@^: Err no, sorry. (no one understands that all the raymarching stuff I'm doing is pretty much useless and just for fun >.>)

There are several reasons why you can't use raymarching the way you (and probably most ppl here) think:

- it's a totally different approach to render a scene, like really different.
- everything is calculated using a shader, which basically means you hardcode your graphics (geometry, reflections and everything). you can't dynamically add or remove anything. Animations are very limited aswell.
- you don't have actual geometry as you know it, you define your geometry with a (or several) mathematical funcion(s), which makes modelling complex objects almost impossible and can look something like this:
Code:
float sdTriPrism( vec3 p, vec2 h )
{
    vec3 q = abs(p);
    return max(q.z-h.y,max(q.x*0.866025+p.y*0.5,-p.y)-h.x*0.5);
}
(that's a hexagonal prism btw.)

the above also counts for materials, textures and everything else except normals (you can easily calculate normals with some additional samples).

so, yeah... except for some really special tasks it's useless.

POTATO-MAN saves the day! - Random