[moksha256]

I've deployed a number of web services and I have to say...Lemmy was incredibly streamlined and easy to deploy. Deployment is entirely automated with docker and an ansible script, and updating requires just 1 command.

Really looking forward to them completing the ActivityPub integration.

[anchpop]

Does lemmy still do the thing where they censor certain words and don't include a way to configure the censorship to try and discourage certain groups from using it?

[calibas]

Looks like it:

https://github.com/LemmyNet/lemmy/blob/master/server/lemmy_u...

That's the most offensive Regex I've ever seen.

[emayljames]

I would seriously question why you would want to attract anybody that used any of the regex statements words, but is as easy as changing it in the code.

[anchpop]

Really? You can't go 10 minutes in many LGBT spaces without hearing someone using the f-slur in an affectionate way. Also banned is "b!tch", which is not really considered that offensive in most of the US between female friends, as well as "pu!sy", which is maybe vulgar but not an offensive term at all (I notice "dick" didn't make cut). Maybe for a site like Hacker News those terms aren't usually appropriate, but I see Lemmy \more as a replacement for Reddit and I'd hoped it would be inclusive to everyone, not just to those in their bubble.

I don't know how their system works, if using these words just means the comment won't be shared or if it flags it for a moderator's attention, but I don't think so. Here's a direct quote from a developer:

> I'll have to think about this. Hard-coding it means I don't have to do a database migration every time someone comes up with a new slur. And putting it in a DB table means someone could very easily remove it by deleting every row of that table, which isn't good. I want to make it very difficult for racist trolls to use the most updated version of Lemmy.

https://github.com/LemmyNet/lemmy/issues/622

In my view this goes against the core point of the fediverse. The whole reason I like the fediverse is because it democratizes control by giving it to the users instead of a small-group of potentially self-interested owners.

(After googling I found out "n!ps" can be used as a racial slur but I associate it with a funny slang for "nipples". I wondered how the AI conference dealt with this but apparently they changed their name a couple years back to avoid connotations with the slur so maybe it's more widespread/well known than I thought.)

(edit: censored slurs. asterisk didn't work because it triggered the markdown so I used !)

[moksha256]

I wasn't aware of this element of Lemmy...thanks for bringing it up. Philosophically speaking, I detest the idea of software limiting a user's speech in such a way...maybe such a filter would be needed, or maybe not, but I would prefer to see the need demonstrated over time through practice instead of bluntly imposed by software by default.

Reminds me of how Mastodon doesn't allow quote-toots...but quote-tweets can be highly effective when used correctly. Should be up to instance owners to regulate such behavior.

[taneq]

Next challenge: Write a regexp to check whether the post was written by an African-American or not, so you know whether to censor the n-word.

[paulmd]

Not possible! Regular expressions aren't context-sensitive.

[BlueTemplar]

You should just detect bit color?