Re: Towards a Unified Open Social Web Spec

On 28/08/24 18:07, Aaron Gray wrote:
> On Wed, 28 Aug 2024 at 16:30, Cristiano Longo 
> <cristianolongo@opendatahacklab.org> wrote:
>
>     More precisely, it is a strict subset of JSON-LD.
>
>     "The serialized JSON form of an Activity Streams 2.0 document MUST be
>     consistent with what would be produced by the standard JSON-LD 1.0
>     Processing Algorithms and API [JSON-LD-API] Compaction Algorithm
>     using,
>     at least, the normative JSON-LD @context definition provided here."
>
>     cfr. https://www.w3.org/TR/activitystreams-core/#jsonld
>
>
>
>     On 28/08/24 15:24, Evan Prodromou wrote:
>     > How so?
>
>
> AFAICT Not everyone follows the 
> `https://www.w3.org/ns/activitystreams` 
> <https://www.w3.org/ns/activitystreams`> @context JSON-LD 
> definition to the letter adding and making variant object types.
However, if the range of a property p is defined of type X from the 
specification (may be the context or other form of specification), and 
an agent send an object of type Y as value of p, then the receiving 
server must assume that Y is a subtype of X.
>
> My idea is that that "extra" compliant "interop" servers may also 
> augment the activitystreams spec with the source or destination agents 
> extensions and deviations based on the agents id or its domain thus 
> determining its id.
>
> It would be nice to have a webfinger agent id field and maybe an 
> augmented context field variant agents could give, falling back to the 
> determinism of the id then the domain augmented by a common 
> repositioryt of deviant agent extensions and deviations.
If I understand (and I suspect, I can't) such an information may be 
encoded in the trasmitted object itself. This would solve problems 
related to changes in the capabilities and representational modes of the 
server which of course will happen in time.
>
> Kind regards,
>
> Aaron
>
>     > Evan
>     >
>     > On 2024-08-28 2:38 a.m., Melvin Carvalho wrote:
>     >> However, ActivityPub did diverge slightly from the JSON-LD
>     standard
>     >> in the end, which complicates interop.
>     >
>
>
>
> -- 
> Aaron Gray - @AaronNGray@fosstodon.org | @aaronngray@threads.net
>
> Independent Open Source Software Engineer, Computer Language 
> Researcher, Information Theorist, Amataur Computer Scientist and 
> Environmentalist and Climate Science Disseminator.
>