The man page reading club: man(1)

This post is part of series

I have been using Linux as my main operating system for the last 14 years, and as my only one for the last 10. Needless to say, I really like it. However, at the end of last year I have decided to use OpenBSD for my VPS. I had 3 main reasons (spoiler: security is not one of them):

As I was expecting, after some time using it I was really pleased with the quality of OpenBSD’s manual pages, to the point that I am sometimes using them (by ssh-ing into my server) also when I am working or writing scripts for my Linux machine.

So I thought that it would be nice to take once in a while one of these manual pages, read it, and share what I learned with other people. Hence I decided to start this blog-series: “The man page reading club”.

Nuclear Apocalypse

I tried to find a nice setting for this series, a funny story to introduce each of these technical posts, but I could not come up with any good idea. Luckily, a couple of weeks ago I read the short post OpenBSD is the Perfect OS post Nuclear Apocalypse and I found it to be exactly what I was looking for!

So here you are, in a cozy underground bunker during a Nuclear Apocalypse. There is enough food and electricity for everyone, but of course no Internet. After a couple of days of social chit-chat with your new roommates you find out that you have pretty much nothing in common with them, so you turn to your trusty laptop and you install OpenBSD 7.1 - the last version released before the apocalypse.

From: (Theo de Raadt)
Subject: Welcome to OpenBSD 7.1!

Says your screen, after you login and type mail.

For more information on how to set up your OpenBSD system, refer to the
"afterboot" man page.

The what?

If you are not familiar with how to read man pages, type "man man"
at a shell prompt and read the entire thing.

Sounds like a good idea, let’s do it!

$ man man


Follow along at

The first interesting thing we see in this manual page is a thing called synopsis:

    man [-acfhklw] [-C file] [-M path] [-m path] [-S subsection]
        [[-s] section] name ...

This explains how to invoke the man command. The parts within square brackets are optional, and the other words such as path and section are going to be explained later in this page.

A short description the tells us that man is used to display manual pages. It is followed by a detailed explanation of what each option does. Let’s look at some of them!

-c  Copy the manual page to the standard output instead of using
    less(1) to paginate it.  This is done by default if the standard
    output is not a terminal device.

This option is not very interesting to use because of the last sentence: it is the default behavior when you need it (for example, when you pipe the output to another program). It is however interesting to read its description, because it reveals a little detail of how man works. Namely, that every time we open a man page we are also calling less!

-h  Display only the SYNOPSIS lines of the requested manual pages.
    Implies -a and -c.

This can be useful for commands with a more complicated syntax.

-k  A synonym for apropos(1).  Instead of name, an expression can be
    provided using the syntax described in the apropos(1) manual.
    By default, it displays the header lines of all matching pages.

This sounds like some kind of search function, like a post-nuclear Google. Indeed, man apropos tells us:

    apropos, whatis - search manual page databases

The page apropos(1) is worth a read, but I won’t write a dedicated post. This utility supports searching by regular expressions and by “mdoc macros”. If you don’t know what they are, worry not: it is all well explained in apropos(1). The page also contains a lot of examples, which is always appreciated.

Now back to man man!

[-s] section
    Only select manuals from the specified section. The currently
    available sections are:

        1   General commands (tools and utilities).
        2   System calls and error numbers.
        3   Library functions.
        3p  perl(1) programmer's reference guide.
        4   Device drivers.
        5   File formats.
        6   Games.
        7   Miscellaneous information.
        8   System maintenance and operation commands.
        9   Kernel internals.

Manual pages are divided into sections: they are the mysterious numbers in parentheses after a command’s name, like in less(1). There can be pages with the same name in different sections, and with this option you can specify which one you want. For example man printf shows the printf(1) page, for the UNIX command with the same name, while man -s 3 printf shows the manual page for the printf() C library function. The syntax for this option is slightly unusual in that the -s itself is optional: man 3 printf does the same as man -s 3 printf.

One might be tempted to think that if no section is specified, the page in the lowest-number section is shown. However:

Within each directory, the search procedes according to the following list
of sections: 1, 8, 6, 2, 3, 5, 7, 4, 9, 3p. The first match found is shown.

The man program also behaves differently depending on the value of some environment variables. For example

MANPAGER Any non-empty value of the environment variable MANPAGER is
         used instead of the standard pagination program, less(1).

PAGER    Specifies the pagination program to use when MANPAGER is not
         defined.  If neither PAGER nor MANPAGER is defined, less(1) is

It looks like the variable MANPAGER is read only by man, while PAGER may be understood by other commands as well. But how do we set environment variables? One way to do this is shown in the EXAMPLES section:

Read a typeset page in a PDF viewer:

    $ MANPAGER=mupdf man -T pdf lpd

Now you should complain, because I did not tell you about the -T option! In fact I skipped this line:

The options -IKOTW are also supported and are documented in mandoc(1).

A quick look at the mandoc(1) page tell us that the most interesting of these options are -T, which we have just seen, and -O, which allows to tune some formatting settings. For example man -O width=50 man shows you a man page using only 50 columns, while man -O tag=EXAMPLES man shows you the man(1) page, but starting from the EXAMPLES section (you can still scroll back up to the beginning of the page).

Back to man man.

    The man utility is compliant with the IEEE Std 1003.1-2008 ("POSIX.1")

    The flags [-aCcfhIKlMmOSsTWw], as well as the environment variables
    MACHINE, MANPAGER, and MANPATH, are extensions to that specification.

This means that all of the stuff we read in this manual page, except for the very basic functionality of man - showing manual pages - is not POSIX-standard, and might be different on other systems. Be sure to check what your version of man does before assuming that it will behave the same as the OpenBSD one! I find it very nice that the OpenBSD man pages always tell you which parts are standard (and thus you can expect to work in the same way on other UNIX-like OSes) and which are OpenBSD-specific. It makes writing portable scripts much easier!


man is a straightforward utility and most of the time you are just going to use it by typing man command. However, reading this page I was still able to learn new stuff - such as the -h and -O options and the fact that apropos supports searching by regular expressions and tags. I hope you have learnt something new too.

As you can see I have skipped a lot of things, including all the parts related to the MANPATH. This post does not want to be a comprehensive tutorial on the man command, just a survey of the subjectively most useful parts of the manual.

For the next post I will either take one of the pages that was referenced here, such as less(1), or dive into more exciting stuff with something like sh(1).

Stay tuned!

Update 2022-09-05

In a more recent blog post I comment on a part of the manual that I skipped, although it is actually useful for writing these blog entries:

When using -c, most terminal devices are unable to show the
markup.  To print the output of man to the terminal with markup
but without using a pager, pipe it to ul(1).  To remove the
markup, pipe the output to col(1) -b instead.

If you are curious, check the new post!

Next in the series: more(1)