From Fedora Project Wiki

Bug Triage Meeting :: 2009-06-09



  • adamw
  • arxs
  • mcepl
  • mpreston
  • rjune
  • SMParrish
  • tk009

Meeting Recap

Long term solution is to use the FAS triageweb integration comphappy is working on as part of the grand metrics project to create an accurate list of active triagers and which components they are working on.

Short term arxs and adamw will edit the page by hand for now, removing non active triagers from the list.

tk009 will be finding out how the components list was generated last cycle to generate a new list for comparison. This will be used to modify the components list as necessary for the F12 cycle.

These items will be on next weeks agenda for follow up.

  • Added agenda item - Discussion on the policy of not triaging kernel bugs.

The consensus during the F11 development cycle was that triaging kernel bugs was beyond our limited resources, and that we could have a greater impact overall if we focused our efforts elsewhere.

We have reevaluated that position for the F12 development cycle. adamw will be taking the lead with the help of a few volunteers to get BugZapper kernel triage back on track.

IRC Transcript

--- Log opened Tue Jun 09 11:01:23 2009
tk009 BugZappers meeting starting now, who all is here 11:01
arxs * here 11:01
* SMParrish At FAD but here 11:01
adamw here sir! 11:01
tk009 =) 11:01
adamw any of our new people around? say hi, don't be shy 11:02
tk009 huge crownd today 11:02
adamw heh 11:02
tk009 Okay, the first and only thing for today is the cleaning up of the active triagers page and the components list 11:03
tk009 in past we moved actriagers to a past triagers list 11:04
tk009 active triagers* that is 11:04
adamw hi, mpreston 11:04
tk009 adamw you mentioned that list now being created by comphappy and FAS yes 11:05
mpreston morning/afternoon/evening 11:05
tk009 hello mpreston 11:05
mcepl re-here 11:05
adamw ah, and brennan's here too 11:05
adamw yay 11:05
adamw now it's a big crowd :) 11:05
tk009 =) 11:05
adamw tk009: shall we reboot? 11:05
tk009 we can yes 11:06
adamw alrighty...from the top :) 11:06
tk009 for comphappy: Okay, the first and only thing for today is the cleaning up of the active triagers page and the components list 11:06
tk009 in past we moved active triagers to a past triagers list 11:06
tk009 adamw you mentioned that list now being created by comphappy and FAS yes 11:07
adamw yeah, i think the 'triagers' group should become the canonical list 11:07
mcepl adamw, tk009: I would add also that I think we should agree on the general part line for F12 11:07
tk009 so comphappy can you shed any light 11:07
adamw but we can purge the wiki page and restart, why not 11:07
adamw mcepl: party line? about what? 11:07
tk009 that confused me as welll 11:07
mcepl (as an additional item on agenda; e.g., do we bug triage kernel bugs in F12 timeframe?) 11:08
adamw on the active triagers list - we should be careful not to take out developers who've marked themselves as their own triagers, best leave those alone 11:08
rjune_ Can you shed a little light as to what the triagers list is? I'm guessing folks that determine which bugs are most important to fix 11:08
adamw rjune_: BugZappers/Components_and_Triagers 11:08
rjune_ ty 11:09
adamw the idea behind emptying the triagers list and asking people to re-add themselves for each release is to make sure it only contains really active people 11:09
comphappy_ sorry computer crashed I think my drive is dieing 11:09
adamw comphappy: eek :\ 11:09
arxs hope you have a good backup :) 11:10
tk009 mcepl: do we want to start triaging kernel bugs or stay away from them? 11:10
tk009 I am still confused 11:10
adamw well, let's just add it as a topic to the agenda 11:10
adamw stick with this one for now... 11:10
tk009 compahappy where did you come in? 11:11
comphappy_ ? 11:11
tk009 did you see any of the discussion? 11:11
comphappy_ nope just when I said computer had crashed 11:11
rjune_ adamw: this afternoon can you answer a question or two for me? 11:11
tk009 for comphappy: Okay, the first and only thing for today is the cleaning up of the active triagers page and the components list 11:11
mcepl tk009: yeah, that's the question 11:11
adamw rjune_: sure sure - it's triage day after the meeting in #fedora-bugzappers, i'll be around all day. 11:11
tk009 comphappy: can the active triagers list be populated by FAS? 11:12
mcepl anyway, concerning the triagers list, could we make a list of people (like me) who are supposed to be triagers and don't have to refill themselves to the list? 11:12
tk009 or a list of zappers in general 11:12
comphappy_ here is an idea 11:12
comphappy_ I just added a FAS authentication layer to triageweb 11:12
comphappy_ this means users can start customizing things, one idea is that they will sign up for components 11:13
adamw mcepl: you mean, people whose jobs involve triaging? 11:13
comphappy_ then you could have a page where you could see people in triagers group and there respective components? 11:13
SmootherFrOgZ /s 21 11:14
adamw that would be nice 11:14
comphappy_ I will be working on traigeweb for the next 12 hours 11:14
mcepl yup 11:14
adamw we also use the page as a list of components that need triage, though, so i'm not sure we can entirely replace the wiki page 11:14
adamw but it seems like triageweb can definitely help us with who's an 'active' triager 11:14
tk009 we need the wiki page 11:14
adamw once we have the 'triagers' groups set up, presumably triageweb could automatically generate a report of people in the group who haven't done any actual triaging for X months, we could use that 11:15
comphappy_ kind of like the packageing page where they have a list of things that are wanted to be packaged 11:15
adamw yeah, exactly 11:15
tk009 what kind of time frame are we looking at for this 11:15
tk009 or is this something that can be done right away 11:16
comphappy_ that page could be up today, I finally got internet down here in sillicon valley this morning 11:16
mpreston as a newbie still trying to decide what components to work with, it is helpful to have a list of component needing triagers 11:16
adamw i think for now we can just go through and clean up the wiki list...actually i think it's small enough that we can do it, y'know, ad hoc 11:16
tk009 I agree 11:16
adamw just reading through that list i can tell you off the top of my head, mostly, who's active and who's not 11:16
arxs let's start 11:17
tk009 s ojust a hand edit job for now cleaning up 11:17
adamw then we can give comphappy some time to see what triageweb is capable of doing here exactly - whether we use it as a reference for updating the wiki page, use a report page on triageweb to entirely replace the wiki page, or have triageweb automatically feed data into the wiki page 11:17
tk009 that works for me 11:18
comphappy_ the automatic feed is unlikely as the mediawiki verion fedora uses does not support it 11:18
adamw for the newer guys, btw: 'triageweb' is a neat system written by comphappy (brennan) for deriving triage-related statistics from bugzilla 11:18
arxs 11:18
adamw please go there and poke it, comphappy welcomes bug / feature requests :) 11:18
comphappy_ and yes it is out of date, I have been trying to push the new version for a few days 11:18
adamw so in theory triageweb should be able to keep track of active triagers and the components they're working on, rather than us having to do it by hand in that wiki page. 11:19
adamw alright, so - does anyone want to take charge of doing a quick update on the wiki page by hand for now? arxs? 11:20
tk009 still would be nice to see who is part of the group on the wiki 11:20
arxs i can do this of course 11:20
adamw okay, congratulations on volunteering ;) 11:21
tk009 hehe 11:21
tk009 that leaves the components 11:21
adamw what about the components? 11:22
tk009 the last time we generated the components list by picking the packages with the most bug reports and adding critical components 11:22
adamw ah - so, do we regenerate the list and see if it's changed? 11:23
tk009 some of those components don't belong on that list I feel 11:23
tk009 yes 11:23
tk009 that and remove yum 11:23
tk009 and anaconda 11:23
tk009 both don't count to us 11:23
tk009 andy does even take part in the bugzappers 11:24
adamw i'm working on that actually 11:24
tk009 doesnt* 11:24
adamw i've been talking with her via email - one of my todos is trying to bring kernel and anaconda triage more into the bugzappers circle 11:24
SMParrish it is safe to remove yum from the list of components. the yum maintainer does it himself and prefers it that way 11:25
adamw current status is that she's working on a wiki page that will describe the lifecycle used by the anaconda folks 11:25
adamw (it's different from the standard one) 11:25
adamw then we'll look at how we can reconcile it with the official lifecycle as defined by us / qa 11:25
adamw i've asked her to come to some meetings, hopefully she'll make it out soon 11:25
adamw SMParrish: i've talked to him about that as well - current status is that he doesn't really mind if people triage bugs, just he requests they get left as NEW and the 'Triaged' keyword used, rather than being changed to ASSIGNED. hence the note on the pag.e 11:26
adamw so...until i fail mightily, i'd rather we still pretend we care about anaconda :) 11:26
SMParrish adamw: your right my bad 11:27
tk009 my feeling is you can't take your ball and walk off and then pretend we are one big happy faily 11:27
tk009 family* 11:27
adamw right, but it's not really at that point yet (afaik) - i don't think andy ever really was _in_ the group, so it's a case of she's not yet brought the ball, and i'm trying to get her to :) 11:28
adamw on the general topic...did we save the query we used to generate the list in the first placE? 11:29
tk009 that was done by jds2001 I believe, and I do not believe so 11:29
adamw it might be tricky to do it in a genuinely comparable way then :\ 11:29
tk009 I could as k him 11:29
adamw that'd be great if you could... 11:30
adamw shall we put that down as an action item for you? try and ask jds2001 about the query, and if you're successful, bring an updated result to the next meeting so we can discuss it there? 11:30
tk009 yes 11:31
tk009 I guess that covers the triager/component page 11:31
adamw great 11:31
tk009 anyone have something t oadd 11:31
tk009 mcepl: I guess the floor is yours 11:32
arxs adamw: do we wait with the update of the wiki page until the bz query is here? 11:33
tk009 no 11:33
tk009 that can be done now 11:33
adamw arxs: no, you can go ahead and revise the triagers part 11:33
mcepl well, we have agreed in the beginning of F11 cycle, that we would rather spend time on something manageable, rather than kernel 11:34
tk009 yes I recall that 11:34
mcepl I think the policy paid off handsomely, we did some good work 11:34
tk009 some allot more than other =) 11:34
mcepl however, just want to be sure we want to continue with this policy of non-internvention-in-kernel for F12 as well. 11:35
adamw i do have something to add on this - it ties in with the anaconda thing 11:35
adamw jlaska asked me to try and improve anaconda and kernel triage within the bugzappers group 11:35
tk009 for me, the kernel is beyound my skill level, that is the reason I leave the alone 11:35
adamw this came out of the 'fedora program' meeting (which, er, i don't know what it is, but hey) 11:36
mcepl so anaconda are not against BugZappers bug triaging anymore? 11:36
adamw i think neither of us (me or jlaska) were actually aware this had been a conscious choice not to triage kernel bugs on the part of bz team 11:36
tk009 it was 11:37
tk009 we decided before you were recruited =P 11:37
adamw mcepl: er, not quite that simple :) the managers of anaconda and kernel are on board, apparently. andy is a bit reluctant, but as i summarized above, we're working on it :) 11:37
mcepl we spent F10 beating ourselves with guilt how much we cannot do enough to fix kernel bugs, and we decided for F11 just not do it. 11:37
adamw ok. well, i'll go back to jlaska with that info because obviously it'll feed in to that issue. for now i'd been working more on the anaconda side of things than kernel anyway 11:37
comphappy_ I can take on some kernel stuff if we want to, as I have done development on it 11:38
tk009 do you have the time? 11:38
comphappy_ this summer yes 11:38
adamw i think if we really wanted to do the kernel we'd have to have at least two or three committed and knowledgeable people with decent chunks of time available 11:38
comphappy_ but it is not a one person job 11:38
adamw so it is a bit of a high bar 11:38
adamw i'd almost think it's an area where RH might want to consider paying someone - after all, we have paid triagers already (mcepl, alindebe) 11:39
tk009 I ma willing to learn 11:39
mcepl probably kernel should (could?) be split into some subcomponents ... adamw could you talk with kernel folks (whoever it is) how they do divide it themveles? 11:39
adamw yeah, i'll do that as part of the process 11:39
comphappy_ I think that the kernel is the place were we can have the biggest impact 11:39
comphappy_ epecially around beta and preview tiems 11:39
tk009 so we are triaging kernel bugs for F12 then 11:40
adamw thing is, both comphappy and tk009 are already doing valuable bugzappers work, so if you started doing kernel triage it may eat into that :\ 11:40
comphappy_ mcepl: I thought there was a wiki page on that? 11:40
adamw comphappy_: yeah, there is, let me see if i can find it 11:40
mcepl on what? 11:41
tk009 kernel triage 11:41
tk009 yes there is 11:41
comphappy_ how kernel triaging is devided 11:41
adamw KernelBugTriage 11:41
adamw there's a table at the bottom which gives a split 11:41
adamw sl y 11:42
adamw so you could look at the kernel as a dozen separate components...which would probably give a more accurate indication of the scale of the task :\ 11:42
comphappy_ I am just saying that I think kernel triaging is very imporant as it contains more of the show stoppers and is more likely to conatain blockers then say xchat 11:42
adamw are any of our new guys interested in suiting up and diving into kernel triage? :) 11:42
rjune_ I will 11:43
mcepl oh cool 11:43
tk009 hey now xcaht is a blocker to me =P 11:43
rjune_ I've done some kernel work already 11:43
adamw rjune_: that would be awesome 11:43
tk009 yes thanks rjune 11:43
adamw tk009: yeah, me too - i can live without a sound driver but i'd be in trouble without xchat :) 11:43
adamw i think for now we should say that we want to try and tackle kernel triage, but with the awareness that we should give up again if it becomes too much of a resultless time sink 11:44
mpreston that isn't a full list, just a list of subsystems... 11:44
comphappy_ if i do not take on the kernel as a component to foucs on I am willing to advise a little on at least the wireless area 11:44
rjune_ adamw: You will most definately need to spend some quality time with me later. 11:44
tk009 no worries rjune 11:44
adamw if it's ok with everyone i'll take it as an action item to try and co-ordinate that, since jlaska had already asked me to 11:44
tk009 kk 11:45
adamw now at least i have a group of people interested in helping out with kernel triage 11:45
adamw so i can draw you guys in on the process of trying to get it going in a reasonable way 11:45
tk009 I think comphappy is right, it could have a ajor impact 11:45
adamw oh yeah, definitely if we get it right :) 11:45
* comphappy_ notes that we are 15 min to to end. Any other agenda items? 11:46
adamw no, we had a light agenda today 11:46
tk009 I have noting else, or if I did I forget it 11:46
mpreston Bugzilla Triage Buttons doesn't seem to be working on the tools page 11:46
adamw we're in 'any other business' already really, since this was added to the agenda during the meeting :) 11:46
comphappy_ thank you everyone who has been submitting RFE and bugs on triageweb 11:47
arxs you need to remove the ;hb=HEAD at the end of the url 11:47
mpreston thanks 11:47
tk009 arxs you can do that s well =) 11:47
adamw yeah i found that too 11:47
adamw i'll fix the page ;) 11:47
tk009 is the script working again? 11:48
tk009 last time I tried it, it didn't work 11:48
adamw yeah 11:48
adamw mcepl did a big update may 23 11:48
adamw and has done small fixes since then 11:48
adamw and greasemonkey now works in firefox 3.5 11:48
arxs it works for me, after the update from mcepl 11:48
adamw so it's all go afaik 11:48
tk009 I will install again in a bit 11:49
tk009 and confirm =P 11:49
arxs it important to remove all older triage scripts 11:49
tk009 yes I did 11:49
adamw let's see, do we have any other business? anyone have anything to bring up? 11:50
mcepl I have monstrous upgrade in the parallel branch, but it is very not done yet ... most of the logic was pushed to data and that to JSON file on the webserver. 11:50
adamw i hope everyone's ready to celebrate release day btw :) 11:50
rjune_ already downloaded and seeding 11:51
adamw naturally, fedora 11 has no bugs, so our work is over anyway 11:51
mcepl yeah, my children will see their father again!!! 11:51
tk009 lol 11:51
tk009 ok I guess that is a wrap on the meeting 11:51
rjune_ Is that who those little people that live in your house are? 11:52
rjune_ adamw: PM for a few? 11:52
tk009 triage day in bugzappers starts now 11:52
adamw rjune_: :) sure, or we can do it in #fedora-bugzappers 11:52
adamw it'll make triage day look busier :) 11:52
rjune_ LOL 11:52
tk009 EOM--------------- (plays the go home music) 11:52
--- Log closed Tue Jun 09 11:53:32 2009

Generated by 2.7 by Marius Gedminas - find it at!