@jwz@mastodon.social 's law in action…

Federation is not the natural unit of organization, that’s a hill I’m ready to die on. The problem is that most people get hung up on the idea of instances because they’ve only experienced the two extremes of siloed networks and the “Dark Forest” of blockchain.

I think the most interesting opportunity here is that ATProto’s strength is that it was built for a “credible exit”, i.e, it allows users to migrate from any centralized network to a decentralized one.

IOW, it would be a lot easier to implement Fediverser on top of ATProto than of ActivityPub.

That’s more for people hosting servers, who are likely to care somewhat.

The whole argument at the top of the thread is “it doesn’t matter how easy/good/performant it is to run a XMPP server, if the people you’ll want to talk to do not have a decent client”.

If I was to make an argument for why someone should switch away from Matrix, I’d just point to how often encryption fails.

That has stopped for me since I started using the element X client (mobile) and Fractal (desktop), which leads me to my “counter argument”, which is: Matrix has always sucked in different ways, and it sucks for everyone equally, but at the least it is slowly and constantly sucking less than it used to. XMPP is adequate (not great) for some (Android, Linux) and disgraceful for others (iOS), but unlike Matrix there is no visible progress for those on the sucky side of XMPP.

That’s like asking why this tractor trailer truck still uses the same old boring shipping containers from 40 years ago.

My point is that most people in 2025 are expecting a lot more from their communications client than what was available in the “good old days” of Gtalk and FB interoperability, and yet most XMPP advocates just bury their heads in the sand and say “it works for me and does everything I need, so there is no reason to add more features”.

What goalposts are being moved?

I’m more concerned with the delays and resource usage, personally.

As you wish. But you do understand that you are part of a very tiny minority and that it’s completely pointless to try to convince people to switch by basing your arguments on the things that align with your values, right?

I think you have a totally false problem analysis if you think the less than ideal iOS support it the thing why XMPP isn’t as popular as Signal for example.

I never said it was the thing, and you are creating a strawman by implying that I’m expecting it be “as popular as Signal”.

What I am saying, plain and simple, is “XMPP might be technically superior in every possible aspect, but this means absolutely jack shit if most if not all of my circles have at 10/20/30% people who will not use it because the client is so feature poor.”

fundamental problems around Matrix

Sorry, this list just re-hashes the usual talking points and cranks the nerdery to 11. “Reliance on JSON” is just an stupid of an argument as criticizing XMPP for using XML. Saying “the main client uses Electron”, presenting it like it is the only alternative (it’s not), focusing on a completely secondary platform (pretty much everyone wants to use a messenger on mobile, so why does it matter to have an optimized client on the desktop?) makes this whole thing seem like it was written by a teenager arguing over racing cars.

in theory the iOS XMPP app experience can be improved relatively easily

Then why isn’t it?! Why is it that the “recommended” client for iOS looks like something from 2005?

Mostly company phones in Europe

Does it matter? It’s not like people everyone carries two phones around.

and I suspect elsewhere too.

Wishful thinking, again.

nice cherry picking of numbers with “western” Europe only…

The point is that there is a non-negligible number of people using iOS everywhere. In Brazil, it’s 17%, In China, it’s ~20%. In Russia, 30%. Saying that “no one uses it outside of the US and Japan” is absurd.

I would tell them to install Monal. It works fine for normal messaging.

If you “know almost no one that uses iOS privately”, you are probably surrounded by people who think that “emoji support” and “reactions” are not important. Try stepping out of this bubble and see how far your can go with Monal. People expect a lot more than “normal messaging”.

And no, I’m not saying this out of hate or because I prefer Matrix. I wish XMPP could be a viable alternative, just like I wish less people were locked on iOS. But no amount of wishing will change the reality. If XMPP enthusiasts really want to make it succeed then unfortunately they need to be ready for an uphill battle and be willing to spend resources to get a compelling alternative working on a hostile platform.

hardly anyone uses iOS.

I keep harping on this point because you are doing wishful thinking. The numbers for iOS usage in Western Europe are around 32%. That’s far from “hardly anyone”. In the UK is over 50%, in Australia it’s over 60%.

If you care about keeping in contact with friends and family, you’ll be bound to find someone on iOS. What are you going to do, tell them to get a new phone just so they can participate in the family group chat?

At this point I almost think not supporting crappy Apple products is a feature.

Yeah, those grapes really are sour…

Last time I tried this I couldn’t even find an iOS client that supported reactions. Hate on the complexity of Matrix all you like, but at least when you ask your family to use it, they get a modern experience and don’t feel like they’re giving up everything vs every other messenger they are used to.

Glad to see I’m not the only one saying that.

May I interest you in !minecraft@level-up.zone instead? I can make you a mod if you want.

There are a lot of architecture astronauts that will keep talking about “we should develop a grand-unifying system”, but most of them are all talk and little result.

When push comes to shove, the developers that actually ship are the ones that look at problem they want to solve and work on that.

Yes, it is possible.

Whether enough people want it is a different story, and most developers are stuck in the app store mentality, so it gets difficult to get enough support for this to move faster.

Wouldn’t you agree that hardcoding one specific community to provide a feature to “try it and see it causes any problems” is yolo’ing?

I’ll code something to spread communities to other instances using an API but that seems like a lot more work.

In my view, the “proper” solution to this (and that would fit right into ActivityPub) would be simply to let the actor that represents the server to post “as:Create” for any new communities that are created and then let the other software follow these if they are interested.

It certainly would be a lot more work and it would still require others to write code on their end to look for this information, but seems like the only implementation that would seem like just another ad-hoc hack.

You think that if I paid you to remove features, you’d do it?

On the other hand, I’m see newcommunities being flooded with posts from piefed.social and thinking the following:

• If this becomes an automatic thing, I’d be less inclined to subscribe to the community.
• How is concentration around piefed.social any different from concentration around LW?

It’s funny to see these contrasting approaches between PieFed and Lemmy development.

Rimu just treats this is a hobby and goes ahead yolo’ing a bunch of these features that abuse the underlying protocol and could bring serious systemic risks if other admins start deploying it, but because the current userbase is small then there is little scrutiny and they all think it’s good to go.

The Lemmy devs are trying for years to get enough support to make a living of their work and therefore a lot more “professional” about what they do, so they would never introduce a feature that could cause Lemmy to be a “bad participant” in the network. But by not taking a more “customer-focused” approach to product development, it takes a long time to bring any functionality that makes it attractive.

PieFed is definitely taking a “Worse is Better” approach and I don’t know what to make of it. It seems to be poised to make it most popular software among the current fediverse users, but at the same time it makes so idiosyncratic decisions that it makes it hard to believe it will be usable if more people started joining in.

https://chat.piefed.social/#narrow/channel/4-developers/topic/Lemmyverse.20Community.20Discovery/near/4966

Wait, so the posts that are showing up on !newcommunities@lemmy.world are not really initiated by the user, but just the piefed server creating a post and impersonating the user that created the community?

@rimu@piefed.social, if I’m understanding this correctly, it’s the second feature (the first one being the import of communities) where you have the server initiating actions but misrepresenting the true actor.