Skip site navigation (1)Skip section navigation (2)
Date:      8 Jan 1997 02:20:36 GMT
From:      peter@spinner.DIALix.COM (Peter Wemm)
To:        freebsd-hackers@freebsd.org
Subject:   Re: #include file xref philosophy
Message-ID:  <5av09k$67a$1@haywire.DIALix.COM>
References:  <199701072241.JAA17826@profane.iq.org>

next in thread | previous in thread | raw e-mail | index | archive | help
In article <Mutt.19970108011911.j@uriah.heep.sax.de>,
	j@uriah.heep.sax.de (J Wunsch) writes:
> As Julian Assange wrote:
> 
>> What is the -current philosophy on intra #include file dependencies?
>> 
>> Is there any reason for not following the posix line and having
>> include files resolve all their own dependencies? I'm talking about
>> things like <dirent.h> needing the code that includes it, include
>> <sys/types.h> before hand.
> 
> IIRC, Bruce once mentioned that Posix allows it to require the
> inclusion of <sys/types.h> (xor another <sys/something.h> which i
> eventually forgot) before other things.
> 
> Either way, the corresponding man pages should at least mention the
> required include files, so your complaint about ``magical knowledge''
> would not be true.

As I understand it, POSIX allows for a system to require <sys/types.h> to
be a prerequisite for other include files, Bruce knows more about more than
I do.

However, the SVIDIII, X/Open XPG (4+), Spec 1170, Single-Unix spec, and the
whatever-it-is-called-this-month line of standards apparently require the
standard headers to be self sufficient.  I believe NetBSD/OpenBSD/Linux
comply with this, I remember NetBSD committing the missing <sys/types.h>
includes some time back.

Cheers,
-Peter



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?5av09k$67a$1>