**** BEGIN LOGGING AT Fri Jan 21 02:59:57 2011 Jan 21 04:08:34 moo Jan 21 06:54:47 pespin: you have to log in via blog.shr-project.org/cgi-bin/mt/mt.cgi - then it works :) Jan 21 06:56:42 mickey|zzZZzz: good question Jan 21 06:57:09 mickey|zzZZzz: less layers without loosing anything looks always correct though :) **** BEGIN LOGGING AT Fri Jan 21 07:52:17 2011 Jan 21 07:58:23 well, the more you diverge from the original code, the more maintenance burden... Jan 21 08:22:23 heyho Jan 21 08:29:13 freesmartphone.org: 03morphis 07msmcomm * ra85a6dfbf881 10/libmsmcomm/ (5 files in 4 dirs): libmsmcomm: implement phonebook extended file info command Jan 21 08:43:57 freesmartphone.org: 03morphis 07msmcomm * ra851f8a46611 10/libmsmcomm/ (4 files in 4 dirs): libmsmcomm: add structure definition for phonebook extended file info event Jan 21 08:56:13 some fun stuff (probably a bug): Jan 21 08:56:23 ~chanstat #openmoko Jan 21 08:56:44 err, that's probably also a bug :-P Jan 21 08:56:54 or not Jan 21 08:57:12 ~chaninfo #openmoko Jan 21 08:57:12 On #openmoko, there have been 1247 Joins, 15 Parts, 268 PublicMsgs and 1243 SignOffs. At the moment, 1 Opped and 121 Total. mickeyl has said the most with a total of 592 messages. Jan 21 08:57:24 ~chaninfo #meego Jan 21 08:57:25 On #meego, there have been 5024 Joins, 150 Parts, 3797 PublicMsgs and 4891 SignOffs. At the moment, 490 Total. mickeyl has said the most with a total of 592 messages. Jan 21 08:57:38 ~chaninfo #kde Jan 21 08:57:38 On #kde, there have been 6189 Joins, 206 Parts, 2711 PublicMsgs and 6054 SignOffs. At the moment, 1 Opped and 333 Total. mickeyl has said the most with a total of 592 messages. Jan 21 08:58:03 highly doubtable that mickey|zzZZzz is that active in all channels :P Jan 21 08:58:13 yep Jan 21 08:58:40 also highly doubtable mickey is the absolute rank #1 poster on >100 chan Jan 21 08:59:39 anybody interested and fluent in 20 lines Perl ibot source? Jan 21 09:01:39 I tracked it down to >> # Step 3:<< +20lines in UserExtra.pl, in https://sourceforge.net/project/showfiles.php?group_id=2241 Jan 21 09:02:10 but I completely fail to understand what's going on, and esp what's wrong Jan 21 09:05:12 I *think* it stored up the nick name from first chan it evaluated with ~chaninfo after reboot, which might happen to be #openmoko, and for #openmoko the above might actually be correct Jan 21 09:08:14 freesmartphone.org: 03morphis 07msmcomm * ree39daddd24c 10/libmsmcomm/ (4 files in 4 dirs): libmsmcomm: add specifications for all know sms subsystem messages Jan 21 09:15:02 meh, I'm tempted to kill apt/ibot Jan 21 09:15:38 just to test Jan 21 09:25:43 freesmartphone.org: 03morphis 07msmcomm * rb8a49d434be3 10/libmsmcomm/msmcomm/ (phonebookmessage.vala phonebookmessagegroup.vala): libmsmcomm: implement urc for phonebook extended file info message Jan 21 09:49:49 freesmartphone.org: 03morphis 07msmcomm * rbb2316d30efc 10/libmsmcomm/msmcomm/phonebookmessage.vala: libmsmcomm: phonebook: add base class for other urc messages Jan 21 09:59:49 freesmartphone.org: 03morphis 07msmcomm * rcdb439f644ea 10/msmcommd/src/phonebookservice.vala: msmcommd: phonebook service: adjust for latest changes to libmsmcomm Jan 21 10:00:28 freesmartphone.org: 03morphis 07msmcomm * r773a83153809 10/libmsmcomm/msmcomm/ (phonebookmessage.vala phonebookmessagegroup.vala): libmsmcomm: phonebook: implement handling of phonebook ready urc again Jan 21 10:05:45 morphis: seems you have good progress with the new style Jan 21 10:06:34 mickey|office: jepp Jan 21 10:07:03 mickey|office: and as playya is starting to work on libmsmcomm progress will raise :) Jan 21 10:08:24 and structure is now a lot cleaner than before Jan 21 10:10:21 mickey|office: I am using now my new Pre Plus as daily phone ... sadly I remembered too late that the frequencies here in europe for umts are different than in the US ... Jan 21 10:10:48 so only gsm gprs/edge Jan 21 10:11:07 oh Jan 21 10:12:02 so I will switch back the pre as daily phone Jan 21 10:12:32 right Jan 21 10:13:17 do you have a rough estimate how long it will take you to get everything but the more esoteric things (CB, USSD, Call Forwarding, etc.) done? Jan 21 10:16:22 I want to finsih the basic things (SMS, State, Misc, Phonebook, Call) in msmcomm as fast as possible Jan 21 10:16:55 but no in the next two weeks as I am moving and have some exams to pass :) Jan 21 10:17:58 playya will mostly work on SMS and I will take care about the other parts Jan 21 10:17:58 which are mostly know and just have to be test and finished in their implementation in libmsmcomm Jan 21 10:18:56 ok Jan 21 10:19:10 fsogsmd integration should be simple Jan 21 10:19:49 I reworked most things in fsogsmd already to suite for the new dbus api of the msmcomm daemon Jan 21 10:20:50 if you want to know a date where I want to finish this basics bit, it should be late february Jan 21 10:25:38 mickey|office: if you have some time, you can help with fsogsmd integration Jan 21 10:36:30 freesmartphone.org: 03morphis 07msmcomm * r36239ad3a238 10/ (2 files in 2 dirs): msmcommd: phonebook service: implement extended file info messages Jan 21 10:41:47 freesmartphone.org: 03morphis 07cornucopia * r4bd5c658de51 10/fsogsmd/src/plugins/modem_qualcomm_palm/Makefile.am: fsogsmd: modem_qualcomm_palm: fix build depenency in Makefile Jan 21 10:55:54 morphis: ok. do you know whether the fso installer already features the dual boot install? Jan 21 10:56:55 mickey|office: puuh Jan 21 10:56:57 don't know Jan 21 10:57:05 but installing it is simple Jan 21 11:07:03 morning! Jan 21 11:07:49 mickey|office, i just saw in the logs, that you asked about the pre dual-boot solution Jan 21 11:08:08 there is a patch for fso-installer to include it, but it isn't applied yet Jan 21 11:12:05 Slyon: what do you think about writing bootr in C? then one could easiely add touchsrceen support Jan 21 11:13:39 Heinervdm, there are two problems: Jan 21 11:14:01 1. i don't know C very well Jan 21 11:14:08 Slyon: ok, thanks, i'll try to patch it then Jan 21 11:14:27 2. we have a very limited enviroment at /boot without much libs Jan 21 11:14:33 but theoretically it should be possible Jan 21 11:14:51 mickey|office, shall i send you the patch? Jan 21 11:16:53 GarthPS send it to me and morphis Jan 21 11:16:53 and afaik garth tested it on his Pre2 Jan 21 11:18:05 mickey|office, patch is here: http://pastebin.com/GZAxe5fT Jan 21 11:26:29 Slyon: excellent, thanks Jan 21 11:27:03 mickey|office, yw Jan 21 11:31:33 lindi-: so, suspend resume test results: as i told, 2 WSODs in beginning, (after 2 and 19 resume), then 1350 suspend/resumes ok, then it went into state similar to yours -> nothing on display, no reaction to power button. given that i had ~17 seconds interval this is like 9 days with 10-minute cycle. Notable moment is that it boots automatically after remove/insert of battery (no need to press 'power'), this probably mean that it 'wakes' from rtc event Jan 21 11:31:33 . so, your bug probably can be reproduced without waiting for several weeks :) Jan 21 11:36:10 gena2x: "great" Jan 21 12:05:39 hello shr people. i am running latest unstable but my gsm does not register. is this a known issue? Jan 21 12:12:51 morphis, i forwarded the patch, which adds the bootr installation to fso-installer to mickey. but you should probably push it to git, as the instroductions at shr wiki depend on it Jan 21 12:17:05 Slyon, you can paste it and I push it if you want Jan 21 12:18:36 playya, http://pastebin.com/GZAxe5fT here it is. But one should probably test it first on a Pre. afaik garthps tested it on a pre2 only Jan 21 12:18:48 ok Jan 21 12:18:56 i already use bootr on my pre Jan 21 12:19:00 good work Jan 21 12:19:04 thanks Jan 21 12:26:29 Hello Jan 21 12:26:48 Is there any new articles about developing in moderm QtCreator? And under contemporaty Ubuntu version? Jan 21 12:27:34 * shade-khv leaves for a while Jan 21 12:27:55 freesmartphone.org: 03Frederik.Sdun 07cornucopia * r7704976d49c0 10/libfsobasics/fsobasics/utilities.vala: libfsobasics: Add some more enum helper Jan 21 12:29:31 freesmartphone.org: 03Frederik.Sdun 07msmcomm * rcfa98d197330 10/libmsmcomm/specs/structures/sms.xml: libmsmcomm:specs: rename _length fields to _len Jan 21 12:29:33 freesmartphone.org: 03Frederik.Sdun 07msmcomm * reb6aaf09854a 10/libmsmcomm/ (6 files in 4 dirs): Jan 21 12:29:33 freesmartphone.org: libmsmcomm:tools: Allow to specify nested structs Jan 21 12:29:33 freesmartphone.org: Nested structs need a _field postfix Jan 21 12:29:33 freesmartphone.org: Regen files Jan 21 12:29:33 freesmartphone.org: 03Frederik.Sdun 07msmcomm * rdc8ea2ae08fa 10/libmsmcomm/msmcomm/callmessage.vala: libmsmcomm: add missing field to CallReturnResponseMessage Jan 21 12:37:04 Slyon, it doesn't apply Jan 21 12:37:11 need to check later Jan 21 12:37:25 hmm, ok, then we should probably wait for GarthPS Jan 21 12:40:21 Slyon: this is untested, but i think it should work as a replace for bootr.sh Jan 21 12:40:23 http://heinervdm.dyndns.info/shr-dist/Pre/bootr/ Jan 21 12:40:50 hi all Jan 21 12:41:21 here it seams to be spam http://docs.openmoko.org/trac/ticket/2476 Jan 21 12:42:06 Heinervdm, wow cool. i will look at it later, as i'm sitting at a lecture atm Jan 21 12:42:11 thanks Jan 21 12:44:05 Heinervdm, i can't access cy8mrln.h, i get a 403 error Jan 21 12:46:56 Slyon, i think cy8mrln.h isn't necessary for bootr Jan 21 12:47:00 Slyon: now Jan 21 12:47:17 playya: i included it for ts Jan 21 12:47:25 yes Jan 21 12:47:49 btw. i "forked" tslib Jan 21 12:48:00 easier to send pull requests Jan 21 12:48:17 Heinervdm, ok dowloaded it and will look at it later today Jan 21 12:49:03 Heinervdm, ah... how do you compile this? do you have a special toolchain? Jan 21 12:49:39 i'm using widk Jan 21 12:50:34 okay, will try to compile it myself then and write some docs Jan 21 12:51:23 i think about replacing the logos on my pre Jan 21 12:51:25 one should try if it works before writing docs :) Jan 21 12:51:28 any nice ideas? Jan 21 12:51:37 * Heinervdm afk Jan 21 12:51:56 yes i will try it first of course Jan 21 13:25:20 freesmartphone.org: 03Frederik.Sdun 07msmcomm * rcddcf88cbc9f 10/libmsmcomm/ (4 files in 3 dirs): libmsmcomm: remove generated test messages Jan 21 13:58:45 freesmartphone.org: 03Frederik.Sdun 07cornucopia * r54f32d8be094 10/libfsobasics/ (fsobasics/utilities.vala tests/testutilities.vala): Jan 21 13:58:45 freesmartphone.org: libfsobasics: make enumToString und enumToNick Jan 21 13:58:45 freesmartphone.org: typesafe delegates Jan 21 13:58:47 freesmartphone.org: 03Frederik.Sdun 07cornucopia * ra5147551a2af 10/fsogsmd/src/ (2 files in 2 dirs): fsogsmd: Catch up with enumToString changes Jan 21 14:44:56 playya: hmm, is the builtin vala support for enumToString not enough? Jan 21 14:52:07 mickey|office, i don't know how vala reacts if there's an unknown value Jan 21 14:52:28 and does vala support from/to nick? Jan 21 14:52:44 unknown value is now being detected Jan 21 14:52:48 what is nick? Jan 21 14:53:08 if you have MyEnum{FOO="foo"} Jan 21 14:53:21 only the "member" Jan 21 14:54:05 should be the same as string marchalling in DBus Jan 21 14:59:57 the nick is like tr '[:upper:]_' '[:lower:]-' on the member Jan 21 15:00:16 without the name prefix Jan 21 15:14:20 i see. i don't think there's a way to get that atm. Jan 21 15:14:30 (not through to_string(), that is) Jan 21 15:17:39 i left it untouched in the first commit, because of compability and now it's there for completeness Jan 21 15:18:09 ok Jan 21 15:20:40 i really like the gobject stuff. they implemented a lot of RTTI stuff Jan 21 15:33:57 true Jan 21 15:34:00 it has its merits Jan 21 15:34:10 now if it wasn't so ridicolously slow... Jan 21 15:34:38 coming from python it's still a speed-boost, but gobject has to be one of the slowest object frameworks ever created ;) Jan 21 15:39:10 mickey|office, hi, mdbus2 crashed on listing network on nojia900 Jan 21 15:39:27 it didn't crash on other phones accroding to other people Jan 21 15:43:59 freesmartphone.org: 03Frederik.Sdun 07msmcomm * r4e5ab9619727 10/libmsmcomm/ (7 files in 5 dirs): Jan 21 15:43:59 freesmartphone.org: libmsmcomm: Allow definition of group id and message Jan 21 15:43:59 freesmartphone.org: id in the XML files Jan 21 15:44:00 freesmartphone.org: 03Frederik.Sdun 07msmcomm * re51915d1393e 10/libmsmcomm/msmcomm/phonebookmessage.vala: libmsmcomm: use libfsobasics to convert data to strings Jan 21 15:44:01 freesmartphone.org: 03Frederik.Sdun 07msmcomm * r17a2256eb49f 10/libmsmcomm/msmcomm/phonebookmessage.vala: libmsmcomm: check sizes before using memcpy Jan 21 15:44:03 freesmartphone.org: 03Frederik.Sdun 07msmcomm * rdd4bb0c78bf8 10/libmsmcomm/ (3 files in 3 dirs): libmsmcomm: Use corect group id for state_status event Jan 21 17:13:35 mickey|office, is it possible to get a \0 in a sms pdu? Jan 21 17:14:02 morphis and I just realized that fsogsmd uses string in this area Jan 21 18:05:51 hello Jan 21 18:07:10 oe git is down :( Jan 21 18:07:18 mrmoku, ping Jan 21 18:07:46 mrmoku, nvm i'll wait for oe to go back online Jan 21 18:43:08 freesmartphone.org: 03morphis 07msmcomm * rbc64c71109c9 10/ (9 files in 6 dirs): msmcomm: sim: add read message to read sim files Jan 21 18:43:11 freesmartphone.org: 03morphis 07msmcomm * redfd0712d5ae 10/libmsmcomm/msmcomm/simmessage.vala: libmsmcomm: sim: correct parsing of imsi from sim field data Jan 21 18:43:15 freesmartphone.org: 03morphis 07msmcomm * r8698af170a40 10/ (2 files in 2 dirs): msmcomm: phonebook service: ship book type with PhonebookInfo structure Jan 21 18:43:37 daniele_athome: pong anyway :) Jan 21 18:43:42 pespin|away: ping Jan 21 18:46:36 freesmartphone.org: 03morphis 07msmcomm * rd2ea027522e5 10/libmsmcomm/msmcomm/simmessage.vala: libmsmcomm: sim: initialize field_data with empty string Jan 21 18:50:47 freesmartphone.org: 03morphis 07msmcomm * raad870846488 10/libmsmcomm/msmcomm/messageutil.vala: libmsmcomm: avoid unnecessary warning due to unused method Jan 21 19:19:03 freesmartphone.org: 03morphis 07msmcomm * red1dd3bd77d6 10/libmsmcomm/ (6 files in 5 dirs): libmsmcomm: add network subsystem message specification Jan 21 20:19:06 playya_: hmm, i don't know offhand. let me reread the specs and i'll tell you on weekend Jan 21 20:19:53 GNUtoo|laptop: (mdbus2) i'm afraid there are some situations where mdbus2 still chokes on empty structures :/ Jan 21 20:19:57 GNUtoo|laptop: try cli-framework Jan 21 20:20:08 >>> gsmnetwork.ListProviders( timeout=100000 ) Jan 21 20:20:42 (cli-framework is actually nicer to work with than mdbus2, since it's a full python commandline) Jan 21 20:21:16 mrmoku: what's your take on my proposal? Jan 21 20:21:20 (libisi) Jan 21 20:22:57 mickeyl: sounds like a sane thing to do - dropping an uneeded layer that is Jan 21 20:25:04 mrmoku: ok, we should experiment with that approach. i wonder though whether it's more important or less important compared to ophoned Jan 21 20:25:06 can't do both... Jan 21 20:25:11 (i.e. not in parallel) Jan 21 20:25:22 mickeyl: you would have to write a vapi for gisi then Jan 21 20:25:25 mickeyl: yeah, ic Jan 21 20:25:39 if you ask me ophoned is what I would prefer you to work on :) Jan 21 20:25:55 maybe GNUtoo|laptop and me can have a look at vala libisi Jan 21 20:25:57 ok. voip would be a big feature for us Jan 21 20:26:05 oh yes :-) Jan 21 20:26:06 right, makes sense Jan 21 20:26:18 and we should ask sre too what he thinks Jan 21 20:26:48 is he actually still working on it? if so at least not pushing commits to the repo :P Jan 21 20:27:04 for sure, i don't want to obsolete his work, it's just that i'm worried about a) the maintenance work while gisi is still progressing and b) the amount of C code necessary to do simple things Jan 21 20:27:12 from what i hear he's busy with studies Jan 21 20:27:23 ok Jan 21 20:27:32 maybe we could just start it in a branch Jan 21 20:27:36 to see where it leads Jan 21 20:27:41 yes Jan 21 20:27:56 ok, i'll keep working on that API proposal then Jan 21 20:28:06 yes, please :) Jan 21 20:29:00 my home telephony and telephony in the company are both sip based... could make very much use of it :D Jan 21 20:29:11 right Jan 21 20:29:27 and we could even use the n900 as a phone pretty soon Jan 21 20:29:36 that too, yes :) Jan 21 20:29:45 [at least via wifi] Jan 21 20:30:16 btw. I'm wondering if the lowlevel plugin is doing too much Jan 21 20:30:20 hmm Jan 21 20:30:27 libisi works better without and manually enabling the modem Jan 21 20:30:37 well, are we sure that ophoned is the right thing to do? Jan 21 20:30:48 yes ;) Jan 21 20:30:58 yes Jan 21 20:31:04 but what about using telepathy? Jan 21 20:31:56 dos1: why would you think otherwise? Jan 21 20:31:57 dos1: that has nothing to do with ophoned or not Jan 21 20:31:57 telepathy could be one backend for ophoned Jan 21 20:31:57 but we want that stuff integrated into one thing Jan 21 20:31:57 gsm + voip Jan 21 20:32:31 mrmoku: i would rather think about ogsmd being backend for telepathy Jan 21 20:32:40 and then... everytime I tried telepathy/empathy on my desktop It did not convince me Jan 21 20:32:41 maybe i understand function of ophoned incorrectly Jan 21 20:32:42 freesmartphone.org: 03morphis 07msmcomm * r06d852edeac9 10/msmcommd/src/ (4 files): msmcommd: add optional timeout for commands Jan 21 20:32:43 freesmartphone.org: 03morphis 07msmcomm * r6c679cfa57a2 10/msmcomm-specs/src/management.vala: msmcomm-specs: add missing TIMEOUT error type Jan 21 20:33:11 and maybe i don't understand telepathy correctly - i never really looked at it Jan 21 20:33:16 dos1: so you want to drop our phone stack? Jan 21 20:33:21 and use telepathy instead? Jan 21 20:33:31 mrmoku: not really Jan 21 20:34:03 mrmoku: i just want to make sure if everything is correctly designed Jan 21 20:34:08 :) Jan 21 20:34:28 as i don't want situation similar to FSO and oFono Jan 21 20:34:40 ? Jan 21 20:34:48 "we don't understand and don't like FSO, and we didn't write it, so we'll write our own stack!" ;) Jan 21 20:34:58 ahh Jan 21 20:35:39 dos1: but I don't think that was the reason for ofono :) Jan 21 20:36:16 morphis: well, yes, but at least that's how looked their argumentation later ;D Jan 21 20:36:25 :) Jan 21 20:37:02 but i understand, ofono is funded by Nokia and Intel, and they have another goals than community Jan 21 20:37:34 and FSO is community based ;) Jan 21 20:38:29 hi Jan 21 20:38:39 what what the reason for ofono then? Jan 21 20:38:53 * onen|openBmap jumps right in the conversation Jan 21 20:38:58 control Jan 21 20:38:59 onen|openBmap: having contorl Jan 21 20:39:06 err, yes control Jan 21 20:39:30 but, as far as I understand it, telepathy was meant to provide such high-level layer for communication, no matter if voice, text or video based Jan 21 20:39:38 mmmh in what sense? Why not git cloning fso, and further develop it. they would still have control? Jan 21 20:40:06 well, there's a couple of things that came together Jan 21 20:40:18 so maybe it's just worth checking? Jan 21 20:40:28 dos1: looking at the architecture right now Jan 21 20:40:39 (if someone already looked at it carefully, then ok, i'm just making sure :)) Jan 21 20:40:47 would mean a lot of work for us Jan 21 20:40:57 a) lead developer having a major NIH syndrome, b) FSO gsm only reference implementation written in Python, c) FSOs APIs being a bit lower level than they wanted. Jan 21 20:41:13 add that to the control thing Jan 21 20:41:15 mickeyl: I see Jan 21 20:41:20 and you have enough reason to start something new Jan 21 20:41:36 mrmoku: to be honest, i really like how messaging app works on Maemo Jan 21 20:41:43 mickeyl: I have an additional though: patents Jan 21 20:41:53 mrmoku: it's nicely integrated between IM, SMS, voice, VoIP and video calls Jan 21 20:42:00 perfectly possible Jan 21 20:42:00 dos1: that we would have to implement then... messaging in Maemo is not open, no? Jan 21 20:42:09 mrmoku: and AFAIK it uses telepathy Jan 21 20:42:34 mrmoku: yup, it's closed. but that doesn't matter Jan 21 20:42:41 mickeyl: doing everything from the start, you are pretty sure of copyright ownership, of no "whatever license" contamination, and easier patent management Jan 21 20:42:54 onen|openBmap: yes, that's right Jan 21 20:43:00 mrmoku: i just want to be sure, that with our architecture it would be quite easy to write such well-integrated app Jan 21 20:43:30 mickeyl: by the way did you see this: http://fosspatents.blogspot.com/2011/01/new-evidence-supports-oracles-case.html Jan 21 20:43:49 onen|openBmap: not yet, no. "fun" Jan 21 20:44:15 mickeyl: it seems someone found it easier to jsut copy/paste some files from java Jan 21 20:44:44 wah, that's tough Jan 21 20:45:09 mickeyl: indeed. this gives an impression of amateurism Jan 21 20:45:20 i can hardly believe that this went through their review Jan 21 20:45:31 freesmartphone.org: 03morphis 07msmcomm * r0d0e8cb86b69 10/msmcommd/src/modemchannel.vala: msmcommd: don't process any command when modem is not active Jan 21 20:45:43 mickeyl: if a company of the size of google did not detect such a thing (if it is indeed true), it means they just simply did not try to detect anything Jan 21 20:47:57 ok got to go Jan 21 20:48:01 bye Jan 21 20:48:16 cu Jan 21 20:48:27 dos1: btw., did you do that ambientlight plugin yet? Jan 21 20:48:31 or shall i? Jan 21 20:48:35 [for n900] Jan 21 20:50:48 mickeyl: dos1 is sunken into telepathy docs :P Jan 21 20:50:51 mickeyl: no, i didn't Jan 21 20:50:52 hehe, ok Jan 21 20:51:02 mrmoku: haha, very funny :P Jan 21 20:51:27 :) Jan 21 20:52:20 /sys/bus/i2c/drivers/tsl2563/2-0029/lux Jan 21 20:52:29 mickeyl: when i started, i didn't know what unit i should use Jan 21 20:52:36 do we need to poll or is it sending out frequently? Jan 21 20:52:37 well Jan 21 20:52:39 cause in palm pre it was something else Jan 21 20:53:02 i know that there are other opinons, but i don't see a better approach than normalized percentage Jan 21 20:53:08 think about the application domain Jan 21 20:53:20 non-specialized apps don't care about lux Jan 21 20:53:25 they care about environment Jan 21 20:53:26 is it dark Jan 21 20:53:27 mid Jan 21 20:53:29 or bright Jan 21 20:53:32 that's all they care about Jan 21 20:53:39 then they adjust something Jan 21 20:53:58 so if we consider multiple devices, i don't think anything else than a percentage is sane Jan 21 20:54:28 specialized apps can still use the sysfs node Jan 21 20:59:27 freesmartphone.org: 03morphis 07msmcomm * r78aefbba7794 10/msmcommd/src/modemcontrol.vala: msmcommd: it shouldn't possible to send commands while the modem is initializing Jan 21 21:00:51 ok, so it looks like we need to poll Jan 21 21:01:48 mickeyl, ok thanks Jan 21 21:02:40 mrmoku, vala libisi? Jan 21 21:07:25 GNUtoo|laptop: might be an idea Jan 21 21:10:54 we want nokia900 bringup the fastest possible Jan 21 21:11:02 will that makes things faster Jan 21 21:11:17 or will that retard the n900 modem stuff? Jan 21 21:11:49 I don't think it will have great impact on speed whatsoever Jan 21 21:11:55 not netagtive... not positive Jan 21 21:12:05 ok Jan 21 21:12:14 if libisi is in C Jan 21 21:12:18 we can just copy ofono code Jan 21 21:12:20 right? Jan 21 21:12:29 that is not that simple Jan 21 21:12:45 what will be the vala equivalent of that? Jan 21 21:12:58 we will still use the gisi c source from ofono Jan 21 21:13:16 but instead of the c stuff in isi we would have vala Jan 21 21:13:39 instead of wrapping the c code via vapi to Modem and such Jan 21 21:15:38 dos1: the telepathy API looks rather complicated :P Jan 21 21:15:42 i have looked at the isi ofono code. we can copy ofono code, but need to do lots of monkeywork to get rid of the ofono data structures and use our data structures. then we have the additional vapi on top and have potential bugs in two layers. by not copying the ofono code but rather writing it in fsogmsd, we skip one layer Jan 21 21:16:14 we can still keep it a seperate library, if you want Jan 21 21:17:05 mickeyl: what about the gisi part? Jan 21 21:18:15 gisi part would be verbosely taken from ofono, no changes except getting it to build standalone Jan 21 21:18:58 ok, so that would be libgisi then... instead of libisi? Jan 21 21:19:07 and be used by the fsogsmd n900 modem plugin directly? Jan 21 21:20:08 that would be my first approach. Jan 21 21:20:21 ok Jan 21 21:20:36 GNUtoo|laptop: what do you think? Jan 21 21:20:37 mickeyl, mrmoku responded on speed, what do you think about that? Jan 21 21:20:42 we would probably have a commandqueue or libgisi abstraction that resembles the libisi abstraction Jan 21 21:20:53 * morphis thought which device will be the first ready for the community ... palm pre or n900 :D Jan 21 21:20:59 I'm not sure what I think yet Jan 21 21:21:04 morphis: gta04 ;) Jan 21 21:21:08 GNUtoo|laptop: well... if we get it working we'll be much faster Jan 21 21:21:10 mrmoku: :) Jan 21 21:21:17 vala to C is an enourmous productivity boost Jan 21 21:21:23 <[Rui]> *sigh* [rms@roque tmp]$ ls | grep -v cache | xargs rm -rf Jan 21 21:21:25 due to the lines of code to write alone Jan 21 21:21:32 mickeyl, then let's do it but I can't help with lower levels yet Jan 21 21:21:47 are we talking about programming speed or execution speed? :P Jan 21 21:22:02 what morphis descirbed speed Jan 21 21:22:07 * mickeyl talks about development speed Jan 21 21:22:10 ahh Jan 21 21:22:16 to be ready for the community Jan 21 21:22:20 to make phonecalls Jan 21 21:22:21 execution speed will not matter significantly Jan 21 21:22:21 * mrmoku was the only one thinking in execution speed then Jan 21 21:22:27 yup Jan 21 21:22:30 ok Jan 21 21:22:34 GNUtoo|laptop: "ready" = use it as your daily phone Jan 21 21:22:45 yes like I did with htc dream Jan 21 21:23:21 again a toolchain upgrade Jan 21 21:23:34 it seem that I won't be able to compile something this night Jan 21 21:25:23 I'm for what bring us a (somewhat) usable phone first Jan 21 21:25:31 if vala is faster let's go vala Jan 21 21:25:32 freesmartphone.org: 03mickey 07cornucopia * r08dfaa79e569 10/fsodeviced/ (4 files in 3 dirs): Jan 21 21:25:32 freesmartphone.org: fsodeviced: ambientlight_n900: new plugin Jan 21 21:25:32 freesmartphone.org: NOTE: Not tested on real hardware yet Jan 21 21:25:37 if C is faster let's go C Jan 21 21:26:26 I even considered using ofono in fsogsmd Jan 21 21:26:32 trough dbus Jan 21 21:26:38 heh Jan 21 21:26:52 what are the downsides of that? Jan 21 21:27:19 switching totally to ofono is not feasible and not desirable, but time is against us Jan 21 21:27:28 we bring up devices when they are obsoletes Jan 21 21:27:48 so I'm really for what's faster whatever it is Jan 21 21:28:03 well Jan 21 21:28:13 including other phones, dirty stuff etc.... Jan 21 21:28:13 I tend to disagree Jan 21 21:28:14 people seem to have nightmares getting it up and running Jan 21 21:28:22 i'm lurking on #ofono Jan 21 21:28:26 :P Jan 21 21:28:32 so we have bugs on yet another level Jan 21 21:28:35 i don't think that makes us faster Jan 21 21:28:38 ok Jan 21 21:28:40 GNUtoo|laptop: I want something good too Jan 21 21:28:44 not just fast Jan 21 21:29:03 basically here's the problem: Jan 21 21:29:08 a device come to the market Jan 21 21:29:10 and fortunatelly n900 is still a bit away from being obsoleted Jan 21 21:29:12 Boys, I know the best solution for your problems ! Jan 21 21:29:12 it's new etc.... Jan 21 21:29:24 it's very simple Jan 21 21:29:27 :D Jan 21 21:29:29 then nobody know a thing about the new device Jan 21 21:29:35 it takes time to know it Jan 21 21:29:55 morphis: hope its not too simple... would be boring then ;) Jan 21 21:30:01 to replace the problematic part etc.... Jan 21 21:30:12 and then when we decide it's wroth Jan 21 21:30:16 it's already late Jan 21 21:30:18 mrmoku: it's simple but not cheap :) Jan 21 21:30:21 and then we have to do the work Jan 21 21:30:28 morphis: ahh... ok Jan 21 21:30:30 what is it? Jan 21 21:30:34 * mrmoku knew there would be a drawback :P Jan 21 21:30:43 * GNUtoo|laptop is interested Jan 21 21:30:49 GNUtoo|laptop: thing about it :) Jan 21 21:30:52 * GNUtoo|laptop tought something not feasible Jan 21 21:30:55 I am waiting for you guys Jan 21 21:31:15 morphis, the palm-pre-2? Jan 21 21:31:19 :) Jan 21 21:31:30 pre-* Jan 21 21:31:36 really, it's not that expensive Jan 21 21:31:44 if you take an old palm-pre instead Jan 21 21:32:18 right Jan 21 21:32:21 basically the best phone would be with AT and NMEA but recent Jan 21 21:32:33 and GNU/Linux with free kernel drivers Jan 21 21:32:50 there is maybe one phone in future ... Jan 21 21:32:55 gta04 Jan 21 21:33:32 nexus one has NMEA,AT,3g,etc....but a bad alsa driver(yes it has an alsa driver but there are buffer underrun) Jan 21 21:33:52 basically when I'm too tired I do useless stuff Jan 21 21:34:01 and I found an alsa driver for it Jan 21 21:34:04 and adapted it Jan 21 21:34:07 but it's not usable Jan 21 21:34:12 it was woth trying tough Jan 21 21:35:23 so how do we start for the vala libisi replacement? Jan 21 21:35:49 first thing would probably a standalone libgisi with vapi Jan 21 21:35:54