Mixing Objective-C and C++: Game Loop Parts

Question

I'm trying to write all of my game in C++ except for drawing and game loop timing. Those parts are going to be in Objective-C for iOS.

Right now, I have ViewController handling the update cycle, but I want to create a GameModel class that ViewController could update. I want GameModel to be in C++. I know how to integrate these two classes. My problem is how to have these two parts interact with the drawing and image loading.

GameModel will keep track of a list of children of type GameObject. These GameObjects update every frame, and then need to pass position and visibility data to whatever class or method will handle drawing.

I feel like I'm answering my own question now (talking it out helps) but would it be a good idea to put all of the visible game objects into an array at the end of the update method, return it, and use that to update graphics inside ViewController?

Answer 1

You want to try and avoid inter-mingling objective-c and c++ as much as possible, it will create code that is very hard to debug and even harder to understand by others potentially joining your project.

It is essential that you ground a sense of framework and work in a modular fashion in respect to how the platform wants to you to interact with objective-c modules and how your c++ modules fit with these.

If you're using c++, that indicates to me that you want to abstract away from objective-c and use c++ as your main language, then do so. Objective-c will only handle the low-tier ios system-specific features.

What purpose do you have for crossing the two languages over in such a complicated way, this speaks to me of bad software design and poor consideration of design methodologies.

I can only advise to keep it straight forward:

//Init In your ViewController.mm just invoke a main engine class(c++) such as:

-(void)viewDidLoad
{
/*unrelated ios-specific init code*/
[super viewDidLoad];

self.context = [[EAGLContext alloc] initWithAPI:kEAGLRenderingAPIOpenGLES2];

engine = new engine();//your engine class, that contains init, update and draw methods

if(!self.context ||  ![EAGLContext setCurrentContext:self.context])
{
    NSLog(@"Failed to create ES context");
}

GLKView * view = (GLKView *)self.view;
view.context = self.context;

GRect screenRect = [[UIScreen mainScreen] bounds];
CGFloat screenWidth = screenRect.size.width;
CGFloat screenHeight = screenRect.size.height;


view.drawableDepthFormat = GLKViewDrawableDepthFormat16;


if(engine != NULL)
    engine->init(screenWidth, screenHeight);
}

//Update

-(void)update {
engine->update();  
}

//Draw

- (void)glkView:(GLKView *)view drawInRect:(CGRect)rect
 {
   engine->draw((GLfloat)view.drawableWidth,(GLfloat)view.drawableHeight);
   [self.context presentRenderbuffer:GL_RENDERBUFFER];
 }

You will need to assert a solid architecture for communicating between your lower-level objective-c and higher-level c++ classes if you wish to use all the features the platform has to offer.