The purpose of this short document is to introduce the reader to what a *nix shell is, and how to maneuver it. We shall be concise and get you using the shell in no time.
Please don't fret and worry, the shell looks intimidating at first but within a few minutes you'll start to get the hang of it. Just remember that it's super important to remember that by using a shell we are communicating more directly with our operating system (OS). Userland, usually maneuvered through a GUI (graphical user interface, pronounced 'goo-ee'), is comprised of programs for interacting with your OS - think games, browsers, notification apps, a GUI file system, etc. - but by using a shell you are provided with a new suite of programs for interacting with your computer one level deeper than GUI Userland.
If you want to accomplish more, stay out of GUI Userland. It's not comfortable there.
- shell - a suite of programs for interacting with your OS, and one level deeper than a GUI Userland. Repeat after me: What you can do in the GUI Userland you can do in the shell! (even browse the 'net!)
- userland - where a computer user exists
- operating system - the core programs that make a computer 'work'. Most people refer to a 'computer' as what they are interacting through, but that is not the case. This is known in geek speak as Userland.
- programs - a set of 'things' that take input and give output - It's the way you speak to your computer (browsers, file systems, games, etc.)
- GUI - Graphical User Interface, that is, a way of interacting with a computer through programs that are more visually appealing than the shell
Before we start we need to clarify one more time what a shell is, and then tell you about commands. A shell is the way to communicate more directly with your computer. It is important to note here that the 'shell' is typically used only when talking about a *nix operating system - that is, a system that is either Unix itself, or mimicks the behaviour of a Unix system. For more info on Unix click here.
You may have heard of the Linux operating system - this is actually a group of projects which sought to replace the Unix OS with a 'free' version that anyone in the world can use. For more on this philosophy you can click here.
So the shell - how does it work? Well, it is actually comprised of small programs that have very niche purposes. Usually these programs are refered to as 'commands', and by issuing commands to your shell you are interacting with your operating system. That's it!
- Your shell uses small programs known as commands which allow you to accomplish what you can in the GUI Userland
- Each command is a small program that accomplishes only one task. Thus, it is important to learn a number of commands to be able to successfully use the shell. That's where this guide comes in!
Hopefully by now you've at least a vague understanding of what I mean when I say 'shell', and how it differs from the typical method of interacting with a computer. Now it's time to learn some commands.
The commands we are going to learn are:
- cd
- ls
- pwd
- nano
- rm
- touch
- mkdir
- man
When you use the shell, you interact with it using something called a terminal. What's a terminal? It's the cool thing you see people in The Matrix and other hacking movies use; you are prompted to input some commands, and the computer does what you want!
The commands listed above will get you up-to-speed so that you can start maneuvering your server with ease; trust me, once you start you can become addicted! Let's jump in, shall we. It'd be best to have your terminal up and running.
As I demonstrate commands for you to try, I'll be using a syntax like so: $___. This means anything after the > I want you to input yourself into your terminal. :)
- pwd
- print working directory
Most of you are familiar that in GUI Userland, you have a file system where you have your Documents, Pictures, Music, etc. folders. These folders are commonly called directories. Well, in the terminal, if you want to know where you are among your folders, you can $ pwd. Give it a go now.
Eg,
$ pwd
$ /home/noah
This means that I am in the subfolder of the folder /home.
But what about looking at what's currently in your directory?
- ls
- list everything in your current directory
So, if I $ ls, I see that in my current folder I have: Mail public_html sent. These can be a mixture of directories and individual files. Take some time now messing about with these two commands and start picturing in your head that you're doing the same thing that's possible in GUI Userland.
But what if I want to move through these folders?
- cd
- change directory
You can $ cd into other directories. You do this by typing the cd command and then an argument, or the value that the program/command will evaluate and manipulate for you. So by doing cd <destination>, you move into that destination!
$ cd public_html
$ ls
$ index.html notes.txt
If I changed into my public_html folder, and then ls in that directory, I see that the shell returns 2 files for me; index.html and notes.txt.
Now, what if I want to move back a directory? You simply cd ... .. Is the syntax that the terminal understand for meaning move back one directory.
$ cd ..
$ ls
$ Mail public_html sent
Are you getting the hang of this now?
So let's say you want to cd back multiple directories; simply add / in between the number of directories you want to change. Let's break this down. cd ../.. means you want to cd back two directories, and cd ../../../.. means you want to cd back four. But what if you want to move forward 4 directories?
cd /path/to/your/destination.
So this means that you simply have to append / in between the destination arguments that you are giving to cd.
Lets say I have a directory structure as so:
/home/noah/
/ | \
directory1 directory2 directory3
/ | \
pics music pdfs
/ \
Sinatra metal
Let's print our working directory. $ pwd. We get /home/noah/. Perfect. Let's just ls real fast and see what we're working with. $ ls.
directory1 directory2 directory3. Brilliant.
If I'm looking for all of my Sinatra tunes (ooooo', Franky!), then instead of cding a bunch of individual times, I can
$ cd directory1/music/Sinatra.
Let's pwd, and we get /home/noah/directory1/music/Sinatra/. Great!
$ ls
You see that blank stuff? It means I've got no files or folders. :( I forgot to upload my Sinatra mp3's to the server!!
Now let's just go back to /home/noah/.
$ cd ../../..
$ pwd
$ /home/naltun
You're doing great!
Now let's say that I wanted to create a directory and add some files to it; it's straight-forward!
$ touch <file>
$ mkdir <direcotry>
Now, you use the touch command to create a file. And, the cool thing is, you can touch as many files as you want in one command; simply append as many arguments onto the touch command.
Let's say I'm writing my autobiography and I want to create a few chapters as separate files:
$ touch ch.1 ch.2 ch.3
Now let's say that I want to add these files to a directory. Before we do that, we should create the directory to hold all of these chapters. For this, we can use to mkdir command. It follows the same conventions as touch.
$ mkdir <directory_1> <directory_2>
Now, let's create that directory...
$ mkdir life_of_noah
Now, you can make yourtouch and mkdir arguments a path to the directory you'd like to touch or mkdir. First, let's make a new directory with mkdir in my /public_html folder.
$ mkdir /home/noah/public_html/scripts
Now let's touch a file in the new /scripts folder.
$ touch /home/noah/public_html/scripts/date.js
Let's pwd and see where we are.
$ pwd
$ /home/naltun
Hmm. I decided that I want to remove that file that I touch'd in the /scripts directory. But how do I do that?
- rm
- remove a file or directory
The rm command follows the same convention as other shell programs/commands; $ command <argument1> <argument2> <etc...>
So let's rm that file.
$ rm /public_html/scripts/date.js
Okay, great! Let's just double check that my date.js has been removed.
$ ls /public_html/scripts
Perfect, we don't receive any output from our command! But I also decided that I want to scrap my /scripts directory altogether. How would I do that? Well, many commands have pre-arguments called flag. A flag tells the command to manipulate the command or the arguments of the command in a very particular manner. Let's see what happens when we try to rm the /scripts directory.
$ cd /public_html
$ rm /scripts
rm: cannot remove 'scripts/': Is a directory
What is this telling us? It's telling us that rm can't be used with a directory argument. But I promise you that rm CAN remove directories! How can I prove it?
- man
- many commands have documentation called manual pages; use man to read the manual for a command
$ man rm
Now using the man command can give you a lot of information at once that can make you stressed; don't worry! The more you man the easier it gets to make sense of the information. Trust me, it's not that difficult; just take it slow and you'll understand what you're reading in no time! So, back to our terminal session.
$ man rm
...
-r, -R, --recursive
remove directories and their contents recursively
...
Okay, cool, so the manual for rm tells us that we can use -r, -R, or --recursive as flags for our rm command to remove a directory. Flags come before the argument to a command. So, any idea how we'd remove our directory?
$ pwd
/home/noah/public_html
$ ls
index.html /scripts
$ rm -r /scripts
$ ls
index.html
YES! We DID IT! Or, YOU did it!
If you've worked through this little tutorial, you should now realize that you've created and removed files and directories, as well as maneuvered all throughout your files and directories. Wicked cool! But there's one last thing to learn now to wrap it all up...
- nano
- a text editor for writing to a document; think of Notepad or Microsoft Word, but in the terminal