**** BEGIN LOGGING AT Wed Jan 09 02:59:59 2013 Jan 09 05:38:00 is there an easy way to change the uboot splash screen with two penguins to something else? Must I recompile uboot to make that work? Jan 09 05:49:10 xenome: That's not uboot, it's the kernel. Jan 09 05:51:28 ah, is there a way to change that w/o having to rebuild the kernel? Jan 09 05:51:40 does that get pulled from the initramfs? Jan 09 05:58:25 xenome: it is part of the kernel Jan 09 05:59:24 and your options are, leave as is, or turn them off Jan 09 05:59:39 how can I turn them off? quiet? Jan 09 05:59:54 recompile the kernel Jan 09 06:00:06 oh, ok so no boot option Jan 09 06:00:08 bummer Jan 09 06:00:35 if I'm going to recompile, should I get the ubuntu source for my kernel or get adventurous and try mainline Jan 09 06:00:52 i suppose I'm in a pickle because I need to use the TI kernel for maximum driver support Jan 09 06:50:44 xenome: We actually remove the penguins on our master kernels, I believe. You could file a bug asking for that patch to be included in linux-ti-omap4 as well, if you're using Ubuntu's omap4 kernels. Jan 09 06:51:16 (At least, I haven't seen a penguin on my x86 framebuffer in years, so I assume we're disabling it... Or we just draw over over it so fast I don't notice) Jan 09 07:02:17 Hi, all! I want to ask you about installing ubuntu on the Nexus 7.. Can I install there my own image of ubuntu? Without the installation progress. Jan 09 07:09:33 Does anyone know what is the full list of packages that contain all the 3d drivers for the nexus 7? Jan 09 07:18:55 nvidia-tegra3? Jan 09 09:44:05 what should contain image file for nexus 7? rootfs? Jan 09 13:36:50 hello everyone Jan 09 13:38:46 I have installed ubuntu on my Nexus 7 32 GB. But it shows that disk space for "/" is only 6 GB. Have I installed the wrong image? I have used the Desktop Installer to install the image automatically to my Nexus 7. Jan 09 13:39:22 no, you havent, we dont offer specific images for the differently sized devices yet Jan 09 13:39:40 so currently the image for 8G is the one used on all devices Jan 09 13:41:41 ogra_: oh. ok. is there any way to use the remaining space? maybe by creating a new partition? Jan 09 13:44:19 you could repackage the image before flashing, but there is a high risk that fastboot will corrupt it while flashing Jan 09 13:46:32 xnox, ogra_ - is there a bug number for ubiquity not accepting text entry on the user details screen? Jan 09 13:46:45 xnox, bug 1093050 Jan 09 13:46:46 ogra_: can't you repartition emmc? Jan 09 13:46:47 Launchpad bug 1093050 in ubuntu-nexus7 "OnBoard doesn't work on text boxes during initial setup" [Medium,Incomplete] https://launchpad.net/bugs/1093050 Jan 09 13:46:52 hrw, nope Jan 09 13:47:22 ogra_: but that's still emmc? Jan 09 13:47:24 hrw, the GPT is hardcoded in the bootloader, the bootloader is a signed blob we cant replace Jan 09 13:47:31 o fsck Jan 09 13:47:50 ogra_: fsresize possible? Jan 09 13:48:05 you can resize them, but then you get massive FS errors Jan 09 13:48:15 ogra_, for me onboard appears but typing does nothing Jan 09 13:48:15 i tried that when rolling the first images Jan 09 13:48:27 brendand, right, a reboot should fix that Jan 09 13:48:49 there is some kind of race on the first boot or so Jan 09 13:49:01 which doesnt seem to show up on subsequent boots into the installer Jan 09 13:49:48 ogra_, ah strange Jan 09 13:50:11 well, metacity worked fine, the bug started showing with the switch to compiz Jan 09 13:50:50 lovely compiz Jan 09 13:58:56 xnox, seen that ? https://bugzilla.redhat.com/show_bug.cgi?id=859373 ... http://lists.fedoraproject.org/pipermail/package-announce/2012-November/090998.html ? Jan 09 13:58:59 bugzilla.redhat.com bug 859373 in ntfs-3g "Danger in dual booting Windows 8 and Linux" [High,Closed: errata] Jan 09 13:59:17 seems ntfs3g needs a fix Jan 09 13:59:43 ogra_: ok thanks Jan 09 14:01:50 btw, i did experience the same problem with ubiquity nt accepting text entry in the oem setup screen Jan 09 14:01:54 ogra_: yes, we did add a similar erratra for wubi & quantal. I think we might be recent enough as we do bail out at mounting windows8 (hence the complaints) Jan 09 14:02:03 i was able to reproduce the error Jan 09 14:02:25 xnox, ah, k, i just saw a news article from today that claimed we dont have the fix in ubuntu and debian Jan 09 14:02:39 ogra_: not sure now, I'll double check and cherry pick as well if we are missing that bit. Jan 09 14:02:55 ogra_: i'll ping you when I check. Jan 09 14:02:59 just selecting "Show Password" in the wifi setup screen would bring the problem. after selecting "Show Password", onscreen would not work. reboot was needed. Jan 09 14:03:18 would it be possible to use e.g. LVM on the partition to split it into one filesystem for the original image and another that is created on the first boot or so? hmm, or could the ubuntu installer load itself into RAM and create the actual file system while installing instead of having a premade fs image? (just speculating here...) Jan 09 14:03:41 xnox, the last upload was in sep. 2012 ... i doubt it has the fix Jan 09 14:03:50 yeah. Jan 09 14:04:53 heh, seems FAT has the same issue on win8 ... but no way to determine if the FS is unclean or not ... Jan 09 14:06:04 Rjs, how would having the installer in ram help ? Jan 09 14:06:16 ... if you cant change the partitioning :) Jan 09 14:06:40 Rjs, we were pondering LVM, but thats kind of a hack Jan 09 14:06:56 ogra_: then the installer could create a full-size filesystem from scratch on the actual device, so it would just use whatever size partition the device has? Jan 09 14:07:05 (vs properly having a single partition across the whole MMC) Jan 09 14:07:19 Rjs, see above, hardcoded partition table Jan 09 14:07:25 you cant really change it Jan 09 14:07:36 ogra_: hmm, does it have a hardcoded 8gb partition on all of the devices? Jan 09 14:07:42 no Jan 09 14:07:57 it has a 6G partition as "userdata" on the 8G device Jan 09 14:08:12 all other devices have bigger userdata partitions Jan 09 14:08:44 6G is the smalles common denominator across all devices Jan 09 14:09:22 ogra_: so if the installer loaded itself into RAM, could it then use mkfs to create the filesystem on the userdata partition (and, say, extract a tar to it to have the initial contents of the file system)? Jan 09 14:10:03 ogra_: then the mkfs on the installer could select the filesystem size based on the size of th userdata partition on the actual device that it is currently installing to? Jan 09 14:10:19 no Jan 09 14:10:49 partitions formatted from the initrd (or wherever you sit in ram) end up with a broken filesystem that loses data over time Jan 09 14:11:06 you filesystem has to be created by the android mkfs Jan 09 14:11:47 (we spent about a week to come up with the cleanest way that doesnt eat your data when we designed the image type) Jan 09 14:12:28 what you can do, and what the usb-creator installation will do, is repacking the image before flashing top the right size using the android-fstools Jan 09 14:12:55 ogra_: oh, sorry, I didn't know that... is it because of some sort of android kernel oddity? (the userdata partition is not just a standard block device where the OS can write whatever data it wants to?) Jan 09 14:13:29 it is because of the GPT partition table that is hardcoded in the bootloader Jan 09 14:14:50 it is hard to tell what exactly is the issue sinbce the whole bootloader is a big binary blob Jan 09 14:14:59 and you cant replace it Jan 09 14:15:42 ogra_: hmm, but after the kernel gets loaded, /dev/mmcblk0p3 (or whatever number the userdata partition has) isn't just a block device that you could access from the initrd to create whatever filesystem you want there? Jan 09 14:16:19 or does the bootloader need something special written on the userdata partition? (so you couldn't for example use a completely different filesystem type for it?) Jan 09 14:16:21 it is, but filesystems created with standard fs tools fail Jan 09 14:16:33 feel free to research it :) Jan 09 14:17:15 hmm, I could try if/when I get the time :) (I just recently bought a nexus7 for testing, but haven't yet had time to install anything on it) Jan 09 14:17:27 working images can be created using the make_ext4fs tool from android-fsutils ... the source code is there, you could research what the mkfs dpoes differently from a normal mkfs :) Jan 09 14:18:44 we used to use the "format from ram" approach in the preinstalled pandaboard images we had earlier Jan 09 14:20:15 ogra_: hmm, ok... maybe I'll try to look into it if I get the time and energy (probably not in a while)... thanks for the information :) Jan 09 15:18:51 jay, https://launchpad.net/~marvin24/+archive/tegrarcm Jan 09 15:19:02 * marvin24 celebrates his first package Jan 09 15:20:12 * ogra_ applauds Jan 09 15:20:15 congrats ! Jan 09 15:21:12 ogra_: now tell me how to build it for different ubuntu releases ;-) Jan 09 15:21:26 you need to change the changelog ... Jan 09 15:21:52 and re-upload for each release (after rolling a fresh source package indeed) Jan 09 15:21:58 cg :) does whoopsie (or how is that "Error occured, send report" thing called) produce logs? Jan 09 15:22:21 apport does, whoopsie just commits them Jan 09 15:22:28 in /var/crash/ Jan 09 15:22:28 ogra_: ok, I hoped that would be more easy using some launchpad interface Jan 09 15:22:45 there might be an easy way i dont use :) Jan 09 15:22:50 because something crashes on my n7, so that this window shows up pretty frequently Jan 09 15:23:19 * Tassadar is looking to /var/crash Jan 09 15:23:29 the .crasdh files in /var/crash have pretty recognizable names Jan 09 15:23:40 yeah Jan 09 15:23:42 dbus a cups Jan 09 15:24:18 *and Jan 09 15:28:23 cups' error log says that some ssl keys are missing http://pastebin.com/rQfak24W Oo Jan 09 15:37:50 seems that "make-ssl-cert generate-default-snakeoil" in /etc/ssl/certs fixed it Jan 09 15:39:03 can somebody confirm that this is happens for all n7's? I mean the "System error occured" window with "report" option? Jan 09 15:41:13 if something crashes it should happen on all n7's Jan 09 15:41:24 happens on your desktop too :) Jan 09 15:42:18 well i am multi-booting it, it _may_ be because of that, but I don't think so Jan 09 15:42:34 anyway, if I won't find bug report about this, I'll create one Jan 09 15:43:05 just click report :) Jan 09 15:43:22 gksudo Jan 09 15:43:49 plus, when I managed to somehow type my password in, the report process crashes :x Jan 09 15:44:02 well, not crashes, "fails" or something Jan 09 15:44:49 gksudo ? Jan 09 15:44:53 i dont unbderstand Jan 09 15:45:08 unable to enter password to gksudo Jan 09 15:45:30 gksudo shouldnt be used by anything on the desktop Jan 09 15:45:31 it like "disables" rest of the screen, including the keyboard Jan 09 15:45:46 if it pops up *anywhere* thats a bug Jan 09 15:46:07 apps needing admin privs need to use pkexec Jan 09 15:46:13 which doesnt have such issues Jan 09 15:46:39 well, I am not sure it is gksudo, i just can't enter anything and it is exatly like https://bugs.launchpad.net/ubuntu-nexus7/+bug/421660 Jan 09 15:46:40 Launchpad bug 421660 in ubuntu-defaults-nexus7 (Ubuntu) "gksu's and gksudo's modal password prompt prevents OnBoard's virtual keyboard input, causing accessibility issues" [High,New] Jan 09 15:47:37 Tassadar, https://bugs.launchpad.net/ubuntu-nexus7/+bug/421660/comments/10 Jan 09 15:48:30 okay Jan 09 15:49:32 so if its actually gksudo it would be good to find out which component uses it and file a bug against it Jan 09 15:57:18 OK. sbuild --host arm64 now works with raring + my raring-bootstrap. :-) I'll run a build and see how much breakage we have left Jan 09 16:42:07 ogra_: seems changing the distro in changelog is not enough, do I also need to bump the package version? Jan 09 16:42:28 hmm. you shouldnt need to Jan 09 16:47:12 dput says: "Package has already been uploaded to ppa on ppa.launchpad.net" Jan 09 16:47:26 the source.changes shows different distributions Jan 09 16:47:41 ypou need to remove the local .upload file :) Jan 09 16:57:12 xnox, so ayan would like to work on the usb-creator installation bits ... Jan 09 16:57:40 \o/ ok =) Jan 09 16:57:43 :) Jan 09 16:58:02 ayan, if you run into any issues dont hesitate to ask here ;) Jan 09 16:58:16 I can help. I should make usb-creator trunk to be usable. But then new bits can be added. Jan 09 17:15:44 ogra_: will do. Jan 09 17:33:34 ayan: it's fairly nice architecture: sepearate frontends & backends. I guess nexus7 images should be as a new backend & then the UI can be hooked up to it. Jan 09 17:39:29 which reminds me about udisks2 port Jan 09 17:42:47 xnox: ya, maybe this is a good time to port usb-creator to udisks2? Jan 09 17:47:29 it's long overdue. usb-creator is the last package to use udisks1 on the desktop cd, everything else was ported already. Jan 09 17:47:42 * ayan nods. Jan 09 17:47:47 i took a brief look at it a while ago. Jan 09 17:47:54 if you don't mind, i can take a stab at it today. Jan 09 17:49:29 For the Ubuntu nexus7 image, to suspend the device does it just use pm-utils, pm-suspend function or something custom for the device? Jan 09 18:09:11 Jef91: it should use pm-utils as far as I know the N7 provides the common kernel interfaces for suspend Jan 09 18:09:41 kk morphis thanks Jan 09 19:11:00 hrmm I have audio working at startup on my nexus 7 Jan 09 19:11:05 without doing a suspend/resume dance Jan 09 19:11:13 now I just need to figure out how I accomplished this Jan 09 19:16:37 nice yer here too Jan 09 19:16:38 lol Jan 09 19:17:30 Hey you ubuntu-arm wizard devs, trying to install Jef91 custom image he made (dope btw u should check it out) and having the same issue trying to install the old image on a 32gb with 3g Nexus7 **** ENDING LOGGING AT Wed Jan 09 19:52:32 2013 **** BEGIN LOGGING AT Wed Jan 09 19:54:19 2013 Jan 09 19:55:28 afk. Jan 09 21:59:51 hello! Jan 09 21:59:59 I've been having this error: http://paste.ubuntu.com/1514650/ Jan 09 22:00:05 and I've been wondering how to fix it. Jan 09 22:00:22 trying to compile the pegasus driver for ARM **** ENDING LOGGING AT Thu Jan 10 02:59:59 2013