From owner-cvs-all@FreeBSD.ORG Thu Feb 19 14:09:01 2004 Return-Path: Delivered-To: cvs-all@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id DC61016A4CF; Thu, 19 Feb 2004 14:09:00 -0800 (PST) Received: from shaft.techsupport.co.uk (shaft.techsupport.co.uk [212.250.77.214]) by mx1.FreeBSD.org (Postfix) with ESMTP id 8381543D1F; Thu, 19 Feb 2004 14:09:00 -0800 (PST) (envelope-from setantae@submonkey.net) Received: from cpc2-cdif3-6-0-cust204.cdif.cable.ntl.com ([81.103.67.204] helo=shrike.submonkey.net ident=mailnull) by shaft.techsupport.co.uk with esmtp (TLSv1:DHE-RSA-AES256-SHA:256) (Exim 4.30; FreeBSD) id 1AtwLq-000Dbr-Ix; Thu, 19 Feb 2004 22:08:58 +0000 Received: from setantae by shrike.submonkey.net with local (Exim 4.30; FreeBSD) id 1AtwLo-000939-Cj; Thu, 19 Feb 2004 22:08:56 +0000 Date: Thu, 19 Feb 2004 22:08:56 +0000 From: Ceri Davies To: Maxim Sobolev Message-ID: <20040219220856.GI380@submonkey.net> Mail-Followup-To: Ceri Davies , Maxim Sobolev , Kris Kennaway , cvs-ports@FreeBSD.ORG, cvs-all@FreeBSD.ORG, ports-committers@FreeBSD.ORG References: <200402191122.i1JBMdHd026435@repoman.freebsd.org> <20040219112932.GA11187@xor.obsecurity.org> <4034A2C4.7030501@portaone.com> <20040219115641.GA11791@xor.obsecurity.org> <4034A78F.9080702@portaone.com> <4034A883.80907@portaone.com> <20040219122529.GA12202@xor.obsecurity.org> <4034ACD3.8010907@portaone.com> Mime-Version: 1.0 Content-Type: multipart/signed; micalg=pgp-sha1; protocol="application/pgp-signature"; boundary="qFgkTsE6LiHkLPZw" Content-Disposition: inline In-Reply-To: <4034ACD3.8010907@portaone.com> X-PGP: finger ceri@FreeBSD.org User-Agent: Mutt/1.5.4i Sender: Ceri Davies cc: cvs-ports@FreeBSD.ORG cc: ports-committers@FreeBSD.ORG cc: cvs-all@FreeBSD.ORG cc: Kris Kennaway Subject: Re: cvs commit: ports/net/asterisk Makefile X-BeenThere: cvs-all@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list List-Id: CVS commit messages for the entire tree List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Thu, 19 Feb 2004 22:09:01 -0000 --qFgkTsE6LiHkLPZw Content-Type: text/plain; charset=us-ascii Content-Disposition: inline Content-Transfer-Encoding: quoted-printable On Thu, Feb 19, 2004 at 02:32:19PM +0200, Maxim Sobolev wrote: > Kris Kennaway wrote: >=20 > >On Thu, Feb 19, 2004 at 02:13:55PM +0200, Maxim Sobolev wrote: > > > > > >>>>No, as I said, this is a supported configuration. Moreover, I'm sure > >>>>it's common for people to move their ports collection from /usr/ports > >>>>to some other location and replace it with a symlink (in fact I've > >>>>done that myself), and this works for all 10364 ports except yours, > >>>>prior to this commit. > >>> > >>> > >>>I still think that you aren't quite correct. You (and everyone who wan= t=20 > >>>to move /usr/ports over) should have set PORTSDIR to its real location= =20 > >>>(that is /a/ports in bento scripts), which would allow=20 > >>>'${WRKDIRPREFIX}${PORTSDIR}/' in ports Makefiles instead of much uglie= r=20 > >>>${WRKDIR}/../../../' (or eaquially ugly construct involving .CURDIR).= =20 > >>>The latter is worse because in this case it is impossible to put ports= =20 > >>>directory (e.g. asterisk in this case) anywhere in the file system not= =20 > >>>in the ${PORTSDIR} and have it building just fine, which is possible= =20 > >>>with the former. IMO, this is much common and useful feature than=20 > >>>ability to move /usr/ports with the help of symlink. > >> > >>Also my version of behaviour is documented (ans has been for a long=20 > >>time) as the One True Way[tm], so that I'd suggest you to fix bento=20 > >>scripts. > >> > >>http://www.freebsd.org/doc/en_US.ISO8859-1/books/porters-handbook/porti= ng-wrkdirprefix.html > > > > > >That seems to be a bug in the documentation (note that it's internally > >inconsistent because it refers to both ${WRKDIRPREFIX}${PORTSDIR} and > >${WRKDIRPREFIX}${.CURDIR} as the "correct" way to reference a port's > >${WRKDIR} depending on whether that port is the current or another > >one). >=20 > No, there is no inconsistency, please read again. It only tells that if= =20 > you are defining your own WRKDIR in your port, you should define it as >=20 > catfoo/portbar/Makefile: > WRKDIR =3D ${WRKDIRPREFIX}${.CURDIR}/blablabla >=20 > This has nothing to do with reffering other ports. >=20 > >To repeat, 10364 ports don't have a problem with this policy that has > >been deliberately enforced by bento since before I came along. Your 1 > >port did. The numbers are really not on your side for making a > >persuasive argument here, and I'm sure we both have better things to > >do with our time than to continue to debate it. >=20 > Only small fraction of all those 10364 ports reffer to other ports=20 > WRKDIR (30-40 according to a quick greep), 3 of them correctly use=20 > ${WRKDIRPREFIX}${PORTSDIR}, all others probably were "fixed" to=20 > workaround a broken assumption that bento makes. ports(7) currently says: PORTSDIR Location of the ports tree. This is /usr/ports on FreeB= SD and OpenBSD, and /usr/pkgsrc on NetBSD. That's sufficiently ambiguous to be interpreted as support for both sides here, but whatever the outcome of this discussion, I'd suggest that this is updated to reflect the way this is supposed to be, so that this argument never has to happen again. Ceri --=20 --qFgkTsE6LiHkLPZw Content-Type: application/pgp-signature Content-Disposition: inline -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.2.4 (FreeBSD) iD8DBQFANTP4ocfcwTS3JF8RArgKAKCot9yZaH3d6W+phKUI4YZ6HzQfiQCgvtgO N14oRHtPP4EsMenIszWlfN8= =o4Ws -----END PGP SIGNATURE----- --qFgkTsE6LiHkLPZw--