I've been looking around for a while, and I have not been able to find an answer to this for some reason. It seems simple enough, but maybe I just can't find the right function in the library.
I have a scene with a layer that contains a bunch of CCNodes with each one CCSprite in them.
During the application, I move around the position of the main layer, so that I "pan" around a camera in a way. (i.e. I translate the entire layer so that the viewport changes).
Now I want to determine the absolute position of a CCNode in screen coordinates. The position property return the position relative to the parent node, but I really would like this transformed to its actual position on screen.
Also, as an added bonus, it would be awesome if I could express this position as coordinate system where 0,0 maps to the upper left of the screen, and 1,1 maps to the lower right of the screen. (So I stay compatible with all devices)
Edit: Note that the solution should work for any hierarchy of CCNodes preferably.
Every CCNode, and descendants thereof, has a method named convertToWorldSpace:(CGPoint)p
This returns the coordinates relative to your scene.
When you have this coordinate, flip your Y-axis, as you want 0,0 to be in the top left.
CCNode * myNode = [[CCNode alloc] init];
myNode.position = CGPointMake(150.0f, 100.0f);
CCSprite * mySprite = [[CCSprite alloc] init]; // CCSprite is a descendant of CCNode
[myNode addChild: mySprite];
mySprite.position = CGPointMake(-50.0f, 50.0f);
// we now have a node with a sprite in in. On this sprite (well, maybe
// it you should attack the node to a scene) you can call convertToWorldSpace:
CGPoint worldCoord = [mySprite convertToWorldSpace: mySprite.position];
// to convert it to Y0 = top we should take the Y of the screen, and subtract the worldCoord Y
worldCoord = CGPointMake(worldCoord.x, ((CGSize)[[CCDirector sharedDirector] displaySizeInPixels]).height - worldCoord.y);
// This is dry coded so may contain an error here or there.
// Let me know if this doesn't work.
[mySprite release];
[myNode release];
Related
Let's say the width of a level in my game is three times the screen width, and my player starts at the left most edge. How should I go about implementing a camera flyby starting from the right edge at the beginning of this level to scroll through the entire world so the player knows what to expect ahead?
I came across this post here http://www.cocos2d-iphone.org/forum/topic/9568 that seems to be useful, borrowed the block of code in it and put it in my GameWorldLayer, but it didn't work. I'm essentially only seeing a black screen. What's wrong? And what would be a better way of implementing this?
-(void)visit {
CGSize screenDims = [[CCDirector sharedDirector] winSizeInPixels];
CGPoint camPos = gameCamera.position;
float camZoom = gameCamera.zoom;
[[CCDirector sharedDirector] setProjection:kCCDirectorProjectionCustom];
//now set your projection
kmGLMatrixMode(KM_GL_PROJECTION);
//save current projection state
kmGLPushMatrix();
kmGLLoadIdentity();
kmMat4 orthoMat;
kmMat4OrthographicProjection(&orthoMat,
camPos.x -screenDims.width/(2*camZoom),
camPos.x +screenDims.width/(2*camZoom),
camPos.y +screenDims.height/(2*camZoom),
camPos.y -screenDims.height/(2*camZoom),
-1000,
1000);
kmGLMatrixMode(KM_GL_MODELVIEW);
kmGLLoadIdentity();
[super visit];
//put it back
kmGLMatrixMode(KM_GL_PROJECTION);
kmGLPopMatrix();
kmGLMatrixMode(KM_GL_MODELVIEW);
kmGLLoadIdentity();
}
All that is way to complicated for what you want. Position the layer at the end of the level at the beginning and then simply move it to the beginning i.e.
[layer runAction:[CCMoveTo actionWithDuration:10 position:ccp(starty, startx)]];
Where 20 is the time you want it to take for the "flyby" and the position is the start position
I have a standard cocos2d startup layer( HelloWorldLayer). I created another class of type CCNode named "Terrain" for my terrain. Then i add it to my layer in the layer's init:
terrain = [[Terrain alloc] initWithWorld:world AndLevel:0];
[self addChild:terrain z:1];
i add a 'CarObject' class (a CCSprite class), and add a car object to my terrain
car = [[CarObject alloc] initWithWorld:world];
[terrain addChild:car];
-i.e. in both the initWithWorld for terrain and car, i initialize some Box2d code
I then try to center my car object to my screen when i move it, i do this in my update method:
float offsetX = car.position.x;
float offsetY = car.position.y;
[terrain setOffsetX:(int)offsetX andOffsetY:(int)offsetY];
where the setOffsetX.. method is:
- (void) setOffsetX:(int)newOffsetX andOffsetY:(int)newOffsetY {
_offsetX = newOffsetX;
_offsetY = newOffsetY;
CGSize winSize = [CCDirector sharedDirector].winSize;
self.position = CGPointMake(-(_offsetX - winSize.width/2), -(_offsetY - winSize.height/2));
}
When i use a NSLog to see if the terrain position changes, i can see that the position actually chages, but the view does not. What am i doing wrong? am sure it's a dumb mistake!
btw, if i try this in my HelloWorldLayer's update method (instead of [terrain setOffsetX..])
self.position = CGPointMake(self.position.x-1, self.position.y);
the terrain is moving.
Car is a child of Terrain. Car's position is therefore relative to Terrain's position. Since you base Terrain's position on Car's position, which is actually relative to Terrain's position, you may be simply running into the effect that your position updates simply cancel each other out.
If you want to move the Terrain while keeping the Car centered, you shouldn't add the Car as a child of Terrain. Instead add it to the same node as the Terrain (HelloWorldLayer). Then you can move the Car and Terrain independently of each other.
I'm having a bit of a problem solving this trivially-looking feature. In my game, at some point in a level, I stop the main game loop and want to zoom to the player sprite. The player sprite is added to a layer that contains this loop and run the code to zoom to this player.
It looks like this :
-(void) stopGameAndZoomToPlayer {
[gamePlayLoop_ invalidate];
zoomToPlayerLoop_ = [NSTimer scheduledTimerWithTimeInterval:1.0f/60.0f target:self selector:#selector(tickZoomToPlayerLoop:) userInfo:nil repeats:YES];
}
The tick method, is for now implemented like this :
-(void) tickZoomToPlayerLoop:(NSTimer*)timer {
if (self.scale > 4) {
[zoomToPlayerLoop_ invalidate];
}
[self setScale:self.scale+0.001];
CGSize winSize = [[CCDirector sharedDirector]winSize];
CGPoint targetPosition = playerSprite_.position;
float wantedX = ((winSize.width/2-targetPosition.x)/self.scale);
float wantedY = ((winSize.height/2-targetPosition.y)/self.scale);
CGPoint maximumCenter = ccp(wantedX,wantedY);
[self setPosition:maximumCenter];
}
Seems trivial right ? Yet, there's something I must be missing because the zoom, while slow, is offset vis-a-vis the player sprite. The player sprite, at the end of the tickZoomToPlayerLoop timer, appears in the top right corner of the screen ; it is not in the center (but not even centered of the top right corner of the viewport).
The way I see it, this code put the center of the main layer to the position of the player sprite. Then it scales. Actually the position shouldn't be in the zoom loop but I put it here because I'll need to handle the boundaries of the level layer, because as for now the user can see the black void of non playable zone, outside the level. Yet, even if the center of the main layer at the beginning of the loop seems to be the player sprite, when scaling, one can see it is not.
Does the scale is targeted to the center of a CCNode ? Is it offseted in any way ? Does the anchor point plays a part in the way the scaling is done ?
Some more infos :
the anchor point of the main layer is (0.5,0.5) - the default one
the position of the main layer is, at first before the timer, (0,0) - simply added when creating the scene
Scaling to my knowledge is effected by anchor points.
If all you need to do is zoom in way not use CCCamera?
[myLayer.camera setEyeX:0 eyeY:0 eyeZ:zoomValue];
You can change the eyeZ to zoom in and out.
I am rotating main view with 360 degrees, and I have subviews added inside main view, everything works correctly, but with one issue.
What I want to do is when I rotate main view, inner views should not lost their frames/position. Right now, when I rotate main view with infinte repeat count and dynamically if I add subview inside main view, it goes into proper position, but it does not retain its frame.
For example, I am implementing orbit, and for that, I have used entire transparent view as orbit and orbit is rotated from center point to 360 degree infinite times, and User can add many planets as he wants onto orbit, so when planets added on orbit, planets do not retain its frame. Can you suggest any idea?
Thanks in advance.
Well it sounds like you need to add a rotating animation for every subview that you add in your main view. If the main view rotates clockwise your subviews will need to rotate around their center in a counter-clockwise direction.
I guess you're trying to keep the subviews' orientations while rotating.
If I were you, I'd use CAAnimation instead of using a view to rotate.
You may add the animation to every subview, try this:
CAKeyframeAnimation* animation;
animation = [CAKeyframeAnimation animation];
CGMutablePathRef path = CGPathCreateMutable();
CGPathMoveToPoint(path, NULL,imgview.layer.position.x,imgview.layer.position.y);
int p = [self getblank:tag];
float f = 2.0*M_PI - 2.0*M_PI *p/PHOTONUM;
float h = f + 2.0*M_PI *num/PHOTONUM;
float centery = self.view.center.y;
float centerx = self.view.center.x;
float tmpy = centery + RADIUS*cos(h);
float tmpx = centerx - RADIUS*sin(h);
imgview.center = CGPointMake(tmpx,tmpy);
CGPathAddArc(path,nil,self.view.center.x, self.view.center.y,RADIUS,f+ M_PI/2,f+ M_PI/2 + 2.0*M_PI *num/PHOTONUM,0);
animation.path = path;
CGPathRelease(path);
animation.duration = TIME;
animation.repeatCount = 1;
animation.calculationMode = #"paced";
return animation;
I assume you have a stored variable that represents the rotation of the "world", correct? If not, you should.
Then, for each image you add, also store a variable with it that represents its rotation to the world.
For example, if your world is rotated 180°, and you added a cup (which you want to appear right-side up when the world is upside-down) the cup's "offset" to the world rotation would be -180°.
Then, if the world is at 180° and you rotate your world by adding 90°, then the cup's new rotation value would be cup_rotate_offset + world_rotation, or 180°+270°, which is the same as saying 90°, and the top of the world would be facing left and the cup's top would be facing right.
You have to independently track the offset values for each added object.
If I start with a UIImageView, and I add a subview, how do I translate a coordinate in the original UIImageView to a corresponding coordinate (the same place on the screen) in the subview?
UIView provides methods for exactly this purpose. In your case you have two options:
CGPoint newLocation = [imageView convertPoint:thePoint toView:subview];
or
CGPoint newLocation = [subview convertPoint:thePoint fromView:imageView];
They both do the same thing, so pick whichever one feels more appropriate. There's also equivalent functions for converting rects. These functions will convert between any two views on the same window. If the destination view is nil, it converts to/from the window base coordinates. These functions can handle views that aren't direct descendants of each other, and it can also handle views with transforms (though the rect methods may not produce accurate results in the case of a transform that contains any rotation or skewing).
Subtract the subview's frame.origin from the point in the parents view to the same point in the subview's coordinate:
subviewX = parentX - subview.frame.origin.x;
subviewY = parentY - subview.frame.origin.y;
Starting with code like:
UIImageView* superView=....;
UIImageView subView=[
[UIImageView alloc]initWithFrame:CGRectMake(0,0,subViewWidth,subViewHeight)
];
subView.center=CGPointMake(subViewCenterX, subViewCenterY);
[superView addSubview:subView];
The (subViewCenterX, subViewCenterY) coordinate is a point, in superView, where the center of subView is "pinned". The subView can be moved around wrt the superView by moving its center around. We can go, for example
subView.center=CGPointMake(subViewCenterX+1, subViewCenterY);
to move it 1 point to the right. Now lets say we have a point (X,Y) in the superView, and we want to find the corresponding point (x,y) in the subView, so that (X,Y) and (x,y) refer to the same point on the screen. The formula for x is:
x=X+subViewWidth/2-subViewCenterX;
and similarly for y:
y=Y+subViewHeight/2-subViewCenterY;
To explain this, if you draw a box representing the superView, and another (larger) box representing the subView, the difference subViewWidth/2-subViewCenterX is "the width of the bit of the subView box sticking out to the left of the superView"