But the same attack is available on other forms of currency.
Take gold, for instance. If a majority of the people who find gold to be valuable were to agree that they should value Palladium instead and stop using gold, then gold would drop in value and palladium would take over. The people who own palladium could even go around to the gold users and bribe them with some palladium to get them to go along.
Now, there is one important difference between my "gold" example and bitcoins. With bitcoin, there are two populations: the users of bitcoin and the subset of those who are miners. For a change to bitcoin, only miners need to agree -- but that's not as much of a limitation as it seems. For instance, if all the miners formed a secret cabal and agreed to blacklist all existing bitcoins and value only FUTURE mined bitcoins, then what would REALLY happen is that someone else would step in and start mining without that policy, forming a fork in the blockchain. The entire community would wind up deciding which fork to follow. So really, what you need is a strong enough majority of the miners (and those qualified to set up as miners if they wanted to) to ensure that there is no blockchain fork. This is not all that different from my "all users of gold" example.
Gold dropped 9.4% a few days ago. That had little to do with what the majority thought instead people wanted to sell slightly more gold than people wanted to buy and the price dropped.
Ass to messing with the block chain, the real problem is a 51% attack is one that people don't know for a while, start randomly seeding bad transactions and there is no obvious point as to which where good transactions and which where bad ones. People my suggest reversing all transactions but if you traded bit-coins for stuff your now out your bit-coins and your stuff. Net result Bitcoin's probably dies.
Absolutely true. "Introduction of double-spend and other corrupt transactions because you control a significant fraction (perhaps LESS than 51%) of the mining power" is a vulnerability for bitcoin that is NOT present in other forms of value.
But "a majority of the users could redefine what is considered valuable" is a vulnerability (mentioned in the post I was replying to) of EVERY representation of value, and not a particularly interesting or realistic one.
Take gold, for instance. If a majority of the people who find gold to be valuable were to agree that they should value Palladium instead and stop using gold, then gold would drop in value and palladium would take over. The people who own palladium could even go around to the gold users and bribe them with some palladium to get them to go along.
Now, there is one important difference between my "gold" example and bitcoins. With bitcoin, there are two populations: the users of bitcoin and the subset of those who are miners. For a change to bitcoin, only miners need to agree -- but that's not as much of a limitation as it seems. For instance, if all the miners formed a secret cabal and agreed to blacklist all existing bitcoins and value only FUTURE mined bitcoins, then what would REALLY happen is that someone else would step in and start mining without that policy, forming a fork in the blockchain. The entire community would wind up deciding which fork to follow. So really, what you need is a strong enough majority of the miners (and those qualified to set up as miners if they wanted to) to ensure that there is no blockchain fork. This is not all that different from my "all users of gold" example.