The Web

The Web is actually a network itself, but one we call a virtual network, because by itself it has no physical infrastructure to support its connections. Instead it relies on the Internet, as many other systems do, for the delivery of content.

One can think of the Web as a connected set of hypertext (e,g., HTML page) resources that can be retrieved through Universal Resource Locater (URL) hypertext links and viewed in a browser. Thus the network is really a set of resources that are linked together with URLs, such as

http://www.montana.edu

Think of it this way. Each place in a Web page you are viewing that you can click on to visit another page is a hyperlink connecting the current page to the new page. Similarly each <img>, <video>, <style>, and <script> tag that includes a link reference to the associated resource is a hyperlink to that resource. So, if you think of Web pages as nodes and hyperlinks in that page to other resources as links, you can easily see why the Web is referred to as such. You can also see how pages and links tie the entire Web together.

Domain Names

Every computer network needs an addressing scheme for the computers in the network to communicate with each other. The Web as a network is no different. The addressing scheme of the Web is based on domain names (the Web is not the only system that uses domain names for addresses). We are all accustomed to domain names. They have the form

name.com, name.gov, name.org, name.biz, name.us, and so forth

To reach the resources at a domain name, we write something like http://www.montana.edu, which means that we want to send this HyperText Transmission Protocol (HTTP) request to the unique domain name montana.edu on the Internet.

Acquiring a Domain Name

Acquiring a domain name is a simple process. Acquiring a name you might want, not so much. To acquire a domain name you simply visit one of the websites that are in the business of leasing domain names. Among them are godaddy.com and cheapnames.com. They all work in similar fashion. They have a field in which one can type a desired domain name, such as whatsmyname.com. A message returns almost immediately stating whether that domain name is available or not. If it is available, you can reserve it for around $10 per year or less. You are then registered as being the owner of that domain name in a database accessible on the Web.

Subdomains Within a Domain

Once you have a domain name, any subdomain names that can be formed from that domain name by prepending strings of characters separated by periods to that domain name are yours. So, if you wanted whatsmyname.com as your domain name but it was already taken you might find that yname.com is not taken and reserve it. You can then form an address as a subdomain thus:

whatsm.yname.com

You do not need to pay for subdomains.

Acquiring a Server

Having a domain name by itself is not of much use. You have to associate that name with an Internet IP address of a computer on the Web. If you are technically savvy you can set your own computer up with software to make it a server, obtain an IP address for your computer, and then set yourself up on the Web. The usual way for individuals to go about this, however, is to avoid the not insignificant hassle of managing your personal computer as a server. Instead, one obtains an account on an existing server at a company that offers this service as part of their business. To do this, you simply go to the Web hosting site of your choice and follow posted directions on their website for obtaining an account. In the sign-up process you supply your already-acquired domain name as the Web address to be used for your website. The hosting company will check to see that you are the registered owner of that name. It will then provide you with an account on one of their servers. Finally, it will broadcast over the Internet your domain name and the associated IP address of the server on which your account has been established. This will result in all DNS servers updating their domain name and IP address correlation tables with your domain name and your just-acquired server's IP address.

Domain Name Service

A Domain Name Service is a service with a table that is replicated on many computers around the world. The table associates domain names with their IP addresses and vice versa. How a DNS is used is described below.

FTP

You can at this point begin to upload your website files to the server where your website is hosted. This is generally done by a program, such as FileZilla that implements the File Transfer Protocol (FTP).

How the Web Uses the Internet

So now the stage is set. The Web has no physical connection among the servers that host websites. Instead, when a person types a Web address (called a URL, for Universal Resource Locater) into the address bar of a Web browser, such as http://www.montana.edu, the following happens.

• a program on the local computer sends the domain name, montana.edu, to a computer hosting a Domain Name Service (DNS) to retrieve the IP value associated with that domain name. How does the program know where a DNS server is on the Internet? When you connect to the Internet with your computer, among the items automatically configured on your computer for managing the Internet connection are the IP addresses of a few known DNS servers. These can be looked up in the network configuration files on your computer as shown in the following figure. In this case, for example, the program handling the request could first send "montana.edu" to the server 192.168.1.1 as a DNS lookup request.

• Suppose that the IP address associated with domain name "montana.edu" that is returned by the DNS table lookup is 153.90.2.87.
• The program then creates an Internet packet with the destination address in the header of 153.90.2.87.
• The program will fill in the rest of the header values required in Internet packets by the Internet Protocol, including the return IP address (the IP address of this computer), the length of the entire packet in bytes, the number of this packet, the number of total packets for this message, and so forth.
• Finally this program will complete the packet by attaching the packet payload (the information you want to send) at the end of the header, namely the string of characters "http://www.montana.edu". This is such a short message that only one packet will be needed to carry it.
• When this packet reaches the destination computer, the Internet program on that end processes the header and unwraps the packet to determine what kind of information it contains.
• It will discover that the payload is an http packet so it turns the payload contents of the packet over to a different program on the server that handles http requests.
• That http server program checks the request and determines that it contains nothing but a domain name hosted on this computer. So it looks in a particular folder (e.g., www or public_html) for that domain name for a default Web page similar to index.html.
• The index.html page now must be packetized and sent back to the requesting computer (known because its IP address was in the header that came with this packet. Therefore, the index.html page is turned over to a program that creates one or more packets from sequential chunks of bytes that make up the index.html page
• That program also builds a new header for each packet of the index.html page using the original sender's IP address (your computer) as the destination IP address for the return trip.
• The page thus arrives in some number of Internet packets back at the sending computer where it is recognized from header information that this series of arriving packets are in response to an earlier request. The packets are reassembled in their proper order into the index.html page and given to the requesting browser for display.

In all of this the Internet programs and protocols have no reason to be concerned with the contents of the packets, but rather to just deliver each packet as they would any other packet.

In this fashion the Web utilizes the Internet to make its own network connections.