Difference between revisions of "Help:Reading"

From ArchWiki
Jump to: navigation, search
(Official packages: rm self-link (formatting of installation sentences is described in Help:Style)
(Installation of packages: update as agreed in Help_talk:Style#Pkg_and_AUR_templates:_add_icon.2C_make_them_look_different)
Line 57: Line 57:
 
For packages from the [[official repositories]] you will read something like:
 
For packages from the [[official repositories]] you will read something like:
  
:Install {{Pkg|package}}.
+
:Install the {{Pkg|foobar}} package.
  
 
This means that you have to run:
 
This means that you have to run:
  
  # pacman -S package
+
  # pacman -S foobar
  
 
The [[pacman]] article contains detailed explanations to deal with package management in Arch Linux proficiently.
 
The [[pacman]] article contains detailed explanations to deal with package management in Arch Linux proficiently.
Line 69: Line 69:
 
For packages from the [[Arch User Repository]] you will read something like:
 
For packages from the [[Arch User Repository]] you will read something like:
  
:Install {{AUR|package}} from the [[AUR]].
+
:Install the {{AUR|foobar}} package.
  
This means that in general you have to follow the {{AUR|package}} link, download the PKGBUILD archive, extract it, '''verify the content''' and finally run, in the same folder:
+
This means that in general you have to follow the {{AUR|foobar}} link, download the PKGBUILD archive, extract it, '''verify the content''' and finally run, in the same folder:
  
 
  $ makepkg -si
 
  $ makepkg -si
  
The [[AUR]] link will lead you to the [[Arch User Repository]] article, which contains all the detailed explanations and best practices to deal with AUR packages.
+
The [[Arch User Repository]] article contains all the detailed explanations and best practices to deal with AUR packages.
  
 
== Control of systemd units ==
 
== Control of systemd units ==

Revision as of 02:41, 24 May 2015

Because the vast majority of the ArchWiki contains indications that may need clarification for users new to GNU/Linux, this rundown of basic procedures was written both to avoid confusion in the assimilation of the articles and to deter repetition in the content itself.

Regular user or root

Some lines are written like so:

# mkinitcpio -p linux

Others have a different prefix:

$ makepkg -s

The numeral or hash sign (#) indicates that the line is to be entered as root, whereas the dollar sign ($) shows that the line is to be entered as a regular user.

Note: The commands prefixed with # are intended to be executed from a root shell, which can for example be easily accessed with sudo -i. Running sudo command from an unprivileged shell instead of command from a root shell will also work in most cases, with some notable exceptions such as redirection and substitution, which strictly require a root shell. See also sudo.

A notable exception to watch out for:

# This alias makes ls colorize the listing
alias ls='ls --color=auto'

In this example, the context surrounding the numeral sign communicates that this is not to be run as a command; it should be edited into a file instead. So in this case, the numeral sign denotes a comment. A comment can be explanatory text that will not be interpreted by the associated program. Bash scripts denotation for comments happens to coincide with the root PS1.

After further examination, "give away" signs include the uppercase character following the # sign. Usually, Unix commands are not written this way and most of the time they are short abbreviations instead of full-blown English words (e.g., Copy becomes cp).

Regardless, most articles make this easy to discern by notifying the reader:

Append to ~/path/to/file:

# This alias makes ls colorize the listing
alias ls='ls --color=auto

Append, create, edit and source

When prompted to append, add, create or edit, consider it an indication for using a text editor, such as nano, in order to make changes to configuration file(s):

# nano /etc/bash.bashrc

In programs, be it shells or otherwise, sourcing applies settings specified in a file. For Bash, sourcing can be done in a command prompt:

$ source /etc/bash.bashrc

and it can also happen in a file itself:

# This line includes settings from another file
source /etc/bash.bashrc

As a result, sourcing a file after alteration is an implied omission in the case of shell files.

However, not all articles will specify the nature of the changes to be made, nor which file to alter in the first place. This wiki builds on previous knowledge, such as common locations for files that are prone to sporadic editing.

Installation of packages

When an article invites to install some packages in the conventional way, it will not indicate the detailed instructions to do so, but instead it will simply mention the names of the needed packages with a link to the article that explains the generic installation procedure.

Official packages

For packages from the official repositories you will read something like:

Install the foobar package.

This means that you have to run:

# pacman -S foobar

The pacman article contains detailed explanations to deal with package management in Arch Linux proficiently.

Arch User Repository

For packages from the Arch User Repository you will read something like:

Install the foobarAUR package.

This means that in general you have to follow the foobarAUR link, download the PKGBUILD archive, extract it, verify the content and finally run, in the same folder:

$ makepkg -si

The Arch User Repository article contains all the detailed explanations and best practices to deal with AUR packages.

Control of systemd units

When an article invites to start, enable, stop or restart some systemd units (e.g. a service), it will not indicate the detailed instructions to do so, but instead you will read something like:

Start example.service.

This means that you have to run:

# systemctl start example.service

The Start link will lead you to the systemd article, which contains all the detailed explanations to deal with systemd units in Arch Linux proficiently.

System-wide versus user-specific configuration

It is important to remember that there are two different kinds of configurations on a GNU/Linux system. System-wide configuration affects all users. Since system-wide settings are generally located in the /etc directory, root privileges are required in order to alter them. For example, to apply a Bash setting that affects all users, /etc/bash.bashrc should be modified.

User-specific configuration affects only a single user. Dotfiles are used for user-specific configuration. For example, the file ~/.bashrc is the user-specific configuration file. The idea is that each user can define their own settings, such as aliases, functions and other interactive features like the prompt, without affecting other users' preferences.

Note: ~/ and $HOME are shortcuts for the user's home directory, usually /home/username/.

Common shell files

For ease of use, here is a selective listing of basic configuration files and their locations.

Bash

See also: Bash and man bash

Within Bash and other Bourne-compatible shells, such as Zsh, there is even further differentiation in the purposes of the configuration files. Some files only get sourced when Bash is starting as a login shell, whereas other files only do so when Bash is an interactive shell.

When Bash is run in a virtual console, for instance, it is started as a login shell. Bash shells started in a Xorg session, such as those employed by xterm, are interactive shells.

Common files:

  • /etc/bash.bashrc: System-wide settings; sourced only by a login shell
  • ~/.bashrc: Personal shell settings; sourced only by an interactive shell

Zsh

See also: Zsh and man zsh

Common files:

  • /etc/zsh/zprofile: System-wide settings; sourced only by a login shell
  • ~/.zshrc: Personal shell settings; sourced only by an interactive shell

Pseudo-variables in code examples

Some code blocks may contain so-called pseudo-variables, which, as the name says, are not actual variables used in the code. Instead they are generic placeholders and have to be manually replaced with system-specific configuration items before the code may be run or parsed. In the articles that comply with Help:Style/Formatting and punctuation, pseudo-variables are formatted in italics.

For example:

  • Enable the dhcpcd@interface_name.service for the network interface identified from the output of the ip link command.

In this case interface_name is used as a pseudo-variable placeholder in a systemd template unit. All systemd template units, identifiable by the @ sign, require a system-specific configuration item as argument. See Systemd#Using units.

  • The command dd if=data_source of=/dev/sd"X" bs=sector_size count=sector_number seek=partitions_start_sector can be run as root to wipe a partition with the specific parameters.

In this case the pseudo-variables are used to describe the parameters that must be substituted for them. Details on how to gather them are elaborated on in the section Securely wipe disk#Calculate blocks to wipe manually, which features the command.

Tango-view-fullscreen.pngThis article or section needs expansion.Tango-view-fullscreen.png

Reason: Mention other examples, ideally from other device categories (e.g. storage), with links to background articles. The examples are meant to avoid duplicating existing explanations in other articles. (Discuss in Help talk:Reading#)

In case of file examples, pasting pseudo-variables in real configuration files might break the programs that use them.

Ellipses

In most cases ellipses (...) are not part of the actual file content or code output, and instead represent omitted or optional text that is not relevant for the discussed subject.

For example HOOKS="... encrypt ... filesystems ..." or:

/etc/X11/xorg.conf.d/50-synaptics.conf
Section "InputClass"
    ...
    Option      "CircularScrolling"          "on"
    Option      "CircScrollTrigger"          "0"
    ...
EndSection

Be aware though that, in a few instances, ellipses may be a meaningful part of the code syntax: attentive users will be able to easily recognize these cases by the context.