Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!zaphod.mps.ohio-state.edu!uwm.edu!ogicse!milton!hlab From: piggy@gargoyle.uchicago.edu (La Monte Yarroll) Newsgroups: sci.virtual-worlds Subject: Re: GoggleVox Update, plus some other things. Message-ID: <1991Apr11.185440.23383@milton.u.washington.edu> Date: 11 Apr 91 16:38:13 GMT References: <1991Apr9.203748.7430@milton.u.washington.edu> Sender: hlab@milton.u.washington.edu (Human Int. Technology Lab) Organization: University of Chicago Lines: 71 Approved: cyberoid@milton.u.washington.edu In <1991Apr9.203748.7430@milton.u.washington.edu> harry@harlqn.co.uk (Harry Fear nhamm) >(3) Object update accountability. ... >(b) learn new desciptions as painlessly as possible). But there will >always be occasions when you have no choice - you *have* to have a >huge update, such as when you enter a VR session for the first time. >What happens? You might chose to wait for all the zillions of objects >to be described to you, or it might be described to you with a level >of detail proportional to a combination of your point of focus and the >distance you are from the objects. Usage note: I use quotes (") to mark terms which I'm intensionally using as vague technical terms. This is a problem I've been considering lately. The model I've been thinking about comes from a graphics transmission technique I read about some years ago (in DDJ?). The idea of the technique is that you can quickly get a low resolution image on your screen. The image resolution improves over time as you receive more data. This way you can decide early on whether or not you are interested in the particular image, and about quickly if you decide you don't want it. The algorithm for generating this data stream is approximately as follows: Divide your image into a small number of symetric regions (say, 4 squares). Decide on an "average" colour for each region. Transmit these data. Recurse breadthwise. Subsequent colours are coded as changes from the "average" colour already placed in the given region. I've been thinking about 3D generalizations of this algorithm, which seems to be what we want for VR. The obvious extension of dividing space into cubes and picking an "average" colour, seems inadequate to me--it seems that you'll get stuck with a bunch of opaque blobs. Is it perhaps possible to come up with an "average" shape for any spacial region--something like approximating a sphere with an icosahedron? >It seems to me that there must be a well defined accountability, but >I'm not sure where that should lie - if several individuals enter a >VR, that VR must exist on some machine, which is responsible for >telling the users of its changes, just as you should tell that machine >of yours; but should it be this VR machine's duty to tell you about >other individuals present, or should any changes from these >individuals go directly to you (and other users present) in some form >of selective broadcasting? I'm inclined to think that the former >would be more efficient and more manageable. If you do indeed have a "VR manager" in the middle, would it not make sense for this manager to cache info on it's participants? >Just some idle ramblings - ignore them at your leisure. Just some more idle ramblings - ignore them at your leisure. :-) >-- > Harry Fearnhamm, ,---.'\ EMAIL: loki@harlqn.co.uk > Harlequin Ltd, (, /@ )/ ...!ukc!cam-cl!harlqn!loki > Barrington Hall, /( _/ ') VOX: +44 (0)223 872522 > Barrington, \,`---' FAX: +44 (0)223 872519 > Cambridgeshire, DISCLAIMER: Nothing is True. > ENGLAND. Everything is Permitted. La Monte H. Yarroll Preferred: piggy@gargoyle.uchicago.edu Work: piggy@mwc.com Home: piggy@baqaqi.chi.il.us often: postmaster@clout.chi.il.us