=== ogra is now known as Guest54665 === Guest54665 is now known as ogra_ === ogra_ is now known as ogra [11:07] morning [11:07] ogra: getting back to work I believe the main issue for getting images is the u-boot mkimage transition [11:08] for some reason people decided to remove it first before fixing the archive [11:08] to be faster than debian, but this is still being discussed there [11:09] ogra: and something would be good to have a final conclusion is the kernel for omap 3, and trying to use the linaro one [11:09] if not this week, at least let's have a discussion next one [11:10] do pandaboard wlan work on natty? I installed the mavertick packages from TI PPA but they don;t seem to work (no chip found by modprobe) [11:10] rsalveti, i'd like to solve it this week [11:10] depending on workload [11:11] yeah [11:11] for mkimage we're waiting for the MIR [11:11] i agree, it was bad to remove uboot-mkimage first, complain to lool [11:11] janimo: hm, I believe you need to get a new wlan package [11:11] not that the images did build though [11:12] well, we got a broken image for days just because of the mkimage mess [11:12] the package was removed at 12-14 [11:12] yep [11:12] but yeah, still waiting for the MIR [11:13] janimo: try using https://launchpad.net/~tiomap-dev/+archive/omap-trunk [11:14] ogra: another thing is that I got a new x-loader package for omap 3 and omap 4, and will try to update it this week [11:14] the good thing is that I'm building both from the same sources [11:14] so one less package at the archive [11:15] ogra: Why complain to me? [11:16] janimo: http://omappedia.org/wiki/Ubuntu_OMAP_trunk [11:16] rsalveti: thanks. upgrading to that PPA now [11:16] janimo: ti updated most packages for 10.10 [11:16] but natty is using the same kernel [11:16] janimo: you may need a new x-loader [11:16] rsalveti: hmm, does that affect the wlan chip? [11:16] if you're using the latest natty's kernel [11:16] ogra: I didn't trigger the mkimage issues; the packages were automatically imported from Debian [11:17] am using 35- 1001 I think [11:17] janimo: not directly I believe, but it's good to use the new x-loader with latest kernel [11:17] ogra: I just noticed because I have mkimage on my laptop and I saw the uboot-mkimage package as obsolete after an upgrade [11:18] hmm, so removals happen automatically ? [11:18] rsalveti: but uBoot actually uses the kernel image on the first partition, and regardless of what I upgrade to on / that kernel is used still right? [11:18] i thought that needs manual action [11:18] then i dont blame you :) [11:18] I believe it was manually removed [11:18] but it wasn't lool [11:18] ogra: I have no idea whether they happen automatically; I didn't request removal, and I'm not an archive admin [11:19] janimo: yes, but after a kernel update the flash-kernel tool is called [11:19] and then it upgrades the kernel from the first partition [11:19] ah, ok [11:19] did not know that [11:23] lool: ogra: bug 674904 [11:23] Launchpad bug 674904 in uboot-mkimage (Ubuntu) "Please remove uboot-mkimage from the archive (affects: 1) (heat: 77)" [Undecided,Fix released] https://launchpad.net/bugs/674904 [11:26] * ogra rants a bit on that bug [11:27] asac, are you still active in the MIR team (i.e. could you approve u-boot) ? [11:30] ogra: is that upstream u-boot? [11:31] * asac thought we already had some u-boot in main [11:31] I even made a patch for u-boot to ship uboot-mkimage transitional [11:32] http://pastebin.com/g49n8Jde [11:32] asac: it's upstream u-boot [11:32] ogra: bug id please [11:33] lool: can you finally review it? [11:33] asac: https://bugs.launchpad.net/ubuntu/+source/u-boot/+bug/692613 [11:33] Launchpad bug 692613 in u-boot (Ubuntu) "[MIR] u-boot (affects: 1) (heat: 509)" [Undecided,New] [11:33] bug 692613 [11:33] Launchpad bug 692613 in u-boot (Ubuntu) "[MIR] u-boot (affects: 1) (heat: 509)" [Undecided,New] https://launchpad.net/bugs/692613 [11:33] hrw: Ah right; I have it in my TODO for a while, I wanted to fix some small things and fix the i386 build issue as well, but didn't get to it before the break; sorry about that [11:34] lool: happens. next time I will request bugs/reviews probably so others could look too [11:34] asac: uboot-mkimage was in main, and uboot-envtools is in main; uboot-mkimage is gone and was a fork of the upstream u-boot code [11:35] lool: ftbfs on i386 [11:35] https://launchpad.net/ubuntu/+source/u-boot/2010.12~rc2-1ubuntu1 [11:35] hrw: Note that the package name will change too, so I expect we will have to do another round [11:35] asac: I've noted that in the MIR [11:35] lool: does this ship anything besides uboot-mkimage? [11:35] e.g. any u-boot bootloader? [11:35] asac: Yes; it ships u-boot [11:35] asac: but we don't use these images [11:36] the Debian maintainer is semi-convinced that we need to split mkimage in a separate package [11:36] lool: so no: http://paste.ubuntu.com/549796/ [11:36] This is discussed in http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=594937 [11:36] Debian bug 594937 in u-boot "u-boot: Create binary package for each supported machine instead of for each architecture" [Wishlist,Open] [11:36] not a single relevant bootloader for ubuntu ;) [11:37] nah [11:37] asac: It's the Debian package [11:37] couldnt we at least produce omap3 from that source? [11:37] theoretically we could [11:37] asac: I don't see any harm in keeping whatever Debian maintains; we don't care about the bootloader bits [11:37] but we're using the linaro one for that [11:37] right [11:38] for both omap 3 and omap 4 [11:38] I don't see why we should move away of Linaro's u-boot [11:39] we would have two omap source packages again (linaro omap4 and debian omap3) it would somehow defeat the purpose of unifying but techincally we could indeed build the omap3 binary [11:39] and i belive there is work going on in debian to actually do that [11:39] i would prefer to wait for u-boot-tools to arrive from debian ... until then couldnt we just use one of our existing u-boot packages to provide mkimage [11:40] * ogra remembers a discussion about building official debian omap3 kernels, they will for sure also build bootloaders if they do that [11:40] asac: Well, we could split u-boot in two in the way that we anticipate Debian to do it and provide the patch [11:40] right [11:40] thats fine with me [11:41] and by we, I obviously don't specifically mean me :-) [11:42] rsalveti, hi and happy new year :) [11:42] guerby: hey! [11:42] happy new year! [11:42] rsalveti, my pandaboard still freeze every few days, did you get any feedback from TI on https://bugs.launchpad.net/ubuntu/+source/linux-ti-omap4/+bug/690370 ? [11:42] Launchpad bug 690370 in linux-ti-omap4 (Ubuntu) "Strange out of memory on pandaboard (affects: 1) (heat: 202)" [Undecided,New] [11:43] guerby: ti is getting back to work today too, so I expect something this week [11:43] rsalveti, ok great! thanks for your work on this. I'm still hoping to put the pandaboard online :) [11:43] guerby: sure, will try to do some work on this bug this week [11:44] my membership of ubuntu-mobile is due to expire - is this team deprecated now? [11:46] yep [11:46] if you are intrested in arm work, there is ubuntu-armel [11:50] * hrw will look at uboot/i386 ftfbs [11:51] hrw, i doubt thats easily solveable without a package split [11:52] hrw: see u-boot mailing-list [11:52] hrw: I propose two fixes; the maintainer of the board says he is working on sending the patc h [11:52] hrw: I think we should just send two patches [11:52] lool: official uboot ml one? [11:53] hrw: Yes [11:53] hrw: MID 20101220111358.GA28532@bee.dooz.org [11:55] lool: author name would be better... [11:55] gmane lacks mid search [12:46] hrw: It does have it [12:46] hrw: http://mid.gmane.org/20101220111358.GA28532%40bee.dooz.org [12:52] ok, thx [17:35] hi all , i have downloaded http://rcn-ee.net/deb/rootfs/maverick/ubuntu-10.10-r3-minimal-armel.tar.7z [17:36] how to install unity in that [17:36] to create a netbook image [17:39] aksh1: if you're using this rootfs already then you just need to install unity, but it doesn't work that well with GL ES [17:41] rsalveti, then do i need to do any changes ? [17:41] aksh1: you need to install the unity package [17:42] rsalveti, but you told that it doesn't work well [17:42] with GL ES [17:43] aksh1: sure, I wouldn't recommend to use it with gl es [17:43] you can, but it's very slow and crashes a lot [17:43] that will be fixed for natty [17:43] rsalveti, what will be the best [17:44] for now on arm the efl netbook interface is the best option [17:44] or other wm like xfce [17:44] rsalveti, how about lxde [17:44] which is good better performance in xfce4 and lxde [17:45] aksh1: could be, just never tried [17:45] ion surely has the better performance ;) [17:46] I'd use evilwm [17:46] :P [17:46] heh [17:47] evilwm ? [17:47] * ogra always loved wm2 [17:47] evilwm is a very minimal wm [17:47] you can only open a terminal with it hehe [17:48] and move it [17:48] wm2 already has a tasklist [17:48] and shiny window decorations [17:49] ogra@ac100:~$ apt-cache show wm2|grep -i size [17:49] Installed-Size: 116 [17:49] Size: 33096 [17:49] ogra@ac100:~$ apt-cache show evilwm|grep -i size [17:49] Installed-Size: 96 [17:49] Size: 30236 [17:49] hmm, but evilwm wins the size battle [17:49] rsalveti, i am looking a wm with full fledged gui to test utouch [17:49] ogra: ;-) [17:50] i guess you cant have bling without paying for it ;) [17:51] aksh1, then take the efl launcher [18:00] GrueMaster: I get banshee UI coming up over ssh -X and I can use the menus. no crash so far [18:00] * janimo has just started banshee for the first time ever [18:00] Interesting. [18:00] last liune in the log is [Info 19:58:16.396] nereid Client Started [18:01] so right before you see the exception [18:01] this is an up-todate natty on pandaboard [18:03] same here/ [18:03] Will try normal desktop to see if it is a netbook-ui issue. [18:10] Nope, still fails. Different error though, but I suspect the root is the same. [18:12] did you wash your fingers before firing it up ? its very fragile ;) [18:12] GrueMaster: so you can reproduce on panda/natty too? [18:12] This is panda/natty. [18:12] ok [18:12] within seconds from startup? [18:12] yes. [18:13] can you check if it crashes with ssh -X too? [18:13] not that it should be much of a difference [18:13] It doesn't stay up long enough to use the menus. Will try ssh -x. [18:13] just to make sure I can try reproduce without pulling out the HDMI cable [18:13] which means no dual monitor on my workstation :D [18:13] Also, I am not getting any window frames in classic desktop mode. [18:14] (different issue). [18:15] GrueMaster, remove maximus or disable it in the session settings [18:15] i thought i had solved that one in maverick [18:15] is that a maverick->natty upgrade ? [18:15] or a plain natty install [18:16] FYI mine is a natty image made with rootstock [18:17] ah [18:17] dont trust rootstock [18:17] its as close as possible to a real install, but after all not identical [18:19] ogra: alsa-utils is still broken :-) [18:19] another package to fix soon [18:19] yeah [18:19] first images ... then packages [18:19] sure [18:19] i also have a broken shadow to fix and doko slowly gets nasty [18:20] ogra: This is the last natty image prior to mkimage snafu, updated to latest packages. [18:20] k [18:20] thx [18:20] that should have the fixed maximus [18:20] keep the pieces until next week and i'll take a look [18:21] ogra: rootstock is recommened in the wiki, it is no longer to be used? [18:21] janimo: it's ok to use it [18:21] are there better ways to do the same? [18:21] but while reproducing issues, the best option is to use the preinstalled image [18:21] ok [18:21] janimo, rootstock is o for building for unsupported HW, for test and development images, but you will only get 100% identical results on the real images [18:21] s/o/ok/ [18:22] just because the supported images are the preinstalled ones [18:22] I used rsalveti's instructions to have / on an external USB drive so could not really use the preinstalled image [18:22] there is no guarantee that rootstock builds 100% the same as we do with the preinstalled image [18:22] s [18:22] it is much much faster this way that on SD card [18:23] well probabaly not the same, but for individual user space app testing I think there should be no differences [18:23] janimo: sure, while debugging and building stuff the usb disk usage is the best option [18:23] but while reproducing and testing bugs, try using the preinstalled one [18:23] the setup might differ [18:23] i have installed plymouth and updated initrd using update-initramfs -u [18:23] do i need to change boot commandline [18:23] rootstock doesnt track what changes are in oem-config for example [18:23] mem=512M console=ttyS2,115200n8 console=tty0 omapfb.mode=dvi:hd720-24@60 root=/dev/mmcblk0p2 rw rootwait is my commandline [18:24] so if you use rootstock and set up a user from cmdline, your user setup might be different [18:24] the way we build the image with rootstock isnt 100% identicaql to livecd-rootfs so fixes that go in there might not be in rootstock [18:24] etc etc [18:25] aksh1, add quiet and splash as keywords [18:26] and make sure your boot setup uses the uInitrd indeed [18:28] ogra, i am not sure uInitrd updated one is used while booting [18:29] bcoz default initrd will get updated in /boot/ folder and beagleboard has other /boot partition [18:30] any reason why you didnt use one of the official netbook images for the beagle ? [18:30] you wouldnt have to fiddle with that now [18:31] ogra, want to know how to create similar image [18:33] well, the same things i wrote for janimo above apply to these images since they are built with rootstock ... they are "similar" but not identical [18:33] ogra: I looked at the image and maximus is not running while in classic desktop. I have plenty of SD cards, so I will preserve this one for you to eval. [18:33] things like the bootloader setup definitely differ heavily [18:34] GrueMaster, weird, maximus isnt running but you dont get window borders either ? [18:35] nope. [18:35] sounds more like metacity being broken [18:35] unless [18:35] hmm [18:35] i heard roumors that metacity would be dropped in favor of a 2D compiz [18:35] not sure what the status on that is [18:37] ogra: http://members.dsl-only.net/~tdavis/Screenshot.png [18:43] Apparently, compiz is the registered WM currently running. [18:44] aha [18:44] so thats likely a 2D compiz bug then [18:44] i think the code is very very young [18:45] ogra, http://www.viewsonic.com/gtablet/spec.htm [18:45] must be. There is no compiz-config utility installed. grrr, [18:46] davidm, hmm, sounds like the folio you can buy here [18:47] interesting [18:47] ogra, after adding quiet splash it is giving kernel panic [18:47] http://pastebin.com/6xyZksqq [18:47] if its the same, it is overpriced, very bad android adaptions, very bad touchscreen [18:48] I was emailed info on this: http://www.tangent-tycoon.com [18:48] very bad android adaptations is kind of expected [18:48] but not a bad touchscreen [18:48] I asked if we could get some eval units to install a real OS, but no reply. [18:48] GrueMaster: very very ugly hehe :-) [18:49] ubuntu should be a lot better [18:49] 120GB HD. 2Gb ram. Just to run Win7 POS edition. [18:50] ouch [18:50] * rsalveti takes a break [18:50] * ogra thinks the galaxy is the only serious ipad competitor atm [18:51] 1.87 lbs. I think my netbook is lighter. [18:51] yeah [18:51] mine surely is (700grams) [18:51] :) [18:53] Heh. To get the extended battery version, you have to buy it with Win7 Pro. [18:53] $100 difference. [18:53] $595 or $695. === playya__ is now known as playya