In this tutorial, we’ll compare text-mode web browsers in Linux, while going over a brief overview of each one.
2. Introduction to Text-Mode Web Browsers
In this section, we’ll compare different text-mode web browsers and then go over some important installation and configuration information.
2.1. Comparing Features
Different text-mode web browsers come with various features. Some of these features may negatively affect performance or security. This is important to consider when determining which browser to use:
|Lynx||https://lynx.invisible-island.net/current/||No||No||No||Current Session Only||Yes|
If a browser supports inline image display, we can view images within the webpage. Otherwise, image links must be viewed using an external program.
2.2. Installing From Source
Since many package managers don’t have an up-to-date version of the browsers covered, it’s usually best to install from source. Generally, most packages use autotools for installation.
The first step is configuring our project. Let’s begin by listing the various configuration options:
$ ./configure --help Usage: configure [options] [host] Options: [defaults in brackets after descriptions] ...
In this example, we get a list of all available options by running ./configure –help. After determining what options we need, we can install our package:
$ ./configure --prefix="$HOME/.local/" $ make $ make install
First, we execute the configure script with the options we want to use. Here, we only use the –prefix option which sets the install location for our package. The XDG Base Directory Specification recommends that user-specific packages be installed in the $HOME/.local directory.
If we don’t specify an install location, the package typically installs globally in the /usr/local directory.
Finally, we can just run the make and make install commands to compile and install our package.
2.3. Displaying Graphics in TTY
Some text-mode browsers support graphics in the TTY through the use of the framebuffer. However, before we are able to write to the framebuffer, we must ensure our user has the correct read and write privileges:
$ usermod -a -G video "$USER"
Here we use the usermod command to add the current user to the video group, the GID that has permissions over /dev/fb* (Frame Buffer Device). This gives the user permission to read and write to all of the framebuffer device files. After running this, we should be able to use supported browsers to view images within the TTY.
2.4. Setting the Browser Home Page
To set the default home page for our text-mode browsers, we can edit our $HOME/.bashrc file:
$ export WWW_HOME="https://www.baeldung.com"
To declare the environment variable $WWW_HOME, we can use the export command. This sets the home page for all browsers mentioned and typically takes precedence over values set in their respective configuration file.
Lynx is one of the oldest text-mode web browsers still being actively maintained. It’s simple and fast, but lacking in all but the most basic features.
We can begin by downloading and extracting the latest version. After this, we can proceed with the installation of Lynx:
$ ./configure --with-ssl --prefix="$HOME/.local/" $ make $ make install
The default configuration should be fine. Importantly, we just need to run ./configure with the –with-ssl option if we want to use HTTPS.
3.2. Getting Started
To start the Lynx browser, we can run the lynx command in our terminal:
Running lynx by itself will open the browser to the home page if set. If we run lynx with a URL as an argument, it opens up into that instead.
Since Lynx is best used with the keyboard, it’s important we know some basic keybindings:
|Enter, Left Mouse Button||Go to the link under the cursor|
|/, n, N||Search within the document, next occurrence, previous occurrence|
|g, G||Go to URL, edit the current URL|
|o, Middle Mouse Button||Open option menu, open navigation menu|
|Up Arrow, Down Arrow||Next link, previous link|
|Left Arrow, Right Arrow||Back button, forward button|
|k, K||List keybindings|
|q, Q||Quit browser, quit browser without prompt|
|Backspace||Show current session history|
By default, the lynx command won’t allow mouse input, but it can be enabled:
$ lynx -use_mouse -scrollbar
Here we use the -use_mouse option to enable mouse input with lynx. We also enable a mouse controllable scrollbar using -scrollbar.
The default config file for Lynx is /etc/lynx.cfg. Listed below are some basic config settings we may want to adjust:
|MAKE_LINKS_FOR_ALL_IMAGES||If set to TRUE, will not use the ALT value for images and instead provide a web link. We should set this to TRUE if we want to be able to view images in an external program.|
|XLOADIMAGE_COMMAND||Defines the external program to open image links with. If we want to open our images using xdg-open, we can include this line, but keep it blank.|
|SET_COOKIES||We can set this to prevent being asked to accept or deny cookies every time we visit a website. Set to FALSE to automatically deny cookies, and TRUE to automatically accept them.|
There are also some settings saved to the ~/.lynxrc file. While running Lynx we can modify these settings using the options menu.
Settings changed through the options menu are temporary, but some of them are able to be made persistent by selecting Save options to disk. These options will overwrite all of the previous values in our ~/.lynxrc file.
Links is an early text-mode browser with support for many features including displaying graphics in terminal emulators and the TTY.
The INSTALL file includes instructions on installation including a list of libraries that are either required or optional. To begin, we should install any libraries we need like graphics drivers, and run the ldconfig command.
After getting the libraries we need, we can begin the installation process:
$ ./configure --enable-graphics --prefix="$HOME/.local/" ... Configuration results: ... $ make $ make install
Overall, most settings used by configure shouldn’t have to be modified, with the exception of –enable-graphics for graphical mode support.
At the end of the output will be a table listing our configured settings. Once we verify these settings are correct, we can make and install our program.
4.2. Getting Started
To start the browser, we can run the links command:
Running links by itself opens up the browser in text mode. To start links with graphics enabled, we can use the -g flag:
$ links -g
Using the -g flag with links will open up the browser using the first graphics library found.
To specify which driver we want links to use, we should use the -driver option:
$ links -driver 'fb'
Here we run links in graphics mode using -driver to ensure the use of the fb (framebuffer) driver.
Configuring links works from the browser menu. To open it, we can press the <esc> key. While there, we can navigate to the View or Setup tab.
Saved options are located either in .html.cfg for HTML options or in .links.cfg for everything else. Both of these files are typically located in the ~/.links directory.
After downloading the source code and any necessary libraries, we can begin installing:
After that, we just compile and install our program with make.
5.2. Getting Started
To get started, we can simply run the elinks command in our terminal:
Most keybindings are the same between ELinks and the previous browsers we discussed, but there are some important exceptions:
|t, c||Open Tab, Close Tab|
|<, >||Previous Tab, Next Tab|
|e, o||Open Tab Menu, Open Option Menu|
|h||Open History Manager|
It’s also possible to interact with a remote instance of elinks:
$ elinks -remote "www.baeldung.com" "www.baeldung.com/linux"
Here we use -remote to open a series of URLs in new tabs on an already running elinks session.
The settings of ELinks are in the ~/.elinks/elinks.conf configuration file. This file can be edited either directly or through the browser’s options menu.
Let’s look at some settings we may want to adjust:
|ui.sessions.homepage||Set default home page. If Overrides the environment variable $WWW_HOME.|
|connection.ssl.https_by_default||Makes HTTPS the default protocol used when one is not specified.|
To initialize the elinks.conf file, we can set Configuration system -> Saving style to 1 in the options menu and save. This adds any settings not already inside elinks.conf to the end of the file, and adds a comment summarizing what each setting does.
6.1. Getting Started
We can open up the browser using w3m by either passing it a URL or using the -v flag:
$ w3m -v
Using -v with w3m opens up the browser to a basic welcome screen.
Navigation in W3m uses the arrow keys or hjkl to move the cursor. To go to a link, we can hover over it and press <return>. Here are a few other important keybindings:
|o, m||Open Options Menu, Open Navigation Menu|
|<tab>, <shift><tab>||Next link, previous link|
|H||Show Current Keybindings|
|U||Go to URL|
|I||Open Image in External Program|
|Ctrl–H||Show History Page|
In addition to being able to browse webpages, w3m can also be used as a pager:
$ w3m <<< "$(w3m -show-option)" ... -o clear_buffer= Free memory of undisplayed buffers -o decode_cte= Decode Content-Transfer-Encoding when saving ... ≪ ↑ ↓ <*stream*>
In this example, we use w3m to view the output of w3m -show-option.
Since w3m is a pager we’re able to view text just as we would with a command like less. We can do this by passing a file to w3m or by passing a stream of data to be treated as a file, using a here string.
Config files are located inside ~/.w3m/. However, in most situations, it isn’t necessary to edit the configuration files directly. Instead, it’s preferred to use the options menu.
If we want to use Lynx-like keybindings, we can copy the provided keymap.lynx file to our ~/.w3m/ directory:
$ mkdir -p ~/.w3m $ cp /usr/share/doc/w3m/keymap.lynx ~/.w3m/keymap
Likewise, if we want to revert back to the default keybinding, we can simply copy back the keymap.default file.
6.3. Image Display
To start w3m with inline image display enabled, we can use the -o flag:
$ w3m -o display_image=TRUE "https://baeldung.com/linux/"
Here, we run w3m with the -o flag to assign a value to a configuration setting, in this case, display_image.
The default image display method uses w3mimgdisplay, but it also supports other methods.
6.4. Using W3m in Emacs
A unique feature of W3m is its ability to run inside Emacs. To do this, we’ll have to install emacs-w3m. It’s important that the install location’s set to the ../emacs/site-lisp directory. After that, we’ll just need to edit our initialization file to load W3m:
After adding this line to our init file, we can start w3m in Emacs by pressing M-x w3m.
In this article, we compared four different text-mode web browsers and then walked through a brief overview of each one.