Re: [boost] [outcome] Second high level summary of review feedback accepted so far

Subject: Re: [boost] [outcome] Second high level summary of review feedback accepted so far
From: Peter Dimov (lists_at_[hidden])
Date: 2017-05-30 23:19:08

Andrzej Krzemienski wrote:
> > To expand on that a bit:
> >
> >    T* operator->() const noexcept;
> >    Requires: get() != nullptr.
> >    Returns: get().
> >
> > Narrow, a mistake.
> >
>
> You say, you do not like the above, right?  You call this mistake *only* 
> because of this LWG guideline?

For shared_ptr, it is a mistake, period. The function does not require get() 
!= nullptr in order to do its job, so it should not have a precondition. { 
return get(); } is fine.

> >    T* operator->() const noexcept;
> >    Returns: get().
> >
> > Wide, as it should have been.
>
> You say, you like the above, right? But what does it mean? That it returns 
> a `(T*)nullptr` ?

For shared_ptr, it just returns the pointer. For result<>, yes, it would 
return a nullptr when no T is present, not an invalid pointer value.

>  This is the same as your value_if?

Yes, as I wrote,

> > All that said, I'm actually not opposed to outcome<T>::op-> being the 
> > second above (the equivalent of value_if) and op* being effectively 
> > *operator->().