Frugalware 20080516 (-current) Documentation Table of Contents JavaScript must be enabled in your browser to display the table of contents. Copyright (C) 2003-2007 The Frugalware Developer Team. Permission is granted to copy, distribute and/or modify this document under the terms of the GNU Free Documentation License, Version 1.2 or any later version published by the Free Software Foundation; with no Invariant Sections, no Front-Cover Texts, and no Back-Cover Texts. A copy of the license is included in the section entitled "GNU Free Documentation License". 1. Introduction Before you start to read this document, you should know some important things about how to read it. 1.1. Things that you should really read First there are some part of this document that you should really read, to understand how frugalware is working and how to administrate it. Important references to read: * This introduction ;) * How to use pacman-g2. * How to manage service. 1.2. Running console commands Along this document, there are boxed text that shows you a console log. These logs are important and requires quite some attentions since most off the time you are expected to run them and get the same output. $ echo foo bar foo bar This is how a console log look. Lets details it so you understand what means. The echo foo bar part is what you should type and it's the command. The following line foo bar is the ouput of the previous command. You may wonder what differenciate the command from the output. You see that in front of the command there is a $. This indicates that it's a command line, but there is more meaning in this symbol. This symbol can change depending on the user priviledges required to run the command. Here is the list of the common prefix for the console commands: * $ indicates that any user can run the command. Most of the time it means you have to run it with your user. * user$ indicates that a the specified user priviledge are required to run this command. Usually this is necessary for security reasons. You can get an interactive shell for this user, replacing user with the wanted user name, by issuing: $ su - user * # indicates that the root users priviledges are required to run this command. Usually this is required to manage the system configuration. You can get an interactive shell for root running: $ su - 2. About Frugalware Seeing this feast of wonderful code spread in front of me as a working system was a much more powerful experience than merely knowing, intellectually, that all the bits were probably out there. It was as though for years I'd been sorting through piles of disconnected car parts - only to be suddenly confronted with those same parts assembled into a gleaming red Ferrari, door open, keys swinging from the lock and engine gently purring with a promise of power... -- Eric S. Raymond The aim of creating Frugalware was to help you doing your work faster and simpler. We hope you will like it. In this introduction, we would like to answer a few questions which were asked in several interview with VMiklos, the founder of the project. You can reach the full list of articles that have been posted about Frugalware [1]here. 2.1. Short Frugalware is a general purpose linux distribution, designed for intermediate users (who are not afraid of text mode). 2.2. Long What branches does Frugalware have? "We have a -current and a -stable branch. The -current branch is updated daily, and we provide security support for our -stable branch till the next release, for approximately 6 months." What is "The Frugalware Philosophy" about? "Briefly: simplicity, multimedia, design. We try to make Frugalware as simple as possible while not forgetting to keep it comfortable for the user. We try to ship fresh and stable software, as close to the original source as possible, because in our opinion most software is the best as is, and doesn't need patching." What is the license of Frugalware? "The license of Frugalware itself stands for the license of the buildscripts used for building Frugalware. That source is available under the GPL license [2]here. Frugalware originally init scripts written by Patrick J. Volkerding, creator of the Slackware Linux distribution. We GPL our additions, but Patrick J. Volkerding's code is still under the BSD license. Frugalware also has a few side projects, like our pacman-g2 package manager, the Frugalware installer an so on. They are available under the GPL license, too. For more info about the license of the packages included in Frugalware, refer to the /usr/share/doc/*/COPYING files." What package manager does Frugalware use? "We have our own package manager, called pacman-g2. It stands for the second generation of the pacman-g1 package manager, as it was originally based on Judd Vinet's great work. The packages are simple .tar.bz2 files, pacman-g2 is written in C, unlike Slackware's shellscript-based package manager (which may be rather slow sometimes)." How does Frugalware manage updating obsolete packages? "We don't have any standalone program for updating packages as pacman-g2 manages this task too. To update your package database, use pacman-g2 -Sy, and to update your packages according the just synchronized package database, you use pacman-g2 -Su. To install package foo with the necessary dependencies directly from one of our ftp servers, you should issue pacman-g2 -S foo. For more information, refer to the pacman-g2 man page." Is there any community support available for Frugalware? "We have mailing lists, irc channels and forums that can be used to communicate with us or with other users and to get help. You can reach the list of mailing lists available [3]here. The irc channels are on the Freenode network (server: irc.freenode.net), the discussion forums are available [4]here." Is there any commercial support available for Frugalware? "No, there isn't for now, and currently it isn't planned, either." For whom is Frugalware recommended to use? "Frugalware is designed for intermediate users. Installing Frugalware is not a magic, of course, but you should read some documentation if you don't know what a partition, an MBR (Master Boot Record), etc. is." How to become a developer? "Get involved! :) Download the FST (Frugalware Source Tree) using the repoman upd command, which is available in the pacman-tools package. Then start to play with the FrugalBuild scripts, for a skeleton, refer to the /docs/skel directory. Try to improve them, or write a new one for a currently unsupported program. Then open feature requests in the [5]Bug Tracking System and attach your patches. From this point everything will come naturally to you :)" What do developers do? "In short, what they want to, if they play a square game. They may maintain packages: building them if a newer version is available and update the FrugalBuild scripts to work correctly against a newer version. They can contribute a new build script to a previously non-existent package. They write documentation, fix bugs, provides supports, or anything else in connection with the Frugalware community. If you only want to help us, but you don't want to hack, you may help in translating Frugalware to your or other language. And, of course, we happily accept donations :)" Who develops Frugalware? "An amazing group of volunteers, who are motived by the users to do so. They also do it as a hobby, and they are always working on having up to date knowledge to make Frugalware even better for you." Is Frugalware specialized in a certain purpose? "No, it's a general purpose distribution, for desktops, mobile computers and servers." Do you plan to release a live cd? "Yes, you can reach test iso images in our [6]testing-iso directory." Does Frugalware support languages other than English? "Yes, it supports all languages supported by the packages. If the init scripts, the setup or the documentation is not available in your language, then it simply means it haven't translated yet." What about Asian languages? "Frugalware roughly supports Asian languages, but don't expect too much - using UTF8 is not the default where it is possible." What architectures does Frugalware support? "Currently we support x86 (Pentium Pro or higher) and x86_64 (k8, aka. amd64) platforms." 3. Quick reference 3.1. Informations * Package management: pacman-g2 (command line) * No PAM problems/support (a'la Slackware) * Linux kernel 2.6 (no 2.4 support) * i686 or later CPU needed on x86 * The latest documentation is [7]here. 3.2. Features * Supported architectures: i686, x86_64 * Stable releases every 6 months * Security support for stable releases * Text mode installation * Offline installation, netboot install supported * Prebuilt CD/DVD, USB, TFTP images are available * Localization supported whereever it's possible * About 3000 source packages and (as of Februrary 2007) 3200 binary packages supported. 4. Installation 4.1. Choosing installation flavor Depending on your needs, there are different installers with different characteristics. You can choose which fits you the best. 4.1.1. Installing from CDs Which CDs do you need? If you install a server without X, only the first. If you need a graphical system, then you'll need the second CD as well. Note Don't download CDs 3-11 unless you don't have an internet connection! You can install language packs later from FTP servers if you need them. 4.1.2. Installing from DVD If you don't have any Internet connection but you want language packs and other optional packages, you'll need two DVDs. Pros: a full offline installation is possible. Cons: Large amount of data must be downloaded, presumably some unnecessary packages too. 4.1.3. Netinstall This is a small ISO image, currently under 32Mb, which is able to boot up, configure the network and install the system with the selected packages, which are downloaded on-the-fly as required. Pros: Small image size, no wasted bandwidth with downloading outdated or unnecessary packages. Cons: No offline installation possible, high bandwidth or hours of patience required for a full installation. 4.1.4. USB image This is a filesystem image, similar to the network install ISO image. Warning Writing the boot image to a USB stick will destroy all the data on the drive. The following command will install the image to the USB stick on any recent Linux system: # dd if=frugalware---usb.img of=/dev/sdX Important Pay attention to see what /dev/sdX device your USB stick, for example by having a look at the contents of the /dev/disk/by-id/ directory! You can use a similar tool ([8]like this) on Windows systems as well: dd if=frugalware---usb.img of=\\?\Device\Harddisk1\Partition0 \ bs=1M --size --progress Pros: No need to burn any CD. Cons: You have to be able to boot from USB. 4.1.5. TFTP image This is a floppy image, for a very special case: * you want to do a network installation * you don't want to / can't use CDs * you don't want to / can't boot from an USB stick * you can boot from a network card, but your BIOS does not supports so * you have a floppy drive Pros: In some cases this is the only way you can install Frugalware Cons: You need a bootable network card and a working TFTP server 4.1.6. Fwbootstrap (self-contained chroot) This is a tarball which has to be downloaded and unpacked. Mostly useful for developers who can compile packages in this build environment on a non-Frugalware host system. Usage example: 1. Download the tarball $ wget ftp://ftp5.frugalware.org/packages/frugalware/pub/frugalware/\ frugalware-stable-iso/fwchroot--.tar.bz2 2. Unpack it $ tar xvjf fwchroot--.tar.bz2 3. Enter the chroot. $ cd fwchroot-- $ ./fwbootstrap 4. Use it (build a package or two) 5. Exit from the shell and fwbootstrap will unmount the necessary dirs for you. You can get a list of installed packages in the chroot with issuing the pacman-g2 -Q command. 4.1.7. A manual bootstrap So you want a complete Frugalware installed into /mnt/foo. First of all, you must have a running Frugalware where you are able to do # pacman-g2 -Sy core base -r /mnt/foo which installs the core and base pkgs into it. But beware: $ pacman-g2 -Qo /etc/sysconfig/keymap No package owns /etc/sysconfig/keymap $ pacman-g2 -Qo /etc/profile.d/lang.sh No package owns /etc/profile.d/lang.sh $ pacman-g2 -Qo /etc/fstab No package owns /etc/fstab so you have to copy or forge them by hand. 4.2. Obtaining a source media A Frugalware installation media can be obtained from several sources. You can download it freely via HTTP, FTP or rsync. You can also grab it via bittorrent, see [9]Linuxtracker for example. The following examples explains how you can get the iso images. You have to replace respectively $version$, $arch$ and $media$ to get the wanted iso image. Via FTP: $ wget ftp://ftp3.frugalware.org/mirrors/frugalware/pub/frugalware/\ frugalware-$version$-iso/frugalware-$version$-$arch$-$media$.iso Via HTTP: $ wget http://www5.frugalware.org/linux/frugalware/pub/frugalware/\ frugalware-$version$-iso/frugalware-$version$-$arch$-$media$.iso Via rsync: $ rsync -avP rsync://rsync4.frugalware.org/ftp/pub/linux/distributions/\ frugalware/frugalware-$version$-iso/frugalware-$version$-$arch$-$media$.iso ./ More info and the full list of mirrors can be found at our [10]download page. 4.3. Using packages from CD/DVD You have a skeleton system installed from CD/DVD, and you want to use the packages from the media afterwards. There are two methods. First is the easiest, but needs quite a lot of space (and caution not to use pacman-g2 -Scc ;) ): mount the media and install all the .fpm's found in frugalware-i686 (or frugalware-x86_64) dir to /var/cache/pacman/pkg. Second is a bit more challenging, but more usable. Add a new line to /etc/pacman-g2/repos/frugalware before the other Server lines: Server = file:///media/dvd/frugalware-i686 On x86_64, use this one: Server = file:///media/dvd/frugalware-x86_64 The media should be mounted on /media/dvd, or change the Server lines appropriately. Also you can only install packages then from the given media, so you have to insert the first CD if you install a package from the first CD and so on. This is something you should pay attention for. 4.4. The installation process Important Do not worry if you misconfigured something! Just press in the next dialog and you will see the menu. Just go back to the given part and you can reconfigure it. * After downloading and burning the CDs/DVD, insert the first CD/DVD to your CD/DVD drive, and reboot your computer. In the grub menu, you can disable the framebuffer, if a framebuffer with resolution 1024x768 is not suitable for your graphics card or monitor. After that, grub loads the kernel and the initrd image. * At the first dialog, you should select your language. If your language is not on the list, you should choose a language fits for you. You can change these options after installing too. * The next dialog is only a greetings. Just push . Now it is time to select your keyboard type. Pick your one, then hit ! * After selecting your keyboard map, setup searches for installation media automatically. * If you use a netinstall image follow these sub-steps. Otherwise jump to the partitioning point! Note These steps sets up your network options during the install. When you finished installing Frugalware the installer will ask for network options again. Those options will be the installed system's options. 1. Now you should select your connection type. The installer uses the netconfig utility. You can also find the documentation for netconfig in this documentation. See the part called: Networking. 2. After setting up the network you can choose a mirror for downloading the packages. The installer will try other mirrors too. This feature is useful when you have got a fast local mirror or something similar. * The next step is partitioning. Frugalware setup displays a list of your hard disks, you should choose one of them to partition it with a program. You can select the partitioning program in the next dialog, currently fdisk and cfdisk are included. You should create at least one partition with type Linux, and it is recommended to create a swap partition (with type Linux swap). The swap size should be 500-1000MB. When you have finished partitioning, press . * The following list displays your swap partitions, here you can choose which swap partitions are allowed to be used by Frugalware. Then setup formats your swap partitions. If you have no swap partition just press ! * In the next window, you should select your root partition first, then you can choose if you wish to format it or keep the existing filesystem on it. After selecting the root partition, you can setup other Linux partitions, optionally format them, and set their mount points. Using a separate partition is supported for /boot, /home, /var, but not yet for /usr. * After having your Linux partitions mounted, you should do the same with your DOS/Windows ones. Setup will display a list of them, if any exists. You should simply choose a mount point for them here. * Now it is the time to select if you want to use expert menus or not. If you choose expert menu after selecting the categories you will be able to pick packages one-by-one from the selected categories. So if you select apps and base the installer will give you a list of packages in apps, when you finished picking the packages you will see the packages in base. After picking them the installation begins. If you choose the normal menu (it's the default) then you will only see the groups, but not the individual packages. So after picking the groups installation starts. * The next step is to select package categories. If you will not use KDE or GNOME, you may probably want to disable them. In most cases, it is not a good idea to disable other categories. If you selected the expert menu you will see the package list after this dialog. Note If the group list is empty that means you probably misconfigured your network. Please go back and try to fix it. You can also test your connection if you press Alt+F2 and try to ping some servers. * Setup will install the packages your selected from the first CD. When it is done, you will be prompted to insert the next Frugalware install. If you have only one disc, feel free to abort installing packages, you can install anything else from the net later. 5. Basic configuration 5.1. Introduction After the installation of the packages, Frugalware setup will configure your new Frugalware system. If you installed the packages manually, then you'll have to perform those configuration steps manually. Note If any problem occurs, there is a debug console on tty4, you can see that by pressing Alt-F4. You can switch back by hitting Alt-F1. 5.2. GRUB The first step is to install grub onto your hard disk. There are four options here: installing to the MBR, the root partition, a floppy or simply skipping. Installing to the MBR is the good choice if you want Frugalware to manage your computer's booting. The root is a good idea if you want to install grub into your root partition. In this case, grub will not modify your existing boot manager. Floppy is a good idea for example if you don't have any boot manager installed, but you want to leave your MBR unmodified. 5.3. Kernel modules After the installation of grub, the installer will configure your kernel modules. This means that an information dialog appears, but nothing more. 5.4. Accounts and passwords After module configuration, you should change the root password. This is very important as there is no default password. If you skip this step, anybody will be able to login as root. After this step, you can create a regular (also known as non-root) user. It's highly recommended to create one, and log in as a regular user. If a command should be run as root, you should use su or sudo under console, and gksu or kdesu under X. 5.5. Network After this, setup will configure your network settings. Setup simply runs the netconfig utility, which is described in the Networking section. 5.6. Timezone If network installation is done, we should configure the system's time. This means two actions. First, you should decide if the hardware (BIOS) clock is set to Coordinated Universal Time (UTC). If yes, select yes here. If the hardware clock is set to the current local time (this is how most PCs are set up), say no here. If you are not sure what is this, you should answer no here. 5.7. Mouse The next step is to configure your mouse. The configuration will take effect on the console mouse services (gpm) and on the X server. The setting is done by xconfig later. 5.8. Graphical interface If you have installed an X server (by default xorg), the setup will run xconfig. For more information on xconfig, see the section Graphical interface (X11). 6. Pacman-G2 6.1. Basics Frugalware comes with Pacman-G2 package manager. Pacman-G2 is a fork of the not-yet-released cvs version of the complete rewrite of pacman-g1 by Aurelien Foret (the old monolithic pacman-g1 is written by Judd Vinet). See the [11]README for details. If you want to do anything with packages, you'll always have to use the pacman-g2 command. Here are some basic actions with pacman-g2: Actions usually used with remote installation from an FTP server: # pacman-g2 -Sy Updates the package database. Before searching for packages or installing them from an FTP server, you will have to use this command. # pacman-g2 -Su Upgrades all packages that are currently installed but a newer version of the package is available on the FTP server. # pacman-g2 -Syu The combination of the above two, that is the command most users use daily. $ pacman-g2 -Sup Prints the URL of all packages that pacman-g2 should download. This way you can download the packages anywhere and then just copy them to /var/cache/pacman/pkg. This is very useful if you have limited bandwidth at your computer, but you can access high bandwidth elsewhere. # pacman-g2 -S sendmail Installs sendmail with all of its dependencies from the FTP server. If it conflicts with any package, you will be asked if pacman-g2 is allowed to remove them. $ pacman-g2 -Ss perl Searches in the package database (on the FTP server). This example will probably display the perl package and all perl modules. Regular expression based search is also supported. Of course, you can treat packages as normal files, and you can manually add/remove/etc them. Here are some examples: # pacman-g2 -U zsh-4.2.1-1.fpm Adds (or if it's already installed, upgrades) the zsh package, which is located in the current directory. # pacman-g2 -R qt Removes the qt package. $ pacman-g2 -Qs perl Shows every installed packages whose name contains the string perl. Generally, if you want to turn off checking for conflicting files, you should use the -f parameter, and if you want to turn off all dependency checking, you should use the -d switch. $ pacman-g2 -h This displays all the switches we discussed above, and a lot more. Once again, these are only the basics. You can also use pacman-g2 -Sh or similar to get help on a particular task. Note Full documentation for pacman-g2 can be reached by issuing man pacman-g2. 6.2. Apt - pacman-g2 cross reference For those who are familiar with the apt package management tool, here is a quick cross-reference. Action Apt command Pacman command Refresh the package database: apt-get update pacman-g2 -Sy Upgrade currently installed packages: apt-get upgrade pacman-g2 -Su Install a new package: apt-get install foo pacman-g2 -S foo Remove a package: apt-get remove foo pacman-g2 -Rc foo Search in the full package database: apt-cache search foo pacman-g2 -Ss foo Install a package from a file: dpkg -i foo.deb pacman-g2 -A foo.fpm Clean the package cache: apt-get clean pacman-g2 -Sc 7. Networking 7.1. Initializing the network card In most cases, configuring your network card will be done automatically by udev. This means that during every system boot your network card will be detected, and the necessary modules will be loaded. If you want, you can load your network card's module manually by editing the /etc/sysconfig/modules file and put the module in the blacklist by editing /etc/sysconfig/blacklist. Configuring any interface on your card will be the task of the netconfig utility. Initializing your card ends here. 7.2. The netconfig utility Configuring your network settings is done by the netconfig utility. 1. First, we have to give a name to your computer. The name must consist of at least two parts, separated by a dot (.). 2. In the next dialog, you should choose how your machine connects to the network. If you have an internal network card and an assigned IP address, gateway, and DNS, use static to enter these values. If your IP address is assigned by a DHCP server (commonly used by cable modem services, not equal to dsl services), select dhcp. In case you've got a DSL connection (eg. ADSL) chose the dsl option! Finally, if you do not have a network card, select the lo choice. The lo is also the correct choice if you are using a PCMCIA network card. When you set up the network first question will be the interface you want to set up. It is usually eth0, but it can differ when you set up wireless interfaces for example. If you set up a wireless card netconfig will also ask your ESSID and encription key. 1. If you chose static, you must give your IP address, the netmask of your local network, your gateway address (you may leave it blank) and the IP address of your primary name server (you can add more nameservers later by editing the /etc/resolv.conf file) and then the configuration is finished. 2. If you chose dhcp, you can optionally give your dhcp hostname, however, netconfig will not ask more questions about your network, since all other data will be provided by the DHCP server. 3. In dsl part you must give your username, something like someone@provider.net. Then you'll have to specify the network interface (usually eth0) through which the ADSL connecting script will try to communicate with your ADSL modem. Then enter your password twice. 4. If you chose lo, you don't have to answer any questions. 3. Finally, netconfig will write all your network configuration files. If you want to edit your settings by hand, the interface information is stored in the /etc/sysconfig/network directory. There is only one file there called default in most cases. It's because you can set up more then one profile. It's very useful when you got a laptop so you can set up options for all networks you use. 7.3. Basic firewall configuration Frugalware comes with a firewall configuration working out of the box. This allows all outgoing connections, and incoming packets for established connections. It does not allow normal incoming packages for any ports. The firewall configuration is at /etc/sysconfig/firewall. Note You will not find this file if you have not installed iptables package as this is an iptables firewall. Let's see an example: you would like to allow others to ssh into your computer. Edit /etc/sysconfig/firewall, remove the hashmark () from the beginning of the line under the ssh description, and restart the firewall: # service firewall restart The same applies for Apache or any other services. If you would like to have any advanced firewall settings, configure your firewall as root with iptables then save your config as root with: # iptables-save > /etc/sysconfig/firewall Warning It will overwrite your existing configuration! It is strongly recommended to make a backup of /etc/sysconfig/firewall before saving your settings. 8. Graphical interface (X11) 8.1. Configuring your graphics card If you install X, setup will run xconfig, our X configuration utility automatically. First, xconfig will detect your configuration and will create a basic configuration file. Then it will ask you to specify the screen resolution and colour depth. Finally, it will create the real configuration file, with the following extras: * Automatically configures the mouse wheel, if found * Sets keyboard layout according to the selected language * Enables DPMS * Enables users to use the DRI extension After generating the config file, setup will start the X server. You must click the OK button to confirm to xconfig that the configuration was successful. 8.2. 3D acceleration, binary drivers If there is built-in 3d acceleration support for your card in X, xconfig will add the necessary entries to /etc/X11/xorg.conf and X will load the module(s). If you have an ATI or NVIDIA card, you probably need the manufacturer's binary drivers. Obtaining the NVIDIA binary driver is fairly simple: # pacman-g2 -Sy nvidia If you have an ATI card, installation will not be more complicated than a simple # pacman-g2 -Sy fglrx 8.3. Allow root login in KDM/GDM By default, no root login is permitted on the GUI, the recommended way of running graphical programs as root is to use gksu or kdesu. To enable it anyway, the following lines should be edited: For KDM (/usr/share/config/kdm/kdmrc) AllowRootLogin=false modify to AllowRootLogin=true For GDM (/etc/gdm/gdm.conf) AllowRoot=false modify to AllowRoot=true 9. Sound 9.1. Configuring the sound card Frugalware uses the Advanced Linux Sound Architecture (ALSA) subsystem for sound cards. For older applications, the Open Sound System (OSS) compatibility modules are loaded, but Frugalware does not contain native OSS support. Finding and loading the necessary module for your sound card is fairly simple. The process is mostly the same as setting up your network card. During every boot, the hotplug scripts will detect your sound card, but, of course, you can take the automatically loaded module to blacklist, and load it manually by editing /etc/sysconfig/modules. 9.2. Volume configuration with alsamixer By default, your sound card can be very loud. You can use alsamixer to set the volume of your card. Use the < and > keys to mute a channel, up and down keys to set the volume and left or right keys to switch to another channel. You can quit alsamixer by hitting the Esc key. From now, during shutdown, Frugalware saves your settings, but you can store or load them any time with the # service alsa save and the # service alsa load command. 10. Printing Frugalware uses the Common Unix Printing System (CUPS) for handling printers and to manage printing. 10.1. Before you start Here comes a few advice depending on what manufacturer made your printer. 10.1.1. Hewlett-Packard You need hpijs at least, but you can also install hplip for advanced HP support. Also if you have got some priter&scanner machine it's a good idea to use hplip. 10.1.2. Canon Most likely you need one of the bjfilter packages. The following list tell you which package you should use. * bjfilter-2.2: Canon Pixus 550i / 850i / 950i (i550 / i850 / i950) and iP90 Driver * bjfilter-2.4: Canon Pixus 560i / 860i / 960i (i560 / i860 / i960) Driver * bjfilter-2.5: Canon Pixus iP3100 / iP4100 / iP8600 (and Pixma iP1000 / iP1500) Driver * bjfilter: Canon Pixus iP2200 / iP4200 / iP6600D / iP7500 / MP500 Driver Please report us if your printer does not listed or listed, but in the wrong line! 10.1.3. Epson If you own an Epson Color InkJet Printer you need the pipslite package. After installing the package do not forget to restart cups and start the ekp daemon! sudo service cups restart sudo service ekpd start sudo service ekpd add Note Till now nobody confirmed that this package actually works. 10.1.4. Samsung The Samsung printer driver for cups is called splix. After installing it and restarting cups you will find your printer when you add it in cups. 10.2. Configuring the printer 1. Open your favorite Internet browser and go to [12]http://localhost:631. This is the Web interface of CUPS. 2. Select Administration from the top menu. If a username is required, type root, and give your root password. 3. You can do almost everything here in connection with printing. In our example, we will add a new local printer. 4. Click Add Printer, type in a name and optionally fill the Location and Description lines, then click on continue. 5. Select Device, in most cases it is Parallel Port #1 for older models and one of the USB ports for newer ones. I you have got a USB printer cups will write the printer name next to the proper port. 6. On the next page, select your vendor and your printer type (the driver/filter). To set up a remote Windows share with password, give a string like this for location (the share name is the printer's assigned name on the remote system): smb://user:passwd@Netbios_Name_or_ip_address/Share_name Notice that, when you view the printer configuration, the credentials will not be shown but will be used. 10.3. My printer is not listed If your vendor or printer type isn't listed in the wizard, you have to check [13]http://openprinting.org/printer_list.cgi wether if is supported under Linux or not. Usually it's enough to install the proper printer driver (see above) or gutenprint. After installing do not forget to restart cups: # service cups restart If it's not on the webpage mentioned above, then try to google after. If listed but said to be "paperweight", then there is nothing to do. If it is supported and said to be working on the site, then please file a bug report with your printer details. While we fix the bug, you can install the driver (the ppd) by yourself. On the left side, select Printer Listings. Then select your device's vendor and proper type. On the results page, select download PPD. After download, there will be a file named someting_that_ends_with.ppd. Save the PPD file in the directory /usr/share/cups/model/. The PPD file doesn't have to be executable, but it should be world-readable and should have the file extension ".ppd". If you do not want to search ppd, try to install foomatic-filters-ppds package. It has a bunch of ppd files for various printers. Then restart the CUPS service: su -c 'service cups restart'. The driver installation is now completed, now you can add your printer via the web interface. A good howto can be found at [14]http://www.linux-foundation.org/en/OpenPrinting/Database/CUPSDocume ntation. 10.4. Multiple pages on a single sheet This is also known as n-up printing. If an application doesn't support it natively, print the document to a file as PostScript and use psnup: $ psnup -2 print.ps > print2page.ps The first option specifies the number of pages stacked on one physical sheet, the second is the filename of the original one-sided document, and the last is the n-up (two-sided) document. You can then print it with $ cupsdoprint -P nameofprinter foo.ps or open it in your favourite PS viewer. 10.5. Troubleshooting If something goes wrong, check out CUPS log at /var/log/cups. There is a verbose error log and an access log, too. 11. The hotplug subsystem 11.1. udev The /dev directory under Frugalware is a ramdisk. Every device node is created automatically during the system boot by the hotplug subsystem, more specifically, by udev. It means, there won't be unnecessary device nodes in /dev, but it also means, if you create a device node manually, it will exist only until the next shutdown/reboot. If you want to force Frugalware to create a device node "manually" during each boot, you must create a device file under /lib/udev/devices: it will be copied on each boot automatically. The udev needs sysfs, so it will only work with the 2.6.x kernel series. Do not try to run udev on Frugalware with kernel series 2.4.x. 11.2. Pen/Thumbdrives Pendrives (also known as thumbdrives) are well-supported through the hotplug scripts and udev. If you insert a pendrive into the USB slot, udev will create a device node for it in /dev. Most pendrives contain only one partition and their filesystem is vfat. In most cases, the pendrive will behave like a SCSI disc. It means, you can find the pendrive under /dev/sda and its first partition under /dev/sda1. Adding the following line to /etc/fstab: /dev/sda1 /media/pendrive auto defaults,noauto,user 0 0 will allow users mount their pendrive if the device node exists (if the device is inserted into the slot). If you use KDE, Gnome or XFCE4 they will handle automatic mounting of such devices. You should not edit /etc/fstab as automounting will not work for you. For blackbox, fluxbox, englightenment, e17 and other smaller window manager users there is ivman for automounting, but do not expect as fine work as in KDE, Gnome, XFCE4. See also the automounting part of the documentation. 11.3. Digital cameras Tyipcally, there are two types of digital cameras. Some of them support both access methods, others use only one of them. First, most of the cameras can be treated as a pendrive (USB Mass Storage device), you can mount them and copy the pictures from them easily. Other cameras support the Picture Transfer Protocol (PTP). You can grab the pictures from them (and do lots of other actions) with gphoto2, if your model is supported. (If it's not available on your system, a simple su -c 'pacman-g2 -S gphoto2' will install it onto your system.) 11.4. Automounting: D-BUS, HAL and Ivman; Gnome and KDE D-BUS is a simple IPC (inter-process communication) library based on messages. HAL is a hardware abstraction layer which uses D-BUS. Ivman is based on HAL and uses pmount ("policy mount"), which is a wrapper around the standard mount program which permits normal users to mount removable devices without an existing /etc/fstab entry. Ivman is a daemon to automount CD-ROMs and DVDs when inserted in a drive, or play audio CDs or video DVDs automatically. It is 100% userspace, so it is a safe replacement for submount. If you want to change the default settings, all config files are located in /etc/ivman. They are plain XML files, just read them, everything is quite self-explanatory. Automounting also happens with KDE and Gnome, but their respective VFS implementation do that, not ivman. Ivman is useful for other windowing systems where is no support for such a feature. 12. The init scripts, bootup 12.1. About the kernel The Linux kernel is in the kernel package. We're trying to use as few patches as possible to stay close to the vanilla kernel. We also use [15]splashy instead of well known bootsplash. The kernel contains compiled-in support for most IDE controllers, but all low-level SCSI drivers are compiled as a module. If Frugalware's kernel doesn't contain built-in support for your controller, you can compile your own kernel. Don't worry, it's fairly simple. 1. After setup is finished, before hitting ENTER to reboot, switch to tty2 by pressing Alt-F2 and press ENTER to get a shell. 2. Change your root directory to /mnt/target: # chroot /mnt/target 3. The source of your kernel (with additional patches applied) can be found at /usr/src/linux. So go to the /usr/src/linux directory and enter the configuration menu by typing make menuconfig. Inside it, select the driver you don't want to compile as a module anymore, and exit from the menu with saving changes. 4. Compile your kernel with the make command. This may take several minutes. 5. Copy your new kernel to /boot by typing the following command: # cp /usr/src/linux/arch/$yourarch$/boot/bzImage /boot/vmlinuz On x86, $yourarch$ has to be replaced by i386. 12.2. Init scripts and services In Frugalware, init scripts are always called rc.something and they are located in /etc/rc.d. They are used to setup the environement and also allow to manage system services. The services are UNIX daemons that provides various kind of service. The spectrum of their actions are very large. Synchronizing your system clock, running your webserver, running the virus scanner, all of these are services and they offer much much more. The files that allow to manage them can be found in /etc/rc.d, but usually you will prefer to use our utility service. This tool allows you to control the running state of the services. In the following examples we will explain how to alter the running state of a given service. You will have to replace $service_name$ with the wanted service name. As you will see the syntax is simple, and you may get more help looking and the service manual doing: $ man service Important Later in this document you will see how to alter the configuration of these services so that they follow your needs. You should better learn how to control them, but don't be afraid, the syntax is really simple, and you will learn it in less then a minute. 12.2.1. Controlling a service execution Services can be started, restarted and stopped, so that you can control what your system has to offer. To start a service, simply do: # service $service_name$ start To restart a service, simply do: # service $service_name$ restart To stop a service, simply do: # service $service_name$ stop As you can see, controlling a service execution is pretty simple. 12.2.2. Controlling a service execution on system boot Controlling the automatic execution of services on system startup is not much more difficult. To add a service for automatic execution on system startup, simply do: # service $service_name$ add To delete a service for automatic execution on system startup, simply do: # service $service_name$ del To list the runlevels in which the service will be running, simply do: # service $service_name$ list 12.3. System boot, runlevels If you don't pass any extra init=/path/to/init parameters to it, the kernel will start /sbin/init as the final step of the kernel boot sequence. According to the content of /etc/inittab, init will run: 1. each S* script at /etc/rc.d/rcS.d 2. each S* script at /etc/rc.d/rcn.d, where n is the default runlevel. This is set to 4 by default. Here is the list of available runlevels: 0 = halt 1 = single user mode 2 = unused (but configured the same as runlevel 3) 3 = multiuser mode (text mode) 4 = multiuser mode, X11 with KDM/GDM/XDM (default Frugalware runlevel) 5 = unused (but configured the same as runlevel 3) 6 = reboot If X11 is configured, /etc/rc.d/rc.4 will start one of the desktop managers, as configured in /etc/sysconfig/desktop. 12.4. GRUB gfxmenu Frugalware comes with a nice graphical grub menu (thanks to SuSE's gfxmenu developers). If you don't like it, you can disable it by commenting out the gfxmenu initialization line in /boot/grub/menu.lst. So for example: Before: gfxmenu (hd0,5)/boot/grub/message After: #gfxmenu (hd0,5)/boot/grub/message 12.5. Splashy Frugalware uses splashy to display nice splash screen and a progress bar instead of text messages during the boot procedure. Splashy is completely user-space, so there is no need for patching the kernel. If you dislike it or want to switch it off for whatever reason add nosplashy for your kernel parameters in /boot/grub/menu.lst. For example: kernel (hd0,2)/vmlinuz root=/dev/hda5 ro quiet vga=791 nosplashy 13. How to contribute If you find our work and effort worth it, please consider donating. This is not limited to monetary contributions, every donation of a limited valuable resource (including your time, knowledge) is appreciated. Depending on your resources, there are many ways to help us. 13.1. Translation A comprehensive and multi-language documentation is very important. With linguistic and no programming knowledge and some time, you can help us create (by asking) and maintain different translations. 13.2. Application packaging In the [16]Bug Tracking System, there are feature requests for some packages. The process of making packages is well documented, and with some GNU/Linux experience, it isn't difficult. But it takes time, so submitting well-packaged software is a great way to help us and save our time. 13.3. Developing Of course, any skilled help is appreciated in developing core systems, like contributing code to pacman-g2 or the setup. 13.4. Donating hardware By sending us some wanted hardware (see [17]donations), you can make testing packages easier, or speed up the package creation process within a specific architecture. 13.5. Artwork We usually update our artwork (background images, grub splash, desktop manager themes, window manager splashes and so on) for each release. If you are skilled to help, you're welcome. :) 13.6. Support If you have time and knowledge, monitor the forums, read the mailing list posts, hang around on IRC and try to answer the questions, solve the occured problems. 13.7. Find bugs If you find bugs, you can help with submitting well-written bugreports, see the Reporting Bugs section for more info. 14. The Frugalware Bugreporting HOWTO 14.1. Introduction The aim of this HOWTO is to explain how to choose a task name and what to include in a feature request/bugreport to help Frugalware developers speed up the process of fixing a bug or fulfilling a feature request. 14.2. Where The URL of our Bug tracking system is: http://bugs.frugalware.org/ 14.3. General Before opening a task, use the search function, maybe there is a task for your bug/feature. In that case just add a comment about "I can reproduce this, too." or "I would enjoy this feature, too." If you'd like to report outdated package make sure that it isn't listed on [18]this site. When your package is listed please do not report it as we know there is a new version and we will update it as soon as possible. Write bugreports in English, please. This is the only language all developers speak. 14.4. Bugreport Please include the following things, unless you know what you are doing: 1. Description of Problem - never say "does not work", quote the error message 2. Steps to reproduce the problem 3. Actual Results 4. Expected Results 5. How often does this happen? 6. Additional Information The default arch is i686 and the default version is -current. If these are not true, don't forget to change them! If you report a -current installer bug, then maybe -current is not enough, please specify the snapshot date. If you found a security bug, then use the [SEC] prefix in the task name. 14.5. Feature Requests Please don't request more than one package in a feature request. Open a task for every package. (Of course you don't have to open task for dependencies if they are also missing from out packages.) If you request a package, please include: . The name of the application (yes, "more games" is not enough!) . The URL of the application . Optionally a short note about why do you think this package would be interesting for others, too If you have a FrugalBuild for the package already, then after opening the task, upload it as an attachment. In this case, please prefix your task name with [FB], because this way it'll be reviewed sooner. Alternatively, you can post your FrugalBuild to the frugalware-devel mailing list for review, that can be handy if you want to submit more and more buildscript - finally to become a developer if possible. Opening a task for your FrugalBuild is still fine if you want us to maintain it after the initial version is accepted. Please don't link other distribution's buildscripts when you request a package. That information is useless for us in most cases and if you don't include such links, you make our life easier. 14.5.1. Don't request Please don't request custom kernels. We try to use as less patches as possible. See man kernel.sh as a reference on building your own packages using various patchsets. Also a [19]tutorial is available. Really, building such a kernel usually requires a buildscript of only 5 lines! 14.6. Pacman-g2 problems If you get a crash from our package manager then we need a backtrace from gdb. Here are the instructions to get a backtrace: * Find the command line that triggers the crash. For example: pacman-g2 -Sy * Get the pacman-g2 git repo and compile it with debug symbols enabled: $ git clone git://git.frugalware.org/pub/other/pacman-g2/pacman-g2 $ cd pacman-g2 $ sh autogen.sh $ ./configure --enable-debug $ make * Then run pacman-g2 in gdb and get the trace: $ cd src/pacman-g2 $ sudo libtool gdb ./pacman-g2 > run -Sy * When pacman-g2 crashed, get the trace by typing bt. Here is an example: Program received signal SIGSEGV, Segmentation fault. 0x0805035e in pacman_sync (targets=0x0) at sync.c:354 354 *p = 1; (gdb) bt #0 0x0805035e in pacman_sync (targets=0x0) at sync.c:354 #1 0x08054594 in main (argc=2, argv=0xbfee1844) at pacman.c:609 * Attach the output of bt to your bugreport. 14.7. Fixed in git Your feature request / bugreport may be closed with a "Fixed in git ..." message. Git is our source control management software (just like CVS). If your task is not considered to be important, then it will be fixed/implemented only in git, without increasing the package release. This means that it will be automatically included in the next release. 15. Mobile computers 15.1. Battery, buttons, thermal management Notebook users are usually interested in the state of their battery. To get the power button and the lid's sensor of its closed state emit events is also nice. Some notebooks only shut down their continously running fans and operate only if needed if the thermal module is loaded. Usually the following steps are required to enable this functionality: Adding the following lines to /etc/sysconfig/modules to get modules loaded at system startup: battery ac button thermal The next task is to enable the acpid service: # service acpid add Then the easiest way is to reboot, or if you don't want do do so: # modprobe battery # modprobe ac # service hald stop # service dbus stop # service acpid start # service dbus start # service hald start The only remaining task is to start a client: if you're on console, try the acpi command, or the relevant applet of your favorite window manager. 15.2. Conserving power The major consumers of power in a notebook are the LCD (size and brightness level), the CPU, hard drives, wireless transceivers like WiFi, Bluetooth, Infrared and the GPU if you have a powerful one. You can conserve a fair amount of power if you lessen the brightness level of the LCD screen. Some notebooks can remember two settings of this level, one when the equipment operates from battery and for another when powered from AC. The CPUs have some sort of power saving capabilities, the most basic is "CPU throttling". Common on Intel mobile Celeron CPUs, only ACPI is needed. Klaptop has a setting for it, where you can specify the level. Letting the HDD spin down gives little extra battery operating time, but frequent spinups (data access) and spindowns wears the disk. Only useful in situations where there is no frequent need for data on hdd like holding a presentation. 15.3. Hibernation Hibernating your computer can cause data loss or severe filesystem damage if things go wrong. It's highly advised that first, you should consider if hibernating is worth the effort at all. Try it on a fresh installation first, instead of a production system. From kernel/suspend.c: * BIG FAT WARNING ********************************************************* * * If you have unsupported (*) devices using DMA... * ...say goodbye to your data. * * If you touch anything on disk between suspend and resume... * ...kiss your data goodbye. * * If your disk driver does not support suspend... (IDE does) * ...you'd better find out how to get along * without your data. * * If you change kernel command line between suspend and resume... * ...prepare for nasty fsck or worse. * * If you change your hardware while system is suspended... * ...well, it was not good idea. * * (*) suspend/resume support is needed to make it safe. You have been warned. If you are still not discouraged, read on! First, you need to create a swap partition (if you don't have any yet). You have to add an extra resume=/dev/swappart kernel parameter to /boot/grub/menu.lst. For example, on my machine the old line was: kernel (hd0,2)/boot/vmlinuz ro root=/dev/hda3 quiet vga=788 The new line: kernel (hd0,2)/boot/vmlinuz ro root=/dev/hda3 quiet vga=788 resume=/dev/hda2 After the above are done, you must reboot. The hibernation can be started with: echo shutdown > /sys/power/disk; echo disk > /sys/power/state and next time you boot your kernel it should resume. For more info, look at /usr/src/linux/Documentation/power/swsusp.txt. It requires the kernel documentation, which can be installed issuing the pacman-g2 -S kernel-docs command as root. 16. Packages The following sections describe the configuration of some packages. 16.1. acoc In order to use acoc you should start it with $ acoc for example, or you can create an alias like this: alias pacman='acoc pacman' 16.2. amavisd-new From amavisd-new-2.5.2-1 we no longer use a random uid/gid, but dedicated ones. Because of this amavis service will not start if you have it installed before, so you have to correct this by issuing these commands: groupmod -g 40 amavis usermod -u 40 -g 40 amavis chown -R amavis:amavis /var/lib/amavis chown -R amavis:amavis /var/lock/amavis You should chown any other amavis-owned stuff you may have lying around, these are only the default ones. 16.3. apache 16.3.1. How to configure Apache 1. These steps require root privileges, so use su - to get a root shell. 2. The Apache server isn't started by default. You can change this with the # service httpd add command. 3. We don't want to reboot, so start it manually: # service httpd start Starting Apache web server (no SSL) [ OK ] You have finished if you don't need SSL support. 16.3.2. Setting up SSL support for Apache 1. Creating the certifications: # cd /etc/httpd/conf/ # sh mkcert.sh Signature Algorithm ((R)SA or (D)SA) [R]: Here we can accept the default RSA signature algorithm first. Then we have to fill out some fields. There are quite a few fields but you can leave most of them blank. If you enter '.', the field will be left blank. 1) Country Name (2 letter code) [XY]: Give the 2-letter code of our contry (for example US) 2) State or Province Name (full name) [Snake Desert]: We type our state. 3) Locality Name (eg, city) [Snake Town]: The name of our city. 4) Organization Name (eg, company) [Snake Oil, Ltd]: Our organization's name. 5) Organizational Unit Name (eg, section) [Webserver Team]: Our section's name. 6) Common Name (eg, FQDN) [www.snakeoil.com]: Important: Give a real address here, otherwise you'll get warnings in your browser! 7) Email Address (eg, `name@FQDN') [`www@snakeoil.com']: I usually give the email address of the webmaster here. (webmaster@domain.com) 8) Certificate Validity (days) [365]: In most cases, one year will be good. Then, we should choose the version of our certificate: Certificate Version (1 or 3) [3]: The default 3 will be good, so just hit enter. In the next step we can encrypt our private key: Encrypt the private key now? [Y/n]: The keys will not be readable by users, so we can leave this step out. So the following files are created: /etc/httpd/conf/ssl.key/server.key (keep this file private!) /etc/httpd/conf/ssl.crt/server.crt /etc/httpd/conf/ssl.csr/server.csr 2. Enable SSL in /etc/httpd/conf/httpd.conf: Open the file with your favorite editor, and search the followings at about line 1040: # Uncomment this if you want SSL support! # # Include /etc/httpd/conf/ssl.conf # Uncomment them. 3. Now we should restart Apache: # service httpd restart 4. Then we can check if the task was successful: $ elinks https://localhost/ This should show the default homapage, received via SSL :) 16.3.3. Self-signed Apache certificate This must be done as root. # openssl genrsa -des3 -out server.key 1024 Enter "foobar" twice as passphrase. # openssl req -new -key server.key -out server.csr Enter "foobar" when asked for passphrase, answer the questions. Leave "challenge password" "and optional company name" empty. # cp server.key server.key.org # openssl rsa -in server.key.org -out server.key Enter "foobar" when asked for passphrase. # openssl x509 -req -days 365 -in server.csr -signkey server.key -out server.crt # cp server.crt /etc/httpd/conf/ssl.crt/ # cp server.key /etc/httpd/conf/ssl.key/ # service httpd stop # vi /etc/httpd/conf/httpd.conf Uncomment the marked three lines around line 1044 (look for "SSL support"). # service httpd restart Don't forget to open port 443 on your firewall, if any. (Based on [20]How to create a self-signed SSL Certificate..., tested on frugalware-current 2007-02-14.) 16.4. avahi Warning If you have rlocate installed on your system, Avahi will not run and therefore Zeroconf functionality in programs will be disabled. If you want this functionality, then please uninstall rlocate. Also, If you are using iptables, please uncomment this line in /etc/sysconfig/firewall: #-A INPUT -p udp -m udp --dport 5353 -j ACCEPT After that do not forget to restart iptables with: # service firewall restart 16.5. b2evolution After installing this package, please run # /usr/bin/b2evosetup to setup B2evolution. 16.6. bcm43xx-fwcutter Since version 2.6.24, the bcm43xx driver is deprecated, replaced by the b43 and b43legacy modules. The module should be loaded automatically, in case it isn't, you can load it manually: # modprobe b43 or: # modprobe b43legacy You must bring the device up with ifconfig before doing any other configuration steps. # ifconfig ethX up Since the channel must be set manually, first do a scan: # iwlist ethX scan Then you can set it: # iwconfig ethX channel Y Finally set your essid: # iwconfig ethX essid "myessid" Ready! 16.7. bitlbee-skype Please read the README file in the documentation directory of the package on how to fine-tune the configuration file of skyped and on how to generate the SSL certificates for it. 16.8. cairo-clock Cairo-Clock requires the Composite option to be enabled in your Xorg configuration. To enable it, add the following lines to /etc/X11/xorg.conf: Section "Extensions" Option "Composite" "Enable" EndSection 16.9. ccache After you installed ccache, it won't be enabled by default. First, you need to determine who is allowed to use ccache. You have to add each user to the ccache group. If you want to allow using ccache from chrooted builds, then you need to add the fst user: # usermod -a -G ccache fst Second, you need to somehow let the build system to use ccache, and not the compiler directly. If you use makepkg, this is enabled by default (you can disable it with the -B option). If you build manually, then you are on your own, though usually there are two ways to do so: * Tell the configure script to use a different compiler: $ CC=/usr/bin/ccache ./configure * Modify path to use the fake compiler provided by ccache: export PATH=/usr/lib/ccache/bin:$PATH 16.10. cpuspeed After installing cpuspeed, make sure you edit the configuration file before starting it. The configuration file is located in /etc/cpuspeed.conf. Set the correct CPUFreq driver name in the confiuration file by setting the DRIVER value. for eg: if you want to use the p4-clockmod driver, your cpuspeed configuration file should contain: DRIVER="p4-clockmod" For a list of drivers, check this directory /lib/modules/your_kernel_version/kernel/arch/i386/kernel/cpu/cpufreq 16.11. cryptsetup-luks Follow these steps to when using cryptsetup-luks: 16.11.1. Creating # cryptsetup luksFormat /dev/partition # cryptsetup luksOpen /dev/partition label # mke2fs -j /dev/mapper/label # mount /dev/mapper/label /mnt/label 16.11.2. Mounting Of course later you don't have to use luksFormat and mke2fs: # cryptsetup luksOpen /dev/partition label # mount /dev/mapper/label /mnt/label 16.11.3. Umounting # umount /mnt/label # cryptsetup luksClose label 16.11.4. Encrypting your home partition Note You have need to install the sharutils package to do the followings! * List these modules in /etc/sysconfig/modules: aes aes-i586 sha256 dm-crypt * Move all data from /home to a secure place (in this example /media/sda1/home) # cp -arvx /home /media/sda1/ * Umount /home (in this example /dev/hda6) and fill it with random numbers: # umount /home # dd if=/dev/urandom of=/dev/hda6 * Create the encrypted partition: # cryptsetup -y luksFormat /dev/hda6 Here we will be asked for a password which will be necessary to access /home at boot time. * Open the encrypted partition and create its file system (ext3 in this example): # cryptsetup luksOpen /dev/hda6 home # mkfs.ext3 /dev/mapper/home * Mount the home partition and copy the contents of original home: # mount /dev/mapper/home /home # cp -arvx /media/sda1/home /home * Edit the home related line in /etc/fstab: /dev/mapper/home /home ext3 noatime 0 0 * Create /etc/rc.d/rc.crypt script with the following content: #!/bin/sh /usr/sbin/cryptsetup luksOpen /dev/hda6 home /bin/mount /dev/mapper/home /home * Enable it: # ln -s /etc/rc.d/rc.crypt /etc/rc.d/rcS.d/S15rc.crypt You have to delay the splash screen, so that you can type your password before the splash appears: # mv /etc/rc.d/rcS.d/S03rc.splash /etc/rc.d/rcS.d/S15rc.splash (It will ask the password between the lvm and the splash service.) Now the system can be restarted and the password will be asked to access home partition boot-time. Note The English keyboard map will be used at that point of the boot process. 16.12. cyrus-sasl 16.12.1. Configuring This mini-howto helps you to install the saslauthd server using postfix which will authenticate using users and passwords from /etc/{passwd,shadow}. First install the necessary packages: # pacman-g2 -S postfix saslauthd Enable sasl in postfix's config by appending the following lines to /etc/postfix/main.cf: smtpd_sasl_auth_enable = yes smtpd_sasl_local_domain = $myhostname smtpd_sasl_security_options = noanonymous You may want to append broken_sasl_auth_clients = yes as well. Put the following lines to /usr/lib/sasl2/smtpd.conf: pwcheck_method: saslauthd mech_list: PLAIN LOGIN Edit /etc/sysconfig/saslauthd by changing the following lines: SASL_DIE=1 to SASL_DIE=0 and auth_mechanism="" to auth_mechanism="shadow" Now you can start saslauthd by service saslauthd start as well as enabled in by default on startup: service saslauthd add Issue id postfix and see if the daemon group is listed. If not, then add postfix to the daemon group: usermod -G daemon postfix Finally restart postfix: service postfix restart Compeleted! 16.12.2. Verifying We test it using telnet. We need perl to generate the string for the SASL authentication: $ perl -MMIME::Base64 -e 'print encode_base64("vmiklos\0vmiklos\0secret");' dm1pa2xvcwB2bWlrbG9zAHNlY3JldA== Then use telnet: $ telnet host.com 25 Trying ip... Connected to host.com. Escape character is '^]'. 220 host.com ESMTP Postfix ehlo my.dhcp 250-host.com 250-PIPELINING 250-SIZE 10240000 250-VRFY 250-ETRN 250-AUTH LOGIN PLAIN 250-ENHANCEDSTATUSCODES 250-8BITMIME 250 DSN AUTH PLAIN dm1pa2xvcwB2bWlrbG9zAHNlY3JldA== 235 2.0.0 Authentication successful quit 221 2.0.0 Bye Connection closed by foreign host. 16.13. dante 16.13.1. Configuration In most cases you have a socks server (you can create one easily using ssh, see the documentation of the openssh package), and you want to route all traffic through it. Here is the config you need: route { from: 0.0.0.0/0 to: 0.0.0.0/0 via: 127.0.0.1 port = 8080 proxyprotocol: socks_v4 } 16.13.2. Testing it Try for example: $ socksify irssi When you connect to a server, others will see that you're connecting from the server, not from your own host. 16.14. darcs First, please note that darcs comes with a very good HTML documentation, which is available under the /usr/share/doc/darcs-*/manual dir. That's the place where everything is properly documented, not the manpage. Using darcs [subcommand] -h is usable only as a reference, too. If you're completely new to darcs, then start at /usr/share/doc/darcs-*/manual/node4.html. Please also note that in order for the darcs send command to work properly, you must properly configure your mail transport agent to relay outgoing mail. For example, if you are using postfix, you need to edit /etc/postfix/main.cf, see the Using a relay host part of the postfix package documentation for more info. 16.15. dazuko If you got errors saying: dazuko: failed to register then you need to do: # rmmod capability # modprobe dazuko # modprobe capability It will work. 16.16. ddclient Please configure /etc/ddclient/ddclient.conf before running ddclient! Samples for common configurations can be found in: /usr/share/doc/ddclient-$package_version/sample* Additional details and instructions can be found in: /usr/share/doc/ddclient-$package_version/README Once you have finished configuring the ddclient.conf file, you can start ddclient as a daemon by running as root, the following command: # service ddclient start 16.17. dhcp If you are in trouble setting up your dhclient, use the following options. These are quite good defaults: request subnet-mask, broadcast-address, time-offset, \ routers, domain-name, domain-name-servers, \ host-name, netbios-name-servers, netbios-scope; timeout 20; script "/sbin/dhclient-script"; 16.18. dillo Dillo is moving to FLTK2. As of 2006-04-29 it is unstable (both unreleased and has some random erratic bugs, but upstream is working on it), and some KDE stuff has fltk dep, so not a good idea to mess with. Dillo is now "crippled" by removing the new FLTK based download GUI, as it is only this needs FLTK2, but for the next release more FLTK2 expected. 16.19. drupal After installing this package, please run /usr/bin/drupalsetup as root to setup Drupal 16.20. drupal6 To be able to use this package as intended, you will have to: * set up apache to access /var/www/drupal6 from the web the way you like; * install and set up your favourite SQL database (mysql or postgresql; this package DOES NOT depend on any of them); * create and/or grant access to a mysql or postgresql database; * set up your drupal installation itself by entering the correct credentials at the install screen to be able to reach the above-mentioned database. 16.21. dspam To populate the DSPAM database, you need to follow several steps. 1. First create a database. Login to the mysql command prompt. $ mysql -u root -p mysql> CREATE database dspam; 2. Next, you need to create a dspam user. At the same MySQL prompt: mysql> GRANT ALL PRIVILEGES ON dspam.* TO dspam@'localhost' IDENTIFIED BY 'passw d'; Replacing passwd with your chosen password. 3. Optimizing the datebase: If you want a space optimized db do: $ mysql -u dspam dspam -p < /var/lib/dspam/mysql/mysql_objects-space.sql If you want a speed optimized db do: $ mysql -u dspam dspam -p < /var/lib/dspam/mysql/mysql_objects-speed.sql Enter the password you set in the previous step, and the database should be populated. 4. Remember to edit /etc/dspam/dspam.conf accordenly If you want to use the postgresql, sqlite3 or Berekely DB4 backends you can find instructions in the dspam documentation. 16.22. eaccelerator 16.22.1. Setting up eaccelerator In order to use eAccelerator, you must add the following lines to your /etc/php.ini file: extension="extensions/no-debug-non-zts-20060613/eaccelerator.so" eaccelerator.shm_size="16" eaccelerator.cache_dir="/tmp/eaccelerator" eaccelerator.enable="1" eaccelerator.optimizer="1" eaccelerator.check_mtime="1" eaccelerator.debug="0" eaccelerator.filter="" eaccelerator.shm_max="0" eaccelerator.shm_ttl="0" eaccelerator.shm_prune_period="0" eaccelerator.shm_only="0" eaccelerator.compress="1" eaccelerator.compress_level="9" Do not forget to create the cache directory as well: mkdir /tmp/eaccelerator chmod 0777 /tmp/eaccelerator 16.22.2. Configuration Options: eaccelerator.shm_size The amount of shared memory (in megabytes) that eAccelerator will use. "0" means OS default. Default value is "0". eaccelerator.cache_dir The directory that is used for disk cache. eAccelerator stores precompiled code, session data, content and user entries here. The same data can be stored in shared memory also (for more quick access). Default value is "/tmp/eaccelerator". eaccelerator.enable Enables or disables eAccelerator. Should be "1" for enabling or "0" for disabling. Default value is "1". eaccelerator.optimizer Enables or disables internal peephole optimizer which may speed up code execution. Should be "1" for enabling or "0" for disabling. Default value is "1". eaccelerator.debug Enables or disables debug logging. Should be "1" for enabling or "0" for disabling. Default value is "0". eaccelerator.check_mtime Enables or disables PHP file modification checking . Should be "1" for enabling or "0" for disabling. You should set it to "1" if you want to recompile PHP files after modification. Default value is "1". eaccelerator.filter Determine which PHP files must be cached. You may specify the number of patterns (for example "*.php *.phtml") which specifies to cache or not to cache. If pattern starts with the character "!", it means to ignore files which are matched by the following pattern. Default value is "" that means all PHP scripts will be cached. eaccelerator.shm_max Disables putting large values into shared memory by " eaccelerator_put() " function. It indicates the largest allowed size in bytes (10240, 10K, 1M). The "0" disables the limit. Default value is "0". eaccelerator.shm_ttl When eaccelerator fails to get shared memory for new script it removes all scripts which were not accessed at last "shm_ttl" seconds from shared memory. Default value is "0" that means - don't remove any files from shared memory. eaccelerator.shm_prune_period When eaccelerator fails to get shared memory for new script it tryes to remove old script if the previous try was made more then "shm_prune_period" seconds ago. Default value is "0" that means - don't try to remove any files from shared memory. eaccelerator.shm_only Enables or disables caching of compiled scripts on disk. It has no effect on session data and content caching. Default value is "0" that means - use disk and shared memory for caching. eaccelerator.compress Enables or disables cached content compression. Default value is "1" that means enable compression. eaccelerator.compress_level Compression level used for content caching. Default value is "9" which is the maximum value eaccelerator.keys eaccelerator.sessions eaccelerator.content Determine where keys, session data and content will be cached. The possible values are: "shm_and_disk" - cache data in shared memory and on disk (default value) "shm" - cache data in shared memory or on disk if shared memory is full or data size greater then "eaccelerator.shm_max" "shm_only" - cache data in shared memory "disk_only" - cache data on disk "none" - don't cache data eAccelerator API: eaccelerator_put($key, $value, $ttl=0) puts the $value into shard memory for $ttl seconds. eaccelerator_get($key) returns the value from shared memory which was stored by eaccelerator_put() or null if it is not exists or was expired. eaccelerator_rm($key) removres the $key from shared memory eaccelerator_gc() removes all expired keys from shared memory eaccelerator_lock($lock) creates a lock with specified name. The lock can be released by function eaccelerator_unlock() or automatic on the end of request. For Example: eaccelerator_unlock($lock) release lock with specified name eaccelerator_set_session_handlers() install the eaccelerator session handlers. Since PHP 4.2.0 you can install eaccelerator session handlers in "php.ini" by "session.save_handler=eaccelerator". eaccelerator_cache_output($key, $eval_code, $ttl=0) caches the output of $eval_code in shared memory for $ttl seconds. Output can be removed from cache by calling mmcach_rm() with the same $key. For Example: eaccelerator_cache_result($key, $eval_code, $ttl=0) caches the result of $eval_code in shared memory for $ttl seconds. Result can be removed from cache by calling mmcach_rm() with the same $key. For Example: eaccelerator_cache_page($key, $ttl=0) caches the full page for $ttl seconds. For Example: eaccelerator_rm_page($key) removes the page which was cached by eaccelerator_cache_page() with the same $key from cache eaccelerator_encode($filename) returns the encoded bytecode of compiled file $filename eaccelerator_load($code) loads script which was encoded by eaccelerator_encode() 16.23. ejabberd 16.23.1. Creating your SSL keys Generate Key Pair: # cd /etc/jabberd # openssl req -new -x509 -newkey rsa:1024 -days 3650 -keyout privkey.pem -out se rver.pem Note You should enter your domain name as the Common Name for your certificate. Remove pass parse: # openssl rsa -in privkey.pem -out privkey.pem Combine the Private and Public Key: # cat privkey.pem >> server.pem Delete Private Key: # rm privkey.pem Set permissions: # chown root:ejabberd server.pem # chmod 640 server.pem Finally update the config file: * Change the ./ssl.pem string to /etc/ejabberd/server.pam. * Change starttls to tls in the listen section if you want to force users to use SSL. 16.23.2. Creating an administrator Register an account on your ejabberd deployment. An account can be created using a jabber client like pidgin. Add the following lines to you config: {acl, admins, {user, "admin", "example.org"}}. {access, configure, [{allow, admins}]}. This will promote the account created in the previous step to an account with administrator rights. 16.23.3. Testing Add the following line to your /etc/sysconfig/firewall, for example after mysql: # ejabberd -A INPUT -p tcp -m tcp --dport 5222 -j ACCEPT Now you should be able to connect to ejabberd remotely. Start your favourite jabber client on a remote machine (ie. pidgin) and register another account. You should be able to talk to the admin now and vica versa. For more info, please read the Installation and Operation Guide, which can be found at /usr/share/doc/ejabberd-*/guide.html. 16.24. enemy-territory If you got disconnected from servers and getting some #20004 errors, then run as pbweb AS ROOT!!! Then try again :) Regards 16.25. entrace If your keyboard doesn't work that can be because wrong virtual terminal number set in entrance config. Use the following command to set the proper value: # ecore_config -c /etc/entrance_config.cfg -k /entranced/xserver -s "/usr/X11R6/ bin/X -quiet -nolisten tcp vt7" In this example you want to run entrance on vt7. 16.26. festival To test festival, try: $ echo "Frugalware can speak" | festival --tts 16.26.1. To test it with kttsd: 1. Start KTTSD (if not already running): kttsd 2. Send "Frugalware can speak" to KTTSD for speaking in English: $ dcop kttsd KSpeech setText "Frugalware can speak" "en" 3. Speak the text: $ dcop kttsd KSpeech startText 0 16.27. fglrx If hardware acceleration does not work make sure you have something similar in /etc/fstab: tmpfs /dev/shm tmpfs defaults 0 0 If xv output doesn't work in mplayer add the following line to /etx/X11/xorg.conf's Device section: Option "VideoOverlay" "on" On the following page you can find some useful solutions for different fglrx problems: [21]Thinkwiki fglrx. 16.28. firestarter This version comes with a system init script now. You have to run the firestarter executable from the command line (in an X driven console) first to generate the initial start-up scripts. To add it to startup, run this: # chkconfig --del rc.firewall # chkconfig --add rc.firestarter To remove it from startup: # chkconfig --del rc.firestarter # chkconfig --add rc.firewall 16.29. flowplayer Once you have the .flv file you want to share, you need FlowPlayer.swf from /usr/share/flowplayer and the following code sniplet: 16.30. foo2zjs This driver is under constant change, therefore no "stable" branch exists. Also, communications with the author led nowhere, that might explain some weirdness of building it, getting the latest stable version number etc. [22]Mail 16.31. freenx 1. Installation On the server: # pacman-g2 -S freenx On the client: # pacman-g2 -S knx 2. Configuration On the server setup a fake password for the nx user: # passwd nx Now allow the client to connect to the server by copying the file /usr/NX/home/nx/id_dsa from the server to /usr/share/knx/client.id_dsa.key on the client. Also make sure about it's readable by users: # chmod 644 /usr/share/knx/client.id_dsa.key Also don't forget to allow incoming ssh connections (by default port 22) in /etc/sysconfig/firwall, then load the config with the # service firewall restart command. 3. Users By default shell accounts are not allowed to connect to th nx server, you can allow a user to do so by # nxserver --adduser then set a separate (for nx) password for the user: # nxserver --passwd 4. The client Now start knx, the usage of that application is self-explaining. 16.32. fudforum After installing this package, please run /usr/bin/fudforumsetup as root to setup FUDforum 16.33. fuse Fuse is a virtual filesystem "helper" which makes possible to mount unusual things as a filesystem. It is achieved by using a simple program, which runs in user space, to provide data that can be represented by the fuse kernel module as a filesystem. The interpreter program is a less complex one than a kernel-space module, which is much harder to write. In Frugalware, regular users of a given box can mount filesystems by fuse. First as root, let's install the tools needed: # pacman-g2 -S fuse Then you have to add the fuse service to the startup list and start it manually for now: # service fuse add # service fuse start Now, having the base of fuse, we need to install the programs for each specific filesystem type. To get a hint on what is available, you can issue the following command: $ pacman-g2 -Ss fuse The two most used (ftp, ssh) plugins can be installed by running the following command. Beware, the ftp fs is a perl module, and it seems a bit memory hungry / buggy / slow so therefore it might be replaced by CurlFtpFS in the future. # pacman-g2 -S fuseftp sshfs-fuse Then, you can mount a remote dir with sftp access as a regular user doing: $ /sbin/mount.fuse sshfs#YOURUSERNAME@SERVER:/REMOTEDIR /LOCALDIR -o rw,OTHEROP TIONS You can also unmount it as a regular user doing: $ fusermount -u /LOCALDIR 16.34. gammu 16.34.1. Configuring You need to create your ~/.gammurc: [gammu] port = /dev/ttyUSB0 connection = fbus Replace /dev/ttyUSB0 with your serial port device and fbus with the appropriate protocol name if you are not a Nokia user. Check if you have write access to the device, you need to be a member of the uucp group. Once you think you're done, check your setup: $ gnokii --identify It should print your IMEI number so that you'll be able to check if gammu really found your phone or there is a problem. 16.34.2. Creating a backup You probably use gammu to make a backup of your phone. This involves two steps: * Backing up your SMSes $ gammu --backupsms backupsms.txt * The rest of your phone. $ gammu --backup backup.txt You may find an alternative format more human-readable for SMSes: $ gammu --geteachsms > eachsms.txt See the manual page for more tricks! 16.35. gcc 16.35.1. libstdc++ I have heard I should never remove the gcc package. Is this still true? "No, since we split the gcc package, the standard C++ library is a separate package, you have to install gcc only in case you really need a C compiler." 16.35.2. gcjwebplugin gcjwebplugin is a Firefox plugin for running Java applets. It is now included in the libgcj sub-package, though it is not enabled by default. GNU Classpath and libgcj's security implementation is under active development, but it is not ready to be declared secure. Specifically, it cannot run untrusted applets safely. When gcjwebplugin is enabled, it prompts you with a dialog before loading an applet. The dialog tells you that a certain URL would like to load an applet, and asks if you trust the applet. Be aware though that this dialog is mostly informative and doesn't provide much protection: * http and DNS can be spoofed meaning that the URL named in the warning dialog cannot be trusted * someone could create a browser denial-of-service attack by creating a page with hundreds of applet tags, causing gcjwebplugin to create warning dialog after warning dialog. The browser would have to be closed to eliminate the latest dialog * the whitelist is provided as a convenience, but it is unsafe because a domain may change hands from a trusted owner to an untrusted owner. If that domain is in the whitelist then the warning dialog will not appear when loading the new malicious applet. CURRENTLY GCJWEBPLUGIN RUNS WITH NO SECURITY MANAGER. THIS MEANS THAT APPLETS CAN DO ANYTHING A JAVA APPLICATION THAT YOU DOWNLOAD AND RUN COULD DO. BE VERY CAREFUL WHICH APPLETS YOU RUN. DO NOT USE GCJWEBPLUGIN ON YOUR SYSTEM IF YOUR SYSTEM STORES IMPORTANT DATA. THIS DATA CAN BE DESTROYED OR STOLEN. The same warning applies to gappletviewer, which also runs with no security manager (in fact, gcjwebplugin spawns gappletviewer to do the applet loading). When run on the command line, gappletviewer issues a warning on startup and asks you if you want to continue. Even considering the risks involved, you may still want to try gcjwebplugin. GNU Classpath's AWT and Swing implementations are now sufficiently mature that they're able to run many applets deployed on the web. If you're interested in trying gcjwebplugin, you can do so by creating a symbolic link in ~/.mozilla/plugins like so: ln -s /usr/lib/gcj-*/libgcjwebplugin.so ~/.mozilla/plugins/ Type about:plugins in Firefox's URL bar to confirm that the plugin has been loaded. To see gcjwebplugin debugging output, run: $ firefox -g then at the GDB prompt, type (gdb) run 16.36. git 16.36.1. gitweb If you want to set up a web interface for your git repositories, then: * install the gitweb package * edit /etc/gitweb.conf so that $projectroot will point to the repository directory * restart apache so that the gitweb configuration will be included. 16.37. gnome-applets The Mini Commander applet has been replaced by Deskbar. If you want the functionality provided by the Mini Commander applet, please install the deskbar-applet package with pacman 16.38. help2man The most common usage of this applications is something like this: $ help2man -n "" -S Frugalware -N ./ |sed 's/\\(co /(c)/' >.1 16.39. horde-webmail This app does not have any webserver, SQL server nor IMAP server in its depends, which is intentional. Anyway, if you plan to use it, you should set up a webserver and an IMAP server. The SQL server is optional, but it's the most easiest-to-use preferences container. Additionally this app is not configured in any way: there are far too many customizable settings, so the packager cannot know how to set them for your particular needs. Installation instructions can be found in the INSTALL file. 16.40. httpircproxy Don't forget to open port 8080 on your firewall! To do this, add the following two lines to your /etc/sysconfig/firewall: # httpircproxy -A INPUT -p tcp -m tcp --dport 8080 -j ACCEPT for example after MySQL's entry. 16.41. hylafax Welcome to the README! Thanks for taking the time to find it ;-) For an introduction to the wonderful world of HylaFAX(tm), please see http://www.hylafax.org/. Beginners should head directly to the docs: http://www.hylafax.org/content/Documentation If you have a question which you think relates only to the FPM version of HylaFAX post a bug to the Frugalware BTS: http://bugs.frugalware.org/ You should also be aware of the following system modification: FaxMaster is added to /etc/postfix/aliases after installation automatically. The default configuration files can be found under /var/spool/hylafax/config/defaults/. You can copy these files to the /var/spool/hylafax/etc/ directory and modify them there. Enjoy! 16.42. joomla After installing this package, please run /usr/bin/joomlasetup as root to setup Joomla 16.43. k3b If you want to rip a video DVD, install the transcode package as well. 16.44. kernel-xen0 Usage instructions: 1. Get the latest xen0 package. # pacman-g2 -Syu && pacman-g2 -S kernel-xen0 2. Add the following lines to the bottom of /boot/grub/menu.lst title Frugalware 0.5 (Siwenna) - With Xen kernel (hd0,13)/boot/xen.gz noreboot dom0_mem=768M module (hd0,13)/boot/vmlinuz-2.6.16-xen0-fw4 ro root=/dev/hda14 Replace (hd0,13) with the grub reference to your Frugalware partition, and /dev/hda14 with the Linux device name for your partition (use the existing grub entry for frugalware as a reference). 768M should also be replaced with a value of around half of your total system memory. 3. Reboot and select the Xen option at the Grub menu. 4. Hopefully nothing bad will happen. If it does, write down the last few messages that you see, and file a bugreport at http://bugs.frugalware.org 5. If all works OK after the first few boots, you should be able to safely remove noreboot from the kernel command line. If you run into problems later and Xen reboots before you can see the error, just readd noreboot Note on i686, you need the glibc-xen package instead of the normal glibc! 16.45. kexec-tools Warning kexec works just like reboot, so please save your data before using it! Loading the new kernel: # kexec -l /boot/vmlinuz-2.6.18-fw1 --append="ro root=/dev/hda3 quiet vga=792 re sume=/dev/hda2" Booting it: # kexec -e 16.46. kiax If you want to test this package, you can use for example the server of [23]VoipBuster. First, you should register a user name and password with their native (Windows-only) client. After that give iax.voipbuster.com as the server and your just registered username and password. Now you should able to dial (currently the first minute is for free). 16.47. knb To use knb, you need a config file like this: nick idlenick realname Knb nicks keepnick server irc.server.com 6667 channel #channel where idlenick is used till keepnick is used by someone else. The bot will join to #channel on irc.server.com. You need to register that you're the owner for the first time. To do this, join #channel and !new nick!ident@host to give access someone to the bot. Once keepnick is no longer used and knb switched to that nick, you can use !n -yes idlenick to change knb's nick back to idlenick, so that you can change your nick back to keepnick. See the scripts directory on how to re-start your knb from cron automatically. 16.48. kqemu If you want to use kqemu, you need to mount tmpfs on /dev/shm. This is not problematic, qemu prints a usable error message if you miss that. The problem is that you have to do this again and again after each reboot. If you hate this, then just add the following line to your /etc/fstab: tmpfs /dev/shm tmpfs defaults,size=144m 0 0 16.49. kvpnc Howto setup KVpnc for use without root password - sudo 1. Install sudo 2. Edit /etc/sudoers: add an command alias # Cmnd alias specification Cmnd_Alias KVPNC = /usr/bin/kvpnc # User privilege specification ALL ALL=NOPASSWD:KVPNC Warning Do it gently! (As always, when you edit /etc/sudoers.) 16.50. lastfmsubmitd 16.50.1. Configuring Lastfmsubmitd Change your LastFM username and password in /etc/lastfmsubmitd.conf and the MPD server settings in /etc/lastmp.conf before starting the LastFM submit daemon. 16.50.2. Starting the daemon(s) After configuring lastfmsubmitd, you should run the following commands to start the daemons: # service lastfmsubmitd start # service lastmp start 16.51. lineakd After installing lineakd, make sure you create a configuration file before starting it. An example configuration file is located in /etc/lineak/lineakd.conf.example for the user's reference. You can then start the lineak daemon by running the following command: $ lineakd 16.52. lirc After installing lirc you need to take the following steps: 1. Find a lird.conf for your remote control on [24]remotes You can also take a look on /usr/share/remotes directory if you do not have an internet connection. If you do not find your remote controller, try irrecord myremote command. 2. Copy your lircd.conf to /etc/ directory as root. 3. Add evdev to /etc/sysconfig/modules. 4. Load the module with modprobe evdev. 5. Edit /etc/sysconfig/lirc if necessary. $ cat /proc/bus/input/devices | grep -e N -e H will show you the event# you should use. (Default is 2.) 6. Start lircd and lircmd with sudo service lirc start. 16.53. lmsensors Lmsensors is a hardware monitoring tool which is able to read thermal and voltage values and fan speeds from the sensor chips of your motherboard. Before running sensors you have to run sensors-detect as root to initialize them. It will autodetect your hardware and define which kernel modules you need to get it working properly, and tell you how to autoload them during boot. So if you want to use lmsensors try to run # sensors-detect and say YES at end of sensors-detect to write /etc/sysconfig/lm_sensors. Then issue: # service lmsensors start 16.54. lvm2 16.54.1. Creating Here is a mini-HOWTO, a longer one is available [25]here. First if you are on a setup cd, you need to modprobe dm-mod and vgchange -a y The first loads the device-mapper support for the kernel, the later enables the existing volume groups. This is automatically done for you on an installed Frugalware system. You need to decide what physical partitions to use for LVM. In this mini-HOWTO / is /dev/hda1 and we create a big /home partition using /dev/hda2 and /dev/hdc1. Let's initialize them for use by LVM: pvcreate /dev/hda2 /dev/hdc1 Create a volume group titled vg: vgcreate vg /dev/hda2 Extend it with /dev/hdb1: vgextend vg /dev/hdc1 Then we can create a logical volume with a size of 400G titled home: lvcreate -L400G -nhome vg Create a filesystem on it as usual, ie. for ext3: mke2fs -j /dev/vg/home And now the only task is to mount it as usual, ie: mount /dev/vg/home /mnt/target/home 16.54.2. Extending You already saw how to extend a volume group. Extending a logical volume is a bit more complex, but still easy. If you use ext3: umount /mnt/target/home lvextend -L+900M /dev/vg/home resize2fs /dev/vg/home mount /dev/vg/home /mnt/target/home Note According to the manpage of resize2fs, it would have support resizing without umounting, but this does not seem to work. If you use reiserfs: lvextend -L+900M /dev/vg/home resize_reiserfs /dev/vg/home 16.54.3. Removing To remove a logical volume: lvremove /dev/vg/home To remove a physical volume from a volume group: vgreduce vg /dev/hdc1 To remove a volume group: vgremove vg That's it. 16.55. mailman There is no any kind of http server in mailman's depends. It's because they are not needed to get a working mailman. Of cource if you want to provide archives and so don't forget to install a http server. 16.56. man If you like coloured man-pages then you can enable that feature by issuing # chmod +x /etc/profile.d/man-colors.sh It is handled as a configuration file, so feel free to edit the colors in that file if you want. 16.57. mantis You have to GRANT some privileges (at least for the operating user) to be able to use this package, as the installer does not GRANT them. The operating user requires ALTER, SELECT, INSERT, UPDATE and even DELETE privileges, regardless that the latter is not mentioned by upstream. For installation, INDEX, CREATE, DELETE, and DROP privileges are also required - this can be carried out if you provide the (MySQL) superuser's credentials to the installer. Do not forget to rm -rf /var/www/mantis/admin after a successful install to prevent hijacking your bugtracker, and chage the default administrator's password. 16.58. maxima Install the tk package if you intend to use the gui "xmaxima". Detailed documentation can be found using * info maxima * in the directory /usr/share/maxima/$package_version/doc/ 16.59. mb2md How to convert each users mbox from /var/mail to Maildir (under /home/$user)? If you are too lazy to read the complete documentation: cd /var/mail for i in * do echo $i su - $i -c "mb2md -m" rm -v $i done 16.60. mediawiki After installing this package, please run /usr/bin/mediawikisetup as root to setup MediaWiki 16.61. midpssh-http-server Do not forget to open port 8080 on your firewall! To do this, add the following two lines to your /etc/sysconfig/firewall: # midpssh-http-server -A INPUT -p tcp -m tcp --dport 8080 -j ACCEPT for example after MySQL's entry. 16.62. monit You can set up Monit by adding these lines to /etc/inittab: # Run monit in standard run-levels mo:2345:respawn:/usr/sbin/monit -Ic /etc/monit/monitrc 16.63. motion You should edit the settings: videodevice, input, norm, frequency, width, height and target_dir in /etc/motion.conf If the file already exists, it wont be overwritten by the package while upgrading. You can refer /etc/motion-dist.conf for configuring motion. 16.64. munin From munin-1.2.5-2 we no longer use a random uid/gid, but dedicated ones. Because of this munin service will not start if you have it installed before, so you have to correct this by issuing these commands: groupmod -g 47 munin usermod -u 47 -g 47 munin chown -R munin:munin /var/lib/munin chown -R munin:munin /var/www/html/munin chown -R munin:munin /var/log/munin chown -R munin:munin /var/run/munin You should chown any other munin-owned stuff you may have lying around, these are only the default ones. 16.65. mythtv You can configure MythTV this way: 1. Start mysql service and setup mysql database password with mysqladmin -u root password mysqlpassword. 2. Set up the initial database with mysql -u root -p < /usr/share/mythtv/mc.sql and enter mysqlpassword. 3. Run sudo mythtv-setup for tune your tvcard. 4. Start mythtv backend with sudo service mythtv start. 5. Use mythfilldatabase to fill in your database. 6. Finally run mythfrontend and have fun! For more information see MythTV [26]documentation. 16.66. nautilus-share Here is an excerpt from the Samba documentation explaining what you need to do to make this work: USERSHARE Starting with version 3.0.23, a Samba server now supports the ability for non-root users to add user define shares to be exported using the "net usershare" commands. To set this up, first set up your smb.conf by adding to the [global] section: usershare path = /home/usershares Next create the directory /home/usershares, change the owner to root and set the group owner to the UNIX group who should have the ability to create usershares, for example the "users" group. Set the permissions on /home/usershares to 01770. (Owner and group all access, no access for others, plus the sticky bit, which means that a file in that directory can be renamed or deleted only by the owner of the file). Finally, tell smbd how many usershares you will allow by adding to the [global] section of smb.conf a line such as: usershare max shares = 100 to allow 100 usershare definitions. Now, members of the UNIX group "users" can create user defined shares on demand using the commands below. The usershare commands are: net usershare add sharename path [comment] [acl] [guest_ok=[y|n]] - to add or ch ange a user defined share. net usershare delete sharename - to delete a user defined share. net usershare info [-l|--long] [wildcard sharename] - to print info about a user defined share. net usershare list [-l|--long] [wildcard sharename] - to list user defined share s. 16.67. ndiswrapper Ndiswrapper requires .inf and .sys files from a Windows(tm) driver to function. Download these to /root for example, then run: # ndiswrapper -i /root/foo.inf After that you can delete them. They will be copied to the proper location. Once done, please run: # depmod -a Check this [27]list of drivers. You can get your possible hadware with: # lspci -n | egrep 'Class (0280|0200):' | cut -d' ' -f4 Look for that on the above page for your driver. Please have a look at the [28]wiki for the FAQ, HowTos, Tips, Configuration, and installation information. 16.68. nessus-core Install script has generated default certificate use: # /usr/sbin/nessus-mkcert if you need to make a personalized one. Remember to create a user with: # /usr/sbin/nessus-adduser Use /etc/sysconfig/nessusd for setting nessusd starting arguments. 16.69. nessus-core-gtk Install script has generated default certificate use: # /usr/sbin/nessus-mkcert if you need to make a personalized one. Remember to create a user with: # /usr/sbin/nessus-adduser Use /etc/sysconfig/nessusd for setting nessusd starting arguments. 16.70. nss-mdns To enable IPv4 multicast DNS lookups, append mdns4 to the hosts line in /etc/nsswitch.conf. Use mdns6 for IPv6 or mdns for both. 16.71. ooextras Please add /usr/lib/ooextras to your template paths in OpenOffice.org! You can do this under: Tools > Options > OpenOffice.org > Paths Enjoy the new templates! :-) 16.72. openssh 16.72.1. Forwarding ports # ssh -L 8000:localhost:80 server.com After this you can access server.com:80 at localhost:8000 even if server.com:80 is not accessible from your machine. 16.72.2. Socks proxy Many mobile users have the following problem: they have to use an unencrypted wireless lan and they want to access webservers which does not support https. There is an easy solution for this: you transfer data to a server in an ssh tunnel then the data can be transferred to the server unencrypted in a wired network. This is much more secure. Set up the socks proxy on localhost:8080: $ ssh -D 8080 server.com Then configure your webbrowser to use the proxy, for example in firefox, select Manual proxy configuration and then set SOCKS Host to localhost, Port to 8080. Note Don't forget to clear other proxy fields! (HTTP, SSL, FTP, etc.) 16.73. openttd You require the data files of the original Transport Tycoon Deluxe for Windows to play the game. You have to manually copy the following files to /usr/share/openttd/data/ sample.cat trg1r.grf trgcr.grf trghr.grf trgir.grf trgtr.grf If you want to use the included scenarios, run openttd with the following command: $ openttd -g openttd -g /usr/share/openttd/scenario/ 16.74. php You should set cgi.fix_pathinfo=1 in /etc/php.ini in order to use php-cgi. 16.75. phpbb After installing this package, please run /usr/bin/phpbbsetup as root to setup phpBB After upgrading, make sure to run the database update script 16.76. pootle 16.76.1. Introduction Pootle provides a separate webserver, after starting it, you should be able to reach it at http://127.0.0.1:8080/. 16.76.2. Creating users No-one can login to Pootle by default. You should edit /usr/lib/python2.5/site-packages/Pootle/users.prefs, and uncomment the startup user. Its password is startup, too. Then login via the web interface, create an account. Finally edit the configuration file again and comment the startup user. 16.77. postfix 16.77.1. Using a relay host These are the basic steps to set up Postfix to use SMTP Authentication to send mail through a relay host. Set up a password maps file (/etc/postfix/sasl_passwd) as follows: mail.ispserver.com username:password # chown root:root /etc/postfix/sasl_passwd # chmod 600 /etc/postfix/sasl_passwd # postmap /etc/postfix/sasl_passwd Append the following lines to /etc/postfix/main.cf: relayhost = mail.ispserver.com smtp_sasl_auth_enable = yes smtp_sasl_password_maps = hash:/etc/postfix/sasl_passwd smtp_sasl_security_options = Finally reload postfix: # postfix reload That should do it! 16.78. postfixadmin This package relies on correct install of postfix's virtual tables and it needs to be configured by hand before usage. Copy /var/www/postfixadmin/config.inc.php.sample to /var/www/postfixadmin/config.inc.php and edit it to your taste. Below is an excerpt from upstream's INSTALL.TXT. 1. Create the Database Tables In DATABASE_MYSQL.TXT you can find the table structure for MySQL that you need in order to configure Postfix Admin and Postfix in general to work with Virtual Domains and Users. In DATABASE_PGSQL.TXT you can find the table structure for PostgreSQL. 2. Configure Check the config.inc.php file. There you can specify settings that are relevant to your setup. Postfix Admin contains 3 views of administration. There is the Site Admin view, located at http://domain.tld/postfixadmin/admin/. There is the Domain Admin view, located at http://domain.tld/postfixadmin/. And there is the User Admin View, located at http://domain.tld/postfixadmin/users/. In order to do the initial configuration you have to go to the Site Admin view. The default password for the Site Admin view of Postfix Admin is admin/admin. This is specified in the .htpasswd file in the /admin directory. Make sure that the location of the .htpasswd file matches your path. When you have finished, you have to restart apache (service httpd restart) in order to be able to authenticate the master admin. If you fail to do this, anybody will be able to do anything as postfixadmin's administrator without any authentication. 16.79. postgrey To use postgrey, put something along the lines of smtpd_recipient_restrictions = ... reject_unauth_destination check_policy_service inet:127.0.0.1:60000 in your /etc/postfix/main.cf (postfix 2.1 or newer is required.) 16.80. pptpd 1. Preface I was asked to set up VPN using PPTP. A much secure way to setup it up is using IPSec, more details [29]here. Also you could use ssh+pppd, but that's rather problematic on platforms other than Unix. 2. Setting up the server The big problem here is that most outdated HOWTO starts with patching your kernel and ppp. This is no longer needed! Requiements: You need kernel>=2.6.15 or newer (Frugalware 0.4 or higher is OK). Also you need ppp>=2.4.2. Also probably these are already installed on your system, let's see the new package: pptpd. Install it with the usual # pacman-g2 -S pptpd Probably this is done if you're reading this HOWTO :-) Here comes my /etc/pptp.conf: $ grep -v '^\(#\|$\)' /etc/pptpd.conf option /etc/ppp/options.pptpd logwtmp localip 10.0.0.88 remoteip 10.0.0.89-127 10.0.0.88 is the internal address of the server, 10.0.0.89-127 is the range that can be used by the pptp clients. Then let's see that referred /etc/ppp/options.pptpd: $ grep -v '^\(#\|$\)' /etc/ppp/options.pptpd name pptpd refuse-pap refuse-chap refuse-mschap require-mschap-v2 require-mppe-128 proxyarp debug lock nobsdcomp novj novjccomp nologfd After everything works fine, you can remove the "debug" line from the config. Then add at least one user: # cat /etc/ppp/chap-secrets ## client server secret IP addresses mylogin * stupidpassword * The rest is about to allow pptp on the firewall (I'm assuming that you use the default Frugalware configuration: INPUT is on DROP by default, but FORWARD is allowed, OUTPUT too.) Add the following 2 lines to the filter section of /etc/sysconfig/firewall: -A INPUT -p gre -j ACCEPT -A INPUT -p tcp -m tcp --dport 1723 -j ACCEPT If you want to allow a client to access Internet via this pptp server, add the following line to the nat section of the same file (change ethX to the correct network interface): -A POSTROUTING -o ethX -j MASQUERADE Then check if you have PPP support in the kernel enabled: # lsmod | grep ppp_generic If there is no output, enable it: # modprobe ppp_generic # echo "ppp_generic" >> /etc/sysconfig/modules Now we're ready to start: # pptpd -f -o /etc/ppp/options.pptpd If no error messages are reported, omit the -f option so it will go background. Later you can put this to your /etc/rc.d/rc.local. Debug messages will appear in /var/log/messages if you're interested in them. 3. Client side Install the necessary "pptp" package: # pacman-g2 -S pptp Most howto suggets the pptpconfig (http://pptpclient.sourceforge.net/) tool, it's written in PHP and uses GTK+2. You don't want to use graphical tools locally (and install XOrg) for administrating your machine, do you? We can do it by hand, not too complicated. You can name every tunnel you create, I'll use here the "mytunnel" name. Fire up your favorite editor and create the /etc/ppp/peers/mytunnel file with the following contents: $ grep -v '^\(#\|$\)' /etc/ppp/peers/mytunnel name mylogin remotename PPTP file /etc/ppp/options.pptp pty "pptp IP_OF_THE_SERVER --nolaunchpppd " require-mppe Your /etc/ppp/chap-secrets should contain the following line: mylogin PPTP secret * We're ready to start the client: # pppd pty 'pptp server --nolaunchpppd' call mytunnel debug dump logfd 2 nodetac h A lot of debug messages will be printed, check on an other console if you got a new pppx interface or not: # ifconfig ppp0 ppp0 Link encap:Point-to-Point Protocol inet addr:10.0.0.89 P-t-P:10.0.0.88 Mask:255.255.255.255 UP POINTOPOINT RUNNING NOARP MULTICAST MTU:996 Metric:1 RX packets:7 errors:0 dropped:0 overruns:0 frame:0 TX packets:7 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:3 RX bytes:70 (70.0 b) TX bytes:76 (76.0 b) If it seems to be ok, you no longer need the debug messages and pppd can go backround: # pppd pty 'pptp server --nolaunchpppd' call mytunnel That was all. Not so simple but anyone can do it :-) 4. Resources + http://czeh.hu/linuxdoc/vpn-pptp.html - VPN connection using PPTP and Linux by Istvan Czeh (Hungarian) + http://webb.gotdns.com:2080/kernel-mppe/pptp-command.html - pptp-command HOWTO 16.81. pyro You'll find pyro's scripts in /usr/lib/python2.5/site-packages/Pyro/bin 16.82. qemu 16.82.1. QuickStart If you are completely new to qemu, you may find the big list of switches a bit confusing. Most users want to install an operating system from a cdrom image to a virtual harddisk. Here is what you need: $ qemu-img create foo.img 8G $ wget http://server.com/bar.iso $ qemu -hda foo.img -cdrom bar.iso 16.82.2. Tricks It worth to read the full documentation at /usr/share/doc/qemu-*/qemu-doc.html, it really worth to do so. To demonstrate how powerfull qemu is, here are a few cheap tricks: If you want to be able to ssh to the machine, you can use port derirection. For example using the -redir tcp:1022::22 option, qemu:22 will be available at localhost:1022. Note This requires root privileges. You can create a unix socket to control your virtual machine. For example if you are not able to ssh to the machine, you can still properly shut it down: Use the -monitor unix:/tmp/qemu,server,nowait option, then send the sendkey ctrl-alt-delete string to the socket, for example using python: python -c "import socket; sock = socket.socket(socket.AF_UNIX, socket.SOCK_STREA M); \ sock.connect('/tmp/qemu'); \ sock.send('sendkey ctrl-alt-delete\n')" Finally a trick about vnc: using for example the -vnc 0 option, it's possible to reach qemu's display via vnc. This is quite handy if you run qemu on a server (for example in screen), then you can freely attach to and detach from it whenever you want to do so. Really, read the full documentation! :) 16.83. quagga The config files have to be in the /etc/quagga dir and have to be writeable by the quagga user (to be able to save config from the daemon's shell). Neither of the daemons will start till you edit the config files and rename/move them from .conf.sample to .conf (be careful to the uid/gid). You have to enable explicitly the routing daemons to get started from the init script. The config file is /etc/sysconfig/quagga If you have a working firewall, the OSPF daemon will not get working, you have to enable it in the firewall with this rule (maybe other routing daemons also have to be enabled, if you use it, but i could not find anything about that): iptables -A INPUT -p 89 -m state --state NEW -j ACCEPT 16.84. quota-tools To really activate quotas, you'll need to add usrquota to the appropriate partitions as listed in /etc/fstab. Here's an example: /dev/hda2 /home ext2 defaults,usrquota 1 1 When you want quota support for a given partition, some special files have to be created boot-time. This is not done by default. To do so, you need to # touch /var/lib/quota/new then, reboot to create those files. To edit user quotas, use edquota. See man edquota. 16.85. r8169 16.85.1. About the driver This is the r8169 driver from Realtek. This in not the same r8169 presented in Linux kernel. This driver supports: RTL8169S/8110S, RTL8169SB/8110SB, RTL8110SC 16.85.2. Using the driver To use this driver you have to remove the official r8169 if loaded. # rmmod r8169 You can load this module with # modprobe realtek-r8169 It might be a good idea to blacklist r8169 and add realtek-r8169 to /etc/sysconfig/modules, so you do not have to play this game after every reboot. 16.86. rss2email 16.86.1. Configure: Create a new feed database: $ r2e new you@yourdomain.com Subscribe to some feeds: $ r2e add http://www.aaronsw.com/2002/rss2email/updates.rss (That's the feed to be notified when there's a new version of rss2email.) Repeat this for each feed you want to subscribe to. When you run rss2email, it emails you about every story it hasn't seen before. But the first time you run it, that will be every story. To avoid this, you can ask rss2email not to send you any stories the first time you run it: $ r2e run --no-send Then later, you can ask it to email you new stories: $ r2e run You probably want to set this up as a cron job or something. 16.86.2. Customize: There are a few options, described at the top of rss2email.py. If you want to change something, add it to config.py. For example, to be notified every time a post changes, instead of just once per post: $ echo "TRUST_GUID = 0" >>~/.rss2email/config.py And you can ask rss2email to make the emails look as if they were sent when the item was actually posted: $ echo "DATE_HEADER = 1" >>~/.rss2email/config.py 16.87. scratchbox You need to complete the install, running: # /usr/lib/scratchbox/run_me_first.sh Do not forget to create a scratchbox user: # /usr/lib/scratchbox/sbin/sbox_adduser For further documentation about how to setup scratchbox for your development needs have a look at [30]scratchbox documentation. Also note that when you reboot and before trying to run scratchbox, you should run: # service scratchbox start You can also add it to the default runlevel: # service scratchbox add Then to start scratchbox, run: $ /usr/lib/scratchbox/login Note In order to run scratchbox, you have to be in the sbox group. 16.88. screen 16.88.1. Keeping your screen running across reboots You may want to restart your screen session automatically after a reboot. This is the case, for example, when we seed the Frugalware ISOs using a torrent client. Here is what you need: * Set up your ~/.screenrc so that it'll start your application when screen starts up: screen -t seed 0 /bin/sh -c 'cd $HOME/frugalware-torrents; rtorrent' * Run crontab -e and append the following line to your crontab: @reboot screen -d -m You're ready! 16.89. speedtouch Driver for the SpeedTouch USB and SpeedTouch 330 The binaries (modem_run and pppoax) have been installed in /usr/sbin. You will find the documentation and example script files in: /usr/share/doc/speedtouch-pkgver You can start configuring your modem by running /usr/bin/speedtouch-setup Note Read the documentation (/usr/share/doc/speedtouch-pkgver/howto) carefully to use this driver correctly! 16.90. squirrelmail Please start the configure script in the /var/www/squirrelmail directory! 16.91. squirrelmail-check_quota You have to install this plugin with squirrelmail's own ./configure tool. 16.92. squirrelmail-login_notes You have to install this plugin with squirrelmail's own ./configure tool. 16.93. stunnel You need some additional configuration before stunnel will be functional: Adjust the configuration file: # cp /etc/stunnel/stunnel.conf-sample /etc/stunnel/stunnel.conf # vi /etc/stunnel/stunnel.conf Note If something goes wrong, try setting sslVersion to all. Genrate your certificate: # openssl req -new -x509 -days 365 -nodes -config /etc/stunnel/stunnel.cnf -out \ /etc/stunnel/mail.pem -keyout /etc/stunnel/mail.pem Hide the certificate from users: # chmod 600 /etc/stunnel/mail.pem Now you can enable and start the service: # service stunnel add # service stunnel start 16.94. sugarcrm In order to use the sugarcrm, you have to symlink it to somewhere. For example, if you want to use it under http://localhost/sugarcrm, then use: # ln -s /var/www/SugarSuite /var/www/html/sugarcrm After installing this package, please run in a browser http://localhost/sugarcrm/install.php to setup SugarSuite (sugarcrm). 16.95. synaptics First of all, read the INSTALL file for the instructions. For lazy users, here's what i had to do to get the driver working: 1. Make sure you have the evdev kernel module loaded before the x server started (or restart it after you loaded the module). 2. Now open your /etc/X11/xorg.conf and add the following: Load "synaptics" to the "Module" section. 3. Replace the content of your "InputDevice" section to the followings: Section "InputDevice" Identifier "Synaptics Mouse" Driver "synaptics" Option "Device" "/dev/psaux" Option "Protocol" "auto-dev" Option "LeftEdge" "1700" Option "RightEdge" "5300" Option "TopEdge" "1700" Option "BottomEdge" "4200" Option "FingerLow" "25" Option "FingerHigh" "30" Option "MaxTapTime" "180" Option "MaxTapMove" "220" Option "VertScrollDelta" "100" Option "MinSpeed" "0.09" Option "MaxSpeed" "0.18" Option "AccelFactor" "0.0015" Option "SHMConfig" "on" # Option "Repeater" "/dev/ps2mouse" EndSection 4. Add this line to the "ServerLayout" Section: InputDevice "Synaptics Mouse" "CorePointer" 16.96. trac After installing trac you need a few steps to set it up. First of all do not forget to install postgresql/mysql/sqlite according to which database backend you want to use. To create a new trac project, just use the command: $ trac-admin /path/to/myproject initenv You can check the result with: tracd --port 8000 /path/to/myproject Then, fire up a browser and visit http://localhost:8000 For further documentation on trac, how to set up with different HTTP daemons see [31]TracGuide 16.97. udev /lib/udev/devices is the directory where packages or you can place real device nodes, which get copied over to /dev at every boot. This replaces our previous /etc/sysconfig/udev solution. (The change was recommended by upstream.) 16.98. unp unp is a small perl script which makes extraction of any archive files easier. It support several compressors and archiver programs, chooses the right one(s) automatically and extracts one or more files in one go. It detects the right type from the file extensions or the "file" output. unp was written by Andre Karwath (http://www.aka-online.de) in 1997, I (Eduard Bloch ) found it later and improved it to fill my needs. Have fun! 16.99. user-mode-linux 16.99.1. Creating a root image Create a big empty file: # dd if=/dev/zero of=root_fs bs=1M count=1000 Format it: # mke2fs -F -j root_fs Mount it: # mkdir uml # mount root_fs -o loop uml # cd uml Install base and openssh: # mkdir -p var/log tmp # pacman-g2 -Sy base openssh -r ./ Create etc/fstab with the following contents: none /proc proc defaults 0 0 none /sys sysfs defaults 0 0 devpts /dev/pts devpts gid=5,mode=620 0 0 /dev/ubda / ext3 defaults 1 1 Create etc/sysconfig/keymap with the following contents: keymap=us Create etc/profile.d/lang.sh with the following contents: export LANG=en_US export LC_ALL=$LANG We want networking, put the followings to etc/sysconfig/network/default: [eth0] options = 192.168.0.1 gateway = default gw 192.168.0.254 If you want to use multiple virtual machines, use 192.168.0.2, 192.168.0.3 and so on instead. Let's copy in the terminal device and change our root: # cp -a /dev/tty dev/ # chroot ./ Create a regular user: # adduser Remove unnecessary services and enable ssh: # service keymap del # service time del # rm /etc/rc.d/rcS.d/S18rc.time # service sshd add Remove unnecessary packages: # pacman-g2 -R gpm kernel Change /etc/inittab so that ctrl-alt-del will halt (and not reboot the system). Change the line ca::ctrlaltdel:/sbin/shutdown -t5 -r now to ca::ctrlaltdel:/sbin/shutdown -t5 -h now Exit from the chroot and umount: # exit # cd .. # umount uml You're ready, let's register it! 16.99.2. Configuration file You should edit /etc/sysconfig/uml. Each item in the machines array defines a virtual machine. Here is an example: machines=('ubd0=/home/uml/root_fs_0 eth0=tuntap,,,192.168.0.254 mem=128MB con0=n ull,fd:1 con=null') This does the following: * root fs will be /home/uml/root_fs_0 * the IP of the host will be 192.168.0.254 * allocate 128MB of memory * disable console input, console output will be stdout (that'll be logged to /var/log) * disable other consoles (we don't need them, we can use ssh) 16.99.3. Configuring the host network First you need the tun kernel module: # modprobe tun # echo tun >> /etc/sysconfig/module Second, you need NAT. Let's assume you access the external network via the eth0 interface, then edit /etc/sysconfig/network/default and search the end of the [eth0] section. Just append post_up = iptables -t nat -A POSTROUTING -j MASQUERADE to the section. After a # netconfig restart NAT will be enabled. Now you can easily start/stop your machines using the usual service uml start/stop command. 16.100. util-linux-ng 16.100.1. Using tmpfs for /tmp Frugalware does not use tmpfs for /tmp by default. However on servers this can cause problems: if you do not reboot for months, then cleaning /tmp can take some time. Using tmpfs can solve your problem: it's a ramdisk so its content not preserved during a reboot. All you need is to add the following line to your /etc/fstab: tmpfs /tmp tmpfs defaults 0 0 Note You need util-linux >= 2.12-31 for this, otherwise X may not start. 16.101. vim If you want to enable spell check support, you need to: * install the spell files for your language: # pacman-g2 -S vim-spell-xx where xx is code of the requested language. * enable the spell check support for your language (type in vim): :setlocal spell spelllang=xx_yy Some languages need correctly set encoding. If you get a message like: Warning: Cannot find word list "hu.latin1.spl" or "hu.ascii.spl" then you need to set your encoding as well: :set encoding=latin2 The incorrect words are coloured red by default. You can reach a list of suggested words by pressing z= when the cursor is at the given word. If you want to disable the spell check support, type: :setlocal nospell It may be handy to have map function keys in ~/.vimrc to enable / disable the spell check support: set encoding=latin2 map :setlocal spell spelllang=en_gb map :setlocal spell spelllang=hu map :setlocal nospell Note The language code is sometimes in an xx and sometimes is in an xx_yy form. This is something you need to figure out for your language. See the upstream documentation for more info about spell check support: :help spell 16.102. virtualbox If you want to be able to use the VirtualBox guest additions, run this command as root to get the Additions ISO (requires an active Internet connection): # /usr/bin/get-vbox-additions 16.103. wifi-radar Don't forget to change the wifi interface name in /etc/wifi-radar.conf! 16.104. x11vnc Running x11vnc without a password is not recommended. To create one, type: vncpasswd ~/.vnc/passwd Then you can start the VNC server using x11vnc -display :0 -rfbauth ~/.vnc/passwd -forever if are logged in on :0. 16.105. xcache 16.105.1. Installing As PHP Extension? 1. Check /etc/php.ini # cat /usr/share/doc/xcache-$pkgver/xcache.ini >> /etc/php.ini 2. Modify php.ini for your needs: # $EDITOR /etc/php.ini 3. Restart php Warning Use >> with cat, not simply > Please take a look on [32]xcache wiki. 16.106. xchat How should I remote control xmms from xchat? First make sure you really need it - some people think it's a security hole. You need [33]XChat-XMMS plugin from [34]XChat's Scripts + Plugins section, and some other packages (mostly Perl modules) which I'm too lazy to search for, but are available in fpm. Unpack the tarball, copy the .pl script to your XChat dir, and (try to) load it. If it complains about missing Perl modules, install them and try again. (This script has some minor bugs, but was found to be the most useful one amongst the kind. The documentation is a German PDF, which is to be translated to English/Hungarian.) 16.107. xen Warning Xen is unstable software, meaning that it should not be used on your main PC, it may destroy your data. As an example, I destroyed my file system during testing. 16.108. xgl 1. Xgl server Xgl is a hardware accelerated X server. It renders everything that gets drawn to the screen with OpenGL to allow for fancy effects like wobbly windows, translucency, etc. The disadvantage to Xgl is that programs that already use OpenGL will not work on it. Xgl is only recommended if you do not have a graphics card that supports GLX_EXT_texture_from_pixmap. 2. AIGLX Accelerated Indirect GLX ("AIGLX") is an open source project founded by Red Hat and the Fedora Linux community to allow accelerated indirect GLX rendering capabilities to X.Org and DRI drivers. This allows remote X clients to get fully hardware accelerated rendering over the GLX protocol; coincident