5 ms·

>If a POW miner invested the resources to take over the network, breaking the network would certainly crash the price. In practice, us government would pass a

>If a POW miner invested the resources to take over the network, breaking the network would certainly crash the price. In practice, us government would pass a regulation that forces miners to register and to censor sanctioned addresses. Once they realize that miners they control own majority of hashrate, they can add a requirement to ignore chains with sanctioned addresses (from others miners). There's no defense against this, because unlike PoS, you can't really move anywhere else. China has its own ban. Energy prices in europe are insane. Russia is sanctioned and just getting chips in is almost certainly a crime. Under PoS as long as there's any subset of non-censoring validators a fork away from censorship is possible. Unlike mining, staking is extremely mobile globally.

aserafini 4y ago
And in this last scenario, how does someone joining the network know which of the now forked chains is the ‘one true chain’?
- idiotsecant 4y ago
  How do you know literally any chain is the one true chain? You don't, because there is no such thing. There is only the chain which you agree is the one true chain. In practice the software generally is designed to connect to only one chain - a chain fork requires you to install different software. This isn't unique to any PoS or PoW project - it's a feature of literally all decentralized consensus mechanisms. Otherwise they wouldn't be decentralized consensus.
  - aserafini 4y ago
    You know which PoW chain is the one true chain because it is the one with the most work. What is the equivalent mechanism in PoS?
    - idiotsecant 4y ago
      That's not at all true - chains pop up all the time with less work and are relatively successful for a time. Bitcoin cash had substantially less work in the network and was (and maybe still is, I don't know) quite popular. Ethereum classic has much less PoW power and it was quite popular for a bit. If people are using it isn't a 'fake' chain in any sense - just a less popular one. That's sort of how distributed consensus works. What is the 'true' hamburger chain? What is the 'true' flavor of ice cream? If less people like bacon flavored ice cream or vanilla flavored hamburgers it doesn't make them not 'true' representation of the archetype, just unpopular ones.
      - aserafini 4y ago
        The most "popular" chain in PoW is the one that has the most work. How should someone determine the most "popular" chain in PoS based on the chain history alone? Is it the one with the most 'staked'? The problem (which PoW doesn't have) is that alternative versions of valid PoS history can be generated at zero or close to zero cost.
        idiotsecant 4y ago
        >How should someone determine the most "popular" chain in PoS based on the chain history alone? Is it the one with the most 'staked'? Yes. >The problem (which PoW doesn't have) is that alternative versions of valid PoS history can be generated at zero or close to zero cost. This is simply a misunderstanding of how it works. In order for you to propose a version of the chain that has some other history you need network consensus. If you claim something that is not true you get slashed. Thats the whole point. You can go make an entirely disconnected chain if you'd like to do that but you'll be the only one on it unless you can convince someone else to join you in your personal universe where you are a trillionare.
        aserafini 4y ago
        You can generate valid alternative Ethereum PoS histories in private without getting slashed: just don’t propose them to network? This is unlike PoW, where even generating the history is expensive. This leaves a problem: how should new nodes joining the network know if the version of history they’ve been presented with is ‘honest’. The answer for Ethereum PoS is ‘checkpoints‘, but it’s not an ‘objective‘ measure, unlike ‘chain with most work’ in PoW. Even Vitalik admits this, it is why he invented the term ‘weak subjectivity’, he is totally explicit that this is a fundamental weakness of PoS: the valid chain is subjective. The only question that remains is whether the economic incentives that have been designed are good enough to work around this weakness and guarantee whatever properties Ethereum network users have decided is important.

aserafini 4y ago

And in this last scenario, how does someone joining the network know which of the now forked chains is the ‘one true chain’?

idiotsecant 4y ago
How do you know literally any chain is the one true chain? You don't, because there is no such thing. There is only the chain which you agree is the one true chain. In practice the software generally is designed to connect to only one chain - a chain fork requires you to install different software. This isn't unique to any PoS or PoW project - it's a feature of literally all decentralized consensus mechanisms. Otherwise they wouldn't be decentralized consensus.
- aserafini 4y ago
  You know which PoW chain is the one true chain because it is the one with the most work. What is the equivalent mechanism in PoS?
  - idiotsecant 4y ago
    That's not at all true - chains pop up all the time with less work and are relatively successful for a time. Bitcoin cash had substantially less work in the network and was (and maybe still is, I don't know) quite popular. Ethereum classic has much less PoW power and it was quite popular for a bit. If people are using it isn't a 'fake' chain in any sense - just a less popular one. That's sort of how distributed consensus works. What is the 'true' hamburger chain? What is the 'true' flavor of ice cream? If less people like bacon flavored ice cream or vanilla flavored hamburgers it doesn't make them not 'true' representation of the archetype, just unpopular ones.
    - aserafini 4y ago
      The most "popular" chain in PoW is the one that has the most work. How should someone determine the most "popular" chain in PoS based on the chain history alone? Is it the one with the most 'staked'? The problem (which PoW doesn't have) is that alternative versions of valid PoS history can be generated at zero or close to zero cost.
      - idiotsecant 4y ago
        >How should someone determine the most "popular" chain in PoS based on the chain history alone? Is it the one with the most 'staked'? Yes. >The problem (which PoW doesn't have) is that alternative versions of valid PoS history can be generated at zero or close to zero cost. This is simply a misunderstanding of how it works. In order for you to propose a version of the chain that has some other history you need network consensus. If you claim something that is not true you get slashed. Thats the whole point. You can go make an entirely disconnected chain if you'd like to do that but you'll be the only one on it unless you can convince someone else to join you in your personal universe where you are a trillionare.
        aserafini 4y ago
        You can generate valid alternative Ethereum PoS histories in private without getting slashed: just don’t propose them to network? This is unlike PoW, where even generating the history is expensive. This leaves a problem: how should new nodes joining the network know if the version of history they’ve been presented with is ‘honest’. The answer for Ethereum PoS is ‘checkpoints‘, but it’s not an ‘objective‘ measure, unlike ‘chain with most work’ in PoW. Even Vitalik admits this, it is why he invented the term ‘weak subjectivity’, he is totally explicit that this is a fundamental weakness of PoS: the valid chain is subjective. The only question that remains is whether the economic incentives that have been designed are good enough to work around this weakness and guarantee whatever properties Ethereum network users have decided is important.