Thread for your script-related questions

[R2INT]

I am trying to run a Python script in Golly, but it gives me this message:
Could not load the Python library

How should I install the Python library on my Windows 11 laptop?

[PK22]

I am trying to run a Python script in Golly, but it gives me this message:
Could not load the Python library

How should I install the Python library on my Windows 11 laptop?

Find the file 'python3*.dll', but replace '3*' with your version number (e.g if you have v3.13 installed, find python313.dll).
It should be in your Python installation somewhere - mine is at C:\Users\PC\AppData\Local\Programs\Python\Python313\python313.dll, for example.
Golly should be asking you for the location of the python3* dll, so enter the file path for it.

If it does not work, then close down all instances of Golly, and find your gollyprefs file (should be at C:\Users\<username>\AppData\Roaming\Golly\GollyPrefs). Open it in Notepad, find where the python3_lib path is shown in the file, and replace it with the file path to python3*.dll. Then reopen Golly and try running the Python script again. It should work.

[Tawal]

[…)
Is there a search program for finding welds?

Take a look on mkstill from Bellman's program suite.
It has a very special input file but it can be easily generated by a script.
However, there is also CompleteStill from Barrister's program suite, but I' can't run this one due to my poor machine.
Both of them need compilation before use and have a README file.

For mkstill, I wrote a script to convert a pattern in LifeHistory to Bellman's mkstill program and launch it in Golly.
It's a bad idea. Prefer to launch it in a terminal/console (need some modifications before).
By the way, you will get at least the conversion function.
Here is the script : link into this forum

Edit:
Barrister is the successor of Bellman and is more friendly and efficient.

Edit2:
Here is a link to a PDF for Bellman's explanation (and mkstill so and more) : Bellman's suite Notice - PDF

[Resu]

Does anyone use Glue?

[PK22]

Does anyone use Glue?

Probably not - HoneySearch is much faster than Glue, and when I tried to find a copy of Glue a while back so I could test it, I failed.
The only reason that Glue would be better is if it were being run on a horrible PC with one core and barely any RAM, or if you only had access to Windows without WSL - HoneySearch is memory-intensive and parallelised.
Still, it looks pretty interesting - does anyone have a copy of the source code?

[Resu]

What is D2_+1_gO1s2?

[PK22]

What is D2_+1_gO1s2?

It is essentially D2_+1, but with a gutter between the two soups, and they are also skewed from each other by 2 cells.

[Resu]

It is essentially D2_+1, but with a gutter between the two soups, and they are also skewed from each other by 2 cells.

Is there one for other symmetries?

[ThePlayzr]

How do I put an input into qfind and run it in visual studio code?

[Resu]

How do I put an input into qfind and run it in visual studio code?

Put the input into vscode or put qfind into vscode?

[ThePlayzr]

How do I put an input into qfind and run it in visual studio code?

Put the input into vscode or put qfind into vscode?

Put the input into vscode.

[PK22]

When trying to compile HoneySearch (by running make), I get the following command output:

Code: Select all

g++ -c -Wall -Wextra -O3 -march=native -fopenmp -std=c++11 -pedantic hs.cpp -o hs.o
hs.cpp: In member function ‘void HoneyThread::setrule(std::string)’:
hs.cpp:277:42: error: use of deleted function ‘apg::base_classifier<1>& apg::base_classifier<1>::operator=(apg::base_classifier<1>&&)’
  277 |         cfier = apg::classifier(&lt, rule);
      |                                          ^
In file included from hs.cpp:2:
../lifelib/classifier.h:18:11: note: ‘apg::base_classifier<1>& apg::base_classifier<1>::operator=(apg::base_classifier<1>&&)’ is implicitly deleted because the default definition would be ill-formed:
   18 |     class base_classifier {
      |           ^~~~~~~~~~~~~~~
../lifelib/classifier.h:18:11: error: use of deleted function ‘apg::lifetree<unsigned int, 2, apg::lifemeta<unsigned int> >& apg::lifetree<unsigned int, 2, apg::lifemeta<unsigned int> >::operator=(apg::lifetree<unsigned int, 2, apg::lifemeta<unsigned int> >&&)’
In file included from ../lifelib/pattern2.h:2,
                 from hs.cpp:1:
../lifelib/lifetree.h:1897:11: note: ‘apg::lifetree<unsigned int, 2, apg::lifemeta<unsigned int> >& apg::lifetree<unsigned int, 2, apg::lifemeta<unsigned int> >::operator=(apg::lifetree<unsigned int, 2, apg::lifemeta<unsigned int> >&&)’ is implicitly deleted because the default definition would be ill-formed:
 1897 |     class lifetree : public lifetree_generic<I, N, J> {
      |           ^~~~~~~~
../lifelib/lifetree.h:1897:11: error: use of deleted function ‘apg::lifetree_generic<unsigned int, 2, apg::lifemeta<unsigned int> >& apg::lifetree_generic<unsigned int, 2, apg::lifemeta<unsigned int> >::operator=(const apg::lifetree_generic<unsigned int, 2, apg::lifemeta<unsigned int> >&)’
../lifelib/lifetree.h:16:11: note: ‘apg::lifetree_generic<unsigned int, 2, apg::lifemeta<unsigned int> >& apg::lifetree_generic<unsigned int, 2, apg::lifemeta<unsigned int> >::operator=(const apg::lifetree_generic<unsigned int, 2, apg::lifemeta<unsigned int> >&)’ is implicitly deleted because the default definition would be ill-formed:
   16 |     class lifetree_generic : public lifetree_abstract<I> {
      |           ^~~~~~~~~~~~~~~~
../lifelib/lifetree.h:16:11: error: use of deleted function ‘apg::hypertree<unsigned int, 4, apg::lifemeta<unsigned int>, apg::nicearray<long unsigned int, 8>, apg::lifemeta<unsigned int> >& apg::hypertree<unsigned int, 4, apg::lifemeta<unsigned int>, apg::nicearray<long unsigned int, 8>, apg::lifemeta<unsigned int> >::operator=(const apg::hypertree<unsigned int, 4, apg::lifemeta<unsigned int>, apg::nicearray<long unsigned int, 8>, apg::lifemeta<unsigned int> >&)’
In file included from ../lifelib/lifetree_abstract.h:3,
                 from ../lifelib/lifetree.h:4:
../lifelib/hashtrees/hypertree.h:70:11: note: ‘apg::hypertree<unsigned int, 4, apg::lifemeta<unsigned int>, apg::nicearray<long unsigned int, 8>, apg::lifemeta<unsigned int> >& apg::hypertree<unsigned int, 4, apg::lifemeta<unsigned int>, apg::nicearray<long unsigned int, 8>, apg::lifemeta<unsigned int> >::operator=(const apg::hypertree<unsigned int, 4, apg::lifemeta<unsigned int>, apg::nicearray<long unsigned int, 8>, apg::lifemeta<unsigned int> >&)’ is implicitly deleted because the default definition would be ill-formed:
   70 |     class hypertree {
      |           ^~~~~~~~~
../lifelib/hashtrees/hypertree.h:70:11: error: use of deleted function ‘apg::kivtable<apg::nicearray<long unsigned int, 8>, unsigned int, apg::lifemeta<unsigned int> >& apg::kivtable<apg::nicearray<long unsigned int, 8>, unsigned int, apg::lifemeta<unsigned int> >::operator=(const apg::kivtable<apg::nicearray<long unsigned int, 8>, unsigned int, apg::lifemeta<unsigned int> >&)’
In file included from ../lifelib/hashtrees/hypertree.h:26:
../lifelib/hashtrees/kivtable.h:58:12: note: ‘apg::kivtable<apg::nicearray<long unsigned int, 8>, unsigned int, apg::lifemeta<unsigned int> >& apg::kivtable<apg::nicearray<long unsigned int, 8>, unsigned int, apg::lifemeta<unsigned int> >:operator=(const apg::kivtable<apg::nicearray<long unsigned int, 8>, unsigned int, apg::lifemeta<unsigned int> >&)’ is implicitly deleted because the default definition would be ill-formed:
   58 |     struct kivtable : public hh::indextable<kiventry<K, I, V>, K, kivtable<K, I, V>> {
      |            ^~~~~~~~
../lifelib/hashtrees/kivtable.h:58:12: error: use of deleted function ‘hh::indextable<apg::kiventry<apg::nicearray<long unsigned int, 8>, unsigned int, apg::lifemeta<unsigned int> >, apg::nicearray<long unsigned int, 8>, apg::kivtable<apg::nicearray<long unsigned int, 8>, unsigned int, apg::lifemeta<unsigned int> >, 8>& hh::indextable<apg::kiventry<apg::nicearray<long unsigned int, 8>, unsigned int, apg::lifemeta<unsigned int> >, apg::nicearray<long unsigned int, 8>, apg::kivtable<apg::nicearray<long unsigned int, 8>, unsigned int, apg::lifemeta<unsigned int> >, 8>::operator=(const hh::indextable<apg::kiventry<apg::nicearray<long unsigned int, 8>, unsigned int, apg::lifemeta<unsigned int> >, apg::nicearray<long unsigned int, 8>, apg::kivtable<apg::nicearray<long unsigned int, 8>, unsigned int, apg::lifemeta<unsigned int> >, 8>&’
In file included from ../lifelib/hashtrees/kivtable.h:16:
../lifelib/hashtrees/../cpads/include/cpads/indextable.hpp:89:7: note: ‘hh::indextable<apg::kiventry<apg::nicearray<long unsigned int, 8>, unsigned int, apg::lifemeta<unsigned int> >, apg::nicearray<long unsigned int, 8>, apg::kivtable<apg::nicearray<long unsigned int, 8>, unsigned int, apg::lifemeta<unsigned int> >, 8>& hh::indextable<apg::kiventry<apg::nicearray<long unsigned int, 8>, unsigned int, apg::lifemeta<unsigned int> >, apg::nicearray<long unsigned int, 8>, apg::kivtable<apg::nicearray<long unsigned int, 8>, unsigned int, apg::lifemeta<unsigned int> >, 8>::operator=(const hh::indextable<apg::kiventry<apg::nicearray<long unsigned int, 8>, unsigned int, apg::lifemeta<unsigned int> >, apg::nicearray<long unsigned int, 8>, apg::kivtable<apg::nicearray<long unsigned int, 8>, unsigned int, apg::lifemeta<unsigned int> >, 8>&)’ is implicitly deleted because the default definition would be ill-formed:
   89 | class indextable {
      |       ^~~~~~~~~~
../lifelib/hashtrees/../cpads/include/cpads/indextable.hpp:89:7: error: use of deleted function ‘hh::ivector<T, B>& hh::ivector<T, B>::operator=(const hh::ivector<T, B>&) [with T = apg::kiventry<apg::nicearray<long unsigned int, 8>, unsigned int, apg::lifemeta<unsigned int> >; int B = 8]’
In file included from ../lifelib/hashtrees/../cpads/include/cpads/indextable.hpp:2:
../lifelib/hashtrees/../cpads/include/cpads/ivector.hpp:121:20: note: declared here
  121 |     ivector<T, B>& operator=(ivector<T, B> const&) = delete;
      |                    ^~~~~~~~
make: *** [makefile:31: hs.o] Error 1

What is the problem, and how can I fix it?

[vilc]

When trying to compile HoneySearch (by running make), I get the following command output:

Code: Select all

...

What is the problem, and how can I fix it?

It looks like HoneySearch expects an older version of lifelib, one which does not rely upon the special aligned vectors "ivector", whose "operator=" is explicitly deleted for performance reasons. Try to downgrade lifelib to a version at least more ancient than october 27th, 2022, the date of this commit :
https://gitlab.com/hatsya/open-source/c ... f4967476a2
An even older version (from 2018 or 2019) might be required.

[Resu]

What are some search programs that can search for oscillators in isotropic non-totalistic rules?

[Katrina]

What are some search programs that can search for oscillators in isotropic non-totalistic rules?

Logic Life Search and rlifesrc are great for specific periods (mostly for when they are below period 7). For a brute force search with no specific period, use apgsearch.

[rabbit]

nbsearch2 is also a good utility for finding oscillators, works on explosive rules, and allows one to search all symmetries at once using the symmetry "All".

[Resu]

Is there a way to search for fuses?

[dvgrn]

Is there a way to search for fuses?

Well, Ivan Fomichev (codeholic) completed a successful fuse search for a particular type of fuse (c/2) that he was looking for to go with the pufferfish.

It's not hard to modify that search script to hunt for and collect other things. The problem is that if you allow much variety in how the fuse's wick is built, there's just absolutely no end to the number of fuses you can find very quickly. The difficulty isn't finding fuses, it's organizing them.

[Resu]

What does CatForce do?

[Sokwe]

What does CatForce do?

For questions like this, you should first read the CatForce readme and the CatForce wiki page. If there's something you're confused about after reading those, you could ask about it here, presumably with a more specific question.

[Katrina]

In Golly, why is there oplus/"overlay" support in Lua, but not Python? I feel like it'd be useful for both programming languages, and I rarely ever code in Lua (because it is so weird to me compared to most other programming languages).

[Resu]

What versions of gfind are there?
There seems to be a lot.

[Katrina]

Something I've noticed while running LLSSS, qfind, or other breadth-first-search programs:
Running searches with a low probability of finding results take a lot less time than searches with a higher probability, depending on the rule.
I have a theory that this is because of the amount of possible frontends, backends, pushalongs, tagalongs, wicks, etc. that can occur in that speed.
Is there a sufficient way to automatically find a rulespace that can support any spaceship with a specific speed? (For example, my estimated B2(45678)/S2(345678) space in OT for possible (2,1)c/3 ships, but less lenient)
What would that algorithm look like, if it's possible? (Note that I am not referring to finding ships themselves, just specific rule transitions that can lead to possible ships, like B2e/S4w/S5k(?) for c/3 diagonal in INT)

[NooneAtAll3]

does golly include endemicity analyzer?

[dvgrn]

In Golly, why is there oplus/"overlay" support in Lua, but not Python? I feel like it'd be useful for both programming languages, and I rarely ever code in Lua (because it is so weird to me compared to most other programming languages).

This was maybe something of a policy decision by Andrew, back almost a decade ago (2016) when Lua got added to Golly 2.8.

The idea was that Python just plain kept on being a huge hassle for a lot of users -- version 2.x going out of support in favor of 3.x, 32-bit vs. 64-bit, Microsoft Store installs vs. python.org installs vs. various competing embedded installs that show up with other software and then get in the way, admin installs vs. hidden AppData installs -- so we always had a big percentage of the user base who couldn't use Golly as it was intended to be used:

- notice something that you wish worked a little differently
- write a little script that makes Golly dance to the tune that you want
- map a keyboard shortcut to that script, maybe replacing whatever piece of functionality you don't like

Everyone's usage of Golly is a little different, so everyone should be able to customize Golly. But Python problems kept making that difficult for a lot of people.

There will never be versioning difficulties with Lua -- it's baked in to Golly and always Just Works with no additional installs.

So from the Golly Gang's point of view... if someone likes Python and has installed Python and it's working with Golly, and Python is what they write their own private helper scripts in, more power to them.

But scripts that are meant to be shared around with the community ... should ideally be written in Lua, just because they'll end up being so much more portable. So the goal with the Lua overlay was partly to gently push people Lua-ward, so that whatever cool new stuff showed up, everybody would be able to run it.

It hasn't really worked out that way -- yet. I suppose we can continue to do what we've done for the last nine years: just wait and see what happens as T approaches infinity.