|
|
Post by asmeurer on Aug 15, 2008 14:57:39 GMT -5
OK, so I've already suggested a computer generated magicbook puzzle game, and the devs basically said that it would be very hard to do, which I can see. But, how about this: user generated magicbooks. There would be a simple editor where you can drag in the five elements, and choose the number of moves necessary. Of course, you would either have to demonstrate that the puzzle is solvable in the number of moves, or have a computer that could test it, which would basically the chief part of a puzzle generator anyway. You could then post the puzzles on line and have a sharing mechanism, much like Enigmo now has.
The only problem with this is that there would be a finite amount of puzzles possible, so maybe additional elements would need to be added, like the addition of new rows in the middle of the puzzle.
|
|
|
|
Post by aardWolf on Aug 21, 2008 23:12:42 GMT -5
It would be relatively easy to make really complicated magic books if a tool like that were created... Just work backwards. Still, implementation would be very difficult. I would probably make one that filled the whole screen that took one move.  |
|
|
|
Post by thod on Aug 22, 2008 1:38:30 GMT -5
a great idea. I will give it some thought how an algorithm to do that would have to look like.
An early option would be something which uses mainly brute force to generate the puzzles. This would solve the first issue generating them.
Important is that it checks for simpler solutions. As a start it needs to be able to solve all existing on it's own.
Hmmm - intriguing. No worry. If I figure something out I will share.
Thod
|
|
|
|
Post by badjoy on Aug 28, 2008 15:45:03 GMT -5
Actually making puzzles computer generated is not that hard as you think.
As mentioned above you work backwards.
1. you generate random initial rotation.
2. you generate line(s) that vanished this move. (moving existing elements up if needed)
3. you generate move that meses up current order.
4. go to 2 unless you are happy with current number of moves.
How you implement number 3 will factor how complicated the puzzles with outcome.
It does not unnecessarily generate all possible puzzles, but the spectrum should be wide enough to make it worth implementation.
|
|
