency to how they get applied. For example, the first two search results for priority "blocker", issues [22283][] and [22148][], have no indication of what (if anything) they block. Meanwhile, issues [14196][] and [13983][] are both enhancement requests but have their priority set to "major", and issue [22136][] is listed as "critical" even though it is actually a regression!

Yes of course. Rule #1:

Severity levels are not always accurately set when issues are first reported and may not have been updated since. The reviewer of a pull request that closes a Bugzilla issue will evaluate the issue’s severity level and may change it if he or she determines it is inaccurate. I will moderate any disagreements that may arise about severity levels.

Obviously, this isn't perfect, but it's the best we've got for the moment. I expect there will be quite a few kinks to work out as time goes by. The initial trial run will surely provide us with lessons we can apply going forward, and we will continue to refine the process as needed.

I don't blame anyone who files reports like these. The fact is, there is no official guidance anywhere about what distinguishes a "minor" issue from a "normal" one, or a "normal" issue from a "major" one, so people just guess. But treating the output of this guessing process as though it were meaningful data is probably a mistake.

They aren't meaningful because there has been no organized process to make them meaningful. Part of Razvan's job description is to oversee the Bugzilla issues, and this initiative is a product of that. But he's one man, and the issues are legion :-) I predict we'll see the publication of reviewer guidelines down the road, and will eventually ask for a handful of people to assist Razvan in making sure Bugzilla issues have roughly accurate severity levels before a PR is submitted. Those are the most obvious steps I see to improving accuracy.

Reply via email to