More reasons as to why I left the Fediverse

I wrote in a prior piece that I was "over" the Fediverse — I never spent any time attempting to build anything with it due to the immediate cost of being an operator and the amount of risk that entailed. I wanted to build something similar to Bookwyrm but for those who know anything about running ActivityPub servers; the canonical/reference implementation tends to be the only (if not one of the few) running instances and such a task entails taking on a cost that I was not comfortable with (fiscally, emotionally and operationally).

Too many accounts

If I wanted to have a real social media experience, one comparable to what exists in the mainstream landscape, I would have to make multiple accounts. Now, this already exists for things like Tumblr, Instagram, Flickr and what have you but this is done because of lock-in and not for any real reason outside of that. You can use single sign on from yet another locked in vendor but that puts me back to the place of having to manage yet another account. My Bitwarden password manager reports 931 accounts (God knows how many of those are active) and having to add 3 more just to review books, scrobble music and participate on one forum felt counter-intuitive.

I didn't want nor care for it. My Bookwyrm account is effectively defunct to me as I don't update it anymore and the social graph it holds is somehow meaningful but trapped to that specific platform; making it no different from Goodreads or Storygraph. I didn't want another silo to throw information into — I want something that demonstrates actual data portability if not malleability.

Identity and data

There's value in sharding identities such that a work account and personal account can be held. However, I would have to have multiple identities per service. I would have to have a Bookwyrm account for reading books on my career and another for the smut fiction I'm into (not really, lol).

This is what appealed to me the most about ATProto and also the IndieWeb — identity and data storage were separated at the protocol level. In the case of ATProto, your personal data server holds and treats your identity as a fungible resource. In the IndieWeb, identity grants you permission to modify the resources on a site. With ActivityPub and in the most popular implementations of it, your identity is heavily tied to the kind of information you can write — which mirrors legacy social media systems to a deficit and doesn't prove to be really innovative in any right.

Development friction

From my experience, building something for the Fediverse required:

• an always connected-to-the-public-Internet service (if you wanted it to be relatively available)
• isolation in the kind of content you create; apps are aggressively focused on singular content types (mainly due to the difficulty of making parallel media types)
• extreme difficulty testing federation

Seeing each of these, by itself, kept me from building in the space.

Culture

Despite the goal of a community for federated social networking, it seemed like the ecosystem built around the Fediverse was inherently European. This might be due to the large push against "American Big Tech"^[1]. That made it hard for me to find people, outside of what Ro built with PlayVicious, that I would want to virtually hang out with.

It doesn't help that PlayVicious had a hard time existing in the Fediverse and little effort by peers outside of words of commiseration — leaving a sour taste in my mouth personally.

With all that said, I do hope the changes come to ActivityPub, in its next version, that make it more viable as a protocol for social networking and communication.