Plagued Soyjak.Party / The Sharty - The altchan born from the ashes of /qa/; also a containment thread

  • 🐕 I am attempting to get the site runnning as fast as possible. If you are experiencing slow page load times, please report it.
Don't they only check the hashes of known CSAM images? As grim as it might be to think about, this group could be out there making new stuff.
Doesn't look like it — one of the thumbnails is from a very old video. Before you ask how I know that, someone used a clip from that video to get one of my old chatbots banned back in 2016ish.
 
Null would never since he hates us. Remember when /raid/ discovered the Bluefolf scat video? It was featured on the front page for a bit but then someone took it down.

Null features funny stuff. This is still a gossip site to him, not a call to action kind of thing, even if it has turned into one due to censorship everywhere else.

Don't they only check the hashes of known CSAM images? As grim as it might be to think about, this group could be out there making new stuff.

CompSci wizardry time: file hashes can be vastly altered by changing a single pixel, and not even perceptibly, because hashes are extremely random by design. The point of a hash is to be vastly wildly different for any given input so you can see if things are even slightly changed, so a single character in a text file or a single bit flipping in binary will throw it off.

Perceptual hashing, which hashes the image in itself, instead of the binary representation of it (the data 1s and 0s a computer sees), returns similar hashes for similar images. This is a bit harder to fake than "I changed the RGB of a single pixel by 1", but is also defeatable. HSV noise, putting random shit (emojis and stickers, even) on the image, rotating enough, left/right flipping, stretching, tiling, etc, defeat this too, but it takes more effort than clicking a single pixel on a smartphone. On the other hand, this does not seem to be as widely used and may incur costs to hosts.

I remember asking Null about this in particular and he said the fees were pretty fucking stupid, for what really should be a service to humanity.

It's easy to circumvent automated file scans and impossible to manually verify everything that gets uploaded to your file hosting site is legal.

This actually is a really big problem. Media dissemination back in the day was curated and there was much lot less of it.

The double edged sword of letting anyone create and disseminate content is that one would need an entire industry of millions to even begin to check on everything, and that's if you banned encrypted files!
 
Last edited:
Doesn't look like it — one of the thumbnails is from a very old video. Before you ask how I know that, someone used a clip from that video to get one of my old chatbots banned back in 2016ish.
No worries, I actually typed that reply up before I saw on the next page that password locking .7z files can obscure the contents of archives and beat detection.
 
Passwords should prevent the contents of archives from being scanned.
The Mediafire links are not password protected.

Access to the Pastelink containing the Mediafire links are Password protected, however the actual video files aren't encrypted.

The way the whole exploit works is

Video 1 contained the Password, in the comments of video 1 will be a link to video 2
Video 2 will have a comment linking to the Password protected Pastelink
They are directed to watch video 1 to gain the password
Once they unlock the Pastebin they then have a Mediafire link
The mediafire link is not password protected, and neither is the media contained in it.
 
It's still posted almost every single day, are you not reading the ban lists?
The ban lists don't tell you what specifically was banned for, or what time a ban happened.

CompSci wizardry time: file hashes can be vastly altered by changing a single pixel, and not even perceptibly, because hashes are extremely random by design. The point of a hash is to be vastly wildly different for any given input so you can see if things are even slightly changed, so a single character in a text file or a single bit flipping in binary will throw it off.

Perceptual hashing, which hashes the image in itself, instead of the binary representation of it (the data 1s and 0s a computer sees), returns similar hashes for similar images. This is a bit harder to fake than "I changed the RGB of a single pixel by 1", but is also defeatable. HSV noise, putting random shit (emojis and stickers, even) on the image, rotating enough, left/right flipping, stretching, tiling, etc, defeat this too, but it takes more effort than clicking a single pixel on a smartphone. On the other hand, this does not seem to be as widely used and may incur costs to hosts.

I remember asking Null about this in particular and he said the fees were pretty fucking stupid, for what really should be a service to humanity.
The top tier of perceptual hashes is an image embedding. Unfortunately image embeddings are very much reversible. The fact that you have to make sure your hash cannot be reversed while making sure it is still a perceptual hash is very difficult and I would appreciate legislation that allows it so perceptual hashes of CSAM are explicitly allowed.
 
  • DRINK!
Reactions: David Thoughie
Null would never since he hates us. Remember when /raid/ discovered the Bluefolf scat video? It was featured on the front page for a bit but then someone took it down.
Just for the record, @0 0 featured it the 2nd time and it was featured the first time when it was discovered. It was taken down but the contents of it were moved to another thread of the reactions of it.
 
The Mediafire links are not password protected.

Access to the Pastelink containing the Mediafire links are Password protected, however the actual video files aren't encrypted.

The way the whole exploit works is

Video 1 contained the Password, in the comments of video 1 will be a link to video 2
Video 2 will have a comment linking to the Password protected Pastelink
They are directed to watch video 1 to gain the password
Once they unlock the Pastebin they then have a Mediafire link
The mediafire link is not password protected, and neither is the media contained in it.

OK, what the fuckety fuck, the video isn't fucking encrypted? OK, that's laziness. That's actually easy as fuck to automate and the average pedofaggotgiganigger doesn't know how to re-encode or pull tricks with keyframes or whatever.

I also just realized some poor fuck has to actually maintain the hash lists for this shit. 💀
 
OK, what the fuckety fuck, the video isn't fucking encrypted? OK, that's laziness. That's actually easy as fuck to automate and the average pedofaggotgiganigger doesn't know how to re-encode or pull tricks with keyframes or whatever.

I also just realized some poor fuck has to actually maintain the hash lists for this shit. 💀

Unencrypted and no password on some found to be uploaded 3 years ago. It's not just that they're uploading fresh material, they're using files that have existed on the site for years.
 
Unencrypted and no password on some found to be uploaded 3 years ago. It's not just that they're uploading fresh material, they're using files that have existed on the site for years.

I'm not going to pretend that there's no cost incurred with running some kind of a cp scanner, but is it really too much to ask? Are there not open source means of getting hashes and running scans during idle time and marking scanned files as you go?
 
  • DRINK!
Reactions: David Thoughie
I'm not going to pretend that there's no cost incurred with running some kind of a cp scanner, but is it really too much to ask? Are there not open source means of getting hashes and running scans during idle time and marking scanned files as you go?

Mediafire is one of the largest hosting platforms on the planet; there is no excuse for a lack of safety system to prevent this.

Youtube STILL has the channel up.

1714765997708.png
 
Mediafire is one of the largest hosting platforms on the planet; there is no excuse for a lack of safety system to prevent this.
I suppose we can ding them for this, but at least they action reports at all. Unlike the other largest hosting platform mentioned later in your post.
 
  • DRINK!
Reactions: David Thoughie
I suppose we can ding them for this, but at least they action reports at all. Unlike the other largest hosting platform mentioned later in your post.

I don't know if "Molly" is a generic names shared in their staff, however the contact I have within Mediafire has been extremely quick to purge and ban the offending accounts. Within 15 minutes so far. I can't really fault them for their response once finding out about it but the fact that this stuff has existed for years is insane.
 
Yeah, it is a very stupid thing to do; to admit to seeking out and viewing CSAM.
When multiple people have admitted to seeing this on this thread alone, I have to wonder what the hell they thought they accomplished other than potential PTSD and disrespecting the sanctity of victims. I have more pity for the guy who found and reported initially. It comes off as a bizarre and undersocialized brag to say you've seen such a thing and I'd rather my cat leave me a dead rat on the pillow than hear that shit. Congrats on breaking federal law of most developed countries out of morbid curiosity, please get therapy. If you didn't need it before, you definitely will now.

It's not that google and/or the feds doesn't know about this, they are probably involved in this. Youtube runs ai bots to sniff out 5 seconds of copyright material, but they can't sniff out CSAM/child porn? Get the fuck out of here.
Most larger social media companies do use a system that automatically flags content that is CSEM or is related to CSEM using a "digital fingerprint" (hash value) for content that has been previously identified as CSEM and added to a database. Most ISPs in the US also use this and can identify if it has been downloaded to your computer in many cases. The content of the video itself does not include any of this which is probably why it evaded detection initially. Youtube sucks at moderating non-English videos as well which doesn't help.

Considering it is Google I'm sure they use this system (supplied by Missing & Exploited children in tandem with federal information if I remember correctly) and for it to work properly there has to be a hash value associated with the content in question. I'm hoping AI will be an immense boon to detecting content like this in the future without the need of the hash value (by learning and associating the sort of weirdass coded language they use with potential CSEM,) but the assholes who did this seemed to be well aware. Most "gates" to CSEM on the clearnet are intentionally innocous and full of code meant for predators to identify. I'm not defending Youtube because every second the video is up is geniunely retarded.
 
Null features funny stuff. This is still a gossip site to him, not a call to action kind of thing, even if it has turned into one due to censorship everywhere else.
Dont forget too that newfeaturefags hardly look at any of the newer replies for context and as a result shit up the thread easily with a bunch of replies. So it's probably wise to not feature super serious stuff like this lest it gets overrun with newfags drowning out more informal discussion.
 
Dont forget too that newfeaturefags hardly look at any of the newer replies for context and as a result shit up the thread easily with a bunch of replies. So it's probably wise to not feature super serious stuff like this lest it gets overrun with newfags drowning out more informal discussion.

I politely disagree and think that some important things might be worth being put forward, but perhaps in another feature feed. You can have "hahalolfunnygossip" and another feature for "fucked up shit that can only be brought forward here because the elites cover up abusing kids."

But, well, that's not my call. KF is a gossip site per the ooperator.
 
I'm not defending Youtube because every second the video is up is geniunely retarded.
I did a quick drive by post, and I suppose I wasn't as clear as I should be. I'm not trying to say Google's board or whatever is involved, I'm saying I believe some of the people work for google/alphabet. Same with the FBI/CIA/ "feds" i don't think they have orders coming down from the top, but i do believe some peds work for google and the "feds".
And I understood you there, you are just explaining it, that isn't taking up for anyone or anything, that's just explaining.
 
Islamic country.

You will notice a lot of the thumbnails shown here also feature hijabs. Seems to be a common theme. There seems to be a huge demand for hijab porn and a tiny supply. Doesn't really explain why they would mix CSAM and Zoo shit on it though. Maybe it's a case of being severely repressed leading to massive overcompensation by getting real freaky?
I think it's also just poverty and lack of law enforcement alongside a cynical lack of people there 'giving a shit' which is pretty sad and disgusting tbh. Apathy does lead towards the worst crimes coming about until its way too late.

Plus, Indonesia is basically made up of thousands of disconnected islands which I'd imagine would be a bitch to enforce. Any of those obscure islands can be another Little St. James when there's no one patrolling around to make sure any chomo activity going about.

Although if it means anything.

While I've heard of a lot of bad shit coming out of Indonesia, the Philippines on the other hand who are pretty much phenotypically similar (they're both from the same Austronesian genetic stock and geographic region) have a difference in religion.

Philippines is majority Catholic while Indonesia is majority Muslim. The worst things I heard about Philippines is that I fucking hate their annoying ass accents when I call up customer support and also when I'm playing games if they have a mic on.

... But at least no monkey torture or CSAM from them, so 1 Point for Christians I guess.
 
Back