RE: Coordinates and Multiuser

ZB Lucas ([email protected])
Wed, 3 May 1995 13:59:26 -0700 (PDT)


It would seem to me that the two of these concepts are completely
intertwined in the way that we are going to design the system. it is
clear to me that we actually already have most of the technology that we
need to create the environment that we desire, (ie methods for displaying
3d viz/aud, behaviour systems and distributed platform independant IPC)
but the problem lies with the implementation of said technology. i think
what we need to do is map out all of the things that we could possibly
want out of this, not thinking of any kind of limitations, and work from
a standpoint of creating something that fits within the paradigm that we
create. like the Internet itself, we must try to create a paradigm that
will outlive the technology that we currently must use to implement it.

RE: SnowCrash versus MUD topology
i would like to introduce the term Realm. (similar to Domain) a Realm
can be any kind of shared virtual space. At this point i am assuming
that there would have to be a Realm Controller/Coordinater to assist in
the syncing of all processes in a Realm. it is possible to have as many
Realms and Sub-Realms as desired. which is like the mud topology.
however this also works with the metaverse. in snowcrash, the top realm
was the Strip, and everybody else just registered their realms with the
Strip. Hiros house or the BlackSun were actually Realms or Sub-Realms
registered to certain locations and dimensions on the Strip. their
registered size doesnt actually affect how large the Realm is, that is
handled by the new RC/C. The cool aspect of this is the fact that it is
possible to register a portal (which of course is just another object in
a realm) in multiple Realms.

RE: Looking thru Portals
i think that this is best handled by the WWWInline node (someone correct
me if i have misinterpreted the node) because this allows the description
of the portal to be designed by the host that the portal links to. the
portal object is retrieved from the target host with all associated
information (like the hyper-link call itself). this, of course is one
way, unless the target host has a portal back at the portal entry point.
this permits cool registration in many realms to one entry point.

Idea: i still think that it would be a good idea to have object classing
or typing so that browsers would be able to replace objects that it
didnt want to take the time to render with something that is similar
and/or cached (locally or remotely). this meshes well with the idea of
stickfigure rendering for low bandwidth entities. some one with a cheezy
computer could still look at everything and get around, and then prompt
the browser to render something that the user wanted to see. (this is
like turning off the images in NN and forcing a selective download of
individual pictures.) we also need to keep inmind that the most
important part of this is not the graphics but the logical information
that it represents.

all comments welcomed.
thanks

jzbl <URL:http://www.semaphore.com/ezekiels>