From Fedora Project Wiki

Bug Triage Meeting :: 2009-01-27

Attendees

  • tk009
  • John5342
  • jds2001
  • poelcat
  • mcepl
  • SMParrish
  • wwoods

Fedora 11 Goals

  • Continue discussion around what we want to accomplish for Fedora 11
    • Goals that are achievable and measurable
  • Ideas:
    • Help people discover their set of components to follow
    • Build individual relationships with maintainers
    • Each person set a personal goal of how many bugs they plan to triage
    • Identify important components and assign people to them
    • Focus on components related to the feature list
    • Focus on components with the most reported bugs--
    • Reorganize the wiki pages to better attract and retain new people
    • Do a better job of retaining new people--current success rate is low
  • Continue discussion on #fedora-bugzappers and fedora-test-list@redhat.com
    • NEXT MEETING: lock down goals for Fedora 11

Top F10 Bug Components (compliments of wwoods)

  1. kernel
  2. anaconda
  3. xorg-x11-drv-ati
  4. selinux-policy
  5. xorg-x11-drv-i810
  6. NetworkManager
  7. PackageKit
  8. system-control-network
  9. pulseaudio
  10. mkinitrd
  11. gdm
  12. evolution
  13. selinux-policy-targeted
  14. rpm
  15. nautilus
  16. firefox
  17. xorg-x11-server
  18. java
  19. hal
  20. compiz
  21. yum
  22. openoffice
  23. initscripts
  24. glibc
  25. xorg-x11

Last Meeting Follow-up

IRC Transcript

poelcat who is here? 07:02
tk009 =) 07:03
poelcat mcepl: John5342 jds2001 ping 07:03
poelcat tk009: hi 07:03
John5342 poelcat: hi 07:03
mcepl pong 07:03
* SMParrish joins the crowd 07:04
poelcat mcepl: was there a topic you wanted to cover? 07:05
mcepl was it? I was just thinking now about BugZappers/Meetings/Minutes-2009-Jan-13#Goals_for_Fedora_11 07:05
poelcat let's start there :) 07:06
poelcat has anyone thought any more about what we want to accomplish during Fedora 11 ? 07:06
mcepl I thought: 07:08
mcepl 5. help people to discover their set of components to follow 07:08
poelcat that sounds good 07:09
* jds2001 here, sorta 07:09
poelcat do you have ideas on how we can do that? 07:09
mcepl the question behind this is whether we want to support more “whole crowd goes after this set of bugs” (which seems to be prevalent notion here) or “this is a group of small bugmasters over some component(s) helping each other” (which is more style of work) 07:09
mcepl *my style of work 07:09
poelcat that is a good point 07:10
poelcat i think up until now because there is (what appears) to be a small group of people invovled 07:10
mcepl and behind that is my doubts about 1. -- how come we can go on de-duplication in crowd, when it (IMHO) requires a lot of experience with given component (try to make order out of NetworkManager bugs without spending endless hours with them) 07:10
poelcat we just tried to cover everything 07:11
mcepl and covered nothing ... 07:11
mcepl (sorry, cannot resist) 07:11
mcepl ;-) 07:11
poelcat :) 07:11
poelcat yes, it can feel that way sometimes 07:11
SMParrish My feeling is people should concentrate on apps they use and are familiar with. Should also get to know the packagers for those apps as well. Thats why I do only KDE related zapping 07:12
poelcat mcepl: would it make sense to create a list of "important" components and have people pick from tose? 07:12
poelcat those 07:12
mcepl yeah, that’s point 3. and I think it is the heart of the matter 07:12
mcepl (or we cannot make them, but we can suggest these as the most helpful etc) 07:13
poelcat for example, focus on triaging kernel bugs instead of that autogenerated "fever" stuff 07:13
mcepl well, yes, except I don't think we have enough manpower to do much difference in kernel bugs ... but yes 07:13
John5342 i still reckon that fever stuff should be auto assigned 07:14
poelcat John5342: i couldn't get the owner of that thing to respond to our request to do that 07:14
jds2001 John5342: me too, but we don't run it :( 07:14
mcepl well, yes, or that behind every assignment should be somebody specific who is responsible for passing the bug in the right direction (e.g., me & xgl-maint@redhat.com) 07:15
poelcat mcepl: something about that sounds wrong to me though... just because we can't make a difference we shouldn't do anything 07:15
John5342 was wondering if there was a way to auto assign bugs by a particular user considering they all come from that one user 07:15
poelcat granted I have felt that way too :) 07:15
mcepl John5342: oh sorry, you mean fever like PackageMaintainers/FEver ? 07:15
John5342 correct 07:15
jds2001 so supposedly ubuntu has been more successful than we have. 07:16
mcepl and gnome 07:16
John5342 wasnt aware there was another fever 07:16
jds2001 but I can't actually see why. 07:16
jds2001 therefore, I propose what anthropologists call "participant observation" 07:16
mcepl me too (and my managers as well :() 07:16
mcepl yeah, 07:16
poelcat i'd say they are because they have ongoing events centered around triage 07:16
skvidal jds2001: I thought scientists called that 'mucking up the results' 07:17
mcepl so, let's make some 07:17
poelcat and a lot more activity than us... at least in blogs i've seen 07:17
jds2001 skvidal: hehe :) 07:17
John5342 tk009 and i were also chatting after last meeting about more ways to bring more people to the cause and more importantly keeping those who do join. 07:17
jds2001 tk009 and i had a nice chat at fudcon too. 07:18
tk009 tk009 is a nut don't listen to him 07:18
jds2001 lol 07:18
poelcat John5342: do you think our "cause" is well defined and valuable enough? 07:18
John5342 tk009: lol 07:18
tk009 I am still very new here and see everything as a new person would 07:18
tk009 my thoughts are on that alot 07:18
John5342 poelcat: i think one of the bigger issues is that we dont seem to keep the people we do get coming in. 07:19
jds2001 and that's really valuable, actually. 07:19
jds2001 fedorabugs has 1000+ members 07:19
jds2001 package maintainers automatically get it. 07:19
poelcat sometimes I fear that even for the people that are involved they may feel like their efforts are in vain when nothing happens to the bugs they triage 07:19
jds2001 so that accounts for ~800 07:19
jds2001 the delta is 200 potential triagers (and some maintainers are potential triagers as well) 07:20
mcepl poelcat: they need to get into contact with the actual developers -- relationship is the key 07:20
poelcat IOW we have a problem getting more triagers and a potentially larger problem of maintainers ... this has been brought up on fedora-devel before 07:20
tk009 do evs want more triagers? 07:21
tk009 devs* 07:21
tk009 I get the feeling most dont 07:21
John5342 th009 and i were talking about re-organizing the wiki to try and make it easier and simpler for people to understad what they are doing and we should also make it even more clear what help and experience there is 07:21
poelcat mcepl: long term I agree.. how can we form these relationships quickly? what can goals can we set and accomplish for Fedora 11? 07:21
jds2001 skvidal: your thoughts here since you chimed in earlier? :D 07:21
skvidal jds2001: I was just playing peanut gallery, sorry. 07:21
mcepl poelcat: concerning maintainers I think it is inevitable -- we have to stop growing in the terms of number of packages somewhere -- we cannot outgrow Debian ;-) 07:22
jds2001 i know, and you got recruited :D 07:22
mcepl back to point -- what are the manageable and important components which needs to get into order? 07:22
jds2001 basically question is do you personally see triagers as able to have a positive impact, and do we want more? 07:22
mcepl *point 3. 07:22
poelcat let's cover mcepl's point and then jds2001 07:23
poelcat mcepl: how can we get this component list? 07:23
mcepl Releases/11/FeatureList -- and see where can we help? 07:23
poelcat i've asked a few people and not gotten any help 07:23
poelcat mcepl: so create a component list from each of the feature pages? 07:24
mcepl e.g., if you buy me a big bottle of Rhinerhisling I am willing to fight with ibus -- I am afraid (after my experience with scim) that it will be buggy as hell. 07:24
poelcat :) 07:25
* jds2001 ends up typing with IM's that I don't understand the output sometimes :/ 07:25
mcepl SMParrish: you can go through everything new in KDE and make sure there are no bugs missed, that -blocker, -target bugs are up-to-date etc. 07:25
poelcat so we start w/ a list of components from the feature list... what about the rest of the distro... how can we generate a list of say the top 100 or 500 components? 07:25
mcepl poelcat: after F11 GA, I think 07:26
poelcat ? 07:26
poelcat we want to set goals to accomplish by F11 :) 07:26
jds2001 number of bugs filed? 07:26
SMParrish Can do 07:26
mcepl I would rather make now something which makes the visible difference -- 100 components we are not able to cover 07:26
poelcat mcepl: how can we measure a visible difference? :) 07:27
mcepl by skvidal mentioning on fedora-devel how wonderful we helped him? (sorry, I cannot resist) 07:27
poelcat oh i see... you're saying 100 is too big 07:27
mcepl yeah 07:27
skvidal mcepl: you work for lenovo? 07:28
mcepl skvidal: apage, satanas!!! 07:28
mcepl (sorry, was just kidding) 07:28
skvidal ah 07:28
tk009 I decided on a personal goal since we need to act timely 07:29
tk009 since last tuesday 07:29
mcepl skvidal: BTW, you don't have problems with managing yum bugs? 07:29
tk009 I decided I wil ltry to triage 400 bugs this cycle 07:29
mcepl tk009: COOL!!! 07:29
skvidal mcepl: of course I do - we get a ton of them :) 07:29
tk009 these other idea's can be incorporated in to that 07:29
tk009 but I do feel we need to act 07:30
* poelcat starts printing pledge cards ;-) 07:30
poelcat I like it 07:30
mcepl poelcat, jds2001, etc.: would anybody want to help skvidal? ;-) 07:30
jds2001 i know yum fairly well. 07:30
skvidal mcepl: no one wants to help with yum bugs - it's like helping lepers 07:30
jds2001 dunno the internals, but I could learn that fairly quick I think :) 07:30
mcepl skvidal: yeah, that's us -- helping to leprose!!! 07:30
poelcat mcepl: so then we don't worry/focus on critical components like the kernel? 07:31
mcepl skvidal: no seriously, I see on Releases/11/FeatureList only Presto, is there any part of yum which is especially messy? 07:31
* poelcat notes presto is on its way to be dropped 07:32
mcepl poelcat: I don't know, but I think kernel could be chewed upon only in parts (KVM PCI Device Assignment)? 07:32
poelcat no update since 2008-09 07:32
wwoods speaking of "top N components" - has anyone ever made a list of the top ~30 components by number of bugs filed? 07:32
jwb i recall seeing that at one point 07:32
mcepl poelcat: or NM & wifi bugs? (I spent whole morning trying to understand interaction between Firefox and NM ;-)) 07:33
wwoods would be super-helpful for the problem-reporter idea I keep messing with 07:33
poelcat wwoods: good idea... i wonder if bz can generate that automatically 07:33
poelcat wwoods: is dkl nearby to answer that question? 07:34
jds2001 https://bugzilla.redhat.com/report.cgi?x_axis_field=&y_axis_field=component&z_axis_field=&query_format=report-table&short_desc_type=allwordssubstr&short_desc=&classification=Fedora&product=Fedora&long_desc_type=allwordssubstr&long_desc=&bug_file_loc_type=allwordssubstr&bug_file_loc=&status_whiteboard_type=allwordssubstr&status_whiteboard=&keywords_type=allwords&keywords=&deadlinefrom=&deadlineto=&bug_status=NEW&emailtype1=substring&email1=&emailty 07:35
buggbot <http://tinyurl.com/cr2249> (at bugzilla.redhat.com) 07:35
* jds2001 notes 0xFFFF has 14 bugs and none of them are related to 0xFFFF 07:36
jds2001 it's just first in the list. 07:36
poelcat right 07:36
mcepl jds2001: much more important question is -- what is 0xFFFF ? :) 07:37
jds2001 mcepl: i forget, some dwmw2 package i think :) 07:37
poelcat jds2001: does that link point to a report? i just end up at the "Reporting and Charting Kitchen" 07:38
jds2001 The Open Free Fiasco Firmware Flasher 07:38
mcepl jds2001: me too 07:38
jds2001 poelcat: yeah 07:38
jds2001 maybe irc cut it off 07:38
mcepl jds2001: could you post tinyurl, please? 07:39
jds2001 http://tinyurl.com/ck6fpa 07:39
jds2001 you could grab the CSV of that and sort it however. 07:40
poelcat yeah, i'll do that as a next step 07:40
poelcat so what are our ideas for goals so far? 07:40
poelcat i've got 07:40
poelcat 1) set a personal goal-- and meet it 07:40
poelcat 2) ? 07:40
* poelcat is a little fuzzy on what the goal is around what we've just been discussing 07:41
* jds2001 might try to whip up some python to generate that on the fly 07:41
* jds2001 too (fuzzy that is) :) 07:41
kulll__ ^_^ 07:42
John5342 only half mentioned so far but try and identify what people might be good at triaging and individuals concentrate on those 07:42
mcepl poelcat: to get a list of components which we should focus on 07:42
poelcat mcepl: okay... so it is a sub-part of the goal of "focusing on a limited set of components to triage for F11" ? 07:44
poelcat John5342: grow the ActiveTriagers list by ____ people ? 07:44
mcepl yup, components with the highest number of bugs are certainly interest (except for Package Review, which is another issue) 07:45
wwoods sort -k2 -t, -rn FILE.csv -> 07:45
wwoods kernel, anaconda, xorg-x11-drv-ati, selinux-policy, xorg-x11-drv-i810, NM, PK, s-c-n, pulseaudio, mkinitrd, gdm, evolution, selinux-policy-targeted, rpm, nautilus, firefox, xorg-x11-server, java, hal, compiz, yum, ooo, initscripts, glibc, xorg-x11 07:45
wwoods (that's F10, obviously) 07:45
mcepl wwoods: xorg* bugs are mine 07:45
John5342 poelcat: that too (growing active triagers) 07:46
mcepl anaconda is covered by Fedora QA, isn't it? 07:46
wwoods yeah - basically anaconda is alindebe 07:46
mcepl John5342: how would you like to do it? 07:46
John5342 mcepl: getting people in will probably be harder but we need to work on keeping people active rather than letting them fade away so to speak. 07:47
mcepl so, SELinux (always good idea -- leprose like skvidal's yum), NM & s-c-n, PK, PA, etc. 07:47
mcepl John5342: yeah, give them something to do ... 07:47
John5342 mcepl: amongst other things. also give them the best support we can to get them started 07:47
mcepl John5342: yeah, that's what #fedora-bugzappers is for, but let's keep on creating the list -- we are running out of the time soon 07:48
poelcat John5342: is there a way to quantify or make it more specific so that we'll know at the end and along the way how we are doing? 07:48
mcepl poelcat: which of the wwoods's list of components matches releases on Releases/11/FeatureList ??? 07:49
mcepl s/releases/features/ 07:49
John5342 poelcat: quantifying needs a bit more work. all i know is i saw about 4 new people in meetings since starting and only 1 remains 07:49
tk009 there is another 07:49
tk009 fcami 07:49
tk009 =) 07:49
mcepl tk009: yeah 07:50
John5342 tk009: thanks. i stand corrected. thats still only 2 out of 5 07:50
kulll__ elo 07:50
tk009 hello 07:50
tk009 there are 119 days left in this cycle 07:51
poelcat how about if we try to keep discussing our goals for F11 on the list ? 07:51
tk009 we will have to 07:51
poelcat tk009: yes :-/ 07:51
poelcat and then next week really should lock something down? 07:51
John5342 and also discuss in #fedora-bugzappers too if thats any better 07:52
poelcat granted a discussion on the list may end up all over the place w/ others chiming, but such is the nature of things and maybe we'll get something good that way too 07:52
poelcat John5342: good idea 07:53
poelcat quick check on last meeting followups... BugZappers/Meetings/Minutes-2009-Jan-13#Previous_Meeting_Follow-up 07:53
John5342 also the wiki might be a good place to put any more general stuff we come up with during the week 07:53
poelcat anything to report back on? 07:53
John5342 i got most of the F11 housekeeping page done. just tidying. 07:54
poelcat John5342: excellent! 07:54
John5342 jds2001: you still had a FIXME on the old page. i copied it accross. BugZappers/HouseKeeping/Fedora11 07:54
John5342 will update Template to match it once FIXMES have been tidied up. 07:55
John5342 same goes for new list names (transfer to fp.o) 07:55
poelcat John5342: i think the query for rawhide rebase needed to be modified to exclude bugs blocking next release trackers 07:56
mcepl poelcat: could we make sure, that curl -s -L http://tinyurl.com/dl9tb5 |sort -k2 -t, -rn ends up in the meeting record? 07:56
poelcat mcepl: you got it! :) 07:56
mcepl I mean including above the line .... 07:57
John5342 poelcat: thats what i understand too. have copied stuff like that so we can perfect it 07:57
poelcat i'll post the minutes like I usually do ... anyone should feel free to ammend or modify them! 07:57
mcepl sure 07:57
poelcat okay so next week we'll lock down our F11 goals and start to be greeted by more bugs as a result of the F11 alpha ! 07:58
poelcat anything else before we close? 07:58
tk009 we are done =) 07:59
poelcat thanks everyone! 08:00
* John5342 waves goodbye 08:00

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