There are many Unixes out there. The Sun workstations in our labs run Opensolaris.
When you log on at one of the workstations,
you are presented with a (pretty much) blank desktop.
This is the
Gnome Desktop. You can do a lot by clicking on
things like you would with Microsoft Windows, but you really
need a shell terminal (command line) to do real work.
Click on the Terminal icon
at the top of the screen.
This pops up a shell terminal (command line) for you to enter commands in.
Many people think of command-line interfaces to computers as primitive. Get over it. Some tasks are much easier to perform with a command line. GUIs are good enough for the average computer user, but none of you are average computer users. Once you've learned to use the shell right, you'll find it's very flexible, powerful and fast. You might not want to go back to that GUI world!
pwd (print working directory)
tells you the absolute path of the directory you're currently "in". So if
pwd gives you
/usr/includeit means that you're "in" the
include directory,
which is contained in the usr directory, which is
contained in the /, which as we talked about last
class, is the
starting point of the file system.
The directory you'll always start off "in" when you
log on is your home directory. If you execute pwd, you should see that you're "in"
the directory /home/m12xxxx, which is
your home diretory. The way your shell is set up, the
current directory is listed in front of the prompt, but
instead of listing /home/m12xxxx/
it simply lists the ~/ shortcut we talked about
last class.
Remember, in your home directory, you have the right to create, delete
and move whatever files or directories you want to.
To get started on the lab, you're going to copy a file from
my home directory to yours, then for the rest of the lab,
we'll be playing around with that file. That brings us to
our second command: cp
cp - Copy. cp takes two
arguments, the first is the source file, and the second is
the destination file. You can either provide an absolute
path, or a relative path in either argument. For instance,
the following command will copy the file passwd
from the /etc directory into my home directory,
regardless of what my working directory is:
cp /etc/passwd ~Note that this would give the same result:
cp /etc/passwd /home/stahlIf, on the other hand, I'm in my home directory, I can take advantage of
. cp /etc/passwd .Oh, and this too:
cp ../../etc/passwd .We want to copy the file
unix_practice.tar.gz,
located in directory:
~wcbrown/courses/IC221/labs/L01, to
your home directory. Hmm, that seems like a lot of typing.
Now might be a good time to show you some cool shell shortcuts:
- Tab completion: Just because we typically use a command
line interface with Unix doesn't mean that we like to
type! Try this: type in
cp ~wcon the command line, then press the TAB key. See my full user name show up? When you press TAB in the command line, the shell will try to complete the line for you up to the next "/" marker! This makes typing in long paths a breeze. - Copy and Paste: This works a bit differently on
Unix than on Windows machines, but once you use it, I
guarantee you'll get hooked. Go ahead and highlight this
in your browser:
~wcbrown/courses/IC221/labs/L01
Now, move your mouse over to the terminal window and click the middle button (or both the left and right buttons at the same time if you're using a two button mouse). BAM! Can't do that on a Windows machine!
ls.
ls: List contents. If you use ls with no arguments, you get the contents of the current directory. If you provide a directory as an argument, you'll get the contents of that directory. Go ahead and type:
lsYou should see the file. If not, either you're not in your home directory, or you didn't copy the file correctly.
Okay, so now we can answer the subject question. A .tar.gz file (also known as a "tarball") is a zipped archive. It's like a .ZIP file in Windows, but a little bit different. To unzip the file, use this command (and remember to use tab completion!):
gunzip unix_practice.tar.gzUse ls again. What happened? The
.gz was
removed from the end of the file, leaving us with
unix_practice.tar To unpack the tar
file, we'll use the tar command.
tar: Create or uncompress a tape archive.
tar -xvf unix_practice.tarYou should see something that looks like this:
> tar -xvf unix_practice.tar unix_practice/ unix_practice/documents/ unix_practice/documents/mydocument.txt unix_practice/programs/ unix_practice/programs/myprogram.c unix_practice/programs/myprogram unix_practice/doc.txt
What's the -xvf you ask? Ah, great question!
We'll use another command to find out.
man: The Unix manual.
Contains reference information for all commands on a Unix
system. To get information about the tar
command, execute:
man tarUse the spacebar to move down a page, 'b' to move up a page, and 'q' to quit. Want some info about
man? Just
type man man! Notice that the tar
command has tons of different command line options to choose
from, and so does just about every other command on Unix.
Whenever you have a question about a command, this is where
you come.
unix_practice in
your home directory. In order to actually navigate into that
directory, we'll need another command:
cd: Change working directory.
Provide a path, either relative or absolute, and
cd will take you there. If you provide
nothing, and just execute cd by itself, it will
take you back to your home directory (which is a nice little
feature). You'll use the cd
.. will move you up one level in the directory
tree.
Using cd and ls, move through the
unix_practice directory and take a look at
what's there. After you're done, return to your home
directory.
unix_practice directory. There's a
file in that directory called doc.txt. Let's see
what's in it.
cat: Concatenate and print. Prints the contents of a file directly to standard output (the screen) without pausing.
To see the contents of doc.txt we'll execute
this:
cat doc.txt
cat is great if you know the file is relatively
small, but if it's large and you want to read it, then
you'll have a hard time. Try this:
cat /etc/passwd
Yikes! cat doesn't pause when the screen is
full to give you a chance to read everything. It actually
does this for a very good reason, which we'll talk about
later. If you want the shell to pause after a screenful,
you need another command.
more: Display a screenful of
text. Use man more to figure out how it works!
I really meant for doc.txt to be in the
documents folder. We already know how we could
copy it there, but instead of copying, let's go ahead and move
the file.
mv: Move file. Takes two
arguments: the source, and the destination. You can also
use it to rename files. Take a look at the man
page for mv, then move the doc.txt
file into the documents folder. Here's how you
can use mv to rename a file:
mv file.txt renamed_file.txtThis will rename
file.txt to
renamed_file.txt. Literally, you're
moving file.txt to renamed_file.txt
cd to get back to your
unix_practice folder. If you execute
ls in that directory, you'll see that there's
another directory in unix_practice called
programs. Change into that directory.
Go ahead and examine the contents of
myprogram.c if you'd like, using either cat or
more. The first thing to notice is that this
isn't C++, it's C. We'll talk about C in a couple of
classes.
I've already compiled myprogram.c into a
program called myprogram, in the same
directory. To execute, just type:
myprogramWhat happened? Get an error message that looks something like this?
bash: ./myprogram: Permission deniedBut can't I do what I want in my home directory? You betcha. The problem is that you have insufficient permission to run
myprogram, but not for the reason you think. To
learn why, we'll use ls with the -l
(that's an 'ell') option. Type this:
ls -lThis will give you a listing of your directory contents with a bunch of other important info. For example:
-rw------- 1 stahl scs 5848 2006-11-21 13:52 myprogram -rw------- 1 stahl scs 167 2006-11-21 13:52 myprogram.cLet's take a look at what each one of these columns means:
-rw------- 1
stahl
scs 167 2006-11-21 13:52 myprogram.c
Number of links to that file.
-rw------- 1 stahl scs 167 2006-11-21 13:52 myprogram.c
The first item is the user name, the second is the group name. Yours will say something different. More on this later.
-rw------- 1
stahl scs 167 2006-11-21 13:52 myprogram.c
The size of the file in bytes.
-rw------- 1
stahl scs 167 2006-11-21 13:52 myprogram.c
The last time the file was modified.
-rw------- 1
stahl scs 167 2006-11-21 13:52 myprogram.c
Duh.
The one we're really interested in to figure out why we got our error is
this one:
-rw------- 1 stahl scs 167 2006-11-21 13:52 myprogram.c
This portion of the output gives the permissions associated
with that file. We read it by breaking it up into three
parts as follows:
-rw-------
The red section right at the start indicates the type of file.
There are a bunch of options, but we're mainly interested
in these: - for a regular file,
d for a directory, and l for a
link (which we'll talk about later).
The next three characters in green are the file
owner's permissions, followed by the group's permissions in
brown and finally everyone else's permissions in orange.
From left to right, they mean "read, write, and execute"
respectively. Although other options are possible, we're
mainly concerned with these three. So, for permissions:
rwx
|||
|||
||+--->execute permissions active, - otherwise
|+---->write permissions active, - otherwise
+----->read permissions active, - otherwise
Let's look at our program's permissions again, they were:
-rw------- 1 stahl scs 5848 2006-11-21 13:52 myprogramWhich means that I can read and write the file (which happens to be a program), no one else can do anything with it, but I don't have the necessary file permissions to execute it! I can fix this with the next command:
chmod: Change permission modes. This
command can only be executed on files that you
own. Remember, Unix was designed to be a
multi-user system, so I shouldn't be able to just go into
another user's directory willy-nilly and give myself
permissions for anything. Speaking of which,
root can execute chmod on
anything you own.
Here's how we'll use chmod most often:
chmod MODE FILE
FILE is just the file we want to change the
permissions of, and MODE is where we specify
exactly what permissions we want the file to have. Use
u, g, and o to modify
the user, group and other permissions respectively. Specify
which permission you want to add with r,
w and x. To add
permissions for the user, we'd do something like
u+r, which means "for the user (file owner), add read
permissions". Similarly, we can do u+rw which
means, "for the user, add read and write permissions. If we
substitue a - for the +, we can
remove permissions.
"We can modify more than one permission group at
a time like this, for example: go+r. I use that
one all the time. It means "for the group and all others, add
read permissions." Why would I want to give everyone read
access to something I own you ask? Well, you're reading such
a reason now! That's how we set the permissions for web
pages!
Now you have enough to correct the permissions of the file
myprogram. Once you've successfully changed
the permissions, go ahead and execute the file. Your screen
will display some instructions for you!
cd to get back to your home directory. Let's
get rid of that unix_practice directory since you
won't be needing it anymore. Try this:
rm unix_practiceWhat happened? Use
man rm to figure out which
option you need to give rm in order for it to
delete the directory for you.
lab01 and cd to that directory, i.e. do
mkdir lab01 ls cd lab01You're going to create a text file called
ans.txt which, eventually, will be your
submission for this lab. Text files are created and modified
with text editors. The text editor we'll be using is
called "emacs". To launch emacs do
emacs &Emacs looks a bit intimidating, but you'll learn to love it ... trust me! To create a new text file named "ans.txt", click on the "open icon"
the type
ans.txt
and hit Enter. A blank buffer comes up. Type your name and a
few newlines, and click on the save icon. If you go back to
your terminal window and ls, you should see that ans.txt
has been created. Use cat, and you'll see the contents.
Exit emacs (file / Exit Emacs), then relaunch it. When you
go to file / Open file, you should see ans.txt listed in the box.
Double-click on it to open it up.
lab01/ans.txt. This file must
have your name and alpha in the first line!
Copy and paste this text
into ans.txt
and add the answers into the file. Do not forget your
name and alpha!
1. For each of the following utilites, give a one-line description of
what they do:
pwd
cp
ls
tar
man
cd
cat
more
mv
chmod
2. Read the man page for chmod, then answer the following questions:
* what does it mean to have x permissions on a directory?
* What about r permissions?
You can do some experimentation to find out. Try looking at the
IC221 course directory at this path:
~wcbrown/public_html/courses/IC221/ . You'll find some folders
that you have x permissions for, but not r permissions. If you say
"x lets you execute a directory and r lets you read it" you'll get
a zero.
3. Copy the file ~wcbrown/courses/IC221/labs/L01/riddle.tar.gz. Unpack
this and follow the clues to find "the answer". Write the answer in
ans.txt. Hint: You'll need to use the ls/cp/cat/tar/mv/chmod commands
described above.
You will submit this lab, like all your labs, via the class
submit script. Change your current directory to the one
containing lab01: not to lab01
itself, but to its parent. Then do:
/courses/wcbrown/etc/ic221/submit lab01 ← If I (Dr. Brown) get a chance to updating submit scripts, this might change!
... for Dr. Brown's students,
/courses/stahl/etc/ic221/submit... for Dr. Stahl's students,
/courses/crabbe/etc/ic221/submit... for Dr. Crabbe's students. For thorough instructions on the submit script see Submission.pdf.