Statechains: a new Layer 2 protocol.

[Last of the V8s]

Statechains have come back to the realm of dicussion.
After Lightning, Liquid sidechain and Risk sidechain, a new Layer 2 protocol is proposed.

https://twitter.com/SomsenRuben/status/1135866719141076993
http://diyhpl.us/wiki/transcripts/bitcoin-core-dev-tech/2019-06-07-statechains/
many more links in this medium article:
https://medium.com/@RubenSomsen/statechains-non-custodial-off-chain-bitcoin-transfer-1ae4845a4a39

In short, Statechains are novel in the sense that they allow you to change UTXO ownership off-chain, while preserving a large degree of censorship resistance due to the ability to withdraw on-chain. It is non-custodial, which naturally reduces risk and makes it easier for the Statechain entity from a regulatory point of view. The end result is a unique layer two protocol with its own distinct strengths and weaknesses

https://youtu.be/ncpjRsHHUiA?t=5900 @shinobi and @nopara commentary
My summary of that, but please correct me where I am wrong.
-requires schnorr sigs and eltoo, also sighash anyprevout, which may take some time
-adds liquidity to lightning without opening channels from mainchain
-could interact with lightning to split up the locked utxo, and/or remove the federation
-could also use blinded sigs, adding a layer of privacy

I find it absolutely intriguing and full of possibilities for the current direction of bitcoin here.
Do you like the sound of it and are there any major downsides?

[1715171571]

1715171571

[1715171571]

1715171571

[1715171571]

1715171571

[1715171571]

1715171571

[1715171571]

1715171571

[pereira4]

See zack_bitcoin commentary on potential downsides (keywords: retirement attacks):

https://twitter.com/zack_bitcoin/status/1135934900605980672

This is the same capability as lighting + probabilistic payments. Except it has a lower level of security and is vulnerable to retirement attacks.

In any case competing layer 2 protocols are always good, the only problem I see is when they may require controversial updates like segwit.

[Last of the V8s]

This was also addressed in the Block Digest video, and here https://twitter.com/kcalvinalvinn/status/1135987591310798848

So "4.1 level security" = money can be stolen outright.

This is not the case with Statechains. If a previous possessor of a transitory key tries to withdraw on chain, then you're able to replace that transaction just like in lightning (eltoo).

The only way the entity can cheat is if they work with a previous possessor of an transitory key. Even so, they'll only be able to steal that particular UTXO. This becomes much harder with MuSig.

Obviously it's still a trusted setup but it's much more harder to cheat than Liquid style Federated Sidechains.

Gosh, don't they all require that segwit? All the changes required are very controversial and even laughable in some quarters, yes. But we're here, in this quarter.

[squatter]

This is an interesting proposal. I wish I had the technical expertise to judge the "retirement attack" vector. For example, what exactly do these statements imply? What is the risk in real terms?

The only way the entity can cheat is if they work with a previous possessor of an transitory key. Even so, they'll only be able to steal that particular UTXO. This becomes much harder with MuSig.

Yes, that is what 4.1 means. They can steal it at no cost, but we all know who stole it.

[Carlton Banks]

This is an interesting proposal. I wish I had the technical expertise to judge the "retirement attack" vector. For example, what exactly do these statements imply? What is the risk in real terms?

The only way the entity can cheat is if they work with a previous possessor of an transitory key. Even so, they'll only be able to steal that particular UTXO. This becomes much harder with MuSig.

Yes, that is what 4.1 means. They can steal it at no cost, but we all know who stole it.

in essence it means: trust required