Status as of Friday, December 19
Posted: Fri Dec 19, 2014 9:09 am
Well, we've finally made it. As I told Chris, we work the entire year for the 17 days that are coming up. It really is a sad state where we spend most of our time doing things we don't want to do just so that we can get these two weeks to do what we do want to do.
Fortunately, this project is something that we do want to do and the pace will be fast and furious over the next two weeks. First up, there is a release now scheduled for Sunday. Rather than using bullets, I'll just summarize that this release is the most significant that we will have made since the launch of the pool, but it also will probably be the most invisible. A significant amount of code has been refactored to eliminate concurrency errors by moving actions out of database triggers into the mining server.
The most noticeable features for most will be increased profitability, from several ways. First, the system is making suboptimal choices now. We aren't exactly sure why and may never know, but some change we made in the new version fixed this. Second, transaction fees are not being properly credited to miners in all cases, although that's only responsible for less than 0.2% of losses, if that. Third, sometimes concurrency errors would result in miners being credited shares at the lower rate (after a switch occurs to some other coin) rather than at the higher, correct, rate (before a block is found) when a coin switch occurs.
We'll start testing the Poloniex code next week, once this release is finalized and completed on Sunday.
Fortunately, this project is something that we do want to do and the pace will be fast and furious over the next two weeks. First up, there is a release now scheduled for Sunday. Rather than using bullets, I'll just summarize that this release is the most significant that we will have made since the launch of the pool, but it also will probably be the most invisible. A significant amount of code has been refactored to eliminate concurrency errors by moving actions out of database triggers into the mining server.
The most noticeable features for most will be increased profitability, from several ways. First, the system is making suboptimal choices now. We aren't exactly sure why and may never know, but some change we made in the new version fixed this. Second, transaction fees are not being properly credited to miners in all cases, although that's only responsible for less than 0.2% of losses, if that. Third, sometimes concurrency errors would result in miners being credited shares at the lower rate (after a switch occurs to some other coin) rather than at the higher, correct, rate (before a block is found) when a coin switch occurs.
We'll start testing the Poloniex code next week, once this release is finalized and completed on Sunday.