Why are all the URL's ending with a ".com", ".org",...?

Domains on Internet are used to switch the frames ("messages"). When you type an URL, a frame is built and sent.
Where? To the server (online computer with for example, all the HTML documents) designated by the URL. This server is somewhere in the world, so that the frame has to be driven trough Internet.
We are speaking about HIGH level domains. There are 2 types: generic and geographic. Some domains like the ".org", ".net", ".com", ".edu" are global: we call them GENERIC TOP LEVEL DOMAIN (GTLD). But most of them are in relation with local areas: all these names are made off 2 letters pointing out the unique entry for this country. Ex: ".jp" = Japan. That means that most of the servers of an area are geographically close to each other. But there is no restriction to have an ".be" server in a Belgian embassy in Africa.
You also know that the frames are routed trough the Internet to reach the site you designate with the URL you've typed. At each crossroad, the routers have to choose one exit direction. The right direction is choosen by looking the first right-URL suffix which is the domain and searching in his internal table the direction associated with the received domain.
For each domain, there is a DNS database which will contain all the information to route the frames to the right server. The DNS (Domain Naming System) server translates the URL of the target site into the IP address (unique ID of a machine) of the server containing the required pages. The DNS servers compulse big databases with a lot of information in relation with the naming system. There are master DNS servers on the Internet which are responsible of all top level domain names. You can click here to locate them on the Internet.

Ex: You are connected on a Belgian network ".be" and you want to reach a Japanese server ".jp". The router will search in his table if there is an entry for the IP addresses corresponding to ".jp" domain.
If yes, a direction is associated with ".jp" and the frame is sent to the next router in the indicated direction (For example, Netherlands).
If no, there is always a "default" exit for unkown domain. In fact, you must know that each router have not an entry for each existing domain: it would generate a big table and it means a waste of time at each crossroads.
Usually, a table contains more than 2 entries (localhost and default).

Example for a router in ICELAND (".is") which received an IP address corresponding to a ".cu" suffix:

Table:
- Localhost - North Pole - EUROPE - ASIA - AMERICA - DEFAULT (Rest of World)	->  NO EXIT = my network  -> direction 0  -> direction 1  -> direction 2  -> direction 3  -> direction 4

At the first crossroad, the router will search in his table for a".cu" entry. In fact, it knows that ".cu" is Cuba and that it's in the AMERICA area. Exit direction 3 is then choosen and the frame is forwarded.
It's simply like this!

More info is available through these links:

• My RFC source
• DNS-BE Registration Office
• ISPA
• Beltung
• RIPE RIPE Network Coordination Centre for European areas
• Locate the DNS servers on the Internet

• Netcraft List of all registered domain names
  More than 70,884,595 web sites: .com, .net, .org, .be, .de, .uk, .ru, .lu, .nz, .za, etc
• Some.COM registred domains