Why are the std::atomic_{char,schar,etc.} typedefs allowed to be typedefs to a base class of std::atomic<T>, and not to atomic<T> only?

Question

C++11 [atomics.types.generic]p7:

There shall be named types corresponding to the integral specializations of atomic, as specified in Table 145, and a named type atomic_bool corresponding to the specified atomic<bool>. Each named type is a either typedef to the corresponding specialization or a base class of the corresponding specialization. If it is a base class, it shall support the same member functions as the corresponding specialization.

(emphasis added)

What's the rationale for letting std::atomic_char not be identical to std::atomic<char>, and so on for all the other types? What flexibility does this enable, and why is it useful? At first glance I don't understand why these wouldn't be required to be typedefs to std::atomic<T> specializations directly.

Note that N2427 which proposed <atomic> says that the typedefs are to specializations exactly, and not possibly to base classes.

Answer 1

It is probably there because someone already had an implementation of atomics using std::atomic_* types as base classes and complained enough/provided a good argument towards this approach.

In MSVC such an implementation is suggested.

In this it is stated that

Remove the definional base-class relationship between atomic_.... named types and the corresponding specializations of the atomic template class. The base-class relationship is now implementation-dependent. This change ensures compatibility with C. Some member functions and operators, that were formerly inherited from the base class, must be hoisted from the named types to the specializations.