Re: diffutils 2.8.1 ISO C90 compliance patch

[Paul Edwards]

From: "Paul Eggert" <address@hidden>
> > You can say the same about any other non-standard file.
> 
> No, <sys/types.h> is an ISO-standard file, as I already explained.
> You may not like the choice of standard, which is your privilege; but
> it's misleading to claim that <sys/types.h> is nonstandard.

It's not standard C, it's standard Unix.

> > There is no reason to make text processing programs require Posix.
> 
> Sure there is, in many cases, one of them being diff3.  On POSIX-like
> systems, diff3 is far more useful written the way that it is, than it
> would be if it were written the way you suggested, where the user
> would have to awkwardly plug in the output of two diffs by hand.

Yes, there's nothing wrong with providing POSIX extensions,
but the functionality should be there on any C90 environment.

> > It has not been written to cater for non-Posix systems.
> 
> Catering to such systems is not a goal of the GNU project.  See
> <http://www.gnu.org/prep/standards/html_node/System-Portability.html#System-Portability>,
> which says (among other things) "As for systems that are not like
> Unix, such as MSDOS, Windows, VMS, MVS, and older Macintosh systems,
> supporting them is often a lot of work. When that is the case, it is
> better to spend your time adding features that will be useful on GNU
> and GNU/Linux, rather than on supporting other incompatible systems."

In this case the work has already been done.

But yes, I agree that the GNU's goals and mine differ, as I
want to support any environment with a C90 compiler, not
just Unix systems.  It is C90 that is the universal language,
not Unix/Posix.

BFN.  Paul.