Recent searches
Search options
![Screenshot of the change diff for a commit called "enable local timeline in Phanpy clients". There is exactly one changed line, which modifies the API to return "; like Akkoma" at the end of the server's version string. The final output looks like a browser user agent, reading "3.0.0 (compatible; Sharkey [version]; like Akkoma)".](https://enby.life/files/7a7cc08a-838a-4e5d-830f-9d6ef1075125.webp "Screenshot of the change diff for a commit called \"enable local timeline in Phanpy clients\". There is exactly one changed line, which modifies the API to return \"; like Akkoma\" at the end of the server's version string. The final output looks like a browser user agent, reading \"3.0.0 (compatible; Sharkey [version]; like Akkoma)\".")
Hey, let’s be positive.. look on the bright side.. it’s still better than the ActivityPub C2S API 
@benpate@mastodon.social I'm actually not very familiar with that... is it bad?
@hazelnoot My understanding is that C2S was more aspirational than actual. I've never had to use C2S, but have only heard from many others that it's cumbersome, and not really usable as a client API, so (virtually) nobody uses it.
The state of ActivityPub in general seems to be many overlapping, mutually exclusive, and un-adopted standards, so as much as it sucks, it makes sense that mobile APIs would fit that stereotype as well.
I hope this works out for you! :)
@benpate @hazelnoot tbh, the worst part is that there's a lot of federation behavior that is completely unspecified, except by loose implication in the C2S parts.
The division of C2S and S2S feels to me like some people wrote it in this way because it's nice to read. Describe a server to server protocol by talking about what happens
- if a client sends a message to the server and the server then sends it to the next one
- and if a server receives a message from another server
I can see myself doing that. Then the genius manager decision happened: We can call this two protocols and claim twice the fame.