3.8 Summary

HTTP is used to browse Web pages. HTTP is a client/server protocol with which Web browsers make requests to Web servers for resources stored on the server. HTTP resources can be static Web pages or dynamically generated pages, images, or any other kind of file. HTTP/1.0 was standardized by the IETF in 1996. HTTP/1.1 became a Proposed Standard in 1997 and was upgraded to Draft Standard in 1999 as RFC2616.

Every HTTP resource has a URL, or address. URLs are usually found in links in Web pages, but they're also passed around in email and even on paper. An HTTP URL has a path part that typically includes a path with directory names, but in an HTTP URL this path is supposed to be opaque to clients.

HTTP is a request/response protocol. An HTTP request is self-contained and normally generates a single self-contained response. A request always has a request line containing the method, request-URI for the resource being addressed, and the protocol version. A request may also have a number of headers and a body. The response is formatted similarly, with a response line, headers, and optional body.

HTTP defines eight methods: GET, PUT, POST, OPTIONS, HEAD, TRACE, DELETE, and CONNECT. In practice, however, only GET, POST, and CONNECT are widely used by ordinary browsers and servers. PUT and DELETE were designed to allow some limited authoring functionality, but in practice they are not well enough defined to be useful.

HTTP goes beyond request and response with a few interesting mechanisms. These mechanisms improve the performance of more advanced HTTP interactions. WebDAV clients and servers must also be aware of these mechanisms.

HTTP supports caching and proxy servers. A fair amount of the HTTP specification is designed to make sophisticated caching and proxying possible. Caches and proxies must be taken into account when WebDAV applications are being designed.