**** BEGIN LOGGING AT Tue Jul 19 02:59:56 2011
Jul 19 09:31:22 <vnayani> I want  to enable ISIDBG logs for my ISI modem. Are they enabled automatically if i run the oFono with -nd options or any special procedure.?
Jul 19 15:01:37 <fdanis> denkenz: ping
Jul 19 15:02:08 <denkenz> fdanis: pong
Jul 19 15:02:39 <fdanis> call indicator is set to when there is at least one active or held call
Jul 19 15:02:54 <fdanis> so it does not change in the intermediate state
Jul 19 15:03:42 <denkenz> ok, so nevermind call=0
Jul 19 15:03:47 <denkenz> but callheld does go to 2
Jul 19 15:04:02 <fdanis> yes, but this is allowed by specs
Jul 19 15:05:26 <denkenz> Section #?
Jul 19 15:06:18 <fdanis> sorry, not specs, tr002v15 allows this in "6.2.9.10 Three way calling from the HF (SEND 2)"
Jul 19 15:07:18 <fdanis> and we are able to pass the test with PTS when intermediate state is callheld=2
Jul 19 15:12:09 <denkenz> fdanis: But why send callheld=2, shouldn't you send callheld=1?
Jul 19 15:13:34 <denkenz> If the swap is the only case where we fail, then I'm fine with the hack
Jul 19 15:13:36 <denkenz> however,
Jul 19 15:13:37 <denkenz> -	data.status = call || held ? OFONO_EMULATOR_CALL_ACTIVE :
Jul 19 15:13:37 <denkenz> +	data.status = non_mpty || multiparty || held ?
Jul 19 15:13:38 <denkenz> +					OFONO_EMULATOR_CALL_ACTIVE :
Jul 19 15:13:39 <denkenz> and
Jul 19 15:13:47 <denkenz> -		data.status = call ? OFONO_EMULATOR_CALLHELD_MULTIPLE :
Jul 19 15:13:47 <denkenz> +		data.status = (non_mpty || multiparty) ?
Jul 19 15:13:48 <denkenz> +					OFONO_EMULATOR_CALLHELD_MULTIPLE :
Jul 19 15:13:53 <denkenz> I'd just keep them as is
Jul 19 15:14:01 <denkenz> The call condition is much clearer
Jul 19 15:14:17 <fdanis> currently, the indicators are only sent when their value changes, so if we hide the intermediate change, callheld=1 will not be sent (as it is already set to 1)
Jul 19 15:15:55 <fdanis> not sure to understand you, we should do difference between active call and a call part of an active multiparty
Jul 19 15:16:43 <denkenz> So the AG is always sending callheld=1 when we swap?
Jul 19 15:17:44 <fdanis> yes it should (4.31.2 in HFP spec)
Jul 19 15:20:32 <denkenz> Then forcing the update is better
Jul 19 15:21:25 <denkenz> Also, check the logic in drivers/hfpmodem/voicecall.c: ciev_callheld_notify
Jul 19 15:21:36 <denkenz> It sounds like it is wrong if callheld=2 transition is allowed
Jul 19 15:22:26 <fdanis> so adding a "force" parameter to ofono_emulator_set_indicator(), right ?
Jul 19 15:22:53 <denkenz> sounds fine to me yeah
Jul 19 15:23:43 <fdanis> but, I do not understand how I will be able to know that all calls are finished to be updated
Jul 19 15:24:36 <fdanis> driver knows it, as it is parsing the CLCC reply, but voicecall atom don't
Jul 19 15:25:08 <denkenz> most 'real' drivers are stateless, so even the driver doesn't know
Jul 19 15:25:24 <denkenz> e.g. peek at ECAV or XCALLSTAT ones
Jul 19 15:25:24 <fdanis> about hfpmodem, it can interact with device doing this intermediate change
Jul 19 15:26:04 <denkenz> I'm not so sure, take a peek at switch/case '2'
Jul 19 15:26:17 <denkenz> We drop active calls on such a transition
Jul 19 15:28:16 <fdanis> where, please
Jul 19 15:28:32 <denkenz> ciev_callheld_notify
Jul 19 15:28:38 <denkenz> line 1044
Jul 19 15:31:09 <fdanis> ok, but I think this does not follow tr002v15 flowchart
Jul 19 15:32:20 <denkenz> I know, probably an oversight of that particular case
Jul 19 15:33:20 <denkenz> Most of callheld=2 transitions are from 0->2
Jul 19 15:38:39 <fdanis> ok, I see how to hide intermediate state, but I do not see how to know when I should force to send the callheld=1
Jul 19 15:41:36 <denkenz> hah, good point
Jul 19 15:43:46 <denkenz> I still think we need to have transactions in the end, however, lets just go with your approach for now
Jul 19 15:44:17 <denkenz> Can you just add a comment to the last chunk with the else if on what it is trying to accomplish
Jul 19 15:44:37 <denkenz> And keep the call variable, that one is more readable
Jul 19 15:48:34 <fdanis> replacing non_mpty by call ?
Jul 19 15:49:50 <denkenz> -	data.status = call || held ? OFONO_EMULATOR_CALL_ACTIVE :
Jul 19 15:49:50 <denkenz> +	data.status = non_mpty || multiparty || held ?
Jul 19 15:49:50 <denkenz> +					OFONO_EMULATOR_CALL_ACTIVE :
Jul 19 15:49:51 <denkenz>  					OFONO_EMULATOR_CALL_INACTIVE;
Jul 19 15:49:53 <denkenz> Keep this one as is
Jul 19 15:50:06 <denkenz> -		data.status = call ? OFONO_EMULATOR_CALLHELD_MULTIPLE :
Jul 19 15:50:07 <denkenz> +		data.status = (non_mpty || multiparty) ?
Jul 19 15:50:07 <denkenz> +					OFONO_EMULATOR_CALLHELD_MULTIPLE :
Jul 19 15:50:07 <denkenz>  					OFONO_EMULATOR_CALLHELD_ON_HOLD;
Jul 19 15:50:09 <denkenz> Keep this one as is
Jul 19 15:50:28 <denkenz> Add a comment in front of the else if
Jul 19 15:50:30 <denkenz> and the rest is OK
Jul 19 15:50:41 <fdanis> ok, but in "else if (non_mpty > 1 || (non_mpty && multiparty)) " I need to differentiate them
Jul 19 15:51:41 <denkenz> yeah, you need the non_mpty and multiparty variables
Jul 19 15:51:51 <denkenz> I just don't want to grok them in places I do not need to
Jul 19 15:52:07 <fdanis> so use the 3  variables, right ?
Jul 19 15:53:09 <denkenz> yep
Jul 19 15:53:26 <fdanis> ok, I will send you the patch updated
Jul 19 16:25:38 <denkenz> fdanis: So do we pass all of PTS yet?
Jul 19 16:33:01 <fdanis> frederic dalleau uses PTS, and as far as I remember all tests are OK, but I am not sure of the PICS he used
Jul 19 16:33:31 <denkenz> ok, so what is left to do?
Jul 19 16:35:14 <fdanis> I have 2 sets of patches in the pipe: one about your comments on BLDN
Jul 19 16:36:28 <fdanis> and another to block sending of CIEV during AT command process (move it from voicecall atom to emulator atom)
Jul 19 16:37:38 <denkenz> OK, sounds good
Jul 19 16:37:52 <fdanis> I also wrote a patch to support AT+BIA (HFP 1.6), but I do not know if it can be usefull to integrate it now
Jul 19 16:38:19 <fdanis> the main task now will be the integration with audio part
Jul 19 16:38:34 <denkenz> Yep, but we need jhe and Vudentz for that
Jul 19 16:38:55 <denkenz> So we're pretty much done on the protocol side, right?
Jul 19 16:39:19 <fdanis> yes
Jul 19 16:39:24 <denkenz> Have we done ofono vs ofono testing yet? e.g. running hfp client against hfp server?
Jul 19 16:39:34 <fdanis> no
Jul 19 16:39:58 <denkenz> Lets do that since it is likely to uncover issues
Jul 19 16:40:25 <denkenz> After that we can mark this task as done
Jul 19 16:40:36 <fdanis> I think it can be useful top add support of AT+VGM/AT+VGS
Jul 19 16:40:55 <denkenz> Also, feel free to send a TODO patch for HFP 1.6 updates if you think them worthwhile
Jul 19 16:41:13 <denkenz> Yep, but those will probably be in a plugin
Jul 19 16:41:14 <fdanis> ok
Jul 19 16:42:08 <fdanis> I think they should be managed in emulator, as we should only store their value and pass them to audio part when we connect it
Jul 19 16:43:11 <denkenz> perhaps, but the audio pieces need to be figured out first
Jul 19 16:43:37 <denkenz> and we don't have a driver framework for emulators yet, so this might have to be a plugin
Jul 19 16:43:40 <denkenz> But we'll see
Jul 19 16:43:45 <fdanis> ok
Jul 19 16:44:28 <denkenz> we can always implement a dummy VGM/VGS if we need it for testing
Jul 19 16:44:54 <denkenz> I just want to make sure we have the protocol parts down
Jul 19 16:45:53 <fdanis> ok, as far as I know, this does not block any test
Jul 19 16:45:57 <fdanis> do you think I should send a patch for hfpmodem removing case 2 in ciev_callheld_notify ?
Jul 19 16:47:00 <denkenz> It sounds like it is wrong behavior at the moment
Jul 19 16:47:16 <denkenz> So yes, likely it is needed, but please confirm
Jul 19 16:47:52 <denkenz> this is why I want ofono vs ofono testing, we'll likely uncover several other bugs this way
Jul 19 16:48:04 <fdanis> ok
Jul 19 17:03:50 <denkenz> fdanis: For patch 2, it can happen that the sim is hot-swappable
Jul 19 17:06:59 <fdanis> yes, this is managed by OFONO_ATOM_WATCH_CONDITION_UNREGISTERED in sim_watch
Jul 19 17:10:28 <denkenz> Ok, I see what you're doing
Jul 19 17:11:14 <fdanis> and AT+BLDN will no more work as settings are closed in first patch
Jul 19 17:13:29 <denkenz> Yep, I forgot that emulators are created in mode 'offline'
Jul 19 17:13:36 <denkenz> So removing the SIM will remove these as well
Jul 19 17:14:11 <fdanis> sorry, not sure to understand your last sentence
Jul 19 17:16:06 <denkenz> hah ;)
Jul 19 17:16:17 <fdanis> those patches do not change that HFP emulator is created in mode 'offline'
Jul 19 17:16:24 <fdanis> did I miss something
Jul 19 17:16:40 <denkenz> I know, I thought we created the emulators in pre_sim
Jul 19 17:16:50 <denkenz> since we let them go live as soon as a voicecall atom is present
Jul 19 19:28:40 <dienet> hello, I've htc desire, can I use ofono (HFP) with it? Rigth now I'm stuck on ./enable-modem...
**** ENDING LOGGING AT Wed Jul 20 02:59:57 2011