**** BEGIN LOGGING AT Thu Jan 20 02:59:57 2011 Jan 20 07:09:26 is ofono already shipping on some device that's on sale? or is N900 basically the closest to that (the only one that has also vendor contributed support in ofono to make the device work)? Jan 20 07:10:04 just trying to figure it out, so far I think most other supported phones are basically contributed by community members trying to get free stack on their phone Jan 20 09:35:27 Jeevaka: This GIOChannel buffering stuff is nasty. Jan 20 10:04:30 Mirv: you can try mbm and huawei that i know that work fine Jan 20 10:04:42 or any modem listed in drivers/* Jan 20 10:10:38 holtmann: is the patch adding ref count in bluetooth ok for integration ? Jan 20 10:10:58 fdanis: I have not looked into it actually. Jan 20 10:11:00 Let me check. Jan 20 10:11:16 I remember seeing something that Gustavo was fine with it. Jan 20 10:13:51 fdanis: I think it is broken. Jan 20 10:18:36 holtmann: fdalleau and I have no problem to apply it and compile Jan 20 10:18:49 could you give me more details, please ? Jan 20 10:22:44 See my response. Jan 20 10:23:09 It will never increment more than to 1. So if you increment twice it should be a 2, but it stays at 1. Jan 20 10:23:22 Then if you decrement twice, you should be at 0, but you end up with -1. Jan 20 10:32:20 wagi: So about the bearer stuff. You had something there. You wanted something. I wanna make sure we have all of our grounds covered. Jan 20 10:32:45 holtmann: I understand my mistake Jan 20 10:33:15 akiniemi: ping, false alarm with n900 (modem-hw issue, not related to ofono) Jan 20 10:33:18 pessi: ^^^ Jan 20 10:33:37 padovan ask me to increment at the end of function, so I should replace the return by a goto to the increment, ok ? Jan 20 10:35:48 That would do it as well. Jan 20 10:36:00 Still just make both functions void. Jan 20 10:36:06 holtmann: I can only remember somehting in the connman context. Jan 20 10:36:10 holtmann: let me check the notes Jan 20 10:36:22 fdanis: Just ensure not to increment in case of an error. Jan 20 10:39:29 holtmann: I have only found these notes here http://pastebin.com/HVbNXfvF Jan 20 10:40:54 and then the two uses cases with the internal modem (carkit) and the one from the customer Jan 20 10:41:50 This is more about the bearer in 3G. Like to know if you are on GSM or UMTS or LTE. Jan 20 10:42:26 exaxtly. ah, you might refer to the other project I was doing where we tried to log the technologies. Jan 20 10:43:54 Marcel? Jan 20 10:44:14 sjur: Yes. Jan 20 10:44:38 wagi: Denis mentioned to me that you had something there. I am going through Remi's patches now. Jan 20 10:45:11 Marcel: Regarding signal-strength reports, your proposal here does not avoid modem, IPC and ofono to be woken up does it? Jan 20 10:45:24 holtmann: ahhhh, slowly its coming back. just a sec Jan 20 10:49:05 sjur: Does it really makes a different if we report signal strength or not? Jan 20 10:49:39 My concern is more about the modems that are damn stupid. Like the Huawei one Kristen and I talked about yesterday. Where it kept reporting the same strength over and over again. Jan 20 10:50:11 So what I am seeing is that no matter how smart we try to be here, if the modem is damn stupid, we are loosing that game. Jan 20 10:51:17 holtmann: I tried to report the current used technology and realized that most modems are quite crapy. So in the end I wanted something like +CPSB Jan 20 10:51:34 Yes, there are certanly issues with some modems, but even if the modem are trying to be smart (reporting intervals or heuristics) you may still get frequent updates over a longer periode of time (e.g. when traveling with screen off) Jan 20 10:51:36 So Remi's patches should do the right thing for you. Jan 20 10:51:41 yep Jan 20 10:52:12 * wagi loves mail archives Jan 20 10:52:16 sjur: Actually I think we wanna get these reports when screen is off. We can trigger GPS updates or WiFi scans based on that. Jan 20 10:53:20 Marcel: Did you see Akis mail just now? Jan 20 10:53:34 Nope. Jan 20 10:55:14 Marcel: If you have GPS and or potentially WiFi active you're probably not exactly optimizing for battery life time anyway ;-) Jan 20 11:00:04 Marcel: If you use an Agent API GPS and ConnMan could still subscribe to the APIs in certain scenarios. I Jan 20 11:02:21 I don't see any conflict with the use-cases you mention and defining an Agent API. And it would enable modem vendors to meet their promises for power consumptions :-) Jan 20 11:02:25 kvehmanen: ack Jan 20 11:04:21 sjur: ISI modems don't actually allow unsubscribing to RSSI only -- unsubscribe, and you lose all net related indications... :( Jan 20 11:05:31 aki: that's too bad :-(. I guess ISI changes are not so likely? Jan 20 11:05:41 sjur: I hear they might be working on an API to allow configuring the RSSI hysteresis/granularity Jan 20 11:05:53 sjur: allowing also completely quelching the inds Jan 20 11:06:04 sjur: so in principle, I like your idea :) Jan 20 11:07:08 aki: yes ditto, I know we have had similar discussions. But even if you report only with some seconds intervall it might still be too much. Jan 20 11:34:26 demarchi: yes, I was mostly wondering about vendor contributions/support, not such functionality Jan 20 11:35:37 Mirv: Check AUTHORS for example. Jan 20 11:35:54 git shortlog -n -s -e Jan 20 12:11:29 holtmann: thanks, of course. Jan 20 14:58:33 holtmann: denkenz: I prepare a patch to add server support in bluetooth.c, without the authorization support in a first time Jan 20 14:58:54 holtmann: denkenz: Should I directly add multiple client support, or do you prefer it in a later patch ? Jan 20 14:59:15 fdanis: You should plan for multiple clients from the beginning. Jan 20 14:59:34 holtmann: ok Jan 20 15:50:28 balrog-k1n: ping Jan 20 16:28:17 denkenz: ping Jan 20 16:28:47 Jeevaka: pong Jan 20 16:29:43 do we need to change the copyright year for the files to 2008 - 2011? Jan 20 16:30:23 yes, let Marcel handle it, he usually does Jan 20 16:32:00 andrzej patches adds the sim-auth addintions after cdma thinks Jan 20 16:33:23 my view is to keep all the gsm related things above the cdma specific defintions Jan 20 16:33:49 what is your view on this? Jan 20 16:35:12 You mean for the atom definitions? Jan 20 16:36:28 basically not only definitions but also the file additions Jan 20 16:36:40 eg: sim-auth.h in makefile.am Jan 20 16:37:16 ah, that grows organically, we can always re-order them as needed Jan 20 16:38:06 atom definitions as well Jan 20 16:40:10 Hi Jan 20 16:41:14 Out of curiosity, are there any plans to support retrieving the user's MSISDN from the SIM card's 6F40 file when using ISI modems, eventually? Jan 20 16:51:50 vmlemon_: what do you plan to do with it? Jan 20 16:58:26 Nothing in particular - I was mostly thinking about it from a "feature completeness" perspective. Jan 20 17:00:15 (No reason why a device running OFono shouldn't have that functionality, in comparison to a Symbian device, for instance - and it's something that quite a lot of people seem to find useful). Jan 20 17:03:43 External applications could probably be able to send it as part of a vCard, or as a convenience feature when completing forms on Web pages, for example. Jan 20 17:03:56 *or it could be used as Jan 20 17:32:59 vmlemon_: not all the SIM cards have a correct phone number there. but it would be nice to have Jan 20 17:33:43 Makes sense - especially when considering number portability, and inactive SIM cards. Jan 20 17:33:45 is there any logic in modem drivers if to create message waiting or not? Jan 20 17:34:03 denkenz? akiniemi? Jan 20 17:34:44 pessi: originally it was intended for modems to control whether they would ever receive such notifications Jan 20 17:34:47 (When I last ported a number between 2 UK-based networks, the MSISDN on the destination network SIM card wasn't updated, if I remember correctly) Jan 20 17:34:49 e.g. pseudo modems Jan 20 17:35:10 Nowadays I'm thinking we should just roll it into the sms atom Jan 20 17:36:20 I'll add it in isigen, our tester wants to have it Jan 20 17:37:38 On another note, is there any documentation for the PN_SIM resource? I understand that it's deprecated, although it seems that some Nokia handsets expose it externally via USB, and it's apparently implemented instead of PN_UICC on the N900. Jan 20 17:38:15 (I've got the Wireless Modem API documentation, but it appears that there's nothing covering it there). Jan 20 17:39:15 I think you can't use the ISI SIM server if CSD or S60 is running on the device though Jan 20 17:39:23 same problem as SMS reception... Jan 20 17:39:34 Seems to work for me, if I set my N73 to Offline Mode. Jan 20 17:39:44 Other handsets might not have that quirk, though. Jan 20 17:40:01 vmlemon_: SIM server works pretty ok (for low values of ok) Jan 20 17:40:16 receiving sms is a problem, however Jan 20 17:40:25 in fact, this limitation is the technical reason why Nokia phones don't do any SIM related AT commands Jan 20 17:40:36 the ISI AT server cannot use the SIM server... Jan 20 17:40:45 (Curiously, I can also register to the network, make calls and receive SMSes - although I'm uncertain as to whether or not incoming call reception works) Jan 20 17:41:13 incoming call should work, though I have not tested N73 Jan 20 17:42:07 I also had a play with the circuit-switched video calling features on the handset - although it appears that the S60 video calling application crashes during call creation, since there's no server available to handle things. Jan 20 17:42:51 (I see packets from the Call resource in Wireshark, though - but I haven't got around to adding support for them to the dissector that I'm using) Jan 20 17:43:02 courmisch: hm? I though it was because the sim server only gives you cooked access, and at commands want to have raw access to elementary files Jan 20 17:43:52 It appears that the current PN_SIM implementation "short-circuits" various record requests, or otherwise has stub implementations for them Jan 20 17:44:07 pessi: +CRSM maybe. But Nokia does not even do AT+CPIN, AT+CPWD and AT+CLCK Jan 20 17:44:44 The Network resource seems to "cook" things, though Jan 20 17:45:04 As an aside, I've noticed some interesting behaviour with the SMS Resource Jan 20 17:47:01 (Something handles packets with Message ID 0x29 according to my traces - but there's no documentation anywhere, and I can't even find references to 0x29 in the codebase)... Jan 20 17:50:21 vmlemon_: it is a debugging/logging message Jan 20 17:51:12 Thanks pessi! I suppose that I'll probably have to dip into the Symbian Platform codebase later, to see if it has a "friendly name" associated with it. :) Jan 20 17:53:20 hm 0x2c is SMS_MT_TEST_ENV_IND, 0x29 is SMS_SEND_HOLD_IND Jan 20 17:54:20 sometimes you can see the received sms with id 0x2c as far as I can remember Jan 20 17:56:06 I always thought that SMS_PP_ROUTING_NTF was responsible for carrying SMS payloads, but it's interesting to know that they've been observed under other Message IDs... Jan 20 17:56:35 (At least that's my understanding from looking at traces containing OTA Settings payloads, and MMS notifications) Jan 20 17:57:14 yep, the test ind is probably for debugging/logging Jan 20 17:57:23 "Routing Notification" feels like a bit of a misnomer, though Jan 20 17:57:33 the SMS_PP_ROUTING_NTF is sent to proper sms client Jan 20 17:57:50 Aah Jan 20 18:13:27 I see from the udev rules/code that the Qualcomm GOBI (05C6_9205) is a recognized device, but does anyone have any experience using it with oFono yet for CS Voice calls? Jan 20 18:15:00 I'm trying to retrofit a netbook as a dev/test env and the GOBI is one of the few modems whitelisted in the BIOS :( Jan 20 18:17:24 sabotage: gobi doesn't work, it has a neutered at interface and no docs for the proprietary one Jan 20 18:17:39 nod :( Jan 20 18:19:25 Huawei (EM770/770U/770W/EM660) are the only other options, but from backlog... wondering if I want a Huawei (OSPM issues with keeping USB awake) Jan 20 18:19:57 though, in my case, I suppose I don't care about OSPM since is stays plugged in at my desk 99% of the time Jan 20 18:21:28 looks like the following Huawai modems are whitelisted, 1003, 1001, 1404 Jan 20 18:21:46 any recommendations between these denkenz? Jan 20 18:22:35 How is your BIOS hacking skills? ;) Jan 20 18:22:45 heh... Jan 20 18:23:17 looked into that already... problem was with the Phoenix bios repackaging... no linux tools for it AFAICS Jan 20 18:23:52 If your choice is between a gobi and a huawei I wouldn't bother Jan 20 18:24:01 the Huawai 1404 seems to support OFONO_HUAWEI_VOICE according to ofono.rules Jan 20 18:24:22 You still need special firmware Jan 20 18:24:59 denkenz: my other option is Ericsson F3607gw, but according to the wiki (http://ofono.org/wiki/ofono-supported-modems) it will only get me data, not voice Jan 20 18:25:48 F3607 is a good choice and with one line change to the modem driver you can probably have voice Jan 20 18:25:50 But not audio Jan 20 18:26:46 hmm, I wonder if I can disect the Huawai FW from the Lenovo provided windows driver... Jan 20 18:27:12 what good is voice w/out audio ;) Jan 20 18:27:34 exactly, no mini PCI or USB device will give you that Jan 20 18:27:59 There are some huawei devices that will, but they need special firmware and for you to run marcel's special audio tool Jan 20 18:28:11 So highly experimental is an understatement Jan 20 18:29:03 nod Jan 20 18:29:29 * sabotage gives up this pipe-dream, again, and gets back to real work :/ Jan 20 18:29:58 You'd have better luck sticking a Wavecom into your destkop Jan 20 18:30:07 hah! Jan 20 18:30:25 And putting in common mic/speakers into it Jan 20 18:31:25 whatever happened to that asterix server idea we noodled about some, oh, almost 2 years ago now ;) Jan 20 18:31:39 or maybe that what james I was talking to... Jan 20 18:31:46 can't recall now Jan 20 18:35:33 hi denis Jan 20 18:36:40 thank you for your comment. Is http://lists.ofono.org/pipermail/ofono/2011-January/007577.html good. or do I have to submit this patch 1/2 along with the changes you suggested for Patch2/2 Jan 20 18:42:24 * vmlemon_ idly wonders if access to the modem audio resource is firewalled on the basebands of devices other than the N900 Jan 20 18:51:41 (If it wasn't, it could make for an interesting desktop integration system; and it'd be useful for my initial plans of using the basebands of Nokia devices via ISI as a method of providing connectivity to QEMU). Jan 20 19:04:29 FaiyazBaxamusa: Just resubmit v3 Jan 20 19:04:48 partial resubmissions only confuse me Jan 20 19:05:30 Thats fine. I will resubmit both patches as v3. Jan 20 21:21:23 denkenz: ping Jan 20 21:21:56 padovan: pong Jan 20 21:23:20 denkenz: Will Emulator be a generic thing? Our will be one for each kind, e.g. one for HFP, other for DUN, etc. Jan 20 21:23:29 My guess is the first one Jan 20 21:23:40 No it has to be specific Jan 20 21:23:55 since the HFP one has special extensions for indicators Jan 20 21:24:15 But much of the implementation can probably be shared Jan 20 21:24:34 If we can unify them, then we will Jan 20 21:24:41 I see. Jan 20 21:25:03 Hint: "I have no idea" ;) Jan 20 21:25:38 Thanks for the hint ;) Jan 20 21:26:37 I can see the ppp interface by now, just need to rework all the patches and fix pontetial crashes before send the patches. Jan 20 21:27:16 sweet Jan 21 00:42:14 ok, my test sim/account is now giving GSM message 24 messages on all incoming calls... what SS did I fubar to cause this :/ **** ENDING LOGGING AT Fri Jan 21 02:59:57 2011