On Fri, Aug 22, 2008 at 4:51 PM, Kevin Horn <[EMAIL PROTECTED]> wrote: > > > On Fri, Aug 22, 2008 at 4:51 PM, Christopher Arndt <[EMAIL PROTECTED]> > wrote: >> >> Hi Cliff, >> >> I was the person the person that closed this bug today. I was going >> through the ticket today in preparation for an anticipated TG 1.1 >> release and in the process tried to clear out some (!) of the really old >> tickets, that nobody had looked after apparently for years. >> >> I can understand your consternation that I decided that the problem was >> based on a user error. "user error" is actually probably the wrong term, >> I just wanted to convey that the problem can be solved with the correct >> application configuration. That this was not the case when the bug was >> reported is unfortunate, but I wasn't even really involved in TG 3 years >> ago. >> >> The experience of wading through all this old and obviously >> forgotten/neglected ticket certainly made it clear to me as well that >> something is amiss with our ticket handling procedures and I was >> thinking of writing my thoughts/suggestions on this topic to the mailing >> list soon anyway. What strikes me as odd is that you use the moment >> somebody is actually starting to do something about it and handles your >> ticket to condemn our practice. >> >> We still encourage user to enter bugs into the trac but we also >> encourage them (and we always did), to discuss problems on the mailing >> list. Also, the more information somebody provides for a bug report or >> patch, the more likely it is, that it will be looked after in timely >> fashion. I also closed about a dozen of tickets today, where the devs >> asked the reporter for more information, because the ticket had neither >> a proper problem description, information about how to reproduce the >> problem or any example code, and there had been no feedback for months >> or even years. I don't think it is useless to enter tickets into trac, >> the problem seems more that there are so many useless tickets. Which is >> our own fault: we should have more quality control and immediate >> feedback for these. >> A little point here, when I was active as trac admin from time to time (say every two weeks) I went thought all the tickets asking for more feedback, but I didn't dare to close many because I was too new on the project. I even made some ticket lists posted to tg-trunk asking for "adopt a ticket" and such.
>> One other problem is, I think, that most of the tickets at the moment >> don't get assigned to anybody by default, so nobody will be notified. A >> related problem is that the general ticket-notification Google group >> stopped working a year ago. Incidentally, just today I wrote to the >> group-owners to ask them if anything could be done about it. >> >> What we need, IMHO, is a ticket manager, who's sole job is to distribute >> tickets and look after them, i.e. see that they: >> >> 1) have a proper problem description and if not >> 2) get feedback from the reporter >> 3) get assigned to the right developer >> 4) they get addressed in some way* in a timely fashion by checking >> tickets periodically and nagging the one they are assigned to. >> 5) get closed when they are done, are invalid, won't fix or do not get >> feedback after some time. >> >> * by either fixing the problem or asking for feedback or giving and ETA >> or rejecting the ticket. >> >> Fact is, that we got ourselves into this mess with dozens (hundreds?) of >> open and often outdated tickets and we now need to find a way to handle >> this so that the ticket system becomes a really useful development tool >> again. This may also mean closing most of the old tickets >> indiscriminately, even if this will turn off some more people like you. >> >> Chris >> > > Chris, > > Something like your "ticket manager" approach has been tried in the past (at > least twice). > > Both times it was basically presented as having a small group of "trac > managers", who would go through tickets, and ask for feedback, etc. > The idea being that we could free up the time of the devs by doing these > "chores". > Of the people still obviously around on this list, myself and Jorge > (Vargas?) have both done this. > yes It was me. > Here are the problems we had (at least from my POV): > 1) We almost never had any feedback from original posters or even people who > had previously worked on a given ticket. > 2) It's tricky for anyone not developing on TG on a day to day basis to know > who the "right" dev is to assign a ticket to. They change frequently. > 3) The ticket updates list really needs to be working, which I see you have > begun addressing. > 4) It needs to be clear how much "authority" this/these person/persons would > have. After all, we don't want to get in the way. > 5) There's probably some others I have forgotten. > I pretty much agree on that. I remember a specific case of #4 where I updated a couple of tickets just to see my work get undone by a "more senior dev", and before anyone gets offended I agree I was the one wrong because I had the wrong impression of a milestone. > Having one person to do this task would improve several of these issues, > especially if the tickets were all initially assigned to him/her, though > with the current backlog, it might be a bit overwhelming at first. > I'm not really sure if this is the right approach, I remember that the "team of trac admins" ideas was because the "one admin" policy didn't work well. > I would be willing to give this a shot again (I believe I still have all the > necessary trac permissions, etc.), however, I am about to go on a weeks > vacation, so I wouldn't really be able to do much until after next Wed. or > so. > In fact I would really like to be involved in (at least) helping to clean up > the trac database, as its current state really annoys me :) > I'm currently back at being more involved with TG again, as the original posted suggests TG2 gave me a good push, and I'm also willing to take another try as trac-manager, if it's needed, which seems it is. > Kevin Horn -------------- That said I do agree with Chris, that trac tickets are not futile, only the more rare bugs get side tracked, we'll have to evaluate the state of the tickets but I drought , any real critical bugs get sidetracked. One thing you need to know is that when you are facing say a 100 tickets that need to be cleaned up, sometimes you just forget about the date. --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "TurboGears Trunk" group. To post to this group, send email to [email protected] To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/turbogears-trunk?hl=en -~----------~----~----~----~------~----~------~--~---
