## page was renamed from MobileTeam/Meeting/2010/20100824 <> ||<>|| '''August 24th, 2010, 13:00 UTC''' in '''#ubuntu-meeting'''.<
> = Agenda = == Meeting Links == * Link to previous meeting * [[MobileTeam/Meeting/2010/20100817]] * Link to next meeting * [[MobileTeam/Meeting/2010/20100831]] == Action Items from last meeting == * NCommander to unbreak apport retracer (c/o) == Special Items == == Standing Items == * http://people.canonical.com/~pitti/workitems/maverick/canonical-mobile.html * http://people.canonical.com/~pitti/workitems/maverick/canonical-mobile-ubuntu-10.10-beta.html * http://qa.ubuntu.com/reports/team-assigned/canonical-mobile-assigned-bug-tasks.html * Kernel Status (cooloney, mpoirier, lag) * QA Status (GrueMaster) * Bugs needing attention before A3: * Bug #605972 - Need to set hostname to ubuntu during first boot. * Bug #600359 - ureadahead generating oom messages during boot. * Bug #605831 - [omap3] Resolution should be taken from /proc/cmdline if provided * ARM Porting/FTBFS status ([[MichaelCasadevall|NCommander]], [[DavidSugar|dyfet]]) * ARM Image Status ([[OliverGrawert|ogra]], [[MichaelCasadevall|NCommander]]) * Any Other Business * Kubuntu Mobile - Liquid update = Meeting Outcome = {{{ [14:05] #startmeeting [14:05] Meeting started at 08:05. The chair is NCommander. [14:05] Commands Available: [TOPIC], [IDEA], [ACTION], [AGREED], [LINK], [VOTE] [14:05] * NCommander grumbles [14:05] sorry, alarm failed to go off [14:05] oh, i thought you had handed over to persia [14:05] no grumbling :) [14:05] G'day all [14:05] ogra: I took it back because my normal dinner plans were canceled [14:05] ah [14:05] [link] https://wiki.ubuntu.com/MobileTeam/Meeting/2010/20100824 [14:06] LINK received: https://wiki.ubuntu.com/MobileTeam/Meeting/2010/20100824 [14:06] no meow-burgers today ? [14:06] * NCommander throws his cat at ogra's head [14:06] heh [14:06] [topic] Action Item Review [14:06] New Topic: Action Item Review [14:06] [topic] NCommander to unbreak apport retracer (c/o) [14:06] New Topic: NCommander to unbreak apport retracer (c/o) [14:06] * NCommander coughs [14:07] This is poriving to be a damn pain because I can't rebuild the chroots on the porter box [14:07] does it still make sense to re-enable it ? we have beta soon :) [14:07] and my bandwidth is ... limited [14:07] so c/o [14:07] [topic] Standing Items [14:07] New Topic: Standing Items [14:07] [link] http://people.canonical.com/~pitti/workitems/maverick/canonical-mobile.html [14:07] LINK received: http://people.canonical.com/~pitti/workitems/maverick/canonical-mobile.html [14:08] [link] http://people.canonical.com/~pitti/workitems/maverick/canonical-mobile-ubuntu-10.10-beta.html [14:08] LINK received: http://people.canonical.com/~pitti/workitems/maverick/canonical-mobile-ubuntu-10.10-beta.html [14:08] both look just horrid [14:08] please please please guys, get your items done or postponed === serfus_ is now known as serfus [14:08] I should have my 1 item done today. [14:09] \o/ [14:09] NCommander, how about yours ? [14:09] there's been some VERY slow work on subarch, but I have to add additional work items because cjwatson wants me to review with d-boot before we implement in Ubuntu which is slowing things down [14:09] Have new warning message about usage of generic fallback code approved and implemented [1 day] [14:09] so I can strrike two items, but I have to add two more in there place :-/ [14:09] that doesnt look like it should take much time [14:09] hrm, strike one [14:10] ogra: it won't, but I've had to rebase some of my code against sid, and spin new patches to submit to d-boot for review [14:10] Building a sid chroot on an internet connection that acan make dialup look fast is painful. [14:10] erm, wasnt that in flash-kernel ? [14:10] just put it into our tree [14:11] ogra: I need an API change in libd-i [14:11] for the message ? [14:11] cjwatson nixed any changes without discucsion from Debian. [14:11] ogra: for the code that triggers the message [14:11] oh, ok, i thought it was a message inside flash-kernel [14:11] ogra: thats just a warning that shows up on STDERR, we need one that pops up during install as well [14:12] NCommander, btw, asac and linaro have massive probs since your code was added to flash-kernel [14:12] it now runs in any case, even if you are in a build chroot [14:12] ogra: have they filed a bug? :-) [14:12] no, they are working on fixes [14:12] ogra: strikely speaking, that's the correct behavior [14:12] ha ! [14:12] *strictly [14:12] they just have [14:12] Have new warning message about usage of generic fallback code approved and implemented [1 day] [14:12] ugh [14:13] yeah, I should change the time estimate [14:13] https://bugs.launchpad.net/ubuntu/+source/initramfs-tools/+bug/623375 [14:13] Launchpad bug 623375 in initramfs-tools (Ubuntu) "Skipping the bootloader installation when creating rootfs or installation media" [Undecided,New] [14:13] probably looking a full week out [14:13] no, that was mis-pasted :) [14:13] * NCommander looks [14:13] f-k wasnt executed inside chroots before your change eneterd it [14:13] well, it was [14:13] f-k just exitted silently :-) [14:14] if you have a chroot on a properly supported arch, it works as expected. The problem is our buildds aren't quite supported [14:14] [action] NCommander to discuss with linaro and asac on improved-generic-subarch-support [14:14] ACTION received: NCommander to discuss with linaro and asac on improved-generic-subarch-support [14:14] it doesnt work "as expected" [14:15] since it tries to access the NAND of the buildd [14:15] ogra: well .... er, crap [14:15] yeah [14:15] that's a bug [14:15] (if your buildd is omap3) [14:15] It shouldn't try and do that on a generic fallback install [14:15] yeah [14:15] oops [14:16] well, lool and asac are looking into it atm [14:16] ogra: I'll work with asac and linaro to get it sorted. My Beagle is in Texas, so someone else will have to test [14:16] we're simply not affected in livecd-rootfs because update.initramfs is diverted during the build [14:16] * NCommander avoided breaking the world by sheer luck :-/ [14:16] anyway, lets not hold up the meeting [14:16] yeah [14:17] [topic] Kernel Status (cooloney, mpoirier, lag) [14:17] New Topic: Kernel Status (cooloney, mpoirier, lag) [14:17] ^- ericm [14:17] hey [14:17] * Marvel (mvl-dove) [14:17] * FIXED : ABI bumping and module checking errors after the branch sync'ed with LSP 5.3.2 [14:17] * MISC : Discussion of moving forward mvl-dove branch into Maverick [14:17] * Freescale (fsl-imx51) [14:17] * Nothing new this week [14:17] * Texas Instruments (ti-omap) [14:17] * MISC : lag's Panda is broken - awaiting replacement from davidm [14:17] * MISC : Still waiting for TI to release L24.9 to us [14:17] * ogra wasnt done with specs [14:17] * FIXED : B591941 work around applied - working on correct fix [14:17] * FIXED : B608266 GPMC was not initialized before accessing NAND [14:17] sigh [14:17] * FIXED : B608279 now able to read the EDID from userspace [14:17] * RESUMED : B563650 problem can be reproduced on latest 2.6.35-18 with non-UNR file system [14:17] * ON GOING : B605488 still unable to reproduce - will test on full-build when new HW arrives [14:17] .. [14:18] https://bugs.launchpad.net/bugs/608266 - waiting for mobile team to test [14:18] ogra: er, whoops. I took "not hold up" as "move on" :-) [14:18] Launchpad bug 608266 in linux (Ubuntu Maverick) "[regression] no more /dev/mtdblock devices on omap3 in maverick" [Medium,Fix released] [14:18] * NCommander will go back after kernel discussion [14:18] https://bugs.launchpad.net/bugs/608279 - Fix committed [14:18] Launchpad bug 608279 in linux (Ubuntu) "i2c support for edid on armel" [Undecided,Fix committed] [14:18] soryy, i wasnt clear [14:18] * ogra hugs mpoirier [14:18] ouf ... [14:19] .. [14:19] ogra: mpoirier: did you see my email? [14:19] * NCommander hugs mpoirier [14:19] yes, did not read yet. [14:19] about edid parsing at u-boot [14:19] mpoirier: the code is basically from omap 4, so in the future it should be easy to port to omap 3 [14:19] duly noted. [14:19] very interesting. [14:20] rsalveti, did dyfet work on that at all (after all its his spec) [14:20] there might be a little bit of time before aI get a new panda. [14:20] I'll look at it. [14:20] ogra: don't know, he's still sleeping it seems dyfet_sleeping :-) [14:20] hrm [14:20] * NCommander head thunks [14:20] ogra: and he was fixing other bugs [14:21] does anyone want to call dyfet? [14:21] rsalveti, well ... [14:21] could be interesting for him, that's why I added him at the email [14:21] I will test 608266 today and report back. [14:22] I'm getting i/o errors at mtd now [14:22] but should be normal, probably missing a correct partition table [14:22] [ 4.979064] end_request: I/O error, dev mtdblock0, sector 128 [14:22] the table is hardcoded in the kernel iirc [14:22] if you write to an mtdblock you need to zero it first [14:22] I know, but it's probably empty or never created [14:22] like uImage and rootfs partition [14:23] what does /proc/mtd report ? [14:23] if you see partitions there its fine [14:23] ogra: shows fine, because the partition table is in the kernel [14:23] right [14:23] probably it's trying to mount or something like that [14:23] need to check [14:24] yeah [14:24] let's also wait GrueMaster to test [14:25] right [14:25] lag: any news regarding TI tree? [14:25] and file bugs so mpoirier doesnt get bored :) [14:25] sure :-) [14:25] NCommander, so since we touched the EDID spec, there is no need to return to specs [14:25] rsalveti: No, nothing :( [14:25] (and since dyfet_sleeping isnt here) [14:26] rsalveti: We're waiting on sebjan [14:26] ogra: np [14:26] lag: hm, ok [14:26] * NCommander still thinks someone should call dyfet_sleeping [14:26] thanks, anyway :-) [14:26] can I move on? [14:26] rsalveti: There's nothing I can do ... [14:27] NCommander, if the kernel team is happy [14:27] rsajdok: lag [14:27] Doh! [14:27] heh [14:27] lag: np :-) [14:27] rsalveti: lag [14:27] * ogra steals lag's tab key [14:27] ogra: It was you who told me I could use tab [14:27] use tab enough times ;) [14:28] heh [14:28] three tabs and you get what you need ;) [14:28] oh, i started to see a new bug [14:29] * NCommander files a bug on ogra and lag's use of tabs [14:29] seems neither omap nor omap4 reboot anymore if reboot is issued [14:29] at least in initramfs [14:29] ouch [14:29] ogra: ugh. === Ursinha-afk is now known as Ursinha [14:29] i'm not sure yet if thats the kernel or the reboot command from busybox [14:29] ogra: does it work in the userland? [14:29] NCommander, yes [14:30] * NCommander can't see it being a utiluty problem, busybox generally just does a syscall to reboot() [14:30] but not in initramfs [14:30] ooh, ugh [14:30] That's nasty [14:30] and initramfs uses a different reboot [14:30] that's fugly [14:30] ogra: have fun debugging it :-) [14:30] :P [14:31] but did we change busybox? [14:31] thats the issue, i dont think we did [14:31] and we had at least one kernel upload per subarch since i saw it working [14:32] * NCommander should check if it works on dove tomorrow [14:32] so i'm tending towards the kernel has changed [14:32] and busybox calls something it stopped understanding [14:33] something like that [14:33] ogra: Blames the kernel - shock! [14:33] ogra: busybox (1:1.15.3-1ubuntu2) 19 Aug 2010 [14:33] rsalveti, oh, thanks :) [14:33] yeah, that might be it [14:33] doesn't seems related, but anyway [14:34] "* armel seems to build fine without -marm nowadays, so remove it." [14:34] well [14:34] (from the changelog) [14:34] that seems like a bug :-) [14:34] hm, true, was looking at previous upload message [14:35] so it builds ... [14:35] just doesnt run properly :P [14:35] but we can pretend it's a kernel issue so we can keep lag and mpoirier busy [14:35] hehe [14:35] ogra: haha :-) [14:35] Are we still in meeting mode, or are we deep diving into bugs & possible workarounds? [14:35] * lag has enough to do [14:36] GrueMaster, meeting indeed [14:36] can I move on to QA? [14:36] NCommander, move ! [14:36] goes [14:36] [topic] QA Status (GrueMaster) [14:36] New Topic: QA Status (GrueMaster) [14:36] * NCommander does as he is commanded [14:36] Attended QA CoP sprint last week. Lots of daily test tracking ideas were shared. Will review and implement something hopefully before Beta. Need to explore more in detail. [14:36] Received # new platforms in the last week. Working to rearrange office to accomodate by EOD. [14:36] Filed bugs against Banshee and F-Spot. Possibly mono specific, but I don't know. Need test suite that can help narrow down mono specific bugs. [14:36] No new images since Aug 19. [14:36] GrueMaster, yep [14:37] GrueMaster: did you get dove stuff yet? [14:37] ^- davidm too [14:37] GrueMaster, waiting for dyfet_sleeping to fix the FTBFS of telepathy-glib [14:37] I have received a Dove A0 (part of total # new systems received) [14:37] ogra: Is that why Empathy is not installing on the new image? [14:38] yep [14:38] ok [14:38] GrueMaster: good to know. We were still having some issues getting confirmation where they were [14:38] and it is why i asked him about 6 weeks ago to look at it :P [14:38] ogra: just rebuild from archive, it is no longer broken [14:38] davidm: did you get the two boards sent to you? === dyfet_sleeping is now known as dyfet_waking [14:38] and there you go [14:38] * NCommander kicks the retry button [14:38] dyfet_waking, it is, i gave it back right after you pinged me today [14:38] and still fails with the same issues [14:38] * NCommander blinks [14:38] ACK [14:38] * NCommander goes into shock from ia64 and sparc's deaths [14:39] ogra: what version did you build? [14:39] dyfet_waking, the one that failed [14:39] it's been updated in the archive [14:40] 0.11.13 [14:40] dyfet_waking: and what's what failed [14:40] NCommander: Looks like the topic moved to FTBFS. [14:40] NCommander, I have all hardware now [14:40] dyfet_waking, 0.11.13-1ubuntu1 is what fails [14:40] davidm: thanks. I'll inform the powers that be that everything was recieved [14:40] dyfet_waking: are you building the source package? [14:40] Yes [14:40] or are you building the upstream tarball?! [14:41] dyfet_waking: how are you building it? [14:41] dyfet_waking, and the former debian sync also failed already since weeks [14:41] libtelepathy-glib-dev_0.11.13-1ubuntu1_armel.deb [14:41] libtelepathy-glib-dev_0.11.8-1_armel.deb [14:41] libtelepathy-glib-doc_0.11.13-1ubuntu1_all.deb [14:41] libtelepathy-glib-doc_0.11.8-1_all.deb [14:41] libtelepathy-glib0-dbg_0.11.13-1ubuntu1_armel.deb [14:41] libtelepathy-glib0-dbg_0.11.8-1_armel.deb [14:42] libtelepathy-glib0_0.11.13-1ubuntu1_armel.deb [14:42] libtelepathy-glib0_0.11.8-1_armel.deb [14:42] 09:42:02 < dyfet_waking> libtelepathy-glib-dev_0.11.8-1_armel.deb [14:42] 09:42:02 < dyfet_waking> libtelepathy-glib-doc_0.11.13-1ubuntu1_all.deb [14:42] 09:42:02 < dyfet_waking> libtelepathy-glib-doc_0.11.8-1_all.deb [14:42] argh [14:42] guys ! [14:42] use a pastebin [14:42] that was an accident [14:42] sorry :) [14:42] irssi likes to paste when I middleclick [14:42] that was what I got to build this morning on arm on maverick [14:42] https://edge.launchpad.net/ubuntu/+source/telepathy-glib/0.11.13-1ubuntu1 [14:42] dyfet_waking: what platform are you building on? [14:43] beagle...it was slow :) [14:43] given back twice today, still failing on the same tests [14:43] lets discuss this after then... [14:43] dyfet_waking: YOu should use the dove I have available. Much faster. [14:44] [action] dyfet and ogra to discuss telepathy-glib and report back [14:44] ACTION received: dyfet and ogra to discuss telepathy-glib and report back [14:44] anyway, beta is next thu, [14:44] [topic] ARM Porting/FTBFS status (NCommander, dyfet) [14:44] New Topic: ARM Porting/FTBFS status (NCommander, dyfet) [14:44] * rsalveti is using his panda with usb disk, and it's fast [14:44] it needs to be fixed before friday [14:44] I have nothing to report, I haven't had time to work on FTBFS [14:44] how is kde going dyfet_waking ? [14:44] I was going to work from what Michael gave me after waking :) === dyfet_waking is now known as dyfet [14:45] well, please priorize telepathy first [14:45] dyfet_waking: please focus on telepathy [14:45] d'oh [14:45] heh [14:46] anyway [14:46] [topic] ARM Image Status (ogra, NCommander) [14:46] New Topic: ARM Image Status (ogra, NCommander) [14:46] bad bad bad [14:46] very bad [14:46] as you could see above already [14:46] * NCommander has some good though [14:46] but ... [14:46] oem-config has a fix thats pending upload [14:46] wooo [14:46] cool [14:46] if we could fix telepathy-glib, we can get our images going again [14:46] (i need to coordinate with cjwatson about uploading it soon) [14:46] yeah [14:47] well, there is also a filesystem corruption i have seen [14:47] So, I have some news [14:47] We've started building dove images again [14:47] with the preinstalled images [14:47] currently only daily-live/ubuntu-netbook [14:47] lool even sees it *after* jasper has run [14:47] I'll be re-enabling alternates sometimes this week for normal ubuntu, as building ubuntu-netbook alternates really goes south [14:47] ogra: coordinate with ev [14:47] cjwatson, will do [14:48] speaking of ubuntu-netbook going south [14:48] NCommander: Looks like alt-inst images are building too. [14:48] NCommander, we decided to not support alternates anymore [14:48] GrueMaster: no, that was actually a separate issue [14:48] NCommander, can you instead enable server ? [14:48] oh. [14:48] ogra: cjwatson: can i delete these (http://cdimage.ubuntu.com/ubuntu-netbook/ports/daily/) out of www or do we need to do something special [14:48] NCommander, just rm it [14:48] thanks [14:49] * NCommander didn't know if they were safe to RM directly, or if I would break image building, and get fired [14:49] I think I already said to rm -rf it [14:49] indeed, thats why i suggested it :P [14:49] cjwatson: oh, I must have missed that. Sorry [14:49] no problems [14:49] [action] NCommander to clean out the stale dove alternates [14:49] ACTION received: NCommander to clean out the stale dove alternates [14:49] [action] NCommander to smoke test ubuntu-server dove alternates [14:49] ACTION received: NCommander to smoke test ubuntu-server dove alternates [14:50] NCommander, there are other dirs too that could need cleaning [14:50] seems someone tried to build daily-preinstalled in / [14:50] ogra: indeed, but my link to antimony is so damn slow that I don't want to try to cleanout the www folder until I'm state side again [14:50] the lag is so bad that typing something takes about 2-3 secends per char [14:51] yeah, just have a look around, there is other mess to clean up if you're at it :) [14:51] [action] NCommander to flush antimony's expired images [14:51] ACTION received: NCommander to flush antimony's expired images [14:51] :-) [14:51] ... and stale empty dirs :) [14:51] ogra: IT SHALL BE DONE! [14:51] haha [14:52] [topic] ABO [14:52] New Topic: ABO [14:52] stop shouting, move [14:52] bah [14:52] * NCommander wins [14:52] youre to fast :P [14:52] ian_brasil: ^ [14:52] rbelem patched startkde (and this was accepted) to start the plasma-mobile desktop and kubuntu mobile default settings package just needs some marketing text changing now - so we are ready to start building images..well, we still need to change the text in the seed to actually call this new default settings and we need a small change to plasma board to add this to the mobile systray but these are very minor things.. [14:52] ian_brasil: what archs are you planning to build for? [14:53] armel and i386 [14:54] ian_brasil: I ask that you only target armel+omap4 and armel+omap for now. We're somewhat strapped for buildd power. Once we get all the kinks worked out o dove images, we can add it [14:54] well, i wonder if you really want to attempt to run it on omap3 [14:54] (omap4 is on a dedicated builder, omap and dove are sharing one :-/) [14:54] NCommander: Ok..we can do that [14:54] whats the ram requirements ? [14:54] beagle only has 256M [14:55] ogra: I thought there were some omap3 platforms with 512M [14:55] there is the beagle XM but its still not on the market [14:55] ogra: I thought it was out [14:55] essentially all omap3 512M platforms currently are unavailable [14:55] ogra: ugh. [14:55] apart from the touchbook for which we dont have kernel patches [14:56] ian_brasil: if you want to trade omap for dove (which is also not out on the market), that's fine :-) [14:56] the other 512M hw you saw was all prototypes or not on the market yet [14:56] ogra: I never saw any 512M hw directly [14:56] NCommander: thx [14:56] NCommander, you were in prague, no ? [14:56] i'm sure i showed you some :) [14:57] ogra: I was working on omap4 [14:57] :-) [14:57] anyway. I think that's everything [14:57] anyone got anything else [14:57] nope [14:57] #endmeeting [14:57] Meeting finished at 08:57. }}} == Action Items == * NCommander to discuss with linaro and asac on improved-generic-subarch-support * NCommander to clean out the stale dove alternates * NCommander to smoke test ubuntu-server dove alternates * NCommander to flush antimony's expired images == Minutes == * Dove enablement continues * Images very broken due to archive churn, and dbus/oem-configu issues * FTBFS list for KDE in bad shape = Weekly Reports = === Michael Casadevall === * Dove enablement in China continues === David Sugar === * telepathy-glib eventually became buildable with introspection updates. * lots of archive churn, focused on bugs. * blocking edid kernel issues resolved.