fully qualified object locations

[Havoc Pennington]

Robert McQueen wrote:
> Can we add a bus type for bus names? If we have this we can establish
> the convention that a struct of bus name and object path can be
> demarshalled into a proxy objct. At the moment, a type 'o' can't always
> correctly be assumed to refer to an object on the service that sent the
> message. It could be sent alongside a string representing some other
> service, and it's not safe to assume that's a bus name unless we have a
> type that validates them.
> 

Some previous threads:
http://lists.freedesktop.org/archives/dbus/2006-January/003986.html
http://lists.freedesktop.org/archives/dbus/2005-May/002629.html
http://lists.freedesktop.org/archives/dbus/2004-October/001687.html

Not sure this should be crammed into 1.0 ...

btw there are just a few little things left for 1.0, doesn't someone on 
this list have a few days to just knock them down? ;-)

Havoc