The goal of this issue is that when a column is sorted in any report, and when two values are identical for this column, then the result will be sorted by a secondary column. Each report could decide which secondary column to sort by, but the default will be
The current behavior is not ideal as the rows appear to be sorted randomly eg. when all values of the sorted column are
0. Example screenshot
I think this is done by default when the data type is numeric. I might be remembering incorrectly, but for strings it doesn't do this, which caused problems for me during the processed metrics refactor.