From: Edward Diener (eddielee_at_[hidden])
Date: 2003-10-31 19:11:10
Bronek Kozicki wrote:
> On Fri, 31 Oct 2003 14:54:32 +0100, Drazen DOTLIC wrote:
>> You are missing the point IMHO. Probably everyone *wants* to have it
>> on, for exactly the reasons you mention above. But, not all of us
>> have the luxury to do so, because we use other libraries besides
>> boost, and compile our programs as a whole (using same set of flags).
>
> I think that small citation from MSVC71 (ie. "1310") Help could be
> relevant here: snipped...
>
> By providing overloads for both the unsigned short and __wchar_t
> variations of wchar_t, you can create libraries that can easily be
> linked with code compiled with or without /Zc:wchar_t and avoid the
> need
> to provide two different builds of the library (one with and one
> without /Zc:wchar_t enabled).
This would mean that Boost wide character implementations would have to
support __wchar_t and unsigned short for all wide character usage in order
to work with VC++ 7+. Providing overloads for all wide characters usage in
this way seems like an enormous PITA for Boost implementors. A better model
from the implementors point of view would be for Boost implementations to
use wchar_t and for Boost library implementors to provide separate versions
in which wchar_t is compiled with /Zc:wchar_t and one in which this switch
is not present, and have slightly different names for each library.