**** BEGIN LOGGING AT Sat Apr 30 02:59:58 2011 Apr 30 07:17:20 moin Apr 30 07:58:51 freesmartphone.org: 03mok 07cornucopia * r14e9a4b4c1de 10/fsogsmd/ (5 files in 4 dirs): Apr 30 07:58:51 freesmartphone.org: fsogsmd: add stub for a new plugin pdp_nokia_isi Apr 30 07:58:51 freesmartphone.org: Signed-off-by: Klaus Kurzmann Apr 30 08:39:18 moin Apr 30 09:16:51 hi Apr 30 09:37:15 freesmartphone.org: 03mok 07cornucopia * re0b1fcafb629 10/.gitignore: Apr 30 09:37:16 freesmartphone.org: .gitignore: ignore vim swp files Apr 30 09:37:16 freesmartphone.org: Signed-off-by: Klaus Kurzmann Apr 30 09:37:58 mickey|bbl: ping Apr 30 09:40:59 GNUtoo|laptop: hi Apr 30 09:41:08 GNUtoo|laptop: I solved the robotic voice problem :) Apr 30 09:41:23 GNUtoo|laptop: our driver reads some firmware from /etc/firmware/htcleo.adcb Apr 30 09:41:24 ok nice Apr 30 09:41:29 I'm writing the activator Apr 30 09:41:33 for fso ? Apr 30 09:41:37 of course Apr 30 09:41:40 crazy !! Apr 30 09:41:42 :) nice Apr 30 09:41:56 there is something we must look at though: microphone does not seem to work Apr 30 09:42:24 http://www.pastie.org/1850015 Apr 30 09:42:28 it's not complete yet Apr 30 09:42:36 there is a small compilation problem Apr 30 09:42:40 and some error -1 at runtime Apr 30 09:43:18 I just uploaded the binaries for captainigloo to test Apr 30 09:46:52 captainigloo: does it work on desire ? http://htc-linux.org/stuff/gnutoo-qsd-routing.tar.bz2 Apr 30 09:50:14 dcordes, no it won't work Apr 30 09:50:22 ah sorry Apr 30 09:50:27 htc-linux.org Apr 30 09:50:33 that are the binaries from gitorious Apr 30 09:50:39 yep Apr 30 09:50:40 because the binaries on my website don't work Apr 30 09:50:45 but gitorious does Apr 30 09:50:50 yep it is the new gitorious git code Apr 30 10:00:18 ok I'll let the runtime error in Apr 30 10:00:18 it's not problematic Apr 30 10:00:26 it just tries to issue an ioctl twice Apr 30 10:00:35 one for FREE_SMARTPHONE_GSM_CALL_STATUS_OUTGOING Apr 30 10:00:42 one for FREE_SMARTPHONE_GSM_CALL_STATUS_ACTIVE Apr 30 10:00:52 where should I put the code? Apr 30 10:00:58 same gitorious repo? Apr 30 10:05:49 mickey|bbl, ping Apr 30 10:05:59 I must get your permission to publish this thing Apr 30 10:06:15 (the snapdragoon call activator) Apr 30 10:06:21 since I took from your code + apache Apr 30 10:06:35 the easiest way is to license the whole under apache Apr 30 10:06:39 or gplv3 Apr 30 10:06:41 GNUtoo|laptop: cannot wait for you to push it, rebuild image, run it :) Apr 30 10:06:58 GNUtoo|laptop: how can we get the additional commands? Apr 30 10:07:21 additional commands? Apr 30 10:08:20 yop Apr 30 10:08:26 let me try to fix the license and I publish Apr 30 10:08:47 it's the same as the link you give me yesterday ? Apr 30 10:08:57 I'll publish as GPLv3 Apr 30 10:09:13 that is what mickeyl choose and it's compatible with apache 2.0 Apr 30 10:11:56 no Apr 30 10:11:58 not the same Apr 30 10:12:06 something that works automatically Apr 30 10:13:41 GNUtoo|laptop: I think captainigloo is asking about what I uploaded Apr 30 10:14:42 captainigloo: https://gitorious.org/htc-msm-2-6-32/uttilities is http://htc-linux.org/stuff/gnutoo-qsd-routing.tar.bz2 Apr 30 10:15:13 captainigloo: it is an updated (fixed) version of the source tarball from yesterday Apr 30 10:15:37 ok cool Apr 30 10:15:49 i will try later Apr 30 10:15:58 busy with my edje tutorial right now Apr 30 10:16:24 captainigloo: writing or reading? :P Apr 30 10:16:34 captainigloo: looking forward to the result. I think it is a major step to make calls on our phones Apr 30 10:17:47 mrmoku: writing Apr 30 10:18:00 * mrmoku promises to read it then :) Apr 30 10:18:06 I will push very soon Apr 30 10:18:12 mrmoku: in french .... Apr 30 10:18:17 ouch Apr 30 10:18:26 that makes reading a lot more difficult :P Apr 30 10:18:26 GNUtoo|laptop: additional commands: I would like to dump everything that gets written to and read from /dev/msm_audio_ctl Apr 30 10:18:29 GNUtoo|laptop: inside of android Apr 30 10:18:41 but i promise writing a english version Apr 30 10:18:51 GNUtoo|laptop: then we can extend your fso addon by handset, and other routing configurations ! Apr 30 10:18:54 ok :) Apr 30 10:19:23 but you should be able to read the example ;) Apr 30 10:19:52 GNUtoo|laptop: if I understand it correctly, right now there is only HEADSET_SPKR_STEREO Apr 30 10:19:55 mrmoku: http://trac.enlightenment.org/e/browser/trunk/DOCS/tutorials/fr/edje Apr 30 10:19:59 not exactly Apr 30 10:20:02 all tut0x are edc file examples Apr 30 10:20:12 ok, thx Apr 30 10:20:15 basically it won't handle routing at all, only voicecalls Apr 30 10:20:21 routing can be done separately Apr 30 10:20:33 GNUtoo|laptop: what is the difference between routing and voicecalls ? Apr 30 10:20:54 voicecalls activates the voicecalls Apr 30 10:21:02 routing route whatever state somewhere: Apr 30 10:21:05 headphones Apr 30 10:21:07 speakers Apr 30 10:21:08 etc... Apr 30 10:21:16 that's what we should do Apr 30 10:21:32 maybe routing won't be needed if cotulla succeed Apr 30 10:21:37 but maybe voicecall will Apr 30 10:21:53 ok. so voicecall is not a waste of time then Apr 30 10:21:56 routing == the idea of routing != the routing program I wrote Apr 30 10:22:00 no idea Apr 30 10:22:09 anyway it's done Apr 30 10:22:12 so I publish Apr 30 10:22:30 Moin Apr 30 10:22:55 GNUtoo|laptop: ok what about extra voicecall settings ? do you think can we can get handset instead of speaker ? Apr 30 10:23:14 ChristW: moin Apr 30 10:23:24 what is the difference between both Apr 30 10:23:44 mrmoku: Around? Apr 30 10:24:27 GNUtoo|laptop: a voice call in speaker mode: you talk on the speaker. handset mode: you talk with the phone on your ear Apr 30 10:24:40 http://gitorious.org/htc-msm-2-6-32/uttilities/trees/master/fso-voicecall Apr 30 10:24:57 dcordes, yes but hardware wise what's the difference? Apr 30 10:25:03 *different volumes? Apr 30 10:25:13 *different output speaker? Apr 30 10:25:29 GNUtoo|laptop: *different output speaker Apr 30 10:25:33 ah ok Apr 30 10:25:35 that can be done Apr 30 10:25:36 then Apr 30 10:25:37 GNUtoo|laptop: that is for sure. about volume, no idea Apr 30 10:25:41 I'll look later then Apr 30 10:25:56 but as I said Apr 30 10:26:02 better doing 2 separate programs Apr 30 10:26:06 one for voicecalls Apr 30 10:26:09 and one for routing Apr 30 10:26:25 I think I still don't get the exact difference Apr 30 10:26:43 I know wew have routing.c and now we have your fso-voicecall git Apr 30 10:26:46 basically voicecall activate or desactivate the audio to the baseband Apr 30 10:26:51 ok Apr 30 10:26:53 forget about routing.c Apr 30 10:27:00 I'm talking about something else: Apr 30 10:27:09 routing == the idea of routing != the routing program I wrote Apr 30 10:27:24 someone should write a new routing.c Apr 30 10:27:27 that handle routing Apr 30 10:27:32 only routing Apr 30 10:27:36 not voicecalls Apr 30 10:28:00 so voicecalls activate desactivate the path to the baseband for audio Apr 30 10:28:08 then that audio can be routed to speakers Apr 30 10:28:11 or headphones Apr 30 10:28:15 or handset Apr 30 10:28:16 etc... Apr 30 10:29:07 anyway to compile do that: Apr 30 10:29:10 source the devshell Apr 30 10:29:19 go in the vso-voicecall directory Apr 30 10:29:26 sh fso-voicecall.sh Apr 30 10:29:33 and it'll build it Apr 30 10:34:15 mrmoku, mickey|bbl is not on irc, will he come? Apr 30 10:34:33 or maybe I should ask morphis but he's not there either Apr 30 10:34:58 the question is: how to implement the alsa player.... Apr 30 10:42:02 mrmoku, what's up with the n900 modem? Apr 30 10:51:00 TAsn: Around? Apr 30 10:58:54 freesmartphone.org: 03mickey 07libgisi * r75cf79186cf0 10/gisicomm/gisicomm.vala: gisicomm: speed up the subsystem initialization Apr 30 10:59:16 freesmartphone.org: 03mickey 07cornucopia * r396c1675c640 10/fsogsmd/src/plugins/modem_nokia_isi/ (channel.vala unsolicited.vala): fsogsmd: modem_nokia_isi: start sanitizing the startup sequence (more work necessary) Apr 30 11:04:38 GNUtoo|laptop: looks like he's active though :) Apr 30 11:05:06 lol Apr 30 11:05:16 he's hidding maybe Apr 30 11:06:06 morning folks Apr 30 11:06:15 morning mickeyl :) Apr 30 11:06:47 mickeyl: I started to implement a pdp_nokia_isi and have a question Apr 30 11:06:58 what is the correct way to get access to gisicomm? Apr 30 11:07:16 let me think for a second Apr 30 11:07:24 ok Apr 30 11:08:56 what part do you need? Apr 30 11:09:09 gpsd Apr 30 11:09:15 NokiaIsi.isimodem is a singleton that gets you to the modem class Apr 30 11:09:18 err, instance Apr 30 11:09:36 ok Apr 30 11:10:11 alternatively you are allowed to use FsoGsm.theModem from everywhere Apr 30 11:10:17 cast this to the actual modem class Apr 30 11:10:23 and move from there until you get the instances you need Apr 30 11:10:30 ok, what is nicer? Apr 30 11:10:44 use the first approch, if it works Apr 30 11:10:57 i'm not sure whether this is visible from any plugin atm. Apr 30 11:11:00 but you'll see soon ;) Apr 30 11:11:06 ok :) Apr 30 11:12:06 ChristW_away, most of the time. Apr 30 11:13:36 mrmoku: we need to talk about the -224 Apr 30 11:13:46 i must be missing something Apr 30 11:13:48 but i don't see it Apr 30 11:13:56 s/-224/224/ Apr 30 11:14:07 yeah Apr 30 11:14:12 as the packages look correct Apr 30 11:14:17 not important atm., but i'd like to fix/understand that before doing larger changes Apr 30 11:14:18 maybe it's just a message parsing problem? Apr 30 11:14:40 on gisi or gisicomm? Apr 30 11:15:01 dunno Apr 30 11:15:07 I can take a look too Apr 30 11:15:25 ok, thanks, 2nd pair of eyes always good Apr 30 11:15:41 one more thing, you seem to be more leaning towards fsoaudiod for the in-call voice handling? Apr 30 11:16:14 my gut feeling is still it rather belongs to fsogsmd itself, but perhaps you can convince me ;) Apr 30 11:16:53 that's just a feeling too... as fsoaudiod is handling alsa anyway Apr 30 11:17:03 so no strong opinion Apr 30 11:17:14 if you prefer it to reside in fsogsmd I'm fine as well :) Apr 30 11:17:40 have to go for lunch now Apr 30 11:17:41 bbiab Apr 30 11:17:42 mickeyl, mrmoku: sorry to be jumping in, in the middle of the discussion Apr 30 11:18:04 but I think in-call volume handling should be in fsoaudiod Apr 30 11:18:06 and not fsogsmd Apr 30 11:18:15 this way Apr 30 11:18:24 sure Apr 30 11:18:32 but we're talking actual voice data transport Apr 30 11:18:35 not volume control Apr 30 11:18:51 i.e routing modem -> sound card? Apr 30 11:19:02 that's fsogsmd Apr 30 11:19:02 mickeyl, hi Apr 30 11:19:05 yep Apr 30 11:19:08 that's what i was thinking Apr 30 11:19:16 cause that's gsm modem related Apr 30 11:19:17 :) Apr 30 11:19:22 right Apr 30 11:19:31 on n900 we have to manually transport the voice data Apr 30 11:19:37 "manually" ;) Apr 30 11:19:39 hi GNUtoo|laptop Apr 30 11:19:43 though I guess Apr 30 11:19:51 you want to be able to say to fsoaudiod Apr 30 11:20:00 mickeyl, I recorded voice and played it back last night Apr 30 11:20:05 This routing is ok, but mute that, and instead enable "media sound" Apr 30 11:20:08 or something like that. Apr 30 11:20:14 anyhow, gtg. Apr 30 11:20:18 meeting with some friends. Apr 30 11:20:26 TAsn: good point, will take this into account. thanx Apr 30 11:20:32 GNUtoo|laptop: excellent, which format do we deal with? Apr 30 11:20:34 mickeyl, np. Apr 30 11:20:40 let me look the logs Apr 30 11:20:51 aplay -r 8000 voice_data.raw -f S16_LE seem to work Apr 30 11:20:59 http://pastie.org/private/dbugvicytmocjexgnehw Apr 30 11:21:03 diff from the cmt program Apr 30 11:21:21 hmm, 8K sample rate 16 bits Apr 30 11:21:26 just about right Apr 30 11:21:27 yes Apr 30 11:21:40 mickeyl, bottom line in my pov: fsogsmd should take care of adding the input source when relevant. fsoaudiod should manage it. Apr 30 11:21:43 ciao, gtg. Apr 30 11:22:16 although i guess they're not using the full spectrum of those 16 bits Apr 30 11:22:23 would be interesting to see Apr 30 11:22:24 heyho Apr 30 11:22:27 moin morphis Apr 30 11:22:32 hi morphis Apr 30 11:22:42 mickeyl: how is pyside progressing in OE? Apr 30 11:22:42 GNUtoo|laptop: ok, so i guess we should enable recording with a seperate command line parameter Apr 30 11:22:58 morphis: i managed to build apiextractor-native and generatorrunner-native Apr 30 11:23:02 no time yet to continue Apr 30 11:23:13 ok Apr 30 11:23:13 mickeyl, how do we integrate all that in fso? Apr 30 11:23:16 12:27 < GNUtoo|laptop> so voicecalls activate desactivate the path to the baseband for audio Apr 30 11:23:19 12:27 < GNUtoo|laptop> then that audio can be routed to speakers Apr 30 11:23:22 12:27 < GNUtoo|laptop> or headphones Apr 30 11:23:23 but I thought that is exactly what routing.c does Apr 30 11:23:26 12:27 < GNUtoo|laptop> or handset Apr 30 11:23:28 12:27 < GNUtoo|laptop> etc... Apr 30 11:23:34 GNUtoo|laptop: once cmtspeech is done, we move the code into fsogsmd or fsoaudiod Apr 30 11:23:45 but cmtspeech will stay as standalone testing Apr 30 11:23:46 dcordes, yes but it has also a VOICE_START thing Apr 30 11:23:56 dcordes, and it's incomplete Apr 30 11:24:07 GNUtoo|laptop: incomplete = it lacks handset Apr 30 11:24:13 yes Apr 30 11:24:15 GNUtoo|laptop: only speaker is bad Apr 30 11:24:26 GNUtoo|laptop: also: in speaker mode, other side can't hear me. volume problem ? Apr 30 11:24:28 dcordes, I'll fix *later* Apr 30 11:24:39 dcordes, routing problem I bet Apr 30 11:24:47 ok, how can I analyze it ? Apr 30 11:24:48 we need to tell to use the microphone Apr 30 11:24:53 I'll look later Apr 30 11:24:55 GNUtoo|laptop: i want the audio being done before i finish the callhandler, it's much easier in that order Apr 30 11:25:09 I thought about dumping /dev/msm_audio_ctl in android Apr 30 11:25:13 dcordes, look at the android AudioHardware.cpp Apr 30 11:25:36 mickeyl, ok btw I tried the dbus code on nexusone and there is a bug: Apr 30 11:26:24 case FreeSmartphone.GSM.CallStatus.OUTGOING: Apr 30 11:26:25 case FreeSmartphone.GSM.CallStatus.ACTIVE: Apr 30 11:26:30 GNUtoo|laptop: http://android.git.kernel.org/?s=AudioHardware.cpp no result Apr 30 11:26:38 makes the code called twice Apr 30 11:26:45 once for outgoing Apr 30 11:26:48 one for active Apr 30 11:26:51 yes Apr 30 11:26:58 ah ok Apr 30 11:26:59 i didn't bother to check Apr 30 11:27:04 ok Apr 30 11:27:08 http://gitorious.org/htc-msm-2-6-32/uttilities/blobs/master/fso-voicecall/fso-voicecall.vala is what I did Apr 30 11:27:14 for nexusone Apr 30 11:27:16 since it's irrelevant for cmtspeech Apr 30 11:27:22 ok Apr 30 11:27:40 ideally you'd keep a shadow variable Apr 30 11:27:46 dcordes, second Apr 30 11:27:48 that keeps track whether you already enabled / disable the audio path Apr 30 11:28:05 that code looks good Apr 30 11:28:08 you're going to be a vala coder :D Apr 30 11:28:14 dcordes, http://gitorious.org/replicant/android_hardware_msm7k Apr 30 11:28:31 lol I was very tired and did it at night Apr 30 11:28:37 and had a hard time doing it Apr 30 11:28:43 because of the valacod Apr 30 11:28:47 *valadoc Apr 30 11:28:57 i don't bother with valadoc Apr 30 11:29:02 i can read vapis by now Apr 30 11:29:06 that's quicker Apr 30 11:29:07 :) Apr 30 11:29:08 ah ok Apr 30 11:29:49 one thing i'd change in that code is Apr 30 11:30:03 don't just output the error code, but rather use Posix.strerror(Posix.errno)) Apr 30 11:30:12 (after the ioctl) Apr 30 11:30:14 ok thanks Apr 30 11:30:27 so I do a try catch? Apr 30 11:30:36 no, err < 0 checking is correct Apr 30 11:30:41 just add a bit more to the debug statement Apr 30 11:30:42 Posix.strerror(err) then Apr 30 11:31:06 yes, or that. i'm not sure whether ioctl just outputs -1 or the actual errno+ Apr 30 11:31:13 i usually use Posix.errno Apr 30 11:31:18 ok Apr 30 11:31:19 which is set after all posix functions Apr 30 11:31:25 (or so they claim ;) Apr 30 11:31:44 GNUtoo|laptop: thanks Apr 30 11:32:08 GNUtoo|laptop: so you can copy paste from your own git to improve routing.c :) Apr 30 11:32:27 http://gitorious.org/replicant/android_hardware_msm7k/blobs/master/libaudio-qsd8k/AudioHardware.cpp Apr 30 11:32:58 dcordes, yes Apr 30 11:33:31 mickeyl, so how I start to implement this alsa player? Apr 30 11:33:39 I do it from scratch? Apr 30 11:33:43 I pipe to a program? Apr 30 11:33:46 I use a lib? Apr 30 11:33:50 I use an fso lib? Apr 30 11:35:36 GNUtoo|laptop: basically from scratch. It's going to be a bit of work. i have bound most of Alsa in alsa.vapi Apr 30 11:35:51 you need to open/close the soundcard, set the format, and pipe data in/out Apr 30 11:35:52 ah ok Apr 30 11:35:59 ok Apr 30 11:36:17 is there any doc on the vapis or how to read vapis? Apr 30 11:37:14 nothing i can find offhand. but basically, it's just a header file Apr 30 11:37:21 ok Apr 30 11:37:22 you see the class definitions, method definitions, enums, etc. Apr 30 11:37:28 ok Apr 30 11:37:29 you can ignore all stuff in brackets Apr 30 11:37:30 I'll look Apr 30 11:37:33 where is it in git\? Apr 30 11:37:37 s#\## Apr 30 11:37:51 GNUtoo|laptop: n900 alsa player ? Apr 30 11:37:54 alsa vapi is upstream Apr 30 11:38:25 dcordes, yes Apr 30 11:38:33 mickeyl, ok Apr 30 11:39:01 dcordes, maybe you should ask mickeyl to autotoolize my code and put in in fso git Apr 30 11:39:16 (for the qsd audio activator) Apr 30 11:39:45 mickeyl: what do you think about autotoolizing GNUtoo|laptop's code for qsd audio activation and put it in fso git ? Apr 30 11:39:52 fd_a1026 = open("/dev/audience_a1026", O_RDWR); Apr 30 11:39:55 wtf a1026 ? Apr 30 11:40:39 GNUtoo|laptop: do AudioHardware.cpp and routing.c write numeric values to /dev/msm_pcm_out ? Apr 30 11:40:55 dcordes, ? Apr 30 11:41:09 dcordes, I'll look Apr 30 11:41:19 where is the part that makes you think that? Apr 30 11:41:29 GNUtoo|laptop: at the beginning of AudioHardware.cpp , all 'modes' are assigned to a number Apr 30 11:41:36 print here Apr 30 11:41:42 ok Apr 30 11:41:59 i don't have time to do it myself soon, but yet, it would be a natural fit in fso:utilities/qsdaudio or so Apr 30 11:42:05 s/yet/yes/ Apr 30 11:42:06 mickeyl meant: i don't have time to do it myself soon, but yes, it would be a natural fit in fso:utilities/qsdaudio or so Apr 30 11:42:10 gitorious makes my browser eat lots of cpu / mem Apr 30 11:42:19 * mickeyl leaving for bbq soon Apr 30 11:42:28 55 Apr 30 11:42:30 static const uint32_t SND_DEVICE_HEADSET = 2; Apr 30 11:42:30 56 Apr 30 11:42:30 static const uint32_t SND_DEVICE_HEADSET_AND_SPEAKER = 10; Apr 30 11:42:30 57 Apr 30 11:42:33 static const uint32_t SND_DEVICE_FM_HEADSET = 9; Apr 30 11:42:35 58 Apr 30 11:42:38 static const uint32_t SND_DEVICE_FM_SPEAKER = 11; Apr 30 11:42:52 mickeyl: mmm bbq bon appetito Apr 30 11:42:59 btw., the proper solution of course is to do it as a fsoaudio plugin Apr 30 11:43:05 much like i did the qdsp5_router Apr 30 11:43:16 which is using strange ioctls for routing as well Apr 30 11:44:21 GNUtoo|laptop: it would be nice to experiment with the values without recompiling a program Apr 30 11:44:33 GNUtoo|laptop: i.e. send value to device node. but I don't know which Apr 30 11:45:42 dcordes, mioctl Apr 30 11:45:59 GNUtoo|laptop: ? Apr 30 11:46:07 dcordes, you don't write theses values directly to the device node Apr 30 11:46:11 you use ioctls Apr 30 11:46:21 see do_ioctl in: Apr 30 11:46:27 http://gitorious.org/htc-msm-2-6-32/uttilities/blobs/master/fso-voicecall/fso-voicecall.vala Apr 30 11:47:09 or man 2 ioctl Apr 30 11:48:18 GNUtoo|laptop: but why do the programs open the device node ? Apr 30 11:50:39 to issue an ioctl Apr 30 11:50:42 learn aobut ioctl Apr 30 11:50:58 http://en.wikipedia.org/wiki/Ioctl Apr 30 11:52:58 can I do this from a terminal ? Apr 30 11:56:16 ah now I understand why the other side can't hear me Apr 30 11:56:21 routing.c does not touch MIC Apr 30 11:56:41 unlike AudioHardware.cpp : Apr 30 11:56:57 } else if (device == SND_DEVICE_NO_MIC_HEADSET) { Apr 30 11:56:57 out_device = HEADSET_SPKR_STEREO; Apr 30 11:56:58 mic_device = HANDSET_MIC; Apr 30 11:57:16 so if I want routing.c to give me mic I need to implement the mic function Apr 30 11:59:21 s/function/ioctl/ Apr 30 11:59:22 dcordes meant: so if I want routing.c to give me mic I need to implement the mic ioctl Apr 30 12:03:37 GNUtoo|laptop: routing CALL_START HEADSET_SPKR_STEREO Apr 30 12:03:47 now let's see if we find that stupid 224... Apr 30 12:03:49 GNUtoo|laptop: after that I have something like Apr 30 12:04:58 GNUtoo|laptop: HEADSET_SPKR_STEREO (as expected because it's sent in routing.c) Apr 30 12:05:02 ok Apr 30 12:05:11 dcordes, can I look at it later Apr 30 12:05:12 GNUtoo|laptop: and mic_device = HANDSET_MIC Apr 30 12:05:26 GNUtoo|laptop: als expected because it's standard and not touched by routing.c Apr 30 12:06:40 GNUtoo|laptop: of course just ignore my loud thinking, sorry Apr 30 12:06:44 GNUtoo|laptop: did you see 13:42 < mickeyl> btw., the proper solution of course is to do it as a fsoaudio plugin Apr 30 12:06:57 dcordes, yes Apr 30 12:07:07 ok Apr 30 12:07:08 I'll re-ping mickey later about that Apr 30 12:14:13 personally I would just be happy with routing.c if it did this: additionally send "HEADSET_MIC" ioctl when routing to "HEADSET_SPKR_STEREO", add a routing mode called "HANDSET" sending "HANDSET_SPKR" and "HANDSET_MIC" Apr 30 12:14:39 then speaker and normal phone call could work Apr 30 12:16:17 I feel a bit stupid because I can't add it on my own but hey Apr 30 12:16:27