Re: C++ defect reports: how to behave

[Matt Austern <austern at apple dot com>]

On Thursday, June 26, 2003, at 12:46  PM, Giovanni Bajo wrote:

> Joe Buck <jbuck@synopsys.com> wrote:
>
> > > > IMO, we should implement the DR's without needing a separate switch
> > > > for them.
> > >
> > > Just to be sure, you're suggesting to implement also all the defects 
> > > with
> > > "WP" status without a different dialect? Even if they modify the 
> > > semantic
> > > of existing programs?
> >
> > Issues with "WP" status are for the next version of the standard, so 
> > if
> > they conflict with the current standard, they form a different 
> > dialect.
> > But from my scanning of the list of "WP" issues, most are simply
> > clarifications or corrections, some considered not important enough 
> > to go
> > in now, so implementing them does not form a different dialect.
>
> I don't mind either way. If there is consensus about implementing both 
> TC1
> and WP defect reports into -std=c++98, I'll prepare a little patch for
> standards.texi specifying this. I would like to hear a comment from a 
> C++
> maintainer though.

I'm not a C++ maintainer, but I do attend standards committee meetings
regularly.  My advice is to incorporate TC1 defect reports only.  
They're
stable.  Issues in 'DR' and 'WP' status aren't necessarily stable: there
have been at least a few occasions when an issue has gone into 'DR' 
status
and then been changed.

			--Matt