[gentoo-user] Re: sys-apps/gentoo-functions
On Wed, 19 Mar 2014 07:25:38 +0530, Nilesh Govindrajan m...@nileshgr.com wrote: What is sys-apps/gentoo-functions (seems to have appeared since 10 March 2014) and how is getting installed (as part of emerge -uNDv @world) when it's hard masked? http://thread.gmane.org/gmane.linux.gentoo.devel/90613 -- eroen signature.asc Description: PGP signature
[gentoo-user] Re: Where is /etc/conf.d/net.example?
On Sun, 16 Mar 2014 22:15:59 +0200, Alan McKinnon alan.mckin...@gmail.com wrote: You have various choices - an orthodox network manager like wicd or nm - a minimal network manager like connman - /etc/init.d/net* scripts supplied by OpenRc - no manager, do it manually Why doesn't anyone ever mention using dhcpcd for managing connections? It and its accompanying openrc init script are installed on almost every gentoo box anyway. For simple setups it should Just Work(TM) out-of-box. -- eroen signature.asc Description: PGP signature
[gentoo-user] Re: MD5SUM
On Fri, 14 Mar 2014 08:49:51 +0800, Guido Budack gla...@yandex.com wrote: md5sum livedvd-amd64-multilib-20121221.iso And now, any suggestion? You don't specifically state what you are after, but I'll mention some stuff anyway. The gentoo livedvds are mostly meant as a showcase of gentoo's capabilities. While they can be used to install gentoo, just like any other linux live media in recent memory, they are not mainly intended for this. For installing gentoo, the minimal install livecd is recommended for simple scenarios as noted in the handbook. For more complex setups, notably installing without a wired network connection, I and many others recommend using systemrescuecd[1]. If not using the gentoo minimal install cd, please refer to [2] for some minor deviations from the handbook that are often necessary. 1: http://www.sysresccd.org/Download 2: https://wiki.gentoo.org/wiki/Installation_alternatives#Installation_from_non-Gentoo_LiveCDs -- eroen signature.asc Description: PGP signature
[gentoo-user] Re: Installing gentoo - First emerge
USE=-cups bindist fontconfig X a52 aac acpi avx avx2 asf x264 bash-completion -bluetooth branding bzip2 java -cairo crypt curl doc -emacs exif ffmpeg ftp gimp gif gpm gzip hddtemp -ipv6 -kde -gnome -gnome-keyring -emacs -gtk -qt3 -gt4 -qt5 fat ntfs opengl truetype xorg jack jpeg jpeg2k matroska nmap mmx mp3 mp4 mpeg xv xcomposite multilib png pdf spell ssh posix pulseaudio rss sse sse2 sse3 sse4_1 ssl systemd svg threads tiff -xemacs -zsh-completion You will want to reconsider your enabling of bindist and doc. bindist is enabled by default in stage3 tarballs due to a bug [1] in catalyst. It is known to cause numerous issues, and is not useful for end-user systems. doc generally enables generation of documentation end-users will not require, like developer docs and complete project websites for offline reference. If you find yourself requiring documentation for a specific package, you are better off enabling the doc useflag for that specific package in /etc/portage/package.use . Doc generation pulls in many otherwise unneccesary dependencies in order to generate the documentation, and is known to cause many circular dependencies due to eg. documentation generating software needing itself to generate its own documentation. A general piece of advice: You might not want to set a long string of USE flags globally immediately in your first install, rather pick a profile and adjust flags when/if the defaults don't suit you. Always use '-a' with emerge to review the use flags on packages to be installed. 1: https://bugs.gentoo.org/473332 -- eroen signature.asc Description: PGP signature
[gentoo-user] Re: [OT] LiveCDs for Uni students
On Sat, 08 Mar 2014 08:54:38 +0800, Andrew Lowe a...@wht.com.au wrote: Hi all, I'm doing some research on the topic of LiveCD's and was after any input the list may have. I'm a tutor at a Uni in Australia teaching, amongst others, 1st year Engineering students. We teach them C. Last year we had a lab set up and as well, they could ssh into a Linux box from home so that they could do assignments. Now due to a bureaucratic change ssh access is now gone. I would guess that there would be under 1% Linux penetration with respect to home computers and I've tried, in the past, to help students set up a dev environment on Macs - a horrid experience, and lets not even mention trying to easily set up Win* with a dev environemt I'm looking for a lightweight LiveCD that includes a graphical environment and gcc/clang so that we can make it available on our internal network for the students to download/burn and use at home. Does anyone have any ideas/experience in something like this? I've looked at Lubuntu but it lacks gcc. Any thoughts are greatly appreciated. Andrew You know, if you can be a bit flexible in the 'lightweight' department, Gentoo has some really splendid full-featured live dvds. See the release announcement[1] for the last official release (list of included packages at [2]). More recent pre-release versions are available at [3]. The livedvd team has an irc channel #gentoo-ten on Freenode too. If your requirements are more specific, you could leverage catalyst[4], Gentoo's script for generating release media. It should be possible to add or remove packages from the spec files used for the livedvd, ostensibly available at [5]. You may wish to confer with #gentoo-releng on Freenode for specific questions on catalyst usage. 1: http://www.gentoo.org/news/20121221-livedvd.xml 2: http://www.gentoo.org/proj/en/pr/releases/20121221/livedvd-amd64-multilib-20121221.iso.PACKAGES.txt 3: http://releases.gentooligans.com/ 4: http://www.gentoo.org/proj/en/releng/catalyst/ 5: http://git.overlays.gentoo.org/gitweb/?p=proj/releng.git;a=tree;f=releases/ -- eroen signature.asc Description: PGP signature
[gentoo-user] Re:
On Tue, 4 Mar 2014 21:26:27 + (UTC), Grant Edwards grant.b.edwa...@gmail.com wrote: On 2014-03-04, Alan McKinnon alan.mckin...@gmail.com wrote: You need a running OS to install an OS. You get this anywhere you choose but the easiest is to boot from a removeable media (CD, USB, etc). Booting the Gentoo minimal install CD (from CD/DVD or USB flash drive) is the canonical way to get to the point where you start following the steps in the handbook. On one recent install, I used systemrescuecd 4.xx, which gives you a nice minimal XFCE desktop with the Midori web browser you can use to read the handbook (and wander around the web while stuff builds). There was one step -- installing syslog-ng, I think -- where there was a glitch due to an environment variable set by srcd. Immediately after doing the chroot do the following command # export path= See below for more details http://forums.gentoo.org/viewtopic-t-921678-start-0.html http://forums.gentoo.org/viewtopic-t-786386.html Gentoo used to have a page[1] documenting the required deviations from the handbook to install from an arbitrary recent-glibc linux environment. Now it redirects to a wiki page[2] with roughly similar content. When installing from media other than the official gentoo minimal install cd, you should refer to the section currently labelled Installation from non-Gentoo LiveCDs in [2] to avoid surprises. In particular, I have several times seen people run into problems due to not properly adapting the instructions on mounting (thus missing /dev/pts in the chroot) and cleaning the environment prior to chrooting (thus ending up with incomplete PATH or having various variables set that break the build systems for packages bundling gnulib). (Sorry for arrogantly repeating some of the information you provided, but since forums and mailing list entries get archived and are searchable forever and are not updated, I feel it is prudent to provide pointers to resources that have a greater chance of staying relevant.) 1: http://www.gentoo.org/doc/en/altinstall.xml 2: https://wiki.gentoo.org/wiki/Installation_alternatives -- eroen signature.asc Description: PGP signature
[gentoo-user] Re: No static or multilib support for libtcl?
On Tue, 4 Mar 2014 16:35:30 + (UTC), Grant Edwards grant.b.edwa...@gmail.com wrote: Why isn't there a static or mutlilib libtcl.so? When I want to build a static binary that uses libtcl, I have to build my own private copy of libtcl? I don't know about a static one, but versions since dev-lang/tcl-8.5.15-r1 are multilib-enabled through the abi_x86_64 use flag (or the related ABI_X86 use_expand) in testing arches. Unfortunately, this capability is presently difficult to leverage on stable arches, particularly because the dependency on multilib zlib which (indirectly through blockers) turns this into an all-or-nothing affair unless you can live without multilib variants of the other libraries that have until now been provided by the emul-linux-x86-baselibs package. For your requirements, I would suggest you rather build tcl yourself for your development and leave the gentoo packaged version for use with other packages that use it. -- eroen
[gentoo-user] Re: banshee installation without systemd
On Mon, 24 Feb 2014 17:52:13 +0800, William Kenworthy bi...@iinet.net.au wrote: moriah ~ # emerge gnome-base/gnome-settings-daemon -vp These are the packages that would be merged, in order: Calculating dependencies... done! [ebuild N ] gnome-base/gnome-desktop-3.8.4:3/7 USE=introspection -debug 1,019 kB [ebuild N ] x11-themes/gnome-themes-standard-3.8.4 USE=gtk 3,765 kB [ebuild N ] sys-apps/systemd-208-r2:0/1 USE=filecaps firmware-loader gcrypt kmod lzma pam policykit python tcpd xattr -acl -audit -cryptsetup -doc -gudev -http -introspection -qrcode (-selinux) {-test} -vanilla PYTHON_SINGLE_TARGET=python2_7 PYTHON_TARGETS=python2_7 2,335 kB [ebuild N ] sys-apps/gentoo-systemd-integration-2 0 kB [ebuild N ] gnome-base/gnome-settings-daemon-3.8.6.1 USE=colord cups i18n policykit short-touchpad-timeout udev -debug (-openrc-force) (-packagekit) {-test} INPUT_DEVICES=-wacom 1,543 kB [blocks B ] sys-apps/systemd (sys-apps/systemd is blocking sys-fs/eudev-1.4-r1) I have a system like the above ... eudev/openrc with openrc-force in the USE flags and the 13.0 profile (not desktop/gnome etc) As you can see above, something is forcing (-openrc-force) - any ideas on how to get it to honour the openrc-force flag or find out what is causing the problem? BillK openrc-force is masked in /usr/portage/profiles/base/use.mask: # Pacho Ramos pa...@gentoo.org (28 Sep 2013) # This USE flag is available after long dicussion in # http://www.gossamer-threads.com/lists/gentoo/dev/276077 # to let some prople not able to run systemd to skip the dep (#480336). # Enabling this you will get a fully unsupported Gnome setup that # could suffer unexpected problem, don't expect support for it then. openrc-force If you are unfamiliar with gentoo's profiles, most (all the handbook-documented) profiles inherit the base profile which contains this file, which means this mask is in effect on most systems. None of the other profiles currently disable this use-mask. emerge's output indicates to you that this use flag is masked by enclosing it in parenthesis, check the documentation for the --verbose switch in emerge(1)[1]. Use-flag masks from the selected profile (and the opposite, force) override the use-flag settings normally made by users in make.conf and package.use. User modifications to the profile should be made in /etc/portage/profile/, see portage(5)[2]. To un-mask the openrc use flag, you can create the file /etc/portage/profile/use.mask with this line: -openrc-force 1: https://dev.gentoo.org/~zmedico/portage/doc/man/emerge.1.html 2: https://dev.gentoo.org/~zmedico/portage/doc/man/portage.5.html -- eroen signature.asc Description: PGP signature
[gentoo-user] Re: EAPI 4-python
On Mon, 24 Feb 2014 14:30:29 +0100, Fox halfsocial...@gmail.com wrote: Hello, I am trying to install an ebuild that used EAPI=4-python getting the error: API of python.eclass in EAPI=4-python not established I googled the problem but there is not much to read (or at least I could not find much) and what is there is old. So I wonder what is the problem with this and if there is a way to use it as there are many ebiulds (maybe only in overlays, I am not sure) that use it. Quim Afaik the *-python eapis are almost exclusively used by Arfrever's Progress overlay (and, by extension, funtoo). The error message you show seems to be from python.eclass in the main gentoo tree, which does not take un-official eapis into account. The code in question for reference (lines 30-32): if ! has ${EAPI:-0} 0 1 2 3 4 5; then die API of python.eclass in EAPI=\${EAPI}\ not established fi You might be able to use the ebuild stand-alone by also copying the relevant eclasses from whereever you got the ebuild into your local overlay (where I presume you put the ebuild?). However, from previous experience with the Progress overlay, you might want to use the entire overlay though layman in stead. Due to unfortunately incompatible python-implementation dependencies with gentoo proper it's rather an all-or-nothing deal. -- eroen signature.asc Description: PGP signature
[gentoo-user] Re: EAPI 4-python
On Mon, 24 Feb 2014 16:49:56 +0100, Fox halfsocial...@gmail.com wrote: On 02/24/2014 04:32 PM, eroen wrote: On Mon, 24 Feb 2014 14:30:29 +0100, Fox halfsocial...@gmail.com wrote: Hello, I am trying to install an ebuild that used EAPI=4-python getting the error: API of python.eclass in EAPI=4-python not established I googled the problem but there is not much to read (or at least I could not find much) and what is there is old. So I wonder what is the problem with this and if there is a way to use it as there are many ebiulds (maybe only in overlays, I am not sure) that use it. Quim Afaik the *-python eapis are almost exclusively used by Arfrever's Progress overlay (and, by extension, funtoo). The error message you show seems to be from python.eclass in the main gentoo tree, which does not take un-official eapis into account. The code in question for reference (lines 30-32): if ! has ${EAPI:-0} 0 1 2 3 4 5; then die API of python.eclass in EAPI=\${EAPI}\ not established fi You might be able to use the ebuild stand-alone by also copying the relevant eclasses from whereever you got the ebuild into your local overlay (where I presume you put the ebuild?). However, from previous experience with the Progress overlay, you might want to use the entire overlay though layman in stead. Due to unfortunately incompatible python-implementation dependencies with gentoo proper it's rather an all-or-nothing deal. The ebuild is from the ezod overlay. I am trying to use the ROS related packages like wstool, rosdep. etc. They all seem to use this EAPI. I thought that using this overlay would be easier than using pip but apparently it's not is it? Installing python packages though pip (without using virtualenv or such) is not generally a great idea on gentoo, in particular many not-quite-standard packages seem to get confused by gentoo's python-exec and overwrite it, or automagically pull in some other dependency (typically setuptools) which is already installed, and clobber files that have been especially patched for gentoo with non-working vanilla versions. Looking at [1], it appears the overlay maintainer just recently (last 4 commits) moved to use the 4-python eapi. From having made this mistake myself in the past, I would guess he added the eclasses to his overlay locally and forgot to add them to git. Alternatively, an updated layout.conf might not have been commited. You could try creating an eclass/ folder in the overlay and put distutils.eclass and python.eclass from the Progress overlay[2] in there. I notice that the ebuilds in question seem to be quite cookie-cutter (as python ebuilds generally are, bless them!). Functionally migrating them to the gentoo python eclasses should not be a large job. Change `EAPI=4-python` to `EAPI=5`, change `inherit distutils` to `inherit distutils-r1`, add a line `PYTHON_COMPAT=( python2_7 python3_3 )` (or as appropriate). Slightly more involved (but not strictly necessary to make it work locally) is to fix up the dependencies. In particular, python library dependencies should specify the required python implementation, like so: DEPEND=dev-python/pyyaml[${PYTHON_USEDEP}] See the wiki[3] for more information on python ebuilds :-) I CC the ezod overlay maintainer on this email to inform about the issue and facilitate a proper fix, either in documentation form or overlay code. 1: http://git.overlays.gentoo.org/gitweb/?p=user/ezod.git 2: http://code.google.com/p/gentoo-progress/ 3: https://wiki.gentoo.org/wiki/Project:Python/distutils-r1 -- eroen signature.asc Description: PGP signature
[gentoo-user] Re: libpng slot usage
On Sun, 23 Feb 2014 17:54:07 +0100, Fox halfsocial...@gmail.com wrote: Hello, I am trying to compile fltk-1.1.10 both using an ebuild from [1] and with just the downloaded source. I think this version needs libpng 1.2 which is installed in one of the slots. Using the source code and cmake with FLTK_USE_SYSTEM_PNG off it complies fine. I can see that it automatically points the libs to /usr/lib64/libpng12.so.0. The ebuild does not use cmake, it uses autoconf/automake which is also suported. I tried to build the code this way with --disable/enable-localpng with no luck. The problem is the used png.h file. Which is /usr/include/png.h which points to /usr/include/libpng16/png.h but I need the one from version 1.2. I thought slots would allow to have both versions of the library and use them but only one include file is present: $ equery f libpng:1.2 * Searching for libpng:1.2 ... * Contents of media-libs/libpng-1.2.50-r1: /usr /usr/lib64 /usr/lib64/libpng12.so.0 /usr/share /usr/share/doc /usr/share/doc/libpng-1.2.50-r1 /usr/share/doc/libpng-1.2.50-r1/CHANGES.bz2 /usr/share/doc/libpng-1.2.50-r1/README.bz2 /usr/share/doc/libpng-1.2.50-r1/TODO.bz2 Any idea on how to solve this problem? Some how it should be possible maybe not using the system lib like I did with cmake but I can't figure it out how to do it with autoconf/automake. Thank you, Quim [1] http://sources.gentoo.org/cgi-bin/viewvc.cgi/gentoo-x86/x11-libs/fltk/fltk-1.1.10.ebuild?view=log The libpng slots, apart from 0, only install the shared library files, not the headers. They are intended for compatibility with old binaries for which no source code is available, not for building new software. Afaik the common opinion is that different libpng versions are mostly source compatible, and minor patching to other things is preferable to inventing a non-standard way to make the libpng implementation switchable at build-time. Is there a reason you can not use fltk-1.1.10-r2.ebuild [1] in stead, which incorporates a patch [2] for libpng-1.5 (which probably still works with 1.6), as well as various other fixes? 1: http://sources.gentoo.org/cgi-bin/viewvc.cgi/gentoo-x86/x11-libs/fltk/fltk-1.1.10-r2.ebuild 2: http://sources.gentoo.org/cgi-bin/viewvc.cgi/gentoo-x86/x11-libs/fltk/files/fltk-1.1.10-libpng15.patch -- eroen signature.asc Description: PGP signature
[gentoo-user] Re: banshee installation without systemd
On Sun, 23 Feb 2014 18:02:01 +0100, Fox halfsocial...@gmail.com wrote: Hello, after reading the thread about systemd somebody mentioned sys-fs/eudev. I decided used because I had systemd only to used udev and unmerge systemd. Now I can't use Banshee which I use as my music player because of the next dependency tree: banshee - gnome-base/gnome-settings-daemon - sys-apps/gentoo-systemd-integration - sys-apps/systemd and systemd can't be used because it conflicts with eudev. Is there anyway to avoid emerge systemd in this case? Thank you, Quim On a system running ~amd64 with eudev/openrc: eroen@falcon ~ $ emerge -pv media-sound/banshee =gnome-base/gnome-settings-daemon-2.32.1-r2 These are the packages that would be merged, in order: Calculating dependencies . . ... done! [ebuild N ] media-sound/banshee-2.6.1 USE=aac bpm cdda encode mtp {test} udev web -daap -doc -ipod -karma -youtube 3,246 kB [ebuild N ] dev-dotnet/gtk-sharp-beans-2.14.0 21 kB [ebuild N ] dev-dotnet/gtk-sharp-gapi-2.12.10:2 USE=-debug 1,601 kB [ebuild NS]sys-devel/automake-1.10.3:1.10 [1.9.6-r3:1.9, 1.11.6:1.11, 1.12.6:1.12, 1.13.4:1.13, 1.14.1:1.14] 936 kB [ebuild N ]dev-lang/mono-3.2.3 USE=nls -debug -doc -minimal -pax_kernel -xen 0 kB [ebuild N ] dev-dotnet/libgdiplus-2.10.9-r1 USE=cairo 0 kB [ebuild N ] dev-dotnet/glib-sharp-2.12.10:2 USE=-debug 0 kB [ebuild N ] dev-dotnet/gtk-sharp-2.12.10:2 USE=-debug 0 kB [ebuild N ]dev-dotnet/atk-sharp-2.12.10:2 USE=-debug 0 kB [ebuild N ]dev-dotnet/gdk-sharp-2.12.10:2 USE=-debug 0 kB [ebuild N ] dev-dotnet/pango-sharp-2.12.10:2 USE=-debug 0 kB [ebuild N ] dev-dotnet/gio-sharp-0.3 88 kB [ebuild N ] dev-dotnet/gkeyfile-sharp-0.1 20 kB [ebuild N ] media-plugins/gst-plugins-taglib-0.10.31:0.10 0 kB [ebuild N ] dev-dotnet/dbus-sharp-0.7.0-r1 125 kB [ebuild N ] media-plugins/gst-plugins-soundtouch-0.10.23:0.10 0 kB [ebuild N ] media-libs/libsoundtouch-1.8.0 USE=sse2 -static-libs 104 kB [ebuild N ] dev-dotnet/gconf-sharp-2.24.2:2 USE=-debug 412 kB [ebuild N ] dev-dotnet/gnome-sharp-2.24.2:2 USE=-debug 0 kB [ebuild N ]dev-dotnet/art-sharp-2.24.2:2 USE=-debug 0 kB [ebuild N ]dev-dotnet/gnomevfs-sharp-2.24.2:2 USE=-debug 0 kB [ebuild N ] dev-dotnet/glade-sharp-2.12.10:2 USE=-debug 0 kB [ebuild N ] net-libs/webkit-gtk-2.2.5-r200:2 USE=egl gstreamer jit opengl {test} webgl (-aqua) -coverage -debug -geoloc -gles2 -introspection -libsecret -spell 9,181 kB [ebuild N ] media-plugins/gst-plugins-lame-0.10.19:0.10 0 kB [ebuild N ] dev-dotnet/dbus-sharp-glib-0.5.0 94 kB [ebuild N ] gnome-base/gnome-settings-daemon-2.32.1-r2 USE=libnotify -debug -policykit -pulseaudio -smartcard 1,327 kB [ebuild N ] gnome-base/libgnomekbd-2.32.0-r1 USE={test} 402 kB [ebuild N ] gnome-base/gnome-desktop-2.32.1-r2:2 USE=-debug -license-docs PYTHON_TARGETS=python2_7 -python2_6 1,596 kB [ebuild N ] dev-dotnet/notify-sharp-0.4.0_pre20090305 USE=-doc 78 kB [ebuild N ] dev-dotnet/taglib-sharp-2.1.0.0 503 kB [ebuild N ] media-plugins/gst-plugins-gio-0.10.36:0.10 0 kB [ebuild N ] dev-dotnet/mono-addins-0.6.2 USE=gtk 330 kB [ebuild N ] dev-dotnet/gudev-sharp-0.1 101 kB Total: 33 packages (32 new, 1 in new slot), Size of downloads: 20,155 kB For ease of upgrades, you might want to add =gnome-base/gnome-settings-daemon-3 in /etc/portage/package.mask rather than specifying it with a specific version on the command line. -- eroen signature.asc Description: PGP signature
[gentoo-user] Re: banshee installation without systemd
On Sun, 23 Feb 2014 19:47:55 -0600, Canek Peláez Valdés can...@gmail.com wrote: On Sun, Feb 23, 2014 at 7:31 PM, eroen er...@falcon.eroen.eu wrote: On Sun, 23 Feb 2014 18:02:01 +0100, Fox halfsocial...@gmail.com wrote: Hello, after reading the thread about systemd somebody mentioned sys-fs/eudev. I decided used because I had systemd only to used udev and unmerge systemd. Now I can't use Banshee which I use as my music player because of the next dependency tree: banshee - gnome-base/gnome-settings-daemon - sys-apps/gentoo-systemd-integration - sys-apps/systemd and systemd can't be used because it conflicts with eudev. Is there anyway to avoid emerge systemd in this case? Thank you, Quim On a system running ~amd64 with eudev/openrc: eroen@falcon ~ $ emerge -pv media-sound/banshee =gnome-base/gnome-settings-daemon-2.32.1-r2 [ snip emerge output ] For ease of upgrades, you might want to add =gnome-base/gnome-settings-daemon-3 in /etc/portage/package.mask rather than specifying it with a specific version on the command line. That solution is a dead end. GNOME 2 is being removed from the tree[1]. Regards. [1] http://blogs.gentoo.org/eva/2014/02/16/the-future-of-gnome-2/ You probably mean temporary. The expression dead end would imply it makes future migration more difficult in some way. One would hope (mostly for their image's sake) the gnome team does not remove gnome-settings-daemon-2 until at least one of cinnamon-settings-daemon and mate-settings-daemon are included in gentoo proper. -- eroen signature.asc Description: PGP signature
[gentoo-user] Re: --jobs is ignored for unmerging?
On Fri, 21 Feb 2014 16:14:00 +0200, Nikos Chantziaras rea...@gmail.com wrote: I recently needed to unmerge Netbeans and KDE from a machine. So removed the top-level packages of those, and then ran: emerge -a --depclean --jobs 20 However, --jobs is being ignored. So I'm sitting there, watching hundreds of packages being unmerged one by one, taking a long time :-/ Is this normal? Shouldn't the --jobs option result in emerge unmerging multiple packages at the same time? From emerge(1)[1]: -j [JOBS], --jobs[=JOBS] Specifies the number of packages to build simultaneously. (...) In the past, unmerging was almost entirely IO bound, and running several in parallel would have (at best) no significant speedup. I suspect the (fairly recent) preserve-libs feature and detection of unmodified configuration files introduced some cpu-intensive parts, which might be parallelizeable to some extent if you sacrifice some of the recoverability of an aborted unmerge. If you can benchmark low IO utilization during unmerges, I suggest you open a feature request bug on bugs.gentoo.org requesting parallel unmerge. :-) In the mean time, you might be able to speed the process up somewhat by setting FEATURES=-merge-sync in the environment when you do unmerges. (see make.conf(5)[2]) 1: https://dev.gentoo.org/~zmedico/portage/doc/man/emerge.1.html 2: https://dev.gentoo.org/~zmedico/portage/doc/man/make.conf.5.html -- eroen signature.asc Description: PGP signature
[gentoo-user] Re: emerge all packages which depend on P : how to
On Thu, 20 Feb 2014 13:38:06 +0100, Helmut Jarausch jarau...@igpm.rwth-aachen.de wrote: Hi, I have a very simple question. How to emerge (update) all packages which depend on some given package P. I've tried emerge -uv1 `equery -q d P` or emerge -uv1 `qdepends -q -Q P` but both commands (equery and qdepends) generate a list with the version attached like app-editors/kile-2.1.3 which emerge doesn't like (unless there is an '= in front of each name) Is there an easy way to do so without resorting to shell/python scripting? Many thanks for a hint, Helmut How about one of these? emerge -1 $(qdepends -N -C -Q P) emerge -1 $(qdepends -C -Q P | sed 's/^/=/') -- eroen signature.asc Description: PGP signature
[gentoo-user] Re: systemd and USE=static
On Thu, 20 Feb 2014 21:48:01 +, Neil Bothwick n...@digimed.co.uk wrote: I thought I'd have another look at systemd, so switched profiles, made sure I had all the kernel options needed and then did an emerge -uN world. It seems the systemd profile masks the static and static-libs USE flags, which are needed by crytsetup and lvm for my initramfs, which mounts / from a LUKS partition. Forcing an unmask of these flags in /etc/portage/profile did no good because when those packages are built with USE=static they require a virtual/udev with matching flags, which systemd cannot provide. Is there a way around this or have a I found a reason for not using systemd that doesn't involve name calling? :-O You are right, systemd dropped support for static libudev (or any other library for that matter) in version 205.[1] If you want to use systemd provided libudev for lvm c. in an initramfs, you'll need to include the shared library and binaries linked to it. I believe both dracut and genkernel-next support shared libudev, while the current versions of genkernel do not (hence why genkernel is masked on the */systemd profiles). While I have limited experience with genkernel, dracut has dragged in all the shared libraries I have required of it. 1: http://cgit.freedesktop.org/systemd/systemd/commit/?id=5e63ce78b5018ba612e794a610a6f13c5eefade7 -- eroen signature.asc Description: PGP signature
[gentoo-user] Re: EFI-based bootloader for BIOS-based computers (?)
On Wed, 19 Feb 2014 15:39:51 -0800, walt w41...@gmail.com wrote: I just spotted that phrase in the sourceforge newsletter: http://sourceforge.net/projects/cloverefiboot/ and it seems to me like an oxymoron. If that phrase makes logical sense then my definitions of 'BIOS' and 'EFI' need the latest updates :) Until now I thought that EFI is a recent replacement for BIOS based machines. Can anyone clarify the linguistics involved here? The scope of UEFI is somewhat greater than that of traditional BIOSes. Both do various hardware initialization and such, but UEFIs (can) have a number of additional features, including more flexibility in what it can launch from where (eg. network booting without iPXE) and even an interactive shell. See [1] for a less organized list of features. I'm unfamiliar with this project in specific, but I'm going by the line This is EFI-based bootloader for BIOS-based computers created as a replacement to EDK2/Duet bootloader http://www.tianocore.org. I have a box running Duet, which is an UEFI implementation that can be launched by (eg.) the extlinux boot loader on a legacy BIOS system. Once Duet is launched, the system is mostly indistinguishable from a native UEFI system that has booted into it's UEFI firmware. From here, Duet can let the user go through menus to select an EFI executable to launch (a EFI-stub enabled kernel or some sort of boot loader), or it can automatically launch something based on existing configuration. 1: https://en.wikipedia.org/wiki/UEFI#Features -- eroen signature.asc Description: PGP signature
[gentoo-user] Re: systemd as a Profile - practical or not?
On Tue, 18 Feb 2014 11:24:05 -0500, Tanstaafl tansta...@libertytrek.org wrote: Hi all, Ok, before I go and open up a bug requesting this... I know there have to be a lot of people on this list who can answer this question... Is making the use of systemd or not based on a selected Profile, as opposed to manually trying to do it via USE flags etc, a practical request, or not? How is this different from the status quo? There are several systemd profiles available that make use of the files in /usr/portage/profiles/targets/systemd/ to make the proper package settings for using systemd on gentoo. AFAIU, a user should only need to switch the profile, install and configure systemd itself and configure their bootloader to start using systemd. -- eroen signature.asc Description: PGP signature
[gentoo-user] Re: Debian just voted in systemd for default init system in jessie
On Sat, 15 Feb 2014 14:34:34 -0600, Daniel Campbell li...@sporkbox.us wrote: Why didn't they consider runit? It has parallel execution of daemons and is backwards compatible with sysv. It has a few other mini-features as well, iirc. I used for a little while before Arch pushed systemd on their community and it was interesting. I'll just put this link to a forum thread on epoch from late last year, in case any potentially interested party has not seen it yet. It's available in the gentoo package tree, and from the thread it seems to have workable integration. http://forums.gentoo.org/viewtopic-t-975382-highlight-epoch.html -- eroen signature.asc Description: PGP signature
[gentoo-user] Re: WinTV Nova-T
On Sat, 15 Feb 2014 23:16:15 +0100, Silvio Siefke siefke_lis...@web.de wrote: siefke ~ $ lsusb Bus 001 Device 003: ID 0402:9665 ALi Corp. Gateway Webcam Bus 001 Device 008: ID 2040:7070 Hauppauge Nova-T Stick 3 Bus 001 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub Bus 005 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub Bus 004 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub Bus 003 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub Bus 002 Device 002: ID 1d57:0008 Xenta Bus 002 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub It appears (from grepping and skimming drivers/media/usb/dvb-usb/dib0700_devices.c and .../Kconfig) that the driver for your device is controlled by the DVB_USB_DIB0700 kernel configuration option, which seems to be disabled in the .config you linked. Please try enabling this option in the kernel configuration, build and boot to the new kernel, and see if that brings you a bit further along :-) In menuconfig, the option should be found here: - Device Drivers - Multimedia support - Media USB Adapters - Support for various USB DVB devices - DiBcom DiB0700 USB DVB devices -- eroen signature.asc Description: PGP signature
[gentoo-user] Re: to install portage on other gentoo installs
On Wed, 12 Feb 2014 16:40:22 -0800, Edward M edwardm.gentoo.j...@live.com wrote: Howdy, Been busy learning Linux :-) got new email other was getting crowded. I'm planing on installing Gentoo on a few systems and I was wondering to save bandwidth, i could install portage to the other Gentoo installs from my system instead downloading from mirrors? Thanks in advance! Setting up a local rsync mirror for a portage tree is fairly simple, you share the working tree at /usr/portage on a local server to the network and make sure the server is up-to-date before syncing other clients to it. This is documented in [1]. After setting up the mirror, you only need to sync-uri in /etc/portage/repos.conf on the other (local) clients. See portage(5)[2], section 'repos.conf' and make.conf(5)[3], section 'SYNC' for details on client configuration. Please *do* read those sections, as much of the information on the wiki regarding the SYNC variable is outdated. For sharing distfiles between hosts, a http server sharing the distfiles folder and prepending the server's uri to GENTOO_MIRRORS in make.conf[3] might be sufficient for your needs. If the installs are not homogenous, you can even set up http servers on all of them, and add all the other local hosts to GENTOO_MIRRORS. You might want to skim through the FEATURES section in make.conf(5)[3] for using local distfiles mirrors, as some of them can improve the usefulness of the mirror significantly. NFS is another possibility for both portage tree and distfiles and can reduce total disk space needs, but it incurs significant network latency which is especially noticeable for portage trees. I use it myself for virtual machines, but would not suggest it for use over a physical network. Also note that portage has PORTAGE_RO_DISTDIRS[3] as an alternative to GENTOO_MIRRORS for eg. read-only *mounted* network shares which may provide some extra flexibility. 1: https://wiki.gentoo.org/wiki/Project:Infrastructure/Rsync#Setting_up_your_own_local_rsync_mirror 2: https://dev.gentoo.org/~zmedico/portage/doc/man/portage.5.html 3: https://dev.gentoo.org/~zmedico/portage/doc/man/make.conf.5.html -- eroen signature.asc Description: PGP signature
[gentoo-user] Re: User eix-sync permissions problem
On Mon, 10 Feb 2014 14:03:44 -0500, Walter Dnes waltd...@waltdnes.org wrote: On Mon, Feb 10, 2014 at 05:09:55PM +, Stroller wrote On Mon, 10 February 2014, at 4:55 pm, Gleb Klochkov glebiu...@gmail.com wrote: Hi. Try to use sudo with no password for eix-sync. I'd really rather not. Thanks, though. Being in group portage is not enough. That merely lets you do emerges with --pretend. emerge --sync modifies files in /usr/portage. Files and directories in /usr/portage/ are user:group root:root. Therefore you *NEED* root-level permission to modify them. No ifs/ands/ors/buts. The overall easiest method is to (as root)... * emerge sudoers if it's not installed * visudo -f /etc/sudoers.d/001 (or whatever you want to call the file) * set up the file. Here's a fragment from my system, with user waltdnes and machine name i660 waltdnes i660 = (root) NOPASSWD: /usr/sbin/hibernate waltdnes i660 = (root) NOPASSWD: /sbin/fdisk -l I could manually type the command with sudo, but I'm lazy. In my /home/waltdnes/bin directory, I have a file hb #!/bin/bash sync sleep 15 sudo /usr/sbin/hibernate and file fdl #!/bin/bash sudo /sbin/fdisk -l To sync the machine, I could add to /etc/sudoers.d/001 waltdnes i660 = (root) NOPASSWD: /usr/bin/emerge --sync and create (as waltdnes) /home/waltdnes/emsy #!/bin/bash /usr/bin/emerge --sync For security, I strongly recommend that the full path of the executable be specified, as well as any options. Do not use the $* commandline parameter in the sudoers file. It probably works, but is too wide open. eroen@falcon ~ $ wget -O - 'http://mirrors.eu.kernel.org/gentoo/snapshots/portage-20140209.tar.xz' 2/dev/null | tar tvJ | head -n 10 drwxr-xr-x portage/portage 0 2014-02-10 01:31 portage/ -rw-r--r-- portage/portage 1232 2013-03-05 22:31 portage/skel.metadata.xml drwxr-xr-x portage/portage0 2014-02-10 01:31 portage/sec-policy/ drwxr-xr-x portage/portage0 2014-01-12 21:31 portage/sec-policy/selinux-thunderbird/ -rw-r--r-- portage/portage 448 2012-10-13 18:31 portage/sec-policy/selinux-thunderbird/selinux-thunderbird-.ebuild -rw-r--r-- portage/portage 10496 2014-01-12 21:31 portage/sec-policy/selinux-thunderbird/Manifest -rw-r--r-- portage/portage 476 2013-02-23 18:31 portage/sec-policy/selinux-thunderbird/selinux-thunderbird-2.20120725-r11.ebuild -rw-r--r-- portage/portage 475 2012-12-13 11:31 portage/sec-policy/selinux-thunderbird/selinux-thunderbird-2.20120725-r8.ebuild -rw-r--r-- portage/portage 475 2013-08-15 09:01 portage/sec-policy/selinux-thunderbird/selinux-thunderbird-2.20130424-r2.ebuild -rw-r--r-- portage/portage 475 2012-10-04 20:31 portage/sec-policy/selinux-thunderbird/selinux-thunderbird-2.20120725-r5.ebuild For portage's (default-enabled) FEATURES=usersync to work (dropping privileges when syncing as root), /usr/portage must be writeable by portage:portage. The tree snapshots have not always had this permissions setup, so mature installs would require manual intervention at some point, either updating the permissions or disabling usersync. Still, the files are not group-writeable by default, so portage group membership would not be sufficient. I would suggest a solution based on su/sudo, as merely changing the permissions would need to be re-done if the tree is ever synced as root later. -- eroen signature.asc Description: PGP signature
[gentoo-user] Re: media-libs/lilv-0.18.0 failed to compile
On Fri, 31 Jan 2014 03:56:36 +0100, meino.cra...@gmx.de wrote: On AMD64, recent Gentoo Linux, the library media-libs/lilv-0.18.0 failed to compile: snip! If wanted I will post all the files metioned below the above report to the console. But I dont want to pollute the mailing list in beforehand...may be someone has already solved this...?!? Thank you very much for any help in advance! Have a nice weekend! Best regards, mcc There are no open bugs for media-libs/lilv[1], and media-libs/lilv-0.18.0 builds fine here on ~amd64. I suggest making the build.log, config.log and `emerge --info` output available, like suggested in emerge's error output. If you like, you can add the logs as an attachments to an email to the list, but it might be better to open a bug on the Gentoo bug tracker[2]. That way, the maintainer of the package will learn about the issue directly, and can fix it for everyone. :-) [1]: https://bugs.gentoo.org/buglist.cgi?no_redirect=1quicksearch=media-libs%2Flilv [2]: https://bugs.gentoo.org -- eroen signature.asc Description: PGP signature
[gentoo-user] Re: emerge latest in a certain version series of a package
On Wed, 29 Jan 2014 16:09:31 +0200, Thanasis thana...@asyr.hopto.org wrote: on 01/28/2014 10:12 PM eroen wrote the following: On Tue, 28 Jan 2014 21:29:21 +0200, Thanasis thana...@asyr.hopto.org wrote: Is there a way to specify that I want to install (emerge) the latest 3.10.X series of sys-kernel/gentoo-sources as a slot, in parallel with the latest gentoo-sources? Currently, that would be version 3.10.28. I know I can specify it like so, emerge =sys-kernel/gentoo-sources-3.10.28 but then it would not get auto-updated when a newer version of that series (for example 3.10.29) becomes available in portage. Afaik there is no configuration-only way to do this, but you can make it happen by adding your own virtual ebuilds in a local overlay[1], say virtual/thanasis-sources, then adding that to world. You can make one ebuild for each kernel-series you want, each in its own slot, using the =cat/pkg-3.10* syntax for RDEPEND. Find attached a (barely tested) suggestion for virtual/thanasis-sources/thanasis-sources-3.10.ebuild . Thanks eroen, I tested it and it looks like it works. One caveat, =cat/pkg-3.10* also matches cat/pkg-3.107.1, I guess that kernel versions = 3.100 are going to be in the tree, if ever, probably in about a decade from now :P which could become an issue if you wanted, say, the 3.2 series. Why would that be a problem? In the ebuild, you specified the value RDEPEND==sys-kernel/gentoo-sources-${PV}* I think you got it, but then I confused you with a poorly chosen example. The problem I referred to; if you try this exact method to follow the 3.2.x kernels, you will switch to 3.20.0 when that comes around. Similarly, it does not currently work with 3.1.x, and could eventually (in a very long time, like you commented :P ) break for other versions. It would be nice to have a better way to accomplish this, perhaps with some sort of world file variant supporting version ranges. That could also be more intuitive and less prone to unintended consequences (especially in the face of SLOTs) than package.mask-ing newer versions to keep an old version of something around. -- eroen signature.asc Description: PGP signature
[gentoo-user] Re: emerge latest in a certain version series of a package
On Tue, 28 Jan 2014 21:29:21 +0200, Thanasis thana...@asyr.hopto.org wrote: Is there a way to specify that I want to install (emerge) the latest 3.10.X series of sys-kernel/gentoo-sources as a slot, in parallel with the latest gentoo-sources? Currently, that would be version 3.10.28. I know I can specify it like so, emerge =sys-kernel/gentoo-sources-3.10.28 but then it would not get auto-updated when a newer version of that series (for example 3.10.29) becomes available in portage. Afaik there is no configuration-only way to do this, but you can make it happen by adding your own virtual ebuilds in a local overlay[1], say virtual/thanasis-sources, then adding that to world. You can make one ebuild for each kernel-series you want, each in its own slot, using the =cat/pkg-3.10* syntax for RDEPEND. Find attached a (barely tested) suggestion for virtual/thanasis-sources/thanasis-sources-3.10.ebuild . One caveat, =cat/pkg-3.10* also matches cat/pkg-3.107.1, which could become an issue if you wanted, say, the 3.2 series. After setting up an overlay and adding the ebuild, you can add it to world with `emerge --select virtual/thanasis-sources:3.10`. To add a different kernel series, you should only need to rename/copy the ebuild in your overlay and add the new version/slot to world. [1]: https://wiki.gentoo.org/wiki/Overlay/Local_overlay -- eroen thanasis-sources-3.10.ebuild Description: Binary data signature.asc Description: PGP signature
[gentoo-user] Re: cdnpayroll.py
On Sat, 25 Jan 2014 19:10:49 -0700, Joseph syscon...@gmail.com wrote: On 01/25/14 14:21, David Abbott wrote: On Sat, Jan 25, 2014 at 1:15 PM, Joseph syscon...@gmail.com wrote: File ./cdnpayroll.py, line 1328 ^ SyntaxError: EOF while scanning triple-quoted string literal Try changing the shebang from:to #!/usr/bin/python /#!/usr/bin/python2 Thanks David, that was it. It should be /#!/usr/bin/python2 Alternatively, you can make the /usr/bin/python symlink point to a python2 implementation with eg. eselect python set python2.7 I do this on most of my systems, since there are still a lot of old python scripts, that used to be valid in every way until python3 came around, that assume 'python' is compatible with python2. Several packages in Gentoo still have build failures due to this, although their numbers are dwindling. Also, scripts written for python3 that are incompatible with earlier implementations should (IMO) be expected to declare a more specific shebang. PS: for completeness' sake, the /usr/bin/python2 and ...3 symlinks can be changed with the --python2 and --python3 arguments after 'set' in the eselect command. Available implementations can be listed with eselect python list which also optionally takes the --python2 and --python3 arguments. -- eroen signature.asc Description: PGP signature
[gentoo-user] Re: Portage performance dropped considerably
On Sun, 26 Jan 2014 16:35:43 +0200, Nikos Chantziaras rea...@gmail.com wrote: Anyone else noticed this yet? Some portage update seems to have made emerge -uDN @world perform about 10 times slower than before. It used to take seconds, now it takes about 4 minutes only to tell me that there's nothing to update. And it does that every time, even directly in succession and with the caches warm. Is it just me? You don't say when your baseline was, but the complexity of resolving the package tree has increased quite a bit over the last year due to new features like automatic rebuilds of consumers after library updates. Another somewhat common cause of sudden slowdowns is how portage resolves conflicts (like packageA requiring an old version of libraryB), which is rather time-consuming. You can try adding --backtrack=0 to the emerge command to make it stop and print an error message when encountering a conflict rather than look for a solution. Then you can 'help' out by manually resolving any conflicts by adding package versions to /etc/portage/package.mask . Preferably try this *after* running an update, so your system is up-to-date against your local version of the gentoo tree, otherwise normal simple-to-resolve conflicts might cause confusion. ;-) -- eroen signature.asc Description: PGP signature
[gentoo-user] Re: Portage performance dropped considerably
On Sun, 26 Jan 2014 07:09:49 -0800, Greg Turner g...@malth.us wrote: It would help if there were some decent way to get some statistics about where portage is spending all its time (especially, I suspect, within the bash code, but the python level would also be interesting to analyse). Anyone know of a good way of doing that? http://permalink.gmane.org/gmane.linux.gentoo.devel/89518 Also, the attached patch to portage (works with 2.2.7 and 2.2.8 at least, that too by TomWij) makes portage print out some more information on what part of dependency resolution currently takes place, without all the noise from --debug. -- eroen index 7b77edc..cfd7025 100644 --- a/pym/_emerge/depgraph.py +++ b/pym/_emerge/depgraph.py @@ -86,6 +86,10 @@ if sys.hexversion = 0x300: else: _unicode = unicode +from time import gmtime, strftime +def time_print(message): + print(\n%s: %s % (strftime(%Y-%m-%d %H:%M:%S, gmtime()), message), end=) + class _scheduler_graph_config(object): def __init__(self, trees, pkg_cache, graph, mergelist): self.trees = trees @@ -1585,6 +1589,10 @@ class depgraph(object): self._spinner_update() while dep_stack: dep = dep_stack.pop() +#try: +# time_print( Processing: %s % dep.cpv) +#except AttributeError: +# pass if isinstance(dep, Package): if not self._add_pkg_deps(dep, allow_unsatisfied=allow_unsatisfied): @@ -2106,6 +2114,7 @@ class depgraph(object): for dep_root, dep_string, dep_priority in deps: if not dep_string: continue +#time_print(Reducing: %s % dep_string) if debug: writemsg_level(\nParent:%s\n % (pkg,), noiselevel=-1, level=logging.DEBUG) @@ -2154,6 +2163,8 @@ class depgraph(object): if not dep_string: continue +#time_print(Disjunct: %s % dep_string) + if not self._add_pkg_dep_string( pkg, dep_root, dep_priority, dep_string, allow_unsatisfied): @@ -3005,6 +3016,7 @@ class depgraph(object): virtuals = pkgsettings.getvirtuals() args = self._dynamic_config._initial_arg_list[:] + time_print(Adding root packages) for arg in self._expand_set_args(args, add_to_digraph=True): for atom in arg.pset.getAtoms(): self._spinner_update() @@ -3116,20 +3128,24 @@ class depgraph(object): # Now that the root packages have been added to the graph, # process the dependencies. + time_print(Processing dependencies) if not self._create_graph(): return 0, myfavorites + time_print(Checking for slot conflicts) try: self.altlist() except self._unknown_internal_error: return False, myfavorites + time_print(Checking for blocker conflicts) if (self._dynamic_config._slot_collision_info and not self._accept_blocker_conflicts()) or \ (self._dynamic_config._allow_backtracking and slot conflict in self._dynamic_config._backtrack_infos): return False, myfavorites + time_print(Checking for rebuild triggers) if self._rebuild.trigger_rebuilds(): backtrack_infos = self._dynamic_config._backtrack_infos config = backtrack_infos.setdefault(config, {}) @@ -3138,12 +3154,14 @@ class depgraph(object): self._dynamic_config._need_restart = True return False, myfavorites + time_print(Checking if restart is needed) if config in self._dynamic_config._backtrack_infos and \ (slot_operator_mask_built in self._dynamic_config._backtrack_infos[config] or slot_operator_replace_installed in self._dynamic_config._backtrack_infos[config]) and \ self.need_restart(): return False, myfavorites + time_print(Checking if we have to prune rebuilds) if not self._dynamic_config._prune_rebuilds and \ self._dynamic_config._slot_operator_replace_installed and \ self._get_missed_updates(): @@ -3161,10 +3179,12 @@ class depgraph(object): self._dynamic_config._need_restart = True return False, myfavorites + time_print(Checking if restart is needed) if self.need_restart(): # want_restart_for_use_change triggers this return False, myfavorites + time_print(Checking for parameters that change behavior) if --fetchonly not in self._frozen_config.myopts and \ --buildpkgonly in self._frozen_config.myopts: graph_copy = self._dynamic_config.digraph.copy() @@ -3188,6 +3208,7 @@ class depgraph(object): digraph_nodes = self._dynamic_config.digraph.nodes + time_print(Checking for changes that are needed) if any(x in digraph_nodes for x in self._dynamic_config._needed_unstable_keywords) or \ any(x in digraph_nodes for x in @@ -3200,6 +3221,7 @@ class depgraph(object): self._dynamic_config._success_without_autounmask = True return False, myfavorites + time_print(Done resolving!) # We're true here unless we are missing binaries. return (True, myfavorites) @@ -5851,21 +5873,25 @@ class depgraph(object): root_config.root][root_config] = root_config def _resolve_conflicts(self): - + time_print (Trying to accept blocker
[gentoo-user] Re: New dependencies suddenly popping up
On Tue, 21 Jan 2014 10:26:10 + Peter Humphrey pe...@prh.myzen.co.uk wrote: After my daily sync today, portage wanted to emerge two new packages: dev- libs/lzo and x11-misc/makedepend. No other packages were mentioned. Odd, I thought, so: $ equery d lzo * These packages depend on lzo: [...] X11-libs/cairo-1.12.14-r4 (dev-libs/lzo) https://bugs.gentoo.org/477078 cairo is/was built slightly differently depending on whether lzo was installed or not when cairo was built. The most significant difference is that cairo might stop working if lzo is removed after cairo was built*. To resolve this, the Gentoo maintainer recently added lzo to the dependencies to make sure lzo is not removed. If you rebuild cairo after installing lzo, some obscure feature might be enabled, but the maintainer seems to think this is not significant, and did not increase the revision number to force a rebuild everywhere. http://sources.gentoo.org/cgi-bin/viewvc.cgi/gentoo-x86/x11-libs/cairo/cairo-1.12.14-r4.ebuild?r1=1.12r2=1.13 * In most cases, that's prevented by the (now default-on) preserve-libs FEATURE in portage. Non-default setups aren't necessarily as lucky. $ equery d makedepend * These packages depend on makedepend: app-cdr/cdrtools-3.01_alpha17 (x11-misc/makedepend) http://sources.gentoo.org/cgi-bin/viewvc.cgi/gentoo-x86/app-cdr/cdrtools/cdrtools-3.01_alpha17.ebuild?r1=1.13r2=1.14 A similar change (though no bug :-( ), except makedepend is a build-time-only dependency, and not strictly required after cdrtools is built. I will not venture to guess what the consequences of building cdrtools without makedepend installed are, but since you somehow got it installed at some point, there is either a fallback of some sort, or you had it installed for other reasons. Whether this build-time dependency (DEPEND) change causes makedepend to be installed after the fact is governed by the --with-bdeps switch to emerge, check the man page for details. But cairo was last emerged 16 days ago and cdrtools last September, so why have these dependencies suddenly appeared today? Portage has been at 2.2.7 since last year too. Emerge --info doesn't mention lzo. By default, in order to let maintainers fix issues, portage calculates what to install based on the *current* package ebuilds, rather than the ebuild at the time a package was installed. While this can be disabled by the user, maintainers will expect it to work, so best not ;-) I can't think of anything else that might have changed, unless new dependencies have been shoved into the ebuilds. Does that sort of thing happen in a well run family? -- eroen signature.asc Description: PGP signature
[gentoo-user] Re: webcam software
On Sat, 18 Jan 2014 17:02:15 -0500 gottl...@nyu.edu wrote: My main system is a dell latitude E6430s. I am embarrassed to say that, although I have had this system for a while, I just now realized that it has a build in webcam. What software do you recommend and what should I start reading to learn how to use it. thanks, allan ffmpeg/libav works well for capturing input from webcams, at least if the driver uses the video4linux (v4l) framework (I haven't encountered any other kinds). For instance, to capture one frame and store it as a png image (using libav's avconv, ffmpeg would be similar), you can use avconv -f video4linux2 -i /dev/video0 testframe.png mplayer (and its descendants) also play well with v4l, to directly display the webcam's input you can use mplayer -tv driver=v4l2:device=/dev/video0 tv:// (The -tv switch and its argument are probably redundant as the default detection is very good.) -- eroen signature.asc Description: PGP signature
[gentoo-user] Re: eix-sync data comparison
On Sat, 11 Jan 2014 21:44:19 +0100, meino.cra...@gmx.de wrote: On 11/01/2014 22:04, meino.cra...@gmx.de wrote: suppose I would do an eix-sync on two different computers of different platforms -- say AMR and an ordinary PC. Sharing a portage tree (the files in /usr/portage) is perfectly safe, as this is identical for basic gentoo systems (prefix and similar are ...different. Best not go there ;-). However, the eix database is not portable between different systems. I do not have an exhaustive list of what settings cause incompatibility, but at least different ARCH (amd64 or arm) will cause issues since different packages are available to each. Especially with the ARM I need to sync once a day, since I have to limit the amount of software to be recompiled/updated after each sync since the ARM is not *that* fast (for example a kernel compilation take ~8h). While that is not strictly enforced, I myself run an rsync server locally, mostly to ensure that different systems have identical trees, thus preserving my sanity. Instructions for setting up a local rsync server can be found here: https://wiki.gentoo.org/wiki/Project:Infrastructure/Rsync To make a client (the arm box) sync from the local server, just change the sync-uri variable in /etc/portage/repos.conf/gentoo.conf and the SYNC variable (if present) in make.conf . You could even share a portage tree over nfs or similar, which I sometimes do for VMs. Depending on your (networking and storage) hardware, that might even work out favourably performance-wise. You could also generate a squashfs image on the desktop and transfer that to and mount it on the arm box (over http or something using a short shell script), if, say, the arm box is limited on storage space. A squashfs of the gentoo tree is about 70MB. -- eroen signature.asc Description: PGP signature
[gentoo-user] Re: make distclean
On Mon, 16 Dec 2013 15:12:58 -0600 Bruce Hill da...@happypenguincomputers.com wrote: On Mon, Dec 16, 2013 at 08:57:08PM +, Markos Chandras wrote: You could get your running config using zcat /proc/config.gz .config Only if you enabled: CONFIG_IKCONFIG=y CONFIG_IKCONFIG_PROC=y in your running kernel. If only CONFIG_IKCONFIG=y is enabled, you can still use the /usr/src/linux/scripts/extract-ikconfig script to extract the config from an existing kernel image. Regardless, enabling CONFIG_IKCONFIG_PROC=y makes things so much simpler for everyone, with very little to speak against enabling it. -- eroen signature.asc Description: PGP signature
[gentoo-user] Re: @preserved-rebuild gone in a loop
On Sun, 15 Dec 2013 17:37:53 +0100 Benjamin Block b...@zlug.org wrote: Most of the times, when some binary packages on my systems do cause something like this, then I just unemerge the package that keeps recompiling and emerge it again afterwards. This will cause the portage to drop the library-references in question and add new ones. So, this should do the trick: emerge -C app-antivirus/avast4workstation emerge -1 app-antivirus/avast4workstation This will make the message from portage and the old library version go away, yes. It will also cause the program that used the library (/opt/avast4workstation/bin/avastgui in OP's case) crash when you try to run it, due to the old library version not being installed. The correct solution to this is to add the specific (old) version of the library to the dependencies (in the ebuild) of the (binary) package that uses it. This will prevent an upgrade that uninstalls the old library version. Sometimes the maintainer of the library will add a slotted version of it, so that non-binary users of it do not have to use the outdated version. If the binary package is not an ebuild, you can manually add the newer library version to package.mask, or make sure that the slot for the older version is installed if the library is slotted. Better yet (in all cases), get a more recent version of the binary package that is built against the newer version of the library. Complain to the vendor if none is available :-) The preserve-libs feature in portage is intended to let things keep on working short-term for source-distributed packages. In that case, the currently installed program is linked against the old library version, and when the program is rebuilt (with @preserved-rebuild) it will be linked against the newer version. -- eroen signature.asc Description: PGP signature
[gentoo-user] Re: new printer : any thoughts ?
On Tue, 10 Dec 2013 21:01:29 -0500 Philip Webb purs...@ca.inter.net wrote: My ancient printer's ink cartridge has finally dried up the mobo in my regular computer accepts only USB. I don't do much printing, but occasionally need a few pages. The local store has an HP Deskjet 2510 on sale this week. Does anyone have thoughts or suggestions ? I got a cheap (bw) HP LaserJet p1005 half a decade ago. It works well with the drivers from http://foo2zjs.rkkda.com/ , which lets me avoid using the bloated hplip mess (I tried that when I got the printer, it didn't work with network printing and required a running X session). It looks like the printer you mentioned only works with hplip. If hplip is acceptable to you, that is great, otherwise I would suggest looking for a different model that is either natively supported by cups or supported by the foo2X drivers or other similarly non-intrusive drivers. -- eroen signature.asc Description: PGP signature
[gentoo-user] Re: Any good way to pick global USE flag alternatives?
On Mon, 09 Dec 2013 08:25:19 +0200 Nikos Chantziaras rea...@gmail.com wrote: There are some global USE flags that allow users to pick alternative methods of implementing the same thing. For example, packages that offer a GUI might do so through the qt or gtk USE flag. Or audio support, where you can choose alsa or pulseaudio. Now, if I wanted to, for example, always choose PulseAudio support instead of ALSA and a Qt GUI instead of a Gtk+ one, one would think that I could just do: USE=pulseaudio qt -alsa -gtk in my make.conf. But, that doesn't work. That's because some packages don't offer an alternative at all. Some package might only support ALSA for audio, and disabling that would lead to the package not offering any audio output at all. Or it might only offer a Gtk+ GUI, not a Qt one, and disabling the gtk flag will make the package not building its GUI component at all. So what's needed here, is a way to tell Portage to only disable a global USE flag for packages that also offer another one, specified by the user. Like this pseudo make.conf syntax: USE=pulseaudio pulseaudio?(-alsa) qt qt?(-gtk) Is something like this already possible? Right now, the only way to painstakingly go through every single package that comes up on an 'USE=use_flags_here emerge -pDN @world' and insert them individually into package.use. There should be a proper way of doing this. I like this idea, I think it could be especially useful in the context of PYTHON_SINGLE_TARGET. I'm not sure I like all the details in the suggestion, though. In particular, the flags you mention have widely varying meaning throughout the tree, for example mkvtoolnix has wxwidgets and qt4 use flags that enable independent and completely different things. Perhaps it would be better to implement this as a new analogue to USE_EXPAND always expands to the first item in the user configuration that is also in IUSE for the package? This would then exclusively be used for pick one of these options style settings, like the gui of *some* packages, TLS implemetation, PYTHON_SINGLE_TARGET and so on. I think this could be preferable to the many packages that currently just arbitrarily picks one choice when two mutually exclusive flags are enabled. It might be necessary to expand this with some form of light masking for ebuilds that want to *strongly* suggest one choice over another. Example: User sets (in make.conf): TLS_IMPL_PREF=openssl gnutls Ebuild with IUSE=tls_impl_pref_gnutls tls_impl_pref_openssl sees tls_impl_pref_gnutls as disabled and tls_impl_pref_openssl as enabled. Ebuild with IUSE=tls_impl_pref_gnutls sees tls_impl_pref_gnutls as enabled. -- eroen signature.asc Description: PGP signature