Comment on Why do a lot of fediverse instances put their software (i.e. Lemmy/Mastodon) in their name somehow?
rglullis@communick.news 10 months agoHow is a Lemmy app support to present that? How am I supposed to consume that from a mastadon server?
You don’t.
There is no place that says that a client needs to process every message that is received on an actor inbox. It doesn’t mean that one client should support only one specific type of activity, or even servers for that matter.
PeriodicallyPedantic@lemmy.ca 10 months ago
Maybe I don’t understand your position then.
Fediverse doesn’t make any claims for SSO or shared user accounts between server types. And servers aren’t required to interoperate with servers of other types. And clients aren’t required to interoperate with multiple server types.
It’s nice when servers and client do Interop between types (what I’m calling networks for lack of better word), but that’s not really fundamental to the fediverse, and is pretty rare. Afaict the only requirement is that servers of the same type can interoperate with eachother and user accounts from other servers of the same type are addressable.
rglullis@communick.news 10 months ago
That is the problem. Assuming that we need different “server types” is a mistake made by Mastodon that benefitted them in the short term but screwed the developers who were looking at activitypub as a simple protocol for bidirectional exchange of data.
What we need is smarter clients and let the servers be completely dumb relays. Instead of thinking of “Mastodon API” or “Pixelfed API” or “Lemmy API”, we could be looking at a single browser extension that could talk Activity Stream directly with the server, let the client be responsible for signing messages and know how to present the context when/how to serve the different activity types.
PeriodicallyPedantic@lemmy.ca 10 months ago
Isn’t this just the difference between an API and a protocol?
The payload of a message for one social network will be different based on the capabilities of that type. There are API architectures that are discoverable, like HATEOAS, but that only gets you so far (and that example is based on HTTP not Activitypub).
I don’t really see anything wrong, in the absence of a standard body, for each social network to define its own activity type, since they typically have some degree of unique capabilities anyways.
rglullis@communick.news 10 months ago
Maybe? I don’t know. Is that a relevant distinction on a decentralized system where the application logic can live on whatever side of the network?
Because they are constrained by the “client-server” paradigm. If you spend some time working with decentralized apps that assume that data is available to any nodes on the network, all your “protocol” really needs to do is to provide the primitives to query, pull and push the data around. I kinda got to write about it on an old blog post