I think the idea of correlating the time versioned table to the record time of the append table is crucial to get the semantics of the join.
Maybe we can add some "real-world" examples for each join type in a follow up issue to illustrate when to use what join. [ Full content available at: https://github.com/apache/flink/pull/6741 ] This message was relayed via gitbox.apache.org for [email protected]
