Re: Avatar creation

Jon Green ([email protected])
Fri, 5 May 1995 10:52:31 +0100 (BST)


Aaron McMahon makes some interesting remarks about the potential for
modelling your own avatar. (Omitted for brevity.) Okay, here's
how I imagine mechanisms to work Sometime In The Future(tm).
My apologies if this is already painfully evident - call me a
newbie, if it helps soothe furrowed brows :).

Your avatar enters a room. As it serves a VRML room description at
you, the server queries your browser for a VRML avatar. The avatar
description contains - apart from the visual information - a number
of "key points" analoguous to those neat black-and-white roundels
you see on crash dummies (but not visible), which represent the
intersection of discrete elements of the avatar structure (arms,
legs, eyes, head, torso, whatever). The elements are encapsulated
and the relationship between the key points and the elements is
described. The avatar movement can now be described purely in
terms of the key points. The description can be as basic or
complex as it needs to be. The server and client will need to
negociate a suitable scaling.

Your browser also receives the avatar descriptions of other people
in the area (with key point info &c.). As you move within the room,
your browser passes telemetry information on your key points to
the server, which rebroadcasts them (alternative: the server has
identified the other visitors, and you send telemetry directly to all
interested parties). You receive telemetry for the other visitors,
and your browser renders their avatars according to their key point
descriptions.

The telemetry is obtained either by interpolation of movement
(simulating walking) or from VR hardware directly operated (data
gloves and all the other clunky stuff).

(Drawing from Mark's interpolation of a scene from Snow Crash, an
idea of how this all could work:- )
A and B are avatars. A's user wants to give a file to B's user.
A approaches B. [A's user tags a file for transfer, identifies
B as sole permissible target.] A picks a token from a "utility belt"
and holds it outstretched to B. B reaches out and takes the token,
which changes colour to indicate the start of transfer, [FTP transfer
starts] and again to signal completion [the transfer is now complete]
(or failure). B puts the token in its utility belt [places the file in
an "in-tray" for later filing].

-- 
[email protected]          Hyphen home page:             http://www.hyphen.com/
[email protected]           And mine:  http://www.hyphen.com/html/jonsg/
PGP key available on request              Opinions here are mine, not Hyphen's