**** BEGIN LOGGING AT Wed Aug 12 02:59:57 2009 Aug 12 03:24:37 Is shr-unstable buildable at the moment? - my build has been dying on libfsoframeworkd Aug 12 03:24:40 "Too few arguments, method `FsoFramework.Logger.createFromKeyFile' does not take 2 arguments" **** ENDING LOGGING AT Wed Aug 12 05:26:18 2009 **** BEGIN LOGGING AT Wed Aug 12 05:28:14 2009 **** ENDING LOGGING AT Wed Aug 12 05:32:06 2009 **** BEGIN LOGGING AT Wed Aug 12 05:34:16 2009 Aug 12 06:25:48 good morning Aug 12 06:35:35 morning methril Aug 12 06:35:41 but gotta work, bye :) Aug 12 06:39:35 i'm at work Aug 12 06:39:39 :) Aug 12 06:39:41 cu Aug 12 07:16:08 ok, new image is much better :-) Aug 12 07:22:41 DocScrutinizer: you're sarcasm detector needs tuning ;) Aug 12 07:23:36 for gta01 too Aug 12 07:23:37 ? Aug 12 07:23:42 i see some bugs Aug 12 07:40:59 methril|work, still building (gta01) Aug 12 07:41:59 mrmoku: when i've a working image in OE dev i send the patches to the ml Aug 12 07:44:30 methril|work, good Aug 12 07:58:22 morning Aug 12 07:59:10 morning mickeyl Aug 12 08:04:59 mickeyl, hey :-D Aug 12 08:06:06 how's mergin to oe.dev going? Aug 12 08:06:34 mickeyl, methril|work is working on that front Aug 12 08:06:41 ah, good Aug 12 08:06:55 I was struggling with building Aug 12 08:08:07 mickeyl, and would have some question :P Aug 12 08:08:29 mickeyl, SRCPV does not work when a package is *not* on AUTOREV? Aug 12 08:09:19 i don't know much about SRCPV, but afaik SRCPV adds the number of revisions into the PV all the time, that's how it differs from SRCREV Aug 12 08:09:49 mickeyl, it does not do it if you set a fixed rev for a package Aug 12 08:09:58 then it's broken Aug 12 08:10:01 it used to do that Aug 12 08:10:14 yeah, thought that too... Aug 12 08:10:19 the other question Aug 12 08:10:31 task-fso-compliance or task-fso2-compliance Aug 12 08:10:36 both do not fit ;) Aug 12 08:10:41 in how farß Aug 12 08:10:42 ? Aug 12 08:10:57 fso2 has stuff in it that does not (yet) build Aug 12 08:11:06 and fso is missing new stuff (like fso-abyss) Aug 12 08:11:17 hmm Aug 12 08:11:22 task-fso2-compliance builds in .dev Aug 12 08:11:45 fsodeviced? fsonetworkd? Aug 12 08:12:03 yeah Aug 12 08:12:13 fso-abyss should not be in the task Aug 12 08:12:15 okay, will retry then and check for differences Aug 12 08:12:26 but rather RRECOMMENDED or RDEPENDED by frameworkd Aug 12 08:12:32 it is that Aug 12 08:12:39 but it does not get included in the image then Aug 12 08:12:40 hmm Aug 12 08:13:31 RDEPENDS should include it Aug 12 08:13:34 for sure Aug 12 08:14:24 mickeyl, RDEPENDS_${PN}_append_om-gta02 = " fso-abyss" Aug 12 08:14:31 is what we have in frameworkd-config-shr Aug 12 08:14:44 (which leads me to another question...) Aug 12 08:14:59 does it make any sense we have that package in the first place? Aug 12 08:15:46 that depends Aug 12 08:16:16 a distro might want to have a special configuration Aug 12 08:16:22 like, moving logs to a certain place or so Aug 12 08:16:28 or disabling/enabling certain services Aug 12 08:16:51 plus, the config is ARCH = ALL Aug 12 08:16:58 mickeyl, no Aug 12 08:16:59 so it makes sense to seperate it from the main package Aug 12 08:17:05 arr Aug 12 08:17:08 vice versa of course Aug 12 08:17:27 mickeyl, PACKAGE_ARCH_${PN}-config = "${MACHINE_ARCH}" Aug 12 08:17:29 ahh, ok :) Aug 12 08:18:21 i've read some fuzz about fsousaged bugs in the backlog, can someone summarize the problem? Aug 12 08:18:37 libfsobasics changed logger semanthics Aug 12 08:18:46 and libfsoframework needs adjusting Aug 12 08:18:58 fsousaged not working is just a consequence of that Aug 12 08:19:09 Hey there Aug 12 08:19:41 Ainulindale, new gta01 image to test for you Aug 12 08:20:15 Yeah but sadly I won't be able until tonight Aug 12 08:20:18 I forgot it this morning Aug 12 08:20:21 (although I took my lunch) Aug 12 08:20:23 (see my priorities) Aug 12 08:20:33 ok, i have 2 hours now, will look into Aug 12 08:20:50 mickeyl: how's life by the way? Aug 12 08:21:27 Did I tell you that Sabine is very nice (both -looking and as in -funny)? Aug 12 08:21:29 Ainulindale: heya! pretty good, i managed to spend some more time away from the computer :)) Aug 12 08:21:54 Ainulindale: hehe, not yet. thanks, i'm glad she exists, yeah Aug 12 08:22:03 Good :-) Aug 12 08:22:05 mickeyl, thanks :D Aug 12 08:22:19 mickeyl: and among other thing Aug 12 08:22:31 khorben, Louis Montagne & al are at HAR today Aug 12 08:22:37 IIRC Harald & Daniel will be there too Aug 12 08:22:44 If you have things to ask them to talk about :-) Aug 12 08:22:50 (I specifically asked them to talk about HTC) Aug 12 08:22:59 And torture Daniel until he speaks Aug 12 08:23:14 right. Daniel will show progress on the G1, i think Aug 12 08:23:49 we already surpassed the Touch Pro since G1 manages phone audio Aug 12 08:24:01 (the baseband processor, that is) Aug 12 08:26:02 mickeyl: kiss me already :-) Aug 12 08:26:36 d'oh :) Aug 12 08:28:19 actually, i think all of the guys are there but me Aug 12 08:30:29 Heinervdm, ping - you saw my mail about elmdentica? Aug 12 08:31:10 mrmoku: yes Aug 12 08:31:34 i looked for every ticked at the feed but for this :D Aug 12 08:46:44 mickeyl, ahh, and then there are rumours that fsousaged brought back the 'does not suspend after resume without screen tap' bug Aug 12 08:49:14 oh Aug 12 08:49:25 ok, i need to check whether we set busy after resume as we did in fso1 Aug 12 08:50:36 :) Aug 12 08:52:27 freesmartphone.org: 03mickey 07cornucopia * rb241178fa679 10/libfsobasics/ (3 files in 2 dirs): libfsobasics: move prefixForExecutable from libfsoframework to here and add test for it Aug 12 08:53:56 mickeyl: did you do a rerun of generate-manifest-2.6.py, because i patched only the script :) Aug 12 08:56:17 freesmartphone.org: 03mickey 07cornucopia * r216dc62912ef 10/libfsobasics/fsobasics/utilities.vala: libfsobasics: utilities: cache program name result Aug 12 08:56:25 Heinervdm: sure Aug 12 08:56:47 recipes/python/python-2.6-manifest.inc | 99 ++++++++++++++++--------------- Aug 12 09:00:14 freesmartphone.org: 03mickey 07cornucopia * ra00d9675034c 10/libfsoframework/ (5 files in 2 dirs): libfsoframework: catch up with new createLogger syntax Aug 12 09:02:34 mrmoku: is this the one simple patch your users needed so badly for ~4 days? ;) http://git.freesmartphone.org/?p=cornucopia.git;a=commitdiff;h=a00d9675034cddc0575b31964e30d10e016e859d Aug 12 09:03:22 PaulFertser, yep Aug 12 09:03:29 mrmoku: then you have to pick the new version of recipes/python/python-2.6-manifest.inc for shr/import to close #549 ;) Aug 12 09:03:51 * mrmoku looks what is #549 Aug 12 09:04:25 mrmoku: python package requests Aug 12 09:04:37 hmm.. ok Aug 12 09:06:37 wait a bit longer Aug 12 09:06:47 there's some broken merges Aug 12 09:06:55 need to add more to python Aug 12 09:24:22 mickeyl: it forgets to put in oe.dev girepository.bbcalss Aug 12 09:24:29 s/it/you Aug 12 09:24:41 i don't like that being seperate Aug 12 09:24:52 for now i added what it does to vala.bbclass Aug 12 09:25:03 it can stay there until we see non-vala GIR stuff Aug 12 09:25:14 ok, then i remove it from my local repo and i redo libmodulo & other class Aug 12 09:25:47 thanks Aug 12 09:25:57 you're wellcom Aug 12 09:26:02 wellcome Aug 12 09:27:00 mrmoku: you were right btw., task-fso2-compliance didn't build at all until now. not sure which branch i thought i'd build *shrug* Aug 12 09:31:21 freesmartphone.org: 03mickey 07cornucopia * recb54c57f8f5 10/libfsotransport/fsotransport/transport.vala: libfsotransport: catch up with new logger API Aug 12 09:33:41 hi mickeyl Aug 12 09:34:10 I'm trying to get FSO working on h:1, and I may have some questions for you Aug 12 09:34:15 hi Deubeuliou, thanks for the bugreport. the problem is that for Vala projects, dist indeed depends on all Aug 12 09:34:40 ah, yeah. Aug 12 09:34:45 not necessarily all though Aug 12 09:34:50 the #vala guys told me there was a solution Aug 12 09:34:52 but at least on the target that generates the .c and .h Aug 12 09:34:58 what did they say? Aug 12 09:35:27 is was not very clear to me, but they say the makefiles should use $topdir and stuff like thta Aug 12 09:35:44 sorry, it's vague Aug 12 09:35:48 hmm Aug 12 09:35:54 let me try 'make dist' on vala itself Aug 12 09:36:50 btw, I managed to get ogsmd running on h:1 several times, but now it fails: ogsmd gets no response from fso-abyss (timeout) Aug 12 09:37:07 as far as I can see, mdbus sees the buses and fso-abyss starts Aug 12 09:37:47 http://pastebin.com/d6ab04835 Aug 12 09:38:14 sounds like the system activation does not work for you Aug 12 09:38:15 (I'm using SHR's /etc/frameworkd.conf) Aug 12 09:38:41 do you have all the necessary dbus conf files installed? Aug 12 09:38:56 I think so. but I may have too many Aug 12 09:38:58 for system activation you need one in /usr/share/dbus-1/system-services/ iirc Aug 12 09:39:07 I have org.freesmartphone.omuxerd and .ogsmd Aug 12 09:39:15 both starting fso-abyss Aug 12 09:39:48 omuxerd is ok Aug 12 09:40:06 I remove ogsmd, and it's the same Aug 12 09:40:16 (ogsmd's .service file, I mean) Aug 12 09:41:32 does it work, when you manually start fso-abyss first? Aug 12 09:42:33 no Aug 12 09:43:33 mickeyl: one question about one recipes/python --> python-phoneutils, are you planning to add it to OE.dev? Aug 12 09:43:42 maybe I could try wit gsm0710muxerd Aug 12 09:44:26 freesmartphone.org: 03mickey 07cornucopia * rd751bb0cf4eb 10/ (5 files in 5 dirs): all: catch up with new logging API Aug 12 09:44:42 nah, don't mess with the old thing Aug 12 09:45:00 ok :p Aug 12 09:45:09 if fso-abyss doesn't work on manual start, that's the first thing you need to check Aug 12 09:45:10 TAsn: http://trac.shr-project.org/trac/ticket/555 this is done, isn't it? Aug 12 09:45:14 how does it fail? Aug 12 09:45:19 try using cli-framework to talk with it Aug 12 09:45:52 Heinervdm, thinks so, yeah Aug 12 09:45:55 -s Aug 12 09:46:18 methril|work: absolutely, yeah Aug 12 09:46:51 commited Aug 12 09:47:20 I get ERROR:dbus.proxies:Introspect error on org.freesmartphone.DBus:/org/freesmartphone/DBus: dbus.exceptions.DBusException: org.freedesktop.DBus.Error.ServiceUnknown: The name org.freesmartphone.DBus was not provided by any .service files Aug 12 09:47:25 when I start cli-framework Aug 12 09:47:41 you are missing dbus-hlid, but that should not hurt too much Aug 12 09:48:02 you should be able to call stuff on the 'muxer' object from cli-framework Aug 12 09:48:07 simplest call would be Aug 12 09:48:21 >>> muxer.GetVersion() Aug 12 09:48:21 '0.2.1' Aug 12 09:49:29 ok, I think I found the problem Aug 12 09:49:41 I had that one before but not with the same symptoms Aug 12 09:49:45 DBusException: org.freedesktop.DBus.Error.Spawn.ExecFailed: Cannot launch daemon, file not found or permissions invalid Aug 12 09:49:50 Aha Aug 12 09:49:58 so system activation does not work Aug 12 09:49:59 that's a /usr/lib/dbus-1/dbus-... permission problem Aug 12 09:50:18 I know how to fix it but it seems alright, this time, though ... Aug 12 09:50:25 system activation is a pain to setup Aug 12 09:50:45 i know my way on OE, but debian way might be different Aug 12 09:51:22 I'll find Aug 12 09:51:24 thanks Aug 12 09:51:27 np Aug 12 09:51:34 mrmoku: NOTE: Running task 2554 of 2554 (ID: 0, /local/pkg/oe/openembedded/recipes/tasks/task-fso2-compliance.bb, do_build) Aug 12 09:51:35 *phew* Aug 12 09:51:53 i'll be doing more changes to the logger API, but will remember to fix all users next time Aug 12 09:52:01 sorry for the trouble Aug 12 09:52:49 mickeyl, np :) Aug 12 09:57:51 some brave shr-unstable user around? Aug 12 09:59:00 mrmoku: i'm on your testimage from yesterday Aug 12 10:00:11 Heinervdm, hmm, need someone to check upgrade on shr-unstable :) Aug 12 10:00:29 Heinervdm, you still have no ringtone, right? Aug 12 10:00:34 yes Aug 12 10:00:41 upgrade from 8.8? Aug 12 10:01:01 Heinervdm, try update on shr-testing then Aug 12 10:01:09 you have to adjust the feed config though Aug 12 10:01:14 ok Aug 12 10:01:25 to point to tests/mrmoku/testing/feed Aug 12 10:01:35 should be fixed then Aug 12 10:20:23 mickeyl: I may have found a cause to the problem Aug 12 10:20:28 fso-abyss segfaults Aug 12 10:20:40 I launched fso-abyss, then framerworkd Aug 12 10:20:46 and I saw fso-abyss segfaulting Aug 12 10:21:39 ouuuuch Aug 12 10:23:10 DocScrutinizer: morning Aug 12 10:23:14 mickeyl, hi, will you be free this october to give talk? :) Aug 12 10:23:15 * DocScrutinizer kicks C for its missing runtime boundry and reference checks Aug 12 10:23:20 *talks* Aug 12 10:23:27 PaulFertser: mooo Aug 12 10:23:58 mrmoku: my FR crashed on upgrade, so my installition is probably broken Aug 12 10:24:09 DocScrutinizer: little question: so should we somehow mention on the wiki that the user should control battery temperature himself in some reasonable way (not charge while the device exposed to the direct sunlight beams e.g.)? Aug 12 10:24:10 Heinervdm, ouch Aug 12 10:24:22 Heinervdm, any hint on why it crashed? Aug 12 10:25:02 PaulFertser: can't cause any damage to do so. I vote for it Aug 12 10:25:28 mrmoku: no last line over ssh was "Configuring shr-splash-theme-simple", then no reaction on touchscreen and network was down too Aug 12 10:25:43 DocScrutinizer: and i guess we should remove mis-information about an internal temperature protection too. Aug 12 10:26:17 PaulFertser: not as a warning, rather in form of a recommendation to have prolonged bat-live Aug 12 10:26:17 DocScrutinizer: it's puzzling why those smart chip don't have an option to monitor the temperature... Aug 12 10:26:33 yup Aug 12 10:28:26 DocScrutinizer: hm... I really wonder if it might explode or not being charged 0.7A warmed up to ~70C. Aug 12 10:28:38 PaulFertser: well I guess on ~120C the polyfuse trips at <100mA, but anyway. Correct is we don't have decent overtemp control. also a fact is bat never heats when treated nicely. so we just make sure we don't heat it from external source Aug 12 10:28:41 Sup3rkiddo: i'm not sure yet, since we are on holidays somewhere between september and october. where and when would that be? Aug 12 10:29:13 PaulFertser: no, never will explode at that treatment Aug 12 10:29:39 DocScrutinizer: any prooflink? ;) Aug 12 10:29:45 mickeyl, Chennai, IITM Aug 12 10:29:56 own experience ;-) Aug 12 10:30:04 mickeyl, they have a tech festival that is pretty famous in India.. Aug 12 10:30:32 mickeyl, I talked about FSO last year, it would be great if you could come down and talk about it Aug 12 10:31:03 Sup3rkiddo: do you have the exact dates yet? Aug 12 10:31:16 mrmoku: The XServer doesn't start anymore and SSH isn't working too, so i can't test Aug 12 10:31:50 mickeyl, September 30 - Oct 4 Aug 12 10:33:17 any alsa expert here? Aug 12 10:33:21 Heinervdm, double ouch Aug 12 10:33:22 mrmoku: please ping me as soon as "don't upgrade" warning becomes obsolete :-) Aug 12 10:33:30 Sup3rkiddo: bummer, i'm afraid that is completely covered by our vacations :( Aug 12 10:33:53 mickeyl, awwww.................... Aug 12 10:33:53 emmentaler: shoot Aug 12 10:33:53 iirc we're away two weeks from 21th of Sept Aug 12 10:34:15 emmentaler: not that I'm an expert ;-) Aug 12 10:34:26 i had loved to come over, but my wife would kill me if i'd cancel our yearly vacation Aug 12 10:34:40 ok.. i am doing this ATM... record (mic -> speex_encoding -> fifobuffer) play( fifobuffer -> speex_decoding -> play) .. in a select loop Aug 12 10:34:59 now i am getting ugly gaps.. see http://yfrog.com/03audacityinterruptionsp Aug 12 10:35:10 hmm Aug 12 10:35:11 mickeyl: muxer.OpenSession(False, 98, "/dev/ttySAC0", 115200) make fso-abyss segfault Aug 12 10:35:29 but not muxer.GetVersion() Aug 12 10:35:45 counting the time inside functions show that recording and playing takes 1-3ms .. but between playing and recording is a gap.. so select-loop just does nothing Aug 12 10:35:49 mickeyl, aw.. shucks.. India is a great place to vacation btw :P... Aug 12 10:35:57 Deubeuliou: ok, i will take a look later today, thanks. Aug 12 10:36:06 anyone else want to come over? Aug 12 10:36:31 (SetStatus makes it segfault too) Aug 12 10:36:42 might be interesting to ask the fso guys, they're on HAR this week thougth Aug 12 10:36:45 I don't see why, however ... it used to work Aug 12 10:36:56 i might have broken it, will check after dinner Aug 12 10:37:01 errr, lunch, that is Aug 12 10:37:15 mickey|lunch, okie Aug 12 10:40:13 DocScrutinizer any idea what cound be behind this? putting alsa into non_blocking mode makes no difference .. SND_PCM_NONBLOCK or SND_PCM_ASYNC makes no difference Aug 12 10:40:26 emmentaler: I think you have problems to read out sample buffer at recording correctly. Aug 12 10:45:45 hmm i dont think..i am confused that this works without interrupts when I just remove the speex-stuff.. Aug 12 10:46:13 emmentaler: maybe a look at http://www.sfr-fresh.com/unix/privat/twinkle-1.4.2.tar.gz:a/twinkle-1.4.2/src/audio/audio_device.cpp code helps spot your problems Aug 12 10:47:12 emmentaler: reading out audio *realtime* is a bit of tricky Aug 12 10:47:40 yes.. this problem drives me nuts since a couple weeks.. Aug 12 10:47:46 mickey|lunch, ok, my bad.. will you be able to do a video conference.. a couple of hours will do.. they had Guido last year :D.... please ping back once your return.. thanks Aug 12 10:49:30 emmentaler: also it depends on the particular hw you are using for recording, and the driver that is servicing this hw. At twinkle We seen lots of problems with e.g. bluetooth for mic Aug 12 10:49:47 emmentaler: see http://osdir.com/ml/voip.twinkle/2008-02/msg00158.html ff Aug 12 10:52:06 emmentaler: probably the speex thread increases your irq-service-rersponse-time, so you see buffer overruns and thus get the dropouts Aug 12 10:53:32 hmm.. i am plotting the device statuses pretty often so i would see if a buffer overrun ocurrs.. also implemented handling of buffer overruns.. Aug 12 10:53:38 emmentaler: first of all you may fiddle with the size and number of buffers you use for recording. This may help though not radically fix he issue Aug 12 10:54:04 strange Aug 12 10:54:28 so it might be a problem of the particular audo-hw driver you are using Aug 12 10:54:43 (alsa that is) Aug 12 10:55:05 ok.. but then alsa without speex should also cause problems.. not? Aug 12 10:55:21 seen this many times, and most of them went away when updating to next new alsa-revision ;-) Aug 12 10:55:37 dsp -> works, dsp with speex -> works, alsa -> works, alsa with speex -> does not work Aug 12 10:55:47 emmentaler: no, that depends on so many race conditions and cpu-load situation Aug 12 10:56:09 dsp alsa compatibility layer? Aug 12 10:56:27 also tried.. did not work on moko.. Aug 12 10:56:46 which hw you do your tests on? Aug 12 10:57:15 eeepc 1000H and sblive... Aug 12 10:58:00 hmmm, so if you say dsp works, which kind of /dev/dsp is this then? I assume thats alsa OSS compatibility lib Aug 12 10:58:35 yes.. alsa-oss on these computers work.. but on moko it does not work Aug 12 10:58:37 if so I assume your generic alsa recording code is lakey Aug 12 10:59:04 on mko we miss the lib for now it seems Aug 12 10:59:25 anyway on moko OSS is strongly deprecated anyway Aug 12 11:00:13 read http://www.sfr-fresh.com/unix/privat/twinkle-1.4.2.tar.gz:a/twinkle-1.4.2/src/audio/audio_device.cpp, check your code. That's all I have for (not so good) advice Aug 12 11:00:18 yes.. but it was pretty fast developing this to get audio input.. (other devs were waiting for audio) Aug 12 11:01:28 thank you.. this is the first time i could talk about this problem.. chats with alsa devs werent very productive.. Aug 12 11:02:07 reatime audio is something like voodoo.. here stops the "exact science" Aug 12 11:02:18 yes, my experience is exactly same wrt alsadevels Aug 12 11:04:36 emmentaler: I found it's often better to use plughw or even dsnoop for input, rather than plain hw:0.0 Aug 12 11:05:12 just another anecdote to share, dunno if it may help for your issue Aug 12 11:05:30 i am still using plughw Aug 12 11:05:40 will now try just hw:0.0 .. Aug 12 11:06:20 my main problem ist.. how do i find out what causes this delay.. since try/error changing the alsa params does not work.. Aug 12 11:06:55 when plotting the times i find that there is a gap between play() and record() Aug 12 11:07:19 so while playing record is not able to get data? hmm.. Aug 12 11:08:09 emmentaler: you know you need separate threads to service record and play? Aug 12 11:08:51 threads? you mean streams.. Aug 12 11:09:40 well, let me put it this way: you mustn't service record and playback in same loop of code Aug 12 11:10:16 * emmentaler is confused.. Aug 12 11:11:44 while true; record(buffer B); playback(buffer B); endwhile; Aug 12 11:11:50 won't pan out Aug 12 11:15:33 but: while true; select(audioout,audioin); if audioin_ready (record()); if audioout_ready(play()) Aug 12 11:17:18 you see the problem? Aug 12 11:20:30 no... basically we have two ringbuffers.. input and output .. all we do with this is read from input and put that into output.. als long as we do this fast enought that no buffer underruns/overruns this works.. our computers are fast enough to "copy" the data that it "plays" the data.. Aug 12 11:22:34 or am i completely wrong? Aug 12 11:25:02 in theory yes. in RL you are just one task of may that gets a timeslice from central arbiter/scheduler every now and then, and calls to bool audioin_ready are NOT zero return time, even might block (though they shouldn't of course). As may play Aug 12 11:27:10 well, as I told, I'm not an expert. I just can advice you to read the audio_device.cpp code of twinkle (plus the way it's used in overlaying modules), and try to learn about the nasties of alsa by doing so Aug 12 11:30:28 ok.. thank you for your help.. will now compyre twinkles implementation and my impl. but now is lunchtime.. Aug 12 11:33:19 emmentaler: to explain a little better: bool audioin_ready may include a hw-access to physical ringbuffer. Depending on actual audio-hw this ringbuffer may live in a remote headset coupled to your machine via BT or USB. So this function may take quite some time, or even fail completely Aug 12 11:38:05 that's reason to have concurrent threads to service record and play independently Aug 12 11:41:30 this dolds true even more if you insert a coding/decoding stage like e.g. speex between record and play Aug 12 11:41:43 s/ do/ ho/ Aug 12 11:41:43 DocScrutinizer meant: this holds true even more if you insert a coding/decoding stage like e.g. speex between record and play Aug 12 11:42:41 you need double/multiple local buffering then Aug 12 11:43:35 i.e speex needs a third own thread Aug 12 11:44:00 and you need pies between those three threads Aug 12 11:44:05 pipes Aug 12 12:00:13 SHR: 03seba.dos1 07shr-themes * ra7168385442f 10/elementary/elementary-theme-niebiee/ (10 files in 2 dirs): elementary-theme-niebiee: slider and radio theme, initial work on genlist theme, little changes in list and clock themes Aug 12 12:05:03 SHR: 03seba.dos1 07shr-themes * r346347487f63 10/elementary/elementary-theme-niebiee/TODO: elementary-theme-niebiee: update TODO Aug 12 12:05:17 Mirv: if you're around, might you be able to try running the KMS Xorg driver under GDB and get a backtrace? i'm really struggling to find what's wrong, and wondering if it's the same for other people Aug 12 12:08:06 emmentaler: you also could follow a async concept (which I'd recommend. See http://www.alsa-project.org/alsa-doc/alsa-lib/group___p_c_m.html#g0d343597cdce35871a5010b6e6d5535b). Still you need more than one buffer and some pipe-alike implementation for those buffers Aug 12 12:08:43 mrmoku: upgrade worked on shr-unstable 08.08. Aug 12 12:09:15 Heinervdm, and I can reproduce the update problem on testing :( Aug 12 12:15:04 mrmoku: perhaps it doesn't work... it does not register to GSM Aug 12 12:17:19 mrmoku: http://shr.pastebin.com/m24de2fce Aug 12 12:17:44 Heinervdm: do you say it's safe now to run "opkg upgrade" on a 88-image? :-)) Aug 12 12:17:52 DocScrutinizer: no Aug 12 12:18:17 DocScrutinizer: For me it does not register to GSM Aug 12 12:18:34 yep, that's the old error still Aug 12 12:18:57 DocScrutinizer: looked at the pastebin? Aug 12 12:19:08 nope. which one? Aug 12 12:19:17 DocScrutinizer: http://shr.pastebin.com/m24de2fce Aug 12 12:19:28 i have no logs from the old error Aug 12 12:21:40 Deubeuliou, ping Aug 12 12:21:44 Deubeuliou, is dbus.exceptions.DBusException: org.freedesktop.DBus.Error.NoReply: Message did not receive a reply (timeout by message bus) Aug 12 12:21:50 what you see with fso-abyss? Aug 12 12:22:10 yepyep, when I accidentally did a "opkg install fso-abyss" 2 days ago, it borjked ogsmd for me as well. I replaced abyss, libfsobasics and libfsotransport back to original and it seems to be fixed Aug 12 12:24:29 * DocScrutinizer wonders how he could ask stupid questions like "which pastebin"... hmmm Aug 12 12:25:33 prolly becuase I look at N810 kbd when typing, and not at screen ;-) Aug 12 12:27:26 :) Aug 12 12:33:06 SHR: 03seba.dos1 07shr-themes * rd6a01ef0affc 10/elementary/elementary-theme-niebiee/images/ (check_off.png check_on.png radio_off.png radio_on.png): elementary-theme-niebiee: fix radio and checkbox graphics to look good on x11-16 Aug 12 12:37:56 mrmoku: yes Aug 12 12:38:07 bu Ainu told me it was the same on SHR-unstable Aug 12 12:38:21 I'll try -testing as soon as I get those FRs sorted Aug 12 12:39:00 Deubeuliou, any idea yet where that could come from? Aug 12 12:39:15 fso-abyss broken, most likely Aug 12 12:39:35 until rev 67, probably Aug 12 13:03:55 Hello,,,,can I set a 'VISIBLE - BIG ' clock on mainscreen in SHR? Aug 12 13:04:13 how do I have to configure? Aug 12 13:04:24 Heinervdm, do you have /etc/abyss.conf? and what does it contain? Aug 12 13:04:32 (on shr-unstable) Aug 12 13:05:04 mrmoku: http://shr.pastebin.com/m6b93a9ce Aug 12 13:05:12 I didn't recover /etc/abyss.conf Aug 12 13:05:49 though I didn't check if it changed by "opkg install fso-abyss" Aug 12 13:06:18 SHR: 03seba.dos1 07shr-themes * r1a9740bf9031 10/elementary/elementary-theme-niebiee/images/ (bubble_1.png bubble_2.png bubble_3.png bubble_4.png): elementary-theme-niebiee: now bubbles are also blue ;) Aug 12 13:07:21 anyway works now, after recovering/restoring abyss ELF, libfsobasics.so, libfsotransport.so Aug 12 13:07:26 Heinervdm, ok, looks sane Aug 12 13:09:48 abyss exe in 88 tarball is quite some months lder than the one that got installed by "opkg install", and it's some bytes off in filesize as well Aug 12 13:10:45 the libs are only a few hours apart from their newer non-funct versions Aug 12 13:11:49 though I can't tell for sure those libs or abyss is the non-funct borked part. I restored all three from original 88-tarball and it seems to fix the issue Aug 12 13:12:22 * mrmoku has some theory... Aug 12 13:12:44 TAsn: commit for you: Aug 12 13:12:52 SHR: 03seba.dos1 07shr-themes * r8727734c6e15 10/elementary/elementary-theme-niebiee/images/ (bubble_2.png bubble_4.png): elementary-theme-niebiee: make right-sided bubbles little darker Aug 12 13:14:06 SHR: 03seba.dos1 07shr-themes * r2162a43c2ba9 10/elementary/elementary-theme-niebiee/TODO: elementary-theme-niebiee: update TODO Aug 12 13:14:31 dos1: make right-sided boobies look better Aug 12 13:14:43 dos1: oh, there's none? Then add them! Aug 12 13:14:57 NOOOOooOO dos1 will make blue boobies then :P Aug 12 13:15:28 mrmoku: wtf? is he some kind of bdsm activist? Aug 12 13:15:40 PaulFertser, no, but he loves blue ;) Aug 12 13:16:00 mrmoku: blue sky, i assume? ;) Aug 12 13:16:18 PaulFertser, nahh... look at his theme... its more like blue everything :P Aug 12 13:16:42 name of theme: "niebiee" Aug 12 13:16:46 blue in polish: "niebieski" Aug 12 13:16:48 ;) Aug 12 13:17:36 anyway, bbl Aug 12 13:17:44 dos1: that sounds like some swearing word in russian Aug 12 13:18:09 * DocScrutinizer wishes, dos1 would reduce size of (almost) useless topbar of terminal kbd in some theme Aug 12 13:18:17 dos1|away: you theme name will be interpreted as "don't fuck" Aug 12 13:18:52 PaulFertser: sounds like a good naming then ;-D Aug 12 13:20:01 DocScrutinizer: i'd prefer "zaibiees" -- that'd sound like "ever'thing's fucking cool" Aug 12 13:23:50 Heinervdm, could you try to upgrade again and see if fso-abyss works now? Aug 12 13:23:57 ok Aug 12 13:24:48 * mrmoku is crossing fingers Aug 12 13:31:59 mrmoku: no, same error as before Aug 12 13:32:03 shit Aug 12 14:53:12 hello Aug 12 14:53:25 SHR: 03seba.dos1 07shr-themes * rf097e2af1aeb 10/elementary/elementary-theme-niebiee/niebiee.edc: elementary-theme-niebiee: more work on genlist Aug 12 14:55:34 im in trouble with the newest shr unstable version.. most of the applications won't start, because of missing python-codecs, but it's installed Aug 12 14:57:07 error log: Aug 12 14:57:10 http://pastebin.com/m4c7790be Aug 12 14:59:38 can anyone help? Aug 12 15:00:17 also "opkg" doesn't exist, instead "opkg-cl".. dont know why Aug 12 15:00:52 did you use the latest image or use 'opkg upgrade'? Aug 12 15:01:08 latest image Aug 12 15:01:37 first i tried opkg upgrade, but the freerunner wasnt starting anymore Aug 12 15:02:01 im having lunch right now, ill be back in 45 minutes.. Aug 12 15:02:08 bye Aug 12 15:02:15 I just went through that myself and discovered that its currently broken - see the TOPIC for this channel... Aug 12 15:06:39 SHR: 03seba.dos1 07shr-themes * rbfbc9fbe546c 10/elementary/elementary-theme-niebiee/niebiee.edc: elementary-theme-niebiee: little fine-tuning of pager Aug 12 15:06:40 SHR: 03seba.dos1 07shr-themes * r9085d4707e12 10/elementary/elementary-theme-niebiee/niebiee.edc: elementary-theme-niebiee: fix gradient on very small frames Aug 12 15:16:21 freesmartphone.org: 03mickey 07libgsm0710mux * rde2f0448f851 10/gsm0710mux/ (channel.vala consts.vala manager.vala multiplexer.vala): libgsm0710mux: catch up with new logging API Aug 12 15:16:57 mickey|lunch, ahh, does that fix fso-abyss? Aug 12 15:16:59 mickeyl: fixing? ;) Aug 12 15:17:06 *nod* Aug 12 15:17:09 :) Aug 12 15:17:12 * mrmoku buildxs Aug 12 15:17:13