**** BEGIN LOGGING AT Thu Oct 04 02:59:56 2007 Oct 04 03:00:12 That's what the 6gb sd card is for, eh? Oct 04 03:00:17 sagacis: but the good news is going forward, we'll probably only need the .net 2.0 profile, and only a subset of the components Oct 04 03:00:26 True Oct 04 03:00:38 sagacis: so, I'm guessing 10-20MB should be plenty Oct 04 03:02:49 cbrake: Is it all 1.2? Oct 04 03:03:52 sagacis: mono 1.2.5.1 (the latest), which implements both the 1.x and 2.x (much of it) .net profiles Oct 04 03:04:16 goodness for sure Oct 04 03:04:31 sagacis: agreed :-) Oct 04 03:04:38 hm, you can't run openmoko-dialier in an X forwarded ssh session, can you? does it confuse d-bus session stuff? Oct 04 03:06:36 orospakr: I haven't tried the dialer Oct 04 03:06:57 hm, it stays on the display now, but calls go nowhere Oct 04 03:07:01 I wonder where I can get logging info Oct 04 03:08:11 heh, gsmd is spinning and consuming 80% of the CPU. Oct 04 03:10:18 /tmp/gsm.log Oct 04 03:14:13 "If we can design things that are somewhat emotionally engaging, it doesn't have to be as reliable." -- http://www.msnbc.msn.com/id/21102202/ Oct 04 03:14:33 Isn't that apple's core tenet from the beginning? Oct 04 03:17:03 sagacis: hey, it works for politicians :( Oct 04 03:24:22 cesarb: audio should be fixed in the qtopia snapshpt update Oct 04 03:27:59 the handset audio worked flawlessly for me today. Oct 04 03:28:41 Hey, I just called myself on and the phone rang. Oct 04 03:28:48 It sounds like an old analog bell Oct 04 03:32:09 so, if I do a ipkg upgrade on OM2007.2 (august) anyway, where should I get an appropriate kernel image to flash? Oct 04 03:37:13 noob debug board question, why is it showing up as several usbdev1.1x devices instead of ttyUSBx? Oct 04 03:39:49 ftdi_sio Oct 04 03:39:52 modprobe it Oct 04 03:41:33 yep, thank you Oct 04 03:43:37 morning...i'm not able to build fontconfig. Every time I get this: unable to get local issuer certificate Oct 04 03:44:19 NOTE: Task failed: Fetch failed: https://stage.maemo.org/svn/maemo/projects/haf/trunk/fontconfig/device_symbols.h Oct 04 03:46:06 there is a hint: i should use: "--no-check-certificate" - where do i have to add this option? Oct 04 04:41:48 moin Oct 04 04:54:57 (script) wiki RecentChanges: [[Wish_List]] [[P1_Owners]] Oct 04 05:16:10 Which MokoMakefile targets should I use to build and flash my Neo with the the kernel and rootfs images? Oct 04 05:17:13 I usually use "make update-makefile;make update;make setup;make openmoko-devel-image" to build Oct 04 05:17:26 To update everything first, then build the image Oct 04 05:18:10 ah, OK! Oct 04 05:18:12 Then flash with make flash-neo-local; make flash-neo-kernel-local Oct 04 05:21:45 CM, shouldn't you do "make setup" before you do "make update", or it doesn't matter? http://wiki.openmoko.org/wiki/MokoMakefile#Updating_the_environment Oct 04 05:22:54 Hmm.. Don't know actually. Just been using it like that since rwhitby put it on his homepage the first time :) Oct 04 05:23:19 CM, thanks Oct 04 05:26:43 mbuf: Hmm.. Looking at the targets in the makefile I guess i should do the setup first. Shouldn't matter much one you've done make setup at least once Oct 04 05:26:49 setup: check-generation setup-patches setup-bitbake setup-mtn setup-openembedded setup-openmoko setup-config setup-env Oct 04 05:26:52 update: check-generation update-mtn update-patches update-openembedded update-bitbake update-openmoko Oct 04 05:27:07 That's what each command does :) Oct 04 05:27:59 abraxa__: Good morning, how's it going with the mediaplayer? :) Oct 04 05:34:28 CM: Good morning! It's going well, though not as fast as I wished. You know, the all-too-common "too much food on the platter" problem :) Oct 04 05:34:57 Oh yes, I know.. :-P Oct 04 05:35:32 Currently I'm working on a new GUI so that's why there haven't been any commits lately - it'll be one big commit when it's done Oct 04 05:35:45 Was some guy here two days ago asking about a segfault when loading the playlist. Told him to file a bug report, but I don't know what happened with that Oct 04 05:36:10 Ah, nice. Any teaser screenshots? ;) Oct 04 05:36:10 I didn't see any bug reports about that, unfortunately... hm Oct 04 05:36:47 http://dar-clan.de/mediaplayer_compact.jpg / http://dar-clan.de/mediaplayer_expanded.jpg Oct 04 05:37:22 Ohh.. Nice :D Oct 04 05:37:49 Yeah, let's hope GTK+ lets me do it the way I want to ;) Oct 04 05:37:51 What's the three dots left of the control buttons? Oct 04 05:38:21 They're the "expand" button - when clicked it reveals the bottom row of controls Oct 04 05:38:39 I figured that shuffle/repeat/settings/whatev' don't need to be visible at all times ;) Oct 04 05:40:45 *nod* Oct 04 05:40:54 Good call, the less clutter the better :) Oct 04 05:41:25 Yeah hehe Oct 04 05:43:47 abraxa__: http://rafb.net/p/ywQtTt73.html Oct 04 05:44:22 Ahh Oct 04 05:44:34 Empty playlist maybe? Oct 04 05:45:00 Yeah, that's because the player in the latest buildhost snapshot has the bug where it doesn't check if /home/root/playlists even exists before trying to create files there Oct 04 05:45:13 So it segfaults Oct 04 05:45:31 Ah, ok. Good that it's known then :) Oct 04 05:45:43 Yeah, and fixed in SVN for quite a while ;) Oct 04 05:46:02 * CM is #openmoko's information relay station Oct 04 05:46:06 lol Oct 04 05:46:19 Thanks for letting me know btw ;3 Oct 04 05:46:47 np, just trying to keep track of all problems and changes in om :) Oct 04 06:36:29 rwhitby: My first attempt at building with MokoMakefile says: Oct 04 06:36:32 File "/home/gnu/src/moko/bitbake/lib/bb/event.py", line 67, in fire Oct 04 06:36:32 if tmpHandler(event) == Handled: Oct 04 06:36:32 File "tmpHandler(e)", line 31, in tmpHandler Oct 04 06:36:32 ValueError: substring not found Oct 04 06:36:32 make: *** [openmoko-devel-image] Error 1 Oct 04 06:37:37 so, I saw my first iphone today (I'm in canada). Nice looknig piece of kit, and it got me thinking and wondering if anyone else is thinking the same. If FIC want to be anything other than a niche manufacturer, they're going to have to work a bit more on the hardware design :/ Oct 04 06:38:13 I'd never ever ever buy an iphone, because it's so locked down, but it's still a much better looking piece of kit Oct 04 06:38:14 (I'm on Ubuntu) Oct 04 06:38:18 doc|work, the iPhone is just a toy...not the same market Oct 04 06:38:34 "piece of kit"? Oct 04 06:38:39 codyl: I know, but have fic said they only want to be niche? I'd be amazed if they had Oct 04 06:38:40 a shit load of money was put into iphone Oct 04 06:38:50 over the course of what...5 years? Oct 04 06:38:59 guaqua: that doesn't mean it's not competition. Oct 04 06:39:00 doc|work: If FIC wants this product to sell, it'll have to run more than two hours on a charge -- and have a wall-wart that actually charges it. Oct 04 06:39:20 gnu: right, but that's a fundamental thing which will have to be fixed anyway Oct 04 06:39:22 gnu, are you referring to the iphone or neo? Oct 04 06:39:30 gnu: otherwise it's utterly doomed, open or not Oct 04 06:39:37 gnu, i havent had any trouble with my neo's power Oct 04 06:39:39 I'm talking past that Oct 04 06:39:51 gnu: It's a bitbake issue, not makefile, but I don't know what's wrong. Do you have all dependencies? pysqlite2 for example? Oct 04 06:40:04 i think FIC is bigger copr than a niche Oct 04 06:40:09 corp that is Oct 04 06:40:11 ljp: right Oct 04 06:40:28 CM: I loaded all dependencies mentioned on the MokoMakefile web page (and even added a few that were poorly documented) Oct 04 06:40:49 design *is* important Oct 04 06:41:18 my first thought seeing the neo was "I want one but, it's a big... ugly" Oct 04 06:41:24 s/big/bit Oct 04 06:41:30 CM: python-pysqlite2 is already the newest version. Oct 04 06:41:36 a bit random question (but FIC is in Taiwan!): Is there an open-source tool for converting a document from traditional to simplified ? (utf-8) Oct 04 06:42:14 the neo isnt a fashion statement, its a hackers phone Oct 04 06:42:30 lol...its * useful * Oct 04 06:42:33 gnu: Have you tried to rebuild bitbake? make clobber-bitbake;make setup-bitbake Oct 04 06:42:43 ljp: sounds like a niche :/ Oct 04 06:43:03 CM: will try that. Oct 04 06:43:37 CM: same output Oct 04 06:43:43 :( Oct 04 06:44:06 gnu: Maybe XorA knows more, or wait for rwhitby Oct 04 06:47:26 the neo is not th eonly phone FIC produces. they are a big company Oct 04 06:49:21 Couldn't they produce 100.000 neo's per month if they wanted? Oct 04 06:49:21 what would be a stronger fashion statement these days than having a hacker phone :-) Oct 04 06:49:43 tuukkah: To actually be able to call with it too? :-P Oct 04 06:49:44 tuukkah: having an attractive hacker phone? :) Oct 04 06:49:45 are there any specific ports (other than FTP) that MokoMakefile uses to do updates? Oct 04 06:49:46 * CM hides Oct 04 06:50:28 CM: hah Oct 04 06:50:52 what's the likelihood of these gsmd/power issues getting sorted before nov? :/ Oct 04 06:50:56 CM, it wouldn't be a true hacker phone if it worked as an appliance when first shipped:-) Oct 04 06:51:02 CM: Two other folks have had this problem (in Google); but neither one gives a clean answer about how to get past it. Oct 04 06:51:18 gnu: What version of python do you have? Oct 04 06:51:33 * CM is using Python 2.5.1 Oct 04 06:51:38 CM: gnu@om:~/src/moko$ python Oct 04 06:51:38 Python 2.5.1c1 (release25-maint, Apr 12 2007, 21:00:25) Oct 04 06:51:38 [GCC 4.1.2 (Ubuntu 4.1.2-0ubuntu4)] on linux2 Oct 04 06:51:50 Hmm.. Same as me then Oct 04 06:52:12 doc|home, if the phone is not attractive, then hacking just isn't strong enough in you ;-) Oct 04 06:52:15 the level of core devs being active here has been significantly lower than before Oct 04 06:52:27 tuukkah: different kind of attractive :) Oct 04 06:52:27 so i'm guessing they are busy somewhere else Oct 04 06:52:53 that might mean they've got down to business with new hw or there is going to be a massive delay Oct 04 06:53:03 I hope they are debugging the new hardware. If not, we're all fucked. Oct 04 06:53:09 tuukkah: I'm buying one as soon as it's released to the public :) Oct 04 06:53:17 considering they haven't been outright lying to the community before, i'd think the first option is more likely Oct 04 06:53:58 doc|home, the less it looks like yet another nokia phone, the more people will learn about it Oct 04 06:54:05 gnu: What shell btw? Oct 04 06:54:12 CM: bash Oct 04 06:54:30 Hmm.. Were some problems with dash before, but I think they were fixed Oct 04 06:54:32 tuukkah: you know what, the first time I saw it, I thought it looked like a nokia phone released in about 2003. The rounded bottom and top. Oct 04 06:54:40 gnu: Sorry, I'm out of ideas Oct 04 06:54:47 CM: I did the update-alternatives thing to get rid of dash. Oct 04 06:54:56 CM: Thanks for working on it... Oct 04 06:54:57 (script) wiki RecentChanges: [[P1_Owners]] Oct 04 06:56:05 doc|home, really? i have to go see some old photos then Oct 04 06:56:26 tuukkah: yeah, i'm looking now but that's way too long ago for me to remember the model number Oct 04 06:56:44 tuukkah: http://www.phonenews.com/phones/gsm/nokia/images/nokia-3600.jpg Oct 04 06:56:51 but I don't think that was it Oct 04 06:57:11 i think i know what you have in mind now... Oct 04 06:57:54 here we go http://img.hc360.com/home-a/info/images/200704/20070417l52.jpg Oct 04 06:58:05 even has similar chunkiness Oct 04 06:58:26 guaqua: we are here Oct 04 06:58:51 XorA: didn't know you were with fic :) Oct 04 06:58:55 good to know :) Oct 04 06:59:04 doc|home, there's even a current model 7373 Oct 04 06:59:04 guaqua: I started on monday Oct 04 06:59:17 ooh Oct 04 06:59:22 congrats Oct 04 06:59:30 guaqua: other core devs are either on much earned holiday of working hard on code ATM Oct 04 06:59:42 tuukkah: that is a 7373, not sure if it's current or not, I don't really follow phones much other than the neo :) Oct 04 07:00:57 people just get frustrated...not very rational. i hope the ones on holiday get enough of it, no use getting people burnt out Oct 04 07:01:49 doc|home, right, might be a couple of years old, but they still sell it and 7370 Oct 04 07:02:15 fair enough Oct 04 07:02:25 guaqua: Harald has been going through bugzilla a lot lately :) Oct 04 07:03:03 that's a good sign Oct 04 07:06:52 ERROR: '.../moko/openembedded/packages/gcc/gcc-cross-initial_4.1.2.bb' failed Oct 04 07:06:54 :( Oct 04 07:07:13 doc|home, one thing i hear people comment here is how "asian" and "plastic" it looks, and that the design is very foreign to them Oct 04 07:07:35 XorA: Trying my first MokoMakefile build, instantly getting ValueError on event.py line 67 the first time it tries to run bitbake. Any clues? Oct 04 07:08:02 hmm Oct 04 07:08:02 i wonder if this is a boog no-one else has Oct 04 07:10:26 the neo charges while plugged in but switched off right? Oct 04 07:10:39 Psi, seems to charge while on... Oct 04 07:10:46 gnu, the error message is very weird to be caused by that line isn't it? Oct 04 07:10:51 ya, it wont turn on, i think its flat Oct 04 07:11:00 Psi, give it about 10 mins..it does that Oct 04 07:11:08 ok :) Oct 04 07:11:19 tuukkah: The whole thing is about 15 lines, starting "ERROR: no files to build." and ending with a Python stack trace. Oct 04 07:12:08 Psi: Sometimes you have to take out the battery for a minute or so, then put it back in. That's to turn off the gsm modem completely so that it charges faster. Oct 04 07:12:11 Psi, you may need to remove the battery for a while so that the subsystems shut down too and the charge level actually starts to increase Oct 04 07:12:19 :P Oct 04 07:12:21 CM :-) Oct 04 07:12:41 * Psi takes batt out Oct 04 07:12:49 gnu: not a late enough version of sqlite or pysqlite Oct 04 07:12:57 gnu, did you have it all pasted somewhere? Oct 04 07:13:54 tuukkah: http://pastebin.ca/725266 Oct 04 07:15:17 gnu, ok, makes more sense to me there Oct 04 07:15:57 XorA: Have latest sqlite3 and python-pysqlite2 on Ubuntu; do I need something from CVS? Oct 04 07:16:25 gnu: no, ones from fiesty should be enough Oct 04 07:17:22 The "ERROR: no files to build" makes me wonder; where should it be finding .bb files? I don't see any. Oct 04 07:18:01 did you make setup, make update etc? Oct 04 07:18:13 XorA: yep Oct 04 07:18:25 no idea then Im sorry Oct 04 07:18:33 XorA: But it's my first time building openembedded, using bitbake, etc; Oct 04 07:18:44 Not a very friendly environment for newbies. Oct 04 07:19:08 MokoMakefile was supposed to make it easy :-) Oct 04 07:19:17 (like having to debug autoconf before you can run a *%&$^# make) Oct 04 07:19:23 it normally just works, must be something were missing here Oct 04 07:20:14 gnu: do you have any symlinks? Oct 04 07:20:29 Does it always say "NOTE: Parsing finished. 0 cached, 0 parsed, 0 skipped, 0 masked." for you? Oct 04 07:20:31 CM: Nope Oct 04 07:21:12 XorA: Mokomakefile seems to be working; it's bitbake that's scrogged... Oct 04 07:21:38 gnu: well it looks like you dont have a checked out copy of OE Oct 04 07:22:13 XorA: gnu@om:~/src/moko$ ls Oct 04 07:22:13 bitbake errs OE.mtn openembedded patches stamps Oct 04 07:22:13 build Makefile OE.mtn.bz2 openmoko setup-env typescript1 Oct 04 07:22:13 gnu@om:~/src/moko$ ls openembedded Oct 04 07:22:13 argtable conf COPYING.MIT files _MTN README site Oct 04 07:22:14 classes contrib delete.txt MAINTAINERS packages removal.txt Oct 04 07:23:35 good morning Oct 04 07:24:34 gnu: pastebin setup-env Oct 04 07:25:46 http://pastebin.ca/725270 Oct 04 07:29:20 that looks right Oct 04 07:30:12 XorA: strace shows bitbake is reading openembedded/conf/bitbake.conf, as it should... Oct 04 07:30:55 /home/gnu/src/moko no symlink in that path? Oct 04 07:31:22 XorA: No symlink. Oct 04 07:32:32 build/conf/local.conf? Oct 04 07:33:28 http://pastebin.ca/725274 Oct 04 07:36:08 XorA: Aha! It's loading /home/gnu/src/moko/openembedded/conf/distro/angstrom-2007.1.conf, isn't that wrong (should be 2007.2)? Oct 04 07:36:19 gnu: yeah Oct 04 07:36:25 gnu: old version of Mokomakefile Oct 04 07:36:39 morning Oct 04 07:36:40 Bitbake and generating packages gives problem Oct 04 07:36:56 while Oct 04 07:37:00 building Oct 04 07:37:01 XorA: I just downloaded it today -- and did make update-makefile too... Oct 04 07:37:02 make build-package-atmycommand Oct 04 07:37:05 gives provblevm Oct 04 07:37:07 problem Oct 04 07:37:36 ERROR: No providers of build target atmycommand (for []) Oct 04 07:38:01 XorA: There isn't an openembedded/conf/distro/angstrom-2007.2.conf... Oct 04 07:38:24 gnu: sorry I read that as openmok-2007.1 Oct 04 07:38:36 gnu: its right reading angstrom-2007.1 Oct 04 07:39:56 XorA: Hmm, before running MokoMakefile, I checked out the openmoko repository. Maybe I'll move that aside and see if MokoMakefile does it differently? Oct 04 07:40:13 gnu: possibly Oct 04 07:40:14 hullo CM, what about the ts_harvest.out? got it? Oct 04 07:40:27 gnu: because Im confused Oct 04 07:40:36 XorA: Me too! Oct 04 07:43:27 atul, atmycommand.bb in a place where bitbake isn't looking for it? Oct 04 07:43:56 tuukkah : ya i place .bb there in that application folder Oct 04 07:49:44 blip Oct 04 07:59:04 i have a really strange effect with my .bb file. since some days i couldn't build because it did not fetch from svn anymore. someone helped me yesterday to fix that (setting SRCREV="${AUTOREV}"). after that it worked find but the final ipk is now build in a path containing a additional /morgue at the end... any idea why? Oct 04 08:01:44 ...the mystery of bitbake - nobody knows? hm :D Oct 04 08:02:22 no idea Oct 04 08:02:27 http://www.openembedded.org/wiki/DirectoryStructure Oct 04 08:03:06 i just know, i can't build and the problem is with gcc-cross-initial Oct 04 08:04:14 ynezz: thnx... outdated? why that? hm Oct 04 08:08:50 ynezz: ah, understand, than this make build-package-package-index put it there...? Oct 04 08:09:14 dunno Oct 04 08:10:07 i had a version with rev rHEAD (as string) which it found more recent, so r2136 was 'outdated' when it created the package db Oct 04 08:11:30 could be, that move is done by ipkg-make-index Oct 04 08:12:03 i understand the comment in the page this way... one more x-file closed on bb :D Oct 04 08:37:26 Any_Key: Yes, same place as before Oct 04 08:37:40 Any_Key: Sent you some messages about it, just a sec Oct 04 08:38:00 http://claesmogren.com/~cm/neo/ts_harvest.out Oct 04 08:39:03 CM, perfect, lets see Oct 04 08:54:37 (script) wiki RecentChanges: [[MokoMakefile]] [[What_mean_by_no_provider_for_build_target]] Oct 04 09:11:27 * ScaredyCat wonders why irda_tools gets build Oct 04 09:11:30 built Oct 04 09:11:54 utils even Oct 04 09:13:12 hi scaredycat Oct 04 09:13:30 lo woglinde Oct 04 09:13:38 scaredycat bitbake -optionsfordependencygraph sometask Oct 04 09:13:49 than you can see why irda is in Oct 04 09:13:56 ooo.. nice... ta Oct 04 09:14:03 bitbake -g I think it was Oct 04 09:14:48 * ScaredyCat nods Oct 04 09:16:20 roofs ... Oct 04 09:16:23 footfs Oct 04 09:16:27 gah! Oct 04 09:16:30 rootfs Oct 04 09:16:46 which means, it's not really needed - no/ Oct 04 09:16:47 ? Oct 04 09:16:52 <_diego__> footfs is interesting ^^ Oct 04 09:19:01 scaredycat roots includes irda? Oct 04 09:19:32 yes, seems to be a dep for rootfs Oct 04 09:19:39 hm Oct 04 09:19:42 strange Oct 04 09:20:00 "openmoko-devel-image.do_rootfs" -> "irda-utils.do_populate_staging" Oct 04 09:20:33 woglinde: salut Oct 04 09:21:21 hi zecke Oct 04 09:22:36 Bitbake and generating packages gives problem in OpenMoko environment. Oct 04 09:22:47 To build the package I do: Oct 04 09:22:47 make build-package-atmycommand Oct 04 09:23:00 It gives error to me Oct 04 09:23:02 ERROR: No providers of build target atmycommand (for []) Oct 04 09:23:25 can anyone tell me the solutions Oct 04 09:23:28 to this problem Oct 04 09:25:39 atul yes build a package that is in oe like smaba Oct 04 09:25:42 aeh samba Oct 04 09:25:47 I begin too with this environment, it seems that the new packages have to be added in an environment variable Oct 04 09:26:05 than the error is really clear Oct 04 09:26:11 -> ERROR: No providers of build target atmycommand (for []) Oct 04 09:26:31 good morning Oct 04 09:26:34 zween: another Scottish person? Oct 04 09:26:38 atul: there are more details if i remember well in the crash course for developers in the openmoko wiki Oct 04 09:26:43 hi florian Oct 04 09:27:41 atul: e.g. there: http://wiki.openmoko.org/wiki/Application_Development_Crash_Course#Adding_Your_Application_to_the_Image Oct 04 09:28:04 zilux : thx Oct 04 09:28:11 i go throuth it Oct 04 09:28:18 woglinge : thx Oct 04 09:29:00 XorA: Hi. Not scottish, but I study in Edinburgh. Oct 04 09:29:14 atul ah you wanted to integrate a package Oct 04 09:29:17 XorA: Where are you? Oct 04 09:29:24 woglinge : yes Oct 04 09:29:30 zween: just outside Edin Oct 04 09:31:19 XorA: When you come to Edinburgh some time, we could go for a pint. I would actually be interested what other people do with openmoko around here. Oct 04 09:31:47 zween: SpeedEvil is just north of us as well Oct 04 09:32:01 zween: there is a page on wiki details where people are :-D Oct 04 09:32:15 zween: I might make November LUG Oct 04 09:33:39 XorA: dont follow the lug mailing list that much and never been to a meeting. Oct 04 09:36:58 zween: but I work for OpenMoko so thats what I do with one :-D Oct 04 09:38:22 XorA: right. Oct 04 09:38:39 Did anyone manage to set up a google map over moko users/devs in the same way as nslu2 has? Oct 04 09:39:14 XorA: well, I guess this will also be your topic for the november lug then. Ill make sure I come around to this one then. Oct 04 09:58:50 openmoko: 03thomas * r3076 10/trunk/src/target/OM-2007.2/applications/openmoko-today2/ (ChangeLog src/today-pim-summary.c): Oct 04 09:58:50 openmoko: * src/today-pim-summary.c: Oct 04 09:58:50 openmoko: (today_pim_summary_tasks_selection_changed_cb): Update binary name for Oct 04 09:58:50 openmoko: openmoko-tasks Oct 04 10:03:46 is there a way to rebuild ipkg status db? Oct 04 10:05:02 hello all Oct 04 10:05:28 does anybody know are there USB Host drivers inside of kernel? Oct 04 10:05:40 XorA: November EdLug sounds good. Oct 04 10:11:38 openmoko: 03chris * r3077 10/trunk/src/target/OM-2007.2/applications/openmoko-today2/ (ChangeLog src/today-pim-journal.c): Oct 04 10:11:38 openmoko: * src/today-pim-journal.c: Oct 04 10:11:38 openmoko: Fix bug #901, thanks to Roman Moravcik Oct 04 10:12:47 getting error, "ERROR: No providers of build target", trying to build local package using this, http://free-opensource.qvantel.net/mediawiki/index.php/OpenMoko_Hello_World_Debian_20072#bitbake_and_generating_packages Oct 04 10:54:57 (script) wiki RecentChanges: [[PyFeedRead]] [[OpenMoko2007.2]] [[User:Zween]] [[OpenMoko_Local_Groups:_Edinburgh]] [[User:Daltona]] Oct 04 10:55:43 lrg: you got Edlug to move back to a pub yet? Oct 04 10:56:08 XorA: not yet, I need to resubscribe to the list again. Oct 04 10:56:54 XorA: although there are 4 of us now. Bob, Mark, you and I. We could allways Oct 04 10:57:02 XorA: force a vote ! Oct 04 10:57:13 lrg: though dont know if I can make it for sure Oct 04 10:57:37 XorA: np, I will be in India for Dec lug. Oct 04 10:57:57 XorA: going to fins of blackness Oct 04 10:58:22 kanelbulle! Oct 04 10:59:28 mmm, kanelbullens dag idag =) Oct 04 11:00:13 javla sjuit bolla! Oct 04 11:00:41 :D Oct 04 11:02:48 * CM har med sig hembakade kanelbullar till jobbet idag Oct 04 11:03:13 jeddy3: Börjar jobba på HiQ i Stockholm i Januari förresten :) Oct 04 11:03:24 CM, grattis! :D Oct 04 11:03:25 * CM forgot a /msg ;) Oct 04 11:03:34 openmoko: 03chris * r3078 10/trunk/src/target/OM-2007.2/libraries/libmokoui2/ (ChangeLog libmokoui/moko-finger-scroll.c): Oct 04 11:03:34 openmoko: * libmokoui/moko-finger-scroll.c: Oct 04 11:03:34 openmoko: Really don't allow clicks after scrolling has begun, fix crash when Oct 04 11:03:34 openmoko: destroying widget while scrolling timeout is active Oct 04 11:05:03 openmoko: 03thomas * r3079 10/trunk/src/target/OM-2007.2/libraries/libmokojournal2/ (10 files in 3 dirs): Oct 04 11:05:03 openmoko: * Makefile.am: Oct 04 11:05:03 openmoko: * autogen.sh: Oct 04 11:05:03 openmoko: * configure.ac: Oct 04 11:05:03 openmoko: * doc/reference/Makefile.am: Oct 04 11:05:05 openmoko: * doc/reference/libmokojournal-docs.sgml: Oct 04 11:05:09 openmoko: Add gtk-doc support Oct 04 11:19:38 Ui - got my Neo today. :-) Oct 04 11:22:13 Bernte: Ah, nice :) Oct 04 11:23:20 :) Oct 04 11:25:40 Bernte: and what are your first impressions? Oct 04 11:26:05 "Because first impressions last.." ;) Oct 04 11:26:29 * SpeedEvil wonders what image is on recent neos. Oct 04 11:26:36 Bernte: when did you order? Oct 04 11:27:22 Bernte: http://wiki.openmoko.org/wiki/P1_Owners Oct 04 11:27:39 what files need to be backed up before doing an update? "make update-makefile && make setup update openmoko-devel-image" overwrote my setup-env file Oct 04 11:27:52 Bernte seems to go offline for his neo... ;) Oct 04 11:28:12 Bernte: http://wiki.openmoko.org/wiki/Neo1973_Robustness Oct 04 11:28:21 Maybe watching it boot, slowly... ;) Oct 04 11:29:55 good stuff, SpeedEvil, didn't know that page :D Oct 04 11:33:31 * CM hints at abraxa's mockups: http://dar-clan.de/mediaplayer_compact.jpg and http://dar-clan.de/mediaplayer_expanded.jpg Oct 04 11:34:19 abraxa_: i'm trying out mediaplayer .. doesn't work too well yet, or what? Oct 04 11:34:37 torpor: Told him about your problem. It's fixed in svn Oct 04 11:34:46 i created a new playlist by hand (text file editor) .. Oct 04 11:34:55 CM: oh, okay i'll do an update now and check it out .. Oct 04 11:34:58 torpor: It has to do with that it doesn't check for the folder correctly Oct 04 11:35:09 couldn't make it play anything, though Oct 04 11:35:27 torpor: I think the version in OpenMoko is locked at a certain source rev Oct 04 11:35:30 i made a dummy playlist, it loaded it fine. but i couldn't find out hwo to play a track at all .. Oct 04 11:35:37 how do i get the svn version then? Oct 04 11:35:59 torpor: Did you create /home/root/playlists? Oct 04 11:36:00 I'm not sure. He said he's redoing the gui completely Oct 04 11:36:11 i see Oct 04 11:36:27 abraxa_: Shh.. You're interrupting the information relay station! :P Oct 04 11:36:34 xP~ Oct 04 11:36:38 Hehe Oct 04 11:36:55 SpeedEvil: Sorry - was AFK ... Let me check when I ordered, but it must have been 1 month+ ... Oct 04 11:37:02 ScaredyCat: I'd guess the IRDA stuff is actually for OBEX, which is the same in BT... the rootfs dependency might be bogus (if there was a true dependency, it would go through tasks) Oct 04 11:37:25 abraxa_: yes Oct 04 11:37:33 and i put my dummy playlist (copied rom the wiki) in there Oct 04 11:37:41 but there's no play button, no way to start a track, etc. Oct 04 11:37:58 emdete: More or less ;-) I am at work, so no 100% OpenMoko buisness currently ;-) Oct 04 11:38:05 torpor: What do you see then? Oct 04 11:38:45 Just writing the rootfs ... I assume that I do not need to update the kernel yet. Oct 04 11:40:03 abraxa_: booting up now Oct 04 11:40:24 Bernte: It's much quicker than updating the rootfs. Also, you need to have a matching kernel to the rootfs, or the modules won't load Oct 04 11:40:51 Hmm ... Oct 04 11:41:04 zecke: you around? Oct 04 11:41:23 Is OpenMoko-openmoko-devel-image-glibc-ipk-P1-September-Snapshot-20070917-fic-gta01.rootfs.jffs2 the right rootfs? Which one is the matching kernel? Oct 04 11:41:51 Bernte: That's a fairly old one Oct 04 11:41:52 thos: yes Oct 04 11:42:03 thos: fighting with a bios upgrade though :) Oct 04 11:42:19 zecke: this line has your name next to it... #define MOKO_HINT_ENTRY_GET_CLASS(obj) (G_TYPE_INSTANCE_GET_CLASS ((obj), MOKO_TYPE_HINT_ENTRY, MokoGroupComboClass)) Oct 04 11:42:27 CM: Which ones should I take? Oct 04 11:43:18 abraxa_: i start up mediaplayer and i get the msg " gstreamer error: resource not found" Oct 04 11:43:23 it has autoloaded my playlist Oct 04 11:43:31 Bernte: I'm using one I built yesterday, and it works great Oct 04 11:43:42 torpor: try killall pulseaudio Oct 04 11:43:44 and i see no play button, but a button that goes "<" instead of">" Oct 04 11:44:03 torpor: The button is wrong as of now, yes Oct 04 11:44:24 torpor: Resource not found means that it can't load the track Oct 04 11:44:25 ok Oct 04 11:44:35 zecke: sorry, false alarm :-) Oct 04 11:44:38 but i get the track info - riders on the storm, the doors, etc. Oct 04 11:44:42 so its loading something Oct 04 11:44:56 CM: Well - I think I will use a pre-build one before I start compiling the stuff myself. Oct 04 11:45:07 torpor: If you loaded the tracks before then the metadata is saved within the playlist itself Oct 04 11:45:25 Bernte: If you want you can have the ones I've built. No guarantees though.. Oct 04 11:45:40 abraxa_: ok Oct 04 11:45:43 Bernte: They work for me, and you can't brick your Neo Oct 04 11:46:08 torpor: Are you trying to load tracks from SD? Maybe it's not mounted for some reason? Oct 04 11:46:23 Bernte: http://claesmogren.com/~cm/neo/2007.2/ Oct 04 11:46:29 yeah the SD is mounted Oct 04 11:47:29 torpor: Is that the playlist you made? I suggest you make a new one and add the tracks Oct 04 11:47:33 so i should do 'killall pulseaudio' then abraxa_? Oct 04 11:47:59 torpor: That won't help you with loading the files - that's for "resource busy" problems :) Oct 04 11:48:16 NOTE: Task failed: Error: /home/j/hak/openmoko/openembedded/packages/ipkg/ipkg-0.99.163/./lonk-link-name.patch not found. Oct 04 11:48:42 Bernte: I guess you're following http://wiki.openmoko.org/wiki/Flashing_openmoko ? And don't touch your uboot unless you have a debugboard and know what you're doing. :) Oct 04 11:49:10 oops Oct 04 11:49:11 And don't type "nand erase" (the correct is "nand erase rootfs") Oct 04 11:49:12 wtf? Oct 04 11:49:12 root@fic-gta01:~$ mount /media/card Oct 04 11:49:12 mount: special device /dev/mmcblk0p1 does not exist Oct 04 11:49:15 CM: I thought there's a known good version of uboot in NOR flash that you can't overwrite? Oct 04 11:49:32 CM: I will not touch the u-boot, and I _do_ have a debug board ;-) Oct 04 11:49:35 abraxa_: Not in GTA01, but GTA02 is supposed to have that Oct 04 11:49:40 http://wiki.openmoko.org/wiki/User:CesarB/Things_not_to_do_with_your_Neo has more... Oct 04 11:49:41 CM: Ohh okay Oct 04 11:49:48 CM: But thanks for the warning. Oct 04 11:49:58 Bernte: Aww.. I'm jealous! My partition table is borked and i don't have a debugboard Oct 04 11:50:01 torpor: I guess there is your problem ;x Oct 04 11:51:03 rebooting now abraxa_ Oct 04 11:51:13 when i tried yesterday the media card was there (thats where i put the test songs) Oct 04 11:51:22 anyway is there some chance i can update to the latest version? Oct 04 11:51:33 moin Oct 04 11:51:36 CM: Sorry to hear that, and Sweden is just too far away from Switzerland. Oct 04 11:52:03 CM: since you can get into linux, you could in theory find out the environment partition offset (it's hidden somewhere in the OOB data of the first pages of the flash), create a valid environment (with a valid partition table), and flash it directly in the correct offset on the device... (that can, however, overwrite your kernel, so you'd have only one chance :( ) Oct 04 11:52:42 * cesarb really thinks uboot should have the USB device as the default instead of serial... would make it way faster to fix these things Oct 04 11:52:50 cesarb: It would be worth a shot, but I'm not good enough to know how to do that. I'll have to mail it somewhere though Oct 04 11:52:50 * cesarb leaves to work Oct 04 11:53:03 cesarb: I know, see bug 765 :) Oct 04 11:53:15 torpor: I don't think so as OE doesn't use the latest version from SVN from what I know - I can send you a recent build if you want to, though Oct 04 11:53:36 CM: well, I bought mine with a debug board, since I think I'm the only one in Brazil with a neo so far Oct 04 11:53:39 torpor: did you apply http://people.openmoko.org/henryk/ipkg-long-link-name.patch? or where did you get that recipe from? Oct 04 11:53:49 abraxa_: i just wanna help with the mediaplayer however i can.. Oct 04 11:53:51 CM: (and I'm reading the bug list, so I've seen that bug report) Oct 04 11:53:57 * cesarb really leaves to work Oct 04 11:53:59 Henryk: i just did make update Oct 04 11:54:07 cesarb: I should have done that too, but I wasn't planning to play with uboot Oct 04 11:54:19 well.. i have a debugboard here in berlin.. which is somehow in the middle of sweden<->switzerland Oct 04 11:54:31 Henryk: so its a typo in the public version .. Oct 04 11:54:31 torpor: Thanks, I appreciate it :) Oct 04 11:54:47 So, have to get back to writing my PhD thesis. Will continue playing later ;-) Oct 04 11:54:53 roh: I guess just mailing it there and back would work fairly good. Oct 04 11:54:59 torpor: well yes, there's a typo, but at least I copy and pasted the typo to both occurences Oct 04 11:55:06 heh heh. Oct 04 11:55:12 It still boots though, so I'm not desperate to get it fixed. Oct 04 11:55:25 I'd try to fix it "manually" first though :) Oct 04 11:55:40 sorry to interrupt Oct 04 11:55:40 hi 2 all Oct 04 11:55:51 CM so your uboot_env is snafu-ed? Oct 04 11:56:02 can someone help me building uicmoc4-native Oct 04 11:56:17 roh: Yes. For a while I could get it back, but it haven't worked lately Oct 04 11:56:42 i have tried a lot but it fails with http://pastebin.com/m2c873815 everytime Oct 04 11:57:01 nobody96: see http://bugzilla.openmoko.org/cgi-bin/bugzilla/show_bug.cgi?id=747 Oct 04 11:57:07 can you see this Oct 04 11:57:57 awesomo: this? Oct 04 11:58:19 yay. i have a patch for mac os x "dfu-util" Oct 04 11:58:36 what's the right way to make sure it is done right Oct 04 11:58:42 that was fast :) Oct 04 11:58:43 thanks Oct 04 11:59:17 awesomo: I'd suggest posting it to bugzilla, and maybe to the relevant mailinglist :) Oct 04 11:59:31 KTHX Oct 04 11:59:53 abraxa_: okay doing a new test from fresh boot now Oct 04 11:59:58 i still get the gstreamer error .. Oct 04 12:00:20 Is the SD mounted? Oct 04 12:00:21 * CM hopes that official OM people corrects him if he's suggesting wrong things. Oct 04 12:00:31 abraxa_: yes Oct 04 12:00:37 this time it is .. just had to re-seat it. Oct 04 12:00:47 why this patch isnt in the repository? Oct 04 12:00:52 torpor: Can you pastebin the .xspf? Oct 04 12:01:00 CM well.. mailinglist and bugzilla are the right places for patches ;) Oct 04 12:01:57 abraxa_: just a minute, i'm on mac osx, have to re-do network config every time i reboot the neo Oct 04 12:02:07 nibbo_ stinks like gcc3 vs gcc4 .. which gcc are you using? Oct 04 12:02:15 torpor: No need to stress out ;) Oct 04 12:02:29 ok Oct 04 12:03:37 http://rafb.net/p/MqvDaY96.html Oct 04 12:03:40 ^^ abraxa_ Oct 04 12:04:04 http://rafb.net/p/lugWw327.html Oct 04 12:04:27 gstreamer: internal data flow error Oct 04 12:04:54 after you add tasks to projects.openmoko.org, does it take time to view the tasks assigned to you? Oct 04 12:08:13 has anyone encountered build failure on webkit-gtk? Oct 04 12:08:50 torpor: Playlist is valid and if the files really *are* mp3s then there should be no problem... but let me send you the latest SVN build Oct 04 12:09:10 ok Oct 04 12:09:24 also whats to be done about pulseaudio now? how do i restart it properly so it doesn't interfere with our testing/ Oct 04 12:10:03 torpor: It won't with the build I'm sending :) Oct 04 12:10:09 ok Oct 04 12:11:44 torpor: Please save http://abraxa.dyndns.org:81/random/openmoko-mediaplayer to /usr/bin on the Neo and try again Oct 04 12:12:22 is anyone running gcc on the neo yet Oct 04 12:12:28 abraxa_: done. Oct 04 12:12:30 trying now Oct 04 12:12:56 hardware question: on pcb board ar there any rs232 (lv) pins? Oct 04 12:13:16 awesomo: "ipkg install task-sdk-native" should give you all needed to build helloworld.c Oct 04 12:13:35 excellent. Oct 04 12:13:49 Cannot find package task-sdk-native. Oct 04 12:14:02 just flashing to the august snapshot Oct 04 12:14:45 mbuf: Yes, I did before. Was an issue with libicui18n or something like that Oct 04 12:15:14 CM, do I need to wait for any fixes? Oct 04 12:15:18 mbuf: I rebuilt everything from scratch, but just rebuilding glibc and icu might be enough Oct 04 12:15:44 I sync:ed and built a complete image this morning Oct 04 12:15:51 hm, if a file was downloaded and corrupted on the way down, and I remove it manually, Mokomakefile doesn't appear to redownload the missing file, and just blows up with a file not found error. Oct 04 12:16:04 CM, "make update-makefile && make setup && make update" ; "make clean-package-glibc clean-package-icu" ; "make openmoko-devel-image"? Oct 04 12:16:30 mbuf: I can't promise it works, because I haven't tried it Oct 04 12:16:37 mbuf: But it looks good to me. :D Oct 04 12:17:33 mbuf: When I had that problem with webkit-gtk I couldn't build icu either. Someone suggested rebuilding glibc would fix that, but by then I had aready restarted a complete build Oct 04 12:17:42 abraxa_: this version doesn't work Oct 04 12:17:55 some dialog box comes up with a large error message but it disappears too fast for me to see Oct 04 12:18:00 i'll try launching it from the cmd line .. Oct 04 12:18:07 CM, i see, in the above, i should also do "make clean-package-webkit-gtk", i guess Oct 04 12:18:36 root@fic-gta01:~$ DISPLAY=0:0 openmoko-mediaplayer Oct 04 12:18:36 Received SIGSEGV Oct 04 12:18:36 This might be a bug in the OpenMoko Media Player. Oct 04 12:18:45 uhh.. Oct 04 12:18:54 something about gstreamer Oct 04 12:19:05 Ahh Oct 04 12:19:42 Do you have libgstpulse.so in /usr/lib/gstreamer-0.10/ ? Oct 04 12:20:44 nope Oct 04 12:21:30 gst-plugin-pulse Oct 04 12:21:42 torpor: Copy http://abraxa.dyndns.org:81/random/libgstpulse.so there, please (unless you want to compile it yourself) Oct 04 12:21:43 mbuf: Let me know later if that worked :) Oct 04 12:22:05 abraxa_: how about i just do ipkg install gst-plugin-pulse Oct 04 12:22:17 torpor: Works, too ;) Oct 04 12:22:36 * CM hands LaF0rge a kanelbulle. (The cinnamon roll day in Swden today) Oct 04 12:22:38 I just never use ipkg, so... Oct 04 12:22:52 okay it starts, but i still get gstreamer error: internal data flow error on startup Oct 04 12:22:56 but at least the app is starting up now Oct 04 12:22:59 i see gui, etc. Oct 04 12:23:19 CM, doing "make openmoko-devel-image", it is rebuilding glibc now Oct 04 12:23:45 abraxa_: i hit play ("<" button) and it stays turned on 'red', but no sound. Oct 04 12:23:50 torpor: Do you have libgstmad.so in /usr/lib/gstreamer-0.10/ ? Actually, can you just show me the contents of that dir? Oct 04 12:24:17 http://rafb.net/p/RXOuPf57.html Oct 04 12:25:40 torpor: Hm. Oct 04 12:26:45 torpor: Please try a "gst-launch filesrc location=/media/card/riders.mp3 ! playbin ! pulsesink" on the command line Oct 04 12:27:05 WARNING: erroneous pipeline: could not link filesrc0 to playbin0 Oct 04 12:28:08 no sound Oct 04 12:28:32 i got sound!! Oct 04 12:28:34 yay Oct 04 12:30:03 dang this is slow Oct 04 12:30:09 but at least it makes noise. Oct 04 12:30:56 torpor: "gst-launch filesrc location=/media/card/riders.mp3 ! mad ! audioconvert ! audioresample ! pulsesink" Oct 04 12:31:08 awesomo: Using gst-launch or how? Oct 04 12:31:34 abraxa_ : i just reflashed to the august snapshot and sound started working Oct 04 12:31:38 Ahh Oct 04 12:31:47 abraxa_: okay now something is playing but its verys hitty Oct 04 12:31:58 i'm not on the bleeding edge yet..it makes a noise whenever i touch the screen Oct 04 12:32:12 its riders alright, but its stuttering and noisy and doesn't sound like the sample right is correct. Oct 04 12:32:38 oops Oct 04 12:32:40 WARNING: erroneous pipeline: no element "pulsesink" Oct 04 12:32:51 torpor: That's somewhat expected as the CPU is too slow for audioconvert/audioresample Oct 04 12:33:05 jeddy3: ipkg install gst-plugin-pulse Oct 04 12:33:21 abraxa_, ah got it Oct 04 12:33:32 ok Oct 04 12:33:40 i'm gonna restart now abraxa_ and have another go.. Oct 04 12:33:44 oki Oct 04 12:34:06 so what does that test tell us then abraxa_ ? Oct 04 12:34:21 oh wow Oct 04 12:34:24 torpor: gstreamer itself works :) Oct 04 12:34:37 i can scroll by flicking ... like the iphone .. cool Oct 04 12:35:03 ok Oct 04 12:35:06 rebooted now Oct 04 12:35:09 what should i do abraxa_ ? Oct 04 12:35:38 abraxa_, yes that does it, although get the "could not link" issue with playbin, and choppy playback with audioconvert Oct 04 12:38:37 torpor: Please replace your libgstmad.so in /usr/lib/gstreamer-0.10 with this one: http://abraxa.dyndns.org:81/random/libgstmad.so Oct 04 12:39:00 (it has a patch that will eventually land in OE, too) Oct 04 12:44:09 who read/send successfully sms on moko (without console :P)? Oct 04 12:46:33 abraxa_, openmoko-mediaplayer says gstreamer: resource busy or not avaliable Oct 04 12:46:53 abraxa_: doing it Oct 04 12:48:52 jeddy3: "killall pulseaudio" should remedy that Oct 04 12:49:19 abraxa_: i gotta go now but i will try this testing again in a few hours .. Oct 04 12:49:19 abraxa_, sure does :D Oct 04 12:49:32 torpor: Okay, thanks :) Oct 04 12:54:57 (script) wiki RecentChanges: [[User:SNMoore]] Oct 04 12:56:59 abraxa_: it works! with http://abraxa.dyndns.org:81/random/libgstmad.so Oct 04 12:57:17 still lags every now and then but i guess thats a better improvement for the future Oct 04 12:57:51 torpor: Good :) Oct 04 12:58:04 cant go to the neext track tho Oct 04 12:58:28 What happens when you click on the button? Oct 04 12:58:48 nothing Oct 04 12:58:51 lags too Oct 04 12:59:01 But it says 01/03 tracks? Oct 04 12:59:09 yup Oct 04 12:59:15 Hmm. Oct 04 12:59:57 i have multiple copies of the same track in the playlist, maybe that has something to do with it .. Oct 04 13:00:00 but this is great. Oct 04 13:00:05 i'll play with it a bit later and try to contribute some more .. Oct 04 13:00:10 gotta go now. thanks abraxa_ !!! Oct 04 13:00:13 okay Oct 04 13:00:16 Thanks for helping ;3 Oct 04 13:02:27 exit Oct 04 13:06:05 openmoko: 03zecke * r3080 10/trunk/src/target/OM-2007.2/applications/openmoko-feedreader2/ (5 files in 2 dirs): (log message trimmed) Oct 04 13:06:05 openmoko: 2007-10-04 Holger 'Siddhartha' Freyther Oct 04 13:06:05 openmoko: s/WebKitGtk/WebKit/g Oct 04 13:06:05 openmoko: s/webkit_gtk_/webkit_/g Oct 04 13:06:05 openmoko: s/WEBKIT_GTK_/WEBKIT_/g Oct 04 13:06:06 openmoko: * src/Makefile.am: Oct 04 13:06:08 openmoko: * src/callbacks.c: Oct 04 13:08:55 openmoko: 03abraxa * r3081 10/trunk/src/target/OM-2007.2/applications/openmoko-mediaplayer2/src/ (guitools.c guitools.h main.c playback.c): Make sure the error dialog is readable if gst-plugin-pulse is missing and exit gracefully afterwards Oct 04 13:12:43 openmoko: 03zecke * r3082 10/trunk/src/target/OM-2007.2/applications/openmoko-browser2/ (7 files in 2 dirs): (log message trimmed) Oct 04 13:12:43 openmoko: 2007-10-04 Holger 'Siddhartha' Freyther Oct 04 13:12:43 openmoko: s/WebKitGtk/WebKit/g Oct 04 13:12:43 openmoko: s/webkit_gtk_/webkit_/g Oct 04 13:12:43 openmoko: s/WEBKIT_GTK_/WEBKIT_/g Oct 04 13:12:45 openmoko: * src/browser-data.c: Oct 04 13:12:47 openmoko: (browser_page_new): Oct 04 13:21:34 * cb22_ is addicted to news about openmoko :/ Oct 04 13:23:04 * zash` too Oct 04 13:23:10 ADDICTED!! Oct 04 13:23:48 like crack :/ Oct 04 13:25:40 hello Oct 04 13:25:47 Hellos :) Oct 04 13:26:27 do you know if there is any "official" member of the project here ? Oct 04 13:26:27 ar0x`: See http://wiki.openmoko.org/wiki/FAQ (and also SH1, SH1_FAQ, Neo1973, Developer_preview and Neo1973_Hardware#GTA01Bv4 pages) (Hopefully these links answers your question. This autotrigger for new nick and question during first 15 minutes done only once.) Oct 04 13:26:35 There are Oct 04 13:26:58 XorA|gone and roh, for example (and I guess myself, though I'm not a hired member, heh) Oct 04 13:27:11 ok Oct 04 13:27:26 Got any questions? Oct 04 13:27:27 are they also member of the core-team ? Oct 04 13:27:41 abraxa_: ie "you do the job without being paid"? ;P Oct 04 13:27:44 Nope Oct 04 13:28:13 viq: Pretty much, unless google's Summer of Code counts ;) Oct 04 13:28:20 hehe Oct 04 13:30:05 ar0x`: What do you want to know? Oct 04 13:31:27 abraxa_: see my pm Oct 04 13:32:12 ar0x`: You don't seem to have your nick registered, so you can't send PMs just yet Oct 04 13:32:16 CM, "make update-makefile && make setup && make update" ; "make clean-package-glibc clean-package-icu clean-package-webkit-gtk" ; "make openmoko-devel-image" - resulted in package-icu failed build error Oct 04 13:32:22 ar0x`: mickey|portugal is back next week monday, he's here a lot. LaF0rge is also here some times, but usually quiet Oct 04 13:32:22 oups, yep, I just noticed this Oct 04 13:32:32 mbuf: Sucks.. Oct 04 13:33:09 mbuf: That icu error is the same reason as the webkit error. Must mean that there's something more than glibc that has to be rebuilt Oct 04 13:33:21 CM, i see Oct 04 13:35:57 (script) openmoko-neo1973: Harald Welte Re: QVGA for Neo Oct 04 13:36:12 mbuf: What if you try building icu-native first? Oct 04 13:36:21 CM, build error log, http://pastebin.ca/725484 Oct 04 13:36:39 CM, how would i do that? Oct 04 13:37:04 CM: thanks Oct 04 13:37:21 make rebuild-package-icu-native Oct 04 13:38:35 mbuf: That's the exact same error I used to get. Building from scratch gets around it, but it would be nicer to only rebuild the package you need to... Oct 04 13:39:06 CM, i don't have the luxury/download speed to build from scratch :( Oct 04 13:39:25 Oh, ok. Hmm.. Oct 04 13:39:28 CM, build from scratch, you mean, create new moko directory, and do the whole step again? but that will be inefficient Oct 04 13:39:40 *whole build steps Oct 04 13:39:52 mbuf: Yes, very inefficient. Takes 10h on my comp Oct 04 13:40:14 CM, make rebuild-package-icu-native is complete; what next? Oct 04 13:40:38 CM, make openmoko-devel-image? Oct 04 13:40:42 Ok.. That worked. Then I wonder why make rebuild-package-icu doesn't Oct 04 13:40:58 mbuf: I don't know the requirements for icu Oct 04 13:41:02 That's the problem Oct 04 13:41:35 CM, ok Oct 04 13:42:56 CM, should i post to openmoko-devel? Oct 04 13:43:29 I guess so. I'm sure someone there knows where to look Oct 04 13:43:38 Just list all steps you've gone through already Oct 04 13:45:01 * ScaredyCat gnashes Oct 04 13:45:51 openmoko: 03thomas * r3083 10/trunk/src/target/OM-2007.2/libraries/libmokojournal2/ (3 files in 2 dirs): Oct 04 13:45:51 openmoko: * mokojournal/moko-journal.c: Oct 04 13:45:51 openmoko: * mokojournal/moko-journal.h: Oct 04 13:45:51 openmoko: Move documentation in C file Oct 04 13:51:36 mbuf: Maybe something in tzset, tzname, or timezone has changed? I looked some at what the icu configure script checks for and those were at least there. Oct 04 13:52:13 CM, my date is wrong! Oct 04 13:52:30 CM, its not supposed to be Friday, yet Oct 04 13:52:45 mbuf: Hehe.. It's not, but that's not the problem Oct 04 13:52:54 Is the UI meant to be so slow? The reaction time looks bad to me. Oct 04 13:53:15 Bernte: Yes, it's a bit sluggish. There's a lot of debug-output written Oct 04 13:53:15 Bernte: hopefully its all the debug code in there... Oct 04 13:53:25 Bernte: Yeah, OpenMoko folks want to make their customers wait. Duh. ;p Oct 04 13:53:32 abraxa_: Hehe Oct 04 13:54:04 mbuf: make rebuild-package-timezones might be something? Oct 04 13:54:16 abraxa_: after they waited for the neo so long... ;) Oct 04 13:54:40 CM: Are the Trolltech people doing a better job? Is it faster? Oct 04 13:54:54 Bernte: It's a lot faster Oct 04 13:55:13 CM: Thought so (as a KDE user ;-) ) Oct 04 13:55:18 Bernte: ljp is working on a new release right now Oct 04 13:55:30 Bernte: I'm a Gnome user.. :P Oct 04 13:56:03 It's not about GTK vs. QT really. Qtopia has had a lot more time to mature Oct 04 13:56:49 CM: Well - it can be slow as hell ... the only thing I would like to see sooner than later is the GSM phone. All my collegues already smirk that I bought a phone that cannot phone (yea, everyboyd knowns it is under development, but it is still strange). Oct 04 13:57:35 Bernte: I could call with the latest OpenMoko builds and with the latest Qtopia builds Oct 04 13:57:49 CM, clock skew detected, your build may be incomplete Oct 04 13:57:57 Both have trouble with power management though, and some audio issues Oct 04 13:58:01 Bernte: http://lists.openmoko.org/pipermail/community/2007-September/009997.html might be an interesting read for you Oct 04 13:58:11 mbuf: :( Maybe you have to rebuild everything? Oct 04 13:58:26 If you do, don't remove the sources folder Oct 04 13:58:30 CM, how would i do that? Oct 04 13:59:12 mbuf: make clobber cleans out everything Oct 04 13:59:21 CM, ok Oct 04 13:59:36 CM, thanks for your time and help Oct 04 13:59:56 abraxa_: Yes, I was aware ... But I though that "calling" was just ringing, w.o. audio ... was not aware that it was possible to _speak_ while connected. Oct 04 14:00:31 * abraxa_ looks confused Oct 04 14:00:37 Sorry, you lost me there :) Oct 04 14:33:06 hi Oct 04 14:34:49 Hi Oct 04 14:35:13 hi all Oct 04 14:35:40 Anyone know how gta2 is coming along? Oct 04 14:36:35 * abraxa_ points at the topic Oct 04 14:36:52 where can i read about theming/skinning the neo? Oct 04 14:37:00 enjahova: you can ask me Oct 04 14:37:22 well, if i or a friend of mine were to decide to make a new skin Oct 04 14:37:27 what would be the first step? Oct 04 14:37:39 abraxa_, Ok *douh* Oct 04 14:38:02 enjahova: make a 3D model of the existing case. Oct 04 14:38:17 whoops, i mean software theme then Oct 04 14:38:24 gtk-stuffs! Oct 04 14:38:25 enjahova: Then talk to your local patternmaker, or machine it out of a billet. Oct 04 14:38:29 Oh :) Oct 04 14:38:33 :) Oct 04 14:38:38 though that does sound interesting... Oct 04 14:38:48 I want a fake-damascus steel case. Oct 04 14:38:49 enjahova: well your first step would be to design it Oct 04 14:39:10 thos: logical ^^ Oct 04 14:39:23 enjahova: design the look for widgets in each state.. buttons, checkboxes, pressed, unpressed Oct 04 14:39:26 enjahova: etc Oct 04 14:39:43 hm, überheavy steel casing would be nice Oct 04 14:40:00 thos: Is there anything I need to do to use your GTK+ engine for the media player btw? Oct 04 14:40:01 http://www.crandart.co.uk/damasteel.htm Oct 04 14:40:22 Of course - it's stupidly expensive. Oct 04 14:40:28 abraxa_: do you have the engine installed? Oct 04 14:40:32 with lead for extra weight Oct 04 14:40:53 zash`: And Not that good reception for gps and gsm ;) Oct 04 14:41:13 CM: gah, who needs that in a phone? ;) Oct 04 14:41:14 thos: Not at the moment - I'm just wondering cause it seems to eventually be standard to use so I would like to take advantage of it as well :) Oct 04 14:41:30 abraxa_: are you writing media player? Oct 04 14:41:34 100 pounds or so for the billet to machine the case out of. Oct 04 14:41:36 thos: Yeah Oct 04 14:41:44 abraxa_: you shouldn't need to do anything Oct 04 14:42:01 abraxa_: media player currently crashes on startup for me though, so I can't tell Oct 04 14:43:10 thos: Oh okay then. And yeah, that's because OE apparently SRCREV's a version that isn't as bug-free as HEAD is Oct 04 14:43:34 abraxa_: you should get it using autorev then Oct 04 14:43:35 Or, to make it simple: it expects /home/root/playlists to exist and segfaults if it doesn't Oct 04 14:43:52 abraxa_: tell XorA to update it :-) Oct 04 14:43:59 Okay, will do :) Oct 04 14:45:20 thos: I like abraxa_'s mockups: http://dar-clan.de/mediaplayer_compact.jpg Oct 04 14:45:34 CM: That's random hehe Oct 04 14:45:57 Random? Oct 04 14:46:27 Yeah cause I didn't expect that Oct 04 14:52:07 Cell-Coverage coming to mt. everest - maybe China Mobile will buy AT&T and we will finally get coverage indoors in the U.S. Oct 04 14:53:38 China Mobile buying AT&T? Hmm, I would think the government wouldn't like that ;) Oct 04 14:54:15 no - Chinese Government won't care - even if its a bad deal Oct 04 14:54:30 I think he means the US government Oct 04 14:54:37 (script) wiki RecentChanges: [[User:ChrisLord]] [[Talk:Main_Page]] [[GUI_Style_Guidelines]] [[No_Linux]] [[P1_Owners]] Oct 04 14:54:37 ;) Oct 04 14:55:11 I forgot - China is Free and US is not Oct 04 14:55:55 lol, that's a good one ;) Oct 04 14:58:06 si a Oct 04 14:58:14 who owns the aloril scripts? Oct 04 14:58:22 the wiki script is really broken Oct 04 14:58:31 oh? Oct 04 15:00:10 aloril: well I don't think Chris has edited the wiki today for starters Oct 04 15:00:16 aloril: I just made an edit however Oct 04 15:00:20 hmm.. its not in order, but otherwise its changes since it blurbed previous time Oct 04 15:00:46 hmm.. yeah, maybe should not include minor edits (see http://wiki.openmoko.org/wiki/Special:Recentchanges ) Oct 04 15:01:25 I think minor edits are ok to show :) Oct 04 15:01:25 aloril: ah sorry, I thought it was giving the author name Oct 04 15:01:35 chris has been editing his own page Oct 04 15:02:11 it does list changes only once every 2 hours Oct 04 15:02:46 thos: so just wait for > 1h for your edit to be announced here too ;-) Oct 04 15:03:06 * aloril thinks announcing every edit immediate would be too annoying Oct 04 15:03:37 it could be improved various ways though (like highlighting edits by openmoko company people for exapmle) Oct 04 15:04:24 aloril: does it announce minor edits? Oct 04 15:04:37 aloril: I just assumed it announced the author of the edit Oct 04 15:04:37 yes Oct 04 15:04:40 Vegar: yes Oct 04 15:05:02 ok Oct 04 15:05:07 oh, you mentioned it above Oct 04 15:05:13 I should start reading the backlog Oct 04 15:06:02 thos: Do you happen to know of a way to make openmoko-today process X requests synchronously? --synchronous doesn't seem to work, unfortunately Oct 04 15:06:49 abraxa_: which one of those was supposed to say asynchronous? ;-) Oct 04 15:07:09 abraxa_: first of all, what's the problem Oct 04 15:08:20 thos: I'm trying to debug it because I want to resolve bug #863 - making X process the calls from openmoko-today synchronously would make it a tad easier Oct 04 15:08:33 ...or wait... do I have to start X with --synchronous then? Oct 04 15:09:46 abraxa_: I think you want to start openmoko-today with --sync ? Oct 04 15:10:57 abraxa_: chris is just about to comment on the bug Oct 04 15:11:03 abraxa_: I think he is thinking along the same lines as you Oct 04 15:11:09 thos: Oh, --sync it is... okay Oct 04 15:11:37 I'm curious what he has to say since I have an idea as well :) Oct 04 15:27:11 hey guys Oct 04 15:27:31 quick qestion to the network-people here Oct 04 15:28:01 Network people? Oct 04 15:28:32 abraxa_: probably best to wait for the question :) Oct 04 15:28:35 when i type "ping www.honda.com" on my neo1973 Oct 04 15:28:49 i get Host name lookup failure Oct 04 15:29:09 RalphEichelberge, you have a resolv.conf? Oct 04 15:29:10 I have /etc/resolf.conf set Oct 04 15:29:12 :D Oct 04 15:29:20 RalphEichelberge: have you checked the usual regarding dns entires? I'm guessing it will be in /etc/resolv.conf Oct 04 15:29:23 root@fic-gta01:/var/lib/ipkg$ cat /etc/resolf.conf Oct 04 15:29:23 nameserver 82.150.193.2 Oct 04 15:29:23 nameserver 82.150.192.2 Oct 04 15:29:30 Can you ping those two servers by ip? Oct 04 15:29:30 Can you ping those? Oct 04 15:29:31 root@fic-gta01:/var/lib/ipkg$ nslookup www.honda.com 82.150.193.2 Oct 04 15:29:31 Server: cdns2.hotze.com Oct 04 15:29:31 Address: 82.150.193.2 Oct 04 15:29:31 Name: www.honda.com Oct 04 15:29:31 Address: 164.109.25.248 Oct 04 15:29:42 i can even look up :) Oct 04 15:29:42 and if you ping using ip numbers? Oct 04 15:29:46 aah :) Oct 04 15:29:51 it works too Oct 04 15:30:01 but it doesnt lookup Oct 04 15:30:07 RalphEichelberge: if you type, nslookup Oct 04 15:30:14 and then type www.google.com Oct 04 15:30:16 or what ever. Oct 04 15:30:19 Does it resolve? Oct 04 15:30:25 root@fic-gta01:/var/lib/ipkg$ ping 164.109.25.248 Oct 04 15:30:25 PING 164.109.25.248 (164.109.25.248): 56 data bytes Oct 04 15:30:25 64 bytes from 164.109.25.248: icmp_seq=0 ttl=236 time=124.4 ms Oct 04 15:30:25 64 bytes from 164.109.25.248: icmp_seq=1 ttl=236 time=123.3 ms Oct 04 15:30:44 root@fic-gta01:/var/lib/ipkg$ ping www.honda.com Oct 04 15:30:44 ping: www.honda.com: Host name lookup failure Oct 04 15:30:44 root@fic-gta01:/var/lib/ipkg$ Oct 04 15:30:55 strange hu? Oct 04 15:30:59 RalphEichelberge: can you check via nslookup? Oct 04 15:31:06 yes, actually strange... Oct 04 15:31:10 see above Oct 04 15:31:14 yes Oct 04 15:31:22 ahh, gotcha Oct 04 15:31:25 missed that Oct 04 15:31:38 I'm guessing you'll have an nsswitch.conf file Oct 04 15:31:40 anything is missing Oct 04 15:31:57 what does it have on the hosts line? Oct 04 15:32:02 yes Oct 04 15:32:03 should be Oct 04 15:32:08 hosts: files dns Oct 04 15:32:18 yes Oct 04 15:32:43 do i need any daemon running? Oct 04 15:32:51 this almost certainly won't work, but what happens when you type wget www.honda.com Oct 04 15:33:07 and is it just honda that is causing the problem or all domains? Oct 04 15:33:22 same Oct 04 15:33:39 no - just an example Oct 04 15:34:11 everything works fine with ip's but ipkg-data is with donainnames Oct 04 15:34:18 so i have to set up dns Oct 04 15:35:01 and if you type host www.honda.com Oct 04 15:35:02 ? Oct 04 15:35:34 host? Oct 04 15:35:39 yes Oct 04 15:35:46 hmmm, may be that is not in the neo distro Oct 04 15:35:59 host not found Oct 04 15:36:14 Ralph: If all fails you can just hardcode the hosts in /etc/hosts Oct 04 15:36:17 RalphEichelberge: I was thinking that would be the case. Oct 04 15:36:42 RalphEichelberge: do you have nscd running? Oct 04 15:36:51 lets see Oct 04 15:37:14 no Oct 04 15:38:04 * wibbit shrugs Oct 04 15:38:25 but nscd dont run on my host eather Oct 04 15:38:41 maybe its called something else? Oct 04 15:39:16 nscd isn't that common to be running on single host environments. Mainly used when using nis/ldap etc Oct 04 15:40:31 RalphEichelberge: the only thing I can think of right now, would be to run an strace, to see if any thing stood out from a ping Oct 04 15:40:50 oh ! did stuff on this page: http://wiki.openmoko.org/wiki/USB_Networking Oct 04 15:41:05 argh, it isn't possible to use the GPS with 2007.2, is it? Oct 04 15:41:13 mayby i should stop someting here? Oct 04 15:41:22 orospakr: It is Oct 04 15:41:29 RalphEichelberge: did you try nslookup without specifying the nameserver? Oct 04 15:41:31 abraxa_: ooh, how? Oct 04 15:42:10 btw, I just discovered that I have to restart the bind9 process running on my laptop when usb0 comes up, otherwise it won't answer the neo's resolve requests. Oct 04 15:42:11 orospakr: http://www.openmoko.org.pl/en/node/55 Oct 04 15:42:16 abraxa_: thanks! Oct 04 15:43:04 thos no Oct 04 15:43:07 ill try Oct 04 15:43:20 not working Oct 04 15:43:38 RalphEichelberge: that suggests that there is a problem with /etc/resolv.conf Oct 04 15:43:51 RalphEichelberge: I would delete the file, and re-write it Oct 04 15:44:22 yes - sorry guys - i found it! Oct 04 15:44:33 i misspelled it Oct 04 15:44:41 resolf Oct 04 15:44:45 :( Oct 04 15:45:06 now it works -thx for beeing with me Oct 04 15:45:09 :) Oct 04 15:45:29 no problem Oct 04 15:45:48 i am trying to install J2ME on neo now Oct 04 15:45:58 someone has experince? Oct 04 15:52:46 hm. flash-neo-local appears to flash both the rootfs and the kernel. nice. Oct 04 15:55:08 thos: Till what time are you guys usually at work? Oct 04 15:55:46 hello Oct 04 15:55:52 Hellos! Oct 04 15:56:54 abraxa_: between five and six Oct 04 16:00:15 * RalphEichelberge macht Feierabendl Oct 04 16:01:57 thos: Ohh, good to know - better chance to discuss stuff with you guys :) Oct 04 16:03:21 anyone knows if it's possible to remove menu in midoribrowser? Oct 04 16:05:02 openmoko: 03thomas * r3084 10/trunk/src/target/OM-2007.2/applications/openmoko-dialer2/ (ChangeLog src/moko-contacts.c src/moko-keypad.c): Oct 04 16:05:02 openmoko: * src/moko-contacts.c: (moko_contacts_init): Oct 04 16:05:02 openmoko: * src/moko-keypad.c: (moko_keypad_set_display_text): Oct 04 16:05:02 openmoko: Fix compiler warnings Oct 04 16:08:07 jeddy3 i guess not Oct 04 16:08:45 kristian-m, no seems not :| Oct 04 16:10:07 jeddy3 it'd be best if it could remove all menues on "toggle fullscreen" Oct 04 16:10:25 yeah that'd be great Oct 04 16:11:59 openmoko: 03thomas * r3085 10/trunk/src/target/OM-2007.2/applications/openmoko-dialer2/ (ChangeLog src/moko-contacts.c): Oct 04 16:11:59 openmoko: Patch by: Roman Moravcik Oct 04 16:11:59 openmoko: * src/moko-contacts.c: (moko_contacts_fuzzy_lookup): Fix Oct 04 16:11:59 openmoko: for a bug 802, (add_number): Fix accessing of digits array. Oct 04 16:12:00 This might drive some more traffic to OpenMoko again: http://tirania.org/blog/archive/2007/Oct-04.html Oct 04 16:12:29 counter Oct 04 16:12:29 http://londerings.novalis.org/wlog/index.php?title=Neo1973_counter Got it, get yours from openmoko.com (and read topic) (1814;252) Oct 04 16:19:17 hmm, the next fullstop Oct 04 16:19:40 openmoko really has changed over the last months :) Oct 04 16:20:08 re Oct 04 16:20:11 does anyone know where i get gecho for gentoo ? Oct 04 16:21:24 gecho or gecko? Oct 04 16:21:49 which: no gecho in (/bin:/usr/bin) Oct 04 16:22:00 What's gecho? Oct 04 16:22:07 dont know either Oct 04 16:22:18 make qemu needs it Oct 04 16:22:18 heh Oct 04 16:22:36 ( cd build/qemu && openmoko/flash.sh ../tmp/deploy/glibc/images/fic-gta01 ) Oct 04 16:22:37 /usr/bin/gmake Oct 04 16:22:37 which: no gecho in (/bin:/usr/bin) Oct 04 16:22:37 which: no pngtopnm in (/bin:/usr/bin) Oct 04 16:22:37 Splash needs 'pngtopnm' and 'ppmtorgb3' - make sure that they're in $PATH. Oct 04 16:22:37 make: *** [flash-qemu-local] Fehler 255 Oct 04 16:30:44 thos: 11 digits for a phone number? (ref #802) Are you sure that's the max? Oct 04 16:31:39 LarstiQ: no, but I believe that's what is used for the searching Oct 04 16:31:51 LarstiQ: so i.e. maximum search resolution is 11 Oct 04 16:32:05 LarstiQ: at least, that's what I have assumed, I didn't look in detail Oct 04 16:32:06 thos: mkay Oct 04 16:35:46 openmoko: 03thomas * r3086 10/trunk/src/target/OM-2007.2/applications/openmoko-dialer2/ (ChangeLog src/moko-contacts.c): * src/moko-contacts.c: (moko_contacts_add_contact): Fix a possible crasher Oct 04 16:35:51 abraxa_, gecho is the windows prog, pngtopnm the linux prog Oct 04 16:35:58 its working now Oct 04 16:36:05 Ahh Oct 04 16:36:08 Good to know :) Oct 04 16:36:50 ( cd build/qemu && /sbin/mkdosfs -C -F 32 -v openmoko/openmoko-sd.image 500000 ) Oct 04 16:36:50 /bin/sh: /sbin/mkdosfs: Datei oder Verzeichnis nicht gefunden Oct 04 16:36:50 make: *** [build/qemu/openmoko/openmoko-sd.image] Fehler 127 Oct 04 16:36:50 josephine openmoko # which mkdosfs Oct 04 16:36:50 /usr/sbin/mkdosfs Oct 04 16:36:51 nice :) Oct 04 16:41:48 thos: Is Chris still around? If so I would appreciate if he could have a quick look at bug #863 Oct 04 16:42:35 openmoko: 03thomas * r3087 10/trunk/src/target/OM-2007.2/applications/openmoko-dialer2/ (ChangeLog src/moko-keypad.c): * src/moko-keypad.c: (moko_keypad_init): Remove hard coded padding from keypad Oct 04 16:44:09 abraxa_: I poked him Oct 04 16:44:13 Thanks :) Oct 04 16:44:33 huh, my self-made image with MokoMakefile appears to hang with the progress bar *almost* at the end. :( Oct 04 16:44:55 and I can't reach it over usb. Oct 04 16:45:58 orospakr: Sounds like the gsmd issue... Oct 04 16:46:09 I can't power it down, either. Oct 04 16:46:42 orospakr: http://bugzilla.openmoko.org/cgi-bin/bugzilla/show_bug.cgi?id=788 Oct 04 16:47:06 hm, I had to pull the battery, just like the bug said Oct 04 16:47:17 abraxa_: he's going to commit your patch :-) Oct 04 16:47:30 thos: Yay! Oct 04 16:47:53 hmm... I hope the order of all the items in the shipment tracking page for my NEO isn't quite correct... In the order the items are listed it goes from .us -> .nl -> .us -> .nl :) Oct 04 16:48:03 (all within a few hours... :P) Oct 04 16:48:28 Mek: I've seen that before Oct 04 16:48:52 abraxa_: I pulled this morning, so I assume the bug is still there? Oct 04 16:52:20 probably billing info received->arrived in my city will be shorter than arrived in my city->package delivered... unless they plan on delivering it in the middle of the night :P Oct 04 16:54:34 openmoko: 03chris * r3088 10/trunk/src/target/OM-2007.2/applications/openmoko-today2/ (ChangeLog src/today-task-manager.c): Oct 04 16:54:34 openmoko: * src/today-task-manager.c: Oct 04 16:54:34 openmoko: Fix crashing on closing applications from task manager (hopefully), Oct 04 16:54:34 openmoko: thanks to patch from Soeren Apel Oct 04 16:54:38 (script) wiki RecentChanges: [[My_Account/it]] [[Talk:My_Account/it]] [[Talk:MokoMakefile]] [[Group_Order]] [[Main_Page/ee]] [[User:ThomasWood]] [[No_Linux]] [[Widgets]] Oct 04 16:57:04 orospakr: Yeah, it's still unresolved cause Harald seems overwhelmed with work :) Oct 04 16:57:25 oh well, I'll put my Neo aside for now and get some work done. :) Oct 04 16:57:36 orospakr: back to the X0? :0 Oct 04 16:57:48 sshhh! ;) Oct 04 16:57:51 lol... Oct 04 16:58:01 I haven't forgotten. Oct 04 16:58:33 http://www.jesc.ch/openmoko Oct 04 16:58:40 just curious.. Oct 04 16:58:44 what is that thing? Oct 04 16:59:24 uh Oct 04 16:59:33 they claim to sell GTA02 units. Oct 04 17:00:15 they claim to take pre-orders for GTA02 units Oct 04 17:00:54 it was linked from http://wiki.openmoko.org/wiki/Group_Order Oct 04 17:00:59 I think it was announced on the community-list earlier Oct 04 17:02:48 ah, fair enough. Oct 04 17:03:25 http://lists.openmoko.org/pipermail/device-owners/2007-August/000323.html Oct 04 17:03:31 there's the mail Oct 04 17:03:42 i wonder if i should order through switzerland or straight from fic :/ Oct 04 17:04:38 I just got my neo! Oct 04 17:04:44 now... how do I charge the battery? Oct 04 17:05:33 paron_work: connect it to an usb host Oct 04 17:05:45 that easy, eh? Oct 04 17:05:58 hehe, it booted straight into a kernel panic :D Oct 04 17:06:16 paron_work: did you flash a rootfs? Oct 04 17:06:24 abraxa_: btw, I was seeing the issue on boot, not shutdown. Oct 04 17:06:27 nope! :D Oct 04 17:06:35 I just excitedly got it out of the box and assembled it Oct 04 17:06:42 paron_work: I recommend reading http://wiki.openmoko.org/wiki/SH1 and http://wiki.openmoko.org/wiki/Getting_Started_with_your_Neo1973 Oct 04 17:06:53 paron_work: it comes with a kernel image, but no rootfs flashed. Oct 04 17:06:57 LarstiQ: thanks! Oct 04 17:07:08 this pen is awesome Oct 04 17:07:20 for everything but using the neo. :) Oct 04 17:10:21 this is the best thing that's ever happened to me, lol Oct 04 17:10:26 * paron_work is so excited Oct 04 17:10:45 orospakr: I saw it on boot, too Oct 04 17:17:24 Actually I was wondering about that myself... the Neo pouch has a pen holder sewed in on the side, however the stylus is way too long to fit... Oct 04 17:17:42 it has =O Oct 04 17:17:44 ? Oct 04 17:33:14 For those who wonder what I mean: http://abraxa.dyndns.org:81/random/neo-pouch.jpg - the stylus that comes with the Neo is even longer... Oct 04 17:36:52 abraxa_: what's that thing you have connected to your neo? Oct 04 17:42:14 bye Oct 04 17:43:11 Vegar: Oh, that's just a 2.5mm speaker plug Oct 04 17:43:38 ah, ok Oct 04 17:44:13 connected to a Bose hifi system :) Oct 04 17:45:42 Something like that, yep ;) Oct 04 17:57:15 abraxa_; i'm right, you are the developer of mediaplayer? Oct 04 17:58:08 fgau: Yup - what's up? Oct 04 17:59:54 great, i have one hint. if i add a track to the playlist, i have no response if i click to the file Oct 04 18:00:24 i wish you can add somthing like a checkbox Oct 04 18:00:40 so it's clear, the track is added Oct 04 18:01:14 what do you mean? Oct 04 18:01:32 Actually, current SVN does show a "file added" dialog for single files as well but that'll be improved later :) Oct 04 18:01:53 Having to click a dialog away for every single file would be too annoying Oct 04 18:02:44 oh, i'm not uptodate .( Oct 04 18:03:08 That's fine, I don't expect anyone to always run SVN HEAD ;) Oct 04 18:03:33 Thanks for the hint btw Oct 04 18:03:36 abraxa_: but you have made a nice app !!! Oct 04 18:05:01 Thanks, though it's not as nice as it will be - I just made it usable for now so people have something to play with while I can keep working on it, hehe Oct 04 18:06:54 * fgau will clean-package-mediaplayer and build the new one Oct 04 18:07:14 I'm not sure that'll help Oct 04 18:09:13 what must i do to build the new on? Oct 04 18:42:14 wtf? now my neo that arrived in eindhoven where I live earlier today is listed as "in transit to Koeln, DE"... Oct 04 18:43:31 Mek: don't worry, the states are often completely wrong :) Oct 04 18:43:35 thats better than amazon.com sending books to Austria instead of Australia Oct 04 18:43:51 :) Oct 04 18:43:53 hehe Oct 04 18:52:03 is anybody working on handwriting recognition for openmoko? Oct 04 18:52:16 i couldn't find anything on the projects page Oct 04 18:52:24 Not that I know of Oct 04 18:53:25 what about language support, namely chinese Oct 04 18:53:29 is there anywhere i can read about that? Oct 04 18:53:42 qtopia has handwriting recognition Oct 04 18:54:24 ljp: did that happen? (amazon, austria) Oct 04 18:54:25 Doubtful, though FIC's mobile division does have staff highly specialized on Chinese IMEs Oct 04 18:54:58 (script) wiki RecentChanges: [[MokoMakefile]] Oct 04 18:56:01 but, chinese fonts are well supported? Oct 04 18:56:11 cjb_ie: yes. my books went from US->Austria->Australia Oct 04 18:56:18 enjahova: support for Chinese et al is going to need support for UCS2 -> UTF-8 and a better approach to i18n throughout Oct 04 18:56:35 s/support/sms support/ Oct 04 18:56:35 Writchie meant: enjahova: sms support for Chinese et al is going to need support for UCS2 -> UTF-8 and a better approach to i18n throughout Oct 04 18:56:43 enjahova: I'm sure that's being thought about and worked on when they have the time Oct 04 18:57:35 apps which don't handle i18n and gettext() etc. will not survive the long term Oct 04 18:57:58 but right now it's a bit more important to just get the plumbing to work Oct 04 18:58:08 Writchie: Though sadly UTF8 is still a PITA with glib Oct 04 18:58:35 I don't think so Oct 04 18:58:36 i18n will always be somewhat of a PITA unfortunately Oct 04 18:59:13 mjr: You don't think so about... what? Oct 04 18:59:13 at least gettext is now lgpl Oct 04 18:59:35 that the unicode jazz is much of a PITA Oct 04 19:01:02 mjr: Well, I think so because tasks that are simple in ASCII become difficult/slow with UTF8 and glib - e.g. extracting a file's file name and/or file extension Oct 04 19:01:13 hm, i have no experience with this kind of stuff. but i'd really like to get into il8n and chinese<->english interaction Oct 04 19:01:47 abraxa_, umm, wy's that any more difficult with utf-8? Oct 04 19:01:55 abraxa_: UTF8 is pretty elegant compared to wide chacters Oct 04 19:02:03 why Oct 04 19:02:27 because you don't have to change everything everywhere Oct 04 19:02:42 all those strwcpy etc. Oct 04 19:02:43 I don't have a problem with wides, but there are handiness issues with C, yes Oct 04 19:03:05 generally higher level languages are goodness in this respect too ;] Oct 04 19:03:34 much better today than a decade ago Oct 04 19:03:47 mjr: Cause I need to make sure the string is valid UTF8, then iterate over each and every char - and glib can only do that from front->back, not back->front, etc... so it's a PITA and a performance eater in my opinion Oct 04 19:04:09 so does openmoko use utf8? and if i'm developing an app, can i do it all using utf8 Oct 04 19:04:11 but anyway, if that wasn't clear, with utf-8, basename and extension separation goes _exactly_ like in ascii Oct 04 19:04:12 True, high-level languages make that easier but they're not exactly suited for embedded systems most of the time Oct 04 19:04:26 especially if it will be dictionary type app, where text gets copy pasted in and out Oct 04 19:04:56 i think i need to do some basic reading, any suggestions? like, where do i start so i can follow this conversation... Oct 04 19:05:04 abraxa_, you don't have to use the glib tools for everything, you know Oct 04 19:05:13 mjr: What else can I use? Oct 04 19:05:25 abraxa_, _do exactly as you'd do with ASCII_, as I said Oct 04 19:05:26 there are some practical issues for openmoko - like how many 'characters' will fit in an SMS message Oct 04 19:05:46 which depends on the over the air character set. Oct 04 19:06:06 s/set/encoding/ Oct 04 19:06:06 Writchie meant: which depends on the over the air character encoding. Oct 04 19:06:43 anyway, validation sure is some overhead when you get an untrusted byte string, but to call such little things performance eaters, pfft Oct 04 19:06:48 mjr: But what it the file name contains multibyte chars? Imagine I walk from back to front to find the dot (.) and find one - but it really is just one of the bytes that belongs to a multibyte char... that can get nasty quickly imo Oct 04 19:07:24 abraxa_, read up on UTF-8. All octets in valid UTF-8 that have the highest bit set to 0 are representing ASCII characters. Oct 04 19:08:06 Hmm Oct 04 19:08:20 enjahova: UTF8 is pretty much defacto now at UI level - everything will be converted to that Oct 04 19:08:35 enjahova: don't forget right to left issues as well Oct 04 19:08:39 and yeah, enjahova, the toolkits use utf-8 so openmoko does too Oct 04 19:09:09 mjr: Maybe it is easier then than I thought... guess I should do more reading indeed ;) Oct 04 19:09:46 thanks guys, so utf8 is what i'll become familiar with :D Oct 04 19:09:47 indeed; they _have_ thought it through rather well, also wrt. backwards compatibility Oct 04 19:11:02 yeah, i was well impressed when i read the utf-8 manpage Oct 04 19:11:22 Makes me wonder why glib makes things so complicated then... Oct 04 19:11:32 having it work on POSIX pathnames, utf-8 strings sorted in byte order being also code-point ordered Oct 04 19:12:03 quite nice for stuff drafted on a napkin in a bar Oct 04 19:12:09 hehe Oct 04 19:12:12 or diner, methinks Oct 04 19:12:55 mjr: it was influenced by an early ad-hoc chinese encoding - i forget the name Oct 04 19:12:59 ...oh and it's self-synchronizing too Oct 04 19:13:16 mjr: a derivative of gb Oct 04 19:15:36 sorry, placemat, http://www.cl.cam.ac.uk/~mgk25/ucs/utf-8-history.txt Oct 04 19:15:42 oh, and the canonical encoding for any utf-8 char is considered to the only correct one Oct 04 19:15:57 this is to make equality comparison robust Oct 04 19:16:31 Though with comparison, you have to remember to normalize too. As with all Unicode. Oct 04 19:16:56 so if you're worrying about something like SQL injection, say, you only need to check for the byte value of ' and not any other way it might possibly be encoded. Oct 04 19:17:42 looking for ascii markers is of course trivial, yes Oct 04 19:18:39 though that wasn't in the first rev of utf-8 Oct 04 19:19:01 Does somebody know what could be wrong when I get the following error message while flashing a qemu-openmoko: "U-boot failed to finish writing in 300 seconds, giving up."? Oct 04 19:19:12 not sure, I just know that the first rev wasn't self-synchronizing Oct 04 19:30:18 nice.. latest kernels resume out of sleep mode nicley on incomming call. Oct 04 19:32:45 folken: OM or qtopia? Oct 04 19:33:22 writchie the kernel should be same Oct 04 19:34:44 i meant what it was testing on - also did the alsa driver resume correctly? Oct 04 19:35:57 after all - PM is presently only problem keeping it from being usable as a phone Oct 04 19:36:19 ...and gsmd ;) Oct 04 19:36:37 except qtopia doesn't use gsmd Oct 04 19:37:02 Qtopia... this is #openmoko ;p Oct 04 19:37:39 openmoko is also the distribution Oct 04 19:37:40 qtopia is fine as a testbed/reference point Oct 04 19:37:54 * zecke is amused Oct 04 19:39:58 * ScaredyCat slaps zecke a few times... Oct 04 19:40:22 ScaredyCat: do you want to get back to my ignore list? or why are you not amused? Oct 04 19:40:47 Writchie: om Oct 04 19:40:58 just seeing if you're still amused after a couple of slaps Oct 04 19:40:58 tx Oct 04 19:41:10 Writchie: yes the sound level was correct. Oct 04 19:41:25 kool Oct 04 19:42:08 and no fumbling with the matchbox session Oct 04 19:45:00 whats PM Oct 04 19:45:21 power management Oct 04 19:45:24 oh ok Oct 04 20:27:05 soooo... has that gsm patch been merged yet? ;) Oct 04 20:31:10 orospakr: No; "ipkg update && ipkg upgrade" produces no improvement... Oct 04 20:33:20 gnu: I don't think it would, actually -- the packages in the repository aren't rebuilt every time someone commits. The only way to do get those changes is actually to rebuild your own image. Oct 04 20:34:27 orospakr: I would hope that every checkin doesn't immediately dump that change to every end-user. But I'd also hope that there's somebody pushing stable changes into the repositories -- particularly for the top ten bugs that keep everybody from using the platform! Oct 04 20:34:50 orospakr: what's the sense of having a repository if it doesn't serve this function? Oct 04 20:35:08 gnu: yeah, that is exactly my feeling on the subject as well. Oct 04 20:35:24 <|R> i put myself in red-pill mode, go there, click install openssh ( http://mg.pov.lt/770/reflash-n800.html ) then open, and it says fails... what kind of newb error am i doing? ;) Oct 04 20:35:32 <|R> oops sorry, wrong window :P Oct 04 20:35:41 To be honest, I haven't entirely figured out the role of the repository in the OE world. Oct 04 20:35:50 I assume it's not too far different from Debian. Oct 04 20:36:24 Stuff does go into it (ipkg upgrade does change things every few days)... Oct 04 20:37:09 I wonder what the process/policy is for that. Oct 04 20:37:24 Non-working... Oct 04 20:37:32 Har! Oct 04 20:38:15 orospakr: Did you get one of the early phones? I seem to recall you talking about 'em. Oct 04 20:40:11 gnu: I got mine yesterday. It took a few months :P Oct 04 20:40:13 I'm trying (yesterday & today) to build my own copy of the sources with MokoMakefile -- what a clusterfrog. Spends half its time waiting for downloads and the other half maxing out the CPU on compiles. (Occasionally it overlaps those two.) Oct 04 20:40:38 MokoMakefile worked out OK for me, except when NetworkManager decided to toast my wireless halfway through the nite and bomb the build. Oct 04 20:40:53 although, the code is broken and gsmd hangs the phone on boot. D'oh! :P Oct 04 20:40:53 ouch Oct 04 20:40:56 do you really need 12gb? Oct 04 20:41:03 ive been cleaning my hdd in preperation Oct 04 20:41:12 enjahova: pretty close Oct 04 20:41:15 I long for the days when you could download the *source tree* of something, disconnect from the net, and build it. Oct 04 20:41:16 * ScaredyCat throws down the gauntlet to oe ... try filling that disk biatch! Oct 04 20:41:16 altho, ill be doing it on a VM ontop of my macbook... =\ Oct 04 20:41:19 9 GB may be more accurate, with default settings. Oct 04 20:41:34 gnu: yeah, the problem here is that you're building a big aggregation of software, an entire distro. Oct 04 20:41:39 for better or worse... Oct 04 20:41:40 9gb isn't enough for a full build Oct 04 20:41:57 is that 9gb of source? Oct 04 20:42:11 No Oct 04 20:42:11 With every Debian, Fedora or Ubuntu ISO I download, I also download the *source ISOs*. I have this theory that the build process is repeatable. Oct 04 20:42:19 9gb of source + intermidiate binaries Oct 04 20:42:22 oh ok Oct 04 20:42:31 * ScaredyCat just upgraded lappy hdd Oct 04 20:42:39 * orospakr generally dislikes so-called projects that consist of random bits of Free code glued together in a non-repeatable fashion. Oct 04 20:42:49 gnu: totally. Oct 04 20:43:00 orospakr: it is repeatable.. Oct 04 20:43:06 yeah, OpenMoko is. Oct 04 20:43:09 but only in the same nano second Oct 04 20:43:16 heh Oct 04 20:43:26 I was referring to that problem in general. Oct 04 20:43:43 * ScaredyCat is now building without rm_work Oct 04 20:43:45 It's actually quite hard to pin down the build environment needed to rebuild a distro from scratch. Oct 04 20:44:16 gnu: the problem is the build tries to grab the latest versions Oct 04 20:44:27 Best you can probably do is to test it on itself; and on an unpatched commonly available other distro. Oct 04 20:44:45 ScaredyCat: ...and there's no way to tell it "No, I want Version XXX"? Oct 04 20:45:01 so if i want to do app development, quickest path is to go with MokoMakeFile right? Oct 04 20:45:09 there is, but there are so many files Oct 04 20:45:21 ScaredyCat: I keep seeing people complaining "the build has been broken for 10 days". They can't just build the one from 10 days ago, and make their patches there, then port 'em forward? Oct 04 20:45:48 ScaredyCat: so how do you specify a version? Oct 04 20:45:50 you can't just change 1 rev no and do it Oct 04 20:45:57 srcrev Oct 04 20:46:07 in the .bb files Oct 04 20:46:26 but then of course there's the conflicting svn stuff too... Oct 04 20:46:27 I'm seeing problems getting the initial build going. I fail to download some project... Oct 04 20:46:31 lots of fun Oct 04 20:46:55 Is there some way to tar up a complete snapshot and host it on a reliable server to retrieve with ftp? Oct 04 20:47:03 enjahova: That's the rumor. I'm trying it out now. See the wiki page on MokoMakefile -- and patch it if it doesn't work for you. Oct 04 20:47:06 generally builds go ok, it's just the build environment can be a little sensitive to people not testing their patches Oct 04 20:47:27 gnu, thanks, i'm going to try when i get home :) Oct 04 20:47:42 build environment? Oct 04 20:48:03 oe Oct 04 20:48:03 Somehow, I don't like the idea that OE appears to not be a centralised distro, like debian. Each OE-using project has different revs of everything. Oct 04 20:48:11 He means the "source code"... :-) Oct 04 20:48:23 and forgive my noobness, whats OE? Oct 04 20:48:30 OpenEmbedded Oct 04 20:48:30 openembedded Oct 04 20:48:40 the build system used to build the image from the source Oct 04 20:48:47 thats like the OS? Oct 04 20:48:56 no Oct 04 20:48:56 it's a very good idea... but i'm not 100% convinced about the execution... Oct 04 20:49:10 but then, we are using oe dev.. Oct 04 20:49:16 which is a little insane Oct 04 20:49:17 it's like what's used to put the OS together from hundreds of programs from various sources. Oct 04 20:49:28 oh ok Oct 04 20:49:32 that makes some sense Oct 04 20:49:58 it's a steep learning curve Oct 04 20:50:07 like a cliff Oct 04 20:50:28 lol Oct 04 20:50:48 some of the docs aren't very helpful either.. Oct 04 20:50:58 "we haven't written this bit yet" Oct 04 20:51:19 but every question gets answered with RTFM Oct 04 20:51:44 well you guys seem helpful enough so far :D Oct 04 20:51:45 aye tis true, or a cryptic reply... Oct 04 20:52:04 but there are days, even weeks when the build works fine Oct 04 20:52:29 speaking of docs - dbus isn't much better nor even glib/gtk for that matter Oct 04 20:53:10 I've never understood that weird layout stuff with gtk Oct 04 20:54:23 for your first build.. you're going to need 7 hours or so Oct 04 20:54:49 a bottle of single malt also helps! Oct 04 20:55:08 only the first half helps Oct 04 20:55:17 the 2nd half gets in the way ;) Oct 04 20:55:17 (script) wiki RecentChanges: [[SMSTool]] [[PyPenNotes]] Oct 04 20:55:38 you need to find the right spot on the balmer curve Oct 04 20:55:59 http://xkcd.com/323/ Oct 04 20:56:26 Writchie: lol Oct 04 20:56:31 ScaredyCat: You should be reaching the Ballmer peak! Oct 04 20:56:39 Hey, I type too slow Oct 04 20:56:58 I could use some Ballmer Curve assistance..... Oct 04 20:57:19 * ScaredyCat hands sagacis some water Oct 04 20:57:50 Problem with Balmer peak is no WiFi at the pubs... Oct 04 20:58:08 you go to the wrong pubs then :) Oct 04 20:58:13 * fgau hands sagacis some german beer Oct 04 20:58:34 * sagacis thanks fgau and scowls at ScaredyCat Oct 04 21:01:04 * mwester crashes the party Oct 04 21:03:33 * cb22 crashes, but into a wall Oct 04 21:03:49 * sagacis 's neo crashes Oct 04 21:04:11 * cb22 thinks sagacis is lucky to have a neo Oct 04 21:04:37 * sagacis would have lower blood pressure without a neo. Oct 04 21:04:49 sagacis: how that? Oct 04 21:04:57 * cb22 thinks sagacis should give it to cb22... i'll pay for shipping :) Oct 04 21:05:04 :D Oct 04 21:05:21 I want to use my neo, and keep running into problems. Oct 04 21:05:30 givit to me, i pay for the neo also ;) Oct 04 21:05:35 * mwester finds that 30 minutes with "Quake III Arena" does a good job of stress relief... Oct 04 21:06:16 hallo Oct 04 21:06:19 mwester: i see it gets more & more important to get 3d graph accel on the neo Oct 04 21:06:19 paging roh Oct 04 21:06:26 * ScaredyCat finds 30 minutes.... never mind.... Oct 04 21:06:29 :P Oct 04 21:06:52 whois roh Oct 04 21:06:57 errm... Oct 04 21:07:08 ~seen roh Oct 04 21:07:13 !bot? Oct 04 21:07:34 roh is currently on #oe #openmoko. Has said a total of 44 messages. Is idling for 9h 5m 27s, last said: 'nibbo_ stinks like gcc3 vs gcc4 .. which gcc are you using?'. Oct 04 21:07:44 laggy bot Oct 04 21:07:46 * mwester observes that the magic incantations uttered by ww finally accomplish their goal... Oct 04 21:08:11 * ww thanks apt Oct 04 21:09:06 * ww waves Oct 04 21:14:11 hm, how would I go about using the hardware flow control patch with MokoMakefile? Oct 04 21:21:55 orospakr: Don't worry about that patch if you are just trying to get a recent image built, there are work-arounds that avoid the problem already in place in the recent images. Oct 04 21:22:40 mwester, I don't suppose I could do one of those workarounds with MokoMakefile? Oct 04 21:23:02 what are working webbrowsers on openmoko? Oct 04 21:23:02 The workarounds are already commited, so just build with MokoMakefile, and you'll have them. Oct 04 21:23:32 ~seen LaF0rge Oct 04 21:23:34 laf0rge is currently on #openmoko (9h 2m 39s), last said: 'all: thanks. '. Oct 04 21:23:46 mwester, I just updated and build about 15 minutes ago, and there's no change. Oct 04 21:23:50 s/build/built/ Oct 04 21:23:50 orospakr meant: mwester, I just updated and built about 15 minutes ago, and there's no change. Oct 04 21:24:17 Ok, do you have a line in your /etc/init.d/gsmd file that reads something like "stty -F $GSMDEV -echo" ? Oct 04 21:24:44 mwester, I wonder where that is in the source tree Oct 04 21:24:57 It would be in the image that you built Oct 04 21:25:48 Kero: try midori or openmoko-browser2 Oct 04 21:25:49 mwester, excuse the silly questions, but how should I modify it? I've never heard of loopback mounting a jffs2 image before... Oct 04 21:26:38 build/tmp/rootfs is a copy of what will end up on your Neo if you flash the image -- so no loopback mounting is required. Just cd to that location, and poke about. But treat it as a read-only location, if you make changes, they will be overwritten and won't end up in the image. Oct 04 21:26:55 But that's a great way to see what will be in the image, without flashing the image. Oct 04 21:26:56 and what's this Origyn Web Browser (OWB) that's mentioned at linuxdevices ? Oct 04 21:27:09 what mickeyl's job description again? Oct 04 21:27:12 orospakr it can be written to a ramdrive somehow Oct 04 21:27:45 Kero: it is a WebKit fork by sand-labs (iirc) and they implement the WebCore porting layer as a porting layer Oct 04 21:28:31 orospakr http://wiki.openmoko.org/wiki/Userspace_root_image#Mounting_the_JFFS2_image_on_a_loop_back_.28MTD_RAM.29_device Oct 04 21:28:34 mwester, http://pastie.caboo.se/103893 Oct 04 21:29:04 it looks like the workaround is already there. Oct 04 21:29:11 perhaps my hang issue is a different bug entirely. Oct 04 21:29:21 orospakr: Yes, you have the gsmd startup script there that "tiptoes" around the flow-control problem. Oct 04 21:29:27 How can I disable the splash logo? Oct 04 21:29:42 Put "psplash=false" on the boot args command line. Oct 04 21:30:00 zecke: neither "fork" nor "sand-labs" mentioned in that article. and they do have a Neo screenshot :) Oct 04 21:30:40 * Writchie heads off to find the Ballmer peak Oct 04 21:31:01 hm, how is that done without a keyboard-equipped bootloader? Oct 04 21:31:01 it's also factually wrong, midori uses webkit, too, no? Oct 04 21:31:08 * sagacis searches at his desk for the Ballmer peak Oct 04 21:31:25 orospakr: USB cable to a host Oct 04 21:31:30 so yeah.. in the latest "official" build.. dialer blows up on me Oct 04 21:31:33 Kero: it can use webkit, and the author of midori hangs around in #webkit Oct 04 21:31:33 ah, right. Oct 04 21:31:45 i've not done an update yet (will do so later) Oct 04 21:31:47 (tho that doesn't tell which one of the two was first) Oct 04 21:32:29 daMaestro: I perceive that it's all become a bit unstable recently. Probably a good thing after such a long quiet spell, it means that development is happening somewhere. Oct 04 21:32:51 yeah, i know Oct 04 21:32:56 was just casually reporting my results Oct 04 21:33:38 I have a build started, I'll see if it works for me. What's the failure? Oct 04 21:34:13 I have a very fresh build just completed. Oct 04 21:34:26 hot out of the oven, then... Oct 04 21:34:43 but your neo is not behaving, or is it your host? Oct 04 21:35:26 mwester I'm not sure yet. Oct 04 21:35:34 * sagacis still hasn't run memtest86 Oct 04 21:37:02 mwester I built from a clean directory. All new get. I'll try that one. Maybe I'll try qemu first.... Oct 04 21:39:44 Have we tried "make flash-neo-rootfs-local" before? Oct 04 21:51:40 hm, I think the default serial settings are wrong in the USB console; typing stuff too fast makes characters come up wrong or not at all, etc. Oct 04 21:52:16 Hey, my build booted! Oct 04 21:52:30 congrats! Oct 04 21:52:57 Now, can I add my neod patch to the build and still boot? I'll have to try after helping my mother-in-law cut tile. Oct 04 21:53:23 grrr Oct 04 21:54:07 whatever Oct 04 21:54:15 * sagacis can't even get the /nick working. Oct 04 21:55:52 something is spitting out crap again (on Neo console) after I did some stuff with libgsmd by hand a few days ago. Time for a new image! Oct 04 22:05:56 yeah, there is definitely something wrong with the console over USB; characters and lines in both directions are getting dropped. Oct 04 22:07:50 orospakr: I've always seen that, but not when I first connect. Oct 04 22:08:02 it happens pretty soon. Oct 04 22:08:02 It's after I type soemthing or printenv. I'm not sure what the trigger is. Oct 04 22:08:13 what's the quickest way to append something to the boot args? Oct 04 22:08:31 * sagacis blew right past the Ballmer peak and is trying to slide back to the left on the graph. Oct 04 22:08:44 orospakr: careful, slow typing. Oct 04 22:09:05 whenever I try cut/paste or something similar, I get dropped characters. Oct 04 22:09:15 I just type each character slowly in a new connection. Oct 04 22:09:33 something is stil fscked with `make *qemu` Oct 04 22:12:15 Kero: Does it still give you the old image? Oct 04 22:12:38 hi 2 all Oct 04 22:13:20 sagacis: did that, since the flash.sh script figures the wrong image from the set. Removing the old ones helped. Oct 04 22:13:46 aha! Oct 04 22:13:56 the problem was that X itself was broken. Oct 04 22:14:15 sagacis: but it fails on flash-qemu-local and when I get impatient and kill it,, there's still qemus running. See any subsequent attempt die in weird ways :) Oct 04 22:14:27 no server '/usr/bin/Xipaq'. my question is, why the heck is xinit set to Xipaq?! Oct 04 22:14:30 Ah. Not great Oct 04 22:14:30 fresh image refuses to launch X. Oct 04 22:14:37 I have little problem :/ my moko is too lasy to charge its own battery :/ why? Oct 04 22:14:41 Kero, yeah, that's what I'm seeing. Oct 04 22:14:42 I used qemu manually when I was at that stage. Oct 04 22:14:46 orospakr: iPAQ ? Oct 04 22:14:55 Now, I'm converting to the commit w/o testing metaphor used so often on openmoko Oct 04 22:15:13 yoyo: Welcome to neo ownership Oct 04 22:15:19 there's an iPAQ on my desk here (Switched off for ages) Oct 04 22:15:21 Kero: X not starting because xinit thinks the Neo is an Ipaq. Oct 04 22:15:26 iPAQ. :P Oct 04 22:15:46 sagacis: i have moko ~2 mounts :) Oct 04 22:16:13 * mwester goes to check the rootfs... Oct 04 22:16:14 orospakr: so, does the Neo have an identity crisis, or is OE being obnoxious? Oct 04 22:17:07 * sagacis is reflashing with the modified rootfs and hopes there's no small limit on reflashing Oct 04 22:17:37 Kero: is yours trying to start xipaq as well? Oct 04 22:17:59 would there be a method to perform the ipkg commands into build/tmp/rootfs before flashing the Neo? Oct 04 22:18:05 this is a fresh build, done with MokoMakefile, from this morning, with no customisations, so I *assume* it's nothing I've done. Oct 04 22:19:05 orospakr: no idea. would have to setup usb stuff for it. lemme see if my laptop-kernel has the gadgetfs stuff around... Oct 04 22:19:22 Kero: if you have a stock distro, then you should. Oct 04 22:19:52 orospakr: for soem reason, debian unstable doesn't :) So had to compile myself. Oct 04 22:20:03 it's there, atm Oct 04 22:20:11 Kero, weird. Oct 04 22:20:16 I guess Ubuntu adds it. Oct 04 22:20:54 DukeOfURL: that's what the oe tasks does -- just add your ipkgs to that file, bump the PR, and rebuild. Oct 04 22:21:14 huh, what actually starts the X server? Oct 04 22:21:20 I can't figure out where the bad config even is. Oct 04 22:21:41 argh, stupid busybox grep has no -r Oct 04 22:21:55 oros och Oct 04 22:22:35 oros in /etc/init.d/ you dont need -r Oct 04 22:22:41 yeah, I know Oct 04 22:22:50 OK, weird: /usr/bin/Xfbdev is a directory. Oct 04 22:22:54 an empty directory. Oct 04 22:22:56 also: Oct 04 22:23:03 root@(none):/etc/X11$ grep -i xipaq * Oct 04 22:23:03 Xserver:XSERVER=/usr/bin/Xipaq Oct 04 22:23:15 both of those things are mighty suspect. :P Oct 04 22:23:26 hm yes Oct 04 22:23:55 oh, btw Oct 04 22:23:56 The x startup file has a number of conditionals, so the XSERVER=/usr/bin/Xipaq is probably ok. Oct 04 22:24:07 But the file being a directory is very bad. Oct 04 22:24:22 what is the significance of "Angstrom"? Oct 04 22:24:28 Can you see if it is also a directory in your rootfs in your OE build envionment? Oct 04 22:24:46 build/tmp/rootfs/usr/bin/Xfbdev Oct 04 22:24:49 I'm seeing a lot of stuff refer to it, and the welcome text displayed by on the VT is a big "Angstrom" in ASCII art. Oct 04 22:24:57 (Angstrom is the distro name) Oct 04 22:25:24 no, actually! It's an ELF executable. Oct 04 22:25:24 what was that key-combo to switch to qemu monitor? Oct 04 22:25:36 for ARM, of course. Oct 04 22:25:41 I wonder what turned it into a directory... Oct 04 22:25:50 orospakr: But it's a directory on the neo after you flashed it? Oct 04 22:25:54 yup! Oct 04 22:25:58 Sigh. Oct 04 22:25:59 weird eh? Oct 04 22:26:01 hm Oct 04 22:26:08 didnt you erase the flash? Oct 04 22:26:13 Yes, and a bugger to fix. I thought we nailed that problem. Oct 04 22:26:18 maybee thats a problem Oct 04 22:26:37 Are you running 64-bit distro? Oct 04 22:26:50 I used the MokoMakefile flash-neo-local target. Oct 04 22:26:57 mwester, no. Oct 04 22:27:05 Ubuntu Gutsy, x86. Oct 04 22:27:21 Gutsy... that's fairly new, then. Oct 04 22:27:25 yes. Oct 04 22:27:32 mwester Others are building on Gusty OK, I think. Oct 04 22:27:46 s/Gusty/Gutsy Oct 04 22:28:15 Anyway,this is almost certainly a problem that has plagued certain special individuals (selected by the gods at random, it seems), and is a result of the "fakeroot-native" package misbehaving. Oct 04 22:28:18 ok no voltage in battery no problem :/ and another ~10 HH of waiting :/ Oct 04 22:28:22 si ja Oct 04 22:29:13 mwester: dang. any other info I should provide you with? Oct 04 22:29:14 Can you find the specific directory name of your fakeroot-native, down in build/tmp/work/*/*fakeroot* ? Oct 04 22:29:58 http://pastie.caboo.se/103911 Oct 04 22:30:36 Ok... Oct 04 22:31:51 * sagacis boots with modified rootfs Oct 04 22:32:12 Yay! My patch doesn't break the rootfs! I just have a broken build directory. Oct 04 22:33:41 mwester, anything else? Oct 04 22:33:50 I have to run off in a few minutes, but I'll be back before too long. Oct 04 22:34:08 Alas, no. If I think of anything, I'll let you know :( Oct 04 22:34:31 thanks. :) Oct 04 22:36:44 with mild success, I'm outta here for a bit Oct 04 22:41:05 ok, I give up. There's CONFIG_GADGEFS 1 in config-host.h, but qemu does not have gadgetfs. Oct 04 22:42:22 got a borked image on my Neo, will try the image on the Neo tomorrow. Oct 04 22:45:25 will be travelling, tho, news on Tue only ;) Oct 04 22:46:58 hey Oct 04 22:54:57 (script) wiki RecentChanges: [[User:Kero]] Oct 04 22:55:04 ah, my neo is back from germany to the netherlands, and "now in transit for delivery" :) Oct 04 22:57:39 so December it is for GTA02...yummy Oct 04 23:00:31 website says to get "openmoko-devel-image-fic-gta01-2007********.rootfs.jffs2"... I'm assuming http://buildhost.openmoko.org/OM2007.2/tmp/deploy/glibc/images/fic-gta01/openmoko-devel-image-fic-gta01.jffs2 is equivalent? Oct 04 23:23:53 mwester, so, any ideas? Oct 04 23:25:01 Anybody here know gconf? Oct 04 23:44:01 How do I submit a patch for .bb files? Oct 04 23:44:56 sagacis: attach the diff to the oe bug tracker Oct 04 23:45:13 No but. New functionality. Oct 04 23:45:19 s/bug/bug/ Oct 04 23:45:21 err Oct 04 23:45:33 slow bot Oct 04 23:45:38 well, OE or OM bugtracker ;) Oct 04 23:46:28 I want to add steps to the postinstall for openmoko2, which is where gconf is initialized Oct 04 23:46:57 OM bugtracker then Oct 04 23:47:00 orospakr: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=366067 Oct 04 23:47:17 Looks to me like nobody has figured this problem out, and it seems to affect more than just OE. Oct 04 23:48:04 mwester, I can't find a bug on the tracker, although I did see mention of a problem with "fakeroot-native or x86_64" causing build issues in an #oe IRC log somewhere. Oct 04 23:48:04 in what releases is the dialer known to work? Oct 04 23:49:29 orospakr_: the original bug was in the OE bug database, and was resolved by upgrading to the 1.8 version of fakeroot. Since then there have been no *identified* problems, so you are probably the first to encounter this failure with OM. Oct 04 23:49:46 paroneayea: I did a update before 3-4 hours and it works fine Oct 04 23:50:00 mwester, have you heard of a "files become directories" issue specifically before? Oct 04 23:50:05 it's the first time it works for me Oct 04 23:50:14 So, open a new bug report in the OM database -- that won't get it fixed, of course, but it might help others identify the issue when they encounter it. Oct 04 23:50:25 hm, fair enough Oct 04 23:50:25 paroneayea: I'm about to try the dialer on a new build. Oct 04 23:50:30 orospakr_: yes, that's what I'm trying to tell you -- we have seen this before. Oct 04 23:50:38 I just wanted to scare up any known info first before filing. Oct 04 23:50:55 orospakr_: did you read the debian bug report that I posted the URL for? Oct 04 23:51:02 mwester, I may have missed it Oct 04 23:51:05 Gotta remember to remove the battery panel applet Oct 04 23:51:18 sagacis: what's wrong with the battery panel applet? Oct 04 23:51:22 orospakr: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=366067 Oct 04 23:51:47 Maybe it should be fixed, but before that was crashing the gsmd for some reason. Oct 04 23:52:23 orospakr_: here's another one - same problem, slightly different symptoms. Oct 04 23:52:39 orospakr: Sorry, I mean crashing the panel applet. Oct 04 23:52:51 One of the OE folks tracked it down to the internal cache in fakeroot getting corrupted, but nobody knows why. Oct 04 23:53:51 http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=425007 <--- yet another report of this problem. Oct 04 23:54:19 It sure would be nice if someone had a solution for the problem instead of just lots of bug reports for it... :( Oct 04 23:54:25 orospakr: if you didn't do "nand erase rootfs" (or flash the august snapshot uboot or a more recent uboot) before reflashing the rootfs, you will get filesystem corruption on the neo Oct 04 23:55:09 orospakr: and having to type s l o w l y on the uboot usb console is a bug which has already been fixed in recent uboots (more recent than the august snapshot one, around the time the gsm passthrough was commited) Oct 04 23:57:55 which uboot do you suggest? Oct 04 23:58:12 btw, I didn't see any errors in the ring buffer which would suggest filesystem corruption, unless it is silent Oct 05 00:00:03 orospakr_: the fs corruption manifests itself in crazy things like files missing, things which should be files or directories being something else, and so on Oct 05 00:00:40 yikes. I'd best upgrade that uboot, then. still, which one should I use? Oct 05 00:00:50 orospakr_: so far, I'm still using the august snapshot uboot (too lazy to break that small piece of paper with the red cross), but I'm thinking of doing some things to uboot's first few pages... Oct 05 00:01:11 orospakr_: as long as you remember to always run "nand erase rootfs" before flashing the rootfs, the corruption doesn't happen Oct 05 00:01:42 orospakr_: the bug was a miscalculation in the amount of space which should be erased automatically when flashing the rootfs (it wasn't erasing enough) Oct 05 00:04:58 And do not mistype that as 'nand erase' Oct 05 00:09:43 ^^ Oct 05 00:10:44 SpeedEvil: I heard about that particular misadventure. :) Oct 05 00:10:54 u-boot should have a confirmation prompt. Oct 05 00:11:05 "THIS WILL BRICK THE DEVICE." Oct 05 00:11:51 orospakr_: uboot follows the unix traditions Oct 05 00:12:00 orospakr_: never ask for confirmation Oct 05 00:12:11 orospakr_: never print anything on success, only on failure Oct 05 00:12:27 orospakr_: assume the user know what he's doing Oct 05 00:12:32 orospakr_: and so on Oct 05 00:12:54 orospakr_: besides, it's not meant to be used by mere mortals... Oct 05 00:13:16 orospakr_: which would be safely using the "padded cell" of the GUIs Oct 05 00:13:24 that's the success of windows: assume the user never knows what he does :) Oct 05 00:14:00 thomasg: Or, assume the user REALLY meant to do something entirely different Oct 05 00:14:03 and just do it Oct 05 00:14:07 :) Oct 05 00:14:15 thomasg: windows is very inconsistent at that... there are lots of places where it lets the user shot himself in the foot with no warning at all Oct 05 00:14:22 "Do you really want to shut down now?" Oct 05 00:14:25 "No!!!" Oct 05 00:14:29 "Shutting down..." :) Oct 05 00:14:45 thomasg: and other places where it doesn't trust the user at all (what was that one, nine prompts to delete a single icon? Oct 05 00:14:46 ) Oct 05 00:15:04 it's almost like it was written by more than one person Oct 05 00:15:10 cesarb: imho it's goot this way: it's gonna be painful if windows would trust it's users :) Oct 05 00:15:14 *good Oct 05 00:16:02 http://www.flickr.com/photo_zoom.gne?id=151250154&size=o <- classic example Oct 05 00:16:37 hehe. I really should take a look at vista, looks funny Oct 05 00:16:41 mwester GSMD still doesn't work on boot for me. I'm going to add the gsmd stop/start to the matchbox session again. Oct 05 00:21:08 cesarb: lol, nice one. I didn't realize the vista wanna-be-auth crap was that bad =) Oct 05 00:21:57 CoreDump|home: ah, that particular example has already been fixed, since it was from one of the betas... but I've heard there are other equally ridiculous situations still remaining ;-) Oct 05 00:22:26 yay, I tink I've verified that my problem was the NAND issue, not the fakeroot issue. Oct 05 00:22:32 I have to admit,windows has a certain entertainment value =) Oct 05 00:22:36 UAC is a very bizarre "you are the administrator, but not quite" scheme... you theorically can do everything, but the OS stops and prompts you when you try Oct 05 00:23:15 not to mention the REAL root acount is SYSTEM, which you don't actually have access to... so even as administrator you cannot do some things (unless you use your administrator powers to make SYSTEM run something for you) Oct 05 00:23:50 I bet in the next major windows version, they implement unix-like user and permission management to finally get it right Oct 05 00:23:53 (and, historically, even normal users could find things like unprotected registry entries which would be read by SYSTEM as a list of things to do (simplifying a bit)... Oct 05 00:24:03 hehe Oct 05 00:24:26 CoreDump|home: and the everyone on unix will be using acls Oct 05 00:24:34 actually doesn Oct 05 00:24:36 t ntfs use acls? Oct 05 00:24:38 windows' overkill security system actually works against it. the permission scheme is so baroque and flexible most people fail to configure it correctly (even people from ms itself) Oct 05 00:25:02 not to mention "Bypass Traverse Checking"... Oct 05 00:25:33 unknown_lamer: I have yet to see a need for ACLs in my home setup =) Oct 05 00:25:34 http://blogs.technet.com/markrussinovich/archive/2005/10/19/the-bypass-traverse-checking-or-is-it-the-change-notify-privilege.aspx Oct 05 00:25:35 I called my other phone from the neo and hung up after one ring. The ring continued until timed out. I'm probably leaving voicemail as I type. Oct 05 00:25:36 sagacis: can you be more clear on how it's failing? Can you post your /etc/gsm.log file? Oct 05 00:25:59 Same thing. ERROR 23 the first time gmsd is started. OK the second. Oct 05 00:26:17 CoreDump|home: ACLs are very useful for sharing a directory or file to another user without having to setup special groups. Oct 05 00:26:21 CoreDump|home: they come in handy if you run a kerberos server and cross domain with your internet host who has an afs share Oct 05 00:26:48 CoreDump|home: and then work on projects with other members of the coop (letting them write to your shared repos and such) Oct 05 00:26:54 hm, the Dialer complains that I'm not registered to the network, and I'm seeing the antenna ? mark in the notification area. gsm.log: http://pastie.caboo.se/103938 Oct 05 00:28:17 I just updated the paste with two new lines (regarding a timeout) that just showed up. Oct 05 00:28:19 unknown_lamer: that sounds very painful =) I guess ACLs are a good thing for enterprise use then. Too bad im too lazy to read up on them ;) Oct 05 00:28:53 sagacis: Ok. You probably should open a new bug report on that as it's a new problem not previously observed (or if it was, nobody else has opened a bug report as far as I can tell). Oct 05 00:28:57 modem dead. :( Oct 05 00:29:03 CoreDump|home: it is actually very trivial and quite convenient Oct 05 00:29:10 CoreDump|home: POSIX ACLs are quite simple (unlike the windows monstrosity). Unfortunately, the page on the project to add them to linux (which had a good explanation) disappeared from the net long ago. Oct 05 00:29:31 CoreDump|home: they are similar to unix permissions except that you have one field -- the list Oct 05 00:29:36 CoreDump|home: then you add users and groups to that list Oct 05 00:29:57 CoreDump|home: the world permission is simulated by a magic anyuser group Oct 05 00:29:57 mwester: I've seen the "no answer to CFUN/COPS" before when you ran them too fast and gsmd got confused, but it doesn't seem to be the case Oct 05 00:30:07 CoreDump|home: much much simpler than having three arbitrary groupings of permissions Oct 05 00:30:38 but the original permissions (U/G/O) still work? Oct 05 00:30:44 sometimes Oct 05 00:30:51 CoreDump|home: yeah Oct 05 00:30:53 afs ignores them I dunno what other filesystems that use acls do Oct 05 00:31:28 unknown_lamer: posix ACLs keep user/group/other, and adds mask and the ability to add arbitrary users Oct 05 00:31:45 (there's also some "default" stuff for directories) Oct 05 00:32:47 I tried clicking on the GSM icon and selecting "power up GSM antenna" and the bar with my battery, bluetooth, etc applets disappeared Oct 05 00:32:52 is this a known issue? Oct 05 00:33:02 anyone know if the little amount of GSM radio noise during calls is due to hardware design or just something in the audio system which isnt setup properly yet. Oct 05 00:33:02 * paroneayea is just trying to figure out how GSM works on here Oct 05 00:33:03 cesarb: are posix acls recursive wrt groups? Oct 05 00:33:03 paroneayea: yep Oct 05 00:33:06 Sounds about right :) Oct 05 00:33:14 paroneayea: it's a very common thing Oct 05 00:33:29 so, the gsm antenna is not on when you boot up? Oct 05 00:33:32 you have to turn it on? Oct 05 00:33:33 paroneayea: just don't try to touch the icons bar, and if you are lucky they won't crash ;-) Oct 05 00:33:35 Psi: probably both. Oct 05 00:33:40 heh Oct 05 00:34:15 paroneayea: it's on when you boot... it's only that gsmd often gets confused and fails to do things (like registering to the network) correctly Oct 05 00:34:19 There's a designed-in amount of noise, because people respond poorly to talking into a completely silent digital phone. The amount of noise may need tweaking. Oct 05 00:34:20 hopefully its fixed in gta02 if its hardware Oct 05 00:34:39 Psi: As I understand it, it's all software controlled. Oct 05 00:34:51 unknown_lamer: don't remember... I never played with the "default" part, which is where any recursion would be Oct 05 00:36:33 if I manually submit the AT commands, AT+COPS=? does not return. Oct 05 00:37:04 orospakr_: who's your carrier? Oct 05 00:37:10 this sometimes works better: AT+COPS? Oct 05 00:37:14 Rogers. Oct 05 00:37:20 They're apparently known to work with the Neo. Oct 05 00:37:21 n8 Oct 05 00:37:26 The SIM card is from 2005. Oct 05 00:37:36 (and known good as of today) Oct 05 00:37:52 orospakr_: and then AT+COPS=0 Oct 05 00:38:08 it doesn't reply anymore. *restarts gsmd* Oct 05 00:38:12 orospakr_: AT+COPS=? for some reason only works after you've registered to the network Oct 05 00:38:23 what command actually registers? Oct 05 00:38:35 ie., what is the minimal conversation necessary to place a call? Oct 05 00:38:38 orospakr_: and when it works it sometimes takes several minutes (on most phones) Oct 05 00:38:48 orospakr_: AT+CFUN and AT+COPS Oct 05 00:38:54 orospakr_: register is AT+COPS=0 Oct 05 00:39:09 I think there is a wiki page on that Oct 05 00:39:10 orospakr_: on libgsmd-tool -m shell, you use "O" and "r" for them Oct 05 00:39:24 orospakr_: I think it's +CFUN=1 and +COPS=0 Oct 05 00:42:20 how can I see if my sim card is registered, and such? Oct 05 00:43:49 paroneayea: Use 'libgsmd-tool -m shell' and type "Q" Oct 05 00:43:53 AT+COPS=0 Oct 05 00:43:53 STR=`AT+COPS=0' Oct 05 00:43:53 RSTR=`+CME ERROR: 3' Oct 05 00:43:53 yeah, I saw that on the wiki. Oct 05 00:43:53 (I'm using -m atcmd to interact directly, because it appears I can do nothing from -m shell) Oct 05 00:44:18 orospakr_: What does AT+CFUN give you? Oct 05 00:44:34 AT+CFUN Oct 05 00:44:34 STR=`AT+CFUN' Oct 05 00:44:34 RSTR=`+CME ERROR: 3' Oct 05 00:44:34 ERROR reding from gsm_fd Oct 05 00:44:34 root@fic-gta01:~$ Oct 05 00:44:41 whoa, it made libgsmd-tool quit. Oct 05 00:45:01 or rather, the daemon itself. Oct 05 00:45:05 Yup. Oct 05 00:45:17 * orospakr_ retries with a fresh gsmd. Oct 05 00:45:18 start with '/etc/init.d/gsmd start' Oct 05 00:45:26 yup, that's what I've been doing. Oct 05 00:45:35 Give it a minute or two Oct 05 00:45:37 sagacis: all it says is "# # Signal strength" Oct 05 00:45:40 Look at /tmp/gsm.log Oct 05 00:45:47 sagacis: before I submit AT+CFUN? Oct 05 00:45:55 s/before/wait a minute or two before/ Oct 05 00:45:56 orospakr_ meant: sagacis: wait a minute or two before I submit AT+CFUN? Oct 05 00:46:14 paroneayea: That's the command name. Try "P" Oct 05 00:46:59 sagacis: does nothin Oct 05 00:47:23 paroneayea: 'O' Oct 05 00:47:29 paroneayea: yeah, that's basically the situation I'm in, so I'm manually submitting the AT commands in the hopes of routing out the trouble. Oct 05 00:47:29 that's a letter Oct 05 00:47:35 s/routing/rooting/ Oct 05 00:47:35 orospakr_ meant: paroneayea: yeah, that's basically the situation I'm in, so I'm manually submitting the AT commands in the hopes of rooting out the trouble. Oct 05 00:47:38 This is what I saw. Oct 05 00:49:16 orospakr_: it's AT+CFUN=1 Oct 05 00:49:26 orospakr_: or AT+CFUN? if you want to know the current value Oct 05 00:49:28 paroneayea: what does your /tmp/gsm.log have? Oct 05 00:49:35 orospakr_: or AT+CFUN=? if you want to know the possible values Oct 05 00:49:44 orospakr_: (almost all commands follow that scheme) Oct 05 00:49:54 OK, it replied STR=`AT+CFUN=1' when I typed AT+CFUN=1, and nothing else. Oct 05 00:50:13 =? replies nothing at all. Oct 05 00:50:18 orospakr_: look at the gsmd log, gsmd gets more confused than usual when you use atcmd Oct 05 00:50:23 I assume the gsm modem itself is trying to associate/register Oct 05 00:50:37 orospakr_: it's not responding--you'll have to restart gsmd Oct 05 00:50:38 orospakr_: you probably just got to an state where gsmd no longer can correctly associate the answers to the commands Oct 05 00:50:56 orospakr_: which is why I prefer -m shell instead of -m atcmd ;-) Oct 05 00:51:41 cesarb: I was hoping to get the cause of the problem from the horse's mouth. :) Oct 05 00:51:59 orospakr_: what I'd do: restart gsmd, wait a little, open libgsmd-tool -m shell, wait a little, type 'O' and enter, wait a little, type 'r' and enter, wait a little -- all the time looking at the gsmd log at /tmp (I've done that tens of times when trying to debug gsmd...) Oct 05 00:52:23 that's what I still do every time I boot. :( Oct 05 00:52:43 orospakr_: but, no matter what you do, always look at /tmp/gsmd.log (keep a tail -f on it on a separate ssh connection, for instance), that's the most reliable way to know what's actually going on Oct 05 00:52:54 yeah, I'm tail -f ing Oct 05 00:53:04 this is basically what I did before, although now I'll be a little more patient. Oct 05 00:55:17 (script) wiki RecentChanges: [[ClipArt]] [[Neo1973]] [[User:Thseiler]] Oct 05 00:56:52 OK, gsmd submitted AT+CFUN=1 and is waiting for a reply... Oct 05 00:57:45 sigh: Thu Oct 4 19:58:35 2007 <1> atcmd.c:562:atcmd_submit() submitting command `AT+CFUN=1' Oct 05 00:57:46 Thu Oct 4 20:00:29 2007 <1> gsmd.c:123:alive_interval_tmr_cb() interval expired, starting next alive inquiry Oct 05 00:57:57 and of course it doesn't reply to the alive query. Oct 05 00:58:06 I guess the GSM model doesn't like my SIM card/network. Oct 05 00:58:12 s/model/modem/ Oct 05 00:58:12 orospakr_ meant: I guess the GSM modem doesn't like my SIM card/network. Oct 05 00:58:13 So, it seems to be locking up at cfun=1 Oct 05 00:58:21 yeah, that's what I saw before. Oct 05 00:58:36 orospakr_: did you try without a SIM? Oct 05 00:58:39 nope Oct 05 00:58:42 * orospakr_ does so Oct 05 00:59:10 * cesarb tests most of the time without the SIM, since I only have one, and changing from the razr to the neo all the time gets tiring fast Oct 05 01:00:32 I"m pretty sure it won't register without a sim. ;-) Oct 05 01:00:48 for sure, but we might learn something. :) Oct 05 01:01:56 The call history is kinda funny. To the openmoko today screen, ALL calls are missed calls. Oct 05 01:04:50 sagacis: CFUN appears to hang the same way without the SIM. Oct 05 01:06:15 how come that applets bar appears sometimes at the top of the UI and not others? Oct 05 01:07:01 yeah, interval is expiring the same way. Oct 05 01:07:10 no apparent difference at all. Oct 05 01:08:05 sagacis: btw, shouldn't the phone still register with a network without a sim to be able to make emergency calls? Oct 05 01:09:51 this really is weird; I assume this problem has nothing to do with any of the software running on the application processor. Oct 05 01:10:51 Mek: It needs a sim to talk at all, but a disabled sim can make emergency calls. Oct 05 01:10:53 Mek: not in the UK Oct 05 01:11:14 * orospakr_ decides to put his SIM back in. Oct 05 01:11:24 orospakr_: I'd assume it does, but it could be similar to using a 3G Cingular (AT&T) sim in the US Oct 05 01:11:39 what happens if you do that? Oct 05 01:11:55 I don't remember. It didn't work, so I e-bay'd an old sim card. Oct 05 01:12:03 I could try, but give me a few Oct 05 01:12:16 hm, I have some old dead ones (ie., no subscription for like 5 years) that I could try. Oct 05 01:12:21 hmm.. okay... my non-neo phone seemed to connect to some network when I had my sim removed... but I didn't want to call emergency services just to check it :P Oct 05 01:12:35 Really? Oct 05 01:12:58 My phone doesn't connect anywhere. Says "Need SIM card". Oct 05 01:13:26 it says "Need SIM card", but after a while the not-connected-icon changed into a connected-icon... Oct 05 01:13:27 Gotta move to the living room... Oct 05 01:17:51 sagacis: btw, this SIM card works in a Nokia 8390. Oct 05 01:22:20 the webkit browser is really cool Oct 05 01:24:29 so out of curiosity Oct 05 01:24:35 did anyone's phone work "out of the box"? Oct 05 01:24:42 besides the flash Oct 05 01:24:51 I hope to find out tomorrow... Oct 05 01:24:52 I mean, dialing & etc Oct 05 01:25:03 ah, OK, apparently there's a TI GSM firmware update in the works that can handle newer SIM cards. Oct 05 01:25:41 that's *real* dangerous Oct 05 01:26:11 I sure hope only TI/FIC can flash the GSM module Oct 05 01:26:31 ah, regulatory/licensing nonsense? Oct 05 01:26:36 well, if that's so, that's a bummer. Oct 05 01:27:07 I don't hope so, but it is so Oct 05 01:27:21 no, I mean we wouldn't want the black hats to be able to hack the radio Oct 05 01:27:30 (well, some outlets maybe also, I dunno) Oct 05 01:28:28 the whole "oh noes hackers could blow up the GSM network if they had access" thing is FUD. Anyone can disrupt the GSM network by building a simple jammer. Oct 05 01:28:39 I have no firm opinion on that. However, it would be nice if one could upgrade the firmware at least from a signed binary. Oct 05 01:28:47 DukeOfURL: if it's anything like what I've read about the razr, the modem firmware is signed (probably with a shared key etched in the silicon) Oct 05 01:28:54 orospakr_, those are readily available on-line too... Oct 05 01:29:01 there you go. Oct 05 01:29:36 orospakr_: well, it's not so much about total disruption as about unfair advantages. the whole gsm protocol seems to be pretty cooperative Oct 05 01:29:52 hm, I'm not convinced... Oct 05 01:29:58 Henryk, it is Oct 05 01:30:01 no, it's about eavesdropping and data theft Oct 05 01:30:10 DukeOfURL: you can change several hundred control bits, change the "flex" (which is a filesystem thingy with even more control stuff and the files stored in the phone), and generally make a bricking mess, but you cannot change the firmware to one which didn't come from motorola Oct 05 01:30:15 DukeOfURL: GSM obviously includes crypto. Oct 05 01:30:17 ... right? Oct 05 01:30:34 gsm includes weak crypto Oct 05 01:30:38 As in: If you set these bits or those bits when registering with a base station you'll get preferred treatment because you must be a first-responder etc. Oct 05 01:30:47 ah, security by obscurity. Oct 05 01:30:49 mjr: the crypto strength depends on the SIM, doesn't it? Oct 05 01:30:50 now that's quality! Oct 05 01:30:57 the gsm firmware does the crypto... Oct 05 01:31:11 mjr: older A5 were weaker, newer A5/8 is stronger Oct 05 01:31:14 so yeah, I suppose there's no hope of ever making the GTA01 units work with new SIM cards. Oct 05 01:31:22 how troublesome. Oct 05 01:31:37 orospakr_, I wouldn't say that. Oct 05 01:31:47 HOLYSHIT IT DIALS Oct 05 01:31:53 yea! Oct 05 01:31:55 AND GETS CALLS Oct 05 01:32:02 mjr: hm, how would it be fixed if the modem can't be flashed? Oct 05 01:32:04 I DON'T EVEN KNOW WHAT I DID DIFFERENT! :D Oct 05 01:32:06 paroneayea: congrats! Oct 05 01:32:10 I don't know if it will be possible, and I'm not saying it'd necessarily be very practical, though Oct 05 01:32:19 orospakr_, sending it in to be flashed is an option Oct 05 01:32:23 yay! Oct 05 01:32:23 You tried again Oct 05 01:32:24 That was my experience all teh way Oct 05 01:32:46 after 15 different shutdowns, powerups, shutdowns, powerups Oct 05 01:32:49 it works :) Oct 05 01:32:50 paroneayea: that's my problem. I can boot and nothing works. After the next boot I can make and receive calls. Intermittent problems. Oct 05 01:33:09 BUT Oct 05 01:33:14 that means that it's a working phone :) Oct 05 01:33:26 even if the software is buggy currently :) Oct 05 01:33:32 paroneayea: are you in a fringe reception area? Oct 05 01:33:36 If FIC will provide the service. Or with luck, some FIC guys might be persuaded to take the necessary equipment into conferences to upgrade thingies etc... But I'm just saying how it _could_ be possible. I don't know how it'll end up. Oct 05 01:33:41 paroneayea: I added '/etc/init.d/gsmd restart' to my /etc/matchbox/session before openmoko-dialer2 Oct 05 01:34:28 That helps it come up more often on boot Oct 05 01:34:49 parag0n: there were (and obviously are) still some timing problems. I hade a reliable image in the past, but unfortunately didn't make a backup and it got deleted Oct 05 01:35:04 sagaci2: ahhhhh Oct 05 01:35:07 I will try that Oct 05 01:35:15 s/parag0n/paroneayea/ (gna, too many nicks, sorry) Oct 05 01:35:36 mjr: well, with commercial phones, the service is provided by third-parties all over the world... Oct 05 01:35:44 so, I'm assuming that contacts aren't read off the sim card yet? Oct 05 01:35:57 you can see them in libgsmd Oct 05 01:35:57 mjr: repair shops which get the materials from the materials (and are supposed to not let other people get hold of them *ahem*) Oct 05 01:36:08 So, gsmd reads them, but openmoko-contacts2 doesn't. Oct 05 01:36:17 s/from the materials/from the manufacturer/ Oct 05 01:36:17 cesarb meant: mjr: repair shops which get the materials from the manufacturer (and are supposed to not let other people get hold of them *ahem*) Oct 05 01:36:23 cesarb, yeah I noted earlier (well, some outlets maybe also, I dunno) Oct 05 01:36:29 sagaci2: I think the modem firmware does this, e.g. you see contact names in the AT dialogue Oct 05 01:37:05 cesarb, presumably would require the outlets to first be persuaded to stock the Neo :] Oct 05 01:37:51 WTF? Rebooting the neo while plugged on USB crashes the gnome battery applet Oct 05 01:38:05 cool. Oct 05 01:38:31 I always knew gnome sucks :P Oct 05 01:39:32 You should use Windows Oct 05 01:39:41 yeah, I really did Oct 05 01:39:49 s/did/do/ Oct 05 01:39:49 Henryk meant: yeah, I really do Oct 05 01:39:54 Henryk: the battery applet has been identified by someone this week (or was it the last one? time passes so fast) as the cause of all (or most) panel crashes Oct 05 01:40:01 nah, I haven't touched windows for a few years, KDE and OSX are good enough for me :) Oct 05 01:40:08 Henryk: so, it's no surprise that it took the panel down that easily Oct 05 01:42:08 cesarb: hmm, at least it doesn't crash the whole panel and a dialogue pops up offering to restart it. what's more infuriating is that some applet quite regularly refuses to freeze when I want to suspend. First it was multiload-applet-2, then I removed that and it was the volume applet Oct 05 02:10:22 I can't load the media player. I get an error dialog that immediately disappears. Oct 05 02:13:33 sagacis: I can't help but notice that's kinda like observing that an automobile's CD player doesn't work, when the auto lacks tires, windows, most of the body, part of the frame, and the engine falls out all the time. :D Oct 05 02:14:00 But, the CD player worked the last time I got in the car.... Oct 05 02:14:05 hehe! Oct 05 02:14:32 But perhaps this time the engine actually starts, instead? Oct 05 02:14:57 Nope. Oct 05 02:15:07 So, I thought I could sit in the driveway and listen to music. Oct 05 02:16:49 I got it too. The error message reads "Error: gstreamer failed to create the PulseAudio sink. Please make sure gstreamer and its modules are properly installed (esp. gst-plugin-pulse). Oct 05 02:18:20 Henryk: Wow, you read fast! Oct 05 02:18:49 No such element or plugin 'gst-plugin-pulse' Oct 05 02:19:54 'ipkg install gst-plugin-pulse' worked Oct 05 02:20:12 Now I have a media player! Thanks, HenryK Oct 05 02:21:27 Hmmm... should gst-plugin-pulse be added to the image then? Oct 05 02:21:35 sagacis: trick 17: Start the program in gdb. Then, when it segfaults, it will pause in gdb and you can read the message Oct 05 02:22:10 * sagacis files that one away. Of course, he normally does that in his day job, so he must be getting sleepy. Oct 05 02:22:21 * mwester wishes the openmoko programs would print a nice message to a log file instead of all of them just seg faulting when they encounter an unexpected event. Oct 05 02:23:00 * mwester _FIXES_ sloppy programming that results in seg faults for his day job. Oct 05 02:23:25 mp3's play kinda skippy Oct 05 02:23:38 * sagacis tells people to fix it. Oct 05 02:23:46 someone else noted that. Don't know why, unless the dialer is chewing up cpu. Oct 05 02:23:50 * sagacis and then fixes it again later. :( Oct 05 02:24:01 I'll top it. Oct 05 02:24:17 midori's running too. Oct 05 02:24:51 A little better without midori and dialer Oct 05 02:25:23 But still skipping? Oct 05 02:25:27 Yup. Oct 05 02:25:30 Pretty bad Oct 05 02:25:37 sagacis: ssh in and run top Oct 05 02:25:43 Is the tune on the jffs or on the SD card? Oct 05 02:26:00 SD card. Oct 05 02:26:09 It should be async, but I haven't checked Oct 05 02:26:19 I noticed that there are quite some differences in mp3 performance between different programs Oct 05 02:27:00 openmoko-mediaplayer takes most of the cpu (48%) Oct 05 02:27:08 pulseaudio takes 25% Oct 05 02:27:28 Good grief! What is the rate for the mp3 encoding? Oct 05 02:27:31 Xfbdev 10% (to show the player, with no display updates) Oct 05 02:27:45 Don't know. I'lll have to go look. Just a song I copied over to try Oct 05 02:28:41 Regardless, I would expect that a 200MHz ARM should be able to play an mp3. Heck, my NSLU2 at 266MHz can easily, without any significant CPU load, play an MP3 to a USB sound device. Oct 05 02:29:25 128K, 44100 Hz, 2 channels Oct 05 02:29:30 * mwester wonders if the plugin is doing something "not nice", like alignment faults, or attempting floating point... Oct 05 02:29:41 yes, I think mplayer takes something like 30% CPU for MP3 playback (direct on alsa, no pulseaudio) Oct 05 02:29:58 The player is taking the hit, not pulsaudio (which hosts the plugin, I'd guess) Oct 05 02:30:30 It's skipping pretty bad Oct 05 02:30:33 badly Oct 05 02:32:32 hmm, yes, ogg playback with openmoko-mediaplayer is also skipping badly Oct 05 02:32:49 64% for openmoko-mediap, 20% for pulse Oct 05 02:32:59 10% for Xfbdev Oct 05 02:33:02 CPU's staying at 75%, too Oct 05 02:33:17 Henryk: That's pretty much all cpu you have Oct 05 02:33:48 Why is the play button backward? Oct 05 02:34:11 sagacis: I wondered about that too Oct 05 02:34:20 ok, mp3 playback is also skipping, though not so badly Oct 05 02:34:22 Should the mediaplayer app be doing much of anything when playing? gst should be doing all the real work. Oct 05 02:35:21 what is openmoko-mediaplayer built on? Oct 05 02:35:30 sagacis: I'm guessing visualization or something? Oct 05 02:35:31 Marshmellows Oct 05 02:35:41 hehehehe Oct 05 02:35:45 Henryk: It's not doing any viz Oct 05 02:35:53 * sagacis is getting loopy Oct 05 02:35:53 sounds to me like highly-under-optimised code Oct 05 02:36:26 Pass every packet through the gui? Oct 05 02:36:33 another data point: madplay on command line, direct on soundcard without pulse: same mp3 file takes 20% cpu, plays with no skips Oct 05 02:36:42 That's more like it! Oct 05 02:36:58 I've used madplay on the device before (when I first got it) and was duly impressed Oct 05 02:37:28 ogg123 uses 76%, plays with no skips Oct 05 02:38:19 should use libmad Oct 05 02:38:51 * raster makes mental notes for a todo list he likely will never have time to get to Oct 05 02:39:40 however, it sounds distorted. the playback was sounding _much_ better in the past Oct 05 02:40:03 * sagacis suggests raster posts the list on the wiki so others can whittle away at it Oct 05 02:40:09 ls Oct 05 02:40:11 oops Oct 05 02:40:48 Ah, that's more like it: Restoring an old alsa state file makes it sound better again Oct 05 02:44:02 sagacis: i'll get to it Oct 05 02:44:10 i should just make my own wiki page Oct 05 02:44:19 "raster's crap to do - please help if you can" Oct 05 02:44:20 :) Oct 05 02:44:24 Put it on your todo list :D Oct 05 02:44:30 i will Oct 05 02:44:43 "Pick up the clothes at the cleaners" Oct 05 02:45:01 "Clip my toenails" Oct 05 02:45:13 "Change media player in Openmoko" Oct 05 02:45:15 Maybe stick with programming tasks. ;-) Oct 05 02:47:38 hahahaha Oct 05 02:49:42 I tried to build jamvm, but the sourceforge project doesn't have the tar.gz any more. They are using bz2. Oct 05 02:52:14 *sigh*, I seem to have lost gllin. :( Oct 05 02:52:32 Yup. I never had it. :( Oct 05 02:53:09 someone posted a link one day, and I fetched it, but alas, I can't find it anywhere. Oct 05 02:53:15 something like totally_not_gllin.tar.gz :( Oct 05 02:53:28 * orospakr grumbles about lawyers. Oct 05 02:55:37 (script) wiki RecentChanges: [[OpenMoko_Local_Groups:_Sydjylland]] [[OpenMoko_local_Groups:_Copenhagen]] [[OpenMoko_local_Groups:_Aalborg]] [[OpenMoko_local_Groups:_Sydjylland]] [[OpenMoko_Local_Groups]] Oct 05 02:56:01 orospakr: save some grumbling for the Openmoko/FIC team; they too own at least some of the responsibility for the current pickle we're in with regard to the GPS device. Oct 05 02:56:14 mwester, are you sure? Oct 05 02:56:28 thats being fixed Oct 05 02:57:18 * orospakr decides to take the brazen approach in the short term: LOL CAN HAS GLLIN WAREZ PLZ KTHXBYE Oct 05 02:57:19 so we're told Oct 05 02:57:26 orospakr: From my experience with vendor relationships, there were things that were done that do not make sense. Oct 05 02:57:53 I'll not elaborate on this issue, it makes roh very angry when I point out my views on the subject ;) Oct 05 02:58:18 well, I don't want to be to harsh with OpenMoko/FIC. This project has been pretty hard on them, by all accounts. Oct 05 02:58:40 but hammerhead, well, they deserve maximum wrath for shipping a proprietary driver with a particularly evil EULA. Oct 05 02:58:42 ;) Oct 05 02:58:53 they're getting noticed... Oct 05 02:59:08 They are all adults, and should be able to take the criticism in the spirit its intended, and perhaps the next ones will be better. Oct 05 02:59:50 And it would also be nice if someone there would stand up tall, and announce to us all what, if anything, will be done about the two major unrevealed flaws in the P1. **** ENDING LOGGING AT Fri Oct 05 02:59:57 2007