[sounds like] the coordinator would see the history of all transactions, since the private keys are not blinded like ecash tokens. Of course you can use a new key for each coin, but you could still be identified by IP address.

"the coordinator would see the history of all transactions, since the private keys are not blinded like ecash tokens"

The coordinator does not see the transactions he is signing. You give him something called a "sighash," which means you hash the tx and have him sign that hash. Consequently, he signs the hash "blindly," i.e. without knowing the preimage, and thus he cannot see the history of the transactions.

"Of course you can use a new key for each coin"

I do

"but you could still be identified by IP address"

The operator does not see your ip address because you communicate with him using nostr dms, and you use a different nostr key for every message

Heidi says... so the operator can "rob you but not rug you"?

Super says... Rugging = robbing. The operator can help someone doublespend their money, which is a form of robbing/rugging one of the would-be recipients. You simply have to trust them not to rob you via that method, just like an ecash mint. The nice thing is, if the operator merely shuts down, no one loses their money -- which is an improvement over ecash imo

Jeroen says... How does that work? Did not watch the full video yet so if it's explained i'll find it out later.

Super says...

• users "refresh" their coins every time they receive money, just like with ecash
• during that refresh, the operator signs a timelocked withdrawal transaction for that user
• if the operator ever goes down, the user can broadcast that transaction when the timelock expires to recover their funds
• to stop "prior holders" from broadcasting the txs he signed for them, decrementing timelocks are used: each "new" holder has a smaller "wait time" than prior holders, ensuring the latest holder can withdraw first

E says... This is also significant because it can be argued that the operator is not really a custodian. If the operator disappears, the user has custody. It makes it less risks because if you have a catastopic hardware issue, no users will be terrible upset with you they will just be annoyed they have to fall back to onchain. Is this correct?

Super says... I agree with everything except this: if someone did argue that "the operator is not really a custodian," I would reply that you can never know this for sure. Each coin is held in a 2 of 2 multisig and the operator is "supposed to" only have 1 key, but you don't know the history of the coins you received. If the operator deposited those coins onto his own statechain, or if someone ever sent them to him, or if a prior holder of the coins colludes with him, then he could easily have both keys. But in theory -- if he's an honest operator and uses the software as I wrote it -- then he never has more than 1 key so he is never more than a "collaborative" custodian

This would have to be used for higher amounts correct? As you at some point you might want to consolidate those UTXOs to basechain. I would not think you can make payments for 5 dollar values then right?

Since the last holder of the coins has to create one or possibly two base layer transactions to withdraw them, I don't think it makes much sense to create statechain utxos with a value lower than the cost of two transaction fees. And let's suppose transaction fees are $5 apiece. You might think "ok then the minimum value of a statechain utxo should be $10." But that would mean half the utxo gets consumed as fees (assuming a cooperative withdrawal) or up to the entire thing (if he has to do a unilateral withdrawal). If you want to keep 99% of your money (or 98% in case of a unilateral withdrawal) then the minimum reasonable size of a statechain utxo is $500, that way you'll still have $495 in case of a cooperative withdrawal (99% of $500) or $490 in case of a unilateral withdrawal (98% of $500).

"I would not think you can make payments for 5 dollar values then right?"

Under these assumptions, that sounds like a reasonable conclusion. Of course, right now, transaction fees are only like 50 cents or less, so utxos as small as $50 make good economic sense for now.

StatechainJS: fast & free bitcoin transactions -- like ecash but a bit safer