Hacker News Reader

Comment by paulgb | Hacker News Reader

Something I've always considered a curious decision about Bitcoin is the decision for halvings to occur as discrete events rather than as an exponential decay occurring on each block. Having them be discrete events is probably easier for someone to verify by eyeball, I guess, but it has the disadvantage that it creates moments in time where some large portion of the hardware in the market can suddenly no longer mine honestly and profitably.

This decision seems to greatly increase an attacker's chances at obtaining a huge supply of hardware that is too inefficient to mine profitably, but can be used profitably for a short time to attack the network.

I suppose Satoshi never imagined there would be specialty hardware used to mine, and that whatever commodity hardware mined bitcoin could find another use.

View on HN

This is the only argument I have found against Bitcoin that have substance. I can see this as a problem, but so far there have always been enough miners that was profitable. One day that might not be true though - hopefully it wont.

One reason for the halving cycle could be the way it kinda shocks the price and that makes people speculate and talk about Bitcoin like crazy. This is the perfect way to have a wrecking ball breaking the old broken system. Or it might just have been easier to implement.

Not the fact that the value proposition is poor because its inputs are destructive to society whilst providing questionable incremental value over existing currencies?

I don't follow your attack logic. If mining hardware becomes inefficient why would using it to "attack the network" be profitable? Can you say more about what attack you're thinking of?

He absolutely did imagine speciality hardware would be used to mine:

"At first, most users would run network nodes, but as the network grows beyond a certain point, it would be left more and more to specialists with server farms of specialized hardware."

> If mining hardware becomes inefficient why would using it to "attack the network" be profitable? Can you say more about what attack you're thinking of?

The idea is that under normal day-to-day circumstances, a big deterrent to dishonest mining is that miners are deeply invested in the hardware and attacking the network would devalue that investment (if it craters the market). Normally if you have a big mining operation, the winning strategy is not to “defect”.

But halvings change the dynamic, because they take a bunch of hardware that was profitable one day and make it unprofitable the next. So for a brief period there's a bunch of hardware floating around that can't be used profitably to mine, and so has no value to honest miners, but has temporary value to dishonest ones if they can accumulate enough of it.

I don't think this will be feasible for a while, but a few halvings from now it could be.

> He absolutely did imagine speciality hardware would be used to mine

I stand corrected, I hadn't seen that post: https://satoshi.nakamotoinstitute.org/emails/cryptography/2/...

Satoshi era was consciously operating under a “gentlemen’s agreement” not to do mining farms

But I don’t think we really need to elevate him to a deity level, 2010 bitcoin client had a lot of dumb stuff in it that everyone promptly scrapped when Satoshi went away

Can you demonstrate this at all? Bitcoin has been mined for over a decade and never came even close. When miners drop off, mining difficulty decreases.

If you're talking about a nuclear war scenario, we will have much bigger problems than bitcoin mining being slow.

> Bitcoin has been mined for over a decade and never came even close.

The problem with using past performance of the Bitcoin network as an indicator of future performance is that the exponential nature of halvings means that each decade is a vastly less subsidized environment for the network than the one that came before it. Over the last decade the network has given ballpark 4.2mm BTC to miners for their services; there are now about 1.7mm BTC available for miners for all time.

Transaction fees are supposed to compensate for the decline in newly minted coins. Using very bad back-of-the-envelope math, transaction fees are about $2 right now, with about 2,300 transactions per block. That works out to a couple billion dollars over the last 10 years, which isn't peanuts.

(Again, bad estimate that uses USD and today's numbers as if they'd been the same since the beginning.)

The reward has only gone down in absolute BTC numbers - in terms of purchasing power I'm pretty sure the reward is trending up.

Expecting this trend to continue forever means expecting the value of Bitcoin to double every four years. It’s basically the wheat and chessboard problem: possible the first few times, then quickly impossible https://en.m.wikipedia.org/wiki/Wheat_and_chessboard_problem

Reply on news.ycombinator.com