3 ms·
>attacker must acquire a large amount of ETH The issue with proof-of-stake: How do you define consensus on what ETH is? Without PoW, an attacker can cheaply g
by aeternum 7y ago
>attacker must acquire a large amount of ETH
The issue with proof-of-stake: How do you define consensus on what ETH is? Without PoW, an attacker can cheaply generate an alternate history in which they control a large amount of ETH. If the alternate history is accepted, they benefit, and if not, they risked nothing.
- MadWombat 7y agoI would expect that this is because the rest of the network needs to agree with your version of history. And they will only accept your vote if you stake the currency they already recognize, not the currency you made up in your version. So to make the network accept a history version where you have a lot of money you need a lot of money.
- aeternum 7y agoA peer has no way of knowing which 'network' is real. Since there is no proof of work, it is effectively free to generate any number of alternate histories or blockchains (this is known as 'costless simulation'). It becomes trivial to generate the longest chain. This is a real problem, and some PoW currencies like NXT have proposed solving it by using an out-of-band solution like asking friends or trusted nodes whether you are on the right root blockchain. This defeats the very purpose of a decentralized currency. If you're going to trust a core set of nodes, just use a centralized database with M-of-N access control.