# Awesome bit-by-bit review of Solana's whitepaper and several blog posts - by Viktor Shoup.

Link - https://www.shoup.net/papers/poh.pdf

## TLDR:
* Solana is just a PERMISSIONED append-only ledger based on PBFT consensus that is run on a top-end hardware only
* There is no evidence that Proof-of-History really helps to achieve consensus or improves its performance.

## Cool quotes:
* "We therefore conclude that, in terms of consensus protocols, there simply is no “clock problem” that needs to be solved — in our opinion, this is yet another red herring"
* "Indeed, it is not clear to us that the communication complexity of Solana’s consensus protocol is any better than that of any of the existing consensus protocols in the literature"
* "While it is well known that Tendermint is not optimistically responsive, other protocols are, including classical PBFT [CL99], HotStuff [YMR+18], SBFT [GAG+19], and ICC [CDH+21]. These optimistically responsive protocols solve perfectly well the problem of avoiding unnecessary delays without the need for performing all of the intensive computations required for proof of history"
* "In any case, this is not germane to our question of how using proof of history as a trusted source of time helps in improving the performance of proof-of-stake consensus protocols"
* "We observed that one interesting and perhaps novel idea in Solana’s consensus protocol is that of using exponentially increasing lockout periods, during which time a node cannot vote on a different branch. This idea seems worthwhile to explore; however, its motivation, benefits, and security properties remain unclear"

p.s.
Watch "The Chopping Block" with Anatoly Yakovenko [here](https://www.youtube.com/watch?v=0KGm0FJ5LWg).