From ivoras at freebsd.org Mon Sep 1 09:42:25 2014 From: ivoras at freebsd.org (Ivan Voras) Date: Mon, 1 Sep 2014 11:41:42 +0200 Subject: pkg crashing? Message-ID: Just a quick note: # pkg upgrade postgresql90-client-9.0.17 postgresql90-contrib-9.0.16 postgresql90-server-9.0.16 Updating FreeBSD repository catalogue... FreeBSD repository is up-to-date. All repositories are up-to-date. pkg: No packages available to upgrade matching 'postgresql90-contrib-9.0.16' have been found in the repositories Child process pid=86122 terminated abnormally: Segmentation fault: 11 Maybe something's missing when a package cannot be upgraded. From rodrigc at freebsd.org Mon Sep 1 10:01:46 2014 From: rodrigc at freebsd.org (Craig Rodrigues) Date: Mon, 1 Sep 2014 03:01:43 -0700 Subject: pkg crashing? In-Reply-To: References: Message-ID: On Mon, Sep 1, 2014 at 2:41 AM, Ivan Voras wrote: > Just a quick note: > > # pkg upgrade postgresql90-client-9.0.17 postgresql90-contrib-9.0.16 > postgresql90-server-9.0.16 > Updating FreeBSD repository catalogue... > FreeBSD repository is up-to-date. > All repositories are up-to-date. > pkg: No packages available to upgrade matching > 'postgresql90-contrib-9.0.16' have been found in the repositories > Child process pid=86122 terminated abnormally: Segmentation fault: 11 > > Maybe something's missing when a package cannot be upgraded. What version of pkg are you running? Next time provide the version of pkg that you are running, when reporting a problem. With pkg 1.3.7, I don't get that crash. I get this: # pkg upgrade postgresql90-client-9.0.17 postgresql90-contrib-9.0.16 Updating FreeBSD repository catalogue... FreeBSD repository is up-to-date. All repositories are up-to-date. pkg: postgresql90-client-9.0.17 is not installed, therefore upgrade is impossible pkg: No packages available to upgrade matching 'postgresql90-client-9.0.17' have been found in the repositories # -- Craig From ivoras at freebsd.org Mon Sep 1 10:20:34 2014 From: ivoras at freebsd.org (Ivan Voras) Date: Mon, 1 Sep 2014 12:19:51 +0200 Subject: pkg crashing? In-Reply-To: References: Message-ID: On 1 September 2014 12:01, Craig Rodrigues wrote: > On Mon, Sep 1, 2014 at 2:41 AM, Ivan Voras wrote: >> Just a quick note: >> >> # pkg upgrade postgresql90-client-9.0.17 postgresql90-contrib-9.0.16 >> postgresql90-server-9.0.16 >> Updating FreeBSD repository catalogue... >> FreeBSD repository is up-to-date. >> All repositories are up-to-date. >> pkg: No packages available to upgrade matching >> 'postgresql90-contrib-9.0.16' have been found in the repositories >> Child process pid=86122 terminated abnormally: Segmentation fault: 11 >> >> Maybe something's missing when a package cannot be upgraded. > > What version of pkg are you running? > Next time provide the version of pkg that you are running, when > reporting a problem. > > With pkg 1.3.7, I don't get that crash. I get this: > > # pkg upgrade postgresql90-client-9.0.17 postgresql90-contrib-9.0.16 > Updating FreeBSD repository catalogue... > FreeBSD repository is up-to-date. > All repositories are up-to-date. > pkg: postgresql90-client-9.0.17 is not installed, therefore upgrade is > impossible > pkg: No packages available to upgrade matching > 'postgresql90-client-9.0.17' have been found in the repositories > # I don't force pkg version so it's always the latest, 1.3.7. It does seem that it could be a local issue, I have some bad experiences with pkg recently. What's up with pkg and dependencies lately? I've just installed libreoffice on this machine which had icu52 installed, and the install passed without error AND without replacing icu52 with icu53, so the program naturally crashed with Shared object "libicuuc.so.53" not found, required by "libcomphelper.so" I've just upgraded icu manually (with "pkg upgrade icu"), but it *didn't upgrade any packages depending on it*, so I'm now left with binaries depending on a missing library: # ldd `which vmtoolsd` vmtoolsd: libvmtools.so.0 => /usr/local/lib/libvmtools.so.0 (0x800823000) libkvm.so.5 => /lib/libkvm.so.5 (0x800a9b000) libcrypt.so.5 => /lib/libcrypt.so.5 (0x800ca3000) libthr.so.3 => /lib/libthr.so.3 (0x800ec2000) libgmodule-2.0.so.0 => /usr/local/lib/libgmodule-2.0.so.0 (0x8010e5000) libgobject-2.0.so.0 => /usr/local/lib/libgobject-2.0.so.0 (0x8012e8000) libffi.so.6 => /usr/local/lib/libffi.so.6 (0x801535000) libgthread-2.0.so.0 => /usr/local/lib/libgthread-2.0.so.0 (0x80173c000) libglib-2.0.so.0 => /usr/local/lib/libglib-2.0.so.0 (0x80193d000) libpcre.so.3 => /usr/local/lib/libpcre.so.3 (0x801c35000) libintl.so.9 => /usr/local/lib/libintl.so.9 (0x801e99000) libiconv.so.3 => /usr/local/lib/libiconv.so.3 (0x8020a3000) libicui18n.so.52 => not found (0) libicuuc.so.52 => not found (0) libicudata.so.52 => not found (0) libstdc++.so.6 => /usr/lib/libstdc++.so.6 (0x80239f000) libm.so.5 => /lib/libm.so.5 (0x8026a6000) libgcc_s.so.1 => /lib/libgcc_s.so.1 (0x8028c7000) libc.so.7 => /lib/libc.so.7 (0x802ad5000) libicui18n.so.52 => not found (0) libicuuc.so.52 => not found (0) libicudata.so.52 => not found (0) # pkg which `which vmtoolsd` /usr/local/bin/vmtoolsd was installed by package open-vm-tools-nox11-1280544_4,1 # pkg search open-vm-tools open-vm-tools-1280544_6,1 open-vm-tools-nox11-1280544_6,1 # pkg info|grep open-vm open-vm-tools-nox11-1280544_4,1 Open VMware tools for FreeBSD VMware guests # pkg -v 1.3.7 # pkg upgrade open-vm-tools-nox11 Updating FreeBSD repository catalogue... FreeBSD repository is up-to-date. All repositories are up-to-date. Updating database digests format: 100% pkg: plexhometheater has a missing dependency: lame The following 1 packages will be affected (of 0 checked): Installed packages to be UPGRADED: open-vm-tools-nox11: 1280544_4,1 -> 1280544_6,1 The operation will free 5 kB. 431 kB to be downloaded. Proceed with this action? [y/N]:y Fetching open-vm-tools-nox11-1280544_6,1.txz: 100% 431 kB 442.1k/s 00:01 Checking integrity... done (0 conflicting) [1/1] Upgrading open-vm-tools-nox11 from 1280544_4,1 to 1280544_6,1: 100% # ldd `which vmtoolsd` /usr/local/bin/vmtoolsd: libvmtools.so.0 => /usr/local/lib/libvmtools.so.0 (0x800823000) libgmodule-2.0.so.0 => /usr/local/lib/libgmodule-2.0.so.0 (0x800a9b000) libgobject-2.0.so.0 => /usr/local/lib/libgobject-2.0.so.0 (0x800c9e000) libffi.so.6 => /usr/local/lib/libffi.so.6 (0x800eeb000) libgthread-2.0.so.0 => /usr/local/lib/libgthread-2.0.so.0 (0x8010f2000) libglib-2.0.so.0 => /usr/local/lib/libglib-2.0.so.0 (0x8012f3000) libiconv.so.3 => /usr/local/lib/libiconv.so.3 (0x8015eb000) libpcre.so.3 => /usr/local/lib/libpcre.so.3 (0x8018e7000) libintl.so.9 => /usr/local/lib/libintl.so.9 (0x801b4b000) libkvm.so.5 => /lib/libkvm.so.5 (0x801d55000) libcrypt.so.5 => /lib/libcrypt.so.5 (0x801f5d000) libthr.so.3 => /lib/libthr.so.3 (0x80217c000) libicui18n.so.53 => /usr/local/lib/libicui18n.so.53 (0x80239f000) libicuuc.so.53 => /usr/local/lib/libicuuc.so.53 (0x802813000) libicudata.so.53 => /usr/local/lib/libicudata.so.53 (0x802ba1000) libstdc++.so.6 => /usr/lib/libstdc++.so.6 (0x80428b000) libm.so.5 => /lib/libm.so.5 (0x804592000) libgcc_s.so.1 => /lib/libgcc_s.so.1 (0x8047b3000) libc.so.7 => /lib/libc.so.7 (0x8049c1000) This is the same type of problem I've complained in my post about php - but this is a completely unrelated system. From rodrigc at freebsd.org Mon Sep 1 16:45:28 2014 From: rodrigc at freebsd.org (Craig Rodrigues) Date: Mon, 1 Sep 2014 09:45:25 -0700 Subject: pkg crashing? In-Reply-To: References: Message-ID: On Mon, Sep 1, 2014 at 3:19 AM, Ivan Voras wrote: > I don't force pkg version so it's always the latest, 1.3.7. It does > seem that it could be a local issue, I have some bad experiences with > pkg recently. OK. You should still report the version every time you report a problem, because pkg development is fast moving. In the past few months, there are a lot of minor releases which fix problems (and break a few things along the way :). It is also useful to provide information about your system such as if you upgraded only with packages from pkg.freebsd.org, or if you built any ports yourself. You might want to see if doing: pkg update -f pkg upgrade -y gets your system into a working state. -- Craig From bapt at FreeBSD.org Mon Sep 1 19:55:26 2014 From: bapt at FreeBSD.org (Baptiste Daroussin) Date: Mon, 1 Sep 2014 21:55:20 +0200 Subject: [HEADSUP] pkg(8) is now the only package management tool Message-ID: <20140901195520.GB77917@ivaldir.etoilebsd.net> Hi all, The ports tree has been modified to only support pkg(8) as package management system for all supported version of FreeBSD. if you were still using pkg_install (pkg_* tools) you will have to upgrade your system. The simplest way is cd /usr/ports/ports-mgmt/pkg make install then run pkg2ng You will have lots of warning, don't be scared, they are expected, pkg_* databases used to get easily mangled. pkg2ng is most of the time able to deal with it. If however you encounter a problem then please report to pkg at FreeBSD.org A tag has been applied to the ports tree if you need to get the latest ports tree before the EOL of pkg_install: https://svn.FreeBSD.org/ports/tags/PKG_INSTALL_EOL A branch has been created if some committers want to provides updates on the for pkg_install users: https://svn.FreeBSD.org/ports/branches/pkg_install Please note that this branch is not officially maintained and that we strongly recommend that you do migrate to pkg(8) Best regards, Bapt on behalf of portmgr -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 181 bytes Desc: not available URL: From michelle at sorbs.net Tue Sep 2 00:19:30 2014 From: michelle at sorbs.net (Michelle Sullivan) Date: Tue, 02 Sep 2014 02:19:19 +0200 Subject: [HEADSUP] pkg(8) is now the only package management tool In-Reply-To: <20140901195520.GB77917@ivaldir.etoilebsd.net> References: <20140901195520.GB77917@ivaldir.etoilebsd.net> Message-ID: <54050D07.4010404@sorbs.net> Baptiste Daroussin wrote: > Hi all, > > The ports tree has been modified to only support pkg(8) as package management > system for all supported version of FreeBSD. > > if you were still using pkg_install (pkg_* tools) you will have to upgrade your > system. > > The simplest way is > cd /usr/ports/ports-mgmt/pkg > make install > then run > pkg2ng > > You will have lots of warning, don't be scared, they are expected, pkg_* > databases used to get easily mangled. pkg2ng is most of the time able to deal > with it. > > If however you encounter a problem then please report to pkg at FreeBSD.org > > A tag has been applied to the ports tree if you need to get the latest ports > tree before the EOL of pkg_install: > https://svn.FreeBSD.org/ports/tags/PKG_INSTALL_EOL > > A branch has been created if some committers want to provides updates on the > for pkg_install users: > https://svn.FreeBSD.org/ports/branches/pkg_install > > Please note that this branch is not officially maintained and that we strongly > recommend that you do migrate to pkg(8) > > Best regards, > Bapt on behalf of portmgr > And for the portsnap users? -- Michelle Sullivan http://www.mhix.org/ From bugzilla-noreply at freebsd.org Tue Sep 2 01:14:33 2014 From: bugzilla-noreply at freebsd.org (bugzilla-noreply at freebsd.org) Date: Tue, 02 Sep 2014 01:14:33 +0000 Subject: [Bug 193195] lang/ruby21 deleted lang/ruby19 In-Reply-To: References: Message-ID: https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=193195 Steve Wills changed: What |Removed |Added ---------------------------------------------------------------------------- Status|In Discussion |Issue Resolved Resolution|--- |Unable to Reproduce -- You are receiving this mail because: You are on the CC list for the bug. From sfourman at gmail.com Tue Sep 2 01:39:36 2014 From: sfourman at gmail.com (Sam Fourman Jr.) Date: Mon, 1 Sep 2014 20:39:34 -0500 Subject: [HEADSUP] pkg(8) is now the only package management tool In-Reply-To: <54050D07.4010404@sorbs.net> References: <20140901195520.GB77917@ivaldir.etoilebsd.net> <54050D07.4010404@sorbs.net> Message-ID: > > And for the portsnap users? > > In short, this change doesn't directly effect portsnap users. Portsnap is a tool that used to obtain a copy of the ports tree. Portsnap is only one way, another way to get a copy of the ports tree is by using subversion and checking it out by using the svn command. pkg(8) is a package management tool, and to make use of most packages having a copy of the ports tree is not required. > > > -- > Michelle Sullivan > http://www.mhix.org/ > > _______________________________________________ > freebsd-ports at freebsd.org mailing list > http://lists.freebsd.org/mailman/listinfo/freebsd-ports > To unsubscribe, send any mail to "freebsd-ports-unsubscribe at freebsd.org" > -- Sam Fourman Jr. From michelle at sorbs.net Tue Sep 2 01:51:35 2014 From: michelle at sorbs.net (Michelle Sullivan) Date: Tue, 02 Sep 2014 03:51:31 +0200 Subject: [HEADSUP] pkg(8) is now the only package management tool In-Reply-To: References: <20140901195520.GB77917@ivaldir.etoilebsd.net> <54050D07.4010404@sorbs.net> Message-ID: <540522A3.9050506@sorbs.net> Sam Fourman Jr. wrote: >> And for the portsnap users? >> >> >> > In short, this change doesn't directly effect portsnap users. > Sure about that? > Portsnap is a tool that used to obtain a copy of the ports tree. > try this: portsnap fetch update && cd /usr/ports/ports-mgmt/pkg && make install If you *haven't* install pkg first... > Portsnap is only one way, another way to get a copy of the ports tree is by > using subversion and checking it out by using the svn command. > Not much good if you haven't installed svn already... > pkg(8) is a package management tool, and to make use of most packages > having a copy of the ports tree is not required. > > Correct, take a 9.2 install disk, install it, portsnap and then install pkg on it... Oh wait, you can't.. pkg_install is broken, and 9.2 install disks don't have pkg in the BaseOS.... -- Michelle Sullivan http://www.mhix.org/ From aberg010 at my.hennepintech.edu Tue Sep 2 02:17:07 2014 From: aberg010 at my.hennepintech.edu (Andrew Berg) Date: Mon, 1 Sep 2014 21:16:49 -0500 Subject: [HEADSUP] pkg(8) is now the only package management tool In-Reply-To: <540522A3.9050506@sorbs.net> References: <20140901195520.GB77917@ivaldir.etoilebsd.net> <54050D07.4010404@sorbs.net> <540522A3.9050506@sorbs.net> Message-ID: <54052891.5000104@my.hennepintech.edu> On 2014.09.01 20:51, Michelle Sullivan wrote: >>> And for the portsnap users? >>> >> In short, this change doesn't directly effect portsnap users. >> > Sure about that? I'm sure of it. Your issue is with the tree itself, not the tool used to fetch it. > Correct, take a 9.2 install disk, install it, portsnap and then install > pkg on it... Oh wait, you can't.. pkg_install is broken, and 9.2 > install disks don't have pkg in the BaseOS.... Use the ports tree tarball included, or fetch it (either during or after installation). It is not impossible to get an old version of the ports tree with only the 9.2 base system. I don't see how this is anything more than an inconvenience. Also, 9.3 is out and the 9.2 EOL is not far away. Not sure why you would be doing a new install with 9.2. From julian at freebsd.org Tue Sep 2 02:20:57 2014 From: julian at freebsd.org (Julian Elischer) Date: Mon, 01 Sep 2014 19:20:49 -0700 Subject: [HEADSUP] pkg(8) is now the only package management tool In-Reply-To: References: <20140901195520.GB77917@ivaldir.etoilebsd.net> <54050D07.4010404@sorbs.net> Message-ID: <54052981.7010502@freebsd.org> On 9/1/14, 6:39 PM, Sam Fourman Jr. wrote: >> And for the portsnap users? >> >> > In short, this change doesn't directly effect portsnap users. > > Portsnap is a tool that used to obtain a copy of the ports tree. > > Portsnap is only one way, another way to get a copy of the ports tree is by > using subversion and checking it out by using the svn command. > > pkg(8) is a package management tool, and to make use of most packages > having a copy of the ports tree is not required. But it is if you don't want the options that a pkg is built with. We need to do a lot of pkg munging for that reason, generating our own versions (which is ok, that's not a complaint, just a fact of life). I've warmed to pkg after using it a bit, and many of its initial shortcomings have been fixed. But one thing I'd like to request (a very minor thing).. Could the packing list have some newlines inserted into it to make it more humanly readable? Our old tools for auditing and controlling (old style) packages would print out that information. The new tools we need to write will need to do similar. We did an experiment at work here and wrote a small script that parsed it and then rewrote it back to the package with newlines added and pkg handled it just fine, so it should be a very minor thing to add some newlines when generating it in the first place. I don't think anything else needs to be changed. > >> >> -- >> Michelle Sullivan >> http://www.mhix.org/ >> >> _______________________________________________ >> freebsd-ports at freebsd.org mailing list >> http://lists.freebsd.org/mailman/listinfo/freebsd-ports >> To unsubscribe, send any mail to "freebsd-ports-unsubscribe at freebsd.org" >> > > From fullermd at over-yonder.net Tue Sep 2 02:27:04 2014 From: fullermd at over-yonder.net (Matthew D. Fuller) Date: Mon, 1 Sep 2014 21:26:55 -0500 Subject: [HEADSUP] pkg(8) is now the only package management tool In-Reply-To: <540522A3.9050506@sorbs.net> References: <20140901195520.GB77917@ivaldir.etoilebsd.net> <54050D07.4010404@sorbs.net> <540522A3.9050506@sorbs.net> Message-ID: <20140902022655.GI43581@over-yonder.net> On Tue, Sep 02, 2014 at 03:51:31AM +0200 I heard the voice of Michelle Sullivan, and lo! it spake thus: > > Correct, take a 9.2 install disk, install it, portsnap and then > install pkg on it... Oh wait, you can't.. pkg_install is broken, > and 9.2 install disks don't have pkg in the BaseOS.... So what? The pkg port uses _ITSELF_ to register. The "pkg" in the base system isn't pkg, it just a bootstrap to fetch the pkg pkg (which them uses itself to register too). If you're using the pkg _PORT_, it's not even involved in the first place. -- Matthew Fuller (MF4839) | fullermd at over-yonder.net Systems/Network Administrator | http://www.over-yonder.net/~fullermd/ On the Internet, nobody can hear you scream. From michelle at sorbs.net Tue Sep 2 02:27:46 2014 From: michelle at sorbs.net (Michelle Sullivan) Date: Tue, 02 Sep 2014 04:27:41 +0200 Subject: [HEADSUP] pkg(8) is now the only package management tool In-Reply-To: <54052891.5000104@my.hennepintech.edu> References: <20140901195520.GB77917@ivaldir.etoilebsd.net> <54050D07.4010404@sorbs.net> <540522A3.9050506@sorbs.net> <54052891.5000104@my.hennepintech.edu> Message-ID: <54052B1D.3040607@sorbs.net> Andrew Berg wrote: > On 2014.09.01 20:51, Michelle Sullivan wrote: > >>>> And for the portsnap users? >>>> >>>> >>> In short, this change doesn't directly effect portsnap users. >>> >>> >> Sure about that? >> > I'm sure of it. Your issue is with the tree itself, not the tool used to fetch it. > > >> Correct, take a 9.2 install disk, install it, portsnap and then install >> pkg on it... Oh wait, you can't.. pkg_install is broken, and 9.2 >> install disks don't have pkg in the BaseOS.... >> > Use the ports tree tarball included, or fetch it (either during or after > installation). It is not impossible to get an old version of the ports tree > with only the 9.2 base system. I don't see how this is anything more than an > inconvenience. > Actually it's an inconvenience for someone like me and you. Not for many freebsd users, and certainly not for me 6 months ago if I hadn't been writing my own ports.... oh and what was it, 1.3.6 -> 1.3.7? broke shit... (badly) ... > Also, 9.3 is out and the 9.2 EOL is not far away. Not sure why you would be > doing a new install with 9.2. > Try getting yourself a FreeBSD server at Softlayer... They still install 7.x for Christ's sake (amongst others - but last time I checked, on new servers, 8.4, 9.0, 9.1, 10.0*) * the 10.0 is the original release, completely unpatched. Look I'm not saying the change isn't for the better, I'm saying not supporting older systems until you're sure 99% of the userbase is upgraded is not a bad thing, what I am saying is deliberately breaking all older systems (some without *major pain*) when the new system has just had a major issue, and not everyone had time to upgrade is a *bad thing* ... (not had time - because an EOL message is not a 'It will not work after this date' message it is a 'you're unsupported after this date and things *might* not work as expected' - even Windows XP didn't got his root... they EOL'd XP, then they stated for 2 or was it 3 years, that after 'x' date there would not be any new security patches... but you can still get software for XP, some is even patched... FreeBSD... Sept 1, 2014, you're not on pkg, you're fucked.) -- Michelle Sullivan http://www.mhix.org/ From julian at freebsd.org Tue Sep 2 02:40:03 2014 From: julian at freebsd.org (Julian Elischer) Date: Mon, 01 Sep 2014 19:39:54 -0700 Subject: [HEADSUP] pkg(8) is now the only package management tool In-Reply-To: <54052891.5000104@my.hennepintech.edu> References: <20140901195520.GB77917@ivaldir.etoilebsd.net> <54050D07.4010404@sorbs.net> <540522A3.9050506@sorbs.net> <54052891.5000104@my.hennepintech.edu> Message-ID: <54052DFA.4030808@freebsd.org> On 9/1/14, 7:16 PM, Andrew Berg wrote: > On 2014.09.01 20:51, Michelle Sullivan wrote: >>>> And for the portsnap users? >>>> >>> In short, this change doesn't directly effect portsnap users. >>> >> Sure about that? > I'm sure of it. Your issue is with the tree itself, not the tool used to fetch it. > >> Correct, take a 9.2 install disk, install it, portsnap and then install >> pkg on it... Oh wait, you can't.. pkg_install is broken, and 9.2 >> install disks don't have pkg in the BaseOS.... > Use the ports tree tarball included, or fetch it (either during or after > installation). It is not impossible to get an old version of the ports tree > with only the 9.2 base system. I don't see how this is anything more than an > inconvenience. > Also, 9.3 is out and the 9.2 EOL is not far away. Not sure why you would be > doing a new install with 9.2. sigh.. when are we as a project, all going to learn that reality in business is that you often need to install stuff that is old. Its not always your choice. The custommers require it.. You should try arguing with someone like Bank of Americas security and operations department some day about whether they want to suddenly upgrade 300 machines for no real reason (from their perspective). On that topic, 10.0 is slightly broken from that perspective because as you install it, it upgrades pkg to a new version that was not in 10.0, so you can no longer build a 10.0 machine that matches the 10.0 machines you installed at the custommer site when 10.0 first came out, that they qualified as acceptible.. Well you MAY get the mostly same result, but the 'pkg' you have is a different one so the image checks out as different' (Imaginary hooters sound and theoretical security alerts trigger etc.) (oh and it interacts badly with the installer designed to run with the previous version.. The first part of the install works fine, and then half way through the install, things go strange when pkg upgrades itself.) 10.0 is past but we should think about how to prevent that in 10.1 etc. I guess the pkg config file in the install needs to be locked down to the release until the install is completed. We should make sure the base install only installs the pkg in the release and doesn't upgrade itself without asking first... (luckily that last issue doesn't affect most business customers who use their own install schemes). > _______________________________________________ > freebsd-current at freebsd.org mailing list > http://lists.freebsd.org/mailman/listinfo/freebsd-current > To unsubscribe, send any mail to "freebsd-current-unsubscribe at freebsd.org" > > From michelle at sorbs.net Tue Sep 2 03:00:01 2014 From: michelle at sorbs.net (Michelle Sullivan) Date: Tue, 02 Sep 2014 04:59:56 +0200 Subject: [HEADSUP] pkg(8) is now the only package management tool In-Reply-To: <54052DFA.4030808@freebsd.org> References: <20140901195520.GB77917@ivaldir.etoilebsd.net> <54050D07.4010404@sorbs.net> <540522A3.9050506@sorbs.net> <54052891.5000104@my.hennepintech.edu> <54052DFA.4030808@freebsd.org> Message-ID: <540532AC.2070508@sorbs.net> Julian Elischer wrote: > > You should try arguing with someone like Bank of Americas security and > operations > department You work for the same company as me? > some day about whether they want to suddenly upgrade 300 machines > for no real reason (from their perspective). > -- Michelle Sullivan http://www.mhix.org/ From aberg010 at my.hennepintech.edu Tue Sep 2 03:03:24 2014 From: aberg010 at my.hennepintech.edu (Andrew Berg) Date: Mon, 1 Sep 2014 22:02:59 -0500 Subject: [HEADSUP] pkg(8) is now the only package management tool In-Reply-To: <54052B1D.3040607@sorbs.net> References: <20140901195520.GB77917@ivaldir.etoilebsd.net> <54050D07.4010404@sorbs.net> <540522A3.9050506@sorbs.net> <54052891.5000104@my.hennepintech.edu> <54052B1D.3040607@sorbs.net> Message-ID: <54053363.2030606@my.hennepintech.edu> On 2014.09.01 21:27, Michelle Sullivan wrote: > Actually it's an inconvenience for someone like me and you. Not for > many freebsd users, and certainly not for me 6 months ago if I hadn't > been writing my own ports.... oh and what was it, 1.3.6 -> 1.3.7? broke > shit... (badly) ... There were instructions for upgrading 1.3.6 to 1.3.7 alongside a notice that things would not be good if the instructions were not followed and an explanation of the issue. I think these kinds of notices need to reach more people, but of course, that is easier said than done. BTW, from what I have observed, 1.3.x issues have affected Poudriere users the most, binary package users a bit less (but still significantly), and pure ports users very little. >> Also, 9.3 is out and the 9.2 EOL is not far away. Not sure why you would be >> doing a new install with 9.2. >> > Try getting yourself a FreeBSD server at Softlayer... They still > install 7.x for Christ's sake (amongst others - but last time I checked, > on new servers, 8.4, 9.0, 9.1, 10.0*) Fair enough. > (not had time - because an EOL message is not a 'It will not > work after this date' message it is a 'you're unsupported after this > date and things *might* not work as expected' No, it means "we're not supporting this any more, so we don't care if there are new vulnerabilities or things stop working". I'm not going to dictate to other people what their upgrade schedule should be, but anyone running unsupported versions of software should not have any expectation that the ecosystem around it will be accommodating. The ports tree already requires a lot work to make sure everything works on supported versions of FreeBSD, and I see no reason whatsoever for anyone to put effort into making it work on EOL versions. From aberg010 at my.hennepintech.edu Tue Sep 2 03:03:29 2014 From: aberg010 at my.hennepintech.edu (Andrew Berg) Date: Mon, 1 Sep 2014 22:03:14 -0500 Subject: [HEADSUP] pkg(8) is now the only package management tool In-Reply-To: <54052DFA.4030808@freebsd.org> References: <20140901195520.GB77917@ivaldir.etoilebsd.net> <54050D07.4010404@sorbs.net> <540522A3.9050506@sorbs.net> <54052891.5000104@my.hennepintech.edu> <54052DFA.4030808@freebsd.org> Message-ID: <54053372.6020009@my.hennepintech.edu> On 2014.09.01 21:39, Julian Elischer wrote: > sigh.. when are we as a project, all going to learn that reality in > business is > that you often need to install stuff that is old. Its not always your > choice. > The custommers require it.. > You should try arguing with someone like Bank of Americas security and > operations > department some day about whether they want to suddenly upgrade 300 > machines > for no real reason (from their perspective). FreeBSD minor version upgrades are meant to be non-disruptive. However, I will admit that I have not performed any such upgrades in a critical environment, so if you think they are disruptive, please enlighten me with the details. Also, there are options out there for getting support for extended periods if you need it. Some companies are built around providing support for things that the original developers have long abandoned because some businesses need it. From michelle at sorbs.net Tue Sep 2 03:09:11 2014 From: michelle at sorbs.net (Michelle Sullivan) Date: Tue, 02 Sep 2014 05:09:07 +0200 Subject: [HEADSUP] pkg(8) is now the only package management tool In-Reply-To: <54053363.2030606@my.hennepintech.edu> References: <20140901195520.GB77917@ivaldir.etoilebsd.net> <54050D07.4010404@sorbs.net> <540522A3.9050506@sorbs.net> <54052891.5000104@my.hennepintech.edu> <54052B1D.3040607@sorbs.net> <54053363.2030606@my.hennepintech.edu> Message-ID: <540534D3.603@sorbs.net> Andrew Berg wrote: > On 2014.09.01 21:27, Michelle Sullivan wrote: > >> Actually it's an inconvenience for someone like me and you. Not for >> many freebsd users, and certainly not for me 6 months ago if I hadn't >> been writing my own ports.... oh and what was it, 1.3.6 -> 1.3.7? broke >> shit... (badly) ... >> > There were instructions for upgrading 1.3.6 to 1.3.7 alongside a notice that > things would not be good if the instructions were not followed and an > explanation of the issue. I think these kinds of notices need to reach more > people, but of course, that is easier said than done. > BTW, from what I have observed, 1.3.x issues have affected Poudriere users the > most, binary package users a bit less (but still significantly), and pure ports > users very little. > I am a poudriere user... across 8.4, 9.0, 9.1, 9.2, 9.3, 10.0 on both i386 and amd64 :/ > >>> Also, 9.3 is out and the 9.2 EOL is not far away. Not sure why you would be >>> doing a new install with 9.2. >>> >>> >> Try getting yourself a FreeBSD server at Softlayer... They still >> install 7.x for Christ's sake (amongst others - but last time I checked, >> on new servers, 8.4, 9.0, 9.1, 10.0*) >> > Fair enough. > > >> (not had time - because an EOL message is not a 'It will not >> work after this date' message it is a 'you're unsupported after this >> date and things *might* not work as expected' >> > No, it means "we're not supporting this any more, so we don't care if there are > new vulnerabilities or things stop working". I'm not going to dictate to other > people what their upgrade schedule should be, but anyone running unsupported > versions of software should not have any expectation that the ecosystem around > it will be accommodating. > That's my point - there was a patch waiting to submit that knowingly broke pkg_install at midnight on the day after the EOL... the EOL shouldn't be an EOL - because it was really a 'portsnap after this date before you upgrade and you're screwed it won't work any more at all...' > The ports tree already requires a lot work to make sure everything works on > supported versions of FreeBSD, and I see no reason whatsoever for anyone to put > effort into making it work on EOL versions. > Some of us have production systems that span 6.0->10.0 (and most version in between) and are fighting fires with minimal help just trying to keep ahead.... -- Michelle Sullivan http://www.mhix.org/ From yaneurabeya at gmail.com Tue Sep 2 03:14:25 2014 From: yaneurabeya at gmail.com (yaneurabeya at gmail.com) Date: Mon, 1 Sep 2014 20:14:22 -0700 Subject: [HEADSUP] pkg(8) is now the only package management tool In-Reply-To: <54053363.2030606@my.hennepintech.edu> References: <20140901195520.GB77917@ivaldir.etoilebsd.net> <54050D07.4010404@sorbs.net> <540522A3.9050506@sorbs.net> <54052891.5000104@my.hennepintech.edu> <54052B1D.3040607@sorbs.net> <54053363.2030606@my.hennepintech.edu> Message-ID: <8E432427-B3A8-4BF7-B54F-DE9B8CA268DC@gmail.com> On Sep 1, 2014, at 20:02, Andrew Berg wrote: > On 2014.09.01 21:27, Michelle Sullivan wrote: >> Actually it's an inconvenience for someone like me and you. Not for >> many freebsd users, and certainly not for me 6 months ago if I hadn't >> been writing my own ports.... oh and what was it, 1.3.6 -> 1.3.7? broke >> shit... (badly) ... > There were instructions for upgrading 1.3.6 to 1.3.7 alongside a notice that > things would not be good if the instructions were not followed and an > explanation of the issue. I think these kinds of notices need to reach more > people, but of course, that is easier said than done. > BTW, from what I have observed, 1.3.x issues have affected Poudriere users the > most, binary package users a bit less (but still significantly), and pure ports > users very little. > >>> Also, 9.3 is out and the 9.2 EOL is not far away. Not sure why you would be >>> doing a new install with 9.2. >>> >> Try getting yourself a FreeBSD server at Softlayer... They still >> install 7.x for Christ's sake (amongst others - but last time I checked, >> on new servers, 8.4, 9.0, 9.1, 10.0*) > Fair enough. > >> (not had time - because an EOL message is not a 'It will not >> work after this date' message it is a 'you're unsupported after this >> date and things *might* not work as expected' > No, it means "we're not supporting this any more, so we don't care if there are > new vulnerabilities or things stop working". I'm not going to dictate to other > people what their upgrade schedule should be, but anyone running unsupported > versions of software should not have any expectation that the ecosystem around > it will be accommodating. > The ports tree already requires a lot work to make sure everything works on > supported versions of FreeBSD, and I see no reason whatsoever for anyone to put > effort into making it work on EOL versions. Installing pkgng on FreeBSD 7.x isn?t impossible, but it does require jumping through some hoops because xz not being present until 8.x. These directions aren?t complete (welcome to feedback if anyone runs into issues), but they?re a start: https://github.com/yaneurabeya/scratch/blob/master/docs/cheatsheets/freebsd . Cheers! -Garrett -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 496 bytes Desc: Message signed with OpenPGP using GPGMail URL: From aberg010 at my.hennepintech.edu Tue Sep 2 03:26:12 2014 From: aberg010 at my.hennepintech.edu (Andrew Berg) Date: Mon, 1 Sep 2014 22:25:58 -0500 Subject: [HEADSUP] pkg(8) is now the only package management tool In-Reply-To: <540534D3.603@sorbs.net> References: <20140901195520.GB77917@ivaldir.etoilebsd.net> <54050D07.4010404@sorbs.net> <540522A3.9050506@sorbs.net> <54052891.5000104@my.hennepintech.edu> <54052B1D.3040607@sorbs.net> <54053363.2030606@my.hennepintech.edu> <540534D3.603@sorbs.net> Message-ID: <540538C6.8050505@my.hennepintech.edu> On 2014.09.01 22:09, Michelle Sullivan wrote: > That's my point - there was a patch waiting to submit that knowingly > broke pkg_install at midnight on the day after the EOL... the EOL > shouldn't be an EOL - because it was really a 'portsnap after this date > before you upgrade and you're screwed it won't work any more at all...' As Peter outlined, this EOL was announced long ago, and it was mentioned at least once that it was to allow breaking changes. There really would be no reason to drop support for it in the ports tree if there were no plans to make changes. From michelle at sorbs.net Tue Sep 2 03:30:14 2014 From: michelle at sorbs.net (Michelle Sullivan) Date: Tue, 02 Sep 2014 05:30:08 +0200 Subject: [HEADSUP] pkg(8) is now the only package management tool In-Reply-To: <540538C6.8050505@my.hennepintech.edu> References: <20140901195520.GB77917@ivaldir.etoilebsd.net> <54050D07.4010404@sorbs.net> <540522A3.9050506@sorbs.net> <54052891.5000104@my.hennepintech.edu> <54052B1D.3040607@sorbs.net> <54053363.2030606@my.hennepintech.edu> <540534D3.603@sorbs.net> <540538C6.8050505@my.hennepintech.edu> Message-ID: <540539C0.7010008@sorbs.net> Andrew Berg wrote: > On 2014.09.01 22:09, Michelle Sullivan wrote: > >> That's my point - there was a patch waiting to submit that knowingly >> broke pkg_install at midnight on the day after the EOL... the EOL >> shouldn't be an EOL - because it was really a 'portsnap after this date >> before you upgrade and you're screwed it won't work any more at all...' >> > As Peter outlined, this EOL was announced long ago, and it was mentioned at > least once that it was to allow breaking changes. There really would be no > reason to drop support for it in the ports tree if there were no plans to make > changes. > The point is the EOL was not an EOL, it was a deadline, either switch or you're screwed, and it was communicated as an EOL not as a "here's a deadline, switch or you're screwed" -- Michelle Sullivan http://www.mhix.org/ From matthew at FreeBSD.org Tue Sep 2 06:27:44 2014 From: matthew at FreeBSD.org (Matthew Seaman) Date: Tue, 02 Sep 2014 07:27:22 +0100 Subject: [HEADSUP] pkg(8) is now the only package management tool In-Reply-To: <54052981.7010502@freebsd.org> References: <20140901195520.GB77917@ivaldir.etoilebsd.net> <54050D07.4010404@sorbs.net> <54052981.7010502@freebsd.org> Message-ID: <5405634A.9040803@FreeBSD.org> On 02/09/2014 03:20, Julian Elischer wrote: > We did an experiment at work here and wrote a small script that parsed > it and then rewrote it back to the package with newlines added and > pkg handled it just fine, so it should be a very minor thing to add > some newlines when generating it in the first place. > I don't think anything else needs to be changed. Have you tried looking at the output of 'pkg info -R pkgname' ? I think that should have as many new lines inserted as you could with for. Yes, the +MANIFEST and similar files are minified by turning all sequences of whitespace into single spaces. It seems to make a difference somewhere, although I'm not exactly clear where myself. Cheers, Matthew -- Dr Matthew J Seaman MA, D.Phil. PGP: http://www.infracaninophile.co.uk/pgpkey -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 971 bytes Desc: OpenPGP digital signature URL: From julian at freebsd.org Tue Sep 2 09:06:40 2014 From: julian at freebsd.org (Julian Elischer) Date: Tue, 02 Sep 2014 02:06:31 -0700 Subject: [HEADSUP] pkg(8) is now the only package management tool In-Reply-To: <540532AC.2070508@sorbs.net> References: <20140901195520.GB77917@ivaldir.etoilebsd.net> <54050D07.4010404@sorbs.net> <540522A3.9050506@sorbs.net> <54052891.5000104@my.hennepintech.edu> <54052DFA.4030808@freebsd.org> <540532AC.2070508@sorbs.net> Message-ID: <54058897.3080102@freebsd.org> On 9/1/14, 7:59 PM, Michelle Sullivan wrote: > Julian Elischer wrote: >> You should try arguing with someone like Bank of Americas security and >> operations >> department > You work for the same company as me? in a past life, they were a customer. > >> some day about whether they want to suddenly upgrade 300 machines >> for no real reason (from their perspective). >> From julian at freebsd.org Tue Sep 2 09:08:38 2014 From: julian at freebsd.org (Julian Elischer) Date: Tue, 02 Sep 2014 02:08:31 -0700 Subject: [HEADSUP] pkg(8) is now the only package management tool In-Reply-To: <54053372.6020009@my.hennepintech.edu> References: <20140901195520.GB77917@ivaldir.etoilebsd.net> <54050D07.4010404@sorbs.net> <540522A3.9050506@sorbs.net> <54052891.5000104@my.hennepintech.edu> <54052DFA.4030808@freebsd.org> <54053372.6020009@my.hennepintech.edu> Message-ID: <5405890F.8080804@freebsd.org> On 9/1/14, 8:03 PM, Andrew Berg wrote: > On 2014.09.01 21:39, Julian Elischer wrote: >> sigh.. when are we as a project, all going to learn that reality in >> business is >> that you often need to install stuff that is old. Its not always your >> choice. >> The custommers require it.. >> You should try arguing with someone like Bank of Americas security and >> operations >> department some day about whether they want to suddenly upgrade 300 >> machines >> for no real reason (from their perspective). > FreeBSD minor version upgrades are meant to be non-disruptive. However, I will > admit that I have not performed any such upgrades in a critical environment, so > if you think they are disruptive, please enlighten me with the details. > Also, there are options out there for getting support for extended periods if > you need it. Some companies are built around providing support for things that > the original developers have long abandoned because some businesses need it. It's not how disruptive they are technically. it's how many months of shakedown testing you have to go through before they allow you to put new software on any production system. > _______________________________________________ > freebsd-stable at freebsd.org mailing list > http://lists.freebsd.org/mailman/listinfo/freebsd-stable > To unsubscribe, send any mail to "freebsd-stable-unsubscribe at freebsd.org" > From ivoras at freebsd.org Tue Sep 2 09:44:31 2014 From: ivoras at freebsd.org (Ivan Voras) Date: Tue, 2 Sep 2014 11:43:48 +0200 Subject: pkg crashing? In-Reply-To: References: Message-ID: On 1 September 2014 18:45, Craig Rodrigues wrote: > On Mon, Sep 1, 2014 at 3:19 AM, Ivan Voras wrote: >> I don't force pkg version so it's always the latest, 1.3.7. It does >> seem that it could be a local issue, I have some bad experiences with >> pkg recently. > > OK. You should still report the version every time you report a problem, > because pkg development is fast moving. In the past few months, there > are a lot of > minor releases which fix problems (and break a few things along the way :). > > It is also useful to provide information about your system such as if you > upgraded only with packages from pkg.freebsd.org, > or if you built any ports yourself. Most are from pkg.freebsd.org but some I've installed from ports. But why should it matter? The pkg should record dependencies in both cases, right? Also, I appear to have had a similar (?) problem with chromium, on a *third* system: # pkg update Updating FreeBSD repository catalogue... FreeBSD repository is up-to-date. Updating FreeBSD_new_xorg repository catalogue... FreeBSD_new_xorg repository is up-to-date. All repositories are up-to-date. # pkg install chromium Updating FreeBSD repository catalogue... FreeBSD repository is up-to-date. Updating FreeBSD_new_xorg repository catalogue... FreeBSD_new_xorg repository is up-to-date. All repositories are up-to-date. The following 1 packages will be affected (of 0 checked): Installed packages to be REINSTALLED: chromium-36.0.1985.143_1 [FreeBSD] (direct dependency changed) The operation will free 816 KB. 28 MB to be downloaded. Proceed with this action? [y/N]: y pkg: http://pkg.FreeBSD.org/freebsd:10:x86:64/latest/All/chromium-36.0.1985.143_1.txz: Not Found Last week it wasn't in the repo databases but this week it is, BUT it's the wrong version. The version on the repo file server is chromium-37.0.2062.94.txz . Doing "pkg update -f" brought in the correct version. This version still crashes in the protobufs library, but this is probably because of the way chromium is built (http://comments.gmane.org/gmane.os.freebsd.chromium/1412, https://code.google.com/p/chromium/issues/detail?id=175508#c5). From haramrae at gmail.com Tue Sep 2 10:01:17 2014 From: haramrae at gmail.com (Alban Hertroys) Date: Tue, 2 Sep 2014 12:01:13 +0200 Subject: [HEADSUP] pkg(8) is now the only package management tool In-Reply-To: <5405890F.8080804@freebsd.org> References: <20140901195520.GB77917@ivaldir.etoilebsd.net> <54050D07.4010404@sorbs.net> <540522A3.9050506@sorbs.net> <54052891.5000104@my.hennepintech.edu> <54052DFA.4030808@freebsd.org> <54053372.6020009@my.hennepintech.edu> <5405890F.8080804@freebsd.org> Message-ID: On 2 September 2014 11:08, Julian Elischer wrote: > On 9/1/14, 8:03 PM, Andrew Berg wrote: >> >> On 2014.09.01 21:39, Julian Elischer wrote: >>> >>> sigh.. when are we as a project, all going to learn that reality in >>> business is >>> that you often need to install stuff that is old. Its not always your >>> choice. >>> The custommers require it.. >>> You should try arguing with someone like Bank of Americas security and >>> operations >>> department some day about whether they want to suddenly upgrade 300 >>> machines >>> for no real reason (from their perspective). >> >> FreeBSD minor version upgrades are meant to be non-disruptive. However, I >> will >> admit that I have not performed any such upgrades in a critical >> environment, so >> if you think they are disruptive, please enlighten me with the details. >> Also, there are options out there for getting support for extended periods >> if >> you need it. Some companies are built around providing support for things >> that >> the original developers have long abandoned because some businesses need >> it. > > > It's not how disruptive they are technically. > it's how many months of shakedown testing you have to go through before they > allow you to put new software on any production system. Just adding here, in commercial environments things don't change quickly or easily. Whether this applies to the current issue with pkg is not for me to say. For example, certain commercial upstream software vendors require to go through a certification process before they even consider supporting the new software you intend to use with theirs. Admittedly we haven't run into this issue in relation to FreeBSD, but we certainly have with Firefox. As an example, the last version of Firefox that Information Builders' WebFOCUS 7.7 supports is 3.6.7 (currently available versions are 31 or 32!) and for Internet Explorer that's 7 (currently at 11). If you run into any kind of problem, the standard answer is to use a browser that they support. Good luck with that! Firefox 3.6.7 was released on July 20, 2010; over 4 years ago. In such cases you're more or less required to keep an old system around that still has such old packages, if only to see if you can reproduce any issues you encounter (with modern versions of your software) on those old versions. With the deprecation of the old pkg_* tools you run into a conflict; You can either update packages that are _not_ under certification for such a vendor and get security updates and fixes using the new pkg, or you have to stick with the certified software and _not_ get any security updates or fixes. It gets more interesting if you have to deal with manufacturing processes (something we're looking to use FreeBSD for to replace our current OpenVMS systems before they go out of support), as often automatons write data to external databases and such software resides in PLC's. Manufacturing equipment tends to age and the kind of external databases they support is limited to what was available when they were new and the capabilities of the PLC involved. I can totally understand that at some point it starts to get impossible to maintain two separate packaging systems and I understand that you think 2 years is enough time to shake things out, but software vendors aren't that quick. For many, 2 years is a short time. Just saying... -- If you can't see the forest for the trees, Cut the trees and you'll see there is no forest. From robbak at robbak.com Tue Sep 2 10:25:20 2014 From: robbak at robbak.com (Robert Backhaus) Date: Tue, 2 Sep 2014 20:25:18 +1000 Subject: [HEADSUP] pkg(8) is now the only package management tool In-Reply-To: <540539C0.7010008@sorbs.net> References: <20140901195520.GB77917@ivaldir.etoilebsd.net> <54050D07.4010404@sorbs.net> <540522A3.9050506@sorbs.net> <54052891.5000104@my.hennepintech.edu> <54052B1D.3040607@sorbs.net> <54053363.2030606@my.hennepintech.edu> <540534D3.603@sorbs.net> <540538C6.8050505@my.hennepintech.edu> <540539C0.7010008@sorbs.net> Message-ID: On 2 September 2014 13:30, Michelle Sullivan wrote: > Andrew Berg wrote: > > On 2014.09.01 22:09, Michelle Sullivan wrote: > > > >> That's my point - there was a patch waiting to submit that knowingly > >> broke pkg_install at midnight on the day after the EOL... the EOL > >> shouldn't be an EOL - because it was really a 'portsnap after this date > >> before you upgrade and you're screwed it won't work any more at all...' > >> > > As Peter outlined, this EOL was announced long ago, and it was mentioned > at > > least once that it was to allow breaking changes. There really would be > no > > reason to drop support for it in the ports tree if there were no plans > to make > > changes. > > > > The point is the EOL was not an EOL, it was a deadline, either switch or > you're screwed, and it was communicated as an EOL not as a "here's a > deadline, switch or you're screwed" > > -- > Michelle Sullivan > http://www.mhix.org/ > The point is the EOL was *actually* an EOL: a deadline, either switch or you're screwed, and it was communicated as an EOL: a "here's a deadline, switch or you're screwed" From mva at freebsd.org Tue Sep 2 10:53:10 2014 From: mva at freebsd.org (Marcus von Appen) Date: Tue, 02 Sep 2014 12:52:56 +0200 Subject: [HEADSUP] pkg(8) is now the only package management tool In-Reply-To: References: <20140901195520.GB77917@ivaldir.etoilebsd.net> <54050D07.4010404@sorbs.net> <540522A3.9050506@sorbs.net> <54052891.5000104@my.hennepintech.edu> <54052DFA.4030808@freebsd.org> <54053372.6020009@my.hennepintech.edu> <5405890F.8080804@freebsd.org> Message-ID: <20140902125256.Horde.uv31ztwymThxUZ-OYPQoBw1@webmail.df.eu> Alban Hertroys : > On 2 September 2014 11:08, Julian Elischer wrote: >> On 9/1/14, 8:03 PM, Andrew Berg wrote: >>> >>> On 2014.09.01 21:39, Julian Elischer wrote: >>>> >>>> sigh.. when are we as a project, all going to learn that reality in >>>> business is >>>> that you often need to install stuff that is old. Its not always your >>>> choice. >>>> The custommers require it.. >>>> You should try arguing with someone like Bank of Americas security and >>>> operations >>>> department some day about whether they want to suddenly upgrade 300 >>>> machines >>>> for no real reason (from their perspective). >>> >>> FreeBSD minor version upgrades are meant to be non-disruptive. However, I >>> will >>> admit that I have not performed any such upgrades in a critical >>> environment, so >>> if you think they are disruptive, please enlighten me with the details. >>> Also, there are options out there for getting support for extended periods >>> if >>> you need it. Some companies are built around providing support for things >>> that >>> the original developers have long abandoned because some businesses need >>> it. >> >> >> It's not how disruptive they are technically. >> it's how many months of shakedown testing you have to go through before they >> allow you to put new software on any production system. > > Just adding here, in commercial environments things don't change > quickly or easily. Whether this applies to the current issue with pkg > is not for me to say. > > For example, certain commercial upstream software vendors require to > go through a certification process before they even consider > supporting the new software you intend to use with theirs. > > Admittedly we haven't run into this issue in relation to FreeBSD, but > we certainly have with Firefox. As an example, the last version of > Firefox that Information Builders' WebFOCUS 7.7 supports is 3.6.7 > (currently available versions are 31 or 32!) and for Internet Explorer > that's 7 (currently at 11). > If you run into any kind of problem, the standard answer is to use a > browser that they support. Good luck with that! > Firefox 3.6.7 was released on July 20, 2010; over 4 years ago. > > In such cases you're more or less required to keep an old system > around that still has such old packages, if only to see if you can > reproduce any issues you encounter (with modern versions of your > software) on those old versions. > > With the deprecation of the old pkg_* tools you run into a conflict; > You can either update packages that are _not_ under certification for > such a vendor and get security updates and fixes using the new pkg, or > you have to stick with the certified software and _not_ get any > security updates or fixes. > > > It gets more interesting if you have to deal with manufacturing > processes (something we're looking to use FreeBSD for to replace our > current OpenVMS systems before they go out of support), as often > automatons write data to external databases and such software resides > in PLC's. Manufacturing equipment tends to age and the kind of > external databases they support is limited to what was available when > they were new and the capabilities of the PLC involved. > > I can totally understand that at some point it starts to get > impossible to maintain two separate packaging systems and I understand > that you think 2 years is enough time to shake things out, but > software vendors aren't that quick. For many, 2 years is a short time. > It also should be noted that everyone had enough time to raise those issues in the time between tthe announcement and now. No one did. Now that it is gone, they are brought up, while they should have been long time ago instead. It can't work that way. My 2 cents in this discussion :-). Cheers Marcus From michelle at sorbs.net Tue Sep 2 11:47:37 2014 From: michelle at sorbs.net (Michelle Sullivan) Date: Tue, 02 Sep 2014 13:47:32 +0200 Subject: [HEADSUP] pkg(8) is now the only package management tool In-Reply-To: <20140902125256.Horde.uv31ztwymThxUZ-OYPQoBw1@webmail.df.eu> References: <20140901195520.GB77917@ivaldir.etoilebsd.net> <54050D07.4010404@sorbs.net> <540522A3.9050506@sorbs.net> <54052891.5000104@my.hennepintech.edu> <54052DFA.4030808@freebsd.org> <54053372.6020009@my.hennepintech.edu> <5405890F.8080804@freebsd.org> <20140902125256.Horde.uv31ztwymThxUZ-OYPQoBw1@webmail.df.eu> Message-ID: <5405AE54.60809@sorbs.net> Marcus von Appen wrote: > Alban Hertroys : > >> >> I can totally understand that at some point it starts to get >> impossible to maintain two separate packaging systems and I understand >> that you think 2 years is enough time to shake things out, but >> software vendors aren't that quick. For many, 2 years is a short time. >> > > It also should be noted that everyone had enough time to raise those > issues > in the time between tthe announcement and now. No one did. Now that it is > gone, they are brought up, while they should have been long time ago > instead. It can't work that way. > > My 2 cents in this discussion :-). Actually I brought it up as soon as I found the EOL was a deadline for breaking pkg_* tools, was told, "too late now" - that was more than 2 weeks ago, less than 2 months ago (forget the date) ... I'm happy with an EOL and working to upgrade everything, I'm not happy that the EOL was not actually an EOL and it was actually a deadline. -- Michelle Sullivan http://www.mhix.org/ From theraven at FreeBSD.org Tue Sep 2 12:08:33 2014 From: theraven at FreeBSD.org (David Chisnall) Date: Tue, 2 Sep 2014 13:08:20 +0100 Subject: [HEADSUP] pkg(8) is now the only package management tool In-Reply-To: <5405AE54.60809@sorbs.net> References: <20140901195520.GB77917@ivaldir.etoilebsd.net> <54050D07.4010404@sorbs.net> <540522A3.9050506@sorbs.net> <54052891.5000104@my.hennepintech.edu> <54052DFA.4030808@freebsd.org> <54053372.6020009@my.hennepintech.edu> <5405890F.8080804@freebsd.org> <20140902125256.Horde.uv31ztwymThxUZ-OYPQoBw1@webmail.df.eu> <5405AE54.60809@sorbs.net> Message-ID: <0B55DEDF-7268-4E7A-8971-36AB68D0C170@FreeBSD.org> On 2 Sep 2014, at 12:47, Michelle Sullivan wrote: > I'm not happy that the EOL was > not actually an EOL and it was actually a deadline. I'm not sure what you think the difference is. The EOL says 'the FreeBSD project no longer supports this configuration'. If you are not relying on us for support (i.e. using an old or forked ports tree, or building your own packages), then things will continue to work. If you are expecting (unpaid, volunteer) support from the project in the form of packages and a useable ports tree, then you need to use a supported configuration. If being able to use the ports tree without installing pkg(8) is sufficiently valuable to you, then I can put you in touch with some companies that will backport things to a copy of the ports tree for your use (although the price tag will scale with the number of ports that you want to support). If, however, your complaint is that it's hard to get new software certified for your system *then this change has absolutely no effect on you!* If you're not worried about upgrading ports at all, then you can just stick with the ports tree version from the time of your release. If you're able to upgrade ports, then upgrading the pkg port should not be an issue. David From allbery.b at gmail.com Tue Sep 2 13:37:30 2014 From: allbery.b at gmail.com (Brandon Allbery) Date: Tue, 2 Sep 2014 09:37:26 -0400 Subject: [HEADSUP] pkg(8) is now the only package management tool In-Reply-To: <20140902125256.Horde.uv31ztwymThxUZ-OYPQoBw1@webmail.df.eu> References: <20140901195520.GB77917@ivaldir.etoilebsd.net> <54050D07.4010404@sorbs.net> <540522A3.9050506@sorbs.net> <54052891.5000104@my.hennepintech.edu> <54052DFA.4030808@freebsd.org> <54053372.6020009@my.hennepintech.edu> <5405890F.8080804@freebsd.org> <20140902125256.Horde.uv31ztwymThxUZ-OYPQoBw1@webmail.df.eu> Message-ID: On Tue, Sep 2, 2014 at 6:52 AM, Marcus von Appen wrote: > It also should be noted that everyone had enough time to raise those issues > in the time between tthe announcement and now > If this is an issue that needed to be brought up, then FreeBSD has apparently never before been used in an enterprise??? -- brandon s allbery kf8nh sine nomine associates allbery.b at gmail.com ballbery at sinenomine.net unix, openafs, kerberos, infrastructure, xmonad http://sinenomine.net From bdrewery at FreeBSD.org Tue Sep 2 13:46:54 2014 From: bdrewery at FreeBSD.org (Bryan Drewery) Date: Tue, 02 Sep 2014 08:46:03 -0500 Subject: [HEADSUP] pkg(8) is now the only package management tool In-Reply-To: <54052B1D.3040607@sorbs.net> References: <20140901195520.GB77917@ivaldir.etoilebsd.net> <54050D07.4010404@sorbs.net> <540522A3.9050506@sorbs.net> <54052891.5000104@my.hennepintech.edu> <54052B1D.3040607@sorbs.net> Message-ID: <5405CA1B.2040307@FreeBSD.org> On 9/1/2014 9:27 PM, Michelle Sullivan wrote: > oh and what was it, 1.3.6 -> 1.3.7? broke > shit... (badly) ... What broke? I am not aware of any new regressions in 1.3.7. -- Regards, Bryan Drewery -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 488 bytes Desc: OpenPGP digital signature URL: From mva at freebsd.org Tue Sep 2 14:30:27 2014 From: mva at freebsd.org (Marcus von Appen) Date: Tue, 02 Sep 2014 16:22:05 +0200 Subject: [HEADSUP] pkg(8) is now the only package management tool In-Reply-To: References: <20140901195520.GB77917@ivaldir.etoilebsd.net> <54050D07.4010404@sorbs.net> <540522A3.9050506@sorbs.net> <54052891.5000104@my.hennepintech.edu> <54052DFA.4030808@freebsd.org> <54053372.6020009@my.hennepintech.edu> <5405890F.8080804@freebsd.org> <20140902125256.Horde.uv31ztwymThxUZ-OYPQoBw1@webmail.df.eu> Message-ID: <20140902162205.Horde._7jSRVzORy7dvrf149pEyw1@webmail.df.eu> Brandon Allbery : > On Tue, Sep 2, 2014 at 6:52 AM, Marcus von Appen wrote: > >> It also should be noted that everyone had enough time to raise those issues >> in the time between tthe announcement and now >> > > If this is an issue that needed to be brought up, then FreeBSD has > apparently never before been used in an enterprise??? I'm tempted to ask, if the enterprise has SLAs to ensure continuity, even after the official support has ended? ;-) Cheers Marcus From yaneurabeya at gmail.com Tue Sep 2 15:21:41 2014 From: yaneurabeya at gmail.com (Garrett Cooper) Date: Tue, 2 Sep 2014 08:21:38 -0700 Subject: [HEADSUP] pkg(8) is now the only package management tool In-Reply-To: <5405AE54.60809@sorbs.net> References: <20140901195520.GB77917@ivaldir.etoilebsd.net> <54050D07.4010404@sorbs.net> <540522A3.9050506@sorbs.net> <54052891.5000104@my.hennepintech.edu> <54052DFA.4030808@freebsd.org> <54053372.6020009@my.hennepintech.edu> <5405890F.8080804@freebsd.org> <20140902125256.Horde.uv31ztwymThxUZ-OYPQoBw1@webmail.df.eu> <5405AE54.60809@sorbs.net> Message-ID: <1D2B4A91-E76C-43A0-BE75-D926357EF1AF@gmail.com> > On Sep 2, 2014, at 4:47, Michelle Sullivan wrote: > > Marcus von Appen wrote: >> Alban Hertroys : >> >>> >>> I can totally understand that at some point it starts to get >>> impossible to maintain two separate packaging systems and I understand >>> that you think 2 years is enough time to shake things out, but >>> software vendors aren't that quick. For many, 2 years is a short time. >> >> It also should be noted that everyone had enough time to raise those >> issues >> in the time between tthe announcement and now. No one did. Now that it is >> gone, they are brought up, while they should have been long time ago >> instead. It can't work that way. >> >> My 2 cents in this discussion :-). > > Actually I brought it up as soon as I found the EOL was a deadline for > breaking pkg_* tools, was told, "too late now" - that was more than 2 > weeks ago, less than 2 months ago (forget the date) ... I'm happy with > an EOL and working to upgrade everything, I'm not happy that the EOL was > not actually an EOL and it was actually a deadline. Hi Michelle, One subtle point that I wanted to ask for clarification is you thought the EOL announcement for pkg_install was going to be "pkg_install is no longer going to be supported, but you can still use it", instead of "pkg_install support is going to be removed from the tree" -- is that correct? You'd probably hate to do this, but forking the sources and changing from portsnap to a git or svn backed ports tree that downloads a tarball snapshot might be the best resolution to this issue now... Thanks! -Garrett From michelle at sorbs.net Tue Sep 2 15:40:41 2014 From: michelle at sorbs.net (Michelle Sullivan) Date: Tue, 02 Sep 2014 17:40:37 +0200 Subject: [HEADSUP] pkg(8) is now the only package management tool In-Reply-To: <1D2B4A91-E76C-43A0-BE75-D926357EF1AF@gmail.com> References: <20140901195520.GB77917@ivaldir.etoilebsd.net> <54050D07.4010404@sorbs.net> <540522A3.9050506@sorbs.net> <54052891.5000104@my.hennepintech.edu> <54052DFA.4030808@freebsd.org> <54053372.6020009@my.hennepintech.edu> <5405890F.8080804@freebsd.org> <20140902125256.Horde.uv31ztwymThxUZ-OYPQoBw1@webmail.df.eu> <5405AE54.60809@sorbs.net> <1D2B4A91-E76C-43A0-BE75-D926357EF1AF@gmail.com> Message-ID: <5405E4F5.4090902@sorbs.net> Garrett Cooper wrote: >> On Sep 2, 2014, at 4:47, Michelle Sullivan wrote: >> >> Marcus von Appen wrote: >> >>> Alban Hertroys : >>> >>> >>>> I can totally understand that at some point it starts to get >>>> impossible to maintain two separate packaging systems and I understand >>>> that you think 2 years is enough time to shake things out, but >>>> software vendors aren't that quick. For many, 2 years is a short time. >>>> >>> It also should be noted that everyone had enough time to raise those >>> issues >>> in the time between tthe announcement and now. No one did. Now that it is >>> gone, they are brought up, while they should have been long time ago >>> instead. It can't work that way. >>> >>> My 2 cents in this discussion :-). >>> >> Actually I brought it up as soon as I found the EOL was a deadline for >> breaking pkg_* tools, was told, "too late now" - that was more than 2 >> weeks ago, less than 2 months ago (forget the date) ... I'm happy with >> an EOL and working to upgrade everything, I'm not happy that the EOL was >> not actually an EOL and it was actually a deadline. >> > > Hi Michelle, > One subtle point that I wanted to ask for clarification is you thought the EOL announcement for pkg_install was going to be "pkg_install is no longer going to be supported, but you can still use it", instead of "pkg_install support is going to be removed from the tree" -- is that correct? > 100% correct! (thank you for being one of the few to see the subtle but *very* important difference) > You'd probably hate to do this, but forking the sources and changing from portsnap to a git or svn backed ports tree that downloads a tarball snapshot might be the best resolution to this issue now... > This is my only option - however, I suspect I'm already f**ked - my build servers kicked off at 4am and the non pkg jails automatically converted themselves to pkg.. the pkg jails obviously continued... however I now have a repo that contains half pkg versions in the same directory structure and indexes as the pkg_install structure... Time to rebuild everything from scratch I think - second time in a year.. I'm guessing my boss is going to tell me, use RPM, no wasting more time on it... only time will tell... you'll know the result if you see future posts and patches from me. Michelle -- Michelle Sullivan http://www.mhix.org/ From allbery.b at gmail.com Tue Sep 2 15:47:52 2014 From: allbery.b at gmail.com (Brandon Allbery) Date: Tue, 2 Sep 2014 11:47:49 -0400 Subject: [HEADSUP] pkg(8) is now the only package management tool In-Reply-To: <5405E4F5.4090902@sorbs.net> References: <20140901195520.GB77917@ivaldir.etoilebsd.net> <54050D07.4010404@sorbs.net> <540522A3.9050506@sorbs.net> <54052891.5000104@my.hennepintech.edu> <54052DFA.4030808@freebsd.org> <54053372.6020009@my.hennepintech.edu> <5405890F.8080804@freebsd.org> <20140902125256.Horde.uv31ztwymThxUZ-OYPQoBw1@webmail.df.eu> <5405AE54.60809@sorbs.net> <1D2B4A91-E76C-43A0-BE75-D926357EF1AF@gmail.com> <5405E4F5.4090902@sorbs.net> Message-ID: On Tue, Sep 2, 2014 at 11:40 AM, Michelle Sullivan wrote: > This is my only option - however, I suspect I'm already f**ked - my > build servers kicked off at 4am and the non pkg jails automatically > converted themselves to pkg.. the pkg jails obviously continued... > however I now have a repo that contains half pkg versions in the same > directory structure and indexes as the pkg_install structure... > So, the flip side of enterprise software management is that you probably should not be integrating a rolling release like ports into what is supposed to be a stable verified environment in the first place. *Especially* not via cron jobs with no supervision. At the very least, your jails should be working from a local ports tree (or packages via poudriere), with cherry-picking of locally tested patches. -- brandon s allbery kf8nh sine nomine associates allbery.b at gmail.com ballbery at sinenomine.net unix, openafs, kerberos, infrastructure, xmonad http://sinenomine.net From michelle at sorbs.net Tue Sep 2 15:53:57 2014 From: michelle at sorbs.net (Michelle Sullivan) Date: Tue, 02 Sep 2014 17:53:52 +0200 Subject: [HEADSUP] pkg(8) is now the only package management tool In-Reply-To: References: <20140901195520.GB77917@ivaldir.etoilebsd.net> <54050D07.4010404@sorbs.net> <540522A3.9050506@sorbs.net> <54052891.5000104@my.hennepintech.edu> <54052DFA.4030808@freebsd.org> <54053372.6020009@my.hennepintech.edu> <5405890F.8080804@freebsd.org> <20140902125256.Horde.uv31ztwymThxUZ-OYPQoBw1@webmail.df.eu> <5405AE54.60809@sorbs.net> <1D2B4A91-E76C-43A0-BE75-D926357EF1AF@gmail.com> <5405E4F5.4090902@sorbs.net> Message-ID: <5405E810.8000400@sorbs.net> Brandon Allbery wrote: > On Tue, Sep 2, 2014 at 11:40 AM, Michelle Sullivan > wrote: > > >> This is my only option - however, I suspect I'm already f**ked - my >> build servers kicked off at 4am and the non pkg jails automatically >> converted themselves to pkg.. the pkg jails obviously continued... >> however I now have a repo that contains half pkg versions in the same >> directory structure and indexes as the pkg_install structure... >> >> > > So, the flip side of enterprise software management is that you probably > should not be integrating a rolling release like ports into what is > supposed to be a stable verified environment in the first place. > *Especially* not via cron jobs with no supervision. At the very least, your > jails should be working from a local ports tree (or packages via > poudriere), with cherry-picking of locally tested patches. > > The roll until they get a stable base (using Jenkins as the controller) - they've been rolling since a patch to DBIx-SearchBuilder (that I created and submitted), which out came the DBD::Pg update to 3.3.0 and the subsequent blacklisting of it for RT 4.x and then the tcl breakage around mid August until 2 days ago... So yeah not that stupid. -- Michelle Sullivan http://www.mhix.org/ From christer.solskogen at gmail.com Tue Sep 2 18:35:20 2014 From: christer.solskogen at gmail.com (Christer Solskogen) Date: Tue, 02 Sep 2014 20:32:19 +0200 Subject: how to rsync from pkg.freebsd.org In-Reply-To: <20140313102947.GE90364@ithaqua.etoilebsd.net> References: <20140313102947.GE90364@ithaqua.etoilebsd.net> Message-ID: On 13.03.2014 11:29, Baptiste Daroussin wrote: > On Thu, Mar 13, 2014 at 03:37:45PM +0800, Jason wrote: >> hi all: >> >> I'm from china.we have thousands of machines running on FreeBSD. >> Because the link between china and usa is poor,the pkg instal is >> very slow and often failed... >> so we want to build our own pkgng mirror. >> Does the pkg.freebsd.org offer any rsync interface so we can access? > > This is being worked on, we are experimenting a couple of different ways and > will soon communicate on a viable way to do it. > > Stay tuned :) > I've stayed, and tuned. Is this possible now? -- chs From mat at mat.cc Wed Sep 3 14:24:34 2014 From: mat at mat.cc (Mathieu Arnold) Date: Wed, 03 Sep 2014 16:24:27 +0200 Subject: [HEADSUP] pkg(8) is now the only package management tool In-Reply-To: <5405AE54.60809@sorbs.net> References: <20140901195520.GB77917@ivaldir.etoilebsd.net> <54050D07.4010404@sorbs.net> <540522A3.9050506@sorbs.net> <54052891.5000104@my.hennepintech.edu> <54052DFA.4030808@freebsd.org> <54053372.6020009@my.hennepintech.edu> <5405890F.8080804@freebsd.org> <20140902125256.Horde.uv31ztwymThxUZ-OYPQoBw1@webmail.df.eu> <5405AE54.60809@sorbs.net> Message-ID: +--On 2 septembre 2014 13:47:32 +0200 Michelle Sullivan wrote: | Marcus von Appen wrote: |> Alban Hertroys : |> |>> |>> I can totally understand that at some point it starts to get |>> impossible to maintain two separate packaging systems and I understand |>> that you think 2 years is enough time to shake things out, but |>> software vendors aren't that quick. For many, 2 years is a short time. |>> |> |> It also should be noted that everyone had enough time to raise those |> issues |> in the time between tthe announcement and now. No one did. Now that it is |> gone, they are brought up, while they should have been long time ago |> instead. It can't work that way. |> |> My 2 cents in this discussion :-). | | Actually I brought it up as soon as I found the EOL was a deadline for | breaking pkg_* tools, was told, "too late now" - that was more than 2 | weeks ago, less than 2 months ago (forget the date) ... I'm happy with | an EOL and working to upgrade everything, I'm not happy that the EOL was | not actually an EOL and it was actually a deadline. I still don't see what you have to say about what EOL mean, it's *End Of Life* meaning after, it is dead, and won't exist any more. -- Mathieu Arnold From michelle at sorbs.net Wed Sep 3 14:36:34 2014 From: michelle at sorbs.net (Michelle Sullivan) Date: Wed, 03 Sep 2014 16:36:29 +0200 Subject: [HEADSUP] pkg(8) is now the only package management tool In-Reply-To: References: <20140901195520.GB77917@ivaldir.etoilebsd.net> <54050D07.4010404@sorbs.net> <540522A3.9050506@sorbs.net> <54052891.5000104@my.hennepintech.edu> <54052DFA.4030808@freebsd.org> <54053372.6020009@my.hennepintech.edu> <5405890F.8080804@freebsd.org> <20140902125256.Horde.uv31ztwymThxUZ-OYPQoBw1@webmail.df.eu> <5405AE54.60809@sorbs.net> Message-ID: <5407276D.3050200@sorbs.net> Mathieu Arnold wrote: > I still don't see what you have to say about what EOL mean, it's *End Of > Life* meaning after, it is dead, and won't exist any more. > > Ahh so all those Windows XP servers are dead and don't work anymore... -- Michelle Sullivan http://www.mhix.org/ From mat at mat.cc Wed Sep 3 14:56:18 2014 From: mat at mat.cc (Mathieu Arnold) Date: Wed, 03 Sep 2014 16:56:14 +0200 Subject: [HEADSUP] pkg(8) is now the only package management tool In-Reply-To: <5407276D.3050200@sorbs.net> References: <20140901195520.GB77917@ivaldir.etoilebsd.net> <54050D07.4010404@sorbs.net> <540522A3.9050506@sorbs.net> <54052891.5000104@my.hennepintech.edu> <54052DFA.4030808@freebsd.org> <54053372.6020009@my.hennepintech.edu> <5405890F.8080804@freebsd.org> <20140902125256.Horde.uv31ztwymThxUZ-OYPQoBw1@webmail.df.eu> <5405AE54.60809@sorbs.net> <5407276D.3050200@sorbs.net> Message-ID: <9DAF63F1F61E0FC2F6B237BD@ogg.in.absolight.net> +--On 3 septembre 2014 16:36:29 +0200 Michelle Sullivan wrote: | Mathieu Arnold wrote: |> I still don't see what you have to say about what EOL mean, it's *End Of |> Life* meaning after, it is dead, and won't exist any more. |> |> | Ahh so all those Windows XP servers are dead and don't work anymore... Not at all, but you don't update them any more. -- Mathieu Arnold From michelle at sorbs.net Wed Sep 3 15:17:51 2014 From: michelle at sorbs.net (Michelle Sullivan) Date: Wed, 03 Sep 2014 17:17:48 +0200 Subject: [HEADSUP] pkg(8) is now the only package management tool In-Reply-To: <9DAF63F1F61E0FC2F6B237BD@ogg.in.absolight.net> References: <20140901195520.GB77917@ivaldir.etoilebsd.net> <54050D07.4010404@sorbs.net> <540522A3.9050506@sorbs.net> <54052891.5000104@my.hennepintech.edu> <54052DFA.4030808@freebsd.org> <54053372.6020009@my.hennepintech.edu> <5405890F.8080804@freebsd.org> <20140902125256.Horde.uv31ztwymThxUZ-OYPQoBw1@webmail.df.eu> <5405AE54.60809@sorbs.net> <5407276D.3050200@sorbs.net> <9DAF63F1F61E0FC2F6B237BD@ogg.in.absolight.net> Message-ID: <5407311C.5070903@sorbs.net> Mathieu Arnold wrote: > +--On 3 septembre 2014 16:36:29 +0200 Michelle Sullivan > wrote: > | Mathieu Arnold wrote: > |> I still don't see what you have to say about what EOL mean, it's *End Of > |> Life* meaning after, it is dead, and won't exist any more. > |> > |> > | Ahh so all those Windows XP servers are dead and don't work anymore... > > Not at all, but you don't update them any more. > > Actually you do, but it does nothing... just like freebsd-update on any EOL release... Microsoft didn't release a patch that would change the base system so you can no longer install any software... They just stopped providing updates, then later stopped providing security updates. -- Michelle Sullivan http://www.mhix.org/ From mat at mat.cc Wed Sep 3 15:21:14 2014 From: mat at mat.cc (Mathieu Arnold) Date: Wed, 03 Sep 2014 17:21:10 +0200 Subject: [HEADSUP] pkg(8) is now the only package management tool In-Reply-To: <5407311C.5070903@sorbs.net> References: <20140901195520.GB77917@ivaldir.etoilebsd.net> <54050D07.4010404@sorbs.net> <540522A3.9050506@sorbs.net> <54052891.5000104@my.hennepintech.edu> <54052DFA.4030808@freebsd.org> <54053372.6020009@my.hennepintech.edu> <5405890F.8080804@freebsd.org> <20140902125256.Horde.uv31ztwymThxUZ-OYPQoBw1@webmail.df.eu> <5405AE54.60809@sorbs.net> <5407276D.3050200@sorbs.net> <9DAF63F1F61E0FC2F6B237BD@ogg.in.absolight.net> <5407311C.5070903@sorbs.net> Message-ID: <562F5E76969DAAFC2106C9EC@ogg.in.absolight.net> +--On 3 septembre 2014 17:17:48 +0200 Michelle Sullivan wrote: | Mathieu Arnold wrote: |> +--On 3 septembre 2014 16:36:29 +0200 Michelle Sullivan |> wrote: |> | Mathieu Arnold wrote: |> |> I still don't see what you have to say about what EOL mean, it's *End |> |> Of Life* meaning after, it is dead, and won't exist any more. |> |> |> |> |> | Ahh so all those Windows XP servers are dead and don't work anymore... |> |> Not at all, but you don't update them any more. |> |> | Actually you do, but it does nothing... just like freebsd-update on any | EOL release... | | Microsoft didn't release a patch that would change the base system so | you can no longer install any software... They just stopped providing | updates, then later stopped providing security updates. You can still go and fetch/build software. You can't do it using the FreeBSD App Store though. -- Mathieu Arnold From lars.engels at 0x20.net Wed Sep 3 18:34:30 2014 From: lars.engels at 0x20.net (Lars Engels) Date: Wed, 3 Sep 2014 20:34:23 +0200 Subject: [HEADSUP] pkg(8) is now the only package management tool In-Reply-To: <5407311C.5070903@sorbs.net> References: <54052DFA.4030808@freebsd.org> <54053372.6020009@my.hennepintech.edu> <5405890F.8080804@freebsd.org> <20140902125256.Horde.uv31ztwymThxUZ-OYPQoBw1@webmail.df.eu> <5405AE54.60809@sorbs.net> <5407276D.3050200@sorbs.net> <9DAF63F1F61E0FC2F6B237BD@ogg.in.absolight.net> <5407311C.5070903@sorbs.net> Message-ID: <20140903183422.GF57121@e-new.0x20.net> On Wed, Sep 03, 2014 at 05:17:48PM +0200, Michelle Sullivan wrote: > Mathieu Arnold wrote: > > +--On 3 septembre 2014 16:36:29 +0200 Michelle Sullivan > > wrote: > > | Mathieu Arnold wrote: > > |> I still don't see what you have to say about what EOL mean, it's *End Of > > |> Life* meaning after, it is dead, and won't exist any more. > > |> > > |> > > | Ahh so all those Windows XP servers are dead and don't work anymore... > > > > Not at all, but you don't update them any more. > > > > > Actually you do, but it does nothing... just like freebsd-update on any > EOL release... > > Microsoft didn't release a patch that would change the base system so > you can no longer install any software... They just stopped providing > updates, then later stopped providing security updates. Same for FreeBSD and pkg_*. Stay with status quo, use pkg_* with the last tagged version of the ports tree that works with pkg_* or switch to pkg and be happy like the rest of us. -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 618 bytes Desc: not available URL: From franck.porcher at gmail.com Wed Sep 3 23:21:39 2014 From: franck.porcher at gmail.com (Franck Porcher) Date: Wed, 3 Sep 2014 13:21:17 -1000 Subject: FreeBSD 10.0 install - pkg database error when trying to install packages from the commercial DVD Message-ID: Hello FreeBSD, I'm sorry to bother you with the following question, but I find myself stuck in trying to (post)install FreeBSD packages directly from the commercial DVD. The install went just fine, and I can use teh computer. But I get the following error when selecting "Packages > CD/DVD" from the bsdconfig command ncurses-based menu : *No pkg(8) database found!* pkg has been boostrapped as indicated, and there are 2 databases in /var/db/pkg, namely : local.sqlite (148K) repo-FreeBSD_install_cdrom.sqlite (16M) and the later seems populated... So, would someone have an idea what such msg actually means, what I'm doing wrong, and how I could overcome the situation. Much thanks in advance for any direction to look for (I have installed FreeBSD since release 6 and cannot remember ever having been met by this msg in trying to install precompiled packages) Franck From walkerindarkness at gmail.com Thu Sep 4 05:42:02 2014 From: walkerindarkness at gmail.com (Fervent Dissent) Date: Thu, 4 Sep 2014 13:42:00 +0800 Subject: FreeBSD Port: x11-drivers/xf86-input-keyboard In-Reply-To: References: Message-ID: This is the first update I've done since switching to the new xorg. The first time neither the keyboard or mouse drivers were correctly chosen. Now the mouse upgraded fine, but the keyboard still defaults to the old xorg driver. After working on it I confirmed by manually downloading both versions. My keyboard only worked after 'pkg add xf86-input-keyboard-1.8.0_2.txz'. The version pkg fetches is 11120kB, the one I added is 11124kB. I am using all packages on my system. Could this be fixed? I also tried to build from ports, but that also built the wrong version? I just wanted to report this incase others have problems. X.Org X Server 1.12.4 Release Date: 2012-08-27 [ 10560.924] X Protocol Version 11, Revision 0 [ 10560.924] Build Operating System: FreeBSD 10.0-RELEASE-p3 amd64 [ 10560.925] Current Operating System: FreeBSD satellite 10.0-STABLE FreeBSD 10.0-STABLE #0 r267603: Wed Jun 18 20:00:27 CST 2014 sara at satellite:/usr/obj/usr/src/sys/TWILIGHT amd64 [ 10560.925] Build Date: 28 August 2014 02:53:46PM [ 10561.397] (II) Using input driver 'mouse' for 'Mouse0' [ 10561.398] (**) Option "CorePointer" [ 10561.398] (**) Mouse0: always reports core events [ 10561.398] (**) Option "Protocol" "auto" [ 10561.398] (**) Option "Device" "/dev/sysmouse" [ 10561.398] (**) Mouse0: Protocol: "auto" [ 10561.398] (**) Mouse0: always reports core events [ 10561.398] (==) Mouse0: Emulate3Buttons, Emulate3Timeout: 50 [ 10561.398] (**) Option "ZAxisMapping" "4 5 6 7" [ 10561.398] (**) Mouse0: ZAxisMapping: buttons 4, 5, 6 and 7 [ 10561.398] (**) Mouse0: Buttons: 7 [ 10561.398] (II) XINPUT: Adding extended input device "Mouse0" (type: MOUSE, id 6) [ 10561.399] (**) Mouse0: (accel) keeping acceleration scheme 1 [ 10561.399] (**) Mouse0: (accel) acceleration profile 0 [ 10561.399] (**) Mouse0: (accel) acceleration factor: 2.000 [ 10561.399] (**) Mouse0: (accel) acceleration threshold: 4 [ 10561.399] (II) Mouse0: SetupAuto: hw.iftype is 4, hw.model is 0 [ 10561.399] (II) Mouse0: SetupAuto: protocol is SysMouse [ 10561.399] (II) LoadModule: "kbd" [ 10561.400] (II) Loading /usr/local/lib/xorg/modules/input/kbd_drv.so [ 10561.401] (II) Module kbd: vendor="X.Org Foundation" [ 10561.401] compiled for 1.7.7, module version = 1.8.0 [ 10561.401] Module class: X.Org XInput Driver [ 10561.401] ABI class: X.Org XInput driver, version 7.0 [ 10561.401] (EE) module ABI major version (7) doesn't match the server's version (16) [ 10561.401] (II) UnloadModule: "kbd" [ 10561.401] (II) Unloading kbd [ 10561.401] (EE) Failed to load module "kbd" (module requirement mismatch, 0) [ 10561.401] (EE) No input driver matching `kbd' -- !CAUTION! Sent from Gmail Mobile !CAUTION! From se at freebsd.org Thu Sep 4 06:04:51 2014 From: se at freebsd.org (Stefan Esser) Date: Thu, 04 Sep 2014 08:04:30 +0200 Subject: FreeBSD Port: x11-drivers/xf86-input-keyboard In-Reply-To: References: Message-ID: <540800EE.6070201@freebsd.org> Am 04.09.2014 um 07:42 schrieb Fervent Dissent: > This is the first update I've done since switching to the new xorg. The > first time neither the keyboard or mouse drivers were correctly chosen. Now > the mouse upgraded fine, but the keyboard still defaults to the old xorg > driver. After working on it I confirmed by manually downloading both > versions. My keyboard only worked after 'pkg add > xf86-input-keyboard-1.8.0_2.txz'. The version pkg fetches is 11120kB, the > one I added is 11124kB. I am using all packages on my system. > > Could this be fixed? > > I also tried to build from ports, but that also built the wrong version? I > just wanted to report this incase others have problems. This is caused by your use of the vt console driver with a Unicode keymap containing characters >= 0x100. I have annalyzed this and created a PR with patch for the port: https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=193192 As a workaround, you may start the X server with a US keyboard by issuing kbdcontrol -l us just before starting X. You'll need to select an appropriate keymap under X, then. Regards, STefan PS: This is a problem with the xorg-input-keyboard port and if it is not fixed within that port, there will be a huge number of users affected, when 10.1 is released. This must be fixed! From ivoras at freebsd.org Thu Sep 4 15:44:04 2014 From: ivoras at freebsd.org (Ivan Voras) Date: Thu, 4 Sep 2014 17:43:21 +0200 Subject: Dependancy tracking not working??? Message-ID: Hello, As a continuation of my previous (and unanswered) post about upgrading php, I've made a clean virtual machine for testing and installed php5 with several extensions on it using pkg. I've waited a few days for some changes to the PHP port to happen, and now have a slightly obsolete version of PHP and its extensions which I'd like to upgrade. The situation is this: # pkg info|grep php php5-5.4.31 PHP Scripting Language php5-curl-5.4.31 The curl shared extension for php php5-gd-5.4.31 The gd shared extension for php php5-session-5.4.31 The session shared extension for php php5-xml-5.4.31 The xml shared extension for php php5-xmlrpc-5.4.31 The xmlrpc shared extension for php php5-zlib-5.4.31 The zlib shared extension for php Note that my installed version of php5 is "5.4.31". A repo search shows this: # pkg search php5 ... php5-5.4.31_1 php5-bcmath-5.4.31_1 php5-bz2-5.4.31_1 ... I.e. the most recent version of php5 is "5.4.31_1" Running "pkg upgrade php5", however, does NOT result in all the dependencies on PHP to be upgraded: # pkg upgrade php5 Updating FreeBSD repository catalogue... FreeBSD repository is up-to-date. All repositories are up-to-date. pkg: plexhometheater has a missing dependency: lame The following 1 packages will be affected (of 0 checked): Installed packages to be UPGRADED: php5: 5.4.31 -> 5.4.31_1 2 MB to be downloaded. Proceed with this action? [y/N]: Dependency information is correctly recorded: # pkg info -r php5 php5-5.4.31: php5-xml-5.4.31 php5-session-5.4.31 php5-zlib-5.4.31 php5-gd-5.4.31 php5-xmlrpc-5.4.31 php5-curl-5.4.31 So... how can I upgrade php and all things which depend on it? This is only a single example, I've complained of the same problems the last couple of days, on real, non-test servers. This time, I have a test server and can offer access to it to whomever wants to fix it. (this is, as before, pkg 1.3.7 on FreeBSD 10). From ivoras at freebsd.org Fri Sep 5 14:36:50 2014 From: ivoras at freebsd.org (Ivan Voras) Date: Fri, 5 Sep 2014 16:36:07 +0200 Subject: pkg upgrade -f refusing to upgrade packages??? Message-ID: Hello, I seem to keep either finding problems in pkg or completely missing the point of it :( Here's another issue: # pkg info -g 'cups*' cups-1.7.3 cups-base-1.7.3_1 cups-client-1.7.3 cups-filters-1.0.55 cups-image-1.7.2 cups-pstoraster-8.15.4_8 Right, so my interpretation of the output above is that I have that set of packages installed on this system. Now, I would like to force-upgrade them all: # pkg upgrade -f `pkg info -g 'cups*'` Updating FreeBSD repository catalogue... FreeBSD repository is up-to-date. Updating FreeBSD_new_xorg repository catalogue... FreeBSD_new_xorg repository is up-to-date. All repositories are up-to-date. The following 3 packages will be affected (of 0 checked): Installed packages to be REINSTALLED: cups-1.7.3 [FreeBSD] cups-pstoraster-8.15.4_8 [FreeBSD] cups-base-1.7.3_1 [FreeBSD] (options changed) 9 MB to be downloaded. Proceed with this action? [y/N]: y So... what happened to the other packages? It looks like pkg is completely ignoring 3 other packages from the original list. Indeed, trying to upgrade just this one package missed by "pkg upgrade above" says: # pkg upgrade -f cups-image-1.7.2 Updating FreeBSD repository catalogue... FreeBSD repository is up-to-date. Updating FreeBSD_new_xorg repository catalogue... FreeBSD_new_xorg repository is up-to-date. All repositories are up-to-date. Checking integrity... done (0 conflicting) Your packages are up to date. Erm, no, this package is NOT up to date: # pkg search cups-image cups-image-1.7.3_1 cups-image-1.7.3_1 I presume the package is listed twice because it appears in two repositories. (but this problem is NOT specific to multiple-repositories issues, the same happens with e.g. p5* packages which are present only in the official repo). (this is pkg 1.3.7 on FreeBSD 10) From listjm at club-internet.fr Fri Sep 5 16:57:45 2014 From: listjm at club-internet.fr (Juan =?iso-8859-1?b?UmFt824=?= Molina Menor) Date: Fri, 05 Sep 2014 18:52:01 +0200 Subject: pkg upgrade -f refusing to upgrade packages??? Message-ID: <5409EA31.9050506@club-internet.fr> > Hello, > > I seem to keep either finding problems in pkg or completely missing > the point of it :( > > > > Now, I would like to force-upgrade them all: > > # pkg upgrade -f `pkg info -g 'cups*'` As far as I know and the manual says, the -f option forces the reinstallation or upgrade of the whole set of packages, this ignoring your provided pkg-names. > Updating FreeBSD repository catalogue... > FreeBSD repository is up-to-date. > Updating FreeBSD_new_xorg repository catalogue... > FreeBSD_new_xorg repository is up-to-date. > All repositories are up-to-date. > The following 3 packages will be affected (of 0 checked): ?of 0 checked? This is maybe not related to your issue, but at least could remove one source of confusion. Did you issued a "pkg update -f" after upgrading to pkg 1.3.7? Hope it helps, Juan From adamw at adamw.org Fri Sep 5 19:14:45 2014 From: adamw at adamw.org (Adam Weinberger) Date: Fri, 5 Sep 2014 15:14:35 -0400 Subject: poudriere nginx.conf.sample patch Message-ID: <464A3D56-AEB6-429A-9CF8-C5B477A391EF@adamw.org> Hi, The nginx.conf.sample code sample creates errors on modern nginx. It just needs a simple reformatting. https://people.freebsd.org/~adamw/poudriere-nginx.patch I tried to make an account on fossil.etoilebsd.net but I keep getting 502 bad gateway errors. # Adam -- Adam Weinberger adamw at adamw.org http://www.adamw.org From adamw at adamw.org Fri Sep 5 21:23:48 2014 From: adamw at adamw.org (Adam Weinberger) Date: Fri, 5 Sep 2014 17:23:44 -0400 Subject: poudriere wiki patch Message-ID: This adds some poudriere tips to the wiki. Gives some information about what sorts of -v VERSION options you can give, and gives step-by-step instructions to copying an existing ports tree and jail for testing. https://people.freebsd.org/~adamw/poudriere-wiki.patch # Adam -- Adam Weinberger adamw at adamw.org http://www.adamw.org From bugzilla-noreply at freebsd.org Sat Sep 6 01:59:09 2014 From: bugzilla-noreply at freebsd.org (bugzilla-noreply at freebsd.org) Date: Sat, 06 Sep 2014 01:59:09 +0000 Subject: [Bug 192922] ports-mgmt/pkg: "pkg install" raises jemalloc assertion for bitmap_get() In-Reply-To: References: Message-ID: https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=192922 Mark Linimon changed: What |Removed |Added ---------------------------------------------------------------------------- Component|bin |Individual Port(s) Version|11.0-CURRENT |Latest Assignee|freebsd-bugs at FreeBSD.org |pkg at FreeBSD.org Product|Base System |Ports Tree Summary|"pkg install" raises |ports-mgmt/pkg: "pkg |jemalloc assertion for |install" raises jemalloc |bitmap_get() |assertion for bitmap_get() --- Comment #1 from Mark Linimon --- The pkg(8) command in base is just a kickstarter. Reclassify PR and assign. -- You are receiving this mail because: You are the assignee for the bug. From rodrigc at freebsd.org Sun Sep 7 23:04:02 2014 From: rodrigc at freebsd.org (Craig Rodrigues) Date: Sun, 7 Sep 2014 16:03:59 -0700 Subject: make -DNO_ROOT to create chroot, problem installing into chroot with pkg Message-ID: Hi, I am using pkg 1.3.7. I did the following as a regular user, not root: rm -fr /tmp/package cd /usr/src make buildworld make buildkernel make -DNO_ROOT -DDB_FROM_SRC installworld DESTDIR=/tmp/package make -DNO_ROOT -DDB_FROM_SRC installkernel DESTDIR=/tmp/package make -DNO_ROOT -DDB_FROM_SRC distribution DESTDIR=/tmp/package This created an installed world under /tmp/package Then I did: pkg -c /tmp/package install -y devel/kyua I got: pkg: chroot failed! Then I tried the same command under sudo: sudo pkg -c /tmp/package install -y devel/kyua I got: pkg: /var/db/pkg wrong user or group ownership (expected 0/0 versus actual 818/0) Is there a way to install packages into chroot without being root? -- Craig From rodrigc at freebsd.org Sun Sep 7 23:48:47 2014 From: rodrigc at freebsd.org (Craig Rodrigues) Date: Sun, 7 Sep 2014 16:48:44 -0700 Subject: pkg upgrade -f refusing to upgrade packages??? In-Reply-To: References: Message-ID: On Fri, Sep 5, 2014 at 7:36 AM, Ivan Voras wrote: > Hello, > > I seem to keep either finding problems in pkg or completely missing > the point of it :( > > Here's another issue: > > # pkg info -g 'cups*' > cups-1.7.3 > cups-base-1.7.3_1 > cups-client-1.7.3 > cups-filters-1.0.55 > cups-image-1.7.2 > cups-pstoraster-8.15.4_8 > > Right, so my interpretation of the output above is that I have that > set of packages installed on this system. Now, I would like to > force-upgrade them all: > > # pkg upgrade -f `pkg info -g 'cups*'` > That invocation is wrong. If you type: pkg help upgrade You will see: "pkg upgrade is used for upgrading packaged software distributions." If you read the man page, you will see that pkg upgrade does not take individual package names as arguments. So you can either do one of the following two options: OPTION 1: Type: pkg upgrade -> this will upgrade all the packages on your system to the latest package set. This is preferred. OPTION 2: Type: pkg info -o -g 'cups*' cups-client-1.7.3 print/cups-client cups-image-1.7.3_1 print/cups-image Looking at the second column, type: pkg install print/cups-client print/cups-image This will download the latest cups-client and cups-image and install/upgrade them. I found the usage of "pkg upgrade" and "pkg update" a bit confusing at first. -- Craig From vas at mpeks.tomsk.su Mon Sep 8 02:24:37 2014 From: vas at mpeks.tomsk.su (Victor Sudakov) Date: Mon, 8 Sep 2014 09:24:32 +0700 Subject: pkg upgrade -f refusing to upgrade packages??? In-Reply-To: References: Message-ID: <20140908022432.GA413@admin.sibptus.tomsk.ru> Craig Rodrigues wrote: [dd] > > If you read the man page, you will see that pkg upgrade does not take > individual package names as arguments. Are you sure about that? Quoting the man page for 1.3.7: SYNOPSIS pkg upgrade [-fInFqUy] [-r reponame] [-Cgix] [ ...] pkg upgrade [--{force,no-install-scripts,dry-run,fetch-only}] [--{quiet,no-repo-update,yes}] [--repository reponame] [--{case-sensitive,glob,case-insensitive,regex}] [ ...] -- Victor Sudakov, VAS4-RIPE, VAS47-RIPN sip:sudakov at sibptus.tomsk.ru From matthew at FreeBSD.org Mon Sep 8 05:13:20 2014 From: matthew at FreeBSD.org (Matthew Seaman) Date: Mon, 08 Sep 2014 06:12:56 +0100 Subject: pkg upgrade -f refusing to upgrade packages??? In-Reply-To: <20140908022432.GA413@admin.sibptus.tomsk.ru> References: <20140908022432.GA413@admin.sibptus.tomsk.ru> Message-ID: <540D3AD8.1030608@FreeBSD.org> On 08/09/2014 03:24, Victor Sudakov wrote: > Craig Rodrigues wrote: >> > If you read the man page, you will see that pkg upgrade does not take >> > individual package names as arguments. > Are you sure about that? Quoting the man page for 1.3.7: > > SYNOPSIS > pkg upgrade [-fInFqUy] [-r reponame] [-Cgix] > [ ...] > > pkg upgrade [--{force,no-install-scripts,dry-run,fetch-only}] > [--{quiet,no-repo-update,yes}] [--repository reponame] > [--{case-sensitive,glob,case-insensitive,regex}] > [ ...] pkg upgrade being able to take package name arguments is new in 1.3.x Cheers, Matthew -- Dr Matthew J Seaman MA, D.Phil. PGP: http://www.infracaninophile.co.uk/pgpkey -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 971 bytes Desc: OpenPGP digital signature URL: From geeohgeegeeoh at gmail.com Mon Sep 8 05:41:45 2014 From: geeohgeegeeoh at gmail.com (George Michaelson) Date: Mon, 8 Sep 2014 15:41:44 +1000 Subject: how do I get apache24 to stay installed with mod_mpm_event and mod_python? Message-ID: I have hand-built apache24 with mod_mpm_event and mod_python. I selected the 'all modules as shared libs' option in make configure, in a belief this would externalize the mod_ stuff so I had an otherwise "normal" apache install. It would seem thats not so. When I update/upgrade, I get overwritten by pkg apache24 which does not have this module, nor the mod_python dependency I need. This then requires me to hand install again from /usr/ports/www/apache24 having removed work/.install*done Whats the canonical recipe to tell pkg and port to "play nice" so you can have non-standard builds of things? I tried looking, but the question "how do I make pkg and port play nice" didn't seem to work. Whats the search term I should have used? Seems to me like a plausible FAQ. From ggm at pobox.com Mon Sep 8 05:42:17 2014 From: ggm at pobox.com (George Michaelson) Date: Mon, 8 Sep 2014 15:42:16 +1000 Subject: xemacs (zombie port, but worked) Message-ID: Is it possible to have xemacs restored as a port/pkg, and also have it not collide with gnu emacs? I have to try and support a user semi-addicted to the xemacs flavour. It was stable, and worked. Did it have to die? (pkg removed it) From m.seaman at infracaninophile.co.uk Mon Sep 8 06:16:48 2014 From: m.seaman at infracaninophile.co.uk (Matthew Seaman) Date: Mon, 08 Sep 2014 07:16:34 +0100 Subject: xemacs (zombie port, but worked) In-Reply-To: References: Message-ID: <540D49C2.20807@infracaninophile.co.uk> On 08/09/2014 06:42, George Michaelson wrote: > Is it possible to have xemacs restored as a port/pkg, and also have it > not collide with gnu emacs? > > I have to try and support a user semi-addicted to the xemacs flavour. > It was stable, and worked. > > Did it have to die? (pkg removed it) This is really a question for the freebsd-ports@ list. xemacs was removed not specifically because of pkg(8) but because no-one cared enough about it to bring it up to the expected current standards for a port. It's not going to magically reappear unless someone with an interest in having it in the ports tree steps up and knocks it into shape. That's not hugely difficult, but it does require some skills with using subversion plus ability with shell and makefile programming. Cheers, Matthew -- Dr Matthew J Seaman MA, D.Phil. PGP: http://www.infracaninophile.co.uk/pgpkey JID: matthew at infracaninophile.co.uk -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 971 bytes Desc: OpenPGP digital signature URL: From m.seaman at infracaninophile.co.uk Mon Sep 8 06:25:59 2014 From: m.seaman at infracaninophile.co.uk (Matthew Seaman) Date: Mon, 08 Sep 2014 07:25:49 +0100 Subject: how do I get apache24 to stay installed with mod_mpm_event and mod_python? In-Reply-To: References: Message-ID: <540D4BED.5060008@infracaninophile.co.uk> On 08/09/2014 06:41, George Michaelson wrote: > Whats the canonical recipe to tell pkg and port to "play nice" so you > can have non-standard builds of things? You can hint to pkg(8) to tell it which repository it should install a package from by using pkg-annotate(8) -- see the 'WORKING WITH MULTIPLE REPOSITORIES' section pkg-repository(5). Beyond that, you can use pkg-lock(8) to force pkg(8) not to overwrite your customised installs with the defaults from the pkgrepo. Having to use pkg-lock(8) to override the logic of the dependency solver is not ideal, but sometimes needs must. Right now, installing packages from ports is not treated entirely equivalently as using a binary package repository. That's an area which is currently under development: handling various different types of package sources coherently. Cheers, Matthew -- Dr Matthew J Seaman MA, D.Phil. PGP: http://www.infracaninophile.co.uk/pgpkey JID: matthew at infracaninophile.co.uk -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 971 bytes Desc: OpenPGP digital signature URL: From geeohgeegeeoh at gmail.com Mon Sep 8 06:46:41 2014 From: geeohgeegeeoh at gmail.com (George Michaelson) Date: Mon, 8 Sep 2014 16:46:39 +1000 Subject: xemacs (zombie port, but worked) In-Reply-To: <540D49C2.20807@infracaninophile.co.uk> References: <540D49C2.20807@infracaninophile.co.uk> Message-ID: thanks. I've re-asked there without a cc so this doesn't get stupid. -G On Mon, Sep 8, 2014 at 4:16 PM, Matthew Seaman wrote: > On 08/09/2014 06:42, George Michaelson wrote: >> Is it possible to have xemacs restored as a port/pkg, and also have it >> not collide with gnu emacs? >> >> I have to try and support a user semi-addicted to the xemacs flavour. >> It was stable, and worked. >> >> Did it have to die? (pkg removed it) > > This is really a question for the freebsd-ports@ list. xemacs was > removed not specifically because of pkg(8) but because no-one cared > enough about it to bring it up to the expected current standards for a > port. It's not going to magically reappear unless someone with an > interest in having it in the ports tree steps up and knocks it into > shape. That's not hugely difficult, but it does require some skills > with using subversion plus ability with shell and makefile programming. > > Cheers, > > Matthew > > -- > Dr Matthew J Seaman MA, D.Phil. > > PGP: http://www.infracaninophile.co.uk/pgpkey > JID: matthew at infracaninophile.co.uk > From ivoras at freebsd.org Mon Sep 8 09:37:41 2014 From: ivoras at freebsd.org (Ivan Voras) Date: Mon, 8 Sep 2014 11:36:58 +0200 Subject: pkg upgrade -f refusing to upgrade packages??? In-Reply-To: References: Message-ID: On 8 September 2014 01:48, Craig Rodrigues wrote: > > > On Fri, Sep 5, 2014 at 7:36 AM, Ivan Voras wrote: >> >> Hello, >> >> I seem to keep either finding problems in pkg or completely missing >> the point of it :( >> >> Here's another issue: >> >> # pkg info -g 'cups*' >> cups-1.7.3 >> cups-base-1.7.3_1 >> cups-client-1.7.3 >> cups-filters-1.0.55 >> cups-image-1.7.2 >> cups-pstoraster-8.15.4_8 >> >> Right, so my interpretation of the output above is that I have that >> set of packages installed on this system. Now, I would like to >> force-upgrade them all: >> >> # pkg upgrade -f `pkg info -g 'cups*'` > > > That invocation is wrong. > > If you type: > > pkg help upgrade > > You will see: > > "pkg upgrade is used for upgrading packaged software distributions." > > If you read the man page, you will see that pkg upgrade does not take > individual package names as arguments. Really? My man page says otherwise: pkg upgrade [-fInFqUy] [-r reponame] [-Cgix] [ ...] ... pkg upgrade compares the versions of all or specific packages installed on the system to what is available in the configured package reposito- ries. Any out-of-date packages are added to a work list for processing. Note the "...or specific packages...". I think what you say was true for early versions of pkg, but sometime recently this has changed (for the better). From vas at mpeks.tomsk.su Mon Sep 8 10:30:56 2014 From: vas at mpeks.tomsk.su (Victor Sudakov) Date: Mon, 8 Sep 2014 17:30:52 +0700 Subject: "pkg search" gives different results, or host does not see a package Message-ID: <20140908103052.GA12090@admin.sibptus.tomsk.ru> Colleagues, There are two boxes running FreeBSD 9.3 using the same repository (built with poudriere). However, the output of "pkg search" is different: [root at admin ~] pkg search postgre postgresql92-client-9.2.9 postgresql92-server-9.2.9 root at pg02-sibptus ~] pkg search postgre postgresql90-client-9.0.18 postgresql90-server-9.0.18 postgresql92-client-9.2.9 postgresql92-server-9.2.9 "pkg -vv" reports the same repository on both hosts. I have already tried "pkg update -f" on the host "admin", which did not improve the situation. Still: [root at admin ~] pkg install -n postgresql90-server Updating sibptus repository catalogue... sibptus repository is up-to-date. All repositories are up-to-date. pkg: No packages available to install matching 'postgresql90-server' have been found in the repositories [root at admin ~] [root at pg02-sibptus ~] pkg install -n postgresql90-server Updating sibptus repository catalogue... sibptus repository is up-to-date. All repositories are up-to-date. The following 2 packages will be affected (of 0 checked): New packages to be INSTALLED: postgresql90-server: 9.0.18 postgresql90-client: 9.0.18 The process will require 14 MB more space. 3 MB to be downloaded. [root at pg02-sibptus ~] What gives? -- Victor Sudakov, VAS4-RIPE, VAS47-RIPN sip:sudakov at sibptus.tomsk.ru From radek.krejca at starnet.cz Mon Sep 8 12:40:09 2014 From: radek.krejca at starnet.cz (=?iso-8859-2?Q?Radek_Krej=E8a?=) Date: Mon, 8 Sep 2014 14:40:03 +0200 Subject: A lot of pkg problems Message-ID: Hello, I am trying to install my computer today (new hardware) and I cannot install a lot of packages, eg. firefox, thunderbird, chromium, xfce, libreoffice..... All is falling on message: Checking integrity... done (1 conflicting) pkg: Cannot solve problem using SAT solver: cannot install package firefox~www/firefox, remove it from request? [Y/n]: All the same... I have deleted all packages and problem is still the same. My conf file is: # $FreeBSD: release/10.0.0/etc/pkg/FreeBSD.conf 258710 2013-11-28 14:24:26Z gjb $ FreeBSD: { url: "pkg+http://pkg.FreeBSD.org/${ABI}/latest", mirror_type: "srv", signature_type: "fingerprints", fingerprints: "/usr/share/keys/pkg", enabled: yes } Radek From brooks at freebsd.org Mon Sep 8 15:48:59 2014 From: brooks at freebsd.org (Brooks Davis) Date: Mon, 8 Sep 2014 15:48:58 +0000 Subject: make -DNO_ROOT to create chroot, problem installing into chroot with pkg In-Reply-To: References: Message-ID: <20140908154858.GB35236@spindle.one-eyed-alien.net> On Sun, Sep 07, 2014 at 04:03:59PM -0700, Craig Rodrigues wrote: > Hi, > > I am using pkg 1.3.7. > > I did the following as a regular user, not root: > > rm -fr /tmp/package > cd /usr/src > make buildworld > make buildkernel > make -DNO_ROOT -DDB_FROM_SRC installworld DESTDIR=/tmp/package > make -DNO_ROOT -DDB_FROM_SRC installkernel DESTDIR=/tmp/package > make -DNO_ROOT -DDB_FROM_SRC distribution DESTDIR=/tmp/package > > This created an installed world under /tmp/package > > Then I did: > > pkg -c /tmp/package install -y devel/kyua > > I got: > > pkg: chroot failed! > > Then I tried the same command under sudo: > > sudo pkg -c /tmp/package install -y devel/kyua > > I got: > > pkg: /var/db/pkg wrong user or group ownership (expected 0/0 versus actual > 818/0) > > Is there a way to install packages into chroot without > being root? If you don't mind the ownership being wrong and there being a few extra +FOO files tar works. It would be great for someone to teach package to install without root and to update a METALOG file. That's not 100% of the solution, but it's a solid 80-90% solution. -- Brooks -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 181 bytes Desc: not available URL: From jmmv at freebsd.org Mon Sep 8 17:07:07 2014 From: jmmv at freebsd.org (Julio Merino) Date: Mon, 8 Sep 2014 13:06:38 -0400 Subject: make -DNO_ROOT to create chroot, problem installing into chroot with pkg In-Reply-To: <20140908154858.GB35236@spindle.one-eyed-alien.net> References: <20140908154858.GB35236@spindle.one-eyed-alien.net> Message-ID: On Mon, Sep 8, 2014 at 11:48 AM, Brooks Davis wrote: > > If you don't mind the ownership being wrong and there being a few extra > +FOO files tar works. It would be great for someone to teach package to > install without root and to update a METALOG file. That's not 100% of > the solution, but it's a solid 80-90% solution. (This is just my completely uninformed opinion as I don't know the internals of pkg.) There are other issues to be addressed. Teaching pkg to install without root means changing pkg to not use chroot: i.e. to make pkg be able to deal with the concept of a "destdir" for package installations. That is probably easy: just prefixing a bunch of destdir to the locations being touched. However, the tricky part here is dealing with any package-specific post-install scripts to ensure they understand destdir, which in turn means that any tools executed by the scripts must also be capable of dealing with a destdir. Also, the scripts (being potentially-untrusted code) cannot be guaranteed to behave correctly on the outside-of-the-chroot system. From brooks at freebsd.org Mon Sep 8 18:36:48 2014 From: brooks at freebsd.org (Brooks Davis) Date: Mon, 8 Sep 2014 18:36:47 +0000 Subject: make -DNO_ROOT to create chroot, problem installing into chroot with pkg In-Reply-To: References: <20140908154858.GB35236@spindle.one-eyed-alien.net> Message-ID: <20140908183647.GD35236@spindle.one-eyed-alien.net> On Mon, Sep 08, 2014 at 01:06:38PM -0400, Julio Merino wrote: > On Mon, Sep 8, 2014 at 11:48 AM, Brooks Davis wrote: > > > > If you don't mind the ownership being wrong and there being a few extra > > +FOO files tar works. It would be great for someone to teach package to > > install without root and to update a METALOG file. That's not 100% of > > the solution, but it's a solid 80-90% solution. > > (This is just my completely uninformed opinion as I don't know the > internals of pkg.) There are other issues to be addressed. > > Teaching pkg to install without root means changing pkg to not use > chroot: i.e. to make pkg be able to deal with the concept of a > "destdir" for package installations. That is probably easy: just > prefixing a bunch of destdir to the locations being touched. This should be pretty easy given that most of the process is extracting files from the tarball. > However, the tricky part here is dealing with any package-specific > post-install scripts to ensure they understand destdir, which in turn > means that any tools executed by the scripts must also be capable of > dealing with a destdir. Also, the scripts (being potentially-untrusted > code) cannot be guaranteed to behave correctly on the > outside-of-the-chroot system. I believe the majority of packages don't suffer from post-install scripts hence the suggestion that extracting in the right place without root would solve 80-90% of the problem (and probably take no more than 10% of the work). I could live with the pain of not having scripts run during install. The correct long term fix as proposed by bapt is the do anyway with most scripts in favor of common actions and if any significant scripts remain add the ability to run them on first boot. -- Brooks -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 181 bytes Desc: not available URL: From rodrigc at freebsd.org Mon Sep 8 19:35:38 2014 From: rodrigc at freebsd.org (Craig Rodrigues) Date: Mon, 8 Sep 2014 12:35:35 -0700 Subject: make -DNO_ROOT to create chroot, problem installing into chroot with pkg In-Reply-To: <20140908154858.GB35236@spindle.one-eyed-alien.net> References: <20140908154858.GB35236@spindle.one-eyed-alien.net> Message-ID: On Mon, Sep 8, 2014 at 8:48 AM, Brooks Davis wrote: > > If you don't mind the ownership being wrong and there being a few extra > +FOO files tar works. It would be great for someone to teach package to > install without root and to update a METALOG file. That's not 100% of > the solution, but it's a solid 80-90% solution. > > I've opened this feature request against pkg: https://github.com/freebsd/pkg/issues/1002 Please add any comments/suggestions there, since you have very valuable input. -- Craig From marck at rinet.ru Mon Sep 8 20:54:08 2014 From: marck at rinet.ru (Dmitry Morozovsky) Date: Tue, 9 Sep 2014 00:53:55 +0400 (MSK) Subject: upgrading ports/packages: mixing portupgrade/portmaster with binary dependencies Message-ID: Dear colleagues, (and, yes, I know I should plan, prepare and deploy poudriere server for the most correct answer to my question ;-P) is there a shortcut way to source-upgrade ports which configured differently comparing to the master default, having install dependencies (which are usually needed only for building, not for running) from default pkg repository? something like old (pre-pkg era) portupgrade -a -PP ; portupgrade -a And, after all of this dance, it would be great to run ``pkg autoremove'' and see installed package list much shorter ;) Thanks in advance! -- Sincerely, D.Marck [DM5020, MCK-RIPE, DM3-RIPN] [ FreeBSD committer: marck at FreeBSD.org ] ------------------------------------------------------------------------ *** Dmitry Morozovsky --- D.Marck --- Wild Woozle --- marck at rinet.ru *** ------------------------------------------------------------------------ From matthew at FreeBSD.org Mon Sep 8 23:25:48 2014 From: matthew at FreeBSD.org (Matthew Seaman) Date: Mon, 08 Sep 2014 23:34:04 +0100 Subject: upgrading ports/packages: mixing portupgrade/portmaster with binary dependencies In-Reply-To: References: Message-ID: <540E2EDC.2040803@FreeBSD.org> On 08/09/2014 21:53, Dmitry Morozovsky wrote: > (and, yes, I know I should plan, prepare and deploy poudriere server for the > most correct answer to my question ;-P) > > is there a shortcut way to source-upgrade ports which configured differently > comparing to the master default, having install dependencies (which are usually > needed only for building, not for running) from default pkg repository? > > something like old (pre-pkg era) portupgrade -a -PP ; portupgrade -a > > And, after all of this dance, it would be great to run ``pkg autoremove'' and > see installed package list much shorter ;) Not yet. This sort of thing is definitely on the drawing board though. At the moment I guess it would be a small matter of writing some scripts to work out what build deps are needed, use pkg to install them from the repo and mark them for automatic removal while doing so, and then do a build from ports of the software you want to apply special options to. Or you could take the easy way out and run a poudriere instance to build just the ports you want custom options on. Well, except that will also build everything those ports depend on as well. But you should be able to use your custom repo in parallel with the official repos. Just be a bit careful to match the revision of the ports tree you build your custom stuff with to the revision used to build the latest official package set. Cheers, Matthew -- Dr Matthew J Seaman MA, D.Phil. PGP: http://www.infracaninophile.co.uk/pgpkey -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 971 bytes Desc: OpenPGP digital signature URL: From ggm at pobox.com Tue Sep 9 02:21:49 2014 From: ggm at pobox.com (George Michaelson) Date: Tue, 9 Sep 2014 10:43:59 +1000 Subject: how do I get apache24 to stay installed with mod_mpm_event and mod_python? In-Reply-To: <540D4BED.5060008@infracaninophile.co.uk> References: <540D4BED.5060008@infracaninophile.co.uk> Message-ID: I went with pkg lock for now. its simple. thanks for cluestick. On Mon, Sep 8, 2014 at 4:25 PM, Matthew Seaman wrote: > On 08/09/2014 06:41, George Michaelson wrote: >> Whats the canonical recipe to tell pkg and port to "play nice" so you >> can have non-standard builds of things? > > You can hint to pkg(8) to tell it which repository it should install a > package from by using pkg-annotate(8) -- see the 'WORKING WITH MULTIPLE > REPOSITORIES' section pkg-repository(5). > > Beyond that, you can use pkg-lock(8) to force pkg(8) not to overwrite > your customised installs with the defaults from the pkgrepo. Having to > use pkg-lock(8) to override the logic of the dependency solver is not > ideal, but sometimes needs must. > > Right now, installing packages from ports is not treated entirely > equivalently as using a binary package repository. That's an area which > is currently under development: handling various different types of > package sources coherently. > > Cheers, > > Matthew > > -- > Dr Matthew J Seaman MA, D.Phil. > > PGP: http://www.infracaninophile.co.uk/pgpkey > JID: matthew at infracaninophile.co.uk > From jmmv at freebsd.org Tue Sep 9 21:22:43 2014 From: jmmv at freebsd.org (Julio Merino) Date: Tue, 9 Sep 2014 16:51:40 -0400 Subject: make -DNO_ROOT to create chroot, problem installing into chroot with pkg In-Reply-To: <20140908183647.GD35236@spindle.one-eyed-alien.net> References: <20140908154858.GB35236@spindle.one-eyed-alien.net> <20140908183647.GD35236@spindle.one-eyed-alien.net> Message-ID: On Mon, Sep 8, 2014 at 2:36 PM, Brooks Davis wrote: > I believe the majority of packages don't suffer from post-install > scripts hence the suggestion that extracting in the right place without > root would solve 80-90% of the problem (and probably take no more than > 10% of the work). I could live with the pain of not having scripts run > during install. The correct long term fix as proposed by bapt is the do > anyway with most scripts in favor of common actions and if any > significant scripts remain add the ability to run them on first boot. Cool; glad to hear. This sounds like a good plan. From papowell at astart.com Wed Sep 10 15:34:04 2014 From: papowell at astart.com (Patrick Powell) Date: Wed, 10 Sep 2014 08:33:56 -0700 Subject: make -DNO_ROOT to create chroot, problem installing into chroot with pkg In-Reply-To: References: <20140908154858.GB35236@spindle.one-eyed-alien.net> <20140908183647.GD35236@spindle.one-eyed-alien.net> Message-ID: <54106F64.3050305@astart.com> On 09/09/14 13:51, Julio Merino wrote: > On Mon, Sep 8, 2014 at 2:36 PM, Brooks Davis wrote: >> I believe the majority of packages don't suffer from post-install >> scripts hence the suggestion that extracting in the right place without >> root would solve 80-90% of the problem (and probably take no more than >> 10% of the work). I could live with the pain of not having scripts run >> during install. The correct long term fix as proposed by bapt is the do >> anyway with most scripts in favor of common actions and if any >> significant scripts remain add the ability to run them on first boot. > Cool; glad to hear. This sounds like a good plan. > _______________________________________________ > freebsd-pkg at freebsd.org mailing list > https://lists.freebsd.org/mailman/listinfo/freebsd-pkg > To unsubscribe, send any mail to "freebsd-pkg-unsubscribe at freebsd.org" > Having dealt with 'run on first boot' installation issues, I don't think that this is a good idea. Have you ever tried to debug one of these little beasts? And make sure that the script is removed so that it does not run again? And deal with adding a whole new infrastructure to deal with this SEPARATE from the current FreeBSD configuration? I lived through the headaches dealing with startup scripts having to be added to things like /etc/rc.local and finally seeing the consensus of using /usr/local/etc/rc.d/xxxx for separate scripts... And now you want to add something like: /usr/local/etc/run_on_first_boot/xxxx? Please reconsider adding 'run on first boot' capabilities. From mikej at mikej.com Wed Sep 10 16:34:49 2014 From: mikej at mikej.com (Michael Jung) Date: Wed, 10 Sep 2014 12:32:53 -0400 Subject: "poudriere buk" - why does it need to fail on single package with missing port =?UTF-8?Q?origin=3F?= Message-ID: Why does "poudriere bulk" need to bail because of a single package that has a missing port origin and not build other packages in the package list? If I remove x11/mate from my package list poudriere rocks on building all my other ports. The missing origin is a RUN_DEPENDS in the x11/mate Makefile, but that dependency should only be for x11/mate right? I can't seem to find a good explanation in the general man pages or searching the web. Educate me please ;-) --mikej Example: poudriere 3.0.17/11.0-CURRENT #1 r264318M amd64 [root at bsd11 /usr/local/poudriere/ports/default/x11/mate]# poudriere bulk -j 10stable -f /usr/local/etc/charon-list ====>> Creating the reference jail... done ====>> Mounting system devices for 10stable-default ====>> Mounting ports/packages/distfiles ====>> Mounting ccache from: /var/cache/ccache ====>> Mounting packages from: /usr/local/poudriere/data/packages/10stable-default ====>> Logs: /usr/local/poudriere/data/logs/bulk/10stable-default/2014-09-10_12h18m15s ====>> Appending to make.conf: /usr/local/etc/poudriere.d/10stable-make.conf /etc/resolv.conf -> /usr/local/poudriere/data/build/10stable-default/ref/etc/resolv.conf ====>> Starting jail 10stable-default ====>> Loading MOVED ====>> Calculating ports order and dependencies ====>> MOVED: databases/db41 renamed to databases/db48 ====>> MOVED: databases/db42 renamed to databases/db48 ====>> Error: Invalid port origin 'archivers/mate-file-archiver' not found. ====>> Cleaning up ====>> Umounting file systems [root at bsd11 /usr/local/poudriere/ports/default/x11/mate]# AFTER remove x11/mate from the list "charon-list" [root at bsd11 /usr/local/poudriere/ports/default/x11/mate]# poudriere bulk -j 10stable -f /usr/local/etc/charon-list ====>> Creating the reference jail... done ====>> Mounting system devices for 10stable-default ====>> Mounting ports/packages/distfiles ====>> Mounting ccache from: /var/cache/ccache ====>> Mounting packages from: /usr/local/poudriere/data/packages/10stable-default ====>> Logs: /usr/local/poudriere/data/logs/bulk/10stable-default/2014-09-10_12h20m14s ====>> Appending to make.conf: /usr/local/etc/poudriere.d/10stable-make.conf /etc/resolv.conf -> /usr/local/poudriere/data/build/10stable-default/ref/etc/resolv.conf ====>> Starting jail 10stable-default ====>> Loading MOVED ====>> Calculating ports order and dependencies ====>> MOVED: databases/db41 renamed to databases/db48 ====>> MOVED: databases/db42 renamed to databases/db48 ====>> Sanity checking the repository ====>> Deleting old version: liboil-0.3.17_1.txz ====>> Deleting old version: p5-DBD-Pg-3.4.0.txz ====>> Deleting old version: p5-Net-SSLeay-1.65.txz ====>> Deleting stale symlinks ====>> Deleting empty directories ====>> Cleaning the build queue ====>> Building 490 packages using 8 builders ====>> Starting/Cloning builders ====>> Hit CTRL+t at any time to see build progress and stats From bugzilla-noreply at freebsd.org Wed Sep 10 17:50:15 2014 From: bugzilla-noreply at freebsd.org (bugzilla-noreply at freebsd.org) Date: Wed, 10 Sep 2014 17:50:15 +0000 Subject: [Bug 193530] New: ports-mgmt/pkg - Error on pkg search -Q required-by Message-ID: https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=193530 Bug ID: 193530 Summary: ports-mgmt/pkg - Error on pkg search -Q required-by Product: Ports Tree Version: Latest Hardware: Any OS: Any Status: Needs Triage Severity: Affects Many People Priority: --- Component: Individual Port(s) Assignee: pkg at FreeBSD.org Reporter: adamw at FreeBSD.org [root at lockup ~] pkg search -Q required-by gettext pkg: sqlite error while executing SELECT p.name, p.origin, p.version, 0 FROM main.packages AS p INNER JOIN main.deps AS d ON p.id = d.package_id WHERE d.name = SPLIT_UID('name', ?1) AND d.origin = SPLIT_UID('origin', ?1); in file pkgdb_iterator.c:277: SQL function split_*() called with invalid arguments. (70)[root at lockup ~] pkg info pkg pkg-1.3.7 Name : pkg Version : 1.3.7 Installed on : Wed Aug 27 10:48:54 EDT 2014 Origin : ports-mgmt/pkg Architecture : freebsd:10:x86:64 Prefix : /usr/local [root at lockup ~] uname -a FreeBSD lockup 10.1-PRERELEASE FreeBSD 10.1-PRERELEASE #4: Tue Sep 9 10:31:25 EDT 2014 root at lockup:/usr/obj/usr/src/sys/GENERIC amd64 -- You are receiving this mail because: You are the assignee for the bug. From mexas at bris.ac.uk Sun Sep 14 08:46:22 2014 From: mexas at bris.ac.uk (Anton Shterenlikht) Date: Sun, 14 Sep 2014 09:45:48 +0100 (BST) Subject: firefox: pkg: Cannot solve problem using SAT solver Message-ID: <201409140845.s8E8jmkC000795@mech-as221.men.bris.ac.uk> I'm in a jail under amd64 9.2-RELEASE. # pkg info -xo pkg- pkg-1.3.7 ports-mgmt/pkg # pkg check -srda Checking all packages: 100% # # pkg install firefox Updating FreeBSD repository catalogue... FreeBSD repository is up-to-date. All repositories are up-to-date. Checking integrity... done (1 conflicting) pkg: Cannot solve problem using SAT solver: cannot install package firefox~www/firefox, remove it from request? [Y/n]: y Checking integrity... done (0 conflicting) The most recent version of packages are already installed # Have I forgotten something? Thanks Anton From matthew at FreeBSD.org Sun Sep 14 08:59:03 2014 From: matthew at FreeBSD.org (Matthew Seaman) Date: Sun, 14 Sep 2014 09:58:44 +0100 Subject: firefox: pkg: Cannot solve problem using SAT solver In-Reply-To: <201409140845.s8E8jmkC000795@mech-as221.men.bris.ac.uk> References: <201409140845.s8E8jmkC000795@mech-as221.men.bris.ac.uk> Message-ID: <541558C4.507@FreeBSD.org> On 14/09/2014 09:45, Anton Shterenlikht wrote: > # pkg install firefox > Updating FreeBSD repository catalogue... > FreeBSD repository is up-to-date. > All repositories are up-to-date. > Checking integrity... done (1 conflicting) > pkg: Cannot solve problem using SAT solver: > cannot install package firefox~www/firefox, remove it from request? [Y/n]: y > Checking integrity... done (0 conflicting) > The most recent version of packages are already installed > # > > Have I forgotten something? You're the second person to report this. It does seem to be a problem with the firefox port in the FreeBSD repository. Looks like it conflicts with it's own dependencies. I don't know if this week's packages have been released yet (should be some time around now usually) -- all I can suggest though is that you wait for an updated package to appear and then try again. Other than building firefox yourself that is. Cheers, Matthew -- Dr Matthew J Seaman MA, D.Phil. PGP: http://www.infracaninophile.co.uk/pgpkey -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 971 bytes Desc: OpenPGP digital signature URL: From radek.krejca at starnet.cz Sun Sep 14 11:02:49 2014 From: radek.krejca at starnet.cz (=?utf-8?B?UmFkZWsgS3JlasSNYQ==?=) Date: Sun, 14 Sep 2014 13:02:45 +0200 Subject: firefox: pkg: Cannot solve problem using SAT solver In-Reply-To: <541558C4.507@FreeBSD.org> References: <201409140845.s8E8jmkC000795@mech-as221.men.bris.ac.uk> <541558C4.507@FreeBSD.org> Message-ID: Hello, > I don't know if this week's packages have been released yet (should be > some time around now usually) -- all I can suggest though is that you > wait for an updated package to appear and then try again. Other than > building firefox yourself that is. > > Cheers, > > Matthew This is not only firefox problem. I know (from my head, because I im installing at this moment from ports): php5-gd thunderbird firefox xfce4 (some parts) chromium ...... Its a lot of packages. Radek From radek.krejca at starnet.cz Sun Sep 14 11:03:23 2014 From: radek.krejca at starnet.cz (=?iso-8859-2?Q?Radek_Krej=E8a?=) Date: Sun, 14 Sep 2014 13:03:20 +0200 Subject: firefox: pkg: Cannot solve problem using SAT solver In-Reply-To: References: <201409140845.s8E8jmkC000795@mech-as221.men.bris.ac.uk> <541558C4.507@FreeBSD.org> Message-ID: And more: Checking integrity... done (1 conflicting) pkg: Cannot solve problem using SAT solver: cannot install package phpeclipse~java/phpeclipse From radek.krejca at starnet.cz Sun Sep 14 16:59:16 2014 From: radek.krejca at starnet.cz (=?utf-8?B?UmFkZWsgS3JlasSNYQ==?=) Date: Sun, 14 Sep 2014 18:59:11 +0200 Subject: firefox: pkg: Cannot solve problem using SAT solver In-Reply-To: <541558C4.507@FreeBSD.org> References: <201409140845.s8E8jmkC000795@mech-as221.men.bris.ac.uk> <541558C4.507@FreeBSD.org> Message-ID: > > Have I forgotten something? > > You're the second person to report this. It does seem to be a problem > with the firefox port in the FreeBSD repository. Looks like it > conflicts with it's own dependencies. It looks, that problem should be in jpeg-turbo, because some packages need jpeg and others jpeg-turbo. But I dont know if is the reason. And this situation is about 3 weeks old. Radek From kaltheat at googlemail.com Mon Sep 15 13:54:47 2014 From: kaltheat at googlemail.com (kaltheat) Date: Mon, 15 Sep 2014 15:54:39 +0200 Subject: Disable pkg cache Message-ID: <148799807a5.d2f58d1b193594.6509254985329847815@googlemail.com> Hi, is it possible to disable local caching of packages? I have my own local repo and don't want to "pollute" each server in my environment with pkg cache. Regards, kaltheat From matthew at freebsd.org Mon Sep 15 14:49:47 2014 From: matthew at freebsd.org (Matthew Seaman) Date: Mon, 15 Sep 2014 15:49:28 +0100 Subject: Disable pkg cache In-Reply-To: <148799807a5.d2f58d1b193594.6509254985329847815@googlemail.com> References: <148799807a5.d2f58d1b193594.6509254985329847815@googlemail.com> Message-ID: <5416FC78.4040104@freebsd.org> On 09/15/14 14:54, kaltheat wrote: > is it possible to disable local caching of packages? > I have my own local repo and don't want to "pollute" each server in my environment with pkg cache. Not -- the cache can't be disabled. However, your concern about 'pollution' is misplaced. The package cache won't overwrite package foo-1.0.0.txz from repoA with foo-1.0.0.txz from repoB --- it knows exactly which repository each cached package came from and it can keep multiple copies from separate sources simultaneously. The only reason it would supercede any package is when it sees there is an update available in the repository catalogue with a different checksum. Even then it keeps the older version of the package n hand. Eg: This is on my desktop. You can see there are 4 different packages for xscreensaver in the cache, covering three different versions of the port. The fourth will be due to a change in a dependency triggerring a rebuild by poudriere. # ls -l /var/cache/pkg/xscreensaver* -rw-r--r-- 1 root wheel 3969844 Aug 29 00:49 /var/cache/pkg/xscreensaver-5.29_2-32189fcfe9.txz -rw-r--r-- 1 root wheel 3971808 Aug 5 16:41 /var/cache/pkg/xscreensaver-5.29_2-45c078f215.txz lrwxr-xr-x 1 root wheel 34 Aug 29 10:43 /var/cache/pkg/xscreensaver-5.29_2.txz@ -> xscreensaver-5.29_2-32189fcfe9.txz -rw-r--r-- 1 root wheel 4010252 Aug 31 00:44 /var/cache/pkg/xscreensaver-5.29_3-7cedad5fcc.txz lrwxr-xr-x 1 root wheel 34 Sep 1 10:03 /var/cache/pkg/xscreensaver-5.29_3.txz@ -> xscreensaver-5.29_3-7cedad5fcc.txz -rw-r--r-- 1 root wheel 4008892 Sep 11 01:09 /var/cache/pkg/xscreensaver-5.29_4-45227659e8.txz lrwxr-xr-x 1 root wheel 34 Sep 11 09:22 /var/cache/pkg/xscreensaver-5.29_4.txz@ -> xscreensaver-5.29_4-45227659e8.txz PS. If you're worried about the cache getting too big, that's not usually a problem and even if it is, 'pkg clean' will sort thing out pronto. Cheers, Matthew -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 949 bytes Desc: OpenPGP digital signature URL: From kaltheat at googlemail.com Tue Sep 16 06:14:38 2014 From: kaltheat at googlemail.com (kaltheat) Date: Tue, 16 Sep 2014 08:14:36 +0200 Subject: Disable pkg cache In-Reply-To: <5416FC78.4040104@freebsd.org> References: <148799807a5.d2f58d1b193594.6509254985329847815@googlemail.com> <5416FC78.4040104@freebsd.org> Message-ID: <1487d193309.11bd56190294561.6586089392236130421@googlemail.com> ---- On Mon, 15 Sep 2014 16:49:28 +0200 Matthew Seaman wrote ---- >On 09/15/14 14:54, kaltheat wrote: >> is it possible to disable local caching of packages? >> I have my own local repo and don't want to "pollute" each server in my environment with pkg cache. > >Not -- the cache can't be disabled. However, your concern about >'pollution' is misplaced. The package cache won't overwrite package >foo-1.0.0.txz from repoA with foo-1.0.0.txz from repoB --- it knows >exactly which repository each cached package came from and it can keep >multiple copies from separate sources simultaneously. The only reason >it would supercede any package is when it sees there is an update >available in the repository catalogue with a different checksum. Even >then it keeps the older version of the package n hand. Well, what I meant with "pollution" was: I already have all packages on the package building server. I don't need packages locally on each server. 'pkg clean' would be a workaround, but it would be fine if pkg can be configured to not create a cache at all ... So I need to file a feature request, right? >PS. If you're worried about the cache getting too big, that's not >usually a problem and even if it is, 'pkg clean' will sort thing out pronto. To me it's a kind of aesthetic concern: I don't want to have redundancies if they aren't necessary. Regards, kaltheat From kaltheat at googlemail.com Tue Sep 16 07:29:29 2014 From: kaltheat at googlemail.com (kaltheat) Date: Tue, 16 Sep 2014 09:29:27 +0200 Subject: Disable pkg cache In-Reply-To: <1487d193309.11bd56190294561.6586089392236130421@googlemail.com> References: <148799807a5.d2f58d1b193594.6509254985329847815@googlemail.com> <5416FC78.4040104@freebsd.org> <1487d193309.11bd56190294561.6586089392236130421@googlemail.com> Message-ID: <1487d5db8b1.af034ce6300807.6233164589363762316@googlemail.com> ---- On Tue, 16 Sep 2014 08:14:36 +0200 kaltheat wrote ---- >---- On Mon, 15 Sep 2014 16:49:28 +0200 Matthew Seaman wrote ---- >>On 09/15/14 14:54, kaltheat wrote: >>> is it possible to disable local caching of packages? >>> I have my own local repo and don't want to "pollute" each server in my environment with pkg cache. >> >>Not -- the cache can't be disabled. However, your concern about >>'pollution' is misplaced. The package cache won't overwrite package >>foo-1.0.0.txz from repoA with foo-1.0.0.txz from repoB --- it knows >>exactly which repository each cached package came from and it can keep >>multiple copies from separate sources simultaneously. The only reason >>it would supercede any package is when it sees there is an update >>available in the repository catalogue with a different checksum. Even >>then it keeps the older version of the package n hand. > >Well, what I meant with "pollution" was: >I already have all packages on the package building server. I don't need packages locally on each server. 'pkg clean' would be a workaround, but it would be fine if pkg can be configured to not create a cache at all ... >So I need to file a feature request, right? > >>PS. If you're worried about the cache getting too big, that's not >>usually a problem and even if it is, 'pkg clean' will sort thing out pronto. > >To me it's a kind of aesthetic concern: I don't want to have redundancies if they aren't necessary. > Ah, they will work on it ... https://github.com/freebsd/pkg/issues/547 From vaclavhosek at volny.cz Wed Sep 17 08:47:44 2014 From: vaclavhosek at volny.cz (=?ISO-8859-2?Q?V=E1clav_Ho=B9ek?=) Date: Wed, 17 Sep 2014 10:47:37 +0200 Subject: Registering installation for dialog4ports-0.1.5_2,*** [fake-pkg] Error code 74 Message-ID: <54194AA9.8070604@volny.cz> I apologise because I am new in freebsd. I can install serviio as a jail in freenas V9.2.1.5 using instructins in http://www.mysticg.com/2013/01/07/freenas-jail-with-serviio/ and till end of August it works fine, but in September I cannot update packages using command: cd /usr/ports/ports-mgmt/portmaster/ && make install clean after this command is invoked it ends with message: Registering installation for dialog4ports-0.1.5_2 *** [fake-pkg] Error code 74 Stop in /usr/ports/ports-mgmt/dialog4ports. *** [install] Error code 1 and installation ends. I suppose that package dialog4ports-0.1.5_2 is missing, but can anybody help me in installation? The whole log is in attachment. Thanks Vaclav Hosek --- Tato zpr?va neobsahuje viry ani jin? ?kodliv? k?d - avast! Antivirus je aktivn?. http://www.avast.com -------------- next part -------------- Ports tree is already up to date. root at serviio:/ # pkg -v 1.2.5 root at serviio:/ # cd /usr/ports/ports-mgmt/portmaster/ && make install clean ===> Building/installing dialog4ports as it is required for the config dialog ===> Cleaning for dialog4ports-0.1.5_2 ===> License BSD2CLAUSE accepted by the user ===> dialog4ports-0.1.5_2 depends on file: /usr/local/sbin/pkg - found => dialog4ports-0.1.5.tar.gz doesn't seem to exist in /usr/ports/distfiles/. => Attempting to fetch http://m1cro.me/dialog4ports/dialog4ports-0.1.5.tar.gz dialog4ports-0.1.5.tar.gz 100% of 10 kB 60 kBps 00m01s ===> Fetching all distfiles required by dialog4ports-0.1.5_2 for building ===> Extracting for dialog4ports-0.1.5_2 => SHA256 Checksum OK for dialog4ports-0.1.5.tar.gz. ===> Patching for dialog4ports-0.1.5_2 ===> Configuring for dialog4ports-0.1.5_2 ===> Building for dialog4ports-0.1.5_2 Warning: Object directory not changed from original /usr/ports/ports-mgmt/dialog4ports/work/dialog4ports-0.1.5 cc -O2 -pipe -fno-strict-aliasing -std=gnu99 -fstack-protector -Wsystem-headers -Werror -Wall -Wno-format-y2k -W -Wno-unused-parameter -Wstrict-prototypes -Wmissing-prototypes -Wpointer-arith -Wreturn-type -Wcast-qual -Wwrite-strings -Wswitch -Wshadow -Wunused-parameter -Wcast-align -Wchar-subscripts -Winline -Wnested-externs -Wredundant-decls -Wold-style-definition -Wno-pointer-sign -c dialog4ports.c cc -O2 -pipe -fno-strict-aliasing -std=gnu99 -fstack-protector -Wsystem-headers -Werror -Wall -Wno-format-y2k -W -Wno-unused-parameter -Wstrict-prototypes -Wmissing-prototypes -Wpointer-arith -Wreturn-type -Wcast-qual -Wwrite-strings -Wswitch -Wshadow -Wunused-parameter -Wcast-align -Wchar-subscripts -Winline -Wnested-externs -Wredundant-decls -Wold-style-definition -Wno-pointer-sign -c mixedlist.c gzip -cn dialog4ports.1 > dialog4ports.1.gz cc -O2 -pipe -fno-strict-aliasing -std=gnu99 -fstack-protector -Wsystem-headers -Werror -Wall -Wno-format-y2k -W -Wno-unused-parameter -Wstrict-prototypes -Wmissing-prototypes -Wpointer-arith -Wreturn-type -Wcast-qual -Wwrite-strings -Wswitch -Wshadow -Wunused-parameter -Wcast-align -Wchar-subscripts -Winline -Wnested-externs -Wredundant-decls -Wold-style-definition -Wno-pointer-sign -o dialog4ports dialog4ports.o mixedlist.o -lncursesw -lm -ldialog ===> Staging for dialog4ports-0.1.5_2 ===> Generating temporary packing list install -s -o root -g wheel -m 555 dialog4ports /usr/ports/ports-mgmt/dialog4ports/work/stage/usr/local/bin/dialog4ports install -o root -g wheel -m 444 dialog4ports.1.gz /usr/ports/ports-mgmt/dialog4ports/work/stage/usr/local/man/man1 ====> Compressing man pages (compress-man) ===> Installing for dialog4ports-0.1.5_2 ===> Checking if dialog4ports already installed ===> Registering installation for dialog4ports-0.1.5_2 *** [fake-pkg] Error code 74 Stop in /usr/ports/ports-mgmt/dialog4ports. *** [install] Error code 1 Stop in /usr/ports/ports-mgmt/dialog4ports. ===> Options unchanged ===> License BSD2CLAUSE accepted by the user ===> portmaster-3.17.7 depends on file: /usr/local/sbin/pkg - found => portmaster-3.17.7.tar.gz doesn't seem to exist in /usr/ports/distfiles/. => Attempting to fetch http://distcache.FreeBSD.org/local-distfiles/bdrewery/portmaster/portmaster-3.17.7.tar.gz portmaster-3.17.7.tar.gz 100% of 43 kB 569 kBps 00m00s ===> Fetching all distfiles required by portmaster-3.17.7 for building ===> Extracting for portmaster-3.17.7 => SHA256 Checksum OK for portmaster-3.17.7.tar.gz. ===> Patching for portmaster-3.17.7 ===> Configuring for portmaster-3.17.7 ===> Building for portmaster-3.17.7 /usr/bin/sed -e 's#/usr/local#/usr/local#g' /usr/ports/ports-mgmt/portmaster/work/freebsd-portmaster-94ed670/portmaster > /usr/ports/ports-mgmt/portmaster/work/portmaster /usr/bin/sed -e 's#/usr/local#/usr/local#g' /usr/ports/ports-mgmt/portmaster/work/freebsd-portmaster-94ed670/files/portmaster.rc.sample > /usr/ports/ports-mgmt/portmaster/work/portmaster.rc.sample /usr/bin/sed -e 's#/usr/local#/usr/local#g' /usr/ports/ports-mgmt/portmaster/work/freebsd-portmaster-94ed670/files/portmaster.8 > /usr/ports/ports-mgmt/portmaster/work/portmaster.8 ===> Staging for portmaster-3.17.7 ===> Generating temporary packing list install -o root -g wheel -m 555 /usr/ports/ports-mgmt/portmaster/work/portmaster /usr/ports/ports-mgmt/portmaster/work/stage/usr/local/sbin/portmaster install -o root -g wheel -m 0644 /usr/ports/ports-mgmt/portmaster/work/portmaster.rc.sample /usr/ports/ports-mgmt/portmaster/work/stage/usr/local/etc install -o root -g wheel -m 444 /usr/ports/ports-mgmt/portmaster/work/freebsd-portmaster-94ed670/files/portmaster.8 /usr/ports/ports-mgmt/portmaster/work/stage/usr/local/man/man8 ====> Compressing man pages (compress-man) ===> Installing for portmaster-3.17.7 ===> Checking if portmaster already installed ===> Registering installation for portmaster-3.17.7 *** [fake-pkg] Error code 74 Stop in /usr/ports/ports-mgmt/portmaster. *** [install] Error code 1 Stop in /usr/ports/ports-mgmt/portmaster. root at serviio:/usr/ports/ports-mgmt/portmaster # From bdrewery at FreeBSD.org Thu Sep 18 20:15:55 2014 From: bdrewery at FreeBSD.org (Bryan Drewery) Date: Thu, 18 Sep 2014 15:15:46 -0500 Subject: Poudriere code moved to https://github.com/freebsd/poudriere Message-ID: <541B3D72.8050502@FreeBSD.org> Hi, Poudriere's code no longer is based in fossil. It has moved to https://github.com/freebsd/poudriere. Fossil will not be kept synced. Please submit patches as pull requests and issues on github going forward (rather than the old fossil site). This conversion is not compatible with the existing repository that I had on github.com/bdrewery/poudriere. The one I had was only a convenience and did not have tickets/commit-hashes converted. The new conversion does have these migrated. There are no hashes in common. Any checkouts you have will need to have customizations rebased onto the new repository. If you have no modifications to your own local repository than you can just recheckout to the new one. Given the old had a 'trunk' head then this should work: # git remote add upstream https://github.com/freebsd/poudriere.git # git remote update upstream # git checkout -b master upstream/master If you have made modifications then you will need to rebase onto the new one. Look in your 'git log' and find the last upstream commit which I will call UPSTREAM_COMMIT. Then rebase all of your work onto the new master: # git remote add upstream https://github.com/freebsd/poudriere.git # git remote update upstream # git rebase --onto upstream/master UPSTREAM_COMMIT Don't forget to git push -f to your own remote if needed. -- Regards, Bryan Drewery -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 488 bytes Desc: OpenPGP digital signature URL: From bdrewery at FreeBSD.org Thu Sep 18 20:21:24 2014 From: bdrewery at FreeBSD.org (Bryan Drewery) Date: Thu, 18 Sep 2014 15:21:16 -0500 Subject: Registering installation for dialog4ports-0.1.5_2, *** [fake-pkg] Error code 74 In-Reply-To: <54194AA9.8070604@volny.cz> References: <54194AA9.8070604@volny.cz> Message-ID: <541B3EBC.6000303@FreeBSD.org> On 9/17/2014 3:47 AM, V?clav Ho?ek wrote: > > Ports tree is already up to date. > root at serviio:/ # pkg -v > 1.2.5 This is the problem. Update to 1.3.x first. Run 'pkg bootstrap -f' or upgrade from ports-mgmt/pkg. -- Regards, Bryan Drewery -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 488 bytes Desc: OpenPGP digital signature URL: From bdrewery at FreeBSD.org Thu Sep 18 20:39:01 2014 From: bdrewery at FreeBSD.org (Bryan Drewery) Date: Thu, 18 Sep 2014 15:15:46 -0500 Subject: Poudriere code moved to https://github.com/freebsd/poudriere Message-ID: <541B3D72.8050502@FreeBSD.org> Hi, Poudriere's code no longer is based in fossil. It has moved to https://github.com/freebsd/poudriere. Fossil will not be kept synced. Please submit patches as pull requests and issues on github going forward (rather than the old fossil site). This conversion is not compatible with the existing repository that I had on github.com/bdrewery/poudriere. The one I had was only a convenience and did not have tickets/commit-hashes converted. The new conversion does have these migrated. There are no hashes in common. Any checkouts you have will need to have customizations rebased onto the new repository. If you have no modifications to your own local repository than you can just recheckout to the new one. Given the old had a 'trunk' head then this should work: # git remote add upstream https://github.com/freebsd/poudriere.git # git remote update upstream # git checkout -b master upstream/master If you have made modifications then you will need to rebase onto the new one. Look in your 'git log' and find the last upstream commit which I will call UPSTREAM_COMMIT. Then rebase all of your work onto the new master: # git remote add upstream https://github.com/freebsd/poudriere.git # git remote update upstream # git rebase --onto upstream/master UPSTREAM_COMMIT Don't forget to git push -f to your own remote if needed. -- Regards, Bryan Drewery -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 488 bytes Desc: OpenPGP digital signature URL: From rodrigc at freebsd.org Thu Sep 18 20:58:33 2014 From: rodrigc at freebsd.org (Craig Rodrigues) Date: Thu, 18 Sep 2014 13:58:30 -0700 Subject: Registering installation for dialog4ports-0.1.5_2, *** [fake-pkg] Error code 74 In-Reply-To: <541B3EBC.6000303@FreeBSD.org> References: <54194AA9.8070604@volny.cz> <541B3EBC.6000303@FreeBSD.org> Message-ID: On Thu, Sep 18, 2014 at 1:21 PM, Bryan Drewery wrote: > > Run 'pkg bootstrap -f' or upgrade from ports-mgmt/pkg. > In pkg 1.3.7, "man pkg" documents 'pkg bootstrap', but "pkg help" does not. Should it? -- Craig From bdrewery at FreeBSD.org Thu Sep 18 21:02:54 2014 From: bdrewery at FreeBSD.org (Bryan Drewery) Date: Thu, 18 Sep 2014 16:02:46 -0500 Subject: Registering installation for dialog4ports-0.1.5_2, *** [fake-pkg] Error code 74 In-Reply-To: References: <54194AA9.8070604@volny.cz> <541B3EBC.6000303@FreeBSD.org> Message-ID: <541B4876.4060709@FreeBSD.org> On 9/18/2014 3:58 PM, Craig Rodrigues wrote: > > > On Thu, Sep 18, 2014 at 1:21 PM, Bryan Drewery > wrote: > > > Run 'pkg bootstrap -f' or upgrade from ports-mgmt/pkg. > > > In pkg 1.3.7, "man pkg" documents 'pkg bootstrap', but "pkg help" does not. > Should it? > > -- > Craig IMHO pkg help needs 50% of the output removed. It is too long. -- Regards, Bryan Drewery -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 488 bytes Desc: OpenPGP digital signature URL: From johnqmetro at hotmail.com Sat Sep 20 00:47:50 2014 From: johnqmetro at hotmail.com (John Metro) Date: Fri, 19 Sep 2014 20:46:42 -0400 Subject: Problem with pkg Makefile Message-ID: That bit in the Makefile that prints out ... You are about to convert your system to pkg while you have ports/packagesinstalled with the old pkg_install tools. To switch to pkg: 1) Install ports-mgmt/pkg 2) Convert your package database by running pkg2ng *** Error code 1 .. should be removed from the makefile for pkg. It REALLY gets in the way if you are trying to upgrade and switch to pkgng for the first time from ports. From tfransosi at gmail.com Sat Sep 20 01:09:18 2014 From: tfransosi at gmail.com (Thiago Farina) Date: Fri, 19 Sep 2014 22:09:17 -0300 Subject: Problem with pkg Makefile In-Reply-To: References: Message-ID: On Fri, Sep 19, 2014 at 9:46 PM, John Metro wrote: > That bit in the Makefile that prints out ... > You are about to convert your system to pkg while you have ports/packagesinstalled with the old pkg_install tools. > To switch to pkg: 1) Install ports-mgmt/pkg 2) Convert your package database by running pkg2ng > *** Error code 1 > .. should be removed from the makefile for pkg. It REALLY gets in the way if you are trying to upgrade and switch to pkgng for the first time from ports. What should be removed? Does it say? -- Thiago Farina From jeffreybouquet at yahoo.com Sun Sep 21 00:10:31 2014 From: jeffreybouquet at yahoo.com (Jeffrey Bouquet) Date: Sat, 20 Sep 2014 17:10:22 -0700 Subject: Pkg feature req. already exits some other way? Message-ID: <1411258222.80076.YahooMailBasic@web140901.mail.bf1.yahoo.com> I am accustomed to portmaster batching a large number of ports at the end of a pipe. ... ... egrep -v 'someIGNport|someBKNport|someBIGport' | xargs -J % pkg install % Instead of a queue formed, This will update port port port And the ports not-found not-found will be left for next time. Proceed? One of the latter(not found), inexplicably halts the pkg command. This takes a magnitude of time away from other tasks I usually schedule in the meanwhile. Or someone is maybe coding it into portmaster soon... or portupgrade, or has tested portupgrade in a manner that does work in that context. From erik.eliassen at alvion.se Sun Sep 21 12:45:19 2014 From: erik.eliassen at alvion.se (Erik Eliassen) Date: Sun, 21 Sep 2014 14:44:03 +0200 Subject: install pkg problem Message-ID: <541EC813.7030607@alvion.se> Hello, I get the "[pre-everything] Error code 1" when I'm trying to install ports-mgmt/pkg. # uname -a FreeBSD 9.2-RELEASE-p12 FreeBSD 9.2-RELEASE-p12 #0: Mon Sep 15 18:46:46 UTC 2014 root at amd64-builder.daemonology.net:/usr/obj/usr/src/sys/GENERIC amd64 I've done the following # portsnap fetch update # echo "WITH_PKGNG=yes" >> /etc/make.conf # make -C /usr/ports/ports-mgmt/pkg install clean /You are about to convert your system to pkg while you have ports/packages// //installed with the old pkg_install tools.// // //To switch to pkg:// // 1) Install ports-mgmt/pkg// // 2) Convert your package database by running pkg2ng// // //*** [pre-everything] Error code 1// // //Stop in /usr/ports/ports-mgmt/pkg./ # Is it safe to just run /usr/sbin/pkg to install pkg? I don't know how to proceed with installing pkg. Thanks for any help Best Regards Erik Eliassen From vas at mpeks.tomsk.su Sun Sep 21 15:38:47 2014 From: vas at mpeks.tomsk.su (Victor Sudakov) Date: Sun, 21 Sep 2014 22:38:41 +0700 Subject: portaudit -C Message-ID: <20140921153841.GA75919@admin.sibptus.tomsk.ru> Colleagues, portaudit says, "The portaudit tool is now obsolete, please remove portaudit and use the command 'pkg audit' instead." portaudit had a very useful function: "portaudit -C": -C Print a vulnerability report for the port in the current working directory. Mostly useful for port developers. How do I do it with "pkg audit"? -- Victor Sudakov, VAS4-RIPE, VAS47-RIPN sip:sudakov at sibptus.tomsk.ru From vas at mpeks.tomsk.su Sun Sep 21 15:46:21 2014 From: vas at mpeks.tomsk.su (Victor Sudakov) Date: Sun, 21 Sep 2014 22:46:18 +0700 Subject: pkg: Unable to update repository Message-ID: <20140921154618.GA76022@admin.sibptus.tomsk.ru> Colleagues, What does this error message mean? # pkg update -f Updating sibptus repository catalogue... Fetching meta.txz: 100% 848 B 0.9k/s 00:01 Fetching digests.txz: 100% 29 kB 30.3k/s 00:01 Fetching packagesite.txz: 100% 93 kB 95.5k/s 00:01 Processing new repository entries: 100% pkg: Unable to update repository sibptus # Why is that and how do I fix it? I tried to "rm /var/db/pkg/repo-sibptus.sqlite" but it did not help. -- Victor Sudakov, VAS4-RIPE, VAS47-RIPN sip:sudakov at sibptus.tomsk.ru From jeffreybouquet at yahoo.com Sun Sep 21 20:35:46 2014 From: jeffreybouquet at yahoo.com (Jeffrey Bouquet) Date: Sun, 21 Sep 2014 13:35:38 -0700 Subject: [worrkaround ] Pkg feature req. already exists some other way? In-Reply-To: <1411258222.80076.YahooMailBasic@web140901.mail.bf1.yahoo.com> Message-ID: <1411331738.18898.YahooMailBasic@web140903.mail.bf1.yahoo.com> >>> See workaround below, after >>> the latter four of these six quoted marks... -------------------------------------------- On Sat, 9/20/14, Jeffrey Bouquet via freebsd-pkg wrote: Subject: Pkg feature req. already exits some other way? To: pkg at freebsd.org Date: Saturday, September 20, 2014, 5:10 PM I am accustomed to portmaster batching a large number of ports at the end of a pipe.??? ...? ... egrep -v ' someIGNport|someBKNport|someBIGport'? ? | xargs -J % pkg install % ? ? ? ? ? ? ? ? ? ? ? Instead of a queue formed, This will update ???port ???port ???port And the ports ???not-found ??not-found ? will be left for next time. Proceed? One of the latter(not found), inexplicably halts the pkg command.???This takes a magnitude of time away from other tasks I usually schedule in the meanwhile.? >>>Or someone is maybe coding it into portmaster soon... or >>>>portupgrade, or has tested >>>>portupgrade in a manner that does work in that context. So... .... egrep -v 'someIGNport|someBKNport' | sort | uniq | xargs -J % pkg rquery %do % | lookat (or less or) Then... pkg install port port port port port port port proceeds with way speedier than what I knew yesterday at this time. I suspect though that it is not upgrading with packages the majority of ports which need upgrading, but it woudl be nicer to have a message about each term sent to pkg, whether it found a remote package, or recieved the port name and did not find a remote package. Then one could pipe those results into a delay file or a portmaster/portupgrade file or an ignore-for-later file ... etc And that is only the first usage of that pipe, it maybe could be improved, ...as well as the pkg-rquery man page maybe, with examples, which I used from "man pkg-query" as a template. From rodrigc at freebsd.org Sun Sep 21 21:08:05 2014 From: rodrigc at freebsd.org (Craig Rodrigues) Date: Sun, 21 Sep 2014 14:08:03 -0700 Subject: portaudit -C In-Reply-To: <20140921153841.GA75919@admin.sibptus.tomsk.ru> References: <20140921153841.GA75919@admin.sibptus.tomsk.ru> Message-ID: On Sun, Sep 21, 2014 at 8:38 AM, Victor Sudakov wrote: > Colleagues, > > portaudit says, "The portaudit tool is now obsolete, please remove > portaudit and use the command 'pkg audit' instead." > > portaudit had a very useful function: "portaudit -C": > > -C Print a vulnerability report for the port in the current working > directory. Mostly useful for port developers. > > How do I do it with "pkg audit"? > Hi, You can do something like this: cd /usr/ports/net/asterisk11 make -d x check-vulnerable or cd /usr/ports/net/asterisk11 pkg audit `make -V PKGNAME` -- Craig From rodrigc at freebsd.org Sun Sep 21 21:18:10 2014 From: rodrigc at freebsd.org (Craig Rodrigues) Date: Sun, 21 Sep 2014 14:18:08 -0700 Subject: Registering installation for dialog4ports-0.1.5_2, *** [fake-pkg] Error code 74 In-Reply-To: <541B4876.4060709@FreeBSD.org> References: <54194AA9.8070604@volny.cz> <541B3EBC.6000303@FreeBSD.org> <541B4876.4060709@FreeBSD.org> Message-ID: On Thu, Sep 18, 2014 at 2:02 PM, Bryan Drewery wrote: > On 9/18/2014 3:58 PM, Craig Rodrigues wrote: > > > > > > On Thu, Sep 18, 2014 at 1:21 PM, Bryan Drewery > > wrote: > > > > > > Run 'pkg bootstrap -f' or upgrade from ports-mgmt/pkg. > > > > > > In pkg 1.3.7, "man pkg" documents 'pkg bootstrap', but "pkg help" does > not. > > Should it? > > > > -- > > Craig > > IMHO pkg help needs 50% of the output removed. It is too long. > There are two separate issues here. (1) pkg help is missing documentation about bootstrap -> "man pkg" documents bootstrap, but "pkg help" does not. This is a bug. (2) pkg help output is too long -> You can't escape the fact that pkg is growing and has a lot of features. Consistently documenting them is important. -> Maybe pkg can follow the example of git. If you type "git help", it only lists a subset of the 'most popular' commands. "git help -a" lists all the commands -- Craig From michelle at sorbs.net Sun Sep 21 23:53:53 2014 From: michelle at sorbs.net (Michelle Sullivan) Date: Mon, 22 Sep 2014 01:53:43 +0200 Subject: [HEADSUP] pkg(8) is now the only package management tool In-Reply-To: <54050D07.4010404@sorbs.net> References: <20140901195520.GB77917@ivaldir.etoilebsd.net> <54050D07.4010404@sorbs.net> Message-ID: <541F6507.70904@sorbs.net> Michelle Sullivan wrote: > Baptiste Daroussin wrote: > >> Hi all, >> >> The ports tree has been modified to only support pkg(8) as package management >> system for all supported version of FreeBSD. >> >> if you were still using pkg_install (pkg_* tools) you will have to upgrade your >> system. >> >> The simplest way is >> cd /usr/ports/ports-mgmt/pkg >> make install >> then run >> pkg2ng >> >> So despite being told 'use the quarterly, patches can be applied to it if requested' and updating ports I maintain asking specifically for the patches to be merged... Still all broken (though patches applied to HEAD)... Nice one... -- Michelle Sullivan http://www.mhix.org/ From tfransosi at gmail.com Mon Sep 22 00:56:49 2014 From: tfransosi at gmail.com (Thiago Farina) Date: Sun, 21 Sep 2014 21:56:48 -0300 Subject: Registering installation for dialog4ports-0.1.5_2, *** [fake-pkg] Error code 74 In-Reply-To: References: <54194AA9.8070604@volny.cz> <541B3EBC.6000303@FreeBSD.org> <541B4876.4060709@FreeBSD.org> Message-ID: On Sun, Sep 21, 2014 at 6:18 PM, Craig Rodrigues wrote: > On Thu, Sep 18, 2014 at 2:02 PM, Bryan Drewery wrote: > >> On 9/18/2014 3:58 PM, Craig Rodrigues wrote: >> > > > There are two separate issues here. > (1) pkg help is missing documentation about bootstrap > -> "man pkg" documents bootstrap, but "pkg help" does not. This > is a bug. > > (2) pkg help output is too long > -> You can't escape the fact that pkg is growing and has a lot of > features. > Consistently documenting them is important. > -> Maybe pkg can follow the example of git. > If you type "git help", it only lists a subset of the 'most > popular' commands. > "git help -a" lists all the commands If someone is interested in helping out with this. The code can be found here -> https://github.com/freebsd/pkg/blob/master/src/main.c#L144 Regards, -- Thiago Farina From vas at mpeks.tomsk.su Mon Sep 22 01:26:20 2014 From: vas at mpeks.tomsk.su (Victor Sudakov) Date: Mon, 22 Sep 2014 08:26:12 +0700 Subject: portaudit -C In-Reply-To: References: <20140921153841.GA75919@admin.sibptus.tomsk.ru> Message-ID: <20140922012612.GA81996@admin.sibptus.tomsk.ru> Craig Rodrigues wrote: > > > > portaudit says, "The portaudit tool is now obsolete, please remove > > portaudit and use the command 'pkg audit' instead." > > > > portaudit had a very useful function: "portaudit -C": > > > > -C Print a vulnerability report for the port in the current working > > directory. Mostly useful for port developers. > > > > How do I do it with "pkg audit"? > > > > Hi, > > You can do something like this: > > cd /usr/ports/net/asterisk11 > make -d x check-vulnerable # make -d x check-vulnerable make: illegal argument to d option -- x We need a different "x". BTW the "check-vulnerable" target is not documented in ports(7), I had to hunt for in in /usr/ports/Mk/* Maybe we need some "pretty-print-vulnerable" target once portaudit is now obsolete? > > or > > cd /usr/ports/net/asterisk11 > pkg audit `make -V PKGNAME` This requires that net/asterisk11 be already installed, doesn't it? -- Victor Sudakov, VAS4-RIPE, VAS47-RIPN sip:sudakov at sibptus.tomsk.ru From rodrigc at freebsd.org Mon Sep 22 08:13:17 2014 From: rodrigc at freebsd.org (Craig Rodrigues) Date: Mon, 22 Sep 2014 01:13:14 -0700 Subject: portaudit -C In-Reply-To: <20140922012612.GA81996@admin.sibptus.tomsk.ru> References: <20140921153841.GA75919@admin.sibptus.tomsk.ru> <20140922012612.GA81996@admin.sibptus.tomsk.ru> Message-ID: On Sun, Sep 21, 2014 at 6:26 PM, Victor Sudakov wrote: > > # make -d x check-vulnerable > make: illegal argument to d option -- x > > I'm not sure what version of FreeBSD you are using, but in FreeBSD 10 and higher, "make -d x check-vulnerable" works. > We need a different "x". > > BTW the "check-vulnerable" target is not documented in ports(7), I had > to hunt for in in /usr/ports/Mk/* > There are *lots* of things in /usr/ports/Mk/* that are not documented in ports(7). :) > Maybe we need some "pretty-print-vulnerable" target once portaudit is > now obsolete? > That's a good question to post on the freebsd-ports at freebsd.org mailing list to get feedback on. > > > > > or > > > > cd /usr/ports/net/asterisk11 > > pkg audit `make -V PKGNAME` > > This requires that net/asterisk11 be already installed, doesn't it? > No. -- Craig From vas at mpeks.tomsk.su Mon Sep 22 08:56:34 2014 From: vas at mpeks.tomsk.su (Victor Sudakov) Date: Mon, 22 Sep 2014 15:56:26 +0700 Subject: portaudit -C In-Reply-To: References: <20140921153841.GA75919@admin.sibptus.tomsk.ru> <20140922012612.GA81996@admin.sibptus.tomsk.ru> Message-ID: <20140922085626.GA93280@admin.sibptus.tomsk.ru> Craig Rodrigues wrote: > > > > > # make -d x check-vulnerable > > make: illegal argument to d option -- x > > > > > I'm not sure what version of FreeBSD you are using, but in FreeBSD 10 and > higher, > "make -d x check-vulnerable" works. 9.3-RELEASE in this case. > > We need a different "x". > > > > BTW the "check-vulnerable" target is not documented in ports(7), I had > > to hunt for in in /usr/ports/Mk/* > > > > There are *lots* of things in /usr/ports/Mk/* that are not documented in > ports(7). :) Most of them are of the "no user serviceable parts inside" kind. [dd] > > > > > > cd /usr/ports/net/asterisk11 > > > pkg audit `make -V PKGNAME` > > > > This requires that net/asterisk11 be already installed, doesn't it? > > > > No. $ $ pwd /var/poudriere/ports/default/www/squid $ pkg audit `make -V PKGNAME` 0 problem(s) in the installed packages found. $ I see the phrase "installed packages" there. -- Victor Sudakov, VAS4-RIPE, VAS47-RIPN sip:sudakov at sibptus.tomsk.ru From rodrigc at freebsd.org Mon Sep 22 09:25:14 2014 From: rodrigc at freebsd.org (Craig Rodrigues) Date: Mon, 22 Sep 2014 02:25:11 -0700 Subject: portaudit -C In-Reply-To: <20140922085626.GA93280@admin.sibptus.tomsk.ru> References: <20140921153841.GA75919@admin.sibptus.tomsk.ru> <20140922012612.GA81996@admin.sibptus.tomsk.ru> <20140922085626.GA93280@admin.sibptus.tomsk.ru> Message-ID: On Mon, Sep 22, 2014 at 1:56 AM, Victor Sudakov wrote: > > > > > > > > > cd /usr/ports/net/asterisk11 > > > > pkg audit `make -V PKGNAME` > > > > > > This requires that net/asterisk11 be already installed, doesn't it? > > > > > > > No. > > $ > $ pwd > /var/poudriere/ports/default/www/squid > $ pkg audit `make -V PKGNAME` > 0 problem(s) in the installed packages found. > $ > > I see the phrase "installed packages" there. > The message from 'pkg audit' is wrong. It parses /var/db/pkg/vuln.xml to list the vulnerabilities. If you give it an exact package name, it searches vuln.xml for that. The package does not need to be installed. For example: pkg audit phpmyfaq-1.5.2 -- Craig From vas at mpeks.tomsk.su Mon Sep 22 10:15:45 2014 From: vas at mpeks.tomsk.su (Victor Sudakov) Date: Mon, 22 Sep 2014 17:15:38 +0700 Subject: portaudit -C In-Reply-To: References: <20140921153841.GA75919@admin.sibptus.tomsk.ru> <20140922012612.GA81996@admin.sibptus.tomsk.ru> <20140922085626.GA93280@admin.sibptus.tomsk.ru> Message-ID: <20140922101538.GA94283@admin.sibptus.tomsk.ru> Craig Rodrigues wrote: > > > > > > > > > > cd /usr/ports/net/asterisk11 > > > > > pkg audit `make -V PKGNAME` > > > > > > > > This requires that net/asterisk11 be already installed, doesn't it? > > > > > > > > > > No. > > > > $ > > $ pwd > > /var/poudriere/ports/default/www/squid > > $ pkg audit `make -V PKGNAME` > > 0 problem(s) in the installed packages found. > > $ > > > > I see the phrase "installed packages" there. > > > > The message from 'pkg audit' is wrong. Good to know, thank you. > It parses /var/db/pkg/vuln.xml to > list the vulnerabilities. > If you give it an exact package name, it searches vuln.xml for that. The > package does not need to be installed. > > For example: > > pkg audit phpmyfaq-1.5.2 > -- Victor Sudakov, VAS4-RIPE, VAS47-RIPN sip:sudakov at sibptus.tomsk.ru From mikej at mikej.com Mon Sep 22 12:51:27 2014 From: mikej at mikej.com (Michael Jung) Date: Mon, 22 Sep 2014 08:50:55 -0400 Subject: [CFT] SSP Package Repository available In-Reply-To: <53F4CE0E.8040106@FreeBSD.org> References: <523D79CD.2090302@FreeBSD.org> <53F4CE0E.8040106@FreeBSD.org> Message-ID: On 2014-08-20 12:34, Bryan Drewery wrote: > On 9/21/2013 5:49 AM, Bryan Drewery wrote: >> Ports now support enabling Stack Protector [1] support on FreeBSD 10 >> i386 and amd64, and older releases on amd64 only currently. >> >> Support may be added for earlier i386 releases once all ports properly >> respect LDFLAGS. >> >> To enable, just add WITH_SSP=yes to your make.conf and rebuild all >> ports. >> >> The default SSP_CLFAGS is -fstack-protector, but -fstack-protector-all >> may optionally be set instead. >> >> Please help test this on your system. We would like to eventually >> enable >> this by default, but need to identify any major ports that have >> run-time >> issues due to it. >> >> [1] https://en.wikipedia.org/wiki/Buffer_overflow_protection >> > > We have not had any feedback on this yet and want to get it enabled by > default for ports and packages. > > We now have a repository that you can use rather than the default to > help test. We need your help to identify any issues before switching > the > default. > > This repository is available for: > > head > 10.0 > 9.1,9.2,9.3 > > It is not available for 8.4. If someone is willing to test on 8.4 I > will > build a repository for it. > > Place this in /usr/local/etc/pkgs/repos/FreeBSD_ssp.conf: > > FreeBSD: { enabled: no } > FreeBSD_ssp: { > url: "pkg+http://pkg.FreeBSD.org/${ABI}/ssp", > mirror_type: "srv", > signature_type: "fingerprints", > fingerprints: "/usr/share/keys/pkg", > enabled: yes > } > > Once that is done you should force reinstall packages from this > repository: > > pkg update > pkg upgrade -f > > Thanks for your help! > Bryan Drewery > On behalf of portmgr. I have been building (poudriere) and running some 1100+ ports WITH_SSP_PORT=yes under 10-STABLE and CURRENT without issue. This is using both our local repository and the ssp repository listed above. --mikej From bdrewery at FreeBSD.org Mon Sep 22 16:11:41 2014 From: bdrewery at FreeBSD.org (Bryan Drewery) Date: Mon, 22 Sep 2014 11:11:31 -0500 Subject: [HEADSUP] pkg(8) is now the only package management tool In-Reply-To: <541F6507.70904@sorbs.net> References: <20140901195520.GB77917@ivaldir.etoilebsd.net> <54050D07.4010404@sorbs.net> <541F6507.70904@sorbs.net> Message-ID: <54204A33.3070700@FreeBSD.org> On 9/21/2014 6:53 PM, Michelle Sullivan wrote: > Michelle Sullivan wrote: >> Baptiste Daroussin wrote: >> >>> Hi all, >>> >>> The ports tree has been modified to only support pkg(8) as package management >>> system for all supported version of FreeBSD. >>> >>> if you were still using pkg_install (pkg_* tools) you will have to upgrade your >>> system. >>> >>> The simplest way is >>> cd /usr/ports/ports-mgmt/pkg >>> make install >>> then run >>> pkg2ng >>> >>> > > So despite being told 'use the quarterly, patches can be applied to it > if requested' and updating ports I maintain asking specifically for the > patches to be merged... > > Still all broken (though patches applied to HEAD)... Nice one... > Which ports and PR? -- Regards, Bryan Drewery -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 488 bytes Desc: OpenPGP digital signature URL: From listjm at club-internet.fr Tue Sep 23 08:24:10 2014 From: listjm at club-internet.fr (Juan =?iso-8859-1?b?UmFt824=?= Molina Menor) Date: Tue, 23 Sep 2014 10:14:50 +0200 Subject: [CFT] SSP Package Repository available Message-ID: <54212BFA.2050702@club-internet.fr> On 9/21/2013 5:49 AM, Bryan Drewery wrote: > Ports now support enabling Stack Protector [1] support on FreeBSD 10 > i386 and amd64, and older releases on amd64 only currently. Hi! I have updated my 10.1-BETA2 (i386) system to use the SSP package repository, together with the FreeBSD_new_xorg repository. It?s a modest set-up, with 390 packages. Force-upgrading was seamless and after a reboot all seems to work as well as before. Two questions though: Can I keep this repository till we are informed of the switch to SSP by default? (that is to say: would it be timely updated as it is the case now?) Can I test the upcoming linux_base-c6 packages with this SSP repository? Best regards, Juan From gibblertron at gmail.com Tue Sep 23 16:58:48 2014 From: gibblertron at gmail.com (Patrick Gibson) Date: Tue, 23 Sep 2014 09:58:45 -0700 Subject: Poudriere hooks? Message-ID: Is there any way to run a hook within a jail when doing a bulk build? We use the enterprise version of Phusion Passenger (www/rubygem-passenger), and I have a script to download the source from Phusion into the distfiles directory, and patch the port to use the modified version. But as we're moving to a Poudriere-based build system, I haven't figured out a good way to do this within the jailed environment during a bulk build. From matthew at FreeBSD.org Tue Sep 23 19:28:28 2014 From: matthew at FreeBSD.org (Matthew Seaman) Date: Tue, 23 Sep 2014 20:28:00 +0100 Subject: Poudriere hooks? In-Reply-To: References: Message-ID: <5421C9C0.60003@FreeBSD.org> On 23/09/2014 17:58, Patrick Gibson wrote: > Is there any way to run a hook within a jail when doing a bulk build? We > use the enterprise version of Phusion Passenger (www/rubygem-passenger), > and I have a script to download the source from Phusion into the distfiles > directory, and patch the port to use the modified version. But as we're > moving to a Poudriere-based build system, I haven't figured out a good way > to do this within the jailed environment during a bulk build. If you use one of the SVN based methods of checking out the ports tree, then you can apply your patches to the checked-out tree that poudriere generates, and -- at the cost of occasionally having to do some merging -- your modified port will be preserved across updates. If you run 'poudriere ports -l' it will tell you where the ports tree that poudriere uses is. Cheers, Matthew -- Dr Matthew J Seaman MA, D.Phil. PGP: http://www.infracaninophile.co.uk/pgpkey -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 971 bytes Desc: OpenPGP digital signature URL: From papowell at astart.com Wed Sep 24 03:20:33 2014 From: papowell at astart.com (Patrick Powell) Date: Tue, 23 Sep 2014 20:20:23 -0700 Subject: Suggestion on how to add search order to multiple reponsitories Message-ID: <54223877.1000302@astart.com> Currently there does not appear to be simple way to specify the order that repositories searched for packages, or at least there is none documented as such that I can find. Suppose we add a 'repo_order' field to the repository specification, i.e. - FreeBSDMyStuff: { URL: http://myserver/${ABI}/latest ENABLED: yes MIRROR_TYPE: SRV repo_order: 1 } FreeBSD: { URL: http://pkg.freebsd.org/${ABI}/latest ENABLED: yes MIRROR_TYPE: SRV repo_order: 2 } When doing a search for packages, the found/discovered packages would be sorted by the value of the order field. By default, if there is no value for the order field then this entry would come last. If there are two entries with no or the same order value then they would be sorted on the repo name, as is currently done. Also, when displaying repositories, i.e. - pkg -v -v, you would sort the display entries by the repo_order field as well. Just to jumpstart this effort I am attaching a set of patches for pkg-1.4.0.pre-alpha15. This implements the basic functionality to add repo_order to the repository definitions and adds support to display the repo_order when doing 'pkg search'. I would happy to add this ordering functionality to other parts of the pkg support, but I need a bit of help with the current code. From my understanding/reading/crystal ball gazing/looking at the code entrails, it appears that when doing a 'pkg install X' the X is first used to determine the set of packages in question. Once these have been determined there is a rather intimidating (recursive as well?) procedure which is used to determine any dependencies. Somewhere in this process there is a place where the listed/named repositories are searched for candidates for dependencies. It appears at a casual reading that the repositories are searched in the reverse order they were put in the reponsitory list and candidates found during the search are then added to the dependency list as they are found. This has the effect of having entries found in earlier listed repositories overwrite those found in later listed respositories (I thought this was VERY clever!). Rather than modify this code, I would suggest adding a 'repository ordering' step or routine. Just after the repositories are found/listed, the repository list should be sorted (stably) using the name and the repo_order field. That is, all of the repositories with a lower repo_order value would be moved ahead of those with lower repo_order values, while preserving the order for those with the same repo_order value. If this is done and the algorithm for searching the repositories is as I have outlined it, then you should get the desired repository priority effect. And perhaps no other code changes would be required. I hope. Perhaps. Maybe. -------------- next part -------------- *** pkg-1.4.0.pre-alpha15/docs/pkg_printf.3 Mon Sep 15 13:18:26 2014 --- pkg-1.4.0.pre-alpha15.orig/docs/pkg_printf.3 Sun Jun 8 12:59:52 2014 *************** *** 644,655 **** Valid only during operations when one version of a package is being replaced by another. .Vt struct pkg * - .It Cm %Z - Repository Order (repo_order) [integer] value. - Value is set in the repository information and is limited to 0-100 (default 100). - When selecting packages for installation, repositories with lower order numbers - are searched before packages with higher order numbers. - .Vt struct pkg * .It Cm %a Autoremove flag [boolean] .Vt struct pkg * --- 644,649 ---- *** pkg-1.4.0.pre-alpha15/docs/pkg.conf.5 Mon Sep 15 13:33:02 2014 --- pkg-1.4.0.pre-alpha15.orig/docs/pkg.conf.5 Sun Sep 14 15:11:40 2014 *************** *** 271,279 **** .Va hw.ncpu is used. Default: 0. - .It Cm REPO_ORDER: integer [1-100, default 100] - Repositories with lower repo_order values are examined before reponsitories with - larger repo_order values. .El .Sh REPOSITORY CONFIGURATION To use a repository you will need at least one repository --- 271,276 ---- *************** *** 422,438 **** .Sy REPOS_DIR search path, with individual repository configuration files in the same directory processed in alphabetical order. ! This search order is modified by the ! .Sy repo_order ! value; repositories with smaller ! .Sy repo_order ! values are searched befor repositories with larger ! .Sy repo_order ! values. ! Packages found earlier in the search order take precedence, ! meaning that ! if the same package is available from several repositories ! the first one found in the search order will be used. This behaviour may be overridden per-package by adding a .Sy repository annotation to the installed package. --- 419,427 ---- .Sy REPOS_DIR search path, with individual repository configuration files in the same directory processed in alphabetical order. ! Earlier files take precedence, meaning that packages will be downloaded ! from them preferentially where the same package is available from several ! repositories. This behaviour may be overridden per-package by adding a .Sy repository annotation to the installed package. *************** *** 455,478 **** fingerprints: "/usr/share/keys/pkg", mirror_type: "srv" } - - # can be in the same or separate file - FreeBSD_my_mods: { - url: "pkg+http://my.server.private/${ABI}/latest", - enabled: true, - signature_type: "fingerprints", - fingerprints: "/usr/share/keys/pkg", - mirror_type: "srv" - repo_order: 20 - } .Ed - .Pp - The - .Sy FreeBSD_my_mods - repository (repo_order 20) will be searched before the - .Sy FreeBSD - repository (repo_order default 100). - .Pp Example for pkg.conf .Bd -literal -offset indent pkg_dbdir: "/var/db/pkg" --- 444,450 ---- *** pkg-1.4.0.pre-alpha15/libpkg/pkg.c Mon Sep 15 13:58:21 2014 --- pkg-1.4.0.pre-alpha15.orig/libpkg/pkg.c Fri Sep 12 16:55:14 2014 *************** *** 41,111 **** static ucl_object_t *manifest_schema = NULL; ! static struct pkg_key_x pkg_keys_x[] = { ! { PKG_ORIGIN, "origin", UCL_STRING }, ! { PKG_NAME, "name", UCL_STRING }, ! { PKG_VERSION, "version", UCL_STRING }, ! { PKG_COMMENT, "comment", UCL_STRING }, ! { PKG_DESC, "desc", UCL_STRING }, ! { PKG_MTREE, "mtree", UCL_STRING }, ! { PKG_MESSAGE, "message", UCL_STRING }, ! { PKG_ARCH, "arch", UCL_STRING }, ! { PKG_MAINTAINER, "maintainer", UCL_STRING }, ! { PKG_WWW, "www", UCL_STRING }, ! { PKG_PREFIX, "prefix", UCL_STRING }, ! { PKG_REPOPATH, "repopath", UCL_STRING }, ! { PKG_CKSUM, "sum", UCL_STRING }, ! { PKG_OLD_VERSION, "oldversion", UCL_STRING }, ! { PKG_REPONAME, "reponame", UCL_STRING }, ! { PKG_REPOURL, "repourl", UCL_STRING }, ! { PKG_DIGEST, "digest", UCL_STRING }, ! { PKG_REASON, "reason", UCL_STRING }, ! { PKG_FLATSIZE, "flatsize", UCL_INT }, ! { PKG_OLD_FLATSIZE, "oldflatsize", UCL_INT }, ! { PKG_PKGSIZE, "pkgsize", UCL_INT }, ! { PKG_LICENSE_LOGIC, "licenselogic", UCL_INT }, ! { PKG_AUTOMATIC, "automatic", UCL_BOOLEAN }, ! { PKG_LOCKED, "locked", UCL_BOOLEAN }, ! { PKG_ROWID, "rowid", UCL_INT }, ! { PKG_TIME, "time", UCL_INT }, ! { PKG_ANNOTATIONS, "annotations", UCL_OBJECT }, ! { PKG_LICENSES, "licenses", UCL_ARRAY }, ! { PKG_CATEGORIES, "categories", UCL_ARRAY }, ! { PKG_UNIQUEID, "uniqueid", UCL_STRING }, ! { PKG_OLD_DIGEST, "olddigest", UCL_STRING }, ! { PKG_REPO_ORDER, "repo_order", UCL_INT } }; - struct pkg_key_x *get_pkg_key_by_enum_tag( int enum_tag ) - { - struct pkg_key_x *found = 0; - int i; - for( i = 0; i < sizeof(pkg_keys_x)/sizeof(pkg_keys_x[0]); ++i ){ - struct pkg_key_x *test = &pkg_keys_x[i]; - if( test->enum_tag == enum_tag ){ - found = test; - break; - } - } - return( found ); - } - - - struct pkg_key_x *get_pkg_key_by_name( const char * name ) - { - struct pkg_key_x *found = 0; - int i; - for( i = 0; i < sizeof(pkg_keys_x)/sizeof(pkg_keys_x[0]); ++i ){ - struct pkg_key_x *test = &pkg_keys_x[i]; - if( ! strcmp( test->name , name ) ){ - found = test; - break; - } - } - return( found ); - } - - int pkg_new(struct pkg **pkg, pkg_t type) { --- 41,80 ---- static ucl_object_t *manifest_schema = NULL; ! struct pkg_key pkg_keys[PKG_NUM_FIELDS] = { ! [PKG_ORIGIN] = { "origin", UCL_STRING }, ! [PKG_NAME] = { "name", UCL_STRING }, ! [PKG_VERSION] = { "version", UCL_STRING }, ! [PKG_COMMENT] = { "comment", UCL_STRING }, ! [PKG_DESC] = { "desc", UCL_STRING }, ! [PKG_MTREE] = { "mtree", UCL_STRING }, ! [PKG_MESSAGE] = { "message", UCL_STRING }, ! [PKG_ARCH] = { "arch", UCL_STRING }, ! [PKG_MAINTAINER] = { "maintainer", UCL_STRING }, ! [PKG_WWW] = { "www", UCL_STRING }, ! [PKG_PREFIX] = { "prefix", UCL_STRING }, ! [PKG_REPOPATH] = { "repopath", UCL_STRING }, ! [PKG_CKSUM] = { "sum", UCL_STRING }, ! [PKG_OLD_VERSION] = { "oldversion", UCL_STRING }, ! [PKG_REPONAME] = { "reponame", UCL_STRING }, ! [PKG_REPOURL] = { "repourl", UCL_STRING }, ! [PKG_DIGEST] = { "digest", UCL_STRING }, ! [PKG_REASON] = { "reason", UCL_STRING }, ! [PKG_FLATSIZE] = { "flatsize", UCL_INT }, ! [PKG_OLD_FLATSIZE] = { "oldflatsize", UCL_INT }, ! [PKG_PKGSIZE] = { "pkgsize", UCL_INT }, ! [PKG_LICENSE_LOGIC] = { "licenselogic", UCL_INT }, ! [PKG_AUTOMATIC] = { "automatic", UCL_BOOLEAN }, ! [PKG_LOCKED] = { "locked", UCL_BOOLEAN }, ! [PKG_ROWID] = { "rowid", UCL_INT }, ! [PKG_TIME] = { "time", UCL_INT }, ! [PKG_ANNOTATIONS] = { "annotations", UCL_OBJECT }, ! [PKG_LICENSES] = { "licenses", UCL_ARRAY }, ! [PKG_CATEGORIES] = { "categories", UCL_ARRAY }, ! [PKG_UNIQUEID] = { "uniqueid", UCL_STRING }, ! [PKG_OLD_DIGEST] = { "olddigest", UCL_STRING }, }; int pkg_new(struct pkg **pkg, pkg_t type) { *************** *** 289,302 **** while ((attr = va_arg(ap, int)) > 0) { ! struct pkg_key_x *key = get_pkg_key_by_enum_tag( attr ); ! if (!key) { ! pkg_emit_error("Bad enum value %d to pkg_vset", attr ); return (EPKG_FATAL); } ! ! obj = ucl_object_find_key(pkg->fields, key->name); ! switch (key->type) { case UCL_STRING: if (obj == NULL) { *va_arg(ap, const char **) = NULL; --- 258,270 ---- while ((attr = va_arg(ap, int)) > 0) { ! if (attr >= PKG_NUM_FIELDS || attr <= 0) { ! pkg_emit_error("Bad argument on pkg_get"); return (EPKG_FATAL); } ! ! obj = ucl_object_find_key(pkg->fields, pkg_keys[attr].name); ! switch (pkg_keys[attr].type) { case UCL_STRING: if (obj == NULL) { *va_arg(ap, const char **) = NULL; *************** *** 356,368 **** ucl_object_t *o; while ((attr = va_arg(ap, int)) > 0) { ! struct pkg_key_x *key = get_pkg_key_by_enum_tag( attr ); ! if (!key) { ! pkg_emit_error("Bad enum value %d to pkg_vset", attr ); return (EPKG_FATAL); } ! ! switch (key->type) { case UCL_STRING: str = va_arg(ap, const char *); data = str; --- 324,335 ---- ucl_object_t *o; while ((attr = va_arg(ap, int)) > 0) { ! if (attr >= PKG_NUM_FIELDS || attr <= 0) { ! pkg_emit_error("Bad argument on pkg_get"); return (EPKG_FATAL); } ! ! switch (pkg_keys[attr].type) { case UCL_STRING: str = va_arg(ap, const char *); data = str; *************** *** 373,401 **** } if (data == NULL) ! ucl_object_delete_key(pkg->fields, key->name); else ucl_object_replace_key(pkg->fields, ucl_object_fromstring_common(data, strlen(data), 0), ! key->name, strlen(key->name), false); free(buf); break; case UCL_BOOLEAN: ucl_object_replace_key(pkg->fields, ucl_object_frombool((bool)va_arg(ap, int)), ! key->name, strlen(key->name), false); break; case UCL_INT: ucl_object_replace_key(pkg->fields, ucl_object_fromint(va_arg(ap, int64_t)), ! key->name, strlen(key->name), false); break; case UCL_OBJECT: case UCL_ARRAY: o = va_arg(ap, ucl_object_t *); ucl_object_replace_key(pkg->fields, o, ! key->name, strlen(key->name), false); break; default: (void) va_arg(ap, void *); --- 340,368 ---- } if (data == NULL) ! ucl_object_delete_key(pkg->fields, pkg_keys[attr].name); else ucl_object_replace_key(pkg->fields, ucl_object_fromstring_common(data, strlen(data), 0), ! pkg_keys[attr].name, strlen(pkg_keys[attr].name), false); free(buf); break; case UCL_BOOLEAN: ucl_object_replace_key(pkg->fields, ucl_object_frombool((bool)va_arg(ap, int)), ! pkg_keys[attr].name, strlen(pkg_keys[attr].name), false); break; case UCL_INT: ucl_object_replace_key(pkg->fields, ucl_object_fromint(va_arg(ap, int64_t)), ! pkg_keys[attr].name, strlen(pkg_keys[attr].name), false); break; case UCL_OBJECT: case UCL_ARRAY: o = va_arg(ap, ucl_object_t *); ucl_object_replace_key(pkg->fields, o, ! pkg_keys[attr].name, strlen(pkg_keys[attr].name), false); break; default: (void) va_arg(ap, void *); *** pkg-1.4.0.pre-alpha15/libpkg/pkg.h.in Mon Sep 15 13:17:10 2014 --- pkg-1.4.0.pre-alpha15.orig/libpkg/pkg.h.in Sun Sep 14 15:11:40 2014 *************** *** 266,272 **** PKG_CATEGORIES, PKG_UNIQUEID, PKG_OLD_DIGEST, ! PKG_REPO_ORDER, } pkg_attr; typedef enum { --- 266,272 ---- PKG_CATEGORIES, PKG_UNIQUEID, PKG_OLD_DIGEST, ! PKG_NUM_FIELDS, /* end of fields */ } pkg_attr; typedef enum { *** pkg-1.4.0.pre-alpha15/libpkg/pkg_checksum.c Mon Sep 15 13:13:54 2014 --- pkg-1.4.0.pre-alpha15.orig/libpkg/pkg_checksum.c Sat Sep 6 19:01:46 2014 *************** *** 127,132 **** --- 127,133 ---- pkg_checksum_generate(struct pkg *pkg, char *dest, size_t destlen, pkg_checksum_type_t type) { + const char *key; unsigned char *bdigest; size_t blen; struct pkg_checksum_entry *entries = NULL; *************** *** 150,163 **** return (EPKG_FATAL); for (i = 0; recopies[i] != -1; i++) { ! struct pkg_key_x *key = get_pkg_key_by_enum_tag(recopies[i]); ! if( !key ){ ! pkg_emit_error("pkg_checksum_generate: no key for enum value '%d'", recopies[i]); ! } else { ! if ((o = ucl_object_find_key(pkg->fields, key->name))){ ! pkg_checksum_add_entry(key->name, ucl_object_tostring(o), &entries); ! } ! } } while (pkg_options(pkg, &option) == EPKG_OK) { --- 151,159 ---- return (EPKG_FATAL); for (i = 0; recopies[i] != -1; i++) { ! key = pkg_keys[recopies[i]].name; ! if ((o = ucl_object_find_key(pkg->fields, key))) ! pkg_checksum_add_entry(key, ucl_object_tostring(o), &entries); } while (pkg_options(pkg, &option) == EPKG_OK) { *** pkg-1.4.0.pre-alpha15/libpkg/pkg_config.c Mon Sep 15 13:09:57 2014 --- pkg-1.4.0.pre-alpha15.orig/libpkg/pkg_config.c Sun Sep 14 15:11:40 2014 *************** *** 424,430 **** const char *signature_type = NULL, *fingerprints = NULL; const char *key; const char *type = NULL; - int repo_order = 100; int use_ipvx; pkg_debug(1, "PkgConfig: parsing repository object %s", rname); --- 424,429 ---- *************** *** 513,529 **** use_ipvx = ucl_object_toint(cur); if (use_ipvx != 4 && use_ipvx != 6) use_ipvx = 0; - } else if (strcasecmp(key, "repo_order") == 0) { - pkg_debug(1, "repo %s has 'repo_order'", rname); - if (cur->type != UCL_INT) { - pkg_emit_error("Expecting an integer for the " - "'%s' key of the '%s' repo", - key, rname); - return; - } - repo_order = ucl_object_toint(cur); - if( repo_order >= 100 ){ repo_order = 100; } - if( repo_order < 0 ){ repo_order = 0; } } } --- 512,517 ---- *************** *** 558,564 **** } r->enable = enable; - r->repo_order = repo_order; if (mirror_type != NULL) { if (strcasecmp(mirror_type, "srv") == 0) --- 546,551 ---- *** pkg-1.4.0.pre-alpha15/libpkg/pkg_manifest.c Mon Sep 15 13:03:46 2014 --- pkg-1.4.0.pre-alpha15.orig/libpkg/pkg_manifest.c Sun Sep 7 23:23:24 2014 *************** *** 99,105 **** { "option_descriptions", PKG_OPTION_DESCRIPTIONS, UCL_OBJECT, pkg_obj}, { "origin", PKG_ORIGIN, UCL_STRING, pkg_string}, { "path", PKG_REPOPATH, UCL_STRING, pkg_string}, - { "repo_order", PKG_REPO_ORDER, UCL_INT, pkg_int}, { "repopath", PKG_REPOPATH, UCL_STRING, pkg_string}, { "pkgsize", PKG_PKGSIZE, UCL_INT, pkg_int}, { "prefix", PKG_PREFIX, UCL_STRING, pkg_string}, --- 99,104 ---- *************** *** 883,888 **** --- 882,888 ---- ucl_object_t *map, *seq, *submap; ucl_object_t *top = ucl_object_typed_new(UCL_OBJECT); const ucl_object_t *o; + const char *key; int recopies[] = { PKG_NAME, PKG_ORIGIN, *************** *** 904,917 **** pkg_debug(4, "Emitting basic metadata"); for (i = 0; recopies[i] != -1; i++) { ! struct pkg_key_x *key = get_pkg_key_by_enum_tag( recopies[i] ); ! if( !key ){ ! pkg_emit_error("pkg_emit_obj: bad enum value %d", recopies[i] ); ! continue; ! } ! if ((o = ucl_object_find_key(pkg->fields, key->name))) ucl_object_insert_key(top, ucl_object_ref(o), ! key->name, strlen(key->name), false); } if (comment) ucl_object_insert_key(top, ucl_object_fromstring_common(comment, 0, --- 904,913 ---- pkg_debug(4, "Emitting basic metadata"); for (i = 0; recopies[i] != -1; i++) { ! key = pkg_keys[recopies[i]].name; ! if ((o = ucl_object_find_key(pkg->fields, key))) ucl_object_insert_key(top, ucl_object_ref(o), ! key, strlen(key), false); } if (comment) ucl_object_insert_key(top, ucl_object_fromstring_common(comment, 0, *************** *** 1054,1060 **** map = ucl_object_typed_new(UCL_OBJECT); /* Add annotations except for internal ones. */ while ((o = ucl_iterate_object(annotations, &it, true))) { - const char *key; if ((key = ucl_object_keyl(o, &key_len)) == NULL) continue; /* Internal annotations. */ --- 1050,1055 ---- *** pkg-1.4.0.pre-alpha15/libpkg/pkg_printf.c Mon Sep 15 12:48:15 2014 --- pkg-1.4.0.pre-alpha15.orig/libpkg/pkg_printf.c Fri Sep 12 16:32:43 2014 *************** *** 111,117 **** * W * X * Y ! * Z pkg repo_order * * a pkg autoremove flag * --- 111,117 ---- * W * X * Y ! * Z * * a pkg autoremove flag * *************** *** 786,800 **** PP_ALL, &format_home_url, }, - [PP_PKG_REPO_ORDER] = - { - 'Z', - '\0', - false, - true, - PP_ALL, - &format_repo_order, - }, [PP_LITERAL_PERCENT] = { '%', --- 786,791 ---- *************** *** 1884,1903 **** return (string_val(sbuf, url, p)); } - - /* - * %Z -- Repo Order. integer - */ - struct sbuf * - format_repo_order(struct sbuf *sbuf, const void *data, struct percent_esc *p) - { - const struct pkg *pkg = data; - int64_t repo_order; - - pkg_get(pkg, PKG_REPO_ORDER, &repo_order); - return (int_val(sbuf, repo_order, p)); - } - /* * %% -- Output a literal '%' character */ --- 1875,1880 ---- *** pkg-1.4.0.pre-alpha15/libpkg/pkgdb.c Mon Sep 15 12:42:09 2014 --- pkg-1.4.0.pre-alpha15.orig/libpkg/pkgdb.c Wed Sep 10 23:17:52 2014 *************** *** 1069,1075 **** const char sql[] = "BEGIN IMMEDIATE TRANSACTION"; psql = sql; ! pkg_debug(4, "Pkgdb: pkgdb_transaction_begin running '%s'", sql); ret = sqlite3_prepare_v2(sqlite, sql, strlen(sql) + 1, &stmt, NULL); } else { --- 1069,1075 ---- const char sql[] = "BEGIN IMMEDIATE TRANSACTION"; psql = sql; ! pkg_debug(4, "Pkgdb: running '%s'", sql); ret = sqlite3_prepare_v2(sqlite, sql, strlen(sql) + 1, &stmt, NULL); } else { *************** *** 1078,1084 **** strlcat(sql, savepoint, sizeof(sql)); psql = sql; ! pkg_debug(4, "Pkgdb: pkgdb_transaction_begin running '%s'", sql); ret = sqlite3_prepare_v2(sqlite, sql, strlen(sql) + 1, &stmt, NULL); } --- 1078,1084 ---- strlcat(sql, savepoint, sizeof(sql)); psql = sql; ! pkg_debug(4, "Pkgdb: running '%s'", sql); ret = sqlite3_prepare_v2(sqlite, sql, strlen(sql) + 1, &stmt, NULL); } *************** *** 1108,1114 **** const char sql[] = "COMMIT TRANSACTION"; psql = sql; ! pkg_debug(4, "Pkgdb: pkgdb_transaction_commit running '%s'", sql); ret = sqlite3_prepare_v2(sqlite, sql, strlen(sql) + 1, &stmt, NULL); } else { --- 1108,1114 ---- const char sql[] = "COMMIT TRANSACTION"; psql = sql; ! pkg_debug(4, "Pkgdb: running '%s'", sql); ret = sqlite3_prepare_v2(sqlite, sql, strlen(sql) + 1, &stmt, NULL); } else { *************** *** 1118,1124 **** psql = sql; ! pkg_debug(4, "Pkgdb: pkgdb_transaction_commit running '%s'", sql); ret = sqlite3_prepare_v2(sqlite, sql, strlen(sql) + 1, &stmt, NULL); } --- 1118,1124 ---- psql = sql; ! pkg_debug(4, "Pkgdb: running '%s'", sql); ret = sqlite3_prepare_v2(sqlite, sql, strlen(sql) + 1, &stmt, NULL); } *************** *** 1148,1154 **** const char sql[] = "ROLLBACK TRANSACTION"; psql = sql; ! pkg_debug(4, "Pkgdb: pkgdb_transaction_rollback running '%s'", sql); ret = sqlite3_prepare_v2(sqlite, sql, strlen(sql) + 1, &stmt, NULL); } else { --- 1148,1154 ---- const char sql[] = "ROLLBACK TRANSACTION"; psql = sql; ! pkg_debug(4, "Pkgdb: running '%s'", sql); ret = sqlite3_prepare_v2(sqlite, sql, strlen(sql) + 1, &stmt, NULL); } else { *************** *** 1157,1163 **** strlcat(sql, savepoint, sizeof(sql)); psql = sql; ! pkg_debug(4, "Pkgdb: pkgdb_transaction_rollback running '%s'", sql); ret = sqlite3_prepare_v2(sqlite, sql, strlen(sql) + 1, &stmt, NULL); } --- 1157,1163 ---- strlcat(sql, savepoint, sizeof(sql)); psql = sql; ! pkg_debug(4, "Pkgdb: running '%s'", sql); ret = sqlite3_prepare_v2(sqlite, sql, strlen(sql) + 1, &stmt, NULL); } *************** *** 1979,1985 **** for (i = 0; i < 2; i++) { /* Clean out old shlibs first */ ! pkg_debug(4, "Pkgdb: pkgdb_reanalyse_shlibs running '%s'", sql[i]); if (sqlite3_prepare_v2(db->sqlite, sql[i], -1, &stmt_del, NULL) != SQLITE_OK) { --- 1979,1985 ---- for (i = 0; i < 2; i++) { /* Clean out old shlibs first */ ! pkg_debug(4, "Pkgdb: running '%s'", sql[i]); if (sqlite3_prepare_v2(db->sqlite, sql[i], -1, &stmt_del, NULL) != SQLITE_OK) { *************** *** 2201,2207 **** assert(db != NULL); ! pkg_debug(4, "Pkgdb: pkgdb_unregister_pkg running '%s'", sql); if (sqlite3_prepare_v2(db->sqlite, sql, -1, &stmt_del, NULL) != SQLITE_OK){ ERROR_SQLITE(db->sqlite, sql); --- 2201,2207 ---- assert(db != NULL); ! pkg_debug(4, "Pkgdb: running '%s'", sql); if (sqlite3_prepare_v2(db->sqlite, sql, -1, &stmt_del, NULL) != SQLITE_OK){ ERROR_SQLITE(db->sqlite, sql); *************** *** 2247,2253 **** sql_to_exec = sql; } ! pkg_debug(4, "Pkgdb: sql_exec executing '%s'", sql_to_exec); if (sqlite3_exec(s, sql_to_exec, NULL, NULL, &errmsg) != SQLITE_OK) { ERROR_SQLITE(s, sql_to_exec); sqlite3_free(errmsg); --- 2247,2253 ---- sql_to_exec = sql; } ! pkg_debug(4, "Pkgdb: executing '%s'", sql_to_exec); if (sqlite3_exec(s, sql_to_exec, NULL, NULL, &errmsg) != SQLITE_OK) { ERROR_SQLITE(s, sql_to_exec); sqlite3_free(errmsg); *************** *** 2271,2277 **** assert(s != NULL && sql != NULL); ! pkg_debug(4, "Pkgdb: get_pragma running '%s'", sql); if (sqlite3_prepare_v2(s, sql, -1, &stmt, NULL) != SQLITE_OK) { if (!silence) ERROR_SQLITE(s, sql); --- 2271,2277 ---- assert(s != NULL && sql != NULL); ! pkg_debug(4, "Pkgdb: running '%s'", sql); if (sqlite3_prepare_v2(s, sql, -1, &stmt, NULL) != SQLITE_OK) { if (!silence) ERROR_SQLITE(s, sql); *************** *** 2303,2309 **** assert(s != NULL && sql != NULL); ! pkg_debug(4, "Pkgdb: get_sql_string running '%s'", sql); if (sqlite3_prepare_v2(s, sql, -1, &stmt, NULL) != SQLITE_OK) { ERROR_SQLITE(s, sql); return (EPKG_OK); --- 2303,2309 ---- assert(s != NULL && sql != NULL); ! pkg_debug(4, "Pkgdb: running '%s'", sql); if (sqlite3_prepare_v2(s, sql, -1, &stmt, NULL) != SQLITE_OK) { ERROR_SQLITE(s, sql); return (EPKG_OK); *************** *** 2382,2388 **** "path TEXT UNIQUE);" ); ! pkg_debug(4, "Pkgdb: pkgdb_integrity_append running '%s'", sql); if (sqlite3_prepare_v2(db->sqlite, sql, -1, &stmt, NULL) != SQLITE_OK) { ERROR_SQLITE(db->sqlite, sql); --- 2382,2388 ---- "path TEXT UNIQUE);" ); ! pkg_debug(4, "Pkgdb: running '%s'", sql); if (sqlite3_prepare_v2(db->sqlite, sql, -1, &stmt, NULL) != SQLITE_OK) { ERROR_SQLITE(db->sqlite, sql); *************** *** 2391,2397 **** pkg_get(p, PKG_UNIQUEID, &puid); ! pkg_debug(4, "Pkgdb: pkgdb_integrity_append test conflicts for %s", puid); while (pkg_files(p, &file) == EPKG_OK) { const char *uid; const char *pkg_path = pkg_file_path(file); --- 2391,2397 ---- pkg_get(p, PKG_UNIQUEID, &puid); ! pkg_debug(4, "Pkgdb: test conflicts for %s", puid); while (pkg_files(p, &file) == EPKG_OK) { const char *uid; const char *pkg_path = pkg_file_path(file); *************** *** 2403,2409 **** if (sqlite3_step(stmt) != SQLITE_DONE) { ! pkg_debug(4, "Pkgdb: pkgdb_integrity_append running '%s'", sql_conflicts); if (sqlite3_prepare_v2(db->sqlite, sql_conflicts, -1, &stmt_conflicts, NULL) != SQLITE_OK) { ERROR_SQLITE(db->sqlite, sql_conflicts); --- 2403,2409 ---- if (sqlite3_step(stmt) != SQLITE_DONE) { ! pkg_debug(4, "Pkgdb: running '%s'", sql_conflicts); if (sqlite3_prepare_v2(db->sqlite, sql_conflicts, -1, &stmt_conflicts, NULL) != SQLITE_OK) { ERROR_SQLITE(db->sqlite, sql_conflicts); *************** *** 2474,2480 **** * Select paths that are both in integritycheck and local table but their * UIDs are different */ ! pkg_debug(4, "Pkgdb: pkgdb_integrity_append running '%s'", sql_integrity_prepare); if (sqlite3_prepare_v2(db->sqlite, sql_integrity_prepare, -1, &stmt, NULL) != SQLITE_OK) { --- 2474,2480 ---- * Select paths that are both in integritycheck and local table but their * UIDs are different */ ! pkg_debug(4, "Pkgdb: running '%s'", sql_integrity_prepare); if (sqlite3_prepare_v2(db->sqlite, sql_integrity_prepare, -1, &stmt, NULL) != SQLITE_OK) { *************** *** 2495,2501 **** * Conflict found on path, so find the corresponding local * package */ ! pkg_debug(4, "Pkgdb: pkgdb_integrity_append running '%s'", sql_local_conflict); ret = sqlite3_prepare_v2(db->sqlite, sql_local_conflict, -1, &stmt_conflicts, NULL); if (ret != SQLITE_OK) { --- 2495,2501 ---- * Conflict found on path, so find the corresponding local * package */ ! pkg_debug(4, "Pkgdb: running '%s'", sql_local_conflict); ret = sqlite3_prepare_v2(db->sqlite, sql_local_conflict, -1, &stmt_conflicts, NULL); if (ret != SQLITE_OK) { *************** *** 2523,2529 **** /* * Now match local and remote conflicting packages */ ! pkg_debug(4, "Pkgdb: pkgdb_integrity_append running '%s'", sql_conflicts); ret = sqlite3_prepare_v2(db->sqlite, sql_conflicts, -1, &stmt_conflicts, NULL); if (ret != SQLITE_OK) { --- 2523,2529 ---- /* * Now match local and remote conflicting packages */ ! pkg_debug(4, "Pkgdb: running '%s'", sql_conflicts); ret = sqlite3_prepare_v2(db->sqlite, sql_conflicts, -1, &stmt_conflicts, NULL); if (ret != SQLITE_OK) { *************** *** 2579,2585 **** "AND i.uid = ?1 AND " "i.uid != p.name || '~' || p.origin"; ! pkg_debug(4, "Pkgdb: pkgdb_integrity_conflict_local running '%s'", sql_conflicts); ret = sqlite3_prepare_v2(db->sqlite, sql_conflicts, -1, &stmt, NULL); if (ret != SQLITE_OK) { ERROR_SQLITE(db->sqlite, sql_conflicts); --- 2579,2585 ---- "AND i.uid = ?1 AND " "i.uid != p.name || '~' || p.origin"; ! pkg_debug(4, "Pkgdb: running '%s'", sql_conflicts); ret = sqlite3_prepare_v2(db->sqlite, sql_conflicts, -1, &stmt, NULL); if (ret != SQLITE_OK) { ERROR_SQLITE(db->sqlite, sql_conflicts); *************** *** 2619,2625 **** }; while ((attr = va_arg(ap, int)) > 0) { ! pkg_debug(4, "Pkgdb: pkgdb_vset running '%s'", sql[attr]); if (sqlite3_prepare_v2(db->sqlite, sql[attr], -1, &stmt, NULL) != SQLITE_OK) { ERROR_SQLITE(db->sqlite, sql[attr]); --- 2619,2625 ---- }; while ((attr = va_arg(ap, int)) > 0) { ! pkg_debug(4, "Pkgdb: running '%s'", sql[attr]); if (sqlite3_prepare_v2(db->sqlite, sql[attr], -1, &stmt, NULL) != SQLITE_OK) { ERROR_SQLITE(db->sqlite, sql[attr]); *************** *** 2700,2706 **** "UPDATE files SET sha256 = ?1 WHERE path = ?2"; int ret; ! pkg_debug(4, "Pkgdb: pkgdb_file_set_cksum running '%s'", sql_file_update); ret = sqlite3_prepare_v2(db->sqlite, sql_file_update, -1, &stmt, NULL); if (ret != SQLITE_OK) { ERROR_SQLITE(db->sqlite, sql_file_update); --- 2700,2706 ---- "UPDATE files SET sha256 = ?1 WHERE path = ?2"; int ret; ! pkg_debug(4, "Pkgdb: running '%s'", sql_file_update); ret = sqlite3_prepare_v2(db->sqlite, sql_file_update, -1, &stmt, NULL); if (ret != SQLITE_OK) { ERROR_SQLITE(db->sqlite, sql_file_update); *************** *** 3102,3108 **** } sbuf_finish(sql); ! pkg_debug(4, "Pkgdb: pkgdb_stats running '%s'", sbuf_data(sql)); ret = sqlite3_prepare_v2(db->sqlite, sbuf_data(sql), -1, &stmt, NULL); if (ret != SQLITE_OK) { ERROR_SQLITE(db->sqlite, sbuf_data(sql)); --- 3102,3108 ---- } sbuf_finish(sql); ! pkg_debug(4, "Pkgdb: running '%s'", sbuf_data(sql)); ret = sqlite3_prepare_v2(db->sqlite, sbuf_data(sql), -1, &stmt, NULL); if (ret != SQLITE_OK) { ERROR_SQLITE(db->sqlite, sbuf_data(sql)); *** pkg-1.4.0.pre-alpha15/libpkg/pkgdb_iterator.c Mon Sep 15 12:40:03 2014 --- pkg-1.4.0.pre-alpha15.orig/libpkg/pkgdb_iterator.c Wed Aug 27 15:03:11 2014 *************** *** 83,89 **** { "origin", PKG_ORIGIN, PKG_SQLITE_STRING }, { "pkgsize", PKG_PKGSIZE, PKG_SQLITE_INT64 }, { "prefix", PKG_PREFIX, PKG_SQLITE_STRING }, - { "repo_order", PKG_REPO_ORDER, PKG_SQLITE_INT64 }, { "repopath", PKG_REPOPATH, PKG_SQLITE_STRING }, { "repourl", PKG_REPOURL, PKG_SQLITE_STRING }, { "rowid", PKG_ROWID, PKG_SQLITE_INT64 }, --- 83,88 ---- *************** *** 108,114 **** if (pkg->flags & flags) return (EPKG_OK); ! pkg_debug(4, "Pkgdb: load_val running '%s'", sql); if (sqlite3_prepare_v2(db, sql, -1, &stmt, NULL) != SQLITE_OK) { ERROR_SQLITE(db, sql); return (EPKG_FATAL); --- 107,113 ---- if (pkg->flags & flags) return (EPKG_OK); ! pkg_debug(4, "Pkgdb: running '%s'", sql); if (sqlite3_prepare_v2(db, sql, -1, &stmt, NULL) != SQLITE_OK) { ERROR_SQLITE(db, sql); return (EPKG_FATAL); *************** *** 148,154 **** if (pkg->flags & flags) return (EPKG_OK); ! pkg_debug(4, "Pkgdb: load_tag_val running '%s'", sql); if (sqlite3_prepare_v2(db, sql, -1, &stmt, NULL) != SQLITE_OK) { ERROR_SQLITE(db, sql); return (EPKG_FATAL); --- 147,153 ---- if (pkg->flags & flags) return (EPKG_OK); ! pkg_debug(4, "Pkgdb: running '%s'", sql); if (sqlite3_prepare_v2(db, sql, -1, &stmt, NULL) != SQLITE_OK) { ERROR_SQLITE(db, sql); return (EPKG_FATAL); *************** *** 204,210 **** return (EPKG_OK); ! pkg_debug(4, "Pkgdb: pkgdb_load_deps running '%s'", mainsql); ret = sqlite3_prepare_v2(sqlite, mainsql, -1, &stmt, NULL); if (ret != SQLITE_OK) { --- 203,209 ---- return (EPKG_OK); ! pkg_debug(4, "Pkgdb: running '%s'", mainsql); ret = sqlite3_prepare_v2(sqlite, mainsql, -1, &stmt, NULL); if (ret != SQLITE_OK) { *************** *** 253,259 **** return (EPKG_OK); ! pkg_debug(4, "Pkgdb: pkgdb_load_rdeps running '%s'", mainsql); ret = sqlite3_prepare_v2(sqlite, mainsql, -1, &stmt, NULL); if (ret != SQLITE_OK) { --- 252,258 ---- return (EPKG_OK); ! pkg_debug(4, "Pkgdb: running '%s'", mainsql); ret = sqlite3_prepare_v2(sqlite, mainsql, -1, &stmt, NULL); if (ret != SQLITE_OK) { *************** *** 301,307 **** if (pkg->flags & PKG_LOAD_FILES) return (EPKG_OK); ! pkg_debug(4, "Pkgdb: pkgdb_load_files running '%s'", sql); if (sqlite3_prepare_v2(sqlite, sql, -1, &stmt, NULL) != SQLITE_OK) { ERROR_SQLITE(sqlite, sql); return (EPKG_FATAL); --- 300,306 ---- if (pkg->flags & PKG_LOAD_FILES) return (EPKG_OK); ! pkg_debug(4, "Pkgdb: running '%s'", sql); if (sqlite3_prepare_v2(sqlite, sql, -1, &stmt, NULL) != SQLITE_OK) { ERROR_SQLITE(sqlite, sql); return (EPKG_FATAL); *************** *** 345,351 **** if (pkg->flags & PKG_LOAD_DIRS) return (EPKG_OK); ! pkg_debug(4, "Pkgdb: pkgdb_load_dirs running '%s'", sql); if (sqlite3_prepare_v2(sqlite, sql, -1, &stmt, NULL) != SQLITE_OK) { ERROR_SQLITE(sqlite, sql); return (EPKG_FATAL); --- 344,350 ---- if (pkg->flags & PKG_LOAD_DIRS) return (EPKG_OK); ! pkg_debug(4, "Pkgdb: running '%s'", sql); if (sqlite3_prepare_v2(sqlite, sql, -1, &stmt, NULL) != SQLITE_OK) { ERROR_SQLITE(sqlite, sql); return (EPKG_FATAL); *************** *** 543,549 **** if (pkg->flags & PKG_LOAD_SCRIPTS) return (EPKG_OK); ! pkg_debug(4, "Pkgdb: pkgdb_load_scripts running '%s'", sql); if (sqlite3_prepare_v2(sqlite, sql, -1, &stmt, NULL) != SQLITE_OK) { ERROR_SQLITE(sqlite, sql); return (EPKG_FATAL); --- 542,548 ---- if (pkg->flags & PKG_LOAD_SCRIPTS) return (EPKG_OK); ! pkg_debug(4, "Pkgdb: running '%s'", sql); if (sqlite3_prepare_v2(sqlite, sql, -1, &stmt, NULL) != SQLITE_OK) { ERROR_SQLITE(sqlite, sql); return (EPKG_FATAL); *** pkg-1.4.0.pre-alpha15/libpkg/private/pkg.h Mon Sep 15 12:33:26 2014 --- pkg-1.4.0.pre-alpha15.orig/libpkg/private/pkg.h Sun Sep 14 15:11:40 2014 *************** *** 351,358 **** pkg_repo_flags flags; - /* order for repository */ - int repo_order; /* Opaque repository data */ void *priv; }; --- 351,356 ---- *************** *** 436,450 **** #define PKG_DELETE_NOSCRIPT (1<<2) #define PKG_DELETE_CONFLICT (1<<3) ! ! struct pkg_key_x { ! int enum_tag; const char *name; int type; ! }; ! ! struct pkg_key_x *get_pkg_key_by_enum_tag( int enum_tag ); ! struct pkg_key_x *get_pkg_key_by_name( const char *name ); int pkg_fetch_file_to_fd(struct pkg_repo *repo, const char *url, int dest, time_t *t); --- 434,443 ---- #define PKG_DELETE_NOSCRIPT (1<<2) #define PKG_DELETE_CONFLICT (1<<3) ! extern struct pkg_key { const char *name; int type; ! } pkg_keys[PKG_NUM_FIELDS]; int pkg_fetch_file_to_fd(struct pkg_repo *repo, const char *url, int dest, time_t *t); *** pkg-1.4.0.pre-alpha15/libpkg/private/pkg_printf.h Mon Sep 15 12:24:03 2014 --- pkg-1.4.0.pre-alpha15.orig/libpkg/private/pkg_printf.h Sun Jun 8 12:59:53 2014 *************** *** 141,147 **** PP_PKG_CHECKSUM, PP_PKG_VERSION, PP_PKG_HOME_PAGE, - PP_PKG_REPO_ORDER, PP_PKG_SHORT_CHECKSUM, PP_LAST_FORMAT = PP_PKG_SHORT_CHECKSUM, PP_LITERAL_PERCENT, --- 141,146 ---- *************** *** 226,232 **** _static struct sbuf *format_short_checksum(struct sbuf *, const void *, struct percent_esc *); _static struct sbuf *format_version(struct sbuf *, const void *, struct percent_esc *); _static struct sbuf *format_home_url(struct sbuf *, const void *, struct percent_esc *); - _static struct sbuf *format_repo_order(struct sbuf *, const void *, struct percent_esc *); _static struct sbuf *format_literal_percent(struct sbuf *, __unused const void *, __unused struct percent_esc *); _static struct sbuf *format_unknown(struct sbuf *, __unused const void *, __unused struct percent_esc *); --- 225,230 ---- *** pkg-1.4.0.pre-alpha15/libpkg/repo/binary/init.c Mon Sep 15 12:23:10 2014 --- pkg-1.4.0.pre-alpha15.orig/libpkg/repo/binary/init.c Wed Aug 27 15:03:11 2014 *************** *** 137,143 **** /* apply change */ if (ret == EPKG_OK) { ! pkg_debug(4, "Pkgdb: pkg_repo_binary_apply_change running '%s'", change->sql); ret = sqlite3_exec(sqlite, change->sql, NULL, NULL, &errmsg); if (ret != SQLITE_OK) { pkg_emit_error("sqlite: %s", errmsg); --- 137,143 ---- /* apply change */ if (ret == EPKG_OK) { ! pkg_debug(4, "Pkgdb: running '%s'", change->sql); ret = sqlite3_exec(sqlite, change->sql, NULL, NULL, &errmsg); if (ret != SQLITE_OK) { pkg_emit_error("sqlite: %s", errmsg); *** pkg-1.4.0.pre-alpha15/libpkg/repo/binary/query.c Mon Sep 15 12:22:10 2014 --- pkg-1.4.0.pre-alpha15.orig/libpkg/repo/binary/query.c Wed Aug 27 15:03:11 2014 *************** *** 128,134 **** sbuf_cat(sql, " ORDER BY name;"); sbuf_finish(sql); ! pkg_debug(4, "Pkgdb: pkg_repo_binary_query running '%s' query for %s", sbuf_get(sql), pattern); ret = sqlite3_prepare_v2(sqlite, sbuf_get(sql), sbuf_size(sql), &stmt, NULL); if (ret != SQLITE_OK) { ERROR_SQLITE(sqlite, sbuf_get(sql)); --- 128,134 ---- sbuf_cat(sql, " ORDER BY name;"); sbuf_finish(sql); ! pkg_debug(4, "Pkgdb: running '%s' query for %s", sbuf_get(sql), pattern); ret = sqlite3_prepare_v2(sqlite, sbuf_get(sql), sbuf_size(sql), &stmt, NULL); if (ret != SQLITE_OK) { ERROR_SQLITE(sqlite, sbuf_get(sql)); *************** *** 167,173 **** sbuf_finish(sql); ! pkg_debug(4, "Pkgdb: pkg_repo_binary_shlib_provide running '%s'", sbuf_get(sql)); ret = sqlite3_prepare_v2(sqlite, sbuf_get(sql), -1, &stmt, NULL); if (ret != SQLITE_OK) { ERROR_SQLITE(sqlite, sbuf_get(sql)); --- 167,173 ---- sbuf_finish(sql); ! pkg_debug(4, "Pkgdb: running '%s'", sbuf_get(sql)); ret = sqlite3_prepare_v2(sqlite, sbuf_get(sql), -1, &stmt, NULL); if (ret != SQLITE_OK) { ERROR_SQLITE(sqlite, sbuf_get(sql)); *************** *** 204,210 **** sbuf_finish(sql); ! pkg_debug(4, "Pkgdb: pkg_repo_binary_shlib_require running '%s'", sbuf_get(sql)); ret = sqlite3_prepare_v2(sqlite, sbuf_get(sql), -1, &stmt, NULL); if (ret != SQLITE_OK) { ERROR_SQLITE(sqlite, sbuf_get(sql)); --- 204,210 ---- sbuf_finish(sql); ! pkg_debug(4, "Pkgdb: running '%s'", sbuf_get(sql)); ret = sqlite3_prepare_v2(sqlite, sbuf_get(sql), -1, &stmt, NULL); if (ret != SQLITE_OK) { ERROR_SQLITE(sqlite, sbuf_get(sql)); *************** *** 325,339 **** "SELECT id, origin, name, version, comment, " "prefix, desc, arch, maintainer, www, " "licenselogic, flatsize, pkgsize, " ! "cksum, path AS repopath, '%1$s' AS dbname, '%2$s' AS repourl, " ! "%3$d as repo_order " "FROM packages "; if (pattern == NULL || pattern[0] == '\0') return (NULL); sql = sbuf_new_auto(); ! sbuf_printf(sql, multireposql, repo->name, repo->url, repo->repo_order ); /* close the UNIONs and build the search query */ sbuf_cat(sql, "WHERE "); --- 325,338 ---- "SELECT id, origin, name, version, comment, " "prefix, desc, arch, maintainer, www, " "licenselogic, flatsize, pkgsize, " ! "cksum, path AS repopath, '%1$s' AS dbname, '%2$s' AS repourl " "FROM packages "; if (pattern == NULL || pattern[0] == '\0') return (NULL); sql = sbuf_new_auto(); ! sbuf_printf(sql, multireposql, repo->name, repo->url); /* close the UNIONs and build the search query */ sbuf_cat(sql, "WHERE "); *************** *** 342,348 **** sbuf_cat(sql, ";"); sbuf_finish(sql); ! pkg_debug(4, "Pkgdb: pkg_repo_binary_search running '%s'", sbuf_get(sql)); ret = sqlite3_prepare_v2(sqlite, sbuf_get(sql), -1, &stmt, NULL); if (ret != SQLITE_OK) { ERROR_SQLITE(sqlite, sbuf_get(sql)); --- 341,347 ---- sbuf_cat(sql, ";"); sbuf_finish(sql); ! pkg_debug(4, "Pkgdb: running '%s'", sbuf_get(sql)); ret = sqlite3_prepare_v2(sqlite, sbuf_get(sql), -1, &stmt, NULL); if (ret != SQLITE_OK) { ERROR_SQLITE(sqlite, sbuf_get(sql)); *************** *** 423,429 **** } sbuf_finish(sql); ! pkg_debug(4, "binary_repo: pkg_repo_binary_stat running '%s'", sbuf_data(sql)); ret = sqlite3_prepare_v2(sqlite, sbuf_data(sql), -1, &stmt, NULL); if (ret != SQLITE_OK) { ERROR_SQLITE(sqlite, sbuf_data(sql)); --- 422,428 ---- } sbuf_finish(sql); ! pkg_debug(4, "binary_repo: running '%s'", sbuf_data(sql)); ret = sqlite3_prepare_v2(sqlite, sbuf_data(sql), -1, &stmt, NULL); if (ret != SQLITE_OK) { ERROR_SQLITE(sqlite, sbuf_data(sql)); *** pkg-1.4.0.pre-alpha15/src/check.c Mon Sep 15 12:17:49 2014 --- pkg-1.4.0.pre-alpha15.orig/src/check.c Thu Aug 28 11:58:05 2014 *************** *** 253,259 **** bool reanalyse_shlibs = false; bool noinstall = false; int nbpkgs = 0; ! int i, processed = 0, total = 0; int verbose = 0; struct option longopts[] = { --- 253,259 ---- bool reanalyse_shlibs = false; bool noinstall = false; int nbpkgs = 0; ! int i, processed, total; int verbose = 0; struct option longopts[] = { *** pkg-1.4.0.pre-alpha15/src/event.c Mon Sep 15 12:14:22 2014 --- pkg-1.4.0.pre-alpha15.orig/src/event.c Thu Aug 28 11:58:05 2014 *************** *** 412,418 **** { int percent; int64_t transferred; ! time_t elapsed = 0, now = 0; char buf[7]; int64_t bytes_left; int cur_speed; --- 412,418 ---- { int percent; int64_t transferred; ! time_t elapsed, now; char buf[7]; int64_t bytes_left; int cur_speed; *** pkg-1.4.0.pre-alpha15/src/pkgcli.h Mon Sep 15 12:10:10 2014 --- pkg-1.4.0.pre-alpha15.orig/src/pkgcli.h Wed Aug 27 15:03:11 2014 *************** *** 214,226 **** #define INFO_DIRS (1LL<<23) #define INFO_USERS (1LL<<24) #define INFO_GROUPS (1LL<<25) ! #define INFO_REPOURL (1LL<<26) #define INFO_LOCKED (1LL<<27) ! #define INFO_REPO_ORDER (1LL<<28) ! #define INFO_OPTION_DEFAULTS (1LL<<29) ! #define INFO_OPTION_DESCRIPTIONS (1LL<<30) ! #define INFO_LASTFIELD INFO_REPO_ORDER #define INFO_ALL (((INFO_LASTFIELD) << 1) - 1) /* Identifying tags */ --- 214,225 ---- #define INFO_DIRS (1LL<<23) #define INFO_USERS (1LL<<24) #define INFO_GROUPS (1LL<<25) ! #define INFO_REPOURL (1LL<<26) #define INFO_LOCKED (1LL<<27) ! #define INFO_OPTION_DEFAULTS (1LL<<28) ! #define INFO_OPTION_DESCRIPTIONS (1LL<<29) ! #define INFO_LASTFIELD INFO_LOCKED #define INFO_ALL (((INFO_LASTFIELD) << 1) - 1) /* Identifying tags */ *** pkg-1.4.0.pre-alpha15/src/search.c Mon Sep 15 12:08:38 2014 --- pkg-1.4.0.pre-alpha15.orig/src/search.c Mon Aug 4 14:59:40 2014 *************** *** 69,75 **** { "options", 'o' }, { "pkg-size", 'P' }, { "prefix", 'p' }, - { "repo_order", 'Z' }, { "repository", 'R' }, { "required-by", 'r' }, { "shared-libs-required", 'B' }, --- 69,74 ---- *************** *** 177,185 **** case 'o': opt = INFO_OPTIONS; break; - case 'Z': - opt = INFO_REPO_ORDER; - break; case 'P': opt = INFO_PKGSIZE; break; --- 176,181 ---- *************** *** 212,218 **** break; default: usage_search(); ! errx(EX_USAGE, "Unknown modifier option %s", optionarg); /* NOTREACHED */ } return opt; --- 208,214 ---- break; default: usage_search(); ! errx(EX_USAGE, "Unkown modifier option %s", optionarg); /* NOTREACHED */ } return opt; *** pkg-1.4.0.pre-alpha15/src/utils.c Mon Sep 15 12:03:42 2014 --- pkg-1.4.0.pre-alpha15.orig/src/utils.c Wed Aug 27 15:03:11 2014 *************** *** 472,482 **** printf("%-15s: ", "WWW"); pkg_printf("%w\n", pkg); break; - case INFO_REPO_ORDER: - if (print_tag) - printf("%-15s: ", "Repo_order"); - pkg_printf("%Z\n", pkg); - break; case INFO_COMMENT: if (print_tag) printf("%-15s: ", "Comment"); --- 472,477 ---- *** pkg-1.4.0.pre-alpha15/src/which.c Mon Sep 15 12:03:02 2014 --- pkg-1.4.0.pre-alpha15.orig/src/which.c Wed Aug 27 15:03:11 2014 *************** *** 57,66 **** struct pkgdb_it *it = NULL; struct pkg *pkg = NULL; char pathabs[MAXPATHLEN]; ! char *p, *path = 0, *match; int ret = EPKG_OK, retcode = EX_SOFTWARE; int ch; ! int res, pathlen = 0; bool orig = false; bool glob = false; bool search = false; --- 57,66 ---- struct pkgdb_it *it = NULL; struct pkg *pkg = NULL; char pathabs[MAXPATHLEN]; ! char *p, *path, *match; int ret = EPKG_OK, retcode = EX_SOFTWARE; int ch; ! int res, pathlen; bool orig = false; bool glob = false; bool search = false; From papowell at astart.com Wed Sep 24 03:46:31 2014 From: papowell at astart.com (Patrick Powell) Date: Tue, 23 Sep 2014 20:46:29 -0700 Subject: FreeBSD Port: x11-servers/xorg-server In-Reply-To: References: <5414A4B6.5010706@UToledo.edu> <5416B1FE.1010801@dumbbell.fr> <1411306792093-5950838.post@n5.nabble.com> <1411317698961-5950873.post@n5.nabble.com> <541F2CA4.6@hiwaay.net> <1411341142791-5950945.post@n5.nabble.com> <541F64A4.3030002@hiwaay.net> <542096D6.2000403@astart.com> <5420F4C9.7090109@hiwaay.net> Message-ID: <54223E95.6080305@astart.com> On 09/22/14 23:50, Kevin Oberman wrote: > On Mon, Sep 22, 2014 at 9:19 PM, William A. Mahaffey III > wrote: > >> On 09/22/14 16:38, Patrick Powell wrote: >> >>> On 09/21/14 16:52, William A. Mahaffey III wrote: >>> >>>> On 09/21/14 18:12, Robert_Burmeister wrote: >>>> >>>>> William A. Mahaffey III wrote >>>>> >>>>>> On 09/21/14 11:41, Robert_Burmeister wrote: >>>>>> >>>>>>> On 13.09.2014 22:10, Robert Burmeister wrote: >>>>>>>>>>> FreeBSD 10.1 i386 >>>>>>>>>>> xorg-server 1.12.4_9,1 and 1.12.4_1,1 >>>>>>>>>>> Still don't have mouse support after upgrade from 1.12.4_8,1 >>>>>>>>>>> >>>>>>>>>> [ 1786.822] (EE) Failed to load module "mouse" (module does not >>>>>>>>> exist, >>>>>>>>> 0) >>>>>>>>> >>>>>>>> Have you installed x11-drivers/xf86-input-mouse? >>>>>>>> >>>>>>>>> [ 1786.825] (EE) Failed to load module "kbd" (module does not >>>>>>>>> exist, >>>>>>>>> 0) >>>>>>>>> >>>>>>>> And x11-drivers/xf86-input-keyboard? >>>>>>>> _______________________________________________ >>>>>>>> >>>>>>> Installing x11-drivers/xf86-input-mouse and >>>>>>> x11-drivers/xf86-input-keyboard >>>>>>> fixed the problem, however, I don't understand why upgrading >>>>>>> from xorg-server 1.12.4_8,1 to xorg-server 1.12.4_9,1 >>>>>>> would require new drivers, or lose the ones it had. >>>>>>> >>>>>>> I would think these drivers would/should be a dependency for >>>>>>> xorg-server >>>>>>> in the Ports system... >>>>>>> _______________________________________________ >>>>>>> >>>>>> I have had that same problem verbatim the last 2 x-server upgrades I >>>>>> did, & that was the fix, (re?)install the kbd & mouse drivers. I >>>>>> (pkg-)upgraded this A.M., no such issues .... >>>>>> ---------------------------------------------------------------------- >>>>>> >>>>> Even more interesting... >>>>> x11-drivers/xf86-input-mouse and x11-drivers/xf86-input-keyboard >>>>> have xorg-server as a dependency, and so cannot be a circular >>>>> dependency. >>>>> >>>>> I'm guessing that the mouse and keyboard drivers got deleted as >>>>> dependents >>>>> of >>>>> xorg-server during the upgrade, but there are no dependencies in my >>>>> desktop >>>>> build >>>>> process that require that they be put back, even through a complete >>>>> system >>>>> recompile. >>>>> >>>>> I'm thinking 'x11-drivers/xorg-drivers' and 'x11/xorg-minimal' should be >>>>> bumped >>>>> when xorg-server is upgraded. >>>>> >>>>> (When my current recompile is done, I will check that my xorg-drivers >>>>> didn't >>>>> get removed as well.) >>>>> >>>> >>>> I am using pkg, no ports, no recompiling .... FBSD 9.3, BTW .... >>>> >>>> >>>> Just a thought - check to make sure that the x11-drivers/xf86-input-keyboard >>> port on the PKGng server you are using was actually rebuilt for the new >>> version of xorg-server. I had this problem a couple of weeks ago and the X >>> log file hinted that the keyboard driver was not compatible with the >>> version of xorg-server. At the time I thought that this was due to a lag >>> in the PKGng server building the new drivers so I compiled and installed >>> the x11-drivers/xf86-input-keyboard. The problem went away. >>> >>> This appears to be the same sort of issue. Perhaps the driver(s) are not >>> getting rebuilt for the new version of xorg-server? Or perhaps this is >>> related to packages on the 'with_new_xorg' PKGng server versus packages on >>> the standard PKGng server? >>> _______________________________________________ >>> freebsd-x11 at freebsd.org mailing list >>> http://lists.freebsd.org/mailman/listinfo/freebsd-x11 >>> To unsubscribe, send any mail to "freebsd-x11-unsubscribe at freebsd.org" >>> >>> >> Would that (drivers not getting rebuilt right) be a pkg problem or an X11 >> problem (or something else) ? Just checking .... >> >> >> -- >> >> William A. Mahaffey III > > It's a ports issue. The ports system (which is used by pkg) depends on > incrementing version numbers to tell when a port has been updated and needs > to be rebuilt or when a port needs updating because a dependency has been > updated in a way that affects the port. The former is typically the result > of a change to a port that changes hte version number. The later is a bump > of the PORT_REVISION to indicate to the ports system that some change that > did not come from a change upstream, but local to FreeBSD requires a port > rebuild. > > In this case, it does not work. The actual version number has not changed > as the upstream version has not changed. PORT_REVISION would result in the > ports being rebuilt, but that does not play with the definition of > WITH_NEW_XORG. I somehow needs to be bumped when any system sets > WITH_NEW_XORG and I don't see any way in hte current structure to do this. > It is the result of having two parallel ports trees. > > One possible fix is to have code in the Makefile to check WITH_NEW_XORG > and, if it is defined, use a different PORT_REVISION. If the old Xorg > driver gets a bump of PORT_REVESION, the new one would, as well, but I > don't see any reason this could not be done as both numbers are in the same > Makefile and should be only a few lines apart. something like: > . if defined(WITH_NEW_XORG) > PORT_REVISION=2 > . else > PORT_REVISION=1 > . endif > > This may break the index, so I'm not sure it would work as simply as this, > but I bet it could be made to work, > -- > R. Kevin Oberman, Network Engineer, Retired > E-mail: rkoberman at gmail.com > _______________________________________________ > freebsd-x11 at freebsd.org mailing list > http://lists.freebsd.org/mailman/listinfo/freebsd-x11 > To unsubscribe, send any mail to "freebsd-x11-unsubscribe at freebsd.org" > I can't check this out right now, BUT are the keyboard/mouse drivers on the WITH_NEW_XORG repo server built correctly? If that is the case then you can force PKG to fetch them from that REPO and then you can (using some magic I don't understand, setting something in the comment field) force PKG to always fetch from this repo. I have a plan B on this, which is to have a 'repo search order' capability added to PKG. IF you search the WITH_NEW_XORG repo first, THEN search the standard repo AND if you have two packages with the same version, etc, then you get it from the first repository you searched. If the packages on the WITH_NEW_XORG server have the same version/etc as the packages on the standardard server BUT they have been built with NEW_XORG then this should work. I think. Perhaps. Maybe. This idea may be bogus. From wam at hiwaay.net Wed Sep 24 04:23:54 2014 From: wam at hiwaay.net (William A. Mahaffey III) Date: Tue, 23 Sep 2014 23:30:06 -0500 Subject: FreeBSD Port: x11-servers/xorg-server In-Reply-To: <54223E95.6080305@astart.com> References: <5414A4B6.5010706@UToledo.edu> <5416B1FE.1010801@dumbbell.fr> <1411306792093-5950838.post@n5.nabble.com> <1411317698961-5950873.post@n5.nabble.com> <541F2CA4.6@hiwaay.net> <1411341142791-5950945.post@n5.nabble.com> <541F64A4.3030002@hiwaay.net> <542096D6.2000403@astart.com> <5420F4C9.7090109@hiwaay.net> <54223E95.6080305@astart.com> Message-ID: <542248CE.9050202@hiwaay.net> On 09/23/14 22:46, Patrick Powell wrote: > On 09/22/14 23:50, Kevin Oberman wrote: >> On Mon, Sep 22, 2014 at 9:19 PM, William A. Mahaffey III >> >> wrote: >> >>> On 09/22/14 16:38, Patrick Powell wrote: >>> >>>> On 09/21/14 16:52, William A. Mahaffey III wrote: >>>> >>>>> On 09/21/14 18:12, Robert_Burmeister wrote: >>>>> >>>>>> William A. Mahaffey III wrote >>>>>> >>>>>>> On 09/21/14 11:41, Robert_Burmeister wrote: >>>>>>> >>>>>>>> On 13.09.2014 22:10, Robert Burmeister wrote: >>>>>>>>>>>> FreeBSD 10.1 i386 >>>>>>>>>>>> xorg-server 1.12.4_9,1 and 1.12.4_1,1 >>>>>>>>>>>> Still don't have mouse support after upgrade from 1.12.4_8,1 >>>>>>>>>>>> >>>>>>>>>>> [ 1786.822] (EE) Failed to load module "mouse" (module does >>>>>>>>>>> not >>>>>>>>>> exist, >>>>>>>>>> 0) >>>>>>>>>> >>>>>>>>> Have you installed x11-drivers/xf86-input-mouse? >>>>>>>>> >>>>>>>>>> [ 1786.825] (EE) Failed to load module "kbd" (module does not >>>>>>>>>> exist, >>>>>>>>>> 0) >>>>>>>>>> >>>>>>>>> And x11-drivers/xf86-input-keyboard? >>>>>>>>> _______________________________________________ >>>>>>>>> >>>>>>>> Installing x11-drivers/xf86-input-mouse and >>>>>>>> x11-drivers/xf86-input-keyboard >>>>>>>> fixed the problem, however, I don't understand why upgrading >>>>>>>> from xorg-server 1.12.4_8,1 to xorg-server 1.12.4_9,1 >>>>>>>> would require new drivers, or lose the ones it had. >>>>>>>> >>>>>>>> I would think these drivers would/should be a dependency for >>>>>>>> xorg-server >>>>>>>> in the Ports system... >>>>>>>> _______________________________________________ >>>>>>>> >>>>>>> I have had that same problem verbatim the last 2 x-server >>>>>>> upgrades I >>>>>>> did, & that was the fix, (re?)install the kbd & mouse drivers. I >>>>>>> (pkg-)upgraded this A.M., no such issues .... >>>>>>> ---------------------------------------------------------------------- >>>>>>> >>>>>>> >>>>>> Even more interesting... >>>>>> x11-drivers/xf86-input-mouse and x11-drivers/xf86-input-keyboard >>>>>> have xorg-server as a dependency, and so cannot be a circular >>>>>> dependency. >>>>>> >>>>>> I'm guessing that the mouse and keyboard drivers got deleted as >>>>>> dependents >>>>>> of >>>>>> xorg-server during the upgrade, but there are no dependencies in my >>>>>> desktop >>>>>> build >>>>>> process that require that they be put back, even through a complete >>>>>> system >>>>>> recompile. >>>>>> >>>>>> I'm thinking 'x11-drivers/xorg-drivers' and 'x11/xorg-minimal' >>>>>> should be >>>>>> bumped >>>>>> when xorg-server is upgraded. >>>>>> >>>>>> (When my current recompile is done, I will check that my >>>>>> xorg-drivers >>>>>> didn't >>>>>> get removed as well.) >>>>>> >>>>> >>>>> I am using pkg, no ports, no recompiling .... FBSD 9.3, BTW .... >>>>> >>>>> >>>>> Just a thought - check to make sure that the >>>>> x11-drivers/xf86-input-keyboard >>>> port on the PKGng server you are using was actually rebuilt for >>>> the new >>>> version of xorg-server. I had this problem a couple of weeks ago >>>> and the X >>>> log file hinted that the keyboard driver was not compatible with the >>>> version of xorg-server. At the time I thought that this was due to >>>> a lag >>>> in the PKGng server building the new drivers so I compiled and >>>> installed >>>> the x11-drivers/xf86-input-keyboard. The problem went away. >>>> >>>> This appears to be the same sort of issue. Perhaps the driver(s) >>>> are not >>>> getting rebuilt for the new version of xorg-server? Or perhaps >>>> this is >>>> related to packages on the 'with_new_xorg' PKGng server versus >>>> packages on >>>> the standard PKGng server? >>>> _______________________________________________ >>>> freebsd-x11 at freebsd.org mailing list >>>> http://lists.freebsd.org/mailman/listinfo/freebsd-x11 >>>> To unsubscribe, send any mail to "freebsd-x11-unsubscribe at freebsd.org" >>>> >>>> >>> Would that (drivers not getting rebuilt right) be a pkg problem or >>> an X11 >>> problem (or something else) ? Just checking .... >>> >>> >>> -- >>> >>> William A. Mahaffey III >> >> It's a ports issue. The ports system (which is used by pkg) depends on >> incrementing version numbers to tell when a port has been updated and >> needs >> to be rebuilt or when a port needs updating because a dependency has >> been >> updated in a way that affects the port. The former is typically the >> result >> of a change to a port that changes hte version number. The later is a >> bump >> of the PORT_REVISION to indicate to the ports system that some change >> that >> did not come from a change upstream, but local to FreeBSD requires a >> port >> rebuild. >> >> In this case, it does not work. The actual version number has not >> changed >> as the upstream version has not changed. PORT_REVISION would result >> in the >> ports being rebuilt, but that does not play with the definition of >> WITH_NEW_XORG. I somehow needs to be bumped when any system sets >> WITH_NEW_XORG and I don't see any way in hte current structure to do >> this. >> It is the result of having two parallel ports trees. >> >> One possible fix is to have code in the Makefile to check WITH_NEW_XORG >> and, if it is defined, use a different PORT_REVISION. If the old Xorg >> driver gets a bump of PORT_REVESION, the new one would, as well, but I >> don't see any reason this could not be done as both numbers are in >> the same >> Makefile and should be only a few lines apart. something like: >> . if defined(WITH_NEW_XORG) >> PORT_REVISION=2 >> . else >> PORT_REVISION=1 >> . endif >> >> This may break the index, so I'm not sure it would work as simply as >> this, >> but I bet it could be made to work, >> -- >> R. Kevin Oberman, Network Engineer, Retired >> E-mail: rkoberman at gmail.com >> _______________________________________________ >> freebsd-x11 at freebsd.org mailing list >> http://lists.freebsd.org/mailman/listinfo/freebsd-x11 >> To unsubscribe, send any mail to "freebsd-x11-unsubscribe at freebsd.org" >> > I can't check this out right now, BUT are the keyboard/mouse drivers > on the WITH_NEW_XORG > repo server built correctly? If that is the case then you can force > PKG to fetch them from that REPO > and then you can (using some magic I don't understand, setting > something in the comment field) force > PKG to always fetch from this repo. I can't vouch for how they're built, but here's where they came from: [root at kabini1, /etc, 11:25:16pm] 324 % pkg query -g '%n: %R' 'xf86*' xf86-input-keyboard: FreeBSD_new_xorg xf86-input-mouse: FreeBSD_new_xorg xf86-video-ati: FreeBSD_new_xorg xf86-video-intel: FreeBSD_new_xorg xf86-video-mach64: FreeBSD xf86-video-nv: FreeBSD xf86-video-openchrome: FreeBSD xf86-video-r128: FreeBSD xf86-video-vesa: FreeBSD_new_xorg xf86dga: FreeBSD xf86dgaproto: FreeBSD xf86driproto: unknown-repository xf86miscproto: FreeBSD xf86vidmodeproto: FreeBSD [root at kabini1, /etc, 11:25:19pm] 325 % > > I have a plan B on this, which is to have a 'repo search order' > capability added to PKG. > IF you search the WITH_NEW_XORG repo first, THEN search the standard > repo > AND if you have two packages with the same version, etc, then you get > it from the first > repository you searched. Someone posted that very suggestion a week or so ago, generated several replies, & it was rejected (IIRC) as a bad idea for some reason .... I like it myself, but what do I know ???? > > If the packages on the WITH_NEW_XORG server have the same version/etc > as the packages on the > standardard server BUT they have been built with NEW_XORG then this > should work. > > I think. Perhaps. Maybe. This idea may be bogus. > _______________________________________________ > freebsd-x11 at freebsd.org mailing list > http://lists.freebsd.org/mailman/listinfo/freebsd-x11 > To unsubscribe, send any mail to "freebsd-x11-unsubscribe at freebsd.org" > -- William A. Mahaffey III ---------------------------------------------------------------------- "The M1 Garand is without doubt the finest implement of war ever devised by man." -- Gen. George S. Patton Jr. From jmc-freebsd2 at milibyte.co.uk Wed Sep 24 11:53:10 2014 From: jmc-freebsd2 at milibyte.co.uk (Mike Clarke) Date: Wed, 24 Sep 2014 12:52:55 +0100 Subject: FreeBSD Port: x11-servers/xorg-server In-Reply-To: <54223E95.6080305@astart.com> References: <5414A4B6.5010706@UToledo.edu> <54223E95.6080305@astart.com> Message-ID: <1661807.9J09C5k3ze@curlew.lan> On Tuesday 23 Sep 2014 20:46:29 Patrick Powell wrote: > I can't check this out right now, BUT are the keyboard/mouse > drivers on the WITH_NEW_XORG > repo server built correctly? They appear to be built OK, curlew:/home/mike% pkg rquery "%n %v %R" xf86-input-mouse xf86-input-mouse 1.9.0_4 FreeBSD xf86-input-mouse 1.9.0_4 FreeBSD_new_xorg curlew:/home/mike% pkg rquery "%n %v %R" xf86-input-keyboard xf86-input-keyboard 1.8.0_5 FreeBSD xf86-input-keyboard 1.8.0_5 FreeBSD_new_xorg Both repos have the same version but when I ran pkg upgrade it used FreeBSD for both these drivers with the result that I couldn't log in with KDM until I forcibly installed them from FreeBSD_new_xorg > If that is the case then you can > force PKG to fetch them from that REPO > and then you can (using some magic I don't understand, setting > something in the comment field) force > PKG to always fetch from this repo. > > I have a plan B on this, which is to have a 'repo search order' > capability added to PKG. > IF you search the WITH_NEW_XORG repo first, THEN search the > standard repo AND if you have two packages with the same version, > etc, then you get it from the first > repository you searched. Would an alternative approach when upgrading a package be to check to see if the existing package is annotated with a repository tag and use its value to decide which repository to use. This would have worked in my above example. curlew:/home/mike% pkg query "%n %At %Av" xf86-input-mouse xf86-input- keyboard xf86-input-mouse repo_type binary xf86-input-mouse repository FreeBSD_new_xorg xf86-input-keyboard repo_type binary xf86-input-keyboard repository FreeBSD_new_xorg It would however require the repository to be specified by the user when initially installing the package. The instructions in https://lists.freebsd.org/pipermail/freebsd-ports/2013-November/087487.html give the impression that the value of the annotation should influence the choice of repository but this does not appear to happen. -- Mike Clarke From milu at dat.pl Wed Sep 24 21:28:25 2014 From: milu at dat.pl (Maciej Milewski) Date: Wed, 24 Sep 2014 23:28:17 +0200 Subject: Can't compile pkg on mips Message-ID: <54233771.4090803@dat.pl> Hi, I have problem with compiling pkg-1.3.8 on my RSPro mips board: ===> Building for pkg-1.3.8 /usr/bin/make all-recursive Making all in external /usr/bin/make all-am Making all in libpkg Making all in repo Making all in binary Making all in . Making all in src CC pkg-info.o cc1: warnings being treated as errors info.c: In function 'exec_info': info.c:266: warning: implicit declaration of function 'cap_rights_init' info.c:267: warning: passing argument 2 of 'cap_rights_limit' makes integer from pointer without a cast *** [pkg-info.o] Error code 1 make: stopped in /data/builds/usr/ports/ports-mgmt/pkg/work/pkg-1.3.8/src 1 error make: stopped in /data/builds/usr/ports/ports-mgmt/pkg/work/pkg-1.3.8/src *** [all-recursive] Error code 1 make: stopped in /data/builds/usr/ports/ports-mgmt/pkg/work/pkg-1.3.8 1 error make: stopped in /data/builds/usr/ports/ports-mgmt/pkg/work/pkg-1.3.8 *** [all] Error code 2 make: stopped in /data/builds/usr/ports/ports-mgmt/pkg/work/pkg-1.3.8 1 error make: stopped in /data/builds/usr/ports/ports-mgmt/pkg/work/pkg-1.3.8 ===> Compilation failed unexpectedly. Try to set MAKE_JOBS_UNSAFE=yes and rebuild before reporting the failure to the maintainer. *** Error code 1 Stop. make: stopped in /usr/ports/ports-mgmt/pkg *** Error code 1 Stop. make: stopped in /usr/ports/ports-mgmt/pkg However my system isn't quite fresh (it's 10-CURRENT r253582) if that might make the difference. Currently ports are throwing: ===> pkgconf-0.9.7 pkg(8) must be version 1.3.8 or greater, but you have 1.2.7_3. You must upgrade pkg(8) first. and I'm stuck. Any hints are welcome. -- Pozdrawiam, Maciej Milewski From p at pierre-selosse.com Wed Sep 24 22:55:08 2014 From: p at pierre-selosse.com (Pierre) Date: Wed, 24 Sep 2014 22:48:10 +0000 (UTC) Subject: firefox: pkg: Cannot solve problem using SAT solver References: <201409140845.s8E8jmkC000795@mech-as221.men.bris.ac.uk> <541558C4.507@FreeBSD.org> Message-ID: Radek Krej?a writes: > > > > Have I forgotten something? > > > > You're the second person to report this. It does seem to be a problem > > with the firefox port in the FreeBSD repository. Looks like it > > conflicts with it's own dependencies. > > It looks, that problem should be in jpeg-turbo, because some packages need jpeg and others jpeg-turbo. But > I dont know if is the reason. > > And this situation is about 3 weeks old. > > Radek > _______________________________________________ > freebsd-pkg at ... mailing list > https://lists.freebsd.org/mailman/listinfo/freebsd-pkg > To unsubscribe, send any mail to "freebsd-pkg-unsubscribe at ..." > > I got a same issue installing Mate. So I just typed pkg add /var/cache/pkg/ mate-x.x.x, it works pierre From cmeyer1969+freebsd at gmail.com Thu Sep 25 15:20:56 2014 From: cmeyer1969+freebsd at gmail.com (Chris Meyer) Date: Thu, 25 Sep 2014 08:20:54 -0700 Subject: Updating openssl with pkg Message-ID: I'm running 9.2-RELEASE-p12 and using pkg (original system switched to pkg using pkg2ng). The system comes with openssl 0.9.8y. I have done 'pkg install openssl' and 'pkg info' shows openssl-1.0.1_15 is now installed. However, when I do 'openssl version' it still shows version 0.9.8y. How do I get the system to use the new version of openssl when installed using pkg? From tfransosi at gmail.com Thu Sep 25 16:01:26 2014 From: tfransosi at gmail.com (Thiago Farina) Date: Thu, 25 Sep 2014 13:01:23 -0300 Subject: Can't compile pkg on mips In-Reply-To: <54233771.4090803@dat.pl> References: <54233771.4090803@dat.pl> Message-ID: On Wed, Sep 24, 2014 at 6:28 PM, Maciej Milewski wrote: > Hi, > I have problem with compiling pkg-1.3.8 on my RSPro mips board: > ===> Building for pkg-1.3.8 > /usr/bin/make all-recursive > Making all in external > /usr/bin/make all-am > Making all in libpkg > Making all in repo > Making all in binary > Making all in . > Making all in src > CC pkg-info.o > cc1: warnings being treated as errors > info.c: In function 'exec_info': > info.c:266: warning: implicit declaration of function 'cap_rights_init' > info.c:267: warning: passing argument 2 of 'cap_rights_limit' makes > integer from pointer without a cast > *** [pkg-info.o] Error code 1 > Have you tried fixing this warning? As it is treating warnings as errors the compilation fails. Hope that helps, Regards, -- Thiago Farina From milu at dat.pl Thu Sep 25 16:19:54 2014 From: milu at dat.pl (Maciej Milewski) Date: Thu, 25 Sep 2014 18:19:49 +0200 Subject: Can't compile pkg on mips In-Reply-To: References: <54233771.4090803@dat.pl> Message-ID: <057b903c-628f-4843-b7de-c0f7cb9750b4@email.android.com> On 25 wrze?nia 2014 18:01:23 CEST, Thiago Farina wrote: >On Wed, Sep 24, 2014 at 6:28 PM, Maciej Milewski wrote: >> Hi, >> I have problem with compiling pkg-1.3.8 on my RSPro mips board: >> ===> Building for pkg-1.3.8 >> /usr/bin/make all-recursive >> Making all in external >> /usr/bin/make all-am >> Making all in libpkg >> Making all in repo >> Making all in binary >> Making all in . >> Making all in src >> CC pkg-info.o >> cc1: warnings being treated as errors >> info.c: In function 'exec_info': >> info.c:266: warning: implicit declaration of function >'cap_rights_init' >> info.c:267: warning: passing argument 2 of 'cap_rights_limit' makes >> integer from pointer without a cast >> *** [pkg-info.o] Error code 1 >> >Have you tried fixing this warning? As it is treating warnings as >errors the compilation fails. > >Hope that helps, > >Regards, Yes tried removing -Werror from src/Makefile. If that's sth connected with capsicum then my system might be built before that got to the tree. I'll try of I can build and flash newer image for that device. -- Wys?ane z Androida za pomoc? K-9 Mail. Prosze wybaczy? lakoniczno??. From bdrewery at FreeBSD.org Thu Sep 25 16:27:20 2014 From: bdrewery at FreeBSD.org (Bryan Drewery) Date: Thu, 25 Sep 2014 11:27:06 -0500 Subject: Can't compile pkg on mips In-Reply-To: <057b903c-628f-4843-b7de-c0f7cb9750b4@email.android.com> References: <54233771.4090803@dat.pl> <057b903c-628f-4843-b7de-c0f7cb9750b4@email.android.com> Message-ID: <5424425A.9040709@FreeBSD.org> On 9/25/2014 11:19 AM, Maciej Milewski wrote: > > > On 25 wrze?nia 2014 18:01:23 CEST, Thiago Farina wrote: >> On Wed, Sep 24, 2014 at 6:28 PM, Maciej Milewski wrote: >>> Hi, >>> I have problem with compiling pkg-1.3.8 on my RSPro mips board: >>> ===> Building for pkg-1.3.8 >>> /usr/bin/make all-recursive >>> Making all in external >>> /usr/bin/make all-am >>> Making all in libpkg >>> Making all in repo >>> Making all in binary >>> Making all in . >>> Making all in src >>> CC pkg-info.o >>> cc1: warnings being treated as errors >>> info.c: In function 'exec_info': >>> info.c:266: warning: implicit declaration of function >> 'cap_rights_init' >>> info.c:267: warning: passing argument 2 of 'cap_rights_limit' makes >>> integer from pointer without a cast >>> *** [pkg-info.o] Error code 1 >>> >> Have you tried fixing this warning? As it is treating warnings as >> errors the compilation fails. >> >> Hope that helps, >> >> Regards, > Yes tried removing -Werror from src/Makefile. If that's sth connected with capsicum then my system might be built before that got to the tree. I'll try of I can build and flash newer image for that device. > Huh that's strange. The header is included properly. What FreeBSD version is this? -- Regards, Bryan Drewery -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 488 bytes Desc: OpenPGP digital signature URL: From peters at connections.ca Thu Sep 25 19:33:33 2014 From: peters at connections.ca (Peter Sprokkelenburg) Date: Thu, 25 Sep 2014 19:32:22 +0000 Subject: Legacy system Message-ID: <593F4F4552EC764FBA345306F57D4C6C47A8194C@exch01.corp.dsl4u.ca> I have a system that we only use internally and it some how got *missed* on getting upgrade to pkg from pkg_ Attempting to upgrade now give an error - /usr/ports/ports-mgmt/pkg]# make You are about to convert your system to pkg while you have ports/packages installed with the old pkg_install tools. To switch to pkg: 1) Install ports-mgmt/pkg 2) Convert your package database by running pkg2ng *** [pre-everything] Error code 1 Stop in /usr/ports/ports-mgmt/pkg. Any work arounds to get this installed and running on pkg? Haven't been able to find much... ----- Peter Sprokkelenburg http://www.dsl4u.ca From milu at dat.pl Thu Sep 25 20:27:09 2014 From: milu at dat.pl (Maciej Milewski) Date: Thu, 25 Sep 2014 22:27:00 +0200 Subject: Can't compile pkg on mips In-Reply-To: <5424425A.9040709@FreeBSD.org> References: <54233771.4090803@dat.pl> <057b903c-628f-4843-b7de-c0f7cb9750b4@email.android.com> <5424425A.9040709@FreeBSD.org> Message-ID: <54247A94.6090901@dat.pl> On 25.09.2014 18:27, Bryan Drewery wrote: > On 9/25/2014 11:19 AM, Maciej Milewski wrote: >> >> On 25 wrze?nia 2014 18:01:23 CEST, Thiago Farina wrote: >>> On Wed, Sep 24, 2014 at 6:28 PM, Maciej Milewski wrote: >>>> Hi, >>>> I have problem with compiling pkg-1.3.8 on my RSPro mips board: >>>> ===> Building for pkg-1.3.8 >>>> /usr/bin/make all-recursive >>>> Making all in external >>>> /usr/bin/make all-am >>>> Making all in libpkg >>>> Making all in repo >>>> Making all in binary >>>> Making all in . >>>> Making all in src >>>> CC pkg-info.o >>>> cc1: warnings being treated as errors >>>> info.c: In function 'exec_info': >>>> info.c:266: warning: implicit declaration of function >>> 'cap_rights_init' >>>> info.c:267: warning: passing argument 2 of 'cap_rights_limit' makes >>>> integer from pointer without a cast >>>> *** [pkg-info.o] Error code 1 >>>> >>> Have you tried fixing this warning? As it is treating warnings as >>> errors the compilation fails. >>> >>> Hope that helps, >>> >>> Regards, >> Yes tried removing -Werror from src/Makefile. If that's sth connected with capsicum then my system might be built before that got to the tree. I'll try of I can build and flash newer image for that device. >> > Huh that's strange. The header is included properly. > > What FreeBSD version is this? > FreeBSD RSPRO 10.0-CURRENT FreeBSD 10.0-CURRENT #9 r253582+713ffac after manual edit src/Makefile and removing -Werror compilation went few lines further: ===> Building for pkg-1.3.8 /usr/bin/make all-recursive Making all in external /usr/bin/make all-am Making all in libpkg Making all in repo Making all in binary Making all in . Making all in src CCLD pkg pkg-info.o: In function `exec_info': info.c:(.text+0x5ac): undefined reference to `cap_rights_init' pkg-ssh.o: In function `exec_ssh': ssh.c:(.text+0x160): undefined reference to `cap_rights_init' pkg-updating.o: In function `exec_updating': updating.c:(.text+0x334): undefined reference to `cap_rights_init' *** [pkg] Error code 1 -- Pozdrawiam, Maciej Milewski From rodrigc at freebsd.org Thu Sep 25 20:27:16 2014 From: rodrigc at freebsd.org (Craig Rodrigues) Date: Thu, 25 Sep 2014 13:27:14 -0700 Subject: Legacy system In-Reply-To: <593F4F4552EC764FBA345306F57D4C6C47A8194C@exch01.corp.dsl4u.ca> References: <593F4F4552EC764FBA345306F57D4C6C47A8194C@exch01.corp.dsl4u.ca> Message-ID: On Thu, Sep 25, 2014 at 12:32 PM, Peter Sprokkelenburg < peters at connections.ca> wrote: > I have a system that we only use internally and it some how got *missed* > on getting upgrade to pkg from pkg_ > > Attempting to upgrade now give an error - > > /usr/ports/ports-mgmt/pkg]# make > You are about to convert your system to pkg while you have ports/packages > installed with the old pkg_install tools. > > To switch to pkg: > 1) Install ports-mgmt/pkg > 2) Convert your package database by running pkg2ng > > *** [pre-everything] Error code 1 > > Stop in /usr/ports/ports-mgmt/pkg. > > Hi, Can you follow the steps (in order) here: https://www.freebsd.org/doc/handbook/pkgng-intro.html#pkgng-initial-setup and report anything that is unclear in that doc, so we can clean it up? -- Craig From tfransosi at gmail.com Thu Sep 25 20:33:53 2014 From: tfransosi at gmail.com (Thiago Farina) Date: Thu, 25 Sep 2014 17:33:52 -0300 Subject: Can't compile pkg on mips In-Reply-To: <54247A94.6090901@dat.pl> References: <54233771.4090803@dat.pl> <057b903c-628f-4843-b7de-c0f7cb9750b4@email.android.com> <5424425A.9040709@FreeBSD.org> <54247A94.6090901@dat.pl> Message-ID: On Thu, Sep 25, 2014 at 5:27 PM, Maciej Milewski wrote: >> What FreeBSD version is this? >> > FreeBSD RSPRO 10.0-CURRENT FreeBSD 10.0-CURRENT #9 r253582+713ffac > > after manual edit src/Makefile and removing -Werror compilation went few > lines further: > ===> Building for pkg-1.3.8 > /usr/bin/make all-recursive > Making all in external > /usr/bin/make all-am > Making all in libpkg > Making all in repo > Making all in binary > Making all in . > Making all in src > CCLD pkg > pkg-info.o: In function `exec_info': > info.c:(.text+0x5ac): undefined reference to `cap_rights_init' > pkg-ssh.o: In function `exec_ssh': > ssh.c:(.text+0x160): undefined reference to `cap_rights_init' > pkg-updating.o: In function `exec_updating': > updating.c:(.text+0x334): undefined reference to `cap_rights_init' > *** [pkg] Error code 1 > Can you do `grep -Rn cap_rights_limit . `, to figure out where "cap_rights_limit" is? It is missing either a ".a" or ".o" where this symbol is, hence it isn't linking. Not your fault though. -- Thiago Farina From milu at dat.pl Thu Sep 25 21:27:27 2014 From: milu at dat.pl (Maciej Milewski) Date: Thu, 25 Sep 2014 23:27:21 +0200 Subject: Can't compile pkg on mips In-Reply-To: References: <54233771.4090803@dat.pl> <057b903c-628f-4843-b7de-c0f7cb9750b4@email.android.com> <5424425A.9040709@FreeBSD.org> <54247A94.6090901@dat.pl> Message-ID: <542488B9.6090700@dat.pl> On 25.09.2014 22:33, Thiago Farina wrote: > On Thu, Sep 25, 2014 at 5:27 PM, Maciej Milewski wrote: >>> What FreeBSD version is this? >>> >> FreeBSD RSPRO 10.0-CURRENT FreeBSD 10.0-CURRENT #9 r253582+713ffac >> >> after manual edit src/Makefile and removing -Werror compilation went few >> lines further: >> ===> Building for pkg-1.3.8 >> /usr/bin/make all-recursive >> Making all in external >> /usr/bin/make all-am >> Making all in libpkg >> Making all in repo >> Making all in binary >> Making all in . >> Making all in src >> CCLD pkg >> pkg-info.o: In function `exec_info': >> info.c:(.text+0x5ac): undefined reference to `cap_rights_init' >> pkg-ssh.o: In function `exec_ssh': >> ssh.c:(.text+0x160): undefined reference to `cap_rights_init' >> pkg-updating.o: In function `exec_updating': >> updating.c:(.text+0x334): undefined reference to `cap_rights_init' >> *** [pkg] Error code 1 >> > Can you do `grep -Rn cap_rights_limit . `, to figure out where > "cap_rights_limit" is? > > It is missing either a ".a" or ".o" where this symbol is, hence it > isn't linking. > > Not your fault though. Sure I can: /data/builds/usr/ports/ports-mgmt/pkg/work/pkg-1.3.8# grep -Rn cap_rights_limit . ./src/ssh.c:78: if (cap_rights_limit(fd, &rights) < 0 && errno != ENOSYS ) { ./src/ssh.c:79: warn("cap_rights_limit() failed"); ./src/info.c:267: if (cap_rights_limit(fd, &rights) < 0 && errno != ENOSYS ) { ./src/info.c:268: warn("cap_rights_limit() failed"); ./src/updating.c:141: if (cap_rights_limit(fileno(fd), &rights) < 0 && errno != ENOSYS ) { ./src/updating.c:142: warn("cap_rights_limit() failed"); Binary file ./src/pkg-info.o matches Binary file ./src/pkg-ssh.o matches Binary file ./src/pkg-updating.o matches -- Pozdrawiam, Maciej Milewski From peters at connections.ca Thu Sep 25 22:20:24 2014 From: peters at connections.ca (Peter Sprokkelenburg) Date: Thu, 25 Sep 2014 22:20:20 +0000 Subject: Legacy system In-Reply-To: References: <593F4F4552EC764FBA345306F57D4C6C47A8194C@exch01.corp.dsl4u.ca> Message-ID: <593F4F4552EC764FBA345306F57D4C6C47A81B71@exch01.corp.dsl4u.ca> Craig, Thank you for the link. I had some pkgdb corruption that I had to clean up first, that is all sorted and I was able to install pkg following the instructions. The /usr/sbin/pkg REALLY helped. Thanks again! ----- Peter Sprokkelenburg http://www.dsl4u.ca From: crodr001 at gmail.com [mailto:crodr001 at gmail.com] On Behalf Of Craig Rodrigues Sent: Thursday, September 25, 2014 4:27 PM To: Peter Sprokkelenburg Cc: pkg at FreeBSD.org Subject: Re: Legacy system On Thu, Sep 25, 2014 at 12:32 PM, Peter Sprokkelenburg > wrote: I have a system that we only use internally and it some how got *missed* on getting upgrade to pkg from pkg_ Attempting to upgrade now give an error - /usr/ports/ports-mgmt/pkg]# make You are about to convert your system to pkg while you have ports/packages installed with the old pkg_install tools. To switch to pkg: 1) Install ports-mgmt/pkg 2) Convert your package database by running pkg2ng *** [pre-everything] Error code 1 Stop in /usr/ports/ports-mgmt/pkg. Hi, Can you follow the steps (in order) here: https://www.freebsd.org/doc/handbook/pkgng-intro.html#pkgng-initial-setup and report anything that is unclear in that doc, so we can clean it up? -- Craig From bdrewery at FreeBSD.org Thu Sep 25 22:59:16 2014 From: bdrewery at FreeBSD.org (Bryan Drewery) Date: Thu, 25 Sep 2014 17:59:04 -0500 Subject: Can't compile pkg on mips In-Reply-To: <542488B9.6090700@dat.pl> References: <54233771.4090803@dat.pl> <057b903c-628f-4843-b7de-c0f7cb9750b4@email.android.com> <5424425A.9040709@FreeBSD.org> <54247A94.6090901@dat.pl> <542488B9.6090700@dat.pl> Message-ID: <54249E38.5000209@FreeBSD.org> On 9/25/2014 4:27 PM, Maciej Milewski wrote: > On 25.09.2014 22:33, Thiago Farina wrote: >> On Thu, Sep 25, 2014 at 5:27 PM, Maciej Milewski wrote: >>>> What FreeBSD version is this? >>>> >>> FreeBSD RSPRO 10.0-CURRENT FreeBSD 10.0-CURRENT #9 r253582+713ffac >>> >>> after manual edit src/Makefile and removing -Werror compilation went few >>> lines further: >>> ===> Building for pkg-1.3.8 >>> /usr/bin/make all-recursive >>> Making all in external >>> /usr/bin/make all-am >>> Making all in libpkg >>> Making all in repo >>> Making all in binary >>> Making all in . >>> Making all in src >>> CCLD pkg >>> pkg-info.o: In function `exec_info': >>> info.c:(.text+0x5ac): undefined reference to `cap_rights_init' >>> pkg-ssh.o: In function `exec_ssh': >>> ssh.c:(.text+0x160): undefined reference to `cap_rights_init' >>> pkg-updating.o: In function `exec_updating': >>> updating.c:(.text+0x334): undefined reference to `cap_rights_init' >>> *** [pkg] Error code 1 >>> Can you post your config.log please? It generates during the build of pkg in the work dir. -- Regards, Bryan Drewery -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 488 bytes Desc: OpenPGP digital signature URL: From milu at dat.pl Thu Sep 25 23:07:49 2014 From: milu at dat.pl (Maciej Milewski) Date: Fri, 26 Sep 2014 01:07:39 +0200 Subject: Can't compile pkg on mips In-Reply-To: <54249E38.5000209@FreeBSD.org> References: <54233771.4090803@dat.pl> <057b903c-628f-4843-b7de-c0f7cb9750b4@email.android.com> <5424425A.9040709@FreeBSD.org> <54247A94.6090901@dat.pl> <542488B9.6090700@dat.pl> <54249E38.5000209@FreeBSD.org> Message-ID: <5424A03B.8080207@dat.pl> On 26.09.2014 00:59, Bryan Drewery wrote: > On 9/25/2014 4:27 PM, Maciej Milewski wrote: >> On 25.09.2014 22:33, Thiago Farina wrote: >>> On Thu, Sep 25, 2014 at 5:27 PM, Maciej Milewski wrote: >>>>> What FreeBSD version is this? >>>>> >>>> FreeBSD RSPRO 10.0-CURRENT FreeBSD 10.0-CURRENT #9 r253582+713ffac >>>> >>>> after manual edit src/Makefile and removing -Werror compilation went few >>>> lines further: >>>> ===> Building for pkg-1.3.8 >>>> /usr/bin/make all-recursive >>>> Making all in external >>>> /usr/bin/make all-am >>>> Making all in libpkg >>>> Making all in repo >>>> Making all in binary >>>> Making all in . >>>> Making all in src >>>> CCLD pkg >>>> pkg-info.o: In function `exec_info': >>>> info.c:(.text+0x5ac): undefined reference to `cap_rights_init' >>>> pkg-ssh.o: In function `exec_ssh': >>>> ssh.c:(.text+0x160): undefined reference to `cap_rights_init' >>>> pkg-updating.o: In function `exec_updating': >>>> updating.c:(.text+0x334): undefined reference to `cap_rights_init' >>>> *** [pkg] Error code 1 >>>> > Can you post your config.log please? It generates during the build of > pkg in the work dir. > > Attached, might be cut by list rules. -- Pozdrawiam, Maciej Milewski From tskinner at deepwebtech.com Fri Sep 26 01:04:51 2014 From: tskinner at deepwebtech.com (Tim Skinner) Date: Thu, 25 Sep 2014 19:04:49 -0600 Subject: Cant install ports-mgmt/pkg Message-ID: Hello, I've run into a problem trying to install ports-mgmt/pkg. When trying to run make or portmaster I get the following error: You are about to convert your system to pkg while you have ports/packages installed with the old pkg_install tools. To switch to pkg: 1) Install ports-mgmt/pkg 2) Convert your package database by running pkg2ng *** [pre-everything] Error code 1 So it is telling me I cant install pkg without first installing pkg, it would be funny if it wasn't totally blocking my work :) I have two machines in this state that I must have forgotten to update to pkg. I have tried putting WITHOUT_PKGNG=yes in /etc/make.conf but that trick apparently does not work anymore. This system is: FreeBSD proxy2 9.3-RELEASE FreeBSD 9.3-RELEASE #2 r269738: Sun Aug 10 12:01:22 MDT 2014 Do you know of a work around to get pkg installed? Do I need to downgrade to an earlier kernel? Thanks, -Tim Skinner From rodrigc at freebsd.org Fri Sep 26 02:01:14 2014 From: rodrigc at freebsd.org (Craig Rodrigues) Date: Thu, 25 Sep 2014 19:01:12 -0700 Subject: Cant install ports-mgmt/pkg In-Reply-To: References: Message-ID: On Thu, Sep 25, 2014 at 6:04 PM, Tim Skinner wrote: > > You are about to convert your system to pkg while you have ports/packages > installed with the old pkg_install tools. > > To switch to pkg: > 1) Install ports-mgmt/pkg > 2) Convert your package database by running pkg2ng > > *** [pre-everything] Error code 1 > Can you try this: https://lists.freebsd.org/pipermail/freebsd-pkg/2014-September/000687.html -- Craig From tfransosi at gmail.com Fri Sep 26 03:47:08 2014 From: tfransosi at gmail.com (Thiago Farina) Date: Fri, 26 Sep 2014 00:47:07 -0300 Subject: Can't compile pkg on mips In-Reply-To: <542488B9.6090700@dat.pl> References: <54233771.4090803@dat.pl> <057b903c-628f-4843-b7de-c0f7cb9750b4@email.android.com> <5424425A.9040709@FreeBSD.org> <54247A94.6090901@dat.pl> <542488B9.6090700@dat.pl> Message-ID: On Thu, Sep 25, 2014 at 6:27 PM, Maciej Milewski wrote: > On 25.09.2014 22:33, Thiago Farina wrote: >> On Thu, Sep 25, 2014 at 5:27 PM, Maciej Milewski wrote: >>>> What FreeBSD version is this? >>>> >>> FreeBSD RSPRO 10.0-CURRENT FreeBSD 10.0-CURRENT #9 r253582+713ffac >>> >>> after manual edit src/Makefile and removing -Werror compilation went few >>> lines further: >>> ===> Building for pkg-1.3.8 >>> /usr/bin/make all-recursive >>> Making all in external >>> /usr/bin/make all-am >>> Making all in libpkg >>> Making all in repo >>> Making all in binary >>> Making all in . >>> Making all in src >>> CCLD pkg >>> pkg-info.o: In function `exec_info': >>> info.c:(.text+0x5ac): undefined reference to `cap_rights_init' >>> pkg-ssh.o: In function `exec_ssh': >>> ssh.c:(.text+0x160): undefined reference to `cap_rights_init' >>> pkg-updating.o: In function `exec_updating': >>> updating.c:(.text+0x334): undefined reference to `cap_rights_init' >>> *** [pkg] Error code 1 >>> >> Can you do `grep -Rn cap_rights_limit . `, to figure out where >> "cap_rights_limit" is? >> >> It is missing either a ".a" or ".o" where this symbol is, hence it >> isn't linking. >> >> Not your fault though. > Sure I can: > /data/builds/usr/ports/ports-mgmt/pkg/work/pkg-1.3.8# grep -Rn > cap_rights_limit . > ./src/ssh.c:78: if (cap_rights_limit(fd, &rights) < 0 && errno != ENOSYS ) { > ./src/ssh.c:79: warn("cap_rights_limit() failed"); > ./src/info.c:267: if (cap_rights_limit(fd, &rights) < 0 && > errno != ENOSYS ) { > ./src/info.c:268: warn("cap_rights_limit() failed"); > ./src/updating.c:141: if (cap_rights_limit(fileno(fd), &rights) < 0 && > errno != ENOSYS ) { > ./src/updating.c:142: warn("cap_rights_limit() failed"); > Binary file ./src/pkg-info.o matches > Binary file ./src/pkg-ssh.o matches > Binary file ./src/pkg-updating.o matches > Do you have HAVE_CAPSICUM defined? It needs to be defined in order to sys/capability.h to get included. But I think that does not explain the link issue. It is weird that it is not finding something from libc. -- Thiago Farina From milu at dat.pl Fri Sep 26 08:19:00 2014 From: milu at dat.pl (Maciej Milewski) Date: Fri, 26 Sep 2014 10:18:53 +0200 Subject: Can't compile pkg on mips In-Reply-To: References: <54233771.4090803@dat.pl> <057b903c-628f-4843-b7de-c0f7cb9750b4@email.android.com> <5424425A.9040709@FreeBSD.org> <54247A94.6090901@dat.pl> <542488B9.6090700@dat.pl> Message-ID: <5425216D.9090405@dat.pl> On 26.09.2014 05:47, Thiago Farina wrote: > On Thu, Sep 25, 2014 at 6:27 PM, Maciej Milewski wrote: >> On 25.09.2014 22:33, Thiago Farina wrote: >>> On Thu, Sep 25, 2014 at 5:27 PM, Maciej Milewski wrote: >>>>> What FreeBSD version is this? >>>>> >>>> FreeBSD RSPRO 10.0-CURRENT FreeBSD 10.0-CURRENT #9 r253582+713ffac >>>> >>>> after manual edit src/Makefile and removing -Werror compilation went few >>>> lines further: >>>> ===> Building for pkg-1.3.8 >>>> /usr/bin/make all-recursive >>>> Making all in external >>>> /usr/bin/make all-am >>>> Making all in libpkg >>>> Making all in repo >>>> Making all in binary >>>> Making all in . >>>> Making all in src >>>> CCLD pkg >>>> pkg-info.o: In function `exec_info': >>>> info.c:(.text+0x5ac): undefined reference to `cap_rights_init' >>>> pkg-ssh.o: In function `exec_ssh': >>>> ssh.c:(.text+0x160): undefined reference to `cap_rights_init' >>>> pkg-updating.o: In function `exec_updating': >>>> updating.c:(.text+0x334): undefined reference to `cap_rights_init' >>>> *** [pkg] Error code 1 >>>> >>> Can you do `grep -Rn cap_rights_limit . `, to figure out where >>> "cap_rights_limit" is? >>> >>> It is missing either a ".a" or ".o" where this symbol is, hence it >>> isn't linking. >>> >>> Not your fault though. >> Sure I can: >> /data/builds/usr/ports/ports-mgmt/pkg/work/pkg-1.3.8# grep -Rn >> cap_rights_limit . >> ./src/ssh.c:78: if (cap_rights_limit(fd, &rights) < 0 && errno != ENOSYS ) { >> ./src/ssh.c:79: warn("cap_rights_limit() failed"); >> ./src/info.c:267: if (cap_rights_limit(fd, &rights) < 0 && >> errno != ENOSYS ) { >> ./src/info.c:268: warn("cap_rights_limit() failed"); >> ./src/updating.c:141: if (cap_rights_limit(fileno(fd), &rights) < 0 && >> errno != ENOSYS ) { >> ./src/updating.c:142: warn("cap_rights_limit() failed"); >> Binary file ./src/pkg-info.o matches >> Binary file ./src/pkg-ssh.o matches >> Binary file ./src/pkg-updating.o matches >> > Do you have HAVE_CAPSICUM defined? It needs to be defined in order to > sys/capability.h to get included. > > But I think that does not explain the link issue. It is weird that it > is not finding something from libc. > Grep shows that it's defined: Is it autodectected somehow? # grep -nr HAVE_CAPSICUM * config.log:3815:#define HAVE_CAPSICUM 1 config.status:1086:D["HAVE_CAPSICUM"]=" 1" configure:13295:$as_echo "#define HAVE_CAPSICUM 1" >>confdefs.h configure.ac:224: AC_DEFINE(HAVE_CAPSICUM, 1, [Define 1 if you have 'capsicum'.]) configure.bak:13295:$as_echo "#define HAVE_CAPSICUM 1" >>confdefs.h libpkg/ssh.c:31:#ifdef HAVE_CAPSICUM libpkg/ssh.c:134:#ifdef HAVE_CAPSICUM pkg_config.h:17:#define HAVE_CAPSICUM 1 pkg_config.h.in:16:#undef HAVE_CAPSICUM src/ssh.c:31:#ifdef HAVE_CAPSICUM src/ssh.c:58:#ifdef HAVE_CAPSICUM src/ssh.c:76:#ifdef HAVE_CAPSICUM src/audit.c:259:#ifdef HAVE_CAPSICUM src/info.c:35:#ifdef HAVE_CAPSICUM src/info.c:99:#ifdef HAVE_CAPSICUM src/info.c:265:#ifdef HAVE_CAPSICUM src/event.c:40:#ifdef HAVE_CAPSICUM src/event.c:215:#ifdef HAVE_CAPSICUM src/event.c:319:#ifdef HAVE_CAPSICUM src/updating.c:32:#ifdef HAVE_CAPSICUM src/updating.c:84:#ifdef HAVE_CAPSICUM src/updating.c:139:#ifdef HAVE_CAPSICUM -- Pozdrawiam, Maciej Milewski From bdrewery at FreeBSD.org Fri Sep 26 14:35:48 2014 From: bdrewery at FreeBSD.org (Bryan Drewery) Date: Fri, 26 Sep 2014 09:35:27 -0500 Subject: Can't compile pkg on mips In-Reply-To: <5425216D.9090405@dat.pl> References: <54233771.4090803@dat.pl> <057b903c-628f-4843-b7de-c0f7cb9750b4@email.android.com> <5424425A.9040709@FreeBSD.org> <54247A94.6090901@dat.pl> <542488B9.6090700@dat.pl> <5425216D.9090405@dat.pl> Message-ID: <542579AF.8030602@FreeBSD.org> On 9/26/2014 3:18 AM, Maciej Milewski wrote: > On 26.09.2014 05:47, Thiago Farina wrote: >> On Thu, Sep 25, 2014 at 6:27 PM, Maciej Milewski wrote: >>> On 25.09.2014 22:33, Thiago Farina wrote: >>>> On Thu, Sep 25, 2014 at 5:27 PM, Maciej Milewski wrote: >>>>>> What FreeBSD version is this? >>>>>> >>>>> FreeBSD RSPRO 10.0-CURRENT FreeBSD 10.0-CURRENT #9 r253582+713ffac >>>>> >>>>> after manual edit src/Makefile and removing -Werror compilation went few >>>>> lines further: >>>>> ===> Building for pkg-1.3.8 >>>>> /usr/bin/make all-recursive >>>>> Making all in external >>>>> /usr/bin/make all-am >>>>> Making all in libpkg >>>>> Making all in repo >>>>> Making all in binary >>>>> Making all in . >>>>> Making all in src >>>>> CCLD pkg >>>>> pkg-info.o: In function `exec_info': >>>>> info.c:(.text+0x5ac): undefined reference to `cap_rights_init' >>>>> pkg-ssh.o: In function `exec_ssh': >>>>> ssh.c:(.text+0x160): undefined reference to `cap_rights_init' >>>>> pkg-updating.o: In function `exec_updating': >>>>> updating.c:(.text+0x334): undefined reference to `cap_rights_init' >>>>> *** [pkg] Error code 1 >>>>> >>>> Can you do `grep -Rn cap_rights_limit . `, to figure out where >>>> "cap_rights_limit" is? >>>> >>>> It is missing either a ".a" or ".o" where this symbol is, hence it >>>> isn't linking. >>>> >>>> Not your fault though. >>> Sure I can: >>> /data/builds/usr/ports/ports-mgmt/pkg/work/pkg-1.3.8# grep -Rn >>> cap_rights_limit . >>> ./src/ssh.c:78: if (cap_rights_limit(fd, &rights) < 0 && errno != ENOSYS ) { >>> ./src/ssh.c:79: warn("cap_rights_limit() failed"); >>> ./src/info.c:267: if (cap_rights_limit(fd, &rights) < 0 && >>> errno != ENOSYS ) { >>> ./src/info.c:268: warn("cap_rights_limit() failed"); >>> ./src/updating.c:141: if (cap_rights_limit(fileno(fd), &rights) < 0 && >>> errno != ENOSYS ) { >>> ./src/updating.c:142: warn("cap_rights_limit() failed"); >>> Binary file ./src/pkg-info.o matches >>> Binary file ./src/pkg-ssh.o matches >>> Binary file ./src/pkg-updating.o matches >>> >> Do you have HAVE_CAPSICUM defined? It needs to be defined in order to >> sys/capability.h to get included. >> >> But I think that does not explain the link issue. It is weird that it >> is not finding something from libc. >> > Grep shows that it's defined: Is it autodectected somehow? > > # grep -nr HAVE_CAPSICUM > * > > config.log:3815:#define HAVE_CAPSICUM 1 > config.status:1086:D["HAVE_CAPSICUM"]=" 1" > configure:13295:$as_echo "#define HAVE_CAPSICUM 1" >>confdefs.h > configure.ac:224: AC_DEFINE(HAVE_CAPSICUM, 1, [Define 1 > if you have 'capsicum'.]) > configure.bak:13295:$as_echo "#define HAVE_CAPSICUM 1" >>confdefs.h > libpkg/ssh.c:31:#ifdef HAVE_CAPSICUM > libpkg/ssh.c:134:#ifdef HAVE_CAPSICUM > pkg_config.h:17:#define HAVE_CAPSICUM 1 > pkg_config.h.in:16:#undef HAVE_CAPSICUM > src/ssh.c:31:#ifdef HAVE_CAPSICUM > src/ssh.c:58:#ifdef HAVE_CAPSICUM > src/ssh.c:76:#ifdef HAVE_CAPSICUM > src/audit.c:259:#ifdef HAVE_CAPSICUM > src/info.c:35:#ifdef HAVE_CAPSICUM > src/info.c:99:#ifdef HAVE_CAPSICUM > src/info.c:265:#ifdef HAVE_CAPSICUM > src/event.c:40:#ifdef HAVE_CAPSICUM > src/event.c:215:#ifdef HAVE_CAPSICUM > src/event.c:319:#ifdef HAVE_CAPSICUM > src/updating.c:32:#ifdef HAVE_CAPSICUM > src/updating.c:84:#ifdef HAVE_CAPSICUM > src/updating.c:139:#ifdef HAVE_CAPSICUM > > IMHO your world is out of sync. Like you have newer /usr/include than /lib. -- Regards, Bryan Drewery -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 488 bytes Desc: OpenPGP digital signature URL: From marcus at marcuscom.com Fri Sep 26 15:22:17 2014 From: marcus at marcuscom.com (Joe Marcus Clarke) Date: Fri, 26 Sep 2014 11:22:10 -0400 Subject: Why clean /usr/local? Message-ID: <542584A2.7030206@marcuscom.com> It was pointed out on the Tinderbox list that pkg now cleans up directories like /usr/local/bin (i.e., removes them) if all ports are uninstalled. This, of course, ticks off TB. My question is, why remove directories owned by BSD.local.dist? This is a behavior change, and before I go changing TB code, I'd like to get some perspective. Thanks. Joe -- PGP Key : http://www.marcuscom.com/pgp.asc From davor at davor.se Sat Sep 27 16:48:18 2014 From: davor at davor.se (Davor Babic) Date: Sat, 27 Sep 2014 18:48:16 +0200 Subject: Broken repository Message-ID: <5426EA50.1010900@davor.se> Hello, I'm not sure who I'm supposed to contact about this, so I'm sending this message here. It's currently not possible to install the 'ack' package with pkg. This is the output I get when I try to do pkg install ack: pkg: http://pkg.FreeBSD.org/freebsd:10:x86:64/latest/All/ack-2.12.txz: Not Found -- Davor Babic davor at davor.se From bugzilla-noreply at freebsd.org Sun Sep 28 06:32:45 2014 From: bugzilla-noreply at freebsd.org (bugzilla-noreply at freebsd.org) Date: Sun, 28 Sep 2014 06:32:45 +0000 Subject: [Bug 191352] ports-mgmt/pkg v 1.2.7_3 does not honor the HTTP_PROXY_AUTH variable. In-Reply-To: References: Message-ID: https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=191352 --- Comment #6 from vas at mpeks.tomsk.su --- I think this could be related to https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=186398 -- You are receiving this mail because: You are the assignee for the bug. From marquis at roble.com Sun Sep 28 20:14:05 2014 From: marquis at roble.com (Roger Marquis) Date: Sun, 28 Sep 2014 13:07:07 -0700 (PDT) Subject: Why clean /usr/local? In-Reply-To: References: Message-ID: Joe Marcus Clarke wrote: > It was pointed out on the Tinderbox list that pkg now cleans up > directories like /usr/local/bin (i.e., removes them) if all ports are > uninstalled. This, of course, ticks off TB. That can't be correct behavior. /usr/local/bin should NEVER be cleaned up. A number of directories under /usr/local are shared with administrative scripts and related files. These are not part of the pkg system nor should they be. If pkg is actually deleting files or directories not owned by any package it is a bug, a bug severe enough to warrant review of the policies, processes and/or individuals approving checked-in code. Roger From paulbeard at gmail.com Sun Sep 28 20:42:16 2014 From: paulbeard at gmail.com (paul beard) Date: Sun, 28 Sep 2014 13:42:15 -0700 Subject: pkg config options: is this a thing? Message-ID: I have been using pkg pretty exclusively since I have a hardware failure and need to rebuild an existing FreeBSD system. It's now running in VirtualBox on an iMac(!) and works just fine. But I have been running into some issues with pkgs that have config options or variables at build time that I can't or don't know how to manage. My most recent example was installing cups and it's various components and finding that the ssl libs to manage the web UI were not installed or linked properly. I ended up building it all from source to make it work properly. I was having bad flashbacks of my days with another more well-known open source OS with it's myriad incompatible distros and packages, which was one of the reasons I tried and have stuck with FreeBSD since release 4.11. If there is something I am missing please advise. If it were up to me, I would like to see a config screen pop up, just as I see with 'make config' and the choices made there inform the pkg I request and install. Ideally, you don't want more libraries or binaries than you truly need for security reasons, to say nothing of sanity. And of course, you also want stuff to work properly. -- Paul Beard / www.paulbeard.org/ From m.seaman at infracaninophile.co.uk Sun Sep 28 21:09:26 2014 From: m.seaman at infracaninophile.co.uk (Matthew Seaman) Date: Mon, 29 Sep 2014 00:09:12 +0300 Subject: pkg config options: is this a thing? In-Reply-To: References: Message-ID: <542878F8.4040406@infracaninophile.co.uk> On 28/09/2014 23:42, paul beard wrote: > I have been using pkg pretty exclusively since I have a hardware > failure and need to rebuild an existing FreeBSD system. It's now > running in VirtualBox on an iMac(!) and works just fine. > > But I have been running into some issues with pkgs that have config > options or variables at build time that I can't or don't know how to > manage. My most recent example was installing cups and it's various > components and finding that the ssl libs to manage the web UI were not > installed or linked properly. I ended up building it all from source > to make it work properly. I was having bad flashbacks of my days with > another more well-known open source OS with it's myriad incompatible > distros and packages, which was one of the reasons I tried and have > stuck with FreeBSD since release 4.11. > > If there is something I am missing please advise. If it were up to me, > I would like to see a config screen pop up, just as I see with 'make > config' and the choices made there inform the pkg I request and > install. Ideally, you don't want more libraries or binaries than you > truly need for security reasons, to say nothing of sanity. And of > course, you also want stuff to work properly. Yes, this is a problem with using binary pkgs at the moment. pkg(8) is very much a work in progress still. Plans are afoot to introduce measures -- sub-packages and flavours -- that should mean you don't get stuck with packages that are using the wrong options or that have undesirable dependencies quite so often. However right now, the advice is to compile your own packages if what is available on the repos isn't suitable. You can just build stuff out of ports, but certainly if you have several FreeBSD machines to manage, and maybe even if you only have just one, consider trying out poudriere to generate your own package repository. It's easier than you think and once you've a poudriere build system set up, it's almost as convenient as just installing pkgs from the official pkg repositories. Cheers, Matthew -- Dr Matthew J Seaman MA, D.Phil. PGP: http://www.infracaninophile.co.uk/pgpkey JID: matthew at infracaninophile.co.uk -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 971 bytes Desc: OpenPGP digital signature URL: From bapt at FreeBSD.org Mon Sep 29 08:01:15 2014 From: bapt at FreeBSD.org (Baptiste Daroussin) Date: Mon, 29 Sep 2014 10:01:10 +0200 Subject: incorrect dependency registration? In-Reply-To: <15E556C8-EEB6-4218-93F3-F76BFA2E4F29@FreeBSD.org> References: <15E556C8-EEB6-4218-93F3-F76BFA2E4F29@FreeBSD.org> Message-ID: <20140929080109.GI40373@ivaldir.etoilebsd.net> On Mon, Sep 29, 2014 at 11:52:50AM +0400, Dmitry Sivachenko wrote: > Hello, > > I am using FreeBSD-10/stable and pkg-1.3.8. > > Consider py-dateutil port, it depends on py-six. > > First, I installed py33-dateutil (with py33-six), for python3. > > Now I install py-dateutil for python2. Upon installation, I get the following waring: > ===> Installing for py27-dateutil-2.2 > ===> py27-dateutil-2.2 depends on package: py27-six>0 - found > ===> py27-dateutil-2.2 depends on package: py27-setuptools27>0 - found > ===> py27-dateutil-2.2 depends on file: /usr/local/bin/python2.7 - found > ===> Checking if py27-dateutil already installed > ===> Registering installation for py27-dateutil-2.2 > pkg-static: py27-dateutil-2.2: duplicate dependency listing: py33-six-1.5.2, ignoring > > When I try to deinstall py27-six, I get: > > # pkg delete py27-six > Checking integrity... done (0 conflicting) > Deinstallation has been requested for the following 3 packages (of 0 packages in the universe): > > Installed packages to be REMOVED: > py27-six-1.8.0 > py33-dateutil-2.1_3 (depends on py27-six-1.8.0) > py27-dateutil-2.2 (depends on py27-six-1.8.0) > > The operation will free 1 MB. > > Proceed with deinstalling packages? [y/N]: > > (it wants to deinstall both versions of py-dateutil). > > Something looks broken here. > > Some time ago it did not let to install the same port for different python version, now it is possible but deps registration looks broken. I do not have time for this right now, I'll see in a couple of days, but CCing parties that should be interested :) regards, Bapt -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 181 bytes Desc: not available URL: From bdrewery at FreeBSD.org Mon Sep 29 20:49:06 2014 From: bdrewery at FreeBSD.org (Bryan Drewery) Date: Mon, 29 Sep 2014 15:49:00 -0500 Subject: Problem with pkg Makefile In-Reply-To: References: Message-ID: <5429C5BC.6070302@FreeBSD.org> On 9/19/2014 8:09 PM, Thiago Farina wrote: > On Fri, Sep 19, 2014 at 9:46 PM, John Metro wrote: >> That bit in the Makefile that prints out ... >> You are about to convert your system to pkg while you have ports/packagesinstalled with the old pkg_install tools. >> To switch to pkg: 1) Install ports-mgmt/pkg 2) Convert your package database by running pkg2ng >> *** Error code 1 >> .. should be removed from the makefile for pkg. It REALLY gets in the way if you are trying to upgrade and switch to pkgng for the first time from ports. > > What should be removed? Does it say? > I'm late, but I modified it a few days ago to allow you to proceed. I'm not comfortable removing it outright as the user must run pkg2ng after installing pkg. -- Regards, Bryan Drewery -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 488 bytes Desc: OpenPGP digital signature URL: From bdrewery at FreeBSD.org Tue Sep 30 16:07:21 2014 From: bdrewery at FreeBSD.org (Bryan Drewery) Date: Tue, 30 Sep 2014 11:07:12 -0500 Subject: svn commit: r369639 - head/devel/automake In-Reply-To: <201409301539.s8UFdlok059126@svn.freebsd.org> References: <201409301539.s8UFdlok059126@svn.freebsd.org> Message-ID: <542AD530.7030902@FreeBSD.org> On 9/30/2014 10:39 AM, Mathieu Arnold wrote: > Author: mat > Date: Tue Sep 30 15:39:46 2014 > New Revision: 369639 > URL: http://svnweb.freebsd.org/changeset/ports/369639 > QAT: https://qat.redports.org/buildarchive/r369639/ > > Log: > Bump PORTREVISION to solve this error: > aclocal-1.14: error: couldn't open directory '/usr/local/share/aclocal': No such file or directory > > Which happens if the automake package was built before pkg 1.3.8, because > share/aclocal was part of BSD.local.dist, and was not created by this port. > > Sponsored by: Absolight > This indicates a much larger problem to me that needs to be fixed. We can't easily determine what other ports are affected by this since an exp-run wipes out all packages. The only solution I see is to retain BSD.local.dist, or have pkg install the entire mtree itself. -- Regards, Bryan Drewery -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 488 bytes Desc: OpenPGP digital signature URL: From bdrewery at FreeBSD.org Tue Sep 30 16:31:13 2014 From: bdrewery at FreeBSD.org (Bryan Drewery) Date: Tue, 30 Sep 2014 11:31:04 -0500 Subject: svn commit: r369639 - head/devel/automake In-Reply-To: <542AD530.7030902@FreeBSD.org> References: <201409301539.s8UFdlok059126@svn.freebsd.org> <542AD530.7030902@FreeBSD.org> Message-ID: <542ADAC8.80107@FreeBSD.org> On 9/30/2014 11:07 AM, Bryan Drewery wrote: > On 9/30/2014 10:39 AM, Mathieu Arnold wrote: >> Author: mat >> Date: Tue Sep 30 15:39:46 2014 >> New Revision: 369639 >> URL: http://svnweb.freebsd.org/changeset/ports/369639 >> QAT: https://qat.redports.org/buildarchive/r369639/ >> >> Log: >> Bump PORTREVISION to solve this error: >> aclocal-1.14: error: couldn't open directory '/usr/local/share/aclocal': No such file or directory >> >> Which happens if the automake package was built before pkg 1.3.8, because >> share/aclocal was part of BSD.local.dist, and was not created by this port. >> >> Sponsored by: Absolight >> > > This indicates a much larger problem to me that needs to be fixed. We > can't easily determine what other ports are affected by this since an > exp-run wipes out all packages. The only solution I see is to retain > BSD.local.dist, or have pkg install the entire mtree itself. > > Apparently this was just a mistake in a previous commit to devel/automake. https://reviews.freebsd.org/D870 is to create a pkg mtree port. -- Regards, Bryan Drewery -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 488 bytes Desc: OpenPGP digital signature URL: From tijl at FreeBSD.org Tue Sep 30 16:54:57 2014 From: tijl at FreeBSD.org (Tijl Coosemans) Date: Tue, 30 Sep 2014 18:53:42 +0200 Subject: svn commit: r369639 - head/devel/automake In-Reply-To: <542AD530.7030902@FreeBSD.org> References: <201409301539.s8UFdlok059126@svn.freebsd.org> <542AD530.7030902@FreeBSD.org> Message-ID: <20140930185342.3339ecbd@kalimero.tijl.coosemans.org> On Tue, 30 Sep 2014 11:07:12 -0500 Bryan Drewery wrote: > On 9/30/2014 10:39 AM, Mathieu Arnold wrote: >> Author: mat >> Date: Tue Sep 30 15:39:46 2014 >> New Revision: 369639 >> URL: http://svnweb.freebsd.org/changeset/ports/369639 >> QAT: https://qat.redports.org/buildarchive/r369639/ >> >> Log: >> Bump PORTREVISION to solve this error: >> aclocal-1.14: error: couldn't open directory '/usr/local/share/aclocal': No such file or directory >> >> Which happens if the automake package was built before pkg 1.3.8, because >> share/aclocal was part of BSD.local.dist, and was not created by this port. >> >> Sponsored by: Absolight > > This indicates a much larger problem to me that needs to be fixed. We > can't easily determine what other ports are affected by this since an > exp-run wipes out all packages. The only solution I see is to retain > BSD.local.dist, or have pkg install the entire mtree itself. Only empty directories that need to exist can cause this. I don't think that applies to any directory in BSD.local.dist besides share/aclocal. The update to pkg 1.3.8 also added share/applications to devel/desktop-file-utils so maybe that one as well. It forgot to bump PORTREVISION on that port too, as well as devel/automake14: https://svnweb.freebsd.org/ports?view=revision&revision=368803 From bdrewery at FreeBSD.org Tue Sep 30 20:14:31 2014 From: bdrewery at FreeBSD.org (Bryan Drewery) Date: Tue, 30 Sep 2014 15:14:24 -0500 Subject: svn commit: r369639 - head/devel/automake In-Reply-To: <20140930185342.3339ecbd@kalimero.tijl.coosemans.org> References: <201409301539.s8UFdlok059126@svn.freebsd.org> <542AD530.7030902@FreeBSD.org> <20140930185342.3339ecbd@kalimero.tijl.coosemans.org> Message-ID: <542B0F20.3020705@FreeBSD.org> On 9/30/2014 11:53 AM, Tijl Coosemans wrote: > On Tue, 30 Sep 2014 11:07:12 -0500 Bryan Drewery wrote: >> On 9/30/2014 10:39 AM, Mathieu Arnold wrote: >>> Author: mat >>> Date: Tue Sep 30 15:39:46 2014 >>> New Revision: 369639 >>> URL: http://svnweb.freebsd.org/changeset/ports/369639 >>> QAT: https://qat.redports.org/buildarchive/r369639/ >>> >>> Log: >>> Bump PORTREVISION to solve this error: >>> aclocal-1.14: error: couldn't open directory '/usr/local/share/aclocal': No such file or directory >>> >>> Which happens if the automake package was built before pkg 1.3.8, because >>> share/aclocal was part of BSD.local.dist, and was not created by this port. >>> >>> Sponsored by: Absolight >> >> This indicates a much larger problem to me that needs to be fixed. We >> can't easily determine what other ports are affected by this since an >> exp-run wipes out all packages. The only solution I see is to retain >> BSD.local.dist, or have pkg install the entire mtree itself. > > Only empty directories that need to exist can cause this. I don't > think that applies to any directory in BSD.local.dist besides > share/aclocal. The update to pkg 1.3.8 also added share/applications > to devel/desktop-file-utils so maybe that one as well. It forgot to > bump PORTREVISION on that port too, as well as devel/automake14: > https://svnweb.freebsd.org/ports?view=revision&revision=368803 > Thanks, I've bumped those now too. -- Regards, Bryan Drewery -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 488 bytes Desc: OpenPGP digital signature URL: