timbeiko.eth
@tim
Is there a way to encrypt something on Ethereum and “force” decryption at a certain block? Idea is to sidestep not revealing the data. It’s fine if “as of block X, anyone can decrypt”, not just the original sender.
6 replies
0 recast
0 reaction
Cassie Heart
@cassie
Timelock cryptography. Not the timelock contract pattern that OpenZeppelin has but something like an RSA or IQC style timelock puzzle. It’d be just as decryptable off-chain as it would on-chain, but that would be the way to do it. You’d basically set the parameters such that it’s _roughly_ the right height, err h
2 replies
0 recast
1 reaction
tg
@tg
Ah, now we're talking about the dual - forcing the operation to NOT be done until a certain amount of elapsed time. FWIW, I'd consider that whatever time-related bounds proofs that timelock cryptography may provide - to be so fuzzy so as to be totally impractical for any practical applications I can think of.
2 replies
0 recast
0 reaction
Cassie Heart
@cassie
Not necessarily — the speed constraint of IQC is in speed of the fastest linear processing, necessarily by design it cannot be parallelized like Bitcoin miners working in unison to “find the winning hash”, and while there exists ASICs for the Wesolowski VDF it can’t get much faster
2 replies
0 recast
0 reaction
tg
@tg
Let’s say that proof has 3 basic states, (1) proved true (2) proved false (3) neither proved true or false. If there’s a type (1) proof of the claim that totally linear is the computation is the fastest possible way, then I’d love to see it. Bet it’s a type (3) pretending to be a type (1).
2 replies
0 recast
0 reaction
tg
@tg
But either way, it’s a slight of hand — whether linear is fastest and whether asics are faster — Both ultimately irrelevant to the real question, of if there exists more & faster hardware than that which you know about. History of e.g. govt tech, has shown that this is a very difficult assumption to bound.
0 reply
0 recast
0 reaction
Cassie Heart
@cassie
It’s not that linear computation is the fastest way to calculate the timelock, it’s that the construction is such that it is the _only_ way to calculate the timelock, as the cyclical calculations performed are contingent on the evaluation of the prior.
1 reply
0 recast
0 reaction
