QA/Meetings/20120924

Attendees

• adamw (141)
• tflink (63)
• dan408- (21)
• jreznik (8)
• nirik (6)
• robatino (5)
• zodbot (3)
• Cerlyn (3)
• maxamillion (3)
• misc (2)
• Southern_Gentlem (2)
• spoore (1)
• brunowolff (1)
• jskladan (1)

Previous meeting follow-up

• adamw to find out who's writing the release announcement and make sure it calls out the biggest Alpha bugs - this was done successfully
• tflink to draft up a freeze entrance requirements proposal for the list and we can take the idea from there - tflink still working on quantifying freeze readiness
• pschindl to kill 'uncategorized package groups' criterion - this got done and reported on the list
• kparal to refine 'release-blocking package sets' criterion - this is still going on but it looks like we're pretty close
• adamw to refine alpha partitioning criterion - adam didn't get around to this yet, sorry

Release criteria revision

• Let's take a look at the whole Beta criteria list and try to get a set of Beta criteria we are happy with firmly enforcing:
  • adamw will propose partitioning criteria after discussion with anaconda team
  • tflink and adamw propose to replace the specific upgrade methods listed in the Beta upgrade criterion with the phrase "officially supported upgrade method(s)"
  • we agreed 'all kickstart delivery methods' criterion is possibly overstated, adamw will consider potential changes
  • tflink will ask other potentially interested parties to see if they think any of the Beta criteria need to be changed

Naming of TCs/RCs

• We still don't have a proposed scheme that everyone loves, but we agree the goal is to come up with a TC/RC naming scheme as unlikely as possible to confuse people about what each build is

Open floor

• tflink will be announcing a new release of the blocker bug tracking app (also known as Skynet) shortly

Action items

• adamw to refine alpha partitioning criterion
• adamw to draft new partitioning criterion for Beta once we know what will be in anaconda
• adamw to consider revisions to 'kickstart delivery method' criterion
• tflink to ask other interested parties (anaconda team, fesco...) to look over the beta criteria and see if there's anything they feel should be dialled down

IRC Log

adamw	#startmeeting Fedora QA Meeting	15:02
zodbot	Meeting started Mon Sep 24 15:02:00 2012 UTC. The chair is adamw. Information about MeetBot at http://wiki.debian.org/MeetBot.	15:02
zodbot	Useful Commands: #action #agreed #halp #info #idea #link #topic.	15:02
adamw	#meetingname fedora-qa	15:02
zodbot	The meeting name has been set to 'fedora-qa'	15:02
adamw	#topic roll call	15:02
* tflink is here		15:02
dan408-	here somewhat	15:02
adamw	who's about for a QA meeting?	15:02
brunowolff	I'll be lurking for a little while, but have a work meeting soon.	15:02
dan408-	ill be joining from the diner	15:03
dan408-	providing my phone holds a decent charge	15:03
Cerlyn	I'm hear to QA the QA meeting	15:03
adamw	well I'm QA'ing your QA'ing of the QA meeting, and you spelled 'here' wrong	15:03
dan408-	you can HEAR the QA meeting?	15:04
Cerlyn	Clearly that's the fault of the IRC channel and not me.	15:04
dan408-	do you have speech assistance turned on?	15:04
spoore	oh that'd be sweet....can you assign different voices to different nicks?	15:04
dan408-	i'll code that in to the next version of BitchX	15:05
Southern_Gentlem	spoore, no but i am sure festival would love help to do that	15:05
adamw	mine should sound like krusty the klown	15:06
adamw	alrighty	15:06
* dan408- = apu		15:06
tflink	looks like we're missing the brno folks - is today a holiday?	15:07
adamw	dan408: I had you down as stan the coffin salesman from monkey island.	15:07
dan408-	i had you down as marge..	15:07
adamw	tflink: i think someone said something about that on the internal list, let me see	15:07
adamw	dan408: hehe	15:07
tflink	adamw: the 28th is a holiday	15:08
adamw	huh. they said fri 28 is a holiday	15:08
adamw	yeah	15:08
adamw	nothing about today	15:08
adamw	maybe the network's down there?	15:08
misc	nope	15:08
dan408-	holiday is 1 week long!	15:08
adamw	or they all finally decided to go get drunk instead of coming to the meeting?	15:08
jreznik	only Friday is a free day here :)	15:08
misc	adamw: or being drunk then coming to the meeting	15:09
adamw	jreznik: can you go check if there's a QA team in the nearest gutter? :)	15:09
jreznik	adamw: you can't get drunk here - there's still prohibition on going...	15:09
dan408-	that's peculiar	15:09
jreznik	hmm, seems like brno's qa guys are really drinking :) I don't see them :)	15:10
* maxamillion is here-ish (I swear every time I try to make this meeting something happens right in the middle .... going to try and stay focused today)		15:11
adamw	grr	15:11
adamw	the drives on this laptop are losing it, methinks	15:11
adamw	well we'll get by as best we can without kparal and co., i guess	15:11
adamw	#topic previous meeting follow-up	15:12
jreznik	adamw: I pinged jskladan, seems like the only one online...	15:12
adamw	"adamw to find out who's writing the release announcement and make sure it calls out the biggest Alpha bugs" - I done that, got the biggest stuff listed in the mail	15:12
adamw	also thanks to bcotton who made sure it was in big red letters in lots of places	15:12
adamw	#info "adamw to find out who's writing the release announcement and make sure it calls out the biggest Alpha bugs" - this was done successfully	15:13
jreznik	adamw: yep, I asked him to be RED :) btw. thanks! good job	15:13
adamw	" tflink to draft up a freeze entrance requirements proposal for the list and we can take the idea from there "	15:13
adamw	tflink?	15:13
tflink	I'm still struggling with a way to quantify freeze readiness	15:13
tflink	I did do a smoketest build with post-alpha stable	15:13
adamw	#info "tflink to draft up a freeze entrance requirements proposal for the list and we can take the idea from there" - tflink still working on quantifying freeze readiness	15:14
dan408-	what is "post-alpha" stable these days?	15:14
tflink	and I put together a list of components that could potentially cause beta to slip (according to the current release requirements)	15:14
dan408-	3.1 or 3?	15:14
adamw	dan408: neither of those.	15:14
dan408-	ok what	15:14
tflink	https://fedoraproject.org/wiki/User:Tflink/F18_alpha_smoketest1_results	15:14
adamw	dan408: he means a build with all the stuff that's in stable now	15:14
* jskladan hides in the shadows		15:14
tflink	I need to move that wiki page, though	15:15
dan408-	okay i reinstalled last night, used 3	15:15
tflink	it should be F18_beta	15:15
dan408-	i heard something bad about 3.1	15:15
* dan408- double checks		15:15
adamw	dan408: doesn't matter now anyhow, alpha is done.	15:15
jreznik	tflink: also there's https://fedorahosted.org/fesco/ticket/946 - so we can talk about the criteria there with FESCo... they understand this ticket the same way	15:15
dan408-	http://dl.fedoraproject.org/pub/alt/stage/ <-- not here.. is it on the other mirror?	15:15
tflink	If anyone has suggestions on realistic ways to measure release readiness other than 'all potential release blocking features must be testable', I'd love to hear them	15:16
tflink	s/release readiness/freeze entrance readiness/	15:16
dan408-	tflink: how about just basic functionality and usability.. since the design is so radically different?	15:16
dan408-	and i dont mean "should be able to install and xxx"	15:17
adamw	dan408: it's not an official build of any kind so it's not mirrored	15:17
tflink	dan408-: that's even more nebulous than what I have already :)	15:17
maxamillion	and like clock work ... $dayjob duties call ...	15:17
jreznik	dan408-: you have to define "basic functionality" and I think it should be based on beta release criteria	15:17
adamw	maxam: hi, bye :)	15:17
maxamillion	;)	15:17
adamw	#info https://fedorahosted.org/fesco/ticket/946 is a ticket related to freeze entrance criteria topic	15:17
dan408-	tflink: i can get detailed but im trying to keep it short and concise as i dress for work.	15:17
adamw	shall we keep it for open floor, since we have other stuff to get through?	15:18
tflink	the problem I'm having is defining "ready for freeze" without getting needlessly complicated or vague	15:18
tflink	yeah, works for me	15:18
dan408-	k brb	15:18
adamw	tflink: now you know why all those criteria are so damn long =)	15:18
* tflink doesn't want to specify something that's ready for release - just something that has a chance of being released after a 2 week freeze		15:19
tflink	adamw: I already knew the reason behind that :-P	15:19
adamw	#info "pschindl to kill 'uncategorized package groups' criterion" - pschindl is not around today, but this got done and reported on the list	15:19
adamw	#info "kparal to refine 'release-blocking package sets' criterion" - this is still going on but it looks like we're pretty close	15:20
adamw	we've got a criteria topic coming up right after this so no need for discussion of any of these right now btw	15:20
adamw	#info "adamw to refine alpha partitioning criterion" - I didn't get around to this yet, sorry	15:21
jreznik	tflink: adamw: also for checking criteria it would be great to have some pre-TC image with latest anaconda & company... /me has to leave now, will be back in approx. 1 hour :( so lets hope for open floor still going on :) or #fedora-qa :D	15:21
adamw	#action adamw to refine alpha partitioning criterion	15:21
adamw	ok, anything I'm missing that we should follow up on from last week that's not in the agenda to come?	15:21
adamw	#topic release criteria revision	15:23
adamw	so I put this on the list mainly because we went on a long time last week and i thought there may be stuff people still wanted to bring up, about any of the existing proposals	15:23
adamw	if everyone's broadly happy with the current proposals, i figured we could take a look at the current beta criteria together and see if any aside from the ones already undergoing revision might need changes	15:24
adamw	welp, i guess that's what we're doing then =)	15:26
tflink	jreznik_afk: that's what the beta smoketest image is for - it has the latest anaconda build (feature-wise, there was a rebuild for new glade)	15:26
adamw	so it would obviously be good for Beta if we can stand confidently by the criteria we have instead of fudging as much as we did for alpha	15:27
tflink	do we know what kinds of disk usage will be ready for beta (free space, etc.)	15:28
adamw	so what i'm thinking there is we all take a look at https://fedoraproject.org/wiki/Fedora_18_Beta_Release_Criteria and see if there's any of those requirements we might need to change or loosen	15:28
adamw	tflink: i don't yet, but that's definitely something we need to nail down	15:28
tflink	that's one of the bigger potential fudges that I see	15:29
tflink	that and what's going on with upgrade	15:29
adamw	well we have little in the way of partitioning for beta	15:29
adamw	but that's because the alpha criterion was quite strong before	15:29
tflink	but the alpha release requirements still hold for beta	15:29
adamw	now we're weakening the alpha one, we'll need a new beta one, but i did want to see what the beta partitioner will look like before drafting one	15:29
adamw	right, but the alpha criteria is now just 'wiping an entire existing disk must work', remember.	15:30
adamw	criterion*	15:30
tflink	yeah, I was going off of what is currently in the wiki and what it used to be	15:30
adamw	right, that's why i need to get the alpha change done, it keeps confusing people :/	15:30
tflink	under the assumption that we aren't going to release final with just full disk autopart	15:30
adamw	#action adamw to draft new partitioning criterion for Beta once we know what will be in anaconda	15:30
adamw	on upgrades, yeah, that's another significant one	15:31
adamw	i think what we should change there is the text "either via preupgrade or by booting to the installer manually"	15:32
tflink	from what I remember hearing, preupgrade is going to die and be replaced by a single upgrade mechanism	15:32
tflink	something related to the current upgrade work	15:32
adamw	it should say "using the officially recommended upgrade method"	15:32
adamw	imho obviously	15:32
adamw	since what we should really enforce at beta is that our 'official' upgrade method should work.	15:33
tflink	s/method/method(s)/ but yeah, that's more flexible	15:33
adamw	yeah.	15:33
adamw	jumping back a bit there's also 7. and 8.	15:33
tflink	do we have any ETAs on partitioning and upgrade?	15:33
adamw	i could see us fudging on 7, though we've discussed its place quite heavily so maybe not	15:34
adamw	and for 8. too, if some really obscure method didn't work	15:34
tflink	alpha or beta 7 and 8?	15:34
adamw	beta	15:34
tflink	isn't serial console install working for the most part?	15:34
adamw	ETAs, i don't right now, but we should definitely find out	15:34
adamw	i dunno, i haven't tested it :)	15:34
tflink	IIRC, that's the only working method of install for ppc and they've been mostly able to get serial working	15:35
tflink	wait, that's text based. not serial	15:35
adamw	right	15:35
adamw	tbh 7. is probably OK, we seem pretty firm on having that one. 8 is more fudgeable i guess.	15:36
adamw	and the other one that might be a candidate is 13.	15:36
tflink	yeah, I'd be OK with weakening 8 for beta	15:36
adamw	since it seems we don't really care as much about rescue mode as we did.	15:36
tflink	and we don't know if LVM or RAID will be supported for install in beta	15:37
adamw	13 is rescue mode, not partitioning, note.	15:37
tflink	yeah, but it talks about being able to detect and use LVM/RAID	15:37
tflink	what are the other options for fixing a busted install w/o rescue mode?	15:38
adamw	#info adamw/tfink propose to replace specific upgrade methods in the Beta upgrade criterion with 'officially supported upgrade method(s)'	15:38
adamw	tflink: live image.	15:38
adamw	either a fedora one or any one of the several that exist specifically for system rescue purposes...	15:38
tflink	I can see some of the logic there, but do the lives have the right packages to re-install grub?	15:38
adamw	all rescue mode really has going for it is a) you definitely have it right there and b) it auto-mounts your install in the correct hierarchy. but neither of those is really critical.	15:39
tflink	I suppose they would since they install grub during install	15:39
adamw	yeah, they have to.	15:39
tflink	it just seems a little odd to require downloading of another iso if the install breaks	15:39
tflink	rather than being able to use the same iso you installed from	15:39
adamw	mounting the installed system for you is pretty convenient, admittedly, especially if it uses LVM. but even there, there are tools to help. the GNOME Disks tool is pretty useful for e.g.	15:39
adamw	sure, but is 'a little odd' enough for us to delay Beta release for a fourth week? these are the tough questions =)	15:40
adamw	i just want to make sure the criteria are as bulletproof as possible in advance	15:40
adamw	#info 'all kickstart delivery methods' criterion is agreed to be possibly overstated	15:41
tflink	is rescue mode enough to delay beta for the fourth week? Maybe, maybe not. Honestly, it shouldn't be a huge issue @ beta, IMHO	15:41
adamw	#action adamw to consider revisions to 'kickstart delivery method' criterion	15:41
tflink	it should work already	15:41
adamw	OK, well let's leave that one alone for the present	15:42
adamw	looks like enough work to be going on with anyhow	15:42
adamw	anyone else spotted anything in the criteria they think might be overambitious?	15:42
tflink	if we're going to have a rescue mode, it should work by beta - there should be little/no need for it @ final	15:42
tflink	adamw: that would be a good question for the anaconda devs and/or fesco	15:43
adamw	tflink: yeah, that's a point	15:45
adamw	tflink: do you feel like an action? i've had a lot this week ;)	15:46
tflink	sure	15:46
adamw	#action tflink to ask other interested parties (anaconda team, fesco...) to look over the beta criteria and see if there's anything they feel should be dialled down	15:47
adamw	ok, i think that covers the criteria issue	15:47
adamw	#topic Naming of TCs/RCs	15:48
adamw	so i really meant to go through the archives and call out specific proposals for this discussion	15:48
adamw	unfortunately i forgot :/ so we don't really have all the various proposals that have been made to hand	15:48
adamw	still, does anyone have a particular favourite proposal for changing the naming of TCs/RCs or anything?	15:49
adamw	we can always continue this next week if necessary	15:49
tflink	works for me :)	15:50
robatino	kparal and i had a proposal that involved making the names of all proposes lexicographically ordered and adding an "R" suffix for release candidates, i'd have to dig it up in the archives	15:52
robatino	s/proposes/composes	15:52
adamw	yeah, i remember that one	15:52
adamw	C1, C2, C3R, C4, C5R - something like that	15:52
robatino	btw, is it still possible to have tcs after rcs? if so, this would make that less confusing	15:53
adamw	i like it because it's flexible and it's also obscure so should frighten people off and not make them confuse it with an official release, which is a good thing	15:53
adamw	robatino: right now it is being intentionally left a grey area	15:53
adamw	but that scheme would certainly allow for it in a nice way	15:53
tflink	it might be good to add in a PC to make it more clear that it isn't a release	15:54
adamw	pc?	15:54
tflink	I can see how F18 beta C3 could be mistaken for something released	15:54
tflink	pre-compose	15:54
tflink	or something along those lines	15:54
adamw	yeah...maybe needs more tweaking	15:54
adamw	i guess there's the danger people just ignore the bit they don't understand and read 'f18 beta'	15:55
nirik	how about... TC's stay the same and RC's change to "AC" "BC" (alpha compose, beta compose)	15:55
adamw	so we go from Alpha TC4 to AC1?	15:57
nirik	yeah	15:57
robatino	as long as the announcement explicitly says "test compose" or "release candidate" doesn't that avoid the confusion?	15:57
nirik	s/release/alpha/	15:57
adamw	robatino: no-one reads announcements, apparently.	15:58
adamw	i don't think it's people on test list who get confused	15:58
adamw	but the builds do get spread around via forums and irc to an extent	15:58
adamw	it's usually people picking them up through those channels who get confused	15:58
robatino	in that case, i'd think "TC" would be just as confusing	15:59
adamw	well yeah, that's why the topic includes TC *and* RC naming...	15:59
nirik	Fedora-18-Beta-NOT_A_RELEASE!.iso	15:59
adamw	heh	16:00
adamw	so i guess the lesson here is we still don't have a proposal everyone's in love with	16:00
adamw	but we all at least agree on the goal	16:00
adamw	sound about right?	16:00
tflink	yeah	16:00
* nirik nods.		16:01
nirik	I don't care what colour the bike shed is, just that it's a better shade of bright red to let people know what things are what.	16:01
adamw	#agreed we still don't have a proposed scheme that everyone loves, but we agree the goal is to come up with a TC/RC naming scheme as unlikely as possible to confuse people about what each build is	16:01
adamw	BRIGHT BLUE	16:01
adamw	okey dokey	16:02
adamw	#topic open floor	16:02
adamw	anything for open floor? if you wanted to continue the freeze entrance discussion now would be the time	16:02
tflink	does anyone have something to add about that discussion?	16:03
adamw	yellow.	16:03
adamw	that bikeshed should be yellow.	16:03
* tflink will be sending out an announcement about a devel version of the blocker tracking app once he fixes a bug or two - hopefully in the next day or so		16:03
adamw	#info tflink will be announcing a new release of the blocker bug tracking app (also known as Skynet) shortly	16:04
adamw	everyone stock up on tin foil hats	16:04
tflink	hey, skynet isn't planned until the next release :-P	16:04
Southern_Gentlem	skynet already exist in the UK	16:05
tflink	sounds like nothing more to add on the freeze entrance stuff ATM	16:05
adamw	sounds like that's all	16:06
adamw	ayup	16:06
* adamw sets fuse for 1 minute		16:06
adamw	i ran out of quantum fuses	16:06
tflink	oh, any thoughts on starting the blocker review meetings this week?	16:06
* adamw stamps on fuse		16:06
tflink	the list is getting _long_	16:06
adamw	yeah, that might not be a bad idea	16:06
adamw	though we should try to get criteria revisions done or at least proposed ahead of the meeting	16:07
adamw	anyone else?	16:07
adamw	i know we all love blocker meetings	16:07
tflink	they're the highlight of my week - I get lost between releases when we don't have blocker meetings	16:07
tflink	we can skip over bugs that hit criteria under contention	16:08
adamw	i mostly just curl into the foetal position and lie there crying	16:08
adamw	wait, i do that anyway.	16:08
adamw	that's a point.	16:08
tflink	I doubt that we're going to hear much back from the anaconda devs or fesco on what's actually going to make it into beta by wednesday	16:08
tflink	all the more reason to start that conversation now, I suppose	16:08
adamw	sure	16:09
adamw	one good way to start it is to accept a bunch of bugs as blockers so clumens explodes	16:09
adamw	i always like watching that	16:09
tflink	motivation :)	16:09
adamw	well i think we bored everyone else to sleep	16:10
adamw	so, blocker meeting on wednesday it is!	16:10
tflink	and one way to phrase the conversation - "we're starting blocker review this week - if you have issues with the current criteria, now would be a good time to raise them"	16:10
* adamw re-lights fuse		16:11
* tflink runs away		16:11
Cerlyn	I believe you mean fuze	16:11
adamw	i'm pretty sure i don't?	16:12
adamw	anyhow! boom	16:12
adamw	#endmeeting	16:12

Generated by irclog2html.py 2.10.0 by Marius Gedminas - find it at mg.pov.lt!