ConwayLife.com - A community for Conway's Game of Life and related cellular automata
Home  •  LifeWiki  •  Forums  •  Download Golly

Orthogonoid spaceship -- completed!

For discussion of specific patterns or specific families of patterns, both newly-discovered and well-known.

Re: Orthogonoid spaceship -- completed!

Postby chris_c » June 7th, 2017, 11:16 am

On the other thread calcyman wrote:Can anyone find a way to slow-salvo construct the following constellation (in such a way that it doesn't interfere with the eater)?

x = 12, y = 7, rule = B3/S23
7bo$6bobo$6bobo$5b2ob3o$11bo$2o3b2ob3o$2o3b2obo!



The block can be made in two gliders from a beehive but I don't know if it will be easy to teach slmake about this kind of thing:

x = 39, y = 40, rule = B3/S23
o14bo$3o11bobo$3bo10bobo$2b2o9b2ob3o$19bo$13b2ob3o$13b2obo3$bo$obo$obo
$bo2$5b3o$5bo$6bo21$36b3o$36bo$37bo!


Anyway, by flipping the G->MWSS vertically I came up with a cheaper 1-Snark Orthogonoid that has 367 cells and is HashLife friendly. Among similar 1-Snark Orhogonoids it should be difficult to beat but maybe the 2-Snark version is still better:

x = 346, y = 297, rule = B3/S23
97b2o$97bobo$99bo4b2o$95b4ob2o2bo2bo$95bo2bobobobob2o$98bobobobo$99b2o
bobo$103bo2$89b2o$90bo7b2o$90bobo5b2o$91b2o2$74bo$72b3o$71bo$61b2o8b2o
47bo$62bo55b3o$62bobo36b2o14bo$63b2o36bo15b2o$77b2o23b3o$49b2o26b2o25b
o$50bo$19bo30bobo$18bobo7bo22b2o$19bo6b3o48b2o35b2o$25bo51b2o34bo2bo$
25b2o77bo9b2o$103bobo$103b2o2$53b2o$53b2o$5b2o25b2o$6bo25b2o$6bobo$7b
2o3$20b2o$20bobo6b2o$22bo6bo$22b2o6b3o$32bo$114b2o$59b2o53b2o$60bo$60b
obo$18b2o41b2o31bo$18bobo73b3o$20bo76bo$20b2o74b2o$61b2o$61b2o3$2b2o$b
obo$bo84b2o$2o84b2o$66b2o$60b2o3bo2bo$60b2o4b2o$123b2o$123b2o$83b2o32b
2o$10b2o71bo33b2o$10b2o7b2o63b3o$19bo41bo24bo$17bobo39b3o57b2o$17b2o4b
2o33bo53b2o5b2o$b2o20bo15bo18b2o7b2o3b2o38b2o$2bo18bobo15b3o25b2o3b2o$
2bobo16b2o19bo$3b2o36b2o$79b2o$79bo$77bobo$77b2o2$38b2o$38b2o4$9b2o63b
2o$5b2o2b2o63bo$4bobo38b2o28b3o$4bo40bo31bo$3b2o41b3o$48bo5$61b2o$61b
2o$69b2o$69bo$70b3o$72bo2$71bo$70bobo$70bobo$69b2ob3o$75bo$69b2ob3o$
69b2obo2$51b2o8b2o$51bobo7b2o$53bo$53bobo$54b2o4$74b2o$74b2o5$59bo$58b
obo$58bobo$59bo187b2o$56b3o187bobo$56bo183b2o4bo$238bo2bo2b2ob4o$238b
2obobobobo2bo$241bobobobo$241bobob2o$242bo2$255b2o$246b2o7bo$246b2o5bo
bo$253b2o2$271bo$271b3o$274bo$225bo47b2o8b2o$225b3o55bo$228bo14b2o36bo
bo$227b2o15bo36b2o$241b3o23b2o$241bo25b2o26b2o$295bo$293bobo30bo$293b
2o22bo7bobo$230b2o35b2o48b3o6bo$229bo2bo34b2o51bo$230b2o9bo77b2o$240bo
bo$241b2o2$12b2o277b2o$12b2o277b2o$312b2o25b2o$312b2o25bo$15b2o320bobo
$14bobo320b2o$12b3obobo$11bo5b2o$11b2o311b2o$315b2o6bobo$316bo6bo$313b
3o6b2o$313bo$230b2o$230b2o53b2o$285bo$283bobo$251bo31b2o41b2o$249b3o
73bobo$248bo76bo$248b2o74b2o$283b2o$283b2o3$342b2o$342bobo$258b2o84bo$
258b2o84b2o$278b2o$277bo2bo3b2o$278b2o4b2o$221b2o$221b2o$227b2o32b2o$
227b2o33bo71b2o$259b3o63b2o7b2o$259bo24bo41bo$225b2o57b3o39bobo$225b2o
5b2o53bo33b2o4b2o$232b2o38b2o3b2o7b2o18bo15bo20b2o$272b2o3b2o25b3o15bo
bo18bo$303bo19b2o16bobo$303b2o36b2o$265b2o$266bo$266bobo$267b2o2$306b
2o$306b2o4$270b2o63b2o$271bo63b2o2b2o$268b3o28b2o38bobo$268bo31bo40bo$
297b3o41b2o$297bo5$283b2o$283b2o$275b2o$276bo$273b3o$273bo2$274bo$273b
obo$273bobo$271b3ob2o$270bo$271b3ob2o$273bob2o2$283b2o8b2o$283b2o7bobo
$292bo$290bobo$290b2o4$270b2o$270b2o5$286bo$285bobo$285bobo$286bo$287b
3o$289bo28$315bo$316bo$311bo4bo15b2o$312b5o15b2o3$329b2o$329bobo$327bo
bob3o$327b2o5bo$333b2o!
chris_c
 
Posts: 743
Joined: June 28th, 2014, 7:15 am

Re: Orthogonoid working notes

Postby chris_c » June 7th, 2017, 11:28 am

dvgrn wrote:In the absence/presence of the key piece, it would all self-destruct without doing anything. The left-side construction arm ends up doing a complicated NOP operation, and the right-side arm gets the minor adjustment it needs.


I am thinking that the left-side construction arm builds:

1. A far away 180 degree reflector on the construction lane.
2. A 0 degree glider aiming at the the 180 degree reflector.
3. (Soon after 2) An eater on the construction lane.

Eventually the glider returns and destroys the eater. This gives a certain period of time where the gliders on the construction lane will be absorbed without effect.

On the right construction arm the presence of a key piece of junk prevents 3 from happening. In the meantime gliders that encode the adjustments to the hand and elbow as well as the building of the key piece of junk in the child pattern are sent.

Might this work? Any better ideas?

EDIT: Rather, the key piece of junk needs to prevent 3 from happening and make a usable mess near the construction lane instead.
chris_c
 
Posts: 743
Joined: June 28th, 2014, 7:15 am

Re: Orthogonoid working notes

Postby dvgrn » June 7th, 2017, 12:08 pm

chris_c wrote:I am thinking that the left-side construction arm builds:

1. A far away 180 degree reflector on the construction lane.
2. A 0 degree glider aiming at the the 180 degree reflector.
3. (Soon after 2) An eater on the construction lane.

Eventually the glider returns and destroys the eater. This gives a certain period of time where the gliders on the construction lane will be absorbed without effect.

On the right construction arm the presence of a key piece of junk prevents 3 from happening. In the meantime gliders that encode the adjustments to the hand and elbow as well as the building of the key piece of junk in the child pattern are sent.

Might this work?

Tricky! Yes, seems like that will work. The faraway one-time reflector can be as simple as a couple of blocks or a long boat. It may need to be pretty far away, though, so it might be necessary to use Calcyman's Cordership build/launch/shoot-down trick.

-- Come to think of it, is the range of 0-degree gliders wide enough now that the Cordership seed could be built directly on the construction arm? EDIT: Not quite -- somewhere around 136 lanes would be needed, and we "only" have 119. Of course we can build any missing ones if we want to, with an elbow-to-hand then converting the hand to a one-time turner.

Anyway, I guess that's not necessary -- the Cordership could be pointed diagonally backwards just as well. For some reason I was visualizing it as launching in the direction the construction arm is pointing.

chris_c wrote:EDIT: Rather, the key piece of junk needs to prevent 3 from happening and make a usable mess near the construction lane instead.

Also, when the reflected glider comes back, it can't just delete the eater, it has to leave some junk. Or if it does delete the eater, there has to be something behind it that can get turned into an elbow... and that absorbs gliders exactly the same as whatever is left behind after the hand&elbow-adjustment/key-junk-building recipe. So probably simplest if it's just a standard elbow.
dvgrn
Moderator
 
Posts: 3996
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI

Re: Orthogonoid working notes

Postby dvgrn » June 16th, 2017, 10:04 am

chris_c wrote:Anyway, by flipping the G->MWSS vertically I came up with a cheaper 1-Snark Orthogonoid that has 367 cells and is HashLife friendly. Among similar 1-Snark Orhogonoids it should be difficult to beat but maybe the 2-Snark version is still better...

No, I think I like this one the best out of the whole collection so far. I think it's time to get an Orthogonoid actually running. I'll probably compile Orthogonoid367 and Orthogonoid372 with slmake next -- see if the integral is really cheaper than two eaters.

Then --

What's the smallest number of slow SW gliders that can clean up an old Orthogonoid construction arm safely? At 128 spacing this design is just a little bit tight along the NW edge (see below). But it looks as if it will be easy to find cleanups. For example, here are two slow gliders that leave only one blinker just out of reach, to be cleaned up at some point, presumably by a stray NW glider or just a lucky spark:

x = 212, y = 318, rule = LifeHistory
209.A$209.A.A$209.2A33$118.4B$119.4B$120.4B$121.4B$122.4B$123.4B$124.
4B$125.4B$126.4B$127.4B$128.4B$129.4B$130.4B$131.4B$132.4B$124.B.B.B.
B.5B21.B$122.17B18.2B$122.18B.B14.4B$120.21B2A12.4B$121.20B2A11.4B$
121.18B2.B11.4B$123.B.B.B.B.B.B.B2.2B12.4B$137.B2A11.4B$138.A.A9.4B$
136.A.A.3A6.4B$136.2A5.A4.4B$142.2A3.4B$146.4B$145.4B$144.4B$143.4B$
142.4B$141.4B$140.4B$139.4B$138.4B$137.4B$136.4B$135.4B$134.4B$133.4B
$132.4B2.A$131.4B3.A.A$130.4B4.2A$129.4B$128.4B$127.4B$126.4B$125.4B$
124.4B65$56.2A$55.A.A$49.2A4.A$47.A2.A2.2A.4A$47.2A.A.A.A.A2.A$50.A.A
BABAB$50.A.AB2AB$51.AB.2B$54.3B$54.4B6.2A$52.3B2AB6.A$52.3B2AB3.BA.A$
50.10B.B2A$49.13B$48.14B18.A$47.15B18.3A$46.4B2.8B23.A$34.A10.4B5.6B
22.2A8.2A61.4B$34.3A7.4B4.9B21.5B5.A61.4B$37.A5.4B5.2A4.4B22.4B.BA.A
60.4B$36.2A4.4B7.A5.4B14.B4.6B.B2A60.4B$36.9B5.3A7.4B12.2AB.10B61.4B$
38.6B6.A10.4B11.2A12B14.2A44.4B$37.6B19.4B11.B.11B14.A44.4B$37.6B20.
4B12.13B3.4B2.BA.A30.A12.4B$38.6B20.4B9.2B.12B2.5B2.B2A22.A7.A.A10.4B
$37.2B2A4B3.3B14.4B7.2A24B24.3A4.2BAB9.4B$36.2BA2BA3B.6B14.4B6.2A24B
27.A4.2B9.4B$35.4B2A9BA2B14.4B6.B.B.20B27.2A5.4B5.4B$.B.B.B.B.B.B.B.B
.B.B.B.B.B.B.B.B.16BABA16.4B8.20B2.B13.B.7B3.3B3.6B3.4B$50B2A17.4B6.
26B5.B.13B5.3B2.5B2.4B$51B19.4B4.67B$49B22.4B4.21B2A42B$50B22.4B3.21B
2A41B$50B23.4B3.41B2A20B5.2A$B.B.B.B.B.B.B.B.B.B.B.B.B.B.B.B.B.16B24.
4B2.41B2A20B5.A$36.14B25.4B.29B2.2B3.27B2.BA.A$36.13B27.17B.7B2.4B10.
10B2.B3.B.9B.B2A$36.11B30.16B2.6B19.6B9.11B$35.13B30.4B.10B3.3B23.3B
10.11B$34.15B30.14B4.B26.B8.2AB.9B$33.16B31.12B32.2A6.A.AB2.7B$32.17B
32.11B33.A6.A5.8B$33.16B32.11B30.3A6.2A4.8B$34.13B35.9B31.A14.7B$34.
5B2A2B39.9B45.11B$36.3B2A2B39.9B3.2A41.11B$36.8B38.9B3.A42.11B$35.8B
40.9BA.A42.11B$35.8B17.A23.6B2.2A41.2AB2.8B$35.7B16.3A19.10B44.A.AB3.
7B$35.7B15.A21.11B44.A6.7B$36.6B15.2A20.11B.2B40.2A7.6B$36.6B13.4B19.
14B2A48.7B$37.5B12.3B5.B.7B7.12B.B2A48.8B$37.6B10.4B.13B4.B2.13B.B50.
8B$36.6B4.45B52.8B$36.7B2.45B52.6B2.B2A$37.6B2.45B52.7B.BA.A$37.7B.
22B2A21B53.6B4.A$36.31B2A22B52.6B4.2A$36.19B2.2B3.11B2.2B3.7B2A2B.2B
49.6B$35.17B11.10B8.B.3BA2BA3B2A47.8B$31.B3.15B16.6B13.2B2A2B.B2A46.
8B$30.2AB.15B19.3B14.6B2.B47.9B$30.2A18B20.B15.4B51.9B$31.B.3B2A12B
20.2A14.4B50.10B$34.2B2A11B22.A15.2B51.3B2A5B$35.2B2.10B19.3A63.2A3.
4B2A5B$34.2B3.6B.B21.A24.A41.A3.11B$33.B2AB2.4B50.3A39.A.A12B$34.2A3.
2B2AB52.A33.2A4.2A2.8B$41.2A38.2A3.2A7.2A18.A15.A9.7B4.2A$80.B2AB.B2A
B6.4B14.3A15.A.AB7.6B4.A$81.2B2.3B3.B5.3B12.A19.2AB.3B3.6B.BA.A$82.3B
.3B.4B3.4B11.2A20.14B.B2A$74.2A5.7B.13B11.B20.16B$75.A5.23B9.3B19.14B
$75.A.AB.19B.8B4.6B16.16B$76.2AB.29B2.10B11.18B$78.44B3.2B2.20B$78.
37B2A31B$78.37B2A22B.7B$79.60B2.6B$81.58B3.6B$79.59B6.4B$79.2A3.25B4.
13B.4B12.B2A2B$80.A3.20B4.B4.7B.B4.4B14.2A.B2A$77.3A6.15B7.2A15.4B18.
BA.A$77.A8.11B12.A14.4B22.A$85.13B8.3A14.4B23.2A$84.15B7.A15.4B$84.
16B21.4B$84.17B19.4B$84.16B19.4B$86.13B19.4B$86.3B.2B2A5B18.4B$84.4B
2.2B2A3B19.4B$84.2A3.8B18.4B$85.A4.8B16.4B$82.3A5.8B15.4B$82.A8.7B14.
4B$91.7B13.4B$83.A7.6B13.4B$82.A.A6.6B12.4B$82.A.A6.5B12.4B$80.3A.2A
4.6B11.4B$79.A4.B6.6B9.4B$80.3AB2AB3.7B8.4B$82.A.2AB.8B8.4B$86.10B8.
3B$86.6B2A3B5.2AB$86.6B2A2B5.A.AB$86.10B5.A$85.11B2.BA.A$85.12B.B2A$
84.15B$83.16B$80.2B.16B$79.2A18B$79.2AB.17B$80.B.4B.8B2.4B$87.7B4.4B$
88.6B5.4B$90.4B6.4B$92.3BA5.4B$93.BA.A5.4B$94.A.A6.4B$95.A8.4B$96.3A
6.4B$98.A7.4B$107.4B$108.4B$109.4B$110.4B$111.4B$112.4B$113.4B$114.4B
$115.4B$116.4B$117.4B$118.4B$119.4B$120.4B$121.4B$122.4B$123.4B$124.
4B$125.4B$126.4B$127.4B$128.4B$129.4B$130.4B$131.4B$132.4B$124.B.B.B.
B.5B$122.17B$122.18B.B$120.21B2A$121.20B2A$121.18B2.B$123.B.B.B.B.B.B
.B2.2B$137.B2A$138.A.A$136.A.A.3A$136.2A5.A$142.2A!
#C [[ STEP 50 ]]

In case it isn't clear, it seems like this would be a perfect occasion to use the new 0-degree Snarkmaker recipe to bend the construction arm around to do the destruction. The rectangular Orthogonoid will need different tricks, but I'll save that for later.
dvgrn
Moderator
 
Posts: 3996
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI

Re: Orthogonoid working notes

Postby dvgrn » June 21st, 2017, 8:52 am

dvgrn wrote:What's the smallest number of slow SW gliders that can clean up an old Orthogonoid construction arm safely?

The number is apparently less than or equal to 45 gliders:

x = 20705, y = 20719, rule = B3/S23
20703bo$20702bo$20702b3o498$20220bo$20219bo$20219b3o498$19724bo$19723b
o$19723b3o498$19130bo$19129bo$19129b3o498$18721bo$18720bo$18720b3o498$
18219bo$18218bo$18218b3o498$17603bo$17602bo$17602b3o498$17144bo$17143b
o$17143b3o498$16649bo$16648bo$16648b3o498$16165bo$16164bo$16164b3o498$
15588bo$15587bo$15587b3o498$15087bo$15086bo$15086b3o498$14599bo$14598b
o$14598b3o498$14100bo$14099bo$14099b3o498$13632bo$13631bo$13631b3o498$
13141bo$13140bo$13140b3o498$12694bo$12693bo$12693b3o498$12215bo$12214b
o$12214b3o498$11706bo$11705bo$11705b3o498$11177bo$11176bo$11176b3o498$
10730bo$10729bo$10729b3o498$10227bo$10226bo$10226b3o498$9616bo$9615bo$
9615b3o498$9201bo$9200bo$9200b3o498$8666bo$8665bo$8665b3o498$8117bo$
8116bo$8116b3o498$7700bo$7699bo$7699b3o498$7193bo$7192bo$7192b3o498$
6666bo$6665bo$6665b3o498$6172bo$6171bo$6171b3o498$5622bo$5621bo$5621b
3o498$5185bo$5184bo$5184b3o498$4640bo$4639bo$4639b3o498$4089bo$4088bo$
4088b3o498$3659bo$3658bo$3658b3o498$3121bo$3120bo$3120b3o498$2745bo$
2744bo$2744b3o498$2186bo$2184b2o$2185b2o498$1603bo$1602bo$1602b3o498$
1224bo$1223bo$1223b3o498$733bo$732bo$732b3o348$481bo$480bo$480b3o2$
241bo$240bo$240b3o42$187bo$186bo$186b3o32$166bo$165bo$165b3o20$111b2o$
111b2o3$108b2o$108bobo$106bobob3o$106b2o5bo$112b2o88$26b2o$25bobo$19b
2o4bo$17bo2bo2b2ob4o$17b2obobobobo2bo$20bobobobo$20bobob2o$21bo2$34b2o
$25b2o7bo$25b2o5bobo$32b2o2$50bo$50b3o$53bo$4bo47b2o8b2o$4b3o55bo$7bo
14b2o36bobo$6b2o15bo36b2o$20b3o23b2o$20bo25b2o26b2o$74bo$72bobo30bo$
72b2o22bo7bobo$9b2o35b2o48b3o6bo$8bo2bo34b2o51bo$9b2o9bo77b2o$19bobo$
20b2o2$70b2o$70b2o$91b2o25b2o$91b2o25bo$116bobo$116b2o3$103b2o$94b2o6b
obo$95bo6bo$92b3o6b2o$92bo$9b2o$9b2o53b2o$64bo$62bobo$30bo31b2o41b2o$
28b3o73bobo$27bo76bo$27b2o74b2o$62b2o$62b2o3$121b2o$121bobo$37b2o84bo$
37b2o84b2o$57b2o$56bo2bo3b2o$57b2o4b2o$2o$2o$6b2o32b2o$6b2o33bo71b2o$
38b3o63b2o7b2o$38bo24bo41bo$4b2o57b3o39bobo$4b2o5b2o53bo33b2o4b2o$11b
2o38b2o3b2o7b2o18bo15bo20b2o$51b2o3b2o25b3o15bobo18bo$82bo19b2o16bobo$
82b2o36b2o$44b2o$45bo$45bobo$46b2o2$85b2o$85b2o4$49b2o63b2o$50bo63b2o
2b2o$47b3o28b2o38bobo$47bo31bo40bo$76b3o41b2o$76bo5$62b2o$62b2o$54b2o$
55bo$52b3o$52bo2$53bo$52bobo$52bobo$50b3ob2o$49bo$50b3ob2o$52bob2o2$
62b2o8b2o$62b2o7bobo$71bo$69bobo$69b2o4$49b2o$49b2o5$65bo$64bobo$64bob
o$65bo$66b3o$68bo30$111b2o$111b2o3$108b2o$108bobo$106bobob3o$106b2o5bo
$112b2o!
#C [[ STEP 50 ]]

That was found with the dumbest possible greedy algorithm, so I'm sure it can be radically improved. If nothing else, the gliders badly need to be shuffled into a more sensible order, generally NW to SE.
dvgrn
Moderator
 
Posts: 3996
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI

Re: Orthogonoid working notes

Postby dvgrn » June 28th, 2017, 9:41 am

dvgrn wrote:
dvgrn wrote:What's the smallest number of slow SW gliders that can clean up an old Orthogonoid construction arm safely?

The number is apparently less than or equal to 45 gliders...

Last-minute circuitry adjustments are no fun at all. I had a nice 45-glider near-to-far slow salvo all compiled into single-channel form -- Calcyman explained that slmake will cheerfully produce a single-channel recipe if the infile.mc consists of a slow salvo of gliders aimed to miss an initial block.

But then it seemed like a good idea to get the output glider from earlier in the Herschel circuit, so that the Snark on the construction arm could be pre-built, and then removed by a Snark-destroy recipe when it was time to shoot down the parent constructor. So the Orthogonoid construction arm should look like this:

Code: Select all
x = 188, y = 279, rule = LifeHistory
100.4B$101.4B$102.4B$103.4B$101.B.5B$100.10B$100.11B.B$100.12B2A$100.
12B2A$100.10B2.B$100.B.B.B.B2.2B$108.B2A$109.A.A$107.A.A.3A$107.2A5.A
$113.2A36$79.2A$79.2A51$27.2A$26.A.A$20.2A4.A$18.A2.A2.2A.4A$18.2A.A.
A.A.A2.A$21.A.ABABAB$21.A.AB2AB$22.AB.2B$25.3B$25.4B6.2A$23.3B2AB6.A$
23.3B2AB3.BA.A$21.10B.B2A$20.13B$19.14B18.A$18.15B18.3A$17.4B2.8B23.A
$5.A10.4B5.6B22.2A8.2A$5.3A7.4B4.9B21.5B5.A$8.A5.4B5.2A4.4B22.4B.BA.A
$7.2A4.4B7.A5.4B14.B4.6B.B2A$7.9B5.3A7.4B12.2AB.10B$9.6B6.A10.4B11.2A
12B14.2A$8.6B19.4B11.B.11B14.A$8.6B20.4B12.13B3.4B2.BA.A30.A$9.6B20.
4B9.2B.12B2.5B2.B2A22.A7.A.A$8.2B2A4B3.3B14.4B7.2A24B24.3A4.2BAB$7.2B
A2BA3B.6B14.4B6.2A24B27.A4.2B10.A$6.4B2A9BA2B14.4B6.B.B.20B27.2A5.4B
5.3A$B.B.16BABA16.4B8.20B2.B13.B.7B3.3B3.6B3.A2B$21B2A17.4B6.26B5.B.
13B5.3B2.5B2.B2AB$22B19.4B4.67B$20B22.4B4.21B2A42B$21B22.4B3.21B2A41B
$21B23.4B3.41B2A20B5.2A$.B.B.16B24.4B2.41B2A20B5.A$7.14B25.4B.29B2.2B
3.27B2.BA.A$7.13B27.17B.7B2.4B10.10B2.B3.B.9B.B2A$7.11B30.16B2.6B19.
6B8.12B$6.13B30.4B.10B3.3B23.3B10.11B$5.15B30.14B4.B26.B8.2AB.9B$4.
16B31.12B32.2A6.A.AB2.7B$3.17B32.11B33.A6.A5.8B$4.16B32.11B30.3A6.2A
4.8B$5.13B35.9B31.A14.7B$5.5B2A2B39.9B45.11B$7.3B2A2B39.9B3.2A41.11B$
7.8B38.9B3.A42.11B$6.8B40.9BA.A42.11B$6.8B17.A23.6B2.2A41.2AB2.8B$6.
7B16.3A19.10B44.A.AB3.7B$6.7B15.A21.11B44.A6.7B$7.6B15.2A20.11B.2B40.
2A7.6B$7.6B13.4B19.14B2A48.7B$8.5B12.3B5.B.7B7.12B.B2A48.8B$8.6B10.4B
.13B4.B2.13B.B50.8B$7.6B4.45B52.8B$7.7B2.45B52.6B2.B$8.6B2.45B52.7B.B
$8.7B.22B2A21B53.6B$7.31B2A22B52.6B$7.19B2.2B3.11B2.2B3.7B2A2B.2B49.
6B$6.17B11.10B8.B.3BA2BA3B2A47.8B$2.B3.15B16.6B13.2B2A2B.B2A46.8B$.2A
B.15B19.3B14.6B2.B47.9B$.2A18B20.B15.4B51.9B49.2A$2.B.3B2A12B20.2A14.
4B50.10B49.2A$5.2B2A11B22.A15.2B51.3B2A5B$6.2B2.10B19.3A63.2A3.4B2A5B
$5.2B3.6B.B21.A24.A41.A3.11B$4.B2AB2.4B50.3A39.A.A12B$5.2A3.2B2AB52.A
33.2A4.2A2.8B$12.2A38.2A3.2A7.2A18.A15.A9.7B4.2A$51.B2AB.B2AB6.4B14.
3A15.A.AB7.6B4.A$52.2B2.3B3.B5.3B12.A19.2AB.3B3.6B.BA.A$53.3B.3B.4B3.
4B11.2A20.14B.B2A$45.2A5.7B.13B11.B20.16B$46.A5.23B9.3B19.14B34.A$46.
A.AB.19B.8B4.6B16.16B31.3A$47.2AB.29B2.10B11.18B30.A$49.44B3.2B2.20B
31.2A$49.37B2A31B$49.37B2A22B.7B$50.60B2.6B$52.58B3.6B$50.59B6.4B$50.
2A3.25B4.13B.4B12.B2A2B$51.A3.20B4.B4.7B.B4.4B14.2A.B2A20.2A$48.3A6.
15B7.2A15.4B18.BA.A18.A.A5.2A$48.A8.11B12.A14.B2AB22.A18.A7.2A$56.13B
8.3A14.3BA23.2A16.2A$55.15B7.A16.3A$55.16B23.AB57.A$55.17B22.B54.2A.A
.A$55.16B77.A.A.A.A$57.13B75.A2.A.A.A.A.2A$57.3B.2B2A5B75.4A.2A2.A2.A
$55.4B2.2B2A3B81.A4.2A$55.2A3.8B79.A.A$56.A4.8B78.2A$53.3A5.8B$53.A8.
7B$62.7B$54.A7.6B$53.A.A6.6B$53.A.A6.5B$51.3A.2A4.6B$50.A4.B6.6B$51.
3AB2AB3.7B$53.A.2AB.8B$57.10B$57.6B2A3B$57.6B2A2B5.2A$57.10B5.A$56.
11B2.BA.A$56.12B.B2A$55.15B$54.16B$51.2B.16B$50.2A18B$50.2AB.17B$51.B
.4B.8B2.4B$58.7B4.4B$59.6B5.4B$61.4B6.4B$63.3BA5.4B$64.BA.A5.4B$65.A.
A6.4B$66.A8.4B$67.3A6.4B$69.A7.4B$78.4B$79.4B$80.4B$81.4B$82.4B$83.4B
$84.4B$85.4B$86.4B$87.4B$88.4B$89.4B$90.4B$91.4B$92.4B$93.4B$94.4B$
95.4B$96.4B$97.4B$98.4B$99.4B$100.4B$101.4B$102.4B$103.4B$83.B.B.B.B.
B.B.B.B.B.B.5B$83.27B$83.12B5A11B.B$83.11BA4BA12B2A$83.16BA12B2A$83.
11BA3BA11B2.B$84.B.B.B.B.B.B.A.B.B.B.B.B2.2B$108.B2A$109.A.A$107.A.A.
3A$107.2A5.A$113.2A!
#C [[ THUMBNAIL THUMBSIZE 2 ZOOM 3 Y 50 HEIGHT 600 ]]

Just one eater had to move, but of course it was one that participated in a big explosive reduction in the early part of the last cleanup recipe. So my mediocre greedy destruction script only seems to be able to manage 52 gliders now.

Anyway, when the circuitry gets retired, what's left to be destroyed will look like this:

x = 6709, y = 6594, rule = B3/S23
6706bo$6706bobo$6706b2o46$6577bo$6577bobo$6577b2o116$6448bo$6448bobo$
6448b2o126$6319bo$6319bobo$6319b2o117$6190bo$6190bobo$6190b2o126$6061b
o$6061bobo$6061b2o127$5932bo$5932bobo$5932b2o120$5803bo$5803bobo$5803b
2o122$5674bo$5674bobo$5674b2o122$5545bo$5545bobo$5545b2o130$5416bo$
5416bobo$5416b2o132$5287bo$5287bobo$5287b2o128$5158bo$5158bobo$5158b2o
109$5029bo$5029bobo$5029b2o120$4900bo$4900bobo$4900b2o131$4771bo$4771b
obo$4771b2o125$4642bo$4642bobo$4642b2o118$4513bo$4513bobo$4513b2o123$
4384bo$4384bobo$4384b2o123$4255bo$4255bobo$4255b2o129$4126bo$4126bobo$
4126b2o100$3999bo$3997b2o$3998b2o133$3870bo$3868b2o$3869b2o118$3741bo$
3739b2o$3740b2o145$3610bo$3610bobo$3610b2o130$3481bo$3481bobo$3481b2o
127$3352bo$3352bobo$3352b2o127$3223bo$3223bobo$3223b2o117$3094bo$3094b
obo$3094b2o127$2965bo$2965bobo$2965b2o119$2836bo$2836bobo$2836b2o123$
2707bo$2707bobo$2707b2o126$2578bo$2578bobo$2578b2o121$2449bo$2449bobo$
2449b2o113$2320bo$2320bobo$2320b2o125$2191bo$2191bobo$2191b2o101$2062b
o$2062bobo$2062b2o119$1933bo$1933bobo$1933b2o130$1804bo$1804bobo$1804b
2o115$1675bo$1675bobo$1675b2o130$1546bo$1546bobo$1546b2o128$1417bo$
1417bobo$1417b2o155$1288bo$1288bobo$1288b2o151$1159bo$1159bobo$1159b2o
109$1030bo$1030bobo$1030b2o122$901bo$901bobo$901b2o131$772bo$772bobo$
772b2o95$643bo$643bobo$643b2o141$514bo$514bobo$514b2o119$385bo$385bobo
$385b2o117$256bo$256bobo$256b2o123$111b2o$111b2o3$108b2o$108bobo$106bo
bob3o$106b2o5bo$112b2o9$127bo$127bobo$127b2o77$26b2o$25bobo$19b2o4bo$
17bo2bo2b2ob4o$17b2obobobobo2bo$20bobobobo$20bobob2o$21bo2$34b2o$25b2o
7bo$25b2o5bobo$32b2o2$50bo$50b3o$53bo$4bo47b2o8b2o$4b3o55bo$7bo14b2o
36bobo$6b2o15bo36b2o$20b3o23b2o$20bo25b2o26b2o$74bo$72bobo30bo$72b2o
22bo7bobo$9b2o35b2o48b3o6bo$8bo2bo34b2o51bo16bo$9b2o9bo77b2o14b3o$19bo
bo91bo$20b2o91b2o2$70b2o$70b2o$91b2o25b2o$91b2o25bo$116bobo$116b2o3$
103b2o$94b2o6bobo$95bo6bo$92b3o6b2o$92bo$9b2o$9b2o53b2o$64bo$62bobo$
30bo31b2o41b2o$28b3o73bobo$27bo76bo$27b2o74b2o$62b2o$62b2o5$37b2o$37b
2o$57b2o$56bo2bo3b2o$57b2o4b2o$2o$2o$6b2o32b2o$6b2o33bo71b2o$38b3o63b
2o7b2o$38bo24bo41bo$4b2o57b3o39bobo$4b2o5b2o53bo33b2o4b2o$11b2o38b2o3b
2o7b2o18bo15bo20b2o$51b2o3b2o25b3o15bobo18bo$82bo19b2o16bobo$82b2o36b
2o$44b2o$45bo$45bobo$46b2o2$85b2o$85b2o4$49b2o63b2o$50bo63b2o2b2o$47b
3o28b2o38bobo$47bo31bo40bo$76b3o15b2o24b2o$76bo18bo$92b3o$92bo3$62b2o$
62b2o$54b2o$55bo$52b3o$52bo2$53bo$52bobo$52bobo$50b3ob2o$49bo$50b3ob2o
$52bob2o2$62b2o$62b2o7b2o$71bo$69bobo$69b2o4$49b2o$49b2o5$65bo$64bobo$
64bobo$65bo$66b3o$68bo30$111b2o$111b2o3$108b2o$108bobo$106bobob3o$106b
2o5bo$112b2o!
#C [[ X -3300 Y 3200 ZOOM 1.4 STEP 50 AUTOSTART ]]

Can somebody write a better meteor-shower search utility, or should I just leave it as it is? It probably won't make any difference to the size of the Orthogonoid, because it will be adjusted to a power-of-two period anyway, and the population increase will be a fraction of a percent... but this cleanup recipe just seems excessive somehow.
dvgrn
Moderator
 
Posts: 3996
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI

Re: Orthogonoid working notes

Postby simeks » June 28th, 2017, 5:51 pm

dvgrn wrote:Can somebody write a better meteor-shower search utility, or should I just leave it as it is? It probably won't make any difference to the size of the Orthogonoid, because it will be adjusted to a power-of-two period anyway, and the population increase will be a fraction of a percent...

I thought it could be nice to have a utility for this, so I'm working on one...
Here's a sample result using 36 gliders in a 32 lanes wide firing window:

x = 3498, y = 3383, rule = LifeHistory
3495.A.A$3495.2A$3496.A13$3400.A$3398.2A$3399.2A89$3297.A.A$3297.2A$
3298.A88$3197.A$3195.2A$3196.2A94$3099.A.A$3099.2A$3100.A80$3028.A$
3026.2A$3027.2A90$2923.A.A$2923.2A$2924.A89$2822.A.A$2822.2A$2823.A
96$2729.A$2729.A.A$2729.2A89$2629.A$2627.2A$2628.2A90$2529.A$2527.2A$
2528.2A90$2427.A.A$2427.2A$2428.A89$2342.A.A$2342.2A$2343.A90$2249.A.
A$2249.2A$2250.A78$2139.A$2137.2A$2138.2A81$2030.A$2028.2A$2029.2A
105$1943.A.A$1943.2A$1944.A79$1832.A.A$1832.2A$1833.A94$1737.A$1737.A
.A$1737.2A103$1649.A.A$1649.2A$1650.A96$1557.A$1555.2A$1556.2A83$
1448.A.A$1448.2A$1449.A105$1363.A.A$1363.2A$1364.A75$1248.A.A$1248.2A
$1249.A99$1157.A.A$1157.2A$1158.A94$1063.A$1061.2A$1062.2A89$962.A$
960.2A$961.2A93$863.A.A$863.2A$864.A98$771.A.A$771.2A$772.A83$666.A$
664.2A$665.2A76$550.A.A$550.2A$551.A100$460.A.A$460.2A$461.A98$370.A$
368.2A$369.2A95$275.A$273.2A$274.2A83$168.A$166.2A$167.2A102$80.A$78.
2A$79.2A46$26.2A$25.A.A$19.2A4.A$17.A2.A2.2A.4A$17.2A.A.A.A.A2.A$20.A
.A.A.A$20.A.A.2A$21.A2$34.2A$25.2A7.A$25.2A5.A.A$32.2A2$50.A$50.3A$
53.A$4.A47.2A8.2A$4.3A55.A$7.A14.2A36.A.A$6.2A15.A36.2A$20.3A23.2A$
20.A25.2A26.2A$74.A$72.A.A30.A$72.2A22.A7.A.A$9.2A35.2A48.3A6.A$8.A2.
A34.2A51.A16.A$9.2A9.A77.2A14.3A$19.A.A91.A$20.2A91.2A2$70.2A$70.2A$
91.2A25.2A$91.2A25.A$116.A.A$116.2A3$103.2A$94.2A6.A.A$95.A6.A$92.3A
6.2A$92.A$9.2A$9.2A53.2A$64.A$62.A.A$30.A31.2A41.2A$28.3A73.A.A$27.A
76.A$27.2A74.2A$62.2A$62.2A5$37.2A$37.2A$57.2A$56.A2.A3.2A$57.2A4.2A$
2A$2A$6.2A32.2A$6.2A33.A71.2A$38.3A63.2A7.2A$38.A24.A41.A$4.2A57.3A
39.A.A$4.2A5.2A53.A33.2A4.2A$11.2A38.2A3.2A7.2A18.A15.A20.2A$51.2A3.
2A25.3A15.A.A18.A$82.A19.2A16.A.A$82.2A36.2A$44.2A$45.A$45.A.A$46.2A
2$85.2A$85.2A4$49.2A63.2A$50.A63.2A2.2A$47.3A28.2A38.A.A$47.A31.A40.A
$76.3A15.2A24.2A$76.A18.A$92.3A$92.A3$62.2A$62.2A$54.2A$55.A$52.3A$
52.A2$53.A$52.A.A$52.A.A$50.3A.2A$49.A$50.3A.2A$52.A.2A2$62.2A$62.2A
7.2A$71.A$69.A.A$69.2A4$49.2A$49.2A5$65.A$64.A.A$64.A.A$65.A$66.3A$
68.A30$111.2A$111.2A3$108.2A$108.A.A$106.A.A.3A$106.2A5.A$112.2A!
simeks
 
Posts: 324
Joined: March 11th, 2015, 12:03 pm
Location: Sweden

Re: Orthogonoid working notes

Postby dvgrn » June 28th, 2017, 10:02 pm

simeks wrote:
dvgrn wrote:I thought it could be nice to have a utility for this, so I'm working on one...
Here's a sample result using 36 gliders in a 32 lanes wide firing window...

Looks good! In practice a different window will be needed, though -- my last pattern included the eater-tie-eater/block constellation in the right position to mark the allowable edge of the firing range.

The Orthogonoid is working as a puffer now -- should be all done pretty soon. I'll probably just compile the meteor shower recipe I have, since it won't make any difference to the period of the (Hashlife-friendly) spaceship.

Orthogonoid-twotothetwentytwo.mc.gz
Orthogonoid puffer, no cleanup yet -- period 2^22 ticks
(118.05 KiB) Downloaded 31 times

EDIT: Looks like it will take something over 6GB of RAM for Golly to be able to "run away" with this one -- lots of different hash tiles with the signals going back and forth next to each other, as usual. Does anyone have a test system with some unreasonable number of gigs of RAM available? Golly's memory use should stabilize at some point, but I have no idea when -- my best system has only 8GB available.

I'm hopeful that it will be a much more reasonable number of gigabytes for Scorbie's new Demonoid...!
dvgrn
Moderator
 
Posts: 3996
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI

Re: Orthogonoid working notes

Postby biggiemac » June 29th, 2017, 1:28 am

I'm giving golly 10GB of my RAM and still getting 99% garbage collections when trying to run at 2^18.

Edit: Pushed it to 13 GB and still got 99% GCs so probably my 16GB laptop can't run away either.
Physics: sophistication from simplicity.
User avatar
biggiemac
 
Posts: 502
Joined: September 17th, 2014, 12:21 am
Location: California, USA

Re: Orthogonoid working notes

Postby dvgrn » June 29th, 2017, 3:01 am

biggiemac wrote:I'm giving golly 10GB of my RAM and still getting 99% garbage collections when trying to run at 2^18.

Edit: Pushed it to 13 GB and still got 99% GCs so probably my 16GB laptop can't run away either.

Hmm. Not too surprised -- this is an ambitious amount of circuitry, and all the different ways the recipe can fold over on itself add up to a lot of hashtiles. The fact that you were testing the puffer rather than the spaceship would have added a few tiles, though probably not a significant number.

Here's a completed period 2^23 Orthogonoid spaceship to try -- it should have fewer hashtiles than a 2^22 model, though again probably not enough fewer to make any difference.

Orthogonoid-p2^23.mc.gz
Double-wide Orthogonoid, speed c/32768 (allows the recipe to straighten out twice per period)
(110.61 KiB) Downloaded 35 times

My laptop's too slow to run these things through several cycles tonight, but things are looking promising:

Compare: (<)Orthogonoid-2^23.rle (953750 bytes)
   with: (>)Orthogonoid-2^23+4194304-reflected.rle (953750 bytes)

The files are identical

Looks like this one will fit in a 2096822x565 rectangle a lot of the time. That's less than half of the size of the Demonoid if we go by bounding box -- 1,184,704,430 cells in this bounding box versus 3,023,569,640 for the Demonoid.

... Which just goes to show what a silly measurement the bounding box is. The Orthogonoid is much bigger by any other metric, and correspondingly slower.

We can cut the bounding box more or less in half by moving the two halves 2^20 cells closer together, and still have a theoretically Hashlife-friendly Orthogonoid -- it's easy to do, just wait until the recipe is maximally folded over, then move the empty half. But just like the maximally folded linear propagator, it runs slower all the time, because the recipe is always folded:

Orthogonoid-p2^22.mc.gz
Smallest Hashlife-friendly Orthogonoid (until someone does a *lot* of recipe optimization) -- speed c/16384
(106.91 KiB) Downloaded 45 times

I think the minimum period for this particular stream of MWSSes is something like 3,476,016. Technically an Orthogonoid can be squeezed a little smaller than that, because the component recipes actually aren't quite packed as tight as they could be.

Then someone could spend approximately a lifetime figuring out how to improve on slmake's compiled recipes. You can see here and there where the algorithm could be a little more efficient. Of course a better meteor-shower cleanup salvo would shorten things up a little more. Really there are possible improvements to be made all over the place, but even all together I don't think they'll add up to a power of two improvement any time soon.

Anyway, no more optimization for me! I'll probably try taking this recipe minus the cleanup, and see if I can write code to fold it successfully into a square Orthogonoid puffer. Have to re-do the cleanup using self-destruct circuits to get an actual square spaceship. It will have a much higher speed; no idea if that will translate into enough fewer hashtiles to make Golly happy.

EDIT: Since it looks like it won't matter much to Golly anyway, here's a copy of the Orthogonoid adjusted down to minimum period, p3476016. Not sure what the phase with the smallest bounding box or population is yet, but it's around 868,750 by 800, and 469,000 ON cells.

Orthogonoid-p3476016.mc.gz
Minimum adjustment without obsessive optimization -- period 3,476,016
(105.27 KiB) Downloaded 37 times
dvgrn
Moderator
 
Posts: 3996
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI

Re: Orthogonoid working notes

Postby simeks » June 29th, 2017, 5:22 pm

dvgrn wrote:In practice a different window will be needed, though -- my last pattern included the eater-tie-eater/block constellation in the right position to mark the allowable edge of the firing range.

Here's a solution with 32 gliders that saves the MWSS-to-G converter:

x = 2639, y = 2543, rule = LifeHistory
2564.A$2562.2A$2563.2A16$2638.A$2636.2A$2637.2A3$2537.A.A$2537.2A$
2538.A26$2488.A.A$2488.2A$2489.A10$2463.A$2461.2A$2462.2A23$2431.A$
2429.2A$2430.2A107$2348.A$2348.A.A$2348.2A32$2312.A$2311.A$2311.3A19$
2278.A$2277.A$2277.3A18$2256.A$2254.2A$2255.2A22$2242.A$2241.A$2241.
3A143$2103.A$2101.2A$2102.2A38$2035.A$2033.2A$2034.2A35$1975.A$1974.A
$1974.3A69$1912.A$1912.A.A$1912.2A30$1896.A$1895.A$1895.3A34$1865.A$
1863.2A$1864.2A208$1610.A$1608.2A$1609.2A89$1537.A$1535.2A$1536.2A37$
1507.A.A$1507.2A$1508.A49$1436.A$1434.2A$1435.2A78$1328.A.A$1328.2A$
1329.A87$1219.A$1217.2A$1218.2A87$1147.A$1145.2A$1146.2A72$1066.A.A$
1066.2A$1067.A122$924.A$922.2A$923.2A78$839.A$837.2A$838.2A171$666.A$
664.2A$665.2A92$572.A$571.A$571.3A80$492.A$490.2A$491.2A101$389.A$
387.2A$388.2A105$281.A$279.2A$280.2A133$111.2A$111.2A3$108.2A$108.A.A
$106.A.A.3A$106.2A5.A$112.2A88$26.2A$25.A.A$19.2A4.A$17.A2.A2.2A.4A$
17.2A.A.A.A.A2.A$20.A.A.A.A$20.A.A.2A$21.A2$34.2A$25.2A7.A$25.2A5.A.A
$32.2A2$50.A$50.3A$53.A$4.A47.2A8.2A$4.3A55.A$7.A14.2A36.A.A$6.2A15.A
36.2A$20.3A23.2A$20.A25.2A26.2A$74.A$72.A.A30.A$72.2A22.A7.A.A$9.2A
35.2A48.3A6.A$8.A2.A34.2A51.A16.A$9.2A9.A77.2A14.3A$19.A.A91.A$20.2A
91.2A2$70.2A$70.2A$91.2A25.2A$91.2A25.A$116.A.A$116.2A3$103.2A$94.2A
6.A.A$95.A6.A$92.3A6.2A$92.A$9.2A$9.2A53.2A$64.A$62.A.A$30.A31.2A41.
2A$28.3A73.A.A$27.A76.A$27.2A74.2A$62.2A$62.2A5$37.2A$37.2A$57.2A$56.
A2.A3.2A$57.2A4.2A$2A$2A$6.2A32.2A$6.2A33.A71.2A$38.3A63.2A7.2A$38.A
24.A41.A$4.2A57.3A39.A.A$4.2A5.2A53.A33.2A4.2A$11.2A38.2A3.2A7.2A18.A
15.A20.2A$51.2A3.2A25.3A15.A.A18.A$82.A19.2A16.A.A$82.2A36.2A$44.2A$
45.A$45.A.A$46.2A2$85.2A$85.2A4$49.2A63.2A$50.A63.2A2.2A$47.3A28.2A
38.A.A$47.A31.A40.A$76.3A15.2A24.2A$76.A18.A$92.3A$92.A3$62.2A$62.2A$
54.2A$55.A$52.3A$52.A2$53.A$52.A.A$52.A.A$50.3A.2A$49.A$50.3A.2A$52.A
.2A2$62.2A$62.2A7.2A$71.A$69.A.A$69.2A4$49.2A$49.2A5$65.A$64.A.A$64.A
.A$65.A$66.3A$68.A30$111.2A$111.2A3$108.2A$108.A.A$106.A.A.3A$106.2A
5.A$112.2A!
simeks
 
Posts: 324
Joined: March 11th, 2015, 12:03 pm
Location: Sweden

Re: Orthogonoid working notes

Postby dvgrn » June 29th, 2017, 10:46 pm

simeks wrote:Here's a solution with 32 gliders that saves the MWSS-to-G converter...

Yeah, that looks a lot more professional than my pretty much one glider per still life solution. I'll get around to recompiling and incorporating this eventually, if no 31- or 30-glider solutions come along in the meantime.

Luckily the destruction happens after the construction is already done, so this can't be used to reduce the period of existing Orthogonoids. The minimum period will still be 3,476,016 until someone gets around to writing an optimizer that can squeeze the last one or two or three ticks out of all those component elbow operations... or until we replace all those recipes with shorter ones that allow glider triplets, quadruplets, etc. That last might actually allow the Orthogonoid's period to drop below 2^21, I suppose.

EDIT: Statistics for the statistics-minded: minimum bounding box for the p3476016 Orthogonoid is 868,856 by 707, at T=218619 from the posted pattern. Minimum population is 467,746 at T=197084. Of course each minimum happens twice per period, 1738008 ticks apart.
dvgrn
Moderator
 
Posts: 3996
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI

Re: Orthogonoid working notes

Postby Hooloovoo » July 1st, 2017, 5:25 pm

dvgrn wrote:Does anyone have a test system with some unreasonable number of gigs of RAM available? Golly's memory use should stabilize at some point, but I have no idea when -- my best system has only 8GB available.

I'm hopeful that it will be a much more reasonable number of gigabytes for Scorbie's new Demonoid...!


I ran the period 2^23 orthogonoid through Golly on my biggest machine at a step size of 2^12. It stabilized at about 35G of RAM and took about a minute to run through the full period.
Last edited by Hooloovoo on July 1st, 2017, 6:41 pm, edited 1 time in total.
Hooloovoo
 
Posts: 3
Joined: July 11th, 2015, 8:59 pm

Re: Orthogonoid working notes

Postby dvgrn » July 1st, 2017, 5:50 pm

Hooloovoo wrote:I ran the period 2^23 orthoganoid through Golly on my biggest machine at a step size of 2^12. It stabilized at about 35G of RAM and took about a minute to run through the full period.

Thanks! That gives me a good data point for designing a Geminoid variant that Golly can handle with just a gigabyte or two of RAM. Basically it should be okay as long as there aren't any of those deadly back-and-forth streams of data.
dvgrn
Moderator
 
Posts: 3996
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI

Re: Orthogonoid working notes

Postby simsim314 » July 2nd, 2017, 1:28 am

Hey dvgrn Congrats! I was thinking to finish this project myself using calcyman script - but I see you've managed to finish it all by yourself (no surprise).

Have you modified something in calcyman code or maybe you used some additional scripts? If so can you please post them as well?

PS. Maybe you should add the "completed" to the topic name - so people that are not following every message could congrat, and be aware this project is done.
User avatar
simsim314
 
Posts: 1516
Joined: February 10th, 2014, 1:27 pm

Re: Orthogonoid working notes

Postby dvgrn » July 2nd, 2017, 9:43 am

simsim314 wrote:Hey dvgrn Congrats! I was thinking to finish this project myself using calcyman script - but I see you've managed to finish it all by yourself (no surprise).

Yes, the Orthogonoid was one of the really easy projects to finish. Haven't really gotten going on the multi-folded rectangular Orthogonoid yet -- that one will run even slower in Golly than the original. I seem to be putting off the more painful design problems, like a diamond-shaped self-constructor that will actually run well in Hashlife.

If somebody wants to tackle a self-constructor with a 2D loop, it looks like it might work to launch Corderships simultaneously in two directions, and then stop them with following gliders. Corderships are so slow that (if my math is right) you can't easily use them to make a recipe loop that has just two 180-degree reflectors. A 1D loop will end up being only just big enough for the gap between the Cordership-launching trigger glider and the Cordership-stopping following glider, leaving no room for the rest of the recipe unless you add extra one-time switching tricks...!

Until then, the true-period knightship might be an obvious next step, unless someone wants to try out Scorbie's new minimal Hashlife-friendly Demonoid. Or maybe Scorbie's constructor/reflector could be adapted for use in the oblique Geminoid blueprint, to get something more Hashlife-friendly with an adjustable width, with about the same population.

simsim314 wrote:Have you modified something in calcyman code or maybe you used some additional scripts? If so can you please post them as well?

Yes, everything I've been using is organized fairly well in the same thread where the slmake beta is posted. There are a few helper scripts a couple of posts down.

It might make sense to post a patched version of slmake in a new thread, to save people that editing step. Not sure when the next official release might appear -- I'm hoping for one that automatically compiles two single-channel recipes, one for each color for the first output glider, written to singlechannelA.txt and singlechannelB.txt instead of just being dumped to stdout. And there are rumors of other possible improvements.
dvgrn
Moderator
 
Posts: 3996
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI

Re: Orthogonoid spaceship -- completed!

Postby wwei23 » August 10th, 2017, 11:27 am

If you got rid of the deletion tape, would you get a puffer?
Replicator!
x = 3, y = 3, rule = B3/S234y
2bo$3o$bo!
User avatar
wwei23
 
Posts: 642
Joined: May 22nd, 2017, 6:14 pm
Location: The (Life?) Universe

Re: Orthogonoid working notes

Postby dvgrn » August 21st, 2017, 9:45 am

simeks wrote:Here's a solution with 32 gliders that saves the MWSS-to-G converter...

Is there an easy way to apply your searcher to the much smaller new Demonoid cleanup problem? I'd really like to see that thing running...!

wwei23 wrote:If you got rid of the deletion tape, would you get a puffer?

Of course. You don't need to ask questions like this, you can just try it yourself. For example, delete a few MWSSes randomly from the end of the recipe stream and see what happens. Or watch the very end of the cleanup process, and figure out exactly which MWSSes you should delete to get the behavior you want.
dvgrn
Moderator
 
Posts: 3996
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI

Previous

Return to Patterns

Who is online

Users browsing this forum: Sokwe and 2 guests