Really Hard Puzzle

[referee]

This is not mine. I found this over the net, and it's by Robert Abbott, author of logic mazes (and this is a hard one), as well as games like Eleusis, Ultima, and Confusion.

Anyway, click here to get to the puzzle. The goal is moving one of the two moving pieces, which start at squares 01 and 07, to the square marked "Goal".

I thought the labyrinth would enjoy this.
_________________
Jan 21st, 2008: The pillaging continues.
Mar 4th, 2008: Rest in Peace, Gary Gygax. May your dice always roll a natural 20 wherever you are.

Be the Ultimate Ninja! Play Billy Vs. SNAKEMAN today!

[lostdummy]

I finally managed to solve this, but only after finding out that linked web page contains error in one (crucial) rule description ;p

I tried to solve manually using that web version of puzzle, but after I couldn't solve, I decided to do what I usually do when there is puzzle which is complex more due to number of combinations than due to logical setup: I wrote program to brute-force solve it.

But I got surprising result , since my program could not solve it either. After some time searching for error, I finally found on net another reference to this maze, where supposedly original version was posted:

http://forums.unfiction.com/forums/viewtopic.php?t=19834

And while rules in boxes are not worded identically, they are in fact identical ... except for ONE box.

When I fixed that one difference in my program, it solved maze in 16 moves.

Now, I wont point out which box is different, you can consider that as mini-puzzle of type "find the difference" ;p

[Zahariel]

I'm somewhat embarrassed that I can't find the difference. All the "if"s, "red"s, and "green"s appear to be the same, even.

[Scurra]

I can see the difference and I can also see that it would probably have quite a big impact on the final result. I'm just not sure if the programmed version actually does it anyway, even though it doesn't say it does, because I can't actually get far enough to find out! (I'm finding it really hard...)
_________________
_{still Quiz Olympiad champion. Must get a life.
New definitions: COFFEE - someone who is coughed upon}

[referee]

No, there is no difference.
_________________
Jan 21st, 2008: The pillaging continues.
Mar 4th, 2008: Rest in Peace, Gary Gygax. May your dice always roll a natural 20 wherever you are.

Be the Ultimate Ninja! Play Billy Vs. SNAKEMAN today!

[Coyote]

Well--maybe.

I'm thinking a difference might lie in Box 65. In Abbot's original, the rule change happens when a marker moves out of Box 65. In Luc St. Louis' remake, the rule change happens after a marker moves from a red-print box.

How or if that helps--I have no idea. But I'm really liking this puzzle so far! lostdummy, does your computer solution work on the animated remake (despite the faulty wording)? Right now the remake is a handy way to work on the puzzle, but I don't want to waste time with it if it won't allow an actual solution.

[lostdummy]

Coyote, 65 is not box with different rule that prevent solution. Although now that I reread 65, it could have different meaning in Luc's and original bxes, but Luc's script actually interpret it like described on original text - so it does not prevent solution.

And my solution can not be replayed on Luc's web version, since his script also do not follow that rule. But you can use it to follow almost until end of solution - only last step can not be followed on web.

So if you know where/when web version is incorrect, Luc's web version can still be useful even if it does not allow you to reach solution.

[lostdummy]

BTW, I mailed Luc about missing description on his version, and also about fact that his version is not advancing always other box when box 61 is used . On second issue he pointed that it is correct interpretation of rule 60, and then I noticed how in my interpretation of rule 60 I made same error as what is missing in Luc's description on other box - namely, Luc's box 61 is missing text "and ignore rule in other box", similar to part that i missed in rule 60 ;p

I changed my interpretation of rule 60 in program, and I can confirm few things:
- new shortest solution now takes 37 moves
- nowhere in solution is used box 61 while other box is any with persistent instruction (like 60), so missing description in Luc's box 61 is irrelevant (at least for shortest solution)
- therefore, new solution can be reached with Luc's web version also