**** BEGIN LOGGING AT Mon Oct 25 02:59:57 2010
Oct 25 07:41:04 <hob> where can i find the coding standards for ofono?
Oct 25 08:37:50 <emdete> hob: in the doc folder...
Oct 25 08:43:11 <hob> i have ofono-0.32 and i am not able to find any coding-style.txt. there
Oct 25 08:46:09 <emdete> strange, here it available
Oct 25 08:46:26 <emdete> try a `git checkout doc/coding-style.txt`
Oct 25 08:58:45 <emdete> hob: does that help?
Oct 25 08:58:59 <hob> yes it does thanks
Oct 25 15:23:55 <holtmann> denkenz: Did you look into the multiple active context support stuff by any chance?
Oct 25 15:24:20 <denkenz> not yet
Oct 25 16:54:10 <holtmann> pessi: Around?
Oct 25 16:56:31 <holtmann> pessi: When you tried multiple contexts and ISI, did you add multiple GPRS context drivers as well. Or just assumed one driver gets used multiple times?
Oct 25 23:04:25 <raji__> holtmann, denkenz, in meego history plugins are failing to load because introspection always creating 2nd method name as empty string in the xml format. Introspect is getting called automatically, who calls it? 
Oct 25 23:04:48 <holtmann> ???
Oct 25 23:04:57 <holtmann> raji__: What does this mean?
Oct 25 23:06:25 <raji__> which method calls introspect, I tried to grep but did not find a method that calls 'introspect' method
Oct 25 23:07:17 <holtmann> Qt or Python or any client.
Oct 25 23:08:36 <raji__> right now I am not using any of those, when run the ofonod , introspect method define in gdbus/object.c is called,
Oct 25 23:08:46 <raji__> I am trying to understand the call flow
Oct 25 23:09:48 <raji__> does the question make sense?
Oct 25 23:10:26 <holtmann> As I said, most likely when some other app using Python or Qt seems org.ofono it will introspect it.
Oct 25 23:10:34 <holtmann> What is failing here exactly.
Oct 25 23:11:38 <raji__> callhistory plugin, exposes two methods , first method name , signature is right but, second method name is "", signature is correct
Oct 25 23:12:07 <holtmann> What is you GDBusMethodTable structure.
Oct 25 23:12:47 <raji__> static GDBusMethodTable call_history_methods[] = {
Oct 25 23:12:47 <raji__>         { "SetVoiceHistoryRead", "", "", call_history_set_voice_history_read ,G_DBUS_METHOD_FLAG_ASYNC},
Oct 25 23:12:47 <raji__>         { "GetVoiceHistory", "", "a(usqii)", call_history_get_voice_history ,G_DBUS_METHOD_FLAG_ASYNC},
Oct 25 23:12:47 <raji__>         { },
Oct 25 23:12:47 <raji__> };
Oct 25 23:12:58 <raji__> Should I use paste bin?
Oct 25 23:13:18 <holtmann> No need.
Oct 25 23:14:30 <holtmann> Looks fine to me. What is the g_dbus_register call.
Oct 25 23:15:26 <raji__> if (!g_dbus_register_interface(conn,
Oct 25 23:15:26 <raji__>                                        OFONO_MANAGER_PATH,
Oct 25 23:15:26 <raji__>                                        OFONO_CALL_HISTORY_INTERFACE,
Oct 25 23:15:26 <raji__>                                        call_history_methods,
Oct 25 23:15:26 <raji__>                                        call_history_signals,
Oct 25 23:15:26 <raji__>                                        NULL,                  /* Properties */
Oct 25 23:15:28 <raji__>                                        shared_data,           /* Userdata   */
Oct 25 23:15:30 <raji__>                                        NULL))    
Oct 25 23:16:08 <holtmann> Looks fine.
Oct 25 23:16:16 <holtmann> So where is the introspection data that gets messed up?
Oct 25 23:19:02 <holtmann> And are you sure you are compiling this against the right gdbus.h file?
Oct 25 23:21:07 <raji__> for any second method, for example "GetVoiceHistory" here, in the generate_introspect_xml method, method->name is "", 
Oct 25 23:23:06 <holtmann> raji__: Same question. Are you sure you are compiling against the right gdbus.h file?
Oct 25 23:23:44 <raji__> you got me , I am using wrong one
Oct 25 23:23:46 <holtmann> Do it like this ....,G_DBUS_METHOD_FLAG_ASYNC, 0 },
Oct 25 23:24:03 <holtmann> Then your binary layout of the method table is wrong.
Oct 25 23:24:20 <holtmann> And of course the method name is wrong. Since know you are using the second signature field.
Oct 25 23:24:29 <holtmann> On x86 32-bit ints and pointers are the same size.
Oct 25 23:25:27 <raji__> ok, thanks a lot
Oct 25 23:26:55 <holtmann> Sure. No problem.
Oct 25 23:27:02 <holtmann> Now to your API. That is racy.
Oct 25 23:27:24 <raji__> Yea, This is going to be a quick fix for couple of weeks.
Oct 25 23:27:37 <holtmann> Good. As long as you know that.
Oct 25 23:30:36 <raji__> did you have a chance to look at the patch I sent, I will fix the issues now rather than later if you absolutely dont like the way it is implemented, that will save me time
Oct 25 23:35:27 <holtmann> I have to have a look at it. Hopefully I get to it this week.
**** ENDING LOGGING AT Tue Oct 26 03:00:09 2010