what are "view"s?

[burblebee]

what are these "view"s? libconcept? code? undefined term?

"view" is used as a term but never defined - what is it? There are quite a few additions of the term in the wording added in P2321R2 from branch motions-2021-10-lwg-16. If it is a term, we should define it. If we're talking about the concept, it should be marked with \libconcept. There is also a function "view()" in "basic_stringbuf" (in which case it should be in \tcode), but I doubt we're talking about that.

Originally posted by @burblebee in #5024 (comment)

[CaseyCarter]

There's a long-standing library convention of using a concept name to denote either types that model the concept or instances of objects with those types, with the distinction hopefully being clear from context. "Iterators" is the classic example. We've enshrined it as the name of a library.

I drew attention to this fact, and the fact that Ranges was extending the practice to "ranges" and "views" during LWG reviews, but folks were happy to continue the colloquial usage without any kind of formal definition. (Mostly people laughed at me for using the word "colloquial".)

TLDR: "view" refers to either a type that models the view concept or an instance of such a type.

[jwakely]

I don't see a problem here. We also refer to "strings" to mean specialisations of basic_string, and I don't think anybody is confused by that.

The view concept defines what a view is. If it models the concept, it's a view. It would be wrong to replace all uses with code font because that would refer to the concept itself, not things that model it.

[jensmaurer]

Given these opinions of LWG participants, I think there's nothing we should do here editorially. @burblebee , if you still feel there is a defect, I'd suggest to file an LWG issue with intensified rationale.