ZK-gossip for lightning channel announcements

[halseth]

@AdamISZ

Here’s the part that might be interesting to you (it was to me!): I didn’t even consider the proof-creation-by-channel-parties-together idea. My thinking was (a) taproot + musig2 making channel utxos the same as other utxos and (b) 2-party construction of proof is a nightmare, so just make proofs over other utxos that happen not to be channels. In an ideal world it doesn’t make a difference, but I neglected to consider: utxos that are not channels are extra liquidity cost for actual Lightning operators!

So while that’s not a slam dunk argument for “supporting musig 2-party outputs is required”, it’s a pretty good argument, and if the only thing sacrificed is that slightly flaky version of a key image, it … may be OK?

I must confess I’m not understanding what you are saying here Could you explain?