Subject: Re: BugZilla - stale bugs
From: Jesper Skov (jskov@zoftcorp.dk)
Date: Wed Dec 12 2001 - 02:17:03 CST
On Tue, 2001-12-11 at 20:57, Paul Egli wrote:
> Quoting Jesper Skov <jskov@zoftcorp.dk>:
> > I hereby declare bugs that haven't been touched for 2+ months for stale.
>
> That might be a bit hasty, IMHO. Perhaps 2+ binary releases... (?)
A reporter's resposiveness should me measured in time, surely. Albeit,
I'll agree that we'd one at least one release after the bug was fixed.
>
>
> > In the future, I will close the following classes of stale bugs:
> >
> > o QA TO QUALIFY:
> > Apparently whoever filed the bug (and others) are not annoyed
> > by this bug anymore, and they forgot to close it.
> >
> > o OPEN/SUBMIT where someone has requested additional information
> > and not received it:
> > Submitter apparently don't care anymore - and without the extra
> > information, the bug is useless.
>
> This sounds good to me. Though, we might want an organized way of determining
> and keeping track of who has contacted who. Perhaps, simply, the bug's owner is
> responsible; if you want to investigate, take ownership. Once you've gathered
> more info, post your findings and re-assign the bug (or whatever is
> appropriate). I personally don't want to e-mail a user asking for more info, if
> they're already receiving such inquiries from other developers. Which reminds
> me...
>
> For bugs logged by users who are not active in abiword development, I think it
> is important to send an e-mail to the user asking specific questions, not just
> add a comment to the bug report in bugzilla saying you need more info. If a
> church secretary logs a bug, s/he might never visit bugzilla again, but would
> probably be delighted to reply to a friendly e-mail (not the ones bugzilla
> automatically sends out ;-).
I love that - but it's just not practical; it takes too much time. I
don't have an accurate number of the bug flow in Bugzilla, but I suspect
we process about 40-50 bugs a week. Spend an additional 5 minutes of
hand-holding on each, and we lose ~4 hours of developer time [which is
probably about the amount of time I can give AbiWord a week].
If we had some non-developer volunteers willing to step up and do the
work, I would be happy for things to work like that. But they don't...
> > For the latter class, I've considered adding a new CLOSED/NEEDINFO state
> > to bugzilla. Sam, do you know hard it would be to do so? Meanwhile, I'll
> > be using CLOSED/INVALID.
> >
> > Note that this does not mean that we ignore those bugs. If you find one
> > of your bugs closed as stale because you've been busy or out of time,
> > simply reopen it, adding the necessary information to make it a valid
> > bug again.
> >
> > It's simply a way to keep the BugZilla backlog from getting out of hand:
> > skimming the same old set of bugs when looking for stuff to do is very
> > tiring (and rather frustrating, actually). The developers' time is
> > limited - this is a way to forcibly shift some of the BugZilla triage to
> > the users who actually report and care about the bugs.
>
> This also sounds good to me. Anything else we can do to clean up the bug
> database would make it easier for a QA man like me to get back in on the
> project (e.g. fix such annoyances as the one Piotr mentioned in his e-mail.
> There seems to be lots of these quirks lingering in our version of bugzilla).
I'd like Sam to tell us - he should know the effort involved with an
update.
> > Jesper - believing it's time we cleaned up the mess
>
> I agree. I'm sorry I haven't been of much help in the bug database lately.
> Please let me know how I can be of help (meta-QA work on bugzilla, or whatever
> else).
When I pitch in on the bugs, the ones I have to leave alone are Windows
bugs, because I have no chance of testing/QAing them. If you want to
feel heroic, please dig in :)
Jesper
This archive was generated by hypermail 2b25 : Wed Dec 12 2001 - 02:17:06 CST