From Fedora Project Wiki

Revision as of 18:12, 6 July 2009 by Rjune (talk | contribs)

Bug Triage Meeting :: 2009-06-23

Agenda: https://www.redhat.com/archives/fedora-test-list/2009-June/msg00629.html

Attendees

  • adamw (96)
  • rjune_wrk (63)
  • arxs (44)
  • mcepl (38)
  • LinuxCode (13)
  • comphappy (13)
  • SMParrish (9)
  • poelcat (5)
  • Viking-Ice (1)
  • jlaska (1)


Meeting Recap

  • topic - Triage Metrics - Update on the FAS integration and overall status.

Comphappy had FAS integration complete until some backend changes rolled through, this should be completed tonight. He'll have an update next week.

  • topic - Component list update for the F12 cycle - Update on status.

Axrs completed the initial list, there was a overwhelming request for requirements information as well. he'll get that together for next meeting.

  • topic - Kernel Triage - Discuss the feedback from Chuck Ebbert on kernel triage.

Chuck asked one specific question in his email: If there might be a different way to indicate a bug's been triaged rather than simply setting it to ASSIGNED? During the meeting on some of the available options were discussed, none were agreed upon. We will continue this in the mailing list. - this topic is also related to anaconda, and with members of the anaconda team present at the meeting we took the opportunity to discuss how we can better assist them.

  • topic - Bugzilla Legend - Unify bugzilla defs between RHEL and Fedora

This was on the add to agenda page, but is apparently obsolete.

  • topic - Debugging pages

Adamw has been discussing unified debugging proceedure with some Ubuntu developers. This could give both projects more help clarifying debugging procedures and insuring the proper information gets to the developers assigned to bugs.

  • topic - Open Floor

Mecpl needs testers for a new version of the GreaseMonkey script


  • Action Item - comphappy get the fas intigration working until next week, likely tonight, done for Triage Metrics
  • Action Item - arxs will update the list with requirements iformation for next meeting.
  • Action Item - adamw will update on collaboration with Ubuntu on debugging procedures.

IRC Transcript

rjune_wrk #startmeeting 14:59
comphappy Here for 10min 14:59
LinuxCode here 14:59
SMParrish here 14:59
rjune_wrk meeting bot commands are found at #link http://wiki.debian.org/MeetBot 15:01
arxs hello guys 15:01
rjune_wrk arxs: backup chair for me? 15:01
mcepl /me 15:01
arxs rjune_wrk: well 15:01
mcepl no, not me! 15:02
arxs mcepl: no worry, i take it :) 15:02
rjune_wrk #chair arxs 15:02
rjune_wrk #topic - Triage Metrics - Update on the FAS integration and overall status. 15:02
rjune_wrk comphappy: make your quick announcement? 15:02
* poelcat here 15:03
* adamw here too 15:03
rjune_wrk #chair adamw 15:03
arxs hmm, maybe comphappy can't here us as the bus? :) 15:05
comphappy Alright I had fas intigration working last night locally then added some features jlaska wanted that meddled with the backend db and the thing blewup on me I need to take a look tonight. I also got component groups working via fast. 15:05
comphappy Phone slow to typeon 15:05
rjune_wrk comphappy: any chance of it being finished by the next meeting? 15:05
comphappy Likely tonight 15:05
rjune_wrk barring more backend meddling? :-) 15:05
rjune_wrk Anybody else have anything to add? 15:06
* poelcat 15:06
adamw motion: quit listening to jlaska, he only causes trouble ;) 15:06
comphappy Release is coming up soon it needs testing. 15:06
comphappy adamw: Added rawhide features I think you willl like 15:07
poelcat comphappy: how about the bugs/RFEs I filed? 15:07
adamw comphappy: great work, thanks as always 15:07
comphappy poelcat: At least a few are making it in 15:07
poelcat cool, thanks! 15:08
rjune_wrk Just for completeness, what is FAS an acronym for? 15:08
poelcat comphappy: and thanks again for all your work on this 15:08
arxs fedora account system 15:08
comphappy Mostly this is features in the backend I am fixing. Ones that can't really be changed post release 15:09
arxs you can find it on https://admin.fedoraproject.org/accounts/ 15:09
comphappy But there are other fixes as well. 15:09
rjune_wrk #link https://admin.fedoraproject.org/accounts/ 15:09
* comphappy done talking 15:09
rjune_wrk comphappy: go through the full list next week? 15:10
comphappy rjune_wrk: Some no as they are fixes to other packages 15:10
rjune_wrk good enough, thanks. 15:10
rjune_wrk #topic - Component list update for the F12 cycle - Update on status. 15:10
rjune_wrk arxs: your floor 15:11
arxs #action comphappy get the fas intigration working until next week, likely tonight, done for Triage Metrics 15:11
arxs ok, if send a mail to the f-t-l 15:11
comphappy arxs: Looked like a good list to me 15:12
arxs https://www.redhat.com/archives/fedora-test-list/2009-June/msg00757.html 15:12
arxs you can find the list at the wiki under https://fedoraproject.org/wiki/User:Arxs/CPCL 15:12
arxs comphappy: thanks 15:12
arxs but one thing is maybe missing, a DE 15:12
adamw does the proposal _remove_ any components? 15:12
arxs adamw: it don't think to (you mean from the actually components and triagers list?) 15:13
adamw yeah, does the new list miss any components that are on the old list 15:13
* comphappy off to quarterly review. 15:13
arxs no, my plan is (sorry if this was not clear) to bug the CPCL list into the actually components and triagers list, so that we have at the end only more items 15:14
adamw cya comphappy, thanks a gain 15:14
arxs s/bug/get/ 15:15
adamw arxs: ah, ok. we don't want to grow forever, but that's probably good for this cycle, as we've been running out of components for new members :) 15:15
adamw which is a good problem to have, heh 15:15
arxs adamw: of course! :) 15:15
mcepl these components are just symbolic names, right, you don't mean to test virtual components like xorg-x11-drivers for real? 15:16
adamw in that case, looks good to me 15:16
adamw some things we can pare; for xorg for instance we don't really need to consider _all_ drivers critical 15:16
mcepl and I have never seen a bug against xorg-x11-xauth 15:17
arxs mcepl: well, i think we should only take care of nouveau, radeon and intel 15:17
mcepl the real drivers were is life are -ati, -intel, -nouveau, perhaps -vesa, to be sure 15:17
* SMParrish agrees 15:17
adamw openchrome is worth caring about too 15:17
adamw (just about) 15:17
mcepl yeah, I am forgetting ... it is not managed by RH engineers 15:18
adamw aside from that, even if they're totally broken we'd probably still ship :D 15:18
mcepl well, let me not comment on that ;-) 15:18
adamw but in general, looks good 15:18
mcepl (just spent whole day plowing through -ati bugs) 15:18
arxs ok, i going to remove xauth and change the xorg-drivers to ati, nouveau, intel and openchrome 15:19
adamw we might not need to worry about the compose packages as there's a fairly tight loop there between the guys building and the guys using 15:19
adamw that's pungi, mash et al 15:19
arxs that was also my feeling about, i think we should cut off the list after the line 'The following items are unchanged included here from skvidal's list.' 15:20
mcepl what about Pulseaudio? why not? 15:20
arxs mcepl: Pulseaudo is one the 'old' list, and it will remain there 15:21
adamw some are worth having - bash, coreutils, gcc are things we could cover 15:21
mcepl ok 15:21
adamw the others there are important but mostly are so mature they're unlikely to suffer problems serious enough to care about 15:21
arxs adamw: right, but how many bugs are there? (maybe gcc, but i don't check this) 15:22
arxs any other suggestion's? 15:23
SMParrish keep on mind that all dependenices of critical packages will also be critical packages 15:24
arxs SMParrish: good point 15:25
adamw yeah, we could look at the list of deps for anything in the final list 15:25
adamw but i think overall we're good with arxs' list...thanks for your work there :) 15:26
SMParrish what we need to do is generate a list of all the dependencies on the list arcs made to see how many packages were are talking about and how to group them 15:26
SMParrish s/arcs/arxs 15:26
arxs SMParrish: this can be done with "repoquery --wahtrequires" or i'm wrong? 15:27
adamw just repoquery --requires 15:28
arxs i will change the list to the given suggestion by you, and then bring up a second list with all the deps. 15:29
mcepl no, repoquery --whatrequires --alldeps <package> 15:29
adamw uh? i'm not sure we're on the same page :) 15:29
mcepl (plain --requires is for groups) 15:29
adamw --whatrequires foo gives you things that require foo, doesn't it? 15:29
adamw that's not what we want 15:29
adamw we want what foo requires 15:29
mcepl yes 15:30
SMParrish adamw is right 15:30
mcepl sorry, once more 15:30
mcepl yeah, but none of these will give you names of package, a sec 15:30
adamw you need to pipe it a bit 15:31
mcepl --requires --resolve is probably the right combination 15:31
adamw oh yes, that does it - neat 15:32
SMParrish thats it 15:32
adamw i used to pipe that through another --provides query, that way's much better 15:32
adamw (your way i mean :>) 15:32
arxs i will try it, if not "yum remove" show also the deps :) 15:32
rjune_wrk LOL 15:32
mcepl yeah, but it has some side-effects ;-) 15:32
rjune_wrk that everything for this topic? 15:33
SMParrish so to get an idea gdm has the following deps that will have to be critical pkgs as well http://fpaste.org/paste/16920 15:33
arxs rjune_wrk: yes i think so 15:33
adamw they don't necessarily have to be critical, but we should consider whether they are 15:33
mcepl there's something wrong -- freetype-freeworld is not Fedora package 15:33
adamw so arxs, make us another list :D 15:34
rjune_wrk #link https://fedoraproject.org/wiki/User:Arxs/CPCL arxs's list of update componants 15:34
rjune_wrk #action arxs will update the list with requirements iformation for next meeting. 15:34
rjune_wrk #topic - Bugzilla Legend - Unify bugzilla defs between RHEL and Fedora 15:34
rjune_wrk adamw: this one is for you 15:34
SMParrish mcepl: could be because I have rpmfusion repos enabled here 15:34
adamw er, it is? 15:34
mcepl yup, I think so, just emphasizing 15:34
rjune_wrk adamw sent a message to the list about this for beland. Hopefully we can 15:34
rjune_wrk > have an update. 15:34
rjune_wrk yup. 15:34
rjune_wrk It's from the add it to the agenda page. 15:35
* adamw is lost...what message is this? 15:35
rjune_wrk I do not know, there is suprisingly little information on the agenda page to pick it up and track it 15:35
rjune_wrk https://fedoraproject.org/wiki/BugZappers:meeting-agenda-list 15:36
adamw well, er, i sort of abandoned unifying the definitions a while ago 15:36
adamw what i've been working on since is just having the fedora definitions properly defined and linked to from bugzilla 15:36
adamw i'm mostly done with that, as far as I knew :) 15:36
rjune_wrk has that been communicated to beland? 15:36
adamw is this an old addition to the agenda list maybe? 15:36
adamw let me look at the wiki page history 15:36
rjune_wrk could be. if so it needs to be removed. 15:36
adamw yeah, it's old, last change to the page was 18 may 15:37
adamw clean it out :) 15:37
rjune_wrk will do 15:37
rjune_wrk #topic - Debugging pages 15:38
adamw now this one i know where we're at :) 15:38
rjune_wrk good. 15:38
adamw this came out of a qa proposal by viking-ice, but obviously touches on bz stuff as well 15:38
rjune_wrk bz is? 15:39
rjune_wrk bugzilla. 15:39
arxs bugzilla 15:39
adamw bugzappers 15:39
rjune_wrk doh! 15:39
adamw :D 15:39
adamw i think the way the x.org page came out was great, and any contributions anyone can come up with to that or other debugging pages would be awesome 15:39
adamw and if your triage component(s) don't have a page, consider making one 15:39
adamw i've also had a proposal from a guy at ubuntu to work together on upstreaming the content of those pages where possible; anyone see any problems with doing that? 15:40
adamw e.g. we'd merge the content of the ubuntu and fedora wiki pages on debugging x.org, and send it upstream to the x.org wiki 15:40
rjune_wrk adamw: link to the x.org page? 15:40
adamw https://fedoraproject.org/wiki/Xorg/Debugging 15:40
mcepl adamw: no, I think wiki is covered by CC license or something anyway .... but be sure to check versions of packages ... bubuntu is behind us (not that broken ;-)) 15:41
rjune_wrk #link https://fedoraproject.org/wiki/Xorg/Debugging 15:41
rjune_wrk I think it's probably a good idea 15:41
arxs well, the NM page is not ready for release: https://fedoraproject.org/wiki/User:Arxs/NetworkManager/Debugging 15:41
adamw mcepl: that's the good part: debugging *procedures* aren't really distro specific 15:41
arxs but i work on this :) 15:41
adamw arxs: awesome! thanks a lot 15:42
adamw looks like you've got a good start going 15:42
mcepl sure, but I think they haven't immersed themselves into KMS that much, which may make things slightly different 15:42
adamw if you need any help do shout on the list, we got the X page done collaboratively pretty well 15:42
adamw mcepl: not really, it just needs a little bit in the intro to say "on distros X Y Z don't bother with the KMS stuff" 15:42
adamw but in practice, current ubuntu releases have kms enabled by default, they just don't _do_ anything with it 15:43
mcepl adamw: and there is continuation about input devices for X on https://fedoraproject.org/wiki/Xorg/Input_Triage_Algorithm and https://fedoraproject.org/wiki/Input_device_configuration 15:43
mcepl adamw: yeah, but we (and them) have to be aware of it 15:43
rjune_wrk mcepl: KMS is ? 15:43
mcepl kernel modesetting 15:43
arxs Kernel Mode Setting 15:43
adamw mcepl: sure, but i don't think that'd be a problem. that's part of what we could work on together. 15:44
mcepl moving some parts of X drivers to kernel 15:44
mcepl sure 15:44
adamw so for now i've told that guy we'd probably be happy to go with it, and asked a few questions, and i'm waiting for a reply from him 15:44
adamw once we have something concrete to propose, we'll take it to the list 15:44
rjune_wrk wonderful, make it a brief update item next week? 15:45
adamw sounds good 15:45
rjune_wrk #action adamw will update on collaboration with Ubuntu on debugging procedures. 15:45
rjune_wrk #topic - Open Floor 15:45
* LinuxCode raises his hand 15:45
rjune_wrk Yes. 15:46
arxs rjune_wrk: did you know that the diff is from action and agreed ? 15:46
arxs in MeetBot :) 15:46
rjune_wrk action is person X will do something 15:46
rjune_wrk agreed is there is a consensus 15:46
mcepl I have almost ready release of next version of the Greasemonkey script, but I need to finish testing (we did something with arxs yesterday, but he hasn't confirmed me that I have fixed his problems). It is quite radical change, so I don't want to push it on you without testing. OTOH, if nobody volunteers, I just will and pick up the pieces. I need to move on with the stuff. 15:47
arxs rjune_wrk: thanks for the info 15:47
adamw mcepl: sorry i couldn't test for you earlier 15:47
* LinuxCode waits for voice 15:47
rjune_wrk mcepl: You're jumping the gun a little bit, and LinuxCode has something to bring up. 15:47
adamw LinuxCode: you don't need voice, afaik 15:47
mcepl adamw: there is always time for penitence ;-) 15:47
adamw LinuxCode: we're not restricted 15:47
* mcepl shuts up 15:47
LinuxCode adamw, was just going to introduce myself as I just applied to the traiging group 15:48
arxs mcepl: let's talk about it after the meeting? 15:48
LinuxCode triaging 15:48
adamw awesome! thanks for applying 15:48
rjune_wrk LinuxCode: did you send your intro to the list? 15:48
arxs LinuxCode: welcome ! 15:48
LinuxCode rjune_wrk, no, didnt know that was required 15:48
adamw you should actually apply to 'fedorabugs' rather than 'triagers' 15:48
mcepl arxs: yeah, I just wanted to make announcement here ... apparnetly, nobody cares when I speak on #fedora-bugzappers 15:48
LinuxCode but some of you might know me anyway 15:48
arxs mcepl: i care :) 15:48
Viking-Ice What's the story with this ubuntu debugging collaboration? 15:48
mcepl adamw: what's triagers anyway? 15:48
mcepl Viking-Ice: see above 15:49
adamw LinuxCode: see the procedure :) https://fedoraproject.org/wiki/BugZappers/Joining 15:49
mcepl Viking-Ice: do you need a backlog? 15:49
adamw Viking-Ice: i really just filled them in on what i mentioned to you in pm earlier 15:49
adamw Viking-Ice: nothing new, i haven't heard back from the guy yet 15:49
adamw mcepl: i 15:49
adamw mcepl: i'll explain outside the meeting, let's not fill up the logs 15:49
adamw oh darn 15:50
adamw i just realized you got it right and i was wrong, it's triagers not fedorabugs you should apply to, sorry :) 15:50
adamw getting confused 15:50
LinuxCode k so points 5 + 7 I still have to do 15:50
adamw so just send your intro to the list and i'll approve you 15:50
LinuxCode I get on it asap 15:50
rjune_wrk LinuxCode: happy to have you. 15:50
adamw that's right, thanks again 15:50
adamw if you hang out in 15:50
adamw grr 15:50
LinuxCode well, every little helps 15:51
adamw damn laptop keyboard...in #fedora-bugzappers after the meeting, some of us may be able to help you get started 15:51
rjune_wrk mcepl: ok, knock yourself ut. 15:51
rjune_wrk out 15:51
LinuxCode adamw, sure, thomasj suggested I join the bugs group, but then somebody told me that it will slip into triage now 15:51
arxs adamw: did you _not_ take your ibm type m keybord with out while you traveling? 15:52
adamw arxs: surprisingly enough...no :) 15:52
adamw LinuxCode: yeah, triagers was the right group, sorry to confuse 15:52
LinuxCode thats ok, no worries 15:52
adamw great to have you :) 15:53
LinuxCode ;-} 15:53
rjune_wrk we have 8 minutes left, lets get to mcepl and his update and close the meeting. 15:53
adamw i had something to add, too, btw. 15:53
rjune_wrk ok 15:53
rjune_wrk you can be after mcepl 15:53
* arxs listen 15:53
adamw sure 15:53
mcepl rjune_wrk: I think I said my stuff, just come to #fedora-bugzappers and let's beat the bugs out of the thing 15:53
rjune_wrk please test so we have some testing. 15:54
rjune_wrk adamw: all you 15:54
adamw thanks :) 15:54
adamw just on the priority / severity thing 15:54
adamw the restrictions have been made in bugzilla and i updated the 'how to triage' page, so please be setting the 'severity' field for all bugs you triage from now on 15:54
adamw the guidelines are pretty simple - most stuff is 'normal', bad breakage in the package is 'high', bad breakage that has implications distro-wide is 'urgent', really trivial stuff is 'low' 15:55
rjune_wrk triagers get severity, and developers get priority, correct? 15:55
arxs adamw: thanks to make this possible! 15:55
adamw rjune_wrk: yes, according to mcepl's proposal. don't touch priority at all. 15:55
arxs rjune_wrk: right 15:55
adamw so to put this into use, we will be reviewing all bugs marked 'urgent' at the blocker review meetings, to make sure none of those that should be blockers slip through the net 15:56
arxs i can see, from the triage of the last month, no severity war on more then 20 bugs :) 15:56
adamw i'm also going to experiment with doing a weekly review of 'urgent' bugs, sent to f-t-l and f-d-l, to see how that goes: i'll write one this friday, if it works out well, it may be nice for someone else to take that over 15:56
adamw i'll ask for volunteers in the mail :) 15:56
arxs adamw: sounds good 15:57
mcepl the main difference between urgent and high (just to emphasize) is that urgent breaks whole distro, high breaks (w/o workaround) just the particular app 15:57
rjune_wrk Kevin_Kofler: we're wrapping up, you'll have it shortly. 15:57
mcepl normal means there are some workarounds or some parts of app still work 15:57
rjune_wrk and that all should be in the wiki page. 15:58
mcepl and it's all highly dependent on component 15:58
mcepl I think it is 15:58
adamw yeah, i've updated one page but not the legend page yet 15:58
arxs it is 15:58
adamw which takes a slightly more formal approach 15:58
adamw we can do that soon 15:58
rjune_wrk ok, the KDE SIG is here for the room, let's give them a minute to setup. 15:58
adamw yeah, we don't have to be too hard-and-fast on the definitions 15:58
jlaska comphappy: <delayed response> I'm afraid adamw is right ... I do frequently just cause trouble :) 15:58
rjune_wrk bugzappers event in #fedora-bugzappers. 15:58
adamw the important thing is to have a useful continuum 15:58
adamw thanks all! 15:59
rjune_wrk ending in 5 15:59
rjune_wrk #endmeeting 15:59

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