14:58:35 #startmeeting 14:58:35 Meeting started Thu Feb 16 14:58:35 2012 UTC. The chair is NCommander. Information about MeetBot at http://wiki.ubuntu.com/meetingology. 14:58:35 14:58:35 Available commands: #accept #accepted #action #agree #agreed #chair #commands #endmeeting #endvote #halp #help #idea #info #link #lurk #meetingname #meetingtopic #nick #progress #rejected #replay #restrictlogs #save #startmeeting #subtopic #topic #unchair #undo #unlurk #vote #voters #votesrequired 14:58:41 you are early today 14:58:55 hi 14:59:09 o/ 14:59:41 [link] https://wiki.ubuntu.com/ARM/Meeting/2012/20120216 15:00:05 #topic ActionItem Review 15:00:10 #topic rsalveti to follow with riku to see if chromium will get fixed 15:00:42 looks like that was done 15:00:47 Riku and micahg were discussing chromium earlier. 15:00:52 Seems like it's nearly there. 15:00:52 we had an extensive discussion this morning 15:01:13 o/ 15:01:29 and while we are at fixing, 1000 hugs to janimo` for fixing libO in just a few days of work :) 15:01:59 ogasawara, well it was a few days work preceded by a few days of staring at the code dumbfounded 15:02:11 heh 15:02:17 yay janimo` 15:02:19 * ogra_ hugs ogasawara ... 15:02:26 ogra_: You've been upgraded to ogasawara! 15:02:32 she always steals my pings ! 15:02:34 awesome! 15:02:51 ogra_, start lurking the kernel channel and steal some back 15:02:53 anyway, no rsajdok so moving on 15:03:07 janimo`, i'm resident there, she steals them there too 15:03:12 #topic Standing Items 15:03:14 #topic http://status.ubuntu.com/ubuntu-precise/ubuntu-arm.html 15:03:45 * ogra_ guesses we need to postpone everything thats FF cfritical today 15:03:51 *critical 15:03:58 Looks like the TODO items have taken a dip. 15:04:21 I don't think any of my items will be blocked by FF. Sadly. 15:04:44 * ogra_ is still planning to fix the ubuntu-core DL page if i ever find out how the heck infinity manages to make make-web-indiciec being ignored 15:04:57 ogra_: deep voodoo 15:05:03 yeah 15:05:08 maple voodo :) 15:05:13 :P 15:05:30 We can talk later. 15:05:35 I didn't realise you were having issues. 15:05:39 [topic] ARM Server Status (NCommander) 15:05:49 linux-armadaxp exists 15:05:51 yeah, not today though 15:06:00 It even installs. 15:06:06 .... i didnt shout yet since i want to find out myself first :) 15:06:30 working on the d-i portion, and working with cooloney to merge in the lastest bits from Marvell 15:06:40 ready today ? 15:07:13 We didn't quite make feature freeze withthe actual d-i bit but the rest should be there 15:08:03 except images tonight/tomrrow 15:08:08 ready tomorrow ? (else i need to notify about the FF exception in tomorrows release meeting) 15:08:18 ok 15:08:40 ogra_: probably tonight if everything goes as planned.If not,tomorrow 15:08:56 disclaimer 15:09:23 ok 15:09:25 this is for netboot. I do plan to try to build alternates but I don't know ifthey'll work right offthebat 15:09:31 if not tomorrow, please let me know 15:09:40 will do 15:09:42 [topic] Kernel Status (cooloney, ppisati) 15:09:55 new kernel release yesterday (1406.8) 15:10:02 with audio fixes 15:10:10 yea! 15:10:16 now i'm working on the "set mac address" from the cmdline 15:10:27 it's broken since natty 15:10:35 * janimo` is currently building a new ac100 3.0.19 kernel and if all's well it is going to be uploaded soon. 15:10:45 \o/ 15:10:58 janimo`, i think -exp was merged yesterday into the stable branch btw 15:11:12 janimo`: Let me know when it gets in. 15:11:13 ogra_, yes, and two more sound related commits since yesterday 15:11:17 GrueMaster, sure 15:11:23 yup 15:11:25 I want to see if it fixes the start-r board. 15:11:39 GrueMaster, start-r ? 15:11:50 Wrong kernel.. 15:11:53 I may need extra config options if that is some USB dongle 15:11:55 New rev. 15:12:01 GrueMaster, ac100 15:12:11 Oh. 15:12:19 * GrueMaster sips more coffee. 15:13:28 NCommander, move ? 15:13:30 [topic[ ARM Porting/FTBFS status (NCommander, janimo) 15:13:40 bah ninjaed 15:13:50 * NCommander hugs janimo` for libo 15:14:01 * janimo` hugs NCommander 15:14:04 for libo 15:14:16 humm ... upstart 15:14:19 and funeral services will be held for his sanity next thursday at 15:00 UTC. 15:14:25 I am debugging upstart now 15:14:29 ah, cool 15:14:35 The upstart failure isn't arm-specific. 15:14:44 as in just fired up gdb. Not sure why it built on armel 15:14:45 It takes several builds to beat the race on x86 buildds. 15:14:57 infinity: thatbug is STILL around? 15:15:01 We've just been throwing it back until it sticks. :/ 15:15:13 I remember keybuk working on that race when he was still here 15:15:17 infinity, so it is a bug but very subtle. throw it back I guess then :) 15:15:22 janimo`: That said, arm-specific or not, the upstart devs might appreciate someone finding the problem. 15:15:25 infinity, well, let the awesome janimo` look at it :) 15:15:40 janimo`: We can't get it to stick reliably on arm, that's the problem. :P 15:15:54 if we used systemd we could just blame Lennart. Now we have to do actual debugging, grumble 15:16:10 lol 15:16:16 now thats a reason for switching 15:16:19 heh 15:16:26 "because we can whine about lennart" 15:16:46 somehow, I don't think it would make ARM bugs vanish 15:16:54 I think slangasek was looking at this yesterday as well. Saw a post of him mucking with cross-compiling in qemu. 15:17:15 NCommander, no, but we could just just blame him for them ;) 15:17:39 we should persuade Lennart to take over upstream upstart maintainership, then we can blame him without switching 15:17:41 ogra_: but we can blame Scott for upstart so whats the difference :-P 15:18:00 janimo`, awesome idea ! 15:18:07 vorlon's upload was meant to fix the test that still fails so, yeah, catching up with him might be good. 15:18:09 NCommander, we like scott 15:18:19 [action] janimo` to get Lennart to maintain upstart 15:18:19 * meetingology janimo` to get Lennart to maintain upstart 15:18:28 I figured you got libo fixed, that shouldn't be hard 15:18:40 hmm, you have apoint 15:19:01 anyway 15:19:05 [topic] ARM Image Status (ogra, NCommander) 15:19:13 broken today 15:19:13 To soon be existing for arm server 15:19:17 thanks to compiz ftbfs 15:19:27 (across the board) 15:19:44 oh, and an announcement (seems to fit here) 15:19:44 that compiz ftbfs seems like a QT issue 15:19:48 the kde backend is the one failing 15:19:49 canonical will from today on support armhf 15:19:55 ''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''' 15:19:56 \o/ 15:20:05 WOOT! 15:20:05 and turn armel into community supported 15:20:19 so GrueMaster will still test it! 15:20:20 I'm dropping armel+mx5 and armel+omap4 images today. 15:20:21 i guess infinity can tell us which images we still build for armel 15:20:30 Are we dropping image builds for armel? 15:20:34 YEA! 15:20:36 GrueMaster, not all 15:20:40 We'll still build armel+omap and armel+ac100, and drop ac100 when nvidia gives us armhf drivers. 15:20:54 that means there are slots freed for live image testbuilds sometime soonish ? 15:20:56 Ultimately, we should just be building armel+omap for smoketesting. 15:21:09 janimo`: Not really. 15:21:17 janimo`: It means we don't have to wait 16 hours for all the images to build. 15:21:21 janimo`, there are tons of slots once infinity is done with the move to LP buildds for image builds 15:21:21 infinity, you keep shattering my dreams 15:21:31 janimo`: he's good at that 15:21:37 but that will still take a bit 15:21:47 he has no sleep and dreams, so messes with others' 15:22:00 Yes, yes, I'm a bad man. 15:22:02 Moving on. 15:22:06 ++ 15:22:14 [topic] QA Status (GrueMaster, mahmoh) 15:22:51 I wrote a few scripts to do server load testing. One for the lamp stack, one for rabbitmq. 15:23:28 Also working on getting power meter data acquisition going. Works on a panda so far. 15:23:44 nice 15:24:00 do you have any intel based comparison data ? 15:24:02 Tested daily images earlier this week, no sudden deaths. Looking solid. 15:24:25 (i.e. from cking) 15:24:35 Unfortunately, it is just a digital amp meter. Very hard to get power loads across multiple voltages. 15:24:43 ah, k 15:24:56 he is testing with a laptop. 15:25:04 right 15:25:11 And completely different loads. 15:25:18 k 15:25:57 anything else? 15:26:03 [topic] Linaro Updates (rsalveti) 15:26:03 move 15:26:05 er 15:26:07 I put my arm server test system on a kill-a-watt device, but it barely registers. 15:26:12 I ran some kernel burn in yesterday and pushed a PTS script to the workload branch - getting to WI today 15:26:24 oh. lots else :) 15:26:55 still no rsajdok 15:26:56 er 15:27:01 still no rsalveti 15:27:04 [topic] AOB 15:27:07 well, lets skip this then 15:27:15 janimo`: it seems gcc has got stricter again and some new errors are turning up, that compiz issue is one 15:27:19 though i would have loved to hear where compiz stands :( 15:27:22 janimo`: I'd like to get info on how you tested erlang & rabbitmq last cycle. after the meeting 15:27:27 janimo`: and so is digikam and something else 15:27:30 GrueMaster, ok 15:27:44 Riddell, ah, so a bug for Linaro gcc folks I guess? 15:27:53 or for doko .) 15:27:56 janimo`: it's probably not hard to work out the way to keep the compiler happy but it's hard for me with no arm hardware 15:28:06 janimo`: or just drop the kde compiz bits as a workaround 15:28:30 janimo`: only if they enjoy fixing c++ cast problems :) 15:28:32 Riddell, ok, we'll probably do that as a workaround since the alternative is fixing and uploading Qt 15:28:33 Riddell: I thought you had a panda? 15:28:35 which takes longer 15:28:47 janimo`: compiz is a "bug" in kdelibs not qt 15:28:58 Riddell, it's part of their job to fix C++ cast problems :) 15:29:00 GrueMaster: I do, yet to find time/knowledge to set it up 15:29:02 janimo`: Fixing the errors seems saner than disabling features. 15:29:13 * ogra_ agrees that compiz is a bug .... not only in kdelibs 15:29:26 infinity, sure unless fixing it means no images for weeks? 15:29:29 Riddell: Get with me offline, I can get you up and running in minutes. 15:29:53 yes, getting a panda up is a matter of a few minutes nowadays 15:29:58 GrueMaster: I will do next time I find some minutes 15:30:03 (plus download time) 15:32:40 anything else or can I close? 15:32:46 close 15:33:03 #endmeeting