I’ve scheduled a nightly test script using Bitrise build scheduling. For context, this script uses the Bitrise API to download old build artifacts and run them against the current api for regression testing.
When scheduling, Bitrise forces you to assign either a Branch, Git Tag, or Commit Hash. Because I am downloading previously built files, I am not pulling anything from git. In this case, I’m able to assign an arbitrary branch name to submit the build schedule, but it is a bit confusing that one of these values have to be filled out.
A related issue exists with starting a build (without scheduling). When manually kicking off a build with this workflow, I would clear out the Branch value and start the build. However, I noticed when the build began running, it had the master tag assigned to it. This would then be problematic with rolling builds, because if I push a new commit to master, it would kill the manually kicked off build I started (which should not have been automatically assigned to master.
