Hacker News Reader

Hacker News

Comment by TylerE | Hacker News Reader

Parent Full thread

TylerE

There a tons of truly vile fediverse instances.

View on HN

But fediverse isn't run by one man that personally uses his position to push those opinions.

.. and I as a mastodon server admin can just unilaterally block them at the behest and for the benefit of my users.

Yes which is a feature of the fediverse, not a bug.

If the admin blocks an instance it's either a community you as a user won't want to be interacting with, or the user has chosen the wrong instance for them to be on.

I don't know why people say that every instance not federating with every other is a bad thing. It's not. I definitely don't want to interact with the extreme right instances around so I choose one that I feel at home with and that has them all blocked. I think that's the beauty of the system.

No, it's a bug, not feature. Those content should not flow into your timeline if your friends are all nice and vile accounts are banned on your instance. The fact that instance owners not just can, nor just do, but HAVE TO block a bunch of demonized instances as a checkbox obligation and compromise and virtue signaling is a bug.

How my community wants to operate is not beholden to your standards if my instance wants to block the stormfront mastodon instance it's going to and nothing you say is going to make anyone on my instance change their minds about that.

I have not seen any evidence of any major instance ban another instance because it refuses to ban yet another instance. There would be no reason for that. Messages from the controversial instance still won't make it to the instance that demands the 'virtual signaling' even through my instance.. unless some user on my instance essentailly quotes whole posts.

I think the difference is that moderation is easier on the fediverse (as well as on Bluesky). If you want that vile content you kind of have to either be in that circle your self, or you have to explicitly look for it. On Bluesky there are popular mute and blocklists, hateful accounts get mass blocked, and engagements around hateful content is diminished (not proliferated like on Twitter). The Fediverse is even more so, where whole servers are mass blocked by because hateful content is allowed to fester.

Someone has to do it though. A default federate-everything install will, well, federate everything.

Sort of. Somebody still has to follow the other or find a post from outside their server and comment on it before any communication takes place. How likely that is depends on who's on your server.

If you're hosting people who already have a large audience, or are especially vulnerable to harassment, that's probably going to happen sooner or later and it would be wise to craft or copy a denylist before it does.

I run my own Mastodon server and I've never had to take any moderation actions with a few hundred followers. I know the names of a few vile servers and can't recall seeing them among my followers. I suspect I'm fortunate in just not being interesting enough to vile people.

True, but apart from moderation, the engagement driven algorithm on Twitter also does a ton to proliferate vile content. On the federated platform the algorithms tend to be simply chronological and limit the content to account you explicitly follow. Meaning there is way less exposure to undesired content. One of your follows must be the one to repost it (which may warrant an unfollow or mute, etc.) or you must opt into an algorithm that gives you unwanted content (which you may then simply swap out for a better one).

Reply on news.ycombinator.com