The European Union must keep funding free software
pad.public.catWell worth reflecting that relatively little of existing free software seems to come from that sort of funding. It is mostly commercial companies following commoditise-the-complement strategies, consultancies building free tools then supporting them, or research spinoffs.
If EU funding can create and sustain a thriving free software ecosystem I suspect it would require a new funding model that we haven't seen yet. Otherwise, why would they do it and how would the institution detect if it was working?
Some projects funded by NLnet, https://nlnet.nl/project/
ActivityPub <> Discourse, WordPress bcachefs CryptPad F-Droid Jabber/XMPP Jitsi LibreOffice Matrix Mastodon NextCloud PulseAudio QubesOS SourceHutMy project Renderling (https://github.com/schell/renderling) is funded by NLNet and I can say the whole process has been great. It’s changed the way I think about open source and gives me motivation to continue.
That list seems a bit lacking substance and slightly off target. NLNet doesn't claim to fund libreoffice for example - that was a Sun office suite, picked up by the Apache foundation and now controlled by the Document Foundation (mainly through general donations if I understood their financial reports). NLNet seems to be funding things like "encrypted collaborative editing in the browser [using libreoffice]" which is a different kettle of fish. Nice option to have, but fairly niche functionality in the LO suite as far as I know.
And how does all this funding compare to something like the Google's Chrome & Firefox? That is one company controlling the majority of web traffic through OSS that on balance respects user freedom. To me, that is a better funding model with better results.
Funding use cases may well be a good use of time, but given the serious issues the EU has establishing itself in the tech industry, the military and economic crisis they have managed to waddle in to and the general political turmoil that seems to have kicked up I opine it is not the time to be wasting political bandwidth like this note calls for. The US model of letting companies fund and build software (including free software) seems a bit stronger, more flexible and politically easier to coordinate.
We don't need all of Europe to come together and work out who they think is the best team to build web technologies. Google puts a team on it and it probably happens. There are worse ideas out there than calls for government funding but it just doesn't sound effective to me - these continent-spanning governance bodies don't have the bandwidth to pull off this sort of delicate technical work.
NLnet funds tactical feature improvements in selected OSS projects, not Google-scale underwriting of employee teams.
> don't have the bandwidth to pull off this sort of delicate technical work
What's the basis for this claim? There are literally 1,000 technical work engagements listed above, scoped in detail and reviewed by a few humans (not "all of Europe") at NLnet, with a track record of successful delivery to upstream OSS projects.
> What's the basis for this claim?
The title of the piece is "Open Letter to the European Commission".
The EU commission has more important things to focus on than this low impact slush fund aspect of NLnet. Given the lack of impressive successes that you can point at, maybe there is an argument that they are and that NLnet just isn't doing very well. I dunno. If they could manage to get a Google-scale company to take root in Europe it'd achieve a lot more.
Are you comparing nonprofit project budgets of a few thousand euros to corporate VC/investment budgets that are orders of magnitudes larger?
> the lack of impressive success
Successful delivery of hundreds of projects doesn't count? Improvements to widely used office and communication software? Improving critical infrastructure used by journalists and other members of civil society? Mentoring the next generation of developers who can staff future EU technology companies?
What type of tech projects, budget and success do you expect the EU commission to fund? Are there previous successes that could be emulated, from the EU or other governments? Google was created by private investors, as was Skype.
> What type of tech projects, budget and success do you expect the EU commission to fund? Are there previous successes that could be emulated, from the EU or other governments? Google was created by private investors, as was Skype.
I'm not sure why you think the EU should be wasting bandwidth on this given that you are (accurately) identifying that it isn't managing to marshal much capital or following a model that leads to notable successes. The focus should be on incubating more successful strategies. The way to get good OSS ecosystems is a web of companies making money off open source software. That is how we get things like Linux, Chrome or Postgres.
> Successful delivery of hundreds of projects doesn't count?
Not really, no. Especially if the projects are already being funded by a government program - give me a budget and I can fund a thousand github projects. When you tried to list successes, you put together a list with names like XMPP or WordPress on it. And don't misunderstand me, I like XMPP and Wordpress; but they are not strategically useful to a body like the EU.
> Are you comparing nonprofit project budgets of a few thousand euros to corporate VC/investment budgets that are orders of magnitudes larger?
Yep. And it is unimpressive. It doesn't look like a close race to me, efforts via NLnet are feeble compared to the US approach, where they pour in orders of magnitude more effort and get much bigger payoffs.
The EU has a problem - no companies. They should solve that problem, not make random microdonations - there is no particular evidence that is powerful, at least not that I've seen. There is evidence that corporations are powerful at producing OSS though.
> The EU has a problem - no companies. They should solve that problem
This thread is about Horizon Europe, which has a specific charter.
If you want to solve a different problem, define a new investment program and lobby for it to be funded.
nlnet funding is cents in comarison to the cost of these. They give grants, but they don't cover all the costs.
Yes, and now even that tiny pool of funding is going to be cut.
Nyxt browser had also been funded by E.U. at some point, not sure if they still do.
Source?
"FOSS funding vanishes from EU's 2025 Horizon program plans", 20 comments, https://news.ycombinator.com/item?id=41002044
My town had a budget vote not to long ago.
I actually read the budget proposals and look at the year over year breakdown for certain categories.
The amount my town is paying for microsoft365 and GIS software is substantial. While not an eyewatering amount, and I can understand that certain things need to be licensed....its hard to understand why my local town isn't using a FOSS GIS and just setting up a basic proxmox cluster with HA failover to host the darn thing.
Cuz when things need tweaking or break, there is usually no one to call. FOSS solutions are fine when there is some internal evangelist or team pushing it and are willing to support and maintain things (which over time usually comes out of some labor budget). If there is no one like that internally, even very simple issues can put people off using these systems.
Because paying people who know how to do that properly are pretty expensive? The average local MSP is paying $25/hr for people who are either early career or flunked out of college or were too incompetent for a corporate job at a larger employer. AD/O365 is pretty easy to get going on a reasonable scale without much
It would be interesting to see what they got out of the first €27 million they put into it.
PyPy received an EU grant sometime in the late 2000s or early 2010s, although it had to be a different one. If I understood TFA correctly, NLnet is partially funded by this money? NLnet itself distributes money to very cool software things[1]—some moonshots such as Spectrum[2], some very down-to-earth and usable ones such as Nitrokey[3]. A discussion[4] linked in a neighbouring comment also mentions Sourcehut and Bcachefs.
Thank you. I was looking for something like a table with a project, maybe a summary, amount funded, KPIs, and ideally an assessment of the results.
this is a recent (1-2 months old) report on the impact of the NGI0 programme:
https://op.europa.eu/en/publication-detail/-/publication/257...
there was also a fosdem talk about this, if you prefer AV to text, finding it i leave as an exercise to the interested parties...
They seem to have prioritized making the process simple for developers by removing overhead. Apparently you are don't have to write lots of reports about your results for them, they just check that the code you said you'd write is in fact written and runs.
I am not sure what to make of the EU’s agenda on free software. On the one hand, we see provinces of some countries announcing open source adoption strategies. On the other hand, we see controversial regulation on things like AI that basically proposes to pull up the ladder and give the market to incumbents.
It's ridiculously short sighted to cancel this stuff. NLnet in particular is by far one of the most impressive & broadspanning places for innovation & development in the software world in general, and vanguard for open source at large. The benefits are so multi-fold.
Half the projects go no where, a quarter have some meaningful output with varying levels of adoption, and a quarter are amazing & loved bits of work.
Theres folks asking if we are getting out moneys worth. First, $27m just doesn't seem like all that much. Especially for such a broad array of nations. Second, the hit ratio is never going to be perfect; worrying about how every dollar is spent will insure you never ever have good research & development, will insure only lukewarm mild takes get tried. You have to be willing to fund adventerous stuff; that's what VCs for example bank on. Because it's impossible to tell, because trying to be smart ahead of time breeds mediocrity. And this r&d&maintenance funding: it buys not only the fruits of that labor, but it maintains a dynamic & empowered culture across the region. It curates talent & starts efforts & intensives the best parts of it's software world.
NLnet is so exceedingly important. There's so much possible good that humanity so rarely can get up to. NLNet has been one of core & best ways to promote making things better. This vision of human possibility is saintly & sovereign. I hope EU can keep it going.
I was thinking earlier today that self-driving automobiles would be a good use-case for a public-private venture for open source software development. My half-baked idea would be that car and computing companies could contribute their existing software to a open source consortium that would be government-sponsored, but privately run. Some fair valuation would be set for their contributions and the OSS license would require anyone using the software to contribute back their changes. Part of the funding mechanism would be that any self-driving vehicle would have a fairly high “tax” per vehicle (maybe $10K, 20K?) but that this could be prepaid by the manufacturer through software contributions. Yes, I know there are bottomless logistical issues to be worked out, but the idea being that perhaps, if everyone can see exactly what everyone else is doing we can maybe get out of the current pit of semi-working solutions that prevent true autonomous vehicles from being viable.
Or, it might end up that self-driving vehicles are an infrastructure problem and that they need dedicated roadways without incursion by non-self-driven cars. Perhaps we could set them up to run on dedicated schedules on electrified metal causeways with shared ridership.
Not sure if you're implying that the government is supposed to oversee/manage the development of such a project, but I think they're extraordinarily bad at such things. It isn't the role of the gov either.
> car and computing companies could contribute their existing software to a open source consortium that would be government-sponsored, but privately run.
So basically it would just be a normal company, but the government pays them? In what way should the gov decide which company to fund, and why? If it is a commercial/private company, their existence should be justified by their profits. I don't really get it.
In what way should the gov decide which company to fund, and why? If it is a commercial/private company, their existence should be justified by their profits.
It makes sense for the government to fund R&D that the private sector won't fund, possibly because it would be profitable for the overall economy but not for any one company. Obviously a dozen companies are already building self-driving so that doesn't apply.
Even they only use free software as a fall back, that is a worth while investment.
I wonder why the list of signatories at the bottom do not seem to include a single government (local or national).
Governments need to band together and collaborate to develop software.
But instead they tend to buy closed systems that are developed by the same companies churning out crud.
Ideally, governments are competent.
Realistically...competent governments are the exception. And competent-at-long-term-technology-policy governments are virtually non-existent.
I understand the current state but also think it’s possible to change.
We just gotta wine and dine some suits. Sounds easy.
No one ever took a government out to a fancy dinner and drinks to convince them to use open source.
Even if it would save governments money and help ensure civil servants get paid on time (phoenix payroll scandal)?
Never had a meal from a sales rep that swayed my mind
[dupe]
Some more discussion: https://news.ycombinator.com/item?id=40970985
Money is a limited resource and I don't think NGI has proven to be a good investment. They made the same sort of embarrassing mistakes your average VC fund would make like BS blockchain ideas. A lot of uninspiring "research" where some of the deliverables is some open source software that no one ever uses and for which development stops after funding runs out. There seems to be something rotten about the open source "community" funded by the EU. I never going to forget the scene of NLnet members distributing free t-shirts with their logo on it to random attendees of the CCC congres in Berlin and then taking a big group photo for some social proof they can put on their website. All in the name of free and open source software that respects EU democratic values, that is pro-social narcissism for you right there.
>I never going to forget the scene of NLnet members distributing free t-shirts with their logo on it to random attendees of the CCC congres in Berlin and then taking a big group photo for some social proof they can put on their website.
You can pretend that grant money doesn't fund the writing of grant applications, but that won't make it true. If you make your criteria stricter and less reliable, then every project has to spend more overhead on crossing the Ts and dotting the Is to ensure their entire budget doesn't fall out from under them.
As someone often on the receiving end of that money... It's complicated. My organization is doing a lot of great things and it's appreciated both nationally and internationally. Some of that work is funded from EU grants (including H2020 before).
But the problem with research is... it's hard to predict where it takes you. At some point you realize that the project you got funding for is a dead end, but you can do a really cool spinoff that becomes successful but is not technically the same project.
Another problem with research is that a lot of it is exploitation. You can't get a grant for just exploring random things and seeing what happens. But at the same time is a vital part of research - so official grants need to fund a lot of "underground" research, that may or may not mature into an official product. It's hard to estimate is a money is well spent by looking just at the official projects.
... And yes, many of projects are actually wasteful. Bureaucrats are really bad at estimating what counts as a valuable project and what is actually a sham moneygrab. In our case we are known for delivering so we often "win" when we start, but often random shell companies with the hype of the week come dangerously close.