Xref: utzoo comp.unix.internals:828 comp.unix.programmer:326
Path: utzoo!attcan!uunet!aplcen!haven!uflorida!gatech!bloom-beacon!eru!hagbard!sunic!mcsun!ukc!stl!robobar!ronald
From: ronald@robobar.co.uk (Ronald S H Khoo)
Newsgroups: comp.unix.internals,comp.unix.programmer
Subject: Re: Checking return values (was: Trojan Horses, NFS, etc.)
Message-ID: <1990Oct25.075856.4923@robobar.co.uk>
Date: 25 Oct 90 07:58:56 GMT
References: <1885@necisa.ho.necisa.oz> <4315@pkmab.se> <1990Oct24.193712.8693@athena.mit.edu>
Followup-To: comp.unix.programmer
Organization: Robobar Ltd., Perivale, Middx., ENGLAND.
Lines: 25

[ wizard readers: skip now.  followups redirected. ]

scs@adam.mit.edu (Steve Summit) writes:

>			default:
>				fprintf(stderr, "%s: can't open %s: ",
>							progname, name);
>				perror("");

usenet: can't open scs@adam.mit.edu: not a tty.

(Steve, I'm shocked!  *you* of all people :-)

For those of you still puzzled: perror() must be called *immediately* after the
error returns, or the internal error state may be modified by calls
to system calls in the intervening situation.  It's annoying, but

	perror(name);
	fprintf(stderr, "%s: error opening %s (see error message above)\n",
		progname, name);

is more correct.

-- 
ronald@robobar.co.uk +44 81 991 1142 (O) +44 71 229 7741 (H)