Part 1 - Introduction to Cookies
What Is A Cookie?
Apart from being a type of biscuit, a cookie is also a very useful piece of technology for use on the web. One of the problems which many websites need to overcome is that there is no way of directly finding out who is on a website. Although many details about the user (such as their browser, IP address and operating system) are available, the use of dynamic IP addresses (which change every time the user logs on) and IP address sharing (so that many people share the same IP) mean that there is no reliable way of recognising a particular user when they re-visit a website.
Cookies overcome this problem. They basically give the website owner the opportunity to store a little piece of information on a user's computer which they can then retrieve at a later date. Cookies are just tiny text files (only up to 4Kb in size) and a website can write them to the user's computer via the web browser. The same website can then request the cookie from the user and, if it exists, the value stored will be reported back to the website. The cookie can persist on the user's computer, staying there if the browser is closed, the computer is switched off and if the internet connection is changed.
What Use Is A Cookie?
So why would anyone want to store 4000 characters of text on a user's computer? It isn't enough to put anything really worthwhile on there! The power of the cookie, though, is to recognise a site visitor over and over again. To give just a few uses of cookies:
- Many portals and search engines use them to provide customized pages and results to their users, allowing such features as 'My Yahoo' etc.
- Visitor tracking and statistics systems often use them to track visitors. By assigning the visitor a cookie, they will not be counted more than once, so accurate unique visitor statistics can be obtained. Also, if a user has a unique cookie the system can 'follow' them through a website, showing the webmaster exactly where the visitor has been, and in what order.
A cookie is a very basic data file. It has a name and a value and also stores the address of websites which are allowed to access it and an expiry time. Basically, a website will set a cookie and give it a name and value. This name is used by the website to refer to it, and no other website can access the cookie, even if they know it's name. The name should be unique to the website, but it doesn't matter if it clashes with the name of a cookie from another website.
The cookie (as mentioned before) can only store up to 4000 characters of data. This is enough to store lots of information about a user so if, for example, you wanted to store the user preferences for a search engine (much like Google does), you could simply list the preferences in the cookie. If you wanted to store more data, you would have to store a unique ID in the cookie, which matched up with a database record, and you could th
en access the user's data this way.
To retrieve data, the website simply has to request if the user has a cookie with a particular name. If the user does, the value is returned to the script and it can be dealt with however the website owner chooses (for example a name stored in a cookie could be returned, a user ID could be loaded from a database, or a record could be made of a user visiting a site).
Every cookie is assigned an expiry date and time. It is up to the website owner to decide how long the cookie should exist for. Many owners may just choose to set the cookie for an hour, meaning it is only available for the user's single session. This is common in visitor tracking. Other cookies could be set for much longer. Maybe a week or a month (often used for affiliate program tracking) or even several years (often used for user preferences).
Despite much worrying in the news a few years ago, cookies pose no real danger to users. Unless they are really worried about themselves being recognised by a website, they are harmless. The browser actually writes and reads cookies from the computer when requested to by a website, so a malicious website cannot damage the computer.
For webmasters, there are some security concerns. When the cookie is set, the domain(s) which can access it are set. Usually this is just the website who set the cookie. This makes them relatively secure, as you can be sure that your competitor cannot load your cookie from one of your visitors' computers (they cannot even find out if it exisits).
One major security problem with cookies, though, is that they can easily be read by anyone using the computer. They are just a simple text file, so you should not under any circumstances store passwords in cookies. A common way to log people in automatically is to store an encrypted version of their password, which can then be matched with an encrypted version on the server. Another method is to store a unique ID and a unique validation number on the user's system. This is then referenced in a database to the user's account. This way, no actual details are stored and a malicious user cannot simply guess users' IDs (as there is the validation number).
This introduction has covered some of the basics of cookies and how they are used. The next three sections cover the setting and reading of cookies using four of the most common scripting languages available. Each page is a self contained description of how to set and read cookies for that language, so you should now jump ahead to the section for your chosen language.
1999 - 2001 David Gowans