**** BEGIN LOGGING AT Thu Jan 05 02:59:57 2012 Jan 05 07:40:18 moin Jan 05 08:38:32 does someone have git svn checkout of opkg-utils? yocto people would like to move it to their repo with history Jan 05 08:39:13 with svn.openmoko.org down again it's good idea Jan 05 09:19:42 moin JaMa Jan 05 09:38:19 SHR: 03Martin.Jansa 07shr-chroot * re7618236c429 10/ (580 files in 35 dirs): system upgrade Jan 05 10:07:31 hello jama please check shr devel ML ca-certificates missing, patch is added Jan 05 10:10:51 I've seen it, thanks Jan 05 10:14:41 nschle85: and btw please use -M and subject is wrong SOB line should be last Jan 05 10:59:43 jama: sorry i did not remember all the options and the shr wiki is incomlete Jan 05 13:28:31 jama: please subscribe to new jenkins, domains are moved, login via ssh-key and change unix password Jan 05 13:29:41 jama: please do not change any config, i have not synced all configs Jan 05 13:31:05 hmmm autosuspend it's not working in shr-core Jan 05 15:22:24 DocScrutinizer, mrmoku for the fowarder: Jan 05 15:22:58 1) should I change to queues for buffer management implemented with linked list Jan 05 15:23:27 2) what should I do about the buffer sizes, I used the same than the hardware Jan 05 15:23:53 that is to say snd_hwparms_get_buffer_size or something like that on modem + CODEC Jan 05 15:23:57 and I took the biggest one Jan 05 15:24:22 for the transfer buffer Jan 05 15:24:29 that is to say the ring buffer in which I write Jan 05 15:24:41 for the writei/readi I kept the period size Jan 05 15:24:58 that is to say codec writei => codec period size as frames number Jan 05 15:37:14 GNUtoo: hmm Jan 05 15:37:31 GNUtoo: and result was not good? Jan 05 15:37:47 result of what? Jan 05 15:37:59 I don't know but the following branch is almost ok: Jan 05 15:38:13 I mean why do you want to change to queues? Jan 05 15:38:31 remotes/origin/gnutoo/gta04-fsoaudiod-forwarder Jan 05 15:38:38 but that one not: Jan 05 15:38:46 remotes/origin/gta04-fsoaudiod-forwarder+automatic_buffers_size Jan 05 15:38:55 because it fits well the use case Jan 05 15:39:04 simplier to handle etc... Jan 05 15:39:04 for instance Jan 05 15:39:27 if you have a buffer underrun you play silence and have already poped the buffer part Jan 05 15:39:54 if you have a buffer overrun in readi you format silence and send silence to the queue instead Jan 05 15:39:59 so it should work well Jan 05 15:40:01 in theory.... Jan 05 15:40:52 hmm Jan 05 15:41:03 but the thing that is the most problematic is 2) Jan 05 15:41:06 I change buffer size Jan 05 15:41:10 and all change Jan 05 15:41:16 sound quality becomes ba Jan 05 15:41:17 *bad Jan 05 15:41:19 etc... Jan 05 15:41:22 let's see what DocScrutinizer51 thinks when he's done with daywork? Jan 05 15:43:17 basically I think I did a mistake trying to use the same buffer size than the hardware Jan 05 15:43:27 because it's hardware params and not software Jan 05 15:44:53 yeah, that might well be true Jan 05 15:45:31 maybe making the ringbuffer hw buffer size + x would work? Jan 05 15:54:41 takes another errr 2..3h Jan 05 15:55:43 ok Jan 05 18:01:48 GNUtoo: hello do you have a shr wiki account ? Jan 05 18:03:08 yes I do Jan 05 18:06:19 yesterday i made a patch but it was not formatted as it should be, can you update please http://shr-project.org/trac/wiki/Howto%20get%20my%20application%20in%20the%20SHR%20feed#FourthStep:SendRecipeToSHR ? so the git commands are correct and where to put any comment and how it should look like is also right ? Jan 05 18:07:30 the git commands seem correct at first sight Jan 05 18:08:00 please read history from today, what jama told me i dont know where using -M Jan 05 18:08:20 ah for renames? Jan 05 18:08:44 renames ? Jan 05 18:08:51 -M[], --find-renames[=] Jan 05 18:08:56 git format-patch Jan 05 18:09:29 hm i thought using git mv does the right things Jan 05 18:09:37 also sed 's/.net/.org/' Jan 05 18:09:41 in the page Jan 05 18:10:04 it does but for sending patches you need a more readable format Jan 05 18:10:13 hence the -M --find-copies-harder Jan 05 18:10:33 GNUtoo: can you add that to the wiki please ? Jan 05 18:10:41 ok I'll do Jan 05 18:11:00 GNUtoo: thank you :-) Jan 05 18:12:50 I changed .net -> .org Jan 05 18:13:38 GNUtoo: i thought about SMS for N900 implementation, ill try but the problem is i have very little time.. so my progress for GPS is also very low... but implemneting SMS is more easier so its better for me to learn about N900 modem Jan 05 18:14:02 done Jan 05 18:14:11 hah, http://www.extremetech.com/computing/103171-why-nokia-is-using-st-ericssons-novathor-arm-chip-instead-of-qualcomm Jan 05 18:15:39 DocScrutinizer51: is that good ? (i heard that qualcomm chips are hard to handle :-) ) Jan 05 18:15:52 DocScrutinizer, how much does the modem control? Jan 05 18:16:46 err qualcom? STE! Jan 05 18:16:57 no, novathor Jan 05 18:17:08 so STE Jan 05 18:17:19 who builds NovaThor? Jan 05 18:17:27 STE Jan 05 18:17:28 x act ly Jan 05 18:17:49 ME ;-P Jan 05 18:17:54 yes I know Jan 05 18:18:15 anyway how much does the modem control? Jan 05 18:20:56 rah: ping Jan 05 18:21:16 DocScrutinizer51, anyway about alsa did you read what I asked before? Jan 05 18:21:28 GNUtoo: I don't think it controls more than what it should Jan 05 18:21:38 ok Jan 05 18:21:47 nice Jan 05 18:22:33 yes, but it will still tsake another hour til I leave here (office) and arrive at home Jan 05 18:22:53 ok Jan 05 18:23:05 GNUtoo: still finding my way thru all the internal docs Jan 05 18:23:50 I am even amazed this HTC is allegedly using novathor, as for all I know true thor isn't reay yet Jan 05 18:24:15 bbl Jan 05 18:25:18 ok Jan 05 18:25:55 I'll do some coreboot work in the meantime Jan 05 18:33:36 GNUtoo: JaMa told me the subject in the patch was wrong, how do i set the right subject ? Jan 05 18:34:09 the pat i sent yesterday in shr devel ML Jan 05 18:34:13 patch Jan 05 18:36:16 nschle85, look in oe wiki Jan 05 18:36:55 GNUtoo: ill do Jan 05 18:37:06 http://www.openembedded.org/wiki/How_to_submit_a_patch_to_OpenEmbedded Jan 05 18:37:19 says: Jan 05 18:37:20 --subject-prefix="PATCH v2" Jan 05 18:37:31 so the best thing would be to get a wiki account Jan 05 18:37:35 for you Jan 05 18:39:26 GNUtoo: wiki account for what ? Jan 05 18:40:02 in other words do you have a wiki account for the SHR wiki? Jan 05 18:40:09 yes Jan 05 18:40:12 ok Jan 05 18:40:17 nice Jan 05 18:40:26 GNUtoo: but i do not have the information :-) Jan 05 18:40:31 ahh ok Jan 05 18:40:46 and you're not very confident with topics you don't know.... Jan 05 18:40:56 GNUtoo: :-) Jan 05 18:42:13 so ill read the oe wiki and update shr wiki, and ill see what JaMa says about my next patch :-) Jan 05 18:42:20 ok Jan 05 18:42:23 thanks Jan 05 19:00:24 GNUtoo: you got a htc desire(?) ? Jan 05 19:01:38 no Jan 05 19:01:40 why? Jan 05 19:01:48 I've htc dream and nexus one Jan 05 19:02:19 well, was it desire that has novathor? Jan 05 19:02:29 no idea Jan 05 19:02:57 * DocScrutinizer51 scrolls back to own link Jan 05 19:03:22 there are more than one htc desire version Jan 05 19:03:37 but the one I know and that captainigloo has has qualcomm Jan 05 19:03:55 ooh Jan 05 19:04:14 maybe wikipedia would tell Jan 05 19:04:28 (about which desire version has which SOC) Jan 05 19:05:30 btw for who has which phones: http://wiki.freesmartphone.org/index.php/People Jan 05 19:08:12 The NovaThor was shipped in a Chinese model HTC Sensation several weeks back, Jan 05 19:09:14 WEEKEND \o/ Jan 05 19:10:50 hmm... did I miss Friday? :P Jan 05 19:11:36 in italy tomorrow is a holliday Jan 05 19:11:42 here too :) Jan 05 19:11:44 ok Jan 05 19:12:05 this night the magic kings are coming! ;) Jan 05 19:12:56 ah ok Jan 05 19:13:07 * GNUtoo is always lost with religious hollidays Jan 05 19:13:10 ohh Jan 05 19:13:15 yeah here too ;) Jan 05 19:13:35 * mrmoku is always lost with _any_ hollidays ;) Jan 05 19:15:56 I think we should protest to change the hollidays to sometihng more simple: Jan 05 19:15:59 * linux days Jan 05 19:16:04 * free software days Jan 05 19:16:06 etc... Jan 05 19:16:13 that way we would remember it Jan 05 19:17:03 yeah nice Jan 05 19:17:08 or eliminate them all Jan 05 19:17:13 who cares anyway :) Jan 05 19:17:36 If you got a boss who's controlling when you work, believe me you won't forget holidays Jan 05 19:18:01 * mrmoku does not have such a thing... so he does not care ;) Jan 05 19:18:40 dang, today It cut thru all my battery Jan 05 19:18:50 quite unusual Jan 05 19:23:01 tobacco shop at main railway station closed at 19:30 today "for xmas celebrations" o.O Jan 05 19:23:26 mrmoku: fiday should be a free day (free from work) Jan 05 19:23:44 didn't realize it was a greek or russian shop Jan 05 19:24:22 DocScrutinizer: do you have a link ? (a speak a little russian) Jan 05 19:51:28 link to what? Jan 05 19:53:30 to the tobaco shop probably :P Jan 05 19:53:49 GNUtoo: so what's been the question? size of fifo buffer for audio? " times size of larger of both hw interfaces' chunksize Jan 05 19:54:49 the paper sign at the door said "am 5. wegen Weihnachtsfeier um 19:30 geschlossen" - no Russian Jan 05 19:55:18 but who's celebrating xmas on 6th of Jan? The orthodox Jan 05 19:55:28 so either Greek or Russian Jan 05 19:56:23 GNUtoo: just read your icon moved ticket Jan 05 19:56:26 strange Jan 05 19:56:27 GNUtoo: so what's been the question? size of fifo buffer for audio? TWO times size of larger of both hw interfaces' chunksize Jan 05 19:57:00 DocScrutinizer: so (dmix buffer + dsnoop buffer) * 2 ? Jan 05 19:57:25 ahh Jan 05 19:57:29 MAX(dmix_buffer, dsnoop_buffer) * 2 Jan 05 19:57:29 of _larger_ of both Jan 05 19:57:30 ok Jan 05 19:57:33 yeah :) Jan 05 19:57:41 back Jan 05 19:57:50 GNUtoo: ^^ Jan 05 19:58:03 you're using dmix/dsnoop on cellmo IF? Jan 05 19:58:16 yes Jan 05 19:58:32 to allow people to play with the modem audio Jan 05 19:58:35 like for instance: Jan 05 19:58:41 then you got a proper ALSA audio device for both ends of your cmtspeech process? Jan 05 19:58:52 for *all 4 ends* Jan 05 19:59:07 cmtspeech thing still has bugs Jan 05 19:59:18 specially for the readi->cmt Jan 05 19:59:27 so the question is that I did this: Jan 05 19:59:43 I got the buffer size trough hwparams_get_buffersize Jan 05 19:59:51 and use that buffer size Jan 05 19:59:59 I guess that's plain wrong Jan 05 20:00:09 DocScrutinizer: this is on gta04 Jan 05 20:00:11 haha i just received a call during upgrade oooops , still going though !!! Jan 05 20:00:16 (shr core) Jan 05 20:00:18 probably, though I first have to see what ALSA naming been like Jan 05 20:00:34 M00R1Z: hehe, nice Jan 05 20:00:35 mrmoku: gathered as much Jan 05 20:00:59 mrmoku: so you actually got a ALSA card for option PCM? Jan 05 20:01:10 mrmoku: upgrading trough ssh maybe the internet dropouts, we'll see Jan 05 20:01:13 yes Jan 05 20:01:34 GNUtoo: so probably you won't need anything homegrown Jan 05 20:02:09 a mere arecord --foo --bar --boo|aplay -D optionALSAaudio will doo Jan 05 20:02:11 DocScrutinizer: you mean we should launch alsaloop? Jan 05 20:02:36 if there's something called alsaloop that sounds like the exactly right tool for it Jan 05 20:02:44 there is Jan 05 20:03:05 you probably should have it on your desktop even Jan 05 20:03:09 I think I once seen a pcm_plugin loop Jan 05 20:03:16 as it's part of the alsa-tools Jan 05 20:03:25 alsa-utils that is Jan 05 20:03:39 mickeyl, ^^^ Jan 05 20:03:51 so I should abandon the work on forwarding and fix n900 instead? Jan 05 20:03:59 jr@halley:~> alsa Jan 05 20:04:01 alsadebug.sh alsamixer Jan 05 20:04:16 Name : alsa-utils Jan 05 20:04:16 Version : 1.0.24.2-3 Jan 05 20:04:37 using old software? ;) Jan 05 20:04:48 rather: Jan 05 20:04:51 jr@halley:~> whichgrep "*alsa*" Jan 05 20:04:53 /home/jr/bin/alsadebug.sh Jan 05 20:04:55 /usr/bin/alsamixer Jan 05 20:05:34 mrmoku: DocScrutinizer51: can you help me please ? I am setting up a new build server and first build gives following error: http://norman-schleicher.de/jenkins/job/shr-core-nokia900-shr-image/34/console Jan 05 20:05:38 yeah, this is not the *most recent* distro here Jan 05 20:06:02 btw still I'd like to finish learning alsa so I wondered about buffer size and why didn't writei act as blocking Jan 05 20:06:11 nschle85: I got No F*ng Clue abiut such things ;-D Jan 05 20:06:25 nschle85: WARNING: you have different bblayers.conf, please sync it from common directory or call update-shr-core-conffiles to replace all config files with new versions Jan 05 20:06:34 nschle85, ask someone that knows oe like mrmoku JaMa or me Jan 05 20:07:01 GNUtoo: there are, like, 3 different methods to move data to/from ALSA Jan 05 20:07:07 DocScrutinizer: alsaloop allows create a PCM loopback between a PCM capture device and a PCM playback device Jan 05 20:07:15 blocking, nonblocking, and callback Jan 05 20:07:19 readi/writei. mmap and callback Jan 05 20:07:33 callback is not a good idea Jan 05 20:08:02 I heard that it messed up signals handlers Jan 05 20:08:05 mrmoku: internet dropout, i'm gonna try downloading and reflashing again , we'll see keep you up to date Jan 05 20:08:06 and similar things Jan 05 20:08:22 what's non-blocking? Jan 05 20:08:30 I tought that writei/readi were blocking Jan 05 20:08:45 maybe writen readn Jan 05 20:08:47 let me look Jan 05 20:10:38 GNUtoo: IIRC you could - lats say write - some data to ALSA and wait until it got moved to ALSA's internal (hw) buffer, or you try move it to ALSA and immediately the call returns with -EAGAIN when there's not enough free space in hw buffer to move your data there Jan 05 20:10:49 let's* say... Jan 05 20:10:55 ok Jan 05 20:11:22 I'm not all sure about it Jan 05 20:11:40 but that's what it *should* provide Jan 05 20:12:16 what's the functions to read/write non-blocking Jan 05 20:12:24 readn is for non interleaved Jan 05 20:12:51 ah it's something selectable Jan 05 20:12:56 the funcitons are the same Jan 05 20:13:01 let's see where it's done Jan 05 20:13:23 GNUtoo: man 2 write: >> Jan 05 20:13:24 ah flag for open mode: Jan 05 20:13:25 EAGAIN The file descriptor fd refers to a file other than a socket and has been marked non-blocking (O_NONBLOCK), Jan 05 20:13:26 SND_PCM_NONBLOCK Jan 05 20:13:26 and the write would block. Jan 05 20:13:27 << Jan 05 20:13:52 yes but my question was how to activate blocking/nonblocking not how to use it Jan 05 20:14:12 obviously by setting O_NONBLOCK Jan 05 20:14:18 or not setting it Jan 05 20:14:20 ok Jan 05 20:15:01 mrmoku: GNUtoo: DocScrutinizer51: bblayers.conf is now identical but error still exists: http://norman-schleicher.de/jenkins/job/shr-core-nokia900-shr-image/35/console Jan 05 20:15:14 nschle85: still NFC Jan 05 20:15:19 lol Jan 05 20:15:33 s/ DocScrutinizer /mrmoku/ Jan 05 20:16:41 DocScrutinizer: ok, ill never ask again you for Fu**g problem :-) Jan 05 20:16:53 hehe Jan 05 20:17:40 really, don't ask me about bitbake etc, I have NO idea about all that Jan 05 20:18:39 nschle85, he's a low level person(electronic engineer + knows well low level computing) and has no clue about the openembedded build system Jan 05 20:19:24 nschle85, http://norman-schleicher.de/jenkins/job/shr-core-nokia900-shr-image/35/console -> try to cleansstate util-linux-native Jan 05 20:19:39 oops Jan 05 20:19:51 I mean do that: Jan 05 20:19:51 oops ? Jan 05 20:19:58 bitbake -c cleansstate -b virtual:native:/var/lib/jenkins/jobs/shr-core/workspace/shr-core/openembedded-core/meta/recipes-core/util-linux/util-linux_2.20.1.bb Jan 05 20:20:17 then add ncurses-native to the dependencies of the recipe Jan 05 20:20:19 in DEPENDS Jan 05 20:20:28 and restart to build Jan 05 20:20:39 I guess the other failures have the same issues Jan 05 20:21:52 and if it works what is shown ? missing dependenciy in recipe ? Jan 05 20:22:03 yes Jan 05 20:22:12 if it works submit a patch Jan 05 20:22:49 GNUtoo: for FIFO: it's basically irrelevant how *large* the FIFO is, as long as it's not too small, so don't worry too much about all that, and start with a size 5 times that of what you think it should be, and probably you're fine for a POC Jan 05 20:23:34 no flags were set Jan 05 20:23:47 and now I'm already afk again, getting some dinner etc Jan 05 20:23:53 when it was too small indeed I got some problems Jan 05 20:23:56 like no sound Jan 05 20:23:59 or interupted sound Jan 05 20:24:11 like 1/4 second sound Jan 05 20:24:15 wait more than one second Jan 05 20:24:21 and again 1/4 second sound Jan 05 20:24:42 thanks a lot Jan 05 20:25:45 GNUtoo: may i run bitbake shr-image again to reproduce the problem after bitbake -c cleansstate .... Jan 05 20:25:47 ? Jan 05 20:28:24 nschle85, no, try to fix it Jan 05 20:28:52 try to add virtual-native as a dependency of the recipe for its native part Jan 05 20:29:02 native can be achieved by 2 ways: Jan 05 20:29:20 * having a file named foo-native_version.bb Jan 05 20:29:27 with foo beeing the recipe name Jan 05 20:30:01 * having BBCLASS_EXTENDS += "native" or something like that in the recipe and different overrides for native/non-native Jan 05 20:30:09 an override is something like: Jan 05 20:30:23 SRC_URI_nokia900 Jan 05 20:30:27 it's the nokia900 Jan 05 20:32:11 GNUtoo: ups i did bitbake shr-image but problem still exists :-) Jan 05 20:32:22 nschle85, show your diff Jan 05 20:32:31 git diff where you changed stuff Jan 05 20:32:37 like in openembedded-core Jan 05 20:32:42 or in meta-openembedded Jan 05 20:32:55 if you didn't try to fix it yet.... Jan 05 20:32:58 then try it now Jan 05 20:33:19 no changed stuff i bb cleaned and bb shr-image again Jan 05 20:33:41 so that did not change the result Jan 05 20:33:55 now ill do what you said Jan 05 20:33:59 yes that's normal Jan 05 20:34:00 ok Jan 05 20:41:14 GNUtoo: here the native version does not build Jan 05 20:41:33 ah? do you have the log? Jan 05 20:41:46 yes Jan 05 20:41:56 so ncurses-native fails...hmmm Jan 05 20:42:50 please wait i send the link Jan 05 20:43:54 http://norman-schleicher.de/jenkins/job/shr-core-nokia900-shr-image/ws/shr-core/tmp-eglibc/work/x86_64-linux/readline-native/readline-native-6.2-r0/temp/log.do_install.8076 Jan 05 20:43:56 ok Jan 05 20:47:10 GNUtoo: is libhistory missing libtermcap ? Jan 05 20:47:26 no idea Jan 05 20:47:38 could you look? Jan 05 20:47:58 ill pastebin so we can talk about line numbers Jan 05 20:48:12 ok Jan 05 20:48:50 http://pastebin.com/Eq9p7TBk Jan 05 20:49:30 GNUtoo: see 20. Jan 05 20:49:50 and 21 Jan 05 20:49:56 yes seem so Jan 05 20:50:13 but it's readline Jan 05 20:50:21 so try to do that: Jan 05 20:50:26 find the termcap recipe name Jan 05 20:50:38 and add termcap-native dep to readline-native Jan 05 20:50:47 or maybe it is lacking on the buildhost? Jan 05 20:50:52 could also be that Jan 05 20:50:58 check if you have all required libs Jan 05 20:51:53 GNUtoo: i think installing the libs on host would solve but its not what we want i think Jan 05 20:52:53 so adding termcap native ad dependency is much better Jan 05 20:53:10 s/ad/as/g Jan 05 20:53:11 nschle85 meant: so asding termcap native as dependency is much better Jan 05 20:53:33 it depends Jan 05 20:53:38 +apt no this i did not mean Jan 05 20:53:55 if you have termcap in the required deps to apt-get before building then you should install it Jan 05 20:54:02 if not then you should use the native recipe Jan 05 20:54:28 GNUtoo: ill add it as dependency< Jan 05 20:54:30 y Jan 05 20:54:42 and test Jan 05 20:55:45 ok Jan 05 21:01:48 GNUtoo: i cannot find any termcap receipe ? do you find one ? Jan 05 21:03:20 let me look Jan 05 21:06:53 I can't find it either Jan 05 21:07:21 GNUtoo: so it has to be in the first checks Jan 05 21:08:31 GNUtoo: how a bug report has to be written ? Jan 05 21:09:59 I think you should write to oe or oe-core ml Jan 05 21:10:09 else you could create such recipe Jan 05 21:10:17 but before check that it's not in the requirement list Jan 05 21:10:18 in oe.dev there was a termcap recipe Jan 05 21:10:23 like thie thing you apt-get Jan 05 21:10:28 Heinervdm, thanks a lot!!!! Jan 05 21:12:33 Heinervdm: hello :-) Jan 05 21:12:43 nschle85: hi :) Jan 05 21:13:13 Heinervdm: so we should import it and test again .-) Jan 05 21:13:47 Heinervdm: i forgot: happy new year :-) Jan 05 21:14:35 nschle85: thank you, happy new year to you too :) Jan 05 21:14:48 importing the reciepe is easy, just copy it Jan 05 21:15:01 but i can't understand that it is not in oe-core... Jan 05 21:15:29 there is only a vte-termcap Jan 05 21:15:37 i know.. pushing a patch is more complicated :-) Jan 05 21:15:49 what does it mean ? Jan 05 21:16:27 no idea Jan 05 21:16:42 what does vte mean Jan 05 21:16:54 mrmoku`: i kept trying to do the internet-dropout-upgrade and it finished it , after lets say 10 times, and i cpumd reboot ! ! ! Jan 05 21:17:02 ? Jan 05 21:17:06 cpumd = could Jan 05 21:18:05 nschle85, happy new year Jan 05 21:18:18 nschle85, it's a terminal thing Jan 05 21:18:22 maybe a lib Jan 05 21:18:23 nschle85: i think it means virtual-terminal but i'm not sure Jan 05 21:18:25 or something like that Jan 05 21:18:35 perhaps it provides the same Jan 05 21:25:50 i can't compile dbus... Jan 05 21:26:11 | pseudo: I can't seem to make LD_PRELOAD go away. Sorry. Jan 05 21:26:13 | pseudo: LD_PRELOAD: libpseudo.so Jan 05 21:27:05 ups i copiled today from scratch Jan 05 21:27:41 so you have a new problem Jan 05 21:28:33 i'm having this problem since a week... Jan 05 21:28:48 no idea what i can do Jan 05 21:29:04 which receipe is it ? Jan 05 21:29:11 dbus Jan 05 21:29:19 openembedded-core/meta/recipes-core/dbus/dbus_1.4.16.bb Jan 05 21:29:43 cou can scan jenkins logs Jan 05 21:31:44 HeinervdmOff: NOTE: Running task 615 Jan 05 21:32:06 http://norman-schleicher.de/jenkins/job/shr-core-nokia900-shr-image/33/consoleFull Jan 05 21:32:14 dbus is task 1252 of shr-lite-image Jan 05 21:33:02 M00R1Z: and it works now? Jan 05 21:33:20 please find it in logfile Jan 05 21:33:55 so do you have logs of broken dbus builds ? Jan 05 21:35:05 mrmoku: i booted , and got an error (which i couldnt read) rebooting now, but at least i get to the desktop Jan 05 21:35:33 Heinervdm, problems with pseudo it seems Jan 05 21:36:02 nschle85: your build did not arrive at do_install of dbus Jan 05 21:36:31 GNUtoo: yes, but i don't know what i can do about this Jan 05 21:37:21 maybe ask in #yocto Jan 05 21:38:18 mrmoku: booted: no icons on desktop (just al before)rightclick refresh-> error msg: "..en directory '/home/root/.e/e/appshadow' :No.." i can't read begin and end Jan 05 21:38:21 ok, but i will do this tomrow Jan 05 21:38:28 mrmoku: retry or abort? Jan 05 21:38:32 will go to sleep now Jan 05 21:38:36 gn8 Jan 05 21:38:58 Heinervdm: hmm ok good night Jan 05 21:39:36 good night Jan 05 21:40:57 M00R1Z: no icons is a known bug :-) Jan 05 21:41:08 there is the mode toggle icon in the toolbar Jan 05 21:41:18 toggling the mode will bring the icons back Jan 05 21:41:24 it's the square one Jan 05 21:41:30 mrmoku: GREAt that means i'm doing right ! ! !, i'll try that right away (though i used refresh before) Jan 05 21:41:41 (i know thats the one i removed hahah) Jan 05 21:42:11 good evening Jan 05 21:42:31 ege: good evening :-) Jan 05 21:42:37 can someone tel me, where the information for sharing GPRS connecton is stored? Jan 05 21:42:53 M00R1Z: alternative... the L ... and then right click on the desktop and refresh :) Jan 05 21:42:54 it uses the wrong settings for iptables Jan 05 21:43:41 mrmoku: thats what i did, thats when i got the errorr msg, i did retry which doesnt do anything so i did abord and now i have some icons(all question marks though) Jan 05 21:44:02 hmm... that's the error GNUtoo has Jan 05 21:44:05 strange Jan 05 21:44:11 ooops Enlightenment error : SIGABRT'd Jan 05 21:44:39 F1 recover or F2 Exit Jan 05 21:44:49 i press F1 Jan 05 21:44:58 back at first desktop error Jan 05 21:45:36 ege: i am not an expert but did you added a shr ticket ? Jan 05 21:45:38 ege: I think that's hardcoded... what's wrong? Jan 05 21:45:41 abort and now i'm back at desktop with some icons icons: applications background config dicts . . . Jan 05 21:46:18 M00R1Z: if you have ssh then /etc/init.d/xserver-nodm stop; rm -rf ~/.e; /etc/init.d/xserver-nodm start Jan 05 21:46:22 might be worth a try Jan 05 21:46:48 sec mrmoku Jan 05 21:46:54 mrmoku, I've seen now, that it uses dhcp.... Jan 05 21:47:30 which has nothing to do with iptables though Jan 05 21:47:38 I'm used to use fixed IP's with different range (because of conficts) and wondered why it doesn't work Jan 05 21:48:09 mrmoku: done! , now i'm at the first boot menu (like when you just flashed) set language etc Jan 05 21:48:20 if I switch to share GPRS connection, it FR has to add forwarding rules. Jan 05 21:48:37 M00R1Z: just exit it Jan 05 21:49:01 It does this only for net 192.168.0.0/24 Jan 05 21:49:32 hmm Jan 05 21:49:40 yeah, that is indeed wrong Jan 05 21:50:07 mrmoku: how? i pressed next at language Jan 05 21:50:20 hmm, good question. I can change to use dhcp when I want to share GPRS. Jan 05 21:50:21 M00R1Z: ahh, sorry, that one you have to do Jan 05 21:50:28 I just did not realize that before... Jan 05 21:50:30 M00R1Z: the SHR wizard you can just exit Jan 05 21:50:41 mrmoku: done! shr wizzard i'll exit DONE! Jan 05 21:50:45 ege: yeah, but 192.168.0 is wrong Jan 05 21:51:06 my screen is locked and i cant get passed Jan 05 21:51:22 dhcp uses same range Jan 05 21:51:40 yeah, but the usb if not Jan 05 21:52:01 ahh, good point, that has been changed... Jan 05 21:52:20 damn, hardcoding that is plain wrong :/ Jan 05 21:53:15 is it possible to just collect the actual used ip and then generating the rules and dhcp with that information? Jan 05 21:53:26 yeah, something like that Jan 05 21:56:17 mrmoku, there are some more strange settings in dhcp.conf.... Jan 05 21:56:25 opt router 202.50.168.192 # address of interface Jan 05 21:56:27 mrmoku: stuck in sleep can't wake up i'll reboot? msg on black screen -> (1091.32500) soc-audio: resume work item may be lost Jan 05 21:56:42 mrmoku: square brackets Jan 05 21:56:56 I thing FR should be router and the given IP ist not "alive" atm Jan 05 21:57:01 mrmoku: ssh still works though Jan 05 21:58:53 ege: the ip for the outgoing connection via GPRS is set correctly Jan 05 21:59:08 ege: just dhcp and iptables forwarding from the local net is wrong Jan 05 21:59:10 (and hardcoded) Jan 05 21:59:53 opt router %s # address of interface Jan 05 22:01:07 mrmoku, that's strange. the 202.50.268.192 ip belongs to TREADWELL-NZ Jan 05 22:01:21 mrmoku: i just did " shutdown now" in ssh , and my last flash was my desktop WITH ALL THE RIGHT ICONS !!! rebooting now Jan 05 22:01:27 but I'm connected to sunrise.ch Jan 05 22:01:50 and I got a router with private network (10.64......) for ppp0 Jan 05 22:02:10 hmm Jan 05 22:02:21 so I don't know where this ip comes from Jan 05 22:03:59 if you're connected via GPRS you should have a public ip Jan 05 22:04:27 mrmoku: not at all :( Jan 05 22:04:47 mrmoku: reboot and everything seems fine ill check it and thanks and i'll be back with good news hopefully Jan 05 22:05:02 PaulFertser: no? Jan 05 22:05:03 heh Jan 05 22:05:11 mrmoku: here those suckers provide proper address only to organisations only when specifically asked (and payed) for. Jan 05 22:05:13 M00R1Z: ok, good luck Jan 05 22:05:36 mrmoku, no, sadly not any more Jan 05 22:05:37 PaulFertser: HI! Jan 05 22:05:40 PaulFertser: gah, that is mad Jan 05 22:05:56 they "save" IP Addresses and make NAT for there customers Jan 05 22:05:58 DocScrutinizer51: hey hey :) Jan 05 22:06:07 sunrise does and some others too Jan 05 22:06:16 mrmoku: those fuckers also refuse to provide IPv6 either Jan 05 22:06:43 PaulFertser: and there they don't have the excuse of missing addresses ;) Jan 05 22:07:39 some providers think of using ipv6 in mobile and nat them to ipv4 internet.... Jan 05 22:08:09 just to have enough IPs for mobile connections Jan 05 22:11:10 my provider just stopped offering public ipv4 addresses and uses private now Jan 05 22:11:33 should gta02 support ipv6 with gprs? Jan 05 22:12:20 lindi-: i think that doesn't depend on the modem. Jan 05 22:12:40 PaulFertser: I would hope so but there' still a chance that it does Jan 05 22:12:46 PaulFertser: how's life in Moskow these days? Jan 05 22:14:28 DocScrutinizer51: those street meetings were "groundbreaking" and i'm glad finally something started to change. Other than that, strange unusually warm weather is confusing for me. Jan 05 22:15:24 lindi-: those USA operators are deploying v6 for LTE devices. I have no idea about the software stack there. Jan 05 22:17:21 DocScrutinizer51: and how is it in DE? Jan 05 22:17:47 rarely noticing anything but my daily work Jan 05 22:18:06 nabend Jan 05 22:18:24 yo mickeyl Jan 05 22:18:38 absolutely the correct moment :) Jan 05 22:18:44 uh oh :) Jan 05 22:18:46 * mickeyl runs Jan 05 22:18:51 :P Jan 05 22:19:15 connection sharing in fsonetworkd is broken Jan 05 22:19:18 hah Jan 05 22:19:20 because the ip address changed Jan 05 22:19:57 lindi-: i think pppd should work with ipv6 just fine. Jan 05 22:20:39 mrmoku: i see, we should figure out the proper netmask from the interface Jan 05 22:20:48 instead of hardcoding to 192.168.0.0 Jan 05 22:20:51 yup Jan 05 22:21:22 (both necessary for the iptables commands and for the udhcp conf) Jan 05 22:21:36 question is if the proper interface is always usb though Jan 05 22:21:56 IIRC wifi tethering is one of you favourites :) Jan 05 22:22:00 +r Jan 05 22:22:27 that's true, i just have seen very little devices that actuall do master mode these days Jan 05 22:22:37 s/little/few/ Jan 05 22:22:37 mickeyl meant: that's true, i just have seen very few devices that actuall do master mode these days Jan 05 22:22:50 PaulFertser: but ppp does not go over the air? Jan 05 22:22:53 yeah Jan 05 22:23:32 mrmoku: actually the interface to use is already a parameter in the dbus call or is it? Jan 05 22:23:57 oh and there is already var ip = FsoFramework.Net.ipv4AddressForInterface( iface ); Jan 05 22:23:58 Who uses GTA as wireless AP should know how to fix that :) Jan 05 22:24:08 should be simple to get the network mask as well Jan 05 22:24:11 mickeyl: yeah, but that is the gprs one, no? Jan 05 22:24:20 ah right, we need another one, true Jan 05 22:24:28 so we need to extend the signature Jan 05 22:24:30 lindi-: yes, but i assumed that somehow the over-the-air layer passes any ppp frames. Jan 05 22:24:44 btw. there is another error in the dhcp template too Jan 05 22:24:53 because it sets the listen interface to the gprs one Jan 05 22:24:58 so what about config file which hast option "guess" and hardcoded ip Jan 05 22:25:40 Using wireless in ibss to tether is better than nothing though. Jan 05 22:25:52 mickeyl: hmm Jan 05 22:25:54 Description: Start sharing the internet connection of this device with any clients connected to the specified network interface via DHCP. Jan 05 22:26:08 mrmoku: so it's the right interface Jan 05 22:26:10 that tells me iface should indeed be the local one Jan 05 22:26:41 ah, i remember Jan 05 22:26:49 it's the one that you were looking for Jan 05 22:26:52 the one for remote clients Jan 05 22:27:05 since the local one is irrelevant, as we just use the default route implicitly Jan 05 22:27:45 ahh Jan 05 22:27:54 right and the ppp0 is not involved Jan 05 22:27:58 so that means we already have almost all the information we need Jan 05 22:27:59 ya Jan 05 22:28:04 indeed :) Jan 05 22:28:09 we just need to gather the netmask and substitute it Jan 05 22:28:16 mrmoku: iliwi goes to idle without finding networks (just like before) and ffalarms doesnt start up tangogps ok but it seems that my gps doesnt pickup any satelites (ill test this later since my qtmoko doesnt find any either) Jan 05 22:28:22 or rather substitute the config entries and the commands Jan 05 22:28:39 M00R1Z: ok Jan 05 22:28:59 mickeyl: yeah Jan 05 22:29:00 imho it would also be beneficial to have some easy means to bridge "wan" and "lan" interfaces together. In case the device connects to something that can distribute IPs to a subnet. Jan 05 22:29:29 mrmoku: killall connmand iliwi fixes that Jan 05 22:30:32 I'm leaving now... Do you need a bug report for that GPRS-sharing thing? Jan 05 22:30:33 mrmoku: i can try to work on that Jan 05 22:30:45 E.g. my phone connects to the internet and then creates a L2 OpenVPN tunnel to my VPS. Now both my phone and all devices bridged to this OpenVpn connection can do SLAAC and access the Internet. Jan 05 22:30:54 mickeyl: cool Jan 05 22:31:04 ege: I don't think so... the problem is quite clear :) Jan 05 22:31:18 mrmoku, cool :) Jan 05 22:31:22 mickeyl, hi Jan 05 22:31:30 PaulFertser: that is step 2 :-P Jan 05 22:31:33 gn8 Jan 05 22:31:37 hi GNUtoo Jan 05 22:31:42 PaulFertser: might be an interesting new fsonetworkd plugin :) Jan 05 22:31:47 ege: gnight and thx Jan 05 22:32:05 mickeyl, what should we do for forwarding? DocScrutinizer thinks we should use alsaloop instead of coding something ourselves Jan 05 22:32:33 nschle85: pong Jan 05 22:33:01 GNUtoo: i don't fancy running an additional binary for that stuff... what speaks against using the code? Jan 05 22:33:14 and saving the additional process Jan 05 22:33:24 ? Jan 05 22:33:28 I don't fancy? Jan 05 22:33:39 GNUtoo: he prefers not to Jan 05 22:33:41 * mrmoku too Jan 05 22:33:42 ok Jan 05 22:33:58 my code is not ready yet but can work Jan 05 22:34:10 it'll take quite some time in finishing it Jan 05 22:34:13 and polishing it Jan 05 22:34:32 and it may not be as good as the alsaloop code Jan 05 22:34:52 another option would be just to use the alsaloop code, no? Jan 05 22:34:58 and integrate it into the plugin Jan 05 22:35:02 what license is it? Jan 05 22:35:03 yes Jan 05 22:35:10 it's part of the alsa-utils Jan 05 22:35:34 GNUtoo: maybe try if alsaloop gives good results first? Jan 05 22:35:41 good idea Jan 05 22:35:45 I'll try later Jan 05 22:41:08 *shrug* Jan 05 22:41:10 * rah -> zzz Jan 05 22:45:10 mrmoku: pidgin ok navit ok settings ok sync time ok midori not ok eve not ok ffphonelog ok claws ok contacts ok enjoy not ok dialer ok Jan 05 22:47:21 gnight Jan 05 22:47:39 you all lost me Jan 05 22:47:50 * pespin__ getting closer to finish eflvala upgrade :D Jan 05 22:49:08 rah: pingpong Jan 05 22:49:13 :-) Jan 05 22:49:31 whatever you do for forwarding audio data from one PCM interface to another (modem) PCM interface, you'll want a dedicated process for that (unless you implement it directly in kernel) - and definitely you will want to do some special scheduling for that particular process Jan 05 22:50:13 rah: ups ok ill try ping you tomorrow again :-) Jan 05 22:50:18 bye Jan 05 22:51:32 and I can't see why anybody would want to implement that tiny process in vala, or run it any other way than as a separate autonomous process Jan 05 22:58:14 M00R1Z: not soo bad :) Jan 05 22:59:20 * mrmoku bed Jan 05 22:59:21 gnight all Jan 05 23:01:53 mrmoku: thanks ! ! i really appreciate your helps ! gonna get some sleep now thanks and later ! Jan 06 01:24:05 good night all **** ENDING LOGGING AT Fri Jan 06 02:59:57 2012