ownCloud has been forked twice.
The first time, in 2016, a significant part of the community and engineering team left to build Nextcloud.
The reasons were real: governance concerns, transparency gaps, disagreements about the relationship between company and community.
People left. The project continued.
oCIS was already taking shape as a ground-up rewrite in Go, and ownCloud 10 kept shipping.
The second fork, in 2025, was after the Kiteworks acquisition in late 2023.
Different context, different reasons, but the same underlying lesson: communication isn’t just about what you say. It’s about what the other person hears. Even well-intentioned decisions fracture communities if words, tone, or timing are off.
What followed was two and a half years of building by a small, dedicated team.
oCIS went from version 4.0 (the state at acquisition) through five production releases to 8.0.1 today.
The Bavarian school cloud (ByCS) runs on it, serving millions of students.
The Desktop Client went from version 4.x to the completely rewritten 6.0 line.
The architecture proved itself at scale.
There was a fear that the code would get closed off and that the community would get hollowed out after the Kiteworks acquisition.
While we kept shipping publicly to our open source repos, we were so focused on building, we did not find and commit the necessary resources into the community.
We will publish a “Lessons Learned” document as part of our OSPO manifesto suite.
It will say, in public, that projects diverge when voices feel unheard and governance becomes unbalanced.
It names what went wrong without blaming anyone.
But more importantly, it commits to doing things differently, with specific mechanisms rather than vague intentions.
Tomorrow: we killed our own CLA.
Note:
This is part 2 of this blog post series. See the first post, A (re)-introduction to the ownCloud community.
