Looking for feedback: Independent collective for AT Protocol infrastructure builders

Discussion
1

Hello, I’m :palm_tree:Tree (@tree.fail). Some of you may know me from Bluesky and my never ending rants about how atproto is not decentralized enough yet … :slight_smile:

In 2023 I created @atscan.net which was explorer or service scanning #atproto ecosystem and monitoring PDSes, but I had to shut it down. My idea was too big for myself - the main problem was coordination with others, where I failed. Building something alone is really hard, and I know others struggle with the same thing.

So here’s an idea: what if we created an independent collective for people building critical AT Protocol infrastructure? SDKs, indexing services, security tools, PDS implementations, documentation - whatever the ecosystem needs. Projects stay completely independent, but the collective helps with coordination, fundraising, accounting, legal work, and shared resources. Those who build decide how it works. Everything transparent.

Think Argot for Ethereum, but for #atproto. Neutral, non-profit, focused on long-term sustainability and real decentralization.

Right now this is just me and a concept:

I’m looking for founding members - maybe 3-10 people who want to shape this from the start. If there’s interest, we can immediately draft a constitution together and start coordination.

But I need your honest feedback: Would you actually join this? Does it make sense? What’s wrong with it? What would make it useful for you?

I really need your help here. Any feedback, thoughts, concerns, or even if you think this is a terrible idea - please tell me. Would you join? Does something feel off? What’s missing? Even small comments help a lot.

Also if you have ideas for a good name, I’d love to hear them!

Thanks for reading :folded_hands:

2

Just clarifications how this collective idea relates to this forum and the broader AT Protocol community. They’re not competing - we can cooperate and benefit each other. This forum is exactly what it should be - a broad, open space for everyone. And honestly, I believe this forum and community (including AtmosphereConf) is also critical infrastructure that we need to take care of.

The collective I’m proposing is something more focused and compact. Think of it like guilds within a larger community. We have this global atproto dev community (which is great!), but then we also have focused collectives for specialized work. This collective would be specifically for people working on critical infrastructure projects - tighter coordination, shared resources, long-term sustainability. Not everyone needs that.

3

Hi tree! Seen you :wink:

I’m interested but I’m sort of, uhh… stuck, I guess? I’m one of the few people who don’t write Typescript, Go, or Rust so as far as being a contributing member to any of that, I’d be right out (not unless there’s a few Perl hounds out there…).

However what I can bring to the table is infrastructure; I’ve been running my own projects on a hilariously over-engineered setup that could easily be turned into easy hosting for projects up to, well, however big you want it to be really. I can bring ~3 decades worth of sysadmin/devops experience as well. I know that spinning up a VPS is easy (and cheap) enough, but I could at least take away the “stupid work” of hosting something so people can focus more on developing and don’t have to deal with the “administrative details” so much.

So yeah count me interested :slight_smile: I guess this’ll have to get fleshed out more but overall I don’t think it’s a terrible idea. As long as it doesn’t turn into something where if 2 people are working on something that’s essentially similar one of them gets frozen out. But I guess that falls under the details that need hashed :slight_smile:

2 Likes
4

Hey @angrydutchman.peedee.es! This is great actually.

So I’m thinking we need two kinds of founding members here:

  • #1 People with critical infrastructure projects they want to bring in (SDKs, indexers, PDS stuff, etc.)
  • #2 People who can help with shared resources - marketing, hosting, devops, coordination, legal work, fundraising, documentation

You fit perfectly in #2 and honestly that’s just as important as #1. Your infrastructure/devops experience is exactly what multiple projects would benefit from.

Honestly though, most hardest to find will be the non-tech people for category #2. Like, we tech folks live in completely different worlds from the marketing/legal/fundraising people.. its always a struggle :slight_smile: (If you’re someone like that and want to help us nerds, please reach out!)

1 Like
5

Welcome Tree!

We created the AT Community Fund (fiscally hosted by Raft Foundation, US 501c3) to host the first conference and pool funds for a few things but there need to be more entities.

I’m on the record saying any single foundation model is not a good pattern.

Will be interesting to see as the Swiss association running the PLC gets set up what it takes on and how it does coordination.

Maybe one more thing to think about: getting REALLY clear on what the needs are – eg “we provide shared legal and moderation for independent PDS” is more narrowly focused…but also that’s like 2000+ servers today!

I don’t see the AT Community Fund / @atprotocol.dev joining such a thing: there should be more new entities with specific focus. Happy to help share what we’ve learned.

I would encourage folks to start with / explore fiscal hosting before setting up an entity, unless the goal actually is to have a particular structure and “home” somewhere – e.g. a desire for a Brazilian entity (insert whatever jurisdiction you like).

Fiscal Hosting also shouldn’t be forever, this 2026 conference is probably at the limit of funds that I’m comfortable flowing through a fiscal host, never mind having it be a US based entity.

4 Likes
6

@bmann.ca Thanks for the comments. My main focus is on creating a community of people who will help each other, and creating a legal entity and fiscal hosting is, in my opinion, just one of many different types of help (needs).

What I mean is that the needs this initiative will have, and be able to secure, will depend mainly on what kind of people and projects get involved.

You and AT Community Fund / AT Protocol Developers / ATmosphereConf can have own legal entity and still be involved in this initiative; I don’t see any contradiction. You might just need a different kind of help. This is about a loosely-coupled connection where projects and people have their own autonomy. At the same time, it’s not a universal project for everyone; not everyone shares the same principles, and I definitely agree that there should be more such “guilds”, because of decentralization and subtraction :slight_smile:

7

Hey Tree,

Love to see the energy. There definitely need to be more folks banding together to do things. But a need isn’t sufficient by itself. It’s important to have a banner or mission that helps to bind an org together, even if it’s a loose one.

So it’s great that you want to get people together, and i hope more people in the ATmosphere band together, but it’s important to know why and for what purpose. Is it just run open source projects? Is it to grow particular parts of the ecosystem? What do y’all hope to accomplish? And, for example, what would motivate funders to support such an effort?

1 Like
8

Yay! Anyway, yeah I’d be up to contribute to that, I guess it also depends a little bit on what “the ideal” situation would be re: hosting stuff. I think I’ll go tinker with that regardless of what happens, and we can see at some point how to fold that up into things.

1 Like
9

Thank you!

I feel that the basic mission and purpose are already written in that conceptual document. It’s about building “critical infrastructure” while maintaining certain principles, such as neutrality, openness, transparency, and so on.

I understand that “Critical infrastructure” isn’t an entirely exact term, but I think we will be able to mutually determine what open-source critical infrastructure is and what isn’t. For example, in my opinion, this automatically excludes projects like Tangled, Leaflet, Streamplace, Slices, and similar, which certainly have great benefits for the AT Protocol ecosystem as a whole, but are definitely not critically important. In contrast, there are SDKs, independent PDS/relay/appView implementations, explorers like pdsls and other developer tools that definitely are.

What we are building
What we are building1770×964 146 KB

I’m not entirely sure how to specify it further, because the concrete needs and goals will truly depend on the projects and people who join.

On a large scale - a functional, sustainable, and Bluesky PBC-independent ecosystem

On a small scale - greater opportunities and mutual help for contributors who wants the former

10

Hi :palm_tree:Tree , may be you want to look into the Eurosky initiative? Sound like we have at least some overlap :wink:

Check out https://www.eurosky.social for more info, if you have the time :slight_smile:

11

@schuman.de and WG Eurosky although the team hasn’t really adopted it yet :stuck_out_tongue: