the utter failure of bugzilla (and us?)

[Marco Martin]

On Wednesday 25 May 2011, Aaron J. Seigo wrote:
> hi ...
> 
> so, congratulations everyone, plasma is #3 in bugs.kde.org with over 1300
> reports:
> 
> 	https://bugs.kde.org/weekly-bug-summary.cgi
> 
> trawling through some right now, it seems that probably ~5% are actual
> issues, 5-15% more are things that could be improved, and the rest is
> mostly duplicates, downstreams/upstream or flawed reports.
> 
> thing is, with 1300 reports, doing 1 per minute would take 21+ hours. and
> with bugzilla it is a LOT more than 1 minute per bug. probably closer to
> 10 for full inspection and resolution. which means 210 hours, at best.
> 
> that we got here is beyond me. i think the issues are obvious:
> 
> * we don't have a sufficient bug triage team
> * the user community is enabled to stuff any crap they wish into the
> database, making the signal-to-noise ratio absolutely horrid
> * bugzilla is slow
> * bugzilla has horrifically bad usability

yes, failure is on both side and i take full responsibility, because me (and 
many others i think) really trascured it in the last months (for us, having so 
much more eggs on the basket than the past doesn't help for sure).

now, we can go over them and have a very relaxed policy to just discards bugs 
that seem to not make sense, think-is-duplicate-but-can-t quite find-of-what, 
bugs that are opinions on usability/prettyness/whtever...

we could manage for the 4.7 release the most recent 100-something probably.

now, bugzilla is just drop dead unusable for sure..

what about trying to kickstart  with an installation of another bugtracker 
just for plasma and see how it goes to migrate the rest of kde?


fromm a quick google found, of the opensource ones:
http://webissues.mimec.org/
trac (that does everything but everything in a not so well fashion?)
http://www.mantisbt.org/ (only experience i had with it, is the cmake bug 
tracker, has been more horrible than bugzilla)
http://www.thebuggenie.com/ 


Cheers,
Marco Martin