[Cryptography] More efficient block-chain ledger: Micali's "Algorand" ?

[James A. Donald]

Algorand sounds like an enormous improvement, and I think it is an 
enormous improvement, since proof of stake is better incentive aligned 
than proof of computational power.

But it still has that pesky bandwidth problem.  The authors cheerfully 
say that the computational burden is minimal, and it is, but every user 
downloading every transaction of every other user is not minimal.

It would be a good system for proving public keys - you would know that 
the person you are communicating with sees the same mapping between 
public keys and phone numbers as you do, or the same mapping between 
blog addresses and public keys, or the same mapping between email 
addresses and public keys.

But for money purposes, it still hits the problem that we would like to 
handle a *lot* of transactions.