From owner-freebsd-ports@FreeBSD.ORG Sat Jul 31 12:26:28 2010 Return-Path: Delivered-To: freebsd-ports@FreeBSD.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id 67C69106564A; Sat, 31 Jul 2010 12:26:28 +0000 (UTC) (envelope-from ehaupt@critical.ch) Received: from mx.critical.ch (cl-8.zrh-02.ch.sixxs.net [IPv6:2001:1620:f00:7::2]) by mx1.freebsd.org (Postfix) with ESMTP id D2C4E8FC25; Sat, 31 Jul 2010 12:26:27 +0000 (UTC) Received: from beaver.home.critical.ch (80-218-148-45.dclient.hispeed.ch [80.218.148.45]) (authenticated bits=0) by mx.critical.ch (8.14.3/8.14.3/critical-1.0) with ESMTP id o6VCQNpb087959; Sat, 31 Jul 2010 14:26:23 +0200 (CEST) (envelope-from ehaupt@critical.ch) Date: Sat, 31 Jul 2010 14:26:13 +0200 From: Emanuel Haupt To: Doug Barton Message-Id: <20100731142613.7bb36c2e.ehaupt@critical.ch> In-Reply-To: References: <201007250132.o6P1Whjh069411@repoman.freebsd.org> <20100726135113.0ee77eea.ehaupt@critical.ch> <20100726233950.822f4f13.ehaupt@FreeBSD.org> X-Mailer: Sylpheed 3.0.3 (GTK+ 2.20.1; amd64-portbld-freebsd8.1) Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Cc: freebsd-ports@FreeBSD.org, Emanuel Haupt Subject: Re: Inverting OPTIONS (Was: Re: cvs commit: ports/net-im/libpurple Makefile distinfo) X-BeenThere: freebsd-ports@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Porting software to FreeBSD List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Sat, 31 Jul 2010 12:26:28 -0000 Doug Barton wrote: > On Mon, 26 Jul 2010, Emanuel Haupt wrote: > > > Doug Barton wrote: > >> On Mon, 26 Jul 2010, Emanuel Haupt wrote: > >> > >>> Doug Barton wrote: > >>>> ports/net-im/libpurple/files/patch-libpurple_protocols_oscar_oscar.c > >>>> has been added to the latest version, so the patch is no longer > >>>> needed. > >>> > >>> Also, the PLIST is broken if the PERL option is specified: > >>> > >>> http://people.freebsd.org/~ehaupt/buildlogs/libpurple-2.7.2.log > >>> > >>> Maybe from time to time a QAT exp run with inverted OPTIONS values > >>> would be revealing. > >> > >> As I said on IRC yesterday, there is no reason to expect that > >> simply inverting OPTIONS would work. A lot of OPTIONS are not > >> binary, and the defaults are chose to work together, often in > >> harmony with the dependencies that they trigger. > >> > >> Put another way, it does not logically follow that if "A, B, and ! > >> C" works that "!A, !B, and C" would also work. > > > > I didn't assume that. It would be an easy way to spot simple > > mistakes such as incomplete package lists, unfetchable optional > > distfiles, things like that. > > Well then why don't you explain in more detail what you're > suggesting, because one of us is confused. :) Inverting OPTIONS setting in an exp run every now and then could screen *simple* mistakes. For instance conditionally defined patchfiles, distfiles, extra patches that don't apply anymore and so on. There is no way this can ensure that those OPTIONS would even make the port build but that would not be the intention. Emanuel