So, about std::swap and std::exchange:

They can use the pure attribute with parameter:
template <typename T>
[[pure(is_pure_move_constructible_v<T> && is_pure_move_assignable_v<T>)]]
void swap(T& a, T& b) noexcept(is_nothrow_move_constructible_v<T> &&
is_nothrow_move_assignable_v<T>) { /* ... */ }
Analogue for std::exchange

Tymi.

On Wed, 5 Mar 2025, 07:04 Tymi, <tymi.cpp_at_[hidden]> wrote:

> Okay, just to be clear, this is a proposal about C++. I don't want to mix
> C into it, as well, it is not a C proposal. Argument about C is not very in
> place for it.
> C unsequenced and reproducible are a C thing, I want a pure attribute in
> C++ that is distinct from C's unsequenced and reproducible.
>
> Note that I never mentioned C in my original post for a reason.
>
> Tymi. (Accidentally sent on wrong mail)
>
> On Wed, 5 Mar 2025, 05:53 Tymi, <tymi.cpp_at_[hidden]> wrote:
>
>> Okay, just to be clear, this is a proposal about C++. I don't want to mix
>> C into it, as well, it is not a C proposal. Argument about C is not very in
>> place for it.
>> C unsequenced and reproducible are a C thing, I want a pure attribute in
>> C++ that is distinct from C's unsequenced and reproducible.
>>
>> Note that I never mentioned C in my original post for a reason.
>>
>> Tymi.
>>
>> On Wed, 5 Mar 2025, 05:29 Jan Schultke, <janschultke_at_[hidden]>
>> wrote:
>>
>>> > Similarly, std::swap for an array of ints is reproducible, but
>>> std::swap isn't.
>>>
>>> I meant std::sort for an array of ints is reproducible.
>>>
>>> Sorry, just woke up, brain not fully booted yet.
>>>
>>