Page 70 of 178

Re: Oscillator Discussion Thread

Posted: April 3rd, 2021, 1:29 pm
by HartmutHolzwart
No chance for a p31 gun?

Re: Octagon oscillators

Posted: April 4th, 2021, 5:48 pm
by Hdjensofjfnen
Intriguingly, two stabilizations of Octagon 5 have shown up on Catagolue in C4_1:

Code: Select all

x = 19, y = 19, rule = B3/S23
8b3o$9bo2$7bobobo2$6bo5bo$5bob5obo$3bo2bo5bo2bo$o5bo2bo2bo5bo$2obo2bobobobo2b
ob2o$o5bo2bo2bo5bo$3bo2bo5bo2bo$5bob5obo$6bo5bo2$7bobobo2$9bo$8b3o!

Code: Select all

x = 17, y = 17, rule = B3/S23
5b2ob2o$5b2obo$5bo4bo$6b5o$5b7o$4bo7bob3o$2b3o7b2ob2o$o2b2o3bo3b2o$2ob2o2bobo
2b2ob2o$3b2o3bo3b2o2bo$2ob2o7b3o$3obo7bo$5b7o$6b5o$6bo4bo$8bob2o$7b2ob2o!
Octagon "sides" for Octagons 3, 4, and 5 can be combined, for instance:

Code: Select all

x = 19, y = 15, rule = B3/S23
6b2ob2o$5bo3bo$6bo4bo2$6bo5bo$2b2obob5obob2o$o5bo5bo5bo$o5bo5bo5bo$o5b
o5bo5bo$2b2obob5obob2o$6bo5bo2$7bo4bo$9bo3bo$8b2ob2o!

Re: Oscillator Discussion Thread (Octagon oscillators)

Posted: April 4th, 2021, 6:58 pm
by Sokwe
hotcrystal0 wrote:
March 28th, 2021, 10:25 am
Hdjensofjfnen wrote:
April 4th, 2021, 5:48 pm
Discussion moved to the oscillator discussion thread. Other octogon-type stabilizations are given in jslife:

Code: Select all

x = 111, y = 25, rule = B3/S23
4bo8bo46bo7bo18bo$4b3o4b3o46b2o5b2o14b2obobo4b2obob2o$7bo2bo26b3o20b2o
5b2o14b2obobo4bob2ob2o3b2o$6b2o2b2o26bo22bo5bo18bob2o13bo$2o16b2o9b2o
52b4o2bo$bo6b2o8bo11bo6b3o22b5o15bo4b18o$bobo3bo2bo5bobo11bobo3bo3bo4b
3o13bo5bo4b3o8b3obo17bo$2b2o2bo4bo3bobo13b2o2bo5bo2b3o13bo7bo2b3o11b3o
18bo$5bo6bob2o18bo7bo13bo2bo9bo17bo19bo2bo$5bo6bo21bo7bo11bo2bobo4bo4b
o17bo3b2o2b2ob2o2b2o3bob2o$2b2o2bo4bo5b2o15bo7bo11bo4bo3bobo3bo17bo3b
2o2b2ob2o2b2o3bo$bobo3bo2bo6bo13b2o2bo5bo2b3o7bo2bobo4bo4bo13b2o2bo19b
o$bo6b2o8b3o9bobo3bo3bo4b3o8bo2bo9bo13b2o2bo3b2o9b2o3bo$2o18bo9bo6b3o
20bo7bo2b2o14bo3b2o9b2o3bo$6b2o2b2o17b2o30bo5bo3bobo13bo19bo$7bo2bo25b
o3bo21b5o6bo13bo3b2o2b2ob2o2b2o3bo$4b3o4b3o21bo5bo31b2o9b2obo3b2o2b2ob
2o2b2o3bob2o$4bo8bo21b7o21bo3b2o15bo2bo19bo2bo$62b2o3bo20bo17bo$37b3o
28b3o18bo15bo$37bo2bo29bo19b15o$39b2o$91bo11bo$90b2o3b2ob2o3b2o$95b2ob
2o!

Re: Thread for your unsure discoveries

Posted: April 10th, 2021, 2:02 pm
by EvinZL
I found a new? periodic reflector mechanism:

Code: Select all

x = 27, y = 17, rule = B3/S23
4bo$2o2b2ob3o$2o4b4o$4b2o3$9b2o4b3o3b3o$10b2o$9bo3b2o3bobo3b2o$13bo4bo
bo4bo$12bo5bobo5bo$13bo4bobo4bo$14b3o5b3o2$17b2ob2o$16bobobobo$17bo3bo
!
Sparks made with blocks also work, but they need some help:

Code: Select all

x = 31, y = 27, rule = B3/S23
4bo$2o2b2ob3o$2o4b4o8b2o$4b2o8b2o2bo$14bobobo$17bo$9b2o$10b2o2b2o3b2o$
9bo3bo2bo2b2o$13bo10b6o$13bo2bo6bo2bo3bo$13b2ob2o6bo5bo$13bo2bo3b2o3b
3obo$13bobo4b2o4bo$14bo$29b2o$16b2o6b2o3b2o$11b2o3b2o6b2o$11b2o$27bo$
15bo4b2o4bobo$12bob3o3b2o3bo2bo$11bo5bo6b2ob2o$11bo3bo2bo6bo2bo$12b6o
10bo$21b2o2bo2bo$21b2o3b2o!
QBS, Jason's p22, and 56P27 also work, but don't have supporting sparkers (I haven't checked jslife yet)

EDIt: How did I forget about unix

EDIT2: 56P27 works for 56P27 and it makes a nice glider shuttle

Code: Select all

x = 74, y = 95, rule = B3/S23
26b2o$26b2o5$24b2o$25b2o$21bo4bo$21bo3bo$21b2o7b3o$30bo2$11b2o21bo$11b
2o5b2o11bob2o$17b2obo11b2o5b2o$17bo21b2o2$21bo$19b3o7b2o$26bo3bo$25bo
4bo$25b2o$26b2o3$11b2o$12bo2b2o$12bobobo7b2o$13bo10b2o2$15b2o$15b2o2$
59b2o$59b2o2$13b3o$13bobo$15bo$9b2o9bo36b3o$10b2o7b2o36bobo$19bo39bo$
53b2o9bo$2o20b2o30b2o7b2o$2o4b3o14bo39bo$6bo14b3o4b2o$6b2o20b2o14b2o
20b2o$44b2o4b3o14bo$10bo39bo14b3o4b2o$9b2o7b2o30b2o20b2o$9bo9b2o$14bo
39bo$14bobo25bobo8b2o7b2o$14b3o26b2o8bo9b2o$43bo14bo$58bobo$58b3o2$13b
2o$13b2o2$57b2o$57b2o2$48b2o10bo$48b2o7bobobo$57b2o2bo$61b2o3$46b2o$
47b2o$43bo4bo$43bo3bo$43b2o7b3o$52bo2$33b2o21bo$33b2o5b2o11bob2o$39b2o
bo11b2o5b2o$39bo21b2o2$43bo$41b3o7b2o$48bo3bo$47bo4bo$47b2o$48b2o5$46b
2o$46b2o!
EDIT3: Shoutouts to Rich's p16 for being able to reflect gliders in two different ways (no, the glider does not bonk into rich's p16 before getting into that position)

Code: Select all

x = 22, y = 22, rule = B3/S23
4bo$2o2b2ob3o$2o4b4o$4b2o3$9b2o$10b2o3bo$9bo3b2obo$13bo3bo$11bo5bo$11b
o5bo2bo$11bo7bobo$13b4o2b2o2$13b4o2b2o$11bo7bobo$11bo5bo2bo$11bo5bo$
13bo3bo$13b2obo$15bo!

Re: Oscillator Discussion Thread

Posted: April 10th, 2021, 3:08 pm
by hotdogPi
P19 or P38 dot sparker needed...

Code: Select all

x = 18, y = 20, rule = B3/S23
12b2o$12b2o4$10bo$10b2o$6b2o2b4o$3b5ob2obob2o$2bo2b2o3b3o2bobo$obo2b3o
3b2o2bo$2b2obob2ob5o$4b4o2b2o$6b2o$7bo4$4b2o$4b2o!

Re: Oscillator Discussion Thread

Posted: April 10th, 2021, 4:03 pm
by Sokwe
hotdogPi wrote:
April 10th, 2021, 3:08 pm
P19 or P38 dot sparker needed...
Interesting.

Here's a suggestion that I've been making for years: start with two honey farm + eater -> glider + junk reactions with 180-degree symmetry. For example:

Code: Select all

x = 27, y = 16, rule = B3/S23
25b2o$25bo$17b3o3bobo$16bo3bo2b2o$15bo5bo$15bo5bo$15bo5bo$7b3o6bo3bo$
6bo3bo6b3o$5bo5bo$5bo5bo$5bo5bo$2b2o2bo3bo$bobo3b3o$bo$2o!
Now add sparks and catalysts. Maybe you can get the honey farms to reappear in their original locations, giving a gun.

Edit:
EvinZL wrote:
April 10th, 2021, 2:02 pm
I found a new? periodic reflector mechanism:
Moved to the oscillator discussion thread.

Re: Oscillator Discussion Thread

Posted: April 11th, 2021, 8:58 am
by hotdogPi
p76 dual pi hassler (P4 can probably be improved):

Code: Select all

x = 53, y = 38, rule = B3/S23
9b2ob2o$9b2obo$5b2o6bo$2b2o2b7obo$bob2o7bobo$o4b3o3b2ob2o$ob2o4b2ob2o
3bo$b2o3bo6bobo$3b2obob2o3b3o8bobo$3b2obob2o3b3o8bobo$b2o3bo6bobo8b3o$
ob2o4b2ob2o3bo$o4b3o3b2ob2o$bob2o7bobo$2b2o2b7obo$5b2o6bo$9b2obo$9b2ob
2o3$39b2ob2o$40bob2o$39bo6b2o$38bob7o2b2o$38bobo7b2obo$37b2ob2o3b3o4bo
$36bo3b2ob2o4b2obo$26b3o8bobo6bo3b2o$26bobo8b3o3b2obob2o$26bobo8b3o3b
2obob2o$37bobo6bo3b2o$36bo3b2ob2o4b2obo$37b2ob2o3b3o4bo$38bobo7b2obo$
38bob7o2b2o$39bo6b2o$40bob2o$39b2ob2o!
Interestingly, the pi heptominoes can be offset by 0, 1, or 2 cells vertically, and it will still work.

Re: Oscillator Discussion Thread

Posted: April 11th, 2021, 9:20 am
by wwei47
hotdogPi wrote:
April 11th, 2021, 8:58 am
p76 dual pi hassler (P4 can probably be improved):

Code: Select all

x = 29, y = 36, rule = B3/S23
bo3bo$obobobo$obobobo$bobobo$3bo$2b2o$b3o$b3o8bobo$b3o8bobo$b3o8b3o$2b
2o$3bo$bobobo$obobobo$obobobo$bo3bo5$23bo3bo$22bobobobo$22bobobobo$23b
obobo$25bo$25b2o$14b3o8b3o$14bobo8b3o$14bobo8b3o$25b3o$25b2o$25bo$23b
obobo$22bobobobo$22bobobobo$23bo3bo!

Re: Oscillator Discussion Thread

Posted: April 11th, 2021, 9:50 am
by JP21
A very weird monomerization:

Code: Select all

x = 20, y = 33, rule = B3/S23
3b2o$3bobo$2obobo$obobo$2bo$bo2bo10bo$bo10bo$bo3bo5bo3bo$bo3bo5b3o$bo
10bo$bo2bo$2bo$obobo$2obobo$3bobo$3b2o3$12bo$12bo$12bo2$7b2o7b2o$7b11o
$6bo4b3o4bo$7bobobobobobo$9bo5bo$5bob3obobob3obo$5bo4bobobo4bo$6b3obob
obob3o$8bob2obo2bo$13bobo$13b2o!

Re: Oscillator Discussion Thread

Posted: April 11th, 2021, 9:57 am
by praosylen
JP21 wrote:
April 11th, 2021, 9:50 am
A very weird monomerization:

Code: Select all

x = 20, y = 33, rule = B3/S23
3b2o$3bobo$2obobo$obobo$2bo$bo2bo10bo$bo10bo$bo3bo5bo3bo$bo3bo5b3o$bo
10bo$bo2bo$2bo$obobo$2obobo$3bobo$3b2o3$12bo$12bo$12bo2$7b2o7b2o$7b11o
$6bo4b3o4bo$7bobobobobobo$9bo5bo$5bob3obobob3obo$5bo4bobobo4bo$6b3obob
obob3o$8bob2obo2bo$13bobo$13b2o!
Nice! Very slight reduction:

Code: Select all

x = 24, y = 31, rule = B3/S23
3b2o$3bobo$2obobo$obobo$2bo$2b2o10bo$2b3o8bob2o$2b3o7b3o$2b3o$2b3o$2b
2o$2bo$obobo$2obobo$3bobo$3b2o4$11b3o2$5bo3bo5bo3bo$4bobo11bobo$4bobo
11bobo$b2obobob9obobob2o$b2obo15bob2o$4bo15bo$4bobo4bobo4bobo$5b2o4b2o
bo3b2o$14bo$14b2o!
(Many congrats to hotdogPi for the initial oscillator!)

Re: Oscillator Discussion Thread

Posted: April 11th, 2021, 10:07 am
by hotdogPi
It turns out that the standard heavyweight emulator works. When I tried it the first time, it didn't work (probably) because it was flipped, but I didn't think to simply flip the emulator.

Code: Select all

x = 27, y = 36, rule = B3/S23
3b2o$3bobo$2obobo$obobo$2bo$b2o$3o$3o8bobo$3o8bobo$3o8b3o$b2o$2bo$obob
o$2obobo$3bobo$3b2o5$22b2o$21bobo$21bobob2o$22bobobo$24bo$24b2o$13b3o
8b3o$13bobo8b3o$13bobo8b3o$24b3o$24b2o$24bo$22bobobo$21bobob2o$21bobo$
22b2o!
About the honey farm + eater reaction mentioned by Sokwe: I tried several configurations, and it returned nothing. However, I can't get all of them. It only checks the leftmost boundary and puts sparks there and in the corresponding position on the right. I used a trick to delete the glider early for one of the runs, but if the eater is farther left than any of the active region, it won't put the sparks in the correct place.

EDIT: Flipping the emulator doesn't cause it to fail. I don't know why it didn't work when I tried it the first time.

Re: Oscillator Discussion Thread

Posted: April 11th, 2021, 10:52 am
by Kazyan
Awesome! With any luck, a small p76 gun will come out of this.

Re: Oscillator Discussion Thread

Posted: April 11th, 2021, 4:25 pm
by MathAndCode
Has this bun hassling mechanism been made into a p14 oscillator?

Code: Select all

x = 45, y = 16, rule = B3/S23
5b2o24b2o$4bo2bo22bo2bo$5b3o22b3o2$2b2o5b2o16b2o5b2o$bobo4bo2bo2b2o10b
obo4bo2bo2b2o$bo6bobo4bo2bo7bo6bobo4bo2bo$2o7bo5bobobo5b2o7bo5bobobo$
12b2obo2bo18b2obo2bo$12bo2bo21bo2bo$9bo4bo19bo4bo$9b5o20b5o2$11bo24bo
$10bobo22bobo$11bo24bo!
The difficulty stems from the tendency for one of the weak sparks at generation six to hit a catalyst on the other side.

Re: Oscillator Discussion Thread

Posted: April 11th, 2021, 8:50 pm
by Sokwe
JP21 wrote:
April 11th, 2021, 9:50 am
A very weird monomerization:

Code: Select all

x = 20, y = 33, rule = B3/S23
3b2o$3bobo$2obobo$obobo$2bo$bo2bo10bo$bo10bo$bo3bo5bo3bo$bo3bo5b3o$bo
10bo$bo2bo$2bo$obobo$2obobo$3bobo$3b2o3$12bo$12bo$12bo2$7b2o7b2o$7b11o
$6bo4b3o4bo$7bobobobobobo$9bo5bo$5bob3obobob3obo$5bo4bobobo4bo$6b3obob
obob3o$8bob2obo2bo$13bobo$13b2o!
Clever! Here is a smaller form that makes some of the sparks a bit more accessible (Edit: replaced with slightly smaller form):

Code: Select all

x = 16, y = 30, rule = B3/S23
b2o$obo$obob2o$bobobo$3bo$2b2o$b3o$b3o8bobo$b3o8bobo$b3o8b3o$2b2o$3bo$
bobobo$obob2o$obo$b2o$5b2o$6bo$5bo$4bob3o3b3o$4bobo2bo$3b2o2b2o$4bobo
4b5o$4bob2obo4b2o$5bo3bo5bo$6b2obob2o$7bobobo$7bobob2o$6b2obo2bo$10b2o
!
This might be useful for finding a gun.

Edit: here's a form that's shorter, but has worse clearance:

Code: Select all

x = 23, y = 28, rule = B3/S23
b2o$obo$obob2o$bobobo$3bo$2b2o$b3o$b3o8bobo$b3o8bobo$b3o8b3o$2b2o$3bo$
bobobo$obob2o$obo$b2o3$5bo$5b3o4b3o$3b2o3bo$2bo3bobo9b2o$3b5o3b5o3bo$
8b3o6bobob2o$3b2ob2obo7bobo2bo$4bobo3bo3b2obob2o$4bobo10bo$5bo10b2o!
Here's a form with a minimum population of 94 (the dimer is still smaller at 82 cells):

Code: Select all

x = 23, y = 28, rule = B3/S23
b2o$obo$obob2o$bobobo$3bo$2b2o$b3o$b3o8bobo$b3o8bobo$b3o8b3o$2b2o$3bo$
bobobo$obob2o$obo$b2o4$7b2o3b3o$7bo$4b2obo10b2o$2bo2bob2o2b5o3bo$2b2ob
o3b2o6bobob2o$5bo11bobo2bo$5b2o7b2obob2o$17bo$10b3o3b2o!
hotdogPi wrote:
April 11th, 2021, 10:07 am
About the honey farm + eater reaction mentioned by Sokwe: I tried several configurations, and it returned nothing. However, I can't get all of them. It only checks the leftmost boundary and puts sparks there and in the corresponding position on the right. I used a trick to delete the glider early for one of the runs, but if the eater is farther left than any of the active region, it won't put the sparks in the correct place.
Thanks for trying. I still consider this a very promising idea, if anyone wants to write a script without this spark placement restriction.
MathAndCode wrote:
April 11th, 2021, 4:25 pm
Has this bun hassling mechanism been made into a p14 oscillator?

Code: Select all

x = 45, y = 16, rule = B3/S23
5b2o24b2o$4bo2bo22bo2bo$5b3o22b3o2$2b2o5b2o16b2o5b2o$bobo4bo2bo2b2o10b
obo4bo2bo2b2o$bo6bobo4bo2bo7bo6bobo4bo2bo$2o7bo5bobobo5b2o7bo5bobobo$
12b2obo2bo18b2obo2bo$12bo2bo21bo2bo$9bo4bo19bo4bo$9b5o20b5o2$11bo24bo
$10bobo22bobo$11bo24bo!
The difficulty stems from the tendency for one of the weak sparks at generation six to hit a catalyst on the other side.
It's been looked at before, and can be extended to a (known) wick:

Code: Select all

x = 13, y = 53, rule = B3/S23
6b2o$5bo2bo$5b3o$2o9b2o$o2bo5bo2bo$b3o5b3o2$b3o5b3o$o2bo5bo2bo$2o9b2o$
5b3o$5bo2bo$6b2o$2o9b2o$o2bo5bo2bo$b3o5b3o2$b3o5b3o$o2bo5bo2bo$2o9b2o$
6b2o$5bo2bo$5b3o$2o9b2o$o2bo5bo2bo$b3o5b3o2$b3o5b3o$o2bo5bo2bo$2o9b2o$
5b3o$5bo2bo$6b2o$2o9b2o$o2bo5bo2bo$b3o5b3o2$b3o5b3o$o2bo5bo2bo$2o9b2o$
6b2o$5bo2bo$5b3o$2o9b2o$o2bo5bo2bo$b3o5b3o2$b3o5b3o$o2bo5bo2bo$2o9b2o$
5b3o$5bo2bo$6b2o!
As far as I know, no finite solution has been found.

Re: Oscillator Discussion Thread

Posted: April 12th, 2021, 8:49 am
by Naszvadi
I wonder if speed wires can be looped somehow... Eventually omniperiodicity would have been proven.

A true period 34 agar oscillator example:

Code: Select all

x = 204, y = 204, rule = B3/S23:T204,204
bo192b2obo$2o195bo2b4o$197bobo$4o192b2obo2b2o$4bo194bobo$5o194bobo$7bo
192b2obo$bob5o195bo$bo201bo$bo2b6o192b2o$bobo6bo$bobo2b5o$2b2obo7bo$5b
o2b6o$5bobo$4b2obo2b6o$7bobo6bo$7bobo2b5o$8b2obo7bo$11bo2b6o$11bobo$
10b2obo2b6o$13bobo6bo$13bobo2b5o$14b2obo7bo$17bo2b6o$17bobo$16b2obo2b
6o$19bobo6bo$19bobo2bob3o$20b2obobo5bo$23bo2b6o$23bobo$22b2obo2b6o$25b
obo6bo$25bobo2b5o$26b2obo7bo$29bo2b6o$29bobo$28b2obo2b6o$31bobo6bo$31b
obo2b5o$32b2obo7bo$35bo2b6o$35bobo$34b2obo2b6o$37bobo6bo$37bobo2b5o$
38b2obo7bo$41bo2b6o$41bobo$40b2obo2b6o$43bobobo4bo$43bobob6o$44b2obo7b
o$47bo2b6o$47bobo$46b2obo2b6o$49bobo6bo$49bobo2b5o$50b2obo7bo$53bo2b6o
$53bobo$52b2obo2b6o$55bobo6bo$55bobo2b5o$56b2obo7bo$59bo2b6o$59bobo$
58b2obo2b6o$61bobo6bo$61bobo2b5o$62b2obo7bo$65bo2b6o$65bobo$64b2obobob
5o$67bobo6bo$67bobo2b5o$68b2obo7bo$71bo2b6o$71bobo$70b2obo2b6o$73bobo
6bo$73bobo2b5o$74b2obo7bo$77bo2b6o$77bobo$76b2obo2b6o$79bobo6bo$79bobo
2b5o$80b2obo7bo$83bo2b6o$83bobo$82b2obo2b6o$85bobo6bo$85bobo2b5o$86b2o
bo7bo$89bo2bob4o$89bobobo$88b2obo2b6o$91bobo6bo$91bobo2b5o$92b2obo7bo$
95bo2b6o$95bobo$94b2obo2b6o$97bobo6bo$97bobo2b5o$98b2obo7bo$101bo2b6o$
101bobo$100b2obo2b6o$103bobo6bo$103bobo2b5o$104b2obo7bo$107bo2b6o$107b
obo$106b2obo2b6o$109bobo6bo$109bobo2b5o$110b2obobo5bo$113bob7o$113bobo
$112b2obo2b6o$115bobo6bo$115bobo2b5o$116b2obo7bo$119bo2b6o$119bobo$
118b2obo2b6o$121bobo6bo$121bobo2b5o$122b2obo7bo$125bo2b6o$125bobo$124b
2obo2b6o$127bobo6bo$127bobo2b5o$128b2obo7bo$131bo2b6o$131bobo$130b2obo
2b6o$133bobo6bo$133bobobob4o$134b2obo7bo$137bo2b6o$137bobo$136b2obo2b
6o$139bobo6bo$139bobo2b5o$140b2obo7bo$143bo2b6o$143bobo$142b2obo2b6o$
145bobo6bo$145bobo2b5o$146b2obo7bo$149bo2b6o$149bobo$148b2obo2b6o$151b
obo6bo$151bobo2b5o$152b2obo7bo$155bo2b6o$155bobo$154b2obo2bob4o$157bob
obo4bo$157bobo2b5o$158b2obo7bo$161bo2b6o$161bobo$160b2obo2b6o$163bobo
6bo$163bobo2b5o$164b2obo7bo$167bo2b6o$167bobo$166b2obo2b6o$169bobo6bo$
169bobo2b5o$170b2obo7bo$173bo2b6o$173bobo$172b2obo2b6o$175bobo6bo$175b
obo2b5o$176b2obo7bo$179bo2b6o$179bobobo$178b2obob7o$181bobo6bo$181bobo
2b5o$182b2obo7bo$185bo2b6o$185bobo$184b2obo2b6o$187bobo6bo$187bobo2b5o
$188b2obo7bo$191bo2b6o$191bobo$190b2obo2b6o$193bobo6bo$193bobo2b5o!

Re: Oscillator Discussion Thread

Posted: April 12th, 2021, 9:10 am
by hotdogPi
Does anyone know if the dual pi hassler above, at 82 cells, is the smallest known p76? (I would imagine it is; what else do we have besides Herschel tracks and glider loops, especially since 76/2=38 and 76/4=19 are unknown periods?)

Re: Oscillator Discussion Thread

Posted: April 12th, 2021, 11:01 am
by dvgrn
hotdogPi wrote:
April 12th, 2021, 9:10 am
Does anyone know if the dual pi hassler above, at 82 cells, is the smallest known p76? (I would imagine it is; what else do we have besides Herschel tracks and glider loops, especially since 76/2=38 and 76/4=19 are unknown periods?)
Yup, I just took a quick look at my copies of DRH-oscillators, jslife, and the jslife supplement, and there's literally no competition -- "literally" as in, period 76 is not represented in these collections.

It's easy enough to build a bumper-based glider loop using p4 bumpers, but because of the way the rephasing works at each 90-degree turn, you need eight bumpers to close a loop. So you end up with some variant of this:

Code: Select all

x = 89, y = 53, rule = B3/S23
15b2o24b2o$7bo7bo9bo6bo9bo7bo$7b3o2b2obo7b3o6b3o7bobo3b3o$10bo3bo7bo
12bo7b2o2bo$9bo12b2o10b2o12bo$10b3o33b2o$18bo20bo6bo$17bobo18bobo$11b
2o3bo2bo18bo2bo3b2o$11b2o4b2o20b2o4b2o2$b2o$2bo$2bobo59b2o$3b2o51bo7bo
9bo$8bo14b2o31b3o3bobo7b3o$7b2o13b2o21bo13bo2b2o7bo$6bo2bo14bo21b2o10b
obo10b2o$7b2o36b2o12b2o$11bobo45b2o6bo$o11b2o14b3o29b2o4bobo$3o9bo17bo
29b2o3bo2bo$3bo25bo30b2o4b2o$2b2o82b2o$8b2o76bo$5b2ob2o74bobo$3bobo77b
obo$2bobo74b2ob2o$2bo76b2o$b2o82b2o$21b2o4b2o30bo25bo$20bo2bo3b2o29bo
17bo9b3o$20bobo4b2o29b3o14b2o11bo$21bo6b2o45bobo$28b2o12b2o36b2o$16b2o
10bobo10b2o21bo14bo2bo$17bo7b2o2bo13bo21b2o13b2o$14b3o7bobo3b3o31b2o
14bo$14bo9bo7bo51b2o$23b2o59bobo$86bo$86b2o2$42b2o4b2o20b2o4b2o$42b2o
3bo2bo18bo2bo3b2o$48bobo18bobo$42bo6bo20bo$41b2o33b3o$40bo12b2o10b2o
12bo$41bo2b2o7bo12bo7bo3bo$38b3o3bobo7b3o6b3o7bob2o2b3o$38bo7bo9bo6bo
9bo7bo$46b2o24b2o!
No doubt there's a way to shorten the loop a little more, but it's not going to get anywhere near Sokwe's reduced monomerization at 16x30.

The interesting thing will be whether someone can come up with a p4 glider splitter, maybe a p4 variant of Jormungant's periodic splitter, that allows for a p76 gun smaller than the current 121x65 pseudo-period gun. Or a true-period p76 smaller than {whatever the smallest p76N Herschel loop is} would still be new territory, I suppose.

The top of Jormungant's Bandersnatch-based splitter is perfectly p4 compatible. I haven't tried a search yet for p4s that can suppress the reaction in the southeast, replacing the blinker here:

Code: Select all

x = 44, y = 41, rule = LifeHistory
18.2A$17.A.A23.B$B15.A2.A.2A.A17.2B$2B12.3A.A.A.A.3A14.3B$3B10.A4.A.B
.A4.A12.4B$4B9.5A.ABA.5A11.4B$.4B9.AB.A.A.A.A.BA11.4B$2.4B8.ABAB.B.B.
BABA10.4B$3.4B7.2B.B2AB2AB.2B9.4B$4.4B6.2A4BA4B2A8.4B$5.4B6.B2A5B2AB
8.4B$6.4B9.3B11.4B$7.4B9.A11.4B$8.4B8.B5.B4.4B$9.4B4.2B.3B2.9B$10.4B
2.7B.9B$11.4B.16B$12.20B$12.21B$12.20B$11.21B$10.23B$9.2A9BA13B$9.2A
10BA8B.4B$10.9B3A5B5.4B$11.16B6.4B$11.16B2D5.4B$12.8B2A7BD5.4B$11.8BA
2BA4B2D2B5.4B$10.10BABA5BD.3B4.4B$10.11BA5BD6B4.4B$11.15B3DBC2A6.4B$
6.2A3.17B.3B8.4B$6.A3.17B14.3B$8.A.2AB.14B14.2B$7.2AB2AB5.B2.B2.B2AB
15.B$10.B6.3B5.2A$7.2A.2A5.B2AB$8.A.A7.2A$8.A.A$9.A!
(No doubt there's something smaller than that p4 fountain, but I'll leave it to an oscillator expert to pick out the best option.)

Re: Oscillator Discussion Thread

Posted: April 12th, 2021, 11:44 am
by dvgrn
dvgrn wrote:
April 12th, 2021, 11:01 am
{whatever the smallest p76N Herschel loop is}...
As of 2012, I think the smallest known true-period p76 gun was this, adapted from a period-doubled version with a smaller bounding box from Jason Summers' guns1j collection:

Code: Select all

x = 164, y = 64, rule = B3/S23
32bo49bo49bo$30b3o20b2o6b2o17b3o20b2o6b2o17b3o11bo$4b2o23bo23bo2bo4b2o
16bo23bo2bo4b2o16bo14b3o$5bo23b2o19b2obob2o22b2o19b2obob2o22b2o16bo14b
o$5bobo41bobobo45bobobo42b2o12b3o$6b2o2b2o37bo3bob2o42bo3bob2o52bo$10b
2o35b3o2bo2b2o40b3o2bo2b2o42b2o8b2o$46bo5b2o42bo5b2o38bobo$33b2o11b2o
35b2o11b2o46bo$33b2o48b2o73b2o$24bo46bobo47b3o16bo3b2o12b2o$21b3o20bo
26bo2bo18b2o23bo4bo16bobob2o$17b2obo22b2o22b2obo2bo18bo2bo21b2o3bo20bo
$17bo4bo19bo2b2o20bo2bo21bo2bo20b2o$18b3o21b2o4bo19b2o22b5o20bo2bo$46b
3o19b2o25b2ob2o18b2o$47b2o48bo$4b2o52b2o48b2o32b2o$3bobo16b2o34b2o11b
2o35b2o11b2o20bo$3bo18bobo46bo49bo19bo$2b2o20bo47b3o47b3o16b2o$8bo9b2o
4b2o48bo49bo20b2o$8b2o8bo2bo124bo$7bob2o9b2o32b2o48b2o37b3o$8bo3bobo
39bo49bo38bo$8b4o21b2o20b3o25b2o20b3o$8bob2o3bo18bo22bo26bo22bo$13b2o
16b3o47b3o$7bo2bo20bo49bo72b3o$8bo145bo2bo5bo$154bob2o3bobo$2o160b2o$b
o$bobo4b2obo$2b2o4bo2bo145bo$9b3o72bo49bo20bo2bo$82b3o47b3o16b2o$58bo
22bo26bo22bo18bo3b2obo$58b3o20b2o25b3o20b2o21b4o$22bo38bo49bo39bobo3bo
$20b3o37b2o48b2o32b2o9b2obo$19bo124bo2bo8b2o$19b2o20bo49bo48b2o4b2o9bo
$23b2o16b3o47b3o47bo20b2o$24bo19bo49bo46bobo18bo$22bo20b2o11b2o35b2o
11b2o34b2o16bobo$22b2o32b2o48b2o52b2o$68bo48b2o$46b2o18b2ob2o25b2o19b
3o$45bo2bo20b5o22b2o19bo4b2o21b3o$48b2o20bo2bo21bo2bo20b2o2bo19bo4bo$
22bo20bo3b2o21bo2bo18bo2bob2o22b2o22bob2o$20b2obobo16bo4bo23b2o18bo2bo
26bo20b3o$6b2o12b2o3bo16b3o47bobo46bo$6b2o73b2o48b2o$21bo46b2o11b2o35b
2o11b2o$21bobo38b2o5bo42b2o5bo$5b2o8b2o42b2o2bo2b3o40b2o2bo2b3o35b2o$
6bo52b2obo3bo42b2obo3bo37b2o2b2o$3b3o12b2o42bobobo45bobobo41bobo$3bo
14bo16b2o22b2obob2o19b2o22b2obob2o19b2o23bo$19b3o14bo16b2o4bo2bo23bo
16b2o4bo2bo23bo23b2o$21bo11b3o17b2o6b2o20b3o17b2o6b2o20b3o$33bo49bo49b
o!
Smallest by population would just use regular Fx77s instead of those big bounding-box-saving welds.

Not sure if anyone has invented any signal-splitting technology since then, that can improve on that. (?)

Re: Oscillator Discussion Thread

Posted: April 12th, 2021, 12:16 pm
by hotdogPi
p39 partial. Unfortunately, a p13 domino sparker (and I don't know of any) won't work, since the next spark 13 generations later will interact with it. However, if we can get something working, and a π → glider (which shouldn't be difficult), we have a p39 glider gun.

Code: Select all

x = 36, y = 36, rule = B3/S23
19bo2bo2bo$19b7o2$19b7o$19bo2bo2bo8$7bo$7bo5bo9b3o$9b2o2b3o10bo$b2o6bo
b5o6bobobo$b2o7b2o9b2ob4o$10bo3b3o4bob3o$11b3obo4b3o3bo$9b4ob2o9b2o7b
2o$10bobobo6b5obo6b2o$10bo10b3o2b2o$11b3o9bo5bo$29bo8$11bo2bo2bo$11b7o
2$11b7o$11bo2bo2bo!

Re: Oscillator Discussion Thread

Posted: April 12th, 2021, 4:26 pm
by Sokwe
hotdogPi wrote:
April 12th, 2021, 12:16 pm
p39 partial.
Another interesting partial result! Here are demonstrations of the p38 and p39 engines supported by 4 glider streams each:

Code: Select all

x = 206, y = 78, rule = B3/S23
175bobo$175b2o$96bo79bo$95bo$95b3o$66bo$64b2o$65b2o2$166bo$165bo$165b
3o$86bo$86bobo$86b2o115bo$55bobo145bobo$55b2o146b2o$56bo2$157bo$155b2o
$77bo78b2o$76bo89bo$76b3o86bobo$47bo117bobo25bobo$45b2o117b2ob3o23b2o$
46b2o122bo23bo$164b2ob3o$164b2obo$52b2o92bo$52b2o92bobo$67bo78b2o$67bo
bo67b2o$67b2o69b2o44bo$36bobo98bo45bo$36b2o13b3o119b2o8b3o$37bo13bo
121bo$51b2o16b2o77b2o12b2o7bobo$69b2o77bobo10b2o8b2o$27b2o111b2o8b2o
10bobo$27b2o16b2o92bobo7b2o12b2o$46bo13bo78bo$44b3o13b2o65b3o8b2o$59bo
bo67bo45bo$29b2o97bo44b2o$28bobo143b2o$30bo134b2o$44b2o118bobo$44b2o
120bo$145bob2o$143b3ob2o$50b2o66bo23bo$51b2o65b2o23b3ob2o$50bo66bobo
25bobo$19b3o123bobo$21bo124bo$20bo134b2o$156b2o$155bo2$41bo$41b2o65b2o
$40bobo64bobo$10b2o97bo$9bobo$11bo$145b3o$147bo$146bo2$31b2o$32b2o$31b
o$3o$2bo$bo134bo$136b2o$135bobo!

Re: Oscillator Discussion Thread

Posted: April 12th, 2021, 8:20 pm
by hotdogPi
I will not be running the program any more. Whenever I run the program (and don't end it immediately or near-immediately), Golly will not close normally, and when I either start a new file or have to quit, I have to force quit.

One time, I decided to keep it going on the same file for as long as possible. When I reached only 2.5 GB left of free space, my computer gave me a warning. (It's normally at 55 GB.) I closed it, and it went back to 10 GB. I opened Golly, made a change, and closed without saving, which got me to 12 GB. I decided to restart my computer, and it took from 3 PM EDT to now for it to start up. (Fortunately, since it was the second time it happened, combined with a way to see how much space was available by starting to run diagnostics, I knew what was happening. The first time, I had to send it in to repair for several days, since my X key refused to work entirely, and I thought the two were related when they weren't.)

Re: Oscillator Discussion Thread

Posted: April 12th, 2021, 9:17 pm
by dvgrn
hotdogPi wrote:
April 12th, 2021, 8:20 pm
I will not be running the program any more. Whenever I run the program (and don't end it immediately or near-immediately), Golly will not close normally, and when I either start a new file or have to quit, I have to force quit.

One time, I decided to keep it going on the same file for as long as possible...
What is "the program" exactly? I'm having trouble locating the reference, unless you mean just running Golly -- but in that case, I don't see what you mean by "keep it going on the same file".

Re: Oscillator Discussion Thread

Posted: April 12th, 2021, 9:23 pm
by hotdogPi
dvgrn wrote:
April 12th, 2021, 9:17 pm
hotdogPi wrote:
April 12th, 2021, 8:20 pm
I will not be running the program any more. Whenever I run the program (and don't end it immediately or near-immediately), Golly will not close normally, and when I either start a new file or have to quit, I have to force quit.

One time, I decided to keep it going on the same file for as long as possible...
What is "the program" exactly? I'm having trouble locating the reference, unless you mean just running Golly -- but in that case, I don't see what you mean by "keep it going on the same file".
It's the Python code I posted earlier, except it keeps getting modified depending on what I'm doing with it. Here is the latest version (note that my variable names r_up and r_down haven't changed despite not always being R-pentominoes, and the comments haven't changed to go along with changes, either; it's not ready for publishing). One thing I have corrected since my previous post containing the code is that the spark on the right automatically updates to the correct location based on initial pattern size.

Code: Select all

import golly as g
import time

g.select([0,0,1,1])
#g.autoupdate(1)
r_up = g.parse('b2o$o2bo$obo$2o!')
r_down = g.parse('2b2o$bobo$o2bo$b2o!')
g.clear(0)
g.clear(1)
g.putcells(r_down, 0, 0)
width = g.getrect()[2]
height = g.getrect()[3]
single_cell = g.parse('o!')
for x in range(-18,21):
    for y in range(-15,16):
        recent_pops = list(range(100,109))
        current_gen = None
        g.clear(0)
        g.clear(1)
        g.putcells(r_down, 0, 0)
        g.putcells(r_up, x, y)
        g.show(str((x,y)))
        #time.sleep(.2)
        g.update()
        initial = g.getcells(g.getrect())
        initial_box = g.getrect()
        initial_box[0] -= 1
        initial_box[1] -= 1
        initial_box[2] += 2
        initial_box[3] += 2
        if initial_box[2] <= 6:
            initial_box = [-50,-50,100,100]
        for generation in range(75):
            #g.update()
            #g.show('Generation ' + str(generation))
            if generation == 0:
                g.run(4)
                pass
            else:
                g.clear(0)
                g.clear(1)
                g.putcells(current_gen)
            g.run(1)
            if g.getcells(g.getrect()) == current_gen:
                break
            recent_pops.append(int(g.getpop()))
            recent_pops.pop(0)
            if recent_pops[-1] < 8: #2 blocks + 2 of any growth
                break
            if max(recent_pops) == min(recent_pops):
                break
            #if g.getcell(1,9) == 0: #block broken
            #    break
            min_x = g.getrect()[0]
            g.select([min_x,-10+y,1,40])
            g.copy()
            edge = g.getclip()[3::2]
            current_gen = g.getcells(g.getrect())[:]
            if not edge:
                continue
            #if max(edge) - len(edge) + 1 == min(edge):
            #    continue
            while max(edge) - len(edge) + 1 != min(edge):
                edge.remove(max(edge))
            #if generation == 0:
            #    g.putcells(single_cell,min_x,min(edge)-21)
            #    g.putcells(single_cell,-min_x+x+5,-min(edge)+y+29)
            #    #g.update()
            #    #time.sleep(.5)
            #    current_gen = g.getcells(g.getrect())[:]
            #    continue
            if len(edge) == 1:
                sparks = [(-2,-1),(-1,0),(-1,1),(0,1),(1,2)]
            if len(edge) == 2:
                sparks = [(-1,),(0,),(2,),(-2,-1),(-1,0),(1,2),(2,3)]
            if len(edge) == 3:
                sparks = [(-1,),(0,),(1,),(2,),(3,),(-1,0),(0,1),(2,3),(3,4)]
            if len(edge) >= 4:
                sparks = [(-1,),(0,),(1,),(2,),(3,),(4,)] #not all yet
            for spark in sparks:
                g.clear(0)
                g.clear(1)
                g.putcells(current_gen)
                for spark1 in spark:
                    g.putcells(single_cell,min_x-2,spark1+y+min(edge)-10)
                    g.putcells(single_cell,-min_x+x+1+width,-spark1-min(edge)+9+height)
                    x1, y1 = min_x-5,spark1+y+min(edge)-10
                #g.update()
                for generation1 in range(200):
                    g.run(1)
                    if generation1 == 40:
                        gen40 = g.getcells(g.getrect())
                    if generation1 == 42 and g.getcells(g.getrect()) == gen40:
                        break
                    if generation1 == 120:
                        gen120 = g.getcells(g.getrect())
                    if generation1 == 122 and g.getcells(g.getrect()) == gen120:
                        break
                    #if g.getcell(1,9) == 0: #block broken
                    #    break
                    if g.getcell(x1, y1) == 1: #hits sparker
                        break
                    if int(g.getpop()) < 8: #2 blocks + 2 R predecessors
                        break
                    if g.getcells(initial_box) == initial:
                        if (x, y) == (8,5):
                            g.warn('Already found: %s, %s, %s' % (x, y, generation))
                            pass
                        else:
                            g.warn('Found!: %s, %s, %s' % (x, y, generation))
                            g.update()
                            g.exit('Found!')
                        
If you run this, to avoid issues, uncomment the last g.update() and you should be fine without requiring a force quit, as it no longer tries to run it as fast as possible.

"Same file" does refer to Golly; I'm running it on untitled, and leaving untitled or quitting entirely takes a really long time to where I have to force quit, probably because Golly allows unlimited undos and it has to store them all.

Re: Oscillator Discussion Thread

Posted: April 12th, 2021, 9:28 pm
by dvgrn
hotdogPi wrote:
April 12th, 2021, 9:23 pm
It's the Python code I posted earlier...

...

g.clear(0)
g.clear(1)
I think you'll find that the only problem here is that you don't call g.new("") anywhere in your script.

If you do that, then Golly won't try to record and remember in its Undo buffer every single individual step that your script takes. As it is, with your clear(0) and clear(1) syntax avoiding a g.new() call but having a similar effect, that's basically what Golly is doing.

You can see why remembering every individual script action could cause a lot of trouble with big temporary file sizes, on a long run of a script.

Re: Oscillator Discussion Thread

Posted: April 12th, 2021, 9:32 pm
by wwei47
hotdogPi wrote:
April 12th, 2021, 9:23 pm
It's the Python code I posted earlier, except it keeps getting modified depending on what I'm doing with it. Here is the latest version (note that my variable names r_up and r_down haven't changed despite not always being R-pentominoes, and the comments haven't changed to go along with changes, either; it's not ready for publishing). One thing I have corrected since my previous post containing the code is that the spark on the right automatically updates to the correct location based on initial pattern size.
Try putting g.new("") at the beginning. Golly is trying to keep the entire undo log since it thinks you're modifying a pattern, and that's a lot of files and running g.new("") tells it that it's creating some new kind of pattern, and that should stop it from logging every change. I don't know if that'll work, but it's worth a try.
At least one new post has been made to this topic. You may wish to review your post in light of this.
I guess I can still mention how it's a lot faster when it's not trying to record every step.
EDIT: dr says this P3 is unknown, and it also happens to work in Life, too:

Code: Select all

x = 15, y = 17, rule = B2n3/S23-q
11b2o$10bobo$10bo$9b2o2bo$7bo3b3o$7b4o$4bo5bob2o$4b4o2b2obo$8bobobo$4b
2o2bobobobo$3bo2bobobo2b2o$3bob2obobo$2obo5bo$bo2b5o$bo$2b3o2b2o$4bo2b
2o!

Code: Select all


*****  Period 3 at gen 6
p3 r9 4x5 ...2A .C..1 C00A. .C...	<- UNKNOWN
u9 r20 8x8 ......3. .....2@3 ....1A@. ....0.1. ..1A@A.. A.1..... A@A..... .1C.....	<- unknown
Change counts: 1 1 3 6 4  4 4 6 9 4
Sizes: 1x1 1x1 2x3 3x4 4x2  3x2 2x3 3x3 4x5 2x3
Gen 0.  Rows 35 - 49.  Cols 30 - 45.
,
,
,,,,,,,.oo..
,,,,,,o...oo
,,,,,,oooo...
,,,,,....o.oo.
,,,,ooo..o1.o.
,......o.o.o..
,..oo..o.o.o.o
,.o..o.o.o.
,.o.oo.o.o
,.o.....o
,..oooo
,......
,