> Nowadays it would be a complete waste of time to try to improve MD5 Not as a t...

adrian_b · on Sept 20, 2022

I agree about the usefulness of thought experiments, but there are many other much more useful thought experiments, even in the domain of hashes.

A large number of people have thought for several years about how to improve MD5, and the results were SHA-1 and RIPEMD-160.

It is very instructive to study the evolution from MD4 to MD5 and to SHA-1 and RIPEMD-160, but it is very unlikely that attempting 30 years later to do something better than those, without completely changing a hash algorithm structure that is now well understood as being inappropriate, can teach you anything.

bombcar · on Sept 20, 2022

The useful thought experiments are often to realize WHY the "trick to fix MD5" you came up with doesn't work.

bawolff · on Sept 20, 2022

People are giving you shit, but people thinking about this sort of thing has yielded useful and interesting things like https://marc-stevens.nl/research/papers/C13-S.pdf

Of course in production, just use sha256, but there is nothing wrong with thinking about unorthodox solutions.