GB-137 Index of library names: Exposition-only private members

[wg21bot]

Are all exposition-only private members needed in the index of library names?

The index of library names includes move_only_function::is-callable-from, an exposition-only variable template that is only used in the immediately following subclause. Does that really need to be indexed? There should at least be some introductory text for the index explaining that italic names are for exposition purposes only, not part of the library API.

[jensmaurer]

See also cplusplus/draft#5894

[tkoeppe]

Editorial meeting decision: accept with modifications: move those exposition-only names out of the index of library names and into the main index, which fixes the immediate issue. We should also add front matter to the general index to explain exposition-only names, and we should add the same explanation to the index of concept names.

[tkoeppe]

What about index entries such as "operator* -- repeat_view::iterator"? That's a non-exposition-only member of an exposition-only (nested) type. Should it remain in the library index? If so, the library index will need explanatory wording even if we move the top-level names out.

[tkoeppe]

I'm leaving everything in the library index that's not a private data member.

[tkoeppe]

Final disposition: Accepted.