To conclude and close up the loose ends on this project I started a few weeks ago, I'm adding the rest of the content that was involved.


I finalized the viewer program that I used in assembling the puzzle, and it can be downloaded (in all its unoptimized glory) here: The main feature I added was color and keyboard coding to help in hiding the pieces, which is useful in examining the internal structure of the puzzle. Someone asked me about source code and I'd rather hope you read somewhere on backtracking search instead. In case you still want to see what I wrote, I also put in the package an optimized version of the search program and code, without any of that task distribution nonsense. Unfortunately the code is heavily tied to my graphics framework, so just ignore that part of it.

If you're interested in seeing how the original search program worked, I'm archiving it as, though it's visually quite dull. The jackpot work unit was Work106.txt.

I have to admit that even with such a nice solution model like this, putting it all together took more than an hour. Take a fly-by in the viewer and you can see that finding the proper assembling order is not that obvious. Now, I was also asked if I could now solve other (dis)entanglement puzzles too, and the answer is that with a little modification to the code, perhaps yes, but I think I've had more than my share (and shame) of wooden puzzles, so better we just call it quits here, ok?

Last Updated (Tuesday, 04 December 2007 03:27)