First Release Fault handling in stream publication and consumption aligns with fault transport facilities of latest gRS2. Faults are published and abort publication unless they are contingencies (cf. StreamContingency annotation). Streams can be conveniently consumed with callbacks (cf. Callback interface and Streams#consume()) and callbacks can explicitly skip elements or stop consumption (cf. Iteration class). Like callbacks, fault handlers can control iteration directly (cf. Iteration). FaultResponse is no longer needed. StreamSkipException is replaced internally with StreamSkipSignal, and StreamStopSignal is also added to model signals from callbacks and fault handlers. New stream implementations as well as code that consumes streams can now be tested with dedicated facilities. Lookahead streams handle null values from underlying stream without going in infinite loop (bug fix) Bug fixing on RSStream ResultSet client timeout increased to 5 minutes