From: John Draper Subject: Re: DESIGN: 3D computing vs. virtual reality: what's the difference? Date: 12 Dec 1995 16:31:02 GMT Message-ID: <4akao6$2ru@stc06.ctd.ornl.gov> Organization: Oak Ridge National Laboratory John Costella wrote: >> DESIGN: 3D computing vs. virtual reality: what's the difference? >> This seems to be the question of the moment. >There will be widespread disagreement on what the various terms >mean. With that caveat, let me offer my humble interpretation. (The >terms are of course highly cross-connected, so the following linear >textual description is quite disordered!) I offer some perspectives on VR terms as published by the National Research Council (NRC), not to disagree or criticize but rather to point out some alternatives. You anticipated some disagreement; I think that any discussion of this area can be made difficult by the different ways in which terms are used at various institutions. This is a confusing part of the field that is probably unavoidable given its state of development. As the NRC says, "There are currently no precise and generally accepted definitions of the terms being used in our area of interest." Your terms are as good as anyone else's. I quote the recent National Research Council book, "Virtual Reality: Scientific and Technical Challenges" (Washington, DC: National Academy Press, 1995) not because I accept it as an authority but because it provides a different lexicon. >VIRTUAL WORLDS >============== >In my mind, "virtual worlds" is the most general, encompassing >term covering the sort of stuff we discuss here on sci-vw. In the National Research Council's book, the authors use the term "synthetic environment" as their "most general" term. Synthetic environments (SE) include virtual environments, teleoperators, and augmented reality. >my lexicon, *any* constructed or simulated system that is a >metaphor for an *action* in the real world, is a virtual world. The NRC says that a VR system consists of a human user, a human-machine interface (HMI), and a computer. The purpose of the system is to "sense, manipulate, and transform the state of the human operator...or to modify the state of information stored in the computer." (This definition seems a little sparse, and so may include a lot of technology that many of us wouldn't call VR.) >A word processing document is not a virtual world. A spreadsheet >is not a virtual world. An image of a building is not a virtual >world. All of these things are just digital representations of other, >more traditional objects in the real world. However, a windowed >operating system *is* a (very primitive) virtual world. It is a >"desktop metaphor"---translate that to a "virtual desktop world", >if you like. If it were just a bitmap of a wooden desk, it would >not be a world. But we *interact* with a windowed system in a way >that, to some extent, simulates the way we interact with a real >desk. Note that the interface devices (mouse, keyboard) are *not* >really natural ways of interfacing. But there is still a virtual >world in there. Ok, one criticism...your definition of a virtual world hangs on the concept that a virtual world presents a metaphor. Is a spreadsheet that uses a notebook metaphor and allows one to "page" through multiple windows a VW but a spreadsheet that just opens one window not a VW? And, generally speaking, a mouse is a metaphor for a finger, so are all mouse-driven interfaces VW's? And finally, some VR interfaces allow people to interact with computer-generated worlds in a way that is NOT metaphorical...a driving simulator, for example. Are these not VW's? >And then we have all of the "higher lifeforms" of virtual worlds, >such as what most people here are working with or on. Again, I >draw the distinction between *objects* and *actions*. A 3d >rendering of some object (real, simulated or imagined) is *not* >a virtual world. A movie is not a virtual world. Images of PET >and MRI scans are not virtual worlds. But if you can *move* >the object around, or move around it; if you can *make* it do >things; if you can *interact* with it;---then you have a virtual >world. According the the NRC, some of these applications are SE's but not VR, but rather "teleoperators." "A teleoperator system consists of a human operator, a human-machine system, and a telerobot," they say, and "The purpose of such a system is to facilitate the human operator's ability to sense, maneuver in, and manipulate the environment." (I'm not sure all three are absolute requirements, myself: some teleoperators only maneuver and sense.) >I realise that the meaning I attach to "virtual world" encompasses >many, many areas of computing and technology, whose creators >would probably not consider to have anything to do with "VR". >I see in "virtual worlds" a term that *links* our field with many >others---a concept, rather than a specific technology. I think that it may be the requirements for the HMI that forge the strongest links among all these worlds. Just an opinion. Dr. John V. Draper, Ph.D. Robotics & Process Systems Division Oak Ridge National Laboratory e-mail: draperjv@ornl.gov