From owner-freebsd-current Mon Jul 29 11:06:12 1996 Return-Path: owner-current Received: (from root@localhost) by freefall.freebsd.org (8.7.5/8.7.3) id LAA13197 for current-outgoing; Mon, 29 Jul 1996 11:06:12 -0700 (PDT) Received: from godzilla.zeta.org.au (godzilla.zeta.org.au [203.2.228.19]) by freefall.freebsd.org (8.7.5/8.7.3) with SMTP id LAA13188 for ; Mon, 29 Jul 1996 11:06:09 -0700 (PDT) Received: (from bde@localhost) by godzilla.zeta.org.au (8.6.12/8.6.9) id EAA08896; Tue, 30 Jul 1996 04:02:41 +1000 Date: Tue, 30 Jul 1996 04:02:41 +1000 From: Bruce Evans Message-Id: <199607291802.EAA08896@godzilla.zeta.org.au> To: bde@zeta.org.au, wollman@lcs.mit.edu Subject: Re: `const char rcsid[]' vs -traditional Cc: current@freebsd.org Sender: owner-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk >> Many sources no longer compile with non-ANSI compilers because they >> begin with >> const char rcsid[] = ... >So? It's broken for non-ANSI compilers. The correct way to fix this is to move the declarations after the #includes so that happens to be included, but this is too large a change for me. It's inconsistent with the care taken in other parts of the sources. E.g., /usr/src/lib/libc/gen/getosreldate.c, which was written by someone named wollman, begins with a `const char rcsid[]' but then uses an ugly `#if __STDC__' to use a new-style header for ANSI compilers without (further) breaking things for non-ANSI compilers. Bruce