BTW, I deliberately left a very small off-by-one error in my calculations. See if you can find it!
https://petertodd.org/2023/v3-transactions-review
My last post of the year: V3 Transactions review.
tl;dr: Replace -By-Fee is half the cost, and ephemeral anchor outputs pose a significant mining decentralization risk.
Happy New Year! 🎊
Discussion
May be that I miss something of LN protocol so apologies for that .. Let's assume that due to a pendind htlc a channel got closed (non collaborative), who close that? The peer nodes who has the channel on it? Who pays for the closure ?on this kind of multisig transactions could both members apply for cpfp ? And what about rbf? could both apply for that? ... I'm wondering if the one who need to pay more for the non collaborative closure it is always the child one :)
One of the design decisions in RBF is deciding who pays for commitment transactions. It's quite feasible to have two different sets, taking the tx fees from different sides depending on who decides to broadcast.