3 ms·

Did anyone solve the "no stake" problem in "proof of stake"?

• DennisP 9y ago
  The solution in Casper: if you approve more than one block at the same height, someone can submit both your approvals to the winning chain, which destroys your stake on that chain. The submitter gets a reward. There's a long delay to withdraw your stake so plenty of time for someone to submit.
  • xorcist 9y ago
    What if the alternate chain is kept secret until it is sure to be victorious? In a situation where this type of conspiracy takes place it makes economic sense to join them.
    • DennisP 9y ago
      It's fine for your favored chain to win. And maybe you can get a group of validators to cooperate and make that happen. But they better have bet on only that chain, or they will lose their stake. If the validators are conspiring on a secret chain, and they're betting only on the secret blocks, the public blocks won't be finalized.
• ralexstokes 9y ago
  You generally get around the "nothing at stake" problem by requiring validators (consensus-forming nodes) to submit a deposit that is subject to deletion upon proof of bad behavior. For example, if I get two conflicting messages signed by you in Ethereum's Casper network, I can present that data as proof to the Casper smart contract where you lose your deposit.