Boost logo

Boost :

Subject: Re: [boost] Changes to VS 2012 config
From: Beman Dawes (bdawes_at_[hidden])
Date: 2013-01-14 15:58:44


On Wed, Nov 28, 2012 at 3:59 PM, Beman Dawes <bdawes_at_[hidden]> wrote:
> On Wed, Nov 28, 2012 at 1:36 PM, Eric Niebler <eric_at_[hidden]> wrote:
>> On 11/28/2012 8:44 AM, Marshall Clow wrote:
>>> Just so people know - I'm updating the Visual Studio configuration to
>>> more accurately reflect the capabilities of VS 2012.
>>>
>>> The first step was in http://svn.boost.org/trac/boost/changeset/81613
>>> ; the next step will be to update the compiler features.
>>>
>>> If you see any problems, please let me know.
>>
>> This is great Marshall. Thanks for doing this. Will there be any effort
>> made to support the rolling CTP releases of the MS compiler? The latest
>> CTP has variadic templates, for instance (although they're a little buggy).
>
> Historically we didn't add config support until a compiler was
> actually released. But if Microsoft is going to be adding features
> just via rolling CTP releases, with no formal release for several
> years, I'd like to see us add the CTP features based on _MSC_FULL_VER.
> Note that the CTP released didn't bump _MSC_VER.
>
> > ... Many messages later...

Summary: there was a lot of unhappiness with the proposal to treat the
CTP features as enabled since these features are not yet supported and
are known to have a relatively high number of bugs. OTOH, some folks
do want to use the shinny new C++ CTP features, if only to test.

Someone (sorry, I can't find the message) suggested that we do report
the features as present, but only when _MSC_FULL_VER is greater or
equal the CTP full version number &&
defined(BOOST_ENABLE_MSVC_2012_NOV_CTP). That will allow those of us
who do want the feature macros to be useable to do so, without getting
in other folks way.

If there aren't strong objections, I'd like go ahead with that
proposal. The actual name of the macro is subject to change, of
course, if someone has a better name.

IIRC, the feature macros involved are:

BOOST_NO_CXX11_EXPLICIT_CONVERSION_OPERATORS
BOOST_NO_CXX11_FUNCTION_TEMPLATE_DEFAULT_ARGS
BOOST_NO_CXX11_HDR_INITIALIZER_LIST
BOOST_NO_CXX11_RAW_LITERALS
BOOST_NO_CXX11_UNIFIED_INITIALIZATION_SYNTAX
BOOST_NO_CXX11_VARIADIC_TEMPLATES

Comments?

--Beman


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