BlueZ configuration file/security question

[Claudio Takahasi]

On 11/30/06, Thiago Macieira <thiago at kde.org> wrote:
> Claudio Takahasi wrote:
> >thanks!!! I think the last option of using peer to peer connection
> >fits better, because Marcel wants keep the service paths/interfaces
> >freely definable and without well-known names for ServiceAgents.
>
> It doesn't make sense to have freely defined interfaces. If they are
> freely defined, they serve no purpose, because they can't be called from
> hcid. The daemon can only call what it knows how to call (method names,
> parameter types, return values, etc.)
>
> --
>  Thiago Macieira - thiago (AT) macieira.info - thiago (AT) kde.org
>   PGP/GPG: 0x6EF45358; fingerprint:
>   E067 918B B660 DBD1 105C 966C 33F5 F005 6EF4 5358
>
>
>

For ServiceAgents there is one mandatory interface called
"org.bluez.ServiceAgent" that contains the methods: Start, Stop and
Release. But ServiceAgents can define more interfaces like
"org.bluez.Headset".

When the ServiceAgent starts, the first procedure is register the
service with the hcid sending a
org.bluez.Manager.RegisterService(path, name, description) message.
The parameter "path" is the path where the messages must tbe sent if
someone wants use this service. When the hcid receives this message,
it uses the path value and add a new object path. All clients that are
listenning for signals in the manager interface will receive a
ServiceRegistered(path) signal. This is how the clients are notified
about the new services available. The link informed in the first email
shows an overview of this procedure.

The main reason for have this kind of design is because the hcid must
control the service availability and there are some hard tasks hiden
from the ServiceAgents.

BR,
Claudio.


-- 
---------------------------------------------------------
Claudio Takahasi
Instituto Nokia de Tecnologia - INdT