**** BEGIN LOGGING AT Wed Jan 12 02:59:58 2011
Jan 12 10:37:26 <akiniemi> In src/network.c, what is NETWORK_REGISTRATION_MODE_MANUAL_AUTO for?
Jan 12 10:43:11 <demarchi> denkenz: what's the correct coding style in a if with only one line, but this line must wrap because of the 80-char limit?
Jan 12 10:43:18 <demarchi> denkenz: should i use braces or not?
Jan 12 10:50:20 <nbertrand> akiniemi: manual/auto (AT+COPS=4) mode means that if the manual registration fail  then it will be selected automatically
Jan 12 11:05:41 <akiniemi> nbertrand: Ah I see. It's not used anywhere, though.
Jan 12 13:01:06 <miojala> I'm having trouble with RadioSettings. For example set-gsm-band says method doesn't exist. Neither I cannot see the interface in question. Could somebody point out what's wrong?
Jan 12 13:17:21 <Jeevaka> miojala: may i know the output of running the ./list-modems script
Jan 12 13:21:12 <miojala> Jeevaka: Basic phonesim usage, but I can dump it in pastebin if you need
Jan 12 13:21:29 <demarchi> denkenz: holtmann: i need one more item to work on...
Jan 12 13:21:40 <demarchi> denkenz: holtmann: what's a good one from the TODO?
Jan 12 13:24:50 <Jeevaka> miojala: send me the pastebin link
Jan 12 13:26:08 <Jeevaka> miojala: phonesim doesn't support radio settings
Jan 12 13:26:36 <inz> Is there a way to hang up active call without activating a held call through oFono?
Jan 12 13:27:02 <inz> VoiceCall.Hangup seems to cause the held call to be activated
Jan 12 13:27:10 <miojala> jeevaka: right, so I need a HW which will do the trick. What triggers oFono to show RadioSettings interface?
Jan 12 13:28:09 <miojala> jeevaka: http://pastebin.com/me94cR6B
Jan 12 13:28:39 <Jeevaka> do you have any hardware with you
Jan 12 13:29:56 <miojala> Jeevaka: for example AT modem on serial cable, but I just need to figure out how to set it up with this udev system
Jan 12 13:30:13 <Jeevaka> miojala: introducing the support in phonesim can also be done
Jan 12 13:30:52 <Jeevaka> in ./plugins/phonesim.c there is a support done for ctm case
Jan 12 13:31:46 <miojala> thank you, I will look into this
Jan 12 13:32:16 <Jeevaka> you can refer ./drivers/ifxmodem/radio-settings.c
Jan 12 13:34:40 <kvehmanen> hello folks
Jan 12 13:35:01 <kvehmanen> I'm having issues with putting a conf call on hold
Jan 12 13:35:46 <kvehmanen> it seems atmodems perform the hold, but call status for all calls remains "active"
Jan 12 13:35:57 <kvehmanen> so there's not way to observe whether the conf call is on hold on not
Jan 12 13:50:53 <akiniemi> miojala: If there is a driver that upon probe() calls register, then the atom will appear.
Jan 12 13:51:29 <akiniemi> miojala: It's also possible that the driver only supports a subset of the functionality
Jan 12 13:51:44 <akiniemi> miojala: Like the isimodem one would support setting the rat mode, but not the band
Jan 12 13:52:15 <akiniemi> inz: how did the call originally go on hold?
Jan 12 13:59:51 <miojala> akiniemi: What USB modem would you recommend that oFono works with? I mean something that I could fetch from store and use as a reference HW.
Jan 12 14:01:19 <Gzajac> padovan: do you want the AT command document to be sent to you ?
Jan 12 14:08:55 <inz> akiniemi, I started another call
Jan 12 14:10:00 <inz> akiniemi, i.e. VCM.Dial; VCM.Dial; VC.Hangup on second call
Jan 12 14:10:36 <akiniemi> inz: at least on ISI, if you put the first call on hold first, then dial, the first call will stay held
Jan 12 14:11:00 <akiniemi> inz: if you dialing causes the first call to be held, then disconnecting will un-hold
Jan 12 14:11:09 <pessi> on isimodem, you can also disable the unhold-on-hangup feature
Jan 12 14:11:12 <akiniemi> inz: pessi might know if there's a spec on this
Jan 12 14:11:41 <akiniemi> miojala: denkenz might have a recommendation
Jan 12 14:12:06 <akiniemi> miojala: I use an N900 as a USB modem (with varying degrees of success ;) and a mini-PCI one as a backup
Jan 12 14:13:02 <akiniemi> miojala: the other one is a Dell 5530, aka Ericsson MBM F3705g
Jan 12 14:13:24 <akiniemi> miojala: There might be a USB adaptor available as well
Jan 12 14:14:14 <inz> akiniemi, pessi, do you know if it works the same way on AT modems (don't have any to test with)
Jan 12 14:15:04 <miojala> akiniemi: thanks, I'll try those options
Jan 12 14:16:23 <pessi> inz: I suppose the AT modems try to follow 22.030 section 6.5.5
Jan 12 14:16:52 <pessi> so they do not retrieve a call automatically
Jan 12 14:17:22 <pessi> perhaps isimodem driver should disable this automatic retreive-after-hangup feature
Jan 12 14:17:32 <miojala> akiniemi: great, RadioSettings interface found using n900 as modem
Jan 12 14:23:39 <kvehmanen> hmm, I guess the above problem issue with confcalls is a modem issue (e.g. I should get XCALLSTATs for each call when putting a multicall on hold)
Jan 12 14:33:29 <padovan> Gzajac: sure
Jan 12 14:50:30 <inz> pessi, akiniemi, btw, VCM.SwapCalls returns Error.Failure on N900/isiusb, even though the calls are swapped.
Jan 12 15:20:57 <pessi> inz: looks like all call control requests fail
Jan 12 15:35:23 <denkenz> demarchi: Example?
Jan 12 15:59:59 <pessi> akiniemi: ping?
Jan 12 16:11:49 <demarchi> denkenz: i've found some examples that don't use braces
Jan 12 16:11:55 <demarchi> so, i think this is the correct
Jan 12 16:12:43 <demarchi> denkenz: http://codepad.org/cckzlAtF
Jan 12 16:15:22 <denkenz> still confused about braces
Jan 12 16:15:42 <denkenz> what braces am I looking at? :)
Jan 12 16:16:23 <demarchi> denkenz: so, in this example there are no braces
Jan 12 16:16:33 <demarchi> denkenz: this was my question
Jan 12 16:16:43 <denkenz> ah for splitting on multiple lines?
Jan 12 16:16:52 <denkenz> Yeah you do it the usual way
Jan 12 16:17:00 <demarchi> should i put braces because the statement was wrapped because of the 80-char limit?
Jan 12 16:17:06 <denkenz> no
Jan 12 16:17:09 <demarchi> ok
Jan 12 16:17:20 <demarchi> i think it was another project that required braces
Jan 12 16:17:32 <demarchi> i don't remember what project though
Jan 12 16:18:26 <demarchi> ow... found it... it's webkit
Jan 12 16:23:09 <andras> my emergency call handling patches have been sitting on the mailing list since 20th of December last year, could you Denis have a peek at it?
Jan 12 16:34:10 <denkenz> andras: I have, I hate them, but I need to talk to Marko about them first
Jan 12 16:39:07 <fdanis> denkenz: holtmann: for the HFP AG emulator, should I do the same than for HFP HF code (Bluetooth link is managed by Bluez) or should I write Bluetooth code for link establishment in oFono ?
Jan 12 16:41:54 <denkenz> fdanis: For now do the link establishment in oFono
Jan 12 16:42:06 <denkenz> holtmann: You were fine applying those btio patches right?
Jan 12 16:42:34 <holtmann> denkenz: Yes.
Jan 12 16:44:40 <andras> denkenz:  this whole emergency stuff is a bit 'ugly' thing, but is needed
Jan 12 16:45:03 <denkenz> andras: Oh no doubt about it
Jan 12 16:45:19 <denkenz> however, I'm 99% convinced the current way is not going to work
Jan 12 16:45:20 <andras> denkenz: I tested the patches with both ISI and IFX modem
Jan 12 16:45:23 <denkenz> too many race conditions
Jan 12 16:45:43 <denkenz> We need to simplify the design, but I need the actual usecases first
Jan 12 16:49:50 <fdanis> denkenz: so I need btio in ofono, right ?
Jan 12 16:50:14 <denkenz> fdanis: Yep, check the mailing list archives.  Zhenhua Zhang already prepared one for submission to oFono
Jan 12 16:52:36 <andras> denkenz: the main idea is the user must succeed making an emergency call no matter what the modem state is: online/offline, transitioning from online to offline or vice versa
Jan 12 16:52:59 <denkenz> yeah, I question that assertion ;)
Jan 12 16:53:48 <andras> in addition, the modem should be left in the state last set by the user
Jan 12 16:54:22 <andras> this is the most one can expect
Jan 12 16:54:39 <andras> do you think is too much?
Jan 12 16:55:56 <denkenz> I think you need to let the user have some control
Jan 12 16:56:14 <denkenz> For instance, I'm in France where my voicemail dial string is 123
Jan 12 16:56:34 <denkenz> I move to weird country A and dial my voicemail
Jan 12 16:56:34 <fdalleau> denkenz:  I rebased Zhenhua Zhang work on latest head, should I resubmit the whole thing ?  I just noticed his btio is no more up to date, I can sync if necessary
Jan 12 16:56:41 <denkenz> suddenly I'm on emergency hotline
Jan 12 16:57:24 <denkenz> Or the other case, I let my kid play with my phone and put it in offline mode so he/she doesn't dial emergency call from the lock screen
Jan 12 16:57:43 <denkenz> I don't want to get slapped with $1000 bill for having the firemen come out
Jan 12 16:57:57 <denkenz> fdalleau: Yes, please resubmit
Jan 12 16:59:08 <andras> denkenz: those are valid points
Jan 12 17:01:44 <andras> but it may also make much sense to be able to make emergency calls easily, in a panic situation you may not have the ability to think much about the steps you need to do to make an emergency call
Jan 12 17:02:33 <andras> I think we have to agree on the use cases
Jan 12 17:03:10 <denkenz> exactly, I want to see / talk about the usecases first
Jan 12 17:03:40 <andras> like, when emergency calls shall be possibel, from which modem states
Jan 12 17:04:50 <denkenz> So my current thinking is to decouple the meaning of the Online property from the sim state
Jan 12 17:05:03 <denkenz> that way you can set the modem online even when there's no SIM / PIN locked
Jan 12 17:05:12 <denkenz> But then the UI makes the decision
Jan 12 17:05:18 <denkenz> (really the user)
Jan 12 17:06:37 <Jeevaka> denkenz, holtmann: inorder to test the tty flag patch, I also modified the phonesim to simulate the tty call. is it possible to integrate this changes?
Jan 12 17:06:55 <andras> denkenz: yes, Online has rather the meaning that the radio is turned ON on the modem
Jan 12 17:07:10 <andras> and that can happen also when there is no SIM
Jan 12 17:07:45 <andras> in fact that's the case for emergency calls without SIM
Jan 12 17:08:34 <denkenz> I know ;)
Jan 12 17:09:06 <denkenz> Jeevaka: Feel free to post them
Jan 12 17:09:55 <denkenz> andras: however, someone needs to make the call whether this approach will satisfy the usecases you have
Jan 12 17:10:27 <denkenz> Or whether those usecases are invalid
Jan 12 17:10:55 <Jeevaka> denkenz: there is duplication of the atmodem/voicecall.c whole file for this case alone
Jan 12 17:11:21 <denkenz> Jeevaka: Um no ;)
Jan 12 17:11:34 <denkenz> Jeevaka: Can't you quirk it?
Jan 12 17:12:49 <Jeevaka> i'll try it out
Jan 12 17:16:01 <Jeevaka> denkenz: any comments on "add additional informartion for mandatory general result" and "LaunchBrowser" patch?
Jan 12 17:16:26 <denkenz> Haven't looked at anything yet, been doing administrative items ;)
Jan 12 17:17:14 <andras> denkenz: do we need to define a new modem state due to decoupling the Online property from SIM state, or it's enough to assume Online property is just reflecting the radio state?
Jan 12 17:19:09 <andras> then we need to see when the post_online function can be called
Jan 12 17:19:19 <denkenz> I think we can keep the current modem states
Jan 12 17:28:48 <fdalleau> denkenz: it's gone, as you may notice, we are still experimenting with git send-email, sorry for the mess
Jan 12 17:31:06 <andras> denkenz: let's take the possible usecases one by one:
Jan 12 17:31:29 <andras> 1. No SIM - emergency call OK
Jan 12 17:32:40 <andras> 2. SIM present, but locked - emergency call OK
Jan 12 17:33:24 <andras> 3. SIM present, not locked, but modem in Offline state
Jan 12 17:34:04 <andras> 4. SIM present, not locked, modem in Online state - emergency call OK
Jan 12 17:35:07 <andras> I think these are all the possible combinations
Jan 12 17:35:32 <BertrandAygon> In No SIM case, modem can be online or offline, no?
Jan 12 17:36:02 <andras> correct
Jan 12 17:36:08 <andras> so we have on more case
Jan 12 17:36:29 <BertrandAygon> sorry ;)
Jan 12 17:36:46 <andras> thanks for your remark
Jan 12 17:37:01 <BertrandAygon> And one more question, sorry I am not so familiar with oFono, by locked, you mean PIN locked?
Jan 12 17:37:30 <andras> 1a. No SIM, modem in Offline state - emergency call OK
Jan 12 17:37:54 <andras> 1b, No SIM, modem in Online state - emergency call OK
Jan 12 17:39:16 <denkenz> andras: Why are you concerned about case 3?
Jan 12 17:39:57 <andras> case 3. is where we can argue whether we want to change the modem state automatically to Online or not
Jan 12 17:40:16 <andras> when making the emergency call
Jan 12 17:40:34 <denkenz> Actually, you also have case 2b, Sim present, PIN locked, Offline
Jan 12 17:40:35 <andras> or whether we shall let the user decide
Jan 12 17:40:49 <denkenz> And the automatic onlining is an issue for 1b, 2b and 3
Jan 12 17:42:01 <andras> for 1b and 2b yes
Jan 12 17:42:08 <denkenz> In fact, with ofono today we have no 1a or 2a case
Jan 12 17:42:24 <denkenz> er sorry mixing them up
Jan 12 17:42:24 <andras> I mean is not an issue, because we want automatic Online
Jan 12 17:42:44 <denkenz> I'm not so sure
Jan 12 17:42:57 <andras> but case 3 is with question mark
Jan 12 17:43:16 <denkenz> The firemen counterpoint applies to those as well
Jan 12 17:43:59 <denkenz> If you want a true panic mode, it should be a dedicated red button on the phone hardware
Jan 12 17:45:15 <andras> the more decision responsability is transferred to the user, the easier the implementation, but we have make sure, we are doing the right thing
Jan 12 17:47:10 <andras> so your opinion is that the user must explicitly turn the modem Online when making an emergency call?
Jan 12 17:47:29 <andras> in all the cases
Jan 12 17:48:32 <BertrandAygon> i think you can ask to turn on or off modem before dialing, but you should not ask for PIN code
Jan 12 17:49:19 <crevetor> Hi guys
Jan 12 17:49:36 <crevetor> Do you know if ofono has been tested on an n900 with 2.6.35 ?
Jan 12 17:50:09 <andras> yes, PIN code will never be asked when making emergency call
Jan 12 17:50:22 <denkenz> andras: The user has to make the decision
Jan 12 17:50:41 <denkenz> Lets put it this way, my kid accessed the emergency screen more often than I ever have in my entire life
Jan 12 17:51:47 <courmisch> IIRC, the EU CE requirement is that best efforts be made to route emergency calls
Jan 12 17:52:02 <courmisch> requiring the phone online is not best efforts since we can easily avoid it
Jan 12 17:52:54 <denkenz> routing is a different question entirely
Jan 12 17:53:02 <crevetor> or do you guys know if /sys/devices/platform/gpio-switch should still be there in 2.6.35 ?
Jan 12 17:53:09 <BertrandAygon> do not give your phone to your kid
Jan 12 17:53:20 <courmisch> I meant, get it to work, not route it through the network
Jan 12 17:54:58 <denkenz> Bottom line is this is a user policy decision
Jan 12 17:55:12 <denkenz> if you want failsafe emergency calls, make a phone with a red button
Jan 12 17:55:17 <denkenz> I personally would never buy one
Jan 12 17:55:44 <courmisch> that's a very odd and unwarranted interpretation of the CE requirement
Jan 12 17:58:23 <andras> well, the Nokia phones let you make emergency calls any time, you do not need to turn on the radio explicitly, it is done transparently, and I think that's what I would call 'best effort'
Jan 12 17:59:01 <courmisch> of course. In an emergency, checking that the phone is online is not really the thing everyone thinks first
Jan 12 17:59:03 <holtmann> andras: Do you have a keyboard or are they all touch screen only?
Jan 12 17:59:12 <courmisch> people think "911, err... no, 112"
Jan 12 17:59:31 <holtmann> courmisch: Nobody says that the user needs to check, but it might be good that it gets asked to confirm to make an emergency call.
Jan 12 17:59:47 <denkenz> Last I checked oFono was not Nokia specific
Jan 12 18:00:03 <courmisch> holtmann: maybe, but that's a different issue than online mode. And it's entirely a UI non-oFono issue then
Jan 12 18:00:33 <courmisch> CE requirements are not Nokia specific either
Jan 12 18:00:36 <denkenz> Bottom line is, the current approach is full of races and too limiting
Jan 12 18:14:17 <andras> denkenz: in what sense the current approach is too limiting?
Jan 12 18:15:23 <xtor> holtmann: I got my Sierra modem recognised, and listed by test/list-modems I assume the missing information is ok with the current plugin state? http://pastebin.com/RT7gTvJ0
Jan 12 18:15:31 <xtor> if so, where should I focus right now?
Jan 12 18:15:51 <xtor> s/state/status/
Jan 12 18:16:11 <holtmann> xtor: So do you have 3 modems attached to it? Or just three interfaces?
Jan 12 18:16:42 <holtmann> You can not create separate modems for each interface. It is physically the same modem.
Jan 12 18:20:22 <denkenz> andras: Today we allow online mode in sim not inserted / pin locked cases only when an emergency call happens
Jan 12 18:21:08 <denkenz> andras: In theory some manufacturer might want to have a policy of always being online in those cases (e.g. to receive network broadcast emergency numbers)
Jan 12 18:21:40 <andras> ok, I see
Jan 12 18:23:10 <andras> that means we need the possibility to turn on/off the modem radio independently of the SIM presence and SIM lock state
Jan 12 18:24:04 <andras> I meant by SIM lock  PIN lock
Jan 12 18:24:23 <xtor> holtmann: 1 embedded modem with three different interfaces, two for commands, one for the connection itself
Jan 12 18:25:27 <holtmann> That is still only one modem.
Jan 12 18:25:41 <xtor> yep
Jan 12 18:25:49 <holtmann> So it needs to be show up as ONE modem.
Jan 12 18:27:52 <xtor> how are these devices handled in ofono? which similar plugin should I look at?
Jan 12 18:28:47 <andras> crevetor: I don't know about the ofono status on N900 with the 2.6.35 kernel, I am not using that setup, but I can ask my colleagues tomorrow and give you the answer, if no one replies meantime
Jan 12 18:35:58 <holtmann> xtor: MBM, HSO etc.
Jan 12 18:36:09 <holtmann> xtor: Huawei and others.
Jan 12 18:51:13 <Jeevaka> denkenz: holtmann: I dont see mpty of ofono_call used anywhere apart from where it is set. correct me if i'm wrong
Jan 12 18:52:01 <denkenz> yes, pointless property
Jan 12 18:52:25 <denkenz> half the vendor commands don't even report it
Jan 12 18:54:58 <Jeevaka> can I go ahead and remove the mpty from ofono_call structure and handling done in at_util_parse_clcc?
Jan 12 18:59:51 <denkenz> sure, why not
Jan 12 19:00:11 <Jeevaka> or is there any possibility to make use of this?
Jan 12 19:01:37 <denkenz> If we haven't in two years, I doubt it
Jan 12 19:01:48 <denkenz> Unless someone has an idea how this can be used
Jan 12 19:02:24 <Jeevaka> ok
Jan 12 19:07:11 <holtmann> denkenz: We could remove the D-Bus property, but leave the mpty internal variable in place.
Jan 12 19:07:23 <holtmann> Lets see if someone actually bothered using the property.
Jan 12 19:07:56 <denkenz> The D-Bus property does not use the internal variable
Jan 12 19:08:05 <denkenz> oFono figures this out itself
Jan 12 19:08:55 <holtmann> Does the UI need this property? I bet we talked about this before and I just forgot.
Jan 12 19:09:16 <denkenz> It might for grouping purposes
Jan 12 19:09:30 <denkenz> e.g. drawing a box around mpty calls
Jan 12 19:10:29 <holtmann> So we just should remove the internal mpty code?
Jan 12 19:10:47 <denkenz> I think so
Jan 12 19:11:00 <denkenz> it is never referenced and some drivers just fudge it
Jan 12 19:11:01 <holtmann> ISI is not using that either as it seems.
Jan 12 19:16:29 <pessi> MultiParty?
Jan 12 19:17:11 <pessi> Kai is trying to get conferences working, he probably can comment it
Jan 12 19:17:48 <pessi> at least ifx has some serious trouble with multiparty
Jan 12 19:18:26 <holtmann> pessi: To be honest, I have never tested that part with IFX.
Jan 12 19:18:48 <holtmann> pessi: If you have OFONO_AT_DEBUG=1 trace for the issues, please post them. Then we can get them fixed.
Jan 12 19:19:46 <pessi> I'll ask Kai to send e-mail tomorrow
**** ENDING LOGGING AT Thu Jan 13 02:59:58 2011