My main takeaway from this slide is that as we have often been taught, striving for simple efficient code is what's important. This principle applies to the design of DSL's and systems as well since architects want to ensure the DSL they create isn't bloated with needless specificities that can easily be implemented by the programmer through existing constructs. Another way of seeing the benefit of this principle is thinking about how instead of creating a very specialized function, it's often better to create the underlying components in a way which align with your DSL's purpose(efficient graph operation) and can serve as a primitive to help generate more functionality.
My main takeaway from this slide is that as we have often been taught, striving for simple efficient code is what's important. This principle applies to the design of DSL's and systems as well since architects want to ensure the DSL they create isn't bloated with needless specificities that can easily be implemented by the programmer through existing constructs. Another way of seeing the benefit of this principle is thinking about how instead of creating a very specialized function, it's often better to create the underlying components in a way which align with your DSL's purpose(efficient graph operation) and can serve as a primitive to help generate more functionality.