4 ms·

I'm fairly blockchain-ignorant despite having read up on it quite a bit, but THIS sort of task seems to me a pretty good use of blockchain. It wouldn't need to

I'm fairly blockchain-ignorant despite having read up on it quite a bit, but THIS sort of task seems to me a pretty good use of blockchain. It wouldn't need to necessarily run in an entirely de-centralised set up, but how that would work (eg. sharing amongst a limited number of trustable centralised parties), I don't know. I did mull over something similar for voting a few years back, scrawling sides of A2-paper with notes and diagrams on the subject, but I'll be damned if I can make head or tail of whatever I wrote back then. I was a bit drunk at the time.

simonw 3y ago
Right - you don't need a wasteful proof-of-work blockchain for this. The mechanism used for Certificate Transparency would work great here: https://en.wikipedia.org/wiki/Certificate_Transparency https://en.wikipedia.org/wiki/Certificate_Transparency Same for "real estate on the blockchain" - when people argue for that, the bit they are usually arguing for is a trail of historical records that can't be corrupted. Certificate Transparency solves that really well.
- junofan 3y ago
  How? It’s not designed for consensus.
  - simonw 3y ago
    You don't need consensus for this, unless I've missed something. We are talking about a mechanism to prove that no-one has tampered with the documents.
pdpi 3y ago
This is one of those cases where terminology matters. A blockchain is just a content-addressed linked list (that is: the "pointer" to the next node is actually hash of the contents of the next node, then you have a big hashmap of hash -> node contents). A merkle tree is the same thing in tree form. Git uses the directed acyclic graph version of the same thing. This approach is perfect for this use case. The decentralized Nakamoto consensus proof of work stuff is unnecessary here.