Overriding Supervisorstrategy as a val in concrete Actor

Question

The Actor trait in Scala is defined like so,

/**
* User overridable definition the strategy to use for supervising
* child actors.
*/
def supervisorStrategy: SupervisorStrategy = SupervisorStrategy.defaultStrategy

However i see that all example code of concrete actors are defined as follows,

override val supervisorStrategy = OneForOneStrategy(loggingEnabled = false) {
    ...
  }

What i am trying to understand is why is the supervisorStrategy overridden as a val ?
Why not override it as a def?

Is it because of the memory considerations , like we dont want this evaluation every time the method is called ?

Answer 1

So first up a scala trait is a place to keep abstract definitions that we know would generally need overriding. Hence there is really no point in defining a method as val , except for cases where the method is meant to be called by all concrete extensions of that trait.

Typically when we extends a trait with a def method we have all options open . That is we can override it as a def , a val or a lazy val. The choice depends on the use case more often. In this case we do not expect the supervisor strategy to change each time the superVisorStrategy method is called from the akka subsystem. So once the method has executed we wish to hold on to the value for the future thus saving on some memory. Hence it is defined as a val or a lazy valin most examples.