The Gay Pedophile at the Gates

1. Account 1 month old or younger? Can't post media.
2. Don't have X messages? Can't post media.
3. First message (or several) is premoderated.
4. Give jannies the ability to grant media posting privilege at will so that new users who are obviously genuine and provide value get approved faster.
5. Support gay genocide.
 
@Null, have you given any thought about using perceptual hashing? Idea being when you ban a user for posting an inappropriate image, you (i.e. the forum software) can first grab its phash and store it in a blacklist (then delete the image, that way you're not accumulating CSAM, just hashes), and then whenever someone uploads an attachment, grab its phash and quickly scan against the blacklist. Matches (rather, "very close proximity") would auto-trigger account suspensions without notice (perceptual hashing is surprisingly accurate -- it's what the feds use for their semi-publicly accessible "CSAM matching database"). False positives are extraordinarily rare and presumably a quick email exchange could clear them up.

There's at least one PHP implementation of perceptual hashing, but there's a shit ton of other options too. This one produces 8-byte hashes, meaning a table lookup on that for a match would be fucking fast even with a hundreds of thousands of rows. The short length of the hash is a feature, not a limitation. You don't want every image to have a unique signature; the purpose here is to find visually identical images while allowing for changes in resolution, color balance, watermarks (or absence thereof), compression artifacts, etc.

Obviously integrating this would require a new DB table (though I'd argue just ID and hash would be enough; it doesn't particularly matter who posts matches for whatever hash, and that could be noted in the ban message anyway), a change to integrate the phash function for all attachments (no need to save them on success BTW; just the blacklist is fine) to catch and block/auto-ban matches, and a change to the ban feature to enable it to phash a newly-discovered "naughty" attachment to add to the blacklist.

It's not a perfect solution, but it would require additional effort for our little pervert pedo to weasel past it. He'd have to constantly post different images (keeping track of which ones have caught bans already) and potentially even start editing existing ones to get past the phash match. He probably does have thousands of different images (pedos always do) but the upshot here is nobody else could ever re-use any of them if they ever decide to play copycat.
 
it seems the gay furry pedo has finally did his jail time and is trying another gayop with his discord trannies.
 
big dumb said:
1. Account 1 month old or younger? Can't post media.
2. Don't have X messages? Can't post media.
3. First message (or several) is premoderated.
4. Give jannies the ability to grant media posting privilege at will so that new users who are obviously genuine and provide value get approved faster.
5. Support gay genocide.
Click to expand...
Does the forum even have enough jannies for something like this? I know single lifeless redditors moderate places 10x the size of the Farms on their own, but do the farms have people with the dedication to do that kind of moderation? Also, this smells like it would just cause a slow burn shitpile where the commiepedotroon just ends up flooding support tickets instead.
 
(I have no idea as to the cost or feasibility of this)

What about setting up forced 2FA Sign Up/Login with KiwiAuth (just use an open source otp authenticator, preferably not only windows based.) If you could set registration and the subsequent login to require a new key generation you could also control the flow of signups.

That should help with stopping him and add an extra detection layer for you to filter malicious registers, I.E. 1 KiwiAuth client doesn't support multiple account additions.

In the case of lost keys, users are to make a new account and talk to staff in a private forum. (IP, password, payment receipt, some other form of verification to recover account.)

As long as you're not logged out everytime you visit the site I don't think it'd be that much of a hassle if it helped stop all the bullshit.
 
God damn. Reporting a pest but nothing happens while they focus on people posting harmless memes. Too bad we can't do anything about this. I mean this wood chipper material doesn't seem to be afraid of being doxxed so what else can be done unless someone does something really bad and nobody of us is stupid to do that.
 
I know the National Centre for Missing and Exploited children maintain hash lists for images and I presume videos for CSAM. I don't know how open it is but they have an API for interacting with it and frankly, I'm surprised I couldn't find an easy PHP implementation of it. I don't believe it's meant for real-time querying but perhaps hashes could be kept up to date on the Farm's end. I could do a little research into how people interact with these lists to protect their sites if that's helpful. One of the first results for this stuff of course was Cloudflare. Yes, of course - due to a never established "imminent threat to life" (how imminent can it be when it's well over a year ago?) Cloudflare now lessen your ability to fight child porn. *mad*

Anyway, NCMEC in case useful is here:

I think law enforcement really needs to deal with this guy. If they're ignoring your reports can we get more people reporting him? Some kind of petition demanding that law enforcement actually act on our reasonable suspicion? We can't do anything about him but maybe we can get the authorities to do so with enough of us raising a stink. The Farms isn't as obscure as it once was. It turns up in mainstream conversations now. Mumsnet of all places as a thread on us!
 
Ferryman said:
Does the forum even have enough jannies for something like this? I know single lifeless redditors moderate places 10x the size of the Farms on their own, but do the farms have people with the dedication to do that kind of moderation? Also, this smells like it would just cause a slow burn shitpile where the commiepedotroon just ends up flooding support tickets instead.
Click to expand...
I'd argue you do not give this task to the current Jannies who are already working too hard. Create a sub-set of trusted users who are not Mods/Jannies and have the sole task of the moderation of say 10 new accounts. It would require a new menu option where the "approve user posts with media" is a separate item they can click into and approve/disapprove. They have no authority beyond this minor task - sort of like giving a "tick" (sort of like how we give people a badge for their comments) to the user post which acts as an authorization/denial. Or create a separate Jannies User that only do this function in exchange for an unlimited membership on the forums. To keep the traffic off Null, if these users given the task have problems they can go to a Janny for a solution before going to Null, so the traffic will filter down to acceptable levels to the top for genuine exceptions/complications.

It's probable (as I am not tech savvy) there are technological solutions that will avail themselves at a later date, so the above mechanism is merely a coping mechanism until something better comes along.

This asshole may be the current monkey on Nulls back, but is also won't be the last. Surely there are 30 users on this site that could each handle say 10 new accounts to monitor each week until they pass probation, or a dedicated "Probation Janny" that could do more.

If nulls entry technology can knock out a good chunk of it, then the New Users could knock out a good chunk of it, then only a mere fraction makes its way to genuine Janny's desk and finally Nulls to filter the shit down to acceptable levels of cope.

Perhaps this coupled with a separate "report" flag only for media content so it gets the swift attention it deserves might also be an added feature to expedite the sensitive nature of these infringements that do indeed require a priority over the usual reported whining.
 
Could reformatting images and videos to remove embedded material work? We had every image being converted to the webm WebP file format at one point, so I don't think implementing this would impossible. For example, convert images to a lossy file format to remove anything hidden inside the original file.
 
Last edited:
moocow said:
@Null, have you given any thought about using perceptual hashing?
Click to expand...
considering he implemented it in the 8chan software he wrote, I'm fairly certain josh knows about perceptual hashing (and likely already using it, since he said he already reports stuff).

eta: as mentioned earlier, the next step is to use CLIP (runs in tens of ms on CPU), and if the uploaded image matches "child" flag/withhold it.
 
Is this guy retarded? Literally all he is doing is bringing more attention to his thread where his furfag crimes are on display.

Null said:
When I am awake and monitoring the forum, it will be Day and registrations will be open. When it is night time and I cannot monitor the Kiwi Farms, I will close registration to invite-only.
Click to expand...
We all appreciate you and the farms but don’t burn yourself out Josh.
Stay safe glorious kiwi emperor. :semperfidelis:
 
I remember seeing a xeet of a website that soyteens were trying to raid, but failed due to not being able to pass the captcha, which shows an image of a random person and requires you to type their name. Perhaps that could work?
1762256270120.png1762256285470.png
Here's two images of the captcha. The people in the images can be famous cows such as Nick Rekieta, Foodie Beauty, Chris Chan, etc.
What do you think?
 
You must log in or register to reply here.
Back
Top Bottom