source: draft-ietf-httpbis/latest/p1-messaging.xml @ 104

Last change on this file since 104 was 104, checked in by julian.reschke@…, 12 years ago

Note changes for #16 in Changes section.

  • Property svn:eol-style set to native
File size: 170.0 KB
Line 
1<?xml version="1.0" encoding="utf-8"?>
2<?xml-stylesheet type='text/xsl' href='../myxml2rfc.xslt'?>
3<!DOCTYPE rfc [
4  <!ENTITY MAY "<bcp14 xmlns='http://purl.org/net/xml2rfc/ext'>MAY</bcp14>">
5  <!ENTITY MUST "<bcp14 xmlns='http://purl.org/net/xml2rfc/ext'>MUST</bcp14>">
6  <!ENTITY MUST-NOT "<bcp14 xmlns='http://purl.org/net/xml2rfc/ext'>MUST NOT</bcp14>">
7  <!ENTITY OPTIONAL "<bcp14 xmlns='http://purl.org/net/xml2rfc/ext'>OPTIONAL</bcp14>">
8  <!ENTITY RECOMMENDED "<bcp14 xmlns='http://purl.org/net/xml2rfc/ext'>RECOMMENDED</bcp14>">
9  <!ENTITY REQUIRED "<bcp14 xmlns='http://purl.org/net/xml2rfc/ext'>REQUIRED</bcp14>">
10  <!ENTITY SHALL "<bcp14 xmlns='http://purl.org/net/xml2rfc/ext'>SHALL</bcp14>">
11  <!ENTITY SHALL-NOT "<bcp14 xmlns='http://purl.org/net/xml2rfc/ext'>SHALL NOT</bcp14>">
12  <!ENTITY SHOULD "<bcp14 xmlns='http://purl.org/net/xml2rfc/ext'>SHOULD</bcp14>">
13  <!ENTITY SHOULD-NOT "<bcp14 xmlns='http://purl.org/net/xml2rfc/ext'>SHOULD NOT</bcp14>">
14  <!ENTITY ID-VERSION "latest">
15  <!ENTITY ID-MONTH "December">
16  <!ENTITY ID-YEAR "2007">
17  <!ENTITY caching                "<xref target='Part6' x:rel='#caching' xmlns:x='http://purl.org/net/xml2rfc/ext'/>">
18  <!ENTITY payload                "<xref target='Part3' xmlns:x='http://purl.org/net/xml2rfc/ext'/>">
19  <!ENTITY CONNECT                "<xref target='Part2' x:rel='#CONNECT' xmlns:x='http://purl.org/net/xml2rfc/ext'/>">
20  <!ENTITY content.negotiation    "<xref target='Part3' x:rel='#content.negotiation' xmlns:x='http://purl.org/net/xml2rfc/ext'/>">
21  <!ENTITY diff2045entity         "<xref target='Part3' x:rel='#differences.between.http.entities.and.rfc.2045.entities' xmlns:x='http://purl.org/net/xml2rfc/ext'/>">
22  <!ENTITY entity                 "<xref target='Part3' x:rel='#entity' xmlns:x='http://purl.org/net/xml2rfc/ext'/>">
23  <!ENTITY entity-header-fields   "<xref target='Part3' x:rel='#entity.header.fields' xmlns:x='http://purl.org/net/xml2rfc/ext'/>">
24  <!ENTITY header-cache-control   "<xref target='Part6' x:rel='#header.cache-control' xmlns:x='http://purl.org/net/xml2rfc/ext'/>">
25  <!ENTITY header-expect          "<xref target='Part2' x:rel='#header.expect' xmlns:x='http://purl.org/net/xml2rfc/ext'/>">
26  <!ENTITY header-pragma          "<xref target='Part6' x:rel='#header.pragma' xmlns:x='http://purl.org/net/xml2rfc/ext'/>">
27  <!ENTITY header-warning         "<xref target='Part6' x:rel='#header.warning' xmlns:x='http://purl.org/net/xml2rfc/ext'/>">
28  <!ENTITY idempotent-methods     "<xref target='Part2' x:rel='#idempotent.methods' xmlns:x='http://purl.org/net/xml2rfc/ext'/>">
29  <!ENTITY qvalue                 "<xref target='Part3' x:rel='#quality.values' xmlns:x='http://purl.org/net/xml2rfc/ext'/>">
30  <!ENTITY request-header-fields  "<xref target='Part2' x:rel='#request.header.fields' xmlns:x='http://purl.org/net/xml2rfc/ext'/>">
31  <!ENTITY response-header-fields "<xref target='Part2' x:rel='#response.header.fields' xmlns:x='http://purl.org/net/xml2rfc/ext'/>">
32  <!ENTITY method                 "<xref target='Part2' x:rel='#method' xmlns:x='http://purl.org/net/xml2rfc/ext'/>">
33  <!ENTITY status-codes           "<xref target='Part2' x:rel='#status.codes' xmlns:x='http://purl.org/net/xml2rfc/ext'/>">
34  <!ENTITY status-100             "<xref target='Part2' x:rel='#status.100' xmlns:x='http://purl.org/net/xml2rfc/ext'/>">
35  <!ENTITY status-1xx             "<xref target='Part2' x:rel='#status.1xx' xmlns:x='http://purl.org/net/xml2rfc/ext'/>">
36  <!ENTITY status-414             "<xref target='Part2' x:rel='#status.414' xmlns:x='http://purl.org/net/xml2rfc/ext'/>">
37]>
38<?rfc toc="yes" ?>
39<?rfc symrefs="yes" ?>
40<?rfc sortrefs="yes" ?>
41<?rfc compact="yes"?>
42<?rfc subcompact="no" ?>
43<?rfc linkmailto="no" ?>
44<?rfc editing="no" ?>
45<?rfc-ext allow-markup-in-artwork="yes" ?>
46<?rfc-ext include-references-in-index="yes" ?>
47<rfc obsoletes="2068, 2616" category="std"
48     ipr="full3978" docName="draft-ietf-httpbis-p1-messaging-&ID-VERSION;"
49     xmlns:x='http://purl.org/net/xml2rfc/ext' xmlns:ed="http://greenbytes.de/2002/rfcedit">
50<front>
51
52  <title abbrev="HTTP/1.1">HTTP/1.1, part 1: URIs, Connections, and Message Parsing</title>
53
54  <author initials="R." surname="Fielding" fullname="Roy T. Fielding" role="editor">
55    <organization abbrev="Day Software">Day Software</organization>
56    <address>
57      <postal>
58        <street>23 Corporate Plaza DR, Suite 280</street>
59        <city>Newport Beach</city>
60        <region>CA</region>
61        <code>92660</code>
62        <country>USA</country>
63      </postal>
64      <phone>+1-949-706-5300</phone>
65      <facsimile>+1-949-706-5305</facsimile>
66      <email>fielding@gbiv.com</email>
67      <uri>http://roy.gbiv.com/</uri>
68    </address>
69  </author>
70
71  <author initials="J." surname="Gettys" fullname="Jim Gettys">
72    <organization>One Laptop per Child</organization>
73    <address>
74      <postal>
75        <street>21 Oak Knoll Road</street>
76        <city>Carlisle</city>
77        <region>MA</region>
78        <code>01741</code>
79        <country>USA</country>
80      </postal>
81      <email>jg@laptop.org</email>
82      <uri>http://www.laptop.org/</uri>
83    </address>
84  </author>
85 
86  <author initials="J." surname="Mogul" fullname="Jeffrey C. Mogul">
87    <organization abbrev="HP">Hewlett-Packard Company</organization>
88    <address>
89      <postal>
90        <street>HP Labs, Large Scale Systems Group</street>
91        <street>1501 Page Mill Road, MS 1177</street>
92        <city>Palo Alto</city>
93        <region>CA</region>
94        <code>94304</code>
95        <country>USA</country>
96      </postal>
97      <email>JeffMogul@acm.org</email>
98    </address>
99  </author>
100
101  <author initials="H." surname="Frystyk" fullname="Henrik Frystyk Nielsen">
102    <organization abbrev="Microsoft">Microsoft Corporation</organization>
103    <address>
104      <postal>
105        <street>1 Microsoft Way</street>
106        <city>Redmond</city>
107        <region>WA</region>
108        <code>98052</code>
109        <country>USA</country>
110      </postal>
111      <email>henrikn@microsoft.com</email>
112    </address>
113  </author>
114
115  <author initials="L." surname="Masinter" fullname="Larry Masinter">
116    <organization abbrev="Adobe Systems">Adobe Systems, Incorporated</organization>
117    <address>
118      <postal>
119        <street>345 Park Ave</street>
120        <city>San Jose</city>
121        <region>CA</region>
122        <code>95110</code>
123        <country>USA</country>
124      </postal>
125      <email>LMM@acm.org</email>
126      <uri>http://larry.masinter.net/</uri>
127    </address>
128  </author>
129 
130  <author initials="P." surname="Leach" fullname="Paul J. Leach">
131    <organization abbrev="Microsoft">Microsoft Corporation</organization>
132    <address>
133      <postal>
134        <street>1 Microsoft Way</street>
135        <city>Redmond</city>
136        <region>WA</region>
137        <code>98052</code>
138      </postal>
139      <email>paulle@microsoft.com</email>
140    </address>
141  </author>
142   
143  <author initials="T." surname="Berners-Lee" fullname="Tim Berners-Lee">
144    <organization abbrev="W3C/MIT">World Wide Web Consortium</organization>
145    <address>
146      <postal>
147        <street>MIT Computer Science and Artificial Intelligence Laboratory</street>
148        <street>The Stata Center, Building 32</street>
149        <street>32 Vassar Street</street>
150        <city>Cambridge</city>
151        <region>MA</region>
152        <code>02139</code>
153        <country>USA</country>
154      </postal>
155      <email>timbl@w3.org</email>
156      <uri>http://www.w3.org/People/Berners-Lee/</uri>
157    </address>
158  </author>
159
160  <author initials="Y." surname="Lafon" fullname="Yves Lafon" role="editor">
161    <organization abbrev="W3C">World Wide Web Consortium</organization>
162    <address>
163      <postal>
164        <street>W3C / ERCIM</street>
165        <street>2004, rte des Lucioles</street>
166        <city>Sophia-Antipolis</city>
167        <region>AM</region>
168        <code>06902</code>
169        <country>France</country>
170      </postal>
171      <email>ylafon@w3.org</email>
172      <uri>http://www.raubacapeu.net/people/yves/</uri>
173    </address>
174  </author>
175
176  <author initials="J. F." surname="Reschke" fullname="Julian F. Reschke" role="editor">
177    <organization abbrev="greenbytes">greenbytes GmbH</organization>
178    <address>
179      <postal>
180        <street>Hafenweg 16</street>
181        <city>Muenster</city><region>NW</region><code>48155</code>
182        <country>Germany</country>
183      </postal>
184      <phone>+49 251 2807760</phone>   
185      <facsimile>+49 251 2807761</facsimile>   
186      <email>julian.reschke@greenbytes.de</email>       
187      <uri>http://greenbytes.de/tech/webdav/</uri>     
188    </address>
189  </author>
190
191  <date month="&ID-MONTH;" year="&ID-YEAR;"/>
192
193<abstract>
194<t>
195   The Hypertext Transfer Protocol (HTTP) is an application-level
196   protocol for distributed, collaborative, hypermedia information
197   systems. HTTP has been in use by the World Wide Web global information
198   initiative since 1990. This document is Part 1 of the seven-part specification
199   that defines the protocol referred to as "HTTP/1.1" and, taken together,
200   obsoletes RFC 2616.  Part 1 provides an overview of HTTP and
201   its associated terminology, defines the "http" and "https" Uniform
202   Resource Identifier (URI) schemes, defines the generic message syntax
203   and parsing requirements for HTTP message frames, and describes
204   general security concerns for implementations.
205</t>
206</abstract>
207
208<note title="Editorial Note (To be removed by RFC Editor)">
209  <t>
210    This version of the HTTP specification contains only minimal editorial
211    changes from <xref target="RFC2616"/> (abstract, introductory paragraph,
212    and authors' addresses).  All other changes are due to partitioning the
213    original into seven mostly independent parts.  The intent is for readers
214    of future drafts to able to use draft 00 as the basis for comparison
215    when the WG makes later changes to the specification text.  This draft
216    will shortly be followed by draft 01 (containing the first round of changes
217    that have already been agreed to on the mailing list). There is no point in
218    reviewing this draft other than to verify that the partitioning has been
219    done correctly.  Roy T. Fielding, Yves Lafon, and Julian Reschke
220    will be the editors after draft 00 is submitted.
221  </t>
222  <t>
223    Discussion of this draft should take place on the HTTPBIS working group
224    mailing list (ietf-http-wg@w3.org). The current issues list is
225    at <eref target="http://www3.tools.ietf.org/wg/httpbis/trac/report/11"/>
226    and related documents (including fancy diffs) can be found at
227    <eref target="http://www3.tools.ietf.org/wg/httpbis/"/>.
228  </t>
229</note>
230</front>
231<middle>
232<section title="Introduction" anchor="introduction">
233<t>
234   This document will define aspects of HTTP related to overall network
235   operation, message framing, interaction with transport protocols, and
236   URI schemes. Right now it only includes the extracted relevant sections
237   of <xref target="RFC2616"/>.
238</t>
239<section title="Purpose" anchor="intro.purpose">
240<t>
241   The Hypertext Transfer Protocol (HTTP) is an application-level
242   protocol for distributed, collaborative, hypermedia information
243   systems. HTTP has been in use by the World-Wide Web global
244   information initiative since 1990. The first version of HTTP,
245   referred to as HTTP/0.9, was a simple protocol for raw data transfer
246   across the Internet. HTTP/1.0, as defined by <xref target="RFC1945"/>, improved
247   the protocol by allowing messages to be in the format of MIME-like
248   messages, containing metainformation about the data transferred and
249   modifiers on the request/response semantics. However, HTTP/1.0 does
250   not sufficiently take into consideration the effects of hierarchical
251   proxies, caching, the need for persistent connections, or virtual
252   hosts. In addition, the proliferation of incompletely-implemented
253   applications calling themselves "HTTP/1.0" has necessitated a
254   protocol version change in order for two communicating applications
255   to determine each other's true capabilities.
256</t>
257<t>
258   This specification defines the protocol referred to as "HTTP/1.1".
259   This protocol includes more stringent requirements than HTTP/1.0 in
260   order to ensure reliable implementation of its features.
261</t>
262<t>
263   Practical information systems require more functionality than simple
264   retrieval, including search, front-end update, and annotation. HTTP
265   allows an open-ended set of methods and headers that indicate the
266   purpose of a request <xref target="RFC2324"/>. It builds on the discipline of reference
267   provided by the Uniform Resource Identifier (URI) <xref target="RFC1630"/>, as a location
268   (URL) <xref target="RFC1738"/> or name (URN) <xref target="RFC1737"/>, for indicating the resource to which a
269   method is to be applied. Messages are passed in a format similar to
270   that used by Internet mail <xref target="RFC822"/> as defined by the Multipurpose
271   Internet Mail Extensions (MIME) <xref target="RFC2045"/>.
272</t>
273<t>
274   HTTP is also used as a generic protocol for communication between
275   user agents and proxies/gateways to other Internet systems, including
276   those supported by the SMTP <xref target="RFC821"/>, NNTP <xref target="RFC3977"/>, FTP <xref target="RFC959"/>, Gopher <xref target="RFC1436"/>,
277   and WAIS <xref target="WAIS"/> protocols. In this way, HTTP allows basic hypermedia
278   access to resources available from diverse applications.
279</t>
280</section>
281
282<section title="Requirements" anchor="intro.requirements">
283<t>
284   The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
285   "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this
286   document are to be interpreted as described in <xref target="RFC2119"/>.
287</t>
288<t>
289   An implementation is not compliant if it fails to satisfy one or more
290   of the &MUST; or &REQUIRED; level requirements for the protocols it
291   implements. An implementation that satisfies all the &MUST; or &REQUIRED;
292   level and all the &SHOULD; level requirements for its protocols is said
293   to be "unconditionally compliant"; one that satisfies all the &MUST;
294   level requirements but not all the &SHOULD; level requirements for its
295   protocols is said to be "conditionally compliant."
296</t>
297</section>
298
299<section title="Terminology" anchor="intro.terminology">
300<t>
301   This specification uses a number of terms to refer to the roles
302   played by participants in, and objects of, the HTTP communication.
303</t>
304<t>
305  <iref item="connection"/>
306  <x:dfn>connection</x:dfn>
307  <list>
308    <t>
309      A transport layer virtual circuit established between two programs
310      for the purpose of communication.
311    </t>
312  </list>
313</t>
314<t>
315  <iref item="message"/>
316  <x:dfn>message</x:dfn>
317  <list>
318    <t>
319      The basic unit of HTTP communication, consisting of a structured
320      sequence of octets matching the syntax defined in <xref target="http.message"/> and
321      transmitted via the connection.
322    </t>
323  </list>
324</t>
325<t>
326  <iref item="request"/>
327  <x:dfn>request</x:dfn>
328  <list>
329    <t>
330      An HTTP request message, as defined in <xref target="request"/>.
331    </t>
332  </list>
333</t>
334<t>
335  <iref item="response"/>
336  <x:dfn>response</x:dfn>
337  <list>
338    <t>
339      An HTTP response message, as defined in <xref target="response"/>.
340    </t>
341  </list>
342</t>
343<t>
344  <iref item="resource"/>
345  <x:dfn>resource</x:dfn>
346  <list>
347    <t>
348      A network data object or service that can be identified by a URI,
349      as defined in <xref target="uri"/>. Resources may be available in multiple
350      representations (e.g. multiple languages, data formats, size, and
351      resolutions) or vary in other ways.
352    </t>
353  </list>
354</t>
355<t>
356  <iref item="entity"/>
357  <x:dfn>entity</x:dfn>
358  <list>
359    <t>
360      The information transferred as the payload of a request or
361      response. An entity consists of metainformation in the form of
362      entity-header fields and content in the form of an entity-body, as
363      described in &entity;.
364    </t>
365  </list>
366</t>
367<t>
368  <iref item="representation"/>
369  <x:dfn>representation</x:dfn>
370  <list>
371    <t>
372      An entity included with a response that is subject to content
373      negotiation, as described in &content.negotiation;. There may exist multiple
374      representations associated with a particular response status.
375    </t>
376  </list>
377</t>
378<t>
379  <iref item="content negotiation"/>
380  <x:dfn>content negotiation</x:dfn>
381  <list>
382    <t>
383      The mechanism for selecting the appropriate representation when
384      servicing a request, as described in &content.negotiation;. The
385      representation of entities in any response can be negotiated
386      (including error responses).
387    </t>
388  </list>
389</t>
390<t>
391  <iref item="variant"/>
392  <x:dfn>variant</x:dfn>
393  <list>
394    <t>
395      A resource may have one, or more than one, representation(s)
396      associated with it at any given instant. Each of these
397      representations is termed a `variant'.  Use of the term `variant'
398      does not necessarily imply that the resource is subject to content
399      negotiation.
400    </t>
401  </list>
402</t>
403<t>
404  <iref item="client"/>
405  <x:dfn>client</x:dfn>
406  <list>
407    <t>
408      A program that establishes connections for the purpose of sending
409      requests.
410    </t>
411  </list>
412</t>
413<t>
414  <iref item="user agent"/>
415  <x:dfn>user agent</x:dfn>
416  <list>
417    <t>
418      The client which initiates a request. These are often browsers,
419      editors, spiders (web-traversing robots), or other end user tools.
420    </t>
421  </list>
422</t>
423<t>
424  <iref item="server"/>
425  <x:dfn>server</x:dfn>
426  <list>
427    <t>
428      An application program that accepts connections in order to
429      service requests by sending back responses. Any given program may
430      be capable of being both a client and a server; our use of these
431      terms refers only to the role being performed by the program for a
432      particular connection, rather than to the program's capabilities
433      in general. Likewise, any server may act as an origin server,
434      proxy, gateway, or tunnel, switching behavior based on the nature
435      of each request.
436    </t>
437  </list>
438</t>
439<t>
440  <iref item="origin server"/>
441  <x:dfn>origin server</x:dfn>
442  <list>
443    <t>
444      The server on which a given resource resides or is to be created.
445    </t>
446  </list>
447</t>
448<t>
449  <iref item="proxy"/>
450  <x:dfn>proxy</x:dfn>
451  <list>
452    <t>
453      An intermediary program which acts as both a server and a client
454      for the purpose of making requests on behalf of other clients.
455      Requests are serviced internally or by passing them on, with
456      possible translation, to other servers. A proxy &MUST; implement
457      both the client and server requirements of this specification. A
458      "transparent proxy" is a proxy that does not modify the request or
459      response beyond what is required for proxy authentication and
460      identification. A "non-transparent proxy" is a proxy that modifies
461      the request or response in order to provide some added service to
462      the user agent, such as group annotation services, media type
463      transformation, protocol reduction, or anonymity filtering. Except
464      where either transparent or non-transparent behavior is explicitly
465      stated, the HTTP proxy requirements apply to both types of
466      proxies.
467    </t>
468  </list>
469</t>
470<t>
471  <iref item="gateway"/>
472  <x:dfn>gateway</x:dfn>
473  <list>
474    <t>
475      A server which acts as an intermediary for some other server.
476      Unlike a proxy, a gateway receives requests as if it were the
477      origin server for the requested resource; the requesting client
478      may not be aware that it is communicating with a gateway.
479    </t>
480  </list>
481</t>
482<t>
483  <iref item="tunnel"/>
484  <x:dfn>tunnel</x:dfn>
485  <list>
486    <t>
487      An intermediary program which is acting as a blind relay between
488      two connections. Once active, a tunnel is not considered a party
489      to the HTTP communication, though the tunnel may have been
490      initiated by an HTTP request. The tunnel ceases to exist when both
491      ends of the relayed connections are closed.
492    </t>
493  </list>
494</t>
495<t>
496  <iref item="cache"/>
497  <x:dfn>cache</x:dfn>
498  <list>
499    <t>
500      A program's local store of response messages and the subsystem
501      that controls its message storage, retrieval, and deletion. A
502      cache stores cacheable responses in order to reduce the response
503      time and network bandwidth consumption on future, equivalent
504      requests. Any client or server may include a cache, though a cache
505      cannot be used by a server that is acting as a tunnel.
506    </t>
507  </list>
508</t>
509<t>
510  <iref item="cacheable"/>
511  <x:dfn>cacheable</x:dfn>
512  <list>
513    <t>
514      A response is cacheable if a cache is allowed to store a copy of
515      the response message for use in answering subsequent requests. The
516      rules for determining the cacheability of HTTP responses are
517      defined in &caching;. Even if a resource is cacheable, there may
518      be additional constraints on whether a cache can use the cached
519      copy for a particular request.
520    </t>
521  </list>
522</t>
523<t>
524  <iref item="upstream"/>
525  <iref item="downstream"/>
526  <x:dfn>upstream</x:dfn>/<x:dfn>downstream</x:dfn>
527  <list>
528    <t>
529      Upstream and downstream describe the flow of a message: all
530      messages flow from upstream to downstream.
531    </t>
532  </list>
533</t>
534<t>
535  <iref item="inbound"/>
536  <iref item="outbound"/>
537  <x:dfn>inbound</x:dfn>/<x:dfn>outbound</x:dfn>
538  <list>
539    <t>
540      Inbound and outbound refer to the request and response paths for
541      messages: "inbound" means "traveling toward the origin server",
542      and "outbound" means "traveling toward the user agent"
543    </t>
544  </list>
545</t>
546</section>
547
548<section title="Overall Operation" anchor="intro.overall.operation">
549<t>
550   The HTTP protocol is a request/response protocol. A client sends a
551   request to the server in the form of a request method, URI, and
552   protocol version, followed by a MIME-like message containing request
553   modifiers, client information, and possible body content over a
554   connection with a server. The server responds with a status line,
555   including the message's protocol version and a success or error code,
556   followed by a MIME-like message containing server information, entity
557   metainformation, and possible entity-body content. The relationship
558   between HTTP and MIME is described in &diff2045entity;.
559</t>
560<t>
561   Most HTTP communication is initiated by a user agent and consists of
562   a request to be applied to a resource on some origin server. In the
563   simplest case, this may be accomplished via a single connection (v)
564   between the user agent (UA) and the origin server (O).
565</t>
566<figure><artwork type="drawing">
567       request chain ------------------------&gt;
568    UA -------------------v------------------- O
569       &lt;----------------------- response chain
570</artwork></figure>
571<t>
572   A more complicated situation occurs when one or more intermediaries
573   are present in the request/response chain. There are three common
574   forms of intermediary: proxy, gateway, and tunnel. A proxy is a
575   forwarding agent, receiving requests for a URI in its absolute form,
576   rewriting all or part of the message, and forwarding the reformatted
577   request toward the server identified by the URI. A gateway is a
578   receiving agent, acting as a layer above some other server(s) and, if
579   necessary, translating the requests to the underlying server's
580   protocol. A tunnel acts as a relay point between two connections
581   without changing the messages; tunnels are used when the
582   communication needs to pass through an intermediary (such as a
583   firewall) even when the intermediary cannot understand the contents
584   of the messages.
585</t>
586<figure><artwork type="drawing">
587       request chain --------------------------------------&gt;
588    UA -----v----- A -----v----- B -----v----- C -----v----- O
589       &lt;------------------------------------- response chain
590</artwork></figure>
591<t>
592   The figure above shows three intermediaries (A, B, and C) between the
593   user agent and origin server. A request or response message that
594   travels the whole chain will pass through four separate connections.
595   This distinction is important because some HTTP communication options
596   may apply only to the connection with the nearest, non-tunnel
597   neighbor, only to the end-points of the chain, or to all connections
598   along the chain. Although the diagram is linear, each participant may
599   be engaged in multiple, simultaneous communications. For example, B
600   may be receiving requests from many clients other than A, and/or
601   forwarding requests to servers other than C, at the same time that it
602   is handling A's request.
603</t>
604<t>
605   Any party to the communication which is not acting as a tunnel may
606   employ an internal cache for handling requests. The effect of a cache
607   is that the request/response chain is shortened if one of the
608   participants along the chain has a cached response applicable to that
609   request. The following illustrates the resulting chain if B has a
610   cached copy of an earlier response from O (via C) for a request which
611   has not been cached by UA or A.
612</t>
613<figure><artwork type="drawing">
614          request chain ----------&gt;
615       UA -----v----- A -----v----- B - - - - - - C - - - - - - O
616          &lt;--------- response chain
617</artwork></figure>
618<t>
619   Not all responses are usefully cacheable, and some requests may
620   contain modifiers which place special requirements on cache behavior.
621   HTTP requirements for cache behavior and cacheable responses are
622   defined in &caching;.
623</t>
624<t>
625   In fact, there are a wide variety of architectures and configurations
626   of caches and proxies currently being experimented with or deployed
627   across the World Wide Web. These systems include national hierarchies
628   of proxy caches to save transoceanic bandwidth, systems that
629   broadcast or multicast cache entries, organizations that distribute
630   subsets of cached data via CD-ROM, and so on. HTTP systems are used
631   in corporate intranets over high-bandwidth links, and for access via
632   PDAs with low-power radio links and intermittent connectivity. The
633   goal of HTTP/1.1 is to support the wide diversity of configurations
634   already deployed while introducing protocol constructs that meet the
635   needs of those who build web applications that require high
636   reliability and, failing that, at least reliable indications of
637   failure.
638</t>
639<t>
640   HTTP communication usually takes place over TCP/IP connections. The
641   default port is TCP 80 (<eref target="http://www.iana.org/assignments/port-numbers"/>), but other ports can be used. This does
642   not preclude HTTP from being implemented on top of any other protocol
643   on the Internet, or on other networks. HTTP only presumes a reliable
644   transport; any protocol that provides such guarantees can be used;
645   the mapping of the HTTP/1.1 request and response structures onto the
646   transport data units of the protocol in question is outside the scope
647   of this specification.
648</t>
649<t>
650   In HTTP/1.0, most implementations used a new connection for each
651   request/response exchange. In HTTP/1.1, a connection may be used for
652   one or more request/response exchanges, although connections may be
653   closed for a variety of reasons (see <xref target="persistent.connections"/>).
654</t>
655</section>
656</section>
657
658<section title="Notational Conventions and Generic Grammar" anchor="notation">
659
660<section title="Augmented BNF" anchor="notation.abnf">
661<t>
662   All of the mechanisms specified in this document are described in
663   both prose and an augmented Backus-Naur Form (BNF) similar to that
664   used by <xref target="RFC822"/>. Implementors will need to be familiar with the
665   notation in order to understand this specification. The augmented BNF
666   includes the following constructs:
667</t>
668<t>
669   name = definition
670  <list>
671    <t>
672      The name of a rule is simply the name itself (without any
673      enclosing "&lt;" and "&gt;") and is separated from its definition by the
674      equal "=" character. White space is only significant in that
675      indentation of continuation lines is used to indicate a rule
676      definition that spans more than one line. Certain basic rules are
677      in uppercase, such as SP, LWS, HT, CRLF, DIGIT, ALPHA, etc. Angle
678      brackets are used within definitions whenever their presence will
679      facilitate discerning the use of rule names.
680    </t>
681  </list>
682</t>
683<t>
684   "literal"
685  <list>
686    <t>
687      Quotation marks surround literal text. Unless stated otherwise,
688      the text is case-insensitive.
689    </t>
690  </list>
691</t>
692<t>
693   rule1 | rule2
694  <list>
695    <t>
696      Elements separated by a bar ("|") are alternatives, e.g., "yes |
697      no" will accept yes or no.
698    </t>
699  </list>
700</t>
701<t>
702   (rule1 rule2)
703  <list>
704    <t>
705      Elements enclosed in parentheses are treated as a single element.
706      Thus, "(elem (foo | bar) elem)" allows the token sequences "elem
707      foo elem" and "elem bar elem".
708    </t>
709  </list>
710</t>
711<t>
712   *rule
713  <list>
714    <t>
715      The character "*" preceding an element indicates repetition. The
716      full form is "&lt;n&gt;*&lt;m&gt;element" indicating at least &lt;n&gt; and at most
717      &lt;m&gt; occurrences of element. Default values are 0 and infinity so
718      that "*(element)" allows any number, including zero; "1*element"
719      requires at least one; and "1*2element" allows one or two.
720    </t>
721  </list>
722</t>
723<t>
724   [rule]
725  <list>
726    <t>
727      Square brackets enclose optional elements; "[foo bar]" is
728      equivalent to "*1(foo bar)".
729    </t>
730  </list>
731</t>
732<t>
733   N rule
734  <list>
735    <t>
736      Specific repetition: "&lt;n&gt;(element)" is equivalent to
737      "&lt;n&gt;*&lt;n&gt;(element)"; that is, exactly &lt;n&gt; occurrences of (element).
738      Thus 2DIGIT is a 2-digit number, and 3ALPHA is a string of three
739      alphabetic characters.
740    </t>
741  </list>
742</t>
743<t>
744   #rule
745  <list>
746    <t>
747      A construct "#" is defined, similar to "*", for defining lists of
748      elements. The full form is "&lt;n&gt;#&lt;m&gt;element" indicating at least
749      &lt;n&gt; and at most &lt;m&gt; elements, each separated by one or more commas
750      (",") and &OPTIONAL; linear white space (LWS). This makes the usual
751      form of lists very easy; a rule such as
752    </t>
753    <t>
754         ( *LWS element *( *LWS "," *LWS element ))
755    </t>
756    <t>
757      can be shown as
758    </t>
759    <t>
760         1#element
761    </t>
762    <t>
763      Wherever this construct is used, null elements are allowed, but do
764      not contribute to the count of elements present. That is,
765      "(element), , (element) " is permitted, but counts as only two
766      elements. Therefore, where at least one element is required, at
767      least one non-null element &MUST; be present. Default values are 0
768      and infinity so that "#element" allows any number, including zero;
769      "1#element" requires at least one; and "1#2element" allows one or
770      two.
771    </t>
772  </list>
773</t>
774<t>
775   ; comment
776  <list>
777    <t>
778      A semi-colon, set off some distance to the right of rule text,
779      starts a comment that continues to the end of line. This is a
780      simple way of including useful notes in parallel with the
781      specifications.
782    </t>
783  </list>
784</t>
785<t>
786   implied *LWS
787  <list>
788    <t>
789      The grammar described by this specification is word-based. Except
790      where noted otherwise, linear white space (LWS) can be included
791      between any two adjacent words (token or quoted-string), and
792      between adjacent words and separators, without changing the
793      interpretation of a field. At least one delimiter (LWS and/or
794      separators) &MUST; exist between any two tokens (for the definition
795      of "token" below), since they would otherwise be interpreted as a
796      single token.
797    </t>
798  </list>
799</t>
800</section>
801
802<section title="Basic Rules" anchor="basic.rules">
803<x:anchor-alias value="OCTET"/>
804<x:anchor-alias value="CHAR"/>
805<x:anchor-alias value="UPALPHA"/>
806<x:anchor-alias value="LOALPHA"/>
807<x:anchor-alias value="ALPHA"/>
808<x:anchor-alias value="DIGIT"/>
809<x:anchor-alias value="CTL"/>
810<x:anchor-alias value="CR"/>
811<x:anchor-alias value="LF"/>
812<x:anchor-alias value="SP"/>
813<x:anchor-alias value="HT"/>
814<x:anchor-alias value="CRLF"/>
815<x:anchor-alias value="LWS"/>
816<x:anchor-alias value="TEXT"/>
817<x:anchor-alias value="HEX"/>
818<x:anchor-alias value="token"/>
819<x:anchor-alias value="separators"/>
820<x:anchor-alias value="comment"/>
821<x:anchor-alias value="ctext"/>
822<x:anchor-alias value="quoted-string"/>
823<x:anchor-alias value="qdtext"/>
824<x:anchor-alias value="quoted-pair"/>
825<t>
826   The following rules are used throughout this specification to
827   describe basic parsing constructs. The US-ASCII coded character set
828   is defined by ANSI X3.4-1986 <xref target="USASCII"/>.
829</t>
830<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"/>
831    OCTET          = &lt;any 8-bit sequence of data&gt;
832    CHAR           = &lt;any US-ASCII character (octets 0 - 127)&gt;
833    UPALPHA        = &lt;any US-ASCII uppercase letter "A".."Z"&gt;
834    LOALPHA        = &lt;any US-ASCII lowercase letter "a".."z"&gt;
835    ALPHA          = UPALPHA | LOALPHA
836    DIGIT          = &lt;any US-ASCII digit "0".."9"&gt;
837    CTL            = &lt;any US-ASCII control character
838                     (octets 0 - 31) and DEL (127)&gt;
839    CR             = &lt;US-ASCII CR, carriage return (13)&gt;
840    LF             = &lt;US-ASCII LF, linefeed (10)&gt;
841    SP             = &lt;US-ASCII SP, space (32)&gt;
842    HT             = &lt;US-ASCII HT, horizontal-tab (9)&gt;
843    &lt;"&gt;            = &lt;US-ASCII double-quote mark (34)&gt;
844</artwork></figure>
845<t>
846   HTTP/1.1 defines the sequence CR LF as the end-of-line marker for all
847   protocol elements except the entity-body (see <xref target="tolerant.applications"/> for
848   tolerant applications). The end-of-line marker within an entity-body
849   is defined by its associated media type, as described in &payload;.
850</t>
851<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="CRLF"/>
852    CRLF           = CR LF
853</artwork></figure>
854<t>
855   HTTP/1.1 header field values can be folded onto multiple lines if the
856   continuation line begins with a space or horizontal tab. All linear
857   white space, including folding, has the same semantics as SP. A
858   recipient &MAY; replace any linear white space with a single SP before
859   interpreting the field value or forwarding the message downstream.
860</t>
861<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="LWS"/>
862    LWS            = [CRLF] 1*( SP | HT )
863</artwork></figure>
864<t>
865   The TEXT rule is only used for descriptive field contents and values
866   that are not intended to be interpreted by the message parser. Words
867   of *TEXT &MAY; contain characters from character sets other than ISO-8859-1
868   <xref target="ISO-8859"/> only when encoded according to the rules of
869   <xref target="RFC2047"/>.
870</t>
871<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="TEXT"/>
872    TEXT           = &lt;any OCTET except CTLs,
873                     but including LWS&gt;
874</artwork></figure>
875<t>
876   A CRLF is allowed in the definition of TEXT only as part of a header
877   field continuation. It is expected that the folding LWS will be
878   replaced with a single SP before interpretation of the TEXT value.
879</t>
880<t>
881   Hexadecimal numeric characters are used in several protocol elements.
882</t>
883<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="HEX"/>
884    HEX            = "A" | "B" | "C" | "D" | "E" | "F"
885                   | "a" | "b" | "c" | "d" | "e" | "f" | DIGIT
886</artwork></figure>
887<t>
888   Many HTTP/1.1 header field values consist of words separated by LWS
889   or special characters. These special characters &MUST; be in a quoted
890   string to be used within a parameter value (as defined in
891   <xref target="transfer.codings"/>).
892</t>
893<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="token"/><iref primary="true" item="Grammar" subitem="separators"/>
894    token          = 1*&lt;any CHAR except CTLs or separators&gt;
895    separators     = "(" | ")" | "&lt;" | "&gt;" | "@"
896                   | "," | ";" | ":" | "\" | &lt;"&gt;
897                   | "/" | "[" | "]" | "?" | "="
898                   | "{" | "}" | SP | HT
899</artwork></figure>
900<t>
901   Comments can be included in some HTTP header fields by surrounding
902   the comment text with parentheses. Comments are only allowed in
903   fields containing "comment" as part of their field value definition.
904   In all other fields, parentheses are considered part of the field
905   value.
906</t>
907<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="comment"/><iref primary="true" item="Grammar" subitem="ctext"/>
908    comment        = "(" *( ctext | quoted-pair | comment ) ")"
909    ctext          = &lt;any TEXT excluding "(" and ")"&gt;
910</artwork></figure>
911<t>
912   A string of text is parsed as a single word if it is quoted using
913   double-quote marks.
914</t>
915<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="quoted-string"/><iref primary="true" item="Grammar" subitem="qdtext"/>
916    quoted-string  = ( &lt;"&gt; *(qdtext | quoted-pair ) &lt;"&gt; )
917    qdtext         = &lt;any TEXT except &lt;"&gt;&gt;
918</artwork></figure>
919<t>
920   The backslash character ("\") &MAY; be used as a single-character
921   quoting mechanism only within quoted-string and comment constructs.
922</t>
923<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="quoted-pair"/>
924    quoted-pair    = "\" CHAR
925</artwork></figure>
926</section>
927</section>
928
929<section title="Protocol Parameters" anchor="protocol.parameters">
930
931<section title="HTTP Version" anchor="http.version">
932<t>
933   HTTP uses a "&lt;major&gt;.&lt;minor&gt;" numbering scheme to indicate versions
934   of the protocol. The protocol versioning policy is intended to allow
935   the sender to indicate the format of a message and its capacity for
936   understanding further HTTP communication, rather than the features
937   obtained via that communication. No change is made to the version
938   number for the addition of message components which do not affect
939   communication behavior or which only add to extensible field values.
940   The &lt;minor&gt; number is incremented when the changes made to the
941   protocol add features which do not change the general message parsing
942   algorithm, but which may add to the message semantics and imply
943   additional capabilities of the sender. The &lt;major&gt; number is
944   incremented when the format of a message within the protocol is
945   changed. See <xref target="RFC2145"/> for a fuller explanation.
946</t>
947<t>
948   The version of an HTTP message is indicated by an HTTP-Version field
949   in the first line of the message. HTTP-Version is case-sensitive.
950</t>
951<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="HTTP-Version"/>
952       HTTP-Version   = "HTTP" "/" 1*DIGIT "." 1*DIGIT
953</artwork></figure>
954<t>
955   Note that the major and minor numbers &MUST; be treated as separate
956   integers and that each &MAY; be incremented higher than a single digit.
957   Thus, HTTP/2.4 is a lower version than HTTP/2.13, which in turn is
958   lower than HTTP/12.3. Leading zeros &MUST; be ignored by recipients and
959   &MUST-NOT; be sent.
960</t>
961<t>
962   An application that sends a request or response message that includes
963   HTTP-Version of "HTTP/1.1" &MUST; be at least conditionally compliant
964   with this specification. Applications that are at least conditionally
965   compliant with this specification &SHOULD; use an HTTP-Version of
966   "HTTP/1.1" in their messages, and &MUST; do so for any message that is
967   not compatible with HTTP/1.0. For more details on when to send
968   specific HTTP-Version values, see <xref target="RFC2145"/>.
969</t>
970<t>
971   The HTTP version of an application is the highest HTTP version for
972   which the application is at least conditionally compliant.
973</t>
974<t>
975   Proxy and gateway applications need to be careful when forwarding
976   messages in protocol versions different from that of the application.
977   Since the protocol version indicates the protocol capability of the
978   sender, a proxy/gateway &MUST-NOT; send a message with a version
979   indicator which is greater than its actual version. If a higher
980   version request is received, the proxy/gateway &MUST; either downgrade
981   the request version, or respond with an error, or switch to tunnel
982   behavior.
983</t>
984<t>
985   Due to interoperability problems with HTTP/1.0 proxies discovered
986   since the publication of <xref target="RFC2068"/>, caching proxies &MUST;, gateways
987   &MAY;, and tunnels &MUST-NOT; upgrade the request to the highest version
988   they support. The proxy/gateway's response to that request &MUST; be in
989   the same major version as the request.
990</t>
991<t>
992  <list>
993    <t>
994      <x:h>Note:</x:h> Converting between versions of HTTP may involve modification
995      of header fields required or forbidden by the versions involved.
996    </t>
997  </list>
998</t>
999</section>
1000
1001<section title="Uniform Resource Identifiers" anchor="uri">
1002<t>
1003   URIs have been known by many names: WWW addresses, Universal Document
1004   Identifiers, Universal Resource Identifiers <xref target="RFC1630"/>, and finally the
1005   combination of Uniform Resource Locators (URL) <xref target="RFC1738"/> and Names (URN)
1006   <xref target="RFC1737"/>. As far as HTTP is concerned, Uniform Resource Identifiers are
1007   simply formatted strings which identify--via name, location, or any
1008   other characteristic--a resource.
1009</t>
1010
1011<section title="General Syntax" anchor="general.syntax">
1012<t>
1013   URIs in HTTP can be represented in absolute form or relative to some
1014   known base URI <xref target="RFC1808"/>, depending upon the context of their use. The two
1015   forms are differentiated by the fact that absolute URIs always begin
1016   with a scheme name followed by a colon. For definitive information on
1017   URL syntax and semantics, see "Uniform Resource Identifiers (URI):
1018   Generic Syntax and Semantics," <xref target="RFC2396"/> (which replaces <xref target="RFC1738"/>
1019   and <xref target="RFC1808"/>). This specification adopts the
1020   definitions of "URI-reference", "absoluteURI", "relativeURI", "port",
1021   "host","abs_path", "rel_path", and "authority" from that
1022   specification.
1023</t>
1024<t>
1025   The HTTP protocol does not place any a priori limit on the length of
1026   a URI. Servers &MUST; be able to handle the URI of any resource they
1027   serve, and &SHOULD; be able to handle URIs of unbounded length if they
1028   provide GET-based forms that could generate such URIs. A server
1029   &SHOULD; return 414 (Request-URI Too Long) status if a URI is longer
1030   than the server can handle (see &status-414;).
1031</t>
1032<t>
1033  <list>
1034    <t>
1035      <x:h>Note:</x:h> Servers ought to be cautious about depending on URI lengths
1036      above 255 bytes, because some older client or proxy
1037      implementations might not properly support these lengths.
1038    </t>
1039  </list>
1040</t>
1041</section>
1042
1043<section title="http URL" anchor="http.url">
1044<t>
1045   The "http" scheme is used to locate network resources via the HTTP
1046   protocol. This section defines the scheme-specific syntax and
1047   semantics for http URLs.
1048</t>
1049<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="http_URL"/>
1050http_URL = "http:" "//" host [ ":" port ] [ abs_path [ "?" query ]]
1051</artwork></figure>
1052<t>
1053   If the port is empty or not given, port 80 is assumed. The semantics
1054   are that the identified resource is located at the server listening
1055   for TCP connections on that port of that host, and the Request-URI
1056   for the resource is abs_path (<xref target="request-uri"/>). The use of IP addresses
1057   in URLs &SHOULD; be avoided whenever possible (see <xref target="RFC1900"/>). If
1058   the abs_path is not present in the URL, it &MUST; be given as "/" when
1059   used as a Request-URI for a resource (<xref target="request-uri"/>). If a proxy
1060
1061   receives a host name which is not a fully qualified domain name, it
1062   &MAY; add its domain to the host name it received. If a proxy receives
1063   a fully qualified domain name, the proxy &MUST-NOT; change the host
1064   name.
1065</t>
1066</section>
1067
1068<section title="URI Comparison" anchor="uri.comparison">
1069<t>
1070   When comparing two URIs to decide if they match or not, a client
1071   &SHOULD; use a case-sensitive octet-by-octet comparison of the entire
1072   URIs, with these exceptions:
1073  <list style="symbols">
1074    <t>A port that is empty or not given is equivalent to the default
1075        port for that URI-reference;</t>
1076    <t>Comparisons of host names &MUST; be case-insensitive;</t>
1077    <t>Comparisons of scheme names &MUST; be case-insensitive;</t>
1078    <t>An empty abs_path is equivalent to an abs_path of "/".</t>
1079  </list>
1080</t>
1081<t>
1082   Characters other than those in the "reserved" set (see
1083   <xref target="RFC2396"/>) are equivalent to their ""%" HEX HEX" encoding.
1084</t>
1085<t>
1086   For example, the following three URIs are equivalent:
1087</t>
1088<figure><artwork type="example">
1089   http://example.com:80/~smith/home.html
1090   http://EXAMPLE.com/%7Esmith/home.html
1091   http://EXAMPLE.com:/%7esmith/home.html
1092</artwork></figure>
1093</section>
1094</section>
1095
1096<section title="Date/Time Formats" anchor="date.time.formats">
1097<section title="Full Date" anchor="full.date">
1098<t>
1099   HTTP applications have historically allowed three different formats
1100   for the representation of date/time stamps:
1101</t>
1102<figure><artwork type="example">
1103   Sun, 06 Nov 1994 08:49:37 GMT  ; RFC 822, updated by RFC 1123
1104   Sunday, 06-Nov-94 08:49:37 GMT ; obsolete RFC 850 format
1105   Sun Nov  6 08:49:37 1994       ; ANSI C's asctime() format
1106</artwork></figure>
1107<t>
1108   The first format is preferred as an Internet standard and represents
1109   a fixed-length subset of that defined by <xref target="RFC1123"/> (an update to
1110   <xref target="RFC822"/>). The other formats are described here only for
1111   compatibility with obsolete implementations.
1112   HTTP/1.1 clients and servers that parse the date value &MUST; accept
1113   all three formats (for compatibility with HTTP/1.0), though they &MUST;
1114   only generate the RFC 1123 format for representing HTTP-date values
1115   in header fields. See <xref target="tolerant.applications"/> for further information.
1116</t>
1117<t><list><t>
1118      <x:h>Note:</x:h> Recipients of date values are encouraged to be robust in
1119      accepting date values that may have been sent by non-HTTP
1120      applications, as is sometimes the case when retrieving or posting
1121      messages via proxies/gateways to SMTP or NNTP.
1122</t></list></t>
1123<t>
1124   All HTTP date/time stamps &MUST; be represented in Greenwich Mean Time
1125   (GMT), without exception. For the purposes of HTTP, GMT is exactly
1126   equal to UTC (Coordinated Universal Time). This is indicated in the
1127   first two formats by the inclusion of "GMT" as the three-letter
1128   abbreviation for time zone, and &MUST; be assumed when reading the
1129   asctime format. HTTP-date is case sensitive and &MUST-NOT; include
1130   additional LWS beyond that specifically included as SP in the
1131   grammar.
1132</t>
1133<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"/>
1134    HTTP-date    = rfc1123-date | rfc850-date | asctime-date
1135    rfc1123-date = wkday "," SP date1 SP time SP "GMT"
1136    rfc850-date  = weekday "," SP date2 SP time SP "GMT"
1137    asctime-date = wkday SP date3 SP time SP 4DIGIT
1138    date1        = 2DIGIT SP month SP 4DIGIT
1139                   ; day month year (e.g., 02 Jun 1982)
1140    date2        = 2DIGIT "-" month "-" 2DIGIT
1141                   ; day-month-year (e.g., 02-Jun-82)
1142    date3        = month SP ( 2DIGIT | ( SP 1DIGIT ))
1143                   ; month day (e.g., Jun  2)
1144    time         = 2DIGIT ":" 2DIGIT ":" 2DIGIT
1145                   ; 00:00:00 - 23:59:59
1146    wkday        = "Mon" | "Tue" | "Wed"
1147                 | "Thu" | "Fri" | "Sat" | "Sun"
1148    weekday      = "Monday" | "Tuesday" | "Wednesday"
1149                 | "Thursday" | "Friday" | "Saturday" | "Sunday"
1150    month        = "Jan" | "Feb" | "Mar" | "Apr"
1151                 | "May" | "Jun" | "Jul" | "Aug"
1152                 | "Sep" | "Oct" | "Nov" | "Dec"
1153</artwork></figure>
1154<t>
1155      <x:h>Note:</x:h> HTTP requirements for the date/time stamp format apply only
1156      to their usage within the protocol stream. Clients and servers are
1157      not required to use these formats for user presentation, request
1158      logging, etc.
1159</t>
1160</section>
1161</section>
1162
1163<section title="Transfer Codings" anchor="transfer.codings">
1164<t>
1165   Transfer-coding values are used to indicate an encoding
1166   transformation that has been, can be, or may need to be applied to an
1167   entity-body in order to ensure "safe transport" through the network.
1168   This differs from a content coding in that the transfer-coding is a
1169   property of the message, not of the original entity.
1170</t>
1171<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="transfer-coding"/><iref primary="true" item="Grammar" subitem="transfer-extension"/>
1172    transfer-coding         = "chunked" | transfer-extension
1173    transfer-extension      = token *( ";" parameter )
1174</artwork></figure>
1175<t>
1176   Parameters are in  the form of attribute/value pairs.
1177</t>
1178<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"/>
1179    parameter               = attribute "=" value
1180    attribute               = token
1181    value                   = token | quoted-string
1182</artwork></figure>
1183<t>
1184   All transfer-coding values are case-insensitive. HTTP/1.1 uses
1185   transfer-coding values in the TE header field (<xref target="header.te"/>) and in
1186   the Transfer-Encoding header field (<xref target="header.transfer-encoding"/>).
1187</t>
1188<t>
1189   Whenever a transfer-coding is applied to a message-body, the set of
1190   transfer-codings &MUST; include "chunked", unless the message is
1191   terminated by closing the connection. When the "chunked" transfer-coding
1192   is used, it &MUST; be the last transfer-coding applied to the
1193   message-body. The "chunked" transfer-coding &MUST-NOT; be applied more
1194   than once to a message-body. These rules allow the recipient to
1195   determine the transfer-length of the message (<xref target="message.length"/>).
1196</t>
1197<t>
1198   Transfer-codings are analogous to the Content-Transfer-Encoding
1199   values of MIME <xref target="RFC2045"/>, which were designed to enable safe transport of
1200   binary data over a 7-bit transport service. However, safe transport
1201   has a different focus for an 8bit-clean transfer protocol. In HTTP,
1202   the only unsafe characteristic of message-bodies is the difficulty in
1203   determining the exact body length (<xref target="message.length"/>), or the desire to
1204   encrypt data over a shared transport.
1205</t>
1206<t>
1207   The Internet Assigned Numbers Authority (IANA) acts as a registry for
1208   transfer-coding value tokens. Initially, the registry contains the
1209   following tokens: "chunked" (<xref target="chunked.transfer.encoding"/>),
1210   "gzip" (&payload;), "compress" (&payload;), and "deflate" (&payload;).
1211</t>
1212<t>
1213   New transfer-coding value tokens &SHOULD; be registered in the same way
1214   as new content-coding value tokens (&payload;).
1215</t>
1216<t>
1217   A server which receives an entity-body with a transfer-coding it does
1218   not understand &SHOULD; return 501 (Unimplemented), and close the
1219   connection. A server &MUST-NOT; send transfer-codings to an HTTP/1.0
1220   client.
1221</t>
1222
1223<section title="Chunked Transfer Coding" anchor="chunked.transfer.encoding">
1224<t>
1225   The chunked encoding modifies the body of a message in order to
1226   transfer it as a series of chunks, each with its own size indicator,
1227   followed by an &OPTIONAL; trailer containing entity-header fields. This
1228   allows dynamically produced content to be transferred along with the
1229   information necessary for the recipient to verify that it has
1230   received the full message.
1231</t>
1232<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"/>
1233    Chunked-Body   = *chunk
1234                     last-chunk
1235                     trailer
1236                     CRLF
1237
1238    chunk          = chunk-size [ chunk-extension ] CRLF
1239                     chunk-data CRLF
1240    chunk-size     = 1*HEX
1241    last-chunk     = 1*("0") [ chunk-extension ] CRLF
1242
1243    chunk-extension= *( ";" chunk-ext-name [ "=" chunk-ext-val ] )
1244    chunk-ext-name = token
1245    chunk-ext-val  = token | quoted-string
1246    chunk-data     = chunk-size(OCTET)
1247    trailer        = *(entity-header CRLF)
1248</artwork></figure>
1249<t>
1250   The chunk-size field is a string of hex digits indicating the size of
1251   the chunk-data in octets. The chunked encoding is ended by any chunk whose size is
1252   zero, followed by the trailer, which is terminated by an empty line.
1253</t>
1254<t>
1255   The trailer allows the sender to include additional HTTP header
1256   fields at the end of the message. The Trailer header field can be
1257   used to indicate which header fields are included in a trailer (see
1258   <xref target="header.trailer"/>).
1259</t>
1260<t>
1261   A server using chunked transfer-coding in a response &MUST-NOT; use the
1262   trailer for any header fields unless at least one of the following is
1263   true:
1264  <list style="numbers">
1265    <t>the request included a TE header field that indicates "trailers" is
1266     acceptable in the transfer-coding of the  response, as described in
1267     <xref target="header.te"/>; or,</t>
1268
1269    <t>the server is the origin server for the response, the trailer
1270     fields consist entirely of optional metadata, and the recipient
1271     could use the message (in a manner acceptable to the origin server)
1272     without receiving this metadata.  In other words, the origin server
1273     is willing to accept the possibility that the trailer fields might
1274     be silently discarded along the path to the client.</t>
1275  </list>
1276</t>
1277<t>
1278   This requirement prevents an interoperability failure when the
1279   message is being received by an HTTP/1.1 (or later) proxy and
1280   forwarded to an HTTP/1.0 recipient. It avoids a situation where
1281   compliance with the protocol would have necessitated a possibly
1282   infinite buffer on the proxy.
1283</t>
1284<t>
1285   A process for decoding the "chunked" transfer-coding
1286   can be represented in pseudo-code as:
1287</t>
1288<figure><artwork type="code">
1289    length := 0
1290    read chunk-size, chunk-extension (if any) and CRLF
1291    while (chunk-size &gt; 0) {
1292       read chunk-data and CRLF
1293       append chunk-data to entity-body
1294       length := length + chunk-size
1295       read chunk-size and CRLF
1296    }
1297    read entity-header
1298    while (entity-header not empty) {
1299       append entity-header to existing header fields
1300       read entity-header
1301    }
1302    Content-Length := length
1303    Remove "chunked" from Transfer-Encoding
1304</artwork></figure>
1305<t>
1306   All HTTP/1.1 applications &MUST; be able to receive and decode the
1307   "chunked" transfer-coding, and &MUST; ignore chunk-extension extensions
1308   they do not understand.
1309</t>
1310</section>
1311</section>
1312
1313</section>
1314
1315<section title="HTTP Message" anchor="http.message">
1316
1317<section title="Message Types" anchor="message.types">
1318<t>
1319   HTTP messages consist of requests from client to server and responses
1320   from server to client.
1321</t>
1322<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="HTTP-message"/>
1323    HTTP-message   = Request | Response     ; HTTP/1.1 messages
1324</artwork></figure>
1325<t>
1326   Request (<xref target="request"/>) and Response (<xref target="response"/>) messages use the generic
1327   message format of <xref target="RFC822"/> for transferring entities (the payload
1328   of the message). Both types of message consist of a start-line, zero
1329   or more header fields (also known as "headers"), an empty line (i.e.,
1330   a line with nothing preceding the CRLF) indicating the end of the
1331   header fields, and possibly a message-body.
1332</t>
1333<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="generic-message"/><iref primary="true" item="Grammar" subitem="start-line"/>
1334     generic-message = start-line
1335                       *(message-header CRLF)
1336                       CRLF
1337                       [ message-body ]
1338     start-line      = Request-Line | Status-Line
1339</artwork></figure>
1340<t>
1341   In the interest of robustness, servers &SHOULD; ignore any empty
1342   line(s) received where a Request-Line is expected. In other words, if
1343   the server is reading the protocol stream at the beginning of a
1344   message and receives a CRLF first, it should ignore the CRLF.
1345</t>
1346<t>
1347   Certain buggy HTTP/1.0 client implementations generate extra CRLF's
1348   after a POST request. To restate what is explicitly forbidden by the
1349   BNF, an HTTP/1.1 client &MUST-NOT; preface or follow a request with an
1350   extra CRLF.
1351</t>
1352</section>
1353
1354<section title="Message Headers" anchor="message.headers">
1355<t>
1356   HTTP header fields, which include general-header (<xref target="general.header.fields"/>),
1357   request-header (&request-header-fields;), response-header (&response-header-fields;), and
1358   entity-header (&entity-header-fields;) fields, follow the same generic format as
1359   that given in <xref target="RFC822" x:fmt="of" x:sec="3.1"/>. Each header field consists
1360   of a name followed by a colon (":") and the field value. Field names
1361   are case-insensitive. The field value &MAY; be preceded by any amount
1362   of LWS, though a single SP is preferred. Header fields can be
1363   extended over multiple lines by preceding each extra line with at
1364   least one SP or HT. Applications ought to follow "common form", where
1365   one is known or indicated, when generating HTTP constructs, since
1366   there might exist some implementations that fail to accept anything
1367   beyond the common forms.
1368</t>
1369<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"/>
1370    message-header = field-name ":" [ field-value ]
1371    field-name     = token
1372    field-value    = *( field-content | LWS )
1373    field-content  = &lt;the OCTETs making up the field-value
1374                     and consisting of either *TEXT or combinations
1375                     of token, separators, and quoted-string&gt;
1376</artwork></figure>
1377<t>
1378   The field-content does not include any leading or trailing LWS:
1379   linear white space occurring before the first non-whitespace
1380   character of the field-value or after the last non-whitespace
1381   character of the field-value. Such leading or trailing LWS &MAY; be
1382   removed without changing the semantics of the field value. Any LWS
1383   that occurs between field-content &MAY; be replaced with a single SP
1384   before interpreting the field value or forwarding the message
1385   downstream.
1386</t>
1387<t>
1388   The order in which header fields with differing field names are
1389   received is not significant. However, it is "good practice" to send
1390   general-header fields first, followed by request-header or response-header
1391   fields, and ending with the entity-header fields.
1392</t>
1393<t>
1394   Multiple message-header fields with the same field-name &MAY; be
1395   present in a message if and only if the entire field-value for that
1396   header field is defined as a comma-separated list [i.e., #(values)].
1397   It &MUST; be possible to combine the multiple header fields into one
1398   "field-name: field-value" pair, without changing the semantics of the
1399   message, by appending each subsequent field-value to the first, each
1400   separated by a comma. The order in which header fields with the same
1401   field-name are received is therefore significant to the
1402   interpretation of the combined field value, and thus a proxy &MUST-NOT;
1403   change the order of these field values when a message is forwarded.
1404</t>
1405</section>
1406
1407<section title="Message Body" anchor="message.body">
1408<t>
1409   The message-body (if any) of an HTTP message is used to carry the
1410   entity-body associated with the request or response. The message-body
1411   differs from the entity-body only when a transfer-coding has been
1412   applied, as indicated by the Transfer-Encoding header field (<xref target="header.transfer-encoding"/>).
1413</t>
1414<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="message-body"/>
1415    message-body = entity-body
1416                 | &lt;entity-body encoded as per Transfer-Encoding&gt;
1417</artwork></figure>
1418<t>
1419   Transfer-Encoding &MUST; be used to indicate any transfer-codings
1420   applied by an application to ensure safe and proper transfer of the
1421   message. Transfer-Encoding is a property of the message, not of the
1422   entity, and thus &MAY; be added or removed by any application along the
1423   request/response chain. (However, <xref target="transfer.codings"/> places restrictions on
1424   when certain transfer-codings may be used.)
1425</t>
1426<t>
1427   The rules for when a message-body is allowed in a message differ for
1428   requests and responses.
1429</t>
1430<t>
1431   The presence of a message-body in a request is signaled by the
1432   inclusion of a Content-Length or Transfer-Encoding header field in
1433   the request's message-headers. A message-body &MUST-NOT; be included in
1434   a request if the specification of the request method (&method;)
1435   does not allow sending an entity-body in requests. A server &SHOULD;
1436   read and forward a message-body on any request; if the request method
1437   does not include defined semantics for an entity-body, then the
1438   message-body &SHOULD; be ignored when handling the request.
1439</t>
1440<t>
1441   For response messages, whether or not a message-body is included with
1442   a message is dependent on both the request method and the response
1443   status code (<xref target="status.code.and.reason.phrase"/>). All responses to the HEAD request method
1444   &MUST-NOT; include a message-body, even though the presence of entity-header
1445   fields might lead one to believe they do. All 1xx
1446   (informational), 204 (no content), and 304 (not modified) responses
1447   &MUST-NOT; include a message-body. All other responses do include a
1448   message-body, although it &MAY; be of zero length.
1449</t>
1450</section>
1451
1452<section title="Message Length" anchor="message.length">
1453<t>
1454   The transfer-length of a message is the length of the message-body as
1455   it appears in the message; that is, after any transfer-codings have
1456   been applied. When a message-body is included with a message, the
1457   transfer-length of that body is determined by one of the following
1458   (in order of precedence):
1459</t>
1460<t>
1461  <list style="numbers">
1462    <x:lt><t>
1463     Any response message which "&MUST-NOT;" include a message-body (such
1464     as the 1xx, 204, and 304 responses and any response to a HEAD
1465     request) is always terminated by the first empty line after the
1466     header fields, regardless of the entity-header fields present in
1467     the message.
1468    </t></x:lt>
1469    <x:lt><t>
1470     If a Transfer-Encoding header field (<xref target="header.transfer-encoding"/>)
1471     is present, then the transfer-length is
1472     defined by use of the "chunked" transfer-coding (<xref target="transfer.codings"/>),
1473     unless the message is terminated by closing the connection.
1474    </t></x:lt>
1475    <x:lt><t>
1476     If a Content-Length header field (<xref target="header.content-length"/>) is present, its
1477     decimal value in OCTETs represents both the entity-length and the
1478     transfer-length. The Content-Length header field &MUST-NOT; be sent
1479     if these two lengths are different (i.e., if a Transfer-Encoding
1480     header field is present). If a message is received with both a
1481     Transfer-Encoding header field and a Content-Length header field,
1482     the latter &MUST; be ignored.
1483    </t></x:lt>
1484    <x:lt><t>
1485     If the message uses the media type "multipart/byteranges", and the
1486     transfer-length is not otherwise specified, then this self-delimiting
1487     media type defines the transfer-length. This media type
1488     &MUST-NOT; be used unless the sender knows that the recipient can parse
1489     it; the presence in a request of a Range header with multiple byte-range
1490     specifiers from a 1.1 client implies that the client can parse
1491     multipart/byteranges responses.
1492    <list style="empty"><t>
1493       A range header might be forwarded by a 1.0 proxy that does not
1494       understand multipart/byteranges; in this case the server &MUST;
1495       delimit the message using methods defined in items 1, 3 or 5 of
1496       this section.
1497    </t></list>
1498    </t></x:lt>
1499    <x:lt><t>
1500     By the server closing the connection. (Closing the connection
1501     cannot be used to indicate the end of a request body, since that
1502     would leave no possibility for the server to send back a response.)
1503    </t></x:lt>
1504  </list>
1505</t>
1506<t>
1507   For compatibility with HTTP/1.0 applications, HTTP/1.1 requests
1508   containing a message-body &MUST; include a valid Content-Length header
1509   field unless the server is known to be HTTP/1.1 compliant. If a
1510   request contains a message-body and a Content-Length is not given,
1511   the server &SHOULD; respond with 400 (bad request) if it cannot
1512   determine the length of the message, or with 411 (length required) if
1513   it wishes to insist on receiving a valid Content-Length.
1514</t>
1515<t>
1516   All HTTP/1.1 applications that receive entities &MUST; accept the
1517   "chunked" transfer-coding (<xref target="transfer.codings"/>), thus allowing this mechanism
1518   to be used for messages when the message length cannot be determined
1519   in advance.
1520</t>
1521<t>
1522   Messages &MUST-NOT; include both a Content-Length header field and a
1523   transfer-coding. If the message does include a
1524   transfer-coding, the Content-Length &MUST; be ignored.
1525</t>
1526<t>
1527   When a Content-Length is given in a message where a message-body is
1528   allowed, its field value &MUST; exactly match the number of OCTETs in
1529   the message-body. HTTP/1.1 user agents &MUST; notify the user when an
1530   invalid length is received and detected.
1531</t>
1532</section>
1533
1534<section title="General Header Fields" anchor="general.header.fields">
1535<t>
1536   There are a few header fields which have general applicability for
1537   both request and response messages, but which do not apply to the
1538   entity being transferred. These header fields apply only to the
1539   message being transmitted.
1540</t>
1541<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="general-header"/>
1542    general-header = Cache-Control            ; &header-cache-control;
1543                   | Connection               ; <xref target="header.connection"/>
1544                   | Date                     ; <xref target="header.date"/>
1545                   | Pragma                   ; &header-pragma;
1546                   | Trailer                  ; <xref target="header.trailer"/>
1547                   | Transfer-Encoding        ; <xref target="header.transfer-encoding"/>
1548                   | Upgrade                  ; <xref target="header.upgrade"/>
1549                   | Via                      ; <xref target="header.via"/>
1550                   | Warning                  ; &header-warning;
1551</artwork></figure>
1552<t>
1553   General-header field names can be extended reliably only in
1554   combination with a change in the protocol version. However, new or
1555   experimental header fields may be given the semantics of general
1556   header fields if all parties in the communication recognize them to
1557   be general-header fields. Unrecognized header fields are treated as
1558   entity-header fields.
1559</t>
1560</section>
1561</section>
1562
1563<section title="Request" anchor="request">
1564<t>
1565   A request message from a client to a server includes, within the
1566   first line of that message, the method to be applied to the resource,
1567   the identifier of the resource, and the protocol version in use.
1568</t>
1569<!--                 Host                      ; should be moved here eventually -->
1570<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="Request"/>
1571     Request       = Request-Line              ; <xref target="request-line"/>
1572                     *(( general-header        ; <xref target="general.header.fields"/>
1573                      | request-header         ; &request-header-fields;
1574                      | entity-header ) CRLF)  ; &entity-header-fields;
1575                     CRLF
1576                     [ message-body ]          ; <xref target="message.body"/>
1577</artwork></figure>
1578
1579<section title="Request-Line" anchor="request-line">
1580<t>
1581   The Request-Line begins with a method token, followed by the
1582   Request-URI and the protocol version, and ending with CRLF. The
1583   elements are separated by SP characters. No CR or LF is allowed
1584   except in the final CRLF sequence.
1585</t>
1586<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="Request-Line"/>
1587     Request-Line   = Method SP Request-URI SP HTTP-Version CRLF
1588</artwork></figure>
1589
1590<section title="Method" anchor="method">
1591<t>
1592   The Method  token indicates the method to be performed on the
1593   resource identified by the Request-URI. The method is case-sensitive.
1594</t>
1595<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="Method"/><iref primary="true" item="Grammar" subitem="extension-method"/>
1596    Method         = token
1597</artwork></figure>
1598</section>
1599
1600<section title="Request-URI" anchor="request-uri">
1601<t>
1602   The Request-URI is a Uniform Resource Identifier (<xref target="uri"/>) and
1603   identifies the resource upon which to apply the request.
1604</t>
1605<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="Request-URI"/>
1606    Request-URI    = "*"
1607                   | absoluteURI
1608                   | ( abs_path [ "?" query ] )
1609                   | authority
1610</artwork></figure>
1611<t>
1612   The four options for Request-URI are dependent on the nature of the
1613   request. The asterisk "*" means that the request does not apply to a
1614   particular resource, but to the server itself, and is only allowed
1615   when the method used does not necessarily apply to a resource. One
1616   example would be
1617</t>
1618<figure><artwork type="example">
1619    OPTIONS * HTTP/1.1
1620</artwork></figure>
1621<t>
1622   The absoluteURI form is &REQUIRED; when the request is being made to a
1623   proxy. The proxy is requested to forward the request or service it
1624   from a valid cache, and return the response. Note that the proxy &MAY;
1625   forward the request on to another proxy or directly to the server
1626   specified by the absoluteURI. In order to avoid request loops, a
1627   proxy &MUST; be able to recognize all of its server names, including
1628   any aliases, local variations, and the numeric IP address. An example
1629   Request-Line would be:
1630</t>
1631<figure><artwork type="example">
1632    GET http://www.example.org/pub/WWW/TheProject.html HTTP/1.1
1633</artwork></figure>
1634<t>
1635   To allow for transition to absoluteURIs in all requests in future
1636   versions of HTTP, all HTTP/1.1 servers &MUST; accept the absoluteURI
1637   form in requests, even though HTTP/1.1 clients will only generate
1638   them in requests to proxies.
1639</t>
1640<t>
1641   The authority form is only used by the CONNECT method (&CONNECT;).
1642</t>
1643<t>
1644   The most common form of Request-URI is that used to identify a
1645   resource on an origin server or gateway. In this case the absolute
1646   path of the URI &MUST; be transmitted (see <xref target="general.syntax"/>, abs_path) as
1647   the Request-URI, and the network location of the URI (authority) &MUST;
1648   be transmitted in a Host header field. For example, a client wishing
1649   to retrieve the resource above directly from the origin server would
1650   create a TCP connection to port 80 of the host "www.example.org" and send
1651   the lines:
1652</t>
1653<figure><artwork type="example">
1654    GET /pub/WWW/TheProject.html HTTP/1.1
1655    Host: www.example.org
1656</artwork></figure>
1657<t>
1658   followed by the remainder of the Request. Note that the absolute path
1659   cannot be empty; if none is present in the original URI, it &MUST; be
1660   given as "/" (the server root).
1661</t>
1662<t>
1663   The Request-URI is transmitted in the format specified in
1664   <xref target="general.syntax"/>. If the Request-URI is encoded using the "% HEX HEX" encoding
1665   <xref target="RFC2396"/>, the origin server &MUST; decode the Request-URI in order to
1666   properly interpret the request. Servers &SHOULD; respond to invalid
1667   Request-URIs with an appropriate status code.
1668</t>
1669<t>
1670   A transparent proxy &MUST-NOT; rewrite the "abs_path" part of the
1671   received Request-URI when forwarding it to the next inbound server,
1672   except as noted above to replace a null abs_path with "/".
1673</t>
1674<t>
1675  <list><t>
1676      <x:h>Note:</x:h> The "no rewrite" rule prevents the proxy from changing the
1677      meaning of the request when the origin server is improperly using
1678      a non-reserved URI character for a reserved purpose.  Implementors
1679      should be aware that some pre-HTTP/1.1 proxies have been known to
1680      rewrite the Request-URI.
1681  </t></list>
1682</t>
1683</section>
1684</section>
1685
1686<section title="The Resource Identified by a Request" anchor="the.resource.identified.by.a.request">
1687<t>
1688   The exact resource identified by an Internet request is determined by
1689   examining both the Request-URI and the Host header field.
1690</t>
1691<t>
1692   An origin server that does not allow resources to differ by the
1693   requested host &MAY; ignore the Host header field value when
1694   determining the resource identified by an HTTP/1.1 request. (But see
1695   <xref target="changes.to.simplify.multi-homed.web.servers.and.conserve.ip.addresses"/>
1696   for other requirements on Host support in HTTP/1.1.)
1697</t>
1698<t>
1699   An origin server that does differentiate resources based on the host
1700   requested (sometimes referred to as virtual hosts or vanity host
1701   names) &MUST; use the following rules for determining the requested
1702   resource on an HTTP/1.1 request:
1703  <list style="numbers">
1704    <t>If Request-URI is an absoluteURI, the host is part of the
1705     Request-URI. Any Host header field value in the request &MUST; be
1706     ignored.</t>
1707    <t>If the Request-URI is not an absoluteURI, and the request includes
1708     a Host header field, the host is determined by the Host header
1709     field value.</t>
1710    <t>If the host as determined by rule 1 or 2 is not a valid host on
1711     the server, the response &MUST; be a 400 (Bad Request) error message.</t>
1712  </list>
1713</t>
1714<t>
1715   Recipients of an HTTP/1.0 request that lacks a Host header field &MAY;
1716   attempt to use heuristics (e.g., examination of the URI path for
1717   something unique to a particular host) in order to determine what
1718   exact resource is being requested.
1719</t>
1720</section>
1721
1722</section>
1723
1724
1725<section title="Response" anchor="response">
1726<t>
1727   After receiving and interpreting a request message, a server responds
1728   with an HTTP response message.
1729</t>
1730<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="Response"/>
1731    Response      = Status-Line               ; <xref target="status-line"/>
1732                    *(( general-header        ; <xref target="general.header.fields"/>
1733                     | response-header        ; &response-header-fields;
1734                     | entity-header ) CRLF)  ; &entity-header-fields;
1735                    CRLF
1736                    [ message-body ]          ; <xref target="message.body"/>
1737</artwork></figure>
1738
1739<section title="Status-Line" anchor="status-line">
1740<t>
1741   The first line of a Response message is the Status-Line, consisting
1742   of the protocol version followed by a numeric status code and its
1743   associated textual phrase, with each element separated by SP
1744   characters. No CR or LF is allowed except in the final CRLF sequence.
1745</t>
1746<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="Status-Line"/>
1747    Status-Line = HTTP-Version SP Status-Code SP Reason-Phrase CRLF
1748</artwork></figure>
1749
1750<section title="Status Code and Reason Phrase" anchor="status.code.and.reason.phrase">
1751<t>
1752   The Status-Code element is a 3-digit integer result code of the
1753   attempt to understand and satisfy the request. These codes are fully
1754   defined in &status-codes;. The Reason-Phrase is intended to give a short
1755   textual description of the Status-Code. The Status-Code is intended
1756   for use by automata and the Reason-Phrase is intended for the human
1757   user. The client is not required to examine or display the Reason-Phrase.
1758</t>
1759<t>
1760   The first digit of the Status-Code defines the class of response. The
1761   last two digits do not have any categorization role. There are 5
1762   values for the first digit:
1763  <list style="symbols">
1764    <t>
1765      1xx: Informational - Request received, continuing process
1766    </t>
1767    <t>
1768      2xx: Success - The action was successfully received,
1769        understood, and accepted
1770    </t>
1771    <t>
1772      3xx: Redirection - Further action must be taken in order to
1773        complete the request
1774    </t>
1775    <t>
1776      4xx: Client Error - The request contains bad syntax or cannot
1777        be fulfilled
1778    </t>
1779    <t>
1780      5xx: Server Error - The server failed to fulfill an apparently
1781        valid request
1782    </t>
1783  </list>
1784</t>
1785<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"/>
1786   Status-Code    = 3DIGIT
1787   Reason-Phrase  = *&lt;TEXT, excluding CR, LF&gt;
1788</artwork></figure>
1789</section>
1790</section>
1791
1792</section>
1793
1794
1795<section title="Connections" anchor="connections">
1796
1797<section title="Persistent Connections" anchor="persistent.connections">
1798
1799<section title="Purpose" anchor="persistent.purpose">
1800<t>
1801   Prior to persistent connections, a separate TCP connection was
1802   established to fetch each URL, increasing the load on HTTP servers
1803   and causing congestion on the Internet. The use of inline images and
1804   other associated data often require a client to make multiple
1805   requests of the same server in a short amount of time. Analysis of
1806   these performance problems and results from a prototype
1807   implementation are available <xref target="Pad1995"/> <xref target="Spe"/>. Implementation experience and
1808   measurements of actual HTTP/1.1 (<xref target="RFC2068" x:fmt="none">RFC 2068</xref>) implementations show good
1809   results <xref target="Nie1997"/>. Alternatives have also been explored, for example,
1810   T/TCP <xref target="Tou1998"/>.
1811</t>
1812<t>
1813   Persistent HTTP connections have a number of advantages:
1814  <list style="symbols">
1815      <t>
1816        By opening and closing fewer TCP connections, CPU time is saved
1817        in routers and hosts (clients, servers, proxies, gateways,
1818        tunnels, or caches), and memory used for TCP protocol control
1819        blocks can be saved in hosts.
1820      </t>
1821      <t>
1822        HTTP requests and responses can be pipelined on a connection.
1823        Pipelining allows a client to make multiple requests without
1824        waiting for each response, allowing a single TCP connection to
1825        be used much more efficiently, with much lower elapsed time.
1826      </t>
1827      <t>
1828        Network congestion is reduced by reducing the number of packets
1829        caused by TCP opens, and by allowing TCP sufficient time to
1830        determine the congestion state of the network.
1831      </t>
1832      <t>
1833        Latency on subsequent requests is reduced since there is no time
1834        spent in TCP's connection opening handshake.
1835      </t>
1836      <t>
1837        HTTP can evolve more gracefully, since errors can be reported
1838        without the penalty of closing the TCP connection. Clients using
1839        future versions of HTTP might optimistically try a new feature,
1840        but if communicating with an older server, retry with old
1841        semantics after an error is reported.
1842      </t>
1843    </list>
1844</t>
1845<t>
1846   HTTP implementations &SHOULD; implement persistent connections.
1847</t>
1848</section>
1849
1850<section title="Overall Operation" anchor="persistent.overall">
1851<t>
1852   A significant difference between HTTP/1.1 and earlier versions of
1853   HTTP is that persistent connections are the default behavior of any
1854   HTTP connection. That is, unless otherwise indicated, the client
1855   &SHOULD; assume that the server will maintain a persistent connection,
1856   even after error responses from the server.
1857</t>
1858<t>
1859   Persistent connections provide a mechanism by which a client and a
1860   server can signal the close of a TCP connection. This signaling takes
1861   place using the Connection header field (<xref target="header.connection"/>). Once a close
1862   has been signaled, the client &MUST-NOT; send any more requests on that
1863   connection.
1864</t>
1865
1866<section title="Negotiation" anchor="persistent.negotiation">
1867<t>
1868   An HTTP/1.1 server &MAY; assume that a HTTP/1.1 client intends to
1869   maintain a persistent connection unless a Connection header including
1870   the connection-token "close" was sent in the request. If the server
1871   chooses to close the connection immediately after sending the
1872   response, it &SHOULD; send a Connection header including the
1873   connection-token close.
1874</t>
1875<t>
1876   An HTTP/1.1 client &MAY; expect a connection to remain open, but would
1877   decide to keep it open based on whether the response from a server
1878   contains a Connection header with the connection-token close. In case
1879   the client does not want to maintain a connection for more than that
1880   request, it &SHOULD; send a Connection header including the
1881   connection-token close.
1882</t>
1883<t>
1884   If either the client or the server sends the close token in the
1885   Connection header, that request becomes the last one for the
1886   connection.
1887</t>
1888<t>
1889   Clients and servers &SHOULD-NOT;  assume that a persistent connection is
1890   maintained for HTTP versions less than 1.1 unless it is explicitly
1891   signaled. See <xref target="compatibility.with.http.1.0.persistent.connections"/> for more information on backward
1892   compatibility with HTTP/1.0 clients.
1893</t>
1894<t>
1895   In order to remain persistent, all messages on the connection &MUST;
1896   have a self-defined message length (i.e., one not defined by closure
1897   of the connection), as described in <xref target="message.length"/>.
1898</t>
1899</section>
1900
1901<section title="Pipelining" anchor="pipelining">
1902<t>
1903   A client that supports persistent connections &MAY; "pipeline" its
1904   requests (i.e., send multiple requests without waiting for each
1905   response). A server &MUST; send its responses to those requests in the
1906   same order that the requests were received.
1907</t>
1908<t>
1909   Clients which assume persistent connections and pipeline immediately
1910   after connection establishment &SHOULD; be prepared to retry their
1911   connection if the first pipelined attempt fails. If a client does
1912   such a retry, it &MUST-NOT; pipeline before it knows the connection is
1913   persistent. Clients &MUST; also be prepared to resend their requests if
1914   the server closes the connection before sending all of the
1915   corresponding responses.
1916</t>
1917<t>
1918   Clients &SHOULD-NOT;  pipeline requests using non-idempotent methods or
1919   non-idempotent sequences of methods (see &idempotent-methods;). Otherwise, a
1920   premature termination of the transport connection could lead to
1921   indeterminate results. A client wishing to send a non-idempotent
1922   request &SHOULD; wait to send that request until it has received the
1923   response status for the previous request.
1924</t>
1925</section>
1926</section>
1927
1928<section title="Proxy Servers" anchor="persistent.proxy">
1929<t>
1930   It is especially important that proxies correctly implement the
1931   properties of the Connection header field as specified in <xref target="header.connection"/>.
1932</t>
1933<t>
1934   The proxy server &MUST; signal persistent connections separately with
1935   its clients and the origin servers (or other proxy servers) that it
1936   connects to. Each persistent connection applies to only one transport
1937   link.
1938</t>
1939<t>
1940   A proxy server &MUST-NOT; establish a HTTP/1.1 persistent connection
1941   with an HTTP/1.0 client (but see <xref target="RFC2068"/> for information and
1942   discussion of the problems with the Keep-Alive header implemented by
1943   many HTTP/1.0 clients).
1944</t>
1945</section>
1946
1947<section title="Practical Considerations" anchor="persistent.practical">
1948<t>
1949   Servers will usually have some time-out value beyond which they will
1950   no longer maintain an inactive connection. Proxy servers might make
1951   this a higher value since it is likely that the client will be making
1952   more connections through the same server. The use of persistent
1953   connections places no requirements on the length (or existence) of
1954   this time-out for either the client or the server.
1955</t>
1956<t>
1957   When a client or server wishes to time-out it &SHOULD; issue a graceful
1958   close on the transport connection. Clients and servers &SHOULD; both
1959   constantly watch for the other side of the transport close, and
1960   respond to it as appropriate. If a client or server does not detect
1961   the other side's close promptly it could cause unnecessary resource
1962   drain on the network.
1963</t>
1964<t>
1965   A client, server, or proxy &MAY; close the transport connection at any
1966   time. For example, a client might have started to send a new request
1967   at the same time that the server has decided to close the "idle"
1968   connection. From the server's point of view, the connection is being
1969   closed while it was idle, but from the client's point of view, a
1970   request is in progress.
1971</t>
1972<t>
1973   This means that clients, servers, and proxies &MUST; be able to recover
1974   from asynchronous close events. Client software &SHOULD; reopen the
1975   transport connection and retransmit the aborted sequence of requests
1976   without user interaction so long as the request sequence is
1977   idempotent (see &idempotent-methods;). Non-idempotent methods or sequences
1978   &MUST-NOT; be automatically retried, although user agents &MAY; offer a
1979   human operator the choice of retrying the request(s). Confirmation by
1980   user-agent software with semantic understanding of the application
1981   &MAY; substitute for user confirmation. The automatic retry &SHOULD-NOT; 
1982   be repeated if the second sequence of requests fails.
1983</t>
1984<t>
1985   Servers &SHOULD; always respond to at least one request per connection,
1986   if at all possible. Servers &SHOULD-NOT;  close a connection in the
1987   middle of transmitting a response, unless a network or client failure
1988   is suspected.
1989</t>
1990<t>
1991   Clients that use persistent connections &SHOULD; limit the number of
1992   simultaneous connections that they maintain to a given server. A
1993   single-user client &SHOULD-NOT; maintain more than 2 connections with
1994   any server or proxy. A proxy &SHOULD; use up to 2*N connections to
1995   another server or proxy, where N is the number of simultaneously
1996   active users. These guidelines are intended to improve HTTP response
1997   times and avoid congestion.
1998</t>
1999</section>
2000</section>
2001
2002<section title="Message Transmission Requirements" anchor="message.transmission.requirements">
2003
2004<section title="Persistent Connections and Flow Control" anchor="persistent.flow">
2005<t>
2006   HTTP/1.1 servers &SHOULD; maintain persistent connections and use TCP's
2007   flow control mechanisms to resolve temporary overloads, rather than
2008   terminating connections with the expectation that clients will retry.
2009   The latter technique can exacerbate network congestion.
2010</t>
2011</section>
2012
2013<section title="Monitoring Connections for Error Status Messages" anchor="persistent.monitor">
2014<t>
2015   An HTTP/1.1 (or later) client sending a message-body &SHOULD; monitor
2016   the network connection for an error status while it is transmitting
2017   the request. If the client sees an error status, it &SHOULD;
2018   immediately cease transmitting the body. If the body is being sent
2019   using a "chunked" encoding (<xref target="transfer.codings"/>), a zero length chunk and
2020   empty trailer &MAY; be used to prematurely mark the end of the message.
2021   If the body was preceded by a Content-Length header, the client &MUST;
2022   close the connection.
2023</t>
2024</section>
2025
2026<section title="Use of the 100 (Continue) Status" anchor="use.of.the.100.status">
2027<t>
2028   The purpose of the 100 (Continue) status (see &status-100;) is to
2029   allow a client that is sending a request message with a request body
2030   to determine if the origin server is willing to accept the request
2031   (based on the request headers) before the client sends the request
2032   body. In some cases, it might either be inappropriate or highly
2033   inefficient for the client to send the body if the server will reject
2034   the message without looking at the body.
2035</t>
2036<t>
2037   Requirements for HTTP/1.1 clients:
2038  <list style="symbols">
2039    <t>
2040        If a client will wait for a 100 (Continue) response before
2041        sending the request body, it &MUST; send an Expect request-header
2042        field (&header-expect;) with the "100-continue" expectation.
2043    </t>
2044    <t>
2045        A client &MUST-NOT; send an Expect request-header field (&header-expect;)
2046        with the "100-continue" expectation if it does not intend
2047        to send a request body.
2048    </t>
2049  </list>
2050</t>
2051<t>
2052   Because of the presence of older implementations, the protocol allows
2053   ambiguous situations in which a client may send "Expect: 100-continue"
2054   without receiving either a 417 (Expectation Failed) status
2055   or a 100 (Continue) status. Therefore, when a client sends this
2056   header field to an origin server (possibly via a proxy) from which it
2057   has never seen a 100 (Continue) status, the client &SHOULD-NOT;  wait
2058   for an indefinite period before sending the request body.
2059</t>
2060<t>
2061   Requirements for HTTP/1.1 origin servers:
2062  <list style="symbols">
2063    <t> Upon receiving a request which includes an Expect request-header
2064        field with the "100-continue" expectation, an origin server &MUST;
2065        either respond with 100 (Continue) status and continue to read
2066        from the input stream, or respond with a final status code. The
2067        origin server &MUST-NOT; wait for the request body before sending
2068        the 100 (Continue) response. If it responds with a final status
2069        code, it &MAY; close the transport connection or it &MAY; continue
2070        to read and discard the rest of the request.  It &MUST-NOT;
2071        perform the requested method if it returns a final status code.
2072    </t>
2073    <t> An origin server &SHOULD-NOT;  send a 100 (Continue) response if
2074        the request message does not include an Expect request-header
2075        field with the "100-continue" expectation, and &MUST-NOT; send a
2076        100 (Continue) response if such a request comes from an HTTP/1.0
2077        (or earlier) client. There is an exception to this rule: for
2078        compatibility with <xref target="RFC2068"/>, a server &MAY; send a 100 (Continue)
2079        status in response to an HTTP/1.1 PUT or POST request that does
2080        not include an Expect request-header field with the "100-continue"
2081        expectation. This exception, the purpose of which is
2082        to minimize any client processing delays associated with an
2083        undeclared wait for 100 (Continue) status, applies only to
2084        HTTP/1.1 requests, and not to requests with any other HTTP-version
2085        value.
2086    </t>
2087    <t> An origin server &MAY; omit a 100 (Continue) response if it has
2088        already received some or all of the request body for the
2089        corresponding request.
2090    </t>
2091    <t> An origin server that sends a 100 (Continue) response &MUST;
2092    ultimately send a final status code, once the request body is
2093        received and processed, unless it terminates the transport
2094        connection prematurely.
2095    </t>
2096    <t> If an origin server receives a request that does not include an
2097        Expect request-header field with the "100-continue" expectation,
2098        the request includes a request body, and the server responds
2099        with a final status code before reading the entire request body
2100        from the transport connection, then the server &SHOULD-NOT;  close
2101        the transport connection until it has read the entire request,
2102        or until the client closes the connection. Otherwise, the client
2103        might not reliably receive the response message. However, this
2104        requirement is not be construed as preventing a server from
2105        defending itself against denial-of-service attacks, or from
2106        badly broken client implementations.
2107      </t>
2108    </list>
2109</t>
2110<t>
2111   Requirements for HTTP/1.1 proxies:
2112  <list style="symbols">
2113    <t> If a proxy receives a request that includes an Expect request-header
2114        field with the "100-continue" expectation, and the proxy
2115        either knows that the next-hop server complies with HTTP/1.1 or
2116        higher, or does not know the HTTP version of the next-hop
2117        server, it &MUST; forward the request, including the Expect header
2118        field.
2119    </t>
2120    <t> If the proxy knows that the version of the next-hop server is
2121        HTTP/1.0 or lower, it &MUST-NOT; forward the request, and it &MUST;
2122        respond with a 417 (Expectation Failed) status.
2123    </t>
2124    <t> Proxies &SHOULD; maintain a cache recording the HTTP version
2125        numbers received from recently-referenced next-hop servers.
2126    </t>
2127    <t> A proxy &MUST-NOT; forward a 100 (Continue) response if the
2128        request message was received from an HTTP/1.0 (or earlier)
2129        client and did not include an Expect request-header field with
2130        the "100-continue" expectation. This requirement overrides the
2131        general rule for forwarding of 1xx responses (see &status-1xx;).
2132    </t>
2133  </list>
2134</t>
2135</section>
2136
2137<section title="Client Behavior if Server Prematurely Closes Connection" anchor="connection.premature">
2138<t>
2139   If an HTTP/1.1 client sends a request which includes a request body,
2140   but which does not include an Expect request-header field with the
2141   "100-continue" expectation, and if the client is not directly
2142   connected to an HTTP/1.1 origin server, and if the client sees the
2143   connection close before receiving any status from the server, the
2144   client &SHOULD; retry the request.  If the client does retry this
2145   request, it &MAY; use the following "binary exponential backoff"
2146   algorithm to be assured of obtaining a reliable response:
2147  <list style="numbers">
2148    <t>
2149      Initiate a new connection to the server
2150    </t>
2151    <t>
2152      Transmit the request-headers
2153    </t>
2154    <t>
2155      Initialize a variable R to the estimated round-trip time to the
2156         server (e.g., based on the time it took to establish the
2157         connection), or to a constant value of 5 seconds if the round-trip
2158         time is not available.
2159    </t>
2160    <t>
2161       Compute T = R * (2**N), where N is the number of previous
2162         retries of this request.
2163    </t>
2164    <t>
2165       Wait either for an error response from the server, or for T
2166         seconds (whichever comes first)
2167    </t>
2168    <t>
2169       If no error response is received, after T seconds transmit the
2170         body of the request.
2171    </t>
2172    <t>
2173       If client sees that the connection is closed prematurely,
2174         repeat from step 1 until the request is accepted, an error
2175         response is received, or the user becomes impatient and
2176         terminates the retry process.
2177    </t>
2178  </list>
2179</t>
2180<t>
2181   If at any point an error status is received, the client
2182  <list style="symbols">
2183      <t>&SHOULD-NOT;  continue and</t>
2184
2185      <t>&SHOULD; close the connection if it has not completed sending the
2186        request message.</t>
2187    </list>
2188</t>
2189</section>
2190</section>
2191</section>
2192
2193
2194<section title="Header Field Definitions" anchor="header.fields">
2195<t>
2196   This section defines the syntax and semantics of all standard
2197   HTTP/1.1 header fields. For entity-header fields, both sender and
2198   recipient refer to either the client or the server, depending on who
2199   sends and who receives the entity.
2200</t>
2201
2202<section title="Connection" anchor="header.connection">
2203  <iref primary="true" item="Connection header" x:for-anchor=""/>
2204  <iref primary="true" item="Headers" subitem="Connection" x:for-anchor=""/>
2205<t>
2206   The Connection general-header field allows the sender to specify
2207   options that are desired for that particular connection and &MUST-NOT;
2208   be communicated by proxies over further connections.
2209</t>
2210<t>
2211   The Connection header has the following grammar:
2212</t>
2213<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="Connection"/><iref primary="true" item="Grammar" subitem="connection-token"/>
2214    Connection = "Connection" ":" 1#(connection-token)
2215    connection-token  = token
2216</artwork></figure>
2217<t>
2218   HTTP/1.1 proxies &MUST; parse the Connection header field before a
2219   message is forwarded and, for each connection-token in this field,
2220   remove any header field(s) from the message with the same name as the
2221   connection-token. Connection options are signaled by the presence of
2222   a connection-token in the Connection header field, not by any
2223   corresponding additional header field(s), since the additional header
2224   field may not be sent if there are no parameters associated with that
2225   connection option.
2226</t>
2227<t>
2228   Message headers listed in the Connection header &MUST-NOT; include
2229   end-to-end headers, such as Cache-Control.
2230</t>
2231<t>
2232   HTTP/1.1 defines the "close" connection option for the sender to
2233   signal that the connection will be closed after completion of the
2234   response. For example,
2235</t>
2236<figure><artwork type="example">
2237    Connection: close
2238</artwork></figure>
2239<t>
2240   in either the request or the response header fields indicates that
2241   the connection &SHOULD-NOT;  be considered `persistent' (<xref target="persistent.connections"/>)
2242   after the current request/response is complete.
2243</t>
2244<t>
2245   An HTTP/1.1 client that does not support persistent connections &MUST;
2246   include the "close" connection option in every request message.
2247</t>
2248<t>
2249   An HTTP/1.1 server that does not support persistent connections &MUST;
2250   include the "close" connection option in every response message that
2251   does not have a 1xx (informational) status code.
2252</t>
2253<t>
2254   A system receiving an HTTP/1.0 (or lower-version) message that
2255   includes a Connection header &MUST;, for each connection-token in this
2256   field, remove and ignore any header field(s) from the message with
2257   the same name as the connection-token. This protects against mistaken
2258   forwarding of such header fields by pre-HTTP/1.1 proxies. See <xref target="compatibility.with.http.1.0.persistent.connections"/>.
2259</t>
2260</section>
2261
2262<section title="Content-Length" anchor="header.content-length">
2263  <iref primary="true" item="Content-Length header" x:for-anchor=""/>
2264  <iref primary="true" item="Headers" subitem="Content-Length" x:for-anchor=""/>
2265<t>
2266   The Content-Length entity-header field indicates the size of the
2267   entity-body, in decimal number of OCTETs, sent to the recipient or,
2268   in the case of the HEAD method, the size of the entity-body that
2269   would have been sent had the request been a GET.
2270</t>
2271<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="Content-Length"/>
2272    Content-Length    = "Content-Length" ":" 1*DIGIT
2273</artwork></figure>
2274<t>
2275   An example is
2276</t>
2277<figure><artwork type="example">
2278    Content-Length: 3495
2279</artwork></figure>
2280<t>
2281   Applications &SHOULD; use this field to indicate the transfer-length of
2282   the message-body, unless this is prohibited by the rules in <xref target="message.length"/>.
2283</t>
2284<t>
2285   Any Content-Length greater than or equal to zero is a valid value.
2286   <xref target="message.length"/> describes how to determine the length of a message-body
2287   if a Content-Length is not given.
2288</t>
2289<t>
2290   Note that the meaning of this field is significantly different from
2291   the corresponding definition in MIME, where it is an optional field
2292   used within the "message/external-body" content-type. In HTTP, it
2293   &SHOULD; be sent whenever the message's length can be determined prior
2294   to being transferred, unless this is prohibited by the rules in
2295   <xref target="message.length"/>.
2296</t>
2297</section>
2298
2299<section title="Date" anchor="header.date">
2300  <iref primary="true" item="Date header" x:for-anchor=""/>
2301  <iref primary="true" item="Headers" subitem="Date" x:for-anchor=""/>
2302<t>
2303   The Date general-header field represents the date and time at which
2304   the message was originated, having the same semantics as orig-date in
2305   <xref target="RFC822"/>. The field value is an HTTP-date, as described in <xref target="full.date"/>;
2306   it &MUST; be sent in rfc1123-date format.
2307</t>
2308<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="Date"/>
2309    Date  = "Date" ":" HTTP-date
2310</artwork></figure>
2311<t>
2312   An example is
2313</t>
2314<figure><artwork type="example">
2315    Date: Tue, 15 Nov 1994 08:12:31 GMT
2316</artwork></figure>
2317<t>
2318   Origin servers &MUST; include a Date header field in all responses,
2319   except in these cases:
2320  <list style="numbers">
2321      <t>If the response status code is 100 (Continue) or 101 (Switching
2322         Protocols), the response &MAY; include a Date header field, at
2323         the server's option.</t>
2324
2325      <t>If the response status code conveys a server error, e.g. 500
2326         (Internal Server Error) or 503 (Service Unavailable), and it is
2327         inconvenient or impossible to generate a valid Date.</t>
2328
2329      <t>If the server does not have a clock that can provide a
2330         reasonable approximation of the current time, its responses
2331         &MUST-NOT; include a Date header field. In this case, the rules
2332         in <xref target="clockless.origin.server.operation"/> &MUST; be followed.</t>
2333  </list>
2334</t>
2335<t>
2336   A received message that does not have a Date header field &MUST; be
2337   assigned one by the recipient if the message will be cached by that
2338   recipient or gatewayed via a protocol which requires a Date. An HTTP
2339   implementation without a clock &MUST-NOT; cache responses without
2340   revalidating them on every use. An HTTP cache, especially a shared
2341   cache, &SHOULD; use a mechanism, such as NTP <xref target="RFC1305"/>, to synchronize its
2342   clock with a reliable external standard.
2343</t>
2344<t>
2345   Clients &SHOULD; only send a Date header field in messages that include
2346   an entity-body, as in the case of the PUT and POST requests, and even
2347   then it is optional. A client without a clock &MUST-NOT; send a Date
2348   header field in a request.
2349</t>
2350<t>
2351   The HTTP-date sent in a Date header &SHOULD-NOT;  represent a date and
2352   time subsequent to the generation of the message. It &SHOULD; represent
2353   the best available approximation of the date and time of message
2354   generation, unless the implementation has no means of generating a
2355   reasonably accurate date and time. In theory, the date ought to
2356   represent the moment just before the entity is generated. In
2357   practice, the date can be generated at any time during the message
2358   origination without affecting its semantic value.
2359</t>
2360
2361<section title="Clockless Origin Server Operation" anchor="clockless.origin.server.operation">
2362<t>
2363   Some origin server implementations might not have a clock available.
2364   An origin server without a clock &MUST-NOT; assign Expires or Last-Modified
2365   values to a response, unless these values were associated
2366   with the resource by a system or user with a reliable clock. It &MAY;
2367   assign an Expires value that is known, at or before server
2368   configuration time, to be in the past (this allows "pre-expiration"
2369   of responses without storing separate Expires values for each
2370   resource).
2371</t>
2372</section>
2373</section>
2374
2375<section title="Host" anchor="header.host">
2376  <iref primary="true" item="Host header" x:for-anchor=""/>
2377  <iref primary="true" item="Headers" subitem="Host" x:for-anchor=""/>
2378<t>
2379   The Host request-header field specifies the Internet host and port
2380   number of the resource being requested, as obtained from the original
2381   URI given by the user or referring resource (generally an HTTP URL,
2382   as described in <xref target="http.url"/>). The Host field value &MUST; represent
2383   the naming authority of the origin server or gateway given by the
2384   original URL. This allows the origin server or gateway to
2385   differentiate between internally-ambiguous URLs, such as the root "/"
2386   URL of a server for multiple host names on a single IP address.
2387</t>
2388<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="Host"/>
2389    Host = "Host" ":" host [ ":" port ] ; <xref target="http.url"/>
2390</artwork></figure>
2391<t>
2392   A "host" without any trailing port information implies the default
2393   port for the service requested (e.g., "80" for an HTTP URL). For
2394   example, a request on the origin server for
2395   &lt;http://www.example.org/pub/WWW/&gt; would properly include:
2396</t>
2397<figure><artwork type="example">
2398    GET /pub/WWW/ HTTP/1.1
2399    Host: www.example.org
2400</artwork></figure>
2401<t>
2402   A client &MUST; include a Host header field in all HTTP/1.1 request
2403   messages . If the requested URI does not include an Internet host
2404   name for the service being requested, then the Host header field &MUST;
2405   be given with an empty value. An HTTP/1.1 proxy &MUST; ensure that any
2406   request message it forwards does contain an appropriate Host header
2407   field that identifies the service being requested by the proxy. All
2408   Internet-based HTTP/1.1 servers &MUST; respond with a 400 (Bad Request)
2409   status code to any HTTP/1.1 request message which lacks a Host header
2410   field.
2411</t>
2412<t>
2413   See Sections <xref target="the.resource.identified.by.a.request" format="counter"/>
2414   and <xref target="changes.to.simplify.multi-homed.web.servers.and.conserve.ip.addresses" format="counter"/>
2415   for other requirements relating to Host.
2416</t>
2417</section>
2418
2419<section title="TE" anchor="header.te">
2420  <iref primary="true" item="TE header" x:for-anchor=""/>
2421  <iref primary="true" item="Headers" subitem="TE" x:for-anchor=""/>
2422<t>
2423   The TE request-header field indicates what extension transfer-codings
2424   it is willing to accept in the response and whether or not it is
2425   willing to accept trailer fields in a chunked transfer-coding. Its
2426   value may consist of the keyword "trailers" and/or a comma-separated
2427   list of extension transfer-coding names with optional accept
2428   parameters (as described in <xref target="transfer.codings"/>).
2429</t>
2430<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="TE"/><iref primary="true" item="Grammar" subitem="t-codings"/>
2431    TE        = "TE" ":" #( t-codings )
2432    t-codings = "trailers" | ( transfer-extension [ accept-params ] )
2433</artwork></figure>
2434<t>
2435   The presence of the keyword "trailers" indicates that the client is
2436   willing to accept trailer fields in a chunked transfer-coding, as
2437   defined in <xref target="chunked.transfer.encoding"/>. This keyword is reserved for use with
2438   transfer-coding values even though it does not itself represent a
2439   transfer-coding.
2440</t>
2441<t>
2442   Examples of its use are:
2443</t>
2444<figure><artwork type="example">
2445    TE: deflate
2446    TE:
2447    TE: trailers, deflate;q=0.5
2448</artwork></figure>
2449<t>
2450   The TE header field only applies to the immediate connection.
2451   Therefore, the keyword &MUST; be supplied within a Connection header
2452   field (<xref target="header.connection"/>) whenever TE is present in an HTTP/1.1 message.
2453</t>
2454<t>
2455   A server tests whether a transfer-coding is acceptable, according to
2456   a TE field, using these rules:
2457  <list style="numbers">
2458    <x:lt>
2459      <t>The "chunked" transfer-coding is always acceptable. If the
2460         keyword "trailers" is listed, the client indicates that it is
2461         willing to accept trailer fields in the chunked response on
2462         behalf of itself and any downstream clients. The implication is
2463         that, if given, the client is stating that either all
2464         downstream clients are willing to accept trailer fields in the
2465         forwarded response, or that it will attempt to buffer the
2466         response on behalf of downstream recipients.
2467      </t><t>
2468         <x:h>Note:</x:h> HTTP/1.1 does not define any means to limit the size of a
2469         chunked response such that a client can be assured of buffering
2470         the entire response.</t>
2471    </x:lt>
2472    <x:lt>
2473      <t>If the transfer-coding being tested is one of the transfer-codings
2474         listed in the TE field, then it is acceptable unless it
2475         is accompanied by a qvalue of 0. (As defined in &qvalue;, a
2476         qvalue of 0 means "not acceptable.")</t>
2477    </x:lt>
2478    <x:lt>
2479      <t>If multiple transfer-codings are acceptable, then the
2480         acceptable transfer-coding with the highest non-zero qvalue is
2481         preferred.  The "chunked" transfer-coding always has a qvalue
2482         of 1.</t>
2483    </x:lt>
2484  </list>
2485</t>
2486<t>
2487   If the TE field-value is empty or if no TE field is present, the only
2488   transfer-coding  is "chunked". A message with no transfer-coding is
2489   always acceptable.
2490</t>
2491</section>
2492
2493<section title="Trailer" anchor="header.trailer">
2494  <iref primary="true" item="Trailer header" x:for-anchor=""/>
2495  <iref primary="true" item="Headers" subitem="Trailer" x:for-anchor=""/>
2496<t>
2497   The Trailer general field value indicates that the given set of
2498   header fields is present in the trailer of a message encoded with
2499   chunked transfer-coding.
2500</t>
2501<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="Trailer"/>
2502    Trailer  = "Trailer" ":" 1#field-name
2503</artwork></figure>
2504<t>
2505   An HTTP/1.1 message &SHOULD; include a Trailer header field in a
2506   message using chunked transfer-coding with a non-empty trailer. Doing
2507   so allows the recipient to know which header fields to expect in the
2508   trailer.
2509</t>
2510<t>
2511   If no Trailer header field is present, the trailer &SHOULD-NOT;  include
2512   any header fields. See <xref target="chunked.transfer.encoding"/> for restrictions on the use of
2513   trailer fields in a "chunked" transfer-coding.
2514</t>
2515<t>
2516   Message header fields listed in the Trailer header field &MUST-NOT;
2517   include the following header fields:
2518  <list style="symbols">
2519    <t>Transfer-Encoding</t>
2520    <t>Content-Length</t>
2521    <t>Trailer</t>
2522  </list>
2523</t>
2524</section>
2525
2526<section title="Transfer-Encoding" anchor="header.transfer-encoding">
2527  <iref primary="true" item="Transfer-Encoding header" x:for-anchor=""/>
2528  <iref primary="true" item="Headers" subitem="Transfer-Encoding" x:for-anchor=""/>
2529<t>
2530   The Transfer-Encoding general-header field indicates what (if any)
2531   type of transformation has been applied to the message body in order
2532   to safely transfer it between the sender and the recipient. This
2533   differs from the content-coding in that the transfer-coding is a
2534   property of the message, not of the entity.
2535</t>
2536<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="Transfer-Encoding"/>
2537  Transfer-Encoding       = "Transfer-Encoding" ":" 1#transfer-coding
2538</artwork></figure>
2539<t>
2540   Transfer-codings are defined in <xref target="transfer.codings"/>. An example is:
2541</t>
2542<figure><artwork type="example">
2543  Transfer-Encoding: chunked
2544</artwork></figure>
2545<t>
2546   If multiple encodings have been applied to an entity, the transfer-codings
2547   &MUST; be listed in the order in which they were applied.
2548   Additional information about the encoding parameters &MAY; be provided
2549   by other entity-header fields not defined by this specification.
2550</t>
2551<t>
2552   Many older HTTP/1.0 applications do not understand the Transfer-Encoding
2553   header.
2554</t>
2555</section>
2556
2557<section title="Upgrade" anchor="header.upgrade">
2558  <iref primary="true" item="Upgrade header" x:for-anchor=""/>
2559  <iref primary="true" item="Headers" subitem="Upgrade" x:for-anchor=""/>
2560<t>
2561   The Upgrade general-header allows the client to specify what
2562   additional communication protocols it supports and would like to use
2563   if the server finds it appropriate to switch protocols. The server
2564   &MUST; use the Upgrade header field within a 101 (Switching Protocols)
2565   response to indicate which protocol(s) are being switched.
2566</t>
2567<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="Upgrade"/>
2568    Upgrade        = "Upgrade" ":" 1#product
2569</artwork></figure>
2570<t>
2571   For example,
2572</t>
2573<figure><artwork type="example">
2574    Upgrade: HTTP/2.0, SHTTP/1.3, IRC/6.9, RTA/x11
2575</artwork></figure>
2576<t>
2577   The Upgrade header field is intended to provide a simple mechanism
2578   for transition from HTTP/1.1 to some other, incompatible protocol. It
2579   does so by allowing the client to advertise its desire to use another
2580   protocol, such as a later version of HTTP with a higher major version
2581   number, even though the current request has been made using HTTP/1.1.
2582   This eases the difficult transition between incompatible protocols by
2583   allowing the client to initiate a request in the more commonly
2584   supported protocol while indicating to the server that it would like
2585   to use a "better" protocol if available (where "better" is determined
2586   by the server, possibly according to the nature of the method and/or
2587   resource being requested).
2588</t>
2589<t>
2590   The Upgrade header field only applies to switching application-layer
2591   protocols upon the existing transport-layer connection. Upgrade
2592   cannot be used to insist on a protocol change; its acceptance and use
2593   by the server is optional. The capabilities and nature of the
2594   application-layer communication after the protocol change is entirely
2595   dependent upon the new protocol chosen, although the first action
2596   after changing the protocol &MUST; be a response to the initial HTTP
2597   request containing the Upgrade header field.
2598</t>
2599<t>
2600   The Upgrade header field only applies to the immediate connection.
2601   Therefore, the upgrade keyword &MUST; be supplied within a Connection
2602   header field (<xref target="header.connection"/>) whenever Upgrade is present in an
2603   HTTP/1.1 message.
2604</t>
2605<t>
2606   The Upgrade header field cannot be used to indicate a switch to a
2607   protocol on a different connection. For that purpose, it is more
2608   appropriate to use a 301, 302, 303, or 305 redirection response.
2609</t>
2610<t>
2611   This specification only defines the protocol name "HTTP" for use by
2612   the family of Hypertext Transfer Protocols, as defined by the HTTP
2613   version rules of <xref target="http.version"/> and future updates to this
2614   specification. Any token can be used as a protocol name; however, it
2615   will only be useful if both the client and server associate the name
2616   with the same protocol.
2617</t>
2618</section>
2619
2620<section title="Via" anchor="header.via">
2621  <iref primary="true" item="Via header" x:for-anchor=""/>
2622  <iref primary="true" item="Headers" subitem="Via" x:for-anchor=""/>
2623<t>
2624   The Via general-header field &MUST; be used by gateways and proxies to
2625   indicate the intermediate protocols and recipients between the user
2626   agent and the server on requests, and between the origin server and
2627   the client on responses. It is analogous to the "Received" field of
2628   <xref target="RFC822"/> and is intended to be used for tracking message forwards,
2629   avoiding request loops, and identifying the protocol capabilities of
2630   all senders along the request/response chain.
2631</t>
2632<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"/>
2633   Via =  "Via" ":" 1#( received-protocol received-by [ comment ] )
2634   received-protocol = [ protocol-name "/" ] protocol-version
2635   protocol-name     = token
2636   protocol-version  = token
2637   received-by       = ( host [ ":" port ] ) | pseudonym
2638   pseudonym         = token
2639</artwork></figure>
2640<t>
2641   The received-protocol indicates the protocol version of the message
2642   received by the server or client along each segment of the
2643   request/response chain. The received-protocol version is appended to
2644   the Via field value when the message is forwarded so that information
2645   about the protocol capabilities of upstream applications remains
2646   visible to all recipients.
2647</t>
2648<t>
2649   The protocol-name is optional if and only if it would be "HTTP". The
2650   received-by field is normally the host and optional port number of a
2651   recipient server or client that subsequently forwarded the message.
2652   However, if the real host is considered to be sensitive information,
2653   it &MAY; be replaced by a pseudonym. If the port is not given, it &MAY;
2654   be assumed to be the default port of the received-protocol.
2655</t>
2656<t>
2657   Multiple Via field values represents each proxy or gateway that has
2658   forwarded the message. Each recipient &MUST; append its information
2659   such that the end result is ordered according to the sequence of
2660   forwarding applications.
2661</t>
2662<t>
2663   Comments &MAY; be used in the Via header field to identify the software
2664   of the recipient proxy or gateway, analogous to the User-Agent and
2665   Server header fields. However, all comments in the Via field are
2666   optional and &MAY; be removed by any recipient prior to forwarding the
2667   message.
2668</t>
2669<t>
2670   For example, a request message could be sent from an HTTP/1.0 user
2671   agent to an internal proxy code-named "fred", which uses HTTP/1.1 to
2672   forward the request to a public proxy at p.example.net, which completes
2673   the request by forwarding it to the origin server at www.example.com.
2674   The request received by www.example.com would then have the following
2675   Via header field:
2676</t>
2677<figure><artwork type="example">
2678    Via: 1.0 fred, 1.1 p.example.net (Apache/1.1)
2679</artwork></figure>
2680<t>
2681   Proxies and gateways used as a portal through a network firewall
2682   &SHOULD-NOT;, by default, forward the names and ports of hosts within
2683   the firewall region. This information &SHOULD; only be propagated if
2684   explicitly enabled. If not enabled, the received-by host of any host
2685   behind the firewall &SHOULD; be replaced by an appropriate pseudonym
2686   for that host.
2687</t>
2688<t>
2689   For organizations that have strong privacy requirements for hiding
2690   internal structures, a proxy &MAY; combine an ordered subsequence of
2691   Via header field entries with identical received-protocol values into
2692   a single such entry. For example,
2693</t>
2694<figure><artwork type="example">
2695    Via: 1.0 ricky, 1.1 ethel, 1.1 fred, 1.0 lucy
2696</artwork></figure>
2697<t>
2698        could be collapsed to
2699</t>
2700<figure><artwork type="example">
2701    Via: 1.0 ricky, 1.1 mertz, 1.0 lucy
2702</artwork></figure>
2703<t>
2704   Applications &SHOULD-NOT;  combine multiple entries unless they are all
2705   under the same organizational control and the hosts have already been
2706   replaced by pseudonyms. Applications &MUST-NOT; combine entries which
2707   have different received-protocol values.
2708</t>
2709</section>
2710
2711</section>
2712
2713<section title="IANA Considerations" anchor="IANA.considerations">
2714<t>
2715   TBD.
2716</t>
2717</section>
2718
2719<section title="Security Considerations" anchor="security.considerations">
2720<t>
2721   This section is meant to inform application developers, information
2722   providers, and users of the security limitations in HTTP/1.1 as
2723   described by this document. The discussion does not include
2724   definitive solutions to the problems revealed, though it does make
2725   some suggestions for reducing security risks.
2726</t>
2727
2728<section title="Personal Information" anchor="personal.information">
2729<t>
2730   HTTP clients are often privy to large amounts of personal information
2731   (e.g. the user's name, location, mail address, passwords, encryption
2732   keys, etc.), and &SHOULD; be very careful to prevent unintentional
2733   leakage of this information via the HTTP protocol to other sources.
2734   We very strongly recommend that a convenient interface be provided
2735   for the user to control dissemination of such information, and that
2736   designers and implementors be particularly careful in this area.
2737   History shows that errors in this area often create serious security
2738   and/or privacy problems and generate highly adverse publicity for the
2739   implementor's company.
2740</t>
2741</section>
2742
2743<section title="Abuse of Server Log Information" anchor="abuse.of.server.log.information">
2744<t>
2745   A server is in the position to save personal data about a user's
2746   requests which might identify their reading patterns or subjects of
2747   interest. This information is clearly confidential in nature and its
2748   handling can be constrained by law in certain countries. People using
2749   the HTTP protocol to provide data are responsible for ensuring that
2750   such material is not distributed without the permission of any
2751   individuals that are identifiable by the published results.
2752</t>
2753</section>
2754
2755<section title="Attacks Based On File and Path Names" anchor="attack.pathname">
2756<t>
2757   Implementations of HTTP origin servers &SHOULD; be careful to restrict
2758   the documents returned by HTTP requests to be only those that were
2759   intended by the server administrators. If an HTTP server translates
2760   HTTP URIs directly into file system calls, the server &MUST; take
2761   special care not to serve files that were not intended to be
2762   delivered to HTTP clients. For example, UNIX, Microsoft Windows, and
2763   other operating systems use ".." as a path component to indicate a
2764   directory level above the current one. On such a system, an HTTP
2765   server &MUST; disallow any such construct in the Request-URI if it
2766   would otherwise allow access to a resource outside those intended to
2767   be accessible via the HTTP server. Similarly, files intended for
2768   reference only internally to the server (such as access control
2769   files, configuration files, and script code) &MUST; be protected from
2770   inappropriate retrieval, since they might contain sensitive
2771   information. Experience has shown that minor bugs in such HTTP server
2772   implementations have turned into security risks.
2773</t>
2774</section>
2775
2776<section title="DNS Spoofing" anchor="dns.spoofing">
2777<t>
2778   Clients using HTTP rely heavily on the Domain Name Service, and are
2779   thus generally prone to security attacks based on the deliberate
2780   mis-association of IP addresses and DNS names. Clients need to be
2781   cautious in assuming the continuing validity of an IP number/DNS name
2782   association.
2783</t>
2784<t>
2785   In particular, HTTP clients &SHOULD; rely on their name resolver for
2786   confirmation of an IP number/DNS name association, rather than
2787   caching the result of previous host name lookups. Many platforms
2788   already can cache host name lookups locally when appropriate, and
2789   they &SHOULD; be configured to do so. It is proper for these lookups to
2790   be cached, however, only when the TTL (Time To Live) information
2791   reported by the name server makes it likely that the cached
2792   information will remain useful.
2793</t>
2794<t>
2795   If HTTP clients cache the results of host name lookups in order to
2796   achieve a performance improvement, they &MUST; observe the TTL
2797   information reported by DNS.
2798</t>
2799<t>
2800   If HTTP clients do not observe this rule, they could be spoofed when
2801   a previously-accessed server's IP address changes. As network
2802   renumbering is expected to become increasingly common <xref target="RFC1900"/>, the
2803   possibility of this form of attack will grow. Observing this
2804   requirement thus reduces this potential security vulnerability.
2805</t>
2806<t>
2807   This requirement also improves the load-balancing behavior of clients
2808   for replicated servers using the same DNS name and reduces the
2809   likelihood of a user's experiencing failure in accessing sites which
2810   use that strategy.
2811</t>
2812</section>
2813
2814<section title="Proxies and Caching" anchor="attack.proxies">
2815<t>
2816   By their very nature, HTTP proxies are men-in-the-middle, and
2817   represent an opportunity for man-in-the-middle attacks. Compromise of
2818   the systems on which the proxies run can result in serious security
2819   and privacy problems. Proxies have access to security-related
2820   information, personal information about individual users and
2821   organizations, and proprietary information belonging to users and
2822   content providers. A compromised proxy, or a proxy implemented or
2823   configured without regard to security and privacy considerations,
2824   might be used in the commission of a wide range of potential attacks.
2825</t>
2826<t>
2827   Proxy operators should protect the systems on which proxies run as
2828   they would protect any system that contains or transports sensitive
2829   information. In particular, log information gathered at proxies often
2830   contains highly sensitive personal information, and/or information
2831   about organizations. Log information should be carefully guarded, and
2832   appropriate guidelines for use developed and followed. (<xref target="abuse.of.server.log.information"/>).
2833</t>
2834<t>
2835   Proxy implementors should consider the privacy and security
2836   implications of their design and coding decisions, and of the
2837   configuration options they provide to proxy operators (especially the
2838   default configuration).
2839</t>
2840<t>
2841   Users of a proxy need to be aware that they are no trustworthier than
2842   the people who run the proxy; HTTP itself cannot solve this problem.
2843</t>
2844<t>
2845   The judicious use of cryptography, when appropriate, may suffice to
2846   protect against a broad range of security and privacy attacks. Such
2847   cryptography is beyond the scope of the HTTP/1.1 specification.
2848</t>
2849</section>
2850
2851<section title="Denial of Service Attacks on Proxies" anchor="attack.DoS">
2852<t>
2853   They exist. They are hard to defend against. Research continues.
2854   Beware.
2855</t>
2856</section>
2857</section>
2858
2859<section title="Acknowledgments" anchor="ack">
2860<t>
2861   This specification makes heavy use of the augmented BNF and generic
2862   constructs defined by David H. Crocker for <xref target="RFC822"/>. Similarly, it
2863   reuses many of the definitions provided by Nathaniel Borenstein and
2864   Ned Freed for MIME <xref target="RFC2045"/>. We hope that their inclusion in this
2865   specification will help reduce past confusion over the relationship
2866   between HTTP and Internet mail message formats.
2867</t>
2868<t>
2869   The HTTP protocol has evolved considerably over the years. It has
2870   benefited from a large and active developer community--the many
2871   people who have participated on the www-talk mailing list--and it is
2872   that community which has been most responsible for the success of
2873   HTTP and of the World-Wide Web in general. Marc Andreessen, Robert
2874   Cailliau, Daniel W. Connolly, Bob Denny, John Franks, Jean-Francois
2875   Groff, Phillip M. Hallam-Baker, Hakon W. Lie, Ari Luotonen, Rob
2876   McCool, Lou Montulli, Dave Raggett, Tony Sanders, and Marc
2877   VanHeyningen deserve special recognition for their efforts in
2878   defining early aspects of the protocol.
2879</t>
2880<t>
2881   This document has benefited greatly from the comments of all those
2882   participating in the HTTP-WG. In addition to those already mentioned,
2883   the following individuals have contributed to this specification:
2884</t>
2885<t>
2886   Gary Adams, Harald Tveit Alvestrand, Keith Ball, Brian Behlendorf,
2887   Paul Burchard, Maurizio Codogno, Mike Cowlishaw, Roman Czyborra,
2888   Michael A. Dolan, Daniel DuBois, David J. Fiander, Alan Freier, Marc Hedlund, Greg Herlihy,
2889   Koen Holtman, Alex Hopmann, Bob Jernigan, Shel Kaphan, Rohit Khare,
2890   John Klensin, Martijn Koster, Alexei Kosut, David M. Kristol,
2891   Daniel LaLiberte, Ben Laurie, Paul J. Leach, Albert Lunde,
2892   John C. Mallery, Jean-Philippe Martin-Flatin, Mitra, David Morris,
2893   Gavin Nicol, Ross Patterson, Bill Perry, Jeffrey Perry, Scott Powers, Owen Rees,
2894   Luigi Rizzo, David Robinson, Marc Salomon, Rich Salz,
2895   Allan M. Schiffman, Jim Seidman, Chuck Shotton, Eric W. Sink,
2896   Simon E. Spero, Richard N. Taylor, Robert S. Thau,
2897   Bill (BearHeart) Weinman, Francois Yergeau, Mary Ellen Zurko,
2898   Josh Cohen.
2899</t>
2900<t>
2901   Thanks to the "cave men" of Palo Alto. You know who you are.
2902</t>
2903<t>
2904   Jim Gettys (the current editor of this document) wishes particularly
2905   to thank Roy Fielding, the previous editor of this document, along
2906   with John Klensin, Jeff Mogul, Paul Leach, Dave Kristol, Koen
2907   Holtman, John Franks, Josh Cohen, Alex Hopmann, Scott Lawrence, and
2908   Larry Masinter for their help. And thanks go particularly to Jeff
2909   Mogul and Scott Lawrence for performing the "MUST/MAY/SHOULD" audit.
2910</t>
2911<t>
2912   The Apache Group, Anselm Baird-Smith, author of Jigsaw, and Henrik
2913   Frystyk implemented RFC 2068 early, and we wish to thank them for the
2914   discovery of many of the problems that this document attempts to
2915   rectify.
2916</t>
2917</section>
2918
2919</middle>
2920<back>
2921<references>
2922
2923<reference anchor="Part2">
2924   <front>
2925      <title abbrev="HTTP/1.1">HTTP/1.1, part 2: Message Semantics</title>
2926      <author initials="R." surname="Fielding" fullname="Roy T. Fielding" role="editor">
2927         <organization abbrev="Day Software">Day Software</organization>
2928         <address><email>fielding@gbiv.com</email></address>
2929      </author>
2930      <author initials="J." surname="Gettys" fullname="Jim Gettys">
2931         <organization>One Laptop per Child</organization>
2932         <address><email>jg@laptop.org</email></address>
2933      </author>
2934      <author initials="J." surname="Mogul" fullname="Jeffrey C. Mogul">
2935         <organization abbrev="HP">Hewlett-Packard Company</organization>
2936         <address><email>JeffMogul@acm.org</email></address>
2937      </author>
2938      <author initials="H." surname="Frystyk" fullname="Henrik Frystyk Nielsen">
2939         <organization abbrev="Microsoft">Microsoft Corporation</organization>
2940         <address><email>henrikn@microsoft.com</email></address>
2941      </author>
2942      <author initials="L." surname="Masinter" fullname="Larry Masinter">
2943         <organization abbrev="Adobe Systems">Adobe Systems, Incorporated</organization>
2944         <address><email>LMM@acm.org</email></address>
2945      </author>
2946      <author initials="P." surname="Leach" fullname="Paul J. Leach">
2947         <organization abbrev="Microsoft">Microsoft Corporation</organization>
2948         <address><email>paulle@microsoft.com</email></address>
2949      </author>
2950      <author initials="T." surname="Berners-Lee" fullname="Tim Berners-Lee">
2951         <organization abbrev="W3C/MIT">World Wide Web Consortium</organization>
2952         <address><email>timbl@w3.org</email></address>
2953      </author>
2954      <author initials="Y." surname="Lafon" fullname="Yves Lafon" role="editor">
2955         <organization abbrev="W3C">World Wide Web Consortium</organization>
2956         <address><email>ylafon@w3.org</email></address>
2957      </author>
2958      <author initials="J. F." surname="Reschke" fullname="Julian F. Reschke" role="editor">
2959         <organization abbrev="greenbytes">greenbytes GmbH</organization>
2960         <address><email>julian.reschke@greenbytes.de</email></address>
2961      </author>
2962      <date month="&ID-MONTH;" year="&ID-YEAR;"/>
2963   </front>
2964   <seriesInfo name="Internet-Draft" value="draft-ietf-httpbis-p2-semantics-&ID-VERSION;"/>
2965   <x:source href="p2-semantics.xml" basename="p2-semantics"/>
2966</reference>
2967
2968<reference anchor="Part3">
2969   <front>
2970      <title abbrev="HTTP/1.1">HTTP/1.1, part 3: Message Payload and Content Negotiation</title>
2971      <author initials="R." surname="Fielding" fullname="Roy T. Fielding" role="editor">
2972         <organization abbrev="Day Software">Day Software</organization>
2973         <address><email>fielding@gbiv.com</email></address>
2974      </author>
2975      <author initials="J." surname="Gettys" fullname="Jim Gettys">
2976         <organization>One Laptop per Child</organization>
2977         <address><email>jg@laptop.org</email></address>
2978      </author>
2979      <author initials="J." surname="Mogul" fullname="Jeffrey C. Mogul">
2980         <organization abbrev="HP">Hewlett-Packard Company</organization>
2981         <address><email>JeffMogul@acm.org</email></address>
2982      </author>
2983      <author initials="H." surname="Frystyk" fullname="Henrik Frystyk Nielsen">
2984         <organization abbrev="Microsoft">Microsoft Corporation</organization>
2985         <address><email>henrikn@microsoft.com</email></address>
2986      </author>
2987      <author initials="L." surname="Masinter" fullname="Larry Masinter">
2988         <organization abbrev="Adobe Systems">Adobe Systems, Incorporated</organization>
2989         <address><email>LMM@acm.org</email></address>
2990      </author>
2991      <author initials="P." surname="Leach" fullname="Paul J. Leach">
2992         <organization abbrev="Microsoft">Microsoft Corporation</organization>
2993         <address><email>paulle@microsoft.com</email></address>
2994      </author>
2995      <author initials="T." surname="Berners-Lee" fullname="Tim Berners-Lee">
2996         <organization abbrev="W3C/MIT">World Wide Web Consortium</organization>
2997         <address><email>timbl@w3.org</email></address>
2998      </author>
2999      <author initials="Y." surname="Lafon" fullname="Yves Lafon" role="editor">
3000         <organization abbrev="W3C">World Wide Web Consortium</organization>
3001         <address><email>ylafon@w3.org</email></address>
3002      </author>
3003      <author initials="J. F." surname="Reschke" fullname="Julian F. Reschke" role="editor">
3004         <organization abbrev="greenbytes">greenbytes GmbH</organization>
3005         <address><email>julian.reschke@greenbytes.de</email></address>
3006      </author>
3007      <date month="&ID-MONTH;" year="&ID-YEAR;"/>
3008   </front>
3009   <seriesInfo name="Internet-Draft" value="draft-ietf-httpbis-p3-payload-&ID-VERSION;"/>
3010   <x:source href="p3-payload.xml" basename="p3-payload"/>
3011</reference>
3012
3013<reference anchor="Part6">
3014   <front>
3015      <title abbrev="HTTP/1.1">HTTP/1.1, part 6: Caching</title>
3016      <author initials="R." surname="Fielding" fullname="Roy T. Fielding" role="editor">
3017         <organization abbrev="Day Software">Day Software</organization>
3018         <address><email>fielding@gbiv.com</email></address>
3019      </author>
3020      <author initials="J." surname="Gettys" fullname="Jim Gettys">
3021         <organization>One Laptop per Child</organization>
3022         <address><email>jg@laptop.org</email></address>
3023      </author>
3024      <author initials="J." surname="Mogul" fullname="Jeffrey C. Mogul">
3025         <organization abbrev="HP">Hewlett-Packard Company</organization>
3026         <address><email>JeffMogul@acm.org</email></address>
3027      </author>
3028      <author initials="H." surname="Frystyk" fullname="Henrik Frystyk Nielsen">
3029         <organization abbrev="Microsoft">Microsoft Corporation</organization>
3030         <address><email>henrikn@microsoft.com</email></address>
3031      </author>
3032      <author initials="L." surname="Masinter" fullname="Larry Masinter">
3033         <organization abbrev="Adobe Systems">Adobe Systems, Incorporated</organization>
3034         <address><email>LMM@acm.org</email></address>
3035      </author>
3036      <author initials="P." surname="Leach" fullname="Paul J. Leach">
3037         <organization abbrev="Microsoft">Microsoft Corporation</organization>
3038         <address><email>paulle@microsoft.com</email></address>
3039      </author>
3040      <author initials="T." surname="Berners-Lee" fullname="Tim Berners-Lee">
3041         <organization abbrev="W3C/MIT">World Wide Web Consortium</organization>
3042         <address><email>timbl@w3.org</email></address>
3043      </author>
3044      <author initials="Y." surname="Lafon" fullname="Yves Lafon" role="editor">
3045         <organization abbrev="W3C">World Wide Web Consortium</organization>
3046         <address><email>ylafon@w3.org</email></address>
3047      </author>
3048      <author initials="J. F." surname="Reschke" fullname="Julian F. Reschke" role="editor">
3049         <organization abbrev="greenbytes">greenbytes GmbH</organization>
3050         <address><email>julian.reschke@greenbytes.de</email></address>
3051      </author>
3052      <date month="&ID-MONTH;" year="&ID-YEAR;"/>
3053   </front>
3054   <seriesInfo name="Internet-Draft" value="draft-ietf-httpbis-p6-cache-&ID-VERSION;"/>
3055   <x:source href="p6-cache.xml" basename="p6-cache"/>
3056</reference>
3057
3058<reference anchor="RFC2616">
3059   <front>
3060      <title>Hypertext Transfer Protocol -- HTTP/1.1</title>
3061      <author initials="R." surname="Fielding" fullname="R. Fielding">
3062         <organization>University of California, Irvine</organization>
3063         <address><email>fielding@ics.uci.edu</email></address>
3064      </author>
3065      <author initials="J." surname="Gettys" fullname="J. Gettys">
3066         <organization>W3C</organization>
3067         <address><email>jg@w3.org</email></address>
3068      </author>
3069      <author initials="J." surname="Mogul" fullname="J. Mogul">
3070         <organization>Compaq Computer Corporation</organization>
3071         <address><email>mogul@wrl.dec.com</email></address>
3072      </author>
3073      <author initials="H." surname="Frystyk" fullname="H. Frystyk">
3074         <organization>MIT Laboratory for Computer Science</organization>
3075         <address><email>frystyk@w3.org</email></address>
3076      </author>
3077      <author initials="L." surname="Masinter" fullname="L. Masinter">
3078         <organization>Xerox Corporation</organization>
3079         <address><email>masinter@parc.xerox.com</email></address>
3080      </author>
3081      <author initials="P." surname="Leach" fullname="P. Leach">
3082         <organization>Microsoft Corporation</organization>
3083         <address><email>paulle@microsoft.com</email></address>
3084      </author>
3085      <author initials="T." surname="Berners-Lee" fullname="T. Berners-Lee">
3086         <organization>W3C</organization>
3087         <address><email>timbl@w3.org</email></address>
3088      </author>
3089      <date month="June" year="1999"/>
3090   </front>
3091   <seriesInfo name="RFC" value="2616"/>
3092</reference>
3093
3094<reference anchor="RFC1436">
3095<front>
3096<title abbrev="Gopher">The Internet Gopher Protocol (a distributed document search and retrieval protocol)</title>
3097<author initials="F." surname="Anklesaria" fullname="Farhad Anklesaria">
3098<organization>University of Minnesota, Computer and Information Services</organization>
3099<address>
3100<postal>
3101<street>100 Union Street SE</street>
3102<street>Room 152 Shepherd Labs</street>
3103<city>Minneapolis</city>
3104<region>MN</region>
3105<code>55455</code>
3106<country>US</country></postal>
3107<phone>+1 612 625 1300</phone>
3108<email>fxa@boombox.micro.umn.edu</email></address></author>
3109<author initials="M." surname="McCahill" fullname="Mark McCahill">
3110<organization>University of Minnesota, Computer and Information Services</organization>
3111<address>
3112<postal>
3113<street>100 Union Street SE</street>
3114<street>Room 152 Shepherd Labs</street>
3115<city>Minneapolis</city>
3116<region>MN</region>
3117<code>55455</code>
3118<country>US</country></postal>
3119<phone>+1 612 625 1300</phone>
3120<email>mpm@boombox.micro.umn.edu</email></address></author>
3121<author initials="P." surname="Lindner" fullname="Paul Lindner">
3122<organization>University of Minnesota, Computer and Information Services</organization>
3123<address>
3124<postal>
3125<street>100 Union Street SE</street>
3126<street>Room 152 Shepherd Labs</street>
3127<city>Minneapolis</city>
3128<region>MN</region>
3129<code>55455</code>
3130<country>US</country></postal>
3131<phone>+1 612 625 1300</phone>
3132<email>lindner@boombox.micro.umn.edu</email></address></author>
3133<author initials="D." surname="Johnson" fullname="David Johnson">
3134<organization>University of Minnesota, Computer and Information Services</organization>
3135<address>
3136<postal>
3137<street>100 Union Street SE</street>
3138<street>Room 152 Shepherd Labs</street>
3139<city>Minneapolis</city>
3140<region>MN</region>
3141<code>55455</code>
3142<country>US</country></postal>
3143<phone>+1 612 625 1300</phone>
3144<email>dmj@boombox.micro.umn.edu</email></address></author>
3145<author initials="D." surname="Torrey" fullname="Daniel Torrey">
3146<organization>University of Minnesota, Computer and Information Services</organization>
3147<address>
3148<postal>
3149<street>100 Union Street SE</street>
3150<street>Room 152 Shepherd Labs</street>
3151<city>Minneapolis</city>
3152<region>MN</region>
3153<code>55455</code>
3154<country>US</country></postal>
3155<phone>+1 612 625 1300</phone>
3156<email>daniel@boombox.micro.umn.edu</email></address></author>
3157<author initials="B." surname="Alberti" fullname="Bob Alberti">
3158<organization>University of Minnesota, Computer and Information Services</organization>
3159<address>
3160<postal>
3161<street>100 Union Street SE</street>
3162<street>Room 152 Shepherd Labs</street>
3163<city>Minneapolis</city>
3164<region>MN</region>
3165<code>55455</code>
3166<country>US</country></postal>
3167<phone>+1 612 625 1300</phone>
3168<email>alberti@boombox.micro.umn.edu</email></address></author>
3169<date month="March" year="1993"/>
3170</front>
3171<seriesInfo name="RFC" value="1436"/>
3172</reference>
3173
3174<reference anchor="RFC1630">
3175<front>
3176<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>
3177<author initials="T." surname="Berners-Lee" fullname="Tim Berners-Lee">
3178<organization>CERN, World-Wide Web project</organization>
3179<address>
3180<postal>
3181<street>1211 Geneva 23</street>
3182<city/>
3183<region/>
3184<code/>
3185<country>CH</country></postal>
3186<phone>+41 22 7673755</phone>
3187<facsimile>+41 22 7677155</facsimile>
3188<email>timbl@info.cern.ch</email></address></author>
3189<date month="June" year="1994"/></front>
3190<seriesInfo name="RFC" value="1630"/>
3191</reference>
3192
3193<reference anchor="RFC1738">
3194  <front>
3195    <title>Uniform Resource Locators (URL)</title>
3196    <author initials="T." surname="Berners-Lee" fullname="Tim Berners-Lee">
3197      <organization>CERN, World-Wide Web project</organization>
3198      <address><email>timbl@info.cern.ch</email></address>
3199    </author>
3200    <author initials="L." surname="Masinter" fullname="Larry Masinter">
3201      <organization>Xerox PARC</organization>
3202      <address><email>masinter@parc.xerox.com</email></address>
3203    </author>
3204    <author initials="M." surname="McCahill" fullname="Mark McCahill">
3205      <organization>University of Minnesota, Computer and Information Services</organization>
3206      <address><email>mpm@boombox.micro.umn.edu</email></address>
3207    </author>
3208    <date month="December" year="1994"/>
3209  </front>
3210  <seriesInfo name="RFC" value="1738"/>
3211</reference>
3212
3213<reference anchor="RFC1945">
3214  <front>
3215    <title abbrev="HTTP/1.0">Hypertext Transfer Protocol -- HTTP/1.0</title>
3216    <author initials="T." surname="Berners-Lee" fullname="Tim Berners-Lee">
3217      <organization>MIT, Laboratory for Computer Science</organization>
3218      <address><email>timbl@w3.org</email></address>
3219    </author>
3220    <author initials="R.T." surname="Fielding" fullname="Roy T. Fielding">
3221      <organization>University of California, Irvine, Department of Information and Computer Science</organization>
3222      <address><email>fielding@ics.uci.edu</email></address>
3223    </author>
3224    <author initials="H.F." surname="Nielsen" fullname="Henrik Frystyk Nielsen">
3225      <organization>W3 Consortium, MIT Laboratory for Computer Science</organization>
3226      <address><email>frystyk@w3.org</email></address>
3227    </author>
3228    <date month="May" year="1996"/>
3229  </front>
3230  <seriesInfo name="RFC" value="1945"/>
3231</reference>
3232
3233<reference anchor="RFC2045">
3234  <front>
3235    <title abbrev="Internet Message Bodies">Multipurpose Internet Mail Extensions (MIME) Part One: Format of Internet Message Bodies</title>
3236    <author initials="N." surname="Freed" fullname="Ned Freed">
3237      <organization>Innosoft International, Inc.</organization>
3238      <address><email>ned@innosoft.com</email></address>
3239    </author>
3240    <author initials="N.S." surname="Borenstein" fullname="Nathaniel S. Borenstein">
3241      <organization>First Virtual Holdings</organization>
3242      <address><email>nsb@nsb.fv.com</email></address>
3243    </author>
3244    <date month="November" year="1996"/>
3245  </front>
3246  <seriesInfo name="RFC" value="2045"/>
3247</reference>
3248
3249<reference anchor="RFC1123">
3250  <front>
3251    <title>Requirements for Internet Hosts - Application and Support</title>
3252    <author initials="R." surname="Braden" fullname="Robert Braden">
3253      <organization>University of Southern California (USC), Information Sciences Institute</organization>
3254      <address><email>Braden@ISI.EDU</email></address>
3255    </author>
3256    <date month="October" year="1989"/>
3257  </front>
3258  <seriesInfo name="STD" value="3"/>
3259  <seriesInfo name="RFC" value="1123"/>
3260</reference>
3261
3262<reference anchor="RFC822">
3263  <front>
3264    <title abbrev="Standard for ARPA Internet Text Messages">Standard for the format of ARPA Internet text messages</title>
3265    <author initials="D.H." surname="Crocker" fullname="David H. Crocker">
3266      <organization>University of Delaware, Dept. of Electrical Engineering</organization>
3267      <address><email>DCrocker@UDel-Relay</email></address>
3268    </author>
3269    <date month="August" day="13" year="1982"/>
3270  </front>
3271  <seriesInfo name="STD" value="11"/>
3272  <seriesInfo name="RFC" value="822"/>
3273</reference>
3274
3275<reference anchor="WAIS">
3276<front>
3277<title>WAIS Interface Protocol Prototype Functional Specification (v1.5)</title>
3278<author initials="F." surname="Davis" fullname="F. Davis">
3279<organization>Thinking Machines Corporation</organization>
3280</author>
3281<author initials="B." surname="Kahle" fullname="B. Kahle">
3282<organization>Thinking Machines Corporation</organization>
3283</author>
3284<author initials="H." surname="Morris" fullname="H. Morris">
3285<organization>Thinking Machines Corporation</organization>
3286</author>
3287<author initials="J." surname="Salem" fullname="J. Salem">
3288<organization>Thinking Machines Corporation</organization>
3289</author>
3290<author initials="T." surname="Shen" fullname="T. Shen">
3291<organization>Thinking Machines Corporation</organization>
3292</author>
3293<author initials="R." surname="Wang" fullname="R. Wang">
3294<organization>Thinking Machines Corporation</organization>
3295</author>
3296<author initials="J." surname="Sui" fullname="J. Sui">
3297<organization>Thinking Machines Corporation</organization>
3298</author>
3299<author initials="M." surname="Grinbaum" fullname="M. Grinbaum">
3300<organization>Thinking Machines Corporation</organization>
3301</author>
3302<date month="April" year="1990"/></front>
3303<seriesInfo name="Thinking Machines Corporation" value=""/>
3304</reference>
3305
3306<reference anchor="RFC1808">
3307  <front>
3308    <title>Relative Uniform Resource Locators</title>
3309    <author initials="R." surname="Fielding" fullname="Roy T. Fielding">
3310      <organization>University of California Irvine, Department of Information and Computer Science</organization>
3311      <address><email>fielding@ics.uci.edu</email></address>
3312    </author>
3313    <date month="June" year="1995"/>
3314  </front>
3315  <seriesInfo name="RFC" value="1808"/>
3316</reference>
3317
3318<reference anchor='RFC3977'>
3319   <front>
3320      <title>Network News Transfer Protocol (NNTP)</title>
3321      <author initials='C.' surname='Feather' fullname='C. Feather'>
3322         <organization>THUS plc</organization>
3323         <address>
3324            <email>clive@demon.net</email>
3325         </address>
3326      </author>
3327      <date year='2006' month='October' />
3328   </front>
3329</reference>
3330
3331<reference anchor="RFC2047">
3332  <front>
3333    <title abbrev="Message Header Extensions">MIME (Multipurpose Internet Mail Extensions) Part Three: Message Header Extensions for Non-ASCII Text</title>
3334    <author initials="K." surname="Moore" fullname="Keith Moore">
3335      <organization>University of Tennessee</organization>
3336      <address><email>moore@cs.utk.edu</email></address>
3337    </author>
3338    <date month="November" year="1996"/>
3339  </front>
3340  <seriesInfo name="RFC" value="2047"/>
3341</reference>
3342
3343<reference anchor="RFC821">
3344<front>
3345<title>Simple Mail Transfer Protocol</title>
3346<author initials="J.B." surname="Postel" fullname="Jonathan B. Postel">
3347<organization>University of Southern California (USC)/Information Sciences Institute</organization>
3348<address>
3349<postal>
3350<street>4676 Admiralty Way</street>
3351<city>Marina del Rey</city>
3352<region>CA</region>
3353<code>90291</code>
3354<country>US</country></postal>
3355<phone>+1 213 822 1511</phone></address></author>
3356<date month="August" year="1982"/></front>
3357<seriesInfo name="STD" value="10"/>
3358<seriesInfo name="RFC" value="821"/>
3359</reference>
3360
3361<reference anchor="RFC4288">
3362  <front>
3363    <title>Media Type Specifications and Registration Procedures</title>
3364    <author initials="N." surname="Freed" fullname="N. Freed">
3365      <organization>Sun Microsystems</organization>
3366      <address>
3367        <email>ned.freed@mrochek.com</email>
3368      </address>
3369    </author>
3370    <author initials="J." surname="Klensin" fullname="J. Klensin">
3371      <organization/>
3372      <address>
3373        <email>klensin+ietf@jck.com</email>
3374      </address>
3375    </author>
3376    <date year="2005" month="December"/>
3377  </front>
3378  <seriesInfo name="BCP" value="13"/>
3379  <seriesInfo name="RFC" value="4288"/>
3380</reference>
3381
3382<reference anchor="RFC959">
3383<front>
3384<title abbrev="File Transfer Protocol">File Transfer Protocol</title>
3385<author initials="J." surname="Postel" fullname="J. Postel">
3386<organization>Information Sciences Institute (ISI)</organization></author>
3387<author initials="J." surname="Reynolds" fullname="J. Reynolds">
3388<organization/></author>
3389<date month="October" year="1985"/></front>
3390<seriesInfo name="STD" value="9"/>
3391<seriesInfo name="RFC" value="959"/>
3392</reference>
3393
3394<reference anchor="RFC1737">
3395<front>
3396<title abbrev="Requirements for Uniform Resource Names">Functional Requirements for Uniform Resource Names</title>
3397<author initials="L." surname="Masinter" fullname="Larry Masinter">
3398<organization>Xerox Palo Alto Research Center</organization>
3399<address>
3400<postal>
3401<street>3333 Coyote Hill Road</street>
3402<city>Palo Alto</city>
3403<region>CA</region>
3404<code>94304</code>
3405<country>US</country></postal>
3406<phone>+1 415 812 4365</phone>
3407<facsimile>+1 415 812 4333</facsimile>
3408<email>masinter@parc.xerox.com</email></address></author>
3409<author initials="K." surname="Sollins" fullname="Karen Sollins">
3410<organization>MIT Laboratory for Computer Science</organization>
3411<address>
3412<postal>
3413<street>545 Technology Square</street>
3414<city>Cambridge</city>
3415<region>MA</region>
3416<code>02139</code>
3417<country>US</country></postal>
3418<phone>+1 617 253 2673</phone>
3419<email>sollins@lcs.mit.edu</email></address></author>
3420<date month="December" year="1994"/></front>
3421<seriesInfo name="RFC" value="1737"/>
3422</reference>
3423
3424<reference anchor="USASCII">
3425  <front>
3426    <title>Coded Character Set -- 7-bit American Standard Code for Information Interchange</title>
3427    <author>
3428      <organization>American National Standards Institute</organization>
3429    </author>
3430    <date year="1986"/>
3431  </front>
3432  <seriesInfo name="ANSI" value="X3.4"/>
3433</reference>
3434
3435<reference anchor="ISO-8859">
3436<front>
3437<title>Information technology - 8-bit single byte coded graphic - character sets</title>
3438<author>
3439<organization>International Organization for Standardization</organization>
3440</author>
3441<date year="1987-1990"/>
3442</front>
3443<annotation>
3444  Part 1: Latin alphabet No. 1, ISO-8859-1:1987.
3445  Part 2: Latin alphabet No. 2, ISO-8859-2, 1987.
3446  Part 3: Latin alphabet No. 3, ISO-8859-3, 1988.
3447  Part 4: Latin alphabet No. 4, ISO-8859-4, 1988.
3448  Part 5: Latin/Cyrillic alphabet, ISO-8859-5, 1988.
3449  Part 6: Latin/Arabic alphabet, ISO-8859-6, 1987.
3450  Part 7: Latin/Greek alphabet, ISO-8859-7, 1987.
3451  Part 8: Latin/Hebrew alphabet, ISO-8859-8, 1988.
3452  Part 9: Latin alphabet No. 5, ISO-8859-9, 1990.
3453</annotation>
3454</reference>
3455
3456<reference anchor="RFC1900">
3457  <front>
3458    <title>Renumbering Needs Work</title>
3459    <author initials="B." surname="Carpenter" fullname="Brian E. Carpenter">
3460      <organization>CERN, Computing and Networks Division</organization>
3461      <address><email>brian@dxcoms.cern.ch</email></address>
3462    </author>
3463    <author initials="Y." surname="Rekhter" fullname="Yakov Rekhter">
3464      <organization>cisco Systems</organization>
3465      <address><email>yakov@cisco.com</email></address>
3466    </author>
3467    <date month="February" year="1996"/>
3468  </front>
3469  <seriesInfo name="RFC" value="1900"/>
3470</reference>
3471
3472<reference anchor="Pad1995">
3473<front>
3474<title>Improving HTTP Latency</title>
3475<author initials="V.N." surname="Padmanabhan" fullname="Venkata N. Padmanabhan"><organization/></author>
3476<author initials="J.C." surname="Mogul" fullname="Jeffrey C. Mogul"><organization/></author>
3477<date year="1995" month="Dec"/>
3478</front>
3479<seriesInfo name="Computer Networks and ISDN Systems" value="v. 28, pp. 25-35"/>
3480<annotation>
3481  Slightly revised version of paper in Proc. 2nd International WWW Conference '94: Mosaic and the Web, Oct. 1994,
3482  which is available at <eref target="http://www.ncsa.uiuc.edu/SDG/IT94/Proceedings/DDay/mogul/HTTPLatency.html"/>.
3483</annotation>
3484</reference>
3485
3486<reference anchor="Tou1998" target="http://www.isi.edu/touch/pubs/http-perf96/">
3487<front>
3488<title>Analysis of HTTP Performance</title>
3489<author initials="J." surname="Touch" fullname="Joe Touch">
3490  <organization>USC/Information Sciences Institute</organization>
3491  <address><email>touch@isi.edu</email></address>
3492</author>
3493<author initials="J." surname="Heidemann" fullname="John Heidemann">
3494  <organization>USC/Information Sciences Institute</organization>
3495  <address><email>johnh@isi.edu</email></address>
3496</author>
3497<author initials="K." surname="Obraczka" fullname="Katia Obraczka">
3498  <organization>USC/Information Sciences Institute</organization>
3499  <address><email>katia@isi.edu</email></address>
3500</author>
3501<date year="1998" month="Aug"/>
3502</front>
3503<seriesInfo name="ISI Research Report" value="ISI/RR-98-463 (original report dated Aug.1996)"/>
3504</reference>
3505
3506<reference anchor="RFC1305">
3507<front>
3508<title>Network Time Protocol (Version 3) Specification, Implementation</title>
3509<author initials="D." surname="Mills" fullname="David L. Mills">
3510<organization>University of Delaware, Electrical Engineering Department</organization>
3511<address>
3512<postal>
3513<street/>
3514<city>Newark</city>
3515<region>DE</region>
3516<code>19716</code>
3517<country>US</country></postal>
3518<phone>+1 302 451 8247</phone>
3519<email>mills@udel.edu</email></address></author>
3520<date month="March" year="1992"/>
3521<abstract>
3522<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>
3523<seriesInfo name="RFC" value="1305"/>
3524</reference>
3525
3526<reference anchor="Spe" target="http://sunsite.unc.edu/mdma-release/http-prob.html">
3527<front>
3528<title>Analysis of HTTP Performance Problems</title>
3529<author initials="S." surname="Spero" fullname="Simon E. Spero">
3530  <organization/>
3531</author>
3532<date/>
3533</front>
3534</reference>
3535
3536<reference anchor="RFC2068">
3537  <front>
3538    <title abbrev="HTTP/1.1">Hypertext Transfer Protocol -- HTTP/1.1</title>
3539    <author initials="R." surname="Fielding" fullname="Roy T. Fielding">
3540      <organization>University of California, Irvine, Department of Information and Computer Science</organization>
3541      <address><email>fielding@ics.uci.edu</email></address>
3542    </author>
3543    <author initials="J." surname="Gettys" fullname="Jim Gettys">
3544      <organization>MIT Laboratory for Computer Science</organization>
3545      <address><email>jg@w3.org</email></address>
3546    </author>
3547    <author initials="J." surname="Mogul" fullname="Jeffrey C. Mogul">
3548      <organization>Digital Equipment Corporation, Western Research Laboratory</organization>
3549      <address><email>mogul@wrl.dec.com</email></address>
3550    </author>
3551    <author initials="H." surname="Nielsen" fullname="Henrik Frystyk Nielsen">
3552      <organization>MIT Laboratory for Computer Science</organization>
3553      <address><email>frystyk@w3.org</email></address>
3554    </author>
3555    <author initials="T." surname="Berners-Lee" fullname="Tim Berners-Lee">
3556      <organization>MIT Laboratory for Computer Science</organization>
3557      <address><email>timbl@w3.org</email></address>
3558    </author>
3559    <date month="January" year="1997"/>
3560  </front>
3561  <seriesInfo name="RFC" value="2068"/>
3562</reference>
3563
3564<reference anchor="RFC2119">
3565  <front>
3566    <title>Key words for use in RFCs to Indicate Requirement Levels</title>
3567    <author initials="S." surname="Bradner" fullname="Scott Bradner">
3568      <organization>Harvard University</organization>
3569      <address><email>sob@harvard.edu</email></address>
3570    </author>
3571    <date month="March" year="1997"/>
3572  </front>
3573  <seriesInfo name="BCP" value="14"/>
3574  <seriesInfo name="RFC" value="2119"/>
3575</reference>
3576
3577<reference anchor="RFC2145">
3578  <front>
3579    <title abbrev="HTTP Version Numbers">Use and Interpretation of HTTP Version Numbers</title>
3580    <author initials="J.C." surname="Mogul" fullname="Jeffrey C. Mogul">
3581      <organization>Western Research Laboratory</organization>
3582      <address><email>mogul@wrl.dec.com</email></address>
3583    </author>
3584    <author initials="R.T." surname="Fielding" fullname="Roy T. Fielding">
3585      <organization>Department of Information and Computer Science</organization>
3586      <address><email>fielding@ics.uci.edu</email></address>
3587    </author>
3588    <author initials="J." surname="Gettys" fullname="Jim Gettys">
3589      <organization>MIT Laboratory for Computer Science</organization>
3590      <address><email>jg@w3.org</email></address>
3591    </author>
3592    <author initials="H.F." surname="Nielsen" fullname="Henrik Frystyk Nielsen">
3593      <organization>W3 Consortium</organization>
3594      <address><email>frystyk@w3.org</email></address>
3595    </author>
3596    <date month="May" year="1997"/>
3597  </front>
3598  <seriesInfo name="RFC" value="2145"/>
3599</reference>
3600
3601<reference anchor="Nie1997">
3602<front>
3603<title>Network Performance Effects of HTTP/1.1, CSS1, and PNG</title>
3604<author initials="H.F.." surname="Nielsen" fullname="H.F. Nielsen">
3605  <organization/>
3606</author>
3607<author initials="J." surname="Gettys" fullname="J. Gettys">
3608  <organization/>
3609</author>
3610<author initials="E." surname="Prud'hommeaux" fullname="E. Prud'hommeaux">
3611  <organization/>
3612</author>
3613<author initials="H." surname="Lie" fullname="H. Lie">
3614  <organization/>
3615</author>
3616<author initials="C." surname="Lilley" fullname="C. Lilley">
3617  <organization/>
3618</author>
3619<date year="1997" month="Sep"/>
3620</front>
3621<seriesInfo name="Proceedings of ACM SIGCOMM '97, Cannes France" value=""/>
3622</reference>
3623
3624<reference anchor="RFC2396">
3625  <front>
3626    <title abbrev="URI Generic Syntax">Uniform Resource Identifiers (URI): Generic Syntax</title>
3627    <author initials="T." surname="Berners-Lee" fullname="Tim Berners-Lee">
3628      <organization abbrev="MIT/LCS">World Wide Web Consortium</organization>
3629      <address><email>timbl@w3.org</email></address>
3630    </author>
3631    <author initials="R.T." surname="Fielding" fullname="Roy T. Fielding">
3632      <organization abbrev="U.C. Irvine">Department of Information and Computer Science</organization>
3633      <address><email>fielding@ics.uci.edu</email></address>
3634    </author>
3635    <author initials="L." surname="Masinter" fullname="Larry Masinter">
3636      <organization abbrev="Xerox Corporation">Xerox PARC</organization>
3637      <address><email>masinter@parc.xerox.com</email></address>
3638    </author>
3639    <date month="August" year="1998"/>
3640  </front>
3641  <seriesInfo name="RFC" value="2396"/>
3642</reference>
3643
3644<reference anchor="RFC2324">
3645<front>
3646<title abbrev="HTCPCP/1.0">Hyper Text Coffee Pot Control Protocol (HTCPCP/1.0)</title>
3647<author initials="L." surname="Masinter" fullname="Larry Masinter">
3648<organization>Xerox Palo Alto Research Center</organization>
3649<address>
3650<postal>
3651<street>3333 Coyote Hill Road</street>
3652<city>Palo Alto</city>
3653<region>CA</region>
3654<code>94304</code></postal>
3655<email>masinter@parc.xerox.com</email></address></author>
3656<date month="April" day="1" year="1998"/>
3657<area>General</area>
3658<keyword>control protocol</keyword>
3659<keyword>coffee</keyword>
3660</front>
3661<seriesInfo name="RFC" value="2324"/>
3662</reference>
3663
3664</references>
3665
3666<section title="Internet Media Type message/http and application/http" anchor="internet.media.type.http">
3667<iref item="Media Type" subitem="message/http" primary="true"/>
3668<iref item="message/http Media Type" primary="true"/>
3669<iref item="Media Type" subitem="application/http" primary="true"/>
3670<iref item="application/http Media Type" primary="true"/>
3671<t>
3672   In addition to defining the HTTP/1.1 protocol, this document serves
3673   as the specification for the Internet media type "message/http" and
3674   "application/http". The message/http type can be used to enclose a
3675   single HTTP request or response message, provided that it obeys the
3676   MIME restrictions for all "message" types regarding line length and
3677   encodings. The application/http type can be used to enclose a
3678   pipeline of one or more HTTP request or response messages (not
3679   intermixed). The following is to be registered with IANA <xref target="RFC4288"/>.
3680</t>
3681<t>
3682  <list style="hanging" x:indent="12em">
3683    <t hangText="Media Type name:">
3684      message
3685    </t>
3686    <t hangText="Media subtype name:">
3687      http
3688    </t>
3689    <t hangText="Required parameters:">
3690      none
3691    </t>
3692    <t hangText="Optional parameters:">
3693      version, msgtype
3694      <list style="hanging">
3695        <t hangText="version:">
3696          The HTTP-Version number of the enclosed message
3697          (e.g., "1.1"). If not present, the version can be
3698          determined from the first line of the body.
3699        </t>
3700        <t hangText="msgtype:">
3701          The message type -- "request" or "response". If not
3702          present, the type can be determined from the first
3703          line of the body.
3704        </t>
3705      </list>
3706    </t>
3707    <t hangText="Encoding considerations:">
3708      only "7bit", "8bit", or "binary" are permitted
3709    </t>
3710    <t hangText="Security considerations:">
3711      none
3712    </t>
3713  </list>
3714</t>
3715<t>
3716  <list style="hanging" x:indent="12em">
3717    <t hangText="Media Type name:">
3718      application
3719    </t>
3720    <t hangText="Media subtype name:">
3721      http
3722    </t>
3723    <t hangText="Required parameters:">
3724      none
3725    </t>
3726    <t hangText="Optional parameters:">
3727      version, msgtype
3728      <list style="hanging">
3729        <t hangText="version:">
3730          The HTTP-Version number of the enclosed messages
3731          (e.g., "1.1"). If not present, the version can be
3732          determined from the first line of the body.
3733        </t>
3734        <t hangText="msgtype:">
3735          The message type -- "request" or "response". If not
3736          present, the type can be determined from the first
3737          line of the body.
3738        </t>
3739      </list>
3740    </t>
3741    <t hangText="Encoding considerations:">
3742      HTTP messages enclosed by this type
3743      are in "binary" format; use of an appropriate
3744      Content-Transfer-Encoding is required when
3745      transmitted via E-mail.
3746    </t>
3747    <t hangText="Security considerations:">
3748      none
3749    </t>
3750  </list>
3751</t>
3752</section>
3753
3754<section title="Tolerant Applications" anchor="tolerant.applications">
3755<t>
3756   Although this document specifies the requirements for the generation
3757   of HTTP/1.1 messages, not all applications will be correct in their
3758   implementation. We therefore recommend that operational applications
3759   be tolerant of deviations whenever those deviations can be
3760   interpreted unambiguously.
3761</t>
3762<t>
3763   Clients &SHOULD; be tolerant in parsing the Status-Line and servers
3764   tolerant when parsing the Request-Line. In particular, they &SHOULD;
3765   accept any amount of SP or HT characters between fields, even though
3766   only a single SP is required.
3767</t>
3768<t>
3769   The line terminator for message-header fields is the sequence CRLF.
3770   However, we recommend that applications, when parsing such headers,
3771   recognize a single LF as a line terminator and ignore the leading CR.
3772</t>
3773<t>
3774   The character set of an entity-body &SHOULD; be labeled as the lowest
3775   common denominator of the character codes used within that body, with
3776   the exception that not labeling the entity is preferred over labeling
3777   the entity with the labels US-ASCII or ISO-8859-1. See &payload;.
3778</t>
3779<t>
3780   Additional rules for requirements on parsing and encoding of dates
3781   and other potential problems with date encodings include:
3782</t>
3783<t>
3784  <list style="symbols">
3785     <t>HTTP/1.1 clients and caches &SHOULD; assume that an RFC-850 date
3786        which appears to be more than 50 years in the future is in fact
3787        in the past (this helps solve the "year 2000" problem).</t>
3788
3789     <t>An HTTP/1.1 implementation &MAY; internally represent a parsed
3790        Expires date as earlier than the proper value, but &MUST-NOT;
3791        internally represent a parsed Expires date as later than the
3792        proper value.</t>
3793
3794     <t>All expiration-related calculations &MUST; be done in GMT. The
3795        local time zone &MUST-NOT; influence the calculation or comparison
3796        of an age or expiration time.</t>
3797
3798     <t>If an HTTP header incorrectly carries a date value with a time
3799        zone other than GMT, it &MUST; be converted into GMT using the
3800        most conservative possible conversion.</t>
3801  </list>
3802</t>
3803</section>
3804
3805<section title="Conversion of Date Formats" anchor="conversion.of.date.formats">
3806<t>
3807   HTTP/1.1 uses a restricted set of date formats (<xref target="full.date"/>) to
3808   simplify the process of date comparison. Proxies and gateways from
3809   other protocols &SHOULD; ensure that any Date header field present in a
3810   message conforms to one of the HTTP/1.1 formats and rewrite the date
3811   if necessary.
3812</t>
3813</section>
3814
3815<section title="Compatibility with Previous Versions" anchor="compatibility">
3816<t>
3817   It is beyond the scope of a protocol specification to mandate
3818   compliance with previous versions. HTTP/1.1 was deliberately
3819   designed, however, to make supporting previous versions easy. It is
3820   worth noting that, at the time of composing this specification
3821   (1996), we would expect commercial HTTP/1.1 servers to:
3822  <list style="symbols">
3823     <t>recognize the format of the Request-Line for HTTP/0.9, 1.0, and
3824        1.1 requests;</t>
3825
3826     <t>understand any valid request in the format of HTTP/0.9, 1.0, or
3827        1.1;</t>
3828
3829     <t>respond appropriately with a message in the same major version
3830        used by the client.</t>
3831  </list>
3832</t>
3833<t>
3834   And we would expect HTTP/1.1 clients to:
3835  <list style="symbols">
3836     <t>recognize the format of the Status-Line for HTTP/1.0 and 1.1
3837        responses;</t>
3838
3839     <t>understand any valid response in the format of HTTP/0.9, 1.0, or
3840        1.1.</t>
3841  </list>
3842</t>
3843<t>
3844   For most implementations of HTTP/1.0, each connection is established
3845   by the client prior to the request and closed by the server after
3846   sending the response. Some implementations implement the Keep-Alive
3847   version of persistent connections described in <xref x:sec="19.7.1" x:fmt="of" target="RFC2068"/>.
3848</t>
3849
3850<section title="Changes from HTTP/1.0" anchor="changes.from.1.0">
3851<t>
3852   This section summarizes major differences between versions HTTP/1.0
3853   and HTTP/1.1.
3854</t>
3855
3856<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">
3857<t>
3858   The requirements that clients and servers support the Host request-header,
3859   report an error if the Host request-header (<xref target="header.host"/>) is
3860   missing from an HTTP/1.1 request, and accept absolute URIs (<xref target="request-uri"/>)
3861   are among the most important changes defined by this
3862   specification.
3863</t>
3864<t>
3865   Older HTTP/1.0 clients assumed a one-to-one relationship of IP
3866   addresses and servers; there was no other established mechanism for
3867   distinguishing the intended server of a request than the IP address
3868   to which that request was directed. The changes outlined above will
3869   allow the Internet, once older HTTP clients are no longer common, to
3870   support multiple Web sites from a single IP address, greatly
3871   simplifying large operational Web servers, where allocation of many
3872   IP addresses to a single host has created serious problems. The
3873   Internet will also be able to recover the IP addresses that have been
3874   allocated for the sole purpose of allowing special-purpose domain
3875   names to be used in root-level HTTP URLs. Given the rate of growth of
3876   the Web, and the number of servers already deployed, it is extremely
3877   important that all implementations of HTTP (including updates to
3878   existing HTTP/1.0 applications) correctly implement these
3879   requirements:
3880  <list style="symbols">
3881     <t>Both clients and servers &MUST; support the Host request-header.</t>
3882
3883     <t>A client that sends an HTTP/1.1 request &MUST; send a Host header.</t>
3884
3885     <t>Servers &MUST; report a 400 (Bad Request) error if an HTTP/1.1
3886        request does not include a Host request-header.</t>
3887
3888     <t>Servers &MUST; accept absolute URIs.</t>
3889  </list>
3890</t>
3891</section>
3892</section>
3893
3894<section title="Compatibility with HTTP/1.0 Persistent Connections" anchor="compatibility.with.http.1.0.persistent.connections">
3895<t>
3896   Some clients and servers might wish to be compatible with some
3897   previous implementations of persistent connections in HTTP/1.0
3898   clients and servers. Persistent connections in HTTP/1.0 are
3899   explicitly negotiated as they are not the default behavior. HTTP/1.0
3900   experimental implementations of persistent connections are faulty,
3901   and the new facilities in HTTP/1.1 are designed to rectify these
3902   problems. The problem was that some existing 1.0 clients may be
3903   sending Keep-Alive to a proxy server that doesn't understand
3904   Connection, which would then erroneously forward it to the next
3905   inbound server, which would establish the Keep-Alive connection and
3906   result in a hung HTTP/1.0 proxy waiting for the close on the
3907   response. The result is that HTTP/1.0 clients must be prevented from
3908   using Keep-Alive when talking to proxies.
3909</t>
3910<t>
3911   However, talking to proxies is the most important use of persistent
3912   connections, so that prohibition is clearly unacceptable. Therefore,
3913   we need some other mechanism for indicating a persistent connection
3914   is desired, which is safe to use even when talking to an old proxy
3915   that ignores Connection. Persistent connections are the default for
3916   HTTP/1.1 messages; we introduce a new keyword (Connection: close) for
3917   declaring non-persistence. See <xref target="header.connection"/>.
3918</t>
3919<t>
3920   The original HTTP/1.0 form of persistent connections (the Connection:
3921   Keep-Alive and Keep-Alive header) is documented in <xref target="RFC2068"/>.
3922</t>
3923</section>
3924
3925<section title="Changes from RFC 2068" anchor="changes.from.rfc.2068">
3926<t>
3927   This specification has been carefully audited to correct and
3928   disambiguate key word usage; RFC 2068 had many problems in respect to
3929   the conventions laid out in <xref target="RFC2119"/>.
3930</t>
3931<t>
3932   Transfer-coding and message lengths all interact in ways that
3933   required fixing exactly when chunked encoding is used (to allow for
3934   transfer encoding that may not be self delimiting); it was important
3935   to straighten out exactly how message lengths are computed.
3936</t>
3937<t>
3938   The use and interpretation of HTTP version numbers has been clarified
3939   by <xref target="RFC2145"/>. Require proxies to upgrade requests to highest protocol
3940   version they support to deal with problems discovered in HTTP/1.0
3941   implementations (<xref target="http.version"/>)
3942</t>
3943<t>
3944   Proxies should be able to add Content-Length when appropriate.
3945</t>
3946<t>
3947   Transfer-coding had significant problems, particularly with
3948   interactions with chunked encoding. The solution is that transfer-codings
3949   become as full fledged as content-codings. This involves
3950   adding an IANA registry for transfer-codings (separate from content
3951   codings), a new header field (TE) and enabling trailer headers in the
3952   future. Transfer encoding is a major performance benefit, so it was
3953   worth fixing <xref target="Nie1997"/>. TE also solves another, obscure, downward
3954   interoperability problem that could have occurred due to interactions
3955   between authentication trailers, chunked encoding and HTTP/1.0
3956   clients.(Section <xref target="transfer.codings" format="counter"/>, <xref target="chunked.transfer.encoding" format="counter"/>,
3957   and <xref target="header.te" format="counter"/>)
3958</t>
3959</section>
3960
3961<section title="Changes from RFC 2616" anchor="changes.from.rfc.2616">
3962<t>
3963  Remove reference to non-existant identity transfer-coding value tokens.
3964  (Sections <xref format="counter" target="transfer.codings"/> and
3965  <xref format="counter" target="message.length"/>)
3966</t>
3967<t>
3968  Clarification that the chunk length does not include
3969  the count of the octets in the chunk header and trailer.
3970  (<xref target="chunked.transfer.encoding"/>)
3971</t>
3972</section>
3973
3974</section>
3975</back>
3976</rfc>
Note: See TracBrowser for help on using the repository browser.