From owner-freebsd-ports Tue Jul 25 20:34:52 2000 Delivered-To: freebsd-ports@freebsd.org Received: from bsa-1-as01-7-a57.gd.uol.com.br (bsa-1-as01-7-a57.gd.uol.com.br [200.197.118.57]) by hub.freebsd.org (Postfix) with ESMTP id C651E37B636 for ; Tue, 25 Jul 2000 20:34:42 -0700 (PDT) (envelope-from lioux@uol.com.br) Received: (qmail 747 invoked by uid 1001); 26 Jul 2000 03:34:59 -0000 From: "Mario Sergio Fujikawa Ferreira" Date: Wed, 26 Jul 2000 00:34:37 -0300 To: CHOI Junho Cc: Mario Sergio Fujikawa Ferreira , freebsd-ports@freebsd.org Subject: Re: Ghostscript dependencies Message-ID: <20000726003437.B312@Fedaykin.here> References: <20000720135834.A50109@Fedaykin.here> <86hf9jllis.fsf@gradius.myhome> <20000721184047.A68315@Fedaykin.here> <86ittyqot4.fsf@gradius.myhome> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.2.4i In-Reply-To: <86ittyqot4.fsf@gradius.myhome>; from cjh@kr.FreeBSD.org on Sat, Jul 22, 2000 at 10:25:37PM +0900 Sender: owner-freebsd-ports@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org On Sat, Jul 22, 2000 at 10:25:37PM +0900, CHOI Junho wrote: > >>>>> "MSFF" == Mario Sergio Fujikawa Ferreira writes: > > MSFF> On Fri, Jul 21, 2000 at 09:23:17PM +0900, CHOI Junho wrote: > > > I think it's the reason that these ghostscript-related ports doesn't > > have dependency on only one ghostscript version. > > MSFF> I understand your claim. > MSFF> However, we should try to think about someway to handle > MSFF> this. IMHO, dependencies SHOULD be handled whenever possible. > MSFF> Sometimes, even when not. :) > MSFF> That's one of the main purposes of this chibang we call > MSFF> ports tree. > MSFF> Naive users, will install gv to find out it does > MSFF> not work. Damn FreeBSD, it is broken. I will try that > MSFF> graphic interface from RH. > MSFF> This is no joke. > > But RH or many Linux distribution has no 'freedom of choice'. It just > install only one version, so it's not a problem for them. I was not arguing those were better just how some prospective user might react. :) > MSFF> For you and all other expert users all around > MSFF> who chose an specific version of ghostscript, the proposed > MSFF> modification will not produce any problems. You already > MSFF> have a binary called gs on your path, the build dependecy > MSFF> check will ignore building ghostcript6 although not > MSFF> adding the dependency correctly as I would have wished for. > MSFF> However, for those who just want to get it working > MSFF> whatever it is (i.e., ghostview and family) the modification > MSFF> is beneficial IMHO. > > Yes, but it applies only in case of installing from > ports. Packages(.tgz) have a dependency list which have exact version > numbers of depending packages. If we wrote RUN_DEPENDS to gs6 on our > Makefile, it records gs-6.01 on @pkgdep, so it will install gs6 when > installing ghostview, even he has installed other ghostscripts. Not if my idea follows, it will check in the DEPENDENCY_RUN database and thing the declared gs as already available. If the user does not install a GS-family port 1st then the default goes. I am sorry, can't please everyone. :) I am accepting suggestions on this one. > MSFF> One possible solution for this would be building > MSFF> a inverted list indexing with our depedency checks: > > [...] > > It can be some solution. But for now, I think the following is prefer: > > - add pkg/MESSAGE saying 'please install any of ghostscript ports'. For now, I agree, such a change would not happen in less than a month anyway. > But, the situation like ghostscript can have a problem for users. If > we have a concept of a group of related packages(e.g. ghostscript = > gs2, gs55, gs6, ko-gs-ft, ko-gs-httf, ja-vfgs etc...) and make related > application such as gv depend on this group(ghostscript). That's the idea behing inverted list indexing. All of them install gs so that gs indexes them as per the last email. MSFF> When I try to be install ghostview, I will MSFF>check the apropriate database (_RUN) and I will find MSFF>there is a RUN_DEPENDS gs there pointing at MSFF>korean/*ghostscript. So it will know who to depend on MSFF>when adding both REQUIRED_BY and @pkgdep. Besides, we can use this as integrity checking; oops we already have a gs under PREFIX, do not go there, use FORCE_PKG_REGISTER if you must; or, try a different PREFIX. Regards, Mario Ferreira To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-ports" in the body of the message