Re: RFD : libg++/gcc/egcs upgrades needed for libc6 (READ ME)
On Tue, Feb 24, 1998 at 04:23:11PM +0100, Brederlow wrote:
> Marcus Brinkmann <Marcus.Brinkmann@ruhr-uni-bochum.de> writes:
>
> > On Sun, Feb 22, 1998 at 03:28:43PM -0700, Galen Hazelwood wrote:
>
> > We will come to a point, where gcc 2.7.2 will not be able to compile all
> > packages (C++ comes to mind. Only recently gcc 2.7 was able to compile
> > libgtkmm, because of substantial changes in libgtkmm. Others may follow).
>
> > I don't know when this will be. Hopefully gcc 2.8 will then be usable.
>
> I think quite a lot of packages will break for some reason or
> another when switching compilers. My experiences with egcs shows that
> the difference in warnings and errors the compiler gives breaks
> several packages.
>
> > Is it a bug in the standard compiler if a package can't be compiled or a bug
> > in the package? If a package can be compiled with egcs (and it follows the
> > standard of that language), but can't be compiled with gcc, what is to fix?
>
> If it compiles without any warnings and warnings enabled under egcs I
> think it will (maybe with some warnings) compile under gcc as well. If
> something gives a warning under egcs and an error under gcc (or other
> way) its a fault of the package and not the compiler. If it compiles
> without warnings on gcc but fails on egcs, then one should recompile
> with warnings on under gcc and see if they are the once that fail on
> egcs (or otherway around).
I hope you refer to gcc 2.8 here, and not 2.7. gcc 2.7 really has problems
that actually lead to wrong or unnecessarily complicated C++ programs. But
surely they will merge together, and then you are right, the code should be
-Wall clean, than we have probably few problems to port it to other
platforms and compilers.
> Packages shouldn't depend on a specific compiler, because that
> decreases portability. The sooner we start to use different compilers
> the sooner we will find those problems.
In general, this is wise. In respect to gcc 2.7, I think we should be
careful here.
Thank you,
Marcus
--
"Rhubarb is no Egyptian god." Debian GNU/Linux finger brinkmd@
Marcus Brinkmann http://www.debian.org master.debian.org
Marcus.Brinkmann@ruhr-uni-bochum.de for public PGP Key
http://homepage.ruhr-uni-bochum.de/Marcus.Brinkmann/ PGP Key ID 36E7CD09
--
TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to
debian-devel-request@lists.debian.org .
Trouble? e-mail to templin@bucknell.edu .
Reply to: