Introductions and kick off

WG Private Data
1

Hello! It’s time to kick off the Private Data Working Group!

The goal of this group is to coordinate on implementations, documentation, interop of private data solutions for AT Proto. This may be off protocol implementations that interop, it may lead to a proposal or RFC of making private data part of the protocol – for now, it’s a space for collaborating together with people who want to implement some kind of private data.

I’ve also created (and enabled wiki mode) on this What is Private Data? post:

Feel free to comment or edit the wiki directly. Some of the work we need to do is development shared definitions and language of what we mean, and what users expect, and that will be part of the output of this group.

I’ve created a Discourse group that anyone can join @PrivateDataWG . If we want to discuss things a little more privately we can message there, we can tag the group as a member to notify folks that we want them to pay attention, and it’s used as an indicator that you actively want to participate and be invited for various things. I don’t know if it is needed aside from this public forum category, we’ll see!

A note: all are welcome here. The sessions that I’ve held in Seattle and New York both started by defining and disambiguating between Private Data and E2EE Messaging, and then talked about use cases and shared language.

Protocol talk is also welcome here, and I’m sure we’ll see our fair share of encryption and cryptography talk, as well as network architectures. But, this should be informed by use cases that we are trying to make possible, and features we’re trying to add.

I’ll also start a separate thread on planning an initial video meeting in a couple of weeks.

Please write an introduction about yourself, what project(s) you’re working on, what use cases you’re interested, and anything else relevant.

5 Likes
2

Hi there! My name is Boris Mann, my interest in Private Data is unlocking a host of other applications that need data to be private, as well as lots of potential user & community use cases for the core Bluesky app.

I think Private Data is existential for being able to take on platforms, which offer “private data” by locking data in private databases. Facebook Groups, Marketplace, Instagram Friends Only are just some of the use cases / killer apps that I’d like to enable.

My two main projects right now are:

  1. Working with Ink & Switch in a COO role, who are also makers of the Automerge (local first data + sync) and Keyhive (e2ee encrypted data with local first access control, designed to add these properties to Automerge).

  2. ATProto Community Fund: helping to grow the ecosystem of applications on ATProto, organizing events, and supporting builders and creators.

I’ll help with coordinating, convening, and note taking for the working group (as well as communications and getting the word out). If anyone is interested, I could use a partner to pair with on helping support the running of the working group!

3 Likes
3

Hey there! I’m Zicklag and I’m working on Roomy, a “gardenable” group chat app at Muni Town.

I don’t have the capacity to really be involved here at the moment but since we are working on an ATProto integrated app that has private data I figured I’d drop a quick introduction.

Because Bluesky isn’t designed directly for realtime, and because we have a focus on local-first features, Roomy isn’t able to leverage the native ATProto repo for much other than public uploads, so our approach for private data, and everything else, is to store it off the PDS in a custom data store.

That said we’re interested in inter-operating with other private-data solutions that might develop, possibly to store private chat archives or other integrations that would bring value to Roomy or other apps.

4 Likes
4

Hi friends, I’m Tessa, the ceo and cofounder of Germ Network. Our multi-identity, phone-number-free E2EE messenger is live for iOS and our atproto integration is in beta. Super enjoyed our sesh in Seattle and definitely want to keep tabs on this conversation and contribute however I can! Hi from SF :slight_smile:

5 Likes
5

Hello! I’m Tony, working on blebbit.app, a platform for communities. I’m looking to support the spectrum from reddit / discord / groups use-cases and that requires the private data concept, or what I’m calling “permissioned spaces” to differentiate from E2EE data use-cases. One goal I have in mind is to make it possible for appviews to get sophisticated permission capabilities like Google Workspace without having to implement it themselves. To that end, I am working on a fork of the official PDS, lexicon, and all that jazz.

There is a proposal.md at the root, but this work is still quite far from being a proposal

4 Likes
6

I’m @ngerakines.me, software developer and ATProtocol platform thinker. I made Smoke Signal and a handful of apps, tools, and libraries in the atmosphere.

Very interested in private data and how it impacts communities.

3 Likes
7

Hi everyone, Wesley here. CTO / technical co-founder of https://cosmik.network/ working on https://semble.so/ - a social knowledge curation tool on atproto. Interested in private data because we’ll want to use it for Semble when available.

3 Likes
8

Hello everyone! I am Miguel, a software engineer based out of Arizona. I am currently in the early stages of building out a long-form publishing platform called Offprint. Private data would be an amazing use case for us as we will have pay-walled content that we’d prefer not to host ourselves and allow user portability of this data.

I’ll be lurking through this discussion as I’m not very versed in cryptography or the like, but I do have opinions and insight into how private data could be used within platforms like Offprint.

2 Likes
9

Heya @btrs.co, welcome! No specific knowledge is required!

One of the things that would be specifically helpful is for folks to start writing up use cases for their specific app.

Would you be comfortable writing a new thread describing a handful of Offprint use cases / features enabled by private data?

eg private drafts stored on PDS so we don’t have to store drafts in a centralized server backend

(That’s a little too brief but hope you get the idea)

1 Like
10

Hey :waving_hand: I’m Brooke, I’m working on a project called pckt.blog. It’s a blogging platform designed to make starting and writing a blog as easy and straightforward as possible.

We have been considering options for integrating ATProto to rely less on a centralized server.

One idea and probably one of the first things we would like to integrate would be using private records to store drafts for posts.

Could also be useful for storing blog specific settings like themes configs, editor preferences, etc.

Another idea would be to create a versioning system that could provide snapshot records of drafts for comparison or rollbacks.

Interested in learning more about ATProto !!

3 Likes
11

Hi, i’am @willem.dobs.nl software developer /architect of a platform providing farmers with a dataspace for storing all their data ( kind of a PDS ).

As a side project i’am now working on a ATProto PDS implementation (custorium) for private data, with the ultimate goal of replacing the dataspace in our platform with a ATProto based one

2 Likes
12

G’day all! I’m Ian Preston creator of Peergos (book.peergos.org). I was in the early private bluesky incubator, setup by Jack Dorsey and Parag Agrawal, for my work on Peergos. I rejected their initial framing of public-only data, knowing that a true social web replacement needed private data as a first class citizen. This is why I started working on Peergos in the first place in 2013, after the Snowden revelations (progress has been much slower without a billionaire’s backing! Though we are bootstrapped).

Happily, after Jay reviewed Peergos in depth in the ecosystem review, she adopted the major properties of portable identity and portable data and social graphs (obviously excluding privacy), and user controlled feeds. Because of this and the common foundation of a restricted dag-cbor subset (now called dasl), Peergos could be basically a drop in addition to atproto for private data. The main difference is we use different dids, and the lowest level of access control to ciphertext is enforced by the PDS. To actually drop-in unchanged, atproto would need to support Ed25519 identity keys. You can read more about our proposal here: Private, non-shared data in repo? · bluesky-social/atproto · Discussion #3363 · GitHub

The use cases I’m interested in are Facebook replacements, charging for content, private events, private surveys, private apps.

5 Likes
13

Hi all,

I’m Gautam a software developer based in San Diego, CA. I’m interested in Private data since I think it’s needed for opening up potential applications that can not fully be built on @Proto at the moment. I’m, also, interested in building application with E2EE enabled, and have private data is going to be an important step in that direction.

2 Likes
14

Hey Everyone, my name is Gaetan!

I’m a Digital Artist/Animator and Creative Technologist. I’m not building anything yet, but I’m really into the idea of decentralized networks and excited to see where this all leads. For right now, I’m here to learn as much as I can from the community and see where someone with my skills could fit in.

2 Likes
15

Felicitous greetings, everyone! I’m Dave, a London-based software developer, and the nerd-like organism behind Skiff.

Since attending the Ahoy meet-up in Hamburg at the end of April, I have thrown myself into ATProto and have been learning Rust in the process, which now means I am thoroughly bored with writing C# in my day job!

Over the past couple of weeks, I’ve been experimenting with the Bluesky app source in an attempt to launch a social network for the bear subculture within the LGBTQ+ community; therefore, I believe private content/data is vital for the future of ATProto. With that in mind, I recently wrote an RFC on the subject to get the conversation moving. Please feel free to take a look and open an issue if you think I’m spouting nonsense, or tell me directly here; no well-formed argument will be ignored! :smiley:

Apologies for the preview, I can’t seem to get it to show the link alone.

https://github.com/knasher/rfcs/blob/main/atproto/001-private-content.md

1 Like
16

Welcome Dave! Thanks for posting!

Dave is one of the few people who has attempted to write down a specification around Private Data. I’d like to have people read it and use it as an example to discuss for the first kick off meeting – as well as asynch issues and discussion as Dave as suggested.

Dave – I’m going to also make a dedicated topic for your thread.

P.S. I removed the preview, in the future people can put angle brackets around URLs you want to not embed <https://example.com> https://example.com.

1 Like
17

Hey Wesley, welcome!

As an app developer who probably mostly wants to consume / implement the spec, could you please start a thread with your use cases and any thoughts on must haves / requirements that you can think of in context of Semble?

18

Hi all! I’m Paul Frazee, CTO at Bluesky and one of the early AT designers. I don’t know how rigorously I’ll be participating but I hope to share thoughts and feedback at times!

4 Likes
19

Hey everybody. I’m Ted Han, part of the ATProto Community Fund, and doing the EiR thing here.

I am focused on the ATProto app ecosystem, AppListings and app capabilities generally speaking. I view private data and access control as a key and critical feature space for apps wishing to build on or integrate ATProto.

3 Likes
20

Hey folks. I’m Kimberly Grommes. (Currently unaffiliated.) I have a couple apps I’d like to build on ATProto and at least one of those would benefit from private data so I’d like to help move this forward.

2 Likes