Page 35 of 68

Re: Thread for basic questions

Posted: January 20th, 2018, 11:22 am
by dvgrn
Macbi wrote:
bestazy wrote:Hello all!
May I ask here..
What's the real story behind the Conway’s Glider that was drilled into the casing for the MacBook Pro’s Retina display?
Eric S. Raymond proposed it as a symbol for hackers (in the sense of computer enthusiasts rather then system infiltrators) so it's probably a reference to that.
Several levels of speculation on this exact question were posted in various places several years ago (2012). I haven't found anywhere that any of the original decision-makers have come forward with the "real story" since then.

Re: Thread for basic questions

Posted: January 21st, 2018, 3:45 pm
by danny
Are there any stable, totalistic rules with gliders as (or more) common as Life's?

EDIT: Yes, move works. Anything between Move and Life as far as general commonness.

Re: Thread for basic questions

Posted: January 21st, 2018, 11:10 pm
by Majestas32
Other than rules where Life's glider works, I don't think so but Move maybe

Re: Thread for basic questions

Posted: January 24th, 2018, 9:34 pm
by 77topaz
Hmm... a related sequence can be generated by looking at the period of the oscillating pattern resulting from a 1xN line in the rule Wild Seas (B2c3-cekq4ikt5i8/S2-in3-acky4aijry5eiky6i), which supports Margolus mechanisms with four different media:

Code: Select all

N,period
1,0
2,0
3,2
4,2
5,2
6,0
7,6
8,6
9,12
10,4
11,28
12,14
13,14
14,0
15,14
16,14
17,84
18,12
19,372
20,62
21,248
22,8
23,504
24,126
25,252
26,28
27,420
28,30
29,30
30,0
31,30
32,30
33,420
34,28
35,2044
36,1022
37,12264
38,24
39,504
40,126
41,7812
42,124
43,253828
44,4094
45,32752
46,16
47,16368
48,2046
49,85932
50,252
51,18396
52,1022
53,4088
54,56
55,917448
56,32766
57,327660
58,60
59,1860
60,62
61,62
62,0
63,62
64,62
Every N divisible by four will have the same period as A160657(N/4), because a 1x4N line in Wild Seas undergoes the same Margolus oscillation as a 2x4N block in the 2x2 rule. The behaviour of the intermediate values, however, appears less predictable. (EDIT: added more terms)

Re: Thread for basic questions

Posted: January 24th, 2018, 9:47 pm
by BlinkerSpawn
77topaz wrote:Every N divisible by four will have the same period as A160657(N/4), because a 1x4N line in Wild Seas undergoes the same Margolus oscillation as a 2x4N block in the 2x2 rule. The behaviour of the intermediate values, however, appears less predictable.
Actually, all even N are equivalent to their 2x2 blocks of the same length:

Code: Select all

x = 42, y = 2, rule = B2c3-cekq4ikt5i8/S2-in3-acky4aijry5eiky6i
18o6b18o$24b18o!
Odd N don't have a particular known pattern because they typically fall into non-2x2 configurations, such as this one, or shrink into shorter patterns:

Code: Select all

x = 7, y = 11, rule = B2c3-cekq4ikt5i8/S2-in3-acky4aijry5eiky6i
3bo2$3bo2$ob3obo2$ob3obo2$3bo2$3bo!

Re: Thread for basic questions

Posted: January 24th, 2018, 9:53 pm
by 77topaz
BlinkerSpawn wrote:
77topaz wrote:Every N divisible by four will have the same period as A160657(N/4), because a 1x4N line in Wild Seas undergoes the same Margolus oscillation as a 2x4N block in the 2x2 rule. The behaviour of the intermediate values, however, appears less predictable.
Actually, all even N are equivalent to their 2x2 blocks of the same length:

Code: Select all

RLE
Yes, that's true, and also for one live cell and three dead ones in a 2x2 block, or two live cells diagonally adjacent to each other. However, the A160657 sequence I was referring to only cover 2x4N blocks, not all 2x2N. (EDIT: The short sequence in the quote at the top of dvgrn's post that I linked previously does contain all the even-numbered terms of the Wild Seas sequence, except the ones equal to zero.)
BlinkerSpawn wrote:Odd N don't have a particular known pattern because they typically fall into non-2x2 configurations, such as this one, or shrink into shorter patterns:

Code: Select all

RLE
I wonder if there does exist some mathematical way to predict those terms also, though. It seems to be related to what dvgrn was discussing in the post I quoted.

EDIT 2: Something else which is interesting is, while a 1x2N line and a 2x2N rectangle while always have the same period for the same N, the behaviour of odd lengths is different: 1x(odd) lines, despite not quite fitting the Margolus mechanisms, always end up forming a single large oscillator, while 2x(odd) rectangles typically break up into smaller patterns.

Re: Thread for basic questions

Posted: January 25th, 2018, 8:34 pm
by AforAmpere
How can you find partials with WLS? Can oscillator partials be found as well?

Re: Thread for basic questions

Posted: January 26th, 2018, 12:21 pm
by googleplex
I found an alternate way to make a block laying switch engine from a switch engine and a beehive:

Code: Select all

#C Generated by copy.sh/life
x = 30, y = 8, rule = B3/S23
b2o$o2bo$b2o$$25bobo$24bo$25bo2bo$27b3o!

Re: Thread for basic questions

Posted: January 26th, 2018, 2:03 pm
by googleplex
If I have a still life, how do I find its ID?

Re: Thread for basic questions

Posted: January 26th, 2018, 2:04 pm
by Majestas32
There really *should* be a script for finding the apgcode of a still life by now but I can't actually find one

Re: Thread for basic questions

Posted: January 26th, 2018, 2:46 pm
by dvgrn
Majestas32 wrote:There really *should* be a script for finding the apgcode of a still life by now but I can't actually find one
There's one by biggiemac on the script request thread.

Re: Thread for basic questions

Posted: January 26th, 2018, 4:49 pm
by Apple Bottom
dvgrn wrote:There's one by biggiemac on the script request thread.
And a straightforward modification that also handles extended ("greedy") apgcodes for larger patterns.

Re: Thread for basic questions

Posted: January 28th, 2018, 10:08 am
by AforAmpere
Can it be proved that a 3-cell C/2 diagonal is impossible without B0? If so, I need to change the color in the Wiki to the optimal blue. I am almost 100% sure there can't be one smaller than 4 cells, but I don't know how to verify.

Re: Thread for basic questions

Posted: January 28th, 2018, 11:03 am
by Macbi
AforAmpere wrote:Can it be proved that a 3-cell C/2 diagonal is impossible without B0? If so, I need to change the color in the Wiki to the optimal blue. I am almost 100% sure there can't be one smaller than 4 cells, but I don't know how to verify.
Yeah. LLS proves that there are no solutions where the three cells fit in an 11 by 11 bounding box. So one of the cells has to be at least distance 4 away from the other two. So it can't even interact with them in the two generations of the period. So there's no way it can move itself in one particular direction.

What do the colours on the wiki mean?

Re: Thread for basic questions

Posted: January 28th, 2018, 11:07 am
by AforAmpere
Blue is 3-cells, or optimal in a lot of cases, green is 4 cells, yellow is above 4, and orange is above 25. I think blue was intended to mean optimal originally by muzik, so I'd like to change all provably optimal ships to blue.

Re: Thread for basic questions

Posted: January 28th, 2018, 11:13 am
by Macbi
It would be good if we could sharpen the bounding-box-logic I used above. I used a really big bounding box because it was easy to check with LLS (anything in period 2 is easy) and meant that the proof was really obvious. But with larger numbers of cells and periods the logic becomes more complicated. It would be good to have some definitive function f, such that if there is no (x,y)c/p ship with at most n cells in a bounding box of side f(x,y,p,n) then there is no (x,y)c/p ship with n cells at all.

EDIT: I should mention that actually doing this with LLS is complicated even more by the fact that the command "-b x y" searches for patterns where the entire evolution is in a x by y box. So you have to think to yourself "if the starting pattern were in a x' by y' box, then what x by y box would the entire evolution fit in?".

Re: Thread for basic questions

Posted: January 28th, 2018, 12:03 pm
by Majestas32
For c/2 diagonal, it seems intuitive that a 4x5 bounding box is all we need to check for the other cell to be not distance 3+ from the other cluster (exploiting the fact that it's a spaceship and not some other kind of pattern to be run in 2 generations). Ditto applies to (2,1)c/2 in B0.

Re: Thread for basic questions

Posted: February 3rd, 2018, 4:01 pm
by Naszvadi
I have no clue where to announce retrocomputing AND CGoL related stuff. Also interested in restoration projects of vintage CA-related IT stuff.

Well, I've just finished a Xenix286 autoinstaller script (using only qemu-0.14 XOR recent qemu, sleep, bash and netcat), needed only 8 minutes to a full install with Xenix Games in a virtual AT 615x4x17 (20 Megabytes) hard disk. The latter contains a game of life implementation, here is a screenshot:

Code: Select all

 ------------------------------------------------------------
|                                    *         **            |    L  I  F  E
|                                              **            | Generation   772
|                                                            | Population    55
|                                                            | Births        10
|                                                            | Deaths        10
|                    **                                      | Frequency      1
|   ***             *  *                                     |
|                    * *                                     | Menus coming soon
|                     *                                      |
|                                                            | New Features:
|               **                                           | S - file save
|               **                                           | R - file restore
|                                    *                     **|
|*                    *              *                     * |
|*         *         * *             *                      *|
|         * *        * *                                     |
|         * *   **    *          ***   ***                   |
|          *   *  *                                          |
|               * *                  *                       |
|                *                   *                       |
 ------------------------------------------------------------
command ?


Re: Thread for basic questions

Posted: February 3rd, 2018, 4:09 pm
by Macbi
That's super neat. Did menus ever arrive? :lol:

What does "frequency" do?

Re: Thread for basic questions

Posted: February 3rd, 2018, 4:18 pm
by danny
Sorry to ask a question so soon after the last, but does the Sprotsmanship badge (Contribute one third of a trillion objects to a different rule or symmetry) apply to just one rule or all rules searched?

Re: Thread for basic questions

Posted: February 6th, 2018, 3:40 pm
by Naszvadi
Macbi wrote:That's super neat. Did menus ever arrive? :lol:

What does "frequency" do?
answers: nope, don't know.

Well, I've successfully "ported" from GNU sed to AT&T System V UNIX' sed - a rule-110 CA simulator. Yes, in pure sed. A screenshot is:

Code: Select all

# (echo 1;while echo;do :;done|head)| ./ca3.sed
11
111
1101
11111
110001
1110011
11010111
111111101
1100000111
11100001101
110100011111
# uname -a
attunix attunix 4.0 2.1 i386 386/AT
# ./ca3.sed
00010111

00111101

01100111
quit
11101101

# 
Is anybody interested in it? With this the Turing-completeness of sed become clear - if arbitrary resources are available and line length has no upper bound.

Re: Thread for basic questions

Posted: February 11th, 2018, 9:46 am
by gameoflifemaniac
Is Calcyman still working on the 0E0P metacell?

Re: Thread for basic questions

Posted: February 11th, 2018, 4:14 pm
by 77topaz
gameoflifemaniac wrote:Is Calcyman still working on the 0E0P metacell?
You might have more luck if you just messaged him directly, rather than only making a post in a thread like that.

Re: Thread for basic questions

Posted: February 12th, 2018, 7:57 pm
by AforAmpere
Is there a cheap way to synthesize this?

Code: Select all

x = 11, y = 9, rule = B3/S23
3$3b3o$3bo2bo$7bo$4bobo!
On that note, where do a lot of three-glider syntheses come from that are used in things like the Synthesizing Oscillators thread? Is there a database?

Re: Thread for basic questions

Posted: February 12th, 2018, 8:01 pm
by 77topaz
AforAmpere wrote:Is there a cheap way to synthesize this?

Code: Select all

RLE
On that note, where do a lot of three-glider syntheses come from that are used in things like the Synthesizing Oscillators thread? Is there a database?
There does exists such a database (which can be found in the "enumerating three-glider collisions" thread).