From:	Martin Str|mberg <ams AT ludd DOT luth DOT se>
Message-Id:	<200004232116.XAA24384@father.ludd.luth.se>
Subject:	Re: EILSEQ in errno.h
In-Reply-To:	<200004232111.RAA20298@indy.delorie.com> from Eli Zaretskii at "Apr 23, 2000 05:11:53 pm"
To:	eliz AT is DOT elta DOT co DOT il (Eli Zaretskii)
Date:	Sun, 23 Apr 2000 23:16:58 +0200 (MET DST)
Cc:	djgpp-workers AT delorie DOT com (DJGPP-WORKERS)
X-Mailer:	ELM [version 2.4ME+ PL54 (25)]
MIME-Version:	1.0
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

According to Eli Zaretskii:
> > Date: Sun, 23 Apr 2000 10:39:43 -0400
> > From: DJ Delorie <dj AT delorie DOT com>
> > 
> > > The C99 standard specifies a third error value for errno.h, EILSEQ.
> > > Should we insert in between EDOM and ERANGE (like it is in the
> > > standard document), and thereby pushing the values of ERANGE and the
> > > rest up one step?
> > 
> > Changing error values would break third party precompiled libraries
> > that return error values.  Unless there's a really compelling reason
> > to change existing values, don't.
> 
> I agree with DJ.  There's no reason for us to change errno values we
> use now, since C99 doesn't require any specific numeric values, not
> does it say that the values should be ordered in any particular way.
> 

Yes, it makes sense, but the errno.h file will be a little confusing:
"...
#define EDOM 	1
#define ERANGE 	2
#define EILSEQ 	39

extern int errno;

#ifndef __STRICT_ANSI__

#define E2BIG 	3
#define EACCES 	4
..."


Right,

							MartinS

- Raw text -