NailLegProcessorDivide

I agree server throughput shouldnt be a problem. I think it would be nice to have a bit more feedback than "5 hours" (I know this was an example not...

Hi, yes, I created a basic server and client that exchange jobs over a rest API mostly as a proof of concept, I havent worked on it in a little...

Just got done running N=18, it was a lot quicker than I initially feared as it sped up over time due to how the first instance of polycubes worked. I...

Very cool, Im going to need to see if I can find time to take a look at how this compares to my current implementation.

wow. just started looking at @dzamlo's rust version and agree its extremely fast. would love to see it integrated into this repo some how. ~ 3x the current DFS solution...

I did some more optimisations after @datdenkikniet's patch reusing already allocated structures (small win) and swapping the `Vec3D`s with a separate `Vec3dBool` that stores an fixed array of 64x64 u64s...

r9 7900X so "only" 12c/24t. I'll see if I can make a gitlab to push with. the code is a bit nasty and I dont like that it hardcodes the...

https://gitlab.com/dzamlo/polycubes2/-/merge_requests/2 hopefully that link should work I forgot to mention there was quite a bit of perf to gain in unsafely accessing the Vec3dBool as well. making vec bool 32x32x32...

``` number of polycubes for n=17: 457409613979, 74301822 ms real 1238m21.828s user 28524m59.824s sys 9m41.177s ``` Finally, N=17 in a bit over 20.5hrs getting the same number as @hidny. a...

I've made a fairly naive networked solution at https://gitlab.com/dzamlo/polycubes2/-/merge_requests/4 @dzamlo if you could update your repo to an MIT or other permissive licence as well it would be nice to...