[Announce] Bluetooth PBAP and IrMC / Carkit support is coming to N900... - Page 37 - maemo.org

Active Topics

Porting apps to Leste (32)
to Maemo 7 / Leste by Arno_11 - 4 days, 14 hrs ago
[ANNOUNCE] CSSU-thumb thread - stable Thumb2 on N900 (2,264)
to Maemo 5 / Fremantle by panjgoori - 6 days, 8 hrs ago
more...

Page 37 of 57

Thread Tools

luf	2012-07-22 , 18:27
Posts: 141 \| Thanked: 313 times \| Joined on May 2012 @ Czech Republic	#361

I clean up the code and patches and add the PBAP call history to the new version. There are also fixes for buggy behaviour of VW Touch Adapter.

I'm using it so I hope it's stable enough so I hope you'll try it also.

Changelog:
obexd (0.46-1maemo2+cssu1)
* Fix missing 0.vcf in phonebook_pull
* Fix IrMC to use full path
* Fix bugs in VW Touch Adapter
- removing X-SKYPE from VCARD
- add empty line after PHOTO;VALUE=uri: (not needed by RFC)
* Add initial support for call history (ich, och, mch, cch)

obexd (0.46-1maemo1+cssu1)
* New upstream release.
* backport for glib, dbus and bluez from maemo

obexd (0.45-1maemo1+cssu1)
* use maemo patches from 0.37-0maemo1
* backport for EDS from maemo
* read default addressbook if no addressbook present
* backport for bluez from maemo

Sources and other .deb packages are accessible at http://merlin1991.at/~luf/pbap/

Newer version is available. You can find more info here http://talk.maemo.org/showpost.php?p...&postcount=378

Attached file:
obexd_0.46-1maemo2+cssu1_armel.deb (94.6 KB)

Attached Files

	evolution-data-server-addressbook_1.4.2.1-20120717-maemo1+cssu1_armel.deb (71.2 KB, 271 views)
	libebook_1.4.2.1-20120717-maemo1+cssu1_armel.deb (106.0 KB, 224 views)
	libedata-book_1.4.2.1-20120717-maemo1+cssu1_armel.deb (62.7 KB, 236 views)
	libedataserver_1.4.2.1-20120717-maemo1+cssu1_armel.deb (84.1 KB, 224 views)

Last edited by luf; 2012-08-30 at 21:59. Reason: Newer version available

Quote & Reply |

The Following User Says Thank You to luf For This Useful Post:
hschmitt

luf	2012-07-22 , 18:32
Posts: 141 \| Thanked: 313 times \| Joined on May 2012 @ Czech Republic	#362

Originally Posted by Acurus_

1. Edit: Nokia PC Suite (7.1.180.64): Contacts can be opened and syncronized with Outlook!

Does it mean that Outlook synchronization doesn't work with older version?

Quote & Reply |

	Acurus_	2012-07-22 , 18:38
	Posts: 209 \| Thanked: 156 times \| Joined on Jun 2012	#363

It worked before I have installed your *.deb's. And it is working after installing your *.deb's.

Quote & Reply |

hschmitt	2012-07-23 , 03:55
Posts: 159 \| Thanked: 122 times \| Joined on Nov 2009	#364

Originally Posted by Acurus_

If I am right, it should be problematic, to chance something in the N9 like it has been changed for the N900 here. Because the PBAP inside the N9 is not "that open" like in the N900. Otherwise PBAP is a part of the N9, but it does not work with every device.

You can debug the behaviour of your N9. In the file "/usr/share/dbus-1/services/obexd.service" add the "--debug" option to the obexd call. Then you can see debug messages in "/var/log/syslog". From these messages you can tell how far the headset gets with querying PBAP.

I wrote in a previous post that IRMC is not possible with N9. Actually, I have to revise my statement. It might be possible.
obexd has a plugin mechanism. At the moment all available pugins are compiled into obexd but it can load plugins dynamic libraries, too.
Therefore obexd tries to load each "lib*.so" file in directory "/usr/lib/obex/plugins/". I renamed the irmc plugin to irmc2 applied the patch to fix it and compiled a libirmc2.so using scratchbox as a obexd plugin.
There are two obstacles.
First, I don't know whether aegis restricts obexd to load via "dlopen" a dynamic library?
Second, I don't know whether I compiled my libirmc2.so correctly?
Since I cannot kill and restart obexd, I cannot test it without a reboot. And if one of the obstacles occurs the device would be bricked. Currently, I have no time to reset my device.
If someone has a spare device to test it, please pn me. I will sent the compiled lib or if someone wants to compile it himself, I can share the sources.

__________________
Your packages are waiting too long to get tested? HowTo encourage at TMO.

Quote & Reply |

The Following User Says Thank You to hschmitt For This Useful Post:
crnkoj

hschmitt	2012-07-28 , 15:37
Posts: 159 \| Thanked: 122 times \| Joined on Nov 2009	#365

Originally Posted by luf

I'm sorry I have no N9. I think you don't use PBAP but IrMC profile. There is only missing '/' in two places. If you read whole previous page we talked about it with hschmitt (and hschmitt sent patch to the obexd team).

Hi luf,
the patch is now upstream:
http://git.kernel.org/?p=bluetooth/o...17a368b87f49a1
There are two more patches related to irmc.
Maybe you could integrate them and see, whether it still works.
http://git.kernel.org/?p=bluetooth/o...7de40772b8c80c
http://git.kernel.org/?p=bluetooth/o...6f94045cb2454a

Sadly, the patches came into obexd right after version 0.47 was published, but for 0.48 IRMC should work.

__________________
Your packages are waiting too long to get tested? HowTo encourage at TMO.

Quote & Reply |

The Following 2 Users Say Thank You to hschmitt For This Useful Post:
crnkoj, luf

crnkoj	2012-07-30 , 15:51
Posts: 32 \| Thanked: 4 times \| Joined on Mar 2012	#366

Originally Posted by hschmitt

Hi luf,
the patch is now upstream:
http://git.kernel.org/?p=bluetooth/o...17a368b87f49a1
There are two more patches related to irmc.
Maybe you could integrate them and see, whether it still works.
http://git.kernel.org/?p=bluetooth/o...7de40772b8c80c
http://git.kernel.org/?p=bluetooth/o...6f94045cb2454a

Sadly, the patches came into obexd right after version 0.47 was published, but for 0.48 IRMC should work.

Thanks for the info, ill try the debug option (sadly havent had any time at all to try it till now, will do it now) and see what i can understand from it. I still strongly believe the headset is using PBAP and not IRMC.

Edit: so this is the syslog with debug on on obexdand with bluetooth debugging (i guess on on, i didnt change it): http://pastebin.com/fUh1CBJR . I believe the key to my issue lies within lines 5 and 6 of the pastebin and possibly line 9? But i have no clue as to how to repair/mod the thing to make this nokia phonebook to work. Does it only want to download phone number from the SIM card, or am i understanding it wrongly?
edit2: here a syslog of it when i receive and do a call: http://pastebin.com/G1hkmY8F

Last edited by crnkoj; 2012-07-30 at 18:33.

Quote & Reply |

luf	2012-07-31 , 10:51
Posts: 141 \| Thanked: 313 times \| Joined on May 2012 @ Czech Republic	#367

Originally Posted by hschmitt

Hi luf,
the patch is now upstream:
http://git.kernel.org/?p=bluetooth/o...17a368b87f49a1
There are two more patches related to irmc.
Maybe you could integrate them and see, whether it still works.
http://git.kernel.org/?p=bluetooth/o...7de40772b8c80c
http://git.kernel.org/?p=bluetooth/o...6f94045cb2454a

It's great that fixes for IrMC are in upstream.

I'm sorry I don't want to use patch for recognizing correct phone books. As you can see I'm using completely different way in obexd for maemo N900 (I added also call history).

BTW it seems to me that it's some kind of cleanup of code instead of fixing some issues (or am I wrong?),
I also want to use only packages from debian testing as it's tested.

Quote & Reply |

The Following User Says Thank You to luf For This Useful Post:
hschmitt

luf	2012-07-31 , 11:04
Posts: 141 \| Thanked: 313 times \| Joined on May 2012 @ Czech Republic	#368

Originally Posted by crnkoj

Thanks for the info, ill try the debug option (sadly havent had any time at all to try it till now, will do it now) and see what i can understand from it. I still strongly believe the headset is using PBAP and not IRMC.

How did you enable the debug and how is the syslog configured?
(I think you can try --debug='*' instead of only --debug)
I miss more info in the syslog.

I get something like:

Code:

Jul 18 21:25:44 Nokia-N900 obexd[1618]: src/obex.c:obex_session_start() 
Jul 18 21:25:44 Nokia-N900 obexd[1618]: src/obex.c:cmd_connect() 
Jul 18 21:25:44 Nokia-N900 obexd[1618]: CONNECT(0x0), (null)(0xffffffff)
Jul 18 21:25:44 Nokia-N900 obexd[1618]: src/obex.c:cmd_connect() Selected driver: Phonebook Access server
Jul 18 21:25:44 Nokia-N900 obexd[1618]: CONNECT(0x0), (null)(0x0)
Jul 18 21:25:44 Nokia-N900 obexd[1618]: src/obex.c:cmd_setpath() 
Jul 18 21:25:44 Nokia-N900 obexd[1618]: SETPATH(0x5), (null)(0xffffffff)

As you can see you logged only some of the lines ...

Quote & Reply |

The Following User Says Thank You to luf For This Useful Post:
crnkoj

hschmitt	2012-07-31 , 11:12
Posts: 159 \| Thanked: 122 times \| Joined on Nov 2009	#369

Originally Posted by luf

It's great that fixes for IrMC are in upstream.

I'm sorry I don't want to use patch for recognizing correct phone books. As you can see I'm using completely different way in obexd for maemo N900 (I added also call history)...

Wow that's cool. Maybe I have to reactivate my N900 as primary phone.

Originally Posted by luf

...
BTW it seems to me that it's some kind of cleanup of code instead of fixing some issues (or am I wrong?),..

Yes it is just cleanup. And the first one is only replacing strings with macros. The one not committed by me is a lot restructuring and I hoped, that you could compile and test it once, so that their is no bug in it (no need to include those in your distributed packages).

Originally Posted by luf

...I also want to use only packages from debian testing as it's tested.

like IRMC

Actually, I doubt that IRMC is tested regularly, because it was not working since 2011-02-17!

__________________
Your packages are waiting too long to get tested? HowTo encourage at TMO.

Quote & Reply |

luf	2012-07-31 , 11:54
Posts: 141 \| Thanked: 313 times \| Joined on May 2012 @ Czech Republic	#370

Originally Posted by hschmitt

Wow that's cool. Maybe I have to reactivate my N900 as primary phone.

I did it only in synchronous way and I tested it only with pbap-client from my computer.
Unfortunately no HF I tested using it (no idea if I miss something or just HFs don't use PBAP call history - ich, och, mch, cch).
I tested it with VW Touch Adapted (Burry manufacturer) and Parrot CK3100 (no idea what firmware - brother took it to vacation for 14 days).

Actually, I doubt that IRMC is tested regularly, because it was not working since 2011-02-17!

That's true. Ok. I'll try the last patch. But I'm working on asynchronous rtcom-eventlogger interface right now so I have no clue when I try the patch.