20100126

January 26th, 2010, 13:00 UTC in #ubuntu-meeting.

Agenda

Action Items from January 19th, 2009

  • JamieBennett to complete imx51 backportng documentation on boot-speed.

  • cooloney to investigate lucid kernel patches that may need back-porting.
  • ericm and NCommander to investigate what causes the gnome-panel crash/restart cycle.
  • asac, ericm, and NCommander to talk about Thumb2 issues after the meeting and report back.
  • cooloney to ping fsl for more reliable chip rev checking method.
  • cooloney to ping fsl about suspend.
  • NCommander to investigate ericm's xorg naming bug.
  • persia, asac, Gruemaster to discuss meeting rotation schedule and report back.

Current Items

  • Readjustment of the meeting schedule to better fit current participants -- persia

Standing Items

Action Items

  • JamieBennett to complete imx51 backporting documentation on boot-speed.

  • cooloney to investigate lucid kernel patches that may need back-porting.
  • NCommander to ensure that dove gets devtmpfs lucid bits enabled.
  • NCommander to raise apport-retracer for armel from the dead.
  • NCommander to investigate ericm's xorg naming bug
  • persia to suggest and coordinate discussion on meeting schedule
  • asac and ogra to decide what to do with bootloader specs
  • asac to open bugs and assign for issues identified in thumb2review
  • asac and dmart to finish thumb2 review for universe
  • JamieBennett, plars and asac to decide what to do wrt bootchartgui seeding

Minutes

  • ogra informed the team that after recent work on the iMX51 image including kernel patches that were back-ported by cooloney, there is a noticeable boot speed increase.
  • gnome-panel crashes still persisted in the images despite work being done by plars to find the issue. A bug was filed (https://launchpad.net/bugs/458109) and investigations continue.

  • NCommander took up the action item to look into crash issues with the apport retracer which can be flaky at times. It is believed that the ARM fakeroot implementation causes this but no more is known at this time.
  • Talk drifted to Thumb2 issues again, this time asac reported the outcome of talks with dmart from ARM. asac explained that there are three possible area's where Thumb2 could have issues, 1) user space apps; solution is to work around issues, 2) kernel/toolchain; eliminate bad instructions, and 3) hardware; fix bad hardware. Of the three we are seeing on the dove images, the team decided that ideally we would see a change in hardware but the most likely route would be kernel and toolchain workarounds.
  • vldr instructions continue to be a problem on dove. The only possible outcomes at the moment are rebuild the archive for generic ARM, losing all iMX51 performance improvements or find a work-around. Obviously the latter is preferred but the solution escapes the team at this time.
  • ogra informed the team that uboot is to be dropped this cycle in favour of continuing with redboot. uboot has proved to be a little too much to implement for this cycle.
  • Discussion moved to meeting schedules again. persia took the action item to investigate a good time to reschedule the meetings after this cycle.
  • Team work items look to be in good shape with the current status being just above the trend line. There was talk that by the end of the week they would be in even better shape as both asac and JamieBennett planned to close multiple items. There is an outstanding issue with the bootloader spec so asac and ogra took up the action item to sort that out.

  • iMX51 kernels were reported to be in good shape with both plars and GrueMaster indicating that testing had gone well, with plars' new testing methods contributing. He invited people to help fill in the test plan at https://wiki.ubuntu.com/ARM/Testing/ArmPairwiseResult

  • GrueMaster indicated that he had issues with ubiquity last week but plars mentioned that more recent images seemed to fix it.

  • JamieBennett mentioned that the Canola media player would not be ready for Lucid but it continues to be an interesting prospect for Lucid+1.

  • The fail to build list for ARM looks in great shape at this time. Thumb2 utilising packages will still need to be reviewed on a package-by-package basis and asac took up the action to do this.
  • ARM images are looking good even though there are still issues with dove. The team hope to resolve that one way or another very soon.
  • ogra asked if bootchart could be removed from the seeds now. plars and JamieBennett indicated that they would like to see it stay for now and took up an action item to discuss this later.

  • persia has been working with rbelem to get Ubuntu Liquid into shape with the consensus that they are progressing well and aiming for a alpha-3 release.

Weekly Reports

Jamie Bennett (JamieBennett)

This Week

  • Sucked into other projects, no direct work on my own blueprints.

Future

  • Get netbook-launcher-efl into main this week.
  • Finish casper speed-up work.
  • Investigate kernel patches that may help boot speed on ARM devices.
  • Write monthly public team report.
  • Work on Mythbuntu.

David Sugar (dyfet)

This Week

  • A few failed to builds for arm, including busybox
  • Changed ARM default apps focus since we are not getting canola in time
  • Tried a new canola build on arm
  • Struggled with Lucid Dove

Future

  • Learn entire image build process
  • More ftb for arm and otherwise as part of lucid package QA
  • Get Dove back up with Lucid

Alexander Sack (asac)

Lightweight Browser

  • Finalize Firefox 3.6 packaging and collect early feedback from community, fix package transition and get package in archive
  • build a chromium with thumb2+neon build in PPA for benchmark
  • finalize chromium packaging licensing; get first stab uploaded to NEW queue.
  • discuss benchmarking with mozilla
  • work on startup benchmark and try a few javascript benchmarks

uboot spec

  • escalate uboot bugs to freescale

dove port

  • various discussions thumb2/breakage with dove

Team/Communication

  • Release Team Meeting
  • collect input on Alpha-3 planning and communicate to managers
  • work on sprint agenda

Other Contributions

  • discuss uboot
  • sponsor/review fontconfig merge
  • MIR processing (uboot-imx)
  • sponsor/review network-manager, network-manager-applet and modemmanager
  • cleanup work items/blueprints for alpha-3
  • discuss openoffice and python on arm with doko

Paul Larson (plars)

This Week

  • spent quite a bit of time trying to determine the cause of the hard hangs we were seeing with dove. Narrowed it down to pybootchartgui, and eventually python in general. With some, but not all scripts, python seemed to be triggering it.
  • Tested a new build of python - seems to have fixed the hang problem we were seeing with dove, at least the one that was affecting boot wrt pybootchartgui
  • Created virtual milestones, inserted entries for the testcases into iso tracker database.
    • These will be available after the hardy .4 testing is concluded via iso tracker

Future

  • Bringup testing and results posting
  • work on suspend/resume tests

Steve Kowalik (StevenK)

This Week

Attended LCA 2010, in Wellington, New Zealand.

Tobin Davis

This week

  • Spent several hours getting alternate image installed on dove and modified to where I can test for other failures. Python issues kept locking the system up, nearly halting all progress. Found that same drive could be booted on imx51, so used it to apply updates and tweak boot scripts until environment was semi-stabilized.
  • Found another program that segfaults on dove but not on imx51. Reported it as bug #510954
  • Posted bug against Lucid on Dove for the way it currently displays monospace fonts. Trying to type commands in a terminal window is nearly impossible due to the font display issues. Same drive & image looks fine on imx51.

  • Ran OLVER Lib tests over the weekend on imx51. Two sets of results; one with OLVER built against Karmic, one with OLVER built against Lucid.

Future

  • need to update OLVER tests based on email correspondence received from LSB Engineer. Some settings were incorrect, plus there are bug fixes in the test suite that are not in our snapshot.
  • daily testing.
  • Prepare for PDX Sprint.

Michael Casadevall

  • Lost most of last week due to sickness
  • Worked on porting likewise to ARM
  • Continued investigation of Dove thumb2 issues
  • Worked with ericm on testing newer Dove boards for stability
  • Got JTAG up and running on Dove Y0/Y1

Future

  • Dove stability investigation continues
  • Prepare for PDX Sprint

Meeting Log

13:03 < NCommander> Action Item Review
13:03 < NCommander> JamieBennett to complete imx51 backportng documentation on boot-speed
13:03 < JamieBennett> Last week was swallowed up by something else so no progress on that
13:03 -!- ericm_ [n=ycmiao@112.65.48.82] has joined #ubuntu-meeting
13:03 < NCommander> cooloney to investigate lucid kernel patches that may need back-porting
13:04 < asac> [ACTION] JamieBennett to complete imx51 backportng documentation on boot-speed
13:04 < ogra> Bug 512321 has a fix and shoves off 2 seconds :)
13:04 < ubottu> Launchpad bug 512321 in linux-fsl-imx51 "please backport devtmpfs to the lucid linux-imx51 kernel tree" [High,New] https://launchpad.net/bugs/512321
13:04 < asac> cooloney cant attend today, so we should move his item forward
13:04 -!- hyperair [n=hyperair@ubuntu/member/hyperair] has quit [Read error: 110 (Connection timed out)]
13:04 < JamieBennett> :)
13:04 < asac> [ACTION] cooloney to investigate lucid kernel patches that may need back-porting
13:04 < ogra> my imx51 install is booting incredibly fast atm :)
13:04 < JamieBennett> cool
13:04 < ogra> NCommander, make sure thats enabled in dove too ^^^
13:04 < davidm> G'day NCommander
13:04 < NCommander> hey DavidLevin
13:04 < NCommander> ...
13:04 < NCommander> tab complete failure
13:04 < ogra> heh
13:05 < NCommander> hey davidm
13:05 < davidm> bummber about mootbot
13:05 < NCommander>  * ericm and NCommander to investigate what causes the gnome-panel crash/restart cycle
13:05 < asac> [ACTION] NCommander to ensure that dove gets devtmpfs lucid bits backported too
13:05  * persia has begun the process of poking appropriate people
13:05  * NCommander takes actions
13:06 < ogra> asac, no need to backport, just enabling
13:06 < ogra> devtmpfs is in .32
13:06 < asac> NCommander: so the gnome-panel crashes are gone?
13:06 < asac> plars: ?
13:06 < plars> no, I saw a gnome panel crash yesterday
13:06 < NCommander> asac, unfortunately no. Last week was a wash for me due to sickness so I didn't look into it
13:06 < plars> that was on imx51
13:06 < asac> plars: a crash ... or constant crashing?
13:06 < NCommander> Seems to be unrelated to our general hanging issue on Dove
13:06 < asac> NCommander: no problem
13:06 < ogra> plars, uuh
13:06  * ogra hasnt seen any yet
13:06 < plars> err, on dove actually
13:06 < plars> sorry
13:07 < ogra> phew
13:07 < plars> on imx51 it was one of the few things that was fine
13:07  * ogra swipes sweat off forehead
13:07 < plars> I think the root of most of my imx51 trouble yesterday was with the usb bug though
13:07 < NCommander> ogra, bah, you should try having some Thumb2 issues. They're a load of fun
13:07 < ogra> which still isnt clear to me its USB though
13:07 < ogra> since the issues obviously affected the livefs on the SD too
13:07 < asac> plars: do we have a bug for gnome-panel crashes?
13:08 < NCommander> ericm_, anything else to add
13:08 < plars> asac: I didn't get a chance yesterday, was going to go looking and make sure it's in today, but I thought there was one for it already
13:08 < asac> thanks
13:08 < GrueMaster> asac bug 512515
13:08 < asac> [ACTION] plars to file a bug on gnome-panel crashes
13:08 < ubottu> Bug 512515 on http://launchpad.net/bugs/512515 is private
13:09 < ogra> bah
13:09 < plars> or possibly 458109... not sure if it's the same problem though
13:09 < asac> GrueMaster: apport hasnt kicked in yet?
13:09 < asac> can you open it?
13:09 < plars> bug 458109
13:09 < ubottu> Launchpad bug 458109 in gnome-panel "gnome-panel crashed with SIGSEGV in free()" [Medium,Incomplete] https://launchpad.net/bugs/458109
13:09 < asac> ah
13:09 < ogra> why cant i see 512515 ?
13:09 < persia> 512515 is *very* private.
13:09 < ogra> isnt ubuntu.armel subbed ?
13:09 < plars> nor can I
13:09 < NCommander> asac, apport on ARM been somewhat twichy. The retracer crashed while I was out sick, and I haven't kicked it hard enough yet
13:09 < asac> ogra: i assume its a crash report and apport didnt came a round yet
13:09 < ogra> asac, still the reported needs to sub -armel
13:09 < GrueMaster> yes
13:09 < persia> Ah, that explains the *very* private-ness.
13:09 < asac> right. so if you report crashes from a test install you can definitly open them up
13:10 < ogra> *reporter
13:10 < NCommander> [ACTION] NCommander to raise apport-retracer for armel from the dead
13:10 < asac> manually
13:10 < persia> Well, depends on what one did with the test install :)
13:10 < ogra> yeah
13:10 < asac> heh. you get the point ;)
13:11 < asac> ok
13:11 < NCommander> ** asac, ericm, and NCommander to talk about Thumb2 issues after the meeting and report back.
13:11 < ogra> what was the outcome ?
13:11 < asac> the outcome is that we have three knobs to tune:
13:12 < asac> a) user space apps (try to workaround)
13:12 < asac> b) kernel/toolchain (try to eliminate bad instructions)
13:12 < NCommander> a is bust
13:12 < ogra> a) would affect all of wrmel, no ?
13:12 < ogra> *armel
13:12 < asac> c) hardware (fix hardware)
13:12  * ogra votes for b
13:12 < asac> ogra: well. so atm dove images start again
13:12 < asac> the hangups were caused by python
13:12 < ogra> right
13:12 < asac> the new python and recreating the .pyc's fixed it
13:12 < NCommander> b is semi-difficult to do. I'm tempted to go with c unless we *really* want Dove Y series hardware to work
13:13 < asac> so a) is kind of done, but flaky
13:13 < ogra> well, c means we need to replace the world
13:13 < asac> NCommander is working on b) with ericm (thats my understanding) ... and we are investigating hardware (but that probably takes a bit)
13:13 < ogra> which we'll surely do over time anyway
13:13 < ogra> but that requires speed
13:13 < asac> ogra: a relatively small world though ;)
13:13 < ogra> indeed
13:13 < asac> e.g. just our boards
13:13 < ogra> still takes shipping time etc
13:14 < ogra> and throws us back even more
13:14 < asac> NCommander: i thought you had an idea how to do that
13:14 < asac> what happened to it?
13:14 < NCommander> asac, on a or b?
13:14 < asac> b
13:14 < ogra> b
13:15 < ogra> semi indicates it easy for 50% ;)
13:15 < ogra> *it's
13:15 < NCommander> b is still fermenting, but I'm not sure we can handle the busted vldr instructions
13:15 < NCommander> We could change the toolchain to not use vldr, but that requires rebuilding the world
13:15 < ogra> i thought you discussed a solution with dmart yesterday
13:15 < asac> right. what was the outcome?
13:16 < NCommander> asac, ogra, just bounced ideas around. I don't think we came to a definate plan
13:16 < ogra> hmm, k
13:16 < asac> ok. lets try to keep a) going for now
13:16 < ogra> and c
13:16 < asac> i will talk to dmart and ericm about b)
13:16 < asac> and c) is ongoing anyway
13:17 < ogra> right
13:17 < NCommander> indeed
13:17 < dmart> From my pov, if vldr can't be worked around, the only other options are to change the toolchain, or build in ARM
13:17 < asac> NCommander: so you are off the hook and can do other things until further notice ;)
13:17 < NCommander> woo!
13:17 < dmart> NCommander, wasn't there an erratum patch for the vldr problems, or does it not work even with the patch.
13:18 < NCommander> dmart, the patch requires that the instruction is excuted and then faults
13:18 < ogra> dmart, build in ARM means we lose all improvements on imx51 too, no ?
13:18 < asac> yes ogra
13:18 -!- hyperair [n=hyperair@ubuntu/member/hyperair] has joined #ubuntu-meeting
13:18 < NCommander> dmart, its my understanding that our board hangs just executing the vldr instruction will cause the hang. (think Intel style f00f bug)
13:18 < ogra> thats what i feared
13:18 < dmart> ogra, well, yes, probably :/
13:19 < dmart> NCommander, I see. Important to clarify the situation on newer hardware then...
13:19  * NCommander has been doing research into the f00f bug
13:19 < NCommander> I'm curious if we can use a similar technique to work around the issue, but most of the proposed fixes abuses things on Intel processors that don't exist on ARM
13:20 < asac> ok. anything else on this that shouldnt be discussed offline?
13:20  * NCommander has nothing
13:20 < NCommander>  ** NCommander to investigate ericm's xorg naming bug.
13:20 < NCommander> ...
13:20  * NCommander can't even remember what the naming bug was ...
13:20 < ogra> you skipped two items
13:21 < asac> heh
13:21 < asac> guess thats a carry forward ;)
13:21 < ogra> no
13:21 < asac> [ACTION] * NCommander to investigate ericm's xorg naming bug.
13:21 < ogra> ah
13:21 < asac> huh? he says he cant even remember ;)
13:21 < ogra> i thought you mean the skipped ones
13:22 < NCommander> ogra, i skipped cooloney's ones since he's not here
13:22 < asac> right.
13:22 < ogra> doesnt matter
13:22 < ogra> i can comment on both
13:22 < NCommander> ogra, oh, :-)
13:22 < JamieBennett> xorg was wrongly named no?
13:22 < NCommander> In that case
13:22 < NCommander>  ** cooloney to ping fsl for more reliable chip rev checking method.
13:22 < ogra> right
13:22 < ogra> the method is reliable, the bootloaders werent
13:22 < asac> [ACTION] * cooloney to ping fsl for more reliable chip rev checking method.
13:22 < asac> hasnt that happened?
13:22 < asac> right. i thought we have the fix for uboot even
13:22 < ogra> we have bugfixes for uboot but will likely stay with redboot anyway
13:23 < ogra> so that item is gone
13:23 < JamieBennett> boo ;)
13:23 < asac> sure ... but that item is done in any case
13:23 < asac> scratch that action ;)
13:23 < ogra> right
13:23 < ogra> same for the next one
13:23 < ogra> suspend/resume works reliable on imx51 now
13:23 < NCommander>  ** persia, asac, Gruemaster to discuss meeting rotation schedule and report back.
13:23 < ogra> all fixed :)
13:24 < asac> we decided that we discuss that during sprint
13:24 < persia> Oh.
13:24 < persia> GrueMaster and I discussed it briefly about 12 hours ago.
13:24 < persia> (but asac was asleep)
13:24 < ogra> evil you !
13:24 < persia> We decided all times were bad for someone.
13:24 < persia> But I'll get into that more in my Current Items item.
13:25 < NCommander> * Current Items
13:25 < NCommander> **  Readjustment of the meeting schedule to better fit current participants -- persia
13:25 < NCommander> persia, ok, now you can go :-)
13:25  * ogra has no concerns about bad times but we need to make sure to match a proper frequency
13:25 < persia> So, basically no matter what time is selected, it's going to be bad for some folk.
13:25 < asac> right ... i would say, persia should send out suggestions
13:25 < ogra> i.e. i wouldnt change the time in the middle of y release cycle
13:25 < asac> and then we discuss that during sprint
13:25 < ogra> s/y/a/
13:25 < persia> As a result, we need a weighted analysis of attendees and times, to make it least bad for the smallest number of folk.
13:26 < persia> So, I'll volunteer to send something to ubuntu-mobile@ asking for timezones and availability of people who want to commit to attend the meeting, and collect the responses privately.
13:26 < asac> right. so i see two things are to be decided:
13:26 < ogra> but please regard my above comment too
13:26 < asac> 1. what meeting time to use next
13:26 < persia> I'll send out an a summarised mail (no identities) with the weighted analysis, and good times.
13:26 < asac> 2. how often do we want to change the meeting time (hint: not too often, to help community)
13:27 < ogra> 2 ++
13:27 < persia> I think rotation is worse than non-rotation.
13:27  * ogra proposes with a release
13:27 < persia> Even if a meeting time is bad for someone, they can probably rearrange things.
13:27 < persia> Rotation makes for lots of rearragement.
13:27 < asac> yes
13:28 < ogra> well, if it stays for 6 months it shouldnt be to hard to do though
13:28 < asac> but having 9/10 regular atttendeed suffer isnt that great
13:28 < persia> Personally, I'm up for reviewing the time anytime there are significant complaints.
13:28 < persia> That doesn't mean it will change, but as we get new participants, and drop old ones, the weighted analysis may suggest better fits.
13:29 < asac> right
13:29 < asac> [ACTION] persia to suggest and coordinate discussion on meeting schedule
13:29 < persia> So, shall I proceed with this plan?
13:29 < persia> Excellent.
13:29 < asac> moving on ...
13:29 < asac> NCommander: ?
13:30 < NCommander> ** Standing Items
13:30 < NCommander> **  * http://macaroni.ubuntu.com/~pitti/workitems/canonical-mobile.html
13:30 < NCommander> ** Kernel Status (cooloney, ericm)
13:30 < ogra> looks good
13:30 < ogra> pretty close to trend
13:31 < JamieBennett> and some are bug linked which will go away soon
13:32 < asac> that url should be using people.canonical.com now
13:32 < asac> [LINK] http://people.canonical.com/~pitti/workitems/canonical-mobile.html
13:32 < ogra> someone didnt update the wiki :)
13:32 < asac> [LINK] http://people.canonical.com/~pitti/workitems/canonical-mobile-lucid-alpha-3.html
13:33  * NCommander fixes the wiki
13:33 < NCommander> ericm_, how's the Dove kernel ATM?
13:34 < asac> so the itesm start to lack behind
13:34 -!- mok0 [n=mok@ubuntu/member/mok0] has quit ["Leaving."]
13:34 < asac> i will elminate a bunch of the firefox items this week and also get the webservice-email spec on track
13:34 < ogra> what do we do about the uboot spec ?
13:35 < ogra> we had one for redboot taht was swapped for it
13:35 -!- mok0 [n=mok@ubuntu/member/mok0] has joined #ubuntu-meeting
13:35 < ogra> do we swap back ? do we just drop it ?
13:35 < ogra> davidm, asac ?
13:35 < asac> ogra: we should set the status to blocked
13:35 < asac> and communicate the deficiencies to fsl
13:36 < ogra> well, i dont see a way that gets unblocked ever
13:36 < asac> if things get fixed early enough then thats ok
13:36 < ogra> at least for the speed issues
13:36 < ogra> these are uboot design issues
13:36 < asac> otherwise thats the reason we dont switch to uboot and we can postpone the items
13:36 < asac> ogra: uboot works for dove
13:36 < ogra> ok, so no swapping back of the redboot spec then
13:36 < asac> i think we would be fine if we get the same feature set
13:36 < ogra> uboot works for dove but we dont have any alternative :)
13:36 < asac> ogra: lets discuss that after meeting
13:36 < plars> is it the mac address bug that is killing it?
13:37 < asac> [ACTION] asac and ogra to decide what to do with bootloader specs
13:37 < ogra> nor did anyone ever research speed on dove wrt different bootloaders
13:37 < ogra> plars, 23 sec vs 11 sec initiaslization speed and the issue that we still need to pull the kernel off SD
13:37 -!- swe3tdave [n=swe3tdav@ubuntu/member/swe3tdave] has joined #ubuntu-meeting
13:37 < asac> ogra: i am 98% sure that thats a mmc driver bug
13:37 < ogra> which would turn image creation into a horrible thing
13:38 < ogra> asac, no, did you read my mail ?
13:38 < persia> I'm entirely certain it's a driver bug.
13:38 < ogra> no
13:38 < asac> ogra: the loading speed?
13:38 < ogra> did you read my analysis ?
13:38 < persia> I know that there exist i.MX51 implementations that can boot from NVRAM/NVROM
13:38 < ogra> the leoading isnt slow
13:38 < asac> ogra: i read your mail (if it was the mail i think)
13:38 < asac> ogra: the unpacking?
13:38 < ogra> the one i sent today
13:38 < ogra> yes
13:38 < asac> ok
13:38 < ogra> unpacking, initalizing of HW etc
13:39 < asac> i will double check. i think the mail i read was from yesterday night
13:39 < ogra> these are the steps that slow down
13:39 < ogra> i mailed one this morning
13:39 < asac> ogra: for me mmc load alone is slow
13:39 < asac> but lets discuss that after i read you mail
13:39 < ogra> subject: "uboot pro/con list, follow up measuring"
13:39 < asac> lets move on.
13:39 < ogra> right
13:39 < asac> anything on specs=?
13:40 < asac> [topic] Kernel Status (cooloney, ericm)
13:40 < asac> so cooloney isnt here ... ericm?
13:40 < ogra> imx51 looks very good apart from the issues plars sees though
13:41 < ogra> (but that could as well be bootloader induced)
13:41 < plars> planning to retry with sata today, hopefully things will look much better
13:41 < ogra> i'll upload a fresh redboot today, we still use the karmic binary
13:41 < ogra> (since i focused on uboot)
13:43 < asac> ok
13:43 < asac> lets move on
13:43 < asac> [topic] QA Status (GrueMaster, plars)
13:44  * ogra thinks ericm_ fell asleep
13:44 < plars> started working on trying to do pairwise testing
13:44 < asac> great
13:44 < plars> but the first one I started with was the thing I ran into problems with yesterday
13:44 < asac> did you create a wiki to track that run?
13:44 < plars> so not much progress there yet
13:44 < plars> yes
13:44 < plars> but keep in mind this is only a temporary location
13:44 < plars> https://wiki.ubuntu.com/ARM/Testing/ArmPairwiseResults
13:44 < asac> plars: that means we already discovered a new bug because of this? that sounds too optimal ;)
13:44 < plars> the tests are already on iso.qa.ubuntu.com
13:45 < plars> but are not visible until after hardy .4 testing
13:45 < asac> very good
13:45 < plars> but anyone should feel free to contribute if you have some spare cycles to try an install
13:45 < plars> poke me if you have questions
13:45 < asac> GrueMaster: any overview/status on daily testing for last week?
13:46 < GrueMaster> Last week I spent too much time trying to get dove to a working point.
13:46 < GrueMaster> Not much went on with IMX51.
13:47  * GrueMaster is gathering a bug list.
13:47 < asac> thanks!
13:48 < GrueMaster> I filed a bug on keytool, which is part of the openjava packages.  It segfaults on dove, but not on imx51.
13:48 -!- mhall119|work [n=mhall@ubuntu/member/mhall119] has quit [Read error: 60 (Operation timed out)]
13:48 < asac> right. i remember that
13:49 < ogra> ubiquity doesnt work on imx51
13:49 < GrueMaster> Bug 510954
13:49 < ubottu> Launchpad bug 510954 in openjdk-6 "keytool segfaults on dove but not on imx51" [High,Confirmed] https://launchpad.net/bugs/510954
13:49 < ogra> at least it didnt for me on sunday
13:49 < ogra> i didnt research it though
13:49 < plars> ogra: it worked for me yesterday - up until flash-kernel died
13:49 < ogra> right
13:49 < GrueMaster> I got ubiquity to install yesterday on IMX51 after doing an apt-get update on the live image.
13:49 < ogra> so it was probably a one time issue with the friday image i had
13:50 < GrueMaster> The issue was a missing package.
13:50 < ogra> plymouth was fixed today btw
13:50 < asac> ok so ubiquity is fixed.
13:50 < asac> was there a bug open?
13:50 < asac> (for any)
13:50 < ogra> nope
13:50 < ogra> but if both are fixed, no need to bother
13:51 -!- mhall119|work [n=mhall@ubuntu/member/mhall119] has joined #ubuntu-meeting
13:51 < asac> ok ... lets move on ... not much time left ;)
13:51 < asac> [ACTION] ARM Application status (JamieBennett, dyfet)
13:51 < JamieBennett> Nothing major to report
13:52 < asac> i guess not much happend there last week.
13:52 < dyfet> Some of the more ambitious ideas were not realizable in this cycle.  I proposed a much smaller subset of seed changes for Jamie to consider.
13:52 < JamieBennett> dyfet: this is the canola spec right?
13:52 < dyfet> Yes
13:52 < dyfet> At least I assumed so ;)
13:52 < asac> yes, so i recall that we discussed to not go for canola by default this cycle, but rather just ensure its in the archive
13:52 < JamieBennett> So canola isn't possible for this cycle but it could be a reality for lucid+1
13:53 < JamieBennett> and we are definately interested in that
13:53 < asac> yes
13:53 < dyfet> yes
13:53 < asac> ok thanks. sorry, time is running out ...
13:53 < asac> [TOPIC] ARM Porting/FTBFS status (NCommander, dyfet)
13:53 < asac> i think that looks quite ok on the ftbfs list
13:53 < asac> i failed to assign the issues identified in the thumb2 review
13:53 < ogra> whats the number of the bug you filed for likewise NCommander ?
13:53 < asac> so ...
13:53 < NCommander> I've done some porting of likewise to armel, but it was movre involved than I expected. Should have an initial build done soon.
13:54 < NCommander> ogra, no bug yet, that slipped my mind
13:54 < asac> [ACTION] asac to open bugs and assign for issues identified in thumb2review
13:54 < ogra> gnome-power-manager and squd need some attention
13:54 < asac> [LINK] https://wiki.ubuntu.com/ARM/Thumb2PackageReviewList
13:54 < ogra> and there is still libv4l
13:54 < ogra> *squid
13:54 < dmart> asac, did you want to continue with the review of the universe packages in the Thumb2 list at some point?  I did a few, but not very many.
13:55 < ogra> NCommander, please file it :)
13:55 < asac> dmart: yes, we can do another sprint later this week i guess
13:55 < NCommander> ogra, will do
13:55 < ogra> thanks
13:55 < dmart> asac, good plan, please suggest a time.
13:55 < asac> [ACTION] asac and dmart to finish thumb2 review for universe
13:55 < asac> lets discuss time after meeting
13:55 < asac> ... ok hurrying ;)
13:55 < dmart> sure
13:56 < asac> [TOPIC] ARM Image Status (ogra, persia)
13:56 < ogra> already covered for imx51 above
13:56 < persia> Things are finally looking really nice.
13:56 < ogra> all fine ...
13:56 < asac> thats what we want to hear ;)
13:56 < ogra> no idea about dove though
13:56 < ogra> since i didnt test that myself
13:56 < persia> The issues are almost entirely details of image construction (previously covered), and the application stacks are mostly just working.
13:56 < asac> afaik, dove images start and work ... thats more than we thought would happen last week
13:56 < ogra> right
13:56 < asac> thanks to plars for tracking it down to pythong and pybootchart
13:57 < asac> and thanks to doko for getting new python in ;)
13:57 < GrueMaster> None of the alt. images are building atm.
13:57 < ogra> btw, JamieBennett do you still beed bootchart on the images
13:57 < ogra> *need
13:57 -!- jMyles [n=justin@rrcs-24-97-206-244.nys.biz.rr.com] has joined #ubuntu-meeting
13:57 < ogra> the apport popup is annoying on imx51 :)
13:57 < asac> i think we should keep it there for now ... might be interesting after the debconf fixes went it
13:57 < asac> in
13:57 < persia> GrueMaster: Hrm.  I missed that.  I'll see if I can't sort it.
13:57 < JamieBennett> ogra: no
13:57 < ogra> and it doesnt seem anyone makes efforts to fix it
13:57 < asac> hmm
13:57 < asac> ok
13:57 < ogra> so lets unseed it
13:57 < plars> I would like to keep bootchart if possible
13:58 < ogra> meh, k
13:58 < plars> not critical, but nice to have
13:58 < persia> plars: Can't you install with image modification if you need it?
13:58 < ogra> plars, when can we remove it ?
13:58 < plars> sure
13:58 < JamieBennett> plars: it's helpful I agree
13:58 < ogra> well, we need to remove it anyway at some point
13:58 < persia> Or can we stick it in the pool, so it's available for install?
13:58 < ogra> that should be possible
13:58  * persia can provide preseeding hints to help get it installed at install-time
13:58 < ogra> a job for StevenK :)
13:58 < JamieBennett> persia: I'm using it in the live-cd ;)
13:59 < asac> yes, its for live images
13:59 < persia> JamieBennett: But you said you didn't need it, so I'm ignoring your use case :)
13:59 < JamieBennett> but I have my investigations done :)
13:59 < asac> do we capture results from it in a daily fashion?
13:59 < ogra> no
13:59 < JamieBennett> no
13:59 < ogra> it was only used for the livefs bootspeed issues (yet)
14:00 < asac> [ACTION] JamieBennett, plars and asac to decide what to do wrt bootchartgui seeding
14:00 < asac> [TOPIC] Ubuntu Liquid
14:00 < asac> persia: rbelem?
14:00 < rbelem> asac, persia and are working on plasma-mobile and kdm-mobile
14:00 < rbelem> asac, plasma-mobile is already in revu
14:00 < asac> right. i saw that.
14:01 < rbelem> and some minor fixes are needed
14:01 < asac> so you are making good progress on those two.
14:01 < asac> what will come after that?
14:01 < rbelem> yep
14:01 < rbelem> we will work on kdebase-workspace
14:01 < persia> We still have a kwin module to get in, and then we should be able to push a base -meta and -default-settings
14:01 < persia> There's stuff not exposed in the monolithic KDE libraries that we need to investigate more.
14:02 < rbelem> to expose some kcontrol libs
14:02 -!- pendulum_ [n=pendulum@pdpc/supporter/active/Pendulum] has joined #ubuntu-meeting
14:02 < asac> ok. i assume thats coordinated with riddell?
14:02 < rbelem> kdm-mobile depens on it
14:02 -!- mjeanson [n=mjeanson@219.40-ppp.3menatwork.com] has joined #ubuntu-meeting
14:02 < rbelem> yep
14:02 < rbelem> we will make the changes and send to him
14:02 < persia> Well, kinda.  Until we finish the investigation we don't have a lot to propose, but we're definitely working with #kubuntu-devel
14:02 -!- jsalisbury [n=chatzill@65.206.2.67] has joined #ubuntu-meeting
14:03 -!- jsalisbury [n=chatzill@65.206.2.67] has quit [Remote closed the connection]
14:03 < rbelem> after that we should work to trim kdm
14:03 < asac> ok. what milestone is liquid aiming for? I assume the kdebase-workspace changes are somewhat bound to the alpha-3 milestone as kubuntu is going for LTS?
14:03 < rbelem> yep
14:04 -!- jsalisbury [n=chatzill@65.206.2.67] has joined #ubuntu-meeting
14:04 < rbelem> our milestone is alpha3
14:04 -!- mjeanson [n=mjeanson@219.40-ppp.3menatwork.com] has left #ubuntu-meeting []
14:04 < persia> For all the KDE stuff.  We might end up with some tweaks all the way to FF to get other stuff clean.
14:04 < asac> ok. so try to align your priorities so that things that need to be changed in software that is shipped by others is done around alpha-3
14:04 < persia> PRecisely.
14:04 < asac> if liquid-only things come a later its not a big deal imo
14:05 < rbelem> cool!
14:05 < rbelem> :-)
14:05 < persia> Well, we want to get at least -meta and -default-settings in by FF.
14:05 < persia> But they might not be 100% feature-complete at that point.
14:05 < persia> (package add being more invasive than bugfix stuff, even large bugfix, if nobody else installs the things)
14:05 < asac> sure. but if you have to decide what to do first, i would suggest to do the changes that have impact on kubuntu first ;)
14:06 -!- DavidLevin [n=dslevin@67-210-33-186.ul.warwick.net] has quit [Client Quit]
14:06 < persia> Absolutely.  kdm/kwin are the big things now, because they need to build against stuff that we might need to expose from kdebase-workspace (as above)
14:06 < asac> well. not if those packages have no impact on other products imo ... but yes, FF should be the goal
14:06 < asac> ok. but feels its on track
14:06 < rbelem> :-)
14:06 < asac> if anything is needed just shout ;)
14:07 < persia> We will.
14:07 < asac> sorry, to cut off but we are overdue ;)
14:07 < asac> [TOPIC] AOB
14:07 < ogra> doesnt include beer :)
14:07 < ogra> or does it ? :)
14:07  * NCommander thanks asac for taking over
14:07 < asac> NCommander: np
14:08 < asac> one thing from me: lets try to add our acitivites to the wiki _before_ the meeting ;)
14:08  * asac slaps himself
14:08 < ogra> meh
14:08 -!- shriekout_ [n=shriekou@124.216.208.61] has joined #ubuntu-meeting
14:08  * ogra totally forgot about it ... 
14:08 -!- shriekout_ [n=shriekou@124.216.208.61] has quit [Read error: 104 (Connection reset by peer)]
14:09 < ogra> i'm missing JamieBennett'S nice remonders
14:09 < JamieBennett> :)
14:09 < ogra> *reminders
14:09 < asac> thougth we had one ping
14:09 < asac> guess jamie was too busy last week  to do that on friday
14:09 < asac> also i think i failed to send out the meeting minutes last week ... is that correct?
14:09 < JamieBennett> I'll start them back up on fridays
14:09 < JamieBennett> asac: yes
14:09  * asac hides
14:10 < JamieBennett> I wrote them on the wiki though
14:10 < ogra> there are always the logs
14:10 < asac> JamieBennett: will you send both? or want me to still send last weeks minutes?
14:10 < JamieBennett> asac: I can do both
14:10 < persia> logs generally hit the TLDR barrier.  Minutes are better.
14:10 < asac> thanks. /me owe's JamieBennett a cookie ;)
14:10 < asac> yes, minutes are much bettter
14:10 < asac> especially if they are so well prepared as ours
14:10 < asac> on the wiki ;)
14:11 < JamieBennett> :)
14:11 < asac> ok thaks all

ARM/Meeting/2010/20100126 (last edited 2011-07-28 17:58:17 by davidm)