Search code, repositories, users, issues, pull requests...
TentHash is a high-quality, reasonably fast, large-output hash. Its target applications are data fingerprinting, content-addressable systems, and other use cases that don't tolerate hash collisions.
Importantly, TentHash is explicitly not intended to stand up to attacks. Its robustness against collisions is only meaningful under non-adversarial conditions. In other words, like a good tent, it will protect you from the elements, but will do very little to protect you from attackers.
Also like a good tent, it is compact (a full implementation is around 50 lines of straightforward code) and you can take it anywhere (no special hardware instructions needed).
When I started work on TentHash I was unable to find any hashes that met all four of these criteria, and TentHash aims to fill that gap.
The table below is a comparison of TentHash to a selection of other hashes with outputs large enough to be used as data fingerprints. Some cryptographic hashes are also included at the bottom for reference.
