Article up to date as of 10 November 2026; the situation is evolving. The facts below are those reported by the press and the public sources cited, and nothing more.
If you publish software built with WinDev and sell it to customers, you have probably seen, since June 2026, the concern among WinDev developers about a proposed fee per client workstation. This article sums up what is known, what is not, and the options open to a software vendor, from the most cautious to the most radical.
What is known, and what is not
The facts, stated with the caution they deserve:
- Since June 2026, a proposed fee per installation at the end customer has been discussed in the WinDev community. According to exchanges reported by the press, the amount mentioned would be around €250 per workstation. As of the Next INpact article of 10 June 2026, PC SOFT had not officially confirmed this project. Source: Next INpact (in French).
- PC SOFT was acquired in November 2024 by a fund linked to the Canadian group Constellation Software.
- French MP Philippe Latombe filed a written question (no. 16261) on the subject, published in the Journal officiel of 23 June 2026. Source: questions.assemblee-nationale.fr.
- The Wx Alliance association launched a petition.
What is not known: whether the fee will be introduced, in what form, on what date and with what exceptions. Everything that follows therefore reasons about a risk, not a certainty.
Why a vendor is more exposed than an in-house user
A company using a WinDev application internally counts its workstations. A software vendor multiplies its installations by the number of its customers, without always controlling the count, and must pass on or absorb any new cost within a selling price already negotiated. A per-workstation fee, if it came, would hit its margin and its business model directly, not just its IT budget.
That is why the question deserves a structured answer, even while the project remains a rumour.
Option 1: wait
Waiting for PC SOFT to make a statement is a legitimate option, and it is the one most vendors are taking today.
- Advantage: no investment committed on a hypothesis.
- Risk: if the fee is confirmed with a short notice period, you will be subject to the vendor’s calendar instead of choosing your own. A migration is not decided in three months.
- Effort: none, provided you use this time for the one thing that costs nothing and serves in every scenario: mapping your application and your database.
Option 2: freeze the version
Stay on the WinDev version you own, without subscribing to the next ones.
- Advantage: controlled short-term cost, no disruption for your customers.
- Risks: growing incompatibilities with operating systems, drivers and libraries; increasing difficulty to recruit; no answer to the demand for web or mobile. And freezing does not necessarily protect against a fee applied to existing installations, a point that remains unknown.
- Effort: low, but it is a debt you postpone, not a decision that removes it.
Option 3: migrate progressively
Run the WinDev application and its replacement side by side, and move the scope module by module: data first, then headless processes, then screens. This is the method I describe in Migrating a WinDev application to .NET without regression.
- Advantages: no blank day for your customers; each step produces evidence and can be delivered; risk is spread over time; you regain control of the calendar.
- Risks: a coexistence period to manage (two applications, two databases, synchronisations); a temptation never to finish if steering slackens.
- Effort: from several months to more than a year depending on the size of the application, but spread out, and you can pause or slow down at each milestone.
For a vendor, this option has an extra merit: the new platform can be designed from the start for the web or mobile access your customers are asking for.
Option 4: rewrite
Start from a blank page on another technology stack, delivering a complete new version.
- Advantage: a coherent architecture, without coexistence compromises.
- Risks: the longest and most uncertain project; a single switch for all your customers; business rules forgotten because they were never read anywhere. It is the option that produces the most regressions.
- Effort: the highest, and concentrated at the end, when nobody can yet see a benefit.
How to choose
The four options are not mutually exclusive: you can wait while mapping, freeze while preparing a progressive migration. In practice, most vendors end up combining them over time: a freeze to stabilise costs this year, a mapping to know where they stand, then a progressive migration started on the module that hurts most, often the one customers want to reach from a browser.
What matters is the order of decisions. Mapping comes first because it costs little and informs everything else. The choice between freezing and migrating comes second, once the exposure is known. The rewrite from scratch should only be considered when the mapping shows that almost nothing of the existing application deserves to be kept, which is rarer than it seems. What separates a vendor at ease from a vendor under pressure is not having migrated already, it is knowing its exposure precisely: number of installations, runtime dependencies, size of the HFSQL database, modules actually used, critical business rules.
That knowledge fits in a short remote audit, whose scope you will find on the services page. It gives you a roadmap and comparable scenarios, whether the fee is one day confirmed or not.
To go further: the WinDev migration page gathers the six-step method, the figures from completed projects and a man-day estimation simulator.