Hacker News Reader

Comment by Rhapso | Hacker News Reader

Honestly, I have yet to see a "meta-coin" that seemed viable. While ideas like creating virtual stocks and property ownership in a block chain are cool they all hinge on actually trusting the person backing those currencies/coins. The meta-programming layer here is nice it gives better transaction control, it does not let us do anything new.

View on HN

Well a cryptocurrency which actually allowed you to pay for distributed computing, that did useful computations - would be a very interesting thing.

"Miners" would have to buy generic hardware that did generic calculations, which actually would have "intrinsic" value since they wouldn't be busy-work to keep things secure, they'd be executing useful computation for a client.

This is difficult, because you need to check the computation. (The whole point of distributed cryptocurrency is that you do not trust your peers.) Only those computations that are hard to compute but easy to check can be used by cryptocurrencies, not any general useful computations.

The only one I know of that fits this criteria is Primecoin, its proof of work function is to mine prime numbers.

I haven't heard of any others, but I'd be curious to know about them.

> Well a cryptocurrency which actually allowed you to pay for distributed computing, that did useful computations

This is a bit like saying, "I wish the computations in TLS/SSL did something useful besides securing the connection". The security IS useful!

Interesting that you pick that, it would be more meaningful if TLS/SSL was not vulnerable to man in the middle and certificate spoofing (respectively). The security is useful, but is it sufficient? Sure searching for hash collisions provides a use, but would it not be interesting to devise a way to provide proof of work for a problem that had more useful results. A good example would be solving satisfiability problems posed in previous blocks, it is easy to calculate the difficulty of the problem and the solving solution is not much better then trying random solutions and hoping and the results is instantly verifiable. It would let people encode meaningful problems to be used as proof of work.

Reply on news.ycombinator.com