Right now I'm trying to get my (new and improved) program to implement supermoves, except I'm cheating. To determine if a huge jump is possible, the program uses this algorithm:
Let F=number of available freecells
Let C=number of available empty columns
Longest card run moveable = (C*C+C)/2 + FC + F + 1
Although this is probably the way it's supposed to be calculated, right?
If the move is cool, the program just picks up the whole lot of cards and moves them over.
What about super-super-moves that require the use of non-empty columns? Are those just treated as two or more separate moves? Right now I project that the biggest problem will be moving runs to empty columns. As for notation, my program's notation is going to look like a nightmare at best. (grin) Until I tell it how to translate. (And figure out for myself how it works!)
Danny, you sound like the most likely to be able to answer this, but: how do you create the Microsoft deals using C++? I found the source somewhere, but I'm pretty sure it was solely for C, and I'm not familiar at all how seeding works. Could you point out where that file is, and possibly what kind of results to expect? This way I could use it in conjunction with my program to see how fast it solves the Microsoft 31999.
How about that Kasparov v. X3D Fritz matchup? The two of 'em so far? That's what inspired me to start writing this program!
Received on Thu Nov 13 2003 - 22:47:16 IST