Re: "by modifying the attachments list in your first post." Only attachment was a zip file. Can that file be replaced by a new attachment?
My notion of levels:
Program contains 10 "seed grids" of known (provable) solutions. Each new play starts with one of these ten picked at random. The pick logic is such that a given grid will not be picked twice in succession, thus making it harder to remember all ten patterns. Beyond that, the grid is further "dinked" before a game begins.
For kindergartners, the "play pattern" is easy to spot, For example:
0 3 0
3 3 3
0 3 0
For grade school kids, the pattern is not as obvious because the grid contains no threes. Adjacent twos, for example, may or may not be part of a given "cross."
0 2 2
2 2 2
2 2 0
Junior high patterns are similar to the above, but there are no ones, and which "x" to play first is harder to spot.
High school grids are further "dinked" before they are displayed. In this case, one of the grid positions is pre-decremented thereby obscuring any visible pattern.
For college seniors, a whole series of positions are arbitrarily decremented before their game begins. This one is the meanest of all. In fact, I can rarely beat it, even though I know the dinking logic.
Now, having said all of that, I did just find that the dinker for the first three levels is not doing its thing. I think when you see the fixed version you will perhaps agree that the grids appear to be randomly organized. Meanwhile, run the college level grids and see what I mean by "mean."
Thanks again for finding my bug. Someday I'll write a program that's bug free. Hopefully.
