Message-Id: <200006051915.PAA04340@qnx.com>
Subject: Re: ANSI C and stdio.h
To: djgpp-workers AT delorie DOT com
Date: Mon, 5 Jun 2000 15:15:37 -0400 (EDT)
From: "Alain Magloire" <alain AT qnx DOT com>
In-Reply-To: <200006051822.OAA29585@envy.delorie.com> from "DJ Delorie" at Jun 05, 2000 02:22:23 PM
X-Mailer: ELM [version 2.5 PL0b1]
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Reply-To: djgpp-workers AT delorie DOT com
Errors-To: nobody AT delorie DOT com
X-Mailing-List: djgpp-workers AT delorie DOT com
X-Unsubscribes-To: listserv AT delorie DOT com
Precedence: bulk

> 
> 
> > Problems like this do not appear in simple "hello world" type of 
> > programs, but rather creep in  when you are doing advanced or large or
> > medium or big projects with interdepencies etc ... when you
> > move the enviroment around, this is where you hit the brick wall.
> 
> I can't think of a single program that includes stdio.h but not
> stdarg.h (or vararg.h), and then tries to use "va_list" as something
> other than the "usual" meaning.  That's the kind of program our
> current headers would cause problems with.  If anyone *does* code like
> that, they're asking for trouble even though the program is
> technically 100% compliant.

Maybe it's me who got this backward or I've missed a couple of threads.
To recap what I've understood in this thread, ANSI C says that
<stdarg.h> should not be part of <stdio.h> because it polluates
the namespace etc ... (I do not have the rationale).  Most OS/platforms
go around this, for functions like vfprintf(..) by including their
own header .i.e solaris <sys/va_list.h> or define va_list Watcomm etc ..
Some other just includes <stdarg.h>  DJGPP, GNU/Linux/Glibc (?)

And on those platforms if people takes things for granted
code like this :

#include <stdio.h> 
void warnings (char *s, ...) 
{ 
    char *t; 
    va_list ap; 
 
    va_start(ap, s); 
    while (t = va_arg(ap, char *))
    { 
	/* do stuff */
    } 
    va_end(ap); 
}

Will work compile under DJGPP, GNU/Linux, since <stdio.h> include <stdarg.h>
but fail to compile under for example Solaris.

This is probably picknitting, because I doubt a portable application will call
va_arg() without including <stdarg.h> but one may get use to just include
<stdio.h>

...

> Compliant programs (except pedantic examples like mine above) work
> just fine.  My point is that some common broken programs should work
> also.  I'd rather be lenient to the clueless user if I can do that
> without causing grief for other users.
> 
> Thus, there are two cases:
> 
> 1. Broken programs that work under DJGPP.
> 
> 2. Valid programs that break under DJGPP.
> 
> I'd like to maximize the first while minimizing the second.


Yes, fine goals.  I just hope to make (1) work, you don't
deviate too much from the std or carry a lot of baggages.

BTW, I'm talking in general and  not targetting stdio.h/stdarg.h issue
in particular, it just seems weird that to minimize FAQ one would prefer
to deviate from the std.  In my env.  if I go this way Q/A would have
my head on a platter the next day ;-)

Later.

-- 
au revoir, alain
----
Aussi haut que l'on soit assis, on n'est toujours assis que sur son cul !!!