@[email protected] Thanks for the thread! Coming myself from a linked data background and having adopted a simple use of JSON-LD as Linked Open Usable Data (LOUD), I never understood (and still don't understand) what problems people have with JSON-LD in AP and AS. I am much in favour of an open world approach. It is quite powerful if people share their extensions and try to find and reuse solutions by others. In the end, we'd create shared data models together: a social act for the social web.
News from fediverse
@trwnh 1.) love this, looking forward to the blog post. 2.) i'm not 100% convinced of the analogy of open-world : closed world :: AP : "one" "social" "network" , but it resonates a lot with my thinking on platforms lately. i think the fediverse thinks of itself as ONE OPEN platform, rather than multiple overlapping platforms (that could include closed platforms, too, in every sense of closed including the economic!) with no global guarantees, periodt.
@by_[email protected] yeah i'm not saying AP is "open-world" but rather it straddles the line
AS2 requiring the AS2 context is a bit weird from an LD perspective because it introduces weird "supremacy" conflicts especially with the "MUST NOT override" requirement
i've thought that perhaps jsonld context should only ever be a "progressive enhancement" to json, and that new apis or interchange formats should instead use expanded form, and processors should expand any compacted json(ld) before using it
@trwnh oh interesting i didn't realize that you meant openworld/closedwforld that literally in the RDF sense, i thought you meant more in the protocol-design sense (of like "drop all unfamiliar properties" as is conventional for all JSON protocols versus "here is how you cautiously parse or preserve for others what you don't know")
the main contention is a disconnect between #ActivityPub as a spec and #fediverse as a protocol/network. a lot of problems cited were with the fediverse as implemented, wishful thinking about what could be changed in spec, many backwards-incompatible, mostly in service of making fediverse impl less painful.
there is a recurring refrain about implementers deciding they don't care to implement AP as specified, and that this indicates a problem with the spec, not a problem with implementers.
2/?