source: draft-ietf-httpbis/latest/p1-messaging.xml @ 35

Last change on this file since 35 was 35, checked in by fielding@…, 12 years ago

Use obsoletes instead of updates and seven-part instead of eight.

  • Property svn:eol-style set to native
File size: 176.0 KB
Line 
1<?xml version="1.0" encoding="utf-8"?>
2<!DOCTYPE rfc [
3  <!ENTITY MAY "<bcp14 xmlns='http://purl.org/net/xml2rfc/ext'>MAY</bcp14>">
4  <!ENTITY MUST "<bcp14 xmlns='http://purl.org/net/xml2rfc/ext'>MUST</bcp14>">
5  <!ENTITY MUST-NOT "<bcp14 xmlns='http://purl.org/net/xml2rfc/ext'>MUST NOT</bcp14>">
6  <!ENTITY OPTIONAL "<bcp14 xmlns='http://purl.org/net/xml2rfc/ext'>OPTIONAL</bcp14>">
7  <!ENTITY RECOMMENDED "<bcp14 xmlns='http://purl.org/net/xml2rfc/ext'>RECOMMENDED</bcp14>">
8  <!ENTITY REQUIRED "<bcp14 xmlns='http://purl.org/net/xml2rfc/ext'>REQUIRED</bcp14>">
9  <!ENTITY SHALL "<bcp14 xmlns='http://purl.org/net/xml2rfc/ext'>SHALL</bcp14>">
10  <!ENTITY SHALL-NOT "<bcp14 xmlns='http://purl.org/net/xml2rfc/ext'>SHALL NOT</bcp14>">
11  <!ENTITY SHOULD "<bcp14 xmlns='http://purl.org/net/xml2rfc/ext'>SHOULD</bcp14>">
12  <!ENTITY SHOULD-NOT "<bcp14 xmlns='http://purl.org/net/xml2rfc/ext'>SHOULD NOT</bcp14>">
13  <!ENTITY ID-VERSION "latest">
14  <!ENTITY ID-MONTH "December">
15  <!ENTITY ID-YEAR "2007">
16  <!ENTITY caching                "<xref target='Part6' x:rel='#caching' xmlns:x='http://purl.org/net/xml2rfc/ext'/>">
17  <!ENTITY payload                "<xref target='Part3' xmlns:x='http://purl.org/net/xml2rfc/ext'/>">
18  <!ENTITY CONNECT                "<xref target='Part2' x:rel='#CONNECT' xmlns:x='http://purl.org/net/xml2rfc/ext'/>">
19  <!ENTITY content.negotiation    "<xref target='Part3' x:rel='#content.negotiation' xmlns:x='http://purl.org/net/xml2rfc/ext'/>">
20  <!ENTITY diff2045entity         "<xref target='Part3' x:rel='#differences.between.http.entities.and.rfc.2045.entities' xmlns:x='http://purl.org/net/xml2rfc/ext'/>">
21  <!ENTITY entity                 "<xref target='Part3' x:rel='#entity' xmlns:x='http://purl.org/net/xml2rfc/ext'/>">
22  <!ENTITY entity-header-fields   "<xref target='Part3' x:rel='#entity.header.fields' xmlns:x='http://purl.org/net/xml2rfc/ext'/>">
23  <!ENTITY header-cache-control   "<xref target='Part6' x:rel='#header.cache-control' xmlns:x='http://purl.org/net/xml2rfc/ext'/>">
24  <!ENTITY header-expect          "<xref target='Part2' x:rel='#header.expect' xmlns:x='http://purl.org/net/xml2rfc/ext'/>">
25  <!ENTITY header-pragma          "<xref target='Part6' x:rel='#header.pragma' xmlns:x='http://purl.org/net/xml2rfc/ext'/>">
26  <!ENTITY header-warning         "<xref target='Part6' x:rel='#header.warning' xmlns:x='http://purl.org/net/xml2rfc/ext'/>">
27  <!ENTITY idempotent-methods     "<xref target='Part2' x:rel='#idempotent.methods' xmlns:x='http://purl.org/net/xml2rfc/ext'/>">
28  <!ENTITY qvalue                 "<xref target='Part3' x:rel='#quality.values' xmlns:x='http://purl.org/net/xml2rfc/ext'/>">
29  <!ENTITY request-header-fields  "<xref target='Part2' x:rel='#request.header.fields' xmlns:x='http://purl.org/net/xml2rfc/ext'/>">
30  <!ENTITY response-header-fields "<xref target='Part2' x:rel='#response.header.fields' xmlns:x='http://purl.org/net/xml2rfc/ext'/>">
31  <!ENTITY method                 "<xref target='Part2' x:rel='#method' xmlns:x='http://purl.org/net/xml2rfc/ext'/>">
32  <!ENTITY status-codes           "<xref target='Part2' x:rel='#status.codes' xmlns:x='http://purl.org/net/xml2rfc/ext'/>">
33  <!ENTITY status-100             "<xref target='Part2' x:rel='#status.100' xmlns:x='http://purl.org/net/xml2rfc/ext'/>">
34  <!ENTITY status-1xx             "<xref target='Part2' x:rel='#status.1xx' xmlns:x='http://purl.org/net/xml2rfc/ext'/>">
35  <!ENTITY status-414             "<xref target='Part2' x:rel='#status.414' xmlns:x='http://purl.org/net/xml2rfc/ext'/>">
36]>
37<?rfc toc="yes" ?>
38<?rfc symrefs="yes" ?>
39<?rfc sortrefs="yes" ?>
40<?rfc compact="yes"?>
41<?rfc subcompact="no" ?>
42<?rfc linkmailto="no" ?>
43<?rfc editing="no" ?>
44<?rfc-ext allow-markup-in-artwork="yes" ?>
45<?rfc-ext include-references-in-index="yes" ?>
46<rfc obsoletes="2068, 2616, 2617" category="std"
47     ipr="full3978" docName="draft-ietf-httpbis-p1-messaging-&ID-VERSION;"
48     xmlns:x='http://purl.org/net/xml2rfc/ext' xmlns:ed="http://greenbytes.de/2002/rfcedit">
49<front>
50
51  <title abbrev="HTTP/1.1">HTTP/1.1, part 1: URIs, Connections, and Message Parsing</title>
52
53  <author initials="R." surname="Fielding" fullname="Roy T. Fielding" role="editor">
54    <organization abbrev="Day Software">Day Software</organization>
55    <address>
56      <postal>
57        <street>23 Corporate Plaza DR, Suite 280</street>
58        <city>Newport Beach</city>
59        <region>CA</region>
60        <code>92660</code>
61        <country>USA</country>
62      </postal>
63      <phone>+1-949-706-5300</phone>
64      <facsimile>+1-949-706-5305</facsimile>
65      <email>fielding@gbiv.com</email>
66      <uri>http://roy.gbiv.com/</uri>
67    </address>
68  </author>
69
70  <author initials="J." surname="Gettys" fullname="Jim Gettys">
71    <organization>One Laptop per Child</organization>
72    <address>
73      <postal>
74        <street>21 Oak Knoll Road</street>
75        <city>Carlisle</city>
76        <region>MA</region>
77        <code>01741</code>
78        <country>USA</country>
79      </postal>
80      <email>jg@laptop.org</email>
81      <uri>http://www.laptop.org/</uri>
82    </address>
83  </author>
84 
85  <author initials="J." surname="Mogul" fullname="Jeffrey C. Mogul">
86    <organization abbrev="HP">Hewlett-Packard Company</organization>
87    <address>
88      <postal>
89        <street>HP Labs, Large Scale Systems Group</street>
90        <street>1501 Page Mill Road, MS 1177</street>
91        <city>Palo Alto</city>
92        <region>CA</region>
93        <code>94304</code>
94        <country>USA</country>
95      </postal>
96      <email>JeffMogul@acm.org</email>
97    </address>
98  </author>
99
100  <author initials="H." surname="Frystyk" fullname="Henrik Frystyk Nielsen">
101    <organization abbrev="Microsoft">Microsoft Corporation</organization>
102    <address>
103      <postal>
104        <street>1 Microsoft Way</street>
105        <city>Redmond</city>
106        <region>WA</region>
107        <code>98052</code>
108        <country>USA</country>
109      </postal>
110      <email>henrikn@microsoft.com</email>
111    </address>
112  </author>
113
114  <author initials="L." surname="Masinter" fullname="Larry Masinter">
115    <organization abbrev="Adobe Systems">Adobe Systems, Incorporated</organization>
116    <address>
117      <postal>
118        <street>345 Park Ave</street>
119        <city>San Jose</city>
120        <region>CA</region>
121        <code>95110</code>
122        <country>USA</country>
123      </postal>
124      <email>LMM@acm.org</email>
125      <uri>http://larry.masinter.net/</uri>
126    </address>
127  </author>
128 
129  <author initials="P." surname="Leach" fullname="Paul J. Leach">
130    <organization abbrev="Microsoft">Microsoft Corporation</organization>
131    <address>
132      <postal>
133        <street>1 Microsoft Way</street>
134        <city>Redmond</city>
135        <region>WA</region>
136        <code>98052</code>
137      </postal>
138      <email>paulle@microsoft.com</email>
139    </address>
140  </author>
141   
142  <author initials="T." surname="Berners-Lee" fullname="Tim Berners-Lee">
143    <organization abbrev="W3C/MIT">World Wide Web Consortium</organization>
144    <address>
145      <postal>
146        <street>MIT Computer Science and Artificial Intelligence Laboratory</street>
147        <street>The Stata Center, Building 32</street>
148        <street>32 Vassar Street</street>
149        <city>Cambridge</city>
150        <region>MA</region>
151        <code>02139</code>
152        <country>USA</country>
153      </postal>
154      <email>timbl@w3.org</email>
155      <uri>http://www.w3.org/People/Berners-Lee/</uri>
156    </address>
157  </author>
158
159  <date month="&ID-MONTH;" year="&ID-YEAR;"/>
160
161<abstract>
162<t>
163   The Hypertext Transfer Protocol (HTTP) is an application-level
164   protocol for distributed, collaborative, hypermedia information
165   systems. HTTP has been in use by the World Wide Web global information
166   initiative since 1990. This document is Part 1 of the seven-part specification
167   that defines the protocol referred to as "HTTP/1.1" and, taken together,
168   obsoletes RFC 2616 and RFC 2617.  Part 1 provides an overview of HTTP and
169   its associated terminology, defines the "http" and "https" Uniform
170   Resource Identifier (URI) schemes, defines the generic message syntax
171   and parsing requirements for HTTP message frames, and describes
172   general security concerns for implementations.
173</t>
174</abstract>
175</front>
176<middle>
177<section title="Introduction" anchor="introduction">
178<t>
179   This document will define aspects of HTTP related to overall network
180   operation, message framing, interaction with transport protocols, and
181   URI schemes. Right now it only includes the extracted relevant sections
182   of <xref target="RFC2616"/> and <xref target="RFC2617"/>.
183</t>
184<section title="Purpose" anchor="intro.purpose">
185<t>
186   The Hypertext Transfer Protocol (HTTP) is an application-level
187   protocol for distributed, collaborative, hypermedia information
188   systems. HTTP has been in use by the World-Wide Web global
189   information initiative since 1990. The first version of HTTP,
190   referred to as HTTP/0.9, was a simple protocol for raw data transfer
191   across the Internet. HTTP/1.0, as defined by RFC 1945 <xref target="RFC1945"/>, improved
192   the protocol by allowing messages to be in the format of MIME-like
193   messages, containing metainformation about the data transferred and
194   modifiers on the request/response semantics. However, HTTP/1.0 does
195   not sufficiently take into consideration the effects of hierarchical
196   proxies, caching, the need for persistent connections, or virtual
197   hosts. In addition, the proliferation of incompletely-implemented
198   applications calling themselves "HTTP/1.0" has necessitated a
199   protocol version change in order for two communicating applications
200   to determine each other's true capabilities.
201</t>
202<t>
203   This specification defines the protocol referred to as "HTTP/1.1".
204   This protocol includes more stringent requirements than HTTP/1.0 in
205   order to ensure reliable implementation of its features.
206</t>
207<t>
208   Practical information systems require more functionality than simple
209   retrieval, including search, front-end update, and annotation. HTTP
210   allows an open-ended set of methods and headers that indicate the
211   purpose of a request <xref target="RFC2324"/>. It builds on the discipline of reference
212   provided by the Uniform Resource Identifier (URI) <xref target="RFC1630"/>, as a location
213   (URL) <xref target="RFC1738"/> or name (URN) <xref target="RFC1737"/>, for indicating the resource to which a
214   method is to be applied. Messages are passed in a format similar to
215   that used by Internet mail <xref target="RFC822"/> as defined by the Multipurpose
216   Internet Mail Extensions (MIME) <xref target="RFC2045"/>.
217</t>
218<t>
219   HTTP is also used as a generic protocol for communication between
220   user agents and proxies/gateways to other Internet systems, including
221   those supported by the SMTP <xref target="RFC821"/>, NNTP <xref target="RFC977"/>, FTP <xref target="RFC959"/>, Gopher <xref target="RFC1436"/>,
222   and WAIS <xref target="WAIS"/> protocols. In this way, HTTP allows basic hypermedia
223   access to resources available from diverse applications.
224</t>
225</section>
226
227<section title="Requirements" anchor="intro.requirements">
228<t>
229   The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
230   "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this
231   document are to be interpreted as described in RFC 2119 <xref target="RFC2119"/>.
232</t>
233<t>
234   An implementation is not compliant if it fails to satisfy one or more
235   of the &MUST; or &REQUIRED; level requirements for the protocols it
236   implements. An implementation that satisfies all the &MUST; or &REQUIRED;
237   level and all the &SHOULD; level requirements for its protocols is said
238   to be "unconditionally compliant"; one that satisfies all the &MUST;
239   level requirements but not all the &SHOULD; level requirements for its
240   protocols is said to be "conditionally compliant."
241</t>
242</section>
243
244<section title="Terminology" anchor="intro.terminology">
245<t>
246   This specification uses a number of terms to refer to the roles
247   played by participants in, and objects of, the HTTP communication.
248</t>
249<t>
250  <iref item="connection"/>
251  <x:dfn>connection</x:dfn>
252  <list>
253    <t>
254      A transport layer virtual circuit established between two programs
255      for the purpose of communication.
256    </t>
257  </list>
258</t>
259<t>
260  <iref item="message"/>
261  <x:dfn>message</x:dfn>
262  <list>
263    <t>
264      The basic unit of HTTP communication, consisting of a structured
265      sequence of octets matching the syntax defined in <xref target="http.message"/> and
266      transmitted via the connection.
267    </t>
268  </list>
269</t>
270<t>
271  <iref item="request"/>
272  <x:dfn>request</x:dfn>
273  <list>
274    <t>
275      An HTTP request message, as defined in <xref target="request"/>.
276    </t>
277  </list>
278</t>
279<t>
280  <iref item="response"/>
281  <x:dfn>response</x:dfn>
282  <list>
283    <t>
284      An HTTP response message, as defined in <xref target="response"/>.
285    </t>
286  </list>
287</t>
288<t>
289  <iref item="resource"/>
290  <x:dfn>resource</x:dfn>
291  <list>
292    <t>
293      A network data object or service that can be identified by a URI,
294      as defined in <xref target="uri"/>. Resources may be available in multiple
295      representations (e.g. multiple languages, data formats, size, and
296      resolutions) or vary in other ways.
297    </t>
298  </list>
299</t>
300<t>
301  <iref item="entity"/>
302  <x:dfn>entity</x:dfn>
303  <list>
304    <t>
305      The information transferred as the payload of a request or
306      response. An entity consists of metainformation in the form of
307      entity-header fields and content in the form of an entity-body, as
308      described in &entity;.
309    </t>
310  </list>
311</t>
312<t>
313  <iref item="representation"/>
314  <x:dfn>representation</x:dfn>
315  <list>
316    <t>
317      An entity included with a response that is subject to content
318      negotiation, as described in &content.negotiation;. There may exist multiple
319      representations associated with a particular response status.
320    </t>
321  </list>
322</t>
323<t>
324  <iref item="content negotiation"/>
325  <x:dfn>content negotiation</x:dfn>
326  <list>
327    <t>
328      The mechanism for selecting the appropriate representation when
329      servicing a request, as described in &content.negotiation;. The
330      representation of entities in any response can be negotiated
331      (including error responses).
332    </t>
333  </list>
334</t>
335<t>
336  <iref item="variant"/>
337  <x:dfn>variant</x:dfn>
338  <list>
339    <t>
340      A resource may have one, or more than one, representation(s)
341      associated with it at any given instant. Each of these
342      representations is termed a `varriant'.  Use of the term `variant'
343      does not necessarily imply that the resource is subject to content
344      negotiation.
345    </t>
346  </list>
347</t>
348<t>
349  <iref item="client"/>
350  <x:dfn>client</x:dfn>
351  <list>
352    <t>
353      A program that establishes connections for the purpose of sending
354      requests.
355    </t>
356  </list>
357</t>
358<t>
359  <iref item="user agent"/>
360  <x:dfn>user agent</x:dfn>
361  <list>
362    <t>
363      The client which initiates a request. These are often browsers,
364      editors, spiders (web-traversing robots), or other end user tools.
365    </t>
366  </list>
367</t>
368<t>
369  <iref item="server"/>
370  <x:dfn>server</x:dfn>
371  <list>
372    <t>
373      An application program that accepts connections in order to
374      service requests by sending back responses. Any given program may
375      be capable of being both a client and a server; our use of these
376      terms refers only to the role being performed by the program for a
377      particular connection, rather than to the program's capabilities
378      in general. Likewise, any server may act as an origin server,
379      proxy, gateway, or tunnel, switching behavior based on the nature
380      of each request.
381    </t>
382  </list>
383</t>
384<t>
385  <iref item="origin server"/>
386  <x:dfn>origin server</x:dfn>
387  <list>
388    <t>
389      The server on which a given resource resides or is to be created.
390    </t>
391  </list>
392</t>
393<t>
394  <iref item="proxy"/>
395  <x:dfn>proxy</x:dfn>
396  <list>
397    <t>
398      An intermediary program which acts as both a server and a client
399      for the purpose of making requests on behalf of other clients.
400      Requests are serviced internally or by passing them on, with
401      possible translation, to other servers. A proxy &MUST; implement
402      both the client and server requirements of this specification. A
403      "transparent proxy" is a proxy that does not modify the request or
404      response beyond what is required for proxy authentication and
405      identification. A "non-transparent proxy" is a proxy that modifies
406      the request or response in order to provide some added service to
407      the user agent, such as group annotation services, media type
408      transformation, protocol reduction, or anonymity filtering. Except
409      where either transparent or non-transparent behavior is explicitly
410      stated, the HTTP proxy requirements apply to both types of
411      proxies.
412    </t>
413  </list>
414</t>
415<t>
416  <iref item="gateway"/>
417  <x:dfn>gateway</x:dfn>
418  <list>
419    <t>
420      A server which acts as an intermediary for some other server.
421      Unlike a proxy, a gateway receives requests as if it were the
422      origin server for the requested resource; the requesting client
423      may not be aware that it is communicating with a gateway.
424    </t>
425  </list>
426</t>
427<t>
428  <iref item="tunnel"/>
429  <x:dfn>tunnel</x:dfn>
430  <list>
431    <t>
432      An intermediary program which is acting as a blind relay between
433      two connections. Once active, a tunnel is not considered a party
434      to the HTTP communication, though the tunnel may have been
435      initiated by an HTTP request. The tunnel ceases to exist when both
436      ends of the relayed connections are closed.
437    </t>
438  </list>
439</t>
440<t>
441  <iref item="cache"/>
442  <x:dfn>cache</x:dfn>
443  <list>
444    <t>
445      A program's local store of response messages and the subsystem
446      that controls its message storage, retrieval, and deletion. A
447      cache stores cacheable responses in order to reduce the response
448      time and network bandwidth consumption on future, equivalent
449      requests. Any client or server may include a cache, though a cache
450      cannot be used by a server that is acting as a tunnel.
451    </t>
452  </list>
453</t>
454<t>
455  <iref item="cacheable"/>
456  <x:dfn>cacheable</x:dfn>
457  <list>
458    <t>
459      A response is cacheable if a cache is allowed to store a copy of
460      the response message for use in answering subsequent requests. The
461      rules for determining the cacheability of HTTP responses are
462      defined in &caching;. Even if a resource is cacheable, there may
463      be additional constraints on whether a cache can use the cached
464      copy for a particular request.
465    </t>
466  </list>
467</t>
468<t>
469  <iref item="upstream"/>
470  <iref item="downstream"/>
471  <x:dfn>upstream</x:dfn>/<x:dfn>downstream</x:dfn>
472  <list>
473    <t>
474      Upstream and downstream describe the flow of a message: all
475      messages flow from upstream to downstream.
476    </t>
477  </list>
478</t>
479<t>
480  <iref item="inbound"/>
481  <iref item="outbound"/>
482  <x:dfn>inbound</x:dfn>/<x:dfn>outbound</x:dfn>
483  <list>
484    <t>
485      Inbound and outbound refer to the request and response paths for
486      messages: "inbound" means "traveling toward the origin server",
487      and "outbound" means "traveling toward the user agent"
488    </t>
489  </list>
490</t>
491</section>
492
493<section title="Overall Operation" anchor="intro.overall.operation">
494<t>
495   The HTTP protocol is a request/response protocol. A client sends a
496   request to the server in the form of a request method, URI, and
497   protocol version, followed by a MIME-like message containing request
498   modifiers, client information, and possible body content over a
499   connection with a server. The server responds with a status line,
500   including the message's protocol version and a success or error code,
501   followed by a MIME-like message containing server information, entity
502   metainformation, and possible entity-body content. The relationship
503   between HTTP and MIME is described in &diff2045entity;.
504</t>
505<t>
506   Most HTTP communication is initiated by a user agent and consists of
507   a request to be applied to a resource on some origin server. In the
508   simplest case, this may be accomplished via a single connection (v)
509   between the user agent (UA) and the origin server (O).
510</t>
511<figure><artwork type="drawing">
512       request chain ------------------------&gt;
513    UA -------------------v------------------- O
514       &lt;----------------------- response chain
515</artwork></figure>
516<t>
517   A more complicated situation occurs when one or more intermediaries
518   are present in the request/response chain. There are three common
519   forms of intermediary: proxy, gateway, and tunnel. A proxy is a
520   forwarding agent, receiving requests for a URI in its absolute form,
521   rewriting all or part of the message, and forwarding the reformatted
522   request toward the server identified by the URI. A gateway is a
523   receiving agent, acting as a layer above some other server(s) and, if
524   necessary, translating the requests to the underlying server's
525   protocol. A tunnel acts as a relay point between two connections
526   without changing the messages; tunnels are used when the
527   communication needs to pass through an intermediary (such as a
528   firewall) even when the intermediary cannot understand the contents
529   of the messages.
530</t>
531<figure><artwork type="drawing">
532       request chain --------------------------------------&gt;
533    UA -----v----- A -----v----- B -----v----- C -----v----- O
534       &lt;------------------------------------- response chain
535</artwork></figure>
536<t>
537   The figure above shows three intermediaries (A, B, and C) between the
538   user agent and origin server. A request or response message that
539   travels the whole chain will pass through four separate connections.
540   This distinction is important because some HTTP communication options
541   may apply only to the connection with the nearest, non-tunnel
542   neighbor, only to the end-points of the chain, or to all connections
543   along the chain. Although the diagram is linear, each participant may
544   be engaged in multiple, simultaneous communications. For example, B
545   may be receiving requests from many clients other than A, and/or
546   forwarding requests to servers other than C, at the same time that it
547   is handling A's request.
548</t>
549<t>
550   Any party to the communication which is not acting as a tunnel may
551   employ an internal cache for handling requests. The effect of a cache
552   is that the request/response chain is shortened if one of the
553   participants along the chain has a cached response applicable to that
554   request. The following illustrates the resulting chain if B has a
555   cached copy of an earlier response from O (via C) for a request which
556   has not been cached by UA or A.
557</t>
558<figure><artwork type="drawing">
559          request chain ----------&gt;
560       UA -----v----- A -----v----- B - - - - - - C - - - - - - O
561          &lt;--------- response chain
562</artwork></figure>
563<t>
564   Not all responses are usefully cacheable, and some requests may
565   contain modifiers which place special requirements on cache behavior.
566   HTTP requirements for cache behavior and cacheable responses are
567   defined in &caching;.
568</t>
569<t>
570   In fact, there are a wide variety of architectures and configurations
571   of caches and proxies currently being experimented with or deployed
572   across the World Wide Web. These systems include national hierarchies
573   of proxy caches to save transoceanic bandwidth, systems that
574   broadcast or multicast cache entries, organizations that distribute
575   subsets of cached data via CD-ROM, and so on. HTTP systems are used
576   in corporate intranets over high-bandwidth links, and for access via
577   PDAs with low-power radio links and intermittent connectivity. The
578   goal of HTTP/1.1 is to support the wide diversity of configurations
579   already deployed while introducing protocol constructs that meet the
580   needs of those who build web applications that require high
581   reliability and, failing that, at least reliable indications of
582   failure.
583</t>
584<t>
585   HTTP communication usually takes place over TCP/IP connections. The
586   default port is TCP 80 <xref target="RFC1700"/>, but other ports can be used. This does
587   not preclude HTTP from being implemented on top of any other protocol
588   on the Internet, or on other networks. HTTP only presumes a reliable
589   transport; any protocol that provides such guarantees can be used;
590   the mapping of the HTTP/1.1 request and response structures onto the
591   transport data units of the protocol in question is outside the scope
592   of this specification.
593</t>
594<t>
595   In HTTP/1.0, most implementations used a new connection for each
596   request/response exchange. In HTTP/1.1, a connection may be used for
597   one or more request/response exchanges, although connections may be
598   closed for a variety of reasons (see <xref target="persistent.connections"/>).
599</t>
600</section>
601</section>
602
603<section title="Notational Conventions and Generic Grammar" anchor="notation">
604
605<section title="Augmented BNF" anchor="notation.abnf">
606<t>
607   All of the mechanisms specified in this document are described in
608   both prose and an augmented Backus-Naur Form (BNF) similar to that
609   used by RFC 822 <xref target="RFC822"/>. Implementors will need to be familiar with the
610   notation in order to understand this specification. The augmented BNF
611   includes the following constructs:
612</t>
613<t>
614   name = definition
615  <list>
616    <t>
617      The name of a rule is simply the name itself (without any
618      enclosing "&lt;" and "&gt;") and is separated from its definition by the
619      equal "=" character. White space is only significant in that
620      indentation of continuation lines is used to indicate a rule
621      definition that spans more than one line. Certain basic rules are
622      in uppercase, such as SP, LWS, HT, CRLF, DIGIT, ALPHA, etc. Angle
623      brackets are used within definitions whenever their presence will
624      facilitate discerning the use of rule names.
625    </t>
626  </list>
627</t>
628<t>
629   "literal"
630  <list>
631    <t>
632      Quotation marks surround literal text. Unless stated otherwise,
633      the text is case-insensitive.
634    </t>
635  </list>
636</t>
637<t>
638   rule1 | rule2
639  <list>
640    <t>
641      Elements separated by a bar ("|") are alternatives, e.g., "yes |
642      no" will accept yes or no.
643    </t>
644  </list>
645</t>
646<t>
647   (rule1 rule2)
648  <list>
649    <t>
650      Elements enclosed in parentheses are treated as a single element.
651      Thus, "(elem (foo | bar) elem)" allows the token sequences "elem
652      foo elem" and "elem bar elem".
653    </t>
654  </list>
655</t>
656<t>
657   *rule
658  <list>
659    <t>
660      The character "*" preceding an element indicates repetition. The
661      full form is "&lt;n&gt;*&lt;m&gt;element" indicating at least &lt;n&gt; and at most
662      &lt;m&gt; occurrences of element. Default values are 0 and infinity so
663      that "*(element)" allows any number, including zero; "1*element"
664      requires at least one; and "1*2element" allows one or two.
665    </t>
666  </list>
667</t>
668<t>
669   [rule]
670  <list>
671    <t>
672      Square brackets enclose optional elements; "[foo bar]" is
673      equivalent to "*1(foo bar)".
674    </t>
675  </list>
676</t>
677<t>
678   N rule
679  <list>
680    <t>
681      Specific repetition: "&lt;n&gt;(element)" is equivalent to
682      "&lt;n&gt;*&lt;n&gt;(element)"; that is, exactly &lt;n&gt; occurrences of (element).
683      Thus 2DIGIT is a 2-digit number, and 3ALPHA is a string of three
684      alphabetic characters.
685    </t>
686  </list>
687</t>
688<t>
689   #rule
690  <list>
691    <t>
692      A construct "#" is defined, similar to "*", for defining lists of
693      elements. The full form is "&lt;n&gt;#&lt;m&gt;element" indicating at least
694      &lt;n&gt; and at most &lt;m&gt; elements, each separated by one or more commas
695      (",") and &OPTIONAL; linear white space (LWS). This makes the usual
696      form of lists very easy; a rule such as
697    </t>
698    <t>
699         ( *LWS element *( *LWS "," *LWS element ))
700    </t>
701    <t>
702      can be shown as
703    </t>
704    <t>
705         1#element
706    </t>
707    <t>
708      Wherever this construct is used, null elements are allowed, but do
709      not contribute to the count of elements present. That is,
710      "(element), , (element) " is permitted, but counts as only two
711      elements. Therefore, where at least one element is required, at
712      least one non-null element &MUST; be present. Default values are 0
713      and infinity so that "#element" allows any number, including zero;
714      "1#element" requires at least one; and "1#2element" allows one or
715      two.
716    </t>
717  </list>
718</t>
719<t>
720   ; comment
721  <list>
722    <t>
723      A semi-colon, set off some distance to the right of rule text,
724      starts a comment that continues to the end of line. This is a
725      simple way of including useful notes in parallel with the
726      specifications.
727    </t>
728  </list>
729</t>
730<t>
731   implied *LWS
732  <list>
733    <t>
734      The grammar described by this specification is word-based. Except
735      where noted otherwise, linear white space (LWS) can be included
736      between any two adjacent words (token or quoted-string), and
737      between adjacent words and separators, without changing the
738      interpretation of a field. At least one delimiter (LWS and/or
739      separators) &MUST; exist between any two tokens (for the definition
740      of "token" below), since they would otherwise be interpreted as a
741      single token.
742    </t>
743  </list>
744</t>
745</section>
746
747<section title="Basic Rules" anchor="basic.rules">
748<x:anchor-alias value="OCTET"/>
749<x:anchor-alias value="CHAR"/>
750<x:anchor-alias value="UPALPHA"/>
751<x:anchor-alias value="LOALPHA"/>
752<x:anchor-alias value="ALPHA"/>
753<x:anchor-alias value="DIGIT"/>
754<x:anchor-alias value="CTL"/>
755<x:anchor-alias value="CR"/>
756<x:anchor-alias value="LF"/>
757<x:anchor-alias value="SP"/>
758<x:anchor-alias value="HT"/>
759<x:anchor-alias value="CRLF"/>
760<x:anchor-alias value="LWS"/>
761<x:anchor-alias value="TEXT"/>
762<x:anchor-alias value="HEX"/>
763<x:anchor-alias value="token"/>
764<x:anchor-alias value="separators"/>
765<x:anchor-alias value="comment"/>
766<x:anchor-alias value="ctext"/>
767<x:anchor-alias value="quoted-string"/>
768<x:anchor-alias value="qdtext"/>
769<x:anchor-alias value="quoted-pair"/>
770<t>
771   The following rules are used throughout this specification to
772   describe basic parsing constructs. The US-ASCII coded character set
773   is defined by ANSI X3.4-1986 <xref target="USASCII"/>.
774</t>
775<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="OCTET"/><iref primary="true" item="Grammar" subitem="CHAR"/><iref primary="true" item="Grammar" subitem="UPALPHA"/><iref primary="true" item="Grammar" subitem="LOALPHA"/><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="HT"/>
776    OCTET          = &lt;any 8-bit sequence of data&gt;
777    CHAR           = &lt;any US-ASCII character (octets 0 - 127)&gt;
778    UPALPHA        = &lt;any US-ASCII uppercase letter "A".."Z"&gt;
779    LOALPHA        = &lt;any US-ASCII lowercase letter "a".."z"&gt;
780    ALPHA          = UPALPHA | LOALPHA
781    DIGIT          = &lt;any US-ASCII digit "0".."9"&gt;
782    CTL            = &lt;any US-ASCII control character
783                     (octets 0 - 31) and DEL (127)&gt;
784    CR             = &lt;US-ASCII CR, carriage return (13)&gt;
785    LF             = &lt;US-ASCII LF, linefeed (10)&gt;
786    SP             = &lt;US-ASCII SP, space (32)&gt;
787    HT             = &lt;US-ASCII HT, horizontal-tab (9)&gt;
788    &lt;"&gt;            = &lt;US-ASCII double-quote mark (34)&gt;
789</artwork></figure>
790<t>
791   HTTP/1.1 defines the sequence CR LF as the end-of-line marker for all
792   protocol elements except the entity-body (see <xref target="tolerant.applications"/> for
793   tolerant applications). The end-of-line marker within an entity-body
794   is defined by its associated media type, as described in &payload;.
795</t>
796<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="CRLF"/>
797    CRLF           = CR LF
798</artwork></figure>
799<t>
800   HTTP/1.1 header field values can be folded onto multiple lines if the
801   continuation line begins with a space or horizontal tab. All linear
802   white space, including folding, has the same semantics as SP. A
803   recipient &MAY; replace any linear white space with a single SP before
804   interpreting the field value or forwarding the message downstream.
805</t>
806<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="LWS"/>
807    LWS            = [CRLF] 1*( SP | HT )
808</artwork></figure>
809<t>
810   The TEXT rule is only used for descriptive field contents and values
811   that are not intended to be interpreted by the message parser. Words
812   of *TEXT &MAY; contain characters from character sets other than ISO-8859-1
813   <xref target="ISO-8859"/> only when encoded according to the rules of RFC 2047
814   <xref target="RFC2047"/>.
815</t>
816<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="TEXT"/>
817    TEXT           = &lt;any OCTET except CTLs,
818                     but including LWS&gt;
819</artwork></figure>
820<t>
821   A CRLF is allowed in the definition of TEXT only as part of a header
822   field continuation. It is expected that the folding LWS will be
823   replaced with a single SP before interpretation of the TEXT value.
824</t>
825<t>
826   Hexadecimal numeric characters are used in several protocol elements.
827</t>
828<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="HEX"/>
829    HEX            = "A" | "B" | "C" | "D" | "E" | "F"
830                   | "a" | "b" | "c" | "d" | "e" | "f" | DIGIT
831</artwork></figure>
832<t>
833   Many HTTP/1.1 header field values consist of words separated by LWS
834   or special characters. These special characters &MUST; be in a quoted
835   string to be used within a parameter value (as defined in
836   <xref target="transfer.codings"/>).
837</t>
838<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="token"/><iref primary="true" item="Grammar" subitem="separators"/>
839    token          = 1*&lt;any CHAR except CTLs or separators&gt;
840    separators     = "(" | ")" | "&lt;" | "&gt;" | "@"
841                   | "," | ";" | ":" | "\" | &lt;"&gt;
842                   | "/" | "[" | "]" | "?" | "="
843                   | "{" | "}" | SP | HT
844</artwork></figure>
845<t>
846   Comments can be included in some HTTP header fields by surrounding
847   the comment text with parentheses. Comments are only allowed in
848   fields containing "comment" as part of their field value definition.
849   In all other fields, parentheses are considered part of the field
850   value.
851</t>
852<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="comment"/><iref primary="true" item="Grammar" subitem="ctext"/>
853    comment        = "(" *( ctext | quoted-pair | comment ) ")"
854    ctext          = &lt;any TEXT excluding "(" and ")"&gt;
855</artwork></figure>
856<t>
857   A string of text is parsed as a single word if it is quoted using
858   double-quote marks.
859</t>
860<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="quoted-string"/><iref primary="true" item="Grammar" subitem="qdtext"/>
861    quoted-string  = ( &lt;"&gt; *(qdtext | quoted-pair ) &lt;"&gt; )
862    qdtext         = &lt;any TEXT except &lt;"&gt;&gt;
863</artwork></figure>
864<t>
865   The backslash character ("\") &MAY; be used as a single-character
866   quoting mechanism only within quoted-string and comment constructs.
867</t>
868<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="quoted-pair"/>
869    quoted-pair    = "\" CHAR
870</artwork></figure>
871</section>
872</section>
873
874<section title="Protocol Parameters" anchor="protocol.parameters">
875
876<section title="HTTP Version" anchor="http.version">
877<t>
878   HTTP uses a "&lt;major&gt;.&lt;minor&gt;" numbering scheme to indicate versions
879   of the protocol. The protocol versioning policy is intended to allow
880   the sender to indicate the format of a message and its capacity for
881   understanding further HTTP communication, rather than the features
882   obtained via that communication. No change is made to the version
883   number for the addition of message components which do not affect
884   communication behavior or which only add to extensible field values.
885   The &lt;minor&gt; number is incremented when the changes made to the
886   protocol add features which do not change the general message parsing
887   algorithm, but which may add to the message semantics and imply
888   additional capabilities of the sender. The &lt;major&gt; number is
889   incremented when the format of a message within the protocol is
890   changed. See RFC 2145 <xref target="RFC2145"/> for a fuller explanation.
891</t>
892<t>
893   The version of an HTTP message is indicated by an HTTP-Version field
894   in the first line of the message.
895</t>
896<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="HTTP-Version"/>
897       HTTP-Version   = "HTTP" "/" 1*DIGIT "." 1*DIGIT
898</artwork></figure>
899<t>
900   Note that the major and minor numbers &MUST; be treated as separate
901   integers and that each &MAY; be incremented higher than a single digit.
902   Thus, HTTP/2.4 is a lower version than HTTP/2.13, which in turn is
903   lower than HTTP/12.3. Leading zeros &MUST; be ignored by recipients and
904   &MUST-NOT; be sent.
905</t>
906<t>
907   An application that sends a request or response message that includes
908   HTTP-Version of "HTTP/1.1" &MUST; be at least conditionally compliant
909   with this specification. Applications that are at least conditionally
910   compliant with this specification &SHOULD; use an HTTP-Version of
911   "HTTP/1.1" in their messages, and &MUST; do so for any message that is
912   not compatible with HTTP/1.0. For more details on when to send
913   specific HTTP-Version values, see RFC 2145 <xref target="RFC2145"/>.
914</t>
915<t>
916   The HTTP version of an application is the highest HTTP version for
917   which the application is at least conditionally compliant.
918</t>
919<t>
920   Proxy and gateway applications need to be careful when forwarding
921   messages in protocol versions different from that of the application.
922   Since the protocol version indicates the protocol capability of the
923   sender, a proxy/gateway &MUST-NOT; send a message with a version
924   indicator which is greater than its actual version. If a higher
925   version request is received, the proxy/gateway &MUST; either downgrade
926   the request version, or respond with an error, or switch to tunnel
927   behavior.
928</t>
929<t>
930   Due to interoperability problems with HTTP/1.0 proxies discovered
931   since the publication of RFC 2068 <xref target="RFC2068"/>, caching proxies MUST, gateways
932   &MAY;, and tunnels &MUST-NOT; upgrade the request to the highest version
933   they support. The proxy/gateway's response to that request &MUST; be in
934   the same major version as the request.
935</t>
936<t>
937  <list>
938    <t>
939      <x:h>Note:</x:h> Converting between versions of HTTP may involve modification
940      of header fields required or forbidden by the versions involved.
941    </t>
942  </list>
943</t>
944</section>
945
946<section title="Uniform Resource Identifiers" anchor="uri">
947<t>
948   URIs have been known by many names: WWW addresses, Universal Document
949   Identifiers, Universal Resource Identifiers <xref target="RFC1630"/>, and finally the
950   combination of Uniform Resource Locators (URL) <xref target="RFC1738"/> and Names (URN)
951   <xref target="RFC1737"/>. As far as HTTP is concerned, Uniform Resource Identifiers are
952   simply formatted strings which identify--via name, location, or any
953   other characteristic--a resource.
954</t>
955
956<section title="General Syntax" anchor="general.syntax">
957<t>
958   URIs in HTTP can be represented in absolute form or relative to some
959   known base URI <xref target="RFC1808"/>, depending upon the context of their use. The two
960   forms are differentiated by the fact that absolute URIs always begin
961   with a scheme name followed by a colon. For definitive information on
962   URL syntax and semantics, see "Uniform Resource Identifiers (URI):
963   Generic Syntax and Semantics," RFC 2396 <xref target="RFC2396"/> (which replaces RFCs
964   1738 <xref target="RFC1738"/> and RFC 1808 <xref target="RFC1808"/>). This specification adopts the
965   definitions of "URI-reference", "absoluteURI", "relativeURI", "port",
966   "host","abs_path", "rel_path", and "authority" from that
967   specification.
968</t>
969<t>
970   The HTTP protocol does not place any a priori limit on the length of
971   a URI. Servers &MUST; be able to handle the URI of any resource they
972   serve, and &SHOULD; be able to handle URIs of unbounded length if they
973   provide GET-based forms that could generate such URIs. A server
974   &SHOULD; return 414 (Request-URI Too Long) status if a URI is longer
975   than the server can handle (see &status-414;).
976</t>
977<t>
978  <list>
979    <t>
980      <x:h>Note:</x:h> Servers ought to be cautious about depending on URI lengths
981      above 255 bytes, because some older client or proxy
982      implementations might not properly support these lengths.
983    </t>
984  </list>
985</t>
986</section>
987
988<section title="http URL" anchor="http.url">
989<t>
990   The "http" scheme is used to locate network resources via the HTTP
991   protocol. This section defines the scheme-specific syntax and
992   semantics for http URLs.
993</t>
994<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="http_URL"/>
995http_URL = "http:" "//" host [ ":" port ] [ abs_path [ "?" query ]]
996</artwork></figure>
997<t>
998   If the port is empty or not given, port 80 is assumed. The semantics
999   are that the identified resource is located at the server listening
1000   for TCP connections on that port of that host, and the Request-URI
1001   for the resource is abs_path (<xref target="request-uri"/>). The use of IP addresses
1002   in URLs &SHOULD; be avoided whenever possible (see RFC 1900 <xref target="RFC1900"/>). If
1003   the abs_path is not present in the URL, it &MUST; be given as "/" when
1004   used as a Request-URI for a resource (<xref target="request-uri"/>). If a proxy
1005
1006   receives a host name which is not a fully qualified domain name, it
1007   &MAY; add its domain to the host name it received. If a proxy receives
1008   a fully qualified domain name, the proxy &MUST-NOT; change the host
1009   name.
1010</t>
1011</section>
1012
1013<section title="URI Comparison" anchor="uri.comparison">
1014<t>
1015   When comparing two URIs to decide if they match or not, a client
1016   &SHOULD; use a case-sensitive octet-by-octet comparison of the entire
1017   URIs, with these exceptions:
1018  <list style="symbols">
1019    <t>A port that is empty or not given is equivalent to the default
1020        port for that URI-reference;</t>
1021    <t>Comparisons of host names &MUST; be case-insensitive;</t>
1022    <t>Comparisons of scheme names &MUST; be case-insensitive;</t>
1023    <t>An empty abs_path is equivalent to an abs_path of "/".</t>
1024  </list>
1025</t>
1026<t>
1027   Characters other than those in the "reserved" and "unsafe" sets (see
1028   RFC 2396 <xref target="RFC2396"/>) are equivalent to their ""%" HEX HEX" encoding.
1029</t>
1030<t>
1031   For example, the following three URIs are equivalent:
1032</t>
1033<figure><artwork type="example">
1034   http://abc.com:80/~smith/home.html
1035   http://ABC.com/%7Esmith/home.html
1036   http://ABC.com:/%7esmith/home.html
1037</artwork></figure>
1038</section>
1039</section>
1040
1041<section title="Date/Time Formats" anchor="date.time.formats">
1042<section title="Full Date" anchor="full.date">
1043<t>
1044   HTTP applications have historically allowed three different formats
1045   for the representation of date/time stamps:
1046</t>
1047<figure><artwork type="example">
1048   Sun, 06 Nov 1994 08:49:37 GMT  ; RFC 822, updated by RFC 1123
1049   Sunday, 06-Nov-94 08:49:37 GMT ; RFC 850, obsoleted by RFC 1036
1050   Sun Nov  6 08:49:37 1994       ; ANSI C's asctime() format
1051</artwork></figure>
1052<t>
1053   The first format is preferred as an Internet standard and represents
1054   a fixed-length subset of that defined by RFC 1123 <xref target="RFC1123"/> (an update to
1055   RFC 822 <xref target="RFC822"/>). The second format is in common use, but is based on the
1056   obsolete RFC 850 <xref target="RFC1036"/> date format and lacks a four-digit year.
1057   HTTP/1.1 clients and servers that parse the date value &MUST; accept
1058   all three formats (for compatibility with HTTP/1.0), though they &MUST;
1059   only generate the RFC 1123 format for representing HTTP-date values
1060   in header fields. See <xref target="tolerant.applications"/> for further information.
1061</t>
1062<t><list><t>
1063      <x:h>Note:</x:h> Recipients of date values are encouraged to be robust in
1064      accepting date values that may have been sent by non-HTTP
1065      applications, as is sometimes the case when retrieving or posting
1066      messages via proxies/gateways to SMTP or NNTP.
1067</t></list></t>
1068<t>
1069   All HTTP date/time stamps &MUST; be represented in Greenwich Mean Time
1070   (GMT), without exception. For the purposes of HTTP, GMT is exactly
1071   equal to UTC (Coordinated Universal Time). This is indicated in the
1072   first two formats by the inclusion of "GMT" as the three-letter
1073   abbreviation for time zone, and &MUST; be assumed when reading the
1074   asctime format. HTTP-date is case sensitive and &MUST-NOT; include
1075   additional LWS beyond that specifically included as SP in the
1076   grammar.
1077</t>
1078<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="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"/>
1079    HTTP-date    = rfc1123-date | rfc850-date | asctime-date
1080    rfc1123-date = wkday "," SP date1 SP time SP "GMT"
1081    rfc850-date  = weekday "," SP date2 SP time SP "GMT"
1082    asctime-date = wkday SP date3 SP time SP 4DIGIT
1083    date1        = 2DIGIT SP month SP 4DIGIT
1084                   ; day month year (e.g., 02 Jun 1982)
1085    date2        = 2DIGIT "-" month "-" 2DIGIT
1086                   ; day-month-year (e.g., 02-Jun-82)
1087    date3        = month SP ( 2DIGIT | ( SP 1DIGIT ))
1088                   ; month day (e.g., Jun  2)
1089    time         = 2DIGIT ":" 2DIGIT ":" 2DIGIT
1090                   ; 00:00:00 - 23:59:59
1091    wkday        = "Mon" | "Tue" | "Wed"
1092                 | "Thu" | "Fri" | "Sat" | "Sun"
1093    weekday      = "Monday" | "Tuesday" | "Wednesday"
1094                 | "Thursday" | "Friday" | "Saturday" | "Sunday"
1095    month        = "Jan" | "Feb" | "Mar" | "Apr"
1096                 | "May" | "Jun" | "Jul" | "Aug"
1097                 | "Sep" | "Oct" | "Nov" | "Dec"
1098</artwork></figure>
1099<t>
1100      <x:h>Note:</x:h> HTTP requirements for the date/time stamp format apply only
1101      to their usage within the protocol stream. Clients and servers are
1102      not required to use these formats for user presentation, request
1103      logging, etc.
1104</t>
1105</section>
1106</section>
1107
1108<section title="Transfer Codings" anchor="transfer.codings">
1109<t>
1110   Transfer-coding values are used to indicate an encoding
1111   transformation that has been, can be, or may need to be applied to an
1112   entity-body in order to ensure "safe transport" through the network.
1113   This differs from a content coding in that the transfer-coding is a
1114   property of the message, not of the original entity.
1115</t>
1116<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="transfer-coding"/><iref primary="true" item="Grammar" subitem="transfer-extension"/>
1117    transfer-coding         = "chunked" | transfer-extension
1118    transfer-extension      = token *( ";" parameter )
1119</artwork></figure>
1120<t>
1121   Parameters are in  the form of attribute/value pairs.
1122</t>
1123<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"/>
1124    parameter               = attribute "=" value
1125    attribute               = token
1126    value                   = token | quoted-string
1127</artwork></figure>
1128<t>
1129   All transfer-coding values are case-insensitive. HTTP/1.1 uses
1130   transfer-coding values in the TE header field (<xref target="header.te"/>) and in
1131   the Transfer-Encoding header field (<xref target="header.transfer-encoding"/>).
1132</t>
1133<t>
1134   Whenever a transfer-coding is applied to a message-body, the set of
1135   transfer-codings &MUST; include "chunked", unless the message is
1136   terminated by closing the connection. When the "chunked" transfer-coding
1137   is used, it &MUST; be the last transfer-coding applied to the
1138   message-body. The "chunked" transfer-coding &MUST-NOT; be applied more
1139   than once to a message-body. These rules allow the recipient to
1140   determine the transfer-length of the message (<xref target="message.length"/>).
1141</t>
1142<t>
1143   Transfer-codings are analogous to the Content-Transfer-Encoding
1144   values of MIME <xref target="RFC2045"/>, which were designed to enable safe transport of
1145   binary data over a 7-bit transport service. However, safe transport
1146   has a different focus for an 8bit-clean transfer protocol. In HTTP,
1147   the only unsafe characteristic of message-bodies is the difficulty in
1148   determining the exact body length (<xref target="message.length"/>), or the desire to
1149   encrypt data over a shared transport.
1150</t>
1151<t>
1152   The Internet Assigned Numbers Authority (IANA) acts as a registry for
1153   transfer-coding value tokens. Initially, the registry contains the
1154   following tokens: "chunked" (<xref target="chunked.transfer.encoding"/>), "identity" (section
1155   3.6.2), "gzip" (&payload;), "compress" (&payload;), and "deflate"
1156   (&payload;).
1157</t>
1158<t>
1159   New transfer-coding value tokens &SHOULD; be registered in the same way
1160   as new content-coding value tokens (&payload;).
1161</t>
1162<t>
1163   A server which receives an entity-body with a transfer-coding it does
1164   not understand &SHOULD; return 501 (Unimplemented), and close the
1165   connection. A server &MUST-NOT; send transfer-codings to an HTTP/1.0
1166   client.
1167</t>
1168
1169<section title="Chunked Transfer Coding" anchor="chunked.transfer.encoding">
1170<t>
1171   The chunked encoding modifies the body of a message in order to
1172   transfer it as a series of chunks, each with its own size indicator,
1173   followed by an &OPTIONAL; trailer containing entity-header fields. This
1174   allows dynamically produced content to be transferred along with the
1175   information necessary for the recipient to verify that it has
1176   received the full message.
1177</t>
1178<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"/>
1179    Chunked-Body   = *chunk
1180                     last-chunk
1181                     trailer
1182                     CRLF
1183
1184    chunk          = chunk-size [ chunk-extension ] CRLF
1185                     chunk-data CRLF
1186    chunk-size     = 1*HEX
1187    last-chunk     = 1*("0") [ chunk-extension ] CRLF
1188
1189    chunk-extension= *( ";" chunk-ext-name [ "=" chunk-ext-val ] )
1190    chunk-ext-name = token
1191    chunk-ext-val  = token | quoted-string
1192    chunk-data     = chunk-size(OCTET)
1193    trailer        = *(entity-header CRLF)
1194</artwork></figure>
1195<t>
1196   The chunk-size field is a string of hex digits indicating the size of
1197   the chunk. The chunked encoding is ended by any chunk whose size is
1198   zero, followed by the trailer, which is terminated by an empty line.
1199</t>
1200<t>
1201   The trailer allows the sender to include additional HTTP header
1202   fields at the end of the message. The Trailer header field can be
1203   used to indicate which header fields are included in a trailer (see
1204   <xref target="header.trailer"/>).
1205</t>
1206<t>
1207   A server using chunked transfer-coding in a response &MUST-NOT; use the
1208   trailer for any header fields unless at least one of the following is
1209   true:
1210  <list style="numbers">
1211    <t>the request included a TE header field that indicates "trailers" is
1212     acceptable in the transfer-coding of the  response, as described in
1213     <xref target="header.te"/>; or,</t>
1214
1215    <t>the server is the origin server for the response, the trailer
1216     fields consist entirely of optional metadata, and the recipient
1217     could use the message (in a manner acceptable to the origin server)
1218     without receiving this metadata.  In other words, the origin server
1219     is willing to accept the possibility that the trailer fields might
1220     be silently discarded along the path to the client.</t>
1221  </list>
1222</t>
1223<t>
1224   This requirement prevents an interoperability failure when the
1225   message is being received by an HTTP/1.1 (or later) proxy and
1226   forwarded to an HTTP/1.0 recipient. It avoids a situation where
1227   compliance with the protocol would have necessitated a possibly
1228   infinite buffer on the proxy.
1229</t>
1230<t>
1231   A process for decoding the "chunked" transfer-coding
1232   can be represented in pseudo-code as:
1233</t>
1234<figure><artwork type="code">
1235    length := 0
1236    read chunk-size, chunk-extension (if any) and CRLF
1237    while (chunk-size &gt; 0) {
1238       read chunk-data and CRLF
1239       append chunk-data to entity-body
1240       length := length + chunk-size
1241       read chunk-size and CRLF
1242    }
1243    read entity-header
1244    while (entity-header not empty) {
1245       append entity-header to existing header fields
1246       read entity-header
1247    }
1248    Content-Length := length
1249    Remove "chunked" from Transfer-Encoding
1250</artwork></figure>
1251<t>
1252   All HTTP/1.1 applications &MUST; be able to receive and decode the
1253   "chunked" transfer-coding, and &MUST; ignore chunk-extension extensions
1254   they do not understand.
1255</t>
1256</section>
1257</section>
1258
1259</section>
1260
1261<section title="HTTP Message" anchor="http.message">
1262
1263<section title="Message Types" anchor="message.types">
1264<t>
1265   HTTP messages consist of requests from client to server and responses
1266   from server to client.
1267</t>
1268<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="HTTP-message"/>
1269    HTTP-message   = Request | Response     ; HTTP/1.1 messages
1270</artwork></figure>
1271<t>
1272   Request (<xref target="request"/>) and Response (<xref target="response"/>) messages use the generic
1273   message format of RFC 822 <xref target="RFC822"/> for transferring entities (the payload
1274   of the message). Both types of message consist of a start-line, zero
1275   or more header fields (also known as "headers"), an empty line (i.e.,
1276   a line with nothing preceding the CRLF) indicating the end of the
1277   header fields, and possibly a message-body.
1278</t>
1279<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="generic-message"/><iref primary="true" item="Grammar" subitem="start-line"/>
1280     generic-message = start-line
1281                       *(message-header CRLF)
1282                       CRLF
1283                       [ message-body ]
1284     start-line      = Request-Line | Status-Line
1285</artwork></figure>
1286<t>
1287   In the interest of robustness, servers &SHOULD; ignore any empty
1288   line(s) received where a Request-Line is expected. In other words, if
1289   the server is reading the protocol stream at the beginning of a
1290   message and receives a CRLF first, it should ignore the CRLF.
1291</t>
1292<t>
1293   Certain buggy HTTP/1.0 client implementations generate extra CRLF's
1294   after a POST request. To restate what is explicitly forbidden by the
1295   BNF, an HTTP/1.1 client &MUST-NOT; preface or follow a request with an
1296   extra CRLF.
1297</t>
1298</section>
1299
1300<section title="Message Headers" anchor="message.headers">
1301<t>
1302   HTTP header fields, which include general-header (<xref target="general.header.fields"/>),
1303   request-header (&request-header-fields;), response-header (&response-header-fields;), and
1304   entity-header (&entity-header-fields;) fields, follow the same generic format as
1305   that given in <xref target="RFC822" x:fmt="sec" x:sec="3.1"/> of RFC 822 <xref target="RFC822"/>. Each header field consists
1306   of a name followed by a colon (":") and the field value. Field names
1307   are case-insensitive. The field value &MAY; be preceded by any amount
1308   of LWS, though a single SP is preferred. Header fields can be
1309   extended over multiple lines by preceding each extra line with at
1310   least one SP or HT. Applications ought to follow "common form", where
1311   one is known or indicated, when generating HTTP constructs, since
1312   there might exist some implementations that fail to accept anything
1313   beyond the common forms.
1314</t>
1315<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"/>
1316    message-header = field-name ":" [ field-value ]
1317    field-name     = token
1318    field-value    = *( field-content | LWS )
1319    field-content  = &lt;the OCTETs making up the field-value
1320                     and consisting of either *TEXT or combinations
1321                     of token, separators, and quoted-string&gt;
1322</artwork></figure>
1323<t>
1324   The field-content does not include any leading or trailing LWS:
1325   linear white space occurring before the first non-whitespace
1326   character of the field-value or after the last non-whitespace
1327   character of the field-value. Such leading or trailing LWS &MAY; be
1328   removed without changing the semantics of the field value. Any LWS
1329   that occurs between field-content &MAY; be replaced with a single SP
1330   before interpreting the field value or forwarding the message
1331   downstream.
1332</t>
1333<t>
1334   The order in which header fields with differing field names are
1335   received is not significant. However, it is "good practice" to send
1336   general-header fields first, followed by request-header or response-header
1337   fields, and ending with the entity-header fields.
1338</t>
1339<t>
1340   Multiple message-header fields with the same field-name &MAY; be
1341   present in a message if and only if the entire field-value for that
1342   header field is defined as a comma-separated list [i.e., #(values)].
1343   It &MUST; be possible to combine the multiple header fields into one
1344   "field-name: field-value" pair, without changing the semantics of the
1345   message, by appending each subsequent field-value to the first, each
1346   separated by a comma. The order in which header fields with the same
1347   field-name are received is therefore significant to the
1348   interpretation of the combined field value, and thus a proxy &MUST-NOT;
1349   change the order of these field values when a message is forwarded.
1350</t>
1351</section>
1352
1353<section title="Message Body" anchor="message.body">
1354<t>
1355   The message-body (if any) of an HTTP message is used to carry the
1356   entity-body associated with the request or response. The message-body
1357   differs from the entity-body only when a transfer-coding has been
1358   applied, as indicated by the Transfer-Encoding header field (<xref target="header.transfer-encoding"/>).
1359</t>
1360<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="message-body"/>
1361    message-body = entity-body
1362                 | &lt;entity-body encoded as per Transfer-Encoding&gt;
1363</artwork></figure>
1364<t>
1365   Transfer-Encoding &MUST; be used to indicate any transfer-codings
1366   applied by an application to ensure safe and proper transfer of the
1367   message. Transfer-Encoding is a property of the message, not of the
1368   entity, and thus &MAY; be added or removed by any application along the
1369   request/response chain. (However, <xref target="transfer.codings"/> places restrictions on
1370   when certain transfer-codings may be used.)
1371</t>
1372<t>
1373   The rules for when a message-body is allowed in a message differ for
1374   requests and responses.
1375</t>
1376<t>
1377   The presence of a message-body in a request is signaled by the
1378   inclusion of a Content-Length or Transfer-Encoding header field in
1379   the request's message-headers. A message-body &MUST-NOT; be included in
1380   a request if the specification of the request method (&method;)
1381   does not allow sending an entity-body in requests. A server &SHOULD;
1382   read and forward a message-body on any request; if the request method
1383   does not include defined semantics for an entity-body, then the
1384   message-body &SHOULD; be ignored when handling the request.
1385</t>
1386<t>
1387   For response messages, whether or not a message-body is included with
1388   a message is dependent on both the request method and the response
1389   status code (<xref target="status.code.and.reason.phrase"/>). All responses to the HEAD request method
1390   &MUST-NOT; include a message-body, even though the presence of entity-header
1391   fields might lead one to believe they do. All 1xx
1392   (informational), 204 (no content), and 304 (not modified) responses
1393   &MUST-NOT; include a message-body. All other responses do include a
1394   message-body, although it &MAY; be of zero length.
1395</t>
1396</section>
1397
1398<section title="Message Length" anchor="message.length">
1399<t>
1400   The transfer-length of a message is the length of the message-body as
1401   it appears in the message; that is, after any transfer-codings have
1402   been applied. When a message-body is included with a message, the
1403   transfer-length of that body is determined by one of the following
1404   (in order of precedence):
1405</t>
1406<t>
1407  <list style="numbers">
1408    <x:lt><t>
1409     Any response message which "&MUST-NOT;" include a message-body (such
1410     as the 1xx, 204, and 304 responses and any response to a HEAD
1411     request) is always terminated by the first empty line after the
1412     header fields, regardless of the entity-header fields present in
1413     the message.
1414    </t></x:lt>
1415    <x:lt><t>
1416     If a Transfer-Encoding header field (<xref target="header.transfer-encoding"/>) is present and
1417     has any value other than "identity", then the transfer-length is
1418     defined by use of the "chunked" transfer-coding (<xref target="transfer.codings"/>),
1419     unless the message is terminated by closing the connection.
1420    </t></x:lt>
1421    <x:lt><t>
1422     If a Content-Length header field (<xref target="header.content-length"/>) is present, its
1423     decimal value in OCTETs represents both the entity-length and the
1424     transfer-length. The Content-Length header field &MUST-NOT; be sent
1425     if these two lengths are different (i.e., if a Transfer-Encoding
1426     header field is present). If a message is received with both a
1427     Transfer-Encoding header field and a Content-Length header field,
1428     the latter &MUST; be ignored.
1429    </t></x:lt>
1430    <x:lt><t>
1431     If the message uses the media type "multipart/byteranges", and the
1432     ransfer-length is not otherwise specified, then this self-elimiting
1433     media type defines the transfer-length. This media type
1434     UST NOT be used unless the sender knows that the recipient can arse
1435     it; the presence in a request of a Range header with ultiple byte-range
1436     specifiers from a 1.1 client implies that the lient can parse
1437     multipart/byteranges responses.
1438    <list style="empty"><t>
1439       A range header might be forwarded by a 1.0 proxy that does not
1440       understand multipart/byteranges; in this case the server &MUST;
1441       delimit the message using methods defined in items 1, 3 or 5 of
1442       this section.
1443    </t></list>
1444    </t></x:lt>
1445    <x:lt><t>
1446     By the server closing the connection. (Closing the connection
1447     cannot be used to indicate the end of a request body, since that
1448     would leave no possibility for the server to send back a response.)
1449    </t></x:lt>
1450  </list>
1451</t>
1452<t>
1453   For compatibility with HTTP/1.0 applications, HTTP/1.1 requests
1454   containing a message-body &MUST; include a valid Content-Length header
1455   field unless the server is known to be HTTP/1.1 compliant. If a
1456   request contains a message-body and a Content-Length is not given,
1457   the server &SHOULD; respond with 400 (bad request) if it cannot
1458   determine the length of the message, or with 411 (length required) if
1459   it wishes to insist on receiving a valid Content-Length.
1460</t>
1461<t>
1462   All HTTP/1.1 applications that receive entities &MUST; accept the
1463   "chunked" transfer-coding (<xref target="transfer.codings"/>), thus allowing this mechanism
1464   to be used for messages when the message length cannot be determined
1465   in advance.
1466</t>
1467<t>
1468   Messages &MUST-NOT; include both a Content-Length header field and a
1469   non-identity transfer-coding. If the message does include a non-identity
1470   transfer-coding, the Content-Length &MUST; be ignored.
1471</t>
1472<t>
1473   When a Content-Length is given in a message where a message-body is
1474   allowed, its field value &MUST; exactly match the number of OCTETs in
1475   the message-body. HTTP/1.1 user agents &MUST; notify the user when an
1476   invalid length is received and detected.
1477</t>
1478</section>
1479
1480<section title="General Header Fields" anchor="general.header.fields">
1481<t>
1482   There are a few header fields which have general applicability for
1483   both request and response messages, but which do not apply to the
1484   entity being transferred. These header fields apply only to the
1485   message being transmitted.
1486</t>
1487<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="general-header"/>
1488    general-header = Cache-Control            ; &header-cache-control;
1489                   | Connection               ; <xref target="header.connection"/>
1490                   | Date                     ; <xref target="header.date"/>
1491                   | Pragma                   ; &header-pragma;
1492                   | Trailer                  ; <xref target="header.trailer"/>
1493                   | Transfer-Encoding        ; <xref target="header.transfer-encoding"/>
1494                   | Upgrade                  ; <xref target="header.upgrade"/>
1495                   | Via                      ; <xref target="header.via"/>
1496                   | Warning                  ; &header-warning;
1497</artwork></figure>
1498<t>
1499   General-header field names can be extended reliably only in
1500   combination with a change in the protocol version. However, new or
1501   experimental header fields may be given the semantics of general
1502   header fields if all parties in the communication recognize them to
1503   be general-header fields. Unrecognized header fields are treated as
1504   entity-header fields.
1505</t>
1506</section>
1507</section>
1508
1509<section title="Request" anchor="request">
1510<t>
1511   A request message from a client to a server includes, within the
1512   first line of that message, the method to be applied to the resource,
1513   the identifier of the resource, and the protocol version in use.
1514</t>
1515<!--                 Host                      ; should be moved here eventually -->
1516<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="Request"/>
1517     Request       = Request-Line              ; <xref target="request-line"/>
1518                     *(( general-header        ; <xref target="general.header.fields"/>
1519                      | request-header         ; &request-header-fields;
1520                      | entity-header ) CRLF)  ; &entity-header-fields;
1521                     CRLF
1522                     [ message-body ]          ; <xref target="message.body"/>
1523</artwork></figure>
1524
1525<section title="Request-Line" anchor="request-line">
1526<t>
1527   The Request-Line begins with a method token, followed by the
1528   Request-URI and the protocol version, and ending with CRLF. The
1529   elements are separated by SP characters. No CR or LF is allowed
1530   except in the final CRLF sequence.
1531</t>
1532<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="Request-Line"/>
1533     Request-Line   = Method SP Request-URI SP HTTP-Version CRLF
1534</artwork></figure>
1535
1536<section title="Method" anchor="method">
1537<t>
1538   The Method  token indicates the method to be performed on the
1539   resource identified by the Request-URI. The method is case-sensitive.
1540</t>
1541<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="Method"/><iref primary="true" item="Grammar" subitem="extension-method"/>
1542    Method         = token
1543</artwork></figure>
1544</section>
1545
1546<section title="Request-URI" anchor="request-uri">
1547<t>
1548   The Request-URI is a Uniform Resource Identifier (<xref target="uri"/>) and
1549   identifies the resource upon which to apply the request.
1550</t>
1551<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="Request-URI"/>
1552    Request-URI    = "*" | absoluteURI | abs_path | authority
1553</artwork></figure>
1554<t>
1555   The four options for Request-URI are dependent on the nature of the
1556   request. The asterisk "*" means that the request does not apply to a
1557   particular resource, but to the server itself, and is only allowed
1558   when the method used does not necessarily apply to a resource. One
1559   example would be
1560</t>
1561<figure><artwork type="example">
1562    OPTIONS * HTTP/1.1
1563</artwork></figure>
1564<t>
1565   The absoluteURI form is &REQUIRED; when the request is being made to a
1566   proxy. The proxy is requested to forward the request or service it
1567   from a valid cache, and return the response. Note that the proxy &MAY;
1568   forward the request on to another proxy or directly to the server
1569   specified by the absoluteURI. In order to avoid request loops, a
1570   proxy &MUST; be able to recognize all of its server names, including
1571   any aliases, local variations, and the numeric IP address. An example
1572   Request-Line would be:
1573</t>
1574<figure><artwork type="example">
1575    GET http://www.w3.org/pub/WWW/TheProject.html HTTP/1.1
1576</artwork></figure>
1577<t>
1578   To allow for transition to absoluteURIs in all requests in future
1579   versions of HTTP, all HTTP/1.1 servers &MUST; accept the absoluteURI
1580   form in requests, even though HTTP/1.1 clients will only generate
1581   them in requests to proxies.
1582</t>
1583<t>
1584   The authority form is only used by the CONNECT method (&CONNECT;).
1585</t>
1586<t>
1587   The most common form of Request-URI is that used to identify a
1588   resource on an origin server or gateway. In this case the absolute
1589   path of the URI &MUST; be transmitted (see <xref target="general.syntax"/>, abs_path) as
1590   the Request-URI, and the network location of the URI (authority) &MUST;
1591   be transmitted in a Host header field. For example, a client wishing
1592   to retrieve the resource above directly from the origin server would
1593   create a TCP connection to port 80 of the host "www.w3.org" and send
1594   the lines:
1595</t>
1596<figure><artwork type="example">
1597    GET /pub/WWW/TheProject.html HTTP/1.1
1598    Host: www.w3.org
1599</artwork></figure>
1600<t>
1601   followed by the remainder of the Request. Note that the absolute path
1602   cannot be empty; if none is present in the original URI, it &MUST; be
1603   given as "/" (the server root).
1604</t>
1605<t>
1606   The Request-URI is transmitted in the format specified in
1607   <xref target="general.syntax"/>. If the Request-URI is encoded using the "% HEX HEX" encoding
1608   <xref target="RFC2396"/>, the origin server &MUST; decode the Request-URI in order to
1609   properly interpret the request. Servers &SHOULD; respond to invalid
1610   Request-URIs with an appropriate status code.
1611</t>
1612<t>
1613   A transparent proxy &MUST-NOT; rewrite the "abs_path" part of the
1614   received Request-URI when forwarding it to the next inbound server,
1615   except as noted above to replace a null abs_path with "/".
1616</t>
1617<t>
1618  <list><t>
1619      <x:h>Note:</x:h> The "no rewrite" rule prevents the proxy from changing the
1620      meaning of the request when the origin server is improperly using
1621      a non-reserved URI character for a reserved purpose.  Implementors
1622      should be aware that some pre-HTTP/1.1 proxies have been known to
1623      rewrite the Request-URI.
1624  </t></list>
1625</t>
1626</section>
1627</section>
1628
1629<section title="The Resource Identified by a Request" anchor="the.resource.identified.by.a.request">
1630<t>
1631   The exact resource identified by an Internet request is determined by
1632   examining both the Request-URI and the Host header field.
1633</t>
1634<t>
1635   An origin server that does not allow resources to differ by the
1636   requested host &MAY; ignore the Host header field value when
1637   determining the resource identified by an HTTP/1.1 request. (But see
1638   <xref target="changes.to.simplify.multi-homed.web.servers.and.conserve.ip.addresses"/>
1639   for other requirements on Host support in HTTP/1.1.)
1640</t>
1641<t>
1642   An origin server that does differentiate resources based on the host
1643   requested (sometimes referred to as virtual hosts or vanity host
1644   names) &MUST; use the following rules for determining the requested
1645   resource on an HTTP/1.1 request:
1646  <list style="numbers">
1647    <t>If Request-URI is an absoluteURI, the host is part of the
1648     Request-URI. Any Host header field value in the request &MUST; be
1649     ignored.</t>
1650    <t>If the Request-URI is not an absoluteURI, and the request includes
1651     a Host header field, the host is determined by the Host header
1652     field value.</t>
1653    <t>If the host as determined by rule 1 or 2 is not a valid host on
1654     the server, the response &MUST; be a 400 (Bad Request) error message.</t>
1655  </list>
1656</t>
1657<t>
1658   Recipients of an HTTP/1.0 request that lacks a Host header field &MAY;
1659   attempt to use heuristics (e.g., examination of the URI path for
1660   something unique to a particular host) in order to determine what
1661   exact resource is being requested.
1662</t>
1663</section>
1664
1665</section>
1666
1667
1668<section title="Response" anchor="response">
1669<t>
1670   After receiving and interpreting a request message, a server responds
1671   with an HTTP response message.
1672</t>
1673<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="Response"/>
1674    Response      = Status-Line               ; <xref target="status-line"/>
1675                    *(( general-header        ; <xref target="general.header.fields"/>
1676                     | response-header        ; &response-header-fields;
1677                     | entity-header ) CRLF)  ; &entity-header-fields;
1678                    CRLF
1679                    [ message-body ]          ; <xref target="message.body"/>
1680</artwork></figure>
1681
1682<section title="Status-Line" anchor="status-line">
1683<t>
1684   The first line of a Response message is the Status-Line, consisting
1685   of the protocol version followed by a numeric status code and its
1686   associated textual phrase, with each element separated by SP
1687   characters. No CR or LF is allowed except in the final CRLF sequence.
1688</t>
1689<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="Status-Line"/>
1690    Status-Line = HTTP-Version SP Status-Code SP Reason-Phrase CRLF
1691</artwork></figure>
1692
1693<section title="Status Code and Reason Phrase" anchor="status.code.and.reason.phrase">
1694<t>
1695   The Status-Code element is a 3-digit integer result code of the
1696   attempt to understand and satisfy the request. These codes are fully
1697   defined in &status-codes;. The Reason-Phrase is intended to give a short
1698   textual description of the Status-Code. The Status-Code is intended
1699   for use by automata and the Reason-Phrase is intended for the human
1700   user. The client is not required to examine or display the Reason-Phrase.
1701</t>
1702<t>
1703   The first digit of the Status-Code defines the class of response. The
1704   last two digits do not have any categorization role. There are 5
1705   values for the first digit:
1706  <list style="symbols">
1707    <t>
1708      1xx: Informational - Request received, continuing process
1709    </t>
1710    <t>
1711      2xx: Success - The action was successfully received,
1712        understood, and accepted
1713    </t>
1714    <t>
1715      3xx: Redirection - Further action must be taken in order to
1716        complete the request
1717    </t>
1718    <t>
1719      4xx: Client Error - The request contains bad syntax or cannot
1720        be fulfilled
1721    </t>
1722    <t>
1723      5xx: Server Error - The server failed to fulfill an apparently
1724        valid request
1725    </t>
1726  </list>
1727</t>
1728<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"/>
1729   Status-Code    = 3DIGIT
1730   Reason-Phrase  = *&lt;TEXT, excluding CR, LF&gt;
1731</artwork></figure>
1732</section>
1733</section>
1734
1735</section>
1736
1737
1738<section title="Connections" anchor="connections">
1739
1740<section title="Persistent Connections" anchor="persistent.connections">
1741
1742<section title="Purpose" anchor="persistent.purpose">
1743<t>
1744   Prior to persistent connections, a separate TCP connection was
1745   established to fetch each URL, increasing the load on HTTP servers
1746   and causing congestion on the Internet. The use of inline images and
1747   other associated data often require a client to make multiple
1748   requests of the same server in a short amount of time. Analysis of
1749   these performance problems and results from a prototype
1750   implementation are available <xref target="Pad1995"/> <xref target="Spe"/>. Implementation experience and
1751   measurements of actual HTTP/1.1 (RFC 2068) implementations show good
1752   results <xref target="Nie1997"/>. Alternatives have also been explored, for example,
1753   T/TCP <xref target="Tou1998"/>.
1754</t>
1755<t>
1756   Persistent HTTP connections have a number of advantages:
1757  <list style="symbols">
1758      <t>
1759        By opening and closing fewer TCP connections, CPU time is saved
1760        in routers and hosts (clients, servers, proxies, gateways,
1761        tunnels, or caches), and memory used for TCP protocol control
1762        blocks can be saved in hosts.
1763      </t>
1764      <t>
1765        HTTP requests and responses can be pipelined on a connection.
1766        Pipelining allows a client to make multiple requests without
1767        waiting for each response, allowing a single TCP connection to
1768        be used much more efficiently, with much lower elapsed time.
1769      </t>
1770      <t>
1771        Network congestion is reduced by reducing the number of packets
1772        caused by TCP opens, and by allowing TCP sufficient time to
1773        determine the congestion state of the network.
1774      </t>
1775      <t>
1776        Latency on subsequent requests is reduced since there is no time
1777        spent in TCP's connection opening handshake.
1778      </t>
1779      <t>
1780        HTTP can evolve more gracefully, since errors can be reported
1781        without the penalty of closing the TCP connection. Clients using
1782        future versions of HTTP might optimistically try a new feature,
1783        but if communicating with an older server, retry with old
1784        semantics after an error is reported.
1785      </t>
1786    </list>
1787</t>
1788<t>
1789   HTTP implementations &SHOULD; implement persistent connections.
1790</t>
1791</section>
1792
1793<section title="Overall Operation" anchor="persistent.overall">
1794<t>
1795   A significant difference between HTTP/1.1 and earlier versions of
1796   HTTP is that persistent connections are the default behavior of any
1797   HTTP connection. That is, unless otherwise indicated, the client
1798   &SHOULD; assume that the server will maintain a persistent connection,
1799   even after error responses from the server.
1800</t>
1801<t>
1802   Persistent connections provide a mechanism by which a client and a
1803   server can signal the close of a TCP connection. This signaling takes
1804   place using the Connection header field (<xref target="header.connection"/>). Once a close
1805   has been signaled, the client &MUST-NOT; send any more requests on that
1806   connection.
1807</t>
1808
1809<section title="Negotiation" anchor="persistent.negotiation">
1810<t>
1811   An HTTP/1.1 server &MAY; assume that a HTTP/1.1 client intends to
1812   maintain a persistent connection unless a Connection header including
1813   the connection-token "close" was sent in the request. If the server
1814   chooses to close the connection immediately after sending the
1815   response, it &SHOULD; send a Connection header including the
1816   connection-token close.
1817</t>
1818<t>
1819   An HTTP/1.1 client &MAY; expect a connection to remain open, but would
1820   decide to keep it open based on whether the response from a server
1821   contains a Connection header with the connection-token close. In case
1822   the client does not want to maintain a connection for more than that
1823   request, it &SHOULD; send a Connection header including the
1824   connection-token close.
1825</t>
1826<t>
1827   If either the client or the server sends the close token in the
1828   Connection header, that request becomes the last one for the
1829   connection.
1830</t>
1831<t>
1832   Clients and servers &SHOULD-NOT;  assume that a persistent connection is
1833   maintained for HTTP versions less than 1.1 unless it is explicitly
1834   signaled. See <xref target="compatibility.with.http.1.0.persistent.connections"/> for more information on backward
1835   compatibility with HTTP/1.0 clients.
1836</t>
1837<t>
1838   In order to remain persistent, all messages on the connection &MUST;
1839   have a self-defined message length (i.e., one not defined by closure
1840   of the connection), as described in <xref target="message.length"/>.
1841</t>
1842</section>
1843
1844<section title="Pipelining" anchor="pipelining">
1845<t>
1846   A client that supports persistent connections &MAY; "pipeline" its
1847   requests (i.e., send multiple requests without waiting for each
1848   response). A server &MUST; send its responses to those requests in the
1849   same order that the requests were received.
1850</t>
1851<t>
1852   Clients which assume persistent connections and pipeline immediately
1853   after connection establishment &SHOULD; be prepared to retry their
1854   connection if the first pipelined attempt fails. If a client does
1855   such a retry, it &MUST-NOT; pipeline before it knows the connection is
1856   persistent. Clients &MUST; also be prepared to resend their requests if
1857   the server closes the connection before sending all of the
1858   corresponding responses.
1859</t>
1860<t>
1861   Clients &SHOULD-NOT;  pipeline requests using non-idempotent methods or
1862   non-idempotent sequences of methods (see &idempotent-methods;). Otherwise, a
1863   premature termination of the transport connection could lead to
1864   indeterminate results. A client wishing to send a non-idempotent
1865   request &SHOULD; wait to send that request until it has received the
1866   response status for the previous request.
1867</t>
1868</section>
1869</section>
1870
1871<section title="Proxy Servers" anchor="persistent.proxy">
1872<t>
1873   It is especially important that proxies correctly implement the
1874   properties of the Connection header field as specified in <xref target="header.connection"/>.
1875</t>
1876<t>
1877   The proxy server &MUST; signal persistent connections separately with
1878   its clients and the origin servers (or other proxy servers) that it
1879   connects to. Each persistent connection applies to only one transport
1880   link.
1881</t>
1882<t>
1883   A proxy server &MUST-NOT; establish a HTTP/1.1 persistent connection
1884   with an HTTP/1.0 client (but see RFC 2068 <xref target="RFC2068"/> for information and
1885   discussion of the problems with the Keep-Alive header implemented by
1886   many HTTP/1.0 clients).
1887</t>
1888</section>
1889
1890<section title="Practical Considerations" anchor="persistent.practical">
1891<t>
1892   Servers will usually have some time-out value beyond which they will
1893   no longer maintain an inactive connection. Proxy servers might make
1894   this a higher value since it is likely that the client will be making
1895   more connections through the same server. The use of persistent
1896   connections places no requirements on the length (or existence) of
1897   this time-out for either the client or the server.
1898</t>
1899<t>
1900   When a client or server wishes to time-out it &SHOULD; issue a graceful
1901   close on the transport connection. Clients and servers &SHOULD; both
1902   constantly watch for the other side of the transport close, and
1903   respond to it as appropriate. If a client or server does not detect
1904   the other side's close promptly it could cause unnecessary resource
1905   drain on the network.
1906</t>
1907<t>
1908   A client, server, or proxy &MAY; close the transport connection at any
1909   time. For example, a client might have started to send a new request
1910   at the same time that the server has decided to close the "idle"
1911   connection. From the server's point of view, the connection is being
1912   closed while it was idle, but from the client's point of view, a
1913   request is in progress.
1914</t>
1915<t>
1916   This means that clients, servers, and proxies &MUST; be able to recover
1917   from asynchronous close events. Client software &SHOULD; reopen the
1918   transport connection and retransmit the aborted sequence of requests
1919   without user interaction so long as the request sequence is
1920   idempotent (see &idempotent-methods;). Non-idempotent methods or sequences
1921   &MUST-NOT; be automatically retried, although user agents &MAY; offer a
1922   human operator the choice of retrying the request(s). Confirmation by
1923   user-agent software with semantic understanding of the application
1924   &MAY; substitute for user confirmation. The automatic retry &SHOULD-NOT; 
1925   be repeated if the second sequence of requests fails.
1926</t>
1927<t>
1928   Servers &SHOULD; always respond to at least one request per connection,
1929   if at all possible. Servers &SHOULD-NOT;  close a connection in the
1930   middle of transmitting a response, unless a network or client failure
1931   is suspected.
1932</t>
1933<t>
1934   Clients that use persistent connections &SHOULD; limit the number of
1935   simultaneous connections that they maintain to a given server. A
1936   single-user client &SHOULD-NOT; maintain more than 2 connections with
1937   any server or proxy. A proxy &SHOULD; use up to 2*N connections to
1938   another server or proxy, where N is the number of simultaneously
1939   active users. These guidelines are intended to improve HTTP response
1940   times and avoid congestion.
1941</t>
1942</section>
1943</section>
1944
1945<section title="Message Transmission Requirements" anchor="message.transmission.requirements">
1946
1947<section title="Persistent Connections and Flow Control" anchor="persistent.flow">
1948<t>
1949   HTTP/1.1 servers &SHOULD; maintain persistent connections and use TCP's
1950   flow control mechanisms to resolve temporary overloads, rather than
1951   terminating connections with the expectation that clients will retry.
1952   The latter technique can exacerbate network congestion.
1953</t>
1954</section>
1955
1956<section title="Monitoring Connections for Error Status Messages" anchor="persistent.monitor">
1957<t>
1958   An HTTP/1.1 (or later) client sending a message-body &SHOULD; monitor
1959   the network connection for an error status while it is transmitting
1960   the request. If the client sees an error status, it &SHOULD;
1961   immediately cease transmitting the body. If the body is being sent
1962   using a "chunked" encoding (<xref target="transfer.codings"/>), a zero length chunk and
1963   empty trailer &MAY; be used to prematurely mark the end of the message.
1964   If the body was preceded by a Content-Length header, the client &MUST;
1965   close the connection.
1966</t>
1967</section>
1968
1969<section title="Use of the 100 (Continue) Status" anchor="use.of.the.100.status">
1970<t>
1971   The purpose of the 100 (Continue) status (see &status-100;) is to
1972   allow a client that is sending a request message with a request body
1973   to determine if the origin server is willing to accept the request
1974   (based on the request headers) before the client sends the request
1975   body. In some cases, it might either be inappropriate or highly
1976   inefficient for the client to send the body if the server will reject
1977   the message without looking at the body.
1978</t>
1979<t>
1980   Requirements for HTTP/1.1 clients:
1981  <list style="symbols">
1982    <t>
1983        If a client will wait for a 100 (Continue) response before
1984        sending the request body, it &MUST; send an Expect request-header
1985        field (&header-expect;) with the "100-continue" expectation.
1986    </t>
1987    <t>
1988        A client &MUST-NOT; send an Expect request-header field (&header-expect;)
1989        with the "100-continue" expectation if it does not intend
1990        to send a request body.
1991    </t>
1992  </list>
1993</t>
1994<t>
1995   Because of the presence of older implementations, the protocol allows
1996   ambiguous situations in which a client may send "Expect: 100-continue"
1997   without receiving either a 417 (Expectation Failed) status
1998   or a 100 (Continue) status. Therefore, when a client sends this
1999   header field to an origin server (possibly via a proxy) from which it
2000   has never seen a 100 (Continue) status, the client &SHOULD-NOT;  wait
2001   for an indefinite period before sending the request body.
2002</t>
2003<t>
2004   Requirements for HTTP/1.1 origin servers:
2005  <list style="symbols">
2006    <t> Upon receiving a request which includes an Expect request-header
2007        field with the "100-continue" expectation, an origin server &MUST;
2008        either respond with 100 (Continue) status and continue to read
2009        from the input stream, or respond with a final status code. The
2010        origin server &MUST-NOT; wait for the request body before sending
2011        the 100 (Continue) response. If it responds with a final status
2012        code, it &MAY; close the transport connection or it &MAY; continue
2013        to read and discard the rest of the request.  It &MUST-NOT;
2014        perform the requested method if it returns a final status code.
2015    </t>
2016    <t> An origin server &SHOULD-NOT;  send a 100 (Continue) response if
2017        the request message does not include an Expect request-header
2018        field with the "100-continue" expectation, and &MUST-NOT; send a
2019        100 (Continue) response if such a request comes from an HTTP/1.0
2020        (or earlier) client. There is an exception to this rule: for
2021        compatibility with RFC 2068, a server &MAY; send a 100 (Continue)
2022        status in response to an HTTP/1.1 PUT or POST request that does
2023        not include an Expect request-header field with the "100-continue"
2024        expectation. This exception, the purpose of which is
2025        to minimize any client processing delays associated with an
2026        undeclared wait for 100 (Continue) status, applies only to
2027        HTTP/1.1 requests, and not to requests with any other HTTP-version
2028        value.
2029    </t>
2030    <t> An origin server &MAY; omit a 100 (Continue) response if it has
2031        already received some or all of the request body for the
2032        corresponding request.
2033    </t>
2034    <t> An origin server that sends a 100 (Continue) response &MUST;
2035    ultimately send a final status code, once the request body is
2036        received and processed, unless it terminates the transport
2037        connection prematurely.
2038    </t>
2039    <t> If an origin server receives a request that does not include an
2040        Expect request-header field with the "100-continue" expectation,
2041        the request includes a request body, and the server responds
2042        with a final status code before reading the entire request body
2043        from the transport connection, then the server &SHOULD-NOT;  close
2044        the transport connection until it has read the entire request,
2045        or until the client closes the connection. Otherwise, the client
2046        might not reliably receive the response message. However, this
2047        requirement is not be construed as preventing a server from
2048        defending itself against denial-of-service attacks, or from
2049        badly broken client implementations.
2050      </t>
2051    </list>
2052</t>
2053<t>
2054   Requirements for HTTP/1.1 proxies:
2055  <list style="symbols">
2056    <t> If a proxy receives a request that includes an Expect request-header
2057        field with the "100-continue" expectation, and the proxy
2058        either knows that the next-hop server complies with HTTP/1.1 or
2059        higher, or does not know the HTTP version of the next-hop
2060        server, it &MUST; forward the request, including the Expect header
2061        field.
2062    </t>
2063    <t> If the proxy knows that the version of the next-hop server is
2064        HTTP/1.0 or lower, it &MUST-NOT; forward the request, and it &MUST;
2065        respond with a 417 (Expectation Failed) status.
2066    </t>
2067    <t> Proxies &SHOULD; maintain a cache recording the HTTP version
2068        numbers received from recently-referenced next-hop servers.
2069    </t>
2070    <t> A proxy &MUST-NOT; forward a 100 (Continue) response if the
2071        request message was received from an HTTP/1.0 (or earlier)
2072        client and did not include an Expect request-header field with
2073        the "100-continue" expectation. This requirement overrides the
2074        general rule for forwarding of 1xx responses (see &status-1xx;).
2075    </t>
2076  </list>
2077</t>
2078</section>
2079
2080<section title="Client Behavior if Server Prematurely Closes Connection" anchor="connection.premature">
2081<t>
2082   If an HTTP/1.1 client sends a request which includes a request body,
2083   but which does not include an Expect request-header field with the
2084   "100-continue" expectation, and if the client is not directly
2085   connected to an HTTP/1.1 origin server, and if the client sees the
2086   connection close before receiving any status from the server, the
2087   client &SHOULD; retry the request.  If the client does retry this
2088   request, it &MAY; use the following "binary exponential backoff"
2089   algorithm to be assured of obtaining a reliable response:
2090  <list style="numbers">
2091    <t>
2092      Initiate a new connection to the server
2093    </t>
2094    <t>
2095      Transmit the request-headers
2096    </t>
2097    <t>
2098      Initialize a variable R to the estimated round-trip time to the
2099         server (e.g., based on the time it took to establish the
2100         connection), or to a constant value of 5 seconds if the round-trip
2101         time is not available.
2102    </t>
2103    <t>
2104       Compute T = R * (2**N), where N is the number of previous
2105         retries of this request.
2106    </t>
2107    <t>
2108       Wait either for an error response from the server, or for T
2109         seconds (whichever comes first)
2110    </t>
2111    <t>
2112       If no error response is received, after T seconds transmit the
2113         body of the request.
2114    </t>
2115    <t>
2116       If client sees that the connection is closed prematurely,
2117         repeat from step 1 until the request is accepted, an error
2118         response is received, or the user becomes impatient and
2119         terminates the retry process.
2120    </t>
2121  </list>
2122</t>
2123<t>
2124   If at any point an error status is received, the client
2125  <list style="symbols">
2126      <t>&SHOULD-NOT;  continue and</t>
2127
2128      <t>&SHOULD; close the connection if it has not completed sending the
2129        request message.</t>
2130    </list>
2131</t>
2132</section>
2133</section>
2134</section>
2135
2136
2137<section title="Header Field Definitions" anchor="header.fields">
2138<t>
2139   This section defines the syntax and semantics of all standard
2140   HTTP/1.1 header fields. For entity-header fields, both sender and
2141   recipient refer to either the client or the server, depending on who
2142   sends and who receives the entity.
2143</t>
2144
2145<section title="Connection" anchor="header.connection">
2146  <iref primary="true" item="Connection header" x:for-anchor=""/>
2147  <iref primary="true" item="Headers" subitem="Connection" x:for-anchor=""/>
2148<t>
2149   The Connection general-header field allows the sender to specify
2150   options that are desired for that particular connection and &MUST-NOT;
2151   be communicated by proxies over further connections.
2152</t>
2153<t>
2154   The Connection header has the following grammar:
2155</t>
2156<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="Connection"/><iref primary="true" item="Grammar" subitem="connection-token"/>
2157    Connection = "Connection" ":" 1#(connection-token)
2158    connection-token  = token
2159</artwork></figure>
2160<t>
2161   HTTP/1.1 proxies &MUST; parse the Connection header field before a
2162   message is forwarded and, for each connection-token in this field,
2163   remove any header field(s) from the message with the same name as the
2164   connection-token. Connection options are signaled by the presence of
2165   a connection-token in the Connection header field, not by any
2166   corresponding additional header field(s), since the additional header
2167   field may not be sent if there are no parameters associated with that
2168   connection option.
2169</t>
2170<t>
2171   Message headers listed in the Connection header &MUST-NOT; include
2172   end-to-end headers, such as Cache-Control.
2173</t>
2174<t>
2175   HTTP/1.1 defines the "close" connection option for the sender to
2176   signal that the connection will be closed after completion of the
2177   response. For example,
2178</t>
2179<figure><artwork type="example">
2180    Connection: close
2181</artwork></figure>
2182<t>
2183   in either the request or the response header fields indicates that
2184   the connection &SHOULD-NOT;  be considered `persistent' (<xref target="persistent.connections"/>)
2185   after the current request/response is complete.
2186</t>
2187<t>
2188   HTTP/1.1 applications that do not support persistent connections &MUST;
2189   include the "close" connection option in every message.
2190</t>
2191<t>
2192   A system receiving an HTTP/1.0 (or lower-version) message that
2193   includes a Connection header MUST, for each connection-token in this
2194   field, remove and ignore any header field(s) from the message with
2195   the same name as the connection-token. This protects against mistaken
2196   forwarding of such header fields by pre-HTTP/1.1 proxies. See <xref target="compatibility.with.http.1.0.persistent.connections"/>.
2197</t>
2198</section>
2199
2200<section title="Content-Length" anchor="header.content-length">
2201  <iref primary="true" item="Content-Length header" x:for-anchor=""/>
2202  <iref primary="true" item="Headers" subitem="Content-Length" x:for-anchor=""/>
2203<t>
2204   The Content-Length entity-header field indicates the size of the
2205   entity-body, in decimal number of OCTETs, sent to the recipient or,
2206   in the case of the HEAD method, the size of the entity-body that
2207   would have been sent had the request been a GET.
2208</t>
2209<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="Content-Length"/>
2210    Content-Length    = "Content-Length" ":" 1*DIGIT
2211</artwork></figure>
2212<t>
2213   An example is
2214</t>
2215<figure><artwork type="example">
2216    Content-Length: 3495
2217</artwork></figure>
2218<t>
2219   Applications &SHOULD; use this field to indicate the transfer-length of
2220   the message-body, unless this is prohibited by the rules in <xref target="message.length"/>.
2221</t>
2222<t>
2223   Any Content-Length greater than or equal to zero is a valid value.
2224   <xref target="message.length"/> describes how to determine the length of a message-body
2225   if a Content-Length is not given.
2226</t>
2227<t>
2228   Note that the meaning of this field is significantly different from
2229   the corresponding definition in MIME, where it is an optional field
2230   used within the "message/external-body" content-type. In HTTP, it
2231   &SHOULD; be sent whenever the message's length can be determined prior
2232   to being transferred, unless this is prohibited by the rules in
2233   <xref target="message.length"/>.
2234</t>
2235</section>
2236
2237<section title="Date" anchor="header.date">
2238  <iref primary="true" item="Date header" x:for-anchor=""/>
2239  <iref primary="true" item="Headers" subitem="Date" x:for-anchor=""/>
2240<t>
2241   The Date general-header field represents the date and time at which
2242   the message was originated, having the same semantics as orig-date in
2243   RFC 822. The field value is an HTTP-date, as described in <xref target="full.date"/>;
2244   it &MUST; be sent in RFC 1123 <xref target="RFC1123"/>-date format.
2245</t>
2246<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="Date"/>
2247    Date  = "Date" ":" HTTP-date
2248</artwork></figure>
2249<t>
2250   An example is
2251</t>
2252<figure><artwork type="example">
2253    Date: Tue, 15 Nov 1994 08:12:31 GMT
2254</artwork></figure>
2255<t>
2256   Origin servers &MUST; include a Date header field in all responses,
2257   except in these cases:
2258  <list style="numbers">
2259      <t>If the response status code is 100 (Continue) or 101 (Switching
2260         Protocols), the response &MAY; include a Date header field, at
2261         the server's option.</t>
2262
2263      <t>If the response status code conveys a server error, e.g. 500
2264         (Internal Server Error) or 503 (Service Unavailable), and it is
2265         inconvenient or impossible to generate a valid Date.</t>
2266
2267      <t>If the server does not have a clock that can provide a
2268         reasonable approximation of the current time, its responses
2269         &MUST-NOT; include a Date header field. In this case, the rules
2270         in <xref target="clockless.origin.server.operation"/> &MUST; be followed.</t>
2271  </list>
2272</t>
2273<t>
2274   A received message that does not have a Date header field &MUST; be
2275   assigned one by the recipient if the message will be cached by that
2276   recipient or gatewayed via a protocol which requires a Date. An HTTP
2277   implementation without a clock &MUST-NOT; cache responses without
2278   revalidating them on every use. An HTTP cache, especially a shared
2279   cache, &SHOULD; use a mechanism, such as NTP <xref target="RFC1305"/>, to synchronize its
2280   clock with a reliable external standard.
2281</t>
2282<t>
2283   Clients &SHOULD; only send a Date header field in messages that include
2284   an entity-body, as in the case of the PUT and POST requests, and even
2285   then it is optional. A client without a clock &MUST-NOT; send a Date
2286   header field in a request.
2287</t>
2288<t>
2289   The HTTP-date sent in a Date header &SHOULD-NOT;  represent a date and
2290   time subsequent to the generation of the message. It &SHOULD; represent
2291   the best available approximation of the date and time of message
2292   generation, unless the implementation has no means of generating a
2293   reasonably accurate date and time. In theory, the date ought to
2294   represent the moment just before the entity is generated. In
2295   practice, the date can be generated at any time during the message
2296   origination without affecting its semantic value.
2297</t>
2298
2299<section title="Clockless Origin Server Operation" anchor="clockless.origin.server.operation">
2300<t>
2301   Some origin server implementations might not have a clock available.
2302   An origin server without a clock &MUST-NOT; assign Expires or Last-Modified
2303   values to a response, unless these values were associated
2304   with the resource by a system or user with a reliable clock. It &MAY;
2305   assign an Expires value that is known, at or before server
2306   configuration time, to be in the past (this allows "pre-expiration"
2307   of responses without storing separate Expires values for each
2308   resource).
2309</t>
2310</section>
2311</section>
2312
2313<section title="Host" anchor="header.host">
2314  <iref primary="true" item="Host header" x:for-anchor=""/>
2315  <iref primary="true" item="Headers" subitem="Host" x:for-anchor=""/>
2316<t>
2317   The Host request-header field specifies the Internet host and port
2318   number of the resource being requested, as obtained from the original
2319   URI given by the user or referring resource (generally an HTTP URL,
2320   as described in <xref target="http.url"/>). The Host field value &MUST; represent
2321   the naming authority of the origin server or gateway given by the
2322   original URL. This allows the origin server or gateway to
2323   differentiate between internally-ambiguous URLs, such as the root "/"
2324   URL of a server for multiple host names on a single IP address.
2325</t>
2326<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="Host"/>
2327    Host = "Host" ":" host [ ":" port ] ; <xref target="http.url"/>
2328</artwork></figure>
2329<t>
2330   A "host" without any trailing port information implies the default
2331   port for the service requested (e.g., "80" for an HTTP URL). For
2332   example, a request on the origin server for
2333   &lt;http://www.w3.org/pub/WWW/&gt; would properly include:
2334</t>
2335<figure><artwork type="example">
2336    GET /pub/WWW/ HTTP/1.1
2337    Host: www.w3.org
2338</artwork></figure>
2339<t>
2340   A client &MUST; include a Host header field in all HTTP/1.1 request
2341   messages . If the requested URI does not include an Internet host
2342   name for the service being requested, then the Host header field &MUST;
2343   be given with an empty value. An HTTP/1.1 proxy &MUST; ensure that any
2344   request message it forwards does contain an appropriate Host header
2345   field that identifies the service being requested by the proxy. All
2346   Internet-based HTTP/1.1 servers &MUST; respond with a 400 (Bad Request)
2347   status code to any HTTP/1.1 request message which lacks a Host header
2348   field.
2349</t>
2350<t>
2351   See sections <xref target="the.resource.identified.by.a.request" format="counter"/>
2352   and <xref target="changes.to.simplify.multi-homed.web.servers.and.conserve.ip.addresses" format="counter"/>
2353   for other requirements relating to Host.
2354</t>
2355</section>
2356
2357<section title="TE" anchor="header.te">
2358  <iref primary="true" item="TE header" x:for-anchor=""/>
2359  <iref primary="true" item="Headers" subitem="TE" x:for-anchor=""/>
2360<t>
2361   The TE request-header field indicates what extension transfer-codings
2362   it is willing to accept in the response and whether or not it is
2363   willing to accept trailer fields in a chunked transfer-coding. Its
2364   value may consist of the keyword "trailers" and/or a comma-separated
2365   list of extension transfer-coding names with optional accept
2366   parameters (as described in <xref target="transfer.codings"/>).
2367</t>
2368<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="TE"/><iref primary="true" item="Grammar" subitem="t-codings"/>
2369    TE        = "TE" ":" #( t-codings )
2370    t-codings = "trailers" | ( transfer-extension [ accept-params ] )
2371</artwork></figure>
2372<t>
2373   The presence of the keyword "trailers" indicates that the client is
2374   willing to accept trailer fields in a chunked transfer-coding, as
2375   defined in <xref target="chunked.transfer.encoding"/>. This keyword is reserved for use with
2376   transfer-coding values even though it does not itself represent a
2377   transfer-coding.
2378</t>
2379<t>
2380   Examples of its use are:
2381</t>
2382<figure><artwork type="example">
2383    TE: deflate
2384    TE:
2385    TE: trailers, deflate;q=0.5
2386</artwork></figure>
2387<t>
2388   The TE header field only applies to the immediate connection.
2389   Therefore, the keyword &MUST; be supplied within a Connection header
2390   field (<xref target="header.connection"/>) whenever TE is present in an HTTP/1.1 message.
2391</t>
2392<t>
2393   A server tests whether a transfer-coding is acceptable, according to
2394   a TE field, using these rules:
2395  <list style="numbers">
2396    <x:lt>
2397      <t>The "chunked" transfer-coding is always acceptable. If the
2398         keyword "trailers" is listed, the client indicates that it is
2399         willing to accept trailer fields in the chunked response on
2400         behalf of itself and any downstream clients. The implication is
2401         that, if given, the client is stating that either all
2402         downstream clients are willing to accept trailer fields in the
2403         forwarded response, or that it will attempt to buffer the
2404         response on behalf of downstream recipients.
2405      </t><t>
2406         <x:h>Note:</x:h> HTTP/1.1 does not define any means to limit the size of a
2407         chunked response such that a client can be assured of buffering
2408         the entire response.</t>
2409    </x:lt>
2410    <x:lt>
2411      <t>If the transfer-coding being tested is one of the transfer-codings
2412         listed in the TE field, then it is acceptable unless it
2413         is accompanied by a qvalue of 0. (As defined in &qvalue;, a
2414         qvalue of 0 means "not acceptable.")</t>
2415    </x:lt>
2416    <x:lt>
2417      <t>If multiple transfer-codings are acceptable, then the
2418         acceptable transfer-coding with the highest non-zero qvalue is
2419         preferred.  The "chunked" transfer-coding always has a qvalue
2420         of 1.</t>
2421    </x:lt>
2422  </list>
2423</t>
2424<t>
2425   If the TE field-value is empty or if no TE field is present, the only
2426   transfer-coding  is "chunked". A message with no transfer-coding is
2427   always acceptable.
2428</t>
2429</section>
2430
2431<section title="Trailer" anchor="header.trailer">
2432  <iref primary="true" item="Trailer header" x:for-anchor=""/>
2433  <iref primary="true" item="Headers" subitem="Trailer" x:for-anchor=""/>
2434<t>
2435   The Trailer general field value indicates that the given set of
2436   header fields is present in the trailer of a message encoded with
2437   chunked transfer-coding.
2438</t>
2439<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="Trailer"/>
2440    Trailer  = "Trailer" ":" 1#field-name
2441</artwork></figure>
2442<t>
2443   An HTTP/1.1 message &SHOULD; include a Trailer header field in a
2444   message using chunked transfer-coding with a non-empty trailer. Doing
2445   so allows the recipient to know which header fields to expect in the
2446   trailer.
2447</t>
2448<t>
2449   If no Trailer header field is present, the trailer &SHOULD-NOT;  include
2450   any header fields. See <xref target="chunked.transfer.encoding"/> for restrictions on the use of
2451   trailer fields in a "chunked" transfer-coding.
2452</t>
2453<t>
2454   Message header fields listed in the Trailer header field &MUST-NOT;
2455   include the following header fields:
2456  <list style="symbols">
2457    <t>Transfer-Encoding</t>
2458    <t>Content-Length</t>
2459    <t>Trailer</t>
2460  </list>
2461</t>
2462</section>
2463
2464<section title="Transfer-Encoding" anchor="header.transfer-encoding">
2465  <iref primary="true" item="Transfer-Encoding header" x:for-anchor=""/>
2466  <iref primary="true" item="Headers" subitem="Transfer-Encoding" x:for-anchor=""/>
2467<t>
2468   The Transfer-Encoding general-header field indicates what (if any)
2469   type of transformation has been applied to the message body in order
2470   to safely transfer it between the sender and the recipient. This
2471   differs from the content-coding in that the transfer-coding is a
2472   property of the message, not of the entity.
2473</t>
2474<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="Transfer-Encoding"/>
2475  Transfer-Encoding       = "Transfer-Encoding" ":" 1#transfer-coding
2476</artwork></figure>
2477<t>
2478   Transfer-codings are defined in <xref target="transfer.codings"/>. An example is:
2479</t>
2480<figure><artwork type="example">
2481  Transfer-Encoding: chunked
2482</artwork></figure>
2483<t>
2484   If multiple encodings have been applied to an entity, the transfer-codings
2485   &MUST; be listed in the order in which they were applied.
2486   Additional information about the encoding parameters &MAY; be provided
2487   by other entity-header fields not defined by this specification.
2488</t>
2489<t>
2490   Many older HTTP/1.0 applications do not understand the Transfer-Encoding
2491   header.
2492</t>
2493</section>
2494
2495<section title="Upgrade" anchor="header.upgrade">
2496  <iref primary="true" item="Upgrade header" x:for-anchor=""/>
2497  <iref primary="true" item="Headers" subitem="Upgrade" x:for-anchor=""/>
2498<t>
2499   The Upgrade general-header allows the client to specify what
2500   additional communication protocols it supports and would like to use
2501   if the server finds it appropriate to switch protocols. The server
2502   &MUST; use the Upgrade header field within a 101 (Switching Protocols)
2503   response to indicate which protocol(s) are being switched.
2504</t>
2505<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="Upgrade"/>
2506    Upgrade        = "Upgrade" ":" 1#product
2507</artwork></figure>
2508<t>
2509   For example,
2510</t>
2511<figure><artwork type="example">
2512    Upgrade: HTTP/2.0, SHTTP/1.3, IRC/6.9, RTA/x11
2513</artwork></figure>
2514<t>
2515   The Upgrade header field is intended to provide a simple mechanism
2516   for transition from HTTP/1.1 to some other, incompatible protocol. It
2517   does so by allowing the client to advertise its desire to use another
2518   protocol, such as a later version of HTTP with a higher major version
2519   number, even though the current request has been made using HTTP/1.1.
2520   This eases the difficult transition between incompatible protocols by
2521   allowing the client to initiate a request in the more commonly
2522   supported protocol while indicating to the server that it would like
2523   to use a "better" protocol if available (where "better" is determined
2524   by the server, possibly according to the nature of the method and/or
2525   resource being requested).
2526</t>
2527<t>
2528   The Upgrade header field only applies to switching application-layer
2529   protocols upon the existing transport-layer connection. Upgrade
2530   cannot be used to insist on a protocol change; its acceptance and use
2531   by the server is optional. The capabilities and nature of the
2532   application-layer communication after the protocol change is entirely
2533   dependent upon the new protocol chosen, although the first action
2534   after changing the protocol &MUST; be a response to the initial HTTP
2535   request containing the Upgrade header field.
2536</t>
2537<t>
2538   The Upgrade header field only applies to the immediate connection.
2539   Therefore, the upgrade keyword &MUST; be supplied within a Connection
2540   header field (<xref target="header.connection"/>) whenever Upgrade is present in an
2541   HTTP/1.1 message.
2542</t>
2543<t>
2544   The Upgrade header field cannot be used to indicate a switch to a
2545   protocol on a different connection. For that purpose, it is more
2546   appropriate to use a 301, 302, 303, or 305 redirection response.
2547</t>
2548<t>
2549   This specification only defines the protocol name "HTTP" for use by
2550   the family of Hypertext Transfer Protocols, as defined by the HTTP
2551   version rules of <xref target="http.version"/> and future updates to this
2552   specification. Any token can be used as a protocol name; however, it
2553   will only be useful if both the client and server associate the name
2554   with the same protocol.
2555</t>
2556</section>
2557
2558<section title="Via" anchor="header.via">
2559  <iref primary="true" item="Via header" x:for-anchor=""/>
2560  <iref primary="true" item="Headers" subitem="Via" x:for-anchor=""/>
2561<t>
2562   The Via general-header field &MUST; be used by gateways and proxies to
2563   indicate the intermediate protocols and recipients between the user
2564   agent and the server on requests, and between the origin server and
2565   the client on responses. It is analogous to the "Received" field of
2566   RFC 822 <xref target="RFC822"/> and is intended to be used for tracking message forwards,
2567   avoiding request loops, and identifying the protocol capabilities of
2568   all senders along the request/response chain.
2569</t>
2570<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"/>
2571   Via =  "Via" ":" 1#( received-protocol received-by [ comment ] )
2572   received-protocol = [ protocol-name "/" ] protocol-version
2573   protocol-name     = token
2574   protocol-version  = token
2575   received-by       = ( host [ ":" port ] ) | pseudonym
2576   pseudonym         = token
2577</artwork></figure>
2578<t>
2579   The received-protocol indicates the protocol version of the message
2580   received by the server or client along each segment of the
2581   request/response chain. The received-protocol version is appended to
2582   the Via field value when the message is forwarded so that information
2583   about the protocol capabilities of upstream applications remains
2584   visible to all recipients.
2585</t>
2586<t>
2587   The protocol-name is optional if and only if it would be "HTTP". The
2588   received-by field is normally the host and optional port number of a
2589   recipient server or client that subsequently forwarded the message.
2590   However, if the real host is considered to be sensitive information,
2591   it &MAY; be replaced by a pseudonym. If the port is not given, it &MAY;
2592   be assumed to be the default port of the received-protocol.
2593</t>
2594<t>
2595   Multiple Via field values represents each proxy or gateway that has
2596   forwarded the message. Each recipient &MUST; append its information
2597   such that the end result is ordered according to the sequence of
2598   forwarding applications.
2599</t>
2600<t>
2601   Comments &MAY; be used in the Via header field to identify the software
2602   of the recipient proxy or gateway, analogous to the User-Agent and
2603   Server header fields. However, all comments in the Via field are
2604   optional and &MAY; be removed by any recipient prior to forwarding the
2605   message.
2606</t>
2607<t>
2608   For example, a request message could be sent from an HTTP/1.0 user
2609   agent to an internal proxy code-named "fred", which uses HTTP/1.1 to
2610   forward the request to a public proxy at nowhere.com, which completes
2611   the request by forwarding it to the origin server at www.ics.uci.edu.
2612   The request received by www.ics.uci.edu would then have the following
2613   Via header field:
2614</t>
2615<figure><artwork type="example">
2616    Via: 1.0 fred, 1.1 nowhere.com (Apache/1.1)
2617</artwork></figure>
2618<t>
2619   Proxies and gateways used as a portal through a network firewall
2620   &SHOULD-NOT;, by default, forward the names and ports of hosts within
2621   the firewall region. This information &SHOULD; only be propagated if
2622   explicitly enabled. If not enabled, the received-by host of any host
2623   behind the firewall &SHOULD; be replaced by an appropriate pseudonym
2624   for that host.
2625</t>
2626<t>
2627   For organizations that have strong privacy requirements for hiding
2628   internal structures, a proxy &MAY; combine an ordered subsequence of
2629   Via header field entries with identical received-protocol values into
2630   a single such entry. For example,
2631</t>
2632<figure><artwork type="example">
2633    Via: 1.0 ricky, 1.1 ethel, 1.1 fred, 1.0 lucy
2634</artwork></figure>
2635<t>
2636        could be collapsed to
2637</t>
2638<figure><artwork type="example">
2639    Via: 1.0 ricky, 1.1 mertz, 1.0 lucy
2640</artwork></figure>
2641<t>
2642   Applications &SHOULD-NOT;  combine multiple entries unless they are all
2643   under the same organizational control and the hosts have already been
2644   replaced by pseudonyms. Applications &MUST-NOT; combine entries which
2645   have different received-protocol values.
2646</t>
2647</section>
2648
2649</section>
2650
2651<section title="IANA Considerations" anchor="IANA.considerations">
2652<t>
2653   TBD.
2654</t>
2655</section>
2656
2657<section title="Security Considerations" anchor="security.considerations">
2658<t>
2659   This section is meant to inform application developers, information
2660   providers, and users of the security limitations in HTTP/1.1 as
2661   described by this document. The discussion does not include
2662   definitive solutions to the problems revealed, though it does make
2663   some suggestions for reducing security risks.
2664</t>
2665
2666<section title="Personal Information" anchor="personal.information">
2667<t>
2668   HTTP clients are often privy to large amounts of personal information
2669   (e.g. the user's name, location, mail address, passwords, encryption
2670   keys, etc.), and &SHOULD; be very careful to prevent unintentional
2671   leakage of this information via the HTTP protocol to other sources.
2672   We very strongly recommend that a convenient interface be provided
2673   for the user to control dissemination of such information, and that
2674   designers and implementors be particularly careful in this area.
2675   History shows that errors in this area often create serious security
2676   and/or privacy problems and generate highly adverse publicity for the
2677   implementor's company.
2678</t>
2679</section>
2680
2681<section title="Abuse of Server Log Information" anchor="abuse.of.server.log.information">
2682<t>
2683   A server is in the position to save personal data about a user's
2684   requests which might identify their reading patterns or subjects of
2685   interest. This information is clearly confidential in nature and its
2686   handling can be constrained by law in certain countries. People using
2687   the HTTP protocol to provide data are responsible for ensuring that
2688   such material is not distributed without the permission of any
2689   individuals that are identifiable by the published results.
2690</t>
2691</section>
2692
2693<section title="Attacks Based On File and Path Names" anchor="attack.pathname">
2694<t>
2695   Implementations of HTTP origin servers &SHOULD; be careful to restrict
2696   the documents returned by HTTP requests to be only those that were
2697   intended by the server administrators. If an HTTP server translates
2698   HTTP URIs directly into file system calls, the server &MUST; take
2699   special care not to serve files that were not intended to be
2700   delivered to HTTP clients. For example, UNIX, Microsoft Windows, and
2701   other operating systems use ".." as a path component to indicate a
2702   directory level above the current one. On such a system, an HTTP
2703   server &MUST; disallow any such construct in the Request-URI if it
2704   would otherwise allow access to a resource outside those intended to
2705   be accessible via the HTTP server. Similarly, files intended for
2706   reference only internally to the server (such as access control
2707   files, configuration files, and script code) &MUST; be protected from
2708   inappropriate retrieval, since they might contain sensitive
2709   information. Experience has shown that minor bugs in such HTTP server
2710   implementations have turned into security risks.
2711</t>
2712</section>
2713
2714<section title="DNS Spoofing" anchor="dns.spoofing">
2715<t>
2716   Clients using HTTP rely heavily on the Domain Name Service, and are
2717   thus generally prone to security attacks based on the deliberate
2718   mis-association of IP addresses and DNS names. Clients need to be
2719   cautious in assuming the continuing validity of an IP number/DNS name
2720   association.
2721</t>
2722<t>
2723   In particular, HTTP clients &SHOULD; rely on their name resolver for
2724   confirmation of an IP number/DNS name association, rather than
2725   caching the result of previous host name lookups. Many platforms
2726   already can cache host name lookups locally when appropriate, and
2727   they &SHOULD; be configured to do so. It is proper for these lookups to
2728   be cached, however, only when the TTL (Time To Live) information
2729   reported by the name server makes it likely that the cached
2730   information will remain useful.
2731</t>
2732<t>
2733   If HTTP clients cache the results of host name lookups in order to
2734   achieve a performance improvement, they &MUST; observe the TTL
2735   information reported by DNS.
2736</t>
2737<t>
2738   If HTTP clients do not observe this rule, they could be spoofed when
2739   a previously-accessed server's IP address changes. As network
2740   renumbering is expected to become increasingly common <xref target="RFC1900"/>, the
2741   possibility of this form of attack will grow. Observing this
2742   requirement thus reduces this potential security vulnerability.
2743</t>
2744<t>
2745   This requirement also improves the load-balancing behavior of clients
2746   for replicated servers using the same DNS name and reduces the
2747   likelihood of a user's experiencing failure in accessing sites which
2748   use that strategy.
2749</t>
2750</section>
2751
2752<section title="Proxies and Caching" anchor="attack.proxies">
2753<t>
2754   By their very nature, HTTP proxies are men-in-the-middle, and
2755   represent an opportunity for man-in-the-middle attacks. Compromise of
2756   the systems on which the proxies run can result in serious security
2757   and privacy problems. Proxies have access to security-related
2758   information, personal information about individual users and
2759   organizations, and proprietary information belonging to users and
2760   content providers. A compromised proxy, or a proxy implemented or
2761   configured without regard to security and privacy considerations,
2762   might be used in the commission of a wide range of potential attacks.
2763</t>
2764<t>
2765   Proxy operators should protect the systems on which proxies run as
2766   they would protect any system that contains or transports sensitive
2767   information. In particular, log information gathered at proxies often
2768   contains highly sensitive personal information, and/or information
2769   about organizations. Log information should be carefully guarded, and
2770   appropriate guidelines for use developed and followed. (<xref target="abuse.of.server.log.information"/>).
2771</t>
2772<t>
2773   Proxy implementors should consider the privacy and security
2774   implications of their design and coding decisions, and of the
2775   configuration options they provide to proxy operators (especially the
2776   default configuration).
2777</t>
2778<t>
2779   Users of a proxy need to be aware that they are no trustworthier than
2780   the people who run the proxy; HTTP itself cannot solve this problem.
2781</t>
2782<t>
2783   The judicious use of cryptography, when appropriate, may suffice to
2784   protect against a broad range of security and privacy attacks. Such
2785   cryptography is beyond the scope of the HTTP/1.1 specification.
2786</t>
2787</section>
2788
2789<section title="Denial of Service Attacks on Proxies" anchor="attack.DoS">
2790<t>
2791   They exist. They are hard to defend against. Research continues.
2792   Beware.
2793</t>
2794</section>
2795</section>
2796
2797<section title="Acknowledgments" anchor="ack">
2798<t>
2799   This specification makes heavy use of the augmented BNF and generic
2800   constructs defined by David H. Crocker for RFC 822 <xref target="RFC822"/>. Similarly, it
2801   reuses many of the definitions provided by Nathaniel Borenstein and
2802   Ned Freed for MIME <xref target="RFC2045"/>. We hope that their inclusion in this
2803   specification will help reduce past confusion over the relationship
2804   between HTTP and Internet mail message formats.
2805</t>
2806<t>
2807   The HTTP protocol has evolved considerably over the years. It has
2808   benefited from a large and active developer community--the many
2809   people who have participated on the www-talk mailing list--and it is
2810   that community which has been most responsible for the success of
2811   HTTP and of the World-Wide Web in general. Marc Andreessen, Robert
2812   Cailliau, Daniel W. Connolly, Bob Denny, John Franks, Jean-Francois
2813   Groff, Phillip M. Hallam-Baker, Hakon W. Lie, Ari Luotonen, Rob
2814   McCool, Lou Montulli, Dave Raggett, Tony Sanders, and Marc
2815   VanHeyningen deserve special recognition for their efforts in
2816   defining early aspects of the protocol.
2817</t>
2818<t>
2819   This document has benefited greatly from the comments of all those
2820   participating in the HTTP-WG. In addition to those already mentioned,
2821   the following individuals have contributed to this specification:
2822</t>
2823<figure><artwork type="text/plain">
2824    Gary Adams                  Ross Patterson
2825    Harald Tveit Alvestrand     Albert Lunde
2826    Keith Ball                  John C. Mallery
2827    Brian Behlendorf            Jean-Philippe Martin-Flatin
2828    Paul Burchard               Mitra
2829    Maurizio Codogno            David Morris
2830    Mike Cowlishaw              Gavin Nicol
2831    Roman Czyborra              Bill Perry
2832    Michael A. Dolan            Jeffrey Perry
2833    David J. Fiander            Scott Powers
2834    Alan Freier                 Owen Rees
2835    Marc Hedlund                Luigi Rizzo
2836    Greg Herlihy                David Robinson
2837    Koen Holtman                Marc Salomon
2838    Alex Hopmann                Rich Salz
2839    Bob Jernigan                Allan M. Schiffman
2840    Shel Kaphan                 Jim Seidman
2841    Rohit Khare                 Chuck Shotton
2842    John Klensin                Eric W. Sink
2843    Martijn Koster              Simon E. Spero
2844    Alexei Kosut                Richard N. Taylor
2845    David M. Kristol            Robert S. Thau
2846    Daniel LaLiberte            Bill (BearHeart) Weinman
2847    Ben Laurie                  Francois Yergeau
2848    Paul J. Leach               Mary Ellen Zurko
2849    Daniel DuBois               Josh Cohen
2850</artwork></figure>
2851<t>
2852   Thanks to the "cave men" of Palo Alto. You know who you are.
2853</t>
2854<t>
2855   Jim Gettys (the current editor of this document) wishes particularly
2856   to thank Roy Fielding, the previous editor of this document, along
2857   with John Klensin, Jeff Mogul, Paul Leach, Dave Kristol, Koen
2858   Holtman, John Franks, Josh Cohen, Alex Hopmann, Scott Lawrence, and
2859   Larry Masinter for their help. And thanks go particularly to Jeff
2860   Mogul and Scott Lawrence for performing the "MUST/MAY/SHOULD" audit.
2861</t>
2862<t>
2863   The Apache Group, Anselm Baird-Smith, author of Jigsaw, and Henrik
2864   Frystyk implemented RFC 2068 early, and we wish to thank them for the
2865   discovery of many of the problems that this document attempts to
2866   rectify.
2867</t>
2868<t>
2869   Based on an XML translation of RFC 2616 by Julian Reschke.
2870</t>
2871</section>
2872
2873</middle>
2874<back>
2875<references>
2876
2877<reference anchor="Part2">
2878  <front>
2879    <title abbrev="HTTP/1.1">HTTP/1.1, part 2: Message Semantics</title>
2880 
2881    <author initials="R." surname="Fielding" fullname="Roy T. Fielding" role="editor">
2882      <organization abbrev="Day Software">Day Software</organization>
2883      <address>
2884        <email>fielding@gbiv.com</email>
2885      </address>
2886    </author>
2887 
2888    <author initials="J." surname="Gettys" fullname="Jim Gettys">
2889      <organization>One Laptop per Child</organization>
2890      <address>
2891        <email>jg@laptop.org</email>
2892      </address>
2893    </author>
2894   
2895    <author initials="J." surname="Mogul" fullname="Jeffrey C. Mogul">
2896      <organization abbrev="HP">Hewlett-Packard Company</organization>
2897      <address>
2898        <email>JeffMogul@acm.org</email>
2899      </address>
2900    </author>
2901 
2902    <author initials="H." surname="Frystyk" fullname="Henrik Frystyk Nielsen">
2903      <organization abbrev="Microsoft">Microsoft Corporation</organization>
2904      <address>
2905        <email>henrikn@microsoft.com</email>
2906      </address>
2907    </author>
2908 
2909    <author initials="L." surname="Masinter" fullname="Larry Masinter">
2910      <organization abbrev="Adobe Systems">Adobe Systems, Incorporated</organization>
2911      <address>
2912        <email>LMM@acm.org</email>
2913      </address>
2914    </author>
2915   
2916    <author initials="P." surname="Leach" fullname="Paul J. Leach">
2917      <organization abbrev="Microsoft">Microsoft Corporation</organization>
2918      <address>
2919        <email>paulle@microsoft.com</email>
2920      </address>
2921    </author>
2922     
2923    <author initials="T." surname="Berners-Lee" fullname="Tim Berners-Lee">
2924      <organization abbrev="W3C/MIT">World Wide Web Consortium</organization>
2925      <address>
2926        <email>timbl@w3.org</email>
2927      </address>
2928    </author>
2929 
2930    <date month="&ID-MONTH;" year="&ID-YEAR;"/>
2931  </front>
2932  <seriesInfo name="Internet-Draft" value="draft-fielding-p2-semantics-&ID-VERSION;"/>
2933  <x:source href="p2-semantics.xml" basename="p2-semantics"/>
2934</reference>
2935
2936<reference anchor="Part3">
2937  <front>
2938    <title abbrev="HTTP/1.1">HTTP/1.1, part 3: Message Payload and Content Negotiation</title>
2939 
2940    <author initials="R." surname="Fielding" fullname="Roy T. Fielding" role="editor">
2941      <organization abbrev="Day Software">Day Software</organization>
2942      <address>
2943        <email>fielding@gbiv.com</email>
2944      </address>
2945    </author>
2946 
2947    <author initials="J." surname="Gettys" fullname="Jim Gettys">
2948      <organization>One Laptop per Child</organization>
2949      <address>
2950        <email>jg@laptop.org</email>
2951      </address>
2952    </author>
2953   
2954    <author initials="J." surname="Mogul" fullname="Jeffrey C. Mogul">
2955      <organization abbrev="HP">Hewlett-Packard Company</organization>
2956      <address>
2957        <email>JeffMogul@acm.org</email>
2958      </address>
2959    </author>
2960 
2961    <author initials="H." surname="Frystyk" fullname="Henrik Frystyk Nielsen">
2962      <organization abbrev="Microsoft">Microsoft Corporation</organization>
2963      <address>
2964        <email>henrikn@microsoft.com</email>
2965      </address>
2966    </author>
2967 
2968    <author initials="L." surname="Masinter" fullname="Larry Masinter">
2969      <organization abbrev="Adobe Systems">Adobe Systems, Incorporated</organization>
2970      <address>
2971        <email>LMM@acm.org</email>
2972      </address>
2973    </author>
2974   
2975    <author initials="P." surname="Leach" fullname="Paul J. Leach">
2976      <organization abbrev="Microsoft">Microsoft Corporation</organization>
2977      <address>
2978        <email>paulle@microsoft.com</email>
2979      </address>
2980    </author>
2981     
2982    <author initials="T." surname="Berners-Lee" fullname="Tim Berners-Lee">
2983      <organization abbrev="W3C/MIT">World Wide Web Consortium</organization>
2984      <address>
2985        <email>timbl@w3.org</email>
2986      </address>
2987    </author>
2988 
2989    <date month="&ID-MONTH;" year="&ID-YEAR;"/>
2990  </front>
2991  <seriesInfo name="Internet-Draft" value="draft-fielding-p3-payload-&ID-VERSION;"/>
2992  <x:source href="p3-payload.xml" basename="p3-payload"/>
2993</reference>
2994
2995<reference anchor="Part6">
2996  <front>
2997    <title abbrev="HTTP/1.1">HTTP/1.1, part 6: Caching</title>
2998 
2999    <author initials="R." surname="Fielding" fullname="Roy T. Fielding" role="editor">
3000      <organization abbrev="Day Software">Day Software</organization>
3001      <address>
3002        <email>fielding@gbiv.com</email>
3003      </address>
3004    </author>
3005 
3006    <author initials="J." surname="Gettys" fullname="Jim Gettys">
3007      <organization>One Laptop per Child</organization>
3008      <address>
3009        <email>jg@laptop.org</email>
3010      </address>
3011    </author>
3012   
3013    <author initials="J." surname="Mogul" fullname="Jeffrey C. Mogul">
3014      <organization abbrev="HP">Hewlett-Packard Company</organization>
3015      <address>
3016        <email>JeffMogul@acm.org</email>
3017      </address>
3018    </author>
3019 
3020    <author initials="H." surname="Frystyk" fullname="Henrik Frystyk Nielsen">
3021      <organization abbrev="Microsoft">Microsoft Corporation</organization>
3022      <address>
3023        <email>henrikn@microsoft.com</email>
3024      </address>
3025    </author>
3026 
3027    <author initials="L." surname="Masinter" fullname="Larry Masinter">
3028      <organization abbrev="Adobe Systems">Adobe Systems, Incorporated</organization>
3029      <address>
3030        <email>LMM@acm.org</email>
3031      </address>
3032    </author>
3033   
3034    <author initials="P." surname="Leach" fullname="Paul J. Leach">
3035      <organization abbrev="Microsoft">Microsoft Corporation</organization>
3036      <address>
3037        <email>paulle@microsoft.com</email>
3038      </address>
3039    </author>
3040     
3041    <author initials="T." surname="Berners-Lee" fullname="Tim Berners-Lee">
3042      <organization abbrev="W3C/MIT">World Wide Web Consortium</organization>
3043      <address>
3044        <email>timbl@w3.org</email>
3045      </address>
3046    </author>
3047 
3048    <date month="&ID-MONTH;" year="&ID-YEAR;"/>
3049  </front>
3050  <seriesInfo name="Internet-Draft" value="draft-fielding-p6-cache-&ID-VERSION;"/>
3051  <x:source href="p6-cache.xml" basename="p6-cache"/>
3052</reference>
3053
3054<reference anchor="RFC1436">
3055<front>
3056<title abbrev="Gopher">The Internet Gopher Protocol (a distributed document search and retrieval protocol)</title>
3057<author initials="F." surname="Anklesaria" fullname="Farhad Anklesaria">
3058<organization>University of Minnesota, Computer and Information Services</organization>
3059<address>
3060<postal>
3061<street>100 Union Street SE</street>
3062<street>Room 152 Shepherd Labs</street>
3063<city>Minneapolis</city>
3064<region>MN</region>
3065<code>55455</code>
3066<country>US</country></postal>
3067<phone>+1 612 625 1300</phone>
3068<email>fxa@boombox.micro.umn.edu</email></address></author>
3069<author initials="M." surname="McCahill" fullname="Mark McCahill">
3070<organization>University of Minnesota, Computer and Information Services</organization>
3071<address>
3072<postal>
3073<street>100 Union Street SE</street>
3074<street>Room 152 Shepherd Labs</street>
3075<city>Minneapolis</city>
3076<region>MN</region>
3077<code>55455</code>
3078<country>US</country></postal>
3079<phone>+1 612 625 1300</phone>
3080<email>mpm@boombox.micro.umn.edu</email></address></author>
3081<author initials="P." surname="Lindner" fullname="Paul Lindner">
3082<organization>University of Minnesota, Computer and Information Services</organization>
3083<address>
3084<postal>
3085<street>100 Union Street SE</street>
3086<street>Room 152 Shepherd Labs</street>
3087<city>Minneapolis</city>
3088<region>MN</region>
3089<code>55455</code>
3090<country>US</country></postal>
3091<phone>+1 612 625 1300</phone>
3092<email>lindner@boombox.micro.umn.edu</email></address></author>
3093<author initials="D." surname="Johnson" fullname="David Johnson">
3094<organization>University of Minnesota, Computer and Information Services</organization>
3095<address>
3096<postal>
3097<street>100 Union Street SE</street>
3098<street>Room 152 Shepherd Labs</street>
3099<city>Minneapolis</city>
3100<region>MN</region>
3101<code>55455</code>
3102<country>US</country></postal>
3103<phone>+1 612 625 1300</phone>
3104<email>dmj@boombox.micro.umn.edu</email></address></author>
3105<author initials="D." surname="Torrey" fullname="Daniel Torrey">
3106<organization>University of Minnesota, Computer and Information Services</organization>
3107<address>
3108<postal>
3109<street>100 Union Street SE</street>
3110<street>Room 152 Shepherd Labs</street>
3111<city>Minneapolis</city>
3112<region>MN</region>
3113<code>55455</code>
3114<country>US</country></postal>
3115<phone>+1 612 625 1300</phone>
3116<email>daniel@boombox.micro.umn.edu</email></address></author>
3117<author initials="B." surname="Alberti" fullname="Bob Alberti">
3118<organization>University of Minnesota, Computer and Information Services</organization>
3119<address>
3120<postal>
3121<street>100 Union Street SE</street>
3122<street>Room 152 Shepherd Labs</street>
3123<city>Minneapolis</city>
3124<region>MN</region>
3125<code>55455</code>
3126<country>US</country></postal>
3127<phone>+1 612 625 1300</phone>
3128<email>alberti@boombox.micro.umn.edu</email></address></author>
3129<date month="March" year="1993"/>
3130</front>
3131<seriesInfo name="RFC" value="1436"/>
3132</reference>
3133
3134<reference anchor="RFC1630">
3135<front>
3136<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>
3137<author initials="T." surname="Berners-Lee" fullname="Tim Berners-Lee">
3138<organization>CERN, World-Wide Web project</organization>
3139<address>
3140<postal>
3141<street>1211 Geneva 23</street>
3142<city/>
3143<region/>
3144<code/>
3145<country>CH</country></postal>
3146<phone>+41 22 7673755</phone>
3147<facsimile>+41 22 7677155</facsimile>
3148<email>timbl@info.cern.ch</email></address></author>
3149<date month="June" year="1994"/></front>
3150<seriesInfo name="RFC" value="1630"/>
3151</reference>
3152
3153<reference anchor="RFC1738">
3154<front>
3155<title>Uniform Resource Locators (URL)</title>
3156<author initials="T." surname="Berners-Lee" fullname="Tim Berners-Lee">
3157<organization>CERN, World-Wide Web project</organization>
3158<address>
3159<postal>
3160<street>1211 Geneva 23</street>
3161<city/>
3162<region/>
3163<code/>
3164<country>CH</country></postal>
3165<phone>+41 22 7673755</phone>
3166<facsimile>+41 22 7677155</facsimile>
3167<email>timbl@info.cern.ch</email></address></author>
3168<author initials="L." surname="Masinter" fullname="Larry Masinter">
3169<organization>Xerox PARC</organization>
3170<address>
3171<postal>
3172<street>3333 Coyote Hill Road</street>
3173<city>Palo Alto</city>
3174<region>CA</region>
3175<code>94034</code>
3176<country>US</country></postal>
3177<phone>+1 415 812 4365</phone>
3178<facsimile>+1 415 812 4333</facsimile>
3179<email>masinter@parc.xerox.com</email></address></author>
3180<author initials="M." surname="McCahill" fullname="Mark McCahill">
3181<organization>University of Minnesota, Computer and Information Services</organization>
3182<address>
3183<postal>
3184<street>100 Union Street SE, Shepherd Labs</street>
3185<street>Room 152</street>
3186<city>Minneapolis</city>
3187<region>MN</region>
3188<code>55455</code>
3189<country>US</country></postal>
3190<phone>+1 612 625 1300</phone>
3191<email>mpm@boombox.micro.umn.edu</email></address></author>
3192<date month="December" year="1994"/>
3193</front>
3194<seriesInfo name="RFC" value="1738"/>
3195</reference>
3196
3197<reference anchor="RFC1945">
3198<front>
3199<title abbrev="HTTP/1.0">Hypertext Transfer Protocol -- HTTP/1.0</title>
3200<author initials="T." surname="Berners-Lee" fullname="Tim Berners-Lee">
3201<organization>MIT, Laboratory for Computer Science</organization>
3202<address>
3203<postal>
3204<street>545 Technology Square</street>
3205<city>Cambridge</city>
3206<region>MA</region>
3207<code>02139</code>
3208<country>US</country></postal>
3209<phone/>
3210<facsimile>+1 617 258 8682</facsimile>
3211<email>timbl@w3.org</email></address></author>
3212<author initials="R.T." surname="Fielding" fullname="Roy T. Fielding">
3213<organization>University of California, Irvine, Department of Information and Computer Science</organization>
3214<address>
3215<postal>
3216<street/>
3217<city>Irvine</city>
3218<region>CA</region>
3219<code>92717-3425</code>
3220<country>US</country></postal>
3221<phone/>
3222<facsimile>+1 714 824 4056</facsimile>
3223<email>fielding@ics.uci.edu</email></address></author>
3224<author initials="H.F." surname="Nielsen" fullname="Henrik Frystyk Nielsen">
3225<organization>W3 Consortium, MIT Laboratory for Computer Science</organization>
3226<address>
3227<postal>
3228<street>545 Technology Square</street>
3229<city>Cambridge</city>
3230<region>MA</region>
3231<code>02139</code>
3232<country>US</country></postal>
3233<phone/>
3234<facsimile>+1 617 258 8682</facsimile>
3235<email>frystyk@w3.org</email></address></author>
3236<date month="May" year="1996"/>
3237</front>
3238<seriesInfo name="RFC" value="1945"/>
3239</reference>
3240
3241<reference anchor="RFC2045">
3242<front>
3243<title abbrev="Internet Message Bodies">Multipurpose Internet Mail Extensions (MIME) Part One: Format of Internet Message Bodies</title>
3244<author initials="N." surname="Freed" fullname="Ned Freed">
3245<organization>Innosoft International, Inc.</organization>
3246<address>
3247<postal>
3248<street>1050 East Garvey Avenue South</street>
3249<city>West Covina</city>
3250<region>CA</region>
3251<code>91790</code>
3252<country>US</country></postal>
3253<phone>+1 818 919 3600</phone>
3254<facsimile>+1 818 919 3614</facsimile>
3255<email>ned@innosoft.com</email></address></author>
3256<author initials="N.S." surname="Borenstein" fullname="Nathaniel S. Borenstein">
3257<organization>First Virtual Holdings</organization>
3258<address>
3259<postal>
3260<street>25 Washington Avenue</street>
3261<city>Morristown</city>
3262<region>NJ</region>
3263<code>07960</code>
3264<country>US</country></postal>
3265<phone>+1 201 540 8967</phone>
3266<facsimile>+1 201 993 3032</facsimile>
3267<email>nsb@nsb.fv.com</email></address></author>
3268<date month="November" year="1996"/>
3269</front>
3270<seriesInfo name="RFC" value="2045"/>
3271</reference>
3272
3273<reference anchor="RFC1123">
3274<front>
3275<title>Requirements for Internet Hosts - Application and Support</title>
3276<author initials="R." surname="Braden" fullname="Robert Braden">
3277<organization>University of Southern California (USC), Information Sciences Institute</organization>
3278<address>
3279<postal>
3280<street>4676 Admiralty Way</street>
3281<city>Marina del Rey</city>
3282<region>CA</region>
3283<code>90292-6695</code>
3284<country>US</country></postal>
3285<phone>+1 213 822 1511</phone>
3286<email>Braden@ISI.EDU</email></address></author>
3287<date month="October" year="1989"/></front>
3288<seriesInfo name="STD" value="3"/>
3289<seriesInfo name="RFC" value="1123"/>
3290</reference>
3291
3292<reference anchor="RFC822">
3293<front>
3294<title abbrev="Standard for ARPA Internet Text Messages">Standard for the format of ARPA Internet text messages</title>
3295<author initials="D.H." surname="Crocker" fullname="David H. Crocker">
3296<organization>University of Delaware, Dept. of Electrical Engineering</organization>
3297<address>
3298<postal>
3299<street/>
3300<city>Newark</city>
3301<region>DE</region>
3302<code>19711</code>
3303<country>US</country></postal>
3304<email>DCrocker@UDel-Relay</email></address></author>
3305<date month="August" day="13" year="1982"/></front>
3306<seriesInfo name="STD" value="11"/>
3307<seriesInfo name="RFC" value="822"/>
3308</reference>
3309
3310<reference anchor="WAIS">
3311<front>
3312<title>WAIS Interface Protocol Prototype Functional Specification (v1.5)</title>
3313<author initials="F." surname="Davis" fullname="F. Davis">
3314<organization>Thinking Machines Corporation</organization>
3315</author>
3316<author initials="B." surname="Kahle" fullname="B. Kahle">
3317<organization>Thinking Machines Corporation</organization>
3318</author>
3319<author initials="H." surname="Morris" fullname="H. Morris">
3320<organization>Thinking Machines Corporation</organization>
3321</author>
3322<author initials="J." surname="Salem" fullname="J. Salem">
3323<organization>Thinking Machines Corporation</organization>
3324</author>
3325<author initials="T." surname="Shen" fullname="T. Shen">
3326<organization>Thinking Machines Corporation</organization>
3327</author>
3328<author initials="R." surname="Wang" fullname="R. Wang">
3329<organization>Thinking Machines Corporation</organization>
3330</author>
3331<author initials="J." surname="Sui" fullname="J. Sui">
3332<organization>Thinking Machines Corporation</organization>
3333</author>
3334<author initials="M." surname="Grinbaum" fullname="M. Grinbaum">
3335<organization>Thinking Machines Corporation</organization>
3336</author>
3337<date month="April" year="1990"/></front>
3338<seriesInfo name="Thinking Machines Corporation" value=""/>
3339</reference>
3340
3341<reference anchor="RFC1808">
3342<front>
3343<title>Relative Uniform Resource Locators</title>
3344<author initials="R." surname="Fielding" fullname="Roy T. Fielding">
3345<organization>University of California Irvine, Department of Information and Computer Science</organization>
3346<address>
3347<postal>
3348<street/>
3349<city>Irvine</city>
3350<region>CA</region>
3351<code>92717-3425</code>
3352<country>US</country></postal>
3353<phone>+1 714 824 4049</phone>
3354<facsimile>+1 714 824 4056</facsimile>
3355<email>fielding@ics.uci.edu</email></address></author>
3356<date month="June" year="1995"/>
3357</front>
3358<seriesInfo name="RFC" value="1808"/>
3359</reference>
3360
3361<reference anchor="RFC1036">
3362<front>
3363<title abbrev="Standard for USENET Messages">Standard for interchange of USENET messages</title>
3364<author initials="M." surname="Horton" fullname="M. Horton">
3365<organization>AT&amp;amp;T Bell Laboratories</organization></author>
3366<author initials="R." surname="Adams" fullname="R. Adams">
3367<organization>Center for Seismic Studies</organization></author>
3368<date month="December" year="1987"/></front>
3369<seriesInfo name="RFC" value="1036"/>
3370</reference>
3371
3372<reference anchor="RFC977">
3373<front>
3374<title abbrev="Network News Transfer Protocol">Network News Transfer Protocol</title>
3375<author initials="B." surname="Kantor" fullname="Brian Kantor">
3376<organization>University of California, San Diego (UCSD)</organization></author>
3377<author initials="P." surname="Lapsley" fullname="Phil Lapsley">
3378<organization>University of California, Berkeley (UCB)</organization></author>
3379<date month="February" year="1986"/></front>
3380<seriesInfo name="RFC" value="977"/>
3381</reference>
3382
3383<reference anchor="RFC2047">
3384<front>
3385<title abbrev="Message Header Extensions">MIME (Multipurpose Internet Mail Extensions) Part Three: Message Header Extensions for Non-ASCII Text</title>
3386<author initials="K." surname="Moore" fullname="Keith Moore">
3387<organization>University of Tennessee</organization>
3388<address>
3389<postal>
3390<street>107 Ayres Hall</street>
3391<street>Knoxville TN 37996-1301</street></postal>
3392<email>moore@cs.utk.edu</email></address></author>
3393<date month="November" year="1996"/>
3394<area>Applications</area>
3395<keyword>Amercian Standard Code for Information Interchange</keyword>
3396<keyword>mail</keyword>
3397<keyword>multipurpose internet mail extensions</keyword>
3398</front>
3399<seriesInfo name="RFC" value="2047"/>
3400</reference>
3401
3402<reference anchor="RFC821">
3403<front>
3404<title>Simple Mail Transfer Protocol</title>
3405<author initials="J.B." surname="Postel" fullname="Jonathan B. Postel">
3406<organization>University of Southern California (USC)/Information Sciences Institute</organization>
3407<address>
3408<postal>
3409<street>4676 Admiralty Way</street>
3410<city>Marina del Rey</city>
3411<region>CA</region>
3412<code>90291</code>
3413<country>US</country></postal>
3414<phone>+1 213 822 1511</phone></address></author>
3415<date month="August" year="1982"/></front>
3416<seriesInfo name="STD" value="10"/>
3417<seriesInfo name="RFC" value="821"/>
3418</reference>
3419
3420<reference anchor="RFC1590">
3421<front>
3422<title>Media Type Registration Procedure</title>
3423<author initials="J." surname="Postel" fullname="Jon Postel">
3424<organization>USC/Information Sciences Institute</organization>
3425<address>
3426<postal>
3427<street>4676 Admiralty Way</street>
3428<city>Marina del Rey</city>
3429<region>CA</region>
3430<code>90292</code>
3431<country>US</country></postal>
3432<phone>+1 310 822 1511</phone>
3433<facsimile>+1 310 823 6714</facsimile>
3434<email>Postel@ISI.EDU</email></address></author>
3435<date month="November" year="1996"/>
3436</front>
3437<seriesInfo name="RFC" value="1590"/>
3438</reference>
3439
3440<reference anchor="RFC959">
3441<front>
3442<title abbrev="File Transfer Protocol">File Transfer Protocol</title>
3443<author initials="J." surname="Postel" fullname="J. Postel">
3444<organization>Information Sciences Institute (ISI)</organization></author>
3445<author initials="J." surname="Reynolds" fullname="J. Reynolds">
3446<organization/></author>
3447<date month="October" year="1985"/></front>
3448<seriesInfo name="STD" value="9"/>
3449<seriesInfo name="RFC" value="959"/>
3450</reference>
3451
3452<reference anchor="RFC1700">
3453<front>
3454<title abbrev="Assigned Numbers">Assigned Numbers</title>
3455<author initials="J." surname="Reynolds" fullname="Joyce K. Reynolds">
3456<organization>USC/Information Sciences Institute</organization>
3457<address>
3458<postal>
3459<street>4676 Admiralty Way</street>
3460<city>Marina del Rey</city>
3461<region>CA</region>
3462<code>90292-6695</code>
3463<country>US</country></postal>
3464<phone>+1 310 822 1511</phone>
3465<email>jkrey@isi.edu</email></address></author>
3466<author initials="J." surname="Postel" fullname="Jon Postel">
3467<organization>USC/Information Sciences Institute</organization>
3468<address>
3469<postal>
3470<street>4676 Admiralty Way</street>
3471<city>Marina del Rey</city>
3472<region>CA</region>
3473<code>90292-6695</code>
3474<country>US</country></postal>
3475<phone>+1 310 822 1511</phone>
3476<email>postel@isi.edu</email></address></author>
3477<date month="October" year="1994"/>
3478<abstract>
3479<t/></abstract></front>
3480<seriesInfo name="STD" value="2"/>
3481<seriesInfo name="RFC" value="1700"/>
3482</reference>
3483
3484<reference anchor="RFC1737">
3485<front>
3486<title abbrev="Requirements for Uniform Resource Names">Functional Requirements for Uniform Resource Names</title>
3487<author initials="L." surname="Masinter" fullname="Larry Masinter">
3488<organization>Xerox Palo Alto Research Center</organization>
3489<address>
3490<postal>
3491<street>3333 Coyote Hill Road</street>
3492<city>Palo Alto</city>
3493<region>CA</region>
3494<code>94304</code>
3495<country>US</country></postal>
3496<phone>+1 415 812 4365</phone>
3497<facsimile>+1 415 812 4333</facsimile>
3498<email>masinter@parc.xerox.com</email></address></author>
3499<author initials="K." surname="Sollins" fullname="Karen Sollins">
3500<organization>MIT Laboratory for Computer Science</organization>
3501<address>
3502<postal>
3503<street>545 Technology Square</street>
3504<city>Cambridge</city>
3505<region>MA</region>
3506<code>02139</code>
3507<country>US</country></postal>
3508<phone>+1 617 253 2673</phone>
3509<email>sollins@lcs.mit.edu</email></address></author>
3510<date month="December" year="1994"/></front>
3511<seriesInfo name="RFC" value="1737"/>
3512</reference>
3513
3514<reference anchor="USASCII">
3515  <front>
3516    <title>Coded Character Set -- 7-bit American Standard Code for Information Interchange</title>
3517    <author>
3518      <organization>American National Standards Institute</organization>
3519    </author>
3520    <date year="1986"/>
3521  </front>
3522  <seriesInfo name="ANSI" value="X3.4"/>
3523</reference>
3524
3525<reference anchor="ISO-8859">
3526<front>
3527<title>Information technology - 8-bit single byte coded graphic - character sets</title>
3528<author>
3529<organization>International Organization for Standardization</organization>
3530</author>
3531<date year="1987-1990"/>
3532</front>
3533<annotation>
3534  Part 1: Latin alphabet No. 1, ISO-8859-1:1987.
3535  Part 2: Latin alphabet No. 2, ISO-8859-2, 1987.
3536  Part 3: Latin alphabet No. 3, ISO-8859-3, 1988.
3537  Part 4: Latin alphabet No. 4, ISO-8859-4, 1988.
3538  Part 5: Latin/Cyrillic alphabet, ISO-8859-5, 1988.
3539  Part 6: Latin/Arabic alphabet, ISO-8859-6, 1987.
3540  Part 7: Latin/Greek alphabet, ISO-8859-7, 1987.
3541  Part 8: Latin/Hebrew alphabet, ISO-8859-8, 1988.
3542  Part 9: Latin alphabet No. 5, ISO-8859-9, 1990.
3543</annotation>
3544</reference>
3545
3546<reference anchor="RFC1900">
3547<front>
3548<title>Renumbering Needs Work</title>
3549<author initials="B." surname="Carpenter" fullname="Brian E. Carpenter">
3550<organization>CERN, Computing and Networks Division</organization>
3551<address>
3552<postal>
3553<street>1211 Geneva 23</street>
3554<country>CH</country></postal>
3555<phone>+41 22 7674967</phone>
3556<facsimile>+41 22 7677155</facsimile>
3557<email>brian@dxcoms.cern.ch</email></address></author>
3558<author initials="Y." surname="Rekhter" fullname="Yakov Rekhter">
3559<organization>cisco Systems</organization>
3560<address>
3561<postal>
3562<street>170 West Tasman Drive</street>
3563<city>San Jose</city>
3564<region>CA</region>
3565<code>95134</code>
3566<country>US</country></postal>
3567<phone>+1 914 528 0090</phone>
3568<email>yakov@cisco.com</email></address></author>
3569<date month="February" year="1996"/>
3570</front>
3571<seriesInfo name="RFC" value="1900"/>
3572</reference>
3573
3574<reference anchor="Pad1995">
3575<front>
3576<title>Improving HTTP Latency</title>
3577<author initials="V.N." surname="Padmanabhan" fullname="Venkata N. Padmanabhan"><organization/></author>
3578<author initials="J.C." surname="Mogul" fullname="Jeffrey C. Mogul"><organization/></author>
3579<date year="1995" month="Dec"/>
3580</front>
3581<seriesInfo name="Computer Networks and ISDN Systems" value="v. 28, pp. 25-35"/>
3582<annotation>
3583  Slightly revised version of paper in Proc. 2nd International WWW Conference '94: Mosaic and the Web, Oct. 1994,
3584  which is available at <eref target="http://www.ncsa.uiuc.edu/SDG/IT94/Proceedings/DDay/mogul/HTTPLatency.html"/>.
3585</annotation>
3586</reference>
3587
3588<reference anchor="Tou1998" target="http://www.isi.edu/touch/pubs/http-perf96/">
3589<front>
3590<title>Analysis of HTTP Performance</title>
3591<author initials="J." surname="Touch" fullname="Joe Touch">
3592  <organization>USC/Information Sciences Institute</organization>
3593  <address><email>touch@isi.edu</email></address>
3594</author>
3595<author initials="J." surname="Heidemann" fullname="John Heidemann">
3596  <organization>USC/Information Sciences Institute</organization>
3597  <address><email>johnh@isi.edu</email></address>
3598</author>
3599<author initials="K." surname="Obraczka" fullname="Katia Obraczka">
3600  <organization>USC/Information Sciences Institute</organization>
3601  <address><email>katia@isi.edu</email></address>
3602</author>
3603<date year="1998" month="Aug"/>
3604</front>
3605<seriesInfo name="ISI Research Report" value="ISI/RR-98-463 (original report dated Aug.1996)"/>
3606</reference>
3607
3608<reference anchor="RFC1305">
3609<front>
3610<title>Network Time Protocol (Version 3) Specification, Implementation</title>
3611<author initials="D." surname="Mills" fullname="David L. Mills">
3612<organization>University of Delaware, Electrical Engineering Department</organization>
3613<address>
3614<postal>
3615<street/>
3616<city>Newark</city>
3617<region>DE</region>
3618<code>19716</code>
3619<country>US</country></postal>
3620<phone>+1 302 451 8247</phone>
3621<email>mills@udel.edu</email></address></author>
3622<date month="March" year="1992"/>
3623<abstract>
3624<t>This document describes the Network Time Protocol (NTP), specifies its normal structure and summarizes information useful for its implementation. NTP provides the mechanisms to synchronize time and coordinate time distribution in a large, diverse internet operating at rates from mundane to lightwave. It uses a returnable-time design in which a distributed subnet of time servers operating in a self-organizing, hierarchical-master-slave configuration synchronizes local clocks within the subnet and to national time standards via wire or radio. The servers can also redistribute reference time via local routing algorithms and time daemons.</t></abstract></front>
3625<seriesInfo name="RFC" value="1305"/>
3626</reference>
3627
3628<reference anchor="Spe" target="http://sunsite.unc.edu/mdma-release/http-prob.html">
3629<front>
3630<title>Analysis of HTTP Performance Problems</title>
3631<author initials="S." surname="Spero" fullname="Simon E. Spero">
3632  <organization/>
3633</author>
3634<date/>
3635</front>
3636</reference>
3637
3638<reference anchor="RFC2068">
3639<front>
3640<title abbrev="HTTP/1.1">Hypertext Transfer Protocol -- HTTP/1.1</title>
3641<author initials="R." surname="Fielding" fullname="Roy T. Fielding">
3642<organization>University of California, Irvine, Department of Information and Computer Science</organization>
3643<address>
3644<postal>
3645<street/>
3646<city>Irvine</city>
3647<region>CA</region>
3648<code>92717-3425</code>
3649<country>US</country></postal>
3650<facsimile>+1 714 824 4056</facsimile>
3651<email>fielding@ics.uci.edu</email></address></author>
3652<author initials="J." surname="Gettys" fullname="Jim Gettys">
3653<organization>MIT Laboratory for Computer Science</organization>
3654<address>
3655<postal>
3656<street>545 Technology Square</street>
3657<city>Cambridge</city>
3658<region>MA</region>
3659<code>02139</code>
3660<country>US</country></postal>
3661<facsimile>+1 617 258 8682</facsimile>
3662<email>jg@w3.org</email></address></author>
3663<author initials="J." surname="Mogul" fullname="Jeffrey C. Mogul">
3664<organization>Digital Equipment Corporation, Western Research Laboratory</organization>
3665<address>
3666<postal>
3667<street>250 University Avenue</street>
3668<city>Palo Alto</city>
3669<region>CA</region>
3670<code>94301</code>
3671<country>US</country></postal>
3672<email>mogul@wrl.dec.com</email></address></author>
3673<author initials="H." surname="Nielsen" fullname="Henrik Frystyk Nielsen">
3674<organization>MIT Laboratory for Computer Science</organization>
3675<address>
3676<postal>
3677<street>545 Technology Square</street>
3678<city>Cambridge</city>
3679<region>MA</region>
3680<code>02139</code>
3681<country>US</country></postal>
3682<facsimile>+1 617 258 8682</facsimile>
3683<email>frystyk@w3.org</email></address></author>
3684<author initials="T." surname="Berners-Lee" fullname="Tim Berners-Lee">
3685<organization>MIT Laboratory for Computer Science</organization>
3686<address>
3687<postal>
3688<street>545 Technology Square</street>
3689<city>Cambridge</city>
3690<region>MA</region>
3691<code>02139</code>
3692<country>US</country></postal>
3693<facsimile>+1 617 258 8682</facsimile>
3694<email>timbl@w3.org</email></address></author>
3695<date month="January" year="1997"/>
3696<abstract>
3697<t>The Hypertext Transfer Protocol (HTTP) is an application-level protocol for distributed, collaborative, hypermedia information systems. It is a generic, stateless, object-oriented protocol which can be used for many tasks, such as name servers and distributed object management systems, through extension of its request methods. A feature of HTTP is the typing and negotiation of data representation, allowing systems to be built independently of the data being transferred.</t>
3698<t>HTTP has been in use by the World-Wide Web global information initiative since 1990. This specification defines the protocol referred to as "HTTP/1.1".</t></abstract></front>
3699<seriesInfo name="RFC" value="2068"/>
3700</reference>
3701
3702<reference anchor="RFC2119">
3703<front>
3704<title abbrev="RFC Key Words">Key words for use in RFCs to Indicate Requirement Levels</title>
3705<author initials="S." surname="Bradner" fullname="Scott Bradner">
3706<organization>Harvard University</organization>
3707<address>
3708<postal>
3709<street>1350 Mass. Ave.</street>
3710<street>Cambridge</street>
3711<street>MA 02138</street></postal>
3712<phone>- +1 617 495 3864</phone>
3713<email>-</email></address></author>
3714<date month="March" year="1997"/>
3715<area>General</area>
3716<keyword>keyword</keyword>
3717</front>
3718
3719<seriesInfo name="BCP" value="14"/>
3720<seriesInfo name="RFC" value="2119"/>
3721</reference>
3722
3723<reference anchor="RFC2145">
3724<front>
3725<title abbrev="HTTP Version Numbers">Use and Interpretation of HTTP Version Numbers</title>
3726<author initials="J.C." surname="Mogul" fullname="Jeffrey C. Mogul">
3727<organization>Western Research Laboratory</organization>
3728<address>
3729<postal>
3730<street>Digital Equipment Corporation</street>
3731<street>250 University Avenue</street>
3732<street>Palo Alto</street>
3733<street>California</street>
3734<street>94305</street>
3735<country>USA</country></postal>
3736<email>mogul@wrl.dec.com</email></address></author>
3737<author initials="R.T." surname="Fielding" fullname="Roy T. Fielding">
3738<organization>Department of Information and Computer Science</organization>
3739<address>
3740<postal>
3741<street>University of California</street>
3742<street>Irvine</street>
3743<street>CA 92717-3425</street>
3744<country>USA</country></postal>
3745<facsimile>+1 (714) 824-4056</facsimile>
3746<email>fielding@ics.uci.edu</email></address></author>
3747<author initials="J." surname="Gettys" fullname="Jim Gettys">
3748<organization>MIT Laboratory for Computer Science</organization>
3749<address>
3750<postal>
3751<street>545 Technology Square</street>
3752<street>Cambridge</street>
3753<street>MA 02139</street>
3754<country>USA</country></postal>
3755<facsimile>+1 (617) 258 8682</facsimile>
3756<email>jg@w3.org</email></address></author>
3757<author initials="H.F." surname="Nielsen" fullname="Henrik Frystyk Nielsen">
3758<organization>W3 Consortium</organization>
3759<address>
3760<postal>
3761<street>MIT Laboratory for Computer Science</street>
3762<street>545 Technology Square</street>
3763<street>Cambridge</street>
3764<street>MA 02139</street>
3765<country>USA</country></postal>
3766<facsimile>+1 (617) 258 8682</facsimile>
3767<email>frystyk@w3.org</email></address></author>
3768<date month="May" year="1997"/>
3769<area>Applications</area>
3770<keyword>HTTP</keyword>
3771<keyword>hypertext transfer protocol</keyword>
3772<abstract>
3773<t>
3774   HTTP request and response messages include an HTTP protocol version
3775   number.  Some confusion exists concerning the proper use and
3776   interpretation of HTTP version numbers, and concerning
3777   interoperability of HTTP implementations of different protocol
3778   versions.  This document is an attempt to clarify the situation.  It
3779   is not a modification of the intended meaning of the existing
3780   HTTP/1.0 and HTTP/1.1 documents, but it does describe the intention
3781   of the authors of those documents, and can be considered definitive
3782   when there is any ambiguity in those documents concerning HTTP
3783   version numbers, for all versions of HTTP.
3784</t></abstract></front>
3785<seriesInfo name="RFC" value="2145"/>
3786</reference>
3787
3788<reference anchor="Nie1997">
3789<front>
3790<title>Network Performance Effects of HTTP/1.1, CSS1, and PNG</title>
3791<author initials="H.F.." surname="Nielsen" fullname="H.F. Nielsen">
3792  <organization/>
3793</author>
3794<author initials="J." surname="Gettys" fullname="J. Gettys">
3795  <organization/>
3796</author>
3797<author initials="E." surname="Prud'hommeaux" fullname="E. Prud'hommeaux">
3798  <organization/>
3799</author>
3800<author initials="H." surname="Lie" fullname="H. Lie">
3801  <organization/>
3802</author>
3803<author initials="C." surname="Lilley" fullname="C. Lilley">
3804  <organization/>
3805</author>
3806<date year="1997" month="Sep"/>
3807</front>
3808<seriesInfo name="Proceedings of ACM SIGCOMM '97, Cannes France" value=""/>
3809</reference>
3810
3811<reference anchor="RFC2396">
3812<front>
3813<title abbrev="URI Generic Syntax">Uniform Resource Identifiers (URI): Generic Syntax</title>
3814<author initials="T." surname="Berners-Lee" fullname="Tim Berners-Lee">
3815<organization abbrev="MIT/LCS">World Wide Web Consortium</organization>
3816<address>
3817<postal>
3818<street>MIT Laboratory for Computer Science, NE43-356</street>
3819<street>545 Technology Square</street>
3820<city>Cambridge</city>
3821<region>MA</region>
3822<code>02139</code></postal>
3823<facsimile>+1(617)258-8682</facsimile>
3824<email>timbl@w3.org</email></address></author>
3825<author initials="R.T." surname="Fielding" fullname="Roy T. Fielding">
3826<organization abbrev="U.C. Irvine">Department of Information and Computer Science</organization>
3827<address>
3828<postal>
3829<street>University of California, Irvine</street>
3830<city>Irvine</city>
3831<region>CA</region>
3832<code>92697-3425</code></postal>
3833<facsimile>+1(949)824-1715</facsimile>
3834<email>fielding@ics.uci.edu</email></address></author>
3835<author initials="L." surname="Masinter" fullname="Larry Masinter">
3836<organization abbrev="Xerox Corporation">Xerox PARC</organization>
3837<address>
3838<postal>
3839<street>3333 Coyote Hill Road</street>
3840<city>Palo Alto</city>
3841<region>CA</region>
3842<code>94034</code></postal>
3843<facsimile>+1(415)812-4333</facsimile>
3844<email>masinter@parc.xerox.com</email></address></author>
3845<date month="August" year="1998"/>
3846<area>Applications</area>
3847<keyword>uniform resource</keyword>
3848<keyword>URI</keyword>
3849</front>
3850<seriesInfo name="RFC" value="2396"/>
3851</reference>
3852
3853<reference anchor="RFC2616">
3854  <front>
3855    <title>Hypertext Transfer Protocol -- HTTP/1.1</title>
3856    <author initials="R." surname="Fielding" fullname="R. Fielding">
3857      <organization>University of California, Irvine</organization>
3858      <address><email>fielding@ics.uci.edu</email></address>
3859    </author>
3860    <author initials="J." surname="Gettys" fullname="J. Gettys">
3861      <organization>W3C</organization>
3862      <address><email>jg@w3.org</email></address>
3863    </author>
3864    <author initials="J." surname="Mogul" fullname="J. Mogul">
3865      <organization>Compaq Computer Corporation</organization>
3866      <address><email>mogul@wrl.dec.com</email></address>
3867    </author>
3868    <author initials="H." surname="Frystyk" fullname="H. Frystyk">
3869      <organization>MIT Laboratory for Computer Science</organization>
3870      <address><email>frystyk@w3.org</email></address>
3871    </author>
3872    <author initials="L." surname="Masinter" fullname="L. Masinter">
3873      <organization>Xerox Corporation</organization>
3874      <address><email>masinter@parc.xerox.com</email></address>
3875    </author>
3876    <author initials="P." surname="Leach" fullname="P. Leach">
3877      <organization>Microsoft Corporation</organization>
3878      <address><email>paulle@microsoft.com</email></address>
3879    </author>
3880    <author initials="T." surname="Berners-Lee" fullname="T. Berners-Lee">
3881      <organization>W3C</organization>
3882      <address><email>timbl@w3.org</email></address>
3883    </author>
3884    <date month="June" year="1999"/>
3885  </front>
3886  <seriesInfo name="RFC" value="2616"/>
3887</reference>
3888
3889<reference anchor="RFC2617">
3890<front>
3891<title abbrev="HTTP Authentication">HTTP Authentication: Basic and Digest Access Authentication</title>
3892<author initials="J." surname="Franks" fullname="John Franks">
3893<organization>Northwestern University, Department of Mathematics</organization>
3894<address>
3895<postal>
3896<street/>
3897<city>Evanston</city>
3898<region>IL</region>
3899<code>60208-2730</code>
3900<country>US</country></postal>
3901<email>john@math.nwu.edu</email></address></author>
3902<author initials="P.M." surname="Hallam-Baker" fullname="Phillip M. Hallam-Baker">
3903<organization>Verisign Inc.</organization>
3904<address>
3905<postal>
3906<street>301 Edgewater Place</street>
3907<street>Suite 210</street>
3908<city>Wakefield</city>
3909<region>MA</region>
3910<code>01880</code>
3911<country>US</country></postal>
3912<email>pbaker@verisign.com</email></address></author>
3913<author initials="J.L." surname="Hostetler" fullname="Jeffery L. Hostetler">
3914<organization>AbiSource, Inc.</organization>
3915<address>
3916<postal>
3917<street>6 Dunlap Court</street>
3918<city>Savoy</city>
3919<region>IL</region>
3920<code>61874</code>
3921<country>US</country></postal>
3922<email>jeff@AbiSource.com</email></address></author>
3923<author initials="S.D." surname="Lawrence" fullname="Scott D. Lawrence">
3924<organization>Agranat Systems, Inc.</organization>
3925<address>
3926<postal>
3927<street>5 Clocktower Place</street>
3928<street>Suite 400</street>
3929<city>Maynard</city>
3930<region>MA</region>
3931<code>01754</code>
3932<country>US</country></postal>
3933<email>lawrence@agranat.com</email></address></author>
3934<author initials="P.J." surname="Leach" fullname="Paul J. Leach">
3935<organization>Microsoft Corporation</organization>
3936<address>
3937<postal>
3938<street>1 Microsoft Way</street>
3939<city>Redmond</city>
3940<region>WA</region>
3941<code>98052</code>
3942<country>US</country></postal>
3943<email>paulle@microsoft.com</email></address></author>
3944<author initials="A." surname="Luotonen" fullname="Ari Luotonen">
3945<organization>Netscape Communications Corporation</organization>
3946<address>
3947<postal>
3948<street>501 East Middlefield Road</street>
3949<city>Mountain View</city>
3950<region>CA</region>
3951<code>94043</code>
3952<country>US</country></postal></address></author>
3953<author initials="L." surname="Stewart" fullname="Lawrence C. Stewart">
3954<organization>Open Market, Inc.</organization>
3955<address>
3956<postal>
3957<street>215 First Street</street>
3958<city>Cambridge</city>
3959<region>MA</region>
3960<code>02142</code>
3961<country>US</country></postal>
3962<email>stewart@OpenMarket.com</email></address></author>
3963<date month="June" year="1999"/>
3964</front>
3965<seriesInfo name="RFC" value="2617"/>
3966</reference>
3967
3968<reference anchor="RFC2324">
3969<front>
3970<title abbrev="HTCPCP/1.0">Hyper Text Coffee Pot Control Protocol (HTCPCP/1.0)</title>
3971<author initials="L." surname="Masinter" fullname="Larry Masinter">
3972<organization>Xerox Palo Alto Research Center</organization>
3973<address>
3974<postal>
3975<street>3333 Coyote Hill Road</street>
3976<city>Palo Alto</city>
3977<region>CA</region>
3978<code>94304</code></postal>
3979<email>masinter@parc.xerox.com</email></address></author>
3980<date month="April" day="1" year="1998"/>
3981<area>General</area>
3982<keyword>control protocol</keyword>
3983<keyword>coffee</keyword>
3984</front>
3985<seriesInfo name="RFC" value="2324"/>
3986</reference>
3987
3988</references>
3989
3990<section title="Internet Media Type message/http and application/http" anchor="internet.media.type.http">
3991<iref item="Media Type" subitem="message/http" primary="true"/>
3992<iref item="message/http Media Type" primary="true"/>
3993<iref item="Media Type" subitem="application/http" primary="true"/>
3994<iref item="application/http Media Type" primary="true"/>
3995<t>
3996   In addition to defining the HTTP/1.1 protocol, this document serves
3997   as the specification for the Internet media type "message/http" and
3998   "application/http". The message/http type can be used to enclose a
3999   single HTTP request or response message, provided that it obeys the
4000   MIME restrictions for all "message" types regarding line length and
4001   encodings. The application/http type can be used to enclose a
4002   pipeline of one or more HTTP request or response messages (not
4003   intermixed). The following is to be registered with IANA <xref target="RFC1590"/>.
4004</t>
4005<t>
4006  <list style="hanging" x:indent="12em">
4007    <t hangText="Media Type name:">
4008      message
4009    </t>
4010    <t hangText="Media subtype name:">
4011      http
4012    </t>
4013    <t hangText="Required parameters:">
4014      none
4015    </t>
4016    <t hangText="Optional parameters:">
4017      version, msgtype
4018      <list style="hanging">
4019        <t hangText="version:">
4020          The HTTP-Version number of the enclosed message
4021          (e.g., "1.1"). If not present, the version can be
4022          determined from the first line of the body.
4023        </t>
4024        <t hangText="msgtype:">
4025          The message type -- "request" or "response". If not
4026          present, the type can be determined from the first
4027          line of the body.
4028        </t>
4029      </list>
4030    </t>
4031    <t hangText="Encoding considerations:">
4032      only "7bit", "8bit", or "binary" are permitted
4033    </t>
4034    <t hangText="Security considerations:">
4035      none
4036    </t>
4037  </list>
4038</t>
4039<t>
4040  <list style="hanging" x:indent="12em">
4041    <t hangText="Media Type name:">
4042      application
4043    </t>
4044    <t hangText="Media subtype name:">
4045      http
4046    </t>
4047    <t hangText="Required parameters:">
4048      none
4049    </t>
4050    <t hangText="Optional parameters:">
4051      version, msgtype
4052      <list style="hanging">
4053        <t hangText="version:">
4054          The HTTP-Version number of the enclosed messages
4055          (e.g., "1.1"). If not present, the version can be
4056          determined from the first line of the body.
4057        </t>
4058        <t hangText="msgtype:">
4059          The message type -- "request" or "response". If not
4060          present, the type can be determined from the first
4061          line of the body.
4062        </t>
4063      </list>
4064    </t>
4065    <t hangText="Encoding considerations:">
4066      HTTP messages enclosed by this type
4067      are in "binary" format; use of an appropriate
4068      Content-Transfer-Encoding is required when
4069      transmitted via E-mail.
4070    </t>
4071    <t hangText="Security considerations:">
4072      none
4073    </t>
4074  </list>
4075</t>
4076</section>
4077
4078<section title="Tolerant Applications" anchor="tolerant.applications">
4079<t>
4080   Although this document specifies the requirements for the generation
4081   of HTTP/1.1 messages, not all applications will be correct in their
4082   implementation. We therefore recommend that operational applications
4083   be tolerant of deviations whenever those deviations can be
4084   interpreted unambiguously.
4085</t>
4086<t>
4087   Clients &SHOULD; be tolerant in parsing the Status-Line and servers
4088   tolerant when parsing the Request-Line. In particular, they &SHOULD;
4089   accept any amount of SP or HT characters between fields, even though
4090   only a single SP is required.
4091</t>
4092<t>
4093   The line terminator for message-header fields is the sequence CRLF.
4094   However, we recommend that applications, when parsing such headers,
4095   recognize a single LF as a line terminator and ignore the leading CR.
4096</t>
4097<t>
4098   The character set of an entity-body &SHOULD; be labeled as the lowest
4099   common denominator of the character codes used within that body, with
4100   the exception that not labeling the entity is preferred over labeling
4101   the entity with the labels US-ASCII or ISO-8859-1. See &payload;.
4102</t>
4103<t>
4104   Additional rules for requirements on parsing and encoding of dates
4105   and other potential problems with date encodings include:
4106</t>
4107<t>
4108  <list style="symbols">
4109     <t>HTTP/1.1 clients and caches &SHOULD; assume that an RFC-850 date
4110        which appears to be more than 50 years in the future is in fact
4111        in the past (this helps solve the "year 2000" problem).</t>
4112
4113     <t>An HTTP/1.1 implementation &MAY; internally represent a parsed
4114        Expires date as earlier than the proper value, but &MUST-NOT;
4115        internally represent a parsed Expires date as later than the
4116        proper value.</t>
4117
4118     <t>All expiration-related calculations &MUST; be done in GMT. The
4119        local time zone &MUST-NOT; influence the calculation or comparison
4120        of an age or expiration time.</t>
4121
4122     <t>If an HTTP header incorrectly carries a date value with a time
4123        zone other than GMT, it &MUST; be converted into GMT using the
4124        most conservative possible conversion.</t>
4125  </list>
4126</t>
4127</section>
4128
4129<section title="Conversion of Date Formats" anchor="conversion.of.date.formats">
4130<t>
4131   HTTP/1.1 uses a restricted set of date formats (<xref target="full.date"/>) to
4132   simplify the process of date comparison. Proxies and gateways from
4133   other protocols &SHOULD; ensure that any Date header field present in a
4134   message conforms to one of the HTTP/1.1 formats and rewrite the date
4135   if necessary.
4136</t>
4137</section>
4138
4139<section title="Compatibility with Previous Versions" anchor="compatibility">
4140<t>
4141   It is beyond the scope of a protocol specification to mandate
4142   compliance with previous versions. HTTP/1.1 was deliberately
4143   designed, however, to make supporting previous versions easy. It is
4144   worth noting that, at the time of composing this specification
4145   (1996), we would expect commercial HTTP/1.1 servers to:
4146  <list style="symbols">
4147     <t>recognize the format of the Request-Line for HTTP/0.9, 1.0, and
4148        1.1 requests;</t>
4149
4150     <t>understand any valid request in the format of HTTP/0.9, 1.0, or
4151        1.1;</t>
4152
4153     <t>respond appropriately with a message in the same major version
4154        used by the client.</t>
4155  </list>
4156</t>
4157<t>
4158   And we would expect HTTP/1.1 clients to:
4159  <list style="symbols">
4160     <t>recognize the format of the Status-Line for HTTP/1.0 and 1.1
4161        responses;</t>
4162
4163     <t>understand any valid response in the format of HTTP/0.9, 1.0, or
4164        1.1.</t>
4165  </list>
4166</t>
4167<t>
4168   For most implementations of HTTP/1.0, each connection is established
4169   by the client prior to the request and closed by the server after
4170   sending the response. Some implementations implement the Keep-Alive
4171   version of persistent connections described in <xref x:sec="19.7.1" x:fmt="sec" target="RFC2068"/> of RFC
4172   2068 <xref target="RFC2068"/>.
4173</t>
4174
4175<section title="Changes from HTTP/1.0" anchor="changes.from.1.0">
4176<t>
4177   This section summarizes major differences between versions HTTP/1.0
4178   and HTTP/1.1.
4179</t>
4180
4181<section title="Changes to Simplify Multi-homed Web Servers and Conserve IP Addresses" anchor="changes.to.simplify.multi-homed.web.servers.and.conserve.ip.addresses">
4182<t>
4183   The requirements that clients and servers support the Host request-header,
4184   report an error if the Host request-header (<xref target="header.host"/>) is
4185   missing from an HTTP/1.1 request, and accept absolute URIs (<xref target="request-uri"/>)
4186   are among the most important changes defined by this
4187   specification.
4188</t>
4189<t>
4190   Older HTTP/1.0 clients assumed a one-to-one relationship of IP
4191   addresses and servers; there was no other established mechanism for
4192   distinguishing the intended server of a request than the IP address
4193   to which that request was directed. The changes outlined above will
4194   allow the Internet, once older HTTP clients are no longer common, to
4195   support multiple Web sites from a single IP address, greatly
4196   simplifying large operational Web servers, where allocation of many
4197   IP addresses to a single host has created serious problems. The
4198   Internet will also be able to recover the IP addresses that have been
4199   allocated for the sole purpose of allowing special-purpose domain
4200   names to be used in root-level HTTP URLs. Given the rate of growth of
4201   the Web, and the number of servers already deployed, it is extremely
4202   important that all implementations of HTTP (including updates to
4203   existing HTTP/1.0 applications) correctly implement these
4204   requirements:
4205  <list style="symbols">
4206     <t>Both clients and servers &MUST; support the Host request-header.</t>
4207
4208     <t>A client that sends an HTTP/1.1 request &MUST; send a Host header.</t>
4209
4210     <t>Servers &MUST; report a 400 (Bad Request) error if an HTTP/1.1
4211        request does not include a Host request-header.</t>
4212
4213     <t>Servers &MUST; accept absolute URIs.</t>
4214  </list>
4215</t>
4216</section>
4217</section>
4218
4219<section title="Compatibility with HTTP/1.0 Persistent Connections" anchor="compatibility.with.http.1.0.persistent.connections">
4220<t>
4221   Some clients and servers might wish to be compatible with some
4222   previous implementations of persistent connections in HTTP/1.0
4223   clients and servers. Persistent connections in HTTP/1.0 are
4224   explicitly negotiated as they are not the default behavior. HTTP/1.0
4225   experimental implementations of persistent connections are faulty,
4226   and the new facilities in HTTP/1.1 are designed to rectify these
4227   problems. The problem was that some existing 1.0 clients may be
4228   sending Keep-Alive to a proxy server that doesn't understand
4229   Connection, which would then erroneously forward it to the next
4230   inbound server, which would establish the Keep-Alive connection and
4231   result in a hung HTTP/1.0 proxy waiting for the close on the
4232   response. The result is that HTTP/1.0 clients must be prevented from
4233   using Keep-Alive when talking to proxies.
4234</t>
4235<t>
4236   However, talking to proxies is the most important use of persistent
4237   connections, so that prohibition is clearly unacceptable. Therefore,
4238   we need some other mechanism for indicating a persistent connection
4239   is desired, which is safe to use even when talking to an old proxy
4240   that ignores Connection. Persistent connections are the default for
4241   HTTP/1.1 messages; we introduce a new keyword (Connection: close) for
4242   declaring non-persistence. See <xref target="header.connection"/>.
4243</t>
4244<t>
4245   The original HTTP/1.0 form of persistent connections (the Connection:
4246   Keep-Alive and Keep-Alive header) is documented in RFC 2068. <xref target="RFC2068"/>
4247</t>
4248</section>
4249
4250<section title="Changes from RFC 2068" anchor="changes.from.rfc.2068">
4251<t>
4252   This specification has been carefully audited to correct and
4253   disambiguate key word usage; RFC 2068 had many problems in respect to
4254   the conventions laid out in RFC 2119 <xref target="RFC2119"/>.
4255</t>
4256<t>
4257   Transfer-coding and message lengths all interact in ways that
4258   required fixing exactly when chunked encoding is used (to allow for
4259   transfer encoding that may not be self delimiting); it was important
4260   to straighten out exactly how message lengths are computed.
4261</t>
4262<t>
4263   The use and interpretation of HTTP version numbers has been clarified
4264   by RFC 2145. Require proxies to upgrade requests to highest protocol
4265   version they support to deal with problems discovered in HTTP/1.0
4266   implementations (<xref target="http.version"/>)
4267</t>
4268<t>
4269   Proxies should be able to add Content-Length when appropriate.
4270</t>
4271<t>
4272   Transfer-coding had significant problems, particularly with
4273   interactions with chunked encoding. The solution is that transfer-codings
4274   become as full fledged as content-codings. This involves
4275   adding an IANA registry for transfer-codings (separate from content
4276   codings), a new header field (TE) and enabling trailer headers in the
4277   future. Transfer encoding is a major performance benefit, so it was
4278   worth fixing <xref target="Nie1997"/>. TE also solves another, obscure, downward
4279   interoperability problem that could have occurred due to interactions
4280   between authentication trailers, chunked encoding and HTTP/1.0
4281   clients.(Section <xref target="transfer.codings" format="counter"/>, <xref target="chunked.transfer.encoding" format="counter"/>,
4282   and <xref target="header.te" format="counter"/>)
4283</t>
4284</section>
4285</section>
4286</back>
4287</rfc>
Note: See TracBrowser for help on using the repository browser.