View/print PDF file
	UNIX at Fermilab

Chapter Contents

Chapter 4: Information Resources
  4.1 The Fermilab Helpdesk
  4.2 UNIX On-Line Help
    4.2.1 man Pages
    4.2.2 Finding the Right Command
    4.2.3 Vendor Product Documentation
  4.3 The Internet
    4.3.1 The World Wide Web
    4.3.2 UNIX Help on the Web
    4.3.3 Newsgroups
  4.4 The Info Utility
  4.5 Other Users: WWW Directories, finger and who

Chapter 4: Information Resources

---

This chapter introduces you to the information resources available to UNIX users. We include instructions for communicating with the Fermilab Computing Division Helpdesk. Standard UNIX on-line help is available via the man pages. We discuss the World Wide Web and newsgroups, which are very rich sources of information on a virtually unlimited set of topics. A few utilities that allow you to get information about vendor products and other users are also covered.

4.1 The Fermilab Helpdesk

The Fermilab Computing Division Helpdesk is available to answer questions related to the supported computer systems and software on site. We encourage you to send all your computer-related questions directly to the Helpdesk for tracking and logging purposes.

Keep in mind that the first priority of the Helpdesk is to maintain central systems and networks, and to ensure that Fermilab-supported software is available and usable. Therefore a request which impacts only one individual may not receive immediate attention.

The Helpdesk is in service Monday through Friday, 9:00 a.m. to 5:00 p.m. You are encouraged to use the Web interface or email for all communications.

Helpdesk Email Address

During business hours:

helpdesk@fnal.gov

During off-hours:

operator@fnal.gov

Helpdesk Web Page

You can request help and/or keep track of actions taken on your requests from this page. Go to http://csdserver1.fnal.gov/HelpDesk/cd/.

Helpdesk Phone Number

630-840-2345

During off-hours, you can leave a phone message, or "escape" to Data Center Services (Operations) for requests requiring immediate attention.

Helpdesk Location

Wilson Hall, 8th floor, northeast corner office.

4.2 UNIX On-Line Help

4.2.1 man Pages

On-line help for UNIX system commands and utilities is in the form of man pages (man stands for manual) which consist of an on-line version of the UNIX documentation set (often called the UNIX Programmer's Reference Manual). You access the man pages with the man command.

Note that the man pages differ in many instances between UNIX platforms.

The man Command

When you need help on a known command, use the general man command format:

% man [<part>] <topic>  

where <topic> is generally a UNIX command. man is really the on-line manual which is divided into several parts. <part> is a digit between 0 and 9. If you know in advance which part contains the information you want, you can speed the search by specifying it. More often than not you will just enter:

% man <topic> 

The word print in man entries usually means display on the screen. Don't be confused by this. Several options are available with the man utility, described under man man. The man command normally displays complete manual pages that you select by name. One-line summaries can be selected by either by keyword (-k option), or by the name of an associated file (-f). These options are described in section 4.2.2 Finding the Right Command.

A typical initial man screen can be seen by issuing the command:

% man ls 

where ls is the UNIX command to list files in a directory.

ls(1) User Commands ls(1) NAME ls - list contents of directory SYNOPSIS ls [ -abcCdfFgilLmnopqrRstux1 ] [ names ] AVAILABILITY SUNWcsu DESCRIPTION For each directory argument, ls lists the contents of the directory; for each file argument, ls repeats its name and any other information requested. The output is sorted alphabetically by default. When no argument is given, the current directory is listed. When several arguments are given, the arguments are first sorted appropriately, but /tmp/mpa002Zf

Man pages are typically formatted with the UNIX text processing utility nroff (or groff). These utilities are covered in most UNIX texts. If you find that the man page is unformatted, run setup groff, and then rerun the man command.

Note that built-in shell commands are described under the topic corresponding to their shell.1 (See section 6.1.1 Programs, Commands and Processes for information on built-in commands.) For example, to get information on the command alias for your current shell, you would enter man <shell> (e.g., man bash) and search there for information on alias using the /<pattern> function described below.

There is an alternate, "quick and dirty" method to verify the format of a command and get a listing of its options. It doesn't work with all commands, but is usually worth a try. Simply enter the command with an illegal option (try / or ? or .). For example,

% ls -/ 

will produce the output:

ls: illegal option -- / 
usage: ls -RadCLHxmnlogrtucpFbqisf [files]  

The man Command for AFS Commands

For AFS commands, it works a little differently. AFS commands are discussed in section 8.4 AFS File System Commands. The man page for an AFS command is found by entering:

% man fs_<command> 

Note the underscore (_) between fs and the rest of the AFS command; the underscore is only used with the man command.

Manipulating man Pages

man displays the information using your $PAGER environment variable, which under FUE is set to less (see section 6.4.4 Filters). Therefore, man pages are normally piped to less. The command man less will give you more information about manipulating man via the less filter. less gives you one page at a time and lets you enter commands at the prompt to control what it does after each page. For example, you page forward with the Spacebar and page backward using b.

Once in the man environment, you can search for patterns by entering the /<pattern> option at the command line. The first instance of the string <pattern> will appear in the top line of the screen. To find additional instances of the pattern in the text, simply enter a slash (/).

To exit from man enter q (for quit).

Printing man Pages

To print a man page, you can use the pipe feature (|) along with recommended print formatting and printing commands. These features are covered in Chapter 11: Printing. As a suggestion, pipe the output of the man command to a2ps -m (to convert man pages to PostScript format) and then pipe that output to the print command flpr:

% man <command> | a2ps -m | flpr [<options>] 

This formats the output nicely in landscape, two pages to a sheet.

4.2.2 Finding the Right Command

If you don't know exactly what command you need, try the -k option with a keyword.

% man -k <keyword>  

This displays the man page name, the section number in the UNIX documentation, and a short description for each man page whose name line contains <keyword>. For example to find a search utility, enter:

% man -k search 

The system will output several records similar to the following:

conflict(8) - search for alias/password conflicts /usr/local/lib/mh/conflict [-m glookbib(1) - search bibliographic databases glookbib [ -v ] [ -istring ] [ -tn ident(1) - identify files ident [ -q ] [ \&.\|.\|. ] ident searches for all occu lkbib(1) - search bibliographic databases lkbib [ -v ] [ -ifields ] [ -pfilename lsearch(l) - See if a list contains a particular element lsearch ?mode? list pat lsearch(n) - See if a list contains a particular element lsearch ?mode? list pat zgrep(1) - search possibly compressed files for a regular expression zgrep [ gre .. ..

Some UNIX systems have an additional utility, apropos, which can be used to locate commands by keyword lookup:

% apropos <keyword>  

apropos <keyword> is equivalent to:

% man -k <keyword>  

The -f <filename> option for man prints the manual entry summaries which might pertain to the given filename(s). Any leading pathname components are stripped from the filename before the filename is matched against the summaries. Here is an example using the -f option, followed by the output:

% man -f /etc/passwd  

passwd (1)              - change login password and password attributes 
passwd (4)              - password file 

The summaries are gotten from the whatis database. You can run whatis command(s) to look up a given command and obtain the header line from the manual section. You can then run man to get more information on the command. If the line starts name(section)... you can do man section name to get the documentation for it.

4.2.3 Vendor Product Documentation

Most vendor product documentation is now available on-line:

Linux

info (Or go to http://www.redhat.com/docs/)

IRIX

insight

Solaris (SunOS)

answerbook (only usable on Sun workstations at Fermilab)

4.3 The Internet

The Internet is a global network of networks that provides access to hundreds of thousands of computers around the world. As the reach of the network has grown, so has the number of services accessible. The main tools that allow the user to navigate through the Internet are:

Web browsers

to browse the World Wide Web (see section 4.3.1 The World Wide Web)

Email

to send electronic mail (see Chapter 12: Email on UNIX)

telnet, ssh

to log into remote hosts (see section 3.3 Connecting from One Kerberized Machine to Another)

ftp

to send/retrieve data files to/from remote hosts (again see section 3.3 Connecting from One Kerberized Machine to Another)

News

to scan internet news groups (see section 4.3.3 Newsgroups)

There are two ways to reference an Internet host: an alphabetic name and a series of numbers. The alphabetic sequence is called the host name (e.g., fsui02.fnal.gov) and the numeric one is called the IP address (e.g., 131.225.18.178). At Fermilab all host names end with fnal.gov, where this suffix is called the domain name. Since hosts may change their IP addresses, it is a good practice to always use the host name.

An introduction to making the Internet work for you is The Whole Internet: The Next Generation, published by O'Reilly & Associates.

4.3.1 The World Wide Web

Our primary way of delivering information to you is via the Web. There are a number of reasons why the Web has become the defacto standard for information delivery at Fermilab, within the HEP community, and in the world:

• The Web is very strong at pulling together information from different places, and of different formats and types (native HTML2, PostScript, newsgroups, and so on), and making it appear as a seamless whole.
• It is relatively easy to make information available on the Web. See The World Wide Web at Fermilab.
• Web browsers are available for every supported platform at Fermilab, making it an excellent fit to the distributed environment in which we now work.

Accessing the Web

Web browsers work best on workstations that support graphics, so assuming you have a graphics terminal, try either Netscape or Mozilla. To invoke them, enter:

% netscape [&] 

or

% mozilla [&]  

The ampersand (&) is discussed in section 6.4.2 Standard Input and Output Redirection. In order to use an X application (which these products are), you must have defined your $DISPLAY variable correctly. See section 9.6 Some Important Variables.

If you don't have a graphics terminal, then you can use the line-mode program, lynx. To invoke it, enter:

% lynx 

This section is not intended to provide detailed instruction on the use of any particular Web browser. Once you get any of them running (at least the graphical ones) there is more information than you will ever want available under Help. Or go to http://www.google.com and search. In addition, at http://www.fnal.gov/docs/products/www/mailcap_mime.html will find information on the files that control what filetypes the browser recognizes. These files are called .mailcap and mime.types.

Web Basics

Web browsers find information based on URLs (Universal Resource Locators) which are like addresses and which take you to a Web site, the top page of which is often called a home page.

Each home page can have several layers or pages that it links to, thereby creating a whole Web site. But the home page is generally the first place you would look. It is like looking at the cover of a book and its table of contents at the same time.

The native WWW addresses are of the form: http://address/ or http://address/something.extension The first part is the protocol, http in this case. (A protocol is a set of rules computers observe to exchange information. http stands for the HyperText Transfer Protocol; think of it as the identifier of a Web page.) Instead of http you might also see https, ftp, file, gopher, mailto, and news. Next you'll see a colon and two forward slashes (except for mailto, which has a different format). Next comes the computer's address in the format described in section 4.3. A computer address, or domain name, is used by computers in routing data across the many networks that make up the Net. Finally, you often see a directory path at that computer plus a file at the end of the path. Web page files usually end in "html", for HyperText Markup Language, although you may also see extensions such as "htm", "shtm", "shtml", or "asp". HTML is the simple scripting language that tells browsers how to display the various elements of a Web page such as links, body text, header text, inline graphics, and external files.

Many HTML files contain links to other documents. Sometimes links are text; sometimes they are images. If a link consists of text, it is underlined and may be in color. You can tell your cursor is on a link when a URL appears at the bottom of the screen. If a link consists of an image, you'll see a URL when you move your mouse pointer across it.

Several Web search engines can help you find information you are seeking, and they vary in the number of URLs they contain in their database, how deep they go into Web sites indexing information, what they index, and how frequently they "crawl" or "walk" the Web in surveying sites. There's a valuable Swiss site at URL http://cuiwww.unige.ch/meta-index.html that collects several WWW search tools. A favorite one is http://www.google.com.

Subject directories make it increasingly easy to find information about broad subjects. An excellent directory is one called Yahoo at URL http://www.yahoo.com/, a well-constructed directory of hundreds of thousands of Web pages.

Writing your Own Web Pages

If you want to start writing Web pages in HTML format, see the Web link on the Computing Division home page. Look under Web Publishers at Fermilab and Web Page Design. The Computing Division provides templates that we ask you to use for official Fermilab web pages.

A note for Web page providers on AFS systems: Set the permissions for system:anyuser to rl on directories containing files that you want to make accessible via a Web browser. See section 8.6 File and Directory Permissions for information on AFS directory permissions.

4.3.2 UNIX Help on the Web

The man pages can sometimes be cryptic, unwieldy, or both. As an alternative, set a bookmark in your Web browser to UNIXhelp for Users at http://www.geek-girl.com/Unixhelp/. Here you'll find easy-to-follow instructions on the use of many UNIX features, organized into four categories: Tasks, Commands, Concepts, and Utilities.

Also look at The UNIX Reference Desk at http://www.geek-girl.com/unix.html. The resources in this document are divided into the following classes: General, Texinfo Pages, Applications, Programming, IBM AIX Systems, HP-UX Systems, Unix for PCs, Sun Systems, X Window System, Networking, Security, Humor.

4.3.3 Newsgroups

Usenet News (or NetNews) is a way of communicating "articles" among people world-wide. In general, information in newsgroups is volatile information, whereas information in Web pages is of longer term. We have a server here at Fermilab which receives articles from elsewhere and posts the articles originating here. Fermilab has its own newsgroups named fnal.xxx. CERN's groups are prefixed with cern and SLAC's with slac. General information and especially important information is posted in fnal.announce. NALCAL, seminar announcements, and the like are posted in fnal.announce.seminars. UNIX discussion articles are posted in fnal.comp.unix. There are many more newsgroups, both Fermilab groups and others, that you might find of interest.

In order to make a newsreader available on UNIX, enter (or put in your login script):

% setup news 

A number of readers will be made available to you. Line-mode browser commands are nn, rn, and trn. X-based browser commands are xrn and knews. News may also be read from most WWW browsers and some mail readers. The readers keep track of the newsgroups that you are interested in ("subscribed to") as well as which articles in each newsgroup you have read; all of the UNIX readers cooperatively maintain this information, so you can use different readers at different times without losing this information.

4.4 The Info Utility

Info is a facility available to the system support people to communicate with you about events regarding the Fermilab computing systems (shutdowns, for example), or other systems-related information that is newly available. To get a list of the Info messages, enter the command Info. To read an item, enter the command with the nametag of the item found on the left side of the Info list:

% Info <nametag>  

Note the capital I! If the item is more than one page, press the space bar to continue. Press q to quit.

4.5 Other Users: WWW Directories, finger and who

WWW Directories

From the Fermilab at Work page, directories are available to point you to information about Fermilab personnel and the high energy physics community at large. These directories typically contain general information such as email addresses, phone numbers, and office locations. For locating Fermilab employees, see the Fermilab Telephone Directory at http://www-tele.fnal.gov/telephone/.

finger

finger is used to find out about other users. It searches for matching account names and first or last names, if known. Depending on the vendor implementation, it may display the name of the person associated with each account, the login name, the home directory and login shell, the contents of the file .plan in the person's home directory, and possibly other information such as waiting mail, and time of last login. If the person is logged in, it also may display information about his or her current sessions.

Note that each vendor has a different implementation of finger. In addition, for security reasons many sites disable the output of finger over the network. It is therefore unwise to rely on the format, content, or even the availability of finger as a tool for finding out about users or their accounts.

The format of the finger command is:

% finger [<options>] [<name>...]  

where <name> can be a part of a personal name or a username. If you specify the option -m, then <name> is matched only to account name and not the rest of the personal name.

We encourage you to create a .plan file. It is just a text file in which you might include information such as your office location, phone numbers, mail station, home address, schedule, emergency numbers, and so on.

finger can often be used to look up users on a remote machine by specifying the name in the standard internet form user@host. This form works on VMS machines with MultiNet running, but in this case <name> must be the username; otherwise not much useful information is obtained.

who

The command who lists certain information about the users on your system.

% who  

If used with the -q option, only the names of the logged in users and the number of users are displayed.

The who am i form identifies the invoking user. The command format is:

% who am i  

There are a number of options which you can read about in the man pages.

1

Some platforms provide man pages for built-in commands, however in general you may find it easier to look in a reference book!

2

HTML stands for HyperText Markup Language, the standard language for documents accessed on the Web.

View/print PDF file

CRL Product Home Page | CRL Manual Home Page | Computing Division | Fermilab at Work | Fermilab Home