Yes!
#Fediverse is not fully decentralized; it's centralized around instances. Recentralized if you will.
This actually matters practically as it's caused issues, for example with the traffic surges as the centralized servers around the internet reach out for page previews to cache all at once.
And in the debates over defederation.
#ActivityPub is not a decentralized system design, although it could have been, and I think some opportunities were missed there.
Yeah, it's a step, but I want us to be clear that it's not a tremendous leap, and that's not just trying to cast shade on #Mastodon, but it's also to highlight remaining, practical problems with retaining this level of centralization.
The system *could have* been much more decentralized, so we wouldn't have to rely on the whims of instance owners, and since it's not, we do.
For another example, because of the centralized design choices of #ActivityPub we have #privacy issues where we rely on third party instances to operate in good faith for everything from deleting content through not sharing beyond our privacy settings.
These things are really important to me, and they come directly from the federated vs decentralized issue.
I really think so much comes down simply to a well-done UI that more or less hides machinery from the user.
For example, in my mind public key encryption should be core to all of this, and yet various instant messenger clients have managed to make end to end encryption accessible to masses without them having to know the term "signing key" at all.
FWIW, I really think mass adoption just comes down to creating a user experience that isn't actually all that big a hurdle to overcome.
But maybe I'm too optimistic :)
@volkris @SrRochardBunson It's not just UI/UX.
I tried to build a true P2P messenger with a sleek UI/UX.
Problem is that I found myself fighting the tech zeitgeist.
If you say the words "P2P", "cryptographic" and "decentralized" together, people automatically believe you're talking about blockchain or cryptocurrency.
Try and explain that you're doing something else, you get met with even more skepticism.
This will take time and socialization.
Number 1, I wouldn't say those words :)
The whole point is to not bother the user with all that stuff, so why bring it up in the marketing?/
Number 2, Mastodon's success thus far shows that "decentralized" is not the dealbreaker, even though I'd argue that the platform is not.
@volkris My app never even got out of closed alpha. I was talking to (mostly) technically proficient people who *should* know better.
Nevertheless, the zeitgeist is the zeitgeist -- if a cultural narrative has taken hold, it's nearly impossible to fight it.
As for Mastodon's success, "decentralization" is not the dealbreaker *now* because more people understand the risks of walled gardens. That wasn't the case 5 years ago.
Until now, the Fediverse was a tough sell.
@volkris I'm not sure you remember this, but 5 years ago, the orthodox opinion was that decentralization was evil.
Every time I tried to bring it up, "progressive" folks said that I was being a tech bro that just wanted to destroy marginalized communities.
Much in these discussions is overly flavored by the elephant in the room, Mastodon. Their dominance is in 'post-facto interoperability', where everyone follows the leaders protocol implementations to remain compatible. But Mastodon is a Microblogging app, and makes choices accordingly.
ActivityPub is much more universal and, in theory, flexible & extensible. When AS/AP reached W3C recommendation status they weren't a batteries-included spec, but foundations to build upon.
Unfortunately building upon foundations in order to evolve the Fediverse into a heterogenous social networking environment requires deep collaboration and co-creation throughout the entire ecosystem.
That didn't happen much in our grassroots movement. It is also where FOSS community is weak: spend time beyond your project to strengthen the technology base you rely on. It should be a win-win, but is seen as beyond scope and "doing the chores" (meet/discuss! specs! docs! 😬)
Some efforts were made and some stuff happened. For a time there were W3C SocialCG meetups, and there's SocialHub dev community. The latter has a huge backlog of open issues.
Nearly all evolution since W3C standardization is someone's code in someone's repo somewhere. I call it ad-hoc interoperability.
https://notes.smallcircles.work/_FF2CKvcRy62RrygWkOWHA
It's also called Protocol Decay: https://www.ietf.org/archive/id/draft-iab-protocol-maintenance-08.html#name-protocol-decay
Luckily FEP is a bit more active now: https://codeberg.org/fediverse/fep
But richly late..
With corporate interest on the fedi + critical mass maybe reached, each and every company thinking of launching a federated product or integration will conclude the incompleteness of the specs and the absence of a "technology substrate" (people and processess to evolve the specs + ecosystem).
And either they'll create their custom extension flavors, or they'll organize technology steering themselves, alone or with partners. And FOSS community can fall in line behind that.
My time doing advocacy and community work made me realize a fundamental weakness of the whole FOSS movement.
FOSS is strong at project level esp. doing the technical work. It outputs deliverables with the 4 freedoms attached.
But that's ultra-weak protection, with entire process and R&D transparent only enabling a 5th freedom of hypercapitalism to corrupt.
And FOSS movement can't deal with ecosystems and technology substrates. We're lacking the organizational powers.
This might change. Ironically enough the Fediverse's paradigm of heterogenous social networking provides the opportunity to overcome our weakness.
But this involves an evolution of our ecosystem that goes way beyond the limited approach that dominates (cram everything into some microblogging domain and copy features of corporate Social Media).
With AS/AP is a Linked Data protocol. Everyone can develop AP vocab extensions for their domains and they'd be building blocks.
A good example of the huge opportunity that lies in front of us is in the space of forge federation.
Now the term is wrong. A forge is an app. The domain is more relevant. And that top-level domain is no less than Software Development!
Imagine what that means: software development processes tied deeply into the fabric of the fediverse itself.
Every online web app where multiple people collaborate is a social app, and can be a fediverse solution.
https://fedi.foundation/2022/09/social-networking-reimagined
Oh gosh I so strongly disagree with SO MUCH in that article.
From the paternalistic, top down tone through the overlooking of different talents of different people--technical vs social--I think it is really flying off in the wrong direction.
If that sort of attitude is to guide Fediverse development, then this whole platform is doomed as a pipe dream.
Thanks for the honest feedback. Yea, main takeaway is that a different approach is possible.. mostly as a (technology) vision to change course somewhat. The peopleverse that allows the different talents their full expression requires a level of socio-technical support that bolting everything on top of a Microblogging domain will never offer.
@volkris @SrRochardBunson SSB and Manyverse are much more de-centralized. Those services pretty much offer what you're looking for.
Problem is they're too much of a paradigm shift for mass adoption.
True decentralization will take steps.