Thanks. I'm just wondering why the example didn't show this extension method. It seems more idiomatic to just call it on the receiver instead of passing a parameter.
It's only more idiomatic if you're coming at this from an object-oriented perspective where anything "dispatchy" should be in receiver position. But once you let go of that notion, it's entirely to just have a function that does a type switch.
I am fine with applying either OO / FP approaches. This example is a sealed class, it is just my opinion that the OO version reads better. And the added benefit of having the function namespaced / bound to the class for IDE support (autocomplete) and documentation.
