Re: [boost] [review] Review of Outcome (starts Fri-19-May)

Subject: Re: [boost] [review] Review of Outcome (starts Fri-19-May)
From: Andrzej Krzemienski (akrzemi1_at_[hidden])
Date: 2017-05-19 13:48:25

2017-05-19 15:08 GMT+02:00 Niall Douglas via Boost <boost_at_[hidden]>:

> >> The constexpr variables are already there. So basically do we prefer:
> >>
> >> 1. .ensure_empty(), .ensure_value(), .ensure_error() and
> >> .ensure_exception()
> >>
> >> 2. Or .ensure(empty), .ensure(value), .ensure(error) and
> .ensure(exception)
> >>
> >> I figure the latter looked nicer. It's same difference to the compiler,
> >> simple overload matching is constant time.
> >>
> > Just one note. If I use namspace prefixes, the notation with "constants"
> > becomes longer:
> >
> > `o.ensure_empty()` becomes `o.ensure(boost::outcome::empty)`
> >
> > One could respond to this "just import anything from namespace
> > `boost::outcome` into the scope", but that is imposing on me a certain
> > style of programming, which I not necessarily want to adapt.
>
> Yes that's a good point.
>
> Also, after a few nights of sleeping on it, I'm not keen on .ensure_XXX().
>
> Would people be okay with:
>
> * o.check() <= (void) o.value()
>
> * o.check_error() <= (void) o.error()
>
> * o.check_exception() <= (void) o.exception()
>

Given that this "ensure" functionality would be used rarely, and some even
suggest there is no use case for it, I would be satisfied with this syntax
`(void)o.value()`.