Hello all!
Maybe some of you still remember this thread...

The issue was that some Bluetooth dongles do not contain a valid
Bluetooth MAC address. Well, this is against the spec and so the dongles
are bad but sometimes users cannot avoid to get such. So I still see the
need that we do as good as we can to help those users and get those bad
dongles working.

Long story short, Marcel's bdaddr application contained in the latest
release of bluez-utils works, at least for the two dongles I have with
ISS chipsets. So the BDADDR can be set now.

I wanted an automatic solution so that when I plug the dongle in it is
automatically assigned "an" address. So first I created an udev rule
that calls an application when the specific USB device with the
vendorid:deviceid I know to be bad is inserted - in my case 1131:1001
(plaease let me know more!).

Then I created a small shell script which sets a "more correct" address
than the default 11:11:11:11:11:11. I assume in this script that an
ethernet device is present in the system as eth0 and use its MAC for the
Bluetooth dongle too.

I did this for two reasons:
1. On the same machine I always want to have the same BDADDR, so random
was no choice.
2. The ethernet addresses also have to follow the same rules as
Bluetooth MAC addresses, i.e. they have to be unique. So the probability
of a collision is very low.

I have attached a tarball of my udev rule and script to this post which
works quite nicely for me. If anyone else finds it useful I would be
happy to receive comments ;)

And finally I would like to ask, how can we try to get this into
distributions? Shall I change anything in the scripts? Can we get it
into some bluez package?

Thanks!

Cheers
  nils


-- 
kernel concepts GbR      Tel: +49-271-771091-12
Dreisbachstr. 24         Fax: +49-271-771091-19
D-57250 Netphen          Mob: +49-176-21024535
--