From Fedora Project Wiki

Bug Triage Meeting :: 2009-02-24


  • comphappy
  • poelcat
  • tk009
  • John5342
  • mcepl
  • bigdufstuff
  • adamw
  • beland

Meeting Summary & Action Items

  • We are officially done setting goals for Fedora 11
    • Goal will be triaging bugs for key components
    • BugZappers/components
    • For Fedora 12 maybe we can create goals that are more measurable
  • mcepl will create queries and RSS feeds for key components to be triaged
  • comphappy will:
    • have metrics reporting working by this weekend
    • merge greasemonkey script created by mcepl to create triager signature into main Fedora triage greasemonkey script
    • make sure download link for greasemonkey script is working and accessible
  • adamw will arrange and coordinate the first bug triage day

IRC Transcript

Tk009_afk Hello, who's here? :) 07:01
beland Greetings. 8) 07:01
John5342 hi all 07:01
Tk009_afk Hello 07:01
Tk009_afk Feeling better john? 07:02
John5342 still rather drugged up but back home now 07:02
John5342 so tons better 07:02
Tk009_afk Well that is progress 07:02
Tk009_afk Looks like just us 07:03
John5342 dont seem to have many of us this week 07:04
Tk009_afk Hello big 07:04
bigdufstuff hi 07:04
* poelcat here 07:04
poelcat mcepl: jds2001 ping 07:05
Tk009_afk None of the people we need are here adamw, iarlyy, comphappy and mcelp 07:05
mcepl pong 07:05
poelcat who is leading the meeting? 07:05
Tk009_afk I am I think 07:05
poelcat Tk009_afk: thanks for sending the reminder :) 07:05
poelcat okay, go for it 07:05
Tk009_afk :) 07:06
mcepl I thought so ... :) 07:06
Tk009_afk I didn't have much on the agenda 07:06
poelcat do we have predefiend urls/queries for the key components? 07:07
poelcat and rss feeds too 07:07
John5342 we all know we dont need a big agenda to fill an hour :) 07:07
Tk009_afk The signature script and bug triage days 07:07
mcepl and mentoring 07:07
* poelcat noted from reading last week's minutes we still seem unable to arrive at hard goals for F11 07:08
poelcat so maybe they are as hard as they are going to get 07:08
Tk009_afk Sorry I am on my blackberry my typing is a bit slow if someone wants to take the lead 07:08
beland and goals 07:08
Tk009_afk Did you have something to add on the goal 07:09
poelcat ideally we would pick something that is measurable so we can tell if we meet our goal 07:09
poelcat but i think we have tried long enough that it is time to move on 07:09
poelcat maybe we'll have a better idea for F12 07:10
poelcat and learn something from what we do for F11 07:10
poelcat so anyone opposed to saying we are done trying to set goals for F11 ? 07:10
beland Well, I was wondering what to put on the wiki page... 07:10
beland It sounded like "try to triage bugs as they come in" was not sustainable. 07:10
poelcat beland: historically it has not 07:11
* poelcat waits 60 more seconds, then moving on 07:11
mcepl poelcat: just jinging in with those queries and RSS feeds -- I will make sure it happens. 07:12
poelcat mcepl: excellent! 07:12
* poelcat was thinking one to capture all components and then individual ones for each component 07:12
Tk009_afk I think we have done what we can on goals. After 5 weeks maybe we have to go with what we have 07:13
beland BugZappers/FindingBugs has queries, only for all components in a given release. 07:13
beland What have you come up with so far? 07:13
mcepl poelcat: meaning, that components on BugZappers/ActiveTriagers should have query and RSS feed to such query? 07:13
poelcat yes 07:13
poelcat so when someone new shows up we can just say "here start with the bugs in this URL" 07:13
mcepl or BugZappers/components ? 07:13
mcepl "yes" is not a good answer for "either ... or ..." question ... 07:14
poelcat mcepl: you are correct, the second one 07:14
poelcat mcepl: BugZappers/components 07:14
poelcat i forgot we made that the canonical location 07:14
poelcat we should probably say something about Fedora 11 at the top of that page 07:15
* poelcat will do it unless there is a volunteer 07:15
beland As a motivational and organizational technique, I think it would be helpful to have a table that simply lists the number of NEW bugs for each component. 07:15
poelcat next topic... signature script 07:15
poelcat beland: great idea... does anyone know how to generate a table like that automatically? 07:16
poelcat maybe tweaking one of the reports like this one? 07:16
beland (Or maybe only each "key" component as listed at BugZappers/components) 07:16
Tk009_afk Comphappy is working on that I believe 07:16
Tk009_afk I don't have the link on hand 07:17
poelcat Tk009_afk: when i last saw it he was listing all components 07:17
poelcat i saw mail that someone updated the greasemonkey script to add the signature 07:17
beland If per-component is the best way to organize things, a reasonable goal would be to zero out the number of NEW bugs in each component, cycling through them one by one. 07:18
comphappy tk009_afk, I am, but it is taking a while. 07:18
poelcat did this get merged into the git repo that comphappy was originally working from? 07:18
poelcat comphappy: do we have an ETC ? 07:18
--- Log closed Tue Feb 24 07:18:48 2009
--- Log opened Tue Feb 24 07:19:00 2009
-!- Irssi: #fedora-meeting: Total of 141 nicks [1 ops, 0 halfops, 0 voices, 140 normal] 07:19
-!- Irssi: Join to #fedora-meeting was synced in 21 secs 07:19
mcepl beland: sure, that's what I do for my xorg+firefox bugs, but it leads even more to my model of "bugzappers as stewards of particular componetns" rather than "horde running through bugzilla looking for a mischief" 07:19
Tk009_afk That is a idea that would work with bug triage days where we are all here 07:20
poelcat_ comphappy: do you have an estimate as to when the metrics site will be ready? 07:20
beland People should also be able to self-start just by looking at the wiki, and if there's a particular place people are encouraged to start, that's a little less intimidating than "There are 7,000 NEW bugs. Good luck" 07:21
beland I see that most people on the active triagers list haven't actually chosen a component. 07:22
comphappy poelcat_: I dont see why it would not be up by the end of the week, but it will take some time to generate useful data 07:22
Tk009_afk Our goal is bugs on the component list about 100 new bugs a day 07:22
comphappy BZ cannot give me historical data on # of bugs of different status 07:22
beland If you are only looking at "how many bugs are in the queue at the end of the week" vs. "how many bugs have gone through the queue all week", bugzilla reports will do, and I'm happy with that. 07:23
beland 07:23
beland bstring&email1=&emailtype2=substring&email2=&bugidtype=include&bug_id=&votes=&chfieldfrom=&chfieldto=Now&chfieldvalue=&format=table&action=wrap&field0-0-0=noop&type0-0-0=noop&value0-0-0= 07:23
buggbot <> (at 07:23
beland lists NEW counts for all of the components. 07:23
poelcat_ what topic are we discussing right now? 07:23
beland Oops, that was more than a mouthful. 07:23
comphappy beland: that does not show history, which is what I am trying to do 07:24
* poelcat_ notes we have about three going on 07:24
beland Right. 07:24
beland But if it's taking a long time and is a lot of work, we can use reports in the meantime. 07:24
Tk009_afk I thought we were on goals still 07:24
beland Well, it's worthwhile to figure out how we want to measure our progress toward meeting goals. 07:25
beland It sounds like there were some fuzzy goals decided on at previous meetings? 07:25
Tk009_afk Many meetings 07:25
Tk009_afk Can't force people to do what they don't want to 07:26
poelcat_ beland: we haven't been able to arrive at anything concrete so in the interest of not spending any more time I motioned to move on (early in the meeting) 07:26
* poelcat_ thinks six weeks is enough time to try :) 07:27
Tk009_afk Agreed 07:27
poelcat_ and discussion can continue on the list, but hasn't 07:27
beland Well, what do you think of the goal of zeroing out the queue for each component on a rotating basis? 07:27
Tk009_afk I did try 07:27
beland I'd like to try doing some triage, and it would be nice to have something concrete to work on. 07:27
poelcat_ Tk009_afk: yes, thanks for starting the thread 07:28
comphappy I had thought that we had agreed to stabalized NEW count for our core components? 07:28
Tk009_afk We have 07:28
poelcat_ comphappy: that is the high level goal, how will we know if we have succeeded? 07:28
Tk009_afk That is the goal we set and voted on 07:28
beland OK, that's a good goal... 07:30
beland So for tracking purposes, why don't I use Bugzilla reports to get NEW numbers for each listed component, and post that on the /components page. 07:31
poelcat_ Tk009_afk: you are correct, i should not have said we don't have something concrete... we do, i was just hoping we could better measure it 07:31
beland We can turn the /Goals page into a redirect. 07:31
* poelcat_ is trying to move the meeting forward, but we are 30 minutes in and still on this topic 07:31
beland We can tell if we've met the goal by checking the report against the numbers stored in the wiki. 07:32
poelcat_ beland: would you be willing to start a mail thread on fedora-test-list ? 07:32
* poelcat_ promises to respond there 07:32
beland Why? 07:32
John5342 beland: so we can move onto next topic 07:33
beland Well, I'll do what I proposed and send mail about it, unless there are objections 07:33
poelcat_ beland: sounds good, thanks 07:34
poelcat_ what is the status of the greasemonkey script for adding the signature and is it in the git repo? 07:34
poelcat_ anyone? 07:35
mcepl it should be 07:36
mcepl it was working for me and somebody at #fedora-bugzappers but not 100% sure now 07:37
poelcat_ mcepl: what is the git url so we can put it in the notes? 07:37
mcepl 07:37
poelcat_ :) 07:37
poelcat_ mcepl: should it be merged into the main fedora greasemonkey script that comphappy orginally worked on? 07:38
poelcat_ seems strange to have two 07:38
adamw morning. sorry, guys 07:38
poelcat_ adamw: hi 07:39
mcepl my stuff is at;a=summary 07:39
mcepl poelcat_: I thought comphappy was promising to sync my repo with the official greasemonkey one. Don't know what's the status 07:39
poelcat_ i really think we need one canonical location to send people to 07:39
poelcat_ comphappy: ? 07:39
comphappy we are going to have gm as an rpm package if I can get the stupid mozillia profiles to cooperate 07:40
comphappy and then the scripts will be shipped as a package that people update 07:40
poelcat_ comphappy: what is the interim plan? 07:41
comphappy I can sync it, but I need to know when there is a change, as it is not automatic 07:41
poelcat_ couldn't we just have people refresh from git for now? 07:41
adamw please don't make people learn how to work git just to download a script for triaging 07:42
adamw we're working with volunteer bug triagers here, not developers 07:42
poelcat_ adamw: it's only two commands 07:42
mcepl poelcat_: I think I can work a little bit more hard on my git-foo and push just scripts which are of interest to Fedora folks. Not sure whether it is possible 07:42
adamw true, but...uh. 07:42
poelcat_ adamw: well, maybe you have a point 07:42
comphappy peolcat: not really, when you think about it most of them do not have any git tools installed 07:43
poelcat_ i'm just trying to say that we need one place to point people all the time 07:43
mcepl the problem is that I have different access bits than real Fedora volunteers and it is pretty complicated for me to pretend I am not a RH employee to develop the scripts. 07:43
John5342 but you dont need to install git to install from git repo. just a link to gitweb 07:43
mcepl and so I try to have always plain URL just with Javascript file available (see 07:43
poelcat_ we should have a siple place for people to get one fedora gm script 07:44
mcepl but it is my understanding of the world, which needs to be fixed for Fedora volunteers 07:44
comphappy I think I linked to the link in trac but I am not sure it ended up working 07:44
mcepl and I don't mind if (after testing etc.) my scripts will be incldued into one overall Fedora scripts 07:44
poelcat_ comphappy: mcepl so what is our go forward plan? :) 07:47
comphappy I will sync the repo, and make sure our download link is working 07:47
comphappy I have a lot on my plate today, so it will be this evening before I get to it 07:48
mcepl and I will help with any bugs which show up 07:48
poelcat_ sounds great :) 07:49
poelcat_ next topic... triage days 07:49
poelcat_ where are we at with them ? 07:50
adamw need to start having them 07:50
adamw preferably later 07:50
adamw :P 07:50
adamw (in the day) 07:50
poelcat_ has someone volunteered to lead or coordinate them? 07:50
John5342 well i know qa is doing something similar with testing. wonder if we can organise something similar with triage. 07:51
adamw i don't think so. i'd do it if no one else wants to, although it doesn't have to be the same person each work, and it doesn't necessarily need that much 'leadership' 07:51
adamw yeah, there are weekly test days, it's working quite well lately 07:51
adamw s/work/week/ 07:52
John5342 i can help but am already fully booked till at least middle of april 07:52
adamw the other thing is to decide what exactly to do in them 07:53
poelcat_ adamw: lead away 07:53
mcepl just get some component and get rid of bug triage it? 07:53
comphappy I think that we are finding the busy people are all in bugzappers... 07:53
adamw i liked comphappy's idea about using them to train people 07:53
John5342 certainly triage day can be both productive _and_ educational 07:54
adamw but if everyone generally agrees it's a good idea and promises to try and show up :), i can start putting it together 07:54
mcepl yeah, but still my offer for mentorees stand ... if there is somebody who is serious and preferably somewhere around desktop components, I will gladly hold his hands. I have currently only one fcami and I think I can do more. 07:54
John5342 as for tips for how to make it work we might be able to talk to the people handling qa. 07:54
adamw mcepl: noted. you could maybe mail that guy who mailed the list recently? 07:55
mcepl Was it the one I replied to? 07:56
comphappy before the meeting ends, Is this what we want to monitor our goal? note it is missing a few comp, I wanted to make something before the meeting ended 07:56
adamw mcepl: no...Lalit Dhiri, topic "BugZappers" 07:56
adamw comphappy: we'd want it as a graph with X as time and Y as quantity and New and Assigned as lines, I think 07:57
beland comphappy: I think it's missing a few components, but you've got the idea. 07:57
adamw object: make the lines flat or down 07:57
mcepl adamw: you mean ? 07:57
adamw mcepl: yup 07:58
comphappy yes, I will do that in the metrics one, but i was talking about the here and now 07:58
John5342 comphappy: seems like a good interim solution 07:58
poelcat_ comphappy: yes, but only NEW 07:58
beland For triage days, we should pick a day, throw a page up on the wiki with instructions for how to get on IRC and listing willing IRC mentors, and make a list announcement. 07:59
poelcat_ if our goal is to have triaged (not in NEW) the identified components 07:59
adamw beland: yes, that's basically my aim. 07:59
mcepl adamw: I will take a look at it 08:00
comphappy I will go ahead and add the rest of the comp, and remove assigned from the table 08:00
comphappy It looks like we are out of time. 08:00
beland Let me know when you are done, and I will copy the current numbers into the wiki. 08:00
beland I'll be on #fedora-bugzappers. 08:01
mcepl adamw: do you know whether he has Linux at all? (meaning that lalit guy)? 08:01
comphappy mcepl: based on his sig I would say yes 08:02
* ltinkl zaps the zappers 08:03
mcepl oh yes, OKL 08:03
poelcat_ we're overtime now... we can move to #fedora-bugzappers 08:03
poelcat_ thanks everyone! 08:03
beland Tata 08:03

Generated by 2.7 by Marius Gedminas - find it at!