What’s that about Gitlab? I use it hosted, and it can be self-hosted, so I’m not sure what the issue is.
Honestly, I don’t find a lot of value in the fediverse generally. I guess it’s kinda cool that things connect together, but URLs also get the job done pretty well, and cookies and password managers handle logins and stuff between platforms. The real value is in being an alternative to the big, centralized services.
I’m not here because of the fediverse, I’m here because Reddit pissed me off almost a year ago and this is a suitable-enough replacement. Seeing my posts on other platforms is honestly a little odd (in theory) because having a post from one context (say lemmy) appear in another (say git hosting) is often not wanted. So I want separation between different types of social media (e.g. I avoid Mastodon because I don’t like that form), not more sharing. But that’s not a real concern, it’s just not why I use it.
That said, I have no idea what kind of social media project you’re working on, so maybe federation is the perfect fit. The fediverse is certainly a good way to quickly build an audience and generate content.
Note about my projectFor my social media project, I’m looking at p2p, because I’m more worried about scaling and longevity. Lemmy can get expensive to host due to duplication, so a big instance going down due to funding can fragment communities (though old data will live on any instance federated with that community). My project will use user devices for most data storage and retrieval, and a handful of (hopefully community funded) storage instances for availability and backups. There’s a lot less duplication, so hosting will be a fraction of what Lemmy costs since most data will be served by people near you. The only infra I’ll need is: - relay nodes to connect people - mostly short-lived STUN connections, but TURN will be supported for people behind CGNAT - backup storage nodes - I plan to only store less popular, older content And since I’ll only be handling text at first (pictures and whatnot will just be URLs), it should be really cheap. And that’s it. If I lose interest, someone probably already has their own storage nodes (lots of data hoarders out there) and can easily take over.