**** BEGIN LOGGING AT Wed Sep 15 02:59:57 2010 Sep 15 08:53:30 morning Sep 15 16:31:27 * dos11 is installing SHR on N900 once again to check, if guide is 100% correct :D Sep 15 16:31:54 dos1: with updated watchdog and xserver-common? Sep 15 16:32:01 mrmoku: yup Sep 15 16:32:05 great Sep 15 16:32:22 works fine without r&d :) Sep 15 16:32:30 and what does shr looks like on th eN900? awsome or not? :) Sep 15 16:32:46 GarthPS: it boots faaaaaaast :) Sep 15 16:32:53 lol hom much ? Sep 15 16:32:56 though still lots of stuff to do Sep 15 16:33:10 did not measure yet Sep 15 16:33:13 phon already works ? Sep 15 16:33:16 no Sep 15 16:33:25 no phone, no gps, no wifi via fso Sep 15 16:33:31 :) Sep 15 16:33:34 or not Sep 15 16:33:35 but looks interesting :) Sep 15 16:33:38 br Sep 15 16:33:39 b Sep 15 16:34:24 GarthPS: http://www.youtube.com/watch?v=ywkWbb_BetI Sep 15 16:35:51 wo 22"! :p Sep 15 16:35:56 thx dos1 Sep 15 16:36:42 sor 18" booting Sep 15 16:36:45 sorry Sep 15 16:37:39 DocScrutinizer: how was called that kernel module for battery status on n900? Sep 15 16:38:10 not bq27x00_battery one, that better one ;) Sep 15 16:38:28 err what? Sep 15 16:38:39 there's no better one Sep 15 16:39:32 DocScrutinizer51: you said some time ago to GNUtoo that bq27x00 is just stupid sysfs interface for i2c data, or something like that Sep 15 16:39:45 and mentioned some other Sep 15 16:39:56 there's a number of scripts based on my bq27kfdetail script and using I2Ctools to access chip directly Sep 15 16:40:39 bq24150 - built by jacekowski Sep 15 16:40:45 (found in logs) Sep 15 16:41:41 or is it only for charging? Sep 15 16:42:06 (i mean - for using sysfs node instead of i2ctools to initiate charging) Sep 15 16:53:19 freesmartphone.org: 03Frederik.Sdun 07cornucopia * r6430e572a81b 10/tools/mdbus2/src/main.vala: Sep 15 16:53:19 freesmartphone.org: mdbus2: strip busnames, paths, interfaces to avoid Sep 15 16:53:19 freesmartphone.org: assertions on malformed parameters Sep 15 16:53:19 freesmartphone.org: 03Frederik.Sdun 07cornucopia * r2e922f3c8273 10/tools/mdbus2/ (data/mdbus2 src/main.vala): mdbus2: Add pretty print Sep 15 16:53:19 freesmartphone.org: 03Frederik.Sdun 07cornucopia * r5b468125336a 10/ (5 files in 4 dirs): Merge branch 'master' of fso:cornucopia Sep 15 17:04:19 dos1: bq27x00 is battery gauge, bq24150 is LiIon charger Sep 15 17:05:51 comparing to OM Neo FR, bq27200 is the gta02-smart-battery while bq24150 does basically same (for battery) as FR PCF50633 Sep 15 17:26:21 DocScrutinizer: ok, thanks for explaining :) Sep 15 17:50:51 GNUtoo|laptop: just found this patch within the meego patches :) : Sep 15 17:50:52 # Fix wl1251 scanning while associated linux-2.6.36-wl1251-fix-trigger-scan-timeout-usage.patch \ Sep 15 17:51:13 mrmoku, ok Sep 15 17:51:16 nice Sep 15 17:54:50 mrmoku: we need udev, watchdog, gtk-immodule-xim, autoloading g_nokia and screen calibration on first boot in default image Sep 15 17:55:04 mrmoku: i think those are absolutely basics Sep 15 17:56:06 dos1: ack Sep 15 17:56:37 mrmoku: could you take care of them? Sep 15 17:57:47 freesmartphone.org: 03seba.dos1 07cornucopia * r8b7652ea9828 10/fsogsmd/conf/nokia_n900/fsogsmd.conf: fsogsmd: use dummy modem on n900 Sep 15 17:58:20 GNUtoo|laptop: weren't you looking into adding udev for nokia900 only? Sep 15 17:58:40 mrmoku, I wonder what's the best approach Sep 15 17:58:48 maybe look at kernel version Sep 15 17:59:18 like if (! kenrel_has_devtmpfs) Sep 15 17:59:22 add_udev Sep 15 17:59:32 I'll try to do it very soon Sep 15 18:01:19 ok Sep 15 18:04:20 mrmoku, shouldn't we move the installation page on shr wiki? Sep 15 18:04:47 GNUtoo|laptop: no opinion on that, probably yes Sep 15 18:04:53 ok Sep 15 18:04:59 because dream installation is there Sep 15 18:05:07 and it's specific to SHR Sep 15 18:05:15 for instance there is a debian install method too Sep 15 18:05:21 and it's very different Sep 15 18:05:26 and SHR != FSO Sep 15 18:07:19 mrmoku, what images do we have: Sep 15 18:07:24 *om-gta01 Sep 15 18:07:27 *om-gta02 Sep 15 18:07:30 *htcdream Sep 15 18:07:32 *n900 Sep 15 18:07:37 *palm-pre? Sep 15 18:07:39 *ezx? Sep 15 18:09:11 mrmoku: about ts calibration: i think there is bug in /etc/X11/Xsession.d/89xTs_Calibrate_xinput_calibrator Sep 15 18:09:57 it runs xinput_calibrator_once.sh only, if /etc/pointercal.xinput exists Sep 15 18:13:31 GNUtoo|laptop: dunno which of those work though Sep 15 18:13:49 dos1: heh and the calibrator refuses to run when it exists :P Sep 15 18:13:58 yes but what images do we have? Sep 15 18:14:11 pre and ezx have current up to date images? Sep 15 18:14:14 mrmoku: but it does ". /etc/pointercal.xinput" Sep 15 18:14:20 mrmoku: and 89xTs... does not Sep 15 18:15:30 * DocScrutinizer is wondering why anybody would want g_nokia Sep 15 18:15:45 maybe we should use the standard one Sep 15 18:18:56 aiui g_nokia is a xinetd for USB, loading g_ether or whatever else is needed, after connection of a host is detected and user has selected mass storage or nokia_suite from menu on N900 Sep 15 18:19:21 dos1: that's correct Sep 15 18:21:29 also my comment to ts-calib on first boot is unchanged: bad idea, as devices usually come with a pretty perfect calibration Sep 15 18:21:57 afaik Sep 15 18:22:06 dos1: you need calibration only if you need it, and if you need it you will run xinput_calibrator_once.sh by hand Sep 15 18:22:34 Heinervdm: ok then... so we need to provide somehow correct calibration with image :P Sep 15 18:22:42 dos1: yes Sep 15 18:23:12 Heinervdm: wait... Sep 15 18:23:14 i think the default configuration should go into xorg.conf Sep 15 18:23:16 Heinervdm: no, i don't think that's correct Sep 15 18:23:27 Heinervdm: now when i'm calibrating it by hand, it's not used at all Sep 15 18:23:35 (i mean - my own calibration) Sep 15 18:23:55 for FR that script is working... Sep 15 18:24:33 Heinervdm: FR has some modification of this script which uses fsoraw Sep 15 18:25:40 dos1: but the pointercal.xinput exists now? Sep 15 18:26:09 Heinervdm: when pointercal.xinput exists, nothing is called at all, even commands which are in pointercal.xinput Sep 15 18:27:32 strange Sep 15 18:28:20 Heinervdm: after removing that "if" from 89xTs... it works Sep 15 18:33:13 GNUtoo|laptop: built images for all of those apart from ezx Sep 15 18:33:20 DocScrutinizer: about g_nokia - there is no g_ether built AFAIK Sep 15 18:33:26 ok Sep 15 18:33:35 dos1, if you build your knerel.... Sep 15 18:33:39 and modprobing just g_nokia makes USB networking working Sep 15 18:33:48 I'll try g_ether Sep 15 18:33:59 I was struggeling with DSP's Sep 15 18:34:12 GNUtoo|laptop: yup, we probably want g_ether, as it's more standard Sep 15 18:34:15 DocScrutinizer: and on maemo there is only g_nokia loaded... even with active USB network Sep 15 18:34:30 but now we have only g_nokia Sep 15 18:38:05 what about compiling and flashing your own kenrel? Sep 15 18:38:12 maybe before udev Sep 15 18:38:22 and before everything I should continue to fight multiboot Sep 15 18:39:26 GNUtoo|laptop: do you have a pointer to some multiboot docs? Sep 15 18:39:58 no but here what I know Sep 15 18:40:09 from #nitdroid and #nitdroid-help Sep 15 18:40:18 you must use the uname -a name Sep 15 18:40:23 in multiboot Sep 15 18:40:33 but where to put the kernel? Sep 15 18:40:39 btw I restored maemo Sep 15 18:40:47 I had to flash a fiasco image Sep 15 18:40:49 with -F Sep 15 18:40:53 and not a kenrel image Sep 15 18:41:01 so I am now able to try multiboot Sep 15 18:41:12 kenrel goes into: Sep 15 18:41:16 /boot/multiboot Sep 15 18:41:19 or something like that Sep 15 18:41:30 I'll try to find if there is some docs Sep 15 18:41:41 googling for multiboot is difficult :P Sep 15 18:41:58 http://wiki.maemo.org/index.php?title=Advanced_booting Sep 15 18:42:13 the docs on this subject is not great Sep 15 18:42:53 btw I was able to mount the flashes Sep 15 18:43:04 the internal os in ubifs Sep 15 18:43:22 the external 25/30G are fat extmmc Sep 15 18:43:42 I must find a way to document everything on SHR wiki Sep 15 18:58:22 mmhh... my kernel has g_nokia _and_ g_ether :) Sep 15 18:59:08 dos1: i've to admit Jeremy's reply is actually funny. Sep 15 18:59:18 ;) Sep 15 19:03:34 dos1: strange, I'm quite sure I've seen g_ether somewhere on N900 Sep 15 19:03:53 either maemo or power-kernel, or meego Sep 15 19:04:22 aah, mrmoku found it already :-) Sep 15 19:20:19 DocScrutinizer, do you use multiboot? Sep 15 19:20:53 nope Sep 15 19:21:03 not yet Sep 15 19:21:23 DocScrutinizer: no, with the meego n900 kernel I get both... not on maemo Sep 15 19:21:27 at least not loaded Sep 15 19:22:00 Nokia-N900-51-1:/etc/multiboot.d# find /lib/modules -name g_\* Sep 15 19:22:00 /lib/modules/2.6.28-omap1/g_nokia.ko Sep 15 19:22:00 /lib/modules/2.6.28-omap1/g_zero.ko Sep 15 19:22:00 /lib/modules/2.6.28-omap1/g_file_storage.ko Sep 15 19:22:00 /lib/modules/2.6.28-omap1/g_softupd.ko Sep 15 19:22:19 :nod: so prolly it was meego where I've seen g_ether. Would take me by surprise if they'd use g_nokia there Sep 15 19:22:20 which certainly does not mean we cannot use g_ether instead though :) Sep 15 19:22:31 hehe, yes Sep 15 19:24:06 for the find, a modinfo g_ether also didn't succeed here ;-) Sep 15 19:25:10 ok Sep 15 19:25:28 mrmoku, did you succeed with multiboot? Sep 15 19:26:10 GNUtoo|laptop: did not yet further investigate... Sep 15 19:26:15 have built a kernel meanwhile though Sep 15 19:26:21 ok Sep 15 19:26:28 did you add all booting support? Sep 15 19:26:38 ext2/3 Sep 15 19:26:43 let me check Sep 15 19:26:44 omap hsmmc Sep 15 19:26:49 mmc block devices Sep 15 19:27:00 framebuffer console for testings Sep 15 19:27:31 meego config... ext2,3 btrfs :P Sep 15 19:37:08 GNUtoo|laptop: do you know where the multiboot source is? Sep 15 19:53:36 mrmoku, maybe apt-get source Sep 15 19:53:47 too bad I'm too tired theses days Sep 15 20:05:50 Hi, is ubifs good choice for Freerunner? I saw SHR has ubifs images - are there any problems or does it work flawlessly? Sep 15 20:06:50 radekp: I never tried ubifs on FR myself... Jama was playing with it a lot I think... and I remember Weiss had some problems Sep 15 20:07:41 mrmoku: ahh thanks i will talk to Martin then, is he still here? Sep 15 20:07:55 radekp: he is in US for vacations Sep 15 20:08:10 mrmoku: do you know for how long? Sep 15 20:08:29 radekp: forgot :/ Sep 15 20:08:39 oki, thanks Sep 15 20:09:15 yw Sep 15 20:25:00 gena2x: hi, tested your patch, it works great, 12mA in suspend with qtmoko - so it's same as 2.6.29 now Sep 15 20:27:24 radekp: did you measure that with a multimeter? Sep 15 20:28:08 lindi-: no just with current_now sysfs node Sep 15 20:29:29 radekp: with 2.6.29 I see 3 mA there Sep 15 20:29:49 (with everything off) Sep 15 20:29:56 lindi-: but that is with modem turned on in deep sleep mode Sep 15 20:30:17 radekp: can you try without modem? Sep 15 20:30:32 lindi-: hmm i would have to reboot Sep 15 20:30:43 why? Sep 15 20:30:48 but i can try... Sep 15 20:34:19 btw i wonder if really everything is properly turned off during suspend (whatever kernel) Sep 15 20:35:08 some old featured phones could last on one battery charge for 14 days and they even dont have turned off CPU and display Sep 15 20:35:41 because they show time on display Sep 15 20:35:58 and they have crappy 700mAh battery Sep 15 20:36:59 freerunner should last at least more then 30days if we expect same modem and if we take into account that is has twice as big battery Sep 15 20:38:04 radekp: if the time is updated only once every 60 seconds the cpu can easily sleep in between Sep 15 20:39:13 but still it's more then FR which should consume 0 Sep 15 20:39:45 maybe they have much better modem... or i have no explanation Sep 15 20:40:03 with 3 mA the battery should in theory last 16 days Sep 15 20:43:59 lindi-: i have 9mA with 2.6.34 and modem turned off Sep 15 20:44:19 lindi-: but it's same as with 2.6.29 here Sep 15 20:44:23 odd Sep 15 20:44:30 I consistently get 3 mA Sep 15 20:44:38 lindi-: i wonder how you can have 3 mA - which kernel do you use? Sep 15 20:44:52 lindi-: maybe i can flash your kernel to check it Sep 15 20:44:58 wiki.openmoko.org/wiki/user:lindi has the git hash and three patches Sep 15 20:47:00 lindi-: oki, but i think i will leave it for tomorrow ;-) Sep 15 21:46:55 <[Rui]> playya_: hi Sep 15 21:47:24 <[Rui]> playya_: I'm not sure I understood your message, I saw those characters right enough?... Sep 15 21:56:58 mrmoku|away, ping Sep 15 21:58:02 <[Rui]> daniele_athome: hi, how's mokosuite going? Sep 15 21:58:14 <[Rui]> got a good set of screenshots? :) Sep 15 21:58:15 hi [Rui] very well thank you :) Sep 15 21:58:29 [Rui], someone from time to time post something to scap Sep 15 21:58:45 <[Rui]> daniele_athome: how stable/working is it? Sep 15 21:58:51 [Rui], of course: http://wiki.openmoko.org/wiki/Mokosuite Sep 15 21:58:58 radekp: yep, I found UBIFS (on NAND) to be very unstable.. Sep 15 21:59:08 almost every unclean shutdown hosed the filesystem Sep 15 21:59:22 well it has call log, contacts (not manageable yet, sqlite db must be created before) Sep 15 21:59:26 phone calls work Sep 15 21:59:43 window manager is pretty much stable Sep 15 21:59:52 but the version found on shr is already old i'm sorry :( Sep 15 22:00:01 <[Rui]> daniele_athome: still a different db than opim? Sep 15 22:00:05 i've made some fixes and modifications these days Sep 15 22:00:14 [Rui], don't worry i'm already porting to opimd :) Sep 15 22:00:21 it was just a test Sep 15 22:00:43 <[Rui]> daniele_athome: ah nice, tell me when you get that part , then I might try it out Sep 15 22:01:21 <[Rui]> daniele_athome: I'm currently still recovering my hard disk so my devel disk is outside the portable case so my old disk is accessible while I get my stuff on the new one, so I can't git it yet Sep 15 22:06:02 Weiss: which version have you tested? the one in 2.6.32 or 2.6.34? Sep 15 22:06:16 Weiss: maybe its fixed now? Sep 15 22:07:12 <[Rui]> Weiss: does ubifs work well on the µ-sd card? better than ext2? Sep 15 22:07:50 is ubifs a wear leveling filesystem? Sep 15 22:08:04 i thought so Sep 15 22:08:19 [Rui]: it's flash filesystem for MTD devices, uSD card cant use it, if am i right... Sep 15 22:08:47 <[Rui]> radekp: I thought it would help with µ-sd weardown... Sep 15 22:09:33 yeah flash fs sorry Sep 15 22:10:16 radekp is right, it wouldn't make sense because a flash filesystem would handle block corruption and other things, which is already handled by the uSD itself Sep 15 22:10:18 [Rui]: i dont think so - uSD do not expose wear leveling logic - they behave as normal harddisks Sep 15 22:10:27 <[Rui]> ok Sep 15 22:10:57 <[Rui]> didn't know that Sep 15 22:11:11 <[Rui]> it's always good to learn new stuff :) Sep 15 22:18:40 mrmoku|away, i've posted everything on shr-devel :) we'll talk tomorrow Sep 15 22:18:43 goodnight guys :) **** BEGIN LOGGING AT Wed Sep 15 23:16:14 2010 Sep 15 23:17:50 Weiss: which can be quite hard to guarantee on Freerunner ;) Sep 15 23:19:35 hehe **** ENDING LOGGING AT Thu Sep 16 02:59:57 2010