sipa changed the topic of #bitcoin-wizards to: This channel is for discussing theoretical ideas with regard to cryptocurrencies, not about short-term Bitcoin development | http://bitcoin.ninja/ | This channel is logged. | For logs and more information, visit http://bitcoin.ninja
Ylbam has quit [Quit: Connection closed for inactivity]
PaulCapestany has quit [Ping timeout: 252 seconds]
in zurich, i had a conversation over dinner with instagibbs and maaku about multi-party payment channels in lightning
instagibbs had mentioned that someone (poon?) had sent to a mailing list a writeup about why multi-party doesn't work for lightning payment channels
my contention was that you could probably make a system where if one party cheats that you could check which signature was used for cheating and then boot them out of the multisig pool or something
.... sorry for all the non-details :).
haha, no worries
I'm very excited about this, though. Sharing the 'posted bonds' in lightning among multiple channels seems big to me in terms of the practicality of the system.
(or among more than two parties to a single channel.)
AaronvanW has quit [Remote host closed the connection]
kanzure: yes, but our back of napkin result was that it would take k*O((n-1)!) signatures per channel update, where n is the number of parties and k the number of signatures normally required for a bidirectional channel
so it might make sense for n=3, but I'm doubtful for other reasons. n=4 and above is probably out of the question
actually k*O(n! - 1)
ThomasV has joined #bitcoin-wizards
maaku: what about the scenario where updates only have to be signed by (1) participants who lose value in the update and (2) the distinguished 'hub' node?
Do you think that could improve on that?
gwillen: maybe.. i might have to whiteboard that
but yes that would change it considerably
gwillen: i'm not sure it's a simple thing to do that at least with the primritives bitcoin script has available, but someone should try
kanzure, I do not doubt that an ok solution exists, but the only one I think works without "big" softforks don't scale well
jl2012 and I discussed this a few days ago in scrollback
atgreen has quit [Ping timeout: 272 seconds]
notices email, reads
I multiway payment channel with the locktime approach is more efficient than LN approach
jl2012, I did more thinking and don't get how csv would work, but it's possible I'm an idiot. The initial funding transaction gets published... can't csv that