Status as of Friday, April 3

Great news for today:

• I discovered a bug that caused two work restarts to be sent when only one was needed. This bug is present in the current production code. Even if we weren't taking any other steps to reduce work restarts, this fix alone would result in significant improvements in hashing speed.

I'm hoping Chris will certify the new release any day now. There are no known issues. Our only delay is that he is having trouble setting Kires's miners up. The only reason we need Kires's miners is to make sure that the pool is measuring the delay correctly with an extremely powerful real-world miner. We're pretty sure that the simulator has taken care of all the other testing.