15:03:03 #startmeeting Fedora QA meeting 15:03:03 Meeting started Mon Apr 25 15:03:03 2016 UTC. The chair is adamw. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:03:03 Useful Commands: #action #agreed #halp #info #idea #link #topic. 15:03:03 The meeting name has been set to 'fedora_qa_meeting' 15:03:05 nirik++ 15:03:14 #meetingname fedora-qa 15:03:14 The meeting name has been set to 'fedora-qa' 15:03:16 .fas jdulaney 15:03:16 handsome_pirate: jdulaney 'John Dulaney' 15:03:20 #topic Roll call 15:03:25 * garretraziel is here 15:03:28 .fas jdulaney 15:03:28 handsome_pirate: jdulaney 'John Dulaney' 15:03:29 ahoy everyone, who's here for qa meeting times? 15:03:33 Y'all see my shiny new email? 15:03:35 :) 15:03:41 * tflink is here 15:03:58 * kparal is here 15:04:15 * brunowolff is here 15:04:17 handsome_pirate: nice 15:05:06 * pschindl is here 15:05:09 hi everyone 15:05:24 sorry if i make no sense this morning, i got in from lfnw 8 hours ago 15:05:25 hi adamw 15:05:50 adamw: you could join maxamillion for his broken coffee maker :) 15:06:03 heh, the adams are on poor form this morning huh 15:06:05 #topic Previous meeting follow-up 15:06:41 #info "adamw to get dennis' thoughts on https://fedorahosted.org/rel-eng/ticket/6383" - damnit i still somehow didn't do that, dgilmore, could you put your notes in that ticket? we've talked about it in IRC I know 15:08:07 #info "adamw to tweak i18n test day page a bit so it's clearer that tests are found on the app page" - I did that, the test day is all done now and went pretty well I think 15:08:20 "sumantro to work on test cases for the media writer Test Day, adamw to guide and help out" - i'd say that went pretty well right sumantro? any notes? 15:08:46 yes adamw , went pretty well. !! 15:09:25 in bodhi now for update testing and most of the bugs are fixed 15:09:32 great 15:09:59 #info "sumantro to work on test cases for the media writer Test Day, adamw to guide and help out" - this was done, we had a good test day with a lot of response and bugs found, and an update for the tool is in testing now 15:10:14 any other previous meeting follow-up i missed that isn't covered on the agenda? 15:11:54 okey dokey 15:11:59 #topic Live Media Writer status 15:12:03 oh, before i forget... 15:12:09 #chair handsome_pirate tflink 15:12:09 Current chairs: adamw handsome_pirate tflink 15:12:41 so, the big news here is - AIUI - that live media writer will no longer be the 'default download' for F24, due to legal issues with providing the sources for windows 15:13:05 wait 15:13:06 but it's still "officially supported" for Linux, I believe 15:13:08 i believe the rewrite is still going ahead and we'll link to the new version from the docs and so forth, but the download page will just give you an ISO as before 15:13:09 How does that work? 15:13:16 so the blocker bugs still apply, I'd say 15:13:30 s/supported/recommended 15:13:36 handsome_pirate: http://lists.fedoraproject.org/archives/list/desktop@lists.fedoraproject.org/message/6LJHMCE4HT4MYBBAGK64R5TE3UHSZTLX/ 15:13:44 kparal: yeah, i agree 15:14:20 handsome_pirate: basically for compliance we need to be sure we can provide the correct sources for all the bits in the 'stack' of the windows build, which includes python and stuff and is apparently hard, there's lots of discussion in that thread 15:14:57 the reason this hasn't been a problem up till now is that technically fedora doesn't distribute the tool for windows, the windows download links have always just pointed to someone's personal space 15:15:07 ah 15:15:16 but if we start putting it as the official download on fedoraproject.org, legal requires us to be a lot more strict about compliance 15:15:34 anyhow, yeah, we don't need to get too far into it here, just know about it :) 15:15:41 Couldn't we cross build? 15:15:43 nice tl;dr 15:15:50 kparal: well its not going to be built and shipped and supported by fedora 15:15:57 * satellit sorry late 15:15:59 kparal: it will be mbriza that is doing it all 15:16:22 dgilmore: I understood that decision being only about the windows build 15:16:23 dgilmore: sure, but that was the same before and we still documented its use in the wiki and stuff 15:16:35 right, aiui the linux build is still gonna be in the repos, right? 15:16:53 it makes no sense to go back to old LUC 15:16:54 kparal: right but the linux side is just a yum install 15:17:22 * satellit add a flag to go to the old gui LUC for windows? It did work 15:17:35 kparal: so its not really providing a download, and more docs 15:17:56 ok, so I think we agree 15:18:10 windows stuff is not blocking, linux stuff is 15:18:13 satellit: there will be windows builds of the new one, just not provided by fedora, the same as the old one was not provided by fedora 15:18:40 kparal: well not really blocking, as fixes will need to go in f22 and f23 15:18:50 yeah, we now have a concept for that 15:18:52 which puts it in weird space 15:19:02 it's Acceptd 15:19:08 it's AcceptedPreviousRelease 15:19:37 I have adjusted the SOPs some time ago, after weeks and months of asking for feedback on the mailing list 15:19:49 yeah, we have a whole process for it now 15:19:50 yay process 15:20:06 things are so much better defined than, say, the F13 days 15:20:11 alright, i think we beat this horse 15:20:45 #topic Workstation graphical upgrade planning 15:20:52 oh damn i meant to ping mclasen and hughsie for this one 15:20:57 lemme see if i can find them 15:21:26 meant to let them know in advance, forgot 15:21:30 #agreed adamw's an idiot 15:22:50 hi hughsie, thanks for coming 15:22:51 adamw: maybe you'll show up in www.irclogsfromlastnight.com 15:22:56 kparal: haha :) 15:23:01 if that doesn't exist yet, make it 15:23:25 #info kparal to do www.irclogsfromlastnight.com 15:23:31 =) 15:23:46 okay, so quick summary for anyone who's not up to speed 15:24:20 graphical upgrades for Workstation are an F24 Change: https://fedoraproject.org/wiki/Changes/GraphicalSystemUpgrades 15:25:03 the workstation folks have been working on this and have it in early testable shape, but it's unlikely to be the case that all the bits will be in stable repositories by beta go/no-go, i.e. meeting the requirements for blocker bugs 15:25:52 * hughsie-afk did the gnome-software 3.20.2 tarball today if that helps 15:26:05 just waiting for kalev to do the new libhif fix for the dnf mark blocker 15:26:10 so we kinda have to decide if we consider 'graphical upgrade is not yet testable from official repos' as a beta blocker, and if not, we should come up with a coherent plan for how we talk about it at beta time, and how we test it and deploy it i guess between beta and final 15:26:45 adamw, the -> stable bit is the part that doesn't make sense to me 15:26:52 is that thing going to be in official upgrade docs? as the first option? or marked as experimental? 15:26:54 updates-testing makes sense 15:27:29 i believe the idea is that for f24 final it will be the official, recommended, documented, supported etc etc mechanism for Workstation upgrades 15:27:53 as of *right now* it is not, the wiki page (which is still the reference doc for this afaik) just documents dnf-system-upgrade 15:28:13 so this sort of falls slightly between a release validation/criteria-ish thing and a Change-ish thing 15:28:18 in that case all release criteria should apply, and we generally require everything to be in stable updates 15:28:47 hughsie-afk: the thinking behind that requirement is kind of tied to things that are more clearly *bugs* 15:29:07 i.e. we don't it to be the case that you have to know to update your f23 system from updates-testing before running an upgrade or else the upgrade will explode and eat all your data, or whatever 15:29:07 adamw, right 15:29:21 well it's not the problem in this case, but if it is not in stable updates, somebody might use an older broken version to upgrade 15:29:27 adamw, no requirement to update before upgrading any more 15:29:27 in this case, there is no older version 15:29:30 yeah 15:29:47 so i don't like fudging things, but otoh, the reasoning behind the requirements doesn't really apply in this case 15:29:59 for me the biggest concern is making sure we do get enough testing of the new mechanism before final 15:30:08 if it helps, i've also emailed kalev asking to take over the libhif blocker 15:31:11 so ignoring process *for the moment*, just looking at the issue on its merits, i'd be fine i think if we have the graphical method testable (but requiring manual steps to enable/use it) at Beta and we can document that, and maybe schedule a test day for it between beta and final 15:31:37 we should do a test day, agreed 15:31:48 adamw, atm, the manual steps are "enable my copr; update gnome-software; wait" 15:31:56 * hughsie-afk is up for that too if that helps 15:32:04 on the process front i guess i'd say that this is more of a 'late Change implementation' (which comes under FESCo's purview) than a 'beta release blocker' 15:32:18 but my on-the-merits evaluation might be influencing that 15:32:27 hughsie-afk: do you also have to teach it to consider 24 a valid target still? 15:32:33 not that it matters that much, just curious 15:32:57 what does everyone else think? 15:33:11 it would be nice to let fesco agree to what we decide here 15:33:25 Yeah 15:33:25 adamw, yes, if you don't want to wait 15:33:36 * handsome_pirate is +1 fesco on this one 15:33:46 hughsie-afk: has fesco discussed it? 15:33:57 they're supposed to review Change progress... 15:34:04 It kind of makes me nervous 15:34:16 also, it's marked as self contained change, but it seems to me to be a system-wide one, even affecting older fedora releases 15:34:19 adamw, i don't think so; i'm not the best person to ask 15:34:29 ok, gimme a sec to look at recent fesco meetings 15:34:30 so one more reason for fesco to bless it 15:34:45 * hughsie-afk tries to spend more time on bugs and features than process... 15:35:11 as far as I know, it has not been discussed on the last FESCo meeting 15:35:37 there were discussed only Changes, not marked as "100% completed" 15:36:36 yes, and you punted them to 04-29, which seems...late. :P 15:36:42 but this was actually marked as ON_QA so wasn't considered 15:36:42 https://bugzilla.redhat.com/show_bug.cgi?id=1308538 15:36:54 hughsie-afk: the only problem with that is then everyone *else* has to spend time on the process... 15:37:22 adamw, i'll just do what i'm told 15:37:24 "just trust everything to work out OK in the end" did not historically work out 100% 15:37:36 so, OK 15:38:35 adamw: I can open this topic on the FESCo meeting this Friday 15:39:38 proposed #agreed QA considers the state of the GraphicalSystemUpgrades Change to be a 'late Change' situation not a release criteria/validation situation and we'll file a FESCo ticket asking them to evaluate it and see if they're OK with where it's at. we'll proceed under the assumption that it's going ahead, and try to ensure it's easily testable at beta release time and organize a Test Day for it shortly after Beta 15:39:55 works for me 15:41:24 just for the record, the Changes policy does state that "if a change is to be enabled by default, it must be so enabled at Change Completion deadline" , and that was the deadline that occurred on 2016-02-23, so i'd say this is definitely late by that definition and shouldn't have been marked MODIFIED at that time 15:41:24 ack 15:41:25 but oh well 15:42:04 obviously we shouldn't have been offering F24 Alpha upgrades to F23 users at that point but it should have been in F23 and possible to use with a single dconf flip or whatever 15:43:27 I've got to run 15:43:40 But, I note that this whole thing makes me nervous 15:43:50 yeah 15:43:55 I'll go with what fesco decides, though 15:44:00 See y'all later 15:44:06 anyone else got an ack/nack? 15:44:24 ack 15:44:30 thanks handsome 15:44:59 adamw, would a gsetting flip be what you want in the future rather than editing the .json file? 15:45:46 * hughsie-afk can do that if you file an upstream bug 15:46:11 hughsie-afk: i kinda think so, yeah? a 'offer unstable releases' switch or something, which would cause it to consider 'Under Development' releases as valid targets 15:46:18 dunno if you'd ever want to offer rawhide, easy enough to filter it out if not 15:46:37 will do 15:46:48 well, rawhide is a valid target 15:47:03 i'd just have a key you can set to "25" or something 15:47:03 why exclude rawhide, let's keep it in 15:47:05 #agreed QA considers the state of the GraphicalSystemUpgrades Change to be a 'late Change' situation not a release criteria/validation situation and we'll file a FESCo ticket asking them to evaluate it and see if they're OK with where it's at. we'll proceed under the assumption that it's going ahead, and try to ensure it's easily testable at beta release time and organize a Test Day for it shortly after Beta 15:47:10 kparal, agreed 15:47:12 okay, we can figure the details out anyway 15:47:35 sumantro: do you want to take on organizing the test day? it should be a fairly simple one 15:47:48 though it'll likely be pretty popular 15:47:51 yes sure adamw 15:47:54 cool thanks :) 15:47:58 :) 15:48:14 #action sumantro to organize Workstation graphical upgrade test day, adamw to help out 15:48:43 #action adamw to file ticket on easier way to enable upgrade to branched/rawhide (rather than editing a json file) 15:49:02 #action adamw to file fesco ticket requesting them to evaluate Workstation graphical upgrade Change status 15:49:08 thanks a lot hughsie 15:49:18 anything else on this one? we still have at least one more topic to squeeze in so make it quick :) 15:49:39 nothing here 15:50:20 alrighty 15:50:30 #topic Fedora 24 status and test planning 15:51:01 OK, so we got a new nominated compose with i386 and other significant fixes 15:51:13 only problem with it is that Workstation live lost the live image compose lottery 15:51:37 still, we can test everything else, so everyone please do; we still have a lot of beta coverage to complete 15:52:25 * kparal will try to convince some people 15:53:04 #info go/no-go is this Thursday, everyone please help test the current nominated compose, treat it as a late TC/RC 15:53:17 i'll try and work on the server tests, though I can't do AD testing :/ 15:54:14 sumantro: have you tried validation testing yet? 15:54:41 no not yet. but can give it a shot. wont be a biggie . 15:54:55 yeah, if you and arvind could help that would be great, it's the most important thing for this week 15:55:17 sumantro: just ask in #fedora-qa if you miss anything 15:55:25 sure , I am all in it. just tell me what your plan is. :) 15:56:08 sumantro: keep the 'current' pages bookmarked - https://fedoraproject.org/wiki/Test_Results:Current_Installation_Test , Current_Base_Test, Current_Server_Test, Current_Desktop_Test, Current_Cloud_Test - and https://www.happyassassin.net/testcase_stats/24/ is also handy to see what tests most need running (it tells you when each test was last run) 15:56:17 testing instructions are included in the pages 15:56:33 you can submit results by editing the pages or by using relval - 'dnf install relval', 'relval report-results' 15:56:49 same instructions go for everyone else who hasn't done validation testing before :) 15:57:25 Got it adamw 15:57:28 cool! 15:57:47 ok, i don't think we have time for the 'test day planning' topic and anyhow it was mostly the upgrade thing i was gonna suggest there 15:57:50 so let's do a very quick: 15:57:51 #topic Open floor 15:57:59 we have blocker review in 3 minutes (did i remember to announce it?) 15:58:00 I'd recommended submitting results by hand by editing the wiki if you're new, with scripts it's easier to mess stuff up more 15:58:09 kparal: funny i was gonna say the opposite 15:58:14 kparal: it's easy to mess up wiki syntax 15:58:28 alright, both arguments are there, pick your choice :-> 15:58:30 relval should be easier for a newbie, it's quite hard to do anything really silly with relval report-results , it can't mass-submit or anything 15:58:40 report-auto you could do bad things with, but that is gone now :) 15:58:43 adamw: ok, I thought it could 15:58:48 nah, one result at at ime 15:59:23 I just would like to ask who is going to be on Go/No-Go & Readiness meeting on Thursday from QA ? 16:00:14 Me jkurik 16:00:23 timing ? 16:00:28 me too 16:00:58 timing: https://apps.fedoraproject.org/calendar/meeting/3877/?from_date=2016-04-25 & https://apps.fedoraproject.org/calendar/meeting/3878/?from_date=2016-04-25 16:01:09 depends on yout timezone 16:01:13 ok, we're over time, so thanks, everyone 16:01:15 thanks 16:01:24 i'll start blocker review meeting up immediately and leave roll call open for a few minutes 16:01:30 see you in #fedora-blocker-review! 16:01:34 #endmeeting