**** BEGIN LOGGING AT Tue Oct 20 02:59:57 2009
Oct 20 03:20:30 <denkenz> holtmann: Currently half the plugins use ofono_info for OFONO_AT_DEBUG function, the other half uses DBG.  Lets just pick one.
Oct 20 07:37:32 <holtmann> denkenz: ofono_info is the right one. Otherwise the OFONO_AT_DEBUG is pointless.
Oct 20 07:52:01 <tmzt> holtmann: any response on my questions from last night? particulurly the length of modem type
Oct 20 07:52:15 <tmzt> and what I should call the qualcomm cdma modem
Oct 20 07:52:44 <holtmann> tmzt: Not really right now. I will look into it, but since you have GIT up and running. You can just go ahead and do everything by yourself :)
Oct 20 07:53:52 <tmzt> right
Oct 20 07:54:13 <tmzt> the length is limited to 16 though
Oct 20 07:54:33 <tmzt> yeah, my plugin works
Oct 20 07:54:54 <tmzt> I need a null sim implementation to use for sim functions
Oct 20 07:55:07 <tmzt> also need to use unicode for sms instead of pdu
Oct 20 07:55:31 <holtmann> Forget about the text mode SMS. That will fall flat on its face at some point.
Oct 20 07:56:22 <tmzt> binary?
Oct 20 07:56:38 <tmzt> with 0x000x1a termination?
Oct 20 07:56:50 <tmzt> 0x001a
Oct 20 07:57:57 <tmzt> the length problem prevents me from having the type string for the modem the same as the filename
Oct 20 07:58:06 <tmzt> which seems to be the convention
Oct 20 07:58:24 <tmzt> filename is qualcomm_msm_cdma
Oct 20 14:31:36 <denkenz> holtmann: I disagree about ofono_info bit, but I know why you did it this way
Oct 20 14:31:44 <denkenz> holtmann: What we really need is zoned debugging
Oct 20 16:17:16 <ygu51> denkenz: After connecting to freerunner, I can't find /dev/ttySAC0. How do I get it?
Oct 20 16:17:49 <ygu51> denkenz: My OS is now om2008.12
Oct 20 16:18:08 <denkenz> I have no idea
Oct 20 16:18:35 <denkenz> Poke around in the telephony daemon to see what port it opens?
Oct 20 16:18:36 <ygu51> Do you have usb0 after connecting to FR?
Oct 20 16:19:07 <denkenz> usb0 on the desktop, yes
Oct 20 16:19:31 <ygu51> And you also have /dev/ttySAC0 on the desktop?
Oct 20 16:20:01 <denkenz> eh? no :)
Oct 20 16:20:08 <ygu51> Need I add some parameter to the boot loader?
Oct 20 16:20:18 <denkenz> ttySAC0 will only appear on the phone
Oct 20 16:20:39 <ygu51> Oh, I see. 
Oct 20 18:14:33 <padovan> denkenz: ping
Oct 20 18:14:42 <denkenz> padovan: pong
Oct 20 18:17:10 <padovan> denkenz: zhenhua is working on voicecall now, right? what features does it includes? Is phone status information there?
Oct 20 18:17:16 <padovan> I need a issue to work on.
Oct 20 18:17:56 <denkenz> padovan: He's working on getting basic calling features in, then multiparty / enhanced call notifications
Oct 20 18:18:07 <denkenz> padovan: If you wanna help, start with the netreg driver
Oct 20 18:18:39 <tmzt> this is telepathy stuff?
Oct 20 18:18:44 <padovan> denkenz: what is netreg?
Oct 20 18:18:46 <tmzt> ah, isi
Oct 20 18:20:26 <denkenz> tmzt: No, hfp
Oct 20 18:20:38 <tmzt> yeah, I read netlink
Oct 20 18:20:40 <denkenz> padovan: Network registration, e.g. registered/not, roaming, signal strength
Oct 20 18:21:41 <padovan> ah.. phone status information :)
Oct 20 18:23:40 <padovan> denkenz: so it will be a new driver or will be into hfpmodem?
Oct 20 18:25:58 <denkenz> it will go into hfpmodem
Oct 20 18:26:40 <denkenz> netreg driver should be quite simple, since you don't need to support register / unregister / operator scan
Oct 20 18:27:01 <denkenz> However, HFP doesn't support numeric operator query, so mcc/mnc will need to be faked I think
Oct 20 18:28:38 <padovan> Ok.
Oct 20 18:28:53 <padovan> let me look to that now.
**** ENDING LOGGING AT Tue Oct 20 20:16:14 2009

**** BEGIN LOGGING AT Tue Oct 20 22:16:15 2009
Oct 20 22:36:51 <tmzt> git calls topgit automatically?
Oct 20 22:37:04 <tmzt> on ofono repository I mean
**** ENDING LOGGING AT Tue Oct 20 22:56:54 2009

**** BEGIN LOGGING AT Tue Oct 20 23:36:15 2009
Oct 20 23:36:34 <denkenz> It should work like I described, if it doesn't, it might be a bug
Oct 20 23:41:40 <tmzt> ofonod[2881]: csms_query_cb got result: 0
Oct 20 23:41:41 <tmzt> ofonod[2881]: Final response: ERROR
Oct 20 23:41:41 <tmzt> ofonod[2881]: SMS not supported by this modem.  If this is in error please submit patches to support this hardware
Oct 20 23:41:44 <tmzt> ofonod[2881]: src/sms.c:sms_remove() atom: 0x68e18
Oct 20 23:41:54 <tmzt> I don't see the AT command that failed
Oct 20 23:42:16 <denkenz> export OFONO_AT_DEBUG=1 , restart ofonod
Oct 20 23:42:22 <tmzt> yeah, I missed it this time
Oct 20 23:42:40 <denkenz> Anyway, you will most likely need a cdma specific SMS driver anyway
Oct 20 23:43:04 <tmzt> AOC
Oct 20 23:43:06 <tmzt> CCWE
Oct 20 23:43:12 <tmzt> CSMS=?
Oct 20 23:43:14 <tmzt> all ERROR
Oct 20 23:43:16 <tmzt> yes
Oct 20 23:43:56 <padovan> denkenz: so, looking at src/modem.c pre_sim is only called if powered is true. :(
Oct 20 23:44:14 <tmzt> which is set in cfun_cb?
Oct 20 23:44:47 <denkenz> padovan: its called whenever powered turns true
Oct 20 23:44:48 <tmzt> would it make sense to have better abstraction than "atmodem" for gsm to make cdma support easier?
Oct 20 23:45:05 <tmzt> I understand that is something that would have to be discussed and I can work around it
Oct 20 23:45:43 <denkenz> tmzt: Naming is the least of your issues right now :)
Oct 20 23:45:46 <tmzt> but if this is intended to support modem cards and other things
Oct 20 23:46:04 <tmzt> well yeah, what are my issues? it's all research?
Oct 20 23:46:33 <denkenz> hehe nod, right now atmodem is a generic 27.007 implementation
Oct 20 23:47:19 <denkenz> holtmann: We need to talk about gprs
Oct 20 23:47:29 <tmzt>         ofono_sms_create(modem, 0, "atmodem", chat);
Oct 20 23:47:32 <tmzt> in atgen.c
Oct 20 23:48:15 <tmzt> so it's more than that, the 27.007 behavior is coded at a very low level of ofono
Oct 20 23:48:17 <denkenz> just hack atgen to have a Type=cdma|gsm attribute
Oct 20 23:48:18 <padovan> denkenz: so I have a deadlock because modem gets powered after receive CIND status values, but we query CIND only if powered is true.
Oct 20 23:49:06 <tmzt> what does the "atmodem" name mean then?
Oct 20 23:49:17 <tmzt> if it's just a string used to register things
Oct 20 23:49:18 <denkenz> it is a driver name
Oct 20 23:49:38 <denkenz> atgen is a 'generic GSM at modem' driver
Oct 20 23:49:54 <tmzt> yes
Oct 20 23:49:58 <tmzt> but what is "atmodem"
Oct 20 23:50:03 <tmzt> where is that registered
Oct 20 23:50:17 <tmzt> I can add sms_cdma.c and ofono_sms_cdma_create or whatever
Oct 20 23:50:18 <denkenz> um, in the driver struct
Oct 20 23:51:21 <tmzt> drivers/atmodem/sms.c:  .name = "atmodem",
Oct 20 23:51:23 <tmzt> ah, got it
Oct 20 23:51:29 <tmzt> it's not src/sms.c
Oct 20 23:52:15 <tmzt> I see it now, I had it completely wrong
Oct 20 23:52:16 <tmzt> good
Oct 20 23:52:35 <tmzt> ofono_sms_driver is what I need to write, not a replacement for src/sms.c
Oct 20 23:54:16 <tmzt> still gsm stuff in there maybe
Oct 20 23:55:26 <tmzt> okay, so it's _probe function
Oct 20 23:55:42 <tmzt> in drivers/atmodem/sms.c
Oct 20 23:59:13 <tmzt> is it possible to register additional drivers in that file to reuse the infrastructure?
Oct 20 23:59:22 <tmzt> is there a better way to do that
Oct 20 23:59:40 <tmzt> for instance, I may not use PDU or some of the stores but most of the code is relevant
Oct 20 23:59:46 <tmzt> but it's in the atmodem driver
Oct 21 00:00:54 <denkenz> depends on how big the change is
Oct 21 00:01:04 <denkenz> In your case you probably need a brand new driver
Oct 21 00:03:33 <tmzt> should I try to write a minimal one and let somebody with more glib experience flesh it out?
Oct 21 00:04:12 <tmzt> as in CMT receive and send but stubbing the probing, SCA, etc.
Oct 21 00:04:14 <denkenz> that sounds good, if changes are minimal we can always put it into generic one
Oct 21 00:04:24 <tmzt> okay, I'll work on that
Oct 21 00:04:44 <tmzt> what are the stores and what do ofono use?
Oct 21 00:05:02 <tmzt> I understand there is no caching on sim but it sim store supported (on gsm)?
Oct 21 00:05:28 <tmzt> and is ME used for those sent over dbus or however it's handled
Oct 21 00:05:41 <tmzt> not sure what MT is other than MS terminated
Oct 21 00:05:51 <denkenz> ofono configures it, with preference to SIM
Oct 21 00:05:51 <tmzt> sd card?
Oct 21 00:06:06 <denkenz> but if possible ofono configures not to even hit the SIM
Oct 21 00:06:13 <denkenz> e.g. direct CMT notification
Oct 21 00:06:41 <tmzt> I think that's what we need
Oct 21 00:06:54 <tmzt> this modem might not support any store at all or it might support EFS
Oct 21 00:07:03 <tmzt> but it's not used on windows ce, the normal os on this phone
Oct 21 00:08:23 <tmzt> this is essentially the document I'm using right now
Oct 21 00:08:26 <tmzt> http://sms.unikom.ac.id/yd1jea/download/S000294D.pdf
Oct 21 00:26:29 <denkenz> padovan: I've no idea what you're talking about.  Unmodified HEAD just works and even calls pre_sim
Oct 21 00:31:56 <zhenhua> denkenz: why we need query CIEV after CMER before modem is powered on?
Oct 21 00:34:34 <zhenhua> denkenz: another question, how to make a USB into serial port TTY
Oct 21 00:35:26 <zhenhua> denkenz: i plug in a iPhone into netbook and see Bus 001 Device 004: ID 05ac:1292 Apple, Inc. iPhone 3G. but how to make a TTY device then?
Oct 21 00:35:58 <tmzt> zhenhua: on linux?
Oct 21 00:36:01 <tmzt> what are you asking?
Oct 21 00:36:02 <zhenhua> yeah
Oct 21 00:36:22 <zhenhua> tmzt: i am asking how to turn a phone into tty
Oct 21 00:36:28 <tmzt> you expect to get access to the modem from your laptop?
Oct 21 00:36:32 <zhenhua> tmzt: when i connect the phone via data cable
Oct 21 00:36:32 <tmzt> in an iPhone #g
Oct 21 00:36:34 <tmzt> 3G
Oct 21 00:36:36 <zhenhua> right
Oct 21 00:36:46 <zhenhua> no matter what kind of phone
Oct 21 00:36:55 <tmzt> look into tethering and come back if there is a way to do it I guess
Oct 21 00:37:12 <zhenhua> i tried modprobe usbserial but it doesn't work
Oct 21 00:37:39 <tmzt> and lsusb -v -v ?
Oct 21 00:37:43 <tmzt> pastebin
Oct 21 00:38:49 <zhenhua> lots of text from lsusb -vv
Oct 21 00:38:56 <zhenhua> what is pastebin?
Oct 21 00:39:06 <tmzt> pastebin.com
Oct 21 00:39:26 <zhenhua> thx.let me check.
Oct 21 00:40:20 <tmzt> I mean paste the output there
Oct 21 00:40:25 <tmzt> not here
Oct 21 00:40:29 <zhenhua> oh, i can submit question there
Oct 21 00:40:29 <tmzt> then paste the url here
Oct 21 00:43:18 <zhenhua> http://pastebin.com/m126cddfd
Oct 21 00:46:24 <tmzt> there's no serial port there
Oct 21 00:46:31 <tmzt> you are going to have to do more research
Oct 21 00:46:47 <tmzt> you might have better luck with linux on the iPhone itself
Oct 21 00:46:48 <zhenhua> i see. i assume someone here already done that
Oct 21 00:47:21 <zhenhua> i have two phones, both of them don't have serial port
Oct 21 00:47:56 <zhenhua> they act as storage, but it may turn into a serial
Oct 21 00:48:17 <zhenhua> i tried usb_modeswitch, but it doesn't work for me. http://www.draisberghof.de/usb_modeswitch/
Oct 21 00:49:59 <zhenhua> padovan: why we need CIEV query before SLC is connected?
Oct 21 00:51:14 <zhenhua> padovan: hi padovan
Oct 21 00:52:06 <padovan> zhenhua: hi :)
Oct 21 00:52:13 <padovan> zhenhua: what's SLC?
Oct 21 00:52:23 <zhenhua> padovan: serivce level connection.
Oct 21 00:53:24 <zhenhua> padovan: according to spec page 17-19, we don't query CIEV during that
Oct 21 00:53:36 <zhenhua> because CIND result should be the same as CIEV does
Oct 21 00:53:37 <padovan> zhenhua: are you talking my patch, right?
Oct 21 00:54:32 <zhenhua> yeah
Oct 21 00:55:11 <zhenhua> padovan: i plan to query CIEV in each driver, like netreg and voicecall
Oct 21 00:55:33 <zhenhua> because each only care part of that, like call, callsetup, callheld, roam, battchg
Oct 21 00:55:36 <padovan> zhenhua: I missed that information :-(. Need to put that after SLC.
Oct 21 00:55:43 <tmzt> zhenhua: what's the other phone?
Oct 21 00:56:20 <zhenhua> tmzt: sorry, another phone is LG KM900, but i don't have cable now. :(
Oct 21 00:57:08 <padovan> zhenhua: so, I'll hold that patch and continue to do the netreg driver
Oct 21 00:58:18 <tmzt> zhenhua: that would be a better choice
Oct 21 00:58:34 <tmzt> but you may not get the full gsm at command set through that port
Oct 21 00:58:51 <tmzt> is it a pda or touchscreen phone?
Oct 21 00:59:41 <zhenhua> touchscreen phone
Oct 21 01:00:00 <zhenhua> padovan: that looks fine for me.
Oct 21 01:00:33 <zhenhua> tmzt: i'd like to try atgen driver but now i couldn't
Oct 21 01:00:33 <padovan> denkenz: I know it works, but the problem I have is: I need to power the modem to have pre_sim called, but who powers the modem is the pre_sim function.
Oct 21 01:05:13 <padovan> zhenhua: I expect to finish netreg soon, I need first to solve the pre_sim problem.
Oct 21 01:06:27 <zhenhua> padovan: no problems
Oct 21 01:19:27 <zhenhua> tmzt: have you uses atgen on any phone?
Oct 21 01:28:06 <tmzt> I'm trying to
Oct 21 01:28:13 <tmzt> but it's a cdma phone
Oct 21 01:30:38 <zhenhua> how do you connect to the modem
Oct 21 01:31:34 <tmzt> I talk to it directly
Oct 21 01:32:36 <zhenhua> via what? ip address or tty device
Oct 21 01:36:13 <denkenz> padovan: No, core powers the modem up ;)
Oct 21 01:40:14 <padovan> denkenz: right, I'll check it tomorrow or the day after tomorrow, thanks.
Oct 21 02:01:26 <holtmann> denkenz: Around? I am sitting in the Tokyo Limosine bus to me hotel and it has WiFi :)
Oct 21 02:33:38 <tmzt> zhenhua: tty, but linux and ofono are running on the phone
Oct 21 02:33:41 <denkenz> holtmann: heh
Oct 21 02:34:28 <holtmann> denkenz: So I have finished phonesim.git during my flight to NRT. I dropped my changes to use autoconf. Geez Qt is painful.
Oct 21 02:34:45 <denkenz> yeah, qmake has too much magic in it
Oct 21 02:34:52 <denkenz> Most Trolls don't know how it works ;)
Oct 21 02:35:40 <holtmann> So about GPRS.
Oct 21 02:36:05 <denkenz> so both proposals on the ML that I've reviewed so far are a bit painful
Oct 21 02:36:30 <denkenz> But GPRS is a pain in the ass too, the AT command setup is just fucked up
Oct 21 02:37:15 <denkenz> So I'm thinking we need to start really small first
Oct 21 02:37:36 <denkenz> With one atom doing CGREG / CGATT
Oct 21 02:38:00 <holtmann> Yes. I agree.
Oct 21 02:38:03 <denkenz> The others doing CGDCONT / CGACT / CGDATA or the CGDCONT / ENAP crap
Oct 21 02:39:23 <denkenz> The problem is we might need several primary contexts, so not sure if we wanna make them into a proper atom 
Oct 21 02:39:33 <denkenz> Similar to how connman network / device is split up
Oct 21 02:40:00 <holtmann> Maybe.
Oct 21 02:40:23 <denkenz> Other approach is to keep it id based, like voicecalls today
Oct 21 02:40:43 <holtmann> We can go either way. Keeping it simple and just have MBM hardware working. And then maybe phonet will lead us the way. What we learn from there helps to redefine the API properly.
Oct 21 02:41:15 <denkenz> Problem is their proposal is pretty basic too
Oct 21 02:41:31 <denkenz> Both proposals actually push the storage of the context into the driver
Oct 21 02:41:40 <denkenz> And there onto the modem, which is wrong
Oct 21 02:42:16 <holtmann> I know that modems tend to store these informations, but that is wrong. We have to store them and re-program on next start.
Oct 21 02:42:35 <holtmann> For example we base this around the IMSI and so just switching the hardware or the SIM card needs to reprogram it properly.
Oct 21 02:43:01 <holtmann> Using the wrong APN can result in expensive roaming charges otherwise.
Oct 21 02:43:01 <denkenz> yah, exactly, but half the API assumes cind is setup into the modem
Oct 21 02:44:32 <denkenz> So for now I'm thinking second atom will have cont_activate(struct primary_context *) and cont_deactivate(struct primary_context *)
Oct 21 02:45:09 <denkenz> And the callback will have to provide an id of some sort once it is active
Oct 21 02:45:28 <holtmann> Sounds good to me.
Oct 21 02:46:49 <denkenz> The tricky bits become something like +CGEV: NW DEACT <PDP_type>, <PDP_addr>, [<cid>]
Oct 21 02:47:14 <denkenz> I'm not sure how ENAP stuff interacts with this
Oct 21 02:49:06 <holtmann> It doesn't. At least I haven't seen it. It has its own ones. Or also LOWER_UP signals via the Ethernet interface.
Oct 21 02:49:45 <denkenz> so that's the question, what do we do there
Oct 21 02:50:02 <holtmann> We have to stuff that into an atom so it can be replaced.
Oct 21 02:50:25 <denkenz> can we set those properties on the network iface from userspace?
Oct 21 02:50:27 <holtmann> I am almost there. Logging off now. Talk to you after I checked in and had a shower.
Oct 21 02:50:35 <denkenz> nod
**** ENDING LOGGING AT Wed Oct 21 02:59:59 2009