[01:20] apparently a 2GB USB stick isn't big enough. :( [01:20] irony: it ran out of disk space copying imx51 files. [01:21] well, of poetic interest anyway, if not irony. [01:24] We usually recommend a minimum of 4G, but I've heard of 2G installs working before. That's a bit of an unexpected result. [02:27] You really do need 4G [02:27] unless you're just using base [02:28] Didn't in 9.04, but that might have been a fluke, or a side effect of a less complete port. [02:29] (lucid being the first time that all of main compiled) [02:29] persia : You may be interested in this ... a new piece of hardware arrived yesterday at the office. I can't reveal the manufacturer, but the form factor will be familiar to you (clamshell PC) .. I was surprised to see what it had when I booted. 1330x768 screen, resistive touch, 128GB flash SSD and a quad core A9 operating at 800Mhz. [02:30] In any case, it's got lucid loaded on it from the factory .. but something that most certainly is _not_ netbook remix as a front end. [02:30] Can you share the screen dimensions, or the DPI? [02:32] heh. OK. Maybe you can bring it to UDS, if you're lucky. [02:32] If it's small enough, I want :) [02:32] If it's large, just not enough pixels for me. [02:33] It's about 12cm in diagonal, widescreen [02:33] Indeed, I'm *very* interested. [02:33] Stacking the SL-3200 on top of it, it's a bit larger, but only by about enough to make the keyboard more comfortable to use [02:33] I'll ask the boss if it's okay for me to bring the thing to UDS [02:34] if I remove all the stickers and external identification, I think it might be [02:34] Does it have the wonderful keyspacing that the 3200 used, or does it have the annoying "large" keys of e.g. the Netwalker? [02:34] neither... hard to describe the keyboard [02:35] Please do ask if you can bring it. That sounds nearly ideal for me, assuming other bits are in place. [02:35] The Netwalker is lovely, but a bit larger than I like, and the keyboard is just plain annoying. [02:36] the closest keyboard I can describe is the UMID [02:36] (but I don't mind waiting ~6 months for retail, to make sure I have the spare funds) [02:36] I know this thing must be floating around Akiahabra [02:36] UMID is a little better than Netwalker, but still not enough space *between* the keys for my hands. [02:37] Yeah, but you'd have to tell me what it was for me to find it there. I didn't see them on Saturday whilst wondering about. [02:37] This isn't the device -- but it's DAMNED close in form factor : http://www.dynamism.com/notebooks/umid-mbook-m1.shtml [02:38] Right. That's the same base as the Onkyo I've been resisting buying. [02:38] Onkyo jammed in more flash (64) and an improved audio handler, and went for the faster chip. [02:39] http://www.slashgear.com/globalfoundries-28nm-arm-cortex-a9-promises-speed-frugality-boost-1581731/ (first wafer of a test chip .. there were two and the first wafer failed (thus the business card on top) ) [02:39] But 1366x768 lets me run 2 terminals and an IRC client, which makes me actually productive, whereas 1024x600 is just a bit too small (or too blocky) [02:41] Yeah, I'm intrigued. I'll bug you about it in a couple weeks. [02:41] (but have to rush off for a bit now) [02:42] See you later Emmett :) [04:04] ikepanhc, 你们太淡定了 === Guest91645 is now known as NCommander [07:18] * NCommander waves [07:42] saeed: you aorund? [07:42] *around [07:45] NCommander: hey there [07:46] GoodMorning All [07:47] ericm: ping [07:47] eggonlea, pong [07:48] eggonlea, NCommander mentioned that LCD support on AV-D1 is not stable ? [07:48] Have to use webchat because the previous proxy is broken. :( [07:48] eggonlea, no problem [07:48] what do you mean by "not stable"? [07:48] NCommander, your turn [07:48] NCommander, are you around? [07:55] eggonlea: yup [07:55] * NCommander coughs === hrw|gone is now known as hrw [08:05] morning [08:05] hrw: heya [08:06] first day at new work ;D [08:07] working from home i guess? [08:14] Stskeeps: sure [08:14] Stskeeps: I work from home for over 3 years now [08:14] * Stskeeps is only up at about 5 months so far [08:14] am happy i have to travel once in a while to work with real people though [08:16] Luck People :( [08:17] *Lucky [08:18] and yesterday I changed one thing - now I have one machine with ubuntu @home [08:18] wife's laptop has kubuntu 10.04-rc === XorA|gone is now known as XorA [09:53] saeed, btw - I've fixed the hibernation issue and updated the LP bug description, please help check [10:59] * ogra ponders to do a mass give-back for universe ftbfs to see if something more survives [10:59] NCommander, what did you want to know about touchscreens ? [11:01] and what kind of touchscreens... usb ones or i2c/spi or uart connected ones [11:01] no idea, he pinged me in another channel asking to pick my brain about them === ericm_ is now known as ericm|ubuntu [11:51] ericm: good job [11:51] ericm|ubuntu: can you please send me the fixed kernel deb? [11:52] guys: I'm running totem with mp4 movie, the pulse audio consumes ~60% of the cpu time!!! [12:19] saeed, the link is on the LP bug description [12:19] saeed, that's possibly pulseaudio issue [12:20] ogra: hi. i'd like to mount a ubuntu minimal FS on qemu using NFS mount. have you already tried that with the kernel you provide here: https://wiki.ubuntu.com/ARM/RootfsFromScratch [12:20] ndec, no, NFS might be modular in that kernel, not sure ... might be that lool tried it before though [12:21] ogra: how did you build the kernel? which config? i can add nfs, if I know which source and which config was used [12:22] its the linux package source [12:22] ogra: i am planning to use qemu to build arm packages [12:22] std ubuntu 2.6.32 release kernel [12:23] ogra: with the versatile_defconfig? [12:23] right [12:24] well, merged with the distro config, amitk should be able to tell you what to run to merge the configs [12:24] there is some script iirc [12:24] ericm|ubuntu: the had this issue in the past with speex, and it supposed to be fixed afair [12:26] saeed, I guess that's because pulseaudio is now realtime priority (for less latency) in lucid, and could possibly eat up more CPU and become unresponsive [12:27] * ericm|ubuntu 's rhythmbox crashed [12:28] ndec: are you experimenting or need something that is more or less 100% reliable? you could also try to use a qemu-arm-static chroot to build packages. [12:29] ndec: with that you can then later with some tweaking use a fast cross compiler/bash from host etc. [12:30] asac: i don't have too much to build for now, so fast is not the biggest requirement. i just need something that works ;-) I am on travel away from my board and I want to build ARM packages [12:31] ericm|ubuntu: hibernate works for me, without adding the resume= ofcourse [12:32] ndec: yeah. the chroot is really simple though ... just install qemu-kvm-extras-static annd then run build-arm-chroot /path/to/chroot ... then you can just chroot /path/to/chroot and are in a arm chroot [12:32] but testing our full qemu is also good ;) [12:32] i think persia included qemu-arm-static support in sbuild and pbuilder [12:33] that should be the easiest to cross build packages [12:33] * ogra doesnt know the magic runes for that though ... i think its just sbuild --arch=armel though [12:34] asac: qemu-arm-static is using syscall emulation, right? [12:34] right [12:35] as long as you dont plan to build anything mono related it should work fine for everything [12:35] ogra: i don't know about sbuild --arch=armel [12:35] ogra: I just prepare all my source packages, and using sbuild --arch=armel? is that what you use? [12:36] no, i use native builds :) [12:36] saeed, do you know the current hibernation implementation on dove will work with HIGHMEM (which I doubt) [12:36] i rarely travel without HW in my bag ... ARM boards are so small :) [12:37] haha [12:37] random arm board and assorted wires in bag is something that should be a problem at airport security [12:37] ...never been for me thou [12:37] neither for me [12:37] i always have enough ubuntu CDs with me to bribe the guys ;) [12:38] and stickers ... stickers work even better ;) [12:38] maybe I look innocent enough to make the security people not care what the bag x-ray looks like [12:38] ogra: well right now I am away without my arm board ;-( [12:38] ndec, well, an armel chroot should get you going then [12:39] ndec, the reciepe asac gave you above should be fine ... "install qemu-kvm-extras-static annd then run build-arm-chroot /path/to/chroot ... then you can just chroot /path/to/chroot and are in a arm chroot" [12:40] thopugh i'd recommend approx or some other package proxy running locally .... [12:40] saves you bandwith if you can re-use the already downloaded packages [12:40] ok. i will try that. [12:43] hmm, i wonder how well ubuntuone works on my beagle [12:44] oh, it causes a firefox race [12:44] "firefox is already running, please close the running firefox process first" [12:44] blah [12:45] clikcing OK brings up firefox with my ubuntuone account :P [13:07] ogra, so, you're aware of rootstock hangs? [13:07] I seem to be getting segfaults instead of hangs now [13:08] yes, as i mentioned in bug 532733 [13:08] Launchpad bug 532733 in qemu-kvm (Ubuntu Lucid) (and 2 other projects) "apt/dpkg in qemu-system-arm hangs if a big task is installed (affects: 4) (dups: 1) (heat: 36)" [High,Incomplete] https://launchpad.net/bugs/532733 [13:08] has nothing to do with rootstock but with qemu [13:09] * XorA found with gcc 4.5 and glibc 2.10.1 there is at least one TLS bug with qemu that causes a hang [13:09] I grabbed two patches to fix it from master git [13:09] its gcc 4.4 though [13:10] but I dont know if thats the same bug you guys are seeing or not [13:10] and we use eglibc [13:10] ogra: which means your probably using an equiv glibc then [13:10] eglibc basically being glibc-next [13:10] oh, didn't see that [13:11] http://repo.or.cz/w/qemu.git?a=search&h=HEAD&st=commit&s=cp15 top two commits on that search [13:12] ogra, sorry, hadn't checked my email yet doay [13:12] * cwillu_at_work reruns the test [13:20] ha [13:20] finally [13:20] * ogra listens to his ubuntuone musicstore music on the beagle === ericm|ubuntu is now known as ericm-afk === rsalveti_ is now known as rsalveti [15:47] ogra: do we support using tslib out of the box and blacklisting evtouch, or is it only evtouch we support? [15:58] evtouch doesnt need to be blacklisted [15:58] its not installed unless you tell it to [15:58] tslib might conflict with usbtouchscreen kernel side though [15:58] since all touchscreen support is supposed to be handled by that now [15:59] oh.. I missed talks === jmcgee|gone is now known as jmcgee [16:01] ogra: it isn't, its being loaded automatically on a touchscreen device [16:01] evtouch ? [16:02] ogra: yeah, at least on ARM [16:02] we neverf install it by default [16:02] its not even in main to my knowledge [16:02] ogra: odd. It is on a ubuntu-netbook image out of the box [16:03] surely not [16:03] apt-cache show xserver-xorg-input-evtouch|grep Filename [16:03] Filename: pool/universe/x/xf86-input-evtouch/xserver-xorg-input-evtouch_0.8.8-3build1_i386.deb [16:03] unless we build from universe which would very much surprise me [16:03] dpkg -l | grep touch [16:03] ii libts-0.0-0 1.0-7build1 touch screen library [16:03] ii tsconf 1.0-7build1 touch screen library common files [16:03] its not there [16:04] XorA, yes, i know :) === bjf is now known as bjf[afk] [16:04] but NCommander doesnt [16:04] ogra: xserver-xorg-input-evdev 1:2.3.2-5ubuntu1 [16:04] ogra: sorry, my brain went tothe wrong place [16:04] evdev != evtouch [16:04] ogra: I'm not caffinated yet; I just got up ;-) [16:05] so evdev is loaded for your touchscreen ? [16:05] ogra: the reason I ask is that the touchscreen on the avenger is causing evdev to segfault, but tslib. I was wondering if there was a way we could blacklist evdev until it can be fixed [16:05] and when evdev segfaults, it takes the entire X stack with it [16:05] well, i guess you have to do that on a kernel/udev level [16:06] NCommander: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=578053 [16:06] Debian bug 578053 in xserver-xorg-input-tslib "xserver-xorg-input-tslib: consider xorg.conf.d instead of udev rule" [Serious,Fixed] [16:06] and make sure that evdev either doesnt consider the device an input device or make sure the kernel exports the right stuff to make evdev actually recognize a touchscreen device [16:07] zumbi, that wont prevent evdev [16:07] ogra_cmpc: i dunno about ubuntu, but debian is moving xorg out of udev [16:08] zumbi, i doubt that [16:08] which seems that it causing the segfault [16:08] upstream mives into udev [16:08] *moves [16:08] it only just started with the recent release [16:08] everything should be handled by kernel or udev [16:09] ogra_cmpc: true, sorry i was misleading [16:09] with the option to use xorg.conf.d snippets for overriding buggy settings [16:10] the question is if the snippet kicks in before evdev probes or not [16:10] ogra_cmpc: read Julien Cristau on the bug report [16:10] and the actual bug is that the kernel doesnt export the proper constraints for the device so evdev can handle it as touchscreen [16:13] zumbi, yes, i see that, still its a kernel bug [16:13] using udev rules or xorg.conf.d snippets are both workarounds [16:18] ogra_cmpc: thanks, let me know if you make any progress (that needs to be incorporated to debian xf86-input-tslib package) === hrw is now known as hrw|gone === bjf[afk] is now known as bjf [19:02] ndec: the config is in the source tree, the git tree, or more easily in the .deb; the /boot/config-foo file has the config [19:13] ojn : You in? [19:13] (and are you in Austin at the moment?) [19:29] Martyn: In CA until friday night [19:30] thanks :) [19:36] Hi, can I use bb files without haveing internet connection on my Beagleboard? === rsalveti_ is now known as rsalveti [19:42] ogra_cmpc: mk-sbuild or pbuilder-dist take --arch : pbuilder and sbuild don't. === JaMa is now known as JaMa|Zzz === jmcgee is now known as jmcgee|gone