IRC log of #maemo-devel for Wednesday, 2010-09-15

« Tuesday, 2010-09-14 Index Thursday, 2010-09-16 »
DocScrutinizerHardware PCM card 0 'RX51' device 0 subdevice 000:00
DocScrutinizerdefinition00:00
DocScrutinizerI'd bet it has 'rate 48000'00:00
MNZentire asound.conf is pasted00:00
DocScrutinizeractually I think I've seen this00:01
DocScrutinizerasound conf is just 10%00:01
MNZor you mean that other thing in /usr//lib/*00:01
DocScrutinizeryep, exactly00:01
MNZ/usr/share/alsa/alsa.conf00:02
DocScrutinizerdefaults.pcm.dmix.rate 4800000:02
MNZyeah found it00:03
MNZbut that's for default00:03
MNZI'm using hw:0,0 directly00:03
MNZoh wait00:03
MNZdefault_s_. I was thinking of the 'default' device00:03
MNZstill, didn't change anything :/00:04
DocScrutinizertry using hw:0.0 with 4410000:05
MNZyeah that works, have already tried00:05
MNZwithout -r or setting rate00:05
DocScrutinizerso it's obviously dmix itself00:05
MNZit negotiates and gets 44.100:05
MNZyeah00:05
DocScrutinizerall this is SOOO F... obfuscated XP00:06
DocScrutinizerand THAT'S the real reason why PA was invented00:06
MNZwell but why write PA when you can fix ALSA :S00:07
MNZor is it unfixable? I don't think so00:07
DocScrutinizerI also don't think so, but then the NIH syndrome kicks in usually00:08
DocScrutinizerwell, anyway my prev quote from http://www.alsa-project.org/alsa-doc/alsa-lib/pcm_plugins.html regarding dmix plugin properties might suggest dmix can't use different sampling rates00:10
DocScrutinizer...dynamically00:10
villagerPA wasn't written just to mix sounds together, you know00:10
DocScrutinizerso >>defaults.pcm.dmix.rate 48000<< is what we got, and what we may change to get another fixed sampling rate00:11
MNZvillager, yeah, it does software gain too :/ what's PA for again?00:11
villagerpersonally I only use PA to route sounds over wlan from my laptop to my desktop pc, which is connected to my high-quality speakers00:13
villagerbut someone once told me that it's cool that with PA he can turn on his bluetooth headset and the sounds on his laptop automatically gets routed to the headset instead of the internal speakers, without needing to restart the apps00:14
* DocScrutinizer mumbles, discarding a pretyped answer to a comment that seems not to come00:14
DocScrutinizerALSA even cooler: it can route to both BT headsaet and internal speakers concurrently, and even level both to arbitrary volume X-P00:15
villagerwith some config? then PA can do that too00:15
villagerbut I do agree that alsa does it with less cpu00:16
DocScrutinizervor audio streaming over IP I prefer other protocols than exactly PA, I have to say00:16
DocScrutinizerbut meh, granted that's one of PA's generic domains00:17
DocScrutinizerhighly usefull for an embedded device with memory and CPU restrictions00:17
DocScrutinizerI bet it was the killer argument why Nokia *needed* PA for maemo00:18
DocScrutinizeras well as for all standard desktop distros...00:18
villagerI figured it was so the music could turn off when you talk on the phone...00:19
DocScrutinizerthe point is simply: ALSA has been supported by Suse from very beginning, and since Suse is called Novel now, there's probably some resentiment against alsa00:20
MNZthe state reason is for making the audio system APE Centric.... Application Processing Engine Centric... it still doesn't really make sense00:20
MNZs/state/stated/00:20
DocScrutinizerthe music turn-off during phone calls has been solved here 2 h ago00:21
MNZthis dmix thing, I bet it would take minimal hacking to get it to do what we want and save 10% cpu for most usage00:21
villagerin a way probably unknown (or impractical) to nokia?00:21
MNZvillager, not really. dmix + softvol plugins + plugin written by Doc.00:22
MNZin contrast with 3 PA plugins00:22
MNZactually, you wouldn't really need Doc's plugin for this specific purpose00:23
MNZthe hook system in alsa is enough00:23
*** edisson has quit IRC00:23
villagerwell I'd guess they made a list of pros and cons or something from what they knew00:24
villagerthen again, perhaps they were ubuntu users and liked PA...00:25
DocScrutinizerMNZ: you're a breeze to watch you learn :-D00:26
MNZDocScrutinizer, :)00:26
DocScrutinizerobviously, absolutely. BUNTKUH users00:26
DocScrutinizersee nokia maemo scratchbox is working ootb only on buntkuh00:27
MNZok I need a quick evaluation because I'm kind stuck on what to do next :/00:29
MNZfilters are ready and work. De-emph kicks in if speaker output starts, EQ filter needs gui00:29
DocScrutinizerMNZ: you're incredible00:30
MNZnext step seems to me to start actually creating the setup for apps to automagically bypass PA (and still calls work)00:30
MNZDocScrutinizer, dude stop flattering me :P00:30
MNZand think more of what to do next :P I want to actually see something finished to the end for a change00:31
DocScrutinizermaybe you can have a look how to plug in a hook plugin to anything different than mere hw slaves, I got stuck on that detail back when, or I abandoned project while still on my way to solve that00:32
MNZis that going to be a problem? by what I've read so far about hooks it seems we won't need anything complicated00:34
DocScrutinizerpcm.t2pcm {    type hooks;     slave { pcm "hw:0,0" } ...00:37
DocScrutinizerwe'd probably like to use >slave.pcm music-dmix00:37
DocScrutinizerinstead of hw:0.000:37
DocScrutinizerhmm, maybe pcm.ringerpcm {  worked, and I just forgot00:39
DocScrutinizer   slave { pcm "default" }00:39
MNZthere are no such limitations listed in the docs, but of course these are ALSA docs, so who knows :/00:40
DocScrutinizerI'm rather sure I've seen 'date;sleep5;date' working on 'aplay -D ringerpcm foo'00:48
*** JPohlmann has joined #maemo-devel00:55
*** mirr0r has quit IRC01:01
*** mirr0r has joined #maemo-devel01:01
*** SpeedEvil has quit IRC01:03
MNZhrmm I just got alsa-lib source... there's pcm_dmix_{generic/x86/x86_64} !01:03
DocScrutinizermhm01:04
MNZso looks like it's not optimized for arm01:05
DocScrutinizeron a sidenote: for speaker protection it's most sensible to use that http://people.openmoko.org/joerg/unsorted/arkanoid-wav/16bit-44ks/Arkanoid_PSID.wav  style of ringtones, as they are produced with small speakers in mind (no bass and still good enough :-P)01:05
DocScrutinizerMNZ: check out open embedded / openmoko, there's alsa dmix as well, and maybe somebody bothered to optimize for arm01:07
*** _0x47 has quit IRC01:09
*** JPohlmann has quit IRC01:09
MNZlol @ ringtone DocScrutinizer01:14
*** SpeedEvil has joined #maemo-devel01:17
*** SpeedEvil has quit IRC01:34
*** SpeedEvil has joined #maemo-devel01:35
MNZGood news, my cup of mint tea is ready!01:37
MNZnow where are those openmoko repos01:38
DocScrutinizerhmm, might want a darjeeling01:38
DocScrutinizernah, Ceylon first flush01:38
DocScrutinizerumm, go for SHR01:38
DocScrutinizerhttp://shr-project.org01:38
DocScrutinizeror openembedded.org I guess01:39
MNZI don't get it...01:51
MNZso openmoko/openembedded etc git repos just contain patches and build scripts?01:52
*** nervewax has joined #maemo-devel02:02
nervewaxHi I got directed here by the maemo wiki02:03
SpeedEvil?02:05
nervewaxyeah see i'm going into Digital Media at uni and I am wondering if anyone can put me in the right direction for app development02:09
MNZnervewax, ask on #maemo for a broader audience02:19
*** nervewax has quit IRC02:52
*** ShadSEC has joined #maemo-devel03:22
*** csaavedra has quit IRC03:46
MNZgood night maemo devs!04:09
*** MNZ has quit IRC04:10
*** shinkamui has quit IRC06:19
*** shinkamui has joined #maemo-devel06:20
*** n4v4r3d has joined #maemo-devel06:56
*** n4v4r3d has left #maemo-devel06:56
*** kamui__ has joined #maemo-devel07:17
*** shinkamui has quit IRC07:21
*** tealbird has quit IRC07:22
*** DocScrutinizer has quit IRC07:33
*** DocScrutinizer has joined #maemo-devel07:33
*** tealbird has joined #maemo-devel07:50
*** mauso has joined #maemo-devel08:10
*** amigadave has joined #maemo-devel09:37
*** ShadSEC has quit IRC09:52
*** appnss has joined #maemo-devel09:53
*** appnss has quit IRC10:06
*** ShadSEC has joined #maemo-devel10:07
*** lcuk2 has joined #maemo-devel10:10
*** MNZ has joined #maemo-devel10:10
*** lcuk has quit IRC10:12
*** swc|666 has quit IRC10:30
*** ShadSEC has quit IRC11:13
*** ShadSEC has joined #maemo-devel11:13
*** appnss has joined #maemo-devel11:32
*** ShadSEC has quit IRC11:33
*** appnss has quit IRC11:47
*** mauso has quit IRC12:35
*** TheXception has joined #maemo-devel12:47
*** silbo has joined #maemo-devel12:47
TheXceptionhi12:47
MNZahoy12:48
*** air has joined #maemo-devel12:50
*** silbo has quit IRC12:57
*** lcuk2 is now known as lcuk13:01
*** lcuk has joined #maemo-devel13:01
*** achipa has joined #maemo-devel13:11
*** csaavedra has joined #maemo-devel13:12
*** appnss has joined #maemo-devel13:21
*** Ian-- has quit IRC13:22
*** _0x47 has joined #maemo-devel13:25
TheXceptionwas there any work done running free pascal compiler (fpc) on maemo (n900)? or did someone already try it?13:25
lcukooh13:25
lcukpascal13:25
*** Ian-- has joined #maemo-devel13:28
TheXceptioni don't need it myself but to port another project13:30
lcukTheXception, I saw you were havign an issue last night with compiling something else13:31
lcukdid you get past that, or does the app require direct ASM fpx stuff?13:31
*** silbo has joined #maemo-devel13:31
TheXceptionyes it's still the problem finding the correct hardware floating support13:31
*** csaavedra has quit IRC13:33
TheXceptionaccording to cat /proc/cpuinfo it is supported13:33
TheXceptionbut i then get a SIGILL and corrupted stack13:33
TheXceptionwhen activating hardware float13:34
TheXceptionmabe someone knows how to display last asm instruction on arm with gdb?13:34
TheXceptiongot it13:54
*** lizardo has joined #maemo-devel13:55
*** TheXception is now known as TheXception|off14:05
*** lmoura has quit IRC14:10
*** appnss is now known as ShadSEC14:11
*** JPohlmann has joined #maemo-devel14:27
MNZah, just about ready to start coding now15:11
*** JPohlmann has quit IRC15:15
*** qknight_ is now known as qknight15:31
*** csaavedra has joined #maemo-devel15:34
*** JPohlmann has joined #maemo-devel15:43
*** achipa has quit IRC15:47
*** lmoura has joined #maemo-devel15:57
*** ShadSEC has quit IRC16:15
*** appnss has joined #maemo-devel16:19
*** silbo has quit IRC16:28
*** TheXception|off is now known as TheXception16:46
*** ruskie has quit IRC17:05
*** ruskie has joined #maemo-devel17:06
*** JPohlmann has quit IRC17:08
*** lmoura has quit IRC17:12
*** amigadave has quit IRC17:13
*** lmoura has joined #maemo-devel17:17
*** _0x47 has quit IRC17:30
*** TheXception is now known as TheXception|off17:52
*** alvaro__ has quit IRC17:59
*** alvaro__ has joined #maemo-devel18:02
*** silbo_ has joined #maemo-devel18:18
*** pH5 has joined #maemo-devel18:36
*** silbo_ has quit IRC19:08
*** silbo_ has joined #maemo-devel19:10
*** shinkamui has joined #maemo-devel19:17
*** kamui__ has quit IRC19:21
*** TheXception|off is now known as TheXception19:25
*** MNZ has quit IRC19:42
*** MNZ has joined #maemo-devel19:47
*** Ian-- has quit IRC19:56
MNZDocScrutinizer, anything new about dmix and friends?20:52
DocScrutinizernope. Anything pending?20:52
MNZnot really20:52
MNZbeen trying to figure out how things work in the alsa plugin domain. I was under the (naive) impression that it would be simpler than this :D20:53
MNZmaking dmix do that thing we talked about should be a matter of few lines (I think...) it's just a question of where20:54
DocScrutinizer:nod:20:54
MNZAnd there's a hell of a lot of config flying around between functions in a simple aplay -> plug -> dmix setup :/20:55
DocScrutinizerI once added a patch to file plugin, to read out file content paced by input stream rather than unpaced 'max rate', but 'upstream' thought that's not how file plugin should work and rejected it. Anyway was really simple20:56
*** mirr0r has quit IRC21:00
*** csaavedra has quit IRC21:01
*** mirr0r has joined #maemo-devel21:01
DocScrutinizeranyway that's been really spoiling my day to find a '"arecord -D myinfile:'path/to/au' recordfile.wav" to finish in zero time, despite a several MB sized .au was `recorded' into recordfile.wav21:02
*** silbo_ has quit IRC21:16
*** edisson has joined #maemo-devel21:40
*** Milhouse has quit IRC21:49
*** silbo_ has joined #maemo-devel21:54
*** MNZ has quit IRC21:55
*** onen|openBmap has joined #maemo-devel21:59
*** csaavedra has joined #maemo-devel22:12
*** Milhouse has joined #maemo-devel22:16
*** Milhouse has joined #maemo-devel22:16
*** fecub has joined #maemo-devel23:34
*** fecub has quit IRC23:36
*** fecub has joined #maemo-devel23:37
*** SpeedEvil has quit IRC23:45
*** silbo_ has quit IRC23:48
*** onen|openBmap has quit IRC23:50
*** lizardo has quit IRC23:54
*** SpeedEvil has joined #maemo-devel23:57
« Tuesday, 2010-09-14 Index Thursday, 2010-09-16 »

Generated by irclog2html.py 2.15.1 by Marius Gedminas - find it at mg.pov.lt!