lightning-dev
LN without SegWit: less efficient or less secure?
Posted on: January 16, 2017 22:12 UTC
In an email conversation, Andrés G.
Aragoneses quoted a paragraph from the last version of the Lightning paper, which talked about a stop-gap solution to be described in another paper by Rusty Russell. The three levels from the Scaling Bitcoin Hong Kong talk were discussed, but it was mentioned that they are not tightly defined or analyzed and should not be taken too seriously. Level 1 uses CLTV only, while level 2 uses CLTV and CSV, with second and third party malleability still possible. Level 3(a) adds segwit, making malleability irrelevant, but outsourcing of channel monitoring is difficult. Level 3(b) adds SIGHASH_NOINPUT, making it easier to get someone else to monitor the blockchain for you. The lightning paper itself was designed around "level 0" when there wasn't even an OP_CLTV. Andrés suggests that everyone has been focused on 3(a), but variations of level 2 may exist.