**** BEGIN LOGGING AT Sat Jan 14 02:59:57 2012 Jan 14 08:45:30 Hi there, see: http://talk.maemo.org/showpost.php?p=1148886&postcount=668. Is the proposition of sponsoring server for few months still actual? Jan 14 11:34:55 mickeyl, ping Jan 14 11:37:14 freesmartphone.org: 03GNUtoo 07gnutoo/alsaloop * ra144d424ec1a 10cornucopia/fsoaudiod/src/plugins/gsmvoice_alsa_forwarder/alsaloop.c: Jan 14 11:37:14 freesmartphone.org: fsoaudiod: gsmvoice_alsa_forwarder: make alsaloop exit without exiting fsoaudiod Jan 14 11:37:15 freesmartphone.org: Without that fix fsoaudiod quits when alsaloop exits Jan 14 11:37:15 freesmartphone.org: Signed-off-by: Denis 'GNUtoo' Carikli Jan 14 11:37:15 mrmoku, ping Jan 14 11:49:41 freesmartphone.org: 03GNUtoo 07gnutoo/alsaloop * re053d3545a4f 10cornucopia/fsoaudiod/src/plugins/gsmvoice_alsa_forwarder/plugin.vala: Jan 14 11:49:41 freesmartphone.org: fsoaudiod: gsmvoice_alsa_forwarder: add mutex to protect threads and add more logging Jan 14 11:49:41 freesmartphone.org: Signed-off-by: Denis 'GNUtoo' Carikli Jan 14 12:08:03 GNUtoo: pong Jan 14 12:08:05 http://www.pastie.org/private/d2g6818lmevyx54todfgva Jan 14 12:08:20 I've a problem with my current alsaloop Jan 14 12:08:29 I start the first call Jan 14 12:08:31 it works fine Jan 14 12:08:34 I stop it Jan 14 12:08:40 and restart a call Jan 14 12:08:42 silence Jan 14 12:08:48 because alsaloop exits Jan 14 12:09:04 -P plughw:0 -C hw:1 -S 0 -f S16_LE -r 8000 -c 1 -T -1 -A 3 -t 500000 -vvvv Jan 14 12:09:11 here's my alsaloop.cfg Jan 14 12:10:28 *alsaloop.conf Jan 14 12:10:45 mrmoku, do you see something in the pastie? Jan 14 12:14:29 GNUtoo: yeah Jan 14 12:14:44 on the first call this comes after the release: Jan 14 12:14:46 *do you see something strange Jan 14 12:14:51 2012-01-03T05:26:05.438358Z [INFO] AlsaloopForwarder <>: Alsaloop exited Jan 14 12:14:57 yes it's normal Jan 14 12:15:03 on the second call Alsaloop exits just before the call is active Jan 14 12:15:08 yes Jan 14 12:15:13 It's not normal Jan 14 12:15:20 ahh, yeah, you even wrote it :) Jan 14 12:15:21 I wonder why it does that Jan 14 12:15:41 * mrmoku pulls the code Jan 14 12:16:31 * GNUtoo wants the GTA04 supported quickely that's why he ask for some help Jan 14 12:17:38 the configs: http://www.pastie.org/private/mn0sj6r438inxlozj5za0g Jan 14 12:18:15 how to see if there was a segfault? Jan 14 12:18:18 catchsegv? Jan 14 12:18:22 gdb? Jan 14 12:18:37 or would the whole thing segfault instead Jan 14 12:19:05 I guess you can't catch a segv Jan 14 12:20:01 ah yes it can ignore it Jan 14 12:20:02 hmmm Jan 14 12:21:54 GNUtoo: hmm... as far as I see alsoloop does *not* ignore SIGSEGV Jan 14 12:22:04 ok Jan 14 12:22:44 so maybe I change to void->int and look at the exit code Jan 14 12:22:44 ? Jan 14 12:22:51 hmmm no Jan 14 12:22:55 won't work Jan 14 12:23:14 since if an exit was called it woul exit fsoaudiod too Jan 14 12:26:57 GNUtoo: one suboptimal thing is... when start_forwarder exits by itself (without calling stop) alsaLoopThread won't be reset to null Jan 14 12:27:17 that's why you get the last Jan 14 12:27:18 2012-01-03T05:26:11.761234Z [ERROR] AlsaloopForwarder <>: FIXME: Are multiple calls at the same time supported by the modem driver? Jan 14 12:27:21 I think Jan 14 12:27:52 no, it's for another reason: it's because of that: Jan 14 12:28:17 case FreeSmartphone.GSM.CallStatus.OUTGOING: Jan 14 12:28:18 case FreeSmartphone.GSM.CallStatus.ACTIVE: Jan 14 12:28:24 but still it's a problem Jan 14 12:28:34 yeah, that's why you call start *twice* Jan 14 12:28:40 so I'll set it to null too Jan 14 12:28:42 no idea Jan 14 12:28:55 that won't help to make it *not* exit though Jan 14 12:28:58 *I've no idea why it's necessary Jan 14 12:29:04 mickeyl, knows Jan 14 12:29:13 for the OUTGOING and ACTIVE Jan 14 12:29:35 I'll fix the forgetten = null Jan 14 12:30:02 what happens if a thread calls exit? Jan 14 12:30:15 no idea Jan 14 12:33:55 Threads terminate by explicitly calling pthread_exit, by letting the function return, or by a call to the function exit which will terminate the process including any threads. Jan 14 12:34:36 so if alsaloop exits fsoaudiod would too Jan 14 12:34:45 that's fixed Jan 14 12:34:49 I removed some exit Jan 14 12:35:54 http://git.freesmartphone.org/?p=cornucopia.git;a=blobdiff;f=fsoaudiod/src/plugins/gsmvoice_alsa_forwarder/alsaloop.c;h=df0f2ff4277085004aa4812f3b37325336ba3b28;hp=52ee31234dfebcae03475d739e6b19fcaffe8133;hb=a144d424ec1a6e6856e67e714e67bfc4b965f86d;hpb=bac004e1fd693f02689ef8e650e77a70595d851b Jan 14 12:40:41 GNUtoo: make those too a return and check for the value? Jan 14 12:40:46 s/too/two/ Jan 14 12:40:46 mrmoku meant: GNUtwo: make those too a return and check for the value? Jan 14 12:40:51 heh Jan 14 12:40:59 s/ too / two / Jan 14 12:41:09 ok good idea Jan 14 12:41:29 freesmartphone.org: 03GNUtoo 07gnutoo/alsaloop * ra6cb4e19ec46 10cornucopia/fsoaudiod/src/plugins/gsmvoice_alsa_forwarder/plugin.vala: Jan 14 12:41:29 freesmartphone.org: fsoaudiod: gsmvoice_alsa_forwarder: fix restart when thread exits by itself Jan 14 12:41:29 freesmartphone.org: Thanks to mrmoku on #openmoko-cdevel Freenode irc channel for noticing it: Jan 14 12:41:29 freesmartphone.org: Jan 14 13:26:40 GNUtoo: one suboptimal thing is... when start_forwarder exits by itself (without calling stop) alsaLoopThread won't be reset to null Jan 14 12:41:29 freesmartphone.org: Signed-off-by: Denis 'GNUtoo' Carikli Jan 14 12:41:34 hopefully it gives some insight :) Jan 14 12:52:23 2012-01-03T06:11:05.859103Z [INFO] AlsaloopForwarder <>: Alsaloop exited with status 0 Jan 14 12:52:26 doesn't help much Jan 14 12:55:48 freesmartphone.org: 03GNUtoo 07gnutoo/alsaloop * r0e4a0f3824c9 10cornucopia/fsoaudiod/src/plugins/gsmvoice_alsa_forwarder/ (alsaloop.c plugin.vala): Jan 14 12:55:48 freesmartphone.org: fsoaudiod: gsmvoice_alsa_forwarder: alsaloop.c: switch from exit to return Jan 14 12:55:48 freesmartphone.org: without that the exit code could make fsoaudiod exit instead of just the forwarder. Jan 14 12:55:48 freesmartphone.org: Thanks mrmoku on #openmoko-cdevel on Freenode IRC server for the idea: Jan 14 12:55:48 freesmartphone.org: Jan 14 13:40:23 GNUtoo: make those too a return and check for the value? Jan 14 12:55:48 freesmartphone.org: Signed-off-by: Denis 'GNUtoo' Carikli Jan 14 12:58:29 maybe gdb? Jan 14 12:58:40 or adding printf Jan 14 13:00:24 removing OUTGOING is much better CPU usage wise Jan 14 13:00:43 it doesn't try to use the sound card when it's not ready Jan 14 13:01:25 mrmoku, it exits at that line: Jan 14 13:02:37 921 from here: Jan 14 13:02:47 http://git.freesmartphone.org/?p=cornucopia.git;a=blob;f=fsoaudiod/src/plugins/gsmvoice_alsa_forwarder/alsaloop.c;h=b0d0c7a99fe94bf735023d93b60cb570f4ff9a0c;hb=refs/heads/gnutoo/alsaloop Jan 14 13:33:27 GNUtoo: hmm Jan 14 13:33:49 I'm at no means a threading expert :/ Jan 14 13:34:24 pthread_join is waiting for the thread to finish? Jan 14 13:34:37 pthread_join - wait for termination of another thread Jan 14 13:34:40 yeah, it does Jan 14 13:36:23 yes it's the join thing Jan 14 13:36:32 *classic join Jan 14 13:38:31 hmm... you don't get the exit code of the threads Jan 14 13:38:47 maybe add some output of it in the pthread_join loop? Jan 14 13:38:58 maybe I get it Jan 14 13:39:02 quit = 1; Jan 14 13:39:09 then send_to_all(signal); Jan 14 13:39:12 but then.... Jan 14 13:39:14 pthread[k].exitcode Jan 14 13:39:22 maybe quit is still one next time Jan 14 13:39:51 ahh, that might make sense :-) Jan 14 13:39:57 indeed Jan 14 13:39:57 as alsaloop is a one time thing Jan 14 13:40:02 hehe, good catch :) Jan 14 13:40:05 as it's global..... Jan 14 13:40:11 and I only call functions Jan 14 13:40:15 and it never exits Jan 14 13:40:18 so it should be that Jan 14 13:40:26 so quit = 0 in forwarder_start and all works :-) Jan 14 13:40:29 yes Jan 14 13:49:02 freesmartphone.org: 03GNUtoo 07gnutoo/alsaloop * rd2a6427b75a6 10cornucopia/fsoaudiod/src/plugins/gsmvoice_alsa_forwarder/alsaloop.c: (log message trimmed) Jan 14 13:49:02 freesmartphone.org: fsoaudiod: gsmvoice_alsa_forwarder: make succesive calls work. Jan 14 13:49:02 freesmartphone.org: Without that fix alsaloop exited right after starting after Jan 14 13:49:02 freesmartphone.org: the first call: that was caused by the global quit variable: Jan 14 13:49:03 freesmartphone.org: it was set to one after the first call to make alsaloop exit, Jan 14 13:49:03 freesmartphone.org: but it stayed one on the successive calls because it was initialized Jan 14 13:49:04 freesmartphone.org: only when alsaloop started for the first time. Jan 14 13:49:28 mickeyl, hi Jan 14 13:51:46 now that it works here's what's left: Jan 14 13:51:55 * convert fsoaudiod to GPLv2 or later Jan 14 13:52:02 instead of LGPLv2.1 Jan 14 13:52:24 * integrate the config files Jan 14 13:52:47 * make review the fsogsmd dbus signal handling by mickeyl Jan 14 13:53:12 then some non-related stuff are needed still for beeing able to use it as a phone: Jan 14 13:53:18 * migrate to the last 3.2 Jan 14 13:53:29 * handle the rfkill resources Jan 14 13:53:39 and it should be pretty good at that poing Jan 14 13:53:41 *point Jan 14 13:54:09 :)( Jan 14 13:54:17 don't we already have the latest 3.2 ? Jan 14 13:54:24 I thought slyon bumped it Jan 14 13:54:27 no we don't Jan 14 13:54:29 ah? Jan 14 13:54:44 because there is a 3.2 based on released 3.2 kernel from neil Jan 14 13:54:51 with improved things Jan 14 13:55:04 yeah, his 3.2 stable branch Jan 14 13:55:07 * mrmoku checks Jan 14 13:55:14 yes that one Jan 14 13:55:30 meta-openmoko: gta04: switch to neil's 3.2-gta04 stable branch Jan 14 13:55:42 3fbfffe30426a5d39f5f0817d65e3e3afcea6017 Jan 14 13:55:45 ok nice Jan 14 13:55:52 I'll opkg upgrade then Jan 14 13:56:02 GNUtoo: btw. great work :-D Jan 14 13:56:56 thanks but let's try it in real conditions before saying that Jan 14 13:57:01 ahh... we have to re-add fsoaudiod for gta04 now Jan 14 13:57:06 like when it's deployed to users Jan 14 13:57:20 we dropped it for the release Jan 14 13:57:23 ok Jan 14 13:58:08 ah and last thing Jan 14 13:58:14 we need fsodeviced scenarios Jan 14 13:58:20 because sh spk.sh is not good Jan 14 14:00:04 I did a stereoout one Jan 14 14:00:12 not committed yet Jan 14 14:00:38 wanted to do a headset one next... until I discovered that I have no working headset for gta04 Jan 14 14:01:30 ok lol Jan 14 14:02:43 so we need a fsoaudiod.conf for gta04 that enables only your forwarder plugin Jan 14 14:02:50 and scenarios for fsodeviced Jan 14 14:03:57 hmmm Jan 14 14:03:59 2012-01-03T07:22:31.786260Z [INFO] AlsaloopForwarder <>: Alsaloop exited with status 0 Jan 14 14:03:59 Segmentation fault Jan 14 14:04:03 quite rare tough Jan 14 14:22:08 mrmoku, latency is too high with my current settings Jan 14 14:22:14 with recording->modem Jan 14 14:22:28 we need better settings than that: Jan 14 14:22:48 -P hw:1 -C plughw:0 -S 0 -f S16_LE -r 8000 -c 1 -T -1 -A 3 -t 500000 Jan 14 14:23:15 does someone have free calls? Jan 14 14:23:24 if so could he try to do it Jan 14 14:23:49 because each call cost so it would cost a lot at the end for the latency test Jan 14 14:25:35 GNUtoo: there are some gratis numbers you can call? Jan 14 14:25:36 GNUtoo: my image just finished and I have free calls Jan 14 14:25:46 mrmoku, ok nice Jan 14 14:26:10 lindi-, I would need some fast echo thing Jan 14 14:27:03 anyway I need to do some university work Jan 14 14:27:19 so it sould be good if someone else would continue the work Jan 14 14:49:04 heyho Jan 14 14:50:44 hey morphis Jan 14 15:02:39 angelox|laptop: you are still without a suitable phone for aurora development? Jan 14 15:03:34 morphis: yes, unfortunately Jan 14 15:03:50 sad :( Jan 14 15:05:42 http://lwn.net/Articles/468659/ Jan 14 15:06:09 for STE andridiot == linux Jan 14 15:06:20 morphis: yep, but next month i'm going to buy a samsung galaxy 5... 84EUR phone, 163EUR in Brazil.. Jan 14 15:06:43 oh great Jan 14 15:06:46 * DocScrutinizer moos at mrmoku & GNUtoo Jan 14 15:06:57 & morphis Jan 14 15:08:57 DocScrutinizer: why that? the drivers and the other CAIF work from STE seems to be very unrelated to android and suites pretty well into the kernel Jan 14 15:10:03 yes, I oly meant internally when they say "linux" they actually always think of the already existing andridiot devices that use Thor chips Jan 14 15:10:44 and andridiot is a bit less picky about GPL issues Jan 14 15:10:50 I guess Jan 14 15:11:14 yes Jan 14 15:11:35 linux == android seems to be a very common term these days Jan 14 15:12:09 there's a problem with http://igloocommunity.org/gitweb/?p=bsp/b2r2lib.git;a=blob;f=src/blt_b2r2.c;h=02f3bc13cce0000afc8b6ff683924cb5f23a94bd;hb=HEAD forbidding proper gfx in proper linux, AIUI Jan 14 15:12:32 e.g for snowball board Jan 14 15:13:20 Good day all, is there any alternative for http://code.google.com/p/android-on-freerunner/ this address is forbiiden for us(iranian) Jan 14 15:16:34 hmm, I'm not sure android on freerunner is an interesting topic or viable alternative - anyway what do you mean by 'alternative' to the URL? another URL with same content? Jan 14 15:17:20 DocScrutinizer yes Jan 14 15:19:01 wana install android on FR , so http://code.google.com/p/android-on-freerunner/ is required for sure yes? Jan 14 15:20:18 hmm, dunno if that will help: Jan 14 15:20:23 jr@halley:~> host code.google.com Jan 14 15:20:24 code.google.com is an alias for code.l.google.com. Jan 14 15:20:26 code.l.google.com has address 173.194.65.102 Jan 14 15:20:27 code.l.google.com has address 173.194.65.113 Jan 14 15:20:29 code.l.google.com has address 173.194.65.138 Jan 14 15:20:30 code.l.google.com has address 173.194.65.139 Jan 14 15:20:32 code.l.google.com has address 173.194.65.100 Jan 14 15:20:33 code.l.google.com has address 173.194.65.101 Jan 14 15:20:58 so? Jan 14 15:21:23 do you mean ? http://173.194.65.101/p/android-on-freerunner/ ? Jan 14 15:21:26 no does not work Jan 14 15:21:54 well, maybe use a open proxy? Jan 14 15:22:22 maybe the only way .. Jan 14 15:22:41 maybe /join #android-on-freerunner Jan 14 15:22:54 do you think debian + android are Good choices Jan 14 15:22:57 dunno if that's a living channel Jan 14 15:22:59 tried there too Jan 14 15:24:14 debian for sure is a good choice usually - andridiot.. not so much, in my book Jan 14 15:24:24 DocScrutinizer which os do you offer on FR ? which of them has more and variable applications ? Jan 14 15:24:57 hmm, I always used SHR since it became somewhat operating Jan 14 15:25:45 DocScrutinizer but methink android has more variable applications ? Jan 14 15:25:46 then otoh I prefer decent programs over 'apps' Jan 14 15:26:41 otoh? Jan 14 15:28:22 DocScrutinizer otoh? anda http://code.google.com/p/android-on-freerunner/ is this page owner USA ? Jan 14 15:29:08 DocScrutinizer, hi Jan 14 15:29:16 ~otoh Jan 14 15:29:18 methinks otoh is On The Other Hand. It's YAFA. Jan 14 15:30:09 DocScrutinizer what do you mean of > I prefer decent programs over 'apps' ? and why ? Jan 14 15:30:12 (owner) obviously google Jan 14 15:31:23 alabd: I don't like andridiot for a number of reasons. And I don't like the 100.000 fart apps it has, while decent stuff is missing Jan 14 15:33:12 for SHR you can .configure && make basically any arbitrary linux app, without too much adaptions to make it work or even compile on target. For andridiot that's a bit different Jan 14 15:34:21 s/work or even compile on/work on or even compile for/ Jan 14 15:34:22 DocScrutinizer meant: for SHR you can .configure && make basically any arbitrary linux app, without too much adaptions to make it work on or even compile for target. For andridiot that's a bit different Jan 14 15:37:51 alabd, maybe use tor? if it's blocked get some bridges address Jan 14 15:38:25 get a gmail account and send a mail to bridges@torproject Jan 14 15:38:30 only gmail account work Jan 14 15:39:17 GNUtoo: thanks no need Jan 14 15:39:20 ok Jan 14 15:39:45 DocScrutinizer is owner of google USA? Jan 14 15:40:40 no idea Jan 14 15:41:09 anda how do you compere debian and shr ? which one why? Jan 14 15:41:10 probably google isn't 'owned' by anybody Jan 14 15:42:12 DocScrutinizer of course Google is company that is in USA Jan 14 15:42:37 shr uses enlightenment and has some finger friendly telephony, settings and other applications Jan 14 15:42:54 debian can look very different depending on what GUI you install Jan 14 15:43:15 what GUIs are there ? Jan 14 15:43:16 you can even have the same software than on SHR but with older version Jan 14 15:44:00 I only tried illume on debian so I don't know.... Jan 14 15:44:08 and I run SHR Jan 14 15:44:26 but debian has desktop apps and SHR not Jan 14 15:45:19 shr has some desktop apps Jan 14 15:45:27 debian has more packages Jan 14 15:45:38 but some debian packages are not obtimized for the freerunner Jan 14 15:45:46 shr has some obtimized packages Jan 14 15:46:05 alabd, can't you try both? Jan 14 15:46:42 and android on freerunner lacks the possibility to run packages with native code btw, unless you recompile them for armv4 Jan 14 15:48:42 GNUtoo: you mean , few percent of android apps are availble on SHR? Jan 14 15:49:20 sorry Jan 14 15:49:30 GNUtoo: you mean , few percent of android apps are availble on fr-android? Jan 14 15:50:08 alabd, I mean you don't have access to the market but even if you install FDroid, a free version of the market, not all apps will work, only the ones that are pure java Jan 14 15:50:39 http://f-droid.org/repository/browse/ Jan 14 15:50:46 you will have a subset of that Jan 14 15:51:21 ohum thanks have tested SHR before but not debian and android on FR Jan 14 15:51:29 ok Jan 14 15:51:40 debian is long to install.... Jan 14 15:51:48 SHR need long time to boot and has not some software that android has .. Jan 14 15:52:01 alabd, what kind of software? Jan 14 15:52:07 SHR is fast to boot nowadays Jan 14 15:52:24 GNUtoo: for example on android we have call manager Jan 14 15:52:45 what's that? Jan 14 15:52:56 you mean for the missed calls etc...? Jan 14 15:53:07 there is ffphonelog in SHR Jan 14 15:54:24 no Jan 14 15:54:31 it can reject some numbers .. Jan 14 15:54:40 ah ok Jan 14 15:55:23 GNUtoo: can we install all free apps on android site ? Jan 14 15:55:50 you will need to recompile some of them Jan 14 15:56:03 and firefox won't work on armv4 Jan 14 15:57:37 note that there is a free SDK for android in the replicant project Jan 14 15:57:49 you may need that Jan 14 15:58:15 but I guess you would need to pass the armv4 flags for gcc Jan 14 15:58:27 to be able to run the binaries on the freerunner Jan 14 15:58:42 alabd, also there should be an irc channel for android on freerunner Jan 14 15:58:53 #android-on-freerunner Jan 14 15:59:06 maybe they know better than me Jan 14 15:59:58 GNUtoo: yes but there are few people Jan 14 16:00:01 anyway Jan 14 16:00:07 have SDK Jan 14 16:00:21 but what should it be used for Jan 14 16:00:41 for recompiling for armv4 the packages Jan 14 16:01:24 angelox|laptop: if you are going to buy a Galaxy S, we already have modem support for it in fsogsmd Jan 14 16:01:31 so only OE support is missing Jan 14 16:01:38 GNUtoo: for recompiling we need electronic board why ? Jan 14 16:02:03 alabd, ? what do you mean by electronic board? Jan 14 16:02:09 SDK Jan 14 16:02:19 because it contains a toolchain Jan 14 16:02:22 for android Jan 14 16:02:30 and it contains the java part of the toolchain Jan 14 16:02:36 that is not present elsewhere Jan 14 16:02:45 but android is designed for arm or not ? Jan 14 16:03:42 I'll explain Jan 14 16:04:11 do you know i386,i486,i586,i686 for the desktop/laptops Jan 14 16:04:40 if you try to run an i686 binary on an i386 cpu, it won't work Jan 14 16:04:42 right? Jan 14 16:04:48 arm has something similar Jan 14 16:05:09 the more recents CPU are like i686 Jan 14 16:05:15 and the freerunner is like i386 Jan 14 16:05:27 and android was made for more recent CPU Jan 14 16:05:56 so people doing the packages assume like i486 minimum but the freerunner is like i386 Jan 14 16:06:07 so their packages won't work on the freerunner Jan 14 16:06:32 a very nice explanation :-) Jan 14 16:06:38 basically it will make an error like: illegal instruction Jan 14 16:06:43 thanks Jan 14 16:08:05 thanks Jan 14 16:08:51 so armv4,armv5,armv6,armv7 are like i386,i486,i586,i686... Jan 14 16:16:03 GNUtoo: so maybe it's better to install android +debian + shr on Fr beside each other , will install SHR on fr memory , and what memory is required for debian + android +their common apps Jan 14 16:21:49 maybe SHR+debian Jan 14 16:21:59 or maybe do that: Jan 14 16:22:05 install android on memory Jan 14 16:22:13 then get 2 microsd card Jan 14 16:22:16 one for debian Jan 14 16:22:25 (you need something like 4GB ) Jan 14 16:22:31 and one for SHR Jan 14 16:50:14 GNUtoo: you mean its not possible to install debian + SHR on one memory ? Jan 14 16:50:44 it is Jan 14 16:50:56 but I'm not sure if it's possible to install android on microsd Jan 14 16:51:40 GNUtoo: so why did you say one for debian and one for SHR ? Jan 14 16:52:19 because if you want android+debian+shr you need 3 install type Jan 14 16:52:23 android goes on NAND Jan 14 16:52:36 and debian+shr can't go on nand if there is already android Jan 14 16:52:48 debian is too big for NAND Jan 14 16:52:58 but qtmoko which uses debian can go on NAND Jan 14 16:53:06 it's based on qt for framebuffer Jan 14 16:53:20 and SHR can go on both too but if the NAND is already taken by android... Jan 14 16:55:12 GNUtoo: am saying why you said 2 memory for debian and shr why not one memory for both ? Jan 14 16:57:02 if you partitionate well and have a system for choosing them..... Jan 14 16:57:19 so you would need uboot Jan 14 16:57:41 or to play with QI's files in /boot/ Jan 14 17:01:17 GNUtoo: so it can be done for sure , is one 3.75 gig enough for debian ? Jan 14 17:01:39 you could try Jan 14 17:01:47 it depend on how much apps you want etc... Jan 14 17:01:50 no wana buy should be sure Jan 14 17:02:06 ah ok Jan 14 17:02:10 hmmm Jan 14 17:02:12 I'm not sure Jan 14 17:02:19 you should ask people running debian Jan 14 17:02:20 like lindi- Jan 14 17:02:32 remebr when had SHR on nand it was enough for me Jan 14 17:03:08 yes but I don't know well debian Jan 14 17:03:14 so ask debian people Jan 14 17:07:46 GNUtoo: has SHR and have used it before on FR and now after a while FR stops at booting page , what do you think ? Jan 14 17:08:26 I don't understand Jan 14 17:08:37 what is booting *page* Jan 14 17:08:58 note that if you use SHR on microsd you should follow the install guide Jan 14 17:09:00 that is to say: Jan 14 17:09:04 tar xvjpf won't work Jan 14 17:09:17 GNUtoo, I guess it's boot splash image what he means Jan 14 17:09:25 you need tar xvjpf ./tarball.tar.bz2 --numeric-owner Jan 14 17:09:26 gafternoon btw :) Jan 14 17:09:30 hi pespin Jan 14 17:09:39 alabd, you need to wait, the first boot is very long Jan 14 17:09:50 alabd, but be sure you used --numeric-owner Jan 14 17:10:04 else some stuff like telephony won't work Jan 14 17:10:19 no that stops ...this is not first time am using that shr ...the page that has SHR image and a line at bottom ..is booting .. Jan 14 17:10:52 alabd, is it on NAND or microsd? Jan 14 17:10:56 nand Jan 14 17:11:17 ok Jan 14 17:11:20 then wait a bit Jan 14 17:11:22 it can be long Jan 14 17:12:19 man am saying it' unormal , some hours is enough Jan 14 17:12:31 ah ok Jan 14 17:12:34 says..please wait,booting.. Jan 14 17:12:40 I tought you just installed it Jan 14 17:12:46 no Jan 14 17:12:46 like 5 min ago Jan 14 17:12:49 it was ok Jan 14 17:12:52 before Jan 14 17:12:58 ask mrmoku Jan 14 17:13:14 after a some weeks wanted use FR .. Jan 14 17:13:28 alabd, do you use QI or uboot? Jan 14 17:14:19 GNUtoo: don't know how to be sure Jan 14 17:14:30 it's rather simple: Jan 14 17:14:50 do you see some images before the splash boot screen? Jan 14 17:15:04 no Jan 14 17:15:05 like something that ressemble to what's in NOR Jan 14 17:15:08 then you have QI Jan 14 17:15:10 then do that Jan 14 17:15:14 remove the battery Jan 14 17:15:24 k Jan 14 17:15:29 wait a second Jan 14 17:15:35 k Jan 14 17:15:36 that won't work Jan 14 17:15:38 hmmm Jan 14 17:15:58 because you are on NAND it's difficult to pass argument to the bootloader Jan 14 17:16:05 s/bootloader/kernel/ Jan 14 17:16:06 GNUtoo meant: because you are on NAND it's difficult to pass argument to the kernel Jan 14 17:16:21 hmmm Jan 14 17:17:19 GNUtoo: what do you mean of splash screen ? Jan 14 17:18:38 http://www.datamation.com/imagesvr_ce/4754/developer-apocalypse.jpg Jan 14 17:19:45 splash screen ? this GNUtoo http://wiki.shr-project.org/trac/raw-attachment/wiki/LogoContest/alphalog-shr-3.png ? Jan 14 17:24:01 alabd, let me look Jan 14 17:27:23 morphis: no, i'm going to buy a Samsung Galaxy 5 (less expensive) Jan 14 17:36:23 GNUtoo: with battery fr does not turns on at all , Jan 14 18:14:34 freesmartphone.org: 03morphis 07aurora * r73eab9f61ba5 10/aurora-daemon/aurora/ (20 files in 3 dirs): Jan 14 18:14:34 freesmartphone.org: aurora-daemon: drop source of the old python implementation Jan 14 18:14:34 freesmartphone.org: Signed-off-by: Simon Busch Jan 14 18:14:38 freesmartphone.org: 03morphis 07aurora * r200b3e4cbf0b 10/aurora-daemon/ (527 files in 9 dirs): Jan 14 18:14:38 freesmartphone.org: aurora-daemon: drop imported symbian stuff from qt-components Jan 14 18:14:38 freesmartphone.org: Signed-off-by: Simon Busch Jan 14 18:14:38 freesmartphone.org: 03morphis 07aurora * r0635567f6047 10/aurora-daemon/ (165 files in 28 dirs): Jan 14 18:14:38 freesmartphone.org: aurora-daemon: drop QML stuff Jan 14 18:14:38 freesmartphone.org: Signed-off-by: Simon Busch Jan 14 18:14:39 freesmartphone.org: 03morphis 07aurora * re54a3246c2b7 10/aurora-daemon/ (14 files in 2 dirs): Jan 14 18:14:39 freesmartphone.org: aurora-daemon: remove all QML relevant code from daemon Jan 14 18:14:40 freesmartphone.org: Signed-off-by: Simon Busch Jan 14 18:14:40 freesmartphone.org: 03morphis 07aurora * r18ffbc8a173b 10/aurora-daemon/src/bin/applicationinfo.cpp: Jan 14 18:14:44 freesmartphone.org: aurora-daemon: reduce debug output Jan 14 18:14:44 freesmartphone.org: Signed-off-by: Simon Busch Jan 14 18:14:44 freesmartphone.org: 03morphis 07aurora * r9f34bca29a18 10/aurora-daemon/ (68 files in 9 dirs): Jan 14 18:14:44 freesmartphone.org: aurora-daemon: restructure source code layout Jan 14 18:14:44 freesmartphone.org: Signed-off-by: Simon Busch Jan 14 18:14:44 freesmartphone.org: 03morphis 07aurora * ra300cd20a2ea 10/aurora-daemon/ (10 files in 2 dirs): Jan 14 18:14:57 freesmartphone.org: 03morphis 07aurora * r24c7730cac33 10/aurora-daemon/components/CellularNetworkStatusIndicator.qml: Jan 14 18:14:57 freesmartphone.org: aurora-daemon: components: update image source of network indicator whenever something changes Jan 14 18:14:57 freesmartphone.org: Signed-off-by: Simon Busch Jan 14 18:15:00 freesmartphone.org: 03morphis 07aurora * r6e2116cd2435 10/aurora-daemon/runtime/applicationinfo.h: Jan 14 18:15:00 freesmartphone.org: aurora-daemon: runtime: expose path of application to QML Jan 14 18:15:00 freesmartphone.org: Signed-off-by: Simon Busch Jan 14 18:15:00 freesmartphone.org: 03morphis 07aurora * r35679250b1bf 10/aurora-daemon/applications/launcher/main.qml: Jan 14 18:15:01 freesmartphone.org: aurora-daemon: launcher: don't use global variable currentApp for local assignment Jan 14 18:15:01 freesmartphone.org: Signed-off-by: Simon Busch Jan 14 18:15:02 freesmartphone.org: 03morphis 07aurora * r4047cf386bc2 10/aurora-daemon/components/ (4 files): Jan 14 18:15:02 freesmartphone.org: aurora-daemon: components: adjust cellular and power status indicators Jan 14 18:15:03 freesmartphone.org: Signed-off-by: Simon Busch Jan 14 18:15:04 freesmartphone.org: 03morphis 07aurora * rb7b420feda03 10/aurora-daemon/components/ (Button.qml Makefile.am TabGroup.qml qmldir): Jan 14 18:15:04 freesmartphone.org: aurora-daemon: components: add tab group and buttom components Jan 14 18:15:05 freesmartphone.org: Signed-off-by: Simon Busch Jan 14 18:15:05 freesmartphone.org: 03morphis 07aurora * r8ccf806f5fb6 10/aurora-daemon/ (7 files in 2 dirs): Jan 14 18:15:06 freesmartphone.org: aurora-daemon: applications: phone: implement first bits of the phone application Jan 14 18:15:45 alabd, what do you mean? Jan 14 18:16:02 freesmartphone.org: Signed-off-by: Simon Busch Jan 14 18:16:02 freesmartphone.org: 03morphis 07aurora * ra1becebb0b70 10/aurora-daemon/components/ (PageStackWindow.qml Window.qml enums.h plugin.cpp plugin.h): Jan 14 18:16:02 freesmartphone.org: aurora-daemon: components: some more work regarding page stack handling Jan 14 18:16:02 freesmartphone.org: Signed-off-by: Simon Busch Jan 14 18:16:03 freesmartphone.org: 03morphis 07aurora * r26ed961699ca 10/aurora-daemon/applications/phone/ (7 files): Jan 14 18:16:03 freesmartphone.org: aurora-daemon: bring back all phone application pages to get it working again Jan 14 18:16:04 freesmartphone.org: Signed-off-by: Simon Busch Jan 14 18:16:28 mrmoku, ping Jan 14 18:16:47 GNUtoo FR does not wakeup with battery at all Jan 14 18:17:02 only with usb cable/power Jan 14 18:17:09 alabd, ah then let it charge Jan 14 18:18:13 angelox|laptop: I pushed all aurora stuff now Jan 14 18:18:18 did it before GNUtoo Jan 14 18:18:21 angelox|laptop: I will reintrage your settings app as next step Jan 14 18:18:43 alabd, maybe your battery is dead? Jan 14 18:19:01 GNUtoo maybe how to be sure Jan 14 18:19:42 maybe reading the sysfs for the battery but I'm not sure Jan 14 18:19:46 maybe charge it fully Jan 14 18:19:53 and look in how much time it discharges Jan 14 18:23:17 GNUtoo but have tested fr with nokia BL-5C battery , the same result , hangs on wait for booting ... Jan 14 18:23:44 alabd, ahh you're still talking about booting Jan 14 18:23:57 yes :P Jan 14 18:23:58 I tought you had another problem not related to booting Jan 14 18:24:05 no Jan 14 18:24:09 ok Jan 14 18:24:20 do you have a microsd? Jan 14 18:24:28 try to enable the messages on the kernel Jan 14 18:24:31 with that guide: Jan 14 18:24:38 no Jan 14 18:24:48 http://wiki.openmoko.org/wiki/Qi Jan 14 18:24:53 hmmm Jan 14 18:25:11 are you able to compile yourself a version of QI? Jan 14 18:25:16 or do you need me to do it? Jan 14 18:25:33 it's for modifying the cmdline arguments Jan 14 18:26:14 how to be sure fr uses QI Jan 14 18:27:25 http://en.wikipedia.org/wiki/File:U-boot_openmoko_freerunner.jpg Jan 14 18:27:28 that's u-boot Jan 14 18:27:54 if you have no menu you have QI Jan 14 18:28:04 maybe it's more clear than the last time Jan 14 18:28:19 yes seems it's QI Jan 14 18:29:15 so read the QI page Jan 14 18:29:26 wait a seconfd Jan 14 18:29:38 you don't have already a microsd Jan 14 18:29:44 then where were we Jan 14 18:29:57 are you able to compile yourself a version of QI? Jan 14 18:30:14 didn't do it before Jan 14 18:30:41 ok so I'll do it Jan 14 18:30:55 let me find the source Jan 14 18:31:25 thanks Jan 14 18:33:02 hmmm what's that in QI? Jan 14 18:33:03 .commandline_board_debug = " loglevel=8", Jan 14 18:33:08 how to activate that? Jan 14 18:37:21 freesmartphone.org: 03morphis 07aurora * r5d9352e69dae 10/aurora-daemon/configure.ac: Jan 14 18:37:21 freesmartphone.org: aurora-daemon: don't depend on QtWebKit for now Jan 14 18:37:21 freesmartphone.org: Signed-off-by: Simon Busch Jan 14 18:39:59 http://gnutoo.homelinux.org/downloads/people/alabd/qi-ubi-s3c2442-1.0.2-r7+gitr1+aa934abb5e8771100240e344411ddebacca33ece.udfu Jan 14 18:40:01 try that Jan 14 18:40:06 tell me if it prints more messages Jan 14 18:41:14 GNUtoo you ask me to install new SHR ? Jan 14 18:41:34 what do you mean? Jan 14 18:41:45 the link is a modified QI Jan 14 18:41:51 with loglevel=8 Jan 14 18:42:00 to make it print the messages Jan 14 18:42:02 during boot Jan 14 18:42:19 so we would be able to debug the reason why your image refuses to boot Jan 14 18:42:28 ah wait a second Jan 14 18:42:34 I forgott the userspace part Jan 14 18:42:55 ah no it's there Jan 14 18:43:03 "console=tty0 " Jan 14 18:45:36 GNUtoo thanks wait long time ago installed SHR don't remebr command ..should read Jan 14 18:45:57 alabd, also we don't support old SHR anymore Jan 14 18:46:04 so maybe install a newer shr-core Jan 14 18:46:29 no am talking about working with udfu files Jan 14 18:47:50 yes look in wiki, in the flashing argticle Jan 14 18:47:55 *flashing article Jan 14 18:48:16 alabd, but also what I said is still valid: we don't support old SHR anymore, we don't have the manpower to do so.... Jan 14 18:49:53 don't think it's old Jan 14 18:50:14 ok Jan 14 18:50:17 it's shr-core then Jan 14 18:50:31 http://wiki.openmoko.org/wiki/Flashing_the_Neo_FreeRunner Jan 14 18:50:51 yes at the end there is the commands Jan 14 18:51:00 like for bootloader, etc... Jan 14 18:53:23 # dfu-util -a u-boot -R -D *.udfu Jan 14 18:53:25 ? Jan 14 18:57:17 yes Jan 14 18:58:07 GNUtoo: before that how to go in FR command line ? Jan 14 18:58:22 ? Jan 14 18:58:34 you must boot in NOR Jan 14 18:58:40 to be able to flash it Jan 14 18:59:20 yes but should run dfu-util -a u-boot -R -D *.udfu in my desktop terminal ? Jan 14 18:59:39 yes but before connect your freerunner to USB Jan 14 18:59:42 and boot on NOR Jan 14 18:59:48 so you're on NOR uboot Jan 14 18:59:58 and then quickely execute the command Jan 14 19:02:40 GNUtoo: ok done Jan 14 19:02:51 the power off and boot normally Jan 14 19:03:56 please wait ,booting .. Jan 14 19:04:23 yes but does it say other stuff? Jan 14 19:04:29 like messages? Jan 14 19:04:35 do you have a debug board btw? Jan 14 19:04:57 GNUtoo: thanks seems shr is loading Jan 14 19:05:10 have debug board but not here Jan 14 19:05:11 ah? Jan 14 19:05:13 ok Jan 14 19:05:25 I didn't change a lot Jan 14 19:05:33 GNUtoo: how did you compiled that qi ? Jan 14 19:05:35 I only changed loglevel=4 to loglevel=8 Jan 14 19:05:41 I did that: Jan 14 19:05:55 :P Jan 14 19:05:56 cd /home/gnutoo/embedded/oe/oe-core/oetmps/shr/work/om_gta02-oe-linux-gnueabi/qi-ubi/qi-ubi-1.0.2-r7+gitr1+aa934abb5e8771100240e344411ddebacca33ece/git/src/cpu/s3c2442 Jan 14 19:06:02 vim gta02.c Jan 14 19:06:11 I changed loglevel=4 to loglevel=8 Jan 14 19:06:25 in .commandline_board Jan 14 19:06:32 then I went there: Jan 14 19:06:48 /home/gnutoo/embedded/oe/oe-core/oetmps/shr/work/om_gta02-oe-linux-gnueabi/qi-ubi/qi-ubi-1.0.2-r7+gitr1+aa934abb5e8771100240e344411ddebacca33ece/temp Jan 14 19:06:51 and I ran: Jan 14 19:07:08 ./run.do_compile.24248 Jan 14 19:07:19 basically it was previously built with openembedded Jan 14 19:07:28 but this was necessary to rebuild it with the change Jan 14 19:08:04 then I scp qi-ubi-s3c2442-1.0.2-r7+gitr1+aa934abb5e8771100240e344411ddebacca33ece.udfu from /home/gnutoo/embedded/oe/oe-core/oetmps/shr/work/om_gta02-oe-linux-gnueabi/qi-ubi/qi-ubi-1.0.2-r7+gitr1+aa934abb5e8771100240e344411ddebacca33ece/git/image to my router Jan 14 19:08:40 GNUtoo: thanks man , wana back sms and contacts and then going to remove shr on nand and install android Jan 14 19:09:11 ok Jan 14 19:10:43 but is there anyway to export them ? Jan 14 19:12:51 freesmartphone.org: 03morphis 07aurora * r57d8e543ebcf 10/aurora-daemon/applications/phone/main.qml: Jan 14 19:12:52 freesmartphone.org: aurora-daemon: applications: phone: load correct page on startup Jan 14 19:12:52 freesmartphone.org: Signed-off-by: Simon Busch Jan 14 19:17:37 GNUtoo: is it possible to upgrade SHR without removing apps and sms's ? Jan 14 19:18:43 alabd, yes but I don't know how to do it Jan 14 19:18:48 (for sms) Jan 14 19:19:00 for apps I don't think it's possible but you could save a list of apps Jan 14 19:19:03 and reinstall tehm Jan 14 19:19:05 *them Jan 14 19:24:08 GNUtoo: how to make list ? Jan 14 19:25:00 ? Jan 14 19:25:06 note that there is a tool named Jan 14 19:25:10 opimd-cli Jan 14 19:31:11 GNUtoo: does shr have such software like android call manager Jan 14 19:31:43 I'm not aware of it Jan 14 19:31:58 but you could write a very simple python program that does that Jan 14 19:32:26 just know PHP man :P Jan 14 19:35:02 then write it with php-cli? Jan 14 19:37:54 GNUtoo: don't know , this is not my ob maybe because didn't worked with fr programming Jan 14 19:45:52 alabd, opkg list_installed > list_of_installed_packages.txt Jan 14 19:46:29 this way you have a list of installed packages in your image :) Jan 14 19:47:18 pespin: hi thanks man , Jan 14 19:47:25 GNUtoo, any idea about why bluez doesn't find bt adapter in gta02? I'll ask in #bluez Jan 14 19:47:52 I just run it in foreground to see logs and I can't see any kind of "big error" Jan 14 19:48:04 http://trac.shr-project.org/trac/attachment/ticket/1639/bluez.log Jan 14 19:48:28 maybe kernel problem? Jan 14 19:48:41 GNUtoo: i know you are lack of time, but what can cause this problem ? http://norman-schleicher.de/jenkins/job/pandaboard-angstrom/13/console Jan 14 19:48:45 pespin, possible let me look Jan 14 19:54:17 GNUtoo, maybe fso is not unblocking the bt device correctly? rfkill? I don't know much weel how all this work Jan 14 19:55:08 yes rfkill is quite possible Jan 14 19:55:23 cd /sys/class/rfkill/ Jan 14 19:55:28 and look Jan 14 19:55:46 pespin, I know what it is Jan 14 19:56:05 pespin, opkg remove connman Jan 14 19:56:10 it was removed from the images Jan 14 19:56:16 it should work again after that Jan 14 19:56:38 GNUtoo, # opkg list_installed | grep connman Jan 14 19:56:38 fsogsmd-connman - 1:0.5.0+gitr28+39d5f492e4eb6881e6ed25412310714f5dc4f625-r6.9 Jan 14 19:56:38 libeconnman0-7x1 - 2:1.1.0+svnr66238-r1.0 Jan 14 19:56:55 opkg remove fsogsmd-connman? Jan 14 19:57:10 nschle85, there are many errors, which one do you want to solve first? Jan 14 19:57:25 it seems connman itself is not installed Jan 14 19:57:25 pespin, remove what provides connmand Jan 14 19:57:39 pespin, do you have connmand in the path? Jan 14 19:57:54 GNUtoo, I don't have connmand running neither, and connman package is not installed Jan 14 19:58:06 which connmand shows nothing Jan 14 19:58:58 GNUtoo: the first is a blocvk of messages like this: NOTE: consider defining a PREFERRED_PROVIDER entry to match task-xserver NOTE: multiple providers are available for runtime kernel-module-autofs4 (linux-kexecboot, linux-palmpre, linux-dummy, linux) Jan 14 20:00:16 so is this causing the problems ? Jan 14 20:00:23 nschle85, no idea Jan 14 20:00:39 maybe you have some build setup issues? Jan 14 20:00:42 GNUtoo: ERROR: Function 'do_compile' failed (see /var/lib/jenkins/jobs/pandaboard-angstrom/workspace/setup-scripts/build/tmp-angstrom_2010_x-eglibc/work/pandaboard-angstrom-linux-gnueabi/linux-palmpre-2.6.24+gitr1+0b1014d99fc4992ece1605285e728c44d801a7a7-r0/temp/log.do_compile.30639 for further information) Jan 14 20:01:04 so linux-palmpre is not my kernel for panda Jan 14 20:01:05 ah ok Jan 14 20:01:11 indeed Jan 14 20:01:15 try to debug it Jan 14 20:01:24 or your setup is wrong Jan 14 20:01:32 or something is wrong in linux-palmpre Jan 14 20:01:43 because morphis is not that good in openembedded, he makes some errors Jan 14 20:01:51 and that can cause stuff like that Jan 14 20:02:10 Jama fixed yesterdday messges like that for hpveer Jan 14 20:02:15 he's very good in modem tough Jan 14 20:03:02 http://git.shr-project.org/git/?p=meta-smartphone.git;a=commit;h=5e47a01b2ab7a092ce5c85c352771e97d35f31ce Jan 14 20:05:51 GNUtoo: so ill do the same for palmpre :-) Jan 14 20:05:51 GNUtoo, # cat /sys/class/rfkill/rfkill0/state Jan 14 20:05:51 1 Jan 14 20:05:59 that means powered up right? Jan 14 20:06:19 right Jan 14 20:06:25 pespin, yes Jan 14 20:06:33 but it depend on the API used Jan 14 20:07:17 hmm ok Jan 14 20:07:20 and # cat /sys/class/rfkill/rfkill0/soft Jan 14 20:07:20 0 Jan 14 20:07:28 mean "soft blocked" = 0 ? Jan 14 20:07:38 so no soft blocked Jan 14 20:07:45 yes Jan 14 20:07:56 0 may be not blocked Jan 14 20:08:02 I don't remember Jan 14 20:08:07 I always do Jan 14 20:08:14 echo 1 > /sys/class/rfkill/rfkill0/state Jan 14 20:08:28 1 is like in Documentation of the kernel Jan 14 20:08:30 let me find it Jan 14 20:08:34 I always use rfkill utility app in my netbook for all this Jan 14 20:09:31 Documentation/ABI/stable/sysfs-class-rfkill Jan 14 20:09:41 it's in oe-core too Jan 14 20:09:46 but maybe not in shr-feeds Jan 14 20:09:58 ask it to JaMa Jan 14 20:10:05 so we would debug rfkill easier Jan 14 20:10:07 yeah it's not in feeds Jan 14 20:10:12 *make rfkill debugging easier Jan 14 20:10:24 ok I'll ask him when he's here :) Jan 14 20:10:46 for soft: Jan 14 20:10:53 1: active The transmitter is turned off by software. Jan 14 20:11:58 ok, so it's not turned off Jan 14 20:12:37 so you don't have connmand for sure? Jan 14 20:13:18 GNUtoo, it seems the kernel finds he device too http://trac.shr-project.org/trac/attachment/ticket/1639/dmesg-bluetooth.txt, but no hci0 is shown Jan 14 20:14:05 ps -ef | grep connman shows nothing Jan 14 20:14:33 pespin, what image? Jan 14 20:14:38 what's your setup? Jan 14 20:14:45 shr-core 017 I guess Jan 14 20:14:51 connman[tab] show nothing I guess Jan 14 20:14:53 latest Jan 14 20:14:57 ok Jan 14 20:15:00 I've shr-core too Jan 14 20:15:04 should be lastest Jan 14 20:15:04 no, nothing Jan 14 20:15:14 yep, feeds are from latest Jan 14 20:15:24 gta02 machine Jan 14 20:15:43 downloaded from official shr build dir Jan 14 20:15:51 and I'm not the only one having this bt problem Jan 14 20:15:59 may I ask in #bluez? Jan 14 20:20:13 if you want Jan 14 20:20:33 but I'm not sure the problem comes from bluez Jan 14 20:20:45 unless Jan 14 20:20:54 unless bluez is messing with rfkill too Jan 14 20:21:08 so you'd better ask Jan 14 20:21:14 it is connected to it somehow Jan 14 20:21:18 because it's quite possible according to the log Jan 14 20:21:25 when I echoed 1 to the sysnode I got this in bt output: Jan 14 20:21:26 bluetoothd[391]: src/rfkill.c:rfkill_event() RFKILL event idx 0 type 2 op 2 soft 0 hard 0 Jan 14 20:22:00 yes I'm talking exactly about that Jan 14 20:28:37 GNUtoo: so ill do the same for pa Jan 14 20:29:09 *palmpre I guess Jan 14 20:39:13 yes you guessed right :-) Jan 14 20:39:31 zub, hi, I just pushed first 2 patches from you Jan 14 20:40:20 zub, I'm not quite ok with the 3rd one though. Could you resend it with only the Toolbar part? Jan 14 20:41:18 and I'm not sure about the object_item changes as you changed all the formatting Jan 14 20:42:46 else I'll do it myself when I have time to finish the api updates Jan 14 20:46:02 mrmoku, ping Jan 14 20:47:14 zub, and btw, in the ToolbarItem class, the ToolbarItem returned by nex/prev_get() should be unowned Jan 14 21:13:56 GNUtoo: a somewhat late followup; but the situation for GTA04 is not at all similar to N900. while many people have an N900, few are interested in running any alternative systems on it -- people got it precisely because they like the default system it ships with! Jan 14 21:14:10 also, N900 is a dead platform, while GTA04 is just coming to live. Jan 14 21:14:52 antrik, by many people has it I mean many SHR devs Jan 14 21:22:26 GNUtoo: for N900 a nice usable OS exists, so why install shr ? for gta04 shr is the only option, thats the difference i think Jan 14 21:23:48 nschle85, I mean many SHR devs got an n900 for free to work on it and no one worked on it enough to make it usable(I didn't have this chance btw) Jan 14 21:24:18 GNUtoo: ups, i paid money for it Jan 14 21:24:40 GNUtoo: ahh ok now i understand Jan 14 21:24:45 me too I paid money Jan 14 21:24:57 I was visiting a company that now employ me Jan 14 21:25:05 so I missed the oportunity Jan 14 21:25:24 so at the end seeing that it was so promizing .... I bought it Jan 14 21:25:51 I mean the part that it was very promizing is also part of the equation Jan 14 21:25:59 for supporting it well Jan 14 21:26:01 GNUtoo: ill contiue at N900 dev until my wife can the russian IP tV with panda board :-) Jan 14 21:26:31 ok Jan 14 21:26:52 so the panda board is no platform i can use in future because its "occupied" by my wife :-) Jan 14 21:27:09 nschle85, if I work on n900, will you work on it too? Jan 14 21:27:32 GNUtoo: yes, and ill work on it even if you do not Jan 14 21:28:33 ok Jan 14 21:28:41 GNUtoo: but my problem is time, I can do something after 8pm for some hours, and Jan 14 21:28:42 what would you prefer me to do? Jan 14 21:28:43 the audio? Jan 14 21:29:02 because I could add it in the todo list Jan 14 21:29:09 *in my todo list Jan 14 21:29:24 GNUtoo: i cannot fix the audio, and fixing audio would make it usable as a phone :-) Jan 14 21:30:14 GNUtoo: but if i want to start, i see problems, which should be fixed, so i do different things .... Jan 14 21:31:04 so I'll work on audio if I get some time slot for that device Jan 14 21:31:06 but you are right, SMS is more important than GPS Jan 14 21:31:19 SMS is easier Jan 14 21:31:30 because the most important is the thing you care more about it Jan 14 21:31:37 but since you were somehow stuck with GPS Jan 14 21:31:40 I proposed SMS Jan 14 21:31:56 since you have no implementation headache Jan 14 21:32:05 GNUtoo: true Jan 14 21:32:06 you just have to understand a bit how isi work Jan 14 21:32:13 which is also required for GPS Jan 14 21:32:32 so it's easier Jan 14 21:33:02 http://redmine.replicant.us/projects/replicant/wiki/SamsungModems Jan 14 21:33:05 oops Jan 14 21:33:11 wrong channel Jan 14 21:33:28 zub, look at last patches I'll push now since I fixed some of the stuff you send me too (I did update all stuff using the C APi diff since last version I had installed here) Jan 14 21:56:17 has the GTA04 headset pinout thing ever been discussed on the ML?... Jan 14 21:59:14 not sure Jan 14 21:59:21 but it has been discussed here on IRC Jan 14 21:59:35 antrik, could you share the work you did at FSOSHRCON somewhere Jan 14 21:59:38 like on a wiki? Jan 14 22:00:09 because it contains great infos on how to record a call Jan 14 22:00:21 and how to play it back Jan 14 22:05:17 mrmoku: I for my part think that spawning an existing external program doing exactly the right thing (such as alsaloop) is a *very* good thing to do Jan 14 22:05:27 this is how code reuse is supposed to work on UNIX Jan 14 22:05:56 unfortunately, most of the kids calling themself UNIX developers nowadays have forgotten about this... Jan 14 22:07:01 antrik: you don't need to be confined to the limits of unix Jan 14 22:07:06 :) Jan 14 22:11:13 fsoaudiod: symbol lookup error: /usr/lib/cornucopia/modules/fsoaudio/gsmvoice_alsa_cmtspeechdata.so: undefined symbol: cmt_speech_init Jan 14 22:11:26 I guess I'll need to build Jan 14 23:17:57 nschle85: I don't see much point in adding a pandaboard+UMTS stick architecture to SHR. it's neither usable as a real phone, nor convenient as a development machine. PC+UMTS stick on the other hand might be a good idea... Jan 14 23:23:06 antrik: ++ on code re-use and spawning a dedicated full process Jan 14 23:23:31 pespin: thanks, I'll look into it tomorrow Jan 14 23:24:08 also I mentioned just a few days ago here that you want a new process anyway, to run it under RT scheduling Jan 14 23:24:59 which maybe could also be done with a thread, burt incredibly more clumsy Jan 14 23:25:11 but* Jan 14 23:27:11 the whole unix philosophy is based on filter cahins aka cmdline tools with sdin and stdout. And on 'everything's a file' - preferably a plaintext file Jan 14 23:27:31 chains* Jan 14 23:29:08 guess how fsoraw got inspired Jan 14 23:29:57 and why I 'invented' ACI Jan 14 23:45:26 lindi-: indeed... which is why I'm interested in the Hurd :-) but I also want to keep (and expand on) the *good* ideas from UNIX! :-) Jan 14 23:50:11 antrik: no need to go there :) Jan 14 23:50:22 antrik: you can build everything you want on top of linux Jan 14 23:53:18 building on top results in ugly, complex, fragile, inflexible solutions. to do it right, I'd have to make heavy changes to Linux itself -- which is a painful process Jan 14 23:54:23 antrik: and building everything completely from scratch isn't fragile? ;) Jan 14 23:54:35 I'd much rather build upon a system offering more customisation possibilities from the ground up Jan 14 23:54:49 antrik: well linux can be customized, you just edit it :) Jan 14 23:56:53 antrik: a lot of new exciting things are happening there all the time Jan 14 23:57:18 no, I also have to make sure it works, which is painful with a monolithic system; I have to forward port with every single version, which happens all the time with a monolithic system; I have to use customized kernels on every system I want to use the features on, which is painful with a monolithic system; or I can try to push my ideas upsteam, which is even more painful with a monolithic system... Jan 14 23:57:57 antrik: hard to say, I haven't ever modified hurd Jan 14 23:58:26 but I have some patches in linux and a many other in-house patches Jan 14 23:58:55 I'm not saying new things *can't* be done on Linux. architectures never differ in what can be done, strictly speaking. they diffen in how *hard* it is to do certain things -- and thus in what is actually realistic and what is not Jan 14 23:59:03 antrik: you can also test many new ideas completely in userland Jan 14 23:59:08 antrik: with just minimal helpers from linux Jan 14 23:59:14 antrik: like cgroups Jan 15 00:01:22 I can also run a whole operating system in a single process and customise it for every use case. such a lovely simple architecture! Jan 15 00:01:48 antrik: well no need to go to the extremes Jan 15 00:01:49 I do see some downsides though, believe it or not Jan 15 00:02:24 in my book, Linux is pretty extreme in being more monolithic even than most other mainsteam systems nowadays... Jan 15 00:03:13 antrik: hard to say, those are typically non-free? Jan 15 00:03:48 indeed. the free ones are not mainsteam ;-) Jan 15 00:04:33 generally speaking, I believe address space boundaries are a good idea. this applies both to application stacks (such as alsaloop running in a separate process...) and to low-level system functionality Jan 15 00:05:01 antrik: speaking of uncommon designs, have you looked at jnode? Jan 15 00:05:24 antrik: afaik they don't rely on MMU but rather bytecode validation Jan 15 00:05:41 I thus prefer a system that both uses address space boundaries in the default setup, and encourages using them in extensions and applications Jan 15 00:06:04 I don't like the VM approach Jan 15 00:06:21 it's a case of the inner platform syndrome Jan 15 00:06:38 antrik: surely you'll want a free software JIT over having the non-free microcode in the cpu do all the branch prediction magic? Jan 15 00:07:34 branch predication is not done in microcode last time I heard Jan 15 00:07:35 I see it more as runtime profiling and runtime recompilation Jan 15 00:08:18 antrik: not sure if microcode is the right term but inside the cpu anyway Jan 15 00:09:13 ah, wikipedia talks about predication and prediction Jan 15 00:10:07 http://en.wikipedia.org/wiki/Branch_predictor is more what I had in mind Jan 15 00:11:41 I can understand why some people like the VM approach; but I don't like locking everything into such a sealed system; it's too limiting for me taste Jan 15 00:11:50 my taste Jan 15 00:12:28 locking? Jan 15 00:14:35 I guess if you have a non-free VM you can't change much indeed Jan 15 00:15:48 that's not what I mean. either I have to run VM stuff and non-VM stuff beside each other (resulting in terrible integration), or I can *only* run stuff that works in that particular VM. both scenarios are not very attractive to me Jan 15 00:16:20 there is a reason why Android introduced an NDK... Jan 15 00:17:10 antrik: why should the integration be terrible? Jan 15 00:18:20 of course they can be but I see no reason why they should be Jan 15 00:20:56 gnucash uses C and scheme. it wasn't a VM system but the integration was already confusing :) Jan 15 00:21:07 building bridges between two distinct system designs is extremely complex. I'm not firm enough on theory to calculate the exact complexity, but I think it's rather obvious really :-) Jan 15 00:22:09 (it's also obvious from looking at all existing practical evidence I'm aware off...) Jan 15 00:22:15 of Jan 15 00:23:00 err... scheme *is* a VM system -- though quite different in nature from the Java one... Jan 15 00:23:39 but that's beside the point I'd say :-) Jan 15 00:30:14 antrik: depends on the scheme interpreter :) Jan 15 00:30:24 antrik: well, depends on what VM means, I mean't a JIT here mostly Jan 15 00:50:17 I don't mind JIT at all. I mind VMs creating their own environment -- which is obviously the case if the VM is to enforce its own isolation mechanism **** ENDING LOGGING AT Sun Jan 15 02:59:57 2012