**** BEGIN LOGGING AT Fri Jun 18 02:59:57 2010 Jun 18 18:18:12 hello Jun 18 18:22:25 I haven't been able to find clear information about the state of SIM Application Toolkit support in ofono. would anyone here happen to know if its implemented yet? Jun 18 18:52:23 flailingmonkey: It is still a work in progress Jun 18 19:04:12 is it being implemented as an ofono plugin? Jun 18 19:24:49 flailingmonkey: no, it's in the core Jun 18 19:26:14 flailingmonkey: what part are you interested in? Jun 18 19:27:10 it's probably never going to be "implemented", as in fully Jun 18 19:27:54 balrog-k1n: So question about your CFUN=1 patch Jun 18 19:28:26 balrog-k1n: This is only a problem only when first plugged in right? Once we CFUN=4 the modem won't ask for a PIN Jun 18 19:29:03 denkenz: right Jun 18 19:30:05 So I'm a bit hesitant to move the atom creation away from post_sim Jun 18 19:30:18 They will be labeled as online state Jun 18 19:31:08 denkenz: ah, i thought the atoms can be created and go away any time Jun 18 19:31:28 (it's also a problem after AT*E2RESET, which is like reconnecting the modem) Jun 18 19:31:30 They can, but its a bit more tricky with the online state patches Jun 18 19:32:05 I keep thinking of adding possibility of forcing a particular phase on an atom Jun 18 19:32:31 However, in the first boot case, the modem proceeds straight to CFUN=1 anyway Jun 18 19:32:38 Do we really need to bother? Jun 18 19:34:26 not sure what you're asking Jun 18 19:34:51 but it made me wonder that maybe the AT+CFUN after pin authentication is the way it was designed for, not other way Jun 18 19:35:47 you're sending AT+CFUN=1 in post_sim, however after EPEV the modem will go to CFUN=1 anyway Jun 18 19:36:14 you mean automatically? Jun 18 19:36:29 That's what mine does without a PIN Jun 18 19:36:43 just goes straight to searching Jun 18 19:36:56 i'll check what this one does Jun 18 19:37:24 Another possibility is to make mbm utilize the online state Jun 18 19:37:32 So you'd do CFUN=1 from set_online Jun 18 19:38:06 the fact is that when CFUN=1 is executed has some bearing on how other stuff works :p , i don't know why, maybe there's another way to make it work Jun 18 19:38:27 I know :( Jun 18 19:39:26 Honestly I don't have a good answer here Jun 18 19:39:41 The only thing I can suggest is forcing E2RESET when we power up Jun 18 19:39:59 i thought about doing it when powering down, Jun 18 19:40:14 it makes the device be reenumerated on USB, Jun 18 19:40:56 so udev removes it and then creates again Jun 18 19:41:03 denkenz: I am mainly interested as it looks like the MeeGo project will need to support it in the handset UX, and my impression is that they expect it will come from ofono Jun 18 19:41:35 flailingmonkey: Your assertion is correct Jun 18 19:41:42 what part of it does meego need exactly? Jun 18 19:41:58 asking just out of curiousity Jun 18 19:42:04 balrog-k1n: Yuck, doesn't sound like a good solution Jun 18 19:42:10 in the requirements they now have available, it looks like (hold on one sec) Jun 18 19:43:27 I'll get a link but I see it says: 2688 oFono Core features for 3G telephony lists "SATK Mandatory SIM Tool Kit features" Jun 18 19:44:48 sounds like a subset, but frankly I think Nokia could bring some inside experience, unless they licensed an implementation for symbian Jun 18 19:45:04 there are features in STK that depend on hardware (gps, wifi etc) and other feature that depend on integration with browser and other applications Jun 18 19:45:27 so it's not something that you can say ofono "implements" starting from version n Jun 18 19:46:28 oFono will do the heavy lifting Jun 18 19:46:38 UI and integration with other components is still required Jun 18 19:46:52 indeed Jun 18 19:47:20 this is a big tree of features still unfinished for meego's handset UX: http://bugs.meego.com/showdependencytree.cgi?id=3211&hide_resolved=1 Jun 18 19:47:54 you'll find bug #2688 in there Jun 18 19:49:24 I'm still not sure what you're after Jun 18 19:50:40 yes, that bug description is very unprecise Jun 18 19:51:23 indeed, just sharing tangential info to meego requirements Jun 18 19:51:28 "Get IMEI", "Set phone functionality" Jun 18 19:51:39 The bug is essentially a wishlist, randomly organized Jun 18 19:51:59 Much better set of tasks can be found in oFono TODO list (which needs to be updated with Sim Toolkit tasks) Jun 18 19:52:28 It reflects reality a bit better Jun 18 19:52:29 thanks, I can take a look at it. Jun 18 19:53:12 I'm pretty sure that list got put together by someone working on handset, as placeholder Jun 18 19:53:42 I know exactly how it got put together Jun 18 19:55:07 it looks familiar :) Jun 18 19:55:26 ah good :) Jun 18 20:00:07 balrog-k1n: The more I think about it, the more I like the idea of using set_online for CFUN Jun 18 20:00:27 balrog-k1n: For MBM, we can actually assume the modem is 'on' when it is inserted Jun 18 20:00:47 So we can simply use ofono_modem_set_powered(mbm, TRUE) Jun 18 20:01:39 i'm not sure i understand how online is supposed to work yet (it seems to always be False, even though phone is usable) Jun 18 20:02:08 That sounds like a bug, Online should be true if set_online is not implemented Jun 18 20:02:18 Basically how Nokia modems work is as follows: Jun 18 20:02:39 set powered -> they actually trigger some hardware switch on the card to power the modem up Jun 18 20:02:52 pre_sim -> as usual Jun 18 20:03:10 post_sim -> SIM is now available for reading / writing, but no tx/rx circuits active Jun 18 20:03:18 set_online -> rx/tx circuits active Jun 18 20:04:27 ok, i think that applies to most modems, not only Nokia? Jun 18 20:04:37 or should apply Jun 18 20:04:40 pretty much Jun 18 20:04:59 Except in the case of our USB sticks, we have no control over power Jun 18 20:05:13 It just turns on and stays on as long as it is plugged in Jun 18 20:05:22 in theory CFUN should do that Jun 18 20:05:36 Well, CFUN *does* do it for MBM Jun 18 20:05:45 Try sending a CFUN=0, the card jumps off the usb bus Jun 18 20:05:51 yeah Jun 18 20:06:01 But it also triggers udev disconnect Jun 18 20:06:08 so no real point Jun 18 20:06:19 on the old nokia phones if you connect to the AT emulator on bluetooth, AT+CFUN=0 powered off the phone as well Jun 18 20:06:45 Yeah, so I think it is a more flexible system, but connman needs to be changed Jun 18 20:06:56 it also makes things like the calypso work nicely Jun 18 20:07:09 e.g. set_powered can flip all those silly gpio lines Jun 18 20:07:22 and set_online can send CFUN=1 / CFUN=4 or whatever Jun 18 20:07:39 i assume Online should automatically go to the state saved on disk, after sim authentication? Jun 18 20:08:18 I think we let connman do it Jun 18 20:08:29 Basically Online becomes like Powered Jun 18 20:08:39 And Powered takes a slightly different meaning Jun 18 20:08:53 so if you're writing a dialer from scratch you need to flip Online yourself? Jun 18 20:09:01 i.e. no connman Jun 18 20:09:12 Maybe not Dialer, but some settings app Jun 18 20:09:24 but yeah, I want to keep it out of oFono for now Jun 18 20:09:37 Online is meant to be used for Flight mode, and you have to turn off gps/wifi, etc too Jun 18 20:09:49 yeah Jun 18 20:17:22 So that's the thing to try, see if we can get mbm to work in the set_online / set_powered semantic Jun 18 20:17:47 Only bad part is that my mbm just goes immediately to online state when I plug it in Jun 18 20:24:21 denkenz: why's that bad though? the online state needs to be restored by someone to a known state anyway Jun 18 20:26:05 this one stays at CFUN=4 after entering pin, or if pin is disabled Jun 18 20:30:52 interesting Jun 18 20:30:59 do you have your MD300 handy? Jun 18 20:31:02 Does it do the same? Jun 18 20:32:34 balrog-k1n: And yeah, I guess we can do a CFUN=4 right after udev detects the card Jun 18 20:34:07 balrog-k1n: Actually nevermind, I was thinking of the Option / Huawei modems Jun 18 20:34:16 mbm indeed stays in mode 4 Jun 18 20:34:24 or let that connman or whoever look at Online and restore the state? Jun 18 20:34:30 why should offline be default? Jun 18 20:35:37 oFono does not allow the driver to notify the core of online state changes Jun 18 20:36:34 It might be we need to add it, but after looking into it briefly I thought it was too complicated Jun 18 20:36:45 So for now online is controlled by the core Jun 18 22:08:50 denkenz, yang_office: are you ok with the parsing status as part of the stk_command struct? the pro argument is that it gives you information whether the rest of the struct is valid, but it could be separate too Jun 18 22:10:03 but i think the parse_ functions should either return the parsing status as return value, or be converted to void and return the result in the struct field Jun 18 22:10:08 the first idea seems easier Jun 18 22:14:31 I'm fine with the first approach mostly because we need command id / dest / src info anyway **** ENDING LOGGING AT Sat Jun 19 02:59:57 2010