10 ms·
I heard Fedora recently started using this for password hashing in /etc/shadow. My question is - why? Argon2 is proven sturdy, well known, and there's many impl
by c7DJTLrn 5y ago
I heard Fedora recently started using this for password hashing in /etc/shadow. My question is - why? Argon2 is proven sturdy, well known, and there's many implementations of it. I could be wrong but I don't think this is the case for Yescrypt. I had never even heard of it until a few weeks ago. Or why not bcrypt? That's sturdy too.
- nullc 5y agoHere is their published rational: https://fedoraproject.org/wiki/Changes/yescrypt_as_default_hashing_method_for_shadow https://fedoraproject.org/wiki/Changes/yescrypt_as_default_h... It's not that shocking-- there is a lot less need for standardization of password hashing than protocols, and there were a lot of legitimate disagreements and ambiguous tradeoffs in the contest. (e.g. about things like timing sidechannels, the relevance of memory hierarchy, or support for other features like precomputed tables).