Expressing memory ownership in C++

Throughout its history, the C++ language has evolved in its approach to expressing memory ownership. The same syntactic constructs have been, at times, imbued with different assumed semantics. This evolution was partially driven by new features added to the language (it’s hard to talk about shared memory ownership if you don’t have any shared pointers). On the other hand, most of the memory management tools added in C++ 11 and later were not new ideas or new concepts. The notion of a shared pointer has been around for a long time. This language support makes it easier to implement one (and having a shared pointer in the standard library makes most custom implementations unnecessary), but shared pointers were used in C++ long before C++ 11 added them to the standard. The more important change that has occurred was the evolution of the understanding of the C++ community and the emergence of common practices and idioms. It is in this sense...