Note
This repo is in maintenance hiatus/stasis.
This allows any human to aid in the search for a counter-example that disproves the Collatz Conjecture.
However, currently, this repo is just for learning purposes.
Important
Expect breaking API changes at literally any time.
I won't provide support for anything, but you can ask questions if you're curious about stuff.
This isn't meant for production environments, at all.
I won't publish it to any package repository.
Install:
cargo install --git https://github.com/Rudxain/collatz_finder.git
ℹinfo:
colfind help
"What to do if I find a counter-example?"
This is not a stupid question! I would also not know what to do in one of the most epic moments in the history of math and CS!
I suggest you to post your results everywhere! except Wikipedia. To make people take you seriously, you must post the numeral that disproves CC. It'll be H U G E , so I recommend posting a gist containing the full numeral, and then share links to that gist.
Tip for smaller size: hexadecimal is more compact than dec, raw-binary is more compact than hex. If you use hex or raw, disambiguate endianness.
Another tip, ensure nobody else takes credit for it. I know that sounds egotistical, but nobody wants to give countless hours of computing power for free, am I right?
Note
The program assumes that numbers don't diverge to +♾️, it only searches cycles. So it could enter an infinite loop that allocates increasingly more memory, until an OOM panic happens.
It supports negatives. It also supports multiple bases/radices (not anymore, sorry 🙁, I might fix it later).
The rationale behind radix support, is that CC is more interesting and helpful to explore/experiment in bases 2 & 3 and any other base whose factors are 2 and/or 3, and bases whose factors are +-1 offset from 2 and/or 3, so I had to add all bases from 1 to 10 (and hexdec, because it's binary in disguise)