[BudaDude]

Meh, I disagree with point 1. Any platform that is not federated or at the minimum open source will go the same way as Twitter, Reddit, or Myspace. The idea of reddit is bigger than one site, it's a replacement for the internet forums. We don't need another platform lock in.

I applaud the dev work, but this isn't going attract the tech crowd.

[dmje]

Hard disagree on point 1. I mean, yeh, you'll maybe win a hardcore niche of nerds with federated services, but the usability / understandability of federation is absolutely awful, per my AskHN recently [0]. That's not to say it couldn't be improved with some good UI and maybe some thoughts about how to communicate it, but right now your average Reddit / Insta / Facebook user is gunna be running a mile. It's absolutely baffling.

[0] https://news.ycombinator.com/item?id=36330155

[hakfoo]

The thing is that federation is a non-user facing feature if you do it right. The whole point is that you don't need to care where a specific service is. Email, arguably the most successful federated app, exposes the servers as a disambiguation mechanism (which hakfoo do you want to message?) but that's not really that important when you're consuming content.

I could see a "polished" Fediverse app that abstracted away the server concept-- you'd say "I want to read the $subject Lemmy community / $person Mastodon account" and it would aggregate and deduplicate data on multiple servers automagically for a rich feed. Maybe track the emergence or discontinuation of servers and refine the lists.

Then, if you want to be particularly precise about which providers you're working with, you can go in and micro-manage the automatic selections.

[dmje]

100% on this. Maybe we're just too early - I'm still waiting for this kind of UI so I don't have to think too hard about providers or the weirdness of the whole thing.

[predictabl3]

1. Matrix is growing so fast that I can't take seriously people that dismiss federation.

2. Name me something that has outlived reddit and not gone through enshitification (it's a term, I'm not being crass). Name one single product or platform.

[dmje]

1. Yeh - but what sets Matrix aside is that you're abstracted away from the conceptual side of federation. As I said, this is about UI as much as anything else. Non nerds are used to "going to InstaFaceTubeDdit and seeing the stuff they want to read" - not having to wrangle with the conceptual stuff about instances, weird looking addresses and usernames, etc.

2. I mean, Wikipedia, but it's not the point. Yes, us tech people are maybe concerned about things being shit, lock-in, etc - but literally no-one else cares. So our best approach isn't to make something perfect from the tech point of view, but perfect enough that there is mass migration to it - because, as we all know, it ain't about the tech, it's about the momentum, content and growth. And in order to get these into a serious shape, we need migration. And to get migration we need slick, beautiful, easy to use UIs.

[kybernetikos]

Email is one of the most successful Internet applications of all time.

[BudaDude]

HTTP is too. People forget that HTTP is the original fediverse.

[kybernetikos]

Yeah. I think it's funny that people talk about individual sites like hackernews as 'centralized' whereas I tend to think of the ecosystem of forum like sites as a decentralised mass.

Now it would be nice if all the things associated with my identity followed me around between the various decentralised forums, if reputation were more easily assessed and transparently generated, and if all my discussions across the web were accessible to me from a single dashboard.

Still, it surprises me when people say 'decentralised? it'll never take off' since I think it's a big part of why the internet was successful in the first place, and we still have the legacy of that in so much of what we do on the internet. People have been using postal services for centuries. With your phone you can call people in different countries, on different networks, using handsets made by different manufacturers.

If anything, I would say that we started decentralised for good reason, the first generation of attempts to enclose the commons failed because they couldn't generate content as fast as the world. The second attempt was very successful, based on a large influx of money, and network effects but we're starting to see those start to fail for predictable reasons. The lifecycle of centralised services are tied to the lifecycle of corporations and the vagaries of shareholder interest. That will never result in services that can stand the test of time in a meaningful way.

You can tell we're in the pendulum swing back now, because companies are fighting against volunteer labour (anyone who tries to make their systems work with other systems) to make things worse for their own customers. That sort of approach is absurd and not long term sustainable.