And now for something completely different....

@Bri_G I wrote some code that displays the disks next to each other. If I tap on a disk, it rotates clockwise one set of pictures. I found that I can manually solve the puzzle in about 5 minutes. I guess it helps to know in what order the disks should be. Even if I didn’t, then it would take about 30 minutes to go thru all 6 combinations of the 4 disks. So basically the puzzle can be solved manually in less than 30 minutes not knowing the order of the 4 disks.

@dave1707- I still have unfinished versions using the graphics. That’s how I started off the I realised I had better finish the basic solution so I could append the front end. I had a few versions I was going to publish on the forum - but need to finish couple before that.

I think the naked eye can help speed up the solution as you take more in than you think. Also, in this case I think the solution was set up for an quick find as the machine speed when it was designed was relatively very slow.

But the challenge is in the coding. Your configuration is faster than mine - as I said before my layout runs clockwise.

@Bri_G I modified my manual code to keep track of the number of moves before I found the solution. The count was 307 moves.

My original code to find the solution doesn’t use graphics at all. Making one move per draw cycle is too fast to see what’s happening and then to go thru all the combinations would take an awfully long time. That’s why I skipped the graphics and just printed the final answer of how the disks needed to be rotated.

I have to admit one thing, this has kept me busy trying different things.

@Bri_G Can you confirm that the background colours are significant?

@Loopspace - when I first used Basic to resolve this I included colours in my coding. However, when I read the code from the creator of the puzzle I couldn’t see any programming involving the colour. So my latest code in Codea doesn’t use it. My latest code found the solution - so I would say no, you don’t need it. I don’t know if it was a red herring or if the design of the mats didn’t incorporate it in key areas.

I’ll have to try my code without the background colors and see what happens.

I removed the background color and I got the same results as before. The only solution was with the 1,2,3,4 orientation. I think if you’re trying to do it manually, the background color might eliminate trying some solutions.

@Bri_G I did something similar.
p=panda, t=tiger, l=lion, f=frog, r=robot, x=fox, u=pumpkin. I then use p=purple and w=white for the background colors. So I had pp=purple panda or wl=white lion.

@LoopSpace Don’t forget about the background color. Some have a white background, others aren’t.

@dave1707 Oh, I didn’t realise that was significant. Bleugh.

@Loopspace - I adopted a similar pattern but used a single letter to represent an image, found out I had a couple of clashes and then had to bend my ‘rules’ a bit to comply so I ended up with:

P - Panda
T - Tiger
L - Lion
F - frog
B - for roBot
R - for fox (Reynard)
H - for pumpkin (Halloween)

Then represented panda-tiger as PT

Oh by the way I was right and wrong for my calculation - had my binary series hat on should have been 16x48x32x16 ie the number permutations at each position. But doubled 32 to 64 instead of adding 16. Thanks again for the pointer.

When coding jigsaws before, I’ve used a scheme of assigning a letter to each type of edge with the convention that a capital letter matches the corresponding lowercase letter.

Using this convention, I get:

  • a robot-fox
  • b panda-frog
  • c frog-tiger
  • d tiger-panda
  • e fox-lion
  • f robot-lion
  • g panda-robot
  • h pumpkin-panda
  • i panda-fox
  • j fox-tiger
  • k lion-tiger
  • l tiger-robot
  • m frog-fox
  • n robot-frog
  • o lion-frog
  • p pumpkin-fox

With these assignments, I think that the four discs are:

abcdaefg hFiejHkb
mnAcKJli GDcIfmoB
pnEldjMG oLkHNPFg
LhICOpAD JMKOpNLC

All,
I have been asked to give a little more description of the actual puzzle - so I have populated the two arrangement skeletons on the website with images of the discs. Both are equivalent and either orientation works and need only minor changes to the code if you switch orientation. Pick which orientation you prefer and place your discs in the selected orientation and compare the 8 touching sides (4 pairs) to see if they match. Then by switching, rotating discs or flipping the sides find one or more configurations that have all 4 pairs matching. Note - by Matching we are talking mirrored images. If you need further explanation don’t hesitate to post.

p.s. You can’t get too much of a hint from the attached pictures I have used all front images on one and all back images on the other.

p.p.s Printing out the discs and folding/glueing them makes the puzzle much clearer.

@Bri_G I rearranged what’s in my table and I reduced the code from 39 numbered lines to 34. What’s in the table and how it’s arranged is important to the amount of code used to process what’s in it. I find this puzzle a good challenge in how to modify tables to reduce the code needed to solve a problem.

@dave1707 - I made the numbered disc to show that there were 8 faces but, putting two together and rotating to give identical faces opposite each other are mirror images adding a new feature to the testing for matches.

@Bri_G I don’t think assigning a single number to each pair would help. When you compare the two pictures to the two on another disk, you need the mirror image of the original picture. Using a single number would work if you didn’t care about the orientation of the two images. How you create a table will make things either easier or harder. It’s all in how the table is created.

I changed what I have in my table and was able to reduce the size of my code to 39 lines. That was optimizing the code and grouping like code. That’s 39 line numbers. If I count the actual lines of code, there are 49. The table, even though it uses one line number, actually takes 12 lines on the screen.

I’m still trying to find a better way of doing this.

@dave1707 - I spent quite a long time looking into alternative ways of solving the puzzle but usually ran into a brick wall.

@dave1707 - very neat, great demo. Covers most of the work needed to resolve the puzzle. Interesting to see how you set this up - you get to read how other minds think.

A first - never seen break used before in anger.

Thanks for that.

@Bri_G There hasn’t been too much activity on this, so I thought I’d give a demo showing what’s going on. While watching the demo, it gave me an idea on maybe another way to solve it.

Hi All,
Thought I’d post an image from the website header that may help in solving the puzzle. Put two together it may help.