QA/Meetings/20121210

Attendees

• adamw (150)
• j_dulaney (36)
• tflink (22)
• Viking-Ice (21)
• kparal (20)
• jreznik (14)
• nirik (11)
• maxamillion (9)
• zodbot (8)
• jskladan (3)
• wwoods (3)
• satellit (3)
• BobJensen (3)
• pschindl (3)
• mjg59 (2)
• Martix (2)
• jsmith (2)
• mkrizek (1)

Agenda

• Previous meeting follow-up
• Fedora 18 Final status
• Enterprise storage support for F18 / F19
• Test case / criteria revision
• Open floor

Previous meeting follow-up

• adamw to brief #fedora ops and fedora-user-list regulars on fedup - he talked to some people, maybe not everyone
• adamw to put 'enterprise storage support in newui' on next week's agenda - done!

Fedora 18 Final status

• Current schedule was: change deadlines 2012-12-11, go/no-go 2012-01-01, release date 2012-01-08
• Final TC1 was available and viable for testing, but not Secure Boot-enabled
• Blocker list was substantial, at 28 proposed
• fedup GUI still was not available, some fixes for rough edges pending

Enterprise storage support for F18 / F19

• Enterprise/advanced storage types that require interaction to enable/use would not work via UI in in F18, but could be used via kickstart. UI is scheduled to re-appear in F19
• Agreed that this needed to be documented for final release

Test case / criteria revision

• RPM package error proposal positively received, should be put into practice
• Agreed that we couldn't agree on a kickstart criterion, instead agreed to handle kickstart bugs on a case-by-case basis for F18
• pschindl was working on test case adjustments for Final

Open floor

N/A

Action items

• tflink to clarify use of his fedup side repo
• viking-ice to let docs team know about advanced storage for release notes
• adamw to put packaging error criterion revision into production
• kparal to follow up on test@ discussion to explain the plan, and let anaconda team know

IRC Log

adamw	#startmeeting Fedora QA meeting	16:00
zodbot	Meeting started Mon Dec 10 16:00:24 2012 UTC. The chair is adamw. Information about MeetBot at http://wiki.debian.org/MeetBot.	16:00
zodbot	Useful Commands: #action #agreed #halp #info #idea #link #topic.	16:00
adamw	#meetingname fedora-qa	16:00
zodbot	The meeting name has been set to 'fedora-qa'	16:00
adamw	#topic roll call	16:00
* jreznik is here		16:00
* satellit listening		16:00
* jsmith is lurking, and having internet issues		16:00
adamw	welcome to the AGM of the grocery store bicyclists' association	16:00
adamw	first on the order of business: raptor protection	16:01
* jskladan gets of his bike		16:01
* mkrizek is here		16:01
* j_dulaney waves		16:01
* jreznik has kickbike only (but the big one as he's big boy!)		16:01
jskladan	adamw: i'd say - don't let raptors pass the tests, but that would probably by rasistic...	16:01
j_dulaney	Raptors aint getting my bacon	16:02
adamw	speciesist!	16:02
* j_dulaney brandishes rapier		16:02
jskladan	that ^^	16:02
* jsmith files a blocker bug against the raptor		16:02
adamw	aside from the few cyclist-munching incidents and their tendency to arrive with significantly fewer passengers, raptors make fine bus drivers	16:02
adamw	man, this is gonna look super-professional in the logs	16:03
adamw	tflink: ping?	16:03
adamw	i know you're around, stop hiding.	16:03
adamw	don't leave me alone with all these Js.	16:03
* pschindl is here		16:03
adamw	so no viking-ice, no tflink, no kparal? fired, the lot of 'em	16:04
adamw	alrighty	16:04
adamw	#topic Previous meeting follow-up	16:05
adamw	"adamw to brief #fedora ops and fedora-user-list regulars on fedup" - heh.	16:05
tflink	sorry, trying to get the list ready for mini-blocker-review	16:06
adamw	well, I've talked to a few people about it, not sure I hit everyone in the description	16:06
* maxamillion is here		16:06
adamw	.fire tflink	16:06
zodbot	adamw fires tflink	16:06
j_dulaney	Heh.	16:06
tflink	cool, now I can go do other stuff	16:06
* maxamillion runs like hell from the angry bot master		16:06
j_dulaney	My greatest contribution to Fedora QA	16:06
adamw	it will live on long beyond fedora QA.	16:07
adamw	as long as adamw is firing stuff, a little piece of you will survive.	16:07
j_dulaney	Yes!	16:07
adamw	so fenrus and bob from #fedora are kind of up on fedup and grumbling about it to wwoods. not sure about user-list folks. anyone been following that list lately?	16:07
tflink	not so much	16:08
BobJensen	I've nothing on fedup	16:08
tflink	but I keep seeing bugs about people using my siderepo	16:08
adamw	BobJensen: oh sorry bob	16:08
tflink	which makes me think that I need to add something to the testing docs "don't use this for production instances, see the real docs"	16:09
adamw	tflink: that sounds like a plan	16:09
adamw	#action tflink to clarify use of his fedup side repo	16:09
BobJensen	I don't know that we have seen enough users of it honestly	16:09
adamw	BobJensen: see https://fedoraproject.org/wiki/FedUp . it mostly works except when it doesn't.	16:10
j_dulaney	Still?	16:10
tflink	j_dulaney: still? it's brand new code	16:10
adamw	BobJensen: the idea was we wanted to make sure you folks knew about fedup and could help people appropriately who try to use it, or asked 'how do i upgrade'	16:10
j_dulaney	Is it still doing the "not installing kernel" issue/	16:10
adamw	j_dulaney: well, I mean, like all upgraders. i'd say the same about the old ones.	16:10
BobJensen	adamw: I'll add a shortcut here so if people ask I have the link anyhow	16:10
tflink	j_dulaney: pretty much, but that's not a fedup issue	16:11
adamw	BobJensen: there's a 'how can I upgrade' anchor in the wiki page	16:11
j_dulaney	tflink: You know what I think about that, I'll not start flamng here	16:11
* kparal arrives late		16:11
adamw	okay, i'll brief bob a bit more in private and let's move on	16:12
adamw	#info "adamw to brief #fedora ops and fedora-user-list regulars on fedup" - partly done	16:12
adamw	#info "adamw to brief #fedora ops and fedora-user-list regulars on fedup" - done, and we'll come to it in a minute.	16:12
adamw	grrr	16:12
adamw	#undo	16:12
zodbot	Removing item from minutes: <MeetBot.items.Info object at 0x29840e50>	16:12
Martix	what are you doing here? let's /join #fedora-test-day :-P	16:12
adamw	#info "adamw to put 'enterprise storage support in newui' on next week's agenda" - done, and we'll come to it in a minute.	16:13
adamw	oh god, in 'things I forgot to blog about' news...	16:13
adamw	#chair tflink kparal	16:13
zodbot	Current chairs: adamw kparal tflink	16:13
adamw	#topic Fedora 18 Final status	16:13
Martix	adamw: you still can, test week just began	16:14
adamw	so to make sure everyone's on the same page - the change deadline is now tomorrow, 2012-12-11	16:14
adamw	the target release date is still 2012-01-08	16:14
adamw	go/no-go is, conveniently, on new year's day, when we will all be fully alert and committed to such an important decision	16:14
tflink	with 28 proposed blockers and 14 accepted blockers? this should be interesting	16:14
j_dulaney	Heh	16:15
adamw	#info current schedule: change deadlines 2012-12-11, go/no-go 2012-01-01, release date 2012-01-08	16:15
adamw	tflink: fascinating!	16:15
adamw	did you clean up the proposed list for the 'obvious' bugs?	16:15
jreznik	adamw: well, the go/no-go has to be definitely moved to thursday... but I'm not sure how final go/no-go looks like - it's on purpose on tuesday...	16:16
tflink	I got sidetracked this morning and haven't made it through the bugs that needed testing but I have a list of stuff that needs discussion	16:16
maxamillion	adamw: I have a prediction the the attendence to the go/no-go will be low ... pending the intensity of the hang overs	16:16
jreznik	so it would be 2013-01-03	16:17
adamw	tflink: cool.	16:17
adamw	jreznik: ah, okay.	16:17
maxamillion	(that was semi peanut gallery, semi serious)	16:17
adamw	grr	16:17
adamw	#undo	16:17
zodbot	Removing item from minutes: <MeetBot.items.Info object at 0x226770d0>	16:17
adamw	#info current schedule: change deadlines 2012-12-11, go/no-go 2013-01-01, release date 2013-01-08	16:17
adamw	#info go/no-go may move to 2013-01-03	16:18
adamw	this is going to be tough enough without building a time machine. :)	16:18
j_dulaney	Heh	16:18
adamw	#info Beta TC1 is out and in testing	16:18
jreznik	I'm really not sure why it's scheudled on tuesday... looking for veterans!	16:18
adamw	no fires or explosions relating to TC1, right? it more or less works	16:18
adamw	jreznik: hell if i know, i'm usually permanently wasted by this point in the schedule :P	16:19
* nirik has one thing to bring up at some point... lives are sometimes failing to compose... looks like a weird race condition or something.		16:19
j_dulaney	nirik: With live-image-creator?	16:20
nirik	livecd-tools	16:20
nirik	https://bugzilla.redhat.com/show_bug.cgi?id=885385 is the bug.	16:20
* jreznik is reading		16:20
Viking-Ice	so we are suddenly back to 16:00 when last meeting was 17:00 utc?	16:21
nirik	I tweaked with the builders yesterday and they seem fine, so not sure what changed or whats causing it. ;(	16:21
adamw	#info nirik found problems with live composes of Final - https://bugzilla.redhat.com/show_bug.cgi?id=885385	16:21
adamw	Viking-Ice: er...it wasn't? we've been at 1600 the whole time?	16:22
j_dulaney	nirik: It's only part of the time?	16:22
Viking-Ice	adamw, ah see it now blocker bug 17:00	16:22
kparal	Viking-Ice: qa meetings are at 16 utc, blocker bug meetings at 17 utc	16:22
j_dulaney	adamw: It's always 11 am Eastern time, no matter if it is daylight savings or not	16:22
nirik	right. it seems unrelated to which ks is used... for example, the kde's failed this way for tc1... but later I did them over and they worked fine. I really don't get whats going wrong. ;(	16:22
adamw	j_dulaney: yes, that's the idea.	16:22
adamw	by 'whole time' i meant 'since the dst switch', of course.	16:23
j_dulaney	nirik: Then we can probably live with it?	16:23
j_dulaney	nirik: If at first you don't succeed, try again?	16:23
adamw	#info no major showstoppers in TC1, it is viable for testing	16:23
nirik	well, I think that would be pretty poor for trying to compose things...	16:23
nirik	keep throwing it against the buildsystem until it works doesn't inspire confidence.	16:24
j_dulaney	Aye	16:24
jreznik	not good :( to just wait "to be lucky"	16:24
adamw	but it works in a crisis.	16:24
adamw	#info blocker list is considerable, though may be a little bit prunable	16:24
adamw	tflink: how long is the list of bugs to discuss? should we do it inline or after the meeting?	16:25
* kparal is for a separate meeting		16:25
* j_dulaney notes he has a final exam in an hour		16:25
j_dulaney	So, by default I will be getting off in thirty minutes	16:26
tflink	I see 16 bugs on my list of 'ready for discussion'	16:26
adamw	oh boy	16:26
adamw	guess we're going 'after meeting'	16:26
adamw	so aside from the above-noted...any major issues to resolve about 18 final status? or is everyone happy doing the blocker bug grind?	16:26
adamw	or, you know, giving it all up to be a himalayan yak farmer.	16:28
adamw	those are your only choices.	16:28
tflink	define "giving it all up"	16:28
adamw	no. you must trust to the yaks.	16:28
j_dulaney	Do the criteria changes right quick	16:28
jreznik	yak farmer... nice dream	16:28
maxamillion	I've been really happy with F18's stability in the past week on my main workstation but I got lucky and didn't hit any anaconda bugs ... so blocker grind?	16:29
adamw	j_dulaney: that's another topic	16:29
j_dulaney	Can I give it all up to be a pirate?	16:29
adamw	maxam: blocker grind as in, find blocker bugs, test fixes, test TCs/RCs	16:29
adamw	j_dulaney: you're allowed to be a pirate on days off from yak farming.	16:30
adamw	oh yeah, i did put another bullet under this heading	16:30
adamw	tflink: do you know what shape fedup is planned to be in for final?	16:30
adamw	i'm guessing no GUI, but are the rough edges getting knocked off? manually specifying remote location, lack of progress etc	16:31
adamw	and nirik: what was the SB status of TC1? I lost track of that	16:31
nirik	the signed shim is in updates-testing and was not pulled into the compose.	16:32
jreznik	adamw: for GUI - FESCo seems not to require it, what's in GIT is really really very early try	16:32
adamw	nirik: okay, so no real SB support in TC1. might come in next build.	16:32
* nirik nods.		16:32
adamw	#info TC1 is not signed for SB	16:33
nirik	if we pull new shim, we also need new lorax	16:33
adamw	and vice versa?	16:34
j_dulaney	Shouldn't, I thought	16:34
j_dulaney	If I read things correctly	16:35
adamw	istr nirik finding the build failed unless he pulled shim or downgraded something.	16:35
tflink	that was fun, lost network for a few minutes	16:35
adamw	welcome back tflink. how are the yaks?	16:35
nirik	if we pull new lorax, it blows up without new shim	16:35
tflink	not shaven yet	16:35
adamw	right.	16:36
adamw	tflink: so i was just asking what the likely status of fedup for final is	16:36
maxamillion	nirik: is lorax the new pungi?	16:36
adamw	as regards gui, 'smooth' operation, progress reporting	16:36
adamw	maxam: they're both involved.	16:36
tflink	it's getting better but I'm a bit behind on my fedup testing ATM	16:36
maxamillion	adamw: rgtr	16:36
maxamillion	rgr*	16:36
tflink	my understanding is that there are some fixes in the pipeline that should help a few of the pain points	16:37
maxamillion	the whole shim madness is something I barely understand	16:37
jreznik	wwoods: are you around?	16:37
adamw	maxam: i can forward you a gigantic email explanation of it i wrote if you like.	16:37
adamw	#info some fixes for the rough edges of fedup impending	16:37
adamw	okay, let's move on, time's a tickin	16:37
adamw	#topic enterprise storage for F18/F19	16:37
wwoods	I'm around! I'm also about.	16:37
adamw	Viking-Ice: around?	16:37
Viking-Ice	yup	16:37
adamw	wwoods: too late!	16:37
adamw	so this was yours...	16:38
wwoods	Well then I'm outside.	16:38
Viking-Ice	yup	16:38
wwoods	and other prepositions.	16:38
mjg59	(Crying)	16:38
adamw	and, indeed, positions.	16:38
j_dulaney	adamw: I'd like the shim email as well	16:38
j_dulaney	If you don't mind	16:38
Viking-Ice	https://www.redhat.com/archives/anaconda-devel-list/2012-December/msg00025.html	16:38
adamw	mjg59: was that email I CCed you on too outrageously inaccurate or anything? before i send it to anyone else	16:38
jreznik	wwoods: (about dialog) - seems like there's another topic but if you have anything for fedup status - just fire it :)	16:39
mjg59	adamw: No, it seemed reasonable	16:40
adamw	Viking-Ice: what was the outcome of that? iirc, 'advanced' storage types that require interaction to enable aren't going to work in F18	16:40
Viking-Ice	basically there is no UI on newUI for Enterprise/Advantage storage so enterprise admins/users will have to rely ks only	16:40
adamw	mjg59: cool.	16:40
adamw	#info enterprise/advanced storage types that require interaction to enable/use will not work via UI in in F18, but can be used via kickstart. UI is scheduled to re-appear in F19	16:40
Viking-Ice	this ofcourse is something that needs to be mentioned in the release notes	16:41
j_dulaney	+1	16:41
adamw	agreed	16:41
Viking-Ice	not sure if there is specific Anaconda/Installer section in the release notes	16:41
adamw	do you want a #action to let the docs team know about it?	16:42
adamw	or i can take it if you like	16:42
Viking-Ice	no I'll handle it	16:42
adamw	ok	16:42
adamw	#action viking-ice to let docs team know about advanced storage for release notes	16:42
adamw	anything else to cover on this or shall we move on to criteria/test cases?	16:42
* jreznik will recheck RN notes for it		16:42
adamw	#topic Test case / criteria revision	16:45
adamw	so i think my proposal on the RPM package error criterion was positively received, i'll go ahead and put that into production	16:45
adamw	#action adamw to put packaging error criterion revision into production	16:45
j_dulaney	+1	16:46
adamw	kparal's kickstart revision proposal spawned a big discussion which was interesting but might have stranded the proposal a bit	16:46
adamw	any ideas for action there? try and reset the discussion to a list of ks parameters it's plausible to 'support' for f18?	16:46
kparal	I think we should vote in F18 according to our best conscience and target this new criterion to F19	16:47
j_dulaney	Aye, that would be a good move	16:47
kparal	it will be a long discussion yet, I think	16:47
j_dulaney	Actually, I really think that criterion revisions should not take effect until the *next* release	16:48
tflink	ideally, yes	16:48
adamw	viking's said the same thing	16:48
j_dulaney	That's been my problem with all the criterion changes in the middle of this release	16:48
adamw	i don't disagree in theory, but for f18, we've kinda had to rejig things on the run or else we'd have nothing applicable	16:48
jreznik	adamw: +1	16:49
adamw	#action kparal to try and 'reset' kickstart criterion discussion back to the proposed set of 'supported' commands	16:49
kparal	hmm	16:49
Viking-Ice	can we actually do that	16:49
adamw	that okay?	16:49
kparal	I can try	16:49
adamw	Viking-Ice: sorry, actually do what?	16:50
Viking-Ice	dont we just need to say all ks command have to work and test for those	16:50
tflink	I'm not sure that's practical for F18	16:50
Viking-Ice	why try to split it up? ( should be easy to test as well )	16:50
adamw	well the worry is that if we say that we'll find a bunch of fairly obscure bugs and then what?	16:50
adamw	but i suppose we could do it that way around...test it and see how bad it is	16:50
Viking-Ice	let anaconda deal with it	16:50
j_dulaney	Heh	16:50
Viking-Ice	it's their brokenness right	16:50
kparal	adamw: well we can always waive some obscure problems, same as for graphics cards etc	16:51
j_dulaney	Aye	16:51
adamw	kparal: well we can do that because they're 'conditional' breakages	16:51
kparal	but it becomes more tricky than with a defined set	16:51
adamw	we don't have a criterion that says 'all graphics cards have to work'	16:51
j_dulaney	"All KS commands should work unless they are really freaking off the wall"	16:51
adamw	if we did, it'd be kinda hard to waive some graphics cards being broken :)	16:51
Viking-Ice	unless they have spesific ks command that can never break or become backwards incompatible I cant see how we can	16:51
adamw	so if we have a criterion that says 'all kickstart commands have to work', it seems quite hard for us to say 'well except THIS one, obviously!'	16:51
Viking-Ice	so basically we should follow their list ( if exist ) or check for all existing supporting ks commands right?	16:52
adamw	Viking-Ice: that was matt miller's idea, and it seems reasonable, but the problem is right now the documentation of kickstart commands is somewhat outdated and doesn't in any way define a 'core set'	16:52
kparal	that goes back to my long-term complain about criteria being all-or-nothing game	16:52
adamw	so for f18, it's a bit hard	16:52
adamw	sigh, we seem to keep winding up on tangents here	16:52
kparal	I'll try to restart the discussion	16:53
adamw	the task is really just 'quantify the level of kickstart functionality we consider minimal for f18 final release', that does not seem like it should be impossible	16:53
kparal	with just a core set of commands for F18	16:53
kparal	and possible changes for F19	16:53
adamw	and shouldn't require fundamental reimaginings of process or anything	16:53
Viking-Ice	yes I just fail to see why the need to quantify it in the first place	16:53
adamw	what's your alternative for dealing with whether breakages in kickstart should block f18 final?	16:54
adamw	just do it case-by-case?	16:54
adamw	i mean, we could do that.	16:54
adamw	we could just say 'we don't want to fiddle with the criteria temporarily, so we'll just handle these one by one for now.'	16:54
Viking-Ice	I would say it blocks always and questionable items dealt with on case by case bases	16:54
jreznik	case by case could work for F18, especially as we do not know in what state it is...	16:54
adamw	shall we go with that, then? seems straightforward	16:55
tflink	it'll work, I suppose and will probably generate less discussion in the short term	16:56
tflink	may lead to some interesting discussions about what is worth blocking for, though :)	16:56
kparal	case by case is the best way, it just requires more time to discuss	16:56
adamw	okay. well let's hope we don't get too many cases. :)	16:56
adamw	#undo	16:57
zodbot	Removing item from minutes: <MeetBot.items.Action object at 0x1f3f3e10>	16:57
kparal	adamw: should we follow up on that discussion for F19?	16:57
adamw	#agreed instead of trying to write a 'temporary criterion' we'll deal with kickstart proposed blockers on a case-by-case basis for F18 final and try to come up with a longer-term criterion for F19+	16:57
adamw	I think so, yeah.	16:57
kparal	ok	16:57
j_dulaney	kparal That would be a good idea	16:57
adamw	you want to do that now or keep it for post-release?	16:57
tflink	we might want to tell the anaconda devs what we're doing, though	16:57
* j_dulaney needs to be off for his final final		16:57
adamw	yeah, and follow up on test list	16:57
kparal	adamw: I'll write a note there	16:58
j_dulaney	See y'all later.	16:58
adamw	#action kparal to follow up on test@ discussion to explain the plan, and let anaconda team know	16:58
adamw	cya dulaney	16:58
adamw	okey dokey	16:58
adamw	pschindl: are you planning to work on the issues identified in your test case / criteria survey?	16:58
pschindl	adamw: yes	16:59
adamw	cool	17:00
pschindl	I'd like to go through it this week	17:00
adamw	did anyone have any worries / questions about pschindl's proposals that haven't been brought up in the thread?	17:00
adamw	#info pschindl will work on his proposed final test case changes (see thread 'Final criteria/Test cases interconnection')	17:01
adamw	#topic open floor	17:01
adamw	anyone have anything that wasn't covered?	17:01
* satellit needed; a good anaconda tutorial lots of people having problems understanding how to install		17:03
adamw	the docs team are working on updating the install guide for newUI.	17:03
tflink	and people having trouble with patience, aparrently	17:03
satellit	+1	17:04
adamw	tflink: where are we doing the blocker review this week?	17:05
tflink	bugzappers, I think unless someone has a better idea	17:05
* kparal +1		17:05
adamw	okey dokey	17:05
adamw	everyone to #fedora-bugzappers for some thrilling blocker review in two minutes	17:06
adamw	or, you know, head for the yaks	17:06
* adamw sets fuse for two minutes		17:06
adamw	it's a classical physics fuse this week!	17:06
Viking-Ice	It should be done in QA or someother channel then bugzappers	17:06
kparal	it should be done in #fedora-anythingelsethanbugzappers, right?	17:07
Viking-Ice	I still say QA	17:07
kparal	-1	17:08
adamw	thanks folks	17:08
Viking-Ice	but I accept #fedora-anythingelsethanbugzappers as well	17:08
adamw	#endmeeting	17:08

Generated by irclog2html.py 2.11.0 by Marius Gedminas - find it at mg.pov.lt!