Right this moment’s lightning channels use the ln-penalty replace mechanism. Every time an HTLC is staged or resolved, the channel house owners negotiate new dedication transactions for all sides. As a part of the negotiation, channel house owners should surrender the previous state: after buying the brand new state, every supplies the counterparty a breach treatment keyed to the prior state. This breach treatment permits a channel proprietor to publish a justice transaction to say all funds within the channel ought to the opposite occasion broadcast the outdated state. Ln-penalty requires that every occasion has a definite dedication transaction for every state for the reason that funds on the writer’s facet are locked for a while to offer the defender an opportunity to make use of the breach treatment.

Every occasion should preserve all prior breach cures and outdated backups change into poisonous waste. If a node unintentionally publishes an previous state, they forfeit all funds within the channel.

I don’t see how the penalty mechanism may work with extra events. Let’s assume we had a channel opened by Alice, Bob, and Mallory. Mallory opened the channel with 0.6 ₿. At a later stage, Alice has 0.4 ₿, Bob has 0.15 ₿, and Mallory solely holds 0.05 ₿. Mallory tries to unilaterally shut the channel with the preliminary state to reclaim all 0.6 ₿. The breach cures Alice and Bob maintain have been created when Mallory renounced the preliminary state. Does whoever broadcast it get all the cash? That might imply that Alice or Bob is punished for Mallory’s transgression. Is the cash break up in half? That method two channel house owners can steal from a 3rd each time one occasion’s channel capability exceeds half the channel steadiness. Maybe a justice transaction may pay out into the shared custody of Alice and Bob, however even then, how would Alice and Bob be certain that every get not less than the quantity that they might lay declare to within the final state?

I posit that designing an ln-penaly-like replace mechanism that ensures honest outcomes in all potential eventualities for 3 or extra events can be vastly extra difficult and doubtless have considerably extra roundtrips and overhead. I believe that it would require renegotiating all prior breach cures for every state replace. Blissful to be satisfied in any other case, if in case you have a nifty development in thoughts, although. 😉

The Decker-Russell-Osuntokun replace mechanism (“Eltoo”) manages to ratchet the channel state ahead with out asymmetry. As a substitute of getting a separate dedication transaction for every occasion, a symmetric replace transaction is created for every new channel state shared by all events. Replace transactions tie solely to the scriptPubKey and quantity of a previous transaction output as a substitute of a particular outpoint (which requires SIGHASH_ANYPREVOUT). Every Replace transaction spends the output of the Setup transaction or any Replace transaction older than itself. All Replace transactions share the identical enter quantity and pay the identical output quantity to the identical scriptPubKey. The pay out to the channel house owners occurs solely in a later Settlement transaction which solely turns into legitimate after the corresponding Replace transaction has matured for a number of blocks (through OP_CHECKSEQUENCEVERIFY). This provides different channel house owners time to submit any newer Replace transactions if an outdated state was printed.

^{through eltoo: A Easy Layer2 Protocol for Bitcoin}

Whereas there isn’t any punishment for the publication of an outdated state, the Replace transactions don’t pay any charges—a writer should add one other enter to supply charges (and presumably one other output to reclaim change). An unsuccessful attacker loses the transaction price of broadcasting the outdated state. Nonetheless, whereas an older Replace transaction is within the mempool and even after it has already been included in a block, the opposite channel house owners can broadcast newer Replace transactions to set off a extra useful end result for themselves (or skip proper to the latest Replace transaction).

For the reason that channel state is symmetric and there’s no state particular breach treatment, every (sincere) channel proprietor wants solely the most recent state which reduces the quantity of state channel house owners have to preserve monitor off. Outdated channel backups are not poisonous waste that instantly set off lack of all funds, however as a substitute the opposite channel proprietor can simply ratchet ahead to a later state (which can nonetheless lose the negligent channel proprietor some or all funds, or will be the newest state within the case of an sincere counterparty). For the reason that channels are symmetric, the overhead of getting multiparty channels is linear. It could nonetheless be impractical to create channels with huge numbers of channel house owners, since every channel proprietor should signal each Replace transaction, however not less than just a few channel house owners must be very nicely potential. To a level the overhead for updating the channel state may be mitigated through Channel Factories which permit subsets of the channel house owners to have segregated balances they’ll replace amongst themselves.