2 ms·

>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

>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 https://eprint.iacr.org/2020/094.pdf http://fc14.ifca.ai/papers/fc14_submission_82.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 https://www.youtube.com/watch?v=bBC-nXj3Ng4

sph 4y ago
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.
- cowtools 4y ago
  > 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.

sph 4y ago

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.

cowtools 4y ago
> 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.