Quote from: Delek on October 21, 2015, 01:26:40 PM

Thanks for your kindly explanations!

You're welcome.

Quote from: Delek on October 21, 2015, 01:26:40 PM

There's something I don't understand however, let's assume that a bitcoin address makes a valid transaction that spend all of this balance, while checking the validity of that transaction you have to go back in time, in transactions and in blocks.

I know now that the antiquity of particular coins can't be determined (thanks!), but antiquity of an address is just how much blocks you had to analyze to tell that the transaction is ok. Isn't?

Transaction validity further highlights the fees issue. Consider a coinbase transaction with a total output of, say, 25.3 BTC. To be valid, we need that the total of all the fees of the transactions included in the block is at least 0.3 BTC. To check this we'd need to check a large number of seemingly unrelated transactions. Working recursively, we'd soon find ourselves checking transactions from 2010 if not 2009.

Also, you may have noticed that neither shorena nor I have yet used the term "address". Bitcoin addresses don't really exist at the protocol level and are more of a higher-order convenience. At any moment in time, an address could be associated with any number of different unspent transaction outputs. The validity of a transaction which spends the balance of an address relies on the validity of each of these outputs. The "antiquity" of a heavily used address, as you've described the term, will rise and fall with time as the set of unspent outputs it is associated with changes.

If anything, the "antiquity of an address" should probably correspond to the minimum of the numbers of the blocks in which it appears.