Howdy!

I'm working on my game and am having problems figuring out how to determine which direction to turn an enemy ship so that it moves toward the player. Basically, I get the angle between the player and the enemy ship with atan2(). Then I want to take that, factor in the current angle of the ship, and determine whether the path is shorter between the two angles is clockwise or counter-clockwise. Positive turnSpeed turns the ship clockwise; negative values turn the ship counter-clockwise.

Here's what I have:



That magical algorithm is really stumping me. Any help sorting this out is welcome.

_________________
What most people don't understand about "enlightenment" is that it is not an end-goal; but where you find yourself just before taking a new "first step."

I think this snippet (pulled out of a quick java test) should do the trick.


Assuming degrees in the range 0-360 (unlike stupid Flash's -180 to 180).

Hope it works!

_________________
www.littlemonkey.co.nz

If you want some theory, I'll give you that.

The basic idea is to measure the distance to turn each way, and choose whichever is the shortest.


(Assuming +ve is clockwise)

Let D = targetangle - angle
clockwise turning distance C = (D) modulo 360 = D%360
anticlockwise turning distance A = (-D) modulo 360 = (-D)%360

And whichever is smallest we go that way.




Another way of looking at it is like this:

If D is between -180 and 0, then C=D+360 which is between 180 and 360.
Meanwhile A= -D which is between 0 and 180. Therefore A<C and we go anticlockwise.

If D is between -360 and -180, then C=D+360 which is between 0 and 180.
Meanwhile A= -D which is between 180 and 360. Therefore C<A and we go clockwise.

If D is between 0 and 180 then C=D, while A=360-D which is between 180 and 360. Therefore C<A and we go clockwise.

If D is between 180 and 360 then C=D, while A=360-D which is between 0 and 180. Therefore A<C and we go anticlockwise.

That logic can be written like this:

_________________
I ain't pushing no moon buttons.

Also, you don't need to calculate any angles.

You can use the vector cross product to determine which way to move.

Suppose ship is at position (x, y)
Suppose target point is at position (tx, ty)
Suppose the ship is currently pointing along vector (rx, ry)

Calculate displacement:
dx = tx - x
dy = ty - y

Do cross product:
CP = dx * ry - dy * rx

If CP>0, you go one way, if CP<0 you go the other way...

Direction = SGN(dx*ry - dy*rx)


Hope that helps.

_________________
I ain't pushing no moon buttons.

I'm not using vectors, just an X, Y coordinate and an angle. How do I get rx/ry without resorting to vectors?

Also, what does the function SGN() represent?

_________________
What most people don't understand about "enlightenment" is that it is not an end-goal; but where you find yourself just before taking a new "first step."

Well you would use cos and sin

But it depends on where you're measuring angles from

If angle = 0 corresponds with the +y axis, then this would work:



You are probably calculating sin and cos already to get components of velocity/acceleration.

SGN is the sign function.

_________________
I ain't pushing no moon buttons.

Doing this:



Instead. It works now (finally!!!). Thanks so much!

Is SGN a common function in programming languages? I'd never heard of it before. If it's a general function in math, well, you have to remember that my math IQ is below zero.

_________________
What most people don't understand about "enlightenment" is that it is not an end-goal; but where you find yourself just before taking a new "first step."

okay, but you know that measures angles anticlockwise?




Yes it's a common math function. But Java likes latin, and calls it 'signum'.

http://download.oracle.com/javase/1.5.0 ... gnum(float)

Java signum dolor sit.

_________________
I ain't pushing no moon buttons.

Anti-clockwise? Hmm... strange. Maybe that's the HTML5 canvas or JavaScript doing that, but it could just as easily be something I'm doing somewhere else. Weird.

JavaScript doesn't have a function similar to signum or sgn, as far as I know. I might have just missed it (I have a book that references all the methods and properties in the Math library, but it is a bit old).

_________________
What most people don't understand about "enlightenment" is that it is not an end-goal; but where you find yourself just before taking a new "first step."

Oh I know now... it's because the axes of the screen are left handed (+y is downward), so clockwise and anticlockwise appear to be switched.

You could write your own SGN function.

_________________
I ain't pushing no moon buttons.

Oh yeah! That makes perfect sense. Not sure why none of us had thought about that sooner.

_________________
What most people don't understand about "enlightenment" is that it is not an end-goal; but where you find yourself just before taking a new "first step."

Now this got me thinking, what would a programming language be like if all of it's keywords were in Latin?

_________________
I ain't pushing no moon buttons.

Short answer: hard to read (for me at least).

You do think of the strangest things Jasmine, though I admire your dedication to following through.

_________________
www.littlemonkey.co.nz

I feel that new ways of looking at things or combining things is good. Exploring possibilities is the force behind creativity, the imagination, invention, and the intellect. "what if" makes life interesting.

_________________
I ain't pushing no moon buttons.

Hahaha, that's awesome, Jasmine!

_________________
Worlds at War (Current Project) - http://www.awkward-games.com
Ganadu'r, The Eternal Sage (Other Current Project) - http://rpg.naget.com
Programming tutorials and web-design services: http://www.wyrmmage.com