For anyone interested in streaming SQL: I see that Storm’s implementation of streaming SQL using Calcite is making progress. See https://issues.apache.org/jira/browse/STORM-1040 <https://issues.apache.org/jira/browse/STORM-1040>.
In my most recent comment on that case I raised the idea of creating a TCK for streaming SQL. (Basically, a set of tests that live in Calcite but can be used by any engine.) I would love to hear from people using Calcite for streaming SQL — Storm, Samza, potentially Flink, and others — if they would like to collaborate on this. I believe you all should be striving to produce a better implementation of streaming SQL, not to write your own specification! Julian
