source: draft-ietf-httpbis/latest/p1-messaging.xml @ 252

Last change on this file since 252 was 252, checked in by julian.reschke@…, 13 years ago

Tweak "Editorial Note", now pointing to Changes section in appendix.

  • Property svn:eol-style set to native
File size: 191.1 KB
1<?xml version="1.0" encoding="utf-8"?>
2<?xml-stylesheet type='text/xsl' href='../myxml2rfc.xslt'?>
3<!DOCTYPE rfc [
4  <!ENTITY MAY "<bcp14 xmlns=''>MAY</bcp14>">
5  <!ENTITY MUST "<bcp14 xmlns=''>MUST</bcp14>">
6  <!ENTITY MUST-NOT "<bcp14 xmlns=''>MUST NOT</bcp14>">
7  <!ENTITY OPTIONAL "<bcp14 xmlns=''>OPTIONAL</bcp14>">
8  <!ENTITY RECOMMENDED "<bcp14 xmlns=''>RECOMMENDED</bcp14>">
9  <!ENTITY REQUIRED "<bcp14 xmlns=''>REQUIRED</bcp14>">
10  <!ENTITY SHALL "<bcp14 xmlns=''>SHALL</bcp14>">
11  <!ENTITY SHALL-NOT "<bcp14 xmlns=''>SHALL NOT</bcp14>">
12  <!ENTITY SHOULD "<bcp14 xmlns=''>SHOULD</bcp14>">
13  <!ENTITY SHOULD-NOT "<bcp14 xmlns=''>SHOULD NOT</bcp14>">
14  <!ENTITY ID-VERSION "latest">
15  <!ENTITY ID-MONTH "May">
16  <!ENTITY ID-YEAR "2008">
17  <!ENTITY caching                "<xref target='Part6' x:rel='#caching' xmlns:x=''/>">
18  <!ENTITY payload                "<xref target='Part3' xmlns:x=''/>">
19  <!ENTITY media-types            "<xref target='Part3' x:rel='#media.types' xmlns:x=''/>">
20  <!ENTITY content-codings        "<xref target='Part3' x:rel='#content.codings' xmlns:x=''/>">
21  <!ENTITY CONNECT                "<xref target='Part2' x:rel='#CONNECT' xmlns:x=''/>">
22  <!ENTITY content.negotiation    "<xref target='Part3' x:rel='#content.negotiation' xmlns:x=''/>">
23  <!ENTITY diff2045entity         "<xref target='Part3' x:rel='#differences.between.http.entities.and.rfc.2045.entities' xmlns:x=''/>">
24  <!ENTITY entity                 "<xref target='Part3' x:rel='#entity' xmlns:x=''/>">
25  <!ENTITY entity-body            "<xref target='Part3' x:rel='#entity.body' xmlns:x=''/>">
26  <!ENTITY entity-header-fields   "<xref target='Part3' x:rel='#entity.header.fields' xmlns:x=''/>">
27  <!ENTITY header-accept          "<xref target='Part3' x:rel='#header.accept' xmlns:x=''/>">
28  <!ENTITY header-cache-control   "<xref target='Part6' x:rel='#header.cache-control' xmlns:x=''/>">
29  <!ENTITY header-expect          "<xref target='Part2' x:rel='#header.expect' xmlns:x=''/>">
30  <!ENTITY header-pragma          "<xref target='Part6' x:rel='#header.pragma' xmlns:x=''/>">
31  <!ENTITY header-warning         "<xref target='Part6' x:rel='#header.warning' xmlns:x=''/>">
32  <!ENTITY idempotent-methods     "<xref target='Part2' x:rel='#idempotent.methods' xmlns:x=''/>">
33  <!ENTITY qvalue                 "<xref target='Part3' x:rel='#quality.values' xmlns:x=''/>">
34  <!ENTITY request-header-fields  "<xref target='Part2' x:rel='#request.header.fields' xmlns:x=''/>">
35  <!ENTITY response-header-fields "<xref target='Part2' x:rel='#response.header.fields' xmlns:x=''/>">
36  <!ENTITY method                 "<xref target='Part2' x:rel='#method' xmlns:x=''/>">
37  <!ENTITY status-codes           "<xref target='Part2' x:rel='' xmlns:x=''/>">
38  <!ENTITY status-100             "<xref target='Part2' x:rel='#status.100' xmlns:x=''/>">
39  <!ENTITY status-1xx             "<xref target='Part2' x:rel='#status.1xx' xmlns:x=''/>">
40  <!ENTITY status-414             "<xref target='Part2' x:rel='#status.414' xmlns:x=''/>">
42<?rfc toc="yes" ?>
43<?rfc symrefs="yes" ?>
44<?rfc sortrefs="yes" ?>
45<?rfc compact="yes"?>
46<?rfc subcompact="no" ?>
47<?rfc linkmailto="no" ?>
48<?rfc editing="no" ?>
49<?rfc comments="yes"?>
50<?rfc inline="yes"?>
51<?rfc-ext allow-markup-in-artwork="yes" ?>
52<?rfc-ext include-references-in-index="yes" ?>
53<rfc obsoletes="2616" category="std"
54     ipr="full3978" docName="draft-ietf-httpbis-p1-messaging-&ID-VERSION;"
55     xmlns:x=''>
58  <title abbrev="HTTP/1.1, Part 1">HTTP/1.1, part 1: URIs, Connections, and Message Parsing</title>
60  <author initials="R." surname="Fielding" fullname="Roy T. Fielding" role="editor">
61    <organization abbrev="Day Software">Day Software</organization>
62    <address>
63      <postal>
64        <street>23 Corporate Plaza DR, Suite 280</street>
65        <city>Newport Beach</city>
66        <region>CA</region>
67        <code>92660</code>
68        <country>USA</country>
69      </postal>
70      <phone>+1-949-706-5300</phone>
71      <facsimile>+1-949-706-5305</facsimile>
72      <email></email>
73      <uri></uri>
74    </address>
75  </author>
77  <author initials="J." surname="Gettys" fullname="Jim Gettys">
78    <organization>One Laptop per Child</organization>
79    <address>
80      <postal>
81        <street>21 Oak Knoll Road</street>
82        <city>Carlisle</city>
83        <region>MA</region>
84        <code>01741</code>
85        <country>USA</country>
86      </postal>
87      <email></email>
88      <uri></uri>
89    </address>
90  </author>
92  <author initials="J." surname="Mogul" fullname="Jeffrey C. Mogul">
93    <organization abbrev="HP">Hewlett-Packard Company</organization>
94    <address>
95      <postal>
96        <street>HP Labs, Large Scale Systems Group</street>
97        <street>1501 Page Mill Road, MS 1177</street>
98        <city>Palo Alto</city>
99        <region>CA</region>
100        <code>94304</code>
101        <country>USA</country>
102      </postal>
103      <email></email>
104    </address>
105  </author>
107  <author initials="H." surname="Frystyk" fullname="Henrik Frystyk Nielsen">
108    <organization abbrev="Microsoft">Microsoft Corporation</organization>
109    <address>
110      <postal>
111        <street>1 Microsoft Way</street>
112        <city>Redmond</city>
113        <region>WA</region>
114        <code>98052</code>
115        <country>USA</country>
116      </postal>
117      <email></email>
118    </address>
119  </author>
121  <author initials="L." surname="Masinter" fullname="Larry Masinter">
122    <organization abbrev="Adobe Systems">Adobe Systems, Incorporated</organization>
123    <address>
124      <postal>
125        <street>345 Park Ave</street>
126        <city>San Jose</city>
127        <region>CA</region>
128        <code>95110</code>
129        <country>USA</country>
130      </postal>
131      <email></email>
132      <uri></uri>
133    </address>
134  </author>
136  <author initials="P." surname="Leach" fullname="Paul J. Leach">
137    <organization abbrev="Microsoft">Microsoft Corporation</organization>
138    <address>
139      <postal>
140        <street>1 Microsoft Way</street>
141        <city>Redmond</city>
142        <region>WA</region>
143        <code>98052</code>
144      </postal>
145      <email></email>
146    </address>
147  </author>
149  <author initials="T." surname="Berners-Lee" fullname="Tim Berners-Lee">
150    <organization abbrev="W3C/MIT">World Wide Web Consortium</organization>
151    <address>
152      <postal>
153        <street>MIT Computer Science and Artificial Intelligence Laboratory</street>
154        <street>The Stata Center, Building 32</street>
155        <street>32 Vassar Street</street>
156        <city>Cambridge</city>
157        <region>MA</region>
158        <code>02139</code>
159        <country>USA</country>
160      </postal>
161      <email></email>
162      <uri></uri>
163    </address>
164  </author>
166  <author initials="Y." surname="Lafon" fullname="Yves Lafon" role="editor">
167    <organization abbrev="W3C">World Wide Web Consortium</organization>
168    <address>
169      <postal>
170        <street>W3C / ERCIM</street>
171        <street>2004, rte des Lucioles</street>
172        <city>Sophia-Antipolis</city>
173        <region>AM</region>
174        <code>06902</code>
175        <country>France</country>
176      </postal>
177      <email></email>
178      <uri></uri>
179    </address>
180  </author>
182  <author initials="J. F." surname="Reschke" fullname="Julian F. Reschke" role="editor">
183    <organization abbrev="greenbytes">greenbytes GmbH</organization>
184    <address>
185      <postal>
186        <street>Hafenweg 16</street>
187        <city>Muenster</city><region>NW</region><code>48155</code>
188        <country>Germany</country>
189      </postal>
190      <phone>+49 251 2807760</phone>   
191      <facsimile>+49 251 2807761</facsimile>   
192      <email></email>       
193      <uri></uri>     
194    </address>
195  </author>
197  <date month="&ID-MONTH;" year="&ID-YEAR;"/>
201   The Hypertext Transfer Protocol (HTTP) is an application-level
202   protocol for distributed, collaborative, hypermedia information
203   systems. HTTP has been in use by the World Wide Web global information
204   initiative since 1990. This document is Part 1 of the seven-part specification
205   that defines the protocol referred to as "HTTP/1.1" and, taken together,
206   obsoletes RFC 2616.  Part 1 provides an overview of HTTP and
207   its associated terminology, defines the "http" and "https" Uniform
208   Resource Identifier (URI) schemes, defines the generic message syntax
209   and parsing requirements for HTTP message frames, and describes
210   general security concerns for implementations.
214<note title="Editorial Note (To be removed by RFC Editor)">
215  <t>
216    Discussion of this draft should take place on the HTTPBIS working group
217    mailing list ( The current issues list is
218    at <eref target=""/>
219    and related documents (including fancy diffs) can be found at
220    <eref target=""/>.
221  </t>
222  <t>
223    The changes in this draft are summarized in <xref target="changes.since.02"/>.
224  </t>
228<section title="Introduction" anchor="introduction">
230   The Hypertext Transfer Protocol (HTTP) is an application-level
231   protocol for distributed, collaborative, hypermedia information
232   systems. HTTP has been in use by the World-Wide Web global
233   information initiative since 1990. The first version of HTTP, commonly
234   referred to as HTTP/0.9, was a simple protocol for raw data transfer
235   across the Internet with only a single method and no metadata.
236   HTTP/1.0, as defined by <xref target="RFC1945"/>, improved
237   the protocol by allowing messages to be in the format of MIME-like
238   messages, containing metadata about the data transferred and
239   modifiers on the request/response semantics. However, HTTP/1.0 did
240   not sufficiently take into consideration the effects of hierarchical
241   proxies, caching, the need for persistent connections, or name-based
242   virtual hosts. In addition, the proliferation of incompletely-implemented
243   applications calling themselves "HTTP/1.0" necessitated a
244   protocol version change in order for two communicating applications
245   to determine each other's true capabilities.
248   This document is Part 1 of the seven-part specification that defines
249   the protocol referred to as "HTTP/1.1", obsoleting <xref target="RFC2616"/>.
250   HTTP/1.1 remains compatible with HTTP/1.0 by including more stringent
251   requirements that enable reliable implementations and adding only
252   those new features that will either be safely ignored by an HTTP/1.0
253   recipient or only sent when communicating with a party advertising
254   compliance with HTTP/1.1.
255   Part 1 defines those aspects of HTTP/1.1 related to overall network
256   operation, message framing, interaction with transport protocols, and
257   URI schemes.
260   This document is currently disorganized in order to minimize the changes
261   between drafts and enable reviewers to see the smaller errata changes.
262   The next draft will reorganize the sections to better reflect the content.
263   In particular, the sections will be organized according to the typical
264   process of deciding when to use HTTP (URI schemes), overall network operation,
265   connection management, message framing, and generic message parsing.
266   The current mess reflects how widely dispersed these topics and associated
267   requirements had become in <xref target="RFC2616"/>.
270<section title="Purpose" anchor="intro.purpose">
272   Practical information systems require more functionality than simple
273   retrieval, including search, front-end update, and annotation. HTTP
274   allows an open-ended set of methods and headers that indicate the
275   purpose of a request <xref target="RFC2324"/>. It builds on the discipline of reference
276   provided by the Uniform Resource Identifier (URI) <xref target="RFC1630"/>, as a location
277   (URL) <xref target="RFC1738"/> or name (URN) <xref target="RFC1737"/>, for indicating the resource to which a
278   method is to be applied. Messages are passed in a format similar to
279   that used by Internet mail <xref target="RFC2822"/> as defined by the Multipurpose
280   Internet Mail Extensions (MIME) <xref target="RFC2045"/>.
283   HTTP is also used as a generic protocol for communication between
284   user agents and proxies/gateways to other Internet systems, including
285   those supported by the SMTP <xref target="RFC2821"/>, NNTP <xref target="RFC3977"/>, FTP <xref target="RFC959"/>, Gopher <xref target="RFC1436"/>,
286   and WAIS <xref target="WAIS"/> protocols. In this way, HTTP allows basic hypermedia
287   access to resources available from diverse applications.
291<section title="Requirements" anchor="intro.requirements">
293   The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
294   "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this
295   document are to be interpreted as described in <xref target="RFC2119"/>.
298   An implementation is not compliant if it fails to satisfy one or more
299   of the &MUST; or &REQUIRED; level requirements for the protocols it
300   implements. An implementation that satisfies all the &MUST; or &REQUIRED;
301   level and all the &SHOULD; level requirements for its protocols is said
302   to be "unconditionally compliant"; one that satisfies all the &MUST;
303   level requirements but not all the &SHOULD; level requirements for its
304   protocols is said to be "conditionally compliant."
308<section title="Terminology" anchor="intro.terminology">
310   This specification uses a number of terms to refer to the roles
311   played by participants in, and objects of, the HTTP communication.
314  <iref item="connection"/>
315  <x:dfn>connection</x:dfn>
316  <list>
317    <t>
318      A transport layer virtual circuit established between two programs
319      for the purpose of communication.
320    </t>
321  </list>
324  <iref item="message"/>
325  <x:dfn>message</x:dfn>
326  <list>
327    <t>
328      The basic unit of HTTP communication, consisting of a structured
329      sequence of octets matching the syntax defined in <xref target="http.message"/> and
330      transmitted via the connection.
331    </t>
332  </list>
335  <iref item="request"/>
336  <x:dfn>request</x:dfn>
337  <list>
338    <t>
339      An HTTP request message, as defined in <xref target="request"/>.
340    </t>
341  </list>
344  <iref item="response"/>
345  <x:dfn>response</x:dfn>
346  <list>
347    <t>
348      An HTTP response message, as defined in <xref target="response"/>.
349    </t>
350  </list>
353  <iref item="resource"/>
354  <x:dfn>resource</x:dfn>
355  <list>
356    <t>
357      A network data object or service that can be identified by a URI,
358      as defined in <xref target="uri"/>. Resources may be available in multiple
359      representations (e.g. multiple languages, data formats, size, and
360      resolutions) or vary in other ways.
361    </t>
362  </list>
365  <iref item="entity"/>
366  <x:dfn>entity</x:dfn>
367  <list>
368    <t>
369      The information transferred as the payload of a request or
370      response. An entity consists of metainformation in the form of
371      entity-header fields and content in the form of an entity-body, as
372      described in &entity;.
373    </t>
374  </list>
377  <iref item="representation"/>
378  <x:dfn>representation</x:dfn>
379  <list>
380    <t>
381      An entity included with a response that is subject to content
382      negotiation, as described in &content.negotiation;. There may exist multiple
383      representations associated with a particular response status.
384    </t>
385  </list>
388  <iref item="content negotiation"/>
389  <x:dfn>content negotiation</x:dfn>
390  <list>
391    <t>
392      The mechanism for selecting the appropriate representation when
393      servicing a request, as described in &content.negotiation;. The
394      representation of entities in any response can be negotiated
395      (including error responses).
396    </t>
397  </list>
400  <iref item="variant"/>
401  <x:dfn>variant</x:dfn>
402  <list>
403    <t>
404      A resource may have one, or more than one, representation(s)
405      associated with it at any given instant. Each of these
406      representations is termed a `variant'.  Use of the term `variant'
407      does not necessarily imply that the resource is subject to content
408      negotiation.
409    </t>
410  </list>
413  <iref item="client"/>
414  <x:dfn>client</x:dfn>
415  <list>
416    <t>
417      A program that establishes connections for the purpose of sending
418      requests.
419    </t>
420  </list>
423  <iref item="user agent"/>
424  <x:dfn>user agent</x:dfn>
425  <list>
426    <t>
427      The client which initiates a request. These are often browsers,
428      editors, spiders (web-traversing robots), or other end user tools.
429    </t>
430  </list>
433  <iref item="server"/>
434  <x:dfn>server</x:dfn>
435  <list>
436    <t>
437      An application program that accepts connections in order to
438      service requests by sending back responses. Any given program may
439      be capable of being both a client and a server; our use of these
440      terms refers only to the role being performed by the program for a
441      particular connection, rather than to the program's capabilities
442      in general. Likewise, any server may act as an origin server,
443      proxy, gateway, or tunnel, switching behavior based on the nature
444      of each request.
445    </t>
446  </list>
449  <iref item="origin server"/>
450  <x:dfn>origin server</x:dfn>
451  <list>
452    <t>
453      The server on which a given resource resides or is to be created.
454    </t>
455  </list>
458  <iref item="proxy"/>
459  <x:dfn>proxy</x:dfn>
460  <list>
461    <t>
462      An intermediary program which acts as both a server and a client
463      for the purpose of making requests on behalf of other clients.
464      Requests are serviced internally or by passing them on, with
465      possible translation, to other servers. A proxy &MUST; implement
466      both the client and server requirements of this specification. A
467      "transparent proxy" is a proxy that does not modify the request or
468      response beyond what is required for proxy authentication and
469      identification. A "non-transparent proxy" is a proxy that modifies
470      the request or response in order to provide some added service to
471      the user agent, such as group annotation services, media type
472      transformation, protocol reduction, or anonymity filtering. Except
473      where either transparent or non-transparent behavior is explicitly
474      stated, the HTTP proxy requirements apply to both types of
475      proxies.
476    </t>
477  </list>
480  <iref item="gateway"/>
481  <x:dfn>gateway</x:dfn>
482  <list>
483    <t>
484      A server which acts as an intermediary for some other server.
485      Unlike a proxy, a gateway receives requests as if it were the
486      origin server for the requested resource; the requesting client
487      may not be aware that it is communicating with a gateway.
488    </t>
489  </list>
492  <iref item="tunnel"/>
493  <x:dfn>tunnel</x:dfn>
494  <list>
495    <t>
496      An intermediary program which is acting as a blind relay between
497      two connections. Once active, a tunnel is not considered a party
498      to the HTTP communication, though the tunnel may have been
499      initiated by an HTTP request. The tunnel ceases to exist when both
500      ends of the relayed connections are closed.
501    </t>
502  </list>
505  <iref item="cache"/>
506  <x:dfn>cache</x:dfn>
507  <list>
508    <t>
509      A program's local store of response messages and the subsystem
510      that controls its message storage, retrieval, and deletion. A
511      cache stores cacheable responses in order to reduce the response
512      time and network bandwidth consumption on future, equivalent
513      requests. Any client or server may include a cache, though a cache
514      cannot be used by a server that is acting as a tunnel.
515    </t>
516  </list>
519  <iref item="cacheable"/>
520  <x:dfn>cacheable</x:dfn>
521  <list>
522    <t>
523      A response is cacheable if a cache is allowed to store a copy of
524      the response message for use in answering subsequent requests. The
525      rules for determining the cacheability of HTTP responses are
526      defined in &caching;. Even if a resource is cacheable, there may
527      be additional constraints on whether a cache can use the cached
528      copy for a particular request.
529    </t>
530  </list>
533  <iref item="upstream"/>
534  <iref item="downstream"/>
535  <x:dfn>upstream</x:dfn>/<x:dfn>downstream</x:dfn>
536  <list>
537    <t>
538      Upstream and downstream describe the flow of a message: all
539      messages flow from upstream to downstream.
540    </t>
541  </list>
544  <iref item="inbound"/>
545  <iref item="outbound"/>
546  <x:dfn>inbound</x:dfn>/<x:dfn>outbound</x:dfn>
547  <list>
548    <t>
549      Inbound and outbound refer to the request and response paths for
550      messages: "inbound" means "traveling toward the origin server",
551      and "outbound" means "traveling toward the user agent"
552    </t>
553  </list>
557<section title="Overall Operation" anchor="intro.overall.operation">
559   HTTP is a request/response protocol. A client sends a
560   request to the server in the form of a request method, URI, and
561   protocol version, followed by a MIME-like message containing request
562   modifiers, client information, and possible body content over a
563   connection with a server. The server responds with a status line,
564   including the message's protocol version and a success or error code,
565   followed by a MIME-like message containing server information, entity
566   metainformation, and possible entity-body content. The relationship
567   between HTTP and MIME is described in &diff2045entity;.
570   Most HTTP communication is initiated by a user agent and consists of
571   a request to be applied to a resource on some origin server. In the
572   simplest case, this may be accomplished via a single connection (v)
573   between the user agent (UA) and the origin server (O).
575<figure><artwork type="drawing">
576       request chain ------------------------&gt;
577    UA -------------------v------------------- O
578       &lt;----------------------- response chain
581   A more complicated situation occurs when one or more intermediaries
582   are present in the request/response chain. There are three common
583   forms of intermediary: proxy, gateway, and tunnel. A proxy is a
584   forwarding agent, receiving requests for a URI in its absolute form,
585   rewriting all or part of the message, and forwarding the reformatted
586   request toward the server identified by the URI. A gateway is a
587   receiving agent, acting as a layer above some other server(s) and, if
588   necessary, translating the requests to the underlying server's
589   protocol. A tunnel acts as a relay point between two connections
590   without changing the messages; tunnels are used when the
591   communication needs to pass through an intermediary (such as a
592   firewall) even when the intermediary cannot understand the contents
593   of the messages.
595<figure><artwork type="drawing">
596       request chain --------------------------------------&gt;
597    UA -----v----- A -----v----- B -----v----- C -----v----- O
598       &lt;------------------------------------- response chain
601   The figure above shows three intermediaries (A, B, and C) between the
602   user agent and origin server. A request or response message that
603   travels the whole chain will pass through four separate connections.
604   This distinction is important because some HTTP communication options
605   may apply only to the connection with the nearest, non-tunnel
606   neighbor, only to the end-points of the chain, or to all connections
607   along the chain. Although the diagram is linear, each participant may
608   be engaged in multiple, simultaneous communications. For example, B
609   may be receiving requests from many clients other than A, and/or
610   forwarding requests to servers other than C, at the same time that it
611   is handling A's request.
614   Any party to the communication which is not acting as a tunnel may
615   employ an internal cache for handling requests. The effect of a cache
616   is that the request/response chain is shortened if one of the
617   participants along the chain has a cached response applicable to that
618   request. The following illustrates the resulting chain if B has a
619   cached copy of an earlier response from O (via C) for a request which
620   has not been cached by UA or A.
622<figure><artwork type="drawing">
623          request chain ----------&gt;
624       UA -----v----- A -----v----- B - - - - - - C - - - - - - O
625          &lt;--------- response chain
628   Not all responses are usefully cacheable, and some requests may
629   contain modifiers which place special requirements on cache behavior.
630   HTTP requirements for cache behavior and cacheable responses are
631   defined in &caching;.
634   In fact, there are a wide variety of architectures and configurations
635   of caches and proxies currently being experimented with or deployed
636   across the World Wide Web. These systems include national hierarchies
637   of proxy caches to save transoceanic bandwidth, systems that
638   broadcast or multicast cache entries, organizations that distribute
639   subsets of cached data via CD-ROM, and so on. HTTP systems are used
640   in corporate intranets over high-bandwidth links, and for access via
641   PDAs with low-power radio links and intermittent connectivity. The
642   goal of HTTP/1.1 is to support the wide diversity of configurations
643   already deployed while introducing protocol constructs that meet the
644   needs of those who build web applications that require high
645   reliability and, failing that, at least reliable indications of
646   failure.
649   HTTP communication usually takes place over TCP/IP connections. The
650   default port is TCP 80 (<eref target=""/>), but other ports can be used. This does
651   not preclude HTTP from being implemented on top of any other protocol
652   on the Internet, or on other networks. HTTP only presumes a reliable
653   transport; any protocol that provides such guarantees can be used;
654   the mapping of the HTTP/1.1 request and response structures onto the
655   transport data units of the protocol in question is outside the scope
656   of this specification.
659   In HTTP/1.0, most implementations used a new connection for each
660   request/response exchange. In HTTP/1.1, a connection may be used for
661   one or more request/response exchanges, although connections may be
662   closed for a variety of reasons (see <xref target="persistent.connections"/>).
667<section title="Notational Conventions and Generic Grammar" anchor="notation">
669<section title="Augmented BNF" anchor="notation.abnf">
671   All of the mechanisms specified in this document are described in
672   both prose and an augmented Backus-Naur Form (BNF) similar to that
673   used by <xref target="RFC822ABNF"/>. Implementors will need to be familiar with the
674   notation in order to understand this specification. The augmented BNF
675   includes the following constructs:
678   name = definition
679  <list>
680    <t>
681      The name of a rule is simply the name itself (without any
682      enclosing "&lt;" and "&gt;") and is separated from its definition by the
683      equal "=" character. White space is only significant in that
684      indentation of continuation lines is used to indicate a rule
685      definition that spans more than one line. Certain basic rules are
686      in uppercase, such as SP, LWS, HTAB, CRLF, DIGIT, ALPHA, etc. Angle
687      brackets are used within definitions whenever their presence will
688      facilitate discerning the use of rule names.
689    </t>
690  </list>
693   "literal"
694  <list>
695    <t>
696      Quotation marks surround literal text. Unless stated otherwise,
697      the text is case-insensitive.
698    </t>
699  </list>
702   rule1 | rule2
703  <list>
704    <t>
705      Elements separated by a bar ("|") are alternatives, e.g., "yes |
706      no" will accept yes or no.
707    </t>
708  </list>
711   (rule1 rule2)
712  <list>
713    <t>
714      Elements enclosed in parentheses are treated as a single element.
715      Thus, "(elem (foo | bar) elem)" allows the token sequences "elem
716      foo elem" and "elem bar elem".
717    </t>
718  </list>
721   *rule
722  <list>
723    <t>
724      The character "*" preceding an element indicates repetition. The
725      full form is "&lt;n&gt;*&lt;m&gt;element" indicating at least &lt;n&gt; and at most
726      &lt;m&gt; occurrences of element. Default values are 0 and infinity so
727      that "*(element)" allows any number, including zero; "1*element"
728      requires at least one; and "1*2element" allows one or two.
729    </t>
730  </list>
733   [rule]
734  <list>
735    <t>
736      Square brackets enclose optional elements; "[foo bar]" is
737      equivalent to "*1(foo bar)".
738    </t>
739  </list>
742   N rule
743  <list>
744    <t>
745      Specific repetition: "&lt;n&gt;(element)" is equivalent to
746      "&lt;n&gt;*&lt;n&gt;(element)"; that is, exactly &lt;n&gt; occurrences of (element).
747      Thus 2DIGIT is a 2-digit number, and 3ALPHA is a string of three
748      alphabetic characters.
749    </t>
750  </list>
753   #rule
754  <list>
755    <t>
756      A construct "#" is defined, similar to "*", for defining lists of
757      elements. The full form is "&lt;n&gt;#&lt;m&gt;element" indicating at least
758      &lt;n&gt; and at most &lt;m&gt; elements, each separated by one or more commas
759      (",") and &OPTIONAL; linear white space (LWS). This makes the usual
760      form of lists very easy; a rule such as
761      <figure><artwork type="example">
762   ( *<x:ref>LWS</x:ref> element *( *<x:ref>LWS</x:ref> "," *<x:ref>LWS</x:ref> element ))</artwork></figure>
763    </t>
764    <t>
765      can be shown as
766      <figure><artwork type="example">
767   1#element</artwork></figure>
768    </t>
769    <t>
770      Wherever this construct is used, null elements are allowed, but do
771      not contribute to the count of elements present. That is,
772      "(element), , (element) " is permitted, but counts as only two
773      elements. Therefore, where at least one element is required, at
774      least one non-null element &MUST; be present. Default values are 0
775      and infinity so that "#element" allows any number, including zero;
776      "1#element" requires at least one; and "1#2element" allows one or
777      two.
778    </t>
779  </list>
782   ; comment
783  <list>
784    <t>
785      A semi-colon, set off some distance to the right of rule text,
786      starts a comment that continues to the end of line. This is a
787      simple way of including useful notes in parallel with the
788      specifications.
789    </t>
790  </list>
793   implied *LWS
794  <list>
795    <t>
796      The grammar described by this specification is word-based. Except
797      where noted otherwise, linear white space (LWS) can be included
798      between any two adjacent words (token or quoted-string), and
799      between adjacent words and separators, without changing the
800      interpretation of a field. At least one delimiter (LWS and/or
801      separators) &MUST; exist between any two tokens (for the definition
802      of "token" below), since they would otherwise be interpreted as a
803      single token.
804    </t>
805  </list>
809<section title="Basic Rules" anchor="basic.rules">
810<t anchor="core.rules">
811  <x:anchor-alias value="OCTET"/>
812  <x:anchor-alias value="CHAR"/>
813  <x:anchor-alias value="ALPHA"/>
814  <x:anchor-alias value="DIGIT"/>
815  <x:anchor-alias value="CTL"/>
816  <x:anchor-alias value="CR"/>
817  <x:anchor-alias value="LF"/>
818  <x:anchor-alias value="SP"/>
819  <x:anchor-alias value="HTAB"/>
820  <x:anchor-alias value="DQUOTE"/>
821   The following rules are used throughout this specification to
822   describe basic parsing constructs. The US-ASCII coded character set
823   is defined by ANSI X3.4-1986 <xref target="USASCII"/>.
825<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="OCTET"/><iref primary="true" item="Grammar" subitem="CHAR"/><iref primary="true" item="Grammar" subitem="ALPHA"/><iref primary="true" item="Grammar" subitem="DIGIT"/><iref primary="true" item="Grammar" subitem="CTL"/><iref primary="true" item="Grammar" subitem="CR"/><iref primary="true" item="Grammar" subitem="LF"/><iref primary="true" item="Grammar" subitem="SP"/><iref primary="true" item="Grammar" subitem="HTAB"/><iref primary="true" item="Grammar" subitem="DQUOTE"/>
826  <x:ref>OCTET</x:ref>          = %x00-FF
827                   ; any 8-bit sequence of data
828  <x:ref>CHAR</x:ref>           = %x01-7F
829                   ; any US-ASCII character, excluding NUL
830  <x:ref>ALPHA</x:ref>          = %x41-5A | %x61-7A
831                   ; A-Z | a-z
832  <x:ref>DIGIT</x:ref>          = %x30-39
833                   ; any US-ASCII digit "0".."9"
834  <x:ref>CTL</x:ref>            = %x00-1F | %x7F
835                   ; (octets 0 - 31) and DEL (127)
836  <x:ref>CR</x:ref>             = %x0D
837                   ; US-ASCII CR, carriage return (13)
838  <x:ref>LF</x:ref>             = %x0A
839                   ; US-ASCII LF, linefeed (10)
840  <x:ref>SP</x:ref>             = %x20
841                   ; US-ASCII SP, space (32)
842  <x:ref>HTAB</x:ref>           = %x09
843                   ; US-ASCII HT, horizontal-tab (9)
844  <x:ref>DQUOTE</x:ref>         = %x22
845                   ; US-ASCII double-quote mark (34)
847<t anchor="rule.CRLF">
848  <x:anchor-alias value="CRLF"/>
849   HTTP/1.1 defines the sequence CR LF as the end-of-line marker for all
850   protocol elements except the entity-body (see <xref target="tolerant.applications"/> for
851   tolerant applications). The end-of-line marker within an entity-body
852   is defined by its associated media type, as described in &media-types;.
854<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="CRLF"/>
855  <x:ref>CRLF</x:ref>           = <x:ref>CR</x:ref> LF
857<t anchor="rule.LWS">
858  <x:anchor-alias value="LWS"/>
859   HTTP/1.1 header field values can be folded onto multiple lines if the
860   continuation line begins with a space or horizontal tab. All linear
861   white space, including folding, has the same semantics as SP. A
862   recipient &MAY; replace any linear white space with a single SP before
863   interpreting the field value or forwarding the message downstream.
865<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="LWS"/>
866  <x:ref>LWS</x:ref>            = [<x:ref>CRLF</x:ref>] 1*( <x:ref>SP</x:ref> | <x:ref>HTAB</x:ref> )
868<t anchor="rule.TEXT">
869  <x:anchor-alias value="TEXT"/>
870   The TEXT rule is only used for descriptive field contents and values
871   that are not intended to be interpreted by the message parser. Words
872   of *TEXT &MAY; contain characters from character sets other than ISO-8859-1
873   <xref target="ISO-8859-1"/> only when encoded according to the rules of
874   <xref target="RFC2047"/>.
876<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="TEXT"/>
877  <x:ref>TEXT</x:ref>           = %x20-7E | %x80-FF | <x:ref>LWS</x:ref>
878                 ; any <x:ref>OCTET</x:ref> except <x:ref>CTL</x:ref>s, but including <x:ref>LWS</x:ref>
881   A CRLF is allowed in the definition of TEXT only as part of a header
882   field continuation. It is expected that the folding LWS will be
883   replaced with a single SP before interpretation of the TEXT value.
885<t anchor="rule.HEX">
886  <x:anchor-alias value="HEX"/>
887   Hexadecimal numeric characters are used in several protocol elements.
889<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="HEX"/>
890  <x:ref>HEX</x:ref>            = "A" | "B" | "C" | "D" | "E" | "F"
891                 | "a" | "b" | "c" | "d" | "e" | "f" | <x:ref>DIGIT</x:ref>
893<t anchor="rule.token.separators">
894  <x:anchor-alias value="tchar"/>
895  <x:anchor-alias value="token"/>
896  <x:anchor-alias value="separators"/>
897   Many HTTP/1.1 header field values consist of words separated by LWS
898   or special characters. These special characters &MUST; be in a quoted
899   string to be used within a parameter value (as defined in
900   <xref target="transfer.codings"/>).
902<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="token"/><iref primary="true" item="Grammar" subitem="tchar"/><iref primary="true" item="Grammar" subitem="separators"/>
903  <x:ref>separators</x:ref>     = "(" | ")" | "&lt;" | "&gt;" | "@"
904                 | "," | ";" | ":" | "\" | <x:ref>DQUOTE</x:ref>
905                 | "/" | "[" | "]" | "?" | "="
906                 | "{" | "}" | <x:ref>SP</x:ref> | <x:ref>HTAB</x:ref>
908  <x:ref>tchar</x:ref>          = "!" | "#" | "$" | "%" | "&amp;" | "'" | "*"
909                 | "+" | "-" | "." | "^" | "_" | "`" | "|" | "~"
910                 | <x:ref>DIGIT</x:ref> | <x:ref>ALPHA</x:ref>
911                 ; any <x:ref>CHAR</x:ref> except <x:ref>CTL</x:ref>s or <x:ref>separators</x:ref>
913  <x:ref>token</x:ref>          = 1*<x:ref>tchar</x:ref>
915<t anchor="rule.comment">
916  <x:anchor-alias value="comment"/>
917  <x:anchor-alias value="ctext"/>
918   Comments can be included in some HTTP header fields by surrounding
919   the comment text with parentheses. Comments are only allowed in
920   fields containing "comment" as part of their field value definition.
921   In all other fields, parentheses are considered part of the field
922   value.
924<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="comment"/><iref primary="true" item="Grammar" subitem="ctext"/>
925  <x:ref>comment</x:ref>        = "(" *( <x:ref>ctext</x:ref> | <x:ref>quoted-pair</x:ref> | <x:ref>comment</x:ref> ) ")"
926  <x:ref>ctext</x:ref>          = &lt;any <x:ref>TEXT</x:ref> excluding "(" and ")"&gt;
928<t anchor="rule.quoted-string">
929  <x:anchor-alias value="quoted-string"/>
930  <x:anchor-alias value="qdtext"/>
931   A string of text is parsed as a single word if it is quoted using
932   double-quote marks.
934<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="quoted-string"/><iref primary="true" item="Grammar" subitem="qdtext"/>
935  <x:ref>quoted-string</x:ref>  = ( <x:ref>DQUOTE</x:ref> *(<x:ref>qdtext</x:ref> | <x:ref>quoted-pair</x:ref> ) <x:ref>DQUOTE</x:ref> )
936  <x:ref>qdtext</x:ref>         = &lt;any <x:ref>TEXT</x:ref> excluding <x:ref>DQUOTE</x:ref> and "\">
938<t anchor="rule.quoted-pair">
939  <x:anchor-alias value="quoted-pair"/>
940  <x:anchor-alias value="quoted-text"/>
941   The backslash character ("\") &MAY; be used as a single-character
942   quoting mechanism only within quoted-string and comment constructs.
944<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="quoted-text"/><iref primary="true" item="Grammar" subitem="quoted-pair"/>
945  <x:ref>quoted-text</x:ref>    = %x01-09 |
946                   %x0B-0C |
947                   %x0E-FF ; Characters excluding NUL, <x:ref>CR</x:ref> and <x:ref>LF</x:ref>
948  <x:ref>quoted-pair</x:ref>    = "\" <x:ref>quoted-text</x:ref>
952<section title="ABNF Rules defined in other Parts of the Specification" anchor="abnf.dependencies">
953  <x:anchor-alias value="request-header"/>
954  <x:anchor-alias value="response-header"/>
955  <x:anchor-alias value="accept-params"/>
956  <x:anchor-alias value="entity-body"/>
957  <x:anchor-alias value="entity-header"/>
958  <x:anchor-alias value="Cache-Control"/>
959  <x:anchor-alias value="Pragma"/>
960  <x:anchor-alias value="Warning"/>
962  The ABNF rules below are defined in other parts:
964<figure><!-- Part2--><artwork type="abnf2616">
965  <x:ref>request-header</x:ref>  = &lt;request-header, defined in &request-header-fields;&gt;
966  <x:ref>response-header</x:ref> = &lt;response-header, defined in &response-header-fields;&gt;
968<figure><!-- Part3--><artwork type="abnf2616">
969  <x:ref>accept-params</x:ref>   = &lt;accept-params, defined in &header-accept;&gt;
970  <x:ref>entity-body</x:ref>     = &lt;entity-body, defined in &entity-body;&gt;
971  <x:ref>entity-header</x:ref>   = &lt;entity-header, defined in &entity-header-fields;&gt;
973<figure><!-- Part6--><artwork type="abnf2616">
974  <x:ref>Cache-Control</x:ref>   = &lt;Cache-Control, defined in &header-pragma;&gt;
975  <x:ref>Pragma</x:ref>          = &lt;Pragma, defined in &header-pragma;&gt;
976  <x:ref>Warning</x:ref>         = &lt;Warning, defined in &header-warning;&gt;
982<section title="Protocol Parameters" anchor="protocol.parameters">
984<section title="HTTP Version" anchor="http.version">
985  <x:anchor-alias value="HTTP-Version"/>
987   HTTP uses a "&lt;major&gt;.&lt;minor&gt;" numbering scheme to indicate versions
988   of the protocol. The protocol versioning policy is intended to allow
989   the sender to indicate the format of a message and its capacity for
990   understanding further HTTP communication, rather than the features
991   obtained via that communication. No change is made to the version
992   number for the addition of message components which do not affect
993   communication behavior or which only add to extensible field values.
994   The &lt;minor&gt; number is incremented when the changes made to the
995   protocol add features which do not change the general message parsing
996   algorithm, but which may add to the message semantics and imply
997   additional capabilities of the sender. The &lt;major&gt; number is
998   incremented when the format of a message within the protocol is
999   changed. See <xref target="RFC2145"/> for a fuller explanation.
1002   The version of an HTTP message is indicated by an HTTP-Version field
1003   in the first line of the message. HTTP-Version is case-sensitive.
1005<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="HTTP-Version"/>
1006  <x:ref>HTTP-Version</x:ref>   = "HTTP" "/" 1*<x:ref>DIGIT</x:ref> "." 1*<x:ref>DIGIT</x:ref>
1009   Note that the major and minor numbers &MUST; be treated as separate
1010   integers and that each &MAY; be incremented higher than a single digit.
1011   Thus, HTTP/2.4 is a lower version than HTTP/2.13, which in turn is
1012   lower than HTTP/12.3. Leading zeros &MUST; be ignored by recipients and
1013   &MUST-NOT; be sent.
1016   An application that sends a request or response message that includes
1017   HTTP-Version of "HTTP/1.1" &MUST; be at least conditionally compliant
1018   with this specification. Applications that are at least conditionally
1019   compliant with this specification &SHOULD; use an HTTP-Version of
1020   "HTTP/1.1" in their messages, and &MUST; do so for any message that is
1021   not compatible with HTTP/1.0. For more details on when to send
1022   specific HTTP-Version values, see <xref target="RFC2145"/>.
1025   The HTTP version of an application is the highest HTTP version for
1026   which the application is at least conditionally compliant.
1029   Proxy and gateway applications need to be careful when forwarding
1030   messages in protocol versions different from that of the application.
1031   Since the protocol version indicates the protocol capability of the
1032   sender, a proxy/gateway &MUST-NOT; send a message with a version
1033   indicator which is greater than its actual version. If a higher
1034   version request is received, the proxy/gateway &MUST; either downgrade
1035   the request version, or respond with an error, or switch to tunnel
1036   behavior.
1039   Due to interoperability problems with HTTP/1.0 proxies discovered
1040   since the publication of <xref target="RFC2068"/>, caching proxies &MUST;, gateways
1041   &MAY;, and tunnels &MUST-NOT; upgrade the request to the highest version
1042   they support. The proxy/gateway's response to that request &MUST; be in
1043   the same major version as the request.
1046  <list>
1047    <t>
1048      <x:h>Note:</x:h> Converting between versions of HTTP may involve modification
1049      of header fields required or forbidden by the versions involved.
1050    </t>
1051  </list>
1055<section title="Uniform Resource Identifiers" anchor="uri">
1057   URIs have been known by many names: WWW addresses, Universal Document
1058   Identifiers, Universal Resource Identifiers <xref target="RFC1630"/>, and finally the
1059   combination of Uniform Resource Locators (URL) <xref target="RFC1738"/> and Names (URN)
1060   <xref target="RFC1737"/>. As far as HTTP is concerned, Uniform Resource Identifiers are
1061   simply formatted strings which identify--via name, location, or any
1062   other characteristic--a resource.
1065<section title="General Syntax" anchor="general.syntax">
1066  <x:anchor-alias value="absoluteURI"/>
1067  <x:anchor-alias value="authority"/>
1068  <x:anchor-alias value="fragment"/>
1069  <x:anchor-alias value="path-absolute"/>
1070  <x:anchor-alias value="port"/>
1071  <x:anchor-alias value="query"/>
1072  <x:anchor-alias value="relativeURI"/>
1073  <x:anchor-alias value="uri-host"/>
1075   URIs in HTTP can be represented in absolute form or relative to some
1076   known base URI <xref target="RFC1808"/>, depending upon the context of their use. The two
1077   forms are differentiated by the fact that absolute URIs always begin
1078   with a scheme name followed by a colon. For definitive information on
1079   URL syntax and semantics, see "Uniform Resource Identifiers (URI):
1080   Generic Syntax and Semantics," <xref target="RFC2396"/> (which replaces <xref target="RFC1738"/>
1081   and <xref target="RFC1808"/>). This specification adopts the
1082   definitions of "URI-reference", "absoluteURI", "fragment", "relativeURI", "port",
1083   "host", "abs_path", "query", and "authority" from that specification:
1085<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="absoluteURI"/><iref primary="true" item="Grammar" subitem="authority"/><iref primary="true" item="Grammar" subitem="path-absolute"/><iref primary="true" item="Grammar" subitem="port"/><iref primary="true" item="Grammar" subitem="query"/><iref primary="true" item="Grammar" subitem="relativeURI"/><iref primary="true" item="Grammar" subitem="uri-host"/>
1086  <x:ref>absoluteURI</x:ref>   = &lt;absoluteURI, defined in <xref target="RFC2396" x:fmt="," x:sec="3"/>>
1087  <x:ref>authority</x:ref>     = &lt;authority, defined in <xref target="RFC2396" x:fmt="," x:sec="3.2"/>>
1088  <x:ref>fragment</x:ref>      = &lt;fragment, defined in <xref target="RFC2396" x:fmt="," x:sec="4.1"/>>
1089  <x:ref>path-absolute</x:ref> = &lt;abs_path, defined in <xref target="RFC2396" x:fmt="," x:sec="3"/>>
1090  <x:ref>port</x:ref>          = &lt;port, defined in <xref target="RFC2396" x:fmt="," x:sec="3.2.2"/>>
1091  <x:ref>query</x:ref>         = &lt;query, defined in <xref target="RFC2396" x:fmt="," x:sec="3.4"/>>
1092  <x:ref>relativeURI</x:ref>   = &lt;relativeURI, defined in <xref target="RFC2396" x:fmt="," x:sec="5"/>>
1093  <x:ref>uri-host</x:ref>      = &lt;host, defined in <xref target="RFC2396" x:fmt="," x:sec="3.2.2"/>>
1096   HTTP does not place any a priori limit on the length of
1097   a URI. Servers &MUST; be able to handle the URI of any resource they
1098   serve, and &SHOULD; be able to handle URIs of unbounded length if they
1099   provide GET-based forms that could generate such URIs. A server
1100   &SHOULD; return 414 (Request-URI Too Long) status if a URI is longer
1101   than the server can handle (see &status-414;).
1104  <list>
1105    <t>
1106      <x:h>Note:</x:h> Servers ought to be cautious about depending on URI lengths
1107      above 255 bytes, because some older client or proxy
1108      implementations might not properly support these lengths.
1109    </t>
1110  </list>
1114<section title="http URL" anchor="http.url">
1115  <x:anchor-alias value="http-URL"/>
1117   The "http" scheme is used to locate network resources via the HTTP
1118   protocol. This section defines the scheme-specific syntax and
1119   semantics for http URLs.
1121<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="http-URL"/>
1122  <x:ref>http-URL</x:ref> = "http:" "//" <x:ref>uri-host</x:ref> [ ":" <x:ref>port</x:ref> ]
1123             [ <x:ref>path-absolute</x:ref> [ "?" <x:ref>query</x:ref> ]]
1126   If the port is empty or not given, port 80 is assumed. The semantics
1127   are that the identified resource is located at the server listening
1128   for TCP connections on that port of that host, and the Request-URI
1129   for the resource is path-absolute (<xref target="request-uri"/>). The use of IP addresses
1130   in URLs &SHOULD; be avoided whenever possible (see <xref target="RFC1900"/>). If
1131   the path-absolute is not present in the URL, it &MUST; be given as "/" when
1132   used as a Request-URI for a resource (<xref target="request-uri"/>). If a proxy
1133   receives a host name which is not a fully qualified domain name, it
1134   &MAY; add its domain to the host name it received. If a proxy receives
1135   a fully qualified domain name, the proxy &MUST-NOT; change the host
1136   name.
1140<section title="URI Comparison" anchor="uri.comparison">
1142   When comparing two URIs to decide if they match or not, a client
1143   &SHOULD; use a case-sensitive octet-by-octet comparison of the entire
1144   URIs, with these exceptions:
1145  <list style="symbols">
1146    <t>A port that is empty or not given is equivalent to the default
1147        port for that URI-reference;</t>
1148    <t>Comparisons of host names &MUST; be case-insensitive;</t>
1149    <t>Comparisons of scheme names &MUST; be case-insensitive;</t>
1150    <t>An empty path-absolute is equivalent to an path-absolute of "/".</t>
1151  </list>
1154   Characters other than those in the "reserved" set (see
1155   <xref target="RFC2396"/>) are equivalent to their ""%" HEX HEX" encoding.
1158   For example, the following three URIs are equivalent:
1160<figure><artwork type="example">
1168<section title="Date/Time Formats" anchor="date.time.formats">
1169<section title="Full Date" anchor="">
1170  <x:anchor-alias value="HTTP-date"/>
1171  <x:anchor-alias value="obsolete-date"/>
1172  <x:anchor-alias value="rfc1123-date"/>
1173  <x:anchor-alias value="rfc850-date"/>
1174  <x:anchor-alias value="asctime-date"/>
1175  <x:anchor-alias value="date1"/>
1176  <x:anchor-alias value="date2"/>
1177  <x:anchor-alias value="date3"/>
1178  <x:anchor-alias value="rfc1123-date"/>
1179  <x:anchor-alias value="time"/>
1180  <x:anchor-alias value="wkday"/>
1181  <x:anchor-alias value="weekday"/>
1182  <x:anchor-alias value="month"/>
1184   HTTP applications have historically allowed three different formats
1185   for the representation of date/time stamps:
1187<figure><artwork type="example">
1188   Sun, 06 Nov 1994 08:49:37 GMT  ; RFC 822, updated by RFC 1123
1189   Sunday, 06-Nov-94 08:49:37 GMT ; obsolete RFC 850 format
1190   Sun Nov  6 08:49:37 1994       ; ANSI C's asctime() format
1193   The first format is preferred as an Internet standard and represents
1194   a fixed-length subset of that defined by <xref target="RFC1123"/> (an update to
1195   <xref target="RFC822"/>). The other formats are described here only for
1196   compatibility with obsolete implementations.
1197   HTTP/1.1 clients and servers that parse the date value &MUST; accept
1198   all three formats (for compatibility with HTTP/1.0), though they &MUST;
1199   only generate the RFC 1123 format for representing HTTP-date values
1200   in header fields. See <xref target="tolerant.applications"/> for further information.
1203      <x:h>Note:</x:h> Recipients of date values are encouraged to be robust in
1204      accepting date values that may have been sent by non-HTTP
1205      applications, as is sometimes the case when retrieving or posting
1206      messages via proxies/gateways to SMTP or NNTP.
1209   All HTTP date/time stamps &MUST; be represented in Greenwich Mean Time
1210   (GMT), without exception. For the purposes of HTTP, GMT is exactly
1211   equal to UTC (Coordinated Universal Time). This is indicated in the
1212   first two formats by the inclusion of "GMT" as the three-letter
1213   abbreviation for time zone, and &MUST; be assumed when reading the
1214   asctime format. HTTP-date is case sensitive and &MUST-NOT; include
1215   additional LWS beyond that specifically included as SP in the
1216   grammar.
1218<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="HTTP-date"/><iref primary="true" item="Grammar" subitem="rfc1123-date"/><iref primary="true" item="Grammar" subitem="obsolete-date"/><iref primary="true" item="Grammar" subitem="rfc850-date"/><iref primary="true" item="Grammar" subitem="asctime-date"/><iref primary="true" item="Grammar" subitem="date1"/><iref primary="true" item="Grammar" subitem="date2"/><iref primary="true" item="Grammar" subitem="date3"/><iref primary="true" item="Grammar" subitem="time"/><iref primary="true" item="Grammar" subitem="wkday"/><iref primary="true" item="Grammar" subitem="weekday"/><iref primary="true" item="Grammar" subitem="month"/>
1219  <x:ref>HTTP-date</x:ref>    = <x:ref>rfc1123-date</x:ref> | <x:ref>obsolete-date</x:ref>
1220  <x:ref>obsolete-date</x:ref> = <x:ref>rfc850-date</x:ref> | <x:ref>asctime-date</x:ref>
1221  <x:ref>rfc1123-date</x:ref> = <x:ref>wkday</x:ref> "," <x:ref>SP</x:ref> date1 <x:ref>SP</x:ref> time <x:ref>SP</x:ref> "GMT"
1222  <x:ref>rfc850-date</x:ref>  = <x:ref>weekday</x:ref> "," <x:ref>SP</x:ref> date2 <x:ref>SP</x:ref> time <x:ref>SP</x:ref> "GMT"
1223  <x:ref>asctime-date</x:ref> = <x:ref>wkday</x:ref> <x:ref>SP</x:ref> <x:ref>date3</x:ref> <x:ref>SP</x:ref> <x:ref>time</x:ref> <x:ref>SP</x:ref> 4<x:ref>DIGIT</x:ref>
1224  <x:ref>date1</x:ref>        = 2<x:ref>DIGIT</x:ref> <x:ref>SP</x:ref> <x:ref>month</x:ref> <x:ref>SP</x:ref> 4<x:ref>DIGIT</x:ref>
1225                 ; day month year (e.g., 02 Jun 1982)
1226  <x:ref>date2</x:ref>        = 2<x:ref>DIGIT</x:ref> "-" <x:ref>month</x:ref> "-" 2<x:ref>DIGIT</x:ref>
1227                 ; day-month-year (e.g., 02-Jun-82)
1228  <x:ref>date3</x:ref>        = <x:ref>month</x:ref> <x:ref>SP</x:ref> ( 2<x:ref>DIGIT</x:ref> | ( <x:ref>SP</x:ref> 1<x:ref>DIGIT</x:ref> ))
1229                 ; month day (e.g., Jun  2)
1230  <x:ref>time</x:ref>         = 2<x:ref>DIGIT</x:ref> ":" 2<x:ref>DIGIT</x:ref> ":" 2<x:ref>DIGIT</x:ref>
1231                 ; 00:00:00 - 23:59:59
1232  <x:ref>wkday</x:ref>        = "Mon" | "Tue" | "Wed"
1233               | "Thu" | "Fri" | "Sat" | "Sun"
1234  <x:ref>weekday</x:ref>      = "Monday" | "Tuesday" | "Wednesday"
1235               | "Thursday" | "Friday" | "Saturday" | "Sunday"
1236  <x:ref>month</x:ref>        = "Jan" | "Feb" | "Mar" | "Apr"
1237               | "May" | "Jun" | "Jul" | "Aug"
1238               | "Sep" | "Oct" | "Nov" | "Dec"
1241      <x:h>Note:</x:h> HTTP requirements for the date/time stamp format apply only
1242      to their usage within the protocol stream. Clients and servers are
1243      not required to use these formats for user presentation, request
1244      logging, etc.
1249<section title="Transfer Codings" anchor="transfer.codings">
1250  <x:anchor-alias value="parameter"/>
1251  <x:anchor-alias value="transfer-coding"/>
1252  <x:anchor-alias value="transfer-extension"/>
1254   Transfer-coding values are used to indicate an encoding
1255   transformation that has been, can be, or may need to be applied to an
1256   entity-body in order to ensure "safe transport" through the network.
1257   This differs from a content coding in that the transfer-coding is a
1258   property of the message, not of the original entity.
1260<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="transfer-coding"/><iref primary="true" item="Grammar" subitem="transfer-extension"/>
1261  <x:ref>transfer-coding</x:ref>         = "chunked" | <x:ref>transfer-extension</x:ref>
1262  <x:ref>transfer-extension</x:ref>      = <x:ref>token</x:ref> *( ";" <x:ref>parameter</x:ref> )
1264<t anchor="rule.parameter">
1265  <x:anchor-alias value="attribute"/>
1266  <x:anchor-alias value="parameter"/>
1267  <x:anchor-alias value="value"/>
1268   Parameters are in  the form of attribute/value pairs.
1270<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="parameter"/><iref primary="true" item="Grammar" subitem="attribute"/><iref primary="true" item="Grammar" subitem="value"/>
1271  <x:ref>parameter</x:ref>               = <x:ref>attribute</x:ref> "=" <x:ref>value</x:ref>
1272  <x:ref>attribute</x:ref>               = <x:ref>token</x:ref>
1273  <x:ref>value</x:ref>                   = <x:ref>token</x:ref> | <x:ref>quoted-string</x:ref>
1276   All transfer-coding values are case-insensitive. HTTP/1.1 uses
1277   transfer-coding values in the TE header field (<xref target="header.te"/>) and in
1278   the Transfer-Encoding header field (<xref target="header.transfer-encoding"/>).
1281   Whenever a transfer-coding is applied to a message-body, the set of
1282   transfer-codings &MUST; include "chunked", unless the message is
1283   terminated by closing the connection. When the "chunked" transfer-coding
1284   is used, it &MUST; be the last transfer-coding applied to the
1285   message-body. The "chunked" transfer-coding &MUST-NOT; be applied more
1286   than once to a message-body. These rules allow the recipient to
1287   determine the transfer-length of the message (<xref target="message.length"/>).
1290   Transfer-codings are analogous to the Content-Transfer-Encoding
1291   values of MIME <xref target="RFC2045"/>, which were designed to enable safe transport of
1292   binary data over a 7-bit transport service. However, safe transport
1293   has a different focus for an 8bit-clean transfer protocol. In HTTP,
1294   the only unsafe characteristic of message-bodies is the difficulty in
1295   determining the exact body length (<xref target="message.length"/>), or the desire to
1296   encrypt data over a shared transport.
1299   The Internet Assigned Numbers Authority (IANA) acts as a registry for
1300   transfer-coding value tokens. Initially, the registry contains the
1301   following tokens: "chunked" (<xref target="chunked.transfer.encoding"/>),
1302   "gzip", "compress", and "deflate" (&content-codings;).
1305   New transfer-coding value tokens &SHOULD; be registered in the same way
1306   as new content-coding value tokens (&content-codings;).
1309   A server which receives an entity-body with a transfer-coding it does
1310   not understand &SHOULD; return 501 (Not Implemented), and close the
1311   connection. A server &MUST-NOT; send transfer-codings to an HTTP/1.0
1312   client.
1315<section title="Chunked Transfer Coding" anchor="chunked.transfer.encoding">
1316  <x:anchor-alias value="chunk"/>
1317  <x:anchor-alias value="Chunked-Body"/>
1318  <x:anchor-alias value="chunk-data"/>
1319  <x:anchor-alias value="chunk-extension"/>
1320  <x:anchor-alias value="chunk-ext-name"/>
1321  <x:anchor-alias value="chunk-ext-val"/>
1322  <x:anchor-alias value="chunk-size"/>
1323  <x:anchor-alias value="last-chunk"/>
1324  <x:anchor-alias value="trailer-part"/>
1326   The chunked encoding modifies the body of a message in order to
1327   transfer it as a series of chunks, each with its own size indicator,
1328   followed by an &OPTIONAL; trailer containing entity-header fields. This
1329   allows dynamically produced content to be transferred along with the
1330   information necessary for the recipient to verify that it has
1331   received the full message.
1333<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="Chunked-Body"/><iref primary="true" item="Grammar" subitem="chunk"/><iref primary="true" item="Grammar" subitem="chunk-size"/><iref primary="true" item="Grammar" subitem="last-chunk"/><iref primary="true" item="Grammar" subitem="chunk-extension"/><iref primary="true" item="Grammar" subitem="chunk-ext-name"/><iref primary="true" item="Grammar" subitem="chunk-ext-val"/><iref primary="true" item="Grammar" subitem="chunk-data"/><iref primary="true" item="Grammar" subitem="trailer-part"/>
1334  <x:ref>Chunked-Body</x:ref>   = *<x:ref>chunk</x:ref>
1335                   <x:ref>last-chunk</x:ref>
1336                   <x:ref>trailer-part</x:ref>
1337                   <x:ref>CRLF</x:ref>
1339  <x:ref>chunk</x:ref>          = <x:ref>chunk-size</x:ref> [ <x:ref>chunk-extension</x:ref> ] <x:ref>CRLF</x:ref>
1340                   <x:ref>chunk-data</x:ref> <x:ref>CRLF</x:ref>
1341  <x:ref>chunk-size</x:ref>     = 1*<x:ref>HEX</x:ref>
1342  <x:ref>last-chunk</x:ref>     = 1*("0") [ <x:ref>chunk-extension</x:ref> ] <x:ref>CRLF</x:ref>
1344  <x:ref>chunk-extension</x:ref>= *( ";" <x:ref>chunk-ext-name</x:ref> [ "=" <x:ref>chunk-ext-val</x:ref> ] )
1345  <x:ref>chunk-ext-name</x:ref> = <x:ref>token</x:ref>
1346  <x:ref>chunk-ext-val</x:ref>  = <x:ref>token</x:ref> | <x:ref>quoted-string</x:ref>
1347  <x:ref>chunk-data</x:ref>     = 1*<x:ref>OCTET</x:ref> ; a sequence of chunk-size octets
1348  <x:ref>trailer-part</x:ref>   = *(<x:ref>entity-header</x:ref> <x:ref>CRLF</x:ref>)
1351   The chunk-size field is a string of hex digits indicating the size of
1352   the chunk-data in octets. The chunked encoding is ended by any chunk whose size is
1353   zero, followed by the trailer, which is terminated by an empty line.
1356   The trailer allows the sender to include additional HTTP header
1357   fields at the end of the message. The Trailer header field can be
1358   used to indicate which header fields are included in a trailer (see
1359   <xref target="header.trailer"/>).
1362   A server using chunked transfer-coding in a response &MUST-NOT; use the
1363   trailer for any header fields unless at least one of the following is
1364   true:
1365  <list style="numbers">
1366    <t>the request included a TE header field that indicates "trailers" is
1367     acceptable in the transfer-coding of the  response, as described in
1368     <xref target="header.te"/>; or,</t>
1370    <t>the server is the origin server for the response, the trailer
1371     fields consist entirely of optional metadata, and the recipient
1372     could use the message (in a manner acceptable to the origin server)
1373     without receiving this metadata.  In other words, the origin server
1374     is willing to accept the possibility that the trailer fields might
1375     be silently discarded along the path to the client.</t>
1376  </list>
1379   This requirement prevents an interoperability failure when the
1380   message is being received by an HTTP/1.1 (or later) proxy and
1381   forwarded to an HTTP/1.0 recipient. It avoids a situation where
1382   compliance with the protocol would have necessitated a possibly
1383   infinite buffer on the proxy.
1386   A process for decoding the "chunked" transfer-coding
1387   can be represented in pseudo-code as:
1389<figure><artwork type="code">
1390    length := 0
1391    read chunk-size, chunk-extension (if any) and CRLF
1392    while (chunk-size &gt; 0) {
1393       read chunk-data and CRLF
1394       append chunk-data to entity-body
1395       length := length + chunk-size
1396       read chunk-size and CRLF
1397    }
1398    read entity-header
1399    while (entity-header not empty) {
1400       append entity-header to existing header fields
1401       read entity-header
1402    }
1403    Content-Length := length
1404    Remove "chunked" from Transfer-Encoding
1407   All HTTP/1.1 applications &MUST; be able to receive and decode the
1408   "chunked" transfer-coding, and &MUST; ignore chunk-extension extensions
1409   they do not understand.
1414<section title="Product Tokens" anchor="product.tokens">
1415  <x:anchor-alias value="product"/>
1416  <x:anchor-alias value="product-version"/>
1418   Product tokens are used to allow communicating applications to
1419   identify themselves by software name and version. Most fields using
1420   product tokens also allow sub-products which form a significant part
1421   of the application to be listed, separated by white space. By
1422   convention, the products are listed in order of their significance
1423   for identifying the application.
1425<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="product"/><iref primary="true" item="Grammar" subitem="product-version"/>
1426  <x:ref>product</x:ref>         = <x:ref>token</x:ref> ["/" <x:ref>product-version</x:ref>]
1427  <x:ref>product-version</x:ref> = <x:ref>token</x:ref>
1430   Examples:
1432<figure><artwork type="example">
1433    User-Agent: CERN-LineMode/2.15 libwww/2.17b3
1434    Server: Apache/0.8.4
1437   Product tokens &SHOULD; be short and to the point. They &MUST-NOT; be
1438   used for advertising or other non-essential information. Although any
1439   token character &MAY; appear in a product-version, this token &SHOULD;
1440   only be used for a version identifier (i.e., successive versions of
1441   the same product &SHOULD; only differ in the product-version portion of
1442   the product value).
1448<section title="HTTP Message" anchor="http.message">
1450<section title="Message Types" anchor="message.types">
1451  <x:anchor-alias value="generic-message"/>
1452  <x:anchor-alias value="HTTP-message"/>
1453  <x:anchor-alias value="start-line"/>
1455   HTTP messages consist of requests from client to server and responses
1456   from server to client.
1458<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="HTTP-message"/>
1459  <x:ref>HTTP-message</x:ref>   = <x:ref>Request</x:ref> | <x:ref>Response</x:ref>     ; HTTP/1.1 messages
1462   Request (<xref target="request"/>) and Response (<xref target="response"/>) messages use the generic
1463   message format of <xref target="RFC2822"/> for transferring entities (the payload
1464   of the message). Both types of message consist of a start-line, zero
1465   or more header fields (also known as "headers"), an empty line (i.e.,
1466   a line with nothing preceding the CRLF) indicating the end of the
1467   header fields, and possibly a message-body.
1469<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="generic-message"/><iref primary="true" item="Grammar" subitem="start-line"/>
1470  <x:ref>generic-message</x:ref> = <x:ref>start-line</x:ref>
1471                    *(<x:ref>message-header</x:ref> <x:ref>CRLF</x:ref>)
1472                    <x:ref>CRLF</x:ref>
1473                    [ <x:ref>message-body</x:ref> ]
1474  <x:ref>start-line</x:ref>      = <x:ref>Request-Line</x:ref> | <x:ref>Status-Line</x:ref>
1477   In the interest of robustness, servers &SHOULD; ignore any empty
1478   line(s) received where a Request-Line is expected. In other words, if
1479   the server is reading the protocol stream at the beginning of a
1480   message and receives a CRLF first, it should ignore the CRLF.
1483   Certain buggy HTTP/1.0 client implementations generate extra CRLF's
1484   after a POST request. To restate what is explicitly forbidden by the
1485   BNF, an HTTP/1.1 client &MUST-NOT; preface or follow a request with an
1486   extra CRLF.
1490<section title="Message Headers" anchor="message.headers">
1491  <x:anchor-alias value="field-content"/>
1492  <x:anchor-alias value="field-name"/>
1493  <x:anchor-alias value="field-value"/>
1494  <x:anchor-alias value="message-header"/>
1496   HTTP header fields, which include general-header (<xref target="general.header.fields"/>),
1497   request-header (&request-header-fields;), response-header (&response-header-fields;), and
1498   entity-header (&entity-header-fields;) fields, follow the same generic format as
1499   that given in <xref target="RFC2822" x:fmt="of" x:sec="2.1"/>. Each header field consists
1500   of a name followed by a colon (":") and the field value. Field names
1501   are case-insensitive. The field value &MAY; be preceded by any amount
1502   of LWS, though a single SP is preferred. Header fields can be
1503   extended over multiple lines by preceding each extra line with at
1504   least one SP or HTAB. Applications ought to follow "common form", where
1505   one is known or indicated, when generating HTTP constructs, since
1506   there might exist some implementations that fail to accept anything
1507   beyond the common forms.
1509<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="message-header"/><iref primary="true" item="Grammar" subitem="field-name"/><iref primary="true" item="Grammar" subitem="field-value"/><iref primary="true" item="Grammar" subitem="field-content"/>
1510  <x:ref>message-header</x:ref> = <x:ref>field-name</x:ref> ":" [ <x:ref>field-value</x:ref> ]
1511  <x:ref>field-name</x:ref>     = <x:ref>token</x:ref>
1512  <x:ref>field-value</x:ref>    = *( <x:ref>field-content</x:ref> | <x:ref>LWS</x:ref> )
1513  <x:ref>field-content</x:ref>  = &lt;field content&gt;
1514                   ; the <x:ref>OCTET</x:ref>s making up the field-value
1515                   ; and consisting of either *<x:ref>TEXT</x:ref> or combinations
1516                   ; of <x:ref>token</x:ref>, <x:ref>separators</x:ref>, and <x:ref>quoted-string</x:ref>
1519   The field-content does not include any leading or trailing LWS:
1520   linear white space occurring before the first non-whitespace
1521   character of the field-value or after the last non-whitespace
1522   character of the field-value. Such leading or trailing LWS &MAY; be
1523   removed without changing the semantics of the field value. Any LWS
1524   that occurs between field-content &MAY; be replaced with a single SP
1525   before interpreting the field value or forwarding the message
1526   downstream.
1529   The order in which header fields with differing field names are
1530   received is not significant. However, it is "good practice" to send
1531   general-header fields first, followed by request-header or response-header
1532   fields, and ending with the entity-header fields.
1535   Multiple message-header fields with the same field-name &MAY; be
1536   present in a message if and only if the entire field-value for that
1537   header field is defined as a comma-separated list [i.e., #(values)].
1538   It &MUST; be possible to combine the multiple header fields into one
1539   "field-name: field-value" pair, without changing the semantics of the
1540   message, by appending each subsequent field-value to the first, each
1541   separated by a comma. The order in which header fields with the same
1542   field-name are received is therefore significant to the
1543   interpretation of the combined field value, and thus a proxy &MUST-NOT;
1544   change the order of these field values when a message is forwarded.
1548<section title="Message Body" anchor="message.body">
1549  <x:anchor-alias value="message-body"/>
1551   The message-body (if any) of an HTTP message is used to carry the
1552   entity-body associated with the request or response. The message-body
1553   differs from the entity-body only when a transfer-coding has been
1554   applied, as indicated by the Transfer-Encoding header field (<xref target="header.transfer-encoding"/>).
1556<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="message-body"/>
1557  <x:ref>message-body</x:ref> = <x:ref>entity-body</x:ref>
1558               | &lt;entity-body encoded as per <x:ref>Transfer-Encoding</x:ref>&gt;
1561   Transfer-Encoding &MUST; be used to indicate any transfer-codings
1562   applied by an application to ensure safe and proper transfer of the
1563   message. Transfer-Encoding is a property of the message, not of the
1564   entity, and thus &MAY; be added or removed by any application along the
1565   request/response chain. (However, <xref target="transfer.codings"/> places restrictions on
1566   when certain transfer-codings may be used.)
1569   The rules for when a message-body is allowed in a message differ for
1570   requests and responses.
1573   The presence of a message-body in a request is signaled by the
1574   inclusion of a Content-Length or Transfer-Encoding header field in
1575   the request's message-headers. A message-body &MUST-NOT; be included in
1576   a request if the specification of the request method (&method;)
1577   explicitly disallows an entity-body in requests.
1578   When a request message contains both a message-body of non-zero
1579   length and a method that does not define any semantics for that
1580   request message-body, then an origin server &SHOULD; either ignore
1581   the message-body or respond with an appropriate error message
1582   (e.g., 413).  A proxy or gateway, when presented the same request,
1583   &SHOULD; either forward the request inbound with the message-body or
1584   ignore the message-body when determining a response.
1587   For response messages, whether or not a message-body is included with
1588   a message is dependent on both the request method and the response
1589   status code (<xref target="status.code.and.reason.phrase"/>). All responses to the HEAD request method
1590   &MUST-NOT; include a message-body, even though the presence of entity-header
1591   fields might lead one to believe they do. All 1xx
1592   (informational), 204 (No Content), and 304 (Not Modified) responses
1593   &MUST-NOT; include a message-body. All other responses do include a
1594   message-body, although it &MAY; be of zero length.
1598<section title="Message Length" anchor="message.length">
1600   The transfer-length of a message is the length of the message-body as
1601   it appears in the message; that is, after any transfer-codings have
1602   been applied. When a message-body is included with a message, the
1603   transfer-length of that body is determined by one of the following
1604   (in order of precedence):
1607  <list style="numbers">
1608    <x:lt><t>
1609     Any response message which "&MUST-NOT;" include a message-body (such
1610     as the 1xx, 204, and 304 responses and any response to a HEAD
1611     request) is always terminated by the first empty line after the
1612     header fields, regardless of the entity-header fields present in
1613     the message.
1614    </t></x:lt>
1615    <x:lt><t>
1616     If a Transfer-Encoding header field (<xref target="header.transfer-encoding"/>)
1617     is present, then the transfer-length is
1618     defined by use of the "chunked" transfer-coding (<xref target="transfer.codings"/>),
1619     unless the message is terminated by closing the connection.
1620    </t></x:lt>
1621    <x:lt><t>
1622     If a Content-Length header field (<xref target="header.content-length"/>) is present, its
1623     decimal value in OCTETs represents both the entity-length and the
1624     transfer-length. The Content-Length header field &MUST-NOT; be sent
1625     if these two lengths are different (i.e., if a Transfer-Encoding
1626     header field is present). If a message is received with both a
1627     Transfer-Encoding header field and a Content-Length header field,
1628     the latter &MUST; be ignored.
1629    </t></x:lt>
1630    <x:lt><t>
1631     If the message uses the media type "multipart/byteranges", and the
1632     transfer-length is not otherwise specified, then this self-delimiting
1633     media type defines the transfer-length. This media type
1634     &MUST-NOT; be used unless the sender knows that the recipient can parse
1635     it; the presence in a request of a Range header with multiple byte-range
1636     specifiers from a 1.1 client implies that the client can parse
1637     multipart/byteranges responses.
1638    <list style="empty"><t>
1639       A range header might be forwarded by a 1.0 proxy that does not
1640       understand multipart/byteranges; in this case the server &MUST;
1641       delimit the message using methods defined in items 1, 3 or 5 of
1642       this section.
1643    </t></list>
1644    </t></x:lt>
1645    <x:lt><t>
1646     By the server closing the connection. (Closing the connection
1647     cannot be used to indicate the end of a request body, since that
1648     would leave no possibility for the server to send back a response.)
1649    </t></x:lt>
1650  </list>
1653   For compatibility with HTTP/1.0 applications, HTTP/1.1 requests
1654   containing a message-body &MUST; include a valid Content-Length header
1655   field unless the server is known to be HTTP/1.1 compliant. If a
1656   request contains a message-body and a Content-Length is not given,
1657   the server &SHOULD; respond with 400 (Bad Request) if it cannot
1658   determine the length of the message, or with 411 (Length Required) if
1659   it wishes to insist on receiving a valid Content-Length.
1662   All HTTP/1.1 applications that receive entities &MUST; accept the
1663   "chunked" transfer-coding (<xref target="transfer.codings"/>), thus allowing this mechanism
1664   to be used for messages when the message length cannot be determined
1665   in advance.
1668   Messages &MUST-NOT; include both a Content-Length header field and a
1669   transfer-coding. If the message does include a
1670   transfer-coding, the Content-Length &MUST; be ignored.
1673   When a Content-Length is given in a message where a message-body is
1674   allowed, its field value &MUST; exactly match the number of OCTETs in
1675   the message-body. HTTP/1.1 user agents &MUST; notify the user when an
1676   invalid length is received and detected.
1680<section title="General Header Fields" anchor="general.header.fields">
1681  <x:anchor-alias value="general-header"/>
1683   There are a few header fields which have general applicability for
1684   both request and response messages, but which do not apply to the
1685   entity being transferred. These header fields apply only to the
1686   message being transmitted.
1688<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="general-header"/>
1689  <x:ref>general-header</x:ref> = <x:ref>Cache-Control</x:ref>            ; &header-cache-control;
1690                 | <x:ref>Connection</x:ref>               ; <xref target="header.connection"/>
1691                 | <x:ref>Date</x:ref>                     ; <xref target=""/>
1692                 | <x:ref>Pragma</x:ref>                   ; &header-pragma;
1693                 | <x:ref>Trailer</x:ref>                  ; <xref target="header.trailer"/>
1694                 | <x:ref>Transfer-Encoding</x:ref>        ; <xref target="header.transfer-encoding"/>
1695                 | <x:ref>Upgrade</x:ref>                  ; <xref target="header.upgrade"/>
1696                 | <x:ref>Via</x:ref>                      ; <xref target="header.via"/>
1697                 | <x:ref>Warning</x:ref>                  ; &header-warning;
1700   General-header field names can be extended reliably only in
1701   combination with a change in the protocol version. However, new or
1702   experimental header fields may be given the semantics of general
1703   header fields if all parties in the communication recognize them to
1704   be general-header fields. Unrecognized header fields are treated as
1705   entity-header fields.
1710<section title="Request" anchor="request">
1711  <x:anchor-alias value="Request"/>
1713   A request message from a client to a server includes, within the
1714   first line of that message, the method to be applied to the resource,
1715   the identifier of the resource, and the protocol version in use.
1717<!--                 Host                      ; should be moved here eventually -->
1718<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="Request"/>
1719  <x:ref>Request</x:ref>       = <x:ref>Request-Line</x:ref>              ; <xref target="request-line"/>
1720                  *(( <x:ref>general-header</x:ref>        ; <xref target="general.header.fields"/>
1721                   | <x:ref>request-header</x:ref>         ; &request-header-fields;
1722                   | <x:ref>entity-header</x:ref> ) <x:ref>CRLF</x:ref>)  ; &entity-header-fields;
1723                  <x:ref>CRLF</x:ref>
1724                  [ <x:ref>message-body</x:ref> ]          ; <xref target="message.body"/>
1727<section title="Request-Line" anchor="request-line">
1728  <x:anchor-alias value="Request-Line"/>
1730   The Request-Line begins with a method token, followed by the
1731   Request-URI and the protocol version, and ending with CRLF. The
1732   elements are separated by SP characters. No CR or LF is allowed
1733   except in the final CRLF sequence.
1735<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="Request-Line"/>
1736  <x:ref>Request-Line</x:ref>   = <x:ref>Method</x:ref> <x:ref>SP</x:ref> <x:ref>Request-URI</x:ref> <x:ref>SP</x:ref> <x:ref>HTTP-Version</x:ref> <x:ref>CRLF</x:ref>
1739<section title="Method" anchor="method">
1740  <x:anchor-alias value="Method"/>
1742   The Method  token indicates the method to be performed on the
1743   resource identified by the Request-URI. The method is case-sensitive.
1745<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="Method"/><iref primary="true" item="Grammar" subitem="extension-method"/>
1746  <x:ref>Method</x:ref>         = <x:ref>token</x:ref>
1750<section title="Request-URI" anchor="request-uri">
1751  <x:anchor-alias value="Request-URI"/>
1753   The Request-URI is a Uniform Resource Identifier (<xref target="uri"/>) and
1754   identifies the resource upon which to apply the request.
1756<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="Request-URI"/>
1757  <x:ref>Request-URI</x:ref>    = "*"
1758                 | <x:ref>absoluteURI</x:ref>
1759                 | ( <x:ref>path-absolute</x:ref> [ "?" <x:ref>query</x:ref> ] )
1760                 | <x:ref>authority</x:ref>
1763   The four options for Request-URI are dependent on the nature of the
1764   request. The asterisk "*" means that the request does not apply to a
1765   particular resource, but to the server itself, and is only allowed
1766   when the method used does not necessarily apply to a resource. One
1767   example would be
1769<figure><artwork type="example">
1770    OPTIONS * HTTP/1.1
1773   The absoluteURI form is &REQUIRED; when the request is being made to a
1774   proxy. The proxy is requested to forward the request or service it
1775   from a valid cache, and return the response. Note that the proxy &MAY;
1776   forward the request on to another proxy or directly to the server
1777   specified by the absoluteURI. In order to avoid request loops, a
1778   proxy &MUST; be able to recognize all of its server names, including
1779   any aliases, local variations, and the numeric IP address. An example
1780   Request-Line would be:
1782<figure><artwork type="example">
1783    GET HTTP/1.1
1786   To allow for transition to absoluteURIs in all requests in future
1787   versions of HTTP, all HTTP/1.1 servers &MUST; accept the absoluteURI
1788   form in requests, even though HTTP/1.1 clients will only generate
1789   them in requests to proxies.
1792   The authority form is only used by the CONNECT method (&CONNECT;).
1795   The most common form of Request-URI is that used to identify a
1796   resource on an origin server or gateway. In this case the absolute
1797   path of the URI &MUST; be transmitted (see <xref target="general.syntax"/>, path-absolute) as
1798   the Request-URI, and the network location of the URI (authority) &MUST;
1799   be transmitted in a Host header field. For example, a client wishing
1800   to retrieve the resource above directly from the origin server would
1801   create a TCP connection to port 80 of the host "" and send
1802   the lines:
1804<figure><artwork type="example">
1805    GET /pub/WWW/TheProject.html HTTP/1.1
1806    Host:
1809   followed by the remainder of the Request. Note that the absolute path
1810   cannot be empty; if none is present in the original URI, it &MUST; be
1811   given as "/" (the server root).
1814   The Request-URI is transmitted in the format specified in
1815   <xref target="general.syntax"/>. If the Request-URI is encoded using the "% HEX HEX" encoding
1816   <xref target="RFC2396"/>, the origin server &MUST; decode the Request-URI in order to
1817   properly interpret the request. Servers &SHOULD; respond to invalid
1818   Request-URIs with an appropriate status code.
1821   A transparent proxy &MUST-NOT; rewrite the "path-absolute" part of the
1822   received Request-URI when forwarding it to the next inbound server,
1823   except as noted above to replace a null path-absolute with "/".
1826  <list><t>
1827      <x:h>Note:</x:h> The "no rewrite" rule prevents the proxy from changing the
1828      meaning of the request when the origin server is improperly using
1829      a non-reserved URI character for a reserved purpose.  Implementors
1830      should be aware that some pre-HTTP/1.1 proxies have been known to
1831      rewrite the Request-URI.
1832  </t></list>
1837<section title="The Resource Identified by a Request" anchor="">
1839   The exact resource identified by an Internet request is determined by
1840   examining both the Request-URI and the Host header field.
1843   An origin server that does not allow resources to differ by the
1844   requested host &MAY; ignore the Host header field value when
1845   determining the resource identified by an HTTP/1.1 request. (But see
1846   <xref target=""/>
1847   for other requirements on Host support in HTTP/1.1.)
1850   An origin server that does differentiate resources based on the host
1851   requested (sometimes referred to as virtual hosts or vanity host
1852   names) &MUST; use the following rules for determining the requested
1853   resource on an HTTP/1.1 request:
1854  <list style="numbers">
1855    <t>If Request-URI is an absoluteURI, the host is part of the
1856     Request-URI. Any Host header field value in the request &MUST; be
1857     ignored.</t>
1858    <t>If the Request-URI is not an absoluteURI, and the request includes
1859     a Host header field, the host is determined by the Host header
1860     field value.</t>
1861    <t>If the host as determined by rule 1 or 2 is not a valid host on
1862     the server, the response &MUST; be a 400 (Bad Request) error message.</t>
1863  </list>
1866   Recipients of an HTTP/1.0 request that lacks a Host header field &MAY;
1867   attempt to use heuristics (e.g., examination of the URI path for
1868   something unique to a particular host) in order to determine what
1869   exact resource is being requested.
1876<section title="Response" anchor="response">
1877  <x:anchor-alias value="Response"/>
1879   After receiving and interpreting a request message, a server responds
1880   with an HTTP response message.
1882<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="Response"/>
1883  <x:ref>Response</x:ref>      = <x:ref>Status-Line</x:ref>               ; <xref target="status-line"/>
1884                  *(( <x:ref>general-header</x:ref>        ; <xref target="general.header.fields"/>
1885                   | <x:ref>response-header</x:ref>        ; &response-header-fields;
1886                   | <x:ref>entity-header</x:ref> ) <x:ref>CRLF</x:ref>)  ; &entity-header-fields;
1887                  <x:ref>CRLF</x:ref>
1888                  [ <x:ref>message-body</x:ref> ]          ; <xref target="message.body"/>
1891<section title="Status-Line" anchor="status-line">
1892  <x:anchor-alias value="Status-Line"/>
1894   The first line of a Response message is the Status-Line, consisting
1895   of the protocol version followed by a numeric status code and its
1896   associated textual phrase, with each element separated by SP
1897   characters. No CR or LF is allowed except in the final CRLF sequence.
1899<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="Status-Line"/>
1900  <x:ref>Status-Line</x:ref> = <x:ref>HTTP-Version</x:ref> <x:ref>SP</x:ref> <x:ref>Status-Code</x:ref> <x:ref>SP</x:ref> <x:ref>Reason-Phrase</x:ref> <x:ref>CRLF</x:ref>
1903<section title="Status Code and Reason Phrase" anchor="status.code.and.reason.phrase">
1904  <x:anchor-alias value="Reason-Phrase"/>
1905  <x:anchor-alias value="Status-Code"/>
1907   The Status-Code element is a 3-digit integer result code of the
1908   attempt to understand and satisfy the request. These codes are fully
1909   defined in &status-codes;.  The Reason Phrase exists for the sole
1910   purpose of providing a textual description associated with the numeric
1911   status code, out of deference to earlier Internet application protocols
1912   that were more frequently used with interactive text clients.
1913   A client &SHOULD; ignore the content of the Reason Phrase.
1916   The first digit of the Status-Code defines the class of response. The
1917   last two digits do not have any categorization role. There are 5
1918   values for the first digit:
1919  <list style="symbols">
1920    <t>
1921      1xx: Informational - Request received, continuing process
1922    </t>
1923    <t>
1924      2xx: Success - The action was successfully received,
1925        understood, and accepted
1926    </t>
1927    <t>
1928      3xx: Redirection - Further action must be taken in order to
1929        complete the request
1930    </t>
1931    <t>
1932      4xx: Client Error - The request contains bad syntax or cannot
1933        be fulfilled
1934    </t>
1935    <t>
1936      5xx: Server Error - The server failed to fulfill an apparently
1937        valid request
1938    </t>
1939  </list>
1941<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="Status-Code"/><iref primary="true" item="Grammar" subitem="extension-code"/><iref primary="true" item="Grammar" subitem="Reason-Phrase"/>
1942  <x:ref>Status-Code</x:ref>    = 3<x:ref>DIGIT</x:ref>
1943  <x:ref>Reason-Phrase</x:ref>  = *&lt;<x:ref>TEXT</x:ref>, excluding <x:ref>CR</x:ref>, <x:ref>LF</x:ref>&gt;
1951<section title="Connections" anchor="connections">
1953<section title="Persistent Connections" anchor="persistent.connections">
1955<section title="Purpose" anchor="persistent.purpose">
1957   Prior to persistent connections, a separate TCP connection was
1958   established to fetch each URL, increasing the load on HTTP servers
1959   and causing congestion on the Internet. The use of inline images and
1960   other associated data often require a client to make multiple
1961   requests of the same server in a short amount of time. Analysis of
1962   these performance problems and results from a prototype
1963   implementation are available <xref target="Pad1995"/> <xref target="Spe"/>. Implementation experience and
1964   measurements of actual HTTP/1.1 (<xref target="RFC2068" x:fmt="none">RFC 2068</xref>) implementations show good
1965   results <xref target="Nie1997"/>. Alternatives have also been explored, for example,
1966   T/TCP <xref target="Tou1998"/>.
1969   Persistent HTTP connections have a number of advantages:
1970  <list style="symbols">
1971      <t>
1972        By opening and closing fewer TCP connections, CPU time is saved
1973        in routers and hosts (clients, servers, proxies, gateways,
1974        tunnels, or caches), and memory used for TCP protocol control
1975        blocks can be saved in hosts.
1976      </t>
1977      <t>
1978        HTTP requests and responses can be pipelined on a connection.
1979        Pipelining allows a client to make multiple requests without
1980        waiting for each response, allowing a single TCP connection to
1981        be used much more efficiently, with much lower elapsed time.
1982      </t>
1983      <t>
1984        Network congestion is reduced by reducing the number of packets
1985        caused by TCP opens, and by allowing TCP sufficient time to
1986        determine the congestion state of the network.
1987      </t>
1988      <t>
1989        Latency on subsequent requests is reduced since there is no time
1990        spent in TCP's connection opening handshake.
1991      </t>
1992      <t>
1993        HTTP can evolve more gracefully, since errors can be reported
1994        without the penalty of closing the TCP connection. Clients using
1995        future versions of HTTP might optimistically try a new feature,
1996        but if communicating with an older server, retry with old
1997        semantics after an error is reported.
1998      </t>
1999    </list>
2002   HTTP implementations &SHOULD; implement persistent connections.
2006<section title="Overall Operation" anchor="persistent.overall">
2008   A significant difference between HTTP/1.1 and earlier versions of
2009   HTTP is that persistent connections are the default behavior of any
2010   HTTP connection. That is, unless otherwise indicated, the client
2011   &SHOULD; assume that the server will maintain a persistent connection,
2012   even after error responses from the server.
2015   Persistent connections provide a mechanism by which a client and a
2016   server can signal the close of a TCP connection. This signaling takes
2017   place using the Connection header field (<xref target="header.connection"/>). Once a close
2018   has been signaled, the client &MUST-NOT; send any more requests on that
2019   connection.
2022<section title="Negotiation" anchor="persistent.negotiation">
2024   An HTTP/1.1 server &MAY; assume that a HTTP/1.1 client intends to
2025   maintain a persistent connection unless a Connection header including
2026   the connection-token "close" was sent in the request. If the server
2027   chooses to close the connection immediately after sending the
2028   response, it &SHOULD; send a Connection header including the
2029   connection-token close.
2032   An HTTP/1.1 client &MAY; expect a connection to remain open, but would
2033   decide to keep it open based on whether the response from a server
2034   contains a Connection header with the connection-token close. In case
2035   the client does not want to maintain a connection for more than that
2036   request, it &SHOULD; send a Connection header including the
2037   connection-token close.
2040   If either the client or the server sends the close token in the
2041   Connection header, that request becomes the last one for the
2042   connection.
2045   Clients and servers &SHOULD-NOT;  assume that a persistent connection is
2046   maintained for HTTP versions less than 1.1 unless it is explicitly
2047   signaled. See <xref target="compatibility.with.http.1.0.persistent.connections"/> for more information on backward
2048   compatibility with HTTP/1.0 clients.
2051   In order to remain persistent, all messages on the connection &MUST;
2052   have a self-defined message length (i.e., one not defined by closure
2053   of the connection), as described in <xref target="message.length"/>.
2057<section title="Pipelining" anchor="pipelining">
2059   A client that supports persistent connections &MAY; "pipeline" its
2060   requests (i.e., send multiple requests without waiting for each
2061   response). A server &MUST; send its responses to those requests in the
2062   same order that the requests were received.
2065   Clients which assume persistent connections and pipeline immediately
2066   after connection establishment &SHOULD; be prepared to retry their
2067   connection if the first pipelined attempt fails. If a client does
2068   such a retry, it &MUST-NOT; pipeline before it knows the connection is
2069   persistent. Clients &MUST; also be prepared to resend their requests if
2070   the server closes the connection before sending all of the
2071   corresponding responses.
2074   Clients &SHOULD-NOT;  pipeline requests using non-idempotent methods or
2075   non-idempotent sequences of methods (see &idempotent-methods;). Otherwise, a
2076   premature termination of the transport connection could lead to
2077   indeterminate results. A client wishing to send a non-idempotent
2078   request &SHOULD; wait to send that request until it has received the
2079   response status for the previous request.
2084<section title="Proxy Servers" anchor="persistent.proxy">
2086   It is especially important that proxies correctly implement the
2087   properties of the Connection header field as specified in <xref target="header.connection"/>.
2090   The proxy server &MUST; signal persistent connections separately with
2091   its clients and the origin servers (or other proxy servers) that it
2092   connects to. Each persistent connection applies to only one transport
2093   link.
2096   A proxy server &MUST-NOT; establish a HTTP/1.1 persistent connection
2097   with an HTTP/1.0 client (but see <xref target="RFC2068"/> for information and
2098   discussion of the problems with the Keep-Alive header implemented by
2099   many HTTP/1.0 clients).
2103<section title="Practical Considerations" anchor="persistent.practical">
2105   Servers will usually have some time-out value beyond which they will
2106   no longer maintain an inactive connection. Proxy servers might make
2107   this a higher value since it is likely that the client will be making
2108   more connections through the same server. The use of persistent
2109   connections places no requirements on the length (or existence) of
2110   this time-out for either the client or the server.
2113   When a client or server wishes to time-out it &SHOULD; issue a graceful
2114   close on the transport connection. Clients and servers &SHOULD; both
2115   constantly watch for the other side of the transport close, and
2116   respond to it as appropriate. If a client or server does not detect
2117   the other side's close promptly it could cause unnecessary resource
2118   drain on the network.
2121   A client, server, or proxy &MAY; close the transport connection at any
2122   time. For example, a client might have started to send a new request
2123   at the same time that the server has decided to close the "idle"
2124   connection. From the server's point of view, the connection is being
2125   closed while it was idle, but from the client's point of view, a
2126   request is in progress.
2129   This means that clients, servers, and proxies &MUST; be able to recover
2130   from asynchronous close events. Client software &SHOULD; reopen the
2131   transport connection and retransmit the aborted sequence of requests
2132   without user interaction so long as the request sequence is
2133   idempotent (see &idempotent-methods;). Non-idempotent methods or sequences
2134   &MUST-NOT; be automatically retried, although user agents &MAY; offer a
2135   human operator the choice of retrying the request(s). Confirmation by
2136   user-agent software with semantic understanding of the application
2137   &MAY; substitute for user confirmation. The automatic retry &SHOULD-NOT;
2138   be repeated if the second sequence of requests fails.
2141   Servers &SHOULD; always respond to at least one request per connection,
2142   if at all possible. Servers &SHOULD-NOT;  close a connection in the
2143   middle of transmitting a response, unless a network or client failure
2144   is suspected.
2147   Clients that use persistent connections &SHOULD; limit the number of
2148   simultaneous connections that they maintain to a given server. A
2149   single-user client &SHOULD-NOT; maintain more than 2 connections with
2150   any server or proxy. A proxy &SHOULD; use up to 2*N connections to
2151   another server or proxy, where N is the number of simultaneously
2152   active users. These guidelines are intended to improve HTTP response
2153   times and avoid congestion.
2158<section title="Message Transmission Requirements" anchor="message.transmission.requirements">
2160<section title="Persistent Connections and Flow Control" anchor="persistent.flow">
2162   HTTP/1.1 servers &SHOULD; maintain persistent connections and use TCP's
2163   flow control mechanisms to resolve temporary overloads, rather than
2164   terminating connections with the expectation that clients will retry.
2165   The latter technique can exacerbate network congestion.
2169<section title="Monitoring Connections for Error Status Messages" anchor="persistent.monitor">
2171   An HTTP/1.1 (or later) client sending a message-body &SHOULD; monitor
2172   the network connection for an error status while it is transmitting
2173   the request. If the client sees an error status, it &SHOULD;
2174   immediately cease transmitting the body. If the body is being sent
2175   using a "chunked" encoding (<xref target="transfer.codings"/>), a zero length chunk and
2176   empty trailer &MAY; be used to prematurely mark the end of the message.
2177   If the body was preceded by a Content-Length header, the client &MUST;
2178   close the connection.
2182<section title="Use of the 100 (Continue) Status" anchor="use.of.the.100.status">
2184   The purpose of the 100 (Continue) status (see &status-100;) is to
2185   allow a client that is sending a request message with a request body
2186   to determine if the origin server is willing to accept the request
2187   (based on the request headers) before the client sends the request
2188   body. In some cases, it might either be inappropriate or highly
2189   inefficient for the client to send the body if the server will reject
2190   the message without looking at the body.
2193   Requirements for HTTP/1.1 clients:
2194  <list style="symbols">
2195    <t>
2196        If a client will wait for a 100 (Continue) response before
2197        sending the request body, it &MUST; send an Expect request-header
2198        field (&header-expect;) with the "100-continue" expectation.
2199    </t>
2200    <t>
2201        A client &MUST-NOT; send an Expect request-header field (&header-expect;)
2202        with the "100-continue" expectation if it does not intend
2203        to send a request body.
2204    </t>
2205  </list>
2208   Because of the presence of older implementations, the protocol allows
2209   ambiguous situations in which a client may send "Expect: 100-continue"
2210   without receiving either a 417 (Expectation Failed) status
2211   or a 100 (Continue) status. Therefore, when a client sends this
2212   header field to an origin server (possibly via a proxy) from which it
2213   has never seen a 100 (Continue) status, the client &SHOULD-NOT;  wait
2214   for an indefinite period before sending the request body.
2217   Requirements for HTTP/1.1 origin servers:
2218  <list style="symbols">
2219    <t> Upon receiving a request which includes an Expect request-header
2220        field with the "100-continue" expectation, an origin server &MUST;
2221        either respond with 100 (Continue) status and continue to read
2222        from the input stream, or respond with a final status code. The
2223        origin server &MUST-NOT; wait for the request body before sending
2224        the 100 (Continue) response. If it responds with a final status
2225        code, it &MAY; close the transport connection or it &MAY; continue
2226        to read and discard the rest of the request.  It &MUST-NOT;
2227        perform the requested method if it returns a final status code.
2228    </t>
2229    <t> An origin server &SHOULD-NOT;  send a 100 (Continue) response if
2230        the request message does not include an Expect request-header
2231        field with the "100-continue" expectation, and &MUST-NOT; send a
2232        100 (Continue) response if such a request comes from an HTTP/1.0
2233        (or earlier) client. There is an exception to this rule: for
2234        compatibility with <xref target="RFC2068"/>, a server &MAY; send a 100 (Continue)
2235        status in response to an HTTP/1.1 PUT or POST request that does
2236        not include an Expect request-header field with the "100-continue"
2237        expectation. This exception, the purpose of which is
2238        to minimize any client processing delays associated with an
2239        undeclared wait for 100 (Continue) status, applies only to
2240        HTTP/1.1 requests, and not to requests with any other HTTP-version
2241        value.
2242    </t>
2243    <t> An origin server &MAY; omit a 100 (Continue) response if it has
2244        already received some or all of the request body for the
2245        corresponding request.
2246    </t>
2247    <t> An origin server that sends a 100 (Continue) response &MUST;
2248    ultimately send a final status code, once the request body is
2249        received and processed, unless it terminates the transport
2250        connection prematurely.
2251    </t>
2252    <t> If an origin server receives a request that does not include an
2253        Expect request-header field with the "100-continue" expectation,
2254        the request includes a request body, and the server responds
2255        with a final status code before reading the entire request body
2256        from the transport connection, then the server &SHOULD-NOT;  close
2257        the transport connection until it has read the entire request,
2258        or until the client closes the connection. Otherwise, the client
2259        might not reliably receive the response message. However, this
2260        requirement is not be construed as preventing a server from
2261        defending itself against denial-of-service attacks, or from
2262        badly broken client implementations.
2263      </t>
2264    </list>
2267   Requirements for HTTP/1.1 proxies:
2268  <list style="symbols">
2269    <t> If a proxy receives a request that includes an Expect request-header
2270        field with the "100-continue" expectation, and the proxy
2271        either knows that the next-hop server complies with HTTP/1.1 or
2272        higher, or does not know the HTTP version of the next-hop
2273        server, it &MUST; forward the request, including the Expect header
2274        field.
2275    </t>
2276    <t> If the proxy knows that the version of the next-hop server is
2277        HTTP/1.0 or lower, it &MUST-NOT; forward the request, and it &MUST;
2278        respond with a 417 (Expectation Failed) status.
2279    </t>
2280    <t> Proxies &SHOULD; maintain a cache recording the HTTP version
2281        numbers received from recently-referenced next-hop servers.
2282    </t>
2283    <t> A proxy &MUST-NOT; forward a 100 (Continue) response if the
2284        request message was received from an HTTP/1.0 (or earlier)
2285        client and did not include an Expect request-header field with
2286        the "100-continue" expectation. This requirement overrides the
2287        general rule for forwarding of 1xx responses (see &status-1xx;).
2288    </t>
2289  </list>
2293<section title="Client Behavior if Server Prematurely Closes Connection" anchor="connection.premature">
2295   If an HTTP/1.1 client sends a request which includes a request body,
2296   but which does not include an Expect request-header field with the
2297   "100-continue" expectation, and if the client is not directly
2298   connected to an HTTP/1.1 origin server, and if the client sees the
2299   connection close before receiving any status from the server, the
2300   client &SHOULD; retry the request.  If the client does retry this
2301   request, it &MAY; use the following "binary exponential backoff"
2302   algorithm to be assured of obtaining a reliable response:
2303  <list style="numbers">
2304    <t>
2305      Initiate a new connection to the server
2306    </t>
2307    <t>
2308      Transmit the request-headers
2309    </t>
2310    <t>
2311      Initialize a variable R to the estimated round-trip time to the
2312         server (e.g., based on the time it took to establish the
2313         connection), or to a constant value of 5 seconds if the round-trip
2314         time is not available.
2315    </t>
2316    <t>
2317       Compute T = R * (2**N), where N is the number of previous
2318         retries of this request.
2319    </t>
2320    <t>
2321       Wait either for an error response from the server, or for T
2322         seconds (whichever comes first)
2323    </t>
2324    <t>
2325       If no error response is received, after T seconds transmit the
2326         body of the request.
2327    </t>
2328    <t>
2329       If client sees that the connection is closed prematurely,
2330         repeat from step 1 until the request is accepted, an error
2331         response is received, or the user becomes impatient and
2332         terminates the retry process.
2333    </t>
2334  </list>
2337   If at any point an error status is received, the client
2338  <list style="symbols">
2339      <t>&SHOULD-NOT;  continue and</t>
2341      <t>&SHOULD; close the connection if it has not completed sending the
2342        request message.</t>
2343    </list>
2350<section title="Header Field Definitions" anchor="header.fields">
2352   This section defines the syntax and semantics of HTTP/1.1 header fields
2353   related to message framing and transport protocols.
2356   For entity-header fields, both sender and recipient refer to either the
2357   client or the server, depending on who sends and who receives the entity.
2360<section title="Connection" anchor="header.connection">
2361  <iref primary="true" item="Connection header" x:for-anchor=""/>
2362  <iref primary="true" item="Headers" subitem="Connection" x:for-anchor=""/>
2363  <x:anchor-alias value="Connection"/>
2364  <x:anchor-alias value="connection-token"/>
2366   The Connection general-header field allows the sender to specify
2367   options that are desired for that particular connection and &MUST-NOT;
2368   be communicated by proxies over further connections.
2371   The Connection header has the following grammar:
2373<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="Connection"/><iref primary="true" item="Grammar" subitem="connection-token"/>
2374  <x:ref>Connection</x:ref> = "Connection" ":" 1#(<x:ref>connection-token</x:ref>)
2375  <x:ref>connection-token</x:ref>  = <x:ref>token</x:ref>
2378   HTTP/1.1 proxies &MUST; parse the Connection header field before a
2379   message is forwarded and, for each connection-token in this field,
2380   remove any header field(s) from the message with the same name as the
2381   connection-token. Connection options are signaled by the presence of
2382   a connection-token in the Connection header field, not by any
2383   corresponding additional header field(s), since the additional header
2384   field may not be sent if there are no parameters associated with that
2385   connection option.
2388   Message headers listed in the Connection header &MUST-NOT; include
2389   end-to-end headers, such as Cache-Control.
2392   HTTP/1.1 defines the "close" connection option for the sender to
2393   signal that the connection will be closed after completion of the
2394   response. For example,
2396<figure><artwork type="example">
2397    Connection: close
2400   in either the request or the response header fields indicates that
2401   the connection &SHOULD-NOT;  be considered `persistent' (<xref target="persistent.connections"/>)
2402   after the current request/response is complete.
2405   An HTTP/1.1 client that does not support persistent connections &MUST;
2406   include the "close" connection option in every request message.
2409   An HTTP/1.1 server that does not support persistent connections &MUST;
2410   include the "close" connection option in every response message that
2411   does not have a 1xx (informational) status code.
2414   A system receiving an HTTP/1.0 (or lower-version) message that
2415   includes a Connection header &MUST;, for each connection-token in this
2416   field, remove and ignore any header field(s) from the message with
2417   the same name as the connection-token. This protects against mistaken
2418   forwarding of such header fields by pre-HTTP/1.1 proxies. See <xref target="compatibility.with.http.1.0.persistent.connections"/>.
2422<section title="Content-Length" anchor="header.content-length">
2423  <iref primary="true" item="Content-Length header" x:for-anchor=""/>
2424  <iref primary="true" item="Headers" subitem="Content-Length" x:for-anchor=""/>
2425  <x:anchor-alias value="Content-Length"/>
2427   The Content-Length entity-header field indicates the size of the
2428   entity-body, in decimal number of OCTETs, sent to the recipient or,
2429   in the case of the HEAD method, the size of the entity-body that
2430   would have been sent had the request been a GET.
2432<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="Content-Length"/>
2433  <x:ref>Content-Length</x:ref>    = "Content-Length" ":" 1*<x:ref>DIGIT</x:ref>
2436   An example is
2438<figure><artwork type="example">
2439    Content-Length: 3495
2442   Applications &SHOULD; use this field to indicate the transfer-length of
2443   the message-body, unless this is prohibited by the rules in <xref target="message.length"/>.
2446   Any Content-Length greater than or equal to zero is a valid value.
2447   <xref target="message.length"/> describes how to determine the length of a message-body
2448   if a Content-Length is not given.
2451   Note that the meaning of this field is significantly different from
2452   the corresponding definition in MIME, where it is an optional field
2453   used within the "message/external-body" content-type. In HTTP, it
2454   &SHOULD; be sent whenever the message's length can be determined prior
2455   to being transferred, unless this is prohibited by the rules in
2456   <xref target="message.length"/>.
2460<section title="Date" anchor="">
2461  <iref primary="true" item="Date header" x:for-anchor=""/>
2462  <iref primary="true" item="Headers" subitem="Date" x:for-anchor=""/>
2463  <x:anchor-alias value="Date"/>
2465   The Date general-header field represents the date and time at which
2466   the message was originated, having the same semantics as orig-date in
2467   <xref target="RFC2822" x:fmt="of" x:sec="3.6.1"/>. The field value is an HTTP-date, as described in <xref target=""/>;
2468   it &MUST; be sent in rfc1123-date format.
2470<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="Date"/>
2471  <x:ref>Date</x:ref>  = "Date" ":" <x:ref>HTTP-date</x:ref>
2474   An example is
2476<figure><artwork type="example">
2477    Date: Tue, 15 Nov 1994 08:12:31 GMT
2480   Origin servers &MUST; include a Date header field in all responses,
2481   except in these cases:
2482  <list style="numbers">
2483      <t>If the response status code is 100 (Continue) or 101 (Switching
2484         Protocols), the response &MAY; include a Date header field, at
2485         the server's option.</t>
2487      <t>If the response status code conveys a server error, e.g. 500
2488         (Internal Server Error) or 503 (Service Unavailable), and it is
2489         inconvenient or impossible to generate a valid Date.</t>
2491      <t>If the server does not have a clock that can provide a
2492         reasonable approximation of the current time, its responses
2493         &MUST-NOT; include a Date header field. In this case, the rules
2494         in <xref target="clockless.origin.server.operation"/> &MUST; be followed.</t>
2495  </list>
2498   A received message that does not have a Date header field &MUST; be
2499   assigned one by the recipient if the message will be cached by that
2500   recipient or gatewayed via a protocol which requires a Date. An HTTP
2501   implementation without a clock &MUST-NOT; cache responses without
2502   revalidating them on every use. An HTTP cache, especially a shared
2503   cache, &SHOULD; use a mechanism, such as NTP <xref target="RFC1305"/>, to synchronize its
2504   clock with a reliable external standard.
2507   Clients &SHOULD; only send a Date header field in messages that include
2508   an entity-body, as in the case of the PUT and POST requests, and even
2509   then it is optional. A client without a clock &MUST-NOT; send a Date
2510   header field in a request.
2513   The HTTP-date sent in a Date header &SHOULD-NOT;  represent a date and
2514   time subsequent to the generation of the message. It &SHOULD; represent
2515   the best available approximation of the date and time of message
2516   generation, unless the implementation has no means of generating a
2517   reasonably accurate date and time. In theory, the date ought to
2518   represent the moment just before the entity is generated. In
2519   practice, the date can be generated at any time during the message
2520   origination without affecting its semantic value.
2523<section title="Clockless Origin Server Operation" anchor="clockless.origin.server.operation">
2525   Some origin server implementations might not have a clock available.
2526   An origin server without a clock &MUST-NOT; assign Expires or Last-Modified
2527   values to a response, unless these values were associated
2528   with the resource by a system or user with a reliable clock. It &MAY;
2529   assign an Expires value that is known, at or before server
2530   configuration time, to be in the past (this allows "pre-expiration"
2531   of responses without storing separate Expires values for each
2532   resource).
2537<section title="Host" anchor="">
2538  <iref primary="true" item="Host header" x:for-anchor=""/>
2539  <iref primary="true" item="Headers" subitem="Host" x:for-anchor=""/>
2540  <x:anchor-alias value="Host"/>
2542   The Host request-header field specifies the Internet host and port
2543   number of the resource being requested, as obtained from the original
2544   URI given by the user or referring resource (generally an HTTP URL,
2545   as described in <xref target="http.url"/>). The Host field value &MUST; represent
2546   the naming authority of the origin server or gateway given by the
2547   original URL. This allows the origin server or gateway to
2548   differentiate between internally-ambiguous URLs, such as the root "/"
2549   URL of a server for multiple host names on a single IP address.
2551<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="Host"/>
2552  <x:ref>Host</x:ref> = "Host" ":" <x:ref>uri-host</x:ref> [ ":" <x:ref>port</x:ref> ] ; <xref target="http.url"/>
2555   A "host" without any trailing port information implies the default
2556   port for the service requested (e.g., "80" for an HTTP URL). For
2557   example, a request on the origin server for
2558   &lt;; would properly include:
2560<figure><artwork type="example">
2561    GET /pub/WWW/ HTTP/1.1
2562    Host:
2565   A client &MUST; include a Host header field in all HTTP/1.1 request
2566   messages. If the requested URI does not include an Internet host
2567   name for the service being requested, then the Host header field &MUST;
2568   be given with an empty value. An HTTP/1.1 proxy &MUST; ensure that any
2569   request message it forwards does contain an appropriate Host header
2570   field that identifies the service being requested by the proxy. All
2571   Internet-based HTTP/1.1 servers &MUST; respond with a 400 (Bad Request)
2572   status code to any HTTP/1.1 request message which lacks a Host header
2573   field.
2576   See Sections <xref target="" format="counter"/>
2577   and <xref target="" format="counter"/>
2578   for other requirements relating to Host.
2582<section title="TE" anchor="header.te">
2583  <iref primary="true" item="TE header" x:for-anchor=""/>
2584  <iref primary="true" item="Headers" subitem="TE" x:for-anchor=""/>
2585  <x:anchor-alias value="TE"/>
2586  <x:anchor-alias value="t-codings"/>
2588   The TE request-header field indicates what extension transfer-codings
2589   it is willing to accept in the response and whether or not it is
2590   willing to accept trailer fields in a chunked transfer-coding. Its
2591   value may consist of the keyword "trailers" and/or a comma-separated
2592   list of extension transfer-coding names with optional accept
2593   parameters (as described in <xref target="transfer.codings"/>).
2595<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="TE"/><iref primary="true" item="Grammar" subitem="t-codings"/>
2596  <x:ref>TE</x:ref>        = "TE" ":" #( <x:ref>t-codings</x:ref> )
2597  <x:ref>t-codings</x:ref> = "trailers" | ( <x:ref>transfer-extension</x:ref> [ <x:ref>accept-params</x:ref> ] )
2600   The presence of the keyword "trailers" indicates that the client is
2601   willing to accept trailer fields in a chunked transfer-coding, as
2602   defined in <xref target="chunked.transfer.encoding"/>. This keyword is reserved for use with
2603   transfer-coding values even though it does not itself represent a
2604   transfer-coding.
2607   Examples of its use are:
2609<figure><artwork type="example">
2610    TE: deflate
2611    TE:
2612    TE: trailers, deflate;q=0.5
2615   The TE header field only applies to the immediate connection.
2616   Therefore, the keyword &MUST; be supplied within a Connection header
2617   field (<xref target="header.connection"/>) whenever TE is present in an HTTP/1.1 message.
2620   A server tests whether a transfer-coding is acceptable, according to
2621   a TE field, using these rules:
2622  <list style="numbers">
2623    <x:lt>
2624      <t>The "chunked" transfer-coding is always acceptable. If the
2625         keyword "trailers" is listed, the client indicates that it is
2626         willing to accept trailer fields in the chunked response on
2627         behalf of itself and any downstream clients. The implication is
2628         that, if given, the client is stating that either all
2629         downstream clients are willing to accept trailer fields in the
2630         forwarded response, or that it will attempt to buffer the
2631         response on behalf of downstream recipients.
2632      </t><t>
2633         <x:h>Note:</x:h> HTTP/1.1 does not define any means to limit the size of a
2634         chunked response such that a client can be assured of buffering
2635         the entire response.</t>
2636    </x:lt>
2637    <x:lt>
2638      <t>If the transfer-coding being tested is one of the transfer-codings
2639         listed in the TE field, then it is acceptable unless it
2640         is accompanied by a qvalue of 0. (As defined in &qvalue;, a
2641         qvalue of 0 means "not acceptable.")</t>
2642    </x:lt>
2643    <x:lt>
2644      <t>If multiple transfer-codings are acceptable, then the
2645         acceptable transfer-coding with the highest non-zero qvalue is
2646         preferred.  The "chunked" transfer-coding always has a qvalue
2647         of 1.</t>
2648    </x:lt>
2649  </list>
2652   If the TE field-value is empty or if no TE field is present, the only
2653   transfer-coding  is "chunked". A message with no transfer-coding is
2654   always acceptable.
2658<section title="Trailer" anchor="header.trailer">
2659  <iref primary="true" item="Trailer header" x:for-anchor=""/>
2660  <iref primary="true" item="Headers" subitem="Trailer" x:for-anchor=""/>
2661  <x:anchor-alias value="Trailer"/>
2663   The Trailer general field value indicates that the given set of
2664   header fields is present in the trailer of a message encoded with
2665   chunked transfer-coding.
2667<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="Trailer"/>
2668  <x:ref>Trailer</x:ref>  = "Trailer" ":" 1#<x:ref>field-name</x:ref>
2671   An HTTP/1.1 message &SHOULD; include a Trailer header field in a
2672   message using chunked transfer-coding with a non-empty trailer. Doing
2673   so allows the recipient to know which header fields to expect in the
2674   trailer.
2677   If no Trailer header field is present, the trailer &SHOULD-NOT;  include
2678   any header fields. See <xref target="chunked.transfer.encoding"/> for restrictions on the use of
2679   trailer fields in a "chunked" transfer-coding.
2682   Message header fields listed in the Trailer header field &MUST-NOT;
2683   include the following header fields:
2684  <list style="symbols">
2685    <t>Transfer-Encoding</t>
2686    <t>Content-Length</t>
2687    <t>Trailer</t>
2688  </list>
2692<section title="Transfer-Encoding" anchor="header.transfer-encoding">
2693  <iref primary="true" item="Transfer-Encoding header" x:for-anchor=""/>
2694  <iref primary="true" item="Headers" subitem="Transfer-Encoding" x:for-anchor=""/>
2695  <x:anchor-alias value="Transfer-Encoding"/>
2697   The Transfer-Encoding general-header field indicates what (if any)
2698   type of transformation has been applied to the message body in order
2699   to safely transfer it between the sender and the recipient. This
2700   differs from the content-coding in that the transfer-coding is a
2701   property of the message, not of the entity.
2703<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="Transfer-Encoding"/>
2704  <x:ref>Transfer-Encoding</x:ref>       = "Transfer-Encoding" ":" 1#<x:ref>transfer-coding</x:ref>
2707   Transfer-codings are defined in <xref target="transfer.codings"/>. An example is:
2709<figure><artwork type="example">
2710  Transfer-Encoding: chunked
2713   If multiple encodings have been applied to an entity, the transfer-codings
2714   &MUST; be listed in the order in which they were applied.
2715   Additional information about the encoding parameters &MAY; be provided
2716   by other entity-header fields not defined by this specification.
2719   Many older HTTP/1.0 applications do not understand the Transfer-Encoding
2720   header.
2724<section title="Upgrade" anchor="header.upgrade">
2725  <iref primary="true" item="Upgrade header" x:for-anchor=""/>
2726  <iref primary="true" item="Headers" subitem="Upgrade" x:for-anchor=""/>
2727  <x:anchor-alias value="Upgrade"/>
2729   The Upgrade general-header allows the client to specify what
2730   additional communication protocols it supports and would like to use
2731   if the server finds it appropriate to switch protocols. The server
2732   &MUST; use the Upgrade header field within a 101 (Switching Protocols)
2733   response to indicate which protocol(s) are being switched.
2735<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="Upgrade"/>
2736  <x:ref>Upgrade</x:ref>        = "Upgrade" ":" 1#<x:ref>product</x:ref>
2739   For example,
2741<figure><artwork type="example">
2742    Upgrade: HTTP/2.0, SHTTP/1.3, IRC/6.9, RTA/x11
2745   The Upgrade header field is intended to provide a simple mechanism
2746   for transition from HTTP/1.1 to some other, incompatible protocol. It
2747   does so by allowing the client to advertise its desire to use another
2748   protocol, such as a later version of HTTP with a higher major version
2749   number, even though the current request has been made using HTTP/1.1.
2750   This eases the difficult transition between incompatible protocols by
2751   allowing the client to initiate a request in the more commonly
2752   supported protocol while indicating to the server that it would like
2753   to use a "better" protocol if available (where "better" is determined
2754   by the server, possibly according to the nature of the method and/or
2755   resource being requested).
2758   The Upgrade header field only applies to switching application-layer
2759   protocols upon the existing transport-layer connection. Upgrade
2760   cannot be used to insist on a protocol change; its acceptance and use
2761   by the server is optional. The capabilities and nature of the
2762   application-layer communication after the protocol change is entirely
2763   dependent upon the new protocol chosen, although the first action
2764   after changing the protocol &MUST; be a response to the initial HTTP
2765   request containing the Upgrade header field.
2768   The Upgrade header field only applies to the immediate connection.
2769   Therefore, the upgrade keyword &MUST; be supplied within a Connection
2770   header field (<xref target="header.connection"/>) whenever Upgrade is present in an
2771   HTTP/1.1 message.
2774   The Upgrade header field cannot be used to indicate a switch to a
2775   protocol on a different connection. For that purpose, it is more
2776   appropriate to use a 301, 302, 303, or 305 redirection response.
2779   This specification only defines the protocol name "HTTP" for use by
2780   the family of Hypertext Transfer Protocols, as defined by the HTTP
2781   version rules of <xref target="http.version"/> and future updates to this
2782   specification. Any token can be used as a protocol name; however, it
2783   will only be useful if both the client and server associate the name
2784   with the same protocol.
2788<section title="Via" anchor="header.via">
2789  <iref primary="true" item="Via header" x:for-anchor=""/>
2790  <iref primary="true" item="Headers" subitem="Via" x:for-anchor=""/>
2791  <x:anchor-alias value="protocol-name"/>
2792  <x:anchor-alias value="protocol-version"/>
2793  <x:anchor-alias value="pseudonym"/>
2794  <x:anchor-alias value="received-by"/>
2795  <x:anchor-alias value="received-protocol"/>
2796  <x:anchor-alias value="Via"/>
2798   The Via general-header field &MUST; be used by gateways and proxies to
2799   indicate the intermediate protocols and recipients between the user
2800   agent and the server on requests, and between the origin server and
2801   the client on responses. It is analogous to the "Received" field of
2802   <xref target="RFC2822"/> and is intended to be used for tracking message forwards,
2803   avoiding request loops, and identifying the protocol capabilities of
2804   all senders along the request/response chain.
2806<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="Via"/><iref primary="true" item="Grammar" subitem="received-protocol"/><iref primary="true" item="Grammar" subitem="protocol-name"/><iref primary="true" item="Grammar" subitem="protocol-version"/><iref primary="true" item="Grammar" subitem="received-by"/><iref primary="true" item="Grammar" subitem="pseudonym"/>
2807  <x:ref>Via</x:ref> =  "Via" ":" 1#( <x:ref>received-protocol</x:ref> <x:ref>received-by</x:ref> [ <x:ref>comment</x:ref> ] )
2808  <x:ref>received-protocol</x:ref> = [ <x:ref>protocol-name</x:ref> "/" ] <x:ref>protocol-version</x:ref>
2809  <x:ref>protocol-name</x:ref>     = <x:ref>token</x:ref>
2810  <x:ref>protocol-version</x:ref>  = <x:ref>token</x:ref>
2811  <x:ref>received-by</x:ref>       = ( <x:ref>uri-host</x:ref> [ ":" <x:ref>port</x:ref> ] ) | <x:ref>pseudonym</x:ref>
2812  <x:ref>pseudonym</x:ref>         = <x:ref>token</x:ref>
2815   The received-protocol indicates the protocol version of the message
2816   received by the server or client along each segment of the
2817   request/response chain. The received-protocol version is appended to
2818   the Via field value when the message is forwarded so that information
2819   about the protocol capabilities of upstream applications remains
2820   visible to all recipients.
2823   The protocol-name is optional if and only if it would be "HTTP". The
2824   received-by field is normally the host and optional port number of a
2825   recipient server or client that subsequently forwarded the message.
2826   However, if the real host is considered to be sensitive information,
2827   it &MAY; be replaced by a pseudonym. If the port is not given, it &MAY;
2828   be assumed to be the default port of the received-protocol.
2831   Multiple Via field values represents each proxy or gateway that has
2832   forwarded the message. Each recipient &MUST; append its information
2833   such that the end result is ordered according to the sequence of
2834   forwarding applications.
2837   Comments &MAY; be used in the Via header field to identify the software
2838   of the recipient proxy or gateway, analogous to the User-Agent and
2839   Server header fields. However, all comments in the Via field are
2840   optional and &MAY; be removed by any recipient prior to forwarding the
2841   message.
2844   For example, a request message could be sent from an HTTP/1.0 user
2845   agent to an internal proxy code-named "fred", which uses HTTP/1.1 to
2846   forward the request to a public proxy at, which completes
2847   the request by forwarding it to the origin server at
2848   The request received by would then have the following
2849   Via header field:
2851<figure><artwork type="example">
2852    Via: 1.0 fred, 1.1 (Apache/1.1)
2855   Proxies and gateways used as a portal through a network firewall
2856   &SHOULD-NOT;, by default, forward the names and ports of hosts within
2857   the firewall region. This information &SHOULD; only be propagated if
2858   explicitly enabled. If not enabled, the received-by host of any host
2859   behind the firewall &SHOULD; be replaced by an appropriate pseudonym
2860   for that host.
2863   For organizations that have strong privacy requirements for hiding
2864   internal structures, a proxy &MAY; combine an ordered subsequence of
2865   Via header field entries with identical received-protocol values into
2866   a single such entry. For example,
2868<figure><artwork type="example">
2869    Via: 1.0 ricky, 1.1 ethel, 1.1 fred, 1.0 lucy
2872        could be collapsed to
2874<figure><artwork type="example">
2875    Via: 1.0 ricky, 1.1 mertz, 1.0 lucy
2878   Applications &SHOULD-NOT;  combine multiple entries unless they are all
2879   under the same organizational control and the hosts have already been
2880   replaced by pseudonyms. Applications &MUST-NOT; combine entries which
2881   have different received-protocol values.
2887<section title="IANA Considerations" anchor="IANA.considerations">
2889   <cref>TBD.</cref>
2893<section title="Security Considerations" anchor="security.considerations">
2895   This section is meant to inform application developers, information
2896   providers, and users of the security limitations in HTTP/1.1 as
2897   described by this document. The discussion does not include
2898   definitive solutions to the problems revealed, though it does make
2899   some suggestions for reducing security risks.
2902<section title="Personal Information" anchor="personal.information">
2904   HTTP clients are often privy to large amounts of personal information
2905   (e.g. the user's name, location, mail address, passwords, encryption
2906   keys, etc.), and &SHOULD; be very careful to prevent unintentional
2907   leakage of this information.
2908   We very strongly recommend that a convenient interface be provided
2909   for the user to control dissemination of such information, and that
2910   designers and implementors be particularly careful in this area.
2911   History shows that errors in this area often create serious security
2912   and/or privacy problems and generate highly adverse publicity for the
2913   implementor's company.
2917<section title="Abuse of Server Log Information" anchor="abuse.of.server.log.information">
2919   A server is in the position to save personal data about a user's
2920   requests which might identify their reading patterns or subjects of
2921   interest. This information is clearly confidential in nature and its
2922   handling can be constrained by law in certain countries. People using
2923   HTTP to provide data are responsible for ensuring that
2924   such material is not distributed without the permission of any
2925   individuals that are identifiable by the published results.
2929<section title="Attacks Based On File and Path Names" anchor="attack.pathname">
2931   Implementations of HTTP origin servers &SHOULD; be careful to restrict
2932   the documents returned by HTTP requests to be only those that were
2933   intended by the server administrators. If an HTTP server translates
2934   HTTP URIs directly into file system calls, the server &MUST; take
2935   special care not to serve files that were not intended to be
2936   delivered to HTTP clients. For example, UNIX, Microsoft Windows, and
2937   other operating systems use ".." as a path component to indicate a
2938   directory level above the current one. On such a system, an HTTP
2939   server &MUST; disallow any such construct in the Request-URI if it
2940   would otherwise allow access to a resource outside those intended to
2941   be accessible via the HTTP server. Similarly, files intended for
2942   reference only internally to the server (such as access control
2943   files, configuration files, and script code) &MUST; be protected from
2944   inappropriate retrieval, since they might contain sensitive
2945   information. Experience has shown that minor bugs in such HTTP server
2946   implementations have turned into security risks.
2950<section title="DNS Spoofing" anchor="dns.spoofing">
2952   Clients using HTTP rely heavily on the Domain Name Service, and are
2953   thus generally prone to security attacks based on the deliberate
2954   mis-association of IP addresses and DNS names. Clients need to be
2955   cautious in assuming the continuing validity of an IP number/DNS name
2956   association.
2959   In particular, HTTP clients &SHOULD; rely on their name resolver for
2960   confirmation of an IP number/DNS name association, rather than
2961   caching the result of previous host name lookups. Many platforms
2962   already can cache host name lookups locally when appropriate, and
2963   they &SHOULD; be configured to do so. It is proper for these lookups to
2964   be cached, however, only when the TTL (Time To Live) information
2965   reported by the name server makes it likely that the cached
2966   information will remain useful.
2969   If HTTP clients cache the results of host name lookups in order to
2970   achieve a performance improvement, they &MUST; observe the TTL
2971   information reported by DNS.
2974   If HTTP clients do not observe this rule, they could be spoofed when
2975   a previously-accessed server's IP address changes. As network
2976   renumbering is expected to become increasingly common <xref target="RFC1900"/>, the
2977   possibility of this form of attack will grow. Observing this
2978   requirement thus reduces this potential security vulnerability.
2981   This requirement also improves the load-balancing behavior of clients
2982   for replicated servers using the same DNS name and reduces the
2983   likelihood of a user's experiencing failure in accessing sites which
2984   use that strategy.
2988<section title="Proxies and Caching" anchor="attack.proxies">
2990   By their very nature, HTTP proxies are men-in-the-middle, and
2991   represent an opportunity for man-in-the-middle attacks. Compromise of
2992   the systems on which the proxies run can result in serious security
2993   and privacy problems. Proxies have access to security-related
2994   information, personal information about individual users and
2995   organizations, and proprietary information belonging to users and
2996   content providers. A compromised proxy, or a proxy implemented or
2997   configured without regard to security and privacy considerations,
2998   might be used in the commission of a wide range of potential attacks.
3001   Proxy operators should protect the systems on which proxies run as
3002   they would protect any system that contains or transports sensitive
3003   information. In particular, log information gathered at proxies often
3004   contains highly sensitive personal information, and/or information
3005   about organizations. Log information should be carefully guarded, and
3006   appropriate guidelines for use developed and followed. (<xref target="abuse.of.server.log.information"/>).
3009   Proxy implementors should consider the privacy and security
3010   implications of their design and coding decisions, and of the
3011   configuration options they provide to proxy operators (especially the
3012   default configuration).
3015   Users of a proxy need to be aware that they are no trustworthier than
3016   the people who run the proxy; HTTP itself cannot solve this problem.
3019   The judicious use of cryptography, when appropriate, may suffice to
3020   protect against a broad range of security and privacy attacks. Such
3021   cryptography is beyond the scope of the HTTP/1.1 specification.
3025<section title="Denial of Service Attacks on Proxies" anchor="attack.DoS">
3027   They exist. They are hard to defend against. Research continues.
3028   Beware.
3033<section title="Acknowledgments" anchor="ack">
3035   This specification makes heavy use of the augmented BNF and generic
3036   constructs defined by David H. Crocker for <xref target="RFC822ABNF"/>. Similarly, it
3037   reuses many of the definitions provided by Nathaniel Borenstein and
3038   Ned Freed for MIME <xref target="RFC2045"/>. We hope that their inclusion in this
3039   specification will help reduce past confusion over the relationship
3040   between HTTP and Internet mail message formats.
3043   HTTP has evolved considerably over the years. It has
3044   benefited from a large and active developer community--the many
3045   people who have participated on the www-talk mailing list--and it is
3046   that community which has been most responsible for the success of
3047   HTTP and of the World-Wide Web in general. Marc Andreessen, Robert
3048   Cailliau, Daniel W. Connolly, Bob Denny, John Franks, Jean-Francois
3049   Groff, Phillip M. Hallam-Baker, Hakon W. Lie, Ari Luotonen, Rob
3050   McCool, Lou Montulli, Dave Raggett, Tony Sanders, and Marc
3051   VanHeyningen deserve special recognition for their efforts in
3052   defining early aspects of the protocol.
3055   This document has benefited greatly from the comments of all those
3056   participating in the HTTP-WG. In addition to those already mentioned,
3057   the following individuals have contributed to this specification:
3060   Gary Adams, Harald Tveit Alvestrand, Keith Ball, Brian Behlendorf,
3061   Paul Burchard, Maurizio Codogno, Mike Cowlishaw, Roman Czyborra,
3062   Michael A. Dolan, Daniel DuBois, David J. Fiander, Alan Freier, Marc Hedlund, Greg Herlihy,
3063   Koen Holtman, Alex Hopmann, Bob Jernigan, Shel Kaphan, Rohit Khare,
3064   John Klensin, Martijn Koster, Alexei Kosut, David M. Kristol,
3065   Daniel LaLiberte, Ben Laurie, Paul J. Leach, Albert Lunde,
3066   John C. Mallery, Jean-Philippe Martin-Flatin, Mitra, David Morris,
3067   Gavin Nicol, Ross Patterson, Bill Perry, Jeffrey Perry, Scott Powers, Owen Rees,
3068   Luigi Rizzo, David Robinson, Marc Salomon, Rich Salz,
3069   Allan M. Schiffman, Jim Seidman, Chuck Shotton, Eric W. Sink,
3070   Simon E. Spero, Richard N. Taylor, Robert S. Thau,
3071   Bill (BearHeart) Weinman, Francois Yergeau, Mary Ellen Zurko,
3072   Josh Cohen.
3075   Thanks to the "cave men" of Palo Alto. You know who you are.
3078   Jim Gettys (the editor of <xref target="RFC2616"/>) wishes particularly
3079   to thank Roy Fielding, the editor of <xref target="RFC2068"/>, along
3080   with John Klensin, Jeff Mogul, Paul Leach, Dave Kristol, Koen
3081   Holtman, John Franks, Josh Cohen, Alex Hopmann, Scott Lawrence, and
3082   Larry Masinter for their help. And thanks go particularly to Jeff
3083   Mogul and Scott Lawrence for performing the "MUST/MAY/SHOULD" audit.
3086   The Apache Group, Anselm Baird-Smith, author of Jigsaw, and Henrik
3087   Frystyk implemented RFC 2068 early, and we wish to thank them for the
3088   discovery of many of the problems that this document attempts to
3089   rectify.
3096<references title="Normative References">
3098<reference anchor="ISO-8859-1">
3099  <front>
3100    <title>
3101     Information technology -- 8-bit single-byte coded graphic character sets -- Part 1: Latin alphabet No. 1
3102    </title>
3103    <author>
3104      <organization>International Organization for Standardization</organization>
3105    </author>
3106    <date year="1998"/>
3107  </front>
3108  <seriesInfo name="ISO/IEC" value="8859-1:1998"/>
3111<reference anchor="Part2">
3112  <front>
3113    <title abbrev="HTTP/1.1">HTTP/1.1, part 2: Message Semantics</title>
3114    <author initials="R." surname="Fielding" fullname="Roy T. Fielding" role="editor">
3115      <organization abbrev="Day Software">Day Software</organization>
3116      <address><email></email></address>
3117    </author>
3118    <author initials="J." surname="Gettys" fullname="Jim Gettys">
3119      <organization>One Laptop per Child</organization>
3120      <address><email></email></address>
3121    </author>
3122    <author initials="J." surname="Mogul" fullname="Jeffrey C. Mogul">
3123      <organization abbrev="HP">Hewlett-Packard Company</organization>
3124      <address><email></email></address>
3125    </author>
3126    <author initials="H." surname="Frystyk" fullname="Henrik Frystyk Nielsen">
3127      <organization abbrev="Microsoft">Microsoft Corporation</organization>
3128      <address><email></email></address>
3129    </author>
3130    <author initials="L." surname="Masinter" fullname="Larry Masinter">
3131      <organization abbrev="Adobe Systems">Adobe Systems, Incorporated</organization>
3132      <address><email></email></address>
3133    </author>
3134    <author initials="P." surname="Leach" fullname="Paul J. Leach">
3135      <organization abbrev="Microsoft">Microsoft Corporation</organization>
3136      <address><email></email></address>
3137    </author>
3138    <author initials="T." surname="Berners-Lee" fullname="Tim Berners-Lee">
3139      <organization abbrev="W3C/MIT">World Wide Web Consortium</organization>
3140      <address><email></email></address>
3141    </author>
3142    <author initials="Y." surname="Lafon" fullname="Yves Lafon" role="editor">
3143      <organization abbrev="W3C">World Wide Web Consortium</organization>
3144      <address><email></email></address>
3145    </author>
3146    <author initials="J. F." surname="Reschke" fullname="Julian F. Reschke" role="editor">
3147      <organization abbrev="greenbytes">greenbytes GmbH</organization>
3148      <address><email></email></address>
3149    </author>
3150    <date month="&ID-MONTH;" year="&ID-YEAR;"/>
3151  </front>
3152  <seriesInfo name="Internet-Draft" value="draft-ietf-httpbis-p2-semantics-&ID-VERSION;"/>
3153  <x:source href="p2-semantics.xml" basename="p2-semantics"/>
3156<reference anchor="Part3">
3157  <front>
3158    <title abbrev="HTTP/1.1">HTTP/1.1, part 3: Message Payload and Content Negotiation</title>
3159    <author initials="R." surname="Fielding" fullname="Roy T. Fielding" role="editor">
3160      <organization abbrev="Day Software">Day Software</organization>
3161      <address><email></email></address>
3162    </author>
3163    <author initials="J." surname="Gettys" fullname="Jim Gettys">
3164      <organization>One Laptop per Child</organization>
3165      <address><email></email></address>
3166    </author>
3167    <author initials="J." surname="Mogul" fullname="Jeffrey C. Mogul">
3168      <organization abbrev="HP">Hewlett-Packard Company</organization>
3169      <address><email></email></address>
3170    </author>
3171    <author initials="H." surname="Frystyk" fullname="Henrik Frystyk Nielsen">
3172      <organization abbrev="Microsoft">Microsoft Corporation</organization>
3173      <address><email></email></address>
3174    </author>
3175    <author initials="L." surname="Masinter" fullname="Larry Masinter">
3176      <organization abbrev="Adobe Systems">Adobe Systems, Incorporated</organization>
3177      <address><email></email></address>
3178    </author>
3179    <author initials="P." surname="Leach" fullname="Paul J. Leach">
3180      <organization abbrev="Microsoft">Microsoft Corporation</organization>
3181      <address><email></email></address>
3182    </author>
3183    <author initials="T." surname="Berners-Lee" fullname="Tim Berners-Lee">
3184      <organization abbrev="W3C/MIT">World Wide Web Consortium</organization>
3185      <address><email></email></address>
3186    </author>
3187    <author initials="Y." surname="Lafon" fullname="Yves Lafon" role="editor">
3188      <organization abbrev="W3C">World Wide Web Consortium</organization>
3189      <address><email></email></address>
3190    </author>
3191    <author initials="J. F." surname="Reschke" fullname="Julian F. Reschke" role="editor">
3192      <organization abbrev="greenbytes">greenbytes GmbH</organization>
3193      <address><email></email></address>
3194    </author>
3195    <date month="&ID-MONTH;" year="&ID-YEAR;"/>
3196  </front>
3197  <seriesInfo name="Internet-Draft" value="draft-ietf-httpbis-p3-payload-&ID-VERSION;"/>
3198  <x:source href="p3-payload.xml" basename="p3-payload"/>
3201<reference anchor="Part5">
3202  <front>
3203    <title abbrev="HTTP/1.1">HTTP/1.1, part 5: Range Requests and Partial Responses</title>
3204    <author initials="R." surname="Fielding" fullname="Roy T. Fielding" role="editor">
3205      <organization abbrev="Day Software">Day Software</organization>
3206      <address><email></email></address>
3207    </author>
3208    <author initials="J." surname="Gettys" fullname="Jim Gettys">
3209      <organization>One Laptop per Child</organization>
3210      <address><email></email></address>
3211    </author>
3212    <author initials="J." surname="Mogul" fullname="Jeffrey C. Mogul">
3213      <organization abbrev="HP">Hewlett-Packard Company</organization>
3214      <address><email></email></address>
3215    </author>
3216    <author initials="H." surname="Frystyk" fullname="Henrik Frystyk Nielsen">
3217      <organization abbrev="Microsoft">Microsoft Corporation</organization>
3218      <address><email></email></address>
3219    </author>
3220    <author initials="L." surname="Masinter" fullname="Larry Masinter">
3221      <organization abbrev="Adobe Systems">Adobe Systems, Incorporated</organization>
3222      <address><email></email></address>
3223    </author>
3224    <author initials="P." surname="Leach" fullname="Paul J. Leach">
3225      <organization abbrev="Microsoft">Microsoft Corporation</organization>
3226      <address><email></email></address>
3227    </author>
3228    <author initials="T." surname="Berners-Lee" fullname="Tim Berners-Lee">
3229      <organization abbrev="W3C/MIT">World Wide Web Consortium</organization>
3230      <address><email></email></address>
3231    </author>
3232    <author initials="Y." surname="Lafon" fullname="Yves Lafon" role="editor">
3233      <organization abbrev="W3C">World Wide Web Consortium</organization>
3234      <address><email></email></address>
3235    </author>
3236    <author initials="J. F." surname="Reschke" fullname="Julian F. Reschke" role="editor">
3237      <organization abbrev="greenbytes">greenbytes GmbH</organization>
3238      <address><email></email></address>
3239    </author>
3240    <date month="&ID-MONTH;" year="&ID-YEAR;"/>
3241  </front>
3242  <seriesInfo name="Internet-Draft" value="draft-ietf-httpbis-p5-range-&ID-VERSION;"/>
3243  <x:source href="p5-range.xml" basename="p5-range"/>
3246<reference anchor="Part6">
3247  <front>
3248    <title abbrev="HTTP/1.1">HTTP/1.1, part 6: Caching</title>
3249    <author initials="R." surname="Fielding" fullname="Roy T. Fielding" role="editor">
3250      <organization abbrev="Day Software">Day Software</organization>
3251      <address><email></email></address>
3252    </author>
3253    <author initials="J." surname="Gettys" fullname="Jim Gettys">
3254      <organization>One Laptop per Child</organization>
3255      <address><email></email></address>
3256    </author>
3257    <author initials="J." surname="Mogul" fullname="Jeffrey C. Mogul">
3258      <organization abbrev="HP">Hewlett-Packard Company</organization>
3259      <address><email></email></address>
3260    </author>
3261    <author initials="H." surname="Frystyk" fullname="Henrik Frystyk Nielsen">
3262      <organization abbrev="Microsoft">Microsoft Corporation</organization>
3263      <address><email></email></address>
3264    </author>
3265    <author initials="L." surname="Masinter" fullname="Larry Masinter">
3266      <organization abbrev="Adobe Systems">Adobe Systems, Incorporated</organization>
3267      <address><email></email></address>
3268    </author>
3269    <author initials="P." surname="Leach" fullname="Paul J. Leach">
3270      <organization abbrev="Microsoft">Microsoft Corporation</organization>
3271      <address><email></email></address>
3272    </author>
3273    <author initials="T." surname="Berners-Lee" fullname="Tim Berners-Lee">
3274      <organization abbrev="W3C/MIT">World Wide Web Consortium</organization>
3275      <address><email></email></address>
3276    </author>
3277    <author initials="Y." surname="Lafon" fullname="Yves Lafon" role="editor">
3278      <organization abbrev="W3C">World Wide Web Consortium</organization>
3279      <address><email></email></address>
3280    </author>
3281    <author initials="J. F." surname="Reschke" fullname="Julian F. Reschke" role="editor">
3282      <organization abbrev="greenbytes">greenbytes GmbH</organization>
3283      <address><email></email></address>
3284    </author>
3285    <date month="&ID-MONTH;" year="&ID-YEAR;"/>
3286  </front>
3287  <seriesInfo name="Internet-Draft" value="draft-ietf-httpbis-p6-cache-&ID-VERSION;"/>
3288  <x:source href="p6-cache.xml" basename="p6-cache"/>
3291<reference anchor="RFC822ABNF">
3292  <front>
3293    <title abbrev="Standard for ARPA Internet Text Messages">Standard for the format of ARPA Internet text messages</title>
3294    <author initials="D.H." surname="Crocker" fullname="David H. Crocker">
3295      <organization>University of Delaware, Dept. of Electrical Engineering</organization>
3296      <address><email>DCrocker@UDel-Relay</email></address>
3297    </author>
3298    <date month="August" day="13" year="1982"/>
3299  </front>
3300  <seriesInfo name="STD" value="11"/>
3301  <seriesInfo name="RFC" value="822"/>
3304<reference anchor="RFC2045">
3305  <front>
3306    <title abbrev="Internet Message Bodies">Multipurpose Internet Mail Extensions (MIME) Part One: Format of Internet Message Bodies</title>
3307    <author initials="N." surname="Freed" fullname="Ned Freed">
3308      <organization>Innosoft International, Inc.</organization>
3309      <address><email></email></address>
3310    </author>
3311    <author initials="N.S." surname="Borenstein" fullname="Nathaniel S. Borenstein">
3312      <organization>First Virtual Holdings</organization>
3313      <address><email></email></address>
3314    </author>
3315    <date month="November" year="1996"/>
3316  </front>
3317  <seriesInfo name="RFC" value="2045"/>
3320<reference anchor="RFC2047">
3321  <front>
3322    <title abbrev="Message Header Extensions">MIME (Multipurpose Internet Mail Extensions) Part Three: Message Header Extensions for Non-ASCII Text</title>
3323    <author initials="K." surname="Moore" fullname="Keith Moore">
3324      <organization>University of Tennessee</organization>
3325      <address><email></email></address>
3326    </author>
3327    <date month="November" year="1996"/>
3328  </front>
3329  <seriesInfo name="RFC" value="2047"/>
3332<reference anchor="RFC2119">
3333  <front>
3334    <title>Key words for use in RFCs to Indicate Requirement Levels</title>
3335    <author initials="S." surname="Bradner" fullname="Scott Bradner">
3336      <organization>Harvard University</organization>
3337      <address><email></email></address>
3338    </author>
3339    <date month="March" year="1997"/>
3340  </front>
3341  <seriesInfo name="BCP" value="14"/>
3342  <seriesInfo name="RFC" value="2119"/>
3345<reference anchor="RFC2396">
3346  <front>
3347    <title abbrev="URI Generic Syntax">Uniform Resource Identifiers (URI): Generic Syntax</title>
3348    <author initials="T." surname="Berners-Lee" fullname="Tim Berners-Lee">
3349      <organization abbrev="MIT/LCS">World Wide Web Consortium</organization>
3350      <address><email></email></address>
3351    </author>
3352    <author initials="R.T." surname="Fielding" fullname="Roy T. Fielding">
3353      <organization abbrev="U.C. Irvine">Department of Information and Computer Science</organization>
3354      <address><email></email></address>
3355    </author>
3356    <author initials="L." surname="Masinter" fullname="Larry Masinter">
3357      <organization abbrev="Xerox Corporation">Xerox PARC</organization>
3358      <address><email></email></address>
3359    </author>
3360    <date month="August" year="1998"/>
3361  </front>
3362  <seriesInfo name="RFC" value="2396"/>
3365<reference anchor="USASCII">
3366  <front>
3367    <title>Coded Character Set -- 7-bit American Standard Code for Information Interchange</title>
3368    <author>
3369      <organization>American National Standards Institute</organization>
3370    </author>
3371    <date year="1986"/>
3372  </front>
3373  <seriesInfo name="ANSI" value="X3.4"/>
3378<references title="Informative References">
3380<reference anchor="Nie1997" target="">
3381  <front>
3382    <title>Network Performance Effects of HTTP/1.1, CSS1, and PNG</title>
3383    <author initials="H.F.." surname="Nielsen" fullname="H.F. Nielsen">
3384      <organization/>
3385    </author>
3386    <author initials="J." surname="Gettys" fullname="J. Gettys">
3387      <organization/>
3388    </author>
3389    <author initials="E." surname="Prud'hommeaux" fullname="E. Prud'hommeaux">
3390      <organization/>
3391    </author>
3392    <author initials="H." surname="Lie" fullname="H. Lie">
3393      <organization/>
3394    </author>
3395    <author initials="C." surname="Lilley" fullname="C. Lilley">
3396      <organization/>
3397    </author>
3398    <date year="1997" month="September"/>
3399  </front>
3400  <seriesInfo name="ACM" value="Proceedings of the ACM SIGCOMM '97 conference on Applications, technologies, architectures, and protocols for computer communication SIGCOMM '97"/>
3403<reference anchor="Pad1995">
3404  <front>
3405    <title>Improving HTTP Latency</title>
3406    <author initials="V.N." surname="Padmanabhan" fullname="Venkata N. Padmanabhan">
3407      <organization/>
3408    </author>
3409    <author initials="J.C." surname="Mogul" fullname="Jeffrey C. Mogul">
3410      <organization/>
3411    </author>
3412    <date year="1995" month="December"/>
3413  </front>
3414  <seriesInfo name="Computer Networks and ISDN Systems" value="v. 28, pp. 25-35"/>
3415  <annotation>
3416    Slightly revised version of paper in Proc. 2nd International WWW Conference '94: Mosaic and the Web, Oct. 1994,
3417    which is available at <eref target=""/>.
3418  </annotation>
3421<reference anchor="RFC822">
3422  <front>
3423    <title abbrev="Standard for ARPA Internet Text Messages">Standard for the format of ARPA Internet text messages</title>
3424    <author initials="D.H." surname="Crocker" fullname="David H. Crocker">
3425      <organization>University of Delaware, Dept. of Electrical Engineering</organization>
3426      <address><email>DCrocker@UDel-Relay</email></address>
3427    </author>
3428    <date month="August" day="13" year="1982"/>
3429  </front>
3430  <seriesInfo name="STD" value="11"/>
3431  <seriesInfo name="RFC" value="822"/>
3434<reference anchor="RFC959">
3435  <front>
3436    <title abbrev="File Transfer Protocol">File Transfer Protocol</title>
3437    <author initials="J." surname="Postel" fullname="J. Postel">
3438      <organization>Information Sciences Institute (ISI)</organization>
3439    </author>
3440    <author initials="J." surname="Reynolds" fullname="J. Reynolds">
3441      <organization/>
3442    </author>
3443    <date month="October" year="1985"/>
3444  </front>
3445  <seriesInfo name="STD" value="9"/>
3446  <seriesInfo name="RFC" value="959"/>
3449<reference anchor="RFC1123">
3450  <front>
3451    <title>Requirements for Internet Hosts - Application and Support</title>
3452    <author initials="R." surname="Braden" fullname="Robert Braden">
3453      <organization>University of Southern California (USC), Information Sciences Institute</organization>
3454      <address><email>Braden@ISI.EDU</email></address>
3455    </author>
3456    <date month="October" year="1989"/>
3457  </front>
3458  <seriesInfo name="STD" value="3"/>
3459  <seriesInfo name="RFC" value="1123"/>
3462<reference anchor="RFC1305">
3463  <front>
3464    <title>Network Time Protocol (Version 3) Specification, Implementation</title>
3465    <author initials="D." surname="Mills" fullname="David L. Mills">
3466      <organization>University of Delaware, Electrical Engineering Department</organization>
3467      <address><email></email></address>
3468    </author>
3469    <date month="March" year="1992"/>
3470  </front>
3471  <seriesInfo name="RFC" value="1305"/>
3474<reference anchor="RFC1436">
3475  <front>
3476    <title abbrev="Gopher">The Internet Gopher Protocol (a distributed document search and retrieval protocol)</title>
3477    <author initials="F." surname="Anklesaria" fullname="Farhad Anklesaria">
3478      <organization>University of Minnesota, Computer and Information Services</organization>
3479      <address><email></email></address>
3480    </author>
3481    <author initials="M." surname="McCahill" fullname="Mark McCahill">
3482      <organization>University of Minnesota, Computer and Information Services</organization>
3483      <address><email></email></address>
3484    </author>
3485    <author initials="P." surname="Lindner" fullname="Paul Lindner">
3486      <organization>University of Minnesota, Computer and Information Services</organization>
3487      <address><email></email></address>
3488    </author>
3489    <author initials="D." surname="Johnson" fullname="David Johnson">
3490      <organization>University of Minnesota, Computer and Information Services</organization>
3491      <address><email></email></address>
3492    </author>
3493    <author initials="D." surname="Torrey" fullname="Daniel Torrey">
3494      <organization>University of Minnesota, Computer and Information Services</organization>
3495      <address><email></email></address>
3496    </author>
3497    <author initials="B." surname="Alberti" fullname="Bob Alberti">
3498      <organization>University of Minnesota, Computer and Information Services</organization>
3499      <address><email></email></address>
3500    </author>
3501    <date month="March" year="1993"/>
3502  </front>
3503  <seriesInfo name="RFC" value="1436"/>
3506<reference anchor="RFC1630">
3507  <front>
3508    <title abbrev="URIs in WWW">Universal Resource Identifiers in WWW: A Unifying Syntax for the Expression of Names and Addresses of Objects on the Network as used in the World-Wide Web</title>
3509    <author initials="T." surname="Berners-Lee" fullname="Tim Berners-Lee">
3510      <organization>CERN, World-Wide Web project</organization>
3511      <address><email></email></address>
3512    </author>
3513    <date month="June" year="1994"/>
3514  </front>
3515  <seriesInfo name="RFC" value="1630"/>
3518<reference anchor="RFC1737">
3519  <front>
3520    <title abbrev="Requirements for Uniform Resource Names">Functional Requirements for Uniform Resource Names</title>
3521    <author initials="L." surname="Masinter" fullname="Larry Masinter">
3522      <organization>Xerox Palo Alto Research Center</organization>
3523      <address><email></email></address>
3524    </author>
3525    <author initials="K." surname="Sollins" fullname="Karen Sollins">
3526      <organization>MIT Laboratory for Computer Science</organization>
3527      <address><email></email></address>
3528    </author>
3529    <date month="December" year="1994"/>
3530  </front>
3531  <seriesInfo name="RFC" value="1737"/>
3534<reference anchor="RFC1738">
3535  <front>
3536    <title>Uniform Resource Locators (URL)</title>
3537    <author initials="T." surname="Berners-Lee" fullname="Tim Berners-Lee">
3538      <organization>CERN, World-Wide Web project</organization>
3539      <address><email></email></address>
3540    </author>
3541    <author initials="L." surname="Masinter" fullname="Larry Masinter">
3542      <organization>Xerox PARC</organization>
3543      <address><email></email></address>
3544    </author>
3545    <author initials="M." surname="McCahill" fullname="Mark McCahill">
3546      <organization>University of Minnesota, Computer and Information Services</organization>
3547      <address><email></email></address>
3548    </author>
3549    <date month="December" year="1994"/>
3550  </front>
3551  <seriesInfo name="RFC" value="1738"/>
3554<reference anchor="RFC1808">
3555  <front>
3556    <title>Relative Uniform Resource Locators</title>
3557    <author initials="R." surname="Fielding" fullname="Roy T. Fielding">
3558      <organization>University of California Irvine, Department of Information and Computer Science</organization>
3559      <address><email></email></address>
3560    </author>
3561    <date month="June" year="1995"/>
3562  </front>
3563  <seriesInfo name="RFC" value="1808"/>
3566<reference anchor="RFC1900">
3567  <front>
3568    <title>Renumbering Needs Work</title>
3569    <author initials="B." surname="Carpenter" fullname="Brian E. Carpenter">
3570      <organization>CERN, Computing and Networks Division</organization>
3571      <address><email></email></address>
3572    </author>
3573    <author initials="Y." surname="Rekhter" fullname="Yakov Rekhter">
3574      <organization>cisco Systems</organization>
3575      <address><email></email></address>
3576    </author>
3577    <date month="February" year="1996"/>
3578  </front>
3579  <seriesInfo name="RFC" value="1900"/>
3582<reference anchor="RFC1945">
3583  <front>
3584    <title abbrev="HTTP/1.0">Hypertext Transfer Protocol -- HTTP/1.0</title>
3585    <author initials="T." surname="Berners-Lee" fullname="Tim Berners-Lee">
3586      <organization>MIT, Laboratory for Computer Science</organization>
3587      <address><email></email></address>
3588    </author>
3589    <author initials="R.T." surname="Fielding" fullname="Roy T. Fielding">
3590      <organization>University of California, Irvine, Department of Information and Computer Science</organization>
3591      <address><email></email></address>
3592    </author>
3593    <author initials="H.F." surname="Nielsen" fullname="Henrik Frystyk Nielsen">
3594      <organization>W3 Consortium, MIT Laboratory for Computer Science</organization>
3595      <address><email></email></address>
3596    </author>
3597    <date month="May" year="1996"/>
3598  </front>
3599  <seriesInfo name="RFC" value="1945"/>
3602<reference anchor="RFC2068">
3603  <front>
3604    <title abbrev="HTTP/1.1">Hypertext Transfer Protocol -- HTTP/1.1</title>
3605    <author initials="R." surname="Fielding" fullname="Roy T. Fielding">
3606      <organization>University of California, Irvine, Department of Information and Computer Science</organization>
3607      <address><email></email></address>
3608    </author>
3609    <author initials="J." surname="Gettys" fullname="Jim Gettys">
3610      <organization>MIT Laboratory for Computer Science</organization>
3611      <address><email></email></address>
3612    </author>
3613    <author initials="J." surname="Mogul" fullname="Jeffrey C. Mogul">
3614      <organization>Digital Equipment Corporation, Western Research Laboratory</organization>
3615      <address><email></email></address>
3616    </author>
3617    <author initials="H." surname="Nielsen" fullname="Henrik Frystyk Nielsen">
3618      <organization>MIT Laboratory for Computer Science</organization>
3619      <address><email></email></address>
3620    </author>
3621    <author initials="T." surname="Berners-Lee" fullname="Tim Berners-Lee">
3622      <organization>MIT Laboratory for Computer Science</organization>
3623      <address><email></email></address>
3624    </author>
3625    <date month="January" year="1997"/>
3626  </front>
3627  <seriesInfo name="RFC" value="2068"/>
3630<reference anchor="RFC2145">
3631  <front>
3632    <title abbrev="HTTP Version Numbers">Use and Interpretation of HTTP Version Numbers</title>
3633    <author initials="J.C." surname="Mogul" fullname="Jeffrey C. Mogul">
3634      <organization>Western Research Laboratory</organization>
3635      <address><email></email></address>
3636    </author>
3637    <author initials="R.T." surname="Fielding" fullname="Roy T. Fielding">
3638      <organization>Department of Information and Computer Science</organization>
3639      <address><email></email></address>
3640    </author>
3641    <author initials="J." surname="Gettys" fullname="Jim Gettys">
3642      <organization>MIT Laboratory for Computer Science</organization>
3643      <address><email></email></address>
3644    </author>
3645    <author initials="H.F." surname="Nielsen" fullname="Henrik Frystyk Nielsen">
3646      <organization>W3 Consortium</organization>
3647      <address><email></email></address>
3648    </author>
3649    <date month="May" year="1997"/>
3650  </front>
3651  <seriesInfo name="RFC" value="2145"/>
3654<reference anchor="RFC2324">
3655  <front>
3656    <title abbrev="HTCPCP/1.0">Hyper Text Coffee Pot Control Protocol (HTCPCP/1.0)</title>
3657    <author initials="L." surname="Masinter" fullname="Larry Masinter">
3658      <organization>Xerox Palo Alto Research Center</organization>
3659      <address><email></email></address>
3660    </author>
3661    <date month="April" day="1" year="1998"/>
3662  </front>
3663  <seriesInfo name="RFC" value="2324"/>
3666<reference anchor="RFC2616">
3667  <front>
3668    <title>Hypertext Transfer Protocol -- HTTP/1.1</title>
3669    <author initials="R." surname="Fielding" fullname="R. Fielding">
3670      <organization>University of California, Irvine</organization>
3671      <address><email></email></address>
3672    </author>
3673    <author initials="J." surname="Gettys" fullname="J. Gettys">
3674      <organization>W3C</organization>
3675      <address><email></email></address>
3676    </author>
3677    <author initials="J." surname="Mogul" fullname="J. Mogul">
3678      <organization>Compaq Computer Corporation</organization>
3679      <address><email></email></address>
3680    </author>
3681    <author initials="H." surname="Frystyk" fullname="H. Frystyk">
3682      <organization>MIT Laboratory for Computer Science</organization>
3683      <address><email></email></address>
3684    </author>
3685    <author initials="L." surname="Masinter" fullname="L. Masinter">
3686      <organization>Xerox Corporation</organization>
3687      <address><email></email></address>
3688    </author>
3689    <author initials="P." surname="Leach" fullname="P. Leach">
3690      <organization>Microsoft Corporation</organization>
3691      <address><email></email></address>
3692    </author>
3693    <author initials="T." surname="Berners-Lee" fullname="T. Berners-Lee">
3694      <organization>W3C</organization>
3695      <address><email></email></address>
3696    </author>
3697    <date month="June" year="1999"/>
3698  </front>
3699  <seriesInfo name="RFC" value="2616"/>
3702<reference anchor="RFC2821">
3703  <front>
3704    <title>Simple Mail Transfer Protocol</title>
3705    <author initials="J." surname="Klensin" fullname="J. Klensin">
3706      <organization>AT&amp;T Laboratories</organization>
3707      <address><email></email></address>
3708    </author>
3709    <date year="2001" month="April"/>
3710  </front>
3711  <seriesInfo name="RFC" value="2821"/>
3714<reference anchor="RFC2822">
3715  <front>
3716    <title>Internet Message Format</title>
3717    <author initials="P." surname="Resnick" fullname="P. Resnick">
3718      <organization>QUALCOMM Incorporated</organization>
3719    </author>
3720    <date year="2001" month="April"/>
3721  </front>
3722  <seriesInfo name="RFC" value="2822"/>
3725<reference anchor='RFC3977'>
3726  <front>
3727    <title>Network News Transfer Protocol (NNTP)</title>
3728    <author initials='C.' surname='Feather' fullname='C. Feather'>
3729      <organization>THUS plc</organization>
3730      <address><email></email></address>
3731    </author>
3732    <date year='2006' month='October' />
3733  </front>
3734  <seriesInfo name="RFC" value="3977"/>
3737<reference anchor="RFC4288">
3738  <front>
3739    <title>Media Type Specifications and Registration Procedures</title>
3740    <author initials="N." surname="Freed" fullname="N. Freed">
3741      <organization>Sun Microsystems</organization>
3742      <address>
3743        <email></email>
3744      </address>
3745    </author>
3746    <author initials="J." surname="Klensin" fullname="J. Klensin">
3747      <organization/>
3748      <address>
3749        <email></email>
3750      </address>
3751    </author>
3752    <date year="2005" month="December"/>
3753  </front>
3754  <seriesInfo name="BCP" value="13"/>
3755  <seriesInfo name="RFC" value="4288"/>
3758<reference anchor="Spe" target="">
3759  <front>
3760  <title>Analysis of HTTP Performance Problems</title>
3761  <author initials="S." surname="Spero" fullname="Simon E. Spero">
3762    <organization/>
3763  </author>
3764  <date/>
3765  </front>
3768<reference anchor="Tou1998" target="">
3769  <front>
3770  <title>Analysis of HTTP Performance</title>
3771  <author initials="J." surname="Touch" fullname="Joe Touch">
3772    <organization>USC/Information Sciences Institute</organization>
3773    <address><email></email></address>
3774  </author>
3775  <author initials="J." surname="Heidemann" fullname="John Heidemann">
3776    <organization>USC/Information Sciences Institute</organization>
3777    <address><email></email></address>
3778  </author>
3779  <author initials="K." surname="Obraczka" fullname="Katia Obraczka">
3780    <organization>USC/Information Sciences Institute</organization>
3781    <address><email></email></address>
3782  </author>
3783  <date year="1998" month="Aug"/>
3784  </front>
3785  <seriesInfo name="ISI Research Report" value="ISI/RR-98-463"/>
3786  <annotation>(original report dated Aug. 1996)</annotation>
3789<reference anchor="WAIS">
3790  <front>
3791    <title>WAIS Interface Protocol Prototype Functional Specification (v1.5)</title>
3792    <author initials="F." surname="Davis" fullname="F. Davis">
3793      <organization>Thinking Machines Corporation</organization>
3794    </author>
3795    <author initials="B." surname="Kahle" fullname="B. Kahle">
3796      <organization>Thinking Machines Corporation</organization>
3797    </author>
3798    <author initials="H." surname="Morris" fullname="H. Morris">
3799      <organization>Thinking Machines Corporation</organization>
3800    </author>
3801    <author initials="J." surname="Salem" fullname="J. Salem">
3802      <organization>Thinking Machines Corporation</organization>
3803    </author>
3804    <author initials="T." surname="Shen" fullname="T. Shen">
3805      <organization>Thinking Machines Corporation</organization>
3806    </author>
3807    <author initials="R." surname="Wang" fullname="R. Wang">
3808      <organization>Thinking Machines Corporation</organization>
3809    </author>
3810    <author initials="J." surname="Sui" fullname="J. Sui">
3811      <organization>Thinking Machines Corporation</organization>
3812    </author>
3813    <author initials="M." surname="Grinbaum" fullname="M. Grinbaum">
3814      <organization>Thinking Machines Corporation</organization>
3815    </author>
3816    <date month="April" year="1990"/>
3817  </front>
3818  <seriesInfo name="Thinking Machines Corporation" value=""/>
3824<section title="Internet Media Types" anchor="">
3826   In addition to defining HTTP/1.1, this document serves
3827   as the specification for the Internet media type "message/http" and
3828   "application/http". The following is to be registered with IANA <xref target="RFC4288"/>.
3830<section title="Internet Media Type message/http" anchor="">
3831<iref item="Media Type" subitem="message/http" primary="true"/>
3832<iref item="message/http Media Type" primary="true"/>
3834   The message/http type can be used to enclose a single HTTP request or
3835   response message, provided that it obeys the MIME restrictions for all
3836   "message" types regarding line length and encodings.
3839  <list style="hanging" x:indent="12em">
3840    <t hangText="Type name:">
3841      message
3842    </t>
3843    <t hangText="Subtype name:">
3844      http
3845    </t>
3846    <t hangText="Required parameters:">
3847      none
3848    </t>
3849    <t hangText="Optional parameters:">
3850      version, msgtype
3851      <list style="hanging">
3852        <t hangText="version:">
3853          The HTTP-Version number of the enclosed message
3854          (e.g., "1.1"). If not present, the version can be
3855          determined from the first line of the body.
3856        </t>
3857        <t hangText="msgtype:">
3858          The message type -- "request" or "response". If not
3859          present, the type can be determined from the first
3860          line of the body.
3861        </t>
3862      </list>
3863    </t>
3864    <t hangText="Encoding considerations:">
3865      only "7bit", "8bit", or "binary" are permitted
3866    </t>
3867    <t hangText="Security considerations:">
3868      none
3869    </t>
3870    <t hangText="Interoperability considerations:">
3871      none
3872    </t>
3873    <t hangText="Published specification:">
3874      This specification (see <xref target=""/>).
3875    </t>
3876    <t hangText="Applications that use this media type:">
3877    </t>
3878    <t hangText="Additional information:">
3879      <list style="hanging">
3880        <t hangText="Magic number(s):">none</t>
3881        <t hangText="File extension(s):">none</t>
3882        <t hangText="Macintosh file type code(s):">none</t>
3883      </list>
3884    </t>
3885    <t hangText="Person and email address to contact for further information:">
3886      See Authors Section.
3887    </t>
3888                <t hangText="Intended usage:">
3889                  COMMON
3890    </t>
3891                <t hangText="Restrictions on usage:">
3892                  none
3893    </t>
3894    <t hangText="Author/Change controller:">
3895      IESG
3896    </t>
3897  </list>
3900<section title="Internet Media Type application/http" anchor="">
3901<iref item="Media Type" subitem="application/http" primary="true"/>
3902<iref item="application/http Media Type" primary="true"/>
3904   The application/http type can be used to enclose a pipeline of one or more
3905   HTTP request or response messages (not intermixed).
3908  <list style="hanging" x:indent="12em">
3909    <t hangText="Type name:">
3910      application
3911    </t>
3912    <t hangText="Subtype name:">
3913      http
3914    </t>
3915    <t hangText="Required parameters:">
3916      none
3917    </t>
3918    <t hangText="Optional parameters:">
3919      version, msgtype
3920      <list style="hanging">
3921        <t hangText="version:">
3922          The HTTP-Version number of the enclosed messages
3923          (e.g., "1.1"). If not present, the version can be
3924          determined from the first line of the body.
3925        </t>
3926        <t hangText="msgtype:">
3927          The message type -- "request" or "response". If not
3928          present, the type can be determined from the first
3929          line of the body.
3930        </t>
3931      </list>
3932    </t>
3933    <t hangText="Encoding considerations:">
3934      HTTP messages enclosed by this type
3935      are in "binary" format; use of an appropriate
3936      Content-Transfer-Encoding is required when
3937      transmitted via E-mail.
3938    </t>
3939    <t hangText="Security considerations:">
3940      none
3941    </t>
3942    <t hangText="Interoperability considerations:">
3943      none
3944    </t>
3945    <t hangText="Published specification:">
3946      This specification (see <xref target=""/>).
3947    </t>
3948    <t hangText="Applications that use this media type:">
3949    </t>
3950    <t hangText="Additional information:">
3951      <list style="hanging">
3952        <t hangText="Magic number(s):">none</t>
3953        <t hangText="File extension(s):">none</t>
3954        <t hangText="Macintosh file type code(s):">none</t>
3955      </list>
3956    </t>
3957    <t hangText="Person and email address to contact for further information:">
3958      See Authors Section.
3959    </t>
3960                <t hangText="Intended usage:">
3961                  COMMON
3962    </t>
3963                <t hangText="Restrictions on usage:">
3964                  none
3965    </t>
3966    <t hangText="Author/Change controller:">
3967      IESG
3968    </t>
3969  </list>
3974<section title="Tolerant Applications" anchor="tolerant.applications">
3976   Although this document specifies the requirements for the generation
3977   of HTTP/1.1 messages, not all applications will be correct in their
3978   implementation. We therefore recommend that operational applications
3979   be tolerant of deviations whenever those deviations can be
3980   interpreted unambiguously.
3983   Clients &SHOULD; be tolerant in parsing the Status-Line and servers
3984   tolerant when parsing the Request-Line. In particular, they &SHOULD;
3985   accept any amount of SP or HTAB characters between fields, even though
3986   only a single SP is required.
3989   The line terminator for message-header fields is the sequence CRLF.
3990   However, we recommend that applications, when parsing such headers,
3991   recognize a single LF as a line terminator and ignore the leading CR.
3994   The character set of an entity-body &SHOULD; be labeled as the lowest
3995   common denominator of the character codes used within that body, with
3996   the exception that not labeling the entity is preferred over labeling
3997   the entity with the labels US-ASCII or ISO-8859-1. See &payload;.
4000   Additional rules for requirements on parsing and encoding of dates
4001   and other potential problems with date encodings include:
4004  <list style="symbols">
4005     <t>HTTP/1.1 clients and caches &SHOULD; assume that an RFC-850 date
4006        which appears to be more than 50 years in the future is in fact
4007        in the past (this helps solve the "year 2000" problem).</t>
4009     <t>An HTTP/1.1 implementation &MAY; internally represent a parsed
4010        Expires date as earlier than the proper value, but &MUST-NOT;
4011        internally represent a parsed Expires date as later than the
4012        proper value.</t>
4014     <t>All expiration-related calculations &MUST; be done in GMT. The
4015        local time zone &MUST-NOT; influence the calculation or comparison
4016        of an age or expiration time.</t>
4018     <t>If an HTTP header incorrectly carries a date value with a time
4019        zone other than GMT, it &MUST; be converted into GMT using the
4020        most conservative possible conversion.</t>
4021  </list>
4025<section title="Conversion of Date Formats" anchor="">
4027   HTTP/1.1 uses a restricted set of date formats (<xref target=""/>) to
4028   simplify the process of date comparison. Proxies and gateways from
4029   other protocols &SHOULD; ensure that any Date header field present in a
4030   message conforms to one of the HTTP/1.1 formats and rewrite the date
4031   if necessary.
4035<section title="Compatibility with Previous Versions" anchor="compatibility">
4037   It is beyond the scope of a protocol specification to mandate
4038   compliance with previous versions. HTTP/1.1 was deliberately
4039   designed, however, to make supporting previous versions easy. It is
4040   worth noting that, at the time of composing this specification
4041   (1996), we would expect commercial HTTP/1.1 servers to:
4042  <list style="symbols">
4043     <t>recognize the format of the Request-Line for HTTP/0.9, 1.0, and
4044        1.1 requests;</t>
4046     <t>understand any valid request in the format of HTTP/0.9, 1.0, or
4047        1.1;</t>
4049     <t>respond appropriately with a message in the same major version
4050        used by the client.</t>
4051  </list>
4054   And we would expect HTTP/1.1 clients to:
4055  <list style="symbols">
4056     <t>recognize the format of the Status-Line for HTTP/1.0 and 1.1
4057        responses;</t>
4059     <t>understand any valid response in the format of HTTP/0.9, 1.0, or
4060        1.1.</t>
4061  </list>
4064   For most implementations of HTTP/1.0, each connection is established
4065   by the client prior to the request and closed by the server after
4066   sending the response. Some implementations implement the Keep-Alive
4067   version of persistent connections described in <xref x:sec="19.7.1" x:fmt="of" target="RFC2068"/>.
4070<section title="Changes from HTTP/1.0" anchor="changes.from.1.0">
4072   This section summarizes major differences between versions HTTP/1.0
4073   and HTTP/1.1.
4076<section title="Changes to Simplify Multi-homed Web Servers and Conserve IP Addresses" anchor="">
4078   The requirements that clients and servers support the Host request-header,
4079   report an error if the Host request-header (<xref target=""/>) is
4080   missing from an HTTP/1.1 request, and accept absolute URIs (<xref target="request-uri"/>)
4081   are among the most important changes defined by this
4082   specification.
4085   Older HTTP/1.0 clients assumed a one-to-one relationship of IP
4086   addresses and servers; there was no other established mechanism for
4087   distinguishing the intended server of a request than the IP address
4088   to which that request was directed. The changes outlined above will
4089   allow the Internet, once older HTTP clients are no longer common, to
4090   support multiple Web sites from a single IP address, greatly
4091   simplifying large operational Web servers, where allocation of many
4092   IP addresses to a single host has created serious problems. The
4093   Internet will also be able to recover the IP addresses that have been
4094   allocated for the sole purpose of allowing special-purpose domain
4095   names to be used in root-level HTTP URLs. Given the rate of growth of
4096   the Web, and the number of servers already deployed, it is extremely
4097   important that all implementations of HTTP (including updates to
4098   existing HTTP/1.0 applications) correctly implement these
4099   requirements:
4100  <list style="symbols">
4101     <t>Both clients and servers &MUST; support the Host request-header.</t>
4103     <t>A client that sends an HTTP/1.1 request &MUST; send a Host header.</t>
4105     <t>Servers &MUST; report a 400 (Bad Request) error if an HTTP/1.1
4106        request does not include a Host request-header.</t>
4108     <t>Servers &MUST; accept absolute URIs.</t>
4109  </list>
4114<section title="Compatibility with HTTP/1.0 Persistent Connections" anchor="compatibility.with.http.1.0.persistent.connections">
4116   Some clients and servers might wish to be compatible with some
4117   previous implementations of persistent connections in HTTP/1.0
4118   clients and servers. Persistent connections in HTTP/1.0 are
4119   explicitly negotiated as they are not the default behavior. HTTP/1.0
4120   experimental implementations of persistent connections are faulty,
4121   and the new facilities in HTTP/1.1 are designed to rectify these
4122   problems. The problem was that some existing 1.0 clients may be
4123   sending Keep-Alive to a proxy server that doesn't understand
4124   Connection, which would then erroneously forward it to the next
4125   inbound server, which would establish the Keep-Alive connection and
4126   result in a hung HTTP/1.0 proxy waiting for the close on the
4127   response. The result is that HTTP/1.0 clients must be prevented from
4128   using Keep-Alive when talking to proxies.
4131   However, talking to proxies is the most important use of persistent
4132   connections, so that prohibition is clearly unacceptable. Therefore,
4133   we need some other mechanism for indicating a persistent connection
4134   is desired, which is safe to use even when talking to an old proxy
4135   that ignores Connection. Persistent connections are the default for
4136   HTTP/1.1 messages; we introduce a new keyword (Connection: close) for
4137   declaring non-persistence. See <xref target="header.connection"/>.
4140   The original HTTP/1.0 form of persistent connections (the Connection:
4141   Keep-Alive and Keep-Alive header) is documented in <xref target="RFC2068"/>.
4145<section title="Changes from RFC 2068" anchor="changes.from.rfc.2068">
4147   This specification has been carefully audited to correct and
4148   disambiguate key word usage; RFC 2068 had many problems in respect to
4149   the conventions laid out in <xref target="RFC2119"/>.
4152   Transfer-coding and message lengths all interact in ways that
4153   required fixing exactly when chunked encoding is used (to allow for
4154   transfer encoding that may not be self delimiting); it was important
4155   to straighten out exactly how message lengths are computed. (Sections
4156   <xref target="transfer.codings" format="counter"/>, <xref target="message.length" format="counter"/>,
4157   <xref target="header.content-length" format="counter"/>,
4158   see also <xref target="Part3"/>, <xref target="Part5"/> and <xref target="Part6"/>)
4161   The use and interpretation of HTTP version numbers has been clarified
4162   by <xref target="RFC2145"/>. Require proxies to upgrade requests to highest protocol
4163   version they support to deal with problems discovered in HTTP/1.0
4164   implementations (<xref target="http.version"/>)
4167   Transfer-coding had significant problems, particularly with
4168   interactions with chunked encoding. The solution is that transfer-codings
4169   become as full fledged as content-codings. This involves
4170   adding an IANA registry for transfer-codings (separate from content
4171   codings), a new header field (TE) and enabling trailer headers in the
4172   future. Transfer encoding is a major performance benefit, so it was
4173   worth fixing <xref target="Nie1997"/>. TE also solves another, obscure, downward
4174   interoperability problem that could have occurred due to interactions
4175   between authentication trailers, chunked encoding and HTTP/1.0
4176   clients.(Section <xref target="transfer.codings" format="counter"/>, <xref target="chunked.transfer.encoding" format="counter"/>,
4177   and <xref target="header.te" format="counter"/>)
4181<section title="Changes from RFC 2616" anchor="changes.from.rfc.2616">
4183  The CHAR rule does not allow the NUL character anymore (this affects
4184  the comment and quoted-string rules).  Furthermore, the quoted-pair
4185  rule does not allow escaping NUL, CR or LF anymore.
4186  (<xref target="basic.rules"/>)
4189  Clarify that HTTP-Version is case sensitive.
4190  (<xref target="http.version"/>)
4193  Remove reference to non-existant identity transfer-coding value tokens.
4194  (Sections <xref format="counter" target="transfer.codings"/> and
4195  <xref format="counter" target="message.length"/>)
4198  Clarification that the chunk length does not include
4199  the count of the octets in the chunk header and trailer.
4200  (<xref target="chunked.transfer.encoding"/>)
4203  Fix BNF to add query, as the abs_path production in
4204  <xref x:sec="3" x:fmt="of" target="RFC2396"/> doesn't define it.
4205  (<xref target="request-uri"/>)
4208  Clarify exactly when close connection options must be sent.
4209  (<xref target="header.connection"/>)
4214<section title="Change Log (to be removed by RFC Editor before publication)" anchor="change.log">
4216<section title="Since RFC2616">
4218  Extracted relevant partitions from <xref target="RFC2616"/>.
4222<section title="Since draft-ietf-httpbis-p1-messaging-00">
4224  Closed issues:
4225  <list style="symbols">
4226    <t>
4227      <eref target=""/>:
4228      "HTTP Version should be case sensitive"
4229      (<eref target=""/>)
4230    </t>
4231    <t>
4232      <eref target=""/>:
4233      "'unsafe' characters"
4234      (<eref target=""/>)
4235    </t>
4236    <t>
4237      <eref target=""/>:
4238      "Chunk Size Definition"
4239      (<eref target=""/>)
4240    </t>
4241    <t>
4242      <eref target=""/>:
4243      "Message Length"
4244      (<eref target=""/>)
4245    </t>
4246    <t>
4247      <eref target=""/>:
4248      "Media Type Registrations"
4249      (<eref target=""/>)
4250    </t>
4251    <t>
4252      <eref target=""/>:
4253      "URI includes query"
4254      (<eref target=""/>)
4255    </t>
4256    <t>
4257      <eref target=""/>:
4258      "No close on 1xx responses"
4259      (<eref target=""/>)
4260    </t>
4261    <t>
4262      <eref target=""/>:
4263      "Remove 'identity' token references"
4264      (<eref target=""/>)
4265    </t>
4266    <t>
4267      <eref target=""/>:
4268      "Import query BNF"
4269    </t>
4270    <t>
4271      <eref target=""/>:
4272      "qdtext BNF"
4273    </t>
4274    <t>
4275      <eref target=""/>:
4276      "Normative and Informative references"
4277    </t>
4278    <t>
4279      <eref target=""/>:
4280      "RFC2606 Compliance"
4281    </t>
4282    <t>
4283      <eref target=""/>:
4284      "RFC977 reference"
4285    </t>
4286    <t>
4287      <eref target=""/>:
4288      "RFC1700 references"
4289    </t>
4290    <t>
4291      <eref target=""/>:
4292      "inconsistency in date format explanation"
4293    </t>
4294    <t>
4295      <eref target=""/>:
4296      "Date reference typo"
4297    </t>
4298    <t>
4299      <eref target=""/>:
4300      "Informative references"
4301    </t>
4302    <t>
4303      <eref target=""/>:
4304      "ISO-8859-1 Reference"
4305    </t>
4306    <t>
4307      <eref target=""/>:
4308      "Normative up-to-date references"
4309    </t>
4310  </list>
4313  Other changes:
4314  <list style="symbols">
4315    <t>
4316      Update media type registrations to use RFC4288 template.
4317    </t>
4318    <t>
4319      Use names of RFC4234 core rules DQUOTE and HTAB,
4320      fix broken ABNF for chunk-data
4321      (work in progress on <eref target=""/>)
4322    </t>
4323  </list>
4327<section title="Since draft-ietf-httpbis-p1-messaging-01">
4329  Closed issues:
4330  <list style="symbols">
4331    <t>
4332      <eref target=""/>:
4333      "Bodies on GET (and other) requests"
4334    </t>
4335    <t>
4336      <eref target=""/>:
4337      "Updating to RFC4288"
4338    </t>
4339    <t>
4340      <eref target=""/>:
4341      "Status Code and Reason Phrase"
4342    </t>
4343    <t>
4344      <eref target=""/>:
4345      "rel_path not used"
4346    </t>
4347  </list>
4350  Ongoing work on ABNF conversion (<eref target=""/>):
4351  <list style="symbols">
4352    <t>
4353      Get rid of duplicate BNF rule names ("host" -> "uri-host", "trailer" ->
4354      "trailer-part").
4355    </t>
4356    <t>
4357      Avoid underscore character in rule names ("http_URL" ->
4358      "http-URL", "abs_path" -> "path-absolute").
4359    </t>
4360    <t>
4361      Add rules for terms imported from URI spec ("absoluteURI", "authority",
4362      "path-absolute", "port", "query", "relativeURI", "host) -- these will
4363      have to be updated when switching over to RFC3986.
4364    </t>
4365    <t>
4366      Synchronize core rules with RFC5234 (this includes a change to CHAR
4367      which now excludes NUL).
4368    </t>
4369    <t>
4370      Get rid of prose rules that span multiple lines.
4371    </t>
4372    <t>
4373      Get rid of unused rules LOALPHA and UPALPHA.
4374    </t>
4375    <t>
4376      Move "Product Tokens" section (back) into Part 1, as "token" is used
4377      in the definition of the Upgrade header.
4378    </t>
4379    <t>
4380      Add explicit references to BNF syntax and rules imported from other parts of the specification.
4381    </t>
4382    <t>
4383      Rewrite prose rule "token" in terms of "tchar", rewrite prose rule "TEXT".
4384    </t>
4385  </list>
4389<section title="Since draft-ietf-httpbis-p1-messaging-02" anchor="changes.since.02">
4391  Closed issues:
4392  <list style="symbols">
4393    <t>
4394      <eref target=""/>:
4395      "HTTP-date vs. rfc1123-date"
4396    </t>
4397    <t>
4398      <eref target=""/>:
4399      "WS in quoted-pair"
4400    </t>
4401  </list>
Note: See TracBrowser for help on using the repository browser.