[rescue] Is this a POS?

[joshua d boyd]

On Wed, Aug 22, 2001 at 01:58:31AM -0400, Patrick Giagnocavo wrote:
> ok, Joshua, tell me what you think about Radiance (raytracing/radiosity
> renderer) and BMRT (blue moon rendering tools, written by the famous Larry
> Gritz - check the Pixar credits).

BMRT is solid, but I think it has many of the same short comings of
PRMan.  Plus, I wish it supported RiPoints (for real preferably, but even
as simple RiPolys would be close enough) and Subdiv. surfaces.  

But short comings in PRMan, I am refering to the shader system.  While it
is what made PRMan great, I don't entirely think that it is up to snuff
for the future.  For instance, I don't know that it can easily let you
view some things as voxels while other are more traditional surfaces, or
how well it is with letting you treat an object as a solid, then control
how the ray passes through it.  This last could be possible in BMRT,
except I don't believe that the shading language allows for it.

I must admit that I'm not a PRMan guru, so I could be missing things about
it.
 
> I came across a Cray-optimized, raytracer and now can't remember the dang
> name.  There were source and binaries for Linux etc too, and I think it was
> already designed for MPP or PVM parallel processing.  Should be on
> freshmeat (maybe it was called Tachyon???)... ah yes:

Note, in this case when you say Cray optimized, you are talking about the
Crays built out of large numbers of Alpha processors, not the more
traditional vector Crays that we all wish we knew and love.
 
> http://freshmeat.net/projects/tachyon

I'll have to look into it.  Initially, it appears that it still lacks to
difficult but important features for something that would want to try to
make it outside the scientific world.  Specifically focal depths and
motion blur.

-- 
Joshua D. Boyd