Trying to reconcile an inspection batch. Failed checks and reviewed retries are both in the history, which is fine. After restart, the dashboard counts attempts as if they're separate accepted housings. Makes us look nearly finished when we're not.
I need a sensible accepted-item view without throwing away the earlier failed checks. Anybody separated these cleanly?
Then make the reviewed disposition explicit. Keep attempts underneath the item, but derive the accepted count from the authorised disposition, not whichever result arrived last.
I recently had an old completion applied to a new attempt in an offline test. Before changing your totals, check that restart and duplicate delivery cannot attach a result to the wrong attempt in the first place.
Found two issues in the export: repeats with the same attempt ID, plus separate reviewed retries. I had lumped them together as duplicates. They're not.
Use one of each as a worked example for whoever maintains the report. Identical red lines on a screen can mean very different things (repeated delivery versus a genuinely new check).
That is exactly why I kept the event history in my test. Removing the repeated rows would have made the total prettier without showing whether the handler could cope when they arrived again.
Thanks Toby, those two examples made the review much easier. We now have attempts and reviewed item disposition shown separately in the draft. Quality still needs to reconcile the outstanding housings, so I'm not posting a new accepted total yet.
Put that outstanding number beside the total when it is reconciled. Otherwise planning will read the big accepted figure and miss the work still waiting for a decision.