For many, contemplating a move into eCommerce for the first time can be bewildering simply because of the terminology used. The underlying basic concepts of web sites and the Internet, however, are remarkably straightforward once this terminology is understood. This document attempts to explain some of the most commonly encountered terms:

ISP

Attaching directly to the 'real' Internet requires some very expensive communications equipment and communications lines, well beyond the reach of most commercial enterprises. The connection is therefore made by a 3rd party company, called an Internet Service Provider, or ISP. When you connect to the Internet, you are in reality connecting to the ISP's network, which then transparently routes you through to the Internet using their communications equipment.

Web Server

When you request a web page from a web site on the Internet, that page is delivered to you by the web server; in essence, the web server is the computer that is attached to the Internet that contains your web pages and sends them over the Internet to anyone that requests them.

In the same way that connecting your own computer or network directly to the Internet is impractical, so is hosting your own web site on your own network. A web server needs to be accessible 24 hours a day and to have a high-speed reliable connection to the Internet. For these reasons, web servers are normally hosted by an ISP at their own premises.

Note that the ISP that provides a web server to host your web site does not necessarily need to be the ISP that provides the Internet connection to you or your company.

Broadband

Broadband is the generic term for a range of technologies that provide relatively high-speed connections to ISPs.

ADSL

ADSL is a specific type of Broadband technology that uses high-frequency data signals "piggy-backed" onto a normal analogue telephone line. The 'A' in ADSL stands for Asymmetric, which means that the download speed is different from the upload speed. In practice, this means that you receive data from the Internet at a much higher speed that you send it. For web browsing, this is perfect, as you only send a small amount of data, requesting a web page, for example, but receive a larger amount of data - the web page itself, containing large quantities of graphics, etc.

This is another reason why it is impractical to host a web site on your own network, since a web server sends out much more data than it receives and, using an ADSL connection, this would be extremely slow.

Server Farm

An ISP that specialises in providing web servers will host those servers in a purpose-built facility known as a Server Farm or Data Centre. As well as having extremely high-speed Internet connections into the Server Farm, it will typically provide very high levels of fault-tolerance, in terms of redundent mains power supplies (often from different local sub-stations), along with Uninterruptable Power Supplies and diesel generators, redundent Internet connections (often from different local exchanges) and highly efficient air conditioning.

The buildings that house Server Farms are usually physically highly secure as well, without windows and with ant-ramraid bollards, 24 hour security, and so. This is particularly important where web servers are processing eCommerce transactions since the theft of a web server could potentially include the theft of vital credit card data.

IP Address

Every connection to the Internet must have a unique IP (Internet Protocol) address so that data can be sent to the right place in the world. These IP Addresses appear as four sets of numbers, separated by dots, so are of the form 123.123.123.123.

All data sent from one place to another across the Internet is addressed to a specific IP address and from a specific IP address.

So, for example, when you request a page from a web site, your computer needs to know the IP address of the web site to know where to send the request and, conversely, the computer that is running the web site needs to know the IP address that the request came from in order to send the web page back.

Every IP address must be unique across the entire planet and since each of those four numbers can only go up to 255 there are only 4 billion possible IP addresses. This may sound a lot but they are most definitely a finite resource.

Dynamic IP Address

If you have a domestic Internet connection, or a budget commercial Internet connection, for your own computer(s) to connect to the Internet, then your own IP Address will be allocated to you each time you connect to the Internet. Since IP addresses are a relatively expensive commodity, a mass-market ISP probably has more customers than it does IP addresses. Each time you connect, it allocates a free IP address from the pool of free ones; when you disconnect from the Internet, the IP address that you are using goes back into the pool. The upshot of this is that your IP address may be different today to what it was yesterday.

Fixed IP Address (or Static IP Address)

The draw back with Dynamic IP addresses is that it makes it impossible for the receiving computer to know anything about the sending computer, because there is nothing unique about it's IP address. This is fine if the requestor is just asking for a publicly-available web page, since the web server does not care who the requestor is. Whenever security or restricted access becomes an issue, it is therefore essential for the recipient to know who the requestor is, which it can only do if that requestor always has the same IP address.

For example, if you have a system whereby your computer regularly updates data in a database on a web server (for an ecommerce solution, for example), you would not want anyone being able to update that database and a simple user id/password block would not provide adequate protection. The web server would therefore be configured to only accept database updates from your IP address. Naturally, this cannot be done if you have a Dynamic IP address, which changes every day. Your connection to the Internet must, in these circumstances, always have the same IP address. This is achieved by having the ISP allocated a Fixed or Static IP Address so that you always use the same one.

Web servers must, by definition, always have a fixed IP address since anyone browsing to that web site needs to be able to find it.

Furthermore, if anyone is to initiate a connection to your network then they need to be configured with your IP address and so, again, your network would need to have a fixed IP address; this would include any form of remote access (such as home workers or mobile workers), distributed email, etc.

Domain Names

IP Addresses are not particularly friendly for humans to deal with. The BBC's web site address is, for example, 212.58.251.195 but it would not be very helpful if the BBC announced "for further information, please visit our web site at http://212.58.251.195"!

A Domain Name, such as bbc.co.uk, is simply a human-friendly alias to an IP address.

Sub Domains

In reality, one domain name can be associated with multiple IP addresses by prefixing the domain name with an additional name, in which case it becomes a Sub Domain. For example, www.bbc.co.uk is a sub-domain of bbc.co.uk, news.bbc.co.uk is a different sub-domain and can point to a different IP address (and therefore a different web server).

This is often done when the main corporate web site is hosted by a different server (possibly even a different ISP) to the eCommerce web site - www.mysite.co.uk can point to one IP address on a server provided by one ISP whilst shop.mysite.co.uk points to an entirely different IP address on another server, provided by another ISP altogether.

NB: a web site can be at {anything}.yourdomain.co.uk (or .com, .net, etc.), it is only convention that it should "www." but since most people will assume that your web site begins with "www." then anything else is confusing unless it is a separate web site that is linked to by your main web site. A good example of this is the News section of the BBC web site - you start off at www.bbc.co.uk but as soon as you click on "News", you end up at news.bbc.co.uk, but most people will not even notice this.

DNS

When you type a domain name into your web browser, your browser needs to translate that domain name into an IP address. It does this by looking up the domain name (or sub domain) in a directory, called a Domain Name Service, or DNS.

For performance reasons, the DNS that your browser uses is probably hosted by your ISP since, otherwise, every request out to the Internet would require two transactions; one to find the IP address for a given domain name, and one to send the data.

Your ISP's DNS needs to synchronise itself with all of the other DNS servers around the world so that it always contains the latest information on the domain names in use, and which IP address they correspond to. This synchronisation is happening around the clock but is a permeation, as one DNS server refreshes from another. Furthermore, each DNS entry has a timeout associated with it, which is usually 24 hours, which means that once a DNS Server has retrieved an entry for a particular domain, it does not need to get it again for that length of time. This means that if you change a DNS configuration (or more likely, your ISP changes a DNS configuration) because, for example, you have moved a web site from one server to another, or you have registered a new domain name, it can typically take up to 24 hours for that change to be picked up by all of the DNS servers around the world.

'A' Records and 'MX' Records

The DNS entries for a domain will actually be quite complex but there are two parts of a DNS entry that are referred to a lot - the A record and the MX record. Put simplisitically, the A record indicates the IP address of the web site for that domain and the MX record indicates the IP address of the email server for that domain.

This means that the email server for mysite.co.uk can be entirely different to the web server for mysite.co.uk; they can even be provided by different ISPs. This is what makes it possible to move your web site from one ISP to another without affecting your email (or vice versa).

Browser

In order to view web sites, you use a Browser, or web browser to be more precise. Historically, the most commonly used browser is Microsoft's Internet Explore ("IE" for short).

Microsoft Internet Explorer established major predominance because it was always shipped with Microsoft Windows. There is, however, a greater and greater movement away to 3rd party browsers, including

It should be noted that the layout of a web page is described in a text file by the web site and it is down to the browser to render (i.e. draw) the image of that web page as the designer intended. Thus not all browsers will render the same page in the same way. This may be down to the design of the browser, the age of the browser, but more often is caused by bugs in various web browsers!

HTML

The language used by the web designer to describe the content and layout of a web page is HTML (or Hypertext Markup Language).

HTML is plain text but with extra bits which describe the fonts, colours, size, etc. of text and the positioning of images. One of the challenges of web design is ensuring that this HTML produces the desired results for all visitors, regardless of the size of their screen, the type of browser, the type of computer (whether a Mac, a PC or an iPhone, for example) and which fonts, etc. they have installed.

Images used on web pages are not part of the HTML file - the HTML file simply describes where the browser can go and download the image(s) from and where to place them on the page. Thus, for example, the images on a web page do not need to have come from the same web site as the page itself!

Cookies

Cookies are small pieces of data that a web site can download onto the computer of the visitor so that, if they return to that site, they can be identified again. We use a cookie, for example, if you log in to this web site and tick the "remember me" box, we store an identifier in a cookie on your computer so that you do not need to log in again. This is the standard mechanism for identifying returning visitors to a site.