1. 5

  2. 1

    Certainly a famous paper. My own feeling is that once one moves away from implementation details where one can argue that one can be the other, objects can be used to implement abstract data types. This is a use to which objects are particularly suited, as evidenced by so many books and courses using objects for that purpose.

    Where I diverge from Cook’s assumption that both are fundamentally about data representation, is that I am more of the “scandinavian school” perspective that objects in object orientation are more generally about the representation of phenomena, real or imagined. I myself refer to this as the “Nygaard perspective” due to it being the viewpoint of one of the two fathers of object-orientation - the other being Dahl.

    1. 1

      Good stuff, thanks for posting this, I am glad I’ve learned about this paper.

      https://www.tedinski.com/2018/02/27/the-expression-problem.html is an interesting post which explores similar ideas. I particularly like the Data/ADTs/Objects diagram there.