I was going to mention that I just noticed that the puzzles do not have a symmetrical distribution of pre-filled-in squares; which was, I thought, something that was common to all "SuDoKu" puzzles, and differentiates them from more general "Number Place" puzzles, which have the same requirement of having rows, columns, and subgrids having only one number from 1-9 (I've read it was something added by the Japanese) -- this is true of other sudokus in books I have, and those in the newspaper.
I also just noticed there's more than one solution to a hard puzzle I just finished. I see that this has already been brought up, and I don't mean to complain, just to give feedback.
I think that some kind of more complicated algorithm needs to be invented. I believe that creating a program to generate sudoku puzzles was what made it possible to start putting them in the New York Times, from which they spread all over the place a few years ago. (There is an article in the June 2006 issue of Scientific American titled "The Science Behind Sudoku" which I highly recommend; I'd also suggest www.sciam.com , and there are several other links in the article which I'll put at the bottom of my message.)
So, my primary suggestions are, 1) the grids should be symmetrical, so find out a way to do that, and 2) there should be only one solution per puzzle.
It seems like you're figuring this program out by yourself for the most part; in any case, I applaud and highly appreciate your efforts, on this program as well as the rest on the site. Please keep working on it 
Maybe there's somewhere you can find a better algorithm, to build off of, that's already been written...?
In any case, I do enjoy the program... (though it'd be nice if it followed all the rules of traditional sudokus...)
PS- more links: (from the Sci.Am. article)
1st World Sudoku Championship: www.wsc2006.com/eng/index.php
Math Games. Ed Pegg, Jr: www.maa.org/editorial/mathgames/mathgames_09_05_05.html
The Mathematics of Su Doku. Sorendu Gupta: http://theory.tifr.res.in/~sgupta/sudoku
Mathematics of Sudoku. Tom Davis: www.geometer.org/mathcircles
SadMan Software Sudoku techniques: www.simes.clara.co.uk/programs/sudokutechniques.htm
Sudoku, an overview: www.sudoku.com/howtosolve.htm
Sudoku, from Wikipedia: http://en.wikipedia.org/wiki/Sudoku
A Variety of Sudoku Variants: www.sudoku.com/forums/viewtopic.php?t=995
Cheers!
Visit the Community page
Join our forums
Subscribe to our email newsletter
Subscribe with RSS
Follow us on Facebook
Follow us on LinkedIn
Follow us on Twitter
Follow us on Mastodon
Follow us on BlueSky
Symmetry isn't actually a requirement for sudoku. It happens to look nice -- well, I think it looks cool, anyway -- but it is by no means required. Only Nikoli Sudoku (which are all *hand* constructed) have the symmetry requirement.
As for the multiple solutions, this is now spelled out right on the download page. (See the level explanations.) A later release of the base software includes this notification in the program itself as well. I'll be adapting that for portability.
Keep in mind that this software was not developed here. It was made portable here. I made some minor changes to the source code to disable sound (it had issues on some operating systems), get the game saves to be portable instead of saved locally and better explain the shift-number possible solution feature. Other than that, it's base software developed elsewhere.
I won't be doing any major changes to the code. But, it's open source, so you or anyone else are more than welcome to make any changes and submit patches back.
Sometimes, the impossible can become possible, if you're awesome!
do you know where it was developed, so I can have a look at their stuff...? was it at the fltk site mentioned in an earlier comment?
and again thanks for this site; I had a U3 drive, and I downloaded the uninstaller for that (took too long to load each time I put the jump drive in) and now have most of the portable apps on by themselves with the PStart menu -- I like it much better. I'm looking forward to more portable apps whenever they're ready.
I noticed that some puzzles got to points where there was 4 squares that could have 2 numbers in any order. This is a little programing glitch that i know you can't fix but I wanted to point it out.