[geeks] lazy thursday night

[Joshua D Boyd]

On Fri, Mar 15, 2002 at 12:02:25AM -0500, S. Gwizdak  wrote:

> It's for a class I'm taking. Since all models loaded in are defined in
> terms of surface arrays, they are initially reduced to polygons to begin
> with. The first part of the class was spent covering basic 3d translations,
> hidden surface removal, animation, etc... Now we have rather nice programs
> that can handle extrusions and revolutions to create rather complex 
> shapes. (Spheres, toruses, stairways.) So the next step is to do some 
> simple raytracing. Additionally 3DS/Wavefront file formats were figured out
> in the class, so we have a bunch of killer models of really neat things.

Ahh.  I love Wavefront files, for polys at least.  I never got around to 
figuring how to use it for other surfaces (no program seems to export to 
those features anyway).  In truth, you can always tesselate a model down to
where you can't see if it is polys or something else, and in fact, you
pretty much have to for scan line renders.  Still, there is a certain purity
to rendering the surfaces directly, plus in my mind, it would make things like
volumetrics easier, but then, I don't write renderers, so I don't really know.

I keep thinking about doing it.  If I could do motion blurring, then it might
be worth while.  Otherwise, it wouldn't be good enough to be worth bothering
for my purposes.
 
> I'm currently starting the first major rewrite of what I have already to
> make it significantly more organized and efficient. I suspect the second
> major rewrite will occur when I want to add things such as bezier surfaces.
> The class is set up as a bunch of small projects, but currently there is a 
> competition between me and 5 other people to add new features. The latest 
> one: outputting to postscript. I hope to have a bunch of posters of toruses,
> the enterprise, and wireframe naked women.

Dude.  Add motion blur, and you will have beaten the vast majority of free
raytracer authors out there.

-- 
Joshua D. Boyd