A for awesome wrote:I really didn't expect this to work, but I figured it was worth a shot. Surprisingly, it does
Very cute! I love it!
A for awesome wrote:More results from dr
Again, it would be helpful if the results are more organized (or example, see the first post in this topic).
Since oscillators of periods 2-6 are so common, I don't personally bother constructing them. Low period oscillators can also "easily" be found using direct methods like WLS/JLS or ofind.
You should also provide all of the new rotors (including periods 2-6) for the knownrotors file (for example, see the second post in this topic). That way, people can update their knownrotors file to avoid finding your new oscillators.
Finally, you might want to try minimizing the stator of your new oscillators. The best tool for this is probably JLS. Here's how to do it:
- Start JLS and change the period to whatever your oscillator period is (change the width and height if necessary).
- Select the entire grid in JLS and right click (this fills the grid with off cells).
- Copy your oscillator from Golly and paste it into JLS.
- Select the entire grid and use ctrl/cmd + shift + arrow keys to shift the pattern into the center of the grid.
- In the "search" menu select "accept displayed state"
- The rotor cells of the oscillator should be highlighted with a thick black border. Select some static (i.e. stator) cells and press "1" on your keyboard. Do this until all stator cells have a yellow background.
- Select these yellow cells and press "c" to clear them.
- From the "search" menu select "options". Click the "constraints" tab and use the "No more than X on cells in generation 0" option to limit the number of cells.
- Run some searches and play around with the on cells limit until you find the minimum. Remember that the minimum bounding box might not allow for the minimum population.
A for awesome wrote:The period-doubling p10s would suggest a great number of 3n+1 even periods, even maybe p34, but I can't find any rotors that work for any period other than p5.
The 3n+1 phase shift reaction is well known. The p10 form can be seen in o0010.lif from jslife/osc, as well as a p22 form in o0022.lif:
Code: Select all
x = 47, y = 15, rule = B3/S23
9bo2bob2o21bo2bo2bo$5bob7obo21b7o$5b2o28b2o$8b4ob2o19bo3b4ob3o$5b3o5bo
bo15bo2bob2o5bo2bo$4bo4bobo3bo13b5obo3bobo3bo$3bobobobob4o13bo6bo2b2ob
4o$3bobo4bo16bob3ob2o5bo$2obobo2b2o2bo14bobo4bob4o2bo$2obobo3bob2o13b
2o2b2o2bobo2bob2o$4b2obobo18bobo2b3o2bo2bo$6bob2o18bo2bo7b2o2bo$6bo22b
obob2obo3bob2o$5b2o23b2obob2o3bo$39b2o!
So far, nobody has been able to find a p34 using this technique. There are a lot of known phase shift reactions. Just search this topic for "phase shift".
The other p5 period doubler is also known:
Code: Select all
x = 15, y = 14, rule = B3/S23
5b2ob2o$5bo3bo$6b2o$3b3o2b4o$2bo4bo3bo$bob4ob2obob2o$bobo4bo2bobo$2o2b
3obo4bo$bobo4bob3o$bob4ob2o$2bo4bo3b2o$3b3obo4bo$5b2o4bo$11b2o!
For p10 oscillators that only differ in their p5 rotor, jslife typically includes only the smallest example.