**** BEGIN LOGGING AT Thu Dec 25 02:59:57 2008 Dec 25 03:33:57 ok, got ~40pcs 100uF (tantal btw), and ~30 2k2 (plus 30pcs 1k8, ND 30 2K7). But we are heavily under duty to handcraft a soldertip for my iron, which actually can do 0402 soldering. Dec 25 03:34:29 just to keep joyful visitors of 25C3 informed Dec 25 03:35:49 btw: UNsldering is a snap by using a pair of SMT-tweezers Dec 25 03:36:34 (already made the custom tips for this one, to unsolder 0402 ;) Dec 25 03:38:17 nota bene: still it's no promise to do a single rework on any od the 3 days 25C3 Dec 25 03:38:30 s/od/of/ Dec 25 03:38:30 DocScrutinizer meant: nota bene: still it's no promise to do a single rework on any of the 3 days 25C3 Dec 25 09:21:41 Merry Christmas!!!! Dec 25 09:27:41 happy christmas ajuonline Dec 25 09:29:55 :) Dec 25 09:59:08 Merry Christmas Dec 25 10:00:05 what is the problem with sim authentification, it takes longer than the booting procedure... Dec 25 10:34:16 BluesLee: sim auth can only occur after dbus is up and then when sim is up Dec 25 10:34:22 this can take a long time Dec 25 10:34:44 why? Dec 25 10:35:21 it takes 2 secs on a mobile Dec 25 10:35:51 same in qtextended Dec 25 10:36:26 maybe dbus should be started with the booting process Dec 25 10:37:30 so the booting itself takes longer but the sim auth itself is shortened Dec 25 10:38:48 it should simply be done in the background Dec 25 10:38:58 ui should come up asap Dec 25 10:39:04 and be as usable as possible Dec 25 10:39:28 as backend services start - features start to work/get enabled Dec 25 10:39:35 and the user should have some indication of this Dec 25 10:40:20 are you sure? mainly i think of this device as a phone, so the first thing i do is waiting for sim auth .. Dec 25 10:40:43 you will wait - one way or another Dec 25 10:40:52 i know Dec 25 10:41:11 why make OTHER uses take longer just because of 1 usage scenario Dec 25 10:41:20 what if i'm on a plane Dec 25 10:41:21 the overall boot time is very good now, i am impressed Dec 25 10:41:25 and i dont care about my sim Dec 25 10:41:35 i actuallyt dont even want the antennae to be powered up Dec 25 10:41:39 raster, what do you mean? Dec 25 10:41:40 i want to turn on and play a game Dec 25 10:41:43 or watch amovie Dec 25 10:41:45 play music Dec 25 10:41:45 ... Dec 25 10:41:47 sure Dec 25 10:41:56 raster, Uhm, turn on your phone. Dec 25 10:41:57 but i dont live in a plan;-) Dec 25 10:42:01 raster, go ahead. Dec 25 10:42:12 99.9% i am on planet earth, hehe Dec 25 10:42:17 Merry Christmas! Dec 25 10:42:20 raster, what does it matter Dec 25 10:42:21 Sargun: which one? Dec 25 10:42:24 admiral0, you in hawaii Dec 25 10:42:41 no Italy XD Dec 25 10:42:48 oooh klined Dec 25 10:43:12 raster: i am getting familiar with edje http://scap.linuxtogo.org/files/38d7a90d8d0370a078e7a051c99341df.png Dec 25 10:44:23 BluesLee: aaah ... not too bad Dec 25 10:44:27 ah that was you? Dec 25 10:44:31 nice Dec 25 10:44:39 thanx Dec 25 10:45:22 one of the better efforts i've seen Dec 25 10:45:25 raster, i am doing a mockup for the launcher Dec 25 10:45:30 though nothng "amazing" Dec 25 10:45:34 i have to understand more of the logic of edje files, than i will remove the > (application to the right) Dec 25 10:45:36 but loosk decent Dec 25 10:45:43 not offensive :) Dec 25 10:46:10 the guy on the lower right is not offensive?;-) Dec 25 10:46:19 hahahahahah Dec 25 10:47:10 i a playing a little bit but i have more ideas, i dont know what is possible at all Dec 25 10:48:39 raster: is there a desktop file for illume itself? or where does the theme get "Illume" from? Dec 25 10:52:24 admiral0: a mockup for the application launcher? Dec 25 10:52:37 yes Dec 25 10:53:05 i dont use it all, one moment i will test the default one Dec 25 10:53:14 i made some brainstorming with a friend of mine, and now i have several ideas Dec 25 10:54:05 okay, it slides on the bottom Dec 25 10:54:26 let's be honest.. the slider sucks and the icons are boring.. Dec 25 10:55:09 then on SHR icons look just like on the iphone... Dec 25 10:55:26 the default ones? Dec 25 10:55:44 i used black and white icons from gnome Dec 25 10:56:01 but i also tested others with lesser success Dec 25 10:56:58 i would like to put oxygen icons on my distro Dec 25 10:57:16 (but i haven't compiled X yet) Dec 25 10:57:37 oygen? one moment Dec 25 10:59:51 those ones http://scap.linuxtogo.org/files/efdd39c5a0f32c1a76fcbfe549f74f9f.png Dec 25 11:00:54 ? Dec 25 11:01:34 they look alike, but oxygen icons as i know them are not b/w Dec 25 11:01:51 sorry Dec 25 11:01:54 wrong link Dec 25 11:02:06 http://www.kde-look.org/content/preview.php?preview=3&id=79046&file1=79046-1.jpg&file2=79046-2.jpg&file3=79046-3.jpg&name=Oxygen+Icon+Theme+for+KDE3&PHPSESSID=13818c6d62449c8839b74ffc1e76defe Dec 25 11:03:28 exactly Dec 25 11:03:44 okay Dec 25 11:03:53 nice aren't they? Dec 25 11:04:51 they are good yes Dec 25 11:05:10 and what kind of wallpaper do you want to choose? Dec 25 11:09:21 don't know yet Dec 25 11:09:34 i chose an iceberg for now Dec 25 11:10:18 i'm going to lunch.. see you Dec 25 11:10:28 cu Dec 25 12:05:08 <\marco> hi to all Dec 25 12:05:28 <\marco> I've problem with start-stop-daemon and fso-gpsd Dec 25 12:06:09 <\marco> start-stop-daemon create a pid file for the process with a wrong pid Dec 25 12:06:56 <\marco> if the pid is 1245 it writes 1246 Dec 25 12:06:58 <\marco> :/ Dec 25 12:07:54 <\marco> setting manually the right pid.. it works.. Dec 25 12:20:11 A) That's an openmoko, not -cdevel thing Dec 25 12:20:24 B) stay here for more than 2 fucking minutes you idiot. Dec 25 12:26:25 openmoko: Who are you? Dec 25 12:28:11 PaulFertser: i read that you fixed the buzz and also echo, did you used http://lists.openmoko.org/pipermail/community/2008-November/036068.html for the echo supression? Dec 25 12:29:21 PaulFertser: why do you ask? Dec 25 12:29:40 BluesLee: Unfortunately i did neither. :( In fact, i never experienced a buzz. And you know exactly (as we all know) that the buzz fix is purely a hardware issue, that is addressed in a Joerg's SOP. Dec 25 12:30:28 openmoko: I want to have a link between irc nicks, commit logs and mailing lists. I'm not a big fun of 2ch-like things. Dec 25 12:31:36 BluesLee: As to the echo, as of MS4, FSO includes a command to activate the suppression. In fact, adding %N0187 to %CPI handler didn't help me. What helped is to lower the volume of the speaker. Dec 25 12:32:35 PaulFertser: %N0187 helped me today Dec 25 12:32:56 PaulFertser: are you sure that fso does include %N0187 at commands? at least it is not included in shr yet ... Dec 25 12:33:06 BluesLee: fso-frameworkd git does not send %N0187 Dec 25 12:33:40 lindi-: me too, i used the above link and it seems to fix the echo Dec 25 12:33:55 BluesLee: i am discussing this in #122 right now Dec 25 12:34:49 ? on shr? Dec 25 12:35:06 BluesLee: I used slightly trimmed settings from http://wiki.openmoko.org/wiki/Freerunner_Hardware_Issues Dec 25 12:36:26 PaulFertser: okay ... i think that teh above link does a good job concerning echo suppression. i used alsamixer to lower buzz without success Dec 25 12:36:40 lindi-: ticket #122 on fso? Dec 25 12:37:13 i've got it Dec 25 12:37:28 BluesLee: I used my link and i had to use alsamixer while asking another side for testing the echo. I basically used "better set of mixer settings" and lowered the speaker volume. Dec 25 12:38:07 busthe "illume" title.. is in code Dec 25 12:38:12 if no other app is focused Dec 25 12:38:14 :) Dec 25 12:38:20 About %Nxxxx commands: FSO incorporates these for quite som time but not the strongest (0187) one. And not on every %CPI response (it's a hack anyway). Mickey promised to make it configurable soon. Dec 25 12:38:23 BluesLee: sorry :) for u Dec 25 12:38:26 :) Dec 25 12:39:06 BluesLee: i just sent my test report to #122, please refresh :) Dec 25 12:39:07 raster: no problem at all Dec 25 12:39:47 BluesLee: i'll workaroud this in my dialer until frameworkd fixes it Dec 25 12:39:56 mickey should include something we could test Dec 25 12:39:58 lindi-: what dialer do you maintain? Dec 25 12:40:12 BluesLee: I think Joerg confirmed that somebody's already asked Dieter to clarify when the command should be send (we are still not certain what kind of events can reset the setting). Dec 25 12:40:44 openmoko: nice nickname you have ;) Dec 25 12:40:50 lindi-: thanks. Dec 25 12:41:22 openmoko: kapula, it's just a very primitive dialer UI Dec 25 12:41:31 lindi-: written in Python Dec 25 12:41:53 lindi-: good job Dec 25 12:42:29 PaulFertser: who is Dieter? Dec 25 12:42:33 BluesLee: In fact FSO included the %N command from at least 4e797f9d (2008-10-05). Look at ti_calypso/channel.py Dec 25 12:42:48 BluesLee: The hardware hacker hired by OM to solve GSM firmware issues. Dec 25 12:43:17 PaulFertser: fso included %N0187 for a while, then it was removed and changed to other %N commands Dec 25 12:43:56 lindi-: Yes, but the other commands also work and sound not so harsh, according to Mickey. Dec 25 12:44:20 PaulFertser: they don't work as well here Dec 25 12:44:56 PaulFertser: harsh? Dec 25 12:45:00 lindi-: Have you tried to change channel.py instead of sending %N0187 on every %CPI. Dec 25 12:45:27 I think that's the work he used. Dec 25 12:45:27 okay Dec 25 12:45:27 c.append( "%N028B" ) # Long Echo Cancellation: active, -6db Dec 25 12:45:27 c.append( "%N0125" ) # Noise reduction: active, -6db Dec 25 12:45:27 thats included Dec 25 12:45:32 yep Dec 25 12:45:43 PaulFertser: change channel.py to do what? Dec 25 12:46:07 lindi-: To send %N0187 instead of %N028B, %N0125 :) Dec 25 12:46:49 PaulFertser: doesn't it send those only at init time? Dec 25 12:46:54 c = self._commands["init"] Dec 25 12:46:57 is the above fix to include the command in unsolicited.py okay? Dec 25 12:47:10 which fix? Dec 25 12:47:25 http://lists.openmoko.org/pipermail/community/2008-November/036068.html Dec 25 12:47:42 devchannel = self._object.modem.communicationChannel("DeviceMediator") Dec 25 12:47:42 devchannel.enqueue( "%N0187" ) Dec 25 12:47:42 info = {} Dec 25 12:47:53 i would guess it would help Dec 25 12:48:07 PaulFertser: Did they hire dieter especially for this project? Dec 25 12:48:18 BluesLee: should i try it? Dec 25 12:48:19 I hope dieter stays on. Dec 25 12:48:21 BluesLee: It's a hack. It should probably work, but sometimes the modem answers ERROR to to it. Not sure it can do any harm. But it's a hack. Not quite right place and time to send this command. Dec 25 12:48:35 PaulFertser: I think he might be a bit busier with #1024. Dec 25 12:49:00 lindi-: seems to work for me i am struggling with buzzing Dec 25 12:49:25 BluesLee: i am sending %N0187 now from the UI Dec 25 12:49:40 BluesLee: and it seems to work too, i'll remove the workaround from ui when frameworkd gets it Dec 25 12:49:42 The most interesting would be to find out whether changing commands in channel.py to 0187 helps or not. That's important. Dec 25 12:50:04 PaulFertser: they are sent at init time only so it can't be permanent fix Dec 25 12:50:14 PaulFertser: i need to re-issue %N0187 after every incoming call Dec 25 12:51:01 openmoko: I think they hired Dieter for the project of fixing calypso's firmware, when it became obvious that TI won't help. And he's busy with 1024, but i think he can't work full time on it without experimental data, and there's not much of it :( Dec 25 12:51:09 where is the right place in shr to do that? Dec 25 12:51:46 PaulFertser: Is there any way to make experimental data? I'd be willing to pay a night in a motel 6 here Dec 25 12:52:00 btw: Do you know what a motel 6 is Dec 25 12:53:07 lindi-: That's a very important information. Would you mind adding your test results to http://trac.freesmartphone.org/ticket/274 ? And i think it's smth mickey|xmas should be interested. Dec 25 12:53:40 openmoko: No, motel 6 is something new to me :) Dec 25 12:53:57 PaulFertser: Ah, it's a low-cost, somewhat crappy hotel chain Dec 25 12:55:08 openmoko: wrt experimental data: i think Dieter will write to the mailing list when he'll have some ideas as what should be done. He's already aware of everybody's equipment, i think. Dec 25 13:00:20 incredible Dec 25 13:00:48 undocumented at commands falling from heaven Dec 25 13:01:27 Hahah Dec 25 13:01:48 magic numbers 187, is it prime? Dec 25 13:04:32 hi Dec 25 13:04:40 hi;-) Dec 25 13:04:49 :D Dec 25 13:05:07 Dave: hi dave, do you like the new snapshots? Dec 25 13:05:12 Yes! Dec 25 13:05:31 raster: why the heck does e take around 15-20% cpu on the phone? Dec 25 13:05:35 It's funky :p Dec 25 13:05:46 (when doing nothing) Dec 25 13:05:47 ronny: shouldnt Dec 25 13:05:50 ises 1-3% here Dec 25 13:06:02 and that seems mostly polling dbus fso Dec 25 13:06:19 raster: i have also about 15% Dec 25 13:06:43 :x Dec 25 13:06:55 hmm Dec 25 13:07:18 i just killed framewkrd cause it took the rest of thr cpu Dec 25 13:07:36 Dave: no funk there, just blues;-) Dec 25 13:08:15 all i see is once a secopnd its polling for both FSO and qtopia gsm signal status for me Dec 25 13:08:24 what about the other shr users, how much cpu takes e? Dec 25 13:08:25 i dont have a sim card in - so it'll keep polling till someone answers Dec 25 13:08:49 PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND Dec 25 13:08:49 1147 root 10 -10 22784 10m 5944 S 1.9 35.7 349:42.98 enlightenment Dec 25 13:08:55 1.9% Dec 25 13:09:17 ronny: do you use a custom theme? Dec 25 13:09:40 BluesLee: which new snapshot? Dec 25 13:09:41 bumbl: ilumine on shr Dec 25 13:09:57 ronny: i mean custom icons Dec 25 13:09:59 bumbl: screenshot on snap Dec 25 13:10:03 bumbl: no custom icons Dec 25 13:10:10 i use custom icons Dec 25 13:10:11 hmm then Dec 25 13:10:17 bumbl: just a fresh install of shr Dec 25 13:10:45 bumbl: what says htop on your neo? Dec 25 13:10:45 BluesLee: http://scap.linuxtogo.org/files/efdd39c5a0f32c1a76fcbfe549f74f9f.png <- Dec 25 13:10:51 looks good Dec 25 13:11:26 i have to get rid of the non gpl stuff Dec 25 13:11:48 but that will not be so hard Dec 25 13:12:34 hmm I might convert the oxygen icon theme and adapt the background and illume.edj to look as nice as your screenshot does Dec 25 13:12:58 bumbl: yeah do that, i also downloaded the oxygens Dec 25 13:13:32 oh wait Dec 25 13:13:35 bumbl: the background is proprietary stuff, but we could take one wit the om logo there Dec 25 13:13:35 that was my treo Dec 25 13:13:51 neo not treo;-) Dec 25 13:14:05 same image tho Dec 25 13:14:06 same e Dec 25 13:14:08 same everything Dec 25 13:14:16 different hardware Dec 25 13:14:45 raster: does fso gsm work on the treo? Dec 25 13:14:53 PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND Dec 25 13:14:54 1263 root 20 0 24048 15m 6312 R 96.3 12.6 2752:19 python Dec 25 13:14:54 1338 root 10 -10 23100 12m 7680 S 1.0 9.9 6:27.81 enlightenment Dec 25 13:14:54 1362 root 21 1 14516 6768 3820 S 0.0 5.4 0:02.11 elementary_test Dec 25 13:14:56 1317 root 19 -1 9968 2844 1560 S 0.6 2.3 1:34.06 Xglamo Dec 25 13:15:10 BluesLee: hell no. its too memory hungry Dec 25 13:15:24 you have 10% Dec 25 13:15:36 10%? Dec 25 13:15:38 now i have 11% Dec 25 13:15:47 no Dec 25 13:15:47 1% Dec 25 13:16:01 why the hell Dec 25 13:16:02 fso is erating 96.3% Dec 25 13:16:46 raster: you use the default shr? Dec 25 13:17:04 raster: or fso ms4 or ... Dec 25 13:17:31 i dont use shr Dec 25 13:17:38 i use the illume-image i build Dec 25 13:17:48 it uses fso though Dec 25 13:17:48 it would be nice to lower this in shr Dec 25 13:17:54 but doesnt have much else on it Dec 25 13:18:12 i just do the illume-image Dec 25 13:18:16 its simple and works Dec 25 13:18:17 maybe it isnt build well in shr or we use modules which eat up to much cpu Dec 25 13:18:29 one moment Dec 25 13:18:36 i use illume-shr Dec 25 13:18:47 maybe i switch back and test it Dec 25 13:19:27 bbl Dec 25 13:20:08 <1% Dec 25 13:20:16 with illume-theme Dec 25 13:20:36 and sim authentification shot up in some secs Dec 25 13:20:44 rebooting:-) Dec 25 13:20:45 hmm so it is shr theme Dec 25 13:21:04 the only thing that redraws for me is the + on the battery Dec 25 13:21:09 because its recharging Dec 25 13:21:45 raster: with the illume theme everything is fine Dec 25 13:21:47 well shr uses a different battery image and a different clock Dec 25 13:22:41 (e has a redraw debug mode - Dec 25 13:22:43 export REDRAW_DEBUG=1 Dec 25 13:22:55 if thats set before e starts it'll flash any area it redraws black & white Dec 25 13:22:59 several times Dec 25 13:23:07 so you can see just what it's redrawing Dec 25 13:23:44 as i said Dec 25 13:23:46 okay, sim auth takes its time again Dec 25 13:23:47 i maintain the illume-image Dec 25 13:23:49 and defaults Dec 25 13:23:50 :) Dec 25 13:23:56 but htop says 1 % for e Dec 25 13:23:58 and it does just what i'd expect it to Dec 25 13:24:49 ie it uses a little bit of cpu waking either pollign for a non-existant fso (oon my palm as i had to kill python just to make it usable) Dec 25 13:25:22 BluesLee, I saw your theme sketches earlier, transparent illume bar is wicked! Dec 25 13:25:35 and on the freerunner its even less Dec 25 13:25:49 as all it does is wait for dbus singals Dec 25 13:26:01 only thing it polls is bluetooth which is minimal Dec 25 13:26:29 TAsn: its not transparent Dec 25 13:26:53 it just draws itself like the background? Dec 25 13:27:05 that's also quite cool. Dec 25 13:27:21 raster: can you have a look at the clock function here http://pastebin.ca/1293436 Dec 25 13:27:32 it just has a nice image on the top-shelf Dec 25 13:27:42 TAsn: yeah its simple Dec 25 13:27:47 raster, i c, well it's still a nice effect ;] Dec 25 13:27:51 ;] Dec 25 13:27:58 TAsn: took a 480x270 image Dec 25 13:28:06 BluesLee, yeah, I got it... Dec 25 13:28:30 raster: it doesnt like transparency in the topshelf, i tried it, everything is messed up Dec 25 13:28:36 raster, are there any plans to add the ability to change the desktop icon positioning? (something iphone like, or maybe even a better way?) Dec 25 13:29:24 i suspect the clock may be ticking much faster than u think Dec 25 13:29:29 TAsn: iphone is a bad example but i agree there should be something included... i would prefer categories gsm apps, games ... Dec 25 13:29:32 BluesLee, well, you got the same results without it, so I'm don't think he should dwell on it, don't you agree? Dec 25 13:29:55 BluesLee, categories can be nice, though maybe also custom ordering is needed. Dec 25 13:30:09 the iphone is a bad example because it's an annoying way to do so Dec 25 13:30:12 TAsn: i think shr switched because of the analog clock ... Dec 25 13:30:44 but a settings button that enables free position changing can be great. Dec 25 13:30:47 in fact Dec 25 13:30:49 TAsn: and i dont like the analog as many others too Dec 25 13:30:53 i almost guaranteed it Dec 25 13:31:02 its scheduling a timero for 0 seconds form now Dec 25 13:31:03 or so Dec 25 13:31:20 BluesLee: u can do it - if you say its shaped Dec 25 13:31:29 but that will screw qith event catching Dec 25 13:31:38 u coudl do that and turn compositing on and use a compsoitor Dec 25 13:31:39 BluesLee, the analog clock sucks bad. the first thing I did when downloading my first shr image was changing it to a digital one.\ Dec 25 13:31:41 eg like bling Dec 25 13:31:44 or xocmpmgr Dec 25 13:31:49 but u'd have wonders peed drops Dec 25 13:31:53 speed Dec 25 13:32:26 TAsn: i havce no lans to do anything interestig with the icons Dec 25 13:32:40 i c. Dec 25 13:32:41 raster: i didnt do the illume-shr theme and the didgital clock there. i am just wondering what takes so much cpu time there Dec 25 13:32:42 that was a quick drop in to avoid the hell that was sean's "Slider mode" design Dec 25 13:32:53 change to slider mode to see it Dec 25 13:32:57 it seems cool Dec 25 13:33:01 but functions like arse Dec 25 13:33:11 slider mode is hate Dec 25 13:33:18 BluesLee: look at the math Dec 25 13:33:40 second will be 0.0 -> 59.9999 \ Dec 25 13:33:43 do you mean the big blocks you can slide to the side ... Dec 25 13:33:47 ie how many seconds (betwene minutes) Dec 25 13:33:57 okay Dec 25 13:33:59 now lets assume we actually tick ON the second Dec 25 13:34:01 exactly Dec 25 13:34:03 seond == 0 Dec 25 13:34:13 err sorry Dec 25 13:34:16 tick just BEFORE Dec 25 13:34:24 59 Dec 25 13:34:28 err 59.9 Dec 25 13:34:36 s = 60 - 59.9 = 0.1 Dec 25 13:34:42 BUT Dec 25 13:34:45 nono Dec 25 13:34:48 i wqs right first time Dec 25 13:34:52 @ second = 0 Dec 25 13:34:52 :-) Dec 25 13:34:58 s = 60 - 0 Dec 25 13:35:00 == 60 Dec 25 13:35:07 then set a timer for 60 - 60 Dec 25 13:35:09 whihc is.. 0 Dec 25 13:35:20 so thus run immediately again Dec 25 13:35:38 u want to set the timer for sxecond Dec 25 13:35:44 not 60 - second Dec 25 13:36:21 nb Dec 25 13:36:33 i fixed that mistake long ago int he original asu theme Dec 25 13:36:50 seems like someone blindly has copied it from an older version and never done any review/checking Dec 25 13:36:54 nor done any math Dec 25 13:37:01 and everyone keeps coming to ask me Dec 25 13:37:02 :) Dec 25 13:37:57 PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND Dec 25 13:37:58 1263 root 20 0 23912 15m 6312 R 97.3 12.6 1:17.71 python Dec 25 13:37:58 1364 root 20 0 2276 1108 888 R 1.9 0.9 0:00.62 top Dec 25 13:37:58 1359 root 20 0 2616 1144 756 S 0.3 0.9 0:00.62 dropbear Dec 25 13:38:00 1 root 20 0 1524 556 496 S 0.0 0.4 0:03.91 init Dec 25 13:38:03 on freerunner Dec 25 13:38:06 e is at 0% Dec 25 13:38:13 as the battery has stoppe dcharging Dec 25 13:38:15 charging Dec 25 13:38:21 as i'd expect Dec 25 13:38:22 :) Dec 25 13:40:15 okay, thanx Dec 25 13:40:45 quickdev or Ainulindale will have to change that there Dec 25 13:43:33 :) Dec 25 13:44:15 i am only an enduser:-) Dec 25 13:50:22 well raster you are the e guru so that's why everyone keeps asking you Dec 25 13:52:48 raster: do you still work for swisscom? Dec 25 13:53:53 bumbl: yes - still doing stuff for them Dec 25 13:54:03 in fact.. working on sms app now Dec 25 13:54:11 moving to use toolbar for top actions Dec 25 13:54:15 and now.. need to hook it in Dec 25 13:54:19 but will do that tomorrow Dec 25 13:54:27 past pumpkin time here Dec 25 13:54:30 and its xmas Dec 25 13:54:50 something like sms notify in the topshelf? Dec 25 13:54:56 no Dec 25 13:55:05 a full sms reader/editor Dec 25 13:55:14 for swisscom Dec 25 13:55:15 nothng to do with notification Dec 25 13:55:20 i got it Dec 25 13:55:33 this is more a demo app Dec 25 13:55:42 to show how to write such a thing using elementary Dec 25 13:55:47 its not mesnt to be powerful Dec 25 13:55:50 just work and look good Dec 25 13:55:59 hohoho! Merry christmas or whatever festivities everybody is endulging in. Dec 25 13:56:22 i removed to the analog clock temporary Dec 25 13:56:23 already have done the data storage layer Dec 25 13:56:43 and contacts storage Dec 25 13:56:47 and lookup Dec 25 13:56:50 well code-level lookup Dec 25 13:59:07 ah ok Dec 25 13:59:14 where i can i modify the look of gtk apps? Dec 25 13:59:32 so we might get an image powered by swisscom somewhere in the future? Dec 25 13:59:56 you'll need to ask them Dec 25 13:59:57 :) Dec 25 14:00:23 what i know or dont know may or may not be under nda :) Dec 25 14:00:40 haha Dec 25 14:00:54 the scope of whats public expentds to the code work and apps i do - not their future plans :) Dec 25 14:00:57 err Dec 25 14:00:59 extends Dec 25 14:01:01 found media player perfect for openmoko! Dec 25 14:01:02 http://scap.linuxtogo.org/files/cee628fa141da42f4ae1fdafb8ddef91.png Dec 25 14:01:04 ;DDD Dec 25 14:01:18 wow Dec 25 14:01:22 * bumbl wants ipk Dec 25 14:01:26 :p Dec 25 14:02:01 dos1: i think it will be the standard player in shr Dec 25 14:02:04 does it run fluently? Dec 25 14:02:07 dos1: no joke Dec 25 14:02:35 bumbl, BluesLee: it's only X forwarding Dec 25 14:02:36 ;] Dec 25 14:02:45 dos1: bah Dec 25 14:02:47 cheater Dec 25 14:02:51 :D Dec 25 14:03:45 but i want to test kde4.2 with debian Dec 25 14:04:06 wait for kde 4.3 that should be fingerfriendly thanks to nokia Dec 25 14:04:25 if only some kernel guy will fix glamo-mci... Dec 25 14:04:51 in u-boot my 4gb microsd is working nicely... i can list all partitions Dec 25 14:04:55 and folders Dec 25 14:05:11 in linux... it only reads first partition table Dec 25 14:05:50 and prints lot of errors with glamo Dec 25 14:06:24 dos1: Have you tried andy-tracking? Dec 25 14:06:35 well I have a 8gb µsd and i can access all the partitions Dec 25 14:06:48 PaulFertser: yes, but only when first android images was available Dec 25 14:06:59 later not Dec 25 14:07:02 dos1: I mean, does it work with your 4Gb card? Dec 25 14:07:13 PaulFertser: when i was trying - no Dec 25 14:07:30 amarok hanged :D Dec 25 14:08:09 oh, whole X hanged Dec 25 14:08:10 ^^" Dec 25 14:08:11 ummmm Dec 25 14:08:18 that doesnt look very usable Dec 25 14:08:25 hi all Dec 25 14:08:29 as it basically cant fit the ui onto the screen Dec 25 14:09:41 raster: who cares ;] Dec 25 14:09:47 i run it only for fun Dec 25 14:09:47 :D Dec 25 14:10:25 not sure whats fun... but hey Dec 25 14:10:38 maybe the novelty of "shit works when its compiled" wore off on me a long time ago Dec 25 14:10:39 :)O Dec 25 14:11:05 dos1: where is the ipk located? Dec 25 14:11:57 BluesLee: nowhere ;] Dec 25 14:12:42 dos1: try it in landscape mode Dec 25 14:12:53 i'm trying now Dec 25 14:13:38 http://scap.linuxtogo.org/files/b5a8243a76b6d593f5e32673bfb60bde.png Dec 25 14:13:48 raster: i have a question concerning the kbd, if i rotate the screen its okay, but rotating back messes up the keyboards, it only shows some parts of it Dec 25 14:13:50 in landscape it looks beeter ;] Dec 25 14:14:01 i know Dec 25 14:14:14 raster: what is the problem there?= Dec 25 14:14:28 layout handling inside the illume kbd Dec 25 14:14:33 hello everyone. Dec 25 14:14:43 * admiral0 feels fat Dec 25 14:15:07 raster: one more, how can i remove the right > app switcher correctly? Dec 25 14:15:15 admiral0: what have you had for xmas dinner? Dec 25 14:15:28 it was only lunch ^^ Dec 25 14:15:58 raster: its called slipshelf:) Dec 25 14:16:24 lasagna, vittel tonnato (i dunno how to translate it) Dec 25 14:16:43 heh Dec 25 14:16:46 ah Dec 25 14:16:48 mhm Dec 25 14:16:52 raster: for me cyclic switching is okay, so i get some additional space there Dec 25 14:17:16 BluesLee any news with your theme? Dec 25 14:17:28 we had a delicious wild duck here Dec 25 14:18:12 mhm Dec 25 14:18:17 delicious Dec 25 14:18:31 admiral0: i am beginning to understand edje more and more ... Dec 25 14:19:37 Yay! Dec 25 14:19:45 Bluelee will be unstoppable soon :D Dec 25 14:19:49 admiral0: i have to change the wallpaper and the topshelf image and remove the right > app slider and wait for a non cpu consuming digital clock and ... Dec 25 14:20:48 Dave: i wouldnt bet, i have to work from next week on and i am learning very slowly Dec 25 14:22:09 BluesLee: "next_app" button is the image Dec 25 14:22:16 notice it depends on "next" too Dec 25 14:22:26 that was the problem Dec 25 14:22:33 (so you'll need to remvoe all of that and anything that adjust relative to it to adjust to something else)\ Dec 25 14:22:38 time to try Qi and 2.6.28 Dec 25 14:22:47 basically in edje everything is relaitve to something Dec 25 14:22:56 if not specified its relative to the whole edje object Dec 25 14:22:58 raster: i tried to remove the corresponding parts but it didnt build Dec 25 14:23:05 (in this case ther whole slipshelf including the offscreen bits) Dec 25 14:23:13 no - as things are rtelative t6o it Dec 25 14:23:21 you need to change them to be relative to dsomething else Dec 25 14:23:27 as they lost their dependency Dec 25 14:23:44 dos1: i have a 2.6.28 kernel Dec 25 14:23:52 already compiled Dec 25 14:24:12 admiral0: hey, cool, what about the fs? Dec 25 14:24:16 admiral0: and how it's working? Dec 25 14:24:38 it's working Dec 25 14:24:38 ok Dec 25 14:24:40 snoozies! Dec 25 14:24:44 nite! Dec 25 14:25:37 BluesLee:about the fs? Dec 25 14:25:38 dos1: i use andy-tracking for more than a week Dec 25 14:25:39 admiral0: suspend/resume? halting? Dec 25 14:25:57 some problems with halting Dec 25 14:26:00 PaulFertser: i want to try it with my microsd card Dec 25 14:26:02 and video Dec 25 14:26:09 dos1: suspend/resume is ok. Halting doesn't work. Dec 25 14:26:15 admiral0: 2.6.28 kernel with shr? i thought the filesystem /sys has to be modified Dec 25 14:26:24 dos1: bluetooth is unusable after resume. Dec 25 14:26:33 ok, i'm not using bluetooth Dec 25 14:26:47 http://downloads.tuxfamily.org/admiral0repo/ Dec 25 14:26:51 BluesLee: no problem, you can apply my temporary patch to the FSO ms4 and that'll be enough, i think. Dec 25 14:27:10 PaulFertser: and patch to exposure with backlight Dec 25 14:27:11 I use 2.6.28 since rc1 and it is quite nice (not on the freerunner though but on the laptop) Dec 25 14:27:27 bumbl: not relevant :p Dec 25 14:27:33 :p Dec 25 14:27:44 admiral0: what problems with video? Dec 25 14:27:48 admiral0: any advantages? Dec 25 14:27:56 strange flickers.. Dec 25 14:28:09 i read wifi is working more reliable Dec 25 14:28:16 admiral0: and i've no flickers, btw. Dec 25 14:28:20 i noticed a quicker suspend/resume Dec 25 14:28:34 i'll recompile it with better flags Dec 25 14:28:45 *cflags Dec 25 14:28:57 PaulFertser: from where is your kernel? Dec 25 14:29:00 Yeah, suspend is quick Dec 25 14:29:03 maybe only a -mcpu=arm920t Dec 25 14:29:14 dos1: c59a8bdc7ca4b547 Dec 25 14:29:27 PaulFertser: but you compiled it or downloaded? Dec 25 14:29:53 dos1: I compiled it, with default config. Dec 25 14:30:00 me too Dec 25 14:30:27 anyway... if you want to try here's ctorrent that i compiled some time ago http://downloads.tuxfamily.org/admiral0repo/Incoming/unpackaged/ctorrent/ Dec 25 14:32:58 i am also interested in the newer kernel Dec 25 14:33:29 you can find it at the link i passed first Dec 25 14:33:37 uImage-GTA02.bin Dec 25 14:33:49 http://downloads.tuxfamily.org/admiral0repo/uImage-GTA02.bin Dec 25 14:34:02 but i have to change something in the shr fs? Dec 25 14:34:16 i didn't change anything Dec 25 14:34:26 but back up your kernel Dec 25 14:34:35 this one is "violent" Dec 25 14:34:49 it may kick your ass Dec 25 14:34:53 :P Dec 25 14:37:30 admiral0: yours kernel is booting now my qtopia from 512mb card Dec 25 14:37:32 for test Dec 25 14:37:44 BluesLee: for shr you need to apply a patch to FSO. Dec 25 14:37:57 and framebuffer speed is so good Dec 25 14:38:14 didn't test with fb Dec 25 14:38:34 admiral0: i mean text on booting Dec 25 14:38:45 :D Dec 25 14:39:02 it scrolls much betten than in 2.6.24 Dec 25 14:39:02 oh Dec 25 14:39:02 i did some cflags magic Dec 25 14:39:11 WSOD Dec 25 14:39:14 my first Dec 25 14:39:15 :DDD Dec 25 14:39:16 no wait Dec 25 14:39:28 i think it's just fb init Dec 25 14:40:08 i never had WSOD on resume Dec 25 14:40:11 since now Dec 25 14:40:13 ^^ Dec 25 14:41:09 ^^ Dec 25 14:41:54 booted? Dec 25 14:42:01 yes Dec 25 14:42:12 ok Dec 25 14:42:17 but auto-suspend in my qtopia is 10sec Dec 25 14:42:18 ;D Dec 25 14:42:25 ... Dec 25 14:43:44 i must recompile it these days Dec 25 14:43:58 wtf? Dec 25 14:44:02 i need better cflags Dec 25 14:44:08 admiral0: now it doesn't boot Dec 25 14:44:08 ? Dec 25 14:44:16 i doesn't see sd card Dec 25 14:44:18 what does it say? Dec 25 14:44:32 kernel panic Dec 25 14:44:35 not syncing Dec 25 14:44:43 append correct init= Dec 25 14:44:53 mmcblk0p2 isn't correct, blahblah Dec 25 14:44:54 ;) Dec 25 14:45:16 d'oh Dec 25 14:45:25 sd card must be working, because kernel is booted from sd Dec 25 14:45:36 few minutes ago it worked Dec 25 14:45:48 i only rebooted it after WSOD Dec 25 14:45:52 and now it doesn't Dec 25 14:45:53 :D Dec 25 14:46:03 strange Dec 25 14:46:04 ;] Dec 25 14:46:22 very Dec 25 14:46:40 i'd like to try another distro other than shr Dec 25 14:46:51 what do you suggest(no android please) Dec 25 14:46:58 debian Dec 25 14:47:17 already tried Dec 25 14:47:27 and i want something on the rootfs Dec 25 14:48:32 i wan't to use debian with shr apps Dec 25 14:48:36 want* Dec 25 14:48:46 why? Dec 25 14:49:23 debian - apt-get, repos and applications Dec 25 14:49:48 shr - because i very like this stack Dec 25 14:49:51 dos1: have you tried rootdelay=1 ? Dec 25 14:49:52 i am starting another distro... Dec 25 14:50:22 does pacman say anything? (no not the game) Dec 25 14:50:30 arch Dec 25 14:50:31 ;) Dec 25 14:50:45 yeah Dec 25 14:50:59 i have already a kernel and busybox Dec 25 14:52:46 PaulFertser: no, thanks, i forgot about it Dec 25 14:56:05 it boots :) Dec 25 14:57:08 yeah Dec 25 14:58:26 Ainu says that instability of andy-tracking is something that hinders the SHR development. Honestly, i don't why he thinks so. Yes, bluetooth is broken after suspend. It doesn't properly halts yet. But that's about all that's not working. It shouldn't matter too much for SHR development anyway. Seriously, i can't see why not to use it now. Dec 25 15:00:03 PaulFertser: well maybe because it could cause problems in the future as fso does not yet use andy-tracking and maybe because andy included some controversial patches into the kernel Dec 25 15:00:30 what patches? Dec 25 15:00:55 bumbl: Let's be honest. The changes to FSO are really very minimal. And Mickey's already implemented what almost everything necessary in HEAD. So, /sys changes are a non-issue. Dec 25 15:01:51 bumbl: Please clarify, what controversial patches included in andy-tracking, can harm SHR? Dec 25 15:02:58 PaulFertser: wsod patch Dec 25 15:03:16 i never had a wsod Dec 25 15:03:48 i had it once, today, with andy-tracking :D Dec 25 15:04:06 bumbl: it was included in "stable" as well. Moreover, it harms suspend time a little. Even if it's really 15%, does it really matter to the SHR developers? Do they develop in absence of power? What's the matter with that, really? Dec 25 15:04:39 ehem Dec 25 15:04:59 turning off screen was in kernel only few days Dec 25 15:05:20 and then it was deleted Dec 25 15:05:45 Seriously, guys. If you need a reliable phone, get a Nokia 1100. Why bother so much about suspend time etc?.. Dec 25 15:05:56 because of wsod Dec 25 15:06:29 well i would prefer to have e.g one hour longer battery life because if you want to use the neo as a phone Dec 25 15:06:38 you need a certain amount of battery life Dec 25 15:06:41 bumbl: me too Dec 25 15:07:00 * bumbl wants back fso ms 2 Dec 25 15:07:19 back then I had 4 days Dec 25 15:08:08 bumbl: why are you so sure, that the wsod patch steals one hour from you? A wild guess?.. I think you can always revert it then and compile without if you're so picky. Dec 25 15:09:40 well anyway - this is a desicion which Ainulindale has to make Dec 25 15:10:02 Just carry a working 1100 with you and swap a SIM when FR runs out of power. What can be simpler? It's still development, and for the development the time we have, i guess, is more than enough. Dec 25 15:10:46 hmm the carrying of the 1100 is the problem Dec 25 15:10:46 i want debian with shr apps with 2.6.28 kernel without wsod patch :] Dec 25 15:10:53 but i'm too lazy to do it :D Dec 25 15:11:14 at the moment I already take too many gadgets with me Dec 25 15:11:22 I've seen Ainulindale frustrated with andy-tracking. But unfortunately, no details. Dec 25 15:18:58 Qi is booting kernel from my 4gb microsd Dec 25 15:19:12 kernel see all partitions Dec 25 15:19:18 logical on extended too Dec 25 15:19:30 but "Warning: unable to open an initial console." Dec 25 15:19:37 but maybe it is problem with my rootfs Dec 25 15:23:47 dos1: are you sure you're booting kernel from the same partition where you rootfs is? Dec 25 15:23:58 yes Dec 25 15:28:47 So, the new kernel can actually see you card. No errors are output to the console. But it doesn't like your rootfs... I think that can be fixed. What's really important is that Qi and the kernel sees you card ok. Dec 25 15:31:45 PaulFertser: now i'm booting nand rootfs with new kernel Dec 25 15:31:54 and i will check it Dec 25 15:36:39 PaulFertser: it works :D Dec 25 15:36:49 dos1: What was wrong then? Dec 25 15:37:15 i mean reading from sd card from nand rootfs Dec 25 15:37:36 now i will delete existing and install new debian ;] Dec 25 15:38:24 dos1: Do you see flickering like admiral0 ? BTW, if you're going to try to adopt anything (like o-p-p), you can see http://wiki.openmoko.org/wiki/Kernel#GTA02_.2Fsys_mappings_for_andy-tracking_vs_stable for reference. Dec 25 15:39:24 dos1: You can find a patch to FSO ms4 at http://trac.freesmartphone.org/ticket/293 Dec 25 15:40:00 PaulFertser: no, but animations in Illume are slower with this kernel Dec 25 15:40:23 i disabled animations Dec 25 15:40:29 Are you sure that it's not subjective? Dec 25 15:40:36 yes, i'm sure Dec 25 15:41:16 are you sure that you are sure ^^ Dec 25 15:41:18 ? Dec 25 15:41:27 :) Dec 25 15:41:50 admiral0: (starting X again to check :P) Dec 25 15:42:06 dos1: Xglamo? Dec 25 15:42:12 yes Dec 25 15:42:25 it's a bit slower Dec 25 15:42:50 animations in keyboard too Dec 25 15:42:57 but it's usable Dec 25 15:47:05 oh Dec 25 15:47:09 now it isn't working Dec 25 15:47:16 dos1: i guess it'd be good if you raised this question on devel- and then someone can suggest a tool for measurement and then if it's really slower (and not Xglamo related, which is deprecated anyway :p ), it can be fixed. We want a modern kernel anyway. Let SHR developers choose one stable revision and use it, and let other users who has time try to help fix regressions. Dec 25 15:47:18 sd card Dec 25 15:47:34 dos1: and what about initdelay=5 ? Dec 25 15:47:50 no, i have i/o errors with rootfs on nand Dec 25 15:48:13 Oh, that sucks... :( Dec 25 15:48:38 Are you sure the card is ok? Have you tried it in a decent card-reader? Dec 25 15:48:48 yes, card is ok Dec 25 15:48:52 [ 480.685000] glamo-mci glamo-mci.0: Error after cmd: 0x8020 Dec 25 15:48:52 [ 480.685000] mmcblk0: error -110 sending read/write command Dec 25 15:48:52 [ 480.695000] end_request: I/O error, dev mmcblk0, sector 1866827 Dec 25 15:48:52 [ 480.705000] glamo-mci glamo-mci.0: Error after cmd: 0x120 Dec 25 15:49:09 this is in dmesg Dec 25 15:49:14 many lines like that Dec 25 15:49:44 Can you pastebin the whole dmesg? Dec 25 15:53:00 PaulFertser: Hey Dec 25 15:53:05 I'm not frustrated Dec 25 15:53:08 I'm outraged Dec 25 15:53:19 Because the way the versioning is handled isn't normal, coherent, and sane Dec 25 15:53:24 (kernel wise) Dec 25 15:53:36 Ainulindale: Hey. I'm sorry for that... Rage doesn't help, unfortunately. Dec 25 15:53:46 At the present time, new functionnalities imply regression every time Dec 25 15:53:49 And that's just not possible Dec 25 15:54:08 PaulFertser: Outraged. Not feeling rage :-) Dec 25 15:54:33 Ainulindale: why not just choose a one working version and stick to it for some time? Dec 25 15:54:45 PaulFertser: And what do you think we're doing right now? :-) Dec 25 15:55:18 We can keep doing that. That's not a good solution, as new kernel version imply patches, etc. Dec 25 15:55:25 Talk to mwester about that, you'll see what I mean Dec 25 15:55:59 Ainulindale: the problem is that you choosed a bad one ;) I mean, a 2.6.28-smth revision. I don't think any kernel changes from now on will require any patches to SHR at all. Dec 25 15:56:16 PaulFertser: I didn't choose. I didn't even test it. Dec 25 15:56:32 PaulFertser; after reboot it's working Dec 25 15:56:36 PaulFertser: I just read mails, and draw my own conclusions based on user experience I see around, and from people testing for me Dec 25 15:56:40 Because I don't have the time, seriously Dec 25 15:56:41 i'm trying to repeat problem Dec 25 15:56:44 And we can't chase the perfect kernel Dec 25 15:56:48 It should be there. Dec 25 15:57:31 I know what mwester thinks. And he has a lot of valid points. But... Neither SHR nor FSO developers need to upgrade every time HEAD is changed. You shoudn't chase. Just take smth stable and _recent_ and stick to it. Dec 25 15:57:43 It should be there, and it'll be there. Dec 25 15:57:49 And as i just told you Dec 25 15:57:59 Stable has borked things more often than not Dec 25 15:58:05 In fact I never saw a kernel who wasn't borked someway Dec 25 15:58:15 And anyway Dec 25 15:58:19 This is a stupid argument Dec 25 15:58:24 As I have no time to talk, clearly Dec 25 15:58:34 And I won't change my opinion that easily :-) Dec 25 15:58:43 I need to think about that, and now isn't the time Dec 25 15:58:50 Because I'm unpacking my new multi effect for my guitar! Dec 25 15:58:56 And you don't want to rob me of my me time :-) Dec 25 15:59:28 Ainulindale: You rock :) Don't waste your time then. Play! :) Dec 25 15:59:58 But please don't be so mean to a kernel you didn't even try yourself. Dec 25 16:00:29 I wasn't mean Dec 25 16:00:33 Just plain logical Dec 25 16:00:38 Even if it's my own twisted logic Dec 25 16:00:45 If something doesn't work and is handled poorly Dec 25 16:00:53 Then handle it by yourself Dec 25 16:00:55 That's how I roll Dec 25 16:01:14 But i use andy-tracking for a week now, and i can't feel regressions. Dec 25 16:01:35 And i remeber how mwester told me about userspace problems. And there were none in FSO, actually. Dec 25 16:01:42 I use SHR every day Dec 25 16:01:52 And it took me two weeks to find out GPS wasn't working Dec 25 16:03:35 I think, there is no need to fork 2.6.24. If you're happy with it, let it be. Just don't fork, there's no sense in it now anyway. That's what i think. Dec 25 16:03:51 Ok, i won't bother you anymore. I'll try to play my flute instead :) Dec 25 16:08:29 PaulFertser: http://pastebin.com/m362cf2d1 Dec 25 16:08:38 whole dmesg Dec 25 16:09:01 i/o errors was after some time when i do "find /media/mmcblk0p2/" Dec 25 16:11:53 dos1|away: I'm afraid there's something wrong with you partition table. I don't see any errors before it tries to access wrong data. Dec 25 16:13:56 all works until first error is printed Dec 25 16:14:13 then i can't read anything from sd Dec 25 16:14:20 ok, now i'm really away ;] Dec 25 16:50:39 Hey Dec 25 16:51:26 hi Dec 25 16:52:45 quickdev: is it planned to change the sms recipient layout Dec 25 16:52:55 in the efl-ui-new Dec 25 16:54:24 bumbl, not planned, but of course I'm open to better ideas :) Dec 25 16:55:24 i might try to do a mockup Dec 25 17:04:02 hi Dec 25 17:04:59 opkg upgrade on SHR testing give my an md5sum mismatch. Dec 25 17:05:15 is upgrade no possible at the moment? Dec 25 17:06:35 ingoa, i've this problem too Dec 25 17:07:52 ingoa, GNUtoo on which packages? Dec 25 17:08:22 on libframeworkd-glib0 Dec 25 17:13:51 ingoa, please try update / upgrade Dec 25 17:17:57 quickdev: Thank you. its running now Dec 25 17:38:25 dos1: Can you dump your partition table on FR and on another computer (fdisk -l)? "attempt to access beyond end of device" in your log is very suspicious to me. Dec 25 17:43:57 Ainulindale, here? Dec 25 17:44:31 PaulFertser: ok, i must reboot fr Dec 25 17:49:43 PaulFertser: "Unable to read /dev/mmcblk0" - so i must reboot again Dec 25 17:49:44 ;] Dec 25 17:49:54 quickdev: he is playing guitar ;D Dec 25 17:51:02 quickdev: ping Dec 25 17:51:17 he stopped Dec 25 17:51:18 ;] Dec 25 17:52:13 Ainulindale, the cache of shr-testing seems to be borked. Have you anything against rerunning the whole build to have something clean? Dec 25 17:52:41 What do you mean, it seems to be borked? Dec 25 17:53:49 try to compile ophonekitd, it makes errors that indicate that libframeworkd-glib is wrong, but its revision is correct. Could you have a look? Dec 25 17:54:08 Ok. Dec 25 17:54:16 dos1: I think you have some real issue, and it should be debugged. I'm not a kernel dev, so i can't help with deep inspection. But i can suggest to enable MMC_DEBUG in kernel and provide the logs for the failing and probably for a successful cases on openmoko bugtracker. Also it'd be good to know what exact parameters are used with this card when it's working flawlessly (on your laptop, in example; usb cardreaders can't help) Dec 25 17:55:22 PaulFertser: i don't have cardreader in laptop - only usb readers Dec 25 17:56:18 MMC_DEBUG? firstly i must read how to compile kernel for freerunner ;] Dec 25 17:57:11 ingoa, does it works...because after a update && upgrade it still have the error...should I reflash...what image do you have? Dec 25 17:58:28 dos1: compiling kernel is amazingly easy. Just trim ``build'' (the script in the root of git checkout) to point to your cross-toolchain and that's all. Dec 25 17:59:37 quickdev & Ainulindale: we had a diskussion with raster today why enlightenment takes more than 15% of cpu time. the digital clock in the illume shr theme is not implemented correctly, removing the part in illume-shr.edc gives me additional 14% Dec 25 18:00:06 dos1: If you'll see a laptop with a cardreader, you can dump the necessary parameters using it. If you're going to do so, i'll look at the code to actually describe what can be done (i.e. how to get enough debugging info from sdhci driver). Dec 25 18:00:28 quickdev: cleaned, recompiled, and it went through. Dec 25 18:00:33 GNUtoo, which package? Dec 25 18:00:50 quickdev, libframeworkd-glib0 Dec 25 18:00:55 but I have an old image Dec 25 18:00:57 Ainulindale, strange - have you compiled anything except libframeworkd-glib and ophonekitd? What/How did you clean? Dec 25 18:01:05 PaulFertser: now i'm booting linux on my mum's laptop - it has card reader Dec 25 18:01:05 bitbake -c clean libframeworkd-glib ophonekitd Dec 25 18:01:11 bitbake libframeworkd-glib; bitbake ophonekitd Dec 25 18:01:18 dos1: you don't need MMC_DEBUG yet, all the card parameters are available without it, i think. Dec 25 18:02:35 PaulFertser: ok, i have fdisk -l done on freerunner Dec 25 18:02:58 dos1: show it to me then :) Dec 25 18:03:00 http://pastebin.com/m58dd66ce Dec 25 18:05:12 PaulFertser: card on computer will be helpful if it is present in system by /dev/mmcblk?, not /dev/sd?, yes? Dec 25 18:05:30 now on this laptop this is /dev/sdb Dec 25 18:05:38 dos1: Ok, wait a little. Dec 25 18:07:39 PaulFertser: on laptop fdisk -l is the same Dec 25 18:07:48 what do you think about this? http://img101.imageshack.us/img101/9217/illumezo8.png Dec 25 18:08:00 dos1: yes, your partition table looks good... Dec 25 18:08:07 only "Disk identifier: 0x0002d397" is added and /dev/sdb is instead of /dev/mmcblk0 Dec 25 18:11:18 quickdev, so what should I do? reflash? Dec 25 18:11:48 dos1: And what about fdisk -ul ? Dec 25 18:12:01 on FR or on computer? Dec 25 18:12:14 It's in the computer, so do it there. Dec 25 18:12:49 GNUtoo, rm /var/lib/opkg/* ; opkg update ; opkg upgrade should fix it Dec 25 18:12:59 quickdev, thanks a lot Dec 25 18:13:42 GNUtoo, tell us if not :) Dec 25 18:14:02 quickdev, ok Dec 25 18:14:14 quickdev & Ainulindale: here you should find a digital clock which isnt cpu hungry http://www.e17-stuff.org/content/show.php/itask+ng?content=63394 Dec 25 18:16:39 PaulFertser: http://pastebin.com/m2a362176 Dec 25 18:18:36 dos1: the reason i ask is because i want to make sure that it's not a bad partition table. But it really seems to be ok... Dec 25 18:19:12 PaulFertser: so what can i do to debug it? Dec 25 18:19:22 mwester: maybe you know? ;] Dec 25 18:22:09 BluesLee: join #openmoko, some people are interested on your screenshots on linuxtogo Dec 25 18:22:43 BluesLee: Yeah, before they start invading the developer channel. ;) Dec 25 18:23:01 :-) Dec 25 18:25:50 dos1: I think you can't really get debug info from working case without compiling kernel with CONFIG_MMC_DEBUG. We already know which parameters are used on your FR, the log is verbose enough. I can only guess that some others are used on your laptop. But they're unavailable with a usb cardreader. Dec 25 18:29:03 dos1: And you can also try to lower glamo-mci clock frequency, with sd_max_clk= kernel parameter. Dec 25 18:29:19 i was trying it with 2.6.24 Dec 25 18:30:39 dos1: But it doesn't recognise a card at all? You've limited success now with 2.6.28, is it does recognise the card and you're able to read at least partition table. Dec 25 18:31:24 PaulFertser_: it sometimes recognised Dec 25 18:35:46 PaulFertser_: but now i have to flash u-boot and play with it to test Dec 25 18:36:19 dos1: why? I suggest you try sd_max_clk with 2.6.28. Dec 25 18:36:53 PaulFertser_: how can i do it with Qi? ;p Dec 25 18:38:18 dos1: if you're booting from SD, you can specify properties in /boot/append-GTA02. Dec 25 18:38:44 quickdev, doesn't work...there is also another error...but i bet that if I reflash they will go Dec 25 18:38:52 PaulFertser_: i know, but i'm not booting from SD due to i/o errors :D Dec 25 18:39:09 dos1: You can recompile Qi then :) Dec 25 18:39:49 Ainulindale, why are there libframeworkd-glib md5sum mismatch errors? Although we just recompield that lib & refreshed package index? Dec 25 18:40:14 quickdev, mabe it's a problem inside my image Dec 25 18:40:24 quickdev, i've an *old* shr image Dec 25 18:41:19 quickdev, and there was something on the planet telling to reflash...i didn't do it Dec 25 18:41:29 but i'll do it now Dec 25 18:41:50 I didn't do it because I don't know to which image it applied Dec 25 18:44:53 but if I've got an errors like this one: * Package=initscripts, 2 matching providers it must be that i have to reflash Dec 25 18:52:56 quickdev: hai Dec 25 18:53:12 Is there something wrong with that package? Dec 25 18:53:30 just asked Ainulindale about it, let's wait for an answer Dec 25 18:56:32 alright:) Dec 25 18:57:14 dos1: i can somewhat understand glamo-mci.c code. And i can somewhat understand sdhci.c code. In fact, the decisions about parameters (clock, voltage) are coming from the upper layer, that's commond for glamo-mci and sdhci. So, i'm really confused here. Dec 25 18:57:47 quickdev: what about this dir http://shr.bearstech.com/shr-testing/images/om-gta02/ Dec 25 18:58:05 quickdev: it contains kernel+modules Dec 25 18:58:25 quickdev: was it just for testing purposes? Dec 25 18:59:11 BluesLee, here are some images: http://shr.bearstech.com/shr-testing/images/om-gta02/ Dec 25 18:59:17 sorry, wrong url Dec 25 18:59:23 http://shr.bearstech.com/shr-testing/images/neo1973/ Dec 25 19:00:07 quickdev: BTW when will this url change ? Dec 25 19:00:20 to be coherent with unstable ? Dec 25 19:00:38 on the next build I think Dec 25 19:00:54 nice :) Dec 25 19:01:03 hello and merry christmas BTW Dec 25 19:01:25 ronny: are you the same ronny which is on #vala ?$ Dec 25 19:04:34 PaulFertser_: ok, i booted fr with glamo_mci.sd_max_clk=1000000 Dec 25 19:04:37 with nor uboot Dec 25 19:06:34 PaulFertser_: ok, i'm testing ;] Dec 25 19:06:41 find /media/mmcblk0p2/ Dec 25 19:08:30 it looks like it's working Dec 25 19:08:44 quickdev: can you have a look on the digital clock in the next days ... Dec 25 19:09:41 PaulFertser_: whole debian was listed and there was no i/o errors Dec 25 19:09:42 :) Dec 25 19:10:05 so now i'm listing other partitions Dec 25 19:12:03 it's working, it's working!!! Dec 25 19:13:13 BluesLee, yes Dec 25 19:13:51 quickdev: i think i will add the digital clock on lock (aux) Dec 25 19:14:00 dos1: Ha, that's strange. Either the maximum clock is too high for this card, or the voltage is too low... I know how to get the maximum clock the card advertises, but i don't know how to get the voltage yet. And moreover you don't have a device to check which values are actually working good with the card :-/ Dec 25 19:14:46 PaulFertser_: this is sdhc class 4 card Dec 25 19:15:13 at last it's working Dec 25 19:15:22 thanks for trace :) Dec 25 19:17:56 dos1: but it doesn't work with Qi and it doesn't work out of the box with Linux. That should probably be fixed. I think if you could provide an information from a driver working good without fine-tuning (e.g. sdhci), the real issue can be traced. Dec 25 19:20:41 hello,qi doesn't seem to boot lastest SHR... Dec 25 19:21:18 what should I do ? can I reflash uboot if i have all my uboot parameters gone? Dec 25 19:22:55 using the nor bootloader gives: Dec 25 19:23:03 no apmd Dec 25 19:23:12 no fbset Dec 25 19:23:36 no fso-gpsd Dec 25 19:23:49 all theses binaries are lacking...so it doesn't start Dec 25 19:24:27 and ssh didn't start: ssh: connect to host 192.168.0.202 port 22: Connection refused Dec 25 19:24:36 so I can't use the image... Dec 25 19:26:21 dos1: btw, sdhci doesn't even support 2.7V voltage! I think it can support 1.65-1.95 and 2.9-3.4. So it might be those extra 0.2V that matter. Dec 25 19:26:34 GNUtoo: how come? Could you describe in detail what's wrong with Qi? Dec 25 19:27:35 GNUtoo: distribution shouldn't matter to it. The algorithm is clear and simple. What UI indications do you have? Red LED, vibration? Dec 25 19:27:47 PaulFertser_, i'll reboot and tell what's happening Dec 25 19:28:40 s3c2410_udc: debugfs dir creation failled -19 Dec 25 19:28:59 how long should I wait? Dec 25 19:29:21 dos1: i think i can show you how to alter the kernel (for testing purposes) to try that idea of low voltage (i.e. how to provide your sd card with 2.9V and see what happens). BTW, are you sure it's not working with the latest Qi? It uses 3V without any switches. Dec 25 19:30:00 PaulFertser_: i will try latest Qi Dec 25 19:30:00 GNUtoo: Are you booting from SD card? Then please add loglevel=8 to /boot/append-GTA02 Dec 25 19:30:21 PaulFertser_, unfortunately not...from flash Dec 25 19:30:40 in shr, everything in settings is just marked with Unknown, also gps daemon doesn't start Dec 25 19:31:35 Starting freesmartphone.org gpsd compatibility daemon: start-stop-daemon: invalid option -- P Dec 25 19:32:04 I opened the file and tried to remove -P flag, which didn't fix Dec 25 19:32:25 GNUtoo: so, you boot it from flash and your SHR is also on flash. And it boots with the same kernel from NOR u-boot. But no matter how long you wait, it doesn't boot with Qi. And you're sure that you're trying to boot from flash, and you don't have any bootable (with /boot/uImage-GTA02) partitions on your SD card? Dec 25 19:32:57 PaulFertser_, i removed the sd-card Dec 25 19:34:40 GNUtoo: i think you can write the kernel to a first partition on the card and use /boot/append-GTA02 trick to add rootfstype=jffs2 root=/dev/mtdblock6 and loglevel=8 to try to boot from flash with debugging info. Dec 25 19:35:00 GNUtoo: or use a debugging board for serial console ;) Dec 25 19:35:29 PaulFertser_, lol...i don't have debugging board...i'll try from sd Dec 25 19:39:11 dos1: in case you're going to try the kernel with higher SD voltage, here's how to do it: in drivers/mfd/glamo/glamo-core.c remove lower voltages from ocr_avail, let it be only 3.0-3.3 ;) I know some IC don't like high frequencies with low voltages, and it might be your card is buggy in this way. Just a wild guess though. Dec 25 19:48:42 Ainulindale? Wiki Build Instruction refer to: http://shr.bearstech.com/Makefile <- but that's not the SHR makefile -> thus, the FSO common/ directory is downloaded Dec 25 19:52:08 PaulFertser_, lol is fsck activated in fstab? Dec 25 19:52:14 whats the gps module name? Dec 25 19:52:23 quickdev: how do I make an opaque type in C ? Dec 25 19:52:23 shr doesn't find the gps Dec 25 19:52:45 PaulFertser_, because i've Ext2-fs warning: mounting unchecked fs,running e2fsck is recommended Dec 25 19:52:54 ah no it continues to boot Dec 25 19:52:59 jonkristian: yep fso-gpsd is borked Dec 25 19:53:18 ouch...mabe filesystem corruption Dec 25 19:53:21 bumbl: so, is there a workaround? Dec 25 19:53:22 GNUtoo: actually, no. It depends on bootscripts. And for journaled filesystem the repair is activated automatically when mounted read-write. You can also force the check on some distributions with "> /forcefsck" and rebooting. Dec 25 19:53:32 don't know Dec 25 19:53:48 ok thanks a lot Dec 25 19:54:06 ptitjes, that means? Dec 25 19:54:53 ptitjes: do you have used my icon or you still hate me? Dec 25 19:55:08 Hire: I don't hate you!! Dec 25 19:55:28 but isn't it too much neo! Dec 25 19:55:30 :) Dec 25 19:55:37 (you will hate me!) Dec 25 19:55:46 quickdev: data encapsulation in C Dec 25 19:55:54 but I think I found something in google Dec 25 19:56:05 ok ;) Dec 25 19:56:08 i've things such as wrong crc Dec 25 19:56:11 (hiding the content of a struct to its users) Dec 25 19:56:31 ptitjes, have a look at elementary. There it's done with Widget_Data afaik Dec 25 19:56:34 weird :> Dec 25 19:58:13 GNUtoo: may be you should reflash? **** BEGIN LOGGING AT Thu Dec 25 20:01:16 2008 Dec 25 20:05:03 oups did we have a split ?? Dec 25 20:06:18 ptitjes: yes Dec 25 20:07:16 fun this was years I did not see that :) Dec 25 20:09:33 dos1: Do you want to test a kernel with higher voltage? Dec 25 20:09:56 yes, Qi and u-boot is ok, the newest kernel works only with lowered frequency (oldest didn't work at all) and usb cardreaders work ok Dec 25 20:10:39 PaulFertser_: if you send me compiled one, i will test it :D Dec 25 20:11:08 dos1: I'll do in 20 minutes. Dec 25 20:11:14 PaulFertser_: ok Dec 25 20:14:19 dos1: What revision do you have? uname -a please :) Dec 25 20:15:37 PaulFertser_: i have this kernel - http://downloads.tuxfamily.org/admiral0repo/uImage-GTA02.bin Dec 25 20:15:52 when i boot fr, i will tell you what revision it is Dec 25 20:16:30 dos1: I can't guess version from this name, sorry. I hoped to compile the same version you have to avoid compiling and downloading modules. Dec 25 20:17:01 PaulFertser_: i don't have modules installed for it at all ;] Dec 25 20:17:34 i only flashed it for testing Dec 25 20:17:45 dos1: Then i still don't know which recent revision works ok. I use c59a8bdc7ca4b547 (Mon Dec 8). Dec 25 20:18:00 but i'm booting now Dec 25 20:18:23 Ok, so the newest shr image isn't working quite well, what is a good one? Dec 25 20:19:43 PaulFertser_: 2.6.28-GTA02_andy-tracking_e142549d884a17b8-mokodev Dec 25 20:19:46 quickdev: I can't manage this error: redefinition of typedef ‘ThreadContext’ - How should look my typedefs ? Dec 25 20:19:52 The newest? Dec 25 20:19:56 Which newest? Dec 25 20:20:14 LOL Dec 25 20:20:22 Ainulindale: shr-unstable Dec 25 20:20:27 xmas edition Dec 25 20:20:31 :p Dec 25 20:20:38 gps not working with that image Dec 25 20:20:43 fso-gpsd fails to start Dec 25 20:20:44 Well, it's funny to see people installing unstable and complaining about instability =) Dec 25 20:20:58 imo noone complains Dec 25 20:21:08 "complain" is a big word here :-) Dec 25 20:21:10 oh I think I get it! Dec 25 20:21:28 Ainulindale: i am not! Dec 25 20:21:43 just mentioning Dec 25 20:21:46 Ainulindale: the newest image as in the latest one from http://shr.bearstech.com/shr-testing/images/neo1973/ Dec 25 20:21:59 ah so it affects testing too Dec 25 20:22:10 jonkristian: well then it's not new Dec 25 20:22:13 And here it's working well Dec 25 20:22:17 i have a gps problem with unstable Dec 25 20:22:19 Is it instable because of upstream work? :s they need to focus on improving, not dis-improving Dec 25 20:22:23 And I am not complaining, I am simply reporting, and asking for a possible fix. Dec 25 20:22:36 jonkristian: I was talking about unstable there: -) Dec 25 20:22:40 Not testing Dec 25 20:22:40 Ainulindale: :) Dec 25 20:22:46 jonkristian: What's your problem anyway? Dec 25 20:23:50 Ainulindale: i did an opkg update && opkg upgrade and got Package libframeworkd-glib0 md5sum mismatch. Dec 25 20:23:57 Ainulindale: and gps isn't working Dec 25 20:24:37 Does that mean the distro as a whole isn't working or does that mean GPS isn't working? Dec 25 20:24:41 jonkristian: run fso-gpsd and gps will work Dec 25 20:25:08 jonkristian: Please don't dramatize :-p Dec 25 20:25:20 We are aware of the GPS issue Dec 25 20:25:32 Ainulindale: on unstable: tangogps tells me "no GPS found" /etc/init.d/fso-gpsd start/restart fails Dec 25 20:25:39 And for the md5sum mismatch, I have no idea why it does that Dec 25 20:26:17 Ainulindale: I'm sorry if you feel that im dramatising. Dec 25 20:26:27 jonkristian: Don't be sorry Dec 25 20:26:29 I'm just stating Dec 25 20:26:31 run fso-gpsd, no /etc/init.d/fso-gpsd Dec 25 20:26:42 "SHR doesn't work quite well" and "GPS doesn't work quite well on latest SHR" is different :-) Dec 25 20:26:47 ptitjes, can't help you with that, sorry ;) Dec 25 20:26:51 One annoys me, the other annoys me more =) Dec 25 20:26:58 (All in all I'm annoyed quite easily) Dec 25 20:27:49 Obviously :) Dec 25 20:28:15 dos1: ah yes that solves Dec 25 20:28:16 it Dec 25 20:29:03 Ainulindale? Wiki Build Instruction refer to: http://shr.bearstech.com/Makefile <- but that's not the SHR makefile -> thus, the FSO common/ directory is downloaded Dec 25 20:29:24 I know Dec 25 20:29:29 I would never complain about something not working if i was using testing, i am not stupid. lastest shr image is 16-Dec-2008 23:59 , is that not understandable? Excuse me for the argument here... Dec 25 20:29:49 jonkristian: well you're using testing Dec 25 20:29:51 There's no release Dec 25 20:29:55 jonkristian :) Dec 25 20:29:58 There are two types of images Dec 25 20:30:00 Testing, and unstable Dec 25 20:30:13 No release, hence, no supposedly stable image Dec 25 20:30:19 think of them pertatin to the psychiatry terms :) Dec 25 20:30:23 Though one is clearly more usable Dec 25 20:30:43 Ainulindale: yes, that's a thing i don't get;) Dec 25 20:31:11 But I'll stop annoying, for now:D Dec 25 20:31:40 You don't annoy me (or anyone I hope) Dec 25 20:31:50 good:) Dec 25 20:31:52 It's just that these days I see comments such as "SHR is not working" Dec 25 20:31:58 Just because a tiny feature isn't working Dec 25 20:32:02 It's tiring :-) Dec 25 20:32:05 Ainulindale, should I replace the FSO makefile with the SHR one? Dec 25 20:32:17 quickdev: Well, for now you could but we were talking about that with mrmoku_ Dec 25 20:32:24 And someone said he would fix the wiki page Dec 25 20:32:32 Obviously it wasn't done :-) Dec 25 20:32:33 it isn't fixe Dec 25 20:32:36 yeah Dec 25 20:32:43 hehe, that is annoying. I never report unless i have proof. It's like saying the car doesn't run, without even trying to turn the key. Dec 25 20:32:43 Sorry if I'm a bit busy by the way quickdev Dec 25 20:32:56 recovering from christmas, it was awful, thus quite tiring Dec 25 20:33:00 But I'm safe and sound at home now Dec 25 20:33:05 Though with a new multi effect Dec 25 20:33:08 So a bit busy :-p Dec 25 20:33:16 Ainulindale, no problem..but we should fix those "blocker" bugs, that are easy to workaround/fix Dec 25 20:33:22 Indeed Dec 25 20:33:38 I'll be more available tomorrow Dec 25 20:33:44 Or maybe later tonight Dec 25 20:34:14 I've got a problem: http://pastebin.com/m5910378f Dec 25 20:34:17 btw, fso-gpsd worked;) so not /etc/init.d/fso-gpsd , it doesn't take the -P flag Dec 25 20:34:34 hey guys, a simple question: Does http://shr.bearstech.com/repo/shr.git contain the changes of quickdev? Dec 25 20:35:07 Yes. Dec 25 20:35:12 dolf1074: by the way Dec 25 20:35:15 git.shr-project.org Dec 25 20:35:16 because I get these errors, if I compile libframeworkd-phonegui-efl and install that package Dec 25 20:35:17 :-) Dec 25 20:35:18 http://pastebin.com/m3f1f2f70 Dec 25 20:35:38 dolf1074: update libframeworkd-phonegui Dec 25 20:35:49 and your lib Dec 25 20:35:55 on the phone? Dec 25 20:35:59 best way to do that is to remove and install agai,n Dec 25 20:36:03 yes Dec 25 20:36:10 ok Dec 25 20:37:21 Ainulindale: btw, the usb connection disconnects much on the unstable image :p Dec 25 20:37:24 Ainulindale, libframeworkd-phonegui-efl, not libframeworkd-phonegui Dec 25 20:37:26 quickdev: It's OK. Found my way on it... :) Dec 25 20:37:39 quickdev: yeah my tab didn't complete that much :-) Dec 25 20:37:56 brb Dec 25 20:38:36 hmm how to set the year via date - shr always tells me that we have 2010 even when i sync with ntp Dec 25 20:38:50 i just want to know whose idea is SHR-Image ULTRA ÜBER LITE Dec 25 20:39:07 every time i visit that page i cannot stop laughing Dec 25 20:39:45 oh I should move to the new git... Dec 25 20:40:12 admiral0: well check the history Dec 25 20:40:28 Ainulindale: nice we have our own domain Dec 25 20:40:43 Ainulindale: What kind of multi effect? Zoom? ;) Dec 25 20:40:50 PaulFertser_: Boss GT 10 Dec 25 20:41:09 admiral0: even though I know who did it but won't admit it :-) Dec 25 20:41:35 Can anybody else get a gps sat fix with shr - testing ? Dec 25 20:41:42 Ainulindale: House did it, isn't it? Dec 25 20:41:48 dolf1074: I bet that's him Dec 25 20:41:50 Ainulindale you're a genius Dec 25 20:42:02 Sigh. Stating the truth like that. Dec 25 20:42:19 :D Dec 25 20:42:20 * Ainulindale goes in front of the wind to make his cape flutter Dec 25 20:42:27 lol Dec 25 20:42:30 But seriously Dec 25 20:42:37 You may be the first guy to laugh at one of my joke ever. Dec 25 20:42:52 this joke was nice Dec 25 20:43:09 For once. Dec 25 20:43:13 Well anyway Dec 25 20:43:28 I have to install some stuff to play with my multi effect (SHINYYY !) Dec 25 20:44:13 btw, if i flash shr-unstable will i regret it very much? Dec 25 20:44:19 Ainulindale: do you think that the current settings-app is good (regardless of the lack of functionality just the way how the ui is done) Dec 25 20:44:25 Ainulindale: congratulations, the official description is very promising. Happy rocking to you! :) Dec 25 20:44:41 Ainulindale: the software stuff? For standard OSes? Dec 25 20:45:10 bumbl: No, not at all. Dec 25 20:45:21 I think someone should work on that and that person will be... YOU! Dec 25 20:45:24 Thanks for playing :-) Dec 25 20:45:37 PaulFertser_: I have a Mac Mini as a media player on the TV Dec 25 20:45:46 And I may try with wine Dec 25 20:46:00 admiral0: Not that much, but don't expect everything to work out of the box Dec 25 20:46:16 nah I just had a look at the code and thought it would be easier to rewrite as to modify Dec 25 20:46:36 that's enough for me to get it on Dec 25 20:46:36 yes that's what I meant Dec 25 20:49:24 dos1: https://paulfertser.is-a-geek.org/files/uImage-moredrivers-GTA02_a-t_e142549d884a17b8-HIGH-VOLTAGE.bin Dec 25 20:49:58 dos1: As a free bonus you should get there a "netdev" trigger, so you can make your LEDS flash on GPRS activity. Dec 25 20:50:10 hmm well I have to restructure the omoney code and have a look at the fso api Dec 25 20:50:29 PaulFertser_: only gprs, not wifi? :] Dec 25 20:50:32 ok, i will test it Dec 25 20:50:55 dos1: And please provide a dmesg with this one, i want to make sure that the voltage stuff is working as i think it should. Dec 25 20:51:11 PaulFertser_: ok Dec 25 20:51:20 what is this high voltage about? Dec 25 20:51:44 admiral0: about my sd card ;] Dec 25 20:51:56 can someone sent an sms to yourself with the content: 2009 Dec 25 20:52:03 quickdev? Dec 25 20:52:42 admiral0: I've a suspicion that his card can't work reliably with a 2.7V that it advertises. sdhci (and i think USB cardreaders too) can't provide this anyway, they use 3V, i think. That can be the reason the card is working unreliably with FR. But it's just a wild guess. Dec 25 20:53:22 but that wonn't fry sono Sd's? Dec 25 20:53:27 *won't Dec 25 20:53:28 admiral0: Also, Qi uses fixed 3V voltage and works. Not sure about the frequency though. Dec 25 20:53:48 admiral0: mmc stack don't allow to use a voltage that the card doesn't support anyway. Dec 25 20:53:58 perfect Dec 25 20:55:06 i am downloading shr-lite-unstable to test it Dec 25 20:56:03 dos1: you can use netdev trigger as follows: cd /sys/class/leds/gta02-aux:red; echo netdev > trigger; echo usb0 > device_name; echo rx tx > mode; Dec 25 20:58:00 PaulFertser_: with your kernel i have i/o errors immediately after boot Dec 25 20:58:02 :( Dec 25 20:59:24 dos1: can you still provide a dmesg? I'm sorry that it didn't work... Dec 25 20:59:30 does elementary have a file chooser widget yet? Dec 25 21:00:06 bumbl: can you make a test for me ... sent an sms to yourself with the content: 2009 Dec 25 21:00:33 PaulFertser_: ok, give me a moment Dec 25 21:01:52 BluesLee I'll try it as soon as i flash unstable Dec 25 21:02:02 okay Dec 25 21:08:35 ptitjes: yes Dec 25 21:11:15 it works with the default kbd. dont know what i am dooing wrong ... "2008" is okay "2009" is reduced to "200" Dec 25 21:11:54 ronny: nice :) Dec 25 21:12:22 are you programming in vala for the openmoko ? Dec 25 21:17:29 no Dec 25 21:18:03 im mostly anoyed by the crap that happens in openmoko Dec 25 21:18:31 :D Dec 25 21:18:36 untill there is a reasonable sable platform that isnt a pile of crap i wont start developing Dec 25 21:18:44 haha Dec 25 21:18:46 Nice Dec 25 21:19:00 PaulFertser_: http://pastebin.com/me2ab2cb Dec 25 21:19:21 i didn't had i/o errors Dec 25 21:19:24 actually i wanted to start developing some time back Dec 25 21:19:34 but i will test it now Dec 25 21:19:38 but that was around when they killed 2007.x Dec 25 21:19:55 for enlightenment Dec 25 21:20:14 about 1 year after i asked why gtk instead of e Dec 25 21:20:34 im not happy Dec 25 21:21:31 mostly cause of the large collections of stupid decissions Dec 25 21:22:08 PaulFertser_: previously i was booting different entry from u-boot, maybe i made mistake and it booted different kernel Dec 25 21:22:26 PaulFertser_: oh, i have i/o errors now Dec 25 21:23:33 have anyone tried fennec a2 yet? Dec 25 21:23:35 (on the moko) Dec 25 21:23:40 now my main point of unhappyness is fso Dec 25 21:23:51 the api reminds me of java Dec 25 21:23:56 PaulFertser_: http://pastebin.com/m2270f574 Dec 25 21:24:09 there's already an angstrom pack... ;] Dec 25 21:24:18 and there never was a time that anything resembling java was not a pain for me Dec 25 21:24:38 oh nvm, the angstrom pack is for arm 7 Dec 25 21:25:07 also im more & more convinced that android will win Dec 25 21:25:23 admiral0: okay, you may test it. the sms tool cuts the last entered key, i am on the testing branch ... Dec 25 21:25:47 i am on unstable Dec 25 21:26:08 admiral0: you may test it anyways Dec 25 21:26:21 starting Dec 25 21:26:33 i was wondering if something is wrong with my kbds Dec 25 21:26:47 I was never aware it was a contes :P Dec 25 21:26:52 contest Dec 25 21:27:21 Dave: more than one platform -> autoconstest Dec 25 21:27:23 ronny let's hope not Dec 25 21:27:27 Ainulindale, I'm sorry I'm still bugging you on this one, though I think you noticed another user on the ML "complaining" about ophonekitd crashing on wrong frameworkd behavior. I really think it's the time to start thinking about the while (true) fix ;] Dec 25 21:27:33 heh Dec 25 21:27:38 android is little better qt extended Dec 25 21:27:39 dos1: Ok, i see. It's not the voltage then. You can find a maximum frequency that still works with your card. But it's still strange that it doesn't work with full speed. Can you post the results of the tests you did to the kernel mailing list? Dec 25 21:28:06 admiral0: not yet, too much leaks there Dec 25 21:28:30 PaulFertser_: ok, i will do it, but no today Dec 25 21:28:50 admiral0: i thought android would be ready to use in december Dec 25 21:29:03 ronny: that is a pity... come on and build something better! Dec 25 21:29:04 dos1: I'm sorry we did so much today but gained too little... Dec 25 21:29:16 ptitjes: i lack time qand interest Dec 25 21:29:47 ok :) Dec 25 21:29:59 so far all that the openmoko stuff convonced me is, that linux fanboys will never get a customer-ready mobile platform done Dec 25 21:30:57 I'm convinced of the contrary :) Dec 25 21:31:02 well Dec 25 21:31:02 damn.. i have to restart.. dbus didn't start Dec 25 21:31:16 there was 4 years of time for the software to mature Dec 25 21:31:38 ronny: and i'm not sure most "linux fanboys" even care about "customer-ready mobile platform" Dec 25 21:31:42 And I'm convinced google made one more iphone crap... with an OSS costume Dec 25 21:31:45 other vendors yank 8 generations of customer ready phones out in that time Dec 25 21:32:06 ptitjes: well, it will be customer-ready Dec 25 21:32:15 ptitjes i agree Dec 25 21:32:18 but it will be crap Dec 25 21:32:20 ronny: but they're all crap comparing to a really free platform. Dec 25 21:32:47 PaulFertser_: if the real free platform is free of reasonable tools who cares? Dec 25 21:33:12 I want my phone to do things that marketers and business man can't even think about... Dec 25 21:33:16 if my phone ends up as fucking compile-stuff toy, it would have been better to buy a cheap arm thing Dec 25 21:33:23 They can't because their only interest is money Dec 25 21:33:25 ronny: I think we have a reasonable tool: gcc. And a reasonable kernel: Linux. And a reasonable userland: GNU. Dec 25 21:33:37 PaulFertser_: oh, well, my pc has that Dec 25 21:33:42 its a fucking phone Dec 25 21:33:52 :P Dec 25 21:33:53 When I was working in computer science, I enough enough had to experience how they take client for fools Dec 25 21:34:02 if pim, sync & stuff wont work its a useless piece of hardware Dec 25 21:34:05 ronny: it's not mobile, it doesn't have wifi, bluetooth and gps, and GPRS connectivity. Dec 25 21:34:06 cool the new ui *_* Dec 25 21:34:28 Yes, Everyone look at the UI and be distracted! Dec 25 21:34:42 :P Dec 25 21:34:45 PaulFertser_: and ? missing vs broken beyond repair - i wont notice the difference Dec 25 21:34:57 :P Dec 25 21:35:11 Of course you won't... if you're an unsuspecting/naive customer :P Dec 25 21:35:26 me naive? Dec 25 21:35:27 ronny: it's you who thinks it's a fucking phone. But it's a mobile platform for many cool applications. Dec 25 21:35:28 well, toys are nice Dec 25 21:35:36 but i need a fucking phone that jsut works Dec 25 21:35:58 ronny: you point it! Dec 25 21:35:59 ronny: so buy nokia 3210 or sth like that Dec 25 21:36:11 ronny: it's a fucking phone that just works Dec 25 21:36:17 ronny: get a Nokia 1100 then. You need a phone, get it. Standby time is great. And it's very stable. Go for it. Dec 25 21:36:19 :P Dec 25 21:36:29 and om is like, "oh, you want a nice phone, guess ill start with assraping ya with a shootgun" Dec 25 21:36:38 o_O Dec 25 21:36:46 ronny: it's all your imagination, Dude. Dec 25 21:36:47 I wasn't aware Sean rolled like that... Dec 25 21:36:57 Ronny, take the red pill :P Dec 25 21:37:01 I want my phone to have real features and not being in front a thing that always make me say but why didn't they implement that another way Dec 25 21:37:13 PaulFertser_: there was 4 years of fucking time, and even basic pim wont work Dec 25 21:37:22 PaulFertser_: so where the heck is my nice phone Dec 25 21:37:59 i dont care for the extra toys if the basic essential functionality is just missing Dec 25 21:38:21 pointless discussion :) Dec 25 21:39:04 ronny: well I'll have an inference engine in my phone - maybe you can't even understand why I want that... Dec 25 21:39:07 ronny: "other vendors" have orders of magnitude more paid manpower Dec 25 21:39:25 And they make crap. Dec 25 21:39:33 and/or they license the software platform from someone who has Dec 25 21:40:13 mpr: other vendors dont change a never actually working base platform like every 6 months Dec 25 21:40:25 zing! Dec 25 21:40:29 ronny: FR is just a year old (if you take mass-production). So, your 4 years are not corresponding to the reality. Dec 25 21:40:30 ;p Dec 25 21:41:03 PaulFertser_: fr wanst when they started with the software platform Dec 25 21:41:24 also there where hardare platforms before fr Dec 25 21:41:44 so dont try to make the crap look any more nice Dec 25 21:42:01 I agree that om has had no idea what to do with the software, they should've just focused their resources to stable hw and low-level sw all along Dec 25 21:42:20 ronny: the OM headquaters have a very strange vision and opinions. It's true they abadoned a semi-working platform for semi-working qtopia. And made a lot of other useless moves. We should count on the community, i think. Dec 25 21:42:29 Shoot the current staff and hire a whole new team? :P Dec 25 21:43:04 Dave: at least i want to see the faces of the so-called "UI design team" :-E Dec 25 21:43:06 but I don't find it at all surprising that with these resources it takes 5 years to get something together Dec 25 21:43:12 Yeah :P Dec 25 21:43:22 mpr, pretty sad Dec 25 21:43:47 well, at this point i think even killing om and starting a entirely new thing with a few reasonable educated people will get it done in a few months Dec 25 21:44:05 also i dont particulary care for the linux/obenebedded platform Dec 25 21:44:22 you don't care Dec 25 21:44:27 its a learning process Dec 25 21:44:29 eh, feel free to do that if you think it's so easy Dec 25 21:44:45 dos1: Happy holidays, btw :) Dec 25 21:44:48 mpr: its not easy, and im not smart enough Dec 25 21:44:53 and one has to accept it. in the end om provides us an opportunity Dec 25 21:45:08 im just smart enough to see the current shit is pretty fucked up crap Dec 25 21:45:19 but you're smart enough to produce a valid estimate of the difficulty of the task?-) Dec 25 21:45:31 he should be Dec 25 21:45:44 mpr: im just smart enough to estimate its above my head to pull this off Dec 25 21:45:55 PaulFertser_: thanks, happy holidays for you too :) Dec 25 21:46:00 mpr: Well, it's easy, because you would more or less utilise the knowledge and parts of the software created by OM all this time. Even if you claim you start from scratch. Dec 25 21:46:10 i need at least 7-10 years more of experience in the area Dec 25 21:46:18 is gsm audio quality is fixed i am fine, thats my personal view Dec 25 21:46:22 ^if Dec 25 21:46:42 dos1: at /sys/bus/mmc/drivers/mmcblk/mmc1:b368 i have several interesting files: cid, csd, scr. It can probably help if you include their contents in your report. Dec 25 21:46:49 making a functional smartphone from scratch in a few months sounds like a completely off-the-mark estimate to me :) Dec 25 21:46:50 right now im just a student with deep intrest in achitectura things Dec 25 21:47:06 even apple took years and they had 100x the resources of om Dec 25 21:47:24 thats ehough to estimate the current stuff is crappy shit, but not even nearly enough to produce higher quality Dec 25 21:47:29 mpr.. don't mention that shit Dec 25 21:47:46 it was all marketing Dec 25 21:47:47 mpr: but apple released something that qorked well to the public Dec 25 21:48:02 ronny: that's where you're wrong. Dec 25 21:48:12 ronny: almost no geeks like their shit. Dec 25 21:48:17 yes, after years of development w/ more resources Dec 25 21:48:20 no innovative things, just old things that they made seem new Dec 25 21:48:43 the semi solution of om with x+qtextended is one thing i cannot understand Dec 25 21:48:53 why they didnt switched yet Dec 25 21:49:04 the good thing about oss is that when we do get a working software stack done it's reusable Dec 25 21:49:08 they are bug fixing qt stuff Dec 25 21:49:19 so cranking out new linux-based phones is much faster than this first time around Dec 25 21:49:28 PaulFertser_: i dont get your point, please rephase Dec 25 21:49:37 freesmartphone.org: 03mickey 07framework * r42bda5dbf045 10/framework/ (netlink.py services.py): add netlink constants Dec 25 21:50:13 ronny: Nobody likes iphones, you see. It didn't work. At least for the geeks. They didn't produce anything we liked. Dec 25 21:50:39 PaulFertser_: but they produced something that worked Dec 25 21:51:00 from my point of view thats more than om ever had Dec 25 21:51:30 ronny: i dont see your argument here ... they put much more professional effort and money into the iphone Dec 25 21:51:33 PaulFertser_: once again, you transfer your view of the world upon the entire community of Openmoko purchasers and interested parties. Dec 25 21:51:33 ronny: do you have a fr ? and did you try all the distros ? Dec 25 21:51:34 ronny: But most of us don't care about that much. In some sense, it doesn't work. I was almost happy with a 1100 and i don't need an iphone at all. It doesn't work for me. Dec 25 21:51:40 if they produced something that didn't work... well no ONE would have bought them Dec 25 21:51:50 ptitjes: most of em Dec 25 21:51:54 mwester: Sure, that's my weapon ;) Dec 25 21:51:57 PaulFertser_: YOU don't care about that much. Others may feel different. Dec 25 21:52:10 mwester: Oh, you got me mwester :P Dec 25 21:52:28 ronny: didn't you find a distro from which you can call ? (say e.g. SHR ?) Dec 25 21:52:41 you want basic functionality you have it! Dec 25 21:52:49 Ainulindale: still here? Dec 25 21:52:55 And YOUR short experience with the OM device, AFTER it was re-marketed by OM, is NOT representative of the community's experience -- others have every right to be frustrated with OM's representation of the product, and thier failures to deliver on promises. Dec 25 21:52:55 PaulFertser_: if i buy a smartphone that barely manages as much as a 8 year old nokia in terms of communication, im rather disappointed Dec 25 21:53:08 haha, true Dec 25 21:53:12 mwester: Now i give up to ronny ;) mwester, do you have any ideas, why an SD card that works ok in all cardreaders and initializes well on FR starts to produce glamo-mci glamo-mci.0: Error after cmd: 0x4302 and such? Dec 25 21:53:14 ronny: what doesnt work for you? as i said: if the gsm buzz is gone i am fine with the neo Dec 25 21:53:39 BluesLee: basicall calling works, but any extended things like pim dont Dec 25 21:53:52 and thats what i really need from a smartphone Dec 25 21:53:54 ronny: there is no pim in fso/shr yet Dec 25 21:54:00 else a 8 years old nokia would do better Dec 25 21:54:09 ronny: flash qtextended and you are fine with pim Dec 25 21:54:12 PaulFertser_: I would suspect most probably the drive strength set on the glamo, or (possible, but less likely) the voltage to the card may be off. Dec 25 21:54:40 mwester: We actually tried to fix the voltage at 3V, it didn't work. It works on FR with a lowered frequency though. Dec 25 21:55:17 BluesLee: also i didnt get the "free" phone just to let nokia/trolltec claim my ass Dec 25 21:55:31 hmm... so is it the lowered freq or a side effect I wonder? (i.e. some timeouts may be longer with a slower clock) Dec 25 21:55:56 ronny: yeah, thats an argument ... pim will be included "soon" Dec 25 21:56:00 ronny: qt extended is GPL'd Dec 25 21:56:02 ronny: Qt Extended is just as free. Dec 25 21:56:20 dos1: glamo_mci.sd_drive=2 (IIRC) could produce another results, as mwester suggests. Dec 25 21:56:25 gpl is not that "free" Dec 25 21:56:33 ronny: but i hope you see all the possibilities which are missing in an iphone ... Dec 25 21:56:43 BluesLee: fo course Dec 25 21:56:56 BluesLee: but i lack basic functionality Dec 25 21:57:03 and thats a shame Dec 25 21:57:22 I'd still mess with drive strength, if capacitance is higher than normal, both a lower clock and a higher power on the clock (ie drive strenght) will work. Dec 25 21:57:45 mwester: it is working some time after boot, and then (after ~2 minutes) it prints i/o errors Dec 25 21:58:00 mwester, PaulFertser_: i will try sd_drive=2 Dec 25 21:58:19 PaulFertser_: I've seen SD cards randomly die. Dec 25 21:58:25 ronny: maybe you are right but in the end we cannot change it within some days ... Dec 25 21:58:26 It's been HOW many years, and Apple still doesn't know how to add Search to the iPhone? Dec 25 21:58:26 BluesLee: and i am DAMN FUCKING ANNOYED since they switched to e years after i asked them why gtk instead of e Dec 25 21:58:29 hah. Dec 25 21:59:43 ronny: oups they seem to have knocked your suceptibility, so... :( Dec 25 21:59:55 mwester: Actually, i'm interested in OM from the time GTA01 was released. And i always assumed that the geeks will finally get what they want with it -- flexible UI and hackable software. In fact, i assumed it from the beginning. They hired some great Free Software hackers, they did many things in open, etc. I thought they're going to please hackers, not businessmen Dec 25 22:00:04 openmoko: how exactly die? Dec 25 22:00:15 ronny: i guess they had their reasons ... Dec 25 22:00:37 PaulFertser_: low quality SD cards don't have great resiliency. Dec 25 22:00:37 BluesLee: it took them jsut years to realize they fucked up badly Dec 25 22:00:49 Also after a few hundred R/Ws they sometimes stop functioning Dec 25 22:00:53 ronny: and also switching to qtopia gsm is understandable if you dont have alternatives ... Dec 25 22:00:58 openmoko: it's new kingston sdhc Dec 25 22:01:07 openmoko: But this particular card of dos1 is working ok in USB cardreaders. Dec 25 22:01:22 PaulFertser_: Oh, I thought you meant the one on the ML Dec 25 22:01:35 dos1: Yeah, I had a similar problem. I just bought a different card. Dec 25 22:01:47 ronny: yeah they/we lost at least 12-18 months here Dec 25 22:02:36 BluesLee: yeah, and thats most likely the difference betwen good working smartphone, and the crap we have atm Dec 25 22:02:50 mwester: btw, do you know that sdhci doesn't support 2.7V at all? It's 2.9-3.4 or 1.65-1.95 as far as i understand. Dec 25 22:02:59 BluesLee: and there was more of that crap Dec 25 22:03:10 just think how fso was founded Dec 25 22:03:16 ronny: how? Dec 25 22:03:34 3rd party started it, om brought it in later Dec 25 22:03:55 at least they realized they need it Dec 25 22:04:00 ronny: Hopefully the GTA03 will fix it Dec 25 22:04:12 openmoko: the card on the mailing list is in unknown state -- the reporter says it works as a floppy, he seems to be clueless. Dec 25 22:04:25 PaulFertser_: Didn't know that. But shouldn't such a card negotiate a supported voltage in the first place? Dec 25 22:04:28 PaulFertser_: hahaha. Dec 25 22:04:53 openmoko: how will a hardware platform fix software issues? Dec 25 22:04:56 mwester: Suppose a buggy card advertises 2.7V as the lowest it supports. Dec 25 22:05:18 mwester: FR provides it with that, but sdhci does 2.9V instead. Dec 25 22:05:30 As far as i understand. Dec 25 22:05:40 openmoko: also, if they just start to ignore gta02 like they did with gta01 when the fr got out, om wil lbe dead for me Dec 25 22:05:53 yeah, let's hope GTA03 is better, so we all can spend our money on it. Dec 25 22:06:02 PaulFertser_: Yep, I suspect you are right -- we should not trust the cards at all (low quality SD cards, in my experience, are truely *LOW* quality!) Dec 25 22:06:58 PaulFertser_: I wonder if we need to select a far more conservative set of defaults, then offer a user-space utility to "tune" them via a set of benchmarks or tests (and set the correct values post-boot). Dec 25 22:08:05 After all, a lot of the values selected are to benefit power consumption and the GPS. Dec 25 22:10:35 mwester, PaulFertser_: with sd_drive=2 there are i/o errors too Dec 25 22:10:49 mwester: Unfortunately, in this particular case fixing vcc at 3V doesn't help. Let's see what drive strength will give. But it's already tunable via module parameter, and... unfortunately... Dec 25 22:12:36 mwester: in this particular case the partition table is read successfully, btw. Dec 25 22:12:55 it reads whole card successfully Dec 25 22:13:16 dos1: with a lower frequency? Dec 25 22:13:22 but some time after boot it violently hangs Dec 25 22:13:24 no Dec 25 22:13:29 all the time Dec 25 22:13:47 dos1: and only with 2.7V, not with 3V? Dec 25 22:13:55 with lower frequency and 2,7 it works all the time Dec 25 22:14:15 in other cases - only some time after boot Dec 25 22:14:49 mickey|zzZZzz: merry Christmas Dec 25 22:21:08 PaulFertser_: when i'm using sd_max_clk=1000000 it works good Dec 25 22:22:16 I still have the md5sum mismatch...i'll rm /var/lib/opkg Dec 25 22:22:45 dos1: and what about higher frequencies? 1e7? Dec 25 22:23:05 PaulFertser_: i didn't tested it yet Dec 25 22:23:13 maybe i will try from command line Dec 25 22:23:25 because changing it from u-boot is tiring... Dec 25 22:29:57 just to put it in perspective, while 1MHz SD clock is terrible, it is only 1/2 the max speed the SD card will run when the GPS is powered up... Dec 25 22:30:53 mwester: i have gta02v5, if it matters Dec 25 22:31:17 Doesn't matter; kernel will do that regardless of hardware revision. Dec 25 22:31:55 i read that running some clock speed is unstable on v5, and may be stable on v6 Dec 25 22:32:10 Ainulindale: the changes in libframeworkd-glib aren't in the repo of unstable yet? Dec 25 22:33:13 Ainulindale, i still have the problem with libframeworkd-glib0 Dec 25 22:33:21 mwester: but it was about higher frequency Dec 25 22:34:16 Is there a way to still get a compiled version of libframeworkd-glib, because the newer libframeworkd-phonegui-efl won't work with on older libframeworkd-glib Dec 25 22:34:29 mwester, PaulFertser_: it seems to be stable on 10000000 (1e7) Dec 25 22:34:48 but i have: Dec 25 22:34:48 [ 495.120000] glamo-mci glamo-mci.0: command timeout, continuing Dec 25 22:34:48 [ 630.495000] glamo-mci glamo-mci.0: command timeout, continuing Dec 25 22:34:50 in dmesg Dec 25 22:34:54 only this two Dec 25 22:35:17 now third Dec 25 22:35:18 [ 895.820000] glamo-mci glamo-mci.0: command timeout, continuing Dec 25 22:37:19 ping Dec 25 22:37:22 /s/i/o/ Dec 25 22:38:13 mwester, PaulFertser_: it seems to be stable on 10000000 (1e7) Dec 25 22:38:17 but i have: Dec 25 22:38:29 [ 495.120000] glamo-mci glamo-mci.0: command timeout, continuing Dec 25 22:38:50 second one on [ 630.495000] and third on [ 895.820000] Dec 25 22:40:51 mwester: it can be overriden, as it's a module parameter (glamo_mci.sd_slow_ratio=8 is default, that's true). I think the units with capacitor fix applied don't really need it, but it should be verified. Dec 25 22:41:15 PaulFertser_: i don't have capacitior fix Dec 25 22:41:33 i had problems with inet, have you said something interesting? :D Dec 25 22:41:45 dos1: Are you sure? Most MP 05 units has it. Dec 25 22:41:55 dos1: No, it seems i can't say anything interesting :-/ Dec 25 22:42:07 * mwester isn't sure that's true. Dec 25 22:42:08 PaulFertser_: it's FR from first batch Dec 25 22:43:03 Anyone with the cap fix pre-installed by the factory has a very recent gta02, relatively speaking. But the point is that the kernel cannot determine if the cap is present or not, which is a bit of a shame. :( Dec 25 22:43:04 dos1: you should better apply it. If you are unsure in you soldering skills, find a person who can do it. It's not really hard. Nowadays almost everybody does SMT rework. Dec 25 22:44:46 mwester: i think a compromise would be to assume it's preinstalled in >= v5 units with a WARNING in dmesg. Dec 25 22:45:20 PaulFertser_, realy? is it easy to do the fix for the bad sound ? Dec 25 22:45:38 GNUtoo: it's not about fix for bad sound Dec 25 22:45:39 GNUtoo: no, fixing the buzz is harder. Really harder. Dec 25 22:45:56 PaulFertser_, not the buzz...the audio with headphones Dec 25 22:46:17 hmm Dec 25 22:46:21 GNUtoo: not hard for a guy who fixes cell phones on a market-place. Dec 25 22:46:39 GNUtoo: It depends. You want to get a decent bass with headphones, yes? Dec 25 22:46:40 PaulFertser_, because i'd realy like to be able to use my freerunner as an ogg player Dec 25 22:46:50 PaulFertser_, yes i want that Dec 25 22:47:00 i have my fr on warrianty... maybe they will replace it to v7... Dec 25 22:47:02 but... Dec 25 22:47:10 i love *this* fr :o Dec 25 22:47:17 PaulFertser_, but i don't know where to give my freerunner to fix it Dec 25 22:47:22 Audio sound is far better on headphones with om, maybe those settings could be moved to shr. Dec 25 22:47:42 jonkrist1an, it's an hardware problem... Dec 25 22:48:06 i already tried the alsamixer fix but it's not sufficent... Dec 25 22:48:23 GNUtoo: you basically have 2 easy options and one hard. The first easy one: lift the can, short the output caps and solder a pair of 100uF in your headphones jack. The second easy option is to change the caps from 1uF to 4.7uF (they'll fit). Dec 25 22:48:47 GNUtoo: In the second case you should be able to compensate with the equalizer. Dec 25 22:48:56 I just remember i've had terrible audio in the headsets, except for one distro, which i think was om where the audio was quite good. Dec 25 22:49:41 i'll be back soon Dec 25 22:49:56 GNUtoo: And the third one: you change the caps inside the FR, but there's no place inside the can (IIRC) so you have to solder them somewhere else with a shielded wires to minimise risk of RF interference (AFAIK). Dec 25 22:50:03 GNUtoo: choose your poison ;) Dec 25 22:50:52 GNUtoo: i think you can find a cell phone repair workshop nearby. If you'll provide them with decent instructions, they'll do it. Dec 25 22:51:58 GNUtoo: Oh, you can also get a pre-amp with high input impedance and carry it with you ;) Dec 25 22:52:12 So, some more recent gta02 has the audio hw fix? Dec 25 22:53:38 If so, is there any way to determin? Dec 25 22:54:59 jonkrist1an: audio is semi-fixed (4.7uF) in v7 only. And it's not in production yet, AFAIK. Dec 25 22:55:24 semi-fixed <- hardware wise? Dec 25 22:55:33 That really doesn't make any sense Dec 25 22:56:19 lol Dec 25 22:56:48 jonkrist1an: it's still not enough as it still filters too much bass. But you'll be able to somewhat compensate that with equalization (hardware, i suppose). For an unknown reason OM decided to avoid retracing the said part and choosed to place the biggest capacitor that fit. Dec 25 22:57:43 I'm eventually using bluetooth for audio transmitting, so I guess it wouldn't make alot of difference. Dec 25 22:58:50 jonkrist1an: good luck with BT ;) BTW, i've tried some BT headphones with a samsung player. I can't say i liked the sound. Dec 25 22:58:53 PaulFertser_, the pre-amp was discussed on the hardware mailing list but...i didn't have recent news on it...mabe they are busy with the buzz Dec 25 22:59:37 GNUtoo: that's because there's nothing to discuss. Either you do the fix one way or another, or use pre-amp, or leave it as it is. Nothing to discuss, hardware-wise. Dec 25 23:00:07 PaulFertser_, they wanted to do some batch for it... Dec 25 23:00:12 PaulFertser_: just tested a bt dongle for it on my laptop, both with AKG and some other earbuds, i must say the sound is pretty good. Haven't gotten it to work on the fr yet though. Dec 25 23:01:18 GNUtoo: I'm not sure they had any plans. It's more likely just an idea. And you'll need to power it somehow... Dec 25 23:01:30 PaulFertser_, yes Dec 25 23:02:46 GNUtoo: i think shorting caps is good enough. You'll need a 3.5->2.5 converter anyway. Solder the needed caps in it and that's all. Dec 25 23:05:12 PaulFertser_, ok thanks a lot...i'll look for a cellphone repair workshop and i already have 1)the converter 2)something which I don't remember the name that does: microphone,headphone and answer button Dec 25 23:07:40 GNUtoo: most factory converters/adapters are solid, so you won't be able to solder anything in it. You'd better have connectors instead. 2) headset. Please see the correct wirining in the wiki, it depends on the manufacturer and you should probably tune your thing to be FR-compatible. Dec 25 23:08:24 ok Dec 25 23:09:31 GNUtoo: and you should probably prepare. You should know exactly which caps should be shorted and where they are placed. Please read the relevant letter and refer to schematics and PCB placement. Dec 25 23:12:27 ok Dec 25 23:13:33 but first i've to find a phone repair workshop in italy(if it still exists) Dec 25 23:15:57 GNUtoo: in fact, i think almost any electronics repair workshop will fit. Have you disassembled a modern device? It's almost always SMT. Dec 25 23:17:12 GNUtoo: Fortunately, you don't need to do BGA ;) Dec 25 23:18:53 PaulFertser_, ah ok...i'll ask my favorite electronic shop Dec 25 23:20:55 ok because BGA needs special lazer tools...i'll go bye Dec 25 23:26:53 hello there! Dec 25 23:26:58 GNUtoo is gone but i want to clarify anyway. BGA doesn't require laser tools. It requires special pattern (not sure how it's called, maybe template) and a suitable heater. No need for lasers :) Dec 25 23:27:45 when trying to build the kernel for gta01 one gets this error (for a few weeks already): Dec 25 23:27:58 | arch/arm/mach-s3c2440/built-in.o: In function `pmu_callback': Dec 25 23:28:08 build/tmp/work/om-gta01-angstrom-linux-gnueabi/linux-openmoko-2.6.24-oe2+gitrfb42ce6724576fc173faf8abfb04aa2c36d213b7-r0/git/arch/arm/mach-s3c2440/mach-gta02.c:527: undefined reference to `bq27000_charging_state_change' Dec 25 23:28:44 i wondered if gta01 is still "supported" or if everyone lost interest in a working openmoko install on the good old gta01 Dec 25 23:29:01 Ah, that's called "solder masks". Dec 25 23:29:46 unfotunatly i lack the skill to fix this error Dec 25 23:29:58 jonkrist1an: if you're trying to build andy-tracking, you'd better ask on the kernel mailing list. Dec 25 23:30:01 is there any interest in getting the gta01 kernel image to compile again? Dec 25 23:30:16 josch: yes, there should be. Dec 25 23:30:26 great :) Dec 25 23:30:39 i was worried about this issue as it occured for quite some time now Dec 25 23:30:49 i think that kernel-wiase it's up to the community Dec 25 23:30:49 (for gta01) Dec 25 23:30:54 josch: you might mention what branch you're on. Dec 25 23:30:58 josch: in general, i see you're building 2.6.24 kernel. If it's from mwester's repo, ask him here. If it's from official repo, ask on the mailing list. Dec 25 23:31:06 its the same gsm modem - so as suchin theory firmware updates should work Dec 25 23:31:32 Yep, firmware updates work just fine on the gta01 -- I've done one of mine. Dec 25 23:31:40 gps is a lost cause Dec 25 23:31:47 Nope, works well. Dec 25 23:31:51 :) Dec 25 23:31:59 as its closed Dec 25 23:32:16 and most other stuff should track whatver happens for gta02 Dec 25 23:32:17 The GTA01 kernel is build with OABI compat enabled, so we can continue to run the closed gllin libs. Dec 25 23:32:22 except... image size Dec 25 23:32:31 with only 64m flash vs 256m for freerunner Dec 25 23:32:41 Yeah, well, everyone should be using an SD card anyway :) Dec 25 23:32:47 i thnik the images are pushing the "doesnt fit" barrier Dec 25 23:32:50 which basically means you run off sd Dec 25 23:32:58 or you need to go nuking stuff from images yourself Dec 25 23:32:58 IMO, the gta01 currently works *better* than the gta02. Dec 25 23:33:02 ie back to community Dec 25 23:33:22 mwester: its still clsoed Dec 25 23:33:31 Yep. but it's the best we have. Dec 25 23:33:31 change abi's and you're in trouble Dec 25 23:33:33 :) Dec 25 23:33:43 that's why we keep OABI enabled. Dec 25 23:34:17 one little question Dec 25 23:34:21 i meant kernel abi's Dec 25 23:34:24 internal ones Dec 25 23:34:48 nor oabi vs eabi Dec 25 23:34:49 not Dec 25 23:35:07 when gta02 will be abbandoned by om as gta01 now? Dec 25 23:35:13 Ah, yes. but since OM doesn't support the kernel for the GTA01 in the first place, it's pretty easy for the community to continue to add the sym links in /sys that are necessary. Dec 25 23:35:47 mwester: wrt gllin, i've read one guy wrote a tricky linker that supports abi on-the-fly translation exactly for the purposes of running gllin. It worked for him. He described it in great detail, very smart guy. Dec 25 23:35:56 So with enough effort, the community can keep the gta01 alive as long as there are people interested. Dec 25 23:36:07 sure Dec 25 23:36:23 but its community :) - om isnt going to work on the gps blob Dec 25 23:36:26 PaulFertser_: Yeah, but it turns out that it has other problems, and he's (apparently) not continued to support/fix that lib. Dec 25 23:36:32 as they would have to get glln to do it Dec 25 23:36:38 and glln isnt goign to lift a finger Dec 25 23:37:05 but several gta01 bits are entirely in community hands Dec 25 23:37:14 raster: Om dropped support for the GTA01 about 2 seconds after the gta02 shipped -- they abandoned it on 2.6.22, without even the courtesy of an announcement that they would no longer bother themselves with it. Dec 25 23:37:14 mwester: too sad. It's a very interesting hacking project... Dec 25 23:37:16 others will benefit from ongoing mainstream os dev Dec 25 23:37:33 mwester: i know :) Dec 25 23:37:40 raster: So the entire gta01 has been unsupported by OM, the gps is no different. Dec 25 23:37:56 just stating the practical implications Dec 25 23:38:07 ie if kdrive gets an improvment/fix Dec 25 23:38:17 its "supported" implicitly as it has nothing thats hardware-specific Dec 25 23:38:21 raster: you're correct - it's a lot of work, for a very small nubmer of devices.- Dec 25 23:38:27 so an upgrade will "just work" Dec 25 23:38:43 other bits require om to specifically focus on gta01 Dec 25 23:38:50 gps, kernel. Dec 25 23:38:59 battery and pmu too Dec 25 23:39:09 The latter two are the most worrisom. Dec 25 23:39:14 er, worrisome. Dec 25 23:42:59 Oh man, i've just read a report how a guy reflowed a large BGA in his G3 ibook with a propane torch! Dec 25 23:44:58 sry for the delay - the kernel is from the oe branch but this one should be identically to the official repo Dec 25 23:48:01 josch: you are building from oe then? Dec 25 23:48:12 not from the OM kernel git repo? Dec 25 23:49:17 the build is directly pulled from oe git Dec 25 23:52:26 org.openembedded.dev branch Dec 25 23:53:26 Ok. That's where the distros build from, so either FSO, SHR, etc are experiencing failing builds, or you have diverged in some way. Dec 26 00:24:37 hi, josch already mentioned that i couldnt compile a current kernel for gta01. it complains about undefined reference to `bq27000_charging_state_change'. in linux-openmoko.inc this is switched off for gta01 but CONFIG_ARCH_S3C2440 is on - i dont know enough about that but the error is in mach-s3c2440/... - can someone confirm that he can compile a current kernel? Dec 26 00:39:03 :) Dec 26 00:39:24 <[Rui]> wow so many around, hi all! Dec 26 00:39:54 <[Rui]> I could use an autotools hint, from someone who has setup optional dependencies that get added into config.h Dec 26 00:40:07 mwester: are FSO, SHR and friends experiencing faling builds recently? Dec 26 00:40:10 raster: still awake? Dec 26 00:40:19 hei Dec 26 00:40:20 actually Dec 26 00:40:22 just woke up Dec 26 00:40:37 Nice :), I just need to know how you can make a list transparent Dec 26 00:40:39 <[Rui]> I have an AC_DEFINE that doesn't get put into config.h even though it shows up in the configdefs section of config.log Dec 26 00:40:54 list? where? Dec 26 00:41:17 the contacts list of shr Dec 26 00:41:49 (So I mean a etk_tree ;) Dec 26 00:41:51 ) Dec 26 00:41:57 <[Rui]> this doesn't seem to work ok for me: http://pastebin.com/d37908169 Dec 26 00:42:01 ooh etk Dec 26 00:42:03 not sure Dec 26 00:42:10 i expect technically yes Dec 26 00:42:21 but u'd have to modify th scrolled container theme Dec 26 00:42:27 and all the the list items Dec 26 00:43:01 What do you mean by the container theme? Dec 26 00:43:25 a list is in a container (a scrolled view) Dec 26 00:43:40 so if u make list itmes tranlucent Dec 26 00:43:48 then they are only translucent over their conteainer Dec 26 00:43:52 or within it Dec 26 00:44:33 ah okay, so I need to set translucent values. (tried with transparent, but it didn't work :) ) Dec 26 00:46:57 dolf1074: why do you want to make lists translucent? ;>>> Dec 26 00:47:13 To see the background Dec 26 00:47:20 http://alasal.be/openmoko/shr2/contacts/r5/contactlist_r5.jpg Dec 26 00:48:30 dolf1074: are you implementing those mockups? Dec 26 00:48:41 yup (at least trying) Dec 26 00:49:10 I have already the background and the icons next to the 'Jean's Dec 26 00:49:11 good luck :) Dec 26 00:49:31 thanks Dec 26 00:50:12 i must think about what can i do for shr Dec 26 00:50:32 there is enough to do Dec 26 00:50:41 yep Dec 26 00:50:46 porting games :) Dec 26 00:51:12 i must find something up to my skills ;) Dec 26 00:51:31 write new settings app, including wifi and bluetooth settings Dec 26 00:51:35 btw Dec 26 00:51:38 which background? Dec 26 00:51:49 at best u'll get to window bg Dec 26 00:51:52 not desktop bg Dec 26 00:51:59 I need the window bg Dec 26 00:52:02 ok Dec 26 00:52:12 thats all possible Dec 26 00:52:14 then Dec 26 00:52:14 :) Dec 26 00:52:56 beautiful. Les beautiful is that it's 2am Dec 26 00:52:59 byebye Dec 26 00:53:15 dolf1074: is this for shr ? Dec 26 00:53:24 good night Dec 26 00:53:32 arg Dec 26 00:53:37 he's left.. Dec 26 00:53:55 ptitjes: but afaik yes Dec 26 00:54:20 writing new settings app... hmm, good idea Dec 26 00:57:15 ptitjes: wow that would be nice if the ui looks like those mockups!! Dec 26 00:57:37 <[Rui]> ah... found out what my problem was thanks to poutine on #gnu, it seems you really need a description on AC_DEFINE Dec 26 00:58:50 ptitjes: afaik it will look like those (or some newer versions of those) mockups Dec 26 00:59:04 :) Dec 26 00:59:45 i must learn programming in etk Dec 26 00:59:56 maybe python-etk Dec 26 01:00:57 would it be easy for someone, who knows pygtk? ;) Dec 26 01:07:21 * raster suggest elementary - but he's biased Dec 26 01:16:09 raster: i don't know efl as well as you, so your suggestions are helpful to me ;) Dec 26 01:17:04 has elementary bindings in python? Dec 26 01:19:37 hmm... i must see how written is libframeworkd-phonegui-etk and then decide which language and toolkit i will use Dec 26 01:20:17 raster: but elementary-tests looks promising Dec 26 01:21:18 ok, i will think about it later Dec 26 01:22:19 now it's time to sleep... uh, no, debiamkn is still upgrading :/ Dec 26 01:22:35 debian* Dec 26 01:22:56 shitty sd card Dec 26 01:31:28 dos1 libframeworkd-phonegui-efl is written in C Dec 26 01:32:09 dos1: I think you should not think it like this Dec 26 01:33:35 Choose the language you are the more familiar with OR choose the language in which the applications that sits along to your app are Dec 26 01:34:11 ptitjes: i know C... i know python too Dec 26 01:34:43 So go for C (or Vala ;)) if you need it to be coherent with the rest of the apps Dec 26 01:34:59 but app in python for freerunner is easier to develop and test Dec 26 01:35:04 However current settings app is in python Dec 26 01:35:10 :) Dec 26 01:35:28 And there is a binding for both etk and gtk in python Dec 26 01:35:55 i'll look at python-elementary Dec 26 01:36:26 I think it is a good choice to start with Dec 26 01:37:10 (I personnaly choosed Vala-nogui ;)) Dec 26 01:38:28 btw. anybody knows what's about python launcher from optimization team? Dec 26 01:38:32 And I'm happy to annouce that, now for two hours, I calling Prolog from Vala (and thus from C) Dec 26 01:38:42 :) Dec 26 01:39:07 team is gone, but launcher isn't there Dec 26 01:39:31 dos1: no idea what you talk about... :( Dec 26 01:39:42 tichy ? Dec 26 01:40:33 ptitjes: when optimization team started their work, in plans was something called "python launcher" Dec 26 01:41:08 probably a preforking python daemon Dec 26 01:41:14 it would decease a lot starting time of python apps Dec 26 01:41:30 that saves startup time of new python code by importing it into a python process that has already loaded all modules etc Dec 26 01:41:34 cjb: yes, something like this Dec 26 01:41:40 already used by maemo, OLPC, etc Dec 26 01:42:00 http://blogs.gnome.org/johan/2007/01/18/introducing-python-launcher/ Dec 26 01:43:27 so where is it for openmoko? ;) Dec 26 01:45:08 this reminds me when we were sharing JVMs... Dec 26 02:06:03 ok, good night **** ENDING LOGGING AT Fri Dec 26 02:59:57 2008