From Fedora Project Wiki
(New page: = Bug Triage Meeting :: 2008-09-30 = == Attendees == * poelcat * John5342 * ke4qqq * jlaska * jds2001 == NEW->ASSIGNED == * Some people believe a different way should be implemented to i...)
 
 
Line 7: Line 7:
* jlaska
* jlaska
* jds2001
* jds2001
* mcepl


== NEW->ASSIGNED ==
== NEW->ASSIGNED ==

Latest revision as of 19:58, 6 October 2008

Bug Triage Meeting :: 2008-09-30

Attendees

  • poelcat
  • John5342
  • ke4qqq
  • jlaska
  • jds2001
  • mcepl

NEW->ASSIGNED

Status of Greasemonkey Scripts

  • Comphappy was making updates a few weeks ago
  • No one knows the status

Generating New Involvement

  • Do we state the benefits of bug triage anywhere?
  • Creating consistent meeting summaries and sending back to the list
  • Publish "success stories"
  • Need to resolve controvery about NEW->ASSIGNED as that may discourage people from helping out
  • Unclear if weekly meeting reminders help

FEver Bugs

IRC Transcript

poelcat who is here? 07:03
* John5342 is here 07:03
poelcat hi John5342 07:03
* ke4qqq is here 07:03
poelcat ke4qqq: hi 07:04
poelcat jds2001: stickster jlaska anyone else? 07:05
* poelcat runs to grab coffee 07:05
* jlaska waves :) 07:05
stickster poelcat: I'll be around as soon as I can, dealing with release stuff 07:05
ke4qqq mmmm coffee.... 07:06
* ke4qqq runs for coffee too 07:06
* jds2001 here 07:08
poelcat how is everyone feeling about bug triage? 07:08
poelcat how are things going? 07:09
poelcat anything in particular we'd like to cover today? 07:09
jds2001 havent had time really 07:09
jds2001 yes, your whole ML thread :) 07:09
poelcat jds2001: that too :) 07:09
poelcat jds2001: okay the mail thread 07:10
poelcat ke4qqq: John5342 are you aware of what we are referring to? 07:10
poelcat the thread about a "Confirmed" state or not moving from NEW --> ASSIGNED 07:11
John5342 poelcat: yes i am. 07:11
ke4qqq honestly no: at least nothing jumps to mind 07:11
ke4qqq link? 07:11
poelcat John5342: did you have anything thoughts or opinions? :) 07:11
jds2001 ke4qqq: https://www.redhat.com/archives/fedora-test-list/2008-September/msg00493.html 07:12
John5342 poelcat: well i agree something should be done about it but which solution is best is really down to the general opinions of everyone involved 07:12
jds2001 and the long ensuing discussion :) 07:12
poelcat well that thread raises two topics :) 07:13
poelcat 1) I'd like to suggest we use fedora-test-list for more traffic on bugs we aren't sure how to triage 07:13
poelcat i think this is one way to create more awareness and also get help 07:13
poelcat from people that can't or don't otherwise participate 07:14
poelcat 2) i am still not convinced how critical the issue is around NEW --> ASSIGNED 07:14
John5342 poelcat: i was refferering to the second but agree with the first aswel 07:14
poelcat IOW if we were to survey people on a scale of 1 to 5 where 5 is "this incredibly horrible and must change to day" 07:15
poelcat and 1 is "i don't care" 07:15
poelcat what would the average score be? 07:15
jds2001 dunno, we could ask G to hold an election :) 07:16
poelcat it seems like there were 3 or 4 people who would haved voted the issue as a "2" or "3", but not higher 07:16
poelcat jds2001: interesting idea 07:16
* poelcat thinks range voting might be unhelpful though ;-) 07:16
John5342 maybe just collect all the possible solutions and a no change and elect on that? 07:17
poelcat i'm still trying to convene with the bugzilla experts internally to understand what the smartest possibility is 07:17
poelcat and how it might affect other product workflows 07:17
poelcat John5342: great idea! 07:17
* poelcat likes that a lot 07:18
jds2001 but FESCo has to agree on whatever we do, I think. 07:18
poelcat so the current possible enhancements are? 07:18
John5342 but ofcourse if people vote for no change then we also have to teach people what the right way is so we dont get people complaining 07:18
poelcat jds2001: yes 07:18
jds2001 we should probably pre-vet everything by them.... 07:18
poelcat jds2001: at what point? 07:19
* poelcat thinks we can collect survey data and ideas up until the point we recommend a change to BZ 07:20
jds2001 before we go through a charade of an election that FESCo says "no way" to the results of :) 07:20
jds2001 yeah :) 07:20
* poelcat is however still a little reluctant to commit the time to this campaign given all the other things I have going on 07:20
* jds2001 doesn't personally see that happening, though. 07:21
poelcat so possible solutions are: 07:21
poelcat 1) a new STATE 07:21
poelcat 2) a new Flag 07:21
poelcat 3) ? 07:21
John5342 Re-using an existing flag was also suggested 07:21
* jds2001 goes for the flag, the UNCONFIRMED state was specifically eradicated from RHBZ 07:22
-!- Irssi: #fedora-meeting: Total of 118 nicks [1 ops, 0 halfops, 0 voices, 117 normal] 07:22
* jds2001 not sure why, though. 07:22
* stickster thinks he just got done with release duty 07:24
jds2001 oh, i thought that was on Thursday 07:24
poelcat okay, i'll try to get some info from the bugzilla people's this week 07:25
* jds2001 may be going crazy again. 07:25
John5342 and ofcourse if there is a new state it needs to be decided if there is a new one for triaged or accepted 07:25
poelcat has anyone seen comphappy around and/or know if he updated the greasemonkey scripts? 07:25
poelcat John5342: correct 07:25
poelcat oh i forgot 3) "Triaged" keyword 07:25
poelcat anyone re: greasemonkey scripts? 07:26
poelcat if not we can move on 07:26
John5342 havent even got the greasemonkey script working yet. not sure its the scripts fault though 07:27
poelcat John5342: comphappy was going to add some additional buttons 07:28
poelcat okay next topic; any ideas out there on how to grow interest and participation in our meetings and bug triage in general? 07:29
John5342 poelcat: i heard about that but both the old and new ones are giving me no buttons at all at the mo but having problems with some other gm stuff too so probably that 07:29
jds2001 geez not sure what my problem is today 07:33
ke4qqq poelcat: so I don't know how we should attract people - I have tried showing the how little the time committment can be.... 07:33
ke4qqq and also showed them how relaxing triaging is - like popping bubble wrap.....but I haven't gotten any significant buy-in 07:34
jds2001 what did i miss? 07:34
poelcat jds2001: the crickets ;-) 07:34
John5342 poelcat: the uncertainty of if they are helping or hindering might be something to sort. was an issue for me when i started. people might give up before they start 07:34
poelcat jds2001: 07:29:12 < poelcat> okay next topic; any ideas out there on how to grow interest and participation in our meetings and bug triage in general? 07:35
ke4qqq interesting - we talk about the mechanics, but is there something that concretely lists the benefits. 07:35
poelcat ke4qqq: good point... i'm sure we did at one time, but I also know it would take me a bit find where that is 07:36
John5342 the bug state issue on the ML is one thing that will stop some complaints from maintinaers that could put people off. 07:36
* ke4qqq doesn't recall seeing that in the presentation or on the join-type pages. 07:37
poelcat John5342: true... so the idea that bug triage is non-invasive... we're just adding a little more metadata to the bug ? 07:37
poelcat Is it a good idea to send the meeting reminders to fedora-test-list each week or is that just spam/noise? 07:38
John5342 poelcat: effectively. one problem is getting people in in the first place but the other is to get members to do the stuff without fear of just getting in the way 07:38
* poelcat also realizes we could create a better echo affect by posting meeting summaries back to f-t-l 07:39
John5342 poelcat: and a complaint from a maintainer about them doing something differentlyu from the norm is just one thing that can put members off 07:39
poelcat so people realize that we met AND that something is going on 07:39
poelcat John5342: right, make people even more hesitant :-/ 07:39
John5342 exactly 07:40
John5342 new state or tag etc instead of assigned will stop one more reletively common complaint that can cause that 07:41
jds2001 around release time we attract a flamefest, too. 07:41
jds2001 but that's not really the actions of any one person, that's our "triage bot" :D 07:42
poelcat jds2001: which we can minimize if we can get up to date on all the NEW rawhide bugs 07:42
* poelcat continues to find a lot of NEW rawhide bugs that should have "FutureFeature" keyword 07:42
poelcat and if a bug has it, then it won't get rebased 07:43
poelcat i'm also adding it to bugs where it is clear the maintainer wants it kept in rawhide 07:43
jds2001 that makes sense. 07:43
poelcat even though it might not exactly be a pure a RFE 07:43
John5342 poelcat: noticed quite a few of the NEW bugs are from FEver. was wondering if its worth asking the person behind it if FEver can be modified to set appropriate state since all of it should always be ASSIGNED and FutureFeature 07:44
poelcat what are those anyway? 07:45
John5342 they are reminders of upstream updates 07:45
John5342 automated 07:45
* jds2001 noticed a thread about 07:46
jds2001 fever spam and apologiizing for it. 07:46
John5342 http://fedoraproject.org/wiki/PackageMaintainers/FEver 07:46
jds2001 but yes, I think they should ALWAYS be FutureFeature 07:46
* poelcat wonders if they should automatically go to ASSIGNED or if we still need to manually triage them 07:47
poelcat just for a sanity check 07:47
John5342 they are requested by the maintainer and submitted by them too so should always be going to correct maintainer and versions appear to be chacked against current packages so should all be correct 07:48
John5342 so i would have thought they should be ASSIGNED. 07:49
poelcat John5342: do you know who is leading the FEver effort? 07:52
John5342 there is an email address on the page i posted ^^^ 07:52
John5342 [[MailTo(mr DOT ecik AT gmail DOT com)] 07:53
poelcat okay 07:53
poelcat any other topics before we close? 07:54
mcepl ehm 07:55
poelcat mcepl: hi :) 07:56
mcepl poelcat: I still believe that the best way to make bug triaging attractive is to have success stories. So, how are we doing about those Rawhide NEW bugs (and yes, i am very much in fault for not helping more with these)? 07:57
poelcat mcepl: we are keeping the number < 600 07:58
poelcat i usually get to 40 or 50 a week 07:59
poelcat but it doesn't seem like enough to stem the tide :) 07:59
mcepl ooelcat: torn between the two (concerning FEVER) -- probably they should be very much triaged! That should be the point where upstreamed bug (no, it shouldn't be CLOSED/UPSTREAM but something else, but that's another issue) should be checked whether they are not actually fixed by the new upstream release. 07:59
John5342 unfortunately i was away last week so wasnt much help either but hope to get a few done this week 08:00
mcepl well, it is very well enough -- last time I checked we were around 1500 or something, right? 08:00
poelcat mcepl: that is filtering out component = Package Review 08:00
poelcat which I think is a whole other issue 08:00
mcepl so, if we can get agreement with package maintainers -- "you will make us testing upstream build asap, and we will close you upstreamed bugs". But all that is probably when bug triage is working better and we will have nothing todo. 08:01
poelcat as it taints our true bug stats by using a bug tracking tool as a workflow tool 08:01
poelcat mcepl: right 08:01
poelcat mcepl: is use this query to locate new rawhide bugs to triage http://tinyurl.com/6llac8 08:02
mcepl well, it IS workflow tool. And actually, cannot bug triagers who are packagers as well close some of those Package Reviews. I know it is silly and boring, but we should get rid of them sometimes. And nobody is likely to do it in our place. 08:03
poelcat mcepl: i thought we agreed a while back to stay away from the Package Review component because the really don't start work 08:04
poelcat until state change to ASSIGNED 08:04
poelcat funny that should come up again :) 08:04
jds2001 agreed, that triagers who are packagers should be doing reviews :) 08:06
mcepl well, I meant that WE should take over the bugs and make reviews. But maybe later. 08:06
jds2001 well that and sponsors 08:06
jds2001 (which after doing a number of reviews, you can become a sponsor)...... 08:07
* poelcat needs to go 08:09
mcepl which reminds me, when it come up again -- instead of silly CONFIRMED state, don't we want to do something about CLOSED/UPSTREAM? I think it is really silly and we should get rid of it. 08:09
poelcat anything else before we wrapup? 08:09
mcepl OK 08:10
mcepl next time 08:10
poelcat mcepl: can we discuss next week? 08:10
poelcat okay :) 08:10
mcepl sure 08:10
poelcat thanks for coming everyone 08:10
John5342 poelcat: thank you too. 08:10
John5342 see you next week everyone 08:10

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