**** BEGIN LOGGING AT Wed Nov 17 02:59:57 2010 Nov 17 12:02:26 is anything going on with ofono & gps? i've seen a mail from aug this year.. so how is such stuff working in generel? how does ppp work with ofono? Nov 17 14:22:58 holtmann: are you avaiable? Nov 17 14:23:14 emdete: Not really. What is up? Nov 17 14:23:45 :D i wonder why my huawei doesnt do ppp. i wanted to ask if connman already supports that Nov 17 14:25:48 the huawei in question works fine with ofono. it exposes 5 serial lines on usb but these seem to be dedicated to stuff like status updates, full at-support, gps and so on. the one that supports at and is not used by ofono does not react on ppp, so i wonder why Nov 17 14:29:15 emdete: Yes it is supported. Nov 17 14:31:08 holtmann: can you tell me how? is the channel somehow setup for that? Nov 17 14:32:20 Do you have TUN kernel driver loaded. Nov 17 14:34:41 holtmann: it's fix in the kernel Nov 17 14:35:49 holtmann: problem is that no LCP packages are send back. this works fine on anohter serial line of the same modem Nov 17 14:35:58 Hah. Funny. Nov 17 14:36:06 Seems like an issue with Huawei card. Nov 17 14:36:22 can i force ofono to use the other serial? Nov 17 14:36:59 It is udev configuration file. Nov 17 14:37:18 how is that determined frm ofono anyway? it uses first and third... why that? okay second is dead anyway but where does ..............ah, okay. let me check Nov 17 14:37:23 Check with OFONO_AT_DEBUG=1 for the first AT commands. They give some basic settings like A2DIAG or so. Nov 17 14:39:17 yes, that's on all the time here ;) Nov 17 14:41:35 i wonder what the logging tells me. fuser says, that ofono is unsing 2 channels, but logging doesn't distinguish those... Nov 17 14:49:04 great! Nov 17 14:49:15 that works just fine. thanks alot, holtmann ! Nov 17 15:00:15 emdete: sorry, but what works just fine? what did you change? Nov 17 15:02:59 kvalo: gprs with my huawei. i changed ofono 'cause of sim state that is not correctly reported and changed the udev rules to free up ttyUSB0 to ttyUSB4 cause 4 can't do ppp but 0 can Nov 17 15:03:57 emdete: can you send a patch? that way others using the same modem would get it working Nov 17 15:05:21 yes, sure, but to publish this i need to check in depth about the sim state. the udev stuff should be okay Nov 17 15:05:57 i would like to have a correct sim state reporting if possible Nov 17 15:06:54 problem is i don't have sufficient documentation ... :/ Nov 17 15:08:16 yeah, the sim state is tricky because huawei models work differently Nov 17 15:09:06 kvalo: do u have any more information on that? Nov 17 15:11:08 holtmann: i saw your mail about gps. with this huawei it is simple because you switch it on on channel 0 or 4 and nmea comes in on 3 all the time. with another modem the switch-on command must be given in the channel where you want to receive the nmea and the channel is still r/w. how would you solve that? Nov 17 15:15:01 emdete: no more information, I just sent few patches earlier to get huawei working with my modem Nov 17 15:16:11 kvalo: what did you change? Nov 17 15:17:03 emdete: gitk plugins/huawei.c :) Nov 17 15:23:31 kvalo: i just meant high-level ;) Nov 17 15:24:36 hm, why does ofono report the cell-id but leaves out everything else that's important? hm, mcc, mnc, lac should be included in that too. Nov 17 15:28:39 kvalo: does sim-inserted status work for u? Nov 17 15:29:36 kvalo: and: which one do you own? Nov 17 15:30:04 ah, sry, the E1552... ;) Nov 17 15:42:18 emdete: basically I worked on sim state handling. so that it works with my modem both with "warm" and "cold" start Nov 17 15:43:28 kvalo: okay, maybe i should remove my hard-state-fix and try :) Nov 17 15:43:39 kvalo: does gprs work for you? Nov 17 15:44:10 emdete: yes. or it did few weeks ago, I haven't tried with latest git for some time Nov 17 15:47:04 kvalo: can you explain a bit how your modem is designed? is it also using ppp? having several serial dev on usb? Nov 17 15:47:35 yeah. ppp with 3 usb serial ports Nov 17 15:50:10 holtmann: i'm messing with phonesim to implement the tty support. If I understood well I have to set the answers in default.xml, right? Nov 17 15:51:06 so, if my phonesim driver in ofono queries the value of Powered property, i'll have a rule in this file with the answer (0 or 1) Nov 17 15:51:22 is that right? Nov 17 16:06:39 demarchi: yes Nov 17 16:07:16 denkenz: hi Nov 17 16:07:34 denkenz: have you seen my cell-id question? Nov 17 16:24:00 emdete: my connection here sucks, so just in case you didn't see my reply, oFono is reporting ci and lac Nov 17 16:25:56 :D Nov 17 16:26:19 double check the string huawei sends, some of the firmware fuck up the CREG response Nov 17 16:26:37 But otherwise it should work Nov 17 16:26:41 yes & no. it reports cell-id as signal. problem is that mcc, mnc, lac and cellid change at once but may give several signals. so how to cope with that? Nov 17 16:27:34 If you need them to be atomic then you should use GetProperties Nov 17 16:27:47 otherwise they're emitted when they change Nov 17 16:29:09 But even then you have no guarantees, since the MCC/MNC query is done after the CREG query Nov 17 16:46:53 so a GetProperties on the NetworkRegistration will determine the values again in that moment? Nov 17 16:49:08 hm, i have to check how a cell-change into roaming is signaled by a modem. did you try that, denkenz ? Nov 17 16:49:45 doesn't matter really Nov 17 16:49:51 Your CREG is reported separately from COPS Nov 17 16:50:00 And the COPS transaction is 4 commands Nov 17 16:50:08 so either way you have a race Nov 17 16:57:51 denkenz: i changed the udev rule for the EM770W - maybe it's usable for everyone... Nov 17 17:00:53 denkenz: ping Nov 17 17:01:41 :( Nov 17 17:07:06 padovan: fyi my link really sucks here Nov 17 17:09:06 denkenz: wher are u? Nov 17 17:09:47 Dublin @ Meego Conf Nov 17 17:15:26 denkenz: are wetab people around? Nov 17 17:16:03 No idea, I admit I didn't visit the vendor booths all that much Nov 17 17:16:26 :) i got ofono running on it so they may be interested :D Nov 17 17:21:28 nice Nov 17 17:22:31 they use chat scripts with their setup Nov 17 17:26:50 so now they should be paying you ;) Nov 17 17:27:34 denkenz: paying isn't he most important thing. i would appreciate more docu about their stuff ;) Nov 17 17:29:22 denkenz: ofono didn't signal a sms which was a wap-push. isn't that supported? Nov 17 17:34:41 emdete: It is, but you gotta be running the push notification agent Nov 17 17:34:47 otherwise it goes to /dev/null Nov 17 17:35:08 see test/test-push-notification Nov 17 17:37:26 denkenz: not sure on to powerdown phonesim before send the reset command, because phonesim closes the socket when it powers down. Nov 17 17:39:16 current implementation of modem reset does power down and then power up the modem. Nov 17 20:27:15 what needs to go to ofono.rules if I believe I have a generic usb atmodem? Nov 17 20:29:43 this is a BandLuxe device that uses the option driver and works out of the box with NM, modem-manager and ppp on another box Nov 17 20:31:02 or should this just work without a udev rule? Nov 17 21:36:51 denkenz: you there? Nov 17 21:38:01 sorta Nov 17 21:38:23 denkenz: about the Lockdown property task Nov 17 21:39:21 set Lockdown = TRUE means lock _and_ power down and set Lockdown = FALSE means unlock _and_ power up, right? Nov 17 21:40:49 and if the owner of the lock calls Powered(1) with Lockdown = 1 we should power the modem up and release the lock as well. Do you agree on that? Nov 17 21:42:20 honestly what I remember was that these become methods and lock down is only a readonly property Nov 17 21:43:51 TODO doesn't say that btw Nov 17 21:43:57 However, either way, implementing setting Powered with Lockdown set to true is not needed Nov 17 21:45:28 try it as a property Nov 17 21:45:53 So setting Powered with Lockdown = TRUE returns a error even for the owner. Nov 17 21:46:07 yeah Nov 17 21:46:28 Yes, I'm doing it as property. Nov 17 21:46:28 The idea is really to turn the modem off and not allow anyone to turn it on for the duration Nov 17 21:46:52 So in some respect methods make more sense there Nov 17 21:47:19 Yeah, you right. Nov 17 21:47:42 I started this task while I wait comments on modem_reset Nov 17 21:48:43 Have to re-send those patches btw, you did a clean up commit that broke mine patches. Nov 17 21:49:09 nod, let me comment on them, I saw some issues Nov 17 21:50:25 denkenz: I'm figuring out on how to make phonesim send the reset command after it is disabled. Nov 17 21:50:37 Once the socket connection is dropped. Nov 17 21:51:02 You don't really need to Nov 17 21:51:28 Just send the reset and have phonesim driver close the socket and signal the core Nov 17 21:52:31 just as it is now? Nov 17 21:53:27 another question, if the modem is disabled, can I lock it? Nov 17 21:53:44 I don't see why not Nov 17 21:53:59 lockdown should basically follow the same logic as reset Nov 17 21:54:11 if we were online, try to get back to online Nov 17 21:54:12 Yes. Nov 17 21:54:17 etc Nov 18 02:14:44 holtmann: which mbm modem are using for multiple contexts of gprs? **** ENDING LOGGING AT Thu Nov 18 02:59:58 2010