From: Jon M. Taylor <taylorj@ecs.csus.edu>
  To  : ggi-develop@eskimo.com
  Date: Wed, 19 Aug 1998 17:49:51 -0700 (PDT)

Re: LibGGI3D RFC

On Thu, 20 Aug 1998, Olivier Galibert wrote:

> On Wed, Aug 19, 1998 at 03:23:41PM -0700, Jon M. Taylor wrote:
> > 	?  I'm not sure what you mean.  If you mean "and then render another
> > set of triangles to the buffer and use them as a texture again", sure.  Of
> > course, this secondary texture buffer scheme will have to be supported by
> > the display target and the texturing shader, but it could be done.
> 
> Yes, I meant that.  I saw a demo of the results of such approaches, it
> was  impressive.  IIRC, three passes  were done (render, use result as
> texture for second rendering, use  second result as texture for  third
> rendering, show third  rendering),  in realtime (60fps, 3200x768,  not
> fully boosted Infinite Reality).

	Yeah.  Fully possible.  Just write the approprate shader in a
rendering target and pass it the necessary info. 

> Another  use of 3D  cards here (french CS research  center)  is to use
> them  to  compute  the  visibility factors in   a scene  for radiosity
> computing by having  the  hardware render   the scene without   lights
> effects,  encoding each patch with  a unique color,  and then transfer
> back the picture and count the pixels.  I guess it will be possible to
> do that with libggi3d for  the rendering and  libggi2d to get back the
> result?

	Sure.  Write a modified radiosity-without-lights shader (probably
it would do everything in greyscale) and then draw each patch through the
shader and then through another shader which applied a color filter to the
greyscale patch before redering it to the final display target.  Is that
what you were looking for?

Jon

---
'Cloning and the reprogramming of DNA is the first serious step in 
becoming one with God.'
	- Scientist G. Richard Seed