## Search found 966 matches

- July 10th, 2020, 7:53 am
- Forum: Patterns
- Topic: Refactoring DRH's cyclotron with stable components?
- Replies:
**95** - Views:
**37392**

### Re: Refactoring DRH's cyclotron with stable components?

Single-engine cyclotron gun using the 3-block shifter+3 Snark reflector, no rectifiers needed. Good stuff. I made a p846 gun with two copies of the engine and only Snarks and blocks (plus a boat that is never actually cleaned up). This gun is fixed at period 846 and the fact that that is 2 modulo 4...

- July 9th, 2020, 2:04 pm
- Forum: Patterns
- Topic: Refactoring DRH's cyclotron with stable components?
- Replies:
**95** - Views:
**37392**

### Re: Refactoring DRH's cyclotron with stable components?

Is there a way to turn that into a gun? BTW, I am personally less interested in compactification. I find sparse patterns a lot easier to understand. I can't see a way to make a gun. Here is another oscillator with two fewer Snarks: x = 572, y = 503, rule = LifeHistory 503.2A$503.A.A$505.A4.2A$501.4...

- July 9th, 2020, 1:31 pm
- Forum: Patterns
- Topic: Refactoring DRH's cyclotron with stable components?
- Replies:
**95** - Views:
**37392**

### Re: Refactoring DRH's cyclotron with stable components?

Here is a single engine cyclotron made with just Snarks and Bandersnatches. Plenty of room for compactification: x = 571, y = 422, rule = LifeHistory 469.2A$468.A.A$462.2A4.A$460.A2.A2.2A.4A$460.2A.A.A.A.A2.A$463.A.ABAB AB$463.A.AB2AB$464.AB.2B$467.3B$56.2A409.4B6.2A$56.A.A406.3B2AB6.A$ 58.A4.2A400....

- July 3rd, 2020, 4:57 pm
- Forum: Bugs & Errors
- Topic: C++ implementation of CGOL
- Replies:
**10** - Views:
**2602**

### Re: C++ implementation of CGOL

Ok I've changed all the loop conditions and it seems to work now. But there still is a problem when inputting grid sizes with a height larger than the width. Sometimes it makes super tiny grids, sometimes it works normal, and sometimes it crashes. Why? I think I see the reason. You define your arra...

- July 3rd, 2020, 4:15 pm
- Forum: Bugs & Errors
- Topic: C++ implementation of CGOL
- Replies:
**10** - Views:
**2602**

### Re: C++ implementation of CGOL

Also how is: for (int deltax = -1; deltax++ < 2;) any different from for (int deltax = -1; deltax < 2; deltax++) ? I've reverted the 2nd version back and it worked correctly. How? As explained here . The structure of a for loop in C++ is: for (init-statement; condition; iteration_expression) The in...

- July 2nd, 2020, 6:19 pm
- Forum: Bugs & Errors
- Topic: C++ implementation of CGOL
- Replies:
**10** - Views:
**2602**

### Re: C++ implementation of CGOL

But still something's not right, because it went haywire completely! You should investigate the behaviour of these for loops: for (int k = -1; k++ < 2;) { for (int l = -1; l++ < 2;) { if (k||l) neighbors+=current[(i+l+x)%x][(j+k+y)%y]; } } e.g. add something like: std::cout << k << " " << l << std:...

- June 28th, 2020, 8:16 am
- Forum: Patterns
- Topic: Stable signal converters
- Replies:
**257** - Views:
**230202**

### Re: Stable signal converters

Here's a Canada Goose to G with repeat time 504: x = 197, y = 190, rule = LifeHistory 87.2A$86.A.A$80.2A4.A$78.A2.A2.2A.4A$78.2A.A.A.A.A2.A$81.A.ABABAB$81. A.AB2AB$82.AB.2B$85.3B$84.5B6.2A$57.A25.3B2AB6.A$56.A.A23.4B2AB3.BA.A $56.A.A22.10B.B2A$54.3A.2A20.13B$53.A4.B11.2A7.6B3D5B$54.3AB2AB2.2B4. B2AB...

- June 18th, 2020, 4:32 am
- Forum: Patterns
- Topic: 31c/240 caterpillar working notes
- Replies:
**409** - Views:
**306421**

### Re: 31c/240 caterpillar working notes

Another idea: we can treat R2L16F as an R1L16F with a compulsory back rake. There's no need to kill the back rake with an immediate forerake, and if we happen to have a few backrakes left over at the end they can hopefully all be killed by a single forerake and kickbacks. x = 595, y = 807, rule = B3...

- June 17th, 2020, 5:07 pm
- Forum: Patterns
- Topic: 31c/240 caterpillar working notes
- Replies:
**409** - Views:
**306421**

### Re: 31c/240 caterpillar working notes

In the latest update to the script on github the add_R4L8F function looks like this: def add_R4L8F(self): # first backrake self.add_backrake() # second backrake self.add_backrake(2260) # R2L23F: # block puffer self.add_standard_row(10, 3974) # teardrop puffer self.add_standard_row(-7, 5998) The add_...

- June 17th, 2020, 12:30 pm
- Forum: Patterns
- Topic: 31c/240 caterpillar working notes
- Replies:
**409** - Views:
**306421**

### Re: 31c/240 caterpillar working notes

Okay, here's my first attempt to extend chris_c's script. It's wrong, but still very promising. Great, I'm glad you were tempted to try it. I'm also keen to get going but not very good at putting things down once I start... therefore delaying. I debugged your script by changing "while events:" into...

- June 15th, 2020, 10:33 am
- Forum: Patterns
- Topic: 31c/240 caterpillar working notes
- Replies:
**409** - Views:
**306421**

### Re: 31c/240 caterpillar working notes

I'll think about whether it's easier to write a script to do the building for me, this time around... Yeah, let's try and script it. The tricks we are talking about seem very painful to implement using manual copy/paste. Here is my attempt at a script. Fundamentally I see the Silverfish as being a ...

- June 15th, 2020, 6:46 am
- Forum: Patterns
- Topic: 31c/240 caterpillar working notes
- Replies:
**409** - Views:
**306421**

### Re: 31c/240 caterpillar working notes

That gets the rephaser length of the silverfish down to 46 ( 44 if we can shoot down the final block with a backward rake), which obviously is a huge improvement over the 71 rephaser lengths in the old silverfish (it should reduce the length of the ship by 10000 to 12000 cells or so). This seems to...

- June 14th, 2020, 2:26 pm
- Forum: Patterns
- Topic: 31c/240 caterpillar working notes
- Replies:
**409** - Views:
**306421**

### Re: 31c/240 caterpillar working notes

Ah, nice to know people are still working on this. I did try some things about a month ago but because of Over Optimisation Syndrome I quit before getting any concrete improvements to the Silverfish. Here are 3 of my ideas. They range from relatively solid to wishful thinking. 1. An improvement to o...

- May 9th, 2020, 9:01 am
- Forum: Patterns
- Topic: 31c/240 caterpillar working notes
- Replies:
**409** - Views:
**306421**

### Re: 31c/240 caterpillar working notes

What about the head? It wouldn't be surprising if the first three Herschel-pair climbers could be moved back about 400 columns, correspondingly reducing the width of the whole structure by quite a bit. Then again, it wouldn't be surprising if chris_c figured out the math for those first interlockin...

- April 25th, 2020, 6:12 am
- Forum: Patterns
- Topic: Simplified Demonoid in CGOL
- Replies:
**13** - Views:
**4425**

### Re: Simplified Demonoid in CGOL

I came up with a method of initialising a Gosper Glider Gun where the relative timing of the gliders only needs to be correct modulo 30. Hardly revolutionary but it could make P30 self-constructing spaceships a bit cheaper if it saves on OTTs. x = 153, y = 30, rule = B3/S23 99bo$100b2o$99b2o3$o$b2o$...

- April 21st, 2020, 2:39 pm
- Forum: Patterns
- Topic: Simplified Demonoid in CGOL
- Replies:
**13** - Views:
**4425**

### Re: Simplified Demonoid in CGOL

The clearance of an Fx119 based edge shooter is overkill. All we really need is a version of NW31 that is transparent two lanes towards the "inside" (transparency on the output lane is not even a requirement). Hence the following should be possible. In terms of population, the units are just over tw...

- April 21st, 2020, 11:30 am
- Forum: Patterns
- Topic: Simplified Demonoid in CGOL
- Replies:
**13** - Views:
**4425**

### Re: Simplified Demonoid in CGOL

@dvgrn All valid points there. One other thing I realise is that we only need to use Fx119's on one side. The side not needing the clearance can stay as simple NW31's (oh, and get rid of the added Fx77's?). This would probably result in a saving but complicates the construction by bringing circuitry...

- April 21st, 2020, 10:45 am
- Forum: Patterns
- Topic: Simplified Demonoid in CGOL
- Replies:
**13** - Views:
**4425**

### Re: Simplified Demonoid in CGOL

Did you check the 10hd Demonoid? I seem to recall that tookit is even faster (but it has twice as much to build, so it ends up in second place overall.) It's about 700 ticks per glider. I remember the advantage being greater but maybe I added more elbow varieties to the 0hd toolkit than the 10hd. H...

- April 20th, 2020, 4:13 pm
- Forum: Patterns
- Topic: Simplified Demonoid in CGOL
- Replies:
**13** - Views:
**4425**

### Re: Simplified Demonoid in CGOL

I did some rough measurements. The p30 construction arm in Remini seems to realease a glider approximately every 6000 ticks. For single arm construction it appears to be around 3600 ticks. The construction arm in the 0hd-Demonoid releases a glider about every 800 ticks. So the construction method in...

- April 14th, 2020, 7:25 am
- Forum: Patterns
- Topic: Binary slow salvos
- Replies:
**123** - Views:
**19419**

### Re: Binary slow salvos

How is this even gonna be cleaned up? This old post shows that the debris trails from the glider-producing switch engines can be cleaned up with one glider and 11 Corderships. After that it might "only" take another hundred or so gliders to finish the cleanup. Admittedly this part is not very elega...

- April 13th, 2020, 3:25 pm
- Forum: Patterns
- Topic: Binary slow salvos
- Replies:
**123** - Views:
**19419**

### Re: Binary slow salvos

Progress: this one throws a blinker about 45hd away from the centre of the crystal, which should hopefully be sufficient: Looks good. Here is an updated 33 glider RCT that starts off with the sequence A.BA.BA.A that does a 1fd block pull: x = 43980, y = 43798, rule = B3/S23 22006bo$22005bo$22005b3o...

- April 13th, 2020, 1:56 pm
- Forum: Patterns
- Topic: Binary slow salvos
- Replies:
**123** - Views:
**19419**

### Re: Binary slow salvos

By the way, that's not quite a universal set just yet: you also need a mechanism to generate a hand block a sufficient distance from the elbow. I will try. The code for the first part of the search is up here if anyone wants to try deeper searches . It's based on the elbow searching code from years...

- April 13th, 2020, 12:50 pm
- Forum: Patterns
- Topic: Binary slow salvos
- Replies:
**123** - Views:
**19419**

### Re: Binary slow salvos

Fibonacci slow salvos with 3 lanes of spacing and equal parity are universal. There is a push 6 from the right hand side of this pattern and there are plenty of pulls and fires. x = 9814, y = 7867, rule = LifeHistory 4998.3D3.3D3.3D85.3D3.3D3.3D87.D3.3D3.3D72.D11.5D3.D3.D89.D4.3D3.3D3. 3D66.D11.5D3....

- April 13th, 2020, 6:26 am
- Forum: Patterns
- Topic: Binary slow salvos
- Replies:
**123** - Views:
**19419**

### Re: Binary slow salvos

I hacked simsim's code as below. It allows a user definable gap and parity and breaks once it finds 1,000 "options". import golly as g import os A = g.parse("bo$2bo$3o!", -14 - 99, -9 - 100) gap = 3 parity = 0 AB = A + g.evolve(g.transform(A, -128, -128-gap), parity) options = [ (g.parse("2o$2o!"), ...

- April 12th, 2020, 6:51 am
- Forum: Patterns
- Topic: Binary slow salvos
- Replies:
**123** - Views:
**19419**

### Re: Binary slow salvos

I'm not sure what is the limitations you have, but here is an example (looks like it explodes exponentially): [rle] Here is the code (I haven't removed duplicated - just checked for exponential explosion): The code you posted uses: bo$2bo$3o8$5bo$6bo$4b3o! as the AB glider pair. The current RCT mod...