On Thu, 20 Apr 2006, Ulf Lamping wrote:
> Anders Broman wrote:
> > Hi,
> > Take http://bugs.ethereal.com/bugzilla/show_bug.cgi?id=481 as an example
> > last entry was 3 month ago. It's written on 0.10.12 which is quite old now.
> >
> > I don't think we have received other reports about this problem, how can it
> > ever be resolved without new input?
> >
> >
> That's the point.
>
> I've already closed some bugs where the reporter didn't respond for more
> than 2 months (my private threshold value for this :-), but using 1
> month seems to be enough. Usually, the reporter will respond within a
> week or never.
>
> Even if a bug is still in the code, it won't help anything/anyone to
> have a bug report about it, but none of the developers can reproduce it
> and the reporter don't provide any more information which could lead to
> some new ideas.
>
> The reporter can also reopen a bug if he wants to respond anytime later.
>
> Getting rid of "dead bugs" also makes it much easier to keep an eye on
> the remaining ones so I would vote for closing "unresponsive" bugs.
>
> BTW: The severity doesn't make any real difference in my eyes. If it's
> clear what to do but nobody finds the time to do it, just leave it open.
> If it's unclear what to do against it and the reporter (and anyone else)
> doesn't have a good idea for some time, just close it.
I think severity does make a difference, hence the different handling of
open bugs proposed.
If a bug report stays open a long time, it is difficult to fix, otherwise
it already would have been fixed. Now it the chance to gather (slowly
maybe, but still) information, maybe just the information we need to crack
the case. I think bug 481 is such a case, a lingering problem spanning
three releases now (watch new info coming in :).
I find 'datamining' for related info in the bugzilla database a 'pain',
and I wonder how many people trying to raise an issue go looking through
the closed bugs. I'm already glad if they scan the open bug list and
append their observations to the already filed bug report (count the
number of DUPS and you know what I mean).
So IMHO severe bugs deserve their time to ripe until we crack it. If it's
not critical we just have to say so, and adjust the severity accordingly.
Then normal 'bug aging rules' apply and will they be closed when their
'idle time' is up.
Thanx,
Jaap