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

Adjustable high velocity ships discussion

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

Adjustable high velocity ships discussion

Postby simsim314 » April 20th, 2016, 4:32 pm

So now we have an adjustable orthogonal speed ships for speeds < 0.25c (caterloopillar) and universal helix that can move at any speed < 0.5c.

I think it's time to discuss how can we solve oblique and orthogonal ships for high speeds (close to 0.5c).

Oblique

For oblique direction which is not orthogonal I have pretty satisfactory design I call spearhead. You can see illustration here, but in this description I use *WSS to code information instead of gliders. This makes sure no signal will interest any reading heads (notice that the oblique direction doesn't need helix).

It consists of:

1. Three stages reading head.
2. Spear shape. In left side of which are vertically moving reading heads, and on the right dried reading heads.
3. Chain reaction of ignition on the bottom.
4. Self destruct.

1. The Reading head.

- Part A: The reading head is moving vertically forward, converts SL into *WSS that moves horizontally to the right.
- Part B: At some point the horizontally moving *WSS converts into glider that moves forward and right.
- Part C: The glider converted back into SL and additional *WSS which moves to the right as well.

The *WSS from C is coding the slow salvo construction which is now done by *WSS instead of glider.

2. The Spearhead Shape.
The reading heads are located on the left, on the right located the dried heads which are separated well by good margin vertically and horizontally.

The dried construction (dx, dy) is the same as (dx, dy) of the moving SLs.

3. The chain reaction
The chain reaction is the way each dry construction activates the next one on the bottom of the ship.

4. Self destruct.
The self destruct works in the same manner as the reading, but destructively.

----------------------------

Orthogonal

Now all this will not work for orthogonal case, because all the dried heads should be on the same lane.

Here are two suggestions:

Idea1: SL -> G-> Universal Helix -> G -> SL

This is somewhat crazy idea, but it should work. We can make *WSS constellation that converts SL into glider salvo that creates the universal helix that after a while places the SL some distance away while shooting a glider.

This will require hundreds of *WSS but the conversion should be possible, and all the construction design of *WSS salvos is already covered by caterloopillar project.

This is similar to caterloopillar, with only one minor detail: the downstream is much simpler, because we don't need to usethe universal helix.

Idea2: Slow down *WSS stream by conduit

Another idea is to have a conduit as speed regulator: *WSS -> Conduit -> *WSS + Glider

The construction done by gliders coming out of the conduit - is of the universal helix.

The the helix is in the front is made to trigger a set of *WSS salvos, fanouts etc. to construct the conduit int the front. It's somewhat more similar to the regular design of caterpillars, but now we need to construct a whole conduit.

To be more precise on the idea - we will need to have two *WSS streams and two conduits and the coding will be done by the timing differences of the output gliders.

Idea 3: Adaption of oblique design.

We need few things:

1. A destructive conversion of reading head into universal helix. This is the "end phase" of the reading head, i.e. instead of just self destruct it becomes a universal helix.

2. The universal helix in the front is making a seed that moves with required period.

3. Many *WSS salvos resulting from the coding tape, are playing with the glider, using fanout etc.to generate slow salvo recipe of the dried head.
Last edited by simsim314 on April 21st, 2016, 9:04 am, edited 3 times in total.
User avatar
simsim314
 
Posts: 1539
Joined: February 10th, 2014, 1:27 pm

Re: Adjustable high velocity ships discussion

Postby biggiemac » April 20th, 2016, 9:23 pm

A few remarks, questions, etc:

Oblique:
First to make sure I get it. So at the bottom of the design, the newly created reading head displaces the SL by (x,y) and fires a signal. This signal hits a frozen constellation designed to produce a full reading head (not 1 *WSS at a time like the caterloopillar). This reading head is aligned with the displaced SL to perform the reaction again T gens later, for an (x,y)c/T spaceship.

Then, questions. The SL to be read needs to be directly above the newly produced reading head, so the thing that gets it there needs to be considered. Should this be accomplished via glider + *WSS flyby or by a 2 glider collision? In the former, you might need to worry about making sure the *WSS from the parent head safely passes through the child's seed before ignition. In the latter there's a timing challenge but with enough duplication and stalling it should be generally solvable. Also, you mentioned maybe using *WSS for construction, I presume horizontal in the drawing. That seems safer for arbitrary obliqueness than using gliders, because you know you will intersect the growing seed with decent clearance. But that would require a lot of new research into slow *WSS construction and production, correct?

Fast Orthogonal:
First, making sure I get it. The first design scheme is very much like what I had in mind when I left my lengthy comment in the other thread. The other uses a conduit that delays a *WSS by a fixed number of generations - by spacing these conduits out we can make the propagation speed very close to c/2. The conduits are created at the top by a fast puffer helix, where every cycle leaves one conduit as ash. At the bottom, they are cleaned up by the negative helix.

Questions. The conduits are used to make slow constructions. But since both the conduit and the growing construction are in the same place every read cycle, the only control we get over the recipe is via timing. So wouldn't we need glider pairs, probably fired by a pair of such conduits, to have universal construction? Or is there a way for the reading head to include a mechanism that freezes the output to a different lane depending on timing, so a second flyby can ignite it for lane-based construction?

I could maybe see that working if the conduit releases a glider (SE), that intersects a NW glider pair produced by some *WSS flyby that is part of the helix, to produce a NE glider on a controlled construction lane. If that causes the construction glider to hit the *WSS then we use SW + NE pair to make a SE instead.

And then, how does the negative helix work? It also needs its signals to propagate at the spaceship speed. So would it construct the mirrored conduit? Or is there a better way? At nearly c/2, math involving c/2 in the other direction could get very extreme.
Physics: sophistication from simplicity.
User avatar
biggiemac
 
Posts: 503
Joined: September 17th, 2014, 12:21 am
Location: California, USA

Re: Adjustable high velocity ships discussion

Postby simsim314 » April 21st, 2016, 1:34 am

biggiemac wrote:The SL to be read needs to be directly above the newly produced reading head, so the thing that gets it there needs to be considered.


I guess you're talking about the chain reaction on the bottom (the "dehydration" phase). The reaction there is "self maintained" like the helix, this reaction occurs by "itself" not using any of the reading head mechanisms, and just ignite in chain. For anything not diagonal, this will need to occur by sending destructive *WSS to the next "dried" head.

To be more clear: there is some empty space on the bottom, where no reading is occurring, only chain reaction of "dehydration" of the dried heads, that start to move forward.

biggiemac wrote:That seems safer for arbitrary obliqueness than using gliders, because you know you will intersect the growing seed with decent clearance. But that would require a lot of new research into slow *WSS construction and production, correct?


Yes this is exactly the idea. Using *WSS instead of gliders, thus assuring good clearance. No "intersection" between the coding stream and the reading heads is occurring.

About the "research" - first of all Chris Cain already done some of it (not sure how much). Another point is that all the movement recipes in caterloopillar were constructed by simple script that ran for a week in golly, that was adaptation of Glue. The *WSS edge shooting recipes were also done by custom written C++ LifeAPI app, that also ran for a week and found all the monochromatic edge shooters.

So this kind of research is expected to be done for such a project - I really don't worry about it.

biggiemac wrote: So wouldn't we need glider pairs, probably fired by a pair of such conduits, to have universal construction


Yes. I mention this in my post. Another option of coding is correct timing on the same lane of the same conduit, less probable but also possible. Maybe also there is some trick like in demonoid 0-hd, using the same conduit twice to code the next and the previous operation. But the standard glider pair - seems most reasonable to me.

biggiemac wrote:And then, how does the negative helix work?


There is no negative helix - self destruct is part of the conduit. Ignition of self destruct is done by shooting a glider "backwards" toward the conduit to destroy it - remember we have universal constructor and the conduit isn't really moving.
User avatar
simsim314
 
Posts: 1539
Joined: February 10th, 2014, 1:27 pm

Re: Adjustable high velocity ships discussion

Postby simsim314 » April 21st, 2016, 2:01 am

Orthogonal idea 3: Adaption of oblique design.

We need few things:

1. A destructive conversion of reading head into universal helix. This is the "end phase" of the reading head, i.e. instead of just self destruct it becomes a universal helix.

2. The universal helix in the front is making a seed that moves with required period.

3. Many *WSS salvos resulting from the coding tape, are playing with the glider, using fanout etc.to generate slow salvo recipe of the dried head.

EDIT

Looks to me that the spearhead design is the simplest of all. The universal helix is constructed in destructive manner, using *WSS salvo (not slow salvo, and not non destructive).

Also thinking of it - universal diagonality is degenerate case of the oblique, skipping the SL->WSS stage (SL -> glider -> SL + *WSS). So I think as this is kinda winning design anyway, we should start from diagonality first.

Any alternative ideas?

EDIT2 In oblique case there is an issue: the first SL for the salvo. As it moves in oblique direction, glider and *WSS is not an option. It looks like destructive helix is needed there as well. Only for diogonal case we can send simple edgy glider.
User avatar
simsim314
 
Posts: 1539
Joined: February 10th, 2014, 1:27 pm

Re: Adjustable high velocity ships discussion

Postby simsim314 » April 21st, 2016, 5:17 am

So what is needed for this project?

1. Non destructive *WSS salvo + glider -> orthogonal *WSS + SL.
2. Monochromatic same state *WSS Slow salvo recipes with the *WSS from (1) + edge shooting glider.
3. A way to dry out *WSS salvo.
4*. Destructive recipe of universal helix from *WSS salvo. (oblique + orthogonal).
5*. A way to adjust glider lane using *WSS + split a glider (orthogonal).

EDIT Obviously step 4 requires:
1. Synthesis by well separated gliders of universal helix.
2. Destructive conversion of *WSS stream into glider synthesis.

Both of the problems are not trivial at all.
User avatar
simsim314
 
Posts: 1539
Joined: February 10th, 2014, 1:27 pm

Re: Adjustable high velocity ships discussion

Postby simsim314 » April 23rd, 2016, 2:34 pm

Here is something very promising for the non-destructive G->LWSS:

x = 38, y = 123, rule = B3/S23
b3o$3bo$2bo11$26bo$25b3o$24b2obo$24b3o$24b3o$24b3o$25b2o7$7bo$6b3o$6bo
b2o$7b3o$7b3o$7b3o$7b2o3$2b3o$bo2bo$4bo$o3bo$4bo$bobo74$4bo$3b3o$3bob
2o26b3o$4b3o26bo2bo$4b2o27bo$33bo3bo$33bo3bo$33bo$34bobo!


EDIT This one even cleaner but the direction is wrong:

x = 60, y = 94, rule = B3/S23
5$11b3o$13bo$12bo15$37bo$36b3o$36bob2o$37b3o$37b3o$37b3o$37b2o3$39bo$
38b3o$37b2obo$37b3o$37b3o$37b3o$38b2o3$21bo$20b3o$20bob2o$21b3o$21b3o$
21b3o$21b2o33$42b3o$41bo2bo$44bo$40bo3bo$40bo3bo$44bo$41bobo!


EDIT2 Complete cycle of reading Block -> Glider -> LWSS + Block (11 *WSS with recovery 64 ticks), not much larger than caterloopillar's:

x = 117, y = 355, rule = B3/S23
24b2o$24b2o31$24b2o$24b2o31$24b2o$24b2o14$18b3o$18bo2bo7b3o$18bo10bo2b
o$18bo3bo6bo$18bo10bo3bo$19bobo7bo3bo$29bo$30bobo4$33b3o$33bo2bo$33bo$
33bo3bo$33bo3bo$33bo$34bobo64$3o$o2bo$o$o$bobo22$106b3o$105bo2bo$108bo
$104bo3bo$104bo3bo$108bo$105bobo7$87b3o$87bo2bo$87bo$87bo3bo$87bo3bo$
87bo$88bobo2$84bo$83b3o$83bob2o$84b3o$84b3o$84b2o75$84b3o$84bo2bo27bo$
84bo29b3o$84bo28b2obo$85bobo25b3o$113b3o$113b3o$114b2o33$80bo$79b3o$
79bob2o$80b3o$80b3o$80b3o$80b2o15$97b3o$96bo2bo$99bo$95bo3bo$95bo3bo$
99bo$96bobo!
User avatar
simsim314
 
Posts: 1539
Joined: February 10th, 2014, 1:27 pm

Re: Adjustable high velocity ships discussion

Postby simsim314 » April 24th, 2016, 4:51 pm

Time for some lwss slow salvo.

Here is 5 lwss edgy glider:

x = 577, y = 18, rule = LifeHistory
317.A2.A$316.A$316.A3.A$316.4A3$61.A2.A$60.A$2A58.A3.A$2A58.4A$189.A
2.A$188.A$188.A3.A252.A2.A$188.4A252.A$444.A3.A124.A2.A$444.4A124.A$
572.A3.A$572.4A!


And the sequence is:

-2,2,-8,4,6


EDIT And a block splitter for 5 lwss:

x = 577, y = 14, rule = B3/S23
189bo2bo124bo2bo$188bo127bo$188bo3bo123bo3bo124bo2bo$188b4o124b4o124bo
$444bo3bo$444b4o$61bo2bo$60bo$60bo3bo508bo2bo$60b4o508bo$572bo3bo$572b
4o$2o$2o!


-6,-12,-12,-10,-4


I will post the recipes in other direction because the script is working this way. I'll post all block->SL recipes with all 4 variants of block locations together with the script, when the script will finish.
User avatar
simsim314
 
Posts: 1539
Joined: February 10th, 2014, 1:27 pm

Re: Adjustable high velocity ships discussion

Postby simsim314 » April 25th, 2016, 12:11 pm

Good news, no synthesis issues are expected. Notice that to dry out the synth the gliders should be more spread in each direction, but it doesn't look as a real issue.

Here're the synth of two places where there could be an issue:

x = 164, y = 355, rule = B3/S23
24b2o$24b2o31$24b2o$24b2o12$57bobo$58b2o$58bo3$55bobo$56b2o$56bo$87bo$
85b2o$86b2o3$89bo$87b2o$88b2o4$24b2o$24b2o6$80b2o$80bobo$80bo6$18b3o$
18bo2bo7b3o$18bo10bo2bo29b3o$18bo3bo6bo34bo$18bo10bo3bo29bo$19bobo7bo
3bo$29bo$30bobo4$33b3o$33bo2bo$33bo$33bo3bo$33bo3bo$33bo$34bobo64$3o$o
2bo$o160bo$o158b2o$bobo156b2o3$163bo$161b2o$162b2o17$106b3o$105bo2bo$
108bo$104bo3bo$104bo3bo27b3o$108bo29bo$105bobo29bo7$87b3o$87bo2bo$87bo
$87bo3bo$87bo3bo$87bo$88bobo2$84bo59bo$83b3o57b3o$83bob2o56bob2o$84b3o
57b3o$84b3o57b3o$84b2o58b2o75$84b3o$84bo2bo27bo$84bo29b3o$84bo28b2obo$
85bobo25b3o$113b3o$113b3o$114b2o33$80bo$79b3o$79bob2o$80b3o$80b3o$80b
3o$80b2o15$97b3o$96bo2bo$99bo$95bo3bo$95bo3bo$99bo$96bobo!


EDIT The full front synthesis - I hope the close glider pair can be somehow squeezed there using SL trickery, other than that - the synth looks good:

x = 310, y = 418, rule = LifeHistory
C.C$.2C$.C23$285.C.C$285.2C$286.C22$27.C.C$28.2C$28.C3$25.C.C$26.2C$
26.C$307.C$305.2C$306.2C3$309.C$307.2C$308.2C99$5.C.C$6.2C$6.C119$
288.2C$287.2C$289.C24$.C$.2C$C.C13$300.2C$300.C.C$300.C8$32.3C$34.C$
33.C74$262.2C$262.C.C$262.C2$10.C$10.2C$9.C.C!


EDIT2 The full back part synthesis:

x = 368, y = 573, rule = LifeHistory
365.C$363.2C$364.2C3$367.C$365.2C$366.2C43$39.C$40.2C$39.2C3$37.C$38.
2C$37.2C15$17.C$18.C$16.3C99$55.C$53.C.C$54.2C3$53.C$51.C.C$52.2C4$
23.C.C$24.2C$24.C104$302.C$302.C.C$302.2C3$304.C$304.C.C$304.2C13$
320.C.C$320.2C$321.C3$322.C.C$322.2C$323.C51$3C$2.C$.C6$360.3C$360.C$
361.C19$17.C$17.2C$16.C.C$340.C$339.2C$339.C.C80$330.2C$330.C.C$330.C
$362.2C$28.C332.2C$28.2C333.C$27.C.C62$46.2C$47.2C$46.C18$64.2C$63.C.
C$65.C!


We left with drying out the synths - and apply lwss slow salvo recipes to them.

EDIT3 Woops forgot about the self destruct.
User avatar
simsim314
 
Posts: 1539
Joined: February 10th, 2014, 1:27 pm

Re: Adjustable high velocity ships discussion

Postby simsim314 » April 28th, 2016, 8:58 am

Here is few full cycles with self destruct (16 *WSS - even smaller than caterloopillar, unfortunately it will need to be dried out - so it's about order of magnitude larger than the caterloopillar):

x = 953, y = 2906, rule = B3/S23
69bo$68bobo$68bobo$69bo2$64b2o7b2o$63bo2bo5bo2bo$64b2o7b2o2$69bo$68bob
o$68bobo$69bo118$24b2o$24b2o31$24b2o$24b2o31$24b2o$24b2o31$24b2o$24b2o
31$24b2o$24b2o9$72b3o$71bo2bo$74bo$70bo3bo$74bo$71bobo17$24b2o$24b2o$
91b3o$90bo2bo$93bo$89bo3bo$93bo$90bobo8$18b3o$18bo2bo7b3o$18bo10bo2bo$
18bo3bo6bo$18bo10bo3bo$19bobo7bo3bo$29bo$30bobo4$33b3o$33bo2bo$33bo$
33bo3bo$33bo3bo$33bo$34bobo5$104b3o$104bo2bo$104bo$104bo3bo$104bo$105b
obo54$3o$o2bo$o$o$bobo22$106b3o$105bo2bo$108bo$104bo3bo$104bo3bo$108bo
$105bobo7$87b3o$87bo2bo$87bo$87bo3bo$87bo3bo$87bo$88bobo2$84bo$83b3o$
83bob2o$84b3o$84b3o$84b2o74$276bo$84b3o188b3o$84bo2bo27bo159bob2o$84bo
29b3o159b3o$84bo28b2obo159b3o$85bobo25b3o160b2o$113b3o$113b3o$114b2o
33$80bo$79b3o$79bob2o$80b3o$80b3o$80b3o$80b2o15$97b3o$96bo2bo$99bo$95b
o3bo$95bo3bo$99bo$96bobo125$105b3o$105bo2bo7b3o$105bo10bo2bo$105bo3bo
6bo$105bo10bo3bo$106bobo7bo3bo$116bo$117bobo4$120b3o$120bo2bo$120bo$
120bo3bo$120bo3bo$120bo$121bobo64$87b3o$87bo2bo$87bo191b3o$87bo190bo2b
o$88bobo190bo$277bo3bo$281bo$278bobo19$298b3o$297bo2bo$300bo$296bo3bo$
300bo$297bobo3$109b3o$108bo2bo$111bo$107bo3bo$111bo$108bobo22$311b3o$
311bo2bo$311bo$311bo3bo$311bo$312bobo80$313b3o$312bo2bo$315bo$311bo3bo
$311bo3bo$315bo$312bobo7$294b3o$294bo2bo$294bo$294bo3bo$294bo3bo$294bo
$295bobo2$291bo$290b3o$290bob2o$291b3o$291b3o$291b2o74$483bo$291b3o
188b3o$291bo2bo27bo159bob2o$291bo29b3o159b3o$291bo28b2obo159b3o$292bob
o25b3o160b2o$320b3o$320b3o$321b2o33$287bo$286b3o$286bob2o$287b3o$287b
3o$287b3o$287b2o15$304b3o$303bo2bo$306bo$302bo3bo$302bo3bo$306bo$303bo
bo125$312b3o$312bo2bo7b3o$312bo10bo2bo$312bo3bo6bo$312bo10bo3bo$313bob
o7bo3bo$323bo$324bobo4$327b3o$327bo2bo$327bo$327bo3bo$327bo3bo$327bo$
328bobo64$294b3o$294bo2bo$294bo191b3o$294bo190bo2bo$295bobo190bo$484bo
3bo$488bo$485bobo19$505b3o$504bo2bo$507bo$503bo3bo$507bo$504bobo3$316b
3o$315bo2bo$318bo$314bo3bo$318bo$315bobo22$518b3o$518bo2bo$518bo$518bo
3bo$518bo$519bobo80$520b3o$519bo2bo$522bo$518bo3bo$518bo3bo$522bo$519b
obo7$501b3o$501bo2bo$501bo$501bo3bo$501bo3bo$501bo$502bobo2$498bo$497b
3o$497bob2o$498b3o$498b3o$498b2o74$690bo$498b3o188b3o$498bo2bo27bo159b
ob2o$498bo29b3o159b3o$498bo28b2obo159b3o$499bobo25b3o160b2o$527b3o$
527b3o$528b2o33$494bo$493b3o$493bob2o$494b3o$494b3o$494b3o$494b2o15$
511b3o$510bo2bo$513bo$509bo3bo$509bo3bo$513bo$510bobo125$519b3o$519bo
2bo7b3o$519bo10bo2bo$519bo3bo6bo$519bo10bo3bo$520bobo7bo3bo$530bo$531b
obo4$534b3o$534bo2bo$534bo$534bo3bo$534bo3bo$534bo$535bobo64$501b3o$
501bo2bo$501bo191b3o$501bo190bo2bo$502bobo190bo$691bo3bo$695bo$692bobo
19$712b3o$711bo2bo$714bo$710bo3bo$714bo$711bobo3$523b3o$522bo2bo$525bo
$521bo3bo$525bo$522bobo22$725b3o$725bo2bo$725bo$725bo3bo$725bo$726bobo
80$727b3o$726bo2bo$729bo$725bo3bo$725bo3bo$729bo$726bobo7$708b3o$708bo
2bo$708bo$708bo3bo$708bo3bo$708bo$709bobo2$705bo$704b3o$704bob2o$705b
3o$705b3o$705b2o74$897bo$705b3o188b3o$705bo2bo27bo159bob2o$705bo29b3o
159b3o$705bo28b2obo159b3o$706bobo25b3o160b2o$734b3o$734b3o$735b2o33$
701bo$700b3o$700bob2o$701b3o$701b3o$701b3o$701b2o15$718b3o$717bo2bo$
720bo$716bo3bo$716bo3bo$720bo$717bobo125$726b3o$726bo2bo7b3o$726bo10bo
2bo$726bo3bo6bo$726bo10bo3bo$727bobo7bo3bo$737bo$738bobo4$741b3o$741bo
2bo$741bo$741bo3bo$741bo3bo$741bo$742bobo64$708b3o$708bo2bo$708bo191b
3o$708bo190bo2bo$709bobo190bo$898bo3bo$902bo$899bobo19$919b3o$918bo2bo
$921bo$917bo3bo$921bo$918bobo3$730b3o$729bo2bo$732bo$728bo3bo$732bo$
729bobo22$932b3o$932bo2bo$932bo$932bo3bo$932bo$933bobo80$934b3o$933bo
2bo$936bo$932bo3bo$932bo3bo$936bo$933bobo7$915b3o$915bo2bo$915bo$915bo
3bo$915bo3bo$915bo$916bobo2$912bo$911b3o$911bob2o$912b3o$912b3o$912b2o
75$912b3o$912bo2bo27bo$912bo29b3o$912bo28b2obo$913bobo25b3o$941b3o$
941b3o$942b2o33$908bo$907b3o$907bob2o$908b3o$908b3o$908b3o$908b2o15$
925b3o$924bo2bo$927bo$923bo3bo$923bo3bo$927bo$924bobo125$933b3o$933bo
2bo7b3o$933bo10bo2bo$933bo3bo6bo$933bo10bo3bo$934bobo7bo3bo$944bo$945b
obo4$948b3o$948bo2bo$948bo$948bo3bo$948bo3bo$948bo$949bobo64$915b3o$
915bo2bo$915bo$915bo$916bobo30$937b3o$936bo2bo$939bo$935bo3bo$939bo$
936bobo!
User avatar
simsim314
 
Posts: 1539
Joined: February 10th, 2014, 1:27 pm


Return to Patterns

Who is online

Users browsing this forum: No registered users and 2 guests