On 22 Mar, 09:31, Bryan Olson <fakeaddr... at nowhere.org> wrote: > Even a hash function that behaves as a random oracle has > worst-case quadratic-time in the algorithm here In which case inserts are not amortized to O(1).