C++ PATCH: PR 20599 (1/3)

[Gabriel Dos Reis]

Joe Buck <Joe.Buck@synopsys.COM> writes:

| On Mon, Oct 02, 2006 at 10:14:06AM +0200, Gennaro Prota wrote:
| > On 02 Oct 2006 05:51:16 +0200, Gabriel Dos Reis
| > <gdr@integrable-solutions.net> wrote:
| > 
| > >Mark Mitchell <mark@codesourcery.com> writes:
| > >
| > >| Gabriel Dos Reis wrote:
| > >| 
| > >| > I would like to see a flag -std=c++03, in addition to -std=c++0x.
| > >| > Other than, that I suspect your proposed plan is a good one, as long
| > >| > as we all understand its implications.
| > >| 
| > >| To be clear, I assume you want -std=c++03 to be the amended version of
| > >| the standard released in 2003?
| > >
| > >Yes, that is right.
| > 
| > Shouldn't that be the default? FWIW, I've never seen a "C95" switch in
| > any C compiler, though they have usually a C99 one (which I'm not sure
| > I like either).
| 
| Unless I'm missing something, I think it should be the default.  My
| understanding is that "c++03" is just a bug-fix to the standard,
| clarifying things that were not clear or mis-stated in the original.
| I don't see why there should be support for two distinct versions of
| the C++ standard.

And in effect, there are two distinct versions of the C++ standard.

This has implications on the value of __cplusplus and the set of
displayed behaviour.  C++03 some semantics changes -- even when
they are corrections.

-- Gaby