Did anyone measure how fast transactions spread through the Bitcoin network? For example, how long does it take for a newly generated block to be propagated to most active clients?
Mainly asking this question to get some general idea whether an alternative protocol that would aim to have faster transactions is really feasible, or is 10 minutes a good buffer.
The alternative currency Geist Geld is experimental and is trying to approach a maximum block rate (that’s currently achievable), and has apparently settled upon a target block rate of 1 block per 15 seconds. However, the block rate doesn’t necessarily make a substantial difference to how you would use a currency as you could still consider a small transaction accepted even before it’s been included in a transaction block, whereas with a faster block rate you may need to wait just as long.
There has been lots of discussion on how a faster block rate would affect things (eg here, here, here, and here) but I think that ribuck summarises it well in that last link: “If you make the blocks too frequent, you get more cases where two blocks are generated at nearly the same time. Due to network propagation delays, this can cause temporary splits in the block chain. On the other hand, if you make blocks too infrequent, it takes too long for a transaction to become essentially irrevocable.”
That same link mentions a calculated latency of 2.11 seconds.
With a block rate that’s too fast, you may end up with miners consolidating within a part of the network with low latency, creating an unfair distribution of mining returns.