You are viewing a single comment's thread from:
RE: Quantum Resistant Error Correcting Code Based Blind Signatures Can Enhance Voter Privacy On Steemit
What's the drawback in terms of "bloat" - if there is any?
What's the drawback in terms of "bloat" - if there is any?
Cost benefit analysis, is the feature it can provide in terms of enhanced privacy worth the code complexity? I would say yes but someone else might say no. Considering private voting typically relies on blind signature schemes, and a lot of people will ask to have the ability to obfuscate their votes for political or commercial reasons, I can understand that at least in my opinion the benefit outweighs the cost.
For e-commerce I would say it's a critical feature. People wont want to rate a service if they think it could negatively effect their reputation or somehow cost them votes from whales.
I'm asking because in blockchain use, the extra bloat might be critical in terms of scaling.
I had asked on bitcointalk about QC-resistant algorithms for Bitcoin and the signatures (lamport, not a Niederreiter PKC scheme) where several kbytes which would make the network process a small fraction of its current transactions (which are currently 250-300 bytes on average). In other words it would hurt scaling so much that it could render Bitcoin useless.
@alexgr Most lattice based schemes are faster and lighter weight than what we use now.
https://en.wikipedia.org/wiki/NTRU#Performance
This is very promising, hmmm... hopefully patents don't limit its use in cryptocurrencies....
@alexgr The patents are a non-issue, they seem to be defensive only. They have released a GPL reference implementation and it's also part of bouncy castle crypto libs.
From my current knowledge of Graphene which I admit is limited, this particular chunk of code wouldn't have an effect on scaling. Graphene is easily able to scale from what I've seen and this part of the code would be on a layer above all that and could be just a smart contract or I forget what they call their plugin method but I do know they have some sort of operation for extending Steemit.
As far as size of transactions, I am not entirely sure on that and there would have to be some tests before I can tell you anything. The data would have to show if it will work but from the literature it's very efficient. We can't compare the architecture of Graphene to Bitcoin at all as they aren't the same species.
Because I'm not an official Graphene developer I am not intimate with the code base. I cannot tell you with certainty how best to do it. But I am familiar enough with the design (LMAX) and Bitshares, to believe that you can easily scale, but again the gold standard is to try it out on test net and analyze the results.
I'm quite unfamiliar with graphene too...so I don't know the specifics. In any case, Williams answer above indicate it's quite promising in terms of performance. Size remains a bit of a mystery, but we'll see when the time comes I guess...