I run an imageboard on vichan as well.
After a few weeks I started getting automated CP attacks. At first I was upset and just deleted it all and contemplated turning off the server for a few days.
But then I got mad and figured out exactly how their attacking and figured out how to block it.
For what its worth I know what tool they attack with even but that doesn't help much. I also reported all the IPs to the appropriate abuse emails but thats a black hole.
I messaged who ever runs the lefty pol twitter but got no reply.
I want to share my defence methods with leftypol in a way that attackers can't see how I did it (i dont wanna publish it on git or something). Message me back on twitter if your interested in automated CP blocking, it supprisingly wasn't that hard.
Is it something that would make sense to incorporate into the design of other imageboards? I did have one CP post over a month despite keeping a low profile. You can message me at $bbsname @ protonmail.com
>>43907Just started work for the day, I'll message you as soon as I'm home.
I wish I could find that TTS rant by one admin who had a look at vichan's code and decided to just throw in the towel
>>43914Def not listening to a 20min ai video
>>43907writing the email now.
>>43906I know that Magalichan has an automatic nsfw detection system that also blocks all spam of that kind, you should ask them.
We mostly use cloudflare nowdays which blocks illegal content, but it requires some legal scaffholding to enable.
>>43916>writing the email now.FYI $bbsname means the name of my site, not zerobbs, if your mail bounced.
I'm at work right now but I'm thinking I will make a private git repo where we can collaborate privately as image board admins
Im home now.
Literally this morning before I left for work the bot managed to sucessfully post CP to my site again. I thought I had it defeated but I don't obviously. Its new attack gave me more clues and commonalities to block and defend against.
But I think truly our best defence is colaboration privatly where we can share our defences and data.
My defence strat is constantly evolving and growing in scope more than I can easily share over a few txt docs now.
What say you leftypol admins?
Should I start a private invite only git repo for this?
>>43923I emailed it as it was written
$bbsname at proton blah blah
>>43926It's leftybbs of course. Why would i name my textboard admin burner mail $bbsname!? You could also message me on github
https://github.com/zerobbs-dev >>43927I dont think you can message on github. message me on twitter I guess?
Again i think our best option is colabing with as many chill img board admins as possible.
>>43928I can easily make github burners, but i don't think that's an option for twitter. I think you can make private github repos, so i would suggest going with that, if you don't know a better option.
>>43929What I'm saying is I can not message you via github.
You can only communicate via comments and code.
I'll add you to the private repo soon.
Also your way of sharing email makes no sense or im too stupid or both.
I followed you on git.
Stay tuned admins.
Private defence against the internet dark arts coming soon.
>>43930replying to myself:
I won't have time or energy until sunday or monday to set it up. I'm going to try and gather as many admins of chill boards in the git repo so we can give the threats that target our smallish community some sort of defence.
>>43931If you didn't get the mail yet, i made a github organization. We can have a private wiki, soft-fork vichan and invite others and shit.
>>43933today is my day off, gonna get to work on it.
Remember please try to reach out to other admins of other imageboards so we can share data.
Current admin of an altchan here, I recommend getting the user's browser's fingerprint and blocking that fingerprint. They are most likely running some Selenium Python bot that opens Firefox in headless mode, so being able to detect whether or not it matches these conditions
>using firefox
>is headless
Might help with curbing the CSAM spam. However, innocent users who just use firefox just because might be affected by this, which is also why I recommend trying to get more info such as
>what OS the bots are using
>what ports the bots are using
I recommend you use the canvas JS API to find information like this and check every time whether or not a user posts it matches the canvas of a banned user. If it matches, it autobans the user for ban evasion. It's been working like a charm to get hostile users of splinters of my altchan off.
Isn't there some kind of AI program that will automatically detect and remove cp?
Unique IPs: 8