Dear Straight Dope:
What's with the .com/.org/.net at the end of internet addresses? Who decides what suffix each address receives?
SDStaff Karl and Little Ed reply:
The suffixes don’t necessarily mean anything, and you assign them. Or at least your Internet service provider (ISP) does. It’s done that way in case there’s a nuclear war. Seriously. Well, semi-seriously.
Here’s the deal. The Internet was originally a U.S. Department of Defense project. The DoD wanted the net to remain usable in wartime, even if major portions of it were destroyed. The solution was to keep it decentralized. The people who own different pieces of the net maintain their chunk and work cooperatively with their peers to keep the whole thing going. There’s no headquarters, no operations center. No one is in charge.
Nuclear war is less of a threat these days but decentralization still serves the net well. If a major node or trunk goes down (power failure, moron with a backhoe, whatever), the computers automatically route around it.
Decentralization also makes administration of the net much simpler — indeed, it’d be impossible otherwise. Assigning of Internet addresses is the classic example of this.
There’s no central “Internet phone book.” How could there be? With thousands of new users registering every day, no central registry could hope to keep up.
Instead, the universe of net users is divided into a hierarchy of domains, sub-domains, and sub-sub-domains. Each level keeps track of the levels beneath it. Originally the top-level domains were:
- .com — commercial enterprise
- .org — not-for-profit organization
- .net — entity involved in net infrastructure, such as an ISP
- .mil — U.S. military
- .gov — U.S. government (federal)
- .edu — college-level educational institution
Since this was pretty U.S.-centric, country domains such as .uk, .fr, and .de (Deutschland=Germany) were added later.
Who decides whether you’re a dot-com or a dot-org or whatever? Well, various registrars handle the busywork of assigning addresses, making sure there are no duplicates, etc. The .mil domain is overseen by the Department of Defense Network Information Center (http://www.nic.mil), the .gov domain by the General Services Administration (http://www.nic.gov), and the .edu domain by a private company called Network Solutions (http://www.networksolutions.com). In the last case, you have to prove to them that you’re a four-year accredited educational institution — if you do, you’re assigned an .edu subdomain for free. International domains commonly are overseen by each country’s “network information center” at www.nic., e.g., http://www.nic.de, the registrar for Germany.
New subdomain names in the .com, .org, and .net domains are registered by Network Solutions and several other companies. (See http://www.internic.net for a list of registrars.) These firms are simply acting as agents of the Internet gods and have no authority of their own. If you decide you’d rather be a dot-net than a dot-com, that’s strictly up to you when you register. Nobody checks. Many people register in all three domains to avoid spoofing by wiseguys.
Once you register your subdomain name, you become the registrar for any sub-subdomains you care to establish. That’s the beauty of it, see? The domain registrars parcel out the subdomain names, the subdomain registrars parcel out sub-subdomain names, and so on down the line. The names of the levels are separated by dots. Thus in division.company.com, “division” is a part of “company” which is a part of the top-level domain “com.” If division.company.com has a Website, it may be identified by creating yet another level in the hierarchy, www.division.company.com, but the “www” is optional. For example, www.straightdope.com will take you to the online Straight Dope. But we set it up so that straightdope.com will take you there too.
Actually locating a particular computer on the net is also a decentralized process. If one computer is to talk with another on the Internet, each computer has to know the IP (Internet Protocol) address of the other. An IP address is simply a number, which computers find easier to deal with than words. The Domain Name System (DNS) translates domain names into IP numbers. Thus, as I write, www.yahoo.com translates to 220.127.116.11. Put “http://18.104.22.168” into your web browser and Yahoo should come up.
Translating domain names into IP numbers is done by computers called DNS servers. When you type in an address, your computer talks to a DNS server to find out the IP number assigned to the address’s domain name. If the DNS server doesn’t know, it asks another DNS server that does. If it doesn’t know who to ask, it talks with one of the so-called “root” DNS servers, asking something like, “What DNS server knows about .com (or .org, or .uk, or whatever)?” Then it asks the “.com” server, “what server knows about yahoo.com?” Then it can ask the yahoo.com DNS server (which is run by Yahoo) about the IP number for “www.yahoo.com.” Finally you get your answer. Sounds complicated but in practice it’s quick, and it puts the burden of keeping track of a particular Internet address on the party best equipped to do so, namely whoever assigned it in the first place.
The top-level domains .com, .org, .net and so on were assigned in the infancy of the Internet and many feel they’re due for an overhaul. The rest of the world is upset that the U.S. gets so many top level domains while other countries only get one each. Businesses are upset that their favorite names are already taken, and they want more top-level domains, like .store, so they can get their piece of the pie. The buzz in the technical community is that there’s no way the current system is going to remain a useful way to find a particular Website out of the gazillions that will eventually be established. The folks who decide these things are technical types, but the issues are, shall we say, politically charged. Stay tuned for the next exciting episode.
If you want to find out more about the DNS wars, you might start by looking through the archives at Wired (http://www.wired.com). For a sense of the current state of the debate, see http://www.opensrs.org/icann/cindex.shtml. No doubt there are other Websites out there that are all over this topic, for those of your brave enough to find them. Thrill seekers should dig up the story of the man who hijacked the root DNS servers for a day.
Send questions to Cecil via firstname.lastname@example.org.
STAFF REPORTS ARE WRITTEN BY THE STRAIGHT DOPE SCIENCE ADVISORY BOARD, CECIL'S ONLINE AUXILIARY. THOUGH THE SDSAB DOES ITS BEST, THESE COLUMNS ARE EDITED BY ED ZOTTI, NOT CECIL, SO ACCURACYWISE YOU'D BETTER KEEP YOUR FINGERS CROSSED.