frerichs@ux1.cso.uiuc.edu (David J Frerichs) (12/06/90)
recently, I was involved in a discussion of how best to represent the net in a Virtual Reality interface. I will share the concepts we came up with... first, you start off with a base locational structure, like a grid, to define "geographic locations" on the net. It could be a sphere like the surface of the Earth, but not necessarily. Next you assign each server a location (address) in relation to the reference structure. Servers with in a common group (University, Company, etc...) would be located in the same "geographic" area. This could easily be translated from telnet addresses for compatability with current systems. Once we have these things determined, we decide on how someone is going to acess other machines on the net. We could give each traveller an id based on his home server, his specific user tag, and his current location on the net. With this done, it becomes clear (yeah... tell us more) on how to transmit visual, audio, etc... representations of the net to the traveller. The traveller will transmit his net id to all servers within a certain radius of his "geographic" location and they in turn will send back how they look to him according to his access to a particuar server, his distance, and orientation. This could remove alot of the visual processing responsibility from the traveller and put it on the server(who most likely has more power to spare.) Another modification is if we put in the ability to "fly" through the virtual sky over the net, the traveller would transmit his ID to a greater radius of servers,... I'm sure more can be though of... To me this seems to be a simple solution that can be used in a HIGH bandwith net structure (ie fiber optic all the way around)... -dfRERICHS Univ of IL, CU Dept of CompEng IEEE/SigGraph