I need to reproduce an inertial touch move and if I do simple parabolic equation the movement seems different from the one observed on a Android (and iPhone).
What is the equation they use?
I have done experiment to reverse engeneering their movement equation. The key is that they seem to respect scrolling time = K v0 (K constant and v0 initial speed), they respect D=K2*v0^2 like if there was gravity force opposing movement. But the movement speed at scrolling time/2 is signicantly slower than the one obtain with a parrabolic equation. I remarked the curve is always the same for all v0 once X and Y scaled.
After some long experiments with polynomial regression I determine approximative equation of the instant position:
P(t,v0)=v0t-7.1397t^2+25.08/v0T^3-37.43(t^4)/v0^2
v0 being the initial speed in page/s (seems capped at arround 12) ; P the instant position (in pages) ; t the time (in s) from the touchend event So the current speed is
v(t,v0)= v0-14.278t+75.24/v0t^2-149.72*(t^3)/v0^2 and the movement stop when v reaches 0.
At last, it seems better to apply a x2 coefficient between the last measured speed and the starting inertial speed v0 (not sure that they do)
It is not such a big utility but using this will make a slower movement ending and less uncontrolled fast speed at the middle.
I have done a script at https://github.com/utsb-fmm/MobileLikeScroller with the demo here https://utsb-fmm.github.io/MobileLikeScroller/