Quote from: illodin on July 19, 2015, 09:55:33 PM

It wouldn't necessarily need a complete redesign to shut that attack vector. All that's needed is to prevent the mixing masternode from knowing which outputs (of the mixing transaction) belong together with which inputs. The mixing peers could do a key exchange and encrypt their own outputs with random mixing peer's public key, and decrypt the others' outputs they can using their private key, and then pass clear text outputs to the mixing masternode.

Or they could do all this without masternodes at all, as is being done with Bitcoin-based mixing systems, so again, Dash and masternodes are pointless. One of the original criticisms of Dash from Greg Maxwell (inventor of CoinJoin) is that the whole purpose of CoinJoin was that it can work directly on top of Bitcoin. Dash missed this point entirely, in pursuit of an instamine and speculative gains. If you are going to create a whole new system, there are better ways of doing mixing (cryptonote being one of them, which was first proposed by satoshi, CT from Blockstream being the first building block of another)

Quote

I wouldn't rule a complete redesign out either

Okay great. At that point, we can discuss whether or not that complete redesign is also snake oil.