Argues that abandoning Scientific Linux was a strategic mistake because the distribution provided institutional independence and option value that short-term engineering savings could not capture. Scientific Linux began as a community-driven rebuild of Red Hat Enterprise Linux to give large, distributed experiments a stable, redistributable platform compatible across institutions. CERN and Fermilab shifted to CentOS after Red Hat and CentOS joined forces in 2014 to avoid duplicating rebuild work, and Fermilab ultimately retired Scientific Linux rather than produce SL8. That decision removed a ready, credible exit and the small but crucial team, processes, and infrastructure needed to reproduce an Enterprise Linux stack.
Subsequent changes vindicate the warning: Red Hat converted CentOS into CentOS Stream in 2020 and altered how RHEL sources were published in 2023, forcing the community to adopt alternative rebuilds like AlmaLinux or to leave the Red Hat ecosystem entirely. Practical pressures - RHEL 9/10 CPU baselines and the prohibitive cost (≈5.4M CHF) of replacing decades-old accelerator hardware - led CERN to plan migration of over 2,200 front-end systems to Debian 13 by end of 2026. The lesson: maintaining independent, community-backed infrastructure can be cheaper and strategically essential, and institutions should resource projects that preserve the option to operate independently.
Summary generated by AI from the linked article. hn.today is not affiliated with Hacker News or Y Combinator.