Text encodings/locales: missing normative association?

[tkoeppe]

I think we may be missing some normative wording that says that locales are associated with encoding schemes.

Such wording would tell us whether the use of "the implementation-defined encoding scheme associated with the locale *this" in locale::encoding is a definition or a reference.

Originally posted by @tkoeppe in #6359 (comment)

[tkoeppe]

@cor3ntin, @jwakely: Thoughts?

[cor3ntin]

I think there is some allusion to that here http://eel.is/c++draft/library#character.seq.general-1.1 and there http://eel.is/c++draft/library#character.seq.general-1.2.sentence-2 but i agree it could be more explicit.

Also, locale properties are not implementation defined, which is interesting http://eel.is/c++draft/intro#compliance.general-9 (but given they need to be documented it's a distinction seemingly without difference)