2/ This analysis only focuses on the superficial role of randomness in Nakamoto consensus: solving the leader-election problem. But solving the leader-election problem is only half the battle.
-
-
Show this thread
-
3/ There’s a much, much bigger role randomness plays in Nakamoto consensus beyond leader election: "unforgeable costliness". h/t
@NickSzabo4 again for the term.Show this thread -
4/ Randomness is what allows PoW mechanism to approximate energy burnt in the creation of blocks. This burnt energy secures the chain from past AND future attacks. I explored this role randomness plays in Bitcoin in my recent article:https://medium.com/@hugonguyen/bitcoin-chance-and-randomness-ba49a6edf933 …
Show this thread -
5/ Bitcoin is past-proof because it would cost an attacker an enormous amount of energy & money to forge an alternative chain with a higher cumulative difficulty. Thanks to randomness-based PoW.
Show this thread -
6/ Bitcoin is future-proof because it would provide an automated way for network nodes to come to consensus in the case of a network split, whether it's intentional or accidental (e.g., a network partition). Thanks again to randomness-based PoW.
Show this thread -
7/ Naive randomness mechanisms in PoS protocols do not give you either of these protections. PoS inevitably requires the intervention of trusted 3rd parties and significantly increases the social attack surface. In other words, PoS is nowhere near as trust-minimized as PoW.
Show this thread -
8/ In summary, randomness in PoW solves *two* crucial problems: a/ Leader-election b/ Unforgeable costliness Leader-election is only half the battle, arguably the less important half.
Show this thread -
9/ P.S. For a list of my writings on the problems with PoS:https://twitter.com/hugohanoi/status/1021831408489136128 …
Show this thread
End of conversation
New conversation -
-
-
Again, the focus of this post was on only randomness for leader election. We have a whole paper that does a more in othe analysis of all the parts that you talk about :)https://github.com/Mechanism-Labs/MetaAnalysis-of-Alternative-Consensus-Protocols …
-
Ah I see, did not know there was another part
I'm generally trigger-happy because I'm frustrated with all the PoS discussions of randomness that say nothing about unforgeable costliness
Sorry for jumping the gun then. Look forward to part 2! -
>None of the subcommittee selection protocols that we at Mechanism Labs have analyzed satisfied both properties of being unbiasable FWIW, I highly agree with this- hard to guarantee both liveness & safety RE:leader-election randomness. Not sure I'd agree with the compromise tho.
-
I think that's where the dimension of synchrony comes in. I definitely have not even touched upon it in this post because that's a much longer analysis.
-
Favoring liveness over safety or relaxing synchrony assumptions won't make PoS protocols any safer. PoS camp likes to say that Bitcoin also doesn't work under a fully async setting, but they overlook important nuances in the level of security offered in the semi-sync setting.
-
Simply put, under semi-sync setting, PoW strongly satisfies liveness & consistency. Whereas PoS protocols do not & break down under a number of scenarios. See my 2 articles for a few examples of these scenarios & the addition by
@Datavetarenhttps://twitter.com/hugohanoi/status/1021832568314843136 …
End of conversation
New conversation -
-
-
@el33th4xor What do you think about this thread by@hugohanoi?Thanks. Twitter will use this to make your timeline better. UndoUndo
-
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.

