Sponsored Open Source is already balkanized. Like Python Software Foundation can probably only sponsor US based maintainers and can not accept crypto. EU based funds can at least start the discussion on how can we support maintainers of Python infrastructure projects in other countries. Pave the way to debalkanize both supporting people and decision making processes.
On a practical side of things, I would actually add some transparency mechanisms on how the funds are really going to individual maintainers, like with OpenCollective, because EU funds tend to flow into consortia where the traces are usually lead to some company names, and public projects that were developed by those bitrot and fade away in a few years, even if they were open source. That did also happen with Google sponsored projects as well, so sponsorship without transparency is not sustainable, and transparency is required for the community to mature and replace “institutional generators” with community owned.
I just looked down to most recent few board resolutions on that link you posted.
It is reasonably easy for a US based open source foundation to employ in other countries. Indeed, the Drupal Association employed me in the UK for the last four years. They also now have people in Spain and other countries, too.
Also, remember that many (quite possibly most?) open source projects are run from countries other than the US. It is important to not fall into the trap that everything is run from there - it isn’t.
Firstly, thanks for sharing this. There’s a lot in here that I agree with and have been trying to build toward for years. There’s also a few things I’m not sure about (they appear to be taking a_pull_ model of requests for support, they appear also to have a degree of pre-selection based on criteria other than usage). But I’m reading through an nodding a fair bit.
My immediate question is, what’s next and how do we get involved? I’m thinking mainly about the identification of appropriate projects and technologies and on providing a flexible basis for funding individuals and small groups.
Agree. I believe what we are seeing is in this report, and I know of several other similar streams of effort in the same direction, is the natural evolution as open source plays a more important role in underpinning government and institutional digital infrastructure. The next few years are going to see the emergence of several public infrastructure funds focused on open source for the public interest. There will of course be earmarking based on the priorities of the funders, as there is now in trade association style OSS foundations. I think our job is to anticipate this and “pre-design” the systems that can facilitate the money flow. The stronger that a community’s governance is prior to the flow of public funds, I suggest the less that these funds can influence the road map and direction of the project. My hope is that public funding is more aligned with the ethos of open source than other available pathways of financially sustaining communities.
I am going to be running a working group this year through my work at Digital Impact Alliance at the UN Foundation on the topic: “Here is a $ Billion of public money for OSS, now what do we do with it”? If anyone is interested in participating, let me know. (note, at least for now the $Billion is a straw man intended to force us to think about how we could allow money flow to maintaining OSS at scale).
The core challenges I see are: 1. How do centralized institutions with formal contracting and reporting expectations give to decentralized communities. 2. Which projects get funded and what is the criteria. 3. Who governs at the meta and project levels. 3. Prioritization of features within projects. 4. Who does the work + how to volunteer and paid contributors work on the same project.
In my mind this really is the centralized old world trying to interact with the decentralized new world. What are the bridges we need to build between the two. There are many initiatives under way to direct public funds to open source at a greater magnitude, many will start to share what they are doing this year. We need to be ready when the funds do become available to say, “we can show you how to spend it”. I would like to see a world where more OSS projects are truly community and contributor governed building in the interest of all users, including governments. It is going to be fun to see how all of this plays out.
I am always interested to participate in such groups to speed up the progress. Especially if they are handled by UN experts, who are good at moderating people with strong opinions to get constructive outcomes. There is always a risk that theoretical discussion won’t give good results, because -
I won’t draw the conclusions why it happens, so I would just bring a use case in which I am involved. An existing issue from Python Package Index (PyPI), which is 3rd top voted that in short means that PyPI doesn’t provide dependency information, so you need to download and parse your packages recursively to find your dependency tree. PyPI received $300k from PSF sponsors, and yet my PR fixing this issue is laying here without review comments (like other PRs about broken development setup etc). So it is the story of the conflict between people with money and no time, and the person with no money and limited abilities to address more and more things that creep in.
As you may see just throwing money doesn’t solve all the problems. Although I admit that there is lack of funding on my side and no resources allocated to governance. And I am interested to test in practice how to solve roadblocks like this without corporate way of solving conflicts by firing people and finding others to fix it.
Encouraging companies to adopt Open Innovation. That could be penalizing monopolies over ideas (Taxes on patents and proprietary software), and rewarding them for liberating their innovations and software under Open Source licenses.
Goverment support for Demand-side funding programs: That means, empowering citizens to invest in Public Goods projects, emphasizing in FOSS projects.
Although it would be wonderful to establish a global public fund immediately, I expect we probably will go through a “national public funds to local open source ecosystem” phase.
Flexible/agile funding is crucial.
If we want to achieve scalability, our funding models should reflect the nature of software development.
The current models can be perfect for single and long-term projects, but they turn into bottlenecks for these new digital public goods (similar to the Waterfall vs. Agile struggles we had in the past).
It should be much more efficient to look past success (number of dependents, installations, etc.) instead of allocating resources by judging future promises.
In other words, in the long run, we should use those “metrics” not only to find the projects but to distribute the fund itself, with minimum paperwork and bureaucracy.
How to keep these public funds continuous?
The report mentions a lump sum of ten million euros for this fund, which would be an excellent start.
On the other hand, a limited budget cannot permanently fix this problem. Hence, eventually, we need to answer how to generate revenue for these funds to make them continuous.
On what to do next
@benjam Do you think Open Collective could play a role in distributing the funds since you already have a strong community and infrastructure?
@RichardLitt Probably you already thought about it, but it would be great to invite them to Sustain podcast?
I do. We are also seeking funding to reinstate and extend the work that Andrew Nesbitt and I did on Libraries.io, and to offer it as infrastructure to those working on this space. i.e to solve that big ‘open source usage database’ bit in the middle of a lot of the diagrams in the report.
I am going to be running a working group this year through my work at Digital Impact Alliance at the UN Foundation on the topic: “Here is a $ Billion of public money for OSS, now what do we do with it”? If anyone is interested in participating, let me know.
I’m interested, thanks for adding me (email@example.com).
That was too obvious, right? Thank you for your generous words, and you can imagine how happy I am to see these discussions.
One of the reasons I arrive at the “agile public funding” conclusion is the potential growth of the open source ecosystem, a detail that I assume most people overlook.
Once we crack the funding problem and there is a clear signal that there will be a solid, increasing investment in the open source market, I expect to see many new initiatives in this space. Most software developers already dream of building their solutions as open source.
Once these investments start, where do we think this transformation will stop within the software industry, one of the largest industries? How much time will it take to get there, and how much capital will we distribute through these funds? I don’t see anything static, minor, or temporary here.
That’s why, as you said, we can already start “pre-designing” these systems. What kind of principles should they have? Transparent, independent (above politics), predictable (don’t change money flows suddenly or dramatically), minimum bureaucracy, minimum intervention in the market (no double standards, laissez-faire), etc.
Again, thanks to everyone for these discussions/threads! I’m glad if I can contribute. But, more than that, I’d love to play an active role in executing these ideas
Tidelift have deprioritised work on Libraries.io so we’re going to fork, re-instate, rebrand and re-release under a different name.
Libraries.io was intended as a resource for the public but access has gradually been scaled back and services have been removed under their ownership, so we’re going to bring it back to the original vision.
Since we already discussed it under this thread, @heatharensen, allow me to share your FOSS Backstage presentation here, which I just finished watching.
It’s pretty exciting to see that we’re raising these questions. I hope the rest of the community will feel the same, join the conversations, collectively start designing these funding systems, and we can turn these ideas into reality.
What would we do with it if 1 Billion was pledged to support Digital Public Goods?
Let’s imagine that there is, and there might very well be very soon, as open source becomes much more critical to national governments and national strategies, that there might be funds of a large size, a scale that we’re not used to seeing in the open-source world. Will we be ready for it?
Here are my humble answers to the questions mentioned in the presentation.
How should a dedicated public fund distribute its money to an entire open-source ecosystem?
One of the core challenges is scalability. The open-source ecosystem and dependants constantly change. The developers regularly add new features, fix issues, release new versions, and potentially increase the value of their software. Or, if the software stops being used or maintained, its importance in the ecosystem becomes less.
If we want to distribute a certain amount of funds to such a dynamic ecosystem, we cannot do this by manual, static decisions. No one can precisely say, this project should receive x amount of money.
Therefore, the funding process, eventually, needs to be automated to avoid scalability issues. We need to study and develop a standard way to evaluate the value of each open source project and distribute the fund to the projects based on specific metrics.
Without getting too much into details, my answer is that the algorithm should be similar to OSSF’s Criticality Score, but combined with better “usage” data, hopefully through Software Bill of Materials initiatives, for example, CISA SBOM.
Bonus: I recently started to run a personal experiment precisely about this issue and now distributing a tiny fund into three open-source collectives each month based on their Criticality Score. As usual, any feedback is priceless
Once the money starts flowing, how can we keep a good balance between paid maintainers and volunteer contributors? Would introducing money harm these communities?
When projects start receiving money, we can expect that it will create some disruptions. Can we minimize such impacts? Might be. Should we worry about this too much? I’d say no.
The essential detail is that there are already open-source projects with paid employees while still having healthy volunteer communities. As we all know, the big tech companies also have quite a solid open-source projects portfolio:
In most cases, the core maintainers are the for-profit company’s employees with probably above the market average salaries. Yet, there’s still a strong community around the projects that keep making volunteer contributions. More than that, these are probably the healthiest open-source projects in the entire ecosystem.
In other words, once there’s money, it’s up to the core maintainers to manage the community and their expectations. Some will succeed at this stage, and some won’t.
Suppose we wish to guide maintainers in this area. In that case, we can reach out to these existing communities/projects and ask them to share their experiences with us and, therefore, other maintainers; what are the challenges we need to be aware of during such transition?
How can we finance decentralized communities?
Especially if we’re talking about funds equivalent to full-time salaries or more, we should ask core maintainers to set up a company for their open source project and pay directly to that company, not the individuals.
The rest should be similar to any other company; if there’s enough funding for more people to join, the maintainers can ask other volunteers to become employees in that company.
With this approach, the open-source maintainers/projects will not get special (legal) treatment. The path will also be the same for other software companies to join the funding program; open/share your technology, and start generating revenue for your contribution to the digital public goods.
One last detail, the funding structure we describe is similar to the online platforms in that the content creators get paid based on specific success metrics (e.g., YouTube, Spotify). Considering that they already have an experience in all the areas we need to deal with, it would be pretty helpful to get in touch with these platforms and see how we can collaborate in designing these systems for the Digital Public Goods.
I’m pretty curious about the upcoming working group on this topic, and I hope to have more extensive discussions around the public fund options and how to move forward