> It is (AFAIK) not one of the datatypes supported by my Standard Library 
> implementation.  That is my point.  I cannot seamlessly use it with the 
> facilities provided by the C++ Standard Library.

Exactly, that's the only real issue that is fixable in the sense that given
a certain set of trade-offs, the current code as delivered is lacking in a
very critical feature to a lot of C++ developers.

So far, nobody with the expertise has offered up a working traits
specialization for unsigned short, and since the hassle dates to 2002 or so
when g++ first dropped support for non-locale-correct basic_strings of
unsigned short (before adding it back for some reason), it seems to be way
more than just non-trivial to do.

When I find that basic_string<unsigned short> really helps me (STL maps or
hashes being an obvious case), I use it, and #ifdef around it to supply a
suboptimal alternative (usually linear) algorithm knowing it won't get used
much on any of the platforms I support (Windows and gcc 3.4+ certainly are
the majority for me).

But I gave up trying to hide XMLCh from my code. That's my native character
type until I hit a boundary I can't avoid.

-- Scott


---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Reply via email to