> But for just the cost of doubling our space, we can use two Bloom filters! We can optimize the hash function to make it more space efficient. Instead of usi

> But for just the cost of doubling our space, we can use two Bloom filters! We can optimize the hash function to make it more space efficient. Instead of using remainders to locate filter positions, we can use a mersenne prime number mask (like say 31), but in this case I have a feeling the best hash function to use would be to mask with (2^1)-1.