From: John Eagan <76130.2225@CompuServe.COM> Subject: Re: DESIGN: Doors between virtual worlds Date: 10 Aug 92 15:10:26 EDT Bob Pendleton (bobp@hal.com) writes: >> (quoting me, stuff deleted) ...... as part >> of the standards for any sort of networked VR, a set of basic >> attributes for a "world", covering things like Scale, Gravity, >> blablabla. > >Too restrictive. > >No reason why a world has to have gravity. I should be able to define >objects that accelerate anything within say 10 meters of the object in >any direction the object likes. This way I can build a Mobius strip >jogging tracks or hallways that twist all over the place while >maintaining local "up" as perpendicular to your current location. > >I should be able to get rid of gravity completely and just have sticky >walls... > >Virtual worlds are not models of the physical world. But they can be. Last sentence, agreed, completely. In fact, that's one thing that I wish was not such a sticking point in general conceptual discussions of "what VR is and will be". VR is not necessarily a "real world" simulation, and in fact I'd be much more interested in a virtual environment that was _not_ a model of the "real world", for the most part. The most interesting possibilities of VR, to me, lie in the potential for abstract environments that have little to do with the "Real World". That's why I suggested having a packet of attributes sent to a user entering a "portal", to tell the machine at the users end "what's what in _this_ World", such as what the scale is and other parameters (is this particular World the size of a house, or is it a "space" that's space in the astronomical sense (no ground, astronomical dimensions, 0 gravity, etc...) The idea is, to not make it restrictive at all, by making sure that when one enters a World, they get the setup of that World. No gravity? Fine. Open space with a body somewhere in it that has it's own gravity that begins to act upon objects within a certain distance? Great, define the gravity of the World as zero, and then make the gravity attributes of that body in there part of the parameters for that particular object that is that body. The idea, which I didn't elaborate on enough, is to _not_ restrict things, as much as possible, by making discrete Worlds, with their own characteristics, and have those individual discrete Worlds accessible through Rob J.'s "portals", so that the overall "Universe" would be as I described in my earlier post, a relatively tight cluster of some simple objects that mark where the "portals" are. Besides opening things up greatly on a conceptual "World design" level, this sort of approach would get us away from some of the problems being wrestled with here, with suggestions like having 128-bit data values so you can cover anything from microscopic dimensions to light-year distances in one humongous virtual world. Taking the approach of discrete Worlds accessed through portals in a Universe would be a simpler and more flexible way of doing things, I would think, whether on a PC system or a massive internet system. For the "low-end/desktop PC" avenue, things would be simplified by having a Universe with a sort of matrix of objects(simple spheres, icons, whatever) marking Worlds, and then reading in a data file for a particular World upon entry to the associated portal marked by one of those spheres/icons/whatever. For a network VR system, especially on a broad scale, this approach seems, IMHO, like the only _practical_ way of doing it. Individual Worlds in the net (the Universe) would be analogous to internet sites, with small markers like suggested earlier in a 3d grid/matrix, analogous to site addresses in the current net system. ****************************************************************** *John Eagan "humans is funny critters.." * *VR section leader -me * *Computer Art Forum * *CompuServe * *compuserve:76130,2225 * *internet:76130.2225@compuserve.com * ******************************************************************