We are continuing Building a Free Internet of the Future, a monthly series of interviews with recipients and consortium members of NGI Zero (NGI0) grants. Funded by the European Commission, NGI0 supports free software, open data, free hardware and free standards projects. It provides financial and practical support in a myriad of forms, including mentoring, testing, security testing, accessibility, dissemination and more.
This month we are featuring the PeerTube project, a decentralised, peer-to-peer, distributed online video-sharing application that has been in existence for seven years. PeerTube received an NGI0 grant in 2023, along with other people who have developed features for the platform.
We spoke to Pouhiou, one of the co-directors of Framasoft (which developed PeerTube), who acts as a liaison between external parties who have an interest in the project and the development team around tasks and expectations. He is PeerTube’s product owner, focusing on the dissemination aspect of product development.
You have been working on the development and maintenance of this broadcasting tool, a video platform, for seven years. The things we create and the tools we use can help or harm humanity, as we embed our political values in them. What does PeerTube bring to people's daily lives?
I hate to be that guy, but (to us) "The People" doesn't exist. It is usually a trope or a placeholder to hide our own biases ("people like me"). That being said, in our digital space, PeerTube brings the demonstration that another internet has existed (I was privileged enough to enjoy it) and that we can still choose to make it happen. Nowadays, surveillance capitalism companies (Big Tech, etc.) have shaped the way we think of our digital tools. But we can develop another way of connecting with each other and being a society.
PeerTube is strong on very technical features and each implies very different values from those imposed by Google's YouTube or Amazon's Twitch.
When I tell you about free/libre licences, federated protocols, peer-to-peer plus classic broadcasting, voluntary redundancy, remote transcoding, optional content monetisation, etc. I am actually telling you that PeerTube has technically implemented, community-driven trust and autonomy; self-managed spaces that can choose to interact; resilience, volunteer solidarity, efficiency that can be shared; and sharing content without being shaped by attention-grabbing economic models.
Long story short: PeerTube empowers people (yes, them!) to create their video platforms, and be the Google of their own YouTube. PeerTube is not another video platform. PeerTube is a network of independent, self-managed and interconnectable platforms.
Criticism of tech, particularly Big Tech, is often very much based on a vision of a Western “counterculture”, as Payal Arora writes in From Pessimism to Promise: Lessons from the Global South on Designing Inclusive Tech. Are you also working on “decentralised” diversities of approaches and needs when you develop PeerTube?
Full disclosure: I haven't read Payal Arora's book yet, just a synthesis that really makes me want to dig deeper. But yes, our Northern/Western counterculture is full of our own biases and at Framasoft, we are no exception. Such a book is really important and eye-opening.
From what I've gathered, part of the book questions the right-or-wrong moralism that can happen in all activist communities, and that we encountered in FOSS activism. At Framasoft we tend to distance ourselves from binary views such as "Use GNU/Linux or GTFO", "If you use Google you're doomed/wrong/a bad human", etc.
Obviously I'm exaggerating, but in reality we have had to send a public service announcement (PSA, in French here) to enthusiasts who were aggressively recommending PeerTube to content creators without taking into account their needs, use cases, aspirations, etc. Behaving as "PeerTube's witnesses" isn't helpful to anyone.
The point here is that we all need to stay humble and listen to people we aim to serve. We try to do that with PeerTube, with issues, our feedback tool, our forum, our social media. Several features that mainly serve non-Northern/Western users have been added this way, for example with the 144p resolution (useful in countries with slow bandwidth infrastructure).
And even with all that, we know that our work won't cover all needs. If you live in an under-represented country and need to raise awareness on a very popular social network, PeerTube (whose goal is not to grab attention and make content viral) is not the tool for you. And that's OK: we have limited means so we must know the limits of our work.
Over the past seven years, have you seen any changes in behaviour and/or practices in broadcasting via PeerTube?
Yes, lots. First, it was mainly a tool that tech-savvy FOSS enthusiasts were happy to toy with. Then it became popular among content creators that wanted a self-hosted mirror of their YouTube/Twitch channels; and among communities who wanted to create and regulate their safe space (deaf people, queer people, etc.)
Unfortunately, we have also had far right extremists and propagandists, banned from mainstream platforms, who started to use PeerTube and wanted to spread their filth in the federation. Obviously, they are free to use our software, but we have made clear that we have and will keep on working to give the federation tools to isolate them in their hate bubble (and studies show it mainly works).
Nowadays, PeerTube is experiencing increasing success among content creators who publish original content (or exclusive content for their community), alternative media, and institutions: colleges, ministries of education, national television and radio archives, etc. Public structures often need to share video content without attention-grabbing mechanisms or data exploitation. To us, this is a new step in the evolution of PeerTube's audiences.
The funny thing is that it was our initial strategy. We knew our small not-for-profit, which already hosts 20 online services for millions of users, wouldn't be able to develop a YouTube/Twitch alternative in a snap. We decided, seven years ago, to grow PeerTube at a slow but steady pace, and plan improvements so we can widen the circle of users one step at a time.
How have you been financing your projects for the last seven years?
There are two main sources of funding for PeerTube, and a secondary one.
Mainly, PeerTube is funded through grassroots donations. We have had specific crowdfundings (for PeerTube v1 in 2017, v3 in 2020), or we have used (and are still using) Framasoft's funds to finance PeerTube. Framasoft (a French non-profit association) funds come (at ~75 %) from grassroots donations for all of our 50+ projects. If you can and want to support PeerTube and Framasoft, please use support.framasoft.org: it is always needed and welcome.
The other main source of funding for PeerTube has been NGI grants programmes, managed in our case by the NLnet team. These also count as donations; but big donations from a foundation (they manage programmes funded by the European Commission). Such funding was and still is a game changer for small actors such as us: paperwork is minimal, we communicate with pros who know development and FOSS − they bring lots of advice and resources... We know that those programmes are being questioned nowadays, and if they were not to be continued, it would be a terrible blow not only for PeerTube but also for the fediverse and the FOSS community, which in Europe is a great source of digital innovation and independence.
The third source of funding for PeerTube comes from specific developments services. It is a secondary and really smaller source of funding as we have very limited time to devote to such tasks, and we only accept orders for features that match our vision/plans for PeerTube.
That being said, people seldom know that PeerTube has been developed with EUR 650,000 over seven years (2018-2024), with only one paid developer for the first six years. Our current two developers have been assisted by some who have dedicated time on strategy and communication, some external designers, and the many contributions of the community (thanks y'all!).
The result is a great piece of technology, and the support we give (once again: for free) is on par with paid support from established IT companies.
But when you compare the funding of PeerTube to that of "regular" IT companies (not even Big Tech ones), we actually have a meagre budget and an understaffed workforce. I don't think I exaggerate when I say our means over seven years would represent, to those companies, barely six months of a side project.
Don't get me wrong: we know we are very lucky to have been able to gather such means and support for a FOSS project. The problem is FOSS initiatives are dramatically disregarded, and underfunded. A consequence is that great talents usually have to go work for companies that develop toxic software, because those are the ones that get funds and recognition.
I think very few people understand that technology has a cost, and that software developed to serve humans cannot be compared to software developed to exploit them and print money. If we want ethical alternatives to blossom and thrive, we need to find the will and ways to fund them and get them resources. That, in itself, is a huge undertaking.
PeerTube is decentralised, which also means that people other than you are involved and that there are many hearts in the project, just like side features in development. And also there are some other developments funded in the NGI0 Open Calls, such as Remote Transcoding (2022-2024); Popularisation of PeerTube (2023) and others; NGI0 Discovery, by NGI0 Entrust; NGI Commons. Why did you come here and how did you use these funds?
I think it was in 2019 that we got an email from Nlnet, whom we didn't know beforehand. They had noticed our work, and wanted to present to us the grants they manage from EU funds.
Truth be told, we were very sceptical at first. In our experience, applying to (and then getting funds from) national or international programmes is a huge workload, you need to get specialised in how to answer these calls (with specific strategies), build a bulletproof application, etc. and it means dedicating one employee to this in the hope that you get the grant, and get more than the time it takes to manage it.
But we saw that NLnet already had a reputation in the FOSS community and seemed to know the reality of small not-for-profits like us, so we said, “OK, let's talk and see if those programmes are the same overcomplicated programmes that don't fit small structures like ours.”
And NLnet told and showed us and we learned, it wasn't: those programmes and funds are specifically adapted to small structures and managed by people who know the realities of FOSS development, so it changed everything.
As for how the funds were used, it was mainly for paying for development. A small proportion was used for project management, design services, etc. but all in all, I'd say at least 75% of the funds were dedicated to developments, i.e. paying developers.
For PeerTube, it funded features and improvements that are harder to promote through crowdfunding. Improvements in livestreaming, menus, filters, the API (to facilitate plug-in developments), performance, telemetry and stats, accessibility, moderation. And also new features such as two-factor identification, remote transcoding (which is a huge undertaking), users data export and import tool, comment moderation... and soon a complete UI/UX remodel and a mobile app.
Such a financial support also has side benefits: as our roadmap is supported, we get more time and energy for the community. That is how we can take time to perfect and integrate code contributions, and offer quite a high level of (free) support within our forum or the issues of our code repository.
PeerTube can be hosted by individuals, associations, companies, public authorities; it can be offered as a service by what are called instances that can be linked together. How is this an interesting option for minorities who are oppressed and downtrodden?
I think it all boils down to one thing: giving control back to people. PeerTube is a tool meant to serve, and not to exploit content creators. You can administer your own video platform, choose your content policy, choose and enforce the rules for your moderation, pick and choose whether you want your platform to federate or not with others... I have so many examples: there is a PeerTube platform dedicated for deaf people to learn and share in French Sign Language. As this language has been historically oppressed in France, deaf people have developed a culture they want to protect, with needs for safe spaces. That's why they chose not to federate their PeerTube platform with others (and they have a zero tolerance policy about ableism, obviously).
On the other hand, getting the power to "be the Google of your own self-hosted YouTube" means you also get the duties. Choosing what content belongs on your platform, how to treat that conflict or moderation problem, which federation request to omit or to accept... It's hard work! And naturally, people who are oppressed usually have less time and energy to manage their own space.
I, for one, am sad that Queermotion (a PeerTube platform that was managed by and for queer content creators) had to shut down. But I understand that managing a community and a platform was too much work, and could have burned out the nice people who maintained it.
So yes: we mean PeerTube as a tool, hopefully a useful one, to give back power for managing one's own video content. And even though we aim to improve it year after year, and to simplify... managing a PeerTube platform... taking your power back from a tech giant will always mean less comfort and more efforts.
Now, whenever people tell me they’re intent on opening their PeerTube platform, I always give the same advice:
Don't go alone, find your peers.
Consider your platform as your first home: you are getting out of Mommy Google and Daddy Amazon's nest: it is huge and scary.
Now you host your content... but it's OK to go on other platforms to reach people so they can find the way to your home.
Yes, managing your own space is a challenge, but the freedom is worth it. Just imagine that you could stop saying things like "unalive", "seggs", "please subscribe and click the notification bell", "Thanks for your subs/bits", "comment for the algorithm".
Sounds freeing, doesn't it?
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).