**** BEGIN LOGGING AT Fri Oct 22 02:59:58 2010
Oct 22 05:32:16 <matt_chen> holtmann:ping
Oct 22 07:16:42 <holtmann> matt_chen: pong
Oct 22 07:21:15 <matt_chen> haltmann : is the /etc/ofono/modem.conf required ?
Oct 22 07:23:00 <matt_chen> haltmann : I found if the ofonod runs without this file, it can not detect and read the correct path of Device. Such as /dev/ttyUSB0.
Oct 22 07:25:58 <akiniemi> denkenz: no problem
Oct 22 07:27:20 <akiniemi> denkenz: I also removed the check in network.c on strength_notify(). I had to put an ugly hack in the isimodem driver due to that, because it seems RSSI is reported *before* reg status...
Oct 22 07:28:34 <akiniemi> denkenz: Can't think of what can break by removing the check, but if it did, I have some ideas on how to fix it in a more elaborate way. :)
Oct 22 07:29:29 <akiniemi> holtmann: gotcha on DGB use; we'll try to debug all we need in the drivers
Oct 22 07:30:01 <holtmann> matt_chen: No. Detection should be done via udev.
Oct 22 07:30:36 <holtmann> akiniemi: I didn't see it initially as well, but then I had isimodem disabled to save some binary size ;)
Oct 22 07:31:01 <akiniemi> holtmann: ahem ;)
Oct 22 07:31:14 <holtmann> Same as my screwup with the compiler warnings that I disabled when building the stupid 32-bit binaries :(
Oct 22 07:32:18 <holtmann> akiniemi: There was a bug in the USB gadget kernel driver and I couldn't copy to big binaries over scp ;)
Oct 22 07:34:21 <matt_chen> haltmann : USB kernel driver bug ?
Oct 22 07:35:31 <matt_chen> holtmann : which version of ofono is working much stable ? do you have any suggestion ? I am using 0.32 now. ;-)
Oct 22 07:35:50 <holtmann> 0.32 is good. I am using it daily.
Oct 22 07:35:59 <holtmann> However it depends a bit on which modem you wanna use it.
Oct 22 07:36:23 <holtmann> matt_chen: Intel platform driver USB host/device support. Is fixed now.
Oct 22 07:37:29 <matt_chen> holtmann : I am using MeeGo v.1.0
Oct 22 07:37:51 <holtmann> That is a bit out-dated.
Oct 22 07:37:59 <matt_chen> holtmann : the kernel is 2.6.33.3. is this what you mentioned ?
Oct 22 07:38:26 <holtmann> What hardware do you wanna use oFono for. 3G hardware.
Oct 22 07:38:38 * matt_chen look up the 3G module.
Oct 22 07:39:05 <matt_chen> holtmann : Huawei (12d1:1404)
Oct 22 07:40:46 <matt_chen> holtmann : I can make it work, but it just needs to read the correct /dev/ttyUSBx
Oct 22 07:40:57 <holtmann> That is an udev configuration file.
Oct 22 07:41:08 <holtmann> It comes with oFono. plugins/ofono.rules.
Oct 22 07:41:25 <matt_chen> I see.
Oct 22 07:42:10 <matt_chen> but sometimes ttyUSB0 can not work, I need to change the /etc/ofono/modem.conf to use a specific path, such as /dev/ttyUSB1.
Oct 22 07:46:34 <holtmann> It will work.
Oct 22 07:46:43 <holtmann> As I said, it is detected via udev.
Oct 22 07:46:56 <holtmann> You do NOT needed modem.conf.
Oct 22 07:49:43 <matt_chen> ok...so about you mentioned the "Intel platform driver USB host/device support", is there any patch for that ?
Oct 22 07:54:03 <holtmann> Most be. The newer kernels that I have has this fixed. I did not look into the details.
Oct 22 07:54:09 <holtmann> s/Most/Must/
Oct 22 07:54:49 <holtmann> Besides the point btw. It is a dev platform anyway. I was just making excuses for being stupid yesterday ;)
Oct 22 13:08:30 <denkenz> akiniemi: In general we try to only report properties where they make sense, e.g. technology only after you're registered, etc.  But for signal strength it might be OK
Oct 22 13:08:33 <denkenz> Lets see how it works out
Oct 22 13:55:59 <emdete> hi! i try to use ofono with huawei EM770W. the modem properties Online=0 and the simmanager Present=0 (the sim is inserted). is the modem known to work?
Oct 22 13:57:00 <denkenz> is the modem powered?
Oct 22 13:57:03 <emdete> yes
Oct 22 13:57:10 <emdete> hi denkenz btw :)
Oct 22 13:57:25 <emdete> i used some py dbus code from the samples to enable power
Oct 22 13:57:28 <denkenz> and the obvious question, is the sim inserted?
Oct 22 13:57:33 <emdete> yes
Oct 22 13:57:34 <emdete> :D
Oct 22 13:58:07 <denkenz> so Huawei firmware is pretty craptastic about telling us when the sim is ready
Oct 22 13:58:12 <denkenz> Do you have a PIN on your sim by any chance?
Oct 22 13:58:19 <emdete> yes
Oct 22 13:58:46 <emdete> so the sim manager shows up /after/ pin entry?
Oct 22 13:58:48 <denkenz> So last I checked the current huawei drivers don't work well when SIM PIN is enabled
Oct 22 13:59:14 <denkenz> no, but we rely on the modem to tell us a sim is inserted or not
Oct 22 13:59:38 <emdete> have you been involved into wetab development?
Oct 22 13:59:48 <denkenz> nope, no idea what that is
Oct 22 14:00:23 <emdete> a atom-based tab with that modem
Oct 22 14:00:38 <emdete> intel is pushing it a bit ;)
Oct 22 14:01:21 <emdete> http://wetab.mobi/en
Oct 22 14:01:54 <denkenz> nice, but looks like its android
Oct 22 14:01:57 <emdete> megoo based, ofono installed but not used. they enable it successfully with shellscript/chat and wvdial :/
Oct 22 14:02:09 <emdete> no, not really
Oct 22 14:02:18 <emdete> it shall one day run android apps
Oct 22 14:02:31 <emdete> today it's a mix of megoo and gtk-stuff
Oct 22 14:02:44 <denkenz> really? does it run oFono? :)
Oct 22 14:03:08 <emdete> yes, i splitted disk, installed debian, installed ofono and it works - until the point shown above
Oct 22 14:03:32 <emdete> in the wetab-os ofono is installed but not used
Oct 22 14:03:44 <denkenz> interesting
Oct 22 14:04:04 <emdete> yes. nice hw but i need the modem to work, so i tried your ofono :D
Oct 22 14:04:25 <denkenz> Someone just needs to give some love to the huawei driver
Oct 22 14:04:40 <denkenz> Basically everything else works except for the silly SIM pin enabled case
Oct 22 14:04:56 <emdete> really?
Oct 22 14:05:01 <emdete> how can i help?
Oct 22 14:05:26 <denkenz> yeah, we were doing voice calls, sms, gprs with every Huawei dongle around
Oct 22 14:05:50 <denkenz> Well, get familiar with plugins/huawei.c
Oct 22 14:06:03 <denkenz> And figure out why ofono_sim_inserted isn't getting called at the right time
Oct 22 14:06:05 <emdete> using bt for voice calls with that tab would be amazing...
Oct 22 14:06:41 <emdete> could one for a first shot answer with "yes" all the time?
Oct 22 14:06:42 <denkenz> Yep, some Huaweis come with special firmware that does voice routing (Vodafone only)
Oct 22 14:07:13 <denkenz> emdete: something you can try
Oct 22 14:07:30 <denkenz> huawei has been touched and re-touched and nobody made it work nicely yet
Oct 22 14:07:40 <denkenz> I don't have time for it unfortunately :(
Oct 22 14:07:57 <emdete> time? what't that? :D
Oct 22 14:08:10 <denkenz> yeah exactly :)
Oct 22 14:08:54 <emdete> weekend is coming, maybe... are you around at weekend for some questions?
Oct 22 14:09:54 <denkenz> I usually have irc open, so yes but with delayed response time
Oct 22 14:11:06 <emdete> is gps included for that modem in ofono too?
Oct 22 14:11:33 <denkenz> we haven't had time to figure out the gps bits
Oct 22 14:11:53 <denkenz> gps will probably be supported by the usual suspects, but oFono will allow power on / off of the GPS device if it is on board the modem
Oct 22 14:12:15 <denkenz> Its even in our TODO list I think
Oct 22 14:15:43 <kvalo> emdete: I have been trying to get huawei work, but like denkenz said pin query is still broken
Oct 22 14:16:25 <emdete> kvalo: do you know more where the problem can be?
Oct 22 14:17:55 <kvalo> emdete: most probably the sim_state handling in plugins/huawei.c. 
Oct 22 14:18:49 <kvalo> denkenz was saying that huawei should poll the sim state just like some other plugin does. was it mbm?
Oct 22 14:18:57 <emdete> what do you think about (for fast workaround) enable sim insterted always?
Oct 22 14:19:17 <denkenz> it should work of course
Oct 22 14:19:24 <kvalo> yeah, simpin_check() in mbm.c polls the state
Oct 22 14:20:00 <kvalo> emdete: but that most breaks the cold start case
Oct 22 14:20:11 <kvalo> emdete: ie. you plug in the modem when ofono is running
Oct 22 14:20:41 <emdete> let me check...
Oct 22 14:21:59 <kvalo> emdete:just put the sim on some phone and disable pin query. that way you should get the modem working
Oct 22 14:22:41 <emdete> kvalo: that sim doesnt allow that ;)
Oct 22 14:24:17 <kvalo> heh
Oct 22 14:24:36 <denkenz> well he's not even getting to sim_inserted
Oct 22 14:24:44 <denkenz> so something with our SYSINFO polling is not quite right
Oct 22 14:25:53 <kvalo> emdete: can you send logs?
Oct 22 14:26:16 <kvalo> emdete: with at debug
Oct 22 14:28:03 <denkenz> holtmann: Btw, go ahead and make the release, neither of the patch sets I talked about yesterday are ready
Oct 22 14:29:38 <emdete> kvalo: sure - adr?
Oct 22 14:30:53 <denkenz> just pastebin it
Oct 22 14:31:07 <kvalo> emdete: remember OFONO_AT_DEBUG=1
Oct 22 14:31:15 <denkenz> Feel free to scrub personal details if any
Oct 22 14:31:40 <emdete> kvalo: always on ;)
Oct 22 14:32:24 <kvalo> ah, your modem most probably is in HUAWEI_SIM_STATE_INVALID_OR_LOCKED and that's why sim is not inserted
Oct 22 14:32:45 <kvalo> it's totally stupid to combine invalid and locked states into one
Oct 22 14:32:59 <denkenz> not necessarily
Oct 22 14:33:11 <denkenz> invalid does not mean not present
Oct 22 14:37:09 <kvalo> but I think that huawei should do that similar polling like mbm does. the current method of following sim_state just doesn't work properly, too many corner cases
Oct 22 14:42:11 <denkenz> kvalo: which polling?
Oct 22 14:42:19 <denkenz> there are two separate issues here
Oct 22 14:42:37 <denkenz> one is sim is inserted and we can read from it
Oct 22 14:42:50 <denkenz> the other one is pin is accepted and we can read the pin-protected entries
Oct 22 14:43:01 <denkenz> both have race condition issues
Oct 22 14:46:25 <kvalo> denkenz: I mean that we would poll with CPIN or similar command to see when the huawei modem has initialised itself properly
Oct 22 14:46:57 <kvalo> denkenz: currently we are trying to guess that from sim state, but it's not reliable
Oct 22 14:47:43 <kvalo> denkenz: I'm travelling this and next week, so I don't have logs right now to show the problem
Oct 22 14:48:09 <denkenz> I think the reason we use SYSINFO and SIMST is that they're actually supposed to be used for the purpose
Oct 22 14:48:14 <denkenz> At least from the AT command docs
Oct 22 14:48:31 <denkenz> But you might be right and they simply don't work well enough
Oct 22 14:49:07 <emdete> kvalo: http://mister-muffin.de/p/8U-f.txt
Oct 22 14:50:56 <emdete> the properties i get from manager.GetModems are these the same as the one from modem.GetProperties?
Oct 22 14:59:54 <denkenz> yes
Oct 22 15:02:04 <emdete> hm, i added a ofono_sim_inserted_notify after line 588. now i get a Present=True. 
Oct 22 15:02:47 <kvalo> emdete: thanks, just as I suspected:
Oct 22 15:02:50 <kvalo> Oct 22 13:47:16 metab ofonod[1162]: plugins/huawei.c:notify_sim_state() 0
Oct 22 15:03:02 <kvalo> that's the invalid and locked state
Oct 22 15:03:30 <kvalo> should we add a new check for that state to see if the sim is either invalid or locked?
Oct 22 15:06:20 <emdete> should i see an org.ofono.NetworkRegistration in modem
Oct 22 15:06:29 <emdete> properties interfaces?
Oct 22 15:08:44 <denkenz> kvalo: So invalid in 3GPP means that the SIM is present but not responding
Oct 22 15:09:04 <denkenz> kvalo: I think in the mean time you can treat invalid_or_locked as present
Oct 22 15:09:15 <denkenz> emdete: Only after you go online
Oct 22 15:09:23 <denkenz> emdete: test/enable-online
Oct 22 15:11:05 <emdete> YES!
Oct 22 15:11:17 <emdete> we are on! great, guys, thanks alot!
Oct 22 15:12:05 <emdete> have u any experiences with cell-ids in 3g networks?
Oct 22 15:12:38 <emdete> does the huawai report that correctly? i.e. by a cellid > 0x10000?
Oct 22 15:13:32 <denkenz> no idea really
Oct 22 15:13:50 <emdete> sure, but one step forward 
Oct 22 15:14:26 <emdete> ah, sry. thought you meant that my solution for SIM inserted was "not ideal" :D
Oct 22 15:14:50 <emdete> okya, will investigate later, have to leave now
Oct 22 16:36:04 <kvalo> denkenz: what's the preferred way of reporting ofono bugs?
Oct 22 16:36:15 <denkenz> mailing list?
Oct 22 16:36:36 <denkenz> along with a patch that fixes it ;)
Oct 22 16:42:42 <kvalo> haha :)
Oct 22 20:02:35 <emdete> denkenz: i have 2 issues, wasy to fix. interested? 
Oct 22 20:02:52 <denkenz> sure
Oct 22 20:05:09 <emdete> one is the manpage, it mentiones --debug, -d ...nothing behind. all other manpage mostly say stuff like --suffix=SUFFIX. this is especailly ugly, cause the argument isn't checked. i struggled when i used `ofonod --debug --nodetach` which in fact detaches. took me a while to figure that out...
Oct 22 20:05:36 <emdete> the old manpage not even mentioned the parameter in the text below, that's fixed.
Oct 22 20:06:12 <emdete> the other issue is about the init.d-script. there is no dependency on dbus mentioned so ofono starts before and is unable to connect to dbus & fails.
Oct 22 20:06:43 <emdete> so i added "Required-Start:   $syslog dbus" and it works
Oct 22 20:08:00 <denkenz> send both fixes to the mailing list
Oct 22 20:08:11 <emdete> which?
Oct 22 20:08:31 <denkenz> Well the init.d fixes for one
Oct 22 20:08:54 <emdete> :D no: which mailing list? may i post without being subscribed?
Oct 22 20:09:40 <denkenz> ofono@ofono.org
Oct 22 20:09:55 <denkenz> And I think you need to be subscribed
Oct 22 20:10:17 <denkenz> The --debug part is funny, we always run it -n -d
Oct 22 20:10:21 <denkenz> Never the other way around
Oct 22 20:10:46 <emdete> try ;)
Oct 22 20:11:15 <emdete> maybe it's different now. i tested an quite old version.. let me check
Oct 22 20:11:15 <denkenz> ofonod -d -n works
Oct 22 20:11:35 <denkenz> So does --debug -n
Oct 22 20:12:40 <emdete> fine, than that was fixed. anyway the manpage should then use a -d bla|--debug[=bla] or similar
Oct 22 20:14:03 <denkenz> yep, please send a patch for that
Oct 22 20:14:47 <emdete> :P
Oct 22 20:15:26 <denkenz> You found it, might as well get the credit
Oct 22 20:19:10 <emdete> take the credit ;)
Oct 22 20:20:32 <denkenz> git shortlog -s -n
Oct 22 20:20:34 <denkenz>   1954  Denis Kenzior
Oct 22 20:20:34 <denkenz>    609  Marcel Holtmann
Oct 22 20:20:34 <denkenz>    251  Andrzej Zaborowski
Oct 22 20:20:39 <denkenz> I think I have enough ;)
Oct 22 20:22:07 <holtmann> denkenz: Did you get a chance to look into the multiple context stuff?
Oct 22 20:22:15 <holtmann> I am a bit lost in that code.
Oct 22 20:22:29 <denkenz> yeah, its missing some bits
Oct 22 20:22:43 <denkenz> nothing major, but I need a few quiet hours
Oct 22 20:22:57 <holtmann> So do we want a release with it or without it?
Oct 22 20:23:10 <denkenz> Do one without it for now
Oct 22 20:23:22 <holtmann> It would be a nice feature though since we have now hardware to test it on.
Oct 22 20:23:32 <denkenz> If I'm going to touch that part, I also want to fix DeactivateAll
Oct 22 20:23:36 <denkenz> and that takes longer
Oct 22 20:23:45 <holtmann> Okay. Then I am doing a release without it.
Oct 22 20:24:02 <holtmann> I did add two context atom within IFX driver. Does that look sane to you?
Oct 22 20:24:38 <denkenz> yeah that looks good
Oct 22 20:24:54 <denkenz> Just remember that only GPRS2 is going to be used ATM
Oct 22 20:25:06 <holtmann> I know that. Which is fine.
Oct 22 20:25:22 <holtmann> The IFX raw IP gives you "invalid" as interface name anyway ;)
Oct 22 20:25:38 <denkenz> heh
Oct 22 20:25:57 <holtmann> And currently I don't wanna look into the async and buffering for running GAtRawIP in user space.
Oct 22 20:26:10 <holtmann> It does work, but I need some clear head to think about it a bit more.
Oct 22 20:26:44 <denkenz> nod, then push that release and we fix multi context in the next one
Oct 22 20:26:59 <holtmann> And I do wanna do some testing in see how far we would get with using N_GSM upstream line discipline.
Oct 22 20:27:07 <holtmann> Do you have anything else pending?
Oct 22 20:27:18 <holtmann> Like the SIM ready stuff? Or does that need work?
Oct 22 20:28:03 <denkenz> Pekka's set I don't completely agree with
Oct 22 20:28:20 <denkenz> So I think more discussion required
Oct 22 20:28:45 <denkenz> If you feel up to it, you can apply my patch from August
Oct 22 20:28:51 <denkenz> And test with IFX
Oct 22 20:28:59 <denkenz> I'm pretty sure it should work
Oct 22 20:29:20 <holtmann> Nope. We do that in the next release then.
Oct 22 20:29:35 <holtmann> I have looked at IFX a bit too much this week.
Oct 22 20:29:40 <denkenz> agreed, that stuff works anyway
Oct 22 20:29:47 <holtmann> I am just happy that all works just fine.
Oct 22 20:31:13 <denkenz> Do note that with FDN, you also need the patch from phonesim
Oct 22 20:31:26 <denkenz> So you might want to do a phonesim release too
Oct 22 20:31:36 <holtmann> Sounds like a good idea.
Oct 22 20:32:04 <denkenz> That one has the nice scripting features to boot
Oct 22 20:32:20 <denkenz> s/has/will have/
Oct 22 20:32:27 <holtmann> Yep. I am getting to that one now within the next hours and then I call it a week.
Oct 22 20:38:22 <emdete> can i determine if the modem uses 2g or 3g?
Oct 22 20:38:39 <denkenz> emdete: for what purpose
Oct 22 20:38:48 <emdete> ?
Oct 22 20:39:08 <emdete> i jsut need to know 
Oct 22 20:39:24 <denkenz> Not through D-Bus
Oct 22 20:41:37 <emdete> any other way? could it be exposed?
Oct 22 20:42:13 <denkenz> Not without a usecase
Oct 22 20:42:30 <emdete> i want the dev as a testing device
Oct 22 20:42:41 <emdete> need cell information, location, stuff like that
Oct 22 20:42:47 <emdete> develop agps test suites and on
Oct 22 20:43:45 <denkenz> Neighbor Cell Info and aGPS is something that is on the TODO list
Oct 22 20:43:59 <denkenz> however, nobody has done the work yet
Oct 22 20:48:29 <emdete> not that far... i just need cell information of the current cell with 3g/2g differenciation
Oct 22 20:49:59 <denkenz> If that's what you want, then you should be able to grab that from NetworkRegistration interface
Oct 22 20:50:10 <denkenz> see CellId / LocationAreaCode and Technology properties
Oct 22 20:51:00 <emdete> yes, i have to check if the cell-id is in fact an utran for 3g then i'm done
Oct 22 21:12:23 <sabotage> running into wierdness on MeeGo trunk builds... ofonod starts, but it's not visible on the system bus... 
Oct 22 21:12:26 <sabotage> known issue?
Oct 22 21:12:40 * sabotage finally getting back to upstream work
Oct 22 21:12:55 <sabotage> denkenz: ^^^
Oct 22 21:13:20 <sabotage> no obvious errors in debugging output
Oct 22 21:13:35 <denkenz> Sounds like the issue raji reported earlier where ofonod was crashing
Oct 22 21:13:46 <sabotage> but it's not crashing
Oct 22 21:14:08 <denkenz> Are you seeing 'System bus disconnected' in your ofono logs?
Oct 22 21:14:15 <sabotage> gets through drivers/atmodem/sim.c:at_crsm_read_cb() and then stops
Oct 22 21:14:43 <denkenz> are you using upstream?
Oct 22 21:14:45 <sabotage> on Meego 1.1, same thing, except that I can see org.ofono on the system bus and not on 1.2
Oct 22 21:14:46 <denkenz> update phonesim
Oct 22 21:15:04 <sabotage> denkenz: upstream, yes
Oct 22 21:15:19 <sabotage> phonesim, using whatever package is in trunk (1.2)...
Oct 22 21:15:21 <denkenz> However, that does not explain your system bus issue
Oct 22 21:15:32 <sabotage> nod
Oct 22 21:15:41 <sabotage> I see the phonesim driver an plugins load fine
Oct 22 21:15:52 <emdete> sabotage: maybe ofono starts before dbus? i had that here...
Oct 22 21:16:02 <sabotage> I saw that above emdete 
Oct 22 21:16:04 <denkenz> sabotage: You need latest upstream phonesim, or the following change
Oct 22 21:16:06 <denkenz>  <file name="EFest">
Oct 22 21:16:07 <denkenz> -    07
Oct 22 21:16:07 <denkenz> +    00
Oct 22 21:16:07 <denkenz>  </file>
Oct 22 21:16:11 <sabotage> but I've restarted ofonod manuall
Oct 22 21:16:14 <sabotage> +y
Oct 22 21:16:42 * sabotage checks what he's got for phonesim
Oct 22 21:17:28 <sabotage> denkenz: ATM, I'm just using what's packaged in OBS for MeeGo trunk:
Oct 22 21:17:32 <sabotage>  phonesim-1.1-4.139.i586
Oct 22 21:17:33 <sabotage> ofono-0.31-1.2.i586
Oct 22 21:17:51 <sabotage> does martin still own these packages
Oct 22 21:18:09 <denkenz> yes, afaik
Oct 22 21:18:42 <sabotage> sheesh!
Oct 22 21:19:00 <sabotage> Trunk is still on phonesim 1.1 and upstream is up to 1.7+
Oct 22 21:19:24 * sabotage files a bug for martin
Oct 22 21:20:29 * sabotage wishes phonesim had it's own component in BMC
Oct 22 21:43:41 <sabotage> denkenz: ok, now running phonesim from git tip, seeing dbus warnings, but not sure if they are related or not
Oct 22 21:43:43 <sabotage> http://pastebin.ca/1970377
Oct 22 21:43:54 <sabotage> meh
Oct 22 21:43:56 <sabotage> sorry
Oct 22 21:44:06 <sabotage> looks like that is just "scripting" support related
Oct 22 21:44:20 <sabotage> prolly nothing to do with ofono 
Oct 22 21:44:36 <denkenz> that's phonesim
Oct 22 21:44:55 <denkenz> and I'm not getting that, so must be some silly security stuff on meego
Oct 22 21:46:55 <raji__> denkenz, From commit d486ba8f2aa22d429f5d892dcd71a820834039ed dbus registration problems started for history plugins. The Changelog shows lot of changes in that release. any clues?
Oct 22 21:47:11 <sabotage> well I see no meaningful diffs in the ofono.conf dbus config files
Oct 22 21:47:55 <denkenz> raji__: Nope, you might have to git bisect a bit more
Oct 22 21:50:12 <crevetor> quick question
Oct 22 21:50:40 <crevetor> I have compiled the nettime example. Do I need to change anything else for it to be called ?
Oct 22 21:51:09 <denkenz> no, but if it uses debugs you might need to run ofono with debug enabled
Oct 22 21:51:24 <denkenz> And remember it only gets called when the modem and the network actually report NITZ
Oct 22 21:53:13 <crevetor> denkenz: If I get a NET_TIME_IND from isimodem then it should be called right ?
Oct 22 21:53:39 <denkenz> yep
Oct 22 21:54:31 <crevetor> ok
Oct 22 22:00:02 <crevetor> denkenz: I don't need to call example_nettime_init from somewhere ?
Oct 22 22:00:31 <denkenz> nope, done automagically
Oct 22 22:00:41 <denkenz> unless you're not running maintainer mode
Oct 22 22:01:06 <crevetor> Let me check
Oct 22 22:01:25 <denkenz> examples might only be enabled by maintainer mode
Oct 22 22:02:06 <crevetor> ok
Oct 22 23:22:34 <sabotage> denkenz: do ofono and connman now have to be matched at some version now?
Oct 22 23:23:24 <sabotage> I can now see ofono starting, and get it to powerup phonesim, exposing the gui
Oct 22 23:23:57 <sabotage> but the moment I try sending it any method calls (via ofono test scripts)
Oct 22 23:24:09 <sabotage> I see it de-refister from the system bus
Oct 22 23:24:22 <sabotage> followed every time by prop changes in connman
Oct 22 23:25:39 <sabotage> it's probably just connman noting that ofono is gone and droping the data context...
Oct 22 23:26:14 <denkenz> there was a major change over for connman and ofono
Oct 22 23:26:21 <denkenz> Right around 0.26+ time frame
Oct 22 23:26:32 <denkenz> After that there should be no incompatibilities
Oct 22 23:26:46 <sabotage> ok
Oct 22 23:26:55 <sabotage> connman is 0.62.2
Oct 22 23:27:06 <sabotage> ofono is 0.31
Oct 22 23:27:58 <denkenz> that should work
Oct 22 23:28:21 <sabotage> like I said... twerks... until I start trying to *use* it?
Oct 22 23:28:34 <sabotage> and by use it, I mean, just call list-modems
Oct 22 23:28:38 <sabotage> or enable-modem
Oct 22 23:28:44 <sabotage> or introspect with d-feet
Oct 22 23:30:12 <sabotage> I see a NameOwnerChanged on the system bus from "org.ofono" to ":1.100" (or whatever address it happened to have)
Oct 22 23:30:25 <sabotage> and then :1.100 disappears
Oct 22 23:30:38 <sabotage> ofonod is still running, or hung
Oct 22 23:31:28 <sabotage> it still at a poll call in strace until I kill it
Oct 22 23:33:05 <denkenz> no idea, twerks for me
Oct 22 23:33:22 <denkenz> try running ofonod manually with debug on and nodetach
Oct 22 23:33:43 <sabotage> been doing that
Oct 22 23:35:41 <sabotage> the python test tools are giving:
Oct 22 23:35:42 <sabotage> ERROR:dbus.proxies:Introspect error on :1.103:/: dbus.exceptions.DBusException: org.freedesktop.DBus.Error.NoReply: Message did not receive a reply (timeout by message bus)
Oct 22 23:35:56 <denkenz> so its crashing, but for what reason
Oct 22 23:35:58 <sabotage> followed by :
Oct 22 23:35:59 <sabotage> dbus.exceptions.DBusException: org.freedesktop.DBus.Error.ServiceUnknown: The name :1.103 was not provided by any .service files
Oct 22 23:36:33 <sabotage> cant tell
Oct 22 23:36:39 <sabotage> I'm seeing no errors
Oct 22 23:38:15 <denkenz> if you're suddenly jumping off dbus that means oFono or its plugins are trying to send malformed data
Oct 22 23:38:20 <denkenz> hard as hell to debug
Oct 22 23:39:23 <sabotage> you're telling me ;)
Oct 22 23:40:43 <sabotage> http://pastebin.ca/1970424
Oct 22 23:40:44 <denkenz> try disabling external plugins if you have them
Oct 22 23:40:47 <denkenz> Otherwise no idea
Oct 22 23:41:13 <sabotage> line 01 is stable point, ofono reg on sys bus, phonesim gui up and all loaded
Oct 22 23:41:42 <sabotage> line 07 is the moment I invoked list-modems
Oct 22 23:41:58 <sabotage> pretty much meaningless to me
Oct 22 23:42:20 <sabotage> and strace it not flushing anymore past that until I kill ofono
Oct 22 23:44:50 <sabotage> ding ding...
Oct 22 23:45:12 <sabotage> removing voicecallhistory and smshistory seems to have fixed it :/
Oct 22 23:45:34 * sabotage bangs head on wall, over and over and over...
Oct 22 23:47:28 <sabotage> reinstalled smshitory, still works
Oct 22 23:48:22 <sabotage> raji__: FYI... seems in 0.31 version of ofono, having voicecallhistory installed is what is causing ofono to drop off the system bus and hang
Oct 22 23:48:30 <sabotage> hows that git bisect coming along ;)
Oct 22 23:49:48 <raji__> denkenz, sabotage, history plugins fail from this commit 8ef8135a2afdfd8de363a61d8a6ae6e1218a01ec, Add support for GDBus security handlers 
Oct 22 23:50:14 <sabotage> denkenz: thanks for the ideas... I'm off for the weekend, will tackle dialer port to 0.31 on monday I guess now
Oct 22 23:51:11 <raji__> sabotage, that is the problem I saw. there are some changes made to gdbus in ofono that is causing the ofono to drop off when history plugins register
Oct 22 23:52:44 <sabotage> raji__: what's different between smshistory and voicecallhistory then
Oct 22 23:53:15 <sabotage> oh, nm
Oct 22 23:53:21 <sabotage> smshistory is never loaded
Oct 22 23:53:26 <sabotage> due to the version mismatch
Oct 22 23:53:56 <sabotage> that's why having it installed doesn't seem to cause the problem while it does for voicecallhistory
Oct 22 23:54:13 <raji__> Yes
Oct 23 01:12:31 <denkenz> balrog-k1n: around?
Oct 23 01:20:26 <denkenz> balrog-k1n: I really like your tone_send patch, just one question, do we really need a memcpy in the tone_request_run function?
Oct 23 01:21:16 <denkenz> Seems we can save the char, set it to null, send to the driver and restore the original char
Oct 23 01:51:21 <crevetor> Is there a documentation somewhere about the isi messages ?
Oct 23 01:56:15 <crevetor> denkenz: Are you there ?
Oct 23 02:39:07 <denkenz> there is no isi documentation
Oct 23 02:45:07 <balrog-k1n> denkenz: yeah, we could do this, i'll send a follow up patch
Oct 23 02:45:08 <crevetor> :'(
Oct 23 02:45:35 <crevetor> denkenz: there's probably one but it's nokia only I guess..
Oct 23 02:46:07 <crevetor> denkenz: So you have to read the code to understand how to decode the messages I guess.
Oct 23 02:46:51 <denkenz> basically
Oct 23 02:47:42 <Termana> denkenz, that's no true
Oct 23 02:47:49 <Termana> There is isi documentation.
Oct 23 02:48:04 <denkenz> balrog-k1n: Cool, another thing that I thought might be a good idea was to limit the number of tones that we send at once
Oct 23 02:48:27 <Termana> Not in the ofono tree, but you can get documentation from a Nokia website
Oct 23 02:48:33 <crevetor> Termana: where is it ?
Oct 23 02:48:52 <denkenz> balrog-k1n: two reasons, AT command lines are limited, and in case we need to cancel it doesn't take too long
Oct 23 02:48:56 <Termana> crevetor, trying to find the link, it's been a while since I looked
Oct 23 02:49:07 <denkenz> balrog-k1n: But those are probably really far out cases :)
Oct 23 02:49:14 <denkenz> There is wireless modem api
Oct 23 02:49:21 <denkenz> However, that is not the same ISI protocol
Oct 23 02:49:33 <denkenz> It is similar, but not the same
Oct 23 02:49:58 <Termana> Ah.
Oct 23 02:50:17 <Termana> http://www.wirelessmodemapi.com/ is the link anyway. You need to create an account and log in to get it
Oct 23 02:50:45 <denkenz> You might get lucky of course
Oct 23 02:50:46 <Termana> (and accept a EULA, but anyhow)
Oct 23 02:51:07 <balrog-k1n> denkenz: let's limit to 8 tones at once
Oct 23 02:51:23 <denkenz> balrog-k1n: Yep, sounds good to me
Oct 23 02:51:54 <crevetor> Termana: thanks
Oct 23 02:53:19 <denkenz> balrog-k1n: Otherwise I really like the approach, and looks like we can easily support dial strings too
**** ENDING LOGGING AT Sat Oct 23 02:59:57 2010