Post

checkout my presentation on atproto spaces from FediForum last week! I was shooting for a not-too-technical presentation that introduced our design goals for spaces and the current design of the protocol. lots of analogies to existing atproto!

FediForum@fediforum.mastodon.social.ap.brid.gy

@dholms.at came to #FediForum last week to give the first public presentation on "Atproto spaces", the new permissioned data and private spaces protocol from Bluesky. Missed it? We have a recording in our PeerTube channel: https://spectra.video/w/p/6P87exJeckBwbGtAijeKJ1?playlistPosition=2

spectra.videoFediForum October 2026Keynote, invited talk and speed demos from FediForum October 2026. More info: http://fediforum.org/2026-10

14/10/26, 8:03 pm10 reposts45 likes

Six

@six.eurosky.social

The presentation cleared some of my doubts! I couldn't help but notice tho: does the space owner needs to be always one? Is it to work as a source of truth? I guess that passing the ownership wouldn't be a problem but still I wonder if there are downsides in this approach

daniel holmgren 🫠

@dholms.at

it does always need to be one DID, but for most community use cases that will be a "community DID". I fudged it for simplicity in the presentation when I said it was Alice a community DID can be controlled by one or more user accounts & be transferred around

Josh Leslie

@joshmleslie.bsky.social

Hey Daniel. Great presentation. I've been following along for while on your work with Spaces and have been slowly sketching a plan of how it might work in a health record setting. If you have some time, I'd be curious on your thoughts: hats.health

hats.healthHATS — Health records that belong to the patientHATS — health records that belong to the patient. Patients hold their records, institutions ask for what they need, the legal system gets a record of everything, and at the center, builders write the ...

Caesar

@caesar.dev

Great presentation! I'm excited by the progress on Spaces; I think it will open up many more use cases for ATProto.