After the previous Beta launch earlier, there were some lessons that we have learned. The first was that we were having issues with budget orders. This was relatively easy to fix. We have added some missing tests, re-wrote some of the matching logic, and made sure market orders cannot be issues against empty exchange. That was relatively easy.
We have also discovered a more serious issue. The first problem was related to similar issues faced by a canadian exchange shut down in February. We had to re-visit some of our internals and we learned that we were also susceptible to some forms of DDoS attacks. The main reason was in the layer between the front-end facing the Internet and accepting all the external requests, and the dispatch mechanism that directs all the requests to the actual exchange. We have spent the last few weeks re-writing the core of our exchange, again. As expected, it was a massive undertaking, and it set us back about 2.5 months from the initial plans for the launch. Migrating from one internal messaging subsystem over to a different one, and making all necessary adjustments to the testing suites and the codebase, was necessary, and it made the overall internal architecture much better, more resilient to failures, and easier to provide fallback in case of power failures.
New built-in incentives
We have also spend some time implementing internal referral subsystem. This will allow early adopters to spread the word about the exchange, and benefit from lower fees. The idea behind the scheme is to reward early adopters that actively put orders in and recommend their friends and family to use the exchange, by permanently reducing their fees. The earlier you come in, and help us grow, the lower your life-long fees will be, and the more you will benefit.
The current scheme has 6 levels:
Reaching the final, 6th level should be reasonably easy for early adopters, and will grant them 0% for life.