Shattered and Shuffled

Discussion of challenges you have already solved
Post Reply
Tron
Posts: 30
Joined: Fri Oct 22, 2010 6:59 pm

Shattered and Shuffled

Post by Tron »

This is a really nice challenge. It's quite clear what to do, but it's not trivial to implement. I wish there were more challenges of this quality!
User avatar
bsguedes
Posts: 103
Joined: Tue Feb 24, 2009 12:39 am
Location: Porto Alegre

Post by bsguedes »

This was only a warmup, the next round comes soon ;)
Tron
Posts: 30
Joined: Fri Oct 22, 2010 6:59 pm

Post by Tron »

This sounds promising!

Do the filenames of the images mean anything? I didn't look into them when the server accepted the solution shown on the image.

Now if I only could get the hang of Real Simple Access. I have the decoded number, but I can't make heads or tails of it. I'm afraid that the next round of Shattered and Shuffled might show up behind Real Simple Access.
User avatar
bsguedes
Posts: 103
Joined: Tue Feb 24, 2009 12:39 am
Location: Porto Alegre

Post by bsguedes »

In this one they mean nothing, but for the next one I think they may carry some information about the pieces and the shapes because complexity will be slighty harder.

I think adum will put the sequel to this challenge after Shattered and Shuffled instead of Real Simple Access.
ctzsm
Posts: 3
Joined: Fri Jan 27, 2012 5:59 am

Find better solution

Post by ctzsm »

I did it with my more than 700 lines Java program. But the program is not good enough to solve it directly, and I must correct some wrong answer manual. So I did this boring thing more than 3 hours= =, could anyone tell me how to do it easier?
User avatar
laz0r
Posts: 290
Joined: Thu Feb 04, 2010 4:18 pm
Location: Within the depths of Unix

Post by laz0r »

Mine (unoptimised Mathematica) took over 250 hours of CPU time, but I only had to correct five or so mistakes. A little embarrassing…
There is no spoon.
ftfish
Posts: 12
Joined: Thu Aug 20, 2009 1:51 am

Post by ftfish »

solved it with straight forward algorithm with 300 LoC in java.
I've implemented backtracking functionality, but my first choice was always correct so that no backtracking was done :)
running time with GUI was several seconds.
User avatar
bsguedes
Posts: 103
Joined: Tue Feb 24, 2009 12:39 am
Location: Porto Alegre

Post by bsguedes »

250 hours? Well, that was unexpected!
User avatar
laz0r
Posts: 290
Joined: Thu Feb 04, 2010 4:18 pm
Location: Within the depths of Unix

Post by laz0r »

bsguedes wrote:250 hours? Well, that was unexpected!
It only took about 150 to get the required text, but it didn't occur to me to try it as the answer - the hex file names threw me off!
There is no spoon.
Tabun
Posts: 17
Joined: Wed Feb 05, 2014 12:21 pm

Post by Tabun »

This was pretty fun!
I had most difficulty getting the edges in place before filling in the rest. By the time I'd tweaked the checks for shape and color enough to get the edges right, the program could fill in the rest on its own. I had to correct a wrong pick twice, after that it was smooth sailing.

250 hours sounds bizarre! My not-so-very-efficient php console app takes about 8 seconds to load in the pieces and their shape/color information in memory, then it takes about 20 seconds per row to match the pieces.
User avatar
Hippo
Posts: 339
Joined: Sat Feb 01, 2014 12:05 am
Location: Praha 5

Post by Hippo »

Thanks for forcing me to learn how o make interactive graphics in java.
On previous challenges I have used java to transform some picture to another or grab an info from it. Now I have implemented interactive puzzle solver ... and my smaller doughter was ready to help me in chosing best of 2-6 pieces... I was afraid my initial renaming of the pictures would be problem but I have tried to answer what I read on the picture first.

It would not be that difficult to transform my saves to original names anyways.

Nice challenge.

[edit]
I have updated my code to reflect "border compatibilities" and now the solution is readable without interaction, and solving whole puzzle needs about 50 clicks, but seems it is still far away from helping in the next level.
Seems to follow the lines would help in the next level ...
[/edit]
User avatar
yes-man
Posts: 32
Joined: Fri Jan 30, 2009 5:14 pm

Excited for the next one!

Post by yes-man »

This challenge was really fun, so I'm happy to see there is a follow-up.

It took me about 10h to code this in about 600 lines of C# but seeing the pieces being puzzled together in realtime (about 3 minutes) is so rewarding. I've watched it a few times :D

I'm also quite sure that I'm well prepared for the next one because only little changes to some constants in my code are needed to adjust for the new pieces etc .. although I suspect there is more to it this time.

Cya in the 'solved' forum soon ;-]
User avatar
bsguedes
Posts: 103
Joined: Tue Feb 24, 2009 12:39 am
Location: Porto Alegre

Post by bsguedes »

Can you post an animation on gfycat or youtube or something???
I'm the author of both challenges, I'm interested in seeing this animation you've done!

Cheers and congrats on solving the first part!
eulerscheZahl
Posts: 58
Joined: Thu Nov 29, 2012 7:45 pm
Location: Germany

Post by eulerscheZahl »

This was generated with my solver for the 2nd puzzle.
Without the image export for intermediate steps it takes about 10s to run.
I start with a random piece, then always take the longest edge and try to place pieces along the edge. I repeat the filling process in both directions and make sure that I place the same pieces in both cases. If it fails, I split the edge in half and retry with the shortened edge.

gif
Post Reply