>We are in some blocklists flagged as racists and have you ever seen me being racist? Most of the reasons are made up anyway. Probably the reason why your instance is on some blocklists is that you federate with instances like SPC/SPW, Asbestos cafe,... Racism is just a blanket word for a "valid" reason.
@voltrina@anna >"uses a blockbot" immediately defederate this instance
Phantasm (phnt@fluffytail.org)'s status on Friday, 05-Apr-2024 06:41:16 JST
Phantasm@anna >i can't recall a single instance of someone talking shit about Ro and actually showing a concrete example of what he did wrong. Here's one: Majority of The Bad Space's rating has absolutely no reasons given. X amount of instances block this instance is not a valid reason. And if it actually has a reason, they are either wrong, heavily opinionated or outright made up. Usually a combination of these three.
@mint@Inginsub You are right, it would need more than just a media proxy patch and a cron job. Didn't think about that. I have some ideas, but most of them aren't ideal.
I'll add that to the never ending pile of pleromer experiments I have in mind.
@mint@Inginsub This may be a dumb idea, but you could possibly modify the nginx invalidation shell script to move the file to a specific long-term directory and patch the media proxy to also look there. Then if you do some clever atime/ctime stuff you can delete old cached files depending on their age or size with a cron job.
Of course you can do all of that without patching the media proxy and use the default folder, but if you ever wanted to disable that, there could be stale files left in the cache.
EDIT: There's not that much you can do with large file uploads. You could move the cache to a S3 bucket and that would solve the bandwidth limitations, but it will also be really expensive to manage and run. Or you can simply ignore them.
@mint@Moon@sun@sunman@captainepoch Changing the AP ID will also prevent interaction with that account. Backend doesn't recognize it as an account properly in some cases and doesn't mention the account in any posts.
There is an MR in the backend that might be able to fix mangled nicknames like this, but I did not check if it really fixes this or it's for another issue.
@mint@pernia@sneeden I'm inclined this a major fuck up when modifying Pleroma source (maybe deliberate) or the server has full queues (I don't really believe that because the site loads fine and post are sent for other instances). I tried manually fetching objects and users from iex, fetching posts via the search, ostatus_subscribe (this one just returns an empty page and no errors in logs). I give up and don't care.
And since his repo in Pleroma git is dead and the instance version points to non-existent commits in Pleroma's main repo, I can't check anything.
@mint@pernia@munir@sneeden@threat That was probably the local status restart. SPW returned 502 for a few manual object fetches before that.
threat and SPCmovienight indeed migrated before any restarts that I know off (some of the first request made to SPW from my instance). Interestingly some account fetch attempts were made before that but all failed.
@Moon@pernia@sneeden@mint Posts are now federating after restart. Some account migrations failed for whatever reason, but I can deal with that.
These accounts didn't migrate here but migrated elsewhere (appeared as new users without notification): dielan coolboymew kirby mangeurdenuage
Also noticed that the relay does not follow back. Does the same for Club Cyberia, so this is something on my side since there are already instances following the relay.
@kirby @creamqueen got doxxed by the license plate probably. That's the easiest path they could have taken. But they now probably have eyes on Cyberia as a whole.
@coolboymew@XRatedPoetry Some sort of zoom is in the Effects in the Desktop behaviour or whatever it's called in english. Zoom the whole desktop though.