and if you assume the list of tuples is sparse, and you keep them ordered, you can eliminate large groups of tuples (for example in intersection) without ever hashing them by knowing that they were in between two far entries in a skip list