RFC: Libdbus++ - C++ bindings for DBus

[staneker at gris.uni-tuebingen.de]

On Wed, 17 May 2006, Thiago Macieira wrote:

> staneker at gris.uni-tuebingen.de wrote:
> >  ...
> >  dbus::DBusServerConnection cnct("org.test.communication");
> >
> [snip]
> >  dbus::ClientConnection cnct(getenv("DBUS_SESSION_BUS_ADDRESS"),
> > "org.test.communication");
>
> Why is there an argument with an interface name there? Or is that a
> service name?

It is later used in dbus_message_new_method_call as
destination if a message is created on the client side to
call a method. I am not sure, that this is the correct/nice
way, but could be changed easily :-)


> >At the moment, it is like a rmi-system, because the
> >instances on the client and on the server-side do not
> >directly correspond. It would be nice, to create an instance
> >on the server-side, if the instance is created on the
> >client-side, but the problem is, when should the server
> >destroy the instance, if the client died unexpected? How
> >does other bindings (qt/glib) solve this problem?
>
> Someone said that the glib bindings emit a signal when the remote end
> disconnects. The Qt binding doesn't do that. The client side has a loose
> reference to the server side and that's all.

Hm. Ok... I thought about a restricted lifetime of an
instance on the server if there is no further
communication.

Dirk