A decade ago, the European Commission initiated an ambitious effort called Next Generation Internet (NGI) to develop a more human-centric internet that would increase digital sovereignty. The first actual programmes within that initiative were launched in 2018, offering a widely acclaimed combination of targeted funding and in-depth support for digital commons – from security scans and accessibility audits to community building, copyright compliance, user experience (UX) design and technical writers (including the first two NGI Zero programmes, NGI0 PET and NGI0 Discovery).
Meanwhile, subsequent incarnations of NGI Zero and some sibling programmes like NGI Taler, NGI Fediversity and NGI Assure have awarded over 1,500 grants to projects around the world – facilitating over a million hours of development of quality free and open source software in several domains. As part of these efforts, with funding from the European Commission, NGI Zero has been supporting free software, open data, open hardware and open standards projects, such as the ones featured by APC, a member of the NGI Zero consortium led by the NLnet Foundation, in the interview series “Building a Free Internet of the Future.”
After nearly 10 years of uninterrupted implementation of back-to-back NGI calls, in June this year the NLnet Foundation announced a temporary pause on its open calls under the initiative, and in particular the NGI Zero programmes, to prepare for a transition. The European Commission is now aiming for new horizons with a new initiative called Open Internet Stack (OIS), which is a partial continuation of the work done within the NGI initiative – but with more weight on the ability to deploy these “next generation” open technologies at scale. OIS is aimed at providing mature and trustworthy open technical building blocks on which Europe can rely, from chips, mobile operating systems and browsers to end-user services.
In this interview, Michiel Leenaars, project lead of Restack and OIS Rebase as well as NGI Zero, talks about where NGI Zero is today and where this transition will take us tomorrow.
With a call for projects every two months, it is difficult for people outside the NGI0 consortium to imagine the sheer number of requests and applications. How did NLnet cope with this intensity and the years that went with it?
I think it is safe to say it has been somewhat of a demolition run for us, which was only possible because we felt our work resonating with so many people – and there has been a huge sense of urgency with everyone involved. When the Next Generation Internet initiative got started, it really wasn't too long after the Snowden revelations and in fact during the first presidency of Donald Trump, when the transatlantic relationship started showing cracks. And yet, this is what we all critically depended on.
What motivated us and has kept us going is an intrinsic appreciation and deep gratitude towards the awesome FOSS [free and open source software] developers, engineers and researchers out there that have a continuous flow of concrete ideas on how to change the world for the better – and yet pretty much no one is listening to them or willing to fund their work. There is something deeply motivating about empowering competent people that know what they are doing, and know they can make a difference in an area you yourself recognise as important. And when that happens at the scale we run, it feels magic at times.
What has been intense about it, is that when you commit to running such a programme, that involves a whole coalition of partners and recruits an army of projects, it behaves like a tanker. It creates vast momentum and you feel the responsibility of this collective dependency. We are a small foundation, and the opportunities for this kind of programme are extremely rare. And when they come, they only come in big chunks. We all felt it worked, and so that is very rewarding. But you can't stop when you are exhausted after the first hundred projects, because doing so means there won't be another programme. And when you've completed 300 projects and finished the first two programmes you had committed to, the third one has already come out of the pipeline. As a team you don't allow yourself a break or take a holiday, because you know that would take valuable time away from our grantees – and you know they will need all the time they can get to successfully finish their projects. You see those hundreds of people working so hard every day and you see the difference it makes, so you keep going. Over time we were able to slowly grow our team, but one needs a specific kind of person to do our work. And demand continues to outpace us.
I imagine there must have been some hiccups, frustrations, unexpected events, oversights and delays. How did you get through those times?
Obviously, the SARS-CoV-2 pandemic sank a crater into the middle of NGI Zero. This impacted society at large a lot, but it impacted us more directly as well. We doubled down to help the European Commission set up an emergency facility to support the development of FOSS technical solutions in the same vein as NGI. And at the same time there was a lot of pressure to speed up the development of tools to mitigate the lockdowns, such as videoconferencing tools. This double load weighed on the team in an unhealthy manner, but it was of course the logical thing to do given the circumstances. Another thing that might be hard for people to grasp is how much of life really happens to our constituency when you run programmes at this scale. The projects we support are typically one year, but as it turns out, in that year all kinds of major life events can and will take place: people get seriously ill and even die; they have a divorce or get married; they have babies or need to take care of elderly people in their household; their social phobia or other mental health issues worsen and take over their life; co-founders turn out to have very different expectations about how to make the project viable, with fighting as a result; part-time employers fold, changing their economic parameters; they finally take the step to transition and have to deal with invasive medical procedures; they need to make sure they don't become cannon fodder in an unjust war, or have to deal with imprisonment by an anti-democratic government. And that is just from the top of my head. We try to help where we can, but this can be very challenging for all parties involved. Luckily, the European Commission (which leads the Open Internet Stack initiative) has been very supportive and the design of the programmes has given us enough flexibility to deal with any setbacks.
Overall, the biggest frustration is the observation that NGI Zero has been among the most productive funding programmes I'm aware of, delivering high quality in areas that really matter, is extremely cost-effective and fulfils a concrete need – and yet it is discontinued.
There is significant money available, but it is being spent on hypes and big tech, not on securing the basics. It seems many policy makers really can't tell what is good for them or for society.
There are some new programmes that will take over some of the roles of NGI Zero, such as those within the new Open Internet Stack. But these should co-exist rather than replace. The transition is lossy. There is currently no funding in sight yet for projects in some domains that NGI Zero was able to support, such as open data, security research and open education.
Behind NGI0 are a dozen diverse organizsations in many countries, as well as regional representatives (individuals), who carry out a variety of activities and missions. What they do is rarely well understood. Could you explain it?
Indeed, we are very proud to work with a coalition of trusted long-term partners that help our projects – this has really been a major factor in the success of NGI Zero. We had learned throughout the years that even the most talented developers can't do everything by themselves, and appreciate some help. The demands of modern technology at internet scale are just so big, and even worse: continuously changing and expanding.
Our amazing partners are available on demand to complement the skills of the projects when they are needed, whether this involves technical or legal skills, community building expertise, improving UX or helping to create a welcoming and diverse ecosystem.
There are many best practices to follow, and so little time. If I just mention legal issues like copyright compliance (which can easily get messy when having to deal with inherited code or dependencies of doubtful provenance), accessibility requirements, cryptographic export controls, localisation and internationalisation, the need to sometimes engage with standards bodies, packaging for a variety of platforms, setting up continuous integration and testing. Only superhumans can deal with all of that and at the same time keep the timelines of their project – and that is just technical stuff. If you are already very busy churning out running code, it isn't always easy to at the same time focus on building and growing a community, resolve issues among contributors, and so on. And maybe you want to set up a business, or properly set up governance. And just maybe, you need to tackle all of them when your project hits the real world and everyone jumps on it – but you can't do everything all at once.
So in all such cases, it is really enormously valuable to have qualified experts ready to help out. In some areas you really should always involve others to avoid having blind spots, for instance, to help harden security or to improve technical or user-facing documentation.
Of course it is pure luxury to be able to offer those to projects on top of grants, so projects don't have to decide between a security audit, an accessibility scan or more functionality in their project. Not everyone will need every support service, but because we have significant numbers we can afford to have dedicated expertise available nonetheless. People might not appreciate how special it is that very competent people choose to serve others, instead of seeking their own fortune levering those skills. Our partners believe in the vision of these programmes as much as we don, and that is priceless. People might not appreciate how special it is that very competent people choose to serve others, instead of seeking their own fortune by leveraging those skills.
The regional representatives are an interesting innovation, a welcome addition to the original programmes made possible because the European Commission has provided us with extra budget for this after the programmes started through so called "HopOn" grants. In a dozen countries we have a network of what you might call in layman terms "ambassadors", though that probably doesn't do them enough justice. These are community-oriented people that don't deliver services to individual projects like our organisational partners, but instead they are individuals that bridge our programmes with local communities at regional or country level.
Our projects wouldn't be so awesome if we only catered to the usual suspects, that already know where and how to get money. An overwhelming majority of people are unaware of the funding we can give to free and open source projects – as if we were hidden in plain sight. Especially younger people might just not find us, even though that is very much a target audience.
That is where our regional reps come in: they talk with potential grantees at events they attend or organise, engage with local governments on the potential of digital commons, and feed the public debate where possible. And they can do that in Polish, Turkish or Portuguese, and with an understanding of the local ecosystem. This helps to reduce language barriers and allows us to better serve the needs of specific regions than we could ever do without boots on the ground.
The regional representatives also help with outreach and promotion. We have quite a unique portfolio that covers many different areas of technology and offers concrete alternatives, but of course people need to know before they can switch.
The web is big, and it is easy to get lost. We don't have marketing budgets and don't advertise, nor typically do the projects we fund – unless they find it elsewhere. So besides helping haul in new talent, the regional representatives help spread the word and make strategic connections to potential key users in their area.
And of course that also feeds back into new innovation, because people can take those building blocks and improve upon them in new projects.
NGI Zero received funding under the European Union’s “Horizon 2020” and “Horizon Europe” research and innovation programmes. This meant that NGI Zero had a limited lifespan, subject to the decisions of the European Commission. Free and open source software groups had called for the continuation of NGI Zero, and several European newspapers had published articles on this issue. Today marks the start of the transition from NGI to the Open Internet Stack (a new programme from the European Commission). What will continue and what will change?
At the basis of NGI Zero was the foundational study for Next Generation Internet that NLnet and Gartner Europe conducted for the European Commission in 2017. That gave a well-defined vision and analytical frame to work with. Obviously, Open Internet Stack doesn't yet have its own clear dot (or dots) on the horizon – so until it does we'll keep the same vision as guidance for now, because it serves us well.
From our perspective, the core challenges and ideas for how to go about our work have not changed, and are very much compatible with the "DNA" of NLnet. We don't engage with just any third party, only by choice. NGI Zero was about digital sovereignty and building a strong commons from the very start. And the mission isn't done yet by any means.
Open Internet Stack is still shaping up and its final shape and form will take a while to materialise. Like the Next Generation Internet initiative, OIS isn't a single programme – it will be a number of independent programmes, run by a variety of organisations doing different things.
The contours in terms of funding are a bit different though. At the moment, there are two grant-making programmes in the same vein as NGI Zero, coordinated by NLnet: Restack and Open Internet Stack (OIS) Rebase.
There are six other Open Internet Stack programmes where the main weight isn't on grant-making, but lies with the work done by a group of collaborating FOSS projects developing concrete open internet stack components. I tend to call these “scale-up programmes”. Some of those (like CodeSupply and ELFA) are coordinated by NLnet and involve a smaller part of grant-making, others involve ad hoc consortia with sometimes tens of FOSS projects as direct beneficiaries. We are happy to see the vast majority of confirmed programmes involving former NGI grantee projects from our portfolio.
The next stage in pupation will probably have to wait for 2028, when the European Commission has moved to a new Multi-Annual Framework. That is still a long time away, and it is not yet known at this point what the next funding cycle will look like or who will be driving those efforts – let alone what will be their angle and who they target.
Together those knowns and unknowns will shape the Open Internet Stack. So that is very much a story still being written, I cannot give the final word on that just yet.
As the name implies (if you interpret the word stack as "a pile of technology components that together power a set of services or applications"), OIS is expected to be a bit more restrictive in scope than "all of digital commons" (which for instance was the remit of the NGI Zero Commons Fund). For instance, open data efforts like geospatial information systems or Open Food Facts aren't really technology components. Obviously, there isn't just one internet stack – there are server stacks, desktop computer/workstation stacks, mobile stacks, etc.
That leaves a lot of ground to cover still for OIS. Besides the further development of concrete user-facing applications like videoconferencing tools, improvements to operating systems, firmware and network protocols (including standardisation work), instant messaging solutions, email, browsers and many other things we've been funding in NGI Zero will remain in scope. A quick rule of thumb could be whether you can use the functionality output over a network (it doesn't really matter whether that is in a local-first manner or through a classic server-client model, or something else). But that isn't the entire story: projects could also improve the quality, availability, security or functioning of one or more of the open internet stacks available.
So the main takeaway is that the understanding of what is Open Internet Stack is going to evolve over time, also based on learnings at our end and the end of the European Commission (as the driver behind this). Obviously, they have just published an ambitious open source strategy, and no doubt there is more in the pipeline.
Meanwhile, we're actively looking from our end to find alternative sources of funding to continue the excellent work on, for instance, open data, open healthcare and open hardware. We are about to announce an ambitious fund on geospatial commons, thanks to a new dedicated grant from the European Commission on that topic. But for other topics, we are exploring every possible option – this work is just too important and strategic to continue unfunded. Most traditional funding instrumentation is badly designed to support digital commons, and I think we have proven with NGI Zero just how much societal value there is to be had. I’m optimistic that there are many people that want to see this work continued, and then it is only a matter of chance for a few of them to be able to steer budget towards our community.
Xavier Coadic is a consultant for the NGI0 consortium, and a free/libre open source software activist with 15 years of experience in free open source cultures and communities (software, data hardware, wetware, policy makers and political groups, research and development).