v1.0

• New API to configure the disruptor (look at the performance tests to see some samples)
• Removed value type ring buffer
• Improved significantly the perf test runner (better reporting, adjusted duration of tests, print additional system informations, print warnings if insuficient cores or HT is turned on)

10th July - v0.7

• Add batching support on producer barrier
• Setup continuous build and new build targets to generate packages

8th July

• completly restructured performance test projects
• command line options to select test scenarios, implementation, number of runs and number of iterations (num of iterations does not work yet)
• generate a html report at the end of the test run

5th July 2011

• Added Microsoft TPL Dataflow comparison to Uni and Pipe perf tests
• Ported all remaining performance tests. All Java code has now been ported.
• Added a new constructor to BatchConsumer taking 2 actions: onAvailable and onEndOfBatch -> we no longer need to implement IBatchHandler for every use case.

4th July 2011

• ported latency test
• ported latest disruptor (Java) changes
• fixed bugs
• performance (throughput) is now equivalent to the Java version

3rd July 2011

• removed IEntry and replaced it with struct Entry<T>: data processed by the disruptor no longer need to implement a specific interface
• split the implementation in 2: RingBuffer<T> for reference type and ValueTypeRingBuffer<T> for value types.

28th June 2011

• automated build process
• imported several .net performance test tools (ClrProfiler, Perf Monitor) and found one source of allocation. This is now fixed, the disruptor does not create any garbage once started.
• first perf test ported (1 producer, 1 consumer): first results show about 9-10 million msg/sec on my dual core i7 @ 2.67GHz
• perf test does not stop properly all the time, to investigate...

27th June 2011

• code base: all classes have been ported except Histogram.java
• unit tests: all unit tests have been ported but 2 are failing and need to be fixed.
• perf tests: TODO