I watched the talk on "splicing" from this year's Bitcoin conference and was curious to know what other people's thoughts on it are. Will this open up the possibility to coinjoin without it being obvious what we're doing? I'm keen to coinjoin, but my current stack is KYC and I'm wary of bringing unnecessary attention to myself.
1501 sats \ 4 replies \ @nerd2ninja 14 Jun 2023
I put out a bounty for someone to explain this and they delivered: #190666
So yeah, its going to encourage transaction batching which I usually think of as just being good for saving on fees, but I believe it has a coinjoin like effect as well.
reply
53 sats \ 3 replies \ @citizen21 OP 14 Jun 2023
The write-up is a smidge beyond my understanding unfortunately, though it seems well-written.
Would this coinjoin effect be obvious to on-chain analysis firms? Would it raise any suspicion as to intent?
reply
0 sats \ 2 replies \ @nerd2ninja 14 Jun 2023
They might think you're consolidating utxos or batching lightning channel splices.
https://thebitcoinmanual.com/articles/open-multiple-lightning-channels/
reply
0 sats \ 1 reply \ @shibe 14 Jun 2023
I think the issue is that it's possible to track Lightning channels through gossip, because nodes share what UTXO is backing the channel. There are plans to change this alongside Taproot, so that Lightning channels are identical to Taproot single-sig spends, and figuring out a way to avoid sharing the on-chain UTXO with other nodes. But I'm not sure how the second thing will work. The reason you need to tell nodes that you actually have this liquidity to route payments is to prevent DoS attacks (imagine I tell everyone I have 0-fee channels with loads of liquidity, only for them to try to route payments through me and fail). It'd probably require something very complex like Zero Knowledge Proofs to do.
But barring both Taproot implementation and some way to avoid gossiping on-chain UTXOs, it should be trivial for chainalysis to track splices.
reply
73 sats \ 0 replies \ @shibe 15 Jun 2023
Important clarification I forgot to make before the edit timer ran out: Private channels don't get gossiped, so large batch splices of private channels might still be private. I would imagine even a single public UTXO getting in the mix could compromise the privacy of the entire transaction though.
reply