New to game programming

[Autechere]

Hey everyone, im a second year computer science major and im interested in beginning game development. My university offers a concentration in game design to computer science majors and I would like to try and make a game before I decide if its for me or not.

I am pretty fluent in java at this point, so I might as well use java to start out with. What I lack is the knowledge of API's for game development. What I would like to get from you all here is: where do I start? I'd just like some broad advice as to where to concentrate my efforts as a novice game programmer.

[kappa]

Just decide if you'd like to go 2D or 3D then pick one from here.

[ReBirth]

Decide too what library you will use

Games published by our own members! Check 'em out!

[Maximus]

My first post .
For simple 2D games, graphically AWT and Swing is all you're gonna need. Swing to have the main window(jpanel) to draw in, AWT for manipulating transforms and drawing sprites.
http://download.oracle.com/javase/6/docs/api/

[ra4king]

There is no point in drawing in JPanel. JComponent is suggested for very simple games, Canvas is otherwise recommended.

[ReBirth]

I prefer canvas if it still on java2d. Double buffered and I feel it (in my old PC) pretty fast.

[Z-Man]

There is no point in drawing in JPanel. JComponent is suggested for very simple games, Canvas is otherwise recommended.

I picked up Killer Game Programming in Java a day ago and have been reading it. KGPJ draws on a JPanel by drawing to an image then drawing the image to the JPanel (Double buffering I think?) from what I have read so far the book seems to be a little old. So is there a better way of drawing for a 2D game?

[ra4king]

Drawing on an image, then on a JPanel?!? Now that is not efficient at all since a JPanel is already double-buffered. However, to get you used to this habit, start using a Canvas. If this is going to be a desktop app, then there is no need to extend anything. All you have to do is create a Canvas object, add it to the JFrame, and to double buffer, do the following:

1   2   3   4   5   6   7   8   9   10   11   12   13   14   15   16   17   18   19   20   21

//instance variables Canvas canvas; BufferStrategy strategy; //init canvas = (Canvas)jframe.add(new Canvas()); canvas.createBufferStrategy(2); strategy = canvas.getBufferStrategy(); //inside game loop do{     do{         Graphics2D g = (Graphics2D)strategy.getDrawGraphics();                 //drawing code                 g.dispose();     }while(strategy.contentsRestored());         strategy.show(); }while(strategy.contentsLost());

[ReBirth]

It's my first time I see this

1	canvas = (Canvas)jframe.add(new Canvas());

any different from

1   2

canvas = new Canvas(); jframe.add(canvas);

??

[ra4king]

Both are the same. The add() method returns the same object you added.

EDIT: Fixed my code a little: strategy.contentsRestored() is supposed to be inner do while loop

Games published by our own members! Check 'em out!

[ReBirth]

Both are the same. The add() method returns the same object you added.

I just know that thanks

[ra4king]

Glad to help

[Z-Man]

EDIT: I just realized I asked the first question about KGPJ    I guess I didn't understand the answers. I also feel stupid for not realizing that until now

I'm not sure if I should create a new thread for this... but it kind of relates to a question already asked here. I too picked up Killer Game Programming in Java (KGPJ) and used it's "double buffering" strategy. Now I am trying to convert it's double buffering strategy which is: "Have a global Graphics object and a global Image. Have a method called renderGame() that gets a Graphics object from the global image and does all the game drawing on the that Graphics object. Then there is a method called paintGame() which draws the Graphics object from the Image onto the Graphics object of a JPanel (I use a JComponent). So drawing on an image then drawing the image to the screen." My problem is I am having trouble converting this to the suggested Canvas method. Does anyone have tips? One of the problems that comes to my mind is that I handle input through the JComponent with a MouseAdapter and a KeyAdapter, can the same thing be done with a Canvas?

[ReBirth]

I had left canvas long time so see if I can recall...
first, you maybe confused how to get the big image (for your draw) from canvas right?

1   2   3   4

Graphics2D g = (Graphics2D) strategy.getDrawGraphics(); //g.drawAnything g.dispose(); strategy.show();

you get "strategy" object from getBufferStrategy();. Second, canvas can work with those adapters.

[Z-Man]

Thanks that helped, I was a little confused by the do-while part, your explanation helped. It's also nice to know that I can use the adapters.

[biro]

Hello,
I have a question to the above posted code.

}while(strategy.contentsRestored());

For what does this loop wait?

biro

[ra4king]

@biro @Z-man
You can read about the methods here

The do-while part is necessary for when the image gets lost and/or reinitialized (because it uses VolatileImage).

[Z-Man]

So for a 2D game in java would you recommend extending Canvas and controlling the looping, drawing and logic there. Then have a main method somewhere create a JFrame and add your custom Canvas to it. Or is there another method you would suggest.

[ra4king]

That's my recommended method

[ReBirth]

@Z-Man
I use that too
you can also have a canvas -> JPanel -> JFrame too so you can set the preferred size.

[biro]

Hello,
today I wrote a short programm, to play around with canvas and a little with the MouseListener.
For the Graphic part I used ra4kings given code or rather the one from http://ra4king.is-a-geek.net/javadocs/java/awt/image/BufferStrategy.html.
My Programm works fine, as long as I have a while(true) loop to refresh my frame contents. Buf if I remove the loop and add render() (rander is the method that redraws my canvas, it's also the one with the while(true) loop from before) tags, in exchange, to the methods that modify my graphicoutput, then I only see my "image" for a split second and then I only see gray until I do something that activates the render method again.

Hope you can help me with the short Explanation I gave you, if it's not enough tell me and I'll add the code

biro

Edit:
It's me again. I got my Programm so far, that it shows me the correct screen after I drag my mouse. But before I do that (in between the time after my startup and before the first mouseDragged() method starts)  I only see a gray screen, except for a short "blinking" in which I believe to see the correct screen.
Here the important parts of the code:

1   2   3   4   5   6   7   8   9   10   11   12   13   14   15   16   17   18   19   20   21   22   23   24   25   26   27   28   29   30   31   32   33   34   35   36   37   38   39   40   41   42   43   44   45   46   47   48   49   50   51   52   53   54   55   56   57   58   59   60   61   62   63   64   65   66   67

public synchronized void render() {            // Prepare for rendering the next frame    // ...    // Render single frame      BufferedImage g = null; //Green Image for a green background      BufferedImage s = null; //Black Image for a black background      try {          g = ImageIO.read(new File("green.png"));          s = ImageIO.read(new File("schwarz.png"));       } catch (IOException ie) {          ie.printStackTrace();          System.exit(0);       }             do {          // The following loop ensures that the contents of the drawing buffer         // are consistent in case the underlying surface was recreated         do {             // Get a new graphics context every time through the loop            // to make sure the strategy is validated            Graphics graphics = strategy.getDrawGraphics();             // Render to graphics            // ...            for (int x = 0; x < xMax(); x++) {                for (int y = 0; y < yMax(); y++) {                   BufferedImage draw = null;                   int drawx = this.activetilex+x;                   int drawy = this.activetiley+y;                   if (drawx >= this.tilesx || drawy >= this.tilesy || drawx < 0 || drawy < 0) { //Außerhalb des Arrays                     draw = s;                   } else {                      switch (feld[drawx][drawy].getTyp()) {                        case 1: //Grün                          draw = g;                           break;                        default:                           draw = s;                      }                   }                   if (x == 0 && y == 0) {                      BufferedImage newdraw = draw.getSubimage(this.activex, this.activey, this.tilewidth-this.activex, this.tileheigth-this.activey);                      graphics.drawImage(newdraw, 0, 0, null);                   } else if (x == 0) {                      BufferedImage newdraw = draw.getSubimage(this.activex, 0, this.tilewidth-this.activex, this.tileheigth);                      graphics.drawImage(newdraw, 0, y*this.tileheigth-this.activey, null);                   } else if (y == 0) {                      BufferedImage newdraw = draw.getSubimage(0, this.activey, this.tilewidth, this.tileheigth-this.activey);                      graphics.drawImage(newdraw, x*this.tilewidth-this.activex, 0, null);                   } else {                      graphics.drawImage(draw, x*this.tilewidth-this.activex, y*this.tileheigth-this.activey, null);                   }                }             }             // Dispose the graphics            graphics.dispose();             // Repeat the rendering if the drawing buffer contents            // were restored         } while (strategy.contentsRestored());          // Display the buffer         strategy.show();          // Repeat the rendering if the drawing buffer was lost      } while (strategy.contentsLost());            }

This is my render method.

1   2   3   4   5   6   7   8   9   10   11   12   13   14   15

private Canvas board;    private BufferStrategy strategy;    private MapMover mouselistener;    private GuiListener listener;        private int tilewidth = 50; //Breite eines Felds in Pixeln / Width of one field in pixels   private int tileheigth = 50; //Höhe eines Felds in Pixeln / Heigth of one field in pixels   private int showwidth = 800; //Fensterbreite /Windowwidth   private int showheigth = 600;//Fensterhöhe  /Windowheigth   private int activetilex = 0; //Welches Tile sehen wir gerade /Which tile on the x axis do we see   private int activetiley = 0; //dito /Same as above for the y axis   private int activex = 0; //Verschiebung des Tiles noch links / adjustment of the tile to the left   private int activey = 0; //Verschiebung des tiles nach oben /adjustment of the tile to top   private int tilesx = 25; //Anzahl an Tiles auf der x-Achse insgesammt /Sum of tiles on the x-axis   private int tilesy = 15; //Anzahl an Tiles auf der y-Achse insgesammt /Sum of tiles on the y-axis

Here the class attributes, these also have getter and setter methods!

1   2   3   4   5   6   7   8   9   10   11   12   13   14

public Gui() {       super();       initField();       listener = new GuiListener(this);       initGUI();       board.createBufferStrategy(2);       this.strategy = board.getBufferStrategy();             mouselistener = new MapMover(this);       board.addMouseListener(mouselistener);       board.addMouseMotionListener(mouselistener);       this.setVisible(true);       render();    }

And here my startup routine
 
Like I said the render method link in the startup method fabricates a short correct screenblink and than the screen get's gray agains.
Because the MouseListener and MouseMotionListener is doing it's job correct I won't post it now, if you ant it pls tell me.

Many thanks for your help
biro  

[Z-Man]

@ReBirth
I think that Canvas has a setPreferredSize() method as well so if you wanted you could probably cut out the JPanel there. Unless of course there are other things you need the JPanel for.

[ra4king]

You need to start a new thread where the run method has the game loop. So change render() to run(), implement Runnable, and start a new thread at the end of the constructor.

[biro]

Hello,
thanks for your answer.
But how does it help to implement a new Thread for the drawing? because I don't want an endless loop to redraw my graphic content, because it eats to many ressources (with the endlessloop before which is like a new thread, one of my cores was 100% used). Hope you can explain that to me, why I should implement a thread there^^

biro

[ra4king]

You need a constant FPS, which will not use the CPU much. That's what all games use, an endless game loop with a constant refresh rate and it updates the game logic and graphics.

[Sinuath]

you make it sound so simple... been reading a lot of tutorials and that line kind of made everything come together and make sense..  the idea of a video game seems really simple now

[biro]

Hello,
thanks for your answers, but in my case I don't need a constants update loop. I only want to update the screen, when it's need, which is after you drag the screen or you resize the frame and after you start the programm. So why should I use a thread, which burns so many many resources, and when I need a thread, how much should I reduce the the framerate (I believe I should do it with the sleep() command, so how many millisecs. should I let the thread sleep?).

Thanks a lot for your answers
biro   

[Mads]

Hello,
thanks for your answers, but in my case I don't need a constants update loop. I only want to update the screen, when it's need, which is after you drag the screen or you resize the frame and after you start the programm. So why should I use a thread, which burns so many many resources, and when I need a thread, how much should I reduce the the framerate (I believe I should do it with the sleep() command, so how many millisecs. should I let the thread sleep?).

Thanks a lot for your answers
biro   

If you want something that is not a still image, you need an update loop. You're just drawing pictures fast, so it looks like things move. If you don't update you have a picture without any animation.

[biro]

Hello,
my problem is not the animation, at the moment I have no real animation, I only need to update the field, when a user gives a command, that changes the world in some aspects. Then should the world be refreshed. And for that I don't need a loop if I'm not wrong, the problem is, that if I don't use a loop and draw the pictures on the screen I see them for a few milliseconds after these few milliseconds I see a gray window again, and I want to know why.
I know that when I want to add animations I need a constant loop, that refreshs my screen, but at the moment that is not needed
But let's look a bit in the future, to a time when I add animations or some other feature that needs a graphic refresh loop, than how should I drossel it, so that it won't eat 100% of my CPU power? My guess is that I should use sleep() for that, but how do I calculate the right amount of milliseconds my thread should sleep, so that my cpu doesn't need to create such a high output and on the other hand, that my "animation" is still fluid?
Please don't forget to answer the question about doing it at first without a loop and why the correct screen is only shown for a split second.

Greetings
biro

[Dx4]

Hello,
my problem is not the animation, at the moment I have no real animation, I only need to update the field, when a user gives a command, that changes the world in some aspects. Then should the world be refreshed. And for that I don't need a loop if I'm not wrong, the problem is, that if I don't use a loop and draw the pictures on the screen I see them for a few milliseconds after these few milliseconds I see a gray window again, and I want to know why.
I know that when I want to add animations I need a constant loop, that refreshs my screen, but at the moment that is not needed
But let's look a bit in the future, to a time when I add animations or some other feature that needs a graphic refresh loop, than how should I drossel it, so that it won't eat 100% of my CPU power? My guess is that I should use sleep() for that, but how do I calculate the right amount of milliseconds my thread should sleep, so that my cpu doesn't need to create such a high output and on the other hand, that my "animation" is still fluid?
Please don't forget to answer the question about doing it at first without a loop and why the correct screen is only shown for a split second.

Greetings
biro

use double buffering.