Why doesn't POPDecayAnimation use toValue? about pop HOT 14 CLOSED

POPDecayAnimation takes a fromPosition and velocity as initial values. The destination value is derived based on those inputs. Check out the README for an example.

from pop.

Ok, understood ... thanks for the clarification (and the project).

For decaying to a specific value, it looks like I'll have to do math :/

from pop.

Yeah, if you do the math you could adjust the decay constant to land at a specific place.

If you end up doing that (and want to contribute that function back), it's an API worth supporting. Could make the animation use the toValue if set, and fall back to the decay constant if not.

from pop.

@grp it seems not so convenient in use, i.e slide a view from bottom to top.

PS: what is the formula to calculate the velocity / deceleration based on toValue?

from pop.

@lzyy it looks like the current formula is in POPDecayAnimationInternal.h ... so, you could probably reverse it to do it based on toValue

from pop.

This could probably be usefuly to provide an edge. I want to decay animation on flick, but I want it to bounce on some specified edge. I can't see a way of doing that right now. Maybe the toValue could be used for that.

from pop.

I would appreciate something like what's been described here!

from pop.

If you're looking to bounce at an edge, a toValue wouldn't be very useful — it would just change where the decay ends up. Instead, I'd suggest swapping out the POPDecayAnimation for a POPSpringAnimation when you see the decay has passed the edge.

One way to check if you've passed an edge is in the animation delegate method pop_animationDidApply:. When the decay animation applies, you can check if it's past the edge. If it is, then start the spring animation & stop the decay. (You might copy the current velocity from the POPDecayAnimation to the POPSpringAnimation to make sure it feels smooth.)

from pop.

@lzyy I think you'd likely need to derive that formula from the existing POPDecayAnimation formulas. If you write them out as equations, I don't think it'll be that bad.

from pop.

@grp "swapping out the POPDecayAnimation for a POPSpringAnimation" is really inspiring, thanks.

yes, the formula is in POPDecayAnimationInternal.h, if kCAMediaTimingFunctionEaseOut is not sufficient, will use that.

from pop.

Anyone managed to do the reverse math for velocity/deceleration based on target position? I have a hard time with formulas here :/

from pop.

If POPDecayAnimation is just following standard deceleration rules this method should give you the acceleration/deceleration for the desired distance and velocity.

+ (CGFloat)decelerationForDistance:(CGFloat)distance withVelocity:(CGFloat)velocity
{
    if(distance == 0.0)
    {
        return 0.0;
    }
    if(velocity == 0.0)
    {
        return 0.0;
    }

    double acceleration = ((velocity * velocity)) / (distance * 2.0);
    double deceleration = 1.0 / fabs(acceleration);
    return MIN(0.9999999, MAX(0.0, deceleration));
}

I haven't tested it with POP, but try and let us know

from pop.

@hfossli ah now I see, I had some issues adapting calculated acceleration to deceleration value used by POP.

I put together a sample app here based on your function:
https://gist.github.com/pronebird/f62330a395712c41feaa

So basically I want to move my from point A (0, 0) to point B (100, 0) with initial velocity = (50, 0).

So we have:

distance = 100 units
acceleration = 12.5
deceleration factor = 0.999

POP calculated coordinate is: 44.956

decay = <POPDecayAnimation:0x79d03ed0; from = (0.000, 0.000); to = (44.956, 0.000); velocity = (50.000, 0.000); duration = 2.301463; deceleration = 0.999000>

from pop.

Well, my attempt was garbage. Seems like the formulas are quite more complicated looking at the POPDecayAnimationInternal.h. I guess what's complicating this is the rest/threshold-concept.

from pop.