Hello,
TL;DR To what extent should we test Desktop applications? How to automate them?
Currently, we are testing all Desktop applications as required by the release criteria, but we have identified different approaches with different testers which resulted in several problematic bugs in the very last moments of the release cycle. We would like to avoid such a situation in the future, so we would like to discuss this matter a bit more.
The release criteria state that the applications must *withstand basic functionality tests*, but it is not very clear what the basic functionality is and how we shall test it.
Therefore, we would like your view on the following questions:
- What should be tested in the scope of basic functionality in general? - Are there specific features or workflows you would like us to test with specific applications? - What could be set as an absolute required minimum for the Desktop applications to not block Fedora (ergo to make Fedora pass the Go/No go meeting). I would like to work on automating this so that we could have that tested very frequently. Otherwise the automation is difficult and requires a lot of time spent if we want to test for the overall functionality.
Thank you very much for your help.
I think this is about https://pagure.io/fedora-workstation/issue/310 and if so, the working group should discuss whether we want to do this in a small break-out group (couple of GNOME folks, couple of QA folks), or if we want to just invite QA folks to a regular Workstation meeting?
-- Chris Murphy
desktop@lists.stg.fedoraproject.org