Recently I found that @NewtonProtocol spent a lot of space talking about Attestation, Verification, and Replay. At first, I didn’t really understand it. In my understanding, as long as the final outcome is correct, the way it’s achieved in the middle doesn’t seem that important. Who the executor is, what happens during execution—these feel more like implementation details rather than things the protocol truly cares about.
It wasn’t until later that I went back and re-traced the entire execution flow—from entering the Gateway from the Transaction Intent, to Policy Evaluation, then Operator execution, and finally the later Attestation—that I realized where my doubts had been coming from.
$NEWT seems to care not so much whether the result is correct, but why the result is worth trusting. The Transaction Intent doesn’t execute immediately just because it enters the system; it first goes through Policy Evaluation. And even after the Operator completes the task, it doesn’t automatically become the final result just because execution has ended. There still needs to be Attestation, and if necessary, Replay.
Keep reading— and the more I look, the more I find that they’re all answering the same question about this execution: whether it was carried out according to the rules jointly accepted by the whole network.
Only then did I realize that #Newt doesn’t truly record the outcome of a single execution, but the process of an execution. Later, I went over it again carefully and suddenly thought of a question I hadn’t taken seriously before.
Why do many systems focus more on proving the result, while Newton spends so much effort proving the process?
The more I think about it, the more it feels like the designs behind these two approaches actually represent two completely different ways of establishing trust. If you only prove the result, in the end you still need to trust the person who tells you that result. But if the entire execution process can be verified, then what truly needs to be trusted is no longer a particular Operator; instead, it’s the execution path that anyone can repeat and verify.
So I think Newton isn’t really trying to reconstruct the execution flow. What it truly challenges is a default assumption that has existed for many years: is the result correct enough?
At least, in Newton’s view, it doesn’t seem to be enough. Maybe this is precisely the real significance of Attestation, Verification, and Replay. They don’t protect just the result, but the entire process that makes the result hold.
It wasn’t until later that I went back and re-traced the entire execution flow—from entering the Gateway from the Transaction Intent, to Policy Evaluation, then Operator execution, and finally the later Attestation—that I realized where my doubts had been coming from.
$NEWT seems to care not so much whether the result is correct, but why the result is worth trusting. The Transaction Intent doesn’t execute immediately just because it enters the system; it first goes through Policy Evaluation. And even after the Operator completes the task, it doesn’t automatically become the final result just because execution has ended. There still needs to be Attestation, and if necessary, Replay.
Keep reading— and the more I look, the more I find that they’re all answering the same question about this execution: whether it was carried out according to the rules jointly accepted by the whole network.
Only then did I realize that #Newt doesn’t truly record the outcome of a single execution, but the process of an execution. Later, I went over it again carefully and suddenly thought of a question I hadn’t taken seriously before.
Why do many systems focus more on proving the result, while Newton spends so much effort proving the process?
The more I think about it, the more it feels like the designs behind these two approaches actually represent two completely different ways of establishing trust. If you only prove the result, in the end you still need to trust the person who tells you that result. But if the entire execution process can be verified, then what truly needs to be trusted is no longer a particular Operator; instead, it’s the execution path that anyone can repeat and verify.
So I think Newton isn’t really trying to reconstruct the execution flow. What it truly challenges is a default assumption that has existed for many years: is the result correct enough?
At least, in Newton’s view, it doesn’t seem to be enough. Maybe this is precisely the real significance of Attestation, Verification, and Replay. They don’t protect just the result, but the entire process that makes the result hold.