Path: utzoo!utgpu!news-server.csri.toronto.edu!mailrus!wuarchive!cs.utexas.edu!sdd.hp.com!uakari.primate.wisc.edu!dali.cs.montana.edu!milton!mike@x.co.uk From: mike@x.co.uk (Mike Moore) Newsgroups: sci.virtual-worlds Subject: Re: Yo! people! Summary: A personal view of the future of VR Keywords: personal dreamer future Message-ID: Date: 31 Aug 90 11:36:35 GMT References: <31304@unix.cis.pitt.edu> Sender: hlab@milton.u.washington.edu Reply-To: Mike Moore Organization: IXI Limited, Cambridge, UK Lines: 120 Approved: hitl@hardy.u.washington.edu Filip Gieszczykiewicz writes: > > > Greetings. > > I understand that many people are "thrilled" to > post and read this group - but. > > [stuff deleted] > > I'm no PhD. I don't think too many reading this > group are. You have a program, _post it_. > Just a small portion will be fine. Show us what > a VR program LOOKS like. I learn by examples, > give me one. > > You have a system that you do your stuff on, give > us the setup! Gosh, I write stuff on a 386 > with a VGA using C, what do you use??? > > [more stuff deleted] > > It does not have to be RealTime or even have fancy > graphics. It can be just a set of subroutines - or > even just pseudo-code - but it must be REAL... not > THEORY... > > [and more] > > P.S. If you don't like the above, you are probably a > PhD. Some of us just can't stand THEORY... Please note: this is NOT a flame Let's swap pseudo-code! Actually there are a lot of people out there; real, working, money-making companies; who have implementations of VR on all sorts of different machines, and they demo them etc... BUT, I know what you mean, the guys on this BB can go above my head, but then *I* think with my fingers! But, I have something in common with them. We are all dreamers. And we can see this dream becoming reality. Space flight became a reality, and most of the world got bored (not me), and then they wouldn't let me GO OUT THERE! So we work on another one, and see what happens this time. Enough of this rambling, what I propose is this: Imagine a machine, with a programming language that hasn't been designed yet, and an operating system that hasn't been built. Imagine a wondrous piece of hardware that ran in the billions of MIPS with multi-multi- multi-parallel chip logic (around, say, 512K cpus), imagine a superb human interface system and that we are the software engineers/artists. Now, call it a single-user system, connect it up to an optic fibre cable that goes to EVERY home in the world and, finally, sell it for the same comparative price that a colour TV set is now (comparative to today's average earnings say). They say never try to predict, but I sincerely believe this will be a reality in less than 20 years (nice safe margin actually!). Now, let's program! Top-down approach, until, not only have we designed a new language, we have the op sys level that is so vital to the work. Of course by the time you and I get there, at least half a dozen others will be there (some a lot bigger fish than just two poor hackers) and standardization will start to arrive. And you and me? Well, we'll be left clutching our baby, thinking, 'we could always build ourselves a virtual reality!' At least I can GO, this time. They can't stop me, except by pricing it right out of my reach (like the space program). Let's take the current discussion about modelling humans. I don't want AI (that will never happen within 20 years), I want a Virtual Reality which human beings populate, and software artists create. Cities, buildings, anything. We need a co-ordinate system, (x,y,z) seems good enough. We need to define what goes where (like, the position of the individual within the cyberspace is looked after and manipulated by the individual's machine. Whereas the description of a building, the object's within it, the physics present in the environment and the allowable actions of the individual and reactions of objects are defined by the host machine, the designer-artists machine. There is some standard software floating around the entire net, that makes each of the designers' machines into a cohesive whole and allows individuals to travel between them in some understandable way. [maybe (x,y,x) isn't good enough, how does my brain deal with buildings which float in the sky? Especially above my head?! With no ground? Do I need a two-dimensional projection of three-dimensions, like the Earth's surface say? But NOW *I'M* beginning to sound like a theorist!] Anyway I said let's code. So here is level 1 of a top-down approach of the design of a building within Cyberspace (I'm poor enough to own a baby designer's machine say, with only enough disk - or is it holocube? - memory for one reasonably large building, and I've bought a plot of cyberspace real-estate on the outskirts of a cybercity and I have planning permission). Name: Suscatrel Temple Location: 123,2867,987 *in whatever units the net-siting software uses Shape: Ah... Hmmm.... Well, OK, I'm stuck. I need a way to specify what my building looks like. And I need a formalised way of doing it. So we talk theory a while, and then we get back to coding (if we don't get discouraged), and we hope that we'll have something really decent for when that imaginary computer arrives (it will). As it happens, we have some decent software that allows me to design my building. So off I go and plug it into my baby designer machine. Now, I need to model a three-d, inside view of my building. Well, we're getting there. Now I need a three-d modelling system for the objects. OK. Then some decent interface equipment. Ah. We're stuck. There isn't any (not yet). But then that's hardware. What I'm trying to illustrate is that a lot of the software we need already exists. What doesn't, is on it's way, and all that is really needed is some way to glue it all together, then wait twenty years for the hardware to catch up! Meanwhile we can refine, and discuss, and dream. Now if you'll forgive the length of this article, I'm going to unjack and feed the meat. -- --- Mike Moore mike@x.co.uk or mike@ixi-limited.co.uk Usual and obvious disclaimers... etc