My trades were blocked by my own policies—Newton, who exactly do you want me to trust?
How’s the big pie? Can BTC still go up?
Last week, in VaultKit, I configured a strategy: if the collateral ratio falls below a threshold, liquidation is automatically triggered—such a perfect automation with a double insurance policy for the vault. But during testing, a replenishment trade that should have gone through was blocked. A policy evaluation failed popped up on the screen. I stared at it for a while before realizing it was my own Rego rules that trapped me.
It’s both ridiculous and hilarious, but it made me think for an entire night.
Newton’s logic sounds indeed rigorous: sensitive actions are packaged into intents, sent to a group of operators for evaluation, and only after generating an attestation are things allowed to proceed. Swap human approvals for code approvals—move trust from people to mathematics—at least that’s the theory. But I can’t help asking: that bit of discretion the curator has—did it get locked in a cage, or did it just upgrade to an even pricier material and keep squatting outside, still pointing and meddling?
Rego is something Goldman Sachs uses, and First Capital also uses. I admit it’s battle-tested—but guess what? The things those institutions use it to do are exactly what DeFi originally wanted to avoid. Now we’ve gone in circles and invited compliance brains back into the room. Funny, isn’t it?
Then there’s TEE. Newton uses it for boundary verification while keeping things private—sounds pretty great. Privacy has to be preserved, and verifiable execution can’t be lost. But as an old player who’s seen SGX crash and seen supposedly trusted hardware get broken through side channels, I just can’t convince myself to stake my assets and life on a chip vendor’s attestation. Do you call this cryptographically verifiable? Or is it just praying that Intel didn’t leave a backdoor for the NSA?
After digging through public materials, the operator signature aggregation does mention BLS, but how exactly to aggregate, how the quorum is set, how malicious nodes are punished—those details are either hidden in code comments in docs.rs, or simply not explained clearly at all. For a system that claims to be verifiable, refusing to lay out even the signature scheme—doesn’t that feel a bit interesting?
And on top of that: RedStone feeds prices, Credora feeds risk ratings, and Chainalysis handles sanctions and compliance screening. Before my trades can land on-chain, how many off-chain checkpoints do they have to pass? #Newt $NEWT @NewtonProtocol
How’s the big pie? Can BTC still go up?
Last week, in VaultKit, I configured a strategy: if the collateral ratio falls below a threshold, liquidation is automatically triggered—such a perfect automation with a double insurance policy for the vault. But during testing, a replenishment trade that should have gone through was blocked. A policy evaluation failed popped up on the screen. I stared at it for a while before realizing it was my own Rego rules that trapped me.
It’s both ridiculous and hilarious, but it made me think for an entire night.
Newton’s logic sounds indeed rigorous: sensitive actions are packaged into intents, sent to a group of operators for evaluation, and only after generating an attestation are things allowed to proceed. Swap human approvals for code approvals—move trust from people to mathematics—at least that’s the theory. But I can’t help asking: that bit of discretion the curator has—did it get locked in a cage, or did it just upgrade to an even pricier material and keep squatting outside, still pointing and meddling?
Rego is something Goldman Sachs uses, and First Capital also uses. I admit it’s battle-tested—but guess what? The things those institutions use it to do are exactly what DeFi originally wanted to avoid. Now we’ve gone in circles and invited compliance brains back into the room. Funny, isn’t it?
Then there’s TEE. Newton uses it for boundary verification while keeping things private—sounds pretty great. Privacy has to be preserved, and verifiable execution can’t be lost. But as an old player who’s seen SGX crash and seen supposedly trusted hardware get broken through side channels, I just can’t convince myself to stake my assets and life on a chip vendor’s attestation. Do you call this cryptographically verifiable? Or is it just praying that Intel didn’t leave a backdoor for the NSA?
After digging through public materials, the operator signature aggregation does mention BLS, but how exactly to aggregate, how the quorum is set, how malicious nodes are punished—those details are either hidden in code comments in docs.rs, or simply not explained clearly at all. For a system that claims to be verifiable, refusing to lay out even the signature scheme—doesn’t that feel a bit interesting?
And on top of that: RedStone feeds prices, Credora feeds risk ratings, and Chainalysis handles sanctions and compliance screening. Before my trades can land on-chain, how many off-chain checkpoints do they have to pass? #Newt $NEWT @NewtonProtocol