HTTP

HTTP stands for Hypertext Transfer Protocol. It is an application-layer protocol used for transferring hypermedia documents and other resources on the World Wide Web. It follows a client–server model: a client sends an HTTP request to a server, which responds with an HTTP response. The protocol is stateless by default, though state can be maintained using mechanisms such as cookies, sessions, or tokens.

Communication consists of request and response messages. Each request includes a request line with a method, a request-target (usually a URL), and a version, followed by header fields and an optional body. Each response includes a status line with a code and reason phrase, header fields, and an optional body. Common methods are GET, POST, PUT, PATCH, DELETE, HEAD, and OPTIONS. Resources are identified by URIs, and responses indicate the media type via Content-Type.

Versions and features: HTTP/1.0 introduced basic connections; HTTP/1.1 added persistent connections, host headers, chunked transfer encoding, caching, and more. HTTP/2 introduced binary framing, multiplexing, header compression, and server push. HTTP/3 uses QUIC over UDP to reduce latency and provides encrypted connections by default. Deployments increasingly favor TLS-based transport, with HTTPS (HTTP over TLS) serving as the standard secure mode.

Security and intermediaries: HTTPS secures data in transit with encryption, integrity, and authentication. Proxies, gateways, and content delivery networks (CDNs) can cache and forward requests, improving performance and reliability. Cache-Control, ETag, and Last-Modified headers guide reuse, while content negotiation allows servers to select representations based on client capabilities.

Standardization and evolution: HTTP specifications are developed by the IETF and, where applicable, the WHATWG. They are published as RFCs, with ongoing work to improve performance, security, and interoperability.