The goals of the protocol would be to use a dynamic (i.e. non static text)
interactive environment, but to also have each scene/room described in VRML
(Even though VRML is non-interactive this can be very useful in giving people
the ability to visualize a location, and the static scene can change as
various objects and people are introduced or removed from it). To be able
to distinguish the difference between various 'pipes' on the connection, and
to be something which could become an addition to currently existing MUDs
and VESs without large driver overhauls.
If anybody is interested in defining such a protocol contact me. If there
is enough response I will create a list on the whole topic. Whatever the
result I will be doing at least the server end of such a protocol at a
Server I am writing.