About to enter Panama Canal:

Dublin - Samuel Beckett Bridge over the River Liffey

lgmd, I cannot wrap my brain around the time spent on a puzzle with so few pieces. Sounds interesting. What makes it hard?

The photo is blurry as I was blurry-eyed at 4:30 in the morning.

Since pieces in igmd's puzzle are square they can be placed in any of the nine possible locations and in any of four orientations in those locations, there are a total of 95126814720 ( which is 9 ! * 4 ^ 9) possibly ways in which the pieces can be placed, but only one in which the images on the edges of the pieces line up correctly.

A coworker had a similar puzzle, and while two other co-workers were trying to solve it, I wrote a computer program to brute force the solution, which took about 45 minutes to write and debug and about 2 seconds to run.

Bob, thanks for the explanation on why these 9 piece puzzles are so hard, I would never have thought of that. People like you make people like me feel really dumb. LOL

So rob333, now you know why its so hard.

I entered the puzzle into my program and it found a different solution.

I knew i wasn't getting it. I can be dense. And now i appreciate it EVEN MORE ;)

Bob, if you rotate your puzzle solution photo to the left...it is the same solution as lgmd_gaz's. I don't think looking at the puzzle from a different side of the table, so to speak, should count as a different solution. Especially for one without a top/bottom. YMMV. ;)

There are occasions that engineers can so cleverly solve incredibly hard challenges and still miss the easiest ones.

Either way, Bob, I'm most impressed that you would be able to conceptualize an approach and write a program that provides an answer to what for us civilians seems like an unaddressable conundrum. .

Play with this one, it may be easier to see what is going on quicker. This one only took me about 4 hours.

I agree that simply rotating the entire puzzle 90º or 180º doesn't count as a different solution. That makes the total number of puzzle permutations less than the number I calculated above, but I'm not sure how much less. It may simply be that number divided by 4, but that seems too simple of a solution.

In any case, I at first assumed that my program had merely produced a rotated version of the same solution, but the program output showed that the puzzle piece in the bottom center of igmd's image, was placed in the top left in my solution, and there is no rotational transformation which can move a center edge piece to a corner.

I then assumed I had fumbled in encoding the puzzle in my program, so I sliced and diced and pasted the puzzle pieces to correspond to my solution and everything fit.

Here is igmd's solution again:

and here is my solution rotated 90º to the left:

and it is indeed a different solution.

So it is now a find the difference puzzle. Ready, set, go!

Ah, ok. The main elements of the puzzle I checked, were not the parts that differed.

Got it...how interesting!

My husband is very good at puzzles.. He has done several of the 9 piece puzzles in under an hour each. He is not an engineer but a mechanic.