File: crossover.rle

package info (click to toggle)
golly 2.3-1
  • links: PTS
  • area: main
  • in suites: wheezy
  • size: 10,080 kB
  • sloc: cpp: 41,951; python: 6,339; sh: 3,912; perl: 1,172; java: 49; makefile: 47
file content (33 lines) | stat: -rw-r--r-- 2,235 bytes parent folder | download | duplicates (4)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
# A demo of one-way units (diodes) and three-way funnelling units (triodes) in
# the Devore variant of Codd's CA. With two diodes and two triodes arranged as
# shown, a crossover can be made, for signals of any type (4, 5, 6 or 7).
#
# John Devore writes: (part I)
#   "I did the original work as a masters report back in 1973. The text of that
# report may be completely lost, but I have a file of notes I have used in 
# presentations so could reconstruct all the important information again. I was 
# working full time while I worked on my masters degree and I never did follow 
# up appropriately on publishing results. I did all my simulations at that time
# on an IBM 360/50 computer. The component I call a diode was suggested by Rod 
# Bates who was working on his PhD in EE at the time. It takes advantage of the
# fact that the Codd space must include triplicate collision rules for signals 
# colliding at a T junction. A straight-on, a left-hand, and a right-hand 
# collision. Codd had treated these all the same and had to make diodes out of
# extremely elaborate race conditions driving his "gates". By suggesting that 
# either the right or left collision annihilate the colliding signals Rod could 
# create a diode from "wires" arranged in a square with leads at two opposite 
# corners where one lead forms a right-hand T connection with two sides of the 
# square and the other lead forms a left-hand T connection. I discovered that
# the square needed NO sheathing (with no additional rules) -- just a 2x2 array
# of cells with values of 1's -- so the diode was extremely small. Further I 
# found one could add a third lead in the correct configuration to make what I 
# called the "triode" with the terrifically useful property of "funneling" 
# signals from either of two of the leads into the third while being blocked
# from going out the other input lead. Again, this required no rule changes
# beyond the basic diode rule change."
# 
x = 25, y = 16, rule = Devore
18.6B$.8B8.B6AB$B8AB7.BA4BAB$BA5B2AB7.BAB2.BAB$BAB4.BAB7.BAB2.BAB$BAB
4.BAB3.4B2AB2.BAB$BGB4.BAB2.B6AB2.BAB$B.B4.BAB2.BA5B3.BAB$BAB3.B2A4BA
B7.BAB$BAB3.B7AB7.BAB$BAB3.BA4B2AB7.BAB$BA5BAB2.BAB8.BAB$B7AB2.BAB8.B
AB$.7B3.BA10BAB$11.B12AB$12.12B!