the utter failure of bugzilla (and us?)

[Aaron J. Seigo]

On Monday, May 30, 2011 01:06:07 you wrote:
> Am Mittwoch 25 Mai 2011, 19:14:13 schrieb Aaron J. Seigo:

> > * 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
> > 
> > we ought to find a way to improve on this situation, but tackling one or
> > more of the above.
> > 
> > i don't have any answers right now, just frustrations.
> 
> I agree that using bugzilla is a pain.
> 
> Though I think that simply restarting would not work as the problems that
> lead to the pain have to be tackled first. 

agreed. the first two items on that list above have nothing to do with 
bugzilla, and that is precisely the reason.

the tool we are using does not help the situation, either, and we'll end up 
having to address both halves of the problem.
 
> * Voting: Only allow if every user has just one vote per bug or use a
> different system

voting is useless. no matter the system, it is too easy to skew the results by 
"promoting" your pet bug on dot.kde.org, planetkde.org, forums.kde.org, irc 
... all places we see that happening. 

"votes" in the form of duplicate reports is usually far more telling, ime.

> * closed tracker: Imo not the solution, especially if only people of a

i don't think we should have a closed tracker, but bugs.kde.org would benefit 
from much less freedom on the part of the reporting participants. for every 
time that someone comments on someone else's bug report in a way that produces 
useful input, there are 10 times where that same action is used for no good 
(and too often, even harm). marking bugs as duplicates is far more useful than 
open commenting for all. the same goes for re-opening reports: commenting on 
your own bug that it is still reproducable and having a triager/developer re-
open it would eliminate the overwhelming majority of re-openings that are 
utterly wrong and, again too often, done simply as a part of a tantrum.

there's also the reality that we get far more reports than we can handle. 
nobody wins with that. limiting the reporting traffic will help a lot. i got 
over 130 messages from bugzilla over the course of something around 1.5 days 
this past weekend. there is no way i can keep up with that, nor do i expect 
others can or will.

so while a closed tracker probably isn't a good answer, what we have now is 
also problematic because it is too open.

> * Karma system for reporters: To me this sounds quite good, though to make

yes, i quite like the idea of this as it allows us to document relationships 
that get built. it will also reward, and therefore encourage, productive 
behaviour rather than the alternative.

> this work I think there should be totally easy instructions on how to create
> good bug reports.

agreed; dr. konqi's steps do help, and could probably be improved on. also 
more of this kind of page would be very useful:

http://techbase.kde.org/Development/Tutorials/Debugging/How_to_create_useful_crash_reports

and they should be on community.kde.org where reporters are more likely to 
find them.

> Then there is a problem I faced myself.
> There are bugs which are totally valid yet it is _very_ hard or even
> impossible to fix them. Closing them would certainly not be the right way,
> yet marking them somehow, to not always have them listed with the rest
> would be nice.

well, that's what we have WONTFIX for. however, the users of bugs.kde.org have 
grown to feel that they are owed things they are not and too often throw 
tantrums more becoming of a 5 year old when that happens. this is the culture 
around bugs.kde.org that we must correct.

-- 
Aaron J. Seigo
humru othro a kohnu se
GPG Fingerprint: 8B8B 2209 0C6F 7C47 B1EA  EE75 D6B7 2EB1 A7F1 DB43

KDE core developer sponsored by Qt Development Frameworks
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 198 bytes
Desc: This is a digitally signed message part.
Url : http://mail.kde.org/pipermail/plasma-devel/attachments/20110530/85e68ab6/attachment.sig