From owner-cvs-ports@FreeBSD.ORG Wed Jul 27 09:28:31 2011 Return-Path: Delivered-To: cvs-ports@FreeBSD.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id AD79C106564A; Wed, 27 Jul 2011 09:28:31 +0000 (UTC) (envelope-from freebsdports@chillibear.com) Received: from mail.sundive.org (mail.sundive.org [212.13.197.214]) by mx1.freebsd.org (Postfix) with ESMTP id 69DCB8FC14; Wed, 27 Jul 2011 09:28:31 +0000 (UTC) Received: from [91.125.124.138] (helo=[192.168.0.44]) by sundive.org with esmtpsa (TLSv1:DES-CBC3-SHA:168) (Exim 4.72 (FreeBSD)) (envelope-from ) id 1Qm02y-000AFR-24; Wed, 27 Jul 2011 10:05:07 +0100 User-Agent: Microsoft-Entourage/12.30.0.110427 Date: Wed, 27 Jul 2011 10:03:08 +0100 From: Eric To: Sahil Tandon , Doug Barton Message-ID: Thread-Topic: cvs commit: ports/mail/rubygem-mail Makefile distinfo Thread-Index: AcxMPACtvsuOaoy/QkiPws7JUlyWXQ== In-Reply-To: <20110727004348.GE1441@magic.hamla.org> Mime-version: 1.0 Content-type: text/plain; charset="US-ASCII" Content-transfer-encoding: 7bit X-Spam_score: -2.9 X-Spam_score_int: -28 X-Spam_bar: -- X-Spam: No X-bounce-key: sundive.org-1;freebsdports@chillibear.com;1311759004;91fccd9e; Cc: "Philip M. Gollucci" , Frederic Culot , cvs-ports@FreeBSD.org, cvs-all@FreeBSD.org, ports-committers@FreeBSD.org Subject: Re: cvs commit: ports/mail/rubygem-mail Makefile distinfo X-BeenThere: cvs-ports@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: CVS commit messages for the ports tree List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 27 Jul 2011 09:28:31 -0000 > From: Sahil Tandon > Date: Tue, 26 Jul 2011 20:43:50 -0400 > >> On Tue, 2011-07-26 at 17:32:34 -0700, Doug Barton wrote: > >>> On 07/26/2011 10:00, Philip M. Gollucci wrote: >>> Normally, I 100% agree with Doug. But we're not only talking about ruby >>> here, we're talking about rails. I hard a hard time naming gems that >>> don't do this crap. I almost replied yesterday and said add an option, >>> but IMHO either way is fine, and not worth diddling over. >> >> There are at least 3 possible ways to handle this situation: >> >> 1. Force the dependency > > I hope we can avoid this. > >> 2a. Have an OPTION and error out if the user chooses WITHOUT_FOO and foo >> is installed. > > For what (little) it's worth, my vote is for 2a. Well it is probably the cleanest route to reflect the situation. If I have time I will try to create some suitable logic to perform this at the weekend (little snowed under this week), unless Doug (or anyone else) has something in mind and has the time to file a PR? >> 2b. Have an OPTION and patch the distfiles if the user chooses >> WITHOUT_FOO and foo is installed. >> >> Personally I don't care which option is used, but IMO the problem of >> ports silently growing dependencies based on what is installed on the >> system has to be solved. >> I'm still not sure* we are silently growing dependencies in this case. There is an interaction between the the mail gem and the activerecord gem, but the mail gem does not have a hard dependency on activerecord, it uses it if found, if not found it doesn't break but continues working. So there would (should) be no problem with a user adding or removing activerecord as they wanted without affecting the apparent functionality of the mail gem**. Regards Eric * and I will have to confess to being far less experienced in ports internals than Doug here, so will defer to his more experienced view on this matter. ** personally I'm not sure the way the author has done things is ideal since they have created a situation where an apparent bug in the mail gem may or may not be caused by an incompatible future version of activerecord.