QA/Meetings/20111031

From FedoraProject

Jump to: navigation, search

Contents

Attendees

  • adamw (129)
  • tflink (35)
  • j_dulaney (34)
  • kparal (23)
  • kk4ewt (22)
  • brunowolff (21)
  • rdieter (13)
  • zodbot (6)
  • jskladan (5)
  • pjones (3)
  • jsmith (3)
  • cra (2)
  • rbergeron (2)
  • fenrus02 (1)
  • drago01 (1)
  • Dr_Horrible (1)
  • pschindl (1)

Agenda

Previous meeting follow-up

  • Last meeting was mostly blocker review, so no follow-ups

Fedora 16 prep: blocker review

  • Again, the meeting became a blocker review: details are in the log

Action Items

  • N/A

IRC Log

adamw #startmeeting Fedora QA Meeting 15:03
zodbot Meeting started Mon Oct 31 15:03:10 2011 UTC. The chair is adamw. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:03
zodbot Useful Commands: #action #agreed #halp #info #idea #link #topic. 15:03
adamw #meetingname fedora-qa 15:03
zodbot The meeting name has been set to 'fedora-qa' 15:03
adamw #chair tflink 15:03
zodbot Current chairs: adamw tflink 15:03
adamw #chair kparal 15:03
zodbot Current chairs: adamw kparal tflink 15:03
brunowolff http://lists.fedoraproject.org/pipermail/test/2011-October/104153.html repots a regression in an rc. 15:03
* j_dulaney is here 15:03
brunowolff s/repots/reports/ 15:03
* pschindl is here 15:03
* kparal present 15:03
brunowolff Does look like it would be a blocker. 15:03
brunowolff s/does/doesn't/ 15:03
adamw it's from chuck, which means i don't care. 15:04
brunowolff I can't type this morning. 15:04
adamw the dude is programmed to provide utterly useless bug reports. 15:04
jskladan me lurks 15:04
* tflink is still here 15:04
adamw #topic roll call 15:05
adamw now do it all again, suckers! 15:05
* brunowolff is here 15:05
* j_dulaney has his harmonica out and is blowing it 15:05
* kparal does it again 15:05
* jskladan can use the irc command now 15:05
j_dulaney After spending last night stroking my big wood instrument and polishing my flute 15:05
* tflink might be here, isn't sure if it changed in the last 2 minutes 15:06
* adamw covers brunowolff's ears 15:06
adamw tflink: I CAN'T HEAR YOU 15:06
tflink adamw: well, you don't need to shout 15:07
adamw alright, let's get this dog and pony show on the road 15:07
adamw heh 15:07
j_dulaney It's only dirty if you have a dirty mind or don't know that I play standup bass 15:08
adamw #topic Previous meeting follow-up 15:08
adamw why are you here anyway? didn't I fire you?! 15:08
tflink many times 15:08
j_dulaney  ??? 15:08
adamw uh, hey, this is going to be a short section, as there appear to be no actions from the previous meeting 15:08
kparal there were more people fired as I recall :) 15:08
j_dulaney Red Hat sudenly laying off? 15:09
adamw YOU'RE ALL FIRED 15:09
tflink yeah, I think that jskladan was the last once to get fired 15:09
jskladan oh yeah 15:09
j_dulaney Yay, I can go home 15:09
jskladan I did not work on state holiday 15:09
j_dulaney Oh, wait 15:09
kparal the more close to release date, the larger fire mood for adamw :) 15:09
tflink j_dulaney: no, adamw just gets cranky and fires us 15:09
brunowolff Was the quote "Good night tflink. Sleep well. I'll most likely kill you in the morning."? 15:09
adamw you know what a holiday is? it's an opportunity for you to work undisturbed by others! 15:09
j_dulaney LOL 15:10
tflink brunowolff: huh? I missed that one 15:10
adamw okay, so last meeting we mostly just reviewed blockers, and we're gonna have to do it again 15:10
* tflink looks through the logs 15:10
adamw #topic Fedora 16 prep: blocker review 15:10
adamw tflink: it's a Princess Bride joke. 15:10
adamw if you don't get it, you're fired. 15:10
jsmith tflink: "Have fun storming the castle!" 15:10
adamw also, you get to have a drink with me... 15:10
rbergeron HOOORAY 15:11
brunowolff Well I probably should have changed it to "Good night tflink. Sleep well. I'll most likely fire you in the morning." 15:11
kparal is that a threat? 15:11
rbergeron i promise i didn't just hop in because i saw the word drinks 15:11
adamw rbergeron: no, you have an alert on the phrase 'princess bride'. admit it. 15:11
jsmith rbergeron: You saw the word "princess"? 15:11
* jsmith ducks 15:11
tflink adamw: you should really add that to the job description @ hiring time. First Dr. Horrible and now Princess Bride? 15:12
* j_dulaney fingers his big wood instrument some more 15:12
j_dulaney Dr. Horrible is here? 15:12
brunowolff I have a t-shirt with "I have a PhD in horribleness" on it. 15:12
Dr_Horrible I will use my Freeze Ray on you all!!! 15:13
tflink j_dulaney: no, adamw fired me for not having seen Dr. Horrible around beta 15:13
fenrus02 Dr_Horrible, just dont talk about the hammer 15:13
adamw all right, simmer down, children 15:13
adamw #topic https://bugzilla.redhat.com/show_bug.cgi?id=750228 15:13
adamw so, it seems like somehow efidisk.img sometimes gets broken in generation. i'm probably +1 blocker on this, and i assume we can just fix efidisk.img without screwing up anything else 15:14
j_dulaney That seems right 15:14
* j_dulaney is also +1 blocker 15:14
adamw dgilmore: poke! 15:14
kk4ewt +1 15:15
adamw criterion is "Where platform support exists, all dedicated installer images (except efidisk.img, which offers no options) must boot to the graphical boot menu and allow the user to select install options. If no option is selected, the installer should load after a reasonable timeout " from alpha 15:16
tflink +1 15:16
j_dulaney Indeed 15:16
adamw propose #accepted 750228 is a blocker per criterion ""Where platform support exists, all dedicated installer images (except efidisk.img, which offers no options) must boot to the graphical boot menu and allow the user to select install options. If no option is selected, the installer should load after a reasonable timeout " 15:16
kparal +1 15:16
j_dulaney +1 15:17
kk4ewt so with that it cant be a blocker 15:17
tflink ack 15:17
adamw it can, though it's a bit poorly worded 15:17
adamw efidisk.img isn't meant to be excepted from the entire criterion, just the requirement to display options 15:17
adamw so for efidisk the requirement is 'boot to graphical boot menu and load installer after a reasonable timeout' 15:18
kk4ewt +1 15:18
adamw #accepted 750228 is a blocker per criterion ""Where platform support exists, all dedicated installer images (except efidisk.img, which offers no options) must boot to the graphical boot menu and allow the user to select install options. If no option is selected, the installer should load after a reasonable timeout " 15:18
adamw #topic https://bugzilla.redhat.com/show_bug.cgi?id=748272 15:18
tflink #accepted? 15:18
kk4ewt should be #agreed 15:19
adamw guh 15:20
adamw #undo 15:20
zodbot Removing item from minutes: <MeetBot.items.Accepted object at 0x4728bd0> 15:20
adamw #agreed 750228 is a blocker per criterion ""Where platform support exists, all dedicated installer images (except efidisk.img, which offers no options) must boot to the graphical boot menu and allow the user to select install options. If no option is selected, the installer should load after a reasonable timeout " 15:20
adamw #topic https://bugzilla.redhat.com/show_bug.cgi?id=748272 15:20
adamw so, this is just a needsinfo , and people REALLY need to stop adding bugs to the blocker list so damn late... 15:20
adamw cra: *shakes fist* 15:21
tflink it almost sounds like this might have been with an older compose 15:21
j_dulaney Indeed 15:21
j_dulaney I'm not running into UEFI problems with RC2 15:21
adamw tflink: the initial report is from beta 15:21
adamw he claims in comment #4 to still see teh bug with tc2 but provides no new logs 15:22
adamw so i don't think we can do anything with this till we get some logs 15:22
tflink yeah 15:23
pjones it's entirely possible the bootloader isn't installing correctly due to a firmware bug. 15:23
pjones that certainly has been something we've seen on some machines (though it's the first report I've seen on that one) 15:23
adamw pjones: there's a claim that ubuntu 11.10 works 15:24
pjones but yeah - we need logs to know what's going on with any degree of certainty. 15:24
adamw k 15:25
cra adamw: don't shoot the messenger :-) 15:25
adamw cra: i will if he's late ;) 15:25
adamw #agreed more information is needed to determine the status of 748272 15:25
adamw #topic https://bugzilla.redhat.com/show_bug.cgi?id=651305 15:26
adamw so, wow, this is annoying. 15:26
rdieter hi 15:26
adamw rdieter: we're on https://bugzilla.redhat.com/show_bug.cgi?id=651305 15:27
tflink yeah, I'm also -1 on this. it doesn't crash the app and doesn't happen on boot 15:27
adamw what's your read on this? 15:27
adamw tflink: er what? 15:27
adamw tflink: this is the kde upgrade bug 15:27
rdieter finally fully understand this one, can plug in some workaround(s) easy enough in a bit. 15:27
tflink adamw: sorry, I was on a different one 15:28
rdieter in short, upgrader's will hit a case where kdm is non-functional due to awol theme 15:28
rdieter only affects DVD 15:28
adamw is there any way you can fix / workaround it which does not involve us having to respin f16? 15:28
rdieter (which is why I never saw it) 15:28
rdieter adamw: the easiest fix I had in mind was a newer kde-settings pkg with a hard-coded dep 15:29
adamw right, which needs a respin. 15:29
adamw (and boy, are dgilmore's arms tired) 15:29
rdieter adamw: another approach would be to jigger DVD creation to include a couple other pkgs 15:29
kk4ewt +1 blocker 15:29
rdieter ie, add kde themes from the last 1-2 fedora releases. 15:30
adamw as far as respinning's concerned those are the same. 15:30
adamw if we have to respin to fix it the kde-settings fix sounds like the safest. 15:30
rdieter hmm... 15:30
j_dulaney Yeah, I'll say +1 blocker on this, as well, since this does kill KDE and thus hits criteria 15:30
rdieter adamw: what mods can we do that would avoid a respin? 15:31
j_dulaney Curses 15:31
adamw i was hoping you'd have something you could pull out of your ass. =) 15:31
rdieter comps ? 15:31
adamw some kind of fix to *f15* would avoid it 15:31
adamw but that seems unpossible 15:31
rdieter oh, ok, yeah, hard. sorry. 15:31
rdieter I'll fixup kde-settings here in a bit 15:31
adamw sigh. well, we can spin again to fix this, i guess, if the fix only touches kde that limits the scope of possible borkage, which is good. 15:32
kk4ewt in short other than telling kde users with the DVD to enable updates there is not way to fix without res[in 15:32
adamw yeah, seems that way. 15:32
adamw i'm not sure you even get the repo choice on an upgrade. 15:33
kk4ewt so borkage +1 respin 15:33
adamw propose #agreed 651305 is accepted as a blocker per criterion "The installer must be able to successfully complete an upgrade installation from a clean, fully updated default installation (from any official install medium) of the previous stable Fedora release, either via preupgrade or by booting to the installer manually. The upgraded system must meet all release criteria." 15:34
adamw anyone want to argue against? 15:34
adamw kk4ewt: how come the nick change? 15:34
tflink ack 15:34
j_dulaney adamw: What happens if you use (CD, I know) a KDE CD to upgrade a Gnome system? 15:34
kparal ack 15:34
kk4ewt j_dulaney, still missing a dep 15:34
j_dulaney Ok 15:35
j_dulaney +1 15:35
kk4ewt livecds dont upgrade 15:35
adamw er, what? 15:35
adamw yeah 15:35
adamw you can't do that 15:35
j_dulaney That's right 15:35
* j_dulaney forgot 15:35
adamw this affects the case where you use the DVD to upgrade a system that uses KDM as the login manager, basically. 15:35
j_dulaney Indeed 15:36
j_dulaney Still, +1 blokcer 15:36
adamw it doesn't quite meet the criterion 15:36
adamw since the criterion refers to a 'default install', which this is not 15:36
adamw but it does seem like something we really ought to fix 15:36
j_dulaney Indeed 15:36
tflink you mean since it's KDE instead of gnome? 15:36
adamw yeah. to get that you had to change the package selection, hence, not default. 15:37
tflink nvm 15:37
kk4ewt default from any medium which covers it 15:37
adamw that criterion is mostly just a hedge against 'you can't test every damn thing', though. 15:37
adamw kk4ewt: oh, right, if you installed kde live then dvd upgrade you'd hit this.\ 15:37
kk4ewt they have an full updated kde livecd install 15:37
adamw #agreed 651305 is accepted as a blocker per criterion "The installer must be able to successfully complete an upgrade installation from a clean, fully updated default installation (from any official install medium) of the previous stable Fedora release, either via preupgrade or by booting to the installer manually. The upgraded system must meet all release criteria." 15:37
adamw #topic https://bugzilla.redhat.com/show_bug.cgi?id=748921 15:38
tflink now that I'm looking at the right bug, I'm also -1 on this 15:38
tflink since it doesn't crash anything and doesn't happen on boot 15:38
adamw yeah, doesn't meet either criterion 15:39
adamw i'm not super keen on NTH for this either, as selinux-policy upgrades are big and can break anything 15:39
adamw i think we just leave this one to get fixed by upgrades 15:39
tflink yeah 15:39
j_dulaney -1 for blocker 15:40
* j_dulaney doesn't see how it really affects usability, either, since SELinux complaints _can_ be ignored 15:40
kk4ewt -1 15:41
kk4ewt next 15:41
kparal -1 blocker 15:42
adamw propose #agreed 748921 is not a blocker, doesn't meet either the selinux or abrt criteria 15:42
tflink ack 15:42
j_dulaney ack 15:42
brunowolff I have had the time/date applet crash a few times, but it seems to happen rarely when first signing on. I don't know if this is exactly the same bug. 15:42
kparal ack 15:42
brunowolff But since it is rare and the effect is small (you can remove and add the applet from the menu) I wouldn't think it would be a blocker. 15:43
adamw #agreed 748921 is not a blocker, doesn't meet either the selinux or abrt criteria 15:43
brunowolff In my case I am using fallback mode, which may make a difference. 15:43
adamw alright, that's the lot, i think 15:43
adamw does anyone with strong psychic powers know what the hell Chuck Forsberg's "Final Release Candidate Regression" mail is saying? 15:44
adamw #topic Fedora 16 final prep: blocker review: any other blockers 15:44
* j_dulaney (as Dr. Horrible) doesn't 15:44
brunowolff Didn't he have some post about 3TB disk drive partitions a few weeks ago? 15:44
kparal I have one bug to discuss 15:44
j_dulaney brunowolff: I think so 15:45
drago01 adamw: parse error 15:45
tflink it sounds like it's an issue with partitioning but I'm not sure that there's enough info to figure out whether its pilot error or not 15:45
cra sounds like reusing a BIOS BOOT partition fails 15:45
kk4ewt adamw, so i think we may have only on blocker which may be the kde stuff 15:45
brunowolff It was annoying that the spin kickstarts needed a few changes. Unless someone thinks it is critical to have the actual kickstarts on the media, I'll just do this as an update. 15:46
adamw cra: well, we can confirm/deny that easy enough 15:46
tflink I just reused a BIOS boot partition about 5 minutes ago so if there is a failure there, it's not with all machines 15:46
adamw brunowolff: i'm not sure if there's any legal implications to that? 15:46
adamw tflink: ah. good. in that case i'm all for ignoring chuck 15:46
* adamw projects the Spot Beam 15:46
adamw spot: legal q! 15:46
brunowolff Jesse used to care about this, but I didn't understand the full reason. He would be a good person to check with. 15:47
brunowolff It's not like the source won't be available somewhere, as the source tar ball will be on fedorahosted as well as the srpm being in updates. 15:48
adamw yeah. we'd best check with him before doing rc3 spin. okay 15:48
adamw kparal: which bug did you want to discuss? the text upgrade one? 15:48
kparal https://bugzilla.redhat.com/show_bug.cgi?id=742207 15:49
brunowolff If you get an answer from him, I'd like to hear it so I can write it down on the spins wiki. 15:49
adamw k 15:49
adamw #topic https://bugzilla.redhat.com/show_bug.cgi?id=742207 15:49
adamw so i did know about this one at beta 15:49
adamw and i'm still pretty much -1 blocker on it. though i was hoping anaconda team would fix it, i admit. 15:50
adamw kparal: yes, we knew about the weird 'it'll proceed if you select nothing' case too 15:50
kparal basically if you don't select the option properly, it will destroy your data without you wanting to 15:50
adamw destroy data is a rather strong way to describe 'mess up bootloader config' 15:50
kparal ok, overwrite bootloader 15:51
adamw when we talk about destroying data we're usually talking about actual, you know, User Stuff 15:51
kparal ok, destroy bootloader 15:51
kparal to put it into correct words 15:51
kparal I just wanted to highlight that issue 15:51
kparal I am not also sure whether to vote it as a blocker 15:51
adamw yeah, i mean, it sucks, but...text mode upgrade. ehhhh. 15:53
adamw we don't really have criteria for it. 15:53
brunowolff I think it is not a blocker. It is rare to do text upgrades and based on the description requires user error (even if easy to do) to cause an issue. 15:53
kparal actually it's extremely easy to do 15:53
kparal I made the error 15:53
brunowolff Given it's anaconda and we have no time left, I wouldn't make it NTH either. 15:53
kparal I had upgraded countless times 15:53
adamw any other votes? 15:54
tflink I think that I'm of pretty much the same mind brunowolff is, if it was earlier ... maybe 15:54
kk4ewt -1 15:54
brunowolff We could document how to recover from it for the few people who might run into it. 15:54
tflink yeah, definitely document the bug 15:54
kk4ewt +1 brunowolff 15:55
kk4ewt doc in common bugs 15:55
adamw propose #agreed 742207 is not a blocker: text mode upgrade isn't really covered by criteria and the likely impact of this is too small to really worry about 15:55
adamw s/small/narrow/ 15:55
j_dulaney ack 15:55
brunowolff +1 15:55
kk4ewt +1 15:56
tflink adamw: have you looked at https://bugzilla.redhat.com/show_bug.cgi?id=750216 ? 15:56
tflink since you've been working with preupgrade more than most of us have? 15:56
adamw #agreed 742207 is not a blocker: text mode upgrade isn't really covered by criteria and the likely impact of this is too narrow to really worry about 15:57
adamw tflink: not yet 15:57
tflink it might be blocker material if there is no pilot error 15:57
tflink but it sounds like the preupgrade issue that we've had for a while 15:57
adamw it sounds incoherent to me 15:58
tflink and if the fix required something from F15 updates-testing, I'm wondering if it just needs retesting 15:58
adamw the reproduction steps are clearly incomplete 15:58
adamw there's no really useful logs there 15:58
tflink took me a while to figure out what he was talking about and I'm still not sure 15:58
* j_dulaney must go AFK for a few minutes,. will return 15:59
kk4ewt should be tested but not a blocker as its reported at the moment 16:00
adamw yeah, i think we just need to keep getting info from hongqing on that one. 16:00
kk4ewt and the preupgrade was pointed at Alpha Beta or what 16:00
adamw i've tested preupgrade and it works, couple of other people have reported it working too, the actual .ks written there looks correct, so i'm not hugely concerned 16:01
adamw okay, so we're over time 16:01
adamw let's skip along quickly 16:01
adamw #topic proventester / autoqa updates 16:01
adamw anything on the autoqa front, kparal or tflink? or has it been all f16 stuff? 16:01
tflink mostly f16 stuff and trying to get ready to test the next release of autoqa 16:02
kparal nothing big from me 16:02
adamw k 16:03
adamw #topic upcoming qa events 16:03
adamw there ain't any, aside from getting f16 done :) 16:03
adamw #topic open discussion 16:03
adamw anything else? might be stuff i've been missing with the focus on f16, don't be shy 16:03
kk4ewt how does one point preupgrade to the latest spin 16:05
adamw kk4ewt: it uses the repos, it doesn't know about spins. 16:05
kk4ewt so how would one test with preupgrade and pull from the rc2 tree 16:06
adamw one doesn't, really. but one should get the latest packages pushed to stable, which is good enough. 16:07
* adamw sets fuse for 2 mins 16:07
kparal this time we have chairs :) 16:07
adamw *shakes fist* 16:08
adamw you little... 16:08
kparal I was wondering if you get it 16:08
* jskladan *misfire* 16:08
adamw does anyone actually know where dgilmore is today? 16:09
adamw #endmeeting 16:09

Generated by irclog2html.py 2.8 by Marius Gedminas - find it at mg.pov.lt!