|
AOLServer |
Apache |
Microsoft IIS |
Netscape iPlanet 6.0 |
Netscape iPlanet 4.x |
Netscape Enterprise Server 3.6
HTTP is the protocol by which web traffic is handled. It defines how web pages are formatted and in what way they are delivered over the Internet. It also includes any information about the objects that are needed by proxy servers or a user's web browser. Currently, HTTP 1.1 is an IETF Standard. The World Wide Web Consortium is the leading organization dedicated to developing this standard.
The process of calling a web page from a server involves several steps. First, a client makes the request of a web server. The web server receives the request and checks to see if the page exists. If the page does not exist, an error message is returned to the client. If the page does exist, the web server then determines if there are any special processing needs, as would be the case with JSP or ASP pages. Any necessary processing is typically done by a third party application before being returned to the web server. The web server then prepares to send the file by adding any default or custom header information to the beginning of the requested page. Finally, the page is served to the client.
HTTP headers are used to define a variety of qualities about a web object. For instance, if a webmaster is setting the cachability of a web site, HTTP headers such as Cache-Control, Expires, and Last-Modified would be used. HTTP headers can also include custom header information. New applications, such as P3P, often require information about the object to be sent via headers. By adding custome headers, a webmaster can create an HTTP header named Copyright and add site specific copyright information to it. Because headers can be set in a variety of ways, the webmaster must choose the combination of headers and values that are appropriate to the needs of the web site.
It is important to note that an HTTP header is not the same as the META tag. META tags send information within the HTML page, while HTTP headers send information about the HTML page. AOL proxy caches, for instance, do not determine cachability via META tags.
There are two ways to set the HTTP header information. The first method is to set an HTTP header via the web server. All outgoing pages will have a default set of headers attached to them that are readable by other web servers, web caches, and web browsers. The second method is to set an HTTP header for each specific page. With this method, a web page must have been created using a programming language such as JSP or ASP, and then the page is scripted so that it will send the appropriate header information. It is readily apparent that there are two drawbacks with this method. First, all pages must be coded to send the header. Second, the pages cannot be static HTML. This will have some effect on the web server’s performance. The first method – to set the headers via the web server – is preferred and will be covered in greater detail within this section.
Depending on the web server involved, setting HTTP headers may require more technical expertise as the server's configuration files are often modified. This section will provide a general overview and some specific implementation details for a variety of the most commonly used web servers. It will also include links to various sites for more details. This section should be considered an introduction to HTTP headers; webmasters who will be implementing HTTP headers should consult their web server documentation for in-depth instruction.
Instructions for Miscellaneous HTTP Headers
| 
