True, but the same issue happens with reddit as well, for example gaming vs games vs truegaming. Over time those communities either found their niche (gaming sub became mostly memes, games sub became news and discussions and truegaming tried to become a more serious discussing sub). Actually there were way more gaming subs but unless they found their niche they died out. So people gravitating towards specific communities is a natural occurrence.
As for trying to automatically consolidate communities across instances, it sounds like a great idea on paper but seems like technical she moderation headache, because you won’t have a clear source of truth. Let’s say instance A and instance B both have a community called news. The same news article with the same title is posted on both communities on both instances by different users. Assuming we want to consolidate those posts into one, which instance post will be shown or in more technical terms, which instance becomes the source of truth for that post? Who makes that decision? What if there’s also instance C with the same community and the same post but that instance isn’t federated with instance A, how do we consolidate posts? Each community has its own moderators and moderation rules, who is allowed to moderate the post? What if the moderation rules contradict between instances and both instances want to apply the rules independently, are they supposed to split the post?
Maybe there is an elegant solution to all the problems but I don’t see there being one. I’m not against the idea, the problem is you want to solve its something I have given some thought and because of that I just don’t see it working out the way you’re imagining it.
It would be a fundamental change, but communities should be global and not tied to instances. This would allow for the necessary centralization and reduce duplication. It could also be used to ensure communities survive a in instance going down scenario.
…
This would allow for the necessary centralization
The fediverse is a collection of community-owned, ad-free, decentralised, and privacy-centric social networks.
How would that even work? Imagine you spin up a brand new instance and create a new user and want to subscribe to a community. Because there is no one source of truth does the new instance simply not have the posts and comments that were made before the instance was created? If it’s supposed to get historic data as well from where is it getting from? Does it pick a random instance and pull all the posts and comments from that instance?
What if that instance is defederated from another instance with the same community and doesn’t contain the posts and comments from the defederated instance? Does your new instance have to go ask all the posts and comments from all the other instances to rebuild the community dataset on your instance? What if these two instances that are defederated both create the same post with the exact same content? Is that one or two posts?
What if user on one instance changes the name of the post but there’s some weird bug that allows only half the instances to register that change. Did that change actually happen or not? How do you solve the data inconsistencies if there’s no central source of truth?
What about moderation? There’s no central authority to define moderators or moderation policies. How do you verify who is actually a moderator and not someone trying to impersonate a moderator? What if different instances have different moderation policies, how would communities agree on a moderation policy if in essence both instances can claim authority over the community?
And these are still pretty high level questions. It would get more complex if we were to dig deeper into a possible solution. Even if it’s all technically solvable I think the solution would probably be so complex that it becomes unmaintainable which means it becomes unusable.
Setting up a new instance wouldn’t be significantly different than today. The difference would be instead of asking each instance individually for what communities they have you would use a distributed ledger to contain a list of communities with their primary and secondary instances. This would create the sigle source of truth for communities. As communities still have to physically exist somewhere, the designated primary instance would have the master record for the community and you could designate secondary instances for resilience and possibly spread out pulling that information.
Moderation doesn’t change significantly, primary instance admins would still be the fallback, but they could designate any user to be a moderator.
Defederation would be a little messy, but not a ton more than it is now. The primary would be the source of truth, if they don’t accept writes from an instance, then those posts and comments wouldn’t exist, (this is basically the same as one way federation now). If an instance wants to read from a community it’s on that instance to drop anything from instances they don’t federate with from the response from the primary.
As above, the primary instance is the source of truth, if a change doesn’t get there. There could be an eventually consistent cache on other instances for usability.
The difficult part would be how to handle changing the primary instance, or designating the primary for a newly created instance.
I would recommend a button thar switches between local and all for a community. So people can look at it with their particular rule set in play if they wish.
