Re: [-empyre-] Introducing Adam and Fabio

["Fabio Zambetta" <fabio.zambetta@rmit.edu.au>]

Hi All,
thanks for having me as a guest, as well and thanks to Melinda for her
introduction. 
I've been reading what people have discussed so far with a certain
interest, especially the ones who approached SL from their perspective
of gamers (oh yes, I'm an RTS/RPG/strategic game freak myself) ;-)
I'll probably give a very different reading of SL from most of the
people on -empyre- because of very technical background: that doesn't
mean I'm totally extraneous to Art (witness is my current collaboration
with Adam on an Interactive Storytelling project), but that I'll be
trying to give a somehow different perspective.
How did I stumble in SL then?
As others I have built real world applications in VRML (both commercial
and research-related), but I've become very rapidly disillusioned about
its capabilities.
Even being a standard and a direct evolution of Open Inventor (one of my
favorite 3D engines from SGI), VRML 2.0 never reached the masses (for
lots of reasons that I could discuss in another post) :-)
That's what SL done well instead, and that's why it attracted me in the
first place: I thought I may actually start implementing smart 3D
characters in SL that would have a massive testbed to improve their
sensory and behavioral policy.
Unfortunately, the more I started to know about SL...the more I realized
about some of its fatal limitations! :-(
In my case especially excruciating problems are:

1) SL doesn't store object in hierarchies (it doesn't use a scene
graph). This is very problematic for me, because most of the
implementation of virtual senses I need to carry out relies on that.
2) The client/server architecture SL is based on is far from being
state-of-the-art both in many respects.  

As it stands, I'm sort of waiting to see what is going to happen: Once
both client and server are open-sourced I'll be able to make a more
informed decision on how much modification would be necessary to bend SL
to my needs.
I am also investigating alternatives, such as Multiverse (and using more
standard 3D engines such as Torque3D to build very rough initial proofs
of concept).
If this technical delirium has not knocked everybody out...I may
possibly elaborate a bit more on the goals of the projects in another
post... ;-)
Ciao

============================
Fabio Zambetta,  Lecturer, 
BP 215 (Games Graphics Programming) Program Coordinator
Room 14.11.02, School of Comp Sci & IT, RMIT University
GPO Box 2476v Melbourne, Victoria 3001, Australia
Mail: fabio@cs.rmit.edu.au 
Web: http://www.cs.rmit.edu.au/~fabio/
CRICOS provider code: 00122A