16:00:50 #startmeeting 16:00:51 Useful Commands: #action #agreed #halp #info #idea #link #topic. 16:00:51 Meeting started Wed Jul 8 16:00:50 2009 UTC. The chair is jlaska. Information about MeetBot at http://wiki.debian.org/MeetBot. 16:01:04 #meetingtopic Fedora QA meeting 16:01:11 #topic Gathering 16:01:26 Hey folks, it's QA meeting time. 16:01:54 I believe we'll be without adamw today ... and f13 said he may be delayed a bit 16:02:22 dpravec: welcome! 16:02:42 hello :) 16:02:47 just waiting a few more minutes for any other folks to join 16:03:11 Proposed meeting agenda - https://www.redhat.com/archives/fedora-test-list/2009-July/msg00142.html 16:03:28 Notes will land at - https://fedoraproject.org/wiki/QA/Meetings/20090708 16:04:28 Jesse here on my phone 16:04:38 Oxf13: Hey there! ... what client you using? 16:04:54 do we have a wwoods? 16:05:09 some free pos 16:05:23 .leave #fedora-meeting 16:05:26 I'll likely bounce in out as edge flakes out 16:05:33 .part #fedora-meeting 16:05:51 zodbot will be back :) 16:06:14 * jlaska wonders if beland is around 16:07:20 * Viking-Ice alive.. 16:07:30 Viking-Ice: sweet, welcome! 16:08:16 hi, i'm here too. i'm new intern in fedora-qa in Brno. 16:08:20 okay, let's get things moving and we can rearrange the agenda as needed until wwoods arrives 16:08:44 kparal: Hi there, thanks for joining! 16:08:56 * wwoods appears 16:09:46 okay, with wwoods I think we have everyone 16:10:15 #topic Previous meeting follow-up https://fedoraproject.org/wiki/QA/Meetings/20090708#Previous_meeting_follow-up 16:10:21 First up, the sad face for me 16:10:24 crap. we took too long to gather. I have to head to the coffee shop. back in 30 min or so 16:10:42 Oxf13: okay, apologies for the start-up delay :( 16:10:45 * [jlaska] - update [[QA/Goals]] wiki document 16:11:20 no updates on this front for me ... it's still something I'm interested in drafting. 16:11:34 So folks are probably getting tired of seeing this, but I'd like to keep it for next week 16:12:04 for kparal and dpravec, who might not have seen my dragging feet on this topic ... 16:12:30 the background is we are interested in updating our QA/Goals wiki page to be more relevant and actionable 16:12:37 the current page is https://fedoraproject.org/wiki/QA/Goals 16:12:52 I've taken an action item to draft some ideas to this page and reach out for review on fedora-test-list 16:13:10 Unfortunately, I don't have any updates [again] this week 16:13:18 Have you sent that to the test list? or did I miss it? 16:13:42 rjune: the original Goals wiki was drafted by adamw and posted for review back in March (or April) I think 16:14:20 adamw and I have been discussing several different approaches to revamping this page, I just need to get those ideas on 'paper' 16:14:45 ok, so it was before I got back into fedora 16:14:56 possibly 16:15:05 that's been the last month or so. 16:16:04 okay, unless any additional thoughts about last week ... let's dive into the agenda ... 16:16:42 #topic ''Fit n Finish'' Display Configuration Test Day Recap 16:16:48 mclasen: are you around? 16:16:56 I am 16:17:01 sweet! 16:17:17 if you have a moment, was hoping you might give a brief recap of the Display config test day 16:17:30 sort of a what worked, what didn't ... what wouldw you improve on for future test days? 16:17:37 I don't have too much to say about it... 16:17:46 * jlaska gathers some links 16:17:50 there was a bit of participation in the irc channel throughout the day 16:17:57 not too much though 16:18:02 Wiki page - https://fedoraproject.org/wiki/Test_Day:2009-07-07_Fit_and_Finish:Display_Configuration 16:18:16 mostly it was me and to some extent adam testing stuff and filing bugs 16:18:24 and working on some fixes 16:18:39 adam == ajax? 16:18:44 yeah 16:19:17 from the looks of the wiki you've nailed quite a bit of fit'n'finish issues 16:19:26 we probably did not do enough advertisement for wider participation 16:19:31 I would have loved to see more non-desktop devel names besides those issues 16:19:34 mclasen: Noticed that you never included what to include on the report... 16:19:35 but rawhide is also still pretty raw at this point 16:20:07 I think it should be set as a mandatory issue for test days.. 16:20:29 Viking-Ice: there is no clear recipe here 16:20:45 Viking-Ice: mandatory might be a bit strong, but I agree it would be tremendously helpful for contributors 16:20:45 one thing I noticed halfway through the test day 16:20:46 is that field lacking from the Test day template 16:20:55 is that we collected a ton of 'issues' 16:20:59 but no positives 16:21:19 mclasen: from a test perspective ... collecting the issues is a positive :) 16:21:31 were you hoping for more fixing of the problems yesterday as well? 16:21:52 with almost all of the desktop team unavailable, no 16:22:24 mclasen: looking at the reporting bugs was this just cosmetic comments that were wanted for the test day ? 16:22:46 not sure I understand the question 16:23:02 mclasen: I was pretty jazzed about the test cases created ... I think we can solicit more testing throughout the week here if there are areas you think weren't touched on enough by your team 16:23:39 one problem I noticed with the test day template 16:23:40 or rather its naming 16:23:55 mclasen: like the font in the users-witch applet is bolder and out of contest with the rest of the Gnome's top panel etc.. 16:23:56 is that I can't prepare the test day page before I know the schedule 16:24:04 since the date is part of the name... 16:24:13 instead of my app crashed 16:24:19 reports 16:25:02 mclasen: yeah ... I don't think it's a hard and fast rule about having the date in the subject 16:25:09 perhaps we can loosen that restriction in the SOP 16:25:25 Viking-Ice: what I was hoping for is reports like: "I was trying to do give a presentation, so I followed these obvious steps...then this and that unexpected problem occurred" 16:25:54 a crash is an obvious bug to file, not really something you need a test day for... 16:26:32 so, I think we are going to try another fit and finish test day in 2 weeks the earliest 16:26:54 Viking-Ice: I've got a bit on the agenda about the debugging pages, perhaps that's a good point to review your thoughts on the fit'n'finish bug reports 16:27:05 I haven't figured out the next topic yet... 16:27:12 mclasen: that's a good spacing between them ... having the same group do them *every* week is tough to manage 16:27:29 mclasen: just a heads up, we do have a NetworkManager event scheduled for August for the main test day track 16:27:38 perhaps that woudl be good to couple with a fit'n'finish event nearby? 16:27:50 * jlaska looks at https://fedoraproject.org/wiki/QA/Fedora_12_test_days 16:28:42 dpravec: I know you've expressed an interest in desktop testing ... these fit'n'finish test days would be a great way to get a feel for the landscape 16:28:56 okay, mclasen to summarize ... 16:28:57 could be 16:28:58 yes, i will sure join 16:29:14 * Loosen the date restriction on the test day wiki name 16:29:24 * need more advertising for fit'n'finish days 16:29:41 * jlaska guilty of digging out from email most of yesterday :( 16:30:09 mclasen: are there any areas from yesterdays test day that could use more attention from testers throughout the week, who might have had a conflict? 16:30:49 I went through all the use cases we had outlined yesterday 16:31:09 but doing the same with other hardware might of course yield different results... 16:31:21 Viking-Ice raises a good point about perhaps lining up some guidelines for filing bugs on test days ... I don't know if we can make that a requirement, but perhaps something to visit 16:31:25 mclasen: okay 16:31:43 jlaska: one thing that may be useful wrt to bug filing 16:31:54 is to mention the test case in the bug 16:32:14 as I said earlier "I was trying to do ..." <- could basically be a reference to a testcase 16:32:47 right on 16:33:01 yeah, several of us would do that frequently while testing the installation matrix 16:33:25 google is releasing neatx (python based nx server) -- which might be interesting for remote desktop problem solving/testing... http://google-opensource.blogspot.com/2009/07/releasing-neatx-open-source-nx-servier.html 16:33:30 I don't know if the installer guys found it helpful, but I know that several weeks later, when I needed to retest the bug ... it was good to know what the _heck_ I was trying to test 16:35:10 dpravec: sounds like you might be much closer to neatx, I'm not very familiar with how that could be used. Definitely open to the idea if that helps solve problems 16:35:15 anyone else know much about that? 16:35:41 neatx is new, there is freenx nx server already in fedora, but its 'hacky' :) 16:36:31 it is 'vnc on steroids' -- much faster, useable remote desktop 16:36:55 (but the nx technology is based on fork of older xorg, which is problem) 16:37:06 interesting, perhaps a howto/blog is in order to to introduce the topic and how it can help testers? 16:37:16 hrmm, yeah I suspect that would be problematic 16:37:32 debian guys refused to package it... 16:37:34 thanks to this 16:37:45 (possible security problems) 16:37:52 which are hard to fix 16:38:06 okay, well we can come back to this outside the meeting 16:38:12 ok.. 16:38:29 sounds like there might be something there, but potential packaging issues lurk 16:38:49 mclasen: any other thoughts/concerns around yesterdays test day, or should we let you go for now? 16:38:58 it is already in fedora... 16:39:18 dpravec: oh my mistake 16:39:41 jlaska: no immediate thoughts; I'll try to send a summary to the list later, and I'll keep f-test-list in the loop for the next test days... 16:39:53 i might write some wiki page about it 16:39:56 thank you 16:40:01 dpravec: that's a great start! 16:40:10 okay ... next up ... 16:40:14 #topic AutoQA - update from wwoods 16:40:24 wwoods: you've got the floor 16:40:51 okay! So there's the test plan: https://fedoraproject.org/wiki/QA:Rawhide_Acceptance_Test_Plan 16:40:59 there's 4 cases remaining to write 16:41:00 * jlaska still in love with the plan 16:41:19 https://fedorahosted.org/autoqa/query?status=new&status=assigned&status=reopened&summary=~Create+test+case&milestone=israwhidebroken.com&order=priority 16:41:43 the first two - Core package {existence,closure} - require the Critical Path package list/depsolving tool 16:42:12 * mclasen thinks that any rawhide vetting plan needs to be measured against the latest prelink disaster... 16:42:16 last week skvidal wrote a tool to depsolve out a list of packages/groups: http://wwoods.fedorapeople.org/files/critical-path/resolve_out_deps.py 16:42:32 mclasen: no 16:42:42 yes, there should be a test plan that checks things like that 16:42:46 but no, this is not that test plan 16:42:55 * jlaska makes a mental note - prelink disaster 16:43:30 writing (and automating) this test plan will show us a clear path to writing (and automating) a test plan that would cover the prelink disaster 16:43:48 walk before we run, etc. 16:44:28 Hey I'm back 16:44:35 * jlaska tips hat to f13 16:44:37 getting back to *this* plan - I created two groups in f12 comps (critical-path-base and critical-path-gnome) 16:45:21 combining those two groups and @core should cover every use case listed in the critical path proposal: https://fedoraproject.org/wiki/Critical_Path_Packages_Proposal 16:45:41 (they don't seem to have appeared in rawhide comps yet, but they're checked into CVS, so it's only a matter of time) 16:45:45 * jlaska has 1 stone and looking for 2 birds 16:46:29 anyway, working with the assumption that we'll use those three package groups and a soon-to-be-completed tool to depsolve / check the packages required by those groups 16:46:45 I'm writing the Core package {existence,closure} test cases today 16:47:02 and RSN I'll write up the final two basic functional tests - network and yum 16:47:11 for the record I think we should create critical-path do all the *DE's or none.. 16:47:31 and by do I mean to.. 16:47:43 Viking-Ice: that's great. Feel free to create a critical-path-*DE group in comps that fills the use cases described in the proposal 16:47:46 with packages from your DE 16:47:57 the stuff in @core and @critical-path-base should apply to all DEs 16:48:01 that's actually a really cool way to extend this 16:48:17 wwoods: I believe there is a link already for how to recommend + review comps changes right? 16:48:19 but given that a) the critical path proposal requires a login manager, and b) gdm is the default login manager 16:48:28 I created a package group that includes gdm and its deps 16:48:51 setting up a group for DEs that don't use GDM is an exercise for the relevant SIGs 16:49:12 so is testing those paths 16:49:17 right on 16:49:29 http://fedoraproject.org/wiki/PackageMaintainers/CompsXml is general info about editing comps 16:49:30 QA/releng signing up to do "the" critical path is a big big task on its own. 16:49:53 but if I understand correctly, the current proposed solution leaves the door open to contributor assistance in those areas 16:49:56 for the record - current expansion of the current Critical Path definitions are here: http://wwoods.fedorapeople.org/files/critical-path/deplist.txt 16:50:04 it works out to 500 packages 16:50:12 or roughly... 5% of the distro? 16:50:16 which sounds about right 16:50:17 * mclasen wonders how such 'technical' comps groups will interfere with comps groups visibility in the ui 16:50:22 500 of 7000 16:50:29 f13: binary packages 16:50:30 not source. 16:50:41 mclasen: they're set as non-visible 16:50:42 mclasen: we can mark them as not visible in comps 16:50:47 or did mark them. 16:50:49 oh, didn't know that 16:51:05 false 16:51:19 also false 16:51:20 without that, it would make for an unpleasant group screen in the installer 16:51:21 so they don't get installed 16:51:46 a handy link for those interested in learning more about contributing to comps - http://fedoraproject.org/wiki/PackageMaintainers/CompsXml 16:51:56 see http://cvs.fedoraproject.org/viewvc/comps/comps-f12.xml.in?view=markup 16:52:44 (search for critical-path) 16:52:49 wwoods: what's on the plan for this week, are there areas you'd like review or engagement by f-test-list ? 16:53:17 plan for this week: finish writing tests, ask for review from f-t-l 16:53:37 and probably send out an informational message about the critical path deplist to f-d-l 16:54:09 hopefully I'll also have time to start trying to automate one of the simpler test cases in autotest 16:54:18 sweet! 16:54:35 speaking of autotest, f13 and I spoke earlier about some next steps identified by mmcgrath for hosting in infrastructure 16:54:40 so, fingers crossed, next week we may have at least some screenshots of green/red boxes 16:54:43 I've added some action items to the end of https://fedorahosted.org/autoqa/ticket/9 16:54:49 everyone loves green boxes, right? 16:55:01 the red boxes don't I think 16:55:04 be we don't like them :) 16:55:24 .wub 34 16:55:56 wub! 16:56:06 * wwoods doesn't know what's going on there 16:56:10 anyway, that's the AutoQA update 16:56:28 wwoods: sweet, sounds like good measured progress on the test case front 16:56:46 I know how tough those are to slog (word?) through, but the plan + cases are looking really solid 16:57:38 okay, unless any other questions, changing topics shortly ... 16:57:44 #topic Debugging pages ... what's next? 16:58:15 I don't know if beland is around, but he provided some data on f-test-list recently on some components to target as candidates for future Debugging pages 16:58:24 link - https://www.redhat.com/archives/fedora-test-list/2009-June/msg00723.html 16:58:45 Viking-Ice: I figured this would be near and dear to you since much of this spawned from your proposal 16:59:38 I didn't see notes for a BugZapper meeting this week, so I wanted to briefly discuss it here 17:00:08 Well yes added Dracut has been added to the category ( https://fedoraproject.org/wiki/Dracut/Debugging ) and btw I would like people to refrane from using the {{command| }} and {{filename}} in the wiki all this does is adding some gray smug that's not visible in all screens rather just bold it.. 17:00:45 Viking-Ice: yeah I loved the Dracut page you drafted and posted for review on the list, really nice work there 17:01:11 it would be nice, if we knew now what all of our test day topics would be ... we could target Debugging guides for each topic 17:01:12 mclasen: Example for the test day's on how to report the "All bug report" 17:01:19 unfortunately we don't have that yet 17:01:23 on the dracut's debugging page. . 17:02:11 Viking-Ice: did you consult with the dracut team when drafting that page, or is this what you discovered while testing? 17:02:31 I talked to Harald on irc.. 17:02:36 mainly 17:02:38 nice! 17:03:03 I think you just demonstrated how QA can add value here ... in drafting and proposing these pages for review 17:03:39 I know the maintainers are very busy making this stuff work ... and often it's hard to juggle all the wiki docs needed 17:04:17 Viking-Ice: do you have any suggestions on how best to go about building pages for the remaining components beland identified? 17:04:41 Wall all the the info I needed from harald ( and the only thing he provided me with ) was what they wanted on the report.. 17:05:14 In similar fashion to the Dracut's one 17:05:47 Viking-Ice: do you think it's worth filing fedora-qa tickets for these items, and asking for volunteers? 17:06:00 * jlaska would like to ... but worry it might be *too* much 17:06:24 ? 17:06:53 jlaska: imho fill them, but not every day... 17:07:46 i mean, do not repeat... 17:07:56 From what looks to be here https://fedoraproject.org/wiki/Category:Debugging I need to (re)write some of this stuff since i want more ( debugging ) info provided 17:08:18 Viking-Ice: are you comfortable using tickets to trac this stuff? 17:08:32 or do folks consider it too much tracking? 17:09:13 does and change matter talking about the trac system I dont get mails when ticket get filed I think.. 17:09:23 and by and I mean not 17:09:33 Viking-Ice: I belive you need to set that up in the admin page, I can look at that for you 17:10:36 and I don't think we have default owners for the current wiki component, I'd need to dbl check 17:10:49 Yes that was an interesting process how infra handled that I requested the trac system for QA you pushed after it and wwoods got notified and made admin.. 17:11:13 that the trac was ready 17:11:26 #action jlaska - update fedora-qa privs so that viking_ice is on the default bcc 17:11:53 okay, unless folks shout ... I think I'll file some tickets to record what debugging pages we should tackle based on the data beland provided 17:12:09 and then beg for volunteers on the list 17:12:22 ok 17:12:45 #action - jlaska files a bunch of wiki requests for content for debugging pages ... email list after 17:12:48 I'll review fill in if needed and add some info for components 17:12:59 Viking-Ice: that'd be great, thank you 17:13:16 #topic Open discussion - 17:13:23 okay, open floor time 17:13:46 any topics/concerns or issues that need debate? 17:13:53 or favorite jokes? 17:14:03 TCMS? 17:14:20 Three men walk into a bar and a working girl shows up.. 17:14:33 :) 17:14:45 Viking-Ice: okay, let's keep it clean! :D 17:14:50 oh, I know how this one ends, I have the DVD 17:15:01 i thought its all... 17:15:03 :) 17:15:11 jlaska: can you post the pgae you were working on with adam? I'm curious to see it. 17:15:22 that's the old page https://fedoraproject.org/wiki/QA/Goals 17:15:51 f13: re TCMS ... there isn't any news on the nitrate front ... they are still working on their own internal milestones 17:15:53 is there a new page? (I was actually working and not paying attention in here) 17:16:07 rjune: I have a draft on my desktop that's not posted yet 17:16:17 I'll be sure to send something to the list for review once I've got it posted 17:16:37 thanks. 17:16:44 f13: in recent news, mchua has offered help to configure a mediawiki+semantic instance for us to use during F12 17:16:55 f13: the infrastructure team kindly provided a play space and installed the packages needed 17:17:13 mchua and I have been playing phone tag a bit so haven't been able to sync up this past week. 17:17:39 mchua setup the laptop.org semantic extension to improve test results query+collection for them 17:17:42 ok 17:17:47 hurray for interns! 17:17:53 so while not a full fledged TCMS, I think of this as a short-term solution to get us over the hump 17:18:17 definitely, I don't think this falls under her normal responsibilities, so I definitely appreciate the time she can loan! 17:18:55 For Fedora 13, we should have some automation in place, I think we can then meet and discuss as a group how best to proceed with a TCMS 17:19:59 another quickie ... 17:20:14 since folks might have been distracted with FUDCon or holidays ... 17:20:18 https://www.redhat.com/archives/fedora-test-list/2009-July/msg00009.html 17:20:36 I posted feedback from the Test Day survey ... and started to outline some recommendations 17:21:01 I welcome any thoughts to the list 17:21:14 If no other topics ... we'll close things out in 4 minutes ... 17:24:47 T - 15 seconds 17:25:05 Okay gang, thanks for your time and input 17:25:06 maybe one small idea... we could use mumble to do audio meetings. 17:25:15 dpravec: sorry 17:25:19 I'm not familiar 17:25:47 its similar to irc, but its voip app. 17:26:06 it's just similar to ventrilo, teamspeak or skype rooms (?) 17:26:17 dpravec: this topic came up a few weeks back on whether we should use the Fedora talk VOIP server to host meetings 17:26:20 created originally for gamers, similar to teamspeak, but free software 17:26:21 interesting who's going to note down what was talked about ? 17:26:32 Viking-Ice: I'll send notes to the list for review 17:26:39 i can write small wiki page about it 17:26:49 then again we could provide audio file.. 17:26:50 Viking-Ice: fedbot records it 17:27:03 oh, nmind. I'm a tard 17:27:24 there were some concerns around accessibility when this topic last came up 17:27:31 * jlaska finds notes 17:27:42 yes imho mac version is missing :) 17:27:47 jlaska: was referring to this mumble idea.. 17:27:54 http://mumble.sf.net 17:27:56 not just that, but voice is much harder for people to participate in 17:27:59 Viking-Ice: yeah sorry :( 17:28:06 we can combine it to get best 17:28:06 dpravec: Fedora already has a VOIP solution, Fedora Talk 17:28:17 dpravec: http://talk.fedoraproject.org/ 17:28:27 also, voice meetings lose a /lot/ of archival uses, meeting logs and all that 17:29:04 I'm a fan of using the phone at times, but I know IRC is really popular in the fedora space and f13 mentions valid points 17:29:17 agreed 17:29:19 dpravec: we did have consensus around using fedora talk for FAD's 17:29:25 ok 17:29:27 np 17:29:29 which was a forum/format where that worked really well in 17:29:40 so a good idea, I think just trying to figure out where best to apply it 17:29:45 its hard to grep audio file :)) 17:29:51 hehe 17:30:06 okay gang ... going to wrap things up 17:30:11 thanks for your time and input!! 17:30:17 #endmeeting