**** BEGIN LOGGING AT Tue Feb 08 02:59:57 2011
Feb 08 08:24:43 <Jeevaka> holtmann: ping
Feb 08 08:45:30 <OlivierG> fdanis:ping ?
Feb 08 15:35:37 <holtmann> Jeevaka: pong
Feb 08 15:54:01 <fdanis> denkenz: ping
Feb 08 16:07:26 <denkenz> fdanis: pong
Feb 08 16:07:59 <denkenz> Jeevaka: Can you do me a favor and check whether Pekka's sim_ready_notify patches work nicely on IFX?
Feb 08 16:15:00 <Jeevaka> denkenz: I dont have access to the hardware now, can i do it tomorrow?
Feb 08 16:15:35 <denkenz> no hurry
Feb 08 16:15:48 <Jeevaka> holtmann: what to do with the modem selftest patch
Feb 08 16:16:03 <holtmann> I consider reverting it.
Feb 08 16:16:11 <fdanis> denkenz: do you think that patches for Bluetooth server support are acceptable ?
Feb 08 16:16:35 <holtmann> Jeevaka: See I want AT+XSELFTEST=? => check for support. AT+XSELFTEST => run the selftest. That would be clean and simple.
Feb 08 16:16:46 <denkenz> fdanis: I got lost in the server patches / emulator patches
Feb 08 16:16:52 <denkenz> padovan: Whats the latest on this?
Feb 08 16:19:49 <Jeevaka> holtmann: agreed. what is the solution for already integrated patch?
Feb 08 16:21:03 <holtmann> Jeevaka: I most likely revert it. Haven't made up my mind.
Feb 08 16:21:15 <Jeevaka> ok
Feb 08 16:23:23 <denkenz> Jeevaka: So you still want me to apply patch 4?
Feb 08 16:24:37 <Jeevaka> denkenz: I would like. whats your view?
Feb 08 16:25:48 <denkenz> Personally I think IFX should have extended XCALLSTAT much more than they did
Feb 08 16:25:58 <denkenz> Then you could have a nice implementation like with ECAV
Feb 08 16:26:22 <denkenz> Now you have to gather the pieces all over the place
Feb 08 16:27:11 <Jeevaka> denkenz: agree
Feb 08 16:27:36 <denkenz> So I'm waffling on this one
Feb 08 16:28:02 <denkenz> I can follow your patch and it seems fine, but separating the info to xcallstat/ccwa/cring/clip gets funny
Feb 08 16:28:24 <holtmann> Jeevaka: Can I ask a different question here. What does it try fix? Is something not working? Or is this just code cleanup?
Feb 08 16:28:36 <holtmann> Jeevaka: Or do you need it for TTY or something funky?
Feb 08 16:30:09 <Jeevaka> all started with relying purely on ifx AT spec
Feb 08 16:30:19 <denkenz> he's trying to outsmart the call id allocator
Feb 08 16:30:24 <Jeevaka> ;)
Feb 08 16:31:04 <denkenz> Have you tried the active/waiting -> incoming case?
Feb 08 16:31:07 <sameo_> denkenz, holtmann: Just curious, how does the SIM authentication API from Pekka look to you ?
Feb 08 16:31:24 <holtmann> sameo_: Briefly in the right direction.
Feb 08 16:31:39 <holtmann> sameo_: Not that we all have a good idea on how that works on AT modems and how good.
Feb 08 16:31:40 <padovan> denkenz: it's my -v2 series.
Feb 08 16:32:15 <sameo_> holtmann: fwiw, from a wpa_supplicant point of view, it looks good.
Feb 08 16:32:50 <fdanis> padovan: do you see my reply to your comment in my version of patches ?
Feb 08 16:32:56 <holtmann> sameo_: There are a few still opens. Since what we learned is that you can have multiple SIM apps on the card. And you need to pick one to authenticate against.
Feb 08 16:33:01 <Jeevaka> denkenz: I have tested one active and waiting call
Feb 08 16:33:20 <holtmann> sameo_: So in theory you can have your call authenticated with the USIM, but for WiFi, they use a different one.
Feb 08 16:34:10 <Jeevaka> denkenz: verified few cases which involves 2calls
Feb 08 16:35:08 <denkenz> Jeevaka: And IFX only has call states 0..7 right?
Feb 08 16:35:15 <denkenz> with 6 being disconnected and 7 being connected
Feb 08 16:35:25 <Jeevaka> it matches exactly with the one we have in ofono
Feb 08 16:35:38 <Jeevaka> denkenz: yes
Feb 08 16:36:13 <denkenz> yeah, except disconnected is something we invented
Feb 08 16:36:16 <denkenz> so we're abusing that a bit
Feb 08 16:36:19 <denkenz> but fair enough
Feb 08 16:37:30 <Jeevaka> only thing that i'm planning to 4/4 is to modify the create_call to take call id
Feb 08 16:38:23 <Jeevaka> but still open for your comments
Feb 08 16:39:29 <denkenz> meh, I'm fine with it
Feb 08 16:39:44 <denkenz> Do the create_call modification, resubmit and I'll apply
Feb 08 16:39:53 <Jeevaka> ok
Feb 08 16:40:49 <denkenz> fdanis, padovan: So you guys are fine with padovan's v2 series?
Feb 08 16:41:31 <padovan> denkenz: yes.
Feb 08 16:42:45 <fdanis> denkenz: padovan: there is a bug in client_event, I can submit a patch after you integrate it if you prefer
Feb 08 16:43:10 <padovan> fdanis: yes, please submit a delta patch for this ;)
Feb 08 16:45:49 <denkenz> ok, then I will take a look at these today
Feb 08 16:51:05 <sameo_> holtmann: Does the network tell you which one is for call, and which one is for AKA ?
Feb 08 16:53:49 <holtmann> sameo_: Good question. Have not figured that one out yet. There are defaults like USIM so that is easy.
Feb 08 16:54:04 <holtmann> However that can be used for WiFi or not.
Feb 08 16:54:18 <holtmann> We are running a bit blind here.
Feb 08 17:11:53 <darasd> denkenz: hi
Feb 08 17:12:26 <darasd> I've got some cdma signal strength patches ready to submit, and I've implemented CSQ polling in the plugin for our hw
Feb 08 17:13:00 <darasd> do you have any recommendation for frequency of polling for performance vs usability
Feb 08 17:13:02 <darasd> ?
Feb 08 17:13:20 <darasd> i"ve got it set at 5secs for now - but theres no science behind it :)
Feb 08 17:21:03 <denkenz> darasd: Polling CSQ is not a great idea, but if you have no other choice then doing this polling in the driver seems fine
Feb 08 17:21:19 <denkenz> Since this is throwaway driver, 1 second vs 5 makes no real difference to me
Feb 08 17:34:27 <darasd> okey doke! - yeah we're kind of stuck where we don't get unsolicited notifications for this
Feb 08 17:34:45 <darasd> thanks, I'll email that out today
Feb 08 17:35:27 <leiyu> denkenz: hi
Feb 08 17:35:40 <denkenz> leiyu: heya
Feb 08 17:37:17 <leiyu> denkenz: any chance to review CDMA SMS v4 patch I sent out before?
Feb 08 18:36:30 <denkenz> padovan: hm
Feb 08 18:36:31 <denkenz> +       g_hash_table_foreach(adapter_address_hash,
Feb 08 18:36:32 <denkenz> +                               (GHFunc) get_adapter_properties, NULL);
Feb 08 18:37:57 <denkenz> Does this belong in a separate patch along with bluetooth_ref changes?
Feb 08 18:42:50 <padovan> denkenz: yes, it would be better.
Feb 08 18:43:17 <denkenz> Basically you're trying to optimize out the manager GetProperties call
Feb 08 18:43:40 <padovan> yep, that is fdanis_ code actually.
Feb 08 18:43:59 <padovan> Do you want a new patch only for that?
Feb 08 18:46:59 <denkenz> I split it out already
Feb 08 18:47:16 <padovan> Ok. thanks.
Feb 08 18:49:28 <denkenz> why are we calling server_start so many times?
Feb 08 18:49:43 <denkenz> shouldn't we just call server_add_record instead?
Feb 08 18:52:30 <padovan> The point is that we don't know if the server was started or not.
Feb 08 18:53:05 <denkenz> hm, when can that happen? if there are no BT adapters?
Feb 08 18:54:38 <padovan> yep, and server_start is called from g_slist_foreach.
Feb 08 18:55:04 <denkenz> ok fair enough
Feb 08 18:55:17 <padovan> So or we call it directli or create a new funcion that would do basically the same, check the server->io and register the record.
Feb 08 18:56:38 <denkenz> So I don't really like this because you print an error when the listen fails
Feb 08 18:56:51 <denkenz> even though you shouldn't be calling the listen in the first place if you have no adapters
Feb 08 18:59:38 <padovan> denkenz: yeah. We can just supress the print, or I can fix the code to no call server_start if we have no adapters.
Feb 08 19:02:00 <denkenz> hm, are you sure about listen failing?
Feb 08 19:02:07 <denkenz> what happens if you listen and an adapter goes away?
Feb 08 19:02:13 <denkenz> Is your socket closed?
Feb 08 19:03:31 <holtmann> denkenz: It is not.
Feb 08 19:03:38 <holtmann> The listen socket stays.
Feb 08 19:17:58 <denkenz> so why does listen fail if you have no adapters
Feb 08 19:18:00 <denkenz> or does it?
Feb 08 19:23:43 <padovan> denkenz: Ah, yeah, it doesn't fail.
Feb 08 19:23:56 <denkenz> So can we simplify this like, alot?
Feb 08 19:25:53 <padovan> denkenz: I can keep only one call to listen, the others I swap for a add_record
Feb 08 19:26:06 <denkenz> That is what I'm thinking
Feb 08 19:26:13 <holtmann> They only way this fails if you have no rights to bind that port or if l2cap.ko, rfcomm.ko and bluetooth.ko failed to load.
Feb 08 19:26:14 <padovan> listen can be called from bluetooth_register_server
Feb 08 19:27:10 <denkenz> Ok, so I pushed the patch for optimizing the call flow
Feb 08 19:27:21 <denkenz> please fix up patch 1 and resubmit
Feb 08 19:28:22 <padovan> resubmit? what did you push?
Feb 08 19:28:53 <denkenz> I basically split out the get_adapter_properties parts
Feb 08 19:29:40 <padovan> Can't fetch yet.
Feb 08 19:31:48 <denkenz> gah
Feb 08 19:32:51 <padovan> Actually I can, Just changed my git config to fetch from hera.
Feb 08 19:33:00 <padovan> And not from the public git.
Feb 08 19:35:49 <denkenz> want me to email it to you?
Feb 08 19:38:05 <padovan> what?
Feb 08 19:39:06 <denkenz> the patch
Feb 08 19:39:42 <padovan> No, I fetch the internal ofono from kernel.org
Feb 08 19:39:51 <padovan> s/fetch/fetched/
Feb 08 19:40:11 <denkenz> ok
Feb 08 21:26:09 <anandx> holtmann: I need to discuss with you regarding modem selftest patch fix to support latest IFX FW.
Feb 08 21:30:04 <holtmann> anandx: My current plan is to just revert it.
Feb 08 21:37:00 <holtmann> anandx: Do you have any other proposal?
Feb 08 21:40:47 <anandx> Holtmann: I have the fix to support latest FW 1104  . it is one line AT command syntax  change. I have the fix  verified.  can I send the patch officially?
Feb 08 21:41:04 <holtmann> Just send the patch and then we see.
Feb 08 21:41:25 <holtmann> I still thing we need AT+XSELFTEST=? and AT+XSELFTEST commands to do this properly.
Feb 08 21:43:20 <anandx> Holtmann:Will send the patch now for sure. We will work for long term solution. 
Feb 08 21:46:02 <anandx> Holtmann: For your proposal, We might have to work with WSIV & Infineon  to get that in place. It might take a while and this patch is for fix the issue now.
Feb 08 22:01:37 <padovan> denkenz: I fixed the issue, patches should be on the mailing list.
Feb 08 22:01:51 <padovan> Just sent them now.
Feb 08 22:16:26 <denkenz> +               g_error_free(err);
Feb 08 22:16:34 <denkenz> Jeez, I wonder how many places I forgot to put this one
Feb 08 22:17:14 <denkenz> Though why you bother I have no idea ;)
Feb 08 22:17:23 <denkenz> Not like you use err anywhere
Feb 08 22:36:27 <denkenz> padovan: What is cb_data stuff doing in the first patch?
Feb 08 22:37:49 <denkenz> especially inside client_event
Feb 08 22:46:00 <anandx> holtmann: I submitted the fix for self test at command changed in latest fw ver.
Feb 08 23:03:35 <denkenz> padovan: Can client_event be simplified ?
Feb 08 23:03:47 <denkenz> e.g. can't we just always call CancelAuthorization
Feb 08 23:05:36 <denkenz> Keeping the watch around past the authorization stage is not necessary as well
Feb 09 00:00:06 <josch> hi, i'm looking to get a umts stick - which modem should should be inside that is most stable and supported by ofono?
Feb 09 00:44:19 <denkenz> anything by mbm
**** ENDING LOGGING AT Wed Feb 09 02:59:57 2011