=== pizthewiz_ is now known as pizthewiz === Jack87 is now known as Jack87|Away === Jack87|Away is now known as Jack87 [07:38] list ubuntu-arm === Pratik_away is now known as Pratik [08:07] hello , does ubuntu11.10 for panda has no h264enc codec? i run the gst to encode h264 failed, who knows ? [08:20] can anybody help me with ARM NEON? [08:29] Good morning, can anybody help me with vector permutation using ARM NEON? [10:24] TheMuso, hey, i meant the other paz00/alc patches ... i dotn know if all of them made it into the recent version we have [10:26] GrueMaster, I am looking at the two main FTBFS you brought up. thanks [11:42] * ogra_ curses, seems my install is seriously scrwed :( [11:48] * ogra_ curses mono [11:49] ogra_, what does mono do in your install? [11:49] fail the assembly install from the postinst [11:50] but what requires it, it's not in default install right? [11:50] note that my system got upgraded from natty over oneiric to precise, i have a lot of shit installed that shouldnt be there [11:50] * ogra_ is just trying to get rid of mono [11:51] also i hacked the alternatives for nvidiy-tegra, that fails too [11:54] * ogra_ tries to purge everything with -cil in the name, lets see if makes it any better [11:56] silly mono crap .... and nobody to fix it :( [11:56] ogra_: the latest mono at least installed on my machine but its asserting all the time [11:57] right, for me it seems to already breal when the assembly update happens [11:57] luckily we dropped it from the images [11:57] still thats a basd situation [11:57] *bad [12:08] grmbl [12:11] update-alternatives: error: alternative nvidia_drv can't be slave of arm-linux-gnueabi_egl_conf: it is a slave of gl_conf [12:11] i really wonder how i broke that [12:12] ogra_: i fixed it on my machine by removing the slave of gl_conf [12:13] in the nvidia-tegra package ? [12:13] i think the bug are simply that the old nvidia-tegra package added the slave for gl_conf [12:13] so the bug is in the state saved in update-alternatives [12:14] well, thats bad, the new package should have a fix for it [12:16] i think i ran update-alternatives --remove gl_conf [12:16] before configuring the new package [12:16] it will remove gl_conf and all associated slaves [12:19] ogra_: try ask someone who better understand this alternative system [12:19] how the transition from non-multiarch to multiarch are supposed to work [12:19] well, i know enough about alterbnatives, i'm just to lazy to weed through all the postinst scripts :P [12:20] the QA team will have a fun time hitting all these upgrade QA issues [12:20] when testing oneiric -> precise upgrades [12:21] i dont think they test ac100 upgrades [12:34] ogra_, the alternatives issue I saw as well, but was not sure if it is not caused by my broken manual alternatives renaming earlier [12:34] maybe preinst should have something to remove the early names indeed [12:35] but I do not understand alternatives well either [12:35] yeah, something like that ... i fiddled with the alternatives as well, though [12:36] i think the prob is that egl is already a slave of gl ... the driver wants to create a slave for a slave [12:37] * ogra_ purges mesas egl stuff for a test [12:39] hmm, intreasting, that removes gimp [12:40] who would have guessed :) [12:40] im surprised that gimp needs egl [12:42] well, it also installs a ton of QT packages that seem to have been held back [12:44] i wonder if we missed a transition here === Ursinha` is now known as Ursinha [13:33] cooloney: ok, I'm spinning a source package now with the change to changelog (guess I need to retag too :-/) [13:33] I'll dump my branch onto zinc once its done [13:34] cooloney: also need to delete all non-revelant changelog entries [13:35] NCommander: yeah, please send me patch [13:35] let me update the git tree [13:36] cooloney: great [13:36] give me a sec to continue smacking [13:36] argh [13:36] Freaking changelog manager [13:37] * NCommander headthunks [13:37] we have a manager for that ? [13:37] how much does he earn per month ? [13:38] :P [13:39] * NCommander hits ogra_ with a brick [13:40] hey, one day i'll build a house... i already have a good collection of your bricks [13:41] * NCommander his ogra_ with a rock [13:42] *hits [13:42] * ogra_ reboots after that dist upgrade finally finished (that took half my workday, sigh) [13:42] * ogra_ throws a mono stack at NCommander [13:42] * NCommander cuts it in half with his libO foo [13:47] cooloney: http://paste.ubuntu.com/835214/ - plz apply [13:47] * NCommander dumps linux-armadaxp into a PPA to testbuild properly [13:50] wm8650 [13:51] NCommander: meta will be ready soon. thx janimo [13:52] uploading to a ppa now to testbuild (not sure all the necessary changes to get the _all packages to build prpoerly made it in) [13:53] going to go get some food, will return [13:53] cooloney, you're welcome [13:54] janimo: for testing build meta, what's kind of build target should i use? [13:54] fdr clean; fdr binary? [13:54] hmm, I think I just used debuild -us -uc when testing [13:55] as it builds very fast - nothing to actually build but the metadatat under debian/ [13:55] what you say is equivalent I guess [13:55] janimo: yeah, i got it [14:03] NCommander: i saw you patch, did you miss to change 3.0.0 to 3.0.17? [14:08] cooloney: infinity gave a good reason not to include it [14:08] NCommander: i miss that reason, why? [14:09] NCommander: let me change back the meta package from 3.0.17 to 3.0.0 [14:09] cooloney: because its a leftoer when we used 2.6.x, including the sub-minor just means the version number cahnges a LOT and is a pain [14:09] NCommander: looks like we have to stick to the tradition [14:09] yeah [14:18] NCommander: how about i fold your patch to my previous commit [14:18] NCommander: and i will push out this change again [14:18] cooloney: that's fine. you'll probably get a few more patches as I'm not sure this will build onteh buildds (doing a PPA test build to check) [14:20] NCommander: great [14:22] NCommander: meta package is here, git://kernel.ubuntu.com/roc/linux-meta-armadaxp.git [14:22] great [14:22] NCommander: please help to review and upload. [14:23] or the gitweb http://kernel.ubuntu.com/git?p=roc/linux-meta-armadaxp.git/.git;a=summary [14:33] NCommander: kernel git tree is also updated with your patch [14:33] * cooloney gonna go to sleep [14:33] good night, folks [14:44] cooloney: sleep well [14:50] ogra_, are the armhf omap sgx drivers coming? [14:50] janimo, at some point, yeah ... with dropping of el [14:51] NCommander, cooloney wonder why is linux-armadaxp building for i386 too [14:51] TI will only support one [14:51] janimo: arch-all voodoo, whichis why I wanted to do a test build to see if it actually will build properly. [14:51] for the brave that want to run armada kernels on pentiums :) [14:51] ogra_, are they waiting for something or next release will automatically be armhf? [14:51] janimo, next will ... they dont have time and manpower, so we have to wait [14:52] ok [14:52] * ogra_ finally finds the time to test janimo's new kernel ... [14:52] do we wait with switching images over to armhf only until they arrive? [14:52] though thats probably brave right before a meeting :) [14:52] ogra_, heh, as long as you don't suspend it should not be an issue :) [14:52] janimo, nope, we switch when we switch ... up to them to react then [14:53] ah, they may not want to put any effort in it until it is 100% 12.04 is armhf primarily [14:53] might be, we wont wait for them though [14:54] * ogra_ installs [14:54] sure. what about nvidia/ac100? Not sure if they even promised armhf this cycle [14:54] no idea [14:54] likely not [14:55] I'd still switch to armhf by default on ac100 [14:55] same thing for omap3 ... ask rsalveti about that one [14:55] ok, in today's meeting probably [14:55] yeah, me too i guess, though that means a cross grade or some such for me ... i dont want to lose all my data [14:56] nvidia have produced armhf drivers for meego in the past [14:56] * ogra_ reboots into new kernel [14:56] do we have any nvidia person to ask directly? [14:56] * janimo wonders what kind of unbackupable/unrestorable heaps of data may ogra's ac100 hold that he cannot just wipe and reinstall armhf [14:57] i doubt they will support it officially but they may create an armhf build for distribution with a specific distribution [14:57] xranby, well a binary dump as their current driver is would suffuce [15:05] janimo: i think it basically boils down to find someone who have access to nvdeveloper.nvidia.com and ask for a rebuild === zZzZ_krosswindz is now known as krosswindz [15:05] the build will then be uploaded to NV online for use by a specific distribution [15:06] this request can possibly be made by someone with influence in the oem team [15:06] xranby, yes, I know some of them work with nvidia occasionally [15:08] try pull one of those strings, i think thats the only way to get a driver into armhf at this point [15:09] xranby, last I heard we needed a new armel rebuild for the new x11 abi. But I am not sure if that arrived or not [15:10] janimo, that happened two uploads before you touched the package ;) [15:10] janimo: for armel nvidia was kind to at least include a v11 abi version next to the v10 abi version [15:10] in the same upstream tarball [15:11] so i think we have that allready :) [15:11] yep [15:11] * ogra_ is currently using it [15:11] well I saw xabi6 through 11 in the upstream binaries but was not sure the last one was ok or needed a newer rebuild [15:11] and given ui checked the changelog today i can say that it was switched at -0ubuntuX still [15:11] it has issues with the 11 abi though [15:12] i have a lot of composite messups on my screen [15:12] ogra_, yes, I think that is what I remembered, it may not be 'the final abi11 rebuild' [15:12] hopefully [15:12] well, its still usable and i like that it saves so much ram [15:13] so i'll keep it running, hoping the best from nvidia === Jack87 is now known as Jack87|Away === dduffey is now known as dduffey_afk === dduffey_afk is now known as dduffey === dduffey is now known as dduffey_afk [18:48] no usb-creator for arm? [18:50] apparently not. Odd, considering it is just a python script (I think). [18:55] Probably because it depends on i386/amd64 only packages (syslinux). [19:26] GrueMaster: Definitely because of that, in fact. [19:30] GrueMaster: btw i tried applying those two patches that were posted on that analog sound -panda board bug and no love [19:31] Ok, good to know. make sure you update the bug report. Thanks. [19:32] ah you mean i need to create a launchpad account [19:32] ;) [19:32] (totally joking, ill update the bug) [19:32] heh [19:40] comment posted [19:41] and a little more info added === dduffey_afk is now known as dduffey [20:07] btw: Do you have a Panda or a Panda ES? [20:07] pbuckley: ^^^ [20:08] es [20:08] have two of em on my desk right now [20:09] one is my desktop, the other is my development target [20:09] I think the alsaucm rules need to be changed for that system. [20:10] any idea what too? my alsa foo is non-existant other then knowing where the config files exist [20:12] I'm booting up to look. [20:13] cool thank you [20:17] I think there may be a couple of places this needs fixing. First, /lib/udev/rules.d/90-alsa-ucm.rules, then /usr/share/alsa/ucm. [20:22] The 90-alsa-ucm.rules is a simple fix. Just run "sudo sed -i 's/Panda/Panda|PandaES/g' /lib/udev/90-alsa-ucm.rules". [20:23] The /usr/share/alsa/ucm part is a little more involved. Need to copy the Panda directory to PandaES, then edit all the files inside to reflect PandaES instead of Panda. [20:25] vi sweet [20:25] my ears thank you.. trying it now [20:31] I'll file a new bug on alsa-utils & libasound2, then go about getting them fixed. Even if the kernel is still borked, they will need these fixes. [20:33] GrueMaster: Yay, I get to duplicate that code again? :) [20:34] ??? [20:34] GrueMaster: The alsa ucm bits. [20:35] I suppose we can just ship symlinks. [20:35] Oh, no. We can't, [20:35] Cause the board ID is in the files. [20:35] *grumble* [20:35] Oh. Nah, simple fix. Already on it. [20:35] Now I remember why I copied it all. [20:36] I got this one. You can fix ksysguardd. :P [20:36] Heh. [20:36] Or, I can attend more meetings! [20:36] \o/ [20:38] meetings, huzzah! [20:39] Meetings == snooze time. [20:40] $ alsaucm -c PandaES set _verb HiFi [20:40] Im setting defaults [20:40] ALSA lib main.c:276:(execute_sequence) unable to execute cset 'name='HS Left Playback' 1' [20:40] ALSA lib main.c:1460:(set_verb_user) error: failed to initialize new use case: HiFi [20:40] alsaucm: error failed to set _verb=HiFi: No such file or directory [20:40] i probably foobar'd something [20:40] hrmm [20:43] for file in *; do sudo sed -i 's/Panda/PandaES/g' $file; done [20:43] basically just did that [20:43] in the /usr/share/alsa/ucm/PandaES directory [20:43] (ill just wait for you to commit a fix) [20:45] wip. Any chance you can post your new shiney kernel (just the uImage) for me to test here? [20:45] Otherwise I'll have to rebuild it. [20:48] ogra_: afaik those other patches are for the kernel I think. [20:50] yeh, where do you want it? [20:50] i have the deb's too if you want [20:50] Someplace accessible. [20:50] hrm that might tricky.. can you do dcc? [20:51] I might. Not sure if quassel-core can handle it. Give it a whirl, worst case it blows up my rack of servers. :P [20:51] heh [20:52] get it? [20:52] TheMuso: I'm getting ready to patch alsa-libs & alsa-utils for PandaES. Should I just push them into bzr branches or do you want me to push to alsa-project as well? [20:53] 12:52 DCC SEND request sent to gruemaster: linux-image-3.2.0-1405-omap4_3.2.0-1405.7+soundfix_armhf.deb [20:53] pbuckley: I see the request. Not sure how to handle it. One sec. [20:53] /dcc accept? [20:54] hrm. [20:54] TheMuso, oh, ok, i just noticed that there were more patches in the list archive, if its all kernel it should be fine [20:55] unfortunately the deb is 20MB which is a little to big for my email system.. lets see if i can sneak it on a web server [20:55] GrueMaster: Push them into a bzr branch for now. I was actually going to ask you guys about the ucm files for the ARM stuff, we currently carry that. DO you want it all to go upstream? [20:55] pbuckley: No dice. Quassel doesn't support it yet. Give me a minute. [20:55] k [20:56] BTW how is armhf looking for precise, as in supportability etc? [20:56] TheMuso: I think they should, yes. I haven't updated my alsa git trees in ages, but it isn't a big deal. [20:56] armhf is rock solid (with a few cracks). [20:56] ive been having pretty good luck with armhf [20:56] Ok, I know they are not upstream yet. [20:57] I presume such cracks are no big deal, i.e they are not major. [20:58] pbuckley: Ok, send it to my other nick (GrueMaster-web). [20:58] sent [20:59] 12:59 gruemaster-web SEND: 0B of 19MB (0%) - 0.00kB/s - ETA (stalled) - [20:59] how big can attachments be on gmail? [20:59] lets see [21:00] Not sure, but it should be smaller than the dirty jokes my mother sends. Email away. [21:00] lol [21:00] email? [21:00] (address that is) [21:00] gruemaster@gmail.com (what else would I be :P) [21:01] leet [21:01] sending [21:01] my internet here isn't the fastest.. give it like 10min [21:02] np. [21:02] Not like I have nothing else to do. :P [21:05] rsalveti: Linaro is already working on the alsa-ucm??? Good to know. [21:05] k [21:05] it should be sent [21:05] let me know if you need the headers deb too [21:16] GrueMaster: receive the email ok? [21:16] yes, installing now. thx [21:17] no no, thank you :) [21:45] WOOT! Sweet stereo analog on Panda ES is back. [21:46] Dosen't help that they change the control names for every release. [21:48] sweet, so that kernel worked? [21:49] hello. could someone point me to a hardfp build of ubuntu? [21:49] nm [21:49] i am silly. [21:49] Between your kernel and me mucking with the ucm config files, yes. [21:49] nice! thats great news.. that just means im doing something stupid [21:50] I'm going to pasebin the /usr/share/alsa/ucm/PandaES/hifi file for you. Hang on. [21:50] thank you [21:51] http://paste.ubuntu.com/8358000 [21:51] Oops. remove one 0 [21:52] got it.. brb going to reboot with that kernel [21:52] Make the ucm change first. That way you only need one reboot. [22:26] I HAVE SOUND! [22:26] :D [22:27] thanks GrueMaster [22:28] Excellent. [22:30] GrueMaster: the multimedia working group is trying to get it properly fixed [22:30] for all the boards [22:31] Great. [22:31] it's a working in progress atm, but tgall_foo is also trying to get it working at our images [22:31] so then we can simply push a new change to precise [22:31] and get it all working not one day before the release :-) [22:31] yes .. "work in progress" is important to keep in mind [22:32] Feature freeze is also important to keep in mind... [22:32] I'd prefer having any ucm configs in precise by FF. [22:33] The "Work in progress" part bothers me a bit. This is a simple fix. I have it working here out of tree (took 10 minutes). I can easily push it to alsa-lib and alsa-utils and have it working in the next image (assuming the kernel update also lands soon). [22:34] (ive duplicated the success for what its worth) [22:35] GrueMaster: alsa-utils is currently being tested at least for x86 in ~ubuntu-audio-dev/ppa, so what would you prefer be done? Wait till they come out of testing and are uploaded to precise proper, or get them uploaded now, and merged back when the update to 1.0.25 of alsa-utils is done? [22:36] TheMuso: The alsa-utils fix is just a fix for 90-alsa-ucm.rules (and I need to check to see if it even needs fixing). [22:36] The other part is alsa-lib which has the ucm config changes. [22:37] I haven't checked, but I don't think they are upstream. [22:39] GrueMaster: it is being developed upstream but you're right I don't believe it's captured in a current alsa release , wei_feng would know for sure [22:39] * Dr_Who points to #linaro-multimedia tho i don't see feng at the mo [22:39] Dr_Who: I'm updating my alsa-project git trees now and not seeing anything. [22:39] (not counting the drivers). [22:40] I'd recommend talking to Feng, he's the expert [22:59] pbuckley: Looks like the fix is even easier than I thought. No need to change /lib/udev/rules.d/90-alsa-ucm.rules. Just a minor tweek to the ucm configs (sudo sed -i 's/HS/Headset' /usr/alsa/ucm/Panda/*) and it works. [23:00] Plus your kernel patches. [23:02] nice.. hopefully that makes it easier to get into the mainline (right terminology? im new to this whole open source dev thing) [23:02] Yea. [23:04] TheMuso: looking at the ppa, alsa-libs isn't there. No update for it? [23:09] GrueMaster: I accdidentally jumped the gun and its in precise already, may be reverting though as there may be issues with flash of all things... But not sure if the new libc may have had something to do with it or not. Still testing that out here. [23:10] Heh. I just blame flash. [23:10] I'd like to but... [23:10] yea, I know... [23:12] can we all just move to html5 already [23:13] heh. [23:14] I thought the goal was to move to Silverlight? :P [23:15] TheMuso: The flash/libc thing should be fixed in today's upload. [23:15] that was even a joke inside of microsoft (at least the group i was in). (disclaimer, i didn't intentionally get a job there. They acquired a company I was working for) [23:16] TheMuso: It was related to https://bugs.launchpad.net/ubuntu/+source/eglibc/+bug/929713 [23:16] Launchpad bug 929713 in eglibc "audio unreliable with eglibc 2.15" [Critical,Fix released] [23:19] pbuckley: It would be a better joke if it wasn't widely deployed by Netflix. :/ [23:19] yeh i need to bug some of my buddies who went over there to find out the real reason they are using it [23:19] my guess is microsoft gave them a pile of cash [23:20] as per their sop [23:20] I can't get Hulu in Canada so don't know much about it. Is it in the same boat, or did they go with saner tech? [23:21] hulu is flash [23:21] Well, I'm not sure that's "saner", but... It's not Silverlight? [23:22] true.. and they do have a reasonable explanation why they haven't switched to html5. That is if you agree with the whole concept of intellectual property. [23:23] I don't really understand how Flash and Silverlight became video delivery media anyway. [23:23] infinity: Ah ok thanks. [23:23] they were better then the

tags available at the time [23:23] and they have support for drm [23:23] I agree with the concept of IP. I don't agree with the draconian DRM and absurdly high prices. [23:23] We solved the streaming video thing more than a decade ago... And then, due to "if you have a hammer, everything looks like nails" web developers, we started using VECTOR-BASED ANIMATION ENGINES to stream video? [23:23] Brilliant. [23:24] pbuckley: As someone who doesn't live in the US, DRM and I are usually not on good terms. [23:24] infinity: Hey, at least flash isn't doing pixel based rendering on Linux anymore. [23:25] have you tried the x-forward-for thing? [23:25] that gets around alot of the drm right now [23:25] There's absolutely nothing more frustrating than the dozens of links from FB/G+/Twitter/etc to news stories, TV clips, and such that "aren't available in my area". [23:25] (at least for geoip based drm) [23:25] pbuckley: Yeah, I could just proxy. It's not worth the effort. [23:25] the x-forward-for thing is just to prepend something like x-forward-for: [23:26] to the html header [23:26] As in, pretend that I'm proxying for a US client? [23:26] yeh [23:26] Are there ffox plugins to spoof that with minimal fuss? [23:26] i think so [23:26] i had to do it a couple months ago when i was in mexico [23:26] let me see if i can find it in my bookmarks [23:27] I might look into that. An extra 40 bytes on every HTTP request doesn't sound too bad to get around some stupidity. [23:27] https://addons.mozilla.org/en-US/firefox/addon/x-forwarded-for-spoofer/ [23:28] And I'll have to re-pack the xpi. [23:28] Not available for FF10, my ass. [23:28] pbuckley: But thanks. :) [23:28] lol [23:28] np [23:28] pbuckley: I'll fiddle with the manifest later and give it a whirl. [23:28] If this lets me watch Comedy Central clips, I'll be a happy camper. [23:28] Every time someone links to a Daily Show clip, I get progressively more annoyed. [23:29] * pbuckley links to a daily show clip [23:29] * pbuckley hides [23:29] Especially given the irony that, I suspect, Jon Stewart would be more than happy with everyone in the world watching it. [23:42] infinity: This may be your lucky day. I can't reproduce the ksysguardd crash on today's build. [23:55] 0!0@ [23:55] whoops