sodiboo :pride_heart:@sodiboo@gaysex.cloud
honestly why don't they make kernel anti cheats open source? or at least, source available?
they literally rely on hardware attestation to prove that you're running the correct kernel module and nothing else. this is why you can't virtualize them or emulate them. it needs secure boot and a TPM and your VM won't have valid keys for that.
like. you fucks. assuming your software is robust (which it generally is! because, cheaters in these games often rely on HARDWARE level cheating;!!), you could literally publish the source code, let us build from source, and run it on Linux without needing to load arbitrary proprietary kernel code.
adding attack surface in kernel space is still a terrible idea, but at least you'd know the kernel module isn't literally designed to be malicious and actually harmful. but currently, with proprietary kernel level anti cheats on windows as they exist today, you do need to simply trust the developer. wtf
Latte macchiato :blobcoffee: :ablobcat_longlong:@privateger@plasmatrap.com
@sodiboo@gaysex.cloud
disagree tbh. no anticheat is actually “robust” software.
the ways to manipulate stuff on the client are so incredibly vast nowadays that you cannot hope to come close to covering all the possible ways. Most AC heuristics are reactive by analyzing samples of confirmed cheaters and whatnot.
most cheaters are still using "simple" external cheats that somehow grab a handle into the process in an undetected way (be that with a vulnerable driver, software, whatever). Kernel-level anticheats still cannot defeat these reliably on Windows, not even to speak of the DMA cheaters.
Gamers don't care, though, and they won't care probably ever 
Making server-side anticheats costs money, you can just slap an anticheat on a client and feel good about yourself, so there's no incentive for developers either