From owner-freebsd-current Wed Apr 7 2:52:34 1999 Delivered-To: freebsd-current@freebsd.org Received: from ns.oeno.com (ns.oeno.com [194.100.99.145]) by hub.freebsd.org (Postfix) with SMTP id EFEFE14D12 for ; Wed, 7 Apr 1999 02:52:27 -0700 (PDT) (envelope-from will@ns.oeno.com) Received: (qmail 438 invoked by uid 1001); 7 Apr 1999 09:50:27 -0000 Date: 7 Apr 1999 09:50:27 -0000 Message-ID: <19990407095027.435.qmail@ns.oeno.com> From: Ville-Pertti Keinonen To: dillon@apollo.backplane.com Cc: current@freebsd.org In-reply-to: <199904061701.KAA09326@apollo.backplane.com> (message from Matthew Dillon on Tue, 6 Apr 1999 10:01:36 -0700 (PDT)) Subject: Re: aio_read Sender: owner-freebsd-current@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG > If it's a redirected output file you simply make it O_APPEND, at which > point the seek offset in the descriptor becomes irrelevant. As others have pointed out, O_APPEND is much newer than the offset-sharing behavior. It also doesn't fix things for inputs. Of course with buffering, it's hard for it to have sensible semantics, but for some commands, it does work and can even useful. While I doubt many scripts rely on such behavior, it does avoid inconsistencies between internal and external commands, as well as files, ttys and pipes. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-current" in the body of the message