=== ka6sox is now known as ka6sox-away === Amaranth_ is now known as Amaranth === asac_ is now known as asac === ka6sox-away is now known as ka6sox [05:48] what sort of voodoo do I have to do to allow a serial login for maverick on a beagleboard (rev C3)? [05:53] oh,remove "quiet splash"? I've tried adding "console=ttyS2,115200n8 serialtty=ttyS2" [06:14] Kamondelious, You'll need to have some upstart job that generates a getty on ttyS2. jasper-initramfs should generate one on first boot if you're using that sort of image, and passed the console= on first boot. If not, you can probably get one with a quick copy&edit of e.g. /etc/init/tty5.conf [06:18] persia, ahhh ok thanks [06:27] and I'm in, awesome, thanks again persia [06:30] There may be a few other things you have to adjust in a similar vein: maverick is really focused on the desktop/laptop use case. There's a few things in progress for Natty that ought make it simpler (like automatic provision of getty on serial consoles) [06:33] like how the kernel can't find the driver for my usb wifi dongle now? [06:34] No, that's likely a general problem. I doubt the kernel has that driver for any environment/architecture. Please file a bug with `ubuntu-bug linux` and report as much about the dongle as you can. [06:34] it's been working fine for the past couple weeks [06:35] =( [06:35] On an Ubuntu system? [06:35] yeah [06:36] Very odd. That really oughtn't happen. Did you recently get a kernel upgrade? [06:36] not unless it did it by itself [06:37] You would have at least been advised (needs your password). [06:38] I did an upgrade a few days ago, but the system's been rebooted a few times since then [06:39] It ought to have shown up the first time, if it was a regression. [06:39] indeed [06:39] kernel hasn't changed [06:44] hmm [06:45] now the wifi dongle doesn't even show up under lsusb [06:45] that's really odd [06:45] the other devices are there, maybe the wifi dongle died [06:45] that would suck [06:46] Maybe a power thing? How is it attached? [06:47] powered usb hub [06:47] it's been working great for the last couple weeks [06:48] Does a plug/unplug show anything in syslog? [06:49] good question [06:51] but I can't test that tonight, it'll require too much dis-assembly [06:51] Well, if you get messages, you probably ended up with it hibernating unexpectedly. If not, you may want to try in another environment, but I suspect it's gone. [06:52] indeed [06:52] thanks again for your help === hrw|gone is now known as hrw [08:52] What is consolekit? It seems its a system for managing permissions and such for logged in users. Am I right? [08:53] I'm building an embedded system which will run exclusively off daemons/services. Can I be without it? It seems to me that consolekit pops up everywhere... [08:57] sveinse: well i've got an embedded system without consolekit.. more than one even, i think. depends on what distribution you are running and what it's supposed to do. [08:58] LetoThe2nd: maverick [08:58] sveinse: mine don't even run ubuntu :P [08:59] calling a machine running ubuntu embedded feels rather strange to me anyway ;-) [08:59] My problem is pulseaudio ATM. Sound worked before the weekend, now I keep getting connection refused [09:00] LetoThe2nd: Well, that's the choice we made. We found it better to use maverick over OE, since it offers a system more similar to "ordinary"/host linux systems [09:06] sveinse: choose whatever you like, it's your project ;-) i just wanted to point out that i personally don't consider ubuntu being something worth calling embedded, for it brings all that stuff that i don't want to see there or that wouldn't fit into flash anyway (like all that funky gizmo-stuff that you are at the moment struggling with) :-) [09:06] sveinse: therefore - sorry, can't help you with that one. [09:18] * hrw will seek for sponsor to get flash-kernel with efika support into archive [09:19] writing patch and waiting for merge is just waste of time [09:56] ogra_: flash-kernel is still somewhere on a list? [09:56] hrw, indeed [09:57] i'll try to get to it this week [09:57] thx [09:57] i think i'll skip the merge and just apply the waiting patches though [09:57] I have to dig patch from lp again... [09:57] there is nothing overly important in debian we need [09:57] i am subscribed to all the bugs [09:59] I want to upgrade kernel on efika - flash-kernel started... === ka6sox is now known as ka6sox-away === ogra_ is now known as ogra === ogra is now known as Guest94419 === Guest94419 is now known as ogra_ [11:03] hrw, are you going to the emdebian sprint ? [11:04] ogra_: yes, Cambridge, UK one [11:05] i'm looking for a hotel to stay [11:05] i dont see any options on the wiki [11:05] holiday inn express [11:05] got a url or so ? i'll book through the travel agent [11:06] http://mapy.google.pl/local_url?q=http://www.expresscambridge.co.uk/&dq=Holiday+Inn+Express&f=q&source=s_q&hl=pl&aq=3&sll=52.214128,0.144882&sspn=0.173124,0.260925&ie=UTF8&rq=1&ev=zi&split=1&z=12&jsv=310c&mpnum=1001&radius=6.64&vps=3&output=js&oi=miw&sa=X&ct=miw_link&cd=1&cad=homepage,cid:18353814746422007578&ei=lpdGTZilJpH6Oaav7O8E&s=ANYYN7k1RadTteoQeE8HTGY63MdSQHmyCQ [11:06] auch [11:06] ogra_: I booked thought agency [11:06] well, works :) [11:06] ah, cool [11:07] so they know it then, thanks [11:07] no problem [11:08] this hotel is nearest to ARM [11:08] and its ~15-20minutes walk [11:08] good, thought far from downtown then i guess [11:08] i.e. dinner options are likely not great [11:08] oops. google said 7 minutes even [11:09] ogra_: it is near [11:09] you just need to know shortcut ;D [11:09] heh, k [11:09] i'll ask you then [11:09] ogra_: use satellite view [11:09] ogra_: there is a bridge over railway [11:09] and long sidewalk near the water [11:10] i guess i'll find my way around, first the booking :) [11:10] ;) [11:11] ogra_: so you go to represent ubuntu/arm and me same for linaro [11:11] likely [11:11] i'm replacing david [11:12] ok [11:12] I land at 11:40 in luton, then bus to cambridge [11:12] on sunday [11:12] back on saturday 11:20 from STN [11:12] hmm, i will likely just get a flight to heathrow [11:13] longer trip you will have then [11:13] STN has best connection to cambridge [11:13] yep [11:13] direct train [11:14] but no good lufthansa connection to FRA [11:14] use cheaper airline? [11:14] cheaper ? [11:14] LH flies me to LHR for 98€ roundtrip [11:15] even with train thats a cheapo [11:16] ;) [11:16] you have good local airport ;D [11:17] ~curse me [11:17] hotel -> arm is 30-35 minutes by walk. 7 by car :D [11:17] anyway near enough [11:17] no, when i lived in cologne i had a good local airport [11:18] it's germanwings home base [11:18] ok [11:18] they offer CGN->STN for 5€ [11:18] ogra_: http://wiki.openembedded.org/index.php/Oedem/2009#Travel will be useful for you [11:19] @ is €? [11:19] switch to de_DE.UTF-8 finally? [11:20] hmm, i thought i had utf-8 everywhere [11:21] ;D [11:22] € [11:22] does it work now ? [11:23] yes [11:23] AHA [11:23] oops [11:23] sorry, caps [11:25] dunno why there was IRC preset as charset by default === amitk is now known as amitk-afk === amitk-afk is now known as amitk === rsalveti` is now known as rsalveti === ogra_ is now known as ogra [12:49] ogra: so I poked cjwatson, and thought it over. I agree that your right with a new target and not reusing base (which also lets us break out a new seed easily if we need it). I just think we should call it something else beside 'headless' [12:49] we could extend it to headless-chicken ;) [12:49] make a name suggestion, i'm fine with anything thats descriptive [12:50] its janimo's spec he should make the final call [12:50] ogra: probably 'ubuntu-cli'. The reason I don't like headless is it could be confused with netboot or LTSP stuff [12:51] i dont see where that would clash name widse [12:51] *wise [12:53] ltsp doesnt have any headless option or anything that would even sound remotely like headless [12:53] i would probably call it ubuntu-minimal-server or so ... but that would clash even more than headless [12:55] the name i'd like most would be ubuntu-core ... but that would clash with a product description mdz works on === ian_brasil__ is now known as ian_brasil [13:03] NCommander, any name is fine, headless is used by linaro for their current dev image though [13:04] NCommander, I am wondering whether preinstalled should be in the name for further clarification [13:04] * NCommander mutters something about the hardest part of implementing a spec is naming the name things [13:04] as for cli wouldn't that be confused with Mono (CLI) centric build? No need to be overly tesre in naming these images [13:05] NCommander, indeed, that is why I said I am fien with any names but want to get it done, and then polish (naming, etc) [13:05] personally, both ogra and I wantedto call preinstalled images 'oem', but we nuked that for obvious reasons [13:05] indeed [13:05] I thikn headless-preinstalled is pretty descriptive [13:06] * rsalveti agrees [13:06] adding preinstalled will add duplication, preinstalled is in the patch already [13:06] *path [13:06] hm, ok then [13:06] damned finger memory [13:06] well but in project names at least in debian-cd [13:06] ubuntu-mid, toc, kubuntu-mobile , ubuntu-netbook are quite nondescriptive imo [13:07] at least this would convey it is preinstalled and not cd and that it is cmd line only [13:07] anyway I'll continue on it today after pathc piloting is over [13:09] imho ubuntu-core would be great, but that would need signoff from mdz [13:09] (since he refers to -core with product stuff) [13:11] core is vague [13:11] so it minimal and standard. headless is clearer imho as it actually says omething about a techincal characteristic - the other can be interpreded very differently [13:32] * NCommander would actually would have liked to call it 'base' but that was already taken ... [13:39] rsalveti, yo ... how are your CDs ... the .38 kernel seems to be in the i386/amd64 kernel and they are looking good [13:42] apw: our latest image is still from friday, so no images to test atm [13:42] apw: but let me try the latest deb available [13:43] Finished on 2011-01-29 (took 19 hours, 36 minutes, 38.4 seconds) [13:43] wow [13:43] rsalveti, ? [13:43] what took 19 hours? [13:43] the latest kernel for armel [13:43] https://launchpad.net/ubuntu/+source/linux/2.6.38-1.28/+buildjob/2228976 [13:43] rsalveti, oh yeah ... thats why we whine when you want new flavours [13:46] rsalveti, thats why i wanted the linaro kernel ... got tired of the whining ;) [13:46] :-) [13:47] apw, we're waiting for mvo to fix update-manager before we can attempt a new image build [13:47] he's on it though, should be fixed today [13:49] and GrueMaster is also getting his new xm today probably [13:49] ogra, heh he is cutting it fine for the freeze :) [13:49] well ... [13:50] having images before the freeze really helps fixing issues :) [13:50] rsalveti, so i am assuming this kernel will be as good for you as the ones we hand made and tested late last week... in which case we are planning on this being the alpha-2 kernenl [13:50] apw: yeah, should be fine [14:16] is there a userspace usb ethernet gadget driver for armel? the kernel level driver for my platform is horribly broken === rbelem__ is now known as rbelem === zumbi_ is now known as zumbi === davidm_ is now known as dm_tab === zyga is now known as zyga-afk [16:43] apw: 2.6.38-1-omap seems to be working fine, the only annoying thing is the call trace for the erratum, that will probably be fixed later [16:43] once upstream gets the changes [16:46] rsalveti, yeah, nothing major then good :) [16:46] yup :-) [17:02] ogra: rsalveti: alsa-1.0.24 was released today. http://www.alsa-project.org/main/index.php/Changes_v1.0.23_v1.0.24 [17:02] UCM is in. [17:02] cool [17:50] GrueMaster: will that make Natty ? [17:52] TheUni: I am hoping it will get pulled in after A2 this week (too late for A2). [17:53] ok [17:53] I'll hunt down our alsa guy and see. [17:54] have they figured out a sane way for applications to detect available cards yet? [17:55] hmm, seems that what ucm is? [17:59] I think UCM is more for a sane way to configure the hardware than through the drivers directly. For example, the Beagle & Panda boards have audio codecs that contain a lot more connections than are available to the user. UCM allows you to map the correct ports for the board in userspace so apps (pulseaudio) will just work. [18:07] GrueMaster: seems there's a serious lack of information. Alsa has provided a changelog and that's all? I have to hope more is coming [18:10] http://www.slimlogic.co.uk/?p=40 [18:10] TheUni, ^^^ [18:12] It's mostly in alsa-lib. UCM is mainly an API interface. Very little is needed at the driver level. [18:13] There is also a utility in alsa-utils, but apps can interface directly with the ucm api in alsa-libs. === ka6sox-away is now known as ka6sox [18:46] ubuntu maverick on beagleboard c4: the problem when opens webrowser the system is freezing... i've got http://rcn-ee.net/deb/rootfs/maverick/ubuntu-10.10-r3-minimal-armel.tar.7z an then done sudo apt-get install xfce4 gdm xubuntu-gdm-theme xubuntu-artwork xserver-xorg-video-omap3 [18:47] canonical image hasn't that problem - browser is working well here [18:49] sonny80, anything in dmesg during the freeze? are you using swaP? [18:50] yes i add swap [18:50] what web browser? [18:50] for ex. when watching google maps in firefiox or chrome [18:51] did you run the script "get_chrome.sh" in /boot/uboot/tools/ (for me it seem's faster then ubuntu's firefox/chrome in the repo's..) [18:52] i didnt ran this script [18:52] i think that problem not in the browser [18:53] cause canonical image working good [18:53] well, the only difference is the kernel. [18:53] yes... can you fix the problem? :) [18:55] that's why i roll back to angstrom [18:56] sonny80, no idea what the problem actually is, since your 'report' is too limited... i'd say use the "get_chrome.sh" script, compare with the builtin firefox/chrome.. [18:56] then we can go from theire [18:56] midori had the same problem... [18:57] ok what step i need to do to help you with testing? [18:57] run the script... then go to engadget.com, is it smooth? [18:58] (on my bx/c4/xm running maverick it is..) [18:59] ok i need to reinstall ubuntu instead angstrom it will be tommorow [18:59] ah, figured you still had it installed.. [19:00] on angstrom it is really smooth [19:00] next time you should keep the sd card as is, it's impossible to troubleshoot anything... [19:01] angstrom is much lighter embedded os.. ubuntu is getting there.. [19:01] can i send you my SD image with troble? [19:02] why? it sounds like you were using my base maverick image with xfce installed.. [19:02] yep [19:02] i also try lxde, icewm [19:02] i'm using the same image today, debugging dspbridge and sgx stuff.. [19:02] the same difficulties [19:03] lxde pulles in too much gnome last i checked, that'll bog down the system too.. [19:03] xfce is the best choice? [19:03] sofar in my exeperience yes.. keep it simple the less lib's the easier it is for the beagle.. [19:04] ok... another question - is it possible to have 24 bit color depth on beagleboard c4 via DVI out? [19:05] 16 bit is looking not perfect [19:05] sure, but it'll take more cpu cycles... [19:05] the neon unit is currently doing the work.. [19:06] ok... i heard that SGX cannot work with more than 16 bit is it true? [19:07] doesn't matter, for 2d graphics the SGX on the omap3 is actually slower then the neon.. [19:07] hmm... interesting [19:08] ok tommorow i will give you detailed information of my problem [19:09] other thing to check, speed of sd card on the beagle.. "sudo hdparm -tT /dev/mmcblk0p2" [19:09] ok [19:09] wget http://rcn-ee.homeip.net:81/dl/updates/omap-image-builder/tools/setup_sdcard.sh is it neccesarry to run it? [19:10] are you building the image manually then? [19:11] (the script set's up the correct clock speed, in this case your c4 might only be running at 500Mhz vs 720..) [19:11] i repeat step by step instructions from http://elinux.org/BeagleBoardUbuntu [19:11] that's fine, the manual step by step is the same as the script.. [19:12] thank you! tomorrow will be next questions [19:22] rsalveti: Does xloader/uboot for omap4 support the different sysboot modes on the panda? Was wondering if I could configure a different boot mode for experimenting. [19:26] GrueMaster: that depends on what you'd like to do [19:26] and how are you planning to boot your panda [19:27] I was hoping that I could configure mode 1101 (UART/MMC1). Right now it is configured for 101 (USB/MMC1). [19:29] GrueMaster: you can, but then you need to write the x-loader and u-boot with uart [19:29] (would be easier if the board used 0ohm resistors). [19:29] like we have for beagle [19:30] I believe someone already did the work to boot from usb [19:30] using micro usb [19:30] Ok, so not currently supported. That's what I needed to know. [19:33] well, probably need some dev to make it work [19:33] but possible, for sure [19:34] I'll explore it later, after A2. [19:34] and why uart? guess usb should also work fine for you [19:35] automation. [21:09] is anyone familiar with rootstock? i try to build a xubuntu rootfs but it always stalls setting up xulrunner... [21:40] nicofs: I have seen this also, not sure why it happens though [21:48] nicofs, on x86 right? it's qemu crashing.. [21:49] rcn-ee_at_work, but why is it doing that to me?!? [21:49] it does it to everyone on x86 since lucid... run it on your target arm board and it'll work fine. .;) [21:55] rcn-ee_at_work: lol, and it will take 100 times longer most likely :D [21:55] I'm guessing it is of no diff on x64? [21:55] actually my igepv2 is faster then my quad 64bit on building rootstock.. ;) [21:56] haha [21:57] it's native armv7a vs emulated armv7a.. .;) [21:57] oh right [21:57] and with thumb stuff thrown in for extra fun.. ;) [21:57] I thought rootstock just downloaded and installed, didn't realize it had to actually do any arm stuff [21:58] rcn-ee_at_work, if my target arm board had any internet - no problem [22:33] Hi guys, I followed the steps here exactly to boot Ubuntu on Panda; https://wiki.ubuntu.com/ARM/OMAPMaverickInstall, but when it boots, I get stuck at "Enabling additional executable binary formats binfmt-support" on serial terminal, nothing over DVI on monitor though. Any ideas? [22:48] Hi ! :) I'm on this page right now : https://wiki.ubuntu.com/ARM/DeviceSupport and wondering what you guy would recommend for an arm notebook. There was a lot of stuff presented in CES 2011 so perhaps there are some models you heard of that run ubuntu Flawlessly ? I'm always on the move and I'd like to be able to write code for hours without charging power :) [22:52] Aureusz, I've seen folk walking around with Ubuntu on the Sharp Netwalker, Genesi Efika MX, and Toshiba ac100. For all of these, it seems some tweaking of kernels is required. [22:54] I dont mind tweaking kernel, but honsetly, if theres a model that handles things perfectly with an arm ubuntu image... it's better ;p [23:01] anyone know why I'd get stuck at "Enabling additional executable binary formats binfmt-support" on 10.10 boot on Panda? [23:08] Aureusz, There's talk about having some for Natty, but I haven't seen anyone commit to one yet. [23:15] persia, thanks for you answers, I read also that some machines are coming. most would be tablets I think, interesting anyway :) gn [23:17] or bookreaders [23:17] ka6sox, What's the difference between a bookreader and a tablet? [23:18] persia, what the manuf loads on them. [23:18] And if we consider that the goal is to ignore that, and install Ubuntu on everything? [23:18] otherwise...from a hardware standpoint? (well at least the Nook) [23:18] not a bit of difference [23:19] persia, I"m moving my nook color from Android to Ubuntu....slowly because I'm so busy with the FOSS project I work on :( [23:19] I suppose one might make a bookreader that was a convertible or a wearable or some such, but yeah, my thought is that "tablet" is a device category and "bookreader" is an application. [23:20] persia, right... [23:20] ka6sox, Your earlier work on that inspired me to go look up the specs, but it didn't seem to be quite enough, for me (but then, I'm generally not a tablet person) [23:20] since I can probably get a kindle or nook app that runs on most things. [23:20] persia, its CHEAP....and works well. [23:21] its not a 10" device however. [23:21] ka6sox, Indeed, but it has to be *better* than my Netwalker to replace it in my pocket. [23:21] true [23:21] And it's close, but not quite better (but lots cheaper, and better as a new purchase for folk who like tablets) [23:21] for now I was looking @ how to do porting...especially of Android devices to Linux. [23:22] Well, the quick'n'dirty way is to just boot the Android kernel and an Ubuntu userspace. [23:22] But one usually finds one needs to tweak the kernel a bit afterwards. [23:23] thats my first approximation...if I ever get time to make this stupid initramfs. [23:23] The ideal way is to add any necessary hardware support patches to the Ubuntu kernel, and then have a kernel flavour that can be used with Ubuntu on the device. [23:23] not sure if the patches are mainline yet...and BN didn't separate out "patches" it was a source dump off a Windows Machine. [23:26] That's unfortunately common. The monolithic patch is usually extracted by diffing against mainline for the reported version, and then one breaks down the components by inspection, and tries to port them to current trunk. [23:49] persia, I'd LOVE to do that if we could figure out what they started with.