New User Question

[OHAD]

So, I'm a seventh grader, and in the last few years, I have been very interested in Cellular Automatas, especially Life-Like, and of course, the most famous of them- Conway's Game Of Life.
I only signed to the forum today, but I've been looking at the forum for a few weeks. Unfortunatly, I don't know how to use the codes of patterns people are putting in a lot of threads (things like x = 40, y = 17 11bo$10bobbo12bo$5b3obbo14bobo11bo$10bo10boobb3oboo3bo4bo$boo11boboobb
o7b4o3b4o$obbo6bob..). Can I get any hint about how things work in here, so I'll be able to join in the discussions?
I'm pretty sure you have to use it in the Golly program, but it's hard for me to use the program. I haven't found any place to paste in the code.
Thanks ahead.

[dvgrn]

I'm pretty sure you have to use it in the Golly program, but it's hard for me to use the program. I haven't found any place to paste in the code.

What I've been using lately is the "Select All" button for the RLE code in the forums -- then Ctrl+C to copy it to the clipboard -- then switch to Golly and choose File > Open From Clipboard (or Ctrl+Shift+O). This changes the rule if necessary, sets the zoom level right, and drops the pattern into the middle of the screen, all at once.

If you have valid RLE in your clipboard, just a plain old Paste operation will drop the pattern into the current Golly universe, but it only changes the rule if the universe is empty, and then you have to click a location to actually place the pattern. For big patterns that can be confusing -- you just see part of a big boundary rectangle attached to the mouse, and all of the actual pattern may be offscreen until you click and re-center.

For small patterns, though, just paste, move the mouse into the Golly universe and click anywhere -- should work fine.

Welcome to the forums!

[OHAD]

Thank you! I haven't thought it'll be so simple. I'm now looking at some patterns people are putting in the forums, and it seems like a big part of the discoveries in the GoL world comes from this very forum :)
I'm starting a new, simple research about the odds of a random-generated pattern to be a Spark (if I get the defenition right, Spark is a pattern that eventually dies, without leaving any still lifes, oscillators, spaceships, or any infinite-growers. So, it's really easy to proof that when the bounding box for the starting generation is N*N, the number of possible starters is a bit less than (2^(N^2))-(2^((N-1)^2)). Please fix me if I'm wrong. I was putting the right side in the equation- 2^((N-1)^2), because I tried to delete all of the possible combination that don't touch the edges. They have a smaller bounding box. Of course, I didn't count things like the bounding box N*(N-1), so I'll need some help in here.
After solving this equation, I'll start with putting it in all sizes. It gets large pretty soon, so it might be hard.
Any help with this?

EDIT #1: I can't seem to find any simple solution. I'm just trying to find all patterns in the Bounding Box N*N, that can't be expressed with a smaller BB. Maybe I'll start the equation again with N*(N-1), that can make it simpler.

[dvgrn]

... when the bounding box for the starting generation is N*N, the number of possible starters is a bit less than (2^(N^2))-(2^((N-1)^2)). Please fix me if I'm wrong. I was putting the right side in the equation- 2^((N-1)^2), because I tried to delete all of the possible combination that don't touch the edges.

Hmm, you can almost handle this by saying that your N^2 object has at least one ON cell along the top, one on the left edge, one on the right edge, and one along the bottom.

But a corner cell can count for two edges, so that's two to four required ON cells (for anything bigger than the N=1 case). All the other N^2 minus 2-or-3-or-4 cells can be either on or off.

So for each given set of four ON cells at the edges, that's 2^[N^2-4] possible patterns. The problem is that some of these objects will be duplicated in the lists for other sets of ON cells -- not to mention rotations and reflections, if you want to handle those... Enumerating the possibilities for objects with one corner or two adjacent corners ON is also a bit more complicated.

But something along these lines might be worked out eventually. I wouldn't be surprised if this problem is listed somewhere in the On-Line Encyclopedia of Integer Sequences, actually -- might be worth a look there.

After solving this equation, I'll start with putting it in all sizes. It gets large pretty soon, so it might be hard. Any help with this?

This is definitely something Golly would be good for. Really the easiest way might be to pick the largest reasonable N you can work with -- 5x5 is clearly doable, but 6x6 has 68 trillion possible sparks... which is certainly not beyond 21st century computing power, but it will take a lot of CPU time.

Anyway, you could write a Golly script to generate each 5x5 pattern in order, find its bounding box (this is a one-liner for Golly) and run it. If it's a spark, add it to the count for its particular bounding rectangle -- you can get counts for all smaller MxN bounding boxes that way, at no extra cost.