G'day all. Quoting Scott Turner <p.turner at computer.org>: > testR2' a n = (testR2' $! (a+n)) (n-1) Proving, once again, that the associativity of $! is wrong. :-) Cheers, Andrew Bromage