Author: Rusty Russell 2019-01-08 05:50:20
Published on: 2019-01-08T05:50:20+00:00
Matt Corallo, a Bitcoin Core developer, raised concerns about defining a "near the top of the mempool" criteria for Lightning's requirements. While it is suitable for large batched transactions where a single counterparty is not allowed to prevent confirmation, Lightning's requirements are different. It requires certainty that the transaction will confirm by a specific deadline, rather than a high probability of confirmation soon. Rusty, another developer, suggested defining the "top of mempool" as "in the first 4 MSipa," which means the next block, and allowing RBF only if the old package was not in the top and the replacement would be. Rusty believes this approach is more incentive-compatible than the current scheme. However, an attacker could make users pay next-block high fees, but it is still better than the current method of overpaying and hoping. Rusty suggests that the simplest implementation should win.
Updated on: 2023-06-13T15:49:24.552386+00:00