=== e-jat is now known as ejat === Wajih is now known as MaWaLe === highvolt1ge is now known as highvoltage === dholbach_ is now known as dholbach === asac_ is now known as asac === cjwatson_ is now known as cjwatson === ogra_ is now known as ogra === ogasawara__ is now known as ogasawara === Zic is now known as Guest74942 [14:00] Hi there [14:00] Where's NCommander? [14:00] hi [14:01] hey [14:01] ogra, GrueMaster, plars, davidm, JamieBennett, dyfet: meeting? [14:01] hi [14:01] if I must. [14:01] yarg [14:01] yup [14:01] oooh [14:01] * ogra totally forgot [14:01] i'm so ddep into filing bugs atm [14:01] *deep [14:02] hello [14:02] * GrueMaster stragles out to get a fresh cup of caffeine. [14:02] I guess we'll do without NC [14:02] NCommander [14:02] #startmeeting [14:02] Meeting started at 08:02. The chair is lool. [14:02] Commands Available: [TOPIC], [IDEA], [ACTION], [AGREED], [LINK], [VOTE] [14:02] * lool sighs on the lack of a meeting page [14:03] Nor for last week, impressive [14:03] [link] https://wiki.ubuntu.com/MobileTeam/Roadmap [14:03] LINK received: https://wiki.ubuntu.com/MobileTeam/Roadmap [14:04] First action on completing dove z0 d-i support was on ncommander [14:05] But I don't think he worked further on it; I agree we should first confirm whether that's still a target since bjf was saying it might not be supported anymore [14:05] move lubuntu seed to proper location and enable the meta on amd64 (dyfet) [14:05] dyfet: ? [14:06] oh...we want amd64 active for that? Then I will change architecture to any [14:06] That I think would be simplest... [14:06] I don't see why lubuntu should be architecture specific [14:06] I do not recall why this was done either originally [14:06] This decision is usually defered to image building time because these cost QA and space [14:07] dyfet: So what about the other bits? [14:07] dyfet: Moving to proper location (/ubuntu-seeds) [14:07] Agreed...and a quick tiny change :). I had been trying to sync up with the other people using the branch to move the seed. But it is a team maintained seed [14:08] NCommander, is here now [14:08] #startmeeting [14:08] NCommander, There is already a meeting in progress. [14:08] ah [14:08] dyfet: so carry on I guess [14:08] okay [14:08] [action] move lubuntu seed to proper location and enable the meta on amd64 (dyfet, c/o) [14:08] ACTION received: move lubuntu seed to proper location and enable the meta on amd64 (dyfet, c/o) [14:09] setup meeting with ubuntu-moblin and report back here (lool, c/o) [14:09] Had a meeting on Monday where we planned our various efforts [14:09] A short summary is that there are some 4 parallel efforts going on or so [14:09] We're trying to merge bug fixes from private jaunty ppa to public moblin ppa [14:09] we're trying to merge karmic public ppa to karmic proper [14:09] we're trying to fix more bugs [14:10] we're trying to merge moblin 2.1 [14:10] (and we're also trying to make the jaunty bits public in some way) [14:10] chrischeney will be tasked with the jaunty -> any merges [14:10] paulliu keeps working on moblin 2.1 [14:10] rest of OEM folks work on bug fixes and moblin 2.1 [14:10] lool: yes. [14:11] We touched based on a plan for some partial moblin compliance too [14:11] That's it [14:11] Image is quite broken right now due to move to 2.1 [14:11] but that's life and we want moblin 2.1 more than a 100% karmic based image which doesn't seem atteinable anymore [14:11] lool: Update mutter-moblin should brings everything back. [14:11] file question against soyuz on PPA size (paulliu, c/o) [14:12] paulliu: ^ did you file that? [14:12] lool: how come I wasn't invited to the meeting? [14:12] lool: Ah. Not yet. (c/o) [14:12] Seeing as how I'm doing the compliance testing. [14:12] lool: But we haven't have a new PPA. [14:12] GrueMaster: We only discussed how to organize package uploads and this was appended to the meeting, not really part of the agenda to start with [14:13] lool: The current PPA size is large enough. If we want a new PPA we need to have the same size. [14:13] GrueMaster: The only thing we discussed on this topic was basically that we want to keep this in a separate compliance ppa and perhaps base on jaunty [14:13] paulliu: Do we want one? Not sure we actually need one anymore [14:13] paulliu: I guess not [14:13] lool: yes. So let's delete that action. [14:14] plars to update the bug workflow [14:14] plars: ^ [14:14] that has been mostly completed, the only thing I hadn't done yet was add a section for moblin [14:14] Ok thanks [14:14] which I was kind of on the fence about one part, and that would be for the 'moblin' tag [14:15] I was thinking it may not be necessary, since all moblin bugs are supposed to go into ubuntu-moblin [14:15] however, after thinking about it some more, I could see how someone might file a bug against a non-moblin specific package, and it end up being in ubuntu with possibly nothing indicating that it happened on a moblin system [14:15] plars: indeed not necessary for now but doesn't hurt much to plan for next cycle and start adding that perhaps? [14:15] ack [14:15] so I'm thinking it would still be useful to have [14:15] plars: makes sense to have it IMO [14:16] so I'm going to go ahead with that, and see about putting the hook in for it too [14:16] we want the tag to just be 'moblin' right? [14:16] GrueMaster to continue driving moblin compliance issues with the LinuxFoundation. [14:16] plars: yeah [14:16] or is something like 'ubuntu-umr' or 'ubuntu-moblin' more paletable? [14:16] ok [14:17] I talked with Ted Tso and a couple of test developers, found out what needs to be installed for testing compliance. [14:18] There were a lot of missing bits, like locale info and a few other missing packages that the tests depend on. [14:18] I'll add those in to the next test round. [14:18] The list I have is non-distro specific, so once I know what ubuntu packages are needed, I'll request a meta package wrapper to be made. [14:19] lool to add uboot-mkimage to the desktop ship seed to make dove images non-dependent on internet access [14:19] done [14:19] pushed today; need to see if that works as expected [14:19] As to actual compliance, the final specification for 2.0 won't be released until after karmic. [14:20] * GrueMaster is done. [14:20] [topic] UNR status [14:20] New Topic: UNR status [14:20] njpatel: You around? [14:20] njpatel: Any particular comments? [14:21] AFAIK we're up to date on releases [14:21] Except for a tentative humanity upload which I nacked [14:22] Nothing particularly interesting to report [14:22] [link] UNR Status [14:22] LINK received: UNR Status [14:22] Ups [14:22] [link] https://bugs.launchpad.net/ubuntu/karmic/+bugs?field.subscriber=ubuntu-unr [14:22] LINK received: https://bugs.launchpad.net/ubuntu/karmic/+bugs?field.subscriber=ubuntu-unr [14:22] No interesting bugs there [14:22] Biggest issue with icon theme has been resolved by saying we keep what we have [14:23] Apparently there was a new bug filed recently against maximus that isn't listed here. [14:23] #? [14:23] bug #421019 [14:23] If someone has some free cycles, there's an easy bug with translations of window-picker-applet where I outlined the necessary changes and they need implementation [14:23] Launchpad bug 421019 in maximus "maximus crashed with SIGSEGV in g_closure_invoke()" [High,Triaged] https://launchpad.net/bugs/421019 [14:24] Ah, someone has already looked at it. [14:25] does it make sense to report such an obvious bug like flash failing to install from firefox plugin finder service ? Does anyone know a bug report about it ? [14:25] Can't personally load it [14:25] I mean it takes forever [14:25] ops, wrong channel, sorry :| [14:25] GrueMaster: Is it properly milestoned etc.? [14:25] my current concern with unr is that desktop-switcher seems unusable without reeking havoc [14:25] plars: Indeed [14:26] plars triaged this one. [14:26] plars: So we tried clamping the worst issues last week but I still see it as a sore poin [14:26] I wanted to talk to neil about it first before suggesting anything rash [14:26] point [14:26] plars: last times I looked into d-s issues, I gazed and thought we need to rewrite everything [14:26] Problem is that we can't do that for karmic [14:27] It should of course be a lucid spec to use a different mechanism for it [14:27] lool: jason had suggested that it should just be a session option at login, which seems reasonable === robbiew-afk is now known as robbiew [14:27] the on-the-fly switching is a neat trick, but doesn't seem to work well in practice [14:27] * StevenK shores [14:27] plars: Yes; I suggested that too, months ago, but nobody had time to look into it [14:27] StevenK: Hey; we're in UNR status [14:27] topic is d-s [14:27] it's the only really good option I've heard so far === fader|away is now known as fader_ [14:28] d-s is broken? [14:28] plars: Cant do that in karmic though; I had in mind we should do our best to fix the biggest issues in karmic's [14:28] StevenK: it never actually worked [14:28] Sure it has, it's just ... fragile [14:28] plars: Could you come up with the highest prio bugs in your eyes which we should fix for release? [14:28] lool: right, but if d-s can't be stabilized enough, I think it needs to not be there. It's breaking systems [14:28] will do [14:28] plars: It's an option [14:29] plars: I think I'd need a summary of breakage from you if you could produce that [14:29] hey, that would also reduce the image size. Win. [14:29] plars: That'd help convincing release team + design team that we want it out [14:29] GrueMaster: Speaking of which [14:29] GrueMaster: How did that effort go recently? [14:29] I haven't had time to look into it. I was busy withthe conference last week. [14:30] I'm looking into it today. [14:30] We need to wait until after beta for large-ish things anyway [14:30] StevenK: anything WRT UNR? [14:31] Yes, we're oversized. [14:31] Right now? [14:31] We seem to be at 686M [14:31] * StevenK checks again [14:32] Okay, we were ... :-) [14:32] We hit 714M this morning [14:32] wow [14:32] how come? [14:33] I wasn't able to figure that out, sadly [14:33] Ok; probably temporary bogosity [14:33] But it looks like we dropped like a stone after that if we're at 686 [14:33] StevenK: any other bugs / highlights? [14:33] I'll re-instate de after beta [14:33] StevenK: are we in sync WRT seeds and upstream releases? [14:34] StevenK: Well I remember pidgin crept in over the week end and I fixed it but it probably was accepted on monday only [14:34] that might explain the big image [14:34] I've not checked those, but we are probably too late to change either [14:34] StevenK: Can you make sure you double check we're not out of date? [14:34] StevenK: I'd hate if we would be missing any change, even the most trivial ones [14:35] lool: I certainly will, when I'm working again [14:35] Thanks [14:35] [topic] moblin remix status [14:35] New Topic: moblin remix status [14:35] StevenK: hey again! [14:35] * StevenK chuckles [14:35] paulliu, StevenK: So it's quite borken right now [14:35] paulliu: You say it's about to be mostly fixed? [14:36] lool: yes. If you boot with a while screen. You just need to upgrade mutter-moblin. [14:36] [link] https://bugs.launchpad.net/ubuntu/karmic/+bugs?field.subscriber=ubuntu-moblin [14:36] LINK received: https://bugs.launchpad.net/ubuntu/karmic/+bugs?field.subscriber=ubuntu-moblin [14:36] lool: I think it might not included in the built today. [14:36] paulliu: Is it still booting into gnome? [14:37] paulliu: So the current livefs needs a fixed mutter-moblin? [14:37] I'll take care of the sources.list and apt preferences files in livecd-rootfs, but after beta [14:37] StevenK: yes. [14:37] paulliu: When was it uploaded? [14:37] paulliu: We had a build an hour ago or so [14:37] lool: one hour ago. [14:37] note that dues to a casper breakage we dont have ttys on live images [14:37] mutter-moblin 0.40.0-0ubuntu0moblin6 [14:38] that's what we have [14:38] Okay, we can organise a rebuild [14:38] so if you have a white screen there is no way to install anything on tty [14:38] lool: ok. It's ok. [14:38] ogra: Is that one fixed? [14:38] paulliu: Cool [14:38] lool, slangasek said post beta [14:38] paulliu: no rebuild needed then? [14:38] ogra: Ok [14:38] lool: no rebuild needed. [14:38] Anything else on moblin for beta? [14:38] Bug 438678 filed ... [14:38] plars, GrueMaster: ^? [14:39] That tty issue may be related to a kernel bug I filed yesterday. [14:39] GrueMaster, it isnt [14:39] Launchpad bug 438678 in casper "25configure_init uses /dev/ttyN.conf instead of /dev/ttyN for /etc/init/tty*.conf mangling" [Medium,Confirmed] https://launchpad.net/bugs/438678 [14:39] ok [14:39] lool: I've not had time to focus on moblin this week, couldn't say [14:39] Is it only breaks Moblin?? [14:39] nope [14:39] all live images [14:39] No, it would affect everything [14:40] Hmm if it's in the installed system too it certainly should be > medium [14:40] ogra: OK. thanks. That should be the bug mentioned by Kevin Huang today afternoon. [14:40] It's more casper and upstart fallout [14:40] but if moblin leaves you with unusable X it is hard :) [14:40] ogra: He complains that moblin installation is totally broken. [14:40] lool, my focus last week was on moblin compliance, based on Alpha 6. [14:40] I didn't get into any testing yet. [14:40] GrueMaster: Do we have public results? [14:40] paulliu: Could you ask him to file bugs? [14:41] GrueMaster: Are you automating this stuff? [14:41] public results for moblin compliance testing? not yet. [14:41] paulliu: Or at least document what is broken, and how it breaks [14:41] * StevenK prods LP with a sharp stick [14:41] StevenK: Hmm. Not good because it's a secret project I think. A hardware runs nvidia driver. Not Intel. [14:42] The testing is automated. Last week i was in a full week long conference, where I managed to corner some members of the linux foundation that work on the compliance test suite. [14:42] GrueMaster: So this runs from cron on your side? [14:42] GrueMaster: Were you armed or unarmed? [14:42] GrueMaster: What I'd like to see are daily builds of the testsuite against the daily image so that we can see progress on compliance fixing [14:43] No. It is manually started. I really don't have time to give a full education on how the test system works. [14:43] Not possible. [14:43] I don't like reading not possible [14:43] Test suite takes 26 hours to run it's course, not including manual tests. [14:43] You want it done faster, buy me more hardware. [14:43] I don't want to have to rely on you to push the button [14:44] GrueMaster: Well actually I did wonder a couple of times why it's running in your basement instead of in our shiny DCs [14:44] It's my job, it's what I do. [14:44] DC's? [14:44] data ceners [14:44] Datacentre [14:44] *centers [14:45] Some of the tests require hardware that you don't see when running in a VM. [14:45] Like the GL tests. [14:45] We don't need to run them in a vm in the DC [14:45] We have our hw cert labs and our regular DCs which could host a new machine [14:45] But ideally this would run on hw cert machines [14:46] And who would run the manual tests? [14:46] GrueMaster: Could you look at documenting what you run and how you setup a machine to rnu the testsuite? [14:46] And what access it needs to the machine [14:46] GrueMaster: I don't know what the manual tests are [14:46] It is already documented on the moblin.org site, along with some new hints for missing packages I just got last week. [14:47] let's take this offline. [14:47] Does it need root, will it spew 80G of garbage to a disk ... [14:47] Ok; I agree it takes too long to discuss here [14:47] GrueMaster: But I would like to see more transparency on the setup and less dependencies on you pressing abutton [14:47] Do we really want to spend the next 15 minutes with a tutorial on how to run the test suite? [14:48] ... not here [14:48] * StevenK has a bed to visit [14:48] Anything else moblin related? [14:48] GrueMaster: we should talk about seeing if we can make it run in checkbox [14:48] plars, GrueMaster: That sounds like a good UDS topic [14:48] lool: you seem to have some sort of misconception about how the test suite works. Please either read the documentation or try the test out yourself for once. [14:48] +1 [14:48] -1. [14:49] Good luck. [14:49] The test suite is updated too often. [14:49] GrueMaster: I have the impression that thousands of tests must be automated in some way and that having a person kick them manually is not the best way to use our time [14:49] GrueMaster: I'm not saying it's simple, just worth taking a look at [14:49] It is automated. To a point. [14:49] GrueMaster: we already have testsuites in checkbox that are updated several times a day! [14:49] IT TAKES LITTLE TIME TO RUN!!!!! [14:50] * ogra shades his ears [14:50] But you might be at a conference or hit by a bus [14:50] * GrueMaster wants desparately to drop this subject. [14:50] ok, in any case, we have 10 minutes left, we can talk about it offline [14:50] Then someone else will need to spend 15 minutes learning how to run them. [14:50] [topic] armel [14:50] * plars suggests [action] gruemaster to avoid busses for a while [14:50] New Topic: armel [14:50] And another 10 minutes posting the data. [14:50] [link] https://bugs.launchpad.net/ubuntu/karmic/+bugs?field.subscriber=ubuntu-armel [14:50] LINK received: https://bugs.launchpad.net/ubuntu/karmic/+bugs?field.subscriber=ubuntu-armel [14:50] * ogra would like to see someone from the team to test Bug 427289 [14:51] Plus an hour a week justifying the ten minutes spent. [14:51] Launchpad bug 427289 in linux-fsl-imx51 "hardware clock not saved if board power is removed on babbage 2.5" [High,Invalid] https://launchpad.net/bugs/427289 [14:51] amitk would like us to test his pselect patches [14:51] the kernel in todays image should support the charging [14:51] but my battery seems to be bad [14:51] ogra: Ok [14:51] will dist upgrade and test [14:51] you need to change for 8-12h, according to amit the clock should keep the time then [14:51] [action] test fix for RTC battery charging (lool) [14:51] ACTION received: test fix for RTC battery charging (lool) [14:51] a voltmeter would also be nice :) [14:52] and reporting some values [14:52] but not mandatory [14:52] CRAP [14:52] https://launchpad.net/ubuntu/+source/openoffice.org/1:3.1.1-2ubuntu3/+build/1264928 [14:52] imx51 FEC driver has issues with NM and in d-i [14:52] doko_: ^ [14:52] ogra: Bug? [14:53] oo.o didnt build and we will likely ship with a broken one this round [14:53] make[2]: *** [bfd.info] Error 1 [14:53] make[2]: Leaving directory `/build/buildd/openoffice.org-3.1.1/binutils-build/bfd/doc' [14:53] Bug 438687 [14:53] Launchpad bug 438687 in network-manager "FEC driver does not set "DRIVER" property in udev which makes network-manager fail" [High,New] https://launchpad.net/bugs/438687 [14:53] doko_: missing bdep? [14:53] lool, he knows the fixy [14:53] ogra: Is this being pushed? [14:53] but we wont make it in time [14:53] Unless we respin thursday morning [14:53] doko is on it afaik [14:54] in any case i propose to not rip out oo.o this time [14:54] lool: it's missing a b-d for binutils, yes. if it helps anything, I'll have packages ready in about 2h/3h on jocote, built from the same source [14:54] ogra: 438687 > looks good [14:54] it took four people and several hours to get that right last time [14:54] if oo.o doesnt build, lets just release note it but leave it in the image [14:54] ogra: It wasn't installable I think [14:54] you have to make changes in to many places [14:54] Ok anything else on armel? [14:55] lool, ouch, indeed [14:55] and it wont be this time either [14:55] sata still doesn't work when running gdm. [14:55] arch all discrepancy [14:55] I have to review the f-k changes from NCommander but will be post beta now [14:55] lool, so it seems we have to rip it out or not respin [14:55] current .1 image looks quite good [14:56] a respin will screw up the livefs build [14:56] Also, the two failing nics of mine have been given to amitk for debugging. [14:56] [topic] specs review [14:56] New Topic: specs review [14:56] GrueMaster, well, the builtin NIC works now [14:56] I have spres here. [14:56] ogra: is cron disabled? [14:56] ogra, I noticed. [14:57] lool, for images ? i think so [14:57] pitti could tell [14:57] ogra: Ok they are [14:57] ogra: then we just need to pass the word not to respin desktop on arnel [14:57] armel [14:57] yeah [14:57] GrueMaster: LSB results for A6 > didn't see them [14:57] I think there is still some stability issues in the USB driver stack. on imx51. [14:57] GrueMaster: So I dropped it from the spec [14:58] sigh. [14:58] 2 min [14:58] I POSTPONED a bunch of items on paulliu's specs [14:58] I just posted them. [14:58] GrueMaster: I updated the spec again [14:58] I'm still waiting to hear when Wubi is fixed ... [14:58] GrueMaster: Just update the spec next time [14:58] StevenK: :-( [14:58] I had waited to go over them with the linux foundation folks so I could understand some major failures. [14:59] StevenK: it affects desktop too, so quite bad [14:59] lool: Absolutely [14:59] GrueMaster: I dont see how that blocks posting them?! [14:59] lool: let ME update my spec, please. [14:59] lool: I think Agostino needs to be nailed down, I'll try and catch him this week [14:59] [link] http://piware.de/workitems/mobile/karmic/report.html [14:59] LINK received: http://piware.de/workitems/mobile/karmic/report.html [14:59] I needed to see if I was doing anything wrong. there was a major test suite that hung. [15:00] turned out to be a kernel bug. [15:00] GrueMaster: Sorry but you're not the only one to be in charge of the spec [15:00] GrueMaster: I happen to have to report on it [15:00] with the help of ogra BTW [15:00] can you at least ask me what's up before blindly making changes? [15:01] * ogra looks up [15:01] GrueMaster: I checked the wiki, the results were not posted, I updated the spec to match [15:01] GrueMaster: I thought that you could still fix it if that was incorrect -- like I just did [15:01] * GrueMaster gives up. [15:01] ogra: Back in your cage! [15:01] GrueMaster: I'm not going to organize to block on you [15:01] * ogra roars [15:01] :) [15:02] plars: I marked the unr checkbox stuff as implemented since the last action was to merge the test in the main branch [15:02] hey, it's 3 minutes past. meeting over, right? [15:03] Ok ; I don't have any other questions on specs; does anybody want to raise anything? [15:03] GrueMaster: We can run a bit over this week; especially since it's beta [15:03] and there's no TB meeting after us [15:03] ian_brasil: Around? [15:03] [topic] AOB [15:03] New Topic: AOB [15:03] yep [15:04] ian_brasil: Did you want to raise MID stuff? [15:04] just an update [15:04] we created the community project..worked on seeds and i have packages upstream libhildon and hildon-desktop..both fail to build ha ha but i will sort that later [15:04] ian_brasil: Did you settle on a set of goals and a codename? [15:05] not the goals yet but we have a codename [15:05] Eh ok [15:05] Liquid remix [15:05] lool: I saw, thanks. They are both marked implemented now [15:05] ian_brasil: I think at this point it's not reasonnable to do anything in karmic, but you guys could prepare stuff in a PPA for lucid [15:06] that is what we thought [15:06] Ok; any other business? [15:06] NCommander: Still with us? :-) [15:07] #endmeeting [15:07] Meeting finished at 09:07. [15:07] Thanks all [15:07] thanks === imlad|away is now known as imlad [15:56] morning [15:56] yo :) === zoopster is now known as jpugh [16:01] o/ [16:01] o/ [16:02] \o [16:02] gday [16:02] o/ [16:02] o/ [16:02] Welcome to the server team meeting... [16:03]