I implemented the marching cubes algorithm, and after correcting a dozen errors in my 256 cases, and writing some hole-patching code, it works beautifully! I can mesh the entire ribosome quickly and easily. For example, I meshed the RNA in 1kpj.pdb at 4 A resolution in 8 seconds! I don't know if that's fast for other implementations, but it's good enough for me.

Well, I've given up on l-surf for now because it didn't work. Instead, I've started on m-surf, which is trying to do the same thing (meshing molecules), only it uses the Marching Cubes algorithm which is a whole lot more elegant and simple.

I am currently working on a little project called l-surf when I should be doing something else. l-surf is my attempt to to create a surface mesh around a PDB model of a molecule. It's coming along, but slowly.

I wish I could find other slashdotters who are into molecular modeling to collaborate with.