That's OK -- my test suite regimen is:
* at every commit, test everything except the bulk-export tests on Z and JM release
* at every release commit, test everything on Z and JM release
* every night, test everything on Z release and beta, and on JM release and beta
So the only thing that is influenced by breakage on a beta is the nightlies, and I can easily disable JM beta testing there. The nightlies serve as an early-warning system for me to prep for breakage in the next Z/JM release, and sometimes have the side benefit of finding problems in the beta itself. If the betas for JM are only intermittently relevant, I can just structurally disable them in the nightlies -- maybe breaking changes in the beta are signalled adequately by changes in the Z betas.