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

The approach used in IEEE 1278 seems to differ radically from the
function OGP is assumed to permit, from a glance: as you said it lets
otherwise incompatible domains of simulation (call them grids)
exchange just the amount and type of information needed to remain
consistent between themselves.

Transposed to a VW, this would mean you can use the SL client to
connect to an SL grid, then your current simulator connects you
through to other environments (WoW, Solipsis, etc.) and does the
"translation". If the "foreign" domain/grid does not support some
Agent Domain services or something else, then the "local" domain can
then provide its own as a substitute, while the client is tasked with
interpreting any native data that rides through (like 3D
representation and sound format).

Is it right ? Maybe detailing a specific example would help pointing
out what should be and shouldn't be in such a protocol. (We're trying
to minimize the amount or complexity of translation required both at
the gateway and the client, aren't we ?)

On Thu, Feb 19, 2009 at 6:55 AM, Jon Watte <jwatte@gmail.com> wrote:
> Dan Olivares wrote:
>>
>> objection that I've heard so far is that it requires a lot more work
>> (service calls?) to develop a shared scene graph and requires a
>> significant more amount of trust then an A-->B teleport does.
>>
>
> There is no shared scene graph. Each server simulates "its" objects, and
> translates from the "interop" protocol to its internal system at the
> gateway. Think of it like a more modern version of IEEE 1278 (which has been
> a great success as far as interoperable distributed simulation goes).
>
> Sincerely,
>
> jw
>
> _______________________________________________
> mmox mailing list
> mmox@ietf.org
> https://www.ietf.org/mailman/listinfo/mmox
>