Import of XigXag version 1.0 revision 2010.0721.
catseye
11 years ago
0 | <html> | |
1 | <head><title>The Xigxag Automaton</title> | |
0 | <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"> | |
1 | <html xmlns="http://www.w3.org/1999/xhtml" lang="en"> | |
2 | <head> | |
3 | <meta http-equiv="Content-Type" content="text/html; charset=UTF-8" /> | |
4 | <title>The Xigxag Automaton</title> | |
5 | ||
2 | 6 | <style type="text/css"> |
3 | 7 | code { |
4 | 8 | background: pink; |
12 | 16 | padding: 0px; |
13 | 17 | } |
14 | 18 | </style> |
19 | ||
15 | 20 | </head> |
16 | 21 | <body> |
17 | 22 | |
59 | 64 | is the sequence of configurations starting with <var>s</var> and closed under successive |
60 | 65 | applications of the transition relation. |
61 | 66 | Every execution sequences comprises a countably infinite |
62 | number of configurations -- that is, runs never stop. | |
63 | They may, however, be ultimately periodic -- that is, they may reach some | |
67 | number of configurations — that is, runs never stop. | |
68 | They may, however, be ultimately periodic — that is, they may reach some | |
64 | 69 | configuration to which they always return (a "fixed point.") |
65 | 70 | But, as I'll attempt to show later, the number of initial |
66 | 71 | configurations that lead to this is finite (and in fact, pretty small.)</p> |
94 | 99 | |
95 | 100 | <h2>Investigation</h2> |
96 | 101 | |
97 | <p>Sure, it's simple, but I find Xigxag moderately interesting -- interesting enough to | |
102 | <p>Sure, it's simple, but I find Xigxag moderately interesting — interesting enough to | |
98 | 103 | devote this section to proving the following property: <strong>Xigxag has exponential growth |
99 | 104 | almost everywhere</strong>.</p> |
100 | 105 | |
326 | 331 | "programs" (initial configurations) give rise to a countably infinite |
327 | 332 | sequence of different configurations during an execution sequence.</p> |
328 | 333 | |
329 | <p>The fact that Xigxag execution never "halts" is also not helpful -- | |
334 | <p>The fact that Xigxag execution never "halts" is also not helpful — | |
330 | 335 | the same is true for all cellular automata, and this hurdle is usually |
331 | 336 | overcome by attaching a "termination predicate" to the system. That is, |
332 | 337 | if some configuration meets some condition (e.g. contains some substring,) |
337 | 342 | alphabet {<code><</code>, <code>></code>}*, does there exist a Xigxag execution sequence |
338 | 343 | where <var>s</var> occurs as a substring in one of the configurations, |
339 | 344 | but not in the initial configuration? I suspect that if one could show |
340 | that this problem is decidable -- or, stronger, if one could give an | |
341 | algorithm for determining what that initial configuration is -- that | |
345 | that this problem is decidable — or, stronger, if one could give an | |
346 | algorithm for determining what that initial configuration is — that | |
342 | 347 | would imply that Xigxag is not Turing-complete.</p> |
343 | 348 | |
344 | 349 | <p>Another approach one could take is to find a very small Turing |