Development of HTTP was initiated byTim Berners-Lee atCERN in 1989 and summarized in a simple document describing the behavior of a client and a server using the first HTTP version, named 0.9.[2] That version was subsequently developed, eventually becoming the public 1.0.[3]
HTTP/1 was finalized and fully documented (as version 1.0) in 1996.[4] It evolved (as version 1.1) in 1997 and then its specifications were updated in 1999, 2014, and 2022.[5] Its secure variant namedHTTPS is used by more than 85% of websites.[6]
HTTP/2, published in 2015, provides a more efficient expression of HTTP's semantics "on the wire". As of August 2024,[update] it is supported by 66.2% of websites[7][8] (35.3% HTTP/2 + 30.9% HTTP/3 with backwards compatibility) and supported by almost all web browsers (over 98% of users).[9] It is also supported by major web servers overTransport Layer Security (TLS) using anApplication-Layer Protocol Negotiation (ALPN) extension[10] whereTLS 1.2 or newer is required.[11][12]
HTTP/3, the successor to HTTP/2, was published in 2022.[13] As of February 2024,[update] it is now used on 30.9% of websites[14] and is supported by most web browsers, i.e. (at least partially) supported by 97% of users.[15] HTTP/3 usesQUIC instead ofTCP for the underlying transport protocol. Like HTTP/2, it does not obsolete previous major versions of the protocol. Support for HTTP/3 was added toCloudflare andGoogle Chrome first,[16][17] and is also enabled inFirefox.[18] HTTP/3 has lower latency for real-world web pages, if enabled on the server, and loads faster than with HTTP/2, in some cases over three times faster than HTTP/1.1 (which is still commonly only enabled).[19]
Technical overview
HTTP functions as arequest–response protocol in theclient–server model. Aweb browser, for example, may be theclient whereas aprocess, namedweb server, running on a computerhosting one or morewebsites may be theserver. The client submits an HTTPrequest message to the server. The server, which providesresources such asHTML files and other content or performs other functions on behalf of the client, returns aresponse message to the client. The response contains completion status information about the request and may also contain requested content in its message body.
A web browser is an example of auser agent (UA). Other types of user agent include the indexing software used by search providers (web crawlers),voice browsers,mobile apps, and othersoftware that accesses, consumes, or displays web content.
HTTP is designed to permit intermediate network elements to improve or enable communications between clients and servers. High-traffic websites often benefit fromweb cache servers that deliver content on behalf ofupstream servers to improve response time. Web browsers cache previously accessed web resources and reuse them, whenever possible, to reduce network traffic. HTTPproxy servers atprivate network boundaries can facilitate communication for clients without a globally routable address, by relaying messages with external servers.
To allow intermediate HTTP nodes (proxy servers, web caches, etc.) to accomplish their functions, some of theHTTP headers (found in HTTP requests/responses) are managedhop-by-hop whereas other HTTP headers are managedend-to-end (managed only by the source client and by the target web server).
In HTTP/1.0 a separate TCPconnection to the same server is made for every resource request.[21]
In HTTP/1.1 instead a TCP connection can be reused to make multiple resource requests (i.e. of HTML pages, frames, images,scripts,stylesheets, etc.).[22][23]
HTTP/1.1 communications therefore experience lesslatency as the establishment of TCP connections presents considerable overhead, especially under high traffic conditions.[24]
HTTP/2 is a revision of previous HTTP/1.1 in order to maintain the same client–server model and the same protocol methods but with these differences in order:
to use a compressed binary representation of metadata (HTTP headers) instead of a textual one, so that headers require much less space;
to use a singleTCP/IP (usuallyencrypted) connection per accessed server domain instead of 2 to 8 TCP/IP connections;
to use one or more bidirectional streams per TCP/IP connection in which HTTP requests and responses are broken down and transmitted in small packets to almost solve the problem of the HOLB (head-of-line blocking).[note 1]
to add a push capability to allow server application to send data to clients whenever new data is available (without forcing clients to request periodically new data to server by usingpolling methods).[25]
HTTP/2 communications therefore experience much less latency and, in most cases, even higher speeds than HTTP/1.1 communications.
HTTP/3 is a revision of previous HTTP/2 in order to useQUIC + UDP transport protocols instead of TCP. Before that version, TCP/IP connections were used; but now, only the IP layer is used (which UDP, like TCP, builds on). This slightly improves the average speed of communications and to avoid the occasional (very rare) problem of TCP connectioncongestion that can temporarily block or slow down the data flow of all its streams (another form of "head of line blocking").
The termhypertext was coined byTed Nelson in 1965 in theXanadu Project, which was in turn inspired byVannevar Bush's 1930s vision of the microfilm-based information retrieval and management "memex" system described in his 1945 essay "As We May Think".Tim Berners-Lee and his team atCERN are credited with inventing the original HTTP, along with HTML and the associated technology for aweb server and a clientuser interface calledweb browser. Berners-Lee designed HTTP in order to help with the adoption of his other idea: the "WorldWideWeb" project, which was first proposed in 1989, now known as theWorld Wide Web.
The first web server went live in 1990.[26][27] The protocol used had only one method, namely GET, which would request a page from a server.[28] The response from the server was always an HTML page.[2]
In 1991, the first documented official version of HTTP was written as a plain document, less than 700 words long, and this version was named HTTP/0.9, which supported only GET method, allowing clients to only retrieve HTML documents from the server, but not supporting any other file formats or information upload.[2]
HTTP/1.0-draft
Since 1992, a new document was written to specify the evolution of the basic protocol towards its next full version. It supported both the simple request method of the 0.9 version and the full GET request that included the client HTTP version. This was the first of the many unofficial HTTP/1.0 drafts that preceded the final work on HTTP/1.0.[3]
W3C HTTP Working Group
After having decided that new features of HTTP protocol were required and that they had to be fully documented as officialRFCs, in early 1995 the HTTP Working Group (HTTP WG, led byDave Raggett) was constituted with the aim to standardize and expand the protocol with extended operations, extended negotiation, richer meta-information, tied with a security protocol which became more efficient by adding additional methods andheader fields.[29][30]
The HTTP WG planned to revise and publish new versions of the protocol as HTTP/1.0 and HTTP/1.1 within 1995, but, because of the many revisions, that timeline lasted much more than one year.[31]
The HTTP WG planned also to specify a far future version of HTTP called HTTP-NG (HTTP Next Generation) that would have solved all remaining problems, of previous versions, related to performances, low latency responses, etc. but this work started only a few years later and it was never completed.
HTTP/1.0
In May 1996,RFC1945 was published as a final HTTP/1.0 revision of what had been used in previous 4 years as a pre-standard HTTP/1.0-draft which was already used by many web browsers and web servers.
In early 1996 developers started to even include unofficial extensions of the HTTP/1.0 protocol (i.e. keep-alive connections, etc.) into their products by using drafts of the upcoming HTTP/1.1 specifications.[32]
HTTP/1.1
Since early 1996, major web browsers and web server developers also started to implement new features specified by pre-standard HTTP/1.1 drafts specifications. End-user adoption of the new versions of browsers and servers was rapid. In March 1996, one web hosting company reported that over 40% of browsers in use on the Internet used the new HTTP/1.1 header "Host" to enablevirtual hosting, and that by June 1996, 65% of all browsers accessing their servers were pre-standard HTTP/1.1 compliant.[33]
In January 1997,RFC2068 was officially released as HTTP/1.1 specifications.
In June 1999,RFC2616 was released to include all improvements and updates based on previous (obsolete) HTTP/1.1 specifications.
W3C HTTP-NG Working Group
Resuming the old 1995 plan of previous HTTP Working Group, in 1997 anHTTP-NG Working Group was formed to develop a new HTTP protocol named HTTP-NG (HTTP New Generation). A few proposals / drafts were produced for the new protocol to usemultiplexing of HTTP transactions inside a single TCP/IP connection, but in 1999, the group stopped its activity passing the technical problems to IETF.[34]
IETF HTTP Working Group restarted
In 2007, the IETFHTTP Working Group (HTTP WG bis or HTTPbis) was restarted firstly to revise and clarify previous HTTP/1.1 specifications and secondly to write and refine future HTTP/2 specifications (named httpbis).[35][36]
SPDY: an unofficial HTTP protocol developed by Google
In 2009,Google, a private company, announced that it had developed and tested a new HTTP binary protocol namedSPDY. The implicit aim was to greatly speed up web traffic (specially between future web browsers and its servers).
SPDY was indeed much faster than HTTP/1.1 in many tests and so it was quickly adopted byChromium and then by other major web browsers.[37]
Some of the ideas about multiplexing HTTP streams over a single TCP/IP connection were taken from various sources, including the work of W3C HTTP-NG Working Group.
HTTP/2
In January–March 2012, HTTP Working Group (HTTPbis) announced the need to start to focus on a new HTTP/2 protocol (while finishing the revision of HTTP/1.1 specifications), maybe taking in consideration ideas and work done for SPDY.[38][39]
After a few months about what to do to develop a new version of HTTP, it was decided to derive it from SPDY.[40]
In May 2015,HTTP/2 was published asRFC7540 and quickly adopted by all web browsers already supporting SPDY and more slowly by web servers.
2014 updates to HTTP/1.1
In June 2014, the HTTP Working Group released an updated six-part HTTP/1.1 specification obsoletingRFC2616:
InRFC7230 Appendix-A, HTTP/0.9 was deprecated for servers supporting HTTP/1.1 version (and higher):[41]
Since HTTP/0.9 did not support header fields in a request, there is no mechanism for it to support name-based virtual hosts (selection of resource by inspection of the Host header field).Any server that implements name-based virtual hosts ought to disable support for HTTP/0.9. Most requests that appear to be HTTP/0.9 are, in fact, badly constructed HTTP/1.x requests caused by a client failing to properly encode the request-target.
Since 2016 many product managers and developers of user agents (browsers, etc.) and web servers have begun planning to gradually deprecate and dismiss support for HTTP/0.9 protocol, mainly for the following reasons:[42]
it is so simple that an RFC document was never written (there is only the original document);[2]
it has no HTTP headers and lacks many other features that nowadays are required for minimal security reasons;
it has not been widespread since 1999..2000 (because of HTTP/1.0 and HTTP/1.1) and is commonly used only by some very old network hardware, i.e.routers, etc.
In 2020, the first draftsHTTP/3 were published and major web browsers and web servers started to adopt it.
On 6 June 2022, IETF standardized HTTP/3 asRFC9114.[43]
Updates and refactoring in 2022
In June 2022, a batch of RFCs was published, deprecating many of the previous documents and introducing a few minor changes and a refactoring of HTTP semantics description into a separate document.
HTTP is astateless application-level protocol and it requires a reliable network transport connection to exchange data between client and server.[20] In HTTP implementations,TCP/IP connections are used usingwell-known ports (typicallyport 80 if the connection is unencrypted or port 443 if the connection is encrypted, see alsoList of TCP and UDP port numbers).[44][45] In HTTP/2, a TCP/IP connection plus multiple protocol channels are used. In HTTP/3, the application transport protocolQUIC over UDP is used.
Request and response messages through connections
Data is exchanged through a sequence ofrequest–response messages which are exchanged by asession layer transport connection.[20] An HTTP client initially tries to connect to a server establishing a connection (real or virtual). An HTTP(S) server listening on that port accepts the connection and then waits for a client's request message. The client sends its HTTP request message. Upon receiving the request the server sends back an HTTP response message, which includes header(s) plus a body if it is required. The body of this response message is typically the requested resource, although an error message or other information may also be returned. At any time (for many reasons) client or server can close the connection. Closing a connection is usually advertised in advance by using one or more HTTP headers in the last request/response message sent to server or client.[22]
InHTTP/0.9, the TCP/IP connection is always closed after server response has been sent, so it is never persistent.
InHTTP/1.0, as stated in RFC 1945, the TCP/IP connection should always be closed by server after a response has been sent.[note 3]
InHTTP/1.1 a keep-alive-mechanism was officially introduced so that a connection could be reused for more than one request/response. Such persistent connections reduce requestlatency perceptibly because the client does not need to re-negotiate theTCP 3-Way-Handshake connection after the first request has been sent. Another positive side effect is that, in general, the connection becomes faster with time due to TCP'sslow-start-mechanism.
HTTP/1.1 added alsoHTTP pipelining in order to further reduce lag time when using persistent connections by allowing clients to send multiple requests before waiting for each response. This optimization was never considered really safe because a few web servers and manyproxy servers, specially transparent proxy servers placed in Internet /Intranets between clients and servers, did not handle pipelined requests properly (they served only the first request discarding the others, they closed the connection because they saw more data after the first request or some proxies even returned responses out of order etc.). Because of this, only HEAD and some GET requests (i.e. limited to real file requests and so withURLs without query string used as a command, etc.) could be pipelined in asafe andidempotent mode. After many years of struggling with the problems introduced by enabling pipelining, this feature was first disabled and then removed from most browsers also because of the announced adoption of HTTP/2.
HTTP/2 extended the usage of persistent connections by multiplexing many concurrent requests/responses through a single TCP/IP connection.
HTTP/3 does not use TCP/IP connections but QUIC + UDP (see also:technical overview).
Content retrieval optimizations
HTTP/0.9
A requested resource was always sent in its entirety.
HTTP/1.0
HTTP/1.0 added headers to manage resources cached by client in order to allow conditional GET requests; in practice a server has to return the entire content of the requested resource only if its last modified time is not known by client or if it changed since last full response to GET request. One of these headers, "Content-Encoding", was added to specify whether the returned content of a resource was or was notcompressed.
If the total length of the content of a resource was not known in advance (i.e. because it was dynamically generated, etc.) then the header"Content-Length: number" was not present in HTTP headers and the client assumed that when server closed the connection, the content had been sent in its entirety. This mechanism could not distinguish between a resource transfer successfully completed and an interrupted one (because of a server / network error or something else).
HTTP/1.1
HTTP/1.1 introduced:
new headers to better manage the conditional retrieval of cached resources.
chunked transfer encoding to allow content to be streamed in chunks in order to reliably send it even when the server does not know its length in advance (i.e. because it is dynamically generated, etc.).
byte range serving, where a client can request only one or more portions (ranges of bytes) of a resource (i.e. the first part, a part in the middle or in the end of the entire content, etc.) and the server usually sends only the requested part(s). This is useful to resume an interrupted download (when a file is very large), when only a part of a content has to be shown or dynamically added to the already visible part by a browser (i.e. only the first or the following n comments of a web page) in order to spare time, bandwidth and system resources, etc.
HTTP/2, HTTP/3
Both HTTP/2 and HTTP/3 have kept the above mentioned features of HTTP/1.1.
HTTP authentication
HTTP provides multiple authentication schemes such asbasic access authentication anddigest access authentication which operate via a challenge–response mechanism whereby the server identifies and issues a challenge before serving the requested content.
HTTP provides a general framework for access control and authentication, via an extensible set of challenge–response authentication schemes, which can be used by a server to challenge a client request and by a client to provide authentication information.[1]
The authentication mechanisms described above belong to the HTTP protocol and are managed by client and server HTTP software (if configured to require authentication before allowing client access to one or more web resources), and not by the web applications usinga web application session.
Authentication realms
The HTTP Authentication specification also provides an arbitrary, implementation-specific construct for further dividing resources common to a given rootURI. The realm value string, if present, is combined with the canonical root URI to form the protection space component of the challenge. This in effect allows the server to define separate authentication scopes under one root URI.[1]
HTTP application session
HTTP is astateless protocol. A stateless protocol does not require the web server to retain information or status about each user for the duration of multiple requests.
To start an application user session, an interactiveauthentication via web applicationlogin must be performed. To stop a user session alogout operation must be requested by user. These kind of operations do not useHTTP authentication but a custom managed web application authentication.
HTTP/1.1 request messages
Request messages are sent by a client to a target server.[note 4]
Request syntax
A client sendsrequest messages to the server, which consist of:[47]
arequest line, consisting of the case-sensitive request method, aspace, the requested URI, another space, the protocol version, acarriage return, and aline feed, e.g.:
GET/images/logo.pngHTTP/1.1
zero or morerequest header fields (at least 1 or more headers in case of HTTP/1.1), each consisting of the case-insensitive field name, a colon, optional leadingwhitespace, the field value, an optional trailing whitespace and ending with a carriage return and a line feed, e.g.:
Host: www.example.comAccept-Language: en
an empty line, consisting of a carriage return and a line feed;
In the HTTP/1.1 protocol, all header fields exceptHost: hostname are optional.
A request line containing only the path name is accepted by servers to maintain compatibility with HTTP clients before the HTTP/1.0 specification inRFC1945.[48]
Request methods
An HTTP/1.1 request made using telnet. Therequest message,response header section, and response body are highlighted.
HTTP defines methods (sometimes referred to asverbs, but nowhere in the specification does it mentionverb) to indicate the desired action to be performed on the identified resource. What this resource represents, whether pre-existing data or data that is generated dynamically, depends on the implementation of the server. Often, the resource corresponds to a file or the output of an executable residing on the server. The HTTP/1.0 specification[49] defined the GET, HEAD, and POST methods as well as listing the PUT, DELETE, LINK and UNLINK methods under additional methods. However, the HTTP/1.1 specification[50] formally defined and added five new methods: PUT, DELETE, CONNECT, OPTIONS, and TRACE. Any client can use any method and the server can be configured to support any combination of methods. If a method is unknown to an intermediate, it will be treated as an unsafe andnon-idempotent method. There is no limit to the number of methods that can be defined, which allows for future methods to be specified without breaking existing infrastructure. For example,WebDAV defined seven new methods andRFC5789 specified thePATCH method.
Method names are case sensitive.[51][52] This is in contrast to HTTP header field names which are case-insensitive.[53]
GET
The GET method requests that the target resource transfer a representation of its state. GET requests should onlyretrieve data and should have no other effect. (This is also true of some other HTTP methods.)[1] For retrieving resources without making changes, GET is preferred over POST, as they can beaddressed through aURL. This enables bookmarking and sharing and makes GET responses eligible forcaching, which can save bandwidth. TheW3C has published guidance principles on this distinction, saying, "Web application design should be informed by the above principles, but also by the relevant limitations."[54] Seesafe methods below.
HEAD
The HEAD method requests that the target resource transfer a representation of its state, as for a GET request, but without the representation data enclosed in the response body. This is useful for retrieving the representation metadata in the response header, without having to transfer the entire representation. Uses include checking whether a page is available through thestatus code and quickly finding the size of afile (Content-Length).
POST
ThePOST method requests that the target resource process the representation enclosed in the request according to the semantics of the target resource. For example, it is used for posting a message to anInternet forum, subscribing to amailing list, or completing anonline shopping transaction.[55]
PUT
The PUT method requests that the target resource create or update its state with the state defined by the representation enclosed in the request. A distinction from POST is that the client specifies the target location on the server.[56]
DELETE
The DELETE method requests that the target resource delete its state.
CONNECT
The CONNECT method requests that the intermediary establish aTCP/IP tunnel to the origin server identified by the request target. It is often used to secure connections through one or moreHTTP proxies withTLS.[57][58] SeeHTTP CONNECT method.
OPTIONS
The OPTIONS method requests that the target resource transfer the HTTP methods that it supports. This can be used to check the functionality of a web server by requesting '*' instead of a specific resource.
TRACE
The TRACE method requests that the target resource transfer the received request in the response body. That way a client can see what (if any) changes or additions have been made by intermediaries.
PATCH
ThePATCH method requests that the target resource modify its state according to the partial update defined in the representation enclosed in the request. This can save bandwidth by updating a part of a file or document without having to transfer it entirely.[59]
All general-purpose web servers are required to implement at least the GET and HEAD methods, and all other methods are considered optional by the specification.[52]
A request method issafe if a request with that method has no intended effect on the server. The methods GET, HEAD, OPTIONS, and TRACE are defined as safe. In other words, safe methods are intended to beread-only. Safe methods can still haveside effects not seen by the client, such as appending request information to alog file or charging anadvertising account.
In contrast, the methods POST, PUT, DELETE, CONNECT, and PATCH are not safe. They may modify the state of the server or have other effects such as sending anemail. Such methods are therefore not usually used by conformingweb robots or web crawlers; some that do not conform tend to make requests without regard to context or consequences.
Despite the prescribed safety of GET requests, in practice their handling by the server is not technically limited in any way. Careless or deliberately irregular programming can allow GET requests to cause non-trivial changes on the server. This is discouraged because of the problems which can occur whenweb caching,search engines, and other automated agents make unintended changes on the server. For example, a website might allow deletion of a resource through a URL such ashttps://example.com/article/1234/delete, which, if arbitrarily fetched, even using GET, would simply delete the article.[60] A properly coded website would require a DELETE or POST method for this action, which non-malicious bots would not make.
One example of this occurring in practice was during the short-livedGoogle Web Accelerator beta, which prefetched arbitrary URLs on the page a user was viewing, causing records to be automatically altered or deleteden masse. Thebeta was suspended only weeks after its first release, following widespread criticism.[61][60]
A request method isidempotent if multiple identical requests with that method have the same effect as a single such request. The methods PUT and DELETE, and safe methods are defined as idempotent. Safe methods are trivially idempotent, since they are intended to have no effect on the server whatsoever; the PUT and DELETE methods, meanwhile, are idempotent since successive identical requests will be ignored. A website might, for instance, set up a PUT endpoint to modify a user's recorded email address. If this endpoint is configured correctly, any requests which ask to change a user's email address to the same email address which is already recorded—e.g. duplicate requests following a successful request—will have no effect. Similarly, a request to DELETE a certain user will have no effect if that user has already been deleted.
In contrast, the methods POST, CONNECT, and PATCH are not necessarily idempotent, and therefore sending an identical POST request multiple times may further modify the state of the server or have further effects, such as sending multipleemails. In some cases this is the desired effect, but in other cases it may occur accidentally. A user might, for example, inadvertently send multiple POST requests by clicking a button again if they were not given clear feedback that the first click was being processed. Whileweb browsers may showalert dialog boxes to warn users in some cases where reloading a page may re-submit a POST request, it is generally up to the web application to handle cases where a POST request should not be submitted more than once.
Note that whether or not a method is idempotent is not enforced by the protocol or web server. It is perfectly possible to write a web application in which (for example) a database insert or other non-idempotent action is triggered by a GET or other request. To do so against recommendations, however, may result in undesirable consequences, if auser agent assumes that repeating the same request is safe when it is not.
A request method iscacheable if responses to requests with that method may be stored for future reuse. The methods GET, HEAD, and POST are defined as cacheable.
In contrast, the methods PUT, DELETE, CONNECT, OPTIONS, TRACE, and PATCH are not cacheable.
Request header fields allow the client to pass additional information beyond the request line, acting as request modifiers (similarly to the parameters of a procedure). They give information about the client, about the target resource, or about the expected handling of the request.
HTTP/1.1 response messages
A response message is sent by a server to a client as a reply to its former request message.[note 4]
Response syntax
A server sendsresponse messages to the client, which consist of:[47]
zero or moreresponse header fields, each consisting of the case-insensitive field name, a colon, optional leadingwhitespace, the field value, an optional trailing whitespace and ending with a carriage return and a line feed, e.g.:
Content-Type: text/html
an empty line, consisting of a carriage return and a line feed;
In HTTP/1.0 and since, the first line of the HTTP response is called thestatus line and includes a numericstatus code (such as "404") and a textualreason phrase (such as "Not Found"). The response status code is a three-digit integer code representing the result of the server's attempt to understand and satisfy the client's corresponding request. The way the client handles the response depends primarily on the status code, and secondarily on the other response header fields. Clients may not understand all registered status codes but they must understand their class (given by the first digit of the status code) and treat an unrecognized status code as being equivalent to the x00 status code of that class.
The standardreason phrases are only recommendations, and can be replaced with "local equivalents" at theweb developer's discretion. If the status code indicated a problem, the user agent might display thereason phrase to the user to provide further information about the nature of the problem. The standard also allows the user agent to attempt to interpret thereason phrase, though this might be unwise since the standard explicitly specifies that status codes are machine-readable andreason phrases are human-readable.
The first digit of the status code defines its class:
1XX (informational)
The request was received, continuing process.
2XX (successful)
The request was successfully received, understood, and accepted.
3XX (redirection)
Further action needs to be taken in order to complete the request.
4XX (client error)
The request contains bad syntax or cannot be fulfilled.
5XX (server error)
The server failed to fulfill an apparently valid request.
The response header fields allow the server to pass additional information beyond the status line, acting as response modifiers. They give information about the server or about further access to the target resource or related resources.
Each response header field has a defined meaning which can be further refined by the semantics of the request method or response status code.
HTTP/1.1 example of request / response transaction
Below is a sample HTTP transaction between an HTTP/1.1 client and an HTTP/1.1 server running onwww.example.com, port 80.[note 5][note 6]
A client request (consisting in this case of the request line and a few headers that can be reduced to only the"Host: hostname" header) is followed by a blank line, so that the request ends with a double end of line, each in the form of acarriage return followed by aline feed. The"Host: hostname" header value distinguishes between variousDNS names sharing a singleIP address, allowing name-basedvirtual hosting. While optional in HTTP/1.0, it is mandatory in HTTP/1.1. (A "/" (slash) will usually fetch a/index.html file if there is one.)
Server response
HTTP/1.1200OKDate:Mon, 23 May 2005 22:38:34 GMTContent-Type:text/html; charset=UTF-8Content-Length:155Last-Modified:Wed, 08 Jan 2003 23:11:55 GMTServer:Apache/1.3.3.7 (Unix) (Red-Hat/Linux)ETag:"3f80f-1b6-3e1cb03b"Accept-Ranges:bytesConnection:close<html><head><title>An Example Page</title></head><body><p>Hello World, this is a very simple HTML document.</p></body></html>
TheETag (entity tag) header field is used to determine if a cached version of the requested resource is identical to the current version of the resource on the server."Content-Type" specifies theInternet media type of the data conveyed by the HTTP message, while"Content-Length" indicates its length in bytes. The HTTP/1.1webserver publishes its ability to respond to requests for certain byte ranges of the document by setting the field"Accept-Ranges: bytes". This is useful, if the client needs to have only certain portions[62] of a resource sent by the server, which is calledbyte serving. When"Connection: close" is sent, it means that theweb server will close theTCP connection immediately after the end of the transfer of this response.[22]
Most of the header lines are optional but some are mandatory. When header"Content-Length: number" is missing in a response with an entity body then this should be considered an error in HTTP/1.0 but it may not be an error in HTTP/1.1 if header"Transfer-Encoding: chunked" is present. Chunked transfer encoding uses a chunk size of 0 to mark the end of the content. Some old implementations of HTTP/1.0 omitted the header"Content-Length" when the length of the body entity was not known at the beginning of the response and so the transfer of data to client continued until server closed the socket.
A"Content-Encoding:gzip" can be used to inform the client that the body entity part of the transmitted data is compressed by gzip algorithm.
Encrypted connections
The most popular way of establishing an encrypted HTTP connection isHTTPS.[63] Two other methods for establishing an encrypted HTTP connection also exist:Secure Hypertext Transfer Protocol, and using theHTTP/1.1 Upgrade header to specify an upgrade to TLS. Browser support for these two is, however, nearly non-existent.[64][65][66]
Similar protocols
TheGopher protocol is a content delivery protocol that was displaced by HTTP in the early 1990s.
TheSPDY protocol is an alternative to HTTP developed atGoogle, superseded byHTTP/2.
TheGemini protocol is a Gopher-inspired protocol which mandates privacy-related features.
Constrained Application Protocol – a semantically similar protocol to HTTP but used UDP or UDP-like messages targeted for devices with limited processing capability; re-uses HTTP and other internet concepts likeInternet media type and web linking (RFC 5988)[67]
^In practice, these streams are used as multiple TCP/IP sub-connections tomultiplex concurrent requests/responses, thus greatly reducing the number of real TCP/IP connections on server side, from 2..8 per client to 1, and allowing many more clients to be served at once.
^In 2022, HTTP/0.9 support has not been officially completely deprecated and is still present in many web servers and browsers (for server responses only), even if usually disabled. It is unclear how long it will take to decommission HTTP/0.9.
^Since late 1996, some developers of popular HTTP/1.0 browsers and servers (specially those who had planned support for HTTP/1.1 too), started to deploy (as an unofficial extension) a sort of keep-alive-mechanism (by using new HTTP headers) in order to keep the TCP/IP connection open for more than a request/response pair and so to speed up the exchange of multiple requests/responses.[32]
^abHTTP/2 and HTTP/3 have a different representation for HTTP methods and headers.
^HTTP/1.0 has the same messages except for a few missing headers.
^HTTP/2 and HTTP/3 use the same request / response mechanism but with different representations for HTTP headers.
