Unix and Linux System Administration and Shell

Unix and Linux System Administration and Shell Programming
Unix and Linux System Administration
and Shell Programming
version 56 of August 12, 2014
Copyright © 1998, 1999, 2000, 2001, 2002, 2003, 2004, 2005, 2006, 2007, 2009, 2010, 2011, 2012, 2013, 2014 Milo
This book includes material from the http://www.osdata.com/ website and the text book on computer programming.
Distributed on the honor system. Print and read free for personal, non-profit, and/or educational purposes. If you like the
book, you are encouraged to send a donation (U.S dollars) to Milo, PO Box 5237, Balboa Island, California, USA 92662.
This is a work in progress. For the most up to date version, visit the website http://www.osdata.com/ and
http://www.osdata.com/programming/shell/unixbook.pdf — Please add links from your website or Facebook page.
Professors and Teachers: Feel free to take a copy of this PDF and make it available to your class (possibly through
your academic website). This way everyone in your class will have the same copy (with the same page numbers) despite
my continual updates. Please try to avoid posting it to the public internet (to avoid old copies confusing things) and take it
down when the class ends. You can post the same or a newer version for each succeeding class. Please remove old copies
after the class ends to prevent confusing the search engines. You can contact me with a specific version number and class
end date and I will put it on my website.
version 56
page 1
Unix and Linux System Administration and Shell Programming
Unix and Linux Administration and Shell
Programming
chapter 0
This book looks at Unix (and Linux) shell programming and system administration.
This book covers the basic materials needed for you to understand how to administer your own Linux or Unix server, as
well as how to run your own personal desktop version of Linux or Mac OS X.
This book goes beyond the typical material in a shell scripting class and presents material related to either downloading
and compiling existing software (including ports to new hardware and/or operating systems) or for preparing your own
software for release via the internet.
requirements
You need a willingness to learn.
You need a working computer or server or access to one.
The computer needs a working version of Unix, Linux, Mac OS X, AIX, HP/UX, Solaris, etc. (it can be a dual boot
computer).
The new version of Mac OS X 10.9 (Mavericks) is now available on the Mac App Store at www.apple.com as of October
22, 2013. The new version of Mac OS X 10.8.1 (Mountain Lion) is now available on the Mac App Store at www.apple.com
as of August 23, 2012. The new version of Mac OS X 10.8 (Mountain Lion) is now available on the Mac App Store at
www.apple.com as of July 25th, 2012. Tell them you heard about it from www.osdata.com when you register your new
copy.
A working connection to the internet is recommended, but not required, preferably a high speed connection.
You may use LAMP on Linux, MAMP on Mac OS X, or WAMP on WIndows to set up and experiment with a local web
server.
You may want to have a domain name of your own and web hosting to try out controlling a server. You can use
GoDaddy or HostGator or most any other major hosting provider to obtain these servcies for low cost with great telephone
tech support. The OSdata.com website where this book is offered to the public is hosted by Host Gator. You may use any
other hosting service you want.
options
Almost anyone can learn this material. It isn’t complicated. It really helps if you enjoy doing this kind of thing. You will
learn faster and you will enjoy the work. If you don’t enjoy this kind of computer activity, please think carefully about
whether or not you want to spend decades at a job you hate.
Almost anyone can slog through and learn at least some of this material, but an aptitude for this material greatly helps
learning. If you are strong at grammar, then you will probably be able to master. This material. mathematical ability is
useful, but not necessary.
Many portions of this book require root or administrator access. While you learn better if you can actually try out each
command for yourself, you can just read about root material if you don’t have root or administrator access.
version 56
page 2
Unix and Linux System Administration and Shell Programming
Some portions of this book require special software. Most of the software can be downloaded for free. Those with Mac
OS X should have the Developer Tools installed. These are available for free on either the install DVD/CD or from Apple
at http://connect.apple.com/
A static IP address is in general useful and is required for some portions of this book.
proletarian
This book is intentionally proletarian and is specifically intended to allow anyone, regardless of socio-economic position,
to have access to the basic information and the power against the established authorities that such knowledge provides.
This subversive intent is consistent with Linux, which was created by a college student who couldn’t afford the dominant
tools of the time.
This contrasts strongly with Apple’s Macintosh, Mac OS X, and iOS, all of which are specifically targeted for sales to
the trendiest members of the wealthy class. Apple does offer a discount to school children, but that discount is smaller than
the discount Apple offers to large corporations.
This also contrasts with Microsoft’s Windows, which is specifically targeted for sales to large corporations and their
employees.
And this contrasts with Google’s Android, which is specifically targeted for businesses. The users of Android products
are part of Google’s product, not the customer. The customers are businesses who want detailed information on the masses
for purposes of advertising.
This book is intended to make knowledge available to everyone, especially those who are increasingly being shut out of
the mainstream education system.
goals
Please note that at the time this paragraph was written, these goals are not yet implemented in the text. Now is a good
time to make suggestions on modifications of the goals.
The reader will understand the Unix computing environment and history.
The reader will be able to access the Unix system and perform basic operations, including using help features.
The reader will be able to access and manipulate files and directories, including basic and advanced directory and file
management. The reader will be able to use file system utilities.
The reader will be able to design and implement file system security.
The reader will be able to print documents.
The reader will be able to perform system backups and restores.
The reader will be able to troubleshoot system processes.
The reader will be able to perform environment customization.
The reader will become acquainted with networks, web servers, web clients, transaction security, and other basic network
concepts.
The reader will learn how to create a web page that accepts input, program the server response, and interface the input
with a database.
version 56
page 3
Unix and Linux System Administration and Shell Programming
The reader will understand web server functionailty and be able to install and configure Apache.
The reader will be able to work with, understand, and configure DNS functionality.
The reader will be able to set up and maintain his or her own general purpose server, including hosting, server
administration, security, user interactivity, and database integration.
The reader will be familiar with SQL Server architecture and able to efficiently administer and maintain SQL Server
instances and databases.
The reader will be able to secure a web server and its websites.
personal reference book
I strongly recommend that each reader print out this PDF and the manual pages from your computer and place the
resulting pages into a looseleaf binder. That will allow you to organize the chapters in the order that makes the most sense
for you. It will also allow you to make notes in the margins. It will allow you to easily add materials from other sources.
The resulting binder will be personalized to meet your individual needs.
If you have a limited printing budget (ink is expensive), print out portions of this PDF as needed.
chapter contents
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
cool shell tricks
basics of computers
Unix/Linux history
choice of shells
connecting to a shell (Telnet and SSH; terminal emulator)
shell basics (book conventions; root or superuser; starting your shell; login and password; prompt; command
example)
login/logout (login; select system; account name; password; terminal type; logout; exit)
passwd (setting password; local password; periodic changes; 100 most common passwords; secure passwords;
superuser)
command structure (single command; who; failed command; date; options, switches, or flags; universal time;
arguments; options and arguments; operators and special characters)
quick tour of shell commands
man (using man for help; man sections)
cat (creating files; example files for this book; viewing files; combining files)
command separator (semicolon)
less, more, pg
file system basics (graphics examples; directory tree; important directories; home directory; parent and child
directories; absolute paths; relative paths; dots, tildes, and slashes)
pwd
command history
built-in commands
ls
cd
cp
mv
rm (recursive)
shred
mkdir
alias
version 56
page 4
Unix and Linux System Administration and Shell Programming
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
50.
51.
52.
53.
54.
55.
56.
57.
58.
59.
60.
61.
62.
63.
64.
65.
66.
67.
68.
69.
70.
71.
72.
73.
74.
75.
76.
77.
78.
79.
80.
81.
82.
pipes
scripts
sysadmin and root/superuser
sudo
su
who
advanced file systems
major directories
network file system (NFS)
tail
wc
chmod
chown
shell levels and subshells
substitutions
command substitutions
arithmetic substitutions
flow control
management tools
df
du
processes
ps
kill
nice
w
date
uname
uptime
top
lsof
free
vmstat
polkit
defaults (screencapture; Mac Flashback Trojan)
init (init; Linux run levels)
sendmail
ifconfig (view configuration; static IP address)
arp
netstat (view connections; main info; routing address)
route (view connections; routing commands)
ping (test packets; measuring)
nslookup
traceroute (entire route; etiquette)
ftp and sftp
curl
sysstat
at (example; removing a job; timing)
back ups
tar
touch (multiple files; specific time)
find
text processing
basename
sed (fixing end of line; adding line numbers)
awk (remove duplicate lines)
version 56
page 5
Unix and Linux System Administration and Shell Programming
83.
84.
85.
86.
87.
88.
89.
90.
91.
92.
93.
94.
95.
96.
97.
screencapture (from graphic user interface; changing defaults; command line screenshots)
signals
kernel modules
LAMP
mysql
PHP
Perl
Tcl
installing software from source code
computer programming
size of programs
kinds of programs
programming languages
standards and variants
test bed
Appendix:
A.
B.
C.
D.
sommand summaries
computer history
critical reasoning
Forth-like routines
version 56
page 6
Unix and Linux System Administration and Shell Programming
cool shell tricks for Unix, Mac OS X, and Linux
chapter 1
summary
This chapter looks at cool shell tricks for Unix, Linux, and Mac OS X to give you an idea of the power of the shell.
A quick summary of how to get to the shell is included in this chapter (more detailed explanations, including what to do
when things go wrong, are in following chapters.
If you need a primer on computer terminology, please look at the next chapter on basics of computers.
WARNING: Never trust any Terminal/Shell commands you find on the internet. Only run shell
commands you understand. In particular, never run anything that includes sudo anywhere in the command line unless you
are absolutely certain what the command does. When you run a command line with sudo you are giving permission to have
complete (possibly destructive) control of your computer at the root level. And, yes, this advice applies even to this book.
Don’t run any commands with sudo unless you know for sure what you are doing.
cool shell tricks
This chapter has a handful of cool shell tricks. These are intended to show a beginner that a command line shell can be
as fun as any graphic user interface and get across the idea that there is a lot of power in the shell that simply doesn’t exist
in a standard graphic user interface.
definitions
Unix is one of the ground-breaking operating systems from the early days of computing. Mac OS X is built on top of
Unix. Linux is a variation of Unix.
The shell is the command line interface for running Unix (and Mac OS X and Linux) with just typing (no mouse).
operating system The software that provides a computer’s basic tasks, such as scheduling tasks, recognizing input from
a keyboard, sending output to a display screen or printer, keeping track of files and folders (directories), running
applications (programs), and controlling peripherals. Operating systems are explained in more detail for beginners just
below.
Unix Unix (or UNIX) is an interactive multi-user multitasking timesharing operating system found on many types of
computers. It was invented in 1969 at AT&T’s Bell Labs by a team led by Ken Thompson and Dennis Ritchie. Some
versions of Unix include: AIX, A/UX, BSD, Debian, FreeBSD, GNU, HP-UX, IRIX, Linux, Mac OS X, MINIX, Mint,
NetBSD, NEXTSTEP, OpenBSD, OPENSTEP, OSF, POSIX, Red Hat Enterprise, SCO, Solaris, SunOS, System V,
Ubuntu, Ultrix, Version 7, and Xenix.
Linux An open-source version of the Unix operating system.
graphical user interface A graphical user interface (GUI) is a windowing system, with windws, icons, and menus,
operated by a mouse, trackball, touch screen, or other pointing device, used for controlling an operating system and
application programs (apps). The Macintosh, Windows, Gnome, and KDE are famous examples of graphical user
interfaces.
command line interface A command line interface (CLI orcommand line user interface CLUI) is a text only interface,
operated by a keyboard, used for controlling an operating system and programs.
version 56
page 7
Unix and Linux System Administration and Shell Programming
shell The shell is the command line interface for Unix, Linux, and Mac OS X. In addition to intrepetting commands, it is
also a programming language.
shell uses
The Unix shell is a very powerful programming language, with access to hundreds of useful tools that are designed to
work with each other, and access to the very heart of the operating system (although this is only available to the root or
superuser account for security reasons).
Unix (and therefore also Mac OS X and Linux) has more than 200 basic commands (also called tools or utilities) that are
distributed with the standard operating system. This collection and the ease with which they work together is the major
source of the power of Unix. The vast majority of these standard tools are designed to be used from a command line (the
shell).
The shell is most commonly used to control servers. Servers are the computers used to host websites. The most common
operating system for the world’s web servers is Linux. If you learn shell scripting and system administration, you can run
your own server and possibly get a job.
The shell can be used to control a desktop or portable computer. Some tablets and smart phones have a shell. The iPhone
actually has a shell, but it can’t be accessed witout jailbreaking the iPhone.
The shell will often run even when a computer is partly broken. Both Mac OS X and Linux (as well as almost all
versions of Unix) can be run in a special single user mode. This starts up the computer or server with just the command
line shell running. This can be used to service a computer or server, including both diagnosis and repair.
The shell is extremely useful for programming. Even when a programmer uses a graphical integrated development
environment (IDE), the programmer is likely to still heavily use the shell for programming support. Some IDEs even have
shell access built-in.
Shell scripts are widely used by system administrators for performing common tasks.
command line interface
Before the widespread introduction of graphic user interfaces (GUI), computers were controlled either by punched cards,
paper tape, or magnetic tape (a batch system) or a command line interface (CLI) using an interactive terminal (originally,
some variation of a teletype machine from the telegraph technology). The earliest computers were controlled by front panel
lights and switches or even by directly changing the wiring.
The command line interface on interactive terminals was a major advance. Because of limitations of the early hardware
(at a time when a computer’s entire memory might be measured in hundreds of bytes), the first CLIs compressed the
number of characters, using two or three letter abbreviations for commands and single character switches for options to
commands.
The modern Unix/Linux shells carry over this early limitation because there is the need to remain backward compatible
and still run shell scripts that are decades old, but essential to continued operation of legacy systems.
This historical limitation makes it difficult for newcomers to figure out how to use a Unix/Linux shell.
how to find Terminal
You can run these commands by copying and pasting into Terminal, a program that is available for free and preinstalled
on Mac OS X and most versions of Linux. Some of these commands will only work on a particular operating system (this
will be indicated), but most can be run from Mac OS X, any distribution Linux, and any kind of Unix.
On Mac OS X, you will find Terminal by opening the Applications folder (on your main hard drive), then opening the
Utilities folder, then scrolling down the list until you find Terminal or Terminal.app. Drag it to your Dock, because you will
version 56
page 8
Unix and Linux System Administration and Shell Programming
be using it a lot.
On Ubuntu Linux, look in Applications menu > Accessories > Terminal. Single click and hold, then drag to your panel,
because you will be using it a lot.
On some versions of Linux, you can press the CONTROL and the ALT and the F1 keys all at once to bring up Terminal.
In Gnome, use the Applications menu > Accessories > Terminal or the keyboard shortcut of CONTROL and ALT and T
(all at the same time).
version 56
page 9
Unix and Linux System Administration and Shell Programming
In KDE, use the KMenu > System > Terminal Program (Konsole).
In Linux Mint you can use the keyboard shortcut of holding down the CONTROL and ALT and T keys all at once.
As a last resort, you can use the file search feature in Linux or Mac OS X and search for “Terminal”.
warnings
Be careful about any hints you find on the internet. There are people who suggest very destructive commands in the guise
of a useful or fun hint just to trick beginners into destroying their computers.
Be very careful if you use any command that includes sudo , because it runs at the root level, allowing complete access to
the entire computer with all safeguards turned off. Very powerful. Potentially very distructive in a hurry. There are
legitimate hints and cool tricks that use sudo , but be careful to type them exactly as you you see them in the hint (or copy
and paste) and only use sudo hints from trusted soruces.
Watch out for anything that includes the command rm or rm *. That is the remove command versions of it can literally
wipe out all of your hard drives in seconds in such a way that only a very expensive data recovery specialist (thousands of
dollars) can get your data back.
Also watch out for anything that includes the command shred . That is the secure delete and even the most expensive
data recovery specialist in the world can’t get your data back.
cool ASCII art
If your computer is connected to the internet, you can use the shell to watch the entire Star Wars Episode IV in old
fashioned ASCII. Type telnet towel.blinkenlights.nl followed by ENTER or RETURN. If you have IPv6, you get
extra scenes and color support.
$ telnet towel.blinkenlights.nl
play a CD
version 56
page 10
Unix and Linux System Administration and Shell Programming
On Linux you can play a CD from the command line. The following example plays the first track of the CD. Change the
number to play a different track.
$ cdplay play 1
On Linux you can get a free coffee cup holder (eject the CD-ROM tray).
$ eject
version 56
page 11
Unix and Linux System Administration and Shell Programming
basics of computers
chapter 2
summary
This chapter will cover some very basic information on how computers work. I’m not trying to insult anyone’s
intelligence. This material addresses some of the questions asked by novice test readers.
major desktop systems
There are only three major desktop computing systems left in common existence: Windows, Macintosh, and Linux.
Apple Macintosh was the first commercially successful graphical user interface. That is, a computer where one uses a
mouse (or other pointing device) and icons, windows, and menus to control the computer. Since the turn of the century,
Mac OS X has actually been built on top of Unix.
Unix is the last major survivor of the mainframe era (back when computers were so large that they took up an entire
room, or even an entire building). Its big advantage of its compeitors was that it is available in free open source versions
and it runs on an extremely wide variety of computer hardware (including many computers that are no longer used).
Microsoft Windows is loosely based on the Macintosh and for decades dominated the personal computer market.
Linux is a very popular open source variation of Unix. It is the most common operating system for servers and the third
most popular desktop computing operating system. In its early days it was very geeky and very difficult to use, but it now
sports two major graphical user interfaces (Gnome and KDE) and is reasonably easy to use.
FreeBSD is the next most popular open source version of Unix and is commonly used for servers. Solaris is the next
most popular commercial version of Unix. IBM and HP both have their own commercial versions of Unix.
definitions
Unix is one of the ground-breaking operating systems from the early days of computing. Mac OS X is built on top of
Unix. Linux is a variation of Unix.
The shell is the command line interface for running Unix (and Mac OS X and Linux) with just typing (no mouse).
operating system The software that provides a computer’s basic tasks, such as scheduling tasks, recognizing input from
a keyboard, sending output to a display screen or printer, keeping track of files and folders (directories), running
applications (programs), and controlling peripherals. Operating systems are explained in more detail for beginners just
below.
Unix Unix (or UNIX) is an interactive multi-user multitasking timesharing operating system found on many types of
computers. It was invented in 1969 at AT&T’s Bell Labs by a team led by Ken Thompson and Dennis Ritchie. Some
versions of Unix include: AIX, A/UX, BSD, Debian, FreeBSD, GNU, HP-UX, IRIX, Linux, Mac OS X, MINIX, Mint,
NetBSD, NEXTSTEP, OpenBSD, OPENSTEP, OSF, POSIX, Red Hat Enterprise, SCO, Solaris, SunOS, System V,
Ubuntu, Ultrix, Version 7, and Xenix.
Linux An open-source version of the Unix operating system.
graphical user interface A graphical user interface (GUI) is a windowing system, with windws, icons, and menus,
operated by a mouse, trackball, touch screen, or other pointing device, used for controlling an operating system and
application programs (apps). The Macintosh, Windows, Gnome, and KDE are famous examples of graphical user
interfaces.
version 56
page 12
Unix and Linux System Administration and Shell Programming
command line interface A command line interface (CLI orcommand line user interface CLUI) is a text only interface,
operated by a keyboard, used for controlling an operating system and programs.
shell The shell is the command line interface for Unix, Linux, and Mac OS X.
the shell
Unix (and therefore also Mac OS X and Linux) has more than 200 basic commands (also called tools or utilities) that are
distributed with the standard operating system. This collection and the ease with which they work together is the major
source of the power of Unix. The vast majority of these standard tools are designed to be used from a command line (the
shell).
A shell is primarily a command interpreter.
In a graphical user interface such as Macintosh or Windows, the user controls the computer and programs primarily
through pointing and clicking, supplemented by some typing.
In a shell, all of the commands are typed. The shell figures out the meaning of what you typed and then has the computer
do as instructed.
But the shell is much more than just a command interpreter. It is also a complete programming language.
Because the shell is a complete programming language, with sequences, decisions, loops, and functions, it can do things
well beyond pointing and clicking. It can take control of your computer and react to changing circumstances.
Programming languages are used to make all of the computer programs and smart phone apps you’ve ever seen or used.
Your imagination is the only limit on the power of the shell. Anything that can be done with a computer can be done with
the shell.
operating systems
The seven layers of software are (top to bottom): Programs; System Utilities; Command Shell; System Services; User
Interface; Logical Level; and Hardware Level. A Graphics Engine stradles the bottom three layers. Strictly speaking, only
the bottom two levels are the operating system, although even technical persons will often refer to any level other than
programs as part of the operating system (and Microsoft tried to convince the Justice Department that their web browser
application is actually a part of their operating system).
The following are examples of each category:
Programs: Examples of Programs include your word processor, spreadsheet,
graphics programs, music software, games, etc.
System Utilities: Examples of System Utilities include file copy, hard drive
repair, and similar items. On the Macintosh, all the Desk Accessories
(calculator, key caps, etc.) and all of the Control Panels are examples of System
Utilities.
Command Shell: The Command Shell on the Macintosh is the Finder and was
the first commercially available graphic command shell. On Windows, the
Command Shell is a poorly integrated comination of the File Manager and the
Program Manager. The command line (C:\ prompt) of MS-DOS or Bourne
Shell of Unix are examples of the older style text-based command shells.
System Services: Examples of System Services are built-in data base query
languages on mainframes or the QuickTime media layer of the Macintosh.
User Interface: Until the Macintosh introduced Alan Kay’s (inventer of the
personal computer, graphic user interfaces, object oriented programming, and
software agents) ground breaking ideas on human-computer interfaces,
version 56
page 13
Unix and Linux System Administration and Shell Programming
operating systems didn’t include support for user interfaces (other than simple
text-based shells). The Macintosh user interface is called the Macintosh
ToolBox and provides the windows, menus, alert boxes, dialog boxes, scroll bars, buttons, controls, and other user
interface elements shared by almost all programs.
Logical Level of Operating System: The Logical Level of the operating system provides high level functions, such as
file management, internet and networking facilities, etc.
Hardware Level of Operating System: The Hardware Level of the operating system controls the use of physical
system resources, such as the memory manager, process manager, disk drivers, etc.
Graphics Engine: The Graphics Engine includes elements at all three of the lowest levels, from physically displaying
things on the monitor to providing high level graphics routines such as fonts and animated sprites.
Human users normally interact with the operating system indirectly, through various programs (application and system)
and command shells (text, graphic, etc.), The operating system provides programs with services thrrough system programs
and Application Program Interfaces (APIs).
basics of computer hardware
A computer is a programmable machine (or more precisely, a programmable sequential state machine) . There are two basic kinds of
computers: analog and digital.
Analog computers are analog devices. That is, they have continuous states rather than discrete numbered states. An
analog computer can represent fractional or irrational values exactly, with no round-off. Analog computers are almost never
used outside of experimental settings.
A digital computer is a programmable clocked sequential state machine. A digital computer uses discrete states. A
binary digital computer uses two discrete states, such as positive/negative, high/low, on/off, used to represent the binary
digits zero and one.
The French word ordinateur, meaning that which puts things in order, is a good description of the most common
functionality of computers.
what are computers used for?
Computers are used for a wide variety of purposes.
Data processing is commercial and financial work. This includes such things as billing, shipping and receiving,
inventory control, and similar business related functions, as well as the “electronic office”.
Scientific processing is using a computer to support science. This can be as simple as gathering and analyzing raw data
and as complex as modelling natural phenomenon (weather and climate models, thermodynamics, nuclear engineering,
version 56
page 14
Unix and Linux System Administration and Shell Programming
etc.).
Multimedia includes content creation (composing music, performing music, recording music, editing film and video,
special effects, animation, illustration, laying out print materials, etc.) and multimedia playback (games, DVDs, instructional
materials, etc.).
Servers includes web servers. Every website is hosted on a computer called a server. When you connect to a website in
your web browser, your computer connects to a web server that provides your web browser with all of the parts (text,
pictures, Flash, style sheets, JavaScripts, etc.) needed for your web browser to display any particular web page.
parts of a computer
The classic crude oversimplication of a computer is that it contains three
elements: processor unit, memory, and I/O (input/output). The borders
between those three terms are highly ambigious, non-contiguous, and
erratically shifting.
A slightly less crude oversimplification divides a computer into five
elements: arithmetic and logic subsystem, control subsystem, main storage,
input subsystem, and output subsystem.
processor
arithmetic and logic
control
main storage
external storage
input/output overview
input
output
processor
The processor is the part of the computer that actually does the computations. This is sometimes called an MPU (for
main processor unit) or CPU (for central processing unit or central processor unit).
A processor typically contains an arithmetic/logic unit (ALU), control unit (including processor flags, flag register, or
status register), internal buses, and sometimes special function units (the most common special function unit being a
floating point unit for floating point arithmetic).
Some computers have more than one processor. This is called multi-processing.
The major kinds of digital processors are: CISC, RISC, DSP, and hybrid.
CISC stands for Complex Instruction Set Computer. Mainframe computers and minicomputers were CISC processors,
with manufacturers competing to offer the most useful instruction sets. Many of the first two generations of
microprocessors were also CISC.
RISC stands for Reduced Instruction Set Computer. RISC came about as a result of academic research that showed that a
small well designed instruction set running compiled programs at high speed could perform more computing work than a
CISC running the same programs (although very expensive hand optimized assembly language favored CISC).
DSP stands for Digital Signal Processing. DSP is used primarily in dedicated devices, such as MODEMs, digital
cameras, graphics cards, and other specialty devices.
Hybrid processors combine elements of two or three of the major classes of processors.
version 56
page 15
Unix and Linux System Administration and Shell Programming
arithmetic and logic
An arithmetic/logic unit (ALU) performs integer arithmetic and logic operations. It also performs shift and rotate
operations and other specialized operations. Usually floating point arithmetic is performed by a dedicated floating point unit
(FPU), which may be implemented as a co-processor.
An arithmetic/logic unit (ALU) performs integer arithmetic and logic operations. It also performs shift and rotate
operations and other specialized operations. Usually floating point arithmetic is performed by a dedicated floating point unit
(FPU), which may be implemented as a co-processor.
control
Control units are in charge of the computer. Control units fetch and decode machine instructions. Control units may also
control some external devices.
A bus is a set (group) of parallel lines that information (data, addresses, instructions, and other information) travels on
inside a computer. Information travels on buses as a series of electrical pulses, each pulse representing a one bit or a zero
bit (there are trinary, or three-state, buses, but they are rare). An internal bus is a bus inside the processor, moving data,
addresses, instructions, and other information between registers and other internal components or units. An external bus is
a bus outside of the processor (but inside the computer), moving data, addresses, and other information between major
components (including cards) inside the computer. Some common kinds of buses are the system bus, a data bus, an address
bus, a cache bus, a memory bus, and an I/O bus.
main storage
Main storage is also called memory or internal memory (to distinguish from external memory, such as hard drives).
RAM is Random Access Memory, and is the basic kind of internal memory. RAM is called “random access” because the
processor or computer can access any location in memory (as contrasted with sequential access devices, which must be
accessed in order). RAM has been made from reed relays, transistors, integrated circuits, magnetic core, or anything that
can hold and store binary values (one/zero, plus/minus, open/close, positive/negative, high/low, etc.). Most modern RAM is
made from integrated circuits. At one time the most common kind of memory in mainframes was magnetic core, so many
older programmers will refer to main memory as core memory even when the RAM is made from more modern
technology. Static RAM is called static because it will continue to hold and store information even when power is
removed. Magnetic core and reed relays are examples of static memory. Dynamic RAM is called dynamic because it loses
all data when power is removed. Transistors and integrated circuits are examples of dynamic memory. It is possible to have
battery back up for devices that are normally dynamic to turn them into static memory.
ROM is Read Only Memory (it is also random access, but only for reads). ROM is typically used to store thigns that will
never change for the life of the computer, such as low level portions of an operating system. Some processors (or variations
within processor families) might have RAM and/or ROM built into the same chip as the processor (normally used for
processors used in standalone devices, such as arcade video games, ATMs, microwave ovens, car ignition systems, etc.).
EPROM is Erasable Programmable Read Only Memory, a special kind of ROM that can be erased and reprogrammed with
specialized equipment (but not by the processor it is connected to). EPROMs allow makers of industrial devices (and other
similar equipment) to have the benefits of ROM, yet also allow for updating or upgrading the software without having to
buy new ROM and throw out the old (the EPROMs are collected, erased and rewritten centrally, then placed back into the
machines).
Registers and flags are a special kind of memory that exists inside a processor. Typically a processor will have several
internal registers that are much faster than main memory. These registers usually have specialized capabilities for
arithmetic, logic, and other operations. Registers are usually fairly small (8, 16, 32, or 64 bits for integer data, address, and
control registers; 32, 64, 96, or 128 bits for floating point registers). Some processors separate integer data and address
registers, while other processors have general purpose registers that can be used for both data and address purposes. A
processor will typically have one to 32 data or general purpose registers (processors with separate data and address registers
typically split the register set in half). Many processors have special floating point registers (and some processors have
version 56
page 16
Unix and Linux System Administration and Shell Programming
general purpose registers that can be used for either integer or floating point arithmetic). Flags are single bit memory used
for testing, comparison, and conditional operations (especially conditional branching).
external storage
External storage (also called auxillary storage) is any storage other than main memory. In modern times this is mostly
hard drives and removeable media (such as floppy disks, Zip disks, DVDs, CDs, other optical media, etc.). With the advent
of USB and FireWire hard drives, the line between permanent hard drives and removeable media is blurred. Other kinds of
external storage include tape drives, drum drives, paper tape, and punched cards. Random access or indexed access devices
(such as hard drives, removeable media, and drum drives) provide an extension of memory (although usually accessed
through logical file systems). Sequential access devices (such as tape drives, paper tape punch/readers, or dumb terminals)
provide for off-line storage of large amounts of information (or back ups of data) and are often called I/O devices (for
input/output).
input/output overview
Most external devices are capable of both input and output (I/O). Some devices are inherently input-only (also called
read-only) or inherently output-only (also called write-only). Regardless of whether a device is I/O, read-only, or writeonly, external devices can be classified as block or character devices.
A character device is one that inputs or outputs data in a stream of characters, bytes, or bits. Character devices can
further be classified as serial or parallel. Examples of character devices include printers, keyboards, and mice.
A serial device streams data as a series of bits, moving data one bit at a time. Examples of serial devices include printers
and MODEMs.
A parallel device streams data in a small group of bits simultaneously. Usually the group is a single eight-bit byte (or
possibly seven or nine bits, with the possibility of various control or parity bits included in the data stream). Each group
usually corresponds to a single character of data. Rarely there will be a larger group of bits (word, longword, doubleword,
etc.). The most common parallel device is a printer (although most modern printers have both a serial and a parallel
connection, allowing greater connection flexibility).
A block device moves large blocks of data at once. This may be physically implemented as a serial or parallel stream of
data, but the entire block gets transferred as single packet of data. Most block devices are random access (that is,
information can be read or written from blocks anywhere on the device). Examples of random access block devices include
hard disks, floppy disks, and drum drives. Examples of sequential access block devcies include magnetic tape drives and
high speed paper tape readers.
input
Input devices are devices that bring information into a computer.
Pure input devices include such things as punched card readers, paper tape readers, keyboards, mice, drawing tablets,
touchpads, trackballs, and game controllers.
Devices that have an input component include magnetic tape drives, touchscreens, and dumb terminals.
output
Output devices are devices that bring information out of a computer.
Pure output devices include such things as card punches, paper tape punches, LED displays (for light emitting diodes),
monitors, printers, and pen plotters.
Devices that have an output component include magnetic tape drives, combination paper tape reader/punches, teletypes,
version 56
page 17
Unix and Linux System Administration and Shell Programming
and dumb terminals.
version 56
page 18
Unix and Linux System Administration and Shell Programming
Unix and Linux history
chapter 3
summary
This chapter looks at the history of Unix and Linux.
The history can help you understand why things are they way they are in Unix and Linux.
Of particular concern are:
the small storage space of early computers which resulted in short command names and one character options
the switchover from teletype I/O to video terminal I/O
the use of eight bit bytes, ASCII characters, and plain text for interprocess communication
treating files as collections of bytes and devices, directories, and certain kinds of inter-process communications as
files
the use of small, single-purpose programs that can be easily combined together to perform complex tasks
Unix history
When Unix came into existence digital computers had been in commercial use for more than a decade.
Mainframe computers were just starting to be replaced by minicomputers. The IBM 360 mainframe computer and clones
by other manufacturers was still the dominant computer. The VAX, with its virtual memory addressing, was still in
development. Microcomputers were just an experimental idea. The CDC mainframe, a forerunner of the more famous Cray
supercomputers, was still the world’s fastest computer.
COBOL was the most common programming language for business purposes. FORTRAN was the most popular
programming language for scientific and systems programming. PL/I was a common third language. ALGOL was dominate
in the academic community.
Most data entry was still performed using punched cards. Directly connected terminals were typically teletypes (TTY), a
technology originally developed for telegraph systems. New terminals combining a keyboard and a monitor were still rare.
These were also called CRTs (for Cathode Ray Tube) or dumb terminals or smart terminals (depending on the capabilities).
Unlike modern black letters on white background terminals (pioneered with the Apple Lisa and Apple Macintosh) and full
color monitors (popularized with the Atari and Commodre Amiga personal computers), these monitors were a single color
phosphor (usually green) on a dark gray background. Even though disk drives existed, they were expensive and had small
capacities. Magnetic tape was still the most commonly used form of storage of large data sets. Minicomputers were making
use of punched paper tape.
Interactive computing was still rare.
Bell Telephone Laboratories was still using the BESYS operating system from 1957. In 1964, Bell Labs decided to
update their operating system, leading to a 1965 partnership between the Massachusetts Institute of Technology (MIT),
AT&T Bell Labs, and General Electric to create an experimental operating system called Multics (MULTIplexed operating
and Computing System) for the GE-645 mainframe computer. AT&T intended to offer subscription-based computing
services over the phone lines, an idea similar to the modern cloud approach to computing.
While the Multics project had many innovations that went on to become standard approaches for operating systems, the
project was too complex for the time.
Bell Labs pulled out of the Multics project in 1969.
version 56
page 19
Unix and Linux System Administration and Shell Programming
Ken Thompson and Dennis Ritchie offered to design a new operating system using a DEC PDP-7 computer that was
available at the time.
UNICS (UNIplexed operating and Computing System, a pun on MULTICS) was created at AT&T’s Bell Labs in 1969
by a group that included Ken Thompson, Dennis Ritchie, Brian Kernighan, Douglas McIlroy, Michael Lesk and Joe
Ossanna.
This group of researchers, the last of the AT&T employees involved in Multics, decided to attempt the goals of Multics
on a much more simple scale.
“What we wanted to preserve was not just a good environment in which to do programming, but a system around which
a fellowship could form. We knew from experience that the essence of communal computing, as supplied by remoteaccess, time-shared machines, is not just to type programs into a terminal instead of a keypunch, but to encourage close
communication,” according to Dennis Ritchie.
Unix was originally intended as a programmer’s workbench. The original version was a single-user system. As Unix
spread through the academic community, more general purpose tools were added, turning Unix into a general purpose
operating system.
While Ken Thompson still had access to the Multics environment, he wrote simulations on Multics for Unix’s file and
paging system.
Ken Thompson ported the Space Travel computer game from Multics to a Digital Equipment Corporation (DEC) PDP-7
he found at Bell Labs.
Ken Thompson and Dennis Ritchie led a team of Bell Labs researchers (the team included Rudd Canaday) working on
the PDP-7 to develop a hierarchial file system, computer processes, device files, a command-line interpreter, and a few
small utility programs.
Early work was done in a BCPL (Basic Combined Programming Language) a language intended for writing compilers
and system software. BCPL was simplified and revised into the B programming language.
By the end of 1969, UNICS had a primitive kernel, an editor, an assembler, a simple shell command interpreter, and a
few basic command utilities, including rm, cat , and cp.
The first Unix command shell, called the Thompson shell and abbreviated sh, was written by Ken Thompson at AT&T’s
Bell Labs, was much more simple than the famous Unix shells that came along later. The Thompson shell was distributed
with Versions 1 through 6 of Unix, from 1971 to 1975.
In 1970 Dennis Ritchie and Ken Thompson traded the promise to add text processing capabilities to Unix for the use of a
Digital Equipment Corporation (DEC) PDP-11/20. The initial version of Unix, a text editor, and a text formatting program
called roff were all written in PDP-11/20 assembly language.
The PDP-11/20 computer had 24K of RAM. The operating system used 12K. The remaining 12K was split between
application programs and a RAM disk. The file size limit was 64K and the disk size limit was 512K.
Soon afterwards roff evolved into troff with full typesetting capability. The first Unix book, UNIX Programmer’s
Manual, by Ken Thompson and Dennis Ritchie, was published on November 3, 1971. The book described more than 60
commands, including b (compile a B program), chdir (change working directory), chmod , chown , ls, and who .
By 1971, many fundamentally important aspects of Unix were already in place, including file ownership and file access
permissions.
The first commercial Unix system was installed in early 1972 at the New York Telephone Co. Systems Development
Center under the direction of Dan Gielan. Neil Groundwater build an Operational Support System in PDP-11/20 assembly
language.
With funding from the Patent Department, the Unix team upgraded to a DEC PDP-11/45.
version 56
page 20
Unix and Linux System Administration and Shell Programming
In 1972, Dennis Ritchie and Ken Thompson rewrote the B programming language into C. Work started on converting
Unix to C. Unix was originally written in assembly language and B, but by 1973 it had been almost completely converted
to the C language. At the time it was common belief that operating systems must be written in assembly in order to perform
at reasonable speeds. Writing Unix in C allowed for easy portability to new hardware, which in turn led to the Unix
operating system being used on a wide variety of computers.
Ken Thompson invented the pipe in 1972. Pipes allow the output of one process to be used as the input for another. Pipes
allow the Unix philosophy of having many small programs that each do one function very well and then write scripts that
combine these small utility programs to accomplish bigger tasks.
By 1973, Unix was running at sixteen (16) sites, all within AT&T and Western Electric. Unix was presented to the public
in an October 1973 conference and within six months the number of sites running Unix had tripled.
A version of Unix was published in the July 1974 issue of the Coomunications of the ACM, leading to widespread
requests for copies. Under the antitrust consent decree with the U.S. government, Bell Labs was required to give a free copy
to anyone who requested it. Ken Thompson personally signed a note for many of these early copies.
Because C was a complete, platform-independent programming language and Unix’s more than 11,000 lines of code
were written in C, it was possible to easily port Unix to any computer platform. And Unix did quickly spread to almost
every available computer platform, becoming the first true universal operating system.
In fall of 1974, Ken Thompson went to the University of California at berkeley to teach for a year. Ken Thompson, Bill
Joy, and Chuck Haley worked together to create the Berkeley version of Unix, called the Berkeley Software Distribution, or
more commonly BSD.
The souce code for BSD Unix was widely distributed to students and other schools. Students at UC-Berkeley and other
schools around the world worked on improvements and the BSD Unix replaced the Bell Laboratories Unix as the primary
Unix.
BSD Unix added the vi editor, the C shell, the Sendmail email system, virtual memory, and support for TCP/IP
(Transmission Control Protocol/Internet Protocol). Because email was built into the operating system, email was the method
Unix used for sending notifications to the system administrator of system status, reports, problems, etc.
The Fifth Edition of Unix was released in 1975 and licensed to universities for free.
The PWB shell or Mashey shell, abbreviated sh, was a variation of the Thompson shell that had been augmented by John
Mashey and others at Bell Labs. The Mashey shell was distributed with the Programmer’s Workbench Unix in 1976.
The Bourne shell, created by Stephen Bourne at AT&T’s Bell Labs (in New Jersey) as a scripting language, was released
in 1977 as the default shell in the Version 7 Unix release distributed to colleges and universities. It gained popularity with
the publication of The UNIX Programming Environment by Brian W. Kernighan and Rob Pike. The book was the first
commercially published tutorial on shell programming.
The Bourne shell provided process control, variables, regular expressions, flow control, input/output control, and
functions.
In 1977, the University of California, Berkeley, released the Berkeley Software Distribution (BSD) version of Unix,
based on the 6th edition of AT&T Unix.
The C shell, abbreviated csh , was created by Bill Joy, a graduate student at the University of California, Berkeley. With
additional work by Michael Ubell, Eric Allman, Mike O’Brien, and Jim Kulp, it was released in the 2BSD release of BSD
Unix in 1978.
The C shell offered command history, aliases, file name completion, and job control.
It turned out that while the C shell was easier to use than the Bourne shell, it failed miserably as a scripting language.It
version 56
page 21
Unix and Linux System Administration and Shell Programming
became common to use the C shell for everyday interactive computing and to use the Bourne shell for script programming.
The improved C shell, abbreviated tcsh , was created by Ken Greer by September 1975 and merged into the C shell in
December 1983. Ken Greer based the shell on the TENEX operating system (hence, the “t” in tsch ). Mike Ellis, Paul
Placeway, and Christos Zoulas made major contributions to tcsh .
The release of the Seventh Edition of Unix in 1978 led to the divergence of Unix on the System V (Bell Labs) and BSD
(Berkeley) paths. System V was often called sys-five. System V was the for pay version and BSD was the free open source
version.
The Korn shell, abbreviated ksh , was created by David Korn at AT&T’s Bell Labs and announced at USENIX on July
14, 1983. Mike Veach and Pat Sullivan were also early contributors. The Korn shell added C shell features to the Bourne
shell (command history and history substitution, command aliases, and file name completion). The Korn shell also added
arrays and built-in integer arithmetic.
When the AT&T broke up in 1984 into “Baby Bells” (the regional companies operating local phone service) and the
central company (which had the long distance business and Bell Labs), the U.S. government allowed them to start selling
computers and computer software.
AT&T gave academia a specific deadline to stop using “encumbered code” (that is, any of AT&T’s source code
anywhere in their versions of Unix). This led to the development of free open source projects such as FreeBSD, NetBSD,
and OpenBSD, as well as commercial operating systems based on the BSD code.
Meanwhile, AT&T developed its own version of Unix, called System V. Although AT&T eventually sold off Unix, this
also spawned a group of commercial operating systems known as Sys V Unixes.
Unix quickly swept through the commercial world, pushing aside almost all proprietary mainframe operating systems.
Only IBM’s MVS and DEC’s OpenVMS survived the Unix onslaught.
Some of the famous official Unix versions include Solaris, HP-UX, Sequent, AIX, and Darwin. Darwin is the Unix
kernel for Apple’s OS X, AppleTV, and iOS (used in the iPhone, iPad, and iPod).
The BSD variant started at the University of California, Berkeley, and includes FreeBSD, NetBSD, OpenBSD, and
DragonFly BSD.
The original Sun OS was based on BSD, but Solaris V5 was based on System V, release V.
Most modern versions of Unix combine ideas and elements from both Sys-V and BSD. HP-UX and Solaris are primarily
System V, while AIX and Mac OS X are hybrids of both flavors of Unix.
Other Unix-like operating systems include MINIX and Linux.
In 1983, Richard Stallman founded the GNU project, which eventually provided the tools to go with Linux for a
complete operating system.
In 1986, Maurice J. Bach of AT&T Bell Labs published The Design of the UNIX Operating System, which described the
System V Release 2 kernel, as well as some new features from release 3 and BSD.
In 1987, Andrew S. Tanenbaum released MINIX, a simplified version of Unix intended for academic instruction.
bash , which stands for Bourne Again SHell, was created by Brian Fox for the Free Software Foundation and first
released on June 7, 1989. bash combined features from the Bourne shell, the C shell, and the Korn shell. Bash also
introduced name completion for variable names, usernames, host names, commands, and file names, as well as spelling
correction for pathnames in the cd command, arrays of unlimited size, and integer arithmetic in any base between 2 and 64.
bash is now the primary shell in both Linux and Mac OS X.
The Z shell, abbreviated zsh , was written by Paul Flastad in 1990 when he was a student at Princton University.
version 56
page 22
Unix and Linux System Administration and Shell Programming
The Linux operating system was first released on September 17, 1991, by Finnish student Linus Torvalds. With the
permission of Andrew S. Tanenbaum, Linus Torvalds started work with MINIX. There is no MINIX source code left in
Linux. Linux Torvalds wrote Linux using the GNU C compiler running on MINIX on an Intel 80386 processor.
Linus Torvalds started work on open source Linux as a college student. After Mac OS X, Linux is the most widely used
variation of Unix.
Linux is technically just the kernel (innermost part) of the operating system. The outer layers consist of GNU tools. GNU
was started to guarantee a free and open version of Unix and all of the major tools required to run Unix.
In 1992 Unix System Laboratories sued Berkeley Software Design, Inc and the Regents of the University of California to
try to stop the distribution of BSD Unix. The case was settled out of court in 1993 after the judge expressed doubt over the
validity of USL’s intellectual property.
See the appendix for a more detailed summary of the history of computers.
version 56
page 23
Unix and Linux System Administration and Shell Programming
choice of shells
chapter 4
summary
This chapter looks at the shells available for Linux and Unix.
This includes an explanation of what a shell is and a list of the popular shells.
what is a shell?
Unix (and therefore also Mac OS X and Linux) has more than 200 basic commands (also called tools or utilities) that are
distributed with the standard operating system. This collection and the ease with which they work together is the major
source of the power of Unix. The vast majority of these standard tools are designed to be used from a command line (the
shell).
A shell is primarily a command interpreter.
In a graphical user interface such as Macintosh or Windows, the user controls the computer and programs primarily
through pointing and clicking, supplemented by some typing.
In a shell, all of the commands are typed. The shell figures out the meaning of what you typed and then has the computer
do as instructed.
But the shell is much more than just a command interpreter. It is also a complete programming language.
Because the shell is a complete programming language, with sequences, decisions, loops, and functions, it can do things
well beyond pointing and clicking. It can take control of your computer and react to changing circumstances.
Programming languages are used to make all of the computer programs and smart phone apps you’ve ever seen or used.
Your imagination is the only limit on the power of the shell. Anything that can be done with a computer can be done with
the shell.
version 56
page 24
Unix and Linux System Administration and Shell Programming
The term shell is derived from the idea of concentric layers that occur in many seashells. One view is that programs can
be run inside of the shell, including other copies of the shell. These layers can be arbitrarily deep. An inverse view is that
the kernel is the inner most layer, with the shell wrapped around the kernel and application programs wrapped around the
shell.
choice of shells
Most versions of Unix and Linux offer a choice of shells.
The term shell comes from a concept of Unix being a series of layers of software that work together rather than one
single monolithic operating system program. The command line shell was part of the outer shell of a series of programs that
eventually reach down to the kernel or innermost portion of the operating system.
Steve Bourne created the original Bourne shell called sh.
Some other famous shells were C Shell (csh), Korn Shell (ksh), Turbo C Shell (tsch), and the Z Shell (zsh). A shell that
combined the major advntages of each of these early shells was the Bourne Again SHell (BASH).
The default shell in Linux and MacOS X is normally BASH (Bourne Again SHell). The following instruction assumes
the use of BASH.
It is possible to change the shell you are currently using (especially useful if you need to run a line or script from an
alternative shell).
original shell
The original shell, called the Thompson shell and abbreviated sh was written by Ken Thompson at AT&T’s Bell Labs,
was very rudimentary. The Thompson shell was distributed with Versions 1 through 6 of Unix, from 1971 to 1975.
Note that the same sh is also used for other shells, including the Bourne shell.
The PWB shell or Mashey shell, abbreviated sh, was a variation of the Thompson shell that had been augmented by John
Mashey and others at Bell Labs. The Mashey shell was distributed with the Prorgammer’s Workbench Unix in 1976.
version 56
page 25
Unix and Linux System Administration and Shell Programming
Bourne shell
Stephen Bourne created the Bourne shell, abbreviated sh, at AT&T’s Bell Labs in New Jersey as a scripting language.
The Bourne shell was released in 1977 as the default shell in the Version 7 Unix release distributed to colleges and
universities. It gained popularity with the publication of The UNIX Programming Environment by Brian W. Kernighan and
Rob Pike. The book was the first commercially published tutorial on shell programming.
The Bourne shell (or a shell that is compatible and will run Bourne shell scripts) is usually located at /bin/sh.
You will also see references to the Bourne shell as bsh .
The Bourne shell has a command language based on the ALGOL programming language, including process control,
variables, regular expressions, flow control (decisions and loops), powerful input and output control, and functions.
Despite its power, the original Bourne shell was difficult to use and lacked file name completion, command history,
aliases, and had difficulty running multiple background jobs.
C shell
The C shell, abbreviated csh , was created by Bill Joy, a graduate student at the University of California, Berkeley. With
additional work by Michael Ubell, Eric Allman, Mike O’Brien, and Jim Kulp, it was released in the 2BSD release of BSD
Unix in 1978.
The C shell was based on the style of the C programming language.
The C shell was much easier to use than the Bourne shell, but was unable to perform anything other than the most simple
scripts.
The C shell lacked functions, had weak input and output control, and had limited programming ability due to a poorly
written command interpreter (Bill Joy explained that at the time he didn’t know about many standard compiler techniques).
The power of the C shell for everyday interactive use was due to its introduction of command history, aliases, file name
completion, and job control.
It became common to use the C shell for everyday interactive computing and to use the Bourne shell for script
programming.
The improved C shell or TENEX C shell, abbreviated tcsh , was created by Ken Greer by September 1975 and merged
into the C shell in December 1983. Ken Greer based the shell on the TENEX operating system (hence, the “t” in tsch ).
Mike Ellis, Paul Placeway, and Christos Zoulas made major contributions to tcsh .
The Hamilton C Shell was written in 1988 by Nicole Hamilton of Hamilton Laboratories for the OS/2 operating system.
In 1992, the Hamilton C Shell was released for Windows NY. The OS/2 version was discontinued in 2003. The Windows
version, with adaptations for Windows peculiarities, is still in use.
Korn shell
The Korn shell, abbreviated ksh , was created by David Korn at AT&T’s Bell Labs and announced at USENIX on July
14, 1983. Mike Veach and Pat Sullivan were also early contributors.
After years of Unix users having to learn two different shells, one for scripting and oen for interactive control, David
Korn built the Korn shell by adding C shell features to the Bourne shell.
In addition to adding command history, history substitution, aliases, and file name completion from the C shell, the Korn
version 56
page 26
Unix and Linux System Administration and Shell Programming
shell also included arrays and built-in integer arithmetic.
The three major versions of the Korn shell are the official version ksh , the public domain version pdksh , and the desktop
version dtksh
CDE shipped with the Desktop version of the Korn shell.
The POSIX shell is a variant of the Korn Shell. HP shipped the POSIX shell with HP-UX 11.0
BASH shell
bash , which stands for Bourne Again SHell, was created by Brian Fox for the Free Software Foundation and first
released on June 7, 1989.
bash combined features from the Bourne shell, the C shell (both the original C shell and tcsh), and the Korn shell, while
using the syntax of the original Bourne shell, allowing it to run almost all historical Bourne shell scripts. bash is now the
primary shell in both Linux and Mac OS X.
tt>bash introduced name completion for variable names, usernames, host names, and commands (in addition to file
names), spelling correction for pathnames in the cd command, arrays of unlimited size, and integer arithmetic in any base
from 2 to 64.
Z shell
The Z shell, abbreviated zsh , was written by Paul Flastad in 1990 when he was a student at Princton University.
The Z shell combines features of the Bourne Shell, the C shell, and bash .
other shells
The POSIX shell, abbrevaited sh, (POSIX = Portable System Interface) is based on the Korn shell and is the official
IEEE P1003.2 shell. It is the /bin/sh shell on some systems.
The Almquist shell, abbreviated ash , was written by Keneth Almquist as a replacement for the Bourne shell written for
BSD systems. The sh in FreeBSD and NetBSD were based on the Almquist shell.
The Debian Almquist shell, abbreviated dash , is a modern replacement for the Almquist shell in Debian and Ubuntu
Linux. While the default shell in both Debian and Ubuntu Linux are bash , /bin/sh is a symlink to ash .
ash and dash are minimalist shells that interpret scripts and are not intended for interactive use. Their small size and fast
execution time make them good for embedded systems. Because many scripts call for the use of /bin/sh, Debian and
Ubuntu Unix gain speed of execution for scripts, while still allowing bash as the primary interactive shell.
The Public domain Korn shell, abbreviated pdksh , is based on the Korn shell.
The MirBSD Korm shell, abbreviated mksh , is based on the OpenBSD version of the Korn shell and the pdksh and was
released as part of MirOS BSD.
Busybox is a set of tiny utilities, including a version of ash , used in embedded systems.
rc, written by Tom Duff, was the default shell for AT&T’s Bell Labs’ Plan 9 and Version 10 Unix. Plan 9 was an
improved operating system based on Unix, but was not a significant enough improvement to overtake its predecessor. rc is
available on some modern versions of Unix and Unix-like operating systems.
es is an rc-compatible shell written in the mid-1990s for functional programming.
version 56
page 27
Unix and Linux System Administration and Shell Programming
Perl shell, abbreviated psh , is a shell that combines bash and the Perl scripting language. it is available for Unix-like and
Windows operating systems.
Friendly interactive shell, abbreviated fish , was released in 2005.
pysh is a profile in the IPython project that combines a Unix-style shell with a Python scripting language shell.
wish is a windowing shell for Tcl.Tk.
Which Shell to Use
from Google Shell Style Guide
Bash is the only shell scripting language permitted for executables.
Executables must start with #!/bin/bash and a minimum number of flags. Use set to set shell
options so that calling your script as bash <script_name> does not break its functionality.
Restricting all executable shell scripts to bash gives us a consistent shell language that’s installed on all
our machines.
The only exception to this is where you’re forced to by whatever you’re coding for. One example of this
is Solaris SVR4 packages which require plain Bourne shell for any scripts.
version 56
page 28
Unix and Linux System Administration and Shell Programming
connecting to a shell
chapter 5
This chapter looks at how to connect to a shell.
On your local desktop or laptop computer, many of these steps won’t apply, but you will probably need most or all of the
steps when logging in remotely to control a server.
If you are connecting to a remote computer (such as web server), then you can use telnet or SSH .
If you connecting to the shell from a modern graphic user interface, then you need to start up a terminal emulator.
If you connecting to a mainframe computer, minicomputer, or the equivalent, you use any terminal connected to the
system.
NOTE: On a single-user Linux or Mac OS X desktop/laptop computer, you can start the Terminal
program and be automatically signed in (no need for account login or password). On a remote server, you will
always need to login in with at least account name and password.
physical terminal
For old style systems where Unix runs on a mainframe or minicomputer, there will be actual physical terminals that are
connected by wires or MODEM to the mainframe or minicomputer. These systems are increasingly rare.
The terminal may already be at the login prompt. If not, try pressing the RETURN key several times until the login
prompt shows up.
Telnet and SSH
telnet or SSH are the two basic methods for connecting to a remote server.
telnet was the original method. There are telnet client programs available for most computer systems. Unfortunately,
telnet performs all communications in plain text, including passwords, which means that anyone malicious along the path
between you and your server can easily read the information needed to hack into your system using your account. For this
reason, many servers no longer allow telnet access.
SSH is Secure Shell. It has the same basic capabilities as the older telnet, but includes security, including encrypted
communications. There are SSH client programs for most modern computer systems.
You can also use SSH through a terminal emulator, which is how many system administrators now access their servers
from their personal or portable computer.
terminal emulator
As mentioned in the history chapter, the Unix shells were accessed through interactive terminals, originally teletype
machines and later special combinations of keyboards and cathode ray tubes (CRTs).
There are programs for graphic user interfaces that allow you to interact with a shell as if you were using one of those
ancient terminal devices. These are called terminal emulators (they emulate the signals of a terminal).
Some common terminal emulators include eterm, gnome-terminal, konsole, kvt, nxterm, rxvt, terminal, and xterm.
Note that if you are logging in to a web server or other remote computer, you should use an SSH (Secure SHell) client
version 56
page 29
Unix and Linux System Administration and Shell Programming
program instead of a terminal emulator. A terminal emulator program is for gaining shell access to a personal computer or
workstation. With a little bit more knowledge, you can use the shell on a personal computer or workstation as your SSH
client.
special settings
If you are starting up Terminal from a modern operating system, you will almost always be able to start using the shell
without any login name or password.
If you are logging in to a remote server you may need a few pieces of additional information beyond just your login
name and password.
Terminfo: You may need to know the Terminfo name for your terminal. The Terminfo name is used to set the
characteristics of your terminal. A common default is the old DEC VT-100 terminal. On old (early) versions of Unix you
may have to use a Termcap name instead of a Terminfo name.
End of line: On most modern systems, the RETURN key will indicate the end of line. On some older systems you may
find a RET, NEWLINE, ENTER< or other key that serves this purpose. On some systems, that have both a RETURN and
an ENTER key, both will serve the same purpose.
Erase key: On most modern systems, the DELETE key will erase one character at a time. Typically the CONTROL-H
combination (hold down both the H key and the CONTROL key, sometimes marked CNTRL or CTRL) will also work as
an erase key. On some older systems, the # key will also serve as the erase key.
Line kill key: On some older systems the @ key deletes the entire line being entered.
Interrupt key: There is typically a key that will interrupt the execution of running programs.
You can find out these items from the tech support for your web hosting, from the system administrator of a system, or
from the manual for your system.
version 56
page 30
Unix and Linux System Administration and Shell Programming
shell basics
chapter 6
summary
This chapter looks at Unix (and Linux) shell basics.
This includes an explanation of what a shell is, how to get to your shell, and a simple example of how to use your shell.
The short version of this chapter is to start up your terminal emulator (or SSH client), type in your account name at the
login prompt, followed by your password at the password prompt (using the RETURN or ENTER button at the end of
each line). Then type a valid command (such as who ) and an invalid command (such as asdf ) and see the results. The rest
of the commentary helps you through common questions and problems.
book conventions
The following conventions are used in this book.
Examples are displayed in a gray box. In the PDF version of this book the gray box does not appear, but the box is
indented from the regular text. In the on-line version at www.osdata.com the gray box displays correctly. I need to find
someone with a legal copy of Adobe Acrobat to help out. If you are in the Costa Mesa, California, area and can help out
with a legal copy of Adobe Acrobat, please contact me.
Additionally, anything that will appear on the screen (whether you type it or the shell produces it as output) will appear
This will help identify the examples in the PDF version. This also applies to material contained in
ordinary descriptive paragraphs that you will type.
in a monospace type .
$ date
Wed Nov 10 18:08:33 PST 2010
$
Bold face type is used to indicate items that you type.
There will be no indication of the correct locations to type RETURN or ENTER or other similar special characters. These
normally happen at the end of every input line. There will be reminders in the descriptive text near the beginning of the
book, as well as in other descriptive text where there is a need for a special reminder or an unusual convention or an
unusual keystroke.
$ date
Bold italics type is used to indicate items that you must customize, such as file names or user account name. In a
descriptive paragraph, these items will be in ordinary italics.
$ AccountName
Material that the shell will output will be in ordinary monospace type . In some cases where the material will be variable,
the shell output will be in italics.
version 56
page 31
Unix and Linux System Administration and Shell Programming
Wed Nov 10 18:08:33 PST 2010
VariableInformation
$
root or superuser
Do not run your terminal emulator or shell in root or superuser.
Root or superuser has complete control over the entire computer. A regular user account has limitations.
Experienced programmers run BASH as a regular user to protect themselves from destructive mistakes — and they are
already skilled at using BASH.
A beginner running BASH as superuser or root can turn a simple typing mistake into a disaster.
If you already have a regular user account created, make sure you use it. If you only have the superuser or root account
running, immediately create a regular user account and switch to it for all of your learning exercises.
Both the Bourne style shells (including BASH) and the C style shells display a # prompt when you are running the shell
as root. Be careful.
#
starting your shell
NOTE: On a single-user Linux or Mac OS X desktop/laptop computer, you can start the Terminal
program and be automatically signed in (no need for account login or password). On a remote server, you will
always need to login in with at least account name and password.
Some information you may need includes the name or URL of your server or computer (sometimes called the host), your
account name, and your assigned password. if you are working on a remote server (such as a web server), you need to get
this information from your hosting company or system administrator. If you are using a large multi-user system, you need
to get this information from your system administrator. If you are using a personal computer or workstation, you probably
set these for yourself.
If you have just purchased a new personal computer or workstation, ask your salesperson for the defaults or look them up
in the manual. Your install CD or DVD is likely to have a utility that can be used to create accounts and set passwords. The
Mac OS X install DVD has utilities for both purposes.
If you are the first person to work with a large system or computer, refer to the manuals that came with your computer.
You will find information on the root/superuser account. That is a powerful account and should be reserved for special
occasions when that kind of power is needed. Look in the manual section with a title similar to “Getting Started” and then
look for an account named user, guest, tutor, or similar name that suggests an ordinary user account.
You login remotely into a web server through SHH (Secure SHell). There are many SSH programs available. You do not
have to have a matching operating system. You can use an SSH client program on Windows, Macintosh, Linux, or almost
any other operating system to sign into a Linux or UNIX server.
You login into a large multi-user traditional command line UNIX system by typing your user name, ENTER, your
password, and ENTER (in that order). This will enter you into the default shell set up for your account.
On a computer that runs a graphic interface (such as Gnome, KDE, and Mac OS X) you will want to use a terminal
emulator program. Typically a terminal emulator program starts with you already logged-in to the same account that you
version 56
page 32
Unix and Linux System Administration and Shell Programming
have running in the graphic user interface.
On Mac OS X and most versions of Linux, the icon for the Terminal program will look like this:
On a Mac OS X computer, Terminal will be located in the Utilities directory, which is located in the Applications
directory. When you find it, you can drag it onto the Dock to make it easily accessible. Apple purposely hid the UNIX
command line to avoid confusing the typical consumer.
On Linux/KDE, look for Konsole or Terminal in the Utilities menu.
version 56
page 33
Unix and Linux System Administration and Shell Programming
On Linux/Gnome, look for color xterm, regular xterm, or gnome-terminal in the Utilities menu.
On Ubuntu Linux, look in Applications menu > Accessories > Terminal. Single click and hold, then drag to your panel,
because you will be using it a lot.
On some versions of Linux, you can press the CONTROL and the ALT and the F1 keys all at once to bring up Terminal.
In Gnome, use the Applications menu > Accessories > Terminal or the keyboard shortcut of CONTROL and ALT and T
(all at the same time).
In KDE, use the KMenu > System > Terminal Program (Konsole).
In Linux Mint you can use the keyboard shortcut of holding down the CONTROL and ALT and T keys all at once.
As a last resort, you can use the file search feature in Linux or Mac OS X and search for “Terminal”.
You may want to try out the various terminal emulators and decide which works best for you (each offer slightly
different features).
version 56
page 34
Unix and Linux System Administration and Shell Programming
login and password
NOTE: On a single-user Linux or Mac OS X desktop/laptop computer, you can start the Terminal
program and be automatically signed in (no need for account login or password). On a remote server, you will
always need to login in with at least account name and password.
At the login prompt type your account name (in lower case letters) followed by ENTER or RETURN.
login: accountname
Password:
At the password prompt type your password followed by ENTER or RETURN.
Password: password
$
If you encounter problems, see the next chapter for help.
prompt
Once the shell is running it will present a shell prompt. This shell prompt is normally the U.S. dollar sign ( $ ). Other
common prompts are the percent sign ( % ) and the pound sign ( # ). The shell prompt lets you know that the shell is ready
and waiting for your input.
$
version 56
page 35
Unix and Linux System Administration and Shell Programming
The most common prompt in the Bourne shell ( sh or bsh ) and Bourne Again shell ( bash ) and Korn shell ( ksh ) is the
U.S. dollar sign ( $ ).
$
When the the Bourne Again shell ( bash ) or Korn shell ( ksh ) is running as root or superuser, the prompt is changed to
the U.S. number or pound sign ( # ) as a clear reminder that you are using a much more powerful account.
#
The Bourne Again shell ( bash ) shell typically reports the current working directory and user name before the $ prompt,
often inside of square braces. An example:
[admin:~ admin]$
The most common prompt in the C shell ( csh ) and TENEX C shell ( tsch ) is the percent sign ( % ).
%
When running as root or superuser, the C shell ( csh ) shows the U.S. number or pound sign sign ( # ).
#
The Z shell ( zsh ) shows the short machine name followed by the percent sign ( % ).
mac%
When running as root or superuser, the Z shell ( zsh ) shows the short machine name followed by the U.S. number or
pound sign sign ( # ).
mac#
Almost all of the shell examples in this book will use bold face to show what you type and plain fixed space characters
to show shell output. Items that you need to modify according to your local needs are normally going to be presented in
italics.
You are likely to see information before the actual command prompt. It may the current working directory. It may be the
current user name. There are ways to change what is presented. For now, don’t worry about it, just be aware that you are
likely to see some text before the prompt.
example command
You can run a UNIX (or Linux or Mac OS X) command (also called a tool) by typing its name and then the ENTER or
version 56
page 36
Unix and Linux System Administration and Shell Programming
RETURN key.
The following example uses the date command or tool.
$ date
Wed Nov 10 18:08:33 PST 2010
$
The format for the output is: day of the week, month, day of the month, 24 hour time, time zone, year.
failed command
BASH will let you know when you have typed something it doesn’t understand.
Type “asdf” and then the ENTER or RETURN key. You should see a message similar to the following:
$ asdf
-bash: asdf: command not found
$
no news is good news
BASH follows the rule that no news is good news. If a command runs successfully, it won’t generate any output
message.
As an example, when you use the cp command to coy a file, there will be no special notice on success. You will just be
returned to the command prompt ( $), ready for the next command.
version 56
page 37
Unix and Linux System Administration and Shell Programming
login/logout
chapter 7
summary
This chapter looks at login and logout, a pair of Unix (and Linux) commands.
On a personal computer or workstation you probably don’t need to login . You can just start your terminal emulator (see
previous chapter).
On a multi-user UNIX or Linux system, such as a web server, you will need to login . This chapter furthers the
discussion login in more detail than the previous chapter’s introduction, including special cases and solutions to common
problems.
In particular, this chapter discusses choosing the system and terminal type, rare options that you might encounter.
This chapter also discusses the logout command and the importance of always logging out from a running system.
NOTE: On a single-user Linux or Mac OS X desktop/laptop computer, you can start the Terminal
program and be automatically signed in (no need for account login or password). On a remote server, you will
always need to login in with at least account name and password.
login
login is a UNIX command for logging into a UNIX system.
login is a builtin command in csh . There is also an external utility with the same name and functionality.
In many modern systems, the functionality of login is hidden from the user. The login command runs automatically upon
connection. In some old computers you may need to type a special character or even type login to bring up the login
program.
If you login remotely to a server you will be prompted for your user/account name and password. If you start up a
terminal emulator and shell from a graphic user interface you will probably already be authenticated and immediately go to
the shell (although some of the startup activity of login will still be done for you).
The activities that login performs at startup of a shell will be covered in a later chapter, because you need more
familiarity with the shell before you can understand that material.
Starting a shell is covered in the previous chapter about shell basics.
select system or host
If you are signing into a system that combines multiple computers, you will first need to indicate which computer you are
logging into.
You may already be logged in if you are using a terminal emulator on a personal computer or workstation. If so, you
don’t need this step.
You will see a prompt indicating a choice of computer, such as (only one of these will appear):
version 56
page 38
Unix and Linux System Administration and Shell Programming
host:
pad:
request:
system:
You will see only one of these (or similar) choices.
If you see login: instead, then you don’t have to worry about choosing the system.
Enter the identifying name of the computer system you were assigned to use, followed by the ENTER or RETURN key.
Your system administrator can provide you with this information.
On some networked systems you can abbreviate with the first two or three characters of the computer name (as long as it
is enough to uniquely identify which computer is intended).
If you are successful, you will be presented with the login: prompt.
account name
You may already be logged in if you are using a terminal emulator on a personal computer or workstation. If so, you
don’t need this step.
Ask your system administrator for your account name. If this is your own personal system, then use the account name set
up during installation. Many modern install disks include a utility for creating a new account.
At the login: prompt, enter your account name, followed by the ENTER or RETURN key.
Type your account name in lower case letters. UNIX assumes that an account name in ALL CAPITAL LETTERS
indicates an old input/out device that doesn’t support lower case letters (such as an old style teletype, or TTY). Also, if you
use upper case letters for the login, then you can not use lower case letters in the password.
If you enter a valid account name, you will be asked for your password. On many computers, you will be asked for a
password even if you enter an incorrect account name. This is a security measure to help prevent guessing account names.
login: accountname
Password:
password
You may already be logged in if you are using a terminal emulator on a personal computer or workstation. If so, you
don’t need this step.
Ask your system administrator for your password. If this is your own personal system, then use the password set up
during installation. Many modern install disks include a utility for changing the password for an existing account or for
creating a new account with a new password.
At the Password: prompt, enter your password, followed by the ENTER or RETURN key.
Password: password
$
In some cases, a new account doesn’t have a password. Also, it is common for Mac OS X to have no password for the
user account.
version 56
page 39
Unix and Linux System Administration and Shell Programming
If your account doesn’t have a password, just press the ENTER or RETURN key.
Change your password to a secure password at your earliest opportunity. If you don’t know how to do this yet, the
upcoming chapter on passwd (yes, that is the correct spelling) will tell you how. That same chapter gives advice on how to
select a secure password.
Failure will be indicated by an error message. The exact error message varies by system, but will probably be something
similar to:
Login incorrect
The most likely reasons are:
1. Incorrect account name.
2. Incorrect password.
3. You typed the account name in upper case letters (check the CAPS LOCK key).
You should have a new login prompt and can easily start over.
If you logged in in upper case, the shell will display in upper case letters and expect you to type only in upper case
letters.
If you accidentally typed the account name in upper case letters, but haven’t yet typed in the password, many UNIX
systems will display aa backslash character in front of the password prompt.
If you realize you made a mistake on the account name and want to start over, hold down the CONTROL key (often
marked CTRL) and the D key at the same time. This will stop the login program and start over again from the beginning.
Note that at one time there was a security hole because of the Control-D. A hacker would start the login, type Control-D,
and then try to slip in a command to the shell before a new login program could be started. This security hole no longer
exists on shell login, but this same type of timing attack is still used for other security holes.
terminal type
On some older UNIX systems you may be asked to provide the terminal type you are using. Some modern terminal
emulator programs offer the choice of emulating various different terminals from the early UNIX era. These different
terminal types had different specialized capabilities.
Because the early traditional UNIX was created primarily on Digital Equipment Corporation (DEC) computers (first the
PDP, followed by the VAX), the default terminal device was the DEC VT100. Most modern terminal emulator programs act
like the VT100.
On older UNIX systems you may see the terminal type prompt:
Term:
In some cases, the prompt may give the default terminal setting in parenthesis.
Term = (VT100)
If the default terminal setting is correct, simply press the RETURN or ENTER key.
version 56
page 40
Unix and Linux System Administration and Shell Programming
Otherwise, type in the correct terminal type, followed by the ENTER or RETURN key.
If you are uncertain, try the vt100 or UNKNOWN, followed by ENTER or RETURN.
Some common terminal types:
h19
tvi925
vt100
wyse50
After you are logged in, you can set the terminal type (capitalization matters).
Shell
csh
Command
or tcsh
setenv TERM vt100
sh
TERM=vt100; export TERM
ksh , bash ,
or zsh export TERM=vt100
VMS
set term/device=vt100
If you do not know your shell type, use the following command (followed by ENTER or RETURN):
$ echo $SHELL
You will need to enter the appropriate command each time you login.
Alternatively, you can add the appropriate command line to the initialization file in your home directory.
Shell
Login file
csh
.cshrc
tcsh
.cshrc
ksh
.profile
.kshrc
zsh
.zshrc
bash
.bash_profile
remote shell
.rhosts
or .login
VMS users need to modify the login.com file, adding the line $set term/device=vt100 .
logout
When you finish using the shell, type the exit or logout command. This will either display a new login prompt or close
the terminal emulator window.
When you finish using the shell, type the exit or logout command. This will either display a new login prompt or close
the terminal emulator window. if you are logged in remotely to a server, this will break your connection. If you are logged
in physically on a large UNIX system, this will prevent someone else from abusing your account.
Some of the possible variations of this command include: bye , exit , lo, logout, and quit .
version 56
page 41
Unix and Linux System Administration and Shell Programming
Typing CtrlL-D (holding down the CONTROL key and the D key at the same time) will also log you out of most shells.
The Ctrl-D tells the shell it has reached the end-of-file (EOF). Because the shell is a filter and filters terminate when EOF
is reached, the shell terminates and you log off the system or return to the parent process.
While it might not be a big deal to skip this step on a single user computer that you never leave unattended, forgetting to
logout from a remote server is a serious security hole. If you leave your computer or terminal unattended, someone else can
sit down and gain access to all of your files, including the ability to read, modify, or delete them. The imposter can send
spam emails from your user ID. The imposter can even use your computer or account to attempt to hack or damage or
break into any system in the world from your user ID. If this occurs from a school or business account, you will be
responsible for any damage done.
Most modern systems accept either logout or exit .
On Mac OS X, typing logout produces the following result:
$ logout
[Process completed]
On Mac OS X, typing exit produces the following result (notice that the shell automatically runs logout for you):
$ exit
logout
[Process completed]
A successful logout will stop all running foreground or background processes. There is a way to keep processes running,
which is useful for things like starting up your Apache web server and MySQL data base, but the method will be discussed
later.
$ logout
There are suspended jobs.
If you see the message “There are stopped jobs.” it means that you have one or more suspended jobs. The shell is letting
you know in case you forgot about them. If you type logout again without running the jobs in between, the system will go
ahead and log you out and terminate all suspended jobs.
logout is a builtin command in csh .
exit
exit is a built-in shell command (for the shells that support it), while logout is a program. In the csh (C Shell), logout
is an internal built-in command. sh (the original Bourne shell) and ksh (the Korn Shell) do not officially have the logout
command, but many modern versions support it.
exit can be used by itself as a substitute for logout (as mentioned above).
exit can also be used to send a return value
exit . This is particularly useful in scripting.
(often an error code) to the parent process. Simply add an integer following
$ exit n
version 56
page 42
Unix and Linux System Administration and Shell Programming
exit will be covered later in its own chapter.
version 56
page 43
Unix and Linux System Administration and Shell Programming
passwd
chapter 8
summary
This chapter looks at passwd, a Unix (and Linux) command.
passwd is used to change your password.
This chapter also includes the list of the 100 worst (most commonly used) passwords.
As mentioned in the previous chapter, you should change your password from the original default or assigned password
to a secure password that nobody else knows. And please don’t leave the new password on a sticky note attached to your
computer.
syntax:
passwd [option…] [user]
shells:
ash
File Name: passwd
bash
bsh
csh
ksh
Directory: /usr/bin/
sh
tcsh
Type:
zsh
External
setting your password
Type passwd followed by the ENTER or RETURN key.
$ passwd
You will be prompoted to give your current (old) password (to make sure it is really you) and then prompted to enter
your new password twice. For security purposes, the password is typically replaced with asterisks or some other character
so that nobody can read your password over your shoulder. To make sure that you have typed what you thought you typed
you are asked to type the new password twice. The two copies must match before your new password replaces your old
password.
local password
The password set by passwd is your local password. On a single user system, this is probably your only password.
On Mac OS X the use of the passwd may or may not be sufficient to change your password for the entire system. This
depends on which version of Mac OS X you are using. It is best to change your password using the install disc. If you do
not have a copy of the install disc, there are instructions on the internet on how to manually change the password.
On a large system, there may be multiple passwords spread across multiple computers. The passwd command will only
change the password on the one server that you are currently logged into (normally through SSH). You may need to use
yppasswd or a web interface to change your password for the entire system.
You can check for your account or username in /etc/passwd. If it’s not listed there, then don’t use the passwd. Check
with your system administrator.
version 56
page 44
Unix and Linux System Administration and Shell Programming
periodic changes
Whenever you first login into a new system, the first thing you should do is change your password. In particular,
immediately change the initial root password for a new system. Leaving the initial default password is a huge security hole
and hackers do try all of the standard default passwords to see if they can find an easy way into a computer. Even with a
user account, it is common for initial passwords to be generated poorly and be easy for hackers to guess.
Additionally, you want to change your password on a regular basis. It only takes a few months to figure out a password
through brute force attacks. Some systems require that you change your password on a regular basis. Once a month is a
good time period. More often if you suspect that someone saw you typing or there is any other possibility that your
password might have been compromised.
You can set up your account to remind you to change your password on a regular basis. If you are the system
administrator, you can set up these reminders for everyone (and should do so). As system administrator you can even
require that users change their passwords on a regular basis (or they become locked out and have to come to you to beg for
re-entry). As system administrator you can also set up a system that requires (or even suggests) secure passwords.
100 most common passwords
Always avoid the common passwords. These are the most common passwords as of June 2012:
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
password
123456
12345678
1234
qwerty
12345
dragon
pussy
baseball
football
letmein
monkey
696969
abc123
mustang
michael
shadow
master
jennifer
111111
2000
jordan
superman
harley
1234567
trustno1
iloveyou
sunshine
ashley
bailey
passw0rd
123123
654321
version 56
page 45
Unix and Linux System Administration and Shell Programming
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
50.
51.
52.
53.
54.
55.
56.
57.
58.
59.
60.
61.
62.
63.
64.
65.
66.
67.
68.
69.
70.
71.
72.
73.
74.
75.
76.
77.
78.
79.
80.
81.
82.
83.
84.
85.
86.
87.
88.
89.
qazwsx
Football
seinfeld
princess
peanut
ginger
tigger
fuckme
hunter
fuckyou
ranger
buster
thomas
robert
soccer
fuck
batman
test
pass
killer
hockey
babygirl
george
charlie
andrew
michelle
love
jessica
asshole
6969
pepper
lovely
daniel
access
123456789
joshua
maggie
starwars
silver
william
dallas
yankees
666666
hello
amanda
orange
biteme
freedom
computer
sexy
nicole
thunder
heather
hammer
summer
corvette
version 56
page 46
Unix and Linux System Administration and Shell Programming
90.
91.
92.
93.
94.
95.
96.
97.
98.
99.
100.
taylor
fucker
austin
1111
merlin
matthew
121212
golfer
cheese
martin
chelsea
Approximately 4.7% of all users have a password of password. 8.5% have one of the top two passwords. 9.8% (nearly
one tenth) have one of the three top passwords. 14% have one of the top 10 passwords. 40% have one of the top 100
passwords. 79% have one of the top 500 passwords. 91% have one of the top 1,000 passwords.
secure passwords
It is important to have secure passwords.
The more characters, the more secure. A minimum of six or eight characters is barely adequate.
A strong mixture of characters for a password includes at least one capital letter, at least one lower case letter, at least on
digit, and at least one punctuation character. You should avoid repeating any character more than once in the same
password. The special character (such as !@#$%^&*,;) should not be th efirst or last character in the password.
Avoid using any word that occurs in your own or any other natural langauge. Hackers use a dictionary attack that tries
words from the dictionary. Also avoid spelling words backwards, using common misspellings, or using abbreviations.
Avoid using dates that are important to you (someone can easily look up your birthday or anniversary on the world wide
web). Avoid using names of family, friends, or even pets.
secure technique
A technique that generates decent passwords is to use a key phrase and then use the first letter of each word in the
keyword. Sprinkle in digits and special characters (punctuation) and make some of the letters upper case and some lower
case.
Never use the same password for more than one purpose. People have the tendency to reuse the same password over and
over. If a hacker gets your password from one system, the hacker will see if it also works on your bank account and other
systems.
superuser
The super user (root) can use the passwd command to reset any other user’s password. There is no prompt for the current
(old) password.
$ passwd username
The super user (root) can also remove a password for a specific user with the -d option. The disable option then allows
the specified user to login without a password. This applies to Linux and Solaris only.
$ passwd -d username
version 56
page 47
Unix and Linux System Administration and Shell Programming
version 56
page 48
Unix and Linux System Administration and Shell Programming
command structure
chapter 9
summary
This chapter looks at simple Unix/Linux commands to get you started with using the shell.
You will learn the basic format or structure of a shell command.
You can run a UNIX (or Linux or Mac OS X) command (also called a tool) by typing its name and then the ENTER or
RETURN key.
All commands (and file names) are case senitive. ls is a valid command, while LS is unrecognized.
commands and utilities
A quick note on the difference between commands and utilities.
A utility is a program. who , date , and ls are examples of utility programs.
A command is all of the text typed at the command line, the utility name and all of the flags, switches, file names, and
other text.
This subtle distinction between utilities and commands is generally ignored and the two terms are often used
interchangeably. As I type this, I am pretty sure that I have ignored this distinction elsewhere in this book and dread
double-checking every use. You can probably use the terms interchangeably except in cases like tests, papers, or books.
The kernel is loaded from hard drive (or other storage) when the computer is started and remains running until the
computer is turned-off or crashes.
Utilities are also stored on hard drives, but are only loaded into main memory when used.
single command
The most simple form of a UNIX command is just a single command by itself. Many Unix/Linux commands will do
useful work with just the command by itself. The next examples show two commands ( who and date ) by themselves.
The output generated by a single command by itself is called the default behavior of that command.
who
The who command will tell you all of the users who are currently logged into a computer. This is not particularly
informative on a personal computer where you are the only person using the computer, but it can be useful on a server or a
large computing system.
Type who followed by the ENTER or RETURN key.
$ who
admin console Aug 24 18:47
admin ttys000 Aug 24 20:09
$
version 56
page 49
Unix and Linux System Administration and Shell Programming
The format is the login name of the user, followed by the user’s terminal port, followed by the month, day, and time of
login.
failed command
The shell will let you know when you have typed something it doesn’t understand.
Type “asdf” and then the ENTER or RETURN key. You should see a message similar to the following:
$ asdf
-bash: asdf: command not found
$
date
The following example uses the date command or tool.
$ date
Wed Nov 10 18:08:33 PST 2010
$
The format for the output is: day of the week, month, day of the month, 24 hour time, time zone, year.
options, switches, or flags
A command may optionally be followed by one or more options. The options are also commonly called flags or
switches.
Options are usually a single character. Usually the option is preceded by a minus sign or hyphen character.
See the following example.
universal time
Adding the -u flag to the command date will cause it to report the time and date in UTC (Coordinated Universal) time
(also known as Zulu Time and formerly known as Greenwich Mean Time or GMT). The seconds are slightly higher in the
second example because of the passage of time.
$ date
Sat Aug 25 19:09:19 PDT 2012
$
$ date -u
Sun Aug 26 02:09:27 UTC 2012
$
option grouping
You can combine multiple options for a single command.
The following example shows the ls command by itself and with two options, first individually, and then combined.
version 56
page 50
Unix and Linux System Administration and Shell Programming
$ ls
Desktop Downloads Movies Pictures scripts
Documents Library Music Public
$ ls -F
Desktop/ Downloads/ Movies/ Pictures/ scripts/
Documents/ Library/ Music/ Public/
$ ls -a
. .cups Library scripts
.. .vminfo Movies
.DS_Store Desktop Music
.Trash Documents Pictures
-bash_history Downloads Public
$ ls -a -F
./ .cups/ Library/ scripts
../ .vminfo Movies/
.DS_Store Desktop/ Music/
.Trash/ Documents/ Pictures/
-bash_history Downloads/ Public/
$
You can group option by listing them in any order (with a few rare exceptions) and either with their own minus sign or
after a single minus sign. All of the following versions of the ls command work the same way:
$
$
$
$
$
ls
ls
ls
ls
-a -F
-F -a
-aF
-Fa
arguments
Commands may also optionally have arguments. The most common arguments are the names of files.
Technically, the options just mentioned are also arguments, but in common practice options are separated from other
arguments in discussions. Also, technically, the operators mentiond below are also arguments, but again it is useful to
separate operators from other arguments in discussions.
The following shows the difference between the default behavior (no arguments) of who and a version with arguments
who am i.
The default behavior of who lists all users on the computer.
$ who
admin console Aug 24 18:47
admin ttys000 Aug 24 20:09
$
The use of the arguments with who am i causes the command to only lists the one user who typed the command.
The command is who and the arguments are am i.
$ who am i
admin ttys000 Aug 25 17:30
version 56
page 51
Unix and Linux System Administration and Shell Programming
$
one argument
In an upcoming chapter on the cat you will use the command cat with the file name file01.txt to confirm that you
correctly entered your sample file.
just observe this example
do not type this into your shell
$ cat file01.txt
This is a line of text in my first file.
This is another line.
To be, or not to be: that is the question:
1234567890
ABC
XYZ
abc
xyz
$
You may download this sample file from http://www.osdata.com/programming/shell/file01.txt.
two arguments
In the upcoming chapter on the cat you will use the command cat with two file names ( file01.txt and
numberfile.txt ) to confirm that you correctly entered two of your sample files.
just observe this example
do not type this into your shell
$ cat file01.txt numberfile.txt
This is a line of text in my first file.
This is another line.
To be, or not to be: that is the question:
1234567890
ABC
XYZ
abc
xyz
1
2
3
4
5
6
7
8
9
10
11
12
version 56
page 52
Unix and Linux System Administration and Shell Programming
13
14
15
16
17
18
19
20
$
You may download the second sample file from
http://www.osdata.com/programming/shell/numberfile.txt.
options and arguments
You can combine options and arguments on the same command line.
The following example uses the command cat with the -b option and the file name file01.txt .
just observe this example
do not type this into your shell
$ cat -b file01.txt
1 This is a line of text in my first file.
2 This is another line.
3 To be, or not to be: that is the question:
4 1234567890
5 ABC
6 XYZ
7 abc
8 xyz
$
operators and special characters
Command lines may include optional operators or other special characters.
In an upcoming chapter on the cat you will use the command cat with the operator > and the file name file01.txt to
enter your sample file.
just observe this example
do not type this into your shell
$ cat > file01
If you accidentally typed in the above command, hold down the CONTROL key and the D key at the same time to return
to your shell.
version 56
page 53
Unix and Linux System Administration and Shell Programming
version 56
page 54
Unix and Linux System Administration and Shell Programming
quick tour of shell commands
chapter 10
summary
This chapter gives you a quick tour of shell commands.
You should do each of the exercises and observe what happens.
Do not worry about trying to learn how each of the commands works.
The goal here is to give you a basic idea of how a shell works and the kinds of things that the most common shell
commands and operations do.
This quick tour will give you a background to understand the upcoming chapters and their lessons. If you have the need
to use anything you observe in this quick tour, you can always jump ahead to the appropriate chapter and learn the details.
failure
Type asdf (a non-existent command) into the terminal. Remember to press the ENTER or RETURN key after this
command.
$ asdf
-bash: asdf: command not found
$
This shows you what it look like when the shell reports mistakes. Note that you will not always see any reports of a
particular error. Commands that have no output will just return silently to the prompt.
echo text
Use the echo command to send text to the terminal. Remember to press the ENTER or RETURN key after each
command. Note that this will be particularly useful in your future Unix/Linux scripts to report information back from your
scripts.
$ echo hello world
hello world
$
echo text with variable information
Use the echo command with an environment variable. it should use your account name.
$ echo hello $USER
hello admin
$
version 56
page 55
Unix and Linux System Administration and Shell Programming
list directory contents
Use the ls command to list the contents of the current directory.
$ ls
Desktop Movies Send registration
Documents Music Sites
Downloads Pictures
Library Public
$
create a text file
Use the cat command to create a small text file.
Type the command line, followed by RETURN or ENTER, then type each of the six suggested lines, each followed by
the RETURN KEY. After having enetered each line, make sure you are at the beginning of a new (blank or empty) line and
type Ctrl-D (hold downt he CONTROL key and the D key at the same time).
$ cat > names
James
Mary
John
Patricia
Robert
Linda
CONTROL-D
$
The choice of names is based on the most popular names in the United States in the year before this chapter was written.
Bribe me if you want your name put into the example.
check the file was created
Use ls to make sure the file was created properly. It should be added to your directory listing.
$ ls
Desktop Movies Send registration
Documents Music Sites
Downloads Pictures names
Library Public
$
display file contents
Use the cat command to show the contents of your new file.
$ cat names
James
Mary
John
Patricia
version 56
page 56
Unix and Linux System Administration and Shell Programming
Robert
Linda
$
count the number of words
Use the wc command to count the number of words in your new file.
$ wc names
6 6 38 names
$
The format is the number of lines ( 6), followed by the number of words ( 6), followed by the number of characters ( 38),
followed by the name of the file ( names ).
copy a file
Use the cp command to make a copy of a file.
$ cp names saved_names
$
Notice that there is no confirmation of the file copy being made.
This silent behavior is typical of any UNIX shell. The shell will typically report errors, but remain silent on success.
While disconcerting to those new to UNIX or Linux, you become accustomed to it. The original purpose was to save paper.
When UNIX was first created, the terminals were mostly teletype machines and all output was printed to a roll of paper. It
made sense to conserve on paper use to keep costs down.
You can use the ls command to confirm that the copy really was made. You won’t be using up any paper.
$ ls
Desktop Movies Send registration
Documents Music Sites
Downloads Pictures names
Library Public saved_names
$
rename a file
Use the mv command to rename a file.
$ mv saved_names old_names
$
Notice that the shell is once again silent with success. You can use the ls command to confirm that the rename really
was made.
$ ls
version 56
page 57
Unix and Linux System Administration and Shell Programming
Desktop Movies Send registration
Documents Music Sites
Downloads Pictures names
Library Public old_names
$
delete a file
Use the rm (remove) command to delete a file.
$ rm old_names
$
You can use the ls command to confirm that the file was really deleted.
$ ls
Desktop Movies Send registration
Documents Music Sites
Downloads Pictures names
Library
$
current directory
Use the pwd command to determine the current directory. Your version should give the name of your home directory,
which normally matches the name of your user account. The example include the directory as part of the prompt (your
system may not include this) and the tilde ( ~ ) character indicates the home directory.
$ pwd
/Users/admin
admins-power-mac-g5:~ admin$
make a directory
Use the mkdir command to make a new directory (folder).
$ mkdir testdir
admins-power-mac-g5:~ admin$
change directory
Use the cd command to change to your new directory (folder). if your prompt includes the current directory, then you
will see your prompt change to show the new location.
$ cd testdir
admins-power-mac-g5:testdir admin$
version 56
page 58
Unix and Linux System Administration and Shell Programming
confirm directory change
Use the pwd command to confirm that you are now in your new directory.
$ pwd
/Users/admin/testdir
admins-power-mac-g5:testdir admin$
return to home directory
Use the cd command without any additional arguments to return to your home directory from anywhere.
$ cd
admins-power-mac-g5:~ admin$
confirm directory change
Use the pwd command to confirm that you are now back in your home directory.
$ pwd
/Users/admin
admins-power-mac-g5:~ admin$
Now you are ready to dive in and start becoming proficient at using the UNIX or Linux shell.
version 56
page 59
Unix and Linux System Administration and Shell Programming
man
chapter 11
summary
This chapter looks at man , a Unix (and Linux) command.
man is the UNIX equivalent of a help function.
man is the command used to view manual pages.
This chapter show you how to use the man command to get help. Because of huge variation in the various flavors of
UNIX and Linux, the man installed on your computer or server is the best source of detailed information. If you have
trouble with any of the lessons in this book or in general use of your UNIX or Linux system, always refer to the local man
for help.
UNIX was the first operating system distributed with online documentation. This included man (manual pages for each
command, library component, system call, header file, etc.) and doc (longer documents detailing major subsystems, such as
the C programming language and troff .
example of man command with options
Most BASH commands accept options. These options follow a space character and are typed before you press RETURN
or ENTER.
The format is man is followed by the name of the command or tool that you want to view. You will get the manual pages
for the named command or tool.
The following example uses the manual page for the date command or tool.
$ man date
DATE(1) BSD General Commands Manual DATE(1)
NAME
date -- display or set date and time
SYNOPSIS
date [-ju] [-r seconds] [-v [+|-]val[ymwdHMS]] ... [+output_fmt]
date [-jnu] [[[mm]dd]HH]MM[[cc]yy][.ss]
date [-jnu] -f input_fmt new_date [+output_fmt]
date [-d dst] [-t minutes_west]
DESCRIPTION
When invoked without arguments, the date utility displays the current
date and time. Otherwise, depending on the options specified, date will
set the date and time or print it in a user-defined way.
The date utility displays the date and time read from the kernel clock.
When used to set the date and time, both the kernel clock and the hard ware clock are updated.
Only the superuser may set the date, and if the system securelevel (see
securelevel(8)) is greater than 1, the time may not be changed by more
than 1 second.
:
version 56
page 60
Unix and Linux System Administration and Shell Programming
Typing the RETURN or ENTER key will bring up the next line of the manual page.
Typing the SPACE-BAR key will bring up the next page of text.
Typing the UP-ARROW or DOWN-ARROW key will move up or down one line.
Typing q will end viewing the manual page and return to the BASH prompt (the entire manual page will disappear from
view).
The man command or tool will be a very useful reference.
A man page is typically organized:
NAME: Command name and brief description.
SYNOPSIS: The syntax for using the command, along with the flags (options) the command takes.
DESCRIPTION: Details about the command.
ENVIRONMENT: The environment variables used by the command.
EXIT STATUS: Information about how the command reports errors or success.
FILES: File related to the command.
SEE ALSO: related commands.
STANDARDS: The international standards, if any.
Some systems might have an AUTHOR or other sections.
Type man man for information about the local version of man .
man sections
The first seven distributions of UNIX (called V1 UNIX through V7 UNIX and collectively called traditional UNIX)
included a two volume printed manual, which was divided into eight sections.
Some manual pages are located in sections. There are generally eight sections:
1.
2.
3.
4.
5.
6.
7.
8.
General commands
System [kernel] calls
C library functions
Special files (such as devices) and drivers
File formats, conventions, and miscellaneous information
games and screensavers
Macro packages
System administration commands and daemons
Note that on some systems, the system administration commands are in section 1m. This section is also sometimes called
the Maintenance commands.
Note that on some systems, section 7 is Miscellaneous.
You can look at a particular section by adding the section number to the command line.
version 56
page 61
Unix and Linux System Administration and Shell Programming
$ man SECTION-NUMBER commandname
You can use the whatis command to find the sections and then look at the particular section of your choice.
$ whatis crontab
$ whatis crontab
crontab(1) - maintain crontab files for individual users (V3)
crontab(5) - tables for driving cron
$ man 5 crontab
-f option
The -f option is used to show all man page titles for entries that begin with a particular word. man -f file will show all
the manual pages that have the word “file” in the title.
$ man file
-k option
The -k option runs a search on all manual pages for any manual page with a particular word anywhere on the page. The
man -k file will show all the manual pages that have the word “file” anywhere in the manual page. The -k option can
take a while to run.
$ man file
version 56
page 62
Unix and Linux System Administration and Shell Programming
cat
chapter 12
summary
This chapter looks at cat , a Unix (and Linux) command.
The cat (as in concatenate) utility can be used to concatenate several files into a single file.
cat is most often used to obtain the contents of a file for input into a Linux or UNIX shell script.
cat is used concatenate file. The name is an abbreviation of catenate, a synonym of concatenate.
cat was part of the original 1969 version of UNICS (the original name for UNIX).
create names file
If you did not create the names file suggested in the quick tour chapter, please do so now, because you will use this file
in future exercises. If you already created the names file, then you can safely skip this step.
Type the command line, followed by RETURN or ENTER, then type each of the six suggested lines, each followed by
the RETURN KEY. After having enetered each line, make sure you are at the beginning of a new (blank or empty) line and
type Ctrl-D (hold downt he CONTROL key and the D key at the same time).
$ cat > names
James
Mary
John
Patricia
Robert
Linda
CONTROL-D
$
The choice of names is based on the most popular names in the United States in the year before this chapter was written.
Bribe me if you want your name put into the example.
check the file was created
Use ls to make sure the file was created properly. It should be added to your directory listing.
$ ls
Desktop Movies Send registration
Documents Music Sites
Downloads Pictures names
Library Public
$
display file contents
version 56
page 63
Unix and Linux System Administration and Shell Programming
Use the cat command to show the contents of your new file.
$ cat names
James
Mary
John
Patricia
Robert
Linda
$
creating files
One simple use of cat is to create simple files.
Type cat > file01.txt , followed by ENTER or RETURN.
$ cat > file01.txt
The cursor will be at the beginning of a line that has no prompt. You are no longer in the shell, but instead in the cat
utility. There is no cat prompt.
Type the following lines (by convention, all of the input should be in bold, but to make it easier on the eye, it is in italics
here):
This is a line of text in my first file.
This is another line.
To be, or not to be: that is the question:
1234567890
ABC
XYZ
abc
xyz
Once you have typed in these lines, press RETURN to make sure you are at the beginning of a new line.
Press Ctrl-D (hold down the CONTROL key and the D key at the same time). This indicates end-of-file (EOF), which
informs cat that you have no more input. This will return you to the shell.
Note that unlike some operating systems, the “.txt” file extension is not required by UNIX or Linux. It is optional, but the
file extensions are often used to make it easier for a human to distinguish file types.
Now repeat the process with the following file (by convention, all of the input should be in bold, but to make it easier on
the eye, it is in italics here):
There is a lot of typing in this example, but it is important for the upcoming exercises. Type the following lines (by
convention, all of the input should be in bold, but to make it easier on the eye, it is in italics here):
Because of the length of this file, you may want to download a copy from the internet. There is a copy at
http://www.osdata.com/programming/shell/file02.txt — if you know how to download a copy and place it in your home
version 56
page 64
Unix and Linux System Administration and Shell Programming
directory, save yourself some typing time.
$ cat > file02.txt
This is a line of text in my second file.
This is another line.
To be, or not to be: that is the question:
Whether 'tis nobler in the mind to suffer
The slings and arrows of outrageous fortune,
Or to take arms against a sea of troubles,
And by opposing end them? To die: to sleep;
No more; and by a sleep to say we end
The heart-ache and the thousand natural shocks
That flesh is heir to, 'tis a consummation
Devoutly to be wish'd. To die, to sleep;
To sleep: perchance to dream: ay, there's the rub;
For in that sleep of death what dreams may come
When we have shuffled off this mortal coil,
Must give us pause: there's the respect
That makes calamity of so long life;
For who would bear the whips and scorns of time,
The oppressor's wrong, the proud man's contumely,
The pangs of despised love, the law's delay,
The insolence of office and the spurns
That patient merit of the unworthy takes,
When he himself might his quietus make
With a bare bodkin? who would fardels bear,
To grunt and sweat under a weary life,
But that the dread of something after death,
The undiscover'd country from whose bourn
No traveller returns, puzzles the will
And makes us rather bear those ills we have
Than fly to others that we know not of?
Thus conscience does make cowards of us all;
And thus the native hue of resolution
Is sicklied o'er with the pale cast of thought,
And enterprises of great pith and moment
With this regard their currents turn awry,
And lose the name of action. - Soft you now!
The fair Ophelia! Nymph, in thy orisons
Be all my sins remember'd.
1234567890
ABC
XYZ
abc
xyz
Now repeat the process with the following file (by convention, all of the input should be in bold, but to make it easier on
the eye, it is in italics here — note the cheat method below):
$ cat > numberfile.txt
1
2
3
4
5
version 56
page 65
Unix and Linux System Administration and Shell Programming
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Once again, use the RETURN key to enter a new line and then use Ctrl-D to exit cat .
If you want to cheat on making this file, you can use the following command on Mac OS X or BSD systems:
$ jot 20 1 > numberfile.txt
$
You will use these files in many of the exercises in this book.
PC-DOS equivalent
cat is the UNIX equivalent of the MS-DOS or PC-DOS command TYPE . You can add the PC-DOS equivalent to your
shell session with the alias command. To make the change permanent, add the following line to the .bashrc file in your
home directory. Note that if you add this PC-DOS/MS-DOS equivalent, only add the all upper case version, because the
lower case type is an important UNIX command that you will also need.
$ alias TYPE="cat"
view a file with cat
Type cat file 01, followed by the RETURN or ENTER key. You should see the contents of your first file. The
beginning of the file may scroll off the top of your screen (this is normal).
$ cat file01.txt
This is a line of text in my first file.
This is another line.
To be, or not to be: that is the question:
1234567890
ABC
XYZ
abc
xyz
$
version 56
page 66
Unix and Linux System Administration and Shell Programming
combine files with cat
Type cat file 02 numberfile.txt , followed by the RETURN or ENTER key. This will show you the contents of both
files, one immediately after the other. This is the use for which cat was named.
$ cat file02.txt numberfile.txt
This is a line of text in my second file.
This is another line.
To be, or not to be: that is the question:
Whether 'tis nobler in the mind to suffer
The slings and arrows of outrageous fortune,
Or to take arms against a sea of troubles,
And by opposing end them? To die: to sleep;
No more; and by a sleep to say we end
The heart-ache and the thousand natural shocks
That flesh is heir to, 'tis a consummation
Devoutly to be wish'd. To die, to sleep;
…many text lines…
The fair Ophelia! Nymph, in thy orisons
Be all my sins remember'd.
1234567890
ABC
XYZ
abc
xyz
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
$
This should list your new file that combines combines two of the files you created. The beginning of the listing will
scroll off the top of your screen. Don’t worry about it. We’ll see how to view long files in the chapter about less .
display with line numbers
The following example uses the command cat with the -n option to add line numbers to the listing of the file.
Type cat -n file01.txt .
version 56
page 67
Unix and Linux System Administration and Shell Programming
$ cat -n file01.txt
1 This is a line of text in my first file.
2 This is another line.
3 To be, or not to be: that is the question:
4
5 1234567890
6 ABC
7 XYZ
8 abc
9 xyz
$
Notice that the blank line is numbered. You can use the -b option to get line numbers, skipping all blank lines.
Type cat -b file01.txt .
$ cat -b file01.txt
1 This is a line of text in my first file.
2 This is another line.
3 To be, or not to be: that is the question:
4 1234567890
5 ABC
6 XYZ
7 abc
8 xyz
$
empty a file
You can use cat to empty a file by sending /dev/null to the file to be emptied. The null device is bit nothingness.
Start by making a copy of one of the files you just created so that you don’t wipe out your hard work.
$ cp file01.txt emptyfile
$
Use cat to verify that the new file really does exist and has a complete copy of the original file.
$ cat emptyfile.txt
This is a line of text in my first file.
This is another line.
To be, or not to be: that is the question:
1234567890
ABC
XYZ
abc
xyz
$
version 56
page 68
Unix and Linux System Administration and Shell Programming
Now type cat /dev/null > emptyfile.txt . This will leave the file in existence, but empty it of all characters.
$ cat /dev/null > emptyfile.txt
$
Confirm this by using both cat and wc (word count). The cat will simply return to the prompt because there is nothing to
print out. The wc will show zero lines, zero words, and zero characters.
$ cat file01.txt > emptyfile.txt
$ wc names
0 0 0 emptyfile.txt
$
replacing files
The previous example brings up the point that if you use cat to output to an existing file, the previous file will be
completely replaced by the new one. There is no warning at all.
appending files
You can append to the end of an existing file by using two greater than signs (no spaces in between) >>.
Start by making a copy of an existing file and then confirming the copy is correct.
$ cp file01.txt file03.txt
$ cat file03.txt
This is a line of text in my first file.
This is another line.
To be, or not to be: that is the question:
1234567890
ABC
XYZ
abc
xyz
$
Now try the append.
$ cat numberfile.txt >>.file03.txt
$
And confirm that your append worked.
$ cat file03.txt
This is a line of text in my first file.
This is another line.
To be, or not to be: that is the question:
version 56
page 69
Unix and Linux System Administration and Shell Programming
1234567890
ABC
XYZ
abc
xyz
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
$
hidden characters and spaces
There are several switches that can be used to view files with non-printable characters, trailing spaces, and other hidden
characters that can drive you crazy.
On Linux (not on BSD or SYS V or Mac OS X), you can use the -E command to place a $ dollar sign at the end of each
line. This will reveal trailing spces that would not otherwise be visible.
$ cat -E names
James$
Mary$
John$
Patricia$
Robert$
Linda$
$
The -v switch will display non-printing characters. Control characters are preceded by the ^ caret character, so control-X
would be listed as ^X. The delete character (octal 0177) is displayed as ^?. The non-ASCII characters (high bit set) are
displayed with M- (for meta) followed by the character for the low seven bits.
On Linux only, you can combine the -vE with the -e character (show nonprinting characters and the end of line marker).
On Linux only, you can use the -T switch to show tab characters as ^I.
You can use the -t switch to show tab characters as ^I and non-printing characters in the manner described for the -v
switch. This is the equivalent of the Linux only -vT.
On Linux only, you can use the -A switch as a shorter version of the ^vET combination.
version 56
page 70
Unix and Linux System Administration and Shell Programming
squeeze lines
You can use the -s switch to suppress multiple consecutive empty lines. Only one empty line will be displayed.
cat in scripts
While cat is a great tool for qork from the command line, if you find yourself using cat in a script, you should probably
rethink your algorithm. It is rare that you are on the right path when you find yourself using cat inside a shell script. The
use of cat in a shell script is a common sign of amateur hacking and the script probably has serious beginner mistakes and
problems as well.
version 56
page 71
Unix and Linux System Administration and Shell Programming
command separator
chapter 13
This chapter looks at the command separator in UNIX or Linux.
The semicolon ( ; ) is used as a command separator.
You can run more than one command on a single line by using the command separator, placing the semicolon between
each command.
$ date; who am i
Mon Aug 27 19:15:41 PDT 2012
admin ttys000 Aug 27 17:50
$
It does not matter if you have a space before the semicolon or not. Use whichever method is more readable and natural
for you.
$ date ; who am i
Mon Aug 27 19:15:41 PDT 2012
admin ttys000 Aug 27 17:50
$
Each command is processed in order, as if you had typed each individually, with the exception that the line with the
prompt is only displayed once, at the end of the series of commands.
If you forget the semicolon, you will get an error message. The following two examples are from different systems.
$ date who am i
date: illegal time format
$
$ date who am i
date: bad conversion
$
The semicolon is a common terminator or separator in many common programming languages, including Ada, C, Java,
Pascal, Perl, PHP, and PL/I.
Many programmers automatically place the semicolon at the end of any shell command. The shell is fine with this.
$ date;
Mon Aug 27 19:15:41 PDT 2012
$
version 56
page 72
Unix and Linux System Administration and Shell Programming
version 56
page 73
Unix and Linux System Administration and Shell Programming
less, more, pg
chapter 14
summary
This chapter looks at less , more , and pg, a related family of Unix (and Linux) commands.
less , more , and pg are utilities for reading a very large text file in small sections at a time.
pg is the name of the historical utility on BSD UNIX systems.
more is the name of the historical utility on System V UNIX systems.
pg and more are very similar and have almost the same functionality and are used in almost the exact same manner,
except for whether you type pg or more .
less is a more modern version that has the capaibilites of more along with additional new capabilities.
Mac OS X has all three versions installed. It is common to find both more and less on Linux systems. The older versions
are typically included so that old scripts written before the invention of less will still work.
The name less is a pun, from the expression “less is more.”
Try using less with your test file file01.txt .
You are going to try less first.
$ less file01.txt
If that doesn’t work, you are going to try more next.
$ more file01.txt
If that doesn’t work, you are going to try pg last.
$ pg file01.txt
One of the three will work. You can try the others to see if they are available on your system, but for the exercise, use the
first one that works.
$ less file01.txt
This is a line of text in my first file.
This is another line.
To be, or not to be: that is the question:
1234567890
ABC
XYZ
version 56
page 74
Unix and Linux System Administration and Shell Programming
abc
xyz
$
Now we are going to see the real power of less with a long file.
$ less file02.txt
This is a line of text in my second file.
This is another line.
To be, or not to be: that is the question:
Whether 'tis nobler in the mind to suffer
The slings and arrows of outrageous fortune,
Or to take arms against a sea of troubles,
And by opposing end them? To die: to sleep;
No more; and by a sleep to say we end
The heart-ache and the thousand natural shocks
That flesh is heir to, 'tis a consummation
Devoutly to be wish'd. To die, to sleep;
To sleep: perchance to dream: ay, there's the rub;
For in that sleep of death what dreams may come
When we have shuffled off this mortal coil,
Must give us pause: there's the respect
That makes calamity of so long life;
For who would bear the whips and scorns of time,
The oppressor's wrong, the proud man's contumely,
The pangs of despised love, the law's delay,
The insolence of office and the spurns
That patient merit of the unworthy takes,
When he himself might his quietus make
File02.txt
Notice that you are not yet back to the shell prompt. You are still in the less command. You are
only one page into the file. less allows you to view long files one page at a time, instead of
scrolling them off the top of your screen.
Type the SPACE key to see another page of text.
That patient merit of the unworthy takes,
When he himself might his quietus make
With a bare bodkin? who would fardels bear,
To grunt and sweat under a weary life,
But that the dread of something after death,
The undiscover'd country from whose bourn
No traveller returns, puzzles the will
And makes us rather bear those ills we have
Than fly to others that we know not of?
Thus conscience does make cowards of us all;
And thus the native hue of resolution
Is sicklied o'er with the pale cast of thought,
And enterprises of great pith and moment
With this regard their currents turn awry,
And lose the name of action. - Soft you now!
The fair Ophelia! Nymph, in thy orisons
Be all my sins remember'd.
1234567890
ABC
XYZ
abc
xyz
(END)
Notice that you are not yet back to the shell prompt. Even though the entire file contents have been listed, you are still in
the less command.
version 56
page 75
Unix and Linux System Administration and Shell Programming
Typing lower case q will quit the less (or more or pg) program.
less (and the other two utility commands) displays one page at a time. This is in contrast with cat , which displays the
entire file at once.
version 56
page 76
Unix and Linux System Administration and Shell Programming
file system basics
chapter 15
summary
This chapter looks at the basics of the file system on UNIX or Linux.
By 1971, many fundamentally important aspects of UNIX were already in place, including file ownership and file access
permissions.
graphics example
In a graphic user interface, you can see the file system directory structure in some kind of picture version. This varies by
system.
The following example is from Mac OS X. If you use Mac OS X, the UNIX directories are kept hidden in Finder. You
can see them by using Finder’s Go menu, Go To Folder…, and type “private”.
The following example is from Linux.
Do not mess with your system files until you know what you are doing. You can cause your computer to stop working.
Some of the typical Unix/Linux directories are etc, tmp, var, bin, sbin, dev, boot, or home.
some basics
Both UNIX and Linux organize the entire file system into a single collection (called the directory tree). This is in sharp
contrast to Windows dividing the file system into multiple parts, each based on a particular drive or device.
File names are case sensitive. The file names “File” and “file” are different files. This is in sharp contrast with Windows
and Macintosh (where they are alternate names for the same file). Note that this is a Mac OS X variation from the standard
UNIX practice.
A file name that starts with a period is a hidden file. In Mac OS X, you can toggle between showing and hiding hidden
files in Finder by typing Command-Shift-Period.
Neither UNIX nor Linux enforce or recognize file extensions (such as “.txt”). In some operating systems (such as
Windows and traditional Macintosh), the file extensions are recognized by the operating system and are used for such
version 56
page 77
Unix and Linux System Administration and Shell Programming
purposes as deciding which program is used to open the file. You can add file extensions UNIX or Linux file names (and
this is commonly done to help humans organize their files), but the file extension is ignored by the operating system and
has no special meaning.
Most modern Linux and UNIX operating systems support long file names with a wide variety of characters, including the
space character. Shells are set up for old school UNIX, which had a limited number of characters in file names and a very
limited choice of characters in file names. it is best to use these old school limits for files you manipulate through the UNIX
or Linux shell. Mac OS X allows you to put file names inside quotation marks (to allow such things as spaces in file
names). You can also access files through their inode numbers (which will let the shell work on any file name). Still, it is
wise to use only letters, numbers, period, hyphen, and underscore in UNIX or Linux file names 9and particularly to avoid
the space character).
directory tree
A directory in UNIX or Linux is the same as a folder in Macintosh or Windows.
Directories may contain any combination of files and subordinate directories.
This structure is called a tree. It branches out from a root.
The root is the beginning of the filesystem.
A branch is a directory and all of its subdirectories.
When a directory is inside another directory, the one inside is called a child and the outer one is the parent.
A leaf is anything that has no children, such as a file.
Unix and Linux are organized into a single file system tree, under a master directory called root and designated by the
forward leaning slash ( / ). This is in sharp contrast with Windows, where there is a seperate file system tree for each drive
letter.
In UNIX, everything is is a file. Even a directory is just a special file that stores information about other files.
some important directories
The following is a brief description of some of the major directories. Not all systems will have all of these directories.
/ The root directory of the entire file system.
/bin A collection of binary files, better known as programs.
/boot or /kernel The files needed to boot or start your computer.
/dev The devices connected to your computer (remember, everything, even hardware devices, is treated as a file in UNIX
or Linux).
/etc System configuration files, startup procedures, and shutdown procedures.
/home or /users The home directories for each user.
/lib or /Library Library files.
/net Other networked systems (again, treated as files).
/opt Third party software. In older systems this might be /usr/local
version 56
page 78
Unix and Linux System Administration and Shell Programming
/private On Mac OS X, this is the location of the UNIX files. It is normally kept hidden from all users in Finder, but is
visible in your terminal emulator and shell. Mac OS X has links to the major UNIX directories at the root level so that
UNIX and Linux tools can find their files in the standard locations.
/proc Information about processes and devices.
/sbin System binaries (that is, system programs).
/tmp Space for temporary files. Required by all forms of UNIX.
/usr User binaries (programs), libraries, manuals, and docs.
/var Variable files.
Swap Virtual memory on a hard drive. Allows the memory manager to swap some data and program segments to hard
drive to expand the amount of memory available beyond the limits of physical memory.
For a more advanced and complete listing, see major directories.
home and working directory
The home directory is the directory that your system places you in when you first log into your system (start the shell).
The home directory is defined in the /etc/passwd file. This is your personal space on a UNIX or Linux machine.
The tilde character can be used to name your home directory. ~user-name (where user-name is your actual user name)
indicates your home directory.
The working directory is whatever directory you happen to be working in. The working directory is always the same as
the home directory when you start a shell session, but you can change your working directory at any time (and typically do
change it). Commands normally are applied to the current working directory.
The working directory may also be called the current directory, current working directory, or present working directory.
parent and child directory
Any directories underneath a particular directory are called child directories. Any directory can have zero, one, or many
child directories.
The directory above any particular directory is called the parent directory. Any directory has only one parent (with the
exception of root, which has no parent).
absolute paths
Every directory or file can be defined by a complete absolute path. The complete absolute path gives every level of
directory, starting from the root. An absolute path name is sometimes called the full path name.
An example might be the /usr/java/bin . This would be a directory containing Java binaries. In this example, under the
/ root directory there is a directory called usr and inside that directory there is a directory called java and inside that
directory there is a directory called bin and inside that directory there are programs for running Java.
Absolute paths normally start at the root of the file system.
relative paths
Relative paths describe where to find a file or directory from the current working directory.
version 56
page 79
Unix and Linux System Administration and Shell Programming
A single period (or dot) character ( ./ ) indicates that you are describing the path to a directory or file from the current
working directory.
If your current working directory is www and that directory is under /usr , then the file index.html inside the current
working directory can be described by the full or absolute path of /usr/www/index.html or a relative path name of
./index.html .
The single period is called “dot”.
Two periods ( ../ called “dot dot”) take you up one directory to the parent directory.
You can use a series of two dots ( ../../ ) to go up more than one directory level.
dots and tildes and slashes
/ Slash is used as a separator between directory names. When slash is the first character in a path name, it indicates the
root directory.
~ Tilde is used to get to your home directory.
./ Single dot is used to indicate the current working directory.
../ Two dots are used to indicate the parent of the current working directory.
. A single dot in front of a file name (such as .Trash) is used to make a hidden file. Hidden files are only listed when
specifically requested.
hidden files
A file name that starts with a period is a hidden file. In Mac OS X, you can toggle between showing and hiding hidden
files in Finder by typing Command-Shift-Period. In MS-DOS, a hidden file can’t be written to. In UNIX and Linux a
hidden file can be read and written.
Some important hidden files are the files used to configure the working environment, including .cshrc, .login, and
.profile .
Some programs keep application-specific resources in hidden files or hidden directories.
moving around
This will all make a bit more sense in the next chapter as you learn how to actually move around your file system.
version 56
page 80
Unix and Linux System Administration and Shell Programming
pwd
chapter 16
summary
This chapter looks at pwd , a Unix (and Linux) command.
pwd is used to Print the Working Directory. It provides the full path for the current working directory. and is normally
run without any options.
pwd is a builtin command in bash . There is also an external utility with the same name and functionality.
working directory
The working directory is whatever directory you happen to be working in. The working directory is always the same as
the home directory when you start a shell session, but you can change your working directory at any time (and typically do
change it). Commands normally are applied to the current working directory.
The working directory may also be called the current directory, current working directory, or present working directory.
using pwd
Type pwd from the command line, followed by ENTER or RETURN, and you will see the current working directory.
$ pwd
/Users/admin
$
There’s not much else to tell you about this command, but this is an important one because you will continually have the
need to know where you are.
moving around
We will go into more detail about these commands later, but for now you need to know how to move around in your file
system.
The ls command is used to list the contents of the current directory.
Go ahead and type ls followed by ENTER or RETURN and you will see a list of the files and directories in your current
working directory (which should still also be your home directory).
The cd command is used to change directories.
Go ahead and type cd directory-name followed by ENTER or RETURN and you will move to the new directory. You
can confirm this by then typing pwd followed by ENTER or RETURN. You will see that you have successfully changed
working directories.
$ pwd
/Users/admin
$ ls
version 56
page 81
Unix and Linux System Administration and Shell Programming
Desktop Movies Send Registration
Documents Music Sites
Downloads Pictures
Library Public
$ cd Desktop
$ pwd
/Users/admin/Desktop
$
As you move around the directory tree, it is common to forget exactly where you are, especially if you do some other
tasks (such as file editing). pwd provides an easy way to determine where you are and insure that your next actions are
happening in the correct directory.
Use cd command all by itself to return to your home directory.
$ cd
$ pwd
/Users/admin
$
advanced topic
The remainder of this chapter covers an advanced topic. I am placing this here so that you will be easily able to find
them when you are ready for this information.
Eventually you will make use of symlinks and symlinked directories. There will be times when you need toknow the
actual physical path to the current working directory, not just the logical symlinked path.
options
pwd is almost always used without options and never uses any arguments (such as file names). Anything other than legal
options typed after pwd are ignored. Some shells ignore anything other than the legal options typed after pwd without
reporting any errors, while most shells report errors for illegal options. All shells seem to ignore all arguments without
reporting any errors.
The two main options for pwd are -L and -P.
The -L option displays the logical current working directory. This is the default if no options are listed.
$ cd /symlink
$ pwd -L
/symlink
$
The -P option displays the physical current working directory, with all symbolic links resolved.
$ pwd -P
/Users/admin
$
You can use the results of the pwd -P switch to cd to the physical directory.
version 56
page 82
Unix and Linux System Administration and Shell Programming
$ cd -P /Users/admin
$
You can also use the -P switch directly on the cd command to change to the physical directory.
$ cd -P /symlink
$ pwd
/Users/admin
$
You can use the help command to get information about pwd .
$ help pwd
Some early Linux shells allow the --help command to get information about pwd .
$ pwd --help
Some early Linux shells allow the --version option gives the version number for your copy of pwd .
$ pwd --version
version 56
page 83
Unix and Linux System Administration and Shell Programming
command history
chapter 17
summary
This chapter looks at command history in a Unix (and Linux) BASH shell.
Command history is used to recall recent commands and repeat them with modifications.
You will find that you are often repeating a few common commands over and over with variations in the options and
arguments (such as file names). Having the ability to recall a previous example and make a few changes will greatly save
on typing.
arrow keys
Pushing the Up Arrow key will bring back previously typed commands one at a time (from most recently typed to the
first typed).
Pushing the Down Arrow key will move from the oldest to the newest commands one at a time.
The commands brought back by the Up Arrow and Down Arrow keys will appear after the current prompt.
You can press the ENTER or RETURN key to immediately run any selected command.
After using the Up Arrow or Down Arrow keys, your cursor is at the end of the line. You can use the DELETE key to
remove old arguments and then start typing in new ones.
Note that holding down the CONTROL key and pressing the p key at the same time will take you to the previous
command just like the Up Arrow key. This is written ^p.
Holding down the CONTROL key and pressing the n key at the same time will take you to the next command just like
the Down Arrow key. This is written ^n.
Pressing the META key (or on some systems pressing the ESCAPE key), then releasing it, then typing the less than <
key will take you to the beginning of the command history. Pressing the META key (or on many modern systems pressing
the ESCAPE key or ESC key), then releasing it, then typing the greater than > key will also move to the end (most recent)
command history. These are written M-< and M->.
editing a line
As mentioned above, when you use command history, the cursor is at the end of the recalled command.
Either the Left Arrow or ^b (CONTROL-B) will go one character backward (towards the beginning of line).
Either the Right Arrow or ^f (CONTROL-F) will go one character forward (towards the end of line).
Typing inside a line will place characters in front of the cursor. The character inside the cursor remains unchanged.
Using the DELETE key will delete the character in front of the cursor. The character inside the cursor remains
unchanged.
^k (CONTROL-K) will delete everything from the cursor to the end of the line.
version 56
page 84
Unix and Linux System Administration and Shell Programming
^d (CONTROL-D) will delete the character inside (under) the cursor.
M-b will move you one word back (towards beginning of line). In this case, a word is a collection of characters, usually
separated by a space character.
M-f will move you one word forward (towards end of line).
^e (CONTROL-E) will move to the end of the line and ^a (CONTROL-A) will move to the beginning of the line.
Once you have edited your command, use the RETURN or ENTER key to run the new command. You can use the
ENTER or RETURN key from anywhere in the command line. You do not have to be at the end of the command line.
version 56
page 85
Unix and Linux System Administration and Shell Programming
built-in commands
chapter 18
summary
This chapter looks at built-in commands in a Unix (and Linux) BASH shell.
tools
Most of the commands in the UNIX or Linux shell are actually programs. If you look at the usr/bin directory, you will
see the actual programs. Most of these programs were written in the C programming language.
There is a common core of tools/commands that will be available on almost every UNIX or Linux machine, but exactly
how many and which commands/tools are available varies widely.
The good news is that if a command or tool is missing from your system, you can go out and get the source code and
recompile it for your local computer.
built-in
Many of the shells have special built-in commands. These are not separate programs, but are part of the code for the
shell itself.
One example would be the shell command cd that you just saw in the previous quick tour chapter.
There are some built-in commands are only available in selected shells and these can make your scripts shell-dependent.
Some examples of built-in commands include the history command in the C shell, and the export command in the
Bourne shell. The cd command is built-in in both bash and csh .
echo is an example of a command that is built into both bash and csh , but also exists externally as a utility.
overriding built-in commands
You can override any built-in commands by giving the full path name to an external command or utility. If bash finds a
slash character ( ) anywhere in a command, the shell will not run the built-in command, even if the last component of the
specified command matches the name of a builtin command.
As an example, using the command echo will run the version of the command that is built into bash , while specifying
or ./echo will ignore the built-in comamnd and instead run the designated utility.
/bin/echo
Overriding can be used to run alternative versions of commands or to extend the built-in command to add additional
features.
determining builtin or external
You can use the type command to determine if a particular command is a built-in command or an external utility. If the
command is an external utility, you will also be told the path to the external command.
$ type echo
echo is a shell builtin
version 56
page 86
Unix and Linux System Administration and Shell Programming
$ type mkdir
mkdir is /bin/mkdir
$
You can use the which command to locate a program in your path.
$ which echo
/bin/echo
$
You can use the whereis command to locate a program in your path.
$ whereis echo
/bin/echo
$
In csh and tcsh you can use the where command to locate a program in your path.
% where echo
/bin/echo
%
problems
You can use the type command to determine if a particular command is a built-in command or an external utility. If the
command is an external utility, you will also be told the path to the external command.
If something bad happens to your computer, if the shell is still loaded in memory and running, any of the built-in
commands will still work correctly, even if the entire file system (including all hard drives) disappears or becomes
unavailable for any reason.
built in command chart
The following chart shows the built-in commands and external utilities for bash and csh for Mac OS X. This wil be
similar for Linux and other UNIXes.
External commands marked No** under the External column do exist externally, but are implemented using the built-in
command.
Command External
!
No
%
No
.
No
:
No
{
No
}
No
alias
No**
alloc
No
bg
No**
version 56
csh
No
Yes
No
Yes
No
No
Yes
Yes
Yes
bash
Yes
No
Yes
Yes
Yes
Yes
Yes
No
Yes
page 87
Unix and Linux System Administration and Shell Programming
bind
bindkey
break
breaksw
builtin
builtins
case
cd
chdir
command
complete
continue
default
dirs
do
done
echo
echotc
elif
else
end
endif
endsw
esac
eval
exec
exit
export
false
fc
fg
filetest
fi
for
foreach
getopts
glob
goto
hash
hashstat
history
hup
if
jobid
jobs
kill
version 56
No
No
No
No
No
No
No
No**
No
No**
No
No
No
No
No
No
Yes
No
No
No
No
No
No
No
No
No
No
No
Yes
No**
No**
No
No
No
No
No**
No
No
No
No
No
No
No
No
No**
Yes
No
Yes
Yes
Yes
No
Yes
Yes
Yes
Yes
No
Yes
Yes
Yes
Yes
No
No
Yes
Yes
No
Yes
Yes
Yes
Yes
No
Yes
Yes
Yes
No
No
No
Yes
Yes
No
No
Yes
No
Yes
Yes
No
Yes
Yes
Yes
Yes
No
Yes
Yes
Yes
No
Yes
No
Yes
No
Yes
Yes
Yes
Yes
No
Yes
No
No
Yes
Yes
Yes
No
Yes
Yes
No
No
No
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
No
Yes
Yes
No
Yes
No
No
Yes
No
No
No
Yes
Yes
Yes
No
page 88
Unix and Linux System Administration and Shell Programming
No
local
No
log
No
login
Yes
logout
No
ls-F
No
nice
Yes
nohup
Yes
notify
No
onintr
No
popd
No
printenv
Yes
pushd
No
pwd
Yes
read
No**
readonly
No
rehash
No
repeat
No
return
No
sched
No
set
No
setenv
No
settc
No
setty
No
setvar
No
shift
No
source
No
stop
No
suspend
No
switch
No
telltc
No
test
Yes
then
No
time
Yes
times
No
trap
No
true
Yes
type
No
ulimit
No
umask
No**
unalias
No**
uncomplete No
unhash
No
unlimit
No
unset
No
unsetenv
No
limit
version 56
Yes
No
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
No
No
No
Yes
Yes
No
Yes
Yes
Yes
Yes
Yes
No
Yes
Yes
Yes
Yes
Yes
Yes
No
No
Yes
No
No
No
No
No
Yes
Yes
Yes
Yes
Yes
Yes
Yes
No
Yes
No
No
No
No
No
No
No
No
No
No
No
Yes
Yes
Yes
No
No
Yes
No
Yes
No
No
No
Yes
Yes
No
No
No
No
No
Yes
Yes
No
Yes
Yes
Yes
Yes
Yes
Yes
Yes
No
No
No
Yes
No
page 89
Unix and Linux System Administration and Shell Programming
until
wait
where
which
while
version 56
No
No**
No
Yes
No
No
Yes
Yes
Yes
Yes
Yes
Yes
No
No
Yes
page 90
Unix and Linux System Administration and Shell Programming
ls
chapter 19
summary
This chapter looks at ls, a Unix (and Linux) command.
ls is the Unix (and Linux) list command.
The ls command was described in the first UNIX book, UNIX Programmer’s Manual, by Ken Thompson and Dennis
Ritchie, published November 3, 1971.
list directory contents
Use the ls command to list the contents of the current working directory (files and directories).
$ ls
Desktop Movies Send registration
Documents Music Sites
Downloads Pictures
Library Public
$
PC-DOS equivalent
ls -la is the UNIX equivalent of the MS-DOS or PC-DOS command DIR . You can add the PC-DOS equivalent to your
shell session with the alias command. To make the change permanent, add the following line to the .bashrc file in your
home directory. Note that if you add this PC-DOS/MS-DOS equivalent, only add the all upper case version, because the
lower case type is an important UNIX command that you will also need.
$ alias DIR="ls -la"
ls -la | less is the UNIX equivalent of the MS-DOS or PC-DOS command DIR /P.
list all
Type ls with the -A option to get a list of all entries other than . (single dot) and .. (double dot).
$ ls -A
list with indicators
Type ls with the -F option to have special indicators for special files.
$ ls -F
Desktop/ Music/ file01.txt
version 56
page 91
Unix and Linux System Administration and Shell Programming
Documents/ Pictures/ file02.txt
Downloads/ Public/ names
Library/ Send registration@ numberfile.txt
Movies/ Sites/ testdir/
$
A slash / will be added after every pathname that is a directory. An asterisk * will be added after every file that is
executable (including scripts). An at sign @ will be displayed after each symbolic link. An equals sign = will be displayed
after each socket. A percent sign % will be displayed after each whiteout. A vertical bar | will be displayed after each item
than is a FIFO.
ls -F is a builtin command in csh .
hidden files
The -a option will give a list of all files, including hidden files.
$ ls
Desktop Downloads Movies Pictures scripts
Documents Library Music Public
$ ls -a
. .cups Library scripts
.. .vminfo Movies
.DS_Store Desktop Music
.Trash Documents Pictures
-bash_history Downloads Public
$
The hidden or invisible files are hidden to make it less likely that they are accidentally corrupted or deleted.
list specific directory
So far all of the examples have been on the current working directory. If you name a specific directory, you will get a
listing for that directory/ You may use any of the flags you have alrady learned.
$ ls Music
GarageBand iTunes
$
multiple directories
You can list multiple directories, each separated by a space character. This will give the listing for each directory, one at
a time.
$ ls Movies Music
Movies:
iMovie Projects.localized
Music:
GarageBand iTunes
$
version 56
page 92
Unix and Linux System Administration and Shell Programming
separate lines
The -1 option will cause each item on its own line. This is useful for scripts that will do further processing.
$ ls -1
Desktop
Documents
Downloads
Library
Movies
Music
Pictures
Public
scripts
$
option grouping
You can combine multiple options for a single command. For example, if you wanted to show visible and invisble files
and show the file indicators, you can use both the -F and -a options together.
$ ls -a -F
./ .cups/ Library/ scripts
../ .vminfo Movies/
.DS_Store Desktop/ Music/
.Trash/ Documents/ Pictures/
-bash_history Downloads/ Public/
$
You can group option by listing them in any order (with a few rare exceptions) and either with their own minus sign or
after a single minus sign. All of the following versions of the ls command work the same way:
$
$
$
$
$
ls
ls
ls
ls
-a -F
-F -a
-aF
-Fa
list specific file
You can ls a specific file or files. This is primarily useful with other options that report additional information.
$ ls filename
list specific directory
You can ls a specific directory. It will list all of the files in the designated directory.
$ ls directory
version 56
page 93
Unix and Linux System Administration and Shell Programming
You can the -d switch to have ls list just the named directory. This is pretty much useless by itself, but useful when
combined with switches for specific kinds of information.
$ ls -d directory
list files and directories
You can ls a a combined list of specific files and directories.
$ ls filename directory
common error
The most common error is attempting to ls a directory or file that doesn’t exist.
colorized
Use the -G option to colorize the output of ls.
$ ls -G
Desktop Downloads Movies Pictures file01.txt names scripts
Documents Library Music Public forthfunctions
numberfile.txt testdir
$
You can edit your .profile file to set specific colors for directories, sockets, pipes, symlinks, etc. You can also set
colors for executable files. You can even set colors for specific file extensions.
The default for Mac OS X (shown above) is royal blue for directories.
Daniel Sandman wrote:
I have this as an alias in my ~/.zshrc .
alias ls="ls --color=auto -FshX"
list all subdirectory contents
Use the -R switch to show all of the contents of all of the subdirectories. This can be a huge list. The -R stands for
recursive.
$ ls -R
long listing
Use the -l switch for a long listing. Each file or directory is on its own line. One sample line shown:
version 56
page 94
Unix and Linux System Administration and Shell Programming
$ ls -l
drwxr-xr-x 2 admin staff 68 Aug 20 16:43 testdir
$
The order of information
Mode file type (see chart below), file permissions (owner, group, other in the order of read, write, execute), possible
special information.
b Block special file
c Character special file
d Directory
l Symbolic link
s Socket link
p FIFO (pipe)
- Regular file
If the file or directory has extended attributes the permissions field is followed by the @ character.
If the file or directory has extended security information (such as an Access Control List), the permissions field is
followed by the + character.
links The number of links.
user The user name. If the user name is unknown, the numeric ID is shown. Adding the -n switch will force the numeric
ID instead.
group The group name. If the group name is unknown, the numeric ID is shown. Adding the -n switch will force the
numeric ID instead.
size The number of bytes. If the file is a character special or block special file, the major and minor device numbers
replace the size field.
date and time The abbreviated month, day-of-month, hour and minute the file was last modified. If the modification
time is more than six (6) months past or future, then the year replaces the hour and minute fields.
pathname The name of the file or directory. If the file is a symbolic link, the pathname of the linked-to file is preceded
by ->.
The first line (before the individual file entries) is the number of blocks used by the files in the directory.
list in reverse order
Use the -r switch to show the results in reverse order.
$ ls -r
list by time of modification
Use the -t switch to show the results by time of modification (most recent first).
$ ls -t
version 56
page 95
Unix and Linux System Administration and Shell Programming
list by time of last access
Use the -u switch to show the results by time of last access (most recent first).
$ ls -u
list by time of file creation
Use the -U switch to show the results by time of file creation (most recent first).
$ ls -U
show time information
Use the -T switch to show full time information: month, day, hour, minute, second, year. This switch needs to be
comibined with the -l switch. Sample line below:
$ ls -Tl
drwxr-xr-x 2 admin staff 68 Aug 20 16:43:15 2013 testdir
$
show extended information
Use the -@ switch to show extended attribute keys and sizes. This switch needs to be comibined with the -l switch.
$ ls -@l
show Access Control List
Use the -e switch to show the Access Control List (ACL) associated with a file when there is an ACL. This switch needs
to be comibined with the -l switch.
$ ls -el
show number of blocks
Use the -s switch to show the size in blocks for each file. The UNIX default block size is 512 bytes, but thsi can very
different on modern versions of UNIX or Linux.
$ ls -s
advanced information
version 56
page 96
Unix and Linux System Administration and Shell Programming
scripting gotchas
This is an advanced item, but it is placed in this chapter so that you can easily find where to look it up when you need it.
Do not use the output of ls in a BASH for loop.
$ for i in $(ls *.mp3); do # Wrong!
$ command $i # Wrong!
$ done # Wrong!
$ for i in $(ls *.mp3) # Wrong!
$ for i in `ls *.mp3` # Wrong!
If any file name has spaces in it, then it will be word split. So, the MP3 for Alex Clair’s “Too Close” (number four song
on his “The Lateness of the Hour” album) might be 04 Too Close . In that case, your command would run on the files
“04”, “Too”, and “Close”, rather than “04 Too Close”. Additional errors likely to follow. Listen to the song “Too Close”
for free at http://www.thissideofsanity.com/musicbox/musicbox.php?mastersong=485 .
Double quote won’t work either because that will cause the entire output of the ls command to be treated as a single
item. Your loop will run once on all of the file names conactenated together.
$ for i in "$(ls *.mp3)"; # Wrong!
The solution is to run your loop without even using ls, relying on BASH’s filename expansion. Don’t forget to check for
the possibility that there are no qualifying files in the current directory.
$ for i in *.mp3; do
$ [[ -f "$i" ]] || continue
$ command "$i"
$ done
Thanks to Daniel Massimillano for a correction.
version 56
page 97
Unix and Linux System Administration and Shell Programming
cd
chapter 20
summary
This chapter looks at cd, a Unix (and Linux) command.
cd is used to Change Directory.
cd is a builtin command in bash and csh . There is also an external utility with the same name and functionality.
cd is used to change the directory you are working in. You type the command cd followed by a space and then the
directory (folder) that you want to change to.
The chdir (change directory) command, the predecessor of cd, was described in the first UNIX book, UNIX
Programmer’s Manual, by Ken Thompson and Dennis Ritchie, published November 3, 1971.
change directory example
For the purposes of an example, we need to first find a directory to change to. Type the command ls -F.
$ ls -F
Desktop/ Music/ file01.txt
Documents/ Pictures/ file02.txt
Downloads/ Public/ names
Library/ Send registration@ numberfile.txt
Movies/ Sites/ testdir/
$
Your listing will be different than this one. Look for any name that has a slash ( / ) after it and use it as the
directory_name in the following example. If you created the testdir in the quick tour chapter, then use the second example
listed here.
$ cd directory_name
admins-power-mac-g5:directory_name admin$
If you created the testdir in the quick tour chapter, then use the following example:
$ cd testdir
admins-power-mac-g5:testdir admin$
Use the pwd command to confirm that you are now in your new directory.
$ pwd
/Users/admin/testdir
admins-power-mac-g5:testdir admin$
Use the cd command without any additional arguments to return to your home directory from anywhere.
version 56
page 98
Unix and Linux System Administration and Shell Programming
$ cd
admins-power-mac-g5:~ admin$
Use the pwd command to confirm that you are now back in your home directory.
$ pwd
/Users/admin
admins-power-mac-g5:~ admin$
return to home directory
You can always return to your home directory from anywhere by typing cd command all by itself.
$ cd
admins-power-mac-g5:~ admin$
If the HOME variable is set (which is normal), then you can use it to return to your home directory.
$ cd $HOME
admins-power-mac-g5:~ admin$
return to previous directory
To return to the previous directory, type cd ~ admins-power-mac-g5:~ admin$ cd testdir
changing to the testdir directory
admins-power-mac-g5:testdir admin$ cd ../Library
changing to the Library directory
admins-power-mac-g5:Library admin$ cd ~changing back to the testdir directory, the immediately preceding directory
admins-power-mac-g5:testdir admin$
go one level up
Type cd ../ to go up one directory level. The following exampel starts in the testdir directory and goes up one level to
the home directory.
admins-power-mac-g5:testdir admin$ cd ../
changing to the home directory, one level up from the testdir directory
admins-power-mac-g5:~ admin$
You can type a series of ../ to go up multiple directories. cd ../../ will go up two levels. cd ../../../ will go up
three levels. And so on.
common cd errors
version 56
page 99
Unix and Linux System Administration and Shell Programming
The two most common cd errors:
Attempting to cd to a file. You can only cd to a directory.
Attempting to cd to a directory that doesn’t exist. The directory must already exist.
PC-DOS equivalent
cd .. is the UNIX equivalent of the MS-DOS or PC-DOS command CD.. . You can add the PC-DOS equivalent to your
shell session with the alias command. To make the change permanent, add the following line to the .bashrc file in your
home directory.
$ alias CD..="cd .."
advanced topic
The following is an advanced topic. I am placing it in this chapter so that you can easily find it when you need it in the
future.
If you use cd to change to a symlinked directory, pwd will show the logical symlinked path rather than the actual physical
path. You can use the -P option to force the change of directory to be resolved to the actual physical path.
$ cd -P /symlink
$ pwd
$ /Users/admin
$
version 56
page 100
Unix and Linux System Administration and Shell Programming
cp
chapter 21
summary
This chapter looks at cp, a Unix (and Linux) command.
cp is used to copy a file.
copy a file
Use the cp command to make a copy of a file. This example assumes you created the names file in the quick tour
chapter.
$ cp names saved_names
$
Notice that there is no confirmation of the file copy being made.
This silent behavior is typical of any UNIX shell. The shell will typically report errors, but remain silent on success.
While disconcerting to those new to UNIX or Linux, you become accustomed to it. The original purpose was to save paper.
When UNIX was first created, the terminals were mostly teletype machines and all output was printed to a roll of paper. It
made sense to conserve on paper use to keep costs down.
You can use the ls command to confirm that the copy really was made. You won’t be using up any paper.
$ ls
Desktop Movies Send registration
Documents Music Sites
Downloads Pictures names
Library Public saved_names
$
cp makes an exact copy of a file.
common errors
Common errors:
the source is a directory
the source does not exist
attempt to copy a file to itself
overwriting files
The plain cp command will simply replace any existing file with a matching name. No warning.
$ cp names saved_names
$
version 56
page 101
Unix and Linux System Administration and Shell Programming
In the above example, if the file save_names already existed, it will be replaced with a copy of names .
interactive confirmation
You can type cp -i to get an interactive confirmation before replacing a file. Enter n or just the RETURN key to stop
the copy and prevent an overwrite. Enter y to go ahead with the copy and replace the old file.
$ cp -i names saved_names
overwrite saved_names? (y/n [n]) n
not overwritten
$
force overwrites
Use the -f option to force the copy to overwrite any file with a matching name. This is the same as the default action of
the cp command.
$ cp names -f saved_names
copy to a directory
To copy a file to a different directory, name the directory with a leading slash ( /) as the destination:
$ cp sourcefilename directory/
Try this example:
$ cp names testdir/
$ ls testdir
names
$
A copy of the file will be stored in the specified directory with the original name. You can add a new file name after the
directory if you want to change the name of the copy at the same time you create it in the specified directory.
$ cp sourcefilename /directory/newname
Try this example:
$ cp names testdir/newnames
$ ls testdir
names newnames
$
copying an entire directory
version 56
page 102
Unix and Linux System Administration and Shell Programming
To copy an entire source directory to a different directory, name each directory:
$ cp sourcedirectory destinationdirectory/
This does not work on all systems without using the -r or -R switch (see next item).
copying an entire directory and subdirectories
If you want to copy a directory and all of its subdirectories (the directory subtree), then use the recursive switch. The
recursive switch will be either -r or -R. Check the man page for your system.
$ cp -r sourcedirectory destinationdirectory/
Mac OS X supports both the -r and the -R switches for compatibility with legacy scripts, but on Mac OS X the -r option
will not correctly copy special files, symbolic links, or fifos, while the -R option will copy them correctly.
Try this example:
$ cp -r testdir newdir/
$ ls newdir
names newnames
$
copying multiple items to a directory
You can create a list of sources (both files and directories) that are all put into a single destination directory. The last
item listed must be a directory and is the destination. You can optionally use the -r or -R flag (which ever is appropriate
for your system) to recursively copy source directories with all of their subdirectories. SOme systems may require the -r or
-R switch if you include directories as well as files int he source list.
$ cp sourcedirectory1/ sourcedirectory2/sourcefile destinationdirectory/
common error
A common error is that the destination folder already exists.
PC-DOS equivalent
cp -i is the UNIX equivalent of the MS-DOS or PC-DOS command COPY . You can add the PC-DOS equivalent to your
shell session with the alias command. To make the change permanent, add the following line to the .bashrc file in your
home directory. Note that this version adds interactive questioning before replacing a file, in a manner similar to PC-DOS.
Note also that if you add this PC-DOS/MS-DOS equivalent, only add the all upper case version, because the lower case
type is an important UNIX command that you will also need.
$ alias COPY="cp -i"
version 56
page 103
Unix and Linux System Administration and Shell Programming
version 56
page 104
Unix and Linux System Administration and Shell Programming
mv
chapter 22
summary
This chapter looks at mv, a Unix (and Linux) command.
mv is used to move or rename a file.
rename a file
Use the mv command to rename a file. This example assumes you created the names and saved-name files in the quick
tour chapter.
$ mv saved_named old_names
$
Notice that the shell is once again silent with success. You can use the ls command to confirm that the rename really
was made.
$ ls
Desktop Movies Send registration
Documents Music Sites
Downloads Pictures names
Library Public old_names
$
common error
A common error is the source file not existing (often because of a typing or speling error).
interactive confirmation
Use the -i option to have an interactive confirmation before overwriting an existing file. Note that this does not seem to
work on Mac OS X Mountain Lion.
$ mv -i names saved_names
overwrite saved_names? (y/n [n]) n
not overwritten
$
force overwrites
Use the -f option to force rename ( mv) to overwrite any file with a matching name. This is the same as the default action
of the mv command.
$ mv -f names saved_names
version 56
page 105
Unix and Linux System Administration and Shell Programming
prevent overwrites
Use the -n option to prevent rename ( mv) from overwriting any file with a matching name.
$ mv -n names saved_names
PC-DOS equivalent
mv is the UNIX equivalent of the MS-DOS or PC-DOS command REN . You can add the PC-DOS equivalent to your
shell session with the alias command. To make the change permanent, add the following line to the .bashrc file in your
home directory.
$ alias REN="mv"
version 56
page 106
Unix and Linux System Administration and Shell Programming
rm
chapter 23
summary
This chapter looks at rm, a Unix (and Linux) command.
rm removes or deletes files or directories.
rm is used to remove or erase a file.
rm was part of the original 1969 version of UNICS (the original name for UNIX).
delete a file
Use the rm (remove) command to delete a file. This example assumes you created the names saved_names and
old_names files in the quick tour chapter.
$ rm old_names
$
You can use the ls command to confirm that the file was really deleted.
$ ls
Desktop Movies Send registration
Documents Music Sites
Downloads Pictures names
Library
$
safety
Before using rm, especially rm with globbing (such as using wild cards), try the same glob with ls first to make sure you
will be removing the files you intend to remove.
For example, before running rm foo*.txt , run ls foo*.txt . Read the list and be sure that you are going to delete the
files you really intend to delete.
Some UNIX and Linux users modify the rm command by aliasing it with the interactive form, rm -i. This changes the
behavior of rm to always ask before deleting files.
$ alias rm="rm -i"
Unfortunately, this alias can cause some scripts to fail (possibly including system scripts that were provided with your
distribution).
An alternative is to alias to a different, non-conflicting name. In this case, del might be appropriate.
version 56
page 107
Unix and Linux System Administration and Shell Programming
$ alias del="rm -i"
PC-DOS equivalent
rm -i is the UNIX equivalent of the MS-DOS or PC-DOS command DEL . You can add the PC-DOS equivalent to your
shell session with the alias command. To make the change permanent, add the following line to the .bashrc file in your
home directory. Note that this version adds interactive questioning before deleting a file, in a manner similar to PC-DOS.
$ alias DEL="rm -i"
deleting a file name with spaces
Sometimes a file name has space or tab characters in it. This is strongyl discouraged on a server, but is fairly common on
a desktop Linux system. Enclose the file name in quotation marks.
$ rm "sample file"
$
recursive
Type rm followed by the option -r followed by directory name to delete recursively. This means that the directory and all
of its subdirectories will be removed, even for non-empty direcories. This is a fast way to remove directories and files in a
single command. This is also a very dangerous command. Always include the directory name rather than * to prevent
accidentally destroying the wrong directories.
$ rm -r directory-name
$
This is also a very dangerous command. You may want to include the -i interactive flag to confirm that you are really
deleting the correct files.
$ rm -ri directory-name
$
dangerous command
WARNING! Do not test these examples. They will delete all of the files on your ocmputer.
These commands often appear in cruel internet jokes.
$ rm -rf /
$ rm -rf *
advanced information
version 56
page 108
Unix and Linux System Administration and Shell Programming
file won’t delete
The following information is rather advanced and requires knowledge of that hasn’t been introduced in this book yet.
This advanced material is placed here so that you can easily find it when using this book as a reference rather a tutorial. For
now, just note that this material is located here.
Sometimes you will attempt to delete a file and the shell will tell you the file can’t be deleted.
$ rm stubborn_file
rm: cannot remove `stubborn_file': Permission denied
$
You can try to chmod 777 the file or chown the file or directory.
The next line of defense is to use sudo and delete the file as root or superuser.
$ sudo rm stubborn_file
$
If you can’t remove the file as root, then there are a number of different possibilities to consider. The following is a
checklist of some of the possibilities.
It might be a matter of directory permissions. Directory permissions, not file permissions, control who can delete and
create files. If you don’t have write-permissions for the directory, you can’t delete the file.
Interestingly, the converse is also true. A user can delete a file set to 000 and owned by root in a directory that he or she
has access to, even though they can’t even view the file contents. Of course, the sticky bit on a directory sets files so that
they can only be deleted by their owners. It is common to set the sticky bit on certain key directories, such as /tmp/ .
The root or superuser should be able to override any regular permissions and delete any file anyway, so the you will
want to check Access Control Lists (ACLs), extended attributes, and other special features that may be in play. You can try
lsattr to list the attributes on a Linux second extended file system (ext2) and chattr to change the attributes. The a
attribute is append mode and can be removed with chattr -a stubborn_file . The i attribute is immutable and can be
removed with chattr -i stubborn_file . On Mac OS X, you can use the similar chflags command (for UFS, HFS+,
SMB, AFP, and FAT). See the man pages for details on these commands.
Another possibility is that the file may be on Network File System (NFS). Your local root won’t matter. You need root
access to the machine running NFS.
$ rm stubborn_file
$
There may be unusual characters in the file name.
You may be able to use auto-completion to correctly escape the file name. Type the first character of file name, then
press the Tab key. The shell should auto-complete the file name with correct escapes for unusual characters.
You can check to see if there are invisible or non-printing characters or spaces or tabs at the beginning or end of the file
name by comparing the results of ls stubborn_file with the output of ls *stubborn_file*. If you get no result from the
first command and a result from the second command, then there are extra characters. Try using the inum solution
mentioned below.
Another method for checking for the real file name:
version 56
page 109
Unix and Linux System Administration and Shell Programming
$ ls -la stubborn_file | sed -n l
-rw-r--r-- 1 Mac staff 17 Jan 11 02:56 stubborn_file$
$
Another method for checking for the real file name is to use the -B option to force the printing of non-printable
characters. The non-printable characters will be presented as \xxx where xxx is the numeric character code in octal. The -b
option is similar, but will use C programming language escape codes whenever possible.
I have read that Linux supports the -Q option to place the file names within quotation marks, but I have not tested this on
a Linux system.
Once you identify the real file name, you can use quotation marks to give the complete file name.
$ rm "stubborn file"
$
In some cases you may need to use single quotes. An example would be a file name with the character $ in it.
$ rm 'stubborn $file'
$
You can try to remove the file using its inode rather than name. Use the correct inum output from your real session. Note
there may be a long delay before success, so you may want to run this command in the background.
$ ls - i stubborn_file
3982096 stubborn_file
$ find . -inum 3982096 -exec rm -f {} \;
$
version 56
page 110
Unix and Linux System Administration and Shell Programming
shred
chapter 24
summary
This chapter looks at shred , a Linux command.
shred is a secure method for eliminating data, overwriting a file.
$ shred filename
Files are normally deleted by removing their file table entries. The data remains out on the hard drive (which is how file
recovery programs can work).
Writing over the file actually eliminates the data so that nobody can ever read it.
This can take a while for large files or for many files.
Files deleted with shred can not be recovered (unless you previously stored them in a backup somewhere else).
Be especially careful when using shred with any wildcard characters.
You should probably never run shred from root or superuser.
Mac OS X
Mac OS X offers secure empty trash from the graphic interface (Finder menu,
Secure Empty Trash menu item, which is directly under regular Empty Trash).
Mac OS X also has the srm command, which is the secure equivalent of regular rm.
$ srm -rfv -s filename
The -z option will overwrite with zeros. The -s option will overwrite in a single
pass (one pass) of random data. The -m option will perform a Department of Defense
seven-pass deletion. No options will result in a 35-pass file erase, which will take a
long time.
version 56
page 111
Unix and Linux System Administration and Shell Programming
mkdir
chapter 25
summary
This chapter looks at mkdir , a Unix (and Linux) command.
mkdir is used to make a directory.
make a directory
Use the mkdir command to make a new directory (folder).
$ mkdir testdir
admins-power-mac-g5:~ admin$
The example includes an extended prompt that shows the current working directory. Using the ls command will confirm
that the directory now exists in the current working directory.
common errors
mkdir will fail if the directory already exists.
mkdir will fail if a file with the same name already exists.
multiple directories
You can name multiple directories with the mkdir command. Simply list each of the new diretcories after the mkdir with
a space character between each directory name.
$ mkdir directory1 directory2 directory3
admins-power-mac-g5:~ admin$
create with permissions
Use the -m option to create a directory with specific permissions. The old school version of the -m switch used octal
permissions.
The permission chocies are the same as with the chmod command. Replace mode ( a=rwx in the following example) with
an appropriate mode argument. In the following example, themode is a=rwx , which would give all users read, write, and
execute permissions.
$ mkdir -m a=rwx directory1 directory2 directory3
admins-power-mac-g5:~ admin$
This option is often used in shell scripts to lock down temporary directories.
version 56
page 112
Unix and Linux System Administration and Shell Programming
use a path
All of the previous examples involved the current working directory. You can use cd to move to a new directory and
create a directory there. Or you can give an entire path name to a specific directory, using either a relative or absolute path
name to create the directory anywhere in the file system.
$ mkdir directory1/directory2/directory3
admins-power-mac-g5:~ admin$
common error
The most common error in using mkdir with a path is one or more of the intervening directories haven’t been made yet.
mkdir will fail.
creating parent directories
There is a solution to this problem. You can use the -p option to create directories along the path. You can therefore,
create a series of parent and child directories all at once.
If you use the -p option and the directory already exists, the mkdir command will continue on through your path and
create new directories as needed down to the last one you list.
$ mkdir -p directory1/directory2/directory3
admins-power-mac-g5:~ admin$
Intermediate directories are created with the permission bits of rwxrwxrwx (0777) as modified by the current umask, plus
the write and search permission for the owner.
view directories made
You can use the -v to get a list of the directories created. This is most often used with the -p option, but you can use it
by itself.
$ mkdir -v directory1/directory2/directory3
$
$ mkdir -pv directory1/directory2/directory3
$
The -v option is considered nonstandard and should not be used in shell scripts.
spaces in names
To create a directory with space in the directory name, you need to use quoting for the entire name so that the shell treats
it as a single name rather than a series of individual directory names separeted by spaces.
$ mkdir "directory name"
$
version 56
page 113
Unix and Linux System Administration and Shell Programming
While it is common to use spaces in file and directory names in graphic user interfaces (such as Macintosh, OS/2, and
Windows, as well as Gnome or KDE), this may cause failure with older scripts or programs. Some programs may convert
the spaces in a directory name into %20 , such as converting "Name with spaces" into "Name%20with%20spaces". This
might cause confusion or failure.
long options
The long options are:
-m
-p
-v
-Z
--mode=MODE
--parents
-verbose
--context=CONTEXT
--help
--version
advanced information
The following items are advanced topics. They are included here so that you can easily find them when needed.
scripting mkdir
If you create a directory in a script, you will want to check first to see if the script actually exists. The following line of
code will create a directory called tmp
$ test ! -d $HOME/tmp && mkdir $HOME/tmp
This line relies on the short circuit evaluation by the test statement. The first part of the test checks to see if the
specified directory does not exist. If the specified directory already exists, the test fails and further evaluation stops. If the
test passes, then the mkdir command is run.
create an entire directory tree
You can create an entire directory tree all at once. The example will create the following diretcory tree (which will be
confirmed with ls commands):
example
|
------------------------------- | | |
docs notes src
|
applet
|
----------------------------------------- | | | |
css html javascript php
$ mkdir -p example/{src/applet/{css,html,php,javascript},docs,notes}
$ ls example
docs notes src
$ ls example/src
applet
$ ls example/src/applet
css html javascript php
version 56
page 114
Unix and Linux System Administration and Shell Programming
$
The {} curly braces are used to create directories at the same level and the directories at the same level are separated by
the , comma.
create a directory and change to it
You can use the following function to both mkdir a new directory and cd to the newly created directory, all in one
command.
$ function mkdircd () { mkdir -p "$@" && eval cd "\"\$$#\""; }
$
This function will work even if the directory name has spaces.
various operating systems
The mkdir command appears in the Linix, Mac OS X, OS/2, PC-DOS (also called MS-DOS), UNIX, and Windows
operating systems. It also appears in the PHP programming language. In OS/2, PC-DOS, and Windows, the command can
be abbreviated md.
set security context
When running SELinux, you can use the -Z or --context switch to set the security context. You must give the full
context as user:role:type .
An example from the CentOS forum:
$ mkdir --context=system_u:object_r:httpd_sys_content_t:s0 /var/www/example.org
$
history
The mkdir command appeared in Version 1 of AT&T UNIX. In early versions of UNIX, the mkdir command had to be
setuid root because the kernel did not yet have a mkdir syscall. The mkdir command made a directory with the mknod
syscall, followed by linking in the . and .. directories.
The mkdir syscall originated in 4.2 BSD and was added to System V in Release 3.0.
version 56
page 115
Unix and Linux System Administration and Shell Programming
alias
chapter 26
This chapter looks at alias , a Unix (and Linux) command.
alias is used to create short names for long strings of commonly performed commands, scripts, and functions, and their
combinations.
alias is a builtin command in bash and csh . There is also an external utility with the same name and functionality.
example
You can use the alias command to modify the rm so that it always displays an interactive query before deleting a file.
$ alias rm="rm -i"
Unfortunately, this alias can cause some scripts to fail (possibly including system scripts that were provided with your
distribution). Any scripts that rely on the unmodified version of an aliased rm are likely to fail or produce strange results
because of your alias.
A better alternative is to use an alias to a different, non-conflicting name. In this case, del might be appropriate.
$ alias del="rm -i"
overriding aliases
If you have set an alias on a command or utility, you can always run the raw unmodified command or utility by typing
the absolute directory path to the command. For example, if you set an alias to have ls always show hidden files, you can
still run the regular ls by typing its full path name.
$ alias ls="ls -a" #modify the ls command
$ /bin/ls #run the raw ls command
Some system administrators always type out the complete path name for a command when using the root or superuser
account to help prevent mistakes.
version 56
page 116
Unix and Linux System Administration and Shell Programming
pipes
chapter 27
summary
This chapter looks at pipes.
Pipes are most commonly used as a method to chain different utilities together, with the output of one piped tot he input
of the next.
history
Ken Thompson invented the pipe in 1972.
The pipe is a standard mechanism for allowing the output of one program or process to be used as the input for another
program or process.
Pipes allow the UNIX philosophy of having many small programs that each do one function very well and then write
scripts that combine these small utility programs to accomplish bigger tasks.
how to pipe
To pipe the output of one UNIX command or utility to the next, simply place the | (vertical bar character) between the
two processes:
$ cat file01.txt | sort | lp
The example obtains the contents of the designated file, sorts the contents, and then sends the result to the printer.
version 56
page 117
Unix and Linux System Administration and Shell Programming
scripts
chapter 28
summary
This chapter looks at scripts.
Scripts are the power tools for system administrators. Scripts allow you to automate tasks.
shell initialization files
When you start Terminal (or, in some cases, run login ) the shell will run two initialization scripts:
/etc/profile
.profile
Note that these are the file names for BASH. See the login chapter for the default initialization file names for other shells.
The /etc/profile initialization script is maintained by the system administrator and is the first initialization script run
for all users.
Each user has their own .profile script in their home directory.
The .profile includes three basic operations (set terminal type, set path to commands, and set path to man pages) and
any customization created by the individual user.
Before the initialization scripts are run, the shell is considered to be an uninitialized shell.
.profile
The .profile initialization script sets the terminal type:
For example: TERM=vt100
The .profile initialization script sets the path to the directory or directories that contains the commands and utilities.
For example: PATH=/bin:/usr/bin
Notice the colon character ( :) is used as a separator between directories, allowing commands and utilities to be
organized in many directories.
The .profile initialization script sets the path to the directory or directories containing the manual ( man ) pages.
For example: MANPATH=/usr/man:/usr/share/man
interactive and noninteractive
When you directly enter a command or commands and the shell performs the commands for you, that is the interactive
mode.
You can also have the shell run scripts or commands in a noninteractive or batch mode. This is great for a system
administrator automating a server or other large scale Unix/Linux/Mac OS X system.
version 56
page 118
Unix and Linux System Administration and Shell Programming
When an non-interactive shell (or subshell) finishes running a script, it exits.
making a script executable
Scripts are just text files with lists of instructions.
To make a text file into an executable script that you can run, simply type chmod a+x filename .
$ chmod a+x filename
$
Also, add the magic line #!/bin/sh as the first line in the text file. This special magic line tells the shell which program
to use to run the script (in this case, the basic shell program).
Shell Files and Interpreter Invocation
File Extensions
from Google Shell Style Guide
Executables should have no extension (strongly preferred) or a .sh extension. Libraries must have a
.sh extension and should not be executable.
It is not necessary to know what language a program is written in when executing it and shell doesn’t
require an extension so we prefer not to use one for executables.
However, for libraries it’s important to know what language it is and sometimes there’s a need to have
similar libraries in different languages. This allows library files with identical purposes but different
languages to be identically named except for the language-specific suffix.
scripting languages
Some of the scripting languages available on Mac OS X are: sh, bash, perl, PHP, python, and ruby.
Examples of running scripts in these languages:
$
$
$
$
$
$
$
sh scriptfilename
bash scriptfilename
perl scriptfilename
php scriptfilename
python scriptfilename
ruby scriptfilename
shell script example
Create a directory (folder) for your personal scripts called scripts:
$ mkdir scripts
$
version 56
page 119
Unix and Linux System Administration and Shell Programming
Create a shell script called script.sh and save it in the new scripts directory (folder):
#!/bin/sh
echo "Hello World!"
Use the chmod command to make the new file an executable script:
$ chmod u+x /scripts/script.sh
$
Add the scripts directory to your command path:
$ export PATH="$PATH:~/scripts"
$
Run your new script:
$ script.sh
Hello World!
$
You can run your script directly in a specific shell:
$ rbash script.sh
$ sh script.sh
$ bash -x script.sh
php script example
This example assumes that you have created the scripts directory.
Create a php script called script.php and save it in the new scripts directory (folder):
<?php
echo "Hello World!"
?>
Notice that we skip the chmod step.
Run your new script by running the php program with your script as the file to execute:
$ php ~/scripts/script.php
Hello World!
$
version 56
page 120
Unix and Linux System Administration and Shell Programming
Note that the shell does not render HTML, so if you run a web scirpt, you will see raw HTML, CSS, and JavaScript as
plain text in your terminal window.
You can run perl, ruby, and python scripts in the same manner.
There is more complete information in the chapter on php and shell.
When to use Shell
from Google Shell Style Guide
Shell should only be used for small utilities or simple wrapper scripts.
While shell scripting isn’t a development language, it is used for writing various utility scripts
throughout Google. This style guide is more a recognition of its use rather than a suggestion that it be
used for widespread deployment.
Some guidelines:
If you’re mostly calling other utilities and are doing relatively little data manipulation, shell is an
acceptable choice for the task.
If performance matters, use something other than shell.
If you find you need to use arrays for anything more than assignment of ${PIPESTATUS}, you
should use Python.
If you are writing a script that is more than 100 lines long, you should probably be writing it in
Python instead. Bear in mind that scripts grow. Rewrite your script in another language early to
avoid a time-consuming rewrite at a later date.
version 56
page 121
Unix and Linux System Administration and Shell Programming
sysadmin and root/superuser
chapter 29
summary
This chapter looks at system administration and the root or superuser account.
As we look at more advanced features and commands, we need to become aware of the root or superuser account and
capabilities and how to safely access that power.
sysadmin
In the UNIX world, the system administrator is often called the sysadmin.
root/superuser
The root or superuser account has total authority to do anything on the system. This power is great for fixing problems,
but bad because one accidenttally mistyped character could be very destructive. Some systems also have admin accounts of
similar power.
Because of the potential for destructiveness, system administrators typically login with either a normal user account or a
limited admin account for every day work and then switch to superuser or root only when absolutely necessary (and then
immediately switch back).
The UNIX command for temporarily switching to root or superuser power is the sudo command, discussed in the next
chapter.
special powers
The root or superuser account has powers that “mere mortal” accounts don’t have.
The root account has access to commands that effect the entire computer or system, such as the ability to halt or
shutdown the system.
The root account is not affected by read and write file access permissions. The root or superuser account can create,
remove, read, or write any file anywhere on the system.
Some commands have built-in restrictions that the root or superuser can ignore. For example, the system administrator
can change any user’s password without knowing the old password.
version 56
page 122
Unix and Linux System Administration and Shell Programming
sudo
chapter 30
summary
This chapter looks at sudo , a Unix (and Linux) command.
WARNING: Never trust any Terminal/Shell commands you find on the internet. Only run shell
commands you understand. In particular, never run anything that includes sudo anywhere in the command line unless you
are absolutely certain what the command does. When you run a command line with sudo you are giving permission to have
complete (possibly destructive) control of your computer at the root level. And, yes, this advice applies even to this book.
Don’t run any commands with sudo unless you know for sure what you are doing.
sudo
The sudo command allows you to run a single command as another user, including at superuser or root level from a
normal account. You will be asked for the password before the command will actually run.
This keeps you firmly in a normal account (with less danger of catastrophic failures), while still giving easy access to
root or superuser power when really needed.
The sudo program was originally written by Bob Coggeshall and Cliff Spencer in 1980 at the Department of Computer
Science at SUNY/Buffalo.
sudo is a concatenation of su (substitute user) and do (perform an action).
To run a single command as superuser or root, type sudo followed by a command.
$ sudo command
You will normally be asked for your password (exceptions listed below).
sudo can be configured to not require a password (very bad idea other than single user personal systems). sudo can also
be configured to require the root password (rather than the current user’s password).
On Mac OS X the sudo command will fail if your account has no password.
On Mac OS X the sudo commands password prompt will not display anything (not even bullets or asterisks) while you
type your password.
You will not be asked for a password if you use sudo from the root or superuser account. You will not be asked for a
password if you use sudo and the target user is the same as the invoking user.
Some systems have a timer set (usually five minutes). You can run additional sudo commands without a password during
the time period.
run in root shell
To change to in the root shell, type sudo followed by the option -s. The following warning is from Mac OS X (entered a
root shell and then immediately returned to the normal shell). Note the change to the pound sign ( # ) prompt.
version 56
page 123
Unix and Linux System Administration and Shell Programming
$ sudo -s
WARNING: Improper use of the sudo command could lead to data loss
typing when using sudo. Type "man sudo" for more information.
To proceed, enter your password, or type Ctrl-C to abort.
Password:
bash-3.2# exit
$
other users
To run a command as another user, type sudo followed by the option -u followed by the user account name followed by
a command.
$ sudo -u username command
To view the home directory of a particular user:
$ sudo -u username ls ~username
edit files as www
To edit a file (this example is for index.html) as user www:
$ sudo -u www vim ~www/htdocs/index.html
which password
On most systems, you will authenticate with your own password rather than with the root or superuser password. The list
of users authorized to run sudo are in the file /usr/local/etc/sudoers or /etc/sudoers (on Mac OS X, /private/etc/sudoers).
These authorized users are identified in the sudoers file as admin .
The sudoers configuration file offers a wide variety of configuration options, including enabling root commands only
from the invoking terminal; not requiring a password for certain commands; requiring a password per user or per group;
requiring re-entry of a password every time for particular command lines; never requiring re-entry of a password for a
particular command line. The sudoers configuration file can also be set support the passing of arguments or multiple
commands and also supports commands with regular expressions.
timeout
sudo can set timeout limits. This is done with the timeout option. This can be configured globally, per user, or per
application. The timeout can be retained only per tty or globally per user. The user or application only has root
authentication until the timeout occurs.
forgot to use sudo
Sometimes you type a command and forget that you needed to use sudo until you see the error message. You can type
version 56
page 124
Unix and Linux System Administration and Shell Programming
sudo !!
to run the previous command with root privileges.
$ head /etc/passwd
head: /etc/passwd: Permission denied
$ sudo !!
unreadable directories
To view unreadable directories:
$ sudo ls /usr/local/protected
shutdown
To shutdown a server:
$ sudo -r +15 "quick reboot"
saving system file in vim
The ideal method for editing and saving a system file that can only be saved by the root user is to prepend the vim
command with sudo . Then the vim command :w will work because the vim program was launched with root privileges.
$ sudo vim /etc/passwd
$ some editing commands
$ :w
So, what do you do if you start editing the file and then remember that you need root permission to save it? Add !sudo
to the vim save command.
tee %
$ vim /etc/passwd
$ some editing commands
$ :w !sudo tee %
usage listing
To make a usage listing of the directories in the /home partition (note that this runs the commands in a sub-shell to make
the cd and file redirection work):
$ sudo sh -c "cd /home ; du -s * | sort -rn > USAGE"
view sudoers configuration
To view the sudoers current configuration settings, type:
version 56
page 125
Unix and Linux System Administration and Shell Programming
$ sudo -ll
editing sudoers file
Run the visudo command line tool to safely edit the sudoers configuration file. You will be presented with the vi editing
interface (this can be changed by setting the shell EDITOR environment variable to a different text editor, such as emacs .
Any syntax error in the sudoers configuration file will make sudo stop working globally. Therefore, always use visudo
to edit the sudoers file. visudo also provides security locks to prevent multiple simultaneous edits and other possible
security problems.
graphic equivalents
The utilities kdesudo (KDE) and gksudo (Gnome) provide a graphic user interface version of sudo (both are based on
sudo ). Mac OS X Authorization Services provides a graphic user interface with adminstrative privileges (but is not based
on the UNIX sudo ).
start a shell as root
If you need to do extended work as root, you can start up a root shell from your user account:
$ sudo bash
running commands as root
sudo only works for programs, not for any built-in commands. If you attempt it, you will get an error message stating
command not found . The solution is to start a root shell:
$ sudo bash
security
The system can be set up to send a mail to the root informing of unauthorized attempts at using sudo .
The system can be set up to log both successful and unsuccessful attempts to sudo .
Some programs (such as editors) allow a user to run commands via shell escapes, avoiding sudo checks. You can use
functionality to prevent shell escapes.
sudo ’s noexec
sudo never does any validation of the ARGUMENTS passed to a program.
sudo defaults to extensive logging, using the syslogd system log daemon to log all commands issued with sudo into a
cnetral host and local host file. This allows a complete audit trail of system access.
A system can be setup so that all machines in a system use the same sudoers file, allowing better central administration
of a network.
trace
You can’t sudo strace … (fill in the rest of the command any way you want) because sudo can’t gain its privileges
version 56
page 126
Unix and Linux System Administration and Shell Programming
while being traced.
polkit alternative
polkit (formerly PolicyKit) is an alternative control component for system-wide privileges.
SUID/SGID
from Google Shell Style Guide
SUID and SGID are forbidden on shell scripts.
There are too many security issues with shell that make it nearly impossible to secure sufficiently to
allow SUID/SGID. While bash does make it difficult to run SUID, it’s still possible on some platforms
which is why we’re being explicit about banning it.
Use sudo to provide elevated access if you need it.
version 56
page 127
Unix and Linux System Administration and Shell Programming
su
chapter 31
summary
This chapter looks at su, a Unix (and Linux) command.
su is used to switch user accounts.
switch to another user account
You can use the su command to switch another user account. This allows a system administrator to fix things signed in
to a particular user’s account rather than from the root account, including using the user’s settings and privileges and
accesses rather than those of root.
version 56
page 128
Unix and Linux System Administration and Shell Programming
who
chapter 32
summary
This chapter looks at who , a Unix (and Linux) command.
The who command was described in the first UNIX book, UNIX Programmer’s Manual, by Ken Thompson and Dennis
Ritchie, published November 3, 1971.
list of who is using a system
The who command will tell you all of the users who are currently logged into a computer. This is not particularly
informative on a personal computer where you are the only person using the computer, but it can be useful on a server or a
large computing system.
Type who followed by the ENTER or RETURN key.
$ who
admin console Aug 24 18:47
admin ttys000 Aug 24 20:09
$
The format is the login name of the user, followed by the user’s terminal port, followed by the month, day, and time of
login.
which account is being used
The command is who with the arguments am i will tell you which account you are currently using. This can be useful for
a system administrator if the system administrator is using the su command and wants to be sure about which the system
administrator is currently using.
$ who am i
admin ttys000 Aug 25 17:30
$
A related command without any spaces, whoami, lets you know which account you are currently logged in with.
$ whoami
admin ttys000 Aug 25 17:30
$
version 56
page 129
Unix and Linux System Administration and Shell Programming
advanced file systems
chapter 33
summary
This chapter is an advanced look at Unix (and Linux) file systems.
Some of the information in this chapter is a refresher carried over from previous chapters. Also see the chapter on major
directories.
Important reminder: Always use the root or superuser account sparingly. Do most of your work with an ordinary user
account and switch to the root account only when needed. And then switch back as soon as possible. The sudo command
can be used for temporarily switching to the root account to run a particular command.
absolute paths
It is common for a system administrator logged in to the root or superuser accunt to type out the absolute path to a
command or utility. This will skip any aliases. This will also skip over any commands in the user’s defined search path.
This guarantees that the correct command or utility will run.
everything is files
In UNIX, everything is a file, even hardware devices such as printers, hard drives, and other devices. Even UNIX
abstractions such as directories, pipes, and file redirection are treated as files.
Becuase everything is a file, the operating system can provide a common processing and communication interface. This
is essential to the UNIX approach of using lots of small programs or utilities that do one job well and connecting these
programs with shell scripts to do a wide variaety of activities.
man pages
Many forms of UNIX have a man page on filesystems.
$ man filesystems
If your system has this man page, you will find a general overview of UNIX file systems of all types, including
CDROMs, floppy disks, and tape drives.
diskless
Almost every UNIX system has a root directory and file system, which is stored on one or more hard drives connected to
the computer system.
There are diskless nodes. These are computers that don’t have their own local hard drive. Diskless nodes run off a
network server. Mac OS X makes it easy to set up a computer to run in this manner. This approach is often used when an
organization wants central control of all of the computers in the organization, such as a computer lab at a school.
root directory
version 56
page 130
Unix and Linux System Administration and Shell Programming
You can examine the contents of the root directory with the ls -F / command.
$ ls -F /
Applications/ Volumes/ mach_kernel.ctfsys
Desktop DB bin/ net/
Desktop DF cores/ private/
Developer/ dev/ sbin/
Library/ etc@ tmp@
Network/ home/ usr/
System/ mach-Kernal var@
Users/
The above example is for Mac OS X 10.5.8 Leopard.
On many systems you will find a file called unix This is the main UNIX operating system kernel.
On some systems you will find root.home . This is a temporary file created when running Netscape as root. This would
be an example of an application-specific file.
some important directories
The following is a brief description of some of the major directories. Not all systems will have all of these directories.
This list has additional information beyond the similar list in chapter on file system basics.
/ The root directory of the entire file system.
/bin A collection of binary files, better known as programs. Many of the standard system commands are stored here.
/boot or /kernel The files needed to boot or start your computer.
/dev The devices connected to your computer (remember, everything, even hardware devices, is treated as a file in UNIX
or Linux). Examples are the keyboard, disks, and printers.
/etc System configuration files, startup procedures, and shutdown procedures.
/home or /users The home directories for each user.
/lib or /Library Library files used by executable programs.
/net Other networked systems (again, treated as files).
/opt Third party software. In older systems this might be /usr/local
/private On Mac OS X, this is the location of the UNIX files. It is normally kept hidden from all users in Finder, but is
visible in your terminal emulator and shell. Mac OS X has links to the major UNIX directories at the root level so that
UNIX and Linux tools can find their files in the standard locations.
/proc Information about processes and devices.
/sbin System binaries (that is, system programs).
/tmp Space for temporary files. Required by all forms of UNIX. This directory is normally erased on bootup.
/usr User binaries (programs), libraries, manuals, and docs. Common subdirectories include online help (/usr/share),
header files for application development (/usr/include), and system configuartion files related to low-level hardware (such
as /usr/cpu and /usr/gfx).
version 56
page 131
Unix and Linux System Administration and Shell Programming
It is rare that a system administrator would even look at the low level hardware directories, such as /usr/cpu. The system
automatically manages these directories and their contents. It is unwise for a system administrator to modify these
directories and their contents. Messing with these will almost certainly bring the system down and may require reinstalling
the entire system.
/var Variable files. Some common subdirectories include X Windows files (/var/X11), system services files, applications
related files (/var/netscape), system administration files and data (/var/adm or /var/spool), a second temporary directory
(/var/tmp) that is not normally erased on bootup.
Swap Virtual memory on a hard drive. Allows the memory manager to swap some data and program segments to hard
drive to expand the amount of memory available beyond the limits of physical memory.
For a more advanced and complete listing, see the chapter on major directories.
FAQs
You can find a FAQ (Frequently Asked Questions) file for all UNIX variants. Find the appropriate FAQ for your version
of UNIX or Linux and look up the details for your system.
/etc
A key directory for new system administrators to learn is the /etc directory. The /etc directory contains key system
administration files.
Because the /etc files change system behavior and properties, a system administrator spends a lot of time modifying and
updating these files.
A new system administrator can learn a lot about UNIX by studying the files in the /etc directory. Do not change these
files until you have figured out what you are doing. This is not the place for random experimentation.
A few key /etc files a system administrator should know.
/etc/sys_id is the name of the system. It may contain the full domain. Not used on Mac OS X.
/etc/hosts is the summary of the full host names. A system administrator will add to this standard file.
/etc/fstab is a list of file systems to mount on bootup. Mac OS X has /etc/fstab.hd .
/etc/passwd is the password file. It includes a number of items of basic information on user accounts.
/etc/group is the group file. It contains details of all user groups.
/var
Another important directory for system administrators to know is the /var directory, which contains configuration files
for many system services.
For example, the /var/named file configures Domain Name Service and the /var/yp file configures Network
Information Service.
where to find answers
Surprisingly, there won’t be a grand tour of the UNIX file system here.
A UNIX system typically has thousands of files. Maybe tens of thousands of files. A system administrator doesn’t
normally need to know what these files are or where they are located.
version 56
page 132
Unix and Linux System Administration and Shell Programming
A system administrator should know where to find relevant information when a problem occurs.
Many system administrators keep a handy notebook with information on where to find answers to various problems. It is
far more useful to know where to find correct answers than to know try to memorize all answers.
And don’t forget popular search engines as a source for answers. There is very little chance that you are the first system
administrator in the world to encounter a particular problem. A quick internet search should reveal many others asking for
help on the same problem, along with answers.
hidden files and directories
Any file or directory that starts with the period or dot ( . ) character is considered hidden.
In MS-DOS (and Windows) and the original Macintosh, files and directories are hidden by setting a special attribute.
Mac OS X has a hybrid. When using the command line shell, Mac OS X acts exactly like any other version of UNIX.
But when using the Finder, Mac OS X also recognizes the invisible bit. This is how the entire UNIX operating system is
kept out of the hands of the typical Mac OS X user.
Access permissions are separate from invisibility. In MS-DOS, a hidden file can not be written to. In UNIX, you can
freely read and write and otherwise modify a hidden file.
In UNIX, file access is controlled by access permissions.
As a reminder from the ls chapter, the -a option will give a list of all files, including hidden files.
$ ls -a
Each user has a few important hidden files that configure their basic working environment when they login. These
include:
.cshrc
.login
.profile
version 56
page 133
Unix and Linux System Administration and Shell Programming
major directories
chapter 34
summary
This chapter looks at some of the major directories (folders) on Unix/Linux/Mac OS X.
directory listing
The following is a brief description of some of the major directories. Not all systems will have all of these directories.
/ The root directory of the entire file system.
/Applications On Mac OS X, this is the location of the Macintosh programs.
/Applications/Utilities On Mac OS X, this is the location of the graphic user interface based utility programs.
/bin A collection of binary files, better known as programs. These are the programs needed by the operating system that
normal users might use. Many of the standard system commands are stored here. Contrast with /sbin below.
/boot or /kernel The files needed to boot or start your computer. Includes the bootstrap loader, the very first thing loaded
when your computer starts. On Linux, it usually includes the Linux kernel in the compressed file vmlinuz.
/dev The devices connected to your computer (remember, everything, even hardware devices, is treated as a file in UNIX
or Linux). Examples are the keyboard, disks, and printers.
/etc System-wide configuration files, startup procedures, and shutdown procedures.
/home or /users The home directories for each user. Contains such things as user settings, customization files,
documents, data, mail, and caches. The contents of this directory should be preserved during operating system updates or
upgrades.
Every new user account starts with a few basic files such as .login, .cshrc, and .profile . A system administrator may
want to add additional customization such as an introductory README file.
/lib or /Library Shared library files and kernel modules.
/lost+found Files recovered during filesystem repair.
/mnt Mount points for removable media (floppy disks, CD-ROM, Zip drives, etc.), partitions for other operating
systems, network shares, and anything else temporarily mounted to the file system. Linux and UNIX don’t use Windowsstyle drive letters.
/net Other networked systems (again, treated as files).
/opt Optional large applications and third party software. In older systems this might be /usr/local.
/private On Mac OS X, this is the location of the UNIX files. It is normally kept hidden from all users in Finder, but is
visible in your terminal emulator and shell. Mac OS X has links to the major UNIX directories at the root level so that
UNIX and Linux tools can find their files in the standard locations.
/proc A Linux-only directory. Information about processes and devices. The files are illusionary. The files don’t exist on
the disk, but are actually stored in memory. Many Linux utilities derive their information from these files.
version 56
page 134
Unix and Linux System Administration and Shell Programming
/sbin System binaries (that is, system programs). These are programs that general users don’t normally use (although
they can). This directory is not in the PATH for general users. Contrast with /bin listed above.
/root The home directory for the system administrator (superuser or root).
/tmp Space for temporary files. Required by all forms of UNIX. This directory may automatically be cleaned on a
regular basis. This directory is normally erased on bootup.
/usr User binaries (programs), libraries, manuals, and documentation (docs). Common subdirectories include online help
(/usr/share), header files for application development (/usr/include), and system configuartion files related to low-level
hardware (such as /usr/cpu and /usr/gfx).
/var Files that change, such as spool directories, log files, lock files, temporary files, and formatted (on use) manual
pages. Some common subdirectories include X Windows files (/var/X11), system services files, applications related files
(/var/netscape), system administration files and data (/var/adm or /var/spool), a second temporary directory (/var/tmp) that
is not normally erased on bootup.
Swap Virtual memory on a hard drive. Allows the memory manager to swap some data and program segments to hard
drive to expand the amount of memory available beyond the limits of physical memory.
version 56
page 135
Unix and Linux System Administration and Shell Programming
Network File System (NFS)
chapter 35
summary
This chapter looks at Network File System (NFS) on Unix/Linux/Mac OS X.
The Network File System (NFS) allows a directory on one machine to be accessed from another machine. This is an
important feature of UNIX.
The procedure for setting up the Network File System is called mounting.
NFS provides a seemless method to merge different file systems from different physical machines into a large shared
structure. NFS works across different kinds of UNIX or Linux and even different kinds of processors.
With NFS, a user doesn’t need to know or keep track of the actual physical locations of directories and files. The NFS
files and directories appear to be a normal part of the file system and are accessed with the exact same methods as any
local file or directory.
mount/export
The mount command will replace a local directory with a directory from another machine. Anyone viewing the mounted
directory will see the remote directory as if it were local.
The export command is used to make a directory available for other computers to mount it.
The /etc/exports file contains a list of directories to be exported.
mount should be run from root.
availability
Network File System (NFS) was developed by Sun Microsystems in 1984 and is based on the Open Network Computing
Remote Procedure Call (ONC RPC) system.
NFS runs on most UNIX and UNIX-like systems (including Solaris, AIX, Free BSD, HP-UX, Linux, and Mac OS X), as
well as classic Macintosh, Microsoft WIndows, Novell NetWare, and IBM AS/400.
Alternative remote file access protocols include Server Message Block (SMB, also called CIFS), Apple Filing Protocol
(AFS), NetWare Core Protocol (NCP), and OS/400 File Server file system (QFileSvr.400).
version 56
page 136
Unix and Linux System Administration and Shell Programming
tail
chapter 36
summary
This chapter looks at tail , a Unix (and Linux) command.
tail is used to report the last lines of a text file.
syntax:
tail [option…] [file…]
shells:
ash
File Name: tail
bash
bsh
csh
ksh
sh
Directory: /usr/bin/
tcsh
Type:
zsh
External
default
The default operation of tail is the last 10 lines of a text file to standard output (the example is from a file with less
than 10 lines).
$ tail xml2Conf.sh
#
# Configuration file for using the XML library in GNOME applications
#
XML2_LIBDIR="-L/usr/lib"
XML2_LIBS="-lxml2 -lz -lpthread -licucore -lm "
XML2_INCLUDEDIR="-I/usr/include/libxml2"
MODULE_VERSION="xml2-2.7.3"
$
example: most recent user
Here is how to use the tail utility to show the last user created on the current system:
$ tail -1 /etc/passwd
number of lines
You can use the -n option to set the number of lines, blocks, or bytes to output.
The option tail -n1 filename shows the last line.
$ tail -n1 xml2Conf.sh
MODULE_VERSION="xml2-2.7.3"
version 56
page 137
Unix and Linux System Administration and Shell Programming
$
The --lines=N option is equivalent to the -nN option.
An obsolete format (still used on Sun Solaris, where the -n option is not supported) uses just the number of lines as the
option.
$ tail -1 xml2Conf.sh
MODULE_VERSION="xml2-2.7.3"
$
number of characters
The option tail -c8 filename shows the last eight (8) characters (bytes).
$ tail -c8 xml2Conf.sh
-2.7.3"
$
Notice in the example that you only see seven characters. The newline character is the eighth character.
The --bytes=N option is equivalent to the -cN option.
An obsolete format (still used on Sun Solaris, where the -c option is not supported) uses the number of character
followed by a c as the option.
$ tail -8c xml2Conf.sh
-2.7.3"
$
suppressing file names
If you provide tail with more than one file name, then it report the file names before the last lines of each file.
$ tail -n2 *Conf.sh
==>. xml2Conf.sh <.==
XML2_INCLUDEDIR="-I/usr/include/libxml2"
MODULE_VERSION="xml2-2.7.3"
==>. xml1Conf.sh <.==
XML2_INCLUDEDIR="-I/usr/include/libxml1"
MODULE_VERSION="xml2-2.7.1"
$
You can suppress the file name(s) with the --silent option. This is particularly useful if you are going to pipe the
results to other Unix tools.
$ tail -n2 --silent *Conf.sh
XML2_INCLUDEDIR="-I/usr/include/libxml2"
MODULE_VERSION="xml2-2.7.3"
version 56
page 138
Unix and Linux System Administration and Shell Programming
XML2_INCLUDEDIR="-I/usr/include/libxml1"
MODULE_VERSION="xml2-2.7.1"
$
The --quiet and -q options are equivalent to the --silent option.
file (log) monitoring
Use the -f (follow) option to monitor a log file. With the -f option, the tail utility will continue to monitor the file and
send any new lines to standard out (unless redirected or piped elsewhere).
$ tail -f importantlogfile.txt
In most versions of Unix, the --follow or --follow=descriptor options are equivalent to the -f option.
If the file being monitored might be rotated, use the -F (follow) option to monitora file, even if the orginal version of the
file is renamed or removed and a new version of the file (with the same name) is created.
$ tail -F importantlogfile.txt
Use Control-C to interrupt the file monitoring.
You can append the ampersand ( & to make the file monitoring a background process.
GNU Emacs emulates this Unix utility with the auto-revert-tail-mode mode.
In most versions of Unix, the --follow=name --retry option is equivalent to the -F option.
version 56
page 139
Unix and Linux System Administration and Shell Programming
wc
chapter 37
summary
This chapter looks at wc, a Unix (and Linux) command.
wc is used to count lines, characters, and words in a file.
word count command
Use the wc command to count the number of words in your file.
$ wc names
6 6 38 names
$
The format is the number of lines ( 6), followed by the number of words ( 6), followed by the number of characters ( 38),
followed by the name of the file ( names ).
multiple files
You can use the wc command with several file names.
$ wc names file01.txt
6 6 38 names
9 29 134 file01.txt
15 35 172 total
$
The output gives the results for each file (in the order they were listed in the command) and a final line with the totals for
al of the files combined.
count the number of lines
Use the wc command with the -l option to count the number of lines in your file.
$ wc -l names
6 names
$
The output is the number of lines and the name of the file.
count the number of words
Use the wc command with the -w option to count the number of words in your file.
version 56
page 140
Unix and Linux System Administration and Shell Programming
$ wc -w names
6 names
$
The output is the number of words and the name of the file.
count the number of characters
Use the wc command with the -m or -c option to count the number of characters in your file.
Solaris and HP-UX use -m.
Linux uses -c.
Mac OS X accepts both -m and -c.
$ wc -c names
38 names
$
The output is the number of characters and the name of the file.
multiple options
You can use multiple options in a single command. The order does not matter. For example, all of the following
variations will count the number of lines and words:
$ wc -l -w filename
$ wc -w -l filename
$ wc -lw filename
$ wc -wl filename
chmod
version 56
page 141
Unix and Linux System Administration and Shell Programming
chapter 38
summary
This chapter looks at chmod , a Unix (and Linux) command.
chmod is used to change the permissions for a file or directory.
The chmod command was described in the first UNIX book, UNIX Programmer’s Manual, by Ken Thompson and Dennis
Ritchie, published November 3, 1971.
wide open
The chmod 777 filename command will set the permissions so that filename is wide open to everyone.
$ chmod 777 filename
The file owner, the file owner’s group, and other (or world) will all be able to read, write, delete, modify, and execute
(run) the file to their heart’s content. This access includes anyone who gets into your system through the internet.
chown
chapter 39
version 56
page 142
Unix and Linux System Administration and Shell Programming
summary
This chapter looks at chown , a Unix (and Linux) command.
chown is used to change the owner or group of a file.
The chown command was described in the first UNIX book, UNIX Programmer’s Manual, by Ken Thompson and Dennis
Ritchie, published November 3, 1971.
giving away a file
You can use the chown to change the ownership of a file or directory, giving ownership away to someone else.
The following example gives away ownership of the filename to the account newowner.
warning
Just read — do not type
this will give away your file and you have to use root to get it back
$ chown newowner filename
Of course, a system administrator can use root to freely change ownership of files and directories. This is a powerful tool
in the hands of a system administrator.
security leak
This command is often limited to only root (or superuser) because if a computer system or server enforces limits on how
much disk storage space each account is allowed, someone can set a file so they still can read and write it (and possibly
execute it, if applicable), then change ownership to anyone else. This gets the file (or entire directories) out of their storage
limits, but they can still make use of the files.
shell levels and subshells
chapter 40
version 56
page 143
Unix and Linux System Administration and Shell Programming
summary
This chapter looks at shell levels and subshells.
As you advance to more sophisticated operations, such as the material on substitutions and variables, you will find that
the bash shell has levels.
Every program, command, or script that you run can have and modify variables. How does Unix/Linux keep programs,
commands, and scripts from interfering with each others’ variables?
The answer is shell levels. Each time the shell starts a program, command, or script, it starts it in a subshell. The subshell
has its own variables and environment. It can’t interfere with any other program, command, or script, even if they use the
same name for a variable.
Most of the time this system is great It can lead to frustration when attempting to share variables (you will have to use
some work-around). And it can trip up beginners who aren’t expecting this.
Technically, each subshell is run as a child process. We will discuss this more in the chapter on processes.
shell levels
You can use the $SHLVL built-in variable to detemrine the current shell level.
In the following example, we will start up bash in a subshell. Yes, you can start-up another copy of bash from the
current copy of bash .
This example includes some materials that will be covered in more detail in later chapters.
$
1
$
$
2
$
$
1
$
echo $SHLVL
bash
echo $SHLVL
exit
echo $SHLVL
The first echo command shows the current shell level (1).
The bash command starts up another copy of bash in a subshell.
The second echo command shows the subshell level (2).
The exit command closes down the copy of bash running in the child process and returns us to the main copy of bash .
The third echo command shows the original shell level (1).
version 56
page 144
Unix and Linux System Administration and Shell Programming
substitutions
chapter 41
summary
This chapter looks at substitutions in Unix (and Linux).
command substitutions
arithmetic substitutions
command substitutions
chapter 42
summary
version 56
page 145
Unix and Linux System Administration and Shell Programming
This chapter looks at command substitutions in Unix (and Linux).
command substitution
Command substitution is used to assign the output of a command to a variable.
Place the command in sideways ticks ( `) around the command. Do not confuse these with regular single quotation marks
( ').
You can place a simple command, a pipeline, or a command list inside the tick marks.
An example of a simple command:
$ DATE=`date`
An example of a pipeline:
$ CONSOLEUSER=`who | grep console`
An example of a command list:
$ FILEUSAGE=`date ; df`
You can use command substitution to create parameters for other commands.
In the following example, the user name is used for grep word search of the file named names :
$ grep `id -un` names
You can use $( ) as a replacement for ` `. If you nest backticks inside of each other, you need to escape the internal
backticks. You do not need to escape nested $( ).
$ DATEVAR=$(date)
$ echo $DATEVAR
$Mon Aug 26 19:35:47 PDT 2013
$
If there is only one level of nesting, some people use the convention of using $() for the outer expression and backicks
( ``) fr the inner expression.
version 56
page 146
Unix and Linux System Administration and Shell Programming
arithmetic substitutions
chapter 43
summary
version 56
page 147
Unix and Linux System Administration and Shell Programming
This chapter looks at arithmetic substitutions in Unix (and Linux).
arithmetic substitution
You can use arithmetic substitution for quick integer artihmetic.
Place an integer arithmetic expression inside double parenthesis and place a dollar mark before the parenthesized
expression (wow, that’s a lot of words), $(( expression )).
$(( 5 + 3 ))
You may use integer constants or integer variables:
$(( 5 + $z ))
You may post-increment, post-decrement, pre-increment, or pre-decrement variables:
$(( --x + z++ ))
You may use negative integers:
$(( -1 * -2 ))
You may use logical ( !) and bitwise ( ~) negation:
$(( !1 * ~2 ))
The order of precedence (and complete list of possible operations):
version 56
operator
VAR++ VAR-++VAR --VAR
-+
meaning
variable post-increment and pre-increment
variable pre-increment and pre-decrement
unary minus and plus
!~
**
logical negation and bitwise negation
exponentiation
*/%
+<< >>
<= >= < >
multiplication, division, and modulo
addition and subtraction
left bitwise shift and right bitwise shift
comparison operators
== !=
&
equality and inequality
bitwise AND
^
|
bitwise exclusive OR
bitwise OR
page 148
Unix and Linux System Administration and Shell Programming
&&
logical AND
||
expression ? expression : expression
logical OR
C-style conditional evaluation
= *= /= %= += -= <<= >>= &= ^= != assignments
,
separator between expressions
Applying the precedence rules:
$(( ((3 + 5*2) -8) /2 ))
The result for the above expression is 2. It is not 2.5 because this is integer arithmetic. It is not 4 because the
multiplcation has a higher precedence than addition.
You can have a raw expression, which is useful if it includes an assignment operator:
$
$
1
$
$
4
$
n=1
echo $n
(( n += 3 ))
echo $n
flow control
chapter 44
summary
version 56
page 149
Unix and Linux System Administration and Shell Programming
This chapter looks at flow control in Unix (and Linux).
The three basic kinds of instructions in structured programming are: sequence, decision, and loop.
A sequence is a series of instructions performed in the same order every time.
A decision makes a choice between two or more different possible sequences of instructions.
A loop repeats a sequence of instructions repeatedly.
The basic two flow control decision statements are if and case .
The basic three flow control loops are while , for , and select.
if
The basic format of the if statement:
if list1 ; then list2 ; elif list3 ; then list4 else list5 fi
case
The basic format of the case statement:
case expression in
pattern1) list1 ;;
…
patternN) listN ;;
esac
while
The basic format of the while loop:
while expression ; do list ; done
for
The basic format of the for loop:
for var in word1 … wordN ; do list ; done
select
The basic format of the select loop:
select var in word1 … wordN ; do list ; done
version 56
page 150
Unix and Linux System Administration and Shell Programming
management tools
chapter 45
summary
This chapter looks at management tools for UNIX and Linux.
version 56
page 151
Unix and Linux System Administration and Shell Programming
Every major version of UNIX and Linux includes several management tools. Some use a graphic interface, while others
are accessed through the command line. Some management tools will even provide both a command line and a graphic
interface.
Comparison chart
OS
AIX
HP-UX
RHEL
SLES
Solaris
Tru-64
Comprehensive
nmon, topas, lparmon
GlancePlus, MeasureWare/Perfview, Caliper
sysstat, systemTAP, oprofile
systat, SystemTap
SE Toolkit, sysperfstat
Collect, sys_check, HP inisght manager
RAM
CPU
I/O
Network
vmstat, svmon vmstat, sar iostat, filemon netstat, nfsstat
vmstat
top
iostat
netstat
vmstat
top, mpstat iostat
netstat, lperf
vmstat
top,mpstat iostat
netstat, lperf
vmstat
top
iostat
netstat
vmstat
top
iostat
netstat
high availability
Some high availability tools for specific operating systems also have a Linux version.
Veritas for High Availability on Solaris also comes with a Linux client.
PowerHA (formerly known as HACMP) for IBM’s AIX also is available in a Linux version.
df
chapter 46
summary
This chapter looks at df, a Unix (and Linux) command.
version 56
page 152
Unix and Linux System Administration and Shell Programming
df is used to display information on disk usage.
human readable
Use the -h or -H option to see space used in human readable format (such as kilobytes, megabytes, gigabytes, terabytes,
petabytes, etc.). Both options do the same thing.
$ df -h
Filesystem Size Used Avail Capacity Mounted on
/dev/disk0s10 149Gi 117Gi 32Gi 79% /
devfs 111Ki 111Ki 0Bi 100% /dev
fdesc 1.0Ki 1.0Ki 0Bi 100% /dev
map -hosts 0Bi 0Bi 0Bi 100% /net
map auto_home 0Bi 0Bi 0Bi 100% /home
/dev/disk1s2 498Gi 40Gi 457Gi 9% /Volumes/msdos
/dev/disk1s1 434Gi 81Gi 353Gi 19% /Volumes/Mac FreeAgent GoFlex Drive
$
The exact format of the human readable output will vary from system to system, but the columns are usually labelled (as
in the Mac OS X example above). The file system path is a reference to a hard drive, storage device, network, or other
location. The mount point is the location in the directory tree where you can find the file system.
type
Use the -T option to see the type of file system.
$ df -T
You can also provide specified filesystem types after the -T option to only show filesystems of the designated type. Use a
comma separated list to provide multiple filesystem types. Pre-fix with no to show file systems other than the listed types.
As an example, the following command will display all file systems other than those of type NFS and MFS:
$ df -T nonfs,mfs
Use the lsvfs command to find out what types of filesystems are available on your computer or server.
differences between
du
and df
You may notice that du and df sometimes give very different numbers for the disk space on a production server. This
rarely happens on a desktop or workstation. Usually df will output the bigger disk usage. This occurs when an inode is
deallocated.
Some process has created a large temporary file and some other process has deleted it. The file remains in existence until
the first process closes it, but the file system immediately removes access for any other process (to prevent all kinds of
strange errors by accessing or manipulating non-existent files.
Running lsof | grep tmp or lsof | grep deleted will often reveal a large temporary file that accounts for the
difference.
du and df should resolve back to the same results after the first process releases the file. If the problem persists, use fsck
to fix it.
version 56
page 153
Unix and Linux System Administration and Shell Programming
Also note that adding the used and available space will not necessarily be an exact match with the total. One explanation
is that a journaling system will account for up to 5% of the disk space.
blocksize
Use the -b or -P option to see use the default of blocks. This may be needed to override a BLOCKSIZE specification
from the environment. In the early days of UNIX and Linux, the default block size was normally 512-bytes. In more
modern times the size of a block can be 1K, 8K, 32K, or even bigger.
$ df -b
result in 512 byte blocks
Use the -k option to see use 1023-byte blocks. This is 1-Kbyte blocks.
$ df -k
result in Kilo-byte blocks
Use the -m option to see use 1048576-byte blocks. This is 1-Mbyte blocks.
$ df -m
result in Mega-byte blocks
Use the -g option to see use 1073741824-byte blocks. This is 1-Gbyte blocks.
$ df -g
result in Giga-byte blocks
display inodes
Use the -i option to include statistics on the number of free inodes.
$ df -i
On Mac OS X, the additional columns are iused , ifree , and %iused.
local
Use the -l option to only display information about locally-mounted filesystems (no network fielsystems).
$ df -l
long delay filesystems
version 56
page 154
Unix and Linux System Administration and Shell Programming
Use the -n option to display previously obtained statistics. This is used when one or more filesystems are in a state
where there will be a long delay to provide statistics.
$ df -n
du
chapter 47
summary
This chapter looks at du, a Unix (and Linux) command.
du is used to display usage.
version 56
page 155
Unix and Linux System Administration and Shell Programming
display usage
Use the -a option to see an entry for each file in the file hierarchy.
$ du -a
Note that if you run this from root, you will get a large report of every single file on all of the mounted hard drives. best
to run it from a subdirectory.
total
Use the -c option to see a grand total.
$ du -c
kilobytes
Use the -k option to see the results listed in kilobytes.
$ du -k
You can use the alias command to modify the du so that it always displays the file information in kilobytes.
$ alias du=du -k
If you need to use the unmodified version of this now aliased command, simply type a slash ( \ ) in front of the du
command.
$ \du
A better alternative is to use an alternate spelling, such as duk for your alias. Any scripts that rely on the unmodified
version of an aliased du are likely to fail or produce strange results because of your alias.
$ alias duk=du -k
version 56
page 156
Unix and Linux System Administration and Shell Programming
processes
chapter 48
This chapter looks at Unix (and Linux) processes.
The UNIX term process is pretty much the same as the terms job or task on other operating systems.
multi-tasking
UNIX is a multi-processing operating system. This means that it runs more than one task (or job or process) at a time.
version 56
page 157
Unix and Linux System Administration and Shell Programming
On a multi-processor computer it is possible for UNIX to literally run multiple processes at the exact same time. On a
single processor computer different processes are swapped into the main CPU (central processing unit) so fast that it gives
the illusion that multiple processes are running simultaneously.
One of the reasons that multitasking works is that the processor runs much faster than input/output or mass storage
devices (and on modern computers often runs much faster than main memory). Whenever a process has to stop for input,
output, or mass storage, there is a convenient moment for the operating system to switch to a different process that is ready
to run rather than leave the CPU sitting idle waiting for the I/O to occur.
init
The first process to start running on a traditional UNIX computer is init . Some modern variations, such as Linux, may
have a few other processes run first, especially.if the computer or server supports multiple operating systems.
init starts a series of other processes (in UNIX, this is called spawning). Some of these processes only run during bootup, while others are intended to run as long as the computer is running.
By the time a human user can login to the system, many processes are running.
init is ultimate ancestor process of all other processes in a UNIX system.
process ID
Every process on UNIX (or Linux) has a unique process identification number, called the process ID.
A system administrator uses the process IDs to control the processes.
kill
One useful command available for the system administrator is the kill command, which stops a process and removes it
from the system.
An ordinary user can only kill his or her own processes.
A system administrator running as root (including using sudo ) can use kill on any process on the entire system.
See the chapter on kill for more information.
ps
chapter 49
summary
This chapter looks at ps, a Unix (and Linux) command.
ps gives information about running processes.
version 56
page 158
Unix and Linux System Administration and Shell Programming
info on processes
Type ps followed by the option -a (for “all”). An optional pipe to more or less will provide one page of info at a time.
$ ps -a |more
ps can be used to monitor the use of a server or system.
ps can be used to find processes that are stuck, which can then be killed.
monitor processes
ps -ax will provide information on most running processes. You can use the command to establish the baseline
performance for your server and to monitor errant processes.
$ ps -ax
full information
Type ps followed by the -ef options to see most of the available information on processes.
$ ps -ef |less
On a Solaris system you will want to use the -aux options.
$ ps -aux |less
searching for specific content
You can use the following example to monitor a specific string or name from the overall output (where term is replaced
with the specific string or name you want):
$ ps -ef|grep term
daemons
On Linux you use the following to get information on core processes and daemons:
$ /sbin/chkconfig --list
load balancing
Most of the daily work for a system administrator is making sure that system resources aren’t overloaded so that
everythign runs smoothly.
version 56
page 159
Unix and Linux System Administration and Shell Programming
Sometimes a program will hog the CPU. This may be due to an infinite loop, repeatedly forking itself, or other bug, or it
may be due to the nature of the processing being done.
In either case, the system administrator can use the nice command to adjust the CPU priority of each runnign process so
that the entire system runs smoothly.
kill
chapter 50
summary
This chapter looks at kill , a Unix (and Linux) command.
kill is used to send signals to a process, most often to stop processes.
kill is a builtin command in csh . There is also an external utility with the same name and functionality.
version 56
page 160
Unix and Linux System Administration and Shell Programming
which processes can be killed
An ordinary user can only kill his or her own processes.
A system administrator running as root (including using sudo ) can use kill on any process on the entire system.
frozen process
Sometimes a process (particularly web browsers) will freeze up. The kill command can be used to remove the frozen
program from the system.
If a user finds their display locked up because of a frozen program, he or she can login to another computer, then login
to the original system using rlogin, and then use the kill command to stop the offending process using the specific
process ID.
killall
killall is a related command. killall program_name will stop all processes with the program name, so the more
specific kill with a specific process ID is usually used first.
process isolation
Unix does a very good job of keeping processes isolated from each other, so that no rogue process can attack or damage
any other process.
Unlike Windows, a frozen or crashed program on UNIX or Linux shouldn’t be able to bring the entire operating system
to a halt.
Most Unix vendors carefully test their own system software and most of the system software has been running and tested
for decades, so those programs rarely cause problems.
Most bugs occur in third party software and in custom software developed for a particular installation.
remove background process or job
The most common use of kill is simply to remove a process or job. This is done by naming the PID of the process
(whcih is usually obtained with ps).
$ kill PID
This works because the default signal is SIGTERM (terminate the process).
discover signals
You will find a chart of signals with the common signals for Linux, Mac OS X, and Solaris.
You can also discover the list for your system by running kill with the -l (lower case L) option.
$ kill -l
You can determine a specific signal with the -l option and a signal number.
version 56
page 161
Unix and Linux System Administration and Shell Programming
$ kill -l 3
QUIT
$
Some operating systems give the entire signal name with the SIG prefix (such as SIGQUIT) and some give just the
signal name (such as SIGQUIT).
testing
You can test the capabilities of your programs and scripts under abnormal coonditions by using kill to send a specific
signal.
$ kill -s SIGNAME PID
The signal name (SIGNAME) can be the signal number or the signal name. The signal name can be either the signal
name with or without the SIG prefix.
You may also simply enter the signal number or signal name (with or without the SIG prefix) as an option.
$ kill -6 PID
$ kill -SIGIOT PID
$ kill -IOT PID
nice
chapter 51
summary
This chapter looks at nice , a Unix (and Linux) command.
load balancing
Most of the daily work for a system administrator is making sure that system resources aren’t overloaded so that
everythign runs smoothly.
version 56
page 162
Unix and Linux System Administration and Shell Programming
Sometimes a program will hog the CPU. This may be due to an infinite loop, repeatedly forking itself, or other bug, or it
may be due to the nature of the processing being done.
In either case, the system administrator can use the nice command to adjust the CPU priority of each runnign process so
that the entire system runs smoothly.
w
chapter 52
summary
This chapter looks at w, a Unix (and Linux) command.
w gives information about current activity on the system, including what each user is doing.
Not to be confused with who , which displays users, but not the processes they are running.
info
version 56
page 163
Unix and Linux System Administration and Shell Programming
Type w:
$ w
w does not provide information on background processes, which typically account for most of the server usage.
date
chapter 53
summary
This chapter looks at date , a Unix (and Linux) command.
date is used to set the system date and time.
$ date
Mon Sep 3 00:56:17 PDT 2012
$
version 56
page 164
Unix and Linux System Administration and Shell Programming
The format for the output is: day of the week, month, day of the month, 24 hour time, time zone, year.
formatted output
Display the time and/or date with formatting by including formatting options (which can be used to set variables to a
specific format).
$ date '+DATE: %m/%d/%y%nTIME: %H:%M:%S'
DATE: 11/23/13
TIME: 22:12:19
$
Setting a variable with the current date:
$ NOW=$(date +"%m/%d/%Y")
$ echo $NOW
11/23/2013
$
Format specifiers (format string starts with +)
Description
Values or example
Day
Specifier
%a
weekday, abbreviated
Sun
%A
%d
weekday, full
day of the month, two digits, zero filled
Sunday
08
%e
day of the month
8
%j
day of year, zero filled
001Ð366
%u
day of week from Monday to Sunday
1Ð7
%w
day of week from Sunday to Saturday
0Ð6
Week
%U
week number, Sunday as first day of week
00Ð53
%W
week number, Monday as first day of week
00Ð53
%V
ISO standard week of the year
01Ð53
Month
%m
two-digit month number
01Ð12
%h
month name, abbreviated
Nov
%b
month name, localised abbreviation
Nov
%B
locale's full month, variable length
November
Year
%y
two-digit year
00Ð99
%Y
four-digit year
2013
%g
two-digit year corresponding to the %V week number
%G
four-digit year corresponding to the %V week number
Century
%C
two century digits from year
version 56
00Ð99
page 165
Unix and Linux System Administration and Shell Programming
Date
%D
mm/dd/yy
11/24/13
%x
locale's date representation
11/24/2013
%F
%Y-%m-%d
2013-11-24
Hours
%l
hour (12 hour)
4
%I
hour (12 hour), zero-filled
04
%k
hour (24 hour)
4
%H
hour (24 hour), zero-padded
04
%p
locale's upper case AM or PM (blank in many locales)
AM
%P
locale's lower case am or pm
am
Minutes
%M
two-digit minute number
05
Seconds
%s
seconds since 00:00:00 1970-01-01 UTC (Unix epoch)
1385265929
%S
two-digit second number
00Ð60 (Includes 60 to accommodate a leap second)
%N
nanoseconds
000000000Ð999999999
Time
%r
hours, minutes, seconds (12-hour clock)
04:05:29 AM
%R
hours, minutes (24 hour clock)
04:05
%T
hours, minutes, seconds (24-hour clock)
04:05:29
%X
locale's time representation
11:07:26 AM
Date and time
%c
locale's date and time
Sat Nov 04 12:02:33 EST 1989
Time zone
%z
RFC-822 style numeric time zone
-0500
%Z
time zone name; nothing if no time zone is determinable EST, EDT
literals: %n newline %% percent %t horizontal tab
By default, date normally fills numeric fields with zeroes. GNU date, but not BSD date, recognizes a modifier between
the per cent sign (%) and the format specifier:
hyphen (-): do not fill the field
underscore (_): pad the field with spaces
TZ Specifies the time zone, unless overridden by command line parameters. If neither is specified, the setting from
/etc/localtime is used.
setting time and date
Only the root or superuser an set the system date and time. In Mac OS X, you can use the clock system preferences t set
the time and date. In Ubuntu-based Linux, you can click on the clock and select Time and Date settings from the menu or
click on the System menu, select Adminsitration, select Time and Date.
Set the time to noon:
version 56
page 166
Unix and Linux System Administration and Shell Programming
$ date 1200
$
Set the time to 3:30:30 a.m.:
$ date 0330.30
$
Set the date to October 31st (Halloween) at 3:30 a.m.:
$ date 10310330
$
uname
chapter 54
summary
This chapter looks at uname , a Unix (and Linux) command.
uname is used to report some basic system information.
On a Linux machine you can find all of this same information in the appropriate files in the /proc filesystem, but uname
utility may be easier and faster to use.
basic use
Type the uname without any options to get the kernel name.
version 56
page 167
Unix and Linux System Administration and Shell Programming
$ uname
Darwin
system
Use the -s option to get the operating system name.
$ uname -s
Darwin
release info
Use the -r option to get the operating system release information.
$ uname -r
9.8.0
operating system version
Use the -v option to get the operating system version.
$ uname -v
Darwin Kernel Version 9.8.0: Wed Jul 15 16:57:01 PDT 2009; root:xnu1228.15.4~1/RELEASE_PPC
machine
Use the -m option to get the machine hardware name.
$ uname -m
Power Macintosh
When running on an Intel (or compatiable) processor, you may also get an indication as to whether it is a 32-bit or 64-bit
system.
$ uname -m
x86_64
processor
Use the -p option to get the machine processor architecture type.
$ uname -p
powerpc
version 56
page 168
Unix and Linux System Administration and Shell Programming
network node
Use the -n option to get the network node host name (node name).
$ uname -n
local
You can also use the hostname command to get the exact same answer.
all
Use the -a option to get all of the information available from this command.
$ uname -a
Darwin admins-power-mac-g5.local 9.8.0 Darwin Kernel Version 9.8.0: Wed Jul 15
16:57:01 PDT 2009; root:xnu-1228.15.4~1/RELEASE_PPC Power Macintosh
nonstandard
There are a few additional nonstandard options that may be available. Even if available, they may produce unknown as
their output. because these are nonstandard options, you should avaoid using them in a script.
Use the -i option to get hardware platform information.
$ uname -i
x86_64
Use the -o option to get he operating system name.
$ uname -o
GNU/Linux
CentOS and Red hat Enterprise
You ca find the distribution version number for CentOS or red hat Enterprise Linux with the following command:
$ cat /etc/redhat-release
CentOS release 5.4 (Final)
version 56
page 169
Unix and Linux System Administration and Shell Programming
uptime
chapter 55
summary
This chapter looks at uptime, a Unix (and Linux) command.
uptime gives information about your server’s total uptime statistics.
info
Type uptime:
$ uptime
20:03 up 1:43, 2 users, load averages: 0.94 0.82 0.78
$
version 56
page 170
Unix and Linux System Administration and Shell Programming
uptime tells you how long your server has been running. uptime also gives the average system load for the last few
minutes. Use uptime for a quick performance check. Collect daily data to watch for problems from underpowered CPUs or
memory management.
top
chapter 56
summary
This chapter looks at top , a Unix (and Linux) command.
top gives the top processes (the default is by CPU usage).
using top
Type top and ENTER or RETURN.
$ top
You will get updated reports (typically, once per second) on the active processes, load average, CPU usage, shared
version 56
page 171
Unix and Linux System Administration and Shell Programming
libraries, and memory usage (regions, physical, and virtual).
Type q to stop the top program.
lsof
chapter 57
This chapter looks at lsof , a Unix (and Linux) command.
lsof lists open files (and that includes devices, directories, pipes, nodes, sockets, and anything else that UNIX treats as a
file).
lsof will give you information on any opened files (including all of the items UNIX treats as a file).
basic use
Type lsof all by itelf to get a list of all open files belonging to all active processes. The list was huge when creating this
example, so I have edited it to highlight some of the things listed.
$ lsof
COMMAND PID USER FD TYPE DEVICE SIZE/OF NODE NAME
loginwind 24 admin cwd DIR 14,8 1564 2 /
loginwind 24 admin txt REG 14,8 946736 1076582
/System/Library/CoreServices/loginwindow.app
version 56
page 172
Unix and Linux System Administration and Shell Programming
/Contents/MacOS/loginwindow
loginwind 24 admin 0r CHR 3,2 0t0 35264644 /dev/null
loginwind 24 admin 1 PIPE 0x224f3f0 16384
loginwind 24 admin 2 PIPE 0x224f3f0 16384
loginwind 24 admin 3u unix 0x27766e8 0t0 ->0x224cdd0
launchd 68 admin 3u KQUEUE count=0, state=0x1
launchd 68 admin 5u systm 0x25f1264 0t0 [1:1:0]
launchd 68 admin 15 PIPE 0x224f6ac 16384
launchd 68 admin 19r DIR 14,8 1122 248 /Library/Preferences
AirPort 84 admin cwd DIR 14,8 1564 2 /
AirPort 84 admin txt REG 14,8 573072 3265433
/System/Library/CoreServices
/AirPort Base Station Agent.app/Contents/MacOS/AirPort Base Station Agent
Spotlight 88 admin cwd DIR 14,8 1564 2 /
Spotlight 88 admin txt REG 14,8 708848 1067264
/System/Library/CoreServices
/Spotlight.app/Contents/MacOS/Spotlight
UserEvent 89 admin cwd DIR 14,8 1564 2 /
Dock 90 admin cwd DIR 14,8 1564 2 /
Dock 90 admin txt REG 14,8 2384752 1046722
/System/Library/CoreServices
/Dock.app/Contents/MacOS/Dock
Dock 90 admin 4u KQUEUE count=0, state=0x2
ATSServer 91 admin cwd DIR 14,8 1564 2 /
ATSServer 91 admin txt REG 14,8 5787888 1131290
/System/Library/Frameworks/ApplicationServices.framework
/Versions/A/Frameworks/ATS.framework/Versions/A/Support/ATSServer
pboard 92 admin cwd DIR 14,8 1564 2 /
SystemUIS 94 admin cwd DIR 14,8 1564 2 /
Finder 96 admin cwd DIR 14,8 1564 2 /
iTunesHel 115 admin cwd DIR 14,8 1564 2 /
Tex-Edit 146 admin cwd DIR 14,8 1564 2 /
Tex-Edit 146 admin txt REG 14,8 367168 1045618
/System/Library/CoreServices/Encodings
/libTraditionalChineseConverter.dylib
firefox-b 149 admin cwd DIR 14,8 1564 2 /
firefox-b 576 admin 56u IPv4 0x4984e64 0t0 TCP 192.168.0.108:60388>173.194.57.119:http (ESTABLISHED)
The default is one file per line. The FD column gives the file descriptor and the TYPE column gives the file type. The
other columns should make sense.
Some of the common FD values are:
cwd = Current Working Directory
mem = memory mapped file
mmap = memory mapped device
rtd = root directory
txt = text file
NUMBER = file descriptor. The character after the number inidicates the mode in which the file is opened. r = read,
w = write, and u = both read and write. This may be followed by lock information.
asdf
Some of the common TYPE values are:
BLK = block special file
CHR = character special file
DIR = directory
FIFO = First In First Out special file
IPv4 = IPv4 socket
IPv6 = IPv6 socket
LINK = symbolic link file
version 56
page 173
Unix and Linux System Administration and Shell Programming
PIPE = pipe
REG = regular file
unix = UNIX domain socket
find which process opened a file
You can get information on which processes opened a specific file by giving the filename as an argument.
$ lsof /System/Library/Fonts/Helvetica.dfont
COMMAND PID USER FD TYPE DEVICE SIZE/OFF NODE NAME
loginwind 24 admin txt REG 14,8 2402112 10720
/System/Library/Fonts/Helvetica.dfont
ATSServer 90 admin txt REG 14,8 2402112 10720
/System/Library/Fonts/Helvetica.dfont
Tex-Edit 123 admin txt REG 14,8 2402112 10720
/System/Library/Fonts/Helvetica.dfont
firefox-b 576 admin txt REG 14,8 2402112 10720
/System/Library/Fonts/Helvetica.dfont
find any open file by name
To find any open file, including an open UNIX domain socket file, with the name /dev/log, type lsof /dev/log . (from the
man pages)
$ lsof /dev/log
list opened files in directory
To list all the processes that have opened files in a particular directory, use the +d option.
$ lsof +d /u/abe/foo/
To list all the processes that have opened files in a particular directory and all of its child directories (subdirectories), use
the +D option. lsof will recurse through all subdirectories.
$ lsof +D /var/
COMMAND PID USER FD TYPE DEVICE SIZE/OFF NODE NAME
loginwind 24 admin txt REG 14,8 149168128 1137272
/private/var/db/dyld/dyld_shared_cache_ppc
loginwind 24 admin 4u REG 14,8 2512 3589022 /private/var/run/utmpx
launchd 64 admin txt REG 14,8 149168128 1137272
/private/var/db/dyld/dyld_shared_cache_ppc
< listing continues >
list open directory
To list the process that has /u/abe/foo open, type lsof /u/abe/foo . (from the man pages)
$ lsof /u/abe/foo
version 56
page 174
Unix and Linux System Administration and Shell Programming
list by process names
To list all open files by process names starting with particlar strings, use the -c option, followed by the process name.
You can give multiple -c switches on a single command line.
Note that this option does not look for an exact match, but any process that includes the character string as a substring of
the process name. So, sh would find ssh and sh
$ lsof -c Terminal
COMMAND PID USER FD TYPE DEVICE SIZE/OFF NODE NAME
Terminal 168 admin cwd DIR 14,8 918 547359 /Users/admin
Terminal 168 admin txt REG 14,8 10244512 1048543
/usr/share/icu/icudt36b.dat
Terminal 168 admin 0r CHR 3,2 0t0 35027076 /dev/null
Terminal 168 admin 1 PIPE 0x22374b4 16384
Terminal 168 admin 2 PIPE 0x22374b4 16384
< listing continues >
list by particular login names
To list all of the files opened by a specific user, type the -u option.
$ lsof -u jill
To list all of the files opened by several specific users, use a comma delimited list.
$ lsof -u jack,jill
To list all of the files opened by every user other than a specific user, use the ^ character. You can use a comma
delimited list
$ lsof -u ^jack,jill
list by particular process
To list all of the files opened by a particular process, type the -p option.
$ lsof -p 1234
list particular login names, user IDs or process numbers
To list all open files for any and all of: login name abe, or user ID 1234, or process 456, or process 123, or process 789,
type lsof -p 456,123,789 -u 1234,abe . (from the man pages)
$ lsof -p 456,123,789 -u 1234,abe
version 56
page 175
Unix and Linux System Administration and Shell Programming
list by mount point
Sometimes when you attempt to unmount a device or directory, the system will warn you with the “Device or resource
Busy” error.
You can list all of the processes using a mount point and then kill those processes so that you can unmount the device or
directory.
$ lsof /home
An equivalent option is:
$ lsof +D /home/
list by device
To list all open files on device /dev/hd4, type lsof /dev/hd4 . (from the man pages)
$ lsof /dev/hd4
kill process
To kill the process that has /u/abe/foo open (by sending the signal SIGHUP), type kill -HUP `lsof -t /u/abe/foo`.
(from the man pages)
$ kill -HUP `lsof /u/abe/foo`
Notice that those are back ticks.
You can also kill all processes that belong to a specific user by using the -t option to output only the process ID and pass
that result on to kill .
$ kill -9 `lsof -t -u jill `
AND/OR
lsof defaults to logical OR of all options. The following example ( from the man pages) will list all of the files open from all
three listed processes and from both users.
$ lsof -p 456,123,789 -u 1234,abe
Use the -a option to perform a logical AND on the user names, processes, etc. Note that you either OR the entire line or
AND the entire line. You can not mix AND and OR together in a single lsof command. The ^ negation on login name or
user ID, process ID, or process group ID options are evaluated prior to other selection criteria and therefore don’t get
included in AND or OR for lsof . Although the -a is legal in any position, placing it between a pair of items does not
version 56
page 176
Unix and Linux System Administration and Shell Programming
cause just those two items to be ANDed, the entire line is still ANDed.
The following AND example will produce a listing of only UNIX socket files that belong to processes owned by the user
foo.
$ lsof -a -U -ufoo
timed listings
You can gather information at specific time intervals. To list the files at descriptors 1 and 3 of every process running the
lsof command for login abe every 10 seconds, type lsof -c lsof -a -d 1 -d 3 -u abe -r10 . (from the man pages)
$ lsof -c lsof -a -d 1 -d 3 -u abe -r10
Use the +r or -r options for timed repeats. The +r switch will stop when no open files that meet the selected criteria are
open. The -r will continue until interrupted by a signal. The number after the r is the time in seconds for each delay.
Between each cycle, lsof will print a sequence of equal signs ( ======= ).
=======
internet connections
Because UNIX and Linux (and Mac OS X) treat internet connections as files, you can use the -i switch to view all of
your open internet connections.
$ lsof -i
COMMAND PID USER FD TYPE DEVICE SIZE/OFF NODE
SystemUIS 93 admin 10u IPv4 0x2152f48 0t0 UDP
firefox-b 127 admin 75u IPv4 0x43c5270 0t0 TCP
>63.141.192.121:http (CLOSE_WAIT)
Fetch 294 admin 23u IPv4 0x27ffe64 0t0 TCP
>reliant.websitewelcome.com:ftp (ESTABLISHED)
Fetch 294 admin 24u IPv4 0x2d2be64 0t0 TCP
>reliant.websitewelcome.com:36975 (LAST_ACK)
Fetch 294 admin 25u IPv4 0x444a66c 0t0 TCP
>reliant.websitewelcome.com:22271 (TIME_WAIT)
NAME
*:*
192.168.0.108:49816192.168.0.108:50539192.168.0.108:50541192.168.0.108:50542-
internet files
To list all open Internet, x.25 (HP-UX), and UNIX domain files, type lsof -i -U. (from the man pages)
$ lsof -i -U
IPv4 network files by PID
To list all open IPv4 network files in use by the process whose PID is 1234, type lsof -i 4 -a -p 1234 . (from the man
pages)
version 56
page 177
Unix and Linux System Administration and Shell Programming
$ lsof -i 4 -a -p 1234
IPv6 files
To list all open IPv6 network files (assuming your UNIX system supports IPv6), type lsof -i 6. (from the man pages)
$ lsof -i 6
list by port
List all of the processes that are listening to a particular port by using colon ( : ) followed by the port number(s).
$ lsof -i:21
To list all files using any protocol on ports 513, 514, or 515 of host wonderland.cc.purdue.edu, type lsof -i
@wonderland.cc.purdue.edu:513-515 . (from the man pages)
$ lsof -i @wonderland.cc.purdue.edu:513-515
list TCP or UDP connections
List all of the TCP connections:
$ lsof -i tcp
List all of the UDP connections:
$ lsof -i udp
list from default domain
Assuming a default domain of cc.purdue.edu, list all files using any protocol on any port of mace.cc.purdue.edu, type
lsof -i @mace . (from the man pages)
$ lsof -i @mace
socket files
To find an IP version 4 socket file by its associated numeric dot-form address, type lsof -i@128.210.15.17 . (from the
man pages)
$ lsof -i@128.210.15.17
version 56
page 178
Unix and Linux System Administration and Shell Programming
To find an IP version 6 socket file by its associated numeric colon-form address, type lsof -i@[0:1:2:3:4:5:6:7].
(from the man pages)
$ lsof -i@[0:1:2:3:4;5;6:7]
To find an IP version 6 socket file by its associated numeric colon-form address that has a run of zeros in it (such as the
loop-back address), type lsof -i@[::1]. (from the man pages)
$ lsof -i@[::1]
Network File System (NFS)
List all of the Network File System (NFS) files by using the -N switch.
$ lsof -N
To find processes with open files on the NFS file system named /nfs/mount/point whose server is inaccessible (assuming
your mount table supplies the device number for /nfs/mount/point), type lsof -b /nfs/mount/point. (from the man pages)
$ lsof -b /nfs/mount/point
To do the preceding search with warning messages suppressed, type lsof -bw /nfs/mount/point. (from the man pages)
$ lsof -bw /nfs/mount/point
ignore device cache file
To ignore the device cache file, type lsof -Di. (from the man pages)
$ lsof -Di
obtain specific multiple info on each file
You can combine flags to gather specific information. To obtain the PID and command name field for each process, file
descriptor, file device number, and file inode number for each file of each process, type lsof -FpcfDi. (from the man pages)
$ lsof -FpcfDi
using regular expressions
To list the current working directory of prcoesses running a command that is exactly four characters long and has an
upper or lower case “O” or “o” in character position three, type lsof -c /^..o.$/i -a -d cwd . (from the man pages)
version 56
page 179
Unix and Linux System Administration and Shell Programming
$ lsof -c /^..o.$/i -a -d cwd
free
chapter 58
summary
This chapter looks at free , a Linux command.
free is used to show free memory.
free displays the total amount of free and used physical memory and swap space in the system, as well as the buffers
and cache consumed by the kernel.
size
The default is to report the memory size in bytes. Use the -g option to display in Gigabytes. Use the -m option to display
in Megabytes. Use the -k option to display in Kilobytes. Use the -b option to display in Bytes.
total
Use the -t option to view total memory.
version 56
page 180
Unix and Linux System Administration and Shell Programming
low/high
Use the -l option to view low vs. high memory usage.
count
The default is to display the information once and then return to the normal shell. Use the -s n option (where n is the
number of times to display) to keep displaying the memory usage (until interrupted). Use the -c n or --count=n options
(where n is the number of times to display the memory statistics).
Mac OS X
Use the Activity Monitor for this purpose. From the command line, you can get similar results from vm_stat. vm_stat is
a Mach-specific version of the UNIX command vmstat Also:
$ echo -e "\n$(top -l 1 | awk '/PhysMem/';)\n"
vmstat
chapter 59
summary
This chapter looks at vmstat, a Unix (and Linux) command.
vmstat is used to show virtual memory usage.
usage
Type vmstat to see a list of virtual memory statistics. Especially useful in a virtualized enivronment.
$ vmstat
Mac OS X
vm_stat is a Mach-specific version of the UNIX command vmstat:
$
version 56
page 181
Unix and Linux System Administration and Shell Programming
vm_stat
Mach Virtual Memory Statistics: (page size of 4096 bytes)
Pages free: 5538.
Pages active: 142583.
Pages inactive: 76309.
Pages wired down: 37531.
"Translation faults": 4427933.
Pages copy-on-write: 54807.
Pages zero filled: 3086159.
Pages reactivated: 6533.
Pageins: 46357.
Pageouts: 4453.
Object cache: 66465 hits of 76770 lookups (86% hit rate)
polkit
chapter 60
summary
This chapter looks at polkit, a Unix (and Linux) command.
polkit is used to control system-wide privileges.
running privileged commands
polkit can be used to run privileged commands by typing pkexec, followed by the intended command:
$ pkexec cat README
sudo alternative
It is generally recommended to use sudo rather than pkexec, because sudo provides greater flexibity and security.
Mac OS X
version 56
page 182
Unix and Linux System Administration and Shell Programming
While polkit is not part of the official Mac OS X release from Apple, it can be installed on mac OS X.
security
pkexec never does any validation of the ARGUMENTS passed to a program.
graphic programs
X11 (and other graphic) programs won’t run by default with polkit because the $DISPLAY environment variable is not
set. The following information from the pkexec manual describes how to set pkexec so that it can run a graphic program.
To specify what kind of authorization is needed to execute the program
/usr/bin/pk-example-frobnicate as another user, simply write an action
definition file like this
<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE policyconfig PUBLIC
"-//freedesktop//DTD PolicyKit Policy Configuration 1.0//EN"
"http://www.freedesktop.org/standards/PolicyKit/1/policyconfig.dtd">
<policyconfig>
<vendor>Examples for the PolicyKit Project</vendor>
<vendor_url>http://hal.freedesktop.org/docs/PolicyKit/</vendor_url>
<action id="org.freedesktop.policykit.example.pkexec.run-frobnicate">
<description>Run the PolicyKit example program Frobnicate</description>
<description xml:lang="da">Kør PolicyKit eksemplet Frobnicate</description>
<message>Authentication is required to run the PolicyKit example
program Frobnicate</message>
<message xml:lang="da">Autorisering er påkrævet for at afvikle
PolicyKit eksemplet Frobnicate</message>
<icon_name>audio-x-generic</icon_name>
<defaults>
<allow_any>no</allow_any>
<allow_inactive>no</allow_inactive>
<allow_active>auth_self_keep</allow_active>
</defaults>
<annotate key="org.freedesktop.policykit.exec.path">
/usr/bin/pk-example-frobnicate</annotate>
</action>
</policyconfig>
and drop it in the /usr/share/polkit-1/actions directory under a
suitable name (e.g. matching the namespace of the action). Note that in
addition to specifying the program, the authentication message,
description, icon and defaults can be specified. For example, for the
action defined above, the following authentication dialog will be
shown:
[IMAGE][2]
+----------------------------------------------------------+
|
Authenticate
[X] |
+----------------------------------------------------------+
|
|
| [Icon] Authentication is required to run the PolicyKit |
|
example program Frobnicate
|
|
|
|
An application is attempting to perform an
|
|
action that requires privileges. Authentication |
|
is required to perform this action.
|
|
|
|
Password: [__________________________________] |
|
|
| [V] Details:
|
| Command: /usr/bin/pk-example-frobnicate
|
| Run As: Super User (root)
|
| Action: org.fd.pk.example.pkexec.run-frobnicate
|
| Vendor: Examples for the PolicyKit Project
|
|
|
|
[Cancel] [Authenticate] |
+----------------------------------------------------------+
If the user is using the da_DK locale, the dialog looks like this:
version 56
page 183
Unix and Linux System Administration and Shell Programming
[IMAGE][3]
+----------------------------------------------------------+
|
Autorisering
[X] |
+----------------------------------------------------------+
|
|
| [Icon] Autorisering er pŒkr¾vet for at afvikle
|
|
PolicyKit eksemplet Frobnicate
|
|
|
|
Et program fors¿ger at udf¿re en handling der
|
|
kr¾ver privilegier. Autorisering er pŒkr¾vet.
|
|
|
|
Kodeord: [___________________________________] |
|
|
| [V] Detaljer:
|
| Bruger:
Super User (root)
|
| Program: /usr/bin/pk-example-frobnicate
|
| Handling: org.fd.pk.example.pkexec.run-frobnicate
|
| Vendor:
Examples for the PolicyKit Project
|
|
|
|
[AnnullŽr] [Autorisering] |
+----------------------------------------------------------+
Note that pkexec does no validation of the ARGUMENTS passed to PROGRAM.
In the normal case (where administrator authentication is required
every time pkexec is used), this is not a problem since if the user is
an administrator he might as well just run pkexec bash to get root.
However, if an action is used for which the user can retain
authorization (or if the user is implicitly authorized), such as with
pk-example-frobnicate above, this could be a security hole. Therefore,
as a rule of thumb, programs for which the default required
authorization is changed, should never implicitly trust user input
(e.g. like any other well-written suid program).
defaults
chapter 61
summary
This chapter looks at defaults , a Mac OS X command.
defaults is used to read, write, or delete defaults from a command line shell.
defaults options
Mac OS X programs use the defaults system to record user preferences and other information. Most of the information is
more easily accessed through the application’s Preferencess panel, but some information is normally inaccessible to the end
user.
Note: Since applications do access the defaults system while they’re running, you shouldn’t modify the defaults of a
running application.
All programs share the defaults in NSGlobalDomain. If the program doesn’t have its own default, it uses the value from
the NSGlobalDomain.
version 56
page 184
Unix and Linux System Administration and Shell Programming
screencapture defaults
The following methods use Terminal to change the default file format and location where the screenshot is saved from
the graphic user interface.
In Mac S X 10.4 (Tiger) or more recent, the default screencapture format can be changed in Terminal by using the
defaults command. In Mac S X 10.4 (Tiger), the new default does not take effect until you logout and log back in (from
the entire computer, not just from Terminal — a full restart will also work) unless you also use the killall command.
$ defaults write com.apple.screencapture type ImageFormat; killall SystemUIServer
The ImageFormat can be png (Portable Network Graphic), pdf (Portable Document Format), tiff (Tagged Image File
Format), jpg or jpeg (Joint Photographic Experts Group), pict (Macintosh QuickDraw Picture), bmp (Microsoft Windows
Bitmap), gif (Graphics Interchange Format), psd (Adobe Photoshop Document), sgi (Silicon Graphics File Format), or tga
(Truevision Targe File Format).
JPGs are saved at quality 60%.
To change the default location where the screenshot file is saved (the default is Desktop), use the following Terminal
command (where PathName is the full path to a directory.
$ defaults write com.apple.screencapture location PathName; killall SystemUIServer
The normal default location would be reset with the following command (where UserName is the current account’s user
name.
$ defaults write com.apple.screencapture location /Users/UserName/Desktop; killall
SystemUIServer
dock defaults
Use the following commands to change the 3D Dock of Mac OS X Leopard and later back to the 2D look. The killall
restarts the Dock so that the change takes effect right away.
$ defaults write com.apple.dock no-glass -boolean YES; killall Dock
Return to the 3D look with the following commands:
$ defaults write com.apple.dock no-glass -boolean NO; killall Dock
Use the following commands to add a gradient behind an icon in a Dock Stack in Mac OS X:
$ defaults write com.apple.dock mouse-over-hilte-stack -boolean YES; killall Dock
version 56
page 185
Unix and Linux System Administration and Shell Programming
Use the following commands to add a small Exposé button to the upper right of your Mac OS X screen. Clicking on the
Exposé button will show all of the windows from the current application (same as the normal F10 default) and option
clicking the Exposé button will show all windows from all programs (same as the normal F9 default). This will work even
if you reconfigure the F9 and F10 keys to do something else.
$ defaults write com.apple.dock wvous-floater -bool YES; killall Dock
Use the following commands to remove the small Exposé button.
$ defaults write com.apple.dock wvous-floater -bool NO; killall Dock
login defaults
Use the following command to add a message to your Login window. Replace message with a short message. Keep the
message short. As always with the sudo command, type very carefully.
$ sudo defaults write /Library/Preferences/com.apple.loginwindow LoginwindowText
"message"
To remove the login message, type the following:
$ sudo defaults write /Library/Preferences/com.apple.loginwindow LoginwindowText ""
showing dot files
The Mac OS X Finder hides all files that start with the period or dot character. This matches the standard behavior of ls
in UNIX, where a leading dot is used as an indicator of a hidden file.
To have Finder always show all files, including hidden files, type:
$ defaults write com.apple.Finder AppleShowAllFiles YES; killall Finder
To restore Finder to its default behavior, type:
version 56
page 186
Unix and Linux System Administration and Shell Programming
$ defaults write com.apple.Finder AppleShowAllFiles NO; killall Finder
.DS_Store files
To prevent Mac OS X from creating .DS_Store files on network shares (and everywhere else), type:
$ defaults write com.apple.desktopservices DSDontWriteNetworkStores true
To restore Mac OS X from creating .DS_Store files on network shares (and everywhere else), type:
$ defaults write com.apple.desktopservices DSDontWriteNetworkStores false
Mac Flashback Trojan
You can use defaults to determine if your Macintosh has been infected by the Mac Flashback Trojan (which enters
your computer through a Java flaw).
Type the following command (copy and paste into Terminal):
$ defaults read /Applications/Safari.app/Contents/Info LSEnvironment
A clean system will report “The domain/default pair of (/Applications/Safari.app/Contents/Info, LSEnvironment) does not
exist”. Any other result indicates your computer is infected.
If your computer passes the first test, type the following command (copy and paste into Terminal):
$ defaults read ~/.MacOSX/environment DYLD_INSERT_LIBRARIES
A clean system will report “The domain/default pair of (/Users/user-name/.MacOSX/environment,
DYLD_INSERT_LIBRARIES) does not exist”. Any other result indicates your computer is infected.
If your computer is infected, immediately go to F-Secure at http://www.f-secure.com/v-descs/trojandownloader_osx_flashback_i.shtml.
Downloading the latest security patches from Apple at http://support.apple.com/kb/HT5228?
viewlocale=en_US&locale=en_US will help prevent infection.
version 56
page 187
Unix and Linux System Administration and Shell Programming
init
chapter 62
summary
This chapter looks at init , a Linux command.
init is used to change the runlevel.
run levels
Linux has run levels (sometimes written as a single word, runlevel). The non-graphic mode is run level 3. the graphic
mode is run level 5. Sometimes a Linux server will only boot up to the non-graphic level. You can use this command to
temporarily get the graphic interface running (run level 5).
$ init 5
Edit the /etc/inittab file to include id:5:initdefault: to make a permanent change.
version 56
page 188
Unix and Linux System Administration and Shell Programming
A chart of the Linux run levels:
RUN LEVEL DESCRIPTION
0
System is halted
Single-user Mode
1
used for system maintenance
2
Multiuser mode withut networking
rarely used
3
Standard multiuser mode with networking active
4
not defined
Multiuser mode with graphics
5
Like standard multiuser mode with networking active plus the X Window System graphic interface
6
Reboots the system
shuts down all system services
sendmail
chapter 63
summary
This chapter looks at sendmail , a Unix/Linux utility.
sendmail is used to send email.
In 1975 the Sendmail email system was added to the BSD version of UNIX. Because email was built into the operating
system, email was the method UNIX used for sending notifications to the system administrator of system status, reports,
problems, etc.
set-up
Each version of UNIX has its own variation of sendmail , but the default installation almost always works correctly
without modification. A system administrator rarely needs to know how to set-up or modify sendmail .
version 56
page 189
Unix and Linux System Administration and Shell Programming
ifconfig
chapter 64
summary
This chapter looks at ifconfig , a Unix (and Linux) command.
ifconfig is used to configure network cards and interfaces.
Warning: ifconfig will not save changes on AIX after a reboot. On Linux, any changes made with ifconfig are saved
on a reboot.
view configuration
Type ifconfig by itself to view how your network is configured.
$ ifconfig
version 56
page 190
Unix and Linux System Administration and Shell Programming
static IP address
Use ifconfig interface IP-address to set the static IP address for a particular interface
$ ifconfig en0 100.1.1.1
arp
chapter 65
summary
This chapter looks at arp , a Unix (and Linux) command.
arp is used to display and modify the Internet-to-Ethernet address translation tables.
display all
Use the -a option to display all of the current ARP entries. Useful in conjunction with ifconfig and route .
$
?
?
?
$
arp -a
(192.168.0.1) at 0:11:95:75:8:86 on en1 [ethernet]
(192.168.0.106) at 0:d:60:d2:55:3f on en1 [ethernet]
(192.168.0.255) at ff:ff:ff:ff:ff:ff on en1 [ethernet]
version 56
page 191
Unix and Linux System Administration and Shell Programming
netstat
chapter 66
summary
This chapter looks at netstat, a Unix (and Linux) command.
netstat is used to symbolically displays the contents of various network-related data structures.
view connections
Type netstat to see the list of all sockets and server connections.
$ netstat
main info
The main information for webserver administration is in the first few lines, so you can pipe to head :
version 56
page 192
Unix and Linux System Administration and Shell Programming
$ netstat | head
routing addresses
Use the -r option to display the network routing addresses.
$ netstat -r
route
chapter 67
summary
This chapter looks at route , a Unix (and Linux) command.
route is used to manipulate the network routing tables.
view connections
Type route followed by the -v option to list the routing tables. This is close to the same information reported by
netstat -r.
$ route -v
routing commands
The route utility can used for add , flush , delete, change, get , or monitor.
The add command will add a route.
version 56
page 193
Unix and Linux System Administration and Shell Programming
The flush command will remove all routes.
The delete command will delete a specific route.
The change command will change aspects of a route (such as its gateway).
The get command will lookup and display the route for any destination.
The monitor command will continuously report any changes to the routing infrmation base, routing lookup misses, or
suspected network partitionings.
ping
chapter 68
summary
This chapter looks at ping , a Unix (and Linux) command.
ping is used to test a server connection.
test packets
ping sends test packets to a specific server to see if the server responds properly.
Type ping followed by -c count followed by an IP address or domain name.
ping will continue forever unless halted or you give a specific count limit.
$ ping -c 5 www.osdata.com
PING osdata.com (67.18.4.2): 56 data bytes
64 bytes from 192.168.0.1: icmp_seq=0 ttl=127
64 bytes from 192.168.0.1: icmp_seq=1 ttl=127
64 bytes from 192.168.0.1: icmp_seq=2 ttl=127
64 bytes from 192.168.0.1: icmp_seq=3 ttl=127
64 bytes from 192.168.0.1: icmp_seq=4 ttl=127
version 56
time=18.801 ms
time=9.918 ms
time=17.216 ms
time=6.748 ms
time=30.640 ms
page 194
Unix and Linux System Administration and Shell Programming
--- 192.168.0.1 ping statistics -- 5 packets transmitted, 5 packets received, 0% packet loss
round-trip min/avg/max/stddev = 6.748/16.665/30.640/8.295 ms
$
ping will quickly either return a report or will say destination host unreachable on failure.
measuring
You can measure the dropped packet rate and the minimum, mean, maximum, and standard deviation of round-trip
times.
When using ping for testing, start by pinging local host to verify that the local network interface is up and running. Then
ping hosts and gateways further away.
nslookup
chapter 69
summary
This chapter looks at nslookup , a Unix (and Linux) command.
nslookup is used to obtain domain name and IP information about a server.
information
Type nslookup followed by a doman name.
$ nslookup www.osdata.com
Server: 192.168.0.1
Address: 192.168.0.1#53
Non-authoritative answer:
www.osdata.com canonical name = osdata.com.
Name: osdata.com
Address: 67.18.4.2
$
version 56
page 195
Unix and Linux System Administration and Shell Programming
security
You can use nslookup to find the DNS information for a particular host IP in your server access logs.
traceroute
chapter 70
summary
This chapter looks at traceroute , a Unix (and Linux) command.
traceroute is used to obtain the entire path to a server or host.
entire route
Type traceroute followed by a doman name.
$ traceroute www.osdata.com
You can use traceroute to diagnose network problems. When some people can access your website and other can’t,
then traceroute can help you find the break or error along the network path.
etiquette
traceroute puts a load on every server in the path to the selected host. Therefore, only use traceroute when you need
the diagnosis.
version 56
page 196
Unix and Linux System Administration and Shell Programming
ftp and sftp
chapter 71
summary
This chapter looks at ftp and sftp , a Unix (and Linux) command.
ftp is FIle Transfer Protocol.
sftp is Secure FIle Transfer Protocol.
In general, you should use sftp rather than sftp . There is a lot of overlap between the two tools and their commands.
This chapter will primarily discuss examples as ftp . In practice you should substitute sftp whenever possible.
There are many graphic FTP tools available. You should consider using these tools for most file transfers. I personally
use Fetch, but that’s from long time habit. FileZilla is popular, free, and open source, with binaries available for Windows,
Linux, and Mac OS X, and source code available for most other platforms.
For serious scripting, you should probably consider the more powerful cURL (see next chapter).
You will find many existing scripts that use ftp . When you are using the command line, it is often easier and faster to do
quick work with ftp or sftp . If you are at an unfamiliar machine and simply fall back on the command line FTP rather
than try to guess what graphic FTP program is available.
Adobe recommends using command line FTP “for testing when troubleshooting problems with the Dreamweaver or
UltraDev FTP client.”. See web page http://kb2.adobe.com/cps/164/tn_16418.html . This is a recommendation to the mostly
artist, non-geek audience for Adobe’s creative products. Adobe provides step by step instructions for both Windows and
version 56
page 197
Unix and Linux System Administration and Shell Programming
Mac OS X.
sample session
The following sample session shows how a simple ftp session might go. This will give you context for the many
specific details that follow.
$ ftp example.com
Connected to example.com
220 example FTP Server (Version x.x Mon Jan 20 05:32:40 EDT 2014) ready.
User (example.com:(none)): User-name
331 Password required for user-name
Password: password
230 User user-name logged in.
ftp> cd /directory
250 CWD command successful.
ftp> pwd
257 "/directory" is the current directory.
ftp> ls
random.txt
picture.png
226 transfer completed
ftp> ascii
200 Type set to A.
ftp> put filename.txt
200 PORT command successful.
Opening ASCII mode data connection for filename.txt
226 Transfer complete
ftp> quit
221 Goodbye.
$
formatting for this lesson
In the following examples, things you type will be in bold. Anything contained inside square braces [ ] is optional and
anything outside of square braces is required. Braces inside braces indicate options that can be applied to options. The
stroke or vertical bar character | indicates a choice — enter only one of the choices. Elipses … indicate that you can have
one, two, or mroe of the indicated item.
standard FTP
To start an FTP session, type ftp hostname . The host name can be a named URL or an IP address. REMINDER:
whenever possible, substitute sftp for ftp throughout this chapter.
You will be asked for your userid and password.
To end the FTP session type either quit or bye .
Enter commands after the ftp> prompt. to conduct your FTP session. Type help to get a current list of commands
available and type help followed by a command to get help on a specific command.
All FTP commands are case sensitive and in lower case.
anonymous FTP
Some servers allow anonymous ftp access. Use anonymous or ftp as your userid. If you are asked for a logon password,
version 56
page 198
Unix and Linux System Administration and Shell Programming
your own email address or guest will often be acceptable.
help
? will give a list of available commands. If followed by the name of a specific command, it will give a brief description
of that command. Works the same way as help .
help will give a list of available commands. If followed by the name of a specific command, it will give a brief
description of that command. Works the same way as ?.
rhelp will get help from the remote server (which may have a different command set than the local machine). rhelp is
more commonly available than remotehelp .
ftp> rhelp
remotehelp will get help from the remote server (which may have a different command set than the local machine).
ftp> remotehelp [command-name]
connections
ftp will connect to a remote FTP server. This command is given at the normal command shell to start FTP. Provide a
host name (URL or IP address) and an optional port number.
$ ftp host-name [port]
sftp will connect to a remote FTP server using secure SSH. This command is given at the normal command shell to
start SFTP. Provide a host name (URL or IP address) and an optional port number.
$ sftp host-name [port]
open will connect to a specified remote FTP host (by name or IP address). An optional port number may be supplied. If
auto-login option is on (the default), then FTP will attempt to automatically log the user in t the FTP server.
ftp> open host-name [port]
account will supply a supplemental password. Sometimes there is a need for an additional password to access specific
resources (such as a password-protected directory). If no argument is included, the user willbe prompted for an account
password in a non-echoing input mode.
ftp> account [password]
close will terminate the FTP session. You can follow a close command with a open command to connect to a new
host. This is the same as the disconnect command. Any defined macors are erased.
ftp> close
version 56
page 199
Unix and Linux System Administration and Shell Programming
disconnect will terminate the FTP session. You can follow a disconnect command with a open command to connect
to a new host. This is the same as the close command.
bye will terminate the FTP session and exit back to the shell. The same as quit .
ftp> bye
quit will terminate the FTP session and exit back to the shell. The same as bye .
ftp> quit
file transfers
ascii will set the FTP session for ASCII file transfers (the alternative is binary). This is the default method.
binary will set the FTP session for binary file transfers (the alternative is ascii ).
type will set the file transfer type.
form will set the file transfer format. The default format is file.
ftp> form format
mode will set the file transfer mode. The default format is stream.
ftp> mode mode-name
struct will set file transfer structure to struct-name . The default is stream structure.
ftp> struct struct-mode
xferbuf will set the size of the socket send/receive buffer.
ftp> xferbuf size
get will receive a file from the remote server. Name the file after the get command. You may optionally follow with a
second name which will assign the second file name to the local downloaded file. The same as recv .
recv will receive a remote file. Name the file after the recv command. You may optionally follow with a second name
which will assign the second file name to the local downloaded file. The same as get .
mget will get multiple remote files. If glob is on, then you can use wildcards in specifying files. Toggle prompt off if you
don’t want to be prompted for each file downloaded.
fget will get files using a localfile as the source of the file names. That is, put the list of file names into a local file and
version 56
page 200
Unix and Linux System Administration and Shell Programming
then pass that file name to this command.
ftp> fget localfile
glob will toggle metacharacter expansion of local file names.
put will send one file to the remote server. Follow the put command with the file to upload. Optionally include a second
name to assign a different name to the file uploaded to the remote server. It is an error if the file does not exist on your
local machine. The same as send .
send will send one file to the remote server. Follow the send command with the file to upload. Optionally include a
second name to assign a different name to the file uploaded to the remote server. It is an error if the file does not exist on
your local machine. The same as put .
append will append to a file.
ftp> append local-file [remote-file]
mput will send multiple files to the remote server. If glob is on, then you can use wildcards in specifying files. Toggle
prompt off if you don’t want to be prompted for each file uploaded.
delete will delete a remote file.
mdelete will delete multiple remote files.
rename will rename a file on the remote server.
directory control
pwd will print the current working directory on the remote server.
ftp> pwd
lpwd will print the current local working directory.
ftp> lpwd
cd changes the working directory on the remote server. Use lcd to change the local working directory.
ls will list the contents of a remote directory. Works the same as dir . The listing will include any system-dependent
information the server chooses to include. Most UNIX systems will produce the output of the command ls -l. You may
name a specific remote directory. Otherwise, the current remote working directory is listed. You may optionally also name
a local file and the listing will be stored in the designated local file. If no local file is specified or the local file is “ -”, the
output will be sent to the terminal.
ftp> ls [remote-path [local-file]]
dir will list the contents of the remote directory. Works the same as ls.
version 56
page 201
Unix and Linux System Administration and Shell Programming
dir
dir
dir
dir
dir
-C lists the files in wide format.
--1 lists the files in bare format in alphabetic order.
--S lists the files in bare format in alphabetic order.
--r lists the files in reverse alphabetic order.
-R lists all files in the current directory and sub directories.
lcd will display the local working directory if typed by itself. If a path name follows the lcd command, it will change the
local working directory. Use cd to change the remote working directory.
mlst will list the contents of the remote directory in a machine parsable form. It defaults to listing the current remote
working directory.
ftp> mlst [remote-path]
mdir will list the contents of multiple remote directories. Works the same as mls .
mls will list the contents of multiple remote directories. Works the same as mdir .
mkdir will make a directory on the remote server.
rmdir will delete or remove a directory on the remote server.
ftp> rmdir directory-name
modtime will show the last modification time and date for a remote file.
ftp> modtime remote-file
chmod will change file permissions on a remote file.
ftp> chmod mode remote-file
toggles
cr will toggle the stripping of carriage returns during ASCII type file retrieval. records are denoted by the sequence of
carriage return/line feed sequence during ASCII type file transfer. When cr is on, carriage returns are stripped to conform
to the UNIX single linefeed record delimiter. This is the default. records on non-UNIX remote systems may contain single
linefeeds and when an ASCII type file transfer is made, these linefeeds will only be distinguished from a record delimiter if
cr is off. The default is for cr to be on.
ftp> cr [ on | off ]
debug will toggle the debugging mode. If the optional debug value is provided, it is used to set the debugging level.
When debugging is on, FTP prints each command sent to the remote server, preceded by the string “ -->”.
ftp> debug [ on | off | debuglevel ]
version 56
page 202
Unix and Linux System Administration and Shell Programming
gate will toggle the gate-ftp. Specify the optional host[:port] to change the proxy.
ftp> gate [ on | off | gateserver [port] ]
glob will toggle metacharacter expansion of local file names.
hash will toggle printing # for each buffer of 1,024 (1 kilobyte) of data is transferred.
passive will toggle the use of passive transfer mode.
ftp> passive [ on | off | auto ]
sunique will toggle storing of files under unique file names on the remote server. Remote FTP server must support the
FTP protocol STOU command for successful completion. The remote server will report unique names. The defauly is off.
ftp> sunique [ on | off ]
trace will toggle packet tracing.
verbose will toggle the verbose mode. In verbose mode, all responses from the FTP server are displayed to the local
user. When a file transfer completes, verbose mode also provides statistics about the effiency of the file transfer. By default,
verbose is on.
ftp> verbose [ on | off ]
pagers
Some versions of command line FTP will allow you to read the contents of remote files in a local pager (such as less or
more).
page will view a remote file through your local pager.
ftp> page remote-file
less will view a remote file through the pager named less .
ftp> less remote-file
more will view a remote file through the pager named more .
ftp> more remote-file
pdir will list the contents of a remote path through your local pager. You can designate a specific remote path or default
to the remote server current working directory.
version 56
page 203
Unix and Linux System Administration and Shell Programming
ftp> pdir [remote-path]
lpage will view a local file through your local pager.
ftp> lpage local-file
other commands
prompt will force interactive prompting on multiple commands.
system will show the operating system running on the remote server.
ftp> system
status will show the current status.
rstatus will show the current status of the remote server. If a file name is specified, it will show the status of the file on
the remote server.
ftp> rstatus [remote-file]
remotestatus will show the current status of the remote server. If a file name is specified, it will show the status of the
file on the remote server.
ftp> remotestatus [remote-file]
user will send new user information.
! Escape the the shell. Type exit to return to the FTP session. You can follow ! with a specific command. If you type a
command, then all the following text will be considered as arguments for that command.
ftp> ! [command [args]]
ftp> !
$ unix_command
command results
$ exit
ftp>
bell will beep when a command is completed.
literal will send an arbitrary FTP command with an expected one line reponse. The same as quote .
macro will execute the designated macro macro-name . The macro macro-name must have been defined with the macdef
command. Arguments passed to the macro are unglobbed.
version 56
page 204
Unix and Linux System Administration and Shell Programming
ftp> $ macro_name [args]
quote will send an arbitrary command to the remote server. The same as literal. All arguments are sent veratim to the
remote FTP server.
ftp> quote line-to-send [ arg1 arg2 … ]
FileZilla
FileZilla Client is a free, open source FTP client that supports FTP, SFTP, and FTPS (FTP over SSL/TLS). It is available
for many operating systems and there are binaries available for Windows, Linux, and Mac OS X. The source code is hosted
at http://sourceforge.net/projects/filezilla. The binaries can be found at https://filezilla-project.org/download.php.
While FileZilla is intended as a graphic interface for FTP, the client will accept command line.
Use the command filezilla [<FTP URL>] to connect to a server. The form of the URL should be
[protocol://][user[:pass]@]host[:port][/path] . Valid protocols include ftp://, ftps://, ftpes://, or sftp://. The default
is ordinary ftp.
Use the option -c or --site=<string> to connect to the specified Site Manager site.
Use the option -a or --local=<string> to set the local directory from which you will be uploading files.
Putting this together, the following example will connect to the site example.com by ftp protocol using the user name and
password and sets the local folder (directory) to uploads in the current user’s home directory:
$ filezilla ftp://username:password@ftp.example.com --local="~/uploads"
Use the -l or ---logontype=(askinteractive) flag to have FileZilla interactively ask for login information. This
option is useful for scripting.
version 56
page 205
Unix and Linux System Administration and Shell Programming
cURL
chapter 72
summary
This chapter looks at cURL , a Unix (and Linux) command line tool.
curl is used to interact with the internet. Linux also includes the similar wget command tool. You can download wget
for free using fink .
cURL is a command line tool and accompanying library for transferring data over the internet. cURL supports a wide
variety of internet protocols, including DICT, FILE, FTP, FTPS, Gopher, HTTP, HTTPS, IMAP, IMAPS, LDAP, LDAPS,
POP3, POP3S, RTMP, RTSP, SCP, SFTP, SMTP, SMTPS, Telnet, and TFTP.
download a file
You can use curl to download the content of a URL. The following command displays the HTML for Google as the
STDOUT (on your Terminal screen):
$ curl http://www.google.com
redirect a download
You can redirect the results to file using the standard Unix/Linux redirect. You will be given some status information if
you redirect.
version 56
page 206
Unix and Linux System Administration and Shell Programming
$ curl http://www.google.com > google.html
% Total % Received % Xferd Average Speed Time Time Time Current
Dload Upload Total Spent Left Speed
100 10559 0 10559 0 0 26690 0 --:--:-- --:--:-- --:--:-- 323k
You can use the -o (lower case) to save the downloaded file with the name of your choice. You will see progress meter
information as the download progresses. In the example, the download was so fast that only one status report was created.
$ curl -o google.303 www.google.com
% Total % Received % Xferd Average Speed Time Time Time Current
Dload Upload Total Spent Left Speed
100 219 100 219 0 0 800 0 --:--:-- --:--:-- --:--:-- 0
sysstat
chapter 73
summary
This chapter looks at sysstat, a server package.
package
sysstat is a server toolset originally created by Sebastien Godard.
Download sysstat from http://pagesperso-orange.fr/sebastien.godard/download.html.
The most famous tools ar iostat, pidstat, and sar.
iostat reports CPU statistics and input/output statistics for devices, partitions and network filesystems.
mpstat reports individual or combined processor related statistics.
pidstar reports statistics for Linux tasks (processes) : I/O, CPU, memory, etc.
sar collects, reports and saves system activity information (CPU, memory, disks, interrupts, network interfaces, TTY,
kernel tables,etc.)
sadc is the system activity data collector, used as a backend for sar.
sa1 collects and stores binary data in the system activity daily data file. It is a front end to sadc designed to be run from
cron.
sa2 writes a summarized daily activity report. It is a front end to sar designed to be run from cron.
version 56
page 207
Unix and Linux System Administration and Shell Programming
sadf displays data collected by sar in multiple formats (CSV, XML, etc.) This is useful to load performance data into a
database, or import them in a spreadsheet to make graphs.
nfsiostat reports input/output statistics for network filesystems (NFS).
cifsiostat reports CIFS statistics.
at
chapter 74
summary
This chapter looks at at, a Unix (and Linux) command.
at is used to schedule a particular job at a particular time.
example
Type at midnight , followed by ENTER or RETURN.
$ at midnight
You may see the at> prompt (on Mac OS X, there is no prompt).
Type a single command, followed by ENTER or RETURN.
$ who > who.out
Type one command per line.
When finished, hold down the CONTROL key and then the D key (written ^d) to exit at.
job 1 at Fri Jul 13 00:00:00 2012
$
version 56
page 208
Unix and Linux System Administration and Shell Programming
You will see a job number and the time it will run. This job will run all of the commands you entered.
removing an at job
Type atrm (for at remove), followed by the job number to remove an existing at job.
$ atrm 1
timing
You can name a specific time using the YYMMDDhhmm.SS format.
You can also specify noon, midnight, or teatime (4 p.m.).
If the time has already passed, the next day is assumed.
version 56
page 209
Unix and Linux System Administration and Shell Programming
back ups
chapter 75
summary
This chapter looks at backing up systems and data.
While everyone should perform back ups, when a system crashes, everyone will turn to the system administrator to be
rescued. And this is a get yourself fired kind of issue because the system administrator will always get the blame for lack of
back ups.
tools
Your basic tools for back ups are tar , various compression schemes, and cron scripts.
tar stands for Tape ARchive, from back in the days when back ups were put on those large tape machines you see in the
old James Bond movies.
cron is a system for scheduling jobs, including UNIX shell scripts, to run at specific times.
rule of thumb
People are always asking how often they should do their backups, looking for a regualr time period.
The rule of thumb on back-ups is to back up your data before you reach the point where you have too much work to do
to replace the data.
If you start typing an email and the system crashes, you are annoyed, but you can type the email over again. On the other
hand, if you lose all of your emails, you have a problem.
Hence, the simple rule of thumb to back up as soon as you are about to have too much data to create over again.
Notice that this answer wasn’t in the form of a specific time periood (such as daily or weekly).
version 56
page 210
Unix and Linux System Administration and Shell Programming
The reality of any large scale system is that you will be using the cruder method of specific times, such as daily, weekly,
and monthly back ups.
incremental and full
A full back up stores everything on a system.
An incremental backup stores only those things that have changed.
Back ups (especially full backups) take a lot of storage space.
The typical policy is to do incremental backups fairly often (daily, hourly, whatever is appropriate for the demands of
your system) and to do full backups less often (weekly or even monthly).
When you have a failure, you go back to the last full back up and then run every incremental back up. The more
incremental back ups you have to run, the longer it will be before the syetem is restored, with angry users and angry bosses
waiting impatiently.
three location storage
There is the rule of a minimum of three copies of your back ups. You have two copies of your back on site and one
stored at a secure location far away.
The secure storage copy protects you from fires or eathquakes or floods or other disasters that physically destroy your
ability to restore your system. Therefore, the secure back ups need to be kept far enough away that they won’t get
destroyed in the same disaster.
Of course, this also means that there is travel time for using the secure back up.
The local back ups are usually alternated (or rotated, if you have more than just two).
The first full back up is stored on media A. The next full back up is stored on media B. The next full back up is stored on
media A (wiping out the original back up).
version 56
page 211
Unix and Linux System Administration and Shell Programming
tar
chapter 76
summary
This chapter looks at tar , a Unix (and Linux) command.
Use tar for making backups and for retrieving open source software from the internet.
This command is used to archive directories and extract directories from archives. The tar archives can be used for
back-up storage and for transfer to other machines.
tar files are often used to distribute open source programs.
tar is short for Tape ARchive.
Warning: tar does not work the exact same on all versions of UNIX and Linux. Test and confirm local variations
before relying on scripts moved from one platform to another.
create
Creating a new tar archive:
$ tar cvf archivename.tar dirname/
The c indicates create a new archive, the v indicates verbosely list file names, and the f indicates that the following is the
archive file name. The v option may be left out to avoid a long list.
Creating a new gzipped tar archive:
$ tar cvzf archivename.tar.gz dirname/
$ tar cvzf archivename.tgz dirname/
version 56
page 212
Unix and Linux System Administration and Shell Programming
The z indicates the use of gzip compression. The file extensions .tgz and .tar.gz mean the same thing and are
interchangeable.
Creating a new bzip2 tar archive:
$ tar cvzj archivename.tar.bz2 dirname/
$ tar cvzj archivename.tbz dirname/
The j indicates the use of bzip2 compression. The file extensions .tbz and .tar.bz2 mean the same thing and are
interchangeable.
Bzip2 takes longer to compress and decompress than gzip, but also produces a smaller compressed file.
extract
Extracting from an existing tar archive:
$ tar xvf archivename.tar
The x indicates extract files from an archive, the v indicates verbosely list file names, and the f indicates that the
following is the archive file name. The v option may be left out to avoid a long list.
Extracting from a gzipped tar archive:
$ tar xvfz archivename.tgz
$ tar xvfz archivename.tar.gz
The z option indicates uncompress a gzip tar archive.
Extracting from a bzipped tar archive:
$ tar xvfj archivename.tbz
$ tar xvfj archivename.tar.bz2
The j option indicates uncompress a bzip2 tar archive.
tarpipe
It is possible to tar a set of directories (and their files and them pipe the tar to an extraction at a new location, making an
exact copy of an entire set of directories (including special files) in a new location:
$ tar -c "$srcdir" | tar -C "$destdir" -xv
tarbomb
A malicious tar archive could overwrite existing files. The use of absolute paths can spread files anywhere, including
overwriting existing files (even system files) anywhere. The use of parent directory references can also be used to overwrite
existing files.
version 56
page 213
Unix and Linux System Administration and Shell Programming
One can examine the tar archive without actually extracting from it.
The bsdtar program (the default tar on Mac OS X v10.6 or later) refuses to follow either absolute path or parent-directory
references.
viewing
Viewing an existing tar archive (without extracting) with any of the following:
$
$
$
$
tar
tar
tar
tar
tvf archivename.tar
-tf archivename.tar
tf archivename.tar
--list --file=archivename.tar
Viewing an existing gzipped tar archive (without extracting) with any of the following:
$ tar tvfz archivename.tar.gz
$ tar tvfz archivename.tgz
Viewing an existing bzipped tar archive (without extracting) with any of the following:
$ tar tvfj archivename.tar.bz2
$ tar tvfj archivename.tbz
less
You can pipe the output of a tar file to the less command.
You can directly use the less command to open a tar archive. Less will show you the names of the files, the file sizes, the
file permissions, the owner, and the group (the equivalent of ls -l).
$ less archivename.tar
extracting a single file
You can extract a specific single file from a large tar archive.
$ tar xvf archivefile.tar /complete/path/to/file
You can extract a specific compressed file from a large tar archive.
$ tar xvf archivefile.tar.gz /complete/path/to/file
$ tar xvf archivefile.tar.bz2 /complete/path/to/file
extracting a single directory
version 56
page 214
Unix and Linux System Administration and Shell Programming
You can extract a specific single directory from a large tar archive.
$ tar xvf archivefile.tar /complete/path/to/dir/
You can extract a specific compressed directory from a large tar archive.
$ tar xvf archivefile.tar.gz /complete/path/to/dir/
$ tar xvf archivefile.tar.bz2 /complete/path/to/dir/
extracting a few directories
You can extract specific directories from a large tar archive by giving the name of each of the directories.
$ tar xvf archivefile.tar /complete/path/to/dir1/ /complete/path/to/dir2/
You can extract specific compressed directories from a large tar archive.
$ tar xvf archivefile.tar.gz /complete/path/to/dir1/ /complete/path/to/dir2/
$ tar xvf archivefile.tar.bz2 /complete/path/to/dir1/ /complete/path/to/dir2/
extracting a group of files
You can extract a group of files from a large tar archive by using globbing. Specify a pattern (the following example
finds all .php files):
$ tar xvf archivefile.tar --wildcards '*.php'
You may use the gzip and bzip2 options as well.
adding to an existing archive
You can add an additional file to an existing tar archive with the option -r:
$ tar rvf archivefile.tar newfile
You can add an additional directory to an existing tar archive:
$ tar rvf archivefile.tar newdir/
This does not work for adding a file or directory to a compressed archive. You will get the following error message
instead:
tar: Cannot update compressed archives
version 56
page 215
Unix and Linux System Administration and Shell Programming
verifying files
You can verify a new archive file while creating is by using the -W option:
$ tar cvfW archivefile.tar dir/
If you see something similar to the following output, something has changed:
Verify 1/filename
1/filename: Mod time differs
1/filename: Size differs
If you just see the Verify line by itself, the file or directory is fine:
Verify 1/filename
The -W option does not work for compressed files.
For gzip compressed files, find the difference between the gzip archive and the file system:
$ tar dfz 1/filename.tgz
For bzip2 compressed files, find the difference between the bzip2 archive and the file system:
$ tar dfj 1/filename.tbz
version 56
page 216
Unix and Linux System Administration and Shell Programming
touch
chapter 77
summary
This chapter looks at touch , a Unix (and Linux) command.
Use touch to change the date stamp on a file.
The most common use of touch is to force make to recompile a particular file (or group of files).
how to use
The simple form of touch changes a named file’s timestamp to the current date and time:
$ touch filename
multiple files
Use touch to change all of the files in a directory:
$ touch *
specific time
You can use touch to set a specific time by using the -t option. The following command sets a file to the Solstice at the
end of the Mayan calendar (11:12 a.m. December 21, 2012):
$ touch -t201212211112.00 filename
The following command sets a file to the noon on January 1, 2012:
version 56
page 217
Unix and Linux System Administration and Shell Programming
$ touch -t201201011200.00 filename
The format is YYYYMMDDhhmm.ss , YYYY = four digit year, MM - two digit month, DD = two digit day, hh = two digit
hour (24 hour clock), mm = two digit month, and ss = two digit second.
create an empty file
You can use touch to create an empty file. Simply issue the touch command with the desired file name.
$ touch filename
This will create an empty file with the designated file name.
Why would you want to do this? One common reason is to create an expected file. Some scripts will check for the
existence of particular files and choke if they can’t find them. If the script does not need particular data alreazdy stored in
the file (for example, it will be appending data intot he file or it is simply checking for the existence of the file), then touch
is your fast and easy method for creating the expected file (as an empty file).
version 56
page 218
Unix and Linux System Administration and Shell Programming
find
chapter 78
summary
This chapter looks at find , a Unix (and Linux) command.
find is used to find files.
find a file
You can use find to quickly find a specific file. The -x option (this is -xdev on older systems) limits the search to only
those directories on the existing filesystem.
$ find / -name filename -type d -x
Note that you may want to run this command from root or superuser.
The locate command is more recent.
gotchas
Do not use the output of find in a BASH for loop.
$ for i in $(find *.mp3); do # Wrong!
$ command $i # Wrong!
$ done # Wrong!
$ for i in $(find *.mp3) # Wrong!
$ for i in `find *.mp3` # Wrong!
If any file name has spaces in it, then it will be word split. So, the MP3 for Mumford & Sons’s “The Cave” (number two
song on their “Sigh No More” album) might be 02 The Cave . In that case, your command would run on the files “02”,
“The”, and “Cave”, rather than “02 The Cave”. Additional errors likely to follow. Listen to the song “The Cave” for free at
http://www.thissideofsanity.com/musicbox/musicbox.php?mastersong=48 .
version 56
page 219
Unix and Linux System Administration and Shell Programming
Double quote won’t work either because that will cause the entire output of the find command to be treated as a single
item. Your loop will run once on all of the file names conactenated together.
$ for i in "$(ls *.mp3)"; # Wrong!
The solution is to run your loop without even using find , relying on BASH’s filename expansion. Don’t forget to check
for the possibility that there are no qualifying files in the current directory.
$ for i in *.mp3; do # Wrong!
$ [[ -f "$i" ]] || continue
$ command "$i"
$ done
version 56
page 220
Unix and Linux System Administration and Shell Programming
text processing
chapter 79
summary
This chapter looks at text processing in Linux/UNIX.
The original approach of UNIX was to move data between small programs as character-based text. Most UNIX
commands and tools produce text output as a character stream.
A number of special utilities were developed that allowed a UNIX script to modify the text-based character stream as it
passed between programs. That use is still common used in shell scripts.
Some of these tools are also useful for processing and formatting text for humans to read.
text processing tools
There are several text processing tools discussed in this book.
awk
basename
cat
sed
tail
wc
version 56
page 221
Unix and Linux System Administration and Shell Programming
basename
chapter 80
summary
This chapter looks at basename , a Unix (and Linux) command.
The basename command removes the directory and (optionally) file extension (or suffix) from a file name.
removing directory
Remove the diretcory portion of a filename with this command:
$ basename filename
Don’t forget to use some form of escaping as needed.
$ basename "/home/local/bigfile.txt"
bigfile.txt
$
removing file suffix
Remove the file suffix portion of a filename with this command:
$ basename filename suffix
And an example:
$ basename "/home/local/bigfile.txt" ".txt"
bigfile
$
Note that the suffix portion can be any text. This means you can use this command to remove the last portion of any
text:
version 56
page 222
Unix and Linux System Administration and Shell Programming
$ basename "Los Angeles, CA" ", CA"
Los Angeles
$
sed
chapter 81
summary
This chapter looks at sed , a Unix (and Linux) command.
sed is a command line editor.
fixing end of line
Windows and DOS end every line with the old carriage return and line feed combination (\r\n). You cna use sed to
convert Windows/DOS file format to UNIX file format:
$ sed 's/.$//' filename
adding line numbers
You can use sed to add line numbers to all of the non-empty lines in a file:
$ sed '/./=' filename | sed 'N; s/\n/ /'
version 56
page 223
Unix and Linux System Administration and Shell Programming
awk
chapter 82
summary
This chapter looks at awk , a Unix (and Linux) command.
awk is a programming language intended to serve a place between simple shell scripts and full programming languages.
It is Turing complete.
remove duplicate lines
You can use awk to remove duplicate lines from a file.
$ awk '!($0 in array) { array[$0]; print }' filename
version 56
page 224
Unix and Linux System Administration and Shell Programming
screencapture
chapter 83
summary
This chapter looks at screencaptures , a Mac OS X-only command.
screencapture creates an image of the screen or a portion of the screen.
from the graphic user interface
The normal method for obtaining a screen capture is through the graphic user interface. Command-Shift-3 takes a
screenshot of the screen and saves it as a file to the desktop under the name of “Picture 1” (or next available number if
there are already screenshots saved there).
If you have multiple monitors connected, each monitor is saved as a separate picture, named “Picture 1”, “Picture 1(2)”,
“Picture 1(3)”, etc.
With Mac OS X 10.6 (Snow Leopard) the default name changes to “Screen shot YYYY-MM-DD at HH.MM.SS XM”,
where YYY=year, MM=month, DD=day, HH=hour, MM=minute, SS=second, and XM = either AM or PM.
The basic screen capture options:
Command-Shift-3: Take a screenshot of the entire screen and save it as a file on the desktop.
Command-Shift-4, then select an area: Take a screenshot of an area and save it as a file on the desktop.
Command-Shift-4, then space, then click on a window: Take a screenshot of a selected window and save it as a
file on the desktop.
Command-Control-Shift-3: Take a screenshot of the screen and save it to the clipboard.
Command-Control-Shift-4, then select an area: Take a screenshot of an area and save it to the clipboard.
Command-Control-Shift-4, then space, then click on a window: Take a screenshot of a selected window and save
it to the clipboard.
In Mac OS X 5 (Leopard) or more recent, the following keys can be held down when selecting an area (with either
Command-Shift-4 or Command-Control-Shift-4):
Space: Used to lock the size of the selected region and move the selected region as the mouse moves.
Shift: Used to resize only one edge of the slected region.
Option: Used to resize the selected region with its center as the anchor point.
Different versions of Mac OS X have different default file formats for saving the screenshot:
Mac OS X 10.2 (Jaguar): jpg
Mac OS X 10.3 (Panther): pdf
version 56
page 225
Unix and Linux System Administration and Shell Programming
Mac OS X 10.4 (Tiger): png
changing defaults
The following methods use Terminal to change the default file format and location where the screenshot is saved from
the graphic user interface.
In Mac S X 10.4 (Tiger) or more recent, the default screencapture format can be changed in Terminal by using the
defaults command. In Mac S X 10.4 (Tiger), the new default does not take effect until you logout and log back in(from
the entire computer, not just from Terminal — a full restart will also work) unless you also use the killall command.
$ defaults write com.apple.screencapture type ImageFormat
$ killall SystemUIServer
The FileFormat can be png (Portable Network Graphic), pdf (Portable Document Format), tiff (Tagged Image File
Format), jpg or jpeg (Joint Photographic Experts Group), pict (Macintosh QuickDraw Picture), bmp (Microsoft Windows
Bitmap), gif (Graphics Interchange Format), psd (Adobe Photoshop Document), sgi (Silicon Graphics File Format), or tga
(Truevision Targe File Format).
JPGs are saved at quality 60%.
To change the default location where the screenshot file is saved (the default is Desktop), use the following Terminal
command (where pathname is the full path to a directory.:
$ defaults write com.apple.screencapture location PathName
$ killall SystemUIServer
The normal default location would be reset with the following command (where USername is the current account’s user
name.
$ defaults write com.apple.screencapture location /Users/UserName/Desktop
$ killall SystemUIServer
command line screenshots
You can also take screenshots from Terminal.
I needed a screenshot of the CONTROL-TAB selection of a program, but in the graphic user interface, I couldn't
simultaneously run Command-Tab and Command-Shift-4, so I used the following command in Terminal to set a 10 second
delay and save the screenshot selection:
$ screencapture -T 10 -t png controltab.png
version 56
page 226
Unix and Linux System Administration and Shell Programming
You can add this command to your Mac OS X scripts.
The format is screencapture options filenames . List more than one file name if you have more than one monitor.
You can use the options in any combination.
You can use the filename to change the file name from the normal default and to set a relative path to a directory/folder
of your choice.
$ screencapture [-icMPmwsWxSCUt] [files]
The basic use, which takes an immediate screenshot in the default format and stores it with the designated filename (in
this case “Picture1”) in the user’s home directory (not the desktop).
$ screencapture Picture1
Force the screenshot to go to the clipboard (the equivalent of the Command-Shift-Control- choices).
$ screencapture -c [files]
Capture the cursor as well as the screen. This applies only in non-interactive modes (such as a script).
$ screencapture -C [files]
Display errors to the user graphically.
$ screencapture -d [files]
Capture the screenshot interactively by either selection or window (the equivalent of Command-Shift-4). Use the
CONTROL key to cause the screenshot to go to the clipboard. Use the SPACE key to toggle between mouse selection and
window selection modes. Use the ESCAPE key to cancel the interactive screen shot.
$ screencapture -i [file]
Use the -m option to only capture the main monitor. This does not work if the -i option is also set.
version 56
page 227
Unix and Linux System Administration and Shell Programming
$ screencapture -m [file]
Send the screenshot to a new Mail message.
$ screencapture -M [files]
Use the -o option in window capture mode to only capture the window and to not capture the shadow of the window.
$ screencapture - o [file]
After savng the screenshot, open the screen capture output in Preview.
$ screencapture -P [files]
Use -s to only allow mouse selection mode.
$ screencapture -s [files]
Use -w to only allow window selection mode.
$ screencapture -w [file]
Use -W to start interaction in the window selection mode.
$ screencapture -W [file]
Use the -S option in window capture mode to capture the screen rather than the window.
$ screencapture -S [files]
Set the format with the -t option. The Format can be png (Portable Network Graphic), pdf (Portable Document Format),
tiff (Tagged Image File Format), jpg or jpeg (Joint Photographic Experts Group), pict (Macintosh QuickDraw Picture), bmp
(Microsoft Windows Bitmap), gif (Graphics Interchange Format), psd (Adobe Photoshop Document), sgi (Silicon Graphics
File Format), or tga (Truevision Targe File Format)
$ screencapture -tFormat [files]
Set a delay time in seconds. The default is five seconds.
$ screencapture -TSeconds [files]
version 56
page 228
Unix and Linux System Administration and Shell Programming
Prevent the playing of sounds (no camera click sound).
$ screencapture -x [files]
signals
chapter 84
summary
This chapter looks at UNIX and Linux signals.
The first chart shows the signals by name, with their explanation and their numbers by operating system. The second
chart shows the signals by number, with the corresponding name by operating system.
signals chart by name
name
SIGABRT
meaning
Used by abort
Linux
Mac OS X
Solaris
6
6
create core image
SIGALRM
Alarm clock, real-time timer expired (POSIX)
14
14
14
terminate process
SIGBUS
BUS error (BSD 4.2)
7
10
10
create core image
SIGCANCEL Thread cancellation signal used by libthread
Child process has stopped or exited, Child status change alias (POSIX)
SIGCHLD
17
20
36
18
discard signal
SIGCONT
Continue executing, if stopped; Stopped process has been continued
(POSIX)
SIGEMT
EMT instruction; Emulate instruction executed
18
19
25
discard signal
7
7
create core image
SIGFPE
Floating point exception (ANSI)
8
8
8
create core image
SIGFREEZE
SIGHUP
Special signal used by CPR
Hangup (POSIX)
1
1
34
1
terminate process
SIGILL
Illegal instruction (ANSI)
4
4
create core image
version 56
page 229
4
Unix and Linux System Administration and Shell Programming
SIGINFO
status request from keyboard
29
terminate process
SIGINT
Terminal interrupt (ANSI)
2
2
2
terminate process
SIGIO
I/O now possible (BSD 4.2)
29
23
22
discard signal
SIGIOT
IOT Trap (BSD 4.2)
6
6
create core image
now called
SIGABRT
SIGKILL
SIGLOST
SIGLWP
SIGPIPE
Kill (POSIX)
(can’t be caught or ignored)
9
Resource lost
Special signal used by thread library
Write on a pipe with no reader, Broken pipe (POSIX)
37
13
13
33
9
9
terminate process
13
terminate process
SIGPOLL
Pollable event occurred or Socket I/O possible
Profiling alarm clock, profiling timer expired (BSD 4.2)
SIGPROF
27
27
22
29
terminate process
SIGPWR
SIGQUIT
Power failure restart (System V)
Terminal quit (POSIX)
30
3
19
3
3
create core image
SIGRTMAX
SIGRTMIN
SIGSEGV
Lowest priority real-time signal
Highest priority real-time signal
Invalid memory segment access, Segmentation violation (ANSI)
11
11
45
38
11
create core image
SIGSTKFLT
SIGSTOP
SIGSYS
Stack fault
Stop executing (POSIX)
(can’t be caught or ignored)
Bad argument to system call; Non-existent system call invoked
16
19
17
23
stop process
12
12
create core image
SIGTERM
Software termination (ANSI)
15
15
15
terminate process
SIGTHAW
Special signal used by CPR
SIGTRAP
Trace trap (POSIX)
5
5
35
5
create core image
SIGTSTP
Terminal stop signal, User stop requested from TTY (POSIX)
20
18
24
stop process
SIGTTIN
Background process trying to read from TTY control terminal (POSIX)
21
21
26
stop process
SIGTTOU
background process trying to write to TTY control terminal (POSIX)
22
22
27
stop process
SIGURG
Urgent condition on socket (BSD 4.2)
23
16
21
discard signal
SIGUSR1
User defined signal 1 (POSIX)
10
30
16
terminate process
SIGUSR2
User defined signal 2 (POSIX)
12
31
17
terminate process
SIGVTALRM Virtual alarm clock, Virtual timer expired (BSD 4.2)
version 56
26
26
page 230
28
Unix and Linux System Administration and Shell Programming
terminate process
SIGWAITING Process’ LWPs are blocked
SIGWINCH Window size change (BSD 4.3, Sun)
28
28
32
20
discard signal
SIGXCPU
CPU limit exceeded (BSD 4.2)
24
24
30
terminate process
SIGXFSZ
File size limit exceeded (BSD 4.2)
25
25
terminate process
signals chart by number
number
1
Linux
SIGHUP
Mac OS X
SIGHUP
Solaris
SIGHUP
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
SIGINT
SIGQUIT
SIGILL
SIGTRAP
SIGIOT
SIGBUS
SIGFPE
SIGKILL
SIGUSR1
SIGSEGV
SIGUSR2
SIGPIPE
SIGALRM
SIGTERM
SIGSTKFLT
SIGCHLD
SIGCONT
SIGSTOP
SIGINT
SIGQUIT
SIGILL
SIGTRAP
SIGABRT
SIGEMT
SIGFPE
SIGKILL
SIGBUS
SIGSEGV
SIGSYS
SIGPIPE
SIGALRM
SIGTERM
SIGURG
SIGSTOP
SIGTSTP
SIGCONT
SIGINT
SIGQUIT
SIGILL
SIGTRAP
SIGABRT
SIGEMT
SIGFPE
SIGKILL
SIGBUS
SIGSEGV
SIGSYS
SIGPIPE
SIGALRM
SIGTERM
SIGUSR1
SIGUSR2
SIGCHLD
SIGPWR
20
21
22
SIGTSTP
SIGTTIN
SIGTTOU
SIGCHLD
SIGTTIN
SIGTTOU
SIGWINCH
SIGURG
SIGIO
SIGPOLL
23
24
SIGURG
SIGXCPU
SIGIO
SIGXCPU
SIGSTOP
SIGTSTP
25
26
27
28
version 56
SIGXFSZ
SIGXFSZ
SIGCONT
SIGVTALRM SIGVTALRM SIGTTIN
SIGPROF
SIGPROF
SIGTTOU
SIGWINCH SIGWINCH SIGVTALRM
29
30
SIGIO
SIGPWR
SIGINFO
SIGUSR1
SIGPROF
SIGXCPU
31
SIGUSR2
SIGXFSZ
page 231
31
Unix and Linux System Administration and Shell Programming
32
33
34
35
36
37
38
45
version 56
SIGWAITING
SIGLWP
SIGFREEZE
SIGTHAW
SIGCANCEL
SIGLOST
SIGRTMIN
SIGRTMAX
page 232
Unix and Linux System Administration and Shell Programming
kernel modules
chapter 85
summary
This chapter looks at UNIX and Linux kernel modules.
Loadable Kernel Modules (LKM) go by many different names, including kernel loadable module (kld) in FreeBSD and
kernel extension (kext) in Mac OS X, as well as Kernel Loadable Modules (KLM) or Kernel Modules (KMOD).
purpose
Loadable kernel modules allow the flexibility to add additional functionality to an operating system without requiring
recompiling and reloading the entire operating system. This same flexibility can be used to choose between various options
at start-up time.
Linux
Loadable Kernel Modules (LKMs) are loaded (and unloaded) by the modprobe command. Linux LKMs are located in
/lib/modules and have the file extension .ko (for kernel object). Prior to Linux version 2.6 they were identified by the .o
file extension.
Linux views LKMs as derived works of the kernel and allow symbols to be marked as only available to GNU General
Public License (GPL) modules. Loading a proprietary or non-GPL-compatible LKM will set a “taint” flag in the running
module.
LKMs run as part of the kernel, so they can corrupt kernel data structures and produce bugs.
FreeBSD
Kernel modules for FreeBSD are stored within /boot/kernel/ for modules distributed with the OS, or usually
/boot/modules/ for modules installed from FreeBSD ports or FreeBSD packages, or for proprietary or otherwise binaryonly modules. FreeBSD kernel modules usually have the extension .ko . Once the machine has booted, they may be loaded
with the kldload command, unloaded with kldunload , and listed with kldstat. Modules can also be loaded from the
loader before the kernel starts, either automatically (through /boot/loader.conf) or by hand.
Mac OS X
Some loadable kernel modules in OS X can be loaded automatically. Loadable kernel modules can also be loaded by the
kextload command. They can be listed by the kextstat command. Loadable kernel modules are located in application
version 56
page 233
Unix and Linux System Administration and Shell Programming
bundles with the extension .kext . Modules supplied with the operating system are stored in the
/System/Library/Extensions directory; modules supplied by third parties are in various other directories.
binary compatibility
Linux does not provide a stable API or ABI for kernel modules. This means that there are differences in internal structure
and function between different kernel versions, which can cause compatibility problems. In an attempt to combat those
problems, symbol versioning data is placed within the .modinfo section of loadable ELF modules. This versioning
information can be compared with that of the running kernel before loading a module; if the versions are incompatible, the
module will not be loaded.
Other operating systems, such as Solaris, FreeBSD, Mac OS X, and Windows keep the kernel API and ABI relatively
stable, thus avoiding this problem. For example, FreeBSD kernel modules compiled against kernel version 6.0 will work
without recompilation on any other FreeBSD 6.x version, e.g. 6.4. However, they are not compatible with other major
versions and must be recompiled for use with FreeBSD 7.x, as API and ABI compatibility is maintained only within a
branch.
security
While loadable kernel modules are a convenient method of modifying the running kernel, this can be abused by attackers
on a compromised system to prevent detection of their processes or files, allowing them to maintain control over the
system. Many rootkits make use of LKMs in this way. Note that on most operating systems modules do not help privilege
elevation in any way, as elevated privilege is required to load a LKM; they merely make it easier for the attacker to hide the
break-in.
Linux allows disabling module loading via /proc/sys/kernel/modules_disabled . A initramfs system may load
specific modules needed for a machine at boot and then disable module loading. This makes the security very similar to a
monolithic kernel. If an attacker can change the initramfs , they can change the kernel binary.
On Mac OS X, a loadable kernel module in a kernel extension bundle can be loaded by non-root users if the
property is set to True in the bundle’s property list. However, if any of the files in the bundle,
including the executable code file, are not owned by root and group wheel, or are writable by the group or “other”, the
attempt to load the kernel loadable module will fail.
OSBundleAllowUserLoad
version 56
page 234
Unix and Linux System Administration and Shell Programming
LAMP
chapter 86
summary
This chapter looks at LAMP, which is the combination of Linux-Apache-MySQL-PHP (or Perl or Python).
These are all open source programs that provide the ability to have a working modern production web server.
MAMP
The MAMP project achieves the same capabilities using Mac OS X instead of Linux. The MAMP package can be
downloaded and used for free without interfering with any similar built-in Mac OS X services. MAMP is intended for
creating a local duplicate of your production server environment so that you can freely test new software before deploying
it to your production servers.
MAMP can be found at http://www.mamp.info/en/index.html.
WAMP
The WampServer project achieves the same capabilities using Windows instead of Linux. The WampServer package can
be downloaded and used for free without interfering with any similar built-in Windows services. WampServer is intended
for creating a local duplicate of your production server environment so that you can freely test new software before
deploying it to your production servers.
WampServer can be found at http://www.wampserver.com/en/.
version 56
page 235
Unix and Linux System Administration and Shell Programming
mysql
chapter 87
summary
This chapter looks at mysql , a Unix (and Linux) command.
mysql is the most common SQL database on Linux.
interactive queries
If you need to run a single interactive query on a particular MySQL database, use the -e flag.
$ mysql DBname - e "select * from table_foo;"
official keywords
The following is the list of the keywords for the official SQL standard:
DELETE
INTO
ALL
DESC
IS
AND
DISTINCT KEY
ANY
DOUBLE
LANGUAGE
AS
END
LIKE
ASC
MAX
AUTHORIZATION ESCAPE
EXEC
MIN
AVG
EXISTS
MODULE
BEGIN
FETCH
NOT
BETWEEN
FLOAT
NULL
BY
FOR
NUMERIC
CHAR
FOREIGN OF
CHARACTER
FORTRAN ON
CHECK
FOUND
OPEN
CLOSE
FROM
OPTION
COBOL
GO
OR
COMMIT
GOTO
ORDER
CONTINUE
GRANT
PASCAL
COUNT
GROUP
PLI
CREATE
HAVING
PRECISION
CURRENT
IN
PRIMARY
CURSOR
DEC
INDICATOR PRIVILEGES
version 56
REFERENCES
ROLLBACK
SCHEMA
SECTION
SELECT
SET
SMALLINT
SOME
SQL
SQLCODE
SQLERROR
SUM
TABLE
TO
UNION
UNIQUE
UDATE
USER
VALUES
VIEW
WHENEVER
WHERE
page 236
Unix and Linux System Administration and Shell Programming
DECIMAL
DECLARE
DEFAULT
INSERT
INT
INTEGER
PROCEDURE WITH
PUBLIC
WORK
REAL
PHP
chapter 88
summary
This chapter looks at php , a Unix (and Linux) command, including both running shell scripts from a web browser and
running PHP from the shell.
php is used to run a CLI version of PHP.
php script example
This example assumes that you have created the scripts directory in your home directory.
Create a php script called script.php and save it in the new scripts directory (folder):
<?php
echo "Hello World!"
?>
Notice that we skip the chmod step typical for activating shell scripts.
Run your new script by running the php program with your script as the file to execute:
$ php ~/scripts/script.php
Hello World!
$
Note that the shell does not render HTML, so if you run a web scirpt, you will see raw HTML, CSS, and JavaScript as
plain text in your terminal window.
You can run perl, ruby, and python scripts in the same manner.
running shell from PHP
You can run shell commands and shell scripts from PHP in Apache by using either the shell_exec function or backticks.
The shell_exec function in PHP takes a shell command as a string, runs that command (or command list) in shell, and
returns the rsult as a string. You can assign the result to a string variable and then output those results to a web browser,
version 56
page 237
Unix and Linux System Administration and Shell Programming
with or without additional processing. Use the <pre> tag to have the browser use the UNIX new line characters.
<php
$linesofoutput = shell_exec('ls');
echo '<pre>'.$linseofoutput.'</pre>';
?>
You can also use the backtick (or back quote) character ( `).
<php
$linesofoutput = `ls`;
echo '<pre>'.$linseofoutput.'</pre>';
?>
You cannot combine both methods. This is useful, because you might actually want to send the backticks to the shell. If
you are using backticks to run a shell command or command list, you can escape internal backticks to send them to the
shell.
web security
Most system administrators disable shell_exec(). When shell_exec is disbled, the PHP script will be unable to run any
shell commands.
If you enable shell access, you must guarantee that web visitor submitted input is never sent to the command shell
without thorough scrubbing or crackers will be able to take over your server.
The difficulty with that “guarantee” is that there is probably a lot of PHP you didn’t write, such as WordPress, a PEAR
package, off-the-shelf software, etc.
Hence the rule: disable shell_exec() and similar functions for any public facing server. Thanks to Stuart Beeton for
emphasizing this point.
command line PHP
PHP is available both as a CGI binary (for Apache) and a CLI binary (for the shell).
Use the -v option to find out how your CLI copy of PHP is set up.
$ php -v
PHP 5.3.15 with Suhosin-Patch (cli) (built: Jul 31 2012 14:49:18)
Copyright (c) 1997-2012 The PHP Group
Zend Engine v2.3.0, Copyright (c) 1998-2012 Zend Technologies
$
Notice that the above output indicated the cli version.
PHP defaults
The CLI version of PHP does not send out any HTTP headers, because the shell doesn’t understand them and instead
expects plain text.
The CLI version of PHP does not change the working directory to that of the script. For a web server, you want to find
all the files from the current web base. For a script running in the shel, you want the script or operator to make the decision.
version 56
page 238
Unix and Linux System Administration and Shell Programming
This allows a PHP script running in the shell to be generic and simply run in the current working directory.
The CLI version of PHP has several php.ini options overridden by default. These changes are enforced after the php.ini
file is run, so you can’t change them in the configuration file, although you can change them during runtime.
html_errors is FALSE(because the shell is cluttered by uninterpretted HTML tags.
implicit_flush is TRUE because the shell should receive immediate outputs of print, echo, and similar commands
rather than having them held in a buffer.
max_execution_time is 0 (unlimited) because a shell environment might run very much more complex scripts than
those in a typical web-based script.
register_argc_argc is TRUE because the PHP script will need to process the command line arguments. The PHP
variable $argc has the number of arguments passed and the variable $argv is an array of the actual arguments. These
values are also in the $_SERVER array, as $_SERVER['argv'] and $_SERVER['argc'].
output_buffering is hardcoded FALSE< but the output buffering functions are available.
max_input_time is FALSE because PHP CLI does not support GET, POST, or file uploads.
running a file in PHP
Use the -f option to parse and execute a designated file in PHP.
$ php -f filename.php
$
Note that the -f switch is optional. You will get the same result if you leave off the -f option and just give the filename.
You get the exact same results from:
$ php filename.php
$
Note that the .php extension is for your convenience.The CLI PHP will run any file as if it were a PHP file, without
regard to the extension type.
run PHP interactively
Use the -a switch to run PHP interactively. You will receive a secondary prompt and each line of PHP code will be run
as soon as a ; or } is found. Do not include the script tags <? and ?>.
$ php -a
asdf
$
run single line of PHP
Use the -r switch to run some particular PHP code immediately. This is useful inside of a script. Do not include the
script tags <? and ?>.
$ php -r 'echo 'hello world\n'
hello world
$
version 56
page 239
Unix and Linux System Administration and Shell Programming
You need to carefully watch your use of quoting and escaping to avoid problems with the command line variable
substitution done by the shell.
syntax check
Use the -l switch to perform a syntax check on the PHP code.
$ php filename
No syntax error detected in <filename>
$
If there are no errors, you will get the message No syntax errors detected in <filename > on standard output and
the shell return code is 0.
If there are no errors, you will get the message Errors parsing <filename > on standard output and the shell return
code is -1.
This option does not work with the -r option.
use PHP in a pipeline
PHP will accept PHP code from standard input (STDIN). Note this means that the previous commands are creating PHP
code, not passing data for a PHP file to use as input.
$ some_application | some_filter | php | some_command > final_output.txt
$
combinations
You cannot combine any of the three ways of executing PHP code (file, -r, and STDIN). It would be great if there was a
way to accept data from STDIN and name a PHP file to process it, with the output continuing as STDOUT to the next
commadn in the shell pipeline, but this isn’t currently a choice.
arguments
PHP can receive an unlimited number of arguments. Unfortunately, the shell has a limit on the number of total characters
which can be passed to a command.
An argument that starts with a - will be used by the PHP interpreter. Use the argument list separator -- and all of the
arguments that follow will be sent to the PHP script through the $argv variable.
shebang
You can insert the shebang line with the correct location of the CLI PHP program and set the execution attrbutes of the
file. The file needs the normal PHP starting and end tags.
Use the type command to find the location of your CLI copy of PHP.
$ type php
php is /usr/bin/php
$
version 56
page 240
Unix and Linux System Administration and Shell Programming
Add that location to the standard shebang as the first line of the file.
#!/usr/bin/php
Set the file to execute:
$ chmod +x filename.php
$
The shebang line has no effect on Windows, but also causes no harm because it is also a valid PHP comment. Therefore
include the shebang line for true cross-platform PHP scripts.
shebang example
Create a sample script file.
#!/usr/bin/php
<?php
var_dump($argv);
?>
Make the script file excutable and then run it.
$ chmod +x phptest.php
$ ./phptest.php -h -- foo
array(4) {
[0]=>
string(13) "./phptest.php"
[1]=>
string(2) "-h"
[2]=>
string(2) "--"
[3]=>
string(3) "foo"
}
$
version 56
page 241
Unix and Linux System Administration and Shell Programming
Perl
chapter 89
summary
This chapter looks at perl , a Unix (and Linux) command, including both running shell scripts from a web browser and
running Perl from the shell.
perl is used to run a CLI version of Perl.
getting started
Check that Perl is actually running on your system by typing perl -v from the command line. Your exact results may
vary.
$ perl -v
This is perl 5, version 12, subversion 4 (v5.12.4) built for darwin-thread-multi2level
(with 2 registered patches, see perl -V for more detail)
Copyright 1987-2010, Larry Wall
Perl may be copied only under the terms of either the Artistic License or the
GNU General Public License, which may be found in the Perl 5 source kit.
Complete documentation for Perl, including FAQ lists, should be found on
this system using "man perl" or "perldoc perl". If you have access to the
Internet, point your browser at http://www.perl.org/, the Perl Home Page.
$
Use which to determine the location of your local copy of Perl:
$ which perl
This is perl 5, version 12, subversion 4 (v5.12.4) built for darwin-thread-multi2level
/usr/bin/perl
$
This example assumes that you have created the scripts directory in your home directory.
Create a Perl script called script.pl and save it in the new scripts directory (folder):
version 56
page 242
Unix and Linux System Administration and Shell Programming
#!/usr/bin/perl
print "Hello World!";
Notice that we skip the chmod step typical for activating shell scripts.
Run your new script by running the perl program with your script as the file to execute:
$ perl ~/scripts/script.pl
Hello World!
$
brief history
Perl created by Larry Wall in 1986. Perl was originally intended as a utility to create reports. Perl stands for Practical
Extraction and reporting Language.
version 56
page 243
Unix and Linux System Administration and Shell Programming
Tcl
chapter 90
summary
This chapter looks at tcl , a Unix (and Linux) command, including both running shell scripts from a web browser and
running Tcl from the shell.
tcl is used to run a CLI version of Tcl.
getting started
Use which to determine the location of your local copy of Tcl (and to make sure you actually have a copy):
$ which tclsh
/usr/bin/tclsh
$
This example assumes that you have created the scripts directory in your home directory.
Create a tcl script called script.tcl and save it in the new scripts directory (folder):
#!/usr/bin/tclsh
puts "Hello World!"
Run your new script by running the tclsh program with your script as the file to execute:
$ tclsh ~/scripts/script.tcl
Hello World!
$
text-based interactive version
To run a text-only version of tclsh , simply type tclsh (and return) from the BASH command line and you should get a
percent sign ( %) prompt. This means you have sccessfully entered tclsh .
$ tclsh
%
version 56
page 244
Unix and Linux System Administration and Shell Programming
Now that you are in tclsh , type info tclversion to see which version.
% info tclversion
8.5
%
And to see Tcl make use of a simple BASH command, type ls -l:
% ls -l
a regular listing of files
%
And to return back to BASH, type exit :
% exit
$
visual interactive version
Now test out the visual version by typing wish (assuming it is installed and you have X-Windows running).
$ wish
%
The contents of your Terminal window will not change from the text-only version, but you will get an additional display
window that is controlled from the command line.
And return to BASH:
% exit
$
version 56
page 245
Unix and Linux System Administration and Shell Programming
brief history
Tcl was created in the late 1980s by Dr. John Osterhout. Tcl stands for Tool Command Language.
version 56
page 246
Unix and Linux System Administration and Shell Programming
installing software from source code
chapter 91
summary
This chapter looks at installing software from source code.
While it is possible to do a lot with just preinstalled software and precompiled software, those working with large
systems will sooner or later need to download the raw source code and use that source code to create their own copy of the
latest version of some software on their servers or other computer systems.
gcc
The vast majority of open source software is written in the C programming language. The GNU foundation C compiler,
known as gcc , is the most commonly used C compiler.
Apple’s Mac OS X Developer Tools include a copy of gcc , but also include an Apple specific variation that includes
specialized features useful for development of Macintosh, iPhone, and iPad software.
Xcode
Xcode is an Apple integrated development environment. It includes the Instruments analysis tool, iOS Simulator (useful
for testing your software on the Mac before it is ready for deployment to real devices), and the latest Mac OS X and iOS
SDKs.
Fink
Fink (German for the bird finch) is an open source program that automates the process of downloading, configuring,
compiling, and installing hundreds of software packages from the original source code.
manual installation
There are two major methods for manually installing software from source: the standard method (used for most Linux
software) and the Perl method (used for Perl modules).
version 56
page 247
Unix and Linux System Administration and Shell Programming
programming
chapter 92
summary
This chapter is intended as an introduction to computer programming.
Programming is problem solving and writing instructions for a computer.
The principles of programming are independent of the computer programming language used. Different languages have
different strengths and weaknesses, making some kinds of programs easier or more difficult to write, but the basic
principles remain the same regardless of language.
A skilled programmer should be able to switch to a new programming language in a few hours.
On the other hand, beginners should pick one language and learn it before attempting a second language.
This free text book includes information on multiple programming languages. Unless instructed otherwise, you should
concentrate on the language you are learning and skip over the others. Trying to learn the syntax and semantics of multiple
programming languages at the same time as learning the basics of programming is a recipe for utter confusion.
“As long as programmers cherish their freedom not only to design their own clever software, but also to
modify adopted software according to their likings, a proper design discipline remains unlikely. And as long as
companies secretly cherish complexity as an effective protection against being copied, there is little hope for
dramatic improvements of the state of the art.” —Niklaus Wirth
programming as a skill
I would like to strongly suggest that learning the principles of computer programming is far more valuable than learning
a specific programming language.
“Hire a programmer. He’s a programmer. You don’t need to hire a Java programmer. By the way, you should
do that now anyway. Do not hire Java programmers, not because Java programmers are bad. But you should
not be asking, “excuse me, what language do you program in? We’re only hiring Java programmers here.”
Learning a language is a matter of a week. Oh, and to get skillful in the platform is a matter of another week or
two or three or a month. Finding someone who actually knows how to write code, that’s the hard problem, so
you don’t ask, “you must know Sprint.” No, you do not have to know Sprint. “you must know JavaScript.” No,
you don’t have to know JavaScript. “Can you write code?” Show me code in any language. If you can write
code, I’ll hire you.” —Robert C. Martin, The Last Programming Language, approximately 37:09
“Learning how to code in any one particular [computer] language is not going to be worthwhile beyond 10 or
20 years. Learning how to problem-solve using algorithms and how technology works and how it’s built is
going to last a century at least,” —Hadi Partovi, cofounder of code.org, For some, learning coding is a
calculated strategy, Los Angeles Times , Business Section, Saturday, August 2, 2014, p. B1
other
“7. It is easier to write an incorrect program than understand a correct one.” —Alan Perlis, Epigrams on
Programming, ACM’s SIGPLAN Notices Volume 17, No. 9, September 1982, pages 7-13
version 56
page 248
Unix and Linux System Administration and Shell Programming
“93. When someone says ‘I want a programming language in which I need only say what I wish done,’ give
him a lollipop.” —Alan Perlis, Epigrams on Programming, ACM’s SIGPLAN Notices Volume 17, No. 9,
September 1982, pages 7-13
version 56
page 249
Unix and Linux System Administration and Shell Programming
size of programs
chapter 93
summary
The educational goal of this chapter is to introduce the concept of differences in scale or size of programming projects
and how the size and complexity of a program impacts the task of programming.
Programs vary in size. Bad habits learned with small or trivial programs might get you into real trouble with medium or
large programs.
size of programs
Programs are generally divided into three basic sizes: trivial, small, and large.
Trivial programs are programs that a skilled programmer can write in less than two days of coding.
Small programs are programs that one skilled programmer can write in less than one year of full time work.
Large programs are programs that require more than two to five man-years of labor, normally written by programming
teams (which can exceed 1,000 skilled workers).
These estimates are approximate and there are obvious gaps in the gray zone between the sizes. Further, there can be
huge differences in individual abilities.
Larry Ellison wrote the first version of Oracle database by himself in about six months. That is a genius exception. Data
bases typically take large teams (sometimes hundreds of programmers) at least a year.
Bill Gates, copying and pasting from the source code of three working open source versions, took more than six months
to create a bug-filled BASIC compiler and then hired a team of six skilled programmers who spent more than six more
months to get rid of enough bugs to make the compiler somewhat usable (a total of more than three man-years). That is an
idiot exception. A BASIC compiler typically takes a skilled programmer a few hours to create. Note that Bill Gates takes
credit for quickly having created a BASIC compiler, but according to other sources he was sued for having illegally used
open source code for commercial purposes, forcing him to spend a great deal of time attempting to do a project that many
programmers of the day could successfully finish in hours.
impact on good programming practices
Almost every program assigned in a class setting will be trivial, simply because there isn’t enough time in a quarter or
semester for longer programs.
Each programming assignment will concentrate on one or a small number of specific programming concepts.
The artificial nature of school programming assignments cause most students to question the utility of modern
programming practices, especially the time and effort spent on form and documentation.
These practices are the result of decades of real world programming.
Successful programs tend to have a long lifetime. Programmers will have to look at existing source code, figure out what
is going on, and then correctly modify the program to add new features or update existing features to meet changing real
world conditions.
version 56
page 250
Unix and Linux System Administration and Shell Programming
Unix style
The Unix philosophy on programming is to encourage a lot of small tools (programs) that work together to complete a
task (usually threaded together with a shell script). This bottom-up approach to programming encourages the reuse and
sharing of software. The typical Unix (or LINUX) distribution includes hundreds of small standardized tools that allow a
skilled Unix prorgammer (or even a skilled Unix administrator or user) to immediately do useful large scale work.
object-oriented programming
Many students openly question why they are going to so much extra work to create object-oriented projects when
procedural programming will clearly be sufficient and involve less work.
The simple answer is that object-oriented programming is one of the few proven successful methods for cutting costs and
improving reliability of large scale real-world projects.
version 56
page 251
Unix and Linux System Administration and Shell Programming
kinds of programs
chapter 94
summary
The educational goal of this chapter is to introduce the basic different kinds of programming in common use.
The two major kinds of programming are systems programming and applications programming.
list of kinds of programming
There are two basic kinds of programming: system and application.
System programming deals with the use of a computer system. This includes such things as the operating system, device
drivers for input and output devices, and systems utilities.
Application programming deals with the programs directly used by most people.
Application programming is generally divided further into scientific and business programming.
Scientific programming is work in the scientific, engineering, and mathematical fields. Often the programmers are the
researchers who use the programs.
Business programming is work in the data processing field, including large scale business systems, web-based
businesses, and office applications. It is exceedingly rare for these kinds of programs to be created by the person who uses
them.
Another large category of programming is programming for personal or home use. This includes games. Historically,
many advances in computer science occurred in the development of computer and video games.
Embedded systems are programs that are built into specific hardware, such as the computer systems in an automobile or
microwave oven. These programs combine features of operating systems and application program into a single monolithic
system.
Scripting is a simplified version of programming originally intended for use primarily by non-programmers. In practice,
most non-programmers have trouble with scripting languages. Some professional programmers have built very useful,
sometimes intricate systems using scripting languages, especially those contained in office software (such as word
processors or spreadsheets).
version 56
page 252
Unix and Linux System Administration and Shell Programming
programming languages
chapter 95
summary
The educational goals of this chapter are to familiarize the student with the kinds of programming languages and the
basic tools for programming. Key terminology is introduced (the student should know the definition of terms in bold). This
is followed by some C specific information from the the Stanford CS Education Library.
Programming languages vary from low level assemblers to high level languages.
“19. A language that doesn’t affect the way you think about programming, is not worth knowing.” —Alan
Perlis, Epigrams on Programming, ACM’s SIGPLAN Notices Volume 17, No. 9, September 1982, pages 7-13
direct programming
Originally computers were programmed directly in a “language” that the computer understood.
This direct programming could involve directly wiring the program into the computer. In some cases, this involved a
soldering iron. In other cases there was some kind of plug-board ot make it easier to change the programmed instructions.
This method was known as hard wiring.
Large telegraph networks and later large telephone networks became so complex as to essentially be a computer on a
system-wide basis. Many of the ideas (especially logic circuits) that were later necessary to create computers were first
developed for large scale telegraph and telephone systems.
In some early computers the programming could be accomplished with a set of switches. The use of front panel
switches (and corresponding indicator lights) continued as an option on many mainframe and minicomputer systems. Some
microcomputer systems intended for hobbyists and for dedicated systems also had some kind of front panel switches.
Another method was the use of punched cards. This was a technology originally developed for controlling early
industrial age factories, particularly large weaving looms. The designs or patterns for the cloth would be programmed using
punched cards. This made it easy to switch to new designs. Some of the large looms became so complex that they were
essentially computers, although that terminology wasn’t used at the time.
machine code and object code
Both the front panel switch and the punched card methods involved the use of numeric codes. Each numeric code
indicated a different machine instruction. The numbers used internally are known as machine code. The numbers on some
external media, such as punched cards (or disk files) are known as object code.
assembly and assemblers
One of the early developments was a symbolic assembler. Instead of writing down a series of binary numbers, the
programmer would write down a list of machine instructions, using human-readable symbols. A special program, the
assembler, would convert these symbolic instructions into object or machine code.
Assembly languages have the advantage that they are easier to understand than raw machine code, but still give access to
all of the power of the computer (as each assembler symbol translates directly into a specific machine instruction).
version 56
page 253
Unix and Linux System Administration and Shell Programming
Assembly languages have the disadvantage that they are still very close to machine language. These can be difficult for a
human to follow and understand and time-consuming for a human to write. Also, programs written in assembly are tied to
a specific computer hardware and can’t be reused on another kind of computer.
The human readable version of assembly code is known as source code (it is the source that the assembler converts into
object code). All programs written in high level languages are also called source code.
high level languages
High level languages are designed to be easier to understand than assembly languages and allow a program to run on
multiple different kinds of computers.
The source code written in high level languages needs to be translated into object code. The two basic approaches are
compilers and interpetters. Some programming languages are available in both interpretted and compiled versions.
High level languages have usually been designed to meet the needs of some particular kind of programming. For
example, FORTRAN was originally intended for scientific programming. COBOL was originally intended for business
programming and data processing. SQL was originally intended for data base queries. C was originally intended for
systems programming. LISP was originally intended for list processing. PHP was originally intended for web scripting. Ada
was originally intended for embedded systems. BASIC and Pascal were originally intended as teaching languages.
Some high level languages were intended to be general purpose programming languages. Examples include PL/I and
Modula-2. Some languages that were originally intended for a specific purpose have turned into general purpose
programming languages, such as C and Pascal.
compilers
Compilers convert a finished program (or section of a program) into object code. This is often done in steps. Some
compilers convert high level language instructions into assembly language instructions and then an assembler is used to
create the finished object code.
Some compilers convert high level language instructions into an intermediate language. This intermediate language is
platform-independent (it doesn’t matter which actual computer hardware is eventually used). The intermediate language is
then converted into object code for a specific kind of computer. This approach makes it easier to move (or port) a compiler
from one kind of computer to another. Only the last step (or steps) need to be rewritten, while the main complier is reused.
Compiled code almost always runs faster than interpretted code. An optimizing compiler examines a high level program
and figures out ways to optimize the program so that it runs even faster.
C
A C program is considered to be strictly conforming to ANSI C if the program only uses features and libraries as they
are described in the ANSI standard (with no additional optional features or extensions).
A conforming hosted implementation accepts any strictly conforming program. This applies to a program that is
intended to run on an operating system.
A conforming freestanding implementation accepts any strictly conforming program that doesnt use any library
facilities other than those in the header files float.h, limits.h , stdarg.h , and stdef.h.. This applies to a program that is
intended to run on an embedded system or other environment with minimal operating system support (such as no file
system)..
compilers and assemblers
version 56
page 254
Unix and Linux System Administration and Shell Programming
A brief explanation of the difference between compilers and assemblers.
An assembler converts symbolic (human readable text) instructions into their corresponding machine or object
instructions. There is generally a one-to-one correspondence between assembler instructions and machine instructions (at
the macro-machine level).
A compiler converts a high level language into machine or object code. Typically there are many machine instructions
for every high level language instruction. There are some exceptions — some older languages, such as COBOL and
FORTRAN, had several instructions that translated directly into a single machine instruction, but even in those cases, most
of the useful portions of the language were translated into many machine instructions.
An example from the C programming language:
if (x==0) z=3; /* test to see if x is zero, if x is zero, then set z to 3 */
The same example in 8080 assembly language (everything after the semicolon ; is a comment to help you understand):
LXIH $E050 ; point to location of variable x (Load Double Immediate into HL
register pair)
MOVAM ; load value of x into the accumulator (MOVe to A register from Memory)
CMPA ; test the value of the accumulator (CoMPare A register with itself)
JNZ @1 ; if not zero, then skip variable assignment (Jump is Not Zero)
MVIA #3 ; load three (new value for variable z) into accumulator (MoVe Immediate
into A register the number three)
LXIH $E060 ; point to location of variable z (Load Double Immediate into HL
register pair)
MOVMA ; store three into variable z (MOVe to Memory from A register)
@1 NOP ; drop through or skip to here to continue program (No OPeration)
DS $E050 ; reserve memory for variable x (Data Storage)
DS $E060 ; reserve memory for variable z (Data Storage)
The same example in 8080 machine code (the comments after the semicolon ; wouldn’t normally be included, but are
added to help you follow along, in a real case of object code it would be just binary/hexadecimal numbers):
21 ; LXIH
50 ; data address
E0 ; data address
7E ; MOVAM
BF ; CMPA
C2 ; JNZ
0D ; code address
00 ; code address
1E ; MVIA
03 ; data
21 ; LXIH
60 ; data address
E0 ; data address
77; MOVMA
00 ; NOP
and later in memory: the data storage
$E050 ; variable x, unknown contents
$E060 ; variable y, becomes three (3)
You will notice that there is one machine instruction for each assembly instruction (some instructions are followed by
data or addresses), while there are many assembly or machine instructions for one C instruction.
linkers
version 56
page 255
Unix and Linux System Administration and Shell Programming
As programs grow in size, requiring teams of programmers, there is a need to break them up into separate files so that
different team members can work on their individual assignments without interfering with the work of others. Each file is
compiled separately and then combined later.
Linkers are programs that combine the various parts of a large program into a single object program. Linkers also bring
in support routines from libraries. These libraries contain utility and other support code that is reused over and over for lots
of different programs.
Historically, linkers also served additional purposes that are no longer necessary, such as resolving relocatable code on
early hardware (so that more than one program could run at the same time).
loaders
A loader is a program that loads programs into main memory so that they can be run. In the past, a loader would have to
be explicitely run as part of a job. In modern times the loader is hidden away in the operating system and called
automatically when needed.
interpreters
Interpreters convert each high level instruction into a series of machine instructions and then immediately run (or
execute) those instructions. In some cases, the interpreter has a library of routines and looks up the correct routine from the
library to handle each high level instruction.
Interpreters inherently run more slowly than the same software compiled. In the early days of computing this was a
serious problem. Since the mid-1980s, computers have become fast enough that interpreters run fine for most purposes.
Most of the scripting languages common on the web and servers are intereted languages. This includes JavaScript, Perl,
PHP, Python, Ruby.
Note that some modern programming languages (including Java and Python) translate the raw text into an intermediate
numeric code (usually a byte code) for a virtual machine. This method is generally faster than older traditional methods of
interpreting scripts and has the advantage of providing a pkatform-independent stored code.
editors
An editor is a program that is used to edit (or create) the source files for programming. Editors rarely have the advanced
formatting and other features of a regular word processor, but sometimes include special tools and features that are useful
for programming.
Two important editors are emacs and vi from the UNIX world. I personally use Tom Bender’s Tex-Edit Plus, which is
available in multiple different languages (Danish, English, French, German, Italian, Japanese, Spanish).
command line interface
A command line interface is an old-style computer interface where the programmer (or other person) controls the
computer by typing lines of text. The text lines are used to give instructions (or commands) to the computer. The most
famous example of a command line interface is the UNIX shell.
In addition to built-in commands, command line interfaces could be used to run programs. Additional information could
be passed to a program, such as names of files to use and various “program switches” that would modify how a program
operated.
version 56
page 256
Unix and Linux System Administration and Shell Programming
See the information on how to use the shell.
development environment
A development environment is an integrated set of programs (or sometimes one large monolithic program) that is used
to support writing computer software. Development environments typically include an editor, compiler (or compilers),
linkers, and various additional support tools. Development environments may include their own limited command line
interface specifically intended for programmers.
The term “development environment” can also be used to mean the collection of programs used for writing software,
even if they aren’t integrated with each other.
Because there are a huge number of different development environments and a complete lack of any standardization, the
methods used for actually typing in, compiling, and running a program are not covered by this book. Please refer to your
local documentation for details.
The development environment for UNIX, Linux, and Mac OS X are discussed in the chapter on shell programming.
Stanford introduction
Stanford CS Education Library This [the following section until marked as end of Stanford University items] is
document #101, Essential C, in the Stanford CS Education Library. This and other educational materials are available for
free at http://cslibrary.stanford.edu/. This article is free to be used, reproduced, excerpted, retransmitted, or sold so long as
this notice is clearly reproduced at its beginning. Copyright 1996-2003, Nick Parlante, nick.parlante@cs.stanford.edu.
The C Language
C is a professional programmer’s language. It was designed to get in one’s way as little as possible. Kernighan and
Ritchie wrote the original language definition in their book, The C Programming Language (below), as part of their
research at AT&T. Unix and C++ emerged from the same labs. For several years I used AT&T as my long distance carrier
in appreciation of all that CS research, but hearing “thank you for using AT&T” for the millionth time has used up that
good will.
Some languages are forgiving. The programmer needs only a basic sense of how things work. Errors in the code are
flagged by the compile-time or run-time system, and the programmer can muddle through and eventually fix things up to
work correctly. The C language is not like that.
The C programming model is that the programmer knows exactly what they want to do and how to use the language
constructs to achieve that goal. The language lets the expert programmer express what they want in the minimum time by
staying out of their way.
C is “simple’ in that the number of components in the language is small-- If two language features accomplish more-orless the same thing, C will include only one. C’s syntax is terse and the language does not restrict what is “allowed” -- the
programmer can pretty much do whatever they want.
C’s type system and error checks exist only at compile-time. The compiled code runs in a stripped down run-time model
with no safety checks for bad type casts, bad array indices, or bad pointers. There is no garbage collector to manage
memory. Instead the programmer mangages heap memory manually. All this makes C fast but fragile.
Analysis -- Where C Fits
Because of the above features, C is hard for beginners. A feature can work fine in one context, but crash in another. The
programmer needs to understand how the features work and use them correctly. On the other hand, the number of features
is pretty small.
version 56
page 257
Unix and Linux System Administration and Shell Programming
Like most programmers, I have had some moments of real loathing for the C language. It can be irritatingly obedient -you type something incorrectly, and it has a way of compiling fine and just doing something you don’t expect at run-time.
However, as I have become a more experienced C programmer, I have grown to appreciate C’s straight-to-the point style. I
have learned not to fall into its little traps, and I appreciate its simplicity.
Perhaps the best advice is just to be careful. Don’t type things in you don’t understand. Debugging takes too much time.
Have a mental picture (or a real drawing) of how your C code is using memory. That’s good advice in any language, but in
C it’s critical.
Perl and Java are more “portable” than C (you can run them on different computers without a recompile). Java and C++
are more structured than C. Structure is useful for large projects. C works best for small projects where performance is
important and the progammers have the time and skill to make it work in C. In any case, C is a very popular and influential
language. This is mainly because of C’s clean (if minimal) style, its lack of annoying or regrettable constructs, and the
relative ease of writing a C compiler.
Other Resources
The C Programming Language, 2nd ed., by Kernighan and Ritchie. The thin book which for years was the bible for all C
programmers. Written by the original designers of the language. The explanations are pretty short, so this book is better as a
reference than for beginners.
Stanford CS Education Library This [the following section until marked as end of Stanford University items] is
document #101, Essential C, in the Stanford CS Education Library. This and other educational materials are available for
free at http://cslibrary.stanford.edu/. This article is free to be used, reproduced, excerpted, retransmitted, or sold so long as
this notice is clearly reproduced at its beginning. Copyright 1996-2003, Nick Parlante, nick.parlante@cs.stanford.edu.
end of Stanford introduction
version 56
page 258
Unix and Linux System Administration and Shell Programming
standards and variants
chapter 96
summary
The educational goal of this chapter is to make the student aware that there are offcial versions of many programming
languages, but that in practice there are a lot of variations (and that a programmer must be able to adapt to changes and
variations).
Programming languages can meet official standards or come in variants and dialects.
standards and variants
Programming languages have traditionally been developed either by a single author or by a committee.
Typically after a new programming language is released, new features or modifications, called variants, start to pop-up.
The different versions of a programming language are called dialects. Over time, the most popular of these variants become
common place in all the major dialects.
If a programming language is popular enough, some international group or committee will create an official standard
version of a programming language. The largest of these groups are ANSI (Ameican national Standards Institute) and ISO
(International Orgnaization for Standardization).
While variants and dialects may offer very useful features, the use of the non-standard features will lock the program
into a particular development environment or compiler and often will lock the program into a specific operating system or
even hardware platform.
Use of official standaards allows for portability, which is the ability to move a program from one machine or operating
system to another.
While variants were traditionally introduced in an attempt to improve a programming language, Microsoft started the
practice of intentionally creating variants to lock developers into using Microsoft products. In some cases the Microsoft
variants offered no new features, but merely chaanged from the established standard for the sake of being different.
Microsoft lost a lawsuit with Sun Microsystems for purposely creating variants to Java in hopes of killing off Java in favor
of Microsoft languages.
version 56
page 259
Unix and Linux System Administration and Shell Programming
test bed
chapter 97
summary
This chapter looks at a test bed for writing this book.
At the time of this writing, this test bed does not yet exist. The author was asked to put together a list of test bed
requirements, emphasizing saving money as much as possible. Rather than keep this test bed as a secret, it seems that it
would be a good idea to share this information with the readers. It also makes sense to share with the reader the process of
installing each component of the test bed.
Thank you.
stages
There are five basic stages of deployment of the test bed.
1. initial main development computer This is a low cost used computer to get work started, as well as the additional
essential items to support that development.
2. initial server machines These are the first and most important machines for testing server deployment.
3. mobile computing These are the items needed for creating and testing mobile apps.
4. content creation These are the items needed for full content creation. These will be divided into additional stages.
5. additional servers These are the additional desired test platforms for server and desktop testing.
initial main development computer
The items in this section are the absolute minimums to be able to do any useful work. This is not enough to do all of the
important aspects of the job, but is enough to get started and to do very important work.
This is a low cost used computer to get work started, as well as the additional essential items to support that
development.
The initial main development computer is older technology chosen to greatly lower costs,
yet still provide a useful platform for development.
Hardware: Note that this is available only as a used computer — Apple Power Mac G5
“Cipher”:
Processor: DP DC “Quadcore“ 2.5 GHz Power PC 970MP (G5)
Cooling: Panasonic LCS
Graphics: GeForce 7800 GT with 512 MB of DDR RAM
Memory: 16GB of 533 MHz PC2-4200 DDR2 SDRAM
Hard Drive: Two of 500GB 7200 rpm drives
Wireless: AirPort Extreme with Bluetooth 2.0+EDR combo card
Optical Drives: Two of 16x SuperDrive (DVD+R DL/DVD±RW/CD-RW)
OS: Mac OS X 10.5.8 Leopard
Monitor: Apple Cinema Display (30-inch DVI)
Input:Apple extended keyboard and Mighty Mouse
I will supply a set of USB speakers and a USB 1.0 hub. If you can provide a modern USB 3.0 and FireWire 800 hub, that
would be very useful.
version 56
page 260
Unix and Linux System Administration and Shell Programming
The obvious question is why does it have to be a Macintosh, because business persons hate Macintosh.
The answer is that the Macintosh is the only computer that allows access to the UNIX command line and all of the UNIX
tools, including a wide array of programming languages and the BASH sheel, and at the same time allows access to the
mainstream content creation software. Adobe software is not available for Linux (other than viewers). Final Cut Pro, Logic
Pro, and other mainstream content creation software is only available on Macintosh.
The use of the BASH shell and UNIX compilers (and other tools) is absolutely essential for this work. .NET and other
Microsoft proprietary software is useless for building open systems. The Microsoft development software is only useful for
building on Microsoft proprietary systems.
Photoshop won’t run on Linux. The other listed content creation software won’t run on Linux. Several of the programs
listed won’t run on Widnows or Linux, only on Macintosh. GIMP is not an acceptable substitute for everyday graphics
work. GIMP is great for automated content creation on a server, but it is very poor for direct human use.
Software: Note that most of this software is only available used.
Tex-Edit Plus (new-shareware)
Fetch (new)
Apple Developer Tools (download free from Apple - you will have to apply and pay a fee to Apple)
Adobe Creative Suite 4 Design Premium, including Adobe InDesign CS4, Photoshop CS4 Extended, Illustrator CS4,
Flash CS4 Professional, Dreamweaver CS4, Fireworks CS4, and Acrobat 9 Pro with additional tools and services
(only available used)
Apple Final Cut Pro 5 (only available used)
Apple Logic Pro 8 (only available used)
Apple QuickTime Pro
Fink (free download)
Note that most of this software is only available used.
Other: These are the additional items that are essential.
A place indoors to work. The place must have a lock. I must be able to get in and out on a 24-hour a day basis, including
any holidays. And the place must be within a reasonable walking distance of 17th and Newport in Costa Mesa.
Moving on, I also need electricity.
Internet connections. There is a need for two independent internet connections that come from two unrelated sources.
Getting two different internet providers that both lease their lines from the phone company does not count. It really needs to
be two completely unrleated sources (such as DSL or U-verse from the phone company and cable modem from the cable
company. It is vital to have two different sources, because internet connections do invariably go down.
I need a chair to sit in. A folding chair is not acceptable because it causes too much back pain after several hours.
I need a stable flat surface to work on. It does not have to be a desk. It can be a table.
Starting in October I need a reliable $2 an hour (after taxes) for a 40 hour work week. I can work all overtime for free.
I understand that the minimum wage is completely unfair to business owners. I am not asking for the minimum wage.
Optional: It would be useful to have a place to obtain a shower. It would be useful to have the ability to store and cook
foods. None of this is essential and it is probably very bad business to spend the extra money, but it would increase
productivity. It would also be useful to have a pair of glasses that I can see clearly through.
initial server machines
version 56
page 261
Unix and Linux System Administration and Shell Programming
These are the first and most important machines for testing server deployment.
Note that there is the intent to build a quad boot computer that has Macintosh, FreeBSD, Linux, and Windows all on one
hard drive. So, this will keep costs down, because it is one computer rather than four. And the steps for completing this
work will be written down and provided to the public as a how-to guide. A friend, Joe, needs to be paid for creating the
quad-boot partioning.
FreeBSD computer
A computer set up with FreeBSD.
Linux Mint computer
A computer set up with Linux Mint.
OpenVMS computer
An itanium computer set up with a hobbyist licensed copy of OpenVMS. Alternatively, an old Alpha or old VAX can
also be set up with a hobbyist licensed copy of OpenVMS. Given the lower cost for a used machine, the Alpha version is
probably the wisest, but if an Alpha or VAX is used, we will still need the modern itanium version for the later step of
additional servers.
Windows computer
Need to find someone willing to work in the horrid Windows environment who can check that the software works
correctly in that environment. Microsoft’s monopolistic approach may prevent deployment of system-independent software.
mobile computing
These are the items needed for creating and testing mobile apps.
development computer
Hardware: This becomes the primary development computer (although the G5 development computer will remain in use
for testing and other purposes). This is a 15-inch MacBook Pro with Retina Display
Processor: 2.7 GHz Quad-core Intel Core 17
Memory: 16GB 1600MHz DDR3L SDRAM
Storage: Internal 768GB Flash Storage; External Promise Pegasus 12TB (6x2TB) R6 RAID System
Optical: External USB SuperDrive
Display: Internal: 15-inch Retina Display; external: Apple 30 inch Cinema HD Display
Keyboard: Backlit Keyboard (English) and User’s Guide (English)
Software: Preinstalled from Apple: Keynote, Pages, Numbers
Accessories: Apple Thunderbolt to Gigabit Ethernet Adaptor; Mini DisplayPort to VGA Adaptor; Apple Thunderbolt
cable; Thunderbolt to FireWire 800 adaptor; Apple Mighty Mouse
Software: This is additional important software.
Tex-Edit Plus (new-shareware)
Fetch (new)
Apple Developer Tools (download free from Apple - you will have to apply and pay a fee to Apple)
Adobe CS6 Master Collection: Photoshop CS6 Extended; Illustrator CS6; InDesign CS6; Acrobat X Pro; Flash
Professional CS6; Flash Builder 4.6 Premium Edition; Dreamweaver CS6; Fireworks CS6; Adobe Premiere Pro CS6;
After Effects CS6; Adobe Audition CS6; SpeedGrade CS6; Prelude CS6; Encore CS6; Bridge CS6; Media Encoder
version 56
page 262
Unix and Linux System Administration and Shell Programming
CS6
Apple Final Cut Pro X (with Apple Motion 5 and Apple Compressor 4) and Final Cut Pro 7 (only available used)
Apple Logic Studio
Apple QuickTime 7 Pro for Mac OS X
Apple QuickTime MPEG-2 Playback Component for Mac OS X
Unsanity FruitMenu
Unsanity Labels X
Unsanity MenuMaster
Unsanity WindowShades
Fink (free download)
iPhone
An iPhone 5 and development system.
Android
An Android and development system.
iPad
An iPad and development system.
BlackBerry
A BlackBerry and development system.
Windows Phone
A Windows Phone 7 and development system.
Nokia
A Nokia and development system.
content creation
These are the items needed for full content creation. These will be divided into additional stages.
The content creation computer needs to be rented or leased, because a significant upgrade is expected early 2013 and that
is when we need to purchase the long term content creation computer. In the meantime, we need to purchase an external
FireWire drive and a Thunderbolt converter. We will do our work from the FireWire drive and then be easily able to plug it
in to the new computer. Note that this main work drive is in addition to working storage drive(s) and backup drive(s).
additional servers
These are the additional desired test platforms for server and desktop testing. There is no particular order implied at this
time, other than getting to Red hat Enterprise, Ubuntu, and Solaris as soon as possible. We also want an IBM Power or Z
system (even if it is an old one) with FORTRAN, PL/I, COBOL, and C compilers, as well as IBM’s server and system
administration software.
Red Hat Enterprise Linux computer
version 56
page 263
Unix and Linux System Administration and Shell Programming
A computer set up with Red Hat Enterprise Linux.
Ubuntu Linux computer
A computer set up with Ubuntu Linux.
Solaris computer
An UltraSPARC computer set up with Solaris and Oracle Database.
Fedora Linux computer
A computer set up with Fedora Linux.
OpenSUSE Linux computer
A computer set up with OpenSUSE Linux.
CentOS Linux computer
A computer set up with CentOS Linux.
Oracle Linux computer
A computer set up with Oracle Linux and Oracle Database.
Debian Linux computer
A computer set up with Debian Linux.
IBM zEnterprise EC12 computer
An IBM zEnterprise EC12 model HA1 computer running z/OS with Enterprise PL/I, FORTRAN, Enterprise COBOL,
version 56
page 264
Unix and Linux System Administration and Shell Programming
and XL C/C++ compilers and APL interpreter. The HA1 with 101 processor units in 4 books, which include Central
Processor (CP), Internal Coupling facility (ICF), Integrated facility for Linux (IFL), Additional System Assist Processor
(SAP), System z Application Assist Processor (zAAP), and System z Intergrated Information Processor (zIIP) — the exact
mix of the 101 processors to be determined. Water cooled option. 3040 GB of memory. zBX Model 003 attachment (with
IBM POWER7 blade running PowerVM Enterprise Edition, IBM POWER7 blade running AIX, IBM BladeCenter HX5
blade running Red Hat Enterprise Linux, and IBM BladeCenter HX5 blade running SUSE Linux Enterprise Server, up to
112 blades). Trusted Key Entry (TKE) workstation. IBM Crypto Express 4S digital signature cryptography and EAL 5+
certification. IBM zAware. IBM DB2 Amalytics Accelerator. IBM SmartCloud Enterprise+ for System z. IBM System
Storage DS8000. IBM TS1140 tape drives with encryption.
The zEnterprise EC12 is available in five hardware models: H20, H43, H66, H89, HA1A1. This includes the capability to
support over 100 configurable cores.
version 56
page 265
Unix and Linux System Administration and Shell Programming
command summaries
appendix A
This chapter provides a list of UNIX (Linux) command summaries.
shell commands, tools, and utilities
CONTROL
CONTROL-c Kills current process
a
a2ps Fromat an ASCII file for printing on a PostScript printer [Linux]
alias Create a command alias [builtin command in BASH]
apropos Search Help manual pages (man -k)
apt-get Search for and install software packages (Debian/Ubuntu)
aptitude Search for and install software packages (Debian/Ubuntu)
arp Display and modify the Internet-to-Ethernet address translation tables
aspell Spell Checker
at Schedule a job
awk Find and Replace text, database sort/validate/index
b
basename Strip directory and suffix from filenames
bash GNU Bourne-Again SHell
bc Arbitrary precision calculator language
bg Send to background
break Exit from a loop [builtin command in BASH]
builtin Run a shell builtin
bzip2 Compress or decompress file(s)
c
cal Displays a calendar [alternative layout using ncal ]
case Conditionally perform a command
cat Concatenate and print (display) the contents of files
cd Change Directory
cfdisk Partition table manipulator for Linux
chgrp Change group ownership
chmod Change access permissions
chown Change file owner and group
chroot Run a command with a different root directory
chkconfig System services (runlevel)
cksum Print CRC checksum and byte counts
clear Clears the terminal screen
cmp Compare two files
comm Compare two sorted files line by line
command Run a command - ignoring shell functions [builtin command in BASH]
continue Resume the next iteration of a loop [builtin command in BASH]
cp Copy one or more files to another location
cron Daemon to execute scheduled commands
crontab Schedule a command to run at a later time
csplit Split a file into context-determined pieces
cut Divide a file into several parts
d
version 56
page 266
Unix and Linux System Administration and Shell Programming
date Display or change the date and time
dc Desk Calculator
dd Convert and copy a file, write disk headers, boot records
ddrescue Data recovery tool
declare Declare variables and give them attributes [builtin command in BASH]
defaults Set defaults [mac OS X]
df Display drives and free disk space
dict Look up a word at dict.org [Perl client script]
diff Display the differences between two files
diff3 Show differences among three files
dig DNS lookup
dir Briefly list directory contents
dircolors Color setup for ls
dirname Convert a full pathname to just a path
dirs Display list of remembered directories
dmesg Print kernel & driver messages
du Estimate file space usage
e
echo Display message on screen [builtin command in BASH]
egrep Search file(s) for lines that match an extended expression
eject Eject removable media
enable Enable and disable builtin shell commands [builtin command in BASH]
env Environment variables
ethtool Ethernet card settings
eval Evaluate several commands/arguments
exec Execute a command
exit Exit the shell
expect Automate arbitrary applications accessed over a terminal
expand Convert tabs to spaces
export Set an environment variable
expr Evaluate expressions
f
false Do nothing, unsuccessfully
fdformat Low-level format a floppy disk
fdisk Partition table manipulator for Linux
fg Send job to foreground
fgrep Search file(s) for lines that match a fixed string
file Determine file type
find Search for files that meet a desired criteria
fmt Reformat paragraph text
fold Wrap text to fit a specified width
for Loop and expand words and execute commands
format Format disks or tapes
free Display memory usage
fsck File system consistency check and repair
ftp File Transfer Protocol
function Define Function Macros
fuser Identify/kill the process that is accessing a file
g
gawk Find and Replace text within file(s)
getopts Parse positional parameters
grep Search file(s) for lines that match a given pattern
groupadd Add a user security group
groupdel Delete a group
groupmod Modify a group
version 56
page 267
Unix and Linux System Administration and Shell Programming
groups Print group names a user is in
gzip Compress or decompress named file(s)
h
hash Remember the full pathname of a name argument
head Output the first part of file(s)
help Display help for a built-in command [builtin command in BASH]
history Command History
hostname Print or set system name
i
iconv Convert the character set of a file
id Print user and group IDs
if Conditionally perform a command
ifconfig Configure a network interface
ifdown Stop a network interface
ifup Start a network interface up
import Capture an X server screen and save the image to file
init Set run level on Linux
install Copy files and set attributes
j
jobs List active jobs [builtin command in BASH]
join Join lines on a common field
jot Print numeric sequences [BSD and Mac OS X]
k
kill Stop a process from running
killall Kill processes by name
l
less Display output one screen at a time
let Perform arithmetic on shell variables [builtin command in BASH]
ln Make links between files
local Create variables [builtin command in BASH]
locate Find files
logname Print current login name
logout Exit a login shell [builtin command in BASH]
look Display lines beginning with a given string
lpc Line printer control program
lpr Off line print
lprint Print a file
lprintd Abort a print job
lprintq List the print queue
lprm Remove jobs from the print queue
ls List information about file(s)
lsof List open files
m
make Compile or recompile a group of programs
man Help manual
mkdir make directory(s)
mkfifo Make FIFOs (named pipes)
mkisofs Create an hybrid ISO9660/JOLIET/HFS filesystem
mknod Make block or character special files
mmv Mass Move and rename (files)
more Display output one screen at a time
mount Mount a file system
mtools Manipulate MS-DOS files
mtr Network diagnostics (traceroute/ping)
version 56
page 268
Unix and Linux System Administration and Shell Programming
mv Move or rename files or directories
mysql SQL database
n
nano Very simple editor
ncal> Displays a calendar [alternative layout using cal ]
netstat Networking information
nice Set the priority of a command or job
nl Number lines and write files
nohup Run a command immune to hangups
notify-send Send desktop notifications
nslookup Query Internet name servers interactively
o
open Open a file in its default application
op Operator access
p
passwd Modify a user password
paste Merge lines of files
pathchk Check file name portability
ping Test a network connection
pkill Stop processes from running
popd Restore the previous value of the current directory
pr Prepare files for printing
printcap Printer capability database
printenv Print environment variables
printf Format and print data [builtin command in BASH]
ps Process Status
pushd Save and then change the current directory
pwd Print Working Directory
q
quota Display disk usage and limits
quotacheck Scan a file system for disk usage
quotactl Set disk quotas
r
ram ram disk device
rcp Remote copy files between two machines
read Read a line from standard input [builtin command in BASH]
readarray Read from stdin into an array variable [builtin command in BASH]
readonly Mark variables/functions as readonly
reboot Reboot the system
rename Rename files
renice Alter priority of running processes
remsync Synchronize remote files via email
return Return from a shell function
rev Reverse lines of a file
rm Remove files
rmdir Remove directory(s)
route Routing commands
rpm Package manager
rsync Remote file copy (Synchronize file trees)
s
screen Multiplex terminal, run remote shells via SSH
screencapture Create screenshotts [Mac OS X]
scp Secure copy (remote file copy)
sdiff Merge two files interactively
version 56
page 269
Unix and Linux System Administration and Shell Programming
sed Stream Editor
select Accept keyboard input
seq Print numeric sequences
service Run init scripts [System V]
set Manipulate shell variables and functions
sftp Secure File Transfer Program
shift Shift positional parameters
shopt Shell Options
shred Secure delete
shutdown Shutdown or restart the system
sleep Delay for a specified time
slocate Find files
sort Sort text files
source Run commands from a file [also .]
split Split a file into fixed-size pieces
ssh Secure Shell client (remote login program)
strace Trace system calls and signals
su Substitute User identity
sudo Execute a command as another user [root]
sum Print a checksum for a file
suspend Suspend execution of this shell [builtin command in BASH]
symlink Make a new name for a file
sync Synchronize data on disk with memory
sysstat Server toolset
t
tail Output the last part of file
tar Tape ARchiver
tee Redirect output to multiple files
test Evaluate a conditional expression
time Measure program running time
times User and system times
touch Change file timestamps
top List top processes running on the system
traceroute Trace Route to Host
trap Run a command when a signal is set (Bourne)
tr Translate, squeeze, and/or delete characters
troff Format text for print
true Do nothing, successfully
tsort Topological sort
tty Print filename of terminal on stdin
type Describe a command [builtin command in BASH]
u
ulimit Limit user resources [builtin command in BASH]
umask Users file creation mask
umount Unmount a device
unalias Remove an alias [builtin command in BASH]
uname Print system information
unexpand Convert spaces to tabs
uniq Uniquify files
units Convert units from one scale to another
unset Remove variable or function names
unshar Unpack shell archive scripts
until Execute commands (until error)
unzip Uncompress .zip files
uptime Show uptime
version 56
page 270
Unix and Linux System Administration and Shell Programming
useradd Create new user account
userdel Delete a user account
usermod Modify user account
users List users currently logged in
uuencode Encode a binary file
uudecode Decode a file created by uuencode
v
v verbosely list directory contents ( ls -l -b)
vdir verbosely list directory contents ( ls -l -b)
vi ancient text editor
vim modern version of ancient text editor
vmstat Report Virtual Memory STATistics
w
w Information on what each user is doing
wait Wait for a process to complete [builtin command in BASH]
watch Execute/display a program periodically
wc Print byte, word, and line counts
whatis Single line summary of a Unix/Linux command
whereis Search the user’s $path , man pages and source files for a program
which Search the user’s $path for a program file
while Loop
who Print all usernames currently logged in
whoami Print the current user id and name ( id -un)
wget Retrieve web pages or files via HTTP, HTTPS or FTP
write Send a message to another user
x
xargs Execute utility, passing constructed argument list(s)
xdg-open Open a file or URL in the user’s preferred application
y
yes Print a string until interrupted
yum Install, upgrade, or remove the Apache web server
punctuation
. Run a command script in the current shell
! Run the last command again
# Comment or Remark
version 56
page 271
Unix and Linux System Administration and Shell Programming
history of computers
appendix B
Notes on the summary at the end of each year:
Year Year that items were introduced.
Operating Systems Operating systems introduced in that year.
Programming Languages Programming languages introduced. While only a few programming languages are
appropriate for operating system work (such as Ada, BLISS, C, FORTRAN, and PL/I, the programming languages available
with an operating system greatly influence the kinds of application programs available for an operating system.
Computers Computers and processors introduced. While a few operating systems run on a wide variety of computers
(such as UNIX and Linux), most operating systems are closely or even intimately tied to their primary computer hardware.
Speed listings in parenthesis are in operations per second (OPS), floating point operatins per second (FLOPS), or clock
speed (Hz).
Software Software programs introduced. Some major application programs that became available. Often the choice of
operating system and computer was made by the need for specific programs or kinds of programs.
Games Games introduced. It may seem strange to include games in the time line, but many of the advances in computer
hardware and software technologies first appeared in games. As one famous example, the roots of UNIX were the porting
of an early computer game to new hardware.
Technology Major technology advances.
Theory Major advances in computer science or related theory.
antiquity
The earliest calculating machine was the abacus, believed to have been invented in Babylon around 2400 BCE . The
abacus was used by many different cultures and civilizations, including the major advance known as the Chinese abacus
from the 2nd Century BCE .
200s BCE
The Antikythera mechanism, discovered in a shipwreck in 1900, is an early mechanical analog computer from between
150 BCE and 100 BCE . The Antikythera mechanism used a system of 37 gears to compute the positions of the sun and the
moon through the zodiac on the Egyptian calendar, and possibly also the fixed stars and five planets known in antiquity
(Mercury, Venus, Mars, Jupiter, and Saturn) for any time in the future or past. The system of gears added and subtracted
angular velocities to compute differentials. The Antikythera mechanism could accurately predict eclipses and could draw up
accurate astrological charts for important leaders. It is likely that the Antikythera mechanism was based on an astrological
computer created by Archimedes of Syracuse in the 3rd century BCE .
100s BCE
The Chinese developed the South Pointing Chariot in 115 BCE . This device featured a differential gear, later used in
modern times to make analog computers in the mid-20th Century.
400s
version 56
page 272
Unix and Linux System Administration and Shell Programming
The Indian grammarian Panini wrote the Ashtadhyayi in the 5th Century BCE . In this work he created 3,959 rules of
grammar for India’s Sanskrit language. This important work is the oldest surviving linguistic book and introduced the idea
of metarules, transformations, and recursions, all of which have important applications in computer science.
1400s
The Inca created digital computers using giant loom-like wooden structures that tied and untied knots in rope. The knots
were digital bits. These computers allowed the central government to keep track of the agricultural and economic details of
their far-flung empire. The Spanish conquered the Inca during fighting that stretched from 1532 to 1572. The Spanish
destroyed all but one of the Inca computers in the belief that the only way the machines could provide the detailed
information was if they were Satanic divination devices. Archaeologists have long known that the Inca used knotted strings
woven from cotton, llama wool, or alpaca wool called khipu or quipus to record accounting and census information, and
possibly calendar and astronomical data and literature. In recent years archaeologists have figured out that the one
remaining device, although in ruins, was clearly a computer.
1800s
Charles Babbage created the difference engine and
the analytical engine, often considered to be the first
modern computers. Augusta Ada King, the Countess
of Lovelace, was the first modern computer
programmer.
In the 1800s, the first computers were
programmable devices for controlling the weaving
machines in the factories of the Industrial Revolution.
Created by Charles Babbage, these early computers
used Punch cards as data storage (the cards contained
the control codes for the various patterns). These
cards were very similiar to the famous Hollerinth
cards developed later. The first computer programmer
was Lady Ada, for whom the Ada programming
language is named.
In 1822 Charles Babbage proposed a difference
engine for automated calculating. In 1933 Babbage
started work on his Analytical Engine, a mechanical computer with all of the elements of a modern computer, including
control, arithmetic, and memory, but the technology of the day couldn’t produce gears with enough precision or reliability
to make his computer possible. The Analytical Engine would have been programmed with Jacquard’s punched cards.
Babbage designed the Difference Engine No.2. Lady Ada Lovelace wrote a program for the Analytical Engine that would
have correctly calculated a sequence of Bernoulli numbers, but was never able to test her program because the machine
wasn’t built.
George Boole introduced what is now called Boolean algebra in 1854. This branch of mathematics was essential for
creating the complex circuits in modern electronic digital computers.
In 1871, James Clerk Maxwell introduced his thought experiment, Maxwell’s demon, and a possible violation of the
Second Law of Thermodynamics. This thought experiment plays an important role in quantum physics and information
theory.
In 1893, Jacques Salomon Hadamard proved his inequaltiy of determinants, which led to the discovery of Hadamard
matrices, which are the foundation for the Hadamard Gate in quantum computing.
1900s
version 56
page 273
Unix and Linux System Administration and Shell Programming
In the 1900s, researchers started experimenting with both analog and digital computers using vacuum tubes. Some of the
most successful early computers were analog computers, capable of performing advanced calculus problems rather quickly.
But the real future of computing was digital rather than analog. Building on the technology and math used for telephone
and telegraph switching networks, researchers started building the first electronic digital computers.
Leonardo Torres y Quevedo first proposed floating point arithmetic in Madrid in 1914. Konrad Zuse independently
proposed the same idea in Berlin in 1936 and built it into the hardware of his Zuse computer. George Stibitz also
indpendently proposed the idea in New Jersey in 1939.
The first modern computer was the German Zuse computer (Z3) in 1941. In 1944 Howard Aiken of Harvard University
created the Harvard Mark I and Mark II. The Mark I was primarily mechanical, while the Mark II was primarily based on
reed relays. Telephone and telegraph companies had been using reed relays for the logic circuits needed for large scale
switching networks.
The first modern electronic computer was the ENIAC in 1946, using 18,000 vacuum tubes. See below for information on
Von Neumann’s important contributions.
The first solid-state (or transistor) computer was the TRADIC, built at Bell Laboratories in 1954. The transistor had
previously been invented at Bell Labs in 1948.
1925
Theory: Wolfgang Pauli invents the Pauli matrices, which are the basis for the Pauli-X, Pauli-Y, and Pauli-Z Gates in
quantum computing.
1936
Theory: Alan Turing invents the Turing machine in his paper “On Computable Numbers, with an Application to the
Entscheidungsproblem”.
1938
Computers: Zuse Z1 (Germany, 1 OPS, first mechanical programmable binary computer, storage for a total of 64
numbers stored as 22 bit floating point numbers with 7-bit exponent, 15-bit signifocana [one implicit bit], and sign bit);
Konrad Zuse called his floating point hardware “semi-logarithmic notation” and included the ability to handle infinity and
undefined.
1941
The first modern computer was the German Zuse computer (Z3) in 1941.
Computers: Atanasoff-Berry Computer; Zuse Z3 (Germany, 20 OPS, added floating point exceptions, plus and minus
infinity, and undefined)
1942
Computers: work started on Zuse Z4
1943
In 1943 Howard Aiken of Harvard University created the Harvard Mark I. The Mark I was primarily mechanical.
Computers: Harvard Mark I (U.S.); Colossus 1 (U.K., 5 kOPS)
version 56
page 274
Unix and Linux System Administration and Shell Programming
1944
In 1944 Howard Aiken of Harvard University created the Harvard Mark II. While the Mark I was primarily mechanical,
the Mark II was primarily based on reed relays. Telephone and telegraph companies had been using reed relays for the logic
circuits needed for large scale switching networks.
Computers: Colossus 2 (U.K., single processor, 25 kOPS); Harvard Mark II and AT&T Bell Labortories’ Model V (both
relay computers) were the first American computers to include floating point hardware
1945
Plankalkül (Plan Calculus), created by Konrad Zuse for the Z3 computer in Nazi germany, may have been the first
programming language (other than assemblers). This was a surprisingly advanced programming language, with many
features that didn’t appear again until the 1980s.
Programming Languages: Plankalkül
Computers: Zuse Z4 (relay based computer, first commercial computer)
von Neumann architecture
John Louis von Neumann, mathematician (born János von Neumann 28 December 1903 in Budapest, Hungary, died 8
February 1957 in Washington, D.C.), proposed the stored program concept while professor of mathemtics (one of the
orginal six) at Princeton University’s Institute for Advanced Services, in which programs (code) are stored in the same
memory as data. The computer knows the difference between code and data by which it is attempting to access at any given
moment. When evaluating code, the binary numbers are decoded by some kind of physical logic circuits (later other
methods, such as microprogramming, were introduced), and then the instructions are run in hardware. This design is called
von Neumann architecture and has been used in almost every digital computer ever made.
Von Neumann architecture introduced flexibility to computers. Previous computers had their programming hard wired
into the computer. A particular computer could only do one task (at the time, mostly building artillery tables) and had to be
physically rewired to do any new task.
By using numeric codes, von Neumann computers could be reprogrammed for a wide variety of problems, with the
decode logic remaining the same.
As processors (especially super computers) get ever faster, the von Neumann bottleneck is starting to become an issue.
With data and code both being accessed over the same circuit lines, the processor has to wait for one while the other is
being fetched (or written). Well designed data and code caches help, but only when the requested access is already loaded
into cache. Some researchers are now experimenting with Harvard architecture to solve the von Neumann bottleneck. In
Harvard arrchitecture, named for Howard Aiken’s experimental Harvard Mark I (ASCC) calculator [computer] at Harvard
University, a second set of data and address lines along with a second set of memory are set aside for executable code,
removing part of the conflict with memory accesses for data.
Von Neumann became an American citizen in 1933 to be eligible to help on top secret work during World War II. There
is a story that Oskar Morganstern coached von Neumann and Kurt Gödel on the U.S. Constitution and American history
while driving them to their immigration interview. Morganstern asked if they had any questions, and Gödel replied that he
had no questions, but had found some logical inconsistencies in the Constitution that he wanted to ask the Immigration
officers about. Morganstern recommended that he not ask questions, but just answer them.
Von Neumann occassionally worked with Alan Turing in 1936 through 1938 when Turing was a graduate student at
Princeton. Von Neumann was exposed to the concepts of logical design and universal machine proposed in Turing’s 1934
paper “On Computable Numbers with an Application to the Entschiedungs-problem”.
Von Neumann worked with such early computers as the Harvard Mark I, ENIAC,
EDVAC, and his own IAS computer.
version 56
page 275
Unix and Linux System Administration and Shell Programming
Early research into computers involved doing the computations to create tables,
especially artillery firing tables. Von Neumann was convinced that the future of computers
involved applied mathematics to solve specific problems rather than mere table generation.
Von Neumann was the first person to use computers for mathematical physics and
economics, proving the utility of a general purpose computer.
Von Neumann proposed the concept of stored programs in the 1945 paper “First Draft of a Report on the EDVAC”.
Influenced by the idea, Maurice Wilkes of the Cambridge University Mathematical Laboratory designed and built the
EDSAC, the world’s first operational, production, stored-program computer.
The first stored computer program ran on the Manchester Mark I [computer] on June 21, 1948.
Von Neumann foresaw the advantages of parallelism in computers, but because of construction limitations of the time,
he worked on sequential systems.
Von Neumann advocated the adoption of the bit as the measurement of computer memory and solved many of the
problems regarding obtaining reliable answers from unreliable computer components.
Interestingly, von Neumann was opposed to the idea of compilers. When shown the idea for FORTRAN in 1954, von
Neumann asked “Why would you want more than machine language?”. Von Neumann had graduate students hand
assemble programs into binary code for the IAS machine. Donald Gillies, a student at Princeton, created an assembler to do
the work. Von Neumann was angry, claiming “It is a waste of a valuable scientific computing instrument to use it to do
clerical work”.
Von Neumann also did important work in set theory (including measure theory), the mathematical foundation for
quantum theory (including statistical mechanics), self-adjoint algebras of bounded linear operators on a Hilbert space closed
in weak operator topology, non-linear partial differential equations, and automata theory (later applied to cmputers). His
work in economics included his 1937 paper “A Model of General Economic Equilibrium” on a multi-sectoral growth model
and his 1944 book “Theory of Games and Economic Behavior” (co-authored with Morgenstern) on game theory and
uncertainty.
I leave the discussion of von Neumann with a couple of quotations:
“If people do not believe that mathematics is simple, it is only because they do not realize how complicated
life is.”
“Anyone who considers arithmetical methods of producing random numbers is, of course, in a state of sin.”
bare hardware
In the earliest days of electronic digital computing, everything was done on the bare hardware. Very few computers
existed and those that did exist were experimental in nature. The researchers who were making the first computers were
also the programmers and the users. They worked directly on the “bare hardware”. There was no operating system. The
experimenters wrote their programs in machine or assembly language and a running program had complete control of the
entire computer. Often programs and data were entered by hand through the use of toggle switches. Memory locations
(both data and programs) could be read by viewing a series of lights (one for each binary digit). Debugging consisted of a
combination of fixing both the software and hardware, rewriting the object code and changing the actual computer itself.
The lack of any operating system meant that only one person could use a computer at a time. Even in the research lab,
there were many researchers competing for limited computing time. The first solution was a reservation system, with
researchers signing up for specific time slots. The earliest billing systems charged for the entire computer and all of its
resources (regardless of whether used or not) and was based on outside clock time, being billed from the scheduled start to
scheduled end times.
version 56
page 276
Unix and Linux System Administration and Shell Programming
The high cost of early computers meant that it was essential that the rare computers be used as efficiently as possible.
The reservation system was not particularly efficient. If a researcher finished work early, the computer sat idle until the next
time slot. If the researcher’s time ran out, the researcher might have to pack up his or her work in an incomplete state at an
awkward moment to make room for the next researcher. Even when things were going well, a lot of the time the computer
actually sat idle while the researcher studied the results (or studied memory of a crashed program to figure out what went
wrong). Simply loading the programs and data took up some of the scheduled time.
computer operators
One solution to this problem was to have programmers prepare their work off-line on some input medium (often on
punched cards, paper tape, or magnetic tape) and then hand the work to a computer operator. The computer operator would
load up jobs in the order received (with priority overrides based on politics and other factors). Each job still ran one at a
time with complete control of the computer, but as soon as a job finished, the operator would transfer the results to some
output medium (punched tape, paper tape, magnetic tape, or printed paper) and deliver the results to the appropriate
programmer. If the program ran to completion, the result would be some end data. If the program crashed, memory would
be transferred to some output medium for the programmer to study (because some of the early business computing systems
used magnetic core memory, these became known as “core dumps”).
The concept of computer operators dominated the mainframe era and continues today in large scale operations with large
numbers of servers.
device drivers and library functions
Soon after the first successes with digital computer experiments, computers moved out of the lab and into practical use.
The first practical application of these experimental digital computers was the generation of artillery tables for the British
and American armies. Much of the early research in computers was paid for by the British and American militaries.
Business and scientific applications followed.
As computer use increased, programmers noticed that they were duplicating the same efforts.
Every programmer was writing his or her own routines for I/O, such as reading input from a magnetic tape or writing
output to a line printer. It made sense to write a common device driver for each input or putput device and then have every
programmer share the same device drivers rather than each programmer writing his or her own. Some programmers resisted
the use of common device drivers in the belief that they could write “more efficient” or faster or "“better” device drivers of
their own.
Additionally each programmer was writing his or her own routines for fairly common and repeated functionality, such as
mathematics or string functions. Again, it made sense to share the work instead of everyone repeatedly “reinventing the
wheel”. These shared functions would be organized into libraries and could be inserted into programs as needed. In the
spirit of cooperation among early researchers, these library functions were published and distributed for free, an early
example of the power of the open source approach to software development.
Computer manufacturers started to ship a standard library of device drivers and utility routines with their computers.
These libraries were often called a runtime library because programs connected up to the routines in the library at run
time (while the program was running) rather than being compiled as part of the program. The commercialization of code
libraries ended the widespread free sharing of software.
Manufacturers were pressured to add security to their I/O libraries in order to prevent tampering or loss of data.
input output control systems
The first programs directly controlled all of the computer’s resources, including input and output devices. Each individual
program had to include code to control and operate each and every input and/or output device used.
One of the first consolidations was placing common input/output (I/O) routines into a common library that could be
shared by all programmers. I/O was separated from processing.
version 56
page 277
Unix and Linux System Administration and Shell Programming
These first rudimentary operating systems were called an Input Output Control System or IOCS.
Computers remained single user devices, with main memory divided into an IOCS and a user section. The user section
consisted of program, data, and unused memory.
The user remained responsible for both set up and tear down.
Set up included loading data and program, by front panel switches, punched card, magnetic tapes, paper tapes, disk
packs, drum drives, and other early I/O and storage devices. Paper might be loaded into printers, blank cards into card
punch mahcines, and blank or formatted tape into tape drives, or other output devices readied.
Tear down would include unmounting tapes, drives, and other media.
The very expensive early computers sat idle during both set up and tear down.
This waste led to the introduction of less expensive I/O computers. While one I/O computer was being set up or torn
down, another I/O computer could be communicating a readied job with the main computer.
Some installations might have several different I/O computers connected to a single main computer to keep the
expensive main computer in use. This led to the concept of multiple I/O channels.
monitors
As computers spread from the research labs and military uses into the business world, the accountants wanted to keep
more accurate counts of time than mere wall clock time.
This led to the concept of the monitor. Routines were added to record the start and end times of work using computer
clock time. Routines were added to I/O library to keep track of which devices were used and for how long.
With the development of the Input Output Control System, these time keeping routines were centralized.
You will notice that the word monitor appears in the name of some operating systems, such as FORTRAN Monitor
System. Even decades later many programmers still refer to the operating system as the monitor.
An important motivation for the creation of a monitor was more accurate billing. The monitor could keep track of actual
use of I/O devices and record runtime rather than clock time.
For accurate time keeping the monitor had to keep track of when a program stopped running, regardless of whether it
was a normal end of the program or some kind of abnormal termination (such as aa crash).
The monitor reported the end of a program run or error conditions to a computer operator, who could load the next job
waiting, rerun a job, or take other actions. The monitor also notified the computer operator of the need to load or unload
various I/O devices (such as changing tapes, loading paper into the printer, etc.).
1946
The first modern electronic computer was the ENIAC in 1946, using 18,000 vacuum tubes.
Computers: UPenn Eniac (5 kOPS); Colossus 2 (parallel processor, 50 kOPS)
Technology: electrostatic memory
1948
Computers: IBM SSEC; Manchester SSEM
Technology: random access memory; magnetic drums; transistor
version 56
page 278
Unix and Linux System Administration and Shell Programming
Theory: Claude E. Shannon publishes “A Mathematical Theory of Communication”, including Shannon’s source coding
theorem that establishes the limits to possible data compression.
1949
Short Code created in 1949. This programming language was compiled into machine code by hand.
Programming Languages: Short Code
Computers: Manchester Mark 1
Technology: registers
1950s
Some operating systems from the 1950s include: FORTRAN Monitor System, General Motors Operating System, Input
Output System, SAGE, and SOS.
SAGE (Semi-Automatic Ground Environment), designed to monitor weapons systems, was the first real time control
system.
1951
Grace Hopper starts work on A-0.
Computers: Ferranti Mark 1 (first commercial computer); Leo I (frst business computer); UNIVAC I, Whirlwind
1952
Autocode, a symbolic assembler for the Manchester Mark I computer, was created in 1952 by Alick E. Glennie. Later
used on other computers.
A-0 (also known as AT-3), the first compiler, was created in 1952 by Grace Murray Hopper. She later created A-2,
ARITH-MATIC, MATH-MATIC, and FLOW-MATIC, as well as being one of the leaders in the development of
COBOL.Grace Hopper was working for Remington Rand at the time. Rand released the language as MATH-MATIC in
1957.
According to some sources, first work started on FORTRAN.
Programming Languages: Autocode; A-0; FORTRAN
Computers: UNIVAC 1101; IBM 701
Games: OXO (a graphic version of Tic-Tac-Toe created by A.S. Douglas on the EDSAC computer at the University of
Cambridge to demonstrate ideas on human-computer interaction)
1953
Computers: Strela
1954
The first solid-state (or transistor) computer was the TRADIC, built at Bell Laboratories in 1954. The transistor had
previously been invented at Bell Labs in 1948.
FORTRAN (FORmula TRANslator) was created in 1954 by John Backus and other researchers at International Business
Machines (now IBM). Released in 1957. FORTRAN is the oldest programming language still in common use. Identifiers
were limited to six characters. Elegant representation of mathematic expressions, as well as relatively easy input and output.
FORTRAN was based on A-0.
version 56
page 279
Unix and Linux System Administration and Shell Programming
“Often referred to as a scientific language, FORTRAN was the first high-level language, using the first
compiler ever developed. Prior to the development of FORTRAN computer programmers were required to
program in machine/assembly code, which was an extremely difficult and time consuming task, not to mention
the dreadful chore of debugging the code. The objective during its design was to create a programming
language that would be: simple to learn, suitable for a wide variety of applications, machine independent, and
would allow complex mathematical expressions to be stated similarly to regular algebraic notation. While still
being almost as efficient in execution as assembly language. Since FORTRAN was so much easier to code,
programmers were able to write programs 500% faster than before, while execution efficiency was only
reduced by 20%, this allowed them to focus more on the problem solving aspects of a problem, and less on
coding.
“FORTRAN was so innovative not only because it was the first high-level language [still in use], but also
because of its compiler, which is credited as giving rise to the branch of computer science now known as
compiler theory. Several years after its release FORTRAN had developed many different dialects, (due to
special tweaking by programmers trying to make it better suit their personal needs) making it very difficult to
transfer programs from one machine to another.” —Neal Ziring, The Language Guide, University of Michigan
“Some of the more significant features of the language are listed below:” —Neal Ziring, The Language
Guide, University of Michigan
Simple to learn - when FORTRAN was design one of the objectives was to write a language that was
easy to learn and understand.
Machine Independent - allows for easy transportation of a program from one machine to another.
More natural ways to express mathematical functions - FORTRAN permits even severely complex
mathematical functions to be expressed similarly to regular algebraic notation.
Problem orientated language
Remains close to and exploits the available hardware
Efficient execution - there is only an approximate 20% decrease in efficiency as compared to
assembly/machine code.
Ability to control storage allocation -programmers were able to easily control the allocation of storage
(although this is considered to be a dangerous practice today, it was quite important some time ago due
to limited memory.
More freedom in code layout - unlike assembly/machine language, code does not need to be laid out in
rigidly defined columns, (though it still must remain within the parameters of the FORTRAN source
code form).
“42. You can measure a programmer’s perspective by noting his attitude on the continuing vitality of
FORTRAN.” —Alan Perlis, Epigrams on Programming, ACM’s SIGPLAN Notices Volume 17, No. 9,
September 1982, pages 7-13
Programming Languages: FORTRAN
Computers: IBM 650; IBM 704 (vacuum tube computer with floating point); IBM NORC (67 kOPS)
Technology: magnetic core memory
1955
Operating Systems: GMOS (General Motors OS for IBM 701)
Computers: Harwell CADET
batch systems
Batch systems automated the early approach of having human operators load one program at a time. Instead of having a
human operator load each program, software handled the scheduling of jobs. In addition to programmers submitting their
jobs, end users could submit requests to run specific programs with specific data sets (usually stored in files or on cards).
The operating system would schedule “batches” of related jobs. Output (punched cards, magnetic tapes, printed material,
version 56
page 280
Unix and Linux System Administration and Shell Programming
etc.) would be returned to each user.
General Motors Operating System, created by General Motors Research Laboratories in early 1956 (or late 1955) for
thieir IBM 701 mainframe is generally considered to be the first batch operating system and possibly the first “real”
operating system.
The operating system would read in a program and its data, run that program to completion (including outputing data),
and then load the next program in series as long as there were additional jobs available.
Batch operating systems used a Job Control Language (JCL) to give the operating system instructions. These instructions
included designation of which punched cards were data and which were programs, indications of which compiler to use,
which centralized utilities were to be run, which I/O devices might be used, estimates of expected run time, and other
details.
This type of batch operating system was known as a single stream batch processing system.
Examples of operating systems that were primarily batch-oriented include: BKY, BOS/360, BPS/360, CAL, and Chios.
1956
Researchers at MIT begin experimenting with direct keyboard input into computers.
IPL (Information Processing Language) was created in 1956 by A. Newell, H. Simon, and J.C. Shaw. IPL was a low
level list processing language which implemented recursive programming.
Programming Languages: IPL
Operating Systems: GM-NAA I/O
Computers: IBM 305 RAMAC; MIT TX-0 (83 kOPS)
Technology: hard disk
1957
MATH-MATIC was released by the Rand Corporation in 1957. The language was derived from Grace Murray Hopper’s
A-0.
FLOW-MATIC, also called B-0, was created in 1957 by Grace Murray Hopper.
The first commercial FORTRAN program was run at Westinghouse. The first compile run produced a missing comma
diagnostic. The second attempt was a success.
BESYS OS was created by Bell telephone Laboratories for internal use.
The U.S. government created the Advanced Research Project Group (ARPA) in esponse to the Soviet Union’s launching
of Sputnik. ARPA was intended to develop key technology that was too risky for private business to develop.
Programming Languages: FLOW-MATIC; MATH-MATIC
Operating Systems: BESYS OS
Computers: IBM 608
Technology: dot matrix printer
1958
FORTRAN II in 1958 introduces subroutines, functions, loops, and a primitive For loop.
IAL (International Algebraic Logic) started as the project later renamed ALGOL 58. The theoretical definition of the
language is published. No compiler.
version 56
page 281
Unix and Linux System Administration and Shell Programming
LISP (LISt Processing) was created n 1958 and released in 1960 by John McCarthy of MIT. LISP is the second oldest
programming language still in common use. LISP was intended for writing artificial intelligence programs.
“Interest in artificial intelligence first surfaced in the mid 1950. Linguistics, psychology, and mathematics
were only some areas of application for AI. Linguists were concerned with natural language processing, while
psychologists were interested in modeling human information and retrieval. Mathematicians were more
interested in automating the theorem proving process. The common need among all of these applications was a
method to allow computers to process symbolic data in lists.
“IBM was one of the first companies interested in AI in the 1950s. At the same time, the FORTRAN project
was still going on. Because of the high cost associated with producing the first FORTRAN compiler, they
decided to include the list processing functionality into FORTRAN. The FORTRAN List Processing Language
(FLPL) was designed and implemented as an extention to FORTRAN.
“In 1958 John McCarthy took a summer position at the IBM Information Research Department. He was
hired to create a set of requirements for doing symbolic computation. The first attempt at this was
differentiation of algebraic expressions. This initial experiment produced a list of of language requirements,
most notably was recursion and conditional expressions. At the time, not even FORTRAN (the only high-level
language in existance) had these functions.
“It was at the 1956 Dartmouth Summer Research Project on Artificial Intelligence that John McCarthy first
developed the basics behind Lisp. His motivation was to develop a list processing language for Artificial
Intelligence. By 1965 the primary dialect of Lisp was created (version 1.5). By 1970 special-purpose computers
known as Lisp Machines, were designed to run Lisp programs. 1980 was the year that object-oriented concepts
were integrated into the language. By 1986, the X3J13 group formed to produce a draft for ANSI Common
Lisp standard. Finally in 1992, X3J13 group published the American National Standard for Common Lisp.” —
Neal Ziring, The Language Guide, University of Michigan
“Some of the more significant features of the language are listed below:” —Neal Ziring, The Language
Guide, University of Michigan
Atoms & Lists - Lisp uses two different types of data structures, atoms and lists.
Atoms are similar to identifiers, but can also be numeric constants
Lists can be lists of atoms, lists, or any combination of the two
Functional Programming Style - all computation is performed by applying functions to arguments.
Variable declarations are rarely used.
Uniform Representation of Data and Code - example: the list (A B C D)
a list of four elements (interpreted as data)
is the application of the function named A to the three parameters B, C, and D (interpreted as
code)
Reliance on Recursion - a strong reliance on recursion has allowed Lisp to be successful in many areas,
including Artificial Intelligence.
Garbage Collection - Lisp has built-in garbage collection, so programmers do not need to explicitly free
dynamically allocated memory.
“55. A LISP programmer knows the value of everything, but the cost of nothing.” —Alan Perlis, Epigrams
on Programming, ACM’s SIGPLAN Notices Volume 17, No. 9, September 1982, pages 7-13
Programming Languages: FORTRAN II; IAL; LISP
Operating Systems: UMES
Computers: UNIVAC II; IBM AN/FSQ-7 (400 kOPS)
Games:Tennis For Two (developed by William Higinnotham using an osciliscope and an analog computer)
Technology: integrated circuit
1959
version 56
page 282
Unix and Linux System Administration and Shell Programming
COBOL (COmmon Business Oriented Language) was created in May 1959 by the Short Range Committee of the U.S.
Department of Defense (DoD). The CODASYL committee (COnference on DAta SYstems Languages) worked from May
1959 to April 1960. Official ANSI standards included COBOL-68 (1968), COBOL-74 (1974), COBOL-85 (1985), and
COBOL-2002 (2002). COBOL 97 (1997) introduced an object oriented version of COBOL. COBOL programs are divided
into four divisions: identification, environment, data, and procedure. The divisions are further divided into sections.
Introduced the RECORD data structure. Emphasized a verbose style intended to make it easy for business managers to read
programs. Admiral Grace Hopper is recognized as the major contributor to the original COBOl language and as the
inventor of compilers.
LISP 1.5 released in 1959.
“ ‘DYNAMO is a computer program for translating mathematical models from an easy-to-understand
notation into tabulated and plotted results. … A model written in DYNAMO consists of a number of algebraic
relationships that relate the variables one to another.’ Although similar to FORTRAN, it is easier to learn and
understand. DYNAMO stands for DYNAmic MOdels. It was written by Dr. Phyllis Fox and Alexander L.
Pugh, III, and was completed in 1959. It grew out of an earlier language called SIMPLE (for Simulation of
Industrial Management Problems with Lots of Equations), written in 1958 by Richard K. Bennett.” —Language
Finger, Maureen and Mike Mansfield Library, University of Montana.
ERMA (Electronic Recording Method of Accounting), a magnetic ink and computer readable font, was created for the
Bank of America.
Programming Languages: COBOL; DYNAMO; ERMA; LISP 1.5
Operating Systems: SHARE
Computers: IBM 1401
early 1960s
The early 1960s saw the introduction of time sharing and multi-processing.
Some operating systems from the early 1960s include: Admiral, B1, B2, B3, B4, Basic Executive System, BOS/360,
Compatible Timesharing System (CTSS), EXEC I, EXEC II, Honeywell Executive System, IBM 1410/1710 OS, IBSYS,
Input Output Control System, Master Control Program, and SABRE.
The first major transaction processing system was SABRE (Semi-Automatic Business Related Environment), developed
by IBM and American Airlines.
multiprogramming
There is a huge difference in speed between I/O and running programs. In a single stream system, the processor remains
idle for much of the time as it waits for the I/O device to be ready to send or receive the next piece of data.
The obvious solution was to load up multiple programs and their data and switch back and forth between programs or
jobs.
When one job idled to wait for input or output, the operating system could automatically switch to another job that was
ready.
system calls
The first operating system to introduce system calls was University of Machester’s Atlas I Supervisor.
time sharing
The operating system could have additional reasons to rotate through jobs, including giving higher or lower priority to
version 56
page 283
Unix and Linux System Administration and Shell Programming
various jobs (and therefore a larger or smaller share of time and other resources). The Compatible Timesharing System
(CTSS), first dmonstrated in 1961, was one of the first attempts at timesharing.
While most of the CTSS operating system was written in assembly language (all previous OSes were written in assembly
for efficiency), the scheduler was written in the programming language MAD in order to allow safe and reliable
experimentation with different scheduling algorithms. About half of the command programs for CTSS were also written in
MAD.
Timesharing is a more advanced version of multiprogramming that gives many users the illusion that they each have
complete control of the computer to themselves. The scheduler stops running programs based on a slice of time, moves on
to the next program, and eventually returns back to the beginning of the list of programs. In little increments, each program
gets their work done in a manner that appears to be simultaneous to the end users.
mid 1960s
Some operating systems from the mid-1960s include: Atlas I Supervisor, DOS/360, Input Output Selector, Master
Control Program, and Multics.
The Atlas I Supervisor introduced spooling, interrupts, and virtual memory paging (16 pages) in 1962. Segmentation was
introduced on the Burroughs B5000. MIT’s Multics combined paging and segmentation.
The Compatible Timesharing System (CTSS) introduced email.
late 1960s
Some operating systems from the late-1960s include: BPS/360, CAL, CHIPPEWA, EXEC 3, EXEC 4, EXEC 8, GECOS
III, George 1, George 2, George 3, George 4, IDASYS, MASTER, Master Control Program, OS/MFT, OS/MFT-II,
OS/MVT, OS/PCP, and RCA DOS.
1960
ALGOL (ALGOrithmic Language) was released in 1960. Major releases in 1960 (ALGOL 60) and 1968 (ALGOL 68).
ALGOL is the first block-structured labguage and is considered to be the first second generation computer language. This
was the first programming language that was designed to be machine independent. ALGOL introduced such concepts as:
block structure of code (marked by BEGIN and END), scope of variables (local variables inside blocks), BNF (Backus
Naur Form) notation for defining syntax, dynamic arrays, reserved words, IF THEN ELSE, FOR, WHILE loop, the :=
symbol for assignment, SWITCH with GOTOs, and user defined data types. ALGOL became the most popular
programming language in Europe in the mid- and late-1960s.
C.A.R. Hoare invents the Quicksortin 1960.
Programming Languages: ALGOL
Operating Systems: IBSYS
Computers: DEC PDP-1; CDC 1604; UNIVAC LARC (250 kFLOPS)
1961
The Compatible Timesharing System (CTSS), first dmonstrated in 1961, was one of the first attempts at timesharing.
While most of the CTSS operating system was written in assembly language (all previous OSes were written in assembly
for efficiency), the scheduler was written in the programming language MAD in order to allow safe and reliable
experimentation with different scheduling algorithms. About half of the command programs for CTSS were also written in
MAD.
Operating Systems: CTSS, Burroughs MCP
version 56
page 284
Unix and Linux System Administration and Shell Programming
Computers: IBM 7030 Stretch (1.2 MFLOPS)
1962
APL (A Programming Language) was published in the 1962 book A Programming Language by Kenneth E. Iverson and
a subset was first released in 1964. The language APL was based on a notation that Iverson invented at Harvard University
in 1957. APL was intended for mathematical work and used its own special character set. Particularly good at matrix
manipulation. In 1957 it introduced the array. APL used a special character set and required special keyboards, displays,
and printers (or printer heads).
FORTRAN IV is released in 1962.
Simula was created by Ole-Johan Dahl and Kristen Nygaard of the Norwegian Computing Center between 1962 and
1965. A compiler became available in 1964. Simula I and Simula 67 (1967) were the first object-oriented programming
languages.
SNOBOL (StroNg Oriented symBOli Language) was created in 1962 by D.J. Farber, R.E. Griswold, and F.P. Polensky
at Bell Telephone Laboratories. Intended for processing strings, the language was the first to use associative arrays, indexed
by any type of key. Had features for pattern-matching, concatenation, and alternation. Allowed running code stored in
strings. Data types: integer, real, array, table, pattern, and user defined types.
SpaceWarI, the first interactive computer game, was created by MIT students Slug Russel, Shag Graetz, and Alan Kotok
on DEC’s PDP-1.
The first operating system to introduce system calls was University of Machester’s Atlas I Supervisor.
The Atlas I Supervisor introduced spooling, interrupts, and virtual memory paging (16 pages) in 1962. Segmentation was
introduced on the Burroughs B5000. MIT’s Multics combined paging and segmentation.
Operating Systems: GECOS
Programming Languages: APL; FORTRAN IV; Simula; SNOBOL
Computers: ATLAS, UNIVAC 1100/2200 (introduced two floating point formats, single precision and double precision;
single precision: 36 bits, 1-bit sign, 8-bit exponent, and 27-bit significand; double precision: 36 bits, 1-bit sign, 11-bit
exponent, and 60-bit significand), IBM 7094 (followed the UNIVAC, also had single and double precision numbers)
Games: Spacewar! (created by group of M.I.T. students on the DEC PDP-1)
Technology: RAND Corporation proposes the internet
1963
Work on PL/I starts in 1963.
“Data-Text was the “original and most general problem-oriented computer language for social scientists.” It
has the ability to handle very complicated data processing problems and extremely intricate statistical analyses.
It arose when FORTRAN proved inadequate for such uses. Designed by Couch and others, it was first used in
1963/64, then extensively revised in 1971. The Data-Text System was originally programmed in FAP, later in
FORTRAN, and finally its own language was developed.” —Language Finger, Maureen and Mike Mansfield
Library, University of Montana.
Sketchpad, an interactive real time computer drawing system, was created in 1963 by Ivan Sutherland as his doctoral
thesis at MIT. The system used a light pen to draw and manipulate geometric figures on a computer screen.
ASCII (American Standard Code for Information Interchange) was introduced in 1963.
Programming Languages: Data-Text
Computers: DEC PDP-6
Software: Sketchpad
version 56
page 285
Unix and Linux System Administration and Shell Programming
Technology: mouse
1964
BASIC (Beginner’s All-purpose Symbolic Instruction Code) was designed as a teaching language in 1963 by John
George Kemeny and Thomas Eugene Kurtz of Dartmouth College. BASIC was intended to make it easy to learn
programming. The first BASIC program was run at 4 a.m. May 1, 1964.
PL/I (Programming Language One) was created in 1964 at IBM’s Hursley Laboratories in the United Kingdom. PL/I
was intended to combine the scientific abilities of FORTRAN with the business capabilities of COBOL, plus additional
facilities for systems programming. Also borrows from ALGOL 60. Originally called NPL, or New Programming
Language. Introduces storage classes (automatic, static, controlled, and based), exception processing (On conditions), Select
When Otherwise conditional structure, and several variations of the DO loop. Numerous data types, including control over
precision.
RPG (Report Program Generator) was created in 1964 by IBM. Intended for creating commercial and business reports.
APL\360 implemented in 1964.
Bell Telephone Laboratories determines that it needs a new operating system to replace its 1957 BESYS OS. This project
becomes Multics.
Operating Systems: DTSS, TOPS-10
Programming Languages: APL\360; BASIC; PL/I; RPG
Computers: IBM 360; DEC PDP-8; CDC 6600 (first supercomputer, scalar processor, 3 MFLOPS)
Technology: super computing
1965
Multics The Massachusetts Institute of Technology (MIT), AT&T Bell Labs, and General Electric attempted to create an
experimental operating system called Multics for the GE-645 mainframe computer. AT&T intended to offer subscriptionbased computing services over the phone lines, an idea similar to the modern cloud approach to computing.
While the Multics project had many innovations that went on to become standard approaches for operating systems, the
project was too complex for the time.
SNOBOL 3 was released in 1965.
Attribute grammars were created in 1965 by Donald Knuth.
Operating Systems: OS/360; Multics
Programming Languages: SNOBOL 3
Technology: time-sharing; fuzzy logic; packet switching; bulletin board system (BBS); email
1966
ALGOL W was created in 1966 by Niklaus Wirth. ALGOL W included RECORDs, dynamic data structures, CASE,
passing parameters by value, and precedence of operators.
Euler was created in 1966 by Niklaus Wirth.
FORTRAN 66 was released in 1966. The language was rarely used.
ISWIM (If You See What I Mean) was described in 1966 in Peter J. Landin’s article The Next 700 Programming
Languages in the Communications of the ACM. ISWIM, the first purely functional language, influenced functional
programming languages. The first language to use lazy evaluation.
version 56
page 286
Unix and Linux System Administration and Shell Programming
LISP 2 was released in 1966.
Programming Languages: ALGOL W; Euler; FORTRAN 66; ISWIM
Computers: BESM-6
1967
Logo was created in 1967 (work started in 1966) by Seymour Papert. Intended as a programming language for children.
Started as a drawing program. Based on moving a “turtle” on the computer screen.
Simula 67 was created by Ole-Johan Dahl and Kristen Nygaard of the Norwegian Computing Center in 1967.
Introduced classes, methods, inheriteance, and objects that are instances of classes.
SNOBOL 4 (StroNg Oriented symBOli Language) was released in 1967.
CPL (Combined Programming Language) was created in 1967 at Cambridge and London Universities. Combined
ALGOL 60 and functional language. Used polymorphic testing structures. Included the ANY type, lists, and arrays.
Operating Systems: ITS; CP/CMS; WAITS
Programming Languages: CPL; Logo; Simula 67; SNOBOL 4
1968
In 1968 a group of scientists and engineers from Mitre Corporation (Bedford, Massachusetts) created Viatron Computer
company and an intelligent data terminal using an 8-bit LSI microprocessor from PMOS technology.
ALGOL 68 in 1968 introduced the =+ token to combine assignment and add, UNION, and CASTing of types. It
included the IF THEN ELIF FI structure, CASE structure, and user-defined operators.
Forth was created by Charles H. Moore in 1968. Stack based language. The name “Forth” was a reference to Moore’s
claim that he had created a fourth generation programming language.
ALTRAN, a variant of FORTRAN, was released.
ANSI version of COBOL defined.
Edsger Dijkstra wrote a letter to the Communications of the ACM claiming that the use of GOTO was harmful.
Programming Languages: ALTRAN; ALGOL 68; Forth
Computers: DEC PDP-10
Technology: microprocessor; interactive computing (including mouse, windows, hypertext, and fullscreen word
processing)
1969
In 1969 Viatron created the 2140, the first 4-bit LSI microprocessor. At the time MOS was used only for a small number
of calculators and there simply wasn’t enough worldwide manufacturing capacity to build these computers in quantity.
UNIX was created at AT&T Bell Telephone Laboratories (now an independent corporation known as Lucent
Technologies) by a group that included Ken Thompson, Dennis Ritchie, Brian Kernighan, Douglas McIlroy, Michael Lesk
and Joe Ossanna.
This group of researchers, the last of the AT&T employees involved in Multics, decided to attempt the goals of Multics
on a much more simple scale.
Unix was originally called UNICS, for Uniplexed Information and Computing Service, a play on words variation of
version 56
page 287
Unix and Linux System Administration and Shell Programming
Multics, Multiplexed Information and Computing Service (uni- means “one”, multi- means “many”).
“What we wanted to preserve was not just a good environment in which to do programming, but a system around which
a fellowship could form. We knew from experience that the essence of communal computing, as supplied by remoteaccess, time-shared machines, is not just to type programs into a terminal instead of a keypunch, but to encourage close
communication,” according to Dennis Ritchie.
UNIX was originally intended as a programmer’s workbench. The original version was a single-user system. As UNIX
spread through the academic community, more general purpose tools were added, turning UNIX into a general purpose
operating system.
While Ken Thompson still had access to the Multics environment, he wrote simulations on Multics for UNIX’s file and
paging system.
Ken Thompson and Dennis Ritchie led a team of Bell Labs
researchers (the team included Rudd Canaday) working on the
PDP-7 to develop a hierarchial file system, computer processes,
device files, a command-line interpreter, and a few small utility
programs.
This version of UNIX was written in the B language.
At the time, AT&T was prohibited from selling computers or
software, but was allowed to develop its own software and
computers for internal use. AT&T’s consent decree with the
U.S. Justice Department on monopoly charges was interpretted
as allowing AT&T to release UNIX as an open source operating
system for academic use. Ken Thompson, one of the originators
of UNIX, took UNIX to the University of California, Berkeley, where students quickly started making improvements and
modifications, leading to the world famous Berkeley Software Distribution (BSD) form of UNIX.
UNIX quickly spread throughout the academic world, as it solved the problem of keeping track of many (sometimes
dozens) of proprietary operating systems on university computers. With UNIX all of the computers from many different
manufacturers could run the same operating system and share the same programs (recompiled on each processor).
BCPL (Basic CPL) was created in 1969 in England. Intended as a simplified version of CPL, includes the control
structures For, Loop, If Then, While, Until Repeat, Repeat While, and Switch Case.
“BCPL was an early computer language. It provided for comments between slashes. The name is condensed
from “Basic CPL”; CPL was jointly designed by the universities of Cambridge and London. Officially, the “C”
stood first for “Cambridge,” then later for “Combined.” -- Unofficially it was generally accepted as standing for
Christopher Strachey, who was the main impetus behind the language.” —Language Finger, Maureen and Mike
Mansfield Library, University of Montana.
B (derived from BCPL) developed in 1969 by Ken Thompson of Bell Telephone Laboratories for use in systems
programming for UNIX. This was the parent language of C.
SmallTalk was created in 1969 at Xerox PARC by a team led by Alan Kay, Adele Goldberg, Ted Kaehler, and Scott
Wallace. Fully object oriented programming language that introduces a graphic environment with windows and a mouse.
RS-232-C standard for serial communication introduced in 1969.
Space Travel Ken Thompson ported the Space Travel computer game from Multics to a Digital Equipment Corporation
(DEC) PDP-7 he found at Bell Labs.
ARPA creates ARPAnet, the forerunner of the Internet (originally hosted by UCLA, UC Santa Barbara, University of
Utah, and Stanford Research Institute).
version 56
page 288
Unix and Linux System Administration and Shell Programming
Operating Systems: ACP; TENEX/TOPS-20; work started on Unix (initially called UNICS)
Programming Languages: B, BCPL; SmallTalk
Computers: CDC 7600 (36 MFLOPS)
Games: Space Travel (written by Jeremy Ben for Multics; when AT&T pulled out of the Multics project, J. Ben ported
the program to FORTRAN running on GECOS on the GE 635; then ported by J. Ben and Dennis Ritchie in PDP-7
assembly language; the process of porting the game to the PDP-7 computer was the beginning of Unix)
Technology: ARPANET (military/academic precursor to the Internet); RS-232; networking; laser printer (invented by
Gary Starkweather at Xerox)
early 1970s
Some operating systems from the early-1970s include: BKY, Chios, DOS/VS, Master Control Program, OS/VS1, and
UNIX.
mid 1970s
Some operating systems from the mid-1970s include: CP/M, Master Control Program.
late 1970s
Some operating systems from the late-1970s include: EMAS 2900, General Comprehensive OS, VMS (later renamed
OpenVMS), OS/MVS.
1970
Prolog (PROgramming LOGic) was created in 1972 in France by Alan Colmerauer with Philippe Roussel. Introduces
Logic Programming.
Pascal (named for French religious fanatic and mathematician Blaise Pascal) was created in 1970 by Niklaus Wirth on a
CDC 6000-series computer. Work started in 1968. Pascl was intended as a teaching language to replace BASIC, but
quickly developed into a general purpose programming language. Programs compiled to a platform-independent
intermediate P-code. The compiler for pascal was written in Pascal, an influential first in language design.
Forth used to write the program to control the Kitt Peaks telescope.
BLISS was a systems programming language developed by W.A. Wulf, D.B. Russell, and A.N. Habermann at Carnegie
Mellon University in 1970. BLISS was a very popular systems programming language until the rise of C. The original
compiler was noted for its optimizing of code. Most of the utilities for DEC’s VMS operating system were written in
BLISS-32. BLISS was a typeless language based on expressions rather than statements. Expressions produced values, and
possibly caused other actions, such as modification of storage, transfer of control, or looping. BLISS had powerful macro
facilities, conditional execution of statements, subroutines, built-in string functions, arrays, and some automatic data
conversions. BLISS lacked I/O instructions on the assumption that systems I/O would actually be built in the language.
UNIX In 1970 Dennis Ritchie and Ken Thompson traded the promise to add text processing capabilities to UNIX for the
use of a Digital Equipment Corporation (DEC) PDP-11/20. The initial version of UNIX, a text editor, and a text formatting
program called roff were all written in PDP-11/20 assembly language.
UNIX was renamed from the original name UNICS. The author of the name change is lost to history.
ROFF Soon afterwards roff evolved into troff with full typesetting capability. The UNIX Programmer’s Manual was
published on November 3, 1971. The first commercial UNIX system was installed in early 1962 at the New York Telephone
Co. Systems Development Center under the direction of Dan Gielan. Neil Groundwater build an Operational Support
System in PDP-11/20 assembly language.
Quantum Computing Stephen Wiesner invents conjugate coding.
version 56
page 289
Unix and Linux System Administration and Shell Programming
Operating Systems: Unix; RT-11; RSTS-11
Programming Languages: BLISS; Pascal; Prolog
Computers: Datapoint 2200; DEC PDP-11
Software: ROFF
Technology: dynamic RAM; flight data processor
1971
Manufacturers saw the benefit of MOS, starting with Intel’s 1971 release of the 4-bit 4004 as the first commercially
available microprocessor.
UNIX shell This first UNIX command shell, called the Thompson shell and abbreviated sh, was written by Ken
Thompson at AT&T’s Bell Labs, was much more simple than the famous UNIX shells that came along later. The
Thompson shell was distributed with Versions 1 through 6 of UNIX, from 1971 to 1975.
The first UNIX book, UNIX Programmer’s Manual, by Ken Thompson and Dennis Ritchie, was published in 1971 and
included 60 commands.
Computers: Intel 4004 (four-bit microprocessor)
Software: UNIX shell
Games: Computer Space (first commercial vidoe game)
Technology: floppy disk; first electronic calculator (T1)
1972
In 1972 Rockwell released the PPS-4 microprocessor, Fairchild released the PPS-25 microprocessor, and Intel released
the 8-bit 8008 microprocessor. All used PMOS.
C was developed from 1969-1972 by Dennis Ritchie and Ken Thompson of Bell Telephone Laboratories for use in
systems programming for UNIX. C was based on the earlier programming language B.
UNIX In 1972 work started on converting UNIX to C. The UNIX kernel was originally written in assembly language, but
by 1973 it had been almost completely converted to the C language. At the time it was common belief that operating
systems must be written in assembly in order to perform at reasonable speeds. This made Unix the world’s first portable
operating system, capable of being easily ported (moved) to any hardware. This was a major advantage for Unix and led to
its widespread use in the multi-platform environments of colleges and universities. Writing UNIX in C allowed for easy
portability to new hardware, which in turn led to the UNIX operating system being used on a wide variety of computers.
In 1972, the pipe concept was invented.
Pong, the first arcade video game, was introduced by Nolan Bushnell in 1972. His company was called Atari.
Operating Systems: VM/CMS; UNIX rewritten in C
Programming Languages: C
Computers: Intel 8008 (microprocessor); Rockwell PPS-4 (microprocessor); Fairchild PPS-25 (microprocessor)
Games: Pong; Magnavox Odysssey (first home video game console)
Technology: game console (Magnavox Odyssey); first scientific calculator (HP); first 32-bit minicomputer; first arcade
video game
1973
In 1973 National released the IMP microprocessor using PMOS.
In 1973 Intel released the faster NMOS 8080 8-bit microprocessor, the first in a long series of microprocessors that led to
the current Pentium.
version 56
page 290
Unix and Linux System Administration and Shell Programming
Quantum Computing Alexander Holevo publishes a paper showing that n quibts cannot carry more than n classical bits
of information (known as “Holevo’s theorem”).
Charles H. Bennett shows that computation can be done reversibly.
ML (Meta Language) was created in 1973 by R. Milner of the University of Edinburgh. Functional language
implemented in LISP.
Actor is a mathematical model for concurrent computation first published bby Hewitt in 1973.
“Actor is an object-oriented programming language. It was developed by the Whitewater Group in Evanston,
Ill.” —Language Finger, Maureen and Mike Mansfield Library, University of Montana.
UNIX The UNIX kernal was rewritten in the C programming language.
In 1973, UNIX was used at 16 sites worldwide.
ARPA creates Transmission Control Protocol/Internet Protocol (TCP/IP) to network together computers for ARPAnet.
Programming Languages: ML
Computers: National IMP (microprocessor)
Software: Actor
Technology: TCP/IP; ethernet
1974
In 1974 Motorola released the 6800, which included two accumulators, index registers, and memory-mapped I/O.
Monolithic Memories introduced bit-slice microprocessing.
SQL (Standard Query Language) was designed by Donald D. Chamberlin and Raymond F. Boyce of IBM in 1974.
AWK (first letters of the three inventors) was designed by Aho, Weinberger, and Kerninghan in 1974. Word processing
language based on regular expressions.
Alphard (named for the brightest star in Hydra) was designed by William Wulf, Mary Shaw, and Ralph London of
Carnegie-Mellon University in 1974. A Pascal-like language intended for data abstraction and verification. Make use of the
“form”, which combined a specification and an implementation, to give the programmer control over the impolementation
of abstract data types.
“Alphard is a computer language designed to support the abstraction and verification techniques required by
modern programming methodology. Alphard’s constructs allow a programmer to isolate an abstraction,
specifying its behavior publicly while localizing knowledge about its implementation. It originated from studies
at both Carnegie-Mellon University and the Information Sciences Institute.” —Language Finger, Maureen and
Mike Mansfield Library, University of Montana.
“CLU began to be developed in 1974; a second version was designed in 1977. It consists of a group of
modules. One of the primary goals in its development was to provide clusters which permit user-defined types
to be treated similarly to built-in types.” —Language Finger, Maureen and Mike Mansfield Library, University
of Montana.
BSD Work on UNIX at University of California, Berkeley, the initial version of BSD UNIX.
Operating Systems: MVS, BSD UNIX
Programming Languages: Alphard; AWK; CLU; SQL
Computers: Intel 8080 (microprocessor); Motorola 6800 (microprocessor); CDC STAR-100 (100 MFLOPS)
Technology: Telenet (first commercial version of ARPANET)
1975
version 56
page 291
Unix and Linux System Administration and Shell Programming
In 1975 Texas Instruments introduced a 4-bit slice microprocessor and Fairchild introduced the F-8 microprocessor.
Scheme, based on LISP, was created by Guy Lewis Steele Jr. and Gerald Jay Sussman at MIT in 1975.
Tiny BASIC created by Dr. Wong in 1975 runs on Intel 8080 and Zilog Z80 computers.
RATFOR (RATional FORtran) created by Brian Kernigan in 1975. Used as a precompiler for FORTRAN. RATFOR
allows C-like control structures in FORTRAN.
TENEX shell (also known as the TOPS shell) was created by Ken Greer by September 1975 and merged into the C shell
in December 1983. Ken Greer based the shell on the TENEX operating system (hence, the “t” in tsch ).
UNIX Fifth Edition licensed to universities for free.
Quantum Computing R.P. Poplavskii publishes “Thermodynamical models of information processing” (in Russian)
showng the comnputational infeasibility of simulating quantum systems on clasical computers because of the superposition
problem.
Programming Languages: RATFOR; Scheme; TENEX shell; Tiny BASIC
Computers: Altair 880 (first personal computer); Fairchild F-8 (microprocessor); MOS Technology 6502
(microprocessor); Burroughs ILLIAC IV (150 MFLOPS)
Technology: single board computer; laser printer (commercial release by IBM)
1976
Design System Language, a forerunner of PostScript, is created in 1976. The Forth-like language handles three
dimensional databases.
SASL (Saint Andrews Static Language) is created by D. Turner in 1976. Intended for teaching functional programming.
Based on ISWIM. Unlimited data structures.
CP/M, an operating system for microcomputers, was created by Gary Kildall in 1976.
PWB shell The PWB shell or Mashey shell, abbreviated sh, was a variation of the Thompson shell that had been
augmented by John Mashey and others at Bell Labs. The Mashey shell was distributed with the Programmer’s Workbench
UNIX in 1976.
Operating Systems: CP/M
Programming Languages: Design System language; SASIL
Computers: Zilog Z-80 (microprocessor); Cray 1 (250 MFLOPS); Apple I
Software: PWB shell
Technology: inkjet printer; Alan Kay’s Xerox NoteTaker developed at Xerox PARC
1977
Icon, based on SNOBOL, was created in 1977 by faculty, staff, and students at the University of Arizona under the
direction of Ralph E. Griswold. Icon uses some programming structures similar to pascal and C. Structured types include
list, set, and table (dictionary).
OPS5 was created by Charles Forgy in 1977.
FP was presented by John Backus in his 1977 Turing Award lecture Can Programming be Liberated From the von
Neumann Style? A Functional Style and its Algebra of Programs.
Modula (MODUlar LAnguage) was created by Niklaus Wirth, who started work in 1977. Modula-2 was released in
1979.
version 56
page 292
Unix and Linux System Administration and Shell Programming
Bourne shell The Bourne shell, created by Stephen Bourne at AT&T’s Bell Labs as a scripting language, was released in
1977 as the default shell in the Version 7 Unix release distributed to colleges and universities. It gained popularity with the
publication of The UNIX Programming Environment by Brian W. Kernighan and Rob Pike. The book was the first
commercially published tutorial on shell programming.
BSD UNIX The University of California, Berkeley, released the Berkeley Software Distribution (BSD) version of UNIX,
based on the 6th edition of AT&T UNIX.
Programming Languages: Bourne shell; FP; Icon; Modula; OPS5
Computers: DEC VAX-11; Apple II; TRS-80; Commodore PET; Cray 1A
Games: Atari 2600 (first popular home video game consle)
1978
CSP was created in 1978 by C.A.R. Hoare.
“C.A.R. Hoare wrote a paper in 1978 about parallel computing in which he included a fragment of a
language. Later, this fragment came to be known as CSP. In it, process specifications lead to process creation
and coordination. The name stands for Communicating Sequential Processes. Later, the separate computer
language Occam was based on CSP.” —Language Finger, Maureen and Mike Mansfield Library, University of
Montana.
C Shell The C shell, abbreviated csh , was created by Bill Joy, a graduate student at the University of California,
Berkeley. With additional work by Michael Ubell, Eric Allman, Mike O’Brien, and Jim Kulp, it was released in the 2BSD
release of BSD Unix in 1978.
It became common to use the C shell for everyday interactive computing and to use the Bourne shell for script
programming.
UNIX The System V flavor of UNIX started.
Operating Systems: Apple DOS 3.1; VMS (later renamed OpenVMS)
Programming Languages: C shell; CSP
Computers: Intel 8086 (microprocessor)
Games: Space Invaders (arcade game using raster graphics; so popular in Japan that the government has to quadruple its
supply of Yen)
Technology: LaserDisc
1979
Modula-2 was released in 1979. Created by Niklaus Wirth, who started work in 1977.
VisiCalc (VISIble CALculator) was created for the Apple II personal computer in 1979 by Harvard MBA candidate
Daniel Bricklin and programmer Robert Frankston.
REXX (REstructured eXtended eXecutor) was designed by Michael Cowlishaw of IBM UK Laboratories. REXX was
both an interpretted procedural language and a macro language. As a maco language, REXX can be used in application
software.
Work started on C with Classes, the language that eventually became C++.
Programming Languages: C with Classes; Modula-2; REXX
Computers: Motorola MC68000 (microprocessor); Intel 8088 (microprocessor)
Software: VisiCalc
Games: Lunar Lander (arcade video game, first to use vector graphics); Asteroids (vector arcade game); Galaxian (raster
version 56
page 293
Unix and Linux System Administration and Shell Programming
arcade game, color screen); first Multi-USer Dungeon (MUD, written by Roy Trubshaw , a student at Essex University,
forrunner of modern massively multiplayer games); Warrior (first head-to-head arcade fighting game)
Technology: first spreadsheet (VisiCalc); object oriented programming; compact disk; Usenet discussion groups
1980s
Some operating systems from the 1980s include: AmigaOS, DOS/VSE, HP-UX, Macintosh, MS-DOS, and ULTRIX.
The 1980s saw the commercial release of the graphic user interface, most famously the Apple Macintosh, Commodore
Amiga, and Atari ST, followed by MIT’s X Window System (X11) and Microsoft’s Windows.
1980
dBASE II was created in 1980 by Wayne Ratliff at the Jet Propulsion Laboratories in Pasadena, California. The original
version of the language was called Vulcan. Note that the first version of dBASE was called dBASE II.
SmallTalk-80 released.
Quantum Computing Yuri Manin proposes the idea of quantum computing.
Operating Systems: OS-9
Computers: Commodore VIC-20; ZX80; Apple III
Software: dBASE II
Games: Battlezone (vector arcade video game, dual joystick controller and periscope-like viewer); Berzerk (raster arcade
video game, used primative speech synthesis); Centipede (raster arcade video game, used trackball controller); Missile
Command (raster arcade video game, used trackball controller); Defender (raster arcade video game); Pac-Man (raster
arcade video game); Phoenix (raster arcade video game, use of musical score); Rally-X (raster arcade video game, first
game to have a bonus round); Star Castle (vector arcade video game, color provided by transparent plastic screen overlay);
Tempest (vector arcade video game, first color vector game); Wizard of Wor (raster arcade video game)
1981
Relational Language was created in 1981 by Clark and Gregory.
Quantum Computing Tommaso Toffoli introduces the reversible Toffoli gate. TOgether with the NOT and XOR gates
providing a universal set for classical computation on a quantum computer.
Operating Systems: MS-DOS; Pilot
Programming Languages: Relational Language
Computers: 8010 Star; ZX81; IBM PC; Osborne 1 (first portable computer); Xerox Star; MIPS I (microprocessor); CDC
Cyber 205 (400 MFLOPS)
Games: Donkey Kong (raster arcade video game); Frogger (raster arcade video game); Scramble (raster arcade video
game, horizontal scrolling); Galaga (raster arcade video game); Ms. Pac-Man (raster arcade video game); Qix (raster arcade
video game); Gorf (raster arcade video game, synthesized speech); Zork (first adventure game)
Technology: portable PC; ISA bus; CGA video card
1982
ANSI C The American National Standards Institute (ANSI) formed a technical subcommittee, X3J11, to create a
standard for the C language and its run-time libraries.
InterPress, the forerunner of PostScript, was created in 1982 by John Warnock and Martin Newell at Xerox PARC.
Operating Systems: SunOS
Programming Languages: ANSI C; InterPress
version 56
page 294
Unix and Linux System Administration and Shell Programming
Computers: Cray X-MP; BBC Micro; Commodore C64 (first home computer with a dedicated sound chip); Compaq
Portable; ZX Spectrum; Atari 5200; Intel 80286 (microprocessor)
Games: BurgerTime (raster arcade video game); Dig Dug (raster arcade video game); Donkey Kong Junior (raster arcade
video game); Joust (raster arcade video game); Moon Patrol (raster arcade video game, first game with parallax scrolling);
Pole Position (raster arcade video game); Q*bert (raster arcade video game); Robotron 2084 (raster arcade video game, dual
joystick); Time Pilot (raster arcade video game); Tron (raster arcade video game); Xevious (raster arcade video game, first
game promoted with a TV commercial); Zaxxon (raster arcade video game, first game to use axonometric projection)
Technology: MIDI; RISC; IBM PC compatibles
Theory: Quantum cryptography proposed
1983
Ada was first released in 1983 (ADA 83), with major releases in 1995 (ADA 95) and 2005 (ADA 2005). Ada was
created by the U.S. Department of Defense (DoD), originally intended for embedded systems and later intended for all
military computing purposes. Ada is named for Augusta Ada King, the Countess of Lovelace, the first computer
programmer inmodern times.
Concurrent Prolog was created in 1983 by Shapiro.
Parlog was created in 1983 by Clark and Gregory.
C++ was developed in 1983 by Bjarne Stroustrup at Bell Telephone Laboratories to extend C for object oriented
programming.
Turbo Pascal, a popular Pascal compiler, was released.
The University of California at Berkeley released a version of UNIX that included TCP/IP.
tcsh The improved C shell, abbreviated tcsh , created by Ken Greer, was merged into the C shell in December 1983. Ken
Greer based the shell on the TENEX operating system (hence, the “t” in tsch ). Mike Ellis, Paul Placeway, and Christos
Zoulas made major contributions to tcsh .
Korn shell The Korn shell, abbreviated ksh , was created by David Korn at AT&T’s Bell Labs and announced at
USENIX on July 14, 1983. Mike Veach and Pat Sullivan were also early contributors. The Korn shell added C shell
features to the Bourne shell.
In 1983, Richard Stallman founded the GNU project, which eventually provided the tools to go with Linux for a
complete operating system. The GNU project intended to create a complete open source UNIX-like operating system, with
GNU Hurd as the kernel. Richard Stallman created the GNU General Public License (GPL).
Operating Systems: Lisa OS
Programming Languages: Ada; C++; Concurrent Prolog; Korn shell; Parlog; tcsh; Turbo Pascal
Computers: Apple IIe; Lisa; IBM XT; IBM PC Jr; ARM (microprocessor); Cray X-MP/4 (941 MFLOPS)
Games: Dragon’s Lair (raster arcade video game, first video game to use laserdisc video; note that the gambling device
Quarterhorse used the laserdisc first); Elevator Action (raster arcade video game); Gyruss (raster arcade video game, used
the musical score Toccata and Fugue in D minor by Bach); Mappy (raster arcade video game, side scrolling); Mario Bros.
(raster arcade video game); Spy Hunter (raster arcade video game, musical score Peter Gunn); Star Wars (vector arcade
video game, digitized samples from the movie of actor’s voices); Tapper (raster arcade video game); Lode Runner (Apple
][E); Journey (arcade video game includes tape of the song Separate Ways leading to licensed music in video games)
Technology: math coprocessor; PC harddisk
1984
Objective C, an extension of C inspired by SmallTalk, was created in 1984 by Brad Cox. Used to write NextStep, the
operating system of the Next computer.
version 56
page 295
Unix and Linux System Administration and Shell Programming
Standard ML, based on ML, was created in 1984 by R. Milner of the University of Edinburgh.
PostScript was created in 1984 by John Warnock and Chuck Geschke at Adobe.
When the AT&T broke up in 1984 into “Baby Bells” (the regional companies operating local phone service) and the
central company (which had the long distance business and Bell Labs), the U.S. government allowed them to start selling
computers and computer software. UNIX broke into the System V (sys-five) and Berkeley Software Distribution (BSD)
versions. System V was the for pay version and BSD was the free open source version.
AT&T gave academia a specific deadline to stop using “encumbered code” (that is, any of AT&T’s source code
anywhere in their versions of UNIX). This led to the development of free open source projects such as FreeBSD, NetBSD,
and OpenBSD, as well as commercial operating systems based on the BSD code.
Meanwhile, AT&T developed its own version of UNIX, called System V. Although AT&T eventually sold off UNIX,
this also spawned a group of commercial operating systems known as Sys V UNIXes.
UNIX quickly swept through the commercial world, pushing aside almost all proprietary mainframe operating systems.
Only IBM’s MVS and DEC’s OpenVMS survived the UNIX onslaught.
Some of the famous official UNIX versions include Solaris, HP-UX, Sequent, AIX, and Darwin. Darwin is the UNIX
kernel for Apple’s OS X, AppleTV, and iOS (used in the iPhone, iPad, and iPod).
The BSD variant started at the University of California, Berkeley, and includes FreeBSD, NetBSD, OpenBSD, and
DragonFly BSD.
Most modern versions of UNIX combine ideas and elements from both Sys-V and BSD.
Other UNIX-like operating systems include MINIX and Linux.
In 1984, Jim Gettys of Project Athena and Bob Scheifler of the MIT Laboratory for Computer Science collaborated on a
platform-independent graphcis system to link together heterogeneous multi-vendor systems. Project Athena was a joint
project between Digital Equipment Corporation (DEC), IBM, and MIT. X version 1 was released in May 1984.
Operating Systems: GNU project started; MacOS 1, X Window System (X11)
Programming Languages: Objective C; PostScript; Standard ML
Computers: Apple Macintosh; IBM AT; Apple IIc; MIPS R2000 (microprocessor); M-13 (U.S.S.R., 2.4 GFLOPS);
Cray XMP-22
Software: Apple MacWrite; Apple MacPaint
Games: 1942 (raster arcade video game); Paperboy (raster arcade video game, unusual controllers, high resolution
display); Punch-Out (raster arcade video game, digitized voice, dual monitors)
Technology: WYSIWYG word processing; LaserJet printer; DNS (Domain Name Server); IDE interface
1985
Paradox was created in 1985 as a competitor to the dBASE family of relational data base languages.
PageMaker was created for the Apple Macintosh in 1985 by Aldus.
The Intel 80386 was the first Intel x86 microprocessor with a 32-bit instruction set and an MMU with paging.
Quantum Computing David Deutsch suggests the first blueprint for the first universal quantum computer, using
quantum gates. The universal quantum computer can simulate any other quantum computer with at most a polynomial
slowdown.
Operating Systems: GEM; AmigaOS; AtariOS; WIndows 1.0; Mac OS 2
Programming Languages: Paradox
Computers: Cray-2/8 (3.9 GFLOPS); Atari ST; Commodore Amiga; Apple Macintosh XL; Intel 80386
version 56
page 296
Unix and Linux System Administration and Shell Programming
(microprocessor); Sun SPARC (microprocessor)
Software: Apple Macintosh Office; Aldus PageMaker (Macintosh only)
Games: Super Mario Bros.; Tetris (puzzle game; invented by Russian mathematician Alexey Pajitnov to test equipment
that was going to be used for artificial intelligence and speech recognition research); Excitebike (first game with battery
backup for saving and loading game play)
Technology: desktop publishing; EGA video card; CD-ROM; expanded memory (for IBM-PC compatibles; Apple
LaserWriter
1986
Eiffel (named for Gustave Eiffel, designer of the Eiffel Tower) was released in 1986 by Bertrand Meyer. Work started on
September 14, 1985.
“Eiffel is a computer language in the public domain. Its evolution is controlled by Nonprofit International
Consortium for Eiffel (NICE), but it is open to any interested party. It is intended to treat software construction
as a serious engineering enterprise, and therefore is named for the French architect, Gustave Eiffel. It aims to
help specify, design, implement, and change quality software.” —Language Finger, Maureen and Mike
Mansfield Library, University of Montana.
GAP (Groups, Algorithms, and Programming) was developed in 1986 by Johannes Meier, Werner Nickel, Alice
Niemeter, Martin Schönert, and others. Intended to program mathematical algorithms.
CLP(R) was developed in 1986.
In 1986, Maurice J. Bach of AT&T Bell Labs published The Design of the UNIX Operating System, which described the
System V Release 2 kernel, as well as some new features from release 3 and BSD.
Operating Systems: Mach; AIX; GS-OS; HP-UX; Mac OS 3
Programming Languages: CLP(R); Eiffel; GAP
Computers: Apple IIGS; Apple Macintosh Plus; Amstrad 1512; ARM2 (microprocessor); Cray XMP-48
Software: Apple Macintosh Programmer’s Workshop
Games: Metroid (one of first games to have password to save game proogress; first female protagonist in video games,
non-linear game play; RPG)
Technology: SCSI
1987
CAML (Categorical Abstract Machine Language) was created by Suarez, Weiss, and Maury in 1987.
Perl (Practical Extracting and Report Language) was created by Larry Wall in 1987. Intended to replace the Unix shell,
Sed, and Awk. Used in CGI scripts.
HyperCard was created by William Atkinson in 1987. HyperTalk was the scripting language built into HyperCard.
Thomas and John Knoll created the program Display, which eventually became PhotoShop. The program ran on the
Apple Macintosh.
Adobe released the first version of Illustrator, running on the Apple Macintosh.
Minix Andrew S. Tanenbaum released MINIX, a simplified version of UNIX intended for academic instruction.
Operating Systems: IRIX; Minix; OS/2; Windows 2.0; MDOS (Myarc Disk Operating System); Mac OS 4; Mac OS 5
Programming Languages: CAML; HyperTalk; Perl
Computers: Apple Macintosh II; Apple Macintosh SE; Acorn Archimedes; Connection Machine (first massive parallel
computer); IBM PS/2; Commodore Amiga 500; Nintendo Entertainment System
Software: Apple Hypercard; Apple MultiFinder; Adobe Illustrator (Macintosh only; later ported to NeXT, Silicon
version 56
page 297
Unix and Linux System Administration and Shell Programming
Graphics IRIX, Sun Solaris, and Microsoft Windows); Display (which eventually became ImagePro and then Photoshop;
Macintosh only); QuarkXPress (Macintosh only)
Games: Final Fantasy (fantasy RPG); The Legend of Zelda (first free form adventure game); Gran Turismo (auto
racing); Mike Tyson’s Punch Out (boxing sports game)
Technology: massive parallel computing; VGA video card; sound card for IBM-PC compatibles; Apple Desktop Bus
(ADB)
1988
CLOS, an object oriented version of LISP, was developed in 1988.
Mathematica was developed by Stephen Wolfram.
Oberon was created in 1986 by Niklaus Wirth.
Refined version of Design becomes ImagePro for the Apple Macintosh.
Operating Systems: OS/400; Mac OS 6
Programming Languages: CLOS; Mathematica; Oberon
Computers: Cray Y-MP; Apple IIc Plus
Software: ImagePro (which eventually became Photoshop; Macintosh only)
Games: Super Mario Bros. 2; Super Mario Bros 3 (Japanese release); Contra (side-scrolling shooter); Joh Madden
Football (football sports game)
Technology: optical chip; EISA bus
1989
ANSI C The American National Standards Institute (ANSI) completed the official ANSI C, called the American
National Standard X3.159-1989.
HTML was developed in 1989.
Miranda (named for a character by Shakespeare) was created in 1989 by D. Turner. Based on SASL and ML. Lazy
evaluation and embedded pattern matching.
Standard Interchange Languagewas developed in 1989.
BASH, which stands for Bourne Again SHell, was created by Brian Fox for the Free Software Foundation and first
released on June 7, 1989. bash combined features from the Bourne shell, the C shell, and the Korn shell. bash is now the
primary shell in both Linux and Mac OS X.
Operating Systems: NeXtStep; RISC OS; SCO UNIX
Programming Languages: ANSI C; BASH; HTML; Miranda; Standard Interchage Language
Computers: Intel 80486 (microprocessor); ETA10-G/8 (10.3 GFLOPS)
Software: Microsoft Office; first (pre-Adobe) version of Photoshop (Macintosh only)
Games: Sim; SimCity; Warlords (four-player shooter)
Technology: ATA interface, World Wide Web
1990s
Some operating systems from the 1990s include: BeOS, BSDi, FreeBSD, NeXT, OS/2, Windows 95, Windows 98, and
Windows NT.
1990
version 56
page 298
Unix and Linux System Administration and Shell Programming
Haskell was developed in 1990.
Tim Berners-Lee of the European CERN laboratory created the World Wide Web on a NeXT computer.
Z shell The Z shell, abbreviated zsh , was written by Paul Flastad in 1990 when he was a student at Princton University.
In February of 1990, Adobe released the first version of the program PhotoShop (for the Apple Macintosh).
Version 2.0 of Photoshop, code named Fast Eddy, ships in the fall of 1990. Version 1.07 toolbar to right and
original icons designed by John Knoll below.
Operating Systems: BeOS; OSF/1
Programming Languages: Haskell; Z shell
Computers: NEC SX-3/44R (Japan, 23.2 GFLOPS); Cray XMS; Cray Y-MP 8/8-64 (first Cray supercomputer to use
UNIX); Apple Macintosh Classic; Neo Geo; Super Nintendo Entertainment System
Software: Adobe Photoshop (Macintosh only)
Games: Final Fantasy III released in Japan (fantasy RPG); Super Mario Bros 3 (Japanese release); Wing Commander
(space combat game)
Technology: SVGA video card; VESA driver
1991
Python (named for Monty Python Flying Circus) was created in 1991 by Guido van Rossum. A scripting language with
dynamic types intended as a replacement for Perl.
Pov-Ray (Persistence of Vision) was created in 1991 by D.B.A. Collins and others. A language for describing 3D
images.
Visual BASIC, a popular BASIC compiler, was released in 1991.
Linux operating system was first released on September 17, 1991, by Finnish student Linus Torvalds. With the
permission of Andrew S. Tanenbaum, Linus Torvalds started work with MINIX. There is no MINIX source code left in
Linux. Linux Torvalds wrote Linux using the GNU C compiler running on MINIX on an Intel 80386 processor.
Linus Torvalds started work on open source Linux as a college student. After Mac OS X, Linux is the most widely used
variation of UNIX.
Linux is technically just the kernel (innermost part) of the operating system. The outer layers consist of GNU tools. GNU
was started to guarantee a free and open version of UNIX and all of the major tools required to run UNIX.
Operating Systems: Linux kernel; Mac OS 7
Programming Languages: Pov-Ray; Python; Visual BASIC
Computers: Apple PowerBook; PowerPC (microprocessor); PARAM 8000 (India, created by Dr. Vijay Bhatkar,
1GFLOP)
Games: Street Fighter II (shooter); The Legend of Zelda: A Link to the Past (fantasy RPG); Sonic the Hedgehog; Sid
version 56
page 299
Unix and Linux System Administration and Shell Programming
Meyer’s Civilization
Technology: CD-i
1992
Dylan was created in 1992 by Apple Computer and others. Dylan was originally intended for use with the Apple
Newton, but wasn’t finished in time.
Oak, the forerunner of Java, was developed at Sun Microsystems.
In 1992 Unix System Laboratories sued Berkeley Software Design, Inc and the Regents of the University of California to
try to stop the distribution of BSD UNIX. The case was settled out of court in 1993 after the judge expressed doubt over the
validity of USL’s intellectual property.
Operating Systems: Solaris; Windows 3.1; OS/2 2.0; SLS Linux; Tru64 UNIX
Programming Languages: Dylan; Oak
Computers: Cray C90 (1 GFLOP)
Games: Wolfenstein 3D (ffirst fully 3D rendered game engine); Mortal Kombat; NHLPA 93 (multiplayer hockey sports
game); Dune II (first real-time strategy game)
1993
“Vendors such as Sun, IBM, DEC, SCO, and HP modified Unix to differentiate their products. This
splintered Unix to a degree, though not quite as much as is usually perceived. Necessity being the mother of
invention, programmers have created development tools that help them work around the differences between
Unix flavors. As a result, there is a large body of software based on source code that will automatically
configure itself to compile on most Unix platforms, including Intel-based Unix.
Regardless, Microsoft would leverage the perception that Unix is splintered beyond hope, and present
Windows NT as a more consistent multi-platform alternative.” —Nicholas Petreley, “The new Unix alters
NT’s orbit”, NC World
AppleScript, a scripting language for the Macintosh operating system and its application softweare, was
released by Apple Computers.
Version 2.5.1 of Photoshop released in 1993. It was one of the first programs to run native on a PowerPC chip.
First release of Windows version of Photoshop in April 1993. Toolbar for Photoshop 2.5.1 to the right.
Operating Systems: Windows NT 3.1; Stackware Linux; Debian GNU/Linux; Newton
Programming Languages: AppleScript
Computers: Cray EL90; Cray T3D; Apple Newton; Apple Macintosh TV; Intel Pentium (microprocessor,
66MHz); Thinking Machines CM-5/1024 (59.7 GFLOPS); Fujitsu Numerical Wind Tunnel (Japan, 124.50
GFLOPS); Intel Paragon XP/S 140 (143.40 GFLOPS)
Games: Myst (first puzzle-based computer adventure game; CD-ROM game for Macintosh); Doom (made the first
person shooter genre popular, pioneering work in immersive 3D graphics, networked multiplayer gaming, and support for
customized additions and modifications; also proved that shareware distribution could work for game distribution); U.S.
Senator Joseph Lieberman holds Congressional hearings and attempts to outlaw violent games
Technology: MP3
1994
Work continued on Java with a version designed for the internet.
Version 3.0 of Photoshop released in 1994. It included Layers.
Operating Systems: Red Hat Linux
version 56
page 300
Unix and Linux System Administration and Shell Programming
Programming Languages: Java
Computers: Cray J90; Apple Power Macintosh; Sony PlayStation; Fujitsu Numerical Wind Tunnel (Japan, 170.40
GFLOPS)
Games: Super Metroid (RPG)
Technology: DNA computing
Theory: Peter Shor publishes Shor’s algorithm for integer factorization wit h a quantum computer.
1995
Java (named for coffee) was created by James Gosling and others at Sun Microsystems and released for applets in 1995.
Original work started in 1991 as an interactive language under the name Oak. Rewritten for the internet in 1994.
JavaScript (originally called LiveScript) was created by Brendan Elch at Netscape in 1995. A scripting language for web
pages.
PHP (PHP Hypertext Processor) was created by Rasmus Lerdorf in 1995.
Ruby was created in 1995 by Yukihiro Matsumoto. Alternative to Perl and Python.
Delphi, a variation of Object Pacal, was released by Borland
Operating Systems: OpenBSD; OS/390; Windows 95
Programming Languages: Delphi; Java; JavaScript; PHP; Ruby
Computers: BeBox; Cray T3E; Cray T90; Intel Pentium Pro (microprocessor); Sun UltraSPARC (microprocessor)
Software: Microsoft Bob
Games: Chrono Trigger (fantasy RPG); Command & Conquer (real time strategy game)
Technology: DVD; wikis
1996
UML (Unified Modeling Language) was created by Grady Booch, Jim Rumbaugh, and Ivar Jacobson in 1996 by
combining the three modeling languages of each of the authors.
Version 4.0 of Photoshop released in 1996. It included a controversial change in the key commands.
Operating Systems: MkLinux
Programming Languages: UML
Computers: Hitachi SR2201/1024 (Japan, 220.4 GFLOPS); Hitachi/Tsukuba CP=PACS/2048 (Japan, 368.2 GFLOPS)
Games: Tomb Raider and Lara Croft, Pokémon; Quake (first person shooter using new 3D rendering on daughter
boards); Super Mario 64 (3D rendering)
Technology: USB
1997
REBOL (Relative Expression Based Object language) was created by Carl SassenRath in 1997. Extensible scripting
language for internet and distributed computing. Has 45 types that use the same operators.
ECMAScript (named for the European standards group E.C.M.A.) was created in 1997.
Alloy, a structural modelling language, was developed at M.I.T.
Rhapsody The first developer version of Mac OS X released.on August 31, 1997.
To date, the most widely used desktop version of UNIX is Apple’s Mac OS X, combining the ground breaking object
oriented NeXT with some of the user interface of the Macintosh.
version 56
page 301
Unix and Linux System Administration and Shell Programming
Operating Systems: Mac OS 8; Rhapsody
Programming Languages: Alley; ECMAScript; REBOL
Computers: AMD K6 (microprocessor); Intel Pentium II (microprocessor); PalmPilot; Intel ASCI Red/9152 (1.338
TFLOPS)
Software: AOL Instant Messenger
Games: Final fantasy VII; Goldeneye 007 (one of the few successful movie to game transitions; based on 1995 James
Bond movie Goldeneye); Castlevania: Symphony of the Night (2D fantasy RPG)
Technology: web blogging
1998
Version 5.0 of Photoshop released in 1998. It included the History palette.
Operating Systems: Windows 98
Computers: Apple iMac; Apple iMac G3; Intel Xeon (microprocessor)
Games: The Legend of Zelda: Ocarina of Time (fantasy RPG); Metal Gear Solid; Crash Bandicoot: Warped
1999
Mac OS X Mac OS X Server 1.0, intended as a server operating system and not for general desktop use, and
Mac OS X Developer Preview were released on March 16, 1999.
Version 5.5 of Photoshop released in 1999. It was the first web ready version of Photoshop. Toolbar for
Photoshop 5.5 to the right.
Operating Systems: Mac OS 9; Mac OS X Server
Computers: PowerMac; AMD Athlon (microprocessor); Intel Pentium III (microprocessor); BlackBerry; Apple
iBook; TiVo; Intel ASCI Red/9632 (2.3796 TFLOPS)
2000s
Some operating systems from the 2000s include: Mac OS X, Syllable, Windows 2000, Windows Server 2003,
Windows ME, and Windows XP.
2000
D, designed by Walter Bright, is an object-oriented language based on C++, Java, Eiffel, and Python.
C# was created by Anders Hajlsberg of Microsoft in 2000. The main language of Microsoft’s .NET.
RELAX (REgular LAnguage description for XML) was designed by Murata Makoto.
Mac OS X The Public Beta version of Mac OS X was released on September 13, 2000.
Operating Systems: Mac OS 9; Windows ME; Windows 2000
Programming Languages: C#; D; RELAX
Computers: Intel Pentium 4 (microprocessor, over 1 GHz); Sony PlayStation 2; IBM ASCI White (7.226 TFLOPS)
Games: Tony Hawk’s Pro Skater 2 (skateboarding sports game); Madden NFL 2001 (football sports game)
Technology: USB flash drive
2001
AspectJ (Aspect for Java) was created at the Palo Alto Research Center in 2001.
Scriptol (Scriptwriter Oriented Language) was created by Dennis G. Sureau in 2001. New control structuress include for
version 56
page 302
Unix and Linux System Administration and Shell Programming
in, while let, and scan by. Variables and literals are objects. Supports XML as data structure.
Mac OS X 10.0 Mac OS X 10.0 Cheetah was released on March 24, 2001.
Mac OS X 10.1 Mac OS X 10.1 Puma was released on September 25, 2001. It included DVD playback.
Operating Systems: Mac OS X 10.0 Cheetah; Mac OS X 10.1 Puma; Windows XP; z/OS
Programming Languages: AspectJ; Scriptol
Computers: Nintendo GameCube; Apple iPod; Intel Itanium (microprocessor); Xbox
Games: Halo
Technology: blade server
2002
Mac OS X 10.2 Mac OS X 10.2 Jaguar was released on August 24, 2002. It included the Quartz Extreme
graphics engine, a system-wide repository for contact information in the Address Book, and an instant messaging
client iChat.
Operating Systems: Mac OS X 10.2 Jaguar
Computers: Apple eMac; Apple iMac G4; Apple XServe; NEC Earth Simulator (Japan, 35.86 TFLOPS)
2003
Mac OS X 10.3 Mac OS X 10.3 Panther was released on October 24, 2003. It included a brushed metal look,
fast user switching, Exposé (a Window manager), Filevault, Safari web browser, iChat AV (which added
videoconferencing), imporved PDF rendering, and better Microsoft Windows interoperability.
Operating Systems: Windows Server 2003; Mac OS X 10.3 Panther
Computers: PowerPC G5; AMD Athlon 64 (microprocessor); Intel Pentium M (microprocessor)
Games: Final Fantasy Crystal Chronicles
2004
Scala was created February 2004 by Ecole Polytechnique Federale de Lausanne. Object oriented language that
implements Python features in a Java syntax.
Programming Languages: Scala
Computers: Apple iPod Mini; Apple iMac G5; Sony PlayStation Portable; IBM Blue Gene/L (70.72 TFLOPS)
Games: Fable
Technology: DualDisc; PCI Express; USB FlashCard
2005
Job Submission Description Language.
Mac OS X 10.4 Mac OS X 10.4 Tiger was released on April 29, 2005. It included Spotlight (a find routine),
Dashboard, Smart Folders, updated Mail program with Smart mailboxes, QuickTime 7, Safari 2, Automator,
VoiceOver, Core Image, and Core Video.
Operating Systems: Mac OS X 10.4 Tiger
version 56
page 303
Unix and Linux System Administration and Shell Programming
Programming Languages: Job Submission Description Language
Computers: IBM System z9; Apple iPod Nano; Apple Mac Mini; Intel Pentium D (microprocessor); Sun UltraSPARC
IV (microprocessor); Xbox 360
Games: Lego Star Wars
2006
Computers: Apple Intel-based iMac; Intel Core 2 (microprocessor); Sony PlayStation 3
Technology: Blu-Ray Disc
2007
Mac OS X 10.5 Mac OS X 10.5 Leopard was released on October 26, 2007. It included an updated Finder, Time
Machine, Spaces, Boot Camp, full support for 64-bit applications, new features in Mail and iChat, and new security
features.
Operating Systems: Apple iOS (for iPhone); Mac OS X 10.5 Leopard; Windows Vista
Computers: AMD K10 (microprocessor), Apple TV; Apple iPhone; Apple iPod Touch; Amazon Kindle
2008
Operating Systems: Google Android
Computers: Android Dev Phone 1; BlackBerry Storm; Intel Atom (microprocessor); MacBook Air; IBM RoadRunner
(1.026 PFLOPS); Dhruva (India, 6 TFLOPS)
2009
Mac OS X 10.6 Mac OS X 10.6 Snow Leopard was released on August 28, 2009. it included the SquirrelFish
javaScript interpreter, a rewrite of Finder in the Cocoa API, faster Time Machine backups, more reliable disk ejects,
a more powerful version on Preview, a faster version of Safari, Microsoft Exchange Server support for Mail, iCal,
and Address Book, QuickTime X, Grand Central Dispatch for using multi-core processors, and OpenCL support.
Operating Systems: Windows 7; Mac OS X 10.6 Snow Leopard
Computers: Motorola Driod; Palm Pre; Cray XT5 Jaguar (1.759 PFLOPS)
2010
Computers: Apple iPad; IBM z196 (microprocessor); Apple iPhone 4; Kobo eReader; HTC Evo 4G
2011
Mac OS X 10.7 Mac OS X 10.7 Lion was released on July 20, 2011. it included Launchpad, auto-hiding
scrollbars, and Mission Control.
Operating Systems: Mac OS X 10.7 Lion
2012
Mac OS X 10.8 Mac OS X 10.8 Mountain Lion was released on July 25, 2012. It included Game Center,
Notification Center, iCloud support, and more Chinese features.
Operating Systems: Mac OS X 10.8 Mountain Lion, Windows 8
2013
version 56
page 304
Unix and Linux System Administration and Shell Programming
Mac OS X 10.9 Mac OS X 10.9 Mavericks was released on October 22, 2013.
Operating Systems: Mac OS X 10.9 Mavericks, iOS7, WIndows 8.1
Games: Sony PlayStation 4
version 56
page 305
Unix and Linux System Administration and Shell Programming
critical reasoning
appendix C
summary
This appendix looks at critical reasoning.
This is the beginning of the much-promised appendix on critical reasoning. At the moment I am attempting to get the
basic structure laid down. This presentation will be too fast paced to be useful to a student, but will help organize the
themes that must be discussed to adequately present the material.
Students who want to learn this material can use this as an outline. Look up the appropriate topics in Wikipedia and you
will find a brief discussion on each topic. Wikipedia isn’t organized for teaching, but using this appendix as a study outline
and Wikipedia for information should allow you to cobble together your own lessons.
Important Note: While critical reasoning is a great tool for computers, it clashes with most social institutions, including
business, law, and government. Business persons and government officials do not normally react well to critical reasoning
and science. Do not confuse the two realms. Use critical reasoning and mathematcis in your work, but switch to social skills
for all interactions with those in positions of power or authority.
tinkering and thought
The vast majority of programmers, system administrators, data base administrators, and other tech personnel are
tinkerers.
Tinkerers have no real understanding of the objects they work on. Tinkerers use trial and error until things work better.
The best tinkerers are successful because they have a good intuition about what guesses might work and over time they
have collected a set of superstitutions that work most of the time for the most common problems.
The very best programmers, system administrators, data base administrators, and other tech personnel use rational
thought and critical reasoning to understand how a system works and fails and uses a combination of knowledge, skill, and
imagination to craft appropriate solutions.
This chapter is about how to use critical reasoning. The vast majority of readers can safely skip over this material.
Engineering and Business
Engineering and business are fundamentally different. This web page is intended to help business persons successfully
make use of engineers.
I will describe the fundamental difference, outline the basic web services, and discuss how to get engineers to deliver
what your business needs
fundamental difference
The fundamental difference is that business is a social activity and engineering is a mathematical activity. The computer
is a pure mathematical machine.
In engineering (as well as science and mathematics), two plus two (2+2) always equals four (4), for all times and places,
regardless of culture. The answer is the same for Stephen Hawking in modern Great Britain and Euclid in Egypt some two
thousaand years ago.
version 56
page 306
Unix and Linux System Administration and Shell Programming
In business (as well as law and government), two plus two might be equal to five-and-a-quarter, trending up. But if it
reaches six, consider that a bubble and sell before it falls. If someone is rich or powerful enough, they can demand that two
plus two be whatever they need it to be — and nobody dare risk their wrath for failure to obey.
Engineering is absolute and business is relative.
A business person needs to be aware that engineers think in a completely different manner and that their machines do
not answer to money or power. Computers answer only to the laws of mathematics.
The reason that business persons often see disconcerting, strange characters among the geeks is because they are highly
skilled at mathematics and mathematics doesn’t take into account fashion or custom.
business and government
Business, government, law, and most social institutions are all subjective perception.
The clash between subjective business perceptions and the harsh absolutism of mathematics and computers is often
confounding to programmers and other technical staff.
It makes sense to base business on perception because consumerism is based not on actual needs but on selling wants and
desires.
It makes sense to base government and law on perception because that allows those in power to continually adjust their
rulings to meet their own needs.
perception and proof
Reality is perceived. Mathematics is proven.
The existence of optical illusions illustrates the inherent lack of reliability of our perceptions of reality. Different
testmony by witnesses to the same event highlight differences in perception by different observers. Some philosphers have
questioned whether an objective reality even exists.
Mathematics is pure in abstraction. Mathematics has clear definitions and axioms, from which theorems can be proven
true or false.
Philosophers have long examined this distinction.
Imhotep, the chief architect of the first two true pyramids in ancient Egypt, inventor of the scientific method, inventor of
medical triage, and inventor of trigonometry, claimed that mathematics were the fundamental divine laws of physical
reality.
Modern applied mathematics is based on the idea that mathematics models reality.
Computers simulate pure mathematics. Unlike pure math, physical hardware is subject to momentary glitches, such as
electrical spikes, cosmic rays, and distant earthquakes.
objectivity and subjectivity
The goal in critical reasoning is objectivity.
The goal in artwork is subjectivity.
Both are valid goals. Kung Fu-Tze (Confucious), Aristotle, and Buddha all proposed some variation of the belief that
humans need balance.
version 56
page 307
Unix and Linux System Administration and Shell Programming
A few hundred years ago, European artists attempted to match exactly what they saw, which was called realism.
Impressionism moved to attempting to capture the essential impression of a scene. Surrealism bent and warped reality.
Abstract expressionism attempted to capture the abstract essence. Numerous other art movements employed new and
different ways to look at the world and express artistic creativity.
Objectivity calls for a fair examination, an attempt to determine a greater truth.
Objective journalism asks six questions; who, what, when, where, why, and how.
classifications
summary
This chapter looks at classifications.
Classifying is the act of placing things into groups.
A famous example is the Dewey Decimal Classification used by some libraries.
concepts and referents
A concept is a category.
A referent is aspecific member of a category.
Fido is a referent in the concept of dogs. Cats and dogs are referents in the concept of mammals. Mammals and reptiles
are referents in the concept of animals.
species and genus
A species is a smaller, more specific group, while a genus is a larger, more general group.
Using the above example, cats and dogs are species of the genus mammals, while mammals and reptiles are species of
the genus animals.
Note that these older philosophical concepts of species and genus predate the more complex modern biological system of
categories.
abstract and concrete
Each higher layer of genus becomes more abstract. Each lower layer of species becomes more concrete.
Abstract is a relative term.
Concrete is not limited to physical items. For example, specific emotions (sad, happy) are more concrete than emotional
states.
classes and objects
version 56
page 308
Unix and Linux System Administration and Shell Programming
You see this same kind of grouping in object oriented programming. A class is a genus (concept) and the various objects
are the species (referent).
A class defines the characteristics (data) and behaviors (methods) of a group.
An object is a specific individual (instance) of a group.
rules of classification
There are two basic rules for good classifications:
(1) A set of classifications must be used according to a consistent principle that is mutually exclusive and jointly
exhaustive.
(2) Classifications must be made on essential attributes.
mutually exclusive
Mutually exclusive means that any particular species can only be in a single genus.
Green characters and Muppets are not mutually exclusive, because Kermit the Frog qualifies for both genuses.
jointly exhaustive
Jointly exhaustive means that all of the species taken together (jointly) will make up the complete (exhaustive) set of the
genus.
The species felines (cats) and canines (dogs) are not jointly exhaustive for the genus mammals, because there are still
many additional creatures that qualify as mammals, but aren’t felines or canines.
consistent principle
It is important to have a consistent principle for organizing your classifications.
The species of (1) green things and (2) big things and (3) mammals are not consistent because the first measure is color,
the second size, and the third a biological group.
The species of red, green, and blue are consistent because they are all colors.
essential attributes
It is important that the choice of rule for classification be based on an essential attribute.
Organizing books by the colors of the covers would create a bizarre library, while organizing books by subject (such as
the Dewey Decimal Classification) creates a library that is easy to use.
This is because the subject matter (or author) is an essential characteristic of a book, while the color of the cover isn’t.
The color of the cover can change from printing to printing.
levels of organization
The levels of ogranization need to make sense.
As an example, the Dewey Decimal Classification is made up of ten classes. Each class is divided into ten divisions.
version 56
page 309
Unix and Linux System Administration and Shell Programming
Each division is divided into ten sections.
Class 0 is Computer science, information & general works. Division 00 is Computer science, information & general
works. Section 004 is Data processing & computer science, 005 is Computer programming, programs & data, and 006 is
Special computer methods.
000 Computer science, information & general works
000 Computer science, information & general works
000 Computer science, knowledge & general works
001 Knowledge
002 The book (i.e. Meta writings about books)
003 Systems
004 Data processing & computer science
005 Computer programming, programs & data
006 Special computer methods
007 (Unassigned)
008 (Unassigned)
009 (Unassigned)
010 Bibliographies
020 Library & information sciences
030 Encyclopedias & books of facts
040 Biographies
050 magazines, journals & serials
060 Associaitons, organizations & museums
070 News media, journalism & publishing
080 General collections
090 Manuscripts & rare books
100 Philosophy and psychology
200 Religion
300 Social sciences
400 Language
500 Science
600 technology
700 Arts & recreation
800 Literature
900 History & geography
Make sure that your categories have a species-genus relationship. For example, an enzyme is a kind of protein and
therefore they have a species-genus relationship, but an amino acid is is a component part of a protein and therefore they
don’t have a species-genus relationship.
outlines
The outlines that you were taught in school are an example of a proper organization of classifications. Outlines developed
specifically because they are useful tools for organizing a well-reasoned argument, presentation, or essay.
I. first main topic
A. first subtopic
i. first inner topic
a. first innermost topic
b. second innermost topic
ii. second inner topic
a. first innermost topic
b. second innermost topic
B. second subtopic
i. first inner topic
version 56
page 310
Unix and Linux System Administration and Shell Programming
a. first innermost topic
b. second innermost topic
ii. second inner topic
a. first innermost topic
b. second innermost topic
C. third subtopic
i. first inner topic
a. first innermost topic
b. second innermost topic
ii. second inner topic
a. first innermost topic
b. second innermost topic
II. second main topic
A. first subtopic
i. first inner topic
a. first innermost topic
b. second innermost topic
ii. second inner topic
a. first innermost topic
b. second innermost topic
B. second subtopic
i. first inner topic
a. first innermost topic
b. second innermost topic
ii. second inner topic
a. first innermost topic
b. second innermost topic
C. third subtopic
i. first inner topic
a. first innermost topic
b. second innermost topic
ii. second inner topic
a. first innermost topic
b. second innermost topic
III. third main topic
A. first subtopic
i. first inner topic
a. first innermost topic
b. second innermost topic
ii. second inner topic
a. first innermost topic
b. second innermost topic
B. second subtopic
i. first inner topic
a. first innermost topic
b. second innermost topic
ii. second inner topic
a. first innermost topic
b. second innermost topic
C. third subtopic
i. first inner topic
a. first innermost topic
b. second innermost topic
ii. second inner topic
a. first innermost topic
b. second innermost topic
version 56
page 311
Unix and Linux System Administration and Shell Programming
elements
In Western antiquity, the ancient Egyptians and the Greeks developed the idea of four or five basic elements: fire, water,
air, earth, and spirit. The Chinese developed the five basic elements of wood, fire, earth, metal, and water.
The Western elements represented the basic underlying principles of natures: three states of matter (earth = solid, water =
liquid, and air = gas) and energy (fire = energy). The spirit or soul was sometimes included as a fifth element.
In modern Western civilization the word element was purposely reused to describe the fundamental chemical elements
that make up matter.
definition
summary
This chapter looks at definitions.
The first rule of philosophy is to define your terms. The reason is to create clarity and make sure that everyone is
discussing the same thing.
Churchill
There is a famous example from World War II. In American use, the term “table” means to put a matter aside. In the rest
of the English-speaking world, the term “table” means to begin discussion on a matter.
Winston Churchill describes how this became a source of confusion during World War II in his book The Second World
War, Volume 3: The Grand Alliance.
The enjoyment of a common language was of course a supreme advantage in all British and American
discussions. … The British Staff prepared a paper which they wished to raise as a matter of urgency, and
informed their American colleagues that they wished to “table it.” To the American Staff “tabling” a paper
meant putting it away in a drawer and forgetting it. A long and even acrimonious argument ensued before both
parties realized that they were agreed on the merits and wanted the same thing.
dictionaries
Why not just use a dictionary?
Dictionaries serve a different purpose. The definitions in dictionaries are about words. The definitions in philosophy are
about concepts.
Even though we have a different purpose for definition in critical reasoning, a dictionary definition can be a good
starting point for coming up with a useful philosophical definition.
A Webster’s style dictionary (there are many different brands, not one single version) attempts to record how the
common person uses the language. The first was Noah Webster’s 1928 edition of the American Dictionary of the English
Language.
version 56
page 312
Unix and Linux System Administration and Shell Programming
An authorative dictionary attempts to record expert use of the language (such as the works of great writers). Originally
authorative dictionaries recorded how the king or queen wanted the language to be used. The famous example is the Oxford
English Dictionary, which is printed on behalf of British royalty.
A lexicon is a specialized dictionary that records the words and phrases for a particular use, such as a law or science
dictionary. The famous example is Black’s Law Dictionary.
boundaries
One of the most important aspects of a good definition is setting boundaries. A definition divides the universe into that
which fits the definition and that which doesn’t fit the definition.
It is generallly easy to deal with cases that are clearly within a definition and cases that are clearly outside of a definition.
The problem is dealing well with the cases that are near the borders. In computer science, these are sometimes called the
“corner cases”.
precision and accuracy
Precision is how exacting something is. For example, 3.14159 is more precise than three.
Accuracy is how correct something is.
The classic example is:
“2 + 2 = 3.9999” is very precise, but inaccurate.
“Two plus two is less than ten” is accurate, but imprecise.
We want any definition to be accurate.
There are always degrees of precision. In philosophy we want to be as precise as necessary. We want to be precise
enough to avoid confusion, but have no need to go to extremes for precision’s own sake.
clarity
The basic function of a definition in critical reasoning is to be clear. Our goal is to avoid confusion.
Working towards great clarity in definitions actually helps us better understand the topics we are thinking about.
In the last half century or so there has been a lot of work on how language and words powerfully control our thoughts
and even what we are able to think of.
summary
A good definition should summarize the concept.
genus and differentia
A classic test of a good definition for critical reasoning is that it includes a genus and a differentia.
The genus let’s us know the class of concepts.
The differentia (difference) let’s us know how this particular concept difefrs from other similar concepts.
In normal speech it is common to leave the genus implied. For good critical reasoning, it is important to actually state the
version 56
page 313
Unix and Linux System Administration and Shell Programming
genus.
In some cases the genus is essential for distinguishing between multiple meanings of a word.
classical logic
summary
This chapter looks at classical or Aristotelian logic.
Aristotle was a Greek philosopher and scientist who lived from 384 to 322 B.C.E. in Athens, Greece. He was a puil of
Plato, who in turn was a pupil an dfriend of Socrates.
logical argument
Aristotle claimed that all logical argument could be reduced to two premises (a major premise and a minor premise) and
a conclusion.
three laws
Aristotle created three basic laws or principles of logical reasoning:
1. The principle of identity
A thing is itself: A is A.
2. The principle of the excluded middle
A proposition is either true or false: either A or not A.
3. The principle of contradiction
No proposition can be both true and false: A cannot be both A and not A.
version 56
page 314
Unix and Linux System Administration and Shell Programming
engineering
summary
This chapter looks at engineering.
engineering method
There are numerous variations of the engineering method. There is no official version. Most have four steps, although
five and six step versions also exist. In general, all of the versions include the same basic stuff, but divide it up differently
or express it differently.
1.
2.
3.
4.
Define problem.
Brainstorm solutions.
Evaluate and choose one method
Plan implementation
1. The first step is to identify the actual problem that is being solved. The vast majority of projects, especially
those created by government or business, start instead with a statement of a particular solution.
In engineering, it is best to start with a clear, neutral definition of the problem that must be solved.
2. The second step is to look at all possible solutions. This includes examining the existing record to see how
others have solved the same or similar problems.
3. The third step is to evaluate the risks, benefits, costs, advantages, and disadvantages of the possible
solutions and determine which solution best meets the needs and budget.
4. The fourth and final step is to come up with a plan to implement the solution. In computer science, this
centers on the software design.
version 56
page 315
Unix and Linux System Administration and Shell Programming
Forth-like routines for Unix/Linux shell
appendix D
summary
This chapter looks at a set of Forth-like routines for use in the BASH UNIX (Linux, Mac OS X) shell.
You can obtain the latest copy of these functions at http://www.osdata.com/programming/shell/forthfunctions
place the file in an appropriate directory.
warning
At the time this paragraph was written the www.osdata.com was hosted on HostGator and at the time this
paragraph was written HostGator tech support used the last four digits of a credit card as their security method
for identifying the account holder. This is a well-known security flaw. The security hole was made famous
when a hacker used this hole at Apple and Amazon to get Mat Honan’s account information and wipe out his
hard drive. Both Apple and Amazon quickly changed their security procedures, but for some reason HostGator
(at the time this was written) hasn’t bothered to change their security procedures. See
http://www.wired.com/gadgetlab/2012/08/apple-amazon-mat-honan-hacking/ for details.
For this reason, there is the distinct possibility that the forthfunctions file might have been hacked. Please
take the time to read through it and make sure it looks correct and doesn’t have any suspicious or malicious
code inserted. I have a distinctive programming style, so something written by someone else might stand out
even if you don’t understand what I have written. Of course, it is better if you understand the code and can spot
things that look wrong.
I apologize for this security hole, but my website is hosted for free for me and I have no control over the
hosting company chosen. The person providing me with the free hosting doesn’t believe me that this is a
security flaw. This is one of the many hassles of poverty.
basic approach
This supplementary material describes a method for creating Forth-like routines for use in the BASH shell.
The data and return stacks will be stored using a pair of shell variables, one of which is an array and the
other is an ordinary scalar variable.
The Forth-like routines will be created as shell functions.
Everything will be stored in a shell function file that can be loaded manually or loaded as part of the normal
shell start-up.
Eventually I intend to write two sets of C source code, one that people can compile into their own copy of
the official BASH source code and one that is part of an entirely new open source shell project that doesn’t
have the GNU infecting license. This alternative shell will be designed so that it can run on bare metal,
allowing it to be used to deploy embedded systems.
stacks
Forth uses two stacks: the main data stack and the return stack.
For the purposes of the BASH shell, we will implement both stacks with a combination of a top of stack
version 56
page 316
Unix and Linux System Administration and Shell Programming
pointer variable and an array for the stack. The top of stack pointer will hold the array index of the top of the
stack. The stack will build up from the array indexed as one. An empty stack is indicated by having the top of
stack pointer variable point to the zero index of the stack array variable. The bottom of the stack will actually
be the one index of the stack array variable.
$
$
$
$
$
export DataStackPointer=0
declare -a DataStack; export DataStack
export ReturnStackPointer=0
declare -a ReturnStack; export ReturnStack
That’s it for the required variables! And, yes, those could all be placed on one line, but I spread them over
four lines for additional clarity.
One very important difference between the real Forth stacks and the Forth-like function stacks is that the real
Forth stacks deal with binary data and the Forth-like function stacks deal with character strings. Each item on a
real Forth stack is a binary number (or binary bit string). Even characters are stored by their binary
representation. Every item on a Forth-like function stack is a character string. Numbers are stored by their
character string representation. And an individual item on a Forth-like stack can be a character string of
arbitrary length, even an entire book.
direct random access
While Forth insists on the stack always being accessed as a stack, there is nothing that prevents direct
random access to Forth-like stack elements.
A shell script mixing BASH and the Forth-like functions can directly read or write any individual element in
the data or return stacks. For that matter, a mixed shell script can do any manipulations to the stacks. It is the
programmer’s responsibility to maintain stack integrity.
One particular gotcha is that I am not currently emptying out stack locations that are no longer used. This
means that when the stack shrinks, there will be entries in the array that are no longer valid. Don’t be fooled by
their existence. Modify my scripts to delete this information if it presents a security risk of any kind.
functions
These are the functions for a Forth-like working environment. Not all of these functions exist in a real Forth.
Most of these functions are have modifications from standard Forth. Many standard Forth functions are
missing.
In particular, this approach treats the stack as a stack of objects. No matter how much memory space an
object takes up, it only takes up a single stack slot. In a real Forth, data is stored on the stack as raw binary
strings, characters, or numbers and the programmer is responsble for keeping track of how much memory every
item actually uses.
stack functions
The double number versions of Forth stack functions are purposely left out. Our stack elements are variable
size objects and can be any size number, as well as strings or characters.
ShowDataStack
The ShowDataStack function shows the current contents of the data stack in the order of top of stack to
version 56
page 317
Unix and Linux System Administration and Shell Programming
bottom of stack. This function is built for debugging purposes, but I am making it available for anyone to use.
$ ShowDataStack ()
>{
> echo "stack size of " $DataStackPointer;
> if [ "$DataStackPointer" -gt "0" ] ; then
> loopcounter=$DataStackPointer
> while [ $loopcounter -gt 0 ] ;
> do
> echo ${DataStack["$loopcounter"]}
> let "loopcounter=$loopcounter - 1"
> done
> else
> echo "the stack is empty"
> fi
> unset loopcounter #clear out variable
>} # end ShowDataStack
$
push
The push function pushes a single object onto the data stack. Forth doesn’t have this word because simply
typing a number is sufficient to push the number onto the stack. This function will be used by other functions
to manipulate the data stack.
NOTE: This function does not yet have error checking. Need to check for exactly one item to push.
$ push ()
>{
> let "DataStackPointer=$DataStackPointer + 1"; #preincrement
> DataStack["$DataStackPointer"]="$1"; #push data item on top of stack
>} # end push
$
It is very important to notice that the item to be pushed onto the stack comes after the Forth-like function
push , which is very different than the normal order for Forth. This variation appears again in other Forth-like
functions and is done to work within the limits of the expectations of the BASH shell.
Note that push will fail inside a series of piped commands or any other situation with a subshell.
pop
The pop function pops a single object from the data stack. Forth has a different function for this
functionality, but this function is intended for internal use by other functions to manipulate the data stack.
NOTE: This function does not yet have error checking. Need to check for empty stack.
$ pop ()
>{
> let "DataStackPointer=$DataStackPointer - 1" #postdecrement
>} # end pop
$
version 56
page 318
Unix and Linux System Administration and Shell Programming
popout
The popout function pops a single object from the data stack and prints it out for use in a pipe. Forth has a
different function for this functionality, but this function is intended for internal use by other functions to
manipulate the data stack.
Note that the value is printed without a new line character, so it might seem to “disappear” into the beginning
of the prompt.
Use the DOT function to pop and print with a new line.
NOTE: This function does not yet have error checking. Need to check for empty stack.
$ popout ()
>{
> printf ${DataStack["$DataStackPointer"]} #pop data item from top of
stack and print
> let "DataStackPointer=$DataStackPointer - 1" #postdecrement
>} # end popout
$
the power of Forth-like functions
For those who don’t already know the power of Forth, let me use the shell commands pushd , popd , and dirs
as an example. For those who already know the power of Forth, you can either read along and nod about how
great Forth is, or skip ahead to actual implementation of Forth-like functions for the BASH shell.
directory stack
You may have noticed that the C shell and BASH include the pushd , popd , and dirs command to
manipulate a stack of directories.
In case you never noticed these commands or have forgotten what they do, a quick summary:
pushd allows the user or a script to push the current working directory onto a stack of directories and change
to a new named directory. The command acts slightly differently if you leave out a directory name. We won’t
discuss that case because it takes us on a tangent from our current discussion.
You can simulate the normal use of the shell command pushd (named directory) with the Forth-like function
and shell command cd.
push
$ push `pwd`; cd testdir
$
You have to explicitly place the current working directory onto the stack (using pwd and backticks) and you
have to issue a second command to actually change the directory. Don’t worry about the extra steps, because
you will continue to use pushd for its intended purpose. We are showing the value of Forth through analogy to
something you already use to save time and effort.
popd allows the user or a script to pop the directory stack and change to that directory.
You can simulate the normal use of the shell command popd with the Forth-like function pop and shell
command cd.
version 56
page 319
Unix and Linux System Administration and Shell Programming
$ cd `pop`
$
Notice that we used the results of the pop function as the parameter for the shell command cd.
dirs allows the user or a script to see all of the directories on the directory stack.
You can simulate the normal use of the shell command dirs with the Forth-like function ShowDataStack .
$ ShowDataStack
stack size of 1
/Users/admin
$
taking it to the next level
At this point, there is no advantage to using our Forth-like functions rather than using the already existing
shell commands. It is a good idea to continue to use the shell commands for their intended use. Less typing.
You could also use named variables for saving directory names. By using short variable names, you can save
a lot of typing over typing out the full path names. And you don’t have to memorize a bunch of long path
names. Of course, you have to memorize all of your named variables.
The shell commands pushd , popd , and dirs were developed because there are a lot of advantages of a stack
over a bunch of named variables.
With a stack, you don’t have to memorize any alternate short variable names for directory paths. And you
can use the same directory stack for work that moves from an old set of directories to a new set of directories.
There is a lot of power in having a stack and three simple commands for manipulating the stack.
more sophisticated directory stack use
The shell commands pushd and popd only work with the single directory that is at the top of the stack. Yes,
you can see the entire stack with the dirs command and use one of several alternate techniques for picking
one directory out of the middle of the stack.
What if you could easily swap the directory that was on the top of the stack with the directory that was one
down? You could then use the popd command to switch to the directory that was formerly one down from the
top of the stack and the former top of the stack would still be in the stack, ready for use.
As it turns out, Forth has a command called SWAP that does exactly that operation. Forth has a rich set of
stack operation words, that can do things like move (or pick) an aribitrary item from deep in the stack to the
top of the stack, make a duplicate (or dup) of the top of the stack, delete (or drop) the item on the top of the
stack, and other cool things.
The Forth-like functions bring variations of these stack operation words to the shell.
arithmetic computations
Both Forth and the Forth-like functions work on much more than just directories.
A common use of Forth is computations.
version 56
page 320
Unix and Linux System Administration and Shell Programming
Many of the scientific and engineering calculators (which used to be stand-alone devices, but are now
commonly apps) use a system called Reverse Polish Notation or RPN
An RPN calculator is a very powerful and efficient way of performing calculations.
For ordinary calculations, such as adding a series of numbers, RPN only saves a single keystroke. That’s not
enough to justify learning an entirely different method of calculating.
But when the calculations start to become complicated, RPN really shines.
With standard algebraic notation, you suddenly have to start typing a whole bunch of open and close
parenthesis and figure out all of your grouping and keep track of how many layers of parenthesis are currently
still open so that at the end of your calculation all of the open and close parenthesis balance out. Anyone who
has done any C programming knows about this hassle.
With RPN, there are no parenthesis. Once you learn to think in RPN, any arbitrarily complex algebraic
computation has a straight-forward and simple RPN expression!
You dramatically save typing, you avoid the headache of balancing nested parenthesis, and you have
confidence in the correctness of your straight-forward expression.
Well, both Forth and the Forth-like functions use the RPN system for computing.
You bring the power of scientific and engineering calculators to the BASH shell. Well, actually, you will
need to write a few additional functions of your own to have all of the features of your favorite scientific or
engineering calculator, but you have all of the tools available to finish out that task.
This means that you can use RPN calculations in all of your shell scripts.
strings
Both Forth and the Forth-like functions can also do manipulations of characters and character strings.
The Forth-like functions store all information as character strings, even numbers. This is different than real
Forth, but has its own advantages and disadvantages.
BASH and all of the other shells work with streams of characters. Storing everything (even numbers) on a
special string-oriented stack allows our Forth-like functions to play nice with the BASH shell.
This allows us to combine all of the power of the BASH shell with much of the power of Forth.
the power of Forth
The power of Forth lies in three characteristics:
(1) Forth has powerful stack-based processing. The Forth-like functions bring this power to the BASH shell
in a manner compatible with BASH’s expectations.
(2) Forth is extensible. You can build new Forth-like functions by threading together other Forth-like
functions. You won’t have quite the flexibility of real Forth and you won’t have the Forth advantage of all new
Forth words being treated the same as all built-in threaded words. But with modern processors, you will be
pretty close to the power of real Forth.
(3) Forth can run on its own without an operating system. This one can’t be done with the BASH shell. I do
intend to eventually write the code to combine a stripped down shell and Forth as native code for use in
embedded systems. The Forth-like functions are intended as a demo to prove that adding Forth to the shell is a
version 56
page 321
Unix and Linux System Administration and Shell Programming
dramatic improvement in both power and flexiibilty.
Once you learn how to use the Forth-like functions in your shell scripts (and even at the interactive
command line), you will gain a huge amount of added power and possibilities for your shell scripts.
The one big hassle is that you can’t simply import existing Forth programs, which is unfortunate given the
huge library of very useful open source and public domain Forth programs.
I do intend to write a translator, which will convert existing Forth programs into their BASH shell Forth-like
equivalent.
With that introduction to the power of Forth and Forth-like functions, let’s take a look at the actual Forthlike functions.
and now, the functions
dot
The dot function pops a single object from the top of the data stack and echos the result. Forth uses the
period ( . ) for this word, but the period is already used for a different BASH command, creating a namespace
conflict. Hence, the name dot .
NOTE: This function does not yet have error checking. Need to check for empty stack.
$ dot ()
>{
> echo ${DataStack["$DataStackPointer"]} #pop data item from top of
stack and print
> let "DataStackPointer=$DataStackPointer - 1" #postdecrement
>} # end dot
$
DUP
The DUP function is nearly the same functionality as the Forth word DUP (the difference is that this function
works with character strings). The top item on the stack is duplicated, increasing the size of the stack by one.
NOTE: This function does not yet have error checking. Need to check for empty stack.
$ DUP ()
>{
> temporary=${DataStack["$DataStackPointer"]} #make temporary copy of
data item from top of stack
> let "DataStackPointer=$DataStackPointer + 1" #preincrement
> DataStack["$DataStackPointer"]=$temporary #store duplicate
> unset temporary #clear out variable
>} # end DUP
$
qDUP
The qDUP function is nearly the same functionality as the Forth word ?DUP (the difference is that this function
works with character strings). The top item on the stack is duplicated only if it is non-zero, increasing the size
version 56
page 322
Unix and Linux System Administration and Shell Programming
of the stack by one if the item on the top of the stack is non-zero.
NOTE: This function does not yet have error checking. Need to check for empty stack and non-number
object on top of stack.
$ qDUP ()
>{
> if [ ${DataStack["$DataStackPointer"]} -eq 0 ] ; then #test for zero
string on top of stack
> return 0 #if zero, take no action, return OK
> fi
> temporary=${DataStack["$DataStackPointer"]} #make temporary copy of
data item from top of stack
> let "DataStackPointer=$DataStackPointer + 1" #preincrement
> DataStack["$DataStackPointer"]=$temporary #store duplicate
> unset temporary #clear out variable
>} # end qDUP
$
OVER
The OVER function is nearly the same functionality as the Forth word OVER (the difference is that this function
works with character strings). The second item down from the top of the stack is duplicated, increasing the size
of the stack by one.
This function uses all caps, OVER , to prevent conflict with the already existing over UNIX tool for pretty
printing and scrolling source code on a terminal.
NOTE: This function does not yet have error checking. Need to check for empty stack or stack with only
one item in it.
$ OVER ()
>{
> let "TemporaryPointer=$DataStackPointer - 1" #create pointer down
one from top of stack
> temporary=${DataStack["$TemporaryPointer"]} #make temporary copy of
data item one down in stack
> let "DataStackPointer=$DataStackPointer + 1" #preincrement
> DataStack["$DataStackPointer"]=$temporary #store duplicate
> unset temporary #clear out variable
> unset TemporaryPointer #clear out variable
>} # end OVER
$
PICK
The PICK function is nearly the same functionality as the Forth word PICK (the difference is that this function
works with character strings). The nth item down from the top of the stack is duplicated, increasing the size of
the stack by one.
NOTE: This function does not yet have error checking. Need to check that a number was entered. Need to
check for stack that is deep enough.
$ PICK ()
>{
version 56
page 323
Unix and Linux System Administration and Shell Programming
> let "TemporaryPointer=$DataStackPointer - $1" #create pointer down
one from top of stack
> temporary=${DataStack["$TemporaryPointer"]} #make temporary copy of
data item one down in stack
> let "DataStackPointer=$DataStackPointer + 1" #preincrement
> DataStack["$DataStackPointer"]=$temporary #store duplicate
> unset temporary #clear out variable
> unset TemporaryPointer #clear out variable
>} # end PICK
$
The PICK function and the Forth word PICK should rarely be used. If you find yourself using this word often,
then you should refactor your code to be more simple.
DROP
The DROP function is nearly the same functionality as the Forth word DROP (the difference is that this function
works with character strings). Pops a single object from the top of the data stack.
NOTE: This function does not yet have error checking. Need to check for empty stack.
$ DROP ()
>{
> let "DataStackPointer=$DataStackPointer - 1" #postdecrement
>} # end DROP
$
SWAP
The SWAP function is nearly the same functionality as the Forth word SWAP (the difference is that this function
works with character strings). The item on the top of the stack is swapped with the item one down from the top
of the stack.
NOTE: This function does not yet have error checking. Need to check that the stack has at least two items.
$ SWAP ()
>{
> let "TemporaryPointer=$DataStackPointer - 1" #create pointer down
one from top of stack
> temporary=${DataStack["$TemporaryPointer"]} #make temporary copy of
data item one down in stack
> DataStack["$TemporaryPointer"]=${DataStack["$DataStackPointer"]}
#move the top item to 2nd location down
> DataStack["$DataStackPointer"]=$temporary #put former 2nd down on
top of stack
> unset temporary #clear out variable
> unset TemporaryPointer #clear out variable
>} # end SWAP
$
ROT
The ROT function is nearly the same functionality as the Forth word ROT (the difference is that this function
works with character strings). Rotate third item to top of stack.
version 56
page 324
Unix and Linux System Administration and Shell Programming
NOTE: This function does not yet have error checking. Need to check that the stack has at least three items.
$ ROT ()
>{
> let "TemporaryPointer=$DataStackPointer - 2" #create pointer down
two from top of stack
> let "SecondPointer=$DataStackPointer - 1" #create pointer down one
from top of stack
> temporary=${DataStack["$TemporaryPointer"]} #make temporary copy of
data item one down in stack
> DataStack["$TemporaryPointer"]=${DataStack["$SecondPointer"]} #move
the 2nd down item to 3rd location down
> DataStack["$SecondPointer"]=${DataStack["$DataStackPointer"]} #move
the top item to 2nd location down
> DataStack["$DataStackPointer"]=$temporary #put former 3rd down on
top of stack
> unset temporary #clear out variable
> unset TemporaryPointer #clear out variable
> unset SecondPointer #clear out variable
>} # end ROT
$
ROLL
The ROLL function is nearly the same functionality as the Forth word ROLL (the difference is that this function
works with character strings). Rotates the nth object to the top of the stack.
NOTE: This function does not yet have error checking. Need to check for no user input.
$ ROLL ()
>{
> if [ $DataStackPointer -gt "$1" ] ; then #check to make sure enough
items on stack
> let "DestinationPointer=$DataStackPointer - $1"
> SavedItem=${DataStack["$DestinationPointer"]} #save old item
> let "SourcePointer =$DestinationPointer + 1"
> LoopCounter=$1
> while [ $LoopCounter -gt 0 ] ; #move data loop
> do
> DataStack["$DestinationPointer"]=${DataStack["$SourcePointer"]} #move data
> let "DestinationPointer=$DestinationPointer + 1" #post
increment
> let "SourcePointer=$SourcePointer + 1" #post increment
> let "LoopCounter=$LoopCounter - 1"
> done
> DataStack["$DataStackPointer"]=$SavedItem
> unset LoopCounter #clear out variable
> unset SavedItem #clear out variable
> unset SourcePointer #clear out variable
> unset DestinationPointer #clear out variable
> fi # end if good input
>} # end ROLL
$
The ROLL function and the Forth word ROLL should rarely be used. If you find yourself using this word often,
then you should refactor your code to be more simple.
version 56
page 325
Unix and Linux System Administration and Shell Programming
DEPTH
The DEPTH function is nearly the same functionality as the Forth word DEPTH (the difference is that this
function works with character strings). Find existing size of stack and then put that number on top of the stack.
NOTE: This function does not yet have error checking.
$ DEPTH ()
>{
> temporary=$DataStackPointer #save the existing depth of the stack
> let "DataStackPointer=$DataStackPointer + 1" #preincrement
> DataStack["$DataStackPointer"]=$temporary #push depth on top of
stack
>} # end DEPTH
$
archive copy
You can obtain the latest copy of these functions at http://www.osdata.com/programming/shell/forthfunctions
place the file in an appropriate directory.
UNIX and Linux System Administration
and Shell Programming
version history
This is a work in progress. For the most up to date version, visit the website
http://www.osdata.com/unixbook.html and http://www.osdata.com/programming/shell/unixbook.pdf — Please
add links from your website or Facebook page.
56: August 12, 2014. (tail, programming)
55: August 10, 2014. (basename, choice of shells, engineering method, kill, mySQL, programming)
54: July 23, 2014. (touch, LAMP, continue to manually adjust the source file for Adobe Acrobat Pro -- a
task that computers can handle much better than humans)
53: July 14, 2014. (modified for Adobe Acrobat Pro -- for some bizarre reason Adobe falsely claims that
they can’t have both links and proper page breaks in the same document. Obvious bullshit to cover up their
own laziness and/or lack of programming skill -- so have to do the work around of manually handling all of the
page breaks with BR tags)
52: June 24, 2014. (added internal links)
51: June 8, 2014. (classical logic, connecting to a shell, polkit, vmstat)
50: June 6, 2014. (ftp, ls, mkdir, Perl, shell basics, shell levels, Tcl, UNIX history)
49: January 15, 2014. (spelling corrections, mysql, NFS, rm, scripts, shells, sudo)
48: December 25, 2013. (cool tricks, connecting, date, login, shell basics)
47: August 31, 2013. (cat, cp, ls, mkdir)
46: August 28, 2013. (cd, ls, pwd)
45: August 28, 2013. (flow control)
version 56
page 326
Unix and Linux System Administration and Shell Programming
44: August 27, 2013. (php)
43: August 26, 2013. (substitutions)
42: August 21, 2013. (file system basics, df, mv)
41: August 20, 2013. (scripts, cat, cp, ls, wc)
40: August 18, 2013. (tail, scripts, various parts)
39: March 5, 2013. (kernel modules)
38: February 28, 2013. (curl)
37: November 20, 2012. (management tools, signals)
36: October 29, 2012. (minor changes)
35: October 8, 2012. (NFS, processes, kill, nice, critical reasoning)
34: October 3, 2012. (unix hstory; advanced file system)
33: September 28, 2012. (unix history; sendmail)
32: September 25, 2012. (Forth-like routines)
31: September 23, 2012. (Forth-like routines)
30: September 18, 2012. (Forth-like routines, shells)
29: September 11, 2012. (Forth-like routines; lsof)
28: September 10, 2012. (mkdir)
27: September 10, 2012. (Forth-like routines; uname)
26: September 6, 2012. (Forth-like routines; alias; chmod; chown)
25: September 5, 2012. (command summary)
24: September 4, 2012. (Forth-like routines)
23: September 3, 2012. (Forth-like routines; cat)
22: September 3, 2012. (Forth-like routines)
21: September 2, 2012. (cd)
20: September 2, 2012. (built-in commands; computer basics)
19: September 1, 2012. (cool shell tricks)
18: September 1, 2012. (cool shell tricks)
17: August 31, 2012. (cool shell tricks)
16: August 30, 2012. (defaults)
15: August 30, 2012. (less)
14: August 29, 2012. (history)
13: August 28, 2012. (less)
12: August 27, 2012. (test bed; command separator; installing software from source)
11: August 26, 2012. (quick tour)
10: August 25, 2012. (root, sudo)
9: August 25, 2012. (quick tour, shred, init, command structure)
8: August 24, 2012. (sudo, screencapture)
7: August 24, 2012. (tar, command structure)
6: August 23, 2012. (shells)
5: August 22, 2012. (passwd; cat; file system basics)
4: August 22, 2012. (login/logout; man)
3: August 21, 2012. (login/logout)
2: August 21, 2012. (shell basics)
1: August 19, 2012. (imported amd formatted already existing work from the website)
This book includes material from the http://www.osdata.com/ website and the text book on computer
programming.
Distributed on the honor system. Print and read free for personal, non-profit, and/or educational purposes. If
you like the book, you are encouraged to send a donation (U.S dollars) to Milo, PO Box 1361, Tustin,
California, USA 92781.
version 56
page 327
Unix and Linux System Administration and Shell Programming
This book handcrafted on Macintosh computers
using Tom Bender’s Tex-Edit Plus
.
Copyright © 1998, 1999, 2000, 2001, 2002, 2003, 2004, 2005, 2006, 2007, 2009, 2010, 2011, 2012, 2013, 2014
Milo
version 56
page 328
Open as PDF
Similar pages