Making switch-engines

For discussion of specific patterns or specific families of patterns, both newly-discovered and well-known.
User avatar
Extrementhusiast
Posts: 1966
Joined: June 16th, 2009, 11:24 pm
Location: USA

Re: Making switch-engines

Post by Extrementhusiast » November 2nd, 2014, 1:45 pm

simsim314 wrote:Glider + 3 still life (3 cells) -> Switch engine + *WSS:

Code: Select all

x = 24128, y = 101, rule = LifeHistory
4018.2A$4018.2A15$4033.3A2$8.A3999.A3999.A$8.A3999.A3999.A$8.A3999.A
3999.A4$.2A3998.2A3998.2A$A.A3997.A.A3997.A.A$2.A3999.A3999.A13$15.A$
15.A$15.A17567.A3999.A$17583.A3999.A$17583.A3999.A$7984.A36.A$7984.A
36.A9560.3A3997.3A$7984.A36.A3$31.2A15376.3A$31.2A11373.3A$15393.2A$
15393.2A5$24106.3A$17597.A3999.A$11419.A3999.A2177.A3999.A$11408.3A8.
A3999.A2177.A3999.A$11419.A3999.A3$17590.2A3998.2A$11412.2A3998.2A
2175.A.A3997.A.A$11411.A.A3997.A.A2177.A3999.A$11413.A3999.A13$24127.
A$24127.A$24127.A4$24120.2A$24119.A.A$24121.A12$24094.3A!
The third one is invalid, as the glider has to pass through a blinker. The sixth and seventh are identical, and are also invalid, since two of the blinkers interact.
I Like My Heisenburps! (and others)

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

Re: Making switch-engines

Post by simsim314 » November 2nd, 2014, 1:58 pm

Extrementhusiast wrote:The third one is invalid, as the glider has to pass through a blinker. The sixth and seventh are identical, and are also invalid, since two of the blinkers interact.
Very correct, thx for noticing. I'm just posting stuff as they come directly from the search, some of them are invalid. Another note is that in no means this is somehow "full" range of examples, the search was very partial, and it didn't covered all the possible variations, just some that I guessed might work better, in some ranges, while many of the searches I stopped in the middle.

Anyway here is another set of examples of 11 cells switch engine + *WSS.

Code: Select all

x = 16003, y = 86, rule = LifeHistory
5.A$5.A$5.A$4015.A$4015.A$4015.A7$7968.A$7968.A$7968.A$11966.2A$
11966.2A13$3969.3A5$.A3999.A3999.A3999.A3999.A$3A3997.3A3997.3A3997.
3A3997.3A$2.A3999.A3999.A3999.A3999.A6$14.3A14$12002.2A$12002.2A18$
15992.2A$15992.2A4$7996.2A$7996.2A4$15998.3A!

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

Re: Making switch-engines

Post by simsim314 » November 5th, 2014, 10:48 am

Here are 9 more 10 cells linear growth predecessors (7 cell Methuselah + 3 cells blinker / proto block) with 4 more glider emitting switch engines. Raising the number of 10 cells glider emitting switch engine predecessors to 8:

Code: Select all

x = 16246, y = 129, rule = B3/S23
10015b3o34$6050b3o6017b2o$12070b2o$14174bo$8054bo6119bo$8054bo6119bo$
8054bo21$18bo2001bo1999bo1999bo1999bo1997bo1999bo$18bo2001bo1999bo
1999bo1999bo1998b2o1998b2o$18bo2000b2o1998b2o1998b2o1998b2o1997bo1999b
o$2018bo1999bo1999bo1999bo1999b2o1998b2o$20b2o1996bo1999bo1999bo1999bo
1999bo1999bo$19b2o1997bo1999bo1999bo1999bo3$16211b2o$16213bo$16212bo$
16211bo$16211bo$16211bo8$3o4$14164bo$14164b2o$14163bo$14165b2o$14167bo
3$4049b3o$16243b3o34$2002b2o$2002b2o!


For reference here is almost full collection of 7 cells Methuselahs (more than 800 generations):

Code: Select all

x = 1860, y = 2041, rule = B3/S23
1215bo$1216b2o$1215bo353bo2b2o$858bo356b2o352b2o2bo$612bo245bo356bo
353bo$612bo244b2o$352b2o258bo243bo$351bo504bo$354bo259b2o240bo$354bo
258b2o$353bobo4$75bo2$76b3o2$74bo$74b2o516$1017bo$1017bo$1017bo$1015b
2o$1013bo$1015bo$689bo1165bo$688bo1166bo$686bo2bo1165bo$687bo1171bo$
686bobo1168b2o$1528bo329bo$1528bo$1529bo$338bo$338bo1188b2o$338bo1189b
o$341bo962b2o222bo$339b3o963bo$1303bo$1303bo$64bo1238bo$64bo1238bo$64b
o$65b3o$66bo442$206b3o$27bo177bo$27b2o175b2o$25bo$26b2o178bo2$26bo335$
14bo$14bo$14bo$12b2o$13bo$14bo65bo$80b2o$79bo$81b2o$83bo$190bo$191bo$
191bo$192bo$188b3o212$7bo100b3o$8bo$8bo$9bo98b2o$7b2o99bo$107bo$9bo
160$7bo$7bo$3bo3bo$4b2o106b2o$114bo$5bo107bo$112bo$112bo$112bo121$o2$b
2o$bo$2bo$2bobo82$5bo$5bo$5bo$2bo$2b2o2$4bo78$6bo$5b2o$4bo$4bo$4bo$2bo
!
EDITJust for protocol here is 6 + 4 linear growth (the only one came up from search):

Code: Select all

x = 11, y = 14, rule = B3/S23
3bo$b2o$o$o$o7$10bo$8b2o$9bo!

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

Re: Making switch-engines

Post by chris_c » November 6th, 2014, 8:38 pm

I thought I would join in on the switch-engine fun. Here is a 14812x1 quadratic growth that uses three switch engines. Two of the switch engine recipes are based on the recipe I saw in the original Nx1 quadratic growth and the other is based on the loaf + R-pentomino reaction that is mentioned in the first post of this thread. In all cases I have adjusted the position of the taming junk so that it is easily constructable with horizontal LWSS. Originally I intended to make a Nx1 quadratic growth based on switch engine ping-pong but that doesn't look like it will be competitive with three switch engine patterns unless the distance between the ping-pong streams can be reduced quite a bit.

Code: Select all

x = 14812, y = 1, rule = B3/S23
4ob4ob4ob4o2b5o51b3o2b4o2b10ob4o128b4o2b4ob6o54b4o3b3ob5o107b4ob4ob4ob
4o2b5o51b3o2b4o2b10ob4o80b4o2b4ob6o54b4o3b3ob5o70b3ob8o2b5ob3ob3o58b5o
b5o2b4ob3ob4o61b3o3b7o4b5o94b5o2b4o3b3o7b4ob4ob4ob4o2b5o51b3o2b4o2b10o
b4o102b4o2b4ob6o54b4o3b3ob5o40b4ob4ob4ob4o2b5o51b3o2b4o2b10ob4o112b4o
2b4ob6o54b4o3b3ob5o134b3ob8o2b5ob3ob3o58b5ob5o2b4ob3ob4o87b3o3b7o4b5o
94b5o2b4o3b3o20b3ob8o2b5ob3ob3o58b5ob5o2b4ob3ob4o91b3o3b7o4b5o94b5o2b
4o3b3o8b4ob4ob4ob4o2b5o51b3o2b4o2b10ob4o144b4o2b4ob6o54b4o3b3ob5o21b5o
b3o3b4o54b6ob4o2b4o140b4ob10o2b4o2b3o51b5o2b4ob4ob4ob4o115b5ob3o3b4o
54b6ob4o2b4o118b4ob10o2b4o2b3o51b5o2b4ob4ob4ob4o42b5ob3o3b4o54b6ob4o2b
4o118b4ob10o2b4o2b3o51b5o2b4ob4ob4ob4o39b5ob3o3b4o54b6ob4o2b4o108b4ob
10o2b4o2b3o51b5o2b4ob4ob4ob4o128b5ob3o3b4o54b6ob4o2b4o106b4ob10o2b4o2b
3o51b5o2b4ob4ob4ob4o97b5ob3o3b4o54b6ob4o2b4o86b4ob10o2b4o2b3o51b5o2b4o
b4ob4ob4o39b5ob3o3b4o54b6ob4o2b4o84b4ob10o2b4o2b3o51b5o2b4ob4ob4ob4o
39b5ob3o3b4o54b6ob4o2b4o96b4ob10o2b4o2b3o51b5o2b4ob4ob4ob4o614b3ob8o2b
5ob3ob3o58b5ob5o2b4ob3ob4o119b3o3b7o4b5o94b5o2b4o3b3o4893b3ob8o2b5ob3o
b3o58b5ob5o2b4ob3ob4o103b3o3b7o4b5o94b5o2b4o3b3o101b3ob8o2b5ob3ob3o58b
5ob5o2b4ob3ob4o113b3o3b7o4b5o94b5o2b4o3b3o20b3ob8o2b5ob3ob3o58b5ob5o2b
4ob3ob4o117b3o3b7o4b5o94b5o2b4o3b3o21b3ob8o2b5ob3ob3o58b5ob5o2b4ob3ob
4o135b3o3b7o4b5o94b5o2b4o3b3o23b3o3b4o2b5o94b5o4b7o3b3o131b4ob3ob4o2b
5ob5o58b3ob3ob5o2b8ob3o128b5ob3o3b4o54b6ob4o2b4o144b4ob10o2b4o2b3o51b
5o2b4ob4ob4ob4o42b5ob3o3b4o54b6ob4o2b4o144b4ob10o2b4o2b3o51b5o2b4ob4ob
4ob4o8b3o3b4o2b5o94b5o4b7o3b3o99b4ob3ob4o2b5ob5o58b3ob3ob5o2b8ob3o!

User avatar
codeholic
Moderator
Posts: 1147
Joined: September 13th, 2011, 8:23 am
Location: Hamburg, Germany

Re: Making switch-engines

Post by codeholic » November 6th, 2014, 9:12 pm

This is quite a significant improvement! Congrats!
Ivan Fomichev

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

Re: Making switch-engines

Post by simsim314 » November 7th, 2014, 3:23 am

Hey Chris congrats!

Just a point of thought - I'm pretty sure the recipes are overkill in order of magnitude. You can look at my 24 quadratic, it uses debris from R + blinker to make switch engine from MWSS collision.

I would start from making simple searches for collisions of some nice Nx1 debris with MWSS the height of which you can tweak, and look there. Probably 2 MWSS would be more than enough. I would also try to take the 34 cell switch engine, and shoot a glider (or *WSS) into it to make it glider emitting switch engine.

Just notice that the initial bounding box size can't be improved drastically, but the amount of living cells can.

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

Re: Making switch-engines

Post by chris_c » November 7th, 2014, 8:38 am

Thanks guys.

Searching for more natural switch engine recipes would definitely improve the pattern but I was too lazy to search for switch engines completely from scratch.

Another way to reduce the bounding box is to add a delay before the trigger of the eastern switch engine. The pattern would still work if the eastern switch engine was pulled west by 384 cells but was triggered 1536 generations later.

This is one idea to make a delay longer than in my current pattern. Build a hive in the north which is struck by a stray glider from the construction of the western switch engines. The return glider hits an LWSS seed which then strikes the switch engine seed. The trigger glider from the western switch engine is fixed but adjusting the distance between the LWSS seed and the switch engine seed is hopefully enough to replace this degree of freedom.

If all the details can be worked out I think getting below 10,000x1 is probably possible.

Code: Select all

x = 491, y = 101, rule = B3/S23
486bo$486bo$486bo$489bo$326b2o160bobo$326b2o161bo8$486b2o$485bo2bo$
334b2o150b2o$334b2o141bo$477bo$284b2o42b2o147bo$284bobo41b2o$284bo$77b
2o$76bobo$78bo75$3o$2bo$bo!

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

Re: Making switch-engines

Post by simsim314 » November 7th, 2014, 10:05 am

chris_c wrote: 10,000x1 is probably possible
There is a nuance here. The first distance to obstacle which generates switch engine for the quadratic, requires somewhere around 7000 cell diagonally. Making it all work requires some additional adjustments, so don't expect "miracles".

Anyway I think this recipe would work much better for you:

Code: Select all

x = 1246, y = 210, rule = B3/S23
b5o$o4bo$5bo$o3bo$2bo52$1082b2o$1082b2o70$962b2o$961bobo$963bo$1167b2o
$1167bobo$1167bo75$1243b3o$1243bo$1244bo!

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

Re: Making switch-engines

Post by chris_c » November 7th, 2014, 10:39 am

simsim314 wrote: There is a nuance here. The first distance to obstacle which generates switch engine for the quadratic, requires somewhere around 7000 cell diagonally.
Yes, I noticed that the crystal has to be quite long to make the block laying switch engine. Shift everything east by 2 * 723 cells and it all still works, but shifting to the west by the same amount means the crystal is too short when the collision happens.
simsim314 wrote: Anyway I think this recipe would work much better for you:
Thanks! I have also found a few options for G->*WSS so I might see how far these new tricks take us.

Code: Select all

x = 143, y = 33, rule = B3/S23
bo59bo59bo$2bo59bo59bo$3o57b3o57b3o12$17bo$15b2o60bobo$16b2o59b2o$78bo
4$16b3o60b3o$16bo62bo$17bo62bo$140bobo$74b2o64b2o$15b2o56bobo65bo$16b
2o57bo$15bo$141bo$130b3o7b2o$132bo7bobo$131bo!

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

Re: Making switch-engines

Post by chris_c » November 9th, 2014, 9:56 pm

Using the silly delaying method and the simpler synthesis for the lone switch engine I managed to get down to 7,242x1. Over half as small surprisingly! The crystal is already as small as can be allowed to make the block laying switch engines so I'm not sure how much smaller the pattern can be made. I'm pretty happy to quit at this point anyway.

Code: Select all

x = 7242, y = 1, rule = B3/S23
6ob4o2b4o77b4ob4ob4ob4o2b5o51b3o2b4o2b10ob4o128b4o2b4ob6o54b4o3b3ob5o
107b4ob4ob4ob4o2b5o51b3o2b4o2b10ob4o80b4o2b4ob6o54b4o3b3ob5o70b3ob8o2b
5ob3ob3o58b5ob5o2b4ob3ob4o61b3o3b7o4b5o94b5o2b4o3b3o7b4ob4ob4ob4o2b5o
51b3o2b4o2b10ob4o102b4o2b4ob6o54b4o3b3ob5o40b4ob4ob4ob4o2b5o51b3o2b4o
2b10ob4o112b4o2b4ob6o54b4o3b3ob5o134b3ob8o2b5ob3ob3o58b5ob5o2b4ob3ob4o
87b3o3b7o4b5o94b5o2b4o3b3o20b3ob8o2b5ob3ob3o58b5ob5o2b4ob3ob4o91b3o3b
7o4b5o94b5o2b4o3b3o8b4ob4ob4ob4o2b5o51b3o2b4o2b10ob4o144b4o2b4ob6o54b
4o3b3ob5o21b5ob3o3b4o54b6ob4o2b4o140b4ob10o2b4o2b3o51b5o2b4ob4ob4ob4o
115b5ob3o3b4o54b6ob4o2b4o118b4ob10o2b4o2b3o51b5o2b4ob4ob4ob4o42b5ob3o
3b4o54b6ob4o2b4o118b4ob10o2b4o2b3o51b5o2b4ob4ob4ob4o39b5ob3o3b4o54b6ob
4o2b4o108b4ob10o2b4o2b3o51b5o2b4ob4ob4ob4o128b5ob3o3b4o54b6ob4o2b4o
106b4ob10o2b4o2b3o51b5o2b4ob4ob4ob4o97b5ob3o3b4o54b6ob4o2b4o86b4ob10o
2b4o2b3o51b5o2b4ob4ob4ob4o39b5ob3o3b4o54b6ob4o2b4o84b4ob10o2b4o2b3o51b
5o2b4ob4ob4ob4o39b5ob3o3b4o54b6ob4o2b4o96b4ob10o2b4o2b3o51b5o2b4ob4ob
4ob4o438b5o3b3ob9o60b10o4b9o42b4o2b4o2b6o89b4ob5o2b6ob4o21b3o2b10ob3ob
3o420b3o3b8o5b3ob4o19b5o2b3ob5ob7o27b4o2b4ob6ob3o61b3ob3o4b3ob7o!

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

Re: Making switch-engines

Post by simsim314 » November 10th, 2014, 5:55 am

Hey congrats on the improvement.
chris_c wrote:Using the silly delaying method
Actually it's very impressive method. Maybe you can built few such delaying mechanisms to solve the distance issue. If so we might reach somewhere around 2K or even less.

But first the crystal is just way to wasteful. Tell me if this helps (notice the timing of gliders is important - but also notice that they are in opposite direction so it's just matter of correct placement of the blinker, relative gliders state, and correct opposite parity):

Code: Select all

x = 370, y = 718, rule = B3/S23
336bo$336bobo$336b2o334$3b3o2$b2o$obo$2bo39$368bo$367bo$367b3o331$33bo
$33bo$33bo$29b3o$31bo$30bo!
You can tweak the location and the timing of the switch engines pretty much, and still almost always get side emitting glider stream.

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

Re: Making switch-engines

Post by chris_c » November 10th, 2014, 8:42 am

simsim314 wrote:Tell me if this helps...
Hmmm.... maybe, but it looks quite scary to make this from a one cell thick pattern using only gliders. Here is a possible predecessor using only northbound gliders:

Code: Select all

x = 715, y = 132, rule = B3/S23
659bo$380b3o275b2o$382bo275bobo$381bo56$b2o303b2o$obo302b2o$2bo304bo
19$697b2o$687b3o7bobo$689bo7bo$688bo17$712b2o$712bobo$712bo7$311bo8b2o
$311b2o6b2o$310bobo8bo18$335b3o$335bo$336bo!
The gliders heading to the north-west are on lanes that are quite close together. I don't know if you took a close look at how I made the R-pentomino + block constellation in my previous pattern:

Code: Select all

x = 231, y = 1, rule = B3/S23
3o3b8o5b3ob4o19b5o2b3ob5ob7o57b4o2b4ob6ob3o61b3ob3o4b3ob7o!
The result might look nice and simple but it took me hours to find this! The 22-cell pattern on the right emits two gliders on the same lane about 200 generations apart. Something similar would probably be needed to make the two gliders + blinker recipe, but now also there is also the timing dependency between the left and the right recipe. Pretty scary. (EDIT: now I see that the right hand switch engine only interacts with the P2 debris of the left switch engine so the timing dependency is nowhere near as bad as I thought, probably it is buildable with a lot of patience)

I think the simplest way to get the side-emitting switch engine would be to search for junk from 2 gliders NE + 2 gliders NW that can be triggered by a slow L/MWSS from one side to make a switch engine. Probably not quite as efficient but it sounds a lot less daunting to build.

Also, if you want to squeeze every last drop out of this project then I think it would be slightly more efficient (create a longer delay) if the switch engine in the east was triggered by a westbound MWSS like so:

Code: Select all

x = 155, y = 97, rule = B3/S23
20bo$19bobo$19bobo$20bo2$15b2o7b2o$14bo2bo5bo2bo$15b2o7b2o2$20bo$19bob
o$19bobo$20bo19$3o$2bo$bo49$153bo$151b2o$152b2o5$152b3o$152bo$153bo3$
151b2o$152b2o$151bo!

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

Re: Making switch-engines

Post by simsim314 » November 10th, 2014, 10:13 am

chris_c wrote:...but it looks quite scary to make this from a one cell thick pattern using only gliders
Well you don't have to use only gliders, you can collide MWSS into block or blinker and get gliders as well.

Anyway how about something of this kind?

Code: Select all

x = 1811, y = 667, rule = B3/S23
b5o$o4bo$5bo$o3bo$2bo52$1082b2o$1082b2o70$962b2o$961bobo$963bo$1167b2o
$1167bobo$1167bo75$1243b3o$1243bo$1244bo248$570b5o$569bo4bo$574bo$569b
o3bo$571bo52$1647b2o$1647b2o70$1527b2o$1526bobo$1528bo$1732b2o$1732bob
o$1732bo75$1808b3o$1808bo$1809bo!
It's a bit ugly but it will VERY probably work for quadratic. Also there should be many constellations of this kind, so if something doesn't work you can tweak it a bit (also change the MWSS timing) and see if it will work.

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

Re: Making switch-engines

Post by chris_c » November 10th, 2014, 11:59 am

simsim314 wrote: Anyway how about something of this kind?
It should work in principle but I cannot make that with current 1-cell thick components: the R + block constellations are too close together. I would need to make another R + block recipe with different timing so that NE gliders from the eastern recipe do not hit the NW gliders from the western recipe.

I have come up with this design which is a decent compromise: quite efficient and not too far away from current recipes. The number of *WSS is reduced from 16 down to 4. The details still to be worked out are the MWSS recipe and a way of making the 4 gliders. (Hmmm... but the more I look at it the more difficult it seems to make those gliders).

Code: Select all

x = 2032, y = 413, rule = B3/S23
b5o$o4bo$5bo$o3bo$2bo81$2029b2o$233b4o1791b4o$232bo3bo1790b2ob2o$236bo
1791b2o$232bo2bo2$1736b2o$1735b2ob2o$1736b4o$1737b2o96$889bo$889b2o$
888bobo$1097b2o$1096b2o$1098bo4$1119b2o$911b3o205bobo$913bo205bo$912bo
4$667b3o3b8o5b3ob4o19b5o2b3ob5ob7o203b4o2b4ob6ob3o61b3ob3o4b3ob7o4$
912bo$913bo205bo$911b3o205bobo$1119b2o4$1098bo$1096b2o$1097b2o$888bobo
$889b2o$889bo96$1737b2o$1736b4o$1735b2ob2o$1736b2o2$232bo2bo$236bo
1791b2o$232bo3bo1790b2ob2o$233b4o1791b4o$2029b2o81$2bo$o3bo$5bo$o4bo$b
5o!

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

Re: Making switch-engines

Post by simsim314 » November 10th, 2014, 12:50 pm

I don't see why do you need those blinkers there.

If you want the ignition from the other side maybe this can help:

Code: Select all

x = 393, y = 33, rule = B3/S23
389b3o$388b5o$387b2ob3o$388b2o21$5b3o6$2bo$3o$bo!
If you worry about the "messy" side rake - don't, The periods should work out for the quadratic. As you said it's all mod 723 and the period is mod 32 * N, so as long as you fine mod 723 everything else will work out as it should, even with pretty messy side puffer. The worst can happen it will take a bit longer to "reignite" from the messy side puffer, but it will ignite - the messy stuff is somewhat better because they also give more options to play with, you just need to tweak it with steps of 64 (all the places where the gliders intersect each other in "nice" way), and you will get the quadratic.

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

Re: Making switch-engines

Post by chris_c » November 10th, 2014, 8:36 pm

simsim314 wrote:I don't see why do you need those blinkers there.
They were designed to be easy to construct. The horizontal distance between them is quite wide so that the gliders needed to construct them can be easily made with non-conflicting 1-cell thick recipes.
simsim314 wrote: If you want the ignition from the other side maybe this can help:
Thanks. One ignition from either side is probably preferable if both switch engines are going to be triggered by *WSS.
simsim314 wrote:If you worry about the "messy" side rake - don't
I wasn't particularly worried about the messiness, I was worried about being able to make the R-pentomino + block pattern twice in such a short space without having to re-work the recipe. The closest that the current recipe can be repeated is 220 lanes but that doesn't allow for any interaction between the switch engines:

Code: Select all

x = 1524, y = 219, rule = B3/S23
b5o$o4bo$5bo$o3bo$2bo96$b5o$o4bo$5bo$o3bo$2bo114$903b3o3b8o5b3ob4o19b
5o2b3ob5ob7o227b4o2b4ob6ob3o61b3ob3o4b3ob7o19b3o3b8o5b3ob4o19b5o2b3ob
5ob7o27b4o2b4ob6ob3o61b3ob3o4b3ob7o!
Having looked at all this with a fresh pair of eyes I am going to try to make a 1-cell thick pattern for the 2G + blinker switch engine. That is the cheapest one (just 5 northbound gliders required) and I think it shouldn't be too hard to build. Something like this:

Code: Select all

x = 1704, y = 289, rule = B3/S23
b5o$o4bo$5bo$o3bo$2bo236$1635b2o$1341b3o291bobo$1343bo291bo$1342bo22$
1655b2o45bo$1029b3o623bobo43b2o$1031bo623bo45bobo$1030bo20$903b3o3b8o
5b3ob4o19b5o2b3ob5ob7o367b4o2b4ob6ob3o61b3ob3o4b3ob7o!
And if I get really keen I will try to go for a version with 2 x (2G + blinker), 10 gliders in total.

EDIT: Nearly complete now. Should end up about 1,000 cells wide once the MWSS recipe gets done.

Code: Select all

x = 1210, y = 305, rule = B3/S23
b5o$o4bo$5bo$o3bo$2bo148$539b4o6b7ob4o134b3o3b8o5b3ob4o19b5o2b3ob5ob7o
58b4ob5ob3ob3o2b6o11b4o2b4ob6ob3o61b3ob3o4b3ob7o181b4o4b3o3b9o19b3ob3o
3b7o3b3o148$2bo$o3bo$5bo$o4bo$b5o!

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

Re: Making switch-engines

Post by simsim314 » November 11th, 2014, 12:41 am

chris_c wrote:, I was worried about being able to make the R-pentomino + block pattern twice in such a short space without having to re-work the recipe.
Well the recipe of R + B is much more adjustable than that, you can move each side by 2 * N, and place one recipe inside the other, for example:

Code: Select all

x = 1443, y = 279, rule = B3/S23
10b5o$9bo4bo$14bo$9bo3bo$11bo136$b5o$o4bo$5bo$o3bo$2bo134$922b3o3b8o5b
3ob4o19b5o2b3ob5ob7o61b3o3b8o5b3ob4o19b5o2b3ob5ob7o67b4o2b4ob6ob3o61b
3ob3o4b3ob7o45b4o2b4ob6ob3o61b3ob3o4b3ob7o!
EDIT Checked again and this trick still doesn't help.

chris_c wrote:EDIT: Nearly complete now. Should end up about 1,000 cells wide once the MWSS recipe gets done.
Yep it looks much better, it's really efficient, so no "order of magnitude" improvement can be made. Maybe tweaking it a bit will do some more improvements but not by 500%, so great! This also leaves the space for timing mechanism that the quadratic requires.

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

Re: Making switch-engines

Post by Extrementhusiast » November 24th, 2014, 10:14 pm

On a different note, glider (from any direction) + two blinkers + two blocks = perfectly clean switch engine:

Code: Select all

x = 94, y = 21, rule = B3/S23
4b2o26b2o26b2o26b2o$4b2o26b2o26b2o26b2o6$3o2bo23bo2b3o25b3o21b3o$4bo
25bo29bo25bo$4b3o21b3o25b3o2bo23bo2b3o5$4b3o25b3o25b3o25b3o5$8b2o26b2o
26b2o26b2o$8b2o26b2o26b2o26b2o!
Note that these switch engines need a fifth blonk in order to be stabilized.
I Like My Heisenburps! (and others)

User avatar
codeholic
Moderator
Posts: 1147
Joined: September 13th, 2011, 8:23 am
Location: Hamburg, Germany

Re: Making switch-engines

Post by codeholic » November 28th, 2014, 3:31 am

I suspect, that this phrase

Code: Select all

[H]owever, there are five known arks in total, two of which (including Noah's ark) are symmetric.
is totally outdated.
Ivan Fomichev

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

Re: Making switch-engines

Post by simsim314 » November 29th, 2014, 7:00 pm

codeholic wrote:there are five known arks in total
I'm not sure about this phrase. It's possible that if two switch engines are in the same phase it might be true, maybe this is how "Ark" should be defined.

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

Re: Making switch-engines

Post by chris_c » October 3rd, 2017, 7:05 pm

Since the 1x7242 quadratic growth just got a mention I am belatedly showing an improvement. The file was dated Oct 2015 on my machine but I never posted it:

Code: Select all

x = 2596, y = 1, rule = B3/S23
6ob4o2b4o30b4ob9ob3ob3o112b4o3b12o3b3o139b3ob5ob4o5b5o71b3o2b5ob5o11b
4o6b7ob4o134b3o3b8o5b3ob4o19b5o2b3ob5ob7o58b4ob5ob3ob3o2b6o11b4o2b4ob
6ob3o61b3ob3o4b3ob7o181b4o4b3o3b9o19b3ob3o3b7o3b3o57b5ob5o2b3o112b3ob
4o2b3ob3ob9o62b4o4b4ob7o48b7o6b9o13b13ob3o2b4o90b3o6b5ob4ob6o3b6ob5o5b
5o53b11ob3o5b5o47b3ob3ob9ob4o277b3o3b8o5b3ob4o19b5o2b3ob5ob7o319b4o2b
4ob6ob3o61b3ob3o4b3ob7o!

Post Reply