Author: ZmnSCPxj 2020-06-22 08:15:37
Published on: 2020-06-22T08:15:37+00:00
ZmnSCPxj responds to a previous message discussing the use of independent pay-to-preimage transactions as a solution to a specific attack. While acknowledging its technical feasibility, ZmnSCPxj notes that this approach may still have incentives issues. In this type of attack, miners who hide the preimage transaction in their mempool must be accomplices with the attacker. Otherwise, they would share that transaction with some of their peers, and non-miner nodes would get that transaction and be able to gossip them off-chain (and even relay them to other mempools).However, ZmnSCPxj proposes a possible solution that is technically possible with current mempool rules without requiring the cooperation of miners with the attacker. The attacker releases two transactions with nearly equal fees, one near miners, and the other near non-miners. Nodes at the boundaries between those that receive the preimage transaction and the timelock transaction will receive both but will keep and reject one based on which they receive first.This boundary is inviolate, neither can get past that boundary, even if everyone is running 100% unmodified Bitcoin Core code. While not being a mempool expert, ZmnSCPxj believes that this approach could be a potential solution to the attack.
Updated on: 2023-06-14T00:34:11.396773+00:00