Re: FreeBSD unmaintained ports which are currently scheduled for deletion
I found some problems in your patch. - probably distinfo is not correct - net/rubygem-rubytter needs to be updated to 1.5.0 or newer 2013-04-22 3:47 olli hauer wrote: The patch updates the port to version 2.1.0 (ruby 1.9 is supported since 1.3.x) I can't make out why the port was marked as broken almost 2 years ago. When the port was marked as broken, PORTVERSION was 1.9.0 (already newer than 1.3.x). It can't be broken with ruby 1.9. -- `whois vmeta.jp | nkf -w` meta 2013-04-22 3:47 olli hauer wrote: On 2013-04-21 12:55, Koichiro IWAO wrote: 2013-04-21 17:30 に lini...@freebsd.org さんは書きました: portname: net-im/rubygem-termtter description:Terminal based Twitter client maintainer: d...@freebsd.org deprecated because: Does not work with Ruby 1.9 expiration date:2013-05-02 build errors: none. overview: http://portsmon.FreeBSD.org/portoverview.py?category=net-im&portname=rubygem-termtter I strongly disagree to delete net-im/rubygem-termtter. It is mismarked as BROKEN by portmgr for one year even if it perfectly works with ruby 1.9. Would you mind to test the following patch (update to 2.1.0). http://people.freebsd.org/~ohauer/diffs/rubygem-termtter.diff The patch updates the port to version 2.1.0 (ruby 1.9 is supported since 1.3.x) Patch is not tested, I avoid to install ruby on my machines. -- `whois vmeta.jp | nkf -w` meta ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
Re: FreeBSD unmaintained ports which are currently scheduled for deletion
On 2013-04-21 12:55, Koichiro IWAO wrote: > 2013-04-21 17:30 に lini...@freebsd.org さんは書きました: >> portname: net-im/rubygem-termtter >> description:Terminal based Twitter client >> maintainer: d...@freebsd.org >> deprecated because: Does not work with Ruby 1.9 >> expiration date:2013-05-02 >> build errors: none. >> overview: >> http://portsmon.FreeBSD.org/portoverview.py?category=net-im&portname=rubygem-termtter > > I strongly disagree to delete net-im/rubygem-termtter. > It is mismarked as BROKEN by portmgr for one year even if it perfectly works > with ruby 1.9. > Would you mind to test the following patch (update to 2.1.0). http://people.freebsd.org/~ohauer/diffs/rubygem-termtter.diff The patch updates the port to version 2.1.0 (ruby 1.9 is supported since 1.3.x) Patch is not tested, I avoid to install ruby on my machines. ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
Re: CFT: vlc 2.0.6
Juergen Lock wrote: > Port update: > > http://people.freebsd.org/~nox/tmp/vlc-2.0.6-001.patch > > Please give this a good testing again... Works for me so far. I mostly tested with DVDs. Fabian signature.asc Description: PGP signature
Re: FreeBSD unmaintained ports which are currently scheduled for deletion
On 21 April 2013 11:55, Koichiro IWAO wrote: > 2013-04-21 17:30 に lini...@freebsd.org さんは書きました: >> >> portname: net-im/rubygem-termtter >> description:Terminal based Twitter client >> maintainer: d...@freebsd.org >> >> deprecated because: Does not work with Ruby 1.9 >> expiration date:2013-05-02 >> build errors: none. >> overview: >> >> http://portsmon.FreeBSD.org/portoverview.py?category=net-im&portname=rubygem-termtter > > > I strongly disagree to delete net-im/rubygem-termtter. > It is mismarked as BROKEN by portmgr for one year even if it perfectly works > with ruby 1.9. Don't worry, GNATS is checked before any ports are deleted; it won't go anywhere. Most of us have a load of commits queued during the hard freeze, and they haven't gone in yet. Chris ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
emulators/qemu-devel port update (bsd-user, clang)
Hi! I'm currently waiting for redports to come back and build the following update, in case anyone wants to already test it... http://people.freebsd.org/~nox/tmp/qemu-devel-1.4.0_1-001.patch Log: - Update to sson's latest bsd-user patches. - Turns out qemu 1.4.x now builds with clang as is, so remove CLANG_HACK [1] - Bump PORTREVISION. Noticed by: Brad Smith [1] Thanx! :) Juergen ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
Re: [Call for Test] [NEW PORT] net/mobile-broadband-provider-info-20130312
On 21 Apr 2013 14:37, "Chris Whitehouse" wrote: > > On 14/04/2013 14:12, nemysis wrote: >> >> >> http://sourceforge.net/projects/nemysisfreebsdp/files/Not_FreeBSD_Port/mobile-broadband-provider-info.shar > > > I do sometimes use ppp over my mobile phone so I thought I'd try it. I get > muji64# make > "Makefile", line 32: Malformed conditional (${PORT_OPTIONS:MDOCS}) > "Makefile", line 36: if-less endif > make: fatal errors encountered -- cannot continue > > With: > muji64# diff -u Makefile.original Makefile > --- Makefile.original 2013-04-21 13:45:06.0 +0100 > +++ Makefile2013-04-21 14:26:34.0 +0100 > @@ -29,7 +29,8 @@ > .include > > post-install: > -.if ${PORT_OPTIONS:MDOCS} > +#.if ${PORT_OPTIONS:MDOCS} > +.if !defined(NOPORTDOCS) > ${MKDIR} ${DOCSDIR} > ${INSTALL_DATA} ${PORTDOCS:S|^|${WRKSRC}/|} ${DOCSDIR} > .endif > > it makes and installs fine. Is this due to the age of my system? > muji64# uname -a > FreeBSD muji64 9.0-RELEASE-p3 FreeBSD 9.0-RELEASE-p3 #0: Tue Jun 12 02:52:29 UTC 2012 r...@amd64-builder.daemonology.net:/usr/obj/usr/src/sys/GENERIC amd64 Your ports tree is incredibly out of date :) Your patch is fine for you however. Chris ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
Re: CFT: google-earth crash fix (ports/160422 [1])
On Sun, Apr 21, 2013 at 12:25:37PM +0200, Kurt Jaeger wrote: > Hi! > > > > > http://people.freebsd.org/~nox/tmp/google-earth-6.0.3.2197-crashfixp-001.patch > > Works on 9.1-amd64. > Ok as it seems to work for everyone that tested it I just committed the update. If it still crashes for some people it's at least much better than it was... Thanx! :) Juergen ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
Re: [Call for Test] [NEW PORT] net/mobile-broadband-provider-info-20130312
On 14/04/2013 14:12, nemysis wrote: http://sourceforge.net/projects/nemysisfreebsdp/files/Not_FreeBSD_Port/mobile-broadband-provider-info.shar I do sometimes use ppp over my mobile phone so I thought I'd try it. I get muji64# make "Makefile", line 32: Malformed conditional (${PORT_OPTIONS:MDOCS}) "Makefile", line 36: if-less endif make: fatal errors encountered -- cannot continue With: muji64# diff -u Makefile.original Makefile --- Makefile.original 2013-04-21 13:45:06.0 +0100 +++ Makefile2013-04-21 14:26:34.0 +0100 @@ -29,7 +29,8 @@ .include post-install: -.if ${PORT_OPTIONS:MDOCS} +#.if ${PORT_OPTIONS:MDOCS} +.if !defined(NOPORTDOCS) ${MKDIR} ${DOCSDIR} ${INSTALL_DATA} ${PORTDOCS:S|^|${WRKSRC}/|} ${DOCSDIR} .endif it makes and installs fine. Is this due to the age of my system? muji64# uname -a FreeBSD muji64 9.0-RELEASE-p3 FreeBSD 9.0-RELEASE-p3 #0: Tue Jun 12 02:52:29 UTC 2012 r...@amd64-builder.daemonology.net:/usr/obj/usr/src/sys/GENERIC amd64 Chris ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
FreeBSD ports you maintain which are out of date
Dear port maintainer, The portscout new distfile checker has detected that one or more of your ports appears to be out of date. Please take the opportunity to check each of the ports listed below, and if possible and appropriate, submit/commit an update. If any ports have already been updated, you can safely ignore the entry. You will not be e-mailed again for any of the port/version combinations below. Full details can be found at the following URL: http://portscout.freebsd.org/po...@freebsd.org.html Port| Current version | New version +-+ graphics/geoserver | 2.1.1 | 2.3.1 +-+ If any of the above results are invalid, please check the following page for details on how to improve portscout's detection and selection of distfiles on a per-port basis: http://portscout.freebsd.org/info/portscout-portconfig.txt If wish to stop receiving portscout reminders, please contact portsc...@portscout.freebsd.org Thanks. ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
Re: non-destructive ports/packages update
On Sat, Apr 20, 2013 at 06:03:17PM -0700, Perry Hutchison wrote: > Xin Li wrote: > > > On 4/19/13 11:34 PM, Perry Hutchison wrote: > > > I'm looking for a way to move everything connected with ports and > > > packages aside, so that I can start fresh but with the ability to > > > easily roll it back when things go badly (as they surely will). > > > > > > I have in mind to something like this: > > > > > > # cd /usr > > > # mkdir old > > > # mv ports local old > > > # mkdir ports local > > > # cd /var/db > > > # mkdir old > > > # mv ports pkg old > > > # mkdir ports pkg > > > > > > Is there anything else that needs to be saved before fetching a > > > new ports tree and starting to build things (or install prebuilt > > > packages)? > > > > If you use ZFS, it's possible to take snapshot, then install new > > ports, then if something blows up, you can rollback. > > > > With UFS, it's still possible to take snapshot but rollback is not > > atomic. > > I'm aware of filesystem snapshots, but I only want to checkpoint the > ports and packages, not the whole filesystem -- a rollback needs to > be fast, easy, and obviously correct; preserve the failure logs; and > not undo changes that may have been made elsewhere in the meantime. > (BTW I don't use ZFS: the machine doesn't have enough memory, and to > me ZFS -- especially on 8.x -- doesn't yet seem sufficiently proven.) > > > If you use portmaster, it can save packages (I think portupgrade > > can do it too). But this approach depends on the fact that the > > port is well written, and is not atomic in terms of package set. > > And then a rollback requires re-installing the saved packages, which > is surely slower than moving a few directories and/or files around. > > The question is, what (if anything) else -- besides /usr/ports, > /usr/local, /var/db/ports, and /var/db/pkg -- needs to be checkpointed? Some ports might store "run state" in /var/db/ or a similarly named directory. The thing is, the decision whether to save this and restore it or to keep it across runs actually depends on the port: for database management systems such as MySQL, PostgreSQL, etc, you'll probably want to keep the databases even if the ports themselves are reinstalled, rolled back, restored, whatever. For some other systems, you might want to remove the "current state" information of the version that you are about to replace. G'luck, Peter -- Peter Pentchev r...@ringlet.net r...@freebsd.org p.penc...@storpool.com PGP key:http://people.FreeBSD.org/~roam/roam.key.asc Key fingerprint 2EE7 A7A5 17FC 124C F115 C354 651E EFB0 2527 DF13 Thit sentence is not self-referential because "thit" is not a word. signature.asc Description: Digital signature
Re: FreeBSD unmaintained ports which are currently scheduled for deletion
2013-04-21 17:30 に lini...@freebsd.org さんは書きました: portname: net-im/rubygem-termtter description:Terminal based Twitter client maintainer: d...@freebsd.org deprecated because: Does not work with Ruby 1.9 expiration date:2013-05-02 build errors: none. overview: http://portsmon.FreeBSD.org/portoverview.py?category=net-im&portname=rubygem-termtter I strongly disagree to delete net-im/rubygem-termtter. It is mismarked as BROKEN by portmgr for one year even if it perfectly works with ruby 1.9. -- `whois vmeta.jp | nkf -w` meta ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
Re: CFT: google-earth crash fix (ports/160422 [1])
Hi! > > http://people.freebsd.org/~nox/tmp/google-earth-6.0.3.2197-crashfixp-001.patch Works on 9.1-amd64. -- p...@opsec.eu+49 171 3101372 7 years to go ! ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
Re: CFT: google-earth crash fix (ports/160422 [1])
В Sat, 20 Apr 2013 23:07:17 +0200 Juergen Lock пишет: > > >> > > >> Hi Juergen, > > >> > > >> Thanks for the info and the link. I just tried the patch with > > >> GoogleEarth version 6.0.3.2197 on a 10.0-CURRENT amd64 (clang) > > >> and it also seems to work. > > >> > > >> Perhaps we should try to reintroduce version 6.0.3.2197? I do not > > >> remember the reason(s) for going back to version 5.2.1.1588 ... > > >> > > > Hmm is the distfile for 6.0.3.2197 still public? > > > > I would think so, because I fetched it in the common way, i.e. > > > > fetch http://dl.google.com/earth/client/current/GoogleEarthLinux.bin > > > > > > the relevant part of my Makefile looks like this > > > > PORTNAME= google-earth > > PORTVERSION=6.0.3.2197 > > #PORTREVISION= 1 > > PORTEPOCH= 1 > > CATEGORIES= astro deskutils geography > > MASTER_SITES= http://dl.google.com/earth/client/current/:current > > DISTFILES= GoogleEarthLinux.bin:current > > DIST_SUBDIR=${PORTNAME}/${PORTVERSION} > > EXTRACT_ONLY= # none > > > Ok new patch here: > > > http://people.freebsd.org/~nox/tmp/google-earth-6.0.3.2197-crashfixp-001.patch > > Please give this a good testing everyone, especially if it crashed > for you before... > > Thanx again, :) > Juergen It works for me too. Thanks! ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
FreeBSD ports which are currently marked forbidden
As part of an ongoing effort to reduce the number of problems in the FreeBSD ports system, we periodically notify users about ports that are marked as "forbidden" in their Makefiles. Often, these ports are so marked due to security concerns, such as known exploits. An overview of each port, including errors seen on the build farm, is included below. portname: graphics/linux-tiff forbidden because: Vulnerable since 2004-10-13, http://portaudit.freebsd.org/8816bf3a-7929-11df-bcce-0018f3e2eb82.html build errors: none. overview: http://portsmon.FreeBSD.org/portoverview.py?category=graphics&portname=linux-tiff portname: security/sudosh3 forbidden because: Secunia Advisory SA38292, ISS X-Force sudosh-replay-bo (55903), replay() function buffer overflow. build errors: none. overview: http://portsmon.FreeBSD.org/portoverview.py?category=security&portname=sudosh3 portname: x11-toolkits/linux-pango forbidden because: Vulnerable since 2009-05-13, http://portaudit.freebsd.org/4b172278-3f46-11de-becb-001cc0377035.html build errors: none. overview: http://portsmon.FreeBSD.org/portoverview.py?category=x11-toolkits&portname=linux-pango ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
FreeBSD ports which are currently marked broken
As part of an ongoing effort to reduce the number of problems in the FreeBSD ports system, we periodically notify users of ports that are marked as "broken" in their Makefiles. In many cases these ports are failing to compile on some subset of the FreeBSD build environments. The most common problem is that recent versions of -CURRENT include gcc4.2, which is much stricter than older versions. The next most common problem is that compiles succeed on the i386 architecture (e.g. the common Intel PC), but fail on one or more of the other architectures due to assumptions about things such as size of various types, byte-alignment issues, and so forth. In occasional cases we see that the same port may have different errors in different build environments. The script that runs on the build cluster uses heuristics to try to 'guess' the error type to help you isolate problems, but it is only a rough guide. One more note: on occasion, there are transient build errors seen on the build farm. Unfortunately, there is not yet any way for this algorithm to tell the difference (humans are much, much better at this kind of thing.) The errors are listed below. In the case where the same problem exists on more than one build environment, the URL points to the latest errorlog for that type. (By 'build environment' here we mean 'combination of 7.x/8.x/9.x/-current with target architecture'.) (Note: the dates are included to help you to gauge whether or not the error still applies to the latest version. The program that generates this report is not yet able to determine this automatically.) portname: accessibility/yasr broken because: fails to build with new utmpx build errors: none. overview: http://portsmon.FreeBSD.org/portoverview.py?category=accessibility&portname=yasr portname: audio/gdam broken because: does not build build errors: none. overview: http://portsmon.FreeBSD.org/portoverview.py?category=audio&portname=gdam portname: audio/hydrogen broken because: does not install build errors: none. overview: http://portsmon.FreeBSD.org/portoverview.py?category=audio&portname=hydrogen portname: benchmarks/polygraph31 broken because: does not build build errors: none. overview: http://portsmon.FreeBSD.org/portoverview.py?category=benchmarks&portname=polygraph31 portname: biology/dotter broken because: checksum mismatch build errors: none. overview: http://portsmon.FreeBSD.org/portoverview.py?category=biology&portname=dotter portname: cad/meshlab broken because: does not build build errors: none. overview: http://portsmon.FreeBSD.org/portoverview.py?category=cad&portname=meshlab portname: chinese/big5con broken because: fails to build with new utmpx build errors: none. overview: http://portsmon.FreeBSD.org/portoverview.py?category=chinese&portname=big5con portname: chinese/bitchx broken because: patch reject build errors: none. overview: http://portsmon.FreeBSD.org/portoverview.py?category=chinese&portname=bitchx portname: chinese/cxterm broken because: fails to build with new utmpx build errors: none. overview: http://portsmon.FreeBSD.org/portoverview.py?category=chinese&portname=cxterm portname: chinese/hztty broken because: fails to build with new utmpx build errors: none. overview: http://portsmon.FreeBSD.org/portoverview.py?category=chinese&portname=hztty portname: comms/hso-kmod broken because: does not build with USB2, please try comms/uhso-kmod instead build errors: none. overview: http://portsmon.FreeBSD.org/portoverview.py?category=comms&portname=hso-kmod portname: comms/ib-kmod broken because: does not build build errors: none. overview: http://portsmon.FreeBSD.org/portoverview.py?category=comms&portname=ib-kmod portname: comms/uticom broken because: does not compile build errors: none. overview: http://portsmon.FreeBSD.org/portoverview.py?category=comms&portname=uticom portname: converters/pdf2djvu broken because: does not build build errors: http://pointyhat.freebsd.org/errorlogs/amd64-errorlogs/e.10.20130313090402.pointyhat/pdf2djvu-0.5.11_10.log (Mar 14 00:22:50 UTC 2013) overview: http://portsmon.FreeBSD.org/portoverview.py?category=converters&portname=pdf2djvu portname: databases/drizzle broken because: fails to build build errors: none. overview: http://portsmon.FreeBSD.org/portoverview.py?category=databases&portname=drizzle portname: databases/glom broken because: bad-plist build errors: none. overview: http://portsmon.FreeBSD.org/portoverview.py?category=databases&po
FreeBSD unmaintained ports which are currently marked broken
As part of an ongoing effort to reduce the number of problems in the FreeBSD ports system, we periodically notify users of ports that are marked as "broken" in their Makefiles. In many cases these ports are failing to compile on some subset of the FreeBSD build environments. The most common problem is that recent versions of -CURRENT include gcc4.2, which is much stricter than older versions. The next most common problem is that compiles succeed on the i386 architecture (e.g. the common Intel PC), but fail on one or more of the other architectures due to assumptions about things such as size of various types, byte-alignment issues, and so forth. In occasional cases we see that the same port may have different errors in different build environments. The script that runs on the build cluster uses heuristics to try to 'guess' the error type to help you isolate problems, but it is only a rough guide. One more note: on occasion, there are transient build errors seen on the build farm. Unfortunately, there is not yet any way for this algorithm to tell the difference (humans are much, much better at this kind of thing.) The errors are listed below. In the case where the same problem exists on more than one build environment, the URL points to the latest errorlog for that type. (By 'build environment' here we mean 'combination of 7.x/8.x/9.x/-current with target architecture'.) (Note: the dates are included to help you to gauge whether or not the error still applies to the latest version. The program that generates this report is not yet able to determine this automatically.) portname: biology/dotter broken because: checksum mismatch build errors: none. overview: http://portsmon.FreeBSD.org/portoverview.py?category=biology&portname=dotter portname: chinese/big5con broken because: fails to build with new utmpx build errors: none. overview: http://portsmon.FreeBSD.org/portoverview.py?category=chinese&portname=big5con portname: chinese/bitchx broken because: patch reject build errors: none. overview: http://portsmon.FreeBSD.org/portoverview.py?category=chinese&portname=bitchx portname: chinese/hztty broken because: fails to build with new utmpx build errors: none. overview: http://portsmon.FreeBSD.org/portoverview.py?category=chinese&portname=hztty portname: databases/msql broken because: Broken on FreeBSD 9+ build errors: none. overview: http://portsmon.FreeBSD.org/portoverview.py?category=databases&portname=msql portname: deskutils/libopensync-plugin-python-devel broken because: fails to build with recent libopensync build errors: none. overview: http://portsmon.FreeBSD.org/portoverview.py?category=deskutils&portname=libopensync-plugin-python-devel portname: deskutils/msynctool-devel broken because: fails to build with recent libopensync build errors: none. overview: http://portsmon.FreeBSD.org/portoverview.py?category=deskutils&portname=msynctool-devel portname: deskutils/simpleagenda broken because: does not build build errors: none. overview: http://portsmon.FreeBSD.org/portoverview.py?category=deskutils&portname=simpleagenda portname: devel/arm-rtems-binutils broken because: many issues; see https://www.rtems.org/bugzilla/show_bug.cgi?id=2099 build errors: none. overview: http://portsmon.FreeBSD.org/portoverview.py?category=devel&portname=arm-rtems-binutils portname: devel/arm-rtems-gcc broken because: many issues; see https://www.rtems.org/bugzilla/show_bug.cgi?id=2099 build errors: none. overview: http://portsmon.FreeBSD.org/portoverview.py?category=devel&portname=arm-rtems-gcc portname: devel/arm-rtems-gdb broken because: many issues; see https://www.rtems.org/bugzilla/show_bug.cgi?id=2099 build errors: none. overview: http://portsmon.FreeBSD.org/portoverview.py?category=devel&portname=arm-rtems-gdb portname: devel/dsss broken because: does not build build errors: none. overview: http://portsmon.FreeBSD.org/portoverview.py?category=devel&portname=dsss portname: devel/i386-rtems-binutils broken because: many issues; see https://www.rtems.org/bugzilla/show_bug.cgi?id=2099 build errors: none. overview: http://portsmon.FreeBSD.org/portoverview.py?category=devel&portname=i386-rtems-binutils portname: devel/i386-rtems-gcc broken because: many issues; see https://www.rtems.org/bugzilla/show_bug.cgi?id=2099 build errors: none. overview: http://portsmon.FreeBSD.org/portoverview.py?category=devel&portname=i386-rtems-gcc portname: devel/i386-rtem
Re: non-destructive ports/packages update
On 21 Apr 2013 02:13, "Perry Hutchison" wrote: > > Xin Li wrote: > > > On 4/19/13 11:34 PM, Perry Hutchison wrote: > > > I'm looking for a way to move everything connected with ports and > > > packages aside, so that I can start fresh but with the ability to > > > easily roll it back when things go badly (as they surely will). > > > > > > I have in mind to something like this: > > > > > > # cd /usr > > > # mkdir old > > > # mv ports local old > > > # mkdir ports local > > > # cd /var/db > > > # mkdir old > > > # mv ports pkg old > > > # mkdir ports pkg > > > > > > Is there anything else that needs to be saved before fetching a > > > new ports tree and starting to build things (or install prebuilt > > > packages)? > > > > If you use ZFS, it's possible to take snapshot, then install new > > ports, then if something blows up, you can rollback. > > > > With UFS, it's still possible to take snapshot but rollback is not > > atomic. > > I'm aware of filesystem snapshots, but I only want to checkpoint the > ports and packages, not the whole filesystem -- a rollback needs to > be fast, easy, and obviously correct; preserve the failure logs; and > not undo changes that may have been made elsewhere in the meantime. > (BTW I don't use ZFS: the machine doesn't have enough memory, and to > me ZFS -- especially on 8.x -- doesn't yet seem sufficiently proven.) > > > If you use portmaster, it can save packages (I think portupgrade > > can do it too). But this approach depends on the fact that the > > port is well written, and is not atomic in terms of package set. > > And then a rollback requires re-installing the saved packages, which > is surely slower than moving a few directories and/or files around. > > The question is, what (if anything) else -- besides /usr/ports, > /usr/local, /var/db/ports, and /var/db/pkg -- needs to be checkpointed? Are you installing any drivers? You may need to back up /boot too. Chris ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"