From owner-cvs-ports@FreeBSD.ORG Mon Jul 13 17:43:24 2009 Return-Path: Delivered-To: cvs-ports@FreeBSD.org Received: by hub.freebsd.org (Postfix, from userid 1033) id 76E441065672; Mon, 13 Jul 2009 17:43:24 +0000 (UTC) Date: Mon, 13 Jul 2009 17:43:24 +0000 From: Alexey Dokuchaev To: Gerald Pfeifer Message-ID: <20090713174324.GA70528@FreeBSD.org> References: <200907072353.n67NrlPh032677@repoman.freebsd.org> <20090713143026.GA36954@regency.nsu.ru> Mime-Version: 1.0 Content-Type: text/plain; charset=koi8-r Content-Disposition: inline In-Reply-To: User-Agent: Mutt/1.4.2.1i Cc: Alexey Dokuchaev , "Philip M. Gollucci" , cvs-ports@FreeBSD.org, cvs-all@FreeBSD.org, ports-committers@FreeBSD.org Subject: Re: cvs commit: ports/devel/scons Makefile distinfo pkg-plist 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: Mon, 13 Jul 2009 17:43:24 -0000 On Mon, Jul 13, 2009 at 07:08:29PM +0200, Gerald Pfeifer wrote: > On Mon, 13 Jul 2009, Alexey Dokuchaev wrote: > >> - Update to checkpoint release 1.2.0_20090223 > > This change is wrong; PORTREVISION should not be used to encode any > > software version part; its sole purpose for ports infrastructure only. > > For complex vendor versions, DISTVERSION should be used instead. Right > > now, for instance, you've essentially made impossible to bump PORTREVISION, > > shall it be required for any reason, without introducing ugly hacks. > > This sounds very familiar to the status quo I maintained with the lang/gcc > ports for some years which I then resolved a few months ago. Philip, you > may want to have a look at lang/gcc43 for an example how I addressed this > there -- and if you find improvements, let me know. ;-) In your case, Gerald, you seem to deal this the fact that GCC people drop minor .4 version for their snapshots, e.g. for version 4.3.4.%date% snapshot would be 4.3-%date% as opposed to 4.3.4-%date% (the latter would allow you to user DISTVERSION without cryptic VERSIONSTRING hack). In Philip's case, DISTVERSION should be enought, I guess. ./danfe