Status as of Monday, February 16

Post by **Steve Sokolowski** » Mon Feb 16, 2015 9:27 am

Today's status is that the release yesterday was successful. I apologize for not updating it yesterday after the release.

Here are some things that you should see immediately:

All existing coins are now traded on Bittrex. If the Bittrex prices are more favorable than the other exchanges, then those prices are already being used to increase profitability automatically. We saw several instances last night where eGulden was sent to Bittrex because it had a higher price than the other exchanges did.
The website has been optimized to load faster by removing animations on the charts.
There were some display bugs relating to dollar payouts that have been resolved.
Previously, coins that were below network difficulty 1 would not be mined. Now, such coins are mined as long as their difficulties are greater than 0.07, which allows us to take advantage of easy networks. In these cases, we only assign 9 times the network's current difficulty to that coin, so that we don't overload the network and orphan our own blocks.
Miners are now better split across coins so that multiple coins are more often mined at once.
The profitability at the top of the page and at /exchange/instantProfitability now reflects exactly what miners actually earned during that period, rather than an estimate of what miners should earn. While we don't know for sure what will happen, I suspect that profitability will appear to decline, because inefficiencies in ASICs and network latency will be reflected there.

Here's what should occur over the next day and week:

Orphan rates should decline, since we introduced a new method of threading to allow faster database writes. However, orphan rates are computed as the average over two weeks, so the numbers will slowly decline over time.
Chris will add 20 Bittrex-exclusive coins to the pool. Some of the coins will increase profitability.
Chris will complete testing of the data archival software and run it, again increasing performance. It is unknown what effect (other than making backups more reasonable) the increased performance will have.

All in all, the downtime yesterday lasted about four hours. The main reason for the length of the downtime was that we needed to add a new column to a table with 250m rows, which took 2.2 hours. The data archival scripts should reduce such downtime in the future.

All that's left to maximize scrypt profitability to the greatest possible degree (now that bter is offline) is to enable trading with intermediate coins. By two weeks from now, I expect that the trader will be able to compute whether coin -> LTC -> BTC is a more profitable path than coin -> BTC, and if so, perform the better trade. In my research, I determined that such trades can increase profitability by 2%.