enough miners to secure the network to make it unfeasible to compromise, roll back, double-spend, etc.
Also, look at chains that have had problems in that area. Did they disappear? No. The exchanges increased the number of confirmations required before crediting the account for the deposit. They increased their transaction monitoring and imposed withdrawal halts when they see a rise in the risk of a 51% double spend attack.
The saving grace is that a successful 51% attack doesn't give the attacker the ability to do anything other than double spend their own previous transactions. They can't take your coins. In other words, the reward is not that great, and the cost to attempt the attack are high and the risk of loss of funds in flight (due to an exchange halting withdrawals) is also high.
I'm not insinuating that a 51% attack is nothing to worry about, but it certainly would not be fatal to bitcoin. The exchange rate would definitely be impacted but no doubt the brilliant minds in bitcoin would find a way. For example, there is already a fork of the bitcoin core client that changes the mining algo. For emergency recovery if there was, suppose, a state actor that had built a collection of mining farms that controlled 101% of the existing hashrate. Change the algo and reorg from the last block before the attack that state actor at that point has 101% of nothing and no further power to do anything.
Thanks for this thorough response; from this and other comments I see that I wasn't factoring in all the ways in which LN necessitates on-chain transactions. As you point out, the security budget has so far been more than sufficient; I'm just trying to game out if a major government decided the Bitcoin network was an existential threat, but it's hard to predict developments on the scale of decades. The replies here have at least convinced me that LN adoption isn't an obvious future vulnerability at present (which I figured but wanted to be able to explain why), and that there are options should it trend toward being one.
reply