I’m working on a small Chess game and think I’m almost done with the basic model and ready to start writing the GUI.
What I’ve been wondering for a long time is that if the Chess board is represented by a Square array of [8][8], and I create a method to draw a square and somehow establish a relation between that method and the Square model, will I then magically have a visual Chess board?
I mean, say that a square is 10×10 pix, does that mean that the first element (represented by a square) will begin at (0, 0), element two at (10, 0), element/square nine at (0, 10) etc?
When I checked out Graphics2D I noticed that the method to draw a rectangle requires x and y coordinates, what do I do if I want the position to depend on the array element?
Or I am going to have to write down the coordinates of each square on the board?
Another thing that has been bothering me is whether it’s [row][column] or [column][row]?
I’ve been using [column][row] in my code since it seems right if you think of it as (x, y).
When you move up and down you change the row, = y, etc.
I’ve been trying to google around to find out how to apply the MVC-design, what to think about, do’s an dont’s, etc. The only thing I know this far is that model = data logic, view = gui and control = interaction, the conclusion is that either I’m bad at googling or there isn’t any good information targeted at beginners regarding MVC.
Mainly I don’t really understand the relation between M, V and C.
So, let’s get the easy question out of the way: as far as 2D arrays goes, it’s actually up to you to decide whether it’s [row][column] or [column][row]. Java doesn’t care. As long as you treat it consistently throughout your app you can think of it whichever way makes most sense to you. If it helps (and it might not) remember that a 2D array is just an array of arrays.
As for MVC, this is how I think of it: M is the model of the data (the array, for example), V is the visual representation of that data (displayed to the user, as a chessboard perhaps) and C is the controller, which connects the two and transmits changes in one to the other, translating or otherwise interpreting user actions as necessary. If the user drags a chess piece from one square to another, the Controller interprets this gesture, implements the logic that decides if the maneuver is legal and what side-effects it might have (like capturing a piece), and updates both the model and the view as required. The confusion lies in that most UI toolkits, Java’s included, often blur the boundaries between these pieces, so that bits of the Controller end up in the View, or the Model, or both. It’s not necessarily a problem, but it’s something to be aware of. The main thing is to try, as far as possible, to keep a well-defined boundary between the data, the view, and the logic that interfaces the two.
To answer your question about drawing and co-ordinates, it really depends how you implement your chessboard UI. You could implement it entirely using Java Swing, for example, and by making clever use of JLabels, JPanels, and maybe a GridLayout, or perhaps even a JTable, you could almost (almost!) have an automagially updating chessboard. On the other hand, you could implement your own UI classes from scratch, making use of Java2D and the drawing primitives it provides, and then you’d have to do more management of co-ordinates and such.