Let's say we have two fragments: A and B
A is visible, we want to replace it with B
A has an exitTransition, B has an enterTransition
Both transitions define a TransitionPropagation that they use to delay some animations. Those TransitionPropagations collect values that are needed for the start delay with captureValues.
What I expect when I commit my transaction from A to B:
TransitionPropagation.captureValuesis called for all relevant views fromA's layout for theexitTransitionofATransitionPropagation.captureValuesis called for all relevant views fromB's layout for theenterTransitionofB
What I see:
TransitionPropagation.captureValues is called for all relevant views from A's layout for the exitTransition of A and then TransitionPropagation.captureValues is also called on the same Propagation for views that are defined in B's layout.
What is happening here? Why are views from B directed to the exitTransition of A?