pkgsrc-Changes archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]

Re: CVS commit: pkgsrc/textproc



On Fri, May 10, 2013 at 08:57:56PM +0900, OBATA Akio wrote:
> On Fri, 10 May 2013 20:41:23 +0900, Joerg Sonnenberger 
> <joerg%britannica.bec.de@localhost> wrote:
> 
> >On Fri, May 10, 2013 at 08:03:01PM +0900, OBATA Akio wrote:
> >>On Fri, 10 May 2013 19:37:03 +0900, Joerg Sonnenberger 
> >><joerg%britannica.bec.de@localhost> wrote:
> >>
> >>>On Fri, May 10, 2013 at 09:35:24AM +0900, OBATA Akio wrote:
> >>>>On Fri, 10 May 2013 08:59:18 +0900, Joerg Sonnenberger 
> >>>><joerg%britannica.bec.de@localhost> wrote:
> >>>>>>Have you tested for runtime, not just for build with your compiler?
> >>>>>
> >>>>>It is linking with the C++ compiler, so it will get the correct
> >>>>>libraries. That did not happen before when using anything but gcc.
> >>>>
> >>>>So what are you doing?
> >>>>Just break to build on environment that require to link with
> >>>>libstdc++ exactly?
> >>>
> >>>Which environment is that? As I said, the C++ compiler does that. The
> >>>behavior is plainly broken for any compiler than G++ or Clang with
> >>>libstdc++ as backend.
> >>
> >>Do you know whole development environments?
> >>Not only recent GCC, Clang, C++11 spec, and so on.
> >>It is there, then it means someone require it.
> >
> >It is a noop for anyone using G++, so it is not visible for them.
> >
> >>In other words, do you know which environment is broken with
> >>original behavior?
> >
> >Any environment that doesn't have libstdc++?
> 
> Thanks a lot.
> I see that your patches are not required by anyone.
> I'll revert them when I'll update such packages,
> because just lead to maintain such packages harder.

Are you even reading what I am saying? *Nothing* should hardcode
-lstdc++. It is wrong. No excuses. Link with ${CXX} and it will work
with every single C++ compiler.

Joerg



Home | Main Index | Thread Index | Old Index