Reports that a Swiss federal body plans to shift roughly 3,000 computers away from Microsoft products have drawn attention to a wider trend. This guide explains how such migrations are planned, what they involve and where they tend to stall.
Key takeaways
- Public-sector migrations away from proprietary desktop software are usually staged over years rather than executed as a single switch-over date.
- The stated motivations in European public administrations typically combine digital sovereignty, long-term cost control and reduced dependency on a single supplier.
- The hardest part of such a project is rarely the operating system; it is the file formats, macros, line-of-business applications and identity systems built around the incumbent stack.
- Details of the reported Swiss plan — including the exact agency, timetable, software chosen and budget — are not verifiable from the circulating summaries alone.
- Similar programmes elsewhere in Europe have produced both continuing rollouts and well-documented reversals, so the outcome of any individual migration is not predictable from the announcement.
What is actually being described
The claim circulating online is that a Swiss federal government body intends to replace Microsoft software on approximately 3,000 computers. In practice, a statement of that shape can mean several quite different projects, and the distinction matters for anyone trying to learn from it.
Replacing “Microsoft” on a desktop fleet may mean swapping the operating system from Windows to a Linux distribution; replacing the office suite with LibreOffice, OnlyOffice or a comparable package; moving from Microsoft 365 cloud services to self-hosted or European-hosted collaboration tools; or changing the identity and device-management layer built on Active Directory and Entra ID. Some migrations do all four. Many do only one, most commonly the office suite, because it is the layer with the clearest substitute and the least dependency on hardware drivers.
Without primary documentation from the body concerned, the precise scope of the reported Swiss plan cannot be confirmed. The figure of 3,000 computers, if accurate, describes a mid-sized fleet: large enough to require formal change management and a helpdesk plan, small enough for a single administration to run without a multi-year national programme.
Why this is being discussed now
Interest in this kind of story has been rising across Europe for several years, driven by a set of overlapping pressures rather than any single event. Licensing costs for cloud-based productivity suites have shifted from perpetual purchases to recurring subscriptions, which changes how public budgets are planned. Data-protection authorities in several European countries have examined whether cloud office products can be used by public bodies in compliance with regional privacy law. And “digital sovereignty” has become an explicit policy goal in several administrations, meaning a preference for software and hosting that can be inspected, controlled or relocated domestically.
Switzerland is a recurring reference point in these discussions because it has a longstanding public debate about open-source software in government, including legislation touching on whether publicly funded software should be released openly. That context is why a report of this kind spreads quickly, even before its details are confirmed.
The background a newcomer needs
Government IT estates are rarely uniform. A typical administration runs a mixture of desktop computers, specialised workstations, shared terminals and mobile devices, layered on top of applications that may be decades old. Many of those applications assume a particular browser, a particular document format or a particular authentication mechanism.
Microsoft’s position in this environment comes from bundling: an operating system, an office suite, an email and calendaring service, an identity directory and a device-management system that are designed to work together. The convenience of that integration is also the source of the lock-in. When one component is removed, the assumptions the others made about it have to be replaced by something else.
Open-source alternatives exist at every layer, and most are mature. What they generally do not offer is the same single-vendor integration, which means the integration work moves in-house or to a systems integrator. That is the central trade-off in every migration of this type: licence costs fall, and internal engineering and support costs rise, at least during the transition.
Who is affected, and how
Three groups feel a migration most directly. Ordinary staff encounter changed interfaces, different keyboard behaviour and documents that render slightly differently — small frictions that accumulate into resistance if training is thin. Specialist users are affected more sharply: teams that depend on complex spreadsheets with macros, on statistical add-ins, or on sector-specific software certified only for Windows can be blocked outright until an equivalent is found or commissioned.
IT departments carry the largest load. They must inventory every application in use, test alternatives, rebuild deployment images, migrate mail and calendars without losing history, retrain the helpdesk, and support two environments simultaneously during the transition. Suppliers and local integrators are affected too, since a migration redistributes spending from licences towards services and support contracts.
For citizens, the visible effect is usually minimal if the project goes well. The main risk is documents exchanged with the public rendering inconsistently, which is why standardised open formats are typically mandated early in such programmes.
Where informed people disagree
The disagreements are genuine and not resolved by evidence alone.
On cost, one side argues that eliminating recurring licence fees produces durable savings across a fleet’s lifetime. The other argues that total cost of ownership rarely falls once retraining, integration, custom development and productivity loss during transition are counted honestly. Both positions can be supported, because the answer depends heavily on how much bespoke software an organisation runs.
On sovereignty, supporters hold that public bodies should be able to inspect and control the software processing citizens’ data, and should not be exposed to unilateral changes in a foreign supplier’s terms. Sceptics respond that self-hosting shifts risk rather than removing it, and that a smaller in-house team may patch more slowly than a large commercial provider.
On feasibility, the record is mixed. Some European public administrations have run large open-source desktop programmes for years; others have migrated and then reverted, citing interoperability problems and user dissatisfaction. Advocates on each side tend to cite the cases that support their view, and the underlying evaluations are often not published in comparable form.
The practical implications for anyone attempting this
The pattern that recurs in successful projects is incremental. An application inventory comes first, classifying every piece of software as replaceable, replaceable with effort, or blocking. Document formats are standardised next — typically on ODF or PDF for archival and exchange — because format discipline is what makes a later application switch survivable.
Migrations then usually proceed layer by layer, most often starting with the office suite on the existing operating system, so that staff learn one new thing at a time. Pilot groups of willing users come before general rollout. Blocking applications are handled through virtualisation, web front-ends, negotiated replacements or, in some cases, a permanent minority of machines left on the original stack. Training and helpdesk capacity are funded as part of the project rather than absorbed afterwards, since the support spike is predictable and short-lived.
The failure mode is equally consistent: an announcement with a fixed deadline, insufficient inventory work, no plan for the blocking applications, and a support function that discovers the change at the same time as users.
What to watch next
For this specific report, the things worth waiting for are primary sources: an official statement or procurement notice from the body concerned, which would confirm the agency, the scope, the software selected and the timeframe. Until then, the number and the framing should be treated as unconfirmed.
More broadly, the signals that indicate whether this trend is durable are the publication of post-migration evaluations, whether pilot projects expand or quietly lapse, procurement rules that mandate open standards, and the response of incumbent suppliers in the form of sovereign or regionally hosted offerings. Reversals are as informative as rollouts, and both are worth reading in full rather than in summary.
Frequently asked questions
Is Switzerland’s government moving all its computers off Microsoft?
There is no verifiable basis for that broader claim. The circulating report concerns a specific body and a fleet of roughly 3,000 computers, which would be a fraction of the total public-sector estate. National governments generally run many independent IT organisations with separate procurement, so a decision in one body does not commit others. The precise scope has not been confirmed from primary documentation.
What software do public bodies usually move to?
Common destinations include Linux distributions with long-term support for the desktop, LibreOffice or OnlyOffice for documents, and open-source collaboration platforms such as Nextcloud for file sharing, alongside self-hosted mail and calendaring. Identity management is often rebuilt on directory services that support open protocols. The exact combination varies, and many organisations replace only the office suite while keeping the existing operating system.
Does switching to open-source software actually save money?
Licence costs fall, but that is not the whole calculation. Migration requires inventory work, testing, integration engineering, retraining and a period of dual support, and specialist applications may need commissioned replacements. Whether the net effect is a saving depends largely on how much bespoke or vendor-certified software the organisation runs. Independent, comparable evaluations of completed migrations are scarce, which is why the cost question remains contested.
How long does a migration of this size take?
Public-sector desktop migrations of a few thousand machines are typically described in terms of years rather than months. Time is consumed by application inventory, pilot phases, format standardisation and the resolution of blocking applications, not by the software installation itself. Timelines also slip when specialist departments discover dependencies late. Any specific timetable for the reported Swiss project is not known.
What is digital sovereignty in this context?
It refers to a state’s ability to control the software, data and infrastructure used by its public administration, including where data is stored, who can access it and whether the code can be inspected and modified. In practice it drives preferences for open-source software, domestic or regional hosting, and open standards. It is a policy goal rather than a technical specification, and its interpretation varies between administrations.
Have similar migrations failed before?
Yes, and that is part of why the topic recurs. Several European public bodies have run large open-source desktop programmes, and at least some have subsequently reversed course, citing interoperability difficulties, application compatibility and user dissatisfaction. Others have continued for many years. The published record is uneven and the evaluations are not always comparable, so neither outcome should be treated as the default expectation.
Sources and further reading
- Official Swiss federal administration publications and procurement portals, which would carry any confirmed decision, scope and timetable.
- European data-protection authorities, which have published assessments of cloud productivity suites in public-sector use.
- The Document Foundation and other open-source project organisations, for documentation on office suite migration and open document formats.
- Established technology trade press covering public-sector IT procurement, useful for tracking migrations and reversals over time.
Surfaced from the reddit:technology signal “government open-source desktop migration”. AI-assisted draft, editorially reviewed.
