From owner-freebsd-chat Thu Jul 29 16:13:33 1999 Delivered-To: freebsd-chat@freebsd.org Received: from smtp03.primenet.com (smtp03.primenet.com [206.165.6.133]) by hub.freebsd.org (Postfix) with ESMTP id 028F9154E6 for ; Thu, 29 Jul 1999 16:13:26 -0700 (PDT) (envelope-from tlambert@usr06.primenet.com) Received: (from daemon@localhost) by smtp03.primenet.com (8.8.8/8.8.8) id QAA15730; Thu, 29 Jul 1999 16:11:25 -0700 (MST) Received: from usr06.primenet.com(206.165.6.206) via SMTP by smtp03.primenet.com, id smtpd015629; Thu Jul 29 16:11:14 1999 Received: (from tlambert@localhost) by usr06.primenet.com (8.8.5/8.8.5) id QAA12642; Thu, 29 Jul 1999 16:11:11 -0700 (MST) From: Terry Lambert Message-Id: <199907292311.QAA12642@usr06.primenet.com> Subject: Re: What to tell to Linux-centric people?! To: Doug@gorean.org (Doug) Date: Thu, 29 Jul 1999 23:11:11 +0000 (GMT) Cc: tlambert@primenet.com, chat@freebsd.org In-Reply-To: <37A0B958.1E4251D6@gorean.org> from "Doug" at Jul 29, 99 01:28:08 pm X-Mailer: ELM [version 2.4 PL25] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-chat@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org > > Yes, I can. Scripting should be, by definition, twit-proof. No > > exceptions. > > No matter how twit-proof you make something, they will always come up with > a better twit. "If bash did not exist, some twit would invent it"? > > An apt analogy would be to say that the person who dies without > > ever getting into an automobile accident didn't need to use > > seatbelts, and so should have had the option of buying a car > > without them. > > Hmmm... I'm not sure that I understand your metaphor. It sounds like > you're advocating a "One true shell" approach, with no non-POSIX sh things > added to it. While there might be some merit to it, it's not a very unix-y > thing to advocate. If different shells are bad, how are different unices > good? I strongly believe that you can't blame the tool if the craftsman > uses it improperly. "Guns don't kill people, people kill people"? While I agree with the sentiment, one could imagine a gun that would only shoot for its owner, and then only at people not previously designated as "friendlies". One could take this further, and preload the friendlies list with all politicians, police officers, and liquor store owners. It's easy to do a /reductio ad absurdum/ argument on such a basis, and prove the major premise false. To extend your craftsman/tool analogy, I blame the tool for being a poor tool if it leaves tool marks on the work. We can drag in the "x86Open" thread now, with the point that it really doesn't matter if you define a universal ABI that is supported by all x86 UNIX and UNIX clones if you can't turn off the platform specific extensions. You can't write code that even _accidently_ uses a platform specific extension, if your intent is to write portable code. The same reasoning is applicable to shell scripts. > > > In fact, Bash is probably the most POSIX compliant shell going, > > > especially when invoked as sh. > > > > Does this turn off extensions? > > Yes. Then bash is "mostly harmless", but _only_ if it is used exactly this way. Terry Lambert terry@lambert.org --- Any opinions in this posting are my own and not those of my present or previous employers. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-chat" in the body of the message