**** BEGIN LOGGING AT Wed Jan 20 02:59:57 2010
Jan 20 19:53:47 <jprvita> zhenhua: I see you were having trouble to run dbus-1.3 on ubuntu another day..  I've updated the deb packages for that: http://local.profusion.mobi:8081/~jprvita/dbus_1.3.0-0ubuntu1_i386/
Jan 20 19:54:06 <jprvita> it worked on my testing netbook with karmic
Jan 20 20:00:33 <padovan> denkenz: I changed the HFP patch to query for all bluetooth devices and load the as a modem. Still a prototype. Patch is on the mailing list.
Jan 20 20:01:39 <padovan> denkenz: jhe: I need some help with the org.bluez.HandsfreeAgent.Release(). I can't get the rfcomm link disconnected.
Jan 20 20:02:13 <denkenz> jhe, holtmann: What is the final verdict of HandsfreeAgent, per device or per adapter?
Jan 20 20:04:04 <holtmann> denkenz: I haven't thought about it at all.
Jan 20 20:04:13 <holtmann> And I am going to get off to the airport now.
Jan 20 20:05:08 <denkenz> Heh, if you see jhe, figure it out between the two of you already ;)
Jan 20 20:41:10 <Vudentz> jprvita, thanks for the package, it would save me a lot of time, just remember me to buy you a beer when we meet :D
Jan 20 20:41:51 <jprvita> Vudentz: hehe, don't worry, this is the kind of thing I usually do not forget :)
Jan 20 20:51:55 <Vudentz> holtmann, padovan, the device path X adapter path is more a design choice, I just prefer adapter path because it will generate faster responses specially in the device creation where the ui normally tries to connect as soon as the services are discovered.
Jan 20 20:54:51 <padovan> Vudentz: looking to the per-adapter choice: what happens when we have more than on device with HF into one adapter?
Jan 20 21:00:05 <Vudentz> padovan, NewConnection will have to take device path, but by looking at the current code this seems quite natural since ofono wouldn't have to enumerate the devices just the adapters which are less likely to change.
Jan 20 21:03:17 <padovan> Vudentz: So will we register all adapter even if has none devices?
Jan 20 21:03:45 <padovan> and will NewConnection actually create the modem?
Jan 20 21:06:03 <Vudentz> padovan, well you can always enumerate then if that is really important, so you can show the offline modems, iirc ofono wants to be able to disable the modem even if it is offline.
Jan 20 21:07:24 <Vudentz> padovan, but this might make our authorization useless, since even if the use accept the connection ofono may drop the connection anyway since it was disabled.
Jan 20 21:10:53 <padovan> Vudentz: the per-adapter agent sounds better.
Jan 20 21:13:03 <Vudentz> padovan, well there one problem though, we need and extra interface in adapter.
Jan 20 21:16:18 <padovan> which one?
Jan 21 01:04:38 <zhenhua> jprvita: thanks, i would have a try asap. ;-)
Jan 21 01:06:05 <zhenhua> padovan: i vote to per-device solution, i agree with denkenz, per-device would be easier for ofono.
Jan 21 02:30:10 <zhenhua> Vudentz: i could understand per-adapter may be fine for bluez. but in oFono, the thing becomes werid.
Jan 21 02:32:40 <zhenhua> Vudentz: original, ofono modem is the peer of bluez device. but in adapter solution, ofono should create new modem instance in each NewConnection. and hfp plugin becomes a peer with adapter.
Jan 21 02:35:00 <zhenhua> jprvita: great work! my ubuntu works fine with your package. don't forget to remind me buy your a beer when you come to Shanghai next time. :-D
**** ENDING LOGGING AT Thu Jan 21 02:59:56 2010