[saalweachter]

We don't know what the cost of transactions in Bitcoin will be yet.

Bitcoin includes two methods to pay miners for the infrastructure costs of running the network: mining rewards and transaction fees. The idea is to bootstrap the network off of mining rewards, and then switch to relying on transaction fees in the future, as Bitcoin reaches its limit of 21 million coins.

The problem is that mining rewards are currently very, very non-negligible, even if they're decreasing. In 2012, the Bitcoin supply increased by over thirty percent. In 2014 another 1.3 million Bitcoins will be minted and sold by miners to pay for the cost of operating the bitcoin network. At current prices, that's nearly a billion dollars.

As mining rewards continue to decrease, that billion dollars a year is going to need to be made up through transaction fees. Mastercard takes in less than 8 billion a year, so those transaction fees are probably going to be substantial.

[modeless]

You're assuming that mining revenue must never decrease, but there's no reason for that to be true. A more likely scenario is that as block rewards go away, instead of transaction fees going up, overall mining revenue goes down. Some miners will exit, but as long as there's money to be made there will still be miners. Bitcoin users and miners will adjust fees until they reach a compromise that's mutually acceptable.

[saalweachter]

The problem with shedding hardware from the Bitcoin network is that it undermines the core of Bitcoin's security.

Let's imagine that 2025, and we hit the fourth mining-rewards-halving. It's now unprofitable to mine given the existing numbers of miners, and users refuse to accept higher transaction fees, so 50% of miners exit, leaving half the profits to half the miners.

This means that 50% of the Bitcoin hardware is now on the market, being sold on the cheap. What do you need to compromise the security of the Bitcoin network? Why, 51% of the hardware--

[maxerickson]

I think the practicality of the attack will still depend more on the price of bitcoin at that point in time than it will on the sudden availability of all the least efficient mining hardware.

At some point obsolete ASIC rigs should be coming on the market for cheap (but right now the reward isn't anywhere near 'full' in terms of the amount of electricity apparently being used to chase it).

(FWIW, I expect a crisis of confidence long before 2025. Let's see how many miners are left after that happens.)

[Sambdala]

Yeah, this is likely to be the case.

Part of the genius of the block award being so top-heavy in the beginning is that it made the network much more secure than it would have been if it had relied on transactions from the beginning which will allow it to mature and stabilize before supply & demand determines what the market is willing to bear in transaction fees.

In addition, it allowed a decentralized method of distributing the coins that isn't able to be gamed.