Concluding Concepts

🔑 Concepts in C++20: an evolution or a revolution?

This small detour expresses my opinion. First, I present the facts, then I draw my conclusion. The facts are based on what has been presented in this chapter. So which arguments speak for evolution or for revolution?

Evolution

• Concepts promote working with generic code at a higher level of abstraction.
• Concepts give you understandable error messages when compiling a template fails. They provide nothing you could not achieve with the type-traits library, SFINAE, and static_assert.
• auto is a kind of unconstrained placeholder. With C++20, we can use concepts as constrained placeholders.
• With C++14, we could use generic lambdas as a convenient way to define function templates.

Revolution

• Concepts allow us, for the first time, to verify template requirements. Of course, you can also achieve the verification of template parameters with a combination of type-traits library, SFINAE, and static_assert, but this technique is way too advanced to regard it as a general solution.
• Thanks to the abbreviated function templates syntax, defining templates has been radically improved.
• Concepts represent semantic categories, but not syntactic constraints. Instead of a concept such as Addable, which requires that a type supports the + operator, we should think in terms of a concept Number. Number is a semantic category such as Equal or Ordering.

My conclusion

There are many arguments about whether concepts are an evolutionary step or a revolutionary jump. Mainly because of the semantic categories, I’m on the revolution side. Concepts such as Number, Equality, or Ordering remind me of Plato’s world of ideas. It is revolutionary that we can now reason about programming in such categories.