I think I just misunderstood what you were saying, I thought you said 2.96
was in there. I would agree with that stance if any distro pulled what redhat
did especially now that gcc's stance is known I'd have to go debian or
something. I don't think mandrake would make that move though.
I've been wondering - where is this going to leave redhat or any other distro
that chooses/chose 2.96. 2.96 may build object code that isn't compatible
with earlier or later gcc's - doesn't that include libraries?
So now when redhat goes to 7.1 with gcc 3 or whatever will that not also mean
they will have to replace all the libs that were made with 2.96 and c++? And
in replacing all the libs wouldn't that break all the compatbility of apps
and the libs they use. So basically 7.1 will bring the need to do pretty much
a total re-install.
Just a thought
On Sun, 08 Oct 2000, you wrote:
> Sorry, have I used the wrong thread ? There was a discussion as to why
> Mandrake should hold off going to gcc-2.96
>
> Owen
>
> On Sun, 08 Oct 2000, you wrote:
> > At the risk of sounding stupid - what are you talking about?
> > gcc -version gives me
> > 2.95.3
> >
> > On Sat, 07 Oct 2000, you wrote:
> > > Moreover, I've had real problems getting server side C code to run, let
> > > alone C++ code !! with 2.96. In order to get existing code to work (not
> > > even new code !) for web page rendering I had to go back to Mandrake
> > > 2.95. I've been dreading Mandrake going 'up' to 2.96. I've had a bug
> > > report to this effect lodged with Red Hat for some time now, they
> > > currently want to examine the code !!! well, this is open source ;-)
> > >
> > > Owen
> > >
> > > On Sat, 07 Oct 2000, you wrote:
> > > > During the bombing raid of Sat, 7 Oct 2000 13:26:13 +0200, somebody
> > > > heard
> > > >
> > > > Jan Niehusmann mumble in fear:
> > > > > On Tue, Oct 03, 2000 at 05:34:35PM +0800, Geoffrey Lee wrote:
> > > > > > hackgcc is only in contribs. Once upon a time, Chmouel put it in
> > > > > > for main but it was too unstable so we reverted to 2.95.
> > > > >
> > > > > While I prefer 2.95, too, I wonder if this has consequences on
> > > > > binary compatibility. As far as I know, C code is not a problem,
> > > > > but C++ code may be incompatible between different compiler
> > > > > releases.
> > > > >
> > > > > So, a binary released for redhat 7 may not work on mandrake, if it
> > > > > contains C++ code and is linked against shared C++ libraries.
> > > > >
> > > > > Does anyone know if this problem really exists or if 2.95 and 2.96
> > > > > are compatible in this regard?
> > > >
> > > > The official word from the GCC group is...RedHat messed up by
> > > > using 2.96. There's no binary compatibility with 2.95 *and* there's
> > > > no binary compatibility with the following releases
> > > > either...so..RH7's stuff doesn't work for nobody but RH7 (unless it's
> > > > pure C or Fortran [I'm not sure on this last one...memory is foggy so
> > > > late at night]) What really worries me about any thoughts of mdk
> > > > going to 2.96 is the *no forward compatibility" part...please, let's
> > > > stick to 2.95 and wait for official 3.0 :)
> > > >
> > > > BTW, the info comes from
> > > > http://gcc.gnu.org/ml/gcc-announce/2000/msg00003.html which was
> > > > linked to from /. today/yesterday (friday eve)
> > > >
> > > > Arioshi ba
> > > >
> > > > Vox, who has scratched RH from his it-may-be-used list at work
> > > > and home.
> > > >
> > > > --
> > > > Pain is the gift of the gods, and I'm the one they chose as their
> > > > messenger...
> > > > For info about safety in BDSM, visit Vox's Info Center at
> > > > http://www.the-vox.com/
> > > >
> > > > Think of the Linux community as a niche economy isolated by its
> > > > beliefs. Kind of like the Amish, except that our religion requires us
> > > > to use _higher_ technology than everyone else.
> > > > -- Donald B. Marti Jr.
> > > >
> > > > "Happiness is having a large, loving, caring, close-knit family in
> > > > another city." -- George Burns