Timing function updates

[typeoneerror]

Seems like the timing functions have been locked down and removed some of the original parameters from Penner's methods.

CGFloat PRTweenTimingFunctionElasticOut (CGFloat t, CGFloat b, CGFloat c, CGFloat d) {
    CGFloat p = d*.3;
    CGFloat s, a;
    if (t==0) return b;  if ((t/=d)==1) return b+c;
    if (!a || a < ABS(c)) { a=c; s=p/4; }
    else s = p/(2*M_PI) * asin (c/a);
    return (a*pow(2,-10*t) * sin( (t*d-s)*(2*M_PI)/p ) + c + b);
}

This for example is missing the amplitude and period params, so much of the logic testing for a isn't needed if you're going to hard code those values right? Just noticed this because when running Analyze you get "a evaluates to a garbage value." Any reason that you got rid of those parameters?

[domhofmann]

It's a remnant from a quick port job of Penner's curves. I think ultimately we want to leave those parameters in, as they can be quite useful and it bothers me that they're hard coded now. There's a question hanging in the air about best approach to this.

One thought is to modify the functions to look like this:

CGFloat PRTweenTimingFunctionElasticOut (CGFloat t, CGFloat b, CGFloat c, CGFloat d, NSDictionary *options) {
    ...
}

In this case I think we'd expose it through a property on PRTweenOperation (call it timingFunctionOptions or something similar.)

With shorthands, it could be handled in two ways. We'd get chaining for free:

[PRTween tween:&someValue from:0 to:100 duration:1 timingFunction:&PRTweenTimingFunctionElasticOut].timingFunctionOptions = options;

Or we could add/modify the method signature to include an options argument, something like [PRTween tween:from:to:duration:timingFunction:options:]

What are your thoughts?

[typeoneerror]

I'd say add a new method signature (not modify) like the last one that allows you to send options dictionary, and have the original method just pass nil to that method. I think 9.9 times out of 10 people will just use the default values that are currently hard-coded, but I always prefer a "complete" API, ya know? Seems like a low priority update.

[domhofmann]

Working on this and it should be up soon.

As a side note, don't hit E by accident on the issues page.

[typeoneerror]

lol, I had to hit 'e' to test, naturally.

[typeoneerror]

Looking forward to the update. I submitted a quick pull request for a typo and a weird "/" I found inside one of the transition functions.