Published on: 2020-04-17T03:07:37+00:00
In an email thread, Olaoluwa Osuntokun addresses concerns regarding the increased cost of the new commitments. He explains that although the default commitment weight is slightly higher, both parties have the option to recover the additional anchor cost of 660 satoshis if they choose to do so. Osuntokun also points out that force cases in the ideal scenario are only necessary when nodes need to go on-chain to sweep HTLCs, meaning that the extra bytes may be overshadowed by several HTLCs.Osuntokun goes on to discuss the issue of UTXO bloat and suggests using two anchors with special scripts. This approach allows anyone to sweep these outputs several blocks after the commitment transaction has been confirmed, which is more advantageous than repurposing the to_remote output as an anchor. Repurposing the to_remote output only permits a single party in the channel to spend that output and remove its impact on the UTXO set. Regarding the reduction of CSV delay to the to_remote party, Osuntokun proposes leaving this matter open for a comprehensive solution since the dual anchor format is completely independent of the rest of the commitment. Despite the slight increase in cost for the new commitments, Osuntokun believes that it is not worth arguing about, considering the overall benefits. He also emphasizes that reducing the difference in cost should not be the primary focus, as the majority of the effort lies in the upgrade itself.The implementation of anchor outputs is now fully supported in the master branch of lnd, and users can opt into this new format by specifying the --protocol.anchors command line parameter (off by default). During the last spec meeting, concerns were raised about having two anchors at all times, as this adds extra bytes to the commitment transaction and increases the fee burden for force closing while also polluting the UTXO set. However, Osuntokun clarifies that although the new commitments are slightly more expensive, they offer significant advantages.In terms of UTXO bloat, repurposing the to_remote output as an anchor is considered worse because it restricts only one party in the channel from spending that output to remove its impact on the UTXO set. On the other hand, using two anchors allows anyone to sweep these outputs after the commitment transaction has been confirmed. Another factor to consider is whether to reintroduce a CSV delay to the to_remote party to address game theoretical issues related to forcing one party to close early without incurring any cost. Looking ahead, Osuntokun mentions that in the future, users will have the ability to easily update their commitment type as they are concurrently working on a dynamic commitment update protocol. This means that users will no longer be locked into a commitment type when the channel opens but will be able to update it on the fly.
Updated on: 2023-07-31T22:45:35.140218+00:00