HTTP

HyperText Transfer Protocol, the principle protocol that governs the rules of behaviour between a client browser and a server on the World Wide Web. Client browsers issue requests to servers, which either send back a document or a response code, possibly indicating that there was a problem retrieving the document, or that it has moved.

Along with a document, servers send HTTP headers back to the client, often describing how to treat the document (for example, as an HTML page, an image, or a PDF file).

Most resources on the web have URLs that start with http://, an indication that they are to be obtained using this protocol. As this is by far the most common transfer protocol used for the web, most browsers don't require users to enter this portion of a URL, inserting it for you if ommitted. Thus, entering www.primitivetype.com in the address bar of a browser is usually the same as entering http://www.primitivetype.com.

HTTP is a stateless protocol, meaning that the connection between requests is not persistent; with each request to a server, a browser has to establish a new request. This has implications for web programmers, especially for scripts where the server needs to remember who is using a website to implement features such as client logging in. Cookies are used by HTTP servers to place files on a client computer that help it identify it on subsequent requests.

See ALL glossary entries