idnits 2.17.1 

draft-hixie-thewebsocketprotocol-48.txt:

  Checking boilerplate required by RFC 5378 and the IETF Trust (see
  https://trustee.ietf.org/license-info):
  ----------------------------------------------------------------------------

  ** The document seems to lack a License Notice according IETF Trust
     Provisions of 28 Dec 2009, Section 6.b.ii or Provisions of 12 Sep 2009
     Section 6.b -- however, there's a paragraph with a matching beginning.
     Boilerplate error?

     (You're using the IETF Trust Provisions' Section 6.b License Notice from
     12 Feb 2009 rather than one of the newer Notices.  See
     https://trustee.ietf.org/license-info/.)


  Checking nits according to https://www.ietf.org/id-info/1id-guidelines.txt:
  ----------------------------------------------------------------------------

     No issues found here.

  Checking nits according to https://www.ietf.org/id-info/checklist :
  ----------------------------------------------------------------------------

  ** The abstract seems to contain references ([HTML5]), which it shouldn't. 
     Please replace those with straight textual mentions of the documents in
     question.


  Miscellaneous warnings:
  ----------------------------------------------------------------------------

  == The copyright year in the IETF Trust and authors Copyright Line does not
     match the current year

  == The document doesn't use any RFC 2119 keywords, yet has text resembling
     RFC 2119 boilerplate text.

  -- The document date (October 13, 2009) is 5299 days in the past.  Is this
     intentional?


  Checking references for intended status: Proposed Standard
  ----------------------------------------------------------------------------

     (See RFCs 3967 and 4897 for information about using normative references
     to lower-maturity documents in RFCs)

  -- Possible downref: Non-RFC (?) normative reference: ref. 'HTML5'

  -- Possible downref: Non-RFC (?) normative reference: ref. 'ORIGIN'

  ** Obsolete normative reference: RFC 2109 (Obsoleted by RFC 2965)

  ** Obsolete normative reference: RFC 2246 (Obsoleted by RFC 4346)

  ** Obsolete normative reference: RFC 2616 (Obsoleted by RFC 7230, RFC 7231,
     RFC 7232, RFC 7233, RFC 7234, RFC 7235)

  ** Obsolete normative reference: RFC 2965 (Obsoleted by RFC 6265)

  ** Obsolete normative reference: RFC 3490 (Obsoleted by RFC 5890, RFC 5891)

  -- Possible downref: Non-RFC (?) normative reference: ref. 'WEBADDRESSES'

  -- Possible downref: Non-RFC (?) normative reference: ref. 'WSAPI'


     Summary: 7 errors (**), 0 flaws (~~), 2 warnings (==), 5 comments (--).

     Run idnits with the --verbose option for more detailed information about
     the items above.

--------------------------------------------------------------------------------


2	Network Working Group                                         I. Hickson
3	Internet-Draft                                              Google, Inc.
4	Intended status: Standards Track                        October 13, 2009
5	Expires: April 16, 2010

7	                        The Web Socket protocol
8	                  draft-hixie-thewebsocketprotocol-48

10	Status of this Memo

12	   This Internet-Draft is submitted to IETF in full conformance with the
13	   provisions of BCP 78 and BCP 79.

15	   Internet-Drafts are working documents of the Internet Engineering
16	   Task Force (IETF), its areas, and its working groups.  Note that
17	   other groups may also distribute working documents as Internet-
18	   Drafts.

20	   Internet-Drafts are draft documents valid for a maximum of six months
21	   and may be updated, replaced, or obsoleted by other documents at any
22	   time.  It is inappropriate to use Internet-Drafts as reference
23	   material or to cite them other than as "work in progress."

25	   The list of current Internet-Drafts can be accessed at
26	   http://www.ietf.org/ietf/1id-abstracts.txt.

28	   The list of Internet-Draft Shadow Directories can be accessed at
29	   http://www.ietf.org/shadow.html.

31	   This Internet-Draft will expire on April 16, 2010.

33	Copyright Notice

35	   Copyright (c) 2009 IETF Trust and the persons identified as the
36	   document authors.  All rights reserved.

38	   This document is subject to BCP 78 and the IETF Trust's Legal
39	   Provisions Relating to IETF Documents in effect on the date of
40	   publication of this document (http://trustee.ietf.org/license-info).
41	   Please review these documents carefully, as they describe your rights
42	   and restrictions with respect to this document.

44	Abstract

46	   The Web Sockets protocol enables two-way communication between a user
47	   agent running untrusted code running in a controlled environment to a
48	   remote host that has opted-in to communications from that code.  The
49	   protocol consists of an initial handshake followed by basic message
50	   framing, layered over TCP.  The goal of this technology is to provide
51	   a mechanism for browser-based applications that need two-way
52	   communication with servers that does not rely on opening multiple
53	   HTTP connections (e.g. using XMLHttpRequest or <iframe>s and long
54	   polling).

56	Author's note

58	   This document is automatically generated from the same source
59	   document as the HTML5 specification.  [HTML5]

61	   Please send feedback to either the hybi@ietf.org list or the
62	   whatwg@whatwg.org list.

64	Table of Contents

66	   1.  Introduction . . . . . . . . . . . . . . . . . . . . . . . . .  4
67	     1.1.  Background . . . . . . . . . . . . . . . . . . . . . . . .  4
68	     1.2.  Protocol overview  . . . . . . . . . . . . . . . . . . . .  4
69	     1.3.  Design philosophy  . . . . . . . . . . . . . . . . . . . .  6
70	     1.4.  Security model . . . . . . . . . . . . . . . . . . . . . .  6
71	     1.5.  Relationship to TCP/IP and HTTP  . . . . . . . . . . . . .  6
72	     1.6.  Establishing a connection  . . . . . . . . . . . . . . . .  7
73	   2.  Conformance requirements . . . . . . . . . . . . . . . . . . .  8
74	     2.1.  Terminology  . . . . . . . . . . . . . . . . . . . . . . .  8
75	   3.  Web Socket URLs  . . . . . . . . . . . . . . . . . . . . . . .  9
76	     3.1.  Parsing Web Socket URLs  . . . . . . . . . . . . . . . . .  9
77	     3.2.  Constructing Web Socket URLs . . . . . . . . . . . . . . . 10
78	   4.  Client-side requirements . . . . . . . . . . . . . . . . . . . 11
79	     4.1.  Handshake  . . . . . . . . . . . . . . . . . . . . . . . . 11
80	     4.2.  Data framing . . . . . . . . . . . . . . . . . . . . . . . 19
81	     4.3.  Closing the connection . . . . . . . . . . . . . . . . . . 20
82	     4.4.  Handling errors in UTF-8 . . . . . . . . . . . . . . . . . 20
83	   5.  Server-side requirements . . . . . . . . . . . . . . . . . . . 22
84	     5.1.  Minimal handshake  . . . . . . . . . . . . . . . . . . . . 22
85	     5.2.  Handshake details  . . . . . . . . . . . . . . . . . . . . 23
86	     5.3.  Data framing . . . . . . . . . . . . . . . . . . . . . . . 24
87	   6.  Closing the connection . . . . . . . . . . . . . . . . . . . . 25
88	   7.  Security considerations  . . . . . . . . . . . . . . . . . . . 26
89	   8.  IANA considerations  . . . . . . . . . . . . . . . . . . . . . 27
90	     8.1.  Registration of ws: scheme . . . . . . . . . . . . . . . . 27
91	     8.2.  Registration of wss: scheme  . . . . . . . . . . . . . . . 28
92	     8.3.  Registration of the "WebSocket" HTTP Upgrade keyword . . . 29
93	     8.4.  WebSocket-Origin . . . . . . . . . . . . . . . . . . . . . 29
94	     8.5.  WebSocket-Protocol . . . . . . . . . . . . . . . . . . . . 30
95	     8.6.  WebSocket-Location . . . . . . . . . . . . . . . . . . . . 30
96	   9.  Using the Web Socket protocol from other specifications  . . . 32
97	   10. Normative References . . . . . . . . . . . . . . . . . . . . . 33
98	   Author's Address . . . . . . . . . . . . . . . . . . . . . . . . . 35

100	1.  Introduction

102	1.1.  Background

104	   _This section is non-normative._

106	   Historically, creating an instant messenger chat client as a Web
107	   application has required an abuse of HTTP to poll the server for
108	   updates while sending upstream notifications as distinct HTTP calls.

110	   This results in a variety of problems:

112	   o  The server is forced to use a number of different underlying TCP
113	      connections for each client: one for sending information to the
114	      client, and a new one for each incoming message.

116	   o  The wire protocol has a high overhead, with each client-to-server
117	      message having an HTTP header.

119	   o  The client-side script is forced to maintain a mapping from the
120	      outgoing connections to the incoming connection to track replies.

122	   A simpler solution would be to use a single TCP connection for
123	   traffic in both directions.  This is what the Web Socket protocol
124	   provides.  Combined with the Web Socket API, it provides an
125	   alternative to HTTP polling for two-way communication from a Web page
126	   to a remote server.  [WSAPI]

128	   The same technique can be used for a variety of Web applications:
129	   games, stock tickers, multiuser applications with simultaneous
130	   editing, user interfaces exposing server-side services in real time,
131	   etc.

133	1.2.  Protocol overview

135	   _This section is non-normative._

137	   The protocol has two parts: a handshake, and then the data transfer.

139	   The handshake from the client looks as follows:

141	        GET /demo HTTP/1.1
142	        Upgrade: WebSocket
143	        Connection: Upgrade
144	        Host: example.com
145	        Origin: http://example.com
146	        WebSocket-Protocol: sample

148	   The handshake from the server looks as follows:

150	        HTTP/1.1 101 Web Socket Protocol Handshake
151	        Upgrade: WebSocket
152	        Connection: Upgrade
153	        WebSocket-Origin: http://example.com
154	        WebSocket-Location: ws://example.com/demo
155	        WebSocket-Protocol: sample

157	   Once the client and server have both sent their handshakes, and if
158	   the handshake was successful, then the data transfer part starts.
159	   This is a two-way communication channel where each side can,
160	   independently from the other, send data at will.

162	   Data is sent in the form of UTF-8 text.  Each frame of data starts
163	   with a 0x00 byte and ends with a 0xFF byte, with the UTF-8 text in
164	   between.

166	   The Web Socket protocol uses this framing so that specifications that
167	   use the Web Socket protocol can expose such connections using an
168	   event-based mechanism instead of requiring users of those
169	   specifications to implement buffering and piecing together of
170	   messages manually.

172	   The protocol is designed to support other frame types in future.
173	   Instead of the 0x00 byte, other bytes might in future be defined.
174	   Frames denoted by bytes that do not have the high bit set (0x00 to
175	   0x7F) are treated as described above (a stream of bytes terminated by
176	   0xFF).  Frames denoted by bytes that have the high bit set (0x80 to
177	   0xFF) have a leading length indicator, which is encoded as a series
178	   of 7-bit bytes stored in octets with the 8th bit being set for all
179	   but the last byte.  The remainder of the frame is then as much data
180	   as was specified.

182	   The following diagrams summarise the protocol:

184	        Handshake
185	           |
186	          \|/
187	        Frame type byte <-------------------------------------.
188	           |      |                                           |
189	           |      `-- (0x00 .. 0x7F) --> Data... --> 0xFF -->-+
190	           |                                                  |
191	           `-- (0x80 .. 0xFF) --> Length --> Data... ------->-'

193	1.3.  Design philosophy

195	   _This section is non-normative._

197	   The Web Socket protocol is designed on the principle that there
198	   should be minimal framing (the only framing that exists is to make
199	   the protocol frame-based instead of stream-based, and to support a
200	   distinction between Unicode text and binary frames).  It is expected
201	   that metadata would be layered on top of Web Socket by the
202	   application layer, in the same way that metadata is layered on top of
203	   TCP/IP by the application layer (HTTP).

205	   Conceptually, Web Socket is really just a layer on top of TCP/IP that
206	   adds a Web "origin"-based security model for browsers; adds an
207	   addressing and protocol naming mechanism to support multiple services
208	   on one port and multiple host names on one IP address; and layers a
209	   framing mechanism on top of TCP to get back to the IP packet
210	   mechanism that TCP is built on, but without length limits.  Other
211	   than that, it adds nothing.  Basically it is intended to be as close
212	   as possible to just exposing raw TCP/IP to script as possible given
213	   the constraints of the Web. It's also designed in such a way that its
214	   servers can share a port with HTTP servers, by having its handshake
215	   be a valid HTTP Upgrade handshake also.

217	1.4.  Security model

219	   _This section is non-normative._

221	   The Web Socket protocol uses the origin model used by Web browsers to
222	   restrict which Web pages can contact a Web Socket server when the Web
223	   Socket protocol is used from a Web page.  Naturally, when the Web
224	   Socket protocol is used directly (not from a Web page), the origin
225	   model is not useful, as the client can provide any arbitrary origin
226	   string.

228	   This protocol is intended to fail to establish a connection with
229	   servers of pre-existing protocols like SMTP or HTTP, while allowing
230	   HTTP servers to opt-in to supporting this protocol if desired.  This
231	   is achieved by having a strict and elaborate handshake, and by
232	   limiting the data that can be inserted into the connection before the
233	   handshake is finished (thus limiting how much the server can be
234	   influenced).

236	1.5.  Relationship to TCP/IP and HTTP

238	   _This section is non-normative._

240	   The Web Socket protocol is an independent TCP-based protocol.  Its
241	   only relationship to HTTP is that its handshake is interpreted by
242	   HTTP servers as an Upgrade request.

244	   Based on the expert recommendation of the IANA, the Web Socket
245	   protocol by default uses port 80 for regular Web Socket connections
246	   and port 443 for Web Socket connections tunneled over TLS.

248	1.6.  Establishing a connection

250	   _This section is non-normative._

252	   There are several options for establishing a Web Socket connection.

254	   The simplest method is to use port 80 to get a direct connection to a
255	   Web Socket server.  Port 80 traffic, however, will often be
256	   intercepted by HTTP proxies, which can lead to the connection failing
257	   to be established.

259	   The second simplest method is to use TLS encryption and port 443 to
260	   connect directly to a Web Socket server.  This has the advantage of
261	   being more secure; however, TLS encryption can be computationally
262	   expensive.

264	   When a connection is to be made to a port that is shared by an HTTP
265	   server (a situation that is quite likely to occur with traffic to
266	   ports 80 and 443), the connection will appear to the HTTP server to
267	   be a regular GET request with an Upgrade offer.  In relatively simple
268	   setups with just one IP address and a single server for all traffic
269	   to a single hostname, this might allow a practical way for systems
270	   based on the Web Socket protocol to be deployed.  In more elaborate
271	   setups (e.g. with load balancers and multiple servers), a dedicated
272	   set of hosts for Web Socket connections separate from the HTTP
273	   servers is probably easier to manage.

275	2.  Conformance requirements

277	   All diagrams, examples, and notes in this specification are non-
278	   normative, as are all sections explicitly marked non-normative.
279	   Everything else in this specification is normative.

281	   The key words "MUST", "MUST NOT", "REQUIRED", "SHOULD", "SHOULD NOT",
282	   "RECOMMENDED", "MAY", and "OPTIONAL" in the normative parts of this
283	   document are to be interpreted as described in RFC2119.  For
284	   readability, these words do not appear in all uppercase letters in
285	   this specification.  [RFC2119]

287	   Requirements phrased in the imperative as part of algorithms (such as
288	   "strip any leading space characters" or "return false and abort these
289	   steps") are to be interpreted with the meaning of the key word
290	   ("must", "should", "may", etc) used in introducing the algorithm.

292	   Conformance requirements phrased as algorithms or specific steps may
293	   be implemented in any manner, so long as the end result is
294	   equivalent.  (In particular, the algorithms defined in this
295	   specification are intended to be easy to follow, and not intended to
296	   be performant.)

298	   Implementations may impose implementation-specific limits on
299	   otherwise unconstrained inputs, e.g. to prevent denial of service
300	   attacks, to guard against running out of memory, or to work around
301	   platform-specific limitations.

303	   The conformance classes defined by this specification are user agents
304	   and servers.

306	2.1.  Terminology

308	   *Converting a string to ASCII lowercase* means replacing all
309	   characters in the range U+0041 ..  U+005A (i.e.  LATIN CAPITAL LETTER
310	   A to LATIN CAPITAL LETTER Z) with the corresponding characters in the
311	   range U+0061 ..  U+007A (i.e.  LATIN SMALL LETTER A to LATIN SMALL
312	   LETTER Z).

314	   The term "URL" is used in this section in a manner consistent with
315	   the terminology used in HTML, namely, to denote a string that might
316	   or might not be a valid URI or IRI and to which certain error
317	   handling behaviors will be applied when the string is parsed.
318	   [HTML5]

320	3.  Web Socket URLs

322	3.1.  Parsing Web Socket URLs

324	   The steps to *parse a Web Socket URL's components* from a string
325	   /url/ are as follows.  These steps return either a /host/, a /port/,
326	   a /resource name/, and a /secure/ flag, or they fail.

328	   1.   If /protocol/ is specified but is either the empty string or
329	        contains characters that are not in the range U+0021 ..  U+007E,
330	        then fail this algorithm.

332	   2.   If the /url/ string is not an absolute URL, then fail this
333	        algorithm.  [WEBADDRESSES]

335	   3.   Resolve the /url/ string using the resolve a Web address
336	        algorithm defined by the Web addresses specification, with the
337	        URL character encoding set to UTF-8.  [WEBADDRESSES] [RFC3629]

339	        NOTE: It doesn't matter what it is resolved relative to, since
340	        we already know it is an absolute URL at this point.

342	   4.   If /url/ does not have a <scheme> component whose value is
343	        either "ws" or "wss", when compared in an ASCII case-insensitive
344	        manner, then fail this algorithm.

346	   5.   If the <scheme> component of /url/ is "ws", set /secure/ to
347	        false; otherwise, the <scheme> component is "wss", set /secure/
348	        to true.

350	   6.   Let /host/ be the value of the <host> component of /url/,
351	        converted to ASCII lowercase.

353	   7.   If /url/ has a <port> component, then let /port/ be that
354	        component's value; otherwise, there is no explicit /port/.

356	   8.   If there is no explicit /port/, then: if /secure/ is false, let
357	        /port/ be 80, otherwise let /port/ be 443.

359	   9.   Let /resource name/ be the value of the <path> component (which
360	        might be empty) of /url/.

362	   10.  If /resource name/ is the empty string, set it to a single
363	        character U+002F SOLIDUS (/).

365	   11.  If /url/ has a <query> component, then append a single U+003F
366	        QUESTION MARK character (?) to /resource name/, followed by the
367	        value of the <query> component.

369	   12.  Return /host/, /port/, /resource name/, and /secure/.

371	3.2.  Constructing Web Socket URLs

373	   The steps to *construct a Web Socket URL* from a /host/, a /port/, a
374	   /resource name/, and a /secure/ flag, are as follows:

376	   1.  Let /url/ be the empty string.

378	   2.  If the /secure/ flag is false, then append the string "ws://" to
379	       /url/.  Otherwise, append the string "wss://" to /url/.

381	   3.  Append /host/ to /url/.

383	   4.  If the /secure/ flag is false and port is not 80, or if the
384	       /secure/ flag is true and port is not 443, then append the string
385	       ":" followed by /port/ to /url/.

387	   5.  Append /resource name/ to /url/.

389	   6.  Return /url/.

391	4.  Client-side requirements

393	   _This section only applies to user agents, not to servers._

395	   NOTE: This specification doesn't currently define a limit to the
396	   number of simultaneous connections that a client can establish to a
397	   server.

399	4.1.  Handshake

401	   When the user agent is to *establish a Web Socket connection* to a
402	   host /host/, on a port /port/, from an origin whose ASCII
403	   serialization is /origin/, with a flag /secure/, with a string giving
404	   a /resource name/, and optionally with a string giving a /protocol/,
405	   it must run the following steps.  The /resource name/ string must
406	   start with a U+002F SOLIDUS character (/).  [ORIGIN]

408	   1.   If the user agent already has a Web Socket connection to the
409	        remote host (IP address) identified by /host/, even if known by
410	        another name, wait until that connection has been established or
411	        for that connection to have failed.

413	        NOTE: This makes it harder for a script to perform a denial of
414	        service attack by just opening a large number of Web Socket
415	        connections to a remote host.

417	        NOTE: There is no limit to the number of established Web Socket
418	        connections a user agent can have with a single remote host.
419	        Servers can refuse to connect users with an excessive number of
420	        connections, or disconnect resource-hogging users when suffering
421	        high load.

423	   2.   _Connect_: If the user agent is configured to use a proxy when
424	        using the Web Socket protocol to connect to host /host/ and/or
425	        port /port/, then connect to that proxy and ask it to open a
426	        TCP/IP connection to the host given by /host/ and the port given
427	        by /port/.

429	           EXAMPLE: For example, if the user agent uses an HTTP proxy
430	           for all traffic, then if it was to try to connect to port 80
431	           on server example.com, it might send the following lines to
432	           the proxy server:

434	              CONNECT example.com:80 HTTP/1.1
435	              Host: example.com

437	           If there was a password, the connection might look like:

439	              CONNECT example.com:80 HTTP/1.1
440	              Host: example.com
441	              Proxy-authorization: Basic ZWRuYW1vZGU6bm9jYXBlcyE=

443	        Otherwise, if the user agent is not configured to use a proxy,
444	        then open a TCP/IP connection to the host given by /host/ and
445	        the port given by /port/.

447	        NOTE: Implementations that do not expose explicit UI for
448	        selecting a proxy for Web Socket connections separate from other
449	        proxies are encouraged to use a SOCKS proxy for Web Socket
450	        connections, if available, or failing that, to prefer the proxy
451	        configured for HTTPS connections over the proxy configured for
452	        HTTP connections.

454	        For the purpose of proxy autoconfiguration scripts, the URL to
455	        pass the function must be constructed from /host/, /port/,
456	        /resource name/, and the /secure/ flag using the steps to
457	        construct a Web Socket URL.

459	        NOTE: The WebSocket protocol can be identified in proxy
460	        autoconfiguration scripts from the scheme ("ws:" for unencrypted
461	        connections and "wss:" for encrypted connections).

463	   3.   If the connection could not be opened, then fail the Web Socket
464	        connection and abort these steps.

466	   4.   If /secure/ is true, perform a TLS handshake over the
467	        connection.  If this fails (e.g. the server's certificate could
468	        not be verified), then fail the Web Socket connection and abort
469	        these steps.  Otherwise, all further communication on this
470	        channel must run through the encrypted tunnel.  [RFC2246]

472	   5.   Send the following bytes to the remote side (the server):

474	           47 45 54 20

476	        Send the /resource name/ value, encoded as US-ASCII.

478	        Send the following bytes:

480	           20 48 54 54 50 2F 31 2E  31 0D 0A 55 70 67 72 61
481	           64 65 3A 20 57 65 62 53  6F 63 6B 65 74 0D 0A 43
482	           6F 6E 6E 65 63 74 69 6F  6E 3A 20 55 70 67 72 61
483	           64 65 0D 0A

485	        NOTE: The string "GET ", the path, " HTTP/1.1", CRLF, the string
486	        "Upgrade: WebSocket", CRLF, and the string "Connection:

488	         Upgrade", CRLF.

490	   6.   Send the following bytes:

492	           48 6F 73 74 3A 20

494	        Send the /host/ value, converted to ASCII lowercase, and encoded
495	        as US-ASCII.

497	        If /secure/ is false, and /port/ is not 80, or if /secure/ is
498	        true, and /port/ is not 443, then send an 0x3A byte (ASCII :)
499	        followed by the value of /port/, expressed as a base-ten
500	        integer, encoded as US-ASCII.

502	        Send the following bytes:

504	           0D 0A

506	        NOTE: The string "Host: ", the host, and CRLF.

508	   7.   Send the following bytes:

510	           4F 72 69 67 69 6E 3A 20

512	        Send the /origin/ value, converted to ASCII lowercase, encoded
513	        as US-ASCII.  [ORIGIN]

515	        NOTE: The /origin/ value is a string that was passed to this
516	        algorithm.

518	        Send the following bytes:

520	           0D 0A

522	        NOTE: The string "Origin: ", the origin, and CRLF.

524	   8.   If there is no /protocol/, then skip this step.

526	        Otherwise, send the following bytes:

528	           57 65 62 53 6F 63 6B 65  74 2D 50 72 6F 74 6F 63
529	           6F 6C 3A 20

531	        Send the /protocol/ value, encoded as US-ASCII.

533	        Send the following bytes:

535	           0d 0a

537	        NOTE: The string "WebSocket-Protocol: ", the protocol, and CRLF.

539	   9.   If the client has any authentication information or cookies that
540	        would be relevant to a resource accessed over HTTP, if /secure/
541	        is false, or HTTPS, if it is true, on host /host/, port /port/,
542	        with /resource name/ as the path (and possibly query
543	        parameters), then HTTP headers that would be appropriate for
544	        that information should be sent at this point.  [RFC2616]
545	        [RFC2109] [RFC2965]

547	        Each header must be on a line of its own (each ending with a CR
548	        LF sequence).  For the purposes of this step, each header must
549	        not be split into multiple lines (despite HTTP otherwise
550	        allowing this with continuation lines).

552	           EXAMPLE: For example, if the server had a username and
553	           password that applied to |http://example.com/socket|, and the
554	           Web Socket was being opened to |ws://example.com/socket|, it
555	           could send them:

557	              Authorization: Basic d2FsbGU6ZXZl

559	   10.  Send the following bytes:

561	           0d 0a

563	        NOTE: Just a CRLF (a blank line).

565	   11.  Read bytes from the server until either the connection closes,
566	        or a 0x0A byte is read.  Let /header/ be these bytes, including
567	        the 0x0A byte.

569	        If /header/ is not at least two bytes long, or if the last two
570	        bytes aren't 0x0D and 0x0A respectively, then fail the Web
571	        Socket connection and abort these steps.

573	        User agents may apply a timeout to this step, failing the Web
574	        Socket connection if the server does not send back data in a
575	        suitable time period.

577	   12.  If /header/ consists of 44 bytes that exactly match the
578	        following, then let /mode/ be _normal_.

580	           48 54 54 50 2F 31 2E 31  20 31 30 31 20 57 65 62
581	           20 53 6F 63 6B 65 74 20  50 72 6F 74 6F 63 6F 6C
582	           20 48 61 6E 64 73 68 61  6B 65 0D 0A

584	        NOTE: The string "HTTP/1.1 101 Web Socket Protocol Handshake"
585	        followed by a CRLF pair.

587	        Otherwise, let /code/ be the substring of /header/ that starts
588	        from the byte after the first 0x20 byte, and ends with the byte
589	        before the second 0x20 byte.  If there are not at least two 0x20
590	        bytes in /header/, then fail the Web Socket connection and abort
591	        these steps.

593	        If /code/, interpreted as ASCII, is "401", then let /mode/ be
594	        _authenticate_.

596	        Otherwise, fail the Web Socket connection and abort these steps.

598	   13.  If /mode/ is _normal_, then read 41 bytes from the server.

600	        If the connection closes before 41 bytes are received, or if the
601	        41 bytes aren't exactly equal to the following bytes, then fail
602	        the Web Socket connection and abort these steps.

604	           55 70 67 72 61 64 65 3A  20 57 65 62 53 6F 63 6B
605	           65 74 0D 0A 43 6F 6E 6E  65 63 74 69 6F 6E 3A 20
606	           55 70 67 72 61 64 65 0D  0A

608	        NOTE: The string "Upgrade: WebSocket", CRLF, the string
609	        "Connection: Upgrade", CRLF.

611	        User agents may apply a timeout to this step, failing the Web
612	        Socket connection if the server does not respond with the above
613	        bytes within a suitable time period.

615	        NOTE: This step is skipped if /mode/ is _authenticate_.

617	   14.  Let /headers/ be a list of name-value pairs, initially empty.

619	   15.  _Header_: Let /name/ and /value/ be empty byte arrays.

621	   16.  Read a byte from the server.

623	        If the connection closes before this byte is received, then fail
624	        the Web Socket connection and abort these steps.

626	        Otherwise, handle the byte as described in the appropriate entry
627	        below:

629	        -> If the byte is 0x0D (ASCII CR)
630	           If the /name/ byte array is empty, then jump to the headers
631	           processing step.  Otherwise, fail the Web Socket connection
632	           and abort these steps.

634	        -> If the byte is 0x0A (ASCII LF)
635	           Fail the Web Socket connection and abort these steps.

637	        -> If the byte is 0x3A (ASCII :)
638	           Move on to the next step.

640	        -> If the byte is in the range 0x41 .. 0x5A (ASCII A .. Z)
641	           Append a byte whose value is the byte's value plus 0x20 to
642	           the /name/ byte array and redo this step for the next byte.

644	        -> Otherwise
645	           Append the byte to the /name/ byte array and redo this step
646	           for the next byte.

648	        NOTE: This reads a header name, terminated by a colon,
649	        converting upper-case ASCII letters to lowercase, and aborting
650	        if a stray CR or LF is found.

652	   17.  Read a byte from the server.

654	        If the connection closes before this byte is received, then fail
655	        the Web Socket connection and abort these steps.

657	        Otherwise, handle the byte as described in the appropriate entry
658	        below:

660	        -> If the byte is 0x20 (ASCII space)
661	           Ignore the byte and move on to the next step.

663	        -> Otherwise
664	           Treat the byte as described by the list in the next step,
665	           then move on to that next step for real.

667	        NOTE: This skips past a space character after the colon, if
668	        necessary.

670	   18.  Read a byte from the server.

672	        If the connection closes before this byte is received, then fail
673	        the Web Socket connection and abort these steps.

675	        Otherwise, handle the byte as described in the appropriate entry
676	        below:

678	        -> If the byte is 0x0D (ASCII CR)
679	           Move on to the next step.

681	        -> If the byte is 0x0A (ASCII LF)
682	           Fail the Web Socket connection and abort these steps.

684	        -> Otherwise
685	           Append the byte to the /value/ byte array and redo this step
686	           for the next byte.

688	        NOTE: This reads a header value, terminated by a CRLF.

690	   19.  Read a byte from the server.

692	        If the connection closes before this byte is received, or if the
693	        byte is not a 0x0A byte (ASCII LF), then fail the Web Socket
694	        connection and abort these steps.

696	        NOTE: This skips past the LF byte of the CRLF after the header.

698	   20.  Append an entry to the /headers/ list that has the name given by
699	        the string obtained by interpreting the /name/ byte array as a
700	        UTF-8 byte stream and the value given by the string obtained by
701	        interpreting the /value/ byte array as a UTF-8 byte stream.

703	   21.  Return to the "Header" step above.

705	   22.  _Headers processing_: Read a byte from the server.

707	        If the connection closes before this byte is received, or if the
708	        byte is not a 0x0A byte (ASCII LF), then fail the Web Socket
709	        connection and abort these steps.

711	        NOTE: This skips past the LF byte of the CRLF after the blank
712	        line after the headers.

714	   23.  If /mode/ is _normal_, then: If there is not exactly one entry
715	        in the /headers/ list whose name is "websocket-origin", or if
716	        there is not exactly one entry in the /headers/ list whose name
717	        is "websocket-location", or if the /protocol/ was specified but
718	        there is not exactly one entry in the /headers/ list whose name
719	        is "websocket-protocol", or if there are any entries in the
720	        /headers/ list whose names are the empty string, then fail the
721	        Web Socket connection and abort these steps.  Otherwise, handle
722	        each entry in the /headers/ list as follows:

724	        -> If the entry's name is "websocket-origin"
725	           If the value is not exactly equal to /origin/, converted to
726	           ASCII lowercase, then fail the Web Socket connection and
727	           abort these steps.  [ORIGIN]

729	        -> If the entry's name is "websocket-location"
730	           If the value is not exactly equal to a string obtained from
731	           the steps to construct a Web Socket URL from /host/, /port/,
732	           /resource name/, and the /secure/ flag, then fail the Web
733	           Socket connection and abort these steps.

735	        -> If the entry's name is "websocket-protocol"
736	           If there was a /protocol/ specified, and the value is not
737	           exactly equal to /protocol/, then fail the Web Socket
738	           connection and abort these steps.  (If no /protocol/ was
739	           specified, the header is ignored.)

741	        -> If the entry's name is "set-cookie" or "set-cookie2" or
742	        another cookie-related header name
743	           Handle the cookie as defined by the appropriate
744	           specification, with the resource being the one with the host
745	           /host/, the port /port/, the path (and possibly query
746	           parameters) /resource name/, and the scheme |http| if
747	           /secure/ is false and |https| if /secure/ is true.  [RFC2109]
748	           [RFC2965]

750	        -> Any other name
751	           Ignore it.

753	        If /mode/ is _authenticate_, then: If there is not exactly one
754	        entry in the /headers/ list whose name is "www-authenticate",
755	        then fail the Web Socket connection and abort these steps.
756	        Otherwise, handle each entry in the /headers/ list as follows:

758	        -> If the entry's name is "www-authenticate"
759	           Obtain credentials in a manner consistent with the
760	           requirements for handling the |WWW-Authenticate| header in
761	           HTTP, and then close the connection (if the server has not
762	           already done so) and jump back to the step labeled _connect_,
763	           including the relevant authentication headers in the new
764	           request.  [RFC2616]

766	        -> Any other name
767	           Ignore it.

769	   24.  The *Web Socket connection is established*.  Now the user agent
770	        must send and receive to and from the connection as described in
771	        the next section.

773	4.2.  Data framing

775	   Once a Web Socket connection is established, the user agent must run
776	   through the following state machine for the bytes sent by the server.

778	   1.  Try to read a byte from the server.  Let /frame type/ be that
779	       byte.

781	       If no byte could be read because the Web Socket connection is
782	       closed, then abort.

784	   2.  Handle the /frame type/ byte as follows:

786	       If the high-order bit of the /frame type/ byte is set (i.e. if
787	       /frame type/ _and_ed with 0x80 returns 0x80)
788	          Run these steps.  If at any point during these steps a read is
789	          attempted but fails because the Web Socket connection is
790	          closed, then abort.

792	          1.  Let /length/ be zero.

794	          2.  _Length_: Read a byte, let /b/ be that byte.

796	          3.  Let /b_v/ be integer corresponding to the low 7 bits of
797	              /b/ (the value you would get by _and_ing /b/ with 0x7F).

799	          4.  Multiply /length/ by 128, add /b_v/ to that result, and
800	              store the final result in /length/.

802	          5.  If the high-order bit of /b/ is set (i.e. if /b/ _and_ed
803	              with 0x80 returns 0x80), then return to the step above
804	              labeled _length_.

806	          6.  Read /length/ bytes.

808	          7.  Discard the read bytes.

810	       If the high-order bit of the /frame type/ byte is _not_ set (i.e.
811	       if /frame type/ _and_ed with 0x80 returns 0x00)
812	          Run these steps.  If at any point during these steps a read is
813	          attempted but fails because the Web Socket connection is
814	          closed, then abort.

816	          1.  Let /raw data/ be an empty byte array.

818	          2.  _Data_: Read a byte, let /b/ be that byte.  If the client
819	              runs out of resources for buffering the incoming data, or
820	              hits an artificial resource limit intended to avoid
821	              resource starvation, then it must fail the Web Socket
822	              connection and abort these steps.

824	          3.  If /b/ is not 0xFF, then append /b/ to /raw data/ and
825	              return to the previous step (labeled _data_).

827	          4.  Interpret /raw data/ as a UTF-8 string, and store that
828	              string in /data/.

830	          5.  If /frame type/ is 0x00, then *a message has been
831	              received* with text /data/.  Otherwise, discard the data.

833	   3.  Return to the first step to read the next byte.

835	   If the user agent is faced with content that is too large to be
836	   handled appropriately, then it must fail the Web Socket connection.

838	   Once a Web Socket connection is established, the user agent must use
839	   the following steps to *send /data/ using the Web Socket*:

841	   1.  Send a 0x00 byte to the server.

843	   2.  Encode /data/ using UTF-8 and send the resulting byte stream to
844	       the server.

846	   3.  Send a 0xFF byte to the server.

848	   If at any point there is a fatal problem with sending data to the
849	   server, the user agent must fail the Web Socket connection.

851	4.3.  Closing the connection

853	   To *fail the Web Socket connection*, the user agent must close the
854	   Web Socket connection, and may report the problem to the user (which
855	   would be especially useful for developers).  However, user agents
856	   must not convey the failure information to the script that attempted
857	   the connection in a way distinguishable from the Web Socket being
858	   closed normally.

860	   Except as indicated above or as specified by the application layer
861	   (e.g. a script using the Web Socket API), user agents should not
862	   close the connection.

864	4.4.  Handling errors in UTF-8

866	   When a client is to interpret a byte stream as UTF-8 but finds that
867	   the byte stream is not in fact a valid UTF-8 stream, then any bytes
868	   or sequences of bytes that are not valid UTF-8 sequences must be
869	   interpreted as a U+FFFD REPLACEMENT CHARACTER.

871	5.  Server-side requirements

873	   _This section only applies to servers._

875	5.1.  Minimal handshake

877	   NOTE: This section describes the minimal requirements for a server-
878	   side implementation of Web Sockets.

880	   Listen on a port for TCP/IP.  Upon receiving a connection request,
881	   open a connection and send the following bytes back to the client:

883	        48 54 54 50 2F 31 2E 31  20 31 30 31 20 57 65 62
884	        20 53 6F 63 6B 65 74 20  50 72 6F 74 6F 63 6F 6C
885	        20 48 61 6E 64 73 68 61  6B 65 0D 0A 55 70 67 72
886	        61 64 65 3A 20 57 65 62  53 6F 63 6B 65 74 0D 0A
887	        43 6F 6E 6E 65 63 74 69  6F 6E 3A 20 55 70 67 72
888	        61 64 65 0D 0A

890	   Send the string "WebSocket-Origin" followed by a U+003A COLON (:) and
891	   a U+0020 SPACE, followed by the ASCII serialization of the origin
892	   from which the server is willing to accept connections, followed by a
893	   CRLF pair (0x0D 0x0A).  [ORIGIN]

895	      For instance:

897	           WebSocket-Origin: http://example.com

899	   Send the string "WebSocket-Location" followed by a U+003A COLON (:)
900	   and a U+0020 SPACE, followed by the URL of the Web Socket script,
901	   followed by a CRLF pair (0x0D 0x0A).

903	      For instance:

905	           WebSocket-Location: ws://example.com/demo

907	   NOTE: Do not include the port if it is the default port for Web
908	   Socket protocol connections of the type in question (80 for
909	   unencrypted connections and 443 for encrypted connections).

911	   Send another CRLF pair (0x0D 0x0A).

913	   Read data from the client until four bytes 0x0D 0x0A 0x0D 0x0A are
914	   read.  This data must either be discarded or handled as described in
915	   the following section describing the handshake details.

917	   If the connection isn't dropped at this point, go to the data framing
918	   section.

920	5.2.  Handshake details

922	   The previous section ignores the data that is transmitted by the
923	   client during the handshake.

925	   The data sent by the client consists of a number of fields separated
926	   by CR LF pairs (bytes 0x0D 0x0A).

928	   The first field consists of three tokens separated by space
929	   characters (byte 0x20).  The middle token is the path being opened.
930	   If the server supports multiple paths, then the server should echo
931	   the value of this field in the initial handshake, as part of the URL
932	   given on the |WebSocket-Location| line (after the appropriate scheme
933	   and host).

935	   If the first field does not have three tokens, the server should
936	   abort the connection as it probably represents an errorneous client.

938	   The remaining fields consist of name-value pairs, with the name part
939	   separated from the value part by a colon and a space (bytes 0x3A
940	   0x20).  Of these, several are interesting:

942	   Host (bytes 48 6F 73 74)
943	      The value gives the hostname that the client intended to use when
944	      opening the Web Socket.  It would be of interest in particular to
945	      virtual hosting environments, where one server might serve
946	      multiple hosts, and might therefore want to return different data.

948	      The right host has to be output as part of the URL given on the
949	      |WebSocket-Location| line of the handshake described above, to
950	      verify that the server knows that it is really representing that
951	      host.

953	   Origin (bytes 4F 72 69 67 69 6E)
954	      The value gives the scheme, hostname, and port (if it's not the
955	      default port for the given scheme) of the page that asked the
956	      client to open the Web Socket.  It would be interesting if the
957	      server's operator had deals with operators of other sites, since
958	      the server could then decide how to respond (or indeed, _whether_
959	      to respond) based on which site was requesting a connection.

961	      If the server supports connections from more than one origin, then
962	      the server should echo the value of this field in the initial
963	      handshake, on the |WebSocket-Origin| line.

965	   Other fields
966	      Other fields can be used, such as "Cookie" or "Authorization", for
967	      authentication purposes.

969	   Any fields that lack the colon-space separator should be discarded
970	   and may cause the server to disconnect.

972	5.3.  Data framing

974	   NOTE: This section only describes how to handle content that this
975	   specification allows user agents to send (text).  It doesn't handle
976	   any arbitrary content in the same way that the requirements on user
977	   agents defined earlier handle any content including possible future
978	   extensions to the protocols.

980	   The server must run through the following steps to process the bytes
981	   sent by the client:

983	   1.  Read a byte from the client.  Assuming everything is going
984	       according to plan, it will be a 0x00 byte.  If the byte is not a
985	       0x00 byte, then the server may disconnect.

987	   2.  Let /raw data/ be an empty byte array.

989	   3.  _Data_: Read a byte, let /b/ be that byte.

991	   4.  If /b/ is not 0xFF, then append /b/ to /raw data/ and return to
992	       the previous step (labeled _data_).

994	   5.  Interpret /raw data/ as a UTF-8 string, and apply whatever
995	       server-specific processing is to occur for the resulting string.

997	   6.  Return to the first step to read the next byte.

999	   The server must run through the following steps to send strings to
1000	   the client:

1002	   1.  Send a 0x00 byte to the client to indicate the start of a string.

1004	   2.  Encode /data/ using UTF-8 and send the resulting byte stream to
1005	       the client.

1007	   3.  Send a 0xFF byte to the client to indicate the end of the
1008	       message.

1010	6.  Closing the connection

1012	   To *close the Web Socket connection*, either the user agent or the
1013	   server closes the TCP/IP connection.  There is no closing handshake.
1014	   Whether the user agent or the server closes the connection, it is
1015	   said that the *Web Socket connection is closed*.

1017	   When a user agent is to close the Web Socket connection, it must drop
1018	   all subsequent data from the server and must act as if the server had
1019	   immediately closed its side of the connection.

1021	   When a user agent notices that the Web Socket connection is closed,
1022	   it must immediately close its side of the connection.

1024	   Servers may close the Web Socket connection whenever desired.

1026	   User agents should not close the Web Socket connection arbitrarily.

1028	7.  Security considerations

1030	   While this protocol is intended to be used by scripts in Web pages,
1031	   it can also be used directly by hosts.  Such hosts are acting on
1032	   their own behalf, and can therefore send fake "Origin" headers,
1033	   misleading the server.  Servers should therefore be careful about
1034	   assuming that they are talking directly to scripts from known
1035	   origins, and must consider that they might be accessed in unexpected
1036	   ways.  In particular, a server should not trust that any input is
1037	   valid.

1039	   EXAMPLE: For example, if the server uses input as part of SQL
1040	   queries, all input text should be escaped before being passed to the
1041	   SQL server, lest the server be susceptible to SQL injection.

1043	   Servers that are not intended to process input from any Web page but
1044	   only for certain sites should verify the "Origin" header is an origin
1045	   they expect, and should only respond with the corresponding
1046	   "WebSocket-Origin" if it is an accepted origin.  Servers that only
1047	   accept input from one origin can just send back that value in the
1048	   "WebSocket-Origin" header, without bothering to check the client's
1049	   value.

1051	   If at any time a server is faced with data that it does not
1052	   understand, or that violates some criteria by which the server
1053	   determines safety of input, or when the server sees a handshake that
1054	   does not correspond to the values the server is expecting (e.g.
1055	   incorrect path or origin), the server should just disconnect.  It is
1056	   always safe to disconnect.

1058	8.  IANA considerations

1060	8.1.  Registration of ws: scheme

1062	   A |ws:| URL identifies a Web Socket server and resource name.

1064	   URI scheme name.
1065	      ws

1067	   Status.
1068	      Permanent.

1070	   URI scheme syntax.
1071	      In ABNF terms using the terminals from the URI specifications:
1072	      [RFC5234] [RFC3986]

1074	           "ws" ":" hier-part [ "?" query ]

1076	      The path and query components form the resource name sent to the
1077	      server to identify the kind of service desired.  Other components
1078	      have the meanings described in RFC3986.

1080	   URI scheme semantics.
1081	      The only operation for this scheme is to open a connection using
1082	      the Web Socket protocol.

1084	   Encoding considerations.
1085	      Characters in the host component that are excluded by the syntax
1086	      defined above must be converted from Unicode to ASCII by applying
1087	      the IDNA ToASCII algorithm to the Unicode host name, with both the
1088	      AllowUnassigned and UseSTD3ASCIIRules flags set, and using the
1089	      result of this algorithm as the host in the URI.  [RFC3490]

1091	      Characters in other components that are excluded by the syntax
1092	      defined above must be converted from Unicode to ASCII by first
1093	      encoding the characters as UTF-8 and then replacing the
1094	      corresponding bytes using their percent-encoded form as defined in
1095	      the URI and IRI specification.  [RFC3986] [RFC3987]

1097	   Applications/protocols that use this URI scheme name.
1098	      Web Socket protocol.

1100	   Interoperability considerations.
1101	      None.

1103	   Security considerations.
1104	      See "Security considerations" section above.

1106	   Contact.
1107	      Ian Hickson <ian@hixie.ch>

1109	   Author/Change controller.
1110	      Ian Hickson <ian@hixie.ch>

1112	   References.
1113	      This document.

1115	8.2.  Registration of wss: scheme

1117	   A |wss:| URL identifies a Web Socket server and resource name, and
1118	   indicates that traffic over that connection is to be encrypted.

1120	   URI scheme name.
1121	      wss

1123	   Status.
1124	      Permanent.

1126	   URI scheme syntax.
1127	      In ABNF terms using the terminals from the URI specifications:
1128	      [RFC5234] [RFC3986]

1130	           "wss" ":" hier-part [ "?" query ]

1132	      The path and query components form the resource name sent to the
1133	      server to identify the kind of service desired.  Other components
1134	      have the meanings described in RFC3986.

1136	   URI scheme semantics.
1137	      The only operation for this scheme is to open a connection using
1138	      the Web Socket protocol, encrypted using TLS.

1140	   Encoding considerations.
1141	      Characters in the host component that are excluded by the syntax
1142	      defined above must be converted from Unicode to ASCII by applying
1143	      the IDNA ToASCII algorithm to the Unicode host name, with both the
1144	      AllowUnassigned and UseSTD3ASCIIRules flags set, and using the
1145	      result of this algorithm as the host in the URI.  [RFC3490]

1147	      Characters in other components that are excluded by the syntax
1148	      defined above must be converted from Unicode to ASCII by first
1149	      encoding the characters as UTF-8 and then replacing the
1150	      corresponding bytes using their percent-encoded form as defined in
1151	      the URI and IRI specification.  [RFC3986] [RFC3987]

1153	   Applications/protocols that use this URI scheme name.
1154	      Web Socket protocol over TLS.

1156	   Interoperability considerations.
1157	      None.

1159	   Security considerations.
1160	      See "Security considerations" section above.

1162	   Contact.
1163	      Ian Hickson <ian@hixie.ch>

1165	   Author/Change controller.
1166	      Ian Hickson <ian@hixie.ch>

1168	   References.
1169	      This document.

1171	8.3.  Registration of the "WebSocket" HTTP Upgrade keyword

1173	   Name of token.
1174	      WebSocket

1176	   Author/Change controller.
1177	      Ian Hickson <ian@hixie.ch>

1179	   Contact.
1180	      Ian Hickson <ian@hixie.ch>

1182	   References.
1183	      This document.

1185	8.4.  WebSocket-Origin

1187	   This section describes a header field for registration in the
1188	   Permanent Message Header Field Registry.  [RFC3864]

1190	   Header field name
1191	      WebSocket-Origin

1193	   Applicable protocol
1194	      http

1196	   Status
1197	      reserved; do not use outside WebSocket handshake

1199	   Author/Change controller
1200	      IETF

1202	   Specification document(s)
1203	      This document is the relevant specification.

1205	   Related information
1206	      None.

1208	8.5.  WebSocket-Protocol

1210	   This section describes a header field for registration in the
1211	   Permanent Message Header Field Registry.  [RFC3864]

1213	   Header field name
1214	      WebSocket-Protocol

1216	   Applicable protocol
1217	      http

1219	   Status
1220	      reserved; do not use outside WebSocket handshake

1222	   Author/Change controller
1223	      IETF

1225	   Specification document(s)
1226	      This document is the relevant specification.

1228	   Related information
1229	      None.

1231	8.6.  WebSocket-Location

1233	   This section describes a header field for registration in the
1234	   Permanent Message Header Field Registry.  [RFC3864]

1236	   Header field name
1237	      WebSocket-Location

1239	   Applicable protocol
1240	      http

1242	   Status
1243	      reserved; do not use outside WebSocket handshake

1245	   Author/Change controller
1246	      IETF

1248	   Specification document(s)
1249	      This document is the relevant specification.

1251	   Related information
1252	      None.

1254	9.  Using the Web Socket protocol from other specifications

1256	   The Web Socket protocol is intended to be used by another
1257	   specification to provide a generic mechanism for dynamic author-
1258	   defined content, e.g. in a specification defining a scripted API.

1260	   Such a specification first needs to "establish a Web Socket
1261	   connection", providing that algorithm with:

1263	   o  The destination, consisting of a /host/ and a /port/.

1265	   o  A /resource name/, which allows for multiple services to be
1266	      identified at one host and port.

1268	   o  A /secure/ flag, which is true if the connection is to be
1269	      encrypted, and false otherwise.

1271	   o  An ASCII serialization of an origin that is being made responsible
1272	      for the connection.  [ORIGIN]

1274	   o  Optionally a string identifying a protocol that is to be layered
1275	      over the Web Socket connection.

1277	   The /host/, /port/, /resource name/, and /secure/ flag are usually
1278	   obtained from a URL using the steps to parse a Web Socket URL's
1279	   components.  These steps fail if the URL does not specify a Web
1280	   Socket.

1282	   If a connection can be established, then it is said that the "Web
1283	   Socket connection is established".

1285	   If at any time the connection is to be closed, then the specification
1286	   needs to use the "close the Web Socket connection" algorithm.

1288	   When the connection is closed, for any reason including failure to
1289	   establish the connection in the first place, it is said that the "Web
1290	   Socket connection is closed".

1292	   While a connection is open, the specification will need to handle the
1293	   cases when "a Web Socket message has been received" with text /data/.

1295	   To send some text /data/ to an open connection, the specification
1296	   needs to "send /data/ using the Web Socket".

1298	10.  Normative References

1300	   [HTML5]    Hickson, I., "HTML5", October 2009.

1302	   [ORIGIN]   Barth, A., Jackson, C., and I. Hickson, "The HTTP Origin
1303	              Header", September 2009.

1305	   [RFC2109]  Kristol, D. and L. Montulli, "HTTP State Management
1306	              Mechanism", RFC 2109, February 1997.

1308	   [RFC2119]  Bradner, S., "Key words for use in RFCs to Indicate
1309	              Requirement Levels", BCP 14, RFC 2119, March 1997.

1311	   [RFC2246]  Dierks, T. and C. Allen, "The TLS Protocol Version 1.0",
1312	              RFC 2246, January 1999.

1314	   [RFC2616]  Fielding, R., Gettys, J., Mogul, J., Frystyk, H.,
1315	              Masinter, L., Leach, P., and T. Berners-Lee, "Hypertext
1316	              Transfer Protocol -- HTTP/1.1", RFC 2616, June 1999.

1318	   [RFC2965]  Kristol, D. and L. Montulli, "HTTP State Management
1319	              Mechanism", RFC 2965, October 2000.

1321	   [RFC3490]  Faltstrom, P., Hoffman, P., and A. Costello,
1322	              "Internationalizing Domain Names in Applications (IDNA)",
1323	              RFC 3490, March 2003.

1325	   [RFC3629]  Yergeau, F., "UTF-8, a transformation format of ISO
1326	              10646", STD 63, RFC 3629, November 2003.

1328	   [RFC3864]  Klyne, G., Nottingham, M., and J. Mogul, "Registration
1329	              Procedures for Message Header Fields", BCP 90, RFC 3864,
1330	              September 2004.

1332	   [RFC3986]  Berners-Lee, T., Fielding, R., and L. Masinter, "Uniform
1333	              Resource Identifier (URI): Generic Syntax", STD 66,
1334	              RFC 3986, January 2005.

1336	   [RFC3987]  Duerst, M. and M. Suignard, "Internationalized Resource
1337	              Identifiers (IRIs)", RFC 3987, January 2005.

1339	   [RFC5234]  Crocker, D. and P. Overell, "Augmented BNF for Syntax
1340	              Specifications: ABNF", STD 68, RFC 5234, January 2008.

1342	   [WEBADDRESSES]
1343	              Connolly, D. and C. Sperberg-McQueen, "Web addresses in
1344	              HTML 5", May 2009.

1346	   [WSAPI]    Hickson, I., "The Web Sockets API", October 2009.

1348	Author's Address

1350	   Ian Hickson
1351	   Google, Inc.

1353	   Email: ian@hixie.ch
1354	   URI:   http://ln.hixie.ch/