this post was submitted on 23 Jun 2023
72 points (100.0% liked)
/kbin meta
16 readers
2 users here now
Magazine dedicated to discussions about the kbin itself. Provide feedback, ask questions, suggest improvements, and engage in conversations related to the platform organization, policies, features, and community dynamics. ---- * Roadmap 2023 * m/kbinDevlog * m/kbinDesign
founded 1 year ago
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
Kbin.social need not be anti-corporation, but setting standards that fellow federated instances must abide by / putting into place a "collective treaty of federation" or some such that sets the terms of federating with kbin.social (and other signatory instances) would be exceedingly wise. In theory, I have no problem seeing commercial entities as part of the fediverse. In practice, though, I'd want to see strong protections in place to prevent them from turning the fediverse into "Social Network Inc, but hosted on everyone else's dime.
Mostly, one instance creating a massive load of traffic, making the overall load on the smaller instances heavier as they pull in posts to which others are subscribed. I'm not sure if it would be a huge problem, though?
It certainly could be. Each server is responsible for keeping track of activity on the network. It hasn't been an issue so far since all the "big" servers are relatively equal and the overall network is still small. If Kbin suddenly had to process and record every single Reddit comment/thread, it would be a massive drain on CPU resources and immediately cause data issues as Kbin just doesn't have the data capacity that a giant company is capable of building out. Federating all the things may not be viable in the organic future, but we would get there much quicker if a mega server started dumping messages hard.
The thing about federation is that every server basically copies any content that the users on it want to see. So if a comment/post is made on lemmy.world, lemmy.world sends out an update to every other server on which a user is subscribed to the thread/community/user. So each instance that has a subscribed user ends up having to process the new comment/post. If a Meta community came in with say, a million users, now every instance has to process the comments for all those users (that is, if folks on those instances want to see that Meta content).
It is a bit inefficient, but it's just the way a decentralized network has to function. I could see many people thinking that any time you open a thread, the data comes in from the originating instance, i.e. that your home instance doesn't store the data you are viewing. It is unfortunately, and I think it will be a problem in the future as communities grow.