LQWiki - User contributions [en]

File system

2005-12-08T14:12:21Z

Jor: little start of /* Choosing your filesystem */

A '''file system''' is one particular way an [[operating system]] defines a user's files on a physical disk. What is represented on the disk is not a linear collection of files, placed one after the other, but extensive other data indicating structure and error correcting and other metadata stored alongside the files also - the way this metadata is stored comprises the file system.

==About filesystems==
When installing your Linux [[distribution]] there are a few options as to what filesystem to use. To many a newbie this option is of no consequence as they may not fully understand the relative pros and cons of each system. Hence, it is usually best to accept the distribution default, however this may not best suit the demands placed on your system or user needs.

The file hierarchy on a [[UNIX]] computer is usually a composite of several filesystems, held together by the 'master' filesystem [[VFS]]. This allows many different filesystems to meld together into one [[tree]]. For example, the CD filesystem [[ISO 9660]] may be mounted in any [[directory]] on an [[ext3]] [[partition]] and programs using the files will not know where one filesystem ends and the next begins. Similarly, files on remote computers may be mounted in directories on the local system, and from then on can be treated as if they were local.

===List of filesystems===
This list only gives a basic overview of which ones are available, detailed information should be found on the pages of the filesystem types itself.

====Common====
*[[Ext2]] - This has long been the ''default'' filesystem for most [[Linux]] [[distributions]], though many are now moving to a Journaling Filesystem like [[ReiserFS]] or [[ext3]].

====Journaling====
Definiton: See [[Journaling]]

For users this mainly means no more long [[fsck]] checks at startup after a crash, power failure or some other major problem.

On the more technical side this means that the filesystem keeps track of data transactions by writing to a journal. It can either write the whole data or only the [[metadata]]. This means it can recover after a failure by reading the journal, which is a lot faster then scanning the whole disk like [[fsck]] does.
Note however that there still can be some dataloss.

*[[Ext3]] - Adds journaling to [[ext2]] (You can switch between ext2 and ext3 by only remounting)
*[[ReiserFS]] - Uses a [[balanced tree]] to store filenames and files for very fast retrieval.
*[[Reiser4]] - The next generation of reiserFS, written from scratch
*[[XFS]] - [[SGI]]'s filesystem, originally from their [[IRIX]] [[OS]]
* [[JFS]] - [[IBM]]'s Journaled Filesystem.

====Other OS====
* [[FAT]]-based
**[[UMSDOS]] - A [[Unix-like]] filesystem on top of a FAT filesystem, basically allows you to install Linux on a FAT filesystem
** [[FAT12]]/[[FAT16|16]]/[[FAT32|32]] - Used in [[MSDOS]] versions
** [[Windows]] [[FAT]] ([[vfat]]) - Similar to [[MSDOS]] [[FAT]] but adds support of long filenames
*[[NTFS]] - New filesystem used in Windows NT,2000,XP
*[[MINIX]] - Used by the [[MINIX]] OS and the first versions of [[Linux]]
*[[UFS]] - Used by most [[BSD]] derivatives
*[[HFS]] - Used by [[MacOS]] versions
*[[NSS]] - Used by [[Novell]] [[NetWare]] for the on-disk storage (don't confuse it with the networked [[NCP]])
*[[befs]]/bfs - Used by [[BeOS]]

====CD/DVD====
*[[ISO 9660]] - Standard filesystem used on [[CD-ROM]]s
*[[UDF]] - Filesystem for [[CD-RW]]s and [[DVD]]s that allows you to use it like a Harddisk, also knows as ''Packet writing'' or ''DirectCD''.

====Networked====
*[[NFS]] - Network Files Sharing, ''export'' directories and files to another system on a network.
*[[SMB]]/[[CIFS]] - Standard network file sharing method used in MS [[Windows]]
*[[NCP]] - Network filesystem protocol used by [[Novell]] [[NetWare]] clients and severs (uses the [[IPX]] network protocol).
*[[Coda]] - Advanced network filesystem, supports disconnected operation, replication and more.
*[[Intermezzo]] - Network filesystem which supports disconnected operation.
*[[AFS]] - Andrew File System.

====In memory or virtual====
*[[tmpfs]] - Filesystem that stores all files in virtual memory (mem+swap), dynamically allocates the used memory and every thing is lost on a umount (mainly used for [[POSIX]] shared memory on /dev/shm)
*[[procfs]] - Provides information on the running system.
*[[devfs]] - Filesystem that dynamically creates the special character and block device files (been depreciated in 2.6 [[kernels]] in favour of [[udev]]
*[[devpts]] - virtual filesystem to support Unix98 PTY
*[[hugetlbfs]] - Filesystem to help use ''huge'' memory pages supported on some CPU architectures ([[mmap]]-ing files on this FS gives you memory backed by huge pages).
*[[sysfs]] - Provides information about the running system, introduced in [[kernel]] version 2.6 (Supposed to eventually replace [[procfs]]).

====Misc.====
*[[romfs]] - Simple read-only filesystem, mainly used for [[initrd]]
*[[cramfs]] - Similar to romfs, but uses zlib to compress data
*[[lufs]] - Linux Userland Filesystem, allows the filesystem to be implemented in [[userspace]]
**[[CARDFS]]
**[[CEFS]] for mounting Windows CE devices
**[[GNUTELLAFS]] Filesystem to use files from the [[Gnutella]] file sharing network directly.
**[[ftpfs]] Allows mounting of remote [[ftp]] sites in local filesystem, so files on the ftp server can be used directly.
**[[sshfs]] Allows mounting of remote [[sftp]] sites in local filesystem, so files on the sftp server can be used directly.
*[[JFFS]] - A journaling file system specially designed for flashcards, does ''wear leveling'' to increase lifespan of flashmemory.

===Choosing your filesystem===
Stick with the default your [[distribution]] selects, which almost always is [[ext2]]/[[ext3]], unless you really know what you are doing and have very specific requirements (like real-time multimedia editing, high volume web server or database, etc.).

===See also===
* [[Partition_a_hard_drive]]
* [[Create a file system]]
* [[Mount]]
* [[fstab|/etc/fstab]]
* [[Filesystem Hierarchy Standard]]
* [[Permissions|Basic Linux file system attributes]]
* [[Configure Storage Devices]]
* [[RAID]]
* [[LVM]]
* [[Quota]]
* [[Disk space]] ''freeing disk space mostly''

===External links===
* [http://www.tldp.org/HOWTO/Filesystems-HOWTO.html Filesystems HOWTO]

Fish shell

2005-12-08T13:54:50Z

Jor: init

'''fish''' ('''f'''riendly '''i'''nteractive '''sh'''ell) is a [[shell]] which tries to be more user friendly compared to the traditional shells (the [[sh]] or C families). It does this by implementing some helpful features which aren't usually built-in in the traditional shells (but are often available as external scripts and patches).
The main features it adds are:
*syntax highlighting
*tab completion
*easier history usage
*help system for the shell itself
It also has a different (easier?) syntax compared to the traditional shells.

==External links==
*[http://roo.no-ip.org/fish/ Project home page]

Shells

2005-12-08T13:34:16Z

Jor: add fish shell

A '''shell''', also known as a ''command interpreter'', is a specialized program for accepting typed user commands, translating those into programs to run, running those programs, and displaying (or doing something) with the results. A "common" example of a shell is the [[DOS]] shell, called COMMAND.COM, which was the complete user interface before [[Windows]].

Different shells exist, offering different feature sets. Two main families of shell exist: the Bourne shell and its variants ([[sh]], [[bash]], [[ksh]]) and the C shell and its variants ([[csh]], [[tcsh]]). Though many shells have features common to others, the way they make use of those features is unique, so that (for example) Bourne shell conventions don't usually apply to C shells.

*[[bash]] - The Bourne-Again SHell - feature-rich default Linux shell.
*[[csh]] - A shell with C-like syntax, with file name completion and command line editing - usually <tt>[[tcsh]]</tt> on modern Linux.
*[[fish shell|fish]] - 'friendly' interactive shell
*[[ksh]] - Korn SHell - part of the <tt>sh</tt> rather than <tt>csh</tt> family.
*[[sh]] - the original shell, often a symlink to <tt>bash</tt> on modern Linux and the most portable.
*[[tcsh]] - An extended version of [[csh]], with all its features and some additional ones.
*[[zsh]] - One of the newest and most feature-rich shells.

==See also==
*[[Commands]]
*[[Filename Matching]]
*[[Commands common to many shells]]
*[[environment variables]]
*[[shell programming]]
*[[The Shell]] A beginner's guide to the Linux [[CLI]]

User talk:Smiletolead

2005-12-02T15:47:17Z

Jor:

Can anyone explain me the use of symbolic links to libraries. Suppose there is a library libicuio.so.34.0. There will be two symbolic links to it: libicuio.so and libicuio.so.34. What is the use of these two symbolic links.
: Versioning, see [[Library-related_Commands_and_Files#Naming_conventions]].--[[User:Jor|Jor]] 10:47, Dec 2, 2005 (EST)

NVIDIA graphics drivers

2005-10-18T14:20:14Z

Jor: update Debian section

This is a guide on how to install [[NVIDIA]]'s drivers for X-windows, the graphical user interface system on Linux. These drivers are not required, but if you wish to run games such as Quake or Unreal at full speed you should install them. Before you start, take note that the drivers are distributed under a non-opensource license.

==Distribution specific==
=== Fedora Users ===
For [[Fedora]] users, the NVIDIA drivers are available repackaged in RPM, with pre-build kernel modules for easy install and update. Normally you can find the rpms in the [[livna]].org apt/yum respository, but the latest drivers (version 6106) are not yet avaliable as RPMs. See [http://fedorafaq.org/faq/ the Fedora FAQ] for more information on how to install on Fedora Core 2 anyway, or [http://rpm.livna.org/livna-switcher.html here] for more info on the livna nvidia packages.

Drivers before version 6106 will not work with prebuilt Fedora Core 2 kernels, because they do not support [[4stacks|4k Stacks]].

===Debian===
[[Debian]] has packages for these drivers. There are two packages that need to be installed.
One for the binary drivers and libraries and one for the [[kernel module]]. 
For Debian 3.1 (Sarge) and higher they are called <tt>nvidia-glx</tt> and <tt>nvidia-kernel-<version></tt> (both are in the non-free section). Where <version> is one matching your used Debian kernel or <tt>source</tt> to build the module for your specific kernel. See the README.Debian file that comes with them for full instructions. And use the package [[module-assistant]] to help compile the kernel module if you don't have a default Debian kernel. After installing these packages, the X server configuration file needs to be changed, seen the section about that below.

The other optional packages are <tt>nvidia-glx-dev</tt> for when you need to compile an openGL applicating against the nvidia openGL libs and <tt>nvidia-settings</tt> for a [[GUI]] application to configure the various driver options.

In Debian 3.0 (Woody) these packages are called <tt>nvidia-glx-src</tt> and <tt>nvidia-kernel-src</tt> (both are in the contrib section and use [[wget]] to download the actual drivers). 

===SuSE===
For [[SuSE]] users, the NVIDIA drivers are available from [http://www.nvidia.com/object/linux.html NVIDIA], the installation HOWTO is available from [ftp://ftp.suse.com/pub/suse/i386/supplementary/X/XFree86/nvidia-installer-HOWTO here].
Follow the steps described in the HOWTO. If Hardware 3D - acceleration fails but X11 in 2D mode is using the new nvidia driver do the following: install the driver again, the second time don't use the program <tt>sax2</tt> but run the command <tt>switch2nvidia</tt>

===Gentoo===
[[Gentoo]] has ebuilds available which can be installed with: 
 <tt> emerge nvidia-kernel nvidia-glx</tt> 
 
The first ebuild "nvidia-kernel" provides Nvidia's driver for their various cards and nvidia-glx provides 3D acceleration (this is optional).

You have to edit your X11 config file manually using the procedure described [[#Editing the X11 configuration|below]] (note: [[gentoo]] uses [[X.org]]).

===Mandrake/Mandriva===
[[Mandriva]] has Nvidia RPMs available for PowerPack users (ie. those who bought a boxed set or are club members). If you have a PowerPack the drivers will install automatically when you install the OS.

For everyone else the manual procedure is:

1) If you haven't already done so go to http://easyurpmi.zarb.org and setup access to the software repositories for your version.

2) In the Software section in the Mandrake Control Centre, or using the urpmi command as root, install the package named "kernel-source-2.6" (without quotes).

Now follow the rest of the instructions below under "Manual Installation".

==Manual installation==
=== Download ===
The first step is to download some needed files. First, grab the latest version of the drivers from the [http://nvidia.com/content/drivers/drivers.asp NVIDIA Website]. Once you have that downloaded, make sure you have the source for your kernel. You can check what kernel you have by running the command '[[uname]] -r', without the quotes.

=== Install ===
First of all, you need to shut down X. Yes, that's right, we're going to be working from a command line. To do this:

1) Press Ctrl+alt+F1 on the keyboard. Now you should be at a console login screen, login as root.

2) Type the command 'telinit 3' (no quotes) to put your machine into runlevel 3 (no graphical interface). It will tell you about some services its restarting, if it doesn't return you to the command prompt press enter. For more information on this, refer to [[run levels]].

Once you get X shut off, navigate (see [[Directory_Commands]]) to where ever you downloaded the driver file to, and run it. You can run it by typing ''sh filename''. It should check to make sure that it downloaded OK, then start up the installation program.

Go through the first screen, accept the license, and when it asks if you want to try and download a kernel interface from NVIDIA, say no. I've never seen anyone get the kernel interface form the site, so we'll save time and move onto the next screen, where it tells us it will compile the kernel interface on its own. This is where our kernel source comes into play. It should compile and install on its own, so go to the next screen. This should pretty much wrap up anything important in the installation.

=== Editing the X11 configuration ===
The next step is to edit your [[X11]] configuration file, so it loads the nvidia driver. For [[XFree86]] the file is usually found in /etc/X11/ and called [[XF86Config]], but you might also have XF86Config-4, in which case you should edit that instead. Many newer distributions use [[X.org]], whic uses the /etc/X11/Xorg.conf file.

Before we do that, we should back up the file, just in case something goes wrong.

# cp /etc/X11/XF86Config /etc/X11/XF86Config.backup

Then, open the X config file in vi (or whatever editor you prefer, be it [[pico]], [[nano]], [[emacs]] or [[joe]]) by running:

# vi /etc/X11/XF86Config

Look for the '''Module''' section, and make sure it includes the '''Load "glx"''' line. If it doesn't, add it in. Then remove the '''Load "dri"''' and '''Load "GLCore"''' lines, if you have them. If you might want to go back to the old drivers later you can comment out loading dri and GLCore instead of removing them completely by putting a hash at the start of the line. An example configuration:

<tt>
Section "Module"
Load "dbe" # Double buffer extension
Load "type1"
Load "speedo"
Load "freetype"
Load "glx"
<nowiki># Load "dri"</nowiki> 
<nowiki># Load "GLcore"</nowiki> 
EndSection
</tt>

Finally, find the Device setion, and change the line '''Driver "nv"''' or '''Driver"vesa"''' to '''Driver "nvidia"'''. An example device section:

<tt>
Section "Device"
Identifier "NVIDIA"
Driver "nvidia"
VideoRam 65536
# Insert Clocks lines here if appropriate
EndSection
</tt>

To disable the NVidia splash-screen logo when starting X, add the option highlighted below:

<tt>
Section "Device"
Identifier "NVIDIA"
Driver "nvidia"
'''Option "nologo" "1"'''
VideoRam 65536
# Insert Clocks lines here if appropriate
EndSection
</tt>

=== Start X ===
Now all that's left is to start X back up. To do this type the command 'telinit 5' at the console prompt and it should load X and take you to the normal graphical login screen. You should now have your brand-new NVIDIA drivers installed and working. Good job!

=== Uninstalling NVIDIA drivers ===
Did something in your installation go terribly, terrible wrong? No problem! Start your box up in console-only mode, navigate to where you have the nvidia driver installation file saved, and run it with the --uninstall option. ''sh filename.run --uninstall'', and it uninstalls everything it put in. Then it's a simple matter of restoring your XF86Config from the backup we made above.

== Issues ==
* One issue I've had is between the GeForce FX 5200 card and the Asus P4S8X-X motherboard. When you started the X server, it would display garbled text and then hang. This problem can be solved by going into the BIOS and setting your AGP speed down from 8x to 4x.

==See Also==
* [[NVIDIA]]

Openoffice.org

2005-08-19T16:55:44Z

Jor: redirect

#redirect [[OpenOffice]]

Beagle

2005-08-19T16:51:18Z

Jor: init

'''Beagle''' is a desktop search tool. It can index various sources to allow fast searching for information. It currently supports searching the [[filesystem]], email messages, instant messaging history, webbrowser history, [[RSS]] feeds and many other sources are being created and added. Since Beagle is aimed at the [[Gnome]] desktop, it can currently only read the email from [[Evolution]], the browser history from [[Epiphany]] and the IM messages from [[GAIM]], etc.

For searching the filesystem, beagle supports getting extra info from certain filetypes like [[openoffice.org]] or Microsoft Word documents, [[HTML]] files, [[JPG]] images, music files (like [[mp3]] and [[ogg]]) and many other types.

Beagle consists out of a [[daemon]] ''beagled'' which does the indexing of the files and other information and a GUI called [[best]] for searching. For storing information about when the a file was last indexed beagled uses [[extended attributes]] (and falls back to a slower [[sqlite]] database backend when that's not available. It can also use [[inotify]] when that's available on the system to receive notification of when to index changed or new files.

==External links==
*[http://beaglewiki.org Main Beagle website]

3D hardware acceleration

2005-08-19T16:05:49Z

Jor: redirect

#REDIRECT [[3D graphics acceleration]]

NVIDIA graphics drivers

2005-08-19T15:56:20Z

Jor: debian module-assistant

This is a guide on how to install [[NVIDIA]]'s drivers under [[XFree86]]. These drivers are not required, but if you wish to run games such as Quake or Unreal at full speed you should install them. Before you start, take note that the drivers are distributed under a non-opensource license.

==Distribution specific==
=== Fedora Users ===
For [[Fedora]] users, the NVIDIA drivers are available repackaged in RPM, with pre-build kernel modules for easy install and update. Normally you can find the rpms in the [[livna]].org apt/yum respository, but the latest drivers (version 6106) are not yet avaliable as RPMs. See [http://fedorafaq.org/faq/ the Fedora FAQ] for more information on how to install on Fedora Core 2 anyway, or [http://rpm.livna.org/livna-switcher.html here] for more info on the livna nvidia packages.

Drivers before version 6106 will not work with prebuilt Fedora Core 2 kernels, because they do not support [[4stacks|4k Stacks]].

===Debian===
[[Debian]] also got packages for these drivers. There are two needed packages.
One for the binary drivers and libraries and one for the [[kernel module]]. 
For Debian [[stable]] these packages are called <tt>nvidia-glx-src</tt> and <tt>nvidia-kernel-src</tt> (both are in the contrib section and use [[wget]] to download the actual drivers). 
For Debian [[testing]] and [[unstable]] they are called <tt>nvidia-glx</tt> and <tt>nvidia-kernel-<version></tt> (both are in the non-free section). Where <version> is one matching your used Debian kernel or <tt>source</tt> to build the module for your specific kernel. See the README.Debian file that comes with them for full instructions. And use the package and command [[module-assistant]] to help compile the kernel module if you don't have a default Debian kernel.

===SuSE===
For [[SuSE]] users, the NVIDIA drivers are available from [http://www.nvidia.com/object/linux.html NVIDIA], the installation HOWTO is available from [ftp://ftp.suse.com/pub/suse/i386/supplementary/X/XFree86/nvidia-installer-HOWTO here].
Follow the steps described in the HOWTO. If Hardware 3D - acceleration fails but X11 in 2D mode is using the new nvidia driver do the following: install the driver again, the second time don't use the program <tt>sax2</tt> but run the command <tt>switch2nvidia</tt>

===Gentoo===
[[Gentoo]] has ebuilds available which can be installed with: 
 <tt> emerge nvidia-kernel nvidia-glx</tt> 
 
The first ebuild "nvidia-kernel" provides Nvidia's driver for their various cards and nvidia-glx provides 3D acceleration (this is optional).

You have to edit your X11 config file manually using the procedure described [[#Editing the X11 configuration|below]] (note: [[gentoo]] uses [[X.org]]).

==Manual installation==
=== Download ===
The first step is to download some needed files. First, grab the latest version of the drivers from the [http://nvidia.com/content/drivers/drivers.asp NVIDIA Website]. Once you have that downloaded, make sure you have the source for your kernel. You can check what kernel you have by running the command '[[uname]] -r', without the quotes.

=== Install ===
First of all, you need to shut down X. Yes, that's right, we're going to be working from a command line. For more information on this, refer to [[run levels]]. Once you get X shut off, navigate to where ever you downloaded the driver file to, and run it. You can run it by typing ''sh filename''. It should check to make sure that it downloaded OK, then start up the installation program. Go through the first screen, accept the license, and when it asks if you want to try and download a kernel interface from NVIDIA, say no. I've never seen anyone get the kernel interface form the site, so we'll save time and move onto the next screen, where it tells us it will compile the kernel interface on its own. This is where oour kernel source comes into play. It should compile and install on its own, so go to the next screen. This should pretty much wrap up anything important in the installation.

=== Editing the X11 configuration ===
The next step is to edit your [[X11]] configuration file, so it loads the nvidia driver. For [[XFree86]] the file is usually found in /etc/X11/ and called [[XF86Config]], but you might also have XF86Config-4, in which case you should edit that instead. If you use [[X.org]], the file to edit is /etc/X11/Xorg.conf.

Before we do that, we should back up the file, just in case something goes wrong.

# cp /etc/X11/XF86Config /etc/X11/XF86Config.backup

Then, open the X config file in vi (or whatever editor you prefer, be it [[pico]], [[nano]], [[emacs]] or [[joe]]) by running vi /etc/X11/XF86Config. Look for the '''Module''' section, and make sure it includes the '''Load "glx"''' line. If it doesn't, add it in. Then remove the '''Load "dri"''' and '''Load "GLCore"''' lines, if you have them. If you might want to go back to the old drivers later you can comment out loading dri and GLCore instead of removing them completely by putting a hash at the start of the line. An example configuration:

<tt>
Section "Module"
Load "dbe" # Double buffer extension
Load "type1"
Load "speedo"
Load "freetype"
Load "glx"
<nowiki># Load "dri"</nowiki> 
<nowiki># Load "GLcore"</nowiki> 
EndSection
</tt>

Finally, find the Device setion, and change the line '''Driver "nv"''' or '''Driver"vesa"''' to '''Driver "nvidia"'''. An example device section:

<tt>
Section "Device"
Identifier "NVIDIA"
Driver "nvidia"
VideoRam 65536
# Insert Clocks lines here if appropriate
EndSection
</tt>

=== Start X ===
Now all that's left is to start X back up. You should now have your brand-new NVIDIA drivers installed and working. Good job! An easy command is to just type 'startx' at the console prompt.

=== Uninstalling NVIDIA drivers ===
Did something in your installation go terribly, terrible wrong? No problem! Start your box up in console-only mode, navigate to where you have the nvidia driver installation file saved, and run it with the --uninstall option. ''sh filename.run --uninstall'', and it uninstalls everything it put in. Then it's a simple matter of restoring your XF86Config from the backup we made above.

== Issues ==
* One issue I've had is between the GeForce FX 5200 card and the Asus P4S8X-X motherboard. When you started the X server, it would display garbled text and then hang. This problem can be solved by going into the BIOS and setting your AGP speed down from 8x to 4x.

==See Also==
* [[NVIDIA]]

Extended attributes

2005-08-19T15:46:34Z

Jor: init

'''extended attributes''' are name-value pairs associated with a [[file]] or [[directory]] on a [[filesystem]]. The are sometimes also called '''EA''' or '''xattrs'''. These attributes are in addition to the common attributes (the [[permissions]]) each file/directory has in Linux (and other unixes]. With extended attributes you can add any name-value pair to a file or directory. The name and value can be any string you want (the maximum length of the name string, value contents and the maximum number of attributes on each file/directory is limited by the implementation of the filesystem). This kind of information added to a file/directory is usually called [[metadata]].

An example usage of these attributes is the implementation of [[POSIX]] [[Access Control Lists]] (ACL). Also an application like [[Beagle]] uses extended attributes to specify if the file has already been indexed. It could also be used for things like storing the character encoding of a file and the song information of a music/mp3 file (could theoretically also contain a thumbnail for an image file, but the size of the value is usually too restricted, so this can't be used for what [[Mac OS X]] and [[Windows|Windows XP]] call multiple streams/forks of a file).

Currently, in kernel version [[2.6]], [[ext2]],[[ext3]],[[ReiserFS]],[[JFS]],[[XFS]] and [[NFS]] all support extended attributes. It does need the option enabled at compile time. For ext2/ext3 the mount option 'user_xattr' must be used.

==Namespaces==
The name of the attributes are usually on a period-separated format like 'user.foo', the first part designates a namespace. This is used to prevent clashes between attributes of the same name. The common namespaces that are defined are:
* The 'user' namespace. This is protected by the normal unix permission settings on the file (so having write access on the file also allows the user to set an extended attribute) and is meant to be used by the user and any application that is run by the user.
* The 'root' or 'system' namespace, which can only be set with root access. ACL uses this namespace, it stores its access control lists in attributes like 'system.posix_acl_access' and 'system.posix_acl_default'.
* The 'security' namespace. For example, [[SELinux]] uses the 'security.selinux' attribute.

==Usage==
The command line tools to get an
d set these attributes are available in a package called 'attr' in most distributions. This package contains the command line tools 'getfattr' and 'setfattr' (and also a command called 'attr', which is similar to the command on [[IRIX]] and mainly aimed at people using XFS on IRIX and migrating to Linux).

Example of setting a extended attribute:
$ setfattr -n user.foo -v bar test-file

Example of getting an extended attribute:
$ getfattr -n user.foo test-file
Output will look like:
<tt>
# file: test-file
user.foo="bar"
</tt>
Example of removing an attribute:
$ setfattr -x user.foo test-file

==External links==
*[http://acl.bestbits.at/ Linux Extended Attributes and ACLs]

XFS

2005-08-17T17:51:54Z

Jor: max size clarification

'''XFS''' is a [[journaling]] [[filesystem]] originally developed by [[SGI]] for their [[IRIX]] [[operating system]]. It supports things like [[extended attributes]] and [[ACL]], [[quota]]s, B+ trees to speed up file lookup in large dirs, fully 64bits filesystem, so supports extermely large filesystems and files.

The theoretical filesystem and file size is 9 [[exabyte]]. But when run on a 32 bit [[platform]], it is limited due to the Linux kernel block layer implementation to 16 [[terabyte]].

Available in the Linux [[kernel]] as of version [[2.6]] and as an external patch for [[2.4]] kernels.

It also has the option of a [[realtime]] sub-volume (still an experimental feature in the linux version). Meaning a special part on the filesystem for data which needs near realtime responses (like when editing large video/multimedia data). The size of these sub-volumes is is currently limited to a maximum of 4 [[gigabyte]]s.

==Usage==
An XFS filesystem is created by the [[mkfs.xfs]] command, simplest use is just:
# mkfs.xfs /dev/hdXX
Where XX is some existing device/partition. But mkfs accepts many tweak options like enabling realtime partition,block sizes, label and the position of the log (can be on a seperate device)

An XFS filesystem can be managed by a set of tools (xfs_*, usually available in a package named xfsprogs in your [[distribution]]), the most important/useful are:
* xfs_admin -- change certain parameters like the [[label]] and [[UUID]]
* xfs_check -- scan XFS filesystem for errors, kind of like [[fsck]] for XFS
* xfs_copy -- copy or clone XFS filesystem
* xfs_growfs -- expand an existing XFS partition
* xfs_repair -- repair a damaged XFS partition
* xfs_rtcp -- copy a file to a realtime partition on a XFS filesystem

==External links==
*[http://oss.sgi.com/projects/xfs/ Project page]

 
''This article is a [[LQWiki:stub_articles|stub]] and needs to be finished. [[LQWiki:plunging_forward|Plunge forward]] and [[LQWiki:How_to_edit_a_page|help it grow]] !''

LQWiki:License Proposal

2005-08-17T16:39:26Z

Jor:

The current objective of the LQ Wiki is to create the "largest general Linux knowledgebase on the web". To meet this goal, I think we'll need to embrace the fact that the Open Source community consists of documentation that is under multiple licenses. I am proposing that we add the ability to add GFDL content to the LQ Wiki. On a technical level, we'd do this by adding a GFDL namespace. All GFDL content would go here, while the main namespace would remain 100% CC by-sa. Of course, no mixing of content would be allowed. I'd like to get community feedback on this proposal before we implement this. Thoughts? [[User:Jeremy|Jeremy]] 12:00, Aug 13, 2005 (EDT)

:It seems worth a try. Since CC-by-sa is the preferred license, how about one namespace for all content that's under other marginally-acceptable licenses? We can make templates to identify a page's specific license. That would probably work better than creating a new namespace for each one. Being in a separate namespace would be a big clue that the article's license isn't preferred, and the license-identification templates could have a link to a page explaining that fact (in addition to describing the license). --[[User:Yath|Yath]] 12:27, Aug 13, 2005 (EDT)

::I'm in favor of having a separate namespace for each license. This will help us easily identify which license an article in under, will make it clear to the article editor which license any content they use must be licensed under, will help prevent the mixing of licenses and will make exporting article by license easy. I think having a single namespace for all non-CC licenses would quickly get too confusing, and we'd also lose the ability to decide which licenses are indeed acceptable. [[User:Jeremy|Jeremy]] 13:39, Aug 13, 2005 (EDT)

:::Can you clarify whether you feel these non-CC-by-sa-licensed articles would be considered second-class articles, present only for completeness, or would the wiki simply be supporting several licenses simultaneously? If it's the latter, namespaces shouldn't be used for articles at all. That would put the article's license right in the title and be needlessly distracting. --[[User:Yath|Yath]] 16:57, Aug 13, 2005 (EDT)

::::They would not really be second class per-se. In the end, CC by-sa would be "preferred", but not in a way that diminished other content. At first, the only two licenses allowed would be CC by-sa and GFDL. It's absolutely not needlessly distracting as it's imperative that content with different licenses never mix. Just having everything lumped together would not be manageable IMHO and would almost be negligence, as it would seem to encourage breaking the license. [[User:Jeremy|Jeremy]] 17:20, Aug 13, 2005 (EDT)

We've gone ahead and made the change. More information is available in the [[LQWiki:Copyrights]] article. As always, feedback is appreciated. [[User:Jeremy|Jeremy]] 10:28, Aug 17, 2005 (EDT)

:I do find this a bit confusing. Ofcourse, first of all the GFDL itself already is confusing :). Specially when applied to wiki-like pages (what to do with titles, network location, invariant sections and cover texts, but IANAL ;). But I'm more concerned with that it now could be possible to have 2 pages with the same topic/name in both namespaces, which would be very confusing.It seems this is mainly aimed at allowing wikipedia content to be copy-pasted (which personally I don't find very usefull, just linking to it sounds much more usefull to me).As for the first page in this new namespace, [[GFDL:Changelog]], why is this GFDL? AFAIK (but IANAL ;p), quoting is allowed and covered by fair use. These 2 lines are already a proper quote (with a proper attribution link, well the external link does need the 'www.' removed to work, but I'm a bit hesitant to alter GFDL documents :).I think the other content on the page could just be CC by-sa licensed. Also, this namespace should really only be used when an actual large wikepedia page is copied and altered to be more aimed at linux, otherwise the page should just be altered at the wikipedia site.[[User:Jor|Jor]] 12:39, Aug 17, 2005 (EDT)

Ext2

2005-07-12T17:50:33Z

Jor: /* Usage in other OS */

'''Ext2''', also called '''ext2fs''', standing for '''Second Extended Filesystem''', is a [[filesystem]] widely used as a default option by many Linux distributions. It provides standard Unix file semantics such as [[File permissions|file permissions]] and [[symbolic link|symbolic links]].

'''[[Ext3]]''' is an enhancement to the ext2 filesystem that adds [[journalling]] support to avoid having to perform a filesystem check after a power failure or system crash.

==Details==
*Max filename length is 255 characters. Extendable to 1012 if needed. No encoding, though distros seem to be converging on UTF-8.
*Percentage of disk reserved for root to prevent users filling drive.
*Large partition support.
*Filesystem checks can be forced by administrator on boot, plus after a set number of boots.
*Advanced disk check monitoring so known when there may be problems with disk.

==Utilities==
The utilities for the ext2 file system come in the [[e2fsprogs]] package, it consists of tools like [[resize2fs]], [[e2label]], [[tune2fs]] and many more.

==Pros==
* Stable,trusted, been around for a long time

==Cons==
* No journaling

==Usage in other OS==
*Microsoft [[Windows]]
**[http://ext2fsd.sourceforge.net/ ext2fsd]
**[http://uranus.it.swin.edu.au/~jn/linux/explore2fs.htm explore2fs]
**[http://uranus.it.swin.edu.au/~jn/linux/ext2ifs.htm ext2ifs]
**[http://www.fs-driver.org/index.html ext2 IFS for Windows]
{{stub}}

Autoconf

2005-03-29T19:50:54Z

Jor: /* External links */

'''autoconf''' is a collection of tools to generate [[port]]able [[configure script]]s. It is part of the [[GNU build system]] ([[automake]],[[libtool]],etc.). It uses the [[m4]] macro processor the process template files (configure.ac/configure.in) into configure scripts.

The autconf package comes with the following tools:
* autoconf -- command to generate a configure script from a template file
* autoscan -- generate a preliminary configure template file
* autoreconf - update a generated configure script
* autoheader -- generate a configuration header file (config.h)
* autoupdate -- update the template file to a new autoconf version

The input templates, usually named <tt>configure.ac</tt> or <tt>configure.in</tt>, specify which tests are performed in the resulting configure script. These tests are implemented as m4 macros and autoconf comes with a whole bunch of standard scripts and specialized tests/macros can easily be added. There are already a lot of extra macros around, a big list of the can be found [http://www.gnu.org/software/ac-archive/htmldoc/index.html here].

Autoconf also includes a tool to get you started writing your configure.ac file named [[autoscan]]. Like the name suggested, it scans the source files of your project and generates a preliminary template file called configure.scan. This file can then manually be checked/extended and renamed to configure.ac.

==External links==
*[http://www.gnu.org/software/autoconf/ Project website]
*[http://sources.redhat.com/autobook/ AutoBook]

{{stub}}

Transport Tycoon Deluxe

2005-03-18T18:17:24Z

Jor: init

'''Transport Tycoon Deluxe''' is a classic computer game made by Chris Sawyer[http://www.chrissawyer.com/]. It a simulation game in which you control a transportation business. Transport Tycoon was released in 1994, and the 'Deluxe' version in 1995. This version added more vehicles, more game worlds, a scenario editor and many other enhanced features.

Later an unofficial patch (TTDPatch[http://www.ttdpatch.net/]) was created by dedicated fans. It added even more features and made the game work on newer versions of the Windows OS.

==Linux==
The original [[DOS]] version of Transport Tycoon Deluxe can be played in [[DOSBox]]. The dos version patched by TTDPatch is also said to work in DOSBox. The windows version of TTDPatch does not work out-of-the-box in [[Wine]], see the TTDPatch website for more information on that.

Last, but not least: There is also a clone, called [[OpenTTD]] which is an engine that can run on Linux. It requires the data files from the original game.

==See also==
*[[simutrans]]

Talk:Game List

2005-03-18T17:34:33Z

Jor: license types

I would suggest organizing the list into: open source, closed source but native, and under emulation. Organize it into genre after that. [[User:Crazyeddie|Crazyeddie]] 15:25, Jan 22, 2005 (EST)

: Sounds like a good plan to me. [[User:Jeremy|Jeremy]] 18:52, Jan 22, 2005 (EST)

SciYro, since you're doing the actual work, it's your call. I don't think the matter is important enough for a vote or consensus poll or whatever. But it might save some time if we sat down and worked out how we want to organize this thing. I think your way might be better than mine. Remember that non-native games have been declared ontopic, provided that the article focuses on getting emulation running. [[User:Crazyeddie|Crazyeddie]] 16:43, Jan 23, 2005 (EST)

yes, i was planning to put non-native apps into a different page for people to see how well there windows games can be run under some of the windows emulators, as it seemed to me to be a different topic then what games exist for Linux. -- SciYro

Sounds good. Might also be good to have a single page on emulation tips in general, and then break it down into tips for each individual game - either in subsections or seperate articles, depending on length. [[User:Crazyeddie|Crazyeddie]] 17:42, Jan 23, 2005 (EST)

-----------------

Could someone who knows the loki games move them over from [[gaming]] onto this page?, i cant seem to find any relavent info on them besides they are binary, ported to native linux, and cost money (i cant find any descrition on them at all, nor can i play them to see) -- SciYro

Well, for starters, I'm pretty sure Loki went belly-up a long time ago. That might be why there isn't much information out there. [[User:Crazyeddie|Crazyeddie]] 12:45, Mar 10, 2005 (EST)

[http://en.wikipedia.org/wiki/Loki_Games Here's] the wikipedia article. Remember to paraphrase. Looks like they went out of business in 2002. They made ports of existing windows games, and did a few things that made the ports run better on Linux. [[User:Crazyeddie|Crazyeddie]] 12:47, Mar 10, 2005 (EST)

thanks, ill get on bringing those games over to this page then in the next few days. -- SciYro

-----------------
Problem with specifying the license type is that it can soon become pretty complicated, the 3 options given in the introduction don't cover every case very clearly. Example are both the transport tycoon type games I just added. OpenTTD is Open source, GPL licensed but requires the non-free data files from a official Windows Transport Tycoon Deluxe game. Simutrans has got it's own very non free license, but is [free-as-in-beer]. I'm not sure if it's a good idea to put this info on this list page, I would just keep the license information of the page of the game itself. But if you really want this, maybe only make a difference between [[Open Source Initiative]] approved licenses and non-OSI approved licenses (but that still doens't really cover the case when the code is GPL but requires non-free data files to be usefull/playable). --[[User:Jor|Jor]] 12:34, Mar 18, 2005 (EST)

Game List

2005-03-18T17:22:56Z

Jor: tt's

This page is to list games that run natively on [[GNU]]/[[Linux]], that means they don't require a emulator to run. See [[Gaming]] section for more info on gaming in Linux.

The page will be formated as follows: ''Game Name - type - description''
Were type will one of "OSS" if its [[open source]] software, "Binary" if its only a binary version only. If type is preceded by a "$" sign, then the game is not free.

Click on the links for a better description of the game, and a link to its homepage. Some games are ports to GNU/Linux. There may also be games derived directly from other games, such games may or may not be listed on this page, but in there parents page, ''[[Doom]]'' for example has tons of ports, and derived games, many are not listed on this page, but the parent ''Doom'' is.

==Action/Arcade==
Pretty much any game that wont fit anywhere else.
*[[Overkill|0verkill]] - OSS - A bloody 2D death match game, in [[ASCII art]].
*[[Abuse]] - OSS - A popular 2D platform scroller, kill the aliens, and save the world.
*[[Angel Wars]] - OSS - A scrolling shooter game.
*[[Amphetamine]] - OSS - A 2D jump'n run shooter, with some nice visual effects.
*[[Armagetron]] - OSS - The fast bike game from the movie ''Tron''.
*[[Atanks]] - OSS - Another Scorched earth clone.
*[[Bitstream]] - OSS - A action game that will be similar to ''Star Fox''.
*[[BZFlag]] - OSS - A 3D multiplayer tank battle game.
*[[Chromium]] - OSS - A fast paste scrolling space shooter.
*[[Dave Gnukem]] - OSS - Another 2D platform scrolling shooter.
*[[Defendguin]] - OSS - ''Defender'' clone with a Linux theme.
*[[Feuerkraft]] - OSS - Tank battle game.
*[[Freedroid]] - OSS - A clone of a old game from C64, gotta kill the robots.
*[[Galaxium]] - OSS - A 3D take on scrolling space shooters.
*[[GLTron]] - OSS - This is another implementation of the bike game from ''Tron''.
*[[ICBM3D]] - OSS - A 3D ''Missile Command'' styled game.
*[[JumpnBump]] - OSS - A multiplayer game, jump on top of bunnies.
*[[LiquidWar]] - OSS - A unique multiplayer game where you control a liquid army.
*[[Mad Bomber]] - OSS - Catch falling bombs before they hit the ground.
*[[Maelstrom]] - OSS - ''Asteroids'' clone.
*[[Penguin Command]] - OSS - Remember the fun old ''Missile Command'' game?

==FPS==
This is for ''First Person Shoot'' 'em ups (see [[FPS]]).
*[[America's Army]] - Binary - extremely big, and made by the US military.
*[[Black Shades]] - OSS - Protect a VIP from assassins.
*[[Chickens]] - OSS - Kill all the chickens, if 1 passes you, the world ends.
*[[Cube]] - OSS - A bit like quake, with some weird monsters.
*[[Doom]] - OSS - This is a very fun game, tho in its day it was considered to have good graphics.
*[[Doom 3]] - $Binary - This is third installment of Doom, its not free, but has good graphics.
*[[Dreams Eternal]] - OSS - A FPS/puzzle solving game, centered around dreams .. still in development.
*[[Parsec]] - OSS - A multiplayer space combat game.
*[[Quake]] - OSS/$ - This a bit more modern then regular doom, tho only the game engine is free and open sourced, meaning you still need the data files to play, and they are not free.
*[[Quake 2]] - OSS/$ - This is the second installment of Quake, and like its predecessor only the game engine is free and open sourced.
*[[Quake 3 Arena]] - $Binary - This is the third, and most modern version of Quake.
*[[Return to Castle Wolfenstein]] - $Binary - This is a modern game, and the sequel to Wolfenstein.
*[[ROTT|Rise of the Traid]] - OSS - Another relatively old DOS game.
*[[Unreal Tournament]] - $Binary - A extremely popular shoot firt game.
*[[Wolfenstein]] - OSS - This is the old DOS version of Wolfenstein.
*[[Wolfenstein: Enemy Territory]] - Binary - A free multiplayer game based off of [[Return to Castle Wolfenstein]].

==Racing==
Racing games involve racing.
*[[Tux Kart]] - OSS - A simple racing game, reminds you of some of the games from the [[N64]].
*[[Tux Racer]] - OSS - Send [[Tux]] racing down mountain slopes eating hearing.

==RPG/Strategy==
This section is for any type of ''Role Playing Game'' (see [[RPG]] and [[MMORPG]]), including strategy games (see [[RTS]]).
*[[Angband]] - OSS - A dungeon exploration game.
*[[ASC]] - OSS - A turn based strategy game.
*[[Crossfire]] - OSS - A big multiplayer advanture RPG, only downside is bad graphics.
*[[Daimonin]] - OSS - A fairly unique MMOPRG based on [[Crossfire]].
*[[Egoboo]] - OSS - A 3D dungeon crawler game.
*[[Eternal Lands]] - OSS - This is a cross-platform, 3D MMORPG fantasy world.
*[[FreeCiv]] - OSS - A multiplayer turn based strategy game, build a city, then a army.
*[[FreeCNC]] - OSS - A free implementation of the ''Command and Conquer'' game engine.
*[[FreedroidRPG]] - OSS - You, a penguin, must defeat the evil robots of the company "Monopoly Software".
*[[Globulation 2]] - OSS - A very good looking RTS game, with a nice feel to it.
*[[Lincity]] - OSS - Think simcity with bad 2d graphics, but more realistic style of play.
*[[Nethack]] - OSS - A [[roguelike|Rogue-like]] game, in the classic [[ASCII]].
*[[Neverwinter Nights]] - $Binary - A RPG based on ''Dungeons and Dragons''.
*[[Savage]] - $Binary - a game with classic RTS and FPS viewpoints.
*[[Vega Stike]] - OSS - a fun space trading game, only downside is no network or multiplayer support.
*[[Wesnoth]] - OSS - Turn based strategy game, with a fantasy theme.

==Puzzle/Educational==
This section encompasses games to make people think, or solve problems.
*[[Batalla Naval]] - OSS - A networked battleship game.
*[[BillardGL]] - OSS - A 3D pool game.
*[[Crack Attack]] - OSS - A game similar to Tetris.
*[[DROD]] - OSS - Get thru dungeon rooms, solving puzzles, and killing monsters.
*[[FooBillard]] - OSS - Another 3D pool game.
*[[Frozen bubble]] - OSS - A simple, and addictive game.
*[[Gem Drop X]] - OSS - Match up the gems before they fill the screen.
*[[glChess]] - OSS - A 3D chess program, mainly a front for GNU Chess, no network support.
*[[Lbreakout 2]] - OSS - For those that don't remember the old ''Break out'' game, think brick breaking mixed with ping-pong.
*[[Kiki]] - OSS - You, a nano-bot, must get past levels by finding a way to get to the key on each level.
*[[PDA Maze]] - OSS - A 3D maze, originally for a Linux based PDA.
*[Pysol]] - OSS - A collection of over 200 card games to play.
*[[TuxMath]] - OSS - Similar to ''Missile Command'', but you must solve math problems instead.
*[[Xboard]] - OSS - A 2D GNU Chess front end, as well as a few other, very feature rich.
*[[XPuyopuyo]] - OSS - Another game similar to Tetris.

==Simulation==
Simulations are meant for teaching people about certain things, for example on how to operate machines, or demonstrate what-ifs.
*[[CoreWars]] - OSS - Create small programs to crash another in a virtual computer.
*[[FlightGear]] - OSS - A flight simulator, designed to be expandable.
*[[YS Flight Simulation]] - OSS - Another flight simulator, has relatively low CPU requirements.
*[[OpenTTD]] - OSS - Clone of "Transport Tycoon Deluxe" from Microprose
*[[Simutrans]] - [[free-as-in-beer]] - Transport business simulation, similar to Transport Tycoon but not a direct clone.

Talk:Gaming

2005-03-18T17:17:17Z

Jor: new list

Er...what criteria should we use to determine if a game gets listed here? I'm on the verge of going through Portage and seriously buffing out our repertoire...

[[User:EvilSporkMan|EvilSporkMan]] 16:08, Mar 7, 2004 (EST)

----

this section seems to be a bit clustered, would anyone object if i moved all the games listed into a separate page, and use this one for gaming info? -- SciYro

This is a top level page, directly linked to from the main page. I think it should be a list of gaming topics, and a list of games should be made elsewhere. So basically, yes, SciYro. [[User:Crazyeddie|Crazyeddie]] 04:11, Jan 22, 2005 (EST)

: +1 from me. [[User:Jeremy|Jeremy]] 14:44, Jan 22, 2005 (EST)

ok, the new [[Game List]] page has all the linux games from this page (that still exist for linux, and had pages already made on this wiki), ill delete the games list from this page soon, but there is still a lot of clutter from all the various links to console, and stray words on this page ..... so before i go deleting i would like to hear suggestions for a new non-cluttered format to re-write this page with (as the current state is pretty much all links, no info, and all in a almost random order) -- SciYro

: I'm still not sure which list I prefer. I think the [[Game List]] is a bit hard to read because of the 'multi-line' short descriptions some games have, but I can live with that. My suggestion for this page would be:
* some general info about gaming in linux? (I really have no idea what that would be :))
* Link to [[game List]], which should list native games and link to a 'games through emulation' list?
* Link to some [[game development]] page, which links to things like [[SDL]].
* Link to [[OpenGL]] and maybe some [[X]] page for configuring the system for optimum performence --[[User:Jor|Jor]] 12:17, Mar 18, 2005 (EST)

Talk:Gaming

2005-03-18T16:59:07Z

Jor: fix link

Ubuntu Linux

2005-03-18T13:08:22Z

Jor: update releases info

'''Ubuntu Linux''' is a free, [[Debian]]-based, [[GNOME]] oriented [[distribution]], on a 6-month release cycle.

Initial release : Ubuntu 4.10 "The Warty Warthog", released October 2004.

Next release: 5.04 "The Hoary Hedgehog" scheduled for release in April 2005, (preview/test CDs available now).

== Replacing the CD as a source ==

If you have the Warty ISO already then you can loop mount it and add it as a repository. Enable the loop devices by running:
sudo modprobe loop

To enable loop devices automatically on startup edit the file ''/etc/modules'' and add the line:
loop

Now add to ''/etc/fstab'' the line:
/path/to/ISO/image/warty.iso /mnt/ubuntu iso9660 ro,loop,auto 0 0

Mount the ISO image with:
sudo mount /mnt/ubuntu

Start Synaptic, disable the CD entry and add a new repository with the following details:
Type: Binary
URI: file::///mnt/ubuntu
Distro: unstable
Sections: main restricted

== Replace the GNOME foot ==

Ubuntu comes with the Gnome foot on the Applications menu. If you want to change it to look like this: 
http://www.tux.com.au/wiki/images/2/2f/App-button.png

All you need to do is right click the image below, select "Save image as.." to a location you can remember and have rights to. Now all you have to do is run the following commands:
cd /usr/share/pixmaps/
sudo mv gnome-logo-icon-transparent.png gnome-logo-icon-transparent.bak
sudo mv /path/to/Gnome-logo-icon-transparent.png ./gnome-logo-icon-transparent.png

http://www.tux.com.au/wiki/images/a/a8/Gnome-logo-icon-transparent.png

You need to restart Gnome, logout then hit Ctrl+Alt+Backspace. Login and you will now see the Ubuntu logo instead of the Gnome foot.

==External links==
* IRC, the #ubuntu channel on the server irc.freenode.net
* [http://ubuntulinux.org/ Ubuntu Linux homepage]
* [http://www.ubuntulinux.org/ubuntu/ About Ubuntu]
* The [http://www.ubuntuguide.org/ UnOfficial Ubuntu 4.10 start guide]
* The [http://www.ubuntuforums.org/ Ubuntu forum] will let you search but you can't post a question until you create an account.

Ext2

2005-02-21T16:26:08Z

Jor: use in win

'''Ext2''', also called '''ext2fs''', standing for '''Second Extended Filesystem''', is a [[filesystem]] widely used as a default option by many Linux distributions. It provides standard Unix file semantics such as [[File permissions|file permissions]] and [[symbolic link|symbolic links]].

'''[[Ext3]]''' is an enhancement to the ext2 filesystem that adds [[journalling]] support to avoid having to perform a filesystem check after a power failure or system crash.

==Details==
*Max filename length is 255 characters. Extendable to 1012 if needed.
*Percentage of disk reserved for root to prevent users filling drive.
*Large partition support.
*Filesystem checks can be forced by administrator on boot, plus after a set number of boots.
*Advanced disk check monitoring so known when there may be problems with disk.

==Utilities==
The utilities for the ext2 file system come in the [[e2fsprogs]] package, it consists of tools like [[resize2fs]], [[e2label]], [[tune2fs]] and many more.

==Pros==
* Stable,trusted, been around for a long time

==Cons==
* No journaling

==Usage in other OS==
*Microsoft [[Windows]]
**[http://ext2fsd.sourceforge.net/ ext2fsd]
**[http://uranus.it.swin.edu.au/~jn/linux/explore2fs.htm explore2fs]

{{stub}}

Using SSH

2005-02-21T16:00:59Z

Jor: ssh-copy-id

'''[[SSH]]''' or Secure Shell allows you to remotely login to and work on your system. There are several implementations for SSH around, the most commonly used in Linux is [[OpenSSH]], but there are other [[SSH#Implementations|SSH implementations]].

To SSH in Linux from the [[Bash]] command prompt you would type:

$ ssh user@remote1

"user" is the login name and "remote" would be the [[DNS]] name of the computer you want to login to. If this the first time that you have connected to that computer you will be shown the name and a "fingerprint" number, and asked to confirm it. For all future sessions SSH then uses the fingerprint to ensure that the other computer is the same system as before, and not merely claiming the same name.

SSH and [[Telnet]] are almost the same, but in SSH all of the information is [[encrypted]] so that even if someone intercepted the packets, they would not be able to read or use them (theoretically).

To SSH from Windows, you need an application like [[PuTTY]] or TeraTerm.

=== Using scp ===

SSH also provides a utility called [[scp]] (Secure CoPy) to encrypt file transfers from one host to another. The basic syntax of scp works like this:

To move a file to a remote host:

$ scp filename user@remote1:/path/to/destination/dir

So if we wanted to copy the file named "foo" to a remote host "remote1" with the user "tux" and a target directory of "/home/tux/files/" you would use:

$ scp foo tux@remote1:/home/tux/files/

To copy a file from the remote host to the current directory:

$ scp user@remote1:/path/to/file filename

So using the same example, to copy the file "/home/tux/files/foo" on the remote host and rename it as "baz" on our current system we'd do:

$ scp tux@remote1:/home/tux/files/foo baz

This method even allows to securely copy files between remote hosts if you have SSH access to this hosts:

$ scp user1@remote1:foo.txt user2@remote2:foo_from_remote1.txt

will copy "foo.txt" from remote1 to "foo_from_remote1.txt" on remote2.

OpenSSH includes the subsystem [[SFTP]]. It is (quite predictably) an [[FTP]] clone that runs through SSH:

$ sftp tux@remote1
> get foo

=== WinSCP - A Windows implementation of scp ===

WinSCP is a graphical SSH (SFTP and SCP) client for Windows. Its main function is secure copying of files between a local and a remote computer, and it works much like the FTP clients in common use. Beyond this function, WinSCP manages some other actions with files. A plugin for FAR manager is available too. You can find more information and download WinSCP here:

http://sourceforge.net/projects/winscp

=== SSHDos - A DOS implementation of ssh ===

SSHDos is a pure DOS SSH (SSH, SFTP and SCP) client. It's a convenient way of transferring files from linux to dos or the other way around, requiring a minimum of fuss. It also supports terminal emulation, making it a good choice for thin clients on ancient hardware (telnet included if ssh is too heavy).

http://sshdos.sourceforge.net

=== Public key authentication ===

Another cool feature of SSH is that you can setup public/private key authentication, which will allow you to login to remote machines without retyping your password, provided that you hold the private key that is paired with the public key on the remote machine. Using a key to login is also potentially more secure, since the user must have access to a file holding the private key to login, rather than simply being able to guess a password. To create a pair of keys type this:

$ ssh-keygen -t dsa -f ~/.ssh/id_dsa -N

This will create a public/private key pair in your ''~/.ssh'' directory - the keys are the files ''id_dsa'' (the private key) and ''id_dsa.pub'' (the public key). Now you will need to append ''id_dsa.pub'' to the ''~user/.ssh/authorized_keys'' file in your home directory on the remote server. OpenSSH comes with a command 'ssh-copy-id' which can do this for you.

now you can easily:

$ ssh user@remote

and get the shell without typing the password. If the usernames are the same on the local and remote hosts you can even omit the username. There are, of course, some pre-requisites - on your local machine ''/etc/ssh/ssh_config'' should contain the line:

IdentityFile ~/.ssh/id_dsa

and the remote ''/etc/ssh/sshd_config'' should have the lines:

PubkeyAuthentication yes

AuthorizedKeysFile .ssh/authorized_keys

One problem that is tricky to debug is when the permissions of the ''.ssh'' directory on the remote host are not set up correctly. SSH will not use the keys unless the write permissions are removed for group and others; this prevents other people from dumping their public keys into your ''.ssh'' directory and then accessing your account. Set the ''.ssh'' directory permissions to "700".

Another useful trick with SSH - you can suspend your SSH session and resume it.
If you work from console and you SSH to remote1 your shell becomes remote1, so if you need to do something on your local machine you would normally have another terminal open, but you can avoid this by placing the SSH session in background by typing ''suspend'' or ''~'' followed by ''ctrl+z'', and the SSH session will go to the background. To bring back your SSH session on remote1 just type ''fg''.

The following page has a nice tutorial: http://fedoranews.org/dowen/sshkeys/

=== Forwarding X11 ===

You can have the server you login to forward [[X]] to the [[X server]] controlling the display on your local machine, so that you can run programs on remote computers and they'll appear on your screen alongside the programs running on your local machine.

This can be done by adding "ForwardX11 yes" to your ''~/.ssh/config''. SSH will handle setting your display automatically, which includes setting the DISPLAY environment variable (usually to localhost:10, or a higher-numbered display if that is taken), and setting up the [[xauth]] file for access to the [[X Server]] running on the SSH client computer.

For this to work the remote sshd must also be configured to allow X forwarding. This option is called "X11Forwarding yes" and resides in ''/etc/ssh/sshd_config''.

Once these are set you can access remote programs at will, like so:

$ ssh user@remote1:/usr/bin/gedit &

Alternatively to setting "ForwardX11 yes", use the command ''ssh -X remote'' to connect. Also, you can use ''ssh -x remote'' to disable it for a single session if "ForwardX11 yes" is in your ''~/.ssh/config'' file.

In recent openSSH, there is also ''ssh -Y'' (or the option "ForwardX11Trusted yes") . This enables 'trusted' X11 forwarding.
This used to be the default with older versions, but because of a potential security vulnerability (which allowed other X11 clients to 'sniff' or alter data from the remote client) this was changed.
Most people won't need to use this '-Y' option (and should not have "ForwardX11Trusted yes" set in their ssh config file).

Talk:Using SSH

2005-02-09T11:13:41Z

Jor:

I'm not sure why this happens, but with my firefox the page is way over the width on my 1600x1200 monitor. I't probably because long ''italic'' texts or something, but hopefully someone knows why and fixes the problem. 
[[User:Streg|Streg]] 09:43, Mar 7, 2004 (EST)

Should be fixed now, I did a spell check and it ruined the </tt> tag, which I subsequently restored. 
[[User:Bj�rn|Bj�rn]] 09:48, Mar 7, 2004 (EST)

Yes, it works now. It's great that the problem is not with my browser =) 
[[User:Streg|Streg]] 09:50, Mar 7, 2004 (EST)

== ssh -Y: Do you mean: "You should use -X instead?" ==

The article reads: "In recent openSSH, there is also ssh -Y (or the option "ForwardX11Trusted yes") . This enables 'trusted' X11 forwarding. This used to be the default with older versions, but because of a potential security vulnerability (which allowed other X11 clients to 'sniff' or alter data from the remote client) this was changed. Most people won't need to bother with this option and should stick with the default."

Just to make sure I understand right, you mean to say:

"ssh -Y uses "trusted" X11 forwarding. It's insecure, there's a security hole in it. It's better to use -X, because it's secure: nobody can snoop, nobody can interfere."

Specificly, the line "Most people won't need to bother with this option and should stick with the default" is confusing, if you are like me, and don't understand this. By "default," do you mean "you shouldn't be forwarding X11?" Or do you mean, "you should be using -X"?

[[User:LionKimbro|LionKimbro]]
:The latter. Use '-X' and forget about '-Y'. I said 'default' because it can be set/overriden in the config file, which would make -X behave as -Y does. --[[User:Jor|Jor]] 06:13, Feb 9, 2005 (EST)

Using SSH

2005-02-09T11:10:16Z

Jor: more clear?

'''[[SSH]]''' or Secure Shell allows you to remotely login to and work on your system. There are several implementations for SSH around, the most commonly used in Linux is [[OpenSSH]], but there are other [[SSH#Implementations|SSH implementations]].

To SSH in Linux from the [[Bash]] command prompt you would type:

$ ssh user@remote1

"user" is the login name and "remote" would be the [[DNS]] name of the computer you want to login to. If this the first time that you have connected to that computer you will be shown the name and a "fingerprint" number, and asked to confirm it. For all future sessions SSH then uses the fingerprint to ensure that the other computer is the same system as before, and not merely claiming the same name.

SSH and [[Telnet]] are almost the same, but in SSH all of the information is [[encrypted]] so that even if someone intercepted the packets, they would not be able to read or use them (theoretically).

To SSH from Windows, you need an application like [[PuTTY]] or TeraTerm.

=== Using scp ===

SSH also provides a utility called [[scp]] (Secure CoPy) to encrypt file transfers from one host to another. The basic syntax of scp works like this:

To move a file to a remote host:

$ scp filename user@remote1:/path/to/destination/dir

So if we wanted to copy the file named "foo" to a remote host "remote1" with the user "tux" and a target directory of "/home/tux/files/" you would use:

$ scp foo tux@remote1:/home/tux/files/

To copy a file from the remote host to the current directory:

$ scp user@remote1:/path/to/file filename

So using the same example, to copy the file "/home/tux/files/foo" on the remote host and rename it as "baz" on our current system we'd do:

$ scp tux@remote1:/home/tux/files/foo baz

This method even allows to securely copy files between remote hosts if you have SSH access to this hosts:

$ scp user1@remote1:foo.txt user2@remote2:foo_from_remote1.txt

will copy "foo.txt" from remote1 to "foo_from_remote1.txt" on remote2.

OpenSSH includes the subsystem [[SFTP]]. It is (quite predictably) an [[FTP]] clone that runs through SSH:

$ sftp tux@remote1
> get foo

=== WinSCP - A Windows implementation of scp ===

WinSCP is a graphical SSH (SFTP and SCP) client for Windows. Its main function is secure copying of files between a local and a remote computer, and it works much like the FTP clients in common use. Beyond this function, WinSCP manages some other actions with files. A plugin for FAR manager is available too. You can find more information and download WinSCP here:

http://sourceforge.net/projects/winscp

=== SSHDos - A DOS implementation of ssh ===

SSHDos is a pure DOS SSH (SSH, SFTP and SCP) client. It's a convenient way of transferring files from linux to dos or the other way around, requiring a minimum of fuss. It also supports terminal emulation, making it a good choice for thin clients on ancient hardware (telnet included if ssh is too heavy).

http://sshdos.sourceforge.net

=== Public key authentication ===

Another cool feature of SSH is that you can setup public/private key authentication, which will allow you to login to remote machines without retyping your password, provided that you hold the private key that is paired with the public key on the remote machine. Using a key to login is also potentially more secure, since the user must have access to a file holding the private key to login, rather than simply being able to guess a password. To create a pair of keys type this:

$ ssh-keygen -t dsa -f ~/.ssh/id_dsa -N

This will create a public/private key pair in your ''~/.ssh'' directory - the keys are the files ''id_dsa'' (the private key) and ''id_dsa.pub'' (the public key). Now you will need to append ''id_dsa.pub'' to the ''~user/.ssh/authorized_keys'' file in your home directory on the remote server.

now you can easily:

$ ssh user@remote

and get the shell without typing the password. If the usernames are the same on the local and remote hosts you can even omit the username. There are, of course, some pre-requisites - on your local machine ''/etc/ssh/ssh_config'' should contain the line:

IdentityFile ~/.ssh/id_dsa

and the remote ''/etc/ssh/sshd_config'' should have the lines:

PubkeyAuthentication yes

AuthorizedKeysFile .ssh/authorized_keys

One problem that is tricky to debug is when the permissions of the ''.ssh'' directory on the remote host are not set up correctly. SSH will not use the keys unless the write permissions are removed for group and others; this prevents other people from dumping their public keys into your ''.ssh'' directory and then accessing your account. Set the ''.ssh'' directory permissions to "700".

Another useful trick with SSH - you can suspend your SSH session and resume it.
If you work from console and you SSH to remote1 your shell becomes remote1, so if you need to do something on your local machine you would normally have another terminal open, but you can avoid this by placing the SSH session in background by typing ''suspend'' or ''~'' followed by ''ctrl+z'', and the SSH session will go to the background. To bring back your SSH session on remote1 just type ''fg''.

The following page has a nice tutorial: http://fedoranews.org/dowen/sshkeys/

=== Forwarding X11 ===

You can have the server you login to forward [[X]] to the [[X server]] controlling the display on your local machine, so that you can run programs on remote computers and they'll appear on your screen alongside the programs running on your local machine.

This can be done by adding "ForwardX11 yes" to your ''~/.ssh/config''. SSH will handle setting your display automatically, which includes setting the DISPLAY environment variable (usually to localhost:10, or a higher-numbered display if that is taken), and setting up the [[xauth]] file for access to the [[X Server]] running on the SSH client computer.

For this to work the remote sshd must also be configured to allow X forwarding. This option is called "X11Forwarding yes" and resides in ''/etc/ssh/sshd_config''.

Once these are set you can access remote programs at will, like so:

$ ssh user@remote1:/usr/bin/gedit &

Alternatively to setting "ForwardX11 yes", use the command ''ssh -X remote'' to connect. Also, you can use ''ssh -x remote'' to disable it for a single session if "ForwardX11 yes" is in your ''~/.ssh/config'' file.

In recent openSSH, there is also ''ssh -Y'' (or the option "ForwardX11Trusted yes") . This enables 'trusted' X11 forwarding.
This used to be the default with older versions, but because of a potential security vulnerability (which allowed other X11 clients to 'sniff' or alter data from the remote client) this was changed.
Most people won't need to use this '-Y' option (and should not have "ForwardX11Trusted yes" set in their ssh config file).

Talk:RPM

2005-02-08T21:39:19Z

Jor: naming guidelines

I wish there were some notes on naming conventions here.

I was confused for a long time by names such as "perl-SVG-Metadata-0.11-1.0.rh8.dag.i386.rpm", because "dag" wasn't a system component that I knew of. Later I find that [http://dag.wieers.com/home-made/apt/FAQ.php#A1 dag is the name of a distribution of RPMs,] named after it's founder. I think that this kind of information should be in the wiki.

My understanding of RPM filenames is incomplete, though, so I'm not writing something here on filename standards myself.

[[User:LionKimbro|LionKimbro]]

:[[AFAIK]] there is no official common standard for this. The closest thing is the package naming guidlines from the Fedora project, but I'm not sure what that current status is of that. Every RPM distribution just use their own variation on that naming scheme. There isn't a strict guideline yet like there is for the Debian .deb packages. There's also the the specification from the [[LSB]] I believe, but this applies to the 'naming' of the package and not the filename (which can be anything, since it doesn't matter much anyway). --[[User:Jor|Jor]] 16:39, Feb 8, 2005 (EST)

Using SSH

2005-02-08T16:45:35Z

Jor: -Y

'''[[SSH]]''' or Secure Shell allows you to remotely login to and work on your system. There are several implementations for SSH around, the most commonly used in Linux is [[OpenSSH]], but there are other [[SSH#Implementations|SSH implementations]].

To SSH in Linux from the [[Bash]] command prompt you would type:

$ ssh user@remote1

"user" is the login name and "remote" would be the [[DNS]] name of the computer you want to login to. If this the first time that you have connected to that computer you will be shown the name and a "fingerprint" number, and asked to confirm it. For all future sessions SSH then uses the fingerprint to ensure that the other computer is the same system as before, and not merely claiming the same name.

SSH and [[Telnet]] are almost the same, but in SSH all of the information is [[encrypted]] so that even if someone intercepted the packets, they would not be able to read or use them (theoretically).

To SSH from Windows, you need an application like [[PuTTY]] or TeraTerm.

=== Using scp ===

SSH also provides a utility called [[scp]] (Secure CoPy) to encrypt file transfers from one host to another. The basic syntax of scp works like this:

To move a file to a remote host:

$ scp filename user@remote1:/path/to/destination/dir

So if we wanted to copy the file named "foo" to a remote host "remote1" with the user "tux" and a target directory of "/home/tux/files/" you would use:

$ scp foo tux@remote1:/home/tux/files/

To copy a file from the remote host to the current directory:

$ scp user@remote1:/path/to/file filename

So using the same example, to copy the file "/home/tux/files/foo" on the remote host and rename it as "baz" on our current system we'd do:

$ scp tux@remote1:/home/tux/files/foo baz

This method even allows to securely copy files between remote hosts if you have SSH access to this hosts:

$ scp user1@remote1:foo.txt user2@remote2:foo_from_remote1.txt

will copy "foo.txt" from remote1 to "foo_from_remote1.txt" on remote2.

OpenSSH includes the subsystem [[SFTP]]. It is (quite predictably) an [[FTP]] clone that runs through SSH:

$ sftp tux@remote1
> get foo

=== WinSCP - A Windows implementation of scp ===

WinSCP is a graphical SSH (SFTP and SCP) client for Windows. Its main function is secure copying of files between a local and a remote computer, and it works much like the FTP clients in common use. Beyond this function, WinSCP manages some other actions with files. A plugin for FAR manager is available too. You can find more information and download WinSCP here:

http://sourceforge.net/projects/winscp

=== SSHDos - A DOS implementation of ssh ===

SSHDos is a pure DOS SSH (SSH, SFTP and SCP) client. It's a convenient way of transferring files from linux to dos or the other way around, requiring a minimum of fuss. It also supports terminal emulation, making it a good choice for thin clients on ancient hardware (telnet included if ssh is too heavy).

http://sshdos.sourceforge.net

=== Public key authentication ===

Another cool feature of SSH is that you can setup public/private key authentication, which will allow you to login to remote machines without retyping your password, provided that you hold the private key that is paired with the public key on the remote machine. Using a key to login is also potentially more secure, since the user must have access to a file holding the private key to login, rather than simply being able to guess a password. To create a pair of keys type this:

$ ssh-keygen -t dsa -f ~/.ssh/id_dsa -N

This will create a public/private key pair in your ''~/.ssh'' directory - the keys are the files ''id_dsa'' (the private key) and ''id_dsa.pub'' (the public key). Now you will need to append ''id_dsa.pub'' to the ''~user/.ssh/authorized_keys'' file in your home directory on the remote server.

now you can easily:

$ ssh user@remote

and get the shell without typing the password. If the usernames are the same on the local and remote hosts you can even omit the username. There are, of course, some pre-requisites - on your local machine ''/etc/ssh/ssh_config'' should contain the line:

IdentityFile ~/.ssh/id_dsa

and the remote ''/etc/ssh/sshd_config'' should have the lines:

PubkeyAuthentication yes

AuthorizedKeysFile .ssh/authorized_keys

One problem that is tricky to debug is when the permissions of the ''.ssh'' directory on the remote host are not set up correctly. SSH will not use the keys unless the write permissions are removed for group and others; this prevents other people from dumping their public keys into your ''.ssh'' directory and then accessing your account. Set the ''.ssh'' directory permissions to "700".

Another useful trick with SSH - you can suspend your SSH session and resume it.
If you work from console and you SSH to remote1 your shell becomes remote1, so if you need to do something on your local machine you would normally have another terminal open, but you can avoid this by placing the SSH session in background by typing ''suspend'' or ''~'' followed by ''ctrl+z'', and the SSH session will go to the background. To bring back your SSH session on remote1 just type ''fg''.

The following page has a nice tutorial: http://fedoranews.org/dowen/sshkeys/

=== Forwarding X11 ===

You can have the server you login to forward [[X]] to the [[X server]] controlling the display on your local machine, so that you can run programs on remote computers and they'll appear on your screen alongside the programs running on your local machine.

This can be done by adding "ForwardX11 yes" to your ''~/.ssh/config''. SSH will handle setting your display automatically, which includes setting the DISPLAY environment variable (usually to localhost:10, or a higher-numbered display if that is taken), and setting up the [[xauth]] file for access to the [[X Server]] running on the SSH client computer.

For this to work the remote sshd must also be configured to allow X forwarding. This option is called "X11Forwarding yes" and resides in ''/etc/ssh/sshd_config''.

Once these are set you can access remote programs at will, like so:

$ ssh user@remote1:/usr/bin/gedit &

Alternatively to setting "ForwardX11 yes", use the command ''ssh -X remote'' to connect. Also, you can use ''ssh -x remote'' to disable it for a single session if "ForwardX11 yes" is in your ''~/.ssh/config'' file.

In recent openSSH, there is also ''ssh -Y'' (or the option "ForwardX11Trusted yes") . This enables 'trusted' X11 forwarding.
This used to be the default with older versions, but because of a potential security vulnerability (which allowed other X11 clients to 'sniff' or alter data from the remote client) this was changed.
Most people won't need to bother with this option and should stick with the default.

Filesystem encryption

2005-02-06T20:52:26Z

Jor: dm-crypt

'''Filesystem encryption''' in Linux allows you to [[Cryptography|encrypt]] all of the data on a drive using a password or other key. This allows for the protection of sensitive data in case of theft of a computer or disk. The technique is very flexible; for example, the encryption key can be kept on a [[USB Flash drive]], with the contents protected by another passphrase.

==Cryptoloop==
The 'old'/deprecated method is Linux is called cryptoloop, which is similar to the normal [[loop]] [[mount]]ing method in Linux. This requires that the CryptoAPI and Cryptoloop options be enabled in your Linux [[kernel]] (module called 'cryptoloop', with separate modules for each [[cipher]] which should me automatically loaded by cryptoloop). See the [http://www.tldp.org/HOWTO/Disk-Encryption-HOWTO/index.html Disk Encryption HOWTO] for install instructions.

==dm-crypt==
The 'new' method was created because there were several security problems with cryptoloop and the code wasn't considered clean. The new method is based on the existing device-mapper code in the 2.6 kernel. It's available in the kernel since version 2.6.4. The kernel config options needed are device mapper support and dm-crypt. Both options can be found under 'Device drivers','Multi-device support (RAID and LVM)' in the various kernel config frontends. Setting up dm-crypt is a bit more difficult compared to cryptoloop. The program cryptsetup can help a bit with that[http://www.saout.de/misc/dm-crypt/]

== External links ==
*[http://clemens.endorphin.org/Cryptoloop_Migration_Guide Cryptoloop Migration Guide]

Mono

2005-02-06T20:32:40Z

Jor: see also

The '''Mono''' Project is an open development initiative sponsored by [[Novell]] that is working to develop an [[open source]], UNIX version of the [[Microsoft]] .NET development platform. Mono was originally begun by [[Miguel de Icaza|Miguel]] and the folks at [[Ximian]], and then Ximian was later purchased by Novell.

Mono includes a [[compiler]] for the C# language, a Common Language Runtime (CLR) for the Common Language Infrastructure (CLI) and a set of class libraries. The runtime can be embedded into your applications. It implements both ADO.NET and ASP.NET.

An IDE called MonoDevelop is available for Mono. It is a port of SharpDevelop. So far it is still in its early stages, but it is functional.

==Mono Project news==
On the 30th June 2004 Novell and the Mono project developer community released Mono version 1. For more information see the [http://www.mono-project.com/using/relnotes/1.0.html release notes].

==See also==
*[[DotGNU]]

==External links==
*[http://www.mono-project.com/about/index.html About Mono]
*[http://tools.devchannel.org/devtoolschannel/04/04/21/169234.shtml?tid=46 What is Mono and why should you care?]
*[http://www.mono-project.com/about/faq.html Mono FAQ]
*[http://go-mono.com/ Mono website]
*[http://www.monodevelop.com MonoDevelop website]
*[http://software.newsforge.com/article.pl?sid=04/07/15/2359201&tid=27 Why Mono is significant]

DotGNU

2005-02-06T20:32:05Z

Jor: see also

DotGNU is [[GNU]]'s own version of Microsoft's .NET

==See also==
*[[Mono]]

==External links==
*[http://www.dotgnu.org Official website]

Bluetooth

2005-02-03T14:37:07Z

Jor: see also

'''Bluetooth''' is a [[wireless]] network [[protocol]], designed for short range distances, low power and low cost. This makes it specially useful in small, mobile devices like mobile phones, [[PDA]]s and laptops, but is also used in many other types of devices. Also often said to be a protocol used to create a [[Personal Area Network]] (PAN).

It uses the ''open'' or ''license free'' 2.4Ghz radio band and does fast frequency hopping to avoid interference with other devices in that band. It was originally developed by [[Ericsson]], but later a Bluetooth consortium was formed together with other companies (including [[IBM]], [[Intel]] and [[Nokia]]).

Current specification standard is version 1.1, which specifies devices with a maximum range of 10 meters and devices with a range up to a 100 meters. The maximum data transfer rate is 1Mbps. A 2.0 version is in development and will feature higher speeds.

==Linux support==
The Linux [[kernel]] includes a bluetooth protocol [[stack]] (since 2.4) named [[bluez]]. It supports [[USB]], [[PCMCIA]] and [[serial]] bluetooth devices.
In most recent [[distribution]]s these devices are normally recognized automatically and just 'plug and play'. If not, the modules involved with the protocol are called:
*bluetooth -- the base
*l2cap -- named after the protocol L2CAP, which bluetooth uses (can be seen as layer 6/data-link layer from the [[OSI Model]]).
*rfcomm -- Emulates [[serial]] [[RS-232]] [[port]]s on top of L2CAP
Next to these there are several modules for different types bluetooth devices, like [[USB]], serial port or specific chipsets. (hci_usb, hci_uart, bcm203x, etc.)

After the hardware has been set up, it's time to actually use it. You need some software to scan and 'pair' devices. The [[Bluez]] project created several programs to do this (hcitool,hcidump,bluez-pin, etc.). Or you could try the Gnome Bluetooth Subsystem[http://usefulinc.com/software/gnome-bluetooth] or maybe whatever your own desktop/distribution provides.

==Applications==
*[[MultiSync]] -- Synchronize contacts,calender,etc
*[[bluexmms]] -- Control [[XMMS]] with your bluetooth phone [http://linuxbrit.co.uk/bluexmms/]
*[[bluemote]] -- Use your bluetooth phone as a remote control[http://www.geocities.com/saravkrish/progs/bluemote/index.html]
*[[btcid]] -- Phone status display on computer screen[http://0x63.nu/files/btcid/]

==See also==
*[[OpenOBEX]]
*[[Wireless]]

==External links==
*[http://www.bluetooth.com/ Official website]
*[http://www.bluez.org/ Bluez Linux bluetooth protocol stack]

{{stub}}

Bluetooth

2005-02-03T14:35:56Z

Jor: /* Applications */

'''Bluetooth''' is a [[wireless]] network [[protocol]], designed for short range distances, low power and low cost. This makes it specially useful in small, mobile devices like mobile phones, [[PDA]]s and laptops, but is also used in many other types of devices. Also often said to be a protocol used to create a [[Personal Area Network]] (PAN).

It uses the ''open'' or ''license free'' 2.4Ghz radio band and does fast frequency hopping to avoid interference with other devices in that band. It was originally developed by [[Ericsson]], but later a Bluetooth consortium was formed together with other companies (including [[IBM]], [[Intel]] and [[Nokia]]).

Current specification standard is version 1.1, which specifies devices with a maximum range of 10 meters and devices with a range up to a 100 meters. The maximum data transfer rate is 1Mbps. A 2.0 version is in development and will feature higher speeds.

==Linux support==
The Linux [[kernel]] includes a bluetooth protocol [[stack]] (since 2.4) named [[bluez]]. It supports [[USB]], [[PCMCIA]] and [[serial]] bluetooth devices.
In most recent [[distribution]]s these devices are normally recognized automatically and just 'plug and play'. If not, the modules involved with the protocol are called:
*bluetooth -- the base
*l2cap -- named after the protocol L2CAP, which bluetooth uses (can be seen as layer 6/data-link layer from the [[OSI Model]]).
*rfcomm -- Emulates [[serial]] [[RS-232]] [[port]]s on top of L2CAP
Next to these there are several modules for different types bluetooth devices, like [[USB]], serial port or specific chipsets. (hci_usb, hci_uart, bcm203x, etc.)

After the hardware has been set up, it's time to actually use it. You need some software to scan and 'pair' devices. The [[Bluez]] project created several programs to do this (hcitool,hcidump,bluez-pin, etc.). Or you could try the Gnome Bluetooth Subsystem[http://usefulinc.com/software/gnome-bluetooth] or maybe whatever your own desktop/distribution provides.

==Applications==
*[[MultiSync]] -- Synchronize contacts,calender,etc
*[[bluexmms]] -- Control [[XMMS]] with your bluetooth phone [http://linuxbrit.co.uk/bluexmms/]
*[[bluemote]] -- Use your bluetooth phone as a remote control[http://www.geocities.com/saravkrish/progs/bluemote/index.html]
*[[btcid]] -- Phone status display on computer screen[http://0x63.nu/files/btcid/]

==External links==
*[http://www.bluetooth.com/ Official website]
*[http://www.bluez.org/ Bluez Linux bluetooth protocol stack]

{{stub}}

Bluetooth

2005-02-03T14:26:43Z

Jor: bit more linux info

'''Bluetooth''' is a [[wireless]] network [[protocol]], designed for short range distances, low power and low cost. This makes it specially useful in small, mobile devices like mobile phones, [[PDA]]s and laptops, but is also used in many other types of devices. Also often said to be a protocol used to create a [[Personal Area Network]] (PAN).

It uses the ''open'' or ''license free'' 2.4Ghz radio band and does fast frequency hopping to avoid interference with other devices in that band. It was originally developed by [[Ericsson]], but later a Bluetooth consortium was formed together with other companies (including [[IBM]], [[Intel]] and [[Nokia]]).

Current specification standard is version 1.1, which specifies devices with a maximum range of 10 meters and devices with a range up to a 100 meters. The maximum data transfer rate is 1Mbps. A 2.0 version is in development and will feature higher speeds.

==Linux support==
The Linux [[kernel]] includes a bluetooth protocol [[stack]] (since 2.4) named [[bluez]]. It supports [[USB]], [[PCMCIA]] and [[serial]] bluetooth devices.
In most recent [[distribution]]s these devices are normally recognized automatically and just 'plug and play'. If not, the modules involved with the protocol are called:
*bluetooth -- the base
*l2cap -- named after the protocol L2CAP, which bluetooth uses (can be seen as layer 6/data-link layer from the [[OSI Model]]).
*rfcomm -- Emulates [[serial]] [[RS-232]] [[port]]s on top of L2CAP
Next to these there are several modules for different types bluetooth devices, like [[USB]], serial port or specific chipsets. (hci_usb, hci_uart, bcm203x, etc.)

After the hardware has been set up, it's time to actually use it. You need some software to scan and 'pair' devices. The [[Bluez]] project created several programs to do this (hcitool,hcidump,bluez-pin, etc.). Or you could try the Gnome Bluetooth Subsystem[http://usefulinc.com/software/gnome-bluetooth] or maybe whatever your own desktop/distribution provides.

==Applications==
*[[MultiSync]]

==External links==
*[http://www.bluetooth.com/ Official website]
*[http://www.bluez.org/ Bluez Linux bluetooth protocol stack]

{{stub}}

RPC

2005-01-30T16:20:32Z

Jor: init

'''RPC''' stands for '''R'''emote '''P'''rocedure '''C'''all. This term can be a bit confusing because it sometimes is used in different meanings. The first one as a general type of [[protocol]], the other common one is as the name for an actual implementation of such a protocol.

==Protocol==
As a type of protocol, it means a protocol that allows a program (client) to request a service from another program (server) which can be somewhere else on the network. The RPC protocol saves client the hassle of dealing with the networking them self and makes the [[procedure]] call look more like a normal local call.

==Implementation==
The second meaning is the as the name of an implementation, originally developed by [[SUN]], a.k.a ONC RPC or SUN RPC. This was later standardized by the [[IETF]].

The most common 'user' of this implementation is [[NFS]] (But also some other daemons like [[famd]]), but it's not used that often any more. It's been mostly superseded by other protocols and models like [[CORBA]], [[XML-RPC]]/[[SOAP]], etc.

===Linux===
On many Linux distributions it is installed by default.
This can be seen when a program called [[portmap]] is running. Access to portmap and the real services is usually controlled by tcp wrapper and configured in the hosts.allow and hosts.deny files in [[/etc]].
Not many desktop users use NFS or any of the other programs that need it, so it can often be uninstalled without much of a problem.

==See also==
*[[rpcinfo]]
*[[rpcbind]]

==External links==
*[http://www.ietf.org/rfc/rfc1831.txt IETF RFC RPC specification]

Rpcbind

2005-01-30T15:48:35Z

Jor: init

'''rcpbind''' is the name of the daemon for the [[RPC]] protocol that is used on [[SUN]] systems. It provides a mapping from a service name to the [[port]] number it's running on.

On Linux this service is usually provided by a program called [[portmapper]]

==See also==
*[[rpcinfo]]

==External links==
* [http://www.eits.uga.edu/wsg/sun/security/rpcbind.html What rpcbind does]

''This article is a [[LQWiki:stub_articles|stub]] and needs to be finished. [[LQWiki:plunging_forward|Plunge forward]] and [[LQWiki:How_to_edit_a_page|help it grow]] !''

Using multiple monitors

2005-01-10T21:25:37Z

Jor: typo + xorg

This is often called [[DualHead]] or [[MultiHead]].
All information descibed here only applies to [[XFree86]] 4.x.x versions, not the older 3.3.6 versions. It also applies to [[Xorg]] 6.8.x.

==Different methods==
There are a couple of different layouts possible:
#Clone mode -- Every monitors output the exact same thing.
#'Normal' MultiHead -- Every monitor has a different seperate X screen and session
#Combined mode -- The different monitors are combined into one big virtual screen. There are a couple different implementations for this:
##[[Xinerama]] -- The XFree86 extension
##Nvidia's [[TwinView]] -- What the Nvidia drivers call it
##ATI's [[Big Desktop]] -- What the ATI drivers call it
##[[MergedFB]] -- New method supported in some xfree86 drivers, currently mga,sis,ati drivers (These last two only in recent experimental/CVS versions)

==Requirements==
===Hardware===
Obviously multiple [[monitor]]s are needed.
For the [[video card]]s different methods are available:
#Using multiple seperate video cards
#Using a single multihead capable video card. This requires special support in the XFree86 driver.
#A combination of the two above.
Note: It's also possible to use the Voodoo1/2 3D-only add-on cards as an extra output, see the [[Glide]] driver manual page.

==Software==
An [[XFree86]] 4.x.x. version.
A window manager that supports your chosen multihead method.
Especially for Xinerama it's required that the window manager supports this extension (both [[KDE]] and [[Gnome]] support it).
For TwinView and MergedFB it's not needed, because these methods 'fake' one big desktop for the xserver itself.

==Configuration==
This section contains some quick info to configure the multihead setups. 
Some basic knowledge of the config file is needed for this. The [[XF86Config]] page should describe it.

===Multihead===
Each videocard should have a working 'Device' section. (Test this by using it as the single output).
Add a matching BusID option to each Device section so it knows which driver to use for each physical device.
The proper BusID can be found by using [[lspci]] or <tt>XFree86 -scanpci</tt>.
Example BusID line looks like this: 
BusID "PCI:1:0:0"
(Note that even [[AGP]] cards use the 'PCI:' prefix) 
For single videocards that support multihead it depends on the used driver. In most cases these cards have a second entry in lspci/scanpci.
Each head on the card should have a seperate Device section, each with the same 'Driver' option but different 'Identifier and'BusID' options.
A special option needs to be added for multihead configuration. Each Device sections needs a 'Screen' option with a number. For example: 
Screen "0"
This would indicate that this particular device section if for the primary head, the other device sections should have a similar option, but with the number increased. The other configuration options of the driver should be only used in the device section of the primary head.

Each monitor should have a working 'Monitor' section. (Also test these as single output).
Now create a 'Screen' section for every monitor-videocard combination.

After this it's time to set up a multihead 'Layout' section in the [[XF86Config]].
To get normal Multihead, just add Screen options to your normal Layout. Example: 
Screen 0 "Normal"
Screen 1 "TV"
Screen 2 "Fridge-LCD"
etc. 
Now if you start up X, you should have seperate X sessions running on each screen.

===Xinerama===
Almost the same as above. Only need to enable the Xinerama extension.
This can be done either by starting XFree86 with the '+xinerama' option, or adding 'Option "Xinerama" "On" to the 'ServerFlags' section of the [[XF86Config]]. 
And you need to specify the position of the screen relative to another. This can be done in the 'Layout' section.
Choices are: LeftOf,RightOf,Above,Below,Absolute X,Y, Relative X Y
Example: 
Screen 0 "Main"
Screen 1 "Secondary" LeftOf "Main"
Screen 2 "TV" relative "Main" 0 2000
etc.

===Nvidia TwinView===
If you are using an NVIDIA card, the README for the Linux drivers on the NVIDIA.com website have excellent instructions on how to set up Xinerama (currently Appendix I in the README). Also see [[Installing_NVIDIA_drivers]]

===ATI Big Desktop===
Easiest way is to let the fglrxconfig app generate a config file for you. Select 'Big Desktop' in the screen layout question.

===MergedFB===
To enable [[MergedFB]] one only has to add some options to the Device section of the multihead video card. Multiple 'Device','Screen' and 'Monitor' sections aren't needed. It's also not needed the change the 'Layout' section. Everthing is configured in the 'Device' section of the videocard. These options can normally be found in the manual page of the driver. And example piece out of a 'Device' section:
Option "MergedFB" "On"
Option "CRT2HSync" "30-70"
Option "CRT2VRefresh" "50-150"
Option "CRT2Position" "LeftOf"
These options can differ between card drivers, since this is a relatively new method and there isn't a standard option set for it yet.

==Links==
[[Xinerama for Radeon 9200]]

[[Multiple Monitors with Nvidia]]

#Maybe some links to working XF86Config files?

Cpio

2005-01-03T16:42:32Z

Jor: init

'''cpio''' ('''c'''o'''p'''y '''i'''nput and '''o'''utput) is both a command and an archive file format. The cpio command can create archives in several formats, including it's own binary format and the more common [[tar]] format.

It isn't often used in the 'wild'. The tar command and archive format is the most commonly used. But [[rpm]] packages contain a cpio archive. A cpio archive can be extracted from a rpm file with the [[rpm2cpio]] tool.

==See also==
*[[Packing and Unpacking Files]]
*[[ar]]

==External links==
*[http://www.gnu.org/software/cpio/cpio.html Gnu Cpio project page]

LQWiki talk:Ways to help the LinuxQuestions.org Wiki

2005-01-03T16:21:17Z

Jor: typo in page

There's a typo in this article, last words: 'maling list'. So if anybody reads this that has write access to this page... :) --[[User:Jor|Jor]] 11:21, Jan 3, 2005 (EST)

Tmpfs

2004-12-03T16:02:08Z

Jor:

'''tmpfs''' is a [[file system]] that stores all files in [[virtual memory]] (meaning it will also use the [[swap]] space).
It adjusts the actual memory size it uses to fit the files, so it doesn't used a fixed amount of memory. It allows to specify a maximum size with the <tt>size</tt> option when (re)[[mount]]ing. The default size is half the available [[RAM]].

It is mainly used to implement [[POSIX]] shared memory in Linux, by mounting a tmpfs file system on /dev/shm. It is also often mounted on /tmp, /var/tmp or some other personal tmp directory to increase the performance of these often used directories. It is also used by [[udev]] (which mounts a tmpfs on /dev/).

==Example usage==
mounting a tmpfs on /tmp with max size of 512 Megabytes and [[permission]]s of 777: 
mount -t tmpfs tmpfs /tmp -o size=512M,mode=777

Virtual memory

2004-11-25T16:33:48Z

Jor: duplicate info, redirect

#REDIRECT [[Swap]]

Grub version 1

2004-11-01T17:57:02Z

Jor: remove z's, so it's copy-pasteable again

'''GRUB''' (short for '''GR'''and '''U'''nified '''B'''oot Loader) is a [[boot loader]], similar in overall function to [[LILO]]. It is the standard way to boot into an [[operating system]] (or several) used by most recent [[Linux]] [[distributions]], but capable of being loaded on any Linux distro.

==Advocacy==
GRUB may have advantages over your current bootloader, but its your choice.
* You can change your boot menu entries by editing a text file. While LILO requires the text file to be amended and the need to re-run /sbin/lilo.
* You can "test" your boot menu entries in an interpreted environment. This may prevent some minor changes that renders the system unbootable.
* Grub boot floppies can use a kernel size of 1423 Kb and an initrd of 1423 Kb using 3 floppies to boot.

==GRUB boot menu==
The default location is in ''/boot/grub/menu.lst''. The .lst suffix stands for menu list, not menu first. The file contains configuration options (like color, timeout and password settings) and a list of menu entries. The file is often managed by some tool from the [[distribution]] itself, so when kernels are installed or removed the boot menu list gets automatically updated. Another common location is ''/etc/grub.conf'' but this is often a link file to the ''/boot/grub'' file.

===Manual configuration===
Full documentation of the format of the ''menu.lst'' file can be found in the GRUB [[info]] pages. But for a quick start, here are a couple of example entries to get you going.

Basic entry to boot a linux kernel located on the first disk:
<pre>
title Linux kernel 2.6.4
root (hd0,0)
kernel /boot/vmlinuz-2.6.4 root=/dev/hda1 ro
initrd /boot/initrd.img
</pre>
root (hd0,0) says /boot partition is on the first partition first drive or that the SUB-folder /boot to the / partition refers to same structure.

The "root" mentioned after vmlinuz refers to the / entry in your /etc/fstab file.

To boot windows:
<pre>
title Win
rootnoverify (hd0,0)
makeactive
chainloader +1

</pre>

===Debian===
[[Debian]] manages the menu file with ''update-grub'', this script checks which kernels are located in /boot/ and generates a menu entry for them. It adds it's own configuration options to the menu.lst file, these concern the options that are given to the generated menu entries. Usually update-grub is automatically run after a kernel packages has been installed or removed (at least it's the default configuration for the sarge/testing disribution). If not, update-grub can be added as a postinst_hook and a postrm_hook in the ''/etc/kernel-img.conf'' file.

===Red Hat===
[[Red Hat]] also manages the menu.lst file itself, it should be updated automatically when a kernel rpm is installed/removed

==Errors and solutions==
===Error 18===
Error 18: Selected cylinder exceeds maximum supported by BIOS 
This error is returned when a read is attempted at a linear block address beyond the end of the BIOS translated area. This generally happens if your disk is larger than the BIOS can handle (512MB for (E)IDE disks on older machines or larger than 8GB on others.).
In more practical terms this means the BIOS is unable to start executing the kernel because the kernel is not located within the block it can access at boot up time.

This can be circumvented by creating a boot partition at the beginning of the disk that is completely within the first 1023 cylinders of the harddrive. This partition will contain the kernel.
The kernel it self does not suffer from the same limitations as the BIOS so after the BIOS has loaded the kernel the kernel will have no problem accessing the whole harddrive.

Newer BIOSes will automatically translate the harddrives size in a way that it can be completely contained within the first 1023 cylinders and hence modern computers do not suffer from this problem.

==Tips==
For an awesome resource for troubleshooting just about any Grub problem (on any [[OS]]), see this thread on Gentoo's forums: [http://forums.gentoo.org/viewtopic.php?t=122656&highlight=grub+error+collection Gentoo Forums: Grub Error Collection]

===Booting Windows/DOS from a non-first harddisk===
This is a common problem of people dual-booting that want Linux to be on their first disk and Windows on some other one. [[Windows]] only wants to boot from partitions on first disk in the system. But Grub can trick Windows with the ''map'' command, an example entry would look like:
<pre>
title Win
rootnoverify (hd1,0)
hide (hd0,0)
map (hd0) (hd1)
map (hd1) (hd0)
makeactive
chainloader +1
</pre>

===Shutdown the machine from the menu===
Another menu entry that can be handy is one to shutdown the machine: 
<pre>
title Halt!
halt
</pre>

==See also==
*[[Securing GRUB]]
*[[MBR]]
*[[GRUB Howto and Trouble-shooter]]
*[[GRUB Floppies to boot large kernels]]

==External links==
*[http://www.gnu.org/software/grub/ Project website]

SMB

2004-10-27T15:12:53Z

Jor: formatting

'''S'''erver '''M'''essage '''B'''lock ('''SMB''') is the older name for the network protocols used in [[Windows]] networking for file and printer sharing, Windows messaging (not to be confused with [[Microsoft]]'s Instant Messaging software), and authentication.

Microsoft has officially renamed SMB to CIFS ("Common Internet File System"), but most Linux software and documentation still use the term SMB to refer to Windows networking.

Originally SMB did not use [[TCP/IP]] or the [[DNS]] system, instead using a proprietary transport system called [[NETBIOS]], and some adminstrators refer to Windows networking by that name. As of Windows 2000 all SMB is done over TCP/IP, but the older NETBIOS naming convention must still be used in some places, e.g.

\\server\share (Windows)

or

//server/share (Samba)

rather than:

smb://server.domain/share

==See Also==
*[[Samba]]
*[[Lisa]]

GParted

2004-10-27T15:09:27Z

Jor: init

'''GParted''' is a graphical [[partition]] editor. It's a front-end for the libparted library from [[parted]]. It uses the [[GTK]] GUI toolkit.

==External links==
*[http://gparted.sourceforge.net/ Project website]

QtParted

2004-10-27T15:04:52Z

Jor: init

'''QtParted''' is a graphical partition editor, similar to the well known 'Partition Magic' application in [[Windows]]. It is based on [[parted]] and uses the [[Qt]] toolkit.

==External links==
*[http://qtparted.sourceforge.net/ Project website]

Parted

2004-10-27T14:49:30Z

Jor: GUIs

'''[[GNU]] Parted''' is a [[partition]] editor. With parted you can create new partitions and change the size of existing partitions for a number of file systems (see this [http://www.gnu.org/software/parted/parted.html#features list].

Parted has been split in a [[CLI]] editor and a library to make it easier to create alternative interfaces. Several [[GUI]] font-ends have been made for this:
*[[QtParted]] -- Based on the [[Qt]] GUI libraries
*[[GParted]] -- Based on the [[GTK]] GUI libraries

==External links==
*[http://www.gnu.org/software/parted/parted.html Parted homepage]

{{stub}}

Udev

2004-10-27T14:42:28Z

Jor: update sysfs link

'''udev''' is a replacement for [[devfs]], meaning it manages the /dev directory and only creates device files for devices that actually exist in the system. It moves the creation and naming of the devices files to [[userspace]]. The main advantage of this is that device naming schemes can be easily changed and it doesn't depend on static minor/major numbers anymore. It also means devices can be moved around (e.g. different pci slot,adding hard disks,etc) without their name changing. There are different naming schemes available, like one replicating the devfs names or one using the traditional (flat directory structure) names.

Udev been added in the Linux [[2.6]] [[kernel]] versions, and devfs has been marked obsolete.

The only thing that devfs does and udev not, is loading kernel modules when a device is being opened. This isn't seen as a needed feature for udev, since loading of kernel modules is handled automatically by [[hotplug]]. But this does mean that all the modules for all the existing devices are always loaded. If for some weird reason (for example, you want save some memory used by the modules) you don't want that, you'll have to use devfs or some manual script to handle that.

It usually mounts a [[tmpfs]] over the /dev directory, and [[hotplug]] is used to receive notification whenever a device is added/removed. The [[sysfs file system]] is used to get the information about the devices.

==Configuration==
All the configuration files are normally located in /etc/udev.
The different naming policy are configured in files ending with .rules. The udev.conf is the main config file, where the used rules and permissions can be configured.

==External links==
*[http://www.kroah.com/linux/talks/ols_2003_udev_talk/ Slideshow] from a talk given about udev
*[http://archive.linuxsymposium.org/ols2003/Proceedings/All-Reprints/Reprint-Kroah-Hartman-OLS2003.pdf Paper] about udev.

{{stub}}

Udev

2004-10-27T14:40:46Z

Jor: ext links

'''udev''' is a replacement for [[devfs]], meaning it manages the /dev directory and only creates device files for devices that actually exist in the system. It moves the creation and naming of the devices files to [[userspace]]. The main advantage of this is that device naming schemes can be easily changed and it doesn't depend on static minor/major numbers anymore. It also means devices can be moved around (e.g. different pci slot,adding hard disks,etc) without their name changing. There are different naming schemes available, like one replicating the devfs names or one using the traditional (flat directory structure) names.

Udev been added in the Linux [[2.6]] [[kernel]] versions, and devfs has been marked obsolete.

The only thing that devfs does and udev not, is loading kernel modules when a device is being opened. This isn't seen as a needed feature for udev, since loading of kernel modules is handled automatically by [[hotplug]]. But this does mean that all the modules for all the existing devices are always loaded. If for some weird reason (for example, you want save some memory used by the modules) you don't want that, you'll have to use devfs or some manual script to handle that.

It usually mounts a [[tmpfs]] over the /dev directory, and [[hotplug]] is used to receive notification whenever a device is added/removed. [[Sysfs]] is used to get the information about the devices.

==Configuration==
All the configuration files are normally located in /etc/udev.
The different naming policy are configured in files ending with .rules. The udev.conf is the main config file, where the used rules and permissions can be configured.

==External links==
*[http://www.kroah.com/linux/talks/ols_2003_udev_talk/ Slideshow] from a talk given about udev
*[http://archive.linuxsymposium.org/ols2003/Proceedings/All-Reprints/Reprint-Kroah-Hartman-OLS2003.pdf Paper] about udev.

{{stub}}

Sysfs file system

2004-10-27T14:38:01Z

Jor: init

'''Sysfs''' is a memory based [[file system]] for the linux [[kernel]] version 2.6 and up. It provides a means of exporting certain attributes and information (internal kernel data structures) to [[userspace]] as normal [[file]]s/[[symlink]]s in [[directory|directories]]. It is meant to eventually replace the attributes that are in [[procfs]] in previous kernel versions. It shows several directory hierarchies showing the available hardware devices and modules/drivers attributes. It uses a 'one attribute per file' policy, which makes the attributes easier to read, parse and optionally alter.
For kernel (module) programmers, it provides a cleaner and easier to use interface to create these files.

[[udev]] uses sysfs to get the information it needs about the hardware (and [[hotplug]] for the initial creation trigger).

There is also a tool set and [[library]], which provides easy access to the information in sysfs for application programmers [http://linux-diag.sourceforge.net/Sysfsutils.html].

Sysfs system call

2004-10-27T14:19:47Z

Jor: moved from sysfs

'''Sysfs''' is a Linux [[system call]] providing information about the various types of [[file system]]s supported by the currently running [[kernel]].

==External Links==
* http://linux.about.com/library/cmd/blcmdl2_sysfs.htm man page

Sysfs

2004-10-27T14:19:08Z

Jor: disambig

*[[sysfs file system]]
*[[sysfs system call]]

{{msg:disambig}}

Udev

2004-10-27T14:09:35Z

Jor: very tiny, minor tweak

'''udev''' is a replacement for [[devfs]], meaning it manages the /dev directory and only creates device files for devices that actually exist in the system. It moves the creation and naming of the devices files to [[userspace]]. The main advantage of this is that device naming schemes can be easily changed and it doesn't depend on static minor/major numbers anymore. It also means devices can be moved around (e.g. different pci slot,adding hard disks,etc) without their name changing. There are different naming schemes available, like one replicating the devfs names or one using the traditional (flat directory structure) names.

Udev been added in the Linux [[2.6]] [[kernel]] versions, and devfs has been marked obsolete.

The only thing that devfs does and udev not, is loading kernel modules when a device is being opened. This isn't seen as a needed feature for udev, since loading of kernel modules is handled automatically by [[hotplug]]. But this does mean that all the modules for all the existing devices are always loaded. If for some weird reason (for example, you want save some memory used by the modules) you don't want that, you'll have to use devfs or some manual script to handle that.

It usually mounts a [[tmpfs]] over the /dev directory, and [[hotplug]] is used to receive notification whenever a device is added/removed. [[Sysfs]] is used to get the information about the devices.

==Configuration==
All the configuration files are normally located in /etc/udev.
The different naming policy are configured in files ending with .rules. The udev.conf is the main config file, where the used rules and permissions can be configured.

{{stub}}