20100105

January 5th, 2010, 13:00 UTC in #ubuntu-meeting.

Agenda

Action Items from December 22nd, 2009

  • cooloney to ask ericm about 2.6.32 patches for marvell dove
  • NCommander to email Marvell about the date of the next expected kernel drop
  • asac, ogra, persia to make sure .32 backporting for imx51 kernels is documented somewhere (c/o)

Current Items

Standing Items

Action Items

  • NCommander to invite relevant kernel people to the next meeting.
  • asac, ogra, persia to make sure .32 backporting for imx51 kernels is documented somewhere (c/o).
  • ogra, asac, NCommander to blog on planet about ARM work.
  • Work item review
    • ALL, check work items due for Alpha-2 and either complete, postpone or push to Alpha-3.
    • asac, JamieBennett to discuss MIR stalling for 2D launcher.

    • asac to report on Chromiums licensing issue progress.

Minutes

  • asac, ogra, and persia agreed to meet to discuss getting the non arch specific features of the current kernel backported to the .31 kernel if it was deemed worthwhile.
  • ericm got a kernel drop from Marvell and is currently working on re-basing it. Initial indications from ericm are that the new re-based kernel could be read this week. ericm pointed out that the Marvell branch is held at http://kernel.ubuntu.com/git?p=marvell/dove-kernel/.git;a=shortlog;h=refs/heads/marvell-dove-2.6.32.2

  • There is currently a backlog of team work items for Alpha-2. asac urged the team to check specs, try to complete or push them back to Alpha-3/postpone them where necessary. ogra pointed out that we still have until Monday to close work items and that at least his would be in better shape then.
  • NCommander noted that we should have working alternates for Alpha2, although at the moment the dove variant is broken during kernel installation. There was some discussion on whether we should be testing alternates or not due to a lack of man power. The conclusion was that we would produce the images but not officially support them.
  • asac asked what the status of the debconf-communicate slowness in Casper was. JamieBennett said that there was multiple solutions but one stood out as the best.

  • asac asked when all-pairs testing could be done. plars replied that the test pairs are generated and documented on the wiki but detail for them needs to be written on the qa testplan wiki. He went on to explain that the pairs were generated programatically and had to be translated manually into corresponding items.
  • ogra indicated that the uboot spec was on target for being completed by the weekend.
  • JamieBennett raised concern that some MIR's for the 2D launcher were stalled. asac agreed to discuss them after the meeting.

  • asac outlined the status of his spec's, thumb work items are review based ones so will be done at the end, chromium is on track after some hard work and suspend-resume testing was mostly moved to Alpha-3. Discussion petered off on to chromiums licensing issues with asac indicating that he was working on it.
  • StevenK said that he was planning to break armel images by getting them built as the UNE (Ubuntu Netbook Edition) this week. asac raised concern that we should keep the desktop image until at least Alpha-2 release so we have a working image. StevenK agreed. Discussion continued on what will be the final image for ARM devices, UNE with the 2D launcher was decided to be the main focus.
  • asac mentioned that he had gone through the FTBFS (Fail to build package list) and assigned them to either NCommander or dyfet for fixing. NCommander said he had uploaded a fix for ocaml and was thanked for it. Discussion continued on where the focus should be on the FTBFS list, it was concluded that main should be FTB free if possible and universe should be too but focus was on main if a conflict arose.
  • A more community engaged approach to testing was discussed. As the community generally don't have the hardware to test many of our issues it is quite hard. More blogging and general publicising of what we do is needed. ogra pointed to a post at his blog, http://ograblog.wordpress.com/2009/07/18/juggling-your-arms-in-karmic-and-no-more-excuses/, that indicates some engagement is happening but asac wanted more. ogra, asac and NCommander took the action to blog about ARM work.

Weekly Reports (week before the break for Christmas)

Jamie Bennett (JamieBennett)

This week

  • Continued on with the Casper investigations. Talked with interested parties to identify the area's of interest. Possible low hanging fruit is disable_guest_account which could be sped up by about 6.5 seconds.

  • Tracked down the major cause of slowness for Casper, that of debconf-communicate. Straced, timestamped and looked at this with the conclusion that loading the templates.dat file takes around 4 seconds for every debconf-communicate call. Massive win to get this sped up so identified ways of tackling it. Needs implementing.
  • Documented possible fixes for Casper slowness here.

  • Worked on, along with help from the community, pushing build dependencies for netbook-launcher-efl.
  • Little work on the REVU version of netbook-launcher-efl to whip it into shape after the excellent work of mterry.
  • Pushed to get Canola packaged.
  • Initial test of Canola proved that it needs some work before it can be the default media player for Ubuntu ARM. Some bugs submitted.
  • Monthly team report done which can be viewed here.

  • Image testing, internal stuff, tending to sick kids, all good fun.

Future

  • Come up with a working implementation to fix the debconf-communicate issues.
  • Gather monthly information and publish monthly team report.
  • Keep pushing and monitoring netbook-launcher-efl dependencies.
  • If I get time take a look at the Moblin media player.
  • Vacation Tuesday till new year (although I'm sure I'll have some cheeky post-turkey hours to bash some blueprints).

Steve Kowalik

  • Propose a patch for Launchpad for a change to germinate, get it merged and landed.
  • Upload netbook-meta, pulling from the new netbook.lucid branch. Do so multiple times.
  • Merge changes to tasksel, adding netbook, and removed unr and mobile.
  • Remove mobile-meta, and discuss with the LOSAs how to remove its seed branch.
  • Test my changes to livecd-rootfs, uploading it, and organising installation on the buildds.
  • Discover issues that relate to changes with desktop versus une, and fix the seeds. Three times.
  • Push out my changes for the rename to the public branch of cdimage.
  • Merge the changes into the deployed cdimage code.
  • Also fix debian-cd and merge it in too.
  • After an hour or so of poking around, discover why my cron.germinate changes caused it to bail and fix two scripts on rookery. Fix component-overrides generation while I was there.
  • Thursday archive admin duties.

Tobin Davis

  • Continued work on porting LSB tests to arm. Currently focusing on T2C test suites.

Paul Larson

  • Was able to begin testing images on imx51 with successful images beginning on 20091213
  • Collected additional requested debug data for bug #494787 - firefox crashed with SIGSEGV
  • Wrote a script for the bringup testing that uses the bootchart tarball to gather the amount of time it took for boot
  • rebuilt dove kernel with patch from bug #494831
  • created a new install image with the kernel for bug #494831 to test the fix - fix worked, and I was able to create a working dove image to test with
  • made a qemu-arm-static build environment to mess with, see if I can gain any efficiency with it, did some performance comparisons, not impressed with the result
  • Tested a new build of Chromium on ARM, still seeing some SIGILL issues, but able to get further
  • Started working on install testing blueprint

Alexander Sack

  • Vacation.

Oliver Grawert

  • Vacation.

Meeting Log

(13:00:17) NCommander: #startmeeting
(13:00:17) ***asac waves
(13:00:18) MootBot: Meeting started at 07:00. The chair is NCommander.
(13:00:18) MootBot: Commands Available: [TOPIC], [IDEA], [ACTION], [AGREED], [LINK], [VOTE]
(13:00:35) pak33m [n=pak33m@ubuntu/member/pak33m] entered the room.
(13:00:42) NCommander: morning all
(13:00:44) davidm: G'day
(13:00:48) dyfet: need coffee
(13:00:52) GrueMaster: Zzzz
(13:01:02) JamieBennett: Happy New Year !
(13:01:03) moustafa [n=moustafa@modemcable227.204-131-66.mc.videotron.ca] entered the room.
(13:01:08) NCommander: [link] https://wiki.ubuntu.com/MobileTeam/Meeting/2009/20100105
(13:01:09) MootBot: LINK received:  https://wiki.ubuntu.com/MobileTeam/Meeting/2009/20100105
(13:01:12) moustafa left the room.
(13:01:14) ***ogra waves (with a head that doesnt feel like it fits through the door)
(13:01:23) NCommander: [link] http://piware.de/workitems/mobile/lucid-alpha2/report.html
(13:01:24) MootBot: LINK received:  http://piware.de/workitems/mobile/lucid-alpha2/report.html
(13:02:23) NCommander: [topic] Action Item Review
(13:02:23) ogra: i guess we can ignore the ARs this week :)
(13:02:24) MootBot: New Topic:  Action Item Review
(13:03:05) NCommander: [topic] cooloney to ask ericm about 2.6.32 patches for marvell dove
(13:03:06) MootBot: New Topic:  cooloney to ask ericm about 2.6.32 patches for marvell dove
(13:03:34) cooloney_: i think ericm_ just got the .32 patches from marvell
(13:03:57) ogra: NCommander, btw. you should include one of the kernel guys in the meeting ping in the future
(13:04:11) NCommander: ogra, ah, will do
(13:04:12) asac: both please
(13:04:45) NCommander: [topic] NCommander to email Marvell about the date of the next expected kernel drop
(13:04:46) MootBot: New Topic:  NCommander to email Marvell about the date of the next expected kernel drop
(13:05:02) b1ackcr0w left the room ("Ex-Chat").
(13:05:06) NCommander: c/o on this one, although if ericm_ got the 32 patches already, then not needed.
(13:05:26) NCommander: [topic] asac, ogra, persia to make sure .32 backporting for imx51 kernels is documented somewhere (c/o)
(13:05:27) MootBot: New Topic:  asac, ogra, persia to make sure .32 backporting for imx51 kernels is documented somewhere (c/o)
(13:05:38) persia: We failed on this one again.
(13:05:40) asac: what was that about? (probably needs to be carried forward)
(13:05:43) ***ogra didnt do anything here 
(13:06:05) ogra: asac, that was about getting the non arch specific features ported to the .31 kernel
(13:06:07) asac: is that about user space?
(13:06:08) persia: asac: ogra: let's get together about 9:00 UTC tomorrow and do this.
(13:06:15) ogra: so we have all distro features available
(13:06:25) ogra: asac, nope
(13:06:29) asac: hmm. shouldnt we get cooloney_ on board too then?
(13:06:37) ogra: about aufs, apparmor and friends
(13:06:38) persia: We were supposed to give cooloney_ a report :)
(13:06:44) anmar [n=anmar@208.49.128.241] entered the room.
(13:06:45) ogra: right
(13:06:50) ***ericm_ waves to all
(13:06:55) persia: So yeah, tomorrow work for you two?
(13:07:08) NCommander: hey ericm_, did we get a 32 drop from Marvell?
(13:07:10) dmart [n=chatzill@fw-tnat.cambridge.arm.com] entered the room.
(13:07:16) ogra: persia, 1h later ?
(13:07:21) persia: OK.
(13:07:24) ericm_: NCommander, yes - I'm working on the rebase
(13:07:24) ogra: great
(13:07:30) asac: ok
(13:07:32) asac: we can try
(13:07:42) ericm_: hopefully to get a working kernel in these days
(13:08:02) ogra: would be good to have both new kernels there for A2
(13:08:10) NCommander: ericm_, where are the base Marvell branches? I think we may have another thumb mode kernel bug (I sent you an email about it)
(13:08:17) ogra: since they will need heavy testing
(13:08:42) ericm_: NCommander, yes I've got that email - the branch is on kernel.ubuntu.com
(13:08:49) ericm_: let me find the exact URL
(13:08:59) asac: ogra: persia: sent an invite for tomorrow (in case you care)
(13:09:08) ogra: thanks
(13:09:21) ericm_: http://kernel.ubuntu.com/git?p=marvell/dove-kernel/.git;a=shortlog;h=refs/heads/marvell-dove-2.6.32.2
(13:09:22) MootBot: LINK received:  http://kernel.ubuntu.com/git?p=marvell/dove-kernel/.git;a=shortlog;h=refs/heads/marvell-dove-2.6.32.2
(13:09:22) asac: yes. from whati understood cooloney_ will try to get the fsl kernel done this week
(13:09:37) cooloney_: asac: yeah,
(13:09:40) cooloney_: i am working on this
(13:09:41) NCommander: ericm_, awesome. I'll talk to you more about the issues outside this meeting.
(13:09:44) ericm_: asac, hopefully me too
(13:09:52) ericm_: NCommander, no problem
(13:10:13) NCommander: [topic] Current Items for Lucid Alpha 2
(13:10:14) MootBot: New Topic:  Current Items for Lucid Alpha 2
(13:10:26) NCommander: [link] http://piware.de/workitems/mobile/lucid-alpha2/report.html
(13:10:27) MootBot: LINK received:  http://piware.de/workitems/mobile/lucid-alpha2/report.html
(13:10:32) asac: we have a big backlog ...
(13:10:42) NCommander: asac, Christmas holidays unfortunately do that :-/
(13:10:48) asac: yeah
(13:10:55) asac: so part of the backlog is definitly me
(13:11:04) ogra: me too
(13:11:09) NCommander: Any objection if we just run through each Alpha 2 spec individually?
(13:11:16) asac: thats the nature of the items i have ... most are items that are scheduled to happen at alpha-2
(13:11:20) asac: like the thumb spec etc.
(13:11:34) njpatel1 left the room (quit: Read error: 60 (Operation timed out)).
(13:11:47) asac: but ... would be great if everyone could check his spec and see if there are work items that either are DONE or can be POSTPONED
(13:11:51) ogra: but we have until monday ...
(13:12:09) ogra: so a good bunch should be affordable
(13:12:44) NCommander: We should have working alternates for Alpha2, although the dove alternate currently seems to be semi-foobar during kernel installation
(13:13:06) ogra: is the test requirement fulfilled already `
(13:13:08) njpatel [n=njpatel@5ac614b6.bb.sky.com] entered the room.
(13:13:16) NCommander: There was an image on antimony that was preventing alternate images for ports spinning. This has been resolved
(13:13:28) NCommander: ogra, which test requirements? ISO testing?
(13:13:41) ogra: iirc that is supposed to have clearified the community testing first
(13:14:23) ogra: at least thats what i was told at UDS
(13:14:49) ogra: i.e. we will only do it if we have enough community testing since we dont have the resources in the team
(13:15:06) asac: i would suggest to do it... and see how well the testing goes
(13:15:15) ***ogra definately doesnt have additional 6h per image to invest 
(13:15:20) asac: of coures not officially support it until we have the resources
(13:15:31) ogra: we dont have them *now*
(13:15:45) ogra: and the desktop images will suffer through pulling time off them
(13:15:59) ogra: s/them/their testing/
(13:16:09) ogra: which we definately have to do
(13:16:14) asac: but from the technical side, what is missing?
(13:16:22) ogra: alternate testing takes at least 3x as much time
(13:16:23) asac: NCommander: ?
(13:16:29) pak33m left the room (quit: "Leaving.").
(13:16:40) NCommander: [topic] Standing Items
(13:16:41) MootBot: New Topic:  Standing Items
(13:16:49) NCommander: [link] http://piware.de/workitems/mobile/lucid/report.html
(13:16:51) MootBot: LINK received:  http://piware.de/workitems/mobile/lucid/report.html
(13:16:56) NCommander: [topic] ARM Image Status
(13:16:57) MootBot: New Topic:  ARM Image Status
(13:17:03) ogra: ??
(13:17:05) asac: NCommander: whats going on?
(13:17:15) NCommander: oh
(13:17:17) NCommander: bugger
(13:17:21) ***NCommander goes to drink more coffee
(13:17:38) NCommander: asac, on the technical side, not too much. the d-cd script for dove have been fixed, so now our alternates boot
(13:17:45) persia: NCommander already talked about the alternate status.
(13:17:52) asac: ok
(13:18:01) ogra: NCommander, what about the ppp package issues ?
(13:18:08) asac: we fixed that a month ago ;)
(13:18:13) asac: or two?
(13:18:15) persia: Live images have been mostly built, although today's were hit with hplib version skew.
(13:18:26) NCommander: ogra, the fix was deploy to antimony just before Christmas break
(13:18:33) ogra: asac, and the new packages are backported and installed on the builder ?
(13:18:37) asac: yes
(13:18:38) ogra: NCommander, ah, thanks
(13:18:39) StevenK: persia: Was that just a delay with NEW or something else?
(13:18:50) asac: ok ... so to finish this alternate image topic.
(13:18:54) ogra: i know when you fixed it ... just didnt know when it went into hardy
(13:19:00) persia: I thought it was just a build-time delay between arch:all and arch:any
(13:19:06) fader|away is now known as fader_
(13:19:07) NCommander: ogra, its in hardy-cat, I don't believe we did an SRU for it
(13:19:08) asac: can we agree that we have all the bits in place ... and that we can do manual image runs to ensure that they work?
(13:19:21) NCommander: asac, +1
(13:19:22) asac: but we dont enable them until we have a) more builders ... and b) a community testing plan?
(13:19:37) ogra: asac, we can ... and i'd be happy to test them, juct not during alpha testing in general
(13:19:44) NCommander: asac, the images are built on antimony and not on an armel box, so an alternate rebuild is trivial
(13:19:57) ogra: right, building isnt an issue
(13:20:04) asac: ok so a) can be scratchd
(13:20:07) NCommander: asac, I'd say we provide them as a best effort, but I'd lthem to at least have a semi-good chance of working
(13:20:13) asac: so cant we keep them being produced, but just not supported?
(13:20:18) ogra: they build since jaunty :)
(13:20:24) ogra: more or less functional though
(13:20:36) NCommander: ogra, imx51 ones maybe, dove just got their first working ones
(13:20:44) ogra: yeah, imx51 indeed
(13:21:02) persia: StevenK: You're right: it was NEW.
(13:21:20) rgreening [n=rgreenin@198.165.63.165] entered the room.
(13:21:33) StevenK: persia: Right, skew due to NEW I can deal with.
(13:22:18) NCommander: so
(13:22:22) NCommander: go onto ARM image status?
(13:22:37) persia: I thought we were talking about ARM image status.
(13:22:38) asac: wait a bit
(13:22:50) ***ogra didnt do any test install, but the build logs for imx51 look perfect 
(13:22:53) asac: wanted to check a few others.
(13:23:21) asac: JamieBennett: do you have a way forward for figuring out where the time is lost in the debconf-communicate thing?
(13:23:27) persia: ogra: livefs builds failed today (libmpmud0 NEW)
(13:23:32) asac: (casper speedup)
(13:23:35) JamieBennett: asac: Yes, 3 actually
(13:23:47) JamieBennett: But one is the obvious choice
(13:23:56) ogra: persia, didnt look today yet, just returned from the dentist and will lay down after the meeting again
(13:24:00) asac: ok cool. thats on track then
(13:24:13) ogra: persia, but nothing that wouldnt be solved in general ... its not arch specific
(13:24:16) asac: plars: efficient instlal testing ... when will we get the all-pairs generated?
(13:24:32) persia: ogra: Right.  Same happened for powerpc.  Anyway, I'll be glad to give this back to you :)
(13:24:42) ogra: :)
(13:24:43) mvo left the room (quit: Read error: 60 (Operation timed out)).
(13:24:48) plars: the test pairs are generated and documented in the wiki, I just need to write the detail for them on the qa testplan wiki
(13:24:56) plars: then we can create the virtual milestones and link them
(13:25:07) asac: cool
(13:25:13) asac: so when can we start doing all-pairs testing?
(13:25:22) DavidLevin left the room (quit: "Leaving").
(13:25:36) mvo [n=egon@p5B09D539.dip.t-dialin.net] entered the room.
(13:25:36) plars: asac: good target would be somewhere after A2 at the soonest
(13:25:47) plars: the milestones should exist between milestone images
(13:26:21) plars: but A2 is even pretty early to focus *too* hard on it
(13:26:22) asac: ok great.
(13:26:39) plars: doesn't hurt though
(13:26:53) asac: so we assume that paramaters == questions asked during instlal?
(13:27:22) asac: thought it could also be things like: "what install media type" ... "what network connection" etc.
(13:27:27) asac: though testing that might be tricky
(13:27:35) plars: essentially, or could be boot screen options also (once supported), network connection during install, partitioning selection, etc
(13:28:01) plars: asac: for purposes of this type of testing, the network connected or disconnected during install is really what we want to focus on
(13:28:06) asac: ok. lets start with what you have now. we can extend later
(13:28:16) plars: since problems have been exhibited there before
(13:28:18) asac: plars: did you manually generate the all-pairs or do you use a script?
(13:28:24) plars: asac: right, we can certainly extend it all if needs be
(13:28:47) asac: if you have a script that would be great to post on wiki so we can rerun it if we have more paramaters we want to pair
(13:28:54) plars: the pairs were generated programatically, but had to be translated manually (easy) into which items that corresponded with
(13:29:19) leoquant [n=leoquant@ubuntu/member/leoquant] entered the room.
(13:29:33) asac: could you post the tool you used to generate it there?
(13:29:44) Pendulum [n=stephani@pdpc/supporter/active/Pendulum] entered the room.
(13:29:50) asac: ogra: uboot spec is on track for alpha-2 until we know its not on track?
(13:29:56) plars: I think it's linked from the bp wiki page, but I'll check to make sure
(13:29:56) asac: :-P
(13:30:10) ogra: asac, yeah ... will be done latest on the weekend
(13:30:30) ogra: will get 100% of my focus as soon as i feel better
(13:30:36) asac: ok.
(13:30:45) mok0 [n=mok@ubuntu/member/mok0] entered the room.
(13:30:45) ***asac sends a grain of good feeling to ogra
(13:30:50) ogra: :)
(13:30:58) DavidLevin [n=dslevin@67-210-33-186.ul.warwick.net] entered the room.
(13:31:32) NCommander: anything else we want to bring up for Alpha 2?
(13:31:34) ogra: i already am better ... just that anaesthesia going away leaves me with massive pain
(13:31:39) JamieBennett: asac: we have issues with 2D launcher MIR's. We need to convince reviews that upstream will sort out what they call 'minor issues' or provide patches ourself to speed up the process (also two MIR's are unassigned)
(13:31:49) ogra: (and i'm not allowed to put painkillers on top)
(13:31:57) asac: JamieBennett: ok. thanks for letting me know. lets talk about that after meeting
(13:32:01) JamieBennett: ok
(13:32:29) asac: ok the other specs have a backlog as previously pointed out:
(13:32:40) asac: thumb -> work items are review items to happen at alpha-2 stage
(13:32:56) asac: lightweight browser -> took a while to get chromium working, but now we have progress again
(13:33:25) ogra: when will we see it in the archive ?
(13:33:29) asac: suspend-resume -> reduced set of work items (most was moved to alpha-3) should be doable for alpha-2
(13:33:34) mok0: chromium is GREAT
(13:33:59) StevenK: I'd like to see a debian/copyright I'm happy with for Chromium, but that is A LOT of work
(13:34:01) dmart: asac, did we have a working ff build now?
(13:34:04) asac: ogra: i am working on license documentation ... tough task, but progress. i hope we get that to a stage that StevenK is happy to accept it
(13:34:19) ogra: bribe him
(13:34:32) asac: dmart: the patch will get uploaded with .7
(13:34:39) persia: Needs to be a big bribe.  That's almost megabytes of licenses.
(13:34:39) asac: which is happening ... erm ... now ;)
(13:34:47) ***ogra heard rubber boats go well in east australia atm 
(13:34:55) ogra: oh, that was mean, sorry
(13:34:57) asac: persia: i have a tool to generate dep5 format copyrigh tnow
(13:35:02) StevenK: ogra: Hm?
(13:35:07) persia: asac: Yeah, but still :)
(13:35:11) asac: with whitelisting etc. stuff will get down to a reasoanble amount of unknown stuff
(13:35:22) asac: persia: yes. but thats StevenK's main job then ;)
(13:35:33) asac: now that didrocks has started :-P
(13:35:54) asac: ok
(13:35:56) asac: we can move on
(13:36:09) asac: oh ... lib-test ... probably will be late
(13:36:24) asac: not too scary as long as it happens soon after alpha-2
(13:36:25) persia: Maybe it's worth re-milestoning those then.
(13:36:43) asac: persia: its tricky the work item tracker needs a different approach
(13:36:55) persia: Hrm.
(13:36:56) asac: so for split up blueprints we target the next delivery milestone
(13:36:58) asac: and then move ahead
(13:37:12) persia: So, do I need to repeat the stuff about the ARM image status, or is that done?
(13:37:23) asac: otherwise the items we want to see done by alpha-2 wont show up here.
(13:37:39) asac: persia: i think that sank ... maybe repeat ;)
(13:37:49) asac: 14:16 < NCommander> [topic] ARM Image Status
(13:37:50) persia: OK.
(13:38:16) persia: So, ports alternates just got fixed a few hours back (some issue with confusion in a cron job)
(13:38:34) ***ogra just repeats imx51 looks good apart from general non armel related breakage
(13:38:35) persia: As a result, we should start seeing alternate images again soon.
(13:38:35) NCommander: persia, d-cd didn't take well to lpia being rm'ed
(13:38:50) persia: NCommander: Yeah, but I wasn't getting into details :)
(13:38:51) ogra: (desktop that is)
(13:39:09) persia: Live images have been OK for a while, but the livefs builds failed today due to NEW delay (sorted for tomorrow).
(13:39:11) ogra: NCommander, how about dove desktop ?
(13:39:28) NCommander: ogra, they're working aside from the usual livefs breakages
(13:39:44) ogra: whats "usual" ?
(13:39:52) NCommander: ogra, i.e., somethng that would also break imx51
(13:40:02) NCommander: ogra, I guess I should say, no dove specific breakage known at this time
(13:40:03) ogra: and amd64 ?
(13:40:07) ogra: or x86
(13:40:15) ogra: great :)
(13:40:32) ogra: as long as its not armel related i dont care :P
(13:40:32) NCommander: ogra, I'm testing images daily at this point as a lead into alpha2. Rather not miss another milestone with dove
(13:41:01) ogra: you know that they can become shaky this week ... dont count to much on installability
(13:41:12) ogra: everyone tries to get features for A2 ready
(13:41:20) ogra: the archive will be out of sync often
(13:41:34) ***StevenK is planning to break armel images this week and get them building UNE
(13:41:36) NCommander: ogra, I rather catch failure early than during A2 ISO testing.
(13:41:42) NCommander: StevenK, need a hand?
(13:41:43) NCommander: :-)
(13:41:49) persia: Oh, and the last thing I had to say about images is that I'm glad people are returning from holiday so that others are watching them :)
(13:41:55) ogra: right, just make sure to tell them apart from feature development
(13:42:12) asac: StevenK: how about keeping keeping the desktop images until we have a working une image? at least we have a image at alpha-2 then ;)
(13:42:31) ogra: multi package features are likely to trash stuff, especially this week with the last minute uploads
(13:42:37) NCommander: ogra, we still do have some ubiquity weirdness with manual partitioning on dove. That might slide to Alpha 3, I'm grinding my wheels on that one.
(13:42:40) StevenK: asac: I'm happy enough to do that, and davidm agrees with you.
(13:42:42) persia: asac: As long as we have all the packages built correctly, why not change the seeds?
(13:42:49) asac: StevenK: ok cool. thx!
(13:42:58) ogra: NCommander, thats just a bug we can release note ...
(13:43:02) asac: persia: risk minmization
(13:43:05) asac: you never know
(13:43:08) ogra: its an alpha after all, not a final :)
(13:43:09) asac: you can just enable a new seed
(13:43:15) asac: and if that work we disable the old seed ;)
(13:43:17) NCommander: StevenK, is the actual desktop image on dove going to be UNE-2D, or are we going to simply build desktop and ubuntu-netbook-edition-2d ARM
(13:43:28) NCommander: or asac beat me to that question
(13:43:31) persia: new seed?  I thought the plan was just arch-specific stuff to existing seeds.
(13:43:37) StevenK: NCommander: At first, both, and then the former only
(13:43:37) smoser [n=smoser@d14-69-66-169.try.wideopenwest.com] entered the room.
(13:43:43) mvo left the room (quit: Read error: 113 (No route to host)).
(13:43:49) smoser left the room (quit: Read error: 104 (Connection reset by peer)).
(13:43:50) asac: NCommander: UNE-2D is our main target
(13:43:57) smoser [n=smoser@d14-69-66-169.try.wideopenwest.com] entered the room.
(13:44:00) asac: persia: yes. new seed for armel
(13:44:02) asac: not overal
(13:44:09) NCommander: asac, StevenK, TBH, I rather just disable the desktop task on the image build system, and leave alternates for people who want them
(13:44:30) asac: NCommander: that doesnt conflict with what we say
(13:44:31) StevenK: asac: It wasn't going to be a new seed?
(13:44:39) asac: StevenK: your decision ;)
(13:44:45) ***ogra thoughgt it was an on-top seed
(13:44:46) NCommander: If we can build both concurrently, I say thats the way to do it, then just focus on UNE
(13:44:53) asac: i just think having desktop images produced until we have the new images would be great
(13:44:56) StevenK: It will not be a new seed. There.
(13:45:22) mvo [n=egon@p5B09D539.dip.t-dialin.net] entered the room.
(13:45:31) asac: and reduces risk and burn out syndrom potential in case something goes bad
(13:45:32) StevenK: platform -> netbook is hard enough; I don't want to deal with platform -> netbook -> armel-stuff
(13:45:35) persia: \o/ reduced number of seeds
(13:45:43) NCommander: asac, I rather just keep the desktop image around like we build kubuntu and xubuntu armel
(13:45:59) ogra: NCommander, ask IS if you may :P
(13:45:59) NCommander: Maybe turn off the live images to save some time on the livefs builder
(13:46:03) StevenK: NCommander: cdimage space is still a going concern
(13:46:09) ogra: yeah
(13:46:13) ogra: a huge concern
(13:46:17) StevenK: (Since it always is)
(13:46:31) ***StevenK gives the ogra the secret cdimage high five
(13:46:39) ***ogra grins :)
(13:46:55) NCommander: one which will never go away due to images go up, not down. I realize its a concern, but we're simply going to have the same issue when we get a new SoC, or a new image type
(13:47:01) ogra: NCommander, you will always have the ubuntu-desktop task you can install
(13:47:18) asac: so is there anything to discuss? from what i understand we make a seemless transition to the new seed, by keeping desktop images enabled until we have a first working UNE image for armel with 2D launcher, right?
(13:47:21) ogra: even more if we get a new SoC
(13:47:33) ogra: yeah
(13:47:38) asac: ok
(13:47:51) asac: then we all know where we are heading ;)
(13:47:53) NCommander: Are we going to transition people from desktop to UNE on the upgrade usecase?
(13:48:02) ***ogra doesnt think so
(13:48:03) asac: NCommander: unlikely
(13:48:15) NCommander: Do we actually care about the karmic->lucid upgrade for dove and imx51?
(13:48:17) persia: It wouldn't be ideal to force a transitoin.
(13:48:24) ogra: NCommander, we should test them
(13:48:25) asac: desktop packages are still there and in main .... so users just ride the general support wave
(13:48:32) persia: Supporting upgrades is nice, if there is commonly supported hardware.
(13:48:40) NCommander: persia, hence the question.
(13:49:16) plars: would be good to have a metapackage to provide them the packages to update to the all the new UNE stuff, but forcing it on them would probably get some complaining
(13:49:18) NCommander: Anyway, anyone have anything else to bring up on ARM Image Status?
(13:49:21) thekorn left the room (quit: Read error: 104 (Connection reset by peer)).
(13:49:28) thekorn [n=markus@a89-182-201-91.net-htp.de] entered the room.
(13:49:33) ogra: plars, we have to have a task anyway
(13:49:40) asac: yep
(13:49:45) asac: ok so i think we can move on
(13:49:47) asac: NCommander: ^
(13:50:33) NCommander: [topic] Any Other Business
(13:50:34) MootBot: New Topic:  Any Other Business
(13:51:50) mok0: Chromium was mentioned before. I suggest it should be adopted in main
(13:51:52) asac: FTBFS's
(13:52:02) asac: mok0: thats the plan
(13:52:07) NCommander: asac, I saw you assigned a bunch :-)
(13:52:11) asac: depends on various factors ;)
(13:52:14) mok0: asac: ah. Great!
(13:52:33) asac: yep ... so i went through all main FTBFS and assigned them to either ncommander or dyfet
(13:52:42) asac: and both subscribed to both
(13:52:59) asac: dont hesitate to tell me if there are FTBFS you get stuck with
(13:53:02) asac: ;)
(13:53:14) asac: so we can unblock them somehow
(13:53:16) ogra: asac, thanks a lot for apex btw
(13:53:33) asac: for univeres i am through half of the list
(13:53:36) asac: with assignments
(13:53:36) NCommander: asac, thanks for the ocaml upload. Upstream is looking at the bug now as well
(13:53:38) asac: will do the rest
(13:53:54) asac: next ...
(13:54:06) ogra: well, we used to handle universe as a nice to have
(13:54:06) asac: i think its important to get all armel-only build failures fixed. of course main has higher prio
(13:54:09) asac: but universe is also important
(13:54:18) asac: because we are the only ones with hardware to actually work on those
(13:54:20) ogra: do you plan to change that  since you are actively assigning stuff ?
(13:54:27) NCommander: Do we have a general goal to when to have main armel-only FTBFS free?
(13:54:37) ogra: thats what we used to have
(13:54:45) ogra: seems asac wants to widen that goal
(13:54:48) asac: ogra: i assign them because main is almost fixed
(13:54:49) persia: I think universe is all the more important considering the spec to remove binaries for anything that FTBFS.
(13:55:00) asac: its still a lower prio, but its more than a nice to have
(13:55:06) persia: So for *this* release, it's more important than usual to demonstrate that something *can* build.
(13:55:06) asac: at least the armel-only build failures
(13:55:08) NCommander: sorry
(13:55:19) NCommander: I meant an Alpha 2 main-free armel-only FBFS
(13:55:28) ogra: but we give MOTU tools in their hands to work on them
(13:55:34) asac: right. its also more important in the light of the new toolchain
(13:55:40) ScottK: Please fix libtool then as it's causing other FTBFS.
(13:55:41) asac: NCommander: would be great.
(13:55:47) asac: yes. libtool has highest prio
(13:55:47) ogra: and i'd like to see MOTU getting more knowledge here
(13:55:53) asac: i have put that on release team radar too
(13:55:59) asac: and assigned NCommander to it ;)
(13:56:01) ogra: if you assign them to us MOTU wont care at all
(13:56:09) NCommander: asac, I was going to ask for it actually :-)
(13:56:13) asac: ogra: i dont see that realistic for now
(13:56:17) asac: they dont have hardware
(13:56:26) ogra: thats why i wrote all these tools
(13:56:32) ScottK: ogra: Without access to hardware, there isn't much we can do.
(13:56:36) NCommander: dyfet, can you make my shell reappear, or should I work off jocote?
(13:56:38) ogra: and blogged about it ... and set up wiki pages
(13:56:53) ogra: ScottK, you can build in chroots to fix FTBFS
(13:57:01) asac: ogra: well. so most i assigned now are _old_ build failures
(13:57:06) asac: seems noone picked them up
(13:57:08) jsalisbury [n=chatzill@65.206.2.67] entered the room.
(13:57:13) asac: but yes, we should make a new blog post etc.
(13:57:21) ogra: http://ograblog.wordpress.com/2009/07/18/juggling-your-arms-in-karmic-and-no-more-excuses/
(13:57:22) MootBot: LINK received:  http://ograblog.wordpress.com/2009/07/18/juggling-your-arms-in-karmic-and-no-more-excuses/
(13:57:27) dyfet: ncommander: shell?  you mean on my arm?
(13:57:32) NCommander: dyfet, yeah
(13:57:35) NCommander: So who wants to blog to planet?
(13:57:35) ***ogra wonders if anyone reads his posts on planet :P
(13:57:45) ***NCommander has a layer of dust on his blog :-/
(13:58:01) NCommander: two minute warning
(13:58:07) asac: so who has a blog at all? ;)
(13:58:11) asac: and who is on planet?
(13:58:16) NCommander: I'm on planet
(13:58:21) asac: both combined probably reduces the set of candidates ;)
(13:58:23) NCommander: (both Ubuntu & Debian)
(13:58:24) asac: i am on planet too
(13:58:35) ***ogra too
(13:58:40) asac: ok so we have three
(13:58:48) marjo [n=marjomer@pool-98-113-160-46.nycmny.fios.verizon.net] entered the room.
(13:58:49) ogra: and i regulary blog about my armel tools and achievements
(13:58:57) StevenK: Or heating houses
(13:59:07) asac: lets discuss communication offline.
(13:59:10) ogra: heh, yeah :)
(13:59:18) NCommander: [action] ogra, asac, NCommander to blog on planet about ARM work
(13:59:19) MootBot: ACTION received:  ogra, asac, NCommander to blog on planet about ARM work
(13:59:20) NCommander: :-)
(13:59:21) ogra: right, time is up
(13:59:40) NCommander: anything else quickly?
(13:59:59) NCommander: #endmeeting

ARM/Meeting/2010/20100105 (last edited 2011-07-28 17:58:37 by davidm)