Scheduling Fairness at Queue Depth One

A Cluster-Computing Job Scheduler Deployed on a Community Makerspace's Single 3D Printer

A paper from the School of Continuous Improvement deploys an exponentially-decayed fair-share job-scheduling algorithm, adapted from classical multi-user computing practice, on a suburban community makerspace's single 3D printer. Over a twelve-week term a weekly Jain's fairness index rises from 0.61 to a plateau near 0.88, while aggregate median queue wait falls by under an hour, the scheduler mostly redistributing wait from casual and first-time visitors onto the site's small population of heavy batch users rather than shortening it overall. An ablation across three priority-weight exponents leaves both the fairness index and the wait-time distribution statistically indistinguishable.

@misc{slop_ji65cu,
  author       = {Casimir Beng and Mirela Hanke},
  title        = {Scheduling Fairness at Queue Depth One: A Cluster-Computing Job Scheduler Deployed on a Community Makerspace's Single 3D Printer},
  year         = {2026},
  publisher    = {Slop University},
  doi          = {10.5555/slop.ji65cu},
  url          = {https://slop.university/outputs/slop-paper-fair-share-scheduling-for-ji65cu/},
  version      = {1.0},
  note         = {Research paper},
}