This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: wstring commented out in egcs-1.1.2 C++ library
- To: Benjamin Kosnik <bkoz at cygnus dot com>
- Subject: Re: wstring commented out in egcs-1.1.2 C++ library
- From: Jeffrey A Law <law at cygnus dot com>
- Date: Tue, 08 Feb 2000 17:14:57 -0700
- cc: Mike Stump <mrs at windriver dot com>, gcc at gcc dot gnu dot org, libstdc++ at sourceware dot cygnus dot com
- Reply-To: law at cygnus dot com
In message <Pine.SOL.3.91.1000208153321.1276A@cse.cygnus.com>you write:
> > Bzzt. Wrong.
> >
> > Backporting the libstdc++ wstring stuff is trivial.
>
> Bzzt. Wrong. Besides, iostreams in v2 cannot handle wchar's, whereas v3
> does. As others have noted on this thread, then you can use wstringstream.
:-) I stand corrected then.
> > The problem is it relies on a lot of code which doesn't exist on a number
> of
> > popular unix systems. There was a tread on this issue a long time ago.
> > Basically someone needs to write the necessary support code for libiberty
> > that can be included when the host system doesn't provide proper wide cha
> r
> > support.
>
> ... this stuff has been done in v-3, in an nice autoconf'd and automake'd
> way. I am unwilling to port this to v2 -- I think it's easier just to use
> v3. Instead, I'm making v3 play nice as a plug-in replacement to the
> current gcc distros. . .
Very very cool. I wish I had more time to at least follow what's happening
in the V3 world. Sigh.
> As I see it, the only reasons to use v2 over v3 are
>
> 1) v2 MT string stuff has not been fully integrated
> 2) xcompiles dont' work.
>
> within two weeks both of these should be fixed.
Even better.
jeff