Re: Opinions on introduction of string_view in Core

From: Andrey Semashev (andrey.semashev_at_[hidden])
Date: 2021-10-18 00:47:25

On 10/18/21 3:43 AM, Vinnie Falco via Boost wrote:
> On Sun, Oct 17, 2021 at 5:33 PM Gavin Lambert via Boost
> <boost_at_[hidden]> wrote:
>> Otherwise, introducing as a separate library and letting the community
>> decide what is better (with review and then with usage) seems like the
>> way to go.  Preemptively putting it in Boost.Core feels like bypassing
>> community choice, which goes against the spirit of Boost.
> 
> Marshall doesn't want to change it, and if it fails review then users
> of my libraries will continue to encounter headwinds when using
> C++17's std::string_view. I really don't understand what is so
> difficult about this. An increasingly common complaint of Boost is
> that it doesn't play well with the std types that Boost so generously
> helped design.
> 
> Please explain to me what users of Beast, Boost.JSON, eventually
> Boost.URL should do if they want to use C++17 string_view, Marshall
> decides against implementing the change, and there is no consensus to
> accept Boost.StringView2?

Making your libraries explicitly support std::string_view, in addition
to boost::string_view, is one option. I realize this is more work for
you as the maintainer, but this is a valid solution nonetheless.