On April 8, 20226, the French Digital Interministerial Directorate (DINUM) advertisement that will migrate your jobs from Windows to Linux. He ordered all ministries to present a plan by the fall with the aim of eliminating dependencies on non-European software. The announcement in fact goes beyond changing Windows for Linux: it also affects collaborative tools, antivirus, AI, databases, virtualization or telecommunications. It is, on paper, the largest operation to replace proprietary software with free software that a Western State has ever attempted. And if the history of this type of projects teaches us anything, it is that many have ended in failure.
French sovereignty. It is not that France is a lover of free software; What has happened is that the relationship with the United States has changed. Trump’s tariffs accelerated a debate that had been postponed for years: To what extent is it sustainable to depend on the US digital infrastructure? French companies like OVHCloud and Scaleway did not stop growing in 2025, but France has already taken some previous steps recently.
In January 2026, announced the plan to replace Microsoft Teams and Zoom with its own video conferencing platform, called Visiowith the aim that its 2.5 million employees would use it. At the moment 40,000 of them are using it, and it remains to be seen if the deployment ends up being total.

This was LinEx, the Linux distribution derived from Debian that was used in public organizations in Extremadura.
Spain tried it in 2002. The Junta de Extremadura is one of the most famous cases of attempted replace Windows with Linux in public administration. In 2001 it launched LinEx, a Linux distribution based on Debian, and tried to implement it massively in the educational environment and in the health system of the autonomous community.
That was imitated in other Spanish regions: Andalusia had Guadalinex, Valencia had LliureX, Madrid had MAX, Galicia had Galinux, Catalonia had Linkat and Castilla La Mancha had Molinux. All of these projects proposed an alternative to the absolute dominance of Windows on the desktops of public officials, and they all failed, but the biggest failure was the one that promised the most: that of LinEx.
LinEx myths and realities. Although this distribution worked reasonably in the aforementioned education and health environments, it never fully penetrated the general public administration of the Autonomous Community. In 2011 the project was transferred to a state foundation due to budget cuts and by then only 1% of positions of the Extremaduran autonomous administration used free software.
The final blow came when SAP, which managed the community’s medical records system, decided to stop supporting Linux. That made this body return to Windows, and in fact in 2024 the Board formally eliminated the obligation to use gnuLinEx.
Rise and fall of Linux in Munich. another case even more famous At the European level it was Munich. In 2003, the city council of this German city announced that it would migrate 14,000 Windows computers to LiMux, its Debian-based Linux distribution. In 2013 the project seemed a success: there were 12,000 migrated computers and theoretically more than 11 million euros had been saved in licenses and other costs.
However, in 2014, complaints about loss of productivity and debate began. ended sharply: At the end of 2017, the leaders of Munich decided to migrate 29,000 PCs of their employees to Windows 10 from LiMux. The initial migration was never complete, and in many cases there was a mix of Windows and Linux systems to complete the processes, something that seemed inefficient and never managed to eliminate the dependency on Windows and especially on legacy applications.
But there are silent successes. LinEx and LiMux failed in Spain and Germany, but there is precedent to show that abandoning Windows in favor of Linux can work. It proves it GendBuntua version of Ubuntu that was implemented in the French National Gendarmerie. This organization was already a pioneer in the adoption of the OpenOffice.org office suite in 2005, and since 2008 the plan was to abandon Windows in favor of its own Linux distribution. In June 2024, GendBuntu runs on 103,164 jobswhich represents 97% of the IT park of this organization. This has also saved around two million euros per year on licenses, and has reduced the total cost of ownership (TCO) by 40%.
Another promising example: Schleswig-Holstein. This German state began its migration from Windows and Office to Linux and LibreOffice in 2021. In early 2026 had already completed almost 80% of the transition in its 30,000 jobs and according to its data that allowed savings about 15 million euros in licenses only in 2026. A one-time investment of 9 million euros is planned to complete 20% of the process, which is still tied to certain specialized applications and will therefore take a little more time. This is the model that is closest to the French initiative: gradual migration and above all a political will that is maintained among the legislatures that are in power.
What distinguishes success from failure. Cases that work share three characteristics. The first, gradual and phased migration, not sudden and massive. The second, real internal technical support that goes beyond political declarations. And the third (and probably the most important), a sustained will beyond an electroral cycle. Those who fail share three others: trying to migrate everything at once, underestimating the cost of legacy applications and depending on the project not changing government, which certainly contributed, for example, to the failure of LinEx.
A colossal challenge. Installing Linux on a computer is trivial today and it is true that today the learning curve has been significantly reduced and its use is very similar to that of Windows or macOS. The real problem is in the applications that run on top of it. In public administration there is often critical software tailored for Windows, forms that only work in certain browsers (including the old Internet Explorer), management systems that do not have equivalents in Linux or vendors that simply do not offer support for Linux as happened with SAP on LinEx. Each exception is a small crack in these projects, and the challenge of migrating absolutely everything is colossal.
France already has an advantage over its predecessors. DINUM has been developing the call Numerical Suitea set of office and productivity tools that are maintained by the State itself. This eliminates precisely one of the great historical obstacles: the lack of functional alternatives to Microsoft Office. Munich, for example, opted for LibreOffice, and although this suite is fantastic, it continued to generate compatibility problems with documents from other organizations that did use the Microsoft suite. It remains to be seen whether the French solution has really eliminated these problems, but it is true that such conflicts of incompatibilities are increasingly unlikely.
The other great obstacles: officials who do not want changes. Another factor to take into account is that of the officials of these administrations themselves. Accustomed to doing things in a certain way for years with Microsoft solutions, the fact that they now want to change everything can cause a logical rejection. In Munich the resistance of the officials was one of the core factors of final failure. Josef Schmid, then one of the top officials of the German city government, explained in 2014 that employees were “suffering” with the transition. Other leaders confirmed it and they spoke how users “were not happy and essential software for the public sector is mostly available on Windows.”
Image | Anthony Choren |


GIPHY App Key not set. Please check settings