Posted on January 1, 1  •  2 minutes  • 361 words

I was at a party dressed as a goth hottie, hanging out with Matt Godbolt talking about various software exploits people would run on his compiler explorer (godbolt.org) that he’s had to deal with. He ended up asking me if I ever had to deal with any software exploits. I said nah I’m not popular to have people attacking me but I made an AI bot that attacks my own stuff and then e-mails me the successful attempts. Sometimes it gives me a false positive because it writes a faulty-proof of-concept, like it would write undefined behaviour and then email me that it’s own PoC doesn’t work as expected (because clearly any unexpected results means it’s successfully exploited something). He asked me of what kinds of undefined behaviour does it typically come up with and I said it’s just the noob shit like the order that parameters are evaluated. Some random guy (who looked like an annoying guy from my highschool) told me that Python is fixing that instance of undefined behaviour. I told him no it’s just a draft RFC and it’s going to take years to fix. We ended up physically fighting over it until I pushed him to the ground and pulled up the page and showed him that it’s only set to be finalized by 2027 and it’s currently 2024. This was Matt’s party, apparently so he just banned the guy from the event causing every bone in his body to instantly break, he then had to be wheeled out of the event on a stretcher. Matt and I then started talking about how people were supposed to bring a cool piece of software to the party but some people are really lazy like a guy we saw walking past who brought Ghidra. Matt proceeded to break that guy’s bones too. We then started talking about what makes software very well known, like if you were to take the first let’s say 10k most starred repos on GitHub, how many of those are things that you wouldn’t even know about? and would those qualify as cool software for the party? I don’t remember what happened after that.