According to me, POV-SDL has two aspects:
  declarative features to describe objects in scenes
  programmatic features such as loops, macros, includes, etc ...

My opinion is that POV is VERY efficient in the declarative aspect, and less
in the progrmmatic one. I can't see how it would be more compact and easy
than

sphere
{
    0, 1
    pigment
    {
        crackle scale 0.2
        color_map {[0.1 color 0.1*Red] [0.15 color 0.5*Green][1 Blue]}
        scale 2*y
    }
    normal {bumps 0.15 scale 0.05}
    finish {specular 0.3 roughness 0.01]
    scale <1.2, 0.5, 1>
    translate 10*z
}

But POV lacks programmatic features like recent langages have that make them
more efficient in programming, reusability, modularity, readability, etc ...
The effort should be made first on these rather than how to declare a sphere
in the scene, or how should the brackets be, curly or square, which are very
secondary concerns. One should remind the history of ray-tracing and that of
POV.

I totally agree that it is time now to go one step further, and that POV-Sdl
needs refurbishing here, enhancement there, redefinitions and complements
(compliments also ... :o)).

In addition, the isosurfaces are one of the most powerful object type you
can ever think of in ray-tracing (blobs are also powerful, but they lack
desiging tools and the are under-employed).

To finish, I can't imagine making 'reasonably advanced' scenes without a
minimum of programming skills and a minimum of geometry knowledge. For
example, the first thing you must have in mind is the order of
translations/rotations for you objects. However, gurus are always there to
work out complicated algorithms for you, provided the language allows them
to make it reusable, and you to reuse it.


    Bruno

Post a reply to this message