Heya, with recent news of beehaw.org defederating from a few instances, I noticed that we also recently defederated from sh.itjust.works. I’m not in-tune with whether they deserve it or not, but I have noticed that it does have some impacts on our users.
I happened to see this post from a fellow furry, expressing frustration with picking the ‘wrong’ server. They can also no longer see pawb.social posts/communities.
I also recently posted my little heart script over there because they had a general scripts community, and I’ve only just noticed that the edits/updates I’ve been doing on that post are not actually going anywhere - it’s similar to being shadowbanned. The pawb.social version of that post gets updated as normal, but we never sync that version to their server (and subsequently no other server ever gets the updated version). This makes sense now that I know we’re defederated, but nowhere in the UI does it indicate that I’m just shouting into the void. As a side effect, I’m no longer able to keep tabs on that scripting community for tool updates.
- Pawb version: https://pawb.social/post/43513
- sh.itjust.works version: https://sh.itjust.works/post/89693
I suspect that this is a big problem right now because people are migrating and joining servers at random, and they don’t know that the server they’re joining has bad admins. Communities are rapidly getting created, growing, then getting shadowbanned by half the lemmyverse.
I’m not petitioning for anything to change at pawb.social at the moment, and I’m sure that there were good reasons to defederate sh.itjust.works, but it does make me a bit wary that eventually I too might feel like I picked the ‘wrong’ server if a defederation culture becomes common in the lemmyverse. It’s not something I thought I had to think about when making an account. I really doubt anyone reputable will ever defederate us, so it’s really just a matter of who we choose to defederate.
Thoughts?
You must log in or register to comment.
I must say, I do have to agree that care must be taken with defederating. While I certainly understand doing so from instances which are highly toxic, or those that are actively engaging in spamming. Banning very large instances, like SIJW, could start to cause a lot of issues as more communities move over from Reddit. Here’s a list I found on Reddit of some communities that have established themselves (I don’t know of it’s in an official manner or not, but not sure it 100% matters in this case) on Lemmy apparently. Many are on lemmy.ml, but others are on different instances too, such as lemmy.world. That’s definitely going to be a problem soon if we defederated from omw of those.
I support defederating from actively troublesome instances, but only in a reactive way. Proactively defederating could lead to more issues than it solves. Unless we actively had issues from SIJW, I don’t see a reason to defederate from them? (It would help of we could just block certain communities from instances, instead of having to defederate entirely)
Indeed. We’re going to be needing a lot of fancier moderation tools for the fediverse in the near future, things that the Twitter-like microblogging probably didn’t have to worry about.
Defederating from an entire instance is probably still going to be warranted in some situations, when a whole instance is basically rotten in some way. But I could see plenty of fine-grained tools that could be applied, such as:
- blacklist/whitelist specific communities, or communities that meet certain criteria. For example, you could only allow content from communities with a certain number of subscribers or moderators to filter out “noise.”
- Same with users, you could perhaps block content from users that are under X number of days old.
- “Reputation score” is probably pretty gameable, but you could perhaps track how many downvotes your own instance members give and auto-block users or communities that get too many.
- Giving individual users access to these tools in their settings would let them set their own “view” of the fediverse to be more to their liking.
This is early days yet, it’s going to be very interesting seeing all the various approaches being tried out.
Yeah, there definitely needs to be finer controls over moderation tools. Especially a per-community blocking. That would solve many issues from the larger instances, like Lemmy.ml, without cutting off the whole thing. The issue with gatekeeping new users is well… Look at me. I’m not even a week old. I’d prefer not to be locked out from elsewhere just because “I’m too young/on probation.” No one is gonna like that very much. Reputation score is an interesting idea, but it’s just too easy to abuse unfortunately. I like the voting mechanics to help show what is and is not popular, but I do start to grow wary when it comes to actually hiding or filtering content. That’s another one of those things that’s just too easy to game or abuse, at the expense of others seeing the content. It breeds echo chambers too easily.
I do support giving individual users the ability to choose what to see and not see from the wider fediverse. That actually is something I’d consider the best solution overall, and perhaps something that should already exist. But yes, these are absolutely the early days indeed, and things are moving rather rapidly due to Reddit’s rather sudden downfall.
Blocking based on account age would mainly be to filter out bots and spammers that create new accounts, spew out some junk, and then get banned. There have been plenty of subreddits with restrictions like these. I wouldn’t imagine the threshold to need to be longer than a few days to catch most of those.
The fact that the fediverse is hundreds or thousands of individual servers with their own admins means that there’s plenty of room to try out every possible approach, as suits every admins’ fancy. The ones that hit on good combinations will flourish and those that don’t will bleed users. Especially once there are tools for migrating user accounts from server to server more easily. I think providing as many options as possible is the best approach.
