When to Use Tuples vs Structs

Tuples and user-defined structs/classes each have their strengths:

Advantages of tuples:

• Convenient built-in functionality like std::get, std::tuple_cat, std::make_tuple
• No need to define your own type
• Easily return multiple values from a function
• Generic code can operate on tuples without knowing the specific types

Advantages of structs/classes:

• Elements have meaningful names rather than just indices
• Can include methods, constructors, access control, etc.
• Clearer intent and easier to maintain for larger numbers of elements
• Can be forward declared and used in other files

In general, tuples are great for small, temporary groupings of values, especially when returning data from functions. They allow bundling data without defining a dedicated type.

However, when a collection of data has a clear identity and will be used in many places, defining a struct or class is usually better. It makes the code more readable and allows encapsulating related functionality with the data.

As a rule of thumb, if you find yourself using std::get<3>, std::get<8>, etc. with a tuple, consider defining a struct with named fields instead. The resulting code will be clearer and more maintainable.