Decoding Internet Addresses

Computer Life column for May 9, 1998
by
Richard Gordon

Out of the blue, I've received about a half dozen questions about interpreting Internet addresses over the past two weeks.

Most simply, in order to find computers connected to the Internet, the people who use them, and the information stored on them, everything and everybody on the Net has to have a unique address.

Every last one of the 50 million computers connected to the Internet has to have its own unique numeric address. Each of these "IP Numbers" has four parts, separated one from the other with a period. For example, one of the computers I use at work has the address 128.175.8.50.

Just as phone numbers have area codes and exchanges to help organize them, so these IP numbers are organized hierarchically.

For example, 128.175 indicates that my computer is at the University of Delaware. The next part, 8, indicates to which UD local network my computer is connected. And the last part, 50, uniquely identifies my ancient Macintosh.

Most IP addresses also have alphabetic names to make it easier for humans to find them. For example, my Mac is also known as elk.us.udel.edu.

These alphabetic addresses are also hierarchical, this time from right to left. "Elk" is a computer on the "us" network, located at an organization called "udel" that is part of a group "edu." The .edu part is one of several "top-level domain names" that help identify the country or type of organization in which the computer or network is located: .edu, educational; .com, commercial; .ca, Canada; .us, United States; and so on.

If you want to send me e-mail, you have to know my address. Under ordinary circumstances, an e-mail address is a person's individual account code, an @ symbol, then either the "domain name" of the organization where the e-mail gets processed, or the exact name of the computer where the e-mail gets read. For example, my personal e-mail address is richard@inet.net.

Several of my recent questions related came from people who unintentionally have multiple addresses. For example, I actually use a computer named ns1.inet.net to process my e-mail. If I join a mailing list using richard@inet.net as my address but my e-mail actually appears to come from richard@ns1.inet.net, my contributions to that list may get rejected.

As he reads this part, I can just see my father's eyes glaze over.

But this is a common problem for people using some of the big Internet Service Providers. If you subscribe to a mailing list as johndoe@abc.com, how are you to know when an ISP adds a "subdomain name" to an outbound address (e.g., johndoe@mail1.abc.com)? To cut down on unsolicited advertising, many mailing lists only allow postings from subscribers' exact e-mail addresses. So, "johndoe" may need to make special arrangements with a list-owner to allow for his e-mail configuration.

Every last file of information on the Internet has an address, often called a URL, a Uniform Resource Locator. For example, when it goes on line, this column's full address will be http://inet.net/~richard/NJ/19980509.html.

The first part of the address, http, indicates that this is an open Web document. If it were a secure web document, that part would be "https", indicating precautions were in place to safeguard the privacy of information you might provide.

After the colon and double slashes comes the name of the computer where the information is stored. Then comes the full "pathname" of the file: the directory (~richard), subdirectory (NJ), and filename (19980509.html).

A Shortcut

If you've stayed with me this long, you deserve a shortcut.

You can reach many commercial home pages or Web starting places with just one part of the full URL.

For example, try telling Netscape or Internet Explorer to go to "yahoo" or "amazon". You'll find that the programs fill in the file name and the http stuff automatically and will automatically connect you with "www.yahoo.com" and "www.amazon.com".

Copyright © 1998, The News Journal Company

Computer Life Index