**** BEGIN LOGGING AT Tue Jun 07 02:59:57 2011 Jun 07 05:10:59 freaking bs :/ Jun 07 05:11:22 JaMa|Off: I built updated images Jun 07 05:11:33 after upgrade I get Jun 07 05:11:33 root@om-gta02 ~ # fsogsmd Jun 07 05:11:33 Illegal instruction Jun 07 05:26:14 actually complete fsoland is sigill Jun 07 05:29:48 mrmoku: shr-u? or shr-c? Jun 07 05:30:12 if both then maybe it's caused by new vala? Jun 07 05:34:11 and I'll kill build of other machines on buildhost before this is resolved somehow Jun 07 05:34:22 even when it's maybe only armv4t issue Jun 07 05:41:59 JaMa|Wrk: shr-u Jun 07 05:42:24 JaMa|Wrk: at least new vala works fine with armv7a Jun 07 05:43:33 JaMa|Wrk: will install the built image now... to see if it's an upgrade issue Jun 07 05:50:59 SHR: 03Martin.Jansa 07shr-chroot * rcaeeea669135 10/usr/local/portage/dev-python/progressbar/ (ChangeLog Manifest metadata.xml progressbar-2.2.ebuild): progressbar: remove ebuild for 2.2, now there is 2.3 in portage tree Jun 07 05:51:01 SHR: 03Martin.Jansa 07shr-chroot * r1dd31e6bd43f 10/ (725 files in 62 dirs): system upgrade Jun 07 05:51:12 SHR: 03Martin.Jansa 07shr-chroot * ree5278423077 10/ (929 files in 32 dirs): system upgrade Jun 07 06:06:05 JaMa|Wrk: fresh image works fine Jun 07 06:06:17 looks like I did a bad update Jun 07 06:11:04 weird Jun 07 06:11:12 I'll test upgrade in few minutes Jun 07 06:15:14 ok Jun 07 06:15:19 * mrmoku off to a client Jun 07 06:15:20 bbiab Jun 07 06:29:05 no sigill here sofar Jun 07 07:09:50 morning Jun 07 07:10:29 ~dict sigill Jun 07 07:10:31 could not find definition for sigill Jun 07 07:10:39 stupid! Jun 07 07:10:50 ~spell sigill Jun 07 07:10:52 possible spellings for sigill: skill seagull soggily Gill gill scull sill skull Sibilla spill still swill vigil Skell signal Sicily Sisile simile sagely sickle sickly silk squall Gilli Gilly gull silly soil signally Gil Sig skills Gall Gila Jill SGML Sile gall girl kill sail sell silo skill's Jun 07 07:11:17 ~sigill Jun 07 07:11:18 well, sigill is an illegal instruction (means you screwed up somewhere in your code) Jun 07 07:12:23 should get extended to read ", or a descriptive term for something containing not a single working bit" Jun 07 07:12:55 "my brain feels sigill today" Jun 07 07:45:00 JaMa|Wrk: ok was just my personal upgrade breakage then :) Jun 07 08:17:37 Moin all. Jun 07 08:18:35 I've got some news: At the end of the year, my wife and I are going to have a baby! Oh, and I've broken my arm... Jun 07 08:19:53 congrats and sorry Jun 07 08:20:15 Fell off my bike. Jun 07 08:22:16 darn Jun 07 08:26:32 One-handed typing sucks, too! Jun 07 08:30:02 mrmoku: Around? Jun 07 08:31:12 ChristW: not really Jun 07 08:31:12 mrmoku: Now, fsodeviced does not crash when my phone rings, it merely states 'Failed to update scenario 'gsmhandset' to get changes: Device or resource busy' Jun 07 08:31:24 will be back in " hour Jun 07 08:31:32 1 Jun 07 08:31:35 Mkay, will wait. Jun 07 08:31:44 k Jun 07 08:32:06 * ChristW goes back to watching 'the IT crowd'... Jun 07 09:16:31 ChristW: back Jun 07 09:20:51 ChristW: heh, now I read you're becoming daddy :) congrats Jun 07 09:25:04 TAsn: I would want your advise on how to fix scaling of the quick settings screen Jun 07 09:34:24 mrmoku: Thanks! I'm back aswell... Jun 07 09:37:52 ChristW: does it work then? Or is audio broken? Jun 07 09:38:22 No, it doesn't work. No mic and no speaker when I call to my FR> Jun 07 09:38:37 hmm Jun 07 09:40:36 lsof tells me that there's 2 'audio' related processes, fsoaudiod and fsodevice/audio.so Jun 07 09:50:28 ChristW: could you try if it works after '/etc/init.d/fsoaudiod stop' ? Jun 07 09:52:15 Still 'device or resource busy', so it's not that one. Jun 07 09:58:14 Outgoing call (to voicemail) also has no audio after reboot. Jun 07 09:59:01 I tried voicemail before and it worked... had problems on an incoming call now too though Jun 07 10:01:01 Glad I'm not the only one, although I think I saw that pespin had problems as well. Jun 07 10:01:23 but he said they're solved by the upgrade Jun 07 10:01:28 btw. did you upgrade? Jun 07 10:01:30 :P Jun 07 10:01:46 I upgraded this morning. I got a new fsoaudiod. Jun 07 10:01:53 (and others...) Jun 07 10:01:56 ok Jun 07 10:03:07 there was a message about an invalid byte somewhere... let me re-upgrade. No, just the prompt again... Jun 07 10:05:01 you're on tests feed, right? Jun 07 10:05:07 Yes. Jun 07 10:05:53 hmm Jun 07 10:05:54 ouch Jun 07 10:06:19 strange Jun 07 10:07:34 damn... we should really, really push forwards to fsoaudiod... as that helps for n900 too Jun 07 10:09:24 I agree (I guess...) Jun 07 10:11:00 DocScrutinizer: creating correct amixer files is way out of my abilities Jun 07 10:11:28 DocScrutinizer: I can try to integrate them with fsoaudiod instead of using scenarios... but I have nfc of mixer controls Jun 07 10:11:35 and phones have quite a bunch of them :P Jun 07 10:13:00 mrmoku: the needed skills are no different from those needed for writing a scenario Jun 07 10:13:22 DocScrutinizer: I would have problems to write a correct scenario too Jun 07 10:13:39 but scenarios we can get by 'stealing' them from meego/maemo/whatever Jun 07 10:13:52 :nod: that's a task for the audio dudes, that know their hardware Jun 07 10:14:24 mrmoku: I honestly doubt maemo is working with any form of scenario Jun 07 10:14:28 DocScrutinizer: you're the only audio dude I know :-) Jun 07 10:14:46 well PaulFertser has probably good insight in audio hw too Jun 07 10:14:53 yep Jun 07 10:15:50 it's about as complex and a similar process as plugging up your effects chain for your guitar on stage Jun 07 10:16:26 hehe Jun 07 10:16:50 just the effect boxes are incredibly inferior wrt the manual shipping with them Jun 07 10:17:05 to those you can get for stage Jun 07 10:17:55 but then, if we don't get datasheets of the codecs, we're screwed anyway Jun 07 10:19:21 there are datasheets for n900 and gta02, right? Jun 07 10:19:36 if there's an alsa mixer def so you get meaningful alsamixer display, then there has to be some docs available out there as well Jun 07 10:19:43 yep Jun 07 10:21:58 in the end the process can be done like: get component X working, then flip every single switch of mixer forth and back, and eliminate all those that do nothing. The ones remaining are the ones to go to your .amix file Jun 07 10:23:35 so for mic you end with .amix containing controls A, C, H - while for earpiece those are B, D, E Jun 07 10:23:47 hmm, ok Jun 07 10:23:58 so the both are compatible path defs that don't conflict Jun 07 10:24:16 I.E. can get used concurrently Jun 07 10:24:32 that might be doable Jun 07 10:25:03 DocScrutinizer: hopefully there are no switches that can do any harm? :P Jun 07 10:25:05 for speaker you might find .amix containing C, X, Y, and it conflicts with mic on C Jun 07 10:25:42 means you can't use mic and speaker same time Jun 07 10:25:51 unless bot set C to same value Jun 07 10:25:56 both* Jun 07 10:26:01 yeah Jun 07 10:26:23 that's what fsoaudiod needs to implement Jun 07 10:26:43 preferrably on a known syntax of .amix files Jun 07 10:26:58 no, there's usually no harm Jun 07 10:27:35 ok Jun 07 10:30:13 of course I'll help with defining pathes wherever I can Jun 07 10:30:26 ok, great Jun 07 10:30:44 for GTA02 it's simple Jun 07 10:31:30 just go to the diagrams with the red lines, ans write down all the function blocks aka controls the line is touching Jun 07 10:32:34 for maemo I seem to recall a similar thaing to .amix, written for alsaped in some weird config syntax Jun 07 10:33:17 also we got schematics and mixed datasheet for N900 Jun 07 10:33:31 mixer* Jun 07 10:34:03 * DocScrutinizer burps loudly while thinking about digital filers in AIC34 Jun 07 10:34:41 seems the mixer controls for those are text boxes to enter hex, at best Jun 07 10:34:41 PaulFertser: any idea if andy is going to mind if I completely change the qi user interface? ;) Jun 07 10:35:45 DocScrutinizer: the red lines in here: http://wiki.openmoko.org/wiki/Neo1973_audio_subsystem ? Jun 07 10:35:58 * DocScrutinizer wonders if Amdy still is around *somewhere* Jun 07 10:36:10 yep, mrmoku Jun 07 10:36:56 ahh Jun 07 10:36:57 http://people.openmoko.org/joerg/ALSA/doc/WM8753_control_diag__gsmhandset_tx+rx-processed.png Jun 07 10:37:01 seems he *completely* abandoned OM domain and halo Jun 07 10:38:04 I guess that means he won't :) Jun 07 10:38:45 mrmoku: nice, isn't it? Jun 07 10:39:01 PaulFertser: I'm thinking that it'd work a lot more like u-boot. there wouldn't be a graphical menu but it would blink AUX N times when it is at boot entry N for example Jun 07 10:39:22 DocScrutinizer: indeed... just need to understand it :) Jun 07 10:39:27 PaulFertser: so that you could "stop" at the boot loader and think which entry you want to boot Jun 07 10:39:29 brb Jun 07 10:39:49 red is mic->modem path Jun 07 10:41:11 NB both the green and the red path are for a phonecall processed by CPU Jun 07 10:41:22 that's NOT what we normally got Jun 07 10:41:43 a very special scenario I've been asked for by the french guys Jun 07 10:42:04 any idea what they do with it? Jun 07 10:42:27 they *planned* to do an audio-level-crypt-phone Jun 07 10:42:37 which obviously never can work Jun 07 10:44:22 ok Jun 07 10:45:23 not that gta02 can't cook audio to any form of noise, and back. But GSM codec can't transmit that Jun 07 10:47:54 there has been a theoretical approach lately, to treat the GSM audiopath as a blackbox, and have a self-calibrating adaptive multichannel modem probing and adapting to that channel. It may get to a baudrate of like 1000bd when it's *very* good Jun 07 10:48:26 obviously too low for any speech codec to work realtime ofer that data channel Jun 07 10:48:36 over* Jun 07 10:49:42 (on the modem part) just like old 19k modems did Jun 07 10:49:55 or DSL modems do nowadays Jun 07 10:50:19 just more sophisticated Jun 07 10:50:58 still you never will get to more than 25% of the raw datarate the gsm codec itself is transferring data at Jun 07 10:51:25 on EFR that was 14.4kbaud Jun 07 10:51:55 9600baud on fullrate Jun 07 10:52:54 and 9600 is itself close to the limit where good codecs can transmit semi-decent speech Jun 07 10:53:50 you'll never reach that speed on transmitting beeps over GSM audio Jun 07 10:54:21 GSM is known to fail on 2400baud fax, even on 1200baud fax Jun 07 10:54:59 with a bit of luck you might get 300bd half duplex Jun 07 10:56:57 they didn't want to hear about those concerns Jun 07 10:57:21 so I gave them what they asked for, and this nice diagram is what resulted from it Jun 07 10:59:43 the light colors are local "uncrypted" audio, while the saturated colors are what's moving OTA via modem Jun 07 11:00:21 bluegreen is an alternative path not really needed for anything Jun 07 11:01:23 * mrmoku lunch and then back to alsamixer Jun 07 11:01:32 DocScrutinizer: maybe you can just use more than one freerunner in parallel :) Jun 07 11:01:43 the case is quite stackable :) Jun 07 11:02:03 hehe, indeed this *could* work Jun 07 11:02:54 but honestly you're better off getting a data service SIM with a true data enabled dial in number Jun 07 11:03:18 this way you can connect two phones at up to 14400bd Jun 07 11:05:00 this however wasn't a valid alternative for that cryptophone - dunno why Jun 07 11:06:10 you can do normal phonecalls with such a datacard as well. And you for sure can't do crypted calls to arbitrary far ends that are not prepared to receive those, so what's wrong with a datacard? Jun 07 11:07:56 where's the difference between "get freerunner and our audioscrambler" and "get freerunner and a ZRTP-enapled VoIP sw, and a datacard" Jun 07 11:08:15 only diff: #2 works, while #1 is a hoax Jun 07 11:10:18 meh, seems everything gets recycled *eventually* :-D Jun 07 11:10:44 even such a nice useless audio routing diagram Jun 07 11:12:25 :P Jun 07 11:15:45 amyway this paper shows 4 distinct audio pathes, of which gsm->AD_PCM and mic->AD_PCM share one control (AD obviously), and PCM_DA->gsm and PCM_DA->speaker/hp share the DA. (btw tha last one are in fact 2 pathes PCM->headphones and PCM_DA->speakers, maybe a 3rd one PCM_DA->earpiece - I'd have to check, but...) Jun 07 11:17:13 so might as well be 6 paths Jun 07 11:18:32 mrmoku: I hope I gave you a bunch of stuff to ponder about with my monologue ;-D Jun 07 11:19:09 * DocScrutinizer waves Jun 07 11:19:31 DocScrutinizer: http://pastebin.com/UuYJZgnX Jun 07 11:19:36 mrmoku: ping me by name if any more questions Jun 07 11:20:06 DocScrutinizer: those are the controls that influence stereoout speaker as far as my ears can tell :P Jun 07 11:20:28 sounds reasonable Jun 07 11:20:31 good Jun 07 11:21:52 while the equalizer settings are arguable if they belong into a path or not, maybe depends on usecase, e,g, for cal_via_earpiece they may well belong in there, while for dunno what stereo-playback_speakers they probably don't Jun 07 11:22:13 yeah, I will ignore them for now Jun 07 11:22:25 fsoaudiod does not have any api for that anyway Jun 07 11:22:52 you usually access those interactively via arbitrary mixer apps Jun 07 11:23:12 yup Jun 07 11:23:37 fsoaudiod could answer queries about which auxiliary controls belong to a path/device Jun 07 11:24:14 for that you could comment them out in a special way, in .amix Jun 07 11:24:29 amixer accepts comments Jun 07 11:24:51 btw. I'm planning to use alsa lib directly for setting the controls Jun 07 11:24:53 actually it seems to ignore all bogus crap Jun 07 11:25:10 makes no sense to call amixer when we already have nice alsa bindings I think Jun 07 11:25:17 I don't care, as long as you keep the file format as compatible as possible Jun 07 11:25:34 well amix format looks quite reasonable Jun 07 11:25:46 and actually similiare to what we have as current scenario files Jun 07 11:25:58 just with an additional cset in front Jun 07 11:26:24 instead of the control # Jun 07 11:26:40 there's also a simplified form, along set and `amix controls` Jun 07 11:27:10 reminds me. I should update http://lindi.iki.fi/lindi/openmoko/amixer2.txt to match 2.6.34 alsa control names I guess Jun 07 11:27:14 amixer can do both "simple" and complex controls Jun 07 11:28:09 sset Master 50 Jun 07 11:28:30 works as well Jun 07 11:30:20 list names with `amixer scontrols` Jun 07 11:30:50 while `amixer` lists complex controls Jun 07 11:31:55 actually `amixer controls` lists complex Jun 07 11:32:59 I got 10 simple and 15 complex controls here Jun 07 11:33:17 (on my PC) Jun 07 11:34:03 seems switches that also have a volume aren't listed under simple Jun 07 11:34:34 kinda makes sense Jun 07 11:34:45 NB names are also stripped Jun 07 11:35:17 maybe you should ignore the whole simple stuff Jun 07 11:36:18 just keep it in mind, and consider it if it turns out it's not any additional complexity in code to include simple syntax Jun 07 11:37:11 embedd additional (meta)info in "# commentlines" Jun 07 11:37:16 freesmartphone.org: 03morphis 07aurora * r2a746233f4eb 10/aurora/aurora/extensions/agents.py: aurora: use class methods and not non existing global methods Jun 07 11:37:22 ala shebang Jun 07 11:37:56 freesmartphone.org: 03morphis 07cornucopia * r9fa98379c472 10/fsogsmd/src/plugins/modem_qualcomm_palm/ (phonebookhandler.vala unsolicited.vala): (log message trimmed) Jun 07 11:37:56 freesmartphone.org: fsogsmd: modem_qualcomm_palm: fix several critical bugs in modem plugin Jun 07 11:37:56 freesmartphone.org: * When connecting again after a closed connection to the fsogsmd daemon the dbus signal Jun 07 11:37:56 freesmartphone.org: handlers for the msmcomm daemon gets connected again. This leads to handling the Jun 07 11:37:56 freesmartphone.org: incomming signals n times (n = how often a client has reconnected). Jun 07 11:37:56 freesmartphone.org: * When a client reconnects after a closed connection the phonebook handler should not be Jun 07 11:37:57 freesmartphone.org: reinitialized just clean it up! Jun 07 11:37:57 freesmartphone.org: 03morphis 07cornucopia * r98e43a20b597 10/fsogsmd/src/lib/modem.vala: fsogsmd: modem: convert tabs to spaces Jun 07 11:37:58 freesmartphone.org: 03morphis 07cornucopia * r59b3c852a272 10/fsogsmd/src/plugins/modem_qualcomm_palm/ (channel.vala phonebookhandler.vala): fsogsmd: modem_qualcomm_palm: a lot of cleanups for the modem initialization code Jun 07 11:37:58 freesmartphone.org: 03morphis 07cornucopia * r83db748f2a8c 10/fsogsmd/src/plugins/modem_qualcomm_palm/plugin.vala: fsogsmd: modem_qualcomm_palm: remove unused name of at channel Jun 07 11:38:37 freesmartphone.org: 03morphis 07msmcomm * r6599dc245b1c 10/libmsmcomm/ (4 files in 4 dirs): libmsmcomm: rename field from data -> packet_data Jun 07 11:38:38 freesmartphone.org: 03morphis 07msmcomm * r085543b08979 10/libmsmcomm/msmcomm/ (4 files): libmsmcomm: implement first gps relevant message Jun 07 11:38:38 freesmartphone.org: 03morphis 07msmcomm * r181cccc24c50 10/msmcomm-specs/src/ (Makefile.am gps.vala): msmcomm-specs: add gps interface specification Jun 07 11:38:39 freesmartphone.org: 03morphis 07msmcomm * r79886c9338b0 10/libmsmcomm/msmcomm/ (pdsmmessage.vala pdsmmessagegroup.vala): libmsmcomm: some namespace modifications for pdsm messages Jun 07 11:38:39 freesmartphone.org: 03morphis 07msmcomm * r310666037dac 10/msmcommd/src/ (Makefile.am dbusservice.vala gpsservice.vala): msmcommd: implement GPS service class Jun 07 11:39:21 #!fsoaudiod aux_class=EQ Jun 07 11:39:39 treble Jun 07 11:39:46 or even Jun 07 11:41:02 heyho Jun 07 11:41:49 #! iface=MIXER,name='Headphone Playback Switch' Jun 07 11:42:58 also you might want to include a #!fsoaudiod path_priority=3000 Jun 07 11:44:44 #!fsoaudiod system='echo \'hello, this is "debug"" print\'' Jun 07 11:46:15 mrmoku: obviously, when one path gets 'closed', fsoaudiod should restore whatever been there before Jun 07 11:47:57 morphis: hey Jun 07 11:48:21 thus re-establishing previous "scenario" and also running whatever been defined for signalling "continue" to any app that got "resource_not_available" msg to pause Jun 07 11:49:03 ok Jun 07 11:49:34 so you'll need some form of stack Jun 07 11:49:57 if only for v2.0 Jun 07 11:50:01 :-) Jun 07 11:50:02 :) Jun 07 11:50:07 yeah, first get something rolling Jun 07 11:50:37 get sth rolling on the wrong whelels will cause headache later Jun 07 11:51:40 sure, but on the other hand it has to evolve Jun 07 11:51:43 that's why I metion as much as my brain wants to recall, even if we don't need it *right now* Jun 07 11:52:40 design it with stack in mind, and you'll maybe have a completely different structure Jun 07 11:52:53 from beginning Jun 07 11:54:01 keep metacomments in mind and you won't do any direct piping of file->whatever, without parsing it at least by a no-op function Jun 07 11:54:10 mrmoku: you are still working on fsoaudiod? Jun 07 11:55:06 morphis: I started ranting about scenarios < .amix path defs yesterday Jun 07 11:55:56 amix path defs? Jun 07 11:56:09 morphis: usual case is you get 2 or more audiodevices open concurrently (mic + earpiece) Jun 07 11:56:18 yep, amix path defs Jun 07 11:56:19 morphis: I restarted to work on the alsa-hooks thing and router Jun 07 11:56:35 we have damn big problems on gta02 Jun 07 11:56:40 mrmoku: so you started to implemented the streamcontrol_alsa plugin? Jun 07 11:56:43 mrmoku: why? Jun 07 11:56:49 morphis: (2 audiodev) that's why "scenario" won't fly Jun 07 11:56:54 streamcontrol_alsa plugin? Jun 07 11:57:14 mrmoku: the thing we need for the alsa-hooks integration within fsoaudiod Jun 07 11:57:20 DocScrutinizer: ah Jun 07 11:57:53 morphis: read backscroll, search for people.openmoko Jun 07 11:59:30 morphis: then yes :) I'm finding out right now which controls influence what device Jun 07 11:59:33 on gta02 Jun 07 11:59:49 morphis: see http://people.openmoko.org/joerg/ALSA/ACI/dot_amix/ Jun 07 12:00:04 mrmoku: ah as we have so much controls on the om-gta02 Jun 07 12:00:26 and http://people.openmoko.org/joerg/ALSA/ACI/ACI_spec/gsm_earpiece__readme Jun 07 12:01:35 DocScrutinizer: so amix is similar to a scenario file? Jun 07 12:01:45 mrmoku: I hope you're not hardcoding these "which control for which device" bits Jun 07 12:02:03 amix is more finegrained Jun 07 12:02:11 DocScrutinizer: what do you mean by hardcoding? Jun 07 12:02:11 ok Jun 07 12:02:20 and btw can also do relative changes Jun 07 12:02:34 that's good to know :) Jun 07 12:02:48 like "volume=volume-10" Jun 07 12:03:38 The controls influencing the headset speakers are quite the same as the stereoout speaker ones Jun 07 12:04:03 mrmoku: you also want to have $VARs for particular controls' values, e.g. the volume :-) Jun 07 12:05:07 mrmoku: why we can't use the already available scenario files for the om-gta02? Jun 07 12:05:08 so user can change volume, and fsoaudiod can save that to a separate storage of name VAR, then reference it in .amix file (which is kinda r/o) by $VAR Jun 07 12:05:39 morphis: because they mix things Jun 07 12:06:22 morphis: see http://people.openmoko.org/joerg/ALSA/ACI/dot_amix/ Jun 07 12:07:25 those should be complete path specs, and mustn't get any larger Jun 07 12:11:03 i.e. http://people.openmoko.org/joerg/ALSA/ACI/dot_amix/gsm_earpiece.amix shall set up all that's needed to make modem -> earpiece path, and you can't load a second scenario ruining this setup when you decide e.g. to use headset mic for the outbound gsm audio Jun 07 12:12:38 mrmoku: I know - but honestly, I can do this in 1/10 of the time you need, so I suggest you do coding, and I care about path defs needed Jun 07 12:13:00 for coding you don't need a single real pathdef right now Jun 07 12:13:41 you just need to know how ro read files like http://people.openmoko.org/joerg/ALSA/ACI/dot_amix/mic2_gsm.amix Jun 07 12:14:46 btw these are *very* special pathes as they're not linked to a real alsa device in any obvious way Jun 07 12:15:33 DocScrutinizer: hehe, right you are... I'm very slow finding my way through there :P Jun 07 12:16:04 DocScrutinizer: http://paste.pocoo.org/show/402127/ Jun 07 12:16:11 is what are my findings so fare Jun 07 12:16:13 -e Jun 07 12:16:18 as this will quite probably change for other devices anyway (e.g. N900 it really id different), I suggest we use virtual device that does nothing, but threat it as if we'd do a normal audio Jun 07 12:18:08 mrmoku: you could compare to my path defs in .amix - btw a path always has 2 endpoints :-) Jun 07 12:19:02 your findings will differ quite a bit for mic->AD_PCM vs mic->gsm Jun 07 12:21:10 DocScrutinizer: ok, my findings are the path from mplayer playback to the respective speaker :) Jun 07 12:21:44 then I wonder why the earpiece and headset have no EQ bits Jun 07 12:21:54 because I left them out there Jun 07 12:21:58 aaah Jun 07 12:22:53 well, as mentioned I suggest you forget about that part as I can do it more easily than you, and you go for coding rather. Jun 07 12:23:02 ok Jun 07 12:23:12 Just realized you need a per-control priority Jun 07 12:24:52 some device might have low prio 500 on ctl A and prio 8888 on ctl B, while another concurrent path has high prio 1000 on A but also 1000 on all other controls Jun 07 12:25:41 duh, getting complicated now :P Jun 07 12:26:33 a path can override another paths controls only when all of the conflicting controls of new path have higer prio than the corresponding controls of any established path Jun 07 12:27:10 if a path fails on allocating one control, the whole path fails Jun 07 12:51:28 (complicated) that's why we need such a daemon :-) Otherwise we could do all this with native ALSA Jun 07 12:52:11 on a related sidenote: what's the policy about inbound calls during a video-recording? Jun 07 12:52:45 whatever it is, we schall and will be able to handle it in ALSA/fsoaudiod Jun 07 12:52:50 shall* Jun 07 12:53:57 I could image calls via headset are allowed while calls via handset are a nogo Jun 07 12:54:06 imagine* Jun 07 12:54:31 just for the audio aspect of it Jun 07 12:54:52 * mrmoku wonders how video recording pans out on gta02 :P Jun 07 12:55:23 mrmoku, what's wrong with the quick settings screen? Jun 07 12:55:38 (actually, I may have hardcoded some stuff there because there's no text size sync yet) Jun 07 12:59:49 mrmoku: we're not designing for a particular platform, are we? Jun 07 13:01:19 TAsn: they're wrong on horizontal displays Jun 07 13:01:31 the top buttons are almost not visible Jun 07 13:01:42 DocScrutinizer: sure... was j/k Jun 07 13:01:52 mrmoku, just make everything relational. Jun 07 13:02:06 and set min sizes Jun 07 13:02:08 which I didn't Jun 07 13:02:15 ok, will try that Jun 07 13:02:37 though I think for some screens we might want to have a completely different layout Jun 07 13:04:46 yeah Jun 07 13:04:50 that's actually Jun 07 13:04:54 one of the biggest issues with Jun 07 13:04:59 elementary atm Jun 07 13:05:07 is that we didn't yet got to doing Jun 07 13:05:16 port/land handling Jun 07 13:05:33 * DocScrutinizer steals TAsn 's return key Jun 07 13:06:15 does extracting the rootfs image in sd card is enough ? Jun 07 13:06:33 TAsn: but there are plans to do something for it? Jun 07 13:06:40 then I can as well wait :P Jun 07 13:09:39 lxsameer, what device? Jun 07 13:10:26 GNUtoo: bravo Jun 07 13:10:47 lxsameer, you need to fastboot your kernel Jun 07 13:10:57 not to flash but do somehting like that: Jun 07 13:11:02 GNUtoo: i have that, but about rootfs Jun 07 13:11:15 GNUtoo: what does that jiff file Jun 07 13:11:15 about rootfs, just extract in in a microsd Jun 07 13:11:16 but Jun 07 13:11:28 extract it with a card reader Jun 07 13:11:34 not trough the phone's usb Jun 07 13:11:41 else it won't work Jun 07 13:11:47 GNUtoo: yeah i do that with my laptop sd reader Jun 07 13:11:52 ok Jun 07 13:11:57 GNUtoo: but why? Jun 07 13:12:01 did you mean jffs2? Jun 07 13:12:10 because the partitions are not exposed the same way Jun 07 13:12:20 GNUtoo: what is the jiff2 Jun 07 13:12:29 another filesystem for flash Jun 07 13:12:33 forget about it for now Jun 07 13:12:36 it won't work anyway Jun 07 13:12:45 GNUtoo: aha ok thanks Jun 07 13:12:47 basically for the partitions: Jun 07 13:12:54 GNUtoo: one more to go Jun 07 13:12:55 mmcblk0 Jun 07 13:12:58 mmcblk0p1 Jun 07 13:13:10 the second one only is exposed trough usb Jun 07 13:13:17 GNUtoo: i just have a partition with ext3 is it ok ? Jun 07 13:13:20 and you need the first one for formatting in ext3 Jun 07 13:13:21 yes Jun 07 13:13:42 personally I boot like that: Jun 07 13:13:49 fastboot -c 'root=/dev/mmcblk0p1 rw rootwait noinitrd fbcon=rotate:1' boot /home/gnutoo/embedded/kernel/kernel.org/arch/arm/boot/zImage Jun 07 13:14:14 GNUtoo: thanks man let me try Jun 07 13:17:43 GNUtoo: so i have to use mmcblk0p1 when i'm booting through usb and mmcblk0 when i'm booting with flashing Jun 07 13:17:59 no Jun 07 13:18:13 when you partition you do: Jun 07 13:18:17 fdisk /dev/mmcblk0 Jun 07 13:18:27 and you create /dev/mmcblk0p1 with the menus Jun 07 13:18:31 and save it Jun 07 13:18:42 and not forget to put linux as partition type Jun 07 13:18:50 and then format in ext3 Jun 07 13:18:50 GNUtoo: hmmm why mmcblk ? why not sdx1 Jun 07 13:19:06 on your computer it could be sdx1 Jun 07 13:19:09 no idea Jun 07 13:19:18 GNUtoo: aha , you mean in my device Jun 07 13:19:21 but on bravo it will be mmcblk0p1 Jun 07 13:19:41 maybe I should restart explaining from scratch Jun 07 13:19:45 I explained too much Jun 07 13:19:49 and now you're confused Jun 07 13:20:03 GNUtoo: i would be great Jun 07 13:20:07 ok Jun 07 13:20:13 so you've a microsd Jun 07 13:20:23 what is it when you plug it in your computer? Jun 07 13:20:28 how did you format it? Jun 07 13:20:43 with gparted or trough the command line? Jun 07 13:20:45 GNUtoo: sdx and sdx1 and i do a cfdisk on sdx Jun 07 13:21:05 and create a msdox partition table and just one partition Jun 07 13:21:12 ok Jun 07 13:21:15 GNUtoo: then format it with mkfs.ext3 Jun 07 13:21:19 I don't know cfdisk sorry Jun 07 13:21:19 ok Jun 07 13:21:54 GNUtoo: its a great tool like fdisk but with a ncurses interface Jun 07 13:21:57 :-O Jun 07 13:22:07 yes I just read about it Jun 07 13:22:11 I use fdisk usually Jun 07 13:22:14 Okay, I'm here. (courtesy of docscrutinizer) Jun 07 13:22:54 GNUtoo: whats next Jun 07 13:22:56 cfdisk >>>> fdisk Jun 07 13:23:17 * GNUtoo is used to fdisk Jun 07 13:23:57 well, others are used to pencil and paper, so refuse to use PC and editor Jun 07 13:24:28 each man on his own preferences Jun 07 13:24:46 DocScrutinizer, already lost it :P Jun 07 13:24:59 why is fdisk not ok? Jun 07 13:25:17 GNUtoo: fdisk is ok too, but i used to cfdisk Jun 07 13:25:20 mrmoku, plans: not really, I know some of the other cellphone efl users implemented something on their own. Jun 07 13:25:29 fdisk is ok, but cfdisk is way easier to use, and has better edit methods Jun 07 13:25:34 mrmoku, but it's fairly easy to implement. Jun 07 13:25:41 lxsameer, what's the partition type, linux or fat32? Jun 07 13:25:47 mrmoku, just create two themes for apps that need them Jun 07 13:25:58 mrmoku, one called land the other port :) Jun 07 13:26:00 GNUtoo: i format it with mkfs.ext3 so linux Jun 07 13:26:05 DocScrutinizer, no manual resize2fs I guess Jun 07 13:26:12 mrmoku, c'est tout Jun 07 13:26:19 lxsameer, in cfdisk can you choose the partition type? Jun 07 13:26:28 DocScrutinizer, You know I've never been a big fan of the return key :) Jun 07 13:26:36 GNUtoo: yeah Jun 07 13:26:42 GNUtoo: nope, resize2fs isn't supported in cfdisk Jun 07 13:27:13 just nice (semi-+)automatic calculation of partition sizes Jun 07 13:27:22 comprehensive listing Jun 07 13:27:44 GNUtoo: describe it in your own way Jun 07 13:28:12 GNUtoo: sure you can choose parttype, evben get a nice helplist Jun 07 13:28:17 DocScrutinizer, you can set size without a calculator in fdisk too Jun 07 13:28:31 lxsameer, what did you put as partition type? Jun 07 13:28:49 GNUtoo: linux Jun 07 13:29:02 ok nice Jun 07 13:29:03 number? Jun 07 13:29:07 GNUtoo: i just want to repartition it with fdisk Jun 07 13:29:24 no cfdisk is just fine Jun 07 13:29:30 if you choose linux it's ok Jun 07 13:29:39 then you did mkfs.ext3 on it Jun 07 13:29:47 then you have to extract the tarball like that Jun 07 13:29:52 mount the microsd Jun 07 13:29:57 go inside the mounted dir Jun 07 13:29:58 and do Jun 07 13:30:07 tar xvjpf /path/to/tarball.tar.bz2 Jun 07 13:30:30 then umount it Jun 07 13:30:42 put it in the phone Jun 07 13:30:46 and boot with fastboot Jun 07 13:31:28 GNUtoo: does the tarball name is some thing like shr-full-eglibx...............? Jun 07 13:31:38 yes Jun 07 13:31:41 it's that tarball Jun 07 13:31:50 letme try again using fdisk Jun 07 13:32:03 no need Jun 07 13:32:07 you did it just fine Jun 07 13:32:34 why? Jun 07 13:32:38 something is not working? Jun 07 13:33:10 GNUtoo: but i ried 3 kernel that known to work in bravo but every kernel stick in alpharev splashscreen and i think there is something with rootfs Jun 07 13:33:15 TAsn: yeah, that much I would have figured out myself :P Jun 07 13:33:29 lxsameer, what's apharev? Jun 07 13:33:44 normally you should have an SHR splashscreen that is moslty garbage Jun 07 13:34:01 if you don't have that Jun 07 13:34:05 recompile your kernel Jun 07 13:34:12 and remove initramfs from it Jun 07 13:34:17 and boot with noinitramfs Jun 07 13:34:22 and also add framebuffer console Jun 07 13:34:27 to see debug message Jun 07 13:34:45 GNUtoo: a patch for hboot that allow booting native kernel Jun 07 13:35:05 ??? Jun 07 13:35:11 native kernel Jun 07 13:35:12 GNUtoo: alpharev.nl Jun 07 13:35:17 what do you mean by that? Jun 07 13:35:31 GNUtoo: removing initramfs is a good idea Jun 07 13:35:38 what's the difference between an android and a native kernel? Jun 07 13:35:48 GNUtoo: but i think it does not enter to boot process Jun 07 13:35:48 yes you should boot without initramfs Jun 07 13:35:51 else it doesn't work Jun 07 13:37:26 and also don't boot the .fastboot yet Jun 07 13:37:30 only boot the zImage Jun 07 13:37:34 GNUtoo: thanks man you thought me good things Jun 07 13:37:35 or the zImage*.bin Jun 07 13:37:57 GNUtoo: hmm my image is Image..... Jun 07 13:38:05 that's bad Jun 07 13:38:07 won't work Jun 07 13:38:16 change to zImage in openembedded Jun 07 13:38:22 GNUtoo: why? Jun 07 13:38:56 because Image is not supported by fastboot Jun 07 13:39:15 Image was for the windows ce bootloader called haret if I understood well Jun 07 13:39:24 GNUtoo: aha good , it seems that finally i found the problem :P Jun 07 13:39:30 the bootloader has two variables: rootfs partition, and name of kernel Jun 07 13:39:46 indeed Jun 07 13:39:56 that's why I told to add support for framebuffer console Jun 07 13:40:07 GNUtoo: does that z have relation with compress algorithm of image ? Jun 07 13:40:21 basically it works like that Jun 07 13:40:21 historically yes Jun 07 13:40:25 zImage is an executable Jun 07 13:40:31 computationally no Jun 07 13:40:33 with arm assembly at the beginning Jun 07 13:40:52 that arm assembly with initialize the CPU and decompress the rest Jun 07 13:41:04 no idea how Image work Jun 07 13:41:16 I only know uImage and zImage and bzImage Jun 07 13:41:27 bzImage is big zImage btw Jun 07 13:41:30 it's not bz2 Jun 07 13:41:39 GNUtoo: aha thanks man let me try to recompile the kernel with gz algorithm Jun 07 13:41:56 lxsameer,how do you compile your kenrel? Jun 07 13:42:00 errrr what? Jun 07 13:42:03 because usually it's Jun 07 13:42:06 make uImage Jun 07 13:42:08 make zImage Jun 07 13:42:13 make Image Jun 07 13:42:32 DocScrutinizer, did I say something wrong? Jun 07 13:42:42 GNUtoo: bitbake virtual/kernel Jun 07 13:42:54 lxsameer, ok then change the image type in your machine config Jun 07 13:42:55 nah >> let me try to recompile the kernel with gz algorithm Jun 07 13:43:00 ok Jun 07 13:43:24 GNUtoo: ok Jun 07 13:43:33 DocScrutinizer: whats wrong with it? Jun 07 13:43:49 the gz algo isn'z inside kernel Jun 07 13:43:58 DocScrutinizer: i see Jun 07 13:44:15 DocScrutinizer: i mean the algorithm for compressing the kernel image Jun 07 13:44:18 the kernel gets compressed after it's built Jun 07 13:44:25 yep Jun 07 13:44:29 DocScrutinizer: in general setup of kernel configuration Jun 07 13:44:57 sorry, I#m a bit tired, so didn't get it instantly Jun 07 13:45:10 lxsameer, basically you used qsd8.inc Jun 07 13:45:11 DocScrutinizer: ;) Jun 07 13:45:20 just override the kernel image type in your machine config Jun 07 13:45:29 DocScrutinizer: drink a hot tee Jun 07 13:45:33 GNUtoo: yeah Jun 07 13:45:34 just put that after the require qsd8.inc: Jun 07 13:45:45 GNUtoo: sure man Jun 07 13:45:46 coffe, just empty Jun 07 13:46:06 DocScrutinizer: ah coffee , let me make my self one cup :D Jun 07 13:46:13 KERNEL_IMAGETYPE = "zImage" Jun 07 13:47:22 btw that explain how to make a .fastboot: Jun 07 13:47:22 http://blog.mitsutaka.org/2010/11/meego-11-for-nexusone-status-part2-and.html Jun 07 13:47:30 for nexusone Jun 07 13:47:38 so look for something similar for desire Jun 07 13:47:52 GNUtoo: i know about fastboot and mkbootimg Jun 07 13:48:06 mrmoku, :) Jun 07 13:48:16 mrmoku, so it's really not too crazy to handle :) Jun 07 13:49:50 yes but I pointed to it for a reason Jun 07 13:49:55 the details.... Jun 07 13:49:58 for instance: Jun 07 13:50:01 --base 0x20000000 Jun 07 13:50:03 or: Jun 07 13:50:06 noinitrd Jun 07 13:50:08 etc... Jun 07 13:50:13 GNUtoo: so i'll surely read it Jun 07 13:58:48 GNUtoo: but what is the --base and that address ? Jun 07 13:59:23 no idea Jun 07 13:59:31 maybe it's the baseaddress Jun 07 13:59:46 like 0x80008000 for uboot Jun 07 13:59:50 or something like that Jun 07 14:00:07 GNUtoo: hmm Jun 07 14:00:19 lxsameer, btw learn abuout the porting guide I'm writing Jun 07 14:00:45 http://www.shr-project.org/trac/wiki/Device%20Porting%20Guide Jun 07 14:01:26 GNUtoo: nice , sure Jun 07 14:25:21 GNUtoo: is it necessary to have fastboot kernel for fastbootutility or we can use a bin kernel with noinitrd boot parameter ? Jun 07 14:25:51 lxsameer, bin is ok but you cannot flash the .bin Jun 07 14:26:08 and bin is harder to use for users, but at that point it's not a problem Jun 07 14:26:16 you just want to start porting Jun 07 14:26:25 not to worry about flashability or users Jun 07 14:27:03 GNUtoo: aha, i made some progress, my kernel now cause some boot loop :D Jun 07 14:27:12 ah? Jun 07 14:27:35 GNUtoo: i really don't know where is my problem, kernel ? rootfs? bootloader or what :( Jun 07 14:28:04 what does it do now exactly? Jun 07 14:29:10 GNUtoo: after i use fastboot to boot a kernel it stock in a splashscreen of that alpharev tool, but about last one it enter to my andoird slash screen like loop and restarted to recovery mode Jun 07 14:29:23 did you follow all my advises? Jun 07 14:29:28 for instance: Jun 07 14:29:35 * remove initramfs from the knerel Jun 07 14:29:41 * boot with noinitrd Jun 07 14:29:44 * boot the .bin Jun 07 14:29:49 * add framebuffer console Jun 07 14:29:49 GNUtoo: yeah Jun 07 14:29:55 hmmm Jun 07 14:30:04 GNUtoo: last one was booting bin with noinitrrf Jun 07 14:30:06 note that captainigloo knows how to boot your phone Jun 07 14:30:11 he has the same Jun 07 14:30:23 so retry the one that booted Jun 07 14:31:12 GNUtoo: ow nice, Jun 07 14:31:16 captainigloo: ping Jun 07 14:31:42 GNUtoo: do you know about the secuflag Jun 07 14:32:08 no what's that? Jun 07 14:33:56 GNUtoo: thats why i use alpharev Jun 07 14:34:02 GNUtoo: http://alpharev.nl/ Jun 07 14:34:12 GNUtoo: first paragraph Jun 07 14:35:22 ok Jun 07 14:37:44 lxsameer: pong Jun 07 14:38:46 captainigloo: how do you boot a kernel with bravo Jun 07 14:39:37 lxsameer: i wrote a quick tutorial on the meego wiki Jun 07 14:39:40 i give you the link Jun 07 14:40:02 lxsameer: http://wiki.meego.com/ARM/MSMQSD Jun 07 14:40:15 lxsameer: look at the htc desire/bravo part Jun 07 14:40:23 captainigloo: thanks man Jun 07 14:40:35 HTC Desire Meego 1.1 WIP Jun 07 14:42:03 lxsameer: and here the config for a kernel based on cyanogenmod kernel : http://www.deco-crea.com/files/bravo_nico_defconfig Jun 07 14:42:21 captainigloo: thanks man Jun 07 14:42:28 captainigloo: did you boot it your self? Jun 07 14:42:38 yes by hand with fastboot Jun 07 14:42:48 sudo ./fastboot -c 'root=/dev/mmcblk0p1 rw rootfs=ext3 init=/sbin/init rootwait noinitrd' boot zImage Jun 07 14:43:16 IIRC fastboot can be build with shr/oe Jun 07 14:43:28 but i think that i find it in the htc-linux.org website Jun 07 14:43:38 can't remember Jun 07 14:43:51 captainigloo: and can i have your machin and kernel recipes file? Jun 07 14:44:02 i'm using htc-leo machine Jun 07 14:44:10 it's building fine Jun 07 14:44:29 the only thig i had to change is the freemsmartphone configuration path Jun 07 14:44:51 you need to change leo by bravo in /etc/freesmartphone/config Jun 07 14:44:54 or something like that Jun 07 14:45:11 dcordes: build a rootfs for me some time ago Jun 07 14:45:22 based on his htc-leo config Jun 07 14:45:43 lxsameer: if you want i can give you this image Jun 07 14:45:54 captainigloo: rootfs image ? Jun 07 14:45:57 yes rootfs Jun 07 14:46:26 you just need to untar it on the sdcard formatted with ext3 Jun 07 14:47:01 and i can also give you a working kernel :) Jun 07 14:47:18 captainigloo: i have the rootfs Jun 07 14:47:29 captainigloo: but i'll be happy to have kernel image Jun 07 14:47:34 ok i upload it Jun 07 14:47:43 captainigloo: i can't understand about /etc/freesmartphone/config Jun 07 14:48:31 http://www.deco-crea.com/files/zImage.10.2.6.37.cm-kernel Jun 07 14:48:34 here it is Jun 07 14:48:44 let me check i will give you the correct path Jun 07 14:50:10 lxsameer: ok so you need to copy/mv/link whatever the directory Jun 07 14:50:28 /etc/freesmartphone/conf/htcleo with /etc/freesmartphone/conf/bravo Jun 07 14:50:55 captainigloo: the image link is broken Jun 07 14:51:31 image link ? Jun 07 14:51:41 captainigloo: yeah Jun 07 14:52:12 i don't understand Jun 07 14:52:15 wich image link ? Jun 07 14:52:31 http://www.deco-crea.com/files/zImage.10.2.6.37.cm-kernel Jun 07 14:53:32 it works here :( Jun 07 14:54:02 captainigloo: ok letme recheck Jun 07 14:54:30 captainigloo: i read at your article to wait a very long time, how long is that? Jun 07 14:54:42 captainigloo, hi, will you integrate your work in openembedded at some point? Jun 07 14:54:43 or in oe-core Jun 07 14:54:52 *shr-core Jun 07 14:55:10 GNUtoo: yes i should Jun 07 14:55:20 i need to take the time to do something clean Jun 07 14:55:34 i'm palying a bit with oe-core Jun 07 14:55:44 so it's maybe the place to send my work Jun 07 14:56:00 i see there is meta-htc directory in oe-core Jun 07 14:56:16 i guess it's the best place to add my work ? Jun 07 14:56:49 lxsameer: when i start to try linux on my desire maybe january this year Jun 07 14:57:15 ok Jun 07 14:57:21 yes Jun 07 14:57:29 I'm still with oe.dev personally Jun 07 14:57:34 since I've spaces issues Jun 07 14:57:48 captainigloo: no i mean how long does it take to boot a kernel at desire for the first time? Jun 07 14:57:56 I've 1TB and it is not enough for both oe and oe-core Jun 07 14:57:59 ah sorry :) Jun 07 14:58:09 and according to mrmoku shr-core is not ready yet totally Jun 07 14:58:09 lxsameer: about 1 min Jun 07 14:58:12 still some issues Jun 07 14:58:23 GNUtoo: yes Jun 07 14:58:25 captainigloo: aha Jun 07 14:58:38 lxsameer maybe 20second with a black screen Jun 07 14:58:51 and after that you should have shr splash screen Jun 07 14:58:55 captainigloo: when i try to boot a kernel, i stock in the alpharev slashscreen , do you have any idea Jun 07 14:58:58 captainigloo, did you work on your keyboard? Jun 07 14:59:19 lxsameer: yes maybe i five you a wrong kernel Jun 07 14:59:22 give* Jun 07 14:59:33 captainigloo: no i did not try your kernel yet Jun 07 14:59:51 captainigloo: i mean my own kernels ( about 6 build on cm and google one) Jun 07 15:00:49 hum Jun 07 15:01:32 lxsameer: did you change the options described in the meego wiki ? Jun 07 15:01:49 i meean remove initramfs Jun 07 15:01:55 captainigloo: no not yet i'm working on it Jun 07 15:01:58 and set Virtual Terminal Jun 07 15:02:05 captainigloo: yeah i remove that already Jun 07 15:02:15 I'll go soon Jun 07 15:02:32 GNUtoo, captainigloo: at least the n900 image I built with shr-core two or three days ago was not usable Jun 07 15:02:34 GNUtoo: no i don't play with keyboard Jun 07 15:03:03 GNUtoo: i'm wokring on enna for tablet : http://www.deco-crea.com/files/2011-06-04-231438_1024x600_scrot.png Jun 07 15:03:09 captainigloo, for nexusone and bravo and hd2 etc....there is the need of an better keyboxrd Jun 07 15:03:20 you did that: Jun 07 15:03:20 GNUtoo: yes i know for tablet too Jun 07 15:03:44 http://scap.linuxtogo.org/files/097b24c92e7b683d59d819a9ba600975.png Jun 07 15:03:52 and the youtube link that morphis send here some days ago with gnome3 virtual keyboard is really good Jun 07 15:04:06 ah? Jun 07 15:04:15 the efl keyboard you did is nice Jun 07 15:04:24 but it is unusable in console Jun 07 15:04:27 it lacks some keys Jun 07 15:04:28 captainigloo: you kernel image works man Jun 07 15:04:37 like - @ | etc... Jun 07 15:04:40 captainigloo: can i have your config file for it Jun 07 15:04:56 lxsameer: http://www.deco-crea.com/files/bravo_nico_defconfig Jun 07 15:05:05 for instance at the left of space a key shifter could be added Jun 07 15:05:10 captainigloo: ah that one Jun 07 15:05:22 captainigloo: thanks alot man , i have some hop now :D Jun 07 15:05:30 GNUtoo: yes i know, i will try to find some time to work on it Jun 07 15:05:36 lxsameer: :) Jun 07 15:05:41 you're welcome Jun 07 15:05:47 thanks a lot Jun 07 15:05:53 I'll go now Jun 07 15:05:55 ok Jun 07 15:05:59 have a nice day Jun 07 15:06:03 thanks Jun 07 15:07:50 captainigloo: does that config is for team desire that host on github ? Jun 07 15:12:41 lxsameer: yes this one : https://github.com/teamdesire/cm-kernel.git Jun 07 15:13:37 captainigloo: thats it thanks Jun 07 15:40:55 captainigloo: ping Jun 07 16:08:14 lxsameer: pong Jun 07 16:08:52 captainigloo: what does that arm-angstrom-linux-gnueab- tool? Jun 07 16:10:02 does any one tried Debian APT on shr before? Jun 07 16:11:20 it's the toolchain i downloaded from angstrom Jun 07 16:11:24 angstrom sdk Jun 07 16:11:47 http://narcissus.angstrom-distribution.org/ Jun 07 16:12:03 i had build online a sdk for the beagleboard Jun 07 16:12:19 so i get the toolchain from there Jun 07 16:13:40 captainigloo: what is its different? Jun 07 16:14:05 should be a gcc 4.3 or 4.4 Jun 07 16:14:22 but it should be ok with the toolchain build by shr Jun 07 16:14:46 it's the same both are based on oe Jun 07 16:14:54 only gcc version should be different Jun 07 16:15:07 depending on when you build it and preference of the distro Jun 07 16:15:32 i guess :) Jun 07 16:15:38 captainigloo: aha ok thanks man Jun 07 17:04:38 DocScrutinizer: hmm... what is the case for prios for single controls? Jun 07 17:04:57 I don't see why we need them Jun 07 17:05:06 arghh... woke up again Jun 07 17:05:07 bbiab Jun 07 17:20:00 captainigloo: ping Jun 07 17:25:57 dcordes: ping Jun 07 17:29:02 hi I just updated http://shr-project.org/trac/ticket/1244 with a patch shall I post it to fso-userspace and shr-devel ML? Jun 07 17:30:59 DocScrutinizer: tell me if I did not get it yet :) Jun 07 17:31:00 how can i change the resolution and DPI Jun 07 17:31:16 DocScrutinizer: the amixer thing is for enabling certain routes Jun 07 17:31:33 DocScrutinizer: like 'gimme gsm on the stereo speakers' Jun 07 17:31:48 DocScrutinizer: and if I say that I want that regardless of priorites, no? Jun 07 17:32:03 DocScrutinizer: for priorities we have the virtual devices like media, ringtone, alert, ... Jun 07 17:32:06 mrmoku: in 45min, on the road Jun 07 17:32:11 ok :) Jun 07 17:35:40 lxsameer: in e configuration Jun 07 17:35:53 you can change the scalling properties Jun 07 17:35:57 scaling* Jun 07 17:36:09 captainigloo: where ? machine file? Jun 07 17:36:52 directly on the device Jun 07 17:37:15 but you should change the Xorg dpi settings but I don't know how to do that Jun 07 17:37:38 E take into account xorg -dpi=220 for example Jun 07 17:37:48 but i don't know how to do that on shr Jun 07 17:38:11 captainigloo: you talk about /etc/freesmartphone before, i can't find that where is it ? Jun 07 17:38:30 on your device Jun 07 17:38:49 in your rootfs Jun 07 17:39:04 captainigloo: how can i change that in my PC before building the rootfs file? Jun 07 17:39:15 no idea :( Jun 07 17:39:28 well i guess you should create a new machine Jun 07 17:39:42 with configurations files for bravo Jun 07 17:39:56 and set MACHINE=bravo Jun 07 17:40:09 but you need to create all configuration files Jun 07 17:40:23 and should change some scripts ... Jun 07 17:40:26 i don't know Jun 07 17:40:38 i try to do that, but i finally take htcleo Jun 07 17:40:45 captainigloo: i already created that Jun 07 17:40:52 captainigloo: :D Jun 07 17:41:29 captainigloo: hm does your screen rotate when you rotate your phone ? Jun 07 17:41:33 no Jun 07 17:42:10 we need xorg msm driver Jun 07 17:42:26 but it's maybe doable with fbdev Jun 07 17:42:28 i don't know Jun 07 17:42:31 captainigloo: did you try that before? Jun 07 17:42:51 i try to build xorg msm but without success Jun 07 17:42:52 captainigloo: aha thanks a lot my friend Jun 07 17:43:12 i know that you can change orientation staticaly with fbdev Jun 07 17:43:51 captainigloo: aha that would be nice, do you know how? Jun 07 17:44:07 man fbdev ;) Jun 07 17:44:14 Option "Rotate" "string" Jun 07 17:44:53 dcordes: i finally made small changes in elfe Jun 07 17:45:05 captainigloo: ah thanks Jun 07 17:45:17 dcordes: visual feedback when you click on icons Jun 07 17:45:27 dcordes: it's you that ask for that ? Jun 07 17:46:55 hi mrmoku dcordes captainigloo Jun 07 17:46:57 captainigloo: which ones :-) Jun 07 17:47:01 ? Jun 07 17:47:03 what's rules.yaml exactly Jun 07 17:47:15 do I really need one Jun 07 17:47:19 for the nexusone Jun 07 17:47:28 as long as you're using oeventsd Jun 07 17:47:38 you need one Jun 07 17:47:40 I am I guess Jun 07 17:47:46 yup Jun 07 17:47:47 so I take the htcleo's one? Jun 07 17:48:07 should be good as starter Jun 07 17:48:36 as starter? I don't understand hwo it works exactly Jun 07 17:48:44 is it documented somewhere? Jun 07 17:48:51 or should I look at in in more details Jun 07 17:49:08 mrmoku: just click feedback when you clicked on icons Jun 07 17:49:17 and grid layout for all apps Jun 07 17:50:17 and i'm wondering if it's not necessary to create elfe-indicator to replace the top bar Jun 07 17:50:23 which is ugly Jun 07 17:50:26 does any one know how to change the rootfs before creating image? Jun 07 17:50:55 lxsameer: you untar it on a sdcard ? Jun 07 17:51:15 captainigloo: yeah Jun 07 17:51:28 so change it on the sdcard :) Jun 07 17:51:58 or untar it on your hard disk, change it, and copy it on the sdcard Jun 07 17:52:16 captainigloo: great... click feedback is one of the two things I miss most Jun 07 17:52:22 captainigloo: oh you're right, shoot me in the head Jun 07 17:52:28 lxsameer: :P Jun 07 17:52:37 the other one probably is already implemented and my rev is to old or I don't know how to use it :P Jun 07 17:53:00 mrmoku: i guess you are talking about deleting icons ? Jun 07 17:53:09 deleting / moving, exactly Jun 07 17:53:16 i notice that on my tablet Jun 07 17:53:28 mrmoku, can I push an exact copy of htcleo's config? Jun 07 17:53:43 for the rules.yaml Jun 07 17:53:45 mrmoku: it's maybe because the mouse pointer is moving during the long click Jun 07 17:53:57 i will look at it Jun 07 17:54:31 captainigloo: ahh, long click should do it... ok... maybe I try with a stylus Jun 07 17:54:49 GNUtoo: what does the nexus have? Jun 07 17:55:15 mrmoku, no keyboard, and it's pretty similar to the leo Jun 07 17:55:22 it's like a leo running android Jun 07 17:55:31 leo == hd2 == dcordes Jun 07 17:55:34 oops Jun 07 17:55:39 leo == hd2 == dcordes's phone Jun 07 17:55:48 the ' is near enter Jun 07 17:56:09 basically power, volume up-down, trackball and touchscreen Jun 07 17:56:16 and unusable yet software buttons Jun 07 17:56:24 like back,menu,home,search Jun 07 17:56:32 how can i install more apps in my final image ? Jun 07 17:56:48 lxsameer, you mean create an image with more apps? Jun 07 17:56:56 or just install once the image is installed Jun 07 17:57:04 I mean like apt-get for instance Jun 07 17:57:06 GNUtoo: the interesting things for oeventsd are more leds Jun 07 17:57:08 GNUtoo: yeah Jun 07 17:57:16 to indicate charging Jun 07 17:57:21 mrmoku, ah yes: Jun 07 17:57:28 anyway... just copy it as starter Jun 07 17:57:30 *leds, and backlight led Jun 07 17:57:38 can I push as is Jun 07 17:57:39 copied Jun 07 17:57:46 lxsameer, learn about opkg Jun 07 17:57:49 GNUtoo: before creating image, for example i want apt and some other application to be included Jun 07 17:57:50 nothing bad should happen Jun 07 17:57:50 opkg list Jun 07 17:57:54 opkg list_installed Jun 07 17:57:56 and opkg install Jun 07 17:58:08 GNUtoo: ok Jun 07 17:58:14 mrmoku, ok so I'll push, I was just asking for permission Jun 07 17:58:22 to push a not perfect thing Jun 07 17:58:58 lxsameer, we have opkg that is like apt but more lightweight Jun 07 17:59:10 it's a rewrite of ipkg if I understand well Jun 07 17:59:47 GNUtoo: hmm but i have to use opkg in my device Jun 07 17:59:59 GNUtoo: i want to create apps in my PC Jun 07 18:00:08 GNUtoo: and add them to rootfs Jun 07 18:00:13 lxsameer, ok Jun 07 18:00:14 lxsameer, recipes/images/shr-image.inc ;) Jun 07 18:00:31 GNUtoo: did you hit "no matching cipher found: client blowfish-cbc server aes128-ctr,3des-ctr,aes256-ctr,aes128-cbc,3des-cbc,aes256-cbc,twofish256-cbc,twofish-cbc,twofish128-cbc" bug on FR? Jun 07 18:00:37 lxsameer, to do that first write the app on your pc, then package the app for openembedded making a recipe Jun 07 18:00:50 and then just include it in the image like pespin said Jun 07 18:01:06 lindi-, not yet, where is that bug? Jun 07 18:01:20 GNUtoo: thanks man Jun 07 18:01:31 GNUtoo: google found some pastebin page with gnutoo-desktop as hostname :) Jun 07 18:01:39 ah? Jun 07 18:01:42 where is that: Jun 07 18:01:47 s/:/?/ Jun 07 18:01:47 GNUtoo meant: where is that? Jun 07 18:02:00 and gnutoo-desktop is me Jun 07 18:02:15 GNUtoo: http://pastebin.com/C9yK5HqE Jun 07 18:02:39 ahhh Jun 07 18:02:45 that's not the freerunner Jun 07 18:02:47 that's the bug1 Jun 07 18:02:54 ah, found it Jun 07 18:03:02 from buglabs Jun 07 18:03:11 I had blacklisted some ciphers on client side indeed Jun 07 18:03:17 ok Jun 07 18:03:22 I've too many devices Jun 07 18:03:42 and no I didn't solve that ssh bug Jun 07 18:03:58 basically on bug1 ssh became not accesible anymore at some point Jun 07 18:04:04 but it was with ultra old rootfs Jun 07 18:04:06 named 1.4.3 Jun 07 18:04:17 (their last poky rootfs) Jun 07 18:04:28 now I'm trying to port that to openembedded....slowly Jun 07 18:05:18 freesmartphone.org: 03GNUtoo 07framework * r19364a246e21 10/etc/nexusone/frameworkd.conf: Jun 07 18:05:18 freesmartphone.org: Add a frameworkd.conf for the nexusone machine Jun 07 18:05:18 freesmartphone.org: Signed-off-by: Denis 'GNUtoo' Carikli Jun 07 18:05:19 freesmartphone.org: 03GNUtoo 07framework * rc835f64c8e2b 10/etc/freesmartphone/oevents/nexusone/rules.yaml: Jun 07 18:05:19 freesmartphone.org: Add rules.yaml for the nexusone machine Jun 07 18:05:19 freesmartphone.org: Theses rules were copied from the htcleo machine and left intact. Jun 07 18:05:19 freesmartphone.org: Signed-off-by: Denis 'GNUtoo' Carikli Jun 07 18:06:17 basically I add limited support for nexusone in order to save me time Jun 07 18:06:53 when I have to help dcordes, rebasing my changes make me loose a lot of time Jun 07 18:06:53 and helping dcordes and captainigloo is strategical Jun 07 18:07:03 + it's normal to help people Jun 07 18:08:27 * GNUtoo is trying to justify that work so lindi- doens't question the freedom of the work Jun 07 18:08:53 (the main CPU and the baseband share the same ram bus on theses qualcomm soc) Jun 07 18:17:13 mrmoku: ok it should be better now :) Jun 07 18:17:38 but you will still no be able to move icons, only remove them Jun 07 18:17:44 captainigloo, btw I'm doing a porting guide Jun 07 18:18:09 http://www.shr-project.org/trac/wiki/Device Porting Guide Jun 07 18:18:21 maybe you could improve it when you upstream your changes Jun 07 18:19:50 captainigloo: can you give me your resolution please ? Jun 07 18:20:03 480x800 I guess Jun 07 18:23:11 hmm which is the recipe to build shr feed? Jun 07 18:23:32 http://en.wikipedia.org/wiki/HTC_Desire says 480x800 Jun 07 18:23:51 ah, found it Jun 07 18:23:56 task-shr-feed Jun 07 18:24:28 thanks guys Jun 07 18:26:00 captainigloo: nice :) it's mrmoku who asked for it Jun 07 18:26:13 captainigloo: as a condition to make elfe default Jun 07 18:26:14 hi all Jun 07 18:26:31 dcordes, hi Jun 07 18:27:36 dcordes, I already have pushed fso configs for the nexusone, and will push to oe.dev when it's ready, while doing it I'm writing a porting guide Jun 07 18:30:10 dcordes, btw about sound morphis helped me Jun 07 18:30:28 dcordes, and I finally got a PCM software slider Jun 07 18:33:50 captainigloo: great :) Jun 07 18:34:19 then at least one can remove re-add in another position Jun 07 18:36:22 captainigloo, btw the palm pre didn't arrive yet Jun 07 18:36:29 GNUtoo: crazy ! Jun 07 18:36:37 dcordes, ? Jun 07 18:36:49 GNUtoo: great ! :) Jun 07 18:36:52 ah ok Jun 07 18:36:54 lxsameer: any news from your side ? Jun 07 18:36:59 the software slider require something tough Jun 07 18:37:22 * a fake alsa card (easy there is one in the kernel) Jun 07 18:37:28 * to use aplay before using it Jun 07 18:37:38 http://www.pastie.org/2033705 Jun 07 18:37:50 with that you need to play a sound on softvol Jun 07 18:37:53 lxsameer: GNUtoo is adding nexusone in the fso (phone daemon) configuration. maybe he can help you add bravo too. it should be fairly easy Jun 07 18:37:54 for instance Jun 07 18:38:02 aplay -Dsoftvol something.wav Jun 07 18:38:09 you have to do it only once tough Jun 07 18:38:16 like at the first boot Jun 07 18:38:29 GNUtoo: ok. can we add auto-mechanism in oe ? Jun 07 18:38:35 dcordes, I'm already helping with that: http://www.shr-project.org/trac/wiki/Device%20Porting%20Guide Jun 07 18:38:45 dcordes, yes post-install Jun 07 18:38:51 ah nice Jun 07 18:38:58 dcordes, but I want to add all to fsoaudiod Jun 07 18:38:59 a good idea to make such a guide Jun 07 18:39:23 mrmoku: (priorities) there are no implicit prio of virt devices. You need to associate a vdev to a path, to get a prio. A oversimplified example for ctl prio: you open system-sound (high prio) but it includes EQ for some reason. Now you need to mix sth to system-sounds, but It *needs* EQ, while for system-sounds EQ is just a nice-to-have Jun 07 18:40:08 DocScrutinizer: for virt devices prio is handled by fsoaudiod Jun 07 18:41:02 dcordes: hi there Jun 07 18:41:13 dcordes: finally i boot shr in my device Jun 07 18:41:25 lxsameer: great job Jun 07 18:41:32 dcordes: it has some problems that i will work on it Jun 07 18:41:36 dcordes,strange.... Jun 07 18:41:37 and fsoaudiod doesn't have any hardcoded a priori knowledge about prio of devices aka pathes. Zhere's just one source it may gather this knowledge: from path defs Jun 07 18:41:45 PCM appeared without me doing -Dsoftvol Jun 07 18:41:47 dcordes: i'll be happy to add it in fso Jun 07 18:42:02 I had /home/root left tough Jun 07 18:42:11 (ffrom the previous install) Jun 07 18:42:23 GNUtoo: can you help me to add bravo in fso too ? Jun 07 18:42:42 lxsameer, yes but right now I've to commit and push the nexusone Jun 07 18:43:08 GNUtoo: ok i'll wait Jun 07 18:43:24 lxsameer, but you can ask questions still Jun 07 18:43:35 ask and you'll see if I respond fast enough Jun 07 18:43:39 I've splitted screens Jun 07 18:43:45 *a splitted screen Jun 07 18:43:54 with kde4's titled like feature Jun 07 18:44:16 GNUtoo: :) nice, so what should i do to add bravo to fso ( don't kill me please :P) Jun 07 18:44:18 DocScrutinizer: hmm... I didn't think about mixing Jun 07 18:44:40 dcordes: can i have your xorg.conf Jun 07 18:44:45 mrmoku: fsoaudiod doesn't even know about which vdevs are available, a hook sends his own devicename to fsoaudiod and fsoaudiod looks for the matching pathdev in a .amix Jun 07 18:44:46 lxsameer, read that: http://www.shr-project.org/trac/wiki/Device Porting Guide Jun 07 18:44:56 GNUtoo: sure Jun 07 18:45:30 then after you read it, please ask more specific quesitons and complete the guide with the answer Jun 07 18:45:41 I.E looks for $[devname}.amix Jun 07 18:47:32 DocScrutinizer: with devname being something like ringtone, media, systemsound? or gsmmic, stereospeaker.... ? Jun 07 18:47:45 exaczly Jun 07 18:48:02 usually Jun 07 18:48:31 DocScrutinizer: exactly which one? Jun 07 18:48:37 or gsm_2_earpiece, if there's an alternative source Jun 07 18:49:03 I mean some virtual soundcard or a virtual route? Jun 07 18:49:07 so virtual route then Jun 07 18:49:23 route, the devices always represent routes Jun 07 18:49:26 but that's not what we thought the hook to provide Jun 07 18:49:51 there's no path with one endpoint Jun 07 18:50:01 a vdev represents a path Jun 07 18:50:07 DocScrutinizer: http://paste.pocoo.org/show/402302/ Jun 07 18:50:19 is the asound.rc we have in fsoaudiod right now Jun 07 18:50:37 usually your startpoint uis clear, it's PCM Jun 07 18:52:08 pity that morphis is not around... Jun 07 18:52:54 slave dmix looks odd Jun 07 18:53:11 there's not *one* dmix, there may be several Jun 07 18:53:16 named ones Jun 07 18:53:19 GNUtoo: how can i add devtmpfs to my kernel Jun 07 18:53:30 lxsameer, what is your kenrel version? Jun 07 18:54:07 GNUtoo: 2.6.37, is it a kernel config ? Jun 07 18:54:08 slave usually is sth like earpice_dmix or stereospeaker_dmix Jun 07 18:54:23 hmm Jun 07 18:54:27 lxsameer, yes it is Jun 07 18:54:32 or all_chan_dmix Jun 07 18:54:36 make ARCH=arm xconfig and look for it Jun 07 18:54:46 GNUtoo: ok i find it Jun 07 18:55:04 DocScrutinizer: all_chan_dmix is probably the correct thing there then Jun 07 18:55:15 depends Jun 07 18:55:18 GNUtoo: is there any thing that devtmpfs depend to that? Jun 07 18:55:31 for ringtones, likely not Jun 07 18:56:11 for mp3Playback_wired, likely not Jun 07 18:56:17 lxsameer, ctrl + f Jun 07 18:56:22 and also Jun 07 18:56:23 go on top Jun 07 18:56:36 and do that: Jun 07 18:56:37 GNUtoo: ??? Jun 07 18:56:42 if the phone app is playing a ringtone with lets say 'aplay -Dringtone' fsoaudiod will mute the others Jun 07 18:56:53 option Jun 07 18:56:55 where mp3Playback_wired is zo hook up your device to a homestereo without systemtones, ringtones, whatnot Jun 07 18:56:58 and check everything Jun 07 18:57:10 (show name,range,data) Jun 07 18:57:14 and check show all options Jun 07 18:57:22 DocScrutinizer: that's for v3 then :P Jun 07 18:57:30 lxsameer, I'm explaining how to make ARCH=arm xconfig Jun 07 18:57:38 that will let you find devtmpfs Jun 07 18:57:41 that's for basic architecture Jun 07 18:57:48 ctrl+f is fine Jun 07 18:57:50 *find Jun 07 18:57:57 * meant ctrl+f == find Jun 07 18:58:04 I don't care about v967 implemetation milestone Jun 07 18:58:07 GNUtoo: ah sorry Jun 07 18:58:13 np Jun 07 18:58:34 * DocScrutinizer off for dinner Jun 07 18:58:43 ~bon appetit Jun 07 18:58:43 well, bon appetit is smacznego. Guten Appetit. Eet Smakelijk. God Appetitt. Buon Appetito. Buen apetito Bom Apetite. buen apetito Smaklig måltid!. Hyvää ruokahalua. Bo Proveito Mahlzeit! Jun 07 18:58:51 so ping me if sth important happens ;-D Jun 07 18:59:08 DocScrutinizer: will continue thinking and digesting :) Jun 07 18:59:19 k Jun 07 19:01:05 ChristW: I had an idea what could be the root of our problem Jun 07 19:01:30 ChristW: the asound.rc that comes with fsoaudiod Jun 07 19:03:57 dcordes, ok it apear when using the normal default device Jun 07 19:04:00 so no worries Jun 07 19:06:11 lxsameer: you can use the htcdream xorg.conf found in oe Jun 07 19:06:25 lxsameer: find ./recipes/ -name xorg.conf | grep dream Jun 07 19:06:33 dcordes: ok thanks Jun 07 19:06:39 lxsameer: and put wvga resolution Jun 07 19:06:46 lxsameer: it also shows you how to rotate Jun 07 19:07:22 dcordes: great help thanks man Jun 07 19:08:25 GNUtoo: what is g_ether Jun 07 19:08:39 lxsameer: usb networking Jun 07 19:08:52 lxsameer, for having usb networking, it's a kernel module Jun 07 19:09:40 GNUtoo: hmm i can't find a config with that name Jun 07 19:10:07 ChristW: hmm... does not seem to help to move it away :/ Jun 07 19:10:23 lxsameer, look for usb gadget Jun 07 19:10:47 and ethernet gadget Jun 07 19:10:48 GNUtoo: ok Jun 07 19:10:54 and please update the howto with that info Jun 07 19:11:36 for me all this stuff is natural, so I've no idea on how it looks for someone that has not all the knowledge we(the shr devlopers) usually have Jun 07 19:11:44 GNUtoo: ok Jun 07 19:11:54 so your contribution would be very welcomed in that howto Jun 07 19:14:03 hi dcordes Jun 07 19:14:06 I've some problem Jun 07 19:14:14 the display becomes black with new rules.yaml Jun 07 19:14:25 I know how to make it come back tough Jun 07 19:14:46 chvt 1;chvt 2 Jun 07 19:14:52 how to hook that in rules.yaml Jun 07 19:14:57 GNUtoo: rules.yaml ? Jun 07 19:14:58 I've the exact same than leo Jun 07 19:15:00 yes Jun 07 19:15:08 you asked mrmoku to commit that for you Jun 07 19:15:17 in the framework repo Jun 07 19:17:04 GNUtoo: will USB_GADGET be enough or i should build the other drivers too ? Jun 07 19:17:16 lxsameer, it's enough Jun 07 19:17:30 you should do that tough: Jun 07 19:18:02 in usb peripheral controller select your device Jun 07 19:18:07 that is to say msm7k Jun 07 19:18:21 then in usb gadget select ethernet gadget Jun 07 19:18:25 and it should work Jun 07 19:18:30 GNUtoo: aha Jun 07 19:18:32 put all that compiled in the kernel Jun 07 19:18:37 since it's easier at first Jun 07 19:18:39 for debugging Jun 07 19:18:44 else if it's as a module Jun 07 19:18:48 it could not make it in the image Jun 07 19:18:53 if you don't do stuff correctly Jun 07 19:18:56 GNUtoo: - Command('xset -display localhost:0 s blank') Jun 07 19:18:56 - Command('xset -display localhost:0 s activate') Jun 07 19:18:58 leaving you with no shell Jun 07 19:19:03 does probably not work? Jun 07 19:19:06 mrmoku, yes I see that Jun 07 19:19:16 that's for blanking or unblanking? Jun 07 19:19:23 blanking Jun 07 19:19:27 ok Jun 07 19:19:30 so for resuming Jun 07 19:20:08 GNUtoo: maybe just the unblank xset is missing some line up Jun 07 19:20:27 ok Jun 07 19:20:46 yes it seem missing Jun 07 19:21:24 I'll try Jun 07 19:22:28 mrmoku, other question, how what should I do with reguards to connmand Jun 07 19:22:34 it setup the default route on rmnet0 Jun 07 19:22:41 (the 3g interface) Jun 07 19:23:04 and you don't want that? Jun 07 19:23:45 no, wifi breaks, udhcpc breaks etc... Jun 07 19:23:46 with that Jun 07 19:23:55 thanks a lot for the yaml trick Jun 07 19:24:00 works? Jun 07 19:24:05 yes Jun 07 19:24:06 seem so Jun 07 19:24:08 it resumes Jun 07 19:24:14 great Jun 07 19:24:35 dcordes, copy my yaml and test it when I'll have commited it Jun 07 19:25:13 GNUtoo: (connmand) unfortunatelly I have absolutely no idea... probably the correct way would be to let it know about wifi too... let it handle all networking Jun 07 19:25:43 e has that nice connman gadget... maybe we should invest some time into making it work Jun 07 19:25:46 hmmm not a good idea, it also breaks usbnet Jun 07 19:25:55 but I could workarround Jun 07 19:26:02 in /etc/network/interfaces Jun 07 19:26:09 it also breaks dream Jun 07 19:26:38 does any one has access to shr-project site ? Jun 07 19:26:48 yes most of us do Jun 07 19:27:30 GNUtoo: so please activate my account , i don't get the activation mail Jun 07 19:28:18 ah you mean for account activation Jun 07 19:28:31 maybe mrmoku could do that, else wait for the email Jun 07 19:28:50 mails can take time to arrive.... Jun 07 19:29:44 GNUtoo: should i just add the network/interfaces or i should add interface configuration too ? Jun 07 19:30:05 ? Jun 07 19:30:12 * GNUtoo doesn't understand that part: Jun 07 19:30:18 or i should add interface configuration too ? Jun 07 19:30:47 GNUtoo: what should i do about /etc/network/interfaces ? Jun 07 19:31:03 add it in openembedded Jun 07 19:31:12 basically the default one lacks auto usb0 Jun 07 19:31:15 which is crucial Jun 07 19:31:20 so add one for your machine Jun 07 19:31:27 lxsameer: for trac? Jun 07 19:31:46 mrmoku: yes please Jun 07 19:32:05 mrmoku, should I add a default /etc/network/interfaces for shr in openembedded? Jun 07 19:33:20 lxsameer: did you try if it works? Jun 07 19:33:25 I can see you there Jun 07 19:33:26 lxsameer, it's in recipes/netbase/netbase Jun 07 19:33:39 GNUtoo: aha Jun 07 19:34:01 brb Jun 07 19:34:32 mrmoku: it does not work yet Jun 07 19:35:44 lxsameer, you also have to bump PR Jun 07 19:35:58 GNUtoo: ok Jun 07 19:37:38 GNUtoo: hmm i don't get it , that interface file has a auto usb0 already Jun 07 19:37:39 I've updated the howto Jun 07 19:37:50 I added xorg.conf and /etc/network/interaces link Jun 07 19:38:07 lxsameer, the image you're running comes from another device? Jun 07 19:38:21 or are you running the image you built with openembedded? Jun 07 19:39:00 GNUtoo: ??? i'm currently look in to ~/openembedded/recipes/netbase/netbase Jun 07 19:39:00 mrmoku: Jun 07 19:39:05 default * 0.0.0.0 U 0 0 0 rmnet0 Jun 07 19:39:05 default * 0.0.0.0 U 0 0 0 rmnet2 Jun 07 19:39:05 default * 0.0.0.0 U 0 0 0 rmnet1 Jun 07 19:39:12 that is problematic Jun 07 19:39:38 lxsameer, I'll look too Jun 07 19:40:04 # ... or on host side, usbnet and random hwaddr Jun 07 19:40:04 iface usb0 inet static Jun 07 19:40:11 no there is no auto usb0 in: Jun 07 19:40:42 recipes/netbase/netbase/interfaces Jun 07 19:41:01 GNUtoo: hm i just made a directory and create a interfaces in it , is it ok ? Jun 07 19:41:15 yes but it's not enough Jun 07 19:41:19 you need to bump PE Jun 07 19:41:21 oops Jun 07 19:41:23 you need to bump PR Jun 07 19:41:29 and it'll be ok Jun 07 19:41:41 the directory has to have the name of your machine in openembedded Jun 07 19:41:48 GNUtoo: where is that PR Jun 07 19:43:06 netbase_4.21.bb Jun 07 19:43:18 GNUtoo: ow sorry Jun 07 19:48:32 lxsameer: I can't confirm you... what I can do is delete and re-add you Jun 07 19:48:57 mrmoku: thanks Jun 07 19:49:01 GNUtoo: yeah, that looks indeed problematic Jun 07 19:49:30 freesmartphone.org: 03morphis 07cornucopia * rb8cddbd5f357 10/configure.ac: Variable should be named $qtembedded and not $qwssupport within configure.ac Jun 07 19:49:31 freesmartphone.org: 03morphis 07cornucopia * rf82be0c3c3c3 10/configure.ac: Add missing statement to retrieve git version from repoistory Jun 07 19:49:31 freesmartphone.org: 03morphis 07cornucopia * rc25a639268de 10/src/Makefile.am: Add QFsoDBus header file which includes all other headers Jun 07 19:49:31 freesmartphone.org: 03morphis 07cornucopia * rc932d8b87d0b 10/src/ (Makefile.am qfsodbusabstractinterface.cpp): Even generate the moc file for the QFsoDBusAbstractInterface class Jun 07 19:51:47 GNUtoo: wandering rmnet is normal there? Jun 07 19:52:17 mrmoku, we need only one rmnet0 for internet Jun 07 19:52:21 s/internet/3g Jun 07 19:52:32 where does 1 and 2 come from? Jun 07 19:52:54 I guess you can have multiples data connections at the same time Jun 07 19:55:42 ahh ok Jun 07 19:56:00 so you always have those three and just the first one is 3g... ic Jun 07 19:56:08 GNUtoo: should i build a patch for DPI too Jun 07 19:56:46 yes Jun 07 19:57:05 mrmoku, so I disables them in /etc/network/interfaces? Jun 07 20:02:08 GNUtoo: does connman respect that? Jun 07 20:02:45 I think so Jun 07 20:02:55 they told so at buglabs Jun 07 20:02:57 I'll try Jun 07 20:03:06 what should I tell as ip? Jun 07 20:08:21 GNUtoo: connman sucks... at least when it comes to documentation ;) Jun 07 20:08:29 ah ok Jun 07 20:09:03 -I, --nodevice=DEV Specify networking interface to ignore Jun 07 20:09:10 would be another possibility Jun 07 20:09:19 yes Jun 07 20:09:22 how do we do that? Jun 07 20:10:08 you add a /etc/default/connman with: Jun 07 20:10:18 DAEMON="/usr/sbin/connmand -I usb0 -I rmnet1 -I rmnet2 -r" Jun 07 20:10:45 -I usb0 is there by default Jun 07 20:10:47 at least on gta02 Jun 07 20:12:01 (in the initscript directly) Jun 07 20:12:08 GNUtoo: i can't understand the frameworkd.conf part Jun 07 20:12:46 lxsameer, ok basically there are sections with: Jun 07 20:12:50 [foo] Jun 07 20:12:57 GNUtoo: it's an shr initscript override Jun 07 20:12:58 you just have to add that under it: Jun 07 20:13:01 GNUtoo: no where is that file Jun 07 20:13:02 disable = 1 Jun 07 20:13:05 so you could put it in there too Jun 07 20:13:14 mrmoku, ? Jun 07 20:13:24 ah ok Jun 07 20:13:34 sorry I didn't read what you said before Jun 07 20:13:45 lxsameer, in the frameworkd repository Jun 07 20:13:49 you need to make a patch Jun 07 20:13:53 and send it Jun 07 20:14:26 GNUtoo: hmmmm let me try some Jun 07 20:26:01 mrmoku, but rmnet0 should be ignored too Jun 07 20:26:05 I'll add the 3 rmnets Jun 07 20:28:24 GNUtoo: hmmm ok Jun 07 20:28:46 GNUtoo: but how can i add that to my current build Jun 07 20:28:56 lxsameer, you can't Jun 07 20:29:01 you must send a patch Jun 07 20:29:22 GNUtoo: so i have to wait until the patch accepted ? Jun 07 20:29:59 mrmoku, I'm not sure the connman config is a good idea Jun 07 20:30:13 it will make connman machine arch for everyy machine Jun 07 20:30:22 *for every machine we override it for Jun 07 20:30:41 /etc/network/interfaces is not a good idea either Jun 07 20:30:48 since I don't know what ip to set Jun 07 20:30:52 hmmmm Jun 07 20:30:59 lxsameer, yes Jun 07 20:31:06 lxsameer, do the patch now then Jun 07 20:31:12 first do the freesmarphones patches Jun 07 20:31:18 GNUtoo: then add them to the shr initscript in connman files Jun 07 20:31:20 GNUtoo: yeah it seems i have to Jun 07 20:31:35 lxsameer, and don't forget to modify the porting guide Jun 07 20:31:42 mrmoku, ok Jun 07 20:32:08 mrmoku, but then will 3g work? Jun 07 20:32:10 GNUtoo: i don't have account yet Jun 07 20:32:14 how does connman work? Jun 07 20:32:17 ok Jun 07 20:32:35 GNUtoo: what do you mean by freesmartphones patches? Jun 07 20:32:54 GNUtoo: no idea... morphis added a plugin for connman... no idea how that works Jun 07 20:32:55 lxsameer, I mean that the repository is at freesmartphone.org Jun 07 20:33:02 ok Jun 07 20:33:11 trough dbus I guess? Jun 07 20:33:41 I'll have to try then Jun 07 20:34:20 the providers should love people testing the phones instead of preventing them to do so.....they spend 1E each time they need to test 3g Jun 07 20:34:46 s/they/people like us/ Jun 07 20:34:46 GNUtoo meant: the providers should love people testing the phones instead of preventing them to do so.....people like us spend 1E each time they need to test 3g Jun 07 20:34:46 :) Jun 07 20:35:01 GNUtoo: btw. no umts coverage from wind where I go :/ Jun 07 20:35:10 ouch Jun 07 20:35:34 do it your self then with the help of an USRP.... Jun 07 20:35:37 (joke) Jun 07 20:35:45 :P Jun 07 20:36:06 you would have mrmoku on your provider bar Jun 07 20:36:15 or fso Jun 07 20:36:21 or something like that Jun 07 20:36:22 that one I could fake anyway ;) Jun 07 20:36:34 yes but it would be a real one Jun 07 20:36:48 it's easy to fake with the fake fsogsmd plugin Jun 07 20:36:56 my suocera would not see the difference :) Jun 07 20:37:55 lol Jun 07 20:38:00 on her phone she would Jun 07 20:40:33 btw hmmm Jun 07 20:40:51 the indicator for the battery of e doesn't work on the nexusone while apm works Jun 07 20:41:01 wasn't it fixed? Jun 07 20:43:01 hmmm Jun 07 20:43:11 about the connman I'll look at n900 Jun 07 20:44:15 GNUtoo: the bat indicator works via fsodeviced now Jun 07 20:44:26 yes I know Jun 07 20:44:27 what fsodeviced sees should work Jun 07 20:44:44 that's why I don't understand why it doesn't work Jun 07 20:44:45 it does not on gta02 because with the newer kernel bq27x00 does not send uevents anymore Jun 07 20:44:49 ahhh Jun 07 20:44:52 illume cache Jun 07 20:45:14 GNUtoo: should i send patch for cornucopia too Jun 07 20:45:57 ha...now it works after rm -rf .e Jun 07 20:46:33 lxsameer, yes Jun 07 20:46:46 GNUtoo: to where ? Jun 07 20:47:39 both to freesmartphone mailing list Jun 07 20:47:43 their website is at: Jun 07 20:47:58 http://www.freesmartphone.org/ Jun 07 20:48:06 go in wiki Jun 07 20:48:22 then community infrastructure Jun 07 20:49:58 hmmm Jun 07 20:50:07 maybe the rmnet interface comes up by default Jun 07 20:50:12 and are picked because of that Jun 07 20:50:52 GNUtoo: can you please check for my mail into that ML ? it seems that my network have some problems Jun 07 20:51:11 ok Jun 07 20:51:25 ping me in 5 minutes for that Jun 07 20:51:31 GNUtoo: ok Jun 07 20:51:35 since mails are not instantaneous Jun 07 20:53:45 I'll try something for the interfaces and connman Jun 07 21:00:34 GNUtoo: ping :P Jun 07 21:04:56 mail object? Jun 07 21:06:12 GNUtoo: [PATCH] htcbravo Jun 07 21:06:24 GNUtoo: it takes really a long time for the pre ! Jun 07 21:06:32 I've nothing Jun 07 21:06:37 for the htcbravo patch Jun 07 21:06:45 btw i think that other people could be interreseted ;) Jun 07 21:06:46 captainigloo, on another hand I'm polishing nexusone Jun 07 21:06:55 I'm tired of rebasing my configs etc... Jun 07 21:07:15 ok Jun 07 21:08:09 * mrmoku is tired too... not of rebasing but in general :P Jun 07 21:08:14 gnight all :) Jun 07 21:10:16 lol good night Jun 07 21:16:34 DocScrutinizer51: "less <(cat foo)" is not different from "cat foo|less" -- but you can do things like "diff <(cat Jun 07 21:16:52 whoops Jun 07 21:17:10 DocScrutinizer51: "less <(cat foo)" is not different from "cat foo|less" -- but you can do things like "diff <(cat foo) <(cat bar)" Jun 07 21:17:43 (obviously "cat" is totally pointless in all these examples... but that's just for illustration :-) ) Jun 07 21:27:06 freesmartphone.org: 03morphis 07cornucopia * r0a07f7fca53c 10/fsogsmd/src/plugins/modem_qualcomm_palm/channel.vala: fsogsmd: modem_qualcomm_palm: use higher timeout to wait for URC_TYPE_RESET_RADIO_IND Jun 07 23:56:57 ChristW: Dvorak created special keyboard layouts for one-handed typing... ;-) Jun 08 00:41:32 DocScrutinizer51: why is it not possible to get more than 25% of the native GSM bandwidth with a data stream? AIUI the latest generation of analogue modems can get up to 48k over a 56k line... Jun 08 00:42:40 antrik: eeh? Jun 08 00:43:30 I was talking about sending accoustic FSK etc over GSM voice codecs Jun 08 00:45:15 antrik: even if you could *completely* control what data the GSM codec is sending out (by feeding it with nifty audio), you still dunno what exchange side processing does to it Jun 08 00:46:12 oh, right... with a "normal" phone line, there is no codec, just plain PCM I guess... Jun 08 00:46:34 there are echo cancellers and audio improver magic at your carrier's site Jun 08 00:46:49 antrik: exactly Jun 08 00:47:23 and the GSM codec is optimized for voice Jun 08 00:47:58 ever tried transmitting techno music over it? or orchester? Jun 08 00:48:20 the codec is well-defined though I guess? so it should be possible to kinda reverse-engineer the right "sound" input to use the available bandwidth fully?... Jun 08 00:48:35 see above Jun 08 00:49:14 far end won't see what you send but what exchange processing made of it Jun 08 00:49:28 are you saying the providers are decoding the GSM, applying some audio processing to it, and reencoding it again? Jun 08 00:50:21 some do, for some types of connections Jun 08 00:51:20 some may even dynamically adapt the algos, also taking into account type of your MT Jun 08 00:51:46 echo cancellation is very common Jun 08 00:52:30 I see Jun 08 00:54:07 one of the major business secrets of carriers, as it directly affects their product quality Jun 08 00:54:43 and there's not much room to differentiate from competition otherwise, in GSM Jun 08 00:58:18 antrik: also even on intranet calls the one peer might use EFR while other end uses HR codec, so carrier needs to transcode anyway Jun 08 01:02:04 this has been trsted and has failed like 1000 times. It's not possible to send acoustic modem data over GSM codec, at a rate sufficient to have another voice codec work on it in reaktime Jun 08 01:04:15 I always like to say "if you need encrypted voice over plain GSM, get a Chirokee Indian on each end" Jun 08 01:09:29 DocScrutinizer: do you know where I can find the finger friendly layouts described here: http://wiki.openmoko.org/wiki/Illume_keyboard#List_of_layouts Jun 08 01:10:01 sorry no **** ENDING LOGGING AT Wed Jun 08 02:59:56 2011