Published on: 2021-06-20T00:30:09+00:00
The email discusses the proposal of using Fast Forwards in the Lightning Network to reduce the requirement of keeping privkeys online. The current protocol requires 1.5 round trips before a node can safely forward a payment, causing latency issues. The Poon-Dryja commitment transaction improves this by having each commitment transaction with two outputs: "to-local" and "to-remote". The author proposes using fast forwards to indicate that an HTLC is "irrevocably committed" with a single message. However, fast forwards have higher risk and could lead to fast failures. Nodes offering low-latency fast forwards may charge higher off-chain fees to offset the risk. To validate the truth of fast forwards, nodes could continuously make known-failing payments in the background.The Decker-Russell-Osuntokun construction does not require fast forwards if the link-level protocol is designed properly. Each update message includes the necessary signature and new state transaction, making it safe to forward the "add HTLC" immediately. However, the Lightning Network still faces payment latency issues due to commitment transaction synchronization.There is a proposal for the "fast forward" technique, which reuses the revocation path to propagate failed payments. This is possible because the HTLC construction used by Lightning is revocable. However, this increases the risk of unilateral close after fast-forward or fast-fail, requiring additional transactions and fees. Conflicting 'next update' messages can be resolved by coordinating with the counterparty.Overall, optimizing the Lightning Network system is necessary to improve payment latency.
Updated on: 2023-07-31T21:36:32.895299+00:00