Status as of Thursday, October 23
Posted: Thu Oct 23, 2014 9:02 am
Here's a brief update:
- We finally made it into the black. For days, we were mining litecoins with nothing to show for it, but then two blocks were found last night, allowing us to start building up the pool's reserve. Chris computed that we need to hold 2 days more than today's mining: 1 day of which is yesterday's immature customer funds, and another day of profits that are needed to pay out in the case of bad luck.
- The Elacoin blockchain doesn't advance when we're not mining it, so we had 0.1btc caught up in mined Elacoin blocks. We can tell when someone starts mining Elacoins by when the trader is able to sell newly mature blocks. Someone started mining it last night, so our outstanding blocks filtered out.
- We finally figured out why some of the displayed numbers, like the "payout coins owed," are different than the actual balances. The issue is multithreading; Postgres doesn't lock tables on SELECT statements, so we need to do that explicitly. I'm working on a fix for that and we'll probably deploy that the morning of November 1. Unlike last time, this maintenence will take only five minutes. Why do we only release on Saturdays? Because then, if issues come up, I can fix them before Monday arrives.
- I had a significant breakthrough in merge mining last night, and resolved the "Aux POW merkle root mismatch" issue. There is now a new issue that is much further down the dogecoin daemon's list of checks, which I hope to investigate over the weekend. It's possible I could spend one hour and we have merge mining working, since everything else except actually submitting the blocks to the daemon is tested and working.