[josephagoss]

You're mostly right, Fees are not paid per address but per transaction.

If I had a thousand addresses full of these amounts (0.0001) I could spend them all in one transaction into a single address.

The fee would end up being higher than the standard due to the recommendation that fees are per byte.

So I might end up with 0.099 in the final address.

So it's possible to move them if you can combine enough of them together to make the transaction worthwhile.

Of course the part where you're right is that it's unlikely a single person owns a large number of these spam transactions and thus the economics doesn't make sense for them to move at this time.

However, if Bitcoin were worth 100x and the fee remained in dollar terms roughly the same , it would now become much easier to move these spam amounts.

[pmorici]

Transaction fees are really a function of transaction size in bytes not per transaction. Adding an additional input address increases transaction size by 180 bytes. The output address and other common overhead is around 44 bytes. The current minimum cost of block space is 0.0000001 BTC per byte if you are willing to wait several hours for your transaction to confirm. At that rate adding an additional input will cost you 0.0000180 so anything less than that is not worth moving under current conditions. To be confirmed in a timely fashion you need to pay a per byte fee of about 5x so in reality amounts under 0.00009 are currently not worth spending.

http://bitcoin.stackexchange.com/questions/1195/how-to-calcu...

https://bitcoinfees.21.co/#delay