Javascript

Part one was easy, though despite starting at midnight I only placed 1786 for part one. I think my tendency to want to do OOP makes it take longer…

Part two… Well, I figured it was some sort of binary circuit for trying to add binary numbers. So I hoped that the sum of the x registers and the y registers was the expected result of simulating the circuit like in part one. I later verified that it is the expected result.

I didn’t want to try and manually figure out the bad outputs, coffee wasn’t helping, I wanted sleep. So I uh… I wrote logic to randomly create swaps. And then just hoped RNG got me covered. To help my chances, I ran it on 8 different processes.

When I woke up in the morning I discovered 8 stopped processes, each with “a solution” that was different. Turns out, if you just randomly swap wires at some point you get a system that outputs the desired result - but only because you sufficiently screwed it up more to produce the expected result, even if the system itself would not work for other input.

I could probably change the registers to another value, run it, and see if they match, thus ruling out an incorrect set of swaps causing a correct result with the original binary inputs. But at this point I just decided to do it the manual way following advice on here. My brain is fried, I’m stepping away to take a shower and get ready to visit family.

I had really hoped the bruteforce would work, I modified the bruteforce to run even after it finds a match and I’ll let it run while I’m gone today and see if RNG produces any correct result at some point - I just fear the exponential answer timeout will prevent me from submitting these correctly incorrect combinations lol. I might modify it later with my theory above and just run it on a server indefinitely and see if it produces the correct result eventually.

https://blocks.programming.dev/Zikeji/9e4d6e81595d4845b88cf98eb91852d8

Edit:

Created a raw multithreaded bruteforce variant: topaz