RE: size of interpreted code