Hi
Not sure if anyone here's tried this before:
http://www.fabfabbers.com/openscad/
I've now implemented an image-based CSG algorithm (similar in function to OpenCSG, but written from scratch), in WebGL. It was really quite a challenge as all current implementation of this are in desktop OpenGL and rely on things which aren't in WebGL spec, such as depth textures.
I'm thrilled to say it actually works, and now you get lightning fast previews of the object you're creating.
EDIT - Just to add, the caveat to that is just as in the native OpenSCAD software - if you use operations such as hull or minkowski, particularly high up in the CSG tree, then it's still going to have to calculate the geometry. Image-based CSG can only deal with boolean ops.
Marcos
Edited 3 time(s). Last edit at 04/29/2013 03:07PM by marcosscriven.
