Hacker News Reader

Comment by cowtools | Hacker News Reader

>Far from “not exactly how market cap works”, this is more or less “not even close to how market cap works”.

>Econ 101 time: value is subjective. Human beings merely use value as a description - it does not exist inside of any object or inside of any equation. A “store of value” has always been an ill considered term - but this is not the argument against it.

Who cares? The general point here still stands: bitcoin's valuation will have to double every halfening (in terms of some indicator like USD) in order for the cost of security (total block reward) to stay constant in (terms of some indicator like USD).

>What Bitcoin is good for is being internet dollars that can’t be spent twice. That’s a nontrivial feature.

Any cryptocurrency can do this, and many do not have bitcoin's crippled market cap or development structure. The question is this: "Why bitcoin?". Bitcoin's sole advantage over other cryptocurrencies is its name brand.

View on HN

No, it doesn't. Bitcoin's difficulty automatically scales based on how fast the previous blocks were "solved". With that, if miners start dropping off, then the protocol will adjust for the falloff in "hashrate".

the problem is that Bitcoin mining has to be sufficiently popular for the network to be secure. if the amount of miming ever drops too low, it becomes attackable.

Which has happened to many of the lesser altcoins.

Look for a boom in that, with all those underutilized GPU miners.

How would the amount of mining drop too low? There are only two cases:

- Miners fall off the face of the Earth. Bitcoin is still valuable so others rush to their place because there is money to be made.

- Bitcoin stops having any meaningful value due to other factors, so there is no reason to mine.

There is no option that Bitcoin is valuable and everybody stops mining thus opening the network to attack.

>- Miners fall off the face of the Earth. Bitcoin is still valuable so others rush to their place because there is money to be made.

Your logic does not follow. What I am saying is this: You cannot continue to halve the mining rewards while maintaining network security.

At some point you reach a point where the cost/benefit of attacking bitcoin (whether that be a deliberate 51% attack or a selfish mining scheme) outweighs the cost/benefit of mining faithfully.

There may be some irrational agents who continue to mine at a loss, but they will be outnumbered by rational agents with more capital.

Your argument assumes that the Bitcoin price stops rising after halving, which does not make any economical sense. And even when all Bitcoin has been mined, transaction fees will prop miners up (though it hasn't happened yet so it's valid to be doubtful this'll happen in practice)

And also depends what you mean by securing. All clients, miners and not, are capable of checking transaction and discarding "insecure" and invalid chains. I don't need a miner to tell me if your transaction is valid or not.

>transaction fees will prop miners up (though it hasn't happened yet so it's valid to be doubtful this'll happen in practice)

No, they won't. This is a common talking point, but it is rarely accompanied by facts. Consider that fees currently amount to about about a percent of the total miner payout. There is some research that indicates that bitcoin is vulnerable to a class of double-spending attacks known as selfish mining:

https://eprint.iacr.org/2020/094.pdf

http://fc14.ifca.ai/papers/fc14_submission_82.pdf

>And also depends what you mean by securing. All clients, miners and not, are capable of checking transaction and discarding "insecure" and invalid chains. I don't need a miner to tell me if your transaction is valid or not.

No they aren't. You misunderstand the fundamental design of bitcoin. The problem is not that the non-mining nodes cannot verify signatures or something, the problem is that these non-mining nodes will not be able prove to each other that the signatures to not refer to already spent funds. The non-mining nodes cannot conclusively prevent reorgs that can be used for double-spend attacks.

>All clients, miners and not, are capable of checking transaction and discarding "insecure" and invalid chains.

They do this by selecting the chain with the highest amount of PoW associated with it. If the network hashrate is low relative to an attacker, then any attacker will be able to easily overpower the network and decide which chains are invalid.

If you are new, I recommend this video on bitcoin's design: https://www.youtube.com/watch?v=bBC-nXj3Ng4

I am not new to Bitcoin but I appreciate the link to the two papers, which I should probably dive into.

All I've heard about the transaction fees problem is that "no one knows for sure, it's still too early to tell." From a cursory look, the papers suggest it's possible to update the protocol/consensus to detect and thwart that type of attacks, and while Bitcoin is slow moving, it's not static and there's still plenty of time to find a decent fix to that issue.

> no one knows for sure, it's still too early to tell.

This is true, no one knows for sure. But I am betting against bitcoin here. At some point the fees are just too high and it become prohibitive towards most people using it at all. So despite the fact that the fee/tx is higher, the total fees collected are lower than an equivalent cryptocurrency with more L1 bandwidth.

As the article I posted says, the real problem with bitcoin here is that the "L2" systems like the lightning network reduce load on the "L1", but also you need the "L1" to collect enough in fees to keep "L2" secure (and that's with L2 middlemen taking their own fees). It is a very delicate balancing act.

>the papers suggest it's possible to update the protocol/consensus to detect and thwart that type of attacks, and while Bitcoin is slow moving, it's not static and there's still plenty of time to find a decent fix to that issue.

Yes, it is quite possible to make these technical improvements to bitcoin. But it may be socially infeasible to convince bitcoiners to do this. The BTC/BCH fork was quite disastrous: not because it forked the cryptocurrency, but because it forked the community. The remaining BTC community is strongly anti-hardfork (they are opposed to any non-reverse-compatible changes). The part of the community that was proactive about change died with BCH. But attitudes may change, who knows really.

The literal hashrate is irrelevant. What matters is the comparative cost/reward of an attacker performing an attack on the network, which will continue to decrease for the reasons i've stated.

Sure, the protocol will adjust for the falloff in hash-rate as miners ditch their hardware- a perfect opportunity for a malicious entity to buy up their hardware and use it to attack the network.

But what is the comparative reward you're talking about here?

The world is filled with situations where something with lesser economic value secures something with higher value. That's a sign of efficiency.

My house is worth $250K. But you couldn't capture that much value by breaking in, so I haven't spent $250K on a security system. If I did need another $250K to secure my first $250K investment, it wouldn't be a great investment.

Suppose it only takes $3.5B to attack a $7.3T network. So what? Unless your plan is to burn billions of dollars for funsies, you'll need to capture at least $4B worth of value for that attack to be worth it. You think you can find someone to anonymously make a $4B crypto sale to you, who won't know where to hunt you down when you try to 51% away the transaction after the next block in 10 minutes?

You seem to be missing the existence of transaction fees.

Reply on news.ycombinator.com