From: broehl@sunee.waterloo.edu (Bernie Roehl) Subject: Re: TECH/PHIL: World Descriptions Date: Sat, 4 Jul 1992 22:49:46 GMT Message-ID: Organization: University of Waterloo In article <1992Jul4.052915.25360@u.washington.edu> bergen@seas.gwu.edu (Daria Bergen) writes: >Bernie Roehl writes: >> In order to share world descriptions, we need to agree on standards for >> some basic aspects of space and its metrics. In most cases, these >> choices are arbitrary (e.g. left-handed versus right-handed coordinate >> systems, degrees vs radians for angular measurements). > >I'm not sure why everyone should have to agree to a specific >standard. We should be able to design in whatever unit of >measure and coordinate system we are comfortable with. Yes, but in order to share information we need to have some idea of what form it's in. I suppose we could (as you implicitly suggest) have every world assign itself a "handedness" attribute; however, this wouldn't be necessary if we all just picked one and stuck to it. Individual world- builders can, of course, use whatever they like; however, when they *export* their world for others to use, they ensure that it gets flipped to the agreed-upon 'handedness'. >If I develop my code using a RHS and I start >participating in a world which uses a LHS, I just set a >switch and flip z before I send/receive anything. Right, except that instead of labelling each world as one or the other we can just agree to send/receive everything the same way. Makes the code simpler and the standard smaller. >>LIGHT SOURCES, COLORS AND TEXTURES >In my opinion, you should support anything we could >possible ever need or calculate in real time. The RenderMan >standard does a great job in enumerating all attributes of a >scene. Hmm... I'm only slightly familiar with RenderMan. I'll check locally for info; is there a quick summary of it out on the net anywhere? >It would be up to the local machine to decide which >attributes it can use and which to ignore. Yes. >I don't think anyone needs to know what driving my behavior. >I could have a great physically based model controlling me >or a keyframe file, it shouldn't matter to anyone else. I agree, but if you've been following the discussion lately you know there are (at least) two differing views on this. Both have their merits. -- Bernie Roehl, University of Waterloo Electrical Engineering Dept Mail: broehl@sunee.waterloo.edu OR broehl@sunee.UWaterloo.ca BangPath: uunet!watmath!sunee!broehl Voice: (519) 885-1211 x 2607 [work]