Re: The "Social Web" vs the "Fediverse"

On 12/27/2023 11:04 AM, Johannes Ernst wrote:

> the core of my question was intended to be a few abstraction levels higher: is what we can do with the current protocols all there is to “social web” or what other functionality or user experience do we refer to when we say “social web” instead of what we have today?

If the list will forgive my navel-gazing, I would define the social web as the subset of the web where properly "pro-social" interactions take place-- whether pseudonymous or anchored in real-world identities, the participants in a conversation or thread think of each other as specific humans posting in good faith with appropriate levels of accountability, with social norms keeping the signal-to-noise ratio high (and here I would class spam, advertising, inauthentic activity, and anti-social content as four distinct kinds of "noise" being ever more normalized by the rest of the web, which we might call the antisocial web for the sake of this definitional excursis). Getting into the weeds one might argue that open-platform guarantees and social-graph portability mechanisms are debatably also ingredients or requirements, but the heart of my definition of the social web is that there is "quality" (I know it when I see it) and accountability (truly the missing layer of the internet to date). These are the two things you most need when convening people to do things socially (particularly with strangers).

Protocols could, in the best of cases, make some subset of the social web more archiveable, portable, scalable, machine-readable, and/or useful; they might even protect these pro-social gardens from the corrosive effects of the deregulated free market. But the protocols create overlapping and in some ways competing platforms _within_ the social web as people invest in them over time, and what's good for a given protocol or economy of scale isn't necessarily good for the social web _as a whole_. And we should view with healthy suspicion any proposal claiming too confidently to know what's best for the social web and which possible future is the best.

> Personally I believe — and I gave some examples — we could support far more types of “social” activities on the web than supported by either proprietary platforms or the ActivityPub Fediverse (or the IndieWeb for that matter) today.

I agree that the AP protocol as a whole, the Mastodon-flavored and Lemmy-flavored and Pixelfed-flavored and Gab-flavored sub-networks of that protocol, the AT protocol, the Nostr protocol, and the #IndieWeb are all platforms within the socialweb that make certain trade-offs and have certain limitations on what kinds of activities they can represent/interact with "in-feed" and/or natively. That's natural: you will never get to global scale while simultaneously optimizing for all kinds of content, all social structures, and all moderation approaches. This is why none of them is the "best" or most important subset of the social web, and any ranking of them says more about the ranker, their use-cases/priorities, and their social values than about the platforms themselves.

While there are many great FEPs adding new activities to the AP vocabulary (and I'm still waiting for the FEP defining chess moves from that amazing online chess demo at the last FediForum!), other activities don't need to be AP-wide, and can be built into one server or one sub-network as an experiment before becoming FEPs (and maybe some day features of a vNext AP). What I love about AP is that for all its rough edges and verbosity, it feels like it was designed for the very extensibility you seem to be describing, in which more activities get incorporated over time first as "optional" (ignorable, opt-in, sub-network) additions to the data model via FEPs/local @Context entries, and later as "first-class" members of the data model if this CG folds them into vNext.

That said, new activities might also enter the social web through other channels further afield of this CG, bubbling up through AT Protocol, Nostr, #indieWeb, etc. If enough people use those networks, and those networks have a first-class treatment for a given activity, some kind of wrapper or translation for AP servers will inevitably get FEP'd and then roadmapped by the peer pressure exerted by the rest of the social web, right? :D

That's how I understand the relationship between "activities" in the everyday sense (things people want to do with real people in some corner of the social web relatively free of spam and inauthentic activity) and Activities (in the JSON-LD sense defined at W3C for one particular social-web protocol). Hopefully this freewrite is helpful to someone writing a CG/WG charter in Q1-- it's not too late to make a public New Year's Resolution if anyone wants the social web to be their accountability mechanism!

Thanks,
__bumblefudge