The primary benefit to federated services as I see it is that you can have a network of groups, all with their own policies on moderation and who they federate with. Some corners of the Fediverse block other corners of the Fediverse, but each corner has their own policies on what they block. It’s more democratized that way, and if you disagree with one instance’s moderation policies you can move to another while still interacting with the same communities. Alternatively, maybe you prefer a different interface or way of managing your online presence! There’s any number of reasons you would want to use a federated service.

It makes sense that someone who runs their own centralized service wouldn’t see the need for a federated service, because they can run their centralized service the way they want and don’t have to worry about whether the staff of their chosen service agree with them, because as operator they hold sway over that.

One of the other benefits of an open protocol like ActivityPub is that people can hook into it in the future. For example, Pixelfed may be the only software solution in its class at the moment, but if someone wanted they could make their own independent version and it could be interoperable if they used ActivityPub plus the same conventions that Pixelfed does. Sure a platform could always create their own open protocol, but better to use an established open protocol instead of reinventing the wheel for every new service. That’s even setting aside the possibility the some service will come along to unite all of the Fediverse under one app (Kbin is trying something like this I think).

Also they’re totally wrong about the Kbin situation. Kbin is 100% compatible with Lemmy, if a little glitchy at times (early days, after all). It’s only the main instance that’s not federating, and that’s because they are using Cloudflare to deal with the surge in popularity from Reddit. Under normal circumstances even the main instance can federate.

Most of the socials that exist on the fediverse are still in beta or even alpha, it seems quite a bit too much to ask for complete federation out of the box. Mastodon gained traction last year, just think of how Facebook or Instagram were in their last first years of development

The Fediverse needs improvements, of course. But it also needs time and resources, and only with patience and coordination (and interest and money) we will get to a perfect working point :)

‘Better’ is relative. To me its better because no one person or group owns or controls the software. There’s no central authority. Don’t like the instance you’re on? Just move to another. Cant find an instance you like? Host your own. Don’t like the path the developers are taking? Fork the code. As long as the very core remains standard (ActivityPub), all possibilities are on.

There needs to be a return to being patient. Most fedi software is not beyond beta yet. They will develop and they will mature but right now the fediverse is a toddler learning to walk. There are issues but with time they’ll get addressed. We’ve all got so addicted to shiny cool apps and services we’ve become prepared to sacrifice our privacy, our choices and our reason at the altar of a quick dopamine hit.

There’s no big money to throw at these issues and therefore no dedicated team. This means solutions come slower. But they will come and they will be motivated by usefulness not profit. The people developing these things have lives and day jobs. Give them time.

no one has done an especially good job explaining why the fediverse is better than centralized solutions

The default example people use for “a federated service” shouldn’t be Mastodon or Lemmy.

It should be email.

Why is it better that different companies, universities, and other organizations (and even hobbyists) can all set up their own email servers, rather than everyone just using (say) Hotmail?

1. If Hotmail does something you don’t like, you can switch to Gmail or AOL instead, and you can still send emails to your friends on Hotmail.
2. Different providers can specialize in different things, while remaining compatible. Maybe one provider doesn’t prioritize the features you need; but they can’t prevent a competing one from offering them.
3. No one provider can impose censorship or other overextended control onto the whole system. No one provider can break the whole system when it has an outage: Hotmail going down does not prevent Gmail’s servers from exchanging messages with the University of Tübingen’s servers.
4. Different servers can operate in different parts of the world, under different legal systems. Not everyone is ruled by California or Washington state; or the US. A hobbyist operating an email server in Alabama is not required to comply with Dutch or EU law, and a hobbyist operating an email server in Amsterdam is not required to comply with Alabama or US law. People get to live under the law of their own country; and yet the Alabama mail server and the Amsterdam one can talk to each other.
5. The same infrastructure that supports federation also supports extension of the platform. Programmers can build services on top of email, and the medium is transparent to them. And, again, no one provider can tell them “no, you may not build that weird client program, bot, or mailing-list service.” To be honest, you don’t even get Hotmail if email is born as a centralized service. The whole emergence of webmail services could only happen because email is extensible and federated.

I don’t think the interesting issue is why not centralization. There’s tons of better explanations out there, but seriously, just “Elon Musk” is enough to explain why centralization is bad.

But the post does raise an interesting issue IMHO, and it is the lack of good explanation as to why federation between different platforms with different paradigms. Why federation between different Mastodon servers is obvious. Why federation between Mastodon and Calckey and whatever else is obvious too. Same with federation between different Lemmy instances, or between Lemmy and Kbin. It just makes sense. What is not clear is why we want/need/like federation between Lemmy and Mastodon. Sure, you can post in a Lemmy community from Mastodon, but it sucks. You can follow a Lemmy community from Mastodon, but the experience isn’t great either.

I do think there are good reasons for this, but I haven’t thought enough about it to articulate it properly. My thinking is that while a Mastodon-like service federating with a Lemmy-like service doesn’t seem to make much sense, Mastodon federating with a Facebook-like service does make sense. And I’m not sure if a Facebook-like thing federating with Lemmy makes sense, but I can definitely imagine something sitting somewhere in the middle between those two. And also, perhaps more importantly, we don’t want to erect artificial walls between the different ActivityPub services. Sure, the Mastodon-Lemmy integration sucks, and maybe it shouldn’t exist, but probably nobody will use it much, exactly because it sucks. But if we add a thing saying “no you can’t do it”, then we start needing to define borders between different services. Is microblogging different from blogging? What about a Facebook-like wall? Or a tumblr-like feed? Are those different enough from each other to be different services? Who wants to be the one defining those borders? I think the current solution, where anything is possible and integrations that don’t make sense just don’t happen organically, is the best.

But still, that is a way more itneresting question than just “why federation”.

Considering the current reddit issue, it seems self explanatory why a non centralized solution is better.

I don’t get why we still haven’t figured out that dictatorships are bad.

Reddit is a content dictatorship, federation democratizes it.

It’s really just that simple.

[implementations] are only either partially compatible: pixelfed posts can be viewed and boosted on mastodon but won’t display past the first four images

To be precise, this happens for posts between Mastodon and Pleroma too, and it’s how improbability or standards simply is. Take W3C for example, some web sites work on Chromium, some on Firefox, and most on both. Or ProtonMail has its own E2EE scheme but still can communicate with other providers. Or a cars and horses both runs on asphalt and dirt roads, but some configurations do better than the others.

not compatible at all: afaict, despite kbin being activitypub based, there is no connection between it and any other activitypub software

Not sure about Kbin, perhaps due to its immaturity, but group are known to work across implementations specialized for it (count Lemmy in as well), and those focus on microblogging, or both, e.g. Frendica.

some of this can be easily chalked up to very different modes of interaction but if that’s the case, why advertise as being part of the fediverse when that’s only somewhat true? is it just for Buzzword points?

The ability to access to the same infrastructure (i.e. fedi or all interconnected roads) even if less efficiently regardless of what one has is important for a society where people are truly free to choose where to live and who to do business with.

even some sort of shared identity system (hold on while i reinvent openid) would do a lot to mitigate this! but if i want to use pixelfed, i need a pixelfed account in addition to my mastodon account.

Seconded, it would be ideal if implementations are views of the network from the same endpoint. It would require an overspecifying protocol like Matrix instead of underspecifying like ActivityPub or XMPP for a backend to get hold of all necessary data, but that comes with the cost of making optimizations (which require context) incredibly difficult, e.g. in case of PeerTube. IMHO it’s not something reasonably achievable, but technology is just part of the equation:

if anything, it’s maybe kinda worse because twitter and instagram at least don’t pretend that they’re compatible, and i don’t have to pick a Twitter or an Instagram, there’s just one and i know that it’ll be the same one my friends are on.

(As an outsider I don’t see any functional difference between Twitter and Instagram, but let’s say it’s Facebook vs Spotify), as stated you won’t have to choose between the N friends on X and M friends on Y. Though it might be harder to communicate with those on Y if you’re on X, it’s better than abandoning them. It’s better, yet entirely optional, to be on both.

why the fediverse is better than centralized solutions

More importantly, improbability is future proof. New and presumably doper shits come along and share the same large user base. Old shits can stick around and people their can still connect. No more reliance on network effect and oligopolistic power to force feed ads, mass surveil, manipulate or roll out predatory pricing. Sure, not everyone can make use of the same features, even if that has any significance, socializing ranks higher in the priority list.

many of the problems that exist on centralized platforms (content can disappear at any moment, you are at the whims of the admins, etc etc) exist on the fediverse too, and there aren’t a ton of benefits beyond “you can host your own.”

Ditto, although it’s beneficial to view it in a different angle. Self-hostable isn’t just about everyone can just spawn up a server. That’s just P2P communication when it comes to social networking. Federation is more than that: though fediblock is a thing, being able to choose your alliance means you can have an admin respecting you and being able to connect with people with different associations, all while minimizing maintenance effort.

Decentralized system also means socially decentralized redundancy, which allows for data preservation even when intentionally removed (by a third party; right to be forgotten is another issue which requires cooperation on data handling;-)

“you can export your follower list and force everyone to follow you on a new account” is not account migration. until there is any story for migrating content, claiming that account migration exists is misguided at best and actively deceitful at worst.

Agreed. Just let’s not make perfect be the enemy of good. Not saying there’ll every be practical data migration across service providers, given the immense amount of moderation backlog when lack of personal trust from the admin; only wanna point out no centralized service even allow migrating one’s network.

I like blogs like this. Its important to not stay too comfy in one spot and recognize the flaws in software in order to improve. The comments already try to spot the core issues and offer solutions and its great. I share most of the authors points when it comes to federation. Creating an insta account and sharing your pics on twitter is easy as it gets, you just need two accounts. Im aware this is a growing technology, but I wonder how much of it can be improved, and how much is already set in stone.

It is, of course, odd that the author brushed off descentralization so quickly, especially considering that post was written a couple days ago. Has he not been paying attention to Reddit or Twitter? Sure, if you switch to another instance youll still be under an admin, but you can switch to a saner one. You wont be stuck with the only option. Glad to see this being pointed out in the comment section

The kbin bit is also odd to me. I thought that was just a bug?

Good writeup nonetheless. Wish opinions like this become more common, with the hope that itll be translated into improving the platform