If I MANUALLY refresh a Semantic Model from the user interface, I do NOT need to specify which tables to refresh. So WHY do I need to do so in the settings of a Pipeline Task that does the same thing.
This thing is broken. If I remove a table from my Semantic Model, this task fails because it still has that table listed. Plus when I go back in to the task to edit it, I need to re-select the Workspace and Semantic Model. Like it doesn't remember those settings.
1 Comment
- Suggestion: Make this task look similar to this mock-up shown in this forum thread: https://community.fabric.microsoft.com/t5/Data-Engineering/Semantic-Model-Refresh-pipeline-task-shows-no-tables/td-p/4756970#M10770 To describe it: Move the Tables selection control to an area with "Optional" as the header. That communicates to the user that one does not need to specify individual tables to get it to work.
