I heard some new ideas about NFRs [nonfunctional requirements] - One was that the distinguisher is that FRs [functional requirements] are quantifiable and NFRs are not. I'm not sure that I think any requirement is not quantifiable if you try hard enough to make it so.In my opinion, these anecdotes only underscore just how misunderstood requirements are. People at a requirements conference were saying these things? Wow.
A further explained response was that using “testable” as a distinguisher, meant that the requirement can be tested by execution of the system (FR = function or performance) vs something you might test about the environment of the system (NFR = maintainability).
As I've mentioned, a nonfunctional requirement is just as testable and quantifiable as a functional requirement.
Functional requirements state what the system should do. The system either does it or doesn't do it. So sure, functional requirements are testable.
Nonfunctional requirements are simply metrics that we attach to functional requirements. While the system is doing whatever it's supposed to be doing, we can measure things like throughput (performance), how long it takes for a typical user to accomplish her goals (usability), the percentage of time the system is available to deliver any particular functionality (availability), etc. These nonfunctional requirements are by their very nature testable (at least in principle) and quantifiable.