BugZappers/Meetings/Minutes-2008-Oct-14

From FedoraProject

Jump to: navigation, search

Contents

Bug Triage Meeting :: 2008-10-14

Attendees

  • jlaska
  • ke4qqq
  • poelcat
  • jds2001

Last Week Followups

  • jds2001 reports that FTBS is resolved
  • poelcat reports that FEver is still in process
  • poelcat to update BugZapper's wiki with bullets on why bug triage is important to Fedora
    • we have great information about the how of bug triage, but not the why or so what
  • no update from comphappy on greasemonkey scripts

Triaging Package Reviews

  • open to considering in the future--right now can't keep up with existing work load

Next Meeting

  • Tuesday @ 14:00 UTC
  • #fedora-meeting on irc.freenode.net

IRC Transcript

poelcat who is here? 07:01
jlaska Hello! 07:03
poelcat jlaska: hi 07:04
poelcat ping ke4qqq: John5342 jds2001 07:04
* ke4qqq is here 07:05
jlaska poelcat: hey there 07:05
poelcat hi ke4qqq 07:05
-!- Irssi: #fedora-meeting: Total of 104 nicks [1 ops, 0 halfops, 0 voices, 103 normal] 07:05
ke4qqq hi poelcat 07:06
poelcat should we wait a minute or two for others? 07:06
poelcat ke4qqq: hi 07:06
poelcat ke4qqq: meet any potential triagers at the Ohio fest? 07:06
ke4qqq we talked to a few people that we might get involved in triaging, it's in my stack of follow up notes 07:07
ke4qqq I just haven't yet gotten through them yet.  :( 07:08
poelcat let's start with that 07:08
* jlaska loads up https://fedoraproject.org/wiki/BugZappers/Meetings/Minutes-2008-Oct-07 07:08
poelcat ke4qqq last week I think you suggested explaining to people the benefits of doing bug triage 07:08
ke4qqq yes - there is great information about the HOW, but not the WHY or SO WHAT 07:09
poelcat ke4qqq: do you think a simple FAQish page like that would help? 07:10
ke4qqq I think so - or even a small concise blurb in the join section 07:10
poelcat ke4qqq: which page? 07:11
* poelcat will take action to try to make this better 07:11
poelcat are there any specific things you think I should call out? 07:11
ke4qqq if it were me it would be on /BugZappers or BugZappers/HelpWanted 07:11
* jds2001 here. apologies 07:12
poelcat ke4qqq: okay 07:13
poelcat hi jds2001 07:13
poelcat jds2001: https://fedoraproject.org/wiki/BugZappers/Meetings/Minutes-2008-Oct-07#FTBFS_.28Fails_To_Build_From_Source.29 07:13
ke4qqq I'd think I'd call out the huge expanse of the problem (# of tickets per week created or sometihng) and the benefits of triaging that to something more manageable. 07:13
poelcat followup? 07:13
poelcat ke4qqq: noted 07:14
jds2001 poelcat: mdomsch has modiifed his script 07:15
jds2001 to put bugs in assigned. 07:15
poelcat ke4qqq: i'm trying to get some graphs made related to Package Review bugs... not sure if you saw fedora-devel thread 07:16
poelcat and will consider that too 07:16
poelcat jds2001: TA-DONE! 07:16
poelcat jds2001: thanks 07:16
poelcat i followed up on https://fedoraproject.org/wiki/BugZappers/Meetings/Minutes-2008-Oct-07#FEver_Bugs 07:17
poelcat the person (name escapes me) wrote back for clarification 07:17
poelcat i need to follow up again to find out if things have gone live, etc. 07:17
ke4qqq poelcat: I have not seen that thread, but honestly email is suffering for about the past week - I am way behind on a big chunk of list traffic. 07:17
poelcat ke4qqq: discussion started about doing 1 package reivew per fesco member per week to reduce the "backlog" of package reviews 07:18
poelcat i pointed out that there are 1,212 :-/ 07:18
poelcat this was a surprise to some people 07:18
poelcat i also suggested that we might be able to help 07:18
ke4qqq yeah, jds2001 was talking about that at OLF. 07:19
poelcat ideally though if we could collaborate together 07:19
jds2001 ke4qqq: any progress on yours? 07:19
poelcat IOW package reviewers help us triage 07:19
poelcat and we help triage Package Review bugs 07:19
jds2001 ke4qqq: afraid I might get smaked with a wet noodle tomorrow :) 07:19
ke4qqq jds2001: no, not yet - haven't even touched it since ..... 07:19
ke4qqq so essentially triage the package reviews - say by doing rpmlint type things? 07:20
ke4qqq and if it returns 0 errors it is in queue that sponsors look at? 07:21
poelcat ke4qqq: i'm not sure.. jason T had created a draft page suggesting how we could help 07:21
poelcat though frankly i'm reluctant to start a new "front in the war against bugs" 07:21
* ke4qqq goes to look for the thread 07:21
jds2001 bpepple and I were discussing the the car Sunday night that there's a TON of the review process that's very tedious and could be automated. 07:21
poelcat when we are not gainging on the ~780 NEW !Package Review Rawhide bugs 07:22
jds2001 I think that the tedious part is why there aren't many reviews. 07:22
jds2001 tibbs|h: you around by chance? 07:22
* poelcat has a link that page somewhere, will try to dig it out 07:24
poelcat but still relluctant to start a new effort? 07:24
poelcat am I the only one that thinks that might be the wrong idea right now? 07:24
jds2001 at the moment, I don't think so :) 07:25
jds2001 i.e. I'm not sure we should start a new effort, we have a difficult time as is with the existing effort :) 07:26
poelcat anyone seen or heard from comphappy about GM script? 07:26
ke4qqq personally while the reviews are suffering I don't know that it's truly a bugzapper issue.... and it creates something that a new user prolly doesn't have the expertise to do 07:26
poelcat ke4qqq: the only "issue" we could help with would be to sort out the dead bugs 07:27
poelcat or I should say 07:27
poelcat the first place we could start helping 07:27
poelcat IOW wouldn't be nicer to know there are only 999 package reviews instead of 1200 :) 07:27
jds2001 and for that, we need to exclude Merge Reviews. 07:27
poelcat or close them if they are never going to happen! 07:28
* poelcat takes things slightly off topic 07:28
* jds2001 thinks they all need to happen...some day. 07:28
poelcat jds2001: do you think they should even be "Bugs" ? 07:28
jds2001 what other tool do we have for tracking them? 07:29
poelcat well we could put them in a separate product 07:29
poelcat so they aren't part of the "bug count" 07:29
poelcat but maybe there is no value there 07:29
poelcat people don't seem to be alarmed by +10,000 unresolved bugs :) 07:29
ke4qqq well you talked about automation jds2001. So maybe some tool that does that - though BZ is used for just about everything, so it may be a fight that will never be won 07:30
poelcat "when all you have is bugzilla, everything looks like a bug" :) 07:30
ke4qqq heh 07:31
poelcat so... anyone hear from comphappy on GM script? 07:31
poelcat i posted feedback to f-t-l 07:31
poelcat there being no status for GM... we have one more topic from last week which was "is it okay to triage your own bugs?" 07:33
poelcat having had a whole week to think about it what do folks think? 07:34
* poelcat says "it is okay" :) 07:34
poelcat if more people were triaging bugs then maybe not, but here we are 07:34
jds2001 i try not to triage my own, since they may be complete rubbish. 07:36
jds2001 especially in components that I'm not familiar with. 07:36
poelcat jds2001: that is a good distinction 07:37
ke4qqq true 07:37
poelcat i was thinking of really obvious selinux AVCs 07:37
poelcat or anaconda traceback 07:38
jds2001 SaveToBugzilla ftw! :) 07:38
jds2001 sorry, had to 07:38
pjones I don't think there's any way to get good triage without that, actually. 07:39
pjones effectively, that's what happens with the complex bugs even when we have dedicated triagers. 07:39
pjones oh, wait, you mean "your own" from a QA standpoint, not a package-owner standpoint. 07:39
ke4qqq pjones: yes 07:40
poelcat pjones: or "bug reporter" 07:41
poelcat anything else for this week or shall we wrap things up? 07:43
jds2001 i guess that means wrap things up :) 07:45
poelcat thanks for coming everyone! 07:46

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!