Quote from: Reid on Today at 08:14:48 AM

I doubt that. Transaction cost may go cheaper and faster. Many things could happen.

First of all, I wouldn't expect a base layer to be able to handle fast and cheap transactions, because (1) it is seen as a settlement layer like Fedwire, which is also a very slow system used to transfer considerable amounts of money between central banks, and (2), at the same time, bitcoin needs to be slow to sustain decentralization and censorship-resistance through the distribution and increased number of full nodes.

Second of all, the bitcoin block is limited in size and mined roughly every ten minutes. The former implies intense competition between users to get into the closest possible block and, therefore, growing transaction fees. The latter implies that no matter how hard you try and how much you pay, you can't get your transaction to confirm faster than miners solve the mining puzzle.

Third of all, each node has a configuration setting called minRelayTxFee that specifies what minimum feerate a transaction should have to be included in a node's mempool and propagated to a node's peers. The current default value is set to 1 sat/b, but if all nodes of the network change it, for example, to 2 sat/b, 1 sat/b transactions will rarely get into blocks. Of course, you can convince a miner to accept even a zero-fee transaction or mine it on your own, but most of the time miners will strive to accept only well-paid transactions.