Zack Campbell on 23 Aug 2021 22:10:24
This would be particularly helpful when dataflow authors are in the process of making/debugging dataflows, since they're likely to have a smattering of failed/canceled refreshes. It's not easy to keep track of *why* they failed (e.g. "this refresh broke because of a poorly-constructed merge")