Message #2984

Subject: Re: MagicTile Solving
Date: Sun, 16 Mar 2014 13:11:13 -0700

514 -> 539

All but one of the new {4,12} puzzles are solved. The remaining one is very confusing.

I finished solving all {4,4|n} puzzles I know how to do. {4,4|7} E0:1.41:0 was the longest (6372 moves) but thankfully I didn’t get any parity on that one. I did get parity on {4,4|5} E0:1.41:0 though, and it turns out there is an easier way to fix parity than making a big X of turns. Turn one edge, and then find the other edge that is along both a vertical line and a horizontal line from the first edge, and turn that one too. Then you just have to fix the pieces you messed up with 3-cycles.

I also solved two puzzles with very long solutions. Icosahedron F0:2:0 was the longest one with 77239 moves, but I took only 956. I also did two other face-turning Icosahedron puzzles that were very similar. I also did {6,4|3} E0:1:0 which took Eduard 16878 moves, but me 2722. {6,4|3} V2:0:0 is coming soon.


—In, <ed.baumann@…> wrote :

I have also solved now the “little devil” “MT hyp {1 2,4} 4 v1.5:0:0”.
There was a lot of trial and error and I cannot explain how I succeeded. My usual technics didn’t work at all. No construction of a single 3-cycle is possible and hence also no edge-turner. Replying the log file of Michael doesn’t give any insight. There aren’t phases like orient all vertices, bring all edges home, orient the edges. I tried also to label the edges in the “fundamental” view. I didn’t succeed!! Perhaps because there are 24 and not 12 different edges. Each of the 6 2-color edges is 4-fold. So it is terribly difficult to follow where the edges go. I have no idea how Michael reaches 42 twists for this puzzle. Andrey also has solved it (wiki not updated).