EU OS in FOSDEM 2026: A Mexican Perspective
Good old Robert Riemann presented some truly interesting viewpoints on FOSDEM this year regarding the EU OS project. I highly respect his movement; in fact, it was a significant inspiration for us starting Fundación MxOS here in México.
That said, respectfully, I have some bones to pick with his presentation.
Vision: Sovereignty vs. Adoption
To me, the MxOS project is fundamentally about learning. It is a vehicle for México to master the entire supply chain: how to setup an organization, how to package software, how to maintain it, and how to deliver support.
MxOS is a blueprint that should be replicated. It is as much about providing the software as it is about learning the ropes of collaboration. We aim to generate a community of professionals who can provide enterprise-grade support, while simultaneously diving deep into research and development.
We aim to mimic the Linux Foundation's role; serving as an umbrella organization for FOSS projects while collaborating with the global community to contribute more code, more research, and more developers to the ecosystem.
A Tale of Two Philosophies
The "Home User" Disconnect
Riemann suggests that EU OS is not for private home users, claiming users can simply run whatever they want at home.
Personally, I think this is a strategic error. For a national or regional OS to succeed, users must live in it. They must get familiar with it. Users will want to run it at home if it guarantees safety, code quality, and supply chain assurance.
MxOS places the user at the center. We want MxOS to be your go-to distro for everything in México; from your gaming rig to your business workstation. Putting the user at the center is where you draw collaboration. That is where people fall in love with the project. You cannot build a community around a system that people are told not to use personally.
Original Code vs. Integration
This is a key divergence. Robert doesn't believe EU OS should produce original software, viewing it primarily as an integration project.
Conversely, I believe MxOS must be a minimal distribution; a bedrock upon which we build new, sovereignty-focused projects. For example:
- libcfdi:
- Our initiative to integrate with the SAT (Mexican Tax Authority) for the validation, generation, and processing of "facturas".
- Identity:
- A project to harmonize Mexican identifiers like CURP, RFC, and SSN.
- Rural Health:
- Software specifically designed for hospitals and clinics in remote areas.
The Container Lunacy
It seems Dr. Riemann proposes EU OS to be primarily a container-based distribution (likely checking the "immutable" buzzword boxes).
While they have excellent integrations with The Foreman and FreeIPA—integrations MxOS would love to have; we are not container-focused.
Warning
To be clear: I am speaking about the desktop paradigm. The current "container lunacy" assumes we should shove every desktop application into a sandbox and ship the OS as an immutable brick. This approach tries to do away with the shared library paradigm, shifting the burden of library maintenance entirely onto the application developer.
This is resource-intensive and, frankly, lazy. We plan to offer minimal container images for the server world where they belong, but we will not degrade the desktop experience by treating the OS as nothing more than a glorified hypervisor.
The Long Game: 50, 100, and Interstellar Support
Riemann touches on "change aversion" as a problem. I disagree.
I am an experimental guy. I live on the bleeding edge. But I respect users who do not want to relearn their workflow every six months. For a long time, the "shiny and new" cycle was just a Microsoft strategy to sell licenses.
But if we are talking about national sovereignty, we are talking about civilizational timeframes.
In MxOS, we are having the "crazy" conversations: How do we support software for 50 or 100 years?
This isn't just about legacy banking systems (though New York still runs payroll on COBOL). This is about the future. One day, humanity will send probes into interstellar space. That software will need to function for 50, 100, or more years without a sysadmin to reboot it. It must be self-sustaining.
We are building MxOS with that level of archival stability in mind. How do we guarantee that files from 2026 are accessible in 2076? That is the standard we aim for.
The Reality Check: Where is México?
Robert showcased many demos and Proof-of-Concept deployments. I am genuinely glad; and yes, a bit envious; to see EU OS being taken seriously by European authorities.
That is not yet our case.
We have ~100 users in our Telegram channel; a mix of developers, social scientists, and sysadmins. I love that individuals are interested. But so far, the Mexican government and enterprise sectors have been indifferent.
We have presented the project. We are building the tools. We are shouting about sovereignty and supply chain security.
It leaves a bittersweet aftertaste. The developers are ready. The code is being written. The individuals care. Why don't our organizations?
We are doing the work. It's time for the country to match our effort.
Distribution Selection: The Strategic Choice
Dr. Riemann’s analysis of distribution selection (favoring Fedora’s immutable bootc architecture) makes a critical omission. He overlooks that the vast majority of FOSS innovation in this space; FreeIPA, GNOME, bootc itself—flows from Fedora and Red Hat.
This is why MxOS chose CentOS Stream 10.
We know CentOS Stream is the upstream of RHEL. This is where Red Hat, Meta, CERN, AWS, Intel, and IBM collaborate. By basing MxOS on Stream, we are closer to the metal. We aren't just consumers; we are positioned to fix bugs before they even reach Red Hat Enterprise Linux.
CentOS Stream is where the magic happens. It offers true security, quality-focused development, and rigorous QA. It is the obvious choice for a serious fork.
We have made significant progress with our build infrastructure (Koji). We have servers but no datacenter. We are not quite there yet, but we are getting close.
Conclusion
Robert makes a great point that we share: Collaboration is key.
We want standards. We want to agree on the fundamentals. And yes, we want to collaborate with EU OS. But we will do it while keeping the Mexican user—and the Mexican reality—at the very center of our compass.