Re: [mmox] The Story So Far...

[Jon Watte <[e-mail]>]

Charles Krinke wrote:
> In any case, "Can we define a message format for avatar presence that 
> contains fields to allow 3D interactive browsing that is reasonable 
> for the SecondLife and OpenSim teams and also flexible enough to be 
> useful for other teams?"


Re-phrase that as "that is reasonable for all likely adopters of the 
standard" -- why special case Second Life or Open Sim?

So, what is an avatar anyway? I think it has one or more meshes (that 
are not standard), draped over a skeleton (that is not standard), with 
one or more textures applied (that are not standard). Avatars also 
represent users, which have identity. Avatars can move (physically, with 
collision detection), animate (interactions, emotes), voice communicate 
and text chat.

If you can send that information ([identity, mesh, texture, skeleton] as 
set-up, and [movement, animation, interactions, voice and text chat] at 
runtime) then do you have an interoperable avatar?
And, if so, what separates an avatar from another animated, movable 
object? Could the same way that avatars are interoperating be used to 
interoperate other simulated objects? From a remote systems' point of 
view, there's not that much difference between a human-driven avatar, 
and a tumbling rock -- the remote system can't be expected to run the 
simulation for either one, so it needs to have that stream of behavior 
to represent it.

As far as I know Wonderland, Qwaq, Multiverse and other possible 
adopters, the above representations maps fairly straightforward to their 
systems, although any system will have to expect to write adaption code 
to accept "arbitrary" external meshes as avatars.

Sincerely,

jw