The “Human Factor”: Why a Next-Gen Linux Filesystem Is on the Ropes
The first release of the Linux 6.17 kernel has arrived—yet it contains no updates related to the bcachefs file system. And the reason lies not in technical shortcomings.
On August 10, Linus Torvalds announced the release of Linux 6.17-rc1. As is often the case, he voiced irritation over some late-arriving RISC-V patches. But what stands out far more is the absence of bcachefs code, despite developer Kent Overstreet having submitted minor changes at the end of July and claiming that its “experimental” status would be lifted in Linux 6.18. Torvalds, however, declined to accept the patches. It now appears that bcachefs—first merged into the kernel with version 6.7—will not attain “stable” status in the near future.
This omission carries particular weight for Canonical, which plans to ship kernel 6.17 in its upcoming Ubuntu 25.10 Questing Quokka release. The distribution has already entered feature freeze, making it highly likely the kernel will remain unchanged. With strong probability, 6.17 will be the final kernel release of 2025 and a candidate for the next Linux LTS edition.
Meanwhile, bcachefs has become the subject of heated debate on the Linux kernel mailing list. Overstreet once again criticized the rival Btrfs file system, reminding developers that it was shipped with unresolved issues and still lacks trust—citing Red Hat’s decision to drop it back in 2017. Even bcachefs’s slogan—“the Linux file system that won’t eat your data”—was a thinly veiled jab at Btrfs.
This time, however, fellow developers rallied to Btrfs’s defense. Josef Bacik, an engineer at Meta, rebuked Overstreet sharply, calling his behavior unacceptable and reminding him that many contributors already view him as toxic. The discussion quickly devolved into personal attacks rather than technical arguments. Although Overstreet pledged to refrain from further criticism of Btrfs, he was met instead with accusations of dishonesty, suggestions that he seek psychotherapy, and even insinuations of mental health issues.
This is far from the first time that the fate of Linux technologies has hinged less on technical merit than on personal conflicts and the sway of influential figures. History recalls how, in the early 2000s, EVMS and LVM2 competed: the former boasted richer features, yet the latter prevailed—simply because its architecture appealed more to kernel developers. EVMS soon vanished.
The bcachefs situation bears troubling similarities. Despite its technological strengths, the project risks being sidelined due to frayed relationships and community discord. Such an outcome would represent a serious loss for the Linux ecosystem—not on technical grounds, but because of the all-too-human factor.
