Return-Path: Delivered-To: new-httpd-archive@hyperreal.org Received: (qmail 12606 invoked by uid 6000); 6 Oct 1998 21:21:08 -0000 Received: (qmail 12598 invoked from network); 6 Oct 1998 21:21:07 -0000 Received: from zap.ne.mediaone.net (HELO zap.ml.org) (24.128.120.231) by taz.hyperreal.org with SMTP; 6 Oct 1998 21:21:07 -0000 Received: (qmail 1284 invoked by uid 1000); 6 Oct 1998 21:20:49 -0000 From: Ben Hyde MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Date: Tue, 6 Oct 1998 17:20:45 -0400 (EDT) To: new-httpd@apache.org Subject: Binary tarballs In-Reply-To: <19981006204635.28792.qmail@hyperreal.org> References: <19981006204635.28792.qmail@hyperreal.org> X-Mailer: VM 6.43 under 20.4 "Emerald" XEmacs Lucid Message-ID: <13850.33801.937437.901555@zap.ml.org> Sender: new-httpd-owner@apache.org Precedence: bulk Reply-To: new-httpd@apache.org This should not be a show stopper for the pending release. This is the kind of problem that needs by some code which we try and then rework. The current APACI "make install" will do a lot of wonderful things, like write conf files for you. It is sort of wise about where things get installed on some platforms. These good deeds should to be done after the binary tarball is unpacked not before. An install binary tarball script could include switches to discard things (like the sources I believe should always be sent). My first thought was: 1 cd apache-1.3 2 ( CC="gcc" \ 3 CFLAGS=-g \ 4 ./configure \ 5 --prefix=./local \ 6 --enable-module=all \ 7 --enable-shared=max; \ 8 make clean \ 9 make; \ 10 make install \ 11 ) > build.log 2>&1 There is almost no line in there that doesn't offer opportunities for discussion. 1. Presumably this was made with cvs export? 2. do we want variations on C compiler. 3. well? what switches to we want? 4. 5. It is thought provoking how much stuff APACI puts in here and how much nice editing it does to that stuff. (as an aside this a relative prefix doesn't actually work out) 6. well? what modules? 7. I think it's good to share :-) 8. how do we assure some historical litter doesn't appear in a submitted tarball by mistake? 9. :-) 10. let APACI do it's magic. 11. Some record of how the tar ball was built should appear. It probably ought to include some sign of who did the building. A PGP signature over the local directory would be very noble. - ben hyde