Re: address@hidden: Re: CVS commit: pkgsrc/devel/m4]

[Bruno Haible]

Joerg Sonnenberger <address@hidden> wrote:
> > They want to force replacement functions
> > for fseeko down our throat when configure detected that fflush for "r"
> > streams doesn't work. This breaks obviously on DragonFly, but I don't
> > think this test should be allowed at all. *sigh* I hate gnulib.

This is a very unqualified statement. Nobody wants to force particular
functions' behaviour "down your throat" (= into NetBSD's libc).

We respect NetBSD's kernel and libc. We are not forcing a behaviour on you.
Microsoft's Windows developers have decided to spit on ISO C 99 and POSIX.
NetBSD's developers can do the same. When we occasionally report POSIX
compliance problems to OS developers, it is on the assumption that it is
welcome.

In return, we expect that you respect the way that we build our programs.
The GNU m4 developers have decided to use a certain source structure,
based upon "modules" and "replacement functions". This is our business,
not yours.

If you find that GNU m4 compiled on GNU/kNetBSD is smaller than compiled on
NetBSD, that is because glibc needs less replacement functions. But it is
not a real problem, since correct functioning of a program is more important
than an executable's size.

So, please, respect our freedom to organize the source of GNU programs the
way we like.

Bruno