Streamlining Analyses on the Linux Kernel with DUKS

With its remarkably extensive code base, uniquely long lifespan, and undeniable importance to modern society, the Linux kernel is trivially hard to maintain. However, its decentralized development spread over many git trees and mailing lists makes empirically assessing the health of its maintainership model nothing short of a challenge. Off-the-shelf data analysis tools fail to capture crucial nuances exclusive to the kernel development model, such as the current authors who take part in every patch submitted, or how the commit flow between trees changes as new release candidates are created for every merge and stabilization window. We propose the Dashboard for Unified Kernel Statistics (DUKS), an innovative framework that supports multiple visualizations and data analyses previously unsupported for the Linux kernel. Using the Linux kernel mainline as an example, we demonstrate how DUKS could provide valuable insights for understanding the health of the kernel maintainership model. By coupling information from the kernel git trees collected from the Software Heritage repository alongside authorship information shared in mailing lists, we envision DUKS as a cornerstone open-access utility to support analyses on the Linux kernel evolution and maintenance. DUKS demo video: https://youtu.be/2RvUgzdr1fo