(23,5)c/79 spaceship components

For discussion of specific patterns or specific families of patterns, both newly-discovered and well-known.
User avatar
dvgrn
Moderator
Posts: 10612
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI
Contact:

Re: (23,5)c/79 spaceship components

Post by dvgrn » December 3rd, 2014, 8:19 pm

codeholic wrote:Overall, it seems that the monster is going to be the most alien creature I've seen so far in Conway's Game of Life.
Yup, just looking at that lead Herschel up at the top, getting reset by a loaf one time and a glider the next, is enough to give the impression that stranger and more wonderful things will follow. And then they do!

I was happy with one prime number in the 31c/240 spaceship project, but (23,5)/79 has three. I might possibly have a shot at understanding how to do the synchronization and matching-up, based on recent experience with 31c/240... but thankfully that was a lot simpler -- each forerake and backrake glider in the same phase, etc. So I might need an explanation using very short words.
simsim314 wrote:As the currently existing knightships are moving very slowly, and current caterpillars designs are not so "knightship" friendly (the raising helix is usually "edgy" - and works only with orthogonal ships), in this sense biggiemac design is real break-through.
It is indeed: this impressive creature will be the first non-half-baked knightship, you might say. The HBK designs took advantage of a lucky low-level reaction that made the construction enormously easier. But they're one-of-a-kind -- we'll have a very hard time build anything else like them, unless we find some other object or constellation that moves when a glider hits it (and emits a same-direction glider).
biggiemac wrote:If you can give me advice as to how to illustrate this in LifeHistory that would be fantastic, as I get the feeling that my explanations thus far are not in satisfyingly clear language..
You're doing great so far (I guess the words are short enough)... but some kind of diagram with labels and arrows and circled structures would be a wonderful addition.

Attached is one way to use LifeHistory states 3 through 5 to annotate large patterns. It's not the only way -- you can draw freehand in state 4 instead, if you want -- but it works, more or less. There are some pre-made labels that can be cut and pasted... Golly isn't really designed for this kind of annotation/drawing task, so for example you have to be careful not to draw over any state-1 cells with state 4, or you'll ruin your pattern. Marked state-1 cells would have to be converted to state 3 or 5 somehow.
knightship2synth-with-labels.mc.gz
knightship with sample labels (but not good ones)
(78.46 KiB) Downloaded 548 times
The vaguely nice thing about state 4 is that you can still run the pattern and see what happens -- the labels don't get in the way of the action. They don't move, so they quickly become irrelevant, of course, and in a pattern like this the universe quickly gets smeared with state-2 history cells. But if you've marked the labeled object with state 3 or 5, then at least you have the afterimage of the object as a reference.

[Annotations in Conway's Life have been a work in progress for decades -- despite several attempts, there's never been an annotation format supported by a large enough fraction of CA software to really catch on. Maybe someday we'll be able to define labels that move at (23,5)/79, and stay the same size no matter how you zoom.]

EDIT: Meant to ask: if each construction triangle is 50% bigger than the last one, and there are 8 *WSSes left to build, does that mean the knightship's final height will be around 2500+665(1+1.5+1.5^2+1.5^3+...+1.5^9) = ~ 750k? (I'm guessing 77,777.) Or are there subsidiary problems to solve that will take more space?

User avatar
biggiemac
Posts: 515
Joined: September 17th, 2014, 12:21 am
Location: California, USA

Re: (23,5)c/79 spaceship components

Post by biggiemac » December 3rd, 2014, 10:58 pm

Thanks for the LifeHistory tip!

As for the final size, it depends on whether I can get the MWSS prep in the same space. Otherwise, that's my first-order guess as well. Although the amount of space it takes to produce the side LWSS stream has been the constraint thus far, so it could be smaller.

Cleanup should be straightforward, the 263 gives a kickback, the 2s can rephase and then delete the lower one so that the end position gives a vanishing reaction from the 31 or 316. I'll try to see if the last glider stream the left cluster creates can spawn the NW kickback lane to make the back end.

For what it's worth, the primes don't work in our favor. The real numbers are 342 (number of distinct track separations), 28 and 18 (number of half-diagonals between stills), and they don't work well with the base reaction, which changes the track by 95 (19 in common with 342) and the hd by 3..
Physics: sophistication from simplicity.

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

Re: (23,5)c/79 spaceship components

Post by codeholic » December 4th, 2014, 3:10 am

simsim314 wrote:Why exactly? Am I missing something?
Now when I'm thinking about it I can't understand why I thought it was impossible. Maybe I just got confused.
simsim314 wrote:You need three directions - two direction for collision, and another for helix stream output. You can have two glider streams that generate *WSS stream, you can have two *WSS stream that generate glider stream, you can have *WSS + glider stream that generate glider or *WSS helix. But you can't manage with gliders alone.
I don't see a problem. Let's assume helix gliders are moving NW. Then the shuttle glider would go NE-SW, and the output glider could go either SW or NE (in the opposite direction of the shuttle glider).
simsim314 wrote:And gliders limited to c/4 - but other than that, you can use gliders of course.
That's true. That's why I had to build a helix out of spaceships for (23,5)c/79. But I actually considered a glider helix for 31c/240 (but as you know, it appeared one didn't need a helix at all). So maybe glider helices are not that important, as you can (always?) replace them with glider-spaceship collisions.
Ivan Fomichev

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

Re: (23,5)c/79 spaceship components

Post by codeholic » December 4th, 2014, 4:09 am

dvgrn wrote:EDIT: Meant to ask: if each construction triangle is 50% bigger than the last one, and there are 8 *WSSes left to build, does that mean the knightship's final height will be around 2500+665(1+1.5+1.5^2+1.5^3+...+1.5^9) = ~ 750k? (I'm guessing 77,777.) Or are there subsidiary problems to solve that will take more space?
That's in fact the reason why I proposed to spawn new tracks, when old ones drift away too much from the helix, in the original design. Of course, one can build a huge triangular spaceship without doing it, but a linear ship might actually turn out to be an order of magnitude smaller.
Ivan Fomichev

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

Re: (23,5)c/79 spaceship components

Post by codeholic » December 4th, 2014, 4:24 am

biggiemac wrote:Hope this helps!
Thanks for your explanations, it helped a lot to understand your track design!

Have you already thought of a back-end for the spaceship, by the way?
Ivan Fomichev

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

Re: (23,5)c/79 spaceship components

Post by simsim314 » December 4th, 2014, 7:13 am

biggiemac wrote: The pairs can be created from a normal rake + kickback like the pairs for the LWSS, the third relies on the slow salvo inserter posted elsewhere (thanks simsim314!)
You don't need this "crazy" edge shooter, and it costs to you too much. Simple blockic that needs only 4 gliders will work for what you need:

Code: Select all

x = 122, y = 80, rule = B3/S23
65bo$66bo$64b3o19$12bo$4bo5bobo$2bobo6b2o$3b2o$95b2o$95b2o3$90b2o$90b
2o$8b3o$10bo$9bo5$3o$2bo$bo3$79bo$80b2o$79b2o3$90b2o$90b2o3$85b2o$85b
2o7$55b2o$56b2o$55bo$70b2o18b2o$60b2o7bobo19b2o$61b2o8bo18bo$60bo24bo
19b2o$75b2o8b2o17bobo$74bobo7bobo19bo$76bo43bo$90bo8b3o18b2o$90b2o9bo
17bobo$89bobo8bo$114b2o$104b3o8b2o$106bo7bo$105bo$119b2o$120b2o$119bo!
@dvgrn - This is an example where good query API would help a lot. Instead of using N prepared recipes, with good performance but great cost, one can just run a query on "db" of reflectors (I've around 4K of them), and find the one that suites him. For example biggiemac could ask for reflectors that fit into his salvo.

EDIT By the way I've a name for the big triangle design - Pyramid.

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

Re: (23,5)c/79 spaceship components

Post by codeholic » December 4th, 2014, 8:45 am

simsim314 wrote:EDIT By the way I've a name for the big triangle design - Pyramid.
Having in mind exponential growth it exposes with every next step, I'd rather call it the Tower of Hanoi ;)
Ivan Fomichev

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

Re: (23,5)c/79 spaceship components

Post by simsim314 » December 4th, 2014, 9:06 am

@dvgrn - do we have some sort of script for writing labels in LifeHistory?

User avatar
dvgrn
Moderator
Posts: 10612
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI
Contact:

Re: (23,5)c/79 spaceship components

Post by dvgrn » December 4th, 2014, 10:23 am

codeholic wrote:
simsim314 wrote:EDIT By the way I've a name for the big triangle design - Pyramid.
Having in mind exponential growth it exposes with every next step, I'd rather call it the Tower of Hanoi ;)
The triangle made me think along the lines of "Delta" -- maybe Delta23-5... Luckily the name isn't my decision!
simsim314 wrote:@dvgrn - do we have some sort of script for writing labels in LifeHistory?
Not really. Sometimes I use a little one-off Python snippet run from the clipboard:

Code: Select all

import golly as g
from glife.text import make_text
t = make_text("Put label text here", "mono")
g.putcells(t)
If I want the label in the LifeHistory background (2) or marked-OFF (4) color, or want to convert patches of cells to marked-ON (3 or 5), to highlight them or to get an after-image after running the pattern, I use this all-purpose script that works on the current selection, or the whole pattern if there's no selection:

ToChangeColors.py:

Code: Select all

# Switch all cells of a clicked-on state in the selected pattern to the current drawing state

import golly as g

r = g.getselrect()
if len(r)==0:
  r = g.getrect()
  if len(r)==0:
    g.exit("No pattern to change.")

# set fill limits to the pattern's bounding box
minx = r[0]
miny = r[1]
maxx = minx + r[2] - 1
maxy = miny + r[3] - 1

area = g.getcells(r)
if len(area)%2 == 0:
  g.exit("Nothing to do here -- rule is not multistate.")

g.show("Click on a location to choose the cell color to change.")
newstate = g.getoption("drawingstate")
oldstate = newstate

# wait for user to click a cell
g.show("Click a sample of the color you wish to change to the current drawing state, " + str(newstate))
while oldstate == newstate:
   event = g.getevent()
   if event.startswith("click"):
      # event is a string like "click 10 20 left none"
      evt, xstr, ystr, button, mods = event.split()
      x = int(xstr)
      y = int(ystr)
      if x < minx or x > maxx or y < miny or y > maxy:
         g.warn("Click inside the current selection, or inside the pattern's bounding box.")
      else:
         # note that user might have changed drawing state
         newstate = g.getoption("drawingstate")
         oldstate = g.getcell(x, y)
         if oldstate == newstate:
            g.warn("The clicked cell must have a different state\n"+
                   "from the current drawing state.")
   else:
      g.doevent(event)

# tell Golly to handle all further keyboard/mouse events
g.getevent(False)

for i in range(0,len(area)-2,3):
  if area[i+2]==oldstate:
    g.setcell(area[i],area[i+1],newstate)

g.show("")
This goes along with other scripts to switch back and forth between LifeHistory and regular Life, which I have mapped to Alt+H and Alt+J respectively:

ToLifeHistory.py:

Code: Select all

# Switch current pattern to five-state LifeHistory rule
import golly as g
g.setrule("LifeHistory")
ToLife.py:

Code: Select all

# Switch current pattern to standard HashLife algorithm and Conway's Life rule
import golly as g

if g.getrule()[:11]=="LifeHistory" or g.getrule()[:14]=="HistoricalLife":
   # unmark any marked cells
   cells=g.getcells(g.getrect())
   for i in xrange(2,len(cells), 3):
      if cells[i]!=cells[i]%2: g.setcell(cells[i-2],cells[i-1],cells[i]%2)
g.setrule("Life")
g.setalgo("HashLife")
For large labels, like the 72-point ones in the knightship sample pattern, I just type whatever I want in Paint, black letters on white background, copy the image out, and paste straight into Golly.

Then select and use ToChangeColors.py if needed. The same script also works to get rid of state-2 LifeHistory envelope cells -- change them to state 0. It's very inefficient, so for really big pattern conversions it's much faster to use a custom rule to remove or convert states, but this works fine for most purposes.

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

Re: (23,5)c/79 spaceship components

Post by simsim314 » December 4th, 2014, 1:45 pm

dvgrn wrote:For large labels, like the 72-point ones in the knightship sample pattern, I just type whatever I want in Paint, black letters on white background, copy the image out, and paste straight into Golly.
Wow this is great tip! I was trying to do the same trick in python unfortunately GIL made it hell. But instead of using paint one can use GIL functions to place text on image and then paste it in golly (or convert it into golly format).

Anyway here is small script I've made to place mono text from glife.text in state 4 in golly (places the text in location of the selection - if no selection in (0,0):

Code: Select all

import golly as g
from glife.text import make_text

if g.numstates() - 1 < 4:
	g.exit("Please select LifeHistory rule")

val = g.getstring("Enter string to place in golly", "Write here", "String Selection") 

t = make_text(val, "mono")
newt = []
cnt = 0 

for i in t:
	cnt += 1
	newt.append(i)
	
	if cnt % 2 == 0:
		newt.append(4)
		
if len(newt) % 2 == 0:
	newt.append(0)
	
if g.getselrect() == []:
	g.putcells(newt)
else:
	rect = g.getselrect()
	g.putcells(newt, rect[0], rect[1])
As for the other scripts you mentioned, here is my version for them:

The Life to LifeHistory converter I use state 3 conversion. It's very useful many times - because usually you use LifeHistory in conjunction with regular Life, and you want to see where the original pattern was placed.

Code: Select all

import golly as g 

g.setrule("LifeHistory")

cells = g.getcells(g.getrect())

for i in xrange(2, len(cells), 3):
	cells[i] = 3
	
g.putcells(cells)
As for LfeHistoty->Life converter, I use setcell instead putcells - this doesn't ruins the selection.

Code: Select all

# Change the LifeHistory N states to Life states and set the rule to Life
# 0 + 2s -> 0
# 1 + 2s -> 1

import golly as g

if not(g.getrule()[:11]=="LifeHistory" or g.getrule()=="HistoricalLife" or g.getrule()=="CBRLife"):
	g.exit()
	
cells = g.getcells(g.getrect())

for i in xrange(2, len(cells), 3):
   
   x = cells[i - 2]
   y = cells[i - 1]
   val = cells[i]
   
   if val % 2 == 1: 
      g.setcell(x, y, 1)
   else:
      g.setcell(x, y, 0)


g.setrule("B3/S23")
g.setalgo("HashLife")
I can't feel any difference performance wise, for most of the patterns I work with, and I don't "think about it" at all, i.e. it doesn't ruin anything (not the selection nor the view etc.).

I'm also not so sure that erasing + putcells is faster than setcell. Anyway to your script I would add the rect = getselrect() and select(rect).

User avatar
biggiemac
Posts: 515
Joined: September 17th, 2014, 12:21 am
Location: California, USA

Re: (23,5)c/79 spaceship components

Post by biggiemac » December 4th, 2014, 4:02 pm

I like the blockic seed much better, glad you pointed that out. Its synthesis also keeps the gliders on the same hd mod 3 with the NE rake! There will need to be 4 rephasings of the 316-2-190-263 in between each NE glider but that's 8 in total, while the other synthesis (which I posted in an edit at the bottom of the last page - I realize now that's the least likely place for it ever to be seen) appeared to need 9 (+6 more if the p2 blinker was the wrong color - I hadn't checked). There's also still the issue of setting up the kickback, as the beehive fuse needs to be on the right hd, and there's only a 1/3 chance the one I used for LWSS synthesis will still work.

Naming.. I had thought delta at times too, trying to make it into a slope pun with no real success. Tower of Hanoi is clever.

As I had mentioned before, I had wanted to try cycling three clusters to keep it from growing exponentially, but that would require the LWSS to pass through more layers of track, and also incur the glider mess from each startup. The current triangular plan is simpler for just making the knightship happen.

Edit: I was wrong, the blockic synthesis does not in fact keep the NE gliders on the same hd mod 3.. so it cannot be constructed from the 316-2-190-263 track. I can put the second block somewhere nearby and nudge it into place though.
Last edited by biggiemac on December 4th, 2014, 5:32 pm, edited 1 time in total.
Physics: sophistication from simplicity.

User avatar
dvgrn
Moderator
Posts: 10612
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI
Contact:

Re: (23,5)c/79 spaceship components

Post by dvgrn » December 4th, 2014, 5:09 pm

biggiemac wrote:As I had mentioned before, I had wanted to try cycling three clusters to keep it from growing exponentially, but that would require the LWSS to pass through more layers of track, and also incur the glider mess from each startup. The current triangular plan is simpler for just making the knightship happen.
I think I see the problem... vaguely, anyway.

To restart tracks farther to the right using current technology, you'd need support for a lead Herschel -- a trail of junk for it to follow, or a SW-ward glider stream (could maybe be created with a kickback?) But you'd also need a second nearby junk trail/glider stream, plus suppression for early output gliders... maybe more than that, I don't know (?).

-- Or else you'd need a double-density stream of forward LWSSes, to do the same glider -> beehive trick as at the current point of the Tower. If you go that route, you have to duplicate the entire two-triangle Tower that's been constructed so far, just to build those extra LWSSes.

That might possibly look like a good deal after you've added several more *WSSes. Triangle #6 will be significantly bigger than the apex of the Tower plus those initial two triangles, I think. So it seems as if a reset at the halfway point, or just possibly two resets at 1/3 and 2/3s of the helix construction might make a smaller ship overall -- two or three connected triangles about the same size, instead of one big one. Four or more mini-Towers wouldn't be worth the reset expense, it looks like.

How much am I missing here? I'm probably measuring something seriously wrong, but it's hard to tell what...!

User avatar
biggiemac
Posts: 515
Joined: September 17th, 2014, 12:21 am
Location: California, USA

Re: (23,5)c/79 spaceship components

Post by biggiemac » December 4th, 2014, 5:49 pm

The creation of the second cluster uses two NE rakes, one to make a kickback and the second to turn the first output glider of the H into a TL, which soon becomes a hive after 1 extra output for parity. Burning this to give the 2-track in a location that isn't consumed by the hives gives 3 more extra outputs. So any new cluster creation in this manner would clean up one SE stream and produce 4 new ones.

We can't make a SE stream that doesn't intersect with the helix syntheses, which is why I needed to clean them up first. So any attempt to reset the clusters to be near the helix would need to begin with a NE rake + leftover stills, rather than a kickback.

The stills are p158. Perhaps we could use rakes to turn the leftover blocks into hives. To give the other input, a p158 NE rake could do a NE G+ LWSS -> SW G reaction. Turning this into a cluster would cost 3 LWSS at best and sounds like a hassle I don't want to worry about for a first go at this ship.

We miiight be able to reset with only a double-density LWSS stream, if we rake it into something useful AND the first glider intersects the stream to give another useful still without destroying it.

Edit: Even better, we can do it with a single p158 LWSS stream, as long as there exists a NE G + N LWSS -> SW G reaction. We get to use as many NE rakes as we need from the existing track to help set up the cluster. So I'll do a reset if I can.
Last edited by biggiemac on December 4th, 2014, 8:31 pm, edited 1 time in total.
Physics: sophistication from simplicity.

User avatar
dvgrn
Moderator
Posts: 10612
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI
Contact:

Re: (23,5)c/79 spaceship components

Post by dvgrn » December 4th, 2014, 6:01 pm

simsim314 wrote:I can't feel any difference performance wise, for most of the patterns I work with, and I don't "think about it" at all, i.e. it doesn't ruin anything (not the selection nor the view etc.).

I'm also not so sure that erasing + putcells is faster than setcell.
I haven't touched that code since I wrote it half a decade ago. A quick semi-scientific test: running the current knightship pattern with labels for 1000 ticks gives quite a number of extra cells to clean up. The old erase + putcells code takes 23.5 seconds on my machine to convert that pattern to simple two-state.

A nice simple version based on your selection, but using setcell() only if a cell actually needs changing, takes ... 25.72 seconds. See, I knew there was a reason I did it the messy complicated way. It just wasn't a very good reason.

The obvious post-apgsearch thing to do is to write a script that creates a LifeHistoryToLife rule (if it doesn't already exist) that does all the appropriate state transitions. Switch to that rule, run for one tick, and then set the generation number back by one tick and switch to the Life rule. For bonus points, set the step size to match the pre-rule-change step size. That ought to be faster by, what, a couple of orders of magnitude?

If only Golly would actually turn off cells with putcells([X,Y,0]) instead of just leaving them the same color, it would be fairly competitive to build a list of cells that need changing and fix them all at once with putcells().

...

[Apologies for the probably unnecessary tangential discussion here. I've updated the ToLife.py code in my previous post, and probably won't feel the need to discuss it any more.]

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

Re: (23,5)c/79 spaceship components

Post by simsim314 » December 5th, 2014, 5:31 am

dvgrn wrote:The obvious post-apgsearch thing to do is to write a script that creates a LifeHistoryToLife...That ought to be faster by, what, a couple of orders of magnitude?
I would add that post-Random Collision Search would be also adding the rule creation into the script, so no one will know there is such rule :)

Here is the script:

Code: Select all

#Blazingly fast LifeHistory to Life converter
#Creates special rule and run it for one generation then switch to Life 
#Replace 2k + 1-> 1 and 2k -> 0
#Preserves step
import golly as g 
import os

if not(g.getrule()[:11]=="LifeHistory" or g.getrule()=="HistoricalLife" or g.getrule()=="CBRLife"):
   g.exit()
   
def CreateRule():
	fname = os.path.join(g.getdir("rules"), "LifeHistory2Life.rule")
	with open(fname, 'wb') as f:
		f.write('@RULE LifeHistory2Life\n')
		f.write('\n')
		f.write('@TABLE\n')
		f.write('n_states:7\n')
		f.write('neighborhood:vonNeumann\n')
		f.write('symmetries:none\n')
		f.write('\n')
		f.write('var a={0,1,2,3,4,5,6}\n')
		f.write('var b={a}\n')
		f.write('var c={a}\n')
		f.write('var d={a}\n')
		f.write('var f={a}\n')
		f.write('\n')
		f.write('1,a,b,c,d,1\n')
		f.write('2,a,b,c,d,0\n')
		f.write('3,a,b,c,d,1\n')
		f.write('4,a,b,c,d,0\n')
		f.write('5,a,b,c,d,1\n')
		f.write('6,a,b,c,d,0\n')
		
		
CreateRule()
g.setrule("LifeHistory2Life")
g.run(1)
step = g.getstep()
g.setrule("B3/S23")
g.setalgo("HashLife")
g.setstep(step)

And it's really blazing fast. I replaced my LifeHistoryToLife.py with it. I will also post it in golly scripts thread.

EDIT Post in golly scripts thread.

EDIT Notice this script is raising the generation by one. It's also sometimes for some reason shows the pattern in LifeHistory2Life rule instead of the original LifeHistory. I didn't manage to reproduce this bug, but it's sometimes happens.

User avatar
biggiemac
Posts: 515
Joined: September 17th, 2014, 12:21 am
Location: California, USA

Re: (23,5)c/79 spaceship components

Post by biggiemac » December 5th, 2014, 2:30 pm

Well, this is awkward. I made the following cute reaction to set up the NE rake for the MWSS synthesis, convinced that I had done my math correctly:

Code: Select all

x = 517, y = 1397, rule = B3/S23
4$258bo$256bobo$257b2o41$271bobo$272b2o$272bo40$287bo$288bo$286b3o14$
371bo$371bobo$371b2o25$301bo$302b2o$301b2o14$346bobo$346b2o$347bo3$
493bobo$493b2o$494bo20$97bo219bo$98b2o215bobo$97b2o217b2o12$502bo$322b
o179bobo$321bo180b2o$321b3o$185bo$186b2o$185b2o282bo$468bo$468b3o21$
113bo216bobo$111bobo217b2o$112b2o217bo12$477bobo$298bo178b2o$296b2o
180bo$297b2o$201bo$199bobo$200b2o243bo$443b2o$444b2o20$346bo$126bobo
218bo$127b2o216b3o$127bo11$273bo179bo$273b2o177bo$275bo176b3o$273b2o$
273bo$214bobo$215b2o58b2o$215bo59b2o142bo$419bobo$419b2o2$266bo$265bo$
265b3o16$142bo217bo$143bo217b2o$141b3o216b2o11$285bo$286bo142bo$284b3o
140b2o$428b2o2$230bo$231bo12b3o$229b3o11bo2bo$243bo4bo145bobo$248bo
145b2o$248b2o145bo$245bob2o$246bo10$234bobo$234b2o$235bo6$156bo219bo$
157b2o215bobo$156b2o217b2o11$299bo$300b2o101bo$218bo80b2o102bobo$217b
3o183b2o$216bo2b2o$216bo3bo23bo$218bobo24b2o$244b2o124bo$215b2ob2o149b
o$217bo151b3o$257bobo$258b2o$258bo$217b4o$217bobobo$220bo4$220bo$221b
2o$220b2o9$172bo216bobo$170bobo30bo186b2o$171b2o30bobo184bo$203b2o10$
191bo123bo$190bobo120bobo62bobo$189bo3bo120b2o62b2o$190bo2bo185bo$190b
o2bo$191bobo66bo$258bobo$259b2o85bo$344b2o$273bo71b2o$274bo$272b3o8$
236bo$234bobo$235b2o7$198bo$199bo205bo$185bobo9b3o206bo$186b2o216b3o$
186bo8$173bo$171b2o$172b2o$328bobo23bo$329b2o22bo$329bo23b3o$167b3o$
167bo2bo$167bo2bo102bobo$166b4o104b2o$166b2o106bo45bo$165bo154bobo$
166bo120bo32b2o$166bo121b2o$287b2o3$314bo26bo$314b2o24b3o$316bo26bo$
314b2o25b2o$313bobo26bo$249bobo62bo$250b2o60bobo$250bo59bo$310bobo6$
212bo$201bo11b2o204bo$202bo9b2o206b2o$200b3o216b2o$339bo$339bo3$336b2o
$335bobo$174bobo159bo$175b2o$175bo7$289bo31bo$290bo31bo$139bo2bo145b3o
29b3o$139bob2o$141b2o$303bo$301bobo18bobo$143b2o157b2o18b2o$143b2o178b
o4$134bo155bo40b2o$134bobo151b2o40bobo$134b2o129bo23b2o20bobo17bo$266b
o44bo2bo$264b3o43bo3b2o$310bo$284b3o$284bo2bo22bo2b3o$284bo2bo22b2o$
283b4o25b3o$283b2o27b3o$282bo$228bo54bo44b2o$215bo10bobo54bo45b2o104bo
$216b2o9b2o99bo104bobo$215b2o126b2o89b2o$342bobo$106bo237bo$107b2o249b
o$106b2o250b2o$357bobo$190bo$191bo180b3o$189b3o182bo$373bo$387b2o$388b
2o$307b2o78bo$306bobo93b2o$307bo93bobo$403bo$153bo149bo113bo$151bobo
150b2o111b2o$152b2o149b2o111bobo2$431b3o$433bo$432bo4$291bobo$292b2o$
109bobo180bo$109b2o$110bo168bo$280b2o10bo146b2o$279b2o12b2o144b2o$287b
2o$287bo$289b3o10b2o$286bo5bo8bobo$256bo2bo27bo4bo9bo$256bob2o28bo2bo$
258b2o30bo$241bobo41bo$231bo10b2o40bo$229bobo10bo17b2o22b3o$230b2o28b
2o2$122bo$120bobo$121b2o128bo$94bo156bobo$93b3o108bo46b2o$93bob2o108b
2o$92b2o110b2o$92b2o$92b2o$96bo337b2o$95b2o337b2o$93bobo$92bo2bo$93bob
o$93b3o70bobo150bo$93b2obo70b2o111b2o35bobo$93bo2bo70bo111bobo36b2o$
94bo185bo6$97bo$98b2o$97b2o7$429b2o$270bo158b2o$231b3o34bobo$78bo152bo
2bo27b3o4b2o$78bobo149bo3bo27bo2bo$78b2o150b2o25bo8bo$231bobo24bo4b3o
9b2o$232b2o2bo7bobo9b3o15bobo$232bo2bo9b2o28bo$232bo12bo$231b2ob2o$
135bobo94bobo$136b2o95bo$136bo2$220bo$218bobo32bobo$219b2o32b2o$254bo
4$221bo$219b2o203b2o$182bo37b2o202b2o$183bo148bobo$181b3o149b2o$333bo
4$253b2o$252bobo$253bo$113bo$111bobo$112b2o5$205bo$204b3o29bo$204bo2bo
27b3o$206b2o26bo2b2o$206bo27b3o$53bobo147b2o214b2o$53b2o148b4o212b2o$
54bo149bo2bo$205bo28bo25bo$204bo32bo23bo$203b2o29bo3bo20b3o$202bo5bo
26b2ob2o$151bo50b2o3b2o$152bo50b2o$150b3o82b2o$234b3o$232bo2b2o$232b2o
$232bo4$208bobo$209b2o$209bo2$196bo151bo$197b2o23bo126bo$196b2o24bobo
122b3o64b2o$222b2o190b2o4$189bo$188bo$188b3o$126bobo$127b2o$127bo98b2o
$225bobo$226bo$209bo$178bo2b2o26bo$182b2o24bobo$181b2o$210b2o$177b2o2b
2o$183bo26b2o$177b4o$179bo2bo2$274bo134b2o$207b2o66b2o132b2o$207b3o64b
2o$211bo$165bo43b2o$166b2o$165b2o$206bobo$206bobo$207bo3$204b2o$203bob
o$204bo4$362bo$363b2o$362b2o3$189bo214b2o$187bobo214b2o$188b2o2$142bo$
143bo48bo$141b3o46b2o$191b2o5$157bobo39b2o$157b2o39bobo$158bo40bo6$
290bo$288bobo$289b2o$399b2o$181bo14bo202b2o$179bobo14b2o$180b2o13bobo
2$210b3o$212bo$211bo$225b2o$226b2o$225bo$240b2o$239bobo$241bo$255bo$
255b2o121bo$254bobo119bobo$377b2o$269b3o$271bo$270bo$284b2o$285b2o$
284bo109b2o$156bo142b2o93b2o$157b2o5bo133bobo$156b2o5bob2o133bo$136bo
28b2o147bo$135b3o26b3o147b2o$135bob2o174bobo$136bo2bo28b2o$137b2o28b3o
158b3o$137bo192bo$329bo$160bo182b2o$159bo184b2o$159b3o181bo$358b2o$
357bobo$359bo$303bobo67bo$126bo177b2o67b2o$126bobo175bo67bobo$126b2o$
194bobo$195b2o$195bo193b2o$389b2o3$384b2o$384b2o2$156b2o$155bobo$156bo
14$384b2o$384b2o2$149bo$147bobo229b2o$148b2o229b2o2$151b2o$136bo13bobo
$134b2o15bo$135b2o$156bo$156b2o$155bobo161bo$320bo$101bobo66b3o145b3o$
101b2o69bo$102bo68bo38bo$185b2o24bo$186b2o21b3o$185bo$200b2o$199bobo$
201bo177b2o$215bo163b2o$215b2o$214bobo$374b2o$229b3o142b2o$231bo$230bo
$244b2o$245b2o$244bo$259b2o$258bobo$260bo$274bo$274b2o$273bobo2$288b3o
$290bo$289bo$303b2o$304b2o$303bo70b2o$318b2o54b2o$317bobo$319bo$369b2o
$110bo258b2o$110bobo$110b2o4$77bo250b3o$76bo$76b3o$224bo$225b2o$224b2o
7$369b2o$369b2o3$364b2o$364b2o5$324bo$324bo$324bo11$364b2o$364b2o$85bo
bo$85b2o$86bo272b2o$359b2o6$240bo77b3o$238bobo$239b2o10$359b2o$359b2o
3$354b2o$354b2o5$314bo$314bo$314bo11$354b2o$354b2o3$349b2o$349b2o3$
253bobo$254b2o$254bo$308b3o12$349b2o$349b2o3$344b2o$344b2o5$304bo$304b
o$304bo11$344b2o$344b2o3$269bo69b2o$270bo68b2o$268b3o5$298b3o12$339b2o
$339b2o3$334b2o$334b2o5$294bo$294bo$294bo11$334b2o$283bo50b2o$284b2o$
283b2o$329b2o$329b2o6$288b3o12$329b2o$329b2o3$324b2o$324b2o3$284bo$
283bobo$283bo2bo$284b2o12$324b2o$324b2o3$319b2o$319b2o3$279bo$278bobo$
278bo2bo$279b2o12$319b2o$319b2o3$314b2o$314b2o3$274bo$273bobo$273bo2bo
$274b2o12$314b2o$314b2o3$309b2o$309b2o3$269bo$268bobo$268bo2bo$269b2o
12$309b2o$309b2o3$304b2o$304b2o3$264bo$263bobo$263bobo$264bo12$304b2o$
304b2o3$299b2o$299b2o3$259bo$258bobo$258bo2bo$259b2o12$299b2o$132bo
166b2o$133b2o$132b2o$294b2o$294b2o3$254bo$253bobo$253bo2bo$254b2o12$
294b2o$294b2o3$289b2o$289b2o3$249bo$248bobo$248bo2bo$249b2o10$148bo$
146bobo$147b2o140b2o$289b2o3$284b2o$284b2o3$244bo$243bobo$243bo2bo$
244b2o12$284b2o$284b2o3$279b2o$279b2o3$239bo$238bobo$238bo2bo$239b2o5$
232bo$231bobo$161bobo66bo3bo$162b2o67bo2bo$162bo68bo2bo$232bobo2$279b
2o$279b2o3$274b2o$274b2o16$239bo$240bo$238b3o33b2o$274b2o3$269b2o$269b
2o4$8b2o$7bobo204bo$9bo202b2o$23bo189b2o$23b2o$22bobo152bo$178bo$37b3o
136b3o$39bo$38bo$52b2o$53b2o$52bo$67b2o$66bobo200b2o$68bo200b2o$82bo$
82b2o$81bobo180b2o$264b2o$96b3o$98bo$97bo$111b2o$112b2o$111bo$126b2o$
125bobo$127bo$141bo$141b2o$140bobo$253bo$155b3o96b2o$157bo95b2o$156bo$
170b2o$171b2o91b2o$170bo93b2o$185b2o$184bobo$186bo72b2o$200bo58b2o$
190b2o8b2o$189bobo7bobo$191bo$205bo8b3o$205b2o9bo$204bobo8bo$229b2o$
219b3o8b2o$221bo7bo$220bo23b2o18b2o$234b2o7bobo19b2o$235b2o8bo18bo$
234bo24bo19b2o$249b2o8b2o17bobo$248bobo7bobo19bo$250bo43bo$264bo8b3o
18b2o$264b2o9bo17bobo$263bobo8bo$288b2o$278b3o8b2o$280bo7bo$279bo$293b
2o$294b2o$293bo!
..and it turns out I had (through two unrelated but mutually encouraging mistakes) been off by one on my glider lane the whole time. So this won't be useful - the first NE glider stream that meets up with the two reactions cannot actually be synthesized from the existing track (wrong lane mod 3). It turns out all the straightforward 2g reactions to synthesize a SL that the H can burn put the SW glider in the wrong lane for this particular kickback, I need to be more creative.

To remedy this:
The best I can do so far to set up the kickback is 2g -> blinker, g + blinker -> nudged blinker, g + blinker -> loaf (I forgot the intermediate step, which was one of my mistakes).
The blockic NE seed just needs a different pair of 2g -> block reactions to start it off, and will need 5 rephasings between the rakes rather than 4. I should still be able to do the cute double-use of the burning H to produce gliders 2 and 3. I will update with a complete component.

Other notes: Obviously the support streams will need to come from the leftmost 31-2 track. The middle 2-track is just going to stay out of the way here, rephasing as necessary. I think after the MWSS is complete I will look into resetting the clusters closer to the helix.
~~~~~~~~~~~~~~~~~~
Edit: got it! Looking into the 2g+g->block reaction from the LWSS rake prep it turns out the block comes out 9hd over, which saved three rephasings. There are 5 streams that need support from the 31-2 track, but this will work.

Code: Select all

x = 509, y = 953, rule = B3/S23
3$242bo$240bobo$241b2o39$364bo$364bobo$255bobo106b2o$256b2o$256bo31$
343b3o$343bo2bo$343bo2bo2$346b2o$345bo2bo$345bo2bo$345bo2$271bo$272bo$
270b3o2$486bobo$486b2o$487bo2$334bo$332b2o$333b2o13$419bobo$419b2o$
397bobo20bo$397b2o$398bo3$143bo$144bo$142b3o$386bobo$386bo2bo$385bo3b
2o$385bo2$385bo2b3o$357bo27b2o$358b2o27b3o105bo$357b2o28b3o105bobo$
495b2o$285bo$286b2o$285b2o$462bo$461bo$461b3o3$308bo$308bobo90bo$308b
2o90b3o$399bo2b2o$398b5o$398b5o$401bo$401bobo$401b2obo$402bobo$403b3o$
403bobo$401b4o$401b3o$395bo6bo$396b2o$122bo272b2o$123b2o$122b2o3$104bo
$105bo51bo$103b3o52b2o$157b2o229bo$388bobo$366bo21b2o$366bobo$366b2o3$
373bo$371bobo96bobo$372b2o96b2o$471bo$301bo$299bobo$300b2o$286b3o149bo
$286bo2bo146b2o$286bo2bo147b2o$287bo2bo$287b3o$288b2o7$17bo$18bo$16b3o
259bo$276b2o$277b2o$417bo$415bobo$411bo4b2o$409bobo$138bo271b2o$136bob
o$137b2o3$118bo$119b2o52bo$118b2o51bobo$172b2o189bobo$363b2o$341bobo
20bo$341b2o$342bo3$386bobo57bo$387b2o56bo$387bo57b3o$260bo$259b3o52bob
o$258b2o2bo52b2o$258bo3bo52bo$258bob2o150bo$258b2o152bobo$299bo112b2o$
300b2o$299b2o2$259b2o2bo$259bobobo$262bo4$31bo229bobo$32b2o228b2o$31b
2o229bo3$430bobo$431b2o$424bobo4bo$425b2o$151bobo271bo$152b2o$152bo93b
o$245bo$245b3o$134bo$132bobo51bobo$133b2o52b2o150bo$187bo150bo$317bo
20b3o$316bo$316b3o3$402bo$403bo18bo$233bo167b3o16b2o$232b3o186b2o$235b
o94bo$233b2o96bo$234bo94b3o2$386b2obo$384bo4bo21bobo$315bo67bo5bo19bo
4bo$313bobo67bo25bo4bo$314b2o66b2ob2obo24b2o$382b2obo2bo$384b4o21b2o2b
2o$385b3o21b2o2b2o$410bob3o$382bo30bo$380b2o30bo$277bo103b2o$47bo183bo
46bo$45bobo183bo44b3o$46b2o2$446bo$447bo$440bo4b3o$441bo$167bo271b3o$
168bo71bo$166b3o69bobo$239b2o3$147bobo52bo203b2o$148b2o53bo201bobo$
148bo52b3o111bo90bo$313b2o$293bo20b2o$291b2o$292b2o$214bobo$214b2o$
215bo4$344bo$345b2o46bo$344b2o48b2o$393b2o3$328bobo59bo$329b2o58bo$
329bo59b3o2$401b2o$400bobo$401bo$356bo$356bobo$291bo64b2o$60bobo229b2o
$61b2o228b2o$61bo$401b2o$400bobo57bo$402bo58b2o$416bo37bo5b2o$416b2o
37b2o$181bo233bobo36b2o$182b2o69bobo$181b2o71b2o174b3o$254bo177bo$271b
o159bo$163bo106b3o172b2o$164bo51bo52b2o2bo172b2o$162b3o52b2o51b5o170bo
$216b2o52b2o2bo14bo170b2o$271bo2bo14bobo167bobo$271bobo15b2o170bo2$
190bo$189bo$189b3o2$279bo$277b6o$277b2o2b2o$261bo98bo$260bo20bo76bobo$
260b3o14b2o2bo77b2o$277b2o3bo$281bo$279bobo62bo$280bo64bo20bo$343b3o
18b2o$365b2o5$331bobo131b2o$331b2o132b2o$76bo230bo24bo21bo$77bo227bobo
46b2o$75b3o228b2o45bo2bo$354bobo$326b2o25bo2bo$326bobo24bo2bo$328b2o
23bobo$327bo26b3o$325bobo$197bo71bo54b2o$195bobo72bo53b3o$196b2o70b3o
3$177bo$178b2o52bo$177b2o51bobo$231b2o4$166bo293b2o$164b2o294b2o$165b
2o2$349b2o$259bo88bobo$257b2o90bo$237bo20b2o113bobo$235b2o137b2o$236b
2o136bo8$333bo$334b2o$333b2o3$90bo229bobo11bo$91b2o228b2o10bo121b2o$
90b2o229bo9b2ob2o119b2o$330b5o$329b2o2bo$329bob4o9b2o$300b2o27bo4bo8bo
bo$299bo2bo26bo2b2o10bo$298b3o30b2o$210bobo70bo$211b2o71b2o14bobo24bo$
211bo71b2o42bobo$303bo23b2o$302b2o$193bo$191bobo51bobo$192b2o52b2o46bo
$246bo46bo$293b3o3$140bo$140bobo$140b2o$450b2o$450b2o$233bo$135b2o96bo
bo153bo$135bob2o72bo21b2o155bo$137b2o72bobo174b3o$135b2o74b2o109b2o$
133b2o186bobo$322bo$133bobo$134bo9$106bo$104bobo$105b2o2$274b3o34bo
133b2o$273bo2bo35bo132b2o$273bo2bo33b3o$272bo2bo28b2o$226bo45bo2bo28bo
bo$227bo48b2o21bo6b2o$225b3o48b2o19bobo4bobo10b2o$274bo3bo19b2o4bo11bo
bo$277b2o25bo2bo9bo$279bo$206bobo52bo12bo3bo27b2o$207b2o53bo12b3o$207b
o52b3o2$143bo$144bo152bo$142b3o150b2o$296b2o5$208bobo51bobo$208b2o52b
2o139bo36b2o$186bobo20bo53bo140b2o34b2o$186b2o215b2o$110bo76bo$108b2o$
109b2o3$295b2o$294bobo$295bo$175bo$174b3o$174b4o3$119bobo52b4o$120b2o
52b4o$120bo55bo5$435b2o$240bo194b2o$241b2o$240b2o2$92bo193bobo$91b3o
128bo62bo$90b2o2bo128bo51bo9bo4b2o$90bo3bo126b3o52b2o7bo4b4o$90bob2o
181b2o9bo3b4o$90b2o195bobob2o4b3o$157bo30bo101bo8bo$158b2o27b3o81bo26b
o$157b2o27bo2b2o80bobo38b2o$190bo80b2o40b2o$91b2o2bo94b2o120bo$91bobob
o96bo134b2o$94bo94b4o133bobo$182bo6b2o47bo89bo$183b2o52bo104bo$182b2o
53b3o102b2o75bo$93bobo245bobo73bobo$94b2o322b2o$94bo261b3o71b2o$358bo
71b2o$357bo$371b2o$372b2o$177bo193bo$177bobo206b2o$155bo21b2o206bobo$
155bobo229bo$78bo76b2o244bo$77bo323b2o$77b3o320bobo$135bo$136bo278b3o$
134b3o280bo$416bo6$256bo$254bobo$65bo189b2o168b2o$64b3o358b2o$67bo$65b
2o169bo$66bo170b2o52bo128b2o$236b2o51bobo128b2o$290b2o2$173bo$171bobo
72bobo$172b2o72b2o$247bo4$214bo$212b2o$109bo103b2o$63bo46bo$63bo44b3o
4$420b2o$420b2o$152bobo$152b2o$72bo57bobo20bo261b2o$70bobo57b2o283b2o$
71b2o58bo3$149bo$150b2o$149b2o5$46bobo$46b2o$47bo221bobo$270b2o$270bo
2$41b2o$41bobo208bo162b2o$43b2o205bobo51bobo108b2o$42bo208b2o52b2o$40b
obo262bo$39b2o369b2o$39b3o144bobo33bo187b2o$187b2o32bo$187bo33b3o5$
188bo$188bobo$123bo64b2o$124b2o$123b2o3$109bo$108b3o$108bob2o16bo$109b
o2bo14bo$110b2o15b3o280b2o$85bobo22bo299b2o$86b2o$86bo$405b2o$405b2o$
165bo8bo$117b2o44bobo6b2ob2o$116b3o45b2o6b2ob2o$116bo2bo53bob3o$48bo
50bo16b2ob2o51b3o28b3o$49b2o48bobo14b2ob2o81bo2bo$48b2o49b2o14b2ob2o
82b2obo$116bo$117b3o53b2o110bo$118bo54b2o111bo$202b2o80b3o$172bo29b4o$
171bob2ob2o25bo2bo$170b2o4b2o26bobo$170b2ob2o29bobo58bobo52bo$15b2o
155bo93b2o53bo$14bo2bo185bo62bo52b3o$13b3o187b2o200b2o$201bo2bo200b2o$
15bobo181b2o$201bo$18bo381b2o$17b2o381b2o$179bo$180bo$9bo168b3o$8bo$8b
3o$139bo$137bobo50bobo$81b3o54b2o50b2o$81bo2bo106bo$80bo3bo$80b4o$81bo
$158bo$156b2o$101bo55b2o$102bo$81b2ob2o14b3o$81bob3o108b2o204b2o$193bo
bo204b2o$194bo$178bobo$84bo94b2o214b2o$85bo9b2obo80bo215b2o$83b3o7bo4b
o$64bo27bo5bo$62bobo27bo$63b2o26b2ob2obo$91b2obo2bo$93b4o202bo$94b3o
203b2o$299b2o$91bo$89b2o$26bo42bo20b2o189bo$27bo39b2o117b2o94bo51bo$
25b3o40b2o117b2o91b3o52b2o$186bo147b2o$201b2o$200bobo$202bo$216bo178b
2o$216b2o177b2o$215bobo2$230b3o157b2o$232bo157b2o$231bo$245b2o$152bobo
11bo79b2o$153b2o10bo79bo$153bo9b2ob2o92b2o$162b5o92bobo$161b2o2bo95bo$
161bob4o108bo$132b2o27bo4bo108b2o$131bo2bo26bo2b2o108bobo$130b3o30b2o$
115bo173b3o$116b2o14bobo24bo131bo$115b2o42bobo128bo$135bo23b2o143b2o$
134b2o169b2o$194bo109bo$195bo123b2o69b2o$49bo76bo66b3o122bobo69b2o$48b
3o74bo194bo$47b2o2bo73b3o206bo$47bo29bobo254b2o49b2o$47b3o28b2o253bobo
49b2o$47b2obo27bo$48bob2o$48b2o265bo$49bob2o260bobo$48b2ob2o261b2o$48b
2ob2o12bo$50bo14bobo$40bo24b2o228bo48bo$41b2o253b2o46bo$40b2o112b2o
139b2o47bo$153bobo$60b2o92bo$60bob2o$62b2o$60b2o$58b2o$37bo$36bo21bobo
324b2o$36b3o20bo325b2o3$380b2o$380b2o4$143bo$144bo$142b3o$105b3o28b2o$
104bo2bo28bobo$104b2obo23bo6b2o198b3o$108b2o19bobo4bobo10b2o$108b2o20b
2o4bo11bobo$107bo2bo25bo2bo9bo$106bo$106b3o29b2o68bo$107bo101b2o$208b
2o2$93bo286b2o$94bo34bo250b2o$92b3o32b2o$128b2o$375b2o$375b2o3$94bobo$
94b2o$56bo38bo215bo$54bobo252bobo22b3o$55b2o253b2o5$35bo91b2o$33b2o91b
obo$13bo20b2o91bo$11b2o$12b2o2$375b2o$375b2o3$370b2o$79bo290b2o$78b3o
29bo$78bob2o27b3o$77b2o29b2o2bo$77b2o$77b2o251bo$81bo248bo$80b2o26b2o
220bo$78bobo28b3o$77bo2bo31bo$78bobo29b3o$78b3o29b2o112bo$78b2obo140bo
bo$78bo2bo141b2o$79bo28b3o$106bo3bo$106bo$106bobo$370b2o$370b2o$82bo$
83b2o$82b2o281b2o$365b2o2$69bobo25bo$70b2o24bo228bo$70bo25b3o225bobo$
324bo2bo$325b2o3$63bo$63bobo$63b2o3$100b2o$99bobo$100bo$53bo$51b6o26bo
281b2o$51b2o2b2o25b3o280b2o$82b3o$55bo$51b2o2bo26b4o274b2o$51b2o3bo24b
5o274b2o$55bo25b2obo$53bobo$54bo265bo$82b2o235bobo$81bo237bo2bo$320b2o
$80bobo2bo$80bob3obo150bobo$79b3ob2obo151b2o$80bobo155bo$80b3o$81bo4$
78b2o$77bobo$78bo281b2o$360b2o$312b3o$312bo2bo$311bo3bo39b2o$311b4o40b
2o$312bo2$62bo$63bo$61b3o$312b2ob2o$312bob3o2$65bobo$65b2o$66bo248bo$
316bo$314b3o2$33bo$31b2o40b2o$32b2o38bobo$73bo281b2o$355b2o3$350b2o$
300bo49b2o$298b2o$299b2o2$253bo$254bo$69b2o181b3o$70b2o$69bo$84b2o$83b
obo$85bo$99bo$99b2o$98bobo2$113b3o$115bo$114bo235b2o$128b2o220b2o$129b
2o$128bo$143b2o200b2o$142bobo200b2o$144bo$158bo$158b2o$157bobo$41bo$
40bo131b3o$40b3o131bo$173bo155bo$187b2o141b2o$188b2o139b2o$187bo$202b
2o$201bobo$203bo$217bo$217b2o$216bobo$345b2o$231b3o111b2o$233bo40bo$
232bo41bobo$246b2o26b2o64b2o$247b2o91b2o$246bo20bo$261b2o5b2o$260bobo
4b2o$262bo$276bo$276b2o$275bobo2$281bo8b3o$281b2o9bo$280bobo8bo$305b2o
$295b3o8b2o$297bo7bo$296bo23b2o18b2o$310b2o7bobo19b2o$311b2o8bo18bo$
310bo24bo19b2o$325b2o8b2o17bobo$324bobo7bobo19bo$326bo43bo$340bo8b3o
18b2o$340b2o9bo17bobo$339bobo8bo$364b2o18b3o$354b3o8b2o19bo$356bo7bo
20bo$355bo23b2o$369b2o7bobo$370b2o8bo$369bo$384b2o$383bobo$385bo!
Physics: sophistication from simplicity.

User avatar
biggiemac
Posts: 515
Joined: September 17th, 2014, 12:21 am
Location: California, USA

Re: (23,5)c/79 spaceship components

Post by biggiemac » December 5th, 2014, 8:21 pm

By my calculations, if we rephase 9 times, including one rake that deletes the nearer of the p158 blocks and nudges the other, we will filter out the correct half of the above rake to synthesize the MWSS. There will be a blinker left behind, which can be used to produce the resetter.

The rake that messes with the blocks appropriately might need support from the 31-2 track, because it isn't in the correct lane. But since its position within its lane isn't of critical importance, I could possibly just fashion a kickback. This is the only way to deal with the two p158 blocks and filtering the rake correctly at once with just p79 elements, though they can be treated separately with p158 rakes (which I think would be too expensive). I have 9 H to burn on this lower cluster anyway; this would put them to good use!

Edit again, I keep messing up the lane counting. Even better for us, this rake is simply constructable, it's an offset of 12 hd from the previous. So the 4th of the 9 rephasings is a rake and problem solved.
Physics: sophistication from simplicity.

User avatar
Andrew
Moderator
Posts: 919
Joined: June 2nd, 2009, 2:08 am
Location: Melbourne, Australia
Contact:

Re: (23,5)c/79 spaceship components

Post by Andrew » December 5th, 2014, 8:58 pm

dvgrn wrote:If only Golly would actually turn off cells with putcells([X,Y,0]) instead of just leaving them the same color, it would be fairly competitive to build a list of cells that need changing and fix them all at once with putcells().
If you use "copy" mode then putcells will set cell states to 0 -- eg. putcells([X,Y,0],0,0,1,0,0,1,"copy"). (Dave, I'm pretty sure it was you who implemented the mode stuff!)
Use Glu to explore CA rules on non-periodic tilings: DominoLife and HatLife

User avatar
biggiemac
Posts: 515
Joined: September 17th, 2014, 12:21 am
Location: California, USA

Re: (23,5)c/79 spaceship components

Post by biggiemac » December 6th, 2014, 2:47 am

knightship three prep.mc
(193.75 KiB) Downloaded 478 times
Everything on the 316-2-190-263 track for the MWSS synth is taken care of. The rest comes from the 31-2, part two of the the matching challenge. Definitely growing large enough that I want to reset after the synth is completed.
Physics: sophistication from simplicity.

User avatar
biggiemac
Posts: 515
Joined: September 17th, 2014, 12:21 am
Location: California, USA

Re: (23,5)c/79 spaceship components

Post by biggiemac » December 7th, 2014, 6:17 pm

31-2 track back end is easy:

Code: Select all

x = 258, y = 357, rule = B3/S23
$201bo$200bo$200b3o8$257bo$255b2o$235bo20b2o$233b2o$234b2o14$182b3o$
181bo2bo$181b2obo4$181b2o$181b4o$182bo2bo$183bobo$183bobo3$185bo$183bo
bo$184b2o10$169bobo59bo$169b2o60bobo$170bo38bo21b2o$209bobo$209b2o4$
204b2o$204bob2o$206b2o$204b2o$202b2o2$202bobo$203bo11$213b3o$213bo2bo$
199bo12bo3bo$198bobo11b4o$198bobo12bo$199bo4$213b2ob2o$213bob3o5$212b
2ob3o$145bo66b4obo$144bo66b2o2b2o$144b3o3$141bo$140b3o$139b2o2bo$143bo
$142bo51bo$138bo54bobo5bo$138bobo52bobo3b2o$138b2o39bo14bo5b2o$177b2o$
178b2o18$189bo$188bobo$188bobo$189bo5$160bo$149bo9b3o$147bobo8b2o2bo$
148b2o3$158b2o$159b3o$162bo$160b3o$160b2o3$113bobo44bobo12bo$113b2o46b
2o12bobo$114bo46bo13b2o7bo$172bo10bobo$171b3o9bobo$170bo2b2o9bo$170b3o
bo$173b2o2$169bo$168bo$147bo20b3o$146bo$146b3o5$153bo$152bobo$152bobo$
153bo4$179bo$178bobo$178bobo$179bo13$148bo$89bo57bobo$88bo58bobo$88b3o
57bo4$174bo$173bobo$173bobo$174bo$145bo$143b2o$123bo20b2o$121b2o$122b
2o4$117b3o$117bo2bo$117bo2bo$116b4o$116b2o25bo$115bo26bobo$116bo25bobo
$116bo26bo3$70b3o$69bo2bo96bo$69b2obo31bobo61bobo$104bo2bo60bobo$103bo
3b2o60bo$103bo$69b2o$69b4o30bo2b3o18bo$70bo2bo29b2o21b3o$71bobo31b3o
17b2o2bo$71bobo31b3o17bo3bo$125bob2o$125b2o$73bo$71bobo$72b2o2$61b3o
62b2o2bo7bo$61bo2bo30b3o28bobobo6bobo$61bo2bo31bo32bo7bobo$96b3o39bo$
64b2o$63bo2bo62bo$63bo2bo58bo2bobo$63bo54bo11bo33bo$117b3o4b2o2b2o33bo
bo$116bo2b2o42bobo$116b4o44bo$94bobo22bobo$94bobo22bo2bo$119b3o$97bo
20bobo$118bobo$118b3o$52bo60bo$50b2o41bobo18b2o$51b2o41bo18b2o4$133bo$
132bobo$132bobo$133bo$101bobo2bo$102b2o2bobo$89bo12bo3b2o$88b3o68bo$
87b3obo66bobo$90b2o66bobo$90b2o67bo$89b2o$83b3o3bo$84bo$85bo8$100b3o$
98b2o2bo25bo$103bo23bobo$75bo25b2o24bobo$76b2o20b2ob2o25bo$75b2o3$120b
3o31bo$123bo29bobo$120bo2bo29bobo$121b2o31bo$121b2o$120bo2bo$120bo$26b
o95b2o$26bobo$26b2o14$149bo$148bobo$148bobo$149bo4$129bo$130bo$128b3o
6$91bo$89bobo$90b2o2$102bo$100b2o$101b2o2$144bo$143bobo$143bobo$144bo$
bobo$b2o$2bo2$78bo$77b2o$77bobo2$52b3o$52bo$53bo$28b2o$27b2o$29bo$3b2o
$3bobo$3bo!
Physics: sophistication from simplicity.

User avatar
biggiemac
Posts: 515
Joined: September 17th, 2014, 12:21 am
Location: California, USA

Re: (23,5)c/79 spaceship components

Post by biggiemac » December 8th, 2014, 7:32 am

Third synth complete. I had to rework part of the 31-2 track to make the LWSS rake as tight as could be. The cleanup is an interesting challenge. But once again only the 31-2 and 316-2-190-263 tracks remain.
knightship three synth.mc
(219.66 KiB) Downloaded 438 times
Since there are 10 *WSS in the helix, and doing 2 resets would add 2 more, I think doing a reset after the 4th would be the best plan. Resetting after a LWSS synth would be better anyway because the block is left right beside the helix.
Incidentally, if someone could find a collection of NE g + N LWSS -> SW g reactions by lane change I would be very appreciative!
Physics: sophistication from simplicity.

User avatar
dvgrn
Moderator
Posts: 10612
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI
Contact:

Re: (23,5)c/79 spaceship components

Post by dvgrn » December 8th, 2014, 8:13 am

biggiemac wrote:Third synth complete. I had to rework part of the 31-2 track to make the LWSS rake as tight as could be. The cleanup is an interesting challenge. But once again only the 31-2 and 316-2-190-263 tracks remain.
And a line of blinkers, it looks like -- useful for filtering the next forward rake, I think? Or no, I guess it's just junk that's trivial to dispose of.
biggiemac wrote:Since there are 10 *WSS in the helix, and doing 2 resets would add 2 more, I think doing a reset after the 4th would be the best plan. Resetting after a LWSS synth would be better anyway because the block is left right beside the helix.
Hmm, I was still thinking each reset would take two LWSS, same as the current front end. But having gone back and found your edit to this posting, I'm prepared to be wrong. I could definitely use a nice labeled diagram...!
biggiemac wrote:Incidentally, if someone could find a collection of NE g + N LWSS -> SW g reactions by lane change I would be very appreciative!
From my reading of the above linked post, you're interested in k*(NE g) + N LWSS -> SW g, with k as small as possible (2, given the results below) and no sparks that would cause trouble for following LWSSes in a single-density p158 stream. I suppose the ten-lane eastward offset means that sparks won't usually be a problem. Do I have that more or less right? EDIT: Hmm, I had written "clean SW g" above, but now I don't see why that would be a requirement -- or only in the sense that the reaction shouldn't spew gliders in other directions besides SW.

EDIT 2: For example, would these be samples from the collection you're looking for?

Code: Select all

#C 2*(NE G) + N LWSS -> beehive + SW g, return lane 0
x = 100, y = 92, rule = B3/S23
bo$b2o$obo$5bo$5b2o$4bobo2$30b2o$31b2o$30bo$34b2o$35b2o$34bo$60bo$60b
2o$59bobo$64bo$64b2o$63bobo3$77b3o$76bo2bo$79bo$79bo$76bobo29$87b3o$
87bo2bo$87bo$87bo$88bobo29$97b3o$96bo2bo$99bo$99bo$96bobo!

Code: Select all

#C 2*(NE G) + N LWSS -> block+ SW g, return lane 6
x = 96, y = 92, rule = B3/S23
bo$b2o$obo6bo$9b2o$8bobo3$30b2o$31b2o$30bo7b2o$39b2o$38bo2$60bo$60b2o$
59bobo6bo$68b2o$67bobo4$73b3o$72bo2bo$75bo$75bo$72bobo29$83b3o$83bo2bo
$83bo$83bo$84bobo29$93b3o$92bo2bo$95bo$95bo$92bobo!

User avatar
biggiemac
Posts: 515
Joined: September 17th, 2014, 12:21 am
Location: California, USA

Re: (23,5)c/79 spaceship components

Post by biggiemac » December 8th, 2014, 2:24 pm

Those work! I had been hoping that k could be as small as 1 but apparently not. I'll show you what I'm thinking of.

This demonstration is 4 hd off, but should get the point across. I was thinking we transform the blocks left behind in the LWSS synth into something manageable, and then use the other half of the glider stream to interact with the LWSS and produce gliders in the other direction. The lead H would again eat an alternating stream of stills and gliders.

Code: Select all

x = 163, y = 295, rule = B3/S23
105bo$104bobo$104bobo$105bo2$109b2o$108bo2bo$109b2o39$95bo$94bobo$94bo
bo$95bo2$99b2o$98bo2bo$99b2o35$21b2o$22b2o$21bo$25b2o9b2o$26b2o7bobo
47bo$25bo11bo46bobo$40b2o9bo32bobo$39bobo9b2o32bo$41bo8bobo$55bo33b2o$
55b2o8b3o20bo2bo$54bobo10bo21b2o$66bo$69b3o8b2o$71bo9b2o$70bo9bo$84b2o
9b2o$85b2o7bobo$84bo11bo$99b2o$98bobo$100bo2$124b3o$126bo$125bo$128b3o
$130bo$129bo11bo$140b3o$139b2obo$139b3o$140b2o20$133bo$132bobo$132bobo
$76bo56bo$75bobo$75bobo$76bo3$151bo$150b3o$150bob2o$151b3o$151b2o16$
109bo$108bo$108b3o11$161bo$160b3o$159b2obo$159b3o$123bo36b2o$122bobo$
122bobo$66bo56bo$65bobo$65bobo$66bo43$56bo$15bo39bobo$14bobo38bobo$14b
obo39bo$15bo13$18bo$18bobo$18b2o$59bo$59bobo$59b2o15$2b3o$2bo2bo$4b2o
2$o2bo$b2o$4b2o$3b3o$4bo$b2ob2o$b2ob2o$bo3bo$bo3bo$2bo7$7bo$8bo$6b3o!
However, in hindsight I see no reason that the LWSS couldn't just interact with the G to make a still in the right place..

Edit: here - there should be plenty of ways to use a p158 LWSS stream to turn a p158 trail of junk into something useful.

Code: Select all

x = 285, y = 357, rule = B3/S23
230bo$229bobo$229bobo$230bo3$227b2o$227b2o39$220bo$219bobo$219bobo$
220bo3$217b2o$217b2o19$b2o$obo$2bo$16bo$16b2o$15bobo2$30b3o$32bo$31bo$
45b2o$46b2o$45bo$60b2o$59bobo$61bo$75bo$75b2o$74bobo2$89b3o118bo$91bo
117bobo$90bo118bobo$104b2o104bo$105b2o$104bo$119b2o86b2o$118bobo86b2o$
120bo$134bo$134b2o$133bobo2$148b3o$150bo$149bo$163b2o$164b2o$163bo$
178b2o$177bobo$179bo$193bo$193b2o$192bobo2$207b3o$209bo$208bo$212bo$
211b3o$210b2obo$210b3o$211b2o13$200bo$199bobo$199bobo$200bo13$222bo$
221b3o$221bob2o$222b3o$222b2o3$195bo$194bobo$194bobo$195bo23$232bo$
231b3o$230b2obo$230b3o$231b2o29$242bo$241b3o$241bob2o$242b3o$242b2o29$
252bo$251b3o$250b2obo$250b3o$251b2o29$262bo$261b3o$261bob2o$262b3o$
262b2o29$272bo$271b3o$270b2obo$270b3o$271b2o29$282bo$281b3o$281bob2o$
282b3o$282b2o!
From there, we just use more forward rakes to turn a single H into a cluster, like in the creation of the 316-2-190-263 track, and no more LWSS are necessary.

Available 1g + LWSS burnable stills with a measure of the lane change from the NE glider to the SW glider produced when the still burns. Since the direction is reversed, I used a weird way of measuring lane change, it's probably off by 1, but the differences between any two are correct. To the left, glider + still easily created from block -> burnable still with the same lane change info. It's almost comical how none of the lane changes match up.

Code: Select all

x = 217, y = 167, rule = B3/S23
4$204bo$203bobo$203bobo$186b3o15bo$188bo$50bo136bo$49bobo139bo$49bobo
82b3o4b2ob2o44b3o$3o3b2ob2ob2ob2ob2o30bo90b2ob2o43b2obo$6b2ob2ob2ob2ob
2o169b3o$54b2o134b2o$53bo2bo$54b2o3$48b2o$47bobo$49bo15$207b2o$130b2ob
2ob2o48bo20b2o$130b2ob2ob2o48b2o$55bo129bobo22b2o$55bo74b2ob2ob2o71bo
2bo$55bo74b2ob2ob2o52bo18bobo$189b3o18bo$130b2ob2ob2o50b2obo$130b2ob2o
b2o50b3o$189b2o$9b2o$9b2o2$48b2o$47bobo$49bo4$188bo$188b2o$187bobo2$
191bo$190b3o17bo$189b2obo16bobo$189b3o17bobo$190b2o18bo$57bo$56bobo$
56bobo$57bo72b2ob2ob2ob2ob2o$130b2ob2ob2ob2ob2o$52b2o$11b2ob2ob2ob2o
29bo2bo75b2ob2ob2ob2ob2o$11b2ob2ob2ob2o30b2o76b2ob2ob2ob2ob2o3$46b2o$
45bobo$47bo4$208bo$207bobo$207bo2bo$186b2o20b2o$185bobo$187bo2$188bo$
187b3o$186b2obo$131b2ob2o50b3o17b2o$131b2ob2o51b2o17b2o3$57bo$56bobo$
56bobo$57bo2$52b2o150bo$51bo2bo148bobo$12b2ob2ob2o32b2o149bobo$12b2ob
2ob2o2b2o180bo$22b2o163b2o$12b2ob2ob2o23b2o141bobo$12b2ob2ob2o22bobo
143bo$44bo$210b2o$189bo20b2o$188b3o$125b3o3b2ob2ob2ob2o45b2obo$131b2ob
2ob2ob2o45b3o$188b2o12$57bo148bo$56bobo146bobo$56bobo125b2o19bo2bo$57b
o127b2o19b2o$184bo$52b2o135bo$51bo2bo133b3o$12b2ob2ob2ob2o29b2o133b2ob
o$12b2ob2ob2ob2o164b3o$188b2o$12b2ob2ob2ob2o19b2o$12b2ob2ob2ob2o18bobo
$43bo87b2ob2ob2o$131b2ob2ob2o7$213b2o$213b2o3$187b2o$188b2o$187bo3$
188bo15b2o6bo$57bo129b3o13bo2bo4bobo$56bobo127b2obo13bobo6b2o$56bobo
127b3o15bo$57bo129b2o2$52b2o$51bo2bo76b2ob2ob2ob2ob2o$52b2o77b2ob2ob2o
b2ob2o2$8b3o3b2ob2ob2o$14b2ob2ob2o31b2o$52bobo$54bo!
Physics: sophistication from simplicity.

User avatar
biggiemac
Posts: 515
Joined: September 17th, 2014, 12:21 am
Location: California, USA

Re: (23,5)c/79 spaceship components

Post by biggiemac » December 9th, 2014, 6:06 am

4th synth will give it a height of about 10700. Assuming whatever eventually works for a reset takes fewer than 1000 vertical cells, a design with 3 triangles of 4 synths should give a bounding box comparable with the parallel HBK. It will be a fast knightship with a manageable size!
Physics: sophistication from simplicity.

User avatar
dvgrn
Moderator
Posts: 10612
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI
Contact:

Re: (23,5)c/79 spaceship components

Post by dvgrn » December 9th, 2014, 11:22 am

biggiemac wrote:However, in hindsight I see no reason that the LWSS couldn't just interact with the G to make a still in the right place.

Edit: here - there should be plenty of ways to use a p158 LWSS stream to turn a p158 trail of junk into something useful.
...
From there, we just use more forward rakes to turn a single H into a cluster, like in the creation of the 316-2-190-263 track, and no more LWSS are necessary.[/code]
Nice! I won't have the spare brainpower to try to improve on that table for a couple of weeks yet, but I'm glad you can almost laugh at the stubborn lack of good lane-change matches.

That's just the B3/S23 Corollary to Murphy's Law making an appearance. I've found it shows up most often when it would be really nice if a compact Herschel track could connect point A to point B. Whenever I do a search and Hersrch finds half a dozen candidates immediately, I know to be worried: there will turn out to be some subtle thing wrong with every one of them. Either that or, in spite of obsessive double-checking, I'll have miscalculated the distance between A and B.

I suspect this will all sound very familiar... but it seems pretty clear that somewhere in there will be a trick that works, even if it's a little awkward. I'm looking forward to the next LWSS and the reset!

EDIT: You have a "glider + still life easily creatable from block" table -- it looks as if "easily" means block plus two gliders, or three in the case of the blinker (?) One more glider hitting one of the two-beehive stills will produce a block, which is a really good absorber for six adjacent lanes... does that give enough adjustability?

Code: Select all

x = 16, y = 13, rule = B3/S23
14bo$13bobo$13bobo$14bo2$9b2o$8bo2bo$9b2o3$b2o$obo$2bo!

Post Reply