**** BEGIN LOGGING AT Sat May 26 02:59:57 2007 May 26 03:31:18 (script) openmoko-community: Sean Moss-Pultz Re: First impressions of Neo1973 May 26 03:33:13 is Sean refering to P1 or P1+? May 26 03:33:49 first sentence sounds like P1+, second like P1 May 26 04:25:18 (script) wiki RecentChanges: [[Buying_Interest_List]] [[Neo1973_Audio_Subsystem]] May 26 05:10:48 apt_: wantnick apt May 26 05:33:45 morning May 26 06:16:18 (script) openmoko-community: Sean Moss-Pultz Re: HXD8? May 26 06:26:10 counter May 26 06:26:10 http://londerings.novalis.org/wlog/index.php?title=Neo1973_counter (probability guess: 50%) P1: 3 days 02:46:53 (3.116 +-3.1 days) (1283;213) May 26 06:26:19 (script) wiki RecentChanges: [[Wish_List]] May 26 07:59:19 good morning May 26 07:59:46 morning :) May 26 08:14:05 mornin' May 26 08:15:00 OpenMoko's GPS: http://pavelmachek.livejournal.com/39229.html May 26 08:26:18 (script) wiki RecentChanges: [[Bluetooth_Support]] May 26 08:44:55 Hello, everyone. May 26 08:45:24 I searched angstrom wiki and openmoko wiki... but still don't know. May 26 08:45:36 How do I get angstrom to run on neo? May 26 08:46:02 Ipkg update (from phase1 machine) tries to load non-existent http://downloads.openmoko.org/snapshots/20070301/ipk/Packages.gz :-( May 26 08:46:13 Hi, Magon! May 26 08:46:37 LuitvD: check out the ThinkPad x range, for example x40 x30 etc.. 12" May 26 08:46:39 pavelm: Ask koen May 26 08:47:36 webjames: thanks :) May 26 08:47:37 CM: Koen will eat me alive ;-). May 26 08:47:48 webjames: OMG, x-series tablet :D May 26 08:48:26 LuitvD: yeah, pretty cool, well made as well you know that with the thinkpads, metal hinges etc.. May 26 08:48:35 webjames: only thing is that it isn't 12" widescreen May 26 08:49:22 pavelm: Hehe.. He tends to do that ;) May 26 08:49:26 * CM hides May 26 08:49:42 LuitvD: ahh yes, i guess they are quite old now May 26 08:50:22 webjames: though for a nice price it could be interesting May 26 08:50:43 too bad it's a laptop based on the old centrino platform (nape) May 26 08:51:14 LuitvD: i would say less than 300 May 26 08:51:31 webjames: a tablet pc for less then 300? :| May 26 08:51:45 LuitvD: oh maybe not the tablet one May 26 08:51:50 heheh May 26 08:52:56 CM: If there's anyone else who got it working (and not yet eaten by koen), please speak up ;-). May 26 08:53:35 webjames: the x60 tablet is at least 1300 euro's May 26 08:53:44 1400 even May 26 08:53:48 LuitvD: check out: http://www.tabletpcreview.com/default.asp?newsID=275 has fingerprint reader as well May 26 08:53:55 excluding vat May 26 08:54:08 i think that would be an amazing feature to have with the neo May 26 08:54:18 a fingerprint reader that is May 26 08:54:54 LuitvD: http://cgi.ebay.co.uk/Lenovo-ThinkPad-X60-Tablet-6366_W0QQitemZ260121868651QQihZ016QQcategoryZ177QQssPageNameZWDVWQQrdZ1QQcmdZViewItem May 26 08:55:31 that's 600 euros :| May 26 08:55:35 that's not much May 26 08:57:24 webjames: nah, I still prefer that HP Pavilion tx1127 :P May 26 08:58:16 LuitvD: the HP is nice May 26 08:58:19 webjames: Fingerprint reader is basically useless. May 26 08:58:28 pavelm: how so? May 26 08:58:50 webjames: it has the same features as that thinkpad x60 or x40 tablet May 26 08:58:51 webjames: You can't use it for encryption, and when it is not encrypted, it does not matter.... May 26 08:58:59 anyone up for a libgsmd-tool question? May 26 08:58:59 michaelshiloh_: See http://wiki.openmoko.org/wiki/FAQ (Hopefully this link answers your question. This autotrigger for new nick and question during first 15 minutes done only once.) May 26 08:59:00 lol May 26 08:59:00 ...on easily-stolen machine like neo, anyway. May 26 08:59:55 i meant like a pin, login thing, unlock thing May 26 09:00:25 Well, your fingerprints are all over your neo. ;-) May 26 09:00:35 Take them and fool the fingerprint reader ;) May 26 09:01:22 Elrond: yeah, but aparently it reads the electrical signature of the print, so if your finger was cut of it would only work for ~15mins after May 26 09:01:24 pavelm - You tried to a repos of prebuild armv4t binary .ipk, or what? May 26 09:01:46 webjames - All of that can be fooled, really. May 26 09:02:18 Elrond: ah, o well, how about camera with iris recognition? May 26 09:02:30 or facial recognition May 26 09:02:48 Oh, and if your fingerprint is not good on your phone, use your tfid reader and get the fingerprint from your new all-cool ID card ;o) May 26 09:03:08 heh, webjames just found the killer argument why neo needs a camera :) May 26 09:03:27 elrond: No, you just steal the neo. May 26 09:03:36 elrond: and then you read the flash using jtag. May 26 09:03:55 webjames - The basic problem with all these things is: They must be fuzzy. So it still recognizes you after you drank 20 beers. ;o) And that fuzziness makes them weak. May 26 09:04:11 pavelm - Right, that anyway May 26 09:04:12 pavelm: therefore encryption is the answer, but what way to lock/unlock that encryption May 26 09:04:13 elrond: I'd love to try prebuild .ipk's, yes. May 26 09:04:32 webj: Fingerprints _do not work_ for unlocking encryption. May 26 09:04:53 webj: You can't md5 fingerprint and use that as a key! May 26 09:05:11 Because the fingerprint is always a little different. May 26 09:05:35 pavelm: no, but what interesting way are there to do that, passwords are so last year...! :) May 26 09:05:41 warm weather... small blood vessels May 26 09:06:12 webjames - passwords are the current best thing. May 26 09:06:27 ... use them to unlock an RSA key, which then is your SSO key :) May 26 09:06:47 a suitable complex squiggle would work, divvying the screen into segments and taking a series of them as the passkey May 26 09:06:55 would have to be somewhat long though May 26 09:06:58 pavelm - I was googling for prebuild .ipk yesterday. Of course I only got drugs for koen ;o) May 26 09:07:24 mjr - Quite long. May 26 09:07:44 elrond: Prebuilt binaries seem to be on http://angstrom-distribution.org/unstable/feed/ ... May 26 09:09:05 pavelm - Hmm, they're eabi? May 26 09:09:21 elrond: They should be May 26 09:09:30 elrond: Is it as easy as changing /etc/ipkg/base-feed.conf ? May 26 09:09:47 pavelm - openmoko is OABI. May 26 09:10:15 ... at least I think so. May 26 09:10:31 elrond: Are you sure? I think koen runs his stuff on openmoko. May 26 09:10:33 pavelm - Yep, it should be as easy as that. Or just wget individual .ipk :) May 26 09:10:50 elrond: Plus openmoko is incompatible with everyone exactly because of being eabi. May 26 09:10:51 pavelm - koen runs OE and builds everything by hand. May 26 09:11:17 elrond: and uploads it to his http://angstrom-distribution.org/ website ;-). May 26 09:11:20 neo% file hello-world May 26 09:11:29 hello-world: ELF 32-bit LSB executable, ARM, version 1 (ARM), for GNU/Linux 2.4.0, dynamically linked (uses shared libs), for GNU/Linux 2.4.0, not stripped May 26 09:11:49 Looks like OABI to me... let me download some .ipk from angstrom May 26 09:14:19 neo% wget 'http://www.angstrom-distribution.org/unstable/feed/armv4t/base/libpython2.4-1.0_2.4.4-ml1_armv4t.ipk' May 26 09:14:28 neo/1/usr/lib% file libpython2.4.so.1.0 May 26 09:14:32 libpython2.4.so.1.0: ELF 32-bit LSB shared object, ARM, version 1 (SYSV), stripped May 26 09:14:36 EABI, I thihk. May 26 09:14:37 pavelm: lets hope a small dna reader comes along, then we could use an unbreakable one time pad May 26 09:15:01 webjames - Well, you leave your DNA around everywhere. May 26 09:15:10 web: dna is not random. May 26 09:15:32 hmm.. back to the drawing board May 26 09:15:58 pavelm - DNA is at least non-changing, so md5'ing it would work. Main question would be, how many bits you get (excluding the non-random part) May 26 09:16:16 elrond: Actually, http://www.angstrom-distribution.org/unstable/feed/ has both arm and arm-oabi. May 26 09:16:17 ... of course ignoring my first argument, that you leave your dna everywhere. May 26 09:16:22 So pick one ;-) May 26 09:16:46 elrond: I do not think DNA is "non-changing" at the "whole cell" level. Still doubt md5 would work. May 26 09:17:02 elrond: plus it is 700MB :-) May 26 09:17:06 pavelm - The arm-oabi does't look like "for armv4t", just for armv-we-dont-know ;o) May 26 09:17:14 okay so i guess a password/pin would be the best way May 26 09:17:48 elrond: But I still think openmoko is eabi. May 26 09:17:57 elrond: Because openmoko is incompatible with everyone and newer. May 26 09:18:24 have to go, bye bye! May 26 09:18:31 pavelm - Let me download a fresh .ipk from openmoko. :) May 26 09:18:38 Bye pavelm! May 26 09:18:52 pavelm - ONe question: Where you lucky to run gllin on P0? May 26 09:19:09 no gllin on p0 :-( May 26 09:32:05 pavelm - You're back? May 26 09:32:05 * pavelm is back --- from zaurus -- and travelling. May 26 09:32:13 hehe :) May 26 09:32:27 pavelm - Did you try to copy the gllin from P1 to P0? May 26 09:32:39 elrond: did any of the binaries work? May 26 09:33:47 pavelm - You mean from angstrom? Did you try yet. I'm reboot the P0 to try gps with different kernels. May 26 09:34:16 did you guys see this: http://www.spectec.com.tw/sdw823.htm ? May 26 09:34:21 wlan without hw mod :) May 26 09:35:18 ossman - ;) May 26 09:35:25 interesting May 26 09:35:28 Elrond: there was no gllin on phase0, so yes, I had to copy it. May 26 09:36:12 ossman - BTW: I talked to my friend and gave him the SD-(for SPI)-URL from you. His problem was: What to do with the non-needed pins of the SD card. Connect to pull-down, or pull-up? May 26 09:36:35 pavelm - But with the copied one it did not work, right? May 26 09:37:19 elrond: I could not get it to work :-( May 26 09:37:47 Elrond, pull-up if we're talking about dat[1-3] May 26 09:38:19 but I'm unsure exactly how spi uses the pins May 26 09:38:50 ossman - I have not looked at the hw-spec myself... maybe I should do that soon. I want the spi thing to work too. ;) May 26 09:39:13 pavelm - Same here. May 26 09:39:47 if you're hacking linux, then talk to David Brownell May 26 09:40:01 he's been working on it for some time May 26 09:40:07 pavelm - If I were much better at hw things, I'd try to get a P1 and put both next to each other and put probes on the PCB ;) May 26 09:40:22 ossman - No. 8051 uC. May 26 09:41:50 Elrond: I guess you can find enough info in kernel patches. May 26 09:42:05 elrond: look for 'if phase(0)'... May 26 10:05:30 pavelm - Yeah, I'll look at it again. May 26 10:07:06 Bus 001 Device 014: ID ffff:1300 May 26 10:07:14 Hey... That's noname. Cool. ;o) May 26 10:24:04 hi May 26 10:26:18 (script) wiki RecentChanges: [[OpenMoko_compatible_cellphone_providers]] May 26 10:46:31 counter May 26 10:46:31 http://londerings.novalis.org/wlog/index.php?title=Neo1973_counter (probability guess: 50%) P1: 3 days 00:36:43 (3.025 +-3.0 days) (1284;213) May 26 12:26:18 (script) wiki RecentChanges: [[Optical_Character_Recognition]] May 26 12:27:18 someone know which kernel version we should use on devices? 2.6.20.x or 2.6.21.x? May 26 13:15:34 counter May 26 13:15:34 http://londerings.novalis.org/wlog/index.php?title=Neo1973_counter (probability guess: 50%) P1: 2 days 23:22:11 (2.974 +-3.0 days) (1285;213) May 26 13:35:31 hrw did you find out which kernel version? May 26 13:35:36 no May 26 13:38:56 hrw, maybe here: http://wiki.openmoko.org/wiki/Kernel#General seems to indicate 20 May 26 13:39:07 but could be patched to 21 May 26 13:42:36 that page is old May 26 13:43:44 o.. May 26 13:51:57 webjames: 2.6.21 is correct May 26 13:52:09 bye May 26 13:52:20 hrw, cool. bye May 26 14:14:50 ossman: do you have an idea why CMD3 may be failing? i now have the dmesg on my screen with MMC_DEBUG enabled but i have no means to transfer it to the PC May 26 14:15:26 CMD3 is the first command that fails in the sequence (reporting Card Error) but it does write a sane looking "relative address" in the response May 26 14:26:18 (script) wiki RecentChanges: [[User_talk:Maui_Addicted]] [[Wish_List_-_Hardware]] [[Main_Page]] May 26 14:32:41 oh, nevermind, the omap folks forgot to take R6 into account (again!) and treated it as R1, while the omap hardware does distinguish betwen them - now it stops at CMD9 May 26 14:40:51 balrog-kun: troubles with the CSD register? May 26 14:44:14 koen: i don't know, it is likely the host driver's fault May 26 14:45:02 was the big omap merge schedules for .22 or .23? May 26 14:47:10 the arch/ and include/ part has been mostly merged now, and the rest i think depends on when rmk will have time to review. iirc he didn't find time in the .22 merge window May 26 14:47:50 ah, right May 26 14:50:54 beg your pardon, was anyone successful recently with MokoMakefile and qemu? I try to run 'make qemu' and it keeps saying 'U-boot failed to finish writing in 300 seconds, giving up.' May 26 14:52:00 yes May 26 14:52:04 worked for me May 26 14:52:48 any ideas why doesn't it work? Neither in 300, nor in 900, 1800 etc. seconds... May 26 14:54:12 so this is when the qemu is running? May 26 14:55:08 well, as I understand it, when it's trying to flash rootfs image into NAND May 26 14:55:25 mmh May 26 14:58:12 Should it write anything informative after 'Please wait, programming the NAND flash...' ? May 26 14:58:39 i don't know May 26 14:58:52 never looked at the terminal when it was doing it :) May 26 14:59:24 yavannadil: can you try without MokoMakefile? May 26 15:00:16 yavannadil: webjames had exactly the same problem before - qemu hung after the "Please wait..." message and did nothing May 26 15:00:19 i don't know why yet May 26 15:00:27 That's my next idea, but, according to wiki, it means just calling the same flash.sh by hand, if I'm not mistaken... May 26 15:01:12 and without MM it worked? May 26 15:01:18 yavannadil: hmm, yes, maybe you'll also have to call download.sh by hand too because MokoMakefile does someting with the directories May 26 15:01:22 yavannadil: yes May 26 15:01:48 o, thanks, will try and then bother everyone later :) May 26 15:02:05 but then i think there was one person for whom it would also hang without MokoMakefile May 26 15:02:28 hm... May 26 15:02:29 i haven't found any relation between when it happens and the system configuration yet May 26 15:02:42 I run Ubuntu Feisty May 26 15:03:30 same here May 26 15:03:45 which build of openmoko are you using? May 26 15:03:53 just before I leave: is it necessary to put everything into /home/moko? May 26 15:04:14 but yes, it should spit whole lots of stuff after the Please wait message, the script would then detect it and wouldn't fail May 26 15:04:21 don't know, just 'make update-makefile; make update; etc.' May 26 15:04:38 i have it in /home/username/moko May 26 15:04:46 so do I May 26 15:05:47 it failed for me twice before (with different errors), so I try to get latest version for every new attempt May 26 15:05:58 never worked yet :( May 26 15:06:33 and with 1-1.5 week between attempts, it's rather slow :( May 26 15:07:04 well, leaving to try qemu without MM May 26 15:07:26 yavannadil: paste this into a file and run it: http://dpaste.com/11077/ May 26 15:08:24 this is the script that i use, mokomakefile didn't work for me, this did. it also is a lot quicker May 26 15:12:03 you'll have to change the dir (2nd line) May 26 15:15:39 so anybody else interested in that HDX8 stuff? :P May 26 15:17:04 okay here's an updated file, just paste it into a file and make it executable then run it: http://dpaste.com/hold/11078/ May 26 15:18:39 yavannadil: see above, hope this helps May 26 15:33:27 bye bye May 26 15:36:02 <_buz> LuitvD: my guess is hxd8 is a mp4 player thingy May 26 15:36:26 <_buz> lots of flash and a widescreen May 26 15:36:42 <_buz> and wifi May 26 15:37:14 <_buz> maybe this is what zune could have been May 26 15:37:53 <_buz> or then it is a iphone killer May 26 15:37:59 <_buz> if it has gsm May 26 15:42:39 is there any way to try out the gps May 26 15:43:30 bmidgley: see http://wiki.openmoko.org/wiki/Hammerhead_protocol May 26 15:45:30 bmidgley: here is some info http://pavelmachek.livejournal.com/ May 26 16:00:57 Hmm May 26 16:01:07 das qemu need Framebuffer? May 26 16:01:21 when I try to do the "make qmeu" it complaints about missing /dev/fb or /dev/fb0... May 26 16:02:46 Rince it should not need it. it usually uses sdl May 26 16:02:51 hmm May 26 16:03:00 (!) Direct/Util: opening '/dev/fb0' and '/dev/fb/0' failed May 26 16:03:00 --> No such file or directory May 26 16:03:00 (!) DirectFB/FBDev: Error opening framebuffer device! May 26 16:03:00 (!) DirectFB/FBDev: Use 'fbdev' option or set FRAMEBUFFER environment variable. May 26 16:03:09 (!) DirectFB/Core: Could not initialize 'system' core! May 26 16:03:09 --> Initialization error! May 26 16:03:09 Could not initialize SDL - exiting May 26 16:03:25 eh... what do you try there May 26 16:03:39 I just did a "make qemu" May 26 16:03:46 or better: May 26 16:03:47 ( cd build/qemu ; arm-softmmu/qemu-system-arm \ May 26 16:03:47 -M neo -m 130 -usb -show-cursor \ May 26 16:03:47 -mtdblock openmoko/openmoko-flash.image \ May 26 16:03:47 -kernel openmoko/openmoko-kernel.bin ) May 26 16:04:14 ah.. you have directfb installed and it tries to link against it? May 26 16:06:46 eh.. the lines youve postet suggest it tried to execute and youve compiled it successfully against directfb which cannot find the framebuffer May 26 16:07:59 fwiw, i have bluetooth powering on consistently now May 26 16:08:12 had to power on and reset in a script so they happen one after the other May 26 16:09:26 Rince: if your SDL build doesn't support X, or DISPLAY is not set, SDL will probe other devices May 26 16:13:02 balrog-kun yeah... that sounds like it... i just were thinking how qemu can run on directfb May 26 16:15:52 i think other than no mouse input, everything should work May 26 16:26:18 (script) wiki RecentChanges: [[Wish_List_-_Hardware]] [[Current_events]] May 26 16:27:43 Hi, does anyone knowwhere I down libosc8k? May 26 16:42:04 hm May 26 16:42:14 ok, I'll try it another way... thanks ;) May 26 17:05:09 alan cox on the openmoko-apps list...ooh May 26 17:05:12 superstars around ;) May 26 17:05:44 I've been drinking with Alan Cox May 26 17:05:46 nice guy May 26 17:06:49 :) May 26 17:06:57 they mostly are May 26 17:07:11 I've been meaning to meet Ulrich Depper May 26 17:07:17 mostly so I can poke him in the eye May 26 17:50:27 Is it me, or did the buildhost build nothing since the 22th? May 26 18:01:39 Elrond: It's the buildhost not you :) May 26 18:01:54 Elrond: You like something newer? May 26 18:05:13 anyone worked with GtkLayout? May 26 18:07:39 hi May 26 18:22:29 stefan_schmidt - Kernel with patches from revision 2072. ;-) May 26 18:22:38 (gps module fixes) May 26 18:23:12 Elrond: ok, give me a moment (food). May 26 18:23:38 stefan_schmidt - No need to panic. If the buildhost will be fixes within the next days, it's fine. May 26 18:34:39 counter May 26 18:34:39 http://londerings.novalis.org/wlog/index.php?title=Neo1973_counter (probability guess: 50%) P1: 2 days 20:42:39 (2.863 +-2.9 days) (1286;213) May 26 18:41:24 Elrond: http://people.openmoko.org/stefan/uImage-2.6-moko10-r0_0_2079_0-fic-gta01.bin May 26 18:45:20 stefan_schmidt - :-)) May 26 18:48:35 Hmm, is the gsm modem connected to the PMU? May 26 18:49:20 Elrond: No idea, why? May 26 18:51:04 * stefan_schmidt flashes bleeding edge u-boot/kernel/rootfs to reproduce some of the recent bugs May 26 18:52:42 stefan_schmidt - Because a friend wanted to know, if one can really cut the gsm module from the power. The normal sysfs powermanagement-gsm driver just sets a pin on the SoC, so it just sets a pin on the gsm. The module still gets power, it seems. May 26 18:54:08 Elrond: I would need to dig out the schematics, try to understand and would still not be sure what I'm allowed to say. :) May 26 18:54:36 :-) May 26 18:54:47 I'm currently digging through the PMU init code. ;) May 26 18:56:02 Elrond: As far as I know we have full control over the BP, but not sure if that's includes total power off. May 26 18:56:42 Right, and that's the question. ;o) May 26 18:57:39 Elrond: I know :) May 26 19:06:50 stefan_schmidt - http://svn.openmoko.org/trunk/doc/hardware/GTA01Bv4/pmu-voltage.txt doesn't look like we have full control over the power. May 26 19:08:20 Elrond: hmm, it seems. May 26 19:18:38 counter? May 26 19:18:39 "/msg aloril counter?" for actual long message, giving short version here: http://londerings.novalis.org/wlog/index.php?title=Neo1973_counter (probability guess: 50%) P1: 2 days 20:20:39 (2.848 +-2.8 days) (1287;213) May 26 19:19:09 i wonder if the 24th run ever actually happened May 26 19:23:20 zecke: Is there any doc how kernel module building, packaging into rootfs and perhaps autoloading is done? Or is it reads the source? May 26 19:24:40 Hello, sorry for asking things I could read, but it's just quick idea... May 26 19:25:11 Neo1973 doesn't have any hardware blitter, which slows it down a bit... May 26 19:25:42 on PXA, there are some DMA channels available for general purpose, serving as a "copy engines" May 26 19:26:04 doesn't S3C contain something like that as well? May 26 19:26:29 I guess it might be then used as very simple hardware blitter, although all the compositing has to be done in software... . May 26 19:26:42 yeah, s3c can do that May 26 19:26:59 stefan_schmidt: hehe, I'm not certain about the current situation. But there is a bbclass to build external modules (example would include hostap), packaging into rootfs by making the neo machine file RRECOMMEND or RDEPEND it (this might have changed), autoloading by inheriting update-modules and asking the module to be autoloaded (I can search an example) May 26 19:28:00 I'm still not sure about the gain in speed, but maybe someone could try to measure that... it could possibly save few hundreds of cycles... . May 26 19:28:39 stefan_schmidt: packages/hostap/hostap-modules.inc an example for building external modules May 26 19:29:16 zecke: 1) modules are not external 2) modules are already build 3) I take a look at the RRECOMMEND adn RDEPENDS 4) thanks May 26 19:29:23 stefan_schmidt: packages/linux/linux-rp.inc for autoloading modules May 26 19:29:33 zecke: ah, cool May 26 19:29:46 Hopscotch: any idea whether it is real/how it would affect performance? :) (I have never played with these things on PXA) May 26 19:29:55 stefan_schmidt: module_autoload_mmc_block_collie = "mmc_block" this should make the mmc_block_collie package load mmc_block May 26 19:30:16 mmp: haven't tried it either May 26 19:30:36 mmp: and i'm not sure if it is possible to program line sizes and skip values May 26 19:30:36 zecke: ok, sounds easy May 26 19:30:48 mmp: otherwise only full screenwith can be blitted May 26 19:31:01 I'm pretty sure you'd need to reset the DMA thingy for each line May 26 19:31:14 Hopscotch: I think you can tell DMA processor how many data it has to copy... May 26 19:31:20 stefan_schmidt: change MACHINE_EXTRA_RRECOMMENDS in the machine conf May 26 19:31:22 And at 480 lines - every 60Hz, that's a nasty burden. May 26 19:31:35 Not as bad as doing it by hand. May 26 19:31:46 SpeedEvil: I'm not sure, but there are more DMA channels... . That could help a bit... . May 26 19:32:01 * mmp guesses May 26 19:32:03 Hmm. You mean do several lines at once. May 26 19:32:04 Maybe. May 26 19:32:07 yes May 26 19:32:19 and setting up DMAs again only after they are all drawn May 26 19:32:31 there's always the method of painting into offscreen memory and then swapping pages in hardware May 26 19:32:50 Doesn't help much with smooth moves though. May 26 19:35:39 true May 26 19:36:15 but without the ability to skip a number of bytes every line you can't use that dma engine... May 26 19:36:49 A DMA engine that could handle scatter/gather would do. May 26 19:39:05 zecke: ok, I added the sound module to MACHINE_EXTRA_RRECOMMENDS. Is there a recommended way for modules that are not needed for the machine itself but should anyway go into the rootfs? May 26 19:39:54 scratches head, well the next best thing would be distro? May 26 19:40:08 e.g. it is a distro decision to ship ipsec by default or not May 26 19:40:59 zecke: Perhaps, I'm just not sure how mickey|lunch like to have it organized. May 26 19:41:15 zecke: I play a bit more with this stuff. Thanks May 26 19:41:40 hmm orginsation is the enemy of chaos? May 26 19:42:29 SpeedEvil, Hopscotch: well, I don't know what DMA engines on s3c can handle... . But I guess it could be possible to have somehow minimize losses when it comes to line-by-line copying... . May 26 19:42:44 zecke: heh, cu you at ccc camp? :) May 26 19:43:07 zecke: Ah, I think we see us at LinuxTag, right? May 26 19:43:22 hopefully May 26 19:43:32 haha "2 + 2 = 5 for extremely large values of 2" May 26 19:43:42 zecke: :) May 26 19:44:01 zecke: Why hopefully? No time? May 26 19:45:15 hopefully for the camp, for the LinuxTag I might be forced to hack at the LinuxTag May 26 19:45:53 ok May 26 19:46:43 Or overloAded values of + May 26 19:55:33 what is the /home/root/DM2 that pavel machek is referring to May 26 19:55:54 apparently it has gllin and a working earphone alsa state file May 26 19:56:06 morning May 26 19:56:11 hey hrw May 26 19:56:21 bmidgley: DM2 is set of soft on bv4 phones which p0 devs got May 26 19:56:41 hmm I didn't get it May 26 19:56:47 where? May 26 19:56:53 in /home/root May 26 19:57:38 hrw: not in my /home/root (empty) May 26 19:57:56 bad luck? May 26 19:58:21 i though only p1 people had it May 26 19:58:44 only the original P1 image had it May 26 19:58:52 right May 26 19:59:26 I did not back up the image that came with this phone May 26 19:59:36 but it wouldn't work (kernel module mismatch) May 26 20:01:46 I filed a bug for a missing alsa state file May 26 20:01:54 maybe it is in /home/root/DM2/audio/ after all May 26 20:03:13 I also filed a bug because stereoout.state does not work as advertised hmm May 26 20:03:30 pavel refers to play_wav_earphone.state May 26 20:03:33 bmidgley: DM2 was only for factory testing. The alsa state files should be in the normal rootfs May 26 20:03:36 balrog-kun: p1 phase is not yet started in full.. May 26 20:04:14 bmidgley: I'll look into this bug later today May 26 20:04:37 ok May 26 20:05:46 I couldn't match up the alsa controller names with the labels in the wolfson docs May 26 20:05:52 bmidgley: That are the alsa state files we currentyl have in rootfs: http://svn.openmoko.org/trunk/oe/packages/alsa/files/ May 26 20:06:05 bmidgley: yeah, alsa controls are a pain in this phone. :) May 26 20:06:26 bmidgley: That why we let the wolfson guys do the alsa state files. :) May 26 20:06:33 I thought I would write the systembluetooth state but I can't figure out what is going on May 26 20:07:15 alsa controls are pain on WMxxxx codecs May 26 20:07:20 they are too complex May 26 20:07:41 bmidgley: Normally Xora and lrg hanging out here. Both are working for wolfson and should be able to help you perhaps May 26 20:08:05 stefan_schmidt: what timezone? May 26 20:08:11 bmidgley: U.K. May 26 20:09:42 Anybody still have the delievered image on the P1 and can tell me if there are different state file then here? http://svn.openmoko.org/trunk/oe/packages/alsa/files/ May 26 20:12:03 In the DM2 folder I mean May 26 20:13:56 when finally openmoko drop that md-desktop1 crap... May 26 20:15:19 hrw: What exactly you refer to? May 26 20:15:19 stefan_schmidt: there are few others May 26 20:15:24 there are the loopback_* and play_wav_*.state files May 26 20:15:52 WebKit victory is near :} May 26 20:15:54 stefan_schmidt: are you able to use that small up/down icons which are in top right corner to navigate? May 26 20:16:20 hrw: Besides they re to small they work May 26 20:16:40 stefan_schmidt: they work IF you ARE able to select them May 26 20:16:45 hrw, balrog-kun: Can you upload this file somewhere? May 26 20:16:54 hrw: heh, indeed May 26 20:17:16 it is guessing where this time I had to point stylus to press them May 26 20:17:20 hrw: In the end we will not use the desktop but the mainmenu instead May 26 20:17:25 stefan_schmidt: do you know if one can sleep in the GSM village? (at the camp?) May 26 20:17:44 stefan_schmidt: mainmenu is? May 26 20:17:50 zecke: I hope so. We plan to build our tend there. :) May 26 20:17:56 * hrw is lost in openmoko naming May 26 20:18:17 menu, window menu, mainmenu, tasklist, task switcher May 26 20:18:28 stefan_schmidt: sleep with shelter/roof without a tent May 26 20:18:49 zecke: ah, no idea May 26 20:19:02 stefan_schmidt: I'm not certain that doku will join me and having a single tent doesn't sond like fun May 26 20:19:15 hrw: moment May 26 20:19:34 zecke: We will come with a SG30 (30m^2 tent) May 26 20:20:20 hrw: main menu is the app which displays all the app icons and have the fingerwheel to navigate. May 26 20:20:21 stefan_schmidt: try http://www.numenor.art.pl/balrog/alsa/ May 26 20:20:30 balrog-kun: thanks May 26 20:21:52 balrog-kun: Will they stay there for some time or should I copy them over when I like to link them from the bugreport? May 26 20:22:40 I would add to bugtracker anyway May 26 20:23:08 hrw: You mean the files directly? May 26 20:23:14 yes May 26 20:23:26 hrw: sounds like a good idea, indeed. :) May 26 20:23:48 SpeedEvil, Hopscotch: Err, I thought s3c has more than 4 DMA engines; I'm not sure whether this amount can significantly help and not hurt other parts of system... . May 26 20:23:54 I'll take care of this. May 26 20:23:57 what if someone want to read bug and balrog machine will be down? May 26 20:24:39 bmidgley: It seem this files will not help you, too. No system -> bt profile May 26 20:26:28 stefan_schmidt: if it's not very critical you can link to this address May 26 20:26:31 stefan_schmidt: ok. btw, I also checked out play_wav_earphone.state. still not stereo May 26 20:26:42 but at least it turns of the ext speaker May 26 20:27:01 balrog-kun: I'll add them to the bug directly as hrw suggested May 26 20:27:08 stefan_schmidt: okay, cool May 26 20:27:12 (the machine has had zero downtime in last 54 days) May 26 20:27:39 maybe there's a bug somewhere else May 26 20:27:57 I looked at the wolfson docs and couldn't figure out how you'd accidentally get left/right mixed down May 26 20:28:09 shit May 26 20:28:19 my bv3 has charge-die loop again ;( May 26 20:28:56 i'm now playing with the modem and the modem brand responses at http://wiki.openmoko.org/wiki/Hardware:AT_Commands appear to have been fixed, i.e. correct names are reported May 26 20:29:00 :) May 26 20:29:02 do i update the page? May 26 20:29:16 balrog-kun: bv3 does not report, bv4 does May 26 20:29:40 ah okay, i thought someone intentionally obscured the strings :) May 26 20:31:26 is there a reason to not give out your IMEI number? May 26 20:33:26 stefan_schmidt: okay, I will book a place in your tent then May 26 20:33:47 zecke: noted May 26 20:34:42 balrog-kun: The bv4 has a new GSM frimware image May 26 20:34:46 firmware even May 26 20:35:39 stefan_schmidt: are you preparing an official P1 image? May 26 20:35:55 ~lart openmoko bitbake May 26 20:36:11 hrw: why? May 26 20:36:22 zecke: parse time May 26 20:36:24 zecke: no, just try to fix some bugs. mickey|lunch seems a bit overloaded May 26 20:36:31 * apt wallops openmoko bitbake with a main rotation server that needs rehubbing. It won't take long May 26 20:37:12 * stefan_schmidt thinks we need a counter how often hrw uses the lart command a day. ;) May 26 20:38:00 zecke: official P1 images are for puppies :) May 26 20:38:01 stefan_schmidt: ;D May 26 20:38:22 stefan_schmidt: I want to lend my neo to a puppy May 26 20:38:38 zecke: oh May 26 20:39:03 zecke: You should try to get girls without hardware ;) May 26 20:39:39 stefan_schmidt: don't make me cry. I'm working on it though May 26 20:40:19 zecke: She will come, for sure May 26 20:40:50 hehe May 26 20:40:57 balrog-kun: ok, I have added the files to the bug. You can remove them if you like May 26 20:41:54 * zecke flips a coin and decides to shave May 26 20:42:17 hm...how long does it take to build the openmoko-devel-image at average? :) May 26 20:42:17 slider: See http://wiki.openmoko.org/wiki/FAQ (Hopefully this link answers your question. This autotrigger for new nick and question during first 15 minutes done only once.) May 26 20:42:54 slider: 3,5h if you have all sources already and a AMD X2 4600 with 4GB RAM May 26 20:43:24 slider: Without the sources it depends on your internet connection May 26 20:43:26 ~lart openmoko^Wfic for lack of normal ps in rootfs May 26 20:43:26 * apt gives openmoko^Wfic a good seeing to for lack of normal ps in rootfs May 26 20:43:41 2ghz singlecore without sources...so maybe tomorrow i guess :) May 26 20:43:53 hrw-lart-counter++; May 26 20:43:59 hrw: the ps thingy is driving me nuts as well May 26 20:44:05 Hello, everyone. May 26 20:44:20 slider: Let it do his job tonight :) May 26 20:44:33 stefan: oh now , lart counter just overflowed. May 26 20:44:52 matchbox-desktop-2 is much better then 1 May 26 20:44:54 hi Pavel May 26 20:45:12 pavelm: it's a long and I started it just 10 minutes ago :) May 26 20:45:20 CoreDump|home: few other things too but I lack time to play more with neo to complain enough May 26 20:45:49 hrw: complaining is of no use. Reaction times to bugreports are a little bit "high" May 26 20:46:17 hrw: btw, I'll submit my startup-monitor applet for panel-2 to mallum next week May 26 20:46:32 CoreDump|home: sometimes I do some neo related things which are pushed by not only me to coreteam to look at them May 26 20:46:42 CoreDump|home: Just add patches and we gat faster ;) May 26 20:46:59 stefan_schmidt: really? May 26 20:47:04 !ombug 564 May 26 20:47:05 * * Bug 564, Status: RESOLVED (INVALID), Created: Unknown May 26 20:47:06 * * gronslet(AT)gmail.com: MokoMakefile: cmd_mfsl.c:32:21: error: asm/asm.h: No such file or directory May 26 20:47:07 * * http://bugzilla.openmoko.org/cgi-bin/bugzilla/show_bug.cgi?id=564 May 26 20:47:08 stefan_schmidt: did once, didn't help. May 26 20:47:13 !ombug 566 May 26 20:47:15 * * Bug 566, Status: NEW, Created: Unknown May 26 20:47:16 * * openmoko(AT)hrw.one.pl: settings-daemon recipe NEED updating May 26 20:47:17 * * http://bugzilla.openmoko.org/cgi-bin/bugzilla/show_bug.cgi?id=566 May 26 20:47:29 566 is critical even May 26 20:47:50 as it break upgrades totally May 26 20:47:50 ok guys, I'll poke mickey|lunch once he is online again May 26 20:48:47 * hrw want separate /boot partition on neo ;D May 26 20:50:13 o.. new theme has smaller fonts May 26 20:54:54 * hrw will run powertop on neo.. May 26 20:56:45 after recompilation of kernel and dfu May 26 21:20:10 http://blog.haerwu.biz/tmp/grab.png May 26 21:20:51 hrw: looking good May 26 21:23:13 hrw: is the screeni from the FIC phone? May 26 21:23:19 yes May 26 21:23:49 hrw - What app do you use to show this nice list? May 26 21:24:05 matchbox-desktop-2 May 26 21:24:36 Is that in the default image? May 26 21:24:45 no May 26 21:28:07 I builded a openmoko-image today and tried it out on my Universal but the text and folders where twice as big as usual (yes i have a vga screen). Looked like i was running the image on a qvga screen. May 26 21:30:35 * pavelm is back, this time hopefully for real. May 26 21:31:22 pavelm: how goes gps hacking? May 26 21:32:38 hrw: phase1 + gllin -> mostly working gps, so I guess I have other toys right now. May 26 21:33:08 pavelm: bridge it to BT so neo will work as BT-GPS :) May 26 21:34:49 hrw: I played with that already ;-). It works on a pc, and I guess I could make it run on neo, too. May 26 21:35:15 hrw: But gps does not have signal here, so such games are time-consuming. May 26 21:35:38 stefan_schmidt - You still alive? May 26 21:35:39 happens May 26 21:35:53 (Is there any way to get recent history of this irc channel? I've seen discussion that looked interesting, but then my machine crashed). May 26 21:35:57 Hi, Elrond. May 26 21:36:09 Elrond: So what is the result of neo vs. oabi vs. eabi? May 26 21:36:17 Elrond: Always, but give me a moment May 26 21:36:39 pavelm - I did not fully it more. I compiled python + pyserial finally. :-) May 26 21:36:54 pavelm - But now u-boot doesn't like me. May 26 21:37:06 ~logs May 26 21:37:25 ~logs May 26 21:37:28 i guess logs is apt/ibot/infobot/jbot/purl all log daily to http://ibot.rikers.org// where channelname is html encoded ie: %23debian | lines that start with a space are not shown | some channels have stats at http://ibot.rikers.org/stats/.html.gz May 26 21:37:40 94.6% (200.2) : S3C2410 Timer Tick May 26 21:37:51 200 wakeups per second???? May 26 21:37:54 [logs] apt/ibot/infobot/jbot/purl all log daily to http://ibot.rikers.org// where channelname is html encoded ie: %23debian | lines that start with a space are not shown | some channels have stats at http://ibot.rikers.org/stats/.html.gz May 26 21:38:35 thanks, hrw May 26 21:38:41 hrw - That's the normal timer clock, not wakeup. May 26 21:38:44 counter May 26 21:38:44 http://londerings.novalis.org/wlog/index.php?title=Neo1973_counter (probability guess: 50%) P1: 2 days 19:10:36 (2.799 +-2.8 days) (1288;213) May 26 21:39:07 Elrond: I did not get it on x86/amd64 May 26 21:39:35 hrw: we don't use dynticks yet? May 26 21:39:40 CPU0 May 26 21:39:40 0: 275589 XT-PIC-XT timer May 26 21:40:06 hrw: ...unfortunately that means I'll have to wait few hours to see the logs :-(. May 26 21:40:18 100Hz, 2755 secs, looks like 46min uptime, which is close. May 26 21:40:27 ...that would certainly explain why openmoko gets warm while turned on. May 26 21:41:05 pavelm: http://ewi546.ewi.utwente.nl/tmp/hrw/t/#openmoko.log May 26 21:41:13 pavelm: whole month May 26 21:42:11 You don't have permission to access /tmp/hrw/t/#openmoko.log on this server May 26 21:42:31 pavelm: retry May 26 21:42:55 Thanks a lot! May 26 21:43:12 zecke: they are in kernel. but not enabled ;( will reboot May 26 21:44:09 Elrond: ok, here I am May 26 21:44:46 stefan_schmidt - Does that match: 83cbdc0b33d4d90c4e26da198bcbdbe8 uImage-2.6-moko10-r0_0_2079_0-fic-gta01.bin ? May 26 21:45:20 stefan_schmidt - Because I loaded it into RAM and u-boot complains about "Verifying Checksum ... Bad Data CRC". May 26 21:46:02 Elrond: moment May 26 21:48:09 83cbdc0b33d4d90c4e26da198bcbdbe8 uImage-2.6-moko10-r0_0_2079_0-fic-gta01.bin May 26 21:48:29 Elrond: So yes May 26 21:49:10 fscking usb serial May 26 21:50:14 * Elrond is reading dfu code in u-boot. ;-) May 26 21:51:19 Elrond: Le me try to flash it on bv4 May 26 21:52:56 Elrond: Works fine here May 26 21:53:42 BTW, if anybody has problems with dfu and has the neo connected to a hub please try it again directly connected to pc May 26 21:54:06 There are problems reported with a hub between May 26 21:54:27 I added a note to the wiki, but you guys are mostly faster for answers :) May 26 21:54:29 http://wiki.openmoko.org/index.php?title=Dfu-util&curid=2018&diff=8821&oldid=8281 May 26 21:55:39 stefan_schmidt - Okay... Story: I do not flash new kernels. I load them into the neo's RAM using DFU. Then I boot the RAM via u-boot's usbserial. Normally works nicely. This time I had to DFU it two times, because u-boot decided to read some fun from flash after the first DFU. May 26 21:56:38 Elrond: ah, ok May 26 21:57:12 stefan_schmidt - I made a photo of the NEO doing that. ;-) (excluding my second DFU load) May 26 21:57:22 stefan_schmidt - I'll upload it to flickr tomorrow. May 26 22:00:16 ...I have may19 image here.. and I'd like to install some software from angstrom. May 26 22:00:37 ...unfortunately ipkg update gives me 404. May 26 22:04:23 pavelm: normal May 26 22:04:37 I have own local feed May 26 22:05:29 and especially for stefan_schmidt: May 26 22:05:30 hrw: could we point ipkg at angstrom/something else working by default? May 26 22:05:41 ~lart neo for not recognizing my sim card May 26 22:05:41 * apt smacks neo up side the head with a clue-by-4 for not recognizing my sim card May 26 22:05:50 pavelm: openmoko has feed May 26 22:05:59 but in other place May 26 22:06:10 hrw-lart-counter++; May 26 22:06:11 I have it on bv3 but it decided to die again May 26 22:06:45 hrw-lart-counter May 26 22:07:08 pavelm: http://buildhost.openmoko.org/tmp/deploy/ipk/ and subdirs May 26 22:07:11 ~hrw-lart-counter May 26 22:07:22 zecke: heh, I'm not a IRC scripter May 26 22:07:40 stefan_schmidt: well, somehow this works over at the webkit channel ;) May 26 22:07:55 Perhaps a fun job for aloril2 May 26 22:08:19 * Elrond is installing python on "his" neo. ;-) May 26 22:08:34 Elrond: heh May 26 22:09:08 unfortunately angstrom's armv4t packages are incompatible with openmoko May 26 22:10:06 pavelm: openmoko is oabi May 26 22:10:28 ...and http://buildhost.openmoko.org/tmp/deploy/ipk/ does not seem to contain "basic" stuff like bash. May 26 22:10:54 pavelm: build it by yourself? May 26 22:11:24 stefan_schmidt - BTW: Your kernel also does not support the backlight on Bv03. ;-) May 26 22:11:48 hrw: I thought it is eabi. Ok, that means I should be able to "just run debian/arm" in chroot? May 26 22:12:20 pavelm: your and your way.... May 26 22:12:36 it reminded me 'please, do not follow pavelm way' post May 26 22:12:57 pavelm: grab mokomakefile and build what you want May 26 22:13:12 * pavelm is downloading debian/arm now :-) May 26 22:13:29 lamer May 26 22:13:34 i think it should be trivial to recompile all of openmoko with an eabi toolchain May 26 22:13:36 pavelm - If you want python, you can have my *.ipk. :-) May 26 22:13:56 balrog-kun: it is even easier if you just install angstrom on phone May 26 22:14:21 hrw: well, or any other distro May 26 22:14:27 hrw: How do I "just install angstrom" on phone? May 26 22:14:30 i think debian also has en eabi version May 26 22:14:33 hrw: I'd pretty much like to do that. May 26 22:14:43 balrog: Yes, but arm/eabi seems in beta now. May 26 22:14:44 balrog-kun: unofficial only May 26 22:14:57 pavelm: http://www.angstrom-distribution.org/images/ May 26 22:15:28 pavelm: things can stay in "beta" for years on debian ;) May 26 22:15:40 hrw: That gives me nice "not found" page. May 26 22:16:10 moment May 26 22:16:16 victory is near :} May 26 22:16:21 http://www.angstrom-distribution.org/unstable/images/ May 26 22:18:19 hrw: Thanks, I did not know about that one. I attempted to search for it but no luck. May 26 22:18:45 YES! May 26 22:18:58 The GPS on the P0 gives me some bytes! :-) May 26 22:19:01 there are no uclibc EABI ARM images May 26 22:19:11 Elrond: How did you do that? May 26 22:19:28 pavelm - Got a _very_ fresh kernel from stefan_schmidt :-) May 26 22:19:31 * hrw goes to bed - see you guys May 26 22:19:53 pavelm - checkin 2072 fixes a stupid bug in the kernel. It gives 1.5V on a 2.5V line to the gps. ;-) May 26 22:20:20 elrond: Ouch ;-). May 26 22:20:29 Elrond: congrats May 26 22:20:35 Could someone check "http://wiki.openmoko.org/wiki/Binary_compatibility"? I'm pretty new to oabi vs. eabi etc. May 26 22:21:23 pavelm: then please do not create such pages... May 26 22:21:24 elrond: Fast, launch gllin :-). May 26 22:21:37 pavelm: oabi can be hardfloat or softfloat May 26 22:21:44 softfloat has few versions May 26 22:21:54 arm OABI world is nightmare May 26 22:22:18 and endianness does not have nothing to say May 26 22:22:22 hrw: Such page needs to exist.. to explain the nightmare. May 26 22:22:26 arm/eabi on LE May 26 22:22:28 arm/eabi on BE May 26 22:22:41 pavelm: then find proper one and put only link in wiki May 26 22:22:52 Okay, so there are more flavours. Is arm/eabi actually used? May 26 22:23:07 nokia use eabi on tables May 26 22:23:09 hrw: I could not find proper one. Debian listed oabi/eabi/big endian. May 26 22:23:10 * hrw|gone goes to bed May 26 22:23:14 cu May 26 22:23:34 Night hrw :-) May 26 22:23:41 good night! May 26 22:23:51 pavelm - Officially I do not have gllin. ;-) May 26 22:26:23 elrond: Then you'll have to say "I feel like this would get a gps fix if I only had gllin". May 26 22:26:48 pavelm - ;-) May 26 22:26:59 pavelm - BTW: Do you know anything about the cmdline params of gllin? May 26 22:28:18 (script) wiki RecentChanges: [[Buying_Interest_List]] [[Binary_compatibility]] [[Main_Page]] [[Dfu-util]] May 26 22:29:03 elrond: gllin -low 3 is nice to test if communications work May 26 22:29:19 elrond: it also inits the gps; then do cat /tmp/nmeaNP, switch consoles, do gllin cold May 26 22:31:00 pavelm - Ahh. :-) May 26 22:31:24 I should get a longer usb cable to get the neo closer to the window first ;) May 26 22:33:13 elrond: do gllin -low 3 first. That does not need window. May 26 22:39:04 ...flashing angstrom now. May 26 22:45:41 ...missing libxcb :-(. May 26 22:47:00 * LuitvD is gone now May 26 22:47:09 bye y'all May 26 23:01:16 stefan_schmidt - BTW: Another way to get modules in the rootfs: let task-base depend on them. (I just wanted to make some space on my rootfs, and tried to remove unneeded modules and task-base kept me from doing that.) May 26 23:03:40 Elrond: ok, thanks for the hint May 26 23:05:42 Elrond: So... does the gps work now? May 26 23:06:24 pavelm - I have not tried with gllin. I have a .c from ph5, which does some basic tests. And that one worked. I guess, it's like gllin -low 3. May 26 23:06:54 Elrond: well EXTRA_MACHINES_RRECOMMENDS are used by task-base May 26 23:09:49 zecke - Ahh. ;o) May 26 23:13:08 Angstrom-x11-image-test-20070524-fic-gta01.rootfs.jffs2 seems to work better... May 26 23:27:16 ...yay, now i have bash :-) May 26 23:29:30 pavelm - But now gllin wont work any more, I'd guess. ;o) May 26 23:32:12 Elrond: Ouch :-( May 26 23:32:19 Elrond: Back to square one. May 26 23:32:27 * pavelm hates hates hates binary-only crap. May 26 23:36:40 pavelm - Try "~lart binary-only crap" ;o) May 26 23:44:26 wtf, I have almost working scrollbars :) May 26 23:46:31 zecke - What are you going? May 26 23:46:51 GtkLayout for the WebKit FrameView May 26 23:47:12 currently it was only able to scroll by scrolling the renderLayer May 26 23:48:42 Ahh. :) May 26 23:51:15 ~lart binary-only crap. May 26 23:51:27 ...I guess I could create small chroot just for gllin? May 26 23:51:30 * apt strangles binary-only crap. with a doohicky mouse cord May 26 23:51:59 pavelm: you could let it run in qemu on your neo May 26 23:52:01 ...that may be right thing, anyway. gllin _belongs_ in chroot jail. May 26 23:55:11 Anyone tried GPE^2? Screenshots look cool.... May 26 23:56:27 lol, the German Tagesschau speaker named the G8 participants hostake-takers :) May 26 23:56:31 pavelm: build it with OE May 26 23:58:16 "hostage-takers"? May 26 23:59:02 ...I guess I should just carry two fic phones with me. One to run gllin, and second to run the rest of code... May 27 00:01:18 *G* May 27 00:01:19 pavelm: right May 27 00:01:32 zecke - Can you please give the german word too? ;) May 27 00:04:34 so... I got python to work on openmoko (using angstrom), but python-gtk is broken :-( May 27 00:04:45 File "/data/build/koen/OE/build/tmp/angstrom/work/armv4t-angstrom-linux-gnueabi/python-pygtk2-2.10.3-ml3/image/usr/lib/python2.4/site-packages/gtk-2.0/gtk/__init__.py", line 48, in ? May 27 00:04:46 ImportError: No module named cairo May 27 00:05:38 ouch, this is worth than a rpath May 27 00:05:59 pavelm: you might have luck taking this up with mickeyl May 27 00:08:03 pavelm - Get python-cairo installed. May 27 00:08:33 zecke - It's not mickey's job, as pavelm is not using openmoko there. ;) May 27 00:11:00 Elrond: ipkg install python-pycairo seems to be close. May 27 00:12:18 Elrond: well, this python edition says 'ml'. But yes I don't want mickeyl to overload himself even more May 27 00:13:10 any git-format-patch expert around? May 27 00:13:50 elrond: I installed python-pycairo, but still get same message. Do I need to run some kind of python-update-your-modules? May 27 00:16:07 pavelm - In theory not. I don't know angstrom too well. May 27 00:16:35 zecke - What did the Tagesschau speaker say in german? May 27 00:18:12 Elrond: "Die G8 Geiselnehm Entschuldigung die G8 Teilnehmer" May 27 00:18:50 *biggrin* :-) May 27 00:20:44 Good night people :-) May 27 00:21:07 good night, elrond! May 27 00:28:18 (script) wiki RecentChanges: [[User_talk:Idarwin]] May 27 01:28:30 Elrond: did you try your neo in the sun? May 27 01:37:04 Elrond, stefan_schmidt: re getting modules in the rootfs - use RRECOMMENDS instead of RDEPENDS - then ipkg doesn't complain about task-base not being complete if someone wants to ipkg remove those modules. May 27 01:38:25 pavelm: http://logs.nslu2-linux.org/livelogs/ has live logs, current 24 hours and past 24 hours in -prev, and archive in the subdir. **** ENDING LOGGING AT Sun May 27 02:59:56 2007