**** BEGIN LOGGING AT Thu Apr 14 02:59:58 2011
Apr 14 10:25:07 <mr_amit> the patches have been sent to ofono mailing list
Apr 14 10:25:14 <mr_amit> the bug report is also updates
Apr 14 10:25:18 <mr_amit> see: https://bugs.meego.com/show_bug.cgi?id=14784
Apr 14 13:04:13 <pnunes> denkenz: ping
Apr 14 14:27:10 <denkenz> pnunes: pong
Apr 14 14:28:13 <pnunes> denkenz: Following the "simplification" related to the flag 'respond_on_exit', I noticed that the flag was set to true for the command Launch browser.
Apr 14 14:28:46 <pnunes> denkenz: But, for such proactive command, the terminal response "UICC session terminated by the user" is not possible ( table 6.1 TS 102 223).
Apr 14 14:29:31 <pnunes> denkenz: I was about to remove this flag for Launch Browser but I realize that we still need to send an error in case of stkagent exit.
Apr 14 14:30:05 <pnunes> denkenz: It requires also to distinguish in the callback 'session/default_agent_notifiy the type of proactive command.
Apr 14 14:31:44 <pnunes> denkenz: I know that this kind of none-compliancy is not tracked by GCF test so is it something you can consider? 
Apr 14 14:36:16 <denkenz> So respond_on_exit is set because the agent can quit during ConfirmLaunchBrowser 
Apr 14 14:36:28 <denkenz> And I believe that this is an error in the spec
Apr 14 14:36:36 <denkenz> Talk to Jeevaka, he did that part
Apr 14 14:37:34 <denkenz> I think we discussed this with him before, and there is in fact a test that checks for end session during launch browser confirmation
Apr 14 14:37:45 <denkenz> But it has been a while
Apr 14 14:38:35 <pnunes> denkenz: OK, I will clarify with Jeevaka. Thanks.
Apr 14 14:40:12 <Gzajac> denkenz: ping
Apr 14 14:40:20 <denkenz> Gzajac: pong
Apr 14 14:41:39 <Gzajac> denkenz: I have manually applied Gustavo patches into ConnMan to request and release a private network, I am now thinking about the way to call those dbus methods into oFono
Apr 14 14:42:25 <Gzajac> denkenz: so I thought first to use a ConnMan plugin to send dbus message to ConnMan and using some drivers registered into the emulator
Apr 14 14:42:54 <Gzajac> But isn't it faster to directly call the dbus methods into emulator?
Apr 14 14:43:17 <Gzajac> Would that respect the architecture ? or should I use a plugin ?
Apr 14 14:43:53 <denkenz> You have to create some sort of plugin for this
Apr 14 14:44:04 <denkenz> Calling connman directly from oFono is not a good idea
Apr 14 14:44:11 <denkenz> from oFono core that is
Apr 14 14:45:28 <Gzajac> ok so the best idea is to implement a plugin connman like an ofono plugin exists into connman, right?
Apr 14 14:47:28 <denkenz> yes, you need to abstract the private network ip address request somehow
Apr 14 14:47:54 <denkenz> we have to play with naming, but something like:
Apr 14 14:48:42 <denkenz> get_emulator_network(struct emulator_network *en, ofono_emulator_network_cb_t cb, void *data)
Apr 14 14:49:26 <denkenz> and ofono_emulator_network_cb_t(struct ofono_error *error, int fd, const char *ip, const char **dns) or something like that
Apr 14 14:51:32 <denkenz> and we would obviously need changes to GAtPPP as well
Apr 14 14:53:25 <Gzajac> ok so we set the callback as soon as the emulator atom exists into connman plugin
Apr 14 14:53:48 <Gzajac> and then we call it when a data call is needed
Apr 14 14:54:18 <Gzajac> I would have to modify the ppp_net and to pass the fd to GAtPPP
Apr 14 14:55:25 <denkenz> yep
Apr 14 14:55:47 <Gzajac> ok I understand
Apr 14 14:55:58 <Gzajac> I will test it
Apr 14 14:57:40 <pnunes> jeevaka: ping
Apr 14 15:57:04 <Jeevaka> pnunes: pong
Apr 14 15:57:47 <pnunes> jeevaka: According the spec, it appears that the terminal response "UICC session terminated by the user" is not possible for Launch browser.
Apr 14 15:58:00 <pnunes> jeevaka: However, this kind of error is returned when the agent quits during ConfirmLaunchBrowser.
Apr 14 15:58:43 <pnunes> jeevaka: Denkenz thinks that this is an error in the spec. Could you confirm ? is there a test that checks for end session during launch browser confirmation?
Apr 14 16:00:59 <Jeevaka> as per the spec, I agree UICC session terminated by user is not one among the allowed responses for Launch Browser. Let me recollect why we added that
Apr 14 16:16:23 <denkenz> That has to be an error in the spec, probably because the spec mentions that a successful launch browser terminates the session
Apr 14 16:17:11 <Jeevaka> we wanted to keep it as its possible for agent to quit. Spec is not clear in this case. I don't think there was any test case related to this
Apr 14 16:17:57 <Jeevaka> as denkenz said we felt that it was an error in the spec
Apr 14 16:18:49 <denkenz> The spec is written by monkeys anyway
Apr 14 16:19:05 <denkenz> There are so many ambiguities its not funny
Apr 14 16:21:16 <pnunes> denkenz: so should I forget this point?
Apr 14 16:22:11 <denkenz> I say forget this for now
Apr 14 16:22:16 <Jeevaka> we will be still be compliant on this as there is no test case on this ;)
Apr 14 16:22:50 <denkenz> oFono is not even claiming compliance towards class C, since the rest of Class C is braindead
Apr 14 16:23:22 <pnunes> denkenz, jeevaka: OK, fine.
Apr 14 16:23:22 <Jeevaka> nod
Apr 14 23:40:16 <holtmann> We have ofono-0.47 and phonesim-1.13 now.
Apr 14 23:40:25 <holtmann> Please do not use ofono-0.46 since it is buggy.
**** ENDING LOGGING AT Fri Apr 15 02:59:57 2011