Hacker News Reader

Comment by bsdetector | Hacker News Reader

> because of how competition over mining rewards works, [bitcoin] has the characteristic of consuming more and more energy the more it succeeds.

New bitcoin from mining is halved every ~4 years, so every four years miners can afford to spend only half as much electricity to mine from that revenue.

As revenue from new bitcoins tapers off the work expended will self limit to the value of transactions. If you're charged 1% to include your transaction the value of energy used to mine a block will eventually not exceed the fee for including those transactions.

So it doesn't have the characteristic of consuming more and more energy and is self-limiting in how much energy is used. The bitcoin energy problem will take care of itself in time.

View on HN

> New bitcoin from mining is halved every ~4 years, so every four years miners can afford to spend only half as much electricity to mine from that revenue.

Not exactly. Miners are paid out of the sum of block reward and fees multiplied by the market price of BTC. Every ~4 years the contribution of block reward goes down, but that doesn't mean that the price goes down, or that the contribution of fees stays the same.

If it was block reward alone and the "energy problem solved itself" then the blockchain would be completely vulnerable to a 51% attack and it would instantly become worthless.

The expectation is that the contribution of fees will go up as the block reward goes down, although it remains to be seen how much direct fee the market will bear. Currently the actual cost of a BTC transaction is hundreds of dollars - but most of it is socialized via inflation. It is unclear if the market will bear paying hundreds of dollars in transaction costs instead -- and if not, there's no reason the 21M coin limit can't be raised to continue doing exactly what has been happening so far.

> Currently the actual cost of a BTC transaction is hundreds of dollars

I thought... huh, this can't be right. So I did some back of the napkin math. We have about 2500 transactions in a block, a block reward is 6,25 BTC. That comes around a cost of 0.0025 BTC per block, or about 112USD. That's without considering the extra tip from the transaction. So, not really hundreads, but damn' close.

It makes me think increasing the block size really wasn't a bad idea.

> there's no reason the 21M coin limit can't be raised to continue doing exactly what has been happening so far.

Sure, but you'd need a hard fork. Not impossible, but hard to reach the consensus.

I don't buy this argument (just like the point you are responding to is mostly wrong too). Yes it won't consume more and more but it won't consume less and less either. Miners are n't going to mine a loss it will always be SUM(rewards + fees). The network has to be secured or bitcoin loses all perceived value. Difficulty will have to creep upwards as technology improves. The question is will there be a way to mine more difficult problems using less power. That is the only way bitcoin power usages stops growing (slowly).

> The question is will there be a way to mine more difficult problems using less power. That is the only way bitcoin power usages stops growing (slowly).

That's not how it works. The whole point is to spend power (work).

If there was some breakthrough that allowed finding hashes with 10x less electricity, then the network wouldn't burn 10x less electricity. It would instead find 10x more hashes.

Doesn't the security of proof of work consensus depend on the work being fairly expensive?

Yes. An attacker who is prepared to spend the same amount of money as all other network players earn combined can do various attacks.

The goal of the "halve every 4 years, plus transaction fees" model is to make sure the above is never possible for an attacker.

I think they just need more than half?

How can you have more than half without more than everyone else (who by definition have less than half) combined?

They either need to sway more than half the miners to their evil side, or they need to exceed the total of everyone else.

This assumes the price of Bitcoin doesn't keep going up.

Reply on news.ycombinator.com