Index: head/en_US.ISO8859-1/books/handbook/ports/chapter.xml =================================================================== --- head/en_US.ISO8859-1/books/handbook/ports/chapter.xml (revision 47472) +++ head/en_US.ISO8859-1/books/handbook/ports/chapter.xml (revision 47473) @@ -1,1902 +1,1902 @@ Installing Applications: Packages and Ports Synopsis ports packages &os; is bundled with a rich collection of system tools as part of the base system. In addition, &os; provides two complementary technologies for installing third-party software: the &os; Ports Collection, for installing from source, and packages, for installing from pre-built binaries. Either method may be used to install software from local media or from the network. After reading this chapter, you will know: The difference between binary packages and ports. How to find third-party software that has been ported to &os;. How to manage binary packages using pkg. How to build third-party software from source using the Ports Collection. How to find the files installed with the application for post-installation configuration. What to do if a software installation fails. Overview of Software Installation The typical steps for installing third-party software on a &unix; system include: Find and download the software, which might be distributed in source code format or as a binary. Unpack the software from its distribution format. This is typically a tarball compressed with &man.compress.1;, &man.gzip.1;, or &man.bzip2.1;. Locate the documentation in INSTALL, README or some file in a doc/ subdirectory and read up on how to install the software. If the software was distributed in source format, compile it. This may involve editing a Makefile or running a configure script. Test and install the software. If the software package was not deliberately ported, or tested to work, on &os;, the source code may need editing in order for it to install and run properly. At the time of this writing, over &os.numports; third-party applications have been ported to &os;. A &os; package contains pre-compiled copies of all the commands for an application, as well as any configuration files and documentation. A package can be manipulated with the pkg commands, such as pkg install. A &os; port is a collection of files designed to automate the process of compiling an application from source code. The files that comprise a port contain all the necessary information to automatically download, extract, patch, compile, and install the application. The ports system can also be used to generate packages which can be manipulated with the &os; package management commands. Both packages and ports understand dependencies. If a package or port is used to install an application and a dependent library is not already installed, the library will automatically be installed first. While the two technologies are similar, packages and ports each have their own strengths. Select the technology that meets your requirements for installing a particular application. Package Benefits A compressed package tarball is typically smaller than the compressed tarball containing the source code for the application. Packages do not require compilation time. For large applications, such as Mozilla, KDE, or GNOME, this can be important on a slow system. Packages do not require any understanding of the process involved in compiling software on &os;. Port Benefits Packages are normally compiled with conservative options because they have to run on the maximum number of systems. By compiling from the port, one can change the compilation options. Some applications have compile-time options relating to which features are installed. For example, Apache can be configured with a wide variety of different built-in options. In some cases, multiple packages will exist for the same application to specify certain settings. For example, Ghostscript is available as a ghostscript package and a ghostscript-nox11 package, depending on whether or not Xorg is installed. Creating multiple packages rapidly becomes impossible if an application has more than one or two different compile-time options. The licensing conditions of some software forbid binary distribution. Such software must be distributed as source code which must be compiled by the end-user. Some people do not trust binary distributions or prefer to read through source code in order to look for potential problems. Source code is needed in order to apply custom patches. To keep track of updated ports, subscribe to the &a.ports; and the &a.ports-bugs;. Before installing any application, check for security issues related to the application or type pkg audit -F to check all installed applications for known vulnerabilities. The remainder of this chapter explains how to use packages and ports to install and manage third-party software on &os;. Finding Software &os;'s list of available applications is growing all the time. There are a number of ways to find software to install: The &os; web site maintains an up-to-date searchable list of all the available applications, at http://www.FreeBSD.org/ports/. The ports can be searched by application name or by software category. FreshPorts Dan Langille maintains FreshPorts.org which provides a comprehensive search utility and also tracks changes to the applications in the Ports Collection. Registered users can create a customized watch list in order to receive an automated email when their watched ports are updated. SourceForge If finding a particular application becomes challenging, try searching a site like SourceForge.net or GitHub.com then check back at the &os; site to see if the application has been ported. pkg search To search the binary package repository for an application: &prompt.root; pkg search subversion git-subversion-1.9.2 java-subversion-1.8.8_2 p5-subversion-1.8.8_2 py27-hgsubversion-1.6 py27-subversion-1.8.8_2 ruby-subversion-1.8.8_2 subversion-1.8.8_2 subversion-book-4515 subversion-static-1.8.8_2 subversion16-1.6.23_4 subversion17-1.7.16_2 Package names include the version number and in case of ports based on python, the version number of the version of python the package was built with. Some ports also have multiple versions available. In case of subversion there are different versions available, as well as different compile options. In this case, the staticly linked version of subversion. When indicating which package to install, it is best to specify the application by the port origin, which is the path in the ports tree. Repeat the pkg search with to list the origin of each package: &prompt.root; pkg search -o subversion devel/git-subversion java/java-subversion devel/p5-subversion devel/py-hgsubversion devel/py-subversion devel/ruby-subversion devel/subversion16 devel/subversion17 devel/subversion devel/subversion-book devel/subversion-static Searching by shell globs, regular expressions, exact match, by description, or any other field in the repository database is also supported by pkg search. After installing ports-mgmt/pkg or ports-mgmt/pkg-devel, see &man.pkg-search.8; for more details. If the Ports Collection is already installed, there are several methods to query the local version of the ports tree. To find out which category a port is in, type whereis file, where file is the program to be installed: &prompt.root; whereis lsof lsof: /usr/ports/sysutils/lsof Alternately, an &man.echo.1; statement can be used: &prompt.root; echo /usr/ports/*/*lsof* /usr/ports/sysutils/lsof Note that this will also return any matched files downloaded into the /usr/ports/distfiles directory. Another way to find software is by using the Ports Collection's built-in search mechanism. To use the search feature, cd to /usr/ports then run make search name=program-name where program-name is the name of the software. For example, to search for lsof: &prompt.root; cd /usr/ports &prompt.root; make search name=lsof Port: lsof-4.88.d,8 Path: /usr/ports/sysutils/lsof Info: Lists information about open files (similar to fstat(1)) Maint: ler@lerctr.org Index: sysutils B-deps: R-deps: The built-in search mechanism uses a file of index information. If a message indicates that the INDEX is required, run make fetchindex to download the current index file. With the INDEX present, make search will be able to perform the requested search. The Path: line indicates where to find the port. To receive less information, use the quicksearch feature: &prompt.root; cd /usr/ports &prompt.root; make quicksearch name=lsof Port: lsof-4.88.d,8 Path: /usr/ports/sysutils/lsof Info: Lists information about open files (similar to fstat(1)) For more in-depth searching, use make search key=string or make quicksearch key=string, where string is some text to search for. The text can be in comments, descriptions, or dependencies in order to find ports which relate to a particular subject when the name of the program is unknown. When using search or quicksearch, the search string is case-insensitive. Searching for LSOF will yield the same results as searching for lsof. Using <application>pkg</application> for Binary Package Management pkg is the next generation replacement for the traditional &os; package management tools, offering many features that make dealing with binary packages faster and easier. pkg is not a replacement for port management tools like ports-mgmt/portmaster or ports-mgmt/portupgrade. These tools can be used to install third-party software from both binary packages and the Ports Collection, while pkg installs only binary packages. Getting Started with <application>pkg</application> &os; 8.4 and later includes a bootstrap utility which can be used to download and install pkg, along with its manual pages. To bootstrap the system, run: &prompt.root; /usr/sbin/pkg For earlier &os; versions, pkg must instead be installed from the Ports Collection or as a binary package. To install the port, run: &prompt.root; cd /usr/ports/ports-mgmt/pkg &prompt.root; make &prompt.root; make install clean When upgrading an existing system that originally used the older package system, the database must be converted to the new format, so that the new tools are aware of the already installed packages. Once pkg has been installed, the package database must be converted from the traditional format to the new format by running this command: &prompt.root; pkg2ng This step is not required for new installations that do not yet have any third-party software installed. This step is not reversible. Once the package database has been converted to the pkg format, the traditional pkg_* tools should no longer be used. The package database conversion may emit errors as the contents are converted to the new version. Generally, these errors can be safely ignored. However, a list of third-party software that was not successfully converted will be listed after pkg2ng has finished and these applications must be manually reinstalled. To ensure that the &os; Ports Collection registers new software with pkg, and not the traditional packages format, &os; versions earlier than 10.X require this line in /etc/make.conf: WITH_PKGNG= yes By default pkg uses the &os; package mirrors. For information about building a custom package repository, see Additional pkg configuration options are described in &man.pkg.conf.5;. Usage information for pkg is available in the &man.pkg.8; manpage or by running pkg without additional arguments. Each pkg command argument is documented in a command-specific manual page. To read the manual page for pkg install, for example, run either of these commands: &prompt.root; pkg help install &prompt.root; man pkg-install The rest of this section demonstrates common binary package management tasks which can be performed using pkg. Each demonstrated command provides many switches to customize its use. Refer to a command's help or man page for details and more examples. Obtaining Information About Installed Packages Information about the packages installed on a system can be viewed by running pkg info which, when run without any switches, will list the package version for either all installed packages or the specified package. For example, to see which version of pkg is installed, run: &prompt.root; pkg info pkg pkg-1.1.4_1 Installing and Removing Packages To install a binary package use the following command, where packagename is the name of the package to install: &prompt.root; pkg install packagename This command uses repository data to determine which version of the software to install and if it has any uninstalled dependencies. For example, to install curl: &prompt.root; pkg install curl Updating repository catalogue /usr/local/tmp/All/curl-7.31.0_1.txz 100% of 1181 kB 1380 kBps 00m01s /usr/local/tmp/All/ca_root_nss-3.15.1_1.txz 100% of 288 kB 1700 kBps 00m00s Updating repository catalogue The following 2 packages will be installed: Installing ca_root_nss: 3.15.1_1 Installing curl: 7.31.0_1 The installation will require 3 MB more space 0 B to be downloaded Proceed with installing packages [y/N]: y Checking integrity... done [1/2] Installing ca_root_nss-3.15.5_1... done [2/2] Installing curl-7.31.0_1... done Cleaning up cache files...Done The new package and any additional packages that were installed as dependencies can be seen in the installed packages list: &prompt.root; pkg info ca_root_nss-3.15.5_1 The root certificate bundle from the Mozilla Project curl-7.31.0_1 Non-interactive tool to get files from FTP, GOPHER, HTTP(S) servers pkg-1.1.4_6 New generation package manager Packages that are no longer needed can be removed with pkg delete. For example: &prompt.root; pkg delete curl The following packages will be deleted: curl-7.31.0_1 The deletion will free 3 MB Proceed with deleting packages [y/N]: y [1/1] Deleting curl-7.31.0_1... done Upgrading Installed Packages Installed packages can be upgraded to their latest versions by running: &prompt.root; pkg upgrade This command will compare the installed versions with those available in the repository catalogue and upgrade them from the repository. Auditing Installed Packages Occasionally, software vulnerabilities may be discovered in third-party applications. To address this, pkg includes a built-in auditing mechanism. To determine if there are any known vulnerabilities for the software installed on the system, run: &prompt.root; pkg audit -F Automatically Removing Leaf Dependencies Removing a package may leave behind dependencies which are no longer required. Unneeded packages that were installed as dependencies can be automatically detected and removed using: &prompt.root; pkg autoremove Packages to be autoremoved: ca_root_nss-3.13.5 The autoremoval will free 723 kB Proceed with autoremoval of packages [y/N]: y Deinstalling ca_root_nss-3.15.1_1... done Restoring the Package Database Unlike the traditional package management system, pkg includes its own package database backup mechanism. This functionality is enabled by default. To disable the periodic script from backing up the package database, set daily_backup_pkgdb_enable="NO" in &man.periodic.conf.5;. To restore the contents of a previous package database backup, run the following command replacing /path/to/pkg.sql with the location of the backup: &prompt.root; pkg backup -r /path/to/pkg.sql If restoring a backup taken by the periodic script, it must be decompressed prior to being restored. To run a manual backup of the pkg database, run the following command, replacing /path/to/pkg.sql with a suitable file name and location: &prompt.root; pkg backup -d /path/to/pkg.sql Removing Stale Packages By default, pkg stores binary packages in a cache directory defined by PKG_CACHEDIR in &man.pkg.conf.5;. Only copies of the latest installed packages are kept. Older versions of pkg kept all previous packages. To remove these outdated binary packages, run: &prompt.root; pkg clean The entire cache may be cleared by running: &prompt.root; pkg clean -a Modifying Package Metadata Software within the &os; Ports Collection can undergo major version number changes. To address this, pkg has a built-in command to update package origins. This can be useful, for example, if lang/php5 is renamed to lang/php53 so that lang/php5 can now represent version 5.4. To change the package origin for the above example, run: &prompt.root; pkg set -o lang/php5:lang/php53 As another example, to update lang/ruby18 to lang/ruby19, run: &prompt.root; pkg set -o lang/ruby18:lang/ruby19 As a final example, to change the origin of the libglut shared libraries from graphics/libglut to graphics/freeglut, run: &prompt.root; pkg set -o graphics/libglut:graphics/freeglut When changing package origins, it is important to reinstall packages that are dependent on the package with the modified origin. To force a reinstallation of dependent packages, run: &prompt.root; pkg install -Rf graphics/freeglut Using the Ports Collection The Ports Collection is a set of Makefiles, patches, and description files stored in /usr/ports. This set of files is used to compile and install applications on &os;. Before an application can be compiled using a port, the Ports Collection must first be installed. If it was not installed during the installation of &os;, use one of the following methods to install it: Portsnap Method The base system of &os; includes Portsnap. This is a fast and user-friendly tool for retrieving the Ports Collection and is the recommended choice for most users. This utility connects to a &os; site, verifies the secure key, and downloads a new copy of the Ports Collection. The key is used to verify the integrity of all downloaded files. To download a compressed snapshot of the Ports Collection into /var/db/portsnap: &prompt.root; portsnap fetch When running Portsnap for the first time, extract the snapshot into /usr/ports: &prompt.root; portsnap extract After the first use of Portsnap has been completed as shown above, /usr/ports can be updated as needed by running: &prompt.root; portsnap fetch &prompt.root; portsnap update When using fetch, the extract or the update operation may be run consecutively, like so: &prompt.root; portsnap fetch update Subversion Method If more control over the ports tree is needed or if local changes need to be maintained, Subversion can be used to obtain the Ports Collection. Refer to the Subversion Primer for a detailed description of Subversion. Subversion must be installed before it can be used to check out the ports tree. If a copy of the ports tree is already present, install Subversion like this: &prompt.root; cd /usr/ports/devel/subversion &prompt.root; make install clean If the ports tree is not available, or pkg is being used to manage packages, Subversion can be installed as a package: &prompt.root; pkg install subversion Check out a copy of the ports tree. For better performance, replace svn0.us-east.FreeBSD.org with a Subversion mirror close to your geographic location: &prompt.root; svn checkout https://svn0.us-east.FreeBSD.org/ports/head /usr/ports As needed, update /usr/ports after the initial Subversion checkout: &prompt.root; svn update /usr/ports The Ports Collection installs a series of directories representing software categories with each category having a subdirectory for each application. Each subdirectory, also referred to as a ports skeleton, contains a set of files that tell &os; how to compile and install that program. Each port skeleton includes these files and directories: Makefile: contains statements that specify how the application should be compiled and where its components should be installed. distinfo: contains the names and checksums of the files that must be downloaded to build the port. files/: this directory contains any patches needed for the program to compile and install on &os;. This directory may also contain other files used to build the port. pkg-descr: provides a more detailed description of the program. pkg-plist: a list of all the files that will be installed by the port. It also tells the ports system which files to remove upon deinstallation. Some ports include pkg-message or other files to handle special situations. For more details on these files, and on ports in general, refer to the &os; Porter's Handbook. The port does not include the actual source code, also known as a distfile. The extract portion of building a port will automatically save the downloaded source to /usr/ports/distfiles. Installing Ports ports installing This section provides basic instructions on using the Ports Collection to install or remove software. The detailed description of available make targets and environment variables is available in &man.ports.7;. Before compiling any port, be sure to update the Ports Collection as described in the previous section. Since the installation of any third-party software can introduce security vulnerabilities, it is recommended to first check for known security issues related to the port. Alternately, run pkg audit -F before installing a new port. This command can be configured to automatically perform a security audit and an update of the vulnerability database during the daily security system check. For more information, refer to &man.pkg-audit.8; and &man.periodic.8;. Using the Ports Collection assumes a working Internet connection. It also requires superuser privilege. To compile and install the port, change to the directory of the port to be installed, then type make install at the prompt. Messages will indicate the progress: &prompt.root; cd /usr/ports/sysutils/lsof &prompt.root; make install >> lsof_4.88D.freebsd.tar.gz doesn't seem to exist in /usr/ports/distfiles/. >> Attempting to fetch from ftp://lsof.itap.purdue.edu/pub/tools/unix/lsof/. ===> Extracting for lsof-4.88 ... [extraction output snipped] ... >> Checksum OK for lsof_4.88D.freebsd.tar.gz. ===> Patching for lsof-4.88.d,8 ===> Applying FreeBSD patches for lsof-4.88.d,8 ===> Configuring for lsof-4.88.d,8 ... [configure output snipped] ... ===> Building for lsof-4.88.d,8 ... [compilation output snipped] ... ===> Installing for lsof-4.88.d,8 ... [installation output snipped] ... ===> Generating temporary packing list ===> Compressing manual pages for lsof-4.88.d,8 ===> Registering installation for lsof-4.88.d,8 ===> SECURITY NOTE: This port has installed the following binaries which execute with increased privileges. /usr/local/sbin/lsof &prompt.root; Since lsof is a program that runs with increased privileges, a security warning is displayed as it is installed. Once the installation is complete, the prompt will be returned. Some shells keep a cache of the commands that are available in the directories listed in the PATH environment variable, to speed up lookup operations for the executable file of these commands. Users of the tcsh shell should type rehash so that a newly installed command can be used without specifying its full path. Use hash -r instead for the sh shell. Refer to the documentation for the shell for more information. During installation, a working subdirectory is created which contains all the temporary files used during compilation. Removing this directory saves disk space and minimizes the chance of problems later when upgrading to the newer version of the port: &prompt.root; make clean ===> Cleaning for lsof-88.d,8 &prompt.root; To save this extra step, instead use make install clean when compiling the port. Customizing Ports Installation Some ports provide build options which can be used to enable or disable application components, provide security options, or allow for other customizations. Examples include www/firefox, security/gpgme, and mail/sylpheed-claws. If the port depends upon other ports which have configurable options, it may pause several times for user interaction as the default behavior is to prompt the user to select options from a menu. To avoid this, run make config-recursive within the port skeleton to do this configuration in one batch. Then, run make install [clean] to compile and install the port. When using config-recursive, the list of ports to configure are gathered by the all-depends-list target. It is recommended to run make config-recursive until all dependent ports options have been defined, and ports options screens no longer appear, to be certain that all dependency options have been configured. There are several ways to revisit a port's build options menu in order to add, remove, or change these options after a port has been built. One method is to cd into the directory containing the port and type make config. Another option is to use make showconfig. Another option is to execute make rmconfig which will remove all selected options and allow you to start over. All of these options, and others, are explained in great detail in &man.ports.7;. The ports system uses &man.fetch.1; to download the source files, which supports various environment variables. The FTP_PASSIVE_MODE, FTP_PROXY, and FTP_PASSWORD variables may need to be set if the &os; system is behind a firewall or FTP/HTTP proxy. See &man.fetch.3; for the complete list of supported variables. For users who cannot be connected to the Internet all the time, make fetch can be run within /usr/ports, to fetch all distfiles, or within a category, such as /usr/ports/net, or within the specific port skeleton. Note that if a port has any dependencies, running this command in a category or ports skeleton will not fetch the distfiles of ports from another category. Instead, use make fetch-recursive to also fetch the distfiles for all the dependencies of a port. In rare cases, such as when an organization has a local distfiles repository, the MASTER_SITES variable can be used to override the download locations specified in the Makefile. When using, specify the alternate location: &prompt.root; cd /usr/ports/directory &prompt.root; make MASTER_SITE_OVERRIDE= \ ftp://ftp.organization.org/pub/FreeBSD/ports/distfiles/ fetch The WRKDIRPREFIX and PREFIX variables can override the default working and target directories. For example: &prompt.root; make WRKDIRPREFIX=/usr/home/example/ports install will compile the port in /usr/home/example/ports and install everything under /usr/local. &prompt.root; make PREFIX=/usr/home/example/local install will compile the port in /usr/ports and install it in /usr/home/example/local. And: &prompt.root; make WRKDIRPREFIX=../ports PREFIX=../local install will combine the two. These can also be set as environmental variables. Refer to the manual page for your shell for instructions on how to set an environmental variable. Removing Installed Ports ports removing Installed ports can be uninstalled using pkg delete. Examples for using this command can be found in the &man.pkg-delete.8; manpage. Alternately, make deinstall can be run in the port's directory: &prompt.root; cd /usr/ports/sysutils/lsof make deinstall ===> Deinstalling for sysutils/lsof ===> Deinstalling Deinstallation has been requested for the following 1 packages: lsof-4.88.d,8 The deinstallation will free 229 kB [1/1] Deleting lsof-4.88.d,8... done It is recommended to read the messages as the port is uninstalled. If the port has any applications that depend upon it, this information will be displayed but the uninstallation will proceed. In such cases, it may be better to reinstall the application in order to prevent broken dependencies. Upgrading Ports ports upgrading Over time, newer versions of software become available in the Ports Collection. This section describes how to determine which software can be upgraded and how to perform the upgrade. To determine if newer versions of installed ports are available, ensure that the latest version of the ports tree is installed, using the updating command described in either or . On &os; 10 and later, or if the system has been converted to pkg, the following command will list the installed ports which are out of date: &prompt.root; pkg version -l "<" For &os; 9.X and lower, the following command will list the installed ports that are out of date: &prompt.root; pkg_version -l "<" Before attempting an upgrade, read /usr/ports/UPDATING from the top of the file to the date closest to the last time ports were upgraded or the system was installed. This file describes various issues and additional steps users may encounter and need to perform when updating a port, including such things as file format changes, changes in locations of configuration files, or any incompatibilities with previous versions. Make note of any instructions which match any of the ports that need upgrading and follow these instructions when performing the upgrade. To perform the actual upgrade, use either Portmaster or Portupgrade. Upgrading Ports Using <application>Portmaster</application> portmaster The ports-mgmt/portmaster package or port is the recommended tool for upgrading installed ports as it is designed to use the tools installed with &os; without depending upon other ports. It uses the information in /var/db/pkg/ to determine which ports to upgrade. To install this utility as a port: &prompt.root; cd /usr/ports/ports-mgmt/portmaster &prompt.root; make install clean Portmaster defines four categories of ports: Root port: has no dependencies and is not a dependency of any other ports. Trunk port: has no dependencies, but other ports depend upon it. Branch port: has dependencies and other ports depend upon it. Leaf port: has dependencies but no other ports depend upon it. To list these categories and search for updates: &prompt.root; portmaster -L ===>>> Root ports (No dependencies, not depended on) ===>>> ispell-3.2.06_18 ===>>> screen-4.0.3 ===>>> New version available: screen-4.0.3_1 ===>>> tcpflow-0.21_1 ===>>> 7 root ports ... ===>>> Branch ports (Have dependencies, are depended on) ===>>> apache22-2.2.3 ===>>> New version available: apache22-2.2.8 ... ===>>> Leaf ports (Have dependencies, not depended on) ===>>> automake-1.9.6_2 ===>>> bash-3.1.17 ===>>> New version available: bash-3.2.33 ... ===>>> 32 leaf ports ===>>> 137 total installed ports ===>>> 83 have new versions available This command is used to upgrade all outdated ports: &prompt.root; portmaster -a By default, Portmaster will make a backup package before deleting the existing port. If the installation of the new version is successful, Portmaster will delete the backup. Using will instruct Portmaster not to automatically delete the backup. Adding will start Portmaster in interactive mode, prompting for confirmation before upgrading each port. Many other options are available. Read through the manual page for portmaster(8) for details regarding their usage. If errors are encountered during the upgrade process, add to upgrade and rebuild all ports: &prompt.root; portmaster -af Portmaster can also be used to install new ports on the system, upgrading all dependencies before building and installing the new port. To use this function, specify the location of the port in the Ports Collection: &prompt.root; portmaster shells/bash Upgrading Ports Using Portupgrade portupgrade Another utility that can be used to upgrade ports is Portupgrade, which is available as the ports-mgmt/portupgrade package or port. This utility installs a suite of applications which can be used to manage ports. However, it is dependent upon Ruby. To install the port: &prompt.root; cd /usr/ports/ports-mgmt/portupgrade &prompt.root; make install clean Before performing an upgrade using this utility, it is recommended to scan the list of installed ports using pkgdb -F and to fix all the inconsistencies it reports. To upgrade all the outdated ports installed on the system, use portupgrade -a. Alternately, include to be asked for confirmation of every individual upgrade: &prompt.root; portupgrade -ai To upgrade only a specified application instead of all available ports, use portupgrade pkgname. It is very important to include to first upgrade all the ports required by the given application: &prompt.root; portupgrade -R firefox If is included, Portupgrade searches for available packages in the local directories listed in PKG_PATH. If none are available locally, it then fetches packages from a remote site. If packages can not be found locally or fetched remotely, Portupgrade will use ports. To avoid using ports entirely, specify . This last set of options tells Portupgrade to abort if no packages are available: - &prompt.root; portupgrade -PP gnome2 + &prompt.root; portupgrade -PP gnome3 To just fetch the port distfiles, or packages, if is specified, without building or installing anything, use . For further information on all of the available switches, refer to the manual page for portupgrade. Ports and Disk Space ports disk-space Using the Ports Collection will use up disk space over time. After building and installing a port, running make clean within the ports skeleton will clean up the temporary work directory. If Portmaster is used to install a port, it will automatically remove this directory unless is specified. If Portupgrade is installed, this command will remove all work directories found within the local copy of the Ports Collection: &prompt.root; portsclean -C In addition, a lot of out-dated source distribution files will collect in /usr/ports/distfiles over time. If Portupgrade is installed, this command will delete all the distfiles that are no longer referenced by any ports: &prompt.root; portsclean -D To use Portupgrade to remove all distfiles not referenced by any port currently installed on the system: &prompt.root; portsclean -DD If Portmaster is installed, use: &prompt.root; portmaster --clean-distfiles By default, this command is interactive and will prompt the user to confirm if a distfile should be deleted. In addition to these commands, the ports-mgmt/pkg_cutleaves package or port automates the task of removing installed ports that are no longer needed. Building Packages with <application>Poudriere</application> Poudriere is a BSD-licensed utility for creating and testing &os; packages. It uses &os; jails to set up isolated compilation environments. These jails can be used to build packages for versions of &os; that are different from the system on which it is installed, and also to build packages for i386 if the host is an &arch.amd64; system. Once the packages are built, they are in a layout identical to the official mirrors. These packages are usable by &man.pkg.8; and other package management tools. Poudriere is installed using the ports-mgmt/poudriere package or port. The installation includes a sample configuration file /usr/local/etc/poudriere.conf.sample. Copy this file to /usr/local/etc/poudriere.conf. Edit the copied file to suit the local configuration. While ZFS is not required on the system running poudriere, it is beneficial. When ZFS is used, ZPOOL must be specified in /usr/local/etc/poudriere.conf and FREEBSD_HOST should be set to a nearby mirror. Defining CCACHE_DIR enables the use of devel/ccache to cache compilation and reduce build times for frequently-compiled code. It may be convenient to put poudriere datasets in an isolated tree mounted at /poudriere. Defaults for the other configuration values are adequate. The number of processor cores detected is used to define how many builds should run in parallel. Supply enough virtual memory, either with RAM or swap space. If virtual memory runs out, compiling jails will stop and be torn down, resulting in weird error messages. Initialize Jails and Port Trees After configuration, initialize poudriere so that it installs a jail with the required &os; tree and a ports tree. Specify a name for the jail using and the &os; version with . On systems running &os;/&arch.amd64;, the architecture can be set with to either i386 or amd64. The default is the architecture shown by uname. &prompt.root; poudriere jail -c -j 10amd64 -v 10.0-RELEASE ====>> Creating 10amd64 fs... done ====>> Fetching base.txz for FreeBSD 10.0-RELEASE amd64 /poudriere/jails/10amd64/fromftp/base.txz 100% of 59 MB 1470 kBps 00m42s ====>> Extracting base.txz... done ====>> Fetching src.txz for FreeBSD 10.0-RELEASE amd64 /poudriere/jails/10amd64/fromftp/src.txz 100% of 107 MB 1476 kBps 01m14s ====>> Extracting src.txz... done ====>> Fetching games.txz for FreeBSD 10.0-RELEASE amd64 /poudriere/jails/10amd64/fromftp/games.txz 100% of 865 kB 734 kBps 00m01s ====>> Extracting games.txz... done ====>> Fetching lib32.txz for FreeBSD 10.0-RELEASE amd64 /poudriere/jails/10amd64/fromftp/lib32.txz 100% of 14 MB 1316 kBps 00m12s ====>> Extracting lib32.txz... done ====>> Cleaning up... done ====>> Jail 10amd64 10.0-RELEASE amd64 is ready to be used &prompt.root; poudriere ports -c -p local ====>> Creating local fs... done ====>> Extracting portstree "local"... Looking up portsnap.FreeBSD.org mirrors... 7 mirrors found. Fetching public key from ec2-eu-west-1.portsnap.freebsd.org... done. Fetching snapshot tag from ec2-eu-west-1.portsnap.freebsd.org... done. Fetching snapshot metadata... done. Fetching snapshot generated at Tue Feb 11 01:07:15 CET 2014: 94a3431f0ce567f6452ffde4fd3d7d3c6e1da143efec76100% of 69 MB 1246 kBps 00m57s Extracting snapshot... done. Verifying snapshot integrity... done. Fetching snapshot tag from ec2-eu-west-1.portsnap.freebsd.org... done. Fetching snapshot metadata... done. Updating from Tue Feb 11 01:07:15 CET 2014 to Tue Feb 11 16:05:20 CET 2014. Fetching 4 metadata patches... done. Applying metadata patches... done. Fetching 0 metadata files... done. Fetching 48 patches. (48/48) 100.00% done. done. Applying patches... done. Fetching 1 new ports or files... done. /poudriere/ports/tester/CHANGES /poudriere/ports/tester/COPYRIGHT [...] Building new INDEX files... done. On a single computer, poudriere can build ports with multiple configurations, in multiple jails, and from different port trees. Custom configurations for these combinations are called sets. See the CUSTOMIZATION section of &man.poudriere.8; for details after ports-mgmt/poudriere or ports-mgmt/poudriere-devel is installed. The basic configuration shown here puts a single jail-, port-, and set-specific make.conf in /usr/local/etc/poudriere.d. The filename in this example is created by combining the jail name, port name, and set name: 10amd64-local-workstation-make.conf. The system make.conf and this new file are combined at build time to create the make.conf used by the build jail. Packages to be built are entered in 10amd64-local-workstation-pkglist: editors/emacs devel/git ports-mgmt/pkg ... Options and dependencies for the specified ports are configured: &prompt.root; poudriere options -j 10amd64 -p local -z workstation -f 10amd64-local-workstation-pkglist Finally, packages are built and a package repository is created: &prompt.root; poudriere bulk -j 10amd64 -p local -z workstation -f 10amd64-local-workstation-pkglist Ctrlt displays the current state of the build. Poudriere also builds files in /poudriere/logs/bulk/jailname that can be used with a web server to display build information. Packages are now available for installation from the poudriere repository. For more information on using poudriere, see &man.poudriere.8; and the main web site, . Configuring pkg Clients to Use a Poudriere Repository While it is possible to use both a custom repository along side of the official repository, sometimes it is useful to disable the official repository. This is done by creating a configuration file that overrides and disables the official configuration file. Create /usr/local/etc/pkg/repos/FreeBSD.conf that contains the following: FreeBSD: { enabled: no } Usually it is easiest to serve a poudriere repository to the client machines via HTTP. Setup a webserver to serve up the package directory, usually something like: /usr/local/poudriere/data/packages/10amd64. Where 10amd64 is the name of the build. If the URL to the package repository is: http://pkg.example.com/10amd64, then the repository configuration file in /usr/local/etc/pkg/repos/custom.conf would look like: custom: { url: "http://pkg.example.com/10amd64", mirror_type: "http", enabled: yes, } Post-Installation Considerations Regardless of whether the software was installed from a binary package or port, most third-party applications require some level of configuration after installation. The following commands and locations can be used to help determine what was installed with the application. Most applications install at least one default configuration file in /usr/local/etc. In the case where an application has a large number of configuration files, a subdirectory will be created to hold them. Often, sample configuration files are installed which end with a suffix such as .sample. The configuration files should be reviewed and possibly edited to meet the system's needs. To edit a sample file, first copy it without the .sample extension. Applications which provide documentation will install it into /usr/local/share/doc and many applications also install manual pages. This documentation should be consulted before continuing. Some applications run services which must be added to /etc/rc.conf before starting the application. These applications usually install a startup script in /usr/local/etc/rc.d. See Starting Services for more information. Users of &man.csh.1; should run rehash to rebuild the known binary list in the shells PATH. Use pkg info to determine which files, man pages, and binaries were installed with the application. Dealing with Broken Ports When a port does not build or install, try the following: Search to see if there is a fix pending for the port in the Problem Report database. If so, implementing the proposed fix may fix the issue. Ask the maintainer of the port for help. Type make maintainer in the ports skeleton or read the port's Makefile to find the maintainer's email address. Remember to include the $FreeBSD: line from the port's Makefile and the output leading up to the error in the email to the maintainer. Some ports are not maintained by an individual but instead by a mailing list. Many, but not all, of these addresses look like freebsd-listname@FreeBSD.org. Take this into account when sending an email. In particular, ports shown as maintained by ports@FreeBSD.org are not maintained by a specific individual. Instead, any fixes and support come from the general community who subscribe to that mailing list. More volunteers are always needed! If there is no response to the email, use Bugzilla to submit a bug report using the instructions in Writing &os; Problem Reports. Fix it! The Porter's Handbook includes detailed information on the ports infrastructure so that you can fix the occasional broken port or even submit your own! Install the package instead of the port using the instructions in . Index: head/en_US.ISO8859-1/books/handbook/x11/chapter.xml =================================================================== --- head/en_US.ISO8859-1/books/handbook/x11/chapter.xml (revision 47472) +++ head/en_US.ISO8859-1/books/handbook/x11/chapter.xml (revision 47473) @@ -1,1628 +1,1628 @@ The X Window System Synopsis An installation of &os; using bsdinstall does not automatically install a graphical user interface. This chapter describes how to install and configure &xorg;, which provides the open source X Window System used to provide a graphical environment. It then describes how to find and install a desktop environment or window manager. Users who prefer an installation method that automatically configures the &xorg; and offers a choice of window managers during installation should refer to the pcbsd.org website. For more information on the video hardware that &xorg; supports, refer to the x.org website. After reading this chapter, you will know: The various components of the X Window System, and how they interoperate. How to install and configure &xorg;. How to install and configure several window managers and desktop environments. How to use &truetype; fonts in &xorg;. How to set up your system for graphical logins (XDM). Before reading this chapter, you should: Know how to install additional third-party software as described in . Terminology While it is not necessary to understand all of the details of the various components in the X Window System and how they interact, some basic knowledge of these components can be useful: X server X was designed from the beginning to be network-centric, and adopts a client-server model. In this model, the X server runs on the computer that has the keyboard, monitor, and mouse attached. The server's responsibility includes tasks such as managing the display, handling input from the keyboard and mouse, and handling input or output from other devices such as a tablet or a video projector. This confuses some people, because the X terminology is exactly backward to what they expect. They expect the X server to be the big powerful machine down the hall, and the X client to be the machine on their desk. X client Each X application, such as XTerm or Firefox, is a client. A client sends messages to the server such as Please draw a window at these coordinates, and the server sends back messages such as The user just clicked on the OK button. In a home or small office environment, the X server and the X clients commonly run on the same computer. It is also possible to run the X server on a less powerful computer and to run the X applications on a more powerful system. In this scenario, the communication between the X client and server takes place over the network. window manager X does not dictate what windows should look like on screen, how to move them around with the mouse, which keystrokes should be used to move between windows, what the title bars on each window should look like, whether or not they have close buttons on them, and so on. Instead, X delegates this responsibility to a separate window manager application. There are dozens of window managers available. Each window manager provides a different look and feel: some support virtual desktops, some allow customized keystrokes to manage the desktop, some have a Start button, and some are themeable, allowing a complete change of the desktop's look-and-feel. Window managers are available in the x11-wm category of the Ports Collection. Each window manager uses a different configuration mechanism. Some expect configuration file written by hand while others provide graphical tools for most configuration tasks. desktop environment KDE and GNOME are considered to be desktop environments as they include an entire suite of applications for performing common desktop tasks. These may include office suites, web browsers, and games. focus policy The window manager is responsible for the mouse focus policy. This policy provides some means for choosing which window is actively receiving keystrokes and it should also visibly indicate which window is currently active. One focus policy is called click-to-focus. In this model, a window becomes active upon receiving a mouse click. In the focus-follows-mouse policy, the window that is under the mouse pointer has focus and the focus is changed by pointing at another window. If the mouse is over the root window, then this window is focused. In the sloppy-focus model, if the mouse is moved over the root window, the most recently used window still has the focus. With sloppy-focus, focus is only changed when the cursor enters a new window, and not when exiting the current window. In the click-to-focus policy, the active window is selected by mouse click. The window may then be raised and appear in front of all other windows. All keystrokes will now be directed to this window, even if the cursor is moved to another window. Different window managers support different focus models. All of them support click-to-focus, and the majority of them also support other policies. Consult the documentation for the window manager to determine which focus models are available. widgets Widget is a term for all of the items in the user interface that can be clicked or manipulated in some way. This includes buttons, check boxes, radio buttons, icons, and lists. A widget toolkit is a set of widgets used to create graphical applications. There are several popular widget toolkits, including Qt, used by KDE, and GTK+, used by GNOME. As a result, applications will have a different look and feel, depending upon which widget toolkit was used to create the application. Installing <application>&xorg;</application> &xorg; is the implementation of the open source X Window System released by the X.Org Foundation. In &os;, it can be installed as a package or port. The meta-port for the complete distribution which includes X servers, clients, libraries, and fonts is located in x11/xorg. A minimal distribution is located in x11/xorg-minimal, with separate ports available for docs, libraries, and apps. The examples in this section install the complete &xorg; distribution. To build and install &xorg; from the Ports Collection: &prompt.root; cd /usr/ports/x11/xorg &prompt.root; make install clean To build &xorg; in its entirety, be sure to have at least 4 GB of free disk space available. Alternatively, &xorg; can be installed directly from packages with this command: &prompt.root; pkg install xorg <application>&xorg;</application> Configuration &xorg; &xorg; In most cases, &xorg; is self-configuring. Autoconfiguration is the preferred method, and should be tried first. When X is started without a configuration file, the video card and input devices are automatically detected and used. If the system already has a configuration file and you wish to create a new configuration, save a copy of the old one first: &prompt.root; mv /etc/X11/xorg.conf ~/xorg.conf.etc &prompt.root; mv /usr/local/etc/X11/xorg.conf ~/xorg.conf.localetc To generate the configuration file and start the X system, issue this command: &prompt.user; startx If a black and grey grid and an X mouse cursor appear, the configuration was successful. Test the system by moving the mouse and typing text into the windows. To exit the test, switch to the virtual console used to start it by pressing Ctrl Alt Fn (F1 for the first virtual console) and press Ctrl C . If the mouse or keyboard do not work, continue with . Desktop environments like GNOME, KDE or Xfce provide graphical tools to set parameters such as video resolution. If the default configuration works, skip to for examples on how to install a desktop environment. The Ctrl Alt Backspace key combination may also be used to break out of &xorg;. To enable it, you can either type the following command from any X terminal emulator: &prompt.user; setxkbmap -option terminate:ctrl_alt_bksp or create a keyboard configuration file for hald called x11-input.fdi and saved in the /usr/local/etc/hal/fdi/policy directory. This file should contain the following lines: <?xml version="1.0" encoding="iso-8859-1"?> <deviceinfo version="0.2"> <device> <match key="info.capabilities" contains="input.keyboard"> <merge key="input.x11_options.XkbOptions" type="string">terminate:ctrl_alt_bksp</merge> </match> </device> </deviceinfo> The following line will also have to be added to xorg.conf.new, in the ServerLayout or ServerFlags section: Option "DontZap" "off" Caveats The ability to configure optimal resolution is dependent upon the video hardware and the support provided by its driver. At this time, driver support includes: Intel: as of &os; 9.3 and &os; 10.1, 3D acceleration on most Intel graphics, including IronLake, SandyBridge, and IvyBridge, is supported. Support for switching between X and virtual consoles is provided by &man.vt.4;. ATI/Radeon: 2D and 3D acceleration is supported on most Radeon cards up to the HD6000 series. NVIDIA: several NVIDIA drivers are available in the x11 category of the Ports Collection. Install the driver that matches the video card. Optimus: currently there is no switching support between the two graphics adapters provided by Optimus. Optimus implementations vary, and &os; will not be able to drive all versions of the hardware. Some computers provide a BIOS option to disable one of the graphics adapters or select a discrete mode. Using Fonts in <application>&xorg;</application> Type1 Fonts The default fonts that ship with &xorg; are less than ideal for typical desktop publishing applications. Large presentation fonts show up jagged and unprofessional looking, and small fonts are almost completely unintelligible. However, there are several free, high quality Type1 (&postscript;) fonts available which can be readily used with &xorg;. For instance, the URW font collection (x11-fonts/urwfonts) includes high quality versions of standard type1 fonts (Times Roman, Helvetica, Palatino and others). The Freefonts collection (x11-fonts/freefonts) includes many more fonts, but most of them are intended for use in graphics software such as the Gimp, and are not complete enough to serve as screen fonts. In addition, &xorg; can be configured to use &truetype; fonts with a minimum of effort. For more details on this, see the &man.X.7; manual page or . To install the above Type1 font collections from the Ports Collection, run the following commands: &prompt.root; cd /usr/ports/x11-fonts/urwfonts &prompt.root; make install clean And likewise with the freefont or other collections. To have the X server detect these fonts, add an appropriate line to the X server configuration file (/etc/X11/xorg.conf), which reads: FontPath "/usr/local/share/fonts/urwfonts/" Alternatively, at the command line in the X session run: &prompt.user; xset fp+ /usr/local/share/fonts/urwfonts &prompt.user; xset fp rehash This will work but will be lost when the X session is closed, unless it is added to the startup file (~/.xinitrc for a normal startx session, or ~/.xsession when logging in through a graphical login manager like XDM). A third way is to use the new /usr/local/etc/fonts/local.conf as demonstrated in . &truetype; Fonts TrueType Fonts fonts TrueType &xorg; has built in support for rendering &truetype; fonts. There are two different modules that can enable this functionality. The freetype module is used in this example because it is more consistent with the other font rendering back-ends. To enable the freetype module just add the following line to the "Module" section of /etc/X11/xorg.conf. Load "freetype" Now make a directory for the &truetype; fonts (for example, /usr/local/share/fonts/TrueType) and copy all of the &truetype; fonts into this directory. Keep in mind that &truetype; fonts cannot be directly taken from an &apple; &mac;; they must be in &unix;/&ms-dos;/&windows; format for use by &xorg;. Once the files have been copied into this directory, use ttmkfdir to create a fonts.dir, so that the X font renderer knows that these new files have been installed. ttmkfdir is available from the FreeBSD Ports Collection as x11-fonts/ttmkfdir. &prompt.root; cd /usr/local/share/fonts/TrueType &prompt.root; ttmkfdir -o fonts.dir Now add the &truetype; directory to the font path. This is just the same as described in : &prompt.user; xset fp+ /usr/local/share/fonts/TrueType &prompt.user; xset fp rehash or add a FontPath line to xorg.conf. Now Gimp, OpenOffice, and all of the other X applications should now recognize the installed &truetype; fonts. Extremely small fonts (as with text in a high resolution display on a web page) and extremely large fonts (within &staroffice;) will look much better now. Anti-Aliased Fonts anti-aliased fonts fonts anti-aliased All fonts in &xorg; that are found in /usr/local/share/fonts/ and ~/.fonts/ are automatically made available for anti-aliasing to Xft-aware applications. Most recent applications are Xft-aware, including KDE, GNOME, and Firefox. In order to control which fonts are anti-aliased, or to configure anti-aliasing properties, create (or edit, if it already exists) the file /usr/local/etc/fonts/local.conf. Several advanced features of the Xft font system can be tuned using this file; this section describes only some simple possibilities. For more details, please see &man.fonts-conf.5;. XML This file must be in XML format. Pay careful attention to case, and make sure all tags are properly closed. The file begins with the usual XML header followed by a DOCTYPE definition, and then the <fontconfig> tag: <?xml version="1.0"?> <!DOCTYPE fontconfig SYSTEM "fonts.dtd"> <fontconfig> As previously stated, all fonts in /usr/local/share/fonts/ as well as ~/.fonts/ are already made available to Xft-aware applications. If you wish to add another directory outside of these two directory trees, add a line similar to the following to /usr/local/etc/fonts/local.conf: <dir>/path/to/my/fonts</dir> After adding new fonts, and especially new font directories, you should run the following command to rebuild the font caches: &prompt.root; fc-cache -f Anti-aliasing makes borders slightly fuzzy, which makes very small text more readable and removes staircases from large text, but can cause eyestrain if applied to normal text. To exclude font sizes smaller than 14 point from anti-aliasing, include these lines: <match target="font"> <test name="size" compare="less"> <double>14</double> </test> <edit name="antialias" mode="assign"> <bool>false</bool> </edit> </match> <match target="font"> <test name="pixelsize" compare="less" qual="any"> <double>14</double> </test> <edit mode="assign" name="antialias"> <bool>false</bool> </edit> </match> fonts spacing Spacing for some monospaced fonts may also be inappropriate with anti-aliasing. This seems to be an issue with KDE, in particular. One possible fix for this is to force the spacing for such fonts to be 100. Add the following lines: <match target="pattern" name="family"> <test qual="any" name="family"> <string>fixed</string> </test> <edit name="family" mode="assign"> <string>mono</string> </edit> </match> <match target="pattern" name="family"> <test qual="any" name="family"> <string>console</string> </test> <edit name="family" mode="assign"> <string>mono</string> </edit> </match> (this aliases the other common names for fixed fonts as "mono"), and then add: <match target="pattern" name="family"> <test qual="any" name="family"> <string>mono</string> </test> <edit name="spacing" mode="assign"> <int>100</int> </edit> </match> Certain fonts, such as Helvetica, may have a problem when anti-aliased. Usually this manifests itself as a font that seems cut in half vertically. At worst, it may cause applications to crash. To avoid this, consider adding the following to local.conf: <match target="pattern" name="family"> <test qual="any" name="family"> <string>Helvetica</string> </test> <edit name="family" mode="assign"> <string>sans-serif</string> </edit> </match> Once you have finished editing local.conf make sure you end the file with the </fontconfig> tag. Not doing this will cause your changes to be ignored. Finally, users can add their own settings via their personal .fonts.conf files. To do this, each user should simply create a ~/.fonts.conf. This file must also be in XML format. LCD screen Fonts LCD screen One last point: with an LCD screen, sub-pixel sampling may be desired. This basically treats the (horizontally separated) red, green and blue components separately to improve the horizontal resolution; the results can be dramatic. To enable this, add the line somewhere in local.conf: <match target="font"> <test qual="all" name="rgba"> <const>unknown</const> </test> <edit name="rgba" mode="assign"> <const>rgb</const> </edit> </match> Depending on the sort of display, rgb may need to be changed to bgr, vrgb or vbgr: experiment and see which works best. The X Display Manager Seth Kingsley Contributed by X Display Manager &xorg; provides an X Display Manager, XDM, which can be used for login session management. XDM provides a graphical interface for choosing which display server to connect to and for entering authorization information such as a login and password combination. This section demonstrates how to configure the X Display Manager on &os;. Some desktop environments provide their own graphical login manager. Refer to for instructions on how to configure the GNOME Display Manager and for instructions on how to configure the KDE Display Manager. Configuring <application>XDM</application> To install XDM, use the x11/xdm package or port. Once installed, XDM can be configured to run when the machine boots up by editing this entry in /etc/ttys: ttyv8 "/usr/local/bin/xdm -nodaemon" xterm off secure Change the off to on and save the edit. The ttyv8 in this entry indicates that XDM will run on the ninth virtual terminal. The XDM configuration directory is located in /usr/local/lib/X11/xdm. This directory contains several files used to change the behavior and appearance of XDM, as well as a few scripts and programs used to set up the desktop when XDM is running. summarizes the function of each of these files. The exact syntax and usage of these files is described in &man.xdm.1;. XDM Configuration Files File Description Xaccess The protocol for connecting to XDM is called the X Display Manager Connection Protocol (XDMCP) This file is a client authorization ruleset for controlling XDMCP connections from remote machines. By default, this file does not allow any remote clients to connect. Xresources This file controls the look and feel of the XDM display chooser and login screens. The default configuration is a simple rectangular login window with the hostname of the machine displayed at the top in a large font and Login: and Password: prompts below. The format of this file is identical to the app-defaults file described in the &xorg; documentation. Xservers The list of local and remote displays the chooser should provide as login choices. Xsession Default session script for logins which is run by XDM after a user has logged in. Normally each user will have a customized session script in ~/.xsession that overrides this script Xsetup_* Script to automatically launch applications before displaying the chooser or login interfaces. There is a script for each display being used, named Xsetup_*, where * is the local display number. Typically these scripts run one or two programs in the background such as xconsole. xdm-config Global configuration for all displays running on this machine. xdm-errors Contains errors generated by the server program. If a display that XDM is trying to start hangs, look at this file for error messages. These messages are also written to the user's ~/.xsession-errors on a per-session basis. xdm-pid The running process ID of XDM.
Configuring Remote Access By default, only users on the same system can login using XDM. To enable users on other systems to connect to the display server, edit the access control rules and enable the connection listener. To configure XDM to listen for any remote connection, comment out the DisplayManager.requestPort line in /usr/local/lib/X11/xdm/xdm-config by putting a ! in front of it: ! SECURITY: do not listen for XDMCP or Chooser requests ! Comment out this line if you want to manage X terminals with xdm DisplayManager.requestPort: 0 Save the edits and restart XDM. To restrict remote access, look at the example entries in /usr/local/lib/X11/xdm/Xaccess and refer to &man.xdm.1; for further information. Desktop Environments Valentino Vaschetto Contributed by This section describes how to install three popular desktop environments on a &os; system. A desktop environment can range from a simple window manager to a complete suite of desktop applications. Over a hundred desktop environments are available in the x11-wm category of the Ports Collection. GNOME GNOME GNOME is a user-friendly desktop environment. It includes a panel for starting applications and displaying status, a desktop, a set of tools and applications, and a set of conventions that make it easy for applications to cooperate and be consistent with each other. More information regarding GNOME on &os; can be found at http://www.FreeBSD.org/gnome. That web site contains additional documentation about installing, configuring, and managing GNOME on &os;. This desktop environment can be installed from a package: - &prompt.root; pkg install gnome2 + &prompt.root; pkg install gnome3 To instead build GNOME from ports, use the following command. GNOME is a large application and will take some time to compile, even on a fast computer. - &prompt.root; cd /usr/ports/x11/gnome2 + &prompt.root; cd /usr/ports/x11/gnome3 &prompt.root; make install clean For proper operation, GNOME requires /proc to be mounted. Add this line to /etc/fstab to mount this file system automatically during system startup: proc /proc procfs rw 0 0 Once GNOME is installed, configure &xorg; to start GNOME. The easiest way to do this is to enable the GNOME Display Manager, GDM, which is installed as part of the GNOME package or port. It can be enabled by adding this line to /etc/rc.conf: gdm_enable="YES" It is often desirable to also start all GNOME services. To achieve this, add a second line to /etc/rc.conf: gnome_enable="YES" GDM will now start automatically when the system boots. A second method for starting GNOME is to type startx from the command-line after configuring ~/.xinitrc. If this file already exists, replace the line that starts the current window manager with one that starts /usr/local/bin/gnome-session. If this file does not exist, create it with this command: &prompt.user; echo "exec /usr/local/bin/gnome-session" > ~/.xinitrc A third method is to use XDM as the display manager. In this case, create an executable ~/.xsession: &prompt.user; echo "#!/bin/sh" > ~/.xsession &prompt.user; echo "exec /usr/local/bin/gnome-session" >> ~/.xsession &prompt.user; chmod +x ~/.xsession KDE KDE KDE is another easy-to-use desktop environment. This desktop provides a suite of applications with a consistent look and feel, a standardized menu and toolbars, keybindings, color-schemes, internationalization, and a centralized, dialog-driven desktop configuration. More information on KDE can be found at http://www.kde.org/. For &os;-specific information, consult http://freebsd.kde.org. To install the KDE package, type: &prompt.root; pkg install x11/kde4 To instead build the KDE port, use the following command. Installing the port will provide a menu for selecting which components to install. KDE is a large application and will take some time to compile, even on a fast computer. &prompt.root; cd /usr/ports/x11/kde4 &prompt.root; make install clean KDE display manager KDE requires /proc to be mounted. Add this line to /etc/fstab to mount this file system automatically during system startup: proc /proc procfs rw 0 0 The installation of KDE includes the KDE Display Manager, KDM. To enable this display manager, add this line to /etc/rc.conf: kdm4_enable="YES" A second method for launching KDE is to type startx from the command line. For this to work, the following line is needed in ~/.xinitrc: exec /usr/local/bin/startkde A third method for starting KDE is through XDM. To do so, create an executable ~/.xsession as follows: &prompt.user; echo "#!/bin/sh" > ~/.xsession &prompt.user; echo "exec /usr/local/bin/startkde" >> ~/.xsession &prompt.user; chmod +x ~/.xsession Once KDE is started, refer to its built-in help system for more information on how to use its various menus and applications. Xfce Xfce is a desktop environment based on the GTK+ toolkit used by GNOME. However, it is more lightweight and provides a simple, efficient, easy-to-use desktop. It is fully configurable, has a main panel with menus, applets, and application launchers, provides a file manager and sound manager, and is themeable. Since it is fast, light, and efficient, it is ideal for older or slower machines with memory limitations. More information on Xfce can be found at http://www.xfce.org. To install the Xfce package: &prompt.root; pkg install xfce Alternatively, to build the port: &prompt.root; cd /usr/ports/x11-wm/xfce4 &prompt.root; make install clean Unlike GNOME or KDE, Xfce does not provide its own login manager. In order to start Xfce from the command line by typing startx, first add its entry to ~/.xinitrc: &prompt.user; echo "exec /usr/local/bin/startxfce4" > ~/.xinitrc An alternate method is to use XDM. To configure this method, create an executable ~/.xsession: &prompt.user; echo "#!/bin/sh" > ~/.xsession &prompt.user; echo "exec /usr/local/bin/startxfce4" >> ~/.xsession &prompt.user; chmod +x ~/.xsession Installing Compiz Fusion One way to increase the pleasantness of using a desktop computer is by having nice 3D effects. Installing the Compiz Fusion package is easy, but configuring it requires a few steps that are not described in the port's documentation. Setting up the &os; nVidia Driver Desktop effects can cause quite a load on the graphics card. For an nVidia-based graphics card, the proprietary driver is required for good performance. Users of other graphics cards can skip this section and continue with the xorg.conf configuration. To determine which nVidia driver is needed see the FAQ question on the subject. Having determined the correct driver to use for your card, installation is as simple as installing any other package. For example, to install the latest driver: &prompt.root; pkg install x11/nvidia-driver The driver will create a kernel module, which needs to be loaded at system startup. Add the following line to /boot/loader.conf: nvidia_load="YES" To immediately load the kernel module into the running kernel by issuing a command like kldload nvidia, however it has been noted that the some versions of &xorg; will not function properly if the driver is not loaded at boot time. After editing /boot/loader.conf, a reboot is recommended. With the kernel module loaded, you normally only need to change a single line in xorg.conf to enable the proprietary driver: Find the following line in /etc/X11/xorg.conf: Driver "nv" and change it to: Driver "nvidia" Start the GUI as usual, and you should be greeted by the nVidia splash. Everything should work as usual. Configuring xorg.conf for Desktop Effects To enable Compiz Fusion, /etc/X11/xorg.conf needs to be modified: Add the following section to enable composite effects: Section "Extensions" Option "Composite" "Enable" EndSection Locate the Screen section which should look similar to the one below: Section "Screen" Identifier "Screen0" Device "Card0" Monitor "Monitor0" ... and add the following two lines (after Monitor will do): DefaultDepth 24 Option "AddARGBGLXVisuals" "True" Locate the Subsection that refers to the screen resolution that you wish to use. For example, if you wish to use 1280x1024, locate the section that follows. If the desired resolution does not appear in any subsection, you may add the relevant entry by hand: SubSection "Display" Viewport 0 0 Modes "1280x1024" EndSubSection A color depth of 24 bits is needed for desktop composition, change the above subsection to: SubSection "Display" Viewport 0 0 Depth 24 Modes "1280x1024" EndSubSection Finally, confirm that the glx and extmod modules are loaded in the Module section: Section "Module" Load "extmod" Load "glx" ... The preceding can be done automatically with x11/nvidia-xconfig by running (as root): &prompt.root; nvidia-xconfig --add-argb-glx-visuals &prompt.root; nvidia-xconfig --composite &prompt.root; nvidia-xconfig --depth=24 Installing and Configuring Compiz Fusion Installing Compiz Fusion is as simple as any other package: &prompt.root; pkg install x11-wm/compiz-fusion When the installation is finished, start your graphic desktop and at a terminal, enter the following commands (as a normal user): &prompt.user; compiz --replace --sm-disable --ignore-desktop-hints ccp & &prompt.user; emerald --replace & Your screen will flicker for a few seconds, as your window manager (e.g. Metacity if you are using GNOME) is replaced by Compiz Fusion. Emerald takes care of the window decorations (i.e. close, minimize, maximize buttons, title bars and so on). You may convert this to a trivial script and have it run at startup automatically (e.g. by adding to Sessions in a GNOME desktop): #! /bin/sh compiz --replace --sm-disable --ignore-desktop-hints ccp & emerald --replace & Save this in your home directory as, for example, start-compiz and make it executable: &prompt.user; chmod +x ~/start-compiz Then use the GUI to add it to Startup Programs (located in System, Preferences, Sessions on a GNOME desktop). To actually select all the desired effects and their settings, execute (again as a normal user) the Compiz Config Settings Manager: &prompt.user; ccsm In GNOME, this can also be found in the System, Preferences menu. If you have selected gconf support during the build, you will also be able to view these settings using gconf-editor under apps/compiz. Troubleshooting If the mouse does not work, you will need to first configure it before proceeding. In recent Xorg versions, the InputDevice sections in xorg.conf are ignored in favor of the autodetected devices. To restore the old behavior, add the following line to the ServerLayout or ServerFlags section of this file: Option "AutoAddDevices" "false" Input devices may then be configured as in previous versions, along with any other options needed (e.g., keyboard layout switching). As previously explained the hald daemon will, by default, automatically detect your keyboard. There are chances that your keyboard layout or model will not be correct, desktop environments like GNOME, KDE or Xfce provide tools to configure the keyboard. However, it is possible to set the keyboard properties directly either with the help of the &man.setxkbmap.1; utility or with a hald's configuration rule. For example if, one wants to use a PC 102 keys keyboard coming with a french layout, we have to create a keyboard configuration file for hald called x11-input.fdi and saved in the /usr/local/etc/hal/fdi/policy directory. This file should contain the following lines: <?xml version="1.0" encoding="iso-8859-1"?> <deviceinfo version="0.2"> <device> <match key="info.capabilities" contains="input.keyboard"> <merge key="input.x11_options.XkbModel" type="string">pc102</merge> <merge key="input.x11_options.XkbLayout" type="string">fr</merge> </match> </device> </deviceinfo> If this file already exists, just copy and add to your file the lines regarding the keyboard configuration. You will have to reboot your machine to force hald to read this file. It is possible to do the same configuration from an X terminal or a script with this command line: &prompt.user; setxkbmap -model pc102 -layout fr /usr/local/share/X11/xkb/rules/base.lst lists the various keyboard, layouts and options available. &xorg; tuning The xorg.conf.new configuration file may now be tuned to taste. Open the file in a text editor such as &man.emacs.1; or &man.ee.1;. If the monitor is an older or unusual model that does not support autodetection of sync frequencies, those settings can be added to xorg.conf.new under the "Monitor" section: Section "Monitor" Identifier "Monitor0" VendorName "Monitor Vendor" ModelName "Monitor Model" HorizSync 30-107 VertRefresh 48-120 EndSection Most monitors support sync frequency autodetection, making manual entry of these values unnecessary. For the few monitors that do not support autodetection, avoid potential damage by only entering values provided by the manufacturer. X allows DPMS (Energy Star) features to be used with capable monitors. The &man.xset.1; program controls the time-outs and can force standby, suspend, or off modes. If you wish to enable DPMS features for your monitor, you must add the following line to the monitor section: Option "DPMS" xorg.conf While the xorg.conf.new configuration file is still open in an editor, select the default resolution and color depth desired. This is defined in the "Screen" section: Section "Screen" Identifier "Screen0" Device "Card0" Monitor "Monitor0" DefaultDepth 24 SubSection "Display" Viewport 0 0 Depth 24 Modes "1024x768" EndSubSection EndSection The DefaultDepth keyword describes the color depth to run at by default. This can be overridden with the command line switch to &man.Xorg.1;. The Modes keyword describes the resolution to run at for the given color depth. Note that only VESA standard modes are supported as defined by the target system's graphics hardware. In the example above, the default color depth is twenty-four bits per pixel. At this color depth, the accepted resolution is 1024 by 768 pixels. Finally, write the configuration file and test it using the test mode given above. One of the tools available to assist you during troubleshooting process are the &xorg; log files, which contain information on each device that the &xorg; server attaches to. &xorg; log file names are in the format of /var/log/Xorg.0.log. The exact name of the log can vary from Xorg.0.log to Xorg.8.log and so forth. If all is well, the configuration file needs to be installed in a common location where &man.Xorg.1; can find it. This is typically /etc/X11/xorg.conf or /usr/local/etc/X11/xorg.conf. &prompt.root; cp xorg.conf.new /etc/X11/xorg.conf The &xorg; configuration process is now complete. &xorg; may be now started with the &man.startx.1; utility. The &xorg; server may also be started with the use of &man.xdm.1;. Configuration with &intel; <literal>i810</literal> Graphics Chipsets Intel i810 graphic chipset Configuration with &intel; i810 integrated chipsets requires the agpgart AGP programming interface for &xorg; to drive the card. See the &man.agp.4; driver manual page for more information. This will allow configuration of the hardware as any other graphics board. Note on systems without the &man.agp.4; driver compiled in the kernel, trying to load the module with &man.kldload.8; will not work. This driver has to be in the kernel at boot time through being compiled in or using /boot/loader.conf. Adding a Widescreen Flatpanel to the Mix widescreen flatpanel configuration This section assumes a bit of advanced configuration knowledge. If attempts to use the standard configuration tools above have not resulted in a working configuration, there is information enough in the log files to be of use in getting the setup working. Use of a text editor will be necessary. Current widescreen (WSXGA, WSXGA+, WUXGA, WXGA, WXGA+, et.al.) formats support 16:10 and 10:9 formats or aspect ratios that can be problematic. Examples of some common screen resolutions for 16:10 aspect ratios are: 2560x1600 1920x1200 1680x1050 1440x900 1280x800 At some point, it will be as easy as adding one of these resolutions as a possible Mode in the Section "Screen" as such: Section "Screen" Identifier "Screen0" Device "Card0" Monitor "Monitor0" DefaultDepth 24 SubSection "Display" Viewport 0 0 Depth 24 Modes "1680x1050" EndSubSection EndSection &xorg; is smart enough to pull the resolution information from the widescreen via I2C/DDC information so it knows what the monitor can handle as far as frequencies and resolutions. If those ModeLines do not exist in the drivers, one might need to give &xorg; a little hint. Using /var/log/Xorg.0.log one can extract enough information to manually create a ModeLine that will work. Simply look for information resembling this: (II) MGA(0): Supported additional Video Mode: (II) MGA(0): clock: 146.2 MHz Image Size: 433 x 271 mm (II) MGA(0): h_active: 1680 h_sync: 1784 h_sync_end 1960 h_blank_end 2240 h_border: 0 (II) MGA(0): v_active: 1050 v_sync: 1053 v_sync_end 1059 v_blanking: 1089 v_border: 0 (II) MGA(0): Ranges: V min: 48 V max: 85 Hz, H min: 30 H max: 94 kHz, PixClock max 170 MHz This information is called EDID information. Creating a ModeLine from this is just a matter of putting the numbers in the correct order: ModeLine <name> <clock> <4 horiz. timings> <4 vert. timings> So that the ModeLine in Section "Monitor" for this example would look like this: Section "Monitor" Identifier "Monitor1" VendorName "Bigname" ModelName "BestModel" ModeLine "1680x1050" 146.2 1680 1784 1960 2240 1050 1053 1059 1089 Option "DPMS" EndSection Now having completed these simple editing steps, X should start on your new widescreen monitor. Troubleshooting Compiz Fusion I have installed Compiz Fusion, and after running the commands you mention, my windows are left without title bars and buttons. What is wrong? You are probably missing a setting in /etc/X11/xorg.conf. Review this file carefully and check especially the DefaultDepth and AddARGBGLXVisuals directives. When I run the command to start Compiz Fusion, the X server crashes and I am back at the console. What is wrong? If you check /var/log/Xorg.0.log, you will probably find error messages during the X startup. The most common would be: (EE) NVIDIA(0): Failed to initialize the GLX module; please check in your X (EE) NVIDIA(0): log file that the GLX module has been loaded in your X (EE) NVIDIA(0): server, and that the module is the NVIDIA GLX module. If (EE) NVIDIA(0): you continue to encounter problems, Please try (EE) NVIDIA(0): reinstalling the NVIDIA driver. This is usually the case when you upgrade &xorg;. You will need to reinstall the x11/nvidia-driver package so glx is built again.