On Sat, Jan 08, 2005 at 01:39:46AM +0100, Frans Pop wrote: > I've read most of the bug history. And I must say to the submitter: if you > think 7 months is bad, you're in for bigger disappointments. There are > bugs in the BTS that are over 3 years old (some even on issues not as > trivial as this one...).
I'm not talking about any kind of bug report. I'm talking about a submitted bug that has failed to be just tagged upstream and forwarded for almost 8 months now. > Although in principle bug triage _is_ the job of the package > maintainer(s), it is also their good right to prioritize their work, > *especially* as we are all volunteers. I guess no one expects the Debian KDE team to _solve_ the bug, and neither do I. "All" a Debian KDE maintainer should have done is to take a couple of minutes to actually tag the bug and forward it to upstream. By not doing so, the bug has lurked around without any processing from upstream (or anyone else) for months and months. This situation is just too silly, whether the team is busy or not. > IMO the responsibility of a maintainer for bug triage is bigger if > upstream is less accessible to users. For KDE this is clearly _not_ the > case. Upstream has a great BTS themselves, open to all and easy to find. > They even have a nice voting system that allows users to collectively > determine the priority of an issue. You can always tell the submitter should have reported it to upstream directly: so just prevent reportbug from sending upstream bugs to Debian, because otherwise the submitter is absolutely not supposed to know that the maintainers are completely buried in work. > I myself have over the past 5 years or so made the transition from "file > everything in the Debian BTS as that's my distribution" to "hey, where > does this problem come from and let's file upstream if more appropriate". > In the second case I may also submit a bug in Debian's BTS if I feel the > maintainers or other users should be aware of the problem. > Of course packaging related issues should always go to Debian's BTS, but > especially for wishlist and minor issues that are clearly upstream, > submitting directly to KDE's BTS is often the better solution as their > manpower (and ability to judge the severity and relevance of an issue) is > way greater than Debian's KDE team. Some ask why the submitter has not submitted the bug directly to upstream when noticing months after that nothing had happened. Once more, I regret: if the submitter gets no feedback, he'll consider that the upstream team had hard times chasing the bug (some bugs are tricky), certainly not that the Debian guys has left his report in the bottom of the barrel for months. Do you actually understand that? > A last suggestion to the submitter: you could have tried asking here how > best to proceed. Once more, the submitter is absolutely not supposed to know that he should ask how to proceed instead of just reporting the bug itself, like reportbug (and http://www.debian.org/Bugs/Reporting) kindly advise him to. And are you really aware of what the BTS would be if every submitter should ask how to proceed first? Come on, you must be joking, especially since the Debian KDE maintainers don't even seem to have time to complete elementary tasks (like just _processing_ this bug). How can you ever think that they would have the time to reply to submitters' questions asking how to proceed? (I like recursive jokes, but there's a limit) The submitter just does what Debian asks/allows him to do, and he has no way to know in advance if his bug will stay unprocessed (I didn't say solved) for ages or quickly forwarded and eventually solved. He's not supposed to know the guys who will deal with his request, whether they are overly busy or not, or (and that still what I'm thinking about this particular case... hope I have demonstrated it more convincingly) whether they have erratic scheduling practices or not, I'm sorry I have to say that. Oh, I forgot: it is _not_ the submitter's job to monitor every bug he filled, just because the submitter can absolutely not imagine that his report might well _not be processed at all_ for months, years, decad^Wok, trollinit aborted. He'll rather think "naively" that maintainers will do what's best during the week (yes, 2 minutes time, remember) and hope upstream will provide a fix some day (a delay is much more expectable here, as it is where the bottleneck may clearly reside, even if... not always, apparently! :-( ). "When Debian KDE processing times might well take way longer than upstream fixing ones". Nice title, but sad tale. Sorry if I'm becoming more and more sarcastic, but I feel that some are more concerned with finding any excuse for justifying the unjustifiable, rather than actually trying to analyze the reasons and change things so that this will not happen again. Seems like we all like discussions, but I hope you realize that all the cumulated time spent in this discussion by Debian KDE maintainers already represents at the very least 20 times more time than one tag/forward operation would require (ok, now I guess some guy will probably say that the bug submitter is the one to blame, wrong shot). So let's hope this time hasn't been completely wasted. I already gave one element in that sense, but do as you wish: Please process _every_ new bugs (but old ones may also require that as a start) within a week: - if the immediate action required for it clearly take less than a couple of minutes (like tag+forward), please just _do_ it right away! - if you feel this bug won't be able to be fixed by your team within more than 2 months (arbitrary value), please let at least the submitter know The rest, you'll deal with it the best you can, as everyone. Herve
