As discussed in the big 'scenario' thread on resultsdb-users:

https://lists.fedoraproject.org/archives/list/resultsdb-users@lists.fedoraproject.org/thread/DJSIOLA3YEH6HNSJEXVPLIZT5TW4CDED/

I think we agreed to do 'scenario' reporting for at least the Taskotron tests that have a 'scenario' which is more than just the 'test case name'. I believe task-rpmdeplint is one of these, so opening a ticket for it.

The requirement is that there should be an item in the extradata, key should be 'scenario', value should be some kind of string that identifies the scenario for that test run, such that either that value alone or that value together with the 'test case name' will uniquely identify a single test 'scenario' for any given item (and any results that share the same 'scenario' for the same 'item' must be duplicates, i.e. "the same" test that got re-run for some reason).

We didn't absolutely decide whether the test case name should be part of the scenario or not. AIUI for depcheck the extra 'scenario' item besides the test case name is the arch, so the scenario could be e.g. 'depcheck.x86_64' or 'x86_64' depending on whether we include it or not.