If ActivityPub can't survive Meta, it was never going to succeed in the first place
anthoniix@kbin.social to 
Fediverse@kbin.social – 120 points – Corporations don't just sit out on new technologies, and no matter how hard you try you can't force them to. Defederating from Meta's new project preemptively is naive, and will not do much of anything.
Protocols are going to be adopted by corporations, whether we like it or not. SMTP, LDAP, HTTP, IP and 802.11 are all examples of that. If it ends up that meta is able to destroy the fediverse simply by joining it, that is a design flaw on OUR end. Something would then clearly need to be different in order to prevent future abuse of the protocol.
FOSS is propped up by corporations. By for profit corporations. If you want to stop those corporations from killing projects, you put safety guards up to make sure that doesn't happen. You don't just shut them out and put your head in the sand.
“Simply by joining it” is not an accurate representation of what will happen in the slightest. Meta is not some scrappy little Lemmy instance operator relying on donations to keep the lights on, they’re one of the biggest companies in the world who simply do not care about fair competition or open standards, and they have a proven track record of using that position to either buy out or destroy competition.
When Meta have so much money that they can simply outspend any other fediverse platform and become dominant that way, how is that a design flaw on our end? You can make a project as resistant to corporate overreach as you like, infrastructure to run it still costs money and there is no fediverse operator on the face of the earth that is going to be able to outspend Meta when it comes to infrastructure and R&D. How is defederation not an appropriate response when smaller instances are crippled under the inevitable load stemming from Metas users?
Corporations have been embracing, extending and extinguishing FOSS projects in the tech space for decades now, and their demise has rarely been because of a fatal flaw in the projects themselves. It’s been an intentional play by Microsoft, Google et al to ensure that there is no viable open alternative to their walled gardens. Trusting them in any capacity is naïve at best and catastrophic at worst.
I encourage you to read this blog post which outlines these concerns much better than I can: https://ploum.net/2023-06-23-how-to-kill-decentralised-networks.html
When this was linked a previous time, I wrote up a reply to it which I think applies here as well, so I'm gonna shamelessly copy and paste myself 🙂
To your point, many of us will defederate, either out of politics or financial necessity. There's rumors milling around that Meta / Threads will only initially federate with a few trusted larger instances and be monetarily compensated for it (aka those who will make a deal with the devil to moderate - more on my thoughts on that here).
It may come to the point where a lot of us are running on our own smaller "Fediverse", intentionally divorced from Meta and those instances which have federated with Meta and taken their advertisements and paid posts. If this is the case, we must take the bad with the good - we will always be smaller and niche, and our less techno-idealistic friends will not join our tiny Fediverse because the barrier to entry will remain high.
It will follow the EEE flow along with their normal anti-competition tactics. First, they embrace: their interest in federation is only to give them the access to content that will make their platform not look empty, allowing them to put their coffers to work on drawing the majority share of users. Then they will extend: they will make sure their platform is compatible with ingesting other server content but others will be unable to federate their content (they will become "incompatible" later, due to "features"). Then they will extinguish competition: they'll cut off what little engagement is left with those (inbound only) federated servers because they no longer need them and the majority of the remaining users will move to their platform because that is where the activity is.
Then Kbin/lemmy will be just like all the other random phpbb instances that no one really uses. Being naive won't make things any less likely, yet there will always be gullible people who argue that "of course they will embrace the technology" and that everything else is just non-sense/wouldn't have worked anyway/blah.
It doesn't take long for the largest servers to have operating costs that they will happily allow Meta to burden in exchange for nearly any concession. The main problem is that, while Kbin/Lemmy is federated, it is federated in a manner that still places content in silos and allows single servers to "own" those spaces. It hasn't really fixed the problem yet, it just spreads the problem out over a few more servers. Until spaces are universal (every server owns a slice of that community, spreading out the community instead of just the users), it will remain ripe for EEE.
Completely agreed on the concept of needing to federate out entire spaces (magazines / communities) - this is a huge gap currently, especially on kbin where nearly every community (and nearly every user) is on kbin.social .
My thought is that Lemmy and kbin, after fixing and updating core functionality (easier said than done), need to jump on the idea of instances having magazines/communities that pull from multiple other sources, rather than federating each community independently - e.g. I could have an @android which is pulling from @android , @android , etc, and the experience is relatively seamless - if someone drops out, yes we lose a lot of content, but others pick up the slack.
This would require more overhead from admins and instance owners to manage which other communities their own communities are pulling from, but I think it would be worthwhile for a better overall user experience and to help decentralize these communities in general.
I can't edit and have my edits federated out (edits federate if someone is following you, not if you are following them) so I want to point out that those links are somehow pointing to users instead of communities, whoops. I was referring to e.g. https://the.coolest.zone/m/android being a collection including contents on that magazine, contents from https://kbin.social/m/android , contents from https://lemmy.world/c/android , etc.
... so I could host an instance with a magazine that is essentially a digest of federated instances, and I can add/remove based on whatever aligns with my principles?
I don't think I have that right, sorry but if you could explain it a little differently, I'm still trying to understand how the fediverse is.
So, as an example. I am on the.coolest.zone which is where my account is registered. I have a magazine on it called fediverse@kbin.social, which is where I am viewing this thread. In fact, you will be able to view it here: https://the.coolest.zone/m/fediverse@kbin.social
I have only three actual magazines on my own instance - random (this is necessary for all kbin instances to collect uncategorized posts), BestOfBlind, and Android (which was my attempt to create a magazine that collects threads based on hashtag, but it's not working right). Everything else is a magazine which is actually a federated version of either a kbin magazine from another kbin community or a Lemmy community from a Lemmy instance.
The couple of things that seem to be missing or broken right now:
kbin and Lemmy are both very new applications, so these will likely shake themselves out over time, but it's a bit of a rough experience right now. 😅
But kbin.social is fully compatible with Lemmy with almost the same number of users and many more communities (dozen of them has more subscribers than most-subscribed /kbin magazine). Maybe /kbin as a platform is much centralised. Threadiverse, not so much.
I'm afraid we will lose if we accept them until they do something bad. Given their track record, that's just a matter of time. If we let them become important to the fediverse as long as they play nice, the final decision could be disastrous for the fediverse when they stop playing nice.
So the right thing to say seems for me: "No, this is our house. We don’t federate with you."
That is a completely valid take here. My partner who runs our Mastodon instance will be preemptively defederating with Threads (on my suggestion), so I do agree with you, but I realize not everyone in the Fediverse may share that take - it's a weighted scale where one end is "mass adoption of a Web 3.0 decentralized Fediverse" and the other end is "but adoption in which most people are on Threads will be centralization anyway, so we will have already failed."
I think in any case it may not matter, as I believe Meta / Threads will only federate with instances that agree to follow their moderation standards - after all, Meta likely doesn't want porn and Nazis federated to their communities because then they can't run advertisements. As a Fediverse community, we're pretty good at taking care of the Nazi thing, but Mastodon's got an awful lot of porn on it.
It will really depend on which admins take that deal to be beholden to Meta's standards, potentially opening themselves up not only to huge moderation concerns but to a future requirement of taking advertisements. I hope large instances will not. I would prefer to see the Fediverse operate separate from Threads, who will be using the ActivityPub protocol but not part of the larger Fediverse. Similar to how the conservative "truth.social" uses the Mastodon application and ActivityPub but is not part of the Fediverse because it is closed off.
A little off topic, but I was very surprised Reddit didn't pursue a similar approach of "we will lower greatly the costs of API but we will serve advertisements in the API as regular posts, so you must display them and can't strip them out."
Why will Meta care? To their large user base, they are on the "federation" and we'll be the odd ones out. Their users won't care either. They'll just use it. And Meta can spin up any number of servers they want. Any company can.
They might not, and frankly that would be the ideal solution, wouldn't it? They fork Mastodon, have their own isolated little universe, and we have our own little Fediverse over here. But the rumors swirling indicate that they do want to federate with larger instances that will be willing to follow their moderation standards and optionally take advertisements - I explain over here why that is likely the preferable approach for Meta.
Echoing this, please read the linked post. There is a big difference between technology and it's implementation vs the community of users of it and what they are using to do so.
Strongly agree, any instance that doesn't defederate meta should be defederated. Meta will forcefully overrun the fediverse for profit.
It's a design flaw when simply "outspending" other fediverse platforms allows you to dominate them. There are ways to design a system where that's not possible.
The fact that those other fediverse platforms can defederate from the "big money instance" if they don't like what it's doing, for example, is a part of the design of the fediverse that can help counter influence like that. You can't force other instances to federate with you even if you have an enormous amount of money, and even if you did manage it in specific cases other people without that vulnerability can just spin up new instances.
We'll see whether this sort of thing is "enough", I guess, because Meta is coming one way or the other. If it turns out that stronger defenses are needed then there are other technical methods that could be used to strengthen the decentralized nature of the fediverse.
I think defederation is not really that useful in this case, because then your users will just leave and sign up for the platform where they can view where the most content is. Although I do agree with your general premise.
Then it's no different than them just jumping back to Facebook or other corporate owned social media like the current situation. That's not a lost. The lost is if they are allowed to federate in the first place and people get used to it. But even then as the concept of enshittification becomes more well known more users will also be resistant to the idea.
If Meta is actually providing the content that people want, then what's the problem?
If it's not, then people won't leave your instance to go sign up for a different one that's not suiting their needs.
The situation I'm thinking of is one where Meta creates Threads (or whatever it will be called), and then a bunch of people defederate. In that scenario, there will of course be big servers who choose to federate with Threads. Given Meta's reach and influence, they will undoubtedly have one of the bigger instances, so a lot of politicians, journalists and everyday people will go there.
Making it so people can't see that content will just make the fediverse become more centralized, because people will just go to the bigger instances that will allow for them to see that content, or just go sign up for threads. I think that's bad because it creates further centralization, even if they're providing the content that people want.
Even though I know a lot of people disagree, we need all types of content in order for this place to grow. I'm not talking about any far-right nonsense, but even garbage like tabloid fodder and stupid meme bullshit will keep our networks alive and users engaging. The easier it is for the average person to use the better. If the point is not profit, then it must be to allow people to come together and talk about almost whatever with almost whoever, and wherever.
Perhaps I'm misunderstanding something, but it sounds to me like you're still saying "if Meta is providing the content that people want and it's not available elsewhere then people will go to Meta for it."
And that seems fine by me, in that case they're competing by providing the content that people want. However, not everyone wants the same thing. I have no interest in "following" politicians or celebrities. A lot of the sorts of conversations I'm interested in do not benefit from having a huge number of people in them. So even if Meta is some kind of juggernaut there's going to be people (like me) who don't want to participate in a juggernaut. The smaller defederated instances will still be attractive.
I agree with OP's title on the social side of the problem, not the technical one. If we allow them to destroy the Fediverse, then it was already lost to begin with. It's not a matter of technology, it's a matter of whether the key people are able to keep it out of the corporate control in the long run. If they can't, then it was all just a matter of time.
EDIT: I don't imply it's a particularly useful thought. It might help with coping though if it would ever happen. Let's enjoy it while it lasts and hope for the best!
Similarly, if the Earth can't survive Exxon, it was never going to succeed in the first place.
I just have to keep on hammering this point, because it pisses me off so, so much. Many people seem to believe that, since regulatory bodies can be captured, that regulation shouldn't be done. This is called learned helplessness, and it's something malicious people inflict on people they want to exploit.
It isn't sticking your head in the sand to resist assimilation by an evil corporation.
Actually, yes. The reason Exxon is fucking the planet right now is because of weak regulation. If we can't build a system that is resistant to the threat of earth destroying corporations, we were never going to succeed in the first place.
Your post is arguing (by analogy) that we shouldn't even bother trying. But I guess you don't need a suicide note when you can just leave a copy of Atlas Shrugged by your body.
On the contrary, I'm just saying if you build something and it gets co-opted by a corporation it probably wasnt meant to be.
It's like when people talk about politicians being bought out by corporations. If that's something that can even happen, it's the fault of a broken system that would even allow that to happen.
This is a very computer sciencey view, which is why I leapt past the intermediate logic straight to its conclusion. But I'll spell it out.
There is no rules-based system that will actually stand in the way of determined, clever, malicious actors. To put it in CS-style terms, you'll never cover all the contingencies. To put it in more realistic terms, control systems only work within certain domains of the thing being controlled; partly this is because you start getting feedback and second-order effects, and partly it's because there's a ton of stuff about the world you just don't know.
If a system is used as intended, it can work out fine. If someone is determined to break a system, they will.
This is why the world is not driven by rules-based systems, but by politics. We're capable of rich and dynamic responses to problems, even unanticipated problems. Which is to say, the only actual solution to Exxon and Meta is to fight back, not to bemoan the inadequacy of systems.
Indeed, this belief in technocracy is explicitly encouraged by malicious elites, who are aware that they can subvert a technocracy.
What he said is right though. Let me put it this way, politics has a system it's relied on. Ancient Greece has its own style of democracy. Current US has its own style of democracy. The EU has its own kind of system. Here's the thing to consider, the content and state of the system can change over time, but the low level of it - the rule to how the changes can happens or how things operate - rarely changes. Politics can change the rules within the system, but it doesn't typically change or revise the foundation of the system. When revision of the system foundation is so rarely done, the things taking advantage of this foundation obviously don't get solved.
When you say someone who wants to break a system will, it's actually because the base of the system doesn't change so the abuse can keep happening. Let's use US politic as an example. Gerrymandering is a problem. There's no sign of it getting fixed and continue to be a problem even now. The reason is because the current system had made it so that the decision to do so could never come to past, at least not easily. It's a deadlock. If instead the system is revised from the ground up this would be as simple as reasoning during the redesign process that the current method is broken and it isn't good at representing the people so it should not be used. Currently that's not how it's being solved, and it's like trying to fix a problem on your computer without the option to shut down or reset the device.
What he's saying is the system is broken like that and we're not solving it by the most efficient method (mainly due to it being so costly). Even so, sometimes it's just better to scrap and start anew.
That said, I don't think it applies to Fediverse at the moment. It's so new that there are so many ways it could develop and if it fails that doesn't imply the concept of Fediverse is never supposed to succeed. It may just be because the best steps to manage it wasn't taken. Going back to the political analogy, it's like having just the concept of democracy as a framework. But it hasn't been decided yet whether this democracy is going to be dominated by just two parties (like the US), or has many different parties with ranked voting (like the UK). Both are democracies but the foundations and implementations are different. And well, one works better than the other.
Things are politically stagnant because people believe that politics is about systems. Politics is about power, and politics will always be an expression of the dominant power dynamics. Governmental systems are just how power is explained to outsiders; it's a mythology that's told to disguise the real nature of power.
So the question of systems is a red herring, that's been carefully instilled. This has been true for all history: Many kings don't really rule, courtiers do. Only kings who can effectively wield power rule, and they're historically in the minority. This should also be obvious in the US: corporate power is only ever checked in the presence of enormous public action. Not public bitching, public action—general strikes being the most important example.
Or to put it really bluntly, while there's a lot of pageantry in politics, what politics actually is, is power struggles. But they sure don't want people to recognize this, which is why there's so much pageantry and partisanship.
This is also why the government is going so hard against Trump, but letting Pence, Clinton, and Biden slide. It's not because they cooperated—if you or I had security clearances and just took documents out of a SCIF and kept them at home, we'd be in jail. It's because Trump clumsily challenged existing power, namely the federal bureaucracy (which he conspiratorially calls the "deep state"), and he wasn't up to the task.
You fight back by fixing the system, or making a new one.
Do you apply this reasoning to everything in life?
If a house catches on fire, it's because it has weak fire suppression?
Meta is allowed to use the ActivityPub standard just as much as any other standard. This does not mean anyone who decides to use it must interact with others who use it. SMTP will block your mail if you aren't from a larger server, have the right signatures and even then. Servers block HTTP over VPNs often, and there are even rules about referencing content via other servers on HTTP (CORS). Just because a standard is open doesn't mean everything using that standard has to communicate with each other.
The beauty of this is that those running instances can't restrict access of other instances to the fediverse. If Meta does start using ActivityPub, every current instance can block it. Other entities could want to run an instance that federates with Meta who has the resources to do so. Currently the biggest issue is the vast difference in scale between current instances and Meta. But if other entities got into the fediverse that federated with Meta this would still be a decentralized system, just with larger nodes between them. All of this still allows those who run small instances to block these larger instances that are more mainstream and keep it the way they want it.
No trying to be explicitly contrarian, but the EEE strategy (embrace, extend, extinguish) is well known by this point and it always ended up with the open standard not being used anymore and falling into irrelevance (as it happened to XMPP after google and Facebook embraced).
I do think it is a design failure, but it’s one that is necessary for it to be open: anyone can enter the space and build features on top of it. So they bring a lot of people, with features exclusive to them and then lots of people migrate because the experience feels broken if you can’t “florp” a post from someone else. It’s the nature of open source vs closed platform that enables the strategy to exist.
It may not happen this time, and I surely hope you’re right, but it would be a shame for the monopoly to win one more time when we had the chance do to something about it but we didn’t. Bringing more people do the fediverse sounds like a dream, but I’m not holding my breath expecting everything will work for the best. There’s a reason they’re doing this, it’s not because they need more users, they already have all of them.
This isn't even remotely true, there are plenty of counterexamples. TCP/IP, HTTP, HTML, XML, SMTP, PNG, SVG, OpenDocument, OGG, PDF, FLAC, WebM, Vorbis, I could go on at great length. There are a vast number of open standards that are still open and are used extensively as fundamental parts of our everyday lives. Eg, the IEEE standards and RFCs.
How compatible is microsoft office with it?
And how many PDFs are broken by Adobes bs?
I don't use Microsoft Office, so I don't know. Same with PDFs, I don't use Adobe.
There are plenty of programs that handle those formats other than Office and Adobe and I've never had significant problems with any of them.
Regardless, OP said "it always ended up with the open standard not being used anymore." Both of those standards are being used, and even if OpenDocument and PDF had been extinguished it still wouldn't matter because any open standard still being used is sufficient to disprove OP's position.
Emrace-Extend-Extinguish is something worth paying some concern to, but it's not some kind of unstoppable boogeyman. It's failed far more often than it's succeeded.
Same, and neither of us is having a particularly fun time when we get sent a .docx or a strange PDF that refuses to be opened and correctly edited without Adobe Acrobat (not joking, this happened to me once).
Fair point. But both of them suffer from being adopted and then maligned by the corporate entities that picked them up, which is a part of the cycle. First it gets adopted, then "slightly tweaked for features" and then its (usually) unusable on other platforms that dont want to adapt to the corporate vision. Granted, OpenDocument has a different history and both it and PDF are old enough to not actually fear that they will be replaced be corporate "alternatives", but that they did generally follow the cycle but didnt finish it doesnt necessarily disprove it imo.
Absolutely. But that doesnt mean we have to attempt giving it a fair fight. They dont intend to either, its antithetical to the point of EEE.
ODF has been supported natively by Office for years now, and LibreOffice is able to open .docx files just fine.
I've never found a PDF "broken by Adobes bs".
Open, yes, but the formatting will be terrible in my experience.
It has happened to me. Wasnt impossible to solve in the end, but still.
Listen buddy, I don’t think you understand how important it is that I florp my second cousin’s ex boyfriend’s post on the hot new bug patties at McDonalds on sale for $12.99. If I do I get a free bottle of fresh air and a complimentary upper body rinse in the labor hydration station next to the bathrooms.
XMPP was irrelevant before Google and Facebook had anything to do with it.
Agree.
That's fine if Meta wants to use an open standard for it's post data. But just because our platform can talk to them doesn't mean we want to hear what their users are saying!
This post is just...so ignorant and unintentionally malicious. Putting in the idea that Fediverse should co-exist with Meta and not treat them like exile when they enter the space is not the way to go for its survival. And the more people there are who just look at the headline without doing further reading is not going to help the situation.
What are we going to do then, everytime a corporation starts up an instance we defederate? All corporations are essentially evil. If we do that, we'll always just be a niche concept that will always fail to keep up with the needs and wants of users.
We need to be able to prevent bad behavior from taking over the project, while also allowing corporations to join and interact with us.
There's nothing wrong with being a separate entity from big corporates. Mastodon is pretty big already and it's going keep growing. Fast growth isn't the point, it's sustainability and providing a haven for people who don't want enshittification and other corporate bullshit. If someone want that they can just use a corporate owned social media, nothing wrong with that. Reddit has never been the biggest fish in the pond anyway, and the whole thing with trying to pull in users is a part of what made the platform worse. Bringing in corporates is gutting the biggest thing that make Fediverse appealing. And for the record, there isn't that many big companies that'd be a threat anyway. Any corporates that aren't big social media corps isn't really the problem. If Nintendo wants to make an instance to be the second Mii verse I highly doubt many people would be against federating with them.
"If your body can't survive toxic poison, it was never going to survive in the first place"
We should not be federating with Meta or any other corporate poison factories in the first place.
This is not a good comparison. Our bodies are not engineered by anyone, but our software is.
i fail to see why that matters?
You can try to engineer something to be resistant to certain threats, but the human body isn't like that.
With the human body comparison, you can't do much but react to what's happening to you and try to fix it or prevent it from happening in the first place in another human body.
With software, you specifically choose almost every aspect of how it's going to work. This allows you to construct it in certain ways that make it resistant to certain threats and modify it as needed.
OK, I accept your explanation and how you specifically referred to 'ActivityPub'.
This is a misunderstanding of the issue. ActivityPub the standard will of course survive the extension of use to Meta. ActivityPub is just a protocol. There are reasonable concerns about the distortion to the priorities and focus of the projects development long term if Meta gets more involved but those are not unresolvable, and have been seen in other open source projects when big business gets involved.
People's objections are also due to how federating with Meta's software, content and commercial priorities will distort the existing communities in Fediverse/Federation/Threadiverse.
While the technology allows the Fediverse to exist, one federation is not the be all and end all of the fediverse; it is understandable that existing parts of the Fediverse might not want to federate with Meta. There can be multiple federations within the fediverse; it doesn't have to be one joined up mass and it's not likely it's going to be one single complete federation long term. People will naturally form interconnected communities around different ethoses, priorities etc. I think there will likely be one big "primary" part of the fediverse just due to shear size and mass of content; but it doesn't follow that it needs to be linked with Meta's implemenation nor that Meta's implementation would damage a separate fediverse. That is what people are pushing against - they do now want the main threadiverse or other elements of the fediverse to be subsumed and lose it's identity due to being swamped by Meta's social media platforms.
It's about the separation between the fediverse as a technological solution and the fediverse as a philosophical solution.
It's not Meta who is going to want to federate with us, it will be a stealthy instance, for example from a videogame, so popular that everyone will want to federate with it. Including the gamers of our instance. And then this instance (which will be a facade of Meta) will implement a feature in the protocol, for the need of their videogame. Oh something really trivial....then something else. etc. And slowly the federations will diverge toward Meta.
Whats your opinion on Google destroying XMPP.
RIP
more of a question of if activitypub can avoid becoming zucks lapdog and dont allow them to lead the project
Why would I want pushed content from Meta of all places? I don't want to see that.
And no, we don't need to grow the Fediverse that way. It's better to stay independent and develop organically and become the next big thing, not eaten by some overstuffed corporation not able to innovate.
Learn from history. Don't repeat it!
Safety guards like what, do you have some ideas or examples?
I like the concept of copyleft, which has prevented a lot of EEE. As for protocols, the answer is a little more complex. Protocols can't really be copyrighted, so it's essentially going to be what's the easiest to implement, who is using it and what utility it provides.
There have always been competing protocols, and also closed vs open protocols. Most of the time the protocols that win are the open ones, and the trend is that they provide a lot of utility and is easily used by anyone. In my view, the question it will come down to will be: is having a decentralized social network going to provide more utility for the big players, or is the concept doomed because centralization will always provide the biggest monetary incentive?
Something that gives me hope is that social media is not a profitable business venture. This could mean that Meta is exploring the fediverse because it sees something useful in it that doesn't conflict with their business interests, but in fact supports it. The biggest tell to see if this will work out is if other companies start to adopt the protocol, at which point the safety guard is "Well, a lot of big players are using it and if I break activitypub support with them that's bad for business.".