Execution of Old Guns by Variable-Speed Firing Squad

For discussion of specific patterns or specific families of patterns, both newly-discovered and well-known.
Post Reply
User avatar
Freywa
Posts: 877
Joined: June 23rd, 2011, 3:20 am
Location: Singapore
Contact:

Execution of Old Guns by Variable-Speed Firing Squad

Post by Freywa » March 21st, 2015, 2:14 am

There has been talk of updating Dieter and Peter's gun collection on several threads in accordance with the new discoveries of the snark, T flip-flop and boojum (among several new rhodium Herschel conduits). This thread is to collect all these discoveries to produce a single and easily available list of glider guns, not all of them true, for all periods up to 1000.

Fire away!
And not the one on GitHub, that's too far away…
Princess of Science, Parcly Taxel

Code: Select all

x = 31, y = 5, rule = B2-a/S12
3bo23bo$2obo4bo13bo4bob2o$3bo4bo13bo4bo$2bo4bobo11bobo4bo$2bo25bo!

User avatar
Kazyan
Posts: 1247
Joined: February 6th, 2014, 11:02 pm

Re: Execution of Old Guns by Variable-Speed Firing Squad

Post by Kazyan » March 21st, 2015, 3:25 am

Here, have 87 guns. Most of them are mine, the result of making simple loops with the G-to-H/syringe. The p94 is Mike Playle's, and I think a few of these are from Chris; I dunno, I just put improved guns in my folder regardless of source.
Attachments
Munitions Dump.zip
(40.79 KiB) Downloaded 1338 times
Tanner Jacobi
Coldlander, a novel, available in paperback and as an ebook. Now on Amazon.

User avatar
simsim314
Posts: 1823
Joined: February 10th, 2014, 1:27 pm

Re: Execution of Old Guns by Variable-Speed Firing Squad

Post by simsim314 » March 21st, 2015, 7:11 am

I think there is no point to use the old collection up to 1000 now, as the collection is totally obsolete, except of maybe for the low periods.

chris_c
Posts: 966
Joined: June 28th, 2014, 7:15 am

Re: Execution of Old Guns by Variable-Speed Firing Squad

Post by chris_c » March 21st, 2015, 8:27 am

There is a family of p358+8n guns starting from this:

Code: Select all

x = 52, y = 54, rule = B3/S23
28bo11b2o$27bobo10b2o$19bo3b2o2bobo$10b2o6bobo2bo2b2ob2o$10b2o6bobo3bo
bo$19bob4o2bob2o$21bo3bobob2o$20bo3bobo$19bo3bobo$19b2o3bo4$32b2o2b2o$
32b2o2bobo$36bo10b2o$46bo2bo$47b2obo$2o48bo$2o48b2o$35b2o$36bo5bo$33b
3o5bobo$21b2o10bo7bo2bo$21bobo19bo$23bo$23b2o22b2o$27b2o17bo2bo$28bo
17bobo$28bobo16bo$18bo10b2o12b3o$16b3o24b2o$15bo$15b2o$2o48b2o$bo48b2o
$bob2o$2bo2bo$3b2o$18b2o$18b2o4$27bo3b2o$26bobo3bo$25bobo3bo$21b2obobo
3bo$21b2obo2b4obo5b2o$25bobo3bobo5bo$21b2ob2o2bo2bobo2b3o$22bobo2b2o3b
o3bo13bo$10b2o10bobo26bo$10b2o11bo25b3o!
It allows an even smaller p78, plus I also accidentally discovered that the G->H can be overclocked at p74:

Code: Select all

x = 171, y = 74, rule = B3/S23
28bo11b2o85bo11b2o$27bobo10b2o84bobo10b2o$19bo3b2o2bobo88bo3b2o2bobo$
10b2o6bobo2bo2b2ob2o78b2o6bobo2bo2b2ob2o$10b2o6bobo3bobo82b2o6bobo3bob
o$19bob4o2bob2o78b2o7bob4o2bob2o$21bo3bobob2o77b3o9bo3bobob2o$9b3o8bo
3bobo81bobo8bo3bobo$8bo2bo7bo3bobo81bo2bo7bo3bobo$8b2o9b2o3bo82b3o8b2o
3bo$35b2o70b2o$10bo2bo21b2o72b2o$31b5o72b2obo16b2o$9bob2o17bo3bo72bo
18b2ob2o$8b2obo18bo75bo3b2o14bo2bo$9bobo18b2ob2o12b2o58b3o16bo2bo6b3o
7b2o$10b3o19bo13bo2bo57b2o18b2o7bo8bo2bo$9b3o28b2o5b2obo59bo26bo8b2obo
$2o38bobo7bo48b2o7b2o39bo$2o38bo9b2o47b2o7b2o39b2o$6bob3o24b2o67b2o2b
2o24b2o$5b6o25bo71b2o25bo$5bo27b3o67b3o26b3o$5bobo13b2o10bo71b4o11b2o
10bo$8bo12bobo83b3o10bobo$6bobob2o11bo22bo60bob2o11bo$8bo2bo11b2o20bob
o58b2o2bo11b2o$8b3o34bo2bo57b2ob2o$9bo37bo59bobo$108bo$51b2o$31b2o17bo
2bo$32bo17bobo$32bobo16bo103b2o$22bo10b2o12b3o105bobo$20b3o24b2o106bo$
19bo$19b2o$4b2o30b2o16b2o$5bo28b3ob2o14b2o$5bob2o6bo18bobob2o$6bo2bo8b
o13b2obo125bo$7b2o151bobo$12bo3bo2bo11bo128bo2bo$12b2o2bo145bo$12bo3bo
2bo$166b2o$18bobo101bo23b2o17bo2bo$15bo2bobo10bo3b2o86b2o22bo17bobo$
15b2o2b2o9bobo3bo85b2o23bobo16bo$15bo2bo10bobo3bo101bo10b2o12b3o$16b3o
6b2obobo3bo100b3o24b2o$25b2obo2b4obo5b2o90bo$29bobo3bobo5bo90b2o16b2o$
25b2ob2o2bo2bobo2b3o76b2o30bo2bo14b2o$26bobo2b2o3bo3bo13bo65bo30bo2bo
14b2o$14b2o10bobo26bo64bob2o26b5o$14b2o11bo25b3o65bo2bo22b2ob2o$122bob
o24bo$131b2o$124bobo2bo2bo$125bo2b2o4bo$128b2o6bo$128bo2b2ob3o$128bo3b
5o9bo3b2o$128bo16bobo3bo$132bo11bobo3bo$129bo2bo7b2obobo3bo$129b3o8b2o
bo2b4obo5b2o$144bobo3bobo5bo$140b2ob2o2bo2bobo2b3o$141bobo2b2o3bo3bo
13bo$129b2o10bobo26bo$129b2o11bo25b3o!
I have also been working on some lower period guns. I will try to put it all together later.

User avatar
Extrementhusiast
Posts: 1966
Joined: June 16th, 2009, 11:24 pm
Location: USA

Re: Execution of Old Guns by Variable-Speed Firing Squad

Post by Extrementhusiast » March 21st, 2015, 10:29 am

simsim314 wrote:I think there is no point to use the old collection up to 1000 now, as the collection is totally obsolete, except of maybe for the low periods.
What about for p30n, p46n, etc.?
I Like My Heisenburps! (and others)

User avatar
simsim314
Posts: 1823
Joined: February 10th, 2014, 1:27 pm

Re: Execution of Old Guns by Variable-Speed Firing Squad

Post by simsim314 » March 21st, 2015, 10:45 am

Extrementhusiast wrote:What about for p30n, p46n, etc.?
Well there are some interesting guns like p22 etc. but most of the high period above p100 are totally obsolete now.

Anyway I think we need a better gun collection that will include edge shooters, as this is the most common usage of guns.

User avatar
Freywa
Posts: 877
Joined: June 23rd, 2011, 3:20 am
Location: Singapore
Contact:

Re: Execution of Old Guns by Variable-Speed Firing Squad

Post by Freywa » March 21st, 2015, 10:47 am

chris_c wrote:I also accidentally discovered that the G->H can be overclocked at p74
Strange, very strange! The gadget also works at p75 (and the concomitant small glider duplicator of Grant, on the other thread) but not at p76 and p77: you get an LWSS and MWSS from them respectively! Oh, how much shrapnel they produce!
Princess of Science, Parcly Taxel

Code: Select all

x = 31, y = 5, rule = B2-a/S12
3bo23bo$2obo4bo13bo4bob2o$3bo4bo13bo4bo$2bo4bobo11bobo4bo$2bo25bo!

chris_c
Posts: 966
Joined: June 28th, 2014, 7:15 am

Re: Execution of Old Guns by Variable-Speed Firing Squad

Post by chris_c » March 21st, 2015, 10:52 am

Extrementhusiast wrote:
simsim314 wrote:I think there is no point to use the old collection up to 1000 now, as the collection is totally obsolete, except of maybe for the low periods.
What about for p30n, p46n, etc.?
Yes, I am writing a little script that compares areas of different gun families and also compares against the old collection. So far I have added families starting from p345, p358, p418 and p482. The p418 is from Kazyan's MunitionsDump. It is larger than the p482 family once the trombone is extended by 8 cells but it gives a nice p79 and a nice overclocked p75 that both appear to be record breakers:

Code: Select all

x = 213, y = 61, rule = B3/S23
31b2o11bo114b2o11bo$31b2o10bobo113b2o10bobo$43bobo2b2o3bo117bobo2b2o3b
o$42b2ob2o2bo2bobo115b2ob2o2bo2bobo$46bobo3bobo119bobo3bobo$42b2obo2b
4obo116b2obo2b4obo$42b2obobo3bo118b2obobo3bo$46bobo3bo15b2o104bobo3bo
15b2o$47bobo3bo14bo106bobo3bo14bo$48bo3b2o12bobo107bo3b2o12bobo$37bo
24b2o2b2o96b2o24b2o2b2o$35b2o25b2o103b2o21b2o$38bobo123bo2b3o$39b3o
123b2obobo$40bo125b2o2b2o$24b2o48b2o76b2o13b4o31b2o$23bo2bo47b2o7b2o
66bo2bo5b2o5b3o32b2o7b2o$22bob2o38bo18bo66bob2o5bobo49bo$22bo39b5o14bo
bo66bo10bo47bobo$21b2o6b2o32bo2bo14b2o66b2o39b3o16b2o$28bobo5b2o23b3o
2b2o96b2o22bob2obo$30bo5bo24b2o2b3o96bo27b3o$37b3o21bo103b3o27bo$39bo
10b2o9b2obo102bo10b2o14bo$49bobo10b3o112bobo13b2o$49bo13bo78b2o33bo11b
2o2bo$48b2o23bo68b2o32b2o12b3o8bo$72bobob2o6bo106bo8bobob2o6bo$11b2o
57b3obobo5bobo113b3obobo5bobo$11b2o56bo3bo4bo4b2o112bo3bo4bo4b2o$70bo
2bob4o63bo55bo2bob4o$62b2o7bobobo65b3o12b2o32b2o7bobobo$62b2o8bo3b3o
61bo2b2o11bo33b2o8bo3b3o$2o76bo60b2o13bobo49bo$bo23b2o112bo14b2o10bo$b
obo7bo13bo112bo27b3o$2b2o6b3o10bobo102b2o9b3o27bo$10bob2o9b2o10bo92b2o
10bob2obo22b2o$13bo21b3o103b3o39b2o$7b3o2b2o24bo5bo127bo10bo$7b2o2b3o
23b2o5bobo125bobo5b2obo$8bo2bo32b2o6b2o110b3o5b2o5bo2bo$8b5o39bo110b4o
13b2o$10bo38b2obo109b2o2b2o$48bo2bo111bobob2o$49b2o113b3o2bo$34bo107b
2o21b2o$33b3o102b2o2b2o24b2o$34bobo100bobo12b2o3bo$11b2o25b2o97bo14bo
3bobo$7b2o2b2o24bo98b2o15bo3bobo$6bobo12b2o3bo127bo3bobob2o$6bo14bo3bo
bo124bob4o2bob2o$5b2o15bo3bobo122bobo3bobo$23bo3bobob2o118bobo2bo2b2ob
2o$21bob4o2bob2o119bo3b2o2bobo$20bobo3bobo131bobo10b2o$20bobo2bo2b2ob
2o128bo11b2o$21bo3b2o2bobo$29bobo10b2o$30bo11b2o!
So far the new variants are appear to be the smallest gun in 274 cases but I'm sure a lot more is to come.

User avatar
Kazyan
Posts: 1247
Joined: February 6th, 2014, 11:02 pm

Re: Execution of Old Guns by Variable-Speed Firing Squad

Post by Kazyan » March 21st, 2015, 12:14 pm

Trombone-slide guns are period X+8n. We could get a gun family for X of all numbers modulo 8, to cover all the middling-period guns and their factors above 78. So far, we have X mod 8 = 0, 1, and 6.
Tanner Jacobi
Coldlander, a novel, available in paperback and as an ebook. Now on Amazon.

User avatar
dvgrn
Moderator
Posts: 10610
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI
Contact:

Re: Execution of Old Guns by Variable-Speed Firing Squad

Post by dvgrn » March 22nd, 2015, 1:34 am

Kazyan wrote:Trombone-slide guns are period X+8n. We could get a gun family for X of all numbers modulo 8, to cover all the middling-period guns and their factors above 78. So far, we have X mod 8 = 0, 1, and 6.
Belated congratulations on the new G-to-H injector/syringe! I've been traveling for the last couple of days, so it was quite a surprise to see all these new developments.

Here are some fairly reasonable 3 mod 8 and 5 mod 8 ouroborous loops. The version on the left has a longer recovery time due to the addition of the R64 -- it doesn't seem too likely that another eater can be welded in there to stop the FNG from the R64 output, but don't let me stop anyone. The R64 doesn't change the output glider color, but because of the 90-degree turn it does end up changing the mod-8 rating:

Code: Select all

#C [[ THEME 4 HEIGHT 240 AUTOSTART STEP 8 STOP 531 ]]
#C p531+8N and p557+8N bootstrapped glider loops
x = 214, y = 76, rule = B3/S23
177b2o3b2o$175b3obo2b2o$174bo4bo$47b2o3b2o120bo2b2ob4o$45b3obo2b2o119b
2obobobo2bo$44bo4bo124bobobobo$44bo2b2ob4o120bobob2o$43b2obobobo2bo
121bo$44bobobobo$44bobob2o138b2o$45bo133b2o7bo$179b2o5bobo$58b2o126b2o
$49b2o7bo$49b2o5bobo$56b2o4$176b2o$70b2o105bo$70b2o102b3o$46b2o126bo$
47bo$44b3o$44bo7$56b2o$55bobo$55bo153bo$54b2o135b2o14b5o$9bo25b2o11bo
87bo25b2o11bo16bo13bo5bo$9b3o23b2o10bobo86b3o23b2o10bobo15bobo12b3o2bo
$12bo34bobo2b2o3bo81bo34bobo2b2o3bo3bo4b2o15bob2o$11b2o33b2ob2o2bo2bob
o79b2o33b2ob2o2bo2bobo2b3o16b4o2bo$50bobo3bobo118bobo3bobo5bo10b2o3bo
3b2o$46b2obo2b4obo17b2o96b2obo2b4obo5b2o10b2o4b3o$2ob2o41b2obobo3bo19b
o51b2ob2o41b2obobo3bo27bo$2obo46bobo3bo19b3o48b2obo46bobo3bo26bob2o$3b
o47bobo3bo20bo51bo47bobo3bo24b2ob2o$3b3o4b2o40bo3b2o72b3o4b2o40bo3b2o$
b2o3bo3b2o116b2o3bo3b2o$o2b4o120bo2b4o67b2o$2obo15b2o106b2obo15b2o53bo
$bo2b3o12bobo17b2o2b2o83bo2b3o12bobo17b2o2b2o30b3o$bo5bo13bo3b2o11bobo
2b2o83bo5bo13bo3b2o11bobo2b2o32bo$2b5o14b2o2bo2b2o10bo88b5o14b2o2bo2b
2o10bo$4bo21b2o2bo100bo21b2o2bo$29bo126bo$26b3o124b3o46b2o$25bo126bo
49b2o$25b2o13b2o9b2o21b2o76b2o13b2o$40bo9bobo21b2o91bo$41b3o6bo17b2o
98b3o$43bo5b2o17b2o100bo10b2o$180bobo$180bo$70b2o107b2o$15b2o46b2o5b2o
70b2o$14bobo5b2o39b2o76bobo5b2o$14bo7b2o117bo7b2o$13b2o125b2o2$27bo
126bo$23b2obobo121b2obobo$22bobobobo120bobobobo$19bo2bobobob2o116bo2bo
bobob2o$19b4ob2o2bo117b4ob2o2bo$23bo4bo121bo4bo$19b2o2bob3o118b2o2bob
3o33b2o$19b2o3b2o120b2o3b2o35b2o!
EDIT: Here's a surprisingly compact p623 = 7 mod 8 bootstrapped loop. Certainly improves on the gun collection's p623 and p631 and so on (though the old p623 is a Herschel factory, so it should get a few extra points somehow):

Code: Select all

x = 97, y = 69, rule = B3/S23
21b2o$21b2o12b2o$35bo17b2o$33bobo17bo$33b2o19bo11bo$53b2o9b3o$63bo$63b
2o10b2o$75bo$15b2o56bobo$15b2o20bo35b2o13b2o$35b2o14bo36b2o$34b2o14b2o
$33bo$34b5o55b2o$35bo58b2o$16bo73b2o$9bo5bobo72b2o$9b3o4bo$12bo27b2o$
11b2o27bo18b2o$41b3o16bo34b2o$25b2o16bo13b3o35b2o$2ob2o20bobo29bo$2obo
23bo$3bo23b2o22b2o11bo$3b3o4b2o39b2o10bobo$b2o3bo3b2o51bobo2b2o3bo$o2b
4o55b2ob2o2bo2bobo$2obo15b2o45bobo3bobo$bo2b3o12bobo40b2obo2b4obo$bo5b
o13bo40b2obobo3bo$2b5o14b2o43bobo3bo$4bo62bobo3bo$68bo3b2o4$59b2o$59b
2o35bo$44b2o48bobo$43bo2bo48b2o$42bob2o$43bo$41bobo$40bobo13b2o9b2o21b
2o$41bo14bo9bobo21b2o$38b3o16b3o6bo17b2o$37bo14b2o5bo5b2o17b2o$37b2o
13b2o2$86b2o$79b2o5b2o$79b2o3$27b2o$26bobo5b2o$26bo7b2o$25b2o2$39bo$
35b2obobo$34bobobobo$31bo2bobobob2o$31b4ob2o2bo$35bo4bo$31b2o2bob3o$
31b2o3b2o!

User avatar
dvgrn
Moderator
Posts: 10610
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI
Contact:

Re: Execution of Old Guns by Variable-Speed Firing Squad

Post by dvgrn » March 22nd, 2015, 11:06 am

chris_c wrote:I am writing a little script that compares areas of different gun families and also compares against the old collection. So far I have added families starting from p345, p358, p418 and p482.
Once a gun family has been adjusted out far enough, it usually makes sense to rearrange the reflectors so that the gun is only growing in one direction instead of two (for a while, anyway). I assume that will count as a new family?

For example, the p623+8N above can be twisted into a p815+8N that's wider but not quite as tall -- and it won't have to grow any taller until after N=18 (at which point presumably it would be time to twist the loop again.)

Code: Select all

#C [[ VIEWONLY HEIGHT 350 GRID GRIDMAJOR 10 ZOOM 5 ]]
x = 120, y = 67, rule = LifeHistory
44.2A$43.B2A2B10.2A$44.4B10.A17.2A$42.7B6.BA.A17.A$17.2A3.2A18.8B.3B.
B2A19.A11.A$17.2A2.A.3A16.14B20.2A9.3A$21.A4.A15.14B20.B9.A$17.4A.2A
2.A15.13B19.3B9.2A10.2A$17.A2.A.A.A.2A11.2B.14B16.6B5.5B10.A$19.BABAB
A.A11.2A18B11.10B3.4B9.BA.A$20.B2ABA.A11.2AB.18BA2.2B3.11B2.6B4.3B.B
2A13.2A$21.2B.BA13.B2.16B2A14BA13B.7B14.B2AB$20.3B20.7B.6B2A14B2A21B
15.2B$11.2A6.4B19.8B.5BA38B17.2B3.B$12.A6.B2A3B17.7B2.6B5A34B15.4B.B
2A$12.A.AB3.B2A3B17.7B3.6BA39B11.8B2A$13.2AB.10B12.AB.6B11.4B.13B3.B.
19B2.2B2.6B2AB.B$15.13B10.A.A7B12.3B5.B.7B7.29B2A$15.14B10.A8B13.4B
19.6B.24B$15.15B10.B.6B15.2A17.5B.B2.28B$17.8B2.4B12.5B15.A18.2A7.28B
$17.6B5.4B10.6B16.3A16.A8.26B2A$16.9B4.4B8.5B.B2A16.A13.3A16.4B.14B2A
$15.4B4.2A5.4B6.4B3.BA.A29.A19.4B2.11B.B$14.4B5.A7.4B4.4B7.A50.4B2.
10B$4.A8.4B7.3A5.4B2.4B8.2A22.2A11.A14.14B$2.5A5.4B5.3A3.A5.8B32.B2AB
9.A.A14.14B$.A5.A4.4B5.A2.A.A7.6B34.3B9.A.A2.2A3.A6.13B$.A2.3AB2.7B.B
A.A.2A.2A7.4B34.B.B9.2A.2A2.A2.A.A5.13B$2A.A.2B3.7B.B2A.A11.6B33.5B8.
B2.A.A3.A.A5.8B.2B$A2.4A12B3.A10.8B32.6B6.2ABA2.4A.A6.8B$.2A2.BA3B2A
7B3.3A7.4B2.4B31.8B4.2A.A.A3.A8.9B$3.3AB.2B2A7B6.A5.4B4.4B31.13B2.A.A
B2.A7.10B$3.A.B3.10B5.2A4.4B6.4B28.13B5.A.A2B.A5.7B.4B$2A.A8.8B4.9B8.
4B26.15B5.A2B.2A4.8B2.4B$2A.2A7.9B5.6B10.4B25.15B4.3B7.8B4.4B$13.3B2.
4B2.8B11.4B23.17B.B.2B7.9B5.4B$11.5B3.15B10.4B22.32B.6B5.4B$11.2A7.
14B11.4B20.13B2A17B.7B6.4B$12.A8.13B12.4B18.14B2A16B2.6B8.3BA$9.3A10.
10B.B2A11.4B16.2AB3.27B.7B9.ABA$9.A14.3B2AB3.BA.A11.4B14.A2.A4.35B9.
2A$24.3B2AB6.A12.4B10.A.A.2A5.6B3.B2.2B2.19B10.B$26.4B6.2A12.4B9.2A.A
6.6B13.18B$26.3B22.4B11.A5.9B11.18B3.B$23.AB.2B24.4B7.3A5.4B4.2A9.2AB
3.15B.B2A$22.A.AB2AB24.4B5.A7.4B5.A9.A.AB2.18B2A$22.A.ABABAB24.4B4.2A
5.4B7.3A6.A5.12B2A3B.B$21.2A.A.A.A2.A23.9B4.4B3.2A5.A5.2A6.11B2A2B$
22.A2.2A.4A24.6B5.4B4.2A19.10B2.2B$22.A4.A28.8B2.4B28.B.6B3.2B$23.3A.
A2.2A22.15B33.4B2.B2AB$25.2A3.2A22.14B33.B2A2B3.2A$54.13B35.2A$52.2AB
.10B$51.A.AB3.B2A3B$51.A6.B2A3B$50.2A6.4B$59.3B$60.2B.BA$59.B2ABA.A$
58.BABABA.A$56.A2.A.A.A.2A$56.4A.2A2.A$60.A4.A$56.2A2.A.3A$56.2A3.2A!
This can be a bit of an annoyance for the derived lower-period guns with multiple signals per loop, because each such case has to be tested separately for signal collisions. There's often a way to adjust the reflectors to fix a given collision, but for multiple crossings it can get tricky.

With any luck we'll always be able to find shorter loops now, with relatively few signals per loop.

chris_c
Posts: 966
Joined: June 28th, 2014, 7:15 am

Re: Execution of Old Guns by Variable-Speed Firing Squad

Post by chris_c » March 22nd, 2015, 11:50 am

dvgrn wrote:Once a gun family has been adjusted out far enough, it usually makes sense to rearrange the reflectors so that the gun is only growing in one direction instead of two (for a while, anyway). I assume that will count as a new family?
Yeah that's going to be a slightly annoying problem. For example, I've added around 10 gun families so far and still the best p841 gun is a p345 with the trombone extended by 62 cells. That surely cannot be optimal. I will try to get the package tidied up a bit and put it up on github.

chris_c
Posts: 966
Joined: June 28th, 2014, 7:15 am

Re: Execution of Old Guns by Variable-Speed Firing Squad

Post by chris_c » March 22nd, 2015, 1:40 pm

I have put a first stab at my gun area scripts up on github here.

The directory named fixed is a dump of all the guns from Jason Summers gun collection. It doesn't contain any of the recent additions that Kazyan or simsim have been collecting.

The directory named variable contains variable glider guns that I have made or have scraped from the forum (guns by dvgrn and Kazyan I think). In this directory the number in the filename always represents the length of the glider loop and there is a suffix that indicates if any period multiplying method is used.

For example, the file p00482t.lif contains a period tripled p482 gun. At the moment its variants give the smallest gun in 29 cases below 1000. Other suffixes are "d" for doubled and "q" for quadrupled.

Code: Select all

#C compression = 78
x = 98, y = 50, rule = B3/S23
43bo11b2o$42bobo10b2o$34bo3b2o2bobo$33bobo2bo2b2ob2o$19b2o12bobo3bobo$
20bo13bob4o2b2o$20bobo13bo3bobob2o$21b2o12bo3bobo$34bo3bobo5bo$34b2o3b
o5bo$45b2o$47b5o$47b2o3$62b2o$bo59bo2bo$obo59b2obo$obo62bo$bo63b2o$50b
2o31b2o$51bo31bo$22b2o24b3o30bobo$22b2o11b2o2bo8bo28b2o2b2o$35bo3b3o
35b2o$36b3o3bo$29bo8bo2b2o11b2o$27b3o24b2o$26bo62b2o$26b2o61b2o$11b2o$
12bo80b2o$12bob2o77b2o2bo$13bo2bo$14b2o$29b2o$29b2o34b2o$64bobo$64bo$
63b2o32bo$38bo3b2o51bobo$37bobo3bo26b2o15b2o7b2o$36bobo3bo12b2o12bobo
16bo$32b2obobo3bo13bobo11bo15b3o$32b2obo2b4obo13bo10b2o7b2o6bo$36bobo
3bobo12b2o18b2o$32b2ob2o2bo2bobo$33bobo2b2o3bo$21b2o10bobo$21b2o11bo!
Also it is necessary to put a comment in the RLE that indicates the minimum compression of the glider loop. The comment line needs to contain the word "compression" and the number must be the last thing on the line and be preceded by a space.

The output of the (plain python) get_area script is shown in the README. At the moment the stats look as follows:

Code: Select all

*************
Stats
*************
fixed 461
p358 84
p345 62
p482d 43
p654 43
p531 41
p557 41
p482 35
p482t 29
p358d 25
p345d 24
p418 17
p592 13
p698 13
p698d 12
p652 12
p418d 9
p482q 8
p654d 5
p345q 4
p358q 3
p592d 2
variable 525
ie. the old fixed guns appear to be the smallest in 461 cases and the new variable guns are best in 525 cases. At the top of the README it gives what it thinks is the best gun for each period. The number after the underscore is the number of cells that the trombone needs to be lengthened by.

At the moment I wouldn't bet very much on these stats being 100% correct. Please post any inaccuracies, inefficiencies or additions in this thread and I will do my best to keep track of them.

User avatar
dvgrn
Moderator
Posts: 10610
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI
Contact:

Re: Execution of Old Guns by Variable-Speed Firing Squad

Post by dvgrn » March 22nd, 2015, 8:44 pm

chris_c wrote:At the moment I wouldn't bet very much on these stats being 100% correct. Please post any inaccuracies, inefficiencies or additions in this thread and I will do my best to keep track of them.
It looks as if your code currently overestimates the size of some variable guns:

Code: Select all

    for d in range(100):
        area = (x+d) * (y+d)
        for p in divisors(base + 8 * d, compression):
            add_gun(p * factor, area, "p%d%s_%d" % (base, "__dtq"[factor], d))
It would only make sense to adjust the p623+8d that I posted earlier from d=0..18, but the area formula would be just

area = (x+d)*y

Reflector adjustments can be kept inside the current bounding box height until d=19. Seems as if this kind of thing is likely to happen a lot... This formula could be continued for d=19+ by adding another twist to the loop, with some adjustment to the initial width. If a given gun family actually wins with a multi-signal loop, some testing would have to be done to make sure there's actually a configuration with no collisions. Usually that seems to work out okay in practice.

Would it work to add another optional item to the comment line? "grow = [H|W|HW]" or something like that?

If you did that, you'd probably want "max = 18" for the particular gun family I'm using as an example, since the formula stops working above that point. Technically I suppose it could be "grow = 0..18H, 19..100HW", but that syntax would be a little trickier to support.

I suspect that five or six signals in a loop may turn out to be optimal on rare occasions, possibly even higher. It may be worth adding "p" and "s" at least, or just switching to "x2", "x3"... "x6", which might be good to avoid confusion if any guns ever go as high as x7.

chris_c
Posts: 966
Joined: June 28th, 2014, 7:15 am

Re: Execution of Old Guns by Variable-Speed Firing Squad

Post by chris_c » March 22nd, 2015, 9:43 pm

dvgrn wrote: It looks as if your code currently overestimates the size of some variable guns:
OK I added the p623 to the collection. The commit is here. It improves another 28 guns compared to the old collection and is completely orthogonal to all the guns that I have put in already. But I'm not sure I agree that my formula is wrong in this case. Moving the trombone outwards will increase the bounding box in the west and the south so I think the formula is ok.

I agree that it is going to be a pain to deal with multi-bouncing Snark loops. Probably the best solution is to start a new family at the point where the extra Snarks become viable. The problem you mention with compression number and self-intersecting glider loops is something I have thought about but not found any good solutions to. I guess it will be too much trouble to write code that deals perfectly with all possible cases but hopefully my script can be used to pick off the low-hanging fruit and make the number of ad-hoc cases much smaller.
dvgrn wrote: I suspect that five or six signals in a loop may turn out to be optimal on rare occasions, possibly even higher. It may be worth adding "p" and "s" at least, or just switching to "x2", "x3"... "x6", which might be good to avoid confusion if any guns ever go as high as x7.
Maybe you are missing the point about the suffix letter? Adding more signals to the loop reduces the period but the suffix letters correspond to period multiplication, either with semi-snarks or a Herschel multiplicators that I found in a post of yours here. Regardless, your idea of using x2, x3, x4 etc is probably a good one.

My code adds as many signals to the loop provided that it doesn't take the repeat time below the minimum compression:

Code: Select all

def divisors(n, compression):
    for x in range(1, n // compression + 1):
        if n % x == 0:
            yield n // x
But number of signals, x, does not get written out into the gun description. Just laziness on my part.

EDIT: The highest number of signals I could find in a record breaking suggestion is 7 in this p82:

Code: Select all

x = 79, y = 81, rule = B3/S23
28bo11b2o$27bobo10b2o$19bo3b2o2bobo$10b2o6bobo2bo2b2ob2o$10b2o6bobo3bo
bo$19bob4o2bob2o$21bo3bobob2o$20bo3bobo$9b3o7bo3bobo$9bo2bo6b2o3bo$8bo
27bo$11b2o22b2o$7bo3b2o18b2o$7bo23bo$8b3o20b3o$47b2o$10b2o34bo2bo$10b
2o35b2obo$2o7bo2bo37bo$2o6b3o39b2o$7bob2obo22b2o5b3o$6b3o27bo5bo$5bo
27b3o7bo$6bo14b2o10bo$6b2o13bobo$7bo2b2o11bo$8b3o12b2o$9bo13$63b2o$63b
obo$63bo6$24bo44bo$25b2o41bobo$24b2o42bo2bo$70bo2$74b2o$54b2o17bo2bo$
55bo17bobo$55bobo16bo$45bo10b2o12b3o$43b3o24b2o$42bo$42b2o$27b2o48b2o$
28bo48b2o$28bob2o23b3ob3o$29bo2bo6b3o12b4obo2bo$30b2o6bo2bo11bobo$37bo
3bo20bo$37bo17b3o$37bo3bo14b2o2b2o$38bo2bo$39b3o$54bo3b2o$53bobo3bo$
52bobo3bo$48b2obobo3bo$48b2obo2b4obo5b2o$52bobo3bobo5bo$48b2ob2o2bo2bo
bo2b3o$49bobo2b2o3bo3bo13bo$37b2o10bobo26bo$37b2o11bo25b3o!
EDIT2: Oh sorry, there are 8 signals in this p80:

Code: Select all

x = 98, y = 66, rule = B3/S23
16b2o11bo$16b2o10bobo$28bobo2b2o3bo$27b2ob2o2bo2bobo$31bobo3bobo$27b2o
bo2b4obo$27b2obobo3bo$31bobo3bo$32bobo3bo$33bo3b2o47b2o$52bo33b2o$50b
3o13bo$22bo4b2o20bo16b3o$27bobo19b2o18bo$27bob2o37b2o$9b2o17b2o$8bo2bo
5bo10bo$7bob2o6b2o$7bo8bobo35bobo$6b2o46bo$21b2o51b3o$21bo31b2o3bo15bo
$22b3o28b2obobo16bo3b4o$24bo10b2o18bo26b2o$34bobo41bo2bob2o11b2o$34bo
43b2obobo12b2o$33b2o10b2o32bob2o$46bo33b2o$43b3o9b2o$43bo11bo19b2o$53b
obo18bobo$53b2o19bo$73b2o$23b2o$23bo19b2o$21bobo18bobo$21b2o19bo11bo$
41b2o9b3o$16b2o33bo$15b2obo32b2o10b2o$2o12bobob2o43bo$2o11b2obo2bo41bo
bo$14b2o26bo18b2o10bo$15b4o3bo16bobob2o28b3o$23bo15bo3b2o31bo$21b3o51b
2o$43bo46b2o$41bobo35bobo8bo$79b2o6b2obo$69bo10bo5bo2bo$68b2o17b2o$67b
2obo$47b2o19bobo$48bo20b2o4bo$45b3o$10b2o33bo$10b2o47b2o3bo$59bo3bobo$
60bo3bobo$61bo3bobob2o$59bob4o2bob2o$58bobo3bobo$58bobo2bo2b2ob2o$42bo
16bo3b2o2bobo$43bo23bobo10b2o$41b3o24bo11b2o!

User avatar
dvgrn
Moderator
Posts: 10610
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI
Contact:

Re: Execution of Old Guns by Variable-Speed Firing Squad

Post by dvgrn » March 22nd, 2015, 10:41 pm

chris_c wrote:I'm not sure I agree that my formula is wrong in this case. Moving the trombone outwards will increase the bounding box in the west and the south so I think the formula is ok.
Well, you're right of course. I was giving the name of one gun while looking at another. The horizontal-only increase applies to the p815+8N gun family, next message down from the p623+8N.

The p623+8N family is getting pretty big by the time it hits p823: the checkin says (11468, 'p623__25'), 122x94. The p815+8N family brings that down to (8040, 'p815__1'), 121x67.

The next gun improved in the last checkin was +839 (11904, 'p623__27'), 124x96 -- increasing X and Y bounds by two. But the p815+8N gun only has to go up to (8174, 'p815__3'), 123x67, increasing in the X direction only:

Code: Select all

#C p839 gun from the p815+8N family
x = 123, y = 67, rule = B3/S23
47b2o$17b2o3b2o23b2o12b2o$17b2o2bob3o35bo17b2o$21bo4bo32bobo17bo$17b4o
b2o2bo32b2o19bo11bo$17bo2bobobob2o51b2o9b3o$20bobobobo62bo$21b2obobo
62b2o10b2o$25bo75bo$41b2o56bobo$11b2o28b2o20bo35b2o13b2o$12bo7b2o39b2o
14bo36b2o$12bobo5b2o38b2o14b2o$13b2o44bo$60b5o55b2o$61bo58b2o$42bo73b
2o$41bobo72b2o$42bo$66b2o$23b2o41bo18b2o$23bo43b3o16bo34b2o$4bo19b3o
24b2o16bo13b3o35b2o$2b5o14b3o2bo24bobo29bo$bo5bo13bo2bo28bo$bo2b3o12bo
bo2b2o27b2o22b2o11bo$2obo15b2o56b2o10bobo$o2b4o82bobo2b2o3bo$b2o3bo3b
2o76b2ob2o2bo2bobo$3b3o4b2o80bobo3bobo$3bo21bo62b2obo2b4obo$2obo21b3o
60b2obobo3bo$2ob2o23bo63bobo3bo$27b2o64bobo3bo$94bo3b2o$11b2o$12bo$9b
3o$9bo75b2o$85b2o35bo$37b2o31b2o48bobo$30b2o5bobo29bo2bo48b2o$30b2o7bo
26bobob2o$39b2o25b2obo$69bo$26bo39b3o13b2o9b2o21b2o$25bobob2o34bo16bo
9bobo21b2o$25bobobobo33b2o16b3o6bo17b2o$24b2obobobo2bo43b2o5bo5b2o17b
2o$25bo2b2ob4o43b2o$25bo4bo$26b3obo2b2o77b2o$28b2o3b2o70b2o5b2o$105b2o
$55b2o$54bobo5b2o$54bo7b2o$53b2o2$67bo$63b2obobo$62bobobobo$59bo2bobob
ob2o$59b4ob2o2bo$63bo4bo$59b2o2bob3o$59b2o3b2o!
chris_c wrote:I guess it will be too much trouble to write code that deals perfectly with all possible cases but hopefully my script can be used to pick off the low-hanging fruit and make the number of ad-hoc cases much smaller.
Well, the ambitious thing to do might be to define the trombone-slide pieces separately from the main gun body, so that a script can actually assemble p623__2 or p815__3 or their lower-period variants if any, dropping in multiple signals at the right times. LifeHistory might be one reasonably simple encoding -- use state 1 for the main body, state 2 for a sample signal, and state 3 for the trombone slide, or some such.

Then most of the testing could be trivially automated -- run one cycle, and if the pattern is exactly the same as when it started except for an output glider, then you can declare success and move on. For the cases where something goes wrong, it's probably best not to try to handle it algorithmically -- just dump some RLE to an "Errors" folder.

The error pattern might show a correctly adjusted loop, but with just one signal in it. Then someone can look and see if it can be fixed. If not, maybe something has to go in the comments:

#C p903/7 = p129 not possible in this family: skip = 815__11/7

This will probably happen fairly rarely, because usually some other gun family will turn out to have a smaller p129 anyway, or the fixed gun will be smaller, or there will be some tricky custom way to fold the loop that solves the problem, that can be checked in as a fixed gun (?).

P.S. Yes, I was confused about the suffixes -- thanks for the explanation.

User avatar
dvgrn
Moderator
Posts: 10610
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI
Contact:

Re: Execution of Old Guns by Variable-Speed Firing Squad

Post by dvgrn » March 23rd, 2015, 1:37 am

Looks like you checked in the basic p531+8N and p557+8N guns . We can add a Snark loop to each, to get much more efficient guns for higher periods.

Code: Select all

#C p739+8N = (p531+8N) + loop
x = 112, y = 72, rule = LifeHistory
73.2A3.2A$71.3A.A2.2A$70.A4.A$70.A2.2A.4A$69.2A.A.A.A2.A$70.A.ABABAB$
70.A.AB2AB$71.AB.2B$74.3B$74.4B6.2A$72.3B2AB6.A$72.3B2AB3.BA.A$70.10B
.B2A$69.13B$68.14B$67.15B$66.4B2.8B15.2A$65.4B5.6B14.B2AB$64.4B4.9B
14.2B6.B$63.4B5.2A4.4B14.2B4.3B$62.4B7.A5.4B12.14B$61.4B5.3A7.4B11.B
3A10B$18.2A3.2A35.4B6.A10.B3A.B5.3BA3B2A4B2A2B$18.2A2.A.3A32.4B19.A
11BA5BA3BABA3B$22.A4.A30.4B20.BA10BA4BA4B3A2BABA$18.4A.2A2.A29.4B21.
15BA7BA5BA$18.A2.A.A.A.2A27.4B22.15BA9BA3BA$20.BABABA.A27.4B24.13BA
10BA2BAB$21.B2ABA.A26.4B23.2AB2.22BA$22.2B.BA26.4B23.A.AB4.B.4B2A10B$
21.3B28.4B24.A11.3BA6B.2B$12.2A6.4B27.4B24.2A12.9B$13.A6.B2A3B24.4B6.
2A11.A16.11B$13.A.AB3.B2A3B23.4B6.B2AB9.A.A14.12B$14.2AB.10B20.4B8.3B
9.A.A2.2A3.A6.12B$16.13B18.4B8.B.B9.2A.2A2.A2.A.A5.11B$16.14B16.4B9.
5B8.B2.A.A3.A.A5.8B.4B$16.15B14.4B10.6B6.2ABA2.4A.A6.7B4.2A$18.8B2.4B
12.4B11.8B4.2A.A.A3.A8.7B4.A$18.6B5.4B10.4B13.13B2.A.AB2.A7.6B6.3A$
17.9B4.4B8.4B12.13B5.A.A2B.A5.7B8.A$16.4B4.2A5.4B6.4B12.15B5.A2B.2A4.
8B$15.4B5.A7.4B4.4B13.15B4.3B7.8B$14.4B7.3A5.4B2.4B13.17B.B.2B7.9B$4.
A8.4B10.A6.8B14.32B.6B$2.5A5.4B5.2A12.6B14.13B2A17B.7B$.A5.A4.4B5.A2.
A11.4B10.2A2.14B2A16B2.6B$.A2.3AB2.7B.BA.A.2A10.6B9.A2.2AB3.27B.7B$2A
.A.2B3.7B.B2A.A11.8B9.2AB.A4.35B$A2.4A12B3.A10.4B2.4B11.A5.6B3.B2.2B
2.3BABA13B$.2A2.BA3B2A7B3.3A7.4B4.4B7.3A5.6B13.B2A15B$3.3AB.2B2A7B6.A
5.4B6.4B5.A7.9B11.2BA15B3.B$3.A.B3.10B5.2A4.4B8.4B4.2A5.4B4.2A9.2AB3.
15B.B2A$2A.A8.8B4.9B10.9B4.4B5.A9.A.AB2.18B2A$2A.2A7.9B5.6B12.6B5.4B
7.3A6.A5.12B2A3B.B$13.3B2.4B2.8B12.8B2.4B10.A5.2A6.11B2A2B$11.5B3.15B
8.15B25.10B2.2B$11.2A7.14B8.14B28.B.6B3.2B$12.A8.13B8.13B33.4B2.B2AB$
9.3A10.10B.B2A4.2AB.10B33.B2A2B3.2A$9.A14.3B2AB3.BA.A2.A.AB3.B2A3B36.
2A$24.3B2AB6.A2.A6.B2A3B$26.4B6.4A6.4B$26.3B18.3B$23.AB.2B10.2A8.2B.B
A$22.A.AB2AB9.2A7.B2ABA.A$22.A.ABABAB16.BABABA.A$21.2A.A.A.A2.A12.A2.
A.A.A.2A$22.A2.2A.4A12.4A.2A2.A$22.A4.A20.A4.A$23.3A.A2.2A12.2A2.A.3A
$25.2A3.2A12.2A3.2A!
#C [[ THUMBNAIL AUTOSTART STEP 8 THEME 4 STOP 1000 ]]

Code: Select all

#C p765+8N = (p557+8N) + loop
x = 112, y = 75, rule = LifeHistory
76.2A3.2A$74.3A.A2.2A$73.A4.A$73.A2.2A.4A$72.2A.A.A.A2.A$73.A.ABABAB$
73.A.AB2AB$74.AB.2B$77.3B$77.4B6.2A$75.3B2AB6.A$75.3B2AB3.BA.A$73.10B
.B2A$72.13B$71.14B$70.15B$69.4B2.8B$68.4B5.6B$67.4B4.9B$66.4B5.2A4.4B
$65.4B7.A5.4B$64.4B5.3A7.4B$63.4B6.A10.4B$62.4B19.4B$61.4B21.4B$18.2A
3.2A35.4B23.4B$18.2A2.A.3A32.4B25.4B$22.A4.A30.4B27.4B$18.4A.2A2.A29.
4B29.4B$18.A2.A.A.A.2A27.4B31.4B$20.BABABA.A27.4B33.4B$21.B2ABA.A26.
4B35.4B$22.2B.BA26.4B37.4B$21.3B28.4B39.4B8.A$12.2A6.4B27.4B34.2A5.4B
5.5A$13.A6.B2A3B24.4B6.2A11.A16.A5.4B4.A5.A$13.A.AB3.B2A3B23.4B6.B2AB
9.A.A15.A.AB.2BA4B2.B3A2.A$14.2AB.10B20.4B8.3B9.A.A2.2A3.A3.A4.2AB.BA
BA3B3.2B.A.2A$16.13B18.4B8.B.B9.2A.2A2.A2.A.A2.3A4.2BA3BA5B4A2.A$16.
14B16.4B9.5B8.B2.A.A3.A.A5.A3.2BA3BA5BAB2.2A$16.15B14.4B10.6B6.2ABA2.
4A.A5.2A.4BA3BA4B.B3A$18.8B2.4B12.4B11.8B4.2A.A.A3.A7.8BABA4B3.B.A$
18.6B5.4B10.4B13.13B2.A.AB2.A8.7BA2B8.A.2A$17.9B4.4B8.4B12.13B5.A.A2B
.A7.10B7.2A.2A$16.4B4.2A5.4B6.4B12.15B5.A2B.2A6.10B$15.4B5.A7.4B4.4B
13.15B4.3B10.6B.5B$14.4B7.3A5.4B2.4B13.17B.B.2B10.7B4.2A$4.A8.4B10.A
6.8B14.29B4.6B4.A$2.5A5.4B5.2A12.6B14.13B2A16B2.7B5.3A$.A5.A4.4B5.A2.
A11.4B10.2A2.14B2A16B2.7B7.A$.A2.3AB2.7B.BA.A.2A10.6B9.A2.2AB3.27B.4B
A4B$2A.A.2B3.7B.B2A.A11.8B9.2AB.A4.29BABA4B$A2.4A12B3.A10.4B2.4B11.A
5.6B3.B2.2B2.13BA6B4.B$.2A2.BA3B2A7B3.3A7.4B4.4B7.3A5.6B14.19B.B2A$3.
3AB.2B2A7B6.A5.4B6.4B5.A7.9B14.19B2A$3.A.B3.10B5.2A4.4B8.4B4.2A5.4B4.
2A15.17B.B$2A.A8.8B4.9B10.9B4.4B5.A15.17B$2A.2A7.9B5.6B12.6B5.4B7.3A
12.12B2A3B$13.3B2.4B2.8B12.8B2.4B10.A10.2AB.9BA2BA2B$11.5B3.15B8.15B
21.A.AB4.11BA$11.2A7.14B8.14B22.A9.9BAB$12.A8.13B8.13B22.2A8.2B2A7BAB
$9.3A10.10B.B2A4.2AB.10B33.BA9BA$9.A14.3B2AB3.BA.A2.A.AB3.B2A3B36.9BA
$24.3B2AB6.A2.A6.B2A3B37.5BA3B$26.4B6.4A6.4B39.A2B2A4B$26.3B18.3B39.B
ABA5B$23.AB.2B10.2A8.2B.BA37.5B$22.A.AB2AB9.2A7.B2ABA.A36.4B$22.A.ABA
BAB16.BABABA.A36.4B$21.2A.A.A.A2.A12.A2.A.A.A.2A34.4B$22.A2.2A.4A12.
4A.2A2.A34.4B$22.A4.A20.A4.A33.A3B$23.3A.A2.2A12.2A2.A.3A33.A3B$25.2A
3.2A12.2A3.2A34.B3A!
#C [[ THUMBNAIL AUTOSTART STEP 8 THEME 4 STOP 1000 ]]
These both would use area = (x+d)*y for quite a while, up to d=22 and d=24 I believe. So it should be possible to cut over a third off the bounding box of some of those biggest ugliest odd-period p900+ guns. Don't know about lower ranges and multiplied and divided periods, but a few improvements might show up there too.

For example, here's a p997 = p765+8N, N=29. 29 is a little beyond the horizontal-only upper limit of 24, so the height has gone up by 5 as well:

Code: Select all

x = 142, y = 80, rule = LifeHistory
18.2A3.2A$18.2A2.A.3A$22.A4.A$18.4A.2A2.A$18.A2.A.A.A.2A$20.BABABA.A
78.2A3.2A$21.B2ABA.A76.3A.A2.2A$22.2B.BA76.A4.A$21.3B79.A2.2A.4A$12.
2A6.4B78.2A.A.A.A2.A$13.A6.B2A3B77.A.ABABAB$13.A.AB3.B2A3B77.A.AB2AB$
14.2AB.10B76.AB.2B$16.13B78.3B$16.14B77.4B6.2A$16.15B74.3B2AB6.A$18.
8B2.4B73.3B2AB3.BA.A$18.6B5.4B70.10B.B2A$17.9B4.4B68.13B$16.4B4.2A5.
4B66.14B$15.4B5.A7.4B64.15B$14.4B7.3A5.4B62.4B2.8B$4.A8.4B10.A6.4B60.
4B5.6B$2.5A5.4B5.2A12.4B58.4B4.9B$.A5.A4.4B5.A14.4B56.4B5.2A4.4B$.A2.
3AB2.7B.BA.A15.4B54.4B7.A5.4B$2A.A.2B3.7B.B2A17.4B52.4B5.3A7.4B$A2.4A
12B20.4B50.4B6.A10.4B$.2A2.BA3B2A7B21.4B48.4B19.4B$3.3AB.2B2A7B22.4B
46.4B21.4B$3.A.B3.10B23.4B44.4B23.4B$2A.A8.8B23.4B42.4B25.4B$2A.2A7.
9B23.4B40.4B27.4B$13.3B2.4B23.4B38.4B29.4B$11.5B3.4B23.4B36.4B31.4B$
11.2A7.4B23.4B34.4B33.4B$12.A8.4B23.4B32.4B35.4B$9.3A10.4B23.4B30.4B
37.4B$9.A13.4B23.4B28.4B39.4B8.A$24.4B23.4B26.4B34.2A5.4B5.5A$25.4B
23.4B24.4B6.2A11.A16.A5.4B4.A5.A$26.4B23.4B22.4B6.B2AB9.A.A15.A.AB.2B
A4B2.B3A2.A$27.4B23.4B20.4B8.3B9.A.A2.2A3.A3.A4.2AB.BABA3B3.2B.A.2A$
28.4B23.4B18.4B8.B.B9.2A.2A2.A2.A.A2.3A4.2BA3BA5B4A2.A$29.4B23.4B16.
4B9.5B8.B2.A.A3.A.A5.A3.2BA3BA5BAB2.2A$30.4B23.4B14.4B10.6B6.2ABA2.4A
.A5.2A.4BA3BA4B.B3A$31.4B23.4B12.4B11.8B4.2A.A.A3.A7.8BABA4B3.B.A$32.
4B23.4B10.4B13.13B2.A.AB2.A8.7BA2B8.A.2A$33.4B23.4B8.4B12.13B5.A.A2B.
A7.10B7.2A.2A$34.4B23.4B6.4B12.15B5.A2B.2A6.10B$35.4B23.4B4.4B13.15B
4.3B10.6B.5B$36.4B23.4B2.4B13.17B.B.2B10.7B4.2A$37.4B23.8B14.29B4.6B
4.A$38.4B23.6B14.13B2A16B2.7B5.3A$39.4B23.4B10.2A2.14B2A16B2.7B7.A$
40.4B21.6B9.A2.2AB3.27B.4BA4B$41.4B19.8B9.2AB.A4.29BABA4B$42.4B6.A10.
4B2.4B11.A5.6B3.B2.2B2.13BA6B4.B$43.4B5.3A7.4B4.4B7.3A5.6B14.19B.B2A$
44.4B7.A5.4B6.4B5.A7.9B14.19B2A$45.4B5.2A4.4B8.4B4.2A5.4B4.2A15.17B.B
$46.4B4.9B10.9B4.4B5.A15.17B$47.4B5.6B12.6B5.4B7.3A12.12B2A3B$48.4B2.
8B12.8B2.4B10.A10.2AB.9BA2BA2B$49.15B8.15B21.A.AB4.11BA$50.14B8.14B
22.A9.9BAB$51.13B8.13B22.2A8.2B2A7BAB$52.10B.B2A4.2AB.10B33.BA9BA$54.
3B2AB3.BA.A2.A.AB3.B2A3B36.9BA$54.3B2AB6.A2.A6.B2A3B37.5BA3B$56.4B6.
4A6.4B39.A2B2A4B$56.3B18.3B39.BABA5B$53.AB.2B10.2A8.2B.BA37.5B$52.A.A
B2AB9.2A7.B2ABA.A36.4B$52.A.ABABAB16.BABABA.A36.4B$51.2A.A.A.A2.A12.A
2.A.A.A.2A34.4B$52.A2.2A.4A12.4A.2A2.A34.4B$52.A4.A20.A4.A33.A3B$53.
3A.A2.2A12.2A2.A.3A33.A3B$55.2A3.2A12.2A3.2A34.B3A!
#C [[ THUMBNAIL AUTOSTART STEP 8 THEME 4 STOP 1000 ]]
But then of course there's some empty space in the north, so we can do better again for p981+8N and divisors:

Code: Select all

#C p981+8N = (p557+8N) + two loops
x = 112, y = 76, rule = LifeHistory
27.2A3.2A$25.3A.A2.2A42.2A3.2A$24.A4.A44.3A.A2.2A$24.A2.2A.4A39.A4.A$
23.2A.A.A.A2.A39.A2.2A.4A$24.A.ABABAB40.2A.A.A.A2.A$24.A.AB2AB42.A.AB
ABAB$25.AB.2B43.A.AB2AB$28.3B43.AB.2B$28.4B6.2A37.3B$26.3B2AB6.A38.4B
6.2A$26.3B2AB3.BA.A36.3B2AB6.A$10.A13.10B.B2A37.3B2AB3.BA.A$10.3A10.
13B37.10B.B2A$13.A8.14B36.13B$12.2A7.15B35.14B$12.5B3.4B2.8B36.15B$
14.3B2.4B5.6B35.4B2.8B$.2A.2A7.9B4.9B33.4B5.6B$.2A.A8.8B5.2A4.4B31.4B
4.9B$4.A.B3.10B7.A5.4B29.4B5.2A4.4B$4.3AB.2B2A7B4.3A7.4B8.A18.4B7.A5.
4B$2.2A2.BA3B2A7B4.A2.3A5.4B5.5A15.4B5.3A7.4B$.A2.4A12B6.A2.A5.4B4.A
5.A13.4B6.A10.4B$.2A.A.2B3.7B.B2A3.2A2.A.AB.7B2.B3A2.A12.4B19.4B$2.A
2.3AB2.7B.BA.A7.2AB.7B3.2B.A.2A10.4B21.4B$2.A5.A4.4B5.A9.12B4A2.A9.4B
23.4B$3.5A5.4B5.2A8.7B2A3BAB2.2A9.4B25.4B$5.A8.4B14.7B2A2B.B3A10.4B
27.4B$15.4B13.10B3.B.A9.4B29.4B$16.4B11.8B8.A.2A5.4B31.4B$17.4B9.9B7.
2A.2A4.4B33.4B$18.4B7.4B2.3B16.4B35.4B$19.4B5.4B3.5B13.4B37.4B$20.4B
3.4B7.2A12.4B39.4B8.A$21.4B.4B8.A12.4B34.2A5.4B5.5A$22.7B10.3A8.4B6.
2A11.A16.A5.4B4.A5.A$23.5B13.A7.4B6.B2AB9.A.A15.A.AB.2BA4B2.B3A2.A$
23.5B20.4B8.3B9.A.A2.2A3.A3.A4.2AB.BABA3B3.2B.A.2A$22.7B18.4B8.B.B9.
2A.2A2.A2.A.A2.3A4.2BA3BA5B4A2.A$21.4B.4B16.4B9.5B8.B2.A.A3.A.A5.A3.
2BA3BA5BAB2.2A$20.4B3.4B14.4B10.6B6.2ABA2.4A.A5.2A.4BA3BA4B.B3A$19.4B
5.4B12.4B11.8B4.2A.A.A3.A7.8BABA4B3.B.A$18.4B7.4B10.4B13.13B2.A.AB2.A
8.7BA2B8.A.2A$17.4B9.4B8.4B12.13B5.A.A2B.A7.10B7.2A.2A$16.4B11.4B6.4B
12.15B5.A2B.2A6.10B$15.4B13.4B4.4B13.15B4.3B10.6B.5B$14.4B15.4B2.4B
13.17B.B.2B10.7B4.2A$4.A8.4B17.8B14.29B4.6B4.A$2.5A5.4B5.2A12.6B14.
13B2A16B2.7B5.3A$.A5.A4.4B5.A2.A11.4B10.2A2.14B2A16B2.7B7.A$.A2.3AB2.
7B.BA.A.2A10.6B9.A2.2AB3.27B.4BA4B$2A.A.2B3.7B.B2A.A11.8B9.2AB.A4.29B
ABA4B$A2.4A12B3.A10.4B2.4B11.A5.6B3.B2.2B2.13BA6B4.B$.2A2.BA3B2A7B3.
3A7.4B4.4B7.3A5.6B14.19B.B2A$3.3AB.2B2A7B6.A5.4B6.4B5.A7.9B14.19B2A$
3.A.B3.10B5.2A4.4B8.4B4.2A5.4B4.2A15.17B.B$2A.A8.8B4.9B10.9B4.4B5.A
15.17B$2A.2A7.9B5.6B12.6B5.4B7.3A12.12B2A3B$13.3B2.4B2.8B12.8B2.4B10.
A10.2AB.9BA2BA2B$11.5B3.15B8.15B21.A.AB4.11BA$11.2A7.14B8.14B22.A9.9B
AB$12.A8.13B8.13B22.2A8.2B2A7BAB$9.3A10.10B.B2A4.2AB.10B33.BA9BA$9.A
14.3B2AB3.BA.A2.A.AB3.B2A3B36.9BA$24.3B2AB6.A2.A6.B2A3B37.5BA3B$26.4B
6.4A6.4B39.A2B2A4B$26.3B18.3B39.BABA5B$23.AB.2B10.2A8.2B.BA37.5B$22.A
.AB2AB9.2A7.B2ABA.A36.4B$22.A.ABABAB16.BABABA.A36.4B$21.2A.A.A.A2.A
12.A2.A.A.A.2A34.4B$22.A2.2A.4A12.4A.2A2.A34.4B$22.A4.A20.A4.A33.A3B$
23.3A.A2.2A12.2A2.A.3A33.A3B$25.2A3.2A12.2A3.2A34.B3A!
#C [[ THUMBNAIL AUTOSTART STEP 8 THEME 4 STOP 1000 ]]
For this p981, the available half-price adjustment has mysteriously moved to the trombone slide in the northeast. For double loops, in general you get two +8 adjustments for each increase in height, so the bounding box formula gets a little trickier again.

Even so, at some point you get a better deal by replacing the leftmost Snark with a loop -- instead of, or in addition to, the Snark replacement in the northwest. One of the most infuriating things about bounding box optimization is the sad fact that (X+1)(X-1)<X^2... which means that you generally make adjustments by adding new columns, if you can, and avoid adding new rows.

chris_c
Posts: 966
Joined: June 28th, 2014, 7:15 am

Re: Execution of Old Guns by Variable-Speed Firing Squad

Post by chris_c » March 23rd, 2015, 8:41 am

dvgrn wrote:We can add a Snark loop to each, to get much more efficient guns for higher periods.

Code: Select all

#C p739+8N = (p531+8N) + loop

Code: Select all

#C p765+8N = (p557+8N) + loop
I added three new optional fields to the variable gun descriptions: x_slack, y_slack and tentative_comression. For example, the p739 has been checked in as:

Code: Select all

#C compression = 739
#C tentative_compression = 149
#C y_slack = 22
x = 112, y = 72, rule = B3/S23
73b2o3b2o$71b3obo2b2o$70bo4bo$70bo2b2ob4o$69b2obobobo2bo$70bobobobo$
70bobob2o$71bo2$84b2o$75b2o7bo$75b2o5bobo$82b2o4$95b2o$95b2o2$72b2o$
73bo$70b3o23b3o$18b2o3b2o45bo11b3o10bo3b2o4b2o$18b2o2bob3o55bo11bo5bo
3bobo$22bo4bo55bo10bo4bo4b3o2bobo$18b4ob2o2bo69bo7bo5bo$18bo2bobobob2o
68bo9bo3bo$21bobobobo68bo10bo2bo$22b2obobo53b2o25bo$26bo53bobo11b2o$
80bo14bo$12b2o65b2o$13bo7b2o37b2o11bo$13bobo5b2o37b2o10bobo$14b2o56bob
o2b2o3bo$71b2ob2o2bo2bobo$75bobo3bobo$71b2obo2b4obo17b2o$71b2obobo3bo
19bo$75bobo3bo19b3o$76bobo3bo20bo$24b2o51bo3b2o$24bo$25b3o$4bo22bo$2b
5o14b2o45b2o$bo5bo13bo2bo25b2o16b2o$bo2b3o12bobob2o25bo2b2o$2obo15b2ob
o28b2o2bo$o2b4o15bo31bo24bobo$b2o3bo3b2o10b3o26b3o25b2o$3b3o4b2o13bo
24bo29bo$3bo20b2o24b2o13b2o9b2o21b2o$2obo61bo9bobo21b2o$2ob2o61b3o6bo
17b2o$68bo5b2o17b2o2$11b2o$12bo82b2o$9b3o22b2o4b2o46b2o5b2o$9bo17b2o5b
obo2bobo5b2o39b2o$27b2o7bo2bo7b2o$36b4o2$23bo14b2o12bo$22bobob2o10b2o
8b2obobo$22bobobobo18bobobobo$21b2obobobo2bo12bo2bobobob2o$22bo2b2ob4o
12b4ob2o2bo$22bo4bo20bo4bo$23b3obo2b2o12b2o2bob3o$25b2o3b2o12b2o3b2o!
The new formula for the area is:

Code: Select all

        new_x = x + max(d - x_slack, 0)
        new_y = y + max(d - y_slack, 0)
        area = new_x * new_y
The tentative_compression field is used for low periods that are compatible with the core circuitry but may not work because of glider collisions.

The p739 improves 7 guns from the old collection and 12 guns from the p531 family. Also it gives two record breaking sub period guns at p151 and p159 both of which are marked as "tentative" by the script but do in fact work:

Code: Select all

x = 312, y = 72, rule = B3/S23
80b2o3b2o186b2o3b2o$78b3obo2b2o184b3obo2b2o$77bo4bo187bo4bo$77bo2b2ob
4o183bo2b2ob4o$76b2obobobo2bo182b2obobobo2bo$77bobobobo186bobobobo$77b
obob2o187bobob2o$78bo192bo2$91b2o191b2o$82b2o7bo183b2o7bo$82b2o5bobo
183b2o5bobo$89b2o191b2o3$18b2o3b2o$18b2o2bob3o75b2o191b2o$22bo4bo74b2o
191b2o$18b4ob2o2bo$18bo2bobobob2o50b2o191b2o$21bobobobo52bo135b2o3b2o
50bo$22b2obobo49b3o23b3o110b2o2bob3o45b3o23b3o$26bo50bo11b3o10bo3b2o4b
2o106bo4bo44bo11b3o10bo3b2o4b2o$89bo11bo5bo3bobo102b4ob2o2bo56bo11bo5b
o3bobo$12b2o76bo10bo4bo4b3o2bobo97bo2bobobob2o56bo10bo4bo4b3o2bobo$13b
o7b2o81bo7bo5bo100bobobobo71bo7bo5bo$13bobo5b2o81bo9bo3bo101b2obobo71b
o9bo3bo$14b2o87bo10bo2bo106bo71bo10bo2bo$88b2o25bo165b2o25bo$87bobo11b
2o107b2o68bobo11b2o$87bo14bo108bo7b2o59bo14bo$86b2o123bobo5b2o58b2o$
67b2o11bo131b2o46b2o11bo$67b2o10bobo178b2o10bobo$24b2o53bobo2b2o3bo
182bobo2b2o3bo$24bo53b2ob2o2bo2bobo180b2ob2o2bo2bobo$25b3o54bobo3bobo
156bo27bobo3bobo$4bo22bo50b2obo2b4obo17b2o136b2o24b2obo2b4obo17b2o$2b
5o14b2o29bo25b2obobo3bo19bo138b2o23b2obobo3bo19bo$bo5bo13bo28b2o30bobo
3bo19b3o111b2o51bobo3bo19b3o$bo2b3o12bobo29b2o30bobo3bo20bo111bo53bobo
3bo20bo$2obo15b2o63bo3b2o133b3o51bo3b2o$o2b4o7b3o55b2o128bo22bo$b2o3bo
5b2ob2o55b2o126b5o14b2o50b2o$3b3o8bo58b5o121bo5bo13bo50b5o$3bo70bo3bo
120bo2b3o7bo4bobo50bo4bo$2obo53b2o19bo119b2obo9bobo3b2o31b2o18b3o2bo$
2ob2o52bo2b2o12b2ob2o119bo2b4o5bo3bo35bo2b2o16bo2b2o$58b2o2bo13bo122b
2o3bo5bo3bo36b2o2bo16b2o$29bo17bobo11bo24bobo112b3o6bo3bo7bo31bo24bobo
$11b2o16b3o16b2o8b3o25b2o113bo9bobo8b3o26b3o25b2o$12bo19bo15bo8bo29bo
110b2obo10bo12bo24bo29bo$9b3o19b2o24b2o13b2o9b2o21b2o90b2ob2o21b2o24b
2o13b2o9b2o21b2o$9bo62bo9bobo21b2o136bobo18bo9bobo21b2o$73b3o6bo17b2o
143b2o19b3o6bo17b2o$75bo5b2o17b2o107b2o34bo22bo5b2o17b2o$210bo$207b3o$
102b2o103bo87b2o$41b2o4b2o46b2o5b2o130b2o4b2o46b2o5b2o$34b2o5bobo2bobo
5b2o39b2o130b2o5bobo2bobo5b2o39b2o$34b2o7bo2bo7b2o171b2o7bo2bo7b2o$43b
4o189b4o2$30bo14b2o12bo163bo14b2o12bo$29bobob2o10b2o8b2obobo161bobob2o
10b2o8b2obobo$29bobobobo18bobobobo161bobobobo18bobobobo$28b2obobobo2bo
12bo2bobobob2o159b2obobobo2bo12bo2bobobob2o$29bo2b2ob4o12b4ob2o2bo161b
o2b2ob4o12b4ob2o2bo$29bo4bo20bo4bo161bo4bo20bo4bo$30b3obo2b2o12b2o2bob
3o163b3obo2b2o12b2o2bob3o$32b2o3b2o12b2o3b2o167b2o3b2o12b2o3b2o!
The p765 beats 12 guns from the old collection and 11 guns from the p557 family. It doesn't seem to offer any record breakers at subperiods. That brings the total number of record breaking guns to a speculative total of 573.

I have not added your double Snark twist p981 gun yet. I would like to delay thinking about that for at least another couple of days if possible! In particular the new variable guns have not produced anything stellar in the p8N category. What magic can you come up with by combining the new G->H and old p8 technology?

EDIT: Ooops I missed your second to last message. Let me get back to that stuff in a while. Safe to say that you spotted some other guns where the x_slack/y_slack thing will come in useful...

EDIT2: Added the p815 with 18 units of y_slack here. It beats 5 more of the old guns and 9 from the p623 family. It offers no new tentative guns so it seems that my less ambitious tentative_compression hack is working quite well at the moment.

EDIT3: Oh, the p765 is already double adjustable so the problem of area overestimation already exists at the moment, even before I think about adding the p981 :(

EDIT4: I experienced some tentative guns that appear to be nonviable. See here for details.

EDIT5: Added some more guns. p363 and p441 are nice and small. The p595, p673 and p702 are variants of existing guns with extra Snarks added. The total of records in the variable gun collection is over 600 but I would love to find a glider loop of period 4N that is below 500 ticks in length. No luck so far.

User avatar
dvgrn
Moderator
Posts: 10610
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI
Contact:

Re: Execution of Old Guns by Variable-Speed Firing Squad

Post by dvgrn » March 24th, 2015, 3:20 am

chris_c wrote:EDIT4: I experienced some tentative guns that appear to be nonviable. See here for details.
Let's see if I'm understanding this well enough. Starting with the first tentative item, the p117... yes, it does look not quite viable as it stands. However, there's actually just enough adjustability in the size of the added loop to get p117 gliders to squeak past each other, so all we need is a quick game of weld-the-tail-on-the-beehive to be back in business:

Code: Select all

#C [[ THUMBNAIL AUTOSTART STOP 117 ZOOM 5.5 ]]
x = 101, y = 59, rule = LifeHistory
47.2A11.A$46.B2AB9.A.A$47.3B9.A.A2.2A3.A$46.B.B9.2A.2A2.A2.A.A$46.5B
8.B2.A.A3.A.A$28.2A16.6B6.2ABA2.4A.A$27.A.A16.8B4.2A.A.A3.A$21.2A4.A
19.13B2.A.AB2.A15.2A$19.A2.A2.2A.4A13.13B5.A.A2B.A14.A$19.2A.A.A.A.A
2.A12.15B5.A2B.2A11.BA.A$22.A.ABABAB14.15B4.3B12.2A.B2A$22.A.AB2AB14.
17B.B.2B12.B2A2B$23.AB.2B15.23B2A2B2A6.4B$26.3B13.13B2A9B2A2B2AB3.6B$
26.4B6.2A3.14B2A7BA6B2A2.6B$24.3B2AB6.A3.2AB3.17BA7BAB.7B9.2A$9.A14.
3B2AB3.BA.A2.A.BA4.17B2AB2AB2A10B6.2B2AB6.2A$9.3A10.10B.B2A4.2A5.6B3.
B2.2B2.3BA3B2A11B5.4B7.A$12.A8.13B12.A5B14.21B.6B2.BA.A$11.2A7.14B11.
B2A6B14.27B.B2A$11.5B3.15B10.BABA4.2A15.28B$13.3B2.4B2.8B11.4B5.A15.
29B$3.2A7.9B5.6B10.4B7.3A12.29B$3.A8.8B4.9B8.4B10.A10.2AB.3B2.22B$2A.
A.B3.10B5.2A4.2B2A6.4B21.A.AB6.13B.9B$A2.3AB.2B2A7B6.A5.2A2B4.4B22.A
8.12B2.11B$.2A2.BA3B2A7B3.3A7.BA2B2.4B22.2A7.12B3.BAB.B3.4B$3.4A12B3.
A10.8B31.10B6.A.A.2A3.3BA$3.A.2B3.7B.B2A.A11.6B31.4B.6B4.3A.ABA5.ABA$
4.3AB2.7B.BA.A.2A.2A7.4B31.4B3.B.B5.A3.A4.A4.2A$7.A4.4B5.A2.A.A7.6B
29.4B5.3B4.2A2.A.4A5.B$2.5A5.4B5.3A3.A5.8B27.4B5.B2AB9.A$2.A10.4B7.3A
5.4B2.4B25.4B7.2A12.A$4.A9.4B5.A7.4B4.4B23.4B21.2A$3.2A10.3BA4.2A5.4B
6.4B21.4B$16.3BA5B4.4B8.4B19.4B$17.3A3B5.4B10.4B6.A10.4B$17.8B2.4B12.
4B5.3A7.4B$15.15B14.4B7.A5.4B$15.14B16.4B5.2A4.A3B$15.13B18.4B4.6BABA
$13.2AB.10B20.4B5.4B2A$12.A.AB3.B2A3B23.4B2.8B$12.A6.B2A3B24.15B$11.
2A6.4B27.14B$20.3B28.13B$21.2B.BA26.10B.B2A$20.B2ABA.A27.3B2AB3.BA.A$
19.BABABA.A27.3B2AB6.A$17.A2.A.A.A.A.2A26.4B6.2A$17.4A.2A2.A2.A26.3B$
21.A4.2A25.AB.2B$19.A.A30.A.AB2AB$19.2A31.A.ABABAB$49.2A.A.A.A.A2.A$
49.A2.A2.2A.4A$51.2A4.A$57.A.A$58.2A!
If that's an unworkable weld, then we can relocate the loop to get another cousin in this gun family -- bigger than the above version, certainly, but still a big improvement on p345__30:

Code: Select all

#C 5-signal p117 gun, bounding box 84x76
#C [[ THUMBNAIL AUTOSTART Y 14 ZOOM 5.5 ]]
x = 84, y = 101, rule = LifeHistory
33.2A$32.A.A$26.2A4.A$24.A2.A2.2A.4A$24.2A.A.A.A.A2.A$27.A.ABABAB19.
2A$27.A.AB2AB20.A.A$28.AB.2B23.A4.2A$31.3B18.4A.2A2.A2.A$31.4B6.2A9.A
2.A.A.A.A.2A$29.3B2AB6.A12.BABABA.A$29.3B2AB3.BA.A13.B2ABA.A$19.A7.
10B.B2A15.2B.BA$17.3A6.13B16.3B$16.A8.14B7.2A6.4B$15.A.A6.15B8.A6.B2A
3B$14.BA.A5.4B2.8B10.A.AB3.B2A3B14.A$13.3BA5.4B5.6B11.2AB.10B10.3A$
11.4B6.4B4.9B12.13B8.A$9.6B5.4B5.2A4.4B11.14B7.2A$8.7B4.BA2B7.A5.4B
10.15B3.5B$.B.4B.8B2.2A2B5.3A7.4B11.8B2.4B2.3B$2AB.15B2A6.A10.4B10.6B
5.9B7.2A$2A18B19.4B8.9B4.8B8.A$.2B.16B20.4B6.4B4.2A5.10B3.B.A.2A$4.
16B21.4B4.4B5.A6.2BABA2B2A2B.B3A2.A$5.15B22.4B2.4B7.3A3.3B2A2B2A3BAB
2.2A$6.12B.B2A21.8B10.A3.3BA8B4A$6.11B2.BA.A21.6B11.A.2AB.7B3.2B.A$7.
10B5.A22.2B2A7.2A.2A.A.AB.7B2.B3A$7.6B2A2B5.2A20.3BABA7.A.A2.A5.4B4.A
$7.6B2A3B25.4BA3B5.A3.3A5.4B5.5A$7.10B25.4B2.4B5.3A7.4B10.A$3.A.2AB.
8B24.4B4.4B7.A5.4B9.A$.3AB2AB3.7B22.4B6.4B5.2A4.4B10.2A$A4.B6.6B21.3A
B8.4B4.9B$.3A.2A4.6B21.3BA10.4B5.6B$3.2A2.A4.5B20.3BA12.4B2.8B$6.A.A.
5BA2B18.4B14.15B$2.A.2A.A.A4BABAB8.A8.4B16.14B$2.2A.A.BA2B.4BAB6.3A7.
4B18.13B$5.A2.2B2.2A3BAB4.A9.4B20.10B.B2A$5.2A.B3.2A2BA2B4.2A7.4B23.
3B2AB3.BA.A$3.2A2.A.A2.4BA3B.4B6.4B24.3B2AB6.A$2.A2.A2.2A2.11B7.4B27.
4B6.2A$3.2A7.2A2B2A6B5.4B28.3B$12.6A6B4.4B26.AB.2B$13.BA9B3.4B26.A.AB
2AB$16.7B3.4B27.A.ABABAB$15.8B2.4B25.2A.A.A.A.A2.A$14.14B26.A2.A2.2A.
4A$13.16B27.2A4.A$11.B.17B.B30.A.A$10.2A17B.B2A30.2A$10.2A19B2A$11.
18B.2B$9.5B2.13B$9.2A6.10B$10.A7.9B$7.3A8.9B$7.A10.8B3.2A$18.7B3.A.A$
19.6B3.A$16.8B.2B2A$16.10BA2.2A$15.2A10B2A2.A$15.2A9B2.B.A2.A$16.B.9B
2A.A.2A$18.8B.A2.A$19.7B3.2A$20.6B$18.9B$18.2A4.4B$19.A5.3BA$16.3A7.
3BA$16.A10.3AB22$50.2A$50.A$51.3A$53.A!
chris_c wrote:The total of records in the variable gun collection is over 600 but I would love to find a glider loop of period 4N that is below 500 ticks in length. No luck so far.
I'll keep an eye out for p4N -- haven't tripped over any yet either.

Period 8N should be easy to find, since we can keep adding pairs of p8N reflectors, or substituting them for pairs of Snarks, until the timing works out. Just have to start with an even-period loop. Let's see, are either of these small enough to be interesting?

Code: Select all

#C [[ AUTOSTART HEIGHT 250 ZOOM 2 STEP 8 STOP 1000 ]]
x = 227, y = 108, rule = B3/S23
39bo$38bobo$37bo3bo$38bo3bo8bo$39bo3bo5b3o$40bo3bo3bo$41bobo4b2o$42bo$
46bo$45bobo$46b2o2b2o$50b2o18$212bo2$86bo45b2o3b2o71b3obo$86b2o44b2o2b
ob3o59bo8bob2o$85b2obo47bo4bo58b3o6b2obo$84bo2b3o42b4ob2o2bo61bo3bob3o
$83bobobo44bo2bobobob2o59b2o$82bobobo48bobobobo67bo$80b3o2bo50b2obobo
63bo$81bob2o55bo63bobo$67bo14b2o113bo2b2o2b2o$51b2o8bo3b3o11bo3bo42b2o
53b2o8bo3b3o2b2o$51b2o7bobobo13bobo46bo7b2o44b2o7bobobo$3b2o5b2o47bo2b
ob4o10b2o47bobo5b2o52bo2bob4o$4bo5b2o46bo3bo4bo14b2o44b2o58bo3bo4bo$4b
obo52b3obobo16bobo104b3obobo$5b2o54bobob2o10b2o5bo31bo74bobob2o$9b2o
26b2o23bo14b2o5b2o81b2o23bo22b2o5b2o$8bobo27bo75bob3o49bo46b2o5bo$9bo
28bobo75b2obo48bobo49bobo$5bo22bo10b2o75bob2o38bo10b2o49b2o$4b3o19b3o
88b3obo16b2o16b3o57b2o$3bob3o17bo112bo16bo60bobo$2bo3bo18b2o92bo19b3o
13b2o46b2o12bo$bo3bo4b2o58b2o51bo18bo57b2o2bo15b3o$3obo6bo58bobo49bobo
14b2obo57bob2o15bo$b3o7bob2o57bo50b2o14bobob2o57bo16bo3bo$2bo9bo2bo47b
2o7b2o45b2o21bo2bo47b2o7bo16bo2bobo$13b2o10bo37b2o53bobo22b2o10bo37b2o
5b2o19bobo2bo$23bobo2b2o88bo5b2o27bobo2b2o39bo22bo3bo$24b2o2b2o87b2o5b
2o28b2o2b2o37b3o26bo$196bo26b3o$51b2o128b2o13b2o$51b2o2b2o124b2o2b2o$
37bo3b2o12bobo109bo3b2o12bobo$36bobo3bo14bo108bobo3bo14bo$35bobo3bo15b
2o106bobo3bo15b2o$31b2obobo3bo120b2obobo3bo$31b2obo2b4obo118b2obo2b4ob
o$35bobo3bobo121bobo3bobo$31b2ob2o2bo2bobo117b2ob2o2bo2bobo$32bobo2b2o
3bo119bobo2b2o3bo$20b2o10bobo115b2o10bobo$20b2o11bo116b2o11bo13$166bo$
166b3o$169bo$168b2o7$178b2o$171b2o5bobo$171b2o7bo$180b2o2$167bo$166bob
ob2o$166bobobobo$165b2obobobo2bo$166bo2b2ob4o$166bo4bo$167b3obo2b2o$
169b2o3b2o!

User avatar
Kazyan
Posts: 1247
Joined: February 6th, 2014, 11:02 pm

Re: Execution of Old Guns by Variable-Speed Firing Squad

Post by Kazyan » March 24th, 2015, 3:32 am

Here's how you play weld-the-tail-on-the-beehive:

Code: Select all

x = 101, y = 59, rule = LifeHistory
47.2A11.A$46.B2AB9.A.A$47.3B9.A.A2.2A3.A$46.B.B9.2A.2A2.A2.A.A$46.5B
8.B2.A.A3.A.A$28.2A16.6B6.2ABA2.4A.A$27.A.A16.8B4.2A.A.A3.A$21.2A4.A
19.13B2.A.AB2.A15.2A$19.A2.A2.2A.4A13.13B5.A.A2B.A14.A$19.2A.A.A.A.A
2.A12.15B5.A2B.2A11.BA.A$22.A.ABABAB14.15B4.3B12.2A.B2A$22.A.AB2AB14.
17B.B.2B12.B2A2B$23.AB.2B15.23B2A2B2A6.4B$26.3B13.13B2A9B2A2B2AB3.6B$
26.4B6.2A3.14B2A7BA6B2A2.6B$24.3B2AB6.A3.2AB3.17BA7BAB.7B9.2A$9.A14.
3B2AB3.BA.A2.A.BA4.17B2AB2AB2A10B6.2B2AB6.2A$9.3A10.10B.B2A3.3A5.6B3.
B2.2B2.3BA3B2A11B5.4B7.A$12.A8.13B12.A5B14.21B.6B2.BA.A$11.2A7.14B3.
3A5.B2A6B14.27B.B2A$11.5B3.15B2.A2.A4.BABA4.2A15.28B$13.3B2.4B2.8B4.
2A5.4B5.A15.29B$3.2A7.9B5.6B10.4B7.3A12.29B$3.A8.8B4.9B8.4B10.A10.2AB
.3B2.22B$2A.A.B3.10B5.2A4.2B2A6.4B21.A.AB6.13B.9B$A2.3AB.2B2A7B6.A5.
2A2B4.4B22.A8.12B2.11B$.2A2.BA3B2A7B3.3A7.BA2B2.4B22.2A7.12B3.BAB.B3.
4B$3.4A12B3.A10.8B31.10B6.A.A.2A3.3BA$3.A.2B3.7B.B2A.A11.6B31.4B.6B4.
3A.ABA5.ABA$4.3AB2.7B.BA.A.2A.2A7.4B31.4B3.B.B5.A3.A4.A4.2A$7.A4.4B5.
A2.A.A7.6B29.4B5.3B4.2A2.A.4A5.B$2.5A5.4B5.3A3.A5.8B27.4B5.B2AB9.A$2.
A10.4B7.3A5.4B2.4B25.4B7.2A12.A$4.A9.4B5.A7.4B4.4B23.4B21.2A$3.2A10.
3BA4.2A5.4B6.4B21.4B$16.3BA5B4.4B8.4B19.4B$17.3A3B5.4B10.4B6.A10.4B$
17.8B2.4B12.4B5.3A7.4B$15.15B14.4B7.A5.4B$15.14B16.4B5.2A4.A3B$15.13B
18.4B4.6BABA$13.2AB.10B20.4B5.4B2A$12.A.AB3.B2A3B23.4B2.8B$12.A6.B2A
3B24.15B$11.2A6.4B27.14B$20.3B28.13B$21.2B.BA26.10B.B2A$20.B2ABA.A27.
3B2AB3.BA.A$19.BABABA.A27.3B2AB6.A$17.A2.A.A.A.A.2A26.4B6.2A$17.4A.2A
2.A2.A26.3B$21.A4.2A25.AB.2B$19.A.A30.A.AB2AB$19.2A31.A.ABABAB$49.2A.
A.A.A.A2.A$49.A2.A2.2A.4A$51.2A4.A$57.A.A$58.2A!
Tanner Jacobi
Coldlander, a novel, available in paperback and as an ebook. Now on Amazon.

User avatar
dvgrn
Moderator
Posts: 10610
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI
Contact:

Re: Execution of Old Guns by Variable-Speed Firing Squad

Post by dvgrn » March 24th, 2015, 3:53 am

Looks like p125/p250 are no problem either:

Code: Select all

x = 106, y = 62, rule = B3/S23
52b2o11bo$52b2o10bobo$64bobo2b2o3bo$63b2ob2o2bo2bobo$67bobo3bobo$63b2o
bo2b4obo$28b2o33b2obobo3bo$27bobo37bobo3bo15b2o$21b2o4bo40bobo3bo14bo$
19bo2bo2b2ob4o37bo3b2o12bobo$19b2obobobobo2bo51b2o2b2o$22bobobobo54b2o
$22bobob2o45bo$23bo36b2o9b2obo$60b2o8bo4bo$36b2o7b2o27bo20b2o$27b2o7bo
7bo2bo22b4o21b2o7b2o$9bo17b2o5bobo6bob2o57bo$9b3o22b2o7bo58bobo$12bo
29b2o58b2o$11b2o44b2o$57bo$47bo10b3o$3b2o42b2o11bo10b2o$3bo42bobo21bob
o$2obo20b2o44bo$o2b3o4b2o13bo43b2o23bo$b2o3bo3b2o10b3o68bobob2o6bo$3b
4o15bo12b2o54b3obobo5bobo$3bo15b2obo11b2o54bo3bo4bo4b2o$4b3o12bobob2o
11bo53b2o2bob4o$7bo13bo2bo58b2o10bo$2b5o14b2o60b2o12bo$2bo93b2o$4bo12b
o10bo$3b2o13bo7b3o$16b3o6bo$25b2o2$54bo10bo$54b3o8bobo$57bo7b2o$56b2o
2$15b2o$14bobo5b2o$14bo7b2o$13b2o2$27bo38b2o$23b2obobo30b2o5bobo$22bob
obobo30b2o7bo$19bo2bobobobob2o36b2o$19b4ob2o2bo2bo$23bo4b2o25bo$21bobo
30bobob2o$21b2o31bobobobo$51b2obobobobo2bo$51bo2bo2b2ob4o$53b2o4bo$59b
obo$60b2o!
There's even almost room to widen the loop enough so that the problem gliders cross in the other order. Probably the wider loop would work if you used the larger (working) p117 gun as a base, anyway.

chris_c
Posts: 966
Joined: June 28th, 2014, 7:15 am

Re: Execution of Old Guns by Variable-Speed Firing Squad

Post by chris_c » March 24th, 2015, 8:57 am

Good work on tweaking the p577's guys. You made the guns work but at a slight expense compared to the area predicted by the script.

I added the p520 by dvgrn. It is the best variable gun between 520 and 584 and it offers outright record breakers at p544, p272 and p280. But really there is no reason not to add an extra signal to the p560 in the old collection to get an even smaller p280.

Code: Select all

#C 1) new p272 based on p520 by Dave Greene
#C 2) p280 based on the old p560 (also by Dave Greene!)
x = 227, y = 82, rule = B3/S23
56bo11b2o91b2o$55bobo10b2o91b2o$47bo3b2o2bobo140bo$46bobo2bo2b2ob2o
138bobo$46bobo3bobo141bo3bo$47bob4o2bob2o95b2o30bo8bo3bo$49bo3bobob2o
95b2o30b3o5bo3bo$31b2o15bo3bobo134bo3bo3bo$32bo14bo3bobo134b2o4bobo$
32bobo12b2o3bo142bo$33b2o2b2o152bo$37b2o151bobo$52bobo131b2o2b2o$48bob
o9b2o124b2o$48bo11b2o$25b2o21bobo3bo20b2o$16b2o7b2o24b2obo19bo2bo$17bo
57b2obo$17bobo58bo14bo53b2o$18b2o58b2o10bo57bo61bo$63b2o24bobo56bobo
15b2o40b3o$64bo23bo3bo56b2o15b2o39bo$61b3o23bo3bo115b2o$49b2o10bo24bo
3bo90b3o$49bobo33bo3bo3bo89bo$51bo34bobo93bo9b3o$4b2o5b2o14bo23b2o34bo
105b3o10bo$5bo5b2o10b2obobo54bo92b2o27bobo$5bobo16bobob3o51bobo91b2o2b
2o23bobo$6b2o14bo4bo3bo50b2o96bobo23bo4b2o$10b2o10b4obo2bo55b2o94bo11b
2o15bobo$9bobo13bobobo7b2o47bobo96bo7bobo17bo9bo$10bo11b3o3bo8b2o42b2o
5bo89b2o4b3o6bo19b2o$22bo58b2o5b2o77b2o10bo3b3obo4b2o27b3obo$6b2o159bo
bo6b3o5bo3bo31bob2o$4bob2o161bo6bo8bo3bo30b2obo$3bo165b2o15bob3o27bob
3o$6bo180b3o$ob2obo182bo31bo$ob2o199b2o11bo$139b2o5b2o55b2o10bobo$140b
o5b2o67b2o$140bobo76b2o$141b2o26b2o48bobo$145b2o67b2o5bo$144bobo10b2o
8bo3bo42b2o5b2o$141b2o2bo11b2o8bo4bo$169bobobo8bo$140bo3bo25bobobo5b3o
8bo$139bo4bo26bo4bo2bo11b3o$138bobobo29bo3bo2b2o13bo$137bobobo51b2o$
135bo4bo27b2o3b2o2bo$135bo3bo7b2o19bo7bobo$148bo17bobo8b2o2b2o$137b2o
9bobo15b2o13b2o$149b2o4bo$154bobo57b2o$154bobo10bo46bo$155bo10bo2bo42b
obo$46b2o118bo2bo42b2o$43bo3bo120bo$42bo4bo$45bo107b2o$41bo2bo109bo15b
o$41b2o108b3o14b2o$45b3o103bo17b2o$45b2o$44b2o$43b2o4b3o119b2o$44bo4b
3o115b2o2b2o$41b3o5b3o114bobo29b2o$41bo10b3o108bo3bo30b2o$52b3o108b2o$
52b3o36bo70b2obo3b2o$92bo68bo2b3o2bo$90b3o67bobobo5b3o$159bobobo8bo$
157b3o2bo$158bob2o64bo$159b2o63bobo$160bo64b2o!
The p720 appears not to break any records. I think that even among the variable guns we already have it is beaten out by a combination of the families p598, p358q and p482q but I'm not 100% sure on that.

EDIT: Hang on I was confusing myself. It's the p592 that beats the p720 (there is no p598 anyway). Widening out the p592 by 16 cells gives a gun that is smaller in both directions than the p720 above.

User avatar
dvgrn
Moderator
Posts: 10610
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI
Contact:

Re: Execution of Old Guns by Variable-Speed Firing Squad

Post by dvgrn » March 24th, 2015, 12:49 pm

Here's a p640+8N that might be worth running through the system. It doesn't beat the fixed p640, but it's triple-adjustable with a lot of X slack -- 34, it looks like. So it's setting records by p800.

But does that just mean that the old p640 should be adjusted better, and it would beat this new one? Don't know yet.

Code: Select all

#C p640 syringe+HtoG#11 double-adjustable gun
#C [[ AUTOSTART STEP 8 LOOP 3200 ]]
x = 102, y = 66, rule = LifeHistory
43.2B$43.3B$41.2B2AB$41.2B2ABA$42.4B.A8.A$44.A.4B4.3A$23.2A20.AB2A2B
2.A$23.3B20.B2A2B2.2A$21.A3BA20.3B2.4B$21.A4BA20.2B.BAB$22.BABA.A8.A
11.B.ABAB$24.A.ABAB4.3A12.2B2AB.2A$25.A4BA2.A14.7B2A$26.A3BA2.2A14.7B
$26.3B2.4B14.7B$27.2A.BAB16.6B$28.B.ABAB14.8B$29.2B2AB.2A10.4B2.4B$
28.7B2A9.4B4.4B$29.7B9.4B6.4B$29.7B8.4B8.4B$29.6B8.4B10.4B$28.8B6.4B
12.4B28.2B$27.4B2.4B4.4B14.4B26.BAB$26.4B4.4B2.4B16.4B25.3A2B$25.4B6.
8B18.4B23.AB3AB$24.4B8.6B20.4B12.A8.A.2BAB$9.A13.4B9.7B20.4B11.3A4.BA
2B.A$9.3A10.4B10.7B21.4B13.A2.B3ABA$12.A8.4B10.7B2A21.4B11.2A2.2B3A$
11.2A7.4B12.2B2AB.2A22.4B10.4B2.BAB$11.5B3.4B12.B.ABAB19.A6.4B11.BAB.
2B$13.3B2.4B12.2B.BAB18.3A7.4B9.BABA.B$2A.2A7.9B12.BAB2.4B15.A11.4B5.
2A.B2A2B$2A.A8.8B13.3A2B2.2A6.2A7.2A11.4B4.2A7B$3.A.B2.11B13.AB3AB2.A
8.A4.5B12.4B4.7B$3.3A4B2A7B12.A.2BAB4.3A.A.A5.4B15.4B3.7B3.2A5.2A$.2A
2.BA3B2A7B10.BA2B.A9.2A.2A3.6B11.A3.4B3.6B.2B2A5.A$A2.4A12B9.B3ABA7.
2A4.B4.6B9.3A4.16B3.BA.A$2A.A.2B2.8B.B2A7.2B3A3.A4.A.3AB2AB.7B8.A8.7B
2.8B.B2A$.A2.3AB2.7B.BA.A2.2A4.BAB3.3A6.A.2AB2.6B.2B5.2A8.5B4.5B2A2B$
.A5.A3.5B5.A2.A5.2B7.A7.21B8.5B4.5BABAB$2.5A5.4B5.3A2.A11.2A3.B5.17B
9.7B2.7BAB.2B$4.A8.4B7.3A11.8B2.15B2A2B7.17B3.3B$14.4B5.A16.22BA2BA2B
5.4B.B.6B3.B.3B2AB$15.4B4.2A15.23B2A3B4.4B4.7B5.2B2ABA$8.2A6.9B14.19B
3.9B.4B5.7B6.4B.A$9.A7.6B14.21B4.8BA3B5.7B2A7.A.4B$9.A.2A5.6B3.B2.2B
2.25B4.7B2AB7.2B2AB.2A8.AB2A2B$10.A2.A4.34B4.4B3.6BABA7.B.ABAB12.B2A
2B$11.2AB3.27B.6B6.4B.9B7.AB.BAB13.3B$12.14B2A16B2.B.5B5.4B.7B7.B2A2.
4B12.2B$13.13B2A16B7.2A6.10B8.2ABAB2.2A$14.29B8.A8.4B.2B9.A2B3A2.A$
14.17B.4B16.3A6.4B10.A.ABAB4.3A$15.15B4.3B17.A7.4B7.BABA.A8.A$15.15B
5.A2B.2A22.4B5.3A2BA$16.13B5.A.A2B.A23.4B4.BAB2A$18.13B2.A.AB2.A25.4B
5.2AB$17.8B4.2A.A.A3.A27.4B4.BA$17.6B6.2ABA2.4A.A26.4B$17.5B8.B2.A.A
3.A.A26.4B$17.B.B9.2A.2A2.A2.A.A27.4B$18.3B9.A.A2.2A3.A29.3BA$17.B2AB
9.A.A38.3BA$18.2A11.A40.3A!
Might work to add another loop in the NE somehow -- you might start seeing cases where a dozen different periods all fit in exactly the same bounding box.

Anyway, maybe there's some useful technology here that can be applied elsewhere. The key things on my reminder list as I was putting this together were:
  • Replace a single p8N reflector with a 3-reflector loop, or vice versa, and it changes the timing by 2 ticks (one for each added reflector).
  • Swap a Snark with a p8N reflector and you'll get a 4-tick difference in timing. (Very handy!)
Speaking of p8N- and p4N-compatible technology, here's a small p464+8N -- not sure if this is useful, entertaining, or mostly just annoying, since its compression rating is rather compromised:

Code: Select all

#C p464+8N using a semi-Snark as a color-changing reflector
#C [[ AUTOSTART STEP 8 LOOP 4640 ]]
x = 88, y = 69, rule = B3/S23
29bo$29b3o$32bo$31b2o6b2o$38bobo$39bo3$26b2o$26b2o6b2o$34b2o2$43b2o$
43b2o2$28b2o$27bobo$27bo$26b2o17$3b2o5b2o$4bo5b2o70b2o$4bobo75bo$5b2o
73bobo$9b2o27bo22bo14b2o2b2o$8bobo27b3o19b2o14b2o$9bo31bo18bobo$5bo22b
o11b2o11b2o$4b3o19b3o24b2o$3bob3o17bo50bo$2bo3bo18b2o48b3o$bo3bo4b2o
62b2o2bo$3obo6bo61b2o3b2o$b3o7bob2o56bobo2bobo$2bo9bo2bo55b2o3b2o$13b
2o$28b2o$28b2o34b2o$63bobo18b2o$63bo20bobo$62b2o22bo$37bo3b2o43b2o$36b
obo3bo$35bobo3bo$31b2obobo3bo8bo$31b2obo2b4obo5b3o$35bobo3bobo4b3o$31b
2ob2o2bo2bobo4bobo28b2obo$32bobo2b2o3bo5b3o28b2ob3o$20b2o10bobo12b5o
33bo$20b2o11bo6b5obo5bob2o23b2ob3o$40bo3bobo5bobo12bo9bo2bobo$37b2obob
obobobobobo2bo12bo8b2o$37bob2ob2o3b2ob2o2b2o10b3o!

chris_c
Posts: 966
Joined: June 28th, 2014, 7:15 am

Re: Execution of Old Guns by Variable-Speed Firing Squad

Post by chris_c » March 24th, 2015, 2:24 pm

dvgrn wrote:Here's a p640+8N that might be worth running through the system.
dvgrn wrote:Speaking of p8N- and p4N-compatible technology, here's a small p464+8N
Yes both of those look like they will set records. The p640 is good at large periods due to the high amount of slack. The p464 looks like it will set records between 464 and your recent p520.

I will make a proper update once I teach the script that the subperiods all need to be multiples of 8 when p8 oscillating components are present.

Other things on the TODO list are:

- Add the fixed guns that simsim and Kazyan have been collecting.
- Find a formula that works correctly with with doubly adjustable loops.
- Make an automated run of Hersrch to make sure no nice Herschel loops are missed.
- Add semi-snarks to existing fixed guns to see if they bring any improvements.

(The last two items may have to wait though!)

Also I made a few ad-hoc guns using p5 and p7 reflectors starting from p415 and p399 respectively:

Code: Select all

x = 228, y = 73, rule = B3/S23
34b2o$33bo2bo$32bob2obo2b2o128b2o$32bobo2bo3bo127bo2bo$31b2o3bob3o125b
o3bobo$31bo2bo3bo125b6obo$32b4obo125bo5bo$35bo127bob4o2bo$28b2o2b3o11b
2o116bobobob2o$28bo2bo2b3obo8bo118bobo$29bob2o3b2o5b2o120b2obob2o$26b
3obo5b2o5b3o118bo5bo$25bo4bo8b2o2b3o116bo2bobobo2bo9bo$23b3ob2ob2o7b2o
4bo116b2obobobob2o7b3o$22bo3bobo2bo12b2o117bobo4bo8bo$21bo2bo5bo6bo
125bob2obobo8b2o$21bob3o3b2o5bobo125bo3bob2o$19bobo4b3o2bo5b2o2b2o122b
4o8bo$19b2ob2o2bobobo10b2o126b2o2b2obobo$22bo3bobo2bo133b4o8b2o2b2o$
22bobobo2b2o134bo2bob2o9b2o$23bobobobo140bo$24bo2bobo138bobo$27b2o139b
2o3$44b2o$43b2o$8b2o35bo$8bobo$13b2o$11b2obo172b2o$10b2o175bobo$17b2o
168bo$2b2o13b2o$2bobo7b2o136b2o5b2o$4bo7b2o137bo5b2o$4b3ob2o6b2o133bob
o$2b2o6bo4bobo134b2o$bo2b2o4b2o4bo63b2o74b2o60b2o$obo5bo58b2o12bo73bob
o47b2o12bo$ob2ob2obo7bo50b2o10bo76bo48b2o10bo$bo4bobo5bo59b2o2bob4o
128b2o2bob4o$2b3obobobo7bo55bo3bo4bo4b2o122bo3bo4bo4b2o$4b2o3b2o2bo61b
3obobo5bobo65b3o55b3obobo5bobo$11b3ob2obobo56bobob2o6bo51b2o7b2o63bobo
b2o6bo$8bobo3bo2b2ob3o30b2o23bo61bo2bob2obo2bo2bobobo32b2o23bo$8b2obob
2o3b2o3bo30bo85bobo3bob2o2b3obob3o31bo$11bobo7b3o30bobo84bob2o5b2o4bo
4bo30bobo$11bo2bo2b4o23bo10b2o86bob5o2b3o2b2ob2o20bo10b2o$12b2o2bo4bo
20b3o98bobo3b4o2bobo22b3o$13bob2obobobo18bo100b2o2b3o4bobobo21bo$13bo
2bo4bo19b2o100bobo2bob4obob2o20b2o$14bo3b3o5b2o58b2o55bo2bo2bo4bo9b2o
58b2o$15b3o9bo58bobo55b2o5b3o11bo58bobo$18bo8bob2o57bo61b2o13bob2o57bo
$17b2o9bo2bo47b2o7b2o76bo2bo47b2o7b2o$29b2o48b2o86b2o48b2o$44b2o136b2o
$44b2o136b2o2$67b2o136b2o$67b2o2b2o132b2o2b2o$53bo3b2o12bobo117bo3b2o
12bobo$52bobo3bo14bo116bobo3bo14bo$51bobo3bo15b2o114bobo3bo15b2o$47b2o
bobo3bo128b2obobo3bo$47b2obo2b4obo126b2obo2b4obo$51bobo3bobo129bobo3bo
bo$47b2ob2o2bo2bobo125b2ob2o2bo2bobo$48bobo2b2o3bo127bobo2b2o3bo$36b2o
10bobo123b2o10bobo$36b2o11bo124b2o11bo!
These kind of loops are always 7 modulo 8. I don't think we have any others of that type until p623 so these are setting a nice handful of records at the moment. If anyone can move the p5 reflector 5 cells closer to the beehive-with-tail then they would have a record gun at p375.

User avatar
dvgrn
Moderator
Posts: 10610
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI
Contact:

Re: Execution of Old Guns by Variable-Speed Firing Squad

Post by dvgrn » March 25th, 2015, 12:51 am

dvgrn wrote:But does that just mean that the old p640 should be adjusted better, and it would beat this new one? Don't know yet.
Still don't know for sure, but it looks like there should be a lot of ways to improve various p8N periods. Here's another p520+8N with a smaller bounding box and a good bit of slack:

Code: Select all

#C p520+8N gun -- no syringes, just old technology plus Snarks
#C [[ AUTOSTART THEME 4 ZOOM 5 STEP 8 LOOP 5200 ]]
x = 87, y = 63, rule = B3/S23
37b2o3b2o$37b2o2bob3o$41bo4bo$37b4ob2o2bo$37bo2bobobob2o10bo$40bobobob
o10b3o$41b2obobo9b3obo$45bo11bo3bo8bo$58bo3bo5b3o$31b2o26bob3o3bo$32bo
7b2o18b3o4b2o$32bobo5b2o19bo$33b2o30bo$64bobo$65b2o2b2o$69b2o4$43b2o$
43bo$15b2o27b3o$8b2o5b2o29bo$8b2o72bo$42bo6b2o13b2o14b5o$29b2o11b3o6bo
13bo13bo5bo$10b2o16bobo14bo2bo16bobo12b3o2bo$10b2o18bo13b2o3b2o15b2o
15bob2o$4b2o74b4o2bo$4b2o69b2o3bo3b2o$75b2o4b3o$83bo$83bob2o$82b2ob2o
2$35b2o$35b2o25b2o10b2o$47b2o13b2o2b2o6bo$46bo2bo16bobo6b3o$47b2o4b2o
12bo9bo$53bobo14b3o$bo53bo8b2o3bo$b3o51b2o8bo3bo3bo$4bo40b2o15b3o4bo2b
obo$3b2o41bo15bo8bobo2bo$43b3o26bo3bo$30b2o11bo32bo$30b2o2b2o37b3o$34b
obo$25bo9bo$23b3o13bo$22bo9b2o4bobo$22b2o9bo3bo3bo$30b3o5bo3bo$30bo8bo
3bo$2o38bo3bo$2o39bobo$42bo3$17bo8bo$9b2o5bobo8bo$9b2o6bo7b3o!

Post Reply