On Wed, Jul 22, 2020 at 6:34 AM Gašper Ažman via Std-Proposals <
std-proposals_at_[hidden]> wrote:

>
> The *new* version of the wording makes the wording far more clear (could
> have sworn we sent it in a mailing, but I guess we'll have to send it
> again):
>
> https://htmlpreview.github.io/?https://raw.githubusercontent.com/BRevzin/cpp_proposals/master/0847_deducing_this/d0847r5.html
> is the status quo of the paper. There are no design changes or semantic
> woding changes, just Jens', Barry's and my wording refactors.
>
> The relevant bits:
>
> 5b A function parameter declared with an
> *explicit-this-parameter-declaration* is an *explicit this parameter*. An
> explicit this parameter shall not be a function parameter pack
> ([temp.variadic]). An *object member function* is either a non-static
> member function or a static member function with an explicit this parameter.
>
> Change 12.6.4 [over.sub] <https://wg21.link/over.sub>/1:
>
> 1 A *subscripting operator function* is a function named operator[] that
> is a non-static an object member function with exactly one ordinary member
> parameter.
>
> See, now *this* I would call a separable bit of the proposal! :)
Do you know if there are any proposals in-flight, or any proposals recently
rejected, to get rid of the asymmetries around some of the operators and
permit e.g. operator[] and operator() to be free functions?

struct S {
    // Current: operator[] cannot be a non-member function
    friend Element& operator[](S& lhs, int rhs) { return lhs.data_[rhs]; }
};

Or even to permit operators to be static member functions?

struct Comparator {
    // Current: operator() cannot be a static member function
    static bool operator()(int x, int y) const { return x < y; }
};

–Arthur