Probably a two hour window to give me time to get things done. There may be an upgrade to lemmy 0.19.11 however that requires some modification of ZippyBot due to changes in the new version.
There will be downtime while the server is restarted, however hopefully this is brief.
Thanks
Demigodrick
I see I have access to my account back. I’m assuming the upgrade did not go as smoothly as anticipated?
Absolutely did not :(
But you, and .zip, are back!
I was worried.
We all were. (Thread from !fediverse@lemmy.world)
Glad to have you back!
Rip
It might be worth writing a brief post mortem if you get time. I’m glad its back up now.
Thank you for everything, sorry to hear it didn’t go smoothly 😔
Love your work, thanks for the heads up!
Thanks for the notice!
Hi again, Elevator
Finallyyyyyy, saved!
I got the Lemmy equivalent of a jump scare. All of this has made me realize that Lemmy is actually quite fragile since a big instance can just vanish.
Always have an alt ready
On the other hand, most content “from” any individual server still exists elsewhere in the network. I was following this thread even while lemmy.zip was down just by reading it from another instance.
While the entrypoint to the network can go down, nothing really goes away, you just have to jump to another server. All you really lose is the post history tied to your account–and the posts still exist, just under your old name.
I think you need Lemmy.zip working for comments from different instances to sync under Lemmy.zip community posts.
Yeah, I hadn’t realized that but you’re right. I originally come from the old dead kbin.social instance. I just checked an old community and I can see the posts, but new comments there don’t propagate to other instances.
I wonder if there’s any solution to that. Lemmy admins can move threads between communities, but I don’t know if that works across instances. I wonder if a whole community from a dead instance could be moved to another instance by mass-moving all threads.
I think it is more of a limitation of activitypub and DNS. Any work around for this would be cutting out the original server which would likely be a major security risk.
Honestly what it should do is warn users when a instance is down and maybe even lock the community if it is without a local mod for to long. Maybe it could warn after is it down for a hour and then lock after 6 hours.
Biggest problem: if an instance closes, the content on it dies as well (can’t interact with it…)
The instance could also decide to wipe all the content that’s on it and it would replicate this mass deletion to all federated instances, making us lose all the content that’s on it. In a way it’s stil very censorable and centralized
The content is still on other servers. The problem is that search engines don’t really understand federation so they pickup posts on particular instances which creates dead links.
For the second case, the federated instances would also remove the content
That’s not how that works. It would still be cached on other instances. The only way it would be deleted is though a mass deletion but if it gets bad enough the admins could restore the content and then block the delete request.
That’s what I was saying. If admins manually deleted comments/posts from their instance, it would replicate to others
The admins of the other instances could restore comments and posts from backup if it was deemed necessary.
Also things posted on the internet tend to live on
Hope all goes well :)
Nice
Good luck! And thanks.
Thanks for the notice. :))
We should have a “backup community” so we can discuss if something goes wrong with the main instance
Matrix channel anyone? Or just create a backup community on another instance
I hate matrix just because I can’t find a good client with no tracking or company behind it
I was thinking about a community on another instance, yea
I just use the Web client hosted by an instance I trust: https://chat.tchncs.de/
Still down :(
