Re: Standardizing coordinate systems, units of measure

Eric Van Hensbergen (
Fri, 05 Aug 1994 07:59:49 -0400

In message <>, writ
>Eric Van Hensbergen <> writes:
>> coordinate_system {
>> units: 1000.0; // scale: 1 unit = 1000 meters (or a kilometer)
>> horizontal: +x // x goes from left to right
>> verticle: -y // y goes from bottom to top
>> depth: +z // z goes from out to in monitor wise
>> }
>What's the difference between this "coordinate system"
>(as you call it) and a rather limited brand of transform?

Not much really, in fact it could be implemented as a transform matrix in
the software end I suppose, I was just trying to define a file format for it.

>You are still implicitly defining a standard coordinate
>system ("horizontal" relative to what?).

horizontal relative to the screen (sorry if that wasn't clear) x in the
above example x goes from left to right, relative to the screen as you are
looking at it.

>The notion of coordinate system becomes more meaningful
>when your virtual reality is inside other manifolds, where
>the choice of coordinates is more arbitrary -- to the extent
>that different choices might not even be related by the
>projective transforms that graphics hardware can quickly compute.

Maybe I'm not catching what you're trying to say here in this last paragraph.
The point of my proposal is that everyone could make their tools and software
in their own coordinate system, and attatch a cooridinate_system property or
attribute which described their particular coordinate system (using the general
format shown above), and that way people using other coordinate systems would
be able to use their models, objects, whatever by applying the coordinate
system transform to 'import' the model into their coordinate system.