Quote from: bigreddmachine on August 31, 2014, 06:18:03 PM

Quote from: Litoshi on August 31, 2014, 05:56:40 PM

Bigredmachine:

To answer your question about 20 second block times producing more orphans than 60 second blocks:

The main cause of orphans is when two miners find the solution at the same time. Both miners then proceed to the next block, but one is now on a fork. This happens even if the block time is 20 minutes. The more numerous the blocks, the more numerous the orphans.

If there was only one pool mining a coin, and all the miners were on that pool, there would not be any orphans at all. But since there are several pools, and many solo miners, the chance of simultaneous solutions increases geometrically.

The reason for the short block times here, is to get more coins out there. There is no limit to the number of JPC coins.
JPC has a 120 second block time

Thanks for the reply!

To clarify about orphans, what I meant by shorter times causing more orphans is this:

After a block is found, a non-negligible amount of time is spent propagating that block to all the nodes in the network.
Assuming that this takes 5 seconds (my understanding is that it is closer to 10-12 seconds), 25% of a 20 second block time is spent without knowledge that the previous block has been found. If a block with the same height as the not-yet-known block is found during this small time, it will be orphaned (and rightfully so).
This small time spent syncing the network is roughly the same regardless of block times.
Again assuming 5 seconds to sync the network, a 60 second block time would have ~8% of the time spent searching by nodes occur during this sync-up time, and 92% of the time spent searching be useful time spent. A 2.5 minute block time would mean ~3% is wasted.
In essence, as block time is decreased, the number of Orphans/Share is increased.

My questions then I suppose were really these (again, I am genuinely interested in an asnwer, and not trying to FUD):

What are the repercussions of the much higher rate of orphans? Are there repercussions?
Can this result in more transactions being voided?
And finally, since all PoS implementations rely on centralized checkpointing, this means that the operator of the central checkpoint and those nodes closest to that central checkpoint are able to sync much faster than other nodes. What advantages does this give to those nodes in terms of # of Orphans compared to slower-to-sync nodes farther away?

I think you are combining orphans and rejected shares in your reasoning here. Orphans are formed when two miners of different pools find a solution nearly the same time, where as what you are describing with the lag time above, is the increased amount of rejected shares due to the miners still working on a block that has already been solved.