Author: Rusty Russell 2018-10-16 22:30:15
Published on: 2018-10-16T22:30:15+00:00
Rusty Russell proposes a new protocol for side splice-in which treats splice-in and splice-out the same. The new protocol would require inline splice-in to have wait time, but since splice-out doesn't need this, it would not be necessary. The proposed protocol includes four types: `prepare any output with script of specific form`, `splice_in`, `update_splice_in_accept`, and `update_splice_in_reject`. The recipient of `splice_in` checks if the `blockheight` is far enough into the future and validates the claimed output before sending `update_splice_in_accept`. From that point onwards, all commitment transactions have one extra signature. However, Lisa found out that parallel splice does not work with Poon-Dryja channels as the counterparty can spend the old funding txout with a revoked spend. Rusty tried to come up with workarounds but nothing was successful, so he went back to Plan A.
Updated on: 2023-05-25T14:05:17.158997+00:00