High Speed Hashing for Integers and Strings · HackerPapers discussions

High Speed Hashing for Integers and Strings — mentioned by thomasahle

Sun, 30 Apr 2023 05:29:54 GMT

They hardly forgot it. It's just that today we prefer hash functions that are seeded, so we use a random number instead of the "magical" fibonacci constant. See for example multiply shift hashing in https://arxiv.org/abs/1504.06804

Paper and discussion history. A mention is not an endorsement.

High Speed Hashing for Integers and Strings — HN submission

Thu, 11 Aug 2022 17:59:18 GMT

High Speed Hashing for Integers and Strings (2015)

2 points; 0 discussion replies on HN. Paper and source history. A submission is not an endorsement.

High Speed Hashing for Integers and Strings — HN submission

Fri, 12 Feb 2021 10:48:48 GMT

High Speed Hashing for Integers and Strings (2015)

28 points; 13 discussion replies on HN. Paper and source history. A submission is not an endorsement.

High Speed Hashing for Integers and Strings — mentioned by 19eightyfour

Fri, 09 Jun 2017 09:04:02 GMT

I think it is probably universal, but I have to show it to know for sure. I'm going to keep calling it universal until I know more. I think I note it in the README, but some reasons I think it is probably universal are because tifuhash can be parameterized, and it has good independence properties ( passes PractRand ). Showing it is universal or not is another step. I suppose I could experiment, to see if the collision probabilities match the criteria for universal or k-universal. Or maybe I could show it. For showing it, my next step is to read over this paper[1] to see if I can use its…

Paper and discussion history. A mention is not an endorsement.