source: draft-ietf-httpbis/03/draft-ietf-httpbis-p1-messaging-03.xml @ 559

Last change on this file since 559 was 559, checked in by fielding@…, 10 years ago

remove executable and set eol-style for earlier drafts

  • Property svn:eol-style set to native
File size: 181.5 KB
Line 
1<?xml version="1.0" encoding="UTF-8"?>
2<!--
3    This XML document is the output of clean-for-DTD.xslt; a tool that strips
4    extensions to RFC2629(bis) from documents for processing with xml2rfc.
5-->
6<?xml-stylesheet type='text/xsl' href='../myxml2rfc.xslt'?>
7<?rfc toc="yes" ?>
8<?rfc symrefs="yes" ?>
9<?rfc sortrefs="yes" ?>
10<?rfc compact="yes"?>
11<?rfc subcompact="no" ?>
12<?rfc linkmailto="no" ?>
13<?rfc editing="no" ?>
14<?rfc comments="yes"?>
15<?rfc inline="yes"?>
16<!DOCTYPE rfc
17  PUBLIC "" "rfc2629.dtd">
18<rfc obsoletes="2616" category="std" ipr="full3978" docName="draft-ietf-httpbis-p1-messaging-03">
19<front>
20
21  <title abbrev="HTTP/1.1, Part 1">HTTP/1.1, part 1: URIs, Connections, and Message Parsing</title>
22
23  <author initials="R." surname="Fielding" fullname="Roy T. Fielding" role="editor">
24    <organization abbrev="Day Software">Day Software</organization>
25    <address>
26      <postal>
27        <street>23 Corporate Plaza DR, Suite 280</street>
28        <city>Newport Beach</city>
29        <region>CA</region>
30        <code>92660</code>
31        <country>USA</country>
32      </postal>
33      <phone>+1-949-706-5300</phone>
34      <facsimile>+1-949-706-5305</facsimile>
35      <email>fielding@gbiv.com</email>
36      <uri>http://roy.gbiv.com/</uri>
37    </address>
38  </author>
39
40  <author initials="J." surname="Gettys" fullname="Jim Gettys">
41    <organization>One Laptop per Child</organization>
42    <address>
43      <postal>
44        <street>21 Oak Knoll Road</street>
45        <city>Carlisle</city>
46        <region>MA</region>
47        <code>01741</code>
48        <country>USA</country>
49      </postal>
50      <email>jg@laptop.org</email>
51      <uri>http://www.laptop.org/</uri>
52    </address>
53  </author>
54 
55  <author initials="J." surname="Mogul" fullname="Jeffrey C. Mogul">
56    <organization abbrev="HP">Hewlett-Packard Company</organization>
57    <address>
58      <postal>
59        <street>HP Labs, Large Scale Systems Group</street>
60        <street>1501 Page Mill Road, MS 1177</street>
61        <city>Palo Alto</city>
62        <region>CA</region>
63        <code>94304</code>
64        <country>USA</country>
65      </postal>
66      <email>JeffMogul@acm.org</email>
67    </address>
68  </author>
69
70  <author initials="H." surname="Frystyk" fullname="Henrik Frystyk Nielsen">
71    <organization abbrev="Microsoft">Microsoft Corporation</organization>
72    <address>
73      <postal>
74        <street>1 Microsoft Way</street>
75        <city>Redmond</city>
76        <region>WA</region>
77        <code>98052</code>
78        <country>USA</country>
79      </postal>
80      <email>henrikn@microsoft.com</email>
81    </address>
82  </author>
83
84  <author initials="L." surname="Masinter" fullname="Larry Masinter">
85    <organization abbrev="Adobe Systems">Adobe Systems, Incorporated</organization>
86    <address>
87      <postal>
88        <street>345 Park Ave</street>
89        <city>San Jose</city>
90        <region>CA</region>
91        <code>95110</code>
92        <country>USA</country>
93      </postal>
94      <email>LMM@acm.org</email>
95      <uri>http://larry.masinter.net/</uri>
96    </address>
97  </author>
98 
99  <author initials="P." surname="Leach" fullname="Paul J. Leach">
100    <organization abbrev="Microsoft">Microsoft Corporation</organization>
101    <address>
102      <postal>
103        <street>1 Microsoft Way</street>
104        <city>Redmond</city>
105        <region>WA</region>
106        <code>98052</code>
107      </postal>
108      <email>paulle@microsoft.com</email>
109    </address>
110  </author>
111   
112  <author initials="T." surname="Berners-Lee" fullname="Tim Berners-Lee">
113    <organization abbrev="W3C/MIT">World Wide Web Consortium</organization>
114    <address>
115      <postal>
116        <street>MIT Computer Science and Artificial Intelligence Laboratory</street>
117        <street>The Stata Center, Building 32</street>
118        <street>32 Vassar Street</street>
119        <city>Cambridge</city>
120        <region>MA</region>
121        <code>02139</code>
122        <country>USA</country>
123      </postal>
124      <email>timbl@w3.org</email>
125      <uri>http://www.w3.org/People/Berners-Lee/</uri>
126    </address>
127  </author>
128
129  <author initials="Y." surname="Lafon" fullname="Yves Lafon" role="editor">
130    <organization abbrev="W3C">World Wide Web Consortium</organization>
131    <address>
132      <postal>
133        <street>W3C / ERCIM</street>
134        <street>2004, rte des Lucioles</street>
135        <city>Sophia-Antipolis</city>
136        <region>AM</region>
137        <code>06902</code>
138        <country>France</country>
139      </postal>
140      <email>ylafon@w3.org</email>
141      <uri>http://www.raubacapeu.net/people/yves/</uri>
142    </address>
143  </author>
144
145  <author initials="J. F." surname="Reschke" fullname="Julian F. Reschke" role="editor">
146    <organization abbrev="greenbytes">greenbytes GmbH</organization>
147    <address>
148      <postal>
149        <street>Hafenweg 16</street>
150        <city>Muenster</city><region>NW</region><code>48155</code>
151        <country>Germany</country>
152      </postal>
153      <phone>+49 251 2807760</phone>   
154      <facsimile>+49 251 2807761</facsimile>   
155      <email>julian.reschke@greenbytes.de</email>       
156      <uri>http://greenbytes.de/tech/webdav/</uri>     
157    </address>
158  </author>
159
160  <date month="June" year="2008" day="17"/>
161
162<abstract>
163<t>
164   The Hypertext Transfer Protocol (HTTP) is an application-level
165   protocol for distributed, collaborative, hypermedia information
166   systems. HTTP has been in use by the World Wide Web global information
167   initiative since 1990. This document is Part 1 of the seven-part specification
168   that defines the protocol referred to as "HTTP/1.1" and, taken together,
169   obsoletes RFC 2616.  Part 1 provides an overview of HTTP and
170   its associated terminology, defines the "http" and "https" Uniform
171   Resource Identifier (URI) schemes, defines the generic message syntax
172   and parsing requirements for HTTP message frames, and describes
173   general security concerns for implementations.
174</t>
175</abstract>
176
177<note title="Editorial Note (To be removed by RFC Editor)">
178  <t>
179    Discussion of this draft should take place on the HTTPBIS working group
180    mailing list (ietf-http-wg@w3.org). The current issues list is
181    at <eref target="http://www.tools.ietf.org/wg/httpbis/trac/report/11"/>
182    and related documents (including fancy diffs) can be found at
183    <eref target="http://www.tools.ietf.org/wg/httpbis/"/>.
184  </t>
185  <t>
186    The changes in this draft are summarized in <xref target="changes.since.02"/>.
187  </t>
188</note>
189</front>
190<middle>
191<section title="Introduction" anchor="introduction">
192<t>
193   The Hypertext Transfer Protocol (HTTP) is an application-level
194   protocol for distributed, collaborative, hypermedia information
195   systems. HTTP has been in use by the World-Wide Web global
196   information initiative since 1990. The first version of HTTP, commonly
197   referred to as HTTP/0.9, was a simple protocol for raw data transfer
198   across the Internet with only a single method and no metadata.
199   HTTP/1.0, as defined by <xref target="RFC1945"/>, improved
200   the protocol by allowing messages to be in the format of MIME-like
201   messages, containing metadata about the data transferred and
202   modifiers on the request/response semantics. However, HTTP/1.0 did
203   not sufficiently take into consideration the effects of hierarchical
204   proxies, caching, the need for persistent connections, or name-based
205   virtual hosts. In addition, the proliferation of incompletely-implemented
206   applications calling themselves "HTTP/1.0" necessitated a
207   protocol version change in order for two communicating applications
208   to determine each other's true capabilities.
209</t>
210<t>
211   This document is Part 1 of the seven-part specification that defines
212   the protocol referred to as "HTTP/1.1", obsoleting <xref target="RFC2616"/>.
213   HTTP/1.1 remains compatible with HTTP/1.0 by including more stringent
214   requirements that enable reliable implementations and adding only
215   those new features that will either be safely ignored by an HTTP/1.0
216   recipient or only sent when communicating with a party advertising
217   compliance with HTTP/1.1.
218   Part 1 defines those aspects of HTTP/1.1 related to overall network
219   operation, message framing, interaction with transport protocols, and
220   URI schemes.
221</t>
222<t>
223   This document is currently disorganized in order to minimize the changes
224   between drafts and enable reviewers to see the smaller errata changes.
225   The next draft will reorganize the sections to better reflect the content.
226   In particular, the sections will be organized according to the typical
227   process of deciding when to use HTTP (URI schemes), overall network operation,
228   connection management, message framing, and generic message parsing.
229   The current mess reflects how widely dispersed these topics and associated
230   requirements had become in <xref target="RFC2616"/>.
231</t>
232
233<section title="Purpose" anchor="intro.purpose">
234<t>
235   Practical information systems require more functionality than simple
236   retrieval, including search, front-end update, and annotation. HTTP
237   allows an open-ended set of methods and headers that indicate the
238   purpose of a request <xref target="RFC2324"/>. It builds on the discipline of reference
239   provided by the Uniform Resource Identifier (URI) <xref target="RFC1630"/>, as a location
240   (URL) <xref target="RFC1738"/> or name (URN) <xref target="RFC1737"/>, for indicating the resource to which a
241   method is to be applied. Messages are passed in a format similar to
242   that used by Internet mail <xref target="RFC2822"/> as defined by the Multipurpose
243   Internet Mail Extensions (MIME) <xref target="RFC2045"/>.
244</t>
245<t>
246   HTTP is also used as a generic protocol for communication between
247   user agents and proxies/gateways to other Internet systems, including
248   those supported by the SMTP <xref target="RFC2821"/>, NNTP <xref target="RFC3977"/>, FTP <xref target="RFC959"/>, Gopher <xref target="RFC1436"/>,
249   and WAIS <xref target="WAIS"/> protocols. In this way, HTTP allows basic hypermedia
250   access to resources available from diverse applications.
251</t>
252</section>
253
254<section title="Requirements" anchor="intro.requirements">
255<t>
256   The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
257   "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this
258   document are to be interpreted as described in <xref target="RFC2119"/>.
259</t>
260<t>
261   An implementation is not compliant if it fails to satisfy one or more
262   of the MUST or REQUIRED level requirements for the protocols it
263   implements. An implementation that satisfies all the MUST or REQUIRED
264   level and all the SHOULD level requirements for its protocols is said
265   to be "unconditionally compliant"; one that satisfies all the MUST
266   level requirements but not all the SHOULD level requirements for its
267   protocols is said to be "conditionally compliant."
268</t>
269</section>
270
271<section title="Terminology" anchor="intro.terminology">
272<t>
273   This specification uses a number of terms to refer to the roles
274   played by participants in, and objects of, the HTTP communication.
275</t>
276<t>
277  <iref item="connection"/>
278  connection
279  <list>
280    <t>
281      A transport layer virtual circuit established between two programs
282      for the purpose of communication.
283    </t>
284  </list>
285</t>
286<t>
287  <iref item="message"/>
288  message
289  <list>
290    <t>
291      The basic unit of HTTP communication, consisting of a structured
292      sequence of octets matching the syntax defined in <xref target="http.message"/> and
293      transmitted via the connection.
294    </t>
295  </list>
296</t>
297<t>
298  <iref item="request"/>
299  request
300  <list>
301    <t>
302      An HTTP request message, as defined in <xref target="request"/>.
303    </t>
304  </list>
305</t>
306<t>
307  <iref item="response"/>
308  response
309  <list>
310    <t>
311      An HTTP response message, as defined in <xref target="response"/>.
312    </t>
313  </list>
314</t>
315<t>
316  <iref item="resource"/>
317  resource
318  <list>
319    <t>
320      A network data object or service that can be identified by a URI,
321      as defined in <xref target="uri"/>. Resources may be available in multiple
322      representations (e.g. multiple languages, data formats, size, and
323      resolutions) or vary in other ways.
324    </t>
325  </list>
326</t>
327<t>
328  <iref item="entity"/>
329  entity
330  <list>
331    <t>
332      The information transferred as the payload of a request or
333      response. An entity consists of metainformation in the form of
334      entity-header fields and content in the form of an entity-body, as
335      described in Section 4 of <xref target="Part3"/>.
336    </t>
337  </list>
338</t>
339<t>
340  <iref item="representation"/>
341  representation
342  <list>
343    <t>
344      An entity included with a response that is subject to content
345      negotiation, as described in Section 5 of <xref target="Part3"/>. There may exist multiple
346      representations associated with a particular response status.
347    </t>
348  </list>
349</t>
350<t>
351  <iref item="content negotiation"/>
352  content negotiation
353  <list>
354    <t>
355      The mechanism for selecting the appropriate representation when
356      servicing a request, as described in Section 5 of <xref target="Part3"/>. The
357      representation of entities in any response can be negotiated
358      (including error responses).
359    </t>
360  </list>
361</t>
362<t>
363  <iref item="variant"/>
364  variant
365  <list>
366    <t>
367      A resource may have one, or more than one, representation(s)
368      associated with it at any given instant. Each of these
369      representations is termed a `variant'.  Use of the term `variant'
370      does not necessarily imply that the resource is subject to content
371      negotiation.
372    </t>
373  </list>
374</t>
375<t>
376  <iref item="client"/>
377  client
378  <list>
379    <t>
380      A program that establishes connections for the purpose of sending
381      requests.
382    </t>
383  </list>
384</t>
385<t>
386  <iref item="user agent"/>
387  user agent
388  <list>
389    <t>
390      The client which initiates a request. These are often browsers,
391      editors, spiders (web-traversing robots), or other end user tools.
392    </t>
393  </list>
394</t>
395<t>
396  <iref item="server"/>
397  server
398  <list>
399    <t>
400      An application program that accepts connections in order to
401      service requests by sending back responses. Any given program may
402      be capable of being both a client and a server; our use of these
403      terms refers only to the role being performed by the program for a
404      particular connection, rather than to the program's capabilities
405      in general. Likewise, any server may act as an origin server,
406      proxy, gateway, or tunnel, switching behavior based on the nature
407      of each request.
408    </t>
409  </list>
410</t>
411<t>
412  <iref item="origin server"/>
413  origin server
414  <list>
415    <t>
416      The server on which a given resource resides or is to be created.
417    </t>
418  </list>
419</t>
420<t>
421  <iref item="proxy"/>
422  proxy
423  <list>
424    <t>
425      An intermediary program which acts as both a server and a client
426      for the purpose of making requests on behalf of other clients.
427      Requests are serviced internally or by passing them on, with
428      possible translation, to other servers. A proxy MUST implement
429      both the client and server requirements of this specification. A
430      "transparent proxy" is a proxy that does not modify the request or
431      response beyond what is required for proxy authentication and
432      identification. A "non-transparent proxy" is a proxy that modifies
433      the request or response in order to provide some added service to
434      the user agent, such as group annotation services, media type
435      transformation, protocol reduction, or anonymity filtering. Except
436      where either transparent or non-transparent behavior is explicitly
437      stated, the HTTP proxy requirements apply to both types of
438      proxies.
439    </t>
440  </list>
441</t>
442<t>
443  <iref item="gateway"/>
444  gateway
445  <list>
446    <t>
447      A server which acts as an intermediary for some other server.
448      Unlike a proxy, a gateway receives requests as if it were the
449      origin server for the requested resource; the requesting client
450      may not be aware that it is communicating with a gateway.
451    </t>
452  </list>
453</t>
454<t>
455  <iref item="tunnel"/>
456  tunnel
457  <list>
458    <t>
459      An intermediary program which is acting as a blind relay between
460      two connections. Once active, a tunnel is not considered a party
461      to the HTTP communication, though the tunnel may have been
462      initiated by an HTTP request. The tunnel ceases to exist when both
463      ends of the relayed connections are closed.
464    </t>
465  </list>
466</t>
467<t>
468  <iref item="cache"/>
469  cache
470  <list>
471    <t>
472      A program's local store of response messages and the subsystem
473      that controls its message storage, retrieval, and deletion. A
474      cache stores cacheable responses in order to reduce the response
475      time and network bandwidth consumption on future, equivalent
476      requests. Any client or server may include a cache, though a cache
477      cannot be used by a server that is acting as a tunnel.
478    </t>
479  </list>
480</t>
481<t>
482  <iref item="cacheable"/>
483  cacheable
484  <list>
485    <t>
486      A response is cacheable if a cache is allowed to store a copy of
487      the response message for use in answering subsequent requests. The
488      rules for determining the cacheability of HTTP responses are
489      defined in Section 1 of <xref target="Part6"/>. Even if a resource is cacheable, there may
490      be additional constraints on whether a cache can use the cached
491      copy for a particular request.
492    </t>
493  </list>
494</t>
495<t>
496  <iref item="upstream"/>
497  <iref item="downstream"/>
498  upstream/downstream
499  <list>
500    <t>
501      Upstream and downstream describe the flow of a message: all
502      messages flow from upstream to downstream.
503    </t>
504  </list>
505</t>
506<t>
507  <iref item="inbound"/>
508  <iref item="outbound"/>
509  inbound/outbound
510  <list>
511    <t>
512      Inbound and outbound refer to the request and response paths for
513      messages: "inbound" means "traveling toward the origin server",
514      and "outbound" means "traveling toward the user agent"
515    </t>
516  </list>
517</t>
518</section>
519
520<section title="Overall Operation" anchor="intro.overall.operation">
521<t>
522   HTTP is a request/response protocol. A client sends a
523   request to the server in the form of a request method, URI, and
524   protocol version, followed by a MIME-like message containing request
525   modifiers, client information, and possible body content over a
526   connection with a server. The server responds with a status line,
527   including the message's protocol version and a success or error code,
528   followed by a MIME-like message containing server information, entity
529   metainformation, and possible entity-body content. The relationship
530   between HTTP and MIME is described in Appendix A of <xref target="Part3"/>.
531</t>
532<t>
533   Most HTTP communication is initiated by a user agent and consists of
534   a request to be applied to a resource on some origin server. In the
535   simplest case, this may be accomplished via a single connection (v)
536   between the user agent (UA) and the origin server (O).
537</t>
538<figure><artwork type="drawing"><![CDATA[
539       request chain ------------------------>
540    UA -------------------v------------------- O
541       <----------------------- response chain
542]]></artwork></figure>
543<t>
544   A more complicated situation occurs when one or more intermediaries
545   are present in the request/response chain. There are three common
546   forms of intermediary: proxy, gateway, and tunnel. A proxy is a
547   forwarding agent, receiving requests for a URI in its absolute form,
548   rewriting all or part of the message, and forwarding the reformatted
549   request toward the server identified by the URI. A gateway is a
550   receiving agent, acting as a layer above some other server(s) and, if
551   necessary, translating the requests to the underlying server's
552   protocol. A tunnel acts as a relay point between two connections
553   without changing the messages; tunnels are used when the
554   communication needs to pass through an intermediary (such as a
555   firewall) even when the intermediary cannot understand the contents
556   of the messages.
557</t>
558<figure><artwork type="drawing"><![CDATA[
559       request chain -------------------------------------->
560    UA -----v----- A -----v----- B -----v----- C -----v----- O
561       <------------------------------------- response chain
562]]></artwork></figure>
563<t>
564   The figure above shows three intermediaries (A, B, and C) between the
565   user agent and origin server. A request or response message that
566   travels the whole chain will pass through four separate connections.
567   This distinction is important because some HTTP communication options
568   may apply only to the connection with the nearest, non-tunnel
569   neighbor, only to the end-points of the chain, or to all connections
570   along the chain. Although the diagram is linear, each participant may
571   be engaged in multiple, simultaneous communications. For example, B
572   may be receiving requests from many clients other than A, and/or
573   forwarding requests to servers other than C, at the same time that it
574   is handling A's request.
575</t>
576<t>
577   Any party to the communication which is not acting as a tunnel may
578   employ an internal cache for handling requests. The effect of a cache
579   is that the request/response chain is shortened if one of the
580   participants along the chain has a cached response applicable to that
581   request. The following illustrates the resulting chain if B has a
582   cached copy of an earlier response from O (via C) for a request which
583   has not been cached by UA or A.
584</t>
585<figure><artwork type="drawing"><![CDATA[
586          request chain ---------->
587       UA -----v----- A -----v----- B - - - - - - C - - - - - - O
588          <--------- response chain
589]]></artwork></figure>
590<t>
591   Not all responses are usefully cacheable, and some requests may
592   contain modifiers which place special requirements on cache behavior.
593   HTTP requirements for cache behavior and cacheable responses are
594   defined in Section 1 of <xref target="Part6"/>.
595</t>
596<t>
597   In fact, there are a wide variety of architectures and configurations
598   of caches and proxies currently being experimented with or deployed
599   across the World Wide Web. These systems include national hierarchies
600   of proxy caches to save transoceanic bandwidth, systems that
601   broadcast or multicast cache entries, organizations that distribute
602   subsets of cached data via CD-ROM, and so on. HTTP systems are used
603   in corporate intranets over high-bandwidth links, and for access via
604   PDAs with low-power radio links and intermittent connectivity. The
605   goal of HTTP/1.1 is to support the wide diversity of configurations
606   already deployed while introducing protocol constructs that meet the
607   needs of those who build web applications that require high
608   reliability and, failing that, at least reliable indications of
609   failure.
610</t>
611<t>
612   HTTP communication usually takes place over TCP/IP connections. The
613   default port is TCP 80 (<eref target="http://www.iana.org/assignments/port-numbers"/>), but other ports can be used. This does
614   not preclude HTTP from being implemented on top of any other protocol
615   on the Internet, or on other networks. HTTP only presumes a reliable
616   transport; any protocol that provides such guarantees can be used;
617   the mapping of the HTTP/1.1 request and response structures onto the
618   transport data units of the protocol in question is outside the scope
619   of this specification.
620</t>
621<t>
622   In HTTP/1.0, most implementations used a new connection for each
623   request/response exchange. In HTTP/1.1, a connection may be used for
624   one or more request/response exchanges, although connections may be
625   closed for a variety of reasons (see <xref target="persistent.connections"/>).
626</t>
627</section>
628</section>
629
630<section title="Notational Conventions and Generic Grammar" anchor="notation">
631
632<section title="Augmented BNF" anchor="notation.abnf">
633<t>
634   All of the mechanisms specified in this document are described in
635   both prose and an augmented Backus-Naur Form (BNF) similar to that
636   used by <xref target="RFC822ABNF"/>. Implementors will need to be familiar with the
637   notation in order to understand this specification. The augmented BNF
638   includes the following constructs:
639</t>
640<t>
641   name = definition
642  <list>
643    <t>
644      The name of a rule is simply the name itself (without any
645      enclosing "&lt;" and "&gt;") and is separated from its definition by the
646      equal "=" character. White space is only significant in that
647      indentation of continuation lines is used to indicate a rule
648      definition that spans more than one line. Certain basic rules are
649      in uppercase, such as SP, LWS, HTAB, CRLF, DIGIT, ALPHA, etc. Angle
650      brackets are used within definitions whenever their presence will
651      facilitate discerning the use of rule names.
652    </t>
653  </list>
654</t>
655<t>
656   "literal"
657  <list>
658    <t>
659      Quotation marks surround literal text. Unless stated otherwise,
660      the text is case-insensitive.
661    </t>
662  </list>
663</t>
664<t>
665   rule1 | rule2
666  <list>
667    <t>
668      Elements separated by a bar ("|") are alternatives, e.g., "yes |
669      no" will accept yes or no.
670    </t>
671  </list>
672</t>
673<t>
674   (rule1 rule2)
675  <list>
676    <t>
677      Elements enclosed in parentheses are treated as a single element.
678      Thus, "(elem (foo | bar) elem)" allows the token sequences "elem
679      foo elem" and "elem bar elem".
680    </t>
681  </list>
682</t>
683<t>
684   *rule
685  <list>
686    <t>
687      The character "*" preceding an element indicates repetition. The
688      full form is "&lt;n&gt;*&lt;m&gt;element" indicating at least &lt;n&gt; and at most
689      &lt;m&gt; occurrences of element. Default values are 0 and infinity so
690      that "*(element)" allows any number, including zero; "1*element"
691      requires at least one; and "1*2element" allows one or two.
692    </t>
693  </list>
694</t>
695<t>
696   [rule]
697  <list>
698    <t>
699      Square brackets enclose optional elements; "[foo bar]" is
700      equivalent to "*1(foo bar)".
701    </t>
702  </list>
703</t>
704<t>
705   N rule
706  <list>
707    <t>
708      Specific repetition: "&lt;n&gt;(element)" is equivalent to
709      "&lt;n&gt;*&lt;n&gt;(element)"; that is, exactly &lt;n&gt; occurrences of (element).
710      Thus 2DIGIT is a 2-digit number, and 3ALPHA is a string of three
711      alphabetic characters.
712    </t>
713  </list>
714</t>
715<t>
716   #rule
717  <list>
718    <t>
719      A construct "#" is defined, similar to "*", for defining lists of
720      elements. The full form is "&lt;n&gt;#&lt;m&gt;element" indicating at least
721      &lt;n&gt; and at most &lt;m&gt; elements, each separated by one or more commas
722      (",") and OPTIONAL linear white space (LWS). This makes the usual
723      form of lists very easy; a rule such as
724      <figure><artwork type="example"><![CDATA[
725   ( *LWS element *( *LWS "," *LWS element ))]]></artwork></figure>
726    </t>
727    <t>
728      can be shown as
729      <figure><artwork type="example"><![CDATA[
730   1#element]]></artwork></figure>
731    </t>
732    <t>
733      Wherever this construct is used, null elements are allowed, but do
734      not contribute to the count of elements present. That is,
735      "(element), , (element) " is permitted, but counts as only two
736      elements. Therefore, where at least one element is required, at
737      least one non-null element MUST be present. Default values are 0
738      and infinity so that "#element" allows any number, including zero;
739      "1#element" requires at least one; and "1#2element" allows one or
740      two.
741    </t>
742  </list>
743</t>
744<t>
745   ; comment
746  <list>
747    <t>
748      A semi-colon, set off some distance to the right of rule text,
749      starts a comment that continues to the end of line. This is a
750      simple way of including useful notes in parallel with the
751      specifications.
752    </t>
753  </list>
754</t>
755<t anchor="implied.LWS">
756  <iref item="implied *LWS" primary="true"/>
757   implied *LWS
758  <list>
759    <t>
760      The grammar described by this specification is word-based. Except
761      where noted otherwise, linear white space (LWS) can be included
762      between any two adjacent words (token or quoted-string), and
763      between adjacent words and separators, without changing the
764      interpretation of a field. At least one delimiter (LWS and/or
765      separators) MUST exist between any two tokens (for the definition
766      of "token" below), since they would otherwise be interpreted as a
767      single token.
768    </t>
769  </list>
770</t>
771</section>
772
773<section title="Basic Rules" anchor="basic.rules">
774<t anchor="core.rules">
775 
776 
777 
778 
779 
780 
781 
782 
783 
784 
785   The following rules are used throughout this specification to
786   describe basic parsing constructs. The US-ASCII coded character set
787   is defined by ANSI X3.4-1986 <xref target="USASCII"/>.
788</t>
789<figure><iref primary="true" item="Grammar" subitem="OCTET"/><iref primary="true" item="Grammar" subitem="CHAR"/><iref primary="true" item="Grammar" subitem="ALPHA"/><iref primary="true" item="Grammar" subitem="DIGIT"/><iref primary="true" item="Grammar" subitem="CTL"/><iref primary="true" item="Grammar" subitem="CR"/><iref primary="true" item="Grammar" subitem="LF"/><iref primary="true" item="Grammar" subitem="SP"/><iref primary="true" item="Grammar" subitem="HTAB"/><iref primary="true" item="Grammar" subitem="DQUOTE"/><artwork type="abnf2616"><![CDATA[
790  OCTET          = %x00-FF
791                   ; any 8-bit sequence of data
792  CHAR           = %x01-7F
793                   ; any US-ASCII character, excluding NUL
794  ALPHA          = %x41-5A | %x61-7A
795                   ; A-Z | a-z
796  DIGIT          = %x30-39
797                   ; any US-ASCII digit "0".."9"
798  CTL            = %x00-1F | %x7F
799                   ; (octets 0 - 31) and DEL (127)
800  CR             = %x0D
801                   ; US-ASCII CR, carriage return (13)
802  LF             = %x0A
803                   ; US-ASCII LF, linefeed (10)
804  SP             = %x20
805                   ; US-ASCII SP, space (32)
806  HTAB           = %x09
807                   ; US-ASCII HT, horizontal-tab (9)
808  DQUOTE         = %x22
809                   ; US-ASCII double-quote mark (34)
810]]></artwork></figure>
811<t anchor="rule.CRLF">
812 
813   HTTP/1.1 defines the sequence CR LF as the end-of-line marker for all
814   protocol elements except the entity-body (see <xref target="tolerant.applications"/> for
815   tolerant applications). The end-of-line marker within an entity-body
816   is defined by its associated media type, as described in Section 3.3 of <xref target="Part3"/>.
817</t>
818<figure><iref primary="true" item="Grammar" subitem="CRLF"/><artwork type="abnf2616"><![CDATA[
819  CRLF           = CR LF
820]]></artwork></figure>
821<t anchor="rule.LWS">
822 
823   HTTP/1.1 header field values can be folded onto multiple lines if the
824   continuation line begins with a space or horizontal tab. All linear
825   white space, including folding, has the same semantics as SP. A
826   recipient MAY replace any linear white space with a single SP before
827   interpreting the field value or forwarding the message downstream.
828</t>
829<figure><iref primary="true" item="Grammar" subitem="LWS"/><artwork type="abnf2616"><![CDATA[
830  LWS            = [CRLF] 1*( SP | HTAB )
831]]></artwork></figure>
832<t anchor="rule.TEXT">
833 
834   The TEXT rule is only used for descriptive field contents and values
835   that are not intended to be interpreted by the message parser. Words
836   of *TEXT MAY contain characters from character sets other than ISO-8859-1
837   <xref target="ISO-8859-1"/> only when encoded according to the rules of
838   <xref target="RFC2047"/>.
839</t>
840<figure><iref primary="true" item="Grammar" subitem="TEXT"/><artwork type="abnf2616"><![CDATA[
841  TEXT           = %x20-7E | %x80-FF | LWS
842                 ; any OCTET except CTLs, but including LWS
843]]></artwork></figure>
844<t>
845   A CRLF is allowed in the definition of TEXT only as part of a header
846   field continuation. It is expected that the folding LWS will be
847   replaced with a single SP before interpretation of the TEXT value.
848</t>
849<t anchor="rule.HEX">
850 
851   Hexadecimal numeric characters are used in several protocol elements.
852</t>
853<figure><iref primary="true" item="Grammar" subitem="HEX"/><artwork type="abnf2616"><![CDATA[
854  HEX            = "A" | "B" | "C" | "D" | "E" | "F"
855                 | "a" | "b" | "c" | "d" | "e" | "f" | DIGIT
856]]></artwork></figure>
857<t anchor="rule.token.separators">
858 
859 
860 
861   Many HTTP/1.1 header field values consist of words separated by LWS
862   or special characters. These special characters MUST be in a quoted
863   string to be used within a parameter value (as defined in
864   <xref target="transfer.codings"/>).
865</t>
866<figure><iref primary="true" item="Grammar" subitem="token"/><iref primary="true" item="Grammar" subitem="tchar"/><iref primary="true" item="Grammar" subitem="separators"/><artwork type="abnf2616"><![CDATA[
867  separators     = "(" | ")" | "<" | ">" | "@"
868                 | "," | ";" | ":" | "\" | DQUOTE
869                 | "/" | "[" | "]" | "?" | "="
870                 | "{" | "}" | SP | HTAB
871                 
872  tchar          = "!" | "#" | "$" | "%" | "&" | "'" | "*"
873                 | "+" | "-" | "." | "^" | "_" | "`" | "|" | "~"
874                 | DIGIT | ALPHA
875                 ; any CHAR except CTLs or separators
876                 
877  token          = 1*tchar
878]]></artwork></figure>
879<t anchor="rule.comment">
880 
881 
882   Comments can be included in some HTTP header fields by surrounding
883   the comment text with parentheses. Comments are only allowed in
884   fields containing "comment" as part of their field value definition.
885   In all other fields, parentheses are considered part of the field
886   value.
887</t>
888<figure><iref primary="true" item="Grammar" subitem="comment"/><iref primary="true" item="Grammar" subitem="ctext"/><artwork type="abnf2616"><![CDATA[
889  comment        = "(" *( ctext | quoted-pair | comment ) ")"
890  ctext          = <any TEXT excluding "(" and ")">
891]]></artwork></figure>
892<t anchor="rule.quoted-string">
893 
894 
895   A string of text is parsed as a single word if it is quoted using
896   double-quote marks.
897</t>
898<figure><iref primary="true" item="Grammar" subitem="quoted-string"/><iref primary="true" item="Grammar" subitem="qdtext"/><artwork type="abnf2616"><![CDATA[
899  quoted-string  = ( DQUOTE *(qdtext | quoted-pair ) DQUOTE )
900  qdtext         = <any TEXT excluding DQUOTE and "\">
901]]></artwork></figure>
902<t anchor="rule.quoted-pair">
903 
904 
905   The backslash character ("\") MAY be used as a single-character
906   quoting mechanism only within quoted-string and comment constructs.
907</t>
908<figure><iref primary="true" item="Grammar" subitem="quoted-text"/><iref primary="true" item="Grammar" subitem="quoted-pair"/><artwork type="abnf2616"><![CDATA[
909  quoted-text    = %x01-09 |
910                   %x0B-0C |
911                   %x0E-FF ; Characters excluding NUL, CR and LF
912  quoted-pair    = "\" quoted-text
913]]></artwork></figure>
914</section>
915
916<section title="ABNF Rules defined in other Parts of the Specification" anchor="abnf.dependencies">
917 
918 
919 
920 
921 
922 
923 
924 
925<t>
926  The ABNF rules below are defined in other parts:
927</t>
928<figure><!-- Part2--><artwork type="abnf2616"><![CDATA[
929  request-header  = <request-header, defined in [Part2], Section 4>
930  response-header = <response-header, defined in [Part2], Section 6>
931]]></artwork></figure>
932<figure><!-- Part3--><artwork type="abnf2616"><![CDATA[
933  accept-params   = <accept-params, defined in [Part3], Section 6.1>
934  entity-body     = <entity-body, defined in [Part3], Section 4.2>
935  entity-header   = <entity-header, defined in [Part3], Section 4.1>
936]]></artwork></figure>
937<figure><!-- Part6--><artwork type="abnf2616"><![CDATA[
938  Cache-Control   = <Cache-Control, defined in [Part6], Section 16.4>
939  Pragma          = <Pragma, defined in [Part6], Section 16.4>
940  Warning         = <Warning, defined in [Part6], Section 16.6>
941]]></artwork></figure>
942</section>
943
944</section>
945
946<section title="Protocol Parameters" anchor="protocol.parameters">
947
948<section title="HTTP Version" anchor="http.version">
949 
950 
951<t>
952   HTTP uses a "&lt;major&gt;.&lt;minor&gt;" numbering scheme to indicate versions
953   of the protocol. The protocol versioning policy is intended to allow
954   the sender to indicate the format of a message and its capacity for
955   understanding further HTTP communication, rather than the features
956   obtained via that communication. No change is made to the version
957   number for the addition of message components which do not affect
958   communication behavior or which only add to extensible field values.
959   The &lt;minor&gt; number is incremented when the changes made to the
960   protocol add features which do not change the general message parsing
961   algorithm, but which may add to the message semantics and imply
962   additional capabilities of the sender. The &lt;major&gt; number is
963   incremented when the format of a message within the protocol is
964   changed. See <xref target="RFC2145"/> for a fuller explanation.
965</t>
966<t>
967   The version of an HTTP message is indicated by an HTTP-Version field
968   in the first line of the message. HTTP-Version is case-sensitive.
969</t>
970<figure><iref primary="true" item="Grammar" subitem="HTTP-Version"/><iref primary="true" item="Grammar" subitem="HTTP-Prot-Name"/><artwork type="abnf2616"><![CDATA[
971  HTTP-Version   = HTTP-Prot-Name "/" 1*DIGIT "." 1*DIGIT
972  HTTP-Prot-Name = %x48.54.54.50 ; "HTTP", case-sensitive
973]]></artwork></figure>
974<t>
975   Note that the major and minor numbers MUST be treated as separate
976   integers and that each MAY be incremented higher than a single digit.
977   Thus, HTTP/2.4 is a lower version than HTTP/2.13, which in turn is
978   lower than HTTP/12.3. Leading zeros MUST be ignored by recipients and
979   MUST NOT be sent.
980</t>
981<t>
982   An application that sends a request or response message that includes
983   HTTP-Version of "HTTP/1.1" MUST be at least conditionally compliant
984   with this specification. Applications that are at least conditionally
985   compliant with this specification SHOULD use an HTTP-Version of
986   "HTTP/1.1" in their messages, and MUST do so for any message that is
987   not compatible with HTTP/1.0. For more details on when to send
988   specific HTTP-Version values, see <xref target="RFC2145"/>.
989</t>
990<t>
991   The HTTP version of an application is the highest HTTP version for
992   which the application is at least conditionally compliant.
993</t>
994<t>
995   Proxy and gateway applications need to be careful when forwarding
996   messages in protocol versions different from that of the application.
997   Since the protocol version indicates the protocol capability of the
998   sender, a proxy/gateway MUST NOT send a message with a version
999   indicator which is greater than its actual version. If a higher
1000   version request is received, the proxy/gateway MUST either downgrade
1001   the request version, or respond with an error, or switch to tunnel
1002   behavior.
1003</t>
1004<t>
1005   Due to interoperability problems with HTTP/1.0 proxies discovered
1006   since the publication of <xref target="RFC2068"/>, caching proxies MUST, gateways
1007   MAY, and tunnels MUST NOT upgrade the request to the highest version
1008   they support. The proxy/gateway's response to that request MUST be in
1009   the same major version as the request.
1010</t>
1011<t>
1012  <list>
1013    <t>
1014      Note: Converting between versions of HTTP may involve modification
1015      of header fields required or forbidden by the versions involved.
1016    </t>
1017  </list>
1018</t>
1019</section>
1020
1021<section title="Uniform Resource Identifiers" anchor="uri">
1022<t>
1023   URIs have been known by many names: WWW addresses, Universal Document
1024   Identifiers, Universal Resource Identifiers <xref target="RFC1630"/>, and finally the
1025   combination of Uniform Resource Locators (URL) <xref target="RFC1738"/> and Names (URN)
1026   <xref target="RFC1737"/>. As far as HTTP is concerned, Uniform Resource Identifiers are
1027   simply formatted strings which identify--via name, location, or any
1028   other characteristic--a resource.
1029</t>
1030
1031<section title="General Syntax" anchor="general.syntax">
1032 
1033 
1034 
1035 
1036 
1037 
1038 
1039 
1040<t>
1041   URIs in HTTP can be represented in absolute form or relative to some
1042   known base URI <xref target="RFC1808"/>, depending upon the context of their use. The two
1043   forms are differentiated by the fact that absolute URIs always begin
1044   with a scheme name followed by a colon. For definitive information on
1045   URL syntax and semantics, see "Uniform Resource Identifiers (URI):
1046   Generic Syntax and Semantics," <xref target="RFC2396"/> (which replaces <xref target="RFC1738"/>
1047   and <xref target="RFC1808"/>). This specification adopts the
1048   definitions of "URI-reference", "absoluteURI", "fragment", "relativeURI", "port",
1049   "host", "abs_path", "query", and "authority" from that specification:
1050</t>
1051<figure><iref primary="true" item="Grammar" subitem="absoluteURI"/><iref primary="true" item="Grammar" subitem="authority"/><iref primary="true" item="Grammar" subitem="path-absolute"/><iref primary="true" item="Grammar" subitem="port"/><iref primary="true" item="Grammar" subitem="query"/><iref primary="true" item="Grammar" subitem="relativeURI"/><iref primary="true" item="Grammar" subitem="uri-host"/><artwork type="abnf2616"><![CDATA[
1052  absoluteURI   = <absoluteURI, defined in [RFC2396], Section 3>
1053  authority     = <authority, defined in [RFC2396], Section 3.2>
1054  fragment      = <fragment, defined in [RFC2396], Section 4.1>
1055  path-absolute = <abs_path, defined in [RFC2396], Section 3>
1056  port          = <port, defined in [RFC2396], Section 3.2.2>
1057  query         = <query, defined in [RFC2396], Section 3.4>
1058  relativeURI   = <relativeURI, defined in [RFC2396], Section 5>
1059  uri-host      = <host, defined in [RFC2396], Section 3.2.2>
1060]]></artwork></figure>
1061<t>
1062   HTTP does not place any a priori limit on the length of
1063   a URI. Servers MUST be able to handle the URI of any resource they
1064   serve, and SHOULD be able to handle URIs of unbounded length if they
1065   provide GET-based forms that could generate such URIs. A server
1066   SHOULD return 414 (Request-URI Too Long) status if a URI is longer
1067   than the server can handle (see Section 9.4.15 of <xref target="Part2"/>).
1068</t>
1069<t>
1070  <list>
1071    <t>
1072      Note: Servers ought to be cautious about depending on URI lengths
1073      above 255 bytes, because some older client or proxy
1074      implementations might not properly support these lengths.
1075    </t>
1076  </list>
1077</t>
1078</section>
1079
1080<section title="http URL" anchor="http.url">
1081 
1082<t>
1083   The "http" scheme is used to locate network resources via the HTTP
1084   protocol. This section defines the scheme-specific syntax and
1085   semantics for http URLs.
1086</t>
1087<figure><iref primary="true" item="Grammar" subitem="http-URL"/><artwork type="abnf2616"><![CDATA[
1088  http-URL = "http:" "//" uri-host [ ":" port ]
1089             [ path-absolute [ "?" query ]]
1090]]></artwork></figure>
1091<t>
1092   If the port is empty or not given, port 80 is assumed. The semantics
1093   are that the identified resource is located at the server listening
1094   for TCP connections on that port of that host, and the Request-URI
1095   for the resource is path-absolute (<xref target="request-uri"/>). The use of IP addresses
1096   in URLs SHOULD be avoided whenever possible (see <xref target="RFC1900"/>). If
1097   the path-absolute is not present in the URL, it MUST be given as "/" when
1098   used as a Request-URI for a resource (<xref target="request-uri"/>). If a proxy
1099   receives a host name which is not a fully qualified domain name, it
1100   MAY add its domain to the host name it received. If a proxy receives
1101   a fully qualified domain name, the proxy MUST NOT change the host
1102   name.
1103</t>
1104</section>
1105
1106<section title="URI Comparison" anchor="uri.comparison">
1107<t>
1108   When comparing two URIs to decide if they match or not, a client
1109   SHOULD use a case-sensitive octet-by-octet comparison of the entire
1110   URIs, with these exceptions:
1111  <list style="symbols">
1112    <t>A port that is empty or not given is equivalent to the default
1113        port for that URI-reference;</t>
1114    <t>Comparisons of host names MUST be case-insensitive;</t>
1115    <t>Comparisons of scheme names MUST be case-insensitive;</t>
1116    <t>An empty path-absolute is equivalent to an path-absolute of "/".</t>
1117  </list>
1118</t>
1119<t>
1120   Characters other than those in the "reserved" set (see
1121   <xref target="RFC2396"/>) are equivalent to their ""%" HEX HEX" encoding.
1122</t>
1123<t>
1124   For example, the following three URIs are equivalent:
1125</t>
1126<figure><artwork type="example"><![CDATA[
1127   http://example.com:80/~smith/home.html
1128   http://EXAMPLE.com/%7Esmith/home.html
1129   http://EXAMPLE.com:/%7esmith/home.html
1130]]></artwork></figure>
1131</section>
1132</section>
1133
1134<section title="Date/Time Formats" anchor="date.time.formats">
1135<section title="Full Date" anchor="full.date">
1136 
1137 
1138 
1139 
1140 
1141 
1142 
1143 
1144 
1145 
1146 
1147 
1148 
1149<t>
1150   HTTP applications have historically allowed three different formats
1151   for the representation of date/time stamps:
1152</t>
1153<figure><artwork type="example"><![CDATA[
1154   Sun, 06 Nov 1994 08:49:37 GMT  ; RFC 822, updated by RFC 1123
1155   Sunday, 06-Nov-94 08:49:37 GMT ; obsolete RFC 850 format
1156   Sun Nov  6 08:49:37 1994       ; ANSI C's asctime() format
1157]]></artwork></figure>
1158<t>
1159   The first format is preferred as an Internet standard and represents
1160   a fixed-length subset of that defined by <xref target="RFC1123"/> (an update to
1161   <xref target="RFC822"/>). The other formats are described here only for
1162   compatibility with obsolete implementations.
1163   HTTP/1.1 clients and servers that parse the date value MUST accept
1164   all three formats (for compatibility with HTTP/1.0), though they MUST
1165   only generate the RFC 1123 format for representing HTTP-date values
1166   in header fields. See <xref target="tolerant.applications"/> for further information.
1167</t>
1168<t><list><t>
1169      Note: Recipients of date values are encouraged to be robust in
1170      accepting date values that may have been sent by non-HTTP
1171      applications, as is sometimes the case when retrieving or posting
1172      messages via proxies/gateways to SMTP or NNTP.
1173</t></list></t>
1174<t>
1175   All HTTP date/time stamps MUST be represented in Greenwich Mean Time
1176   (GMT), without exception. For the purposes of HTTP, GMT is exactly
1177   equal to UTC (Coordinated Universal Time). This is indicated in the
1178   first two formats by the inclusion of "GMT" as the three-letter
1179   abbreviation for time zone, and MUST be assumed when reading the
1180   asctime format. HTTP-date is case sensitive and MUST NOT include
1181   additional LWS beyond that specifically included as SP in the
1182   grammar.
1183</t>
1184<figure><iref primary="true" item="Grammar" subitem="HTTP-date"/><iref primary="true" item="Grammar" subitem="rfc1123-date"/><iref primary="true" item="Grammar" subitem="obsolete-date"/><iref primary="true" item="Grammar" subitem="rfc850-date"/><iref primary="true" item="Grammar" subitem="asctime-date"/><iref primary="true" item="Grammar" subitem="date1"/><iref primary="true" item="Grammar" subitem="date2"/><iref primary="true" item="Grammar" subitem="date3"/><iref primary="true" item="Grammar" subitem="time"/><iref primary="true" item="Grammar" subitem="wkday"/><iref primary="true" item="Grammar" subitem="weekday"/><iref primary="true" item="Grammar" subitem="month"/><artwork type="abnf2616"><![CDATA[
1185  HTTP-date    = rfc1123-date | obsolete-date
1186  obsolete-date = rfc850-date | asctime-date
1187  rfc1123-date = wkday "," SP date1 SP time SP "GMT"
1188  rfc850-date  = weekday "," SP date2 SP time SP "GMT"
1189  asctime-date = wkday SP date3 SP time SP 4DIGIT
1190  date1        = 2DIGIT SP month SP 4DIGIT
1191                 ; day month year (e.g., 02 Jun 1982)
1192  date2        = 2DIGIT "-" month "-" 2DIGIT
1193                 ; day-month-year (e.g., 02-Jun-82)
1194  date3        = month SP ( 2DIGIT | ( SP 1DIGIT ))
1195                 ; month day (e.g., Jun  2)
1196  time         = 2DIGIT ":" 2DIGIT ":" 2DIGIT
1197                 ; 00:00:00 - 23:59:59
1198  wkday        = "Mon" | "Tue" | "Wed"
1199               | "Thu" | "Fri" | "Sat" | "Sun"
1200  weekday      = "Monday" | "Tuesday" | "Wednesday"
1201               | "Thursday" | "Friday" | "Saturday" | "Sunday"
1202  month        = "Jan" | "Feb" | "Mar" | "Apr"
1203               | "May" | "Jun" | "Jul" | "Aug"
1204               | "Sep" | "Oct" | "Nov" | "Dec"
1205]]></artwork></figure>
1206<t>
1207      Note: HTTP requirements for the date/time stamp format apply only
1208      to their usage within the protocol stream. Clients and servers are
1209      not required to use these formats for user presentation, request
1210      logging, etc.
1211</t>
1212</section>
1213</section>
1214
1215<section title="Transfer Codings" anchor="transfer.codings">
1216 
1217 
1218 
1219<t>
1220   Transfer-coding values are used to indicate an encoding
1221   transformation that has been, can be, or may need to be applied to an
1222   entity-body in order to ensure "safe transport" through the network.
1223   This differs from a content coding in that the transfer-coding is a
1224   property of the message, not of the original entity.
1225</t>
1226<figure><iref primary="true" item="Grammar" subitem="transfer-coding"/><iref primary="true" item="Grammar" subitem="transfer-extension"/><artwork type="abnf2616"><![CDATA[
1227  transfer-coding         = "chunked" | transfer-extension
1228  transfer-extension      = token *( ";" parameter )
1229]]></artwork></figure>
1230<t anchor="rule.parameter">
1231 
1232 
1233 
1234   Parameters are in  the form of attribute/value pairs.
1235</t>
1236<figure><iref primary="true" item="Grammar" subitem="parameter"/><iref primary="true" item="Grammar" subitem="attribute"/><iref primary="true" item="Grammar" subitem="value"/><artwork type="abnf2616"><![CDATA[
1237  parameter               = attribute "=" value
1238  attribute               = token
1239  value                   = token | quoted-string
1240]]></artwork></figure>
1241<t>
1242   All transfer-coding values are case-insensitive. HTTP/1.1 uses
1243   transfer-coding values in the TE header field (<xref target="header.te"/>) and in
1244   the Transfer-Encoding header field (<xref target="header.transfer-encoding"/>).
1245</t>
1246<t>
1247   Whenever a transfer-coding is applied to a message-body, the set of
1248   transfer-codings MUST include "chunked", unless the message is
1249   terminated by closing the connection. When the "chunked" transfer-coding
1250   is used, it MUST be the last transfer-coding applied to the
1251   message-body. The "chunked" transfer-coding MUST NOT be applied more
1252   than once to a message-body. These rules allow the recipient to
1253   determine the transfer-length of the message (<xref target="message.length"/>).
1254</t>
1255<t>
1256   Transfer-codings are analogous to the Content-Transfer-Encoding
1257   values of MIME <xref target="RFC2045"/>, which were designed to enable safe transport of
1258   binary data over a 7-bit transport service. However, safe transport
1259   has a different focus for an 8bit-clean transfer protocol. In HTTP,
1260   the only unsafe characteristic of message-bodies is the difficulty in
1261   determining the exact body length (<xref target="message.length"/>), or the desire to
1262   encrypt data over a shared transport.
1263</t>
1264<t>
1265   The Internet Assigned Numbers Authority (IANA) acts as a registry for
1266   transfer-coding value tokens. Initially, the registry contains the
1267   following tokens: "chunked" (<xref target="chunked.transfer.encoding"/>),
1268   "gzip", "compress", and "deflate" (Section 3.2 of <xref target="Part3"/>).
1269</t>
1270<t>
1271   New transfer-coding value tokens SHOULD be registered in the same way
1272   as new content-coding value tokens (Section 3.2 of <xref target="Part3"/>).
1273</t>
1274<t>
1275   A server which receives an entity-body with a transfer-coding it does
1276   not understand SHOULD return 501 (Not Implemented), and close the
1277   connection. A server MUST NOT send transfer-codings to an HTTP/1.0
1278   client.
1279</t>
1280
1281<section title="Chunked Transfer Coding" anchor="chunked.transfer.encoding">
1282 
1283 
1284 
1285 
1286 
1287 
1288 
1289 
1290 
1291<t>
1292   The chunked encoding modifies the body of a message in order to
1293   transfer it as a series of chunks, each with its own size indicator,
1294   followed by an OPTIONAL trailer containing entity-header fields. This
1295   allows dynamically produced content to be transferred along with the
1296   information necessary for the recipient to verify that it has
1297   received the full message.
1298</t>
1299<figure><iref primary="true" item="Grammar" subitem="Chunked-Body"/><iref primary="true" item="Grammar" subitem="chunk"/><iref primary="true" item="Grammar" subitem="chunk-size"/><iref primary="true" item="Grammar" subitem="last-chunk"/><iref primary="true" item="Grammar" subitem="chunk-extension"/><iref primary="true" item="Grammar" subitem="chunk-ext-name"/><iref primary="true" item="Grammar" subitem="chunk-ext-val"/><iref primary="true" item="Grammar" subitem="chunk-data"/><iref primary="true" item="Grammar" subitem="trailer-part"/><artwork type="abnf2616"><![CDATA[
1300  Chunked-Body   = *chunk
1301                   last-chunk
1302                   trailer-part
1303                   CRLF
1304 
1305  chunk          = chunk-size [ chunk-extension ] CRLF
1306                   chunk-data CRLF
1307  chunk-size     = 1*HEX
1308  last-chunk     = 1*("0") [ chunk-extension ] CRLF
1309 
1310  chunk-extension= *( ";" chunk-ext-name [ "=" chunk-ext-val ] )
1311  chunk-ext-name = token
1312  chunk-ext-val  = token | quoted-string
1313  chunk-data     = 1*OCTET ; a sequence of chunk-size octets
1314  trailer-part   = *(entity-header CRLF)
1315]]></artwork></figure>
1316<t>
1317   The chunk-size field is a string of hex digits indicating the size of
1318   the chunk-data in octets. The chunked encoding is ended by any chunk whose size is
1319   zero, followed by the trailer, which is terminated by an empty line.
1320</t>
1321<t>
1322   The trailer allows the sender to include additional HTTP header
1323   fields at the end of the message. The Trailer header field can be
1324   used to indicate which header fields are included in a trailer (see
1325   <xref target="header.trailer"/>).
1326</t>
1327<t>
1328   A server using chunked transfer-coding in a response MUST NOT use the
1329   trailer for any header fields unless at least one of the following is
1330   true:
1331  <list style="numbers">
1332    <t>the request included a TE header field that indicates "trailers" is
1333     acceptable in the transfer-coding of the  response, as described in
1334     <xref target="header.te"/>; or,</t>
1335
1336    <t>the server is the origin server for the response, the trailer
1337     fields consist entirely of optional metadata, and the recipient
1338     could use the message (in a manner acceptable to the origin server)
1339     without receiving this metadata.  In other words, the origin server
1340     is willing to accept the possibility that the trailer fields might
1341     be silently discarded along the path to the client.</t>
1342  </list>
1343</t>
1344<t>
1345   This requirement prevents an interoperability failure when the
1346   message is being received by an HTTP/1.1 (or later) proxy and
1347   forwarded to an HTTP/1.0 recipient. It avoids a situation where
1348   compliance with the protocol would have necessitated a possibly
1349   infinite buffer on the proxy.
1350</t>
1351<t>
1352   A process for decoding the "chunked" transfer-coding
1353   can be represented in pseudo-code as:
1354</t>
1355<figure><artwork type="code"><![CDATA[
1356    length := 0
1357    read chunk-size, chunk-extension (if any) and CRLF
1358    while (chunk-size > 0) {
1359       read chunk-data and CRLF
1360       append chunk-data to entity-body
1361       length := length + chunk-size
1362       read chunk-size and CRLF
1363    }
1364    read entity-header
1365    while (entity-header not empty) {
1366       append entity-header to existing header fields
1367       read entity-header
1368    }
1369    Content-Length := length
1370    Remove "chunked" from Transfer-Encoding
1371]]></artwork></figure>
1372<t>
1373   All HTTP/1.1 applications MUST be able to receive and decode the
1374   "chunked" transfer-coding, and MUST ignore chunk-extension extensions
1375   they do not understand.
1376</t>
1377</section>
1378</section>
1379
1380<section title="Product Tokens" anchor="product.tokens">
1381 
1382 
1383<t>
1384   Product tokens are used to allow communicating applications to
1385   identify themselves by software name and version. Most fields using
1386   product tokens also allow sub-products which form a significant part
1387   of the application to be listed, separated by white space. By
1388   convention, the products are listed in order of their significance
1389   for identifying the application.
1390</t>
1391<figure><iref primary="true" item="Grammar" subitem="product"/><iref primary="true" item="Grammar" subitem="product-version"/><artwork type="abnf2616"><![CDATA[
1392  product         = token ["/" product-version]
1393  product-version = token
1394]]></artwork></figure>
1395<t>
1396   Examples:
1397</t>
1398<figure><artwork type="example"><![CDATA[
1399    User-Agent: CERN-LineMode/2.15 libwww/2.17b3
1400    Server: Apache/0.8.4
1401]]></artwork></figure>
1402<t>
1403   Product tokens SHOULD be short and to the point. They MUST NOT be
1404   used for advertising or other non-essential information. Although any
1405   token character MAY appear in a product-version, this token SHOULD
1406   only be used for a version identifier (i.e., successive versions of
1407   the same product SHOULD only differ in the product-version portion of
1408   the product value).
1409</t>
1410</section>
1411
1412</section>
1413
1414<section title="HTTP Message" anchor="http.message">
1415
1416<section title="Message Types" anchor="message.types">
1417 
1418 
1419 
1420<t>
1421   HTTP messages consist of requests from client to server and responses
1422   from server to client.
1423</t>
1424<figure><iref primary="true" item="Grammar" subitem="HTTP-message"/><artwork type="abnf2616"><![CDATA[
1425  HTTP-message   = Request | Response     ; HTTP/1.1 messages
1426]]></artwork></figure>
1427<t>
1428   Request (<xref target="request"/>) and Response (<xref target="response"/>) messages use the generic
1429   message format of <xref target="RFC2822"/> for transferring entities (the payload
1430   of the message). Both types of message consist of a start-line, zero
1431   or more header fields (also known as "headers"), an empty line (i.e.,
1432   a line with nothing preceding the CRLF) indicating the end of the
1433   header fields, and possibly a message-body.
1434</t>
1435<figure><iref primary="true" item="Grammar" subitem="generic-message"/><iref primary="true" item="Grammar" subitem="start-line"/><artwork type="abnf2616"><![CDATA[
1436  generic-message = start-line
1437                    *(message-header CRLF)
1438                    CRLF
1439                    [ message-body ]
1440  start-line      = Request-Line | Status-Line
1441]]></artwork></figure>
1442<t>
1443   In the interest of robustness, servers SHOULD ignore any empty
1444   line(s) received where a Request-Line is expected. In other words, if
1445   the server is reading the protocol stream at the beginning of a
1446   message and receives a CRLF first, it should ignore the CRLF.
1447</t>
1448<t>
1449   Certain buggy HTTP/1.0 client implementations generate extra CRLF's
1450   after a POST request. To restate what is explicitly forbidden by the
1451   BNF, an HTTP/1.1 client MUST NOT preface or follow a request with an
1452   extra CRLF.
1453</t>
1454</section>
1455
1456<section title="Message Headers" anchor="message.headers">
1457 
1458 
1459 
1460 
1461<t>
1462   HTTP header fields, which include general-header (<xref target="general.header.fields"/>),
1463   request-header (Section 4 of <xref target="Part2"/>), response-header (Section 6 of <xref target="Part2"/>), and
1464   entity-header (Section 4.1 of <xref target="Part3"/>) fields, follow the same generic format as
1465   that given in Section 2.1 of <xref target="RFC2822"/>. Each header field consists
1466   of a name followed by a colon (":") and the field value. Field names
1467   are case-insensitive. The field value MAY be preceded by any amount
1468   of LWS, though a single SP is preferred. Header fields can be
1469   extended over multiple lines by preceding each extra line with at
1470   least one SP or HTAB. Applications ought to follow "common form", where
1471   one is known or indicated, when generating HTTP constructs, since
1472   there might exist some implementations that fail to accept anything
1473   beyond the common forms.
1474</t>
1475<figure><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"/><artwork type="abnf2616"><![CDATA[
1476  message-header = field-name ":" [ field-value ]
1477  field-name     = token
1478  field-value    = *( field-content | LWS )
1479  field-content  = <field content>
1480                   ; the OCTETs making up the field-value
1481                   ; and consisting of either *TEXT or combinations
1482                   ; of token, separators, and quoted-string
1483]]></artwork></figure>
1484<t>
1485   The field-content does not include any leading or trailing LWS:
1486   linear white space occurring before the first non-whitespace
1487   character of the field-value or after the last non-whitespace
1488   character of the field-value. Such leading or trailing LWS MAY be
1489   removed without changing the semantics of the field value. Any LWS
1490   that occurs between field-content MAY be replaced with a single SP
1491   before interpreting the field value or forwarding the message
1492   downstream.
1493</t>
1494<t>
1495   The order in which header fields with differing field names are
1496   received is not significant. However, it is "good practice" to send
1497   general-header fields first, followed by request-header or response-header
1498   fields, and ending with the entity-header fields.
1499</t>
1500<t>
1501   Multiple message-header fields with the same field-name MAY be
1502   present in a message if and only if the entire field-value for that
1503   header field is defined as a comma-separated list [i.e., #(values)].
1504   It MUST be possible to combine the multiple header fields into one
1505   "field-name: field-value" pair, without changing the semantics of the
1506   message, by appending each subsequent field-value to the first, each
1507   separated by a comma. The order in which header fields with the same
1508   field-name are received is therefore significant to the
1509   interpretation of the combined field value, and thus a proxy MUST NOT
1510   change the order of these field values when a message is forwarded.
1511</t>
1512</section>
1513
1514<section title="Message Body" anchor="message.body">
1515 
1516<t>
1517   The message-body (if any) of an HTTP message is used to carry the
1518   entity-body associated with the request or response. The message-body
1519   differs from the entity-body only when a transfer-coding has been
1520   applied, as indicated by the Transfer-Encoding header field (<xref target="header.transfer-encoding"/>).
1521</t>
1522<figure><iref primary="true" item="Grammar" subitem="message-body"/><artwork type="abnf2616"><![CDATA[
1523  message-body = entity-body
1524               | <entity-body encoded as per Transfer-Encoding>
1525]]></artwork></figure>
1526<t>
1527   Transfer-Encoding MUST be used to indicate any transfer-codings
1528   applied by an application to ensure safe and proper transfer of the
1529   message. Transfer-Encoding is a property of the message, not of the
1530   entity, and thus MAY be added or removed by any application along the
1531   request/response chain. (However, <xref target="transfer.codings"/> places restrictions on
1532   when certain transfer-codings may be used.)
1533</t>
1534<t>
1535   The rules for when a message-body is allowed in a message differ for
1536   requests and responses.
1537</t>
1538<t>
1539   The presence of a message-body in a request is signaled by the
1540   inclusion of a Content-Length or Transfer-Encoding header field in
1541   the request's message-headers. A message-body MUST NOT be included in
1542   a request if the specification of the request method (Section 3 of <xref target="Part2"/>)
1543   explicitly disallows an entity-body in requests.
1544   When a request message contains both a message-body of non-zero
1545   length and a method that does not define any semantics for that
1546   request message-body, then an origin server SHOULD either ignore
1547   the message-body or respond with an appropriate error message
1548   (e.g., 413).  A proxy or gateway, when presented the same request,
1549   SHOULD either forward the request inbound with the message-body or
1550   ignore the message-body when determining a response.
1551</t>
1552<t>
1553   For response messages, whether or not a message-body is included with
1554   a message is dependent on both the request method and the response
1555   status code (<xref target="status.code.and.reason.phrase"/>). All responses to the HEAD request method
1556   MUST NOT include a message-body, even though the presence of entity-header
1557   fields might lead one to believe they do. All 1xx
1558   (informational), 204 (No Content), and 304 (Not Modified) responses
1559   MUST NOT include a message-body. All other responses do include a
1560   message-body, although it MAY be of zero length.
1561</t>
1562</section>
1563
1564<section title="Message Length" anchor="message.length">
1565<t>
1566   The transfer-length of a message is the length of the message-body as
1567   it appears in the message; that is, after any transfer-codings have
1568   been applied. When a message-body is included with a message, the
1569   transfer-length of that body is determined by one of the following
1570   (in order of precedence):
1571</t>
1572<t>
1573  <list style="numbers">
1574    <t>
1575     Any response message which "MUST NOT" include a message-body (such
1576     as the 1xx, 204, and 304 responses and any response to a HEAD
1577     request) is always terminated by the first empty line after the
1578     header fields, regardless of the entity-header fields present in
1579     the message.
1580    </t>
1581    <t>
1582     If a Transfer-Encoding header field (<xref target="header.transfer-encoding"/>)
1583     is present, then the transfer-length is
1584     defined by use of the "chunked" transfer-coding (<xref target="transfer.codings"/>),
1585     unless the message is terminated by closing the connection.
1586    </t>
1587    <t>
1588     If a Content-Length header field (<xref target="header.content-length"/>) is present, its
1589     decimal value in OCTETs represents both the entity-length and the
1590     transfer-length. The Content-Length header field MUST NOT be sent
1591     if these two lengths are different (i.e., if a Transfer-Encoding
1592     header field is present). If a message is received with both a
1593     Transfer-Encoding header field and a Content-Length header field,
1594     the latter MUST be ignored.
1595    </t>
1596    <t>
1597     If the message uses the media type "multipart/byteranges", and the
1598     transfer-length is not otherwise specified, then this self-delimiting
1599     media type defines the transfer-length. This media type
1600     MUST NOT be used unless the sender knows that the recipient can parse
1601     it; the presence in a request of a Range header with multiple byte-range
1602     specifiers from a 1.1 client implies that the client can parse
1603     multipart/byteranges responses.
1604    <list style="empty"><t>
1605       A range header might be forwarded by a 1.0 proxy that does not
1606       understand multipart/byteranges; in this case the server MUST
1607       delimit the message using methods defined in items 1, 3 or 5 of
1608       this section.
1609    </t></list>
1610    </t>
1611    <t>
1612     By the server closing the connection. (Closing the connection
1613     cannot be used to indicate the end of a request body, since that
1614     would leave no possibility for the server to send back a response.)
1615    </t>
1616  </list>
1617</t>
1618<t>
1619   For compatibility with HTTP/1.0 applications, HTTP/1.1 requests
1620   containing a message-body MUST include a valid Content-Length header
1621   field unless the server is known to be HTTP/1.1 compliant. If a
1622   request contains a message-body and a Content-Length is not given,
1623   the server SHOULD respond with 400 (Bad Request) if it cannot
1624   determine the length of the message, or with 411 (Length Required) if
1625   it wishes to insist on receiving a valid Content-Length.
1626</t>
1627<t>
1628   All HTTP/1.1 applications that receive entities MUST accept the
1629   "chunked" transfer-coding (<xref target="transfer.codings"/>), thus allowing this mechanism
1630   to be used for messages when the message length cannot be determined
1631   in advance.
1632</t>
1633<t>
1634   Messages MUST NOT include both a Content-Length header field and a
1635   transfer-coding. If the message does include a
1636   transfer-coding, the Content-Length MUST be ignored.
1637</t>
1638<t>
1639   When a Content-Length is given in a message where a message-body is
1640   allowed, its field value MUST exactly match the number of OCTETs in
1641   the message-body. HTTP/1.1 user agents MUST notify the user when an
1642   invalid length is received and detected.
1643</t>
1644</section>
1645
1646<section title="General Header Fields" anchor="general.header.fields">
1647 
1648<t>
1649   There are a few header fields which have general applicability for
1650   both request and response messages, but which do not apply to the
1651   entity being transferred. These header fields apply only to the
1652   message being transmitted.
1653</t>
1654<figure><iref primary="true" item="Grammar" subitem="general-header"/><artwork type="abnf2616"><![CDATA[
1655  general-header = Cache-Control            ; [Part6], Section 16.2
1656                 | Connection               ; Section 8.1
1657                 | Date                     ; Section 8.3
1658                 | Pragma                   ; [Part6], Section 16.4
1659                 | Trailer                  ; Section 8.6
1660                 | Transfer-Encoding        ; Section 8.7
1661                 | Upgrade                  ; Section 8.8
1662                 | Via                      ; Section 8.9
1663                 | Warning                  ; [Part6], Section 16.6
1664]]></artwork></figure>
1665<t>
1666   General-header field names can be extended reliably only in
1667   combination with a change in the protocol version. However, new or
1668   experimental header fields may be given the semantics of general
1669   header fields if all parties in the communication recognize them to
1670   be general-header fields. Unrecognized header fields are treated as
1671   entity-header fields.
1672</t>
1673</section>
1674</section>
1675
1676<section title="Request" anchor="request">
1677 
1678<t>
1679   A request message from a client to a server includes, within the
1680   first line of that message, the method to be applied to the resource,
1681   the identifier of the resource, and the protocol version in use.
1682</t>
1683<!--                 Host                      ; should be moved here eventually -->
1684<figure><iref primary="true" item="Grammar" subitem="Request"/><artwork type="abnf2616"><![CDATA[
1685  Request       = Request-Line              ; Section 5.1
1686                  *(( general-header        ; Section 4.5
1687                   | request-header         ; [Part2], Section 4
1688                   | entity-header ) CRLF)  ; [Part3], Section 4.1
1689                  CRLF
1690                  [ message-body ]          ; Section 4.3
1691]]></artwork></figure>
1692
1693<section title="Request-Line" anchor="request-line">
1694 
1695<t>
1696   The Request-Line begins with a method token, followed by the
1697   Request-URI and the protocol version, and ending with CRLF. The
1698   elements are separated by SP characters. No CR or LF is allowed
1699   except in the final CRLF sequence.
1700</t>
1701<figure><iref primary="true" item="Grammar" subitem="Request-Line"/><artwork type="abnf2616"><![CDATA[
1702  Request-Line   = Method SP Request-URI SP HTTP-Version CRLF
1703]]></artwork></figure>
1704
1705<section title="Method" anchor="method">
1706 
1707<t>
1708   The Method  token indicates the method to be performed on the
1709   resource identified by the Request-URI. The method is case-sensitive.
1710</t>
1711<figure><iref primary="true" item="Grammar" subitem="Method"/><iref primary="true" item="Grammar" subitem="extension-method"/><artwork type="abnf2616"><![CDATA[
1712  Method         = token
1713]]></artwork></figure>
1714</section>
1715
1716<section title="Request-URI" anchor="request-uri">
1717 
1718<t>
1719   The Request-URI is a Uniform Resource Identifier (<xref target="uri"/>) and
1720   identifies the resource upon which to apply the request.
1721</t>
1722<figure><iref primary="true" item="Grammar" subitem="Request-URI"/><artwork type="abnf2616"><![CDATA[
1723  Request-URI    = "*"
1724                 | absoluteURI
1725                 | ( path-absolute [ "?" query ] )
1726                 | authority
1727]]></artwork></figure>
1728<t>
1729   The four options for Request-URI are dependent on the nature of the
1730   request. The asterisk "*" means that the request does not apply to a
1731   particular resource, but to the server itself, and is only allowed
1732   when the method used does not necessarily apply to a resource. One
1733   example would be
1734</t>
1735<figure><artwork type="example"><![CDATA[
1736    OPTIONS * HTTP/1.1
1737]]></artwork></figure>
1738<t>
1739   The absoluteURI form is REQUIRED when the request is being made to a
1740   proxy. The proxy is requested to forward the request or service it
1741   from a valid cache, and return the response. Note that the proxy MAY
1742   forward the request on to another proxy or directly to the server
1743   specified by the absoluteURI. In order to avoid request loops, a
1744   proxy MUST be able to recognize all of its server names, including
1745   any aliases, local variations, and the numeric IP address. An example
1746   Request-Line would be:
1747</t>
1748<figure><artwork type="example"><![CDATA[
1749    GET http://www.example.org/pub/WWW/TheProject.html HTTP/1.1
1750]]></artwork></figure>
1751<t>
1752   To allow for transition to absoluteURIs in all requests in future
1753   versions of HTTP, all HTTP/1.1 servers MUST accept the absoluteURI
1754   form in requests, even though HTTP/1.1 clients will only generate
1755   them in requests to proxies.
1756</t>
1757<t>
1758   The authority form is only used by the CONNECT method (Section 8.9 of <xref target="Part2"/>).
1759</t>
1760<t>
1761   The most common form of Request-URI is that used to identify a
1762   resource on an origin server or gateway. In this case the absolute
1763   path of the URI MUST be transmitted (see <xref target="general.syntax"/>, path-absolute) as
1764   the Request-URI, and the network location of the URI (authority) MUST
1765   be transmitted in a Host header field. For example, a client wishing
1766   to retrieve the resource above directly from the origin server would
1767   create a TCP connection to port 80 of the host "www.example.org" and send
1768   the lines:
1769</t>
1770<figure><artwork type="example"><![CDATA[
1771    GET /pub/WWW/TheProject.html HTTP/1.1
1772    Host: www.example.org
1773]]></artwork></figure>
1774<t>
1775   followed by the remainder of the Request. Note that the absolute path
1776   cannot be empty; if none is present in the original URI, it MUST be
1777   given as "/" (the server root).
1778</t>
1779<t>
1780   The Request-URI is transmitted in the format specified in
1781   <xref target="general.syntax"/>. If the Request-URI is encoded using the "% HEX HEX" encoding
1782   <xref target="RFC2396"/>, the origin server MUST decode the Request-URI in order to
1783   properly interpret the request. Servers SHOULD respond to invalid
1784   Request-URIs with an appropriate status code.
1785</t>
1786<t>
1787   A transparent proxy MUST NOT rewrite the "path-absolute" part of the
1788   received Request-URI when forwarding it to the next inbound server,
1789   except as noted above to replace a null path-absolute with "/".
1790</t>
1791<t>
1792  <list><t>
1793      Note: The "no rewrite" rule prevents the proxy from changing the
1794      meaning of the request when the origin server is improperly using
1795      a non-reserved URI character for a reserved purpose.  Implementors
1796      should be aware that some pre-HTTP/1.1 proxies have been known to
1797      rewrite the Request-URI.
1798  </t></list>
1799</t>
1800</section>
1801</section>
1802
1803<section title="The Resource Identified by a Request" anchor="the.resource.identified.by.a.request">
1804<t>
1805   The exact resource identified by an Internet request is determined by
1806   examining both the Request-URI and the Host header field.
1807</t>
1808<t>
1809   An origin server that does not allow resources to differ by the
1810   requested host MAY ignore the Host header field value when
1811   determining the resource identified by an HTTP/1.1 request. (But see
1812   <xref target="changes.to.simplify.multi-homed.web.servers.and.conserve.ip.addresses"/>
1813   for other requirements on Host support in HTTP/1.1.)
1814</t>
1815<t>
1816   An origin server that does differentiate resources based on the host
1817   requested (sometimes referred to as virtual hosts or vanity host
1818   names) MUST use the following rules for determining the requested
1819   resource on an HTTP/1.1 request:
1820  <list style="numbers">
1821    <t>If Request-URI is an absoluteURI, the host is part of the
1822     Request-URI. Any Host header field value in the request MUST be
1823     ignored.</t>
1824    <t>If the Request-URI is not an absoluteURI, and the request includes
1825     a Host header field, the host is determined by the Host header
1826     field value.</t>
1827    <t>If the host as determined by rule 1 or 2 is not a valid host on
1828     the server, the response MUST be a 400 (Bad Request) error message.</t>
1829  </list>
1830</t>
1831<t>
1832   Recipients of an HTTP/1.0 request that lacks a Host header field MAY
1833   attempt to use heuristics (e.g., examination of the URI path for
1834   something unique to a particular host) in order to determine what
1835   exact resource is being requested.
1836</t>
1837</section>
1838
1839</section>
1840
1841
1842<section title="Response" anchor="response">
1843 
1844<t>
1845   After receiving and interpreting a request message, a server responds
1846   with an HTTP response message.
1847</t>
1848<figure><iref primary="true" item="Grammar" subitem="Response"/><artwork type="abnf2616"><![CDATA[
1849  Response      = Status-Line               ; Section 6.1
1850                  *(( general-header        ; Section 4.5
1851                   | response-header        ; [Part2], Section 6
1852                   | entity-header ) CRLF)  ; [Part3], Section 4.1
1853                  CRLF
1854                  [ message-body ]          ; Section 4.3
1855]]></artwork></figure>
1856
1857<section title="Status-Line" anchor="status-line">
1858 
1859<t>
1860   The first line of a Response message is the Status-Line, consisting
1861   of the protocol version followed by a numeric status code and its
1862   associated textual phrase, with each element separated by SP
1863   characters. No CR or LF is allowed except in the final CRLF sequence.
1864</t>
1865<figure><iref primary="true" item="Grammar" subitem="Status-Line"/><artwork type="abnf2616"><![CDATA[
1866  Status-Line = HTTP-Version SP Status-Code SP Reason-Phrase CRLF
1867]]></artwork></figure>
1868
1869<section title="Status Code and Reason Phrase" anchor="status.code.and.reason.phrase">
1870 
1871 
1872<t>
1873   The Status-Code element is a 3-digit integer result code of the
1874   attempt to understand and satisfy the request. These codes are fully
1875   defined in Section 9 of <xref target="Part2"/>.  The Reason Phrase exists for the sole
1876   purpose of providing a textual description associated with the numeric
1877   status code, out of deference to earlier Internet application protocols
1878   that were more frequently used with interactive text clients.
1879   A client SHOULD ignore the content of the Reason Phrase.
1880</t>
1881<t>
1882   The first digit of the Status-Code defines the class of response. The
1883   last two digits do not have any categorization role. There are 5
1884   values for the first digit:
1885  <list style="symbols">
1886    <t>
1887      1xx: Informational - Request received, continuing process
1888    </t>
1889    <t>
1890      2xx: Success - The action was successfully received,
1891        understood, and accepted
1892    </t>
1893    <t>
1894      3xx: Redirection - Further action must be taken in order to
1895        complete the request
1896    </t>
1897    <t>
1898      4xx: Client Error - The request contains bad syntax or cannot
1899        be fulfilled
1900    </t>
1901    <t>
1902      5xx: Server Error - The server failed to fulfill an apparently
1903        valid request
1904    </t>
1905  </list>
1906</t>
1907<figure><iref primary="true" item="Grammar" subitem="Status-Code"/><iref primary="true" item="Grammar" subitem="extension-code"/><iref primary="true" item="Grammar" subitem="Reason-Phrase"/><artwork type="abnf2616"><![CDATA[
1908  Status-Code    = 3DIGIT
1909  Reason-Phrase  = *<TEXT, excluding CR, LF>
1910]]></artwork></figure>
1911</section>
1912</section>
1913
1914</section>
1915
1916
1917<section title="Connections" anchor="connections">
1918
1919<section title="Persistent Connections" anchor="persistent.connections">
1920
1921<section title="Purpose" anchor="persistent.purpose">
1922<t>
1923   Prior to persistent connections, a separate TCP connection was
1924   established to fetch each URL, increasing the load on HTTP servers
1925   and causing congestion on the Internet. The use of inline images and
1926   other associated data often require a client to make multiple
1927   requests of the same server in a short amount of time. Analysis of
1928   these performance problems and results from a prototype
1929   implementation are available <xref target="Pad1995"/> <xref target="Spe"/>. Implementation experience and
1930   measurements of actual HTTP/1.1 (RFC 2068) implementations show good
1931   results <xref target="Nie1997"/>. Alternatives have also been explored, for example,
1932   T/TCP <xref target="Tou1998"/>.
1933</t>
1934<t>
1935   Persistent HTTP connections have a number of advantages:
1936  <list style="symbols">
1937      <t>
1938        By opening and closing fewer TCP connections, CPU time is saved
1939        in routers and hosts (clients, servers, proxies, gateways,
1940        tunnels, or caches), and memory used for TCP protocol control
1941        blocks can be saved in hosts.
1942      </t>
1943      <t>
1944        HTTP requests and responses can be pipelined on a connection.
1945        Pipelining allows a client to make multiple requests without
1946        waiting for each response, allowing a single TCP connection to
1947        be used much more efficiently, with much lower elapsed time.
1948      </t>
1949      <t>
1950        Network congestion is reduced by reducing the number of packets
1951        caused by TCP opens, and by allowing TCP sufficient time to
1952        determine the congestion state of the network.
1953      </t>
1954      <t>
1955        Latency on subsequent requests is reduced since there is no time
1956        spent in TCP's connection opening handshake.
1957      </t>
1958      <t>
1959        HTTP can evolve more gracefully, since errors can be reported
1960        without the penalty of closing the TCP connection. Clients using
1961        future versions of HTTP might optimistically try a new feature,
1962        but if communicating with an older server, retry with old
1963        semantics after an error is reported.
1964      </t>
1965    </list>
1966</t>
1967<t>
1968   HTTP implementations SHOULD implement persistent connections.
1969</t>
1970</section>
1971
1972<section title="Overall Operation" anchor="persistent.overall">
1973<t>
1974   A significant difference between HTTP/1.1 and earlier versions of
1975   HTTP is that persistent connections are the default behavior of any
1976   HTTP connection. That is, unless otherwise indicated, the client
1977   SHOULD assume that the server will maintain a persistent connection,
1978   even after error responses from the server.
1979</t>
1980<t>
1981   Persistent connections provide a mechanism by which a client and a
1982   server can signal the close of a TCP connection. This signaling takes
1983   place using the Connection header field (<xref target="header.connection"/>). Once a close
1984   has been signaled, the client MUST NOT send any more requests on that
1985   connection.
1986</t>
1987
1988<section title="Negotiation" anchor="persistent.negotiation">
1989<t>
1990   An HTTP/1.1 server MAY assume that a HTTP/1.1 client intends to
1991   maintain a persistent connection unless a Connection header including
1992   the connection-token "close" was sent in the request. If the server
1993   chooses to close the connection immediately after sending the
1994   response, it SHOULD send a Connection header including the
1995   connection-token close.
1996</t>
1997<t>
1998   An HTTP/1.1 client MAY expect a connection to remain open, but would
1999   decide to keep it open based on whether the response from a server
2000   contains a Connection header with the connection-token close. In case
2001   the client does not want to maintain a connection for more than that
2002   request, it SHOULD send a Connection header including the
2003   connection-token close.
2004</t>
2005<t>
2006   If either the client or the server sends the close token in the
2007   Connection header, that request becomes the last one for the
2008   connection.
2009</t>
2010<t>
2011   Clients and servers SHOULD NOT  assume that a persistent connection is
2012   maintained for HTTP versions less than 1.1 unless it is explicitly
2013   signaled. See <xref target="compatibility.with.http.1.0.persistent.connections"/> for more information on backward
2014   compatibility with HTTP/1.0 clients.
2015</t>
2016<t>
2017   In order to remain persistent, all messages on the connection MUST
2018   have a self-defined message length (i.e., one not defined by closure
2019   of the connection), as described in <xref target="message.length"/>.
2020</t>
2021</section>
2022
2023<section title="Pipelining" anchor="pipelining">
2024<t>
2025   A client that supports persistent connections MAY "pipeline" its
2026   requests (i.e., send multiple requests without waiting for each
2027   response). A server MUST send its responses to those requests in the
2028   same order that the requests were received.
2029</t>
2030<t>
2031   Clients which assume persistent connections and pipeline immediately
2032   after connection establishment SHOULD be prepared to retry their
2033   connection if the first pipelined attempt fails. If a client does
2034   such a retry, it MUST NOT pipeline before it knows the connection is
2035   persistent. Clients MUST also be prepared to resend their requests if
2036   the server closes the connection before sending all of the
2037   corresponding responses.
2038</t>
2039<t>
2040   Clients SHOULD NOT  pipeline requests using non-idempotent methods or
2041   non-idempotent sequences of methods (see Section 8.1.2 of <xref target="Part2"/>). Otherwise, a
2042   premature termination of the transport connection could lead to
2043   indeterminate results. A client wishing to send a non-idempotent
2044   request SHOULD wait to send that request until it has received the
2045   response status for the previous request.
2046</t>
2047</section>
2048</section>
2049
2050<section title="Proxy Servers" anchor="persistent.proxy">
2051<t>
2052   It is especially important that proxies correctly implement the
2053   properties of the Connection header field as specified in <xref target="header.connection"/>.
2054</t>
2055<t>
2056   The proxy server MUST signal persistent connections separately with
2057   its clients and the origin servers (or other proxy servers) that it
2058   connects to. Each persistent connection applies to only one transport
2059   link.
2060</t>
2061<t>
2062   A proxy server MUST NOT establish a HTTP/1.1 persistent connection
2063   with an HTTP/1.0 client (but see <xref target="RFC2068"/> for information and
2064   discussion of the problems with the Keep-Alive header implemented by
2065   many HTTP/1.0 clients).
2066</t>
2067</section>
2068
2069<section title="Practical Considerations" anchor="persistent.practical">
2070<t>
2071   Servers will usually have some time-out value beyond which they will
2072   no longer maintain an inactive connection. Proxy servers might make
2073   this a higher value since it is likely that the client will be making
2074   more connections through the same server. The use of persistent
2075   connections places no requirements on the length (or existence) of
2076   this time-out for either the client or the server.
2077</t>
2078<t>
2079   When a client or server wishes to time-out it SHOULD issue a graceful
2080   close on the transport connection. Clients and servers SHOULD both
2081   constantly watch for the other side of the transport close, and
2082   respond to it as appropriate. If a client or server does not detect
2083   the other side's close promptly it could cause unnecessary resource
2084   drain on the network.
2085</t>
2086<t>
2087   A client, server, or proxy MAY close the transport connection at any
2088   time. For example, a client might have started to send a new request
2089   at the same time that the server has decided to close the "idle"
2090   connection. From the server's point of view, the connection is being
2091   closed while it was idle, but from the client's point of view, a
2092   request is in progress.
2093</t>
2094<t>
2095   This means that clients, servers, and proxies MUST be able to recover
2096   from asynchronous close events. Client software SHOULD reopen the
2097   transport connection and retransmit the aborted sequence of requests
2098   without user interaction so long as the request sequence is
2099   idempotent (see Section 8.1.2 of <xref target="Part2"/>). Non-idempotent methods or sequences
2100   MUST NOT be automatically retried, although user agents MAY offer a
2101   human operator the choice of retrying the request(s). Confirmation by
2102   user-agent software with semantic understanding of the application
2103   MAY substitute for user confirmation. The automatic retry SHOULD NOT
2104   be repeated if the second sequence of requests fails.
2105</t>
2106<t>
2107   Servers SHOULD always respond to at least one request per connection,
2108   if at all possible. Servers SHOULD NOT  close a connection in the
2109   middle of transmitting a response, unless a network or client failure
2110   is suspected.
2111</t>
2112<t>
2113   Clients that use persistent connections SHOULD limit the number of
2114   simultaneous connections that they maintain to a given server. A
2115   single-user client SHOULD NOT maintain more than 2 connections with
2116   any server or proxy. A proxy SHOULD use up to 2*N connections to
2117   another server or proxy, where N is the number of simultaneously
2118   active users. These guidelines are intended to improve HTTP response
2119   times and avoid congestion.
2120</t>
2121</section>
2122</section>
2123
2124<section title="Message Transmission Requirements" anchor="message.transmission.requirements">
2125
2126<section title="Persistent Connections and Flow Control" anchor="persistent.flow">
2127<t>
2128   HTTP/1.1 servers SHOULD maintain persistent connections and use TCP's
2129   flow control mechanisms to resolve temporary overloads, rather than
2130   terminating connections with the expectation that clients will retry.
2131   The latter technique can exacerbate network congestion.
2132</t>
2133</section>
2134
2135<section title="Monitoring Connections for Error Status Messages" anchor="persistent.monitor">
2136<t>
2137   An HTTP/1.1 (or later) client sending a message-body SHOULD monitor
2138   the network connection for an error status while it is transmitting
2139   the request. If the client sees an error status, it SHOULD
2140   immediately cease transmitting the body. If the body is being sent
2141   using a "chunked" encoding (<xref target="transfer.codings"/>), a zero length chunk and
2142   empty trailer MAY be used to prematurely mark the end of the message.
2143   If the body was preceded by a Content-Length header, the client MUST
2144   close the connection.
2145</t>
2146</section>
2147
2148<section title="Use of the 100 (Continue) Status" anchor="use.of.the.100.status">
2149<t>
2150   The purpose of the 100 (Continue) status (see Section 9.1.1 of <xref target="Part2"/>) is to
2151   allow a client that is sending a request message with a request body
2152   to determine if the origin server is willing to accept the request
2153   (based on the request headers) before the client sends the request
2154   body. In some cases, it might either be inappropriate or highly
2155   inefficient for the client to send the body if the server will reject
2156   the message without looking at the body.
2157</t>
2158<t>
2159   Requirements for HTTP/1.1 clients:
2160  <list style="symbols">
2161    <t>
2162        If a client will wait for a 100 (Continue) response before
2163        sending the request body, it MUST send an Expect request-header
2164        field (Section 10.2 of <xref target="Part2"/>) with the "100-continue" expectation.
2165    </t>
2166    <t>
2167        A client MUST NOT send an Expect request-header field (Section 10.2 of <xref target="Part2"/>)
2168        with the "100-continue" expectation if it does not intend
2169        to send a request body.
2170    </t>
2171  </list>
2172</t>
2173<t>
2174   Because of the presence of older implementations, the protocol allows
2175   ambiguous situations in which a client may send "Expect: 100-continue"
2176   without receiving either a 417 (Expectation Failed) status
2177   or a 100 (Continue) status. Therefore, when a client sends this
2178   header field to an origin server (possibly via a proxy) from which it
2179   has never seen a 100 (Continue) status, the client SHOULD NOT  wait
2180   for an indefinite period before sending the request body.
2181</t>
2182<t>
2183   Requirements for HTTP/1.1 origin servers:
2184  <list style="symbols">
2185    <t> Upon receiving a request which includes an Expect request-header
2186        field with the "100-continue" expectation, an origin server MUST
2187        either respond with 100 (Continue) status and continue to read
2188        from the input stream, or respond with a final status code. The
2189        origin server MUST NOT wait for the request body before sending
2190        the 100 (Continue) response. If it responds with a final status
2191        code, it MAY close the transport connection or it MAY continue
2192        to read and discard the rest of the request.  It MUST NOT
2193        perform the requested method if it returns a final status code.
2194    </t>
2195    <t> An origin server SHOULD NOT  send a 100 (Continue) response if
2196        the request message does not include an Expect request-header
2197        field with the "100-continue" expectation, and MUST NOT send a
2198        100 (Continue) response if such a request comes from an HTTP/1.0
2199        (or earlier) client. There is an exception to this rule: for
2200        compatibility with <xref target="RFC2068"/>, a server MAY send a 100 (Continue)
2201        status in response to an HTTP/1.1 PUT or POST request that does
2202        not include an Expect request-header field with the "100-continue"
2203        expectation. This exception, the purpose of which is
2204        to minimize any client processing delays associated with an
2205        undeclared wait for 100 (Continue) status, applies only to
2206        HTTP/1.1 requests, and not to requests with any other HTTP-version
2207        value.
2208    </t>
2209    <t> An origin server MAY omit a 100 (Continue) response if it has
2210        already received some or all of the request body for the
2211        corresponding request.
2212    </t>
2213    <t> An origin server that sends a 100 (Continue) response MUST
2214    ultimately send a final status code, once the request body is
2215        received and processed, unless it terminates the transport
2216        connection prematurely.
2217    </t>
2218    <t> If an origin server receives a request that does not include an
2219        Expect request-header field with the "100-continue" expectation,
2220        the request includes a request body, and the server responds
2221        with a final status code before reading the entire request body
2222        from the transport connection, then the server SHOULD NOT  close
2223        the transport connection until it has read the entire request,
2224        or until the client closes the connection. Otherwise, the client
2225        might not reliably receive the response message. However, this
2226        requirement is not be construed as preventing a server from
2227        defending itself against denial-of-service attacks, or from
2228        badly broken client implementations.
2229      </t>
2230    </list>
2231</t>
2232<t>
2233   Requirements for HTTP/1.1 proxies:
2234  <list style="symbols">
2235    <t> If a proxy receives a request that includes an Expect request-header
2236        field with the "100-continue" expectation, and the proxy
2237        either knows that the next-hop server complies with HTTP/1.1 or
2238        higher, or does not know the HTTP version of the next-hop
2239        server, it MUST forward the request, including the Expect header
2240        field.
2241    </t>
2242    <t> If the proxy knows that the version of the next-hop server is
2243        HTTP/1.0 or lower, it MUST NOT forward the request, and it MUST
2244        respond with a 417 (Expectation Failed) status.
2245    </t>
2246    <t> Proxies SHOULD maintain a cache recording the HTTP version
2247        numbers received from recently-referenced next-hop servers.
2248    </t>
2249    <t> A proxy MUST NOT forward a 100 (Continue) response if the
2250        request message was received from an HTTP/1.0 (or earlier)
2251        client and did not include an Expect request-header field with
2252        the "100-continue" expectation. This requirement overrides the
2253        general rule for forwarding of 1xx responses (see Section 9.1 of <xref target="Part2"/>).
2254    </t>
2255  </list>
2256</t>
2257</section>
2258
2259<section title="Client Behavior if Server Prematurely Closes Connection" anchor="connection.premature">
2260<t>
2261   If an HTTP/1.1 client sends a request which includes a request body,
2262   but which does not include an Expect request-header field with the
2263   "100-continue" expectation, and if the client is not directly
2264   connected to an HTTP/1.1 origin server, and if the client sees the
2265   connection close before receiving any status from the server, the
2266   client SHOULD retry the request.  If the client does retry this
2267   request, it MAY use the following "binary exponential backoff"
2268   algorithm to be assured of obtaining a reliable response:
2269  <list style="numbers">
2270    <t>
2271      Initiate a new connection to the server
2272    </t>
2273    <t>
2274      Transmit the request-headers
2275    </t>
2276    <t>
2277      Initialize a variable R to the estimated round-trip time to the
2278         server (e.g., based on the time it took to establish the
2279         connection), or to a constant value of 5 seconds if the round-trip
2280         time is not available.
2281    </t>
2282    <t>
2283       Compute T = R * (2**N), where N is the number of previous
2284         retries of this request.
2285    </t>
2286    <t>
2287       Wait either for an error response from the server, or for T
2288         seconds (whichever comes first)
2289    </t>
2290    <t>
2291       If no error response is received, after T seconds transmit the
2292         body of the request.
2293    </t>
2294    <t>
2295       If client sees that the connection is closed prematurely,
2296         repeat from step 1 until the request is accepted, an error
2297         response is received, or the user becomes impatient and
2298         terminates the retry process.
2299    </t>
2300  </list>
2301</t>
2302<t>
2303   If at any point an error status is received, the client
2304  <list style="symbols">
2305      <t>SHOULD NOT  continue and</t>
2306
2307      <t>SHOULD close the connection if it has not completed sending the
2308        request message.</t>
2309    </list>
2310</t>
2311</section>
2312</section>
2313</section>
2314
2315
2316<section title="Header Field Definitions" anchor="header.fields">
2317<t>
2318   This section defines the syntax and semantics of HTTP/1.1 header fields
2319   related to message framing and transport protocols.
2320</t>
2321<t>
2322   For entity-header fields, both sender and recipient refer to either the
2323   client or the server, depending on who sends and who receives the entity.
2324</t>
2325
2326<section title="Connection" anchor="header.connection">
2327  <iref primary="true" item="Connection header"/>
2328  <iref primary="true" item="Headers" subitem="Connection"/>
2329 
2330 
2331<t>
2332   The Connection general-header field allows the sender to specify
2333   options that are desired for that particular connection and MUST NOT
2334   be communicated by proxies over further connections.
2335</t>
2336<t>
2337   The Connection header has the following grammar:
2338</t>
2339<figure><iref primary="true" item="Grammar" subitem="Connection"/><iref primary="true" item="Grammar" subitem="connection-token"/><artwork type="abnf2616"><![CDATA[
2340  Connection = "Connection" ":" 1#(connection-token)
2341  connection-token  = token
2342]]></artwork></figure>
2343<t>
2344   HTTP/1.1 proxies MUST parse the Connection header field before a
2345   message is forwarded and, for each connection-token in this field,
2346   remove any header field(s) from the message with the same name as the
2347   connection-token. Connection options are signaled by the presence of
2348   a connection-token in the Connection header field, not by any
2349   corresponding additional header field(s), since the additional header
2350   field may not be sent if there are no parameters associated with that
2351   connection option.
2352</t>
2353<t>
2354   Message headers listed in the Connection header MUST NOT include
2355   end-to-end headers, such as Cache-Control.
2356</t>
2357<t>
2358   HTTP/1.1 defines the "close" connection option for the sender to
2359   signal that the connection will be closed after completion of the
2360   response. For example,
2361</t>
2362<figure><artwork type="example"><![CDATA[
2363    Connection: close
2364]]></artwork></figure>
2365<t>
2366   in either the request or the response header fields indicates that
2367   the connection SHOULD NOT  be considered `persistent' (<xref target="persistent.connections"/>)
2368   after the current request/response is complete.
2369</t>
2370<t>
2371   An HTTP/1.1 client that does not support persistent connections MUST
2372   include the "close" connection option in every request message.
2373</t>
2374<t>
2375   An HTTP/1.1 server that does not support persistent connections MUST
2376   include the "close" connection option in every response message that
2377   does not have a 1xx (informational) status code.
2378</t>
2379<t>
2380   A system receiving an HTTP/1.0 (or lower-version) message that
2381   includes a Connection header MUST, for each connection-token in this
2382   field, remove and ignore any header field(s) from the message with
2383   the same name as the connection-token. This protects against mistaken
2384   forwarding of such header fields by pre-HTTP/1.1 proxies. See <xref target="compatibility.with.http.1.0.persistent.connections"/>.
2385</t>
2386</section>
2387
2388<section title="Content-Length" anchor="header.content-length">
2389  <iref primary="true" item="Content-Length header"/>
2390  <iref primary="true" item="Headers" subitem="Content-Length"/>
2391 
2392<t>
2393   The Content-Length entity-header field indicates the size of the
2394   entity-body, in decimal number of OCTETs, sent to the recipient or,
2395   in the case of the HEAD method, the size of the entity-body that
2396   would have been sent had the request been a GET.
2397</t>
2398<figure><iref primary="true" item="Grammar" subitem="Content-Length"/><artwork type="abnf2616"><![CDATA[
2399  Content-Length    = "Content-Length" ":" 1*DIGIT
2400]]></artwork></figure>
2401<t>
2402   An example is
2403</t>
2404<figure><artwork type="example"><![CDATA[
2405    Content-Length: 3495
2406]]></artwork></figure>
2407<t>
2408   Applications SHOULD use this field to indicate the transfer-length of
2409   the message-body, unless this is prohibited by the rules in <xref target="message.length"/>.
2410</t>
2411<t>
2412   Any Content-Length greater than or equal to zero is a valid value.
2413   <xref target="message.length"/> describes how to determine the length of a message-body
2414   if a Content-Length is not given.
2415</t>
2416<t>
2417   Note that the meaning of this field is significantly different from
2418   the corresponding definition in MIME, where it is an optional field
2419   used within the "message/external-body" content-type. In HTTP, it
2420   SHOULD be sent whenever the message's length can be determined prior
2421   to being transferred, unless this is prohibited by the rules in
2422   <xref target="message.length"/>.
2423</t>
2424</section>
2425
2426<section title="Date" anchor="header.date">
2427  <iref primary="true" item="Date header"/>
2428  <iref primary="true" item="Headers" subitem="Date"/>
2429 
2430<t>
2431   The Date general-header field represents the date and time at which
2432   the message was originated, having the same semantics as orig-date in
2433   Section 3.6.1 of <xref target="RFC2822"/>. The field value is an HTTP-date, as described in <xref target="full.date"/>;
2434   it MUST be sent in rfc1123-date format.
2435</t>
2436<figure><iref primary="true" item="Grammar" subitem="Date"/><artwork type="abnf2616"><![CDATA[
2437  Date  = "Date" ":" HTTP-date
2438]]></artwork></figure>
2439<t>
2440   An example is
2441</t>
2442<figure><artwork type="example"><![CDATA[
2443    Date: Tue, 15 Nov 1994 08:12:31 GMT
2444]]></artwork></figure>
2445<t>
2446   Origin servers MUST include a Date header field in all responses,
2447   except in these cases:
2448  <list style="numbers">
2449      <t>If the response status code is 100 (Continue) or 101 (Switching
2450         Protocols), the response MAY include a Date header field, at
2451         the server's option.</t>
2452
2453      <t>If the response status code conveys a server error, e.g. 500
2454         (Internal Server Error) or 503 (Service Unavailable), and it is
2455         inconvenient or impossible to generate a valid Date.</t>
2456
2457      <t>If the server does not have a clock that can provide a
2458         reasonable approximation of the current time, its responses
2459         MUST NOT include a Date header field. In this case, the rules
2460         in <xref target="clockless.origin.server.operation"/> MUST be followed.</t>
2461  </list>
2462</t>
2463<t>
2464   A received message that does not have a Date header field MUST be
2465   assigned one by the recipient if the message will be cached by that
2466   recipient or gatewayed via a protocol which requires a Date. An HTTP
2467   implementation without a clock MUST NOT cache responses without
2468   revalidating them on every use. An HTTP cache, especially a shared
2469   cache, SHOULD use a mechanism, such as NTP <xref target="RFC1305"/>, to synchronize its
2470   clock with a reliable external standard.
2471</t>
2472<t>
2473   Clients SHOULD only send a Date header field in messages that include
2474   an entity-body, as in the case of the PUT and POST requests, and even
2475   then it is optional. A client without a clock MUST NOT send a Date
2476   header field in a request.
2477</t>
2478<t>
2479   The HTTP-date sent in a Date header SHOULD NOT  represent a date and
2480   time subsequent to the generation of the message. It SHOULD represent
2481   the best available approximation of the date and time of message
2482   generation, unless the implementation has no means of generating a
2483   reasonably accurate date and time. In theory, the date ought to
2484   represent the moment just before the entity is generated. In
2485   practice, the date can be generated at any time during the message
2486   origination without affecting its semantic value.
2487</t>
2488
2489<section title="Clockless Origin Server Operation" anchor="clockless.origin.server.operation">
2490<t>
2491   Some origin server implementations might not have a clock available.
2492   An origin server without a clock MUST NOT assign Expires or Last-Modified
2493   values to a response, unless these values were associated
2494   with the resource by a system or user with a reliable clock. It MAY
2495   assign an Expires value that is known, at or before server
2496   configuration time, to be in the past (this allows "pre-expiration"
2497   of responses without storing separate Expires values for each
2498   resource).
2499</t>
2500</section>
2501</section>
2502
2503<section title="Host" anchor="header.host">
2504  <iref primary="true" item="Host header"/>
2505  <iref primary="true" item="Headers" subitem="Host"/>
2506 
2507<t>
2508   The Host request-header field specifies the Internet host and port
2509   number of the resource being requested, as obtained from the original
2510   URI given by the user or referring resource (generally an HTTP URL,
2511   as described in <xref target="http.url"/>). The Host field value MUST represent
2512   the naming authority of the origin server or gateway given by the
2513   original URL. This allows the origin server or gateway to
2514   differentiate between internally-ambiguous URLs, such as the root "/"
2515   URL of a server for multiple host names on a single IP address.
2516</t>
2517<figure><iref primary="true" item="Grammar" subitem="Host"/><artwork type="abnf2616"><![CDATA[
2518  Host = "Host" ":" uri-host [ ":" port ] ; Section 3.2.2
2519]]></artwork></figure>
2520<t>
2521   A "host" without any trailing port information implies the default
2522   port for the service requested (e.g., "80" for an HTTP URL). For
2523   example, a request on the origin server for
2524   &lt;http://www.example.org/pub/WWW/&gt; would properly include:
2525</t>
2526<figure><artwork type="example"><![CDATA[
2527    GET /pub/WWW/ HTTP/1.1
2528    Host: www.example.org
2529]]></artwork></figure>
2530<t>
2531   A client MUST include a Host header field in all HTTP/1.1 request
2532   messages. If the requested URI does not include an Internet host
2533   name for the service being requested, then the Host header field MUST
2534   be given with an empty value. An HTTP/1.1 proxy MUST ensure that any
2535   request message it forwards does contain an appropriate Host header
2536   field that identifies the service being requested by the proxy. All
2537   Internet-based HTTP/1.1 servers MUST respond with a 400 (Bad Request)
2538   status code to any HTTP/1.1 request message which lacks a Host header
2539   field.
2540</t>
2541<t>
2542   See Sections <xref target="the.resource.identified.by.a.request" format="counter"/>
2543   and <xref target="changes.to.simplify.multi-homed.web.servers.and.conserve.ip.addresses" format="counter"/>
2544   for other requirements relating to Host.
2545</t>
2546</section>
2547
2548<section title="TE" anchor="header.te">
2549  <iref primary="true" item="TE header"/>
2550  <iref primary="true" item="Headers" subitem="TE"/>
2551 
2552 
2553<t>
2554   The TE request-header field indicates what extension transfer-codings
2555   it is willing to accept in the response and whether or not it is
2556   willing to accept trailer fields in a chunked transfer-coding. Its
2557   value may consist of the keyword "trailers" and/or a comma-separated
2558   list of extension transfer-coding names with optional accept
2559   parameters (as described in <xref target="transfer.codings"/>).
2560</t>
2561<figure><iref primary="true" item="Grammar" subitem="TE"/><iref primary="true" item="Grammar" subitem="t-codings"/><artwork type="abnf2616"><![CDATA[
2562  TE        = "TE" ":" #( t-codings )
2563  t-codings = "trailers" | ( transfer-extension [ accept-params ] )
2564]]></artwork></figure>
2565<t>
2566   The presence of the keyword "trailers" indicates that the client is
2567   willing to accept trailer fields in a chunked transfer-coding, as
2568   defined in <xref target="chunked.transfer.encoding"/>. This keyword is reserved for use with
2569   transfer-coding values even though it does not itself represent a
2570   transfer-coding.
2571</t>
2572<t>
2573   Examples of its use are:
2574</t>
2575<figure><artwork type="example"><![CDATA[
2576    TE: deflate
2577    TE:
2578    TE: trailers, deflate;q=0.5
2579]]></artwork></figure>
2580<t>
2581   The TE header field only applies to the immediate connection.
2582   Therefore, the keyword MUST be supplied within a Connection header
2583   field (<xref target="header.connection"/>) whenever TE is present in an HTTP/1.1 message.
2584</t>
2585<t>
2586   A server tests whether a transfer-coding is acceptable, according to
2587   a TE field, using these rules:
2588  <list style="numbers">
2589    <t>The "chunked" transfer-coding is always acceptable. If the
2590         keyword "trailers" is listed, the client indicates that it is
2591         willing to accept trailer fields in the chunked response on
2592         behalf of itself and any downstream clients. The implication is
2593         that, if given, the client is stating that either all
2594         downstream clients are willing to accept trailer fields in the
2595         forwarded response, or that it will attempt to buffer the
2596         response on behalf of downstream recipients.
2597      <vspace blankLines="1"/>
2598         Note: HTTP/1.1 does not define any means to limit the size of a
2599         chunked response such that a client can be assured of buffering
2600         the entire response.</t>
2601    <t>If the transfer-coding being tested is one of the transfer-codings
2602         listed in the TE field, then it is acceptable unless it
2603         is accompanied by a qvalue of 0. (As defined in Section 3.4 of <xref target="Part3"/>, a
2604         qvalue of 0 means "not acceptable.")</t>
2605    <t>If multiple transfer-codings are acceptable, then the
2606         acceptable transfer-coding with the highest non-zero qvalue is
2607         preferred.  The "chunked" transfer-coding always has a qvalue
2608         of 1.</t>
2609  </list>
2610</t>
2611<t>
2612   If the TE field-value is empty or if no TE field is present, the only
2613   transfer-coding  is "chunked". A message with no transfer-coding is
2614   always acceptable.
2615</t>
2616</section>
2617
2618<section title="Trailer" anchor="header.trailer">
2619  <iref primary="true" item="Trailer header"/>
2620  <iref primary="true" item="Headers" subitem="Trailer"/>
2621 
2622<t>
2623   The Trailer general field value indicates that the given set of
2624   header fields is present in the trailer of a message encoded with
2625   chunked transfer-coding.
2626</t>
2627<figure><iref primary="true" item="Grammar" subitem="Trailer"/><artwork type="abnf2616"><![CDATA[
2628  Trailer  = "Trailer" ":" 1#field-name
2629]]></artwork></figure>
2630<t>
2631   An HTTP/1.1 message SHOULD include a Trailer header field in a
2632   message using chunked transfer-coding with a non-empty trailer. Doing
2633   so allows the recipient to know which header fields to expect in the
2634   trailer.
2635</t>
2636<t>
2637   If no Trailer header field is present, the trailer SHOULD NOT  include
2638   any header fields. See <xref target="chunked.transfer.encoding"/> for restrictions on the use of
2639   trailer fields in a "chunked" transfer-coding.
2640</t>
2641<t>
2642   Message header fields listed in the Trailer header field MUST NOT
2643   include the following header fields:
2644  <list style="symbols">
2645    <t>Transfer-Encoding</t>
2646    <t>Content-Length</t>
2647    <t>Trailer</t>
2648  </list>
2649</t>
2650</section>
2651
2652<section title="Transfer-Encoding" anchor="header.transfer-encoding">
2653  <iref primary="true" item="Transfer-Encoding header"/>
2654  <iref primary="true" item="Headers" subitem="Transfer-Encoding"/>
2655 
2656<t>
2657   The Transfer-Encoding general-header field indicates what (if any)
2658   type of transformation has been applied to the message body in order
2659   to safely transfer it between the sender and the recipient. This
2660   differs from the content-coding in that the transfer-coding is a
2661   property of the message, not of the entity.
2662</t>
2663<figure><iref primary="true" item="Grammar" subitem="Transfer-Encoding"/><artwork type="abnf2616"><![CDATA[
2664  Transfer-Encoding       = "Transfer-Encoding" ":" 1#transfer-coding
2665]]></artwork></figure>
2666<t>
2667   Transfer-codings are defined in <xref target="transfer.codings"/>. An example is:
2668</t>
2669<figure><artwork type="example"><![CDATA[
2670  Transfer-Encoding: chunked
2671]]></artwork></figure>
2672<t>
2673   If multiple encodings have been applied to an entity, the transfer-codings
2674   MUST be listed in the order in which they were applied.
2675   Additional information about the encoding parameters MAY be provided
2676   by other entity-header fields not defined by this specification.
2677</t>
2678<t>
2679   Many older HTTP/1.0 applications do not understand the Transfer-Encoding
2680   header.
2681</t>
2682</section>
2683
2684<section title="Upgrade" anchor="header.upgrade">
2685  <iref primary="true" item="Upgrade header"/>
2686  <iref primary="true" item="Headers" subitem="Upgrade"/>
2687 
2688<t>
2689   The Upgrade general-header allows the client to specify what
2690   additional communication protocols it supports and would like to use
2691   if the server finds it appropriate to switch protocols. The server
2692   MUST use the Upgrade header field within a 101 (Switching Protocols)
2693   response to indicate which protocol(s) are being switched.
2694</t>
2695<figure><iref primary="true" item="Grammar" subitem="Upgrade"/><artwork type="abnf2616"><![CDATA[
2696  Upgrade        = "Upgrade" ":" 1#product
2697]]></artwork></figure>
2698<t>
2699   For example,
2700</t>
2701<figure><artwork type="example"><![CDATA[
2702    Upgrade: HTTP/2.0, SHTTP/1.3, IRC/6.9, RTA/x11
2703]]></artwork></figure>
2704<t>
2705   The Upgrade header field is intended to provide a simple mechanism
2706   for transition from HTTP/1.1 to some other, incompatible protocol. It
2707   does so by allowing the client to advertise its desire to use another
2708   protocol, such as a later version of HTTP with a higher major version
2709   number, even though the current request has been made using HTTP/1.1.
2710   This eases the difficult transition between incompatible protocols by
2711   allowing the client to initiate a request in the more commonly
2712   supported protocol while indicating to the server that it would like
2713   to use a "better" protocol if available (where "better" is determined
2714   by the server, possibly according to the nature of the method and/or
2715   resource being requested).
2716</t>
2717<t>
2718   The Upgrade header field only applies to switching application-layer
2719   protocols upon the existing transport-layer connection. Upgrade
2720   cannot be used to insist on a protocol change; its acceptance and use
2721   by the server is optional. The capabilities and nature of the
2722   application-layer communication after the protocol change is entirely
2723   dependent upon the new protocol chosen, although the first action
2724   after changing the protocol MUST be a response to the initial HTTP
2725   request containing the Upgrade header field.
2726</t>
2727<t>
2728   The Upgrade header field only applies to the immediate connection.
2729   Therefore, the upgrade keyword MUST be supplied within a Connection
2730   header field (<xref target="header.connection"/>) whenever Upgrade is present in an
2731   HTTP/1.1 message.
2732</t>
2733<t>
2734   The Upgrade header field cannot be used to indicate a switch to a
2735   protocol on a different connection. For that purpose, it is more
2736   appropriate to use a 301, 302, 303, or 305 redirection response.
2737</t>
2738<t>
2739   This specification only defines the protocol name "HTTP" for use by
2740   the family of Hypertext Transfer Protocols, as defined by the HTTP
2741   version rules of <xref target="http.version"/> and future updates to this
2742   specification. Any token can be used as a protocol name; however, it
2743   will only be useful if both the client and server associate the name
2744   with the same protocol.
2745</t>
2746</section>
2747
2748<section title="Via" anchor="header.via">
2749  <iref primary="true" item="Via header"/>
2750  <iref primary="true" item="Headers" subitem="Via"/>
2751 
2752 
2753 
2754 
2755 
2756 
2757<t>
2758   The Via general-header field MUST be used by gateways and proxies to
2759   indicate the intermediate protocols and recipients between the user
2760   agent and the server on requests, and between the origin server and
2761   the client on responses. It is analogous to the "Received" field defined in
2762   Section 3.6.7 of <xref target="RFC2822"/> and is intended to be used for tracking message forwards,
2763   avoiding request loops, and identifying the protocol capabilities of
2764   all senders along the request/response chain.
2765</t>
2766<figure><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"/><artwork type="abnf2616"><![CDATA[
2767  Via =  "Via" ":" 1#( received-protocol received-by [ comment ] )
2768  received-protocol = [ protocol-name "/" ] protocol-version
2769  protocol-name     = token
2770  protocol-version  = token
2771  received-by       = ( uri-host [ ":" port ] ) | pseudonym
2772  pseudonym         = token
2773]]></artwork></figure>
2774<t>
2775   The received-protocol indicates the protocol version of the message
2776   received by the server or client along each segment of the
2777   request/response chain. The received-protocol version is appended to
2778   the Via field value when the message is forwarded so that information
2779   about the protocol capabilities of upstream applications remains
2780   visible to all recipients.
2781</t>
2782<t>
2783   The protocol-name is optional if and only if it would be "HTTP". The
2784   received-by field is normally the host and optional port number of a
2785   recipient server or client that subsequently forwarded the message.
2786   However, if the real host is considered to be sensitive information,
2787   it MAY be replaced by a pseudonym. If the port is not given, it MAY
2788   be assumed to be the default port of the received-protocol.
2789</t>
2790<t>
2791   Multiple Via field values represents each proxy or gateway that has
2792   forwarded the message. Each recipient MUST append its information
2793   such that the end result is ordered according to the sequence of
2794   forwarding applications.
2795</t>
2796<t>
2797   Comments MAY be used in the Via header field to identify the software
2798   of the recipient proxy or gateway, analogous to the User-Agent and
2799   Server header fields. However, all comments in the Via field are
2800   optional and MAY be removed by any recipient prior to forwarding the
2801   message.
2802</t>
2803<t>
2804   For example, a request message could be sent from an HTTP/1.0 user
2805   agent to an internal proxy code-named "fred", which uses HTTP/1.1 to
2806   forward the request to a public proxy at p.example.net, which completes
2807   the request by forwarding it to the origin server at www.example.com.
2808   The request received by www.example.com would then have the following
2809   Via header field:
2810</t>
2811<figure><artwork type="example"><![CDATA[
2812    Via: 1.0 fred, 1.1 p.example.net (Apache/1.1)
2813]]></artwork></figure>
2814<t>
2815   Proxies and gateways used as a portal through a network firewall
2816   SHOULD NOT, by default, forward the names and ports of hosts within
2817   the firewall region. This information SHOULD only be propagated if
2818   explicitly enabled. If not enabled, the received-by host of any host
2819   behind the firewall SHOULD be replaced by an appropriate pseudonym
2820   for that host.
2821</t>
2822<t>
2823   For organizations that have strong privacy requirements for hiding
2824   internal structures, a proxy MAY combine an ordered subsequence of
2825   Via header field entries with identical received-protocol values into
2826   a single such entry. For example,
2827</t>
2828<figure><artwork type="example"><![CDATA[
2829    Via: 1.0 ricky, 1.1 ethel, 1.1 fred, 1.0 lucy
2830]]></artwork></figure>
2831<t>
2832        could be collapsed to
2833</t>
2834<figure><artwork type="example"><![CDATA[
2835    Via: 1.0 ricky, 1.1 mertz, 1.0 lucy
2836]]></artwork></figure>
2837<t>
2838   Applications SHOULD NOT  combine multiple entries unless they are all
2839   under the same organizational control and the hosts have already been
2840   replaced by pseudonyms. Applications MUST NOT combine entries which
2841   have different received-protocol values.
2842</t>
2843</section>
2844
2845</section>
2846
2847<section title="IANA Considerations" anchor="IANA.considerations">
2848<section title="Message Header Registration" anchor="message.header.registration">
2849<!--AUTOGENERATED FROM extract-header-defs.xslt, do not edit manually-->
2850<!--(START)-->
2851<t>
2852    The Message Header Registry located at <eref target="http://www.iana.org/assignments/message-headers/message-header-index.html"/> should be updated
2853    with the permanent registrations below (see <xref target="RFC3864"/>):
2854</t>
2855<texttable>
2856   <ttcol>Header Field Name</ttcol>
2857   <ttcol>Protocol</ttcol>
2858   <ttcol>Status</ttcol>
2859   <ttcol>Reference</ttcol>
2860
2861   <c>Connection</c>
2862   <c>http</c>
2863   <c>standard</c>
2864   <c>
2865      <xref target="header.connection"/>
2866   </c>
2867
2868   <c>Content-Length</c>
2869   <c>http</c>
2870   <c>standard</c>
2871   <c>
2872      <xref target="header.content-length"/>
2873   </c>
2874
2875   <c>Date</c>
2876   <c>http</c>
2877   <c>standard</c>
2878   <c>
2879      <xref target="header.date"/>
2880   </c>
2881
2882   <c>Host</c>
2883   <c>http</c>
2884   <c>standard</c>
2885   <c>
2886      <xref target="header.host"/>
2887   </c>
2888
2889   <c>TE</c>
2890   <c>http</c>
2891   <c>standard</c>
2892   <c>
2893      <xref target="header.te"/>
2894   </c>
2895
2896   <c>Trailer</c>
2897   <c>http</c>
2898   <c>standard</c>
2899   <c>
2900      <xref target="header.trailer"/>
2901   </c>
2902
2903   <c>Transfer-Encoding</c>
2904   <c>http</c>
2905   <c>standard</c>
2906   <c>
2907      <xref target="header.transfer-encoding"/>
2908   </c>
2909
2910   <c>Upgrade</c>
2911   <c>http</c>
2912   <c>standard</c>
2913   <c>
2914      <xref target="header.upgrade"/>
2915   </c>
2916
2917   <c>Via</c>
2918   <c>http</c>
2919   <c>standard</c>
2920   <c>
2921      <xref target="header.via"/>
2922   </c>
2923</texttable>
2924<t>
2925    The change controller is: "IETF (iesg@ietf.org) - Internet Engineering Task Force".
2926</t>
2927<!--(END)-->
2928</section>
2929</section>
2930
2931<section title="Security Considerations" anchor="security.considerations">
2932<t>
2933   This section is meant to inform application developers, information
2934   providers, and users of the security limitations in HTTP/1.1 as
2935   described by this document. The discussion does not include
2936   definitive solutions to the problems revealed, though it does make
2937   some suggestions for reducing security risks.
2938</t>
2939
2940<section title="Personal Information" anchor="personal.information">
2941<t>
2942   HTTP clients are often privy to large amounts of personal information
2943   (e.g. the user's name, location, mail address, passwords, encryption
2944   keys, etc.), and SHOULD be very careful to prevent unintentional
2945   leakage of this information.
2946   We very strongly recommend that a convenient interface be provided
2947   for the user to control dissemination of such information, and that
2948   designers and implementors be particularly careful in this area.
2949   History shows that errors in this area often create serious security
2950   and/or privacy problems and generate highly adverse publicity for the
2951   implementor's company.
2952</t>
2953</section>
2954
2955<section title="Abuse of Server Log Information" anchor="abuse.of.server.log.information">
2956<t>
2957   A server is in the position to save personal data about a user's
2958   requests which might identify their reading patterns or subjects of
2959   interest. This information is clearly confidential in nature and its
2960   handling can be constrained by law in certain countries. People using
2961   HTTP to provide data are responsible for ensuring that
2962   such material is not distributed without the permission of any
2963   individuals that are identifiable by the published results.
2964</t>
2965</section>
2966
2967<section title="Attacks Based On File and Path Names" anchor="attack.pathname">
2968<t>
2969   Implementations of HTTP origin servers SHOULD be careful to restrict
2970   the documents returned by HTTP requests to be only those that were
2971   intended by the server administrators. If an HTTP server translates
2972   HTTP URIs directly into file system calls, the server MUST take
2973   special care not to serve files that were not intended to be
2974   delivered to HTTP clients. For example, UNIX, Microsoft Windows, and
2975   other operating systems use ".." as a path component to indicate a
2976   directory level above the current one. On such a system, an HTTP
2977   server MUST disallow any such construct in the Request-URI if it
2978   would otherwise allow access to a resource outside those intended to
2979   be accessible via the HTTP server. Similarly, files intended for
2980   reference only internally to the server (such as access control
2981   files, configuration files, and script code) MUST be protected from
2982   inappropriate retrieval, since they might contain sensitive
2983   information. Experience has shown that minor bugs in such HTTP server
2984   implementations have turned into security risks.
2985</t>
2986</section>
2987
2988<section title="DNS Spoofing" anchor="dns.spoofing">
2989<t>
2990   Clients using HTTP rely heavily on the Domain Name Service, and are
2991   thus generally prone to security attacks based on the deliberate
2992   mis-association of IP addresses and DNS names. Clients need to be
2993   cautious in assuming the continuing validity of an IP number/DNS name
2994   association.
2995</t>
2996<t>
2997   In particular, HTTP clients SHOULD rely on their name resolver for
2998   confirmation of an IP number/DNS name association, rather than
2999   caching the result of previous host name lookups. Many platforms
3000   already can cache host name lookups locally when appropriate, and
3001   they SHOULD be configured to do so. It is proper for these lookups to
3002   be cached, however, only when the TTL (Time To Live) information
3003   reported by the name server makes it likely that the cached
3004   information will remain useful.
3005</t>
3006<t>
3007   If HTTP clients cache the results of host name lookups in order to
3008   achieve a performance improvement, they MUST observe the TTL
3009   information reported by DNS.
3010</t>
3011<t>
3012   If HTTP clients do not observe this rule, they could be spoofed when
3013   a previously-accessed server's IP address changes. As network
3014   renumbering is expected to become increasingly common <xref target="RFC1900"/>, the
3015   possibility of this form of attack will grow. Observing this
3016   requirement thus reduces this potential security vulnerability.
3017</t>
3018<t>
3019   This requirement also improves the load-balancing behavior of clients
3020   for replicated servers using the same DNS name and reduces the
3021   likelihood of a user's experiencing failure in accessing sites which
3022   use that strategy.
3023</t>
3024</section>
3025
3026<section title="Proxies and Caching" anchor="attack.proxies">
3027<t>
3028   By their very nature, HTTP proxies are men-in-the-middle, and
3029   represent an opportunity for man-in-the-middle attacks. Compromise of
3030   the systems on which the proxies run can result in serious security
3031   and privacy problems. Proxies have access to security-related
3032   information, personal information about individual users and
3033   organizations, and proprietary information belonging to users and
3034   content providers. A compromised proxy, or a proxy implemented or
3035   configured without regard to security and privacy considerations,
3036   might be used in the commission of a wide range of potential attacks.
3037</t>
3038<t>
3039   Proxy operators should protect the systems on which proxies run as
3040   they would protect any system that contains or transports sensitive
3041   information. In particular, log information gathered at proxies often
3042   contains highly sensitive personal information, and/or information
3043   about organizations. Log information should be carefully guarded, and
3044   appropriate guidelines for use developed and followed. (<xref target="abuse.of.server.log.information"/>).
3045</t>
3046<t>
3047   Proxy implementors should consider the privacy and security
3048   implications of their design and coding decisions, and of the
3049   configuration options they provide to proxy operators (especially the
3050   default configuration).
3051</t>
3052<t>
3053   Users of a proxy need to be aware that they are no trustworthier than
3054   the people who run the proxy; HTTP itself cannot solve this problem.
3055</t>
3056<t>
3057   The judicious use of cryptography, when appropriate, may suffice to
3058   protect against a broad range of security and privacy attacks. Such
3059   cryptography is beyond the scope of the HTTP/1.1 specification.
3060</t>
3061</section>
3062
3063<section title="Denial of Service Attacks on Proxies" anchor="attack.DoS">
3064<t>
3065   They exist. They are hard to defend against. Research continues.
3066   Beware.
3067</t>
3068</section>
3069</section>
3070
3071<section title="Acknowledgments" anchor="ack">
3072<t>
3073   This specification makes heavy use of the augmented BNF and generic
3074   constructs defined by David H. Crocker for <xref target="RFC822ABNF"/>. Similarly, it
3075   reuses many of the definitions provided by Nathaniel Borenstein and
3076   Ned Freed for MIME <xref target="RFC2045"/>. We hope that their inclusion in this
3077   specification will help reduce past confusion over the relationship
3078   between HTTP and Internet mail message formats.
3079</t>
3080<t>
3081   HTTP has evolved considerably over the years. It has
3082   benefited from a large and active developer community--the many
3083   people who have participated on the www-talk mailing list--and it is
3084   that community which has been most responsible for the success of
3085   HTTP and of the World-Wide Web in general. Marc Andreessen, Robert
3086   Cailliau, Daniel W. Connolly, Bob Denny, John Franks, Jean-Francois
3087   Groff, Phillip M. Hallam-Baker, Hakon W. Lie, Ari Luotonen, Rob
3088   McCool, Lou Montulli, Dave Raggett, Tony Sanders, and Marc
3089   VanHeyningen deserve special recognition for their efforts in
3090   defining early aspects of the protocol.
3091</t>
3092<t>
3093   This document has benefited greatly from the comments of all those
3094   participating in the HTTP-WG. In addition to those already mentioned,
3095   the following individuals have contributed to this specification:
3096</t>
3097<t>
3098   Gary Adams, Harald Tveit Alvestrand, Keith Ball, Brian Behlendorf,
3099   Paul Burchard, Maurizio Codogno, Mike Cowlishaw, Roman Czyborra,
3100   Michael A. Dolan, Daniel DuBois, David J. Fiander, Alan Freier, Marc Hedlund, Greg Herlihy,
3101   Koen Holtman, Alex Hopmann, Bob Jernigan, Shel Kaphan, Rohit Khare,
3102   John Klensin, Martijn Koster, Alexei Kosut, David M. Kristol,
3103   Daniel LaLiberte, Ben Laurie, Paul J. Leach, Albert Lunde,
3104   John C. Mallery, Jean-Philippe Martin-Flatin, Mitra, David Morris,
3105   Gavin Nicol, Ross Patterson, Bill Perry, Jeffrey Perry, Scott Powers, Owen Rees,
3106   Luigi Rizzo, David Robinson, Marc Salomon, Rich Salz,
3107   Allan M. Schiffman, Jim Seidman, Chuck Shotton, Eric W. Sink,
3108   Simon E. Spero, Richard N. Taylor, Robert S. Thau,
3109   Bill (BearHeart) Weinman, Francois Yergeau, Mary Ellen Zurko,
3110   Josh Cohen.
3111</t>
3112<t>
3113   Thanks to the "cave men" of Palo Alto. You know who you are.
3114</t>
3115<t>
3116   Jim Gettys (the editor of <xref target="RFC2616"/>) wishes particularly
3117   to thank Roy Fielding, the editor of <xref target="RFC2068"/>, along
3118   with John Klensin, Jeff Mogul, Paul Leach, Dave Kristol, Koen
3119   Holtman, John Franks, Josh Cohen, Alex Hopmann, Scott Lawrence, and
3120   Larry Masinter for their help. And thanks go particularly to Jeff
3121   Mogul and Scott Lawrence for performing the "MUST/MAY/SHOULD" audit.
3122</t>
3123<t>
3124   The Apache Group, Anselm Baird-Smith, author of Jigsaw, and Henrik
3125   Frystyk implemented RFC 2068 early, and we wish to thank them for the
3126   discovery of many of the problems that this document attempts to
3127   rectify.
3128</t>
3129</section>
3130
3131</middle>
3132<back>
3133
3134<references title="Normative References">
3135
3136<reference anchor="ISO-8859-1">
3137  <front>
3138    <title>
3139     Information technology -- 8-bit single-byte coded graphic character sets -- Part 1: Latin alphabet No. 1
3140    </title>
3141    <author>
3142      <organization>International Organization for Standardization</organization>
3143    </author>
3144    <date year="1998"/>
3145  </front>
3146  <seriesInfo name="ISO/IEC" value="8859-1:1998"/>
3147</reference>
3148
3149<reference anchor="Part2">
3150  <front>
3151    <title abbrev="HTTP/1.1">HTTP/1.1, part 2: Message Semantics</title>
3152    <author initials="R." surname="Fielding" fullname="Roy T. Fielding" role="editor">
3153      <organization abbrev="Day Software">Day Software</organization>
3154      <address><email>fielding@gbiv.com</email></address>
3155    </author>
3156    <author initials="J." surname="Gettys" fullname="Jim Gettys">
3157      <organization>One Laptop per Child</organization>
3158      <address><email>jg@laptop.org</email></address>
3159    </author>
3160    <author initials="J." surname="Mogul" fullname="Jeffrey C. Mogul">
3161      <organization abbrev="HP">Hewlett-Packard Company</organization>
3162      <address><email>JeffMogul@acm.org</email></address>
3163    </author>
3164    <author initials="H." surname="Frystyk" fullname="Henrik Frystyk Nielsen">
3165      <organization abbrev="Microsoft">Microsoft Corporation</organization>
3166      <address><email>henrikn@microsoft.com</email></address>
3167    </author>
3168    <author initials="L." surname="Masinter" fullname="Larry Masinter">
3169      <organization abbrev="Adobe Systems">Adobe Systems, Incorporated</organization>
3170      <address><email>LMM@acm.org</email></address>
3171    </author>
3172    <author initials="P." surname="Leach" fullname="Paul J. Leach">
3173      <organization abbrev="Microsoft">Microsoft Corporation</organization>
3174      <address><email>paulle@microsoft.com</email></address>
3175    </author>
3176    <author initials="T." surname="Berners-Lee" fullname="Tim Berners-Lee">
3177      <organization abbrev="W3C/MIT">World Wide Web Consortium</organization>
3178      <address><email>timbl@w3.org</email></address>
3179    </author>
3180    <author initials="Y." surname="Lafon" fullname="Yves Lafon" role="editor">
3181      <organization abbrev="W3C">World Wide Web Consortium</organization>
3182      <address><email>ylafon@w3.org</email></address>
3183    </author>
3184    <author initials="J. F." surname="Reschke" fullname="Julian F. Reschke" role="editor">
3185      <organization abbrev="greenbytes">greenbytes GmbH</organization>
3186      <address><email>julian.reschke@greenbytes.de</email></address>
3187    </author>
3188    <date month="June" year="2008"/>
3189  </front>
3190  <seriesInfo name="Internet-Draft" value="draft-ietf-httpbis-p2-semantics-03"/>
3191 
3192</reference>
3193
3194<reference anchor="Part3">
3195  <front>
3196    <title abbrev="HTTP/1.1">HTTP/1.1, part 3: Message Payload and Content Negotiation</title>
3197    <author initials="R." surname="Fielding" fullname="Roy T. Fielding" role="editor">
3198      <organization abbrev="Day Software">Day Software</organization>
3199      <address><email>fielding@gbiv.com</email></address>
3200    </author>
3201    <author initials="J." surname="Gettys" fullname="Jim Gettys">
3202      <organization>One Laptop per Child</organization>
3203      <address><email>jg@laptop.org</email></address>
3204    </author>
3205    <author initials="J." surname="Mogul" fullname="Jeffrey C. Mogul">
3206      <organization abbrev="HP">Hewlett-Packard Company</organization>
3207      <address><email>JeffMogul@acm.org</email></address>
3208    </author>
3209    <author initials="H." surname="Frystyk" fullname="Henrik Frystyk Nielsen">
3210      <organization abbrev="Microsoft">Microsoft Corporation</organization>
3211      <address><email>henrikn@microsoft.com</email></address>
3212    </author>
3213    <author initials="L." surname="Masinter" fullname="Larry Masinter">
3214      <organization abbrev="Adobe Systems">Adobe Systems, Incorporated</organization>
3215      <address><email>LMM@acm.org</email></address>
3216    </author>
3217    <author initials="P." surname="Leach" fullname="Paul J. Leach">
3218      <organization abbrev="Microsoft">Microsoft Corporation</organization>
3219      <address><email>paulle@microsoft.com</email></address>
3220    </author>
3221    <author initials="T." surname="Berners-Lee" fullname="Tim Berners-Lee">
3222      <organization abbrev="W3C/MIT">World Wide Web Consortium</organization>
3223      <address><email>timbl@w3.org</email></address>
3224    </author>
3225    <author initials="Y." surname="Lafon" fullname="Yves Lafon" role="editor">
3226      <organization abbrev="W3C">World Wide Web Consortium</organization>
3227      <address><email>ylafon@w3.org</email></address>
3228    </author>
3229    <author initials="J. F." surname="Reschke" fullname="Julian F. Reschke" role="editor">
3230      <organization abbrev="greenbytes">greenbytes GmbH</organization>
3231      <address><email>julian.reschke@greenbytes.de</email></address>
3232    </author>
3233    <date month="June" year="2008"/>
3234  </front>
3235  <seriesInfo name="Internet-Draft" value="draft-ietf-httpbis-p3-payload-03"/>
3236 
3237</reference>
3238
3239<reference anchor="Part5">
3240  <front>
3241    <title abbrev="HTTP/1.1">HTTP/1.1, part 5: Range Requests and Partial Responses</title>
3242    <author initials="R." surname="Fielding" fullname="Roy T. Fielding" role="editor">
3243      <organization abbrev="Day Software">Day Software</organization>
3244      <address><email>fielding@gbiv.com</email></address>
3245    </author>
3246    <author initials="J." surname="Gettys" fullname="Jim Gettys">
3247      <organization>One Laptop per Child</organization>
3248      <address><email>jg@laptop.org</email></address>
3249    </author>
3250    <author initials="J." surname="Mogul" fullname="Jeffrey C. Mogul">
3251      <organization abbrev="HP">Hewlett-Packard Company</organization>
3252      <address><email>JeffMogul@acm.org</email></address>
3253    </author>
3254    <author initials="H." surname="Frystyk" fullname="Henrik Frystyk Nielsen">
3255      <organization abbrev="Microsoft">Microsoft Corporation</organization>
3256      <address><email>henrikn@microsoft.com</email></address>
3257    </author>
3258    <author initials="L." surname="Masinter" fullname="Larry Masinter">
3259      <organization abbrev="Adobe Systems">Adobe Systems, Incorporated</organization>
3260      <address><email>LMM@acm.org</email></address>
3261    </author>
3262    <author initials="P." surname="Leach" fullname="Paul J. Leach">
3263      <organization abbrev="Microsoft">Microsoft Corporation</organization>
3264      <address><email>paulle@microsoft.com</email></address>
3265    </author>
3266    <author initials="T." surname="Berners-Lee" fullname="Tim Berners-Lee">
3267      <organization abbrev="W3C/MIT">World Wide Web Consortium</organization>
3268      <address><email>timbl@w3.org</email></address>
3269    </author>
3270    <author initials="Y." surname="Lafon" fullname="Yves Lafon" role="editor">
3271      <organization abbrev="W3C">World Wide Web Consortium</organization>
3272      <address><email>ylafon@w3.org</email></address>
3273    </author>
3274    <author initials="J. F." surname="Reschke" fullname="Julian F. Reschke" role="editor">
3275      <organization abbrev="greenbytes">greenbytes GmbH</organization>
3276      <address><email>julian.reschke@greenbytes.de</email></address>
3277    </author>
3278    <date month="June" year="2008"/>
3279  </front>
3280  <seriesInfo name="Internet-Draft" value="draft-ietf-httpbis-p5-range-03"/>
3281 
3282</reference>
3283
3284<reference anchor="Part6">
3285  <front>
3286    <title abbrev="HTTP/1.1">HTTP/1.1, part 6: Caching</title>
3287    <author initials="R." surname="Fielding" fullname="Roy T. Fielding" role="editor">
3288      <organization abbrev="Day Software">Day Software</organization>
3289      <address><email>fielding@gbiv.com</email></address>
3290    </author>
3291    <author initials="J." surname="Gettys" fullname="Jim Gettys">
3292      <organization>One Laptop per Child</organization>
3293      <address><email>jg@laptop.org</email></address>
3294    </author>
3295    <author initials="J." surname="Mogul" fullname="Jeffrey C. Mogul">
3296      <organization abbrev="HP">Hewlett-Packard Company</organization>
3297      <address><email>JeffMogul@acm.org</email></address>
3298    </author>
3299    <author initials="H." surname="Frystyk" fullname="Henrik Frystyk Nielsen">
3300      <organization abbrev="Microsoft">Microsoft Corporation</organization>
3301      <address><email>henrikn@microsoft.com</email></address>
3302    </author>
3303    <author initials="L." surname="Masinter" fullname="Larry Masinter">
3304      <organization abbrev="Adobe Systems">Adobe Systems, Incorporated</organization>
3305      <address><email>LMM@acm.org</email></address>
3306    </author>
3307    <author initials="P." surname="Leach" fullname="Paul J. Leach">
3308      <organization abbrev="Microsoft">Microsoft Corporation</organization>
3309      <address><email>paulle@microsoft.com</email></address>
3310    </author>
3311    <author initials="T." surname="Berners-Lee" fullname="Tim Berners-Lee">
3312      <organization abbrev="W3C/MIT">World Wide Web Consortium</organization>
3313      <address><email>timbl@w3.org</email></address>
3314    </author>
3315    <author initials="Y." surname="Lafon" fullname="Yves Lafon" role="editor">
3316      <organization abbrev="W3C">World Wide Web Consortium</organization>
3317      <address><email>ylafon@w3.org</email></address>
3318    </author>
3319    <author initials="J. F." surname="Reschke" fullname="Julian F. Reschke" role="editor">
3320      <organization abbrev="greenbytes">greenbytes GmbH</organization>
3321      <address><email>julian.reschke@greenbytes.de</email></address>
3322    </author>
3323    <date month="June" year="2008"/>
3324  </front>
3325  <seriesInfo name="Internet-Draft" value="draft-ietf-httpbis-p6-cache-03"/>
3326 
3327</reference>
3328
3329<reference anchor="RFC822ABNF">
3330  <front>
3331    <title abbrev="Standard for ARPA Internet Text Messages">Standard for the format of ARPA Internet text messages</title>
3332    <author initials="D.H." surname="Crocker" fullname="David H. Crocker">
3333      <organization>University of Delaware, Dept. of Electrical Engineering</organization>
3334      <address><email>DCrocker@UDel-Relay</email></address>
3335    </author>
3336    <date month="August" day="13" year="1982"/>
3337  </front>
3338  <seriesInfo name="STD" value="11"/>
3339  <seriesInfo name="RFC" value="822"/>
3340</reference>
3341
3342<reference anchor="RFC2045">
3343  <front>
3344    <title abbrev="Internet Message Bodies">Multipurpose Internet Mail Extensions (MIME) Part One: Format of Internet Message Bodies</title>
3345    <author initials="N." surname="Freed" fullname="Ned Freed">
3346      <organization>Innosoft International, Inc.</organization>
3347      <address><email>ned@innosoft.com</email></address>
3348    </author>
3349    <author initials="N.S." surname="Borenstein" fullname="Nathaniel S. Borenstein">
3350      <organization>First Virtual Holdings</organization>
3351      <address><email>nsb@nsb.fv.com</email></address>
3352    </author>
3353    <date month="November" year="1996"/>
3354  </front>
3355  <seriesInfo name="RFC" value="2045"/>
3356</reference>
3357
3358<reference anchor="RFC2047">
3359  <front>
3360    <title abbrev="Message Header Extensions">MIME (Multipurpose Internet Mail Extensions) Part Three: Message Header Extensions for Non-ASCII Text</title>
3361    <author initials="K." surname="Moore" fullname="Keith Moore">
3362      <organization>University of Tennessee</organization>
3363      <address><email>moore@cs.utk.edu</email></address>
3364    </author>
3365    <date month="November" year="1996"/>
3366  </front>
3367  <seriesInfo name="RFC" value="2047"/>
3368</reference>
3369
3370<reference anchor="RFC2119">
3371  <front>
3372    <title>Key words for use in RFCs to Indicate Requirement Levels</title>
3373    <author initials="S." surname="Bradner" fullname="Scott Bradner">
3374      <organization>Harvard University</organization>
3375      <address><email>sob@harvard.edu</email></address>
3376    </author>
3377    <date month="March" year="1997"/>
3378  </front>
3379  <seriesInfo name="BCP" value="14"/>
3380  <seriesInfo name="RFC" value="2119"/>
3381</reference>
3382
3383<reference anchor="RFC2396">
3384  <front>
3385    <title abbrev="URI Generic Syntax">Uniform Resource Identifiers (URI): Generic Syntax</title>
3386    <author initials="T." surname="Berners-Lee" fullname="Tim Berners-Lee">
3387      <organization abbrev="MIT/LCS">World Wide Web Consortium</organization>
3388      <address><email>timbl@w3.org</email></address>
3389    </author>
3390    <author initials="R.T." surname="Fielding" fullname="Roy T. Fielding">
3391      <organization abbrev="U.C. Irvine">Department of Information and Computer Science</organization>
3392      <address><email>fielding@ics.uci.edu</email></address>
3393    </author>
3394    <author initials="L." surname="Masinter" fullname="Larry Masinter">
3395      <organization abbrev="Xerox Corporation">Xerox PARC</organization>
3396      <address><email>masinter@parc.xerox.com</email></address>
3397    </author>
3398    <date month="August" year="1998"/>
3399  </front>
3400  <seriesInfo name="RFC" value="2396"/>
3401</reference>
3402
3403<reference anchor="USASCII">
3404  <front>
3405    <title>Coded Character Set -- 7-bit American Standard Code for Information Interchange</title>
3406    <author>
3407      <organization>American National Standards Institute</organization>
3408    </author>
3409    <date year="1986"/>
3410  </front>
3411  <seriesInfo name="ANSI" value="X3.4"/>
3412</reference>
3413
3414</references>
3415
3416<references title="Informative References">
3417
3418<reference anchor="Nie1997" target="http://doi.acm.org/10.1145/263105.263157">
3419  <front>
3420    <title>Network Performance Effects of HTTP/1.1, CSS1, and PNG</title>
3421    <author initials="H.F.." surname="Nielsen" fullname="H.F. Nielsen">
3422      <organization/>
3423    </author>
3424    <author initials="J." surname="Gettys" fullname="J. Gettys">
3425      <organization/>
3426    </author>
3427    <author initials="E." surname="Prud'hommeaux" fullname="E. Prud'hommeaux">
3428      <organization/>
3429    </author>
3430    <author initials="H." surname="Lie" fullname="H. Lie">
3431      <organization/>
3432    </author>
3433    <author initials="C." surname="Lilley" fullname="C. Lilley">
3434      <organization/>
3435    </author>
3436    <date year="1997" month="September"/>
3437  </front>
3438  <seriesInfo name="ACM" value="Proceedings of the ACM SIGCOMM '97 conference on Applications, technologies, architectures, and protocols for computer communication SIGCOMM '97"/>
3439</reference>
3440
3441<reference anchor="Pad1995">
3442  <front>
3443    <title>Improving HTTP Latency</title>
3444    <author initials="V.N." surname="Padmanabhan" fullname="Venkata N. Padmanabhan">
3445      <organization/>
3446    </author>
3447    <author initials="J.C." surname="Mogul" fullname="Jeffrey C. Mogul">
3448      <organization/>
3449    </author>
3450    <date year="1995" month="December"/>
3451  </front>
3452  <seriesInfo name="Computer Networks and ISDN Systems" value="v. 28, pp. 25-35"/>
3453  <annotation>
3454    Slightly revised version of paper in Proc. 2nd International WWW Conference '94: Mosaic and the Web, Oct. 1994,
3455    which is available at <eref target="http://www.ncsa.uiuc.edu/SDG/IT94/Proceedings/DDay/mogul/HTTPLatency.html"/>.
3456  </annotation>
3457</reference>
3458
3459<reference anchor="RFC822">
3460  <front>
3461    <title abbrev="Standard for ARPA Internet Text Messages">Standard for the format of ARPA Internet text messages</title>
3462    <author initials="D.H." surname="Crocker" fullname="David H. Crocker">
3463      <organization>University of Delaware, Dept. of Electrical Engineering</organization>
3464      <address><email>DCrocker@UDel-Relay</email></address>
3465    </author>
3466    <date month="August" day="13" year="1982"/>
3467  </front>
3468  <seriesInfo name="STD" value="11"/>
3469  <seriesInfo name="RFC" value="822"/>
3470</reference>
3471
3472<reference anchor="RFC959">
3473  <front>
3474    <title abbrev="File Transfer Protocol">File Transfer Protocol</title>
3475    <author initials="J." surname="Postel" fullname="J. Postel">
3476      <organization>Information Sciences Institute (ISI)</organization>
3477    </author>
3478    <author initials="J." surname="Reynolds" fullname="J. Reynolds">
3479      <organization/>
3480    </author>
3481    <date month="October" year="1985"/>
3482  </front>
3483  <seriesInfo name="STD" value="9"/>
3484  <seriesInfo name="RFC" value="959"/>
3485</reference>
3486
3487<reference anchor="RFC1123">
3488  <front>
3489    <title>Requirements for Internet Hosts - Application and Support</title>
3490    <author initials="R." surname="Braden" fullname="Robert Braden">
3491      <organization>University of Southern California (USC), Information Sciences Institute</organization>
3492      <address><email>Braden@ISI.EDU</email></address>
3493    </author>
3494    <date month="October" year="1989"/>
3495  </front>
3496  <seriesInfo name="STD" value="3"/>
3497  <seriesInfo name="RFC" value="1123"/>
3498</reference>
3499
3500<reference anchor="RFC1305">
3501  <front>
3502    <title>Network Time Protocol (Version 3) Specification, Implementation</title>
3503    <author initials="D." surname="Mills" fullname="David L. Mills">
3504      <organization>University of Delaware, Electrical Engineering Department</organization>
3505      <address><email>mills@udel.edu</email></address>
3506    </author>
3507    <date month="March" year="1992"/>
3508  </front>
3509  <seriesInfo name="RFC" value="1305"/>
3510</reference>
3511
3512<reference anchor="RFC1436">
3513  <front>
3514    <title abbrev="Gopher">The Internet Gopher Protocol (a distributed document search and retrieval protocol)</title>
3515    <author initials="F." surname="Anklesaria" fullname="Farhad Anklesaria">
3516      <organization>University of Minnesota, Computer and Information Services</organization>
3517      <address><email>fxa@boombox.micro.umn.edu</email></address>
3518    </author>
3519    <author initials="M." surname="McCahill" fullname="Mark McCahill">
3520      <organization>University of Minnesota, Computer and Information Services</organization>
3521      <address><email>mpm@boombox.micro.umn.edu</email></address>
3522    </author>
3523    <author initials="P." surname="Lindner" fullname="Paul Lindner">
3524      <organization>University of Minnesota, Computer and Information Services</organization>
3525      <address><email>lindner@boombox.micro.umn.edu</email></address>
3526    </author>
3527    <author initials="D." surname="Johnson" fullname="David Johnson">
3528      <organization>University of Minnesota, Computer and Information Services</organization>
3529      <address><email>dmj@boombox.micro.umn.edu</email></address>
3530    </author>
3531    <author initials="D." surname="Torrey" fullname="Daniel Torrey">
3532      <organization>University of Minnesota, Computer and Information Services</organization>
3533      <address><email>daniel@boombox.micro.umn.edu</email></address>
3534    </author>
3535    <author initials="B." surname="Alberti" fullname="Bob Alberti">
3536      <organization>University of Minnesota, Computer and Information Services</organization>
3537      <address><email>alberti@boombox.micro.umn.edu</email></address>
3538    </author>
3539    <date month="March" year="1993"/>
3540  </front>
3541  <seriesInfo name="RFC" value="1436"/>
3542</reference>
3543
3544<reference anchor="RFC1630">
3545  <front>
3546    <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>
3547    <author initials="T." surname="Berners-Lee" fullname="Tim Berners-Lee">
3548      <organization>CERN, World-Wide Web project</organization>
3549      <address><email>timbl@info.cern.ch</email></address>
3550    </author>
3551    <date month="June" year="1994"/>
3552  </front>
3553  <seriesInfo name="RFC" value="1630"/>
3554</reference>
3555
3556<reference anchor="RFC1737">
3557  <front>
3558    <title abbrev="Requirements for Uniform Resource Names">Functional Requirements for Uniform Resource Names</title>
3559    <author initials="L." surname="Masinter" fullname="Larry Masinter">
3560      <organization>Xerox Palo Alto Research Center</organization>
3561      <address><email>masinter@parc.xerox.com</email></address>
3562    </author>
3563    <author initials="K." surname="Sollins" fullname="Karen Sollins">
3564      <organization>MIT Laboratory for Computer Science</organization>
3565      <address><email>sollins@lcs.mit.edu</email></address>
3566    </author>
3567    <date month="December" year="1994"/>
3568  </front>
3569  <seriesInfo name="RFC" value="1737"/>
3570</reference>
3571
3572<reference anchor="RFC1738">
3573  <front>
3574    <title>Uniform Resource Locators (URL)</title>
3575    <author initials="T." surname="Berners-Lee" fullname="Tim Berners-Lee">
3576      <organization>CERN, World-Wide Web project</organization>
3577      <address><email>timbl@info.cern.ch</email></address>
3578    </author>
3579    <author initials="L." surname="Masinter" fullname="Larry Masinter">
3580      <organization>Xerox PARC</organization>
3581      <address><email>masinter@parc.xerox.com</email></address>
3582    </author>
3583    <author initials="M." surname="McCahill" fullname="Mark McCahill">
3584      <organization>University of Minnesota, Computer and Information Services</organization>
3585      <address><email>mpm@boombox.micro.umn.edu</email></address>
3586    </author>
3587    <date month="December" year="1994"/>
3588  </front>
3589  <seriesInfo name="RFC" value="1738"/>
3590</reference>
3591
3592<reference anchor="RFC1808">
3593  <front>
3594    <title>Relative Uniform Resource Locators</title>
3595    <author initials="R." surname="Fielding" fullname="Roy T. Fielding">
3596      <organization>University of California Irvine, Department of Information and Computer Science</organization>
3597      <address><email>fielding@ics.uci.edu</email></address>
3598    </author>
3599    <date month="June" year="1995"/>
3600  </front>
3601  <seriesInfo name="RFC" value="1808"/>
3602</reference>
3603
3604<reference anchor="RFC1900">
3605  <front>
3606    <title>Renumbering Needs Work</title>
3607    <author initials="B." surname="Carpenter" fullname="Brian E. Carpenter">
3608      <organization>CERN, Computing and Networks Division</organization>
3609      <address><email>brian@dxcoms.cern.ch</email></address>
3610    </author>
3611    <author initials="Y." surname="Rekhter" fullname="Yakov Rekhter">
3612      <organization>cisco Systems</organization>
3613      <address><email>yakov@cisco.com</email></address>
3614    </author>
3615    <date month="February" year="1996"/>
3616  </front>
3617  <seriesInfo name="RFC" value="1900"/>
3618</reference>
3619
3620<reference anchor="RFC1945">
3621  <front>
3622    <title abbrev="HTTP/1.0">Hypertext Transfer Protocol -- HTTP/1.0</title>
3623    <author initials="T." surname="Berners-Lee" fullname="Tim Berners-Lee">
3624      <organization>MIT, Laboratory for Computer Science</organization>
3625      <address><email>timbl@w3.org</email></address>
3626    </author>
3627    <author initials="R.T." surname="Fielding" fullname="Roy T. Fielding">
3628      <organization>University of California, Irvine, Department of Information and Computer Science</organization>
3629      <address><email>fielding@ics.uci.edu</email></address>
3630    </author>
3631    <author initials="H.F." surname="Nielsen" fullname="Henrik Frystyk Nielsen">
3632      <organization>W3 Consortium, MIT Laboratory for Computer Science</organization>
3633      <address><email>frystyk@w3.org</email></address>
3634    </author>
3635    <date month="May" year="1996"/>
3636  </front>
3637  <seriesInfo name="RFC" value="1945"/>
3638</reference>
3639
3640<reference anchor="RFC2068">
3641  <front>
3642    <title abbrev="HTTP/1.1">Hypertext Transfer Protocol -- HTTP/1.1</title>
3643    <author initials="R." surname="Fielding" fullname="Roy T. Fielding">
3644      <organization>University of California, Irvine, Department of Information and Computer Science</organization>
3645      <address><email>fielding@ics.uci.edu</email></address>
3646    </author>
3647    <author initials="J." surname="Gettys" fullname="Jim Gettys">
3648      <organization>MIT Laboratory for Computer Science</organization>
3649      <address><email>jg@w3.org</email></address>
3650    </author>
3651    <author initials="J." surname="Mogul" fullname="Jeffrey C. Mogul">
3652      <organization>Digital Equipment Corporation, Western Research Laboratory</organization>
3653      <address><email>mogul@wrl.dec.com</email></address>
3654    </author>
3655    <author initials="H." surname="Nielsen" fullname="Henrik Frystyk Nielsen">
3656      <organization>MIT Laboratory for Computer Science</organization>
3657      <address><email>frystyk@w3.org</email></address>
3658    </author>
3659    <author initials="T." surname="Berners-Lee" fullname="Tim Berners-Lee">
3660      <organization>MIT Laboratory for Computer Science</organization>
3661      <address><email>timbl@w3.org</email></address>
3662    </author>
3663    <date month="January" year="1997"/>
3664  </front>
3665  <seriesInfo name="RFC" value="2068"/>
3666</reference>
3667
3668<reference anchor="RFC2145">
3669  <front>
3670    <title abbrev="HTTP Version Numbers">Use and Interpretation of HTTP Version Numbers</title>
3671    <author initials="J.C." surname="Mogul" fullname="Jeffrey C. Mogul">
3672      <organization>Western Research Laboratory</organization>
3673      <address><email>mogul@wrl.dec.com</email></address>
3674    </author>
3675    <author initials="R.T." surname="Fielding" fullname="Roy T. Fielding">
3676      <organization>Department of Information and Computer Science</organization>
3677      <address><email>fielding@ics.uci.edu</email></address>
3678    </author>
3679    <author initials="J." surname="Gettys" fullname="Jim Gettys">
3680      <organization>MIT Laboratory for Computer Science</organization>
3681      <address><email>jg@w3.org</email></address>
3682    </author>
3683    <author initials="H.F." surname="Nielsen" fullname="Henrik Frystyk Nielsen">
3684      <organization>W3 Consortium</organization>
3685      <address><email>frystyk@w3.org</email></address>
3686    </author>
3687    <date month="May" year="1997"/>
3688  </front>
3689  <seriesInfo name="RFC" value="2145"/>
3690</reference>
3691
3692<reference anchor="RFC2324">
3693  <front>
3694    <title abbrev="HTCPCP/1.0">Hyper Text Coffee Pot Control Protocol (HTCPCP/1.0)</title>
3695    <author initials="L." surname="Masinter" fullname="Larry Masinter">
3696      <organization>Xerox Palo Alto Research Center</organization>
3697      <address><email>masinter@parc.xerox.com</email></address>
3698    </author>
3699    <date month="April" day="1" year="1998"/>
3700  </front>
3701  <seriesInfo name="RFC" value="2324"/>
3702</reference>
3703
3704<reference anchor="RFC2616">
3705  <front>
3706    <title>Hypertext Transfer Protocol -- HTTP/1.1</title>
3707    <author initials="R." surname="Fielding" fullname="R. Fielding">
3708      <organization>University of California, Irvine</organization>
3709      <address><email>fielding@ics.uci.edu</email></address>
3710    </author>
3711    <author initials="J." surname="Gettys" fullname="J. Gettys">
3712      <organization>W3C</organization>
3713      <address><email>jg@w3.org</email></address>
3714    </author>
3715    <author initials="J." surname="Mogul" fullname="J. Mogul">
3716      <organization>Compaq Computer Corporation</organization>
3717      <address><email>mogul@wrl.dec.com</email></address>
3718    </author>
3719    <author initials="H." surname="Frystyk" fullname="H. Frystyk">
3720      <organization>MIT Laboratory for Computer Science</organization>
3721      <address><email>frystyk@w3.org</email></address>
3722    </author>
3723    <author initials="L." surname="Masinter" fullname="L. Masinter">
3724      <organization>Xerox Corporation</organization>
3725      <address><email>masinter@parc.xerox.com</email></address>
3726    </author>
3727    <author initials="P." surname="Leach" fullname="P. Leach">
3728      <organization>Microsoft Corporation</organization>
3729      <address><email>paulle@microsoft.com</email></address>
3730    </author>
3731    <author initials="T." surname="Berners-Lee" fullname="T. Berners-Lee">
3732      <organization>W3C</organization>
3733      <address><email>timbl@w3.org</email></address>
3734    </author>
3735    <date month="June" year="1999"/>
3736  </front>
3737  <seriesInfo name="RFC" value="2616"/>
3738</reference>
3739
3740<reference anchor="RFC2821">
3741  <front>
3742    <title>Simple Mail Transfer Protocol</title>
3743    <author initials="J." surname="Klensin" fullname="J. Klensin">
3744      <organization>AT&amp;T Laboratories</organization>
3745      <address><email>klensin@research.att.com</email></address>
3746    </author>
3747    <date year="2001" month="April"/>
3748  </front>
3749  <seriesInfo name="RFC" value="2821"/>
3750</reference>
3751
3752<reference anchor="RFC2822">
3753  <front>
3754    <title>Internet Message Format</title>
3755    <author initials="P." surname="Resnick" fullname="P. Resnick">
3756      <organization>QUALCOMM Incorporated</organization>
3757    </author>
3758    <date year="2001" month="April"/>
3759  </front> 
3760  <seriesInfo name="RFC" value="2822"/>
3761</reference>
3762
3763<reference anchor="RFC3864">
3764  <front>
3765    <title>Registration Procedures for Message Header Fields</title>
3766    <author initials="G." surname="Klyne" fullname="G. Klyne">
3767      <organization>Nine by Nine</organization>
3768      <address><email>GK-IETF@ninebynine.org</email></address>
3769    </author>
3770    <author initials="M." surname="Nottingham" fullname="M. Nottingham">
3771      <organization>BEA Systems</organization>
3772      <address><email>mnot@pobox.com</email></address>
3773    </author>
3774    <author initials="J." surname="Mogul" fullname="J. Mogul">
3775      <organization>HP Labs</organization>
3776      <address><email>JeffMogul@acm.org</email></address>
3777    </author>
3778    <date year="2004" month="September"/>
3779  </front>
3780  <seriesInfo name="BCP" value="90"/>
3781  <seriesInfo name="RFC" value="3864"/>
3782</reference>
3783
3784<reference anchor="RFC3977">
3785  <front>
3786    <title>Network News Transfer Protocol (NNTP)</title>
3787    <author initials="C." surname="Feather" fullname="C. Feather">
3788      <organization>THUS plc</organization>
3789      <address><email>clive@demon.net</email></address>
3790    </author>
3791    <date year="2006" month="October"/>
3792  </front>
3793  <seriesInfo name="RFC" value="3977"/>
3794</reference>
3795
3796<reference anchor="RFC4288">
3797  <front>
3798    <title>Media Type Specifications and Registration Procedures</title>
3799    <author initials="N." surname="Freed" fullname="N. Freed">
3800      <organization>Sun Microsystems</organization>
3801      <address>
3802        <email>ned.freed@mrochek.com</email>
3803      </address>
3804    </author>
3805    <author initials="J." surname="Klensin" fullname="J. Klensin">
3806      <organization/>
3807      <address>
3808        <email>klensin+ietf@jck.com</email>
3809      </address>
3810    </author>
3811    <date year="2005" month="December"/>
3812  </front>
3813  <seriesInfo name="BCP" value="13"/>
3814  <seriesInfo name="RFC" value="4288"/>
3815</reference>
3816
3817<reference anchor="Spe" target="http://sunsite.unc.edu/mdma-release/http-prob.html">
3818  <front>
3819  <title>Analysis of HTTP Performance Problems</title>
3820  <author initials="S." surname="Spero" fullname="Simon E. Spero">
3821    <organization/>
3822  </author>
3823  <date/>
3824  </front>
3825</reference>
3826
3827<reference anchor="Tou1998" target="http://www.isi.edu/touch/pubs/http-perf96/">
3828  <front>
3829  <title>Analysis of HTTP Performance</title>
3830  <author initials="J." surname="Touch" fullname="Joe Touch">
3831    <organization>USC/Information Sciences Institute</organization>
3832    <address><email>touch@isi.edu</email></address>
3833  </author>
3834  <author initials="J." surname="Heidemann" fullname="John Heidemann">
3835    <organization>USC/Information Sciences Institute</organization>
3836    <address><email>johnh@isi.edu</email></address>
3837  </author>
3838  <author initials="K." surname="Obraczka" fullname="Katia Obraczka">
3839    <organization>USC/Information Sciences Institute</organization>
3840    <address><email>katia@isi.edu</email></address>
3841  </author>
3842  <date year="1998" month="Aug"/>
3843  </front>
3844  <seriesInfo name="ISI Research Report" value="ISI/RR-98-463"/>
3845  <annotation>(original report dated Aug. 1996)</annotation>
3846</reference>
3847
3848<reference anchor="WAIS">
3849  <front>
3850    <title>WAIS Interface Protocol Prototype Functional Specification (v1.5)</title>
3851    <author initials="F." surname="Davis" fullname="F. Davis">
3852      <organization>Thinking Machines Corporation</organization>
3853    </author>
3854    <author initials="B." surname="Kahle" fullname="B. Kahle">
3855      <organization>Thinking Machines Corporation</organization>
3856    </author>
3857    <author initials="H." surname="Morris" fullname="H. Morris">
3858      <organization>Thinking Machines Corporation</organization>
3859    </author>
3860    <author initials="J." surname="Salem" fullname="J. Salem">
3861      <organization>Thinking Machines Corporation</organization>
3862    </author>
3863    <author initials="T." surname="Shen" fullname="T. Shen">
3864      <organization>Thinking Machines Corporation</organization>
3865    </author>
3866    <author initials="R." surname="Wang" fullname="R. Wang">
3867      <organization>Thinking Machines Corporation</organization>
3868    </author>
3869    <author initials="J." surname="Sui" fullname="J. Sui">
3870      <organization>Thinking Machines Corporation</organization>
3871    </author>
3872    <author initials="M." surname="Grinbaum" fullname="M. Grinbaum">
3873      <organization>Thinking Machines Corporation</organization>
3874    </author>
3875    <date month="April" year="1990"/>
3876  </front>
3877  <seriesInfo name="Thinking Machines Corporation" value=""/>
3878</reference>
3879
3880</references>
3881
3882
3883<section title="Internet Media Types" anchor="internet.media.type.http">
3884<t>
3885   In addition to defining HTTP/1.1, this document serves
3886   as the specification for the Internet media type "message/http" and
3887   "application/http". The following is to be registered with IANA <xref target="RFC4288"/>.
3888</t>
3889<section title="Internet Media Type message/http" anchor="internet.media.type.message.http">
3890<iref item="Media Type" subitem="message/http" primary="true"/>
3891<iref item="message/http Media Type" primary="true"/>
3892<t>
3893   The message/http type can be used to enclose a single HTTP request or
3894   response message, provided that it obeys the MIME restrictions for all
3895   "message" types regarding line length and encodings.
3896</t>
3897<t>
3898  <list style="hanging">
3899    <t hangText="Type name:">
3900      message
3901    </t>
3902    <t hangText="Subtype name:">
3903      http
3904    </t>
3905    <t hangText="Required parameters:">
3906      none
3907    </t>
3908    <t hangText="Optional parameters:">
3909      version, msgtype
3910      <list style="hanging">
3911        <t hangText="version:">
3912          The HTTP-Version number of the enclosed message
3913          (e.g., "1.1"). If not present, the version can be
3914          determined from the first line of the body.
3915        </t>
3916        <t hangText="msgtype:">
3917          The message type -- "request" or "response". If not
3918          present, the type can be determined from the first
3919          line of the body.
3920        </t>
3921      </list>
3922    </t>
3923    <t hangText="Encoding considerations:">
3924      only "7bit", "8bit", or "binary" are permitted
3925    </t>
3926    <t hangText="Security considerations:">
3927      none
3928    </t>
3929    <t hangText="Interoperability considerations:">
3930      none
3931    </t>
3932    <t hangText="Published specification:">
3933      This specification (see <xref target="internet.media.type.message.http"/>).
3934    </t>
3935    <t hangText="Applications that use this media type:">
3936    </t>
3937    <t hangText="Additional information:">
3938      <list style="hanging">
3939        <t hangText="Magic number(s):">none</t>
3940        <t hangText="File extension(s):">none</t>
3941        <t hangText="Macintosh file type code(s):">none</t>
3942      </list>
3943    </t>
3944    <t hangText="Person and email address to contact for further information:">
3945      See Authors Section.
3946    </t>
3947                <t hangText="Intended usage:">
3948                  COMMON
3949    </t>
3950                <t hangText="Restrictions on usage:">
3951                  none
3952    </t>
3953    <t hangText="Author/Change controller:">
3954      IESG
3955    </t>
3956  </list>
3957</t>
3958</section>
3959<section title="Internet Media Type application/http" anchor="internet.media.type.application.http">
3960<iref item="Media Type" subitem="application/http" primary="true"/>
3961<iref item="application/http Media Type" primary="true"/>
3962<t>
3963   The application/http type can be used to enclose a pipeline of one or more
3964   HTTP request or response messages (not intermixed).
3965</t>
3966<t>
3967  <list style="hanging">
3968    <t hangText="Type name:">
3969      application
3970    </t>
3971    <t hangText="Subtype name:">
3972      http
3973    </t>
3974    <t hangText="Required parameters:">
3975      none
3976    </t>
3977    <t hangText="Optional parameters:">
3978      version, msgtype
3979      <list style="hanging">
3980        <t hangText="version:">
3981          The HTTP-Version number of the enclosed messages
3982          (e.g., "1.1"). If not present, the version can be
3983          determined from the first line of the body.
3984        </t>
3985        <t hangText="msgtype:">
3986          The message type -- "request" or "response". If not
3987          present, the type can be determined from the first
3988          line of the body.
3989        </t>
3990      </list>
3991    </t>
3992    <t hangText="Encoding considerations:">
3993      HTTP messages enclosed by this type
3994      are in "binary" format; use of an appropriate
3995      Content-Transfer-Encoding is required when
3996      transmitted via E-mail.
3997    </t>
3998    <t hangText="Security considerations:">
3999      none
4000    </t>
4001    <t hangText="Interoperability considerations:">
4002      none
4003    </t>
4004    <t hangText="Published specification:">
4005      This specification (see <xref target="internet.media.type.application.http"/>).
4006    </t>
4007    <t hangText="Applications that use this media type:">
4008    </t>
4009    <t hangText="Additional information:">
4010      <list style="hanging">
4011        <t hangText="Magic number(s):">none</t>
4012        <t hangText="File extension(s):">none</t>
4013        <t hangText="Macintosh file type code(s):">none</t>
4014      </list>
4015    </t>
4016    <t hangText="Person and email address to contact for further information:">
4017      See Authors Section.
4018    </t>
4019                <t hangText="Intended usage:">
4020                  COMMON
4021    </t>
4022                <t hangText="Restrictions on usage:">
4023                  none
4024    </t>
4025    <t hangText="Author/Change controller:">
4026      IESG
4027    </t>
4028  </list>
4029</t>
4030</section>
4031</section>
4032
4033<section title="Tolerant Applications" anchor="tolerant.applications">
4034<t>
4035   Although this document specifies the requirements for the generation
4036   of HTTP/1.1 messages, not all applications will be correct in their
4037   implementation. We therefore recommend that operational applications
4038   be tolerant of deviations whenever those deviations can be
4039   interpreted unambiguously.
4040</t>
4041<t>
4042   Clients SHOULD be tolerant in parsing the Status-Line and servers
4043   tolerant when parsing the Request-Line. In particular, they SHOULD
4044   accept any amount of SP or HTAB characters between fields, even though
4045   only a single SP is required.
4046</t>
4047<t>
4048   The line terminator for message-header fields is the sequence CRLF.
4049   However, we recommend that applications, when parsing such headers,
4050   recognize a single LF as a line terminator and ignore the leading CR.
4051</t>
4052<t>
4053   The character set of an entity-body SHOULD be labeled as the lowest
4054   common denominator of the character codes used within that body, with
4055   the exception that not labeling the entity is preferred over labeling
4056   the entity with the labels US-ASCII or ISO-8859-1. See <xref target="Part3"/>.
4057</t>
4058<t>
4059   Additional rules for requirements on parsing and encoding of dates
4060   and other potential problems with date encodings include:
4061</t>
4062<t>
4063  <list style="symbols">
4064     <t>HTTP/1.1 clients and caches SHOULD assume that an RFC-850 date
4065        which appears to be more than 50 years in the future is in fact
4066        in the past (this helps solve the "year 2000" problem).</t>
4067
4068     <t>An HTTP/1.1 implementation MAY internally represent a parsed
4069        Expires date as earlier than the proper value, but MUST NOT
4070        internally represent a parsed Expires date as later than the
4071        proper value.</t>
4072
4073     <t>All expiration-related calculations MUST be done in GMT. The
4074        local time zone MUST NOT influence the calculation or comparison
4075        of an age or expiration time.</t>
4076
4077     <t>If an HTTP header incorrectly carries a date value with a time
4078        zone other than GMT, it MUST be converted into GMT using the
4079        most conservative possible conversion.</t>
4080  </list>
4081</t>
4082</section>
4083
4084<section title="Conversion of Date Formats" anchor="conversion.of.date.formats">
4085<t>
4086   HTTP/1.1 uses a restricted set of date formats (<xref target="full.date"/>) to
4087   simplify the process of date comparison. Proxies and gateways from
4088   other protocols SHOULD ensure that any Date header field present in a
4089   message conforms to one of the HTTP/1.1 formats and rewrite the date
4090   if necessary.
4091</t>
4092</section>
4093
4094<section title="Compatibility with Previous Versions" anchor="compatibility">
4095<t>
4096   It is beyond the scope of a protocol specification to mandate
4097   compliance with previous versions. HTTP/1.1 was deliberately
4098   designed, however, to make supporting previous versions easy. It is
4099   worth noting that, at the time of composing this specification
4100   (1996), we would expect commercial HTTP/1.1 servers to:
4101  <list style="symbols">
4102     <t>recognize the format of the Request-Line for HTTP/0.9, 1.0, and
4103        1.1 requests;</t>
4104
4105     <t>understand any valid request in the format of HTTP/0.9, 1.0, or
4106        1.1;</t>
4107
4108     <t>respond appropriately with a message in the same major version
4109        used by the client.</t>
4110  </list>
4111</t>
4112<t>
4113   And we would expect HTTP/1.1 clients to:
4114  <list style="symbols">
4115     <t>recognize the format of the Status-Line for HTTP/1.0 and 1.1
4116        responses;</t>
4117
4118     <t>understand any valid response in the format of HTTP/0.9, 1.0, or
4119        1.1.</t>
4120  </list>
4121</t>
4122<t>
4123   For most implementations of HTTP/1.0, each connection is established
4124   by the client prior to the request and closed by the server after
4125   sending the response. Some implementations implement the Keep-Alive
4126   version of persistent connections described in Section 19.7.1 of <xref target="RFC2068"/>.
4127</t>
4128
4129<section title="Changes from HTTP/1.0" anchor="changes.from.1.0">
4130<t>
4131   This section summarizes major differences between versions HTTP/1.0
4132   and HTTP/1.1.
4133</t>
4134
4135<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">
4136<t>
4137   The requirements that clients and servers support the Host request-header,
4138   report an error if the Host request-header (<xref target="header.host"/>) is
4139   missing from an HTTP/1.1 request, and accept absolute URIs (<xref target="request-uri"/>)
4140   are among the most important changes defined by this
4141   specification.
4142</t>
4143<t>
4144   Older HTTP/1.0 clients assumed a one-to-one relationship of IP
4145   addresses and servers; there was no other established mechanism for
4146   distinguishing the intended server of a request than the IP address
4147   to which that request was directed. The changes outlined above will
4148   allow the Internet, once older HTTP clients are no longer common, to
4149   support multiple Web sites from a single IP address, greatly
4150   simplifying large operational Web servers, where allocation of many
4151   IP addresses to a single host has created serious problems. The
4152   Internet will also be able to recover the IP addresses that have been
4153   allocated for the sole purpose of allowing special-purpose domain
4154   names to be used in root-level HTTP URLs. Given the rate of growth of
4155   the Web, and the number of servers already deployed, it is extremely
4156   important that all implementations of HTTP (including updates to
4157   existing HTTP/1.0 applications) correctly implement these
4158   requirements:
4159  <list style="symbols">
4160     <t>Both clients and servers MUST support the Host request-header.</t>
4161
4162     <t>A client that sends an HTTP/1.1 request MUST send a Host header.</t>
4163
4164     <t>Servers MUST report a 400 (Bad Request) error if an HTTP/1.1
4165        request does not include a Host request-header.</t>
4166
4167     <t>Servers MUST accept absolute URIs.</t>
4168  </list>
4169</t>
4170</section>
4171</section>
4172
4173<section title="Compatibility with HTTP/1.0 Persistent Connections" anchor="compatibility.with.http.1.0.persistent.connections">
4174<t>
4175   Some clients and servers might wish to be compatible with some
4176   previous implementations of persistent connections in HTTP/1.0
4177   clients and servers. Persistent connections in HTTP/1.0 are
4178   explicitly negotiated as they are not the default behavior. HTTP/1.0
4179   experimental implementations of persistent connections are faulty,
4180   and the new facilities in HTTP/1.1 are designed to rectify these
4181   problems. The problem was that some existing 1.0 clients may be
4182   sending Keep-Alive to a proxy server that doesn't understand
4183   Connection, which would then erroneously forward it to the next
4184   inbound server, which would establish the Keep-Alive connection and
4185   result in a hung HTTP/1.0 proxy waiting for the close on the
4186   response. The result is that HTTP/1.0 clients must be prevented from
4187   using Keep-Alive when talking to proxies.
4188</t>
4189<t>
4190   However, talking to proxies is the most important use of persistent
4191   connections, so that prohibition is clearly unacceptable. Therefore,
4192   we need some other mechanism for indicating a persistent connection
4193   is desired, which is safe to use even when talking to an old proxy
4194   that ignores Connection. Persistent connections are the default for
4195   HTTP/1.1 messages; we introduce a new keyword (Connection: close) for
4196   declaring non-persistence. See <xref target="header.connection"/>.
4197</t>
4198<t>
4199   The original HTTP/1.0 form of persistent connections (the Connection:
4200   Keep-Alive and Keep-Alive header) is documented in <xref target="RFC2068"/>.
4201</t>
4202</section>
4203
4204<section title="Changes from RFC 2068" anchor="changes.from.rfc.2068">
4205<t>
4206   This specification has been carefully audited to correct and
4207   disambiguate key word usage; RFC 2068 had many problems in respect to
4208   the conventions laid out in <xref target="RFC2119"/>.
4209</t>
4210<t>
4211   Transfer-coding and message lengths all interact in ways that
4212   required fixing exactly when chunked encoding is used (to allow for
4213   transfer encoding that may not be self delimiting); it was important
4214   to straighten out exactly how message lengths are computed. (Sections
4215   <xref target="transfer.codings" format="counter"/>, <xref target="message.length" format="counter"/>,
4216   <xref target="header.content-length" format="counter"/>,
4217   see also <xref target="Part3"/>, <xref target="Part5"/> and <xref target="Part6"/>)
4218</t>
4219<t>
4220   The use and interpretation of HTTP version numbers has been clarified
4221   by <xref target="RFC2145"/>. Require proxies to upgrade requests to highest protocol
4222   version they support to deal with problems discovered in HTTP/1.0
4223   implementations (<xref target="http.version"/>)
4224</t>
4225<t>
4226   Transfer-coding had significant problems, particularly with
4227   interactions with chunked encoding. The solution is that transfer-codings
4228   become as full fledged as content-codings. This involves
4229   adding an IANA registry for transfer-codings (separate from content
4230   codings), a new header field (TE) and enabling trailer headers in the
4231   future. Transfer encoding is a major performance benefit, so it was
4232   worth fixing <xref target="Nie1997"/>. TE also solves another, obscure, downward
4233   interoperability problem that could have occurred due to interactions
4234   between authentication trailers, chunked encoding and HTTP/1.0
4235   clients.(Section <xref target="transfer.codings" format="counter"/>, <xref target="chunked.transfer.encoding" format="counter"/>,
4236   and <xref target="header.te" format="counter"/>)
4237</t>
4238</section>
4239
4240<section title="Changes from RFC 2616" anchor="changes.from.rfc.2616">
4241<t>
4242  The CHAR rule does not allow the NUL character anymore (this affects
4243  the comment and quoted-string rules).  Furthermore, the quoted-pair
4244  rule does not allow escaping NUL, CR or LF anymore.
4245  (<xref target="basic.rules"/>)
4246</t>
4247<t>
4248  Clarify that HTTP-Version is case sensitive.
4249  (<xref target="http.version"/>)
4250</t>
4251<t>
4252  Remove reference to non-existant identity transfer-coding value tokens.
4253  (Sections <xref format="counter" target="transfer.codings"/> and
4254  <xref format="counter" target="message.length"/>)
4255</t>
4256<t>
4257  Clarification that the chunk length does not include
4258  the count of the octets in the chunk header and trailer.
4259  (<xref target="chunked.transfer.encoding"/>)
4260</t>
4261<t>
4262  Fix BNF to add query, as the abs_path production in
4263  Section 3 of <xref target="RFC2396"/> doesn't define it.
4264  (<xref target="request-uri"/>)
4265</t>
4266<t>
4267  Clarify exactly when close connection options must be sent.
4268  (<xref target="header.connection"/>)
4269</t>
4270</section>
4271</section>
4272
4273<section title="Change Log (to be removed by RFC Editor before publication)" anchor="change.log">
4274
4275<section title="Since RFC2616">
4276<t>
4277  Extracted relevant partitions from <xref target="RFC2616"/>.
4278</t>
4279</section>
4280
4281<section title="Since draft-ietf-httpbis-p1-messaging-00">
4282<t>
4283  Closed issues:
4284  <list style="symbols"> 
4285    <t>
4286      <eref target="http://www3.tools.ietf.org/wg/httpbis/trac/ticket/1"/>:
4287      "HTTP Version should be case sensitive"
4288      (<eref target="http://purl.org/NET/http-errata#verscase"/>)
4289    </t>
4290    <t>
4291      <eref target="http://www3.tools.ietf.org/wg/httpbis/trac/ticket/2"/>:
4292      "'unsafe' characters"
4293      (<eref target="http://purl.org/NET/http-errata#unsafe-uri"/>)
4294    </t>
4295    <t>
4296      <eref target="http://www3.tools.ietf.org/wg/httpbis/trac/ticket/3"/>:
4297      "Chunk Size Definition"
4298      (<eref target="http://purl.org/NET/http-errata#chunk-size"/>)
4299    </t>
4300    <t>
4301      <eref target="http://www3.tools.ietf.org/wg/httpbis/trac/ticket/4"/>:
4302      "Message Length"
4303      (<eref target="http://purl.org/NET/http-errata#msg-len-chars"/>)
4304    </t>
4305    <t>
4306      <eref target="http://www3.tools.ietf.org/wg/httpbis/trac/ticket/8"/>:
4307      "Media Type Registrations"
4308      (<eref target="http://purl.org/NET/http-errata#media-reg"/>)
4309    </t>
4310    <t>
4311      <eref target="http://www3.tools.ietf.org/wg/httpbis/trac/ticket/11"/>:
4312      "URI includes query"
4313      (<eref target="http://purl.org/NET/http-errata#uriquery"/>)
4314    </t>
4315    <t>
4316      <eref target="http://www3.tools.ietf.org/wg/httpbis/trac/ticket/15"/>:
4317      "No close on 1xx responses"
4318      (<eref target="http://purl.org/NET/http-errata#noclose1xx"/>)
4319    </t>
4320    <t>
4321      <eref target="http://www3.tools.ietf.org/wg/httpbis/trac/ticket/16"/>:
4322      "Remove 'identity' token references"
4323      (<eref target="http://purl.org/NET/http-errata#identity"/>)
4324    </t>
4325    <t>
4326      <eref target="http://www3.tools.ietf.org/wg/httpbis/trac/ticket/26"/>:
4327      "Import query BNF"
4328    </t>
4329    <t>
4330      <eref target="http://www3.tools.ietf.org/wg/httpbis/trac/ticket/31"/>:
4331      "qdtext BNF"
4332    </t>
4333    <t>
4334      <eref target="http://www3.tools.ietf.org/wg/httpbis/trac/ticket/35"/>:
4335      "Normative and Informative references"
4336    </t>
4337    <t>
4338      <eref target="http://www3.tools.ietf.org/wg/httpbis/trac/ticket/42"/>:
4339      "RFC2606 Compliance"
4340    </t>
4341    <t>
4342      <eref target="http://www3.tools.ietf.org/wg/httpbis/trac/ticket/45"/>:
4343      "RFC977 reference"
4344    </t>
4345    <t>
4346      <eref target="http://www3.tools.ietf.org/wg/httpbis/trac/ticket/46"/>:
4347      "RFC1700 references"
4348    </t>
4349    <t>
4350      <eref target="http://www3.tools.ietf.org/wg/httpbis/trac/ticket/47"/>:
4351      "inconsistency in date format explanation"
4352    </t>
4353    <t>
4354      <eref target="http://www3.tools.ietf.org/wg/httpbis/trac/ticket/48"/>:
4355      "Date reference typo"
4356    </t>
4357    <t>
4358      <eref target="http://www3.tools.ietf.org/wg/httpbis/trac/ticket/65"/>:
4359      "Informative references"
4360    </t>
4361    <t>
4362      <eref target="http://www3.tools.ietf.org/wg/httpbis/trac/ticket/66"/>:
4363      "ISO-8859-1 Reference"
4364    </t>
4365    <t>
4366      <eref target="http://www3.tools.ietf.org/wg/httpbis/trac/ticket/86"/>:
4367      "Normative up-to-date references"
4368    </t>
4369  </list>
4370</t>
4371<t>
4372  Other changes:
4373  <list style="symbols"> 
4374    <t>
4375      Update media type registrations to use RFC4288 template.
4376    </t>
4377    <t>
4378      Use names of RFC4234 core rules DQUOTE and HTAB,
4379      fix broken ABNF for chunk-data
4380      (work in progress on <eref target="http://www3.tools.ietf.org/wg/httpbis/trac/ticket/36"/>)
4381    </t>
4382  </list>
4383</t>
4384</section>
4385
4386<section title="Since draft-ietf-httpbis-p1-messaging-01">
4387<t>
4388  Closed issues:
4389  <list style="symbols"> 
4390    <t>
4391      <eref target="http://www3.tools.ietf.org/wg/httpbis/trac/ticket/19"/>:
4392      "Bodies on GET (and other) requests"
4393    </t>
4394    <t>
4395      <eref target="http://www3.tools.ietf.org/wg/httpbis/trac/ticket/55"/>:
4396      "Updating to RFC4288"
4397    </t>
4398    <t>
4399      <eref target="http://www3.tools.ietf.org/wg/httpbis/trac/ticket/57"/>:
4400      "Status Code and Reason Phrase"
4401    </t>
4402    <t>
4403      <eref target="http://www3.tools.ietf.org/wg/httpbis/trac/ticket/82"/>:
4404      "rel_path not used"
4405    </t>
4406  </list>
4407</t>
4408<t>
4409  Ongoing work on ABNF conversion (<eref target="http://www3.tools.ietf.org/wg/httpbis/trac/ticket/36"/>):
4410  <list style="symbols"> 
4411    <t>
4412      Get rid of duplicate BNF rule names ("host" -&gt; "uri-host", "trailer" -&gt;
4413      "trailer-part").
4414    </t>
4415    <t>
4416      Avoid underscore character in rule names ("http_URL" -&gt;
4417      "http-URL", "abs_path" -&gt; "path-absolute").
4418    </t>
4419    <t>
4420      Add rules for terms imported from URI spec ("absoluteURI", "authority",
4421      "path-absolute", "port", "query", "relativeURI", "host) -- these will
4422      have to be updated when switching over to RFC3986.
4423    </t>
4424    <t>
4425      Synchronize core rules with RFC5234 (this includes a change to CHAR
4426      which now excludes NUL).
4427    </t>
4428    <t>
4429      Get rid of prose rules that span multiple lines.
4430    </t>
4431    <t>
4432      Get rid of unused rules LOALPHA and UPALPHA.
4433    </t>
4434    <t>
4435      Move "Product Tokens" section (back) into Part 1, as "token" is used
4436      in the definition of the Upgrade header.
4437    </t>
4438    <t>
4439      Add explicit references to BNF syntax and rules imported from other parts of the specification.
4440    </t>
4441    <t>
4442      Rewrite prose rule "token" in terms of "tchar", rewrite prose rule "TEXT".
4443    </t>
4444  </list>
4445</t>
4446</section>
4447
4448<section title="Since draft-ietf-httpbis-p1-messaging-02" anchor="changes.since.02">
4449<t>
4450  Closed issues:
4451  <list style="symbols"> 
4452    <t>
4453      <eref target="http://www3.tools.ietf.org/wg/httpbis/trac/ticket/51"/>:
4454      "HTTP-date vs. rfc1123-date"
4455    </t>
4456    <t>
4457      <eref target="http://www3.tools.ietf.org/wg/httpbis/trac/ticket/64"/>:
4458      "WS in quoted-pair"
4459    </t>
4460  </list>
4461</t>
4462<t>
4463  Ongoing work on IANA Message Header Registration (<eref target="http://www3.tools.ietf.org/wg/httpbis/trac/ticket/40"/>):
4464  <list style="symbols"> 
4465    <t>
4466      Reference RFC 3984, and update header registrations for headers defined
4467      in this document.
4468    </t>
4469  </list>
4470</t>
4471<t>
4472  Ongoing work on ABNF conversion (<eref target="http://www3.tools.ietf.org/wg/httpbis/trac/ticket/36"/>):
4473  <list style="symbols"> 
4474    <t>
4475      Replace string literals when the string really is case-sensitive (HTTP-Version).
4476    </t>
4477  </list>
4478</t>
4479</section>
4480
4481</section>
4482
4483</back>
4484</rfc>
Note: See TracBrowser for help on using the repository browser.