problem solving

sorry but my question is how does one tackle any given problem

such a broad question....its almost hard to answer. Can you elaborate ?

I suppose I would start with trying to find the root of the problem.

well lets say I am programming a game such as tic tac toe , how would I begin and how would I progress until I finish.are there any good web sites for general game design and problem solving.

1. get your idea on paper... this case tic tac toe
2. separate the elements of the game (marks (X,O), game rules, board, AI, user input, etc)
3. take a big cup of coffee, or red bull.. always helps.
4. sort your elements from what has the broadest effect on the application, and the ones that have the least effects on the application. (Board, marks, user input, game rules, AI)
5. Program in the order from the the most broad to the least
6. Each time you apply a element, test it until you feel confident that is working the way it should
7. Play your game for a while.
8. Set yourself a new project, and start at number 1 again.

just my 2 cents.

sounds good thanks for the input.

Phil, I love you brother, and I hope you keep up this fight to learn programming. But, I have to ask, have you actually gone through a C++ tutorial? Did you understand all of it? If not, you really need to. Let me suggest this one:
http://cplusplus.com/doc/tutorial/

I know you're in a hurry to write a cool game, and to play it. But, nothing truly worth the effort comes easy. When I was young (13/14), I learned to program in BASIC. I wrote crappy code. There was no internet, only BASIC reference manuals. But, i started by writing text adventure games, that had thins interface:

You are in a dark room. You see an exit north and South.
What would you like to do:
1) Go North
2) Go South


And, I had a blast writing those programs. Slowly, I advanced, learned about variables, how they worked, I wrote increasingly more complex programs. But, this was over a span of a few years. You need to show patience, start with a step, then another step. You're trying to run a marathon before you can sprint.

You can have fun with simple programs while you learn too. As you go through the tutorials, try to use what you learn and make a simple game from it. Guess the number, roulette wheel gambling, hangman, etc.

GL.

thanks beer I really appreciate your input.I was also thinking about brushing up on my c++.yes I need to walk before I can run.Is there any specific part of c++ I should learn to do games?

It's not really the parts of C++ that you need to grasp, it's more of understanding the philosophy of programming. Going though the C++ tutorials will help you learn to program, but it will also teach you the C++ syntax, which you can then apply to other API's (SDL, or SFML).


Once you understand the basics of programming structures, you'll be able to apply all of it to making games.

ok sounds good thanks beer

Problem solving is a complex field, and there are a lot of studies and books on how to approach it.

How a person attempts to solve a problem will differ from person to person based on how they think, organize, and even visualize a problem.

That being said, there are a few common things you can do to help yourself solve problems:

1. Write down the problem, be as descriptive as possible without going overboard. This is to get the problem solidly described in your head and on paper. It helps you to build a proper visualization, rather than some vague description.
   
2. Break the problem up into smaller pieces, these pieces should be relatively self contained and you should be able to describe how the different pieces talk to the other pieces to produce the whole problem.
   
3. Repeat for each of the smaller pieces till you have something more manageable.
   

Solving the problem is then just an effort in solving each of the much simpler and smaller problems. Further, since you have an idea how all these pieces fit together you can write solutions that take that into account. That being said, many people make things far more complicated than they need to be. A good example is hierarchical scene culling, its relatively complicated and in the end it turns out that a brute forced array is usually faster.

well as beer recommended I am studying c++.