copyfail challenge: making Perl version smaller

dmitri has asked for the wisdom of the Perl Monks concerning the following question:

Replies are listed 'Best First'.
Re: copyfail challenge: making Perl version smaller by marto (Cardinal) on Apr 30, 2026 at 09:26 UTC
Context: https://copy.fail.	[reply]
Re: copyfail challenge: making Perl version smaller by Corion (Patriarch) on Apr 30, 2026 at 09:22 UTC
My guess would be that you would get better responses if you actually posted your attempt. People familiar with the usual golfing approaches can then spot locations where you can improve the golfing. The actual value of golfing an exploit is very low, but I guess the PR value of having a "really low number of bytes" in the headline is high.	[reply]
Re^2: copyfail challenge: making Perl version smaller by LanX (Saint) on Apr 30, 2026 at 11:58 UTC
> but I guess the PR value of having a "really low number of bytes" in the headline is high. I'm not sure about that PR value. The most likely way for outside attackers is to hide the exploit inside some "harmless" script/module. The message now is that all Python code is potentially dangerous, do we want to push to be in the same spotlight? Cheers Rolf _{(addicted to the Perl Programming Language :) see Wikisyntax for the Monastery}	[reply]
Re^3: copyfail challenge: making Perl version smaller by marto (Cardinal) on Apr 30, 2026 at 12:18 UTC
"The message now is that all Python code is potentially dangerous, do we want to push to be in the same spotlight?" I don't think that's the message at all. Don't blindly trust any code you don't understand or aren't as far as is reasonably practicable satisfied with the chain of trust? Don't curl things and pipe it to stuff without know what it's going to do first?	[reply]
Re^4: copyfail challenge: making Perl version smaller by LanX (Saint) on Apr 30, 2026 at 12:30 UTC
Re^2: copyfail challenge: making Perl version smaller by dmitri (Priest) on Apr 30, 2026 at 19:25 UTC
That's true -- I think I should have posted my code with the original comment.	[reply]
Re: copyfail challenge: making Perl version smaller by LanX (Saint) on Apr 30, 2026 at 16:10 UTC
> Don't run it on your own machine, though, as it will mangle your `su` binary The project's site says the effect is gone after rebooting. And even if there was a lasting effect, what about booting Linux from a stick on an old box for testing? Cheers Rolf _{(addicted to the Perl Programming Language :) see Wikisyntax for the Monastery}	[reply]
Re^2: copyfail challenge: making Perl version smaller by dmitri (Priest) on Apr 30, 2026 at 19:27 UTC
Yes, but until you reboot (or flush pagecache), the su command is bad. And some people do not reboot for a long time. I checked the functionality of the exploit in QEMU.	[reply]
Re: copyfail challenge: making Perl version smaller by stevieb (Canon) on Apr 30, 2026 at 09:05 UTC
At any rate -- I figured, we can make it smaller in Perl! A half hour and a bunch of tokens later, I failed. You're using pronouns like some "AI" deserves one. AI is not intelligent, and tokens don't denote thought, and "we" don't include me if AI is brought in to pretend to be something that thinks for you logically. What a fucking disgrace if this is what golf has come to. -stevieb	[reply]
Re^2: copyfail challenge: making Perl version smaller by harangzsolt33 (Deacon) on Apr 30, 2026 at 19:31 UTC
You may want to watch this movie. It is really an eye-opener about AI... THE THINKING GAME https://www.youtube.com/watch?v=d95J8yzvjbQ	[reply]