Hi all, I’m looking into Table API for my new project. It looks like a sweet spot between DataStream API/SQL.
However, it doesn’t seem like the expressivity of Table API equals to that of DataStream API. My previous Flink projects were building simple pipelines using DataStream API with custom evictor (FF Berlin 17) and custom trigger (FF Berlin 18). I believe these pipelines can be expressed with Table API/UDF/UDAF except the custom windowing components. Do I have no way but to change the table into DataStream to use the custom components at the moment? Best, - Dongwon
