=== prpplague^2 is now known as prpplague === asac_ is now known as asac [03:59] hm- anyone get adobe flash to work on this yet, or is the work porting the flash from 0xdroid up for grabs [04:13] mouse-_, ti's adobe binary was no help? [04:16] I don't see what the point is, it's not really a product... see the little blurb under TI's flash page, if you wanna ship it anywhere, you have to be qualified by an Adobe partner [04:16] it'll never ever go into a partner repo for example :( [04:16] or restricted on armel [04:17] unless digikey sponsor it.. [04:17] and pay for the oem services at canonical L] [04:18] me neither, it should be a straight download from adobe like x86.. but even their it's bloated, so it's a good thing it's that way.. ;) [04:19] Adobe just don't have the manpower to do it on every little arm board [04:19] it's hard enough building every possible combination on the PC :) [04:20] different radeon and nvidia drivers, intel, amd, windows xp upwards, service packs, browser versions.. [04:20] true... i'm just thinking a generic universal armv7 binary.. ;) but without dsp help it would be too slow.. [04:20] most of flash is software anyway [04:20] they have some fancy neon code in there [04:21] yeap, and ffmpeg plays flash videos fine.. [04:21] the amount of help you get for video codecs is minimal and on ARM they decided to use OpenMAX [04:21] what's left that OpenMAX doesn't do, it's all software.. not custom DSP code, just plain C and some assembler [04:22] it links to OpenGLES2 on iMX515 too, but there is a little contention on whether it is even using it [04:22] someone said it may have been pushed to 10.2 [04:33] persia: I recompiled the netwalker kernel in my lucid image, and nothing has crashed yet. :D [04:48] mouse-_: i've gotten it running using binaries from maemo [04:48] (it = flash) [04:50] bernard_, that was a nice catch, that config change had been in their for almost a year.. ;) [05:00] rcn-ee: heh. i only found it because i'm compiling natively on my beagleboard, and the config change killed the dependency tracking. [05:00] rcn-ee: ti's adobe binary ?? it's inaccessible :P [05:00] (now i can do minor changes, and rebuild my kernel on the beagle in 2 minutes) [05:00] btw, has anyone else gotten any errors like this - [ 116.582672] asoc: interface omap-mcbsp-dai-0 hw params failed [05:01] bernard_: nice! thanks for the pointer [05:01] mouse-_: ack. same happens to me on login, but never after that. [05:01] ahh, yeah [05:02] oo, does the maemo flash use the sgx libs [05:04] mouse-_: doesn't look like it. [05:05] mouse-_: http://pastebin.com/brdmvuTc [05:06] mouse-_: all the libraries are in ubuntu except for 2. [05:06] mouse-_: libplayback-1-0 i copied from maemo also. libhildonfm i just created a dummy library for (it doesn't seem to use it unless it needs a file open/save dialog) [05:11] * mouse-_ nods [05:13] anyone noticed console-setup not fixing up the font? === JaMa|Zzz is now known as JaMa|W === hrw|gone is now known as hrw === zyga is now known as zyga-lunch [11:42] rcn-ee: any chance you could pick up this patch in your 2.6-stable series? http://marc.info/?l=linux-omap&m=128744421925804&w=2 [11:45] (davem's taken it for 2.6.37) [11:57] ogra_ac: just curious, does 'ac' stand for toshiba ac100? [11:57] Mike^, yep [11:57] ogra_ac, and how is it? [11:57] great if you can live without the proprietary bits [11:58] (which do the power management, sound and stuff like that) [11:58] you mean no 3d, no video, no PM :) [11:58] its awesome to have a device i can do arm development on though [11:58] right [11:59] you mean native builds? [12:00] yes, thats what we do in ubuntu [12:03] I'm trying to push Nvidia to at least make the proprietary bits independed of the kernel/X11 version, like they do in x86 [12:04] probably having ac100 users posting such requests on their forum will help... [12:08] Mike^, yeah, DKMS modules and a redistribution license for the daemon stuff would be cool [12:09] Mike^, btw, there is an #ac100 channel [12:10] ogra_ac: it's almost the same people :) [12:13] I'd say ac100 is good example why mainlining android is needed [12:13] ugh [12:13] suihkulokki, why ? [12:14] suihkulokki: mainlining android won't help with ac100 until nvidia do something with the drivers [12:14] ogra_ac: less porting of the drivers? [12:14] suihkulokki, i think ac100 is the *best* example why android is not for netbooks [12:14] no need to port any android drivers for ac100 on ubuntu [12:15] you mean it works out of box with ubuntu kernel? [12:15] we a) have the kernel source from toshiba and b) the community maintained ac100 ubuntu kernel doesnt have a single android module [12:15] suihkulokki, http://ac100.gudinna.com/ [12:16] suihkulokki, the community upstream tree is on gitorious since weeks [12:16] (for the kernel) [12:16] its still a fork [12:16] it is openwrt all over again [12:16] suihkulokki: it is not openwrt [12:16] suihkulokki: with current tegra support in kernel it'll remain a fork for a long time [12:17] its a tegra 2.6.29 kernel with toshiba sauce on top [12:17] indeed its not ideal [12:17] (i dont claim that) [12:18] but it surely doesnt need any android bits [12:18] I was under the impression the drivers have android pm hooks inside [12:18] ogra_ac: I think you can run the tegra tree from the android repos, afaik they already have fbdev added [12:18] and https://launchpad.net/ac100 makes some slow progress already [12:19] Mike^, no LCD support for the ac100 yet [12:19] does it have external display connector? [12:19] yes, but that only works with the nvidia bits [12:20] ogra, wow, you left that bare :D [12:20] it has HDMI [12:20] so, there's a chance :) === zyga-lunch is now known as zyga [12:20] but needs some magically trigger from the nvrm_daemon to switch on [12:20] reverse engineering most of the daemon should be possible [12:21] (since you have full acess to the kernel side) [12:21] right [12:21] ogra_ac: so no display at all? [12:21] most stuff is done by just sending ioctls [12:21] hrw, not with the non toshiba source yet [12:21] someone needs to forward port the sauce [12:25] tomato or bbq ? [12:30] hdmi and lcd [12:33] Sounds awesome [12:34] Does the processor work? [12:34] How about RAM? [12:34] :) [12:35] ? [12:35] i'm typing on an ac100, so yes, ram and cpu work :P [12:41] ;) [12:41] The only things that do by the sounds of it ;) [12:41] nope [12:41] i can show you mine at UDS :) [12:41] * Mike^ looking at git://gitorious.org/ac100/kernel.git/drivers/paz00/lcdhdmi-disp.c [12:42] i use it as my main work machine since a few weeks [12:42] lag, there will be plenty ac100's at UDS btw, i know about at least 6 owners now that will bring theirs [12:43] lag, power management, 3D, HW buttons and local sound (usb works) dont work yet [12:45] 3D is unlikely to happen unless nvidia release the driver in a way we can use DKMS with it ... buttons are possible to get to work through a free daemon app that sends the right ioctls to the kernel [12:45] sound is tricky ... and power management too since thats all handled by a proprietary daemon in userspace [12:46] and in android the PM is mostly done through hotplugging one CPU [12:46] which is horribly laggish [12:48] ogra_ac: seems that everything except 3D and video is possible even without the daemon [12:48] well,i'm not sure about brightness control and cpufreq [12:49] the PMIC bits are in the mainline and what's missing is the arch/arm/mach-tegra/board-ac100.c [12:49] that will tie everithing together [12:50] well, it surely needs somone to work on it ... and preferably someone with more kernel background than me ;) [12:50] i'm fine doing one ot the other ugly hack, but i'm not a kernel dev [12:51] if I'll find a time to send you some patches, will you test them? knowing that it may brick your netbook? ;-) [12:52] brick ? [12:52] i can always revert to another kernel ;) [12:53] playing with PMIC may be dangerous :) [12:53] tried kexec? [12:53] ops, wrong window [12:54] i tried kexec though ;) [12:54] didnt work [12:58] ogra_ac: anyway, I have too many stuff to finish before plumbers, so your ac100 is safe for now :) [12:59] Mike^, i'll be at plumbers, lets play there ;) [12:59] no problem :) === bernard_1 is now known as bernard_ [13:03] sure, bernard_ i'll add that smsc patch, it'll make alot of xM users happy... [13:04] rcn-ee: great. thank you! [14:03] Dear ogra (or ogra_ac), would you have some time to help me with our LP PPA, please? It seems some of the team PPA will not build for armel... [14:17] vstehle, whats the exact prob ? [14:19] ogra_ac, the PPA private-stable and stable won't build for armel. Also, the PPA 'private-stable' does not look very private :) [14:20] ogra_ac, for example http://paste.ubuntu.com/517426/. This is on PPA private-stable [14:23] ogra_ac, for the "private/non-private" part of the issue; private-stable has the following URL: http://ppa.launchpad.net/tiomap-dev/private-stable/ubuntu. This is not a private PPA. It should read something more like https://private-ppa.launchpad.net/tiomap-dev/private-stable/ubuntu [14:24] vstehle, hmm, hard to check ... let me log out of LP [14:25] ogra_ac, good idea. I just did, and I can see private-stable [14:25] vstehle, yes, private-stable seems to have been set up wrongly [14:26] https://edge.launchpad.net/~tiomap-dev only shows public ones beyond that [14:27] ogra_ac, would you please be able to fix it? This PPA can even be deleted and re-created if needed; there is nothing in it. But I think we would like to keep the name, so making the changes from the web interface would probably not work... [14:28] well, trying to find out whats wrong atm [14:30] Oh you are "digging in the guts" of LP? Cool :) [14:32] i deleted it, but recreating a new one with that name doesnt offer me to make it private it seems [14:32] i'll ask in #launchpad i think [14:34] vstehle, wrt your rejected upload, line 8 in the paste is wrong ;) [14:50] vstehle, seems there is no LOSA available in #launchpad atm, i'm waiting for an answer there [14:53] ogra_ac, I think private-stable is now private. At least it disappeared from the webpage... [14:53] vstehle, (btw, you can contact them directly there too ) [14:53] vstehle, no, i deleted it since there were already directories for the failed upload [14:54] we need to get it completely wiped and then the help of a LOSA to make sure the newly created one is completely private [14:54] i couldnt make out the difference through the UI so there must be some admin love involved === plars_ is now known as plars-actual [15:16] GrueMaster: hm, lots of people complaining about lack of sound in beagle xm with ubuntu [15:16] GrueMaster: what is needed to make it work? [15:17] asking you as I remember you tested that :-) [15:17] fix! [15:18] I only got it to work in speaker-test. I can't remember off hand what settings they were. Will look to see if I still have the amixer diff. [15:18] Give me a few minutes as I am just up and waiting for coffee. [15:18] np, will also be out for lunch soon [15:19] * ogra_ac knows a gross hack he wont chare here to make sound work on everything with maverick [15:19] *share [15:20] I do know that it is a regression from Lucid. Nothing in the driver changed from what I can tell, but audio worked fine in Lucid on Beagle, and it is the same audio hw. [15:21] GrueMaster, my gross hack doesnt touch alsa ... [15:21] it only touches udev [15:21] GrueMaster, ls /dev/snd to give one more hint [15:22] I'm not set up yet. Will take a little as I am still reorganizing my office. [15:22] GrueMaster, on any maverick system [15:23] the background is that we dont create any devices under /dev/snd anymore, they go directly into /dev [15:23] that way you cant fall back to alsa setup if you want to [15:25] GrueMaster, http://paste.ubuntu.com/517475/ even makes the soundblaster play work on the ac100 [15:25] but as i said, gross hack [15:26] (likely breaks autodetection etc in pulse) [15:26] so, someone modified udev to a new model without testing to see if it broke the old model ? [15:26] might be [15:26] i'd rather think that someone considers the old model deprecated [15:30] hm, weird [15:32] vstehle, so i talked to LOSA in #launchpad ... we will need a new PPA (which means new name) that has to be made private manually by a LOSA before anyone uploads to it [15:33] vstehle, once uploaded to it it seems to not be possible to convert it [15:33] nor to completely delete it [15:34] vstehle, could you discuss a new name with ndec ? [15:34] and tell me the name [15:36] Ok ogra_ac. Having deleted PPA on LP myself, I was sure it would end this way :) [15:38] * rsalveti lunch [15:48] rsalveti: Do we know if the Beagle/XM supports S-Video out? I have no way of testing this. Currently reading an email sent to ubuntu-mobile mailing list. === mturquet1e is now known as mturquette [16:35] lool: I had a chat with hrw on xdeb [16:35] lool: trying to setup an environment [16:35] berco: better take it to #linaro ;) [16:36] hrw: I couldn't find lool on #linaro :) [16:36] I see him there [16:37] hrw: yeah, I see him now, don't if I opened my eyes :) [16:37] hrw, juat because *you* see him there :P [16:38] *just [16:38] ogra_ac: I do not want to bother arm team with boring cross compilation issues ;D [16:39] hrw, i meant your lool radar ;) [16:40] (and we dont mind xdeb discussions here) [16:50] ndec, there is a new alsa-utils in the queue with fixed postinst script that calls alsactl init === hrw is now known as hrw|gone [17:05] GrueMaster: it should support, but I believe nobody tested it [17:11] I have an old (near ancient) TurboTV PCI card that has coax, svideo, and rca jacks. Not sure if the svideo & rca are input, but I think they are. I guess I could turn my serial console monitor into an svideo capture display. [17:12] hehe, I believe this can work, if your card is still working :-) [17:20] As soon as it finishes upgrading to Mav, I'll give it a whirl. [17:38] Heh. So much for that idea. Case is ITX. Only takes half height PCI. [17:38] Hrm [17:40] bah :( [17:40] rsalveti, about bug #663642 [17:40] Launchpad bug 663642 in linux-linaro (Ubuntu) (and 1 other project) "DVI doesn't work at BeagleBoard xM rev A3 (affects: 1) (heat: 12)" [Undecided,New] https://launchpad.net/bugs/663642 [17:41] jcrigby: yup [17:41] just sent the patch to the kernel team [17:41] what does this mean: This patch is not yet upstream because it depends on the xM check [17:41] patches that are still not upstream. [17:41] jcrigby: upstream is lacking code that checks when it's running on xm and when it's not [17:41] if you compare the beagleboard board file you will easily notice it [17:42] ok [17:42] I'm trying to get this into the linaro kernel today so we can get it through before uds [17:44] jcrigby: our kernel: http://kernel.ubuntu.com/git?p=ubuntu/ubuntu-maverick.git;a=blob;f=arch/arm/mach-omap2/board-omap3beagle.c;h=94cb19c07eede995039035a63e5ac53d96e64b7f;hb=HEAD#l282 [17:44] l-o: http://git.kernel.org/?p=linux/kernel/git/tmlind/linux-omap-2.6.git;a=blob;f=arch/arm/mach-omap2/board-omap3beagle.c;h=14f42240ae792bf01fcd289d60d441d9d55eb328;hb=HEAD#l273 [17:45] jcrigby: it's basically the check when the kernel is running on cpu_is_omap363 [17:45] *cpu_is_omap3630 [17:45] hey arm guys :D [17:46] jcrigby: do you know if anyone from linaro has a xM rev A3? [17:46] at least one guy from the beagle ml said it worked for him [17:46] question: is there anything installed on the gnome desktop by default, besides the games and useless stuff like brasero, that you'd consider aren't particularly useful to 99% of users? [17:46] not including language support stuff [17:46] I am just trying to cut about 100MB out of the filesystem if I can [17:46] I only tested with my current rev A, and it worked as before [17:48] rsalveti, I don't know [17:52] jcrigby: as most newer xMs are rev A3 now, this is kind of a high priority bug [17:52] otherwise the user is unable to use the dvi output [17:54] rsalveti, I understand, thats why we want to get it into the linaro kernel [18:39] plars-actual, so you are not the fake plars ? [18:39] ogra: no, plars is clearly the fake! [18:40] now, how can you prove that to us ? [18:40] oh, does that not just mean he's the officer in charge? [18:40] ogra: working from a different location today, my firewall is having fits and not keeping it's config lately, so I can't connect to my bip proxy :( [18:40] oh, thats bad :/ [18:41] slangasek, hey, good that i see you ... does your team care about multivers breakage ? [18:41] could be due to the weekly, sudden power outages I was having for about a month [18:41] ogra_ac: "breakage"? :) [18:41] i just ran into a bug trying to play an mp4 movie [18:41] ogra_ac: I don't believe we're touching anything near multiverse currently, so it's ok if you break it - or are you asking us to fix it? :) [18:41] which works fine until i install ubuntu-restricted-extras [18:42] hmm, ok [18:42] file a bug, subscribe linaro-foundations please [18:42] seems there is a libavcodec52-extras package that replaces the installed one [18:42] which seems broken [18:42] ok, will do [18:43] hmm === Mike^ is now known as Guest60685 [18:43] apparently the gstreamer equivalent is broken too [18:43] weird [18:44] mplayer plays the file, totem complains not having support for mp4 [18:47] mplayer uses different libraries than totem. [18:47] GrueMaster, right, but i see the same breakage in both [18:48] * ogra_ac removes gst-plugins-bad-multivers [18:48] e [18:48] i wonder if there is an x86 binary somewhere in the multiverse codecs [18:49] http://www.dailypspmovie.com/movies/2010/20101021-never-say-no-to-panda.mp4 [18:50] plays fine with the defaults [18:50] but stops playing in anything after ubuntu-restricted-extras was installed [18:54] file a bug. Maybe we can get an armel developer to look at it. :P [18:54] do you know any ? [18:54] Good ones? [18:55] * ogra_ac cant get back into the state where totem played the file [18:55] there are good ones ?!? [18:57] Heh. On alsa-devel mailing list: A new version of the funny game "Catch-the-slider" aka Alsamixer-Qt4 0.5.1 is now available... [18:57] giggle [19:23] ogra_ac: http://qt.gitorious.org/~dfaure/qt/dfaures-clone/commit/ebf9d5dd5174b7c82fec83b56d71a59d5277bd51 [19:23] Is this in our version? [19:25] no idea [19:31] Ok, no it isn't. I'm looking at the source file and comparing with the patch above. [19:31] btw: my dove crashed during the build. [19:32] not enough swap ? [19:32] you need a gig or more afaik [19:33] I have 2.5G [19:33] then it should have worked [19:33] but well [19:33] its dove [19:33] use the panda ;) [19:33] heh [19:34] That will require a bit of extra work. ENOHOOKEDUP [19:34] well [19:35] i uploaded the fix to proposed [19:35] just let the buildds suffer ;) [19:35] which fix? [19:35] -no-neon [19:35] if we are lucky its ready before UDS and i can test with mumble on the ac100 [19:35] Ah. I will try to explore this patch more. Might be a better fix. [19:36] i just set the configure option, threw the package at the wall and wait ... [19:36] If it enables proper runtime detection. [19:36] it doesnt [19:37] check if it opens /proc/self/auxv somewhere [19:37] then it uses hwcaps [19:38] which is what is needed for doing runtime detection properly [19:39] http://www.llvm.org/bugs/attachment.cgi?id=4428 is a good example patch [19:57] Did you break alsa-utils? I'm seeing a lot of bug reports on it failing to install. [20:04] from maveirck-proposed ? [20:07] * ogra_ac doesnt think his last change was even accepted [20:10] * GrueMaster had 17 emails of new bug reports in the last 20 minutes related to this. [20:11] Here come 14 more. [20:12] can you probably point me to one ? [20:12] * ogra_ac didnt get any alsa mails [20:12] This is the original bug. Bug 664645. [20:12] Launchpad bug 664645 in alsa-utils (Ubuntu) "package alsa-utils 1.0.23-2ubuntu3.3 failed to install/upgrade: subprocess installed post-installation script returned error exit status 1 (affects: 14) (dups: 12) (heat: 106)" [High,Confirmed] https://launchpad.net/bugs/664645 [20:12] Every new one is being auto-tagged as a duplicate. [20:15] yeah, i see the issue, damned [20:17] copying code from the initscript wasnt so clever [20:17] bad ogra_ac. [20:18] * ogra_ac curses [20:18] i was supposed to be afk 2h ago [20:21] ogra_ac: sit and fix instead :( [20:21] yes [20:29] another issue on the sound thing? [20:29] ogra_ac: you're just with a lot of bad luck [20:30] rsalveti, nah, to blind [20:30] slangasek, could i ask you a favor and let the alsa-utils package i just uploaded to maverick-proposed in so the alsa guys dont drown in bugs ? [20:30] slangasek, seeing you on the SRU team [20:31] (seems pitti is gone for the day) [21:01] Interesting. Just added the omap ppa to kubuntu image and did apt-get install on ubuntu-omap4-extras-graphics. It recommended linux-headers-2.6-amd64. [21:02] Somethings not right here. [21:03] just recommend [21:03] because of dkms, probably [21:05] Just seems odd. [21:06] yeah, as having a *lot* of video drivers installed by default [21:07] for sure I'm not going to run s3virge, trident or voodoo at my panda [21:07] no, thats valid [21:07] for x86 [21:07] :) [21:07] on arm we should fix that at some point [21:07] it'd be good to clean that [21:07] probably worth a lst minute spec [21:08] though david isnt there to approve it in time [21:08] I'll head it up. Oh, wait... [21:10] lets talk about it in the call [21:11] oh, we have a call today, almost forgot about that [21:11] david asked me to run one [21:12] not sure we have many topics [21:12] but as long as NCommander shows up to entertain us with echos ... [21:13] lol [21:20] he is on vacation too. Not sure he'll make it. [21:21] oh, i didnt know [21:24] I'll find out. [21:25] Nope, he has no network access. [21:25] k [21:27] Well, this is certainly interesting. After installing video drivers from ppa & rebooting, I am getting a good display, albeit at 640x480. Probably due to my HDMI switch. [21:28] you and your toys [21:28] heh. [21:29] Yea, it was the switch. It automatically switches to the next device when signal is lost, and switches back when the signal returns. Of course we have already pulled edid at that time. === JaMa is now known as JaMa|Zzz [23:13] ogra_ac: alsa-utils accepted [23:13] * ogra hugs slangasek [23:13] ogra_ac: which should correct the -proposed regression; but what's the reason for this being a panda-specific hack in the first place? [23:13] i owe you a beer on sunday [23:14] slangasek, the alsa driver for the SDP4430 and Panda doesnt init properly ... it needs a one shot initialization once [23:14] that needs to be fixed in ASoC at some point but we didnt get a better fix yet [23:15] (as many other bits need fixing in ASoC for omap4, teh driver really isnt 100% there yet) [23:15] ok [23:15] so this is a not-for-natty change? [23:16] well, it is for natty until we get a fixed driver [23:16] i dont know how long that will take upstream [23:16] for natty there is code that doesnt need the postinst [23:16] the postinst is only needed on systems that already have alsa.state in place [23:17] ok