Code: Select all
g++ -std=c++11 -O3 -march=native -o ntzfind ntzfind.cppCode: Select all
g++ -std=c++11 -O3 -march=native -o ntzfind ntzfind.cppAll good so farglobins wrote: March 12th, 2020, 2:17 am I downloaded ntzfind today and I cannot get it to run on my computer. I'm using Windows and I have Cygwin downloaded. I've gotten as far as compiling it (although I'm not sure I'm doing that right, since I know nothing about C or C++). I just used the command that was listed in the readme:Code: Select all
g++ -std=c++11 -O3 -march=native -o ntzfind ntzfind.cpp
Try the following command to get usage instructions:Whenever I try to run it (using ./ntzfind, I think?), I get an error that says "floating point exception (core dumped)."
Code: Select all
./ntzfind cYou need to run programs built with Cygwin from the Cygwin terminal, so this error is not unexpected.When I click on the .exe file that was generated by the compilation, it gives me a message that says "The program can't start because cygwin1.dll is missing from your computer. Try reinstalling the program to fix this problem." I have cygwin1.dll on my computer; I went and found the file in C:\cygwin64\bin and it's there. I tried reinstalling Cygwin and nothing changed. How do I fix this?
Code: Select all
x = 23, y = 14, rule = B36/S1258
4$6b5o$7b3o!
Code: Select all
ooo
o.o
oo.
o.o
oo.
oo.
.o.
.oo
o.o
o.o
.oo
...
oo.
...
Code: Select all
ntzfind B36/S1258 p7 k1 w3 u e initrows.txt
This is pretty much all correct, except that the initial partial isn't long enough to generate a valid set of initial rows. As a rough guide you need p-k+1 extra rows in a single phase partial to generate the initial rows across all phases. So providing the set of rows you generated to zfind results in it finding a spaceship tail that moves at c/7, but it doesn't match a valid frontend. For such a short partial you could try selecting the row three ahead of the front row and running get-rows.py with that. It should generate a set of rows which are mostly empty but with that dot spark showing up in the later phases. You'll also want to start with a wider search. Here's the initrows.txt for width 5:globins wrote: March 12th, 2020, 4:10 am I just ran a search for an expansion of this partial spaceship in B36/S1258 and it returned a result instantly, claiming that it already was a spaceship (it's not).Did I do something wrong? Here's what my initrows.txt looked like, which I generated using Sokwe's python script:Code: Select all
x = 23, y = 14, rule = B36/S1258 4$6b5o$7b3o!Code: Select all
ooo o.o oo. o.o oo. oo. .o. .oo o.o o.o .oo ... oo. ...
Code: Select all
.....
.....
.....
.....
.....
.....
.....
.....
.....
....o
.....
....o
.....
....o
Code: Select all
x = 9, y = 9, rule = B36/S1258
2b5o$3b3o2$b7o$2obobob2o$4bo$4bo$2bobobo$2obobob2o!
This part is correct, except as mentioned above you'll want to run a wider search - at least w6 or w7.My command looked likeCode: Select all
ntzfind B36/S1258 p7 k1 w3 u e initrows.txt
Were you perhaps running a width 11 search? As noted in the README that requires a lot of RAM (up to 19GB for some searches). w10 requires somewhere between 2-4 GB. ntzfind grows the main look up table as required and if there is insufficient RAM at some point then it will crash.globins wrote: March 15th, 2020, 11:25 am I set a search to run overnight and when I looked this morning I had a “segmentation fault” message. What do I do about that?
Please see my post on the other thread: https://conwaylife.com/forums/viewtopic ... 234#p91234globins wrote: March 15th, 2020, 2:04 pm You recommended using gfind for larger widths. What's the difference between that and qfind?