Re[2]: [Haskell-cafe] can there be (hash-table using) O(n) version of this (I think currently) n log n algo?