**** BEGIN LOGGING AT Fri Dec 18 02:59:57 2009 Dec 18 05:23:13 Hi, how to get SHR testing image source code ? Dec 18 06:43:24 satish: what exactly do you mean? Dec 18 06:43:54 mrmoku: he's gone anyway, too impatient apparently :D Dec 18 06:43:57 satish: the OE 'source code' used to build shr-testing or source code of apps included in shr-testing Dec 18 06:44:01 PaulFertser: ohh Dec 18 06:44:18 PaulFertser: hmm... irssi/bip still autocompletes him Dec 18 06:44:28 he already sent me few e-mails about that today :) Dec 18 06:44:36 heh Dec 18 06:44:39 :) Dec 18 06:44:43 morning all :P Dec 18 06:44:47 moin Dec 18 06:46:01 morning :D Dec 18 06:46:24 PaulFertser: please... keep the cold :P Dec 18 06:46:36 it is moving towards over here... and we don't want it ;) Dec 18 06:46:43 sissies Dec 18 06:46:48 yup Dec 18 06:46:58 :P Dec 18 06:47:34 ahh... Dec 18 06:47:38 JaMa: got a mail too now :P Dec 18 06:49:15 morning Dec 18 06:50:14 ok... out in the cold and to the kindergarten... Dec 18 06:50:18 PaulFertser: btw.... with bike :P Dec 18 06:50:26 bbl Dec 18 06:53:44 Cool Dec 18 07:09:51 mrmoku|away: probably because I answered only about not using Makefile at all and that he can checkout shr/testing2009 branch manually :) Dec 18 07:17:58 moin Dec 18 07:48:43 PaulFert`: bip bork? Dec 18 07:54:48 DocScrutinizer51: not really borked, just intermediate networking issues... Dec 18 07:55:13 intermittent Dec 18 07:55:15 whatever Dec 18 07:55:43 PaulFertser: yeah. lousy net weather these days Dec 18 07:56:57 DocScrutinizer51: i know back in the days the weather physically affected the majority of dialup links :) Dec 18 07:59:23 yeah. nowadays it's more like "severe DDoS on freenode. Seacable NA-7H dead. Slight sasser in northern hemisphere" Dec 18 08:22:46 <[Rui]> am I still here? can someone tell me the ip addresses of opendns or google dns? my dns is borked Dec 18 08:22:58 fix it :) Dec 18 08:26:06 8.8.8.8 and 8.8.4.4 Dec 18 08:26:33 <[Rui]> let me rephrase that... my ISP's dns is borked :) Dec 18 08:26:45 <[Rui]> 10x Dec 18 08:33:02 [Rui]: dns is for whimps Dec 18 08:33:13 real men know the internet by IP address alone Dec 18 08:33:23 (and can even give you the routes to get there) Dec 18 08:33:24 :) Dec 18 08:33:24 raster: including ipv6? Dec 18 08:33:46 Defiant-: thats for the new adventurers Dec 18 08:41:43 [Rui]: apt-get install bind Dec 18 08:54:24 raster: there's something else besides IP addr ?? o.O Dec 18 09:04:06 heh. kindergarden time finished? Dec 18 09:04:37 ohh yes :P Dec 18 09:04:44 even some daywork already ;) Dec 18 09:04:58 :) Dec 18 09:05:30 * DocScrutinizer51 yawns and checks coffee stash Dec 18 09:09:34 I am still failing to build shr-u and even rm'ing the tmp directory and starting anew hasnt worked, Dec 18 09:09:56 can someone take a look at http://pastebin.com/mdb1e4b2 and advise what I need to do ... Dec 18 09:13:14 BillK_: have you tried to add a LICENSE = "GPL" to that .bb file? Dec 18 09:15:39 but why is that in your depencie chain at all? that package isn't in any DEPENDS Dec 18 09:16:53 Heinervdm: running now - will see if it works ... its a totally Dec 18 09:17:24 standard build - its in there because something recent must be pulling it in. Dec 18 09:17:50 i greped for it, and there is nothing Dec 18 09:18:05 maybe I need to rm the lot and start again? Dec 18 09:18:24 instead of just tmp I mean Dec 18 09:18:45 make update is enough to clean the openembedded dir Dec 18 09:19:18 you're trying to build shr-image or sth else? Dec 18 09:20:09 make image - all defaults except set local.conf to "1" thread as usually something fails when it gets built out-of-order Dec 18 09:22:02 Heinervdm: LICENSE trick didnt work :( Dec 18 09:22:33 task_shr_minimal_gtk -> openmoko_icon_theme_standard2; Dec 18 09:22:42 there is the depends Dec 18 09:22:50 but it's written in a strange way... Dec 18 09:24:19 I have to go now. Need to visit the dentist Dec 18 09:24:46 ok, I hate dentists :) Dec 18 09:25:56 just a checkup, i hope he will just look at my new tooth Dec 18 09:51:46 hmm OOM just killed fsousaged.. while neo is just waiting at home.. Dec 18 09:52:59 and its because /var/volatile is full again :/ -rw-r--r-- 1 root root 16M 2009-12-18 10:53 frameworkd.log Dec 18 09:53:02 -rw-r--r-- 1 root root 46M 2009-12-18 10:53 phonefsod.log Dec 18 09:53:14 and frameworkd is eating 80% mem because of that :/ Dec 18 09:54:40 maybe we should add logrotate? Dec 18 09:54:48 what does it store there Dec 18 09:54:52 and why the leak? Dec 18 09:55:35 https://www.codeaurora.org/gitweb/quic/xwin/?p=xf86-video-msm.git;a=log Dec 18 09:55:50 playya: iirc phonefsod doesn't have log-level setting yet (imho better than logrotate) and mem leak in frameworkd should be fixed too Dec 18 09:56:17 if we get phone audio to work on 2.6.32 with alsa shr on g1 becomes realistic Dec 18 09:56:22 kernel already boots Dec 18 09:57:06 tmzt: and its stored in /var/log (which is linked to /var/volatile to be a bit nicer to uSD/NAND, but maybe its not worth it on uSD Dec 18 09:57:28 JaMa: shouldn't /var/log survive reboots? Dec 18 09:57:30 ah Dec 18 09:58:06 lindi-: it doesn't but some structure is recreated in some volatile-populate.sh Dec 18 09:58:22 lindi-: actually it doesn't. And it's a question of personal prefs Dec 18 09:58:36 lindi-: with cheap uSD i would disable whole /var/volatile.. Dec 18 09:58:42 have to leave, too. bye Dec 18 09:59:49 JaMa: going to fix that soon Dec 18 09:59:50 lindi-: as I have swap file on uSD then its no use to safe few write with /var/volatile and make more writes because of /var/volatile used 64MB of precious ram Dec 18 09:59:58 JaMa: phonefsod needs loglevels :| Dec 18 10:01:16 DocScrutinizer51: but for testing it's not very useful if the logs are lost on reboot? Dec 18 10:01:42 mrmoku: yes.. and seems like with 2.6.31 i have more log entries than with andy-tracking :/ I've run out of volatile space in up 3 days, 1:27 Dec 18 10:03:24 and probably sooner, because I'm not sure when oom was called and how long the mem leak were eating mem after volatile was full Dec 18 10:03:50 lindi-: quite obviously Dec 18 10:04:51 JaMa: (swap) good point Dec 18 10:08:58 netsplit? Dec 18 10:09:18 JaMa: log rotation should make sure you never run out of space? Dec 18 10:10:28 lindi-: probably not implemented for volatile :-P Dec 18 10:12:31 lindi-: if you run it at least twice a day, volatile space would be enough for at least few weeks :), and if you rotate it outside volatile it would be even better to free some RAM.. Dec 18 10:13:20 DocScrutinizer51: imho it should work on volatile the same as on normal fs as volatile is quite transparent to file operations Dec 18 10:15:39 JaMa: of course. I just thought nobody cared as it's volatile anyway Dec 18 10:17:01 yup.. and still there shouldn't be 60MB of logs generated in 3 days even when compressed with logrotate to 5 Dec 18 10:34:34 morning Dec 18 11:02:33 mrmoku: can you run task-shr-feed after testing finish? Dec 18 11:03:01 mrmoku: c_c packages seem still not compiled and I bumped shr-launcher again few mins ago Dec 18 11:18:15 JaMa: feed is running Dec 18 11:18:22 thx Dec 18 11:27:49 dos1: what should i use for input fields with elementary? Entry isn't that nice Dec 18 11:28:08 you're using anchorbox, aren't you? Dec 18 11:28:16 Heinervdm: no, i'm using entries Dec 18 11:28:24 entries wrapped by frame Dec 18 11:29:05 but an empty entry is too smalll.. Dec 18 11:45:13 g'morning Dec 18 11:47:55 hi all Dec 18 11:57:49 I notice there are ubifs images available, but the install page does not mention them - are they usable ? Dec 18 11:58:02 or is jffs2 still the recommended fs ? Dec 18 11:58:09 yann: i think yes, just you have to format nand i think Dec 18 11:59:18 and what about the .ubi and .ubifs files ? what is the difference between them ? Dec 18 11:59:47 yann, look at shr devel list Dec 18 12:00:01 ok, thx Dec 18 12:00:57 sry shr user Dec 18 12:01:17 yann, http://lists.shr-project.org/pipermail/shr-user/2009-December/002731.html Dec 18 12:02:02 thx Dec 18 12:05:49 ok, sounds like still a bit experimental, I'll stick with jffs2 for now :) Dec 18 12:13:07 http://fireforge.net/scm/viewvc.php/trunk/INSTALL?root=hedgewars&view=markup Dec 18 12:13:16 it would be really great to have hedgewars on freerunner! Dec 18 12:13:35 Heinervdm: so make that entry bigger ;P Dec 18 12:13:36 SHR: 03heinervdm 07opimd-utils * r1bddde91858e 10/opimd-dates: opimd-dates: Dec 18 12:13:45 Heinervdm: like i did in GPRS module for shr-settings Dec 18 12:14:08 dos1: can you have a look at that new page? The frame isn't visible Dec 18 12:14:10 about hedgewars - there is interface for touchscreen devices and iPhone build... so on freerunner it should work :P Dec 18 12:15:14 dos1: then make a recipe :) Dec 18 12:15:29 Heinervdm: i'll try :P Dec 18 12:15:47 Heinervdm: i just discorered that there is touchscreen interface included in code ;p Dec 18 12:16:02 expect for freepascal there should be no problem Dec 18 12:16:14 yup, pascal ;x Dec 18 12:16:56 never seen a pascal app in OE Dec 18 12:18:05 but it's handled by cmake... is there recipe for fpc in oe? Dec 18 12:18:07 * dos1 checks... Dec 18 12:20:24 looks like no... Dec 18 12:20:32 Heinervdm: did you come across a fix for that build problem? Dec 18 12:21:13 BillK_: no, i have no idea what can be wrong, that package builds fine for me Dec 18 12:21:55 must be something to do with a missing package in the build order :( Dec 18 12:22:03 as workaround you can remove it temporarly, it's in tast-shr-minmal at the end Dec 18 12:22:48 ok, will try that - tried to bitbake shr-image and that failed at the same point - eventually. Dec 18 12:23:09 BillK_: that's exatly the same as make image Dec 18 12:23:21 so it should fail at the same place Dec 18 12:25:16 thought make image does a minimal? - it built a LOT of packages before the second failure. Dec 18 12:25:34 bitbake shr-image that is Dec 18 12:26:13 not sure, i just use bitbake Dec 18 12:37:17 Hi all... any calendar application that uses opimd yet? Dec 18 12:41:15 SHR: 03mok 07phonefsod * r0b87b731a202 10/ (data/phonefsod.conf src/phonefsod.c): handle different loglevels to not flood the log with debug stuff Dec 18 12:41:23 beniwtv: only opimd-dates from opimd-utils till now, i will write another one when the elementary calendar widget is released Dec 18 12:41:32 * mrmoku lunch Dec 18 12:41:33 bbl Dec 18 12:43:27 Heinervdm: Well, I'll wait then :) Dec 18 12:44:09 Hmm... syncml is not available :( Dec 18 13:20:36 hi! does someone know how to pass this? Dec 18 13:20:37 ERROR: '/home/sylvain/dev/SHR-Project/shr-unstable/openembedded/recipes/openmoko2/openmoko-icon-theme-standard2_svn.bb' failed Dec 18 13:21:15 ERROR: Exception: Message:need more than 1 value to unpack . Dec 18 13:37:33 GarthPS: you can remove the paclage from the task file, but it still cvrapps out in building the actual image :( Dec 18 13:38:05 hmmm - keybourds a bit funnny ..... Dec 18 13:38:21 see backlog. Dec 18 13:40:23 mrmoku: shr-today should be replaces with libphone-ui-shr? Dec 18 13:41:14 * spaetz looks at recipes to see if that has been marked this way Dec 18 13:42:20 spaetz: yup Dec 18 13:42:31 spaetz: shr-today is integrated in libphone-ui-shr now Dec 18 13:42:47 one serpentic piece less ;) Dec 18 14:09:37 SHR: 03heinervdm 07opimd-utils * rcec1da6c8eca 10/opimd-dates: opimd-dates: remove all uses of partial, add replacement for calendar widget (to make switch easier) Dec 18 14:51:39 question: where are the SMS saved to? Dec 18 14:51:50 ^^ Dec 18 14:52:04 Hardy_: to opimd Dec 18 14:52:32 i mean, which path ^^ Dec 18 14:52:38 cant find it at all Dec 18 14:52:52 want to clean up a bit Dec 18 14:53:11 !logs Dec 18 14:53:12 Channel logs for #openmoko-cdevel are archived at: Dec 18 14:53:13 http://hentges.net/tmp/logs/irc/%23openmoko-cdevel Dec 18 14:53:14 Live-logs are available at Dec 18 14:53:16 http://hentges.net/tmp/logs/irc/livelogs/%23openmoko-cdevel.livelog Dec 18 14:53:18 See ?? help-logs for usage instructions Dec 18 14:53:30 Hardy_: sqlite.db Dec 18 14:53:33 because an sms seems to make an error, it always respawn after deletion Dec 18 14:54:02 ow... sqlite.db ow ow ow xD Dec 18 14:55:23 /etc/freesmartphone/opim/sqlite-messages.db Dec 18 14:55:25 IIRC Dec 18 14:55:37 thanks a lot dos1 Dec 18 14:55:49 Hardy_: we weren't lazy to do any effort to get a decent.overkill ;-P Dec 18 14:56:03 Hardy_: but you really shouldn't use database directly, without opimd Dec 18 14:56:22 okay, but i dont know how tu use opimd ^^ Dec 18 14:57:10 well. your problem sounds like a opimd bug anyway Dec 18 14:57:26 no, its just one sms Dec 18 14:57:42 i think its been sended on upgrading Dec 18 14:57:57 that's an easy one for opimd XD Dec 18 14:58:18 lol Dec 18 14:58:28 i just take a look in that file Dec 18 14:58:35 it ca do much higher sophisticated bugs as well Dec 18 14:58:48 perhaps i will recognize something Dec 18 14:59:00 lol.... sounds bad Dec 18 15:00:43 ow cant read anything in that file xD Dec 18 15:01:24 can i just delete that file? Dec 18 15:01:34 dont need the saved sms there Dec 18 15:01:51 ye Dec 18 15:01:54 yes Dec 18 15:02:00 think so Dec 18 15:02:01 nice Dec 18 15:02:05 i just try xD Dec 18 15:02:13 maybe just rename first Dec 18 15:02:38 good idea Dec 18 15:02:59 seems it worked Dec 18 15:03:14 no, loool they appeard later xD loool Dec 18 15:03:45 you need to retart opimd of course Dec 18 15:03:55 err restart Dec 18 15:04:02 aww ok xD Dec 18 15:04:26 though the first one was nonsense. no needed anymore X-P Dec 18 15:11:43 worked Dec 18 15:13:35 is there a shr boot splash around ? Dec 18 15:19:01 hm, I got an error during rootfs flash, and now "dfu-util -l" does not show my device any more, whether in nor or nand boot menu Dec 18 15:20:57 unrelated. at least the first can't cause the second Dec 18 15:22:58 what is autodeletion feature of an elementary window Dec 18 15:23:00 ? Dec 18 15:23:29 hm, it's those damn "unable to enumerate USB device" errors again Dec 18 15:29:54 anyone used dfu-util on win 7 64 bit? Dec 18 15:30:44 dos11: i need some help ;) i have a function in a class where i create a window with an widget in it to choose the day. after the day was clicked i want to delete the widget and the window. Dec 18 15:30:57 I can't get the drivers to work (cannot verify the drivers), even after I have enabled test mode Dec 18 15:31:32 dos11: for the first time everything is fine, but when calling that function the second time, the window can't be deleted anymore because it's already deleted Dec 18 15:33:25 dos1: code is http://shr.pastebin.com/d4f053283 and the function is at line 599 Dec 18 16:40:58 * Heinervdm hates python! Dec 18 16:42:22 ? Dec 18 16:42:44 Heinervdm: what for? speed? lack thereof? syntax? variablename checking? lack thereof? Dec 18 16:43:22 bugs Dec 18 16:44:18 if i call a function twice variables are somehow reused Dec 18 16:44:58 nice Dec 18 16:46:04 I've just been fighting with C strangeness, so it's no better on this side ;) Dec 18 16:46:22 :) Dec 18 16:46:47 apparently you can't rely on integer division to always return the same result Dec 18 16:47:08 that's bad... Dec 18 16:47:16 I had l=malloc(31), memset(l, w()/16+1, 0); Dec 18 16:47:20 w() happens to return 480 Dec 18 16:47:25 480/16+1 happens to be 31 Dec 18 16:47:30 still I get a SEGV Dec 18 16:47:57 telling myself to scr** it and use a malloc(128) fixes it. Dec 18 16:50:43 although I fear that the problem is actually someplace totally different and this just happens to make the segment violation fall into a still-allocated area Dec 18 16:50:52 I really need a _good_ debugging tool :( Dec 18 16:52:48 does python do call by referece or call by value? Dec 18 16:53:51 blindcoder: isn't 480/16 = 30.0 ? Dec 18 16:54:26 DocScrutinizer51: that's why I do +1 Dec 18 16:54:32 DocScrutinizer51: w()/16+1 Dec 18 16:54:59 yeah. still that's 31.0 then Dec 18 16:55:27 the joys of implicit typecasting Dec 18 16:55:54 hrm... Dec 18 16:58:17 blindcoder: btw that code is buggy by design. the malloc should take in account w() same way like memset, no? Dec 18 16:58:51 DocScrutinizer51: indeed, that's what got my attention at that code in the first place. Don't know why I missed that, though. Dec 18 16:59:01 DocScrutinizer51: now both use 128 hard-coded Dec 18 16:59:02 blindcoder: you? here? since when? ;) Dec 18 16:59:14 Wonka: erm... since I got my freerunner months ago? Dec 18 16:59:32 blindcoder: oh... I got too less time, obviously Dec 18 16:59:59 Heinervdm: IMO call by reference with delayed copying if needed. Dec 18 17:00:53 Q-Master: hmm, then i don Dec 18 17:01:03 't understand why my error happens Dec 18 17:01:13 Heinervdm: let me see Dec 18 17:02:04 Q-Master: http://shr.pastebin.com/d5c032c94 line 601 Dec 18 17:02:25 there i create a window, then passing that window to the callback Dec 18 17:02:41 and in the callback i delete the window Dec 18 17:02:55 first call is ok, second fails Dec 18 17:03:34 but when i use a real widget instead of my own class it works... Dec 18 17:05:30 Heinervdm: Are you sure that you're not destroying window during the work of this func? Dec 18 17:06:07 Q-Master: window is still on the screen Dec 18 17:06:22 the calendar disapears but window is still there Dec 18 17:09:05 blindcoder: though I'm absolutely no experienced C coder... isn't malloc() giving zeroed storage by definition? If so, the memset is senseless, no? Dec 18 17:10:13 DocScrutinizer-8, no Dec 18 17:10:20 malloc does not zero the storage Dec 18 17:10:25 calloc(1, size) Dec 18 17:10:26 does Dec 18 17:10:57 blindcoder, if you want to alloc a zero space, using calloc(1, size) instead of malloc(size) and memset is the correct way to go. Dec 18 17:11:08 s/zero/zeroed/ Dec 18 17:11:08 TAsn meant: blindcoder, if you want to alloc a zeroed space, using calloc(1, size) instead of malloc(size) and memset is the correct way to go. Dec 18 17:12:07 that's just my 2 cent (without knowing what you were talking about) :P Dec 18 17:12:16 * TAsn is off again. Dec 18 17:12:18 ciao. Dec 18 17:12:54 TAsn: so what,s the content of a malloc()d storage then? the password the app next door stored there some millisec before? Dec 18 17:16:45 Heinervdm: don't really understand... 8( It should be ok. Dec 18 17:17:06 Q-Master: i'm thinking that too... Dec 18 17:17:49 Heinervdm: try not do destroy that win. just show and hide Dec 18 17:18:15 Q-Master: how do i hide elementary Windows? Dec 18 17:18:33 Heinervdm: dunno really. 8( Dec 18 17:19:24 Q-Master: i can't find sth like hide in the source... Dec 18 17:20:17 Heinervdm: visible? just wild guessing Dec 18 17:20:29 DocScrutinizer51: no Dec 18 17:22:17 TAsn: okay, I'll use that then, but I doubt that that's my problem. thanks anyway Dec 18 17:23:49 Heinervdm: is window ever deleted? Dec 18 17:24:17 Q-Master: in the previus call of the funktion Dec 18 17:24:42 but a new one is created after that Dec 18 17:25:11 Heinervdm: and new creation failed? Dec 18 17:25:25 Q-Master: no, the window is shown Dec 18 17:26:29 somehow on the second creation the delted tag isn't reset Dec 18 17:28:12 Q-Master: i will set that on hold and will wait for raster to integrate the calendar widget... Dec 18 17:31:45 blindcoder: did you try to explicitly cast to integer the result of tha w()/30+1 ? Dec 18 17:33:28 DocScrutinizer51: nope, I just increased the buffer now. The error is now happening inside a allocated memory, but still appearing Dec 18 17:33:33 DocScrutinizer51: I'm not sure where it happens Dec 18 17:33:54 btw what happen if w(j is 485? whatever w() is at all Dec 18 17:34:49 Hi there. Do you know where can I find the source of openmoko-browser2 or how can I change the font size the pages are rendered in? pls Dec 18 17:35:11 DocScrutinizer51: valgrind output: http://shr.pastebin.com/d52c7ee20 Dec 18 17:35:22 DocScrutinizer51: w() is the width of the screen on the FreeRunner :) Dec 18 17:35:25 blindcoder: or is that the reason for the +1? Dec 18 17:35:39 DocScrutinizer51: actually, it's the width of the widget, the widget being as wide as the display Dec 18 17:36:08 DocScrutinizer51: that's a calculation of the maximum number of fixed-width characters that can be displayed on the screen Dec 18 17:36:30 ahh ok then Dec 18 17:37:00 the valgrind output is giving me the creeps, actually Dec 18 17:37:17 I think I'm wrecking havoc in the memory somewhere but can't find where exactly Dec 18 17:43:18 blindcoder: no idea. Not very used to reading valgrind logs Dec 18 18:19:01 DocScrutinizer-8, (content of malloc) yes. Dec 18 18:19:04 DocScrutinizer-8, just random. Dec 18 18:19:40 cool attack vector Dec 18 18:20:07 blindcoder, mind showing me code? I'll take a snappy look as I have to go in a sec :P Dec 18 18:20:09 I think you're wrong on that one for all new linux kernels Dec 18 18:20:31 DocScrutinizer-8, might be, I'm telling you what the docs promise, not what's actually being done on linux :P Dec 18 18:22:30 DocScrutinizer-8, just compiled a test app, on my box, you are right. (zeroes everything) though docs don't promise that. Dec 18 18:23:49 ok, you are not right ;P Dec 18 18:24:19 a = malloc(2000); Dec 18 18:24:26 for (i = 0 ; i < 2000 ; i++) { Dec 18 18:24:26 if (a[i]) printf("%d - %c\n", i, a[i]); Dec 18 18:24:28 :P Dec 18 18:25:19 ok, completely wrong :P Dec 18 18:25:29 (you) Dec 18 18:26:59 so I'm wrong and you got a severe security issue on your system. Better yet you just wrote and compiled the malware to exploit it Dec 18 18:27:14 http://pastebin.com/m73878d8d Dec 18 18:27:57 which means you should wipe clean sensitive buffers as soon as possible :P Dec 18 18:48:28 TAsn: http://scavenger.homeip.net/svn/fltknotes/trunk/src/ Dec 18 18:48:50 and what's the problematic section? Dec 18 18:49:06 yuck, I hate .cxx :P Dec 18 18:49:15 .cpp makes much more sense Dec 18 18:49:30 TAsn: notes_main.cxx lines 212 following Dec 18 18:50:06 TAsn: that's where the error seems to happen. either there, or in fltkclickdown.cxx lines 43 following Dec 18 18:50:42 well, since the code is not really c++, but rather a crude mixture of C and Objects, I use cxx ;) Dec 18 18:52:07 what's the problem btw? Dec 18 18:58:24 TAsn: segfault Dec 18 18:58:31 TAsn: http://shr.pastebin.com/d52c7ee20 Dec 18 18:58:42 that's what valgrind says Dec 18 18:58:44 why don't you gdb it? Dec 18 18:59:00 because I can't make sense of the output. it happens somewhere deep inside FLTK Dec 18 18:59:14 so I probably wreck the memory somewhere Dec 18 18:59:25 gdb Dec 18 18:59:27 and when it segs Dec 18 18:59:28 bt Dec 18 18:59:34 and show where's the last call you did. Dec 18 18:59:49 or put a breakpoint just before where you suspect it happens Dec 18 19:00:13 this seems like too much of a mess (in memory) for me to statically debug it (as I don't have fltk) Dec 18 19:00:48 TAsn: http://shr.pastebin.com/m66c2bd99 Dec 18 19:01:05 none of that actually happens in my code Dec 18 19:01:24 yeah, a mess it is ;) Dec 18 19:01:36 fltktouchscreen.cxx:153 Dec 18 19:01:41 not your code? Dec 18 19:01:55 what needs to be enabled in frameworkd.conf to use the device orientation stuff? Dec 18 19:01:58 that line just hands the input handling to Fl_Group::handle, which is in libfltk again Dec 18 19:02:09 return Fl_Group::handle(event); Dec 18 19:02:11 ok, but look what you pass... Dec 18 19:02:15 up Dec 18 19:02:18 your way to there Dec 18 19:02:21 the integer I get passed from fltk Dec 18 19:03:01 i c. Dec 18 19:03:44 that's why I hate GUIs :P Dec 18 19:04:25 hehe Dec 18 19:04:39 well, a note-taking application for the FR without a GUI is somewhat sad Dec 18 19:05:06 Ainulindale, tell him he's wrong ^ Dec 18 19:05:06 :P Dec 18 19:05:50 hey, gimme a hw-keyboard and I'll be happily noting stuff down in vim :P Dec 18 19:05:53 anyhow, sorry, no idea. Dec 18 19:06:00 thanks for the offer, anyway Dec 18 19:06:32 Sometimes a quick review by someone else solves stupid bugs Dec 18 19:06:46 so either this is not one of this cases, or I'm not the man for the job this time :P Dec 18 19:06:54 anyhow, like I said before Dec 18 19:07:10 malloc() + memset should be replaced by calloc(1, ) Dec 18 19:18:04 hmmm Dec 18 19:18:09 any kind of GUI is wrong Dec 18 19:18:26 blindcoder, there you go... ^ Dec 18 19:19:01 dos1, BAH stupid frameworkd! Doesn't support choosing different prefixes for installation :( Dec 18 19:20:58 TAsn: http://openmoko.opendevice.org/~dos/slider.edj - theme for idle screen in libphone-ui-shr Dec 18 19:21:04 name is wrong, just change it ;P Dec 18 19:21:10 sec Dec 18 19:21:44 TAsn: it's proof-of-concept, imagine red rectangle to be key and blue rectangles to be a padlock Dec 18 19:24:55 TAsn: done now, in malloc.cxx ;) I'm wrapping malloc anyway to check for errors Dec 18 19:25:40 also, #define free(x) free(x); x = NULL; Dec 18 19:27:18 you sholud wrap macros with do_whiles Dec 18 19:27:33 do { free(x); x = NULL; } while (0) Dec 18 19:29:27 dos1, weird :P Dec 18 19:30:15 TAsn: have you used qtopia? Dec 18 19:30:20 TAsn: oh, why that? Dec 18 19:30:21 or maybe just tested it? Dec 18 19:30:26 unfortunately yeah. Dec 18 19:30:32 blindcoder, for example Dec 18 19:30:43 if (x) Dec 18 19:30:44 free(x) Dec 18 19:30:49 else Dec 18 19:30:57 printf("bla"); Dec 18 19:31:00 will break with your code. Dec 18 19:31:02 err, okay, I see. Dec 18 19:31:20 multi expression macros should be wrapped with do whiles :P Dec 18 19:31:26 TAsn: why weird? something like that was in qtopia :P Dec 18 19:31:28 okay, I'll remember that Dec 18 19:31:35 well, differently Dec 18 19:31:38 blindcoder, there's a very good book Dec 18 19:31:41 called "bug free c" Dec 18 19:31:50 this is something like that qtopia thing :P Dec 18 19:31:51 dos1, I know, I was talking about the green/blue :P Dec 18 19:31:57 TAsn: ah Dec 18 19:32:03 but the idea is awesome. Dec 18 19:32:11 TAsn: green/blue rects are there just because i didn't prepare graphics yet :P Dec 18 19:32:25 maybe we need something like Dec 18 19:32:30 a telephone going free Dec 18 19:32:33 or I don't know what :P Dec 18 19:32:44 i.e leading an icon of shr out of jail :P Dec 18 19:34:25 ;) Dec 18 19:38:39 JesusMcCloud: ping Dec 18 19:46:57 dos1, does opimd cache queries? Dec 18 19:47:07 i don't think so Dec 18 19:47:10 ok :P Dec 18 19:47:20 def match(self, entries): Dec 18 19:47:23 in query_manager Dec 18 19:47:38 I thought maybe, though even if it did, it doesn't update them so it's useless for us :P Dec 18 19:48:18 it updates them when new entry arrives Dec 18 19:48:32 just handling updated and eleted entries is missing :P Dec 18 19:49:47 dos1: do you know away to get rid of the text in an elm toolbar? Dec 18 19:49:53 s/away/a way/ Dec 18 19:49:53 mrmoku meant: dos1: do you know a way to get rid of the text in an elm toolbar? Dec 18 19:50:25 hmm... look maybe at elementary-tests Dec 18 19:50:36 it seems toolbars without text are used in Map test Dec 18 19:51:16 Ainulindale: so... tell me when you'll have some time to install new trac theme :) Dec 18 19:51:28 dos1: where? Dec 18 19:51:36 dos1, I'm trying to figure out how to start hacking opimd :P Dec 18 19:51:47 mrmoku: when you click at some marker Dec 18 19:52:13 ag Dec 18 19:52:15 ah* Dec 18 19:52:21 there are labels... Dec 18 19:52:23 :) Dec 18 19:52:29 as it's wayyy too slow atm :P Dec 18 19:52:30 hmm... i think they wasn't there before :x Dec 18 19:52:59 ah, that's not toolbar Dec 18 19:53:00 :D Dec 18 19:53:15 mrmoku: dunno then... Dec 18 19:53:20 ok Dec 18 19:53:31 guess it would need changing the theme... Dec 18 19:53:38 to get rid of them Dec 18 19:53:56 dos1, got a sec? Dec 18 19:55:08 TAsn: it depends ;x Dec 18 19:55:16 I'll take that as a yes :P Dec 18 20:04:05 dos1, respond :P Dec 18 20:10:07 TAsn: in case he does not... have a minute? :P Dec 18 20:20:21 mrmoku, pong Dec 18 20:20:35 JesusMcCloud: two things Dec 18 20:20:47 first... how is your elm_toolbar knowledge? Dec 18 20:21:06 practically not existend Dec 18 20:21:09 ok Dec 18 20:21:11 second :P Dec 18 20:21:18 regarding theming of libphone-ui-shr Dec 18 20:21:24 ..yup?! Dec 18 20:21:32 TAsn had a good argument to _not_ have just one .edj Dec 18 20:21:46 one day we want to be able to mix different backends Dec 18 20:22:04 which means for example you might configure the thing to use libphone-ui-shr just for messages Dec 18 20:22:05 so diffented edje for every bit so you will be able to mix withotu wasting ram? Dec 18 20:22:12 yup Dec 18 20:22:18 one edj for every domain Dec 18 20:22:23 which is right now Dec 18 20:22:25 messages Dec 18 20:22:26 contacts Dec 18 20:22:28 makes sense Dec 18 20:22:29 idle screen Dec 18 20:22:30 dialer Dec 18 20:22:34 notification Dec 18 20:22:36 ... Dec 18 20:22:50 which in turn means we have to go the subdir way Dec 18 20:22:59 yup Dec 18 20:23:10 /usr/share/libphone-ui-shr/themes/default/messages.edj Dec 18 20:23:17 would be sane Dec 18 20:23:19 IMO Dec 18 20:23:21 sounds sane Dec 18 20:23:31 good Dec 18 20:24:02 will do that then... and give you a ping when it works... for you to try :) Dec 18 20:24:12 (won't be today though :P) Dec 18 20:24:50 mrmoku, sounds good... but expect high latency as i will not be home over holidays Dec 18 20:25:25 ok, me neither Dec 18 20:25:31 (i can oracitally hear ma liver pray for mercy...) Dec 18 20:25:38 :P Dec 18 20:26:23 ..if you were gone long enoght all of a sudden people want so spend time with you and are very generous... i'd be a fool not to enjoy it.. Dec 18 20:26:33 how's elementary phonelog progressing? Dec 18 20:26:48 JesusMcCloud: that was the first thing... toolbar :) Dec 18 20:26:50 working on it Dec 18 20:27:03 just use existign elmphonelog as template?! Dec 18 20:27:16 elmphonelog? Dec 18 20:27:24 u dont know it? Dec 18 20:27:42 think I have heard of it... never tried it though Dec 18 20:27:48 it was just great, but diesnt work anymore because sqlite was hardcoded... but it alredy partly supperted opimd Dec 18 20:28:10 also it was written in vala... Dec 18 20:28:17 ohh Dec 18 20:28:21 nice Dec 18 20:28:30 http://code.google.com/p/elmphonelog/ Dec 18 20:29:18 i used it as long as i could... i did not look at the source, but if you're lucky you can get a working phonelog from it within few hours Dec 18 20:29:33 no code in there though Dec 18 20:29:36 no? Dec 18 20:29:40 damn Dec 18 20:30:06 JesusMcCloud: what did you like? Dec 18 20:30:25 and what do you dislike with pyphonelog apart from being dead slow Dec 18 20:30:53 i am a big fan of elementary and elmphonelog was VERY fast Dec 18 20:31:09 also it supported opimd when i came out Dec 18 20:31:46 there is a recipe for elmphonelog, so you can find the source there Dec 18 20:31:57 nice Dec 18 20:33:36 mrmoku, it is in one of my (crappy) youtube vids about neo theme if you want to know what it looks like, i cal look for t Dec 18 20:34:20 http://www.youtube.com/watch?v=8DKwyv_mhkM Dec 18 20:35:52 mrmoku, yup at 5:15 Dec 18 20:35:56 sry 6:45 Dec 18 20:36:13 http://github.com/Leonti/elmphonelog Dec 18 20:36:13 waahrg... i need a cigarette Dec 18 20:36:47 we probably haven't moved the recipe from shr/import to shr/merge... Dec 18 20:37:30 Heinervdm, but elmphonelog doesnt work anymore, because it used hardcoded sql backend without opim interface for that.. it only used opim for contacts Dec 18 20:38:01 JesusMcCloud: i know, because of that we haven't moved it Dec 18 20:38:23 JesusMcCloud: ok, think you will like the new phonelog too then :) Dec 18 20:38:41 mrmoku, coolio... btw how come pyphonelog survived that long? Dec 18 20:39:11 good question :P Dec 18 20:40:00 i mean it was not integratet to lobframeworkd-phonegui, so stand alone, also it did not fit with the other apps and of cource it was slow as hell... would've been more efficient to focer someone to remember all the data and tell it to you... Dec 18 20:40:20 waaahrg stupid keyboard Dec 18 20:47:05 mrmoku, could you do me a favor and send me an email as soon as theming is done? maybe i'll be bored some time over the holidays Dec 18 20:47:52 JesusMcCloud: ok Dec 18 22:41:03 raster: small question before I fall asleep :P Dec 18 22:41:13 can I push a genlist directly into a pager? Dec 18 22:41:20 or do I have to pack it in a box? Dec 18 22:44:02 mrmoku: should work Dec 18 22:45:02 raster: ok, thanks :) Dec 18 22:45:31 and good night :P Dec 19 00:34:44 how can i backup my NAND so i can reflash it later ? Dec 19 01:01:36 dfu-util Dec 19 01:01:58 with -U Dec 19 01:11:56 Make sure you're using a recent u-boot with "dfu-util -U" - there used to be a data-corruption bug (#676, #1843) **** ENDING LOGGING AT Sat Dec 19 02:59:56 2009