Author: Matt Corallo 2020-06-24 08:32:52
Published on: 2020-06-24T08:32:52+00:00
The discussion on the Bitcoin-dev mailing list is about the possibility of a miner hiding a transaction from certain parts of the network. One participant agrees that it is possible in theory, but difficult to achieve in practice. They argue that an attacker who could do this would break any off-chain construction that relies on absolute timeouts and hope that it is difficult to achieve without cooperation from a subset of miners. Another participant suggests using independent pay-to-preimage transactions as a technical solution, but acknowledges that there are incentives issues. They propose releasing two transactions with near-equal fees, one near miners and the other near non-miners, so that nodes at the boundary will receive one or the other first and reject the other. This creates an inviolate boundary between the two transactions that neither can get past. Matt observes that the transaction relay delays and network topology make this attack very real and practical with a double-digit success rate, and suggests that it may be worth fixing in the medium term. He proposes saving the blind-CPFP approach with eltoo and a magic transaction relay metadata that allows specifying which output a transaction spends, but acknowledges that this may be complex.
Updated on: 2023-05-23T03:13:31.666660+00:00