On 2022-05-16 19:30, Nicholas Drozd wrote:
> Pavel, is there any chance you could put your code online? I
> google-translated your old thesis, but that only got me so far :)
it should be possible to download my old code somewhere along the
thesis, but as you experienced with the text - even for me it was easier
to invent it from scratch than to understand myself from a decade ago -
it's really horrible experimental c++. and it's overly complicated and
ineffecient compared to what it can handle in the end => so i started
new version in rust :).
i can opensource my new code, but:
* it will not help you right now - 2 most important parts - computing
nested minima & divisibility conditions are not yet right. i
simplified/optimized that part compared to old code, but it's tricky to
get it right - that was the place i left that code 1.5y ago - i needed
gui to debug it and gui is one of the last things that doesn't have a
success story in rust yet. but I finally hacked one :)
* rado's `busy beaver competition` would lost its `competition` part.
any idea how to frame collaborative effort afterwards?
however, science is about making progress, not winning, and my code is a
good starting point from where to hunt interesting machines - for
example doomsday variation of those machines Shawn is searching for - a
machine that halts, but it's not possible to compute the number. (the
new code has been shaped to make it easier to implement the 'rozbor
pripadov' section from my thesis to make it possible). plus there are
more things i would like to see implemented, but it would probably
exceed my time budget for this "round" - happy to share the load.
and the final quest - it's time to finalize BB(5) and get close to BB(6)
- but there are real monsters that would need yet another proof/tape