Does clipping imply that a 3D model would be redrawn every time the camera angle is changed in, say, a 3D modeling program? How does clipping compare to drawing all the triangles, and then rotating the screen image wrt to how the camera rotates?
Will it be more efficient to rasterize a triangle depending on how the view frustum intersects with the triangle? The two situations on the left would lead to rasterizing more triangles than before.
Is there a quick implementation to find the part in the canvas of a shape at the edge?