oleg@okmij.org wrote:
The others have already pointed out the problem with the imperative solution, which used the mutation of the global state with the new random seed. Imperative approach is indeed often a problem.
As Daniel Fischer pointed out, my immediate problem was that iterateR never finished, even though it did produce results lazily. I missed the subtlety that access to results didn't mean access to the state.
There is a simple solution however. [Snipped ASCII art of head exploding...]
module RList where
[Nice code snipped] So, if I may paraphrase, oleg's solution works by reifying the implicit continuation in my iterateR's recursive definition into a data structure that is explicitly forced with pullR and its callers. Fascinating... Thanks to all who responded. Dale