Re: [grt-talk] Hello

[Nikodemus Siivola]

On Wed, 9 Apr 2003, Simon Adameit wrote:

> FreeBSD

Great! Can you please check out the current CVS HEAD and run:

 ./dist.sh pre0.1

to make a tarkit, unpack it, enter the pre0.1 directorym and verify that

 lisp -load grt.lisp -load examples.lisp -eval '(all-examples :screen t)'

works. Should go without a hitch assuming that you are on CMUCL. If you get
warnings about missing window-component, the images should end up in
examples-direcotry anyways. Comments on the README, INSTALL and LISP are
also urgently welcome.

> Perhaps the texturing system.

Have you looked at the current code? Apart from more patterns / pattern
generators the major addition for 0.2 will be perturbated normals.

> What are your future plans for GRT?

Check out Development Notes on the homepage.

Number one goal is to make GRT a good classical raytracer: this includes a
number of primitives, light sources, volume rendering, efficiency measures,
sophisticated texturing capabilities, some GUI tools that are integrated
into the lisp-environment (color pickers, spline-drawing facility, etc),
OpenGL preview, and animation support.

Once we get there (not all the way necessarily, but far enough that the way
is clear), it's time to start considering the role of alternative rendering
strategies in GRT: backwards raytracing, path tracing, monte-carlo methods,
shader based architecture, scanline rendering, classic radiosity, etc.

Time will show what gets implemented, but in the distant future I hope that
GRT will blossom into a powerfull image synthesis system, not "just" a
raytracer.

For today however, writing a raytracer is hard enough. ;)

Also, a great deal depends on the people who take an interest in GRT. Since
GRT's internals are a great deal more accessible than any other raytracers
that I know of, I hope that once we start getting actual *users* a fair
part of them will actively tweak GRT for their own benefit.

Cheers,

  -- Nikodemus