[00:25] <cyphermox> anyone here willing to help me test a fix for the CD-ROM ejection bug? I should have a package ready any minute that looks okay on my end
[09:27] <elfy> cyphermox: in vbox I got corruption, but power off and restart, it had ejected. In kvm it just seemed to hang - had to poweroff - but again it had ejected
[09:27] <elfy> on a laptop I confirmed bug 1436715
[10:19] <flexiondotorg> The is an update to caja in the upload queue. Could someone let it through please?
[10:19] <flexiondotorg> Just contains a fix for a segfault.
[12:03] <flexiondotorg> Caja is in the upload queue. It fixes a segfault.
[12:03] <flexiondotorg> Would somebody be good enough to let in to the vivid archive please?
[12:04] <brendand> flexiondotorg, see topic
[12:05] <flexiondotorg> When beta 2 is done, will I be permitted to have the updated Caja included in vivid
[12:05] <flexiondotorg> Been stuck onm
[12:06] <flexiondotorg> The motorway for 3 hours so doing this on my phone.
[12:07] <brendand> flexiondotorg, do you have an FFe?
[12:09] <cjwatson> brendand: bug fixes don't require FFe
[12:14] <brendand> cjwatson, indeed - that was the next question (which yes, could have been asked simultaneously)
[12:15] <cjwatson> Well, you can look at the queue yourself :)
[12:45] <cyphermox> good morning!
[12:45] <cyphermox> elfy: thanks, I was aware, I tested the ubuntu iso early after respin.... which means very late last night or very early this morning
[12:46] <cyphermox> this needs more work to make sure all the systemd jobs get started properly, I guess
[13:50] <amjjawad> hello everyone
[13:50] <amjjawad> I have a problem: http://iso.qa.ubuntu.com/qatracker/milestones/336/builds/91028/testcases
[13:50] <amjjawad> installer crashes :(
[13:50] <amjjawad> should I re-spin or it is too late?!
[13:52] <infinity> amjjawad: What would a respin do for you?
[13:53] <amjjawad> infinity, not really sure, it is 1:00am here and I'm super dead tired and sleepy
[13:53] <amjjawad> and can't install anything
[13:53] <amjjawad> it crashes
[13:54] <infinity> amjjawad: It sounds more like a virtualbox bug?  But I suppose it could be a broken ISO somehow.  Let me grab the GNOME ISO quickly and see how it fares here.
[13:55] <amjjawad> infinity, ok, thanks!
[14:04] <ara> infinity, ping
[14:06] <infinity> ara: If it's the usual question, it's happening this morning because if it doesn't, I won't have room to release the beta.
[14:06] <infinity> ara: If it's another question, what's up?
[14:06] <ara> infinity, it was that one :)
[14:07] <ara> infinity, why is it taking so long? anything we can do differently to avoid this from happening again when 14.04.2 gets released?
[14:07] <ara> sorry, .3
[14:08] <ara> infinity, can we add it to the release process steps?
[14:08] <infinity> ara: I'm afraid I don't have a good answer to the why.  It just kept slipping through the cracks.  Will be improving things for .3, yes.
[14:09] <ara> infinity, thanks :)
[14:13] <amjjawad> infinity, same goes to i386 :(
[14:14] <Riddell> can someone add upgrade to iso tracker? or tell me how to add?
[14:15] <apw> amjjawad, that would be expected if it is virtual box blowing up, and if you are saying your hosts virtual box is blowing up ... that sounds very much like a bug in the host not the guest
[14:15] <amjjawad> apw, I can understand but never happened to me as I always test and this is the very first time ..
[14:16] <apw> amjjawad, and had your host stayed the same for the entire time or have you also upgraded that ?
[14:16] <amjjawad> also, not sure if this bug: https://bugs.launchpad.net/ubuntu/+source/casper/+bug/1436715 is critical or not
[14:17] <infinity> amjjawad: Not sure that we'll be able to fix that one for the beta, but it's certainly critical for final release.
[14:17] <amjjawad> I see, thanks infinity but what about the crashing :(
[14:20] <amjjawad> apw, usually, the vb complains if there is a newer version but it didn't for a while now
[14:21] <amjjawad> and I am sure this is the first time apw
[14:22] <apw> amjjawad, people are downloading your image now to test it, takes a little time
[14:23] <amjjawad> apw, it is 1:22am now here and I'm doing my best to stay awake but ... so tired :(
[14:23] <amjjawad> maybe I should sleep and check in the morning
[14:23] <amjjawad> infinity, how many hours left to the release ?
[14:24] <infinity> amjjawad: Undecided.
[14:25] <amjjawad> I can't hold it more ... must sleep so will check later infinity
[14:26] <amjjawad> just saw this infinity : https://lists.ubuntu.com/archives/ubuntu-gnome/2015-March/002882.html
[14:29] <infinity> amjjawad: I'm testing in qemu/kvm right now...
[14:31] <amjjawad> infinity, thanks a lot
[14:36] <apw> amjjawad, i am installing under VB on a vivid host right now
[14:36] <amjjawad> apw, many thanks
[14:40] <stokachu> infinity: wrt 'openstack' looks like I uploaded it a day or two late, it replaces cloud-installer package which is in the archive already
[14:40] <stokachu> infinity: honestly it isn't that bad if it doesn't make it in vivid since its not lts
[14:41] <apw> amjjawad, ok it installed ocmpletely, i think i am seeing the other "reboot does not work" issue
[14:41] <infinity> apw: Yeah, everyone should see that one.
[14:41] <amjjawad> apw, many thanks so that means my Oracle VB is acting naughty?
[14:42] <infinity> Hrm.  Which, to be fair, could occasionally cause a corrupted target filesystem, if it never syncs.
[14:42] <apw> amjjawad, yep forced it off and it boots just fine into GUbuntu
[14:42] <infinity> I really don't want to respin again today...
[14:42] <apw> infinity, if yours boots and installs ok, i'd say its local to amjjawad
[14:43] <infinity> cyphermox: What are the odds that we could fix the "not listening for keypress" issue (at least, I assume that's what it is, we're just blocking forever on a keypress we'll never see)
[14:43] <infinity> apw: yeah, it installed fine, and hard-killing it after the failure to reboot brings up the installed system fine.
[14:43] <cyphermox> pretty good, I'm assuming it's just more of the same systemd unit ordering issue
[14:43] <infinity> amjjawad: ^-- Under kvm.
[14:43] <stokachu> infinity: hmm feature freeze was feb 19th and the package was uploaded on the 19th
[14:43] <cyphermox> maybe the casper job isn't happening at the right place.
[14:44] <infinity> stokachu: Oh, I'm not asking for FFes or anything here, I just literally wanted to know what it was about. :P
[14:44] <amjjawad> infinity, I'm using Oracle VB
[14:44] <stokachu> infinity: ahhhh
[14:44] <infinity> stokachu: It was an idle question, not a solicitation for process.
[14:44] <apw> amjjawad, i am using VB here without issue on a vivid host on those images
[14:44] <stokachu> infinity: ah ok, sorry i thought I messed up somewhere
[14:44] <stokachu> infinity: yea basically its the Ubuntu OpenStack installer renamed from cloud-installer to openstack package
[14:45] <infinity> cyphermox: Ordering seems like a weird thing to blame.  Unless there's a job that says "hey, stop listening to input devices trololol".
[14:45] <amjjawad> apw, then I have no other choice but to trust your words and accept them and blame my machine.
[14:45] <amjjawad> but I don't know if it is wise to mark these images as ready
[14:45] <infinity> cyphermox: I mean, your job is obviously running, since it's popping out CD trays all over the land.  So, it's a question of why isn't plymouth getting control.
[14:45] <amjjawad> giving that just me and lance tested them
[14:46] <cyphermox> infinity: well, it seems likely to me that it's not getting the input because either plymouth is started but not quite started right, or the casper job isn't started at the right moment, or the casper job breaks plymouth by not caching it properly.
[14:46] <amjjawad> at least, that's what the tracker says
[14:46] <apw> amjjawad, well i suspect we may have to get more testing on them indeed if you are normally tester for themm
[14:46] <cyphermox> infinity: so that's what I meant by ordering.
[14:47] <infinity> cyphermox: Fair enough.
[14:47] <infinity> cyphermox: So, historically, the caching issue only led to a lack of display, but the input still worked (so, people whacked [enter] blindly on their "hung" computers, and they'd reboot).
[14:47] <cyphermox> yes
[14:48] <infinity> cyphermox: But that if we're now running it a tiny bit later, I could see how it would just fail to work entirely.
[14:48] <infinity> s/that //
[14:48] <cyphermox> as far as I know, we are
[14:48] <cyphermox> we're running this about as late as it seems we could
[14:48] <amjjawad> apw, I sent more than one email and posted on our social media channels but not sure why we have very few testing :(
[14:48] <infinity> Yeah.  Which, to be fair, seems like the right time to run it.
[14:48] <cyphermox> (from my very limited understanding of it)
[14:48] <cyphermox> yeah
[14:49] <infinity> But will break miserably if the filesystem is effectively already gone.
[14:50] <cyphermox> fwiw, the call to plymouth to display the message happens before eject, and the input listening happens after, so my money is on caching for now
[14:51] <cyphermox> otoh, it could also just display the message, not wait for input and reboot if it was just not getting the request for input
[14:56] <infinity> cyphermox: You know, it could be worse.  It could be the whole system just freezing immediately after eject.
[14:56] <cyphermox> it could be
[14:56] <cyphermox> it doesn't seem to explode as badly if you don't run with splash though
[15:20] <elfy> cyphermox: I thought you probably were - but just in case I mentioned it :)
[15:23] <cyphermox> no problem
[15:26] <cyphermox> elfy: that's what I'm working on right now
[15:26] <elfy> ok :)
[15:32] <infinity> cyphermox: So, I think we need to make a hard decision soon for "try to fix or just release note" on the reboot thing.  If it looks like it'll be architecturally painful (or time consuming) to fix, I think we'll just have to pass and fix it ASAP post-beta so we don't forget.
[15:32] <cyphermox> fixed it!
[15:32] <cyphermox> :D
[15:33] <cyphermox> let me make sure it's really really good
[15:35] <cyphermox> elfy: want to help test the fix?
[15:35] <mdeslaur> infinity: who do I need to bug to get vlc approved, the mate folk?
[15:35] <cyphermox> it's just a matter of editing /sbin/casper-stop and adding /lib/systemd in place of /etc/rc?.d in the caching list
[15:35] <mdeslaur> (not sure what "mozilla" is in the seed there)
[15:37] <infinity> mdeslaur: Looks like mate and myth, according to seeded-in-ubuntu, and myth is lts-only, so just mate.
[15:37] <mdeslaur> infinity: who's in charge of mate?
[15:38] <mdeslaur> yo, mate people, speak up! :)
[15:38] <cyphermox> flexiondotorg: ^
[15:38] <elfy> cyphermox: if possible - always happy to help
[15:39] <cyphermox> I had a good response in console in qemu, so now it's hardware testing
[15:39] <elfy> cyphermox: and this would work from a livesession ?
[15:39] <cyphermox> yeah
[15:39] <cyphermox> you can just sudo vi it
[15:40] <elfy> okey doke - just booting laptop then
[15:40] <infinity> cyphermox: Does this magically give us a visual prompt too, or just make [enter] work blindly again, as before?
[15:40] <cyphermox> maybe it's not quite there though, still no response
[15:41] <cyphermox> infinity: you already do have the visual prompt
[15:41] <infinity> cyphermox: I sure don't.
[15:41] <cyphermox> where?
[15:41] <infinity> cyphermox: At least, not on the xubuntu I tested a while ago.
[15:41] <infinity> cyphermox: Maybe it's a timing thing, but on xubuntu, when you click the "reboot" button from ubiquity, the system basically just hangs there.
[15:42] <cyphermox> I'm still using the dvd I burned for ubuntu last night
[15:42] <cyphermox> in qemu?
[15:42] <infinity> Yeah.
[15:42] <cyphermox> I also get a hang, the display doesn't change
[15:42] <elfy> infinity: I get similar in vbox and kvm - just hangs, but h/w gives the Enter and remove
[15:43] <infinity> elfy: Fun.  So totally a timing thing.
[15:43] <cyphermox> it does fix the rebooting from a VT with no splash, though, on qemu
[15:43] <infinity> cyphermox: Well, I'm about 75% on the way to just calling the current images "good enough" and starting with all the paperwork, but if you can come up with something better than the current state, lemme know.
[15:43] <cyphermox> ok
[15:43] <infinity> I've done no paperwork at all this cycle, so ugh... Whee.
[15:46] <cyphermox> arf
[15:48] <elfy> cyphermox: well - that didn't work, assuming not using vi isn't an issue
[15:52] <cyphermox> no, I noticed, it only seems to help a bit with the vt-only reboot, not with teh splash
[15:56] <infinity> mdeslaur: 3.3MB diff from 2.2.0~rc2 to 2.2.0?  How did you manage that feat?
[16:16] <mdeslaur> infinity: heh, don't blame me :)
[16:53] <infinity> cyphermox: Any news, or am I just calling this good enough?
[16:54] <cyphermox> good enough
[16:54] <cyphermox> it's going to take a few more hours to get it better
[16:54] <cyphermox> and I am hungry :/
[16:54] <infinity> Alright.  Spinning source ISOs and tying off some bows.
[16:55] <infinity> ... after I teach publish-image-set about MATE.
[16:56] <cyphermox> ah?
[16:57] <ianorlin_> argh not much lubuntu testing got done after the last respin
[16:57] <infinity> ianorlin_: All that really should be needed is boot/install/reboot smoketesting to make sure it didn't somehow regress horribly.
[17:02] <flexiondotorg> I've been stranded in the car all day.
[17:02] <flexiondotorg> 'sup with the release?
[17:02] <flexiondotorg> See new iso images. Will do a quick test.
[17:02] <flexiondotorg> Is there a release ETA?
[17:02] <infinity> flexiondotorg: ETA is "later today", but these are the images I intend to ship unless someone decides to pull their flavour (or an image) at the last minute.
[17:03] <flexiondotorg> infinity, Understood. Looking at the tracker it seems things are in better shape than yesterday.
[17:03] <flexiondotorg> :)
[17:03] <infinity> flexiondotorg: Well, MATE looks entirely untested right now, so some smoketesting there would be nice.
[17:04] <infinity> flexiondotorg: Note that there is a reboot bug (or, rather, a failure to reboot), which we're not going to fix before release, but it's bad enough that I think I'll mention it in the release announcement, not just bury it in the release notes.
[17:04] <flexiondotorg> infinity, Will do. But we did extensive testing yesterday. The only issue we encountered appear half (mostly) fixed now.
[17:05] <flexiondotorg> infinity, That is the half fixed thing to which I refer :) I think the current situation is better than no eject prompt.
[17:06] <infinity> flexiondotorg: I'm on the fence about which is worse, but really, we just need to fix it properly before final.
[17:07] <rcj> infinity, cloud images are ready to go
[17:07] <infinity> rcj: Shiny.
[17:08] <rcj> infinity, I'll be watching for the go-ahead to start publication
[17:08] <flexiondotorg> infinity, Downloading now. But, I live in a rural area and connected to the Internet with short wave radio, so.....
[17:09] <infinity> rcj: It'll be several (many?) hours, but we don't have to be perfectly in sync either.  If you push early, you'll be all set up by the time I do the release announcement after all.
[17:09] <infinity> flexiondotorg: I assume zsync helps a bit.
[17:09] <flexiondotorg> I t   d o e s . . .
[17:09] <flexiondotorg> Still 13 mins for me though.
[17:10] <infinity> Ouch.
[17:10] <rcj> infinity, I'll kick things off in ~2 hours to get them going
[17:10] <flexiondotorg> Yeah, but the view is lovely and the cows have nice grass to chew on ;)
[17:13] <flexiondotorg> infinity, This is my first official cycle and I'm still learning.
[17:13] <flexiondotorg> Can you clear something up for me
[17:14] <flexiondotorg> Caja is in the upload queue. It was a sync from Debian to fix a segafault.
[17:14] <infinity> flexiondotorg: It'll get in right after we push out the images, so people will get it on upgrade.
[17:15] <flexiondotorg> Do I have to request any new packages be let in from now on or when Beta2 is over will that stuff just "flow in"?
[17:15] <flexiondotorg> infinity, Thanks.
[17:15] <infinity> flexiondotorg: For bugfixy things, you shouldn't need to make any justifications, we'll just review and let it in.  For larger things that should have had Feature Freeze Exceptions, you might need to make a case.
[17:16] <flexiondotorg> infinity, I'm only doing bug fixes.
[17:16] <infinity> flexiondotorg: Right, then you don't have much to worry about, just upload as usual.
[17:16] <flexiondotorg> infinity, But I understand I can submit package updates that just carry new translation stings until April 9th, right?
[17:18] <infinity> flexiondotorg: You can keep pushing bugfixes pretty much right up until final release.  The non-langpack-deadline is actually a weird thing for just a few packages, and not really worth thinking about.
[17:19] <infinity> flexiondotorg: New master strings were frozen long ago, but for packages that only affect you, you have some leeway if you'd prefer to break translations for a good cause.
[17:19] <flexiondotorg> So, I can request new uploads for a couple of package that just carry new translations early next week?
[17:20] <infinity> flexiondotorg: Yeah.  If you're just adding new translations, that can really happen any time.  It's new master strings that are problematic, as they invalidate all the other translations.
[17:20] <flexiondotorg> infinity, Thanks. Understood. I'm not adding new msater strings.
[17:21]  * infinity goes to find a snack while source ISOs build.
[17:21]  * flexiondotorg haz iso
[17:27] <mitya57> Hi, can someone please reject my gnome-flashback upload from the queue?
[17:28] <mitya57> The changelog entry is wrong + I want to include one more fix.
[17:29] <mitya57> (In vivid)
[17:33] <infinity> mitya57: Done.
[17:33] <mitya57> thanks
[18:13] <jibel> infinity, desktop testing is almost done. The major issue is OEM installations. Sometimes the session doesn't start (X or unity-settings-daemon crash bug 1436861) and the OEM user is not removed ( bug 1436936 )
[18:14] <jibel> bug 1436937
[18:15] <infinity> jibel: That second one looks like another "needs porting to systemd" bug.
[18:16] <infinity> jibel: Can you note whatever you think is relevant in the release notes?
[18:17] <jibel> infinity, OK, we'll just verify that the end user is admin of his system even if OEM is still present.
[18:33] <elfy> infinity: clarifying - we're going with what we have now?
[18:33] <infinity> elfy: Yeah, that's the plan.
[18:34] <elfy> ta :)
[19:13] <flexiondotorg> infinity, I'm also testing the OEM user thing.
[19:20] <ianorlin_> I don't really see a fix for bug 1432843 coming out in a few hours
[19:38] <flexiondotorg> infinity, When do you need confirmations?
[19:41] <infinity> ianorlin: No, and I'm not sure exactly what to make of that one.
[19:42] <infinity> flexiondotorg: Soon.
[19:42] <flexiondotorg> infinity, OK. I'll test faster...
[19:42] <ianorlin> at least the alternate for 32 big lubuntu will work but I don't see how that will get fixed in time
[19:43] <infinity> flexiondotorg: I'm not in a huge rush, still have a bunch of cleanup to do on the master cdimage system.
[19:44] <infinity> ianorlin: Well, the reason I don't know what to make of that bug is that the claim is that Ubuntu i386 has the same problem, which no other testers are seeing.
[19:45] <infinity> ianorlin: So, if it's simply very hardware specific, we can certainly test that and release it still.
[19:45] <ianorlin> infinity: there isn't a boot option for turning zram off is there
[19:46] <ianorlin> infinity: the wierd thing is I had that same hardware install amd64 just fine
[19:48] <infinity> ianorlin: There are no boot options to influence zram, no.
[19:51] <cjwatson> I'm pretty sure "nocompcache" disables it
[19:51] <cjwatson> Used to, at least
[19:53] <infinity> cjwatson: Ahh, for the builtin one, yeah, but they also have zram-config seeded.  I wonder if that could interact poorly.
[19:53] <infinity> I really should have merged those ages ago.
[19:54] <ianorlin> um I just tried with nocompcache on the i386 dialy and the installer started on the same hardware it failed on earlier
[19:55] <infinity> ianorlin: So, probably not an xorg bug.
[19:55]  * infinity wonders why you have zram-config seeded to lubuntu-live
[19:56] <ianorlin> there were certain machines were it made a huge performance difference on the low end when they don't yet have swap so installs would take a really long time if you were say resizing an ntfs partition
[20:02] <infinity> ianorlin: So, this absolutely sounds like something we can fix over the next few days.  But not right now.  If you're comfy with telling people to boot with "nocompcache" for the lubuntu installer, we can release what you have and fix it later.
[20:02] <infinity> How this ever really worked is a mystery to me, though.
[20:04] <infinity> ianorlin: If you're not comfortable with that, we can skip releasing the i386 desktop image for the beta.
[20:24] <flexiondotorg> infinity, cyphermox OEM config is quite busted.
[20:26] <flexiondotorg> infinity, cyphermox I confirm the the oem user is left behind and has admin rights.
[20:26] <flexiondotorg> Or rather sudo rights.
[20:31] <flexiondotorg> infinity, cyphermox I've updated https://bugs.launchpad.net/ubuntu/+source/ubiquity/+bug/1436937
[20:37] <rcj> infinity, cloud images will be live in a bit, I've kicked it off before I have to drop off for a bit.
[20:50] <infinity> darkxst: Around?
[20:51] <darkxst> infinity, morning
[20:53] <infinity> darkxst: Hey.  You have enough context and a willingness to sign off on the gnome images as "good enough"?
[20:54] <infinity> darkxst: amjjawad was apparently having some issues with them in his local virtualbox setup, but apw and I couldn't reproduce in vbox or kvm.
[20:57] <darkxst> infinity, yeh, I not heard of any other reports of that crash
[21:00] <flexiondotorg> infinity, All set for Ubuntu MATE.
[21:01] <infinity> flexiondotorg: Has that powerpc image actually been tested?
[21:01] <flexiondotorg> infinity, Yes, but none of the buggers could be arsed to do it properly.
[21:01] <flexiondotorg> infinity, As far as I can tell depending on the system it works.
[21:01] <infinity> Heh.
[21:02] <infinity> Yeah, it'll be hit and miss, given that it's > 10 year old hardware they're playing with.
[21:02] <flexiondotorg> infinity, There are regressions though. X is dying on iBook G4s which worked fine in Beta 1.
[21:03] <flexiondotorg> I think the new Xorg is busted for PowerPC. But, they are a motivated group and maintain custom versions of mesa etc. So, I'll try and get them engaged.
[21:04] <infinity> flexiondotorg: Yeah, custom versions don't really help anyone, you should see about getting them engaged with upstream or us or both.
[21:04] <flexiondotorg> infinity, I intend to for 15.10.
[21:04] <flexiondotorg> I want them to help with the initrd customisations.
[21:04] <infinity> ianorlin: Alright, it's all on you and your last three unmarked images.  What's the story with them?
[21:04] <flexiondotorg> So, they don't patch post-install.
[21:05] <ianorlin> I think wxl wanted powerpc for dialies but lts only
[21:05] <infinity> ianorlin: Okay, so that just leaves i386 desktop.
[21:06] <ianorlin> nocompcache still failed on a kvm virtual machine so it doesn't solve everything
[21:07] <ianorlin> I sent out a mail asking other people to see if nocompache solves it but no repsonse yet
[21:08] <ianorlin> still think desktop i386 is pretty broken
[21:11] <infinity> ianorlin: We can skip publishing it, up to you.
[21:12] <ianorlin> I think that is best since the alternate works well enough
[21:12] <infinity> ianorlin: Alrighty.
[22:51] <Riddell> infinity: ETA to announce?
[22:53] <infinity> Riddell: Just synced mirrors now, so going to give in some time to settle.  Maybe a couple of hours.
[22:55] <Riddell> thanks, I'll take my laptop home
[23:15] <ianorlin> I have my release notes at https://wiki.ubuntu.com/VividVervet/Beta2/Lubuntu but need to get grocieres so will be afk for  a bit