Update readme, examples and IT tests for table creation. #197
Conversation
Also, deprecate unbounded source.
| * The sequence of operations in bounded and hybrid pipelines are: <i>source > flatMap > keyBy | ||
| * > sum > print</i>. Hybrid pipeline also includes a window operation. <br> | ||
| * The sequence of operations in the unbounded pipeline is: <i>source > keyBy > map > sink</i> | ||
| * <br> |
There was a problem hiding this comment.
In the spirit of maintaining the documentation since this is the example file, could we add what the modified pipelines do?
There was a problem hiding this comment.
I dont think we need this much detail in the javadoc. Users will need to see the code anyway.
There was a problem hiding this comment.
In fact, I will remove this pipeline's operation flow part from Table API example file as well
There was a problem hiding this comment.
Any specific reason to remove the flatmap?
There was a problem hiding this comment.
It assumes the presence of three columns: "name", "number", "ts". This seems like friction for the users. Our examples should primarily highlight how to use the connector. We can assume that they are already familiar with Flink.
Also, deprecate unbounded source.
To clarify, tests have been removed to reduce redundancy, not to hide flaky/failing scenarios.
/gcbrun