> When I see a hash that satisfies the difficulty, I know it must have taken time. The method by which the delay is accomplished is “work”, but the hash is a proof of time, not work.
I disagree with this characterization.
Mostly because there actually is a more proper concept of Proof-of-Time, that doesn't depend on Bitcoin's dynamic difficulty adjustment to space puzzle solutions a certain time apart on average.
It is what Bram Cohen is trying to use in conjunction with Proof-of-Space to make [1] a greener alternative to Bitcoin.
One candidate problem for proving time is modular square roots. In a group of order 2^k-1, we can square element x (k-1) times to obtain x^(2^(k-1)), which is the square root of x^(2^k)=x. Since the k squarings cannot be parallellized, this takes time proportional to k.
>> If the stated problem is to find a conforming hash, all you have to do is to try it once, and bingo, you’ve affected the global hash rate, and for that one attempt you were a participant helping others solve the problem. You did not need to tell others that you did it (unless you actually found a solution), others didn’t need to know about it, but your attempt did affect the outcome. For the whole universe, no less.
>> It is noteworthy that since SHA is progress-free, each attempt could be thought of as a participant joining the effort and immediately leaving. Thus miners join and leave, trillions of times per second.
Is the value of pooled mining that participants in the pool share the nounce they’re calculating, to avoid duplicating work within the pool?
>> Is the value of pooled mining that participants in the pool share the nounce they’re calculating, to avoid duplicating work within the pool?
This is an excellent question. The pools do not need to divide the nonces across participants as far as I can tell, it's not necessary.
The nonce isn't big enough (being only 32 bits) for today's difficulty, so miners change other parts of the block (known as extranonce, which is part of the coinbase transaction). The chance of two participants in the pool (or the universe for that matter) selecting the same nonce/extranonce (of which there are 2^256) is next to nil.
The pool does something different: it isolates a group of participants which then divide the mining reward among themselves, proportional to the hashrate they each have contributed. The "value" is that unlike with "solo mining" everyone gets something, as opposed to there being only one huge winner.
But I don't see how pools are in any way "bad" for the decentralized PoW clock, how the reward is divided afterwards doesn't really matter to the blockchain. If anything may be pools make mining more attractive/profitable and because of them we have the insanely high hash rate?
3 comments
[ 3.2 ms ] story [ 15.8 ms ] threadI disagree with this characterization.
Mostly because there actually is a more proper concept of Proof-of-Time, that doesn't depend on Bitcoin's dynamic difficulty adjustment to space puzzle solutions a certain time apart on average.
It is what Bram Cohen is trying to use in conjunction with Proof-of-Space to make [1] a greener alternative to Bitcoin.
One candidate problem for proving time is modular square roots. In a group of order 2^k-1, we can square element x (k-1) times to obtain x^(2^(k-1)), which is the square root of x^(2^k)=x. Since the k squarings cannot be parallellized, this takes time proportional to k.
[1] https://chia.net/
>> It is noteworthy that since SHA is progress-free, each attempt could be thought of as a participant joining the effort and immediately leaving. Thus miners join and leave, trillions of times per second.
Is the value of pooled mining that participants in the pool share the nounce they’re calculating, to avoid duplicating work within the pool?
This is an excellent question. The pools do not need to divide the nonces across participants as far as I can tell, it's not necessary.
The nonce isn't big enough (being only 32 bits) for today's difficulty, so miners change other parts of the block (known as extranonce, which is part of the coinbase transaction). The chance of two participants in the pool (or the universe for that matter) selecting the same nonce/extranonce (of which there are 2^256) is next to nil.
The pool does something different: it isolates a group of participants which then divide the mining reward among themselves, proportional to the hashrate they each have contributed. The "value" is that unlike with "solo mining" everyone gets something, as opposed to there being only one huge winner.
But I don't see how pools are in any way "bad" for the decentralized PoW clock, how the reward is divided afterwards doesn't really matter to the blockchain. If anything may be pools make mining more attractive/profitable and because of them we have the insanely high hash rate?
Thanks for pointing this out.