Get the latest tech news
Repeatability: As Difficult as it is Important
Reproducibility or repeatability of experimental results is a cornerstone of the scientific method. This should be true for software systems
Reproducibility—or rather, repeatability, if you want to replicate a result starting with the authors’ software artifact rather than just based on your reading of a technical paper—has not historically been a strong suit of the experimental systems research community. For reasons that will become clear, I find myself thinking about this highbrow principle in the context of open source, where the goal is to lower the barrier for others to replicate and use your software—to repeat the process of deploying a system, but for their own purposes rather than just to verify your results. Our Edge Cloud Operations book documents the full scope of the problem space, and describes a specific combination of tools that provide an end-to-end management solution.
Or read this on Hacker News