Do they get adopted by other instances? Are they still accessible from other instances? Can you still post on them from another instance?
Edit: From my understanding every instance that deals with a community has a cached copy. Will that copy disappear after a certain time, because it can’t phone to home anymore?
As of Lemmy 0.18.1, cached copies on other instances do not disappear if the original instance has died.
In theory, it might even be possible to actually clone a cached copy into a new local community. This would require some database hacking, so not recommended unless you’re familiar with Lemmy code and SQL.
I actually wondered about that. The way instances share data and store it is all pretty black box to me.
It’s actually extremely easy to flip a remote community, into being a “local” one.
If I understand the fediverse correctly, if the instance dies (meaning all the background servers are taken away), everything on that instance is gone; accounts, communities, posts, comments. Since the instance is the host, they are the ones holding the data; if they decide to stop, it’s gone.
However, the creator of the community and its members could create new accounts on another instance to rebuild the community.
When that’s the case, wouldn’t it be nice to be able to migrate communities, so when an adminndecides to quit an instance, those communities that want to move can arrange that move.
I agree that would be nice. Hopefully it is something that will eventually get added as a possibility, but I don’t know enough about the background workings to say if it is being considered or not.
I can’t imagine that archiving and instance migration aren’t on the hub roadmap, but they’re probably well beyond the horizon.
Nascent tech rarely prioritizes development for contingencies like it’s own decay/demise.
Bad mojo. 😆
Afaik posts and comments are copied locally to federated instances so if the original went down you could still access what was once there from some other instance.
E.g everything (excluding uploaded images/videos) on https://sopuli.xyz/c/technology@beehaw.org would still be there without beehaw.Once a user on instance A subscribes to a community on instance B, then instance A starts caching posts from the community on B. But to my understanding it doesn’t retroactively fetch all historical posts and comments.
I think it needs the ability for multiple hosts to opt in to co-hosting a community. It could work as a cache and provide some redundancy if something goes down, temporarily or permanently.
The accounts are gone-
But, communities, posts, comments are all replicated. With a single flip of a bit, you can take that cached copy, and turn it into a local community.
Users would need to re-subscribe/follow/etc, but, all of the data would be there.
I’m hoping for account linking so you can have multiple synced accounts between instances so if one goes down due to load or permanently you can seamlessly continue without any issue.
Lemmy is instanced. if the host “dies” everthing else dies with it. even the accounts and everything. as far as i understand it. maybe there will be a few cached posts from that instance but im not sure about that.
This is going to make it impossible for any technical help communities to take root. The fact that the whole thing can just go poof completely turns me off from using something like that.
Exactly, this is why things like stack overflow never came into existence.
I get the concern, but long term persistence is probably a rarity. The internet is still young. If anything a federated group of communities that are linked somehow will last far longer than a single server of even a large corporation. For the weeks that Lemmy et al have been growing, how to best develop communities that connect and last has been an ongoing question.
The internet is still young.
In what world is 40 years “young”?
The internet as most people know it and as companies depend on it isn’t that old.
The difference being discussed here is a single existence vs. potential for redundancy. The best way for something to outlive even the places it’s stored is by repetition. That goes against both how we’ve grown things so far on the internet as well as the talk about competition among instances and the biggest one wins. It’s far better for there to be many groups that share information in some way but are their own entities and aren’t dependent on the rest.
Everything dies.
The existence itself doesn’t.
Are you sure about that?
Yes, 51%
From memory posts are partially archived by Google’s cache, so if they’re indexed correctly people should still be able to search for something and have it as a result? Unsure if that would work if the whole domain is actually gone though
I haven’t bothered playing with instances yet I just made my account yesterday. I need another account if I leave lemmy.world?
If lemmy.world dies, you’re account goes down along with it. You would then need to find a new instance and make a new account.
This is a great question that I also would like to understand better.
The modern-day equivalent of “what happens to a dream deferred?”
Just started learning about the fediverse but I suspect everything goes down with the ship. Sort of creates a development opportunity for community driven backup utilities, though. Interesting problems are fun development challenges :)
I’ve been wondering about this also since I started a new community in the last week and have already invested a fair amount of time into it. I’m hesitant to keep investing time and effort, though, if it can just disappear with no recourse.
I wonder if the activitypub protocol (or, if not the protocol then some other layer) allows for the idea of “community mirrors”. The way that the protocol works at the moment, as I understand it, only the host instance has a complete record of a community’s posts and comments. But if there was a way for a community to designate one or more other instances as “mirrors” which maintain a complete sync of a community’s content (going back all the way to the community’s founding), that would lower the exposure to instances going down.
There would need to be a process (both technical and administrative) for a mirror to be designated as the new host instance should the original host disappear.
This would build in additional resilience into the fediverse model, by taking advantage of its distributed nature.
