When I look at https://lemmy.ml/c/startrek vs https://kbin.social/m/startrek I see two entirely different lists of posts. Why? It’s the same topic, just on different instances. How can we have communities about topics without having them siloed into their own instance-based communities? Is this just related to that 0.18 issue with Lemmy/kbin not talking nicely, or is this how the Fediverse is?
Is it (at least theoretically) possible for me to post an article on https://kbin.social/m/startrek and have it automatically show up on https://lemmy.ml/c/startrek, or are they always going to be two separate communities?
I suspect he just straight up logged out and went off into the sunset…
Nope, still trying to understand things. I want to get away from Reddit, but if this isn’t made for users like me, then that’s okay, it’s not made for users like me. I don’t blame kbin, nor want it to change what it is; I’m just trying to understand the Fediverse and all things related to it, that’s all. Trying to find a good home for the future that isn’t Reddit.
Federation means centralized decentralisation. It aims to strike a balance.
Each community still needs to have an admin, a creator, and moderators. In the future, it will likely become possible for a community to pack up and move between instances, but things still have to have a “source”. This is what enables centralised control, moderation, within a decentralised system. The “home” instance is in control of any given community, but it is in fact hosted on ALL servers that it is federated with.
But this isn’t peer to peer, each copy is just a copy, only the “real” community gets to be a “legit original”, this is how it can delete stuff. This is how you can delete your own stuff. Or edit it, for that matter. Every community has a “center” somewhere on the network, and all others are “spokes” to that “hub”. But each server can be both the hubs and spokes for different things, spreading out the load on the system, and providing redundancy. When a server goes offline, the other “spokes” of any communities on it keep working to a limited extent, however, the “spokes” can no longer talk to each other without the “hub” so comments and posts stop syncing.
If any given server goes to shit, yes, there will be loss, but the system as a whole survives. And it wont be long before the communities that were lost set up new “hubs”. Additionally, the old spokes don’t go anywhere, they will still show up in search, be visible in your user history, available in full as an archive. It just wont be an interactive one.
Without this, you just get peer to peer, along with all it’s suitabilities for illegal activity. In a federation, there’s still someone in control, who can purge criminal or simply unwelcome users. Or to just keep things on topic, to prevent a star trek community from being flooded with bots posting star wars.
See, that must be my confusion. I was thinking that each instance moderated the content that was being ingested by the server from whatever instances its users subscribed to. Like, if I subscribed to !startrek@startrek.website, then it would create a community here on kbin where the moderators of that magazine (I’d assume I would be assigned as the moderator if I were the first to subscribe to that source) would then moderate that magazine based on the kbin instance’s rules. Like a: the instance pulls all the content from the external instance, but it’s up to the instance’s users to moderate the ingested content themselves, kind of deal. I’m learning that I have waaaaay too high expectations for the Fediverse given how young it still is.
This would be monstrously inefficient. No, each community is moderated by its top mod, and any additional mods that they appoint.
Worth noting, is that you can mod communities that are on other instances, an account does not need to be on the “home” instance of a community, in order to be a mod on it.
This way, content does not need to be moderated multiple times, for every instance it is on.
Could you imagine subscribing to a community you like, and suddenly being saddled with the responsibility of monitoring everything that comes through just because you accessed it off-instance? Or worse, having to review the entire history of a community because you just added it from a new instance? No, this would never work.
I don’t know about, “this would never work,” as it’s basically how news forums work today. The news is posted by websites such as CNN.com, Reuters, etc., and then individual forums/websites moderate that content based on their rules. It would be less efficient for small instances to subscribe to big topics, but then there could be a solution about layering moderation (an option to ingest the feed as it gets moderated by a different instance, or to ingest the raw feed, or some other option I didn’t think of in the few seconds I gave it thought).
I mean, it could work, but having moderation be centralised, with the option to start up communities on the same subject with different modding policies, just makes more sense efficiency-wise.
What would be the benefit of every server modding everything that comes in compared to that? And they CAN still do that, by appointing instance mods.
The benefit would be the content. Imagine you post to a magazine with 300,000 users, with new posts every 5 minutes, and hundreds of comments per article; that kind of an experience would be more desirable than posting to a magazine with 500 people, with new posts every 12 hours, and maaaybe 10 comments on the more popular posts, wouldn’t it?
The benefit of every server moderating everything that comes in would be that a post that isn’t suitable for one instance could be perfectly fine for another. Imagine the topic of politics: for some people, discussing abortion might be too sensitive, but others might be totally fine with allowing it. We wouldn’t want to stifle conversations about that subject, though, so maybe it gets through to the individual instances to handle it as they see fit. This way a user can continue interacting with a large community that’s interested in politics, instead of fragmenting that community into half a dozen smaller communities; sure, some posts might be hidden by some instances, and those threads would be less active than thread about more agreeable subjects, but that’s still a lot better than every thread being less active, isn’t it?
But this means there needs to be someone monitoring posts, not just every five minutes, but every five minutes, for every server.
This is completely untenable. An off-instance sub might not even have enough subscribers on that other server, to count on both hands. Yet someone has to mod it? For small communities, there might just be one or two subscribers to it per server.
You’re giving examples with massive usercounts, which wouldn’t work, due to that massive usercount. But low usercount examples also don’t work, due to the low usercount.