From owner-freebsd-commit  Sat Apr  1 20:35:37 1995
Return-Path: commit-owner
Received: (from majordom@localhost) by freefall.cdrom.com (8.6.10/8.6.6) id UAA22429 for commit-outgoing; Sat, 1 Apr 1995 20:35:37 -0800
Received: (from majordom@localhost) by freefall.cdrom.com (8.6.10/8.6.6) id UAA22416 for cvs-usrsbin-outgoing; Sat, 1 Apr 1995 20:35:35 -0800
Received: (from wpaul@localhost) by freefall.cdrom.com (8.6.10/8.6.6) id UAA22398; Sat, 1 Apr 1995 20:34:12 -0800
From: Bill Paul <wpaul>
Message-Id: <199504020434.UAA22398@freefall.cdrom.com>
Subject: Re: cvs commit: src/usr.sbin/ypbind ypbind.c
To: bde@zeta.org.au (Bruce Evans)
Date: Sat, 1 Apr 1995 20:34:12 -0800 (PST)
Cc: cvs-commiters, cvs-usrsbin
In-Reply-To: <199504020336.NAA29929@godzilla.zeta.org.au> from "Bruce Evans" at Apr 2, 95 01:36:18 pm
X-Mailer: ELM [version 2.4 PL24]
Content-Type: text
Content-Length: 382       
Sender: commit-owner@FreeBSD.org
Precedence: bulk

> 
> open() is spec'ed to return -1 for errors.  When doesn't it?  It is also
> spec'ed to return a non-negative integer so checking for < 0 is safer,
> but I prefer -1 for consistency with syscalls like read().
> 
> Bruce
> 

I must be going insane. You're right, of course. But I've seen ypbind
goof up in this situation before -- I know I have! Now I feel *really*
silly.

-Bill