=== Ursinha-afk is now known as Ursinha === Ursinha is now known as Ursinha-afk [01:54] so I'm trying to update my chromebook to 12.10 [01:55] and it is failing due to a checksum mismatch [01:55] W:Failed to fetch [01:55] gzip:/var/lib/apt/lists/partial/ports.ubuntu.com_ubuntu-ports_dists_quantal_main_binary-armhf_Packages [02:02] Try again now (it's been 5 minutes). That usually means a mirror goes wonky, which is usually solved shortly for the mirrors identified by "ports.ubuntu.com" [02:11] I tried at like noon, and was getting it then [02:11] * mjrosenb tries again [02:12] nopepe, still there [02:13] s/pe// [02:41] any other suggestions? [03:07] Maybe delete everything under /var/lib/apt/lists/ and run `apt-get update` again? This should redownload everything, but it won't fix it if it's an issue with the checksums on the server. [03:16] It won't be a mirror issue, that file has been static for months. [03:16] Unless it's actually corrupt on the mirror he's hitting. [03:17] Which is pretty unlikely for ports. [03:18] Nope, looks fine from here. [03:18] mjrosenb: Try persia's "delete the world and try harder" approach. [03:21] * mjrosenb is running apt-get update as i type === heathkid|2 is now known as heathkid [03:29] well, that is strange [03:29] it says there is no new release [03:30] but dist-upgrade wants to upgrade 1,223 packages [03:30] and /etc/issue now says that I am running 12.10 [03:30] Are the nexus7 raring images working yet? [03:31] apt believes you to be running 12.10 with outdated packages (based on the content of /etc/apt/sources.list) [03:31] persia: sounds right. [03:31] * mjrosenb wonders if this is going to attempt to install a new kernel [03:31] and how that is going to go if it does [03:32] It will only do so if you have a kernel package installed for which there is a newer package available in your currently configured archives. === zz_chihchun is now known as chihchun [03:37] I never installed a kernel package [03:38] although I should still get perf on here at some point or other. [03:40] and iirc, the kernel that this is running was just copied from chromium so we'd get nice driver support out of the box [04:45] great, now plymouth is throwing an assertion [04:45] hrmm [04:45] actually, I have absolutely no clue how to recover this system [04:52] <[mbm]> mjrosenb: nexus7? [04:53] <[mbm]> there was an open bug about that one; the console=none in the kernel commandline was causing issues, and plymouth had been disabled until it could be fixed [05:01] [mbm]: no, chromebook. [05:38] although I would not be surpred if this is the same assertion that the nexus 7 is throwing [06:16] so, I'd like to do something fancy like init=/bin/bash [06:16] but I don't actually know how to touch the kernel arguments on here. [08:01] good morning [08:05] morning. === dholbach_ is now known as dholbach [09:31] http://lwn.net/Articles/526920/ [09:31] Support for Tegra 2D hardware [09:32] ppisati, heh, you just missed the person intrested in it most :) [09:33] (though igues marvin24 has seen it already) [09:33] marvin24, http://lwn.net/Articles/526920/ [09:34] ppisati, sadly tegradrm doesnt work with the binary driver yet [09:34] (afaik) [09:35] ogra_, ppisati: yeah, this will take a *long* time [09:35] right, and thanks to our new desktop policy we are bound to the GLES drivers [09:35] so even if it would land tomorrow, we couldnt use it [09:36] cause of unity [09:36] well, compiz, but yeah [09:36] what abount software gles? [09:36] *about* [09:36] i didnt get it working on the nexus yet [09:36] i have it working on the chromebook and dont know why :) [09:37] does anyone know how to change the parameters that the kernel gets on a chromebook? [09:37] ogra_: strange ... [09:37] mjrosenb, probably hrw [09:37] mjrosenb: you have to sign kernel and at that stage you can change cmdline [09:37] easy [09:38] mjrosenb: > [09:38] > [09:38] > [09:38] ops [09:38] mjrosenb: https://plus.google.com/u/0/109993695638569781190/posts/34PYU79eUqP [09:38] so just get a pen then :P [09:39] ogra_: any progres on teaching plymouth to check proper console=? [09:39] hrw, it isnt the console parameter at all :) [09:40] or whatever [09:40] hrw, it is a missing piece of console-setup [09:40] ogra_: tell me more [09:40] if you build your initrd with FRAMEBUFFER=1 set in the initramfs.conf, plymouth works fine [09:41] I do not have initrd [09:41] i'm working on tracking it down to a specific function in console-setup, once i have that well just add it to the plymouth upstart job and it should work [09:41] we'll [09:41] hrw: the issue is I cannot currently boot into ubuntu [09:42] hrw, right, initrd doesnt matter its just that you can fix it by having an initrd that runs console-setup [09:42] mjrosenb: tell me more? [09:42] which means running that console-setup bit from an upstart job before plymouth starts should work as well [09:43] hrw: I attempted to upgrade from 12.04 to 12.10 with the standard do-dist-upgrade, and now it plymouth just thros an assertion, and I cannot continue booting. [09:43] mjrosenb: boot chromium, mount ubuntu, cd ubuntu/etc/init/;rm plymouth*;reboot [09:43] works just fine [09:44] no, the biuts will re-appear on next update [09:44] or s/rm plymouth*/for f in plymouth*;do echo manual > ${f}.override/ [09:44] or s/rm plymouth*/for f in plymouth*;do echo manual > ${f}.override;done/ [09:44] drop the rm's [09:44] iirc, I can boot into chromium by just turning on the validation again? [09:44] yeah [09:45] the above looks fine [09:45] mjrosenb: better is keep chromium kernel in KERN-A and boot ubuntu with KERN-B or KERN-C [09:45] *nod* [09:45] so all you need is change of partition priorities [09:46] works just fine [09:46] hrw: so I didn't touch the kernels, I just ran do-dist-upgrade; apt-get dist-upgrade [09:46] err, maybe it wasn't do-dist-upgrade [09:46] do-release-upgrade? [09:46] mjrosenb: do update to 13.04 [09:46] I tab-completed it [09:46] do-release-upgrade is the right tool [09:46] hrw: is support better in 13.04? [09:47] yes [09:47] mjrosenb: s/better// [09:47] hrw, up0loads packages and fixes all the time :) [09:47] 12.10 does not know anything about chromebook [09:47] hrw: does 12.04? [09:47] you are joking or not? [09:48] the 12.04 that you can install with that weird script that downloads 50 tarballs will work [09:48] (since it is hacked up) [09:49] yeah, that is how I installed it to begin with. [09:50] I used python to calculate offsets and did setup with linaro image [09:50] that's why my chromebook may lack some packages [09:53] I am waiting for 32GB microsd card to arrive so will have pristine ubuntu install on it [09:53] unless courier would bring me another chromebook... [09:54] but this will rather not happen [09:55] is there a microsd card reader other than the full-sized sdcard on the side? [09:57] there are small usb readers [09:57] which sticks out for 3mm from socket [09:58] ahh. I've seen those [09:58] this one that holds the microsd card so it will eject *into* the computer if you could theoretically eject it when it was inside the computer? [10:00] no idea - order not arrived yet [10:02] this mount command is taking quite a while to complete... [10:04] aaand, ^C does not do anything on this terminal :/ [10:13] ok, problem solved [10:13] i assume I need to re-run |sudo cgpt add -i 6 -P 5 -S 1 /dev/mmcblk0| in order to get this machine to want to boot into ubuntu again? [10:20] -T5 [10:21] or -T1 at least [10:21] ok, this time, it finished the kernel initializations, then "mountall: Disconnected from Plymouth" and is now hung there [10:21] what does that do? [10:22] sets 'tries' field [10:22] oh, so it'll boot into ubuntu N times, then go back to cros? [10:23] no, thats matter of priority field [10:24] you have 3 fields: tries, priority, successful [10:24] priority == which kernel boot first and which if first one fail [10:24] successful = this kernel boots at all [10:53] i think i'll go with something like the second one here http://tbreak.com/tech/2012/06/quick-look-apacer-32gb-usb-3-0-memory-sticks/ [10:54] the values arent stellar but still far beyond USB 2.0 [10:54] (for the chromebook rootfs) [10:56] ogra-cb: nice :) [10:57] i havent found a place to buy it yet though [10:57] http://www.shopmania.co.uk/external-memory/p-apacer-ah152-super-mini-usb-drives-16gb-48812413 but 16GB only [10:57] it is small enough to not turn into a lever that opens the netbook case if you press it :) [10:58] I have to connect my usb3.0 harddrive and check speed [10:59] you should see between 80 and 120M/s [10:59] i personally think even 60 are awesome alreadfy though :) [11:00] ogra-cb: it's just a pity you can't use usb 3.0 to boot kernel from :( [11:00] i dont care [11:01] as long as my rootfs runs off a fast device [11:01] and as long as i *can* boot a kernel somehow :) === hrww is now known as hrw [11:31] RaYmAn: you put kernel in KERN-C on internal and boot rootfs on usb3 - fine [11:36] sure [11:39] internal emmc has 59 MB/s average read in gnome-disks bench [11:44] ech. gnome-disks fails to bench my usb3 drive [11:44] on chromebook [11:45] file a bug :) [11:45] hdparm has 92MB/s [11:50] Bug #1081019 already exists [11:50] Launchpad bug 1081019 in gnome-disk-utility (Ubuntu) "Benchmark for Harddisk fail" [Undecided,Confirmed] https://launchpad.net/bugs/1081019 === davidm` is now known as davidm [11:51] same hdd connected to desktop has 147 MB/s average [11:56] I miss "maximize to whole available desktop space" action in WM === chihchun is now known as zz_chihchun [12:04] which WM ? [12:08] kwin === Guest2393 is now known as jrgifford === Ursinha-afk is now known as Ursinha === gatox_ultra is now known as gatox === yofel_ is now known as yofel [14:46] hi. Does anybody know why ext4 fs is prefered over a "flash" fs such as ubifs ? [14:49] fmoreau: for a flash file system, such as ubifs, you need a raw interface to the flash. most device these days instead have a MMC interface where you can only put block filesystems, such as ext4 === Quintasan_ is now known as Quintasan === plars_ is now known as plars === prpplague^2 is now known as prpplague === gatox is now known as gatox_lunch [16:18] o/ [16:39] yippie ! [16:39] slideshow is back [16:41] xnox, hmm, so that onboard behavior is really bad in oem-config since the switch to compiz [16:41] i actually have to reboot to make typing work [16:43] plars, janimo, so i dont get why you need adbd instead of just using ssh over usbnet [16:44] ogra_, no idea if usbnet works I never used it [16:44] it does [16:44] pretty agressively even :) [16:44] ogra_, if the current kernel needs some patch or config options enabled feel free to ping the kernel team :) [16:45] if not then a wiki entry would be nice [16:45] * ogra_ accidentially built it in into one of his test kernels ... made my desktop NM go crazy :) [16:45] I never used this feature [16:45] it should not interfere with normal work though and not be buggy :) [16:45] ogra_: same, never tried usbnet. If you have some pointers I'd like to hear more about it though. adb is nice from the perspective of being a similar tool for what a lot of people would have normally used there, and has some neat features, but I'm not really stuck on it [16:45] it doesnt [16:46] but NM shows you it tries to auto establish the network (it only shows you a wired connection) [16:46] ogra_: will usbnet have issues since the n7 might be on a server, potentially with several other n7's attached doing the same thing? [16:46] it just took me quite aq while to find out that this behavior comes from the attched nexus (which didnt have the NIC up ) [16:47] plars, usbnet just behaves like any other network [16:47] ogra_: So I'd need to be able to bring up those connections and distinguish between them [16:47] just that your device is called usb0 and that the wire is actually a usb connection to the next machine [16:48] ogra_, janimo: so it sounds like we just need a kernel with it turned on, and give it a try [16:48] right [16:48] ogra_, and it shows up depending on whether you connect the nexus to a laptop or not? [16:48] i simply assume you want to issue more than just a reboot via ssh [16:49] janimo, plugging in the USB connection with usbnet *buioltin* will always try to bring up usb0 (and the desktop/laptop side for it) [16:49] indeed you shouldnt build it in :) [16:49] but load it on demand as module [16:49] ogra_: I don't actually have NM running on the ubuntu-server they are attached to currently [16:50] well, you want to attach more of them [16:50] i would actually runs something like dnsmasq [16:50] so you can hand out IPs to them via dhcp [16:50] and preseed usb0 properly for that === gatox_lunch is now known as gatox [17:26] ogra_: why does a reboot help after flashing in oem-config? a reboot is still treated as first boot if oem-config is not completed? [17:26] ogra_: is it a problem with fastboot reboot then? [17:42] xnox, nope. rather with some ubiquity-dm race or onboard [17:43] and yes, oem-config-remove is run last, it removed ubiquity-dm ... as long as that didnt run you will always end up in oem-config on next boot [17:43] *removes [17:44] ogra-cb: right so if I reboot a few times sometimes onboard is fine and sometimes it isn't? but doesn't ubiquity launch onboard anyway ... [17:45] * xnox will have to play with it. [17:45] err no [17:45] if you reboot onboard will always work fine [17:45] it just doesnt if it boots through from the tarball installer [17:46] i could add another reboot to the process but imho we already have to many (especially with the one we will need for the hostname later) [17:56] janimo: so from what I'm reading, it looks like it's could just be CONFIG_USB_ETH that we need to turn on? [17:57] plars, ogra-cb should know as he had a kernel with that enabled. I am not certain what other (if any) options are required === Ursinha is now known as Ursinha-afk === Ursinha-afk is now known as Ursinha [21:03] hrw: ping [21:12] unrelatedy, if I don't want oneconf or update-manager running, what is the right way to turn them off? [21:59] ok, so I'm going to assume that the best way forward at this point is to just go through the chrubuntu setup again, since all of the data I care about is on an external sdcard === Jack87|Away is now known as Jack87 [22:35] mjrosenb: pong [22:36] hrw: should I try to get 12.10 working, or should I just give up, re-install from the crazy tarball-downloading script, then try to update directly to 13.04? [22:41] mjrosenb: I suggest skipping 12.10 [22:41] well, the plann was to get 12.10 booting, then immediately upgrade to 13.04 [22:48] mjrosenb: make a backup, upgrade to 13.04 directly. if fail - restore backup [22:55] hrw: that shouldn't be too hard, the partitions are small enough that I can dd them both onto an external sd card. [23:21] is anyone familiar with the rockchip rk3066 devices? e.g. those chinese hdmi sticks like MK808? [23:23] and does anyone have a complete kernel source? the chinese manufacturers don't seem to care much about the GPL.