An evidence-based and critical analysis of the Fediverse decentralization promises

I don't care if 99% of users are on once instance as long as people have the option to create their own instances and build on tech software and content created by the community.

Tl;dr - fediverse probably won't do too much, and it does have discoverability issues, along with migration issues

Another paper that equates not changing the world with being a complete failure.

A valid viewpoint, I suppose, but some Fedi-things have certainly improved my life, which really, is how these things work: you improve people's lives incrementally, and not by the hundreds of millions at once.

Of course, that means this is a complete failure because we won't accept anything other than massive global success as success anymore because... reasons?

pretty brief paper though ngl, it doesn't do much

They define decentralisation as an even distribution of users? Or did I get that wrong skimming the paper?

This seems arbitrary. Mastodon is a decentralised network, no matter how big Mastodon.social is. Lemmy is equally decentralised, even though there's a dominant actor.

The other hubs in the network don't revolve around mastodon.social/lemmy.world. they connect to each other bilaterally - if the central hubs disappeared over night it wouldn't affect them all that much.

I think the notion that decentralised networks can't have hubs of varying sizes is plain wrong, and a fundamental misunderstanding of what decentralized means.