The Chapo.Chat moderators have already stated that they’re planning on removing the hard-coded slur filter from their instance of Lemmy; it’s the reason you’re not able to say the full text of ACAB. It’s just a bad idea to have something like this hard-coded in, leftist instances are created by fostering a good culture and not by babying the people who run these servers and if we want Lemmy to grow as an actual federated alternative to Reddit then it’s worth bringing up how bad this policy is by the original developers.
As a greater point, what will we do to prevent slurs? Will we even have a slur filter on this instance or just have enough moderation and yelling at wreckers that we don’t have to hear anything like that.
If the filter was less aggressive and matched whole words only it’d probably be fine, though deleting the entire message and responding with a bot explaining why that word causes harm would be much more effective for what we’re doing. We don’t want to just bleep stuff out where it’s super obvious what someone just said, if someone used a slur then that message needs to be unmade entirely and there needs to be immediate education or banning. Deliberately trying to evade the filter to use a slur as a slur should result in being ejected from teh community and should be happening infrequently enough outside of raids that it shouldn’t really be a major concern that someone could use X’s in front of and after a word to evade the filter, normally people should only be actually posting slurs by pure ignorance and anyone posting it otherwise just makes their malice known.
Is that even effective though? We should just be able to trust that people using this platform aren’t foaming at the mouth to say the n word or something and that if someone says a word like tr*p we can just have a struggle session over it like posters do.
Is deleting posts and having a bot post explanations as to why the word is harmful effective? Or are you talking about the current filter?
If the former, absolutely yes. Multiple subs on reddit use that approach and it works really well, people are generally receptive to it and wean themselves off of slurs because they’re posting in good faith and just genuinely didn’t know that word was a problem. It doesn’t require strict filtering, exact matches work jsut fine because anyone trying to evade the filter can just be safely assumed to be posting in bad faith and immediately banned.
If the latter, obviously not. It catches unrelated words like basterd and generally assumes we’re not going to be punishing people for trying to say slurs without triggering the filter. The community, as you said, isn’t foaming at hte mouth to say these words, 99% of hte time when we’re not being raided it’ll be said out of ignorance and all that’s needed to catch those cases is an exact or near-exact match.
If you’re implying it’d be better to have no filter at all, that then puts the onus on the community to intervene every single time someone says a word, struggle about it, deal with cliques that think it’s OK, et cetera. It’s not worth the stress on everyone to get people to stop using slurs, it wasn’t fun on the old subreddit and it wouldn’t be fun now. A filter and bot response can automate that entire struggle session for us and makes it clear what the official stance is on those slurs, and lets the user know that if they continue to use those slurs they’re liable to be banned. Like I’m really, really tired of having to explain to people why the r-slur is bad and I don’t want other ND people to have to defend themselves every time it comes up.
