Conversation

Animation APIs parameterized by eg duration and curve are fundamentally opposed to continuous, fluid interactivity.

152

What’s wrong with additive keyframes? No velocity but still youtube.com/watch?v=M-8_9v source

github.com

GitHub - KevinDoughty/AdditiveSamples: Sample code showing explicit additive animation techniques.

Sample code showing explicit additive animation techniques. - GitHub - KevinDoughty/AdditiveSamples: Sample code showing explicit additive animation techniques.

You have a .5s toggle animation. Interrupt it halfway through using additive anims and it takes full .5s to go half as far.

Indeed, every animation has same duration. But distance travelled > half. Additive path is not linear.

Sure, I know. My point is that v(t) doesn't vary in a predictable or designed way with interruption timing.

I admit my physics was lame, but maybe altering the spring timing function based on velocity could be a solution?

Yeah, that's basically what POPSpringAnimation does. There's no duration in that model though.

POP mutates with much less performance than readonly, GPU & threaded CA, is my (admittedly limited) understanding.

Replying to

It's the same approach UIScrollView deceleration uses. But sure, yes, in certain scenarios. Talking concepts here not impls

9:58 PM · Feb 15, 2015Twitter for iPhone

Replying to

Given set of inputs surely duration is deterministic. Six & one half. Except performance, forward filling, mutate