This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: Release Plan Updates
"H. J. Lu" wrote:
>
> On Sun, Jul 14, 2002 at 11:03:01PM -0700, David S. Miller wrote:
> > From: Mark Mitchell <mark@codesourcery.com>
> > Date: Sun, 14 Jul 2002 23:07:33 -0700
> >
> > --On Sunday, July 14, 2002 10:51:39 PM -0700 "David S. Miller"
> > <davem@redhat.com> wrote:
> >
> > > When we fix bugs for the "stable compiler" after #2 happens, will we
> > > be checking them into both the 3.1.x and 3.2.x branches?
> >
> > No; just the 3.2.x branch. The 3.1.x branch will be unmaintained after
> > that point -- the only way there will be a 3.1.2 release is if someone
> > else gets motivated to make another release. For non-C++ users, 3.2.1
> > will be what 3.1.2 would have been; for C++ users it will be that,
> > except with an incompatible ABI.
> >
> > This sounds like nobody is supposed to care about 3.1.x
> > anymore, right? Essentially whoever is using 3.1.x and
> > doesn't want the c++ ABI change, well they won't get bug
> > fix releases anymore past 3.1.1. Do I understand this
> > correctly?
>
> That is what I understood. If I am right, only Apple wants 3.1.x and
> is ok with its current status. Personall, I believe Apply should
> switch to new gcc 3.2. But they have no plan to do so. To me, that
> implies they have no major problems with C++ in 3.1.x.
The compiler for Jaguar is basically frozen at this point, and
we're not going to hold up the release (and the associated revenue,
ahem) just for the reformulated 3.2. I still have to think about
how this affects our plans, but my instant reaction is that we'll be
skipping 3.2 and going to 3.3 next, with a backwards-compat flag if
that proves to be necessary - compatibility is important for drivers
for instance, but they use a subset of C++ that may not be affected
by the ABI changes, we'll need to analyze them.
Stan