Comments on the BOSS Proof of Concept of the Federal Chancellary
The successful conclusion of the office automation Proof of Concept (PoC) BOSS by the Federal Chancellary and its continuation in the planned rollout to 3000 work stations across the federal administration by the end of 2027 are productive steps forward to realize “digital sovereignty” as defined by the Swiss Government defined by the Swiss government in November 2025. The final strategic evaluation and the technical report on the PoC discuss the achievements of the project, obstacles and possible ways forward.
Why is BOSS so difficult?
There are real challenges in transitioning from Microsoft 365 [M365] to a fully free and open-source office automation software stack, and many of them are highlighted in the two reports. However, the standalone findings in the reports benefit from being contextualized to make them more interpretable:
E.g., when it comes to migrating data out of M365, the technical report states that
The native M365-Backup is not fit for purpose ([the data] stay in the Microsoft Cloud and there is no export artifact). — technical report, p. 10, own translation
While it is important to acknowledge this circumstance and it being an obstacle for the transition, one can draw different conclusions from this isolated fact for the decision of how to proceed with the migration: When focusing on the challenge (and cost) of exporting one’s data out of M365, it can be used as an argument against moving to a free and open source alternative. With the context that the lack of an officially supported data export or migration tool away from M365 is a specific design choice by Microsoft to prevent customers from leaving its ecosystem, the mere existence of this obstacle becomes a reason to proceed with the transition: The impossibility to export one’s own data isn’t innovation but a customer hostile feature which cements vendor dependencies.
Similarly, the finding that
BOSS [the free and open-source office automation] does not cover all M365 functionalities. The users will have to expect limitations and accept them. — strategic report, p. 11, own translation
can be taken in different ways depending on one’s context: One can hold the missing feature parity against the BOSS free and open-source software solution. However, when taking into account that the free and open source alternative has not had the the billions in license payments and support fees which have been paid to Microsoft over the last decades by businesses and administrations in- and outside of Switzerland, the existence of a functional alternative at all is – despite its limits – properly appreciated. This background also opens one up to the idea that with sufficient investment in and funding for the free and open-source solutions in the future, feature parity can be attained. This is where cross-border collaboration as with the ZenDis or Schleswig-Holstein can pay off for all parties involved.
Analogously, the observation that
The Federal Office of Information Technology, Systems and Telecommunication (FOITT) currently does not provide a central Linux management infrastructure. — technical report, p. 9, own translation
and the resulting lack of official support for Linux-based clients, and consequentially, the tight coupling of third party applications to the Microsoft Windows platform, e.g.,
GEVER (Acta Nova): Not fully supported for Linux. No Roadmap for Linux support by the vendor. Acta Nova can be used via the browser but central productivity features like Office and Email integration are not available outside of Windows clients. — technical report, p. 10, own translation
can be used as an arguments against (the Linux client part of) BOSS as a widespread rollout would necessitate costly change and disrupt established workflows. But by taking this perspective one automatically views BOSS as the problem rather than the existing IT infrastructure which is full of undesirable dependencies in the form of proprietary data formats as well as platform/operating system and vendor lock-ins.
The authors of the two reports in the Federal Office of Information Technology, Systems and Telecommunication and the Federal Chancellary are right to be transparent and vocal about the challenges in the transition to BOSS. However, the bare description of the encountered obstacles is not a complete problem analysis yet as it does not make any statements about the causes of the problem. When taking this missing context into account, it becomes clear that many of the problems in the transition are the consequence of past decisions which did not take lock-in effects of technical dependencies seriously or chose to sacrifice flexibility and resilience for a myopic understanding of efficiency. Being cognizant of the path dependence between those past decisions and their impact on the situation now, is important for the discussion on how to proceed strategically in the office automation on the federal level.
The dangers of a self-limiting vision
The additional context does not make the migration problems for BOSS any less real or challenging, but it shows that not committing fully to a system based on free and open-source software, open standards, and open APIs is just kicking the can down the road until the next crisis which necessitates change. To tackle the problem at the root, the vision for the IT infrastructure in the federal administration should be the full substitution of proprietary components by free and open-source software despite the involved risks and required cultural, organizational, and technical changes. Clearly, that is an ambitious goal which can only be realized incrementally, in the long term, by mobilizing additional resources, and with a strong political backing. The example of Schleswig-Holstein shows that despite the obstacles, defining and pursuing such a vision is possible in practice.
Settling for less ambitious, partial or backup solutions introduces new failure modes: Entertaining parallel infrastructures can become costly and introduce interoperability problems whereby the new solution will be blamed, rather than the dependencies of the existing one. If interoperating across systems does not work or introduces unnecessary friction, people will default to the existing system to get work done, thereby relegating the alternative free and open-source system to a niche. Simply put, a limited scale might make the project too small too succeed.
Additionally, limiting the scope of the vision will allow for the introduction of new proprietary dependencies in the non-migrated parts, making a potential future migration even more complicated and expensive. A transition to a fully free and open-source office automation suite would have been less involved a decade ago, as dependencies, habits, and integrations would have been less entrenched. Further delaying the migration is only going to make it more costly in the future.
Only an ambitious vision for free and open source software, open standards, and open APIs in the IT of the federal administration addresses the issue of technological lock-ins effectively, motivates to overcome the by now entrenched dependencies, and prevents too small to succeed failure modes. If this conclusion would be the outcome of reflecting on past mistakes and on how today’s decisions shape the future, the Swiss government would come closer to realizing “the ability to act in the digital space to fulfill state tasks”.
The Internet Society (ISOC) Switzerland Chapter is a non-profit organization that engages on a variety of Internet-related topics, ensuring that it is a place of possibility, opportunity, and progress that benefits people worldwide. We provide technically-grounded advice, policy recommendations, and educational material regarding privacy, security, Free and Open-Source Software, and digital sovereignty. We also organize informative events and debates like the annual Public Policy Sessions and participate in collaborative research projects like the NGI0 Commons Fund.
As a national chapter of the international organization responsible for the .org domain, ISOC CH acts as a gateway between Switzerland and the international digital civil society.
You can consider becoming a member (through the main ISOC web site) following the instructions at https://isoc.ch/membership, or just subscribe to our newsletter (2-3 announcements per year) by sending a message to contact@isoc.ch.