People with 8 year old accounts and 100 posts just HAD to come and give their take, they were that inspired.
Events like this are a fantastic way for surface all sorts of undesirable types to help fill in my [ignore] list. I can tell the difference between someone who's just shitposting (enjoyable, if not also irritating sometimes) and someone who genuinely believes the shit they're saying, and ignoring the latter when they say insane things has always improved my Kiwi Farms Experience™.
That's an interesting thought experiment. What if there was a type of ban that didn't prevent the user from posting, but simply banned anyone who tried to talk to them? Like the forum version of shunning. You'd end up with a user desperately trying to get someone, anyone to interact with them, unable to do anything but plead with an empty room for companionship.
There's all sorts of clever shit you can do with the "social network graph data" you can collect and generate on any given social media or internet forum site. Facebook was probably the first to be so open about what they were analyzing (and not so much what they were
doing with those insights), but the tech is there for all of us to use.
I've been toying with the idea of writing an internet forum software suite as an excuse to learn Erlang, Elixir and Phoenix and this kind of thing attracts my interest. Reddit does the same shit internally too, and you can see all the "we did it reddit!" amateurs whenever they spot a post containing wrongthink and go dumpster diving through the poster's post history (previous words, subreddits, linked sites, etc.).
This is one of those things the current generation of hardware powering these "gimmicky AI" toys is actually really good at -- feed it seemingly meaningless data like who used which reaction stickers on which posts at what time and date, and all sorts of interesting insights can start to wash out of that sauce from a machine that's good at noticing patterns. Who a user tends to engage with most (or least), the types of comments that are most likely to get a reply from him, whether time of day or day of week has any bearing on it, who they like (or hate) the most, etc.
Note the data for all of this is available
now, just waiting for analysis by clever assholes of all kinds. Any site admin or ooperator can obtain it immediately from data on-hand just from running the site. You could use it for reverse-shadowbans like the above (you can post, everyone can see your interactions, but no one can interact with you), for sweeping bans like Null did with HHH's "friends" (lol I still think that particular decision is fucking hilarious and awesome and it's why I'll always vote Null for dear leader/feeder for life), and even for pre-emptively locking (or slowing down) threads, banning new users with few posts who post in tell-tale active/popular threads with tell-tale provocative garbage, all sorts of stuff.
You could also use it for scarier purposes, like visualizing and tracking down entire networks of agitprops, glowies, and the like. All by just observing usage and language patterns. Remember all of these sites keep track of what threads and posts you view (and when, and in what order), ratings you give, posts you make (and who you quote and/or reply to), links you post, all that shit. They have to in order for various site functions to work properly.
Honestly that kind of "thinking" and function implementation is the more interesting part in building forum software these days than the nitty-gritty of "rendering posts formatted in VB syntax" and "performing 2FA logins" and "fetching post attachments for page 938 of thread 1234" and the like.
I'd offer to lend Null a hand with his KF rewrite in Rust if I didn't hate that language with the burning passion of a thousand angry suns, but if I come up with anything useful in the "creepy AI analytics" realm I'll send it his way just in case he's interested too.
