Proposed resolution:

This wording is relative to N5014.

1. Modify [iterator.synopsis], header <iterator> synopsis, as indicated:

   […]
   namespace std: {
     […]
     // [range.iter.op.distance], ranges::distance
    
     template<class T>
       using distance-iterator-t =               // exposition only
         conditional_t<is_array_v<remove_reference_t<T>>,
           decay_t<T>, remove_const_t<remove_reference_t<T>>>;
     
     template<class I, sentinel_for<I> S>
       requires (!sized_sentinel_for<S, I>)
       constexpr iter_difference_t<I> distance(I first, S last); // freestanding
     template<class I, sized_sentinel_for<decay_tdistance-iterator-t<I>> S>
       constexpr iter_difference_t<decay_tdistance-iterator-t<I>> distance(I&& first, S last); // freestanding
     template<range R>
       constexpr range_difference_t<R> distance(R&& r); // freestanding
     […]
   }
   

2. Modify [range.iter.op.distance] as indicated:

   template<class I, sized_sentinel_for<decay_tdistance-iterator-t<I>> S>
     constexpr iter_difference_t<decay_tdistance-iterator-t<I>> distance(I&& first, S last);
   

   -3- Effects: Equivalent to:

   if constexpr (!is_array_v<remove_reference_t<I>>)
     return last - first;
   else
     return last - static_cast<decay_t<I>>(first);
   

[ 2025-08-29; Reflector poll ]

Set priority to 4 after reflector poll.

This is discovered when implementing the resolution LWG 4242. Per LWG 4242, it is intended to allow `ranges::distance` to handle `volatile`-qualified iterator values. However, the uses of `decay_t` (established per LWG 3664) are still problematic, because when sized_sentinel_for<S, decay_t<I>> is modeled, there's no semantic or syntactic requirement that `S` shall work with volatile-qualified `I`.

If we implement the constraint as is, there will still be some underconstrained cases. E.g. When the `operator==` or `operator-` intendedly rejects volatile-qualified iterators. And even when they accept volatile-qualified iterators, the additional semantic requirements imposed by sized_sentinel_for<S, decay_t<I>> are still undesired.

I think we should only decay arrays and keep `volatile` for non-array arguments.