I think the #ActivityPub client-to-server API is extremely important and underrated.

[smallcircles@social.coop]

@evan @cwebber @steve

Which was exactly what I also indicated above, and which aligns to that diagram as well.

[smallcircles@social.coop]

@evan @cwebber @steve

So why don't you use the word REST? I never encountered "read-write API". It is an informal term.

But that is not the point. You can have a REST API, fine. But that says nothing in itself. What does it expose? You might say "Duh.. ActivityPub!" but that is not very informative either. There is the notion of message exchange, and of an addressing mechanism, indicating higher level abstractions that conform to well-known architecture patterns, and would allow us to have more productive communication, delve less in implementation details and confusions of protocol behavior with solution design functionality, for starters.

[evan@cosocial.ca]

@smallcircles @cwebber @steve awesome.

So, would you like me to review your diagram and give comments? I don't know what you're looking for from me in this conversation.

[smallcircles@social.coop]

@evan @cwebber @steve

Not needed. I hope to be able to add some feedback to the AP API repo.

[evan@cosocial.ca]

@smallcircles @cwebber @steve I would personally really appreciate that. I also think it'd be helpful for the ecosystem. I like that you combine a high-level social and technical approach to discussions of ActivityPub and the Social Web with an almost encyclopedic knowledge of the details. It's a rare combination and extremely valuable.

[smallcircles@social.coop]

@evan @cwebber @steve

Thank you, that is nice to hear. I am however not an expert, am but a humble generalist and a person who'd love to be in that Solution developer stakeholder role. Who however does not see the fediverse trend in a direction where I'd adopt the technology for what I have in mind. Drifting away from "the promise" that I read in the #ActivityPub specs in 2017, and which at the time made me decide to lend a helping hand here and there as #SocialHub facilitator and tech advocate.

[evan@cosocial.ca]

@smallcircles what do you have in mind, and how is the Fediverse trending in the wrong direction for it?

[evan@cosocial.ca]

@smallcircles @cwebber @steve I'm fine with that.

However, I think ActivityPub builds up persistent state on the server side which can be read and used by other processors.

For example, when I `Like` something, it goes into my `liked` collection, and the activity goes into the `likes` for that object. The `Like` activity goes into my `outbox` and others' `inbox`. People can review that information and use it.

[evan@cosocial.ca]

@smallcircles @cwebber @steve

Some implementations skip some of these steps, because they are focused only on processing messages as they arrive. So, I am reluctant to overemphasize the message processing at the expense of the personal datastore functionality.

[smallcircles@social.coop]

@evan

So the area where my plans go I call "Residential social networking", geo-fenced but inter-connected social networking circles that cover a city, town, or rural area, and which enable their residents to not only create content on the network, but the dynamic apps and services based on local needs that exist in the area. The intent of a residential social network is to engage people *offline* and in activities that support the local economy. Or rather strengthens the Circles of Sustainability in SX terminology:

https://coding.social/blog/reimagine-social/#circles-of-sustainability

And all this should be a relatively low-code affair, directly accessible already for a first-time dev. This requires having a mature open standards based healthy technology foundation and thriving ecosystem.

I am a developer, though with rusty coding skills these days, and I might have started a fedi app design in 2018 or so. But this would not have led to the desired outcome, just throw one more app-centric software in the mix.

[evan@cosocial.ca]

@smallcircles that sounds cool!