Boost logo

Boost :

Subject: Re: [boost] List of C++ 11 only Boost libraries and their status?
From: M.A. van den Berg (thijs_at_[hidden])
Date: 2014-11-24 15:36:32


On 24 Nov 2014, at 20:03, Pete bartlett <pete_at_[hidden]> wrote:

>
>
>> On 24 Nov 2014, at 18:28, Gottlob Frege <gottlobfrege_at_[hidden]> wrote:
>>
>> On Mon, Nov 24, 2014 at 8:20 AM, Hartmut Kaiser
>> <hartmut.kaiser_at_[hidden]> wrote:
>>>
>>>> For C++ Now 2015 I am intending to present an overview of C++ 11 only
>>>> Boost libraries in the review queue or very near the review queue by
>>>> the time of the conference. I'd appreciate help filling in missing
>>>> gaps in this list:
>>>
>>> *snipped list*
>>>
>>> Technically, all libraries in Boost are C++11. What's your point?
>>
>> C++11 *only*. (or more precisely >= C++11)
>> ie won't work in anything less
>
> I think it *is* worth asking "what's the point". For end users whether a particular library uses c++11 features is more-or-less irrelevant - they have a set of real world compilers to support (or justify upgrade from) and so they would love clear statements about what compilers a particular library works on. Of course it is not trivial for an author to supply such a statement - bits of the library may require more modern c++ than other parts - or the minimum feature set required may be inherited from a library used internally by the implementation and this might also change from release to release...
>
> I believe Stephen Kelly has posted on this list a couple of times about how CMake tries to address these issues.
>
>

If the modularisation of boost takes shape then if would be useful to have C++11 libraries NOT unnecessarily depend on the boost version of C++11 libraries in order to reduce dependencies.

What’s your view on that?

E.g. if I wanted to make a "boost C++11 random library” then that could build on top of the C++11 random lib instead of the boost random lib. This would reduce dependencies and make the library work standalone on a C++11 compiler.
 


Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk