coordination #64126
Updated by tjyrinki_suse about 3 years ago
There's both 1) potentially indirect The first part would be to identify what packages we are actually regression testing indirectly in some tests and decide if it's sufficient or not, and eventually mark them as automated. Go through existing tests and decide what else is being used there, or the other way around take a specific package and look for where it is used. For example, webkit is used in evolution, so it might be enough to check those tests and there is probably no need for additional resources and test. (the above part of description modified from Juraj Hura's idea) The second part would be reviewing and actually using the collected TEST_SUITE_SUFFICIENT data together with making template generator aware of it to have templates automatically acknowledge when build time regression test suite is enough. The current machine readable data for automation is at https://pes.suse.de/QA_Maintenance/Automation_Progress_Tracking_in_QAM/ and so that should be appended - probably with new sections for indirect and build time ones - as needed with packages that could would be now marked as tested automatically, and 2) database of TEST_SUITE_SUFFICIENT information from Updates Squad that could be used likewise. automatically tested.