**** BEGIN LOGGING AT Tue Jul 28 02:59:57 2009 Jul 28 05:32:27 hey mrmoku, whats new? Jul 28 05:32:54 F4t, he, good question :) Jul 28 05:33:57 ive seen u added svnr102 to the repo, any complaints so far? :) Jul 28 05:34:10 F4t, no and works good for me Jul 28 05:34:22 cool Jul 28 05:34:43 im reflashing the device now :P Jul 28 05:35:34 Ainulindale, did you look into the gta01 image? Jul 28 05:36:07 mwester_, I'm not thinking about complex datatypes or OO, don't worry. was just thinking about glib for easy parsing of config files, and maybe for easy linked lists use :) but is also seemed a bad idea when I thought about it... so don't worry ;) Jul 28 05:43:13 TAsn: what's wrong with glib? Jul 28 05:44:54 TAsn: not lightweight enough for SlugOS? Jul 28 06:43:16 Guys, i still can't understand why i should see a blue LED when my battery is actually not full and not charging? Jul 28 06:45:22 PaulFertser, 'cause its pretty? :P Jul 28 06:45:44 F4t: are you sure that's the right behaviour? IMHO it's confusing :-/ Jul 28 06:46:03 PaulFertser, when did u see it anyway? Jul 28 06:46:33 F4t: because mickey_away decided so :S Jul 28 06:46:51 at what stage? i mean, what tiggers it? Jul 28 06:46:59 what does it indicate? Jul 28 06:47:52 F4t: after charge is over charger turns off. And the battery is discharged by GSM modem. But charger _is_ off and the battery _is_ discharging. I can't see why it should be the same as "fully charged". Jul 28 06:53:46 * PaulFertser really should remove that errorneous forcing charging restart from the kernel :-/ Jul 28 07:11:43 PaulFertser, have to agree with you, blue led should only be when charger is plugged... Jul 28 07:12:20 F4t: even with the charger plugged you can have you battery discharging with as much as ~100mA! Jul 28 07:13:06 PaulFertser, i know, but i think if the charger is plugged, and the battery is discharging after it was full, and its not time yet to charge, it should still be blue... Jul 28 07:13:31 F4t: imagine i came home, plugged to the usb, it charged the battery fully to 100%, disabled charger (to not damaged the battery) and then i get a call and talk for an hour. All that time the battery is discharging by the gsm modem. Is discharging. And it's not 100% capacity anymore. Jul 28 07:14:45 hmmm, what i ment is, if it can automatically switch to charging, then while its not, it should be blue... Jul 28 07:14:49 F4t: you can have as little as 80% left there after. Jul 28 07:14:51 if you disable it, then not Jul 28 07:15:22 F4t: isn't it confusing to see the blue led while your battery is only at 80%? Jul 28 07:16:16 PaulFertser, a bit, but its better then starting to guess if you dont see a light because the charger is bad, or the contacts, or stuff like that... Jul 28 07:16:36 there should be some indication that the connection is good, and it will charge when it decides to Jul 28 07:17:12 F4t: i disagree :( Jul 28 07:17:37 mh... there's orange and blue ATM - what about switching both LEDs on? Jul 28 07:17:49 blinking blue? Jul 28 07:18:11 F4t: that'd be much better, i'd agree. Jul 28 07:18:52 well, suggest it to the one that can change it then :) Jul 28 07:18:59 F4t: it's mickey_away Jul 28 07:19:14 F4t: but i need support from other users :D Jul 28 07:19:30 F4t: as he has the same reasoning you described. Jul 28 07:21:52 Wonka: it looks very cool! Beautiful magenta. Jul 28 08:53:36 hello Jul 28 08:55:10 I've a question about opimd Jul 28 08:55:12 :P Jul 28 08:55:40 I supposed messages are stored in /etc/freesmartphone/opim/sqlite-messages.db Jul 28 08:56:01 but it seems not updated with latest received messages... Jul 28 08:57:27 a21_work: As default backend SIM is used Jul 28 08:57:55 Heinervdm: I know... but also sqlite is enabled Jul 28 08:58:12 a21_work: but the message is saved only once Jul 28 08:58:28 and if I delete a sms from sim, I still get it with: opimd-cli m query Jul 28 08:59:05 a21_work: deleting SMS is there since yesterday in the git ;) Jul 28 08:59:17 and not yet build Jul 28 08:59:27 no no, sorry about my bad english Jul 28 08:59:36 I'm saying: Jul 28 09:00:01 my english isn't better :D Jul 28 09:00:03 if I delete a sms from sim, I can see it anyway with opimd command Jul 28 09:00:16 so, I suppose it's double Jul 28 09:00:27 double storage I mean Jul 28 09:00:31 you delete it with an other command? Jul 28 09:00:58 I delete it by the application which uses only sim backend Jul 28 09:01:18 so it's not removed, i suppose, from sqlite Jul 28 09:01:56 --> infact I've many sms in my .db not on sim too Jul 28 09:01:56 opimd caches all messages on startup, and if you delete a message in the background without restarting opimd it's not delteted from the cache Jul 28 09:03:00 aah! here I am! :D so it's a memory only cache? volatile? Jul 28 09:03:07 yes Jul 28 09:03:21 dos1 will change this in near future Jul 28 09:03:32 but currently it's like that Jul 28 09:03:46 try restarting frameworkd Jul 28 09:03:49 so, if I reboot I'll loose all the cached sms not on sim? Jul 28 09:04:09 yes, i think so Jul 28 09:04:20 mmh... let's try! ;) Jul 28 09:05:10 [rebooting] Jul 28 09:06:21 [please wait ;) ] Jul 28 09:14:03 Heinervdm: u r right! :P Jul 28 09:14:16 is there already work being done on a webradio application? I've started one for the webradio I listen to and like to generalise it. Jul 28 09:14:21 cache flushed Jul 28 09:14:50 blindcoder: vagalume? Jul 28 09:14:58 a21_work: vagalume is just for last.fm, no? Jul 28 09:15:16 blindcoder: wouldn't it be nice to integrate something like this into intone? Jul 28 09:15:22 a21_work: I mean arbitrary radios (mp3.webradio.rockantenne.de in my case) Jul 28 09:15:32 i think one musicplayer is enough Jul 28 09:15:42 umh... maybe... I do not use it really... but I think anyway you could expand with that feature if not already done Jul 28 09:15:44 Heinervdm: sure is, that's why I'm asking :) Jul 28 09:16:12 Heinervdm: my app is just a python script starting mplayer in the background and parsing /etc/asound.conf for output modules Jul 28 09:16:17 it shouldn't be that hard, because mplayer can play streams Jul 28 09:16:27 :) Jul 28 09:17:33 I think it'd better as a feature of vagalume... try to contact the developer! Jul 28 09:17:58 a single application for every kind of radios Jul 28 09:18:02 ;) Jul 28 09:18:17 I hate many icons on my desk Jul 28 09:18:22 :P Jul 28 09:18:28 that's right! Jul 28 09:18:54 yeah. vagalume also lacks the possibility to stream via different pcm.* modules yet. I'll try to get in touch with the devs Jul 28 09:19:10 fine! :) Jul 28 09:19:33 this is what I call community and open source Jul 28 09:19:48 ;) Jul 28 09:20:00 I just got my FR last week and I'm currently trying to implement functionality I want/need. I already tried with something 'deeper' (shr-contacts) but I don't get it really so I'm looking elsewhere Jul 28 09:20:32 just unfamiliar with phonegui and which phonegui is used (efl? gtk? none?) Jul 28 09:21:26 mrmoku is the master of libphonegui, so if you want to know something ask him ;) Jul 28 09:21:56 I realised yesterday that my freerunner cant display mms... Jul 28 09:23:39 I wanted to develop mms support for this august... but I found it's a too big project for my spare time on holydays Jul 28 09:23:42 :( Jul 28 09:23:56 I never used mms, so I have no use for it. don't even have a data plan yet Jul 28 09:24:14 But I _need_ extended phonebook support Jul 28 09:24:26 mrmoku pointed be to dos1 already Jul 28 09:24:44 but I think until I know what is what and goes where will take too much time Jul 28 09:25:03 so I hope someone else will do that and leave me with less critical stuff :) Jul 28 09:25:53 blindcoder: the backend can handel all fields you can imagine, only the gui is lacking the support Jul 28 09:26:36 Heinervdm: yes, so I've been told Jul 28 09:28:12 one problem is, that shr-contacts only du small things, most is done by phonegui Jul 28 09:28:35 Heinervdm: I've currently got the phonegui sourcecode open Jul 28 09:29:16 blindcoder: i'm wating for phonegui-efl2 to be finished, after that one can improve it Jul 28 09:29:17 trying to make sense of it Jul 28 09:30:05 ah so there's use to make changes to -efl now anyway? Jul 28 09:30:18 -efl2 Jul 28 09:30:31 -efl2 uses opimd Jul 28 09:30:52 but it crashes very often, i heard :D Jul 28 09:31:23 next thing i'm planing to do is dates support for opimd Jul 28 09:31:28 I see Jul 28 09:31:43 I got the shr Makefile setup from July 21st. Guess -efl2 is not in there yet? Jul 28 09:31:51 but i have to wait for dos1, because he want to change some general things Jul 28 09:32:50 blindcoder: it has to be there, because you can install it Jul 28 09:33:24 blindcoder: shr-unstable/openembedded/recipes/shr/libframeworkd-phonegui-efl2_git.bb Jul 28 09:34:54 Heinervdm: ah, thanks. I didn't run a complete build, maybe that's why it wasn't there. So far I followed the wiki instructions (I'm used to a different distribution build kit) Jul 28 09:35:12 hi Jul 28 09:35:19 are there any shr developers here? Jul 28 09:35:47 Michal1: most are sleeping, but just ask :) Jul 28 09:36:00 would it be possible to put qt libraries in the feed? Jul 28 09:36:31 Michal1: for this you have to wait for mrmoku Jul 28 09:36:48 ok Jul 28 09:38:06 blindcoder: you didn't use the Makefile from SHR? Jul 28 09:40:48 did anyone try litephone yet? Jul 28 09:41:01 I have :-) Jul 28 09:44:15 Michal1, how is it? Jul 28 09:44:52 Heinervdm: I do, but so far I just did bitbake -c build shr-contacts Jul 28 09:44:57 I like it Jul 28 09:45:03 I'm the author though :-) Jul 28 09:45:28 blindcoder: ok, then just do bitbake -c build libframeworkd-phonegui-efl2 too ;) Jul 28 09:45:34 Heinervdm: just running :) Jul 28 09:45:41 Heinervdm: ERROR: Nothing PROVIDES 'libframeworkd-phonegui-efl2_git' Jul 28 09:45:43 oh, lol, i posted a suggestion for you on the wiki, upcoming appointments from some calendar, perhaps even a builtin one Jul 28 09:45:47 :-( Jul 28 09:46:04 blindcoder: without _git everything behind the _ is about versioning Jul 28 09:46:28 Heinervdm: ah, okay. Jul 28 09:46:34 i always though the current SHR is missing a home screen, with easy and fast access to common apps... landing in the menu screen with just buttons is not very phone-like Jul 28 09:46:39 * blindcoder .oO( as I said, used to different distribution build kit :/ ) Jul 28 09:48:32 mmh I'm not able to update frameworkd-config-shr-dev Jul 28 09:48:38 you? Jul 28 09:49:05 a21_work, i dont think anyone can, frameworkd-config-shr with a matching version is missing from the repo for some reason... Jul 28 09:49:16 aah fine Jul 28 09:49:19 tnx Jul 28 09:49:29 a21_work: this pop's up every X months for some reason :D Jul 28 09:49:45 ghghgh Jul 28 09:49:55 voodoo? Jul 28 09:50:02 ;) Jul 28 09:50:08 but because it's a -dev package, i purged it Jul 28 09:50:44 are you able to get a fix with gps? Jul 28 09:51:12 since 21/JUL I'm no longer able to get it Jul 28 09:51:13 tried 2 day's ago and it worked, but others have problems Jul 28 09:51:36 a21_work: theres a thread on the mailing list about that Jul 28 09:51:36 Heinervdm: do u have also a hcsd? Jul 28 09:51:48 hcsd? Jul 28 09:51:53 oh yes Jul 28 09:51:56 hc Jul 28 09:52:14 i'm running from flash Jul 28 09:52:31 mhh... maybe this is the problem? Jul 28 09:53:25 I'm also running from sd card and don't get a gps fix Jul 28 09:53:35 but if that's the reason that would be very strange Jul 28 09:54:13 well.. a regression is always possible Jul 28 09:54:57 sure Jul 28 09:55:29 can't do much checking right now. Only have a windows machine here which tends to crash with the freerunner plugged in Jul 28 09:55:30 a21_work: http://lists.openmoko.org/pipermail/community/2009-July/052061.html Jul 28 09:56:15 Heinervdm: reading Jul 28 09:56:22 :) Jul 28 09:56:31 Michal1, you should talk to mrmoku to add ur app to the shr repo... Jul 28 09:59:52 btw, I've now set up my shr to be usable to me on a day-to-day basis. What is the recommended way of making a backup? take out sd card, put into card reader and run dd? Jul 28 10:02:53 F4t: first I have to ask him to put Qt in the repo Jul 28 10:04:50 odd I know dos1 was asking about GPS suddenly dying and not coming back, GSM has totally gone here now Jul 28 10:05:03 without an upgrade or any change Jul 28 10:13:22 For me GPS is working good, i just got a fix in the room Jul 28 10:13:46 tig|: are you sure that's not userland software issue? Jul 28 10:14:56 PaulFertser: actually it looks like the whole DBUS ophonekit lot has gone titsup here. I will do some more digging in a minute Jul 28 10:15:30 tig|: ah, that's another matter. dos1's problems are a magnitude nastier. Jul 28 10:21:40 Heinervdm: how did you check? tangogps? Jul 28 10:22:51 in settings -> gps I set manual and on, got a Time immediately but Fix: no fix Jul 28 10:23:02 Satellite Details is empty Jul 28 10:25:56 blindcoder: yes tangogps, got gps koordinates Jul 28 10:26:32 Heinervdm: do I need to start a service or something? or just enable gps in settings? Jul 28 10:26:55 blindcoder: gps is enabled automatically whenever some program needs it. Jul 28 10:30:08 blindcoder: i just started tangogps, nothing else Jul 28 10:30:14 okay, waiting for tangogps to report something != 0.000000 - 0.000000 Jul 28 10:30:53 i just tried again Jul 28 10:31:09 went outside and got fix 2D after 3 Minuets Jul 28 10:31:52 phone is sitting at the window, Sat 0/0 Jul 28 10:32:42 I got sat 10/0 directly Jul 28 10:42:43 starting fso-gpsd -N -D 9 gets me lots of Jul 28 10:42:49 Timeout sending data to client Jul 28 10:52:26 mrmoku, here? Jul 28 10:55:42 blindcoder: yay, someone with my problem! Jul 28 10:56:11 blindcoder: check in dmesg if you have rxerr messages Jul 28 10:56:19 blindcoder: i have exactly the same problem Jul 28 10:56:34 blindcoder: when did your gps work last time? Jul 28 10:56:35 dos1: yes, I have 1 rxerr Jul 28 10:56:39 dos1: never ;) Jul 28 10:56:49 dos1: flashed it directly with shr from 21st july Jul 28 10:56:57 blindcoder: it worked at morning 2 days ago Jul 28 10:57:08 blindcoder: but it just suddenly stopped to work few hour later Jul 28 10:57:19 without any change, upgrade or reboot Jul 28 10:57:36 and now it doesn't work at all, with the same symptoms as your Jul 28 10:58:01 rxerr: port=1 ch=0x00, rxs=0x0000000c Jul 28 10:58:08 dos1: same here, got 3 rxerr here Jul 28 10:58:23 PaulFertser: do you know how to debug it somehow? Jul 28 10:58:51 tig|: when your gps worked last time? :x Jul 28 10:58:57 blindcoder: I have two like that and then one ending with a d Jul 28 10:59:13 dos1: good question I have not used it for a while Jul 28 10:59:21 :( Jul 28 10:59:39 i would like to know, if it stopped to work for us at the same time :( Jul 28 10:59:43 I will check the datestamp on the image I flashed with Jul 28 10:59:59 tig|: well, i don't think that's image Jul 28 11:00:04 i even don't think that's software problem Jul 28 11:00:17 eventually kernel, but kernel didn't changed in SHR for looong timwe Jul 28 11:00:22 s/timwe/time/ Jul 28 11:00:22 dos1 meant: eventually kernel, but kernel didn't changed in SHR for looong time Jul 28 11:00:43 dos1: i'm getting rxerr too, but my gps is working ! Jul 28 11:01:01 dos1: kernel 8th July, SHR-Lite 21st July Jul 28 11:01:24 Heinervdm: i don't think that rxerr is all evil :P Jul 28 11:01:51 dos1: i just wanted to tell you :) Jul 28 11:02:24 now you know that i can run with rxerr too :) Jul 28 11:02:38 PaulFertser: before it stopped to work, there was few minutes with my GPS being "noisy" - it got fix, then it said "no fix" but it still reported few satelites used to get fix... Jul 28 11:02:53 and after that it stopped to work, only time is displayed Jul 28 11:03:05 dos1: it'd be nice to 'scope the lines... Jul 28 11:04:06 PaulFertser: scope lines? please explain in more simple english ^^'' Jul 28 11:05:10 dos1: oscilloscope connected to UART lines... Jul 28 11:05:11 moooooin Jul 28 11:05:25 ohh Jul 28 11:05:43 DocScrutinizer: heh, saw a familiar word? ;) Jul 28 11:05:44 well, i don't think i'm able to do that now ;x Jul 28 11:05:57 sure :) Jul 28 11:06:06 DocScrutinizer: btw, i was testing autorestarting charging by PMU. According to CC it was triggered at 77% Jul 28 11:06:15 76% even Jul 28 11:06:17 F4t, now here Jul 28 11:06:21 ack Jul 28 11:06:32 mrmoku, i added static wifi config, wanna test? Jul 28 11:06:39 3989000 uV Jul 28 11:07:02 mrmoku, and once testing is done, ill be marking 0.4 Jul 28 11:07:09 PaulFertser: 75%..80% here Jul 28 11:07:23 F4t, not right now :P Jul 28 11:07:25 DocScrutinizer: my next task is to kill all the shit Werner outlined in his email. Jul 28 11:07:36 huh? Jul 28 11:08:06 F4t, I can build it for others to test though Jul 28 11:08:07 PaulFertser: whee ! ;-) Jul 28 11:08:10 I'm totally out of sync wrt mail Jul 28 11:08:30 mrmoku, i prefer to mark it 0.4 before building... Jul 28 11:08:54 DocScrutinizer: that was from a long time ago :) Jul 28 11:08:55 ill bug someone else then, DocScrutinizer, PaulFertser, anyone want to test wifi static config with mokonnect? :) Jul 28 11:09:04 wpwrak: i think a good compromise wrt reporting charging status is to report it based on bq27000 readings if present and report based on PMU active_charger _register_ bit if not. Jul 28 11:09:49 DocScrutinizer: Subject: "battery charging" when it's not - an analysis Jul 28 11:10:12 F4t: sorry, not now. And i'm a debian user, i lack elementary here. Jul 28 11:10:29 PaulFertser, oh, thanks anyway... Jul 28 11:10:57 PaulFertser: sounds like a plan. maybe add some means for user space to find out in which mode you are, considering that the PMU doesn't have a very clear idea of what it's actually doing :) Jul 28 11:11:02 PaulFertser: first define semantics and use of status:charging, *then* decide how to implement Jul 28 11:11:23 DocScrutinizer: i think charging means actually charging the battery. Jul 28 11:11:50 ok, that's what I thought as well Jul 28 11:12:20 wpwrak: if a gta01 driver is used, it's not bq27000. And if not, it doesn't report status anyway. Jul 28 11:13:06 DocScrutinizer: there's a small window for false positives: when all the current flowing through the charges goes to the GSM modem. I don't think it's common enough to care about. Jul 28 11:13:08 PaulFertser, DocScrutinizer: "charging" when charging, "not charging" when full and discharging with usb cable (waiting for autorestart), "discharging" when really discharging without USB Jul 28 11:13:11 isn't it that way? Jul 28 11:13:30 PaulFertser: ack Jul 28 11:13:31 dos1: i don't like "not charging". Jul 28 11:14:12 dos1: 100%ack Jul 28 11:15:28 dos1: alas now it's a little screwed as it reports "charging" when actually it's "not charging" Jul 28 11:15:40 It's a battery property. If we can't say for sure by quering bq27000 i'd say we should report discharging whenever charger is not enabled. Jul 28 11:15:42 15min after initial charge finished Jul 28 11:16:34 DocScrutinizer: my logs showed that with the correct threshold current it reports "not charging" right after the charge is finished. Jul 28 11:16:38 actually nobody cares to distinguish not-chg from dischg for now Jul 28 11:17:43 DocScrutinizer: i saw errourneous "Charging" after kernel tried (but failed) to trigger charging restart. In fact battery continued to discharge. Jul 28 11:17:45 PaulFertser: (not chg) right, for 15min, untill stupid (luckily) broken charge re-enable of kernel kicks in. Or is this fixed by now? Jul 28 11:18:00 hehe, Jul 28 11:18:18 hmm... battery, gta01, bq27000 in the same discussion.. that's frightening /o\ Jul 28 11:18:23 what I say from beginning Jul 28 11:18:46 rtp: no risk, no fun ;) Jul 28 11:18:53 rtp: you should definetely try my patches i just sent to the ML, they should give a pretty accurate estimate of remaining capacity on gta01. Jul 28 11:19:07 PaulFertser: will you also add an indicator if the bq27000 is confused about the battery state ? Jul 28 11:19:18 rtp: just wait when we add the calypso :) Jul 28 11:19:33 DocScrutinizer: hehe :) Jul 28 11:19:58 PaulFertser: did you fix the unit reported on the sysfs files too ? :) Jul 28 11:20:13 wpwrak: i think there's no need. If bq27000 is there, we count on its readings fully. If not, we report not present for status and ENODEV for all other properties (as was noticed as fixed by larsc) Jul 28 11:20:17 rtp: sure Jul 28 11:20:34 wpwrak: ah, confused. Jul 28 11:20:37 PaulFertser: great :) Jul 28 11:20:52 wpwrak: you mean it doesn't think its readings are accurate. Sure, will do. Jul 28 11:21:26 what's news with qi? need a job! Jul 28 11:21:39 PaulFertser: I've notice something "weird" on my gta01, the current sysfs file is negative when not plugged on usb. maybe it's normal ? Jul 28 11:21:57 echo 0 > //sys/class/i2c-adapter/i2c-0/0-0073/pcf50633-regltr.7/neo1973-pm-gps.0/power_on && Jul 28 11:21:58 sleep 1s && echo 1 > /sys/class/i2c-adapter/i2c-0/0-0073/pcf50633-regltr.7/neo1973-pm-gps.0/power_on && stty -F /dev/ttySAC1 -echo && cat -u /dev/ttySAC1 | grep -v ^$ Jul 28 11:22:04 rtp: hm, here it's negative when it's charged. Jul 28 11:22:28 PaulFertser: it should display data from GPS, yep? Jul 28 11:22:50 PaulFertser: here it just quits after second (sleep) Jul 28 11:22:58 PaulFertser: fwiw, it says now '-53' when I cat the sysfs file. Jul 28 11:23:00 dos1: I have GPS! Jul 28 11:23:01 rtp: so, no, looks like it's wrong. Anyway gta01 current readings are inaccurate a bit more than wholly. Jul 28 11:23:05 PaulFertser: kewl. thanks ! Jul 28 11:23:23 dos1: I changed automatic to manual then turned on and it is working here Jul 28 11:23:37 PaulFertser: :( Jul 28 11:23:47 now it displayed: Jul 28 11:23:49 01,02,ANTARIS ATR062x HW 80040001*26 Jul 28 11:23:51 $GPTXT,01,01,02,ROM CORE , Jul 28 11:23:52 root@(none) /sys/class $ Jul 28 11:24:00 wpwrak: btw i spent quite some time trying to understand wtf sync reading from pcf50633 doesn't work. Still wondering if i solved it properly. Jul 28 11:24:06 and now fsck it, the blasted thing has WSOD Jul 28 11:24:14 wpwrak: (sync adc reading) Jul 28 11:24:38 dos1: looks ok? Jul 28 11:25:03 PaulFertser: but nothing more. it quited after that (i didn't ctrl+c'ed it ;P) Jul 28 11:25:44 dos1: mwester_ is the serial guru here. I still don't understand why and what happens when you "cat" a serial device. So no idea why it could quit. Jul 28 11:25:57 mwester_: ping? :D Jul 28 11:26:38 dos1: concurrent access to the device? Jul 28 11:26:43 rtp: is the current positive when you plug a charger? If yes, i guess it's just a sign mismatch. Jul 28 11:26:43 PaulFertser: hmm, what's the sync reading issue ? Jul 28 11:26:54 wpwrak: please read the kernel ML :) Jul 28 11:26:55 DocScrutinizer: no, frameworkd is stopped Jul 28 11:27:14 X are also stopeed Jul 28 11:27:19 PaulFertser: yeah, it's >0 when plugged Jul 28 11:27:35 rtp: ok, will fix though it's of no real use anyway. Jul 28 11:28:23 PaulFertser: it's not a matter of use, it's a matter of correctness :) Jul 28 11:28:27 DocScrutinizer: htop, init, udevd, portmap, sshd, syslogd, klogd, getty, sshd, sh Jul 28 11:28:30 rtp: sure Jul 28 11:28:34 DocScrutinizer: that's everything running now on my neo :P Jul 28 11:28:46 ah, bunch of new stuff ... Jul 28 11:29:05 once it returned Jul 28 11:29:07 ,01,02,ANTARIS ATR062x HW 80040001*26 Jul 28 11:29:07 dos1: doesn't help a lot. check lsof Jul 28 11:29:08 $GPTXT,01,01,02,ROM CORE Jul 28 11:29:11 and then it quit Jul 28 11:29:18 wpwrak: and it started as "let's make a simple dumb battery driver work on gta02" :) Jul 28 11:29:19 another time Jul 28 11:29:21 ,01,02,ANTARIS ATR062x HW 80040001*26 Jul 28 11:29:22 $GPTXT,01,01,02,ROM CORE ,,,,,,,N Jul 28 11:29:25 and another nothing :x Jul 28 11:29:42 PaulFertser: btw, it's the pcf the responsibl for rtc on gta01, right ? Jul 28 11:29:55 DocScrutinizer: says nothing, only ttySAC2 is used by getty Jul 28 11:29:56 rtp: but it's pcf50606 on gta01. Jul 28 11:30:20 dos1: stty settings for your serial device might be screwed Jul 28 11:30:45 dos1: seen strange things like your issue with calypso tty as well Jul 28 11:31:00 does strace say anything? Jul 28 11:31:06 DocScrutinizer: how to check it? Jul 28 11:31:14 duh, no idea Jul 28 11:31:15 strace cat /dev/whatever Jul 28 11:31:17 sorry Jul 28 11:31:35 tmzt: good idea Jul 28 11:31:40 PaulFertser: I know it's a pcf50606. my question was more pcf50606 vs samsung soc as rtc device Jul 28 11:31:51 DocScrutinizer: ok, i have Jul 28 11:31:53 rtp: fyi dumb battery driver capacity estimation is somewhat accurate during discharge only. Jul 28 11:32:04 rtp: it's pcf Jul 28 11:32:04 device will probably throw some error that aborts cat Jul 28 11:32:30 PaulFertser: the discharge case is the most important one imho :) Jul 28 11:32:40 DocScrutinizer: http://pastebin.com/m7d694487 Jul 28 11:33:15 DocScrutinizer, tmzt: and that's ending of strace: http://pastebin.com/meb64cc Jul 28 11:33:35 PaulFertser: ok. I'll have a look. Looks like the rtc irq stuff is buggy /o\ Jul 28 11:34:32 yeah, that makes no sense Jul 28 11:34:42 rtp: also SD vcc setting is fubared on gta01. Need to fix that too... Jul 28 11:34:42 Hey there Jul 28 11:34:46 PaulFertser: many great adventures start like that. e.g., "let's find a shortcut to china" :) Jul 28 11:35:07 wpwrak: :D at least i learned something during the venture. Jul 28 11:36:03 F4t, will test it now as I'm going downstairs to have lunch :) Jul 28 11:37:02 mrmoku, ok :D Jul 28 11:37:07 PaulFertser: how come ? Jul 28 11:37:17 bbiab Jul 28 11:37:31 rtp: looks like bad merge Jul 28 11:37:54 Heinervdm: can you do stty -F /dev/ttySAC1 --all ? Jul 28 11:38:13 dos1: with or without fix? Jul 28 11:38:30 Heinervdm: just when powered up Jul 28 11:39:11 PaulFertser: dohh... I really need to update my shr repo and starting playing with the kernel. Jul 28 11:39:17 does anyone know what happened to the dollar lately? Jul 28 11:40:03 dos1: invalid argument --all Jul 28 11:40:07 huh? $ Jul 28 11:40:29 hmm Jul 28 11:40:34 i have coreutils installed :P Jul 28 11:41:18 dos1: http://pastebin.com/d14d99509 without --all Jul 28 11:42:05 installing coreutils needs some time Jul 28 11:42:11 rtp: i'm still waiting for the guy who promised to give me a gta01... Jul 28 11:42:52 PaulFertser: still? :( Jul 28 11:43:29 DocScrutinizer: he wrote on ML that he'll do that in several weeks. About a month passed... Jul 28 11:44:22 DocScrutinizer: do you see something strange in stty outputs? Jul 28 11:44:25 PaulFertser: oh... too bad. I though you got it /o\ Jul 28 11:44:44 hmm, not really. Jul 28 11:45:15 Hand to check semantics of read() return code though Jul 28 11:45:25 it reads "" and returns 0 Jul 28 11:45:39 this might mean EOF Jul 28 11:45:40 * Weiss wonders if we should point "HEAD" of git.openmoko.org/kernel.git to "andy-tracking" rather than "master" Jul 28 11:46:25 probabl depending on stty settings Jul 28 11:47:00 http://pastebin.com/m55445b33 now with --all Jul 28 11:48:08 * DocScrutinizer thinks there's some worms in tty-driver Jul 28 11:48:29 dos1: it's exactly the same output Jul 28 11:48:36 as yours Jul 28 11:51:38 I think read returns the amount read Jul 28 11:51:48 to me it seemed as if a stty call to set parameteres had effect only *after* open() until first read(), and it's quite hard to establish that from cmdline Jul 28 11:52:42 tmzt: sure, that's what I'd expect as well Jul 28 11:52:53 If `read' returns at least one character, there is no way you can tell whether end-of-file was reached. But if you did reach the end, the next read will return zero. Jul 28 11:53:08 but tty is rather tricky, with brea and EOF and dunno what else Jul 28 11:53:15 we need to check errno Jul 28 11:53:28 I think it should have been EAGAIN Jul 28 11:53:30 which is not eof Jul 28 11:53:36 yup Jul 28 11:53:52 tmzt, DocScrutinizer: just tell me what i should do :x Jul 28 11:54:05 fix the kernel driver :) Jul 28 11:55:05 tmzt: if only kernel was written in python... ;D Jul 28 11:55:22 cat /dev/tty &; killall -SIGSTOP cat; stty -sane (or whatever); killall -sigcont cat Jul 28 11:55:28 tty layer is confusing Jul 28 11:55:37 you bet it is Jul 28 11:57:27 WTF? Jul 28 11:57:33 DocScrutinizer: i just did stty -F /dev/ttySAC1 sane Jul 28 11:57:34 there's definitely some bug in tty-driver or stty Jul 28 11:57:39 and it works now... Jul 28 11:58:01 :) Jul 28 11:58:28 i'm checking now, what changed in stty --all output Jul 28 11:58:37 bet nottin Jul 28 11:59:12 that's what I seem to remember with my similar problems with calypso-tty Jul 28 12:01:02 DocScrutinizer: http://pastebin.com/m14fdc0b7 Jul 28 12:02:13 but there is still question - why it happened? Jul 28 12:02:20 dos1: pff, that helps a lot, no ;] Jul 28 12:02:32 ;DD Jul 28 12:02:55 as I guessed before, there seems to be a flaw somewhere Jul 28 12:03:30 sth like open() resetting some of the previously set params sometimes Jul 28 12:03:36 or dunno what else Jul 28 12:05:48 maybe stty not able to set all the parameteres to the requested state of given explicitely as a parameter, but does on -sane Jul 28 12:06:30 or stty not correctly reporting all the settings Jul 28 12:06:51 thus spilling them on restore Jul 28 12:06:58 realy no idea Jul 28 12:07:16 I gave up on this after a day, before I got nuts on it Jul 28 12:07:27 what's different? Jul 28 12:07:35 crtcts is -crtcts on both Jul 28 12:07:57 oh, htat's a diff Jul 28 12:08:36 tmzt: http://pastebin.com/m2f36168e Jul 28 12:08:39 a little better :D Jul 28 12:09:43 brkint Jul 28 12:09:44 ? Jul 28 12:09:59 [-]brkint Jul 28 12:10:00 breaks cause an interrupt signal Jul 28 12:10:11 but we don't get INTR, we just get an exit Jul 28 12:10:47 it works from cat now, doesn't work from ogpsd Jul 28 12:10:53 still this might be the cause Jul 28 12:11:51 well, this was the point where I was happy with *any* working uSD-gsm-flashing solution I was able to come up with ;-) Jul 28 12:12:33 nothing else had it open? Jul 28 12:12:47 as you see exactly the same bullshit when trying to talk to modem via cat and all the kinky means suggested in original manual method Jul 28 12:13:06 tmzt: nothing else Jul 28 12:13:12 was gps muxed? Jul 28 12:13:20 muxed? Jul 28 12:13:21 on gta01 Jul 28 12:13:27 shared Jul 28 12:13:29 i have gta02 Jul 28 12:13:39 oh, didn't know this was an issue Jul 28 12:13:48 but right, gta01 doesn't do nmea anyway Jul 28 12:14:07 gllin Jul 28 12:14:31 cya Jul 28 12:14:33 can you restore the first set of outputs and then set -brkint ? Jul 28 12:14:43 the first set of parameteres Jul 28 12:16:09 F4t, had no luck finding my network Jul 28 12:16:27 mrmoku, how about trying it as hidden? Jul 28 12:16:39 F4t, even if it's not hidden? Jul 28 12:16:42 ok Jul 28 12:16:52 it might work... Jul 28 12:17:21 * mrmoku back downstairs Jul 28 12:21:00 tmzt: just tell me how ;) Jul 28 12:23:50 F4t, gives me a NotSupported error from connman ;) Jul 28 12:24:01 in _ConnectHiden line 400 Jul 28 12:24:05 +d Jul 28 12:28:47 dos1: to fix kernel? Jul 28 12:28:59 mrmoku, weird shit Jul 28 12:29:02 oh, restore the stty stuff Jul 28 12:29:20 take the paste bin, stty -F options I think is the command Jul 28 12:29:54 they just need to be on one line (or protected with \ ) Jul 28 12:49:06 PaulFertser, it seems so (not light weight enough for slugos) Jul 28 12:49:20 and just an unnecessary dep. Jul 28 12:50:39 raster, got any idea what's wrong with the code I'm pasting in your query once in a while? :) Jul 28 12:54:01 the python? Jul 28 12:55:08 aye. Jul 28 12:55:16 though same behavior with exactly the same c code Jul 28 12:55:22 please try to fix it Jul 28 12:55:55 as elementary isn't documented and doing trial and error/asking ppl/doing what you said didn't help :) I really suspect it's a bug, though you claim it's not... Jul 28 12:56:05 have u tried using win as parent for adds? Jul 28 12:56:31 I'm writing a widget Jul 28 12:56:35 so I used the parent Jul 28 12:56:42 (in the C code) Jul 28 12:56:46 and in the python yeas Jul 28 12:56:51 do u have the c code version? Jul 28 12:56:51 I tried Jul 28 12:56:56 raster, aye. Jul 28 12:57:04 sec, will up it. Jul 28 12:59:14 http://pastebin.com/m501f4781 Jul 28 12:59:20 one of my many brute force attempts ;) Jul 28 13:00:20 as you see Jul 28 13:00:29 it's a tweaked version of the list widget Jul 28 13:03:46 also tried asking dos1 (who wrote the python code) Jul 28 13:09:04 errr.... ummm.... hmmm Jul 28 13:09:25 you're trying to make your own widget? Jul 28 13:09:44 elm's widget-making apis arent available externally.... Jul 28 13:09:45 bbl Jul 28 13:50:54 bak Jul 28 13:51:00 TAsn: what are you trying to do? Jul 28 13:51:38 raster, editing the list image Jul 28 13:51:50 ??? Jul 28 13:51:57 do you remember what we discussed about? (indexed lists?) Jul 28 13:52:13 I want to add a feature to the list widget Jul 28 13:52:32 "show index" Jul 28 13:52:32 everything is nice and easy Jul 28 13:52:32 except for the issue I'm talking about.. Jul 28 13:52:54 show index... Jul 28 13:52:59 u mean u want like an a-z thing Jul 28 13:53:04 to quickly jump to a spot? Jul 28 13:53:14 yes. Jul 28 13:53:26 wait - i've got that on a funded todo Jul 28 13:53:30 someoen want to pay for it Jul 28 13:53:31 :) Jul 28 13:53:44 raster, :) Jul 28 13:53:47 cool. Jul 28 13:53:49 (sorry to steal the wind in your sails!) Jul 28 13:54:00 but i'll implement it in the scroller Jul 28 13:54:00 raster, then just make it. Jul 28 13:54:15 whatever, I don't care. Jul 28 13:54:20 though please do it ;) Jul 28 13:54:22 as aindex mechanism (be able to add index items with virtual pan regions for them) Jul 28 13:54:33 I want it to be done Jul 28 13:54:36 I don't care how. Jul 28 13:54:36 and the scroller will figure out how many fit and how to pack/show them Jul 28 13:54:37 :) Jul 28 13:54:46 no probs Jul 28 13:54:53 raster, that's exactly what I was trying to do. Jul 28 13:54:56 but its a "want" on a list of things to be done soon Jul 28 13:55:07 :| Jul 28 13:55:51 next time please tell me in advance that you wanna do it yourself... Jul 28 13:56:06 this only has come up in the last 2 weeks or so Jul 28 13:56:06 as we already discussed this Jul 28 13:56:11 oh cool Jul 28 13:56:13 :) Jul 28 13:56:15 then it's not your fault. Jul 28 13:56:24 from clients Jul 28 13:56:41 been punting project proposal back and forth Jul 28 13:56:44 thus... its new Jul 28 13:57:01 ok. Jul 28 13:57:17 it's quite easy to implement Jul 28 13:57:36 as there are nice APIs available Jul 28 13:57:41 that make everything easy ;) Jul 28 13:58:00 but cool. Jul 28 14:00:01 which api's? Jul 28 14:01:03 hm.. forgot the name Jul 28 14:01:07 something for jumping to pos Jul 28 14:01:09 sec Jul 28 14:02:01 elm_list_item_show Jul 28 14:02:18 (what's implemented there actually) Jul 28 14:02:22 makes the hopping easy Jul 28 14:02:28 so there's nothing really hard left. Jul 28 14:03:42 raster: what about other interesting things in your todo? what can we expect soon in elementary? :) Jul 28 14:05:58 ooh that Jul 28 14:06:03 thats yeah - easy Jul 28 14:06:17 scroller already takes care of that too by hopping to a region too Jul 28 14:06:28 the hard work is display of the index Jul 28 14:06:35 eg - if u dont have room for all items.. what do you do? Jul 28 14:08:47 raster, aye. Jul 28 14:08:54 though that's decisions Jul 28 14:09:02 you can either have a scrolling index limited to 10 items Jul 28 14:09:06 and you can scroll through Jul 28 14:09:14 or just decide about a minimal size Jul 28 14:09:20 and put as many indexes as you can shove there. Jul 28 14:10:02 yeah Jul 28 14:10:06 i'm going to do the latter Jul 28 14:10:18 dos1: ummm.. good q. :) Jul 28 14:12:12 anyhow, raster I dropped it Jul 28 14:12:22 so I changed the bug in our trac to upgstream ;) Jul 28 14:12:25 upstream* Jul 28 14:12:29 so please make it ;) Jul 28 14:12:52 :) Jul 28 14:13:06 quick stupid question: looking at the screen, on which side is the microphone? Jul 28 14:13:10 Ainulindale: have some time now? Jul 28 14:13:18 blindcoder, down. Jul 28 14:13:24 whene the mic hole is. Jul 28 14:13:31 TAsn: :P Jul 28 14:13:38 TAsn: I have two holes, one left, one right Jul 28 14:13:49 that's speakers (I think) Jul 28 14:13:55 I'm talking about the big hole Jul 28 14:13:59 TAsn: and apparently terrible recording quality Jul 28 14:13:59 inside it there's a small hole Jul 28 14:14:04 ah Jul 28 14:14:13 okay, I see it Jul 28 14:14:45 thanks Jul 28 14:15:45 is there a way to record something in the shr image? Jul 28 14:16:21 blindcoder, call recorder/dictator Jul 28 14:17:17 TAsn: dictator is probably what I'm looking for, thanks ;) Jul 28 14:18:42 TAsn: about "holes" at sides - left one is speaker, right one is just hole Jul 28 14:18:47 TAsn: both was speakers in gta01 Jul 28 14:18:53 s/was/were/ Jul 28 14:18:53 dos1 meant: TAsn: both were speakers in gta01 Jul 28 14:21:07 cool. Jul 28 14:21:12 ;) Jul 28 14:26:59 mrmoku: no, sorry Jul 28 14:27:01 swamped here Jul 28 14:27:09 Ainulindale: ok Jul 28 14:54:21 mrmoku: I have to answer to a weird and boring document and that drives me nuts Jul 28 14:55:01 Ainulindale: hehe :P Jul 28 14:57:28 ptitjes: I'm looking at libfso-glib... but I don't get it... Jul 28 14:57:36 mrmoku: it's generated Jul 28 14:58:01 Ainulindale: well I'm looking at the generated frameworkd.h :) Jul 28 15:00:31 howdy Jul 28 15:00:35 err freesmartphone.h :P Jul 28 15:06:08 mrmoku: what don't you get ? Jul 28 15:09:04 heh Jul 28 15:11:42 mrmoku: I must leave to go to the doctor... Jul 28 15:11:49 ptitjes: can't see how to do async calls Jul 28 15:12:02 mrmoku: oh Jul 28 15:12:02 but go to the doctor first :) Jul 28 15:12:15 in the traditionnal Gio way Jul 28 15:12:32 * mrmoku lacks the Gio tradition foo ;) Jul 28 15:12:32 call _async with a callback Jul 28 15:12:45 ahh, ok Jul 28 15:13:00 in the callback call _finish to get the results of the async method Jul 28 15:13:40 ok, thanks Jul 28 15:13:43 if you try it you'll see that arguments just do come at the right place ;) Jul 28 15:13:55 np Jul 28 15:17:19 mrmoku, please write a tutorial when you are done Jul 28 15:17:19 ;] Jul 28 15:17:37 TAsn: hehe, lets see ;) Jul 28 15:17:45 gps works! :) Jul 28 15:18:04 dos1, just like that? Jul 28 15:18:04 or did you do something? Jul 28 15:18:04 mrmoku, what about a new unstable image for me to test? Jul 28 15:18:17 (my current one is broken because a dumbass sent a bad opkg patch) Jul 28 15:18:31 TAsn: well, i don't know Jul 28 15:18:47 TAsn: building a light one now Jul 28 15:18:56 TAsn: when i'll finish eating icecream i'll describe you what i did ;d Jul 28 15:19:36 mrmoku, since I don't care about a non-lite one, COOL! :) Jul 28 15:19:44 mrmoku, it's a testing candidate, right? Jul 28 15:20:01 every unstable image is a testing candidate these days :P Jul 28 15:20:40 mrmoku, which in turn means this is a testing candidate, cool :) Jul 28 15:24:04 mrmoku: you're putting it online? I'm into flashing too atm ;) Jul 28 15:24:31 pieterc: it will be online, yeah. Not in the official place though Jul 28 15:24:47 under /tests/mrmoku/unstable Jul 28 15:24:56 ( 3, Jul 28 15:24:56 'read', Jul 28 15:24:56 '+', Jul 28 15:24:58 wtf?! Jul 28 15:25:04 the number got is a fucking '+' Jul 28 15:25:05 ?! Jul 28 15:25:10 weird frameworkd :) Jul 28 15:25:18 mdbus -s org.freesmartphone.ogsmd /org/freesmartphone/GSM/Dev Jul 28 15:25:18 ice org.freesmartphone.GSM.SIM.RetrieveMessagebook all Jul 28 15:32:40 TAsn: you asked for my opinion for some foo the day before? Jul 28 15:34:36 aye. ;) Jul 28 15:34:43 npkg Jul 28 15:35:09 http://wiki.openmoko.org/wiki/Npkg Jul 28 15:35:39 though I'm kinda out. so please pm me/edit that page with your suggestions. Jul 28 15:35:51 ciao and thanks. Jul 28 16:36:19 TAsn: please split the UI from the backend :) Jul 28 16:36:29 TAsn: this is my first suggestion :) Jul 28 16:37:27 TAsn: my second one is: use simple gettext for error messages in UI (that s don't do funky error number computing like it was done in opkg :)) Jul 28 16:37:34 mrmoku: did you succeed ? Jul 28 16:37:48 mrmoku: may I read your tutorial to make some comments ?? Jul 28 16:51:26 ptitjes: was away too... still ongoing :) Jul 28 16:52:18 ptitjes: doing a small test app to see how to do it Jul 28 17:09:27 mrmoku: thanks for that Jul 28 17:09:44 mrmoku: I'm a little bit taken by Vala to make some non-generated C ;) Jul 28 17:11:04 ptitjes: how's current vala regarding the 'two mainloops' problem? Jul 28 17:11:20 I remember mickey_away wanted to add something for that... Jul 28 17:11:37 mrmoku: there is no problems about that IMHO Jul 28 17:12:17 mrmoku: both (if not E then GLib does) provide some mean to execute one iteration and to plug on external FD polls Jul 28 17:12:43 I don't know e's mainloop design Jul 28 17:12:50 but I really like the one from GLib Jul 28 17:13:10 it is really correctly designed and enough API is provided to open it to something else Jul 28 17:13:29 so it would be a glib mainloop that listens to e stuff too... Jul 28 17:13:40 anyway... that's the next step ;) Jul 28 17:14:05 first step is to use libfso-glib for the efl frontends Jul 28 17:14:19 and even before that I have to read the good night story :P Jul 28 17:14:20 bbiab Jul 28 17:18:31 cu Jul 28 17:19:03 mrmoku: I don't understand why none of you want to write that in Vala Jul 28 17:19:33 mrmoku: you'll lose time to write that in C Jul 28 17:40:05 ptitjes, please add to wiki... :) Jul 28 17:40:10 ptitjes: how do you compile vala? I'm new but interested Jul 28 17:40:16 and btw, don't worry about splitting UI from backend... :) Jul 28 17:40:44 I'll have an npkg_error function Jul 28 17:40:53 that can be set according to UI Jul 28 17:40:59 and act accordingly. Jul 28 17:41:26 (fprintf as std, anything else according to UI...) Jul 28 17:42:03 TAsn: write it in Vala, this could be a good exercise for you Jul 28 17:42:08 ptitjes, no. Jul 28 17:42:30 pieterc: get a release tarball, and build as usual Jul 28 17:42:58 i meant, how do you compile vala programming code Jul 28 17:43:26 pieterc: http://www.freesmartphone.org/sources there is vala-0.7.5-fso2 there Jul 28 17:43:31 pieterc: oh Jul 28 17:43:38 pieterc: sorry Jul 28 17:43:50 ptitjes: my bad Jul 28 17:43:55 pieterc: use valac Jul 28 17:44:03 is it easier? Jul 28 17:44:18 man valac is not up-to-date compared to valac --help Jul 28 17:44:43 pieterc: easier ? to code in Vala than C ? Jul 28 17:44:54 to compile Jul 28 17:45:16 to cross compile in particular Jul 28 17:45:18 pieterc: Vala is object-oriented, thus you can program GObjects without the burden of writing lots of C lines Jul 28 17:46:02 oh alright Jul 28 17:46:09 so to compile you do Jul 28 17:46:11 pieterc: Vala produces C code Jul 28 17:46:18 valac project Jul 28 17:46:22 and the bin comes out Jul 28 17:46:23 ? Jul 28 17:46:38 pieterc: then you can compile the C code with whatever C compiler, thus it is extra easy to cross-compile Jul 28 17:46:49 oh okay Jul 28 17:46:57 so it's just a layer above gcc Jul 28 17:47:06 pieterc: valac myfile.vala Jul 28 17:47:23 pieterc: add -C if you would like the C code, which is your case Jul 28 17:47:28 pieterc: exactely Jul 28 17:47:47 pieterc: thus I don't understand why TAsn want to write C code when he can write Vala code Jul 28 17:48:13 ptitjes: is Vala ready for glibless programming? Jul 28 17:48:14 which is clearer, have automated memory management (without garbage collector) Jul 28 17:48:33 PaulFertser: it has two profiles GObject and Posix Jul 28 17:48:35 ptitjes: also does Vala easily allow non-OO approach? Jul 28 17:48:44 ptitjes, me no like vala Jul 28 17:48:47 and now for real Jul 28 17:48:47 ptitjes: i know about the Posix profile. But is it already usable. Jul 28 17:48:47 PaulFertser: non-OO approach ? Jul 28 17:48:58 me no like TAsn Jul 28 17:49:00 ;p Jul 28 17:49:00 more ppl know C than vala Jul 28 17:49:01 therefore more potential devs Jul 28 17:49:04 ptitjes: yes, traditional procedural technique, not object oriented. Jul 28 17:49:05 I know better C than vala Jul 28 17:49:14 which means I won't make noobie mistakes Jul 28 17:49:24 yep I know lots of people that don't want to change their habits too Jul 28 17:49:25 :p Jul 28 17:49:41 PaulFertser: you have functions too Jul 28 17:49:42 this project is too important for allowing these two downfalls to harm it. Jul 28 17:49:44 TAsn, btw. did you test the image? Jul 28 17:49:55 mrmoku, didn't know there is one Jul 28 17:49:58 PaulFertser: but this gives no real interest Jul 28 17:49:58 sec, will d/l Jul 28 17:50:03 though I'm leaving soon Jul 28 17:50:07 and I need a working phone Jul 28 17:50:14 so I'll only test it in a couple of hours Jul 28 17:50:16 TAsn, do it on your own risk then ;) Jul 28 17:50:20 ptitjes: no doubt but what does vala give to a programmer that wants to write a non-glib non-OO program comparing to plain C? Jul 28 17:50:21 ahh Jul 28 17:50:26 mrmoku, i'll just wait Jul 28 17:50:27 TAsn: downfalls ?? you mean "TAsn" and "C" ?? Jul 28 17:50:29 :D Jul 28 17:50:34 have you tested it yourself? Jul 28 17:50:40 TAsn, no ;) Jul 28 17:50:43 PaulFertser, ++. ;) Jul 28 17:50:44 was away too Jul 28 17:50:45 mrmoku, shit. Jul 28 17:51:03 PaulFertser: I thought you were at F9N, when I talked about memory mangement ? Jul 28 17:51:09 ptitjes: i was :) Jul 28 17:51:18 PaulFertser: so you did not listen Jul 28 17:51:20 ptitjes: even asked questions iirc Jul 28 17:51:22 ptitjes: i did Jul 28 17:51:24 * ptitjes spanks PaulFertser Jul 28 17:51:28 ptitjes: i did! Jul 28 17:51:31 :p Jul 28 17:51:32 tztztzz Jul 28 17:51:34 ptitjes: it's just i forget easily Jul 28 17:51:52 PaulFertser: ownership transfer is part of APIs Jul 28 17:51:57 btw: WHERE ARE THE VIDEO-STREAMS??? Jul 28 17:52:11 VIDEOS VIDEOS VIDEOS VIDEOS VIDEOS VIDEOS VIDEOS VIDEOS VIDEOS VIDEOS VIDEOS VIDEOS Jul 28 17:52:11 ptitjes: yeah, yeah, the real decent argument to use Vala, i agree. Jul 28 17:52:15 PaulFertser: when you want that the caller of your function manages the memory you return to him, you say so Jul 28 17:52:25 ptitjes: _now_ i remember. Jul 28 17:52:26 and the vala compiler will take care of that for the caller Jul 28 17:52:43 TAsn: you see? That's a nice reason. Jul 28 17:53:22 thus TAsn will write bad code with segfaults because even as good programmer as he is, he WILL forgot some free or something like this Jul 28 17:53:46 et voilà Jul 28 17:54:33 TAsn: also you'll have to deal with strings and that is easier in vala too. Jul 28 17:54:33 free missing != segfault. thats mere leak Jul 28 17:55:11 DocScrutinizer: yep sorry, but too much free = segfault, right ? Jul 28 17:55:18 premature free() == really bad ;-) Jul 28 17:56:03 * dos1 knows something about that, when he was trying to get USSD working in libframeworkd-phonegui-efl :x Jul 28 17:56:17 PaulFertser, don't take is side. Jul 28 17:56:30 and I never leak! :) (hopefully) Jul 28 17:56:34 TAsn: he provided decent arguments. Now it's your turn. Jul 28 17:56:44 PaulFertser, I already provided mine Jul 28 17:57:09 I don't know Vala, many people don't know Vala -> thus, vala noobs will be deving this, and less people will be helping Jul 28 17:57:16 which is BAD Jul 28 17:57:24 this project is too important for me to allow this to happen. Jul 28 17:57:32 :) Jul 28 17:57:33 TAsn: Vala is easy enough esp. given you have such a qualified expert available any time to help you ;) Jul 28 17:57:47 PaulFertser, lol. Jul 28 17:58:00 TAsn: and once the skeleton is ready i'm sure any C programmer will be able to contribute. Jul 28 17:58:04 still, it's less than appealing (making people learn that) Jul 28 17:58:21 PaulFertser, I don't agree. Jul 28 17:58:30 I wouldn't even consider contributing to a vala project Jul 28 17:58:34 TAsn: i'm a C programmer but i can read Python. If you write something in Vala in a straightforward way everybody'll be able to understand that. Jul 28 17:58:36 and I'm a C programmer ;) Jul 28 17:58:49 PaulFertser, I can read a lot of langs Jul 28 17:59:07 though still Jul 28 17:59:07 not being an expert in an env is bad. Jul 28 17:59:18 isn't vala kind of C plus pp-macros ;-D Jul 28 17:59:35 PaulFertser: thanks for the kind comment Jul 28 17:59:38 let's write it in python! Jul 28 17:59:39 really I appreciate Jul 28 17:59:41 :D Jul 28 17:59:41 :p Jul 28 17:59:52 TAsn: i'm not an expert in anything. Am i bad? Jul 28 18:00:13 TAsn: Vala's syntax is like C Jul 28 18:00:30 TAsn: the only thing you'll have to learn is a bit of object orientation Jul 28 18:00:38 ptitjes, I know. Jul 28 18:00:38 though it's not entirely the same. Jul 28 18:00:38 dos1, +++ :) Jul 28 18:00:45 DocScrutinizer, exactly. Jul 28 18:00:51 hehe, c++, I knew it Jul 28 18:00:53 ptitjes, I know OO Jul 28 18:00:57 TAsn: but if you know how to design correctly procedural programs, then you will feel that natural Jul 28 18:01:09 then your arguments are dumb Jul 28 18:01:18 I used to be a C++ programmer quite some time ago... Jul 28 18:01:21 or boost??? Jul 28 18:01:40 and also (although I'm denying it) Jul 28 18:01:40 a java programmer Jul 28 18:01:43 ptitjes, no they are not. Jul 28 18:01:52 which is "the next c++", at least according to it's developers Jul 28 18:02:06 DocScrutinizer, the next crappy lang? Jul 28 18:02:18 see libboost Jul 28 18:02:43 DocScrutinizer, I know boost. Jul 28 18:02:50 I hate it Jul 28 18:03:06 TAsn: what about compiling python code with Cython or Pyrex? :DDD Jul 28 18:03:12 ;) Jul 28 18:03:15 dos1, ++! :) Jul 28 18:03:39 What about compilig java code with gcj? Jul 28 18:03:47 :P Jul 28 18:03:53 TAsn: i was reading today about cython to find, if i can compile shr-settings to something faster Jul 28 18:03:57 I'm into prog-business for 35Y now, and every 6months they tell me another lang is inevitable next evolutionary step Jul 28 18:04:07 Heinervdm, TAsn: write it in prolog. end of discussion :D Jul 28 18:04:39 dos1: i wil never again write a program in prolog! Jul 28 18:04:48 dos1, you have great ideas :) Jul 28 18:04:55 DocScrutinizer, same as SCM Jul 28 18:05:00 though about SCM they were right Jul 28 18:05:00 hmm... but if you want great amount of developers Jul 28 18:05:05 git is great! :) Jul 28 18:05:08 what about logo? basic? or pascal? Jul 28 18:05:11 Heinervdm: you'll like Flora Jul 28 18:05:15 dos1, I'll have to stick with bash scripting ... Jul 28 18:05:23 ;) Jul 28 18:05:25 I considered pascal Jul 28 18:05:38 TAsn: you want to know why SHR apps evolve so slowly, because C bloats the code Jul 28 18:05:44 I still love modula Jul 28 18:05:46 pascal and delphi were my first programming languages :x Jul 28 18:05:53 ptitjes, I once said the same Jul 28 18:05:56 though Ainulindale proved me wrong Jul 28 18:05:56 ptitjes: when will we switch from stupid oeventsd to something that every prolog geek will adore? ;) Jul 28 18:06:03 TAsn: if you like python, then you get the ease of python with the power of C Jul 28 18:06:13 ptitjes, I don't really like python that much Jul 28 18:06:15 I like c better ;) Jul 28 18:06:19 TAsn: Ainulindale can make you eat snakes Jul 28 18:06:34 ptitjes, yes, the force is strong with this one. Jul 28 18:06:59 TAsn: though you don't want to, right ?? Jul 28 18:07:05 (eat snakes) Jul 28 18:07:11 yes I do Jul 28 18:07:13 if not then code in Vala Jul 28 18:07:23 TAsn: i have great new idea Jul 28 18:07:30 TAsn: write it in PHP! Jul 28 18:07:39 or, for that behalf : s/modula/delphi/ Jul 28 18:07:40 there are lots of php coders! Jul 28 18:07:41 :D Jul 28 18:07:46 ptitjes, but I do. Jul 28 18:07:52 dos1, ! Jul 28 18:07:54 PaulFertser: when some people will join the crew of Vala coders Jul 28 18:07:57 dos1: thats a nice idea, i've written scripts in PHP for over 10 years Jul 28 18:08:03 I have a friend who wrote PHP irc services Jul 28 18:08:03 :) Jul 28 18:08:06 so why can't i? Jul 28 18:08:07 PaulFertser: then we will be able to program correctly SHR-apps Jul 28 18:08:20 PaulFertser: then I'll have some free time to make really interesting things Jul 28 18:08:31 PaulFertser: among which is ologicd Jul 28 18:08:34 ptitjes: i want to learn vala when cornucopia will be more ready Jul 28 18:08:36 DocScrutinizer: i have a real-life delphi experience. It really sucks :S Jul 28 18:08:57 It never got mature, I agree Jul 28 18:09:02 TAsn: I wrote an inference engine in PHP, in 2000 Jul 28 18:09:16 TAsn: though I prefer to write Vala, because it is better Jul 28 18:09:18 :p Jul 28 18:09:29 dos1: I'll be there for you Jul 28 18:09:37 DocScrutinizer: it never could imho, bad language with bad useless "RAD" approach to design UIs. Jul 28 18:09:44 dos1: as I promised some weeks ago Jul 28 18:09:49 i loved delphi, but now I couldn't write anything bigger in it Jul 28 18:10:08 ptitjes: ...you reminded me... Jul 28 18:10:13 ptitjes: i start to think that most of freerunner UI should be written as an emacs mode. Jul 28 18:10:14 ptitjes: ...what about ophonekitd-vala? Jul 28 18:10:25 WHERE ARE THE F9N VIDEO-STREAMS??? Jul 28 18:10:35 VIDEOS VIDEOS VIDEOS VIDEOS VIDEOS VIDEOS VIDEOS VIDEOS VIDEOS VIDEOS VIDEOS VIDEOS Jul 28 18:10:36 PaulFertser: you have ncurses API already in Vala Jul 28 18:10:57 PaulFertser: but if you prefer emacs and that there is some C API, then you can use it in Vala Jul 28 18:11:21 TAsn, with C++ you could count me in as contributor ;) Jul 28 18:11:36 mrmoku, nah. Jul 28 18:11:40 * TAsn hates C++. Jul 28 18:11:42 that's why I said Jul 28 18:11:45 TAsn: and btw, if you program it in Vala, nothing prevents anybody to write plugins/extensions in C Jul 28 18:11:48 I *used* to be a c++ dev Jul 28 18:11:49 :) Jul 28 18:11:49 * dos1 hates everything Jul 28 18:11:59 TAsn: look at what mickey_away do, he mixes C and Vala Jul 28 18:12:08 ptitjes, and if I write in C nothing prevents ppl to write plugins/etxensions in vala Jul 28 18:12:20 ... Jul 28 18:12:20 :) Jul 28 18:12:20 mrmoku: you told you want to learn Vala, go on Jul 28 18:12:30 :) Jul 28 18:12:31 TAsn: for sure Jul 28 18:12:39 TAsn: but you will loose your time Jul 28 18:12:53 0002 mok@mrdenker[pts/5]:~/src/other/openmoko/framework-> git pull Jul 28 18:12:53 git.freesmartphone.org[0: 134.169.172.109]: errno=No route to host Jul 28 18:12:53 fatal: unable to connect a socket (No route to host) Jul 28 18:12:55 damn Jul 28 18:13:00 * dos1 wonders, if in SHR repo is some build-essential package Jul 28 18:13:04 * dos1 checks, if in SHR repo is some build-essential package Jul 28 18:13:07 dos1: as for opkd-vala, I'm not really much on it Jul 28 18:13:37 ptitjes: no, you don't understand. I think that everything needed can be written in elisp with dbus calls to frameworkd. Jul 28 18:14:21 ok moving to dinner Jul 28 18:14:24 * dos1 wonders, why in SHR repo build-essential package is not present Jul 28 18:14:26 ;P Jul 28 18:14:45 PaulFertser: I like the approach, but is it good for embedded ? Jul 28 18:15:21 ptitjes: i tried emacs on my freerunner and it seemed to have the most potential to provide fast, consistent and comprehensible UI. Jul 28 18:18:37 PaulFertser: I'm still wondering if you're kidding or try to start a real project ;D Jul 28 18:19:03 PaulFertser: are there any gta01 specific drivers in andy-tracking which are not upstream? Jul 28 18:19:13 mrmoku, I can't seem to find the image Jul 28 18:19:15 where are they at? Jul 28 18:19:58 DocScrutinizer: johnsu01's already started it. I hope to understand enough of the autofoo soon to properly crosscompile emacs23. I think it can get a good momentum. I'm ready to write some elisp myself. Jul 28 18:20:25 larsc: i'm afraid yes. pcf50606, gta01_battery ;) Jul 28 18:20:35 TAsn, tests/mrmoku/unstable Jul 28 18:20:42 larsc: ASoC driver Jul 28 18:21:11 pcf50606 is not upstream? Jul 28 18:21:18 ASoC eeek Jul 28 18:22:52 larsc: looking at mfd kernel tree atm, yes, it's not there. Jul 28 18:23:22 oh ok Jul 28 18:23:25 DocScrutinizer: i can make gta02 driver support gta01 too. But there won't be 1-1 statefile compatibility nevertheless. Jul 28 18:23:48 understood and supported here Jul 28 18:24:02 oh ;) Jul 28 18:24:05 larsc: have you seen my pcf50633 patches already? Jul 28 18:25:18 PaulFertser: gta01 has a few additional (useless) controls for amp Jul 28 18:25:25 DocScrutinizer: i know Jul 28 18:25:50 if those are nonfunct for 02, who cares Jul 28 18:27:23 DocScrutinizer: they won't be compiled for gta02 anyway. Jul 28 18:27:33 even better Jul 28 18:27:56 unify the code, but keep the individual details Jul 28 18:28:30 though we should try to sync control-names and numbers wherever possible Jul 28 18:28:31 PaulFertser: yep Jul 28 18:28:33 Sure, i can do it. Unfortunately i have other stuff to do too. Jul 28 18:37:39 larsc: should i direct patches to MFD maintainer and Cc linux-kernel@vger.kernel.org ? I can't see a dedicated MFD ML. Jul 28 18:44:47 PaulFertser: That's how I would to it. Jul 28 18:45:20 larsc: ok, thanks Jul 28 18:53:25 larsc: we'll have problems merging bq27000 driver upstream, there's already some (incomplete) solution there.. Jul 28 18:53:36 Saying that SHR apps are bad because they are written in C and they'd be better in vala is as stupid as saying "e ? It's crap, it's C!" Jul 28 18:53:56 Just plain stubborness, out of the blue crap, based on nothing technical Jul 28 18:54:29 'nuff said Jul 28 18:56:11 (and as much as I like functional languages, and as much as I would like to conquer the world with them, they are not appropriate for everything just because I think so) Jul 28 18:56:22 "You claim that emacs sh*ts itself when it gets EAGAIN, and you think Jul 28 18:56:22 that's an emacs bug. And I think you're full of crap. We should NEVER EVER Jul 28 18:56:25 get EAGAIN (due to the SIGCHLD, at least) if the app on the other side Jul 28 18:56:27 wrote data that could be read." Jul 28 18:56:31 (recent message from Linus to Alan Cox) Jul 28 18:56:32 (there are industrial standards, best practices, and developers who know this or that language) Jul 28 18:57:33 PaulFertser: bloody russian! :-) Jul 28 18:58:16 PaulFertser: how are you by the way? :-) Jul 28 18:58:40 dos1: by the way Jul 28 18:58:45 judging from your jabber subtitle Jul 28 18:58:48 You're a utter geek Jul 28 18:58:57 But you have good tastes for TV shows Jul 28 18:59:12 Although you should develop rathen than slack around :-) Jul 28 18:59:54 Ainulindale: i'm ok, thanks. :) hope you have fun there. Jul 28 19:00:31 Not at all Jul 28 19:00:40 I ran from work to Bearstech to meet with mickey_away Jul 28 19:00:45 Eh? Jul 28 19:00:53 A whole day of meeting and reviewing documents Jul 28 19:02:18 And I can't find a single second to help mrmoku Jul 28 19:02:21 And that drives me nuts Jul 28 19:02:43 PaulFertser: bq27x00? Jul 28 19:02:48 larsc: yes Jul 28 19:03:00 Ainulindale: oh, documents. That sucks. Jul 28 19:04:26 PaulFertser: That's for sure Jul 28 19:04:34 PaulFertser: yeay, saw that, I guess we should be able to replace it with our implementation. Jul 28 19:05:14 larsc: :) we should prove that our implementation is better and doesn't break things. Probably we can try to start convincing folks using it right now. :) Jul 28 19:06:39 Also we'd need to check i2c support etc. Jul 28 19:08:55 My point is that most of the actual problems are because of bad error management and memory management Jul 28 19:09:28 Both of these can easily be addressed by programming a level up Jul 28 19:10:01 The whole advantage of code generation is to not repeat common pitfalls Jul 28 19:10:08 Car pollution could be solved by either walking, not taking car, or inventing a new clean energy Jul 28 19:10:28 C is a very good language but it is error-prone as soon as many developers put their hands in the code Jul 28 19:10:32 But we don't do any of that because cars have a purpose which isn't filled by anything else Jul 28 19:10:34 That's all Jul 28 19:10:57 Your reasoning is close minded as ever as long as you disagree with the basis, and "that" is all :-) Jul 28 19:10:59 Ainulindale: that is not a decent argument Jul 28 19:11:04 ptitjes: nor is yours Jul 28 19:11:18 Ainulindale: we will be forced to not taking cars at a point and you are consious of that Jul 28 19:11:42 :-) Jul 28 19:11:44 That said, I leave Jul 28 19:11:47 cu all Jul 28 19:12:34 ptitjes: see you Jul 28 19:13:10 ptitjes, before you leave... Jul 28 19:13:38 how to use for example FREE_SMARTPHONE_PIM_SOURCES Jul 28 19:14:04 it wants a FreeSmartphonePIMSources object or a pointer? Jul 28 19:14:27 just instanciate a dbus proxy with the helper function provided at the end of the .h Jul 28 19:14:56 ahh :) Jul 28 19:20:20 do you see the prototype Jul 28 19:20:40 mrmoku: something like free_smarphone_get_object Jul 28 19:20:45 or something like that Jul 28 19:21:40 Ainulindale: hehe :D when i saw that i automaticly paused to analize, what i saw :D Jul 28 19:22:08 but i'm not sure it's KDE3, cause focus is only on left side of screen Jul 28 19:22:42 :-) Jul 28 19:22:47 but i think it's transparent kicker on bottom of screen :D Jul 28 19:23:07 and i'm sure oowriter uses kde3 icons ;) Jul 28 19:23:59 Alan Cox is dropping maintaining of tty layer. Wow Jul 28 19:24:12 <[Rui]> well, my gcc compile problems must've been some fluke, bitbake -c clean gcc and then it was all ok Jul 28 19:24:28 <[Rui]> but after building elmdentica... man libelm apps look ugly in shr-u :( Jul 28 19:25:38 <[Rui]> PaulFertser: perhaps (I'd guess) because it's so stable? :D Jul 28 19:26:22 [Rui]: no, because Linus says Alan is full of crap: http://lkml.org/lkml/2009/7/28/375 Jul 28 19:26:41 <[Rui]> shudder Jul 28 19:27:36 ptitjes, yep, see it and playing with it :) Jul 28 19:27:46 <[Rui]> *shudder* Jul 28 19:31:29 ptitjes, and for the GType parameter I use a define like FREE_SMARTPHONE_PIM_TYPE_SOURCES? Jul 28 19:33:17 ptitjes, there is no _non_ vala sample code somewhere? :P Jul 28 19:35:01 <[Rui]> PaulFertser: what's worst is an apparent lack of patience at each other.... Jul 28 19:35:07 ptitjes, and I have to cast the return value (GObject *) to something like FreeSmartphonePIMSources * ? Jul 28 19:35:29 [Rui]: a week is long enough... Jul 28 19:37:11 PaulFertser, depends on the problem ;) Jul 28 19:39:39 <[Rui]> PaulFertser: I think he was quite aggressive from the start of the thread Jul 28 19:40:03 [Rui], guess Alan should have get used to it by now :P Jul 28 19:40:09 s/get/got/ Jul 28 19:40:11 mrmoku meant: [Rui], guess Alan should have got used to it by now :P Jul 28 19:40:11 <[Rui]> PaulFertser: in a snow ball kind of way, things ended up with Alan giving up tty maintenance :( Jul 28 19:40:32 ptitjes, works, great :) Jul 28 19:40:55 <[Rui]> yeah, but people don't always have their best days out of bed, and I'm betting they both had the "period" on the same day :) Jul 28 19:40:56 now to test the async variant... Jul 28 19:41:04 :) Jul 28 20:04:16 * dos1 started rewrite of SHR Settings launcher Jul 28 20:04:25 i already have licence :D Jul 28 20:05:24 that means? Jul 28 20:05:52 i dont see a launcher in the shr settings app Jul 28 20:06:40 it would be nice to have it like in the preferences of e Jul 28 20:07:23 ptitjes, does it use GAsyncReadyCallback as GIO does? Jul 28 20:07:59 * mrmoku thinks so... as he said in GIO tradition :P Jul 28 20:08:56 very nice screenshot http://scap.linuxtogo.org/files/1d8b8ecadc17fd3a6fee18843633760f.png Jul 28 20:09:04 BluesLee: run shr-settings Jul 28 20:09:07 BluesLee: that's launcher Jul 28 20:09:09 BluesLee: :P Jul 28 20:09:14 BluesLee: launcher of shr-settings modules Jul 28 20:09:18 okay Jul 28 20:09:51 i found it:-) Jul 28 20:10:37 is this shr? http://scap.linuxtogo.org/files/b9bd3d53357f5dabe3d45f9a58fac6b8.png Jul 28 20:12:26 our new rotating 3D desktop cube ;) Jul 28 20:13:05 hehe Jul 28 20:13:29 that's some Qt4 app Jul 28 20:13:53 and that 3D effect is from qt examples :D Jul 28 20:14:21 are you sure? it looks good, i wonder if its fast enoug Jul 28 20:14:23 ^h Jul 28 20:14:32 i'm sure Jul 28 20:14:49 that calendar looks exactly like Qt's one Jul 28 20:15:01 yes Jul 28 20:15:05 only in white Jul 28 20:15:09 and that interface is doable really easy in qt Jul 28 20:15:20 BluesLee: hmm? Qt's is white ;x Jul 28 20:15:47 qt has few really nice things Jul 28 20:15:48 the one i know was white on black Jul 28 20:15:58 BluesLee: that's KDE's Jul 28 20:16:20 but even KDE 4.0 was using pure Qt calendar, and it was white ;) Jul 28 20:16:37 no, the one from qtopia/qtextended or whatever ... white on black Jul 28 20:16:38 ok, eeeeaaaatiiiiingggg Jul 28 20:16:52 BluesLee: oh. qtopia has different style Jul 28 20:16:52 okay, i won:-) Jul 28 20:17:03 BluesLee: and qt is stylable with css :P Jul 28 20:17:24 i dont care i want it in shr Jul 28 20:17:30 but on that screen is how pure qt calendar without any external style looks Jul 28 20:17:31 can you include it? Jul 28 20:17:45 haha Jul 28 20:17:48 he has gone Jul 28 20:19:15 and him too ;) Jul 28 20:19:23 ptitjes, ping :) Jul 28 20:22:17 ptitjes, undefined reference to `free_smartphone_pim_sources_get_entry_count_async' Jul 28 20:22:25 means my libfso-glib is without async? Jul 28 20:23:26 uhh... yeah Jul 28 20:23:36 * mrmoku sets vala-dbus-binding-tool-native back to AUTOREV :P Jul 28 20:25:23 uuhhh latest scap screenshots look good... Jul 28 20:25:30 does anyone know what is it? Jul 28 20:26:46 F4t: we need a scap with the ability to post comments :) Jul 28 20:29:44 Zorkman, it can be done if u can manage to edite the "model" field.... Jul 28 20:29:53 im pretty sure its free test Jul 28 20:29:53 test=text* Jul 28 20:30:11 F4t: I know, i can do it Jul 28 20:32:30 someone needs to change gpe-scap to request a comment :) Jul 28 20:37:41 F4t: look at latest scap Jul 28 20:38:31 last two images to be more precise Jul 28 20:48:59 Zorkman, F4t: guys, just use brain! Jul 28 20:49:01 it's somethin on Qt4 Jul 28 20:49:05 it's clearly visible Jul 28 20:49:12 there was discussion about Qt4 recently Jul 28 20:49:16 in LitePhone thread Jul 28 20:49:57 i'm pretty sure that's Qalee Jul 28 20:50:33 yes, i was right Jul 28 20:50:35 http://www.openmoko.openmatica.es/QALEE/ Jul 28 20:50:36 :P Jul 28 20:52:18 wasn't better searching for that instead of abusing scap? Jul 28 20:52:23 ;x Jul 28 20:52:51 dos1: haven't been able to keep up with mails lately Jul 28 20:53:08 so don't know how i would be able to find it Jul 28 20:53:51 but that isn't reason for abusing scap with too long model name without model name in it :P Jul 28 20:54:07 oh, there was model name... but with long comment, which broke layout :P Jul 28 20:54:24 hello All! Jul 28 20:55:04 please tell me how to install the xterm on neo? Jul 28 20:55:12 opkg install? :) Jul 28 20:55:44 Qalee looks great! Jul 28 20:55:59 it wrote not found Jul 28 20:56:09 fruct: which distro? Jul 28 20:56:33 dos1 SHR Jul 28 20:56:41 fruct: oh, nice Jul 28 20:57:04 i thought you will say "om2008", and then i would be angry ;D Jul 28 20:57:13 mrmoku: ping Jul 28 20:57:18 mrmoku: add xterm to feed :) Jul 28 20:57:42 dos1: but xterm nothing Jul 28 20:58:07 fruct: you're not native speaker, am i right? :) Jul 28 20:58:09 ;> Jul 28 20:58:18 (i'm not too... don't worry :D) Jul 28 20:58:21 I'm yes I'm from Russia Jul 28 20:58:43 i'm from Poland ;) Jul 28 20:58:57 and I'm somewhere between you) Jul 28 20:58:58 fruct: when we'll add xterm to SHR feed, you'll be able to install it by just "opkg install xterm" Jul 28 21:00:09 ...and noone knows, where exactly... ;) Jul 28 21:00:32 dos1: I see some people on TangoGPS friends Jul 28 21:01:05 In europe many peoples have the neo =) Jul 28 21:01:44 dos1: opkg install xterm - wrote that nothing package =| Jul 28 21:02:11 fruct: cause it isn't now ;P Jul 28 21:02:24 fruct: you'll <- future tense :D Jul 28 21:03:05 it possible to copy from other distributives? Jul 28 21:03:36 or other repository or compile it? Jul 28 21:04:59 fruct: by tomorrow "opkg install xterm" will probably work (if they have added it to the feed) Jul 28 21:05:07 fruct, nah, wait I will build it for you :) Jul 28 21:07:03 Zorkman: you SHR developer? Jul 28 21:07:40 fruct: no, not me, but mrmoku is, and he will add it Jul 28 21:08:22 fruct, building it right now Jul 28 21:08:41 thanks! =) Jul 28 21:12:04 I have start x11vnc by the "xterm -e x11vnc" and don't do it =( Jul 28 21:12:20 fruct, opkg update; opkg install xterm Jul 28 21:20:06 ptitjes, yo, managed to do an async libfso-glib call :D Jul 28 21:32:37 where power config in the SHR? after reboot dont saved the settings-power-Auto-dimmig Jul 28 21:33:15 fruct: reboot doesn't save that Jul 28 21:33:51 I can't set it always OFF? Jul 28 21:36:07 and why opkg have >70% of CPY time? Jul 28 21:36:31 CPU? Jul 28 21:36:42 decompressing is hard word Jul 28 21:36:44 work Jul 28 21:36:44 oh yes cpu! Jul 28 21:37:18 fruct, flooding memory with useless stuff is also hard work Jul 28 21:37:54 thanks for answear! good bye all! Jul 28 21:48:59 yeah, time to say good bye :) Jul 28 21:49:03 gnight all Jul 28 21:49:54 gnight mrmoku Jul 28 22:57:22 * Weiss reports the first non-blank screen with the KMS-based DDX Jul 28 22:57:27 (though this isn't very impressive) Jul 28 23:00:23 not impressive for users, but impressive for some developers) Jul 28 23:00:46 * max_posedon will be lucky have DRI2 + KMS on Neo Jul 28 23:01:07 s/lucky/happy Jul 28 23:02:05 i think the only piece of the puzzle left is some memory mapping voodoo in the kernel Jul 28 23:36:09 mrmoku|away: yeah!! :D Jul 28 23:36:23 mrmoku|away: sorry to come back so late Jul 28 23:36:32 mrmoku|away: does it work as expected ?? Jul 28 23:37:37 mrmoku|away: what was the undefined ??? non-installed rebuild lib ? Jul 28 23:38:54 Weiss:are you working on dri2 & neo? Jul 28 23:39:25 admiral0: yep Jul 28 23:39:35 niiiiiiiiiiiiiiice Jul 28 23:39:57 so how's it going Jul 28 23:39:59 ? Jul 28 23:41:05 i hope to have the userspace bits of KMS working very soon, then there'll be a bit of stabilisation to make it as stable (hopefully a lot more stable) than the fbdev-based driver Jul 28 23:41:17 then it's' on to some fun stuff using acceleration Jul 28 23:44:51 acceleration? would it be able to squeeze more performance from the glamo (and the bottleneck we all know) ? Jul 28 23:45:50 uhhhhhhhhhhhhhh libmodulo can build on vala master Jul 28 23:46:24 i think we can get a lot more out of it - we're barely using any of its power at all. however, it remains to be seen how well it works in practice.. Jul 28 23:46:53 i hope for a faster freerunner Jul 28 23:47:20 :) Jul 28 23:48:00 one thing it'll definitely do is to allow MPEG acceleration and 2D acceleration in Xorg to work together (currently they compete against one another, and by some miracle don't detonate the whole system..) Jul 28 23:48:13 lol Jul 28 23:48:35 what's the stats? Jul 28 23:48:47 s/stats/status/ Jul 28 23:48:47 admiral0 meant: what's the status? Jul 28 23:52:08 the kernel-level KMS stuff works (that's mostly cutting+pasting the contents of glamo-fb.c into a new framework), DRM bits exist for memory management and command queue submission, and right now i'm nearly there with the userspace bits of KMS (to allow Xorg to use it to control the display) Jul 28 23:53:18 one more question:that means that our current 2d acceleration is not the max we can achieve? Jul 28 23:55:06 i think we can make the user experience a lot faster if we're prepared to accept some measure of platform-specific code Jul 28 23:56:07 whyshouldn't we? Jul 28 23:56:40 faster freerunner = happy users Jul 28 23:57:21 well... i'm perfectly willing to write lots of Glamo-specific code because this is an interesting project for me (and potentially leads to further interesting projects on other hardware), even in the knowledge that FR won't be around forever (and nothing else will use Glamo) Jul 28 23:59:10 but not everyone is willing to do that.. e.g. raster is very focussed on general, platform-independent code Jul 28 23:59:56 i think x11 should be modified a little Jul 29 00:00:36 let's be reasonable who needs X11 networking on freerunner? Jul 29 00:01:18 some glamo code in there would be fine Jul 29 00:02:00 X11 isn't really the problem.. more that the high-level stuff doesn't give the low-level stuff much chance to accelerate Jul 29 00:02:27 for example? Jul 29 00:02:41 (could be useful for me :D) Jul 29 00:03:29 our hardware works best when you can form the screen image by compositing smaller images together which individually don't change. say, you could have a menu of icons which pops up, and have the icons stay in VRAM for most of the time, just telling the GPU to draw them in different places Jul 29 00:03:53 at the moment, many such things are entirely re-drawn from scratch each time Jul 29 00:05:06 isn't it the toolkit's fault? Jul 29 00:05:56 exactly Jul 29 00:07:05 that's bad Jul 29 00:07:36 for us, it is, at least Jul 29 00:07:47 not only Jul 29 00:08:12 on desktops too, even if we don't manage to see it :P Jul 29 00:08:42 for other hardware, the overheads involved with accelerating such things outweigh the gains. in our case, the bus to the VRAM is so immensely slow that we have to keep things in VRAM as much as possible Jul 29 00:10:02 ok.. understood what coding style i need for progs on freerunner Jul 29 00:10:33 keep ui s in VRAM a Jul 29 00:11:03 i see the most benefit in writing programs which go lower level than the toolkit, almost driving the hardware directly - that's where DRI comes in Jul 29 00:12:42 that's time consuming Jul 29 00:13:05 it would be better to modify a toolkit Jul 29 00:15:27 perhaps.. but i suspect that at the toolkit level you're too abstracted to know what's really important Jul 29 00:17:37 i'm still learning.. maybe there'll be no overall benefit, or a very small one (some new discoveries about the memory clocks by larsc are slightly worrying in this regard) Jul 29 00:17:41 anyway.. i need to sleep Jul 29 00:21:07 me too Jul 29 00:21:12 good night Jul 29 00:21:40 (i've managed to package netcfg yay! ) **** ENDING LOGGING AT Wed Jul 29 02:59:57 2009