idnits 2.17.1 draft-hixie-thewebsocketprotocol-05.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 (March 23, 2009) is 5506 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' ** 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) Summary: 6 errors (**), 0 flaws (~~), 2 warnings (==), 2 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 March 23, 2009 5 Expires: September 24, 2009 7 The Web Socket protocol 8 draft-hixie-thewebsocketprotocol-05 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 September 24, 2009. 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 This protocol enables two-way communication between a user agent 47 running untrusted code running in a controlled environment to a 48 remote host that understands the protocol. It is intended to fail to 49 communicate with servers of pre-existing protocols like SMTP or HTTP, 50 while allowing HTTP servers to opt-in to supporting this protocol if 51 desired. It is designed to be easy to implement on the server side. 53 Author's note 55 This document is automatically generated from, and is therefore a 56 subset of, the HTML5 specification produced by the WHATWG. [HTML5] 58 Table of Contents 60 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 4 61 2. Conformance requirements . . . . . . . . . . . . . . . . . . . 5 62 3. Client-side requirements . . . . . . . . . . . . . . . . . . . 6 63 3.1. Handshake . . . . . . . . . . . . . . . . . . . . . . . . 6 64 3.2. Data framing . . . . . . . . . . . . . . . . . . . . . . . 12 65 4. Server-side requirements . . . . . . . . . . . . . . . . . . . 14 66 4.1. Minimal handshake . . . . . . . . . . . . . . . . . . . . 14 67 4.2. Handshake details . . . . . . . . . . . . . . . . . . . . 14 68 4.3. Data framing . . . . . . . . . . . . . . . . . . . . . . . 15 69 5. Closing the connection . . . . . . . . . . . . . . . . . . . . 17 70 6. Security considerations . . . . . . . . . . . . . . . . . . . 18 71 7. IANA considerations . . . . . . . . . . . . . . . . . . . . . 19 72 8. Normative References . . . . . . . . . . . . . . . . . . . . . 20 73 Author's Address . . . . . . . . . . . . . . . . . . . . . . . . . 21 75 1. Introduction 77 ** ISSUE ** ... 79 2. Conformance requirements 81 All diagrams, examples, and notes in this specification are non- 82 normative, as are all sections explicitly marked non-normative. 83 Everything else in this specification is normative. 85 The key words "MUST", "MUST NOT", "REQUIRED", "SHOULD", "SHOULD NOT", 86 "RECOMMENDED", "MAY", and "OPTIONAL" in the normative parts of this 87 document are to be interpreted as described in RFC2119. For 88 readability, these words do not appear in all uppercase letters in 89 this specification. [RFC2119] 91 Requirements phrased in the imperative as part of algorithms (such as 92 "strip any leading space characters" or "return false and abort these 93 steps") are to be interpreted with the meaning of the key word 94 ("must", "should", "may", etc) used in introducing the algorithm. 96 Conformance requirements phrased as algorithms or specific steps may 97 be implemented in any manner, so long as the end result is 98 equivalent. (In particular, the algorithms defined in this 99 specification are intended to be easy to follow, and not intended to 100 be performant.) 102 Implementations may impose implementation-specific limits on 103 otherwise unconstrained inputs, e.g. to prevent denial of service 104 attacks, to guard against running out of memory, or to work around 105 platform-specific limitations. 107 The conformance classes defined by this specification are user agents 108 and servers. 110 3. Client-side requirements 112 _This section only applies to user agents, not to servers._ 114 NOTE: This specification doesn't currently define a limit to the 115 number of simultaneous connections that a client can establish to a 116 server. 118 3.1. Handshake 120 When the user agent is to *establish a Web Socket connection* to a 121 host /host/, optionally on port /port/, from an origin /origin/, with 122 a flag /secure/, and with a particular /resource name/, it must run 123 the following steps. 125 NOTE: The /host/ and /origin/ strings will be all-lowercase when this 126 algorithm is invoked. 128 1. If there is no explicit /port/, then: if /secure/ is false, let 129 /port/ be 81, otherwise let /port/ be 815. 131 2. If the user agent is configured to use a proxy to connect to 132 host /host/ and/or port /port/, then connect to that proxy and 133 ask it to open a TCP/IP connection to the host given by /host/ 134 and the port given by /port/. 136 EXAMPLE: For example, if the user agent uses an HTTP proxy 137 for all traffic, then if it was to try to connect to port 80 138 on server example.com, it might send the following lines to 139 the proxy server: 141 CONNECT example.com HTTP/1.1 143 If there was a password, the connection might look like: 145 CONNECT example.com HTTP/1.1 146 Proxy-authorization: Basic ZWRuYW1vZGU6bm9jYXBlcyE= 148 Otherwise, if the user agent is not configured to use a proxy, 149 then open a TCP/IP connection to the host given by /host/ and 150 the port given by /port/. 152 3. If the connection could not be opened, then fail the Web Socket 153 connection and abort these steps. 155 4. If /secure/ is true, perform a TLS handshake over the 156 connection. If this fails (e.g. the server's certificate could 157 not be verified), then fail the Web Socket connection and abort 158 these steps. Otherwise, all further communication on this 159 channel must run through the encrypted tunnel. [RFC2246] 161 5. Send the following bytes to the remote side (the server): 163 47 45 54 20 165 Send the /resource name/ value, encoded as US-ASCII. 167 Send the following bytes: 169 20 48 54 54 50 2f 31 2e 31 0d 0a 55 70 67 72 61 170 64 65 3a 20 57 65 62 53 6f 63 6b 65 74 0d 0a 43 171 6f 6e 6e 65 63 74 69 6f 6e 3a 20 55 70 67 72 61 172 64 65 0d 0a 174 NOTE: The string "GET ", the path, " HTTP/1.1", CRLF, the string 175 "Upgrade: WebSocket", CRLF, and the string "Connection: 176 Upgrade", CRLF. 178 6. Send the following bytes: 180 48 6f 73 74 3a 20 182 Send the /host/ value, encoded as US-ASCII. 184 Send the following bytes: 186 0d 0a 188 NOTE: The string "Host: ", the host, and CRLF. 190 7. Send the following bytes: 192 4f 72 69 67 69 6e 3a 20 194 Send the /origin/ value, encoded as US-ASCII. 196 NOTE: The /origin/ value is a string that was passed to this 197 algorithm. 199 Send the following bytes: 201 0d 0a 203 NOTE: The string "Origin: ", the origin, and CRLF. 205 8. If the client has any authentication information or cookies that 206 would be relevant to a resource accessed over HTTP, if /secure/ 207 is false, or HTTPS, if it is true, on host /host/, port /port/, 208 with /resource name/ as the path (and possibly query 209 parameters), then HTTP headers that would be appropriate for 210 that information should be sent at this point. [RFC2616] 211 [RFC2109] [RFC2965] 213 Each header must be on a line of its own (each ending with a CR 214 LF sequence). For the purposes of this step, each header must 215 not be split into multiple lines (despite HTTP otherwise 216 allowing this with continuation lines). 218 EXAMPLE: For example, if the server had a username and 219 password that applied to |http://example.com/socket|, and the 220 Web Socket was being opened to |ws://example.com:80/socket|, 221 it could send them: 223 Authorization: Basic d2FsbGU6ZXZl 225 However, it would not send them if the Web Socket was being 226 opened to |ws://example.com/socket|, as that uses a different 227 port (81, not 80). 229 9. Send the following bytes: 231 0d 0a 233 NOTE: Just a CRLF (a blank line). 235 10. Read the first 85 bytes from the server. If the connection 236 closes before 85 bytes are received, or if the first 85 bytes 237 aren't exactly equal to the following bytes, then fail the Web 238 Socket connection and abort these steps. 240 48 54 54 50 2f 31 2e 31 20 31 30 31 20 57 65 62 241 20 53 6f 63 6b 65 74 20 50 72 6f 74 6f 63 6f 6c 242 20 48 61 6e 64 73 68 61 6b 65 0d 0a 55 70 67 72 243 61 64 65 3a 20 57 65 62 53 6f 63 6b 65 74 0d 0a 244 43 6f 6e 6e 65 63 74 69 6f 6e 3a 20 55 70 67 72 245 61 64 65 0d 0a 247 NOTE: The string "HTTP/1.1 101 Web Socket Protocol Handshake", 248 CRLF, the string "Upgrade: WebSocket", CRLF, the string 249 "Connection: Upgrade", CRLF. 251 11. Let /headers/ be a list of name-value pairs, initially empty. 253 12. _Header_: Let /name/ and /value/ be empty byte arrays. 255 13. Read a byte from the server. 257 If the connection closes before this byte is received, then fail 258 the Web Socket connection and abort these steps. 260 Otherwise, handle the byte as described in the appropriate entry 261 below: 263 -> If the byte is 0x0d (ASCII CR) 264 If the /name/ byte array is empty, then jump to the headers 265 processing step. Otherwise, fail the Web Socket connection 266 and abort these steps. 268 -> If the byte is 0x0a (ASCII LF) 269 Fail the Web Socket connection and abort these steps. 271 -> If the byte is 0x3a (ASCII ":") 272 Move on to the next step. 274 -> If the byte is in the range 0x41 .. 0x5a (ASCII "A" .. "Z") 275 Append a byte whose value is the byte's value plus 0x20 to 276 the /name/ byte array and redo this step for the next byte. 278 -> Otherwise 279 Append the byte to the /name/ byte array and redo this step 280 for the next byte. 282 NOTE: This reads a header name, terminated by a colon, 283 converting upper-case ASCII letters to lowercase, and aborting 284 if a stray CR or LF is found. 286 14. Read a byte from the server. 288 If the connection closes before this byte is received, then fail 289 the Web Socket connection and abort these steps. 291 Otherwise, handle the byte as described in the appropriate entry 292 below: 294 -> If the byte is 0x20 (ASCII space) 295 Ignore the byte and move on to the next step. 297 -> Otherwise 298 Treat the byte as described by the list in the next step, 299 then move on to that next step for real. 301 NOTE: This skips past a space character after the colon, if 302 necessary. 304 15. Read a byte from the server. 306 If the connection closes before this byte is received, then fail 307 the Web Socket connection and abort these steps. 309 Otherwise, handle the byte as described in the appropriate entry 310 below: 312 -> If the byte is 0x0d (ASCII CR) 313 Move on to the next step. 315 -> If the byte is 0x0a (ASCII LF) 316 Fail the Web Socket connection and abort these steps. 318 -> Otherwise 319 Append the byte to the /name/ byte array and redo this step 320 for the next byte. 322 NOTE: This reads a header value, terminated by a CRLF. 324 16. Read a byte from the server. 326 If the connection closes before this byte is received, or if the 327 byte is not a 0x0a byte (ASCII LF), then fail the Web Socket 328 connection and abort these steps. 330 NOTE: This skips past the LF byte of the CRLF after the header. 332 17. Append an entry to the /headers/ list that has the name given by 333 the string obtained by interpreting the /name/ byte array as a 334 UTF-8 byte stream and the value given by the string obtained by 335 interpreting the /value/ byte array as a UTF-8 byte stream. 337 18. Return to the "Header" step above. 339 19. _Headers processing_: If there is not exactly one entry in the 340 /headers/ list whose name is "websocket-origin", or if there is 341 not exactly one entry in the /headers/ list whose name is 342 "websocket-location", or if there are any entries in the 343 /headers/ list whose names are the empty string, then fail the 344 Web Socket connection and abort these steps. 346 20. Handle each entry in the /headers/ list as follows: 348 -> If the entry's name is "websocket-origin|" 349 If the value is not exactly equal to /origin/, converted to 350 lowercase, then fail the Web Socket connection and abort 351 these steps. 353 -> If the entry's name is "websocket-location|" 354 If the value is not exactly equal to a string consisting of 355 the following components in the same order, then fail the Web 356 Socket connection and abort these steps: 358 1. The string "ws" if /secure/ is false and "wss" if 359 /secure/ is true 361 2. The three characters "://". 363 3. The value of /host/. 365 4. If /secure/ is false and /port/ is not 81, or if /secure/ 366 is true and /port/ is not 815: a ":" character followed 367 by the value of /port/. 369 5. The value of /resource name/. 371 -> If the entry's name is "set-cookie|" or "set-cookie2|" or 372 another cookie-related header name 373 Handle the cookie as defined by the appropriate spec, with 374 the resource being the one with the host /host/, the port 375 /port/, the path (and possibly query parameters) /resource 376 name/, and the scheme |http| if /secure/ is false and |https| 377 if /secure/ is true. [RFC2109] [RFC2965] 379 -> Any other name 380 Ignore it. 382 21. The *Web Socket connection is established*. Now the user agent 383 must send and receive to and from the connection as described in 384 the next section. 386 To *fail the Web Socket connection*, the user agent must close the 387 Web Socket connection, and may report the problem to the user (which 388 would be especially useful for developers). However, user agents 389 must not convey the failure information to the script that attempted 390 the connection in a way distinguishable from the Web Socket being 391 closed normally. 393 3.2. Data framing 395 Once a Web Socket connection is established, the user agent must run 396 through the following state machine for the bytes sent by the server. 398 1. Try to read a byte from the server. Let /frame type/ be that 399 byte. 401 If no byte could be read because the Web Socket connection is 402 closed, then abort. 404 2. Handle the /frame type/ byte as follows: 406 If the high-order bit of the /frame type/ byte is set (i.e. if 407 /frame type/ _and_ed with 0x80 returns 0x80) 408 Run these steps. If at any point during these steps a read is 409 attempted but fails because the Web Socket connection is 410 closed, then abort. 412 1. Let /length/ be zero. 414 2. _Length_: Read a byte, let /b/ be that byte. 416 3. Let /b_v/ be integer corresponding to the low 7 bits of 417 /b/ (the value you would get by _and_ing /b/ with 0x7f). 419 4. Multiply /length/ by 128, add /b_v/ to that result, and 420 store the final result in /length/. 422 5. If the high-order bit of /b/ is set (i.e. if /b/ _and_ed 423 with 0x80 returns 0x80), then return to the step above 424 labeled _length_. 426 6. Read /length/ bytes. 428 7. Discard the read bytes. 430 If the high-order bit of the /frame type/ byte is _not_ set (i.e. 431 if /frame type/ _and_ed with 0x80 returns 0x00) 432 Run these steps. If at any point during these steps a read is 433 attempted but fails because the Web Socket connection is 434 closed, then abort. 436 1. Let /raw data/ be an empty byte array. 438 2. _Data_: Read a byte, let /b/ be that byte. 440 3. If /b/ is not 0xff, then append /b/ to /raw data/ and 441 return to the previous step (labeled _data_). 443 4. Interpret /raw data/ as a UTF-8 string, and store that 444 string in /data/. 446 5. If /frame type/ is 0x00, then *a message has been 447 received* with text /data/. Otherwise, discard the data. 449 3. Return to the first step to read the next byte. 451 If the user agent is faced with content that is too large to be 452 handled appropriately, then it must fail the Web Socket connection. 454 Once a Web Socket connection is established, the user agent must use 455 the following steps to *send /data/ using the Web Socket*: 457 1. Send a 0x00 byte to the server. 459 2. Encode /data/ using UTF-8 and send the resulting byte stream to 460 the server. 462 3. Send a 0xff byte to the server. 464 4. Server-side requirements 466 _This section only applies to servers._ 468 4.1. Minimal handshake 470 NOTE: This section describes the minimal requirements for a server- 471 side implementation of Web Sockets. 473 Listen on a port for TCP/IP. Upon receiving a connection request, 474 open a connection and send the following bytes back to the client: 476 48 54 54 50 2f 31 2e 31 20 31 30 31 20 57 65 62 477 20 53 6f 63 6b 65 74 20 50 72 6f 74 6f 63 6f 6c 478 20 48 61 6e 64 73 68 61 6b 65 0d 0a 55 70 67 72 479 61 64 65 3a 20 57 65 62 53 6f 63 6b 65 74 0d 0a 480 43 6f 6e 6e 65 63 74 69 6f 6e 3a 20 55 70 67 72 481 61 64 65 0d 0a 483 Send the string "WebSocket-Origin" followed by a U+003A COLON (":") 484 followed by the ASCII serialization of the origin from which the 485 server is willing to accept connections, followed by a CRLF pair 486 (0x0d 0x0a). 488 For instance: 490 WebSocket-Origin: http://example.com 492 Send the string "WebSocket-Location" followed by a U+003A COLON (":") 493 followed by the URL of the Web Socket script, followed by a CRLF pair 494 (0x0d 0x0a). 496 For instance: 498 WebSocket-Location: ws://example.com:80/demo 500 Send another CRLF pair (0x0d 0x0a). 502 Read (and discard) data from the client until four bytes 0x0d 0x0a 503 0x0d 0x0a are read. 505 If the connection isn't dropped at this point, go to the data framing 506 section. 508 4.2. Handshake details 510 The previous section ignores the data that is transmitted by the 511 client during the handshake. 513 The data sent by the client consists of a number of fields separated 514 by CR LF pairs (bytes 0x0d 0x0a). 516 The first field consists of three tokens separated by space 517 characters (byte 0x20). The middle token is the path being opened. 518 If the server supports multiple paths, then the server should echo 519 the value of this field in the initial handshake, as part of the URL 520 given on the |WebSocket-Location| line (after the appropriate scheme 521 and host). 523 The remaining fields consist of name-value pairs, with the name part 524 separated from the value part by a colon and a space (bytes 0x3a 525 0x20). Of these, several are interesting: 527 Host (bytes 48 6f 73 74) 528 The value gives the hostname that the client intended to use when 529 opening the Web Socket. It would be of interest in particular to 530 virtual hosting environments, where one server might serve 531 multiple hosts, and might therefore want to return different data. 533 The right host has to be output as part of the URL given on the 534 |WebSocket-Location| line of the handshake described above, to 535 verify that the server knows that it is really representing that 536 host. 538 Origin (bytes 4f 72 69 67 69 6e) 539 The value gives the scheme, hostname, and port (if it's not the 540 default port for the given scheme) of the page that asked the 541 client to open the Web Socket. It would be interesting if the 542 server's operator had deals with operators of other sites, since 543 the server could then decide how to respond (or indeed, _whether_ 544 to respond) based on which site was requesting a connection. 546 If the server supports connections from more than one origin, then 547 the server should echo the value of this field in the initial 548 handshake, on the |WebSocket-Origin| line. 550 Other fields 551 Other fields can be used, such as "Cookie" or "Authorization", for 552 authentication purposes. 554 4.3. Data framing 556 NOTE: This section only describes how to handle content that this 557 specification allows user agents to send (text). It doesn't handle 558 any arbitrary content in the same way that the requirements on user 559 agents defined earlier handle any content including possible future 560 extensions to the protocols. 562 The server should run through the following steps to process the 563 bytes sent by the client: 565 1. Read a byte from the client. Assuming everything is going 566 according to plan, it will be a 0x00 byte. Behaviour for the 567 server is undefined if the byte is not 0x00. 569 2. Let /raw data/ be an empty byte array. 571 3. _Data_: Read a byte, let /b/ be that byte. 573 4. If /b/ is not 0xff, then append /b/ to /raw data/ and return to 574 the previous step (labeled _data_). 576 5. Interpret /raw data/ as a UTF-8 string, and apply whatever 577 server-specific processing should occur for the resulting string. 579 6. Return to the first step to read the next byte. 581 The server should run through the following steps to send strings to 582 the client: 584 1. Send a 0x00 byte to the client to indicate the start of a string. 586 2. Encode /data/ using UTF-8 and send the resulting byte stream to 587 the client. 589 3. Send a 0xff byte to the client to indicate the end of the 590 message. 592 5. Closing the connection 594 To *close the Web Socket connection*, either the user agent or the 595 server closes the TCP/IP connection. There is no closing handshake. 596 Whether the user agent or the server closes the connection, it is 597 said that the *Web Socket connection is closed*. 599 Servers may close the Web Socket connection whenever desired. 601 User agents should not close the Web Socket connection arbitrarily. 603 6. Security considerations 605 ** ISSUE ** ... 607 7. IANA considerations 609 ** ISSUE ** ...(two URI schemes, two ports, HTTP Upgrade keyword) 611 8. Normative References 613 [HTML5] Hickson, I., "HTML5", March 2009. 615 [RFC2109] Kristol, D. and L. Montulli, "HTTP State Management 616 Mechanism", RFC 2109, February 1997. 618 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 619 Requirement Levels", BCP 14, RFC 2119, March 1997. 621 [RFC2246] Dierks, T. and C. Allen, "The TLS Protocol Version 1.0", 622 RFC 2246, January 1999. 624 [RFC2616] Fielding, R., Gettys, J., Mogul, J., Frystyk, H., 625 Masinter, L., Leach, P., and T. Berners-Lee, "Hypertext 626 Transfer Protocol -- HTTP/1.1", RFC 2616, June 1999. 628 [RFC2965] Kristol, D. and L. Montulli, "HTTP State Management 629 Mechanism", RFC 2965, October 2000. 631 Author's Address 633 Ian Hickson 634 Google, Inc. 636 Email: ian@hixie.ch 637 URI: http://ln.hixie.ch/