Are there any plans to replace Blake2b with Blake 3 for hashing in the future?
According to Blake 3 Github it is approximatly 5 times faster than blake2b
I guess it's not an easy task to do because we will probably need a transistion period where both blake2b and blake 3 are valid hashes while all nodes are upgraded.
I'm also unsure what the real world performance improvement will be.
Blake 3 Github
This is very interesting. I guess that whether it gets implemented soon depends on how hard is it to add it to the protocol and what percentage of the processing in a transaction is consumed by hashing operations. Clearly if it's a big part, then using something like this might be a huge improvement.
Some discussion from Discord:
Thanks for the reply. No need for Blake 3 then