Date: Sun, 12 Jul 1998 03:21:40 -0500 (CDT) From: Joel Ray Holveck <joelh@gnu.org> To: tlambert@primenet.com Cc: bakul@torrentnet.com, dchapes@ddm.on.ca, rminnich@Sarnoff.COM, hackers@FreeBSD.ORG Subject: Re: Improvemnet of ln(1). Message-ID: <199807120821.DAA01163@detlev.UUCP> In-Reply-To: <199807120123.SAA17458@usr08.primenet.com> (message from Terry Lambert on Sun, 12 Jul 1998 01:23:19 %2B0000 (GMT)) References: <199807120123.SAA17458@usr08.primenet.com>
next in thread | previous in thread | raw e-mail | index | archive | help
>>> How on earth will issuing a diagnostic make it harder to write >>> scripts? >> Because now you have to filter out (additional) noise. > Consider the "improvement" to the dump(8) command about a year ago... Regrettably, I don't remember this. Could you please refresh my memory? > Especially if it is a deviation in order to protect the kind of people > who type "DIR" in DOS after "DELETE FILE.DAT" to reassure themselves > that the system, indeed, did what they told it to do, and deleted the > file. > Protecting people who don't believe a file is really gone until they > get a directory listing without the file in it is pretty low on the > ladder of UNIX priorities. I trust Unix to do what I tell it to. But I don't mind it reminding me if I may have had one too many as it's doing what I told it to. > I wouldn't object too strongly to a "-w", as has been suggested elsewhere, > so long as the alias was not there by default. I would still object a > little, on the principle that a future version of POSIX might define > a "-w" argument, causing a namespace collision with the FreeBSD version > of the command (and thus breaking scripts, .login's, .cshrc's, etc.). I generally prefer a 'no-warnings' option over a 'enable-warnings' option if the warnings don't change the effects of the command. That is because a user at the keyboard is lazy, and will not, generally, type the extra options. Yes, we can make aliases, and so on, and so forth. On the other hand, those who are scared of breaking something can do the same with the -q (quiet) option I proposed. It comes back to my earlier question: Are there going to be more lossages if we add the warnings, or if we don't? I know for a fact that I've done the same thing that rminnich described in his original post. I also know that I've never written a script that this change would break. How many of us have done each? Happy hacking, joelh -- Joel Ray Holveck - joelh@gnu.org - http://www.wp.com/piquan Fourth law of programming: Anything that can go wrong wi sendmail: segmentation violation - core dumped To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-hackers" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199807120821.DAA01163>