AMA - Do I See a Future for the Federated Web
In this AMA post, Nigel asks me about the about the federated web and whether I think there's a future for it. It's an interesting question as I'm heavily involved in the federated web by running this blog, as well as Fosstodon and 500.Social. Anyway, Nigel asks:
Do you see a long term future for the federated web (like ActivityPub), and if so in what form?
The TL;DR is that I do see a future in the federated web, but maybe not in its current form. Let me explain...
As I alluded above - to me, the "federated web" (aka fedi for the purposes of this post) isn't just Mastodon and its ilk. To me, the backbone the fedi are personal sites, like this one.
So in that respect, yeah, I don't see the fedi going anywhere soon - personal sites seem to be having somewhat of a resurgence over the last few years, and it's fantastic to see. I'm a strong believer in the POSSE mentality, and I think as more people become disenfranchised with centralised social platforms, like Facebook and Twitter, the fedi will only become more important.
But what about decentralised social media?
Services like Mastodon, PixelFed, and to a lesser degree, BlueSky and Threads, definitely make up a fair proportion of the fedi. But I'm not sure how sustainable they are long-term.
I think that the ActivityPub (AP) protocol is fundamentally flawed and as a result of this, scale will be difficult. You see, AP basically caches everything it sees on remote servers locally, so the amount of resources you need grows exponentially as the network grows.
This isn't even tied to number of users on any given server. For example, I could have a single user instance, but with my ~26,000 followers on Mastodon, I'd need a very powerful Mastodon instance, as mine would have to connect to a large proportion of the AP network in order to federate. All of this traffic can easily DDoS a site.
Scale that up to tens, or even hundreds, of millions of users, and I think we will have a resource problem. One solution to that would obviously be more decentralisation, which would be great, but as I mentioned before, if you start getting users with millions of followers, you won't be able to decentralise that and you'd need a mahoosive server to cope with it.
Anyway, I digress...
BlueSky is yet to federate, although they're getting close apparently; then again, their federation protocol, AT, isn't compatible with ActivityPub. Threads on the other hand has integrated with ActivityPub, but I think that's a gimmick, and I'm not sure it will last. Or if it does, the OG Facebook scumbags that are behind Threads will find a way to monetise it.
Bad time. Anyway...
Back to the question
So Nigel, I think the fedi definitely has a future, I'm just not sure it will be in the hands of social media. I think the long-term future of the fedi is actually in our hands - it's personal blogs that are syndicated around the web using services like RSS, email and ActivityPub as a means of distribution.
By using our own sites, there's no Facebook fuckery possible. We, the people, remain in control of the fedi, and that's what it should all be about. We're talking about social networks here and social = people. Not money making organisations.
What do you people think? Will the fedi remain in its current state, or will personal sites become more important? I hope it's the latter.
Want more?
So you've read this post and you're still not satisfied? Ok then, here's some other stuff for you to do: