I could use it in my version of Reversi.

Could you make the project "go" (http://en.wikipedia.org/wiki/Go_%28board_game%29) ... that is a much greater challenge! And I could then make a flash version for the website.

this i great game can I contribute with Visual Basic .NET 2005 to prorgam it can anyone help with this bye

I'd be interested

I would like to do something, but upon deeper thought, Reversi is probably not the thing to do it on.  Min-max algorithm would really be almost trivial.

What I am looking for now is a large scale project that is mathematically related.  Something that will emphasize design principles would be best.  Suggestions are welcome. 

An idea is a combinatorics library, however I'm not certain what could go in there.  The reason I had this idea was a function I wrote which was a bijection between the integers modulo n! and the permutations of n objects.  This made it very easy to loop through all possible permutations of an array, making brute force combinatorics programing much simpler.

I'm not anywhere near being an expert, but that could possibly be because Chess AI has been given a lot more attention?

I'm at an amateurish level at both games, and I don't see what would make Go intrinsically harder to program an AI for.

19x19 is a lot of moves and sub-moves (and sub-sub-moves) to evaluate.

ALL HAIL RECURSION!

We did something similar to that in the newgrounds flash forum a while back, only it was a robot wars thing.

we had a set of methods accesible to each robot that it could use to move, to detect opponnents and the obstacles and to fire its weapons etc, and the winner of each round went forward to a final battle between the robots

I want to move away from AI.  The main reason for this is that AI is a complex field which thouseands of people have dedicated their lives toward researching.  We, amatures, have no hope of doing anything "new", unless someone gets really creative.  I don't see too much of a chance of that happening.

However, we can produce utilities that are useful.  So here is an idea.  When you use php in relation with a database to store content for generated web pages, it is very often to want web interfaces for maintaining this data.  This makes it so that anyone not familiar with databases/php can update the content with ease and no risk of dropping a table.  However, writing these html/php/javascript forms out by hand would be quite tedious.  We can create a program, which given a relation schema, will automatically generate said forms which allow users to insert, update, and remove tuples (rows).

Here is the general idea for such an application:

Enter in the domain of the database, username, and password.
Program looks at the database schema, determines all the tables, and lets the user select which tables to generate forms for.
User selects options on the creation of the forms, for example perhaps a field is to be left out (such as id).
When running the program, the html form will be spit out, as well as corresponding php code which will connect to and modify the database.  Also, javascript code will be generated which checks the html form for virtually all possible errors.

For people who don't know anything about server side scripts (PHP) or relations database management systmes (MySQL) I can provide assistance for getting you into it.  I would like to ask that you know at least one higher level programming language (Fortran, C, C++, Java, Python, Perl, etc), otherwise I think this would be too much to handle.

What I really like about this project is that it is (a) useful (especially to me ), (b) Invovles mutliple languages (Java, PHP, MySQL, HTML, possibly CSS, Javascript) and (c) can be subdivided into parts which will be great for team programming.  It is also a large enough project in which design can be emphasized.

But I'm not set on having this be the project.  I want to hear from everyone what they think about it and suggestions for other projects.  And if you really want to do AI, please speak up (because I really don't).

Sure luca.  Got a suggestion? 

A graph theory library would be interesting.  It would include functions to construct, display, and traverse graphs.  It could be made to handle graphs, mutligraphs, directed graphs, weighted graphs, weighted directed graphs, trees, forests, and many more.  It would include algorithms to find minimal spanning trees, Euler paths/circuits, Hamiltonian paths/circuits, bridges, etc.  And additional requirement could be to design a file format which would handle input into the engine.  For example, Prufer's Algorithm could be used for trees.  There is a ton of different variations on graph, and it would require research to find out what exactly we want to include.  It would also be very expandable, and great for design and multiple people working on it.

I think it would be best if it were GUI, which would probably mean Java.  And just a reminder to the rest of you, other opinions would be nice.

When it comes to coding style, I agree.  However Java has a built in graphics library which is platform independent.  C++ can't even begin to compare with that.  And since I'm on a Linux box and you guys are probably on Windows, Java is the way to go.  Java also has a very impressive api, which would be helpful when doing this sort of thing.  And to top it all off, we can make this an applet so that we can put it on a web page.

mikau, you replied on how much you can do, but not on which project you like, or if you don't like either.  If you don't know an element of the project, then it is perhaps motivation for you to learn it through experience.  I'm not concerned with having people knowledgeable on what we use, only that the people involved have some programming experience.