Having complete tests for this repository would be crazy (because some of the stuff we run could take weeks). But we should create a small-ish runcard per runmode so that we can check that things can run.
Even better if we do it regression-test style, so that we ensure that when results change we know why it is.
Having complete tests for this repository would be crazy (because some of the stuff we run could take weeks). But we should create a small-ish runcard per runmode so that we can check that things can run.
Even better if we do it regression-test style, so that we ensure that when results change we know why it is.