This position requires key competencies such as encapsulation, inheritance, polymorph

so if you apply for a job that has that line as a requirement what exactly are you supposed to know about inheritance, polymorphism and encapsulation?

Do you think it's about reciting definitions or just being aware of the fact that in OO languages such as C# it is possible to implicitly convert an expression of a derived type to an expression of a base type and explictly perform the reverse operation, as well as have derived classes that have members with the same name but different functionality?

No idea about what exactly is understood by the term encapsulation. (I don't know the dictionary definition though I might be aware of the concept). Does it have to do with taking the implementation of a type/type member away from the user as much as possible? Or is it about an expression of a base type 'encapsulating' an expression of a derived type?

Well. I'm gonna nit-pick and say that those are not competencies. Those are concepts. :P

I'd say that you should be able to explain how your language of choice supports these concepts and what the practices for their use is.

Quote:

---

Originally Posted by y99q

so if you apply for a job that has that line as a requirement what exactly are you supposed to know about inheritance, polymorphism and encapsulation?

---

I'd say you were in luck. Those are OOP 101 concepts, so presuming the job involves programming in a formal OO language such as C++, they are just asking for an absolute rock bottom minimum. A lot better than "at least 5-10 years experience with Unified Process", etc.

Encapsulation is the fundamental concept underlying OOP.* Primarily, it refers to the ability of a class to instantiate objects that contain both state and behaviour. An example of "state" in C++ would be a member such as int size. An example of "behaviour" in C++ would be a method such as grow(), which can change the state of the object by increasing int size. Secondarily, encapsulation refers to the possibility of regulating access to the internal state of an object. Eg, rather than expose int size as public, allowing the user to do whatever with an int, you keep it private and expose method grow(). That way, you can validate and delimit the manipulation of int size, and even change int size to a float internally later without changing the public interface grow().

However, don't walk into an interview taking my word for it, lol, because that's generally presented the other way around, such that encapsulation is primarily about "information hiding" and secondarily about internalizing behaviour. Which IMO is bassakwards if information hiding depends on the possibility of internalizing behaviour (which it does in practice).

* maybe the reason you are unsure about the definition is that asking an OO programmer to explain encapsulation might be like asking a fish to explain water ;)

Other key competencies include: ability to breathe, measurable pulse. Candidates who are dead will not be accepted.