Message #904

From: Melinda Green <>
Subject: The first 3^4 solution
Date: Tue, 15 Jun 2010 21:40:57 -0700

Dear Cubists,

Along with his first solution submission, Andrey also claimed to have
written and solved his own implementation about a year earlier than
Don’s first solution around 1989. Unfortunately he never kept a record
of that solution, but then neither had we kept a record of Don’s. Andrey
didn’t even keep a copy of his original code but while discussing the
event he later wrote "I found that it’s more easy to rewrite that old
program than to explain what was it." and included an amazingly short
129 line implementation in C. If his recent record breaking solutions
left any doubt of the veracity of such a claim, running his
implementation leaves no doubt. I thought that it would have to be a
purely symbolic representation, but his visualization gives a good 3D
effect in some very simple character graphics. I can see how someone
might solve this puzzle with such a interface.

I’ve updated the HOF with this new claim and a link to his
implementation. I encourage all the programmers among you to compile and
run it. Here are his instructions for using his command-line interface:

&#42; Program shows four sides of cube, three of them are connected on<br>
  the picture, and the side in top right angle is actually on the<br>
  back (in 3D sense) of left-bottom side. Command is a sequence of<br>
  twists that have structure &#91;O&#124;M&#124;R&#93;(A&#124;B&#124;C&#124;D)(X&#124;Y&#124;Z)&#91;+&#124;-&#124;2&#93; .<br>
&#42; First letter is a selection of layer (none - front, R - rear, M -<br>
  middle, O - whole cube)<br>
&#42; Second - selection of side (A is bottom-left, B - right, C - top,<br>
  D - rear, as they shown on the screen)<br>
&#42; Third - direction of axis (X - to the right, Y - up, Z - back)<br>
&#42; Fourth - direction of rotation (+ or none - clockwise, - - CCW, 2<br>
  - double-turn).<br>
&#42; Also there are commands&#58; P - scramble, S - save, L - load (with<br>
  the fixed file name).

He says that he hasn’t even debugged it because he’s sure that nobody
will seriously try to use it to solve a cube, but I’m not so sure. If
you find any bugs, please let him know.

Well done, Andrey!