[boost] Re: Re: Re: BOOST_STD_EXTENSION_NAMESPACE and dinkumwarestdlib313

From: Jonathan Turkanis (technews_at_[hidden])
Date: 2004-01-26 19:44:50

"Reece Dunn" <msclrhd_at_[hidden]> wrote in message
news:BAY7-F28VkhNPmV6ZD80000662b_at_hotmail.com...
> Jonathan Turkanis wrote:
> >"Reece Dunn" <msclrhd_at_[hidden]> wrote in message:
> > > >If Dinkumware has made this change uniformly in its various
> > > >distributions, it sounds like BOOST_STD_EXTENSION_NAMESPACE
should
> >be
> > > >defined to be stdext in boost/config/stdlib/dinkumware.hpp if
> > > >_CPPLIB_VER is a certain value or higher (313?).
> > >
> > > I have done a little digging:
> > >
> > >
>
>http://msdn.microsoft.com/library/default.asp?url=/library/en-us/vcst
dlib/html/vclrfhashmap_class.asp
> > > says:
> > >
> > >    "In Visual C++ .NET 2003, members of the <hash_map> and
> > >    <hash_set> header files are no longer in the std namespace,
> > >    but rather have been moved into the stdext namespace."
> > >
> > > Also, the online Dinkumware reference
> > >
(http://www.dinkumware.com/manuals/reader.aspx?b=p/&h=hash_map.html)
> >places
> > > hash_xxx in the std namespace. Therefore, it is specific to
VC7.1
> >(unless
> > > Intel has done something similar?).
> >
> >That's interesting. Since the compiler is determined before the
> >standard library, it would still be possible to put the definition
in
> >boost/config/stdlib/dinkumware.hpp, like so
> >
> >     #if defined(BOOST_MSVC) && (BOOST_MSVC >= 1310)
> >         #define BOOST_STD_EXTENSION_NAMESPACE stdext
> >     #endif
> >
> >Maybe one should check _CPPLIB_VER for good measure. This sounds
> >reasonable. What do you think?
>
> I agree. The question is, should the test be BOOST_MSVC >= 1310 or
> BOOST_MSVC == 1310, likewise for the _CPPLIB_VER test. I say this
because a
> future version of VC may move hash_xxx back into the std namespace
for
> future standard library conformance (does TC1 have hash_map, etc.?).
>

TR1 has 'unordered_map', etc. These are similar but not identical to
the Dinkumware versions. My guess is that Dinkumware for VC will keep
hash_map for a while, presumably in stdext, even after adding
<unordered_map>, etc. So I think >= is appropriate.

BTW, for the time being, unordered_map and its companions are in the
namespace std::tr1.

Jonathan