Leaving Microsoft Takes More Than Installing Linux

Key takeaways

• NixOS defines system configuration in code, making shared setups easier to reproduce and review.
• Replacing Windows does not resolve compatibility with documents, business applications, or collaboration services.
• Open source still needs someone responsible for security updates and outage recovery.
• Digital sovereignty depends on being able to change suppliers while keeping public services running.

Imagine a Dutch government office moving from Windows to NixOS to reduce its dependence on Microsoft. It wants more control over its software and less exposure to a vendor’s pricing and support decisions. Then a front-desk PC freezes at 9 a.m. on Monday, and the immediate question becomes who can get it working again.

A configuration file is a useful starting point

NixOS is a Linux distribution with a distinctive approach to managing computers: describe the software and system settings you want in code, then build the environment from that description.

Instead of asking an administrator to click through settings screens and remember every adjustment, you keep a written specification of the machine. That makes it easier to apply a common setup across an agency and track changes over time.

It also makes staff turnover less painful. The next administrator has a record of how the system is configured, rather than a collection of guesses about what the previous person did.

NixOS can also roll back to an earlier system configuration. But configuration rollback and data recovery are different jobs. Reverting an update will not automatically restore a document someone edited incorrectly afterward.

Backups still earn their keep.

Open source does not come with an on-call team

Access to source code gives an organization the ability to inspect and modify its software. It does not put someone on duty when a service fails overnight.

A public agency considering NixOS would need to assign clear responsibilities. One possible arrangement would be:

• An internal team defines operational requirements and security standards.
• An operating contractor tests updates, deploys them, and handles recovery.
• A specialist support provider investigates problems the operating team cannot resolve.

One organization could cover several of those roles. What matters is that nobody has to negotiate ownership of an incident while a public service is unavailable.

Support contracts also need to distinguish response time from recovery time. A promise to call back within an hour says very little about when staff can resume work.

The agreement should cover emergency patches, recovery exercises, and the eventual handoff to another provider. A community forum can be valuable, but voluntary help cannot carry an agency’s contractual obligation to restore service.

The migration bill may be hiding in a spreadsheet

Replacing an operating system sounds like a neatly bounded project. Replacing everything employees depend on during a working day is considerably messier.

Suppose a department uses Excel macros to produce its monthly statistics. Opening the spreadsheet in the new environment is only the first test. The calculations and automated steps must produce the same results.

If they do not, the file has moved. The workflow has not.

Approval systems, authentication certificates, printers, and browser-dependent internal applications need their own checks. NixOS does not, by itself, replace an office suite, email platform, or collaboration service.

That changes the cost comparison. Licensing is one line item alongside document conversion, application changes, staff training, and help-desk support. Running the old and new environments together during a transition can add costs too.

Lower licensing costs could be a benefit. Whether the agency spends less overall depends on compatibility work and the ongoing cost of operating the replacement.

The real test is whether another supplier can take over

An agency could reduce its Microsoft dependence and still end up with very little room to maneuver.

If only one contractor understands the new environment, that contractor becomes a dependency of its own. A price increase or business closure can leave the agency scrambling again.

NixOS configuration files can help make a handoff possible. They cannot explain every decision on their own. The next team also needs to know why settings were chosen, how deployments work, and what to do when they fail.

The agency needs control of its configuration repository and deployment procedures. A replacement provider should be able to use the documentation to rebuild a test environment.

Better still, rehearse that handoff. Watching another team reconstruct the system tells you more than being assured that the documentation is comprehensive.

Success should also be measured beyond the number of computers migrated. Can employees complete representative tasks? How long does it take to deploy a security update? How quickly can work resume after an outage?

For this hypothetical Dutch migration, digital sovereignty would mean having practical choices about who supplies and operates government systems. NixOS could support that goal, but compatibility, support responsibilities, and a tested handoff would determine whether it holds up. The decisive test is whether the next team can keep services running when the current supplier leaves.