3 ms·

Automatizing rewards for solving hard problems with Bitcoin

• kolinko 12y ago
  As for "M of N oracles running a public code" mentioned in the article - Orisi.org offers a framework for that. As for using blockchain directly for contract - there are a few addresses that have funds locked for the first person to find collisions in sha1, sha256 and others: https://bitcointalk.org/index.php?topic=293382.0 https://bitcointalk.org/index.php?topic=293382.0 Finally, there's (or "there will be") Ethereum which seems best for preparing contracts that require no external inputs. Solving hard problems usually requires no such inputs.
  • infruset 12y ago
    Thank you very much for the references! As for Ethereum, I know that it is the goal, certainly the fact that it is Turing-complete argues in favor of this usecase. However, how practical and efficient is it, as in what kind of slowdown in the execution would you get on the blockchain?
    • rwallace 12y ago
      Even if it's orders of magnitude slower, that should be okay, because you're not going to be using it for the hard job of finding a solution, only for the easy job of verifying the solution once found. (I don't know much about Ethereum, so I don't know whether it has any arbitrary limits that would be a problem, but it should be okay.)
      • infruset 12y ago
        Well, suppose what you want to check is a proof of the Riemann conjecture in Coq, this might take some substantial time. So whether it is fast could be relevant.