On Sun, Feb 25, 2007 at 08:36:21PM -0500, Roberto C. Sanchez wrote: > On Sun, Feb 25, 2007 at 07:27:29PM -0500, David Nusinow wrote: > > > > And what he's telling you, and what I'm telling you, is that it's a > > completely crap criterion for those of us who deal with massive packagesets > > like KDE. Simply replying to a bug won't get it fixed any sooner or > > decrease the impact it has on the user. In addition, it distracts us from > > doing what is potentially far more productive work. > > > I have to completely disagree here. When I started in earnest with the > effort to clean up the sasl package, I literally spent three twelve hour > days in a row doing nothing but bug triage. I really am not surprised > that people had not offered to help previously. There were bugs which > had been unanswered for months or years. And that was for a package > with only ~100 bugs. When we finally uploaded the new version, it > closed something like 38 bugs. > > IOW, having the packages bugs properly organized and triaged is critical > to knowing what you are fixing with each upload. IOW, know your distribution. KDE weights 12 or 1300 bugs I think, maybe a bit less, it was 1700 not so long time ago. I'd bet X.org, Gnome, OOO.org, fire^H^H^H^Hiceweasel, ... live the same nightmare. For the record, the libc had 330 bugs opened quite recently, we're down to 280 or so I guess, maybe a bit less if we exclude those who have been tagged wontfix. I'd bet most of them are unanswered, or at least, completely forgotten. I was previously beeing ironic, now I'm not anymore. Such a criterion sucks, and sucks hard. Most of the huge pieces of software around generate a load that is barely bearable for the teams that work on them if you just consider RC bugs, some importants ones, and packaging new upstreams. Sorry, but sasl, if widely used, remains a _small_ package, that covers a quite small amount of features. You can't take it as a general example. And btw, help for bug triaging for any of those kind of packages is vastly appreciated... But here is a newsflash: 100 bugs is fairly easy to reduce. the 5 or 600 bugs the KDE team has closed was a year of work. Yes a damn year, during which there has been many upstream releases (getting a release ready is a week of work, plus the RC that always come with them, so you can add another week on top of this one). So please explain me how a team that has had sometimes less of 2 to 3 _actually active_ members at a given time do manage to keep up with bugs as well ? I'll tell you: they just don't. Sorry to seem pissed, but well, I am. Your mail (and others with the same thoughts) are completely disconnected from reality. Totally. -- ·O· Pierre Habouzit ··O madcoder@debian.org OOO http://www.madism.org
Attachment:
pgpGS7uRwg5pA.pgp
Description: PGP signature