Just checking, but you know that translate does not actually affect the values of your camera coordinates right?

Hi asimes. Yeah, I realise that translate doesn't actually move the camera per se. The numbers I have there already.. can probably be ignored.. It's just there so I'm not putting an empty function call.. and to demonstrate how really confused I am. :S That is to say, what's pasted isn't the only thing I've tried. That was just the last attempt at "I really have no idea anymore so I'll just try random things".
I'm a bit extra upset because I've used the same method before in other programs and I've always managed to nut it out. I don't know what's wrong with my brain this time.

I guess the real problem is I'm not fully grasping everything. Obviously. I'm just embarrassingly unable to compute in my head whats going where, which is common enough and usually when I come back to a problem with a fresh start I get things to fall into place but this has been a bane for a few days.. er. derp me?

1. float angle = atan2(y, x);
2. angle = (angle >= 0) ? 0 : TWO_PI + angle;

 This will give a value in the range 0 to TWO_PI start on the X axis going clockwise

Thanks quarks! I'll give that a go. I'm an abject math failure and it surprises me I'm able to accomplish as much as I can... so I appreciate the help. :)

I'm sorry quarks, but can I ask a few questions? Do I still need to be translating? I assume I still want to translate to one set of points then use the other set in atan2()?

My answer although correct did not go far enough, sorry about that.

I assume the camera is used to define the view to be displayed, in which case you do NOT translate it. Yes calculate the camera's position and direction but then use camera() to set the view.

The correct sequence of actions are

1. calculate the current position, rotation, scale etc of any objects to be drawn
2. calculate the current position and view direction of the camera
3. for each object to be drawn
4.       pushMatrix
5.             apply translate, rotate and scale matrix for this object
6.             draw the object
   
7.       popMatrix
8. use camera() to set the view point (eye) and the direction of view

1 and 2 can be interchanged.
3 - 8 must be done in the draw method (so could 1 and 2 provided you maintain the overall order of actions)
5 and 6 assume that each object has its own co-ordinate system origin (0,0,0) for its vertices etc, and that the position of the object in your 3D world environment is (x,y,z). (This would involve a translate to x,y,z)

HTH
 

quarks, thank-you! You've done much typing. I've applied your suggestion as such:

1. if (cameraFree == false) {
2.     pushMatrix();
3.     translate(cameraX, cameraY);
4.     cameraDir = atan2(objectY[objectSelected], objectX[objectSelected]);
5.     popMatrix();
6.    
7.     cameraDir = (cameraDir >= 0) ? 0 : TWO_PI + cameraDir;      //provided by "quarks"/processing forum
8.    
9.     cameraVel = sqrt(dist(cameraX, cameraY, objectX[objectSelected], objectY[objectSelected])) / 2;
10.     cameraVelX = cos(cameraDir) * cameraVel;
11.     cameraVelY = sin(cameraDir) * cameraVel;
12.     cameraX += cameraVelX;
13.     cameraY += cameraVelY; 
14. }

I appreciate your comments! I have no problem with drawing things in the proper places, etc. as the camera (transformation matrix) moves around. My problem is in trying to get a smooth camera traverse/pan when different objects are selected instead of a hard transition to the object (Which does work fine right now. When I select something the view snaps so its displayed in the center of the screen as it should). As I have it there it still unfortunately doesn't work quite properly. (and I won't lie, I don't understand the syntax you used 100%). The camera does move, but not usually in the right direction and never seems to stop moving.

Yes, it's just a 2-D world. I name everything related to moving the transformation matrix where the view was concerned as 'camera' though, even though I'm aware it's not actually moving a camera object, but the final effect is the same. That's the kind of thing that makes sense to me. Is that a bad idea? haha. I'm more than willing to accept people that tell me some things I do are stupid.

I don't understand the syntax you used 100%

Since I only put two lines of actual code I assume you mean this.
angle = (angle >= 0) ? 0 : TWO_PI + angle;
and just as well you did mention it because it's wrong and should be
angle = (angle >= 0) ? angle : TWO_PI + angle;

So in line 7 of your code should be
cameraDir = (cameraDir >= 0) ? cameraDir : TWO_PI + cameraDir;
sorry about that.

So I will explain this first.
Actual angle               Value returned by atan2
0  ..  180                  0  ..  PI    (positive values)
180  ..  360               -PI  .. 0     (negative values)

So if we have a positive value then no problem but if its negative then we want to add TWO_PI to correct the angle. We could do this with an if-else statement.
if(angle >= 0)
      angle = angle; // no change
else
      angle = TWO_PI + angle;

This can be replaced by my ( corrected code). The general format statement for the ?: operator is
(boolean expression) ? (result if true) : (result if false)


Now back to your last post:
Obviously I can't see all of your code but I assume that when (cameraFree == false) is true then the program should move the camera towards a new location?

The camera does move, but not usually in the right direction and never seems to stop moving.

The direction problem is probably due to my mistake above so try that first to see if it fixes it. Just noticed in line 4 the actual calculation is wrong in that it calculates the angle from the orign (0,0) to the selected object rather than the angle from the camera. I believe line 4 should be
cameraDir = atan2(objectY[objectSelected] - cameraY, objectX[objectSelected] - cameraX);

It will only stop moving when cameraFree == true. You need a way of detecting when the camera is near enough to the selected object. I see that you are applying easing on the velocity so that it approaches zero when the camera nears the destination putting all this together means we have something like this

1. if (cameraFree == false) {
2.     pushMatrix();
3.     translate(cameraX, cameraY);
4.     cameraDir = atan2(objectY[objectSelected] - cameraY, objectX[objectSelected] - cameraX);
5.     popMatrix();
6.    
7.     cameraDir = (cameraDir >= 0) ? cameraDir : TWO_PI + cameraDir;   
8.    
9.     cameraVel = sqrt(dist(cameraX, cameraY, objectX[objectSelected], objectY[objectSelected])) / 2;
10.     if(cameraVel < 1.0){ // need to expriment to get a reasonable value
    
11.             // reached destination
12.             cameraVelX = 0;
13.             cameraVelY = 0;
14.             cameraFree = true;
15.       }
16.       else {
17.             // further to go yet
    
18.           cameraVelX = cos(cameraDir) * cameraVel;
19.           cameraVelY = sin(cameraDir) * cameraVel;
20.           cameraX += cameraVelX;
21.           cameraY += cameraVelY; 
22.       }
23. }

Well quarks, you made this bloke smile, anyway. I deserve not such a grand reply!

Anyway.. Your assertions are correct with cameraFree etc. ..You know I did have a check against the distance to stop the camera moving originally but nothing else was working properly anyway and there's part of my brain that says that if it's right on the spot the camera won't be moving anyway (or will always attempt to return to the center and thus achieve an equilibrium)?

SO! I applied your code and there must be something else wrong somewhere because I achieve the same result as some of my old tries that I thought should be working properly - based on my other sketches I mention where I'd done similar and it worked fine - before I came here. It's.. really difficult to describe. And.. I have no idea what else could be throwing things off. My sketch isn't that complicated yet. D:

If I uploaded it to OpenProcessing (or emailed a zip or something. I don't have webspace anymore or i'd do that) would you be willing to look?


EDIT: Oh wait. I have discovered something. I'm not sure what yet, but yes, it does look like perhaps my other translate() is messing things up, at a guess. I simplified everything and stuff is messed up. - Looking into this now :)