If it later feels like its too less.. its no problem at all to split satoshis into 0.1 and 0.01 more..
This would require a hard fork. Also if 1 satoshi would be worth something significant we might also consider changing mining schedule to go beyond 2140 - that is make the block reward keep halving into fractional satoshis. This would effect how the last few satoshis are mined, but would have no significant impact on the ~21 million total BTC.
Are you sure it would need a fork? I would think that all it would need would be that everyone update his Bitcoin client.
I'm not a technical person. However, if the denominations start to go below satoshi, the new client would implement this and if miners mine the blocks with transactions below satoshi, it would create a fork. The newer clients will agree and the older client would disagree. This would create a fork.
The only real problem is that with miner fees.
If we have mass adoption, rise in price, what will happen to the fees and output priority?
The relay fees would probably decrease. The transaction fee the miners want to accept is totally up to them. I would imagine that the transaction priority relaying limit would decrease and the block size would increase in the near future. The blocks can then have more free transactions.
Changes in the client would be small.
The miners would be the first ones to upgrade their clients, because they are on the cutting edge, so to speak.
People who didn't upgrade simply wouldn't have their wallets updated when they received transactions smaller than their clients could handle... at least until the combined amounts for any single address reached 1 satoshi or more. And they wouldn't be able to send smaller amounts than 1 satoshi.
The latest Bitcoin core clients look for the best chain. They would still do this, even if they couldn't update the blocks with fractions of a satoshi. Why? Because the miners would still be using updated clients that would be indicating the best chain.
I, like you, can see that there might be problems. They should be able to be solved. If in the future, when miners were down to mining only 2 or 3 bitcoins per block, or if the whole block size process is going to be changed like Gavin wants, by then everyone will be using an updated client. Are there people who still use version 0.2.0 (we are a 0.11.2 right now)?
