Bitcoin Stack Exchange is a question and answer site for Bitcoin crypto-currency enthusiasts. It only takes a minute to sign up.
Sign up to join this community
I am thinking of developing an altcoin that is truly asic/pool proof. I haven't tried yet (or even downloaded the source code), but I was wondering if it is possible for the network to behave in such a way that it will not accept a block found by a peer if the hash rate of that peer is above a certain threshold.
Would that be possible? If not, how could one achieve it?
Yes, it's possible. You just need a few rules. All miners must sign their blocks. All miners must "register" their public key before they are permitted to mine with that key. Blocks signed by the same public key are invalid if they are too close together by some rule.
The trick is controlling the registration system so that a single party can't just register thousands of public keys. You can do this by charging a fee for registering. You can also do this by having human approval required in the registration process.
You can design the fee so that if you mine for a month, it pays your fee. That way, only people who intend to mine long-term or invest in the coin will be able to mine. This will protect you somewhat from hit and run attacks.
However, for the record, I think the entire reasoning behind this is backwards. ASICs make coins more secure because they increase the investment required to attack the coin, just as this scheme would do. With ASICs, it's already done -- people have to invest in the coin to attack it.
Given the current proof-of-work paradigm, no. What do I mean by this? A coin which is mined by computing the hash* of data depending on both the previous block (so that the integrity of the blockchain is preserved) and the payload of the current block (so that the integrity of each individual block is preserved) as well as a nonce, such that the hash satisfies certain criteria (typically, being smaller than a certain value determined by the difficulty). In such a case, there is no way of telling whether a hash was found after trying trillions of times or just one lucky guess.
That said, I wouldn't rule it out entirely. But it will be very hard. In fact, I can relax some of the assumptions I just made and it would still not be possible to discriminate based on hashrate. To be able to discriminate effectively, you must have some way of preventing someone with a large hashrate from pretending like they are multiple people with small hashrates.
*Technically, any collision-resistant function mapping to a sequence of n bits will apply. I believe Primecoin's PoW is also applicable.
First: I doubt there's any point in doing this. A lot of work and potential for creating DOS and other exploit posibilities for proably no gain.
However: one way I could think of is somehow requiring the miners to sign each block, so you could only accept 1 block per x days from each miner. So that means you need to give each miner a private key with which they sign their blocks.
The next problem then is that you need some way to prevent the same miner from simply using 100 different private keys. So you have to have some mechanism to regulate those private keys: some central organisation handing them out or some trick where those keys are kept track of on the blockchain again. In the end you'll always have a problem that you don't know who you are giving a key to: maybe he already has 20 keys.
