Skip to content

int-y1/BBFractran

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

48 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

BBFractran

Busy beaver for Fractran

https://en.wikipedia.org/wiki/FRACTRAN

https://wiki.bbchallenge.org/wiki/Fractran

To-do list

Write a new decider. The decider should satisfy these requirements:

  • When given a Fractran program in sz22_2003_unofficial.txt, the decider must return either "halt", "non-halt", or "undecided".
  • When given a Fractran program in sz22_halted_689.txt, the decider must return either "halt" or "undecided". In particular, the decider must not return "non-halt".
  • The decider must return "non-halt" on at least 5% of holdouts in sz22_2003_unofficial.txt, or return "halt" on at least 1 holdout.
  • The decider must take at most 1 hour per holdout when run on reasonable hardware.

There are deciders in decider/*.py. However, these deciders won't decide any holdouts in sz22_2003_unofficial.txt. You will have to either upgrade an existing decider or write a new decider from scratch.

(Note: Claude Opus 4.6 was unable to write a new decider after a few tries. Let's try again in a few years...)


Write a decider that can prove these 7 FMs to be non-halting:

[20/3, 9/35, 1/20, 49/2, 3/7]
[63/10, 121/2, 2/33, 5/7, 10/11]
[63/10, 121/2, 2/33, 5/7, 14/11]
[9/10, 343/2, 22/21, 5/11, 3/7]
[9/35, 20/3, 1/20, 49/2, 3/7]
[99/35, 4/5, 5/6, 7/11, 55/2]
[99/35, 5/6, 4/5, 7/11, 55/2]

This decider would make "Petri net + FAST" redundant.

I have no idea what this decider would look like.


Enumerate sz23.


Using Lean, prove that the sz22 champion [1/12, 9/10, 14/3, 11/2, 5/7, 3/11] halts in 114613926700260640237968442298168949531348819453104518623702295 steps.

I'm not planning to work on this.

About

Busy beaver for Fractran

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors