Repeatability: As Difficult as it is Important
(systemsapproach.org)
We all understand that reproducibility of experimental results is a cornerstone of the scientific method. It should be possible to consistently replicate a finding, and this should be just as true for experimental systems research as for any scientific field. One could also argue that the bar ought to be higher for systems research, since it should be possible for others to not only reuse software artifacts developed for our experiments to replicate the results, but to also perform follow-on experiments.
We all understand that reproducibility of experimental results is a cornerstone of the scientific method. It should be possible to consistently replicate a finding, and this should be just as true for experimental systems research as for any scientific field. One could also argue that the bar ought to be higher for systems research, since it should be possible for others to not only reuse software artifacts developed for our experiments to replicate the results, but to also perform follow-on experiments.