5 ms·
How does the proof of storage work and will proof of time be decentralised across all nodes in the network, or a subset of trusted nodes?
by Cakez0r 9y ago
How does the proof of storage work and will proof of time be decentralised across all nodes in the network, or a subset of trusted nodes?
- marknadal 9y agoThe paper on storage proof is here: https://eprint.iacr.org/2017/893 https://eprint.iacr.org/2017/893 The proof of time is probably best described as being federated. Any peer can join, but it realistically will only be run by a very few peers with specialized hardware. This mistakenly trips up a lot of people as a "bad" thing, when it shouldn't. The only purpose of the proof-of-time servers is to make replay attacks impractical. As in, if the network is 10 years old, an attacker would actually have to spend 10 provable years to manipulate it. What if they collude with the time server? Well, they'd have to collude with all of them, which is unlikely, even if there are only a few. And second, it isn't worth the time server's work. Again, the details are a lot more complicated, but hopefully that is a quick off-the-top-of-my-head explanation that gives some intuition.