**** BEGIN LOGGING AT Mon Aug 03 02:59:57 2009 Aug 03 05:36:24 budfive_: i think the idea of -dbg packages is to give meaningful backtraces. Aug 03 05:36:58 budfive_: and if you install the source separately i guess you'll see it in the debugger too. Aug 03 06:23:38 PaulFertser, (battery mail) oversaw that yesterday... I would *think* many SHR users read community mail. no precise idea though Aug 03 06:24:20 mrmoku: do you think there's a demand for using dumb batteries with FR? Aug 03 06:25:14 PaulFertser, I don't know... once reviving a dead FR was a reasone.. Aug 03 06:25:15 mrmoku: i cincerely hope you'll get a response today on OE ml, seems it's not very active on the weekends, so no wonder we're waiting for monday. Aug 03 06:25:53 PaulFertser, yeah, this build problem is depressing me :( Aug 03 06:26:24 mrmoku: So many people complained about power consumption, it seems that if you can carry 3 additional BL-6Cs with you it becomes a non-issue. Aug 03 06:27:00 PaulFertser, yeah, that would be another reason :) Aug 03 06:27:17 or replace a dead original one someday... Aug 03 06:27:51 mrmoku: exactly Aug 03 06:33:10 PaulFertser: hey. You use xorg on the gta02, not kdrive, righ? Aug 03 06:34:15 PaulFertser: I'm seeing something that may be a bug in xserver-kdrive, and need somebody to do a test for me Aug 03 06:34:51 PaulFertser: it's either in the xserver or in e Aug 03 06:35:56 budfive_: don't you have Xorg on your PC? Aug 03 06:38:18 budfive_: i'm still using kdrive i'm afraid :( Aug 03 06:41:09 budfive_: i use Xorg Aug 03 06:41:23 budfive_: does your e stuff run as non-root? Aug 03 06:49:00 lindi-: yes Aug 03 06:49:27 lindi-: I'm seeing that under some conditions, when XTest is used to create a sequence of fake keystrokes, some letters are transposed Aug 03 06:49:50 lindi-: This happens with the dictionary-based keyboard, for example. Aug 03 06:50:04 lindi-: If you try to type "morning", you get "mornnig" Aug 03 06:50:49 lindi-: it's something about queuing up duplicated letters. If I put delays or XSync() between each letter that is sent, the issue goes away, but XSync() makes it noticeably slower Aug 03 06:51:30 lindi-: I'm seeing the issue when faking letters into the dillo address bar or search bar, but am not seeing the issue when faking them into the vala terminal Aug 03 06:52:58 lindi-: http://pastebin.com/d2f4a383b is a program that types "morning". If I run this with the dillo address bar focused, I see "mornnig" Aug 03 06:53:12 lindi-: I can give you a binary if you like Aug 03 06:55:52 budfive_: key repeat? Aug 03 06:56:06 ah, transposed Aug 03 06:56:09 lindi-: it's not repeated, but the letters are transposed Aug 03 06:56:33 lindi-: it happens when the same letter appears multiple times in one word. The latter repeated letters are moved forward Aug 03 06:56:49 lindi-: morning -> mornnig, becomes -> beecoms Aug 03 06:57:54 lindi-: X has some sort of buffer, so the letters to fake are queued up. Then something goes wrong when you buffer a keystroke that is already in the queue. I don't know how this works internally, but right now I'm wondering whether the kdrive-based X is just buggy Aug 03 07:07:18 lindi-: can you try it? I can give you a precompiled executable of that little test program Aug 03 07:09:04 busy Aug 03 07:10:13 lindi-: ok. no worries Aug 03 07:20:37 cool Aug 03 07:24:40 good morning Aug 03 07:36:18 good morning Aug 03 07:44:43 ptitjes, methril|work: morning :) Aug 03 07:46:13 morning mrmoku Aug 03 07:46:16 how are you ? Aug 03 07:47:36 ptitjes, miserable :( Aug 03 07:47:47 oh! Aug 03 07:48:04 what"s happening ? Aug 03 07:48:07 I'm suffering a very serious OE blues :P Aug 03 07:48:23 oh Aug 03 07:55:25 mrmoku: thios happens so often with de oe repo :) Aug 03 07:57:43 methril|work, well... to defend OE a bit... correctly cross building a complete distro surely is not a lightweight task... Aug 03 07:57:55 though from time to time it really annoys me :( Aug 03 08:33:52 * DocScrutinizer still wonders why we base on OE instead of a decent distro (e.g. redhat). FR isn't exactly an "embedded" device. And we would have been rid of nasties like busybox from very beginning Aug 03 08:34:05 *I HATE BUSYBOX* Aug 03 08:36:56 mrmoku: I got it zecke cured yournblues yesterday? Aug 03 08:38:45 DocScrutinizer, no :( Aug 03 08:39:08 it somehow worked on my laptop, but not on the buildhost Aug 03 08:39:22 and zecke said it should not make any functional difference Aug 03 08:39:38 yay Aug 03 08:42:36 well, sounds like the old nasty issue of "buildhost has to be identical to targetsystem" once again. The nonsense where build checks for tools/libs to exist on buildhost and does different settings for target depending on it Aug 03 08:43:48 we had such nonsense with a pulseaudio lib or sth recently, you remember? Aug 03 08:46:59 is busybox really that bad? Aug 03 08:47:23 I kind of like their approach... Aug 03 08:48:26 BADBADBUSYBOX Aug 03 08:48:33 hehe Aug 03 08:48:44 spaetz: busybox is a crook Aug 03 08:49:01 mrmoku: can't get no sleeves for my records, can't get no laces for my shoes; can't get no fancy notes on my blue guitar, can't get no antidote for blues. Aug 03 08:49:25 admittedly it's not UNIX - one tool for one job - but hey it saves loads of disk space :) Aug 03 08:49:35 what's the use of implementing a 'less' command that isn't any better than 'more'? Aug 03 08:50:33 you ever noticed that e.g less doesn't know how to "/find" in busybox? Aug 03 08:50:50 I don't give a *shit* for BB Aug 03 08:51:14 nah, if I want to search I use nano :) Aug 03 08:51:23 ouuuuch Aug 03 08:51:37 hehe, or I use emacs when having remote-mounted the FR Aug 03 08:52:12 DocScrutinizer: have you already review FSO_Resource page? Do you have anything to add? Aug 03 08:52:24 yup, have Aug 03 08:52:32 but forgot what :-/ Aug 03 08:53:05 basically it's OK Aug 03 08:54:05 you might reference to the right ticket in a footnote for fsoraw e.g. (as that ticket is origin /me ;-P ) Aug 03 08:54:39 I might create a similar ticket for lsar Aug 03 08:55:04 wtf TEST resource??? Aug 03 08:55:40 s/fsoraw-armel/fsoraw/ Aug 03 08:56:58 add a really simple example like "fsoraw -r cpu mokomaze" Aug 03 08:59:21 PaulFertser: now memory starts to suck. Maybe that's it Aug 03 08:59:35 PaulFertser: hey its a wiki ;-) Aug 03 09:00:02 aaah, ref-URL to FSO Aug 03 09:02:35 DocScrutinizer: test resource is there, still present, nothing useful for an end-user but it appears in the list. Aug 03 09:02:42 DocScrutinizer: ticket referenced. Aug 03 09:02:49 DocScrutinizer: -armel done Aug 03 09:18:11 PaulFertser, the problem is not even RRECOMMENDS specific... some DEPENDS do not get installed either Aug 03 09:21:32 mrmoku: oh, that sucks. Aug 03 09:21:47 Hey there. Aug 03 09:21:49 TAsn, oh my opkg expert :P Aug 03 09:22:03 Status: install user not-installed Aug 03 09:22:09 what does that mean? Aug 03 09:22:22 Ainulindale, hey Aug 03 09:22:36 Ainulindale, back in Paris or still working hard? :P Aug 03 09:24:18 back in paris Aug 03 09:26:37 "" Status: install user not-installed "" loool Aug 03 09:27:02 mrmoku, Ainulindale: i think that this have to be with the bitbake update, it could be? Aug 03 09:27:18 mrmoku, Ainulindale: some clasess in dev changed Aug 03 09:27:31 methril|work, do you still have the old svn tree around? Aug 03 09:27:31 What's the problem? Aug 03 09:27:44 Ainulindale, lots of packages don't get stuffed into the image Aug 03 09:27:46 mrmoku: nop Aug 03 09:27:50 OE sucks Aug 03 09:27:53 mrmoku: i updated this weekend Aug 03 09:27:54 :) Aug 03 09:27:58 Ainulindale, stuff like ophonekitd for example :P Aug 03 09:28:07 spaetz: hehe, ack Aug 03 09:28:33 Problem is, is just so mcuheasier to say sth sucks than actually producing anything better :) Aug 03 09:29:03 Ainulindale, and if I compare the opkg status file from a working rootfs and a not working rootfs... for example task-shr-minimal-apps on not working has Aug 03 09:29:14 Status: install user not-installed Aug 03 09:29:14 on working it has Aug 03 09:29:20 Status: install user installed Aug 03 09:29:28 Again? Aug 03 09:29:30 Sheesh Aug 03 09:29:39 Ainulindale, you know that? Aug 03 09:29:59 I meant package not being installed Aug 03 09:30:22 ahh, k Aug 03 09:30:25 * DocScrutinizer adores clear crisp errormsgs like "install user not-installled" - muhahaha Aug 03 09:31:21 Ainulindale, yesterday I tried bitbake from master branch and bitbake from the rework-parser branch Aug 03 09:31:36 master branch failed with three ERROR for some packages Aug 03 09:31:53 then I tried the rework-parser and it built a successfull image Aug 03 09:31:58 best of them are dynamically composed of a few very common words, to not make it too simple to grep Aug 03 09:32:11 though on the buildhost (and for PaulFertser) just using rework-parser branch does not help Aug 03 09:33:03 Ainulindale, on the other hand it is doing 'opkg-cl install task-shr-minimal-apps' Aug 03 09:33:16 and somehow does not install it :( Aug 03 09:33:55 might be an opkg problem as well Aug 03 09:34:43 back in the old days it was mandatory each errormsg had a unique ID-number to be obtained from our private project "icann" Aug 03 09:35:39 so "open() failed [120034]" still was a usefull msg Aug 03 09:36:34 DocScrutinizer, well, what I pasted was no error message. It is the Status: for a package in /usr/lib/opkg/status Aug 03 09:36:46 ooh Aug 03 09:36:54 TAsn, please explain ^^^ :) Aug 03 09:37:17 what says the manpage about semantics of this status? :-P Aug 03 09:40:08 haha Aug 03 09:40:20 to be honest "install user" sounds rather sensible and tempting to me ;-D Maybe a good moment to implement a decent non-root user? :-P Aug 03 09:42:33 hi Aug 03 09:42:45 so is it true that 3G SIM cards don't work with freerunner? Aug 03 09:44:43 doxy2: waht a obsolete urban legend. Update to moko11 gsm-FW. See gsm-flashing wikipage! Aug 03 09:46:25 btw it's never been 3G that borked, it was a SIM-protocol speedup optimization that borked on newer cards supporting it. Aug 03 09:46:33 ok. could the firmware be the reason that requesting the gsm resource fails through frameworkd Aug 03 09:46:39 i Aug 03 09:46:48 yes Aug 03 09:46:50 requesting other resources works Aug 03 09:47:01 if you have a new SIM Aug 03 09:47:17 DocScrutinizer: you just made my day, dude Aug 03 09:47:22 :D Aug 03 09:47:28 np, welcome Aug 03 09:48:21 DocScrutinizer: hehe, it was easy. Some people don't need much :D Aug 03 09:48:22 yup Aug 03 09:48:37 nice teaser to start my day Aug 03 09:49:47 mrmoku: so you say you suspect it's opkg and not OE fault? Aug 03 09:50:06 PaulFertser: yeah, i've been working on getting my freerunner work like i want it for so long that i'm pretty easy by now :D Aug 03 09:50:19 * DocScrutinizer prepares another coffee to celebrate the salvaged soul of doxy2 Aug 03 09:53:41 PaulFertser, no, I just say it *might* be opkgs fault too :P Aug 03 09:54:05 Recent Change Aug 03 09:54:07 http://code.google.com/p/opkg/source/list Aug 03 09:56:16 PaulFertser, 215 is what we have Aug 03 09:59:24 PaulFertser, what makes me think more that its bitbakes fault is that yesterday I managed to build a image Aug 03 09:59:53 waaah, noon (kinda my private midnight) Aug 03 10:00:06 PaulFertser, if you want you could try to build an image with master branch of bitbake (which will fail) and then with the 1.8 branch again Aug 03 10:00:20 that somehow might have been what made me build a working image Aug 03 10:02:59 mrmoku: i don't like shamanism, i can't see how that can help us to solve the issue even if that'd work. Aug 03 10:04:24 PaulFertser, to verify it is bitbake we have to look at? Aug 03 10:04:32 PaulFertser: scientific approach: observe, explain, induce Aug 03 10:04:52 * mrmoku did not want to add that trick to the SHR wiki ;) Aug 03 10:05:10 lol Aug 03 10:55:20 PaulFertser: http://trac.freesmartphone.org/ticket/461 Aug 03 10:57:28 DocScrutinizer: link added Aug 03 10:58:08 please toss link to wikipage ;) Aug 03 11:00:30 aaah, http://wiki.openmoko.org/wiki/Special:Newpages -- of course, what else Aug 03 11:05:36 hi, is what's the program name that can redirect a serial port to the network? can socat do that? Aug 03 11:05:51 freesmartphone.org: 03jluebbe 07framework * rca2c19a7ee58 10/framework/subsystems/ogsmd/modems/ti_calypso/mediator.py: ogsmd: [TI CALYPSO] verify that we got a complete monitor response from the modem Aug 03 11:07:45 Hi All! Aug 03 11:07:45 gnutoo: yes Aug 03 11:07:55 ok thanks a lot I'll read the man Aug 03 11:08:38 Q-Master: dos1|away just came but he's already away :D Aug 03 11:09:19 gnutoo: i've recently had to redirect my serial port to reflash my old dialup modem as the flashing program works only with windows. Socat worked :) Aug 03 11:09:29 PaulFertser: I found a webirc 8) Aug 03 11:11:32 PaulFertser: how to link from ndex-page to fso-resources? Aug 03 11:11:42 PaulFertser: please check my edit in http://wiki.openmoko.org/wiki/FSO_Resources#Automatic_way Aug 03 11:12:08 DocScrutinizer: (index-page) no idea Aug 03 11:13:35 DocScrutinizer: looks ok Aug 03 11:13:49 k Aug 03 11:15:56 PaulFertser: do you think it needs a verbose explanation about multiple allocation and ref-counters, so ppl understand even better the diff? Aug 03 11:17:12 DocScrutinizer: i'm not sure, but otoh i don't know end-users enough to judge :) Aug 03 11:17:31 hehe, same here ;) Aug 03 11:20:05 maybe a word about automatic freeing of resources at process termination (in contrary to explicit reset of policy at regular terminating an app) would be even more enlightening Aug 03 11:20:44 including a "kill -9" example which fails boldly with mdbus method Aug 03 11:21:33 AIUI ;-) correct me if I'm wrong Aug 03 11:24:00 DocScrutinizer: (automatic freeing) it works afaik and a note should be added. Aug 03 11:24:47 you bet it *should* work. It's been the basic idea behind it Aug 03 11:36:10 automatic freeing of resources work Aug 03 11:53:08 spaetz: see http://git.freesmartphone.org/?p=specs.git;a=blob_plain;f=html/usage-intro.html;hb=HEAD, "" the system can detect a client leaving the bus, no matter whether this has been a regular detach or happened due to unforseen circumstances, such as a crash."" -- that's been the reason I requested creation of fsoraw ;-) Aug 03 11:54:27 huh? I'm quite sure frameworkd does that... Aug 03 11:54:36 (at least thats what my memory says...) Aug 03 11:55:09 mrmoku: sure. where's the confusing part in there? Aug 03 11:55:34 DocScrutinizer, in the doc you cited? Aug 03 11:56:41 mrmoku: I don't see any confusing part in that doc Aug 03 11:56:48 it's correct and explains the way fsoraw is working Aug 03 11:58:09 right, whenever fsoraw exits the resource gets freed Aug 03 11:58:27 I understand that. That's why I said, that automatic freeing of resources works :) Aug 03 11:58:38 even if fsoraw would crash Aug 03 12:01:46 raster: ping Aug 03 12:02:54 dos1: i've news from Q-master. He says opimd-contacts work, but search is broken and gives not what you ask. But exiting search by pressing enter works. Aug 03 12:03:39 PaulFertser: that's something with utf8 support (russian chars)... Aug 03 12:03:42 DocScrutinizer, ahh... rereading I see there is "can detect" not "can't detect"... :) Aug 03 12:03:45 yeah Aug 03 12:03:54 PaulFertser: with ascii it works correctly ;x Aug 03 12:04:06 dos1: poo (next generation) Aug 03 12:04:18 dos1: you fixed that yesterday?? Aug 03 12:04:37 DocScrutinizer: big j! Aug 03 12:04:55 DocScrutinizer: utf8? i thought, it doesn't traceback anymore Aug 03 12:05:03 raster: OLA! Aug 03 12:05:39 raster: can I use elm_icon by just packing it into box? Aug 03 12:05:44 yup Aug 03 12:06:09 and... off i run. bbl Aug 03 12:06:09 :) Aug 03 12:06:33 hasta luego, amigo Aug 03 12:06:46 raster: but it does't work here :( Aug 03 12:07:04 but ok, if you said that it should work then probably i'm doing something wrong :P Aug 03 12:07:19 dos1: ;-P Aug 03 12:09:14 * DocScrutinizer away, starting day... Aug 03 12:14:31 * dos1 has big update to opimd-contacts, but that fscking icons don't want to work :( Aug 03 12:35:09 oh, ok, icons work :D now fighting with size_hint_align_set :x Aug 03 12:39:30 is there a size_not_a_hint_align_set()? Aug 03 12:39:42 just wondering Aug 03 12:40:15 dir(elementary.Button) Aug 03 12:40:26 or different widget Aug 03 12:40:29 and you'll see ;p Aug 03 12:40:42 (in python, after import elementary) Aug 03 12:43:36 mrmoku: hrw telling why are problems with the 1.8 git bitbake Aug 03 12:43:44 mrmoku: in oe channel Aug 03 12:43:50 methril|work: uuuhhh Aug 03 12:47:18 methril|work: he does not want to tell us :( Aug 03 12:54:45 mrmoku: so, what happened there? I was too late to join. Aug 03 12:55:02 PaulFertser: problems of hrw are different Aug 03 12:55:07 kernel module related Aug 03 12:55:22 though he told me that svn tree is still up Aug 03 12:55:26 and actually it is Aug 03 12:55:29 was just too stupid Aug 03 12:55:39 mrmoku: can you probably try to ask him specifically about your problem? ;) Aug 03 13:17:56 yay! Aug 03 13:18:14 ok, my ui problems fixed :) Aug 03 13:21:14 dos1: rather fix *my* ui problems! :-P Aug 03 13:21:26 DocScrutinizer: which? ;p Aug 03 13:21:36 hmm, lemme think.... Aug 03 13:21:42 :P Aug 03 13:21:49 please don't! Aug 03 13:21:50 :D Aug 03 13:26:35 looks like opimd-contacts is first contacts app which supports multiple phone numbers in one contact Aug 03 13:26:37 ;) Aug 03 13:29:10 and pictures :D Aug 03 13:30:05 F4t: no, pictures aren't implemented yet ;) Aug 03 13:30:15 oh.. Aug 03 13:30:32 but from opimd-contacts side it would be only two lines to implement (except adding new image), but i'm too lazy ;) Aug 03 13:42:40 hmm, is it normal that after enabling and disabling gprs, i can't make phonecalls anymore in shr-unstable? Aug 03 13:43:27 and if i restart ophonekitd, it hangs after "DEBUG: Initiated etk" Aug 03 13:50:40 mrmoku, Deubeuliou Aug 03 13:50:46 git clone git:// is now enabled on all the repos Aug 03 13:58:12 Ainulindale: I love you :D Aug 03 13:58:37 mrmoku: kiss me O old one Aug 03 13:59:10 Ainulindale: nah... get rid of your beard first ;) Aug 03 13:59:26 :-) Aug 03 14:09:57 SHR: 03seba.dos1 07opimd-utils * r4bed6254ce1b 10/ (opimd-messages opimd-notifier opimd_utils/opimd_reply.py): opimd-reply: add func_status Aug 03 14:26:53 displaying custom fields, adding new contact, editing contact and displaying contact history (messages and calls) Aug 03 14:27:07 that's the only missing things now ;) Aug 03 15:16:59 Ainulindale, still here? Aug 03 15:18:10 Yep Aug 03 15:18:26 I want to use libfso-glib Aug 03 15:18:39 and there are some things I'm not sure about Aug 03 15:18:46 one thing is g_type_init Aug 03 15:18:46 Yes? Aug 03 15:18:59 right now it is called by libframeworkd-glib Aug 03 15:19:27 and then there is (like you already complained) the missing framework handler struct Aug 03 15:21:13 well Aug 03 15:21:15 next thing is the dbus proxies missing... lfg is doing that now Aug 03 15:21:20 Well Aug 03 15:21:24 I just saw the dbus proxies Aug 03 15:21:38 As for the framework handler structs Aug 03 15:21:42 It's just out there for convenience Aug 03 15:21:52 mrmoku: Aug 03 15:22:11 damn Aug 03 15:22:16 you deleted my working directory on the buildhost Aug 03 15:22:29 your working directory? Aug 03 15:22:33 yes Aug 03 15:22:40 doing stuff on the buildhost? Aug 03 15:22:42 me too :P Aug 03 15:22:45 yeah :-) Aug 03 15:22:47 sorry :) Aug 03 15:23:25 Don't move I'm redoing it Aug 03 15:23:39 image build is running though Aug 03 15:24:04 well I'm just building a working directory for libfso-glib Aug 03 15:24:06 nothing more Aug 03 15:24:14 Although it might confuse the build process Aug 03 15:24:29 np :) Aug 03 15:24:42 image is crap anyway :( Aug 03 15:24:58 ahh and finished Aug 03 15:25:48 omg *sigh* Aug 03 15:25:50 Although I think that ptitjes didn't do .h Aug 03 15:25:51 And that is bad Aug 03 15:26:10 Ainulindale, well the .h gets generated Aug 03 15:26:32 FreeSmartphoneGSMPDP* free_smartphone_gsm_pdp_dbus_proxy_new (DBusGConnection* connection, const char* name, const char* path); Aug 03 15:26:37 That's of interest to you Aug 03 15:26:44 Ainulindale, yeah, I know Aug 03 15:26:48 but lfg was hiding that Aug 03 15:27:01 WEll yeah because this was called before every method Aug 03 15:27:01 you did not have to get DBusGConnection in the first place Aug 03 15:27:07 And that should be done as well in libfsoglib Aug 03 15:27:23 That's why I just asked to Deubeuliou to see if he could do that Aug 03 15:27:50 And anyway Aug 03 15:27:55 That's the problem when you autogenerate stuff =) Aug 03 15:28:25 mrmoku: what is lacking is an easy way to register handlers Aug 03 15:28:33 and an easy way to ensure that a specific connection has been built Aug 03 15:28:38 All in all we just miss a .vala thing Aug 03 15:29:34 Ainulindale, and Deubeuliou is working on that? Aug 03 15:29:56 well I don't know Aug 03 15:29:59 ask him :-p Aug 03 15:30:04 Deubeuliou, ping :) Aug 03 15:31:34 He's in front of me right now Aug 03 15:31:40 Ainulindale, ping him then ;) Aug 03 15:31:55 anyway, maybe its not even worth it Aug 03 15:32:26 * mrmoku checks what is needed by a libfso-glib user to listen to signals... Aug 03 15:33:14 There's no method to create proxies Aug 03 15:33:46 what is of interest to me is the new methods defined in the generated .c Aug 03 15:33:51 they should be as well in the .h Aug 03 15:34:11 like which ones? Aug 03 15:34:15 the _async stuff? Aug 03 15:43:54 mrmoku: what we lack is Aug 03 15:44:02 1) being able to create the connection automagically Aug 03 15:44:12 2) the headers to create the proxy interfaces Aug 03 15:44:17 (judging from what I see) Aug 03 15:45:24 free_smartphone_get_proxy? Aug 03 15:46:13 yeah didn't see that Aug 03 15:46:47 ps = (FreeSmartphonePIMSources *)free_smartphone_get_proxy(bus, "org.freesmartphone.opimd", "/org/freesmartphone/PIM/Sources", "org.freesmartphone.PIM.Sources", FREE_SMARTPHONE_PIM_TYPE_SOURCES); Aug 03 15:46:52 yeah just saw that Aug 03 15:46:52 is what I do Aug 03 15:46:56 You're a good boy Aug 03 15:47:01 * Ainulindale pats mrmoku Aug 03 15:47:02 :-) Aug 03 15:47:03 :) Aug 03 15:47:17 so the only thing lacking is 1 Aug 03 15:47:19 Right? Aug 03 15:47:28 what I'm missing is a convenient way to add signals Aug 03 15:47:30 (ptitjes should have written a howto in the first place) Aug 03 15:47:43 without manually calling dbus_g_proxy_add_signal and dbus_g_proxy_connect_signal Aug 03 15:48:34 mrmoku: well Aug 03 15:48:39 are they somewhere in the vala thing? Aug 03 15:48:51 no idea Aug 03 15:48:56 I just looked at the .h Aug 03 15:48:58 (I hate new shiny stuff) Aug 03 15:49:00 (Just for that) Aug 03 15:49:17 (I hate new shiny stuff I don't know ;) Aug 03 15:49:39 mickey|dinner ? Aug 03 15:49:42 Stop eating and help us Aug 03 15:51:47 Hmmm Aug 03 15:52:02 I hate the (missing) doc of new shiny stuff Aug 03 15:52:14 I think he implemented signals as methods mrmoku Aug 03 15:53:18 that or the methods handlers should be register as signal handlers manually Aug 03 15:53:38 http://git.freesmartphone.org/?p=libframeworkd-glib.git;a=blob;f=src/frameworkd-glib-dbus.c;h=58683bf0c5375ded57d06dd53adaf5d71e9f7335;hb=HEAD <= line 206 Aug 03 15:55:24 hmmm Aug 03 15:55:31 I'm under the impression that it connects to every signal Aug 03 15:55:54 lfg? Aug 03 15:56:27 just the ones you set a handler for... Aug 03 15:57:21 Not lfg but libfsoglib Aug 03 15:57:44 Ainulindale, ahh, because your gitlink was lfg... Aug 03 15:58:11 does a signal source get to know about recipients registering for the signal? Aug 03 15:58:17 mrmoku Aug 03 15:58:36 tmp/work/armv4t-angstrom-linux-gnueabi/libfso-glib-0.2.0-gitrx44+9d292508739452b55b80ec40ec57405a5de2159f-r0/git/src/freesmartphone-gsm.c Aug 03 15:58:39 in shr-unstable Aug 03 15:58:46 search keypad Aug 03 15:58:51 freesmartphone.org: 03seba.dos1 07dos/opimd-tracking * r1099a55dc5b9 10framework/framework/subsystems/opimd/ (5 files): opimd: fix imports and move QueryMatcher class into query_manager.py Aug 03 15:58:52 freesmartphone.org: 03seba.dos1 07dos/opimd-tracking * raa021ba78f6e 10framework/framework/subsystems/ogsmd/modems/ti_calypso/mediator.py: Merge branch 'master' of git@git.freesmartphone.org:framework into dos/opimd-tracking Aug 03 15:58:53 freesmartphone.org: 03seba.dos1 07dos/opimd-tracking * r59a37a54a40e 10framework/framework/subsystems/opimd/ (6 files): opimd: use GenericEntry in Contact class and refactor some opimd structures (experimental!) Aug 03 15:59:07 You'll find free_smartphone_gsm_device_dbus_register_object Aug 03 15:59:20 I can't see where this function is called Aug 03 16:00:50 Ok when the iface is inited apparently Aug 03 16:01:14 static void free_smartphone_gsm_device_base_init (FreeSmartphoneGSMDeviceIface * iface) { Aug 03 16:01:26 (8238) Aug 03 16:02:00 and that method is called at the init of the type Aug 03 16:02:50 and connects to all signals for that iface Aug 03 16:02:54 yeah Aug 03 16:03:00 (this is ugly) Aug 03 16:04:30 Now Aug 03 16:04:34 How to give it a freaking listener... Aug 03 16:04:35 apart from that... what I can't see anywhere is where it calls some callback... Aug 03 16:07:45 Ainulindale, it listens to all signals and handles them with _dbus_free_... Aug 03 16:07:50 or am I missing something? Aug 03 16:08:47 I don't know Aug 03 16:09:44 freesmartphone.org: 03seba.dos1 07framework * raec532c0ba24 10/framework/subsystems/opimd/ (pimb_sim_contacts_fso.py pimb_sim_messages_fso.py): opimd: SIM-*-FSO: cast _backend_entry_id to int when deleting and updating Aug 03 16:15:15 freesmartphone.org: 03seba.dos1 07dos/opimd-tracking * rd1a6e562781f 10framework/framework/subsystems/opimd/ (pimb_sim_contacts_fso.py pimb_sim_messages_fso.py): Merge branch 'master' of git@git.freesmartphone.org:framework into dos/opimd-tracking Aug 03 16:27:44 PaulFertser: mwester: http://en.wikibooks.org/wiki/Serial_Programming/Serial_Linux#Line_Discipline ""On some systems, the settings done by stty are reverted to system defaults as soon as the device is closed again."" If system defaults is a subset of the complete set of possible settings, this makes sense with the behaviour observed, no? Aug 03 16:28:09 Ainulindale, still need your workdir? Aug 03 16:29:13 PaulFertser: mwester: ooops -- http://en.wikibooks.org/wiki/Serial_Programming/Serial_Linux#Configuration_with_stty Aug 03 16:41:05 yay! Aug 03 16:41:18 dos1, got your icons working? Aug 03 16:41:26 mrmoku: long time ago ;) Aug 03 16:41:34 mrmoku: now i was fighting with size_hint_align_set ;D Aug 03 16:41:36 dos1, what is working then now? Aug 03 16:41:43 mrmoku: but it already works :) Aug 03 16:41:51 mrmoku: custom fields in opimd-contacts Aug 03 16:41:52 mrmoku: nah Aug 03 16:42:02 so now only s/tel:/icon/ and s/mail:/icon/ and commit :) Aug 03 16:42:59 Ainulindale, I looked at ophonekitd vala source code (for the signal stuff) Aug 03 16:43:04 though it does not help for c clients Aug 03 16:43:48 and I really dislike the fact that fso-glib is listening to all signals :( Aug 03 16:44:07 well so do I Aug 03 16:56:59 mrmoku: I'd consider this a design flaw Aug 03 16:57:23 DocScrutinizer: would be interesting to hear ptitjes comment on that one Aug 03 16:58:58 mrmoku: I'm no expert for dbus, but I seem to remember there's some config that enables/blocks connection of certain clients to certain signals (among others). I'd expect we shoot our own foot by making fso-glib listen/register to *all* signals irrespective of those config settings Aug 03 17:00:01