[filmgirlcw]

Mastodon is successful for sure, but it isn't successful enough to overcome the incumbent in the space, which is what we're talking about. That isn't a bad thing, but it's important to add that context when looking at anything else that is being heralded as "the next X."

That's literally why I added the caveat of saying "Mastodon isn't a failure" -- but it hasn't overcome the incumbent and it won't and that may not have been the goal, but that is certainly the story that was told/sold "this OSS de-centralized app is going to replace Twitter," if not by the founders, than by the press/people excited by the idea.

[apatters]

To think in terms of there even being a next "incumbent in the space" implies a misunderstanding of what is driving Mastodon adoption.

The differentiator behind Mastodon is federation via the ActivityPub protocol. And Mastodon is just one implementation of ActivityPub. More are being built every day.

It's likely that a world with ubiquitous ActivityPub would look very different from the social media landscape of today -- Mastodon is a little like Twitter, Pixelfed is a little like Instagram, Peertube is a little like Youtube, but they all syndicate their content to each other. So as ActivityPub grows the conventional categories blur. What happens if WordPress decides to add ActivityPub support to the Core?

I can't read the future but I think the real metric to look at is ActivityPub adoption. No one is tracking it very well right now to my knowledge. And if it hits a certain critical mass I think it will start to reconfigure the current categories of social media. It will also bring down barriers to entry which may discourage the rise of incumbents similar to today's crop.

[thosakwe]

I love the idea of ActivityPub, but in practice, I think the very complex protocol, which is built on top of JSON-LD (I still don't understand why) of all things, and accepts variable structures for many, if not most fields, which makes it cumbersome to (de)serialize in languages with static typing, will hurt adoption by new projects.

My main gripe with it is that potentially writing a spec-compliant server in an enterprise-y language like Java almost feels like a fool's errand (I tried it in Dart, which is very Java-esque).

On that note, I don't even think that the main focus should be on which protocol is used, but rather the features of platforms in the "fediverse," and reasons other than the blanket term "privacy" that the average person would consider switching.

[mehhh]

There are tons of differing implementations like PeerTube, PixelFed (Instagram-esque), Lemmy (Federated Reddit/HN), Write.as (great for blogging on the fediverse and closed platforms at the same time), Pleroma, etc

[thosakwe]

Yet out of those, which actually implement the full spec (server-to-server, not to mention client-to-server, though I do think I remember reading that Pleroma has it), or even if they implement parts of it, do it in a compliant way? I would imagine the number is small because of the sheer complexity of the spec, and the many ways the same data can appear in different requests. I remember reading one of the Pleroma authors saying that they don't support JSON-LD in their implementation because no library for it existed in Elixir.

Like I said, I'm a fan of ActivityPub and its concept, but I think it's fair to say that spec is more complex than what it needs to be, and would have been better off strictly defining the shape of data it can handle.

[iamnothere]

> Mastodon is successful for sure, but it isn't successful enough to overcome the incumbent in the space, which is what we're talking about. That isn't a bad thing, but it's important to add that context when looking at anything else that is being heralded as "the next X."

Are we talking about that? I don't think so... why must everyone "overcome" the incumbent?

Sure, Facebook is the goto for "general purpose" social networking between friends and family. On the other hand, Facebook is terrible for following and discussing the news, unless you like a huge heap of bias and low-quality clickbait. WT Social is planning to address that niche. Facebook is also terrible if you want to keep up with a huge volume of tech news and follow interesting discussions on those topics with lots of expert contributors. HN fills that niche, but sometimes there is too much politics here, so Lobste.rs exists for people who prefer smaller, more focused conversation.

The "long tail" in the social space is huge. Not every venture needs or wants to be the next Amazon in its space.

> if not by the founders, than by the press/people excited by the idea.

And that's the problem. The press doesn't know what they are talking about.