I went through all the code and made it ISO C90 compliant. It is just the right thing to do. I got sloppy in some places and did variable typing in the code so I didn't have to page up to the top. It is bad, but when trying to solve an issue, it can be distracting. I am still wondering about the utility of Valgrind. It has a problem with ATI drivers, which I can relate to.
I am creating the n-dimensional logic that should be included in this. It mainly just deals with the fact that the visible relationship to position association is just one of 6 possibility dimensions. I have the matrix and I am devising a sensible method to implement it. I will implement GNUbik as a probability array also. That is actually a simpler puzzle, if you get beyond the mind scrambling effects of multiple simultaneous axis and position changes. I see the cube as a combination of 6 sets with 8 elements and a single method of use. The handedness and character of the set elements do not change. I wanted to look at the 6x8 dimensional matrix and see if there are holes, and I am quite certain there are.
The project itself and the game like product is actually very interesting to play. It does the heavy lifting of disposing of possibilities, which I found tedious when playing. I can identify which hints produce what results and apply them out of order to solve the puzzle in fewer moves than the computer. It is all set to measure skill and it is on my list, to add, along with about 30 other features.
I am thinking of a new name for it, as it has gone beyond the original puzzle in some significant ways. I was thinking maybe tanglinks.
I also noted an issue of nested ifs, for, cases, whiles, and else that may have stretched the limits of possibility of the gcc program to resolve. I split the code out to functions and it resolved and I honestly don't know how it was failing and it was so logically complex I don't want to be distracted by the issue.
0 comments:
Post a Comment