I’m using shared transitions to leave an anim state, using GetRelevantAnimTimeRemainingFraction (holy crap that is a long name 
). But it uses the animation from the original state the transition was made for, which isn’t the relevant animation.
Shared transitions seem to be exactly why you’d even create relevant animation functions. This would completely solves the problem of transitions referencing “hard coded” animations.
Hi Troy,
The relevant anim getters aren’t really set up to pull data for shared transitions as they are specific to a state (which ever state is mentioned in the name of the node). When we build the action list for the getter nodes we make sure from the source graph that there is a well defined state we can look at and the node is given all the information it needs to directly access that state.
We’d probably need a new node type to resolve this at runtime rather than have the information compiled into the nodes. I’ve added a feature request (UE-27645) to look at this in future.
If you are considering trying to get this info yourself the dynamic node will probably need to have references back to the baked transition entry (FAnimationTransitionBetweenStates) compiled into it rather than state information, and use that to access the machine description transitions array and grab the state index for the source state.
Thanks for your request!
Benn.
I see. That makes the use cases for shared transition logic to be pretty minimal, currently. Ok, thanks for putting the feature request in.