Show Posts
It decreases the security of the network proportional to how much the time was decreased. E.g. decreasing the block time by half also decreases the security by half because then it would only cost an attacker half as much in order to pull off an attack like a finney attack.
In the case of a Finney attack, the quicker that the average block is mined, the less time the attacker has to execute the double payment and then broadcast his mined block (http://bitcoin.stackexchange.com/questions/4942/what-is-a-finney-attack). It would cost the attacker half as much time, but he would be racing against other miners who would be solving twice as fast. I imagine there are other security issues associated with faster reward time but I don't see how it makes Finney attacks more likely.
The reduced time does not affect how long (time wise) it would take for a transaction to be considered safe. If blocks can be solved with less hash power, then the orphan rate increases and thus lower confirmation numbers become less safe. It would still take 10-20 minutes for a transaction to be considered safe, just that instead of having 1 or 2 confirmations, it becomes having a lot more confirmations
Is this necessarily true? I think it depends on whether the confirmation accuracy scales at the same rate as the number of stale blocks. From what I understand, lower reward time --> more stale blocks --> more stale transactions percolating through the network --> less confidence in confirmations --> need for greater # of confirmations. 1) Is that right? 2) If so, I can't imagine each piece scales proportionally, which is what it seems like you're proposing above (in asserting that safe transaction time wouldn't change).
