Hi! I've been looking at Flink for the last few days and have very much appreciated the concept of Dynamic Tables, it solves a lot of my needs and handles a lot of the complex state tracking that is otherwise painful. I have a question about the composability of the system which the docs don't answer.
The docs use the example of 'SELECT user, COUNT(url) as cnt FROM clicks GROUP BY user', where clicks is a stream coming in of user and the url they've clicked. From such a Table, I can then get a retract stream written into an external system, perhaps outputting (true, User1, 1), ..., (true, User1, 2) indicating that User1's clicked on something. Is there an idiomatic way to convert a retract stream into a semantically equivalent table? Thanks, James
