A light would check the distance between itself and the mesh, minus the widest dimension of the mesh, for an approximation of whether the mesh is in range or not. Then it would go through and light each vertex.
Only calculating lights within the camera frustrum is a good idea, and I can add that right now. As for more advanced sectioning of the level, I think this is something that needs to be done all at once. I mean, the same system that optimizes lights should be used to optimize picking and rendering, etc. That system is going to be some kind of BSP tree, but I think it's too soon to implement.
Only calculating lights within the camera frustrum is a good idea, and I can add that right now. As for more advanced sectioning of the level, I think this is something that needs to be done all at once. I mean, the same system that optimizes lights should be used to optimize picking and rendering, etc. That system is going to be some kind of BSP tree, but I think it's too soon to implement.