To: J3                                                     J3/26-141
From: Brad Richardson & Generics
Subject: Response to UTI 011
Date: 2026-May-26

Reference: 26-007r1

UTI 011 states:

    "Misplaced constraints?

    The constraint on what may or may not appear in a type-attr-spec
    surely belongs there, not here. Unless you want to explain on why
    it may not appear? On checking, I see that it also appears in
    type-attr-spec. It is unacceptable for it to be specified in both
    places!

    The constraint on what may or may not be a coarray surely belongs
    in type-spec or CODIMENSION attribute, not here. Unless you want
    to explain why? (I might guess that the reason is that there are
    restrictions on what type a coarray may have, so having a coarray
    of deferred type would mean applying those restrictions to the
    instantiation argument, and that is not explicitly spelled out.)

Indeed, C747 has already been edited to prohibit extending deferred
types, so C1616 is redundant and should be removed.

[402:9] Delete constraint C1616.

Also, agree that the constraint on coarrays would be more appropriate
with CODIMENSION.

[123:4] Delete "or " from the sentence, and add the following
        immediately preceding the ".", ", or of a deferred type
        (16.4.1.2)", such that the whole constraint now reads:

        "A coarray shall not be of type C_PTR or type C_FUNPTR from the
        intrinsic module ISO_C_BINDING (19.3.2), of type TEAM_TYPE from
        the intrinsic module ISO_FORTRAN_ENV (17.10.2.44), or of a
        deferred type (16.4.1.2)."

[402:10] Delete constraint C1617.

--END--