Received: by taz.hyperreal.com (8.6.12/8.6.5) id GAA24931; Mon, 12 Feb 1996 06:07:10 -0800 Received: from sierra.zyzzyva.com by taz.hyperreal.com (8.6.12/8.6.5) with ESMTP id GAA24923; Mon, 12 Feb 1996 06:07:04 -0800 Received: from zyzzyva.com (localhost [127.0.0.1]) by sierra.zyzzyva.com (8.7.3/8.6.11) with ESMTP id IAA21434 for ; Mon, 12 Feb 1996 08:06:58 -0600 (CST) Message-Id: <199602121406.IAA21434@sierra.zyzzyva.com> To: new-httpd@hyperreal.com Subject: Re: votes on 1.0.2 In-reply-to: drtr's message of Mon, 12 Feb 1996 12:43:00 +0000. X-uri: http://www.zyzzyva.com/ Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Mon, 12 Feb 1996 08:06:58 -0600 From: Randy Terbush Sender: owner-new-httpd@apache.org Precedence: bulk Reply-To: new-httpd@hyperreal.com > >65b.new-make.tar.gz -1 > > This still seems to have some problems with dependencies. > > It would often decide do rebuild everything after making a > > change in one .c file. Painful decision on a Sparc 1+. > > Well, > 1) It doesn't do this for me (under SunOS 4) > 2) The makefile that is run using 65b.new-make.tar.gz should be almost > identical to that run using the old system. So it shouldn't behave > any differently. > > More details please? The package was a bit cumbersome when needing to add a library link flag. For example -lmsql. Changes to Makefile also caused a complete recompile, which may or may not be necessary. We've kicked this carcas around for quite some time. I think what would be helpful in solving the problem is to agree on a concept before anyone spends much of the work to make it happen. Suggestions: I see two (well three) options: 1. Using the GNU style autoconf. 2. Using the Perl style Configuration script. 3. Come up with our own.... I personally would vote for the Perl style since it seems to normally prompt the user for configuration info. We could then step through available modules and descriptions of each. IF we could get a consensus on one of these options, it would be much more attractive to do the work. There is way to much work involved to start in a direction that is not supported by the group. (speaking as someone who tried several iterations of Makefile in the past)