We have tested Event Metrics one feature at a time. But as we move towards release, we should do some more holistic, exploratory testing. Below are suggestions for some use cases that seem worth checking out.
- Large event, one wiki: EM can theoretically track a per-wiki maximum of 50k pages, Commons files or wikidata items. But what is its real limit? Create an event with a number approaching that limit on 1 wiki. This should be fairly easy to do by combining certain large categories. The Wiki Loves Monuments categories, for example, are conveniently grouped by year. One might also do this by combining some of our most into an event (there's a report of Top Editors on Wikistats 2.0).
- Can the event successfully process or does it time out?
- If the latter, how close to the limit do we seem to be able to get?
- What happens when users try to download a Pages Created report with many thousands of results?
- Do very large numbers get abbreviated and displayed properly on the Event Summary page? E.g., what happens when Pageviews or Bytes hit the billions?
- Large event, multiple wikis: The 50k limit is per wiki. What happens when an event combines multiple wikis and tries to track a large number of items on each?
- Category with Participants filtering: For T205734 we will have tested Category filtering without participants. But do these filters work properly when combined? Results should be for thie intersection of the two, using this logic: Participants AND (Category1 OR Category2 OR Category3).