This is interesting, but I don’t yet entirely understand it.
My first thought after trying to read the entire document was that the author seems to suggest that “AT Protocol” is a natural result of the movements they describe, but I find it hard to believe that the “peer-to-peer (p2p) movement” could naturally result in things that “are not meaningfully decentralized, and are not federated”.
I think they meant mostly the decentralised distribution of data.
At the end of “Generic hosting, Centralised product development” it says
Even though product development is centralized, the underlying data and identity remain open and universally accessible as a result of building on atproto. Put another way, ownership is clear for the evolution of a given application, but since the data is open, it can be reused, remixed, or extended by anyone else in the network.
So theoretically everyone can access the data but before it reaches the end users it goes through centralised applications like bsky
This is interesting, but I don’t yet entirely understand it.
My first thought after trying to read the entire document was that the author seems to suggest that “AT Protocol” is a natural result of the movements they describe, but I find it hard to believe that the “peer-to-peer (p2p) movement” could naturally result in things that “are not meaningfully decentralized, and are not federated”.
I think they meant mostly the decentralised distribution of data.
At the end of “Generic hosting, Centralised product development” it says
So theoretically everyone can access the data but before it reaches the end users it goes through centralised applications like bsky