Challenge: Adopt a shared (technology) vision for the Fediverse

[aschrijver]

This topic relates and was moved from: Challenge: Fixing the Fediverse Technology Adoption Lifecycle.

---

Question: Does having a technology vision help adoption?

On Lemmy, where I posted a link to this topic, I received interesting feedback and critique to my observation that the fediverse as a whole is visionless. So I took quite a bit of time to explain in what light I meant that…

@electrodynamica@mander.xyz commented:

My criticism on Fediverse as a whole is that the technology is currently visionless.

Completely disagree. There are many of us who have been working on this for a decade plus who share a solid vision. We just don’t have capital. I think you mean to say the software that becomes popular is for some reason the most visionless of all the software being worked on. Perhaps figure out why that is first… (Lacking capital is the pretty obvious answer IMHO)

In follow-up comment after I asked how they see the vision, @electrodynamica mentions that the open standards exist for quite some time, including oft-forgotten RDF and SPARQL and lists a number of fields, areas of interest, where indeed they may find / have found good use, like healthcare, small business, customer interaction. Asking me for which field to describe a vision first.

Here’s my (yeah, well… lengthy) response:

Well, it is not needed. There’s some things in your response where I can follow up on to demonstrate how I meant the issue I have re:fedi-vision. Mind you, I am passionate about all these technologies, and have spent countless hours advocating their broader use. By creating this community for instance, or moderating SocialHub, maintain fediverse watchlists/delightful lists, and also by creating plenty of issues in project trackers proposing fedi support.

On RDF… I was deep into XML standards at the time it was drafted as a spec, and I was thoroughly excited about the vision of the Semantic Web at the start of the hype. Very sad when it was clear that it failed. RDF et al have their uses, but after many years still not as widespread as originally imagined. I have never seen a single app, where I would say to others “Yes, of course this is a gorgeous delicious piece of work. It is a Linked Data app, after all”. Most good linked data apps I’ve seen were made - and only to be understood - by the academic world. I am sure there are good ones. Likely in specialistic domains. And I’m aware about some widescale application of linked data, like Google Knowledge Graph / schema.org mostly for SEO, Wikidata, etc.

But there hasn’t been mass adoption of Linked Data, and historic experiences have created salted ground in the dev community. On fediverse most devs don’t care much about JSON-LD, just append that @context prop or advocate for going plain JSON altogether. Besides the academic / specialist world, the locus of attention re:linked-data was in the Solid Project, not surprisingly led by Tim Berners-Lee. And there’s some other interesting standards development, like around Verified Credentials with a host of specs (Digital Bazaar, Manu Sporny being influental here). Now in the Solid community I spend a lot of time discussing the very weird, very unclear positioning of the project in the past. And recently their website has improved quite a bit as to what the project is about. But I still get the feeling that in the background there’s a motivation of ‘re-inventing the semantic web’.

To end on RDF / Linked Data itself there was recent discussion Is RDF “hard”? that I also brought to SocialHub as Linked Data: Undersold, Overpromised? that are interesting reads in this regard, I think.

Vision of the Fediverse?

What I meant to say with “My criticism on Fediverse as a whole is that the technology is currently visionless.” is that there is no common outlook on what the fediverse technology is to be used for, capable of, and what its full potential entails. There’s no shared common vision that quickly communicates to someone who never heard of it befofe, that makes them jump with excitement and say “Aha, now that’s fascinating. I am going to jump on this technology as I shouldn’t miss the boat”.

And not only that, because that’s just the technical aspect. More importantly there’s only quite technical-sounding storytelling around the fediverse. If you tell some non-technical person “You should come to the Fediverse” and they say “Why? What is that?” then likely the answer will include ‘federation’, ‘decentralization’, ‘instances’, ‘moderation’. And you can make cases why fediverse is better than traditional social media platforms. All valid reasons to migrate… if you would care about these things.

Back to the developer. The developer base of the fediverse after many years is absolutely tiny. Not thousands, not hundreds, but ~150 active developers. And when a new app announces fediverse support, this mostly means they are now capable of sending a message that pops up as a toot in a Mastodon client, or showing a timeline of toots from remote instances. Somewhere along the lines the imagination of what federation support means, has become limited.

Vision (technical). What is the breadth and scope of what I can achieve on the fediverse as a developer? How will the ecosystem we thus create look like in 5 years? How will all these different development initiatives integrate and intertwine? What crazy potential does this unleash? How will it change the web as a whole?

Vision (social). What can I do on this new social space? How does it change interaction I can have with others online? In what way is this different and unique in comparison to traditional social media I already use? How is it more social? How will this impact my life?

There are very few fedizens that I see pondering these questions, exploring and advocating the boundaries of what’s possible with the technology. Most people take fediverse as-is and effectively say “Come to us, this is great”. The devs that venture in federated apps find it difficult to onboard themselves, they need to overcome significant hurdles before they can count themselves among the elite group of ‘fediverse experts’. The average developer - as Microsoft characterizes them as “the 99% developers” - will likely not persevere and reach this stage. They either stick to contributing to an already established popular app, or they just move on to the next project.

Conclusion?

I leave conclusion to the reader of this rather long explanation. This is how I look upon the current situation, with a lot of advocacy experience under my belt. I see many people thinking that fediverse is a “done deal”, that the tech and ecosystem will only grow and flourish from now on. I am not so sure about that. Most importantly I don’t see strength in the ecosystem and ‘substrate formation’. This makes me fear that fedi will only manage to become interesting… to be exploited by the powers that be (i.e. corporate takeover, similar to the regular web).

At least in your reply I am happy to see that you were thinking of the fields where federation can be applied. Those are use cases, not vision, but it is a ramp up to imagining what an attractive shared vision could look like.

Lastly I want to mention something about Metaverse. Now, I personally find this a rather uninspiring, bloodless concept coming from Facebook and all. The name itself is a bit ‘void of humanity’ and dystopic (very fitting for FB). But what it has going for it is that most devs have an idea of what VR/AR virtual worlds would look like, have gaming experience that show how far you can take your imagination and creativity in such worlds, and hence that is an enormous boost for this future direction. And of course it helps that corporate world pumps big $$$ in PR and marketing into it, is ready to pounce.

We need something similar for fedi. That’s all.

This Lemmy discussion triggered me to send the following toot:

Hi there #Fediverse

Today a very simple question for y’all fine fedizens to ponder if you like:

"What is the Vision of the Fediverse?"

(Try to stick within the char limit)

And one day later I posted a follow-up to it, as I got some very nice responses from various fedizens. Quoting below:

Thanks to all of ya who responded to my question thus far on:

“What is the #Vision of the #Fediverse?”

Some interesting takes, and also the observation that its personal and there are many visions.

Imho we need ask such questions more.

• What do we like, want and need on the fedi?
• Are we going there?
• Do we have shared vision, common cause?

And things like…

• Where might fedi be in 2 years time?
• In 5 years?
• Is that likely to happen?
• What is needed to get there?
• What’s missing?
• How can I help?

The observation “there is no singular vision” by @skipfordj is a valid one.

• But isn’t there shared vision besides that?
• Should there be?
• Maybe shared technology vision, such as in the #SocialHub slogan #SocialNetworkingReimagined?
• Or a shared cultural vision, such as expressed in #UnitedInDiversity
• What technology vision inspires the most progress towards getting the fedi we want?
• And what shared vision allows us to have, retain and strengthen our culture and values, respect our diversity?

So why did I write this toot in the first place?

My observation is one where the #Fediverse technology adoption lifecycle is broken due to the lack of healthy substrate - people and processes to evolve the ecosystem as a whole, maintain open standards, ensure broad #interoperability.

That’s technical vision related. It hampers innovation and the imagination of what fedi can be.

It leads to risks for fedi as a whole. That fedi stalls, or will be exploited.

[aschrijver]

I just responded to interesting follow-up on the technology vision discussion that I posted on SocialHub:

https://socialhub.activitypub.rocks/t/what-is-the-vision-of-the-fediverse/2356/6?u=aschrijver

[aschrijver]

In response to @weex stating on Fediverse Town:

Lots of responses and a fair amount of consensus but I agree that there’s not going to be one vision. That’s why free software’s so good for this, each admin can check the boxes they like and implement their own vision or even let it emerge from community concerns.

I responded with:

I wholly agree there will be no one vision. But what I am aiming is to have people think of an outlook, a perspective, and an insight of what they can achieve with the technology. And this is something that too few people think about.

Right now ActivityPub / ActivityStreams are perceived to be just some low-level open standards, plumbing technology, that you add to your app to add microblogging-like features and now you are part of the Fediverse. We have a mindset where this is all we’ll ever get, if we don’t manage to reposition fedi in our thinking.

The way we think about Fediverse right now is killing the imagination of what we can do with it, what we can achieve, and the great opportunity that it offers. And while the (technological) evolution of the Fediverse grinds on at an excruciating slow pace, this is not the case for many competing technologies (ones that are much less to our liking, and fitting to how we want our online world to be).

Right now we are on a firm path to squandering our unique opportunity, of that I have become convinced. @weex you counted yourself that there’s about ~125 active developers creating the software that upholds the whole fedi. That is nothing in terms of adoption after so many years. What one medium-sized IT company can bring to the table easily.

@weex

implement their own vision or even let it emerge from community concerns.

This is all fine and well, and highly encouraged. But if every implementation does it their own way and without collaborating, feeding back their insights, then fedi will be fragmenting, stalling and eventually languish. Or maybe I should say the potential of the fedi shall disappear, because the currently successful microblogging apps might well continue to be very popular. Individual app adoption lifecycles have less of a problem to bring a project to maturity.

I’m hoping we can collectively form a ‘Peopleverse’ where the re-thinking of what social truly means goes next-level…

Fediverse (technical) → Peopleverse (social)

• Social Networking Reimagined (technical): Strategic direction. Technology vision.
• United in Diversity (social): Desired outcome.

I have to admit. It is getting harder for me to advocate this perspective. I feel that I am bothering people with it, trolling y’all. Am I seeing things wrong? Are we all fine and dandy? Shouldn’t I worry? My only observation is that many of the types of application I’d like to see on the fediverse do not yet exist, nor are they being built right now. And that our tiny development community is discussing single extensions like common Groups support for 2.5 years without a final result yet.

[aschrijver]

Strypey tooted about Perspect3vism, a Web3 company who are developing:

AD4M

The Agent-Centric Distributed Application Meta-ontology or just: Agent-Centric DApp Meta-ontology

• A new meta-ontology for interoperable, decentralized application design
• A spanning-layer to enable seamless integration between Holochain DNAs, blockchains, linked-data structures/ontologies and centralized back-ends
• The basis for turning distinct, monolithic and siloed apps into a global, open and interoperable sense-making network

Though I’m very critical about much of the Web3 space, and deliberately staying well away from it (perfectly happy to be a late majority if any technologies find real adoption), there is something in that Web3 projects do way better than Fediverse projects in general:

• Thinking way out of the box, in radically innovative new directions.
• Productizing their project websites to communicate their vision.

Founder of Perspect3vism is the former tech lead of Holochain Nicolas Luck, and while the website is very buzzword-laden its main message of a new “Human to Human” (H2H) web is I think very inspiring and the overall direction sure is visionary. If we look at features:

Perspect3vism User Features

• No more middle men: Direct and create your own web experiences
• Personalised perspectives: Say goodbye to apps and websites and say hello to your own hyper-personalised Perspectives.
• Self-organise your community: Decide your own rules and standards for coming together.
• Own your media: Retain ownership of your data and online life.
• Drop the silos: All Perspectives are highly interoperable and can ‘talk to each other’.
• Interoperable accounts: Decentralised Digital Identifiers (DIDs), access to multiple web experiences via single log-on.

We can see some overlap and alignment to Fediverse, but also this going well beyond in innovation and ambition. Same is true for their social platform project Flux:

Flux: A social toolkit for the new internet

Create your own decentralized community.

Holistic social infrastructure: Flux interacts with a complete agent centric ontology for social communication agnostic to backend. This allows us greater flexibilty and composability without being technologically restricted.

Decentralized: All communication is completely P2P & decentralized. This protects you from authorities snooping on your conversations whilst also protecting you from censorship and control. You own the experience and data, take charge!

Highly customizable: Each interface element of Flux is made of modular components conforming to a common interface. This allows flux to be highly composable suiting your style, needs or curiosities.

Web3 Integrations: Flux is built to be ready for Web3 with integration to your favourite platforms. Easily leverage Flux as a contact list for crypto payments… Or import proof of asset ownership and leverage for permission management & flexing

Distributed Governance: Integrate with existing DAO tools to build consensus within your community with greater transparency and cohesion.

Asset ownership: Collectively manage community treasuries to build your empire. Charge entry fees for joining communities and deploy these funds as you see best!

There’s a bunch of Web3 takes and terminology that I generally dislike. But again there’s a kind of “completeness of vision” shining through that is generally lacking on the Fediverse. I also feel resistance to the way that monetization is done in Web3 (Wild-West nature of the technology landscape being the cause), but having the option to design sustainable revenue models is a plus. And also monetary incentives (“I want to earn some income from my content creation / my own work”) are attractors of people to these platforms.

All in all more common vision on the Fediverse would be a boon, imho. The creator of Slow Social, a social network built for friends, not influencers puts it well in this HN comment:

Has any social network ever focused on users well-being and been successful? I think it’s not criticism but an interesting idea.

I too, am curious. It seems like Mastodon and other federated networks definitely have user well being and autonomy as a goal. However, for whatever reason, most of the more popular federated networks seem to aspire to be “X, but federated” instead of having drastically different user experiences.

[WisTex]

As a relative outsider who has recently become active in the community, I don’t see the lack of a common vision as a problem. The problem is interoperability.

We can look at the most successful decentralized communication system in existence, which is email, and see what is working and what is not.

It doesn’t matter if you have an open source email client, a proprietary one, a customer relationship management (CRM) system, a ticket system, or marketing system, an autoresponder, etc. The apps are very very diverse and have very different use cases… yet the email still gets sent and delivered to all of these diverse systems.

ActivityPub and other protocols are vital for making these competing visions work together. If we can get the protocols to the point where it does not matter which app you are using, then we will start getting somewhere.

But my impression is that most fediverse platforms want to promote connecting between instances of their own platform, and being compatible with other fediverse platforms is a secondary priority if a priority at all. And even when they are technically compatible, there usually is little user documentation or interface cues to let people know they can connect to competing platforms.

So, from my perspective, interoperability is what needs to be prioritized. The apps themselves can be quite diverse.

[aschrijver]

@WisTex

As a relative outsider who has recently become active in the community, I don’t see the lack of a common vision as a problem. The problem is interoperability.

It is partially technology vision, and partially a product vision of “what the Fediverse could be”, on both of which there looks to be a lack of imagination in general (though, of course, some projects are wildly imaginative and don’t fit this general label).

Wrt technology vision most people look at what is implement… Microblogging apps… and then move ahead with a “How can I cram my non-microblogging app into this?” approach. As a result no one has looked well at how to really build vocabulary extensions and benefit from the Linked Data aspects of ActivityPub before. Many people - exposed to the complexity that’s increased by protocol decay - fall back on doing things the way Mastodon does them, i.e. post-facto interoperability.

Wrt product vision I see people endlessly stuck in talking about Instances, federation, accounts, timelines, and a whole host of features that are 1-to-1 copies of what corporate social media offer. But there is so much more, and there are higher / different levels of abstractions, numerous business domains that can be modeled on the Fediverse. If you consider all the online services on the internet where people interact… they are all forms of social networking. This is a wild potential!

@WisTex

It doesn’t matter if you have an open source email client, a proprietary one, a customer relationship management (CRM) system, a ticket system, or marketing system, an autoresponder, etc. The apps are very very diverse and have very different use cases… yet the email still gets sent and delivered to all of these diverse systems.

This demonstrates a point made above. The talk about instances and federation we do on the fedi… those are implementation details. Like mentioning SMTP all the time. In email indeed you have many different uses. Delta Chat is a good example of how far beyond what most people understand with “email” you can go. Those out-of-the-box uses should be encouraged. This has been and will remain a big part of my fediverse advocacy.

@WisTex

ActivityPub and other protocols are vital for making these competing visions work together. If we can get the protocols to the point where it does not matter which app you are using, then we will start getting somewhere.

I’ve said this myself for a long time, but recently found that I should revise that so as not to raise unreasonable expectations. ActivityPub does not and will not facilitate this. Broad interoperability like that requires massive and prolonged collaboration across all the different domains that are involved.

But that doesn’t mean we can’t come a long way The blog posts by Dennis Schubert of Diaspora* mentioned in Challenge: Healthy substrate formation for the Fediverse made me see that it is better to talk about AS/AP et al as providing a Framework with which to become part of the Fediverse. But your interoperability depends how you collaborate with the ecosystem partners in your own and other domains.

A great example of a new ecoystem forming and in need of ‘substrate’ is forge federation and I advocate forming an Ecosystem Alliance (of which Social Coding will be a part) to evolve this area.

Btw. “federating forges” is a good example of setting your vision appropriately. Federating forges means adding Gitea, Sourcehut, Gitlab etc. to the Fediverse. But what is a “forge”, really? A rather arbitrary collection of online software development tools. A better vision would be to dedicate to “federating software development”, and then you are on the level where Github is also positioning their product portfolio.

@WisTex

But my impression is that most fediverse platforms want to promote connecting between instances of their own platform, and being compatible with other fediverse platforms is a secondary priority if a priority at all. And even when they are technically compatible, there usually is little user documentation or interface cues to let people know they can connect to competing platforms.

Indeed. It is understandable, as it is the easiest way to start and make progress. Documentation is the eternal tech debt in any software project, always And the interoperability with other apps, means overcoming the Protocol Decay, more tech debt. And maintaining it continuously as arbitrary ad-hoc interop elsewhere creates breakages. A situation that will become unmaintainable in the future.

[aschrijver]

On Corporate Takeover

Copying my current take on corporate takeover risk from Social Coding FSDL chatroom:

As I see it the corporate takeover is a gradual trend that has set in. It hasn’t materialized yet, and maybe it never will. But whether it won’t continue is not under the control of the People, but depends on the corporations and the money-making potential they see in the Fediverse. If interests in federated apps continues to grow, there aren’t any real defense lines to avoid a corporate takeover from happening. Of course this ‘looking into the future’ is full of speculation, and the speed with which things may go sour greatly depends on the larger players like Meta and the moves they make. Thing is these corporations hold their cards very close to their chest, preparing their moves.

The Commons OTOH is an open book. For example, the mere news of Meta considering AP has made chaos erupt, and any kind of product feedback that Meta would ever need has been broadly discussed at length, including all the ways they may benefit.

Meta taking longer to roll out AP support, or maybe canceling this? Who knows. Maybe they internally upgraded AP support from minor roadmap thing of one small Chains team, to “we’ll take the lead in Fediverse technology space” and even Meta needs more time to prepare that well.

Thing is: The initiative is theirs, not ours. If it happens it is on their conditions, not ours. The ActivityPub spec is sufficiently broad, flexible and open that “AP-compliance” is not a meaningful concept. AP support is individual projects doing their own thing, most often starting from post-facto interop of Mastodon stuff. Fediverse is project-level fragmented, and its defenses consist of project-level “moats” (which are tiny, except for Mastodon’s maybe).

I gauge the risk such that I no longer advocate (technology) vision publicly, to not inspire the wrong folks.