Re: __cplusplus revisited (on Solaris 8)

[Benjamin Kosnik <bkoz at redhat dot com>]

> That means just passing the contents of these headers through as-is will introduce name conflicts and defeat the
> entire purpose of putting the <c..> headers in the std namespace.  The only way they would work is if the C headers
> put everything in namespace std and only there.

got it thanks.

> I think the include/c headers should be modified to filter out everything not already in the std namespace.  That
> way, they should work under the Solaris 8 circumstances and continue to provide the same functionality they currently
> do.  This would have the effect of rolling "option 4" into "option 1".

Agreed. At least we know the approach now...

> I can certainly work this up.  Unfortunately I seem to have discovered a possible bug in the compiler that seems to
> prevent certain key header files from being wrapped in a namespace, so any results will have to wait until that issue
> is resolved.

Great. Namespace issues are pretty much priority one, so if you are 
running into compiler issues WRT to namespaces both Jason and Martin have 
been very helpful, so don't hesitate to cc them.

thanks,
benjamin