Re: Towards a Unified Open Social Web Spec

On Wednesday 28 August 2024, Melvin Carvalho <melvincarvalho@gmail.com>
wrote:

>
>
> st 28. 8. 2024 v 5:57 odesílatel Aaron Gray <aaronngray@gmail.com> napsal:
>
>> Hi Melvin,
>>
>> Please don't be put off there's still lots of furtile ground within
>> ActivityPub and JSON-LD, and extensions to Activity Streams to bring
>> existing extensions https://swicg.github.io/extensions-policy/ within a
>> meta framework, where the nonstandardized can be "standardised based on
>> source or destination domain and a JSON-LD description.
>>
>
> Hi Aaron,
>
> Thanks for the encouragement. The topic of ActivityPub extensibility came
> up during the 2014 Paris face-to-face, during the Working Group. There was
> a proposed AP extension system, and I argued that JSON-LD already had an
> in-built extension mechanism, so we could leverage that. Fortunately, my
> argument prevailed at the time.
>
> However, ActivityPub did diverge slightly from the JSON-LD standard in the
> end, which complicates interop. This is one reason to revisit and
> potentially create a clean, standard-compliant system that can fully
> realize the promise of standards—interoperability across all the diverse
> parts of the social web.
>
I think only deviance from the standard and usage of JSON ignoring the
@context rather than using JSON-LD or being in alignment with the `
https://www.w3.org/ns/activitystreams' protocol/specification by adding or
modifying features or treating them with different semantics.

> Meta descriptors sound interesting—metadata underpins HTTP and much of the
> web. There’s already significant standardization in this space. Could you
> elaborate on what you mean by meta descriptors and how you see them driving
> standardization and interoperability?
>
I have a long time interest in meta systems, systems of systems, systems
for describing systems including themselves, and also meta languages.

I am developing a means by which rather than programming totally in bespoke
code that descriptive systems can be defined in terms of meta constructs
and manipulated as such, obviously with some augmented bespoke or pragmatic
code to augment corner cases. It is still early days yet but I am hoping to
prototype soon for my own social media, networking, and collaboration
system that will be extensible by design with very little actual
programming other than specification.

Kind regards,

Aaron



>
>> Kind regards,
>>
>> Aaron Gray
>>
>> On Tuesday 27 August 2024, Melvin Carvalho <melvincarvalho@gmail.com>
>> wrote:
>>
>>> Hi All,
>>>
>>> The open social web is making great strides, but we're seeing varied
>>> efforts with different trade-offs.
>>>
>>> ActivityPub is federating at scale, reaching millions with solid
>>> moderation. Nostr, though smaller with 10k-20k DAUs, offers a rich
>>> playground for R&D with advanced features like zaps, encryption, and app
>>> portability. IndieWeb is driven by passionate folks focused on interop,
>>> specs, and running code. Solid blends social with personal data storage,
>>> standards-compliant and hopefully WG-bound soon.
>>>
>>> Yet, we lack a unified data model to bridge these efforts seamlessly.
>>> Some promising bridges exist—Alex Gleason’s “Ditto” between Nostr and
>>> ActivityPub, and Bridgy unifying across systems, even touching Bluesky. But
>>> there's no consistent, extensible, and interoperable spec that allows
>>> everything to just work together.
>>>
>>> The promise of standards has often fallen short—things built outside the
>>> standard, or standards not quite fitting needs. For instance, adding a
>>> second "Nip-05" identifier in Nostr could take ages to agree on, despite
>>> being technically simple. Similar issues linger in Solid, even with its
>>> theoretically compliant system.
>>>
>>> We’re not reaping the full benefits of standardization, though they’re
>>> within reach. Maybe it's time for a few of us to craft a unified W3C social
>>> web spec. We need a flexible template where developers can build freely,
>>> rapid prototyping in a permissionless environment, with specs that don’t
>>> require months of consensus. Backward compatibility, unihibited
>>> development, and a an outlet to unlock new waves of creativity, could be
>>> the result of a clean unified social spec
>>>
>>> Would love to hear your thoughts on this. (yes, I know xkcd!).
>>>
>>> Best,
>>>
>>> Melvin
>>>
>>
>>
>> --
>> 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.
>>
>>
>>