Talk:RAND

Computational Class

Unless I'm mistaken, it looks like a RAND program only has a finite number of memory locations it can affect (that is, the variables it defines) and each of those memory locations can only hold a finite value (7-bit binary integers?) That means the language isn't Turing-complete. The fact that you can save variables on the stack might mean you could simulate a push-down automaton, but since it's only a single stack, you wouldn't be able to simulate a tape with it. --Chris Pressey 04:11, 24 December 2010 (UTC)