Just something I whipped up in a couple days. There's no textures or anything, just a proof of concept to show that raycasting is a feasible way to render a game like Minecraft. I will eventually add textures, dynamic chunk loading, and more. The advantage of raycasting over polygons is that effects like lighting and reflections are trivial to do - just send a ray towards the light and see if it hits anything for the former, or send a ray in the direction of the reflection for the latter.
This is impressive, but what kind of performance could be expected after you implement the rest of minecraft into that? I'm still disappointed from the "lighting update" ages ago...
Rollback Post to RevisionRollBack
You may want to read my post while you still can, some of them have been randomly disappearing lately *cough*
Quote or reply my post if you expect a response, thanks!
If it was the precomputed block based lighting, next to no performance impact, but for dynamic shadows, which would mean casting another ray, there would be a noticeable hit. Keep in mind that this is quite unoptimized, so there is still room for performance improvements.
Added ambient occlusion. I also eliminated two branches from a loop, and it gave me a a 33% decrease in frame time. (That's a 50% increase in frame rate.)
You may want to read my post while you still can, some of them have been randomly disappearing lately *cough*