> A common counterexample to OO inheritance is the relationship between a square and a rectangle. Geometrically, a square is a specialisation of a rectangle: ev

> A common counterexample to OO inheritance is the relationship between a square and a rectangle. Geometrically, a square is a specialisation of a rectangle: every square is a rectangle, not every rectangle is a square. Well yes and no. A square and a rectangle are two kind of objects that have some common properties. In Math, for an object to belong to a set, it must have the properties described by the set. - Squares is the set of polygons with 4 sides of the same length and 4 right angles. - Rectangles is the set of polygons with 4 sides and 4 right angles. - Diamonds is the set of polygons with 4 sides of the same length. So yes, it's difficult to represent this relationship with inheritance. But no, in OOP, you should not consider those properties as data on a class, but more like a trait/interface. Your functions should take as input shapes implementing the expected traits/interfaces. Then what is a shape? It's a collection of points. But a point can be 2D, 3D, n-D. Here you would still not use inheritance but generics. Usually, I only use inheritance to specialize configuration, for example: class Foo { constructor(options) { ... } } class Bar extends Foo { constructor() { super({ some: 'options' }) } }