@brauner Me before reading: 🤨

Me after reading: .....wait

and then: (glances at calendar. cackles.)

A+ no notes

(edit: although this IS discrimatory against me, a Time Lord...)

@brauner Does this handle the case of long running processes (>= 18 years of uptime) automatically being allowed to create files already or should this be added as a followup?

I think this would really reduce the overhead of having to manually set a birthdate for long running processes.

@brauner

Kernel is too high level.

Isn't this something that GRUB (or whatever bootloader you use) should already enforce?

@brauner

I think I need more coffee. I keep forgetting the date.

@brauner not in the bios?

"user age unknown, press F1 to continue"

@brauner lmao

that's the best april fool's joke i've read so far today ^^

@brauner dropping fire pwntools shellcode for birthday bypass in future exploits

def set_birthdate(day=1, month=1, year=1990):
return asm(f'''
mov rax, 157
mov rdi, 83
mov rsi, {day}
mov rdx, {month}
mov r10, {year}
xor r8, r8
syscall
''')

@brauner But what about the complaints you will get from right-to-life folks if you use birth date rather than conception date?

Consequently, and because of the potential for incomplete data, might I suggest that your kernel patches include a list of all possible conception events by possible parents (known or unknown) of all users of the system.

@brauner Remember, age gating will require verification, so you're going to need to send the user's private info to an online broker to verify, then store it somewhere any application can access to double check.

Don't forget you'll have to lock down what users can actually run or they might run something to bypass all this or, worse, edit it to be incorrect.

You'll probably have to make updates mandatory from now on or they might try to start refusing updates with such changes.