From owner-freebsd-stable Wed Jan 15 19:35: 4 2003 Delivered-To: freebsd-stable@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id D556437B401 for ; Wed, 15 Jan 2003 19:34:59 -0800 (PST) Received: from rutger.owt.com (rutger.owt.com [204.118.6.16]) by mx1.FreeBSD.org (Postfix) with ESMTP id 3F4C643ED8 for ; Wed, 15 Jan 2003 19:34:59 -0800 (PST) (envelope-from kstewart@owt.com) Received: from owt-207-41-94-233.owt.com (owt-207-41-94-233.owt.com [207.41.94.233]) by rutger.owt.com (8.9.3/8.9.3) with ESMTP id TAA25906; Wed, 15 Jan 2003 19:34:56 -0800 Content-Type: text/plain; charset="iso-8859-1" From: Kent Stewart To: Stephen Hilton , freebsd-stable@FreeBSD.ORG Subject: Re: Portversion - Portupgrade Date: Wed, 15 Jan 2003 19:34:56 -0800 User-Agent: KMail/1.4.3 References: <3E215BB3.9020400@swissgeeks.com> <200301151237.13627.kstewart@owt.com> <20030115174051.2d3ce171.nospam@hiltonbsd.com> In-Reply-To: <20030115174051.2d3ce171.nospam@hiltonbsd.com> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Message-Id: <200301151934.56555.kstewart@owt.com> Sender: owner-freebsd-stable@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.ORG On Wednesday 15 January 2003 03:40 pm, Stephen Hilton wrote: > On Wed, 15 Jan 2003 12:37:13 -0800 > > Kent Stewart wrote: > > On Wednesday 15 January 2003 11:46 am, Stephen Hilton wrote: > > > On Wed, 15 Jan 2003 10:54:14 -0800 > > > > > > "Kevin Oberman" wrote: > > > > > Date: Sun, 12 Jan 2003 13:12:35 +0100 > > > > > From: Pierrick Brossin > > > > > Sender: owner-freebsd-stable@FreeBSD.ORG > > > > > > > > > > Belphoebe Niressi wrote: > > > > > > > > > Is the "portsdb -Uu" working properly now? I have been using: > > > > > > #> cd /usr/ports > > > #> rm INDEX* > > > #> make index > > > #> pkgdb -Fv > > > #> portsdb -u > > > > I think they are both broken in some respect right now because what > > I am seeing is the following using "make index" and then run > > "portsdb -u" > > > > > > /man7/des_modes.7.gz /nonexistentlocal/man/""/man7/Modes.7.gz > > /nonexistentlocal/man/""/man7/des_modes.7.gz > > /nonexistentlocal/man/""/man7/of.7.gz | /usr/bin/awk '{for (i=3D2;hav= e > > i<=3DNF; i+=3D2) print $i}'" returned non-zero status > > Done. > > > > [Updating the portsdb in /usr/ports ... - 8009 > > port entries found > > .........1000.........2000.........3000.........4000.........5000.. > >.......6000.........7000.........8000 ..... done] > > > > Running "portsdb -Uu I get > > > > > > make_index: no entry for: /usr/ports/net/p5-Net-DNS > > make_index: no entry for: /usr/ports/net/p5-Net-DNS > > Warning: Duplicate INDEX entry: freeciv-imlib-1.13.0 > > done > > ruby# portsdb -u > > [Updating the portsdb in /usr/ports ... - 8002 > > port entries found > > .........1000.........2000.........3000.........4000.........5000.. > >.......6000.........7000.........8000 ..... done] > > > > There are 7 ports that portsdb -U can't account for. > > Regarding this error: > "Warning: Duplicate INDEX entry: freeciv-imlib-1.13.0" > is this an "inside joke" or just a real tough problem to > fix? been seeing this error for ages (I would fix this if > I could, but I couldn't code myself out of a brown paper > bag :-) I don't either. I think Kris made a comment on one of them a while back=20 about the index building process not being able to detect that they=20 were different. > > > I also don't see why you are doing more than > > > > cd /usr/ports > > make indextwice > > portsdb -u > > Thanks for the correction Kent, I was doing a couple of unnecessary > steps. I would usually rm my /usr/ports/INDEX /usr/ports/INDEX.db and > /usr/ports/INDEX.txt files because they were going to be recreated > anyways (I do refuse the INDEX file when cvsuping to save bandwidth). > > > Everything else is something I would do if your installed based > > gets messed up. For example, portsdb -Fv doesn't have anything to > > do with making your INDEX* files. When you run "make index", it > > always wipes INDEX. > > I never referenced the command "portsdb -Fv" I said "pkgdb -Fv", > and I picked up this step from a message thread started by Stacey > Roberts in December 2002 that you had responded to, so I assumed > (ya I know) that this was OK or even a "good thing" to do during > the INDEX's generation process.they are I was looking at something else and missed that part. I maintain I run=20 "kent" because I have so many aliases and shell scripts that do most of=20 the repetitive things. I also try to keep things clean. I used to have=20 a script that did everything from cvsup to installworld and I only had=20 to run mergemaster when it finished. Now, I split them up into tasks =20 that have a single purpose. I got used to doing this during my=20 programming days. You didn't change or add many features before you=20 started debugging what you had done. If you stepped on something you=20 didn't know which change had done it. So, my update ports script only=20 cvsup updates the port structure and recreates my INDEX files. Running=20 pkgdb violates my single purpose rule. Since you really need to update=20 your INDEX files everytime you cvsup ports-all, rebuilding the INDEX=20 files is part of my port cvsup update script. > > As I understand "pkgdb -Fv" it is an interactive check/repair of > your installed ports/packages as referenced by the /var/db/pkg > entries (package registry database) and _Yes_ it has no bearing on > generation of the INDEX/db files. I have done this a couple of times in near past. I may get it right yet.=20 When I see the message, I cut and paste it. I don't think I have ever=20 typed "pkgdb -F" from scratch. When I am looking at something in the=20 vicinity of portsdb, I naturally type portsdb instead of pkgdb. I don't=20 even notice that I have done it and I proof read everything 2 or 3=20 times before I send it. > > Thanks again Kent for your willingness to share your expertise > and all the great pointers and background information I pick up > just reading yours and others contributions to the mailing lists. Everyone zeroes in on different specific's. I think it depends on what=20 you have been bitten by. The buildworld, build[install]kernel,=20 installworld got me during the update from 4.0 to 4.1. I was used to=20 doing a make world and then do the kernel. I couldn't do a "make world"=20 now if I wanted to. In addition, if you have interpreted something=20 differently, you are sure to have some one point that out. The other=20 cool part of the lists is that you don't know what part of the world is=20 going to provide the person that answers your question. Kent > > Regards, > > Stephen Hilton > nospam@hiltonbsd.com --=20 Kent Stewart Richland, WA http://users.owt.com/kstewart/index.html To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-stable" in the body of the message