on 25 Feb 2019 20:09:45
Right now, if the underlying schema of a table that is brought into a dataflow changes (IE, a new column is added to a table, excel sheet, etc), the dataflow will error out. The solution is simple, open the dataflow in edit mode, click done and save.
What makes this a problem is when you have a pivot. If the pivot ends up with more or less columns than when the dataflow was created, it will error out. The fix is the same, but this makes pivots almost worthless, as it cannot change with your data.
This idea is to make dataflows update the column definitions if the schema evolves.