Re: [boost] Re: fix: intel.hpp

From: David Abrahams (dave_at_[hidden])
Date: 2003-04-25 10:25:15

Gennaro Prota <gennaro_prota_at_[hidden]> writes:

> On Fri, 25 Apr 2003 10:16:09 -0400, David Abrahams
> <dave_at_[hidden]> wrote:
>
>>Applied, thanks!
>
> Wait a moment. Have you asked John about this? 

Nope.

> That's not to doubt of
> you, of course

No offense taken; thanks for looking over my shoulder.

> , but the issue is quite tricky. _NATIVE_WCHAR_T_DEFINED is a VC7
> undocumented macro, defined when you use /Zc:wchar_t:
>
> http://lists.boost.org/MailArchives/boost/msg16494.php
>
> I can't verify this but I think that with VC7 it is different from
> _WCHAR_T_DEFINED (which is defined even if wchar_t is a typedef).
> Intel probably follows the same pattern. We went through this matter a
> while ago:
>
> http://lists.boost.org/MailArchives/boost/msg32639.php
>
> And, if I didn't say it before, the issue is quite tricky :-)
>
> http://lists.boost.org/MailArchives/boost/msg32483.php

OK, what if it were changed to:

#  if !(defined(_WCHAR_T_DEFINED) || defined(_NATIVE_WCHAR_T_DEFINED))
#     define BOOST_NO_INTRINSIC_WCHAR_T
#  endif

??

-- 
Dave Abrahams
Boost Consulting
www.boost-consulting.com