Tonight the bitcoin instance for sending rewards crashed and started to show corrupted data. So I stopped sending rewards and will have to investigate it manually later today. I have precise logs about what happen, no coins are lost. I'm very sorry for troubles with delayed payouts.
My confirmed reward stopped going up a few hours ago. I'm guessing that is related.
Thought experiment: if a pool operator in a pool this size claimed technical difficulties making payouts, how many bitcoins could he collect before miners would cut their losses and move their miners to mine solo or in another pool? At 150 GHash/Second, a trick like that right now would allow him to collect more than 2000 BTC/day until people started getting wise to what was happening. Compare that to the less than 50 BTC/day he makes collecting 2% and you can see how that might be a profitable exit strategy if he decided he was done being a pool operator.
This is NOT an accusation of any kind, and I am not pulling my miners out of the pool or suggesting that anyone else do so. It's just that this particular problem makes me wonder what would happen if slush were evil.
I think the test for evilness would be whether after 24 hours or so the pool operator started making manual payments to the biggest miners while he worked on fixing the problem. I think that would be sufficient to prove no ill intent.
