Published on: 2016-09-27T18:01:35+00:00
In a recent email exchange, Rusty Russell discussed the potential risks and benefits of out-of-band distribution of the pre-image. Two examples were mentioned - one involving B (role 2) to B (role 1), and another scenario assuming C receiving the pre-image out-of-band. It was concluded that there is no risk in this process and, in fact, out-of-band distribution of the pre-image is desirable. This is because if one intermediate node blocks the regular distribution, other nodes can still commit the transaction on their channels as soon as they receive the pre-image. This allows the node on the payee-side to enforce being paid by the HTLC mechanism, while the node on the payer-side doesn't mind not having to pay but can still pay voluntarily.On the other hand, if there is no out-of-band distribution of the pre-image, one blocking node can potentially keep all HTLCs on their payer-side locked for an extended period until their time-outs. Eventually, these transactions would be rolled back, resulting in the blocking node being the only one losing funds. The advantage of resolving HTLCs quickly through non-regular distribution of the pre-image incentivizes its use. For example, in Amiko Pay, payer->payee distribution is added alongside payee->payer distribution, but it's voluntary and can be removed without causing harm.Rusty Russell also highlighted a potential risk in a specific scenario where someone pays $4 in fees to C via B. In this case, B can use the H-preimage received from A to fulfill the HTLC A offered, gaining $4 and ignoring C. However, if C somehow obtains the pre-image out-of-band, it can claim the $5 from B and then B can retrieve its $1 from C. The risk for B is that C might wait until the C->B HTLC has expired and then use the B->C HTLC to collect $5, leaving B out-of-pocket. Rusty Russell argued that this game of risks and rewards applies to normal fees as well, especially since it's difficult to determine if apparently distinct nodes are actually identical.In addition to these discussions, Rusty Russell shared a new idea regarding atomically paying multiple parties using source routing and explicit fee payment. The proposal suggests designating all but one recipient as intermediate parties in the route and paying them a high transaction fee. To obfuscate the payment amount, the sender would be the final recipient. The author emphasizes that this approach is not complicated and has not been previously realized.
Updated on: 2023-07-31T19:12:40.977829+00:00