tl;dr - There have been some recent updates regarding changes in the Horizon Europe Program plans for 2025. This would directly and severely impact the NGI0 consortium and NLnet in their ability to further fund open source projects. To be clear, existing NGI contracts for NixOS are not affected and therefore this isn’t an emergency. But budget termination or even a significant budget cut in the NGI program would have repercussions for Nix and stifle the entire FOSS ecosystem in the longer term.
I’m probably saying things that we all know too well, but I feel the need to emphasize them. Open source has driven, is driving, and will continue to drive some of the biggest waves in software history. Therefore this pinches me a bit more to hear that there are plans to divert a substantial portion of public open source infrastructure investments into AI. It’s wild to me that sometimes I still feel like the enormous contributions being made aren’t felt and understood by everyone in and out of the open source ecosystems.
Getting up to speed
Recent discussions with NLnet have reassured us that our current funding based existing contracts is safe, and there is even some buffer. For now, it is business as usual. However, such a buffer won’t last forever, and we should be concerned about the fact that current EU discussions envision a substantial decrease in support for open-source infrastructure initiatives under the new work package for Horizon Europe 2025. This shift could significantly affect the long-term EU support of projects like ours. Concretely, if the draft gets accepted, it would reduce investments into the broader open source ecosystem by two thirds from 27 million EUR in 2023 to 10 million EUR in 2025, and make it much more inefficient to work with.
How we are currently getting involved
We are currently working with a few different groups and individuals to understand how to best support the effort of preventing these plans from coming to fruition, and help change the course to the better. We are also collaborating with other open source projects, organizations, and individuals to try and establish a unified front to emphasize the importance of FOSS. We’ve also joined and posted the letter on the NixOS blog which can be viewed here - Blog | Nix & NixOS.
How to get involved
If you’re interested:
Respond to or engage with social media outreach on the topic
Reach out to your local EU representatives to express the importance of continuing the NGI funding as part of Horizon Europe. (Happy to help write a template if there isn’t one yet)
From what I understand, they can appeal towards the general case and have an additional positive impact on it. The most critical piece is that we formulate a strong message with wide backing for our case for when they start making the 2026 and beyond plans.
One strong message I’d like to see shared is the value provided by ecosystem size and network effects. What we see with NixOS and all the projects NLnet enabled is that we not only have a variety of small and mid-size applications that developers – including professionals in enterprises – can use as building blocks or for reducing operational costs, but that the program managed to build an active community of contributors with substantial mind share, awareness of certain issues, and the skill set for global collaboration. Quite precisely that corresponds to the topics the NGI0 consortium is built around: reproducibility, accessibility, internationalisation, free licensing, participation, and, most importantly, innovation and sharing of ideas.
You just can’t get that by betting everything on one big project.
Arguably, NGI0 is at least a modest success that’s still compounding, and the numbers seem to support that. Cutting it down before reaping the full benefits seems to be a substantial risk of wasting a lot of money.
divert a substantial portion of public open source infrastructure investments into AI
I’m uncertain about what exactly is meant by AI in this context. Does it include open infrastructure for AI? If so, projects like NixOS could be equally eligible for such investments. It’s common for grant money to be reallocated according to the prevailing political debates of the time. In the past decade, a major debate in the EU triggered by events we all know—and consequently a key focus of NGI—was on promoting data protection and sovereignty. While this issue remains unresolved, new events and new concerns triggered other debates and will likely entail a different focus of funding. The key question is whether NixOS can contribute anything valuable within this new area of focus. I believe the answer is an easy Yes.
Another aspect to consider is how the money is invested—whether it can be distributed through cascading calls to individual contributors and directly to open-source ecosystems, or if the EU prefers direct subsidies to EU-owned businesses or projects. Even in the latter case, there might still be a significant open-source component involved (as seen in the “digital commons” call in France’s AI program).
Admittedly, much of this is speculative at this point (maybe not for someone with more information). While it is important for the NixOS community to be vocal about what it considers effective and valuable, it is also wise to be prepared (as I’m sure NLNet is) to adapt to the changing priorities of the time.
, no wonder they’d get to horrible decisions like this