I'll speculate for you Shane.
It is a run-time thing.
LVOOP class wires just carry a pointer that points to a data structure that in turn defines what is in the class and where to find in in core.
When we invoke a method LV uses the pointer to find the best fit for that type of wire based on the inheritance and availablity of the method used to figure which version of the method (from which class) gets called with the pointer found in the LVOOP wire. That VI works based off that wire and is done.
By using pointers and data structures LV can select the proper version of the method to use even at run-time in a built application.
We can alost liken it to how dll work under windows where the file association is looked up in the registry to figure who to call for this type of file but in LVOOP the registry is defined by the class hiarchy.
A virus can atack a PC by modifying the registry to point at itself instead of the proper dll so it can hook-in.
Done specualting and sharing my model of LVOOP.
Aristos COULD straighten me out if he wanted to but NI so closely guards how they implement stuff like this that I would be suprised if he did.
And if you (Stephen) are reading this wildly speuculative post, feel free to "put me in my place".
Ben
PS: Wire branching duplicate the wire and the buffer that goes with it.
PPS: Does the restriction LVOOP methods put on re-entrancy fit with the above? You tell me.
